Abstract S2-04: RANK ligand as a target for breast cancer prevention in BRCA1 mutation carriers

Background: BRCA1 mutation carriers commonly undergo prophylactic mastectomy to reduce their risk of breast cancer. The precise role of chemoprevention with tamoxifen, which reduces the incidence of ER-positive breast cancer in the general population, is uncertain for BRCA1 mutation carriers, where uptake has been modest. The identification of an effective and acceptable prevention therapy therefore remains a 'holy grail' for the field. Precancerous BRCA1mut/+ tissue harbors an aberrant population of luminal progenitor cells and deregulated progesterone signaling has been implicated in BRCA1-associated oncogenesis. Since Receptor Activator of Nuclear Factor-kappa B ligand (RANKL) is a key paracrine effector of progesterone signaling, and RANKL and its receptor RANK contribute to mammary tumorigenesis, we investigated a role for this pathway in the preneoplastic phase of BRCA1 mutation carriers.

Methods: We explored a role for the RANK/RANKL pathway during the preneoplastic phase in freshly isolated (histologically normal) patient specimens from BRCA1 mutation carriers using several approaches. RANK and RANKL expression in breast cancer was evaluated in formalin fixed paraffin embedded (FFPE) archival sections by IHC from the kConFab and the Amgen Tissue Banks. All samples were obtained with patient consent and relevant IRB approval. A role for RANKL inhibition in attenuating tumor onset was studied using the MMTV-cre/Brca1fl/fl/p53+/– mouse model that recapitulates human basal-like breast cancer.

Results: We identified two subsets of luminal progenitors (RANK+ and RANK–) in histologically normal tissue of BRCA1 mutation carriers and found that RANK+ cells are highly proliferative, exhibit grossly aberrant DNA repair and bear a molecular signature similar to that of basal-like breast cancer. Moreover, high levels of RANK expression prevailed in established BRCA1-associated tumors. These data suggest that RANK+ and not RANK– progenitors are a key target population in these women. Notably, inhibition of RANKL signaling by denosumab in 3D breast organoids derived from pre-neoplastic BRCA1mut/+ tissue attenuated progesterone-induced proliferation. Furthermore, inhibition of RANKL with either the RANKL inhibitor OPG-Fc or a RANKL monoclonal antibody in a Brca1-deficient mouse model significantly curtailed mammary tumorigenesis, when compared to controls (p<0.001).

Conclusions: Together these findings identify a targetable pathway in a putative cell of origin population in BRCA1 mutation carriers and implicate RANKL blockade as a promising breast cancer prevention strategy.

Citation Format: Lindeman GJ, Nolan E, Vaillant F, Branstetter D, Pal B, Giner G, Whitehead L, Lok SW, Mann GB, kConFab Consortium, Rohrbach K, Huang L-Y, Soriano R, Smyth GK, Dougall WC, Visvader JE. RANK ligand as a target for breast cancer prevention in BRCA1 mutation carriers [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr S2-04.

Abstract P5-03-02: Expression of RANK and RANK ligand (RANKL) in breast carcinoma and distinct breast epithelial cells from BRCA1 mutation carriers

Background: Breast tumors in BRCA1 mutation carriers likely arise from luminal progenitor (LP) cells, previously shown to exhibit aberrant growth properties. Oophorectomy, and possibly tamoxifen, reduce breast cancer risk in BRCA1 carriers, potentially via inhibition of paracrine mediated signaling to stem/progenitor cells. RANKL is a major paracrine effector of progesterone's mitogenic action in mammary epithelium via its receptor RANK, and has a role in ovarian hormone-dependent activation of stem cells. Here we assessed RANK and RANKL expression in breast tumors and normal breast epithelial subtypes from women with mutations of BRCA1 (mBRCA1) or BRCA2 (mBRCA2).

