Abstract
Lung oncogenesis relies on intracellular cysteine to overcome oxidative stress. Several tumor types, including non-small cell lung cancer (NSCLC), upregulate the system xc- cystine/glutamate antiporter (xCT) through overexpression of the cystine transporter SLC7A11, thus sustaining intracellular cysteine levels to support glutathione synthesis.
Nuclear factor erythroid 2-related factor 2 (NRF2) serves as a master regulator of oxidative stress resistance by regulating SLC7A11, whereas Kelch-like ECH-associated protein (KEAP1) acts as a cytoplasmic repressor of the oxidative responsive transcription factor NRF2. Mutations in KEAP1/NRF2 and p53 induce SLC7A11 activation in NSCLC. Extracellular cystine is crucial in supplying the intracellular cysteine levels necessary to combat oxidative stress. Disruptions in cystine availability lead to iron-dependent lipid peroxidation, thus resulting in a type of cell death called ferroptosis. Pharmacologic inhibitors of xCT (either SLC7A11 or GPX4) induce ferroptosis of NSCLC cells and other tumor types. When cystine uptake is impaired, the intracellular cysteine pool can be sustained by the transsulfuration pathway, which is catalyzed by cystathionine-B-synthase (CBS) and cystathionine g-lyase (CSE). The involvement of exogenous cysteine/cystine and the transsulfuration pathway in the cysteine pool and downstream metabolites results in compromised CD8+ T cell function and evasion of immunotherapy, diminishing immune response and potentially reducing the effectiveness of immunotherapeutic interventions. Pyroptosis is a previously unrecognized form of regulated cell death. In NSCLCs driven by EGFR, ALK, or KRAS, selective inhibitors induce pyroptotic cell death as well as apoptosis. After targeted therapy, the mitochondrial intrinsic apoptotic pathway is activated, thus leading to the cleavage and activation of caspase-3. Consequently, gasdermin E is activated, thus leading to permeabilization of the cytoplasmic membrane and cell-lytic pyroptosis (indicated by characteristic cell membrane ballooning). Breakthroughs in KRAS G12C allele-specific inhibitors and potential mechanisms of resistance are also discussed herein.
keywords
- Solute carrier family 7 member 11 (SLC7A11)
- nuclear factor erythroid 2-related factor 2 (NRF2)
- ferroptosis
- pyroptosis
- KRAS G12C allele-specific inhibitors
- non-small cell lung cancer (NSCLC)
Introduction
The ASCO living clinical practice guidelines are aimed at providing up-to-date evidence based on clinical trials to guide the management of patients with stage IV non-small cell lung cancer (NSCLC). The therapeutic strategies for patients with stage IV NSCLC with versus without driver alterations have critically distinctions. Unfortunately, no cure is currently available for patients with stage IV NSCLC. For patients who cannot be treated with targeted therapies, who have a performance status (PS) of 0 or 1, platinum-based chemotherapy is recommended. Within this group, patients with high PD-L1 expression (TPS > 50%) are recommended to receive pembrolizumab or pembrolizumab in combination with chemotherapy if they have non-squamous histology. Although other treatment options are possible, the ASCO guidelines emphasize that cemiplimab as a single agent is supported by good evidence in patients with high PD-L1 expression, non-squamous NSCLC, and PS of 0–11. For patients with PD-L1 expression that is negative or between 1% and 49% (determined in tumor cells), non-squamous cell carcinoma, and PS of 0–1, the recommended treatment is pembrolizumab in combination with carboplatin and pemetrexed.
In contrast, for patients with stage IV squamous NSCLC with high PD-L1 and PS of 0–1, single-agent pembrolizumab, atezolizumab, or cemiplimab is recommended, whereas other options have not received strong recommendations. The ASCO Living Guidelines endorse pembrolizumab with carboplatin and paclitaxel or nab-paclitaxel for patients with stage IV squamous NSCLC with negative or low positive PD-L1 expression (TPS, 1%–49%)1.
In the KEYNOTE-189 study in patients with metastatic non-squamous NSCLC2, the median progression-free survival (PFS) was 9.0 months in the pembrolizumab plus pemetrexed-platinum group and 4.9 months in the placebo plus pemetrexed-platinum group (hazard ratio 0.49). The median overall survival was 22.0 months in the pembrolizumab plus pemetrexed-platinum group and 10.6 months in the placebo plus pemetrexed-platinum group (hazard ratio 0.56). The median duration of treatment was 7.2 months in the pembrolizumab plus pemetrexed-platinum group and 4.2 months in the placebo plus pemetrexed-platinum group. Fifty-six patients (13.7%) allocated to receive pembrolizumab plus pemetrexed-platinum completed 35 cycles of pembrolizumab therapy2. In the GEMSTONE-302 trial, patients with squamous or non-squamous NSCLC were randomized to receive either sugemalimab (a PD-L1 inhibitor) plus carboplatin-paclitaxel or carboplatin-pemetrexed, respectively, or placebo plus the same carboplatin-based regimens3. The median PFS was 7.8 months in the sugemalimab group vs. 4.9 months in the placebo group (hazard ratio 0.50). The median duration of treatment was 7.2 months with sugemalimab and 4.6 months with placebo.
For patients with stage IV NSCLC and driver alterations, the ASCO Living Guidelines (version 2022.3) recommend standalone osimertinib for patients with EGFR mutation (L858R/exon 19 deletions, with or without concomitant T790M) with PS of 0–2. For other EGFR-targeted treatments, such as a combination of gefitinib plus chemotherapy, dacomitinib, afatinib, erlotinib plus bevacizumab, or erlotinib plus ramucirumab, the level of recommendation is moderate, and no evidence supports the use of single-agent immunotherapy. The guidelines also provide recommendations for ALK, ROS1, BRAF, MET exon 14 skipping mutations, RET rearrangements, NTRK fusions, and HER2 alterations4. For patients with advanced NSCLC and KRAS G12C mutations who have received prior therapy, sotorasib may be considered, although this recommendation is weak4. Sotorasib is a selective inhibitor of KRAS G12C. Recent results from the phase III CodeBreak 200 trial have indicated the efficacy and safety of oral sotorasib (960 mg once daily) in comparison to intravenous docetaxel (75 mg/m2 once every 3 weeks)5. The trial was conducted in patients with advanced NSCLC with KRAS G12C mutations, who had progressed after previous platinum-based chemotherapy and PD-1 or PD-L1 inhibitor treatment. The median PFS was 5.6 months for sotorasib vs. 4.5 months for docetaxel (hazard ratio 0.66, P = 0.0017). The overall response rate was 28.1% for sotorasib and 13.2% for docetaxel. The median overall survival was 10.6 months for sotorasib and 11.3 months for docetaxel5. Treatment-associated adverse events of grade 3 or higher included diarrhea (12%), alanine aminotransferase increase (8%), and increased aspartate aminotransferase (5%) with sotorasib, and neutropenia (9%), fatigue (6%), and febrile neutropenia (5%) with docetaxel5. In other studies, such as the LC-SCRUM-Asia database, the median survival among patients with KRAS G12C was 24.6 months and did not significantly differ from that of patients with other KRAS G12C mutations6. The incidence of KRAS G12C is lower in Asian and Hispanic patients than white patients with NSCLC6,7. A recent study conducted in the Netherlands, using data from a nationwide registry, has demonstrated that Dutch patients with stage IV EGFR-mutant NSCLC bearing an exon 19 deletion have significantly prolonged median overall survival compared to those with the L858R mutation8. The subgroup of patients with exon 19 deletion and brain metastases showed a survival benefit from first-line therapy with osimertinib compared with other TKIs, whereas other subgroups did not experience this benefit. The median survival was 22.8 months, and 31% and 12% of patients survived at 3 and 5 years, respectively8.
