Our research efforts have resolved a mechanism that permits emergence of new mutations in TKI-treated animal models of NSCLC. We attribute this process to the ability of EGFR-specific TKIs to instigate apoptosis of EGFR mutation-bearing cells. The many attempts to find a mechanism that possibly underlies the apparent drug-induced accelerated evolution led us to the response of bacteria to antibiotics. In 1975 Miroslav Radman reported an inducible bacterial DNA mutagenesis system, the SOS response, which might explain the link between genotoxic stress and adaptive mutagenesis of EGFR. He found that released fragments of single stranded DNA act as the sensors that initiate transcriptional programs and mutate the bacterial genome. The major endogenous mechanism of mutagenesis (i.e. mutator) in E. coli is DNA polymerase V (polV), which initiates virtually all SOS mutagenesis. PolV belongs to the group called Y family DNA polymerases, which promote translesion synthesis (TLS) of DNA. These polymerases exhibit low fidelity, thereby increase mutagenesis rates when they are engaged in DNA replication. We previously investigated whether the treatment of lung cancer with TKIs similarly engages hypermutators. Because GAS6 (growth arrest-specific protein 6), AXL’s ligand, is upregulated in cycling drug-persister cells and it binds with externalized phosphatidylserine of apoptotic bodies, we assumed that the GAS6-AXL module acts as a sensor that stimulates SOS-like reactions in response to TKIs. In line with this prediction and with previous reports that associated AXL and GAS6 with intrinsic resistance to TKIs, we found that AXL overexpression can up-regulate low-fidelity DNA polymerases and downregulate DNA repair enzymes. Moreover, simultaneously inhibiting AXL and EGFR completely blocked relapses in animal models. Metabolomic analysis uncovered yet another intrinsic mutator that relates to the dependency of DNA replication on balanced pools of deoxyribonucleotides (dNTPs). By activating MYC and purine synthesis, AXL disbalances the pools of dNTPs, which can influence polymerase proofreading and mutator phenotypes. In conclusion, pharmacological stress-driven mutagenesis might be shared by eukaryotes and unicellular organisms, arguing against prevailing assumptions that mutations occur purely stochastically. Moreover, blocking the mutators and their upstream control might identify new strategies to prevent or significantly delay the onset of cancer relapse post treatment.
In parallel to our attempts to resolve EGFR secondary mutagenesis and delay onset of resistance to TKIs, we studied yet another strategy, namely: enhancing responses of patients with EGFR mutations to immune checkpoint blockers (ICBs). It is well documented that EGFR mutant tumors exhibit relatively low response rates to ICBs. Therefore, understanding mechanisms underlying resistance of EGFR mutant patients to immunotherapy is urgently needed. An important clue as to the mechanism driving resistance of the EGFR mutant group to ICBs has been provided by clinical observations made with patients expressing rare mutant forms of EGFR, who respond relatively well to ICBs. Because treatment outcomes vary by EGFR allele, we assumed that features intrinsic to the tumor cells drive the relatively high resistance of the EGFR+ group. While studying intrinsic factors that potentially underlie immunosuppression, we discovered previously unknown strong physical and functional interactions between phospholipase C gamma (PLC-g) and PD-L1. These observations might guide future attempts to combine anti-PLC drugs and PD-L1 inhibitors for the benefit of patients with lung cancer.
