Application of quantitative proteomic analysis for cancer therapy using "reverse-phase" protein lysate microarrays

Proteomic analysis using quantitative high-throughput technology can provide new insights in cancer therapeutics. It can reveal how cancer cells respond to given therapies by measuring multiple dimensions of information, from which dynamic proteomic responses can be observed. A lack of high throughput proteomic technologies has previously limited such multi-dimensional approaches. We have developed a high-throughput, "reverse-phase" protein microarray system which can handle more than 20,000 lysate features on a single glass slide. Subsequent immunochemical detection methods allow us to monitor protein expression in a quantitative manner as a function of both time and drug dosage. The data generated using this RPA technology has proved to be an excellent reference for theoretical protein network modeling in vitro. Clinical evaluation of drug efficacy based on the data generated by this technology may provide a means to accurately predict the effectiveness of cancer therapies.     

Classification of breast cancer using genetic algorithms and tissue microarrays.

PURPOSE: A multitude of breast cancer mRNA profiling studies has stratified breast cancer and defined gene sets that correlate with outcome. However, the number of genes used to predict patient outcome or define tumor subtypes by RNA expression studies is variable, nonoverlapping, and generally requires specialized technologies that are beyond those used in the routine pathology laboratory. It would be ideal if the familiarity and streamlined nature of immunohistochemistry could be combined with the rigorously quantitative and highly specific properties of nucleic acid-based analysis to predict patient outcome. EXPERIMENTAL DESIGN: We have used AQUA-based objective quantitative analysis of tissue microarrays toward the goal of discovery of a minimal number of markers with maximal prognostic or predictive value that can be applied to the conventional formalin-fixed, paraffin-embedded tissue section. RESULTS: The minimal discovered multiplexed set of tissue biomarkers was GATA3, NAT1, and estrogen receptor. Genetic algorithms were then applied after division of our cohort into a training set of 223 breast cancer patients to discover a prospectively applicable solution that can define a subset of patients with 5-year survival of 96%. This algorithm was then validated on an internal validation set (n=223, 5-year survival=95.8%) and further validated on an independent cohort from Sweden, which showed 5-year survival of 92.7% (n=149). CONCLUSIONS: With further validation, this test has both the familiarity and specificity for widespread use in management of breast cancer. More generally, this work illustrates the potential for multiplexed biomarker discovery on the tissue microarray platform.     

Quantitative analysis of breast cancer tissue microarrays shows high cox-2 expression is associated with poor outcome

Epidemiologic and preclinical studies suggest that cyclooxygenase-2 (Cox-2) may promote tumor growth and spread by affecting angiogenesis and apoptosis in breast cancer. Using a tissue microarray (TMA), we analyzed the expression and subcellular localization of Cox-2 by AQUA and X-tile, our algorithms for quantitative analysis of protein expression and determination of optimal cutpoints. Our TMA consisted of 669 Stage I-III primary breast cancers. The total tumor and subcellular expression of Cox-2 were then correlated with clinicopathologic factors and with survival. Cox-2 expression appeared higher in malignant than in benign tissue and was predominantly membrane/cytoplasmic (i.e. non-nuclear). X-tile determines an optimum cutpoint on a training set then uses this cutpoint on a validation set. This cutpoint was 19.3 (top 44 percent defined as positive) with high nonnuclear Cox-2 expressers having significantly worse survival. Cox-2 expression also was inversely associated with estrogen receptor (ER) and progesterone receptor (PR), and directly associated with nuclear grade. Multivariate analysis showed that Cox-2 remained a significant prognostic factor for survival independent of tumor size, nodal status, ER, Her2/neu, and grade. In summary, Cox-2 is overexpressed in breast neoplasms, is associated with other markers of poor prognosis, and is significantly associated with worse survival independent of known prognostic factors. Furthermore, AQUA and X-tile analysis suggest an optimal cutpoint that may be helpful in future investigations of Cox-2 and specifically, in studies looking at its expression as a predictive biomarker in clinical trials of Cox-2 inhibitors in breast cancer.     

Immunoprofile from tissue microarrays to stratify familial breast cancer patients

