The resistance of tumor cells to chemothera-

The presence of MDR1-expressing tumor cells may be useful as a predictive marker for clinical resistance to combination chemotherapy in ovarian cancer and SCLC. Prospective studies are needed to confirm this hypothesis.

Identification of potential anticancer drug targets through the selection of growth-inhibitory genetic suppressor elementsIn this article by Primiano et al. (Cancer Cell 4,, the authors inadvertently indicated the incorrect concentration of BrdU on page 50 in the second paragraph of the second column; this should read 50 M BrdU.

Selective inhibition of specific genes can be accomplished using genetic suppressor elements (GSEs) that encode antisense RNA, dominant negative mutant proteins, or other regulatory products. GSEs may correspond to partial sequences of target genes, usually identified by trial and error. We have used bacteriophage lambda as a model system to test a concept that biologically active GSEs may be generated by random DNA fragmentation and identified by expression selection. Fragments from eleven different regions of lambda genome, encoding specific peptides or antisense RNA sequences, rendered E. coli resistant to the phage. Analysis of these GSEs revealed some previously unknown functions of phage lambda, including suppression of the cellular lambda receptor by an 'accessory' gene of the phage. The random fragment selection strategy provides a general approach to the generation of efficient GSEs and elucidation of novel gene functions.

To identify human genes required for tumor cell growth, transcriptome-scale selection was used to isolate genetic suppressor elements (GSEs) inhibiting breast carcinoma cell growth. Growth-inhibitory GSEs (cDNA fragments that counteract their cognate gene) were selected from 57 genes, including known positive regulators of cell growth or carcinogenesis as well as genes that have not been previously implicated in cell proliferation. Many GSE-cognate genes encode transcription factors (such as STAT and AP-1) and signal transduction proteins. Monoclonal antibodies against a cell surface protein identified by GSE selection, neural cell adhesion molecule L1CAM, strongly inhibited the growth of several tumor cell lines but not of untransformed cells. Hence, selection for growth-inhibitory GSEs allows one to find potential targets for new anticancer drugs.