Search

Here, we describe a fast, simple method for constructing fulllength cDNA libraries using SMART™ technology. This novel procedure uses the template-switching activity of Moloney murine leukemia virus (MMLV) reverse transcriptase to synthesize and anchor first-strand cDNA in one step. Following reverse transcription, three cycles of PCR are performed using a modified oligo(dT) primer and an anchor primer to enrich the cDNA population for fulllength sequences. Starting with 1 μg human skeletal muscle poly(A)+ RNA, a cDNA library was constructed that contained 3 × 106 independent clones with an average insert size of 2 kb. Sequence analysis of 172 randomly selected clones showed that 77% of cDNA clones corresponding to known genes contained intact open reading frames. The average length of complete open reading frames was 2.4 kb. Furthermore, 86% of the full-length clones retained longer 5′ UTR sequences than the longest 5′ end deposited in the GenBank® database. cDNA libraries generated using this method will be useful for accelerating the collection of mRNA 5′ end sequence information, which is currently very limited in GenBank.

Here, we describe the application of a RecA-based cloning technology to generate full-length cDNA libraries enriched for genes that are differentially expressed between tumor and normal tissue samples. First, we show that the RecA-based method can be used to enrich cDNA libraries for several target genes in a single reaction. Then, we demonstrate that this method can be extended to enrich a cDNA library for many full-length cDNA clones using fragments derived from a subtracted cDNA population. The results of these studies show that this RecA-mediated cloning technology can be used to convert subtracted cDNAs or a mixture of several cDNA fragments corresponding to differentially expressed genes into a full-length library in a single reaction. This procedure yields a population of expression-ready clones that can be used for further high-throughput functional screening.

An efficient cDNA amplification procedure is described for determining of the 5′ and 3′ ends of mRNAs and cloning full-length cDNAs. In this approach, a double-stranded (ds) adaptor is ligated to both ends of a library of ds cDNA by T4 DNA ligase. This adaptor-ligated ds cDNA is then used to selectively amplify 5′- or 3′-cDNA fragments by PCR with a combination of gene-specific and adaptor-specific primers. This is a unified method for 5′and 3′rapid amplification of cDNA ends (RACE) from the same adaptor-ligated ds cDNA template. A specially designed adaptor combines features of “vectorette PCR” and “suppression PCR” technologies that significantly reduce background during amplification. The application of “long and accurate PCR” (LA PCR) technology makes possible the amplification of large RACE products and full-length cDNAs with high fidelity to the original mRNA. We investigated efficacy and limitations of this PCR-based approach for cDNA cloning by amplification of 5′- and 3′-RACE fragments and full-length cDNAs of three members of the abundant human actin gene family (1.3–1.9 kb), the medium abundance transferrin receptor mRNA (5.0 kb) and the low-medium abundance insulin-like growth factor II receptor mRNA (9.1 kb).

We have developed a new strategy for differential screening of genes that are expressed in two or more tissues, and have used it to identify genes that are preferentially expressed in both testis and ovary. In this approach, testis-specific cDNAs were first generated using the suppression subtractive hybridization technique, cloned and arrayed in microplates. The inserts from the subtracted testis-specific library were amplified by PCR and spotted on filters. The dot blots were screened by hybridization using either testis- or ovary-specific subtracted cDNA mixtures as probes. By screening about 2000 putative clones from the subtracted testis-specific library, we identified 14 candidate genes that hybridized to both cDNA probes. Northern blot analysis showed that 12 clones were preferentially or exclusively expressed in testis and ovary. Nucleotide sequence analysis indicated that three of the identified clones represented cDNAs from novel genes.

