A large scale analysis of cDNA in Arabidopsis thaliana: generation of 12,028 non-redundant expressed sequence tags from normalized and size-selected cDNA libraries.

For comprehensive analysis of genes expressed in the model dicotyledonous plant, Arabidopsis thaliana, expressed sequence tags (ESTs) were accumulated. Normalized and size-selected cDNA libraries were constructed from aboveground organs, flower buds, roots, green siliques and liquid-cultured seedlings, respectively, and a total of 14,026 5'-end ESTs and 39,207 3'-end ESTs were obtained. The 3'-end ESTs could be clustered into 12,028 non-redundant groups. Similarity search of the non-redundant ESTs against the public non-redundant protein database indicated that 4816 groups show similarity to genes of known function, 1864 to hypothetical genes, and the remaining 5348 are novel sequences. Gene coverage by the non-redundant ESTs was analyzed using the annotated genomic sequences of approximately 10 Mb on chromosomes 3 and 5. A total of 923 regions were hit by at least one EST, among which only 499 regions were hit by the ESTs deposited in the public database. The result indicates that the EST source generated in this project complements the EST data in the public database and facilitates new gene discovery.     

A Physical Map of Arabidopsis thaliana Chromosome 3 Represented by Two Contigs of CIC YAC, P1, TAC and BAC Clones

We have constructed a physical map of Arabidopsis thaliana chromosome3 by ordering the clones from CIC YAC, P1, TAC and BAC librariesusing the sequences of a variety of genetic and EST markersand terminal sequences of clones. The markers used were 112DNA markers, 145 YAC end sequences, and 156 end sequences ofP1, TAC and BAC clones. The entire genome of chromosome 3, exceptfor the centromeric and telomeric regions, was covered by twolarge contigs, 13.6 Mb and 9.2 Mb long. This physical map willfacilitate map-based cloning experiments as well as genome sequencingof chromosome 3. The map and end sequence information are availableon the KAOS (Kazusa Arabidopsis data Opening Site) web siteat http://www.kazusa.or.jp/arabi/.     

The size and sequence organization of the centromeric region of Arabidopsis thaliana chromosome 4.

We have determined the genome structure of the centromeric region of Arabidopsis thaliana chromosome 4 by sequence analysis of BAC clones obtained by genome walking, followed by construction of a physical map using DNA of a hypomethylated strain. The total size of the centromeric region, corresponding to the recombinant inbred (RI) markers between mi87 and mi167, was approximately 5.3 megabases (Mb). This value is over 3 Mb longer than that previously estimated by the Arabidopsis Genome Initiative (Nature, 408, 796-815, 2000). Although we could not cover the entire centromeric region by BAC clones because of the presence of highly repetitive sequences in the middle (2.7 Mb), the cloned regions spanning approximately 1 Mb at both sides of the gap were newly sequenced. These results together with the reported sequences in the adjacent regions suggest that the centromeric region is principally composed of a central domain of 2.7 Mb, consisting of mainly 180-bp repeats and Athila elements, and upper and lower flanking regions of 1.55 Mb and 1 Mb, respectively. The flanking regions were predominantly composed of various types of transposable elements, except for the upper end moiety in which a large 5S rDNA array (0.65 Mb) and central domain-like sequence are present. Such an organization is essentially identical to the centromeric region of chromosome 5 reported previously.     

The size and sequence organization of the centromeric region of arabidopsis thaliana chromosome 5.

We have determined the size of the centromeric region of Arabidopsis thaliana chromosome 5, which corresponds to the genetically defined centromere by Copenhaver et al. (Science, 286, 2468-2474, 1999) on the basis of restriction analysis. As a large clone gap was present in the previously constructed contig map of the centromeric region, the restriction map of this region was constructed using Asc I, Not I, Apa I and Pme I and genomic DNA from a hypomethylated strain. The size of the centromeric region finally estimated by combination with the sequence data of cloned regions at both sides was 4.35 megabases (Mb). This value is over 2 Mb longer than those estimated in our previous work and also by Copenhaver et al. Combing this centromeric region with the physical map previously constructed, the entire length of chromosome 5 becomes 31 Mb. Although the internal moiety of the centromeric region has not been sequenced yet because of extremely high repetition, the result of sequence analysis from both sides toward the inside strongly suggests that the centromeric region is composed of the central 2.9-Mb domain consisting of mainly 180-bp repeats and Athila retrotransposons and flanking regions containing various types of transposons. On the basis of these observations, a structural model for the centromeric region is discussed.     

