JXB Advance Access originally published online on July 7, 2007
Journal of Experimental Botany 2007 58(11):2839-2849; doi:10.1093/jxb/erm094
RESEARCH PAPER |
Cloning of the PpNHAD1 transporter of Physcomitrella patens, a chloroplast transporter highly conserved in photosynthetic eukaryotic organisms
Departamento de Biotecnología, Universidad Politécnica de Madrid, 28040 Madrid, Spain
* To whom correspondence should be addressed. E-mail: begona.benito{at}upm.es
Received 25 October 2006; Revised 26 March 2007 Accepted 4 April 2007
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
A cDNA that encodes a transporter from the NHAD family was identified in Physcomitrella patens. Computer-based searches using the amino acid sequence of PpNHAD1 revealed that, in addition to being expressed in flowering plants, highly conserved transporters of this family are expressed in red algae, green algae, mosses, liverworts, and photosynthetic stramenopiles, but not in heterotrophic stramenopiles. A chloroplast transit peptide was detected in PpNHAD1 and in most of the related sequences, indicating that PpNHAD1 is a chloroplast transporter. A PpNHAD1-GFP fusion localized to the chloroplast in Physcomitrella protoplasts, and truncation of the N-terminus of the protein dispersed the fluorescence signal outside the chloroplast. PpNHAD1 did not show functional expression in either yeast or bacterial mutants, but truncated proteins with shorter N-termini, PpNHAD1-1 and PpNHAD1-2, could be functionally expressed in bacteria. PpNHAD1-1 alleviated the Li+ intolerance of a Na+-efflux Escherichia coli mutant at acidic pH values. Both PpNHAD1-1 and PpNHAD1-2 reduced the K+ requirements of a K+-influx E. coli mutant more actively at high pH values. PpNHAD1 seems to be an important transporter that mediates ionic homeostasis in chloroplasts from red algae to flowering plants.
Key words: Antiporter, chloroplast, lithium, NHAD, Physcomitrella, potassium, sodium
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The regulation of K+ and Na+ contents and the regulation of cellular pH are closely related processes that are essential for all living cells and for the organelles of eukaryotes. Under this regulation, cells maintain high K+ and low Na+ concentrations, contrary to the high Na+ and low K+ concentrations that prevail in many environments. The resulting concentration gradients across the membranes are created and maintained by K+, Na+, and H+ transporters that play crucial roles in the physiology of all living cells.
Many transporters mediate inward and outward fluxes of K+ and Na+ across the plasma and organelle membranes of plant cells (Véry and Sentenac, 2003; Pardo et al., 2006). Analysis of the genome sequences of Arabidopsis and rice has revealed that some of these transport proteins belong to families with many members, for example, HAK, NHX, or CHX, although for others, for example, SOS1, only one or at most two genes exist. Many of these plant protein families have sequence similarity to pumps, channels, or other types of transporters that have been described in bacteria, fungi, or animals, with which they form large families of transport proteins. The membranes in which plant transporters are expressed and the functions of one or several plant members of each family are currently being studied and compared with their homologues in other organisms. As a result, the basic transport functions of most alkali cation transport proteins are clear, although the details of their physiological functions, especially those of the transporters that are expressed on internal membranes, need more extensive research. By contrast, very little has been reported about the plant transporters with sequence similarity to bacterial NhaD transporters.
NhaD transporters have been identified in Vibrio parahaemolyticus (Nozaki et al., 1998), Vibrio cholerae (Dzioba et al., 2002; Habibian et al., 2005), and Alkalimonas amylolytica (Liu et al., 2005). These transporters mediate Na+ and Li+ effluxes and have been characterized as Na+(Li+)/H+ antiporters. Although this function is shared with other bacterial antiporters such as NhaA and NhaB (Padan et al., 2001), which also exist in V. cholerae (Herz et al., 2003) and V. parahaemolyticus (Kuroda et al., 2005), NhaD is peculiar because it is probably an electroneutral exchanger that mediates import of Na+ at alkaline pH (Dibrov, 2005). Furthermore, NhaD has sequence similarity to the ArsB carrier of Escherichia coli (Meng et al., 2004), and it has been found that V. cholerae cells devoid of NhaD show high resistance to millimolar concentrations of arsenate as compared to wild-type cells (Dibrov, 2005).
In plants, a cDNA encoding an NHAD transporter has been cloned from Populus euphratica, and homology-based computer searches indicate that these types of transporters exist in several plant species. Sequence analysis of the N-terminal amino acids failed to show targeting sequences for localization to internal membranes, which suggests that PeNHAD1 is located at the plasma membrane. Functional expression of the PeNHAD1 cDNA in an nhaA nhaB E. coli double mutant revealed that PeNHAD1 alleviates Na+ and Li+ toxicities at pH 5.5 (Ottow et al., 2005).
The moss Physcomitrella patens has become a suitable model with which most molecular techniques can be applied (Schaefer and Zrÿd, 2001; Schaefer, 2002; Frank et al., 2005). This model plant is being used for alkali cation transport studies (Benito and Rodríguez-Navarro, 2003) and the cloning of the Physcomitrella PpNHAD1 transporter and its localization to the chloroplast envelope are reported here. Computer-based searches in databases using the Physcomitrella sequence as a query revealed that chloroplast NHAD transporters probably exist in all plants, green and red algae, and photosynthetic stramenopiles. PpNHAD1 mediates K+ influx and Li+ efflux in E. coli.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Plant material and growth conditions
The moss P. patens was grown axenically in a medium described elsewhere (Ashton et al., 1979), supplemented with 500 mg l–1 ammonium tartrate, 5 g l–1 glucose (named PPNH4), and, when required, 7 g l–1 agar. Plants were grown in biofermenters or jars in a phytochamber with a discontinuous white light (16/8 h light/dark at 25/18 °C) at a quantum irradiance of 200 µmol m–2 s–1. Moss cultures were propagated by fragmenting old cultures in sterile water and transferring aliquots of this suspension to fresh medium.
