Tissue microarrays for high-throughput molecular profiling of tumor specimens

Tissue microarrays: a new approach for quality control in immunohistochemistry

AIMS

To improve the interpretation of immunohistochemistry (IHC) staining results the use of a tissue microarray technique was established in a routine setting.

METHODS

A tissue microarray was constructed by harvesting 600 microm tissue cores from paraffin wax embedded samples available in a routine pathology department. The punches originating from non-tumorous tissue were placed on host paraffin wax blocks. The microarray contained 12 different tissue samples, with a wide antigen profile and a dimension of 3.5 x 3 mm. One section of the multitissue array was placed as an "internal" positive control on each slide of the patient tissue to undergo identical immunohistochemical procedures.

RESULTS

Using the tissue microarray technique as a tool for internal quality control, the interpretation of immunohistochemical staining of more than 20 different antigens in routine IHC was improved. The tissue microarray did not influence the staining results in conventional IHC or in different automated IHC settings.

CONCLUSION

The regular use of an institution adapted tissue microarray would be useful for internal positive control in IHC to enable different laboratory demands. Furthermore, this technique improves the evaluation of staining results in IHC.

Microarrays as cancer keys: an array of possibilities

Malignant transformation results from accumulation of genetic and epigenetic events. Functional studies of cancer will be crucial to our understanding of its complexity and polymorphism. There is no doubt that emerging genomic and proteomic technologies will facilitate such investigations. Microarray technology is a new and efficient approach to extract data of biomedical relevance for a wide range of applications. In cancer research, it will provide high-throughput and valuable insights into differences in an individual's tumor as compared with constitutional DNA, mRNA expression, and protein expression and activity. Across individuals, comparisons could provide tissue-specific disease signatures that provide diagnosis based on hundreds of informative genes. The resulting product should be a wealth of tumor-associated and tumor-specific biomarkers, which may help in cancer etiology, diagnosis, and therapy and ultimately lead to "molecular nosology" of cancers. This review highlights the recent developments in microarray technologies in cancer research, focuses on the results obtained so far, and describes the eventual use of microarray technology for clinical applications.

Expression profiling of microdissected pancreatic adenocarcinomas

Pancreatic ductal adenocarcinoma is characterized by a paucity of neoplastic cells embedded in a densely desmoplastic stroma. Therefore, laser capture microdissection was performed to obtain homogeneous populations of normal and neoplastic ductal cells. These were subjected to a comparative study of gene expression utilizing human cDNA arrays. A variety of dysregulated genes were identified, comprising cell cycle and growth regulators, invasion regulators, signalling and developmental molecules. In addition to genes already found to be overexpressed in pancreatic cancer, such as TIMP1, MMP7, CD59, rhoC and NDKA, we present evidence to implicate genes which have not previously been reported in this tumour type. These include the overexpressed genes ABL2, Notch4 and SOD1, as well as XRCC1, a DNA repair gene whose transcript was found downregulated. Quantitative real-time RT-PCR (QRT-PCR) was employed to confirm differential expression of ABL2, Notch4 and SOD1 and immunohistochemical analysis was used to verify decreased protein expression of XRCC1 using a custom-built pancreatic tissue array. Combining microarray-derived gene expression profiles of pure pancreatic cell populations, QRT-PCR and pancreas-specific tissue arrays therefore proved to be highly informative in elucidating the molecular pathology of this highly malignant tumour type.

Biomedical applications of protein chips

The development of microchips involving proteins has accelerated within the past few years. Although DNA chip technologies formed the precedent, many different strategies and technologies have been used because proteins are inherently a more complex type of molecule. This review covers the various biomedical applications of protein chips in diagnostics, drug screening and testing, disease monitoring, drug discovery (proteomics), and medical research. The proteomics and drug discovery section is further subdivided to cover drug discovery tools (on-chip separations, expression profiling, and antibody arrays), molecular interactions and signaling pathways, the identification of protein function, and the identification of novel therapeutic compounds. Although largely focused on protein chips, this review includes chips involving cells and tissues as a logical extension of the type of data that can be generated from these microchips.

High-throughput copy number analysis of 17q23 in 3520 tissue specimens by fluorescence in situ hybridization to tissue microarrays

The chromosomal region 17q23 has been shown to be commonly amplified in breast tumors, especially those with poor prognosis. In addition to breast cancer, studies by comparative genomic hybridization have implicated the involvement of 17q23 in other tumor types as well. Here we performed a large-scale survey on the distribution and frequency of the 17q23 copy number increases across different tumor types using fluorescence in situ hybridization on tissue microarrays containing 4788 specimens. A total of 4429 tumor samples representing 166 different tumor categories and 359 normal tissue samples from 40 different tissue categories were analyzed. Successful hybridizations were observed in 3520 of the 4788 specimens (74%). Increased 17q23 copy number was detected in 15% of the evaluable specimens with tumors originating from the lung, mammary gland, and soft tissue being most frequently affected. Interestingly, high-level amplification was detected only in 2% of the tumors and was generally restricted to mammary tumors. In addition, we observed an association between the frequency of increased 17q23 copy number and tumor progression in various tumor types. These results indicate that increased 17q23 copy number occurs frequently in several different tumor types suggesting that increased dosage of genes in this region might play a role in development and progression of many tumor types.

Inadequate formalin fixation decreases reliability of p27 immunohistochemical staining: probing optimal fixation time using high-density tissue microarrays

