Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Jul 7;112(27):8409-14.
doi: 10.1073/pnas.1500223112. Epub 2015 Jun 22.

Selective Resistance to the PARP Inhibitor Olaparib in a Mouse Model for BRCA1-deficient Metaplastic Breast Cancer

Affiliations
Free PMC article

Selective Resistance to the PARP Inhibitor Olaparib in a Mouse Model for BRCA1-deficient Metaplastic Breast Cancer

Linda Henneman et al. Proc Natl Acad Sci U S A. .
Free PMC article

Abstract

Metaplastic breast carcinoma (MBC) is a rare histological breast cancer subtype characterized by mesenchymal elements and poor clinical outcome. A large fraction of MBCs harbor defects in breast cancer 1 (BRCA1). As BRCA1 deficiency sensitizes tumors to DNA cross-linking agents and poly(ADP-ribose) polymerase (PARP) inhibitors, we sought to investigate the response of BRCA1-deficient MBCs to the PARP inhibitor olaparib. To this end, we established a genetically engineered mouse model (GEMM) for BRCA1-deficient MBC by introducing the MET proto-oncogene into a BRCA1-associated breast cancer model, using our novel female GEMM ES cell (ESC) pipeline. In contrast to carcinomas, BRCA1-deficient mouse carcinosarcomas resembling MBC show intrinsic resistance to olaparib caused by increased P-glycoprotein (Pgp) drug efflux transporter expression. Indeed, resistance could be circumvented by using another PARP inhibitor, AZD2461, which is a poor Pgp substrate. These preclinical findings suggest that patients with BRCA1-associated MBC may show poor response to olaparib and illustrate the value of GEMM-ESC models of human cancer for evaluation of novel therapeutics.

