cDNA Microarray Gene Research Articles

Statistical analysis of C‐DNA microarray data for sample clustering and gene identification

PurposeThis purpose of this paper is to describe a fast and easy method of both clustering samples and identifying active genes in cDNA microarray data.Design/methodology/approachThe method relies on alternation of identification of the active genes using a mixture model and clustering of the samples based on Ward hierarchical clustering. The initial‐point of the procedure is obtained by means of a χ2 test. The method attempts to locally minimize the sum of the within cluster sample variances under a suitable Gaussian assumption on the distribution of data.FindingsThis paper illustrates the proposed methodology and its success by means of results from both simulated and real cDNA microarray data. The comparison of the results with those from a related known method demonstrates the superiority of the proposed approach.Research limitations/implicationsOnly empirical evidence of algorithm convergence is provided. Theoretical proof of algorithm convergence is an open issue.Practical implicationsThe proposed methodology can be applied to perform cDNA microarray data analysis.Originality/valueThis paper provides a contribution to the development of successful statistical methods for cDNA microarray data analysis.

Genes that code for T cell signaling proteins establish transcriptional regulatory networks during thymus ontogeny

Gene expression profiling by cDNA microarrays during murine thymus ontogeny has contributed to dissecting the large-scale molecular genetics of T cell maturation. Gene profiling, although useful for characterizing the thymus developmental phases and identifying the differentially expressed genes, does not permit the determination of possible interactions between genes. In order to reconstruct genetic interactions, on RNA level, within thymocyte differentiation, a pair of microarrays containing a total of 1,576 cDNA sequences derived from the IMAGE MTB library was applied on samples of developing thymuses (14-17 days of gestation). The data were analyzed using the GeneNetwork program. Genes that were previously identified as differentially expressed during thymus ontogeny showed their relationships with several other genes. The present method provided the detection of gene nodes coding for proteins implicated in the calcium signaling pathway, such as Prrg2 and Stxbp3, and in protein transport toward the cell membrane, such as Gosr2. The results demonstrate the feasibility of reconstructing networks based on cDNA microarray gene expression determinations, contributing to a clearer understanding of the complex interactions between genes involved in thymus/thymocyte development.

Alteration of apoptotic signaling molecules as a function of time after radiation in human neuroblastoma cells

Ascertaining the time-dependent regulation of induced apoptosis and radioresistance is important to understand the relationship between the level of spontaneous apoptosis in cells and their radiosensitivity. Accordingly, we investigated the time-dependent expression of apoptosis related genes and radioresistance in neuroblastoma cells. Serum-starved human SK-N-MC cells were exposed to low linear energy transfer (LET) radiation (2 Gy) and incubated for 15, 30, 45 min, and 48 h. Radioresistance was investigated by examining the NF kappa B DNA-binding activity, cellular toxicity, DNA fragmentation, and expression of apoptotic signal transduction molecules. NF kappa B DNA binding activity was analyzed using electrophoretic mobility shift assay (EMSA). Cellular toxicity was measured using MTT assay. DNA fragmentation was quantified by labeling with fluorescein-conjugated deoxynucleotides. Microarray analysis was performed using cDNA microarray and relative gene expression was measured as % GAPDH and, subsequently validated using Q-PCR. Induction of NF kappa B analyzed using EMSA showed an increased DNA-binding activity at all time points investigated. Induced DNA fragmentation was observed after 15, 30, and 45 min post-radiation. Relatively, induced fragmentation was reduced after 48 h. Compared to the untreated controls cellular toxicity was induced with low LET radiation after 15, 30, and 45 min. Conversely, cytotoxicity was relatively less at 48 h after low LET radiation. Microarray analysis after low LET radiation revealed time-dependent modulation of apoptosis-related genes that are involved in radio-adaptation, spontaneous apoptosis-related early-responsive genes and late response genes. These results suggest that the time-dependent regulation of apoptotic response may determine the relationship between the level of spontaneous apoptosis in cells and their radiosensitivity.