Immunohistochemical analysis of molecular targets in clinical tissues is increasingly becoming central to our ability to render diagnoses, to predict prognosis, to select patients for appropriate therapies, and to provide surrogate end points for therapeutic monitoring. For example, reduction of immunohistochemical staining for the cyclin-dependent kinase inhibitor p27(Kip1) has been proposed as a potential prognostic biomarker in prostate, breast, and gastrointestinal tumors. We observed that with our standard formalin fixation in rapidly processed (same-day) radical prostatectomy specimens, there is often a gradient of p27(Kip1) staining in normal prostate epithelium, with more staining near the periphery and less staining toward the center of the sample. This raised the hypothesis that the reliability of staining for p27(Kip1) is decreased in inadequately fixed tissues. The implications of this, if true, are that many studies using p27(Kip1) for prognostic purposes may be subject to unpredictable artifacts, and hence unreliable results, if the fixation of the specimens is not well controlled. The objectives of the present study were (1) to formally test the hypothesis that inadequate fixation time is responsible for apparent loss of p27(Kip1) nuclear staining and (2) to test a recently proposed method for improving the uniformity of immmunohistochemical staining using formalin injection. Prostate tissue sections from radical prostatectomy specimens were either processed immediately (zero time fixation) or fixed for 1, 2, 3, or 8 days in 10% neutral buffered formalin before processing into paraffin. To assure identical antigen retrieval and immunohistochemical staining conditions for specimens fixed for different lengths of time, 2 high-density tissue microarrays (TMAs), containing 564 tissue samples (0.6 mm in diameter) were constructed. Based on an estimate of the percentage of nuclei in normal prostatic epithelial secretory cells with strong staining, quality of p27(Kip1) staining was graded in a blinded fashion with respect to fixation time. There was a significant increase in the percentage of cores that were scored as "strong" as fixation time increased from 0 (same-day processing) to 1 or more days (P <.0001). Interestingly, even at 8 days of fixation, there was excellent staining that was superior to the same-day processing. Based on these results, we conclude the following: (1) for large clinical specimens that have been fixed briefly to decrease diagnostic turn-around time, the reliability of interpretation of immunohistochemical staining may be quite limited; (2) for p27(Kip1), decreased antigen staining as a result of the widely held concept of "overfixation" is much less of a problem than "underfixation"; (3) formalin injection produces a marked improvement in staining for several markers, including p27(Kip1); and (4) high-density TMAs, which assure identical test conditions, provide an excellent platform on which to evaluate the effects of tissue fixation on immunohistochemical staining.

Expression of multidrug resistance-associated protein 2 (MRP2) in normal human tissues and carcinomas using tissue microarrays

AIMS

Using tissue microarrays, this study analysed the expression of the multidrug resistance protein, MRP2, by immunohistochemistry with two different MRP2 antibodies. This is the first study to address the expression of MRP2 in various common human neoplasms.

METHODS AND RESULTS

Immunohistochemistry was performed on zinc formalin-fixed tissue to evaluate normal tissues and carcinomas using two antibodies against MRP2 (EAG5, a polyclonal antibody, and M2-lll-6, a monoclonal antibody). Immunostaining was localized in neoplastic cells mainly on the cell membrane with M2-lll-6 and cell membrane and cytoplasm with EAG5. In normal tissues MRP2 was expressed in liver, gastrointestinal tract, and kidney tubular epithelial cells with both antibodies. MRP2 was seen in nine of 22 renal cell carcinomas, eight of 13 gastric carcinomas, 25 of 49 breast carcinomas, 14 of 32 lung carcinomas, 39 of 50 colon carcinomas, and 16 of 17 ovarian carcinomas. There was < 10% variability between the two antibodies. MRP2 expression was highest in moderate to poorly differentiated tumours from colon, lung, gastric, and ovarian carcinomas and in grade 2 and 3 breast and renal carcinomas.

CONCLUSION

The expression of MRP2 in many solid human tumours indicates that inherent drug resistance may play an important role as a biomarker for predictive chemotherapy treatment.

C-myc oncogene amplification in ductal carcinoma in situ of the breast

The c-myc oncogene is frequently activated in invasive breast cancer and has been associated with high nuclear grade, lymph node metastasis and poorer disease outcome. We have examined c-myc oncogene amplification using fluorescence in situ hybridization in a series of 96 pure DCIS. Additionally we assessed amplification and expression of the Her2 and bcl-2 oncogenes. The findings were compared with clinicopathological data, Ki-67 proliferative index and hormone receptor status. We observed c-myc oncogene amplification in 19 tumours (20%). These cases were significantly associated with an average of 38% higher proliferative activity (p = 0.045), a 43% larger tumour size (p = 0.029) and the otherwise rare micropapillary subtype (p = 0.0005). Concluding the c-myc oncogene appears to be involved in the development of a more aggressive phenotype of DCIS.

Molecular Markers in Prostate Cancer: The Role in Preoperative Staging

Radical prostatectomy as a primary treatment for clinically localized prostate cancer has increased dramatically over the past decade due to prostate-specific antigen (PSA) screening and the awareness of the increased incidence of localized disease. Despite the stage migration to increase clinically localized disease, there are still vast numbers of men who harbor occult extraprostatic extension and develop recurrence after surgery. The study of molecular markers in the blood or tissue of surgical patients prior to treatment, called " molecular staging, " is the focus of this review. The reverse transcriptase- polymerase chain reaction (RT-PCR) test for PSA gene expression in peripheral blood or bone marrow has received considerable attention since its first report in 1992. The test detects messenger RNA species for prostate-specific/abundant genes such as PSA and prostate-specific membrane antigen. These messenger RNAs were not detected in normal blood or bone marrow, but were detected in some prostate cancer patients presumably due to circulating prostatic epithelial cells. These prostate epithelial cells are thought to be occult metastases cells, and early studies correlated a positive RT-PCR test with surgical pathology adverse features such as positive margins. Despite the many studies over the past few years, there have been inconsistent results, and the most recent studies have not been able to confirm clinical utility. Bone marrow RT-PCR has been more promising; however, it is still a research tool that needs further study. The study of molecular markers in tissue material, ie, prostate biopsy samples prior to radical prostatectomy, is problematic due to the sampling error inherent in a multifocal heterogeneous tumor such as prostate cancer. The tumor suppressor proteins p53 and p27, Bcl-2 oncoprotein, Ki-67 proliferation index protein, E-cadherin, and microvessel density have been assessed in preradical prostatectomy needle biopsy. Results have been conflicting, and none are yet accepted as a clinically useful marker. Current and future work is focusing on analysis of multiple gene expressions or proteins simultaneously via gene chip or proteomics technology. While these expression profiles might be of value in whole prostate surgical specimens where tissues are well characterized, it is unclear how this new technology will be applied to the needle biopsy samples. Although molecular staging of radical prostatectomy patients has been under study for a decade, all assays remain research tools. Still, this area holds great promise for improving the accuracy of staging and providing a more accurate prognosis of individual men with clinically localized prostate cancer.

Impact of the Human Genome Project on the clinical management of sporadic cancers

The publication of the human genome sequence has provided a new basis for cancer research. Molecular analysis of single cancer genes in isolation may lead to an underestimation of the impact of networks of intertwined genes in molecular cancer pathology. However, new technologies such as DNA microarrays or microchips will enable the detection of global gene-expression profiles--already described for lymphomas, acute leukaemias, and various solid tumours--and may help to overcome some of the limitations of gene function analysis. In addition, DNA microchip data banks may uncover new genes that are relevant to the molecular pathology of specific cancers and trigger detailed analysis of their function. Clinically, microchips will enable us to identify new molecular cancer markers or marker profiles of prognostic and predictive value, since global gene-expression patterns can highlight molecular tumour characteristics that relate to clinically distinct entities within heterogeneous cancers, such as non-Hodgkin lymphomas or breast cancer. However, before its promise can be realised, all molecular information stemming from the Human Genome Project will need to be tested for its clinical relevance in appropriate cancer trials; this presents a formidable but important challenge.

