Imlunestrant a next-generation oral SERD overcomes ESR1 mutant resistance in estrogen receptor-positive breast cancer

Abstract

Estrogen receptor α (ER) is a critical driver of tumorigenesis and tumor progression in most breast cancers. Endocrine therapies (ET) targeting ER are central to treating hormone receptor-positive breast cancer, but resistance poses a clinical challenge. Some resistance mechanisms, particularly those involving estrogen-independent activity such as the ESR1 mutations, rely on ER signaling, supporting the need for next-generation ET. We investigated the preclinical efficacy of imlunestrant, an oral selective ER degrader, in ER-positive breast cancer preclinical models, including models harboring the Y537S ESR1 mutation, an activating mutation. Imlunestrant demonstrated antagonistic activity and effective degradation of both WT and mutant ER, resulting in cell growth suppression. In vivo, imlunestrant outperformed fulvestrant, leading to tumor regression in a patient-derived xenograft harboring the Y537S ESR1 mutation. Cyclic mutiplexed immunofluorescence and transcriptomic analysis revealed enhanced cell cycle arrest and downregulation of estrogen-responsive genes with imlunestrant treatment. Additionally, a genome-wide CRISPR knock-out screen identified several vulnerabilities that were either persistent or acquired after imlunestrant treatment, providing a rationale for future studies of combination treatments with imlunestrant. Collectively, these results highlight the on-target and selective activity of imlunestrant, which can circumvent resistance engendered by the Y537S ESR1 mutation.

Keywords: Breast cancer; Cell biology; Oncology.

Figure 1. Antagonistic and ER degradation activity of imlunestrant.

(A) Cell growth studies depicted as cell number fold change for MCF7 (left) or T47D (right) cells during treatment up to 8 days with DMSO or imlunestrant (IML). Error bars denote ± SD. Two-way ANOVA with Dunnett’s multiple comparisons test. (B) Normalized cell proliferation for MCF7 (left) or T47D (right) cells expressing ER-WT or ER-Y537S and treated with a dose-response of imlunestrant for 5 days. Error bars with average ± SD. Two-way ANOVA with Šidák’s multiple comparisons test. (C) Normalized luciferase signal in MCF7 ERE-LUC cells after hormone deprivation (HD) and treatment with or without E₂ (1 nM) and imlunestrant (0, 50 nM, 500 nM, or 1,000 nM) for 1 hour, (D) 6 hours, or (E) 24 hours. Bar graph with averages. Error bars denote ± SD. Two-way ANOVA with Tukey’s multiple comparisons test. (F) Western blot of whole cell lysates for ER and GAPDH in MCF7 cells in HD media treated with fulvestrant (FULV; 100 nM, 24 hours) or imlunestrant (IML; 100 nM from 0–72 hours), as indicated. Densitometry of ER levels normalized to DMSO and GAPDH. (G) Western blot for ER WT (lower band) and ER mutant (HA-tagged, upper band) and GAPDH in MCF7 ER-Y537S cells in HD media treated with FULV (100 nM, 24 hours) or IML (100 nM, 0 to 72 hours), as indicated. Densitometry of ER mutant levels normalized to DMSO and GAPDH. (H) Colony assay crystal violet staining results from MCF7 or T47D cells with ER WT or ER-Y537S expression and treatment with imlunestrant (0 to 100 nM, as indicated) for 2 weeks in full media. (I) Relative confluency of colony assay crystal violet staining in H. Bar graph with average ± SD. One-way ANOVA with Dunnett’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2. Transcriptional effects of imlunestrant.

(A) Cartoon schematic of the RNA-seq treatment conditions. (B) Volcano plot of the differentially expressed genes following estradiol treatment or (C) doxycycline induction of ER-Y537S in MCF7 cells. Dotted lines denote chosen cutoffs |log₂FC| ≥ 1 and FDR < 10%. Orange is upregulated and blue is downregulated following treatment. n = 2 replicates per group. ESR1 and estrogen-regulated genes are labeled with gene names when significant. (D) Bar plot of the number of differentially expressed genes in MCF7 or (E) T47D cells following IML or FULV treatment in cells expressing WT-ER or ER-Y537S. (F) Dot plot of gene set testing results in ER-Y537S-expressing cells for the Hallmark gene collection using cameraPR (10% FDR threshold per column).