Disruption of the latent transforming growth factor-β binding protein-1 gene causes alteration in facial structure and influences TGF-β bioavailability

Latent transforming growth factor-β binding proteins are a family of extracellular matrix proteins comprising four isoforms (LTBP-1, -2, -3, -4) with different structures, tissue expression patterns and affinity for TGF-β. So far, respective knockout models have highlighted some essential functions for LTBP-2, LTBP-3 and LTBP-4, while the physiological significance of LTBP-1 is only superficially known. Here we report for the first time the generation and characterization of a mouse model lacking both the long and short LTBP-1 isoform. Surprisingly, respective mice are viable and fertile. However, detailed X-ray analysis of the skull revealed a modified facial profile. In addition, the gene disruption induces a reduced biological activity of TGF-β that became evident in an experimental model of hepatic fibrogenesis in which the LTBP-1 knockout animals were less prone to hepatic fibrogenesis. Furthermore, comparative cDNA microarray gene expression profiling of cultured hepatic stellate cells confirmed that respective nulls were less receptive to cellular activation and transdifferentiation into myofibroblasts. Therefore, we conclude that LTBP-1 has essential functions in the control of TGF-β activation.

Transplant reno‐vascular stenoses associated with early erythropoeitin use

This report describes an unusual presentation of severe hypertension (HTN) in a subset of pediatric kidney recipients treated with a steroid avoidance pediatric renal transplantation protocol. The HTN was secondary to atypical, reno-vascular abnormalities (RVA) of the transplanted vasculature, temporally associated with erythropoietin (EPO) use. To investigate the clinical significance underlying this event, a retrospective clinical study of 100 pediatric renal transplants was undertaken (50 steroid-free and 50 matched steroid-based controls), with peripheral blood transcriptional analysis of four RVA patients and controls. Regardless of a higher observed incidence of anemia (p < 0.001) and greater overall EPO usage in the first post-transplant year in steroid-free patients, the incidence of new-onset HTN at one yr was significantly less in the steroid-free cohort (p = 0.03). Nevertheless, early EPO (first week post-transplant) was significantly associated with the combinatory findings of new-onset HTN (p = 0.03) and RVA (p = 0.007). Molecular mechanisms of RVA injury were investigated further by peripheral blood cDNA microarray gene expression profiling. A panel of 42 transcripts differentiated patients with RVA and HTN from three sets of matched controls, with and without HTN and EPO use, with 100% concordance (p < 0.001). The biological processes governed by these significant genes suggest a role for EPO regulation of growth factor receptor ubiquitination as a putative mechanism for renal vascular injury. This study cautions against the use early post-transplant use of EPO in immunosuppression regimens with steroid minimization/avoidance, which may have an increased incidence of post-transplant anemia.

Diagnosis of the Small Round Blue Cell Tumors Using Multiplex Polymerase Chain Reaction

The small round blue cell tumors of childhood, which include neuroblastoma, rhabdomyosarcoma, non-Hodgkin's lymphoma, and the Ewing's family of tumors, are so called because of their similar appearance on routine histology. Using cDNA microarray gene expression profiles and artificial neural networks (ANNs), we previously identified 93 genes capable of diagnosing these cancers. Using a subset of these, together with some additional genes (total 39), we developed a multiplex polymerase chain reaction (PCR) assay to diagnose these cancer types. Blinded testing of 96 new samples (26 Ewing's family of tumors, 29 rhabdomyosarcomas, 24 neuroblastomas, and 17 lymphomas) using ANNs in a complete leave-one-out analysis demonstrated that all except one sample were accurately diagnosed as their respective category. Moreover, using an ANN-based gene minimization strategy in a separate analysis, we found that the top 31 genes could correctly diagnose all 96 tumors. Our results suggest that this molecular test based on a multiplex PCR reaction may assist the physician in the rapid confirmation of the diagnosis of these cancers.

Coronary Artery Aneurysm Induced by Kawasaki Disease in Children Show Features Typical Senescence

Kawasaki disease (KD) causes coronary artery disease (CAD) in children. In addition, a history of KD is suspected to be a risk factor for the development of atherosclerotic heart disease in the future. Histological senescence changes are a common denominator in atherosclerotic lesions in adults, so the present study investigated whether histological senescence changes had already occurred in KD aneurysm. KD coronary aneurysms and internal mammary arteries retrieved from 5 children with KD (3, 4, 5, 6, and 11 years old, respectively) who underwent coronary artery bypass grafting, as well as giant coronary aneurysm size-reducing operations, were analyzed. Senescence-associated strong beta-galactosidase activity was observed in KD aneurysms, but not in the internal mammary arteries. An immunohistochemical analysis of the KD aneurysm using anti-CD31, anti-endothelial nitric oxide synthetase (eNOS), anti-vascular adhesion molecule-1 (VCAM-1), and anti-monocyte chemoattractant protein-1 (MCP-1) showed vascular endothelium CD31 staining, decreased staining of eNOS and strong staining of MCP-1 and VCAM-1. cDNA microarray gene expression profiling revealed increased MCP-1 expression in the KD aneurysm, a finding confirmed by quantitative polymerase chain reaction. Histological features of senescence and active remodeling gene expression show that the KD aneurysm is not a silent vasculitis terminal. The future fate of KD aneurysms, including atherosclerosis, should be monitored carefully.

