Introduction

Peptide transport is characterized by the ability of cells to transport peptides or peptide derivatives across membranes in an energy-dependent manner. Peptide transporters can be classified into three groups, namely the ATP-binding cassette transporters (ABC family), the peptide transporters (PTR family) and the oligopeptide transporters (OPT family). Plant ABC transporters use ATP hydrolysis to energize transport and include the Arabidopsis MRPs (multidrug-resistance-associated proteins), which were shown to transport glutathione (γ-Glu-Cys-Gly), glutathione S-conjugates and glucoronides (reviewed by Blackmore et al. 2001; Detmers et al. 2001; Rea 1998). PTR transporters use the proton gradient to drive transport. PTR proteins were shown to transport a wide variety of nitrogen-containing substrates such as amino acids (Frommer et al. 1994; Zhou et al. 1998), short peptides (di- and tripeptides) (Song et al. 1996; Steiner et al. 1994; Chiang et al. 2004) and nitrate (Tsay et. al. 1993). Although PTR proteins are well characterized from bacteria, fungi and human, plant PTRs are less studied (reviewed by Stacey et al. 2002a; Williams and Miller 2001). Sequence comparisons identified 51 putative PTR genes in the Arabidopsis genome (Stacey et al. 2002a).

Members of the OPT family were first characterized in yeast and likely use the proton gradient to energize transport (Lubkowitz et al. 1997, 1998). In contrast to PTR peptide transporters which recognize di- and tripeptides, the yeast OPT transporters predominantly recognize tetra- and pentapeptides ((Lubkowitz et al. 1997; Lubkowitz et al. 1998), including the endogenous opioids Met-enkephalin (YGGFM) and Leu-enkephalin (YGGFL) (Hauser et al. 2000). The Saccharomyces cerevisiae Opt1p was also shown to be a high-affinity glutathione transporter (Bourbouloux et al. 2000). Database sequence comparisons with Candida albicans CaOpt1p identified nine putative Arabidopsis OPT transporters, AtOPT1 to AtOPT9 (Koh et al. 2002). These AtOPTs share 49–53% sequence similarity to CaOPT1p and show no sequence similarity to ABC or PTR transporters. OPT proteins were also identified in other plants, namely, the Brassica juncea BjGT1 (Bogs et al. 2003) and the Oryza sativa OsGT1 (Zhang et al. 2004). The plant OPTs form a distinct subfamily compared to the fungal OPTs (Yen et al. 2001; Koh et al. 2002; Zhang et al. 2004).

Koh et al (2002) showed that expression of AtOPTs 1, 4, 5, 6 and 7 (of 7 AtOPTs tested) in yeast enabled the uptake of at least one of the synthetic peptides KLLG, KLGL or KLLLG (Koh et al. 2002). More recently, BjGT1 (Bogs et al. 2003), OsGT1 (Zhang et al. 2004), and AtOPT6 (Cagnac et al. 2004) have been reported to enable the uptake of glutathione (GSH and GSSG) and glutathione derivatives (GS conjugates) when expressed in yeast. Subsequent to the identification of the first plant OPT transporters, more extensive sequence comparisons led to the identification of a more extended protein family (Yen et al. 2001). These include the Myxococcus xanthus EspB protein (Cho and Zusman 1999), involved in temporal pattern development during sporulation, and the Zea mays Yellow Stripe 1 (YS1) protein, shown to transport Fe3+-phytosiderophore (Curie et al. 2001) and Fe2+-nicotianamine (NA; Roberts et al. 2004; Schaaf et al. 2004) complexes. There are eight full-length Yellow Stripe-Like (AtYSLs) genes in the Arabidopsis genome (DiDonato et al. 2004), forming a cluster phylogenetically distinct from the AtOPTs. The OPT family of transporters, therefore, is larger and more diverse than was previously recognized. However, no OPT transporter has yet been identified in animals. Although OPT proteins were originally characterized as tetra- and pentapeptide transporters, it now appears that OPT transporters can transport more diverse substrates, all of which are peptide-containing compounds or peptide derivatives.