Figure 3. Imlunestrant treatment leads to decreased proliferation and tumor regression.

(A) Tumor volumes of ER-Y537S PDX treated with vehicle (VEH, black), 5 mg subcutaneous fulvestrant (FULV, purple) per week, or 15mg/kg oral daily imlunestrant (IML, orange). Error bars denote ± SD. Two-way ANOVA with Dunnett’s multiple comparisons test. n = 3 mice/treatment on day 10 and n = 5 mice/treatment on day 28. (B) Body weight of the PDX-bearing mice treated with VEH, FULV, or IML. Error bars denote ± SD. One-way ANOVA with Tukey’s multiple comparisons test. (C) Representative images of IHC staining on ER-Y537S PDX day 10 for ER or PR or Ki67; scale bar: 100 μm. (D) IHC staining index for ER or (E) PR or (F) % Ki67 positive on ER-Y537S PDX day 10 tissue. Box and whisker plot with maximum and minimum. One-way ANOVA with Tukey’s multiple comparisons test, n = 2 mice, 4–5 images from each mouse. (G) Representative images of IHC staining on ER-Y537S PDX tissue treatment day 28 stained for ER or PR or Ki67. 20× magnification, scale bar: 100 μm. (H) IHC staining index for ER or (I) PR or (J) % Ki67 positive on day 28 ER-Y537S PDX tissue. Box and whisker plot with maximum and minimum. One-way ANOVA with Tukey’s multiple comparisons test, n = 2 mice, 4–5 images from each mouse. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4. Multiparametric immunofluorescence analysis.

(A) Representative immunofluorescence images for DAPI, Ki67, p21, Rb phosphorylated at Serines 807/811 (phospho-Rb), and merged staining from ER-Y537S patient-derived xenograft (PDX) treated for 10 days with vehicle, fulvestrant, or imlunestrant. n = 3 mice/treatment. Scale bar: 50 μm. (B) Multivariate proliferation index (MPI) quantification of proliferative fraction, nonproliferative fraction, and arrested fraction of epithelial cells after treatment with VEH (vehicle), FULV (fulvestrant), or IML (imlunestrant) for 10 days. Each point represents 1 mouse. Bar plot with average ± SD. One-way ANOVA with Tukey’s multiple comparisons test. (C) Representative immunofluorescent images after 28 days of treatment. n ≥ 4 mice/treatment. Scale bar: 100 μm. (D) MPI quantification of immunofluorescence after treatment for 28 days. One-way ANOVA with Tukey’s multiple comparisons test. (E) Pseudotime circle-fitted distribution of proliferative cells from ER-Y537S PDX tumors treated with VEH, FULV, or IML for 10 days or (F) 28 days. Scatter plot, each point represents a cell with n ≥ 1,909 cells/treatment. Orange circle indicates G1 cell cycle arrest in IML treated tumors. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5. Transcriptomic effects of imlunestrant.

(A) Bar plot of the number of differentially expressed genes in ER-Y537S patient-derived xenograft (PDX) treated with fulvestrant (FULV) or imlunestrant (IML) versus vehicle (VEH). Bars are orange for upregulated genes and blue for downregulated genes. Treatment duration for 10 days or 28 days. n ≥ 3 mice per group. (B) Heatmap of the row-centered variance-stabilized counts for the top 500 variable genes in ER-Y537S PDX tumors treated for 10 days with VEH, FULV, or IML. k-medoids clustering on 2 clusters with predominantly IML-downregulated genes and IML-upregulated genes following treatment for 10 days or (C) PDX treatment for 28 days. (D) Dot plot of gene set testing results for the Hallmark gene collection using cameraPR (10% FDR threshold per column). (E) Imlunestrant (green) docked to x-ray crystal structure of the ER LBD harboring the Y537S mutation (PDB: 9bu1). Proximity to residues important for ligation (orange) reveals imlunestrant occupation of ligand binding pocket. (F) Ligand binding pose comparison between imlunestrant and a derivative of the aliphatic SERD ICI 164,384 (PDB: 7r62, a derivative of estradiol that is closely related to fulvestrant) in cyan illustrates side arm conformation in relation to D351 (magenta) and S537 mutation (yellow).