Familial breast cancer (BC) is a heterogeneous disease with variable prognosis. The identification of an immunoprofile is important to predict tumor behavior for the routine clinical management of familial BC patients. Using immunohistochemistry on tissue microarrays, we studied 95 familial BCs in order to analyze the expression of some biomarkers involved in different pathways. We used unsupervised hierarchical clustering analyses (HCA), performed using the immunohistochemical score data, to define an immunoprofile able to characterize these tumors. The analyses on 95 and then on a subset of 45 tumors with all biomarkers contemporarily evaluable, revealed the same biomarker and patient clusters. Focusing on the 45 tumors we identified a group of patients characterized by the low expression of estrogen receptor (P = 0.009), progesterone receptor (P < 0.001), BRCA1 (P = 0.005), nuclear Na⁺/H⁺ exchanger regulatory factor 1 (NHERF1) (P = 0.026) and hypoxia inducible factor-1 alpha (P < 0.001), and also by the higher expression of MIB1 (P = 0.043), cytoplasmic NHERF1 (P = 0.004), cytoplasmic BRCT-repeat inhibitor of hTERT expression (P = 0.001), vascular endothelial growth factor (VEGF) (P = 0.024) and VEGF receptor-1 (P = 0.029). This immunoprofile identified a more aggressive tumor phenotype associated also with a larger tumor size (P = 0.012) and G3 grade (P = 0.006), confirmed by univariate and multivariate analyses. In conclusion, the clinical application of HCA of immunohistochemical data could allow the assessment of prognostic biomarkers to be used simultaneously. The 10 protein expression panel might be used to identify the more aggressive tumor phenotype in familial BC and to direct patients towards a different clinical therapy.     

Proteome-wide analysis of head and neck squamous cell carcinomas using laser-capture microdissection and tandem mass spectrometry

Remarkable progress has been made to identify genes expressed in squamous cell carcinomas of the head and neck (HNSCC). However, limited information is available on their corresponding protein products, whose expression, post-translational modifications, and activity are ultimately responsible for the malignant behavior of this tumor type. We have combined laser-capture microdissection (LCM) with liquid chromatography-tandem mass spectrometry (LC-MS/MS) to identify proteins expressed in histologically normal squamous epithelium and matching SCC. The protein fraction from approximately 10,000-15,000 normal and tumor cells was solubilized, digested with trypsin, and the resulting peptides were analyzed by LC-MS/MS. Database searching of the resulting sequence information identified 30-55 proteins per sample. Keratins were the most abundant proteins in both normal and tumor tissues. Among the proteins differentially expressed, keratin 13 was much lower in tumors, whereas heat-shock (Hsp) family members were highly expressed in neoplastic cells. Wnt-6 and Wnt-14 were identified in both normal and tumor tissues, respectively, and placental growth factor (PIGF) was detected only in tumors. Immunohistochemical analysis of HNSCC tissues revealed lack of keratin 13 in tumor tissues, and strong staining in normal epithelia, and high expression of Hsp90 in tumors. Our study, by combining LCM and proteomic technologies, underscores the advantages of this approach to investigate complex changes at the protein level in HNSCC, thus complementing existing and emerging genomic technologies. These efforts may likely result in the identification of new biomarkers for HNSCC that can be used to diagnose disease, predict susceptibility, and monitor progression in individual patients.     

Proteomic analysis of differently archived breast cancer tissues

Currently, the most common practice of human breast tissue preservation is formalin fixation which ensures good quality for histopathological analyses but damages DNA, RNA, and proteins, impairing their usefulness for molecular analysis and biomarker investigations. We investigated the potential value of a non-toxic fixative for sparing proteins preserved in paraffin-embedded breast biopsies. Specimens were fixed in formalin-free fixative prior to paraffin embedding, and then processed for quality and quantity of protein conservation. Similar protein patterns were observed in formalin-free fixative and frozen tissues using mono- and bi-dimensional electrophoresis, as well as western blotting. Protein patterns assessed by mass spectrometric analysis were found to be identical for frozen and formalin-free-fixed tissues. Immunohistochemistry using various antibodies showed comparable results for both tissue storage methods. In conclusion, we believe that formalin-free fixative represents an easy-to-use alternative to formalin for archived tissue and for biomarker investigations, since it simultaneously protects both the histomorphology and the integrity of macromolecules.     

An ultrastructural morphometric analysis of normal human mammary tissue and human breast cancer

An ultrastructural morphometric study of normal and tumourous adult human mammary tissue is presented. The data show a characteristic numerical pattern for the different neoplastic stages of breast tissues. By applying a special data-comparison program a standardized diagnostic and prognostic system appears possible through the morphometric analysis of ultrathin epon sections of surgically obtained material.     

Proteomic analysis of human prostate cancer.

Proteomics is a promising approach in the identification of proteins and biochemical pathways involved in tumorigenesis. In an effort to discover such proteins and pathways that are deregulated in prostate tumorigenesis, cellular proteomes of matched normal prostate epithelial cells and high-grade prostate cancer cells were analyzed by tissue microdissection, two-dimensional electrophoresis, and mass spectrometry. Forty protein alterations were detected in the tumors; however, the majority of these changes were not shared among the 12 neoplasms. In contrast, parallel cDNA microarray analysis identified a number of common gene expression changes. The marked heterogeneity of the observed protein alterations may have significance with regard to tumor biology and research strategies for molecular profiling analyses of human prostate cancer.     