Evaluation of the tissue microarray technique for immunohistochemical analysis in rectal cancer

BACKGROUND

Immunohistochemical staining for tumor-associated proteins is widely used for the identification of novel prognostic markers. Recently, a tissue-conserving, high-throughput technique, tissue microarray, has been introduced. This technique uses 0.6-mm tissue core biopsy specimens, 500 to 1000 of which are brought into a new paraffin array block, which can be sectioned up to 100 times.

METHODS

We evaluated the tissue microarray technique for immunohistochemical analysis in 20 rectal cancers. Immunohistochemical staining was performed for the proliferation marker Ki-67 and the tumor suppressor protein p53 in whole tissue sections and in tissue core biopsy specimens.

RESULTS

The whole tissue sections were assessed by counting all cells in 10 high-power fields (x40), which resulted in a mean fraction of Ki-67-expressing tumor cells of 0.81 (range, 0.54-1.0). p53 expression assessed in whole tissue sections showed nuclear staining in 15 (75%) of 20 rectal carcinomas. For the tissue microarray technique, a median of 3 (range, 3-5) 0.6-mm tissue core biopsy specimens were studied from each of the 20 tumor specimens. The tissue microarray method gave a mean Ki-67 expression of 0.85 (range, 0.50-1.0) in tumor cell nuclei and showed p53 protein expression in the same 15 of 20 tumors as in the whole tissue sections.

CONCLUSION

We conclude that the tissue microarray technique for immunohistochemical staining in rectal cancer yields staining of good quality and expression data for Ki-67 and p53 comparable to those obtained with whole tissue staining. The feasibility of tissue microarray thus enables time- and tissue-preserving studies of multiple markers in large tumor series.

Challenges in developing a molecular characterization of cancer

DNA microarrays are widely used to measure gene expression across thousands of genes in parallel. Recently, considerable efforts have been made to utilize this technology to improve our understanding of cancer and to identify novel diagnostic markers and therapeutic targets. Here, we detail some of the challenges in developing a molecular characterization of cancer and in translating these new discoveries towards clinical utility.

Amplification of PPM1D in human tumors abrogates p53 tumor-suppressor activity

Expression of oncogenic Ras in primary human cells activates p53, thereby protecting cells from transformation. We show that in Ras-expressing IMR-90 cells, p53 is phosphorylated at Ser33 and Ser46 by the p38 mitogen-activated protein kinase (MAPK). Activity of p38 MAPK is regulated by the p53-inducible phosphatase PPM1D, creating a potential feedback loop. Expression of oncogenic Ras suppresses PPM1D mRNA induction, leaving p53 phosphorylated at Ser33 and Ser46 and in an active state. Retrovirus-mediated overexpression of PPM1D reduced p53 phosphorylation at these sites, abrogated Ras-induced apoptosis and partially rescued cells from cell-cycle arrest. Inactivation of p38 MAPK (the product of Mapk14) in vivo by gene targeting or by PPM1D overexpression expedited tumor formation after injection of mouse embryo fibroblasts (MEFs) expressing E1A+Ras into nude mice. The gene encoding PPM1D (PPM1D, at 17q22/q23) is amplified in human breast-tumor cell lines and in approximately 11% of primary breast tumors, most of which harbor wildtype p53. These findings suggest that inactivation of the p38 MAPK through PPM1D overexpression resulting from PPM1D amplification contributes to the development of human cancers by suppressing p53 activation.

Issues and barriers to development of clinically useful tumor markers: a development pathway proposal

There are few tumor markers that are clinically useful in predicting therapeutic responses or patient outcomes despite nearly 20 years of advances in molecular biology. We discuss a variety of issues and barriers that have affected movement of clinical tests from research into clinical practice. Studies of new markers frequently lack clear hypotheses and are generally underpowered to reach statistically valid conclusions. Relevant clinical endpoints may not be possible to evaluate, often leading to suboptimal study designs. Major stumbling blocks exist because studies are rarely comparable. This makes it difficult to determine why results vary from study to study. It also prevents pooling of small datasets for analysis. We propose a tumor marker development pathway that we think will be more efficient and effective. The pathway depends on developing statistically valid study designs, focusing on assay refinement and standardization early in the process, including assay details in publications, and providing data in a format that allows comparison with other studies. The process described should be applicable to development of new technologies that include analysis and interpretation of large, complex datasets. The proposed marker development pathway will require thoughtful refinement and expansion, but it should begin a productive dialog.

Assessment of interlaboratory variation in the immunohistochemical determination of estrogen receptor status using a breast cancer tissue microarray

The determination of tumor cell estrogen receptor (ER) expression status by immunohistochemical analysis has become standard practice, yet assay reproducibility has not been assessed adequately. By using a breast cancer tissue microarray, we examined interlaboratory variability in ER reporting. A 2-fold redundant tissue microarray block was created from 29 breast cancers. Unstained slides were distributed to 5 laboratories, and each laboratory immunostained and scored 1 slide for ER. Interlaboratory agreement ranged from moderate to high (overall kappa = 0.54 for 0-3+ grading; overall kappa = 0.84 for negative vs positive assessment of ER status). When 1 observer scored each of the 5 slides, interlaboratory agreement was slightly better (kappa = 0.63 for 0-3+ scoring; kappa = 0.96 for negative vs positive scoring). One laboratory, which had used an antibody and antigen retrieval method different from the others, demonstrated only fair concordance with the other 4 laboratories, but there was substantial intralaboratory interobserver agreement and excellent agreement with an outside observer reviewing the slide stained in that laboratory. The tissue microarray was an efficient and effective tool for identifying variability in ER reporting and should prove valuable in other external quality assurance programs.

