Based on the conserved sequence of the known plant PCS proteins, four full-length cDNA clones encoding PCS in S. rostrata, designated SrPCS1 (GenBank accession no. DQ010916), SrPCS2 (GenBank accession no. GQ204308), SrPCS3 (GenBank accession no. GQ204309) and SrPCS4 (GenBank accession no. GQ204310), were isolated by RT-PCR amplification of conserved cDNA sequences combined with 3′ and 5′-rapid amplification of cDNA ends (RACE). The calculated molecular masses from the deduced amino acid sequences of SrPCS1, SrPCS2, SrPCS3 and SrPCS4 are 26 kD, 19.8 kD, 55.8 kD, and 19.9 kD, respectively (Figure 1).

Among the four SrPCS cDNAs, the ORF of SrPCS3 is the longest one; its amino acid sequence shares 65% identity to Arabidopsis AtPCS1 [16], 87% identity to soybean GmhPCS1 [7] and 87% identity to L. japonicus LjPCS1 [18]. The identity of SrPCS3 with SrPCS1, SrPCS2 and SrPCS4 is 97.9%, 87.6% and 88.1%, respectively. The N-terminal domains of SrPCS3 and SrPC1, but not SrPCS2 and SrPCS4, contain a catalytic triad of Cys-56, His-162 and Asp-180 (Figure 1), which is essential for the activity [9,10,13,14].

An unrooted phylogenetic tree (Figure 2) was constructed including the majority of the predicted PCS sequences available in the databases. The tree revealed separate clusters, paralleling taxonomic distance, for the PCSs of cyanobacteria, nematodes, yeast, ferns, and the families of vascular plants Poaceae, Allaceae-Typhaceae, Solanaceae, Brassicaceae, and Leguminosae. As shown in Figure 2, SrPCS proteins (in red dashed ellipse of Figure 2) are highly homologous to leguminous soybean GmPCS [7] and Lotus LjPCS1 [18]. The structure and function of PCS genes have only been studied in detail in the model plant Arabidopsis. Two functional PCS genes, AtPCS1 and AtPCS2, which are localized in different chromosomes (At5g44070 and Atlg03980), were found in Arabidopsis [6,10,17,24]. In order to study the structure of S. rostrata PCS gene (SrPCS), we cloned the genomic sequence of SrPCS from the genomic DNA of S. rostrata by genome walking [28] with primers based on the conserved sequence of SrPCS1-4 cDNAs. Unexpectedly, only a 5,112-bp genomic DNA fragment of SrPCS (GenBank accession no. GQ204307) was obtained.

Alignment of the genomic sequence of SrPCS with those of four cDNA sequences of SrPCS1-4 revealed that all four mRNAs (SrPCS1-SrPCS4) were produced from the one SrPCS gene through alternative splicing (Figure 3).

By comparing with the sequence of the canonically spliced form SrPCS3 mRNA, eight exons and seven introns were found in SrPCS gene. As shown in Figure 3, SrPCS1 and SrPCS2 are generated through alternative AG acceptor splice sites. And the SrPCS4 is produced by atypical splice mechanism employing direct repeats at splice sites (CTCC, Figure 3), which is similar to the atypical mRNA processing of p53 transcripts in the human brain [29].

The presence of a single copy of the SrPCS gene in S. rostrata genome was further confirmed by Southern blot analysis (Figure 4). Only a single hybridizing band was detected in DNA samples digested with either EcoRI, BamHI or XbaI, which do not cut within the probe region (Figure 4, lane 1–3). When the genomic DNA was digested with PstI that cut once within the probe region, two hybridizing bands were observed (Figure 4, lane 4). These results indicate that the S. rostrata genome contains only a single copy of the SrPCS gene.

Alternative splicing of pre-mRNA is often regulated by specific factors in response to developmental or environmental cues [30–32]. To find out whether the alternative splicing of SrPCS pre-mRNA is regulated by heavy metals, we investigated SrPCS expression in plants in the presence and absence of Cd²⁺. As in wheat [17] and Brassica juncea [33] plants, SrPCS1–SrPCS4 mRNA expression was very low and could not be detected by RNA blot analysis of total RNA from different organs of S. rostrata. Therefore a more sensitive real time-PCR method was used to estimate the abundance of SrPCS transcripts. Unfortunately, because of very high sequence similarity of the four SrPCS mRNA, only specific primers for SrPCS3 can be designed for real time-PCR analysis.