Methods: RANK and RANKL expression in breast cancer or normal breast tissue samples with mBRCA1, mBRCA2 or wildtype (WT) BRCA1/2 were analyzed in formalin fixed paraffin embedded (FFPE) sections by IHC. kConFab and The Royal Melbourne Hospital Tissue Bank provided the samples used in this analysis; these samples were obtained with relevant IRB approval. RANK expression on normal breast epithelial cells was measured by flow cytometry. Antibodies against human RANK (N-1H8, N-2B10; Amgen) and RANKL (M366; Amgen) were used in both assays. Incidence of IHC staining was scored as a positive IHC signal of any intensity. The overall expression was generated using the H scoring method which is calculated as the staining intensity of the tumor (0-3) multiplied by the percentage of positively staining cells.

Results: Breast tumors from women with mBRCA1 had a higher incidence of RANK expression (68/162; 42%) compared with mBRCA2 (17/113; 15%) or WT (34/314; 11%) and higher overall H score (21.3) compared with mBRCA2 (8.0) or WT (3.4); RANKL expression did not vary greatly between groups: mBRCA1 (13/135; 10%), mBRCA2 (5/114; 4%), WT (23/212; 11%). In normal breast tissue, LP (Lin−EpCAM+CD49f+) and basal/stem cells (Lin−EpCAM−CD49fhi) expressed RANK on their surface. Similar expression patterns were seen in these epithelial subtypes from each BRCA1/2 genotype. Stromal cells (Lin−EpCAM−CD49f−) had minimal RANK expression.

Conclusions: RANK expression intensity and incidence scores are both enriched approximately 4-fold in breast tumors from BRCA1 carriers compared with other genotypes. Also, RANK is normally expressed in breast LP cells as well as the basal/stem cell containing population. Ongoing studies will assess functional regulation of LP or mammary stem cell activity by RANKL and determine if the RANKL/RANK signaling pathway affects the aberrant growth characteristics of these cells from BRCA1 mutation carriers.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-03-02.

P3-04-05: Expression Profiles of RANK and RANKL mRNA and Protein in the Mammary Gland of Female Cynomolgus Monkeys after Long-Term Treatment with Different Menopausal Hormone Therapies

Background: RANK and its ligand (RANKL), key factors for bone remodeling and metastasis, are crucial for the development of the mouse mammary gland during pregnancy. OPG also binds RANKL and inhibits RANKL action. Treatment of mice with progesterone increases RANKL expression in mammary epithelium. RANKL functions as a major paracrine effector of both progesterone-dependent mitogenic action in mammary tissue, and also in ovarian hormone-dependent expansion and regenerative potential of mammary stem cells. RANKL inhibition in vivo reduces progestin-induced mammary preneoplasias and adenocarcinomas. In postmenopausal women, the addition of a progestin to estrogen therapy has been associated with significant increases in mammary proliferation, breast density, and breast cancer risk and recurrence. The goal of this study was to characterize the expression of RANK, RANKL, and OPG in the mammary gland following estrogen+progestin therapy in comparison with estrogen-alone therapy or placebo in a postmenopausal primate model.

Methods: Ovariectomized female cynomolgus monkeys (Macaca fascicularis) were randomized into 3 treatment groups: placebo, oral conjugated equine estrogens at 0.625 mg/d (CEE) and CEE with medroxyprogesterone acetate (MPA) at 2.5 mg/d (CEE+MPA). Treatments were administered in the diet for 2 years. RANK, RANKL, and OPG mRNA were measured in mammary gland samples (n = 28 to 31/group) using quantitative real-time reverse transcriptase polymerase chain reaction (qPCR). Antibodies against human RANK (N-2B10 and N-1H8; Amgen) and human RANKL (M366; Amgen) were used for immunohistochemistry (IHC) of formalin-fixed tissues (n = 20 to 23/group) along with an isotype control. The intensity of IHC staining was scored on a semiquantitative scale (0 = absent, 1 = weak, 2 = moderate, 3 = intense). Incidence was scored as a positive IHC signal (any intensity).