Amplification and overexpression of Elongin C gene discovered in prostate cancer by cDNA microarrays

Functional significance of several oncogenes is mediated by overexpression. To identify overexpressed genes in prostate cancer, we analyzed expression of 1081 transcripts in three prostate cancer cell lines (PC-3, DU145, and LNCaP) using cDNA microarray hybridization. The cDNA microarray analyses were validated by quantitative real-time RT-PCR. On average, 64% of the genes were expressed at detectable levels in the cell lines. Next, the expression profiles were combined with the data on DNA sequence copy number alterations in the cell lines obtained by comparative genomic hybridization. The genes for Elongin C and urokinase type plasminogen-activator, both located in the regions of amplification in the PC-3 cell line (8q21 and 10q22, respectively), were found to be overexpressed in the PC-3. Amplification and overexpression of urokinase type plasminogen-activator in prostate cancer has previously been reported. Here, fluorescence in situ hybridization on tissue microarray showed high-level amplification of the Elongin C gene in 8 (23%) of 35 hormone-refractory carcinomas but in none of the untreated prostate carcinomas (n = 35). Finally, it was shown that the Elongin C gene was overexpressed and amplified also in breast cancer cell line SK-Br-3. The results indicate that Elongin C is a putative target gene for 8q amplification.

Bacterial genomics: the use of DNA microarrays and bacterial artificial chromosomes

Immense amounts of genetic information are contained within microbial genomes. As the number of completely sequenced microbial genomes is increasing, functional and comparative genomic techniques will be employed for sequence analysis and gene characterization. Sequence comparison and expression profiling by DNA microarrays can determine phylogenetic relationships and identify genes while bacterial artificial chromosomes (BACs) allow the study of entire biochemical pathways and permit the expression of bacterial genes in a foreign host.

Sol-gel encapsulated enzyme arrays for high-throughput screening of biocatalytic activity

We developed versatile low-cost arrays of sol-gel-encapsulated enzymes (referred to as solzymes) suitable for repeated assays of bioactivity or enzyme inhibition. Sol-gel microstructures containing active enzymes were stabilized on glass at moderate pH and room temperature without harsh calcination. A multi-well bilayer of polydimethylsiloxane was used to support the solzyme array and contain the reaction medium. Each of the 147 microwells has a working volume of 5 muL and contains 50 mug of immobilized enzyme. The solzyme arrays maintained high activity through repeated applications and exhibited superior thermostability compared to soluble enzymes. Among the enzymes used were lipases, glucose oxidase, and horseradish peroxidase. Twenty different lipases and proteases were also used to prepare a hydrolase array, for which bromthymol blue served as a generic indicator of activity. The relative activities of the encapsulated hydrolases correlated closely with those of the soluble hydrolases, illustrating that sol-gel encapsulation preserved the hierarchy of enzyme activity. The development of solzyme arrays paves the way to higher throughput screening of diverse proteins and enzymes, including those that are available only in trace amounts.

Amplification pattern of 12q13-q15 genes (MDM2, CDK4, GLI) in urinary bladder cancer

The chromosomal region 12q13-q15 is recurrently amplified in bladder cancer. Putative target genes located in this region include MDM2, CDK4, and GLI. To evaluate the involvement of these genes in bladder cancer, we screened a tissue microarray (TMA) containing 2317 samples by fluorescence in situ hybridization (FISH). Amplification was found for MDM2 in 5.1%, for CDK4 in 1.1%, and for GLI in 0.4% of interpretable tumors. Among tumors having amplification of at least one of these 12q13-q15 genes, 76.6% had amplification of MDM2 alone and 6.4% had amplification of CDK4 alone. Coamplifications were seen of MDM2 and CDK4 in 10.6%, and of CDK4 and GLI in 6.4%. Neither coamplifications of all three genes nor isolated GLI amplifications were found. These data suggest a prominent role of MDM2 as a 12q13-q15 amplification target in bladder cancer. However, independent CDK4 amplifications do also occur suggesting either two non-overlapping amplification sites or else a minimal overlapping region between MDM2 and CDK4 perhaps containing another yet unknown oncogene. The frequency of amplification increased significantly from stage pTa to pT1-4 (P<0.04) and from low to high grade (P<0.005). These data are consistent with a high level of genetic instability in invasively growing and high-grade bladder tumors.

Conditional mouse models of sporadic cancer

First-generation mouse tumour models, which used transgenic mice or conventional knockouts, are now being superseded by models that are based on conditional knockouts and mice that carry regulatable oncogenes. In these mice, somatic mutations can be induced in a tissue-specific and time-controlled fashion, which more faithfully mimics sporadic tumour formation. These second-generation models provide exciting new opportunities to gain insight into the contribution of known and unknown genes in the initiation, progression and treatment of cancer, and mimic human cancer better than ever before.

Gene profiling in anaplastic large-cell lymphoma-derived cell lines with cDNA expression arrays

Comparative hybridization of cDNA arrays is a powerful high-throughput tool for the measurement of gene expression between two or more tissues. Moreover, cDNA array technology permits detection of genes involved in carcinogenesis and prospective targets for diagnosis or therapy in the future. We used this technique in cell lines to discover genes possibly involved in the pathogenesis of the large-cell anaplastic lymphoma (ALCL). CD30(+) lymphoma cells are characteristic for this lymphoma entity and were first described 16 years ago. Still, little is known about the biology of this aggressive lymphoma. Identical arrays covering 1176 genes (Clontech Human Cancer 1.2) were hybridized with labeled first-strand cDNA from five ALCL cell lines and stimulated T cells. We found overlapping gene expression patterns in the five cell lines. Three genes, tissue inhibitor of metalloproteinases 1 (TIMP-1), nm23-H4, and interferon-induced 56K protein (IFI-56K) were strongly overexpressed in all ALCL cell lines compared with stimulated T cells. Whereas interleukin-2 precursor, interleukin-9 precursor, interferon-gamma precursor, T cell surface glycoprotein CD3 epsilon subunit, 40S ribosomal protein SA, and protein-tyrosine phosphate 1C were underexpressed, Reverse transcription polymerase chain reaction and Northern blot analysis were employed to confirm the expression patterns of the genes identified by Atlas hybridization.

Reliability of the tissue microarray based FISH for evaluation of the HER-2 oncogene in breast carcinoma

AIMS

Tumour tissue microarray allows the analysis of hundreds of tumour samples simultaneously on a single microscope slide. However, the extremely small tissue samples taken from the original tissue may not always be representative of the entire tumour.

METHODS

The reliability of this new technology was investigated by analysing HER-2 oncogene amplification by fluorescence in situ hybridisation (FISH) from representative slides of the whole tumour and small tissue core biopsies from 29 invasive breast tumours.

RESULTS

The tissue microarray method had high accuracy; in only one of 29 cases (3.4%; 95% confidence interval, 0% to 10%) were the results discordant with whole tumour analysis.