Keywords: BRCA1; PARP; breast cancer; mouse model; resistance.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.
Fig. 1.
Mammary tumor development in chimeric mice derived from KB1P ESCs. (A) ESCs were isolated from our well-validated KB1P mouse model and subsequently tested for performance to generate chimeric mice. (B) Scatter plot showing the performance to generate chimeras of three individual ESC clones (one male, two female). (C) Tumor-free survival of KB1P chimeric mice (red, clone A2, t50 = 266 d, n = 13 mice; and blue, clone C7, t50 = 237 d, n = 6 mice) and KB1P GEMM mice (black; t50 = 209 d, n = 48 mice). Log-rank (Mantel–Cox) P value is indicated. (D) Histopathology of mammary tumors derived from KB1P chimeric and KB1P GEMM female mice. (Upper) H&E staining and (Middle, Lower) immunohistochemical detection of estrogen receptor (ER) and progesterone receptor (PR). (E) Pie diagrams illustrate the distribution of mammary tumor types in KB1P GEMM mice (93% carcinoma; 7% carcinosarcoma) and chimeras from KB1P clone A2 (93% carcinoma; 7% carcinosarcoma) and KB1P clone C7 (100% carcinoma). Carcinomas are shown in blue and carcinosarcomas in orange. (F) CNV analysis of KB1P chimeric and KB1P GEMM mammary tumors. Chromosome numbers are represented on the x axis and log2 ratio is depicted on the y axis.
Fig. S1.
Fig. S1.
XO conversion in female ESCs. (A) DNA CNV analysis of two female ESC clones derived from the KB1P mouse model. Chromosome numbers are represented on the x axis and log2 ratios are depicted on the y axis. (B) FISH analysis of three individual female ESC clones using probes to detect X chromosome number. The numbers in the table represent individual cells for each clone, assessed for X chromosome number. A minimum of 100 cells were assessed per clone. Confocal images of metaphase spreads from two female ESC clones stained with a probe for the X-chromosome and Y-chromosome.
Fig. S2.
Fig. S2.
Southern blot analysis of genomic DNA from KB1P chimeric tumors to determine Brca1 (EcoRV/Stul digestion, exon 14 probe) and Trp53 (Bglll digestion, 5’ Xbal probe) deletion. Fragment sizes: Brca1 WT, 7 kb; Brca1 del, 6 kb; Trp53 WT, 18 kb; Trp53 del, 9.4 kb. Spleen is used as control tissue.
Fig. 2.
Fig. 2.
MET expression in KB1P chimeric mice leads to accelerated mammary tumor development. (A) Overview of the experimental procedure. Female ESCs derived from our KB1P mouse model were equipped with a Col1a1-frt homing cassette, and subsequently the Met proto-oncogene was introduced by recombinase-mediated cassette exchange. (B) Scatter plot illustrating the performance of KB1P ESC clones targeted with the Col1a1-frt homing cassette (Col1a1), and additionally equipped with Met (MET clone A9 and B3) to produce chimeric mice. (C) Mammary tumor-free survival of KB1P chimera (red; t50 = 284 d, n = 13 mice), KB1P-Col1a1 (black; t50 = 299 d, n = 14 mice), and KB1P-MET (blue; t50 = 176 d, n = 11 mice) chimeric mice. Log-rank (Mantel–Cox) P values are indicated. (D) Immunohistochemical staining of formalin-fixed/paraffin-embedded sections from KB1P, KB1P-Col1a1, and KB1P-MET chimeric mammary tumors for H&E, MET, METY1234/1235, E-cadherin, and vimentin. (E) Pie diagrams illustrating the distribution of different mammary tumor types in KB1P (93% carcinoma; 7% carcinosarcoma), KB1P-Col1a1 (100% carcinoma), and KB1P-MET (45% carcinoma; 55% carcinosarcoma) chimeric mice. Carcinomas are shown in blue and carcinosarcomas in orange. (F) Box plot showing relative expression of claudin 3, 4, and 7 in KB1P-MET carcinomas and carcinosarcomas. Mann–Whitney P values are indicated.
Fig. S3.
Fig. S3.
Targeting strategy. (A) Targeting of Col1a1-frt in the Col1a1 locus was performed by homolog recombination in KB1P ESCs. Schematic representation of the Col1a1 locus and the Col1a1-frt targeting construct. The Col1a1-frt vector contains a neomycin resistance cassette flanked by frt sites, followed by an ATG-less hygromycin resistance gene. The 3′ internal probe is indicated. Exons are shown as black boxes and the 3′UTR as a gray box. B, BglII; E, EcoRI; P, PstI; pA, polyadenylation signal; S, SpeI; X, XhoI; Xb, XbaI. (B) Flip-in of the frt-invCag-Met-Luc vector in Col1a1-frt targeted KB1P ESC clones was performed. Expression of Flpe recombinase removes the pGK-neo-pA cassette from the targeted Col1a1 locus and allows for Flpe-mediated integration of the frt-invGag-Met-Luc vector. ESC clones are selected for resistance to hygromycin that is expressed from the CAG promoter present in the