Results: Analysis of mRNA revealed that treatment with CEE+MPA resulted in a significantly greater ratio of RANKL to OPG compared to treatment with placebo (P < 0.01) or CEE (P < 0.05). IHC demonstrated RANK protein was predominately localized to basal cells of ducts and acinar cells, while RANKL protein was localized to luminal cells. RANK protein was detected in mammary glands of animals in all groups: placebo (74%), CEE (55%), and CEE +MPA (53%). RANKL protein was expressed in the animals treated with CEE+MPA (74%) and CEE (15%), but not in placebo-treated animals. Quantitative expression of specific mRNA markers for epithelial proliferation (MKI67) and density (KRT19) was highest for CEE+MPA, intermediate for CEE, and lowest for placebo.

Discussion: Key components of the RANKL/RANK pathway are expressed in the normal postmenopausal primate mammary gland, modulated by long-term estrogen+progestin exposure at clinically relevant doses, and associated with changes in estrogen+progestin-driven proliferation of mammary tissue.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-04-05.

P2-01-02: Pharmacological Inhibition of RANKL Attenuates Mammary Tumor Development and Lung Metastases in the MMTV-neu Transgenic Spontaneous Tumor Model

Background: RANK and its ligand (RANKL), key factors for bone remodeling and metastasis, are crucial for the development of mouse mammary glands during pregnancy. In a mouse model of mammary carcinogenesis induced by the hormone medroxyprogesterone (MPA) and a carcinogen (DMBA), treatment with a RANKL inhibitor RANK-Fc reduces hormone-induced mammary epithelial proliferation, incidence of preneoplasias, and mammary adenocarcinomas. This study assessed the expression of mouse RANK and RANKL in FVB wild-type (WT) and MMTV-neu mice and whether RANKL inhibition could inhibit mammary tumor formation and lung metastases in this model.

Materials and Methods: MMTV-neu mice were treated with the RANKL inhibitor RANK-Fc (10 mg/kg, 3x/week) or vehicle (muFc) beginning at 20 weeks. Mammary tumor formation was determined by palpation and confirmed by histologic examination. The presence of preneoplastic lesions and histotype of tumors was determined on H&E-stained tissues. Entire lungs from tumor-bearing MMTV-neu mice were step-sectioned at 75 μm and assessed histologically for the presence of metastases. Epithelial proliferation was measured by BrdU labeling of the mammary epithelium. The expression of RANK, RANKL, cyclin D1, progesterone receptor (PR), F4/80, FOXP3, and CD3 was determined by immunohistochemistry (IHC).

Results: RANK-Fc treatment of MMTV-neu mice resulted in a significant reduction in total number of mammary tumors and lung metastases quantified at necropsy but did not impact median time to spontaneous mammary tumor onset. RANKL inhibition resulted in reduced pre-neoplasias and corresponding reductions in proliferation and cyclin D1 within normal mammary epithelia but not in adenocarcinomas. Treatment of MMTV-neu mice with RANK-Fc also reduced the incidence and number of lung metastases per mouse. RANK protein was detected in normal mammary epithelium, preneoplastic mammary intraepithelial neoplasia (MIN), the epithelial component of adenocarcinomas and lung metastases, and monocytic element adjacent to primary MMTV-neu tumors. RANKL protein was detected in normal mammary epithelium but not in the epithelia, stromal component or infiltrating cells of adenocarcinomas or lung metastases. Progesterone receptor expression was observed in normal mammary epithelium of MMTV-neu and FVB WT mice but decreased over time in MMTV-neu mice and was not detected in preneoplasias, adenocarcinomas, or lung metastases. CD3 was detected in mammary tumors and lung metastases but neither RANKL nor FOXP3 were present. Staining of lymph nodes was used as a positive control for RANKL and FOXP3. Expression of FOXP3 and RANKL were not measurably colocalized in dual staining of lymph nodes.

Discussion: RANKL is expressed in normal mammary epithelia of MMTV-neu mice but is not detected in any stage of atypia, including MIN, mammary adenocarcinoma, or lung metastases. Selective pharmacological inhibition of RANKL by RANK-Fc treatment attenuated mammary tumor development and lung metastases in the MMTV-neu transgenic spontaneous tumor model.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-01-02.