Bacterial and yeast strains, growth conditions, and plasmids
E. coli strain DH5
was routinely used for plasmid DNA propagation. The E. coli strains EP432 (melBLid, nhaA1, nhaB1, lacZY, thr1), which is deficient in two systems of Na+ efflux, NhaA and NhaB (Pinner et al., 1993), and TKW4205 (thi rha lacZ nagA recA Sr::Tn10 kdpABC5 trkA405 Kup1), which is deficient in the three K+ uptake systems, Kdp1, TrkA, and Kup (Schleyer and Bakker, 1993), were used to test the functional expression of the cloned cDNAs. Functional tests in Saccharomyces cerevisiae were carried out in strains B3.1 (ena1::HIS3::ena4 nha1::LEU2), which is deficient in the ENA1-4 and NHA1 Na+ efflux systems (Bañuelos et al., 1998); AXT3K (ena1::HIS3::ena4 nha1::LEU2 nhx1::KanMX4), which is deficient in the ENA1-4 and NHA1 Na+ efflux systems and also in the NHX1 vacuolar Na+/H+ antiporter (Quintero et al., 2002); and W
6 (trk1::LEU2 trk2::HIS3), which lacks the TRK1 and TRK2 K+ uptake systems (Haro and Rodríguez-Navarro, 2003). Strain EP432 was routinely grown in a modified LB medium (LK) in which KCl substitutes for NaCl (1% tryptone, 0.5% yeast extract, 87 mM KCl); and strain TKW4205 was grown in LB medium (1% tryptone, 0.5% yeast extract, 87 mM NaCl) supplemented with 50 mM K+ (LBK). The pH of LB and LBK media is approximately 7.3; for experiments at pH 5.5, LB or LK media were supplemented with 10 mM MES and the pH adjusted with HCl. Bacterial growth at low K+ concentrations was tested in normal-strength or quarter-strength LB medium, or in a minimal medium (Senn et al., 2001). This minimal medium contained 3 µM K+ and 15 µM Na+, but the concentrations of these cations increased to 100 µM K+ and 9 mM Na+ after the addition of agar. The yeast strains B3.1 and AXT3K were maintained in YPD (2% peptone, 1% yeast extract, 2% glucose), and strain W6 in YPD supplemented with 50 mM KCl. Yeast growth tests were carried out either in arginine phosphate (AP) medium, whose basic formulation does not contain K+, Na+, or 
(Rodríguez-Navarro and Ramos, 1984), or in YPD, with both media supplemented with the required amounts of NaCl, LiCl, or KCl.
The cDNAs were cloned into plasmid pYPGE15 (Brunelli and Pall, 1993) for functional expression tests in yeast and into plasmid pBAD24 (Guzman et al., 1995) for bacterial tests. The yeast vector has a constitutive expression promoter, while in the bacterial vector, the expression is induced by arabinose; tests were carried out at several arabinose concentrations. The growth of bacterial cells in liquid media was monitored by recording absorbance at 600 nm in a Microbiology Reader Bioscreen C workstation (Growth Curves Oy, Finland), which enables several samples to be run at the same time. Microtitre plates containing 300 µl of bacterial suspension per well were incubated at 37 °C for 48 h with periodic shaking.
Recombinant DNA techniques
Manipulation of nucleic acids was performed by standard protocols or, when appropriate, according to the manufacturers’ instructions. PCR was performed in a Perkin-Elmer thermocycler using the Expand-High-Fidelity PCR System (Roche Molecular Biochemicals). DNA sequencing was performed in an automated ABI PRISM 3100 DNA sequencer (Applied Biosystems). PCR amplification of PpNHAD1 cDNA was carried out on double-stranded cDNA synthesized from total RNA by using the cDNA Synthesis System Kit (Roche) and the primers described below. Total Physcomitrella RNA and DNA were prepared using the RNeasy Plant Kit and DNeasy Plant Kit (Qiagen). Full-length cDNAs were obtained by using the 5'/3’ RACE Kit (Roche) according to the manufacturer's instructions. The PpNHAD1 cDNA was amplified from Physcomitrella mRNA by standard RT-PCR methods using a forward primer (PpNhaD-ATG, 5'- CTCTTGTGAAACTACCCATA-3') and a reverse primer (PpNhaD-STOP, 5'-CTACAAATTAAAGCCCTGGA-3') that amplified a fragment which included the ATG and stop triplets. The resulting PCR fragment was first cloned into the PCR2.1-Topo vector using the TOPO TA Cloning Kit (Invitrogen). For expression in yeast, the PCR2.1-Topo construct was digested with XbaI and KpnI and the fragment containing the cDNA was then ligated into the pYPGE15 vector, which had been previously digested with the same enzymes. For E. coli expression, the PCR2.1-Topo construct was digested with EcoRI and the fragment containing the cDNA was then ligated into the pBAD24 vector, which had been previously digested with the same enzyme. The PpNHAD1-1 and PpNHAD1-2 mutant cDNAs encode shorter versions of the PpNHAD1 transporter, with the corresponding proteins starting at Met103 and Met148 residues of the original PpNHAD1 protein, respectively, both excluding the N- terminal putative chloroplast transit peptide. These mutant cDNAs were constructed by PCR amplification using as a forward primer either PpNhaD-1-ATG (5'-GAGCTAGCGACCCGTTGTGTTCAG-3') or PpNhaD-2-ATG (5'-TGCTATGTCAGTGGTGTTCG-3'), and as a reverse primer the PpNhaD-STOP described above.
The PpNHAD1–GFP construct was an in-frame fusion of the 3’ end of the PpNHAD1 ORF to the GFP gene of the plasmid 35S-AdhI::GFP (Rubio-Somoza et al., 2006). To generate this construct, the PpNHAD1 full-length cDNA was amplified using the following primers, which each include a BamHI restriction site (underlined): BamHI-NhaD-ATG: 5'-TCTGGATCCCTTGACTCTTGCCAT-3' and BamHI-NhaD-Rev: 5'-TTGGATCCCTGGAAGAGCCGGTAG-3'. The PCR fragment was inserted into plasmid 35S-AdhI::GFP at the BamHI site, which is at the 5' end of the GFP gene. The shortest version of PpNHAD1, PpNHAD1-2, was also fused to the GFP gene. In this case, the PpNHAD1-2 cDNA was amplified using the forward primer BamHI-NhaD-2-ATG: 5'-CCGGATCCTTGCTATGTCAGTGGT-3' and the reverse primer BamHI-NhaD-Rev described above. This PCR fragment was also cloned into the BamHI site of plasmid 35S-AdhI::GFP.