A sequence-ready contig map of the top arm of Arabidopsis thaliana chromosome 3.

A fine physical map of the top arm of Arabidopsis thaliana chromosome 3 has been constructed by ordering P1, TAC and BAC clones using the sequences of a variety of DNA markers and end-sequences of clones. The marker sequences used in this study were derived from 58 DNA markers, 93 YAC end-sequences, and 807 end-sequences of P1, TAC and BAC clones. The entire top arm of chromosome 3, except for the centromeric and telomeric regions, was covered by a single contig 13.3 Mb long. This fine physical map will facilitate gene isolation by map-based cloning experiments as well as genome sequencing of the top arm of chromosome 3. The map and end-sequence information are available on the web site KAOS (Kazusa Arabidopsis data Opening Site) at [http://www.kazusa.or.jp/arabi/].     

Regional insertional mutagenesis of specific genes on the CIC5F11/CIC2B9 locus of Arabidopsis thaliana chromosome 5 using the Ac/Ds transposon in combination with the cDNA scanning method.

We have recently developed a novel cDNA selection method (the cDNA scanning method) to select cDNAs for expressed genes in specific regions of the genome [Hayashida et al. (1995) Gene 165: 155, Seki et al. (1997) Plant J. 12: 481]. The gene Ds is known to transpose mainly in its neighborhood. By combining the cDNA scanning method with this trait of Ds, we started functional analysis of region-specific expressed genes on the Arabidopsis thaliana genome. DNA fragments of yeast artificial chromosome (YAC) clones CIC5F11 and CIC2B9 on A. thaliana chromosome 5 were used for the selection of region-specific cDNAs. In total, 50 and 68 cDNA clones were selected from CIC5F11 and CIC2B9, respectively. In parallel, we transposed Ds from a donor T-DNA line, which was mapped on the CIC5F11/CIC2B9 locus of chromosome 5, and obtained Ds-transposed lines. To isolate Ds insertion mutants in the 10 specific genes identified by the cDNA scanning method, we carried out PCR-based screening of 100 Ds-transposed lines and found that 2 lines contain Ds mutations in the genes isolated. We also isolated Ds-flanking genomic DNAs by thermal asymmetric interlaced PCR (TAIL-PCR) in 153 Ds transposon-tagged lines. Southern blot analysis showed that 14% of the lines contained the transposed Ds in the CIC5F11/2B9 region. This suggests that this Ac/Ds transposon system is effective for region-specific insertional mutagenesis.     

Physical map-based sizes of the centromeric regions of Arabidopsis thaliana chromosomes 1, 2, and 3.

The sizes of the centromeric regions of Arabidopsis thaliana chromosomes 1, 2, and 3 were determined by construction of their physical maps on the basis of restriction analysis. As the reported centromeric regions contain large gaps in the middle due to highly repetitive sequences, appropriate probes for Southern hybridization were prepared from the sequences reported for the flanking regions and from the sequences of BAC and YAC clones newly isolated in this work, and restriction analysis was performed using DNA of a hypomethylated strain (ddm1). The sizes of the genetically defined centromeric regions were deduced to be 9 megabases (Mb), 4.2 Mb and 4.1 Mb, respectively (chromosome 1, from markers T22C23-t7 to T3P8-sp6; chromosome 2, from F5J15-sp6 to T15D9; chromosome 3, from T9G9-sp6 to T15M14; G. P. Copenhaver et al. Science, 286, 2468-2479, 1999). By combining the sizes of the centromeric regions previously estimated for chromosomes 4 and 5 and the sequence data reported for the A. thaliana genome, the total genome size of A. thaliana was estimated to be approximately 146.0 Mb.     

Structural analysis and complete physical map of Arabidopsis thaliana chromosome 5 including centromeric and telomeric regions.