We previously reported on the identification and characterization of a loss of function mutation in the Arabidopsis AtOPT3 gene (Stacey et al. 2002b). We found that AtOPT3 was crucial for normal embryogenesis, as indicated by the embryo-lethal phenotype of plants homozygous for the opt3 mutant allele. Mutant embryos were arrested very early in development, mostly at the two- to eight-cell stage of embryogenesis. Consistent with the embryo-lethal phenotype associated with opt3, AtOPT3 expression was localized in the endosperm tissues, integuments and, most notably, in the embryos of developing seeds. To date, only two plant OPT members have been reported to show a mutant phenotype, AtOPT3 and the maize ZmYS1, where mutants exhibited a chlorotic phenotype due to a defect in iron uptake (von Wiren et al. 1994; Curie et al. 2001). The lack of observable phenotypes in loss-of-function mutants of other AtOPTs may be the result of functional redundancy among the family members. Knowledge of the expression pattern of individual genes can be helpful in designing multiple knock-out lines, and also in elucidating physiological function. Koh et al. (2002) previously reported on the tissue-specific expression of seven of the nine AtOPTs (AtOPT1 to AtOPT7) by semi-quantitative RT-PCR. We now expand on this effort by analyzing AtOPT promoter-GUS fusion expression in transgenic Arabidopsis. The results showed that the Arabidopsis AtOPTs (AtOPTs 1, 4, 6, 7 and 8) gave comparable expression to that previously reported for AtOPT3 (Stacey et al. 2002). All were preferentially expressed in the vascular tissues of vegetative and reproductive tissues, and none was expressed in root hairs and root tips. Additional unique expression patterns were also found for selected AtOPTs. For example, AtOPT1 was expressed in pollen tubes, AtOPT6 was expressed in ovules and AtOPT8 was expressed in the endosperm and integument of developing seeds. Iron limitation was found to uniquely enhance the expression of AtOPT3.

Materials and methods

Plant growth conditions and transformation

Seeds were surface-sterilized with a 35% bleach solution containing 0.1% Triton X-100 for 10 min, rinsed several times with sterile water, and cold-treated for 2–3 days. For routine growth, seedlings were grown on agar medium containing one-half-strength Murashige and Skoog (MS) salts (Sigma, St. Louis, MO, USA), MS vitamins (Sigma), 1% sucrose (w/v), 0.05% MES, and 25 μg/ml hygromycin. For seed amplification and analysis of mature plants, 10- to 15-day-old seedlings were transferred and grown on Premier Pro-Mix soil (Quakertown, PA, USA). Plants were grown at 22°C under constant fluorescent white light.

For growth in iron-sufficient and iron-limiting conditions, surface-sterilized, cold-treated seeds were germinated on agar medium containing Gamborg’s B-5 salts (Sigma), 1% sucrose (w/v), 0.05% MES, and 25 μg/ml hygromycin. Seedlings were grown until the four- to six-true leaf stage (12–14 days) and transferred either to agar medium containing salts as described by Masrchner et. al. (1982), 0.05% MES and 50 μM Fe(III) EDTA for iron-sufficient plates or 300 μM ferrozine (Sigma) for iron-deficient plates. Seedlings were analyzed after 5 days of growth in iron-sufficient or iron-deficient conditions. All seedlings were grown under yellow 0.118 in. plexiglass (Delvie’s Plastics, Inc, Salt Lake City, UT, USA) at 22°C and constant fluorescent white light. Stable transformation of Arabidopsis (ecotype Columbia, Col-O) was done following the vacuum infiltration procedure for Agrobacterium tumefaciens-mediated T-DNA gene transfer (Bechtold and Pelletier 1998).