The resulting versions of the PpNHAD1–GFP fusion constructs were used to transform Physcomitrella protoplasts following the polyethylene-glycol (PEG)-mediated procedure described in Hohe et al. (2004). For large-scale protoplast isolation, pH-controlled bioreactor cultivation in modified Knop medium was used with a reduced calcium concentration according to the protocol described by Hohe and Reski (2002). After transformation, the protoplasts were maintained in the dark for 24 h in PPNH4 medium supplemented with 3% mannitol and 5% glucose, followed by cultivation in the same medium for 4–6 d under normal growth conditions. The GFP fluorescence signal in Physcomitrella protoplasts was visualized using a confocal ultraspectral Leica microscope TCS-Sp2-AOBS-UV (Easley, SC, USA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Cloning and sequence analysis of PpNHAD1
In a BLAST search of translated Physcomitrella ESTs using as a query the amino acid sequence of PeNHAD1 (AJ561195), several ESTs were identified (BJ179942, BJ182411, BJ183588, BJ189378, BJ941375, BJ962081, and BJ969113) which could encode fragments of the N and C termini of a single NHAD transporter. By RT-PCR, using primers designed from 5' and 3' ESTs, a full-length cDNA was obtained from P. patens that unequivocally encodes an NHAD transporter. The PpNHAD1 cDNA had a length of 1746 bp and could encode a protein of 582 amino acid residues with a theoretical molecular mass of 61.98 kDa. A hydrophobicity profile of the deduced polypeptide (at http://www.ch.embnet.org/software/TMPRED_form.html) predicted 10 membrane-spanning regions.
The translated sequence of PpNHAD1 showed 76% identity with the amino acid sequence of the NHAD transporter of P. euphratica, and 72, 69, and 71% identities with the putative Arabidopsis and rice transporters AtNHD1 (At3g19490), AtNHAD2 (At1g49810), and OsNHAD1 (BAD17583 [GenBank] ), respectively (in all cases the comparison program ignores the N-terminal sequences that do not overlap, these sequences are discussed below).
Putative NHAD transporters were also found in many other photosynthetic organisms (Table 1). In comparison to other plant-type Na+-H+ antiporters, NHAD proteins form a compact cluster that is phylogenetically independent from NHX or SOS1 antiporters (Fig. 1).
|
|
Two truncated versions of the PpNHAD1 cDNA were constructed, PpNHAD1-1 and PpNHAD1-2 so that the N termini of the translated proteins corresponded to the second and third in-frame AUG codons of the cDNA (Fig. 2). This was necessary because the N terminus of plant NHAD transporters is controversial, as described below.
|
Subcellular localization of the PpNHAD1–GFP fusion protein
Analysis of the N-terminal amino acid sequence of PpNHAD1 predicted the presence of a chloroplast transit peptide (TargetP V1.1; (Emanuelsson et al., 2000)). A more precise analysis using the ChloroP 1.1 Server (Emanuelsson et al., 1999) again predicted the presence of a potential chloroplast transit peptide 59 amino acids long. This transit peptide is present in AtNHD1 but not in AtNHD2 and has not been reported to exist in PeNHAD1 (Ottow et al., 2005); in the proteins lacking the transit peptide the first Met residue aligned with those of bacterial NHAD proteins (Fig. 2).
To determine the subcellular localization of PpNHAD1, a translational fusion of PpNHAD1 and GFP was expressed under the control of the 35S promoter in Physcomitrella protoplasts. As shown in Fig. 3A–C, the green fluorescence signal of PpNHAD1-GFP was in patches at the chloroplast periphery whereas the red chlorophyll auto-fluorescence signal filled the chloroplast uniformly. This strongly suggests that PpNHAD1 mainly locates to the chloroplast envelope. Peripheral GFP signals, such as those found for PpNHAD1, have been described for the envelope phosphate transporter of Medicago truncatula (Zhao et al., 2003) and Arabidopsis (Versaw and Harrison, 2002), and the PAA1 copper ATPase of Arabidopsis (Abdel-Ghany et al., 2005). A uniform GFP signal has been reported for the thylakoid PAA2 copper ATPase (Abdel-Ghany et al., 2005).
|
To assess the location of the proteins with short N termini such as PeNHAD1 and AtNHD2, a truncated version of the PpNHAD1–GFP fusion, PpNHAD1-2–GFP was constructed. In this construct, the N terminus of the translated protein corresponded to the Met148 residue of the PpNHAD1 protein. This Met residue aligns with the N termini of the AtNHD2 and PeNHAD1 transporters (Fig. 2). The expression of this truncated fusion protein resulted in a dispersed intracellular green fluorescence signal that was external to the chloroplasts (Fig. 3D–F). The expression of GFP alone also produced a disperse signal with high accumulation in the nucleus (Fig. 3G–I). These data support the idea that the N terminus of PpNHAD1 determines the chloroplastic location of the transporter.
It has been predicted that the PeNHAD1 protein has an ER-signal peptide with a most likely cleavage site between amino acid positions 40 and 41, but not a chloroplast transit peptide or a mitochondrial targeting peptide. In accordance with these predictions, it has been proposed that PeNHAD1 is targeted to the secretory pathway and may localize to the plasma membrane with the function of transporting excess Na+ out of the cytoplasm into the apoplast (Ottow et al., 2005). On comparing the translated sequence of the putative 5’ untranslated region of PeNHAD1 with the sequence of the N terminus of PpNHAD1, it was found that a chloroplast transit peptide could be predicted in PeNHAD1, although a stop codon was present in the mRNA fragment encoding that potential transit peptide (Fig. 2). To investigate whether this stop codon was a sequencing mistake, the genome sequence of Populus trichocarpa, (http://genome.jgi-psf.org/Poptr1/Poptr1.home.html) was analysed, and two genes were identified that could encode proteins which were almost identical to PeNHAD1 (98% and 95% of amino acid sequence identity if the N-termini sequences were truncated up to the Met residue that corresponded to PpNHAD1-2 in Fig. 2, and 97% and 90% of identity when the whole sequences were compared). The conceptual translation of both genes produced proteins of lengths similar to that of PpNHAD1 and whose N termini included a putative chloroplast transit peptide (Fig. 2). These findings raised the possibility of there being a sequencing mistake in the PeNHAD1 cDNA sequence.