Solute carrier family 7 member 11 (SLC7A11)-mediated cystine uptake
Transporters are membrane proteins regulating the shuttling of solutes across cell membranes. They are classified into 2 major superfamilies: the ATP-binding cassette (ABC) and solute carrier transporters. SLC7A11 is overexpressed in lung cancer as well as in other tumors. SLC7A11 encodes the catalytic subunit, xc- cystine/glutamate antiporter (xCT), of the anionic cystine/glutamate exchanger. The protein product of SLC7A11 (light chain) together with SLC3A2 (heavy chain), composes the system xc- transporter. Increased lipid metabolism increases membrane phospholipids, particularly polyunsaturated fatty acids, which are sensitive to lipid peroxidation under oxidative stress. Resistant mesenchymal cells may be vulnerable to ferroptosis induction. Ferroptosis is a non-apoptotic form of cell death resulting from lipid peroxidation when levels of glutathione, in both reduced (GSH) and oxidized (GSSG) forms, are low. SLC7A11 and glutathione peroxidase-4 (GPX4) are central regulators of ferroptosis. Cysteine is the rate-limiting substrate for the synthesis of GSH. Cystine uptake is mediated by the cystine/glutamate antiporter xCT (SLC7A11). Imported cystine supports GSH biosynthesis, ROS removal, and ferroptosis suppression. The SLC7A11 inhibitors sulfasalazine and erastin, and the GPX4 inhibitor RSL3, suppress tumor growth in vitro and in vivo9–11 (Figure 1). The accumulation of lipid peroxidation is also eliminated through ubiquinol or ferroptosis suppressor protein (FSP1)12 or dihydroorotate dehydrogenase (DHODH)13. Tetrahydrobiopterin (BH4) biosynthesis is upregulated under GPX4 inhibition. Dihydrofolate reductase catalyzes the synthesis of BH4, and enzymatic inhibition by methotrexate acts synergistically with GPX4 inhibition14.
Solute Carrier family 7 member 11 (SLC7A11) mediates cystine uptake and is a central regulator of ferroptosis, together with downstream glutathione peroxidase-4 (GPX4). Membrane phospholipids (PL) comprise polyunsaturated fatty acids (PUFAs) that are sensitive to lipid peroxidation and ferroptosis. Sulfasalazine and erastin are SLC7A11 and GPX4 inhibitors, respectively. Alternative ferroptosis scavenger pathways are compromised by: ferroptosis suppressor protein 1 (FSP1), dihydroorotate dehydrogenase (DHODH), and the tetrahydrobiopterin (BH4) biosynthesis pathway. TP53 mediates ferroptosis and irradiation sensitivity by decreasing SLC7A11. Nuclear factor E2-related factor 2 (NRF2), interferon gamma (IFN-γ), and OUT domain-containing ubiquitin aldehyde-binding protein 1 (OTUB1) repress SLC7A11 expression. Yes-associated protein1 (YAP1) upregulates several ferroptosis modulators, such as transferrin receptor (TFRC) and acyl-CoA synthetase long chain 4 (ASCL4), thus favoring ferroptosis. TFRC permits access to ferritin-bound iron for cellular use (ferritinophagy), in a process that allows iron release from ferritin and conversion of ferric iron (Fe3+) to bioactive ferrous iron (Fe2+). Zinc finger E-box binding homeobox 1 (ZEB1) is a mesenchymal marker in drug-tolerant persister cells, which may indicate inactivation of Hippo signaling and subsequent derepression of YAP1. YAP1 subsequently translocates to the nucleus and initiates its transcriptional program. Both YAP1 and ZEB1 are potential predictors of sensitivity to ferroptosis inducers GTP cyclohydrolase (GCH1).
TP53 mediates ferroptosis and radiosensitivity by blocking SLC7A11 expression. In addition, TP53 deficiency produces radio resistance in cancer cells and tumors through SLC7A11-mediated ferroptosis inhibition, thus suggesting that sulfasalazine in combination with irradiation may be beneficial in TP53-mutant NSCLC15. Radiotherapy induces the expression of both SLC7A11 and GPX4 in A549 cells with TP53 knockdown. Treatment with diverse types of ferroptosis inducers, such as erastin or sulfasalazine (blocking SLC7A11-mediated cystine uptake), has been found to restore irradiation-induced lipid peroxidation in TP53 knockdown cells. Immunohistochemical analysis of 4-hydroxy-2-noneal (4-HNE, a lipid peroxidation marker) has indicated that TP53 deletion decreases radiation-induced 4-HNE levels, whereas SLC7A11 deletion in TP53-knockdown tumors restores 4-HNE levels. Furthermore, ferroptosis induction correlates with TP53 activation and favorable responses to radiation in esophageal cancer15. Immunohistochemical analysis of 4-HNE and TP53 in 30 post-radiotherapy tumor samples has revealed a strong correlation between TP53 and 4-HNE levels. Patients with strong TP53 and 4-HNE staining have significantly longer overall survival than those with mild/moderate TP53 and 4-HNE staining15. TP53 is a central negative regulator of SLC7A11-mediated cystine uptake, together with nuclear factor E2-associated factor 2 (NRF2), interferon gamma (IFN-γ), and OUT domain-containing ubiquitin aldehyde-binding protein 1 (OTUB1)16. Activation of the yes-associated protein (YAP) and forkhead box protein M1 (FOXM1) axis has been identified to drive epithelial-to-mesenchymal transition (EMT)-associated EGFR TKI resistance in EGFR mutant NSCLC17. In pancreatic cancer models in which TP53 is also frequently mutated, mutant p53 binds the promoter of the long noncoding RNA LINC00857, thereby inducing its expression. LINC00857 upregulates FOXM1 expression and serves as a scaffold that enhances the interaction between FOXM1 and OTUB1, thus preventing deubiquitination of FOXM1, favoring FOXM1 accumulation, and leading to EMT and metastasis in pancreatic cancer. Atorvastatin (a lipid-lowering reagent) degrades mutant TP53. Atorvastatin inhibits the growth of pancreatic cancer cells in a dose-dependent manner, with half-maximal inhibitory concentrations of 43.04 μM and 27.63 μM in Panc-1 and MIA PaCa-2 cells, respectively. Pancreatic cells incubated with atorvastatin show markedly decreased protein expression of mutant TP53 and mRNA expression of LINC0085718. LINC00857 is upregulated in NSCLC and correlates with poor survival. In this regard, LINC00857 silencing impairs tumor growth in lung cancer cell lines bearing EGFR, KRAS, MET, and LKB1 mutations, in addition to TP53 mutations19. Moreover, LINC00857 regulates MET expression via Y-box binding protein 1 (YBX1) at the transcriptional level19. Previous work has indicated that YBX1 (or YB-1) correlates with poor survival in lung adenocarcinoma. Intriguingly, depletion of YBX1 has been found to decrease metastasis associated in colon cancer-1 (MACC1) promoter activity, thus abolishing MACC1/c-Met signaling in lung adenocarcinoma cells. Western blotting has indicated that YBX1 and MACC1 are overexpressed in several lung cancer cell lines (H358, H460, A549, and H1299)20. Early studies identified MACC1 as a key regulator of HGF-MET signaling, and MACC1 mRNA levels have been found to predict colon cancer metastasis21. In another study, the expression of the DNA/RNA YBX1 protein has been found to predict resistance to endocrine treatments and chemotherapy in patients with breast cancer22. Of interest, FOXM1 is also a direct transcriptional target induced by YAP23.
