SEARCH

SEARCH BY CITATION

Keywords:

  • Lycopersicon esculentum;
  • Solanum tuberosum;
  • MicroTom;
  • mycorrhiza;
  • phosphate transport

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Solanaceous species are among the >200 000 plant species worldwide forming a mycorrhiza, that is, a root living in symbiosis with soil-borne arbuscular-mycorrhizal (AM) fungi. An important parameter of this symbiosis, which is vital for ecosystem productivity, agriculture, and horticulture, is the transfer of phosphate (Pi) from the AM fungus to the plant, facilitated by plasma membrane-spanning Pi transporter proteins. The first mycorrhiza-specific plant Pi transporter to be identified, was StPT3 from potato [Nature414 (2004) 462]. Here, we describe novel Pi transporters from the solanaceous species tomato, LePT4, and its orthologue StPT4 from potato, both being members of the Pht1 family of plant Pi transporters. Phylogenetic tree analysis demonstrates clustering of both LePT4 and StPT4 with the mycorrhiza-specific Pi transporter from Medicago truncatula [Plant Cell, 14 (2002) 2413] and rice [Proc. Natl Acad. Sci. USA, 99 (2002) 13324], respectively, but not with StPT3, indicating that two non-orthologous mycorrhiza-responsive genes encoding Pi transporters are co-expressed in the Solanaceae. The cloned promoter regions from both genes, LePT4 and StPT4, exhibit a high degree of sequence identity and were shown to direct expression exclusively in colonized cells when fused to the GUS reporter gene, in accordance with the abundance of LePT4 and StPT4 transcripts in mycorrhized roots. Furthermore, extensive sequencing of StPT4-like clones and subsequent expression analysis in potato and tomato revealed the presence of a close paralogue of StPT4 and LePT4, named StPT5 and LePT5, respectively, representing a third Pi transport system in solanaceous species which is upregulated upon AM fungal colonization of roots. Knock out of LePT4 in the tomato cv. MicroTom indicated considerable redundancy between LePT4 and other Pi transporters in tomato.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

The macronutrient phosphorus (P) is one of the most immobile and least available nutrients present in soils (Raghothama, 1999; Schachtman et al., 1998). P is taken up by plants as orthophosphate (H2POinline image, Pi) by secondary transport of Pi anions across membrane integral phosphate transporter proteins, driven by the plasma membrane proton pump at the expense of ATP (Daram et al., 1998; Leggewie et al., 1997). ATP hydrolysis enables root cells to take up Pi against a steep concentration gradient of available Pi in the soil solution rarely exceeding 1–10 μm compared to intracellular Pi concentrations in the millimolar range (Marschner, 1995). Plants have developed a number of strategies to support growth under low P conditions, including the formation and elongation of root hairs, and the development of a mycorrhiza, both resulting in an extension of the nutrient absorptive surface area. During Pi deprivation, additional root hairs develop and hairs elongate, thus increasing the plant's Pi acquisition efficiency (Ma et al., 2001; Narang et al., 2000) by enhancing Pi uptake along the so-called ‘direct pathway’ from soil to root at the level of the root epidermis (Smith et al., 2003). Strongly reduced mobility of Pi in the soil and rapid direct Pi uptake into the root lead to the development of a Pi depletion zone around the root hair cylinder and a rapid decline of Pi acquisition over time. While in non-mycorrhizal plants, the extension of the Pi depletion zone is closely related to root hair length (Marschner and Dell, 1994), in mycorrhized plants the depletion zone of Pi by far exceeds the root hair cylinder (Jungk and Claassen, 1989), indicating that Pi not directly available to the plant is being delivered by the fungal hyphae. Thus, the presence of the Pi depletion zone is a major factor contributing to the advantage of forming a mycorrhiza for plants. The arbuscular mycorrhizal (AM) symbiosis occurs in roughly 80% of vascular plants (>200 000 species). In this symbiosis, AM fungi (AMF) extract Pi from a soil volume encompassing the entire ‘mycorrhizosphere’ via their extraradical hyphae in soil areas distant from the root surface. Pi is then transported via the ‘mycorrhizal pathway’ along fungal hyphae to the specialized symbiotic interfaces within the root cortex in exchange for plant-based carbohydrate (Smith et al., 2003). Members of the Pht1 family of plant Pi transporters (Mudge et al., 2002; Rausch and Bucher, 2002) are expressed at both the root–soil interface including root hairs (Chiou et al., 2001; Daram et al., 1998; Gordon-Weeks et al., 2003; Liu et al., 1998a; Muchhal and Raghothama, 1999; Mudge et al., 2002; Zimmermann et al., 2003) and the symbiotic fungus–plant interface around arbuscules or hyphal coils (Harrison et al., 2002; Karandashov et al., 2004; Paszkowski et al., 2002; Rausch et al., 2001). This expression pattern is consistent with a function of Pht1 transporters in the capture of Pi and its uptake along both the direct and the mycorrhizal pathway, respectively. Moreover, AM fungal proteins similar to those of the Pht1 family are expressed in extraradical hyphae (Harrison and Van Buuren, 1995; Maldonado-Mendoza et al., 2001). In the two solanaceous species tomato and potato, the respective orthologous pairs of Pi transporters LePT1 and StPT1, and LePT2 and StPT2, are predominantly expressed in the rhizodermis including root hairs, the former rather constitutively, the latter inducible by Pi deprivation (Daram et al., 1998; Gordon-Weeks et al., 2003; Liu et al., 1998a; Zimmermann et al., 2003). A recently identified third potato Pi transporter gene, StPT3, which is expressed in cortex cells of roots colonized by AMF, is thought to mediate symbiotic Pi uptake at the fungus–root interface in the mycorrhiza (Rausch et al., 2001). The mycorrhiza-responsive StPT3 promoter directs expression similarly in transgenic roots of distantly related plant species including potato, petunia, carrot, Medicago truncatula, and Lotus japonicus, indicating a high degree of conservation of the signal recognition and transduction pathways in AM symbiotic Pi transport (Karandashov et al., 2004). The identification of mycorrhiza-specific rice OsPT11 (Paszkowski et al., 2002) and M. truncatula MtPT4 (Harrison et al., 2002) Pi transporters, both being non-orthologous to StPT3, recently indicated the presence of a second mycorrhiza-specific Pi uptake system in vascular plants.

Here, we report on the characterization of two novel orthologous mycorrhiza-inducible Pi transporters from tomato and potato, designated LePT4 and StPT4, respectively. This finding indicates the presence of two Pi transport systems, exhibiting functional redundance and probably differing in affinity for Pi, in each of these plant species.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Cloning of LePT4 and StPT4, and computational sequence analysis