Cloning of AtOPT promoter–GUS fusions

AtOPT promoter sequences were amplified from Arabidopsis thaliana ecotype Col-O genomic DNA and cloned upstream of the promoterless β-glucoronidase (GUS) gene encoded in the binary vector pCambia1391Z (Hajdukiewicz et al. 1994). Primers and promoter fragment size used for cloning each AtOPT promoter are as follows: AtOPT1, 5′-GAGATCTCGATCTTGTTGGTACAAGCGTGCT-3′ (forward primer) and 5′-GCCTAGGTGTCTTCAAGCTCCTTTTTTCCGT-3′, 3.2 kbp; AtOPT4, 5′-GGTCGACTGTGTTTCTGGAGAGCCATTGTAA-3′ (forward primer) and 5′-GCCTAGGTCCGGTTTAGTTTGTGTGTGAGTG-3′ (reverse primer), 2.1 kbp; AtOPT6, 5′-CTCAAGCTTAAAGGAGGAGCCGATGAAGTGATATTCTT-3′ (forward primer) and 5′-CTGCCTAGGGTTTTAGAAGAGTGAGTGTTCCGTTCTTT-3′ (reverse primer), 2.2 kbp; AtOPT7, 5′-CAGGATCCTGTATTAGCACTAACGAAATCTTT-3′ and 5′-CAGAATTCTTTGGGGAGATGTGAATCTGACAA-3′ (reverse primer), 2.2 kbp; AtOPT8, 5′-GAGATCTGCATACTACATTCTCCCTGGTTAT-3′ (forward primer) and 5′-GCCTAGGCGGTCTTCAGAAACAGAGGATTGG-3′ (reverse primer), 3.4 kbp. Cloning of the AtOPT3 promoter-GUS fusion was previously described (Stacey et al., 2002). Promoter fusions for AtOPT2 (3.5 kbp), AtOPT5 (3.4 kbp) and AtOPT9 (3.0 kbp) were also cloned but gave no or very weak GUS activity in planta.

Histochemical GUS staining and microscopy

Transgenic Arabidopsis plants expressing AtOPT promoter–GUS fusions were stained for GUS using 5-bromo-4-chloro-3-indoyl-β-D-glucoronide (X-Gluc; Gold Biotechnology, Inc., St. Louis, MO, USA) as substrate according to published protocols (Jefferson et al. 1987). Stained seedlings and tissues were fixed in FAA solution (50% ethanol, 5% acetic acid, 3.7% formaldehyde) and destained in 70% ethanol overnight. Additional overnight clearing with lactophenol (1 H2O:1 glycerol:1 lactate:2 phenol, v/v) was done for unfertilized gynoecia and young siliques to visualize internal parts such as ovules, pollen tubes and developing seeds. In certain cases, seeds were dissected from the ovary and cleared with lactophenol to visualize the embryos. Staining patterns were observed and documented using a Nikon microscope (SMZ1500) equipped with a Nikon digital camera (DXM1200). For each AtOPT promoter-GUS fusion, at least 25 independent T1 lines were examined for GUS activity in leaves. Transgenic lines expressing a GUS fusion to AtOPT1, AtOPT4, AtOPT6, AtOPT7 or AtOPT8 gave similar leaf-expression patterns except for differences in expression levels, which ranged from very weak to strong. Three to four lines showing strong expression were selected for each AtOPT promoter fusion and corresponding T2 progenies were analyzed at various stages of development, namely, during germination, at the two–four seedling stage, and adult plants within 1 week of bolting. The expression of each AtOPT–GUS fusion was also analyzed at various stages of flower development as described by Smyth et al. (1990) and, subsequently, in developing seeds. Transgenic lines expressing a GUS fusion to AtOPT2, AtOPT5 or AtOPT9 showed little to no detectable GUS staining in leaves of T1 plants and were not analyzed further.

Quantitative GUS activity assays

GUS activity was measured fluorometrically for intact tissues using 4-methylumbellyferyl-β-D-glucoronide (4-MUG; Gold Biotechnology, Inc., St Louis, MO, USA) as substrate, as described by Jefferson et al. (1987). Fluorescence was measured in a spectrofluorometer equipped with a microplate reader. For each AtOPT promoter–GUS fusion, at least three independent lines were analyzed. Fluorometric assays were done in three trials, with two replicates per trial. Four to five transgenic seedlings were assayed per replicate. GUS activity for each sample was normalized per unit seedling wet weight.