A mechanism of resistance to osimertinib in EGFR-mutant NSCLC cell lines and patient-derived xenograft mice has been associated with SLC7A11 upregulation. The mechanism of resistance involves aldo-keto reductase family 1 member B1 (AKR1B1), which interacts with and activates signal transducer and activation of transcription 3 (STAT3), thereby increasing SLC7A11 expression. Opportunistically suppressing AKR1B1 with the selective inhibitor epalrestat (an antidiabetic drug) recovers the sensitivity of resistant cell lines to EGFR TKIs and delays resistance in lung cancer patient-derived xenograft mice. AKR1B1 is upregulated in all resistant cell models. Interestingly, EGFR TKI therapy-relapsed patients have elevated GSH and GSSG in the blood, and they remain sensitive to TKI. Ectopic expression of AKR1B1 in parental cells induces STAT3 translocation into the nucleus. In addition, AKR1B1 inhibition decreases SLC7A11 transcription, as demonstrated by dual-luciferase reporter analysis. Experimental models have also indicated that gefitinib (EGFR TKI) treatment increases AKR1B1, p-STAT3, and SLC7A11 proteins, whereas this phenomenon is abolished by epalrestat treatment. AKR1B1 enhances glutathione de novo synthesis by upregulating SLC7A11, thus providing resistance against TKIs. Inhibition of AKR1B1 restores sensitivity in EGFR TKI-resistant NSCLC cell lines with recognized resistance traits, including upregulation of AXL, activation of AKT and nuclear factor kB (NF-κB), and delayed acquired resistance in CDX and PDX mouse models24.
Yes-associated protein 1 (YAP1) as a biomarker of ferroptosis
When the Hippo pathway is active, phosphorylation of YAP1 on serine 127 causes cytoplasmic retention of YAP1, thus rendering it inactive. However, in EGFR-mutant cell lines treated with EGFR TKIs, YAP signaling is activated through interleukin-6 (IL-6)-SRC-paxillin, thereby promoting YAP1 phosphorylation at tyrosine 357 and nuclear translocation. High expression of YAP1 mRNA predicts poor PFS in patients with stage IV EGFR-mutant NSCLC25. Interestingly, mesenchymal cancer cells, which are prone to metastasis and resistant to various anticancer treatments, are highly susceptible to ferroptosis. E-cadherin, an adherens junction protein in epithelial cells, suppresses ferroptosis by activating the intracellular merlin (NF2) and Hippo signaling pathway. In contrast, YAP1 upregulates several ferroptosis modulators, such as the transferrin receptor TFRC and acyl-CoA synthetase long chain 4 (ACSL4), thereby promoting ferroptosis26. TFRC is responsible for the cellular absorption of transferrin-bound iron, which is essential for cell survival but also acts as a catalyst for ferroptosis (Figure 1). ACSL4 favors the incorporation of oxidation-sensitive long-chain polyunsaturated fatty acids into phospholipids, thereby providing substrates for lipid peroxidation and ferroptosis (Figure 1). Previous studies have suggested that the mesenchymal therapy-resistant cancer cell state is dependent on the lipid peroxidation pathway, and that ZEB1 is associated with GPX4 dependency27 (recently reviewed by Lee and Roh11). Mechanistically, a loss of E-cadherin inhibits the activity of the tumor-suppressive Hippo pathway, which plays a crucial role in contact inhibition of cell growth, thus preventing cells from growing when they encounter other cells in an E-cadherin-dependent manner, and suppressing ferroptosis. Mutations in E-cadherin render diffuse gastric cancer cells more sensitive to ferroptosis induction, in a process mediated by merlin and downstream upregulation of YAP/TAZ28. According to DepMap (https://depmap.org/portal/), the expression of CDH1, which encodes E-cadherin, is inversely correlated with cellular dependence on GPX4 among gastric cancer cell lines, thus suggesting that cells with little or no E-cadherin expression have relatively stronger reliance on GPX4. CRISPR-Cas9-mediated knockout of E-cadherin in SNU16 cells has been found to increase sensitivity to ferroptosis, even under culturing at high density28. In diffuse gastric cancer models, wild-type E-cadherin suppresses the expression of TFRC and ACSL4. The expression of canonical YAP/TAZ targets, such as CCN1 (CYR61) and CCN2 [also known as connective tissue growth factor (CTGF)], is suppressed. In vitro observations have indicated that imidazole ketone erastin, an erastin derivative, induces ferroptosis in SNU16 cells with defective CDH1 expression. In vivo, increased levels of malondialdehyde (MDA), a lipid peroxidation decomposition product and marker of ferroptosis, have been observed along with increased nuclear accumulation of YAP and elevated ACSL4 expression28.
Compelling evidence suggests that the Hippo pathway causes gemcitabine resistance by inactivating YAP. Under low-crowding conditions, when the Hippo pathway is inactive, YAP leads to down-regulation of several multidrug transporters and cytidine deaminase (CDA), a key enzyme that metabolizes gemcitabine after its uptake. The sustained diminished levels of CDA and efflux pumps support gemcitabine sensitivity, and the maintenance of intracellular drug concentrations enables gemcitabine’s cytotoxic activity. In contrast, under high-crowding conditions, when the Hippo pathway is active, high levels of CDA and efflux pumps decrease intracellular gemcitabine concentrations and cause treatment resistance29. As described above, when cells grow at low density, YAP is localized to the nucleus, and the cells are sensitive to gemcitabine. However, when cells grow at high density, YAP is located in the cytoplasm, and YAP-dependent transcription is impaired, thus resulting in gemcitabine resistance. In pancreatic cancer, high expression of YAP-dependent genes, including AMOTL2, CTGF, and AXL, had been associated with prolonged patient survival. Additionally, in patients with lung cancer with STK11 mutations, high expression of CTGF has been correlated with better survival. Patients with intrahepatic cholangiocarcinoma expressing high levels of CTGF have also been found to have better survival rates than those with tumors lacking CTGF expression29. Moreover, the Hippo-YAP pathway may modulate the efficacy of other chemotherapeutic agents, such as antimetabolites and topoisomerase inhibitors.
Transsulfuration pathway
Growing evidence suggests that transsulfuration is a crucial pathway for cysteine biosynthesis in cancer cells when extracellular sources of cysteine are limited, such as during tumor growth or pharmacological inhibition of the SLC7A11-GPX4 axis. In the transsulfuration pathway, methionine is converted into cysteine. The cystathionine-β-synthase (CBS) enzyme condenses serine with the methionine cycle intermediate homocysteine, thus forming cystathionine, which is later cleaved by cystathionine-γ-lyase (CTH), thereby releasing cysteine30 (Figure 2). Recent studies have indicated a negative correlation between the transcription levels of CBS and SLC7A11 (xCT) in cancer cell lines from the Cancer Cell Line Encyclopedia31 across various cancer types. Hydrogen sulfide (H2S), a gasotransmitter32 produced by both gut microorganisms and epithelial cells from dietary sulfur amino acids (methionine and cystine), has been implicated in ferroptosis regulation. In colon cancer cells, tumor-derived H2S, produced mainly by CBS, inhibits ferroptosis by stabilizing SLC7A11 via sulfhydration at cysteine 9133. Furthermore, interference with SLC7A11 expression has been found to increase the expression of CBS and CTH.
The transsulfuration pathway is a powerful default pathway maintaining cysteine biosynthesis when the cystine supply is depleted or SLC7A11 is pharmacologically inhibited. In the transsulfuration pathway, methionine is converted into cysteine. The cystathionine-β-synthase (CBS) enzyme condenses serine with the methionine cycle intermediate homocysteine, thus forming cystathionine, which is later cleaved by cystathionine-γ-lyase (CTH), thereby releasing cysteine. Hydrogen sulfide (H2S), a gas transmitter, is produced mainly by CBS, and persulfidation of OTUB at cysteine 91 enhances the stability of SLC7A11. Aminooxy acetic acid (AOAA) inhibits CBS and CTH. ASCT2 and LAT1 are amino acid transporters in the plasma membrane S-adenosyl-methionine (SAM) and S-adenosyl-homocysteine (SAH).
