Introduction: The aim of this study was to use a gel-based comparative proteomics approach to identify and quantify proteins differentially expressed between non-small cell lung cancer (NSCLC) cell lines sensitive to gefitinib and those resistant to it. These differentially expressed proteins may serve as biomarkers that distinguish the subpopulation of patients with NSCLC tumors that will respond to treatment with gefitinib. Materials and Methods: We examined thirteen NSCLC cell lines with differing sensitivity to gefitinib: 8 cell lines were sensitive (IC50 < 1μM) and 5 were resistant (IC50 > 10μM). Four of the 8 sensitive cell lines had mutations in EGFR. The technique of difference gel electrophoresis (DIGE) was used to quantify protein expression disparities across the cell lines. An internal standard composed of a pool of all the cell lines was used to facilitate gel matching and improve quantitative precision. Protein spots determined to be differentially expressed between sensitive and resistant cell lines were robotically excised from the gel, digested with trypsin, analyzed by MALDI-TOFMS and/or LC-MS/MS and then identified by database searching. Results: Approximately 1800-2400 protein spots were resolved on each gel. Of these, 940 protein spots were matched across 10 or more of the 13 gels using a pooled internal standard. Twenty five spots were established to be differentially expressed between sensitive and resistant cell lines with a p < 0.01. Twenty spots were increased in the sensitive cell lines (with fold changes ranging from 1.4-5.5) and 5 were increased in the resistant lines (with fold changes ranging from 1.3 - 10.1). Proteins identified as increased in gefitinib sensitive NSCLC lines included: catecohol-O-methyl transferase (COMT), cathepsin B, prohibitin (PHB), gelsolin isoform B, serine proteinase inhibitor clade B, member 1 (Serpin B1), triosephosphate isomerase and cytokeratin 19. Proteins identified as increased in resistant cell lines included: ubiquitin carboxy-terminal hydrolase L1 (UCHL1) and AICAR formyltransferase/IMP cyclohydrolase bifunctional enzyme. Selected proteins are being evaluated as predictive and prognostic markers in human NSCLC tumors using tissue microarrays. Conclusion: Through the application of DIGE and mass spectrometry we were able to identify a panel of putative biomarkers that predict sensitivity and resistance to gefitinib in vitro. Ongoing studies in human NSCLC tumors will determine the clinical utility of these biomarkers.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]