Novel blood biomarkers of human urinary bladder cancer: Osman I, Bajorin DF, Sun T-T, Zhong H, Douglas D, Scattergood J, Zheng R, Han M, Marshall KW, Liew CC, Wayne Marshall and Choong-Chin Liew. New York University School of Medicine, New York, NY

Recent data indicate that cDNA microarray gene expression profile of blood cells can reflect disease states and thus have diagnostic value. We tested the hypothesis that blood cell gene expression can differentiate between bladder cancer and other genitourinary cancers as well as between bladder cancer and healthy controls. We used Affymetrix U133 Plus 2.0 GeneChip (Affymetrix, Santa Clara, CA) to profile circulating blood total RNA from 35 patients diagnosed with one of three types of genitourinary cancer [bladder cancer ( n = 16), testicular cancer ( n = 10), and renal cell carcinoma ( n = 9)] and compared their cDNA profiles with those of 10 healthy subjects. We then verified the expression levels of selected genes from the Affymetrix results in a larger number of bladder cancer patients ( n = 40) and healthy controls ( n = 27). Blood gene expression profiles distinguished bladder cancer patients from healthy controls and from testicular and renal cancer patients. Differential expression of a combined set of seven gene transcripts (insulin-like growth factor-binding protein 7, sorting nexin 16, chondroitin sulfate proteoglycan 6, and cathepsin D, chromodomain helicase DNA-binding protein 2, nell-like 2, and tumor necrosis factor receptor superfamily member 7) was able to discriminate bladder cancer from control samples with a sensitivity of 83% (95% confidence interval, 67–93%) and a specificity of 93% (95% confidence interval, 76–99%). We have shown that the gene expression profile of circulating blood cells can distinguish bladder cancer from other types of genitourinary cancer and healthy controls and can be used to identify novel blood markers for bladder.

Novel Blood Biomarkers of Human Urinary Bladder Cancer

Recent data indicate that cDNA microarray gene expression profile of blood cells can reflect disease states and thus have diagnostic value. We tested the hypothesis that blood cell gene expression can differentiate between bladder cancer and other genitourinary cancers as well as between bladder cancer and healthy controls. We used Affymetrix U133 Plus 2.0 GeneChip (Affymetrix, Santa Clara, CA) to profile circulating blood total RNA from 35 patients diagnosed with one of three types of genitourinary cancer [bladder cancer (n = 16), testicular cancer (n = 10), and renal cell carcinoma (n = 9)] and compared their cDNA profiles with those of 10 healthy subjects. We then verified the expression levels of selected genes from the Affymetrix results in a larger number of bladder cancer patients (n = 40) and healthy controls (n = 27). Blood gene expression profiles distinguished bladder cancer patients from healthy controls and from testicular and renal cancer patients. Differential expression of a combined set of seven gene transcripts (insulin-like growth factor-binding protein 7, sorting nexin 16, chondroitin sulfate proteoglycan 6, and cathepsin D, chromodomain helicase DNA-binding protein 2, nell-like 2, and tumor necrosis factor receptor superfamily member 7) was able to discriminate bladder cancer from control samples with a sensitivity of 83% (95% confidence interval, 67-93%) and a specificity of 93% (95% confidence interval, 76-99%). We have shown that the gene expression profile of circulating blood cells can distinguish bladder cancer from other types of genitourinary cancer and healthy controls and can be used to identify novel blood markers for bladder cancer.

Qualitative assessment of cDNA microarray gene expression data using detrended fluctuation analysis

A number of microarray studies assume gene expression data to be independent of one another. In this report, we provide evidence of correlation in cDNA microarray gene expression data using classical power spectral analysis and the sophisticated detrended fluctuation analysis (DFA). Such correlations are shown to be an outcome of gene's position on the arrays and immune to pre-processing procedures such as normalization. The results presented encourage DFA as a tool for qualitative assessment of microarray gene expression data prior to inferring differential gene expression.

