INTRODUCTION

Top Abstract Introduction Results Discussion Methods References

Molecular markers used in mapping experiments should ideally have a known function. This requirement is fulfilled best by the use of known genes in the form of genomic clones or cDNA clones. Many genetic maps based on molecular markers, such as RFLPs (restriction fragment length polymorphisms) include a number of known genes. However, the number of known genes available for a given organism is usually very limited. To circumvent this problem, many maps contain a large number of randomly selected cDNA clones. cDNAs have the additional advantage that they are frequently single or low copy and thus ideal for genetic mapping (Bernatzky and Tanksley 1986). Furthermore, they show a higher level of conservation between species than genomic clones (Zamir and Tanksley 1988) and allow more efficient cross-mapping in related species for the determination of synteny (Ahn and Tanksley 1993).

In recent years random cDNA clones have attracted considerable interest, as the sequencing of such clones provides a means to obtain a fast catalog of expressed genes from a given organism without sequencing the entire genome. In plants, major efforts are now under way to sequence random cDNAs in Arabidopsis and rice (Sasaki et al. 1994; Delseny et al. 1997). Many thousands of such fragments have been deposited in the respective databases. For ~30%-40% of the cDNA sequences a possible function can be deduced based on homologies to known genes from bacteria, animals, or plants (Delseny et al. 1997).

For all expressed genes, not only the DNA sequence and the possible function should be known but also their map position on the chromosomes. In the long term this will allow merging of the classical genetic map based on mutations with potential candidate genes for these mutants. A common repertoire of mapped cDNA clones will, in the future, enable us to study synteny even between distantly related species (Paterson et al. 1996) for which studies by cross-hybridization are very difficult. The availability of sequence information for mapped cDNA clones should make conclusions from hybridization studies more firm and testable.

Tomato (Lycopersicon esculentum) has one of the most densely populated genetic maps among plants. Currently, >1000 RFLP markers have been published by Tanksley et al. (1992), of which >300 are random cDNA clones. An additional 500 RFLP markers have been localized in reference to this map through the fact that all markers from potato can be readily transferred to tomato on the basis of extensive synteny (Gebhardt et al. 1991; Jacobs et al. 1995). We report here the results of the sequencing of ~90% of the mapped cDNA clones from the high-density tomato map.

	RESULTS

Top Abstract Introduction Results Discussion Methods References

cDNA Sequencing and Analysis

A total of >300 cDNA clones have been mapped onto the high-density map of tomato by Tanksley et al. (1992). These clones were derived from two different libraries. CD clones were generated from leaf tissue and CT clones from epidermal tissue. The clones were in four different vectors. For sequencing, only those clones were used that were generated in vectors with M13 or pBluescript primer sites. This excluded some of the early cDNA clones (Bernatzky and Tanksley 1986) because they were generated in pBR322. A total of 272 cDNA clones could be sequenced in this way from at least one direction. Sequences from another 15 clones were not included in the analysis because they contained obvious cloning artifacts that could not be resolved. From both sides 145 clones were sequenced, and for 73 clones the entire insert was sequenced. The largest clone completely sequenced was 896 bp.

Because these clones had already been placed onto the genetic map of tomato, it was anticipated that most should represent different genes. Nevertheless, for some cDNA clones duplicates could be identified (clones CT71 and CT210, CT72 and CT242, CT75 and CT166, CT88 and CT257, CT115 and CT218, CT154 and CT214, CT223 and CD61, CD18 and CD34). In these cases, the cDNA clones show a complex hybridization pattern upon which it was not possible to state previously that they are derived from the same gene.

Sequence Homologies

For 156 of the 275 analyzed clones (57%), significant nucleic acid and/or protein homologies were found with the respective databases. Of those, 125 clones showed matches with known DNA sequences and 141 clones showed matches with protein sequences with a BLAST score of <10¹⁰. Some sequenced clones showed only homology to other plant sequences at the DNA level and not at the protein level because the cDNA clones were not integrated directionally into the cloning vector and for these clones sequence information was obtained only from the untranslated 3' region with the poly(A) tail. If matches on the DNA and protein level were found, only those were considered significant that matched to the same protein type or corresponding genes. Table 1 shows a summary of the data for the clones with significant homologies. For 25 tomato genes already described we found a direct match with the database.