Loss of antigenicity in stored sections of breast cancer tissue microarrays.

Immunohistochemical characterization of tumor tissues in epidemiological studies is a promising approach to identify breast cancer subtypes with distinct etiology. The recent development of the tissue microarray (TMA) technique allows for standardized, rapid, and cost-effective immunohistochemical characterization of many cases, which is critical in epidemiological studies. Sectioning paraffin blocks at different times results in loss of material, which can be reduced by preparing many sections each time a block is cut. However, data suggest that staining intensity declines in whole sections prepared from conventional paraffin blocks with storage time, resulting in false-negative results. This problem would be accentuated in TMAs because of the limited tissue representation of each case. To evaluate this concern, we prepared a single TMA block from 125 invasive breast carcinomas collected in a population-based case-control study conducted in Poland and compared estrogen receptor (ER-alpha), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression in sections cut and stored for 6 months at room temperature with sections cut from the same TMA block and stained on the same day. Percentage of positive cases for stored versus fresh sections was similar for ER (59.0%) but significantly higher in fresh sections for PR (56.3% versus 64.1%, P = 0.01) and HER2 (45.5% versus 64.4%, P < 0.001). Among cases positive in both stored and fresh sections, the median percentage of immunoreactive cells was significantly reduced and the staining intensity was consistently lower in stored compared with fresh sections. We conclude that loss of immunoreactivity is an important problem in TMAs of breast cancer. Improved methods for sectioning TMAs and storing tissue sections aimed at reducing loss of immunoreactivity are critical for the use of TMAs in epidemiological studies.     

Specific cytoplasmic alpha-fetoprotein binding protein in MCF-7 human breast cancer cells and primary breast cancer tissue

Summary Direct evidence was obtained for the existence of a specific high affinity alpha-fetoprotein (AFP)-binding protein in the cytosol of both MCF-7 human breast cancer cultured cells and primary breast cancer tissue from postmenopausal women using a nitrocellulose blotting assay. Scatchard analysis of the binding data for MCF-7 cells at 37° C revealed the presence of a single class of AFP binding sites with an apparent K_d of 4.5 × 10^–8 M, and 75,000 binding sites per cell. All 9 primary breast cancer cytosols obtained from postmenopausal women also contained measureable levels of this specific AFP-binding protein. The number of AFP molecules specifically bound varied considerably between patients and ranged from 29–250 fmol per mg cytosol protein. Levels of AFP-binding protein levels and estrogen receptor measured in these same breast cancer cytosols showed a positive statistical correlation (r = 0.85). Taken together, the present evidence for the existence of a specific cytoplasmic AFP-binding protein in MCF-7 cells and previously reported evidence forde novo synthesis of free immunoreactive and bound nonimmunoreactive forms of cytoplasmic AFP by MCF-7 cells is consistent with the conclusion that most of the endogenous AFP synthesized in breast cancer cells is rapidly bound to specific cytoplasmic AFP-receptors, and that binding of AFP to these receptors masks its immunoreactivity. The association of AFP synthesis with rapidly growing fetal liver and adult regenerating liver, germ-cell tumors, immature uterus, and breast cancer cells suggests that a positive correlation exists between cytoplasmic AFP-receptor levels and the proliferative capacity of malignant breast tumors, and therefore such measurements may provide useful therapeutic and/or prognostic information in individual patients.     

Ki67 expression in invasive breast cancer: the use of tissue microarrays compared with whole tissue sections

Background

Although the prognostic value of Ki67 in breast cancer is well documented, using optimal cut-points for patient stratification, reproducibility of the scoring and interpretation of the results remains a matter of debate particularly when using tissue microarrays (TMAs). This study aims to assess Ki67 expression assessed on TMAs and their matched whole tissue sections (WTS). Moreover, whether the cut-off used for WTS is reproducible on TMA in BC molecular classes and the association between Ki67 expression cut-off, assessed on TMAs and WTS, and clinicopathological parameters and patient outcome were tested.

Method

A large series (n = 707) of primary invasive breast tumours were immunostained for Ki67 using both TMA and WTS and assessed as percentage staining and correlated with each other, clinicopathological parameters and patient outcome. In addition, MKI67 mRNA expression was correlated with Ki67 protein levels on WTS and TMAs in a subset of cases included in the METABRIC study.

Results

There was moderate concordance in Ki67 expression between WTS and TMA when analysed as a continuous variable (Intraclass correlation coefficient = 0.61) and low concordance when dichotomised (kappa value = 0.3). TMA showed low levels of Ki67 with mean percentage of expression of 35 and 22% on WTS and TMA, respectively. MKI67 mRNA expression was significantly correlated with protein expression determined on WTS (Spearman Correlation, r = 0.52) and to a lesser extent on TMA (r = 0.34) (p < 0.001). Regarding prediction of patient outcome, statistically significant differences were detected upon stratification of patients with tumours expressing Ki67 at 10, 15, 20, 25 or 30% in TMA. Using TMA, ≥20% Ki67 provided the best prognostic cut-off particularly in triple-negative and HER2-positive classes.