Effect of Fenoldopam on Ischemia/Reperfusion-Induced Apoptosis

Recently we demonstrated the effect of fenoldopam on ischemia/reperfusion (I/R) induced NFκB mediated pro-inflammatory signal transduction. However, the effect of fenoldopam on I/R-induced apoptosis is not known. We utilized a rat model of acute ischemic nephropathy to test the hypothesis that fenoldopam attenuates I/R-induced apoptosis. Sprague-Dawley rats were anesthetized by intraperitoneal administration of 50 mg/kg urethane and randomly allocated into 4 groups (n = 6 each): 1) sham-operated, 2) sham operation with infusion of 0.1 μg/kg/min fenoldopam, 3) unilateral renal ischemia followed by 4 h of reperfusion, and 4) I/R with fenoldopam infusion. Kidney samples were fixed and paraffin-embedded to measure apoptosis. Data were compared between groups using ANOVA with Bonferroni correction. RNA was extracted from each left kidney to probe cDNA microarray and measure gene expression as percent of positive control. Compared to the control group, I/R significantly (P < 0.001) induced apoptosis in both the cortex and medulla. Similarly, microarray analysis revealed that I/R induced 73 apoptosis-related genes. Treatment with fenoldopam significantly reduced (P < 0.001) I/R-induced apoptosis both in the cortex and medulla and attenuated all 73 I/R-induced apoptosis-related genes. Data from this rat model of ischemic nephropathy suggest that fenoldopam may attenuate I/R-induced apoptosis and apoptosis-related gene transcription.

Simplifying Complex Microarray Data To Derive Gene Expression Profiles Which Identify Childhood Acute Lymphoblastic Leukaemia Patients at Risk of Relapse.

The optimal treatment of patients with childhood acute lymphoblastic leukaemia (ALL) depends on establishing accurate diagnosis. Our investigations seek to strategically develop the application of microarray gene expression profiling to identify ALL patients with clinically homogenous presentations but which may respond differently to established treatment regimens. We have determined the gene expression profiles of ALL bone marrow (BM) samples taken from patients at diagnosis. Data analysis has focussed on the use of a novel and innovative statistical technology, Gene-RaVE. This series of patent protected algorithms builds a multinomial regression model using Bayesian variable selection. Gene-RaVE leads to the generation of a parsimonious and simple diagnostic gene expression signature, but which provides increased predictive ability over current analysis approaches. We describe our analysis of both Affymetrix (HU133A) and cDNA (10.5K) microarray gene expression profiles generated from diagnostic BM from >100 ALL patients covering the broad ALL subtypes including T and B lineage as well as T cell lymphoma leukaemia. Comparison of gene expression data failed to identify clearly distinguishing profiles between patients identified as ‘standard risk' from ‘medium risk' according to BFM95 clinical criteria. Gene expression profiles from a cohort of ALL patients, identified as ‘standard risk' at diagnosis, were compared on the basis of their overall clinical outcome: relapse within 2 yrs vs non-relapse. Using a range of analyses including t-test, Gene-RaVE, discriminant analysis approaches and principle component analysis, we have discovered that small subsets of genes (<10), all of which included Nedd4BP3 and Ribosomal Protein L38 (RPL38), can be used to distinguish the two outcome groups. Subsequent validation using real time PCR supports the increase in Nedd4BP3 expression in standard risk patients which do not respond well to established treatment regimens. The Gene-RaVE algorithm also provides a generic framework for survival analysis. This approach indicates that the expression of these Nedd4BP3, RPL38 and inositol 1, 4, 5-triphosphate receptor, type 2 can be used to build a survival ‘index' which correlates with the time to a relapse event in standard risk childhood ALL patients. Our results are suggestive of a way forward in the development of an informative, yet efficient diagnostic tool for this childhood malignancy using microarray gene expression analysis technology.

An evaluation of linear RNA amplification in cDNA microarray gene expression analysis

DNA microarray is becoming a common tool across a broad range of disciplines, particularly in the basic and clinical biomedical sciences. However, the limitation of this technology is the requirement for relatively large amount of RNA for labeling and hybridization. T7-based linear amplification technique can overcome this limitation and enables consumption of very low amount of samples. In this study, we utilized a low RNA input fluorescent linear amplification kit from Agilent to amplify 0.2 μg total RNA and compared the gene expression profiles generated from amplified aRNA and total RNA. Our results demonstrated that nanograms total RNA can be amplified reproducibly with this protocol and can generate gene expression profiles comparable with unamplified total RNA. Furthermore, quantitative real-time PCR measurements of a set of 10 genes showed good correlation with amplified aRNA arrays.

