Synergistic targeting of BRCA1 mutated breast cancers with PARP and CDK2 inhibition

Basal-like breast cancers (BLBC) are aggressive breast cancers that respond poorly to targeted therapies and chemotherapies. In order to define therapeutically targetable subsets of BLBC we examined two markers: cyclin E1 and BRCA1 loss. In high grade serous ovarian cancer (HGSOC) these markers are mutually exclusive, and define therapeutic subsets. We tested the same hypothesis for BLBC. Using a BLBC cohort enriched for BRCA1 loss, we identified convergence between BRCA1 loss and high cyclin E1 protein expression, in contrast to HGSOC in which CCNE1 amplification drives increased cyclin E1. In cell lines, BRCA1 loss was associated with stabilized cyclin E1 during the cell cycle, and BRCA1 siRNA led to increased cyclin E1 in association with reduced phospho-cyclin E1 T62. Mutation of cyclin E1 T62 to alanine increased cyclin E1 stability. We showed that tumors with high cyclin E1/BRCA1 mutation in the BLBC cohort also had decreased phospho-T62, supporting this hypothesis. Since cyclin E1/CDK2 protects cells from DNA damage and cyclin E1 is elevated in BRCA1 mutant cancers, we hypothesized that CDK2 inhibition would sensitize these cancers to PARP inhibition. CDK2 inhibition induced DNA damage and synergized with PARP inhibitors to reduce cell viability in cell lines with homologous recombination deficiency, including BRCA1 mutated cell lines. Treatment of BRCA1 mutant BLBC patient-derived xenograft models with combination PARP and CDK2 inhibition led to tumor regression and increased survival. We conclude that BRCA1 status and high cyclin E1 have potential as predictive biomarkers to dictate the therapeutic use of combination CDK inhibitors/PARP inhibitors in BLBC.

Synergistic CDK control pathways maintain cell size homeostasis

To coordinate cell size with cell division, cell size must be computed by the cyclin-CDK control network to trigger division appropriately. Here we dissect determinants of cyclin-CDK activity using a novel high-throughput single-cell in vivo system. We show that inhibitory phosphorylation of CDK encodes cell size information and works synergistically with PP2A to prevent division in smaller cells. However, even in the absence of all canonical regulators of cyclin-CDK, small cells with high cyclin-CDK levels are restricted from dividing. We find that diploid cells of equivalent size to haploid cells exhibit lower CDK activity in response to equal cyclin-CDK enzyme concentrations, suggesting that CDK activity is reduced by DNA concentration. Thus, multiple pathways directly regulate cyclin-CDK activity to maintain robust cell size homeostasis.

Cyclin E1 protein is stabilized in BRCA1 mutated breast cancers leading to synergy between CDK2 and PARP inhibitors

ABSTRACTBasal-like breast cancers (BLBC) are aggressive breast cancers that respond poorly to targeted therapies and chemotherapies. In order to define therapeutically targetable subsets of BLBC we examined two markers: cyclin E1 and BRCA1 loss. In high grade serous ovarian cancer (HGSOC) these markers are mutually exclusive, and define therapeutic subsets. We tested the same hypothesis for BLBC.Using a BLBC cohort enriched for BRCA1 loss, we identified convergence between BRCA1 loss and high cyclin E1 expression, in contrast to HGSOC in which CCNE1 amplification drives increased cyclin E1 gene expression. Instead, BRCA1 loss stabilized cyclin E1 during the cell cycle. Using siRNA we showed that BRCA1 loss leads to stabilization of cyclin E1 by reducing phospho-cyclin E1-T62, and conversely the overexpression of BRCA1 increased phospho-T62. Mutation of cyclin E1-T62 to alanine increased cyclin E1 stability. We showed that tumors with high cyclin E1/BRCA1 mutation in the BLBC cohort had decreased phospho-T62, supporting this hypothesis.Since cyclin E1/CDK2 protects cells from DNA damage and cyclin E1 is elevated in BRCA1 mutant cancers, we hypothesized that CDK2 inhibition would sensitize these cancers to PARP inhibition. CDK2 inhibition induced DNA damage and synergized with PARP inhibitors to reduce cell viability in BRCA1 mutated cell lines. Treatment of BLBC patient-derived xenograft models with combination PARP and CDK2 inhibition led to tumor regression and increased survival. We conclude that BRCA1 status and high cyclin E1 have potential as predictive biomarkers to dictate the therapeutic use of combination CDK inhibitors/PARP inhibitors in BLBC.