Chloroplast NHAD transporters may exist in all plants
A search in genomic and EST databases using the PpNHAD1 sequence as a query revealed that NHAD transporters may exist in all land plants, green and red algae, and photosynthetic stramenopiles. In Table 1, examples of identified sequences in flowering plants, the pteridophyte Ceratopteris, the liverwort Marchantia, the chlorophyte Chlamydomonas, the diatoms Phaeodactylum and Thalassiosira, and the red alga Porphyra yezoensis are listed. In most cases where the available sequence extended over the 5' terminus of the gene, a chloroplast transit peptide was detected. In the case of the diatoms Phaeodactylum and Thalassiosira, the NHAD genes were identified, but the chloroplast transit peptides were not. This result raised the possibility of a different localization of the NHAD transporter in these species. Because diatoms are photosynthetic stramenopiles, a search was carried out for NHAD genes in the genomes of the oomycetes Phytophthora soja, Phytophthora ramorum, and Phytophthora infestans, which are heterotrophic stramenopiles whose genome sequences are almost completed; in all cases the searches yielded negative results.
A remarkable characteristic of all NHAD transporters is their high degree of sequence conservation even in phylogenetically distant species. For example, the Arabidopsis and Oryza NHAD transporter sequences are 82% identical. By contrast, the Oryza sequence that is closest to the putative chloroplast Na+/H+ antiporter CHX23 of Arabidopsis (Song et al., 2004) is only 41% identical.
PpNHAD1-1 increases lithium tolerance in an E. coli Na+(Li+) efflux mutant
Bacterial NhaD transporters, which have shorter N termini than the plant transporters (Fig. 2), significantly increase the resistance to LiCl and NaCl of an E. coli nhaA nhaB mutant (Nozaki et al., 1998; Dzioba et al., 2002; Habibian et al., 2005; Liu et al., 2005), and a similar capacity at pH 5.5 has been found for PeNHAD1 (Ottow et al., 2005). To test whether the full-length PpNHAD1 cDNA or either of the two shorter versions, PpNHAD1-1 or PpNHAD1-2, has a function which is homologous to that of PeNHAD1, they were expressed from vector pBAD24, under the control of the arabinose-responsive promoter PBAD (Guzman et al., 1995) in the E. coli Na+(Li+)-efflux mutant EP432 (Pinner et al., 1993). Tests in solid media supplemented with Na+ or Li+ at pH values from 5.5 to 8.0 revealed that the three versions of PpNHAD1 failed to enhance the growth of the E. coli mutant in the absence of arabinose. By contrast, at 13 mM arabinose, pH 5.5, PpNHAD1-1 suppressed the inability of EP432 to grow in the presence of 50 mM LiCl (Fig. 4A), while PpNHAD1 and PpNHAD1-2 failed to show any effect on Li+ tolerance. In liquid media, pH 5.5, PpNHAD1-1 induced fairly rapid growth rates in EP432 up to 100 mM Li+ (Fig. 4B). As described for PeNHAD1 (Ottow et al., 2005), the ability of PpNHAD1-1 to suppress the mutant's defective growth on Li+ was pH-dependent. The increase of the pH up to 6.0 had an insignificant effect, at pH 7.0 the suppression was weak, and at pH 8.0 no effect was detected. Regarding the effect coincidences of PpNHAD1-1 and PeNHAD1 in strain EP432, it is worth noticing that the N-terminus of the PeNHAD1 transporter expressed in bacteria is shorter than the PpNHAD1-1 and PpNHAD1-2 proteins [41 amino acid residues shorter than PpNHAD1-2; compare Fig. 2 with the sequence after the SP fragment in Fig. 1 in Ottow et al. (2005)]. By contrast with the increase in Li+ tolerance, no significant effects were found of any of the tested cDNAs on the Na+ tolerance of the strain EP432.
|
The function of bacterial NhaD transporters has been tested in everted vesicles of the E. coli EP432 mutant or in another strain that carries an additional mutation in the ChaA gene (Nozaki et al., 1998; Dzioba et al., 2002; Habibian et al., 2005; Liu et al., 2005). Therefore, to investigate further the functional mechanism of PpNHAD1-1, everted membrane vesicles of the E. coli EP432 strain expressing PpNHAD1-1 were prepared by the procedures described in the above-mentioned reports. Although the vesicles were fully active in developing a pH gradient, the gradient was insensitive to the addition of Na+ or Li+ in either choline- or K+-loaded vesicles (results not shown).
Yeast heterologous expression has been used for functional characterization of transport proteins of the chloroplast envelope (Versaw and Harrison, 2002; Seigneurin-Berny et al., 2006). Therefore, the PpNHAD1 cDNA and its two shorter versions, PpNHAD1-1 and PpNHAD1-2, were inserted into a yeast expression vector and the resulting plasmid was transformed into the two S. cerevisiae strains, B3.1 (nha1 ena1-4) and AXT3K (nha1 nhx1 ena1-4), which are defective in Na+ efflux (Bañuelos et al., 1998; Quintero et al., 2002). In no case was it possible to detect an increase in the tolerance of these strains to Na+ or Li+.
PpNHAD1-1 and PpNHAD1-2 mediate K+ uptake in a K+-defective E. coli mutant
To investigate whether PpNHAD1 mediates K+ transport, the PpNHAD1, PpNHAD1-1, and PpNHAD1-2 constructs in pBAD24 were transformed into the E. coli TKW4205 strain, which is deficient in its intrinsic K+ uptake systems. At 13 mM arabinose, PpNHAD1-2 restored the growth of TKW4205 in solid minimal medium at 5 mM K+ (the agar added 9 mM Na+ to the basal medium, which was practically free of Na+). PpNHAD1-1 was effective at 10 mM K+, and PpNHAD1 did not show any effect (Fig. 5A). Similar results were obtained in liquid LB medium without added K+ (this medium had a pH of 7.3, contains 10 mM K+ due to the K+ content of the tryptone and yeast extract components, and 80 mM Na+) (Fig. 5B). Tests at decreasing pH values in this medium, down to 5.5, revealed that growth was scarcely dependent on the pH (data not shown). By contrast, in quarter-strength LB medium (2.5 mM K+, 20 mM Na+), both PpNHAD1-1 and PpNHAD1-2 were still able to suppress the growth defect of TKW4205, but in a pH-dependent manner that completely abolished growth at pH 5.5 (Fig. 5C).
|
Because these results indicated that PpNHAD1-1 and PpNHAD1-2 mediated K+ uptake in bacteria, Rb+ influx was tested, and it was found that both clones increased Rb+ influx with reference to the mutant strain, in the presence of 13 mM arabinose. This clearly occurred at all pH values tested, even at pH 5.5, and further demonstrated that PpNHAD1 is a K+ transporter. However, because the ability of NHAD antiporters to transport K+ has not been reported, the capacity of NHAD transporters to discriminate between K+ and Rb+ is unknown. Therefore, a kinetic study using Rb+ to predict the kinetic characteristics of the K+ influx mediated by PpNHAD1-1 or PpNHAD1-2 was complex and, unfortunately, the accuracy of our Rb+ uptake experiments was too low to support such a study.