The identification of differentially expressed genes enables us to understand the molecular mechanisms associated with disease, conditions of stress, drug treatments and developmental processes. Microarrays provide a powerful tool for studying these complex phenomena. Verification of differentially expressed genes and correlation with biological function, which is usually done by northern blot analysis, RNase protection assay or RT-PCR, is the bottleneck in all these protocols. We developed a new type of cDNA array for high-throughput expression profiling of multiple tissues and blood samples (i) for confirmation analysis of statistically significant number of clinical samples (ii) from limited amount of starting material, and (iii) with detailed clinical data from each individual. In contrast to traditional cDNA arrays, these arrays are spotted with a complex cDNA representing the entire mRNA message expressed in a given tissue or blood sample. cDNAs for these arrays were generated using SMART technology and accurately represent the original mRNA population, producing specific and quantitative signals during hybridization. cDNAs on Disease Profiling Arrays were derived from total RNAs of diseased and normal tissues or different blood fractions of patients. These cDNAs were spotted onto nylon membranes along with positive and negative controls. The arrays were then hybridized with gene-specific probes. Hybridization results revealed disease-related as well as patient-specific gene expression patterns between different disease types for these genes. These studies demonstrate that Disease Profiling Arrays are suitable for high-throughput studies comparing the relative abundance of a target gene, for simultaneously detecting differentially expressed genes in a wide variety of tissues and blood samples, and can be used down-stream from cDNA microarrays for confirmation analysis.

We have established a method for preparing cDNA sublibraries enriched in sequences from specific chromosome regions, called selection of hybrids by affinity capture (SHAC). This procedure can be described in two stages. In the first stage, a particular chromosome region, in this study mouse chromosome 11, was microdissected, followed by PCR amplification with a universal degenerate primer. This material is referred to as the “target” DNA. In the second stage, a mouse liver cDNA library with unique linker–adapter ends, referred to as the “source” cDNA, was hybridized to the biotin-labeled target DNA prepared during the first stage. The resulting DNA duplexes were captured by streptavidin-coated magnetic beads. The cDNAs were released from their biotin-labeled target homologs by alkaline denaturation and recovered by PCR amplification. These cDNAs were referred to as the SHACcDNAs. Specificity of the SHACcDNA to chromosome 11 was verified by FISH analysis. To examine representation of the SHACcDNA, we confirmed the presence of seven genes or single-copy DNA segments known to be localized on mouse chromosome 11, using a dot blot assay. In addition, a second round of SHAC was performed to achieve even higher specificity for the resulting chromosome 11 SHACcDNA. The SHAC technology should facilitate construction of cytogenetically defined cDNA libraries and should assist in the fields of gene discovery and genome mapping.

We have developed preparative in situ hybridization (Prep-I.S.H.) of complex DNA populations to mitotic chromosomes as a means of generating chromosome region-specific cDNA subpopulations. Prep-I.S.H. is a combination of two cytogenetic techniques: hybridization of DNA molecules to mitotic chromosomes, and chromosome microdissection. Here we provide technical details of the procedure and describe its application to the human chromosome 11q22–23 region containing the ataxia telangiectasia genes. Prep-I.S.H. has a number of applications in studies of gene expression and genome organization, including efficient cytogenetic sorting of tissue-specific cDNAs and in dramatically reducing the number of candidate genes to aid in gene discovery. Prep-I. S.H. provides a technical approach to a “positional candidate” strategy for gene discovery that differs from positional cloning by focusing on analysis of the genes in a chromosomal region of interest rather than on detailed analysis of physical and genetic maps.

We have developed RNA expression microarrays (REMs), in which each spot on a glass support is composed of a population of cDNAs synthesized from a cell or tissue sample. We used simultaneous hybridization with test and reference (housekeeping) genes to calculate an expression ratio based on normalization with the endogenous reference gene. A test REM containing artificial mixtures of liver cDNA and dilutions of the bacterial LysA gene cDNA demonstrated the feasibility of detecting transcripts at a sensitivity of four copies of LysA mRNA per liver cell equivalent. Furthermore, LysA cDNA detection varied linearly across a standard curve that matched the sensitivity of quantitative real-time PCR. In REMs with real samples, we detected organ-specific expression of albumin, Hnf-4 and Igfbp-1, in a set of mouse organ cDNA populations and c-Myc expression in tumor samples in paired tumor/normal tissue cDNA samples. REMs extend the use of classic microarrays in that a single REM can contain cDNAs from hundreds to thousands of cell or tissue samples each representing a specific physiological or pathophysiological state. REMs will extend the analysis of valuable samples by providing a common broad based platform for their analysis and will promote research aimed at defining gene functions, by broadening our understanding of their expression patterns in health and disease.