Immunohistochemical evaluation of Cyclin D1 and β-Catenin expression in subtypes of triple-negative breast cancer.

e12553 Background: The purpose of the study was to reveal characteristics of Cyclin D1 and β-Catenin expression in subtypes of triple negative breast cancer (TNBC). Methods: The study included 60 patients diagnosed with verified TNBC (T1N1M0/T2N0M0, ER-/PR-/Her2-). Immunohistochemical staining with CK5/6, EGFR, Cyclin D1, β-Catenin antibodies was performed. Basal-like (BL) subtype was identified when > 25% cells were CK5/6+ and/or EGFR+ (any staining). The expression of β-Catenin was assessed by its location: the membrane, cytoplasm, nucleus. Cells with negative staining on the membrane and cytoplasmic staining were counted and calculated as a percentage of the total number of tumor cells. Nuclear expression of β-Catenin was considered positive with at least 1 positively stained cell in the field of view at x200 magnification. Parametric statistical methods were used. Significance of the difference between the two means was assessed by the Student's t-test. Results: A number of TNBC samples showed Cyclin D1 overexpression. The loss of β-Catenin on the cell membrane and its abnormal accumulation in the cytoplasm was significantly more frequent in TNBC with Cyclin D1 overexpression. These processes were more pronounced in BL cancers. Loss of membrane expression in BL cancers: 29.6±6.1 and 58±7.4%; accumulation in the cytoplasm: 33.4±5.4 and 63.3±7.2%, with low and high Cyclin D1 respectively. Loss of membrane expression in non-BL cancers: 17.6±3.8 and 33.3±4.3%; accumulation in the cytoplasm: 27.3±6.3 and 49.5±8.2%, with low and high Cyclin D1 respectively. Only BL TNBC demonstrated β-catenin translocation to the nucleus (up to 20 stained cells in the field of view): in 33.3% tumors with Cyclin D1 overexpression and in 12% of tumors with its low expression. Conclusions: Levels of expression of β-catenin and Cyclin D1 in TNBC may have predictive value, and the choice of these biomarkers as targets will improve the treatment with new-generation medications.

Ki-67 and Cyclin D1 in subtypes of triple-negative breast cancer with different androgen profiles.

e12552 Background: The purpose of the study was to analyze Ki-67 and Cyclin D1 expression in subtypes of triple negative breast cancer (TNBC) with different androgen profiles. Methods: Tissues of 60 patients with verified TNBC (T1N1M0/T2N0M0, ER-/PR-/Her2-) were studied. Immunohistochemical staining was performed with antibodies to androgen receptors (AR), CK5/6, Ki-67, Cyclin D1, and EGFR. Basal-like (BL) subtype was identified when > 25% cells were CK5/6+ and/or EGFR+ (any staining). Tumors with ≥10% positively stained cells were considered AR+. Parametric statistical methods were used. Significance of the difference between the two means was assessed by the Student's t-test. Results: BL TNBC showed significantly higher Ki-67 expression (81.9±3.1%), compared to other subtypes (70.8±3.1%), p < 0.05. A number of TNBC samples demonstrated Cyclin D1 overexpression ( > 30% stained cells) with no correlation with Ki-67 expression (r = 0.12). High Cyclin D1 levels were less common in BL TNBC (32.4% tumors), compared to other subtypes (43.5% tumors), but its average values were significantly higher (89.9±4.7% vs. 66.5±6.6%, p < 0.05). AR+ tumors were observed in 11 cases (18.3%). Levels of Ki-67 were similar in TNBC with different AR profiles. The percentage of BL tumors was similar in AR+ and AR- cancers (63.6% and 61.3%, respectively). Expression of Cyclin D1 was more frequent in AR+ TNBC (45.5%), vs. 34.7% in АR-. Average levels of Cyclin D1 were significantly higher in AR+ tumors (51.8±13.7%), vs. AR- (31.6±5.1%). Conclusions: TNBC differs in the Cyclin D1 leves; some tumors show its overexpression. Inhibition of cyclin-dependent kinases blocks the cell cycle, which may be useful in the TNBC treatment, which currently has no targets for therapy.