Application of pharmacodynamic assays in a phase Ia trial of Apo2L/TRAIL in patients with advanced tumors

3535

Background: Recombinant human (rh) Apo2L/TRAIL is a dual pro-poptotic receptor agonist (PARA) that induces apoptosis by binding to pro-apoptotic receptors DR4 and DR5, which recruit a death inducing signaling complex upon ligand binding. This results in activation of the effector caspase 3/7, that subsequently cleaves intracellular substrates to execute cellular apoptosis. A Phase1a trial is underway to evaluate the safety and tolerability of rhApo2L/TRAIL in patients with advanced tumors. The aim of this study was to develop and validate high-throughput pharmacodynamic assays to monitor rhApo2L/TRAIL activity in easily accessible patient samples such as serum. Methods: To monitor rhApo2L/TRAIL activity in patients, we optimized assays to measure the release of the apoptotic markers caspase 3/7, cytokeratin 18 (CK18), and genomic DNA (gDNA) in serum. Serum caspase 3/7 levels were monitored using the Caspase Glo kit, which generates a luminescent signal upon cleavage of a caspase 3/7 substrate; cleavage of the caspase substrate CK18 was measured using an optimized form of the M30 ELISA assay; gDNA was measured using a β-actin-specific TaqMan real-time PCR assay. Mice bearing Colo205 xenografts were treated with rhApo2L/TRAIL and sera were collected and assayed for apoptotic markers. Upon validating these assays, we monitored the levels of apoptotic markers in cancer patients who received rhApo2L/TRAIL. Results: We detected transient increases in apoptotic markers in mouse sera 8–24 hr after treatment with rhApo2L/TRAIL. This increase was dose-dependent and correlated with active caspase 3 detected by IHC in Colo205 tumors. In the phase Ia study, increases in serum caspase 3/7 and gDNA levels were observed in >50% of colorectal, lung and sarcoma patients evaluated. Preliminary analyses show the percentage of increase correlates using both analytes and is dose-dependent. Conclusions: These findings support the use of serum-based pharmacodynamic assays as a means to monitor rhApo2L/TRAIL activity in patients with advanced tumors. A complete analysis of all patient serum samples from the ongoing phase Ia trial will be reported.

No significant financial relationships to disclose.

Investigations into biochemical changes due to diurnal variation and estrus cycle in female rats using high-resolution (1)H NMR spectroscopy of urine and pattern recognition

Metabonomic methods utilizing (1)H NMR spectroscopy and pattern recognition analysis (NMR-PR) have been applied to investigate biochemical variation in a control population of female rats over time in relation to diurnal and estrus cycle fluctuations. Urine samples were collected twice daily (6 AM-6 PM and 6 PM-6 AM) from female rats (n = 10) for a period of 10 days. (1)H NMR spectroscopic analysis and PR were performed on each sample. Subtle differences in the endogenous metabolite excretion profiles of urine samples at the various stages of the estrus cycle were observed. The main inherent metabolic clustering in the principal components analysis (PCA) maps was related to interrat variation and was observed in the first two principal components (PCs), accounting for 66% of the variance in these data. Separation of urinary data according to time of sampling (day and night) was achieved in the lower PCs. Some of the differences in the urinary profiles of day and night samples causing this separation were attributed to the increase in metabolic activity of the rat during the night. Individual rat data were also mapped as a function of time, using PCA, to produce a metabolic trajectory, which in a number of cases facilitated separation of one or more stages of the estrus cycle. Several of the fluctuations observed between urine samples collected during the different stages of the estrus cycle may be related to hormone levels. Although variation in metabolite profiles relating to both diurnal and hormonal variation could be detected these perturbations were minor compared with the effects observed due to interrat variation. This is the first time that a hormonal cycle has been described for individuals based on NMR spectroscopic and multivariate analysis of metabolic data and shows the value of metabonomic methods in the investigation of physiological variation and rhythms.