frt-invGag-Met-Luc vector. (C) Cre recombinase-mediated expression of MET and firefly luciferase. The frt-invCag-Met-Luc vector contains Lox66 and Lox71 recombination sites that allow for promoter inversion after Cre recombinase expression resulting in MET and firefly luciferase expression. (D) Southern blot analysis of genomic DNA from KB1P-Col1a1 targeted ESC clones to determine incorporation of the Col1a1-frt homing cassette in the Col1a1 locus using the 3′ internal probe (EcoRI digest). Fragment sizes: Col1a1 WT, 4.6 kb; Col1a1-frt, 3.8 kb. (E) Southern blot analysis of genomic DNA from KB1P-MET ESC clones to determine successful Flp-in of MET using the 3′ internal probe (BglII digest). Fragment sizes: Col1a1 WT, 0.9 kb; Col1a1-frt, 3.3 kb; frt-invCag-Met-Luc, 6.8 kb. (F) Measurement of bioluminescence of two KB1P-MET ESC clones left untreated (Left) or infected with AdCre (Right).
Fig. S4.
Fig. S4.
Analysis of KB1P chimeric mammary tumors. (A) In vivo bioluminescence imaging of a KB1P-MET female chimera at day 82 and day 110. (B) Southern blot analysis of genomic DNA from KB1P-Col1a1 and KB1P-MET chimeric tumors to determine Brca1 (EcoRV/StuI digestion, exon 14 probe) and Trp53 (BglII digestion, 5′ XbaI probe) deletion. Fragment sizes: Brca1 WT, 7 kb; Brca1 del, 6 kb; Trp53 WT, 18 kb; Trp53 del, 9.4 kb. Spleen is used as control tissue.
Fig. S5.
Fig. S5.
RNA-seq analysis of KB1P-MET tumors. Heat map of RNA-sequence analysis of KB1P and KB1P-MET chimeric tumors using the EMT signature.
Fig. 3.
Fig. 3.
KB1P-MET carcinosarcomas respond poorly to clinical PARP inhibitor olaparib. (A) Small fragments of KB1P-Col1a1 and KB1P-MET epithelial and mesenchymal tumors were transplanted in the fourth mammary fat pad of WT recipient syngeneic mice. When tumors had reached a size of 200 mm3, mice were treated. Tumor-bearing KB1P-Col1a1 and KB1P-MET mice were treated with 6 mg/kg cisplatin, administered i.v. on day 0 and day 14. Tumor growth was monitored three times per week to assess treatment efficacy. (B) Tumor-bearing KB1P-Col1a1 and KB1P-MET mice were treated with 50 mg/kg olaparib daily, administered by i.p. injection, for 28 consecutive days. Tumor growth was monitored three times per week to assess treatment efficacy.
Fig. 4.
Fig. 4.
KB1P-MET carcinosarcomas regress upon AZD2461 treatment. (A) Bar graph showing reverse transcriptase-multiplex ligation-dependent probe amplification (RT-MLPA) analysis on mammary tumors from KB1P and KB1P-MET chimeric mice. Abcb1a and Abcb1b mRNA levels are shown relative to the housekeeping gene Hprt. C, carcinoma; CS, carcinosarcoma. (B) Scatter plot showing the correlation between EMT status and Abcb1a or Abcb1b expression levels. Orange circle indicates KB1P chimera; blue circle indicates KB1P-MET carcinoma; blue triangle indicates KB1P-MET carcinosarcoma. (C) Small fragments derived from a KB1P-MET carcinoma and two carcinosarcomas were transplanted into the mammary fat pad of WT recipient syngeneic mice. When tumors had reached a size of 200 mm3, mice were treated with olaparib (50 mg/kg i.p. for 28 d) with or without the Pgp inhibitor tariquidar (2 mg/kg i.p.) or left untreated. Tumor growth was monitored three times per week to assess treatment efficacy. (D) Small fragments derived from a KB1P-Col1a1 carcinoma and KB1P-MET carcinosarcomas were transplanted into the mammary fat pad of WT recipient syngeneic mice. When tumors had reached a size of 200 mm3, mice were treated orally with the PARP inhibitor AZD2461 (100 mg/kg per os for 28 d) or left untreated. Tumor growth was monitored three times per week to assess treatment efficacy.
Fig. 5.
Fig. 5.
ABCB1 expression and EMT score in human basal and claudin-low TNBCs. Box plots showing EMT score (A) and ABCB1 expression levels (B) in basal-like and claudin-low TNBCs. Wilcoxon P values are indicated.
Fig. S6.
Fig. S6.
53BP1 expression and RAD51 foci formation in KB1P-MET carcinosarcomas. (A) Immunohistochemical detection of 53BP1 expression in FFPE sections from KB1P-Col1a1 and KB1P-MET tumors. (B) Analysis of RAD51 expression using confocal microscopy in KB1P-MET carcinosarcomas. Tumors were exposed to 15 Gy irradiation before analysis. IR, irradiated; KP, K14cre;Trp53F/F; NIR, nonirradiated. (C) Quantification of RAD51 foci formation. Box plot representing the percentage of cells with more than five foci. The Mann–Whitney P value is indicated for KB1P-MET donor 3.

Similar articles

See all similar articles

Cited by 34 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback