Targeting PEA3 transcription factors to mitigate small cell lung cancer progression

Abstract

Small cell lung cancer (SCLC) remains a lethal disease with a dismal overall survival rate of 6% despite promising responses to upfront combination chemotherapy. The key drivers of such rapid mortality include early metastatic dissemination in the natural course of the disease and the near guaranteed emergence of chemoresistant disease. Here, we found that we could model the regression and relapse seen in clinical SCLC in vitro. We utilized time-course resolved RNA-sequencing to globally profile transcriptome changes as SCLC cells responded to a combination of cisplatin and etoposide-the standard-of-care in SCLC. Comparisons across time points demonstrated a distinct transient transcriptional state resembling embryonic diapause. Differential gene expression analysis revealed that expression of the PEA3 transcription factors ETV4 and ETV5 were transiently upregulated in the surviving fraction of cells which we determined to be necessary for efficient clonogenic expansion following chemotherapy. The FGFR-PEA3 signaling axis guided the identification of a pan-FGFR inhibitor demonstrating in vitro and in vivo efficacy in delaying progression following combination chemotherapy, observed inhibition of phosphorylation of the FGFR adaptor FRS2 and corresponding downstream MAPK and PI3K-Akt signaling pathways. Taken together, these data nominate PEA3 transcription factors as key mediators of relapse progression in SCLC and identify a clinically actionable small molecule candidate for delaying relapse of SCLC.

Fig. 1. SCLC cellular models allow for in vitro modeling of acquired resistance and response and relapse dynamics.

A Experimental timeline of chemotherapy exposure. B Demonstration of in vitro acquired resistance in H524 with iterative rounds of cisplatin treatment at 1 μM concentration. Viable cell number at each time point was measured in triplicate. C Cisplatin and etoposide dual titrations for determination of treatment concentrations. Estimated IC₅₀ doses by line are as follows: cisplatin 500 nM etoposide 500 nM for H82 and H526, cisplatin 1 μM etoposide 1 μM for H1963. D Response and regrowth curves following 72 h treatment with combination cisplatin and etoposide at experimentally determined 72-h IC₅₀ doses. Viable cell number at each time point was measured in triplicate. E EdU labeling at various time points following cisplatin 500 nM and etoposide 500 nM exposure for 72 h in H82.

Fig. 2. Persister-enriched time points are transcriptionally distance from initial and recovered time points and demonstrate a transient diapause-like state.

A Timeline of sampling for RNA sequencing in cisplatin and etoposide challenge. B Summary of gene set enrichment analysis (GSEA) results comparing each time point to initial. Positive normalized enrichment scores (NES) represent gene sets enriched in the initial time point. C Two-dimensional GSEA (2D GSEA) comparing the intermediate to initial and intermediate to recovered comparisons and demonstrating a high positive correlation. D 2D GSEA comparing the intermediate to initial with various stages of murine embryo development. E Diapause gene signature scoring of each time point.

Fig. 3. Drug-tolerant persisters are enriched in expression of PEA3 transcription factors ETV4 and ETV5.

A Expression of top intermediate-enriched transcription factors in each cell line dataset. B Expression of PEA3 and ELF subgroups of transcription factors plotted against viable cell counts at each time point.

Fig. 4. ETV4 and ETV5 are regulators of regrowth following combination chemotherapy in SCLC.

A Schematic demonstrating workflow for generation of loss-of-function mutant lines. B Cellular growth curves of monoclonal sublines derived from H526 parental line comparing wild-type lines to mutant lines generated from CRISPR-Cas9 by CellTiter-Glo. Each time point was measured in quadruplicate. C Clonogenic regrowth assay in monoclonal sublines following challenge with combination 500 nM cisplatin and etoposide for 72 h. D Clonogenic regrowth assay in stably-transduced H82 and H526 lines expressing the following shRNA: scrambled, ETV4, ETV5, ETV4 and ETV5. Cells were treated with 500 nM cisplatin and etoposide for 72 h and then seeded in 1% methylcellulose for quantification of clonogenic regrowth. Non-parametric t-test was used to determine statistical significance between groups. E Clonogenic regrowth assay in stably-transduced H526 lines expressing the following cDNA: GFP, ETV4, ETV5, ETV4, and ETV5. As above, cells were treated with 500 nM cisplatin and etoposide for 72 h and then seeded in 1% methylcellulose for quantification of clonogenic regrowth. Non-parametric t-test was used to determine statistical significance between groups.

Fig. 5. In vitro evaluation of pan-FGFR inhibitors in preventing recurrence following combination chemotherapy.

A Generalized FGFR intracellular signaling pathway. B Experimental scheme for kinase inhibitor evaluation. C Viable cell counts at 21 days following 500 nM cisplatin and etoposide challenge across three SCLC cell lines: H146, H524, and H526. D Evaluation of single agent cytotoxicity of pan-FGFR inhibitors at 3 day and 7 day time points across three SCLC cell lines: H209, H524, and H526. E Dual titrations of pan-FGFR inhibitors infigratininb and erdafitinib against cabozantinib in H82 (left) and H526 (right).

Fig. 6. LY2874455 alone demonstrates inhibitory activity of downstream MAPK and PI3K-Akt signaling pathways.

A Time course analysis of phosphorylation of Erk1/2 (T202/Y204) and Akt (S473) following exposure to 1 μM LY2874455 in H526. B Immunoblot evaluation of phosphorylation in FRS2 (Y196), Erk1/2 (T202/Y204), and Akt (S473) following 16 h treatment with 1 μM of each pan-FGFR inhibitor: LY2874455, infigratinib, and erdafitinib in H526. C Immunoblot evaluation of ETV4 and ETV5 protein expression following 3 and 7 days of exposure to the pan-FGFR inhibitors LY2874455, infigratinib, and erdafitinib in H82 and H526. D 2D GSEA comparing H526 ETV4/5 double mutant to H526 wildtype and H526 LY2874455 exposed to vehicle exposed comparisons to one another.

Fig. 7. In vivo efficacy of LY2874455 as a single agent and in combination with standard-of-care for treatment of SCLC.

A Experimental scheme for in vivo evaluation of single agent LY2874455 in H526 xenograft model. B H526 xenograft tumor growth comparing daily intraperitoneal administration of 12 mg/kg LY2874455 (n = 10) compared to vehicle (n = 10). Error bars represent standard error of the mean. C Representative photograph of tumors at 14 days of treatment. D Experimental scheme for evaluation of LY2874455 in combination with cisplatin and etoposide. E H526 xenograft tumor growth comparing daily intraperitoneal administration of 12 mg/kg LY2874455 in combination with cisplatin and etoposide (n = 6) compared to only cisplatin and etoposide (n = 7). Error bars represent standard error of the mean. F Representative photograph of tumors at 28 days of treatment. CE = cisplatin and etoposide. G Representative images of hematoxylin and eosin stained sections.