Interestingly, supplementation with H2S and GYY4137 (a slow-releasing H2S donor) restores resistance to 5-FU in cell lines treated with SLC7A11, thus indicating that H2S participates in mediating the effect of SLC7A11 in maintaining chemoresistance33. Persulfidation of specific cysteine residues is involved in the posttranslational modification of multiple proteins, as mediated by endogenous H2S. The CBS-CTH-H2S axis preserves the stability of SLC7A11 through persulfidation of OTUB at cysteine 9133. A synergistic effect has been observed between aminooxyacetic acid (AOAA), an inhibitor of CBS and CTH34, with sulfasalazine and erastin in colon cancer cells33. The combination of AOAA and erastin decreases intracellular GSH levels and increases ferroptosis, with increased levels of MDA.
These findings suggest that reciprocal regulation between SLC7A11 (xCT) and the endogenous H2S produced by CBS and CTH enzymes in the transsulfutation pathway may play a role in modulating ferroptosis sensitivity and the immune response. In addition, H2S has been shown to promote regulatory T cells, a subset of immunosuppressive CD4+ T cells, by activating forkhead box P3 (FOXP3) signaling through persulfidation of Enolase 1 (ENO1) at cysteine 119. H2S also inhibits the migration of CD8+ T cells by increasing the expression of the adaptor-associated kinase AKK1 via ETS-like transcription factor 4 (ELK4) persulfidation at cysteine 25. The decrease in H2S promotes a positive immune tumor microenvironment in colorectal cancer by lowering ENO1 persulfidation in regulatory T cells and ELK4 in CD8+ T cells, thereby enhancing the activity of anti-PD-L1 and anti-CTLA4 therapies in colon cancer35.
ENO1, an important enzyme involved in glycolysis, is also regulated by NRF236. Interest in this enzyme increased after it was found to be crucial in MET activation, given that knockdown of ENO1 prevents hepatocyte growth factor (HGF) activation of MET and downstream signaling in lung cancer cell lines37. Moreover, ENO1 has been demonstrated to trigger MET and activation of low-density lipoprotein receptor-related proteins (LRP5/6-GSK3β)-β-catenin. Inhibition of MET and knockdown of ENO1 decrease LRP5/6 phosphorylation. ENO1-triggered HGF activation has been hypothesized to induce transphosphorylation of LRP5/6 in a Wnt-independent manner. The regulation of ENO1 directly activates HGF-MET signaling and Wnt-LRP5/6, in a manner indirectly driven by an EMT phenotype in studied lung cancer models37. HGF is known to stimulate MET and GSK3-dependent phosphorylation of LRP5/6 in renal proximal tubules in mice, independently of the WNT pathway38. ENO1 triggers HGF activation and induces transphosphorylation of LRP5/6 bound to the co-receptor Frizzled, both of which are bound by the prorenin receptor (PRR), which acts as an adapter between the receptor complex and the ATP-driven pump V-ATPase. Therefore, full activation via LRP6 phosphorylation requires V-ATPase activity39. Likewise, V-ATPase inhibitors (for example, Bafilomycin A, concanamycin, and omeprazole) block activation of the Wnt ligand-independent signaling pathway (Figure 3). Further insights into KRAS-mutant lung cancers and other cancers have revealed their dependence on MET for anchorage-independent growth in 3-D conditions but not monolayer conditions. In vivo, HGF is secreted from the tumor stroma and organs such as the liver. In vitro, HGF increases tumor growth only in anchorage-independent conditions. Consequently, patients with KRAS mutation may exhibit a reliance on MET signaling, specifically downstream of HGF, which is secreted by non-tumor cells and remains undetected under 2-D monolayer culture conditions40. Moreover, EGFR-mediated signaling pathways regulate MET at various levels, including the mRNA expression and protein stability levels, in KRAS-driven cells40,41.
Working model of MET in KRAS-mutant NSCLC. Enolase (ENO1) is an important enzyme that may be necessary for MET activation, through hepatocyte growth factor (HGF) and transphosphorylation of low-density lipoprotein receptor-related proteins (LRP5/6) coupling with the coreceptor Frizzled, which are together bound by the prorenin receptor (PRR), which acts as an adapter between the receptor complex and the ATP-driven pump V-ATPase. In KRAS-mutant NSCLC models growing in anchorage-independent conditions, the translation of MET is enhanced. A combination of MET inhibitors with v-ATPase inhibitors may be a complementary strategy for KRAS G12C inhibitors and MEK inhibitors.
Consequently, in anchorage-independent conditions, the translation of MET is enhanced, thus increasing expression and cell proliferation. The downregulation of MET signaling (through MET inhibitors) suppresses tumor growth40 (Figure 3).
NRF2-hyperactivation in NSCLCs
In murine cells, the expression of oncogenes, such as KRAS G12D, BRAFV619E, and MycERT2, each increases the transcription of NRF2 and genetic targeting of NRF2 pathway limited, KRASG12D-induced tumorigenesis in vivo42. As previously described, NRF2 regulates the expression of enzymes involved in various metabolic pathways, including glycolysis (such as ENO1), glutathione synthesis (such as SLC7A11), the pentose phosphate pathway, fatty acid synthesis, glutaminolysis, one-carbon metabolism, and amino acid metabolism36. Additionally, NRF2 transcriptionally upregulates CBS, thus conferring ferroptosis resistance43. Importantly, NRF2 is a master regulator of the activity of several ferroptosis and lipid peroxidation-related proteins, which are grouped into 3 classes: iron/metal metabolism (for example, ferritin heavy chain-1), intermediate metabolism (for example, AKR1B1), and GSH synthesis/metabolism (for example, SLC7A11 and GPX4)44. The status of NRF2 may be a determinant of the ferroptosis response45, and might be used to gauge and customize therapies in different NSCLC sub-types with or without driver alterations. Furthermore, the activation of NRF2 plays a crucial role in promoting the proliferation and survival of lung tumor spheroid cells. NRF2 prevents the occurrence of ferroptotic death in the inner matrix-deprived spheroid cells, while the targeting of NRF2 and GPX4 leads to the destruction of tumor cells present within the spheroids46. Lung cancer cell lines display a unique dependence on NRF2 hyperactivation for spheroid formation, and the presence of triple mutation (KRAS, TP53, and KEAP1) leads to aggressive proliferation47. However, in mice, the same triple mutation in the pancreas does not cause cancer but instead results in fibrosis48. The PIM1 proto-oncogene is located on chromosome 6p21-p25, a recurrent amplicon in triple-negative breast cancer associated with resistance to chemotherapy49.
The PIM1 kinase regulates c-MET expression via phosphorylation of eukaryotic translation initiation factor 4B (eIF4B), thus promoting lung adenocarcinoma proliferation. Phosphorylated eIF4B triggers c-MET mRNA translation with hyperactivation of the KRAS/ERK, PI3K/AKT, and STAT3 pathways50. Furthermore, PIM1 has been demonstrated to induce resistance to MET inhibitors in cell lines with MET gene amplification, such as EBC-1 (a lung cancer cell line) and MKN45 (a gastric cancer cell line). These findings suggest that co-administration of MET and PIM kinase inhibitors may be a potential strategy to prevent or overcome acquired resistance to MET standalone inhibitors51. Intriguingly, PIM kinase inhibitors target hypoxic cancer cells by decreasing NRF2 activity. Hypoxia upregulates PIM kinases, thereby promoting the nuclear translocation of NRF2, and consequently leading to the transcriptional activation of several ferroptosis and lipid peroxidation programs. Pharmacological inhibition of PIM blocks the nuclear translocation of NRF2 and induces cancer cell destruction52. In different sub-classes of NSCLC, including a broad variety of KRAS mutants, MET and NRF2 might potentially interact at multiple levels involving crucial proteins, such as ENO1, which might also undergo modification through CBS-driven H2S. Additionally, the potential regulatory effects of PIM kinases in regulating MET and NRF2 might enable novel therapeutic strategies based on PIM1 inhibitors.
