Cloning of TaAGOswith full-length cDNA
We performed a TBLASTX analysis in the NCBI EST (expressed sequence tag) database (http://www.ncbi.nlm.nih.gov/dbEST/) with two Arabidopsis Argonaute genes, AGO1 [GenBank: NM_179453] and AGO4 [GenBank: NM_128262]. Two groups of wheat ESTs proved to be highly homologous to Arabidopsis AGO1 and AGO4. Based on the conserved regions of their EST sequences, we designed specific primers (Additional file 1) for cloning the wheat Argonaute genes.
RT-PCR (reverse transcription-polymerase chain reaction) amplification was conducted with the primer combination of TaAGO1-1F and -1R. A 412-bp cDNA fragment (Figure 1-A) was cloned. To elongate the sequence of the wheat Argonaute gene, new primers were designed based on the cloned sequence and used in 5'- and 3'-RACE (Rapid Amplification of cDNA Ends). Although 5'-RACE analysis was performed several times, no satisfactory results were obtained. Therefore, we selected a genome-walking strategy to clone that 5' region. First, the genomic fragment corresponding to the cDNA region (Figure 1-A) was cloned and genome-walking primers (Additional file 1) were designed based on the sequence. Three rounds of genome-walking (Figure 1, GW1-3) were then conducted to obtain the 3958-bp genomic DNA. Finally, the 5'-cDNA region (Figure 1-B) was deduced by assembling exons (http://genes.mit.edu/GENSCAN.html). The 3'-cDNA fragment (Figure 1-C) was analyzed by 3'-RACE. The full-length cDNA was obtained by assembling the three fragments indicated above (Figure 1-A-C), and the cDNA sequence between AGO1-O1 and -O2 was confirmed by RT-PCR cloning and sequencing. This wheat AGO gene (3273 bp long) encodes a putative protein of 868 amino acid residues, which is highly homologous to rice OsAGO1b [Swiss-Prot: Q7XSA2.3], and was designated as TaAGO1b [GenBank: JQ805149].
Using primers TaAGO4-1F and -1R, we cloned a 498-bp wheat cDNA fragment via RT-PCR amplification. Sequencing results showed that this cDNA is very similar to Arabidopsis AGO4. Based on the cloned sequence, we obtained its full-length cDNA by 5'- and 3'-RACE. Sequence analysis indicated that the cDNA from our wheat AGO4 is 3157 bp long and encodes a putative protein of 916 amino acid residues. BLASTX analysis in NCBI revealed that it is highly homologous to Arabidopsis AGO4 [GenBank: AEC07929.1]. Thus, we designated it as TaAGO4 [GenBank: JQ805150].
Phylogenetic analyses of AGO plant proteins, including those from rice, Arabidopsis, and wheat, showed that TaAGO1b and TaAGO4 can be classified into two of the three groups described above (Figure 2). TaAGO1b, assigned to Group I, shares a high degree of homology with OsAGO1b, TaAGO1, OsAGO1a and AtAGO1. TaAGO4, as part of Group III, belongs to the same monophyletic subclass as rice OsAGO4a, OsAGO4b, and OsAGO6; and Arabidopsis AtAGO4, AtAGO6, AtAGO8, and AtAGO9. Three Arabidopsis AGOs (AtAGO2, 3, and 7) and two rice AGOs (OsAGO2 and 3) were sorted into Group II (Figure 2).
Characteristics of TaAGOs
The sequence analysis at ExPaSy (http://www.expasy.org) [13, 14] indicated that TaAGO1b includes 868 amino acids, with a predicted molecular weight of ~97.78 kDa and theoretical pI of 9.29. TaAGO4 is 916 amino acids long and has a theoretical pI of 9.12 and a molecular mass of about 102.10 kDa.
Both TaAGOs contain typical PAZ and PIWI conserved regions (Figure 3). The PAZ domain of TaAGO1b is composed of 114 amino acids (from Residue 207 – 320) and shares 82 to 95% homology with those of AGO1 proteins in other plant species. The PAZ domain of TaAGO4 comprises 76 amino acids (Residue 325 – 400) and has only limited homology with other plant AGO4 members. The C-terminal PIWI domain of TaAGO1b includes 322 amino acids (Residue 497 – 818), and has high similarity with that domain in other plant AGO1 members, e.g., OsAGO1b (97%), OsAGO1a (96%), and AtAGO1 (92%). The PIWI domain of TaAGO4 (Residue 569 – 875) shares 90%, 85%, and 77% homology with that of OsAGO4a, OsAGO4b, and AtAGO4, respectively. Generally, the PIWI domain displays a higher degree of similarity among plant Argonaute members, which supports a previous report of greater conservation of the PIWI domain but a poorly conserved PAZ domain [15].
When we aligned the PIWI domains of plant AGOs, including TaAGOs and paralogs in rice and Arabidopsis, we detected a trio of conserved metal-chelating amino acids -- aspartate, aspartate, and histidine (DDH) -- in both TaAGO1b and TaAGO4 (Figure 4). These particular DDH residues play critical roles during the process of RNA-directed cleavage of target RNAs, acting as a catalytic triad [16]. Thus, the inclusion of a DDH triad in both wheat AGOs suggests that they are functionally similar to previously characterized AGOs in RNA-directed gene silencing. Previous research has shown that the conserved histidine at position 798 (H798) of AtAGO1 is essential for in vitro endonuclease activity [17]; this residue was also detected in TaAGO1b. By comparison, in TaAGO4, as well as in OsAGO4a, OsAGO4b, AtAGO6, and AtAGO8, the histidine at the 798th position was replaced by proline (P), whereas in AtAGO4 and AtAGO9, that site was switched to serine (S) and arginine (R), respectively (Figure 4).
Structural modelling of the PIWI domain, with the Argonaute protein [PDB: 1u04] from Pyrococcus furiosus as template, indicated that TaAGOs can fold to a specific α/β structure that is dominated by a central mixed β-sheet and flanked by two long α-helices. Here, the three aligned DDH residues (Asp, Asp, and His) were spatially close to each other and located in the “slicer” site of the PIWI domain (Figure 5).
Genomic sequence analysis of TaAGO1b and TaAGO4
The genomic DNA sequence was analyzed by DNA amplification or genome-walking, using specific primers. TaAGO1b included 20 introns that varied in length from 72 to 449 bp. By contrast, 18 introns, 78 to 858 bp long, occurred in TaAGO4 (Figure 6). The largest intron (858 bp) in TaAGO4 was located in the region corresponding to the 5' UTR of mRNA.
Spatial expression patterns of TaAGOs
Expression analysis via Semi RT-PCR demonstrated that both TaAGO1b and TaAGO4 were highly expressed in all examined tissues, including the root, stem, leaf, anther, ovule, and mature seed (Figure 7-A), as well as in developing wheat kernels (Figure 7-B). During germination, TaAGO1b and TaAGO4 were ubiquitously expressed in embryonic tissues (Figure 7-C) but only differentially expressed in the endosperm (Figure 7-D). The transcript level of TaAGO4 in the endosperm tissues of germinating wheat seeds was greatly decreased.
Changes in TaAGOexpression in response to vernalization
To investigate the expression patterns of TaAGOs over time, we harvested wheat leaves at the 1- to 8-leaf stages of development and performed quantitative real-time RT-PCR. Expression of TaAGO4 was significantly up-regulated in the 2- and 3-leaf stages, while that of TaAGO1b was not obviously changed during our observation period (Figure 8-A). Vernalization treatment (exposing germinated seeds to 4°C for 30 d under darkness before planting) significantly affected their expression patterns, with TaAGO4 no longer being up-regulated at the 2- and 3-leaf stages but TaAGO1b being significantly up-regulated at the 6- and 7-leaf stages (Figure 8-B).
For further examination of the effects of vernalization treatment, we monitored the expression of TaAGOs during cold accumulation. Both were induced in the shoot tissues, with expression of TaAGO4 being greatly up-regulated at 18 to 30 d after cold treatment began but induction of TaAGO1b being relatively lower (Figure 8-C). However, no obvious induction of TaAGOs was detected in root tissues during the cold accumulation (Figure 8-D). These results indicated that both genes have important roles within the vernalization response, but their functioning might involve different regulatory mechanisms.
In silicomapping of TaAGOs
From a set of wheat aneuploids and deletion stocks, over 16000 ESTs have been mapped in the wheat chromosome / chromosome bins [18]. Those results are very useful for in silico mapping analysis. Through BLASTN searches, we identified mapped ESTs that are homologous to wheat AGO genes (Additional file 2). Based on sequence similarities and positions, we were able to map TAGO1 to the long arm of 7D and TaAGO4 to the short arm of 3A, 3B, and 3D on the wheat chromosomes.