Generation of region- and species-specific expressed gene probes from somatic cell hybrids

Increased expression of T-plastin gene in cisplatin-resistant human cancer cells: identification by mRNA differential display

The cellular resistance to the potent anticancer agent cis-diamminedichloroplatinum(II) (cisplatin) is thought to be mediated by multiple mechanisms. The technique of differential display of mRNAs was applied to various cisplatin-resistant cell lines and the corresponding parental sensitive human bladder, prostatic, and head and neck cancer cells in order to identify genes that underlie cisplatin resistance. Twenty-four clones were confirmed by Northern blot analysis to be expressed differentially between resistant and the corresponding sensitive cells. Partial DNA sequences of the eight clones that showed a threefold or greater increase in expression in either the resistant cells (seven clones) or sensitive cells (one clone) revealed that two were derived from the T-plastin gene and one from the tissue factor gene. The abundance of T-plastin mRNA in cisplatin-resistant T24/DDP10 cell was approximately 12 times that in the parental T24 cells. Transfection of T24/DDP10 cells with a vector encoding full-length T-plastin antisense RNA demonstrated that reduced T-plastin expression was associated with increased sensitivity to cisplatin. These results are consistent with the hypothesis that several mechanisms participate cooperatively in the acquisition of cisplatin resistance in human cancer.

An expression-independent catalog of genes from human chromosome 22

To accomplish large-scale identification of genes from a single human chromosome, exon amplification was applied to large pools of clones from a flow-sorted human chromosome 22 cosmid library. Sequence analysis of more than one-third of the 6400 cloned products identified 35% of the known genes previously localized to this chromosome, as well as several unmapped genes and randomly sequenced cDNAs. Among the more interesting sequence similarities are those that represent novel human genes that are related to others with known or putative functions, such as one exon from a gene that may represent the human homolog of Drosophila Polycomb. It is anticipated that sequences from at least half of the genes residing on chromosome 22 are contained within this exon library. This approach is expected to facilitate fine-structure physical and transcription mapping of human chromosomes, and accelerate the process of disease gene identification.

Cloning mammary cell cDNAs from 17q12-q23 using interspecific somatic cell hybrids and subtractive hybridization

We have cloned human genes that are encoded in the region 17q12-q23 and expressed in breast tissue using interspecific somatic cell hybrids and subtractive hybridization. Two mouse microcell hybrids containing fragments of human chromosome 17 with a nonoverlap region at 17q12-q23 were generated by microcell transfer. Radiolabeled cDNA was synthesized from the hybrid cell containing the 17q12-q23 interval and was subtracted with an excess of RNA from the hybrid cell lacking the interval. Resulting cDNA probes enriched for sequences from 17q12-q23 were used to screen a human premenopausal breast cDNA library, and 60 cDNAs were identified. Three of these cDNAs mapped to the hybrid cell nonoverlap region. These cDNAs were expressed in mammary epithelial cell hybrids, although none appeared to be breast-specific. Sequence analysis of the cDNAs revealed that clone 93A represents a previously unidentified gene, clone 98C has homology to an expressed sequence tag from goat mammary tissue, and clone 200A is identical to the human homologue of the Drosophila melanogaster flightless-I gene. These genes map outside a 1-cM region linked to early onset familial breast cancer but may be useful genetic markers in the 17q12-q23 region.

Chromosome specific c-DNA libraries: reduction of unspecific priming events by purification of heteronuclear RNA

Chromosome specific c-DNA libraries greatly facilitate the isolation of disease associated genes which have been previously linked to particular chromosomes. Recently, several methods have been developed and employed for the isolation of transcribed sequences from specific human chromosomes and chromosome regions. Heteronuclear (hn) RNA from somatic human/rodent cell hybrids has been used as starting material to selectively prime the synthesis of human specific c-DNAs. A drawback of this method is the high number of rodent clones found in these chromosome specific c-DNA libraries. Here, we provide direct evidence that unspecific priming events account for the majority of these rodent clones. Using an Alu consensus primer hn-RNA human specific c-DNA libraries have been established and the specificity of Alu-priming has been evaluated. Using a variety of purification schemes for isolating hn-RNA we have significantly reduced the percentage of unspecific priming events. We also included a comparison of the hn-RNA yield from different somatic hybrids prior and after purification.

Isolation of genes from complex sources of mammalian genomic DNA using exon amplification

Modifications to exon amplification have been instituted that increase its speed, efficiency and reliability. Exons were isolated from target human or mouse genomic DNA sources ranging from 30 kilobases (kb) to 3 megabases (Mb) in complexity. The efficiency was dependent upon the amount of input DNA, and ranged from isolation of an exon for every 20 kb to an exon for every 80 kb of target genomic DNA. In these studies, several novel genes and a smaller number of genes isolated previously that reside on human chromosome 9 have been identified. These results indicate that exon amplification is presently adaptable to large scale isolation of exons from complex sources of genomic DNA.

Identification of new genes by systematic analysis of cDNAs and database construction

The large-scale collection of partial cDNA sequences is becoming a powerful tool in biology. Similarity or motif searches in DNA databases using these partial cDNA sequences have facilitated the discovery of new genes of interest. By collecting and registering large numbers of partial sequences with a well designed non-biased cDNA library, an expression profile of active genes in a particular tissue can be obtained. Tissue-specific or stage-specific genes can be discovered by comparing the profiles from different tissues or from a tissue at different stages of development, respectively. The compilation of such expression profiles enables genes to be mapped to the tissue(s) where they are actively transcribed. The large-scale collation of gene sequences actively expressed in the body into databases complements efforts directed towards the structural analysis of the genome, with the ultimate aim of decoding all the genetic information carried in the human genome. This cDNA strategy is also being widely applied to organisms other than man.

Identification of chromosome-specific sequence-tagged sites by Alu-PCR

Recently, methods have been developed for the isolation of expressed sequences from particular human chromosomes. Using Alu consensus sequences as primers, cDNA synthesis has been initiated from interspecies hybrid cell lines that contain single human chromosomes. Alu consensus sequences have also been utilized to amplify human genomic sequences via polymerase chain reaction (PCR). Here, we describe the use of Alu-PCR to isolate expressed sequences from human chromosomes selectively. Heteronuclear (hn) RNA is transcribed into cDNA by using poly-(dT)15 primer sequences. Subsequently, human specific cDNA sequences are amplified by Alu-PCR and cloned into pBluescript. To verify the chromosomal assignment, cloned PCR products are sequenced, converted into STS markers, and tested on a different somatic hybrid that contains human chromosome 22. The method provides a fast, reliable way to identify expressed sequence tagged sites from selected human chromosomes.