Novel EGFR-mutant mouse models of lung adenocarcinoma reveal adaptive immunity requirement for durable osimertinib response

Abstract

Lung cancers bearing oncogenically-mutated EGFR represent a significant fraction of lung adenocarcinomas (LUADs) for which EGFR-targeting tyrosine kinase inhibitors (TKIs) provide a highly effective therapeutic approach. However, these lung cancers eventually acquire resistance and undergo progression within a characteristically broad treatment duration range. Our previous study of EGFR mutant lung cancer patient biopsies highlighted the positive association of a TKI-induced interferon γ transcriptional response with increased time to treatment progression. To test the hypothesis that host immunity contributes to the TKI response, we developed novel genetically-engineered mouse models of EGFR mutant lung cancer bearing exon 19 deletions (del19) or the L860R missense mutation. Both oncogenic EGFR mouse models developed multifocal LUADs from which transplantable cancer cell lines sensitive to the EGFR-specific TKIs, gefitinib and osimertinib, were derived. When propagated orthotopically in the left lungs of syngeneic C57BL/6 mice, deep and durable shrinkage of the cell line-derived tumors was observed in response to daily treatment with osimertinib. By contrast, orthotopic tumors propagated in immune deficient nu/nu or Rag1^-/- mice exhibited modest tumor shrinkage followed by rapid progression on continuous osimertinib treatment. Importantly, osimertinib treatment significantly increased intratumoral T cell content and decreased neutrophil content relative to diluent treatment. The findings provide strong evidence supporting the requirement for adaptive immunity in the durable therapeutic control of EGFR mutant lung cancer.

Keywords: Adaptive immunity; EGFR; GEMM; Lung adenocarcinoma; Tyrosine kinase inhibitor.

Figure 1.. Induction of T cell-focused gene expression signatures in on-treatment biopsies from EGFR mutant LUAD patients associates with greater time to progression.

Using previously reported RNAseq data from pre- and on-treatment biopsies collected from eight EGFR mutant LUAD patients under informed consent [18], the sums of the 18 genes in the signature described by (A) Ayers et al [44] and the 25 genes in the signature described by (B) Lau et al [45] were calculated for each patient-derived specimen and converted to Z-scores. The resulting values for the Ayers (A) and Lau et al (B) signatures were binned by the time to progression (TTP) values of less than or greater than 12 months where the eight patients (#’s 1–8) exhibited TTP of 6.2, 7, 7, 8.6 13, 13.1. 16 and 16.3 months, respectively (mean = 10.9, median = 10.8 months). The matched Z-scores for each patient at baseline and upon re-biopsy (patient IDs are in parentheses) are graphed and analyzed by paired t-tests. The p values for the two analyses are indicated.

Figure 2.. Transgenic mouse strains for Cre-mediated expression of Egfr^del19 and Egfr^L860R.

Egfr-transgenic mice (Rosa-loxP-STOP-LoxP- EGFR L860R/del19-mC3-WPRE) were created by the Mouse Genetics Core at National Jewish Health. One strain expresses the murine Egfr cDNA encoding an exon 19 deletion and one with an L860R mutation (equivalent to the human L858R mutation). The resulting mice were bred with p53^LoxP mice from Jackson Laboratories (B6.129P2-Trp53^tm1Brn/J; #008462). To induce expression of the oncogenic EGFR transgene, an Adeno-Cre (AdCre) virus was instilled intratracheally (2.5×10⁷ PFU/mouse) to excise the LoxP-STOP-LoxP sequence. (A) Schematic of the recombined Rosa locus and (B) breeding crosses. (C) Gross dissection of multifocal tumors in the right and left lungs resulting from AdCre instillation. Left panel: Egfr^+/L860R;Trp53^+/f lungs at 24 weeks post AdCre. Right panel: Egfr^+/del19;Trp53^+/f lungs at 11 weeks post AdCre. Arrows indicate visible tumors. (D) H&E-stained tumor sections showing adenocarcinoma of the lungs in this model resulting from AdCre installation. Left panels: an Egfr^L860R/L860R;Trp53^+/+ tumor at 14 weeks post AdCre. Right panels: an Egfr^+/del19;Trp53^+/f tumor at 8 weeks post AdCre. Images captured at 100x and 400x. The white boxes in the 100X images indicate positions of the 400x images.

Figure 3.. Murine EGFR-mutant LUAD cell lines are responsive to EGFR inhibitors in vitro.