CONCLUSION

Tumour microarray is a highly reliable method for analysing HER-2 oncogene amplification by FISH in human breast tumours.

Identification of novel targets for cancer therapy using expression proteomics

Although most drugs target proteins, the proteome has remained largely untapped for the discovery of drug targets. The sequencing of the human genome has had a tremendous impact on proteomics and has provided a framework for protein identification. There is currently substantial interest in implementing proteomics platforms for drug target discovery. Although the field is still in the early stages, current proteomic tools include a variety of technologies that could be implemented for large-scale protein expression analysis of cells and tissues, leading to discovery of novel drug targets. Proteomics uniquely allows delineation of global changes in protein expression patterns resulting from transcriptional and post-transcriptional control, post-translational modifications and shifts in proteins between different cellular compartments. Some of the current technologies for proteome profiling and the application of proteomics to the analysis of leukemias by our group are reviewed.

Large-scale gene expression analysis in molecular target discovery

The evolution of simple arrays consisting of a few genes to ones composed of thousands of genes and/or ESTs has allowed investigators unprecedented views of the molecular mechanisms within cells. Due to the enormous quantities of information derived from microarray analysis, new types of problems have surfaced, such as where to store all of the data. The ability to solve database or statistical problems has required the bench biologist to collaborate with database developers, software designers and statisticians to determine solutions for storage, analysis and interpretation of microarray data. The collaborative effort between these extremely diverse disciplines has led to the development of creative database query and gene expression analysis tools, producing significant reductions in the time required by researchers to filter through the datasets and discover the key processes perturbed by the diseases of interest. Both unsupervised and supervised analysis methods have been applied to gene expression data leading to the discovery of novel therapeutic targets and diagnostic markers. Furthermore, tumor classification based on their respective molecular fingerprints has led to the classification of cancer subtypes and the discovery of novel molecular taxonomies that may eventually lead to improved patient stratification and superior therapeutic strategies.

Expression of cyclic adenosine monophosphate response-element binding protein in acute leukemia

Cyclic adenosine monophosphate response-element binding protein (CREB) is a nuclear protein that regulates expression of genes that control cell proliferation, differentiation, and survival. To analyze CREB expression in leukemia cells, we conducted Western blot analysis of bone marrow cells obtained from patients with acute lymphoblastic leukemia, patients with acute myeloid leukemia, and patients without active leukemia. CREB was expressed at a higher frequency in bone marrow cells from patients with acute lymphoid or myeloid leukemia than in patients with leukemia remission or without leukemia. Our results indicate that CREB expression could be a useful marker for leukemia in patients with acute disease and suggest a role for CREB in leukemogenesis.

Telomere length assessment in human archival tissues: combined telomere fluorescence in situ hybridization and immunostaining

A method was developed to assess human telomere lengths at the individual cell level in tissue sections from standard formalin-fixed paraffin-embedded tissues. We coupled this method with immunofluorescence to allow the simultaneous identification of specific cell types. Validation of this in situ quantification method showed excellent agreement with the commonly used telomere repeat fragment-Southern blot method. The assay requires very few cells ( approximately 10 to 15). Thus, small tissue samples, including clinical biopsies, can be easily accommodated. In addition, the cells under study need not be actively cycling and there is no requirement for tissue disaggregation or cell culture. This method provides a more accurate assessment of telomere lengths than Southern blotting because confounding contributions from undesired cell types within tissue samples are avoided. Using this technique, we were able to perform the first comparison of relative telomere lengths in matched tumor versus normal epithelial cells within archival human prostate tissues.

DNA microarrays in clinical oncology

Aberrant gene expression is critical for tumor initiation and progression. However, we lack a comprehensive understanding of all genes that are aberrantly expressed in human cancer. Recently, DNA microarrays have been used to obtain global views of human cancer gene expression and to identify genetic markers that might be important for diagnosis and therapy. We review clinical applications of these novel tools, discuss some important recent studies, identify promising avenues of research in this emerging field of study, and discuss the likely impact that expression profiling will have on clinical oncology.

Understanding cancer metastasis: an urgent need for using differential gene expression analysis

Cancer is a multistep process and occurs as a result of the loss of control of cell division, leading to the initial tumor formation, which is followed by metastatic spread. Recent years have witnessed a vast improvement in the understanding of the molecular mechanisms regulating cell division and their links to tumorigenesis. The process of metastasis involves an intricate interplay between cell adhesion, proteolysis, migration, and angiogenesis. However, there is little knowledge of how these events are coordinately regulated in the tumor cell. Given that the uncontrolled spread of the tumor to distant organs is usually lethal, a study of the molecular mechanisms regulating metastasis assumes great significance. Recently, several technologies have been developed for analyzing differential gene expression. The current review discusses the importance of these technologies in the molecular analyses of metastasis.

Low prevalence of Epstein-Barr virus in incident gastric adenocarcinomas from the United Kingdom

Epstein-Barr virus has been associated with a proportion of typical gastric adenocarcinomas. Here we report that the prevalence of Epstein-Barr virus in gastric adenocarcinomas from the United Kingdom is one of the lowest in the World. Gastric adenocarcinoma is another tumour whose association with Epstein-Barr virus varies with the population studied.

Interactive gene expression pattern in prostate cancer cells exposed to phenolic antioxidants

Dietary phenolic compounds are known to elicite vital cellular responses such as cell cycle arrest, apoptosis and differentiation by activating a cascade of molecular events. As there is an increasing interest to improve the efficacy of these compounds for use as potential chemopreventive agents, we wanted to understand the impact of phenolic compounds on target genes in prostate cancer. In this study we used human cDNA microarrays with 2400 clones consisting of 17 prosite motifs to characterize alterations in gene expression pattern in response to the phenolic antioxidants ellagic acid (EA) and resveratrol (RE). Over a 48-hr exposure of androgen - sensitive LNCaP cells to EA and RE, a total of 593 and 555 genes respectively, showed more than a two fold difference in expression. A distinct set of genes in both EA-and RE-treated cells may represent the signature profile of phenolic antioxidant-induced gene expression in LNCaP cells. Although extensive similarity was found between effects of EA - and RE - responsive genes in prostate cancer cells, out of 246 genes with overlapping responses, 25 genes showed an opposite effect. Quantitative RT-PCR was used to verify and validate the differential expression of selected genes identified from cDNA microarrays. In-depth analysis of the data from this study provided insight into the alterations in the p53 - responsive genes, p300, Apaf-1, NF-kBp50 and p65 and PPAR families of genes, suggesting the activation of multiple signaling pathways that leads to growth inhibition of LNCaP cells. This is a first study to look for changes in a large number of human genes in response to dietary compounds.

