Orthology between genomes of Brachypodium, wheat and rice

Cereals constitute the most important group of cultivated plants, and are known to have diverged from a common paleopolyploid ancestor ~45–47 million years ago (Mya) [1]. Despite this, a remarkable overall structural and functional similarity exists among different cereal genomes [2, 3], although the size of these genomes differs greatly, ranging from 430 Mb in rice (Oryza sativa) to 16,000 Mb in hexaploid wheat (Triticum aestivum). Due to its small size and availability of whole genome sequence, rice has been used as a model system for a variety of experimental studies including map-based cloning [4]. However, recent studies resolved further the dynamic changes in rice genome sequences, thus questioning the utility of rice as a model crop [5], and necessitating the need for search of a more efficient model system.

Brachypodium distachyon, a small temperate grass (sub-family Pooideae) has recently emerged as a better model system for the study of temperate grasses. This is particularly, due to several of its desirable biological features and its phylogenetic position [6, 7]. It is postulated that relative to rice genome, Brachypodium genome will exhibit a much higher level of colinearity and synteny to the genomes of temperate cereal crops. In the present study, the available Brachypodium EST contigs (bEST contigs) and supercontigs were utilized to explore further the utility of the Brachypodium genome as a model for carrying out comparative genomics studies in cereals in general, and for wheat genomics in particular. The relationship of Brachypodium genome with wheat and rice genomes has been examined for this purpose, and improved criteria of sequence similarity search were used for more accurate estimation of similarity [8].

In the present study, EST sequences from Brachypodium were utilized to find out the degree of similarity of Brachypodium genome with EST/genomic sequences of wheat and rice. The orthologous wheat sequences thus identified were also utilized to study the relationship of wheat genome sequences with Brachypodium supercontigs. We have also taken note of the comparisons of chloroplast genomes among eight grass species, which were included in the report on Brachypodium chloroplast genome sequence that was recently worked out [9].

Analysis of mapped wEST contigs that matched bEST contigs

The above 449 bEST contigs were homologous with a corresponding number of wESTs carrying 1,154 bin-mapped loci or regions giving an average of 2.57 loci per wEST contig (Figures 1, 2). The distribution of ortholoci on the three wheat sub-genomes (A, B and D) and among the seven homoeologous groups of chromosomes (Table 1) was non-random (P << 0.05), when the known chromosome lengths and their DNA contents were used as the basis [10]. The distribution of ortholoci on long and short arms of the chromosomes (excluding 37 loci, which could not be assigned to individual arms) was also non-random (P < 0.05). This non-random distribution of ortholoci is, however, based on limited data.

	Sub-genome
Homoeologous group	A	B	D	Total
1	36	46	46	128
2	53	51	62	166
3	43	54	52	149
4	59	61	66	186
5	62	65	49	176
6	57	51	41	149
7	59	66	75	200
Total	369	394	391	1,154

Of the above 449 matched wEST contigs (orthologous to bEST contigs), 77 (17.2%) represented unique loci, and the remaining 372 (82.8%) detected multiple loci with 283 (76.1%) having multiple loci on homoeologous chromosomes and 89 (23.9%) having multiple loci on non-homoeologous chromosomes.

Of the 1,154 orthologous loci with known positions on wheat chromosomes, 1,094 (94.8%) loci were known to have earlier been assigned to 159 chromosome bins defined by deletion break points. The remaining 60 (5.2%) loci could be assigned only to individual chromosomes or their arms. A maximum of 386 loci (35.3%) were mapped in the proximal regions (60% of the arm length from centromere; C-0.60) followed by 331 loci (30.3%) mapped to the distal regions (40% terminal arm length; 0.60–1.00). The remaining 377 loci (34.4%) were mapped to the interstitial bins having proximal and distal regions.

The above 449 mapped wheat orthologs were also used for homology search among Brachypodium supercontigs. The wheat EST contigs located on homoeologous group 4 chromosomes had maximum homology (54.5% of mapped contigs) with the Brachypodium super_1 contig. In contrast, Brachypodium super_0 to 2 contigs had homology with wEST contigs dispersed on all the seven homoeologous groups, although no redundancy for wheat homologues was observed within the above supercontigs (Table 2).

