Dynamic variations in epithelial-to-mesenchymal transition (EMT), ATM, and SLFN11 govern response to PARP inhibitors and cisplatin in small cell lung cancer

Abstract

Small cell lung cancer (SCLC) is one of the most aggressive forms of cancer, with a 5-year survival <7%. A major barrier to progress is the absence of predictive biomarkers for chemotherapy and novel targeted agents such as PARP inhibitors. Using a high-throughput, integrated proteomic, transcriptomic, and genomic analysis of SCLC patient-derived xenografts (PDXs) and profiled cell lines, we identified biomarkers of drug sensitivity and determined their prevalence in patient tumors. In contrast to breast and ovarian cancer, PARP inhibitor response was not associated with mutations in homologous recombination (HR) genes (e.g., BRCA1/2) or HRD scores. Instead, we found several proteomic markers that predicted PDX response, including high levels of SLFN11 and E-cadherin and low ATM. SLFN11 and E-cadherin were also significantly associated with in vitro sensitivity to cisplatin and topoisomerase1/2 inhibitors (all commonly used in SCLC). Treatment with cisplatin or PARP inhibitors downregulated SLFN11 and E-cadherin, possibly explaining the rapid development of therapeutic resistance in SCLC. Supporting their functional role, silencing SLFN11 reduced in vitro sensitivity and drug-induced DNA damage; whereas ATM knockdown or pharmacologic inhibition enhanced sensitivity. Notably, SCLC with mesenchymal phenotypes (i.e., loss of E-cadherin and high epithelial-to-mesenchymal transition (EMT) signature scores) displayed striking alterations in expression of miR200 family and key SCLC genes (e.g., NEUROD1, ASCL1, ALDH1A1, MYCL1). Thus, SLFN11, EMT, and ATM mediate therapeutic response in SCLC and warrant further clinical investigation as predictive biomarkers.

Keywords: ATM; EMT; PARP inhibitor; SCLC; SLFN11.

Figure 1. PDX Models with High SLFN11 and Low ATM Expression Levels Are More Sensitive to Talazoparib

A. Examples of PDX response to single-agent talazoparib and percent change in tumor volume from baseline for individual PDX models. B, and C. Regression analysis of 12 PDX models grouped as having progressive disease (PD, n=6), stable disease (SD, n=4), or partial response (PR, n=2) following treatment with single-agent talazoparib identifies high SLFN11 and low ATM expression and CHK1, IGF1R beta, and IRS1 as markers of sensitivity. D. Analysis of CHK1, ATM, and SLFN11 protein and mRNA expression across response groups. E. EMT score is correlated with PDX resistance to talazoparib. F. Immunolocalization of SLFN11 and ATM in PDX models from the 3 response groups. SLFN11, but not ATM H score, predicts PDX response. G. Myriad HRD score and FoundationOne non-germline mutational burden do not predict PDX response to talazoparib. H. Oncoprint representation of FoundationOne mutations identified in DDR genes or mutations previously shown to occur in SCLC.

Figure 2. SLFN11 Protein Levels Are Correlated with Drug Sensitivity in SCLC

A. A correlation analysis of cisplatin, talazoparib, and olaparib sensitivity (IC₅₀ values) and 171 proteins in 51 SCLC cell lines shows that SLFN11 is the strongest predictor of sensitivity to both cisplatin and PARP inhibition. E-cadherin, but not ATM, is also correlated with drug sensitivity. B. SLFN11 expression is bimodal in SCLC cell lines, showing a switch-like pattern, and these naturally formed groups are correlated with drug sensitivity. Cell lines with higher SLFN11 protein expression levels have greater sensitivity to cisplatin and talazoparib but not olaparib. C. Treatment of H209 and H526 SCLC cell lines with 1 μM cisplatin, 1μM olaparib, or 100 nM talazoparib for 72h reduced SLFN11 levels compared with vehicle-treated cells*, P<0.0005. Treatment of A549, H1944, HCC827 NSCLC cell lines with 1 μM cisplatin for 96h reduced SLFN11 levels, but not Calu6. *, P<0.002. D. Waterfall plot of drug sensitivity in SCLC. A comparison of high SLFN11 protein levels and the IC₅₀ values of 526 cancer drugs in 61 cell lines identified several drugs with similar targets, including PARP1 inhibitors, alkylating agents, TOP1 inhibitors, TOP2A/B inhibitors, and DNA synthesis inhibitors. E. A drug interaction network shows classes of oncology agents that are effective in SCLC cell lines with high SLFN11 levels. F. SLFN11-high SCLC has high expression levels of Type I IFN signaling molecules. IPA analysis identified IFN signaling as the top pathway associated with SLFN11 expression in SCLC patient tumors (P=6.6*10⁻⁶). Heatmap of Type I IFN signaling genes and immunotherapy target genes ranked by association with SLFN11 levels in 70 tumors from treatment-naïve SCLC patients (FDR=0.2).

Figure 3. High SLFN11 and Low ATM Levels Maintain Sensitivity to PARP Inhibition and Chemotherapy

A, and B. Silencing of SLFN11 with siRNA in DMS79 and H209 SCLC cell lines (which have high SLFN11 expression levels, are sensitive to PARP inhibition, and have no ATM mutations), increases the cells’ resistance to cisplatin, talazoparib and olaparib. C, and D. ATM knockdown increases sensitivity to cisplatin, talazoparib, and olaparib. siRNA effectively reduced SLFN11 and ATM. E. ATM inhibition sensitizes cell lines with low SLFN11 expression to PARP inhibition. The ATM inhibitor KU55933 plus talazoparib or olaparib was more effective in SCLC cell lines with lower SLFN11 levels.

Figure 4. SLFN11 is Regulated by PARP1 and EHF

A. Silencing of SLFN11 in H209 and DMS79 is insufficient to induce the expression of H2AX, a marker of DNA damage. Stimulation with olaparib, talazoparib, and cisplatin induced γH2AX in scrambled (SCR) but not SLFN11-knockdown cells. B. Silencing of SLFN11 or ATM does not affect PARP1 levels. Silencing of PARP1 in H209 and H526 cells reduces SLFN11 levels. C. EHF expression is elevated in SCLC tumors compared to normal adjacent tissues and is correlated with SLFN11 expression in tumors from treatment-naïve patients with early-stage SCLC. Silencing of EHF reduces SLFN11 mRNA and protein expression levels.

Figure 5. E-Cadherin Level Predicts Drug Sensitivity in SCLC

A. Heatmap of SCLC cell lines ranked by E-cadherin protein levels demonstrating a bimodal distribution pattern (BI=1.72). Cell lines with EMT scores greater than 0 are classified as mesenchymal. Correlation of E-cadherin expression and the IC₅₀ values of PARP inhibitors, standard-of-care chemotherapy drugs, and BCL2 inhibitors. miR200 family microRNAs are strongly correlated with E-cadherin levels and EMT score. Expression analysis of a subset of genes involved in SCLC progression that are correlated with E-cadherin expression. B. A heatmap of 70 tumor samples from treatment-naïve SCLC patients ranked by EMT score shows the expression analysis of the same subset of genes as in the SCLC cell lines. C. High E-cadherin levels are associated with sensitivity to several drug classes, including PARP inhibitors, chemotherapy drugs, BCL2 inhibitors, and IGF1R inhibitors. D. A drug interaction network shows classes of drugs that are effective in SCLC cell lines with high E-cadherin levels.