NSCLC, particularly KRAS-mutant NSCLC, bears co-mutations in kelch-like ECH-associated protein 1 (KEAP1), a negative regulator of NRF253. KEAP1 eviction and NRF2 activation provoke tumor progression and metastasis in lung adenocarcinoma through the activation of BTB and CNC homology 1 (BACH1), which has been implicated in RAS-driven tumor formation. BACH1 is a leucine zipper transcription factor that represses the expression of heme oxygenase-1 (HO-1), an enzyme believed to inhibit ferroptosis54–56.
Long noncoding RNA LINC00336 increases CBS and ELK4
Another long noncoding RNA, LINC00336, is upregulated in NSCLC. This RNA mechanistically binds the RNA-binding protein ELAV-like RNA-binding protein 1 (ELAVL1). Interestingly, lymphoid-specific helicase (LSH) induces ELAV1 expression through the inactivation of TP53. In turn, ELAVL1 enhances LINC00336 levels. LINC00336 acts as a sponge for microRNA6852, thus increasing CBS mRNA levels, and consequently stimulating tumor proliferation and inhibition of ferroptosis in NSCLC57. LSH levels are elevated in several lung cancer cell lines, and LINC00336 has been found to interact with ELAV1 in H358 and PC9 cells. Moreover, microRNA6852 directly binds LINC0033 and acts as a negative upstream regulator of CBS-mediated ferroptosis inhibition57. LINC00336 might be a clinically relevant readout of biomarkers of ferroptosis inhibition with CBS upregulation in wild-type TP53. Additionally, LINC00857 may serve as a relevant readout in TP53-mutant NSCLC (Figure 4).
Working diagram indicating that NSCLC, primarily KRAS-mutant NSCLC, has several commonalities with NSCLCs with distinct driver alterations or no drivers. The diagram shows that nuclear factor E2-related factor 2 (NRF2) hyperactivation is necessary for NSCLC cells to grow in 3-D spheroid conditions. Additionally, in 3-D conditions, SHOC2, a component in noncanonical RAS pathways, and MET are factors driving KRAS-mutant cell proliferation. Mechanisms of resistance in KRAS G12C cells (H358 parental) or AMG-510 (sotorasib) involve the downregulation of the proapoptotic BIM protein and increased Bcl-x expression. Celastrol is an inhibitor of SHOC2 in KRAS and EGFR-mutant non-small cell lung cancers (NSCLCs) (see text). Treatment with v-ATPase inhibitors and MET inhibitors warrants investigation in preclinical models. Aquaporin 5 (AQP5) is elevated in NSCLC, and may potentially influence intracellular swelling and interconnect with other signaling pathways, as shown in Figure 5. PIM1 kinase is upregulated, in a mechanism of resistance to MET inhibition in cell lines with MET amplification. A major subject of interest is that PIM1 is promoted by cancer hypoxia and is a principal regulator of the master regulator NRF2, which is involved in ferroptosis and many other pathways, including positive, direct regulation of enolase 1 or though the activation of CBS. If PIM1 regulates NRF2, and persister cancer cells are dependent on glutathione 4 peroxidase (GPX4) and methionine extra-supply (transsulfuration pathway), then PIM1 inhibition might serve as a relevant therapeutic strategy. Immunohistochemical analysis of 4-hydroxy-2-nonenal (4-HNE) is indicative of ferroptosis. Long noncoding RNAs, such as LINC00857, have important functions and have been found to be upregulated in cell lines bearing TP53 mutations. In addition, these RNAs are involved in NSCLC containing many other drivers. KRAS-resistant cells exhibit an active KRAS GTP state with downregulation of the GTPase protein regulator of G protein signaling 3 (RGS3). Persister cancer cells might possibly also display high activity of NRF2 and enolase 1. Additional descriptions of the processes depicted in the figure are provided in the main text.
In colon cancer, the persulfidation of ELK4 at cysteine 25 inhibits CD8+ T cell function. This inhibition is mediated through CBS-driven H2S production, leading to acetylation of ELK4. In gastric cancer, ELK4 plays a role in regulating the lysine-specific demethylase 5 (KDM5A)/Praja-2 (PJA2)/kinase suppressor of RAS1 (KSR1) axis. ELK4 negatively regulates PJA2 expression by transcriptionally upregulating KDM5A, which reduces the ubiquitination of KSR1. This, in turn, promotes M2 polarization of macrophages and contributes to the development of gastric cancer. In gastric cancer cells and tumor associated macrophages, ELK4 might possibly activate lysine-specific demethylase 5A (KDM5A), which inhibits the expression of Praja 2 (PJA2), thus decreasing kinase suppressor of Ras 1 (KSR1). The subsequent upregulation of KSR1 promotes M2 polarization of macrophages and gastric cancer development58 (Figure 4). Therapy with the MEK inhibitor selumetinib in KRAS-mutant A427 and A549 NSCLCs increases phosphorylation of KSR-1 (serving as a scaffold for RAF, MEK, and ERK proteins), as well as in receptor tyrosine kinase signaling59. The transsulfuration pathway through ELK4 might possibly enhance KSR-1 function and serve as a barrier to the effects of KRAS selective inhibitors in KRAS-mutant NSCLCs.
Drug-tolerant persister cancer cells evade pyroptotic osmotic lysis
In NSCLC preclinical models, EMT is a mechanism of resistance to targeted therapies with either KRAS G12C or EGFR TKI inhibitors. In several KRAS G12C NSCLC cell lines, primarily those with TP53 and/or LKB1 mutations, newly synthesized KRAS is detected by western blotting within 72 h after treatment with a G12C inhibitor (ARS1620) with downstream ERK phosphorylation60. In addition, in EGFR-mutant NSCLC cells, osimertinib inhibits EGFR and decreases extracellular receptor kinase (ERK) phosphorylation. Re-activation of ERK1/2 is detected in the surviving population of cells after 72 h. Real-time PCR assays have indicated that RNA levels of ZEB are upregulated within 72 h of osimertinib therapy61. These observations suggest that drug-tolerant persister cancer cells may appear very early after standalone targeted therapy. Persister cells, such as A549 (KRAS G12S) or PC9 (EGFR E746-A750 deletion), have been found to be resistant to pyroptosis induced by erlotinib or trametinib, respectively. During treatment, cells exhibit balloon-like bubbles, which are indicative of a pyroptotic morphology. Pyroptosis is a form of lytic cell death that involves the formation of membrane pores, ion gradient imbalance, water inflow, and membrane rupture, thus resulting in catastrophic events. Although these events are observed in most cells, time-lapse imaging has indicated that approximately 5% of viable single cells (persisters) remain after 3 days of treatment62. During the process of pyroptosis, an inflammatory type of cell death, the release of inflammatory cytokines, including high-mobility group B1 (HMGB1), is facilitated through the proteolytic fragmentation of gasdermin D63.