Wheat homoeologous group	bEST mapped contig	Number of Brachypodium supercontigs showing orthology*	Supercontig showing maximum homology
1	55	11 (0–4,7–10,12,14)	2 (32.7%)
2	69	11 (0–6,9,10,14,15)	0 (44.9%)
3	63	11 (0–2,4–6,8–10,12,13)	4 (33.3%)
4	77	10 (0–3,6–8,11,15,189)	1 (54.5%)
5	68	11 (0–3,5–8,12,15,525)	0 (27.9%)
6	54	11 (0–9,11)	5 (42.5%)
7	63	15 (0–9,11,13,14,538)	0 (19.0%)

*Designated numbers for supercontigs are given in a parenthesis in each row

Orthology between bEST contigs and rice genome sequences

The BLASTN results of 3,818 bEST contigs against the rice genome sequences identified as many as 743 matching bEST contigs (see methods), which had homologues distributed on all the 12 chromosomes of rice. On the basis of relative length (Mb) of chromosomes and their arms [11], the ortholoci on 12 rice chromosomes/arms were non-randomly distributed (P <<< 0.05) (Table 3; Figure 3).

Rice Chromosome	1	2	3	4	5	6	7	8	9	10	11	12	Total
Number of loci	122	95	133	52	64	61	52	43	41	23	27	30	743

Conserved orthologous sequences among Brachypodium, wheat and rice

In the present study, 183 orthologous sequences were conserved among all the three species (Brachypodium, wheat and rice). As many as 126 of the 183 orthologous sequences also confirmed known homology between wheat-rice chromosomes. Functional annotation of these 183 orthologous sequences suggested that a majority (137; 74.8%) of these bEST contigs matched with proteins of known functions (see Additional file 1; Figure 4).

Transferability of conserved orthologous SSRs

In order to validate experimentally the conservation of Brachypodium SSRs among the genomes of wheat and rice, primer pairs for SSRs belonging to 12 orthologs were synthesized and used for PCR amplification of the SSRs (Table 4). All the 12 primer pairs gave amplification products in wheat and rice (Figure 5).

Brachypodium contig Id.	Motif	Primer sequence 5'-3'	Tm (°C)	Product size (bp)	Gene class
BDEST01P1_Contig9	(gct)5	L GCCTATGTTTCCGCAGAGAG R CCAGGCAAGAAGTTCCTGAG	59.98	203	Formate dehydrogenase
BDEST01P1_Contig1223	(cca)4	L AGCCAACTCTTGCAGCAAAT R TGTTGCTCCCTCCTTTTGAT	59.67	207	Endo-1,4-beta-glucanase Cel1
BDEST01P1_Contig1335	(tca)4	L GACGAGAGGTTGTGTTGGTG R ACAGGACACCGTCAGAGGAA	60.71	239	Putative uncharac-terized protein
BDEST01P1_Contig1574	(cgc)6	L CAAAACCCTAGCTGCCCTTC R TGCCAGTGCTTCTTGAAATG	59.99	123	Putative 60S ribos-omal protein L13E
BDEST01P1_Contig2089	(ggc)4	L GCTCTTCTCGCCCCTCTACT R CTCCATCTGGAAATCGCAGT	60.22	200	Hypothetical protein
BDEST01P1_Contig2416	(tcaaga)2	L CCGCACCTCAAGGACTACA R TCGGAGGAGATCTTGGTGAG	60.34	194	Succinate dehydrogenase
BDEST01P1_Contig2648	(ggt)4	L AAACCACTTGCCAAAACACC R GCTGCGGTTCTCCATGAC	60.37	248	Putative uncharac-terized protein
BDEST01P1_Contig3139	(tcg)4	L AGTCACCAAGGTCGTCAAGG R CCTTCGCTGCTCCATAGTCT	59.6	224	Putative ribosomal protein
BDEST01P1_Contig3247	(tggtgc)2	L AGTTGGAATGAGGGCATCAG R TTCAAGGCTCTCGAGTAGGG	59.57	214	Putative uncharac-terized protein
BDEST01P1_Contig3321	(gctcgc)2/N36/(ggc)4	L CACTTCGAGTTTCCCGTCAT R TTTTGCAGTGTCCACACCAT	60.01	244	Protein disulfide isomerase 2 precursor
BDEST01P1_Contig3721	(gct)4	L GGACTACTTTGGGGCTCACA R GGATTCATAACTGGCAACCA	59.44	180	Cytosolic 6-phosphogluconate dehydrogenase
BDEST01P1_Contig3747	(tcgcca)2	L AGGTCAACTCGGTCAACGAC R AGGTCAGCCCGTTGTTGTAG	60.17	192	Phenylalanine ammonia-lyase

