Sonal Varma

Mass Spectrometric Profiling of N-Linked Oligosaccharides and Uncommon Glycoform in Mouse Serum with Head and Neck Tumor

N-linked oligosaccharides obtained from total serum of mice with implanted head and neck tumors were analyzed and compared with those from control samples of healthy mice. Methods used include a combination of a derivatization procedure with phenylhydrazine (PHN) and analysis by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). Oligosaccharides were enzymatically released from total serum with PNGaseF and purified by high-performance liquid chromatography (HPLC) on a reversed-phase column. Mass spectra contained ion peaks of labeled oligosaccharides and MS/MS experiments provided useful data for the structural elucidation of these compounds. More than 40 N-glycans with compositions characteristic of high-mannose, hybrid, complex, neutral, and sialylated structures were identified in the serum of tumoral mice. Significant differences between samples were observed with respect to the abundances of high mannose and hybrid glycans. These oligosaccharides showed higher relative intensities in the spectra obtained from the cancer sera. Complex sialylated oligosaccharides had similar abundances in both types of sera, with the exception of fucosylated biantennary disialylated oligosaccharide, which was mostly detected with lower abundance in control samples. In the MALDI spectra, several minor species corresponded to uncommon carbohydrates. These structures have been investigated in detail by MS/MS. Among these novel glycoforms, a few sialylated oligosaccharides without a free reducing end were identified. Also, glycans with an extra 60 u were observed and likely feature the presence of a 2-acetamido-2-deoxyoctose residue attached on antennae of 3- or 6-linked mannose.

Automated Quantitative Analysis of p53, Cyclin D1, Ki67 and pERK Expression in Breast Carcinoma Does Not Differ from Expert Pathologist Scoring and Correlates with Clinico-Pathological Characteristics

There is critical need for improved biomarker assessment platforms which integrate traditional pathological parameters (TNM stage, grade and ER/PR/HER2 status) with molecular profiling, to better define prognostic subgroups or systemic treatment response. One roadblock is the lack of semi-quantitative methods which reliably measure biomarker expression. Our study assesses reliability of automated immunohistochemistry (IHC) scoring compared to manual scoring of five selected biomarkers in a tissue microarray (TMA) of 63 human breast cancer cases, and correlates these markers with clinico-pathological data. TMA slides were scanned into an Ariol Imaging System, and histologic (H) scores (% positive tumor area x staining intensity 0–3) were calculated using trained algorithms. H scores for all five biomarkers concurred with pathologists’ scores, based on Pearson correlation coefficients (0.80–0.90) for continuous data and Kappa statistics (0.55–0.92) for positive vs. negative stain. Using continuous data, significant association of pERK expression with absence of LVI (p = 0.005) and lymph node negativity (p = 0.002) was observed. p53 over-expression, characteristic of dysfunctional p53 in cancer, and Ki67 were associated with high grade (p = 0.032 and 0.0007, respectively). Cyclin D1 correlated inversely with ER/PR/HER2-ve (triple negative) tumors (p = 0.0002). Thus automated quantitation of immunostaining concurs with pathologists’ scoring, and provides meaningful associations with clinico-pathological data.

Immunohistochemical Assessment of Expression of Centromere Protein—A (CENPA) in Human Invasive Breast Cancer

Abnormal cell division leading to the gain or loss of entire chromosomes and consequent genetic instability is a hallmark of cancer. Centromere protein –A (CENPA) is a centromere-specific histone-H3-like variant gene involved in regulating chromosome segregation during cell division. CENPA is one of the genes included in some of the commercially available RNA based prognostic assays for breast cancer (BCa)—the 70 gene signature MammaPrint® and the five gene Molecular Grade Index (MGISM). Our aim was to assess the immunohistochemical (IHC) expression of CENPA in normal and malignant breast tissue. Clinically annotated triplicate core tissue microarrays of 63 invasive BCa and 20 normal breast samples were stained with a monoclonal antibody against CENPA and scored for percentage of visibly stained nuclei. Survival analyses with Kaplan–Meier (KM) estimate and Cox proportional hazards regression models were applied to assess the associations between CENPA expression and disease free survival (DFS). Average percentage of nuclei visibly stained with CENPA antibody was significantly higher (p = 0.02) in BCa than normal tissue. The 3-year DFS in tumors over-expressing CENPA (>50% stained nuclei) was 79% compared to 85% in low expression tumors (<50% stained nuclei). On multivariate analysis, IHC expression of CENPA showed weak association with DFS (HR > 60.07; p = 0.06) within our small cohort. To the best of our knowledge, this is the first published report evaluating the implications of increased IHC expression of CENPA in paraffin embedded breast tissue samples. Our finding that increased CENPA expression may be associated with shorter DFS in BCa supports its exploration as a potential prognostic biomarker.