Interestingly, persister cancer cells show high consumption of methionine (Figure 2), thus suggesting that the transsulfuration pathway is mobilized, and a vicious cycle is generated through the CBS-H2S axis. This increased methionine flux can preserve plasma membrane integrity and avoid pyroptosis. Gasdermin E, which suppresses gasdermin D63, is elevated in many NSCLC cell lines, for example after treatment with trametinib (a MEK inhibitor) in A549 and NCI-H358 (KRAS G12C) cells, or erlotinib in PC9 cells, as well as other forms of targeted therapy in a large panel of NSCLC cell lines64. The metabolic determinants of resistance to pyroptotic cell death are complex, but recent observations62 have clearly indicated a metabolic dependency on methionine and cysteine metabolism through the transsulfuration pathway (Figure 2). Excessive methionine use has been proposed to increase the cellular capacity for methylation of proteins, DNA, and histones. Persister cancer cells have been found to be sensitive to the hypomethylating agent decitabine and 5-aza-2′-deoxicytidine62.
During pyroptosis, water enters through pores and causes cell swelling, osmotic lysis, and rupture of the plasma membrane62. Additionally, aquaporin 5 (AQP5), a water channel protein, is highly expressed in NSCLC cell lines such as A549 and H358. Aquaporins are small transmembrane water channel proteins that regulate water homeostasis. Overexpression of AQP5 in NSCLCs is associated with aggressive tumor proliferation. AQP5 silencing inhibits proliferation and induces apoptosis in A549 cells65. The mechanism underlying AQP5 overexpression, which can cause cell swelling and high intracellular pressure, warrants further investigation. AQP5 overexpression also increases activation of EGFR and ERK1/2 in NSCLC cells66. Inhibition of AQP5 with acetazolamide has been found to inhibit cell proliferation and migration in gastric cancer cells67.
Chemotherapy-induced pyroptosis
Recent studies have indicated that evasion of pyroptosis, a form of programmed cell death, is a mechanism of chemotherapy resistance. This resistance is achieved through upregulation of β5-integrin68, which has been associated with resistance to osimertinib in EGFR-mutant NSCLC cell lines69. In wild-type chemosensitive lung cancer or pancreatic cancer cells, cytotoxic drugs increase sphingomyelinase activity, thus enabling hydrolysis of sphingomyelin to ceramide, increasing ROS production, and leading to the release of thioredoxin-interacting protein (TXNIP) from thioredoxin. TXNIP then binds NLR family pyrin domain containing 3 (NLRP3), thus leading to NLRP3-apoptosis associated speck-like protein (ASC)-procaspase-1 inflammasome assembly and pro-caspase-1 autocleavage. Active caspase 1 cleaves gasdermin D within the linker between its N-terminal (gasdermin-N) and C-terminal (gasdermin-C) domains. The released gasdermin D-N then oligomerizes, forming pores in the plasma membrane and inducing pyroptosis. Chemoresistant cells upregulate β5-integrin expression and augment Src phosphorylation, and subsequently activate STAT3. Importantly, STAT3 then dimerizes and binds the promoter of the sphingolipid enzyme N-acylsphingosine amidohydrolase 2 (ASAH2) and upregulates its expression. ASH2 hydrolyses ceramide to sphingosine, thereby impairing ROS production and chemotherapy-induced pyroptosis68. In cancer cell lines, patient-derived tumor organoids, and orthotopic lung and pancreatic animal models, administration of dasatinib (a Src inhibitor) or ceramidase inhibitor has been found to reactivate pyroptosis in vitro and in vivo68. On the basis of previous studies in EGFR-mutant cell lines, treatment with osimertinib increases the expression of integrins, alpha-v (αv), β3, and β5, because continuous inhibition of EGFR signaling by osimertinib activates Src family kinases, including YES1. However, when osimertinib is combined with dasatinib, an inhibitor of SFK and focal adhesion kinase (FAK), MAPK, and AKT pathways, the effect of osimertinib is enhanced69. Similarly, a recent study has demonstrated that the combination of osimertinib with repotrectinib (a multikinase inhibitor abrogating signaling by SFK members and JAK upstream of STAT3) causes tumor growth inhibition in EGFR-mutant cells, both in vitro and in vivo70.
Interleukin 6 (IL-6) and signal transduction and activation of transcription 3 (STAT3) in NSCLC
An early study revealed that EGFR-mutant cells produce high levels of IL-6, and inhibition of the IL-6/gp-130-JAK signaling pathway decreases phosphorylation levels of STAT371. In addition, phosphorylated STAT3 and IL-6 positivity have been correlated in primary lung adenocarcinomas. Treatment with P6, a JAK inhibitor, represses tumor proliferation and pSTAT3 phosphorylation in EGFR-mutant cells (11–18, H1650, and H1975) but not in KRAS-mutant cells (H460). A study evaluating the mRNA levels of the IL-6 family of cytokines has detected only IL-6 mRNA in the EGFR-mutant cell lines 11–18, H3255, H1650, and H1975. Other cytokines, such as oncostatin, leukemia inhibitory factor, or ciliary neurotrophic factor, were not found. These seminal findings have demonstrated that IL-6 is secreted by EGFR-mutant cells lines, thus leading to activation of the gp130-JAK-STAT3 signaling pathway71. Later, IL-6 was discovered to jeopardize erlotinib activity in EGFR-mutant cells, and the EMT phenotype was found to be promoted by paracrine or autocrine-stimulation of TGF-β. In erlotinib-resistant cells, TGF-β drives the expression of IL-672. Recent studies have reaffirmed that EGFR-mutant tumors, which develop oncogene-independent acquired resistance to EGFR TKIs, exhibit a mesenchymal-like phenotype and have exacerbated secretion of IL-6, thus leading to the suppression of T cell and natural killer functions. In experimental mouse models, IL-6 blockade enhances the antitumor immunity and efficacy of anti-PD-1 antibodies73. In our view, this study has provided the first solution to the challenging problem of using immunotherapy in EGFR-mutant NSCLCs. The rationale for the proposed approach is strong, given the crucial role of IL-6 in EGFR-mutant NSCLC and its effects on mesenchymal transformation and immunosuppression. IL-6 inhibitors have been approved by the FDA. Moreover, metformin has been shown to block IL-6 expression and EMT in EGFR-TKI resistant cells, and Arrieta and colleagues have reported significantly longer median PFS in patients receiving EGFR TKI plus metformin (13.1 months) than EGFR TKI (9.9 months; P = 0.03). The median overall survival was also significantly longer for patients receiving the combination (31.7 months vs. 17.5 months; P = 0.02)74.
TGF-β is a source of IL-6 production, and, as previously described, B7-H4 might also plausibly contribute. Studies have indicated that β2-adrenergic receptors are co-expressed in EGFR-mutant NSCLC cells, thus nullifying LKB1 and enhancing CREB (cyclic adenosine 3′-5′-monophosphate response element-binding protein) and IL-6 activation, and causing resistance to EGFR TKIs. However, notably, seminal findings have indicated that IL-6 secretion stimulated by norepinephrine increases only in EGFR-mutant cell lines but not EGFR wild-type cells, as measured with enzyme linked immunosorbent assays75.