A mutant population of Ac/Ds transposon-trap insertion lines has recently been generated for the tomato cultivar MicroTom (Meissner et al., 2000). Based on analysis of transposon flanking sequences, an Ac insertion into a genomic region exhibiting substantial homology to LePT1 (Y14214, AF022873), LePT2 (AF022874), and other plant Pi transporters was identified. The gene was named Lycopersicon esculentum Pi transporter 4 (LePT4; AY885651). Eventually, a genomic fragment of 1990 bp was cloned by polymerase chain reaction (PCR) encompassing an intron-less open reading frame of 1634 bp encoding a predicted polypeptide of 545 amino acids with a mass of 60.5 kDa and a theoretical pI value of 8.49. In parallel, a cDNA of same length, sharing 100% identity with the corresponding genomic sequence, was cloned by RT-PCR. Finally, we identified by inverse PCR (iPCR) a genomic fragment starting 794 bp upstream and extending 717 bp downstream of the putative initiator ATG, thus including almost 0.8 kbp of LePT4 promoter sequence. The genomic sequence downstream of the ATG was identical to the corresponding cDNA sequence (data not shown). An unrooted phylogenetic tree demonstrates the close relationship between the encoded LePT4 protein and members of four subfamilies (I–IV) of the plant Pht1 family, and high-affinity Pi transporters of fungal origin (Figure 1), whereas the Pi transporter ARAth;Pht2;1 (Daram et al., 1999) was non-homologous to Pht1 proteins. LePT4 clustered with the mycorrhiza-specific MtPT4 from M. truncatula (AY116210) and OsPT11 from rice (AF536971) forming subfamily I, but not with StPT3 from potato (AJ318822). In contrast to LePT4 and StPT3, the orthologous proteins LePT1/StPT1 and LePT2/StPT2, respectively, are closely related to each other. Sequence divergence between LePT4 and StPT3 suggested the existence of orthologous StPT4 and LePT3, respectively. In order to identify the potato StPT4 gene, a sequence comparison using LePT4-, MtPT4-, and OsPT11-coding sequences was performed and highly similar regions differing from the corresponding regions in StPT1, StPT2, and StPT3, respectively, were used to design primers for PCR amplification of the respective 1.164 kbp genomic region. iPCR and 3′RACE subsequently resulted in the cloning of a 2.71 kbp genomic fragment composed of 1.12 kbp 5′ untranslated region followed by a 1.587 kbp intron-less region encoding the StPT4 polypeptide of 529 amino acids (acc. no. AY793559) with a predicted mass of 58.72 kDa, and a theoretical pI value of 8.48. The StPT4 amino acid sequence shared 97.2% identity with the LePT4 polypeptide and hence clustered with subfamily I of mycorrhiza-inducible genes (Figure 1), demonstrating functional conservation.

image

Figure 1. Unrooted phylogenetic tree of plant and fungal Pi transporters. Transporters shown to be strongly upregulated upon mycorrhization are highlighted in gray. Transporters and corresponding plant species are: rice, ORYsa;Pht1;1 to 13 (Paszkowski et al., 2002); Arabidopsis thaliana, ARAth;Pht1;1 to 9, and Pht2;1 (Mudge et al., 2002; Poirier and Bucher, 2002); tobacco, NtPT1 to 4 (Kai et al., 2002); tomato, LePT1 and 2 (Daram et al., 1998; Liu et al., 1998a), LePT4 (this work) and partial LePT5 (this work); potato, StPT1 to 3, and SOLtu;Pht2;1 (Karandashov et al., 2004; Leggewie et al., 1997; Rausch et al., 2001), StPT4 (this work) and partial StPT5 (pStPT5, this work); barrel medic, Medicago truncatula, MtPT1, 2 and 4 (Harrison et al., 2002; Liu et al., 1998b); yeast, PHO84 (Bun-Ya et al., 1991); Glomus versiforme, GvPT (Harrison and Van Buuren, 1995); Neurospora crassa, PHO-5 (Versaw, 1995). Roman numerals indicate four different Pht1 subfamilies thought to differ in evolutionary age (see Discussion).

Download figure to PowerPoint

The other Pht1 family members, LePT4 and StPT4, both contain 12 putative transmembrane domains (TMs) as was predicted by the TMPred program (Hofmann and Stoffel, 1993), hydrophilic N and C termini, and a hydrophilic loop between TM6 and TM7 (data not shown). Scanning and alignment of the protein sequences using ScanProsite (Gattiker et al., 2002), indicated that both transporters share two putative glycosylation sites after the predicted TM7 and within TM10, and three phosphorylation sites, respectively, with all Pht1 family members identified from both tomato and potato (data not shown). Additionally, TM10 was indicated to share high similarity with a WW/rsp5/WWP domain signature (amino acids 418–443 or 402–427 for LePT4 or StPT4, respectively) which has been shown by Chen and Sudol (1995) to bind proteins with particular proline-motifs, [AP]-P-P-[AP]-Y, and thus resembles SH3 domains. The WW domain in LePT4/StPT4 may therefore be involved in protein–protein interactions.

Phosphate transporter gene expression analysis

LePT4 and StPT4 promoters share an overall sequence identity of 60% with almost identical regions of up to 330 bp in length separated by gaps or rather short regions of low similarity in the alignment (Figure 2a and data not shown). Sequence comparisons with the StPT3, OsPT11, and MtPT4 promoter, respectively, did not reveal structural commonalities (data not shown). Especially, unlike in the StPT3 promoter, neither a miniature inverted repeat transposable element (MITE) nor mycorrhiza- and resistance-related (MRR) elements (Rausch et al., 2001) were identified within the two promoter regions. However, putative cis-acting elements, identified recently in a 129 bp promoter region of the StPT3 gene by phylogenetic footprinting (Karandashov et al., 2004), revealed a similar distribution of the same elements in both the StPT4 and LePT4 promoters without the 129 bp region being conserved (data not shown).

image

Figure 2. LePT4 and StPT4 promoter analysis. (a) Schematic view of sequence comparison between the LePT4 and the StPT4 promoter, respectively. Corresponding promoter regions shaded in gray exhibit >90% sequence identity. Corresponding intervening sequences do not display obvious similarity. (b–d) Localization of LePT4 and StPT4 reporter gene activity in AM fungal hyphae-containing root sectors. Magenta dye corresponds to GUS activity in mycorrhizal transgenic hairy roots carrying the LePT4-GUS (b, c) or StPT4-GUS (d, e) chimeric gene, respectively. Individual cells were dissected from hairy roots to colocalize GUS activity and intracellular fungal structures (c, e). Trypan blue staining was used to visualize fungal hyphae (c, e). White arrows indicate intracellular coiled hyphae. Bar in (a) = 50 bp, bar in (b), (c), (e) = 50 μm, bar in (d) = 100 μm.

Download figure to PowerPoint

To characterize the tissue-specific expression of LePT4/StPT4, the promoter regions of both genomic clones were linked with an exact fusion at the initiator ATG to the β-glucuronidase (GUS) reporter gene (Jefferson et al., 1987). These chimeric genes were introduced into potato and tomato hairy roots via Agrobacterium rhizogenes-mediated gene transfer. Potato hairy roots and tomato composite plants, the latter consisting of wild-type shoots and transgenic hairy roots, harboring either of the two chimeric reporter genes were subsequently inoculated with Gigaspora margarita and Glomus intraradices, respectively. GUS activity in mycorrhizal roots was detectable by histochemical staining in distinct areas corresponding to the zonation of fungal root colonization, whereas non-colonized neighboring cells were not stained (Figure 2b,d). Co-localization of Magenta-GUS stain with arbuscules or hyphae was demonstrated using trypan blue staining for fungal structures (Figure 2c,e). Thus, the expression of LePT4 and StPT4 is AM-dependent and is associated with cortex cells colonized by symbiotic AM fungal structures.

To corroborate and extend the GUS expression data (Figure 2b–e), RT-PCR with gene-specific primers was performed. In the two tomato cultivars MicroTom and Moneymaker, LePT4 transcripts were detected in mycorrhized roots, while LePT1 transcripts were well detectable in all organs tested (Figure 3a). In contrast, LePT2 transcripts were only detectable in roots, irrespective of whether a mycorrhiza was established or not (Figure 3a). In potato, a comprehensive semiquantitative RT-PCR analysis of Pht1 transporter gene expression in mycorrhized and non-mycorrhized plants, respectively, demonstrated StPT1 expression in both roots and leaves, and StPT2 expression predominantly in roots, with highest levels of transcripts of both genes in P-deprived roots, independent of whether the roots had formed a mycorrhiza or not (Figure 3b). There was StPT3 transcript abundance in mycorrhized roots, medium levels in non-mycorrhized roots, independent of the P status of the plant, and lowest levels in leaves, respectively (see also Karandashov et al., 2004). In contrast, at this level of sensitivity, StPT4 transcripts were exclusively detectable in mycorrhized roots. Thus the four Pht1 genes of potato are differentially regulated, with StPT4 exhibiting mycorrhiza-specific regulation, similar to LePT4 (Figure 3a), while StPT3 is a mycorrhiza-upregulated gene. However, during the course of this work we observed that cross-annealing of the StPT4-primer set used for the PCR can lead to the amplification of clones corresponding to a close paralogue of StPT4, named StPT5 (see below). To detect expression of the StPT3-orthologue LePT3 in tomato, a cDNA sequence encompassing the 3′-end of StPT3 was used as a radioactive probe on a tomato RNA gel blot (Figure 3d). RNA hybridizing to the StPT3 probe revealed an expression pattern of LePT3 similar to that of StPT3 in potato, but distinct from that of LePT4, that is, transcripts were abundant in mycorrhizas of a lept4 knockout mutant (see below).