Tissue microarray sampling strategy for prostate cancer biomarker analysis

High-density tissue microarrays (TMA) are useful for profiling protein expression in a large number of samples but their use for clinical biomarker studies may be limited in heterogeneous tumors like prostate cancer. In this study, the optimization and validation of a tumor sampling strategy for a prostate cancer outcomes TMA is performed. Prostate cancer proliferation determined by Ki-67 immunohistochemistry was tested. Ten replicate measurements of proliferation using digital image analysis (CAS200, Bacus Labs, Lombard, IL, USA) were made on 10 regions of prostate cancer from a standard glass slide. Five matching tissue microarray sample cores (0.6 mm diameter) were sampled from each of the 10 regions in the parallel study. A bootstrap resampling analysis was used to statistically simulate all possible permutations of TMA sample number per region or sample. Statistical analysis compared TMA samples with Ki-67 expression in standard pathology immunohistochemistry slides. The optimal sampling for TMA cores was reached at 3 as fewer TMA samples significantly increased Ki-67 variability and a larger number did not significantly improve accuracy. To validate these results, a prostate cancer outcomes tissue microarray containing 10 replicate tumor samples from 88 cases was constructed. Similar to the initial study, 1 to 10 randomly selected cores were used to evaluate the Ki-67 expression for each case, computing the 90th percentile of the expression from all samples used in each model. Using this value, a Cox proportional hazards analysis was performed to determine predictors of time until prostate-specific antigen (PSA) recurrence after radical prostatectomy for clinically localized prostate cancer. Examination of multiple models demonstrated that 4 cores was optimal. Using a model with 4 cores, a Cox regression model demonstrated that Ki-67 expression, preoperative PSA, and surgical margin status predicted time to PSA recurrence with hazard ratios of 1.49 (95% confidence interval [CI] 1.01-2.20, p = 0.047), 2.36 (95% CI 1.15-4.85, p = 0.020), and 9.04 (95% CI 2.42-33.81, p = 0.001), respectively. Models with 3 cores to determine Ki-67 expression were also found to predict outcome. In summary, 3 cores were required to optimally represent Ki-67 expression with respect to the standard tumor slide. Three to 4 cores gave the optimal predictive value in a prostate cancer outcomes array. Sampling strategies with fewer than 3 cores may not accurately represent tumor protein expression. Conversely, more than 4 cores will not add significant information. This prostate cancer outcomes array should be useful in evaluating other putative prostate cancer biomarkers.

Expression of Bcl-2 family member Bid in normal and malignant tissues

Bid is the only known Bcl-2 family member that can function as an agonist of proapoptotic Bcl-2-related proteins such as Bax and Bak. Expression of the proapoptotic Bcl-2 family protein Bid was assessed by immunoblotting and immunohistochemical methods in normal murine and human tissues, and in several types of human cancers and tumor cell lines. Bid expression in normal tissues varied widely, with prominent Bid immunostaining occurring in several types of short-lived cells (e.g., germinal center B cells, peripheral blood granulocytes, differentiated keratinocytes) and in apoptosis-sensitive cells (e.g., adult neurons). Analysis of Bid expression by immunostaining of 100 colon, 95 ovarian, and 254 prostate cancers, as well as 59 brain tumors and 50 lymphomas, revealed evidence of altered Bid regulation in some types of cancers. Correlations with clinical outcome data revealed association of higher levels of Bid with longer recurrence-free survival in men with locally advanced (T3 stage) prostate cancer (P=0.04). Immunoblot analysis of Bid protein levels in the NCI's panel of 60 human tumor cell lines revealed a correlation between higher levels of Bid and sensitivity to ribonucleotide reductase (RR)-inhibiting drugs (P<0.0005). Overexpression of Bid in a model tumor cell line by gene transfection resulted in increased sensitivity to apoptosis induction by a RR inhibitor. Taken together, these observations suggest a potential role for Bid in tumor responses to specific chemotherapeutic drugs, and lay a foundation for future investigations of this member of the Bcl-2 family in healthy and diseased tissues.

Importance of vascular phenotype by basic fibroblast growth factor, and influence of the angiogenic factors basic fibroblast growth factor/fibroblast growth factor receptor-1 and ephrin-A1/EphA2 on melanoma progression

The expression of several angiogenic factors and receptors was examined in a series of vertical growth phase cutaneous melanomas using high-throughput tissue microarray technology and immunohistochemistry. The results were correlated with microvessel density, clinicopathological features, and patient survival. Expression of basic fibroblast growth factor (bFGF) was significantly associated with increased microvessel density. Also, we found an independent prognostic importance of vascular phenotype by endothelial cell expression of bFGF; cases with positive vessels had the best prognosis and these tumors revealed a low frequency of vascular invasion (14%) when compared with bFGF-negative vessels (47%). This bFGF-negative phenotype was significantly increased in metastatic lesions. Strong tumor cell expression of FLT-4, ephrin-A1, and EphA2 was associated with increased melanoma thickness, and ephrin-A1 staining was related to decreased survival (P = 0.039). Expression of EphA2 in tumor cells was associated with increased tumor cell proliferation (Ki-67 positivity), indicating possible autocrine growth stimulation. Thus, our findings indicate the presence of phenotypic diversity among tumor-associated vessels, and subgroups defined by bFGF expression may be of clinical importance. bFGF was associated with microvessel density, whereas the ephrin-A1/EphA2 pathway might also be important for tumor cell proliferation and patient survival.

Use of interphase fluorescence in situ hybridization as a powerful diagnostic tool in cytology

Interphase fluorescence in situ hybridization (I-FISH) using labeled nucleic acid probes detects chromosomal and genetic aberrations at a cellular level. I-FISH is a relatively fast and sensitive technique for evaluating a large number of cells and revealing more specific information than other techniques. It has been proven to be an invaluable molecular test in cytologic analyses for the detection of subtle genetic alterations that correlate with disease progression. In this postgenomic era, with the draft of the human genome available and expansion of the knowledge of tumor-specific genetic changes, the application of I-FISH probes in cytologic analysis should be of great value in the early detection, risk assessment, and monitoring of therapy efficacy in cancer. Here, we outline the principle of the I-FISH procedure, present suggestions to efficiently analyze cytologic materials, provide examples of practical applications, and discuss new aspects of the technique.