Previously, we have reported a fine physical map of Arabidopsis thaliana chromosome 5, except for the centromeric and telomeric regions, by ordering clones from YAC, P1, TAC, and BAC libraries of the genome consisting of the two contigs of upper arm and lower arm, 11.6 M bases and 14.2 M bases, respectively. Here, the remaining centromeric and telomeric regions of chromosome 5 are completely characterized by the ordering of clones and PCR amplifications. Chromosome 5 of Arabidopsis thaliana ecotype Columbia is about 28.4 M bases long. The centromeric region is estimated at about 2 M bases long between two 5S-rDNA clusters. The 180-bp repeat region mainly consists of blocks of 180-bp tandem family and various type retroelements dispersed over a 500-kb region. The telomeric regions of chromosome 5 are characterized by PCR cloning, sequencing and hybridization. The telomere repeats at both ends are about 2.5-kb long and interestingly, telomere-associated repeats (approximately 700 bp) are found near both ends of chromosome 5.     

Characterization of expressed genes in the SLL2 region of self-compatible Arabidopsis thaliana.

Self-incompatibility in Brassica species is regulated by a set of S-locus genes: SLG, SRK, and SP11/SCR. In the vicinity of the S-locus genes, several expressed genes, SLL2 and SP2/ClpP, etc., were identified in B. campestris. Arabidopsis thaliana is a self-compatible Brassica relative, and its complete genome has been sequenced. From comparison of the genomic sequences between B. campestris and A. thaliana, microsynteny between gene clusters of Arabidopsis and Brassica SLL2 regions was observed, though the S-locus genes, SLG, SRK, and SP11/SCR were not found in the region of Arabidopsis. Almost all genes predicted in this region of Arabidopsis were expressed in both vegetative and reproductive organs, suggesting that the genes in the SLL2 region might not be related to self-incompatibility. Considering the recent speculation that the S-locus genes were translocated as a single unit between Arabidopsis and Brassica, the translocation might have occurred in the region between the SLL2 and SP7 genes.     

Structural analysis of Arabidopsis thaliana chromosome 5. IX. Sequence features of the regions of 1,011,550 bp covered by seventeen P1 and TAC clones.

In this series of projects sequencing the entire genome of Arabidopsis thaliana chromosome 5, non-redundant P1 and TAC clones have been sequenced according to the fine physical map, and as of May 7, 1999, the sequences of 16.2 Mb representing approximately 60% of chromosome 5 have been accumulated and released at our web site. In parallel, structural features of the sequenced regions have been analyzed by applying a variety of computer programs, and to date we have predicted a total of 2380 potential protein-coding genes in the 10,154,580 bp regions, which are covered by 142 P1 and TAC clones. In this paper, we newly analyzed the structural features of the 1,011,550 bp regions covered by additional 17 P1 and TAC clones, and predicted 298 protein-coding genes. The average density of the genes identified was 1 gene per 3394 bp. Introns were observed in 67% of the genes, and the average number per gene and the average length of the introns were 3.2 and 159 bp, respectively. The gene density became higher than the value estimated in the previously analyzed regions (1 gene per 4,267 bp), as the data in this paper were compiled based on a new standard of gene assignment including the computer-predicted hypothetical genes. The regions also contained 8 tRNA genes when searched by similarity to reported tRNA genes and the tRNA scan-SE program. The sequence data and information on the potential genes are available on the database KAOS (Kazusa Arabidopsis data Opening Site) at http://www.kazusa.or.jp/arabi/.     

Structural analysis of Arabidopsis thaliana chromosome 5. X. Sequence features of the regions of 3,076,755 bp covered by sixty P1 and TAC clones.

In our ongoing project to deduce the nucleotide sequence of Arabidopsis thaliana chromosome 5, non-redundant P1 and TAC clones have been sequenced on the basis of the fine physical map, and as of January, 2000, the sequences of 16.6 Mb representing approximately 60% of chromosome 5 have been accumulated and released at our web site. Along with the sequence determination, structural features of the sequenced regions have been analyzed by applying a variety of computer programs, and we already predicted a total of 2697 potential protein coding genes in the 11,166,130 bp regions, which are covered by 159 P1 and TAC clones. In this paper, we describe the structural features of the 3,076,755 bp regions covered by newly analyzed 60 P1 and TAC clones. A total of 715 potential protein coding genes were identified, giving an average density of the genes identified of 1 gene per 4001 bp. Introns were observed in 80% of the genes, and the average number per gene and the average length of the introns were 4.5 and 147 bp, respectively. These sequence features are nearly identical to those in our latest report in which the data were compiled based on a new standard of gene assignment including the computer-predicted hypothetical genes. The regions also contained 12 tRNA genes when searched by similarity to reported tRNA genes and the tRNA scan-SE program. The sequence data and information on the potential genes are available through the World Wide Web database KAOS (Kazusa Arabidopsis data Opening Site) at http://www.kazusa.or.jp/kaos/.     

