The impact of oral peroxisome proliferator-activated receptor gamma agonist (PPARG) efatutazone exposure through diet on mammary preneoplasia and cancer development was evaluated in genetically engineered mice carrying one vs two disrupted Brca1 alleles targeted to mammary epithelial cells in the background of germ-line Trp53 (p53) haploinsufficiency. Brca1WT/fl11/Cre/p53+/− and Brca1fl11/fl11/Cre/p53+/− female mice were randomly assigned to receive efatutazone-containing diet (30 mg/kg) or equivalent diet without efatutazone at age 2 months. Efatutazone exposure significantly increased the percentage of mice with tertiary branching (Fig. 1A). Palpable mammary cancers appeared significantly earlier in Brca1fl11/fl11/Cre/p53+/− as compared to Brca1WT/fl11/Cre/p53+/− mice, independent of efatutazone exposure (Fig. 1B). Efatutazone significantly reduced mammosphere development in Brca1fl11/fl11/Cre/p53+/− mice (Fig. 1C, D and E). The significant reduction in numbers and prevalence of hyperplastic alveolar nodules (HANs) was more profound in Brca1WT/fl11/Cre/p53+/− as compared to Brca1fl11/fl11/Cre/p53+/− mice (Fig. 1F and G). Palpable subcutaneous lipomas appeared in both efatutazone cohorts beginning at age 7 months, significantly shifting the reason for killing in the two different models (Fig. 2A). Fewer Brca1WT/fl11/Cre/p53+/− mice were killed for palpable mammary gland cancer than Brca1fl11/fl11/Cre/p53+/− mice. Most mice developed one mammary cancer with a range of histology. Four mice developed two palpable mammary cancers: three Brca1fl11/fl11/Cre/p53+/− mice (two control and one efatutazone cohort) and one Brca1WT/fl11/Cre/p53+/− mouse (control cohort). Most mammary cancers were triple negative (Fig. 2B, C, D, E, F and G). The two estrogen receptor alpha (ER)- and progesterone receptor (PR)-positive mammary adenocarcinomas that appeared were unique to the Brca1fl11/fl11/Cre/p53+/− efatutazone cohort (Fig. 2H and I).
The study was prompted by a previous investigation demonstrating that efatutazone exposure beginning at age 4 months in Brca1fl11/fl11/Cre/p53+/− mice, a model which typically generates a range of histology of triple-negative mammary cancers, significantly reduces mammary hyperplasia and induces more differentiated mammary cancers including some ER-positive adenocarcinomas (Nakles et al. 2013). Questions raised by the earlier study included whether or not mice carrying a disruption in only one Brca1 allele (Alothman et al. 2017) might be more responsive to efatutazone and the possibility that earlier administration would be more effective in reducing mammary cancer development (Dong 2013). PPARG has been reported as detectable in over 90% of triple-negative breast cancers developing in women who carry a BRCA1 mutation (Heublein et al. 2017).
Significant differences in efatutazone response were found between the two models. Both models are derived from the same breeding colony with one vs two disrupted Brca1 alleles discriminating between them. The proportion of control mice that developed mammary cancer by age 12 months was not significantly different between the two genotypes; however, the models did exhibit statistically significantly different time-courses of mammary carcinogenesis. Efatutazone was delivered at the same age in both models, raising the possibility that the significant differences in response found between the two models was due to the timing of the intervention relative to the course of mammary carcinogenesis, rather than the underlying genetic difference in full-length Brca1 expression. Even the significant difference in mammosphere development between the models could be related to timing as the published literature suggests that mammosphere development from cancer cells is more profoundly affected by a PPARg agonist as compared to non-cancer cells. Results could argue that for efatutazone to be maximally effective, it may need to be initiated at very early stages of carcinogenesis. Theoretically to test that hypothesis efatutazone could be started even before the end of puberty; however, here we identified a toxicity, lipoma development, prominent in this study but not found when efatutazone exposure was started at age 4 months (Nakles et al. 2013). PPARg activation previously has been suggested as a component of lipoma pathophysiology (Koppen & Kalkhoven 2010). The study indicates that developmental age at exposure may also play a role. Finally, we cannot exclude the possibility that germ-line Trp53 haploinsufficiency also contributed to the susceptibility of the mice for lipoma development.