image

Figure 3. Phosphate transporter transcript abundance in mycorrhizas. (a–c) Comparison of phosphate transporter transcript levels in tomato and potato tissues by RT-PCR. RT-PCR was performed with total RNA from roots and shoot organs, as indicated on top (a and c) or at right (b), respectively, using gene-specific primers for respective cDNAs, as indicated at left. StPT4 primers used in (b) also amplified StPT5 DNA (see Results). PCR was performed with total RNA subjected (+RT) or not (−RT) to reverse transcription in (c). (d) Randomly labeled cDNA probes, as indicated at left, hybridized to corresponding RNA on the gel blot (note that an StPT3 cDNA probe was hybridized to LePT3 RNA). Individual tomato or potato plants were grown at 5 μm Pi (a, c, and d) or at 5 μm Pi (−P) and 1 mm Pi (+P), respectively, in (b). Total RNA was extracted from respective organs of individual plants that were cultured without (−myc) or with (+myc) Glomus intraradices, respectively. RNA samples of two different plants from each genotype were used in (d) to demonstrate reproducibility of the relative signal strengths. 25S rRNA served as a control for equal loading of the gel and RNA integrity in (d). The tomato cultivars used were Moneymaker (Mm) and MicroTom (Mt); lept4-1 designates the LePT4 knockout mutant described in this work (a, c, and d).

Download figure to PowerPoint

Identification of partial clones corresponding to StPT5 and LePT5

The StPT4-primer sets used for iPCR and RT-PCR also amplified, in addition to StPT4 DNA, five DNA fragments corresponding to the novel gene StPT5, the partial sequences of which exhibited reduced identity, ranging from 85 to 91%, to the StPT4 clones (Figure S1). Phylogenetic tree analysis showed that the partial StPT5 (i.e., pStPT5 deduced from StPT5_cDNA_2a in Figure S1) translation product (acc. no. AY885654) clustered with proteins from subfamily I of mycorrhiza-inducible Pi transporters (Figure 1).

Subsequently, StPT4- and StPT5-specific primer pairs were developed and mycorrhiza-upregulated expression of the two putatively distinct genes was demonstrated (Figure 7a). To avoid laborious analysis of allelic sequence variation within the StPT4/StPT5 group and to support the hypothesis that StPT4 and StPT5 are two distinct non-allelic genes in the tetraploid potato genome, StPT5-specific primers were used to amplify cDNA corresponding to the orthologous LePT5 gene from diploid MicroTom tomato. A fragment of 0.4 kbp could be amplified, cloned, and sequenced sharing 68, 61, and 88% sequence identity with LePT1, LePT2, and LePT4, respectively. The partial LePT5 translation product (acc. no. AY885653), sharing 99.3% identity with the corresponding pStPT5 sequence, clustered with proteins from subfamily I of mycorrhiza-inducible Pi transporters (data not shown).

image

Figure 7. Pht 1 phosphate transporter gene expression and proposed evolution in solanaceous species. (a) Phosphate transporter transcript abundance in potato organs. Comparison of phosphate transporter transcript levels in potato tissues by semiquantitative RT-PCR. RT-PCR was performed with 27 cycles with total RNA from roots and shoot organs, as indicated on top, using gene-specific primers for respective cDNAs, as indicated at left. Potato plants were grown at 5 μm Pi. Total RNA was extracted from respective organs of individual plants that were cultured without or with Glomus intraradices (G.i.), respectively. (b) Diagram showing the proposed series of gene duplications which led to the formation of the potato and tomato Pht1 gene families. Ancestral Pi transporter (PT) genes arising after gene duplication are shown. After fixation of the duplicates, the expression pattern of the precursor gene is often partitioned between the two duplicates as a consequence of changes or deletions in promoter sequences (subfunctionalization) (Hurles, 2004). Greatest transcript amounts of respective tomato and potato genes as indicated at bottom were detectable in mycorrhized roots (+myc), Pi-deprived roots (rootP) and shoots (shootP), respectively.

Download figure to PowerPoint

LePT5 gene-specific RT-PCR utilizing the same cDNA libraries which were used for Figure 3(a) revealed differential expression of LePT5 when compared with LePT1, LePT2, and LePT4 (Figure 3a,c), respectively, with high LePT5 transcript abundance in mycorrhized roots of both MicroTom and lept4-1. Trace amounts of transcripts were also detectable in non-mycorrhized roots after 40 PCR cycles (Figure 3c). The expression of StPT5 and LePT5 was not detectable in shoot organs (Figures 3c and 7a). Non-reverse transcribed RNA samples did not give rise to amplification of the LePT5 cDNA indicating absence of genomic DNA in the PCR templates (Figure 3c). Therefore, it can be concluded that both tomato and potato contain three different Pht1 genes, which are strongly upregulated upon AM fungal colonization of their roots (see also Figure 7, and Discussion).

Functional analysis in yeast

For a functional analysis of the LePT4 and StPT4 gene products, the yeast mutants PAM2 (Martinez and Persson, 1998) and EY917 (Wykoff and O'Shea, 2001), which are defective in two and five Pi transporter (PHO) genes, respectively, were used for complementation analysis (data not shown). Transformed PAM2 or EY917 cells in logarithmic growth phase were resuspended in liquid medium containing different concentrations of Pi. The mutant cells carrying the empty expression vector grew generally poorly at Pi concentrations of ≤1 mm, while cells expressing the high-affinity Pi transporter StPT3 grew well (data not shown). Growth rates positively correlated with increasing Pi concentrations, clearly demonstrating that at micromolar concentrations Pi was growth-limiting. Strains carrying the LePT4 or StPT4 cDNA, respectively, or the M. truncatula orthologue MtPT4, generally grew at rates similar to those of the vector controls at millimolar Pi concentrations (data not shown). These results led us to suggest that both LePT4 and StPT4, due to high sequence similarity with their orthologue MtPT4 (Harrison et al., 2002), are functional as plasma membrane Pi transporters in yeast cells probably exhibiting a low affinity for Pi at millimolar concentrations, similar to the endogenous yeast low-affinity Pi uptake system (Tamai et al., 1985).

Functional analysis in planta

Genomic Southern blot analysis revealed the presence of a single copy of LePT4 in the haploid genome of tomato (Figure 4b). To analyze the function of the mycorrhiza-specific LePT4 gene in planta, a mutant of MicroTom tomato designated lept4-1 was characterized. Sequencing revealed that lept4-1 contained the insertion of a transposon, which is based on maize Ac/Ds and contains the luciferase gene (Meissner et al., 1997), 361 bp downstream of the LePT4 initiator ATG (Figure 4a). Transposon insertion resulted in an approximately 500 bp deletion in LePT4. DNA gel blot analysis with appropriately restricted wildtype and lept4-1 genomic DNA, respectively, and detection of the luciferase gene (Figure 4a,b) strongly suggested the presence of two closely linked Ac/Ds transposons inserted in the lept4-1 genome, of which one probably inserted as a truncated version outside of LePT4 within a few kbp upstream of the first insertion. RT-PCR analysis demonstrated expression of both LePT1 and LePT2 in roots of wild-type MicroTom plants and lept4-1 mutants growing on low P, irrespective of the presence or absence of root-colonizing AMF (Figure 4c). In contrast, LePT4 transcripts were only detectable in mycorrhized wild-type roots, while no signal was visible with mycorrhizas of lept4-1. Similarly, DNA gel blot analysis using LePT4 cDNA as a radioactive probe demonstrated high abundance of LePT4 cDNA exclusively in RT-PCR samples from mycorrhized wild type, while no LePT4 cDNA was synthesized from RNA templates from mycorrhized lept4-1, as well as non-mycorrhized wildtype and lept4-1, respectively (Figure 4c).