Cloning of AtOPT3 cDNA and transport assays in S. cerevisiae and Xenopus oocytes

To clone the AtOPT3 cDNA, total RNA isolated from Col-0 plants was used as a template for reverse-transcriptase-based (RT) PCR using random hexamer primers. First-strand cDNA was amplified with Pfu DNA polymerase (Stratagene, La Jolla, CA, USA) using primers 5′-GGCCAACAACGCAACTTTGTCTGGTACTTCA-3′ (forward primer) and 5′-CTGCTGAGCTTAGAAAACGGGACAGCC-3′ (reverse primer). Amplified PCR products were cloned into pGEM T Easy cloning vector (Promega, Madison, WI, USA) and sequenced. For functional expression in S. cerevisiae or Xenopus oocytes, the full-length AtOPT3 cDNA clone was digested with NotI and subcloned into the yeast expression vector pDB20 (Becker et al. 1991) or the oocytes expression vector pGEMHE (Liman et al. 1992), respectively. An artificial Not I site in pGEMHE was created by swapping EcoRI–XbaI sites with EcoRI- and XbaI-digested multicloning sites of pCR 2.1 TOPO cloning vector (Invitrogen, Carlsbad, CA, USA). Transformation of S. cerevisiae BY4730 (MATα leuΔ0 met15Δ0 ura3Δ0) was done following the LiOAc method. Transformed cells were selected by growth on 0.17% (w/v) YNB plates supplemented with 2% (w/v) glucose lacking uracil. Yeast growth assays in the presence of 100 μM KLGL, KLLLG, YGGFL or GSH were performed according to previously described protocol (Koh et al. 2002). AtOPT3 and ScOPT1 (Hauser et al. 2000) expression in Xenopus oocytes was performed as previously described (Rubio et al. 1995). Oocytes, 4 days after injection with 23 ng of AtOPT3 or ScOPT1 cRNA, were voltage-clamped in a bath solution containing 1 mM KCl, 6 mM MgCl2, 1.8 mM CaCl2, 10 mM Mes–tris (pH 5.0), and 1 mM GSH. Osmolality of the bath solution was adjusted to 240 mosmol kg−1 by the addition of D-mannitol.

Results

Expression of Arabidopsis OPTs during germination and early seedling growth

To examine the possible role of AtOPTs in storage protein mobilization during germination, we analyzed AtOPT promoter–GUS fusion activity in germinating seeds prior to radicle protrusion (germination stage) and in newly germinated seedlings (post-germination stage). AtOPT1 was expressed at a low but detectable level in the cotyledons of germinating seedlings (Fig.1a, b), as well as in the cotyledons and prevascular tissues of newly germinated seedlings (Fig. 1c). Similar expression patterns were observed for AtOPT3 and AtOPT4 at the germination (data not shown) and post-germination stages. However, stronger AtOPT3 expression was observed in the prevascular bundles of hypocotyls of postgerminative seedlings (Fig. 1d) when compared to AtOPT4 (Fig. 1e). Likewise, AtOPT6 (Fig. 1f, g) and AtOPT7 (Fig. 1i) were also expressed in cotyledons of germinating seeds but at a much higher level than AtOPT1. Relatively higher expression levels of AtOPT6 (Fig. 1h) and AtOPT7 (Fig. 1j) were also observed in the cotyledons and prevascular tissues of hypocotyls of postgerminative seedlings. In contrast, AtOPT8 was poorly expressed in germinating seeds and seedlings (Fig. 1k), showing detectable GUS staining only after prolonged staining (Fig. 1l). For all AtOPT tested, no detectable expression was observed in young roots or root hairs (Fig. 1c–e, h, j–l). In summary, with the exception of AtOPT8, all the AtOPTs examined showed significant expression in cotyledons of pregerminative seeds and newly germinated seedlings. Moreover, although differences in expression levels were observed among the various AtOPTs in postgerminative seedlings, similar expression patterns were observed, namely, predominant expression in prevascular tissues of cotyledons and hypocotyls and lack of expression in radicles and root hairs.