Tissue microarrays: a powerful tool for high-throughput analysis of clinical specimens: a review of the method with validation data

With recent advances in the knowledge of human molecular genetics, new gene-based disease mechanisms are emerging in many areas of medicine. The study of new prognostic and diagnostic markers in large numbers of clinical specimens is an important step in translating the new findings from basic science to clinical practice. The recently developed tissue microarray technology allows parallel molecular profiling of clinical samples at the DNA, RNA, and protein level. This technique enables pathologists to perform large-scale analyses using immunohistochemistry, fluorescence in situ hybridization, or RNA in situ hybridization substantially faster and at markedly lower costs compared with the conventional approach. This article provides a short review of this technology, focuses on several technical aspects of tissue microarray construction, and addresses the validity of the tissue microarray results for clinical research by reviewing data from recent literature along with the authors' own data.

Adrenocortical carcinoma: clinical, morphologic, and molecular characterization

PURPOSE

To define multimolecular phenotypes of adrenocortical carcinoma (ACC) and to correlate outcome with morphologic and molecular parameters.

PATIENTS AND METHODS

Clinical data were analyzed for 124 patients, histopathologic slides for 67 primary tumors, and tissue specimens for 74 patients (38 primary and 36 metastatic tumors) with ACC and for 38 normal adrenal tissue samples. Molecular expression profiles were investigated by immunohistochemistry. The prognostic significance of 12 gross and histologic parameters in 67 primary ACCs was evaluated. Morphologic and protein expression patterns were correlated with disease-specific survival (DSS). Univariate influence of prognostic factors on DSS was analyzed by log-rank test and multivariate analysis by Cox regression.

RESULTS

The median follow-up period was 4.7 years. Significant predictors of DSS included distant metastasis at time of initial presentation; venous, capsular, and adjacent organ invasion; tumor necrosis, mitotic rate, atypical mitosis, and mdm-2 overexpression. Five-year DSS by number (one to six) of adverse histologic parameters was as follows: one to two, 84%; three to four, 37%; more than four, 9% (P =.005). The phenotype Ki-67(-)p53(-)mdm-2(+)cyclinD1(-)Bcl-2(-)p21(-)p27(+) was observed in 83% of normal and 3% of malignant adrenal tissue (P =.01). Molecular phenotypic expression was more heterogeneous in malignant than in normal (10 v five phenotypes) adrenal tissue.

CONCLUSION

Meticulous morphologic evaluation, mitotic count, and tumor stage are essential in determining prognosis for patients with ACC. Multimolecular phenotyping demonstrates that the molecular complexity and heterogeneity of these neoplasms are such that targeted therapy needs to be patient specific.

Solid-phase profiling of proteins

Protein array technologies refer to any fabrication and use of any arrays containing multiple proteins captured on solid surfaces. This valuable tool is used for high-throughput protein-protein interaction studies, protein marker discovery and other applications. This unit provides basic protocols for biomarker discovery and protein-protein interactions. Different kinds of recently emerged protein array technologies and related detection methods are reviewed.

Methods of molecular analysis: assessing losses and gains in tumours

The study of chromosomal aberrations has facilitated the understanding of tumorigenesis. By applying molecular genetic techniques to regions highlighted by cytogenetic study, many genes important in tumorigenesis have been identified. This review will describe the cytogenetic and molecular cytogenetic techniques used to identify these changes. The clinical information that they can provide, including diagnostic and prognostic information, will also briefly be discussed.

Clinical and functional target validation using tissue and cell microarrays

Expression levels of thousands of genes or proteins can be readily determined using microarray techniques. However, this represents only the first step in understanding the biological and medical significance of these molecules. New high-throughput techniques, such as tissue and cell microarrays, will facilitate clinical and functional analysis of molecular targets.

Subjective differences in outcome are seen as a function of the immunohistochemical method used on a colorectal cancer tissue microarray

Immunohistochemistry is a useful technique to localize antigens in cell preparations and tissue sections and can be helpful in identifying molecular markers that may be predictive of patient outcomes. Subjective assessment of expression and semiquantitative grading systems are the current standards in pathology literature for the analysis of tissue sections. However, expression levels assessed in this manner may be dramatically affected by the method of visualization. Tissue microarray (TMA) is a recently developed technique for the simultaneous high-throughput evaluation of protein expression on tissue samples from large cohorts of patients. The scoring of TMAs has, in general, mirrored the systems utilized for tissue sections. Here, 4 detection systems (avidin-biotin complex, indirect immunofluorescence, peroxidase-labeled polymer conjugate, and the latter with Cyanine-3-Tyramide amplification) were compared using a beta-catenin antibody on a TMA containing a cohort of colorectal cancer specimens. Peroxidase-labeled polymer with or without tyramide enhancement was found to be the most sensitive method, revealing a greater staining intensity and percentage of nuclear staining, without an apparent increase in background. Subjective assessment of expression is highly dependent on the method of visualization and may illustrate why discrepant data is often seen in literature based on immunohistochemistry.

Nucleic acid biotechnology

Driven by advances in the acquisition of genetic sequence information and the ability to manipulate small quantities of nucleic acid, a number of technologies are emerging that exploit nucleic acids for research, diagnostic, and therapeutic utility. In this review, we cover three technologies based on nucleic acids--DNA microarrays, antisense technology, and gene therapy--that are especially promising and may make a substantial impact in the laboratory and in the clinic during the coming years. For each of these areas, an overview of the current status and applications is provided, followed by a discussion of critical issues and challenges to be faced for further advancement of the technology; an emphasis is placed on quantitative and engineering aspects.

Bacterial pathogen genomics and vaccines

Infectious diseases remain a major cause of deaths and disabilities in the world, the majority of which are caused by bacteria. Although immunisation is the most cost effective and efficient means to control microbial diseases, vaccines are not yet available to prevent many major bacterial infections. Examples include dysentery (shigellosis), gonorrhoea, trachoma, gastric ulcers and cancer (Helicobacter pylori). Improved vaccines are needed to combat some diseases for which current vaccines are inadequate. Tuberculosis, for example, remains rampant throughout most countries in the world and represents a global emergency heightened by the pandemic of HIV. The availability of complete genome sequences has dramatically changed the opportunities for developing novel and improved vaccines and facilitated the efficiency and rapidity of their development. Complete genomic databases provide an inclusive catalogue of all potential candidate vaccines for any bacterial pathogen. In conjunction with adjunct technologies, including bioinformatics, random mutagenesis, microarrays, and proteomics, a systematic and comprehensive approach to identifying vaccine discovery can be undertaken. Genomics must be used in conjunction with population biology to ensure that the vaccine can target all pathogenic strains of a species. A proof in principle of the utility of genomics is provided by the recent exploitation of the complete genome sequence of Neisseria meningitidis group B.