image

Figure 4. Copy numbers of LePT4 and luciferase in the wild type and lept4-1 tomato genomes, respectively. (a) Schematic representation of the Ac insertion into LePT4. The figure is not to scale. DNA restriction sites important for the interpretation of the data in (a) and (b) are indicated. (b) Wildtype genomic DNA was digested with EcoRV and SphI. Radioactively labeled LePT4 partial cDNA was subsequently hybridized to the blot (left panel). The number of Ac insertions into lept4-1 was estimated after blotting of genomic DNA from wild-type controls and lept4-1, respectively, digested with the indicated restriction enzymes (RV is EcoRV), and hybridization of a luciferase probe to the blot (central and right panel). The positions of DNA marker fragments in kb are indicated at the left of each panel. (c) RT-PCR performed with total RNA extracted from wild type and lept4-1 roots either colonized or not by Glomus intraradices as indicated at left. cDNAs were amplified with LePT1, LePT2, and LePT4 sequence-specific primers, respectively, over 35 cycles. Amplified fragments from mycorrhized roots were subsequently blotted from the gel to a nylon membrane and hybridized to a radioactively labeled LePT4 probe (lower panel).

Download figure to PowerPoint

Analysis of lept4-1 development under various experimental conditions demonstrated a constitutive growth phenotype in non-mycorrhized roots. Ten-day-old wild-type MicroTom seedlings were grown on vertically oriented 2MS agar plates and developed vigorous root systems with first and second order lateral roots. In contrast, lept4-1 roots remained short with very few laterals representing only 14% of the total number of wild-type laterals (n = 13, P < 0.01, Figure 5a and data not shown). As a consequence, total root length in lept4-1 was 61% lower than in the wild type (P < 0.01). Shoot total biomass was visibly reduced in lept4-1 seedlings, probably as a result of reduced nutrient uptake correlating with reduced absorptive root surface area. Similarly, roots of 6 week-old lept4-1 and MicroTom plantlets, respectively, cultivated in culture glasses on MS (1.25 mm Pi; n ≥ 14) or minimal medium (approximately 120 μm Pi; n = 4) were compared (Figure 5b). A reduction in the number of lateral roots by 48% (P < 0.01) was observed in lept4-1 when grown in minimal medium, while on MS medium, a 25% reduction was not statistically significant (P = 0.065). The reduced number of roots and shorter lateral roots in lept4-1 versus MicroTom added up to a 57% (P < 0.01) or 43% (P < 0.01) reduction, respectively, of total root length. Shoot fresh weights of the two genotypes did not differ significantly (Figure 5c). Similarly, total root length of lept4-1 was significantly shorter than that of controls when plants were grown in a 1:1 sand/soil mixture (R. Nagy, I. Jakobsen and M. Bucher, unpublished data). Similar root phenotypic differences were equally apparent in cultures of MicroTom and lept4-1 transgenic hairy roots containing either a StPT3 promoter-GUS control gene (StPT3-GUS; n = 4) or an StPT4 genomic fragment (gStPT4; n = 4), with the latter containing both the cloned StPT4 promoter and full-length downstream coding sequence. The respective transgenic lept4-1 hairy roots remained significantly shorter [29% (P < 0.01) or 18% (P < 0.01) of StPT3-GUS controls or gStPT4 expressing roots, respectively] with less lateral roots [41% (P < 0.01) or 23% (P < 0.01), respectively] (Figure 5d). Thus, ectopic expression of gStPT4 in hairy roots of lept4-1 did not restore the MicroTom root phenotype, strongly suggesting that the latter was independent from LePT4 and is therefore likely to be the result of the second Ac/Ds transposon insertion (see above).

image

Figure 5. The mutant lept4-1 exhibits altered root morphology. (a) Wild-type MicroTom and lept4-1 seedlings were grown on vertically oriented 2MS agar plates for 10 days, before seedlings were removed from the plates and the roots were carefully spread to allow calculation of laterals and total root length. (b) Shoot fresh weight from 3-week-old MicroTom wild type and lept4-1 seedlings, respectively, which were grown in tissue culture in MS (n ≥ 14) or minimal medium (n = 4), respectively, as indicated. (c) Growth parameters of roots from 3-week-old MicroTom wildtype and lept4-1 seedlings, respectively, as described above, are shown. (d) Transgenic hairy roots, harboring StPT3-GUS or the genomic StPT4 fragment including the promoter region, respectively, were grown for 2–3 weeks. Subsequently, growth parameters as indicated were analyzed. Note that the data obtained with StPT3-GUS and gStPT4 hairy roots, respectively, stem from two different experiments and therefore cannot be compared.

Download figure to PowerPoint

Under low P conditions, AMF often significantly enhance the ability of plants to scavenge for scarce and immobile nutrients. In Table 1, contents of P and micronutrients in shoots of non-mycorrhized or mycorrhized MicroTom and lept4-1 are shown. Plants were cultivated for 6 weeks in the compartmented cuvette system as described below. Except for Fe, levels of P, Zn, and Mn were significantly enhanced in both genotypes when they formed a mycorrhiza. The role of LePT4 activity in Pi transfer to the plant at the symbiotic interface (the mycorrhizal pathway) was directly assessed by measuring 33P-labeled Pi uptake into shoots in two different experimental systems, a compartmented cuvette system (Jansa et al., 2003) and a compartmented pot system (Smith et al., 2003) (Figure 6a,c). Common features of the two systems were (1) the main root + hyphal compartment (RHC) containing a soil/sand mixture, with or without incorporated AMF inoculum containing colonized roots, spores, and soil, and (2) the hyphal compartment (HC) separated by a 21-μm nylon mesh, which allowed AM fungal hyphae, but not roots, to penetrate from the RHC and absorb P. The soil for the HC was mixed with 33P-labeled Pi of high specific activity. In both systems, mycorrhized plants absorbed a high amount of 33P-label when compared with low (cuvette system) or negligible amounts (pot system) in plants with non-mycorrhized roots. Surprisingly, no differences in 33P-label were observed between wildtype and lept4-1 shoot material (Figure 6b,d), indicating normal Pi transfer activity at the symbiotic interface. We conclude therefore that a high degree of redundancy exists between LePT4 and its paralogues such as LePT3 and LePT5 (see also Figure 3c,d). A similar situation is expected to exist with the paralogous proteins StPT3/StPT4/StPT5 in potato.

Table 1.  Nutrient element analysis of mycorrhized (+myc) or non-mycorrhized (−myc) roots of MicroTom wildtype and lept4-1
Plant materialP**Zn*Mn*Fen.s.
  1. Values represent the mean ± standard error (SE). All values are given in mg/kg DW. The different letters for each individual element indicate significant differences, while similar letters indicate non-significant differences. The level of significance between the different values is indicated by the number of stars following the symbol of each element. **P ≤ 0.01; *P ≤ 0.05; n.s., non-significant. There is an exception for Mn analysis, where a differs from b at P ≤ 0.05 (Mn *), and b differs from c at P ≤ 0.01 (**b, **c). n = 5.

wt − myc959.8 ± 21.3a19.97 ± 2.4a21.9 ± 1.4a69.12 ± 4.3a
wt + myc2166.6 ± 105.4b30.6 ± 4.1b26.3 ± 0.9**b86.68 ± 15a
lept4-1 − myc1295.5 ± 162.8a16.3 ± 0.5a17.5 ± 1.1**c50.4 ± 13.1a
lept4-1 + myc2030 ± 104.2b27.8 ± 3.8b25 ± 0.8**b55.8 ± 5.1a
image