Aberrant Cyclin E and Hepatocyte Growth Factor Expression, Microvascular Density, and Micro-Lymphatic Vessel Density in Esophageal Squamous Cell Carcinoma

Cyclin E and hepatocyte growth factor (HGF) have been observed as a multifaceted factor in many cancers, and the assessment of microvascular density (MVD) and micro-lymphatic vessel density (MLVD) has been used to quantify tumor angiogenesis and lymphangiogenesis. The aim of this study was to explore the association between expression of cyclin E, HGF, MVD, and MLVD, and clinicopathologic parameters in esophageal squamous cell carcinoma (ESCC). The expression of cyclin E, HGF, MVD, and MLVD were detected using immunohistochemically anticyclin E, HGF, CD34, and lymphatic vessel endothelial hyaluronan receptor 1 in 168 surgically resected ESCC cases and 30 normal esophageal mucosal samples. The expression levels of cyclin E, HGF, MVD, and MLVD were higher compared to controls. High cyclin E and HGF expression was found more frequently in the tumors larger than 5 cm ( P < .001), with poorer differentiation ( P = .034) and higher tumor node metastasis (TNM) staging ( P = .009) compared to their counterparts. Both MVD and MLVD values were found to be higher in the tumors larger than 5 cm ( P < .001), with poorer differentiation ( P < .001) and higher TNM staging ( P < .001) compared to their counterparts. Furthermore, the expression of MVD and MLVD in both the high cyclin E and high HGF expression groups was significantly higher compared to the low cyclin E and HGF expression groups ( P < .001). This study demonstrated that high cyclin E and HGF expression is closely correlated with tumor size, tumor differentiation degree, and TNM stage in patients with ESCC. These findings proposed that cyclin E and HGF could serve as novel molecular markers for preoperational evaluation of ESCC.

Non-Producing Multiple Myeloma (MM) Is a Distinct Subset Of Non-Secretory MM Characterized By High Cyclin D1 Expression and Decreased Progression Free Survival