Figure 6. Genome-wide CRISPR screen in the presence of imlunestrant treatment.

(A) Cartoon schematic of the genome-wide CRISPR/Cas9-knockout (KO) screen; (step 1) T47D cells are infected with the H3 lentiviral library of gRNAs, (step 2) infected cells are selected using puromycin, (step 3) cells are collected for untreated baseline (Day 0), or treated with DMSO or 1 nM imlunestrant, (step 4) after treatment for 14 or 31 days cells are collected, gRNAs amplified, and sequenced, followed by (step 5) gRNA analysis using MAGeCK algorithm. (B) MAGeCK β score and ranking results for gRNAs enriched or depleted after treatment with DMSO versus Day 0 for 14 days or (C) 31 days. (D) MAGeCK β score and ranking results for gRNAs enriched or depleted after treatment with imlunestrant versus Day 0 for 14 days or (E) 31 days. Scatter plot with each dot represents a gene in the genome-wide gRNA library, red are enriched after treatment (β ≥ +1), blue are depleted after treatment (β ≤ –1), black are insignificant after treatment, genes in the ER pathway are orange, the CDK pathway are purple, and the MTORC1 pathway are green. (F) β scores for individual genes depleted after imlunestrant treatment in the ER pathway, (G) the CDK pathway, or (H) the MTORC1 pathway.

Figure 7. Imlunestrant treatment increases the essentiality of genes related to oxidative phosphorylation.

(A) Nine-square scatter plot of the gene β scores in day 31 IML or day 31 DMSO treated T47D cells compared with day 0. Vertical and horizontal dotted lines denote ± 1 SD of DMSO or IML β scores, respectively, and diagonal dotted lines denote ± 1 SD of the difference in β scores (IML – DMSO). Green are gRNAs enriched in only IML (IML up; IML β score > 1 and DMSO β score < 1) and orange are gRNAs depleted only in IML (IML down; IML β score < –1 and DMSO β score > –1 and < 1). (B) Balloon plot of Hallmark GSEA results for genes in IML up or IML down. Circle size is the number of overlapping genes and circle color is the FDR q-value. (C) Individual gene β scores that are depleted in IML treated cells from the HALLMARK_OXIDATIVE_PHOSPHORYLATION and HALLMARK_REACTIVE_OXYGEN_SPECIES pathways. (D) Fold change for MCF7 (left) or T47D (right) cells during treatment up to 8 days with DMSO, IACS (10 nM), IML (100 nM), or IACS and IML. Growth curve with average ± SD. One-way ANOVA with Tukey’s multiple comparisons test. (E) Normalized cell proliferation for MCF7 (circle) or T47D (square) cells expressing ER-Y537S and treated with a dose-response of IACS for 5 days. Growth curves with average ± SD. Two-way ANOVA with Šidák’s multiple comparisons test. (F) Synergy distribution analysis in MCF7 ER-Y537S cells treated with a dose response matrix of IACS and IML. Loewe synergy score using the synergyfinder tool. Red indicates synergism. (G) Normalized cell proliferation for MCF7 parental (black) or MCF7 LTE-IML (orange) cells treated with IML (left) or IACS (right). Growth curve with average ± SD. Two-way ANOVA with Šidák’s multiple comparisons test. (H) Colony assay crystal violet staining results from MCF7 parental or LTE-IML cells treated with imlunestrant (100 nM), IACS (10 nM), IML and IACS for 2 weeks in full media. (I) Relative confluency of colony assay crystal violet staining in H. Bar graph with average ± SD. Two-way ANOVA with Dunnett’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.