Unlike in Arabidopsis and Brassica juncea, where higher amounts of Arabidopsis AtPCS1 transcript [6] and B. juncea BjPCS1 protein [33] were detected in roots than in leaves, SrPCS3 is more highly expressed in the leaves than roots and stems under normal growth conditions without Cd²⁺ treatment (Figure 5). SrPCS3 expression was significantly reduced in roots after exposure to different concentration of Cd²⁺ for 24 hours. No significant change of SrPCS3 expression was observed in stems, whereas SrPCS3 expression was also significantly down-regulated in leaves by 0.3 mM Cd²⁺ treatment. These observations are contrary to TaPCS1 from wheat [17] and LjPCSs from the model legume Lotus japonicus [18]. Clemens et al. [17] found that TaPCS1 showed a 5- to 10-fold higher expression in wheat roots treated with 0.1 mM Cd²⁺ for 6 h. Ramos et al. [18] investigated the expression of three PCS genes in roots of L. japonicus, and found that the mRNA expression of LjPCS1-8R increased only slightly (2.1-fold) after 6 h of 0.1 mM Cd²⁺ treatment, whereas the mRNA levels of LjPC2-7R and LjPCS3-7N increased 2.3- to 2.9-fold and 2.5- to 3.5-fold between 6 and 96 h, respectively.

To test the functionality of four SrPCS isoforms, the full-length cDNAs of SrPCS1–SrPCS4 were expressed in S. cerevisiae under the control of the inducible pGAL1 promoter. Cd tolerance assay showed that S. cerevisiae cells expressing SrPCS1 or SrPCS3, but not SrPCS2 and SrPCS4, exhibited strong tolerance to Cd²⁺ (600 μM), which completely inhibited the growth of cells transformed with the empty pYES2 plasmid or cells expressing SrPCS2 or SrPCS4 (Figure 6). Growth curve experiments (Figure 7, a–d) also demonstrated that expression of SrPCS1 and SrPCS3 conferred enhanced Cd tolerance to yeast cells (Figure 7, a and c), whereas the expression of SrPCS2 and SrPCS4 did not make yeast cells more Cd tolerant (Figure 7, b and d).

Dose-response analyses showed that SrPCS3-expressing yeast cells tolerated higher Cd²⁺ concentrations than the SrPCS1 expressing cells and control cells (Figure 7, e), which may result from the lack of the C-domain that has been shown to ensure higher PCS activity [11]. Alternative splicing, unlike gene promoter activity that primarily regulates the amount of transcripts, changes the structure of transcripts and their encoding proteins [34]. Therefore alternative splicing is one of the most important mechanisms to generate a large number of mRNAs and protein isoforms from a single or few genes [35]. Several studies have shown that about 20% of plant genes have one or more alternative transcript [36–39]. And based on the alignments of ESTs from other legume species against the Medicago truncatula genome sequence, hundreds of legume genes were recently predicted to be alternatively spliced [39].

In this study, we found that a single copy of the SrPCS gene occurs in S. rostrata genome, and that alternative splicing of the SrPCS pre-mRNA produced four different SrPCS mRNAs and proteins, SrPCS1–SrPCS4 (Figure 1). The SrPCS1 and SrPCS3 proteins, whose N-terminal domains contain a catalytic triad of Cys-56, His-162 and Asp-180 (Figure 1), conferred Cd tolerance when expressed in yeast cells (Figure 6 and 7), whereas the SrPCS2 and SrPCS4 proteins, which contain only a Cys-56 of the catalytic triad and lack the N-terminal domains (Figure 1), did not render yeast cells more Cd tolerant (Figure 6 and 7). This observation is similar to the findings of Ramos et al. [18], who showed that a PCS gene (LjPCS2) in model legume Lotus japonicus can be alternatively spliced to encode two protein isoforms, one (LjPCS2-7N) conferring enhanced Cd tolerance to yeast cells while the other (LjPCS2-7R) did not. SrPCS1 and SrPCS3 characterized in this study have potential applications in genetic engineering of plants for enhancing heavy metal tolerance and phytoremediation of contaminated soils [40]. Although we detected the mRNA expression of the four SrPCS isoforms in the current study, the corresponding proteins of these mRNAs and their physiological functions remain to be characterized. In particular, the presence of SrPCS2 and SrPCS4 proteins, which lack the catalytic triad and the N-terminal domains and do not confer enhanced Cd tolerance to yeast cells, needs to be determined in S. rostrata plants, since they are small variants of the full-length protein (about one-third of the SrPCS3) and may not be stable if translated in plant cells.