Phosphoserine aminotransferase 1 (PSAT) is a key enzyme in the serine synthesis pathway with high PSAT mRNA levels in erlotinib-resistant cell lines76. In addition, the mRNA levels of other serine enzymes, phosphoglycerate dehydrogenase (PHGDH), phosphoserine phosphatase (PSPH), and serine hydroxymethyltransferase 1 (SHMT1), which are the up- and downstream enzymes in the serine synthesis pathway, have been found to be elevated in lung adenocarcinoma, thus negatively regulating interferon regulatory factor 1 (IRF1)77. The protein levels (on the basis of western blot analysis) of PHGDH, PSAT1, and PSPH are elevated in acquired erlotinib-resistant cells and in EGFR-mutant HCC827 cells after short-term treatment with 1 μmol/L erlotinib. PSAT1 depletion eliminates erlotinib resistance, partially by reverting ferroptosis resistance via decreasing the GSH/GSSG ratio. Moreover, PSAT1 interacts with IQ motif-containing GTPase-activating protein 1 (IQGAP1) and recruits STAT376. IQGAP1’s interaction with STAT3 and cell division cycle 42 (CDC42) is required for IL-6 stimulation of pancreatic cancer progression78. In summary, PSAT-IQGAP1-STAT3 is upregulated in erlotinib-resistant cells, and PSAT1 promotes tumor metastasis and EGFR inhibitor resistance. The involvement of NRF2 in the regulation of serine and glycine biosynthesis is highly relevant. NRF2 hyperactivation or mutations in NRF2 or KEAP1 promote activating transcription factor 4 (ATF4). ATF4 is produced from the glycolytic intermediate 3-phosphoglycerate, PHGDH-PSAT, PSPH, and SHMT, thus promoting cysteine and methionine transulfuration signaling. Therefore, a substantial proportion of NSCLCs are activated by NRF2-dependent programs, which confer higher aggressiveness79.
KRAS-mutant NSCLC and allele-specific inhibitors of KRAS G12C
Regulatory proteins should be considered in KRAS-mutant NSCLC, because they might be clinically relevant to understanding the current mechanisms of resistance to KRAS G12C inhibitors and/or other forms of cancer management. Important functions of KSR, CBL and SHOC2 are particularly worthy of consideration80. SHOC2 deletion sensitizes KRAS and EGFR-mutant NSCLC cells to MEK inhibitors. Moreover, SHOC2 deletion blocks MEK inhibitor-induced RAF dimerization and consequently hyperactivates ERK signaling with inhibition of BIM-dependent apoptosis81. Treatment with the MEK inhibitors trametinib or selumetinib, in combination with osimertinib, increase cell proliferation inhibition in SHOC2-depleted PC9 (EGFR-mutant) cells. In addition, SHOC2 limits sensitivity to EGFR TKIs in NSCLC cells. Treatment with the potential SHOC2 inhibitor celastrol mimics the phenotype of SHOC2 depletion. The combination of osimertinib, selumetinib, and celastrol ablates drug-tolerant persister cells82. Importantly, KRAS-mutant cells rely on SHOC2 for ERK signaling under anchorage-independent conditions83. Therefore, the SHOC2 complex and MET may be relevant therapeutic targets in KRAS-mutant NSCLCs (Figure 4). We have previously reviewed the function of SHOC2 as a potential therapeutic target84.
After treatment with a KRAS G12C inhibitor, a mechanism of adaptive resistance has been observed, with new synthesis of KRAS mRNA and expression of KRAS protein, thus suggesting that newly synthesized KRAS G12C undergoes nucleotide exchange to an active, drug-insensitive state before binding the G12C inhibitor.
EGF stimulation preceding G12C inhibitor treatment diminishes the inhibition of KRAS G12C60. The expression of heparin-binding epidermal growth factor (HBEGF), a ligand of epidermal growth factor receptor (EGFR), is shut down after therapy with a KRAS G12C inhibitor but rebounds at 48–72 h, thus resulting in a two-fold increase in secreted HBEGF after G12C inhibitor treatment. Inhibition of EGFR signaling, either by targeting EGFR or SHP2, limits the adaptive reactivation of KRAS-GTP in KRAS G12C-mutant NSCLC cells. In addition, aurora kinase A (AURKA) is upregulated, and AURKA inhibitors suppress the reactivation of KRAS-GTP induced by a KRAS G12C inhibitor alone60. Intriguingly, Casitas B-lineage lymphoma (Cbl) C (CBLC) E ligase has an oncogenic function by positively regulating EGFR activation in lung adenocarcinoma and positively regulating the stability of AURKA85. CBLC knockdown significantly decreases cell viability in H358 cells (45.13% inhibition rate) as well as colony formation. Targeting CBLC combined with paclitaxel further inhibits tumor growth85. Thus, CBLC expression may be an important determinant of the regulation of EGFR and AURKA sensitivity to KRAS G12C inhibitors53.
In a study comparing adagrasib (an oral inhibitor of mutant KRAS G12C protein) in combination with intravenous cetuximab (an anti-EGFR monoclonal antibody) vs. adagrasib alone in previously treated patients with metastatic colorectal cancer with mutated KRAS G12C, a response rate of 46% and median PFS of 6.9 months were observed in the combination therapy group, compared with a response rate of 19% and a median PFS of 5.6 months in the monotherapy group86. Although the data indicate an improvement with the dual inhibition of KRAS G12C and EGFR, some areas warrant further improvement in the management of patients with KRAS G12C mutation. The amplification of KRAS G12C is also relevant in this context. Colorectal cancer cell sublines (106 and RW7213) resistant to sotorasib and cetuximab express higher RAS-GTP levels than parental cells, and treatment only partly inhibits RAS-GTP. Targeted sequencing of the resistant sublines has identified amplification of KRASG12C in RW7213 resistant cells, which have homozygous (through loss of heterozygosity) and clonal KRASG12C mutation. In addition, copy-number analysis of the resistant RW7213 cells has revealed more than 20 copies of KRAS; the findings were further validated with fluorescence in situ hybridization87. Serial blood sample assessments in patients with colorectal cancer treated with adagrasib plus cetuximab, or sotorasib plus panitumumab have demonstrated several acquired alterations at low frequency, except for KRAS G12C amplification, which is a consistent mechanism of resistance observed with clinical progression87. After drug removal, KRAS G12C-amplified signaling switches to oncogene-induced senescence with mTOR hyperactivity that could be considered for treatment with senolytic agents87. Similarly, the novel KRAS G12 C inhibitor AZ’1569 has been used in a panel of KRAS G12C-mutant colorectal cell lines. AZ’1569 resistant cells exhibit amplification of KRAS G12C, as well as EpHA2 and c-MET activation. Similarly, KRAS amplification and AZ’1569 resistance is reversed after drug withdrawal, thereby reinforcing the clinical axiom of using of drug holidays when KRAS amplification is emerging88. Bcl-xL has been found to mediate resistance to AZ’1569, and combination treatment with AZ’1569 and navitoclax (a Bcl-2 inhibitor) has been found to attenuate tumor growth in KRAS G12C colorectal xenografts. BIM upregulation has been noted as well88. Further investigation of the behavior of leucine zipper-like transcriptional regulator 1 (LZTR1), a Golgi protein belonging to the BTB-Kelch superfamily, which interacts with the Cullin 3 (CUL3)-based E3 ubiquitin ligase complex, may prove valuable. LZTR1 promotes the polyubiquitination and degradation of RAS via the ubiquitin-proteasome pathway, thereby inhibiting RAS-MAPK signaling89,90. Therefore, loss of LZTR1 function might possibly contribute to resistance to KRAS G2C inhibitors with or without EGFR inhibitors. The potential LZTR1 function as a “RAS killer protein”90 warrants clarification, because several studies involve LZTR1 in KRAS degradation, as well as the abundance of multiple RAS GTPases, including RIT, MRAS, or SHOC2. LZTR1 depletion mobilizes and accumulates RIT1, KRAS, MRAS, and SHOC2, thus promoting non-canonical RAS signaling through MRAS-SHOC2-PPI activity91. Overexpression and accumulation of EGFR and AXL have been observed in LZTR1-mutant cancers, thus suggesting that co-inhibition targeted therapy may be beneficial92. As previously described, sulfasalazine inhibits SLC7A11 and has been identified to suppress NSCLC proliferation in AXL-expressing cell lines93. Further research is needed to understand the relevance of MET and SHOC2 in KRAS-mutant NSCLC cell lines (Figures 3 and 4), as well as in clinical settings. HGF has been reported to confer resistance to EGFR TKIs by inducing interceptor cross talk with integrin β4 (ITGβ4), EphA2, CCDP1, AXL, and JAK94. Multiple drugs have been developed to directly target KRAS G12C, among which the most well-known are sotorasib (AMG510) and adagrasib (MRTX849); however, many other direct oral inhibitors of KRAS are emerging53,95, such as GFH92595. A first-line trial combining GFH925 plus cetuximab in treating metastatic NSCLC (NCT05756153) is expected to provide further insights into the potential benefits of this combination therapy, and to alert researchers to potential putative mechanisms of resistance that may arise, probably those associated with KRAS G12C amplification.