Fig. 1
figure 1

a-l Histochemical localization of AtOPT promoter–GUS expression during Arabidopsis seed germination (a, b, f) and post-germinative seedling growth (c–e, g, h–l) Representative GUS staining patterns of transgenic Arabidopsis expressing AtOPT1 (a–c), AtOPT3 (d), AtOPT4 (e), AtOPT6 (f–h), AtOPT7 (i, j) and AtOPT8 (k, l) fusions are shown. Bars= 60 μm (a, b, f, g, i), 100 μm (c, d, e, h, j–l)

Full size image

Expression of Arabidopsis OPTs during vegetative growth

In young seedlings (at the two to four true-leaf stage), strong AtOPT1 expression was observed in the vascular tissues of cotyledons (Fig. 2a), hypocotyls (Fig. 2b) rosette leaves (Fig. 2a, c) and roots (Fig. 2d). AtOPT1 was also expressed in trichomes (Fig. 2c), but no detectable expression was observed in root hairs (Fig. 2d) and in root tips (Fig. 2e). Similar expression patterns were observed in cotyledons, rosette leaves and hypocotyl for AtOPT4 (data not shown), AtOPT6 (data not shown) and AtOPT7 (Fig. 2f), although the latter was expressed in all cells of the hypocotyl (Fig. 2g). AtOPT8 was also expressed strongly in vascular tissues of cotyledons (Fig. 2h) and hypocotyl (Fig. 2i). However, unlike the other AtOPTs, AtOPT8 was predominantly expressed in primary veins of rosette leaves (Fig. 2h). Comparable root expression to that of AtOPT1 was observed for AtOPT4, AtOPT6, AtOPT7 and AtOPT8, namely, high expression in the vascular cylinder but no detectable expression in root hairs and root tips (data not shown). AtOPT6 expression was observed in emerging lateral roots for AtOPT6 but not for the other AtOPTs (data not shown).

Fig. 2
figure 2

a–o Histochemical localization of AtOPT promoter-GUS expression during vegetative growth and in inflorescence stems of transgenic Arabidopsis. a–e Representative expression patterns of AtOPT1 expression in whole seedling (a), hypocotyl (b), trichomes and vascular tissues of rosette leaf (c), root (d), and root tip (e). f, g Representative expression patterns of AtOPT7 expression in whole seedling (f) and hypocotyl (g). Note darker GUS staining in older leaves in (a) and (f). h, i Representative expression patterns of AtOPT8 expression in whole seedling (h) and hypocotyl (i) and in rosette leaf (j) and leaf petiole (k) of 25-day-old soil-grown plant. Note strong GUS staining only in primary veins of rosette leaves in (h). l–o Representative expression patterns of promoter–GUS fusions observed for AtOPT1 (l), AtOPT4 (m, n) and AtOPT7 (o) in inflorescence stems. Bars=2 mm (a, f, h, j, k, l, m), 150 μm (b, c, d, e, g, i), 300 μm (n, o)

Full size image

In adult plants, all the AtOPTs showed predominant expression in vascular tissues of cauline (data not shown) and rosette leaves, as well as rosette leaf petioles. Representative GUS staining patterns in rosette leaves and leaf petioles expressing AtOPT8 promoter–GUS are shown in Fig. 2j, and k, respectively. In inflorescence stems, very weak expression was detected for AtOPT1 (Fig. 2l) and AtOPT8 (data not shown), whereas strong expression, predominantly in vascular tissues, was observed for AtOPT4 (Fig. 2m, n), AtOPT6 (data not shown) and AtOPT7 (Fig. 2o). Expression of AtOPT3 in inflorescence stems was not examined.

