Osteopontin Identified as Lead Marker of Colon Cancer Progression Using Expression Profiling on Pooled Samples

Abstract

Background:
     New tumor markers and markers of progression are needed for improved staging and treatment of many cancers. Gene expression profiling techniques offer the opportunity to discover such markers. However, limited quantities of microdissected tumor tissue, as well as variations in expression pattern profiles between individuals, complicate this approach. We hypothesized that these problems could be solved by applying microarray profiling to RNA derived from pooled tumor specimens of similar stage. In addition, we proposed that the pooling process would both reduce the amount of RNA needed per sample, as well as amplify the signal of commonly expressed markers.

Methods:
     Total RNA from human colon tumors (n = 60) of multiple stages (adenomas, Dukes' B, C, D cancers, and liver metastases) were pooled within stages, and compared with pooled normal mucosal specimens (n = 10) using Affymetrix 6,800 oligonucleotide expression arrays, followed by 12,000 element arrays. Hybridization data were analyzed using a novel algorithm we developed and used to infer changes in gene expression, focusing on genes that consistently increased or decreased in their expression through tumor progression.

Results:
     Over 240 candidate tumor markers, and more than 100 markers of tumor progression, were identified. Of these, eleven tumor markers, over- or under-expressed in tumor vs. normal tissue, were validated by northern analysis. The gene most significantly and consistently differentially expressed in conjunction with tumor progression was that of the secreted, integrin-binding protein, osteopontin. Its potential as a progression marker was validated with northern analysis using RNA from an independent set of normal and tumor samples representing all stages (n = 60). Moreover, a significant correlation between osteopontin protein expression and advancing tumor stage was identified using 303 normal and tumor specimens (Spearman coefficient = 0.903, P < 0.0001).

Conclusions:
     Using a novel pooling strategy and gene expression profiling, a large number of markers of potential clinical interest were identified. One of these, osteopontin, was shown to be a highly expressed marker of progression in colon cancer. This study further demonstrates that analysis of pooled samples represents a powerful, cost-effective strategy for identifying multiple genes, within complex biological samples, that co-vary with tumor progression.