Figure 6. Diagram (not to scale) showing the compartmented system's design and plant 33P-labeled Pi (33Pi) uptake. (a) The compartmented cuvette system (Jansa et al., 2003) and (c) the compartmented pot system (Smith et al., 2003) consist of the main root + hyphal compartment (RHC) containing a soil/sand mixture, with or without incorporated AMF inoculum (G.i. is Glomus intraradices), and the hyphal compartment (HC), separated by a 21- or 25-μm nylon mesh, respectively, which allows AM fungal hyphae (shown as lines entering the HC), but not roots, to penetrate from the RHC and absorb P. In the compartmented pot system, the HC was a small plastic tube. The soil substrate in the HC was well mixed with 33P-labeled orthophosphate of high specific activity. There was one plant per cuvette (n = 5) and three plants per pot (n = 15); for simplicity, only a single plant is illustrated in both systems. (b) and (d) Mycorrhizal effects on 33P uptake in mycorrhized or non-mycorrhized MicroTom wild type and lept4-1 as indicated.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Two transport systems involved in mycorrhizal Pi uptake in solanaceous species

Pi uptake at the perifungal interface in a mycorrhiza is mediated by plant Pi transporters and is a major parameter contributing to the mutual benefit of the AM symbiosis (Harrison, 1997; Karandashov and Bucher, 2005; Rausch et al., 2001; Smith and Read, 1997). Recent analysis of the phylogenetic relationship between transporter proteins involved in AM symbiotic Pi transfer revealed that OsPT11 from rice and MtPT4 from M. truncatula are non-orthologous to the potato StPT3 protein (Figure 1; Harrison et al., 2002; Paszkowski et al., 2002). As seen in Figure 1, the orthologous transporters OsPT11, MtPT4, LePT4, LePT5, StPT4, and StPT5 comprise subfamily I, which thus includes members from both dicot and monocot species, similar to subfamily II. In contrast, StPT3 clusters in Pht1 subfamily III with proteins solely from dicotyledonous species, including non-mycorrhizal Arabidopsis (Figure 1). Therefore, subfamilies I and II are evolutionarily older than subfamilies III and IV. Conservation of sequence among proteins from evolutionary distant plant species can indicate conservation of their function, which is substantiated in subfamily I by a very similar expression pattern of the respective genes in mycorrhizas (Figures 2b and 3; Harrison et al., 2002; Karandashov et al., 2004; Paszkowski et al., 2002). Based on the absence of structural commonalities in the promoter regions of StPT3 and OsPT11, Paszkowski et al. (2002) previously concluded that, ‘although cytological and physiological features of the AM symbiosis seem to be conserved, the molecular components may differ significantly between distantly related plant species.’ Our data provide evidence for the concurrent operation of two non-orthologous Pi transport systems in mycorrhizas of a single solanaceous plant species (i.e., LePT3/LePT4 in tomato, and StPT3/StPT4 in potato, respectively) (Figures 1 and 3), while in the monocot rice, only one Pi transporter out of a total of 13 Pht1 family members was expressed in a mycorrhiza-specific way (Paszkowski et al., 2002). It will be of special interest to see whether this discrepancy applies to dicotyledonous and monocotyledonous species in general, and if so, what are the biological and ecological consequences of this genetic diversity. Moreover, sequence and expression data indicated a recent gene duplication event resulting in the generation of LePT5/StPT5 which are paralogous to LePT4/StPT4 (Figures 3c and 7).

Regulation of mycorrhiza-specific Pi transport

Both the LePT4 and StPT4 promoters share a high degree of sequence identity (Figure 2a), and direct gene expression in a way similar to that of the StPT3 promoter, that is, transcripts are highly abundant in mycorrhizas, and GUS activity is predominant in sectors of a mycorrhiza colonized by the AMF (Figure 2b–d; Karandashov and Bucher, 2005; Karandashov et al., 2004; Rausch et al., 2001). The regulatory mechanism(s) controlling StPT3 expression are conserved within evolutionary distant dicotyledonous species and become operative in cortex cells colonized by hyphae, hyphal coils, or arbuscules, respectively (Karandashov et al., 2004). We previously speculated that a transposon-like element may participate in mycorrhiza-specific StPT3 regulation (Rausch et al., 2001). However, sequence analysis did not uncover such an element within the LePT4 or StPT4 promoter region, respectively (data not shown), indicating that this element is not essential for mycorrhiza-specific Pi transporter gene regulation. Its presence may rather be evidence of a high recombination frequency in the respective genic region during StPT3 promoter evolution, eventually leading to mycorrhiza-specific upregulation of the respective gene. Alignment of promoter sequences of mycorrhiza-upregulated Pi transporter genes is not sufficient for the identification of putative regulatory motifs within these promoters. Phylogenetic footprinting is a suitable method to identify conserved motifs in normally non-conserved promoter sequences (Hong et al., 2003; Lenhard et al., 2003) and has recently led to the identification of candidate regulatory elements in promoter sequences of StPT3, LePT4, and MtPT4, respectively (Karandashov et al., 2004). It is not surprising that due to high sequence similarity between the LePT4 and StPT4 promoters (Figure 2a), the spatial distribution of the previously described putative mycorrhiza-specific cis-acting elements is very similar in the two promoters (data not shown). We therefore assume that evolutionary ancient regulatory mechanism(s) forming the basis of mycorrhiza-specific Pi transport are conserved within dicotyledonous species and regulate expression of subfamily I and functionally related subfamily III proteins (i.e., LePT3 and StPT3) (Figure 1). Moreover, the previous observation that the rice OsPT11 promoter does not direct GUS gene expression in transgenic potato roots upon AM fungal colonization suggests that since the evolutionary separation of monocots and dicots the regulatory mechanism(s) involved in AM-specific Pi transport have diverged in angiosperms (Karandashov et al., 2004).

Subfunctionalization within the Pht1 gene family in tomato and potato

An StPT3 cDNA probe hybridized to tomato mRNA on the RNA gel blot under stringent conditions (Figure 3d), reflecting expression of the orthologous LePT3 gene in mycorrhized tomato roots. During the course of this work, LePT3 was cloned and its expression was shown to be upregulated upon mycorrhization similar to StPT3 (Avner Silber, personal communication, Institute of Soil, Water and Environmental Science, the Volcani Center, Bet Dagan, Israel; Rausch et al., 2001). Thus, at least five differentially expressed Pht1 genes exist in both tomato and potato (Figures 3, 7 and S1). For the establishment of the mycorrhizal symbiosis in evolution, changes in root structure and function (Brundrett, 2002; Raven and Edwards, 2001) needed to be associated with changes in Pi transporter gene expression and activity to facilitate Pi transfer at newly established plant interfaces, that is, the root periphery including root hairs and the AM symbiotic interface in root cortex cells. This adaptation most likely contributed to overcoming the strong selection pressure imposed on primitive plants (e.g., by limited water and nutrient supply in the terrestrial habitat) during the colonization of land 460–800 Ma. During evolution and speciation, gene duplication events and ultimate fixation of the duplicated genes in populations of single-copy genes led to the generation of gene families (Hurles, 2004). Subsequent changes in regulatory and coding sequences of the duplicated genes led to the acquisition of novel expression sites and subsequent subfunctionalization (Force et al., 1999). Subfunctionalization within the Pht1 family is likely to have increased the fitness of terrestrial plants. Based on the phylogenetic tree (Figure 1) and the expression analyses (Figures 3 and 7a), we propose that the first of three subsequent events led to the duplication of an ancestral gene to precursors of LePT1/StPT1, 2, and 3 on the one hand, and mycorrhiza-specific LePT4/StPT4, and 5 on the other (Figure 7). Evolutionarily, the most recent duplications eventually led to the divergence of (1) LePT1/StPT1 and LePT3/StPT3, exhibiting constitutive and mycorrhiza-upregulated gene expression (Figure 3b; Karandashov et al., 2004), respectively, and (2) LePT4/StPT4 and LePT5/StPT5, both exhibiting mycorrhiza-upregulated gene expression (Figures 3c and 7). It is tempting to postulate that, along this evolutionary trend, LePT4/StPT4 transporters ultimately evolved being strictly expressed in the mycorrhizal symbiosis. In the future, detailed promoter studies should eventually reveal which regulatory elements are conserved and required to direct tissue-specific or environment-dependent expression of members within the Pht1 gene family.

Functional diversity of Pht1 proteins in mycorrhizal Pi transport

Pht1 proteins share similarity with the proton-coupled Pi symporters from yeast, Neurospora crassa, and the AMF G. versiforme and are unrelated, both with respect to sequence and function, to the low-affinity Pi transporter Pht2;1 (Figure 1), which was originally cloned from Arabidopsis (Daram et al., 1999) and was later shown to be localized to plastids (Ferro et al., 2002; Rausch et al., 2004; Versaw and Harrison, 2002). Apparent Km values of Pht1 transporters for Pi have been calculated from only a few such proteins via their expression in yeast mutants (Daram et al., 1998; Harrison et al., 2002; Leggewie et al., 1997; Rausch et al., 2001), tobacco cells (Mitsukawa et al., 1997), or transgenic rice plants (Rae et al., 2003), respectively, overall ranging from 3 to 668 μm. The Km values calculated for mycorrhiza-specific Pi transporters were 64 μm for StPT3 (Rausch et al., 2001) and 493/668 μm for MtPT4 using two different yeast mutants (Harrison et al., 2002), respectively. In contrast, our detailed functional analysis of LePT4 and StPT4 in yeast did not yield clear-cut results. This was probably due to Pi uptake at millimolar concentrations in PAM2 (Martinez and Persson, 1998) and EY917 (Wykoff and O'Shea, 2001) cells, which could have interfered with the activity of the two plant proteins toward Pi. Significant and reproducible growth improvement with the yeast cells was only observed when the high-affinity Pi transporter StPT3 was used for complementation (data not shown). Therefore, we assume that both LePT4 and StPT4, as well as LePT5 and StPT5, all four sharing high similarity with MtPT4, are low-affinity Pi transporters. The selection for the high-affinity StPT3/LePT3-type system must therefore have occurred after monocotyledonous grasses such as rice diverged from the eudicot lineage about 150 Ma (Wikstrom et al., 2001). Thus, the non-orthologous Pi transport systems residing at the symbiotic interface in tomato and potato facilitate transport over a wide range of interfacial Pi concentrations, suggesting that the perifungal Pi concentration in roots can vary substantially, probably depending on the physiology of the symbiotic fungus. Important factors contributing to variations in perihyphal Pi concentrations are likely to be Pi uptake in extraradical hyphae (Cox and Tinker, 1976; Harrison et al., 2002; Smith et al., 1994), polyphosphate (polyP) homeostasis, hyphal translocation of polyP and subsequent hydrolysis in intraradicle hyphae (Viereck et al., 2004), followed by hyphal Pi efflux dependent on divalent cations (Cairney and Smith, 1993). The coordinated activity of the symbiotic Pi transporters would ensure efficient uptake of fungal Pi, which, as has recently been shown for tomato, can account for up to 100% of total P taken up by the plant (Smith et al., 2003).

Functional redundancy between mycorrhiza-specific Pi transporters

While our work was in progress, functional genomic analysis of two Pht1 transporters from Arabidopsis thaliana demonstrated a 75% reduction in Pi uptake in a loss-of-function mutant lacking both Pht1;1 and Pht1;4 (Shin et al., 2004), corroborating the presumed function of Pht1 proteins in plant Pi uptake (Raghothama, 1999; Rausch et al., 2001; Schachtman et al., 1998). Unfortunately, in Arabidopsis, as in Brassicacean species in general, roots are not colonized by AMF and this species is therefore not a suitable model plant to investigate the molecular basis of symbiotic P uptake. In contrast, tomato is normally a well mycorrhized species in which functional diversity of symbiotic P transfer has recently been studied (Smith et al., 2003, 2004). Thus, analysis of lept4-1 allowed for testing of the functional role of LePT4 in the mycorrhiza. Insertional mutagenesis with maize Ac/Ds transposable elements in the background of the miniature tomato cultivar MicroTom results, on average, in two to three Ds inserts per line (Meissner et al., 2000). Genomic DNA gel blot analysis (Figure 4, and data not shown) suggested that two closely linked insertions are present in lept4-1, with insertion of one element leading to a complete disruption of LePT4 expression (Figure 4). Heterologous expression of the genomic fragment gStPT4 did not restore the mutant phenotype in lept4-1 hairy roots (Figure 5d). We therefore assume that the second insertion may be responsible for the disruption of a gene that is involved in lateral root initiation and elongation. Further genetic analysis of lept4-1 is required to test this hypothesis.

Despite strongly reduced total root length in lept4-1 (Figure 5), which is likely to result in a reduced Pi uptake via the direct pathway (see Introduction), transport of 33P-labeled Pi via the mycorrhizal pathway was unaffected in two different experimental systems (Figure 6) using two isolates of G. intraradices, which is efficient in delivering Pi to the plant symbiont (see e.g. Jansa et al., 2003; Smith et al., 2003). A similar result was obtained under high P conditions in a 1:1 sand-soil substrate (R. Nagy et al., unpublished data). This indicates a high degree of functional redundancy between the two transport systems, LePT4 and LePT5, and probably LePT3, in tomato, and suggests that LePT4 is dispensable for symbiotic Pi transfer under our experimental conditions. Redundancy within mycorrhiza-inducible Pi transporters ensures that symbiotic Pi transfer will be relatively insensitive to mutations and evolutionary robust. Functional diversity in AM symbiosis is probably widespread globally and relates diversity in (1) the fungal contribution to plant P uptake, and (2) dependence of growth of the plant host on mycorrhizal P transfer (Smith et al., 2003, 2004). Hence, detailed analysis of LePT3, LePT4, and LePT5 function should include particular physiological states of either of the two symbiotic partners, or interactions of tomato with different AMF species. In addition, downregulation of paralogous Pi transporters, for example, LePT3, alone and in combination with LePT4 and LePT5 will be needed to further elucidate the role of each Pht1 family member in mycorrhizal Pi uptake in tomato.