Seed testae of Sesbania rostrata were scalpelled and surface-sterilized in 70% alcohol for 1 minute, followed by three rinses with sterile distilled water, and then soaked in 10% sodium hypochlorite for 15 minutes with gentle agitation. After washing three times with sterile distilled water, the seeds were germinated on 1/2 MS medium [41] for 6 days. These seedlings were then transferred to 1/8 liquid MS without sugar to grow in this medium for 15 days, changing medium every three days, and then transferred to 1/4 liquid MS containing different concentration of cadmium for 1–2 days as needed at 24 ± 2 °C under fluorescent light with a 16-h photoperiod.

Total RNA was extracted from seedlings treated with Cd for 2 days as described above using the Trizol reagent (Invirtogen). The cDNA was synthesized from 2 μg of total RNA that was treated by DNaseI before reverse transcription (RT) with the M-MLV (Progema). RT-PCR amplification of the conserved partial sequence of SrPCS cDNAs was performed with a pair of primers (forward primer ZF168, 5′-GCGTTAAGAGACAATGGCGATG-3′; reverse primer ZF169, 5′-GGGAGATGCTAAG-AAAAGGGTACA-3′) based on the sequence of soybean PCS [7]. The amplification parameters were as follows: 5 min at 95 °C, 30 cycles of 40 s at 95 °C, 40 s at 54 °C and 2 min at 72 °C, 10 min at 72 °C in a 20-μl reaction mixtures. The 5′- and 3′- ends of SrPCS cDNAs were isolated with a modified RACE procedure based on template-switching effect and inverse PCR [42]. Genomic DNA from S. rostrata was extracted by a modified CTAB method [43]. SrPCS genomic clone was obtained using the Universal GenomeWalker Kit (Clontech) according to the manufacturer's protocol. PCR products were cloned into pGEM-T easy (Progema) and sequenced. Sequence identity was calculated by the Ident and Sim program in the Sequence Manipulation Suite (http://www.bioinformatics.org/sms/).

Real time RT-PCR was used for the quantification of the SrPCS3 expression using the Premix Ex Taq^™ (Perfect Real Time, TakaRa) with a pair of SrPCS3 specific primers (forward primer ZF332, 5′-GCTTTATACTCTGAGCTG-3′; reverse primer ZF335, 5′-TTAGATGGCAG-TGATGT-3′). As an internal control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was simultaneously amplified to normalize SrPCS3 expression with a pair of GAPDH specific primers (forward primer ZF371, 5′-GCTGGTGCTGATTATGTT-3′; reverse primer, ZF372 5′-GCTCTGGCTTGTATTCCT-3′). PCR cycling conditions were: 15 s at 95 °C; 40 cycles of 5 s at 96 °C, 20 s at 58 °C and 30 s at 72 °C in a 25-μl reaction mix containing 1x SYBR Green I. Each sample was performed in triplicates at least three times. Amplification products were measured and analyzed with an ABI Prism 7900HT sequence detection system and SDS software (Applied Biosystems).

For Southern blot analysis, 25 μg genomic DNA was digested with restriction enzymes and separated on 0.8% agarose gel and transferred to Hybond-N membrane (Amersham) according to Sambrook and Russell [44]. The conserved SrPCS cDNA fragment (nt 47-919 of GenBank GQ204309) was labeled with [α-³²p] dCTP using Random Primer Kit (TakaRa) and used to detect DNA on the membrane as previously described [45].

Four ORFs of SrPCS1-4 cDNAs were amplified by PCR with a pair of SrPCS-specific primers carrying a BamHI and a KpnI restriction enzyme site (forward primer ZF268, 5′-GGTACCATGGCG-ATGGCGGGGTTGTA-3′; reverse primer ZF269, 5′-ATGGGATCCGGAGATGCTAAGAAAAGG-3′), respectively. After digestion with BamHI and KpnI, each of four PCR fragments containing the ORF of SrPCS1-4 was ligated in the yeast expression vector pYES2 (Invitrogen), resulting in four recombinant vectors (pAM26, pAM27, pAM28, and pAM29) that were used to transform Saccharomyces cerevisiae strain INVSc1 (MATK his3v1 leu2 trp1-289 ura3-62) cells according to the supplier’s protocol (Invitrogen). For the Cd tolerance assay, yeast cells containing the empty vector pYES2 or the SrPCS1-4 cDNA constructs were grown in the yeast nitrogen base (YNB) medium supplemented with the appropriate amino acids, 1% galactose and 1% raffinose containing different concentrations of Cd²⁺ at 30 °C.