Comparative genomics among grasses initially focused on the analysis of colinearity (gene order) and synteny (gene content) among DNA markers mapped on individual chromosome at a low resolution (10 cM). This led to the identification of 30 rice-independent linkage blocks involved in the constitution of all cereal genomes and allowed identification of a number of rearrangements within individual genomes [12]. However, due to the availability of whole genome sequence of rice, and substantial partial sequences from other cereal genomes, emphasis shifted to a comparison of nucleic acid sequences. In particular, sequences of ~7000 bin-mapped wESTs were aligned with rice genome sequences [13], allowing improved resolution and discovery of many more rearrangements.

Although rice worked well as a model for all grasses including wheat, and generated useful information, Brachypodium, belonging to subfamily Pooideae (wheat also belongs to Pooideae), is proposed as a better model than rice (subfamily Ehrhartoideae). Recent studies have suggested that relative to rice, Barchypodium is more closely related to wheat and barley and the colinearity between Barchypodium and wheat is better than that between wheat and rice [14, 15]. Chloroplast sequence-based phylogenetic analysis in eight grass species also suggested that Brachypodium is closer to the tribe Triticeae [9]. The possible estimated time of divergence between Brachypodium and Triticeae is also shorter (35 Mya) than that of divergence between wheat and rice (50 Mya) [16] thus supporting the view that Brachypodium is more closely related with the members of Triticeae.

During the present study, orthologous relationship among bEST contigs, wEST contigs and rice genome sequences was studied using improved criteria of sequence comparison. Observation of higher number of bEST contigs showing orthology with rice genome was mainly attributed to the fact that only a small fraction of wheat genome (0.02%) and almost complete rice genome (95%) were used for sequence comparison with the available Brachypodium EST contigs. If we take into account the proportion of the genome used for comparison, it may be concluded that wheat has higher level of orthology with Brachypodium than with rice.

The mapped loci in different deletion bins of a particular chromosome of wheat matched with same or different supercontigs of Brachypodium. For instance, wheat group 4 chromosomes are highly syntenic to Brachypodium super_1 contig (54.5%) than to other supercontigs, although super_1 contig showed homology with other homoeologous groups also. The mapping information of these Brachypodium supercontigs on individual Brachypodium chromosomes will be useful for developing markers specific to the targeted regions of wheat chromosomes.

It was also observed that although D sub-genome of wheat is smaller in size, the orthologous loci mapped on this sub-genome are no fewer than those mapped on sub-genome B, suggesting closer relationship between Brachypodium and Aegilops tauschii, the donor of the D sub-genome of hexaploid wheat.

The relative abundance of orthologous loci on proximal regions of chromosome arms in wheat is in agreement with the earlier studies in wheat and rice [17]. It seems that higher degree of sequence conservation coincides with the low recombination proximal regions, which is understandable, since high recombination in terminal regions will cause reshuffling of genes during evolution [18].

The results of the present study indicate that the availability of whole genome sequence of Brachypodium will be of enormous relevance for comparative genomics, gene annotation and evolutionary, structural and functional genomic studies of large genomes of the Triticeae.