The capacity of PpNHAD1, PpNHAD1-1, and PpNHAD1-2 to transport K+ was also tested in the K+ uptake yeast mutant W6 (Haro and Rodríguez-Navarro, 2003), but none of these clones improved the K+ dependence of the mutant.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In this study, a Physcomitrella patens cDNA was cloned, PpNHAD1, which encodes a transporter of the NHAD family. These transporters have been identified in halophilic and alkaliphilic eubacteria (Nozaki et al., 1998; Dzioba et al., 2002; Herz et al., 2003; Liu et al., 2005; Kurz et al., 2006), but they can also be identified in other bacteria by BLAST searches in databases. Remarkably, these transporters also seem to exist in flowering plants and an NHAD transporter from the tree Populus euphratica has been studied (Ottow et al., 2005).
Sequence analyses of the N-terminus of PpNHAD1 predicted a transit peptide of 59 amino acids. Consistent with this, the expression of a PpNHAD1–GFP fusion in Physcomitrella protoplasts resulted in a GFP signal that localized at the chloroplast periphery, whereas the red chlorophyll autofluorescence signal was uniform. As has already been described above, a comparison of these images with similar ones reported for chloroplastic proteins (Versaw et al., 2002; Zhao et al., 2003; Abdel-Ghany et al., 2005) strongly suggests that PpNHAD1 mainly locates to the chloroplast envelope. Deletion of the N-terminus up to the Met148 residue, which aligns with the N-terminus of the bacterial, PeNHAD1, and AtNHD2 transporters (Fig. 2), delocalized the GFP signal from the chloroplast (Fig. 3). The chloroplastic location of PpNHAD1 is in contrast with a previous study that suggested PeNHAD1 localizes to the plasma membrane (Ottow et al., 2005). However, that proposal was based exclusively on a predictive sequence study. We have now found two genes in the related species P. trichocarpa that encode NHAD transporters with chloroplast signal peptides. The most remarkable difference between the P. trichocarpa and P. euphratica cDNAs is the presence of an in-frame stop codon in the putative 5' UTR of the PeNHAD1 cDNA. These findings raise the possibility of there being a sequencing mistake in the PeNHAD1 cDNA and the prediction that PeNHAD1 may also localize to the chloroplast envelope.
A computer-based search for putative NHAD genes in genomic and EST databases revealed that these genes may be universally present in photosynthetic eukaryotic organisms because NHAD genes were identified in plants, green algae, red algae, and photosynthetic stramenopiles (Table 1). Although a chloroplast transit peptide can be difficult to predict in ESTs due to an absent or insufficient sequence at the 5' end, the transit peptide was almost constantly present in those cases in which it was properly studied (Table 1), suggesting that most NHAD transporters localize to the chloroplasts. Only in the cases of the diatoms Phaeodactylum and Thalassiosira, which are photosynthetic stramenopiles, could the existence of a chloroplast transit peptide not be predicted. However, this predictive failure does not suggest that the NHAD transporters in these organisms do not localize to the chloroplast for two reasons. First, in stramenopiles, among which there are photosynthetic and non-photosynthetic members, NHAD genes are restricted to the former. Second, a transit peptide does not always exist in chloroplast proteins. For example, the putative Na+/H+ exchanger AtCHX23 has been shown to localize to the chloroplast envelope (Song et al., 2004), but this protein lacks a predictable chloroplast transit peptide (results not shown).
Assuming the chloroplastic localization of the NHAD transporters, their presence in the red alga P. yezoensis is especially important because the divergence of red algae from vascular plants and bryophytes occurred more than 1500 Mya (Yoon et al., 2004). By contrast, a homologue of AtCHX23, whose function may be similar to that of NHAD, was not found in P. yezoensis (results not shown).
A complex problem, which cannot be currently solved because the cation transport mechanism is not yet clear, is the concrete function of the NHAD transporters in the chloroplast envelope. It has been proposed that bacterial NhaD transporters mediate an electroneutral Na+/H+ exchange that mediates export of Na+ at acidic pH values and import of Na+ at pH 8.0 (Dibrov, 2005). This mechanism could explain the Na+ and Li+ tolerance induced by PeNHAD1 (Ottow et al., 2005) and the Li+ tolerance induced by PpNHAD1-1 (Fig. 4), because in both cases there was a marked dependence on an external acidic pH. The failure of PpNHAD1-1 to relieve the toxicity of Na+ in strain EP432 while suppressing that of Li+ can be explained by a weak functional expression of PpNHAD1-1 in E. coli. Strain EP432 is defective for the NhaA and NhaB antiporters, which mediate Na+ and Li+ effluxes (Inaba et al., 1994; Seo, 1998), but conserves the ChaA antiporter (Ivey et al., 1993) and the electron transport-linked Na+ pump (Verkhovskaya et al., 1996), which specifically extrude Na+ but not Li+. Therefore, in the absence of the NhaA and NhaB Na+/H+ antiporters, the ChaA antiporter and the electron transport-linked Na+ pump can compensate, at least partially, for the Na+/H+ but not for the Li+/H+ antiport activity. Consequently, the defect in the latter activity can be more visibly suppressed if a transporter that has both activities is expressed in mutant EP432. This might be the case of PpNHAD1, which probably transports Na+ and Li+. Certainly, the NhaA transporters of E. coli and V. cholerae exhibit lower Kms for Li+ than for Na+ (Herz et al., 2003) and growth experiments suggest that AtNHX8 transports only Li+ (Rui et al., 2007). However, overall, a specific Li+ transporter is exceptional because most Na+/H+ antiporters (Häse et al., 2001, and references therein), and specifically NHAD transporters, accept both Na+ and Li+ as substrates (Nozaki et al., 1998; Dzioba et al., 2002; Habibian et al., 2005; Liu et al., 2005).