For preneoplasia reduction, Brca1WT/fl11/Cre/p53+/− mice were more sensitive than Brca1fl11/fl11/Cre/p53+/− mice, but this did not translate to absolute mammary cancer prevention. The possibility of a parallel significant reduction in mammary cancer appearance could not be fully assessed due to the unexpected appearance of the palpable lipomas, found uniquely in this study when efatutazone was initiated at the younger age. This necessitated killing for tumor burden, required by the Institutional Animal Use and Care approved protocol, prior to the scheduled endpoint for evaluation of mammary cancer prevalence (age 12 months) in a significant proportion of the Brca1WT/fl11/Cre/p53+/− mice. In contrast, mammary tumor and lipoma development appeared temporally coincident in the Brca1fl11/fl11/Cre/p53+/− mice that exhibit earlier mammary cancer development. The appearance of lipomas in Brca1fl11/fl11/Cre/p53+/− mice did not compromise assessment of mammary cancer prevalence, which was unchanged by efatutazone exposure.
An alternative or contributing hypothesis for the significant differences found would be a differential action of efatutazone based on the presence or absence of full-length Brca1. Brca1WT/fl11/Cre/p53+/− mice still retain one functional Brca1 allele, and mammary cancer development occurs without loss of the second Brca1 allele (Alothman et al. 2017), as occurs in a portion of women carrying BRCA1 mutation (Roy et al. 2011). PPARG agonists are reported to increase BRCA1 expression in vitro in human breast cancer cells and in vivo in mammary glands of FVB mice (Subbaramaiah et al. 2012). Perhaps significantly, efatutazone is reported to be effective in delaying progression and inducing differentiation in a xenograft model of triple-negative ductal carcinoma in situ (MCFDCIS) with intact BRCA1 and p53 alleles (Ory et al. 2018).
Replicated in this investigation was the novel appearance of ER+/PR+ mammary adenocarcinomas in the efatutazone-exposed Brca1fl11/fl11/Cre/p53+/− mice. While a low-incidence event in both this and the prior study (Nakles et al. 2013), it did appear in two independent and temporally distinct experiments. Results support the possibility that activation of the PPARG pathway can contribute to the appearance of more differentiated mammary cancers retaining expression of ER and PR. None of the cancers expressed pathophysiological levels of Erb-B2 receptor tyrosine kinase 2 (ErbB2/Her2) and the majority of cancers in all groups were triple negative, consistent with previous studies in these and related models (Nakles et al. 2013).
Recognizing the fact that efatutazone was not protective of mammary cancer development, even as it significantly reduced preneoplastic growth, prompts the question of whether or not combining the drug with a second agent such as an antihormonal, a retinoid X receptor RXR ligand, TNF superfamily member 11 (TNFSF11)/RANKL inhibition or agents affecting the cell cycle such as cyclin-dependent kinase (CDK)4/6 inhibitors would be more effective. Time-course studies in which interventions are initiated at different ages in models with different temporal patterns of carcinogenesis on a prospective basis may help to further define the points during carcinogenesis when specific preventive agents can be effective. Direct comparison of the mechanistic action of efatutazone in in vitro and in vivo models with different gene dosages of Brca1 and Trp53 intact could establish if there are specific roles for their respective proteins in moderating response to a PPARG agonist.
