Arabidopsis and Brassica comparative genomics: sequence, structure and gene content in the ABI-Rps2-Ck1 chromosomal segment and related regions.

The region corresponding to the ABI1-Rps2-Ck1 segment on chromosome 4 of Arabidopsis thaliana was sequenced in Brassica oleracea. Similar to A. thaliana, the B. oleracea homolog BoRps2 is present in single copy. The B. oleracea orthologous segment was located on chromosome 4 and can be distinguished by the presence of an N-myristoyl transferase coding gene (N-myr) between the Rps2 and Ck1 (BoCk1a) genes. The N-myr homologs in Arabidopsis are on chromosomes 2 and 5. Additional homologs for Ck1 are located on these two chromosomes. A second Ck1 homolog found on B. oleracea (BoCk1b) chromosome 7 served to define another orthologous segment located in Arabidopsis chromosome 1. The two segments displayed identical gene content and order in both species, namely BoCK1b, a gene encoding a hypothetical protein (BohypothA) and transcription factor eiF4A. High levels of sequence identity were observed for the coding sequences of all genes examined. Although in general larger spacers were found in Brassica than in A. thaliana, this was not always the case. Promoters were poorly conserved, except for several sequence stretches of a few nucleotides. Comparative sequencing revealed microsyntenic changes resulting from chromosomal structural rearrangements, which are often undetectable by genetic mapping.

Fine sequence analysis of 60 kb around the Arabidopsis thaliana AtEm1 locus on chromosome III.

The Arabidopsis thaliana Em1 gene has been mapped to the lower arm of chromosome III. Fine analysis of 60 kb around this gene, based largely on identification and sequencing of cognate cDNAs, has allowed us to identify 15 genes or putative genes. Cognate cDNAs exist for ten of these genes, indicating that they are effectively expressed. Analysis by sequence alignment and intracellular localization prediction programs allows attribution of a potential protein product to these genes which show no obvious functional relationship. Comparison of the true exon/intron structure based on cDNA sequences with that proposed by three commonly used prediction programs shows that, in the absence of further information, the results of these predictions on anonymous genomic sequences should be interpreted with caution. Examination of the non-coding sequence showed the presence of a novel repeated, palindromic element. The results of this detailed analysis show that in-depth studies will be necessary to exploit correctly the complete A. thaliana genome sequence.

Rice lipid transfer protein (LTP) genes belong to a complex multigene family and are differentially regulated.

Several cDNA clones encoding three different lipid transfer proteins (LTPs) have been isolated from rice (Oryza sativa L.) in order to analyse the complexity, the evolution and the expression of the LTP gene family. The mature proteins deduced from three clones exhibited a molecular mass of 9 kDa, in agreement with the molecular mass of other LTPs from plants. The clones were shown to be homologous in the coding region, while the 3' non-coding regions diverged strongly between the clones. The occurrence of at least three small multigene families encoding these proteins in rice was confirmed by Southern blot analysis. When compared with each other and with LTPs from other plants, the cluster including rice LTPs and other cereal LTPs indicated that these genes duplicated rather recently and independently in the different plant phyla. The expression pattern of each gene family was also investigated. Northern blot experiments demonstrated that they are differentially regulated in the different tissues analysed. Components such as salt, salicylic acid and abscisic acid were shown to modulate Ltp gene expression, depending on tissues and gene classes, suggesting a complex regulation of these genes.

The Arabidopsis thaliana cDNA sequencing projects.

Nearly 30000 Arabidopsis thaliana EST (Expressed Sequence Tags) have been produced by a French and an American consortium. Despite redundancy these sequences tag about half the expected Arabidopsis genes. Approximately 40% of the non-redundant EST can be assigned a putative function by simple homology search. This programme allowed the identification of a large number of genes which would have been very difficult to isolate by other classical techniques. It considerably stimulated many areas of plant biology by the rapid discovery a large number of genes, by revealing multigene families and by allowing the analysis of differential expression of the different members. Finally this programme facilitated construction of physical maps of the chromosomes and opened the way for complete sequencing of the Arabidopsis genome and comparative mapping of the major plant crops.

The Arabidopsis thaliana cDNA sequencing projects.

