Molecular classification of cancer: class discovery and class prediction by gene expression monitoring.

Although cancer classification has improved over the past 30 years, there has been no general approach for identifying new cancer classes (class discovery) or for assigning tumors to known classes (class prediction). Here, a generic approach to cancer classification based on gene expression monitoring by DNA microarrays is described and applied to human acute leukemias as a test case. A class discovery procedure automatically discovered the distinction between acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) without previous knowledge of these classes. An automatically derived class predictor was able to determine the class of new leukemia cases. The results demonstrate the feasibility of cancer classification based solely on gene expression monitoring and suggest a general strategy for discovering and predicting cancer classes for other types of cancer, independent of previous biological knowledge.

DNase I hypersensitivity analysis of non-pituitary human prolactin gene expression.

The expression of non-pituitary human PRL is initiated at a unique 5' untranslated exon located approximately 5.7 kb upstream of the pituitary-specific transcriptional start site. Unlike pituitary PRL expression, transcriptional regulation from the upstream promoter does not rely on the POU-homeodomain protein Pit-1. We have used DNase I mapping of chromatin from PRL-producing and non-producing human lymphoblastoid cell lines to identify hypersensitive sites unique to the PRL expressing phenotype. Analysis of 22 kb of 5' flanking DNA revealed DNase I hypersensitive sites in intron A-1 separating the pituitary from non-pituitary specific transcription start site which were only detected in the PRL-producing cell line. Transient transfection showed strong transcriptional activity directed by this region only in the antisense orientation and in a non cell-type specific manner. Transfection experiments with deletion mutants of 5259 bp of the non-pituitary PRL promoter region also revealed promoter activity not restricted to the PRL expressing phenotype. These data suggest that non-pituitary PRL gene expression may be regulated by elements located in intron A-1 and that recapitulation of cell-specific expression requires a unique cellular context and chromatin assembly.

Two enzymes together capable of cysteine biosynthesis are encoded on a cyanobacterial plasmid.

The cyanobacterium Synechococcus sp. PCC 7942 contains two endogenous, genetically cryptic plasmids of 8.0 and 48.5 kb, which have been designated pANS and pANL, respectively. Characterization of the 3.8 kb Ba6 BamHI fragment of pANL identified three open reading frames which were transcriptionally regulated by sulfur availability and the protein CysR. One of these genes, designated srpG, encodes a protein which exhibits 67% amino acid identity to the Escherichia coli enzyme O-acetyl-L-serine (thio)-lyase A. Overlapping the 3' end of srpG is a second gene, designated srpH, which encodes a protein with similarity to the amino-terminal region of serine acetyltransferase enzymes. DNA hybridization results indicate that there is second copy of srpG in Synechococcus sp. PCC 7942, which is consistent with previous isoenzyme studies on O-acetyl-L-serine (thiol)-lyase in cyanobacteria. The introduction of srpG and srpH into E. coli cysKcysM and cysE mutant strains, respectively, results in the complementation of the lesion in cysteine biosynthesis. Additionally, the E. coli cysK cysM strain containing srpG is able to utilize sulfate more efficiently than thiosulfate, indicating that SrpG is probably a type A O-acetyl-L-serine (thiol)-lyase. The possible function of these genes in the adaptation of cyanobacteria to sulfur stress is discussed.