The association of Id-1, MIF and GSTpi with acquired drug resistance in hormone independent prostate cancer cells

Most patients with prostate cancer respond to androgen ablation therapy, but the tumor eventually progresses to an androgen-independent stage with multiple anti-cancer drug resistance. Novel therapeutic strategies for hormone independent prostate cancer need to be developed. The genes for Id-1, MIF and GSTpi were up-regulated in drug resistant cells of hormone independent prostate cancer cells using cDNA microarray gene expression analysis. In this study we aimed to investigate whether constitutive overexpression of these candidate genes in cells can affect cellular resistance to chemotherapeutic agents. The mRNA expression was determined by reverse transcriptase-polymerase chain reaction, and the Id-1, MIF and GSTpi protein contents in these cell lines were measured by Western blotting. Susceptibility of the cells to chemotherapeutic agents including doxorubicin, paclitaxel and cyclophosphamide was determined by microculture tetrazolium bromide assay. The analysis of apoptosis was assessed by annexin V assay. The cytotoxity assay results demonstrated that cells overexpressing the Id-1 gene increased in their resistance to doxorubicin, paclitaxel and cyclophosphamide. MIF expression can drive cells to increase in their resistance to paclitaxel, and GSTpi expression confers drug resistance to doxorubicin and cyclophosphamide. The induction of mdr1 gene expression was noted in up-regulation of MIF and GSTpi. Our findings suggest that overproduction of the Id-1, MIF and GSTpi proteins and coinduction of mdr-1 play an important role in the development of acquired drug resistance to various drugs of prostate cancer cells.

Discovery of 342 putative new genes from the analysis of 5′-end-sequenced full-length-enriched cDNA human transcripts

In this work we describe the process that, starting with the production of human full-length-enriched cDNA libraries using the CAP-Trapper method, led us to the discovery of 342 putative new human genes. Twenty-three thousand full-length-enriched clones, obtained from various cell lines and tissues in different developmental stages, were 5′-end sequenced, allowing the identification of a pool of 5300 unique cDNAs. By comparing these sequences to various human and vertebrate nucleotide databases we found that about 40% of our clones extended previously annotated 5′ ends, 662 clones were likely to represent splice variants of known genes, and finally 342 clones remained unknown, with no or poor functional annotation. cDNA-microarray gene expression analysis showed that 260 of 342 unknown clones are expressed in at least one cell line and/or tissue. Further analysis of their sequences and the corresponding genomic locations allowed us to conclude that most of them represent potential novel genes, with only a small fraction having protein-coding potential.

Transcript AA454543 Is a Novel Prognostic Marker for Hepatocellular Carcinoma after Curative Partial Hepatectomy

BACKGROUND: We have previously reported on the cDNA microarray gene expression profiles of hepatocellular carcinomas (HCCs). Among the genes that show prognostic significance and are overexpressed in tumor compared with adjacent nontumorous liver, transcript AA454543 may have potential for practical use. Our aim is to validate the prognostic significance of transcript AA454543 by alternative research methods and in a separate group of HCC patients. METHODS AND RESULTS: The data of transcript AA454543 derived from microarray analysis of 48 patients having curative partial hepatectomy (group 1) were verified by quantitative reverse transcription polymerase chain reaction (r = 0.618, P < .001). A separate sample set of HCCs obtained from 53 patients (group 2) was examined and the association of AA454543 expression level with overall survival was again validated (P = .027). By Cox regression analysis, transcript AA454543 [hazard ratio (HR) = 3.0, P =.01 7] and pathologic tumor node metastasis (pTNM) stage (HR = 3.3, P = .010) were independent prognostic factors for overall survival. The accuracy of prediction for 3-year overall survival for transcript AA454543 (74.2%, P = .001) and pTNM stage (76.4%, P = .001) was comparable as measured by the area under the receiver operating characteristic curve. CONCLUSION: Transcript AA454543 is a potentially useful molecular prognostic marker for overall survival after curative partial hepatectomy for HCC.