Nearly 30000 Arabidopsis thaliana EST (Expressed Sequence Tags) have been produced by a French-American consortium. Despite redundancy, these sequences tag about half of the expected Arabidopsis genes. Approximately 40% of the non-redundant EST can be assigned a putative function by simple homology search. This programme allowed the identification of a large number of genes which would have been very difficult to isolate by other classical techniques. It considerably stimulated many areas of plant biology by the rapid discovery a large number of genes, by revealing multigene families and by allowing the analysis of differential expression of the different members. Finally this programme facilitated construction of physical maps of the chromosomes and opened the way for complete sequencing of the Arabidopsis genome and comparative mapping of the major plant crops.

Further progress towards a catalogue of all Arabidopsis genes: analysis of a set of 5000 non-redundant ESTs.

Nearly 7000 Arabidopsis thaliana-expressed sequence tags (ESTs) from 10 cDNA libraries have been sequenced, of which almost 5000 non-redundant tags have been submitted to the EMBL data bank. The quality of the cDNA libraries used is analysed. Similarity searches in international protein data banks have allowed the detection of significant similarities to a wide range of proteins from many organisms. Alignment with ESTs from the rice systematic sequencing project has allowed the detection of amino acid motifs which are conserved between the two organisms, thus identifying tags to genes encoding highly conserved proteins. These genes are candidates for a common framework in genome mapping projects in different plants.

Sequencing and mapping the Arabidopsis genome: a weed model for real crops.

Arabidopsis is a crucifer weed with a small genome of about 120 Mbp which has been chosen as a model species for plant molecular genetics. Four years ago, a consortium of nine French laboratories, including ours, initiated a project aimed at mapping the transcribed regions of the genome. The strategy employed was to systematically and randomly sequence cDNA clones isolated from libraries made from different tissues and organs of plants grown under various physiological conditions. The consortium released about 7,000 expressed sequenced tags (ESTs) in the dbEST database corresponding to approximately 3,500 unique genes. In the next phase of the programme, a YAC library with average inserts of 500 kbp has been prepared. We have now started to use the EST information to map the cDNA clones on these YACs. The most recent aspect of Arabidopsis sequencing is the ESSA (European Scientists Sequencing Arabidopsis) project, in which the aim is to describe 2.5 Mbp by the end of 1996. Genomic sequencing has revealed a very high gene density. Comparison of present genomic sequencing results with the EST data suggests that up to half of the genes might already be tagged with an EST. In collaboration with Carlos Quiros' group in Davis we have also analysed the conservation of a 30 kbp locus (Em 1, a late embryogenesis abundant protein gene) on chromosome 3 between Arabidopsis and several Brassica species. Progress on these various aspects will be reviewed. We shall also present some sequence comparisons between Arabidopsis and rice ESTs. These results suggest that it should be possible in the very near future to map a pool of common genes onto many different plant genomes. This should provide a common framework to integrate maps from different species and facilitate mapbased cloning of genes of agronomical importance.

A chlorophyll synthetase gene from Arabidopsis thaliana.

During the course of an Arabidopsis thaliana genome sequencing project, we identified a gene, G4, with a derived amino acid sequence showing homology to the product of the Rhodobacter capsulatus bchG locus which is involved in the esterification of bacteriochlorophyllide with geranylgeraniol. The relationship between this gene and bchG was confirmed by the isolation and analysis of a corresponding full-length cDNA. Comparison of genomic and cDNA sequences indicated that the gene is made up of 14 exons, some of them being very short. Southern and Northern analyses showed that this sequence represents a single-copy gene and its transcript is detected only in green or greening tissues. Both homologies and expression data suggest that this gene encodes a chlorophyll synthetase, one of the last enzymes of chlorophyll biosynthesis, and thus represents a new example of a nuclear gene encoding an enzyme of this pathway in higher plants.

GASA, a gibberellin-regulated gene family from Arabidopsis thaliana related to the tomato GAST1 gene.

A multiple gene family of at least four members, related to a GA-stimulated transcript (GAST1) from tomato, was characterized in Arabidopsis thaliana by analysing four related cDNAs, named GASA1 to GASA4. The corresponding peptides display comparable structural features: (1) a putative signal peptide of 18 to 23 residues; (2) a highly divergent hydrophilic region of about 22 amino acids; (3) a conservative 60 amino acid C-terminal domain containing 12 cysteines. This organization has also bean shown in two related peptides from tomato, GAST1 found in shoots and RSI-1 found in early lateral roots. Southern blot hybridization patterns showed single-copy genes for all four members of the GASA family. Accumulation of the various transcripts, monitored by northern blot hybridization, indicated that the various genes are expressed differentially in plant organs: Specific mRNAs were mostly detected in flower buds and immature siliques in the case of GASA1, in siliques and dry seeds in the case of GASA2 and 3, and in growing roots and flower buds in the case of GASA4. At least two of the GASA genes are activated in GA-deficient mutant ga5, as early as 4 to 8 h after spraying with 50 microM GA3. The complex patterns of expression and regulation of the various genes suggest that the related peptides are involved in a developmental regulation process in Arabidopsis.

Identification of the most represented repeated motifs in Arabidopsis thaliana microsatellite loci.

The major simple sequence repeats present in the Arabidopsis genome were identified by Southern hybridizations with 49 oligonucleotide probes matching all the possible combinations of motifs up to 4 nucleotides long. The method used allowed us to perform all the hybridizations under the same temperature conditions. A good correlation was observed with the data obtained from database analysis, indicating that the method can be useful for identifying the major classes of microsatellite loci in species for which few or no sequence data are available. AG/CT, AAG/CTT, ATG/CAT and GTG/CAC are the major motifs present in the Arabidopsis genome that can be used as convenient probes to isolate microsatellite loci by screening libraries. AAG/CTT is the more frequent of these motifs, and its relative frequency in Arabidopsis is much higher than averagely found in the plant kingdom. About 8% of the cDNA clones from an immature silique library contains AG/CT, AAG/CTT or ATG/CAT microsatellite loci. Several microsatellite loci were isolated by screening genomic and cDNA libraries. Twenty-six tri-nucleotide loci were PCR amplified from four different ecotypes, and polymorphism was observed for 12 of them; 10 loci showing two alleles and 2 loci showing three alleles.

Characterization of a rice gene coding for a lipid transfer protein.

The cloning and sequence analysis of a gene that encodes a lipid transfer protein (LTP) from rice is reported. A genomic DNA library from Oryza sativa was screened using a cDNA encoding a maize LTP. One genomic clone containing the gene (Ltp) was partially sequenced and analyzed. The open reading frame is interrupted by an 89-bp intron. From the results of Southern hybridizations, Ltp appears to be a member of a small multigenic family. Transcripts of the corresponding gene were detected in several tissues including coleoptile, leaf, endosperm, scutellum and root. The transcription start point was determined by primer extension. The deduced amino-acid sequence of the Ltp product is shown to be homologous to LTPs from other crops.

Characterization of the ribosomal DNA units in two related Prunus species (P. Cerasifera and P. Spinosa).

The genetic relationships between two Prunus species, involved in rootstock breeding, were examined at the level of the ribosomal RNA genes. Twenty clones of P. cerasifera, a diploid species, and 12 clones of P. spinosa, a tetraploid wild species, were studied. The use of three heterologous ribosomal DNA probes covering different regions of the ribosomal tandem repeats enabled us to construct restriction maps for EcoRI and BamHI. We identified two unit types (unit I and unit II) in P. cerasifera. In P. spinosa, P. cerasifera units were present in addition to a third ribosomal unit type (unit III). These results appeared to confirm previous cytological studies (Salesses 1973) indicating that one of the genomes in P.spinosa has homology with the one from P. cerasifera.

Fine structure and evolution of the rDNA intergenic spacer in rice and other cereals.

The intergenic spacer of a rice ribosomal RNA gene repeating unit has been completely sequenced. The spacer contains three imperfect, direct repeated regions of 264-253 bp, followed by a related but more highly divergent region. Detailed analysis of the sequence allows the presentation of an evolutionary scenario in which the 264-253-bp repeats are derived from an ancestral 150-bp sequence by deletion and amplification. Comparison of the rice sequence with those of maize, wheat, and rye shows that, despite considerable divergence from the ancestral sequence, several regions have been highly conserved, suggesting that they may play an important role in the structure and/or expression of the ribosomal genes.

Two different Em-like genes are expressed in Arabidopsis thaliana seeds during maturation.

Using a radish cDNA probe, we have isolated and characterized two genomic clones from Arabidopsis thaliana (GEA1 and GEA6) encoding two different proteins that are homologous to the "Early methionine-labelled" (Em) protein of wheat. GEA1 differs from GEA6 and Em clones of wheat in that a sequence coding for 20 amino acid residues is tandemly repeated 4 times. These two genomic clones correspond to two genes named AtEm1 and AtEm6. Sequencing of several cDNA clones showed that both genes are expressed. The transcription start site was determined for both genes by RNase mapping. The site of polyadenylation is variable and there is no obvious consensus sequence for polyadenylation at the 3' ends of the genes. mRNA corresponding to GEA6 is present only in nearly dry and dry seeds, whereas the corresponding to GEA1 appears in immature seeds and is maximum in dry seeds. No expression of either gene could be detected in leaf, stem, or floral buds. Expression of both genes could be detected in immature seeds when the siliques were incubated with abscisic acid (ABA), demonstrating that both genes are ABA responsive. However, examination of the 5' upstream region does not reveal any extensive homology, suggesting that regulation of the two genes differs. In situ hybridization with a GEA1 probe demonstrated that the expression of this gene is essentially located in the provascular tissues of the cotyledons and axis of the dry seed as well as in the epiderm and outer layers of the cortex in the embryo axis.

The cruciferin gene family in radish.

In order to analyse the cruciferin gene family in radish a cDNA library was screened either with heterologous rapeseed probes or by differential screening and sequencing. We could identify six partial cDNA clones belonging to two different groups of cruciferin genes which do not cross-hybridize, and probably three distinct subfamilies. One of these classes corresponds to the previously described cruciferin from rapeseed and Arabidopsis. A gene corresponding to the second group, as well as its border sequences, was isolated from a radish genomic library and analysed in more detail. The cruciferin gene (cruRS) contains three introns and encodes a 479 amino acid protein. The transcription initiation site was determined. The expression of the different group of genes was studied by northern blot analysis: genes of both classes are expressed simultaneously and roughly at the same level between 25 and 35 days after flowering. Cruciferin gene copy number was estimated by Southern blot analysis. There appear to be seven or eight genes in one class and three in the other, located at different loci.

Genome specificity of rDNA spacer fragments from Oryza sativa L.

The intergenic spacer derived from a cloned rDNA unit from a cultivated rice was dissected into several subclones, which were used as probes to analyze sequence homologies between rDNA spacers from wild rice belonging to genome types AA, BB, CC, EE, FF, BBCC, and CCDD. This analysis allowed us to detect several regions with different degrees of homology. A series of 250-260 bp repeats is located in the central part of the AA spacer. This sequence cross-hybridizes with the BB, CC, BBCC, and FF genomes, but is absent in the EE and CCDD genomes. Regions proximal to 25S and 18S sequences are well conserved in all genomes. Finally, two adjacent sequences of 61 bp and 94 bp, located downstream from the repeats, have been found to have a narrow genomic specificity, restricted respectively to AA and FF genomes for the first one and to AA for the second. These data provide new information on the evolution of the ribosomal RNA gene spacer within a complex of related species, and add to our knowledge on the relationship between the various rice genomes.

Multiple mRNA coding for phospholipid-transfer protein from Zea mays arise from alternative splicing.

We have isolated a novel cDNA coding for maize phospholipid-transfer protein. The cDNA sequence is similar to the first one obtained by Tchang et al. [J. Biol. Chem. 263 (1988) 16849-16855] differing only by a mslal number of nucleotide substitutions and insertions. One of these insertions is 74 bp long and is flanked by consensus intron splicing sequences. The protein coded by the two cDNA has identical amino acids except in the C terminus. This difference derived from the presence of the 74-bp insert. The possible existence of an alternative splicing mechanism that could introduce heterogeneity in the sequence of these proteins is proposed.