Introduction Non-Secretory Multiple Myeloma (NSMM) is a well described entity with an estimated prevalence of 3% characterized by typical morphological and pathological MM characteristics and the absence of an M-protein on immunofixation electrophoresis (BJH, 2003, 121:749-757). The serum free light chain (sFLC) assay can detect an abnormal serum free light chain level in up to two-thirds of cases, suggesting that many cases of NSMM are at least minimally secretory. Among the true NSMM cases there is a subset in which no cytoplasmic Immunoglobulin synthesis is detected. This entity of ‘’Non Producing’’ Multiple Myeloma (NPMM) though recognized three decades ago (J Clin Invest. 1985;76(2):765-769) has only been described in a small number of case reports. The purpose of this study is to confirm the existence of and subsequently systemically describe NPMM, through the interrogation of MIRT’s MM Data Base (MMDB). Methods MMDB was interrogated from January 2000 until January 2013. Clinical and laboratory data were systemically reviewed for the MM isotype classification. Results 210 MM patients were identified as NSMM. Flow cytometry data on cytoplasmic immunoglobulin (cIg), performed at initial diagnosis, was available for 197 out of 210 patients. 19 cases revealed no cIg by flow cytometry thus identifying them as NPMM. Review of sFLC assay results revealed that in all cases but one, the free immunoglobulin light chain levels were low and the sFLC ratio was normal, confirming that the negative result of flow cytometry was not due to a technical failure. In one single case elevated sFCL levels were measured, which could be traced to grade 3 renal failure resulting in accumulation of FLC in the serum. The MM diagnoses were confirmed by chart review for each of the 19 cases, which revealed that all cases had symptomatic MM. The basic clinical/laboratory characteristics of the 19 NPMM cases are shown in table 1. Sixteen of the 19 cases had results on immunoglobulin In Situ Hybridization (ISH) and 5 were found to be positive for κ or λ light chains. In two cases a mixed population of light chain positive and negative plasma cells by ISH was identified. α and γ heavy chain positivity was found in 4 and 1 cases respectively. Thus in pathological terms, two categories of “minimally producing” and “ totally non producing” MM can be identified. There was no overlap between the cases with cytoplasmic heavy or light chain positivity by ISH. Osteolytic bone disease by X-rays was evident in 12/19 cases and active focal lesions were seen in 11/19 patients by PET and 16/19 by MRI. Metaphase cytogenetic analysis at initial diagnosis was abnormal in 10 out of 19 available samples, with the t(11;14) translocation found in 3 of them. Baseline Gene Expression Profiling (GEP) was available for 16 cases. 13 (81%) of them belonged to the CD1/CD2 subgroup, and all of them were characterized by markedly high cyclin D1 expression, the hallmark of the t(11;14) translocation in MM. In comparison, only 21% of patients enrolled in the Total Therapy 3 trial belonged to the CD1 or CD2 molecular subgroups (94/441, p<0.001) only 87 had a CCND1 gene expression above the FISH defined t(11;14) translocation threshold (p<0.001). With a median follow up of 68 months the median overall survival was not reached (6 deaths), and progression-free survival was only 8.9 months. To further delineate the pathophysiology of NPMM we performed a comparative genomic study of the GEP of the NPMM with the normally producing-secreting MM. The comparison revealed a list of 128 genes with a p value <0.0001 and a false discovery rate < 0.01. Ingenuity pathway analysis revealed that these genes were closely related with Cellular Development, Cellular Growth and Proliferation and Cellular Morphology. Conclusion NPMM is a very rare entity in MM. It is characterized by an increased expression of cyclin D1, suggesting that the t(11;14) translocation plays a role in its pathophysiology. Furthermore, while data on OS are yet inconclusive, it seems to exert a markedly decreased PFS implying a decreased sensitivity in MM therapy. Disclosures: No relevant conflicts of interest to declare.

p31comet acts to ensure timely spindle checkpoint silencing subsequent to kinetochore attachment

The spindle assembly checkpoint links the onset of anaphase to completion of chromosome-microtubule attachment and is mediated by the binding of Mad and Bub proteins to kinetochores of unattached or maloriented chromosomes. Mad2 and BubR1 traffic between kinetochores and the cytosol, thereby transmitting a “wait anaphase” signal to the anaphase-promoting complex. It is generally assumed that this signal dissipates automatically upon kinetochore-microtubule binding, but it has been shown that under conditions of nocodazole-induced arrest p31comet, a Mad2-binding protein, is required for mitotic progression. In this article we investigate the localization and function of p31comet during normal, unperturbed mitosis in human and marsupial cells. We find that, like Mad2, p31comet traffics on and off kinetochores and is also present in the cytosol. Cells depleted of p31comet arrest in metaphase with mature bipolar kinetochore-microtubule attachments, a satisfied checkpoint, and high cyclin B levels. Thus p31comet is required for timely mitotic exit. We propose that p31comet is an essential component of the machinery that silences the checkpoint during each cell cycle.