Current advances in biomolecular technology allow precise genetic fingerprinting of specific cells responsible for the pathogenesis of human diseases. This study demonstrates the feasibility of enerating target samples from laser capture microdissection (LCM) tissues suitable for hybridization of high-density oligonucleotide arrays for gene expression profiling. RNA was successfully isolated by LCM from three paired specimens of oral cancer and linearly amplified using T7 RNA polymerase. Evaluation of the cDNA revealed that five of five cellular maintenance transcripts are detected. Biotinylated cRNA was generated and hybridized to the human Test 1 GeneChip "probe arrays, which demonstrated that the RNA is of sufficient quality and integrity to warrant further analysis. Subsequent hybridization of the samples to the HuGenFL GeneChip probe arrays revealed that 26.5%-33.0% of the approximately 7000 represented genes are expressed in each of the six samples. These results demonstrate that LCM-generated tissues can generate sufficient quality cRNA for high-density oligonucleotide microarray analysis, an important step in determining comprehensive gene expression profiling using this high-throughput technology.hybridization of LCM-generated RNA to cDNA-based expression microarrays, no report has yet shown the successful hybridization to high-density oligonucleotide mircoarrays. The limited quantity and quality of RNA isolated using LCM continues to be a technical obstacle for in vivo analysis of gene expression using microarrays. While RNA isolated from LCM-harvested tissue can be converted into cDNA libraries, it is generally agreed that the most accurate representation of gene expression for DNA chips is achieved by making the target sample directly from the RNA or by a T7-based linear RNA amplification (2,4,5). Comprehensive gene expression analysis of LCM-isolated pure cell populations will represent a major biomedical advance in our understanding of the pathogenesis of human disease (7). A major challenge in this line of investigation is the ability to generate a sufficient amount and quality of the desireddesired macromolecule from biopsied material. The GeneChip"probe arrays (Affymetrix, Santa Clara, CA, USA) allow the generation of accurate and reproducible mRNA transcript-level data (3,6,9). The HuGeneFL probe array contains probes representing approximately 7000 full-length human genes. The ability to generate target samples from LCM-based cells and tissues for hybridization to DNA chips represents an important advance that demonstrates the feasibility of using this method of sample collection for gene expression analysis on DNA chips.

Human DOC-1/CDK2AP1 gene encodes a growth suppressor protein of 12kDa (p12DOC-1=CDK2AP1). Recently, p12DOC-1=CDK2AP1 has been shown to associate with cell cycle proteins including CDK2 and DNA olymerase a/primase. It negatively regulates CDK2 activities and suppressesDNAreplication. Therefore, identification of other p12DOC-1=CDK2AP1 interacting proteins might clarify its role in the cell cycle regulation and carcinogenesis. The purpose of this study was to identify additional p12DOC-1=CDK2AP1 interacting proteins using the yeast two-hybrid system. Using human p12DOC-1=CDK2AP1 as a bait in a liver cDNA library screening, cDNA clones identical to human DOC-1R transcript were identified. The interaction between p12DOC-1=CDK2AP1 and p14DOC-1R was verified in vitro and in cells. GST pull-down assay and immunoprecipitation experiments confirmed the interaction between the two proteins. The ritical region for p12DOC-1=CDK2AP1's interaction with p14DOC-1R was defined to amino acids 20-25 by using a series of deletion mutants as baits in the yeast two-hybrid system. Our data indicated that p12DOC-1=CDK2AP1 could associate with its homologous protein, p14DOC-1R.

A novel tumor suppressor gene, PTEN/MMAC1, has been recently shown to be mutated in gliomas, breast, prostate, kidney cancers and melanomas. Loss-of-heterozygosity studies in melanoma have suggested the presence of at least one chromosome 10q locus lost early in tumor progression. In this study, we screened 45 melanoma cell lines and 17 paired uncultured metastatic melanoma and peripheral blood specimens for PTEN/MMAC1 alterations using PCR-SSCP and direct sequencing. We found nine melanoma cell lines with homozygous deletions (

To screen genes involved in P15NK4b regulation during cell cycle, differential display method was applied to compare mRNAs from G1 synchronized cells of MLIK6, which overexpressed P15INK4b gene, and itscontrol MLC2. By using thisapproach, 15 cDNA fragments that were preferentially expressed in MLIK6 cells, but not in MLC2 cells, were screened out. A novel gene named P15RS was identified with further analysis. Combining the sequence from DD-PCR, homology analysis against EST database and RACE, a 4404 bp complete cDNA sequence of P15RS was generated. Sequence analysis revealed that P15RS cDNA encoded a 312-aminoacid peptide containing a RAR domain that is involved in regulation of nuclear pre-mRNA, which suggests that P15RS may be a nuclear regulation protein. Genomic sequence analysis demonstrated that human P15RS gene was localized on chromosome 18q12 with seven exons and six introns. Expressing antisense P15RS in MLIK6 cells can up-regulate the expression of cyclinD1 and cyclinE. These data indicate that P15RS may act as a negative regulator in G1 phase.

A novel p53-related gene, p73, was recently isolated and cytogenetically mapped to chromosome region 1p36. Functionally, p73 expression induces p21waf and suppresses tumor cell growth. We mapped p73 using radiation hybrids and localized the gene to an interval that putatively harbors a melanoma tumor suppressor locus. We then analyzed p73 transcripts from 24 melanoma cell lines using reverse transcription-PCR/single strand conformation polymorphism and identified nine polymorphic sequence changes (three novel and six previously published polymorphisms); furthermore, we found evidence of biallelic transcription in our cell lines. However, we did not detect any deleterious mutations. These data suggest that the p73 gene is unlikely to be essential in melanoma tumorigenesis.