Microarray Studies of Gene Expression in Prostate Carcinoma

Microarray analysis has been used in several published studies designed to profile gene expression alterations in prostate carcinoma (Table 1). Importantly, although all of these experiments used microarrays, there are important differences that preclude a simple comparison of the reported results. These include the use of different patient samples, microarrays with different genes represented and variations in experimental and analytical approaches. Magee et al. used oligonucleotide arrays to characterize the expression of 4712 genes in four benign and 11 prostate cancer samples.17 Most of the neoplastic samples represented primary tumors of various Gleason grades, though two metastatic lesions were also characterized. The samples were enriched for epithelial cells using macrodissection procedures. Analyses of the gene-expression profiles identified four genes with significant changes associated with carcinoma, including Hepsin, a type-II membrane-bound serine protease. Luo et al. used microarrays comprised

Normal Prostate

Neoplastic Prostate

Normal Prostate

Labeled cDNA

Neoplastic Prostate

Labeled cDNA

Combine ~ EqualAmounts/

Observation Primary PrimaryTreatment

DietaryChange Treatment +

Chemoprevention AdjuvantTherapy

Observation Primary PrimaryTreatment

DietaryChange Treatment +

Chemoprevention AdjuvantTherapy

Fig. 1. Prostate cancer outcomes determined by microarray expression profiles. Messenger RNA (mRNA) is isolated separately from normal and neoplastic prostate tissues, preferably from specific microdissected cell types. RNA samples are either labeled directly with fluorescent probes or first converted to complementary cDNA followed by hybridization to microarrays of spotted or synthesized DNAs representing genes of interest. Gene expression measurements are determined by analyzing the fluorescent signal intensity for each gene on the microarray and subsequently comparing the signal levels between normal and neoplastic cell types and neoplastic cell types representing different clinical outcomes. Expression data are correlated with known clinical outcomes to determine a profile or fingerprint capable of predicting the risk of local and distant cancer progression. Patients with a minimal risk of progression (i.e. with indolent disease) would receive no primary intervention but could be considered for dietary alteration or chemoprevention trials. Patients at intermediate risk would receive primary curative therapy such as radical prostatectomy or radiation therapy. Patients at high risk would receive primary therapy with the addition of systemic adjuvant or neoadjuvant therapy.

of 6500 spotted cDNAs to profile gene expression in nine benign prostatic hyperplasia (BPH) specimens and 16 prostate cancer samples.18 A common reference standard was used to facilitate comparisons and clearly discernable patterns that discriminated prostate cancer and BPH was evident. The study identified 210 genes with statistically significant expression differences between the two tissue types. This study also identified Hepsin overexpression in cancer epithelium relative to benign epithelium. A study by Dhanasekaran et al. characterized the gene expression signatures of more than 50 normal and cancerous prostate specimens and three prostate cancer cell lines using arrays comprised of 9984 cDNAs.19 Cohorts of genes distinguishing normal prostate, BPH, localized prostate cancer and metastatic hormone-refractory prostate cancer were identified. Examples of genes differentially-expressed between benign and neoplastic prostate tissue include Hepsin, PIM1, IGFBP-5, DAN1, FAT, RAB5A and HEVIN. This study further explored the correlation of Hepsin and PIM1 protein expression using tissue microarrays comprised of 700 prostate-cancer specimens with known clinical attributes. The expression levels of Hepsin and PIM1 were shown to independently correlate with cancer progression. Overall, these microarray-based studies of prostate gene expression demonstrate consistent alterations in a subset of genes that can be grouped into a cancer-associated profile. These results provide important substrates for mechanistic studies designed to evaluate their utility as diagnostic, prognostic and therapeutic targets.

0 0

Post a comment