	DISCUSSION

Top Abstract Introduction Results Discussion Methods References

The availability of 272 sequenced cDNA clones from the tomato map, together with other previously mapped genes of known function (Pillen et al. 1996b), creates a framework of >350 markers for the tomato genome for which at least part of the DNA sequence is known. This information will be sufficient for the generation of PCR-based markers for most regions of the tomato genome. Such sequence-tagged sites (STSs) will function as genetic anchor markers (Inoue et al. 1994) and permit fast and high throughput analysis of loci in large populations (Schumacher et al. 1995). In plant breeding and genetic experiments, STS markers, for example, in the form of CAPs (cleaved amplified polymorphisms), are very useful to follow linked genes in an economical manner through generations (Konieczny and Ausubel 1993). Such markers are especially useful for preselecting recombinants in specific regions of the tomato genome for the purpose of high-resolution mapping of genes targeted for map-based cloning (Alpert and Tanksley 1996). At the same time, they provide starting points for the rapid isolation of large insert clones from tomato DNA libraries, such as yeast artificial chromosomes (YACs) (Martin et al. 1992; Bonnema et al. 1996; Pillen et al. 1996a) or bacterial artificial chromosomes (BACs) (Shizuya et al. 1992) and binary BAC (BiBAC) clones (Hamilton et al. 1996).

Considerable effort is currently spent on the comparative mapping of highly conserved cDNA clones across a wide range of plant taxa. This has revealed that in limited regions, synteny exists even between distantly related plant species (Paterson et al. 1996). Sequence data from mapped cDNA clones will eventually help to reveal synteny between such plant species. For example, using the information from these mapped tomato clones, it will be possible to identify genes or expressed sequence tags (ESTs) with very high sequence homology on the DNA and/or protein level to the model plant organism Arabidopsis thaliana. Such probes are very likely cross-hybridizing between Arabidopsis and tomato. If they are single copy in hybridization on both genomes they can be mapped comparatively in these species and it can be determined whether there is conservation of linkage in certain areas of their genomes. Similarly, such experiments can be expanded to additional plant species for which large numbers of ESTs will be generated in the future. A carefully chosen approach involving sequence comparisons in combination with genetic mapping by hybridization is absolutely necessary for the study of synteny between distantly related plant species to discriminate between orthologous and paralogous genes (Tatusov et al. 1997). Large numbers of cDNA clones are currently sequenced and mapped in some plant genomes. For rice, EST sequences were used for the construction of a high-density genetic map (Kurata et al. 1994). Large numbers of ESTs are mapped onto the genetic map or YAC contig map of A. thaliana (Agyare et al. 1997). With the sequencing of the entire Arabidopsis and rice genomes in the forseeable future, it will be easier to study exclusively orthologous sequences from those two genomes in comparison to data from other plant species.

	METHODS

Top Abstract Introduction Results Discussion Methods References

cDNA Clones

All cDNA clones from the map of Tanksley et al. (1992) that have been cloned in pUC vectors, the pBluescript SK vector, or the pCR II (Invitrogen) vector were used for this study. Insert sizes were confirmed prior to sequencing by PCR using M13 forward and reverse sequencing primers.

Plasmid Preparation and Sequencing

Plasmids were prepared according to standard protocols using Qiagen colums (Qiagen, Hilden, Germany) from 5-ml cultures. Plasmid DNA was sequenced using commercially available sequencing kits and analyzed on ALF (automated laser fluorescence) sequencers (Pharmacia) and ABI sequencers (Perkin Elmer) using either M13 forward (pUC vectors) or SK primers (pBluescript). Most of the clones were also sequenced from the opposite side using M13 reverse (pUC vectors) or KS primers (pBluescript).

Sequence Analysis

Raw sequences were transferred into the sequence analysis program DNAsis (Hitachi) and edited for vector sequences, poly(A) tails, and other cloning artifacts. If sequencing was performed from both sides of a clone, it was determined whether the two sequences overlap and in such cases they were edited and merged into a single file. The edited sequences were analyzed. The DNA sequence and the translated protein sequences were compared to all available DNA and protein sequences using the NIH BLAST server (BLASTN and BLASTX). Matches with a score of <10¹⁰ were considered to be significant. Accession numbers correspond to the respective entries in the GenBank Nucleotide Sequence and GenBank Protein databases, respectively.