The failure of our experiments with vesicles, which contrasts with the simplicity of these experiments with bacterial NhaD transporters (Nozaki et al., 1998; Dzioba et al., 2002; Habibian et al., 2005; Liu et al., 2005), cannot currently be explained. Taking this into consideration, the aforementioned expression problems with PpNHAD1, and the fact that the suppression of the defect of strain EP432 by PeNHAD1 (Ottow et al., 2005) is very weak, it may be concluded that the bacterial expression of plant NHAD transporters has intrinsic difficulties.
It was also found that PpNHAD1-1 and PpNHAD1-2 very effectively suppressed the defect of the E. coli K+ uptake mutant TKW4205 and mediated Rb+ uptake. In quarter-strength LB medium (2.5 mM K+, 20 mM Na+) the suppression of the defect of the TKW4205 mutant was more effective at high (7.3) rather than at low (5.5) pH values. This is in contrast with the Li+ tolerance of the EP432 mutant strain transformed with PpHHAD1-1, which was restricted to low pH values. In other words, the two processes associated to the PpNHAD1 transporter are affected by the pH in an opposite way. It is worth observing that TKW4205 is normal for the Na+ efflux systems and EP432 for the K+ uptake ones.
In respect of the mechanism of PpNHAD1 that could explain the abovementioned results, the first consideration is that the same transporter is unlikely to mediate two independent processes, K+ uptake and Na+ efflux. The second consideration is that K+ uptake has to be electrogenic in the growth tests that were carried out at pH values of 7.3 or 6.5 (Fig. 5). This is necessary because the internal K+ concentration in E. coli is around 250 mM (Epstein and Schultz, 1965) and this involves an uphill transport against a 100-fold concentration gradient that took place in the absence of an outwardly directed H+ gradient. In the case of an electroneutral cation/H+ exchange, as proposed for NHAD transporters (Dibrov, 2005), only in tests at pH values which are much higher than 7.3, could the outwardly directed H+ gradient drive K+ uptake against the aforementioned transmembrane K+ gradient. Two electrogenic transport mechanisms fulfil the aforementioned requirements, 1Na+(Li+)/1K+-1H+ and 1Na+(Li+)/2K+ exchanges, i.e. one Na+(Li+) moves out coupled to the uptake of one K+ plus one H+ or to two K+. These two mechanisms are intrinsically different and can be distinguished by taking into account their pH dependences. In the Na+(Li+)/K+-H+ exchange two driving forces can operate, the H+ gradient and the membrane potential (the transporter moves one net positive charge inside). The joint effect of the two driving forces complicates the response of the transporter because, in bacteria and in particular in E. coli, a decrease in the external pH, which increases the inwardly directed H+ gradient, brings about a depolarization of the membrane potential (Bakker and Mangerich, 1981; Kashket, 1982). As a consequence, in an Na+(Li+)/K+-H+ antiporter, pH changes may enhance or inhibit the transporter depending on the actual conditions of the tests. Furthermore, the membrane potential is also dependent on the external K+ concentration (Kashket and Barker, 1977; Bakker and Mangerich, 1981). In our tests, when the concentration of external K+ is high, as is the case under the conditions for testing Li+ toxicity in EP432 (87 mM K+ in liquid medium), the bacterial cells are probably depolarized. In these conditions, the positive effect of a pH decrease can dominate the negative effect of the decrease in the membrane potential, which is already low. On the contrary, when K+ is at a low concentration, which is the case under the conditions for testing K+ requirements in TKW4205 (2.5 mM K+ in liquid medium), the cells are hyperpolarized with reference to cells at high K+ and the membrane potential can be highly sensitive to pH changes. Now, a decrease in the external pH can produce a negative effect on K+ uptake if the decrease of the membrane potential dominates the positive effect of increasing the concentration of H+. In conclusion, the aforementioned dual effect of pH changes can be explained if the mechanism is a Na+(Li+)/K+-H+ exchange. By contrast, in a 1Na+/2K+ mechanism, H+ is not a substrate for the transporter and the only effect of the decrease in the pH is a decrease in the membrane potential. This may affect the Na+/2K+ exchanger to a greater or lesser extent, depending on the testing conditions, but it always operates in the same direction, a pH decrease inhibits and a pH increase enhances the Na+/2K+ exchange.
The observed dual function of PpNHAD1-1 on the E. coli K+ or Na+ transport mutants and the abovementioned Na+(Li+)/K+-H+ mechanism have precedents in bacteria. The GerN protein of Bacillus cereus (Thackray et al., 2001) complements the Na+ sensitivity of an Na+ efflux mutant and the low-K+ sensitivity of a K+ uptake mutant of E. coli (Southworth et al., 2001). GerN can exchange 1Na+/1K+-1H+, as we propose for PpNHAD1-1, as well as 1Na+/2H+. By contrast, there are not any previous proposals for the Na+(Li+)/H+-K+ mechanism in bacterial NhaD transporters (Dibrov, 2005), which are supposed to be electroneutral Na+/H+ antiporters, as has also been proposed for PeNHAD1 (Ottow et al., 2005). It is worth noting that the negative effect of a pH increase on the growth of EP432 expressing PeNHAD1, which is the basis of the proposal of the electroneutral Na+/H+ exchange, can also be accounted for by the Na+(Li+)/K+-H+ mechanism, as has already been explained.