EMT is a process in which epithelial cancer cells undergo a transformation leading to the loss of cell-cell adhesion and the acquisition of mesenchymal traits. This transformation promotes tumor progression, metastasis, and resistance to chemotherapy and other anti-cancer treatments. ZEB1 is a transcription factor involved in promoting EMT. Recently, RHOJ, a small GTPase, has been discovered to be expressed predominantly in EMT cancer cells. RHOJ interacts with proteins involved in the regulation of nuclear actin and therefore is a crucial regulator of EMT-associated resistance to chemotherapy96. Platinum (II) N-heterocyclic carbene complexes, novel forms of platinum derivatives, have shown promising activity in comparison with that of cisplatin in NSCLC cell lines with down-regulation of mesenchymal markers such as vimentin and tumor proliferation97. HECT, UBA, and WWE domain-containing protein 1 (HUWE1) is activated by HGF in NSCLC cell lines, and subsequently induces the degradation of T lymphoma invasion and metastasis inducing protein (TIAM1) and loss of cell-cell adhesion98 (Figure 5). These findings have reinforced the role of MET as a driving force in lung cancer progression and dissemination. The E3 ligase HUWEI ubiquitylates TIAM1 after HGF treatment in NSCLC cells. Depletion of HUWE1 in H1299, H358, H358, and H596 cells increases TIAM1 protein levels98. Interestingly, HUWE1 has multiple functions: it mediates the degradation of TP53 in a Mdm2-independent manner in NSCLC cells99 and additionally interacts with MCl-1, thereby causing degradation after DNA damage. HUWE1 also associates with the BRCA1-Merit40/RAP80 complex, thus leading to BRCA1 degradation100. Beyond its negative regulation of tumor suppressor genes such as TP53 and BRCA1, HUWE1 also positively regulates oncoproteins such as c-MYC101. RAC activity is controlled by several other guanine nucleotide exchange factors that exchange GDP for GTP and activate RAC1. Disassociation of RAC1 from Wave1 stimulates the activation of the actin regulated protein (ARP2/3) complex and promotes actin polymerization102. Elevated extracellular fluid viscosity, a condition imposed by mechanical loading in cancer cells, induces an ARP2/3-complex-dependent dense actin network that activates transient receptor potential cation vanilloid 4 (TRPV4), and consequently increases calcium influx and causes RHOA/myosin contractibility and cell migration103 (Figure 5). Elevated viscosity achieved by addition of 65 KDa methylcellulose into the cell culture medium enhances cell dissemination from 3D spheroids. Cells pre-exposed to elevated viscosity have been convincingly demonstrated to acquire TRPV4-depedency with HIPPO signaling inhibition and YAP1 upregulation103. Aquaporin 5, together with an Na+/H+ exchanger (NHE1), further contributes to cell swelling and increased membrane tension, thereby activating TRVP4103 (Figure 5; NHE1 is not depicted). Intriguingly, cystic fibrosis transmembrane conductance regulator (CFTR) protein, an epithelial anion channel that transports Cl- across epithelial surfaces, such as the respiratory tract (Figure 5), is diminished in lung cancer. Low CFTR mRNA levels have been associated with poor prognosis in NSCLC. Knockdown of CFTR in NSCLC cells increases the EMT phenotype, whereas CFTR overexpression suppresses cancer progression in vitro and in vivo104,105 (Figure 5). Interestingly, cigarette smoke impairs CFTR function, which is likewise implicated in the progression of chronic bronchitis106. The loss of CFTR induces intracellular Cl- accumulation, thereby inhibiting the chloride ion-sensitive with-no-lysine kinase 1 (WNK1) protein and resulting in the disinhibition of TRPV4107 (Figure 5). Ivacaftor has been shown to prevent CFTR loss in the lungs of mice with pneumonia and thus might be a potential treatment option for patients with acute respiratory distress syndrome107. Moreover, temozolomide has been found to be an efficient WNK1 activator107 (Figure 5).
Various intracellular processes further promote chemotherapy resistance and are refractory to other anticancer modalities. Cystic transmembrane conductance regulator (CFTR) expression is diminished in NSCLC. Mechanistically, low CFTR expression impairs the expression of the chloride ion-sensitive with-no-lysine kinase 1 (WNK1) protein, thereby leading to activation of transient receptor potential cation vanilloid 4 (TRPV4). TRPV4 serves as a mediator of calcium influx, promoting increased cell contractibility and cell migration. Ivacaftor is a potentiator of CFTR and temozolomide that activates WNK1. Acetazolamide blocks aquaporin 5 (AQP5), which is overexpressed in NSCLC and contributes to intracellular swelling. MET activates HECT, UBA, and WWE domain-containing protein 1 (HUWE1) E3 ligase, thus resulting in degradation of T lymphoma invasion and metastasis inducing protein 1 (TIAM1)—a guanine nucleotide exchange factor that regulates the GTPase RAC—and ultimately causing loss of cell adhesion. RAC1 is involved in the actin-related protein 2/3 (ARP2/3) complex dependent actin network.
Conclusions
The outlook for NSCLC has substantially improved in the past 2 decades, partly because of the identification of numerous driver mutations, particularly EGFR mutations, and the recognition of PD-L1 expression as a reliable marker for immunotherapy response in patients without driver mutations. However, despite the benefits of targeted next generation sequencing in tissue or plasma, treatment with single targeted agents, or even combination treatments, ultimately lead to disease progression after a prolonged period of remission. Currently, focus is being placed on cell adoptive therapies, although their efficacy can be decreased by factors such as the tumor microenvironment and overlooked signaling pathways that have high potential for clinical application. While unraveling the intricate workings of cancer cells poses challenges, exploring shared characteristics among different types of NSCLCs and other cancers remains crucial. Although the quest to unravel the secrets of cancer cells may appear daunting, it inspired us to adopt a strategy centered around identifying commonalities among diverse NSCLCs and other cancer types. This review is grounded in fundamental principles that impact the effectiveness of various therapies, leveraging insights gained from studying preclinical cell lines as well as NSCLC patient tissues or blood samples. By presenting working hypotheses derived from our analysis, we aim to advance research and potentially facilitate the translation of these findings into practical applications for the management of NSCLC and other cancers.
Conflict of interest statement
No potential conflicts of interest are disclosed.
Author contributions
Conceived and designed the analysis: Rafael Rosell, Jordi Codony-Servat, Anisha Jain, Eloisa Jantus-Lewintre and María González-Cao.
Collected the data: Rafael Rosell, Eloisa Lewintre, Jordi Codony-Servat, María González-Cao.
Prepared the figures: Rafael Rosell, María González-Cao.
Wrote the paper: Rafael Rosell.
Writing-review & editing: Rafael Rosell, Jordi Codony Servat, Eloisa Jantus-Lewintre, María González-Cao.
Acknowledgements
The concept diagrams were created with Biorender.com.
- Received April 4, 2023.
- Accepted June 5, 2023.
- Copyright: © 2023, Cancer Biology & Medicine
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY) 4.0, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source are credited.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