Claudin-10 Expression Level is Associated with Recurrence of Primary Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) patients with the same clinicopathologic features can have remarkably different disease outcomes after curative hepatectomy. To address this issue, we evaluated the cDNA microarray gene expression profiles of HCCs and identified claudin-10 expression level was associated with disease recurrence. The aim of the current study is to validate the microarray data by an alternative research method applicable for routine practice. Quantitative reverse transcription-PCR (RT-PCR) was used to validate the microarray data on claudin-10 expression level. The assay was repeated on a separate HCC sample set to consolidate the prognostic significance of claudin-10. Claudin-10 expression level by quantitative RT-PCR and by microarray measurement showed a high concordance (r = 0.602, P < 0.001). Quantitative RT-PCR was repeated on a separate HCC sample set and the association of claudin-10 expression with recurrence was again confirmed (hazard ratio, 1.2; 95% confidence interval, 1.0-1.4; P = 0.011). By multivariable Cox regression analysis, claudin-10 expression and pathologic tumor-node-metastasis stage were independent factors for prediction of disease recurrence. Claudin-10 expression of HCC can be used as a molecular marker for disease recurrence after curative hepatectomy.

Modulation of gene expression in MHCC97 cells by interferon alpha

To elucidate the molecular mechanisms of the inhibitory effects of IFN-alpha on tumor growth and metastasis in MHCC97 xenografts. Three thousand international units per milliliter of IFN-alpha-treated and -untreated MHCC97 cells were enrolled for gene expression analysis using cDNA microarray. The mRNA levels of several differentially expressed genes in cDNA microarray were further identified by Northern blot and RT-PCR. A total of 190 differentially expressed genes including 151 IFN-alpha-repressed and 39-stimulated genes or expressed sequence tags from 8,464 known human genes were found to be regulated by IFN-alpha in MHCC97. With a few exceptions, mRNA levels of the selected genes in RT-PCR and Northern blot were in good agreement with those in cDNA microarray. IFN-alpha might exert its complicated anti-tumor effects on MHCC97 xenografts by regulating the expression of functional genes involved in cell metabolism, proliferation, morphogenesis, angiogenesis, and signaling.

Hierarchical Bayes models for cDNA microarray gene expression

cDNA microarrays are used in many contexts to compare mRNA levels between samples of cells. Microarray experiments typically give us expression measurements on 1000-20 000 genes, but with few replicates for each gene. Traditional methods using means and standard deviations to detect differential expression are not satisfactory in this context. A handful of alternative statistics have been developed, including several empirical Bayes methods. In the present paper we present two full hierarchical Bayes models for detecting gene expression, of which one (D) describes our microarray data very well. We also compare the full Bayes and empirical Bayes approaches with respect to model assumptions, false discovery rates and computer running time. The proposed models are compared to existing empirical Bayes models in a simulation study and for a set of data (Yuen et al., 2002), where 27 genes have been categorized by quantitative real-time PCR. It turns out that the existing empirical Bayes methods have at least as good performance as the full Bayes ones.

Differential expression of the Tie-2 receptor and its ligands in human pancreatic tumors

Derangements in the balance of the Tie-2 receptor ligands, angiopoietin-1 and angiopoietin-2 (Ang-1 and Ang-2), have been implicated in the growth and differentiation of several human tumors. To assess the potential role of angiogenesis factors in pancreatic tumorigenesis, we confirmed previously observed oligonucleotide-based microarray data by analyzing their expression in pancreatic tumors using semiquantitative reverse transcription polymerase chain reaction and immunohistochemistry. RNA harvested from tissue samples obtained from patients with normal pancreata, adenocarcinoma of the pancreas, and neuroendocrine tumors of the pancreas (nine each) was grouped by tissue type, pooled, and hybridized to a cDNA microarray and global gene expression patterns were determined. Semiquantitative reverse transcription polymerase chain reaction was applied to individual pool components to specifically determine expression of Tie-2, Ang-1, and Ang-2. A separate set of tumors and matched normal pancreas were then stained for expression of Ang-2 protein using immunohistochemistry. Microarray data showed that Ang-2 was upregulated in neuroendocrine tumors ( approximately 8 times more than normal) and adenocarcinoma of the pancreas ( approximately 5 times more than normal) although Ang-1 and Tie-2 were not differentially expressed. Reverse transcription polymerase chain reaction confirmed these findings showing strong Ang-2 expression in nine of nine neuroendocrine tumors, eight of nine adenocarcinoma of the pancreas, and weak expression in two of nine normal pancreata. Immunohistochemistry showed tumor-specific staining of Ang-2 in 10 of 15 matched adenocarcinoma of the pancreas (67%) and 6 of 8 neuroendocrine tumors (75%). Although Tie-2 and Ang-1 are not differentially expressed in pancreatic tumors, Ang-2 gene and gene product are overexpressed, suggesting a significant role for Ang-2 in pancreatic tumorigenesis. Differential expression of Ang-2 in human tumors may be useful as a diagnostic and therapeutic target.