Expression and gene copy number analysis of ERBB2 oncogene in prostate cancer

An anti-ERBB2 antibody, trastuzumab, has been shown to be highly efficient in the treatment of metastatic breast cancers overexpressing the ERBB2 gene. It has been suggested that overexpression and even amplification of ERBB2 may play a role in the development of prostate cancer. Here, we have analyzed gene copy number and expression of the ERBB2 gene in both androgen-dependent primary and metastatic tumors, as well as recurrent hormone-refractory tumors. The expression levels were compared to the expression of ERBB2 in breast cancers with or without ERBB2 gene amplification. Of 126 prostate tumors, chromogenic in situ hybridization (CISH) revealed only 1 case containing borderline (six to eight copies) amplifications of ERBB2. This hormone-refractory tumor, however, did not express ERBB2 protein. Immunohistochemical staining of ERBB2 protein was negative (0 or 1+ intensity) in all prostate samples (n = 124) analyzed. To quantitate the level of ERBB2 mRNA expression in prostate tumors (n = 34) and cell lines (n = 3), as well as in breast tumors (n = 30) and cell lines (n = 16), real-time reverse transcriptase-polymerase chain reaction (LightCycler) methodology was used. The expression level was similar in all prostate tumor types and corresponded to the level of expression in breast carcinomas without ERBB2 amplification. Breast tumors with ERBB2 amplification expressed, on average, approximately 20 times (P < 0.001) higher mRNA levels than prostate tumors or breast carcinomas without the gene amplification. In conclusion, the expression of ERBB2 in prostate cancer is relatively low, and is not altered during disease progression. Thus, it is unlikely that treatment modalities relying on the overexpression of ERBB2 gene will be useful in treating prostate cancer.

Clinical significance of molecular expression profiles of Hürthle cell tumors of the thyroid gland analyzed via tissue microarrays

Hürthle cell tumors are rare thyroid neoplasms for which disease biology is poorly understood and diagnosis of carcinoma can be challenging. The aim of the study was to characterize molecular expression profiles of Hürthle cell tumors and to determine the clinical significance of identified phenotypes. Paraffin-embedded tissue cores of normal thyroid (n = 18), and histopathologically well-defined Hürthle cell adenomas (n = 27), Hürthle cell tumors of unknown malignant behavior (n = 7), and minimally (n = 14) and widely (n = 21) invasive Hürthle cell carcinomas were arrayed in triplicate on tissue microarrays. Expression profiles of p53, mdm-2, p21, Bcl-2, cyclin D1, and Ki-67 were detected by immunohistochemistry and correlated with clinicopathological data and patient outcome using standard statistical methodology. Median follow-up time was 8 years. High Ki-67 proliferative index was evident only in the clinically aggressive widely invasive Hürthle cell carcinomas and was associated with significantly reduced relapse-free (P = 0.001) and disease-specific (P < 0.001) survival. The molecular phenotype of Hürthle cell tumors, independent of histopathological subtype diagnosis, was characterized by p53(-), mdm-2(+), p21(+/-), cyclin D1(-), and Bcl-2(+/-). Normal thyroid tissue demonstrated a p53(-), mdm-2(-), p21(-), cyclin D1(-), and Bcl-2(+) phenotype. The Bcl-2(+) phenotype was associated with improved relapse-free survival (P = 0.04) and disease-specific survival (P = 0.01) in widely invasive carcinomas and the Ki-67(+)/Bcl-2(-) phenotype was associated with the diagnosis of widely invasive Hürthle cell carcinoma (P < 0.001). This study demonstrates that tissue microarray-based profiling allows identification of molecular markers that are associated with patient prognosis. High Ki-67 proliferative index was associated with adverse outcome in Hürthle cell neoplasms. Together with down-regulation of Bcl-2, high Ki-67 proliferative index may be useful for diagnosing widely invasive Hürthle cell carcinomas. Molecular alterations in the p53 pathway play a role in Hürthle cell tumorigenesis, but other unidentified molecular changes seem to be required to induce the malignant phenotype.

Moving smaller in drug discovery and delivery

Advances in new micro- and nanotechnologies are accelerating the identification and evaluation of drug candidates, and the development of new delivery technologies that are required to transform biological potential into medical reality. This article will highlight the emerging micro- and nanotechnology tools, techniques and devices that are being applied to advance the fields of drug discovery and drug delivery. Many of the promising applications of micro- and nanotechnology are likely to occur at the interfaces between microtechnology, nanotechnology and biochemistry.

Gene-expression profiles in hereditary breast cancer

DNA microarray technology is revolutionizing the way fundamental biologic questions are addressed in the postgenomic era. In this study by Hedenfalk et al., the authors attempted to identify discrete gene expression profiles for patients with known hereditary breast cancers caused by mutations in BRCA1 and BRCA2, both of which increase the lifetime risk of developing breast cancer. The genome-wide perspective identified discrete sets of genes that discriminated between BRCA, BRCA2, and sporadic breast tumors. This commentary discusses some limitations of studying a small number of cases. The authors also address the need for validation on independent tumor sets, and the potential benefit of multivariable-type analyses to consider other potential confounding factors such as tumor grade, receptor status, tumor stage, and treatment information.

Characterization of prostate cell types by CD cell surface molecules

A set of monoclonal antibodies raised against lymphocyte cell surface molecules, the cluster designation (CD) antigens, was used to distinguish the constituent cell types of the prostate. The luminal secretory epithelial, basal epithelial, fibromuscular stromal, nerve sheath, and endothelial cells express distinctive complements of cell surface molecules that were identified by immunohistochemistry using 152 commercially available antibodies. Many of the CD antibodies stained lymphocyte populations in the prostate. These lymphocyte populations were grouped into abundance classes of rare, moderate, and high. Some of these molecules are expressed by multiple cell types, both parenchymal and lymphoid; others are expressed by only one cell type. Distinctive patterns of CD expression, which are most similar to the expression pattern of prostate luminal cells, also characterize a small series of Gleason score 6 prostate cancers. The cell-type specificity of CD molecules increases the prospect of isolating specific cell populations, using such techniques as laser capture microdissection and flow cytometry, for cell-specific molecular studies.