We have previously shown that treatment of the human leukemic cell line K562 with the tumor promoter 12-O-tetradecanoyl phorbol 13-acetate (TPA) leads to a reduction in the expression of two surface glycoproteins, Gp-105 and glycophorin (Fukuda, M. (1981) Cancer Res. 41, 4621-4628). In this report, we examine the mechanism for reduction of glycophorin expression. By immunoprecipitation of metabolically labeled proteins, we found an 11-fold reduction in the incorporation of [3H]glucosamine into glycophorin after 48 h of TPA treatment. This was found to closely correlate with a reduction in [35S]methionine incorporation, suggesting that regulation occurs at the level of protein synthesis and not glycosylation. We also found that TPA decreased the incorporation of [3H]glucosamine into glycophorin in another leukemic cell line, HEL. A time course showed that there was a 3-fold reduction in glycophorin biosynthesis as early as 4 h after TPA treatment. The level then decreased to approximately 10% of the untreated control levels after 12 h of treatment. The reduction in glycophorin biosynthesis was found to be reversible following removal of TPA from the culture medium. By immunoprecipitation of in vitro translation products directed by purified total cellular RNA, we found that there is a corresponding decrease in glycophorin mRNA activity. Glycophorin mRNA activity was extensively reduced as early as 1 h after TPA treatment and by 12 h was nearly undetectable. Interestingly, the size of the primary translation product was found to be 8 kDa larger than the fully processed apoprotein. These results are consistent with the transcriptional regulation of glycophorin expression by TPA.

The tumor-promoting phorbol ester 12-O-tetradecanoyl phorbol 13-acetate (TPA) was found to induce the biosynthesis and deposition of two cytoskeletal components in the human erythroleukemic cell line K562 with apparent Mr = 55,000 and 46,000. A time course showed that maximal changes in these cytoskeletal components occurred as early as 24 h after treatment and were maintained at least as long as 72 h. By immunoprecipitation, one of the induced cytoskeletal components was identified as vimentin. There was approximately a 10-fold induction of vimentin biosynthesis following TPA treatment. The other TPA-induced protein was found to co-migrate with purified actin in sodium dodecyl sulfate-polyacrylamide gels. Neither of these proteins was induced in the closely related human erythroleukemic cell line, HEL. By in vitro protein synthesis directed by total cellular RNA isolated from K562 cells, induction of vimentin biosynthesis was found to correlate with increases in the level of vimentin mRNA activity. A time course showed that vimentin mRNA activity was detectably elevated as early as 3 h after TPA treatment and reached a maximum by 12 h then the level decreased. By RNA dot blot and Northern gel hybridization using a cloned human actin cDNA, we found that there was also an induction of actin mRNA. A time course showed that there was an elevation of actin mRNA as early as 1 h after TPA treatment. The level then increased to a maximum at 6 h, after which the level decreased. These results are consistent with the induction of vimentin and actin gene transcription by TPA in K562 cells.

Human glycophorin A belongs to a family of structurally related cell surface glycoproteins which are expressed in erythroid cells. We have recently isolated several human glycophorin A-specific cDNA clones and have determined their nucleotide sequence (Siebert, P. D., and Fukuda, M. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 1665-1669). In this report by using cDNA and exact sequence synthetic oligodeoxyribonucleotides as hybridization probes, we provide evidence that human glycophorin A and B are encoded by separate and distinct genes present as single copies in the human genome. Furthermore, we show that the expression of the glycophorin A and B genes are coordinately regulated by tumor-promoting phorbol ester, 12-O-tetradecanoylphorbol-13-acetate.