Expression of Arabidopsis OPTs during flower development and fertilization

No detectable AtOPT1 expression was observed in immature flowers (up to stage 11 of flower development; Fig. 3a). Also, no detectable expression was observed for AtOPT3, AtOPT4, AtOPT6, AtOPT7 and AtOPT8 during early stages of flower development (data not shown). At stage 12, detectable expression of AtOPT1 was observed in sepals (Fig. 3b), predominantly in vascular tissues, and subsequently in anthers and pollen grains (Fig. 3c, d). We previously reported that AtOPT3 was expressed in pollen grains (Stacey et al. 2002). Further examination of AtOPT3 expression showed that it was expressed in pollen grains and anthers starting at stage 12 of flower development (data not shown). AtOPT1 expression in stigma and in vascular tissues of the style (Fig. 3e) and petals (data not shown) was observed in only a portion (approximately 50%) of flowers examined. No expression in ovules or any part of the ovary was observed prior to anthesis (Fig. 3e). However, immediately following anthesis, AtOPT1 was expressed along the central septum of the ovary (Fig. 3f).

Fig. 3
figure 3

a–l Histochemical localization of AtOPT promoter–GUS expression during Arabidopsis flower development and fertilization. a–e Representative expression patterns of AtOPT1 promoter–GUS expression in immature flowers and floral buds (a) and in unfertilized flowers at stage 12 of floral development (b–e). At stage 12, AtOPT1 was expressed in pollen grains (c, d) but not in ovules (e). f AtOPT1 expression in fertilized flower at stage 15. Arrow indicates GUS staining along the septum of the ovary. g–l Representative expression patterns of promoter–GUS fusions observed for AtOPT4 (g, h), AtOPT6 (i, j) and AtOPT8 (k, l) in unfertilized (g, i, k) and fertilized (h, j, l) flowers. Arrows in (i) and (j) indicate AtOPT6 expression in ovules and vascular tissue of ovary, respectively. Arrows in (l) indicate AtOPT8 expression in ovules immediately following fertilization. Unfertilized flowers were at stage 12 and fertilized flowers were at stages 14–15. The stages of flower development were as described by Smyth et al. (1990). Bars=300 μm (a, f, h, j, l), 150 μm (b, c, g, i, k), 100 μm (d, e)

Full size image

AtOPT4 (Fig. 3g, h) and AtOPT7 (data not shown) gave comparable expression patterns in unfertilized and fertilized flowers. Both showed relatively lower expression in flowers when compared to AtOPT1, notably weaker expression in sepals (Fig. 3g) and no detectable expression in pollen grains, anthers and ovary (Fig. 3g, h). Like AtOPT1, AtOPT4 and AtOPT7 were expressed in stigma and vascular tissues of style in only a portion of flowers examined. This lack of detectable expression in a portion of flowers could be due to lower expression levels in this subset of flowers and/or possibly to less permeation of the substrate during staining.

AtOPT6 was strongly expressed in vascular tissues of sepals, petals, style, filaments, and ovary but no expression was detected in anthers and pollen grains (Fig.3i, j). The high expression of AtOPT6 in unfertilized gynoecium was unique among the AtOPTs examined, especially expression in ovules (Fig. 3i). Following fertilization, a decrease in overall AtOPT6 expression was observed in the gyneocium, although prominent expression in vascular tissues of the ovary and style was apparent (Fig 3j).

Similar to AtOPT1 and AtOPT3, AtOPT8 was strongly expressed in anthers and pollen grains starting at stage 12 of flower development (Fig. 3k). AtOPT8 gave no detectable expression in ovules. However, strong AtOPT8 expression was observed in ovules immediately following fertilization (Fig. 3l). AtOPT8 gave no detectable expression in other floral organs, although weak GUS staining was observed in vascular tissues of sepals and filaments (data not shown).