Strains and plasmids

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

The plant material was from tomato [L. esculentum cv. MicroTom (Meissner et al., 1997 and cv. Moneymaker], potato (Solanum tuberosum cv. Désirée), and plantain (Plantago lanceolata). AMF were G. margarita (BEG34), Glomus caledonium (BEG20), and G. intraradices (BEG75 and BEG87). The LePT 4 knockout mutant lept4-1 originated from a mutant population generated by Ac/Ds transposable element-mediated mutagenesis in the cv. MicroTom background. Sequencing of Ac/Ds transposon insertion flanking sites allowed the identification of the LePT4 gene. The LePT4 genomic fragment was PCR amplified from MicroTom genomic DNA with the gene-specific primers Lp1, 5′ GGTTTATCAGGGAATTTCTTGTTC-3′ and Lp2, 5′CGTCCTTTTGAATGTTTACACATG-3′. The PCR product was subsequently cloned into modified pBluescript KS(-) (Stratagene, Amsterdam, The Netherlands) harboring T-overhangs. The LePT4 promoter region was cloned using the iPCR technique (Szabados et al., 2002). An NcoI restriction site was introduced at the initiator ATG of the LePT4 coding region via PCR according to standard procedures. The NcoI site enabled the direct fusion of the LePT4 promoter at the initiator ATG to the β-glucuronidase (GUS) marker gene/nopaline synthase (nos) terminator from plasmid 5′AKT1-320X (Lagarde et al., 1996) into the binary vector Bin19 (Bevan, 1984). The yeast strains EY917 (MATαpho84Δ::HIS3 pho87::CgHIS3 pho89::CgHIS3 pho90::CgHIS3 pho91ADE2; Wykoff and O'Shea, 2001) lacking five Pi transporter genes, and PAM2 (Δpho89::TRP1Δpho84::HIS3 ade2 leu2 his3 trp1 ura3; Martinez et al., 1998), which is devoid of high-affinity Pi transport but harbors a single low-affinity Pi uptake system operative under Pi-rich conditions (for review see Borst-Pauwels, 1981), were used for complementation studies with the yeast shuttle vector 181A1NE as described previously (Daram et al., 1998). Cultivation of EY917 on glucose-containing medium causes synthetic lethality. Growth is resumed by growing the cells in galactose-containing media in the presence of a plasmid containing PHO84 under the control of the GAL1 promoter. Escherichia coli (DH5α, Bethesda Research Laboratories, MD, USA) and A. rhizogenes (ATCC 15834, The Belgian Co-ordinated Collections of Micro-organisms, http://www.belspo.be/bccm/bccm.htm) were cultured according to standard techniques. The LePT4 coding region was subcloned into the SmaI site of 181A1NE. Orthologous StPT4 from potato was cloned via iPCR using sequence information from LePT4 and mycorrhiza-specific Pi transporters from barrel medic (M. truncatula), and rice (Oryza sativa). A 1164-bp fragment of StPT4 cDNA was amplified using StPT4-1 (5′-TTCACTGATGCATATGATCTGTTCTG-3) and StPT4-4 (5′-ATTGGGGCCAAAATTGGCAAAGAAG-3′) primers from a potato/G. caledonium cDNA library, cloned into the pPCR-ScriptTM Amp SK(+) plasmid (Stratagene) according to the manufacturer's instruction, and sequenced. The 5′ and 3′ regions of the StPT4 gene including the promoter, were cloned via iPCR and 3′RACE. For iPCR, 5  μg genomic DNA isolated from potato grown in tissue culture were digested with Ecl136II/SnaBI and Ecl136II/EcoRV, self-ligated and used for PCR. Two sets of primers were used for iPCR: StPT4-iPCR1 (5′-GGCCTAAAAGTTTCGAGATAGTAGTG-3′) and StPT4-iPCR2a or StPT4-iPCR2b (5′-GGATGAAATTTTGTCCACGGAGCC-3′ or 5′-GATGAAGTTTTCTCCACGGAGCC-3′, respectively); StPT4-iPCR3 (5′-CGCGGCTTGTTTAGCATTTCCTTC-3′) and StPT4-iPCR4a or StPT4-iPCR4b (5′-GACTAAAGAGCACAAATGGACATTCGC-3′ or 5′-GACTAAAGAACACAAATGGACATTTGC-3′, respectively). For 3′RACE, a potato/G. caledonium cDNA library was generated using T18 Adapter primer (5′-CTCTGAATTCAAGCTTGGATCCT18-3′). PCR was performed with StPT4-4Fw (5′-ATGAAGGGAAGCCATTTGATGTG-3′) and Adapter primer (5′-CTCTGAATTCAAGCTTGGATCC-3′). Amplified fragments were cloned into the pDrive Cloning Vector (Qiagen, Basel, Switzerland) and pGEM®-T vector (Promega; Catalys, Wallisellen, Switzerland), respectively. Sequences covering the StPT4 promoter and the coding regions were subsequently used for further analysis. Corresponding fragments were amplified from potato genomic DNA with primers StPT4-4Fw and StPT4-4Rev (5′-GAGAGAATTCATCAGCCCTTTGTATGCTGCTAAC-3′ and 5′-CTCTGGATCCTATCATTCCCATCCGTCGTC-3′, respectively) containing EcoRI and BamHI sites, respectively, cloned into EcoRI/BamHI restricted pBluescript KS(-) plasmid, and sequenced. Promoter regions were amplified from the plasmid with primers StPT4-4Fw and StPT4-5Rev, the latter containing a terminal BamHI site (5′-CTCTGGATCCGTTATTTTATTTTCTCTGAGATTGGTTGATGTCTGATC-3′) and were ligated into EcoRI/BamHI-restricted Bin19 vector containing a β-glucuronidase (GUS) marker gene/nopaline synthase (nos) terminator construct (Rausch et al., 2001). The StPT4 coding sequence was amplified from the plasmid with primers StPT4-4Rev and StPT4-6Fw, the latter containing a BamHI site (5′-GAGAGGATCCCATGGCCTCAGACAATCTTGTAGTGC-3′) and ligated into BamHI-restricted yeast expression vector 181A1NE.

Computational analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Multiple sequence alignment and topology prediction were performed as described (Daram et al., 1999). Phylogenetic tree analysis was performed with the Phylip drawtree program at the Pasteur Institute (http://bioweb.pasteur.fr/seqanal/interfaces/drawtree.html). The phylogenetic tree (Figure 1) is a reconstructed tree based on sequence similarities. The LePT4 and the StPT4 protein sequences were scanned for patterns using ScanProsite (http://www.expasy.org/tools/scanprosite/) on the ExPASy proteomics server of the Swiss Institute of Bioinformatics (http://www.isb-sib.ch/).

Plant growth conditions for gene expression studies

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Tomato seeds were germinated in sterile conditions on MS medium supplied with 2% sucrose (Murashige and Skoog, 1962). Stock cultures of potato plants were maintained in tissue culture under the same conditions. Two- to three-week-old tomato seedlings or potato plantlets, respectively, were first adapted to conventional greenhouse conditions for 7–10 days, in pots containing sterilized quartz sand. After adaptation, an equal number of plants was transferred either to pots containing a soil/quartz sand mixture (1:10) or to pots containing the same mixture supplemented with roots of P. lanceolata colonized with G. intraradices (BEG 75). Fertilization was carried out with quarter-strength Hoagland medium (Hoagland and Broyer, 1936) containing 500 μm NH4H2PO4 prior to inoculation and 5 μm NH4H2PO4 during AMF colonization using drop irrigation. Plants were harvested 4 or 6 weeks, respectively, after inoculation. Randomly chosen root segments were used for trypan blue staining (Rausch et al., 2001), while the rest was used for RNA isolation (see below).

Extraction of nucleic acids and gel blot analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Total RNA from different organs from tomato (cvs Moneymaker and MicroTom) was isolated using the hot phenol extraction method (Verwoerd et al., 1989). Ten micrograms of total RNA was electrophoretically separated on denaturing formaldehyde gels with 1% agarose. High molecular weight genomic DNA was isolated from young tomato leaves as described previously (Dellaporta et al., 1983). Genomic DNA was digested overnight with EcoRV and SphI and was subsequently separated electrophoretically on a 0.8% agarose gel. Nucleic acids were blotted onto Hybond-NX nylon membrane (Amersham Biosciences, Zurich, Switzerland). Hybridization was carried out with 5x SSC, 5% Denhardt's, and 0.5% SDS (w/v) at 68 or 65°C for RNA or DNA gel blots, respectively, with a final wash using 0.1x SSC, 0.1% SDS at 65°C (Sambrook et al., 1989). Radioactively labeled LePT4 fragments (nucleotides 464–1526), StPT3 3′-end fragments, and the luciferase gene from plasmid pSP-luc+ (Catalys, Wallisellen, Switzerland) were used as probes.