STAT5A is regulated by DNA damage via the tumor suppressor p53

Here we report that the STAT5A transcription factor is a direct p53 transcriptional target gene. STAT5A is well expressed in p53 wild type cells but not in p53-null cells. Inhibition of p53 reduces STAT5A expression. DNA damaging agents such as doxorubicin also induced STAT5A expression in a p53 dependent manner. Two p53 binding sites were mapped in the STAT5A gene and named PBS1 and PBS2; these sites were sufficient to confer p53 responsiveness in a luciferase reporter gene. Chromatin immunoprecipitation experiments revealed that PBS2 has constitutive p53 bound to it, while p53 binding to PBS1 required DNA damage. In normal human breast lobules, weak p53 staining correlated with regions of intense STAT5A staining. Interestingly, in a cohort of triple negative breast tumor tissues there was little correlation between regions of p53 and STAT5A staining, likely reflecting a high frequency of p53 mutations that stabilize the protein in these tumors. We thus reveal an unexpected connection between cytokine signaling and p53.

Dataset of STAT5A status in breast cancer

We analysed STAT5A gene expression in breast cancer using the Oncomine database. We exemplify four representative studies showing that STAT5A is generally downregulated in breast cancer.

Abstract 943: Characterization of RET and integrin signaling in thyroid tumors

Proceedings: AACR 103rd Annual Meeting 2012‐‐ Mar 31‐Apr 4, 2012; Chicago, IL The REarranged in Transfection (RET) tyrosine kinase is important during development of neural crest derived tissues such as the enteric and sympathetic nervous systems, but is also activated in multiple human tumour types. Activating mutations of RET play an important role in thyroid cancers including both papillary thyroid carcinoma (PTC) and medullary thyroid carcinoma (MTC). More broadly, recent studies have also shown that RET may be active in other tumour types such as pancreatic and breast cancers, and that the overexpression and activation of RET may be linked to more aggressive disease. In order to explore this link, we have investigated the expression of RET in a large cohort of over 400 thyroid tumour samples of multiple histologies, ranging from benign lesions through to aggressive, metastatic disease by immunhistochemistry using a thyroid tissue microarray. Our data suggest that RET expression may correlate with more aggressive phenotype across a broader spectrum of tumours than previously predicted. We are currently investigating this association further; first by examining the expression of two major RET protein isoforms, RET9 and RET51, in this panel of tumours, and second, by investigating the coexpression of RET with another group of proteins associated with tumour progression, the integrin family. In preliminary studies, we have evaluated RET coexpression with two integrin subunits, ITGB1 and ITGB3. Our data suggest correlation between tumour aggressiveness and coexpression of RET with these integrin subunits. Together, our data provides insight on the role of RET in increasing aggressiveness over a range of tumours. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 943. doi:1538-7445.AM2012-943

Abstract 3207: Automated quantitative analysis of p53, cyclin D1 and pErk expression in breast carcinoma does not differ from expert pathologist scoring and correlates well with clinico-pathological characteristics

Proceedings: AACR 102nd Annual Meeting 2011‐‐ Apr 2‐6, 2011; Orlando, FL Prognosis and risk assessment of breast cancer patients are currently driven by TNM stage, ER/PR/HER2 expression, tumor grade and lymphovascular invasion (LVI). However there is critical need for improved biomarker assessment platforms to better predict systemic treatment response. One roadblock is the lack of semi-quantitative methods to reliably measure expression, activity and localization of biomarkers in formalin-fixed tumor specimens. The present study assesses reliability of automated IHC scoring compared to manual scoring of routine and non-routine biomarkers (HER2, cyclin D1, p53 and phospho(p)-ERK) on a human breast cancer tissue microarray according to REMARK guidelines, and correlates these markers with clinical-pathological data. Using a triplicate core TMA of formalin-fixed paraffin embedded tissues, we investigated 63 primary invasive breast cancers, for which ER/PR/HER2 status, LVI, grade and recurrence status were recorded. IHC was performed on the TMA for the above biomarkers (pH 6 citrate buffer conditions). Histologic (H) scores (% positive tumor area × staining intensity 0-3) were determined manually by two independent evaluators with resolution of discordant cases by a senior pathologist. Excellent replicability was observed between H scores for each marker compared on replicate slides, as determined by Spearman correlations (0.79-0.82). Each TMA slide was then scanned into the Ariol Imaging System, algorithms were trained for each marker, and H scores were calculated. Pearson correlation coefficients (with data left as continuous) and Kappa statistics (with dichotomized data) were used for inter-method comparisons. Associations between biomarker positivity and clinical data were assessed by Fishers exact test. Excellent concordance between manual and automated Ariol scores was observed for all four markers based on Kappa statistics (0.667-0.813) and Pearson correlation coefficients (0.790-0.885). Distinct proportions of tumor cases showed any positive staining for membranous HER2 (19/63), nuclear p53 (16/56), cyclin D1 (26/57) and pERK (32/59). A statistically significant association of pERK positivity with absence of LVI (p=0.0025) and lymph node negativity (p=0.0006) was observed. In contrast, pERK positivity was associated with high-grade tumors (p=0.0040), consistent with a role of pERK in poorly differentiated high-grade primary tumors. p53 over-expression, characteristic of dysfunctional p53 in breast cancer, was also associated with high tumor grade (p=0.0074). Thus automated quantitation of immunostaining yields objective results that do not differ from pathologists’ scoring, and provide meaningful associations with clinico-pathological data. (Supported by CIHR, PSI, and Queens Dept. Pathol. & Mol. Med.) Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3207. doi:10.1158/1538-7445.AM2011-3207