(A) The three EGFR-mutant murine cell lines were treated for 10 days with increasing concentrations of EGFR-specific TKIs, gefitinib or osimertinib and cell number was quantified with CyQUANT reagent. IC₅₀ values were calculated with Prism 9 and presented in the graph. The data shown are the means of two independent experiments, each performed with technical triplicates. (B-D) The murine EGFR mutant cell lines were treated (2 hrs) with the indicated EGFR targeting TKIs (gefitinib, afatinib, osimertinib) as well as the ALK/MET-targeting TKI, crizotinib as a negative control. Cell lysates were submitted to SDS-PAGE and immunoblotted for phospho-Y1068 and total EGFR, phospho-S483 and total AKT, and phospho-Thr202/Tyr204 and total ERK.

Figure 4.. Therapeutic response of orthotopic murine EGFR tumors to osimertinib in immunocompetent and immunodeficient mice.

Murine EGFRdel19.1 and del19.2 cells (500,000 cells/mouse) were injected into the left lungs of syngeneic C57BL/6 (A, B) or nu/nu (C) or nu/nu and Rag1^−/− (D) mice. After 10 days, the mice were imaged by μCT to generate pre-treatment tumor volumes and randomized into osimertinib (5 mg/kg) or control treatment groups. The tumor-bearing mice were submitted to weekly μCT scans and the data are presented as fold of the initial pre-treatment volumes (means and SEM). For the C57BL/6 experiments, the initial tumor volumes (mean ± SEM) for the diluent and osimertinib-treated groups were 19.9 ± 4.4 mm³ (n=14) and 13.5 ± 2.8 mm³ (n=14) and 11.9 ± 4.5 mm³ (n=8) and 11.6 ± 3.0 mm³ (n=9) for mEGFRdel19.1 and mEGFRdel19.2, respectively. For the Nu/Nu experiments, the initial tumor volumes for the diluent and osimertinib-treated groups were 2.9 ± 0.7 mm³ (n=6) and 9.0 ± 1.6 mm³ (n=9) and 21.7 ± 6.5 mm³ (n=6) and 13.6 ± 4.0 mm³ (n=9) for mEGFRdel19.1 and mEGFRdel19.2, respectively. For the mEGFRdel19.2 experiment in Rag1^−/− mice, the initial tumor volumes for the control and osimertinib-treated groups were 3.8 ± 1.5 mm³ (n=5) and 7.0 ± 2.5 mm³ (n=6), respectively.

Figure 5.. Osimertinib-treatment increases T cell content in orthotopic murine EGFRdel19 tumors.

Murine EGFRdel19.1 (A and C) and 19.2 (B) cells were injected into C57BL/6 mice, tumors were permitted to establish for 14–21 days and then the mice were treated for 4 days with diluent or osimertinib. In A and B, lungs were harvested, inflated with formalin, paraffin-embedded and 5 μm sections were submitted to immunofluorescence staining with anti-CD3 and anti-CD8 antibodies (see Materials and Methods). Total CD3+ cells as well as CD3+CD8+ and CD3+CD8- cells were quantified in 3 distinct fields for each tumor with Cell Profiler 4.2.4 and the resulting values were averaged. The data in A and B are the means and SEM of 4 and 3 independent tumors for control and osimertinib-treated tumors, respectively, and were analyzed by student’s two-way t-tests. (C) Single-cell suspensions were generated from the left mEGFRdel19.1 tumor-bearing lungs and submitted to flow cytometry analysis for innate immune cells using the previously described gating strategy for neutrophils, monocytes and MacA and MacB macrophages [56]. Differences between groups was analyzed with two-way T tests. Data are presented as the mean ± SEM where ** indicates p-value less than 0.01; ns indicates not significant.

Figure 6.. Transcriptional regulation of inflammation-related Hallmark pathways and chemokines and cytokines in response to osimertinib treatment.

The murine EGFR mutant cell lines were treated in vitro for 1 to 6 days with DMSO or osimertinib (100 nM) and RNA was purified and sequenced. (A) For this analysis, the different time points were considered as replicates (n=4) and DMSO vs. alectinib-treated samples analyzed with GSEA using the Hallmark Pathways. The heatmap presents the normalized enrichment scores (NES) in the osimertinib-treated samples (see color bar for relative scores). (B) The RNA expression values for the selected genes were converted to Z-scores and presented in a heatmap format.