In summary, notable differences in tissue-specific expression were observed for the AtOPTs in flowers before and after fertilization. AtOPT1, AtOPT3 and AtOPT8 were strongly expressed in anthers and pollen grains, whereas AtOPT4, AtOPT6 and AtOPT7 were not. In the unfertilized gynoecium, only AtOPT6 was expressed in ovules and very strongly in vascular tissues. Only AtOPT1 was expressed along the septum of ovary, while only AtOPT8 was expressed in fertilized ovules immediately following anthesis.

AtOPT1 is expressed in pollen tubes

The expression of AtOPT1 along the septum of the ovary may be due to its expression in the maternal transmitting tract tissue or in the growing pollen tubes. To address this issue, stained pistils expressing the AtOPT1 promoter–GUS fusion were cleared with lactophenol and examined in more detail (Fig. 4a–c). AtOPT1–GUS staining in the stigma and style appeared to be consistent with pollen tube expression as indicated by GUS-stained strands starting at stigmatic papillae and through the central portion of the style (Fig. 4b). The AtOPT1 expression pattern in the ovary was also consistent with pollen tube expression, being expressed in the central septum and in GUS-stained strands exiting the transmitting tract, which can be presumed to be pollen tubes penetrating ovules (Fig. 4c). Reciprocal crosses between Col-O and plants expressing an AtOPT1 promoter–GUS fusion were performed to confirm that AtOPT1 was indeed expressed in pollen tubes. When AtOPT1 promoter–GUS plants were used as pollen donor, GUS staining was observed in the stigma and in the central septum of the ovary (Fig. 4d). When Col-O plants were used as pollen donor, no staining was observed in the stigma and the central septum (Fig. 4e). Control crosses, where AtOPT1 promoter–GUS plants were manually self-fertilized, showed the expected GUS staining in stigma and central septum (Fig. 4f). Therefore, AtOPT1 was expressed in pollen tubes and not in transmitting tract tissues during pollination.

Fig. 4
figure 4

a–c Expression of AtOPT1 promoter-GUS in pollen tubes as indicated by GUS staining patterns in cleared silique (a), stigma and style (b) and ovary (c). Arrow in (c) indicates stained pollen tube exiting the transmitting tract tissue. d GUS-stained Col-O siliques fertilized with pollen grains from plants expressing AtOPT1 promoter–GUS. e GUS-stained siliques from plants expressing AtOPT1 promoter–GUS fertilized with Col-O pollen grains. f GUS-stained siliques from manually selfed plants expressing OPT1 promoter-GUS. Scale bars=500 μm (a, d–f), 60 μm (b, c)

Full size image

Expression of Arabidopsis OPTs during seed development

The developing seed is a major sink tissue that requires nutrients from other parts of the plant. AtOPT expression patterns were observed during various stages of seed development, namely, following anthesis until seed desiccation, to determine if peptide transport through AtOPTs plays a role in seed development. Only AtOPT8 was expressed in fertilized ovules immediately following anthesis and, therefore, AtOPT8 expression was examined in more detail. AtOPT8–GUS staining was seen in cleared siliques and excised seeds (Fig. 5a–e). AtOPT8 was expressed in endosperm and integument tissues of seeds at the preglobular (Fig. 5a, b) and globular stages (Fig. 5c) of seed development. By early heart stage, no AtOPT8 expression in the endosperm and integument was observed, but rather AtOPT8 expression was only observed in the funiculus (Fig. 5d). No AtOPT8 expression was observed in embryos in all stages of seed development examined. Under similar staining conditions as those used for AtOPT8, no detectable seed expression was observed for AtOPT7 during early embryogenesis (Fig. 5f–h). No detectable expression was also observed for AtOPT1, AtOPT4 and AtOPT6 during early embryogenesis (data not shown). However, like AtOPT8, these AtOPTs were strongly expressed in the vascular tissues of funiculi starting at around the onset of heart stage of embryogenesis, as shown in Fig. 5i for AtOPT7. All the AtOPTs were also expressed in the vascular tissues of siliques, with stronger expression observed starting at the heart stage until the curled cotyledon stage of embryogenesis, as shown in Fig. 5e and j for AtOPT8 and AtOPT7, respectively. