Cyclin E Overexpression Sensitizes Triple-Negative Breast Cancer to Wee1 Kinase Inhibition

Abstract

Purpose: Poor prognosis in triple-negative breast cancer (TNBC) is due to an aggressive phenotype and lack of biomarker-driven targeted therapies. Overexpression of cyclin E and phosphorylated-CDK2 are correlated with poor survival in patients with TNBC, and the absence of CDK2 desensitizes cells to inhibition of Wee1 kinase, a key cell-cycle regulator. We hypothesize that cyclin E expression can predict response to therapies, which include the Wee1 kinase inhibitor, AZD1775.

Experimental design: Mono- and combination therapies with AZD1775 were evaluated in TNBC cell lines and multiple patient-derived xenograft (PDX) models with different cyclin E expression profiles. The mechanism(s) of cyclin E-mediated replicative stress were investigated following cyclin E induction or CRISPR/Cas9 knockout by a number of assays in multiple cell lines.

Results: Cyclin E overexpression (i) is enriched in TNBCs with high recurrence rates, (ii) sensitizes TNBC cell lines and PDX models to AZD1775, (iii) leads to CDK2-dependent activation of DNA replication stress pathways, and (iv) increases Wee1 kinase activity. Moreover, treatment of cells with either CDK2 inhibitors or carboplatin leads to transient transcriptional induction of cyclin E (in cyclin E-low tumors) and result in DNA replicative stress. Such drug-mediated cyclin E induction in TNBC cells and PDX models sensitizes them to AZD1775 in a sequential treatment combination strategy.Conclusions: Cyclin E is a potential biomarker of response (i) for AZD1775 as monotherapy in cyclin E-high TNBC tumors and (ii) for sequential combination therapy with CDK2 inhibitor or carboplatin followed by AZD1775 in cyclin E-low TNBC tumors.

Figure 1.. Cyclin E overexpression sensitizes TNBC to Wee1 inhibition in vitro and in vivo.

(A) Percentage of breast tumors with CCNE1 amplification and/or RNA upregulation in TNBC, all tumors and ER positive (ER+) tumors from TCGA and METABRIC databases. Fisher exact test was used to compare two groups. (B) Kaplan-Meier survival plot of cyclin E alterations (mRNA and DNA) of the TNBC patient cohort from the TCGA database. (C) Cyclin E immunohistochemical analysis of TNBC tumor tissues from two representative patient samples with low (top) or high (bottom) cyclin E. (Scale: 600 μm). (D) Kaplan-Meier survival plot of cyclin E immunohistochemical results (high/low) from 248 TNBC samples from two different patient cohorts. (E) Six TNBC cell lines were immunoblotted (bottom) with the indicated antibodies and subjected to high-throughput survival assay (HTSA) (top) with 0.4 μM AZD1775 for 48 hours. (F) HCC1806-shRNA cyclin E cells were subjected to HTSA (left) with increasing concentrations of AZD1775 for 48 hours as shown in fig. S2C and were immunoblotted (right) with the indicated antibodies. (G) Cyclin E expression in the inducible 76NE6 cells and MDA231 cells treated with (DOX) or without (control, or CNL) 10 ng/mL doxycycline for 10 days. After cyclin E was induced for 24 hours, cells were treated with AZD1775 for 48 hours and subjected to MTT on day 11, as shown in fig. S2E. (H) Immunoblot of three different tumors from TNBC PDX models with the indicated antibodies. (I-K) Mice harboring (I) PDX1, (J) PDX2, or (K) PDX3 models were treated with either vehicle (V) or 50 mg/kg AZD1775 (AZD) for 4 cycles as shown in fig. S4A, presented as (top) relative tumor volume and (bottom) ethical endpoint survival (n=4–6). One-way analysis of variance (ANOVA) was used for multiple-group comparison. Log-rank test was used in survival experiments.

Figure 2.. Wee1 kinase protects the cyclin E-overexpressed cells from DNA replication stress and DNA damage.

(A) Representative ongoing replication forks of DNA tracts pulse-labeled with iododeoxyuridine (IdU) and chlorodeoxyuridine (CIdU) from control cells (CNL), cells treated with 48 hours of 1 μM AZD1775 (AZD), cells with 48-hour induction of cyclin E by 10 ng/mL doxycycline (cyclin E), and cells with both cyclin E induction and AZD1775 treatment for 48 hours (cyclin E + AZD) in MDA231 cyclin E-inducible cells. Scale bar, 10 μm. (B) Analysis of replication fork progression of the ongoing replication forks in the samples as shown in (A). More than 190 DNA fibers were analyzed for each sample. (C) Representative images of RPA, γH2AX, RAD51, and 53BP1 foci in control cells (CNL), cells treated with 48 hours of 1 μM AZD1775 (AZD), cells with 48-hour induction of cyclin E by 10 ng/mL doxycycline (cyclin E), and cells with both cyclin E induction and AZD1775 treatment for 48 hours (cyclin E + AZD) in MDA231 cyclin E-inducible cells. Scale bar, 10 μm. (D) Quantification of foci-positive cell (>5 foci per cell) population in the samples as in (C), n=3. (E) Immunoblot of indicated antibodies of MDA231 cyclin E–inducible cells with (+) or without (−) cyclin E induction by 10 ng/mL doxycycline and with different concentrations (0, 0.5, and 1 μM) of AZD1775 for 48 hours. L, long exposure; S, short exposure. Vinculin is the loading control. (F) The representative histogram images (upper) and the cell-cycle analysis with flow cytometry (lower) in control cells (CNL), cells treated with 48 hours of 1 μM AZD1775 (AZD), cells with 48-hour induction of cyclin E by 10 ng/mL doxycycline (cyclin E), cells with both cyclin E induction and AZD1775 treatment for 48 hours (cyclin E + AZD) in MDA231 cyclin E-inducible cells. Green in histogram image indicates polyploidy. (G) Representative images (left) of normal (CNL) and abnormal (cyclin E + AZD) nuclei and quantification of the frequency of abnormal nuclei in control cells (CNL), cells treated with 48 hours of 1 μM AZD1775 (AZD), cells with 48-hour induction of cyclin E by 10 ng/mL doxycycline (cyclin E), and cells with both cyclin E induction and AZD1775 treatment for 48 hours (cyclin E + AZD) in MDA231 cyclin E-inducible cells. Scale bar, 10 μm. A two-tailed unpaired t-test was used to compare two groups. Error bars represent standard error of the mean. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

Figure 3.. Cyclin E overexpression depends on CDK2 to sensitize cells to Wee1 kinase inhibition.

(A) Kaplan-Meier survival plots of cyclin E^highpCDK2^high and cyclin E^lowpCDK2^low tumors from the TNBC patient cohorts from fig. 1D. (B) Percentages of γH2AX- and RAD51-positive cells of 76NE6 cyclin E-inducible cells cultured with (+) or without (−)10 ng/mL doxycycline (DOX) for 24 hours by IHC staining. (C) 76NE6 cyclin E^R130A inducible cells cultured with (DOX) or without (CNL) 10 ng/mL doxycycline for 24 hours were treated with AZD1775 for 48 hours and subjected to HTSA as in fig. S2C. Inset shows immunoblot of cyclin E expression. (D) Representative images of RPA, γH2AX, RAD51, and 53BP1 foci in MDA231 cyclin E-inducible cells as in (A). Scale bar, 10 μm. (F) Quantification of foci-positive cell (>5 foci per cell) population in the samples as in (D), n=3. (F) Representative ongoing replication forks of DNA tracts pulse-labeled with IdU and CIdU from the parental cells and CDK2 knockout (KO) cells (via CRISPR/CAS9) of MDA231 with or without cyclin E induction for 48 hours by 10 ng/mL doxycycline (DOX). Scale bar, 10 μm. The data in the blue box are presented in fig 2A. (G) Analysis of replication fork progression of the ongoing replication forks in the samples as in (F). More than 89 DNA fibers were analyzed for each sample. The data in the blue box were presented in fig 2B. (H) 76NE6 cyclin E–inducible cells with scrambled shRNA (Scr.) or cyclin E shRNA knockdown (shCyclin E) (left) with (+) or without (−) 10 ng/mL doxycycline (DOX) for 24 hours (left) and MDA231 cyclin E-inducible cells as in (F) (right) were subjected to immunoblot with the indicated antibodies. Actin is the loading control. (I) 76NE6 cyclin E–inducible cells with CDK2 or scramble shRNA were subjected to HTSA as in fig. S2E. (J) Parental and CDK2 knockout cells of MDA231 cyclin E–inducible cells were subjected to HTSA as in fig. S10F. (K) MDA231 cyclin E–inducible cells were concomitantly treated with 6 nM dinaciclib and different concentrations of AZD1775 with or without 10 ng/mL doxycycline for 48 hours as in fig. S10H. The bar graph shows the changes in cell survival between the cells without and with 6nM dinaciclib at absence (CNL) or presence (DOX) of 10 ng/mL doxycycline at the different AZD1775 doses.

Figure 4.. Sequential combination treatment with CDK2 inhibitor followed by AZD1775 in cyclin E–low TNBC is synergistic in vitro and in vivo.

(A) Combination index (CI) of sequential (D→AZD) and concomitant (D+AZD) combination of dinaciclib and AZD1775 in MDA231 cyclin E–inducible cells with (DOX) or without (CNL) 10 ng/mL doxycycline. CI=1.0 indicates an additive effect (red line). (B) Mean CI values of sequential (D→AZD) and concomitant (D+AZD) treatment using dinaciclib and AZD1775 in TNBC cells by HTSA. CI=1.0 indicates an additive effect (red line). The concentrations of inhibitors of this assay are listed in table S5. (C) The clonogenic assay of TNBC cells with dinaciclib (D) and/or AZD1775 (AZD), as single agents or in combination sequentially (D→AZD). (D) Cells were treated with vehicle, dinaciclib (24 hours, D), AZD1775 (48 hours, AZD), or sequential dinaciclib (24 hours) followed by AZD1775 (48 hours) (D→AZD) at concentrations listed in table S6. Cells subjected annexin V assay (top), quantification of the sub-G1 population (middle), or immunoblot with PARP antibody (bottom). Differences between combination treatment and other groups were analyzed statistically for significance. (E) Mean CI values of sequential (CDK2i→AZD) and concomitant (CDK2i+AZD) treatment using CDK2 inhibitor (meriolin5, SNS032, or roscovitine) and AZD1775 in MDA231 and SUM159 by HTSA as in (B). CI=1.0 indicates an additive effect (red line). (F, G, H) Tumor volumes (top) and mouse survival (bottom) of mice harboring xenograft tumors derived from (F) SUM149 cells or patient tumors, (G) PDX1, or (H) PDX2 were treated with 25 mg/kg dinaciclib once per day for 2 days and 50 mg/kg AZD1775 twice per day for 2 days of each 7-day cycle. (Schematic in fig. S12A.) V, mouse treated by vehicle only; D, mouse treated by dinaciclib and vehicle of AZD1775; AZD, mouse treated by AZD1775 and vehicle of dinaciclib; D+AZD, mouse concomitantly treated by dinaciclib and AZD1775; D→AZD, mouse sequentially treated by dinaciclib and AZD1775; N.S., not significant. SUM149 xenografts were treated for 6 cycles; PDX1 and PDX2 were treated for 4 cycles, with the exception of the D+AZD arm and the D→AZD arm in PDX1, which were treated for 6 cycles and 10 cycles respectively. A two-tailed unpaired t-test was use to compare two groups (n=4–20). Error bars represent standard error of the mean. The log-rank (Mantel-Cox) test was used in survival experiments. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

Figure 5.. Transactivation of cyclin E by dinaciclib sensitizes TNBC cells to AZD1775.

(A) TNBC cells were treated with dinaciclib (D) or vehicle (CNL) for 24 hours or with dinaciclib (24 hours) followed by 24 hours in drug-free medium (D+24) at IC₅₀ concentrations (table S6) and subjected to Western blot with the indicated antibodies. (B) Relative cyclin E mRNA levels of TNBC cell lines were quantified by qRT-PCR using the level in MDA231 cells as a baseline (equivalent to 1). (C) Six TNBC cell lines were treated with dinaciclib (D) or vehicle (CNL) for 24 hours at IC₅₀ concentrations (table S6) and subjected to qRT-PCR of cyclin E. (D) Chromatin immunoprecipitation–qPCR of the enhancer domain (binding region) and downstream negative region of CCNE1 by H3K27AC antibody in cell lines treated with dinaciclib as in (C). (E) Representative images (top) and quantification of foci-positive cell (>5 foci per cell, bottom) population of RPA and RAD51 foci in MDA231 cells treated as in (A). RPA+RAD51 indicates cells with both RPA and RAD51 foci. Scale bar, 10 μm. n=3. (F) Parental cells and two cyclin E CRISPR/CAS9 knockout (KO) clones (#3 and #10) of MDA231 were subjected to sequential combination treatment (D→AZD) as in fig. 5B. Mean CI values are shown. CI=1.0 indicates an additive effect (red line). Inserted panels indicate the protein level of cyclin E of parental and KO cells by immunoblot. (G) The CI of MDA231 (left) and SUM149 (right) cells with single and/or double knockdown of CDK2 and/or cyclin E with the D→AZD treatment as in fig. 5B. (H) MDA231 and SUM159 cells were treated with roscovitine, SNS032, or meriolin5 for 24 hours and subjected to immunoblot with indicated antibodies. The concentrations of each drug used for each cell line are listed in table S6. Vinculin is a loading control. (I) Parental cells and two cyclin E CRISPR/CAS9 KO clones (#3 and #10) of MDA231 were subjected to sequential combination treatment as in fig. 5B. Mean CI values are shown. CI=1.0 indicates an additive effect (red line). Ros→AZD, the sequential combination of roscovitine followed by AZD1175; M5→AZD, the sequential combination of meriolin5 followed by AZD1775; SNS→AZD, the sequential combination of SNS032 followed by AZD1775. A two-tailed unpaired t-test was use to compare two groups. Error bars represent standard error of the mean. *, p<0.05; **, p<0.01; ***, p<0.001.

Figure 6.. Sequential combination treatment with carboplatin followed by AZD1775 is synergistic in vitro and in vivo.

(A) MDA231 and SUM159 cells were treated with palbociclib for 72 hours and other agents as indicated for 24 hours and subjected to immunoblot. Dimethyl sulfoxide (DMSO) was the vehicle control for palbociclib. (The concentrations of each drug used for each cell line are listed in table S6.) (B) The mean CI for each agent followed sequentially by AZD1775 for 48 hours in MDA231 and SUM159 cells was determined by CalcuSyn. Palbociclib was administered for 72 hours, and other agents were administered for 24 hours. The red line represents CI=1. (The concentrations of each drug used for each cell line are listed in table S7.) (C) MDA231 cells were treated with carboplatin (carbo) or vehicle (CNL) for 24 hours, or with carboplatin (24 hours) followed 24 hours in drug-free medium (carbo+24) at the same concentrations as in (A) and subjected to immunofluorescence staining with RPA and RAD51 antibodies. Representative images (top) and quantification of foci-positive cell (>5 foci per cell, bottom) population of RPA and RAD51 foci. RPA+RAD51 indicates cells with both RPA and RAD51 foci. Scale bar, 10 μm. n=3. (D) Cell cycle analysis of MDA231 cells treated as in (C) by FACS, n=3. (E) Mice harboring MDA231T xenograft tumors were treated with 0, 15, or 30 mg/kg carboplatin at 1 dose/week for 3 doses. Tumors were harvested at 24 hours after the last dose of carboplatin. Immunoblot of three different tumors for each dosage with the indicated antibodies. (F, G) Tumor volumes (F) and mouse survival (G) of mice harboring MDA231T xenograft tumors were treated with 30 mg/kg carboplatin once per day for 1 day and 50 mg/kg AZD1775 twice per day for 2 days of each 7-day cycle and repeated for 3–7 cycles. (Schematic in fig. S15D.) V, mouse treated only with vehicle; carbo, mouse treated with carboplatin and vehicle of AZD1775; AZD, mouse treated with AZD1775 and vehicle of dinaciclib; carbo+AZD, mouse concomitantly treated with carboplatin and AZD1775; carbo→AZD, mouse sequentially treated by carboplatin and AZD1775, n=5. (H) Working model. A two-tailed unpaired t-test was use to compare two groups. Error bars represent standard error of the mean. The log-rank (Mantel-Cox) test was used in survival experiments. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.