Conclusion

Ki67 expression in breast cancer can be evaluated using TMA although different cut-points are required to emulate results from WTS. A cut-off of ≥20% for Ki67 expression in BC provides the best prognostic correlations when TMAs are used.

     

Hormone therapy failure in human prostate cancer: analysis by complementary DNA and tissue microarrays.

BACKGROUND: The molecular mechanisms underlying the progression of prostate cancer during hormonal therapy have remained poorly understood. In this study, we developed a new strategy for the identification of differentially expressed genes in hormone-refractory human prostate cancer by use of a combination of complementary DNA (cDNA) and tissue microarray technologies. METHODS: Differences in gene expression between hormone-refractory CWR22R prostate cancer xenografts (human prostate cancer transplanted into nude mice) and a xenograft of the parental, hormone-sensitive CWR22 strain were analyzed by use of cDNA microarray technology. To validate the data from cDNA microarrays on clinical prostate cancer specimens, a tissue microarray of specimens from 26 prostates with benign prostatic hyperplasia, 208 primary prostate cancers, and 30 hormone-refractory local recurrences was constructed and used for immunohistochemical detection of protein expression. RESULTS: Among 5184 genes surveyed with cDNA microarray technology, expression of 37 (0.7%) was increased more than twofold in the hormone-refractory CWR22R xenografts compared with the CWR22 xenograft; expression of 135 (2.6%) genes was reduced by more than 50%. The genes encoding insulin-like growth factor-binding protein 2 (IGFBP2) and 27-kd heat-shock protein (HSP27) were among the most consistently overexpressed genes in the CWR22R tumors. Immunohistochemical analysis of tissue microarrays demonstrated high expression of IGFBP2 protein in 100% of the hormone-refractory clinical tumors, in 36% of the primary tumors, and in 0% of the benign prostatic specimens (two-sided P =.0001). Overexpression of HSP27 protein was demonstrated in 31% of the hormone-refractory tumors, in 5% of the primary tumors, and in 0% of the benign prostatic specimens (two-sided P =.0001). CONCLUSIONS: The combination of cDNA and tissue microarray technologies enables rapid identification of genes associated with progression of prostate cancer to the hormone-refractory state and may facilitate analysis of the role of the encoded gene products in the pathogenesis of human prostate cancer.     

Histone deacetylase-1 and -3 protein expression in human breast cancer: a tissue microarray analysis

Summary Impaired histone acetylation was recognized to be involved in carcinogenesis. Furthermore, histone deacetylase (HDAC) inhibitors induce differentiation of breast cancer cells and inhibit tumour growth. These results prompted us to study HDAC-1 and -3 expression in breast tumours to establish their potential therapeutic and prognostic significance.HDAC-1 und HDAC-3 protein expression was analyzed immunohistochemically on a tissue microarray (TMA) containing 600 core biopsies from 200 patients. HDAC-1 and -3 expression was correlated to steroid hormone receptor-, Her2/neu- and proliferation status of tumours as well as to overall and disease free survival.Moderate or strong nuclear immunoreactivity for HDAC-1 was observed in 39.8% and for HDAC-3 in 43.9% of breast carcinomas. HDAC-1 and -3 expression correlated significantly with oestrogen and progesterone receptor expression (both p< 0.001). HDAC-1 expression predicted significantly better disease free survival (DFS: p=0.044), in particular, in patients with small tumours of all differentiation types (DFS: p=0.016). Multivariate analysis demonstrated that HDAC-1 is an independent prognostic marker.Our data suggest that evaluation of HDAC-1 protein expression enables a more precise assessment of the prognosis of breast cancer patients. Thus, HDAC-1 expression analysis might be clinically useful to facilitate an individual, risk-directed, adjuvant systemic therapy in breast cancer patients.     

In situ gene expression analysis of cancer using laser capture microdissection,microarrays and real time quantitative PCR

The simultaneous development of laser capture microdissection (LCM) and high-throughput mRNA analysis platforms has provided a significant technological advance in the world of cancer biology. The combination of such technologies provides a unique and powerful opportunity to directly assess the in situ molecular genetic events that are associated with initiation and progression of human malignancies. Despite these technological advances, the integration of LCM with high-throughput gene expression analysis has been met with various challenges. The goal of this review is to highlight some of the obstacles that we have faced and continue to face as we try to optimally apply LCM to in situ gene expression analysis.