All animal procedures were performed in accordance with federal guidelines and were approved by the Georgetown University Institutional Animal Care and Use Committee. Generation of mouse cohorts, dietary interventions (F3028, rodent diet, grain-based 1/2-in pellets without or with efatutazone (30 mg/kg), Bio-Serv, Frenchtown, NJ, USA) and histology were performed as previously described (Nakles et al. 2013, Alothman et al. 2017). Pathology was read blindly by a board-certified pathologist (B V K). NanoCulture Plates (ORGANOGENIX Inc., Kanagawa, JP) were used to assess 3D mammosphere formation in serum-free EpiCult-B Mouse media (STEMCELL Technologies, Vancouver, CA, USA) (Arai et al. 2016, Alothman 2018). Immunohistochemical (IHC) procedures: ER (primary antibodies: ab75635, 1:50, overnight 4°C, Abcam; 06-935, 1:100, 1 h, room temperature (RT), Millipore Sigma), PR (sc538, 1:350, 1 h, RT, Santa Cruz Biotechnology, Inc.), HER2 (ERBB2) (2165, 1:100, overnight 4°C, Cell Signaling). Before primary antibody exposure, heat-induced epitope retrieval was performed by immersing tissue sections at 98°C for 20 min in 10 mM citrate buffer (pH 6.0) with 0.05% Tween (ER, PR) or 98oC for 20 min in 0.5 M EDTA buffer (pH 8.0) (HER2) followed by treatment with 3% hydrogen peroxide and 10% normal goat serum (10 min each). Following primary antibody exposure, slides were exposed to an HRP-labeled polymer (EnVision+ System- HRP Labeled Polymer Anti-Rabbit, K4003, Dako, 30 min) and 3′-Diaminobenzidine (DAB) (Dako, 5 min). Cancers were read as ER+/PR+ when nuclear localization of both proteins was present. Two independent primary antibodies were used to test ER nuclear localization. Positive (human ER+/PR+/HER2-amplified breast cancer) and negative (no primary antibody) controls were performed with each individual IHC experiment. Statistical analyses were performed using GraphPad Prism and the VassarStats: Statistical Computation Web Site. P < 0.05 was considered significant.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this article.
This work was supported by National Institutes of Health, National Cancer Institute RO1 CA112176 (P A F) and 5P30CA051008 (Histopathology and Tissue and Animal Model Shared Resources) and the King Abdullah Scholarship Program, Ministry of Education, Kingdom of Saudi Arabia.
Author contribution statement
Study conception and design (S J A, P A F), mouse husbandry (P A F, W W), data procurement (S J A, W W, S C, B V K, E S D-C, P A F), data analysis (S J A, S C, P A F), writing manuscript (S J A, P A F), review and editing manuscript (S J A, W W, S C, B V K, E S D-C, P A F).
AlothmanSJWangWGoerlitzDSIslamMZhongXKishoreAAzharRIKallakuryBVFurthPA 2017 Responsiveness of Brca1 and Trp53 deficiency-induced mammary preneoplasia to selective estrogen modulators versus an aromatase inhibitor in Mus musculus. Cancer Prevention Research 10 244–254. (https://doi.org/10.1158/1940-6207.CAPR-16-0268)
AraiKEguchiTRahmanMMSakamotoRMasudaNNakatsuraTCalderwoodSKKozakiK-IItohM 2016 A novel high-throughput 3D screening system for EMT inhibitors: a pilot screening discovered the EMT inhibitory activity of CDK2 inhibitor SU9516. PLoS ONE 11 e0162394. (https://doi.org/10.1371/journal.pone.0162394)
HeubleinSMayrDMeindlAKircherAJeschkeUDitschN 2017 Vitamin D receptor, retinoid X receptor and peroxisome proliferator-activated receptor γ are overexpressed in BRCA1 mutated breast cancer and predict prognosis. Journal of Experimental and Clinical Cancer Research 36 57. (https://doi.org/10.1186/s13046-017-0517-1)
NaklesREKallakuryBVFurthPA 2013 The PPARgamma agonist efatutazone increases the spectrum of well-differentiated mammary cancer subtypes initiated by loss of full-length BRCA1 in association with TP53 haploinsufficiency. American Journal of Pathology 182 1976–1985. (https://doi.org/10.1016/j.ajpath.2013.02.006)
OryVKeitzmanWBoekelmanJKallakuryBWellsteinAFurthPARiegelAT 2018 The PPARγ agonist efatutazone delays invasive progression and induces differentiation of ductal carcinoma in situ (DCIS). Breast Cancer Research and Treatment 169 47–57. (https://doi.org/10.1007/s10549-017-4649-y)
SubbaramaiahKHoweLRZhouXKYangPHudisCALevy KopelovichLDannenbergAJ 2012 Pioglitazone, a PPARg agonist, suppresses CYP19 transcription: evidence for involvement of 15-hydroxyprostaglandin dehydrogenase and BRCA1. Cancer Prevention Research 5 1183–1194. (https://doi.org/10.1158/1940-6207.CAPR-12-0201)