Structural analysis of Arabidopsis thaliana chromosome 3. I. Sequence features of the regions of 4,504,864 bp covered by sixty P1 and TAC clones.

Based on the physical map of Arabidopsis thaliana chromosome 3 previously constructed with CIC YAC, TAC, P1 and BAC clones (Sato, S. et al., DNA Res., 5, 163-168, 1998), a total of 60 P1 and TAC clones were sequenced, and the sequence features of the resulting 4,504,864 bp regions were analyzed by applying various computer programs for similarity search and gene modeling. As a result, a total of 1054 potential protein-coding genes were identified. The average density of the genes identified was 1 gene per 4066 bp. Introns were observed in 77% of the genes, and the average number per gene and the average length of the introns were 3.9 and 156 bp, respectively. These sequence features are essentially identical to those of chromosome 5 in our previous reports, but the gene density was slightly higher than that observed for chromosomes 2 and 4. The regions also contained 10 tRNA genes when searched by similarity to reported tRNA genes and the tRNA scan-SE program. The sequence data and information on the potential genes are available through the World Wide Web database KAOS (Kazusa Arabidopsis data Opening Site) at http://www.kazusa.or.jp/kaos/.     

Analysis of Tag1-like elements in Arabidopsis thaliana and their distribution in other plants.

Analysis of genomic DNA of Arabidopsis Columbia (Col.) ecotype using a transposon Tag1-specific primer showed the presence of Tag1 homologues which was confirmed by Southern hybridization with a Tag1 probe. Further analysis showed that the homologue, 0.75 kb in length, had inverted repeats at both ends, 8-bp duplicated sequences at the site at which it is located and about 80% homology with Tag1, and was randomly distributed in the Arabidopsis genome. Based on these results, we concluded that these elements are non-autonomous variants of Tag1 and we termed this element sTag1. Using the polymerase chain reaction fragment hybridization technique, we found the distribution of such homologues in other plant species.     

Structural analysis of Arabidopsis thaliana chromosome 5. VIII. Sequence features of the regions of 1,081,958 bp covered by seventeen physically assigned P1 and TAC clones.

A total of 17 Pl and TAC clones each representing an assigned region of chromosome 5 were isolated from P1 and TAC genomic libraries of Arabidopsis thaliana Columbia, and their nucleotide sequences were determined. The length of the clones sequenced in this study summed up to 1,081,958 bp. As we have previously reported the sequence of 9,072,622 bp by analysis of 125 P1 and TAC clones, the total length of the sequences of chromosome 5 determined so far is now 10,154,580 bp. The sequences were subjected to similarity search against protein and EST databases and analysis with computer programs for gene modeling. As a consequence, a total of 253 potential protein-coding genes with known or predicted functions were identified. The positions of exons which do not show apparent similarity to known genes were also assigned using computer programs for exon prediction. The average density of the genes identified in this study was 1 gene per 4277 bp. Introns were observed in 74% of the potential protein genes, and the average number per gene and the average length of the introns were 4.3 and 168 bp, respectively. The sequence data and gene information are available on the World Wide Web database KAOS (Kazusa Arabidopsis data Opening Site) at http://www.kazusa.or.jp/arabi/.     

Nucleotide sequence and characterization of a cDNA encoding the acetohydroxy acid isomeroreductase from Arabidopsis thaliana

The primary structure of acetohydroxy acid isomeroreductase from Arabidopsis thaliana was deduced from two overlapping cDNA. The full-length cDNA sequence predicts an amino acid sequence for the protein precursor of 591 residues including a putative transit peptide of 67 amino acids. Comparison of the A. thaliana and spinach acetohydroxy acid isomeroreductases reveals that the sequences are conserved in the mature protein regions, but divergent in the transit peptides and around their putative processing site.