Previous studies directed at identifying paclitaxel resistance genes in a paclitaxel-resistant subclone of the human ovarian cancer cell line SKOV-3 identified a novel cancer testis antigen, Taxol resistance-associated gene 3 (TRAG-3). Because investigation suggested that TRAG-3, located on chromosome Xq28, does not directly participate in the paclitaxel-resistant phenotype, it was hypothesized that TRAG-3 might be linked to a neighboring gene that is directly involved in the drug-resistant phenotype, or alternatively, overexpression of TRAG-3 might be attributable to coregulation with other cancer testis antigens. To distinguish between these two hypotheses, expression of the genes that flank TRAG-3 was evaluated, namely the Centrin 2 gene and several members of the MAGE gene cluster. Northern analysis demonstrates overexpression of MAGE2 but not Centrin 2. Extension of this analysis to other neighboring and non-neighboring representative cancer testis antigens reveals overexpression of MAGE3, MAGE6, MAGE11, and MAGE12, as well as GAGE-2, GAGE-4, GAGE-5, GAGE-6, and GAGE-7 (clustered on Xp11) in SKOV-3(TR), as compared with SKOV-3. In addition, Affymetrix-based analysis of gene expression in SKOV-3 subclones with variable paclitaxel resistance demonstrates MAGE gene overexpression occurs early in the development of the paclitaxel-resistant phenotype, whereas GAGE gene overexpression occurs somewhat later. Evaluation of additional breast and ovarian cancer cell lines reveals MAGE/GAGE overexpression in both paclitaxel- and doxorubicin-resistant cell lines, whereas gemcitabine-resistant subclones of several ovarian cancer cell lines, including SKOV-3(GR), reveals no change in MAGE/GAGE expression. To determine whether MAGE gene overexpression contributes directly to the drug-resistant phenotype, MAGE2 or MAGE6, cDNA was introduced into the paclitaxel-sensitive human ovarian cancer cell line OVCAR8. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide cytotoxicity analysis of both MAGE2 and MAGE6 transfectants demonstrates a 4-fold increase in resistance to paclitaxel and 2-fold increase in resistance to doxorubicin but not to other drugs, such as topotecan and cisplatin, through a nonmultidrug resistance-1 mechanism. MAGE2 or MAGE6 overexpression also induces a growth advantage in OVCAR8-transfected cells. These studies suggest that the in vitro acquisition of paclitaxel and doxorubicin resistance can be associated with increased expression of a variety of both neighboring and non-neighboring cancer testis antigens genes. This does not appear to be a consequence of random genetic instability or genomic amplification of the X chromosome. These antigens, because of limited expression in normal tissues, may be suitable targets for immunotherapy and novel therapeutic strategies in the treatment of chemotherapy-resistant epithelial tumors.