It seems unlikely that the transport mechanism of PpNHAD1 in chloroplasts is very different from that in bacteria. However, it may be difficult to establish the details of chloroplastic functions of NHAD1 transporters, which may be important for understanding the cation homeostatic processes in the chloroplast, only with the data obtained in bacteria. For example, the divergent effects of a pH decrease in the E. coli mutants expressing PpNHAD1, positive for Na+ tolerance but negative for K+ requirements, can be accounted for by the different conditions of the two experiments. This indicates that the chloroplast conditions, including ionic environment and membrane potential, must be determined before applying the findings in E. coli to chloroplasts. Therefore, the eventual elucidation of the physiological function of plant NHAD transporters might need more extensive studies with chloroplasts and the construction of nhad1 mutant plants. These lines of research deserve attention because chloroplasts need to control the pH (Peters and Berkowitz, 1991), a function that cannot be independent of the control of the K+ and Na+ contents. In addition to NHAD, the only transporter thus far identified that can mediate H+ and alkali cation fluxes is AtCHX23 (Song et al., 2004). By contrast with NHAD, CHX23 transporters are poorly conserved. For example, the identity between the Arabidopsis and Oryza NHAD sequences is 82% versus 41% for CHX23. This high conservation further supports the idea that NHAD transporters may have a critical function in the chloroplast.
|
Acknowledgements |
|---|
We would like to thank Marcel Velduizen for his technical assistance. Financial support for this work was provided by the Ministerio de Educación y Ciencia and a FEDER programme of the EU, through grant no. AGL2004-05153, which also funded a fellowship to JB and a Ramon y Cajal contract to BB. Additional financial support was obtained from grant no. R05/10719 of the DGUI-UPM Research Group Programme.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Abdel-Ghany SE, Müller-Moulé P, Niogi KK, Pilon M, Shikanai T. Two P-type ATPases are required for copper delivery in Arabidopsis thaliana chloroplasts. The Plant Cell (2005) 17:1233–1251.
Ashton NW, Cove DJ, Featherstone DR. The isolation and physiological analysis of mutants of the moss Physcomitrella patens. Planta (1979) 144:437–442.[CrossRef][ISI]
Bakker EP, Mangerich WE. Interconversion of components of the bacterial proton motive force by electrogenic potassium transport. Journal of Bacteriology (1981) 147:820–826.
Bañuelos MA, Synchrová H, Bleykasten-Grosshans C, Souciet J-L, Potier S. The Nha1 antiporter of Saccharomyces cerevisiae mediates sodium and potassium efflux. Microbiology (1998) 144:2749–2758.[Abstract]
Benito B, Rodríguez-Navarro A. Molecular cloning and characterization of a sodium-pump ATPase of the moss Physcomitrella patens. The Plant Journal (2003) 36:382–389.[CrossRef][ISI][Medline]
Brunelli JP, Pall ML. A series of yeast/Escherichia coli 
expression vectors designed for directional cloning of cDNAs and cre/lox-mediated plasmid excision. Yeast (1993) 9:1309–1318.[CrossRef][ISI][Medline]
Dibrov P. The sodium cycle in Vibrio cholerae: riddles in the dark. Biochemistry (Moscow) (2005) 70:150–153.[CrossRef][Medline]
Dzioba J, Ostroumov E, Winogrodzki A, Dibrov P. Cloning, functional expression in Escherichia coli and primary characterization of a new Na+/H+ antiporter, NhaD, of Vibrio cholerae. Molecular and Cellular Biology (2002) 229:119–124.
Emanuelsson O, Nielsen H, Brunak S, Heijne Gv. Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. Journal of Molecular Biology (2000) 300:1005–1016.[CrossRef][ISI][Medline]
Emanuelsson O, Nielsen H, Heijne Gv. ChloroP, a neural network-based method for predicting chloroplast transit peptides and their cleavage sites. Protein Science (1999) 8:978–984.[Abstract]
Epstein W, Schultz SG. Cation transport in Escherichia coli. V. Regulation of cation content. Journal of General Physiology (1965) 49:221–234.
Frank W, Decker EL, Reski R. Molecular tools to study Physcomitrella patens. Plant Biology (2005) 7:220–227.[CrossRef][Medline]
Guzman L-M, Belin D, Carson MJ, Beckwith J. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. Journal of Bacteriology (1995) 177:4121–4130.
Habibian R, Dzioba J, Barrett J, Galperin MY, Loewen PC, Dibrov P. Functional analysis of conserved polar residues in Vc-NhaD, Na+/H+ antiporter of Vibrio cholerae. Journal of Biological Chemistry (2005) 280:39637–39643.
Haro R, Rodríguez-Navarro A. Functional analysis of the M2D helix of the TRK1 potassium transporter of Saccharomyces cerevisiae. Biochimica et Biophysica Acta (2003) 1613:1–6.[Medline]
Herz K, Vimont S, Padan E, Berche P. Roles of NhaA, NhaB, and NhaD Na+/H+ antiporters in survival of Vibrio cholerae in a saline environment. Journal of Bacteriology (2003) 185:1236–1244.
Hohe A, Egener T, Lucht JM, Holtorf H, Reinhard C, Schween G, Reski R. An improved and highly standardized transformation procedure allows efficient production of single and multiple targeted gene-knockouts in a moss, Physcomitrella patens. Current Genetics (2004) 44:339–347.[CrossRef][ISI][Medline]
Hohe A, Reski R. Optimisation of a bioreactor culture of the moss Physcomitrella patens for mass production of protoplasts. Plant Science (2002) 163:69–74.
Häse CC, Fedorova ND, Galperin MY, Dibrov PA. Sodium ion cycle in bacterial pathogens: evidence from cross-genome comparisons. Microbiology and Molecular Biology Review (2001) 65:353–370.[CrossRef]
Inaba K, Kuroda T, Shimamoto T, Kayahara T, Tsuda M, Tsuchiya T. Lithium toxicity and Na+(Li+)/H+ antiporter in Escherichia coli. Biological and Pharmaceutical Bulletin (1994) 17:395–398.
Ivey DM, Guffanti AA, Zemsky J, Pinner E, Karpel R, Padan E, Schuldiner S, Krulwich TA. Cloning and characterization of a putative Ca2+/H+ antiporter gene from Escherichia coli upon functional complementation of Na+/H+ antiporter-deficient strains by the overexpressed gene. Journal of Biological Chemistry (1993) 268:11296–11303.
Kashket ER. Stoichiometry of the H+-ATPase of growing and resting aerobic Escherichia coli. Biochemistry (1982) 21:5534–5538.[CrossRef][Medline]
Kashket ER, Barker SL. Effects of potassium ions on the electrical and pH gradients across the membrane of Streptococcus lactis 7982. Journal of Bacteriology (1977) 130:1017–1023.
Kuroda T, Mizushima T, Tsuchiya T. Physiological roles of three Na+/H+ antiporters in the halophilic bacterium Vibrio parahaemolyticus. Microbiological Immunology (2005) 49:711–719.[ISI][Medline]
Kurz M, Brunig ANS, Galinski EA. NhaD type sodium/proton-antiporter of Halomonas elongata: a salt stress response mechanism in marine habitats? Saline Systems (2006) 2:10.[CrossRef][Medline]
Liu J, Xue Y, Wang Q, Wei Y, Swartz TH, Hicks DB, Ito M, Ma Y, Krulwich TA. The activity profile of the NhaD-type Na+(Li+)/H+ antiporter from the soda lake haloalkaliphile Alkalimonas amylolytica is adaptive for extreme environment. Journal of Bacteriology (2005) 187:7589–7595.