RT-PCR

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

RT-PCR was performed as previously described (Bucher et al., 2002), using gene-specific primers for LePT1 (5′-CATTGTTTCTGCAGCATTCAAGG-3′ and 5′-CTTCCCATTGGATTCTGGAACC-3′), LePT2 (5′-GATTCGATCACGCGTATAGATCC-3′ and 5′-CCAGTTATAGTTTCTTGTGATGC-3′), and LePT4 (5′-GAAGGGGAGCCATTTAATGTGG-3′ and 5′-CCATCTTGTGTGTATTGTTGTATC-3′), StPT1 (5′-AATGAATTTGGTTTGTTCAGTAAGG-3′ and 5′-AAACTTAAACAGGACTGTCCTTCC-3′), StPT2 (5′-AATCAATTCGGTCTGTTTTCATGGGAA-3′ and 5′-CAACAAACAAGCTTACACAATACAAAG-3′), StPT3 (5′-GAGACGATGAACGCGTTGGATGAG-3′ and 5′-GGGCGCATTTATGTATTAAACTGG-3′), StPT4 (5′-TTCACTGATGCATATGATCTGTTCTG-3′ and 5′-CGCGGCTTGTTTAGCATTTCCTTC-3′; it has to be noted that during the course of this work this pair was shown to cross-anneal with StPT5 DNA), StPT4-specific (5′-ATGAAGGGAAGCCATTTGATGTG-3′ and 5′-CTCCCTCAAGGCGGATATCGTG-3′) and StPT5/LePT5 (5′-TTGGGGGTAAAGCGTTTACTACT-3′ and 5′-TTCCCTCAAAGCTGAAATGTTCTTC-3′).

Plant hairy root cultures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Agrobacterium rhizogenes was transformed with pBin19 carrying LePT4 and StPT4 promoter-GUS chimeric genes, respectively, via electroporation. Potato hairy roots were generated in sterile conditions essentially as previously described (Karandashov et al., 2004) using wild-type potato plantlets. Tomato composite plants consisting of untransformed shoots and transgenic hairy roots were generated by wounding sterile MicroTom and lept4-1 tomato plantlets with a needle containing agrobacterial suspension. Potato and tomato hairy roots appeared within about 2 weeks. Clonally independent potato hairy roots were selected and inoculated with spores of G. margarita as previously described (Karandashov et al., 2004), whereas tomato hairy roots were inoculated with spores of G. intraradices. Tomato composite plants carrying hairy roots harboring the LePT4-GUS construct were introduced to already established mycorrhizae of G. intraradices on P. lanceolata as described previously (Karandashov et al., 2004). Few weeks after inoculation, histochemical staining for GUS activity and AMF mycelium visualization were performed as previously described (Karandashov et al., 2004).

Determination of growth characteristics

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Shoot dry weight was measured after incubation of the material in an oven for 2 days at 80°C. Dried shoot material was homogenized, incinerated for 8 h at 550°C and subsequently 100 mg ash were solubilized in 2 ml 6 m HCl, briefly heated up to 100°C, purified through Whatman No. 40 ashless filter paper and transferred to double-distilled water to a total volume of 50 ml. The concentrations of P and other elements (Fe, Mn, and Zn) were measured using ICP emission-spectroscopy (Varian Liberty 220; Varian Inc., Palo Alto, CA, USA, equipped with an ultrasonic nebulizer (CETAC U-5000 AT+) according to standard procedures.

In order to study whether the mutation in LePT4 is involved in the root phenotype, growth of MicroTom tomato and lept4-1 was studied in vitro. Plantlets were grown in sterile glass vessels containing 50 ml of either MS (see above) or minimal (Karandashov et al., 2004) nutrient media. After 3 weeks of growth, shoots were detached from roots and were subsequently transferred to new culture vessels after fresh weight determination, where they were inoculated with A. rhizogenes carrying the pBin19 plasmid containing either StPT3 promoter-GUS reporter-nos terminator or StPT4 genomic fragment (promoter + coding sequence)-nos terminator chimeric genes as described above. After detachment of shoots, the roots were extracted from nutrient medium, washed, stained with 0.05% neutral red for 24 h, and scanned. Total and primary root length, respectively, were analyzed using WinRHIZO Pro version 2003b software (Régent Instruments Inc., Québec, Canada). The number of lateral roots was counted manually. Developed hairy roots were aseptically transferred to 500-ml conical flasks, containing 50 ml minimal medium. Roots were grown for 2–3 weeks and then processed as described above. PCR was performed to confirm genetic transformation of hairy root clones.

Measurement of 33P uptake

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

The radioactive P uptake experiment in the compartmented cuvette system was performed as previously described (Jansa et al., 2003) with modifications. One week prior to inoculation, plants from sterile cultures were adapted to climate chamber conditions with a day/night cycle of 15/9 h, temperature settings of 22/18°C, and a relative aerial humidity of 80/70%, respectively. Eight kilograms of commercially available soil supplied with low levels of inorganic nutrients, was labeled with 185 MBq of 33Pi (Hartmann Analytic, Braunschweig, Germany) at a final concentration of 23.1 MBq kg−1 and equally distributed in the radioactive compartments of 20 cuvette units. Ten plant compartments (see Figure 6a) at a distance of 3 cm from the radioactive compartments were inoculated with G. intraradices, while 10 additional compartments were left un-inoculated, serving as controls. Soil moisture was kept constant by using a time-controlled automatic watering facility. Plants were fertilized weekly with half-strength Hoagland nutrient solution lacking phosphate (NH4H2PO4 was replaced with NH4Cl at 1 mm). Shoots and roots were harvested 6 weeks post-inoculation. Plant shoots were used for ICP emission spectroscopy as described above. After neutralization of the extracts with NaOH, radioactivity was measured by β-scintillation counting in a Packard (Packard BioScience, Meriden, CT, USA) 2500 TR counter using the Packard Ultima Gold scintillation cocktail. Plant roots were stained with trypan blue for the presence of AM fungal structures.

The experiment with the compartmented pot system was performed as reported by Smith et al. (2003, 2004) with minor modifications. MicroTom and lept4-1 seeds, respectively, were germinated on sterilized vermiculite. Three 5-day-old seedlings of each genotype were transplanted into compartmented pots for the uptake experiment. The RHC contained 925 g of semi-sterile (15 kGy, 10 MeV electron beam) 1:1 soil/ sand mixture (9 mg P kg−1 soil; Olsen et al., 1954) supplied with nutrients, except P (Pearson and Jakobsen, 1993). Mycorrhizal treatments had 75 g dried inoculum of G. intraradices, BEG 87 (pot soil containing spores and colonized root segments of Trifolium subteraneum) mixed into the 925 g soil, while equivalent amounts of semi-sterile soil were used for the non-mycorrhizal treatments. The HC was a small plastic vial capped with a 25-μm nylon mesh. The HC contained 27 g of inoculum-free soil, labeled with 33Pi (Amersham Biosciences, Cardiff, UK; 5 kBq g−1 soil) and was placed 5 cm below the soil surface with the mesh facing the center of the pot. Pots were watered as required to maintain the moisture at 65% of the water-holding capacity. Plants were grown in climate chambers with a day/night cycle of 16/8 h, and temperature settings of 23/18°C, respectively. Plant material was harvested 6 weeks after sowing. Fresh weight was determined and the plant material was dried for 2 days at 80°C and weighed. Ground, well homogenized shoot samples were digested in 4:1 nitric–perchloric acid mixtures and radioactivity was determined as described above. Randomly chosen root samples were stained using trypan blue and checked for the presence of AM fungal structures.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

We are very grateful to Anette and Anne Olsen, Risø National Laboratory, and Dr Jan Jansa and Richard Ruh, ETH Zurich, for their support in the experiments with the compartmented systems; Dr Christine Rausch for RNA samples; Thomas Flura (ETH Zurich) and Theres Rösch (ETH Zurich) for ICP measurements; Sarah Wegmüller (ETH Zurich) for phylogenetic footprinting analysis and promoter sequence comparison; David Drissner (ETH Zurich) for statistical analysis; and Dr Emmanuel Frossard and Dr Markus Liedgens (ETH Zurich) for the use of the WinRHIZO software.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Strains and plasmids
  8. Computational analysis
  9. Plant growth conditions for gene expression studies
  10. Extraction of nucleic acids and gel blot analysis
  11. RT-PCR
  12. Yeast manipulations
  13. Plant hairy root cultures
  14. Determination of growth characteristics
  15. Measurement of 33P uptake
  16. Acknowledgements
  17. Supplementary Material
  18. References
  19. Supporting Information

Fig. S1.  A compilation of all StPT4 and StPT5 sequences identified during the course of this work.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
TPJ_2364_sm_figureS1.rtf505KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.