Abstract 2370: A novel Src/Ezrin pathway regulates local invasion and metastasis of breast

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC The membrane cytoskeletal crosslinker, ezrin, is frequently over-expressed in human breast cancer, and is required for motility and invasion of epithelial cells. Our group recently showed that ezrin acts co-operatively with the non-receptor tyrosine kinase, Src, in transformation of epithelial cells, requiring phosphorylation of specific tyrosines on ezrin by Src (Srivistava et al. MBC, 16:1481-90, 2005; Naba et al. EMBO J. 27:38-50, 2008). We showed that Src binds to ezrin Y190 via its SH2 domain and that ezrin Y145 and Y477 are phosphorylated by Src. Expression of Y145F or Y477F ezrin mutants, which cannot be phosphorylated by Src, delays cell spreading of epithelial cells on fibronectin substratum. Furthermore, ezrin is required for metastasis in a breast tumor xenograft model (Elliott et al. BCR 7:R365-73, 2005). We therefore examined whether Src/ezrin interaction regulates invasion and metastasis of breast cancer, using a mouse breast tumor (AC2M2) xenograft model. Transplanted tumor cells expressing vector alone rapidly infiltrated into surrounding stroma, underlying abdominal muscle, visceral organs and metastasized to lung and other sites. Interestingly, both Y145F and Y477F ezrin mutants had little effect on primary tumor growth rate. However, Y145F ezrin showed a marked attenuating effect on pulmonary metastasis. Furthermore, transplanted mouse breast carcinoma cells over-expressing Y477F ezrin formed cohesive tumors that were circumscribed by normal stroma with no detectable infiltration into underlying muscle. Marked lymphovascular involvement, a prognostic indicator of relapse in clinically advanced human breast cancers, was evident in control tumors but not in tumors expressing Y477F ezrin. In a preliminary IHC analysis, total ezrin showed strong cytoplasmic and weak membranous staining in the cortical region of control primary tumors, but showed strong membranous and weak cytoplasmic staining in Y477F ezrin-expressing primary tumors. The cohesive phenotype of Y477F ezrin is likely a consequence of a defect in cell adhesion as described by Naba et al. Thus, the precise state of ezrin phosphorylation may elicit different effects on local invasion and metastasis. Our study implicates a role of the Src/ezrin pathway in regulating local invasion and metastasis of breast carcinoma cells, and provides a clinically relevant model for assessing the Src/ezrin pathway as a potential prognostic marker and treatment target for invasive breast cancer. (Funded by CBCRA 17374 and ARC 4823) Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2370.

Abstract P2-06-11: Ramifications of HER2/ER/PR Guidelines from ASCO/CAP for Translational Cancer Research Using a Cohort from a Tertiary Care Centre in Ontario

Background: A transdisciplinary team from basic science, pathology, clinical and biostatistics was assembled to establish a framework with which to take novel laboratory biomarkers and targets to clinical validation. Human epidermal growth factor receptor (HER2), estrogen (ER) and progesterone (PR) receptor are of important prognostic and predictive value and drivers of systemic therapy for breast cancer (BC). As a first step, the current ASCO/CAP guidelines were used to re-assign centrally reviewed tumour specimens and compare to the clinically assigned scores for ER/PR and HER2. Methods: With REB approval, a cohort of 62 cases of non-metastatic invasive BC with banked tumour specimens was assembled between 2005 and 2007. Clinico-pathological information for each case was retrospectively obtained from the medical file and entered into an anonymized database. Full section slides were originally stained by routine immunohistochemistry (IHC). Categorical clinical scores for ER/PR (negative-neg/weak/positive-pos) were compared to the continuous scores assigned in a blinded fashion using ASCO/CAP criteria (% pos/H-score). Categorical clinical scores obtained with duplicate IHC antibody staining of full sections for HER2 (neg/equivocal-eq/pos) were compared to those obtained from IHC assessments of triplicate 6mm cores in a tissue microarray (TMA) that were assigned to be neg/eq/pos using ASCO/CAP criteria. A senior breast pathologist adjudicated discordant specimens. Exact Fisher tests were used to compare the two sets of categorical assessments. Results: Mean age was 43.5 years, (range 29-49). The majority of the cohort (59.7%) had N0 disease and received adjuvant chemotherapy (74.2%); 72.6% of the cohort was alive at the time of this analysis. Score means and ranges of ER/PR are displayed below. Two of 16 clinically ER neg cases (12.5%) were rescored as pos and 0/43 clinically ER pos cases were rescored as neg, P ER/PR scores Conclusions: Systemic therapy recommendations could be impacted in a small but substantive number of cases by the methodology used for biomarker assessment and scoring, particularly near threshold values. This study illustrates that the scoring criteria used may be an important contributor to variability in correlative biomarker studies. Consideration should be given to routine systematic reassessment with continuous scoring for biomarker data proposed for use in correlative science studies. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-06-11.