Meng Y-L, Liu Z, Rosen BP. As(III) and Sb(III) uptake by GlpF and efflux by ArsB in Escherichia coli. Journal of Biological Chemistry (2004) 279:18334–18341.
Nozaki K, Kuroda T, Mizushima T, Tsuchiya T. A new Na+/H+ antiporter, NhaD, of Vibrio parahaemolyticus. Biochimica et Biophysica Acta (1998) 1369:213–220.[Medline]
Ottow EC, Polle A, Brosché M, Kangasjärvi J, Dibrov P, Zörb C, Teichmann T. Molecular characterization of PeNhaD1: the first member of the NhaD Na+/H+ antiporter family of plant origin. Plant Molecular Biology (2005) 58:75–88.[CrossRef][ISI][Medline]
Padan E, Venturi M, Gerchman Y, Dover N. Na+/H+ antiporters. Biochimica et Biophysica Acta (2001) 1505:144–157.[Medline]
Pardo JM, Cubero B, Leidi EO, Quintero FJ. Alkali cation exchangers: roles in cellular homeostasis and stress tolerance. Journal of Experimental Botany (2006) 57:1181–1199.
Peters JS, Berkowitz GA. Studies on the system regulating proton movement across the chloroplast envelope. Effects of ATPase inhibitors , Mg2+, and an amine anesthetic on stromal pH and photosynthesis. Plant Physiology (1991) 95:1229–1236.
Pinner E, Kotler Y, Padan E, Schuldiner S. Physiological role of NhaB, a specific Na+/H+ antiporter in Escherichia coli. Journal of Biological Chemistry (1993) 268:1729–1734.
Quintero FJ, Ohta M, Shi H, Zhu J-K, Pardo JM. Reconstitution in yeast of the Arabidopsis SOS signaling pathway for Na+ homeostasis. Proceedings of the National Academy of Sciences, USA (2002) 99:9061–9066.
Rodríguez-Navarro A, Ramos J. Dual system for potassium transport in Saccharomyces cerevisiae. Journal of Bacteriology (1984) 159:940–945.
Rubio-Somoza I, Martinez M, Diaz I, Carbonero P. HvMCB1, a R1MYB transcription factor from barley with antagonistic regulatory functions during seed development and germination. The Plant Journal (2006) 45:17–30.[CrossRef][ISI][Medline]
Rui A, Qi-Jun C, Mao-Feng C, Ping-Li l Zhao S, Zhi-Xiang Q, Jia C, Xue-Chen W. AtNHX8 a member of the monovalent cation:proton antiporter-1 family in Arabidopsis thaliana, encodes a putative Li+/H+ antiporter. The Plant Journal (2007) 49:718–728.[CrossRef][ISI][Medline]
Schaefer DG. A new moss genetics: targeted mutagenesis in Physcomitrella patens. Annual Review of Plant Biology (2002) 53:477–501.[CrossRef][Medline]
Schaefer DG, Zrÿd J-P. The moss Physcomitrella patens, now and then. Plant Physiology (2001) 127:1430–1438.
Schleyer M, Bakker EP. Nucleotide sequence and 3'-end deletion studies indicate that the K+-uptake protein Kup from Escherichia coli is composed of a hydrophobic core linked to a large and partially essential hydrophilic C terminus. Journal of Bacteriology (1993) 175:6925–6931.
Seigneurin-Berny D, Gravot A, Auroy P, et al. HMA1, a new Cu-ATPase of the chloroplast envelope, is essential for growth under adverse light conditions. Journal of Biological Chemistry (2006) 281:2882–2892.
Senn ME, Rubio F, Bañuelos MA, Rodríguez-Navarro A. Comparative functional features of plant potassium HvHAK1 and HvHAK2 transporters. Journal of Biological Chemistry (2001) 276:44563–44569.
Seo SY. Characterization of the two Na+/H+ antiporters of Escherichia coli. Journal of Microbiology (1998) 36:9–13.[ISI]
Song C, Guo Y, Qiu Q, Lambert G, Galbraith DW, Jagendorf A, Zhu JK. A probable Na+(K+)/H+ exchanger on the chloroplast envelope functions in pH homeostasis and chloroplast development in Arabidopsis thaliana. Proceedings of the National Academy of Sciences, USA (2004) 101:10211–10216.
Southworth TW, Guffanti AA, Moir A, Krulwich TA. GerN, an endospore germination protein of Bacillus cereus, is an Na+/H+-K+ antiporter. Journal of Bacteriology (2001) 183:5896–5903.
Thackray PD, Behravan J, Southworth TW, Moir A. GerN, an antiporter homologue important in germination of Bacillus cereus. Journal of Bacteriology (2001) 183:476–482.
Verkhovskaya ML, Verkhovsky MI, Wikström M. The respiration-driven active sodium transport system in E. coli does not function with lithium. FEBS Letters (1996) 388:217–218.[CrossRef][ISI][Medline]
Versaw WK, Chiou T, Harrison MJ. Phosphate transporters of Medicago truncatula and arbuscular mycorrhizal fungi. Plant and Soil (2002) 244:239–245.[CrossRef][ISI]
Versaw WK, Harrison MJ. A chloroplast phosphate transporter, PHT2;1, influences allocation of phosphate within the plant and phosphate-starvation responses. The Plant Cell (2002) 14:1751–1766.
Véry A-A, Sentenac H. Molecular mechanisms and regulation of K+ transport in higher plants. Annual Review of Plant Biology (2003) 54:575–603.[CrossRef][Medline]
Yoon HS, Hackett JD, Ciniglia C, Pinto G, Bhattacharya D. A molecular timeline for the origin of photosynthetic eukaryotes. Molecular Biology Evolution (2004) 21:809–818.[CrossRef]
Zhao L, Versaw WK, Liu J, Harrison MJ. A phosphate transporter from Medicago truncatula is expressed in the photosynthetic tissues of the plant and located in the chloroplast envelope. New Phytologist (2003) 157:291–302.[CrossRef][ISI]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||