Important Expression Changes In Cap Progression

Expression profiling may also be very important in identifying the driving genetic changes in individual prostate tumors. For example, despite its importance in terms of therapeutics and biologi-

Fig. 4. Transcript and protein expression of myosin light chain kinase (MYLK). (A) Oncomine (www. oncomine.org) was used to visualize the normalized expression of MYLK across eight data sets (Lapointe et al. [40] = 1; Dhanasekaran et al. [26] = 2; Welsh et al. [28] = 3; Luo et al. [29] = 4; LaTulippe et al. [41] = 5; Luo et al. [42] = 6; Singh et al. [47] = 7; and Magee et al. [27] = 8), comparing grossly dissected benign tissue (blue) vs localized prostate cancer (CaP) (red). (B) Tissue microarray analysis of MYLK protein expression. Representative cores of benign (NOR), localized CaP, and metastatic CaP (MET) tissues from a tissue microarray stained for MYLK expression demonstrating stromal expression.

Fig. 4. Transcript and protein expression of myosin light chain kinase (MYLK). (A) Oncomine (www. oncomine.org) was used to visualize the normalized expression of MYLK across eight data sets (Lapointe et al. [40] = 1; Dhanasekaran et al. [26] = 2; Welsh et al. [28] = 3; Luo et al. [29] = 4; LaTulippe et al. [41] = 5; Luo et al. [42] = 6; Singh et al. [47] = 7; and Magee et al. [27] = 8), comparing grossly dissected benign tissue (blue) vs localized prostate cancer (CaP) (red). (B) Tissue microarray analysis of MYLK protein expression. Representative cores of benign (NOR), localized CaP, and metastatic CaP (MET) tissues from a tissue microarray stained for MYLK expression demonstrating stromal expression.

cal behavior, in breast cancer profiling studies, ERBB2 does not appear on the list of the most upregulated genes between normal breast tissue and breast cancer. Most commonly, studies attempt to identify the genes that are most differentially expressed (often using a i-test) between two classes of samples. In these types of analyses, ERBB2 is not highly significant, because it is not upregulated in all cases of cancer, even though it is very highly expressed when it is upregulated. A similar situation has recently been discovered during trials evaluating the epidermal growth factor receptor (EGFR) inhibitor, gefitinib, in the treatment of non-small cell lung cancer. During the trial, only 20% of patients had major objective responses and several groups have now determined that most patients experiencing a response have tumors that demonstrate mutations associated with the ATP-binding site of EGFR, the target of gefitinib (89,90). Again, if researchers were mining lung cancer microarray data for genes differentiating between lung cancer and benign, the downstream targets of activated EGFR would be missed, because EGFR is mutated in only 10 to 20% of cases. Recent work has confirmed that signatures in individual tumors can reveal oncogene activation or amplification when compared with mouse models (91,92). For example, a study by Ellwood-Yen et al. used expression signatures from mouse tumors driven by prostate specific expression of Myc to identify a subset of human CaPs that have a "Myc-like" signature (9). Taken together, these results indicate that directed analysis of microarray data may be able to identify genetic alterations that drive CaP progression as well as identify potential therapeutic targets. Recently, we developed a bioinformatics algorithm termed COPA to identify genes with marked-over-expression in a fraction of cancer cases from DNA microarray data. This led to the discovery of recurrent gene fusions in the majority of prostate cancers, where the 5' region of the androgen regulated gene TMPR552 is fused to the ETS transcription factors ERG or ETV1 in cases with outlier expression of the respective gene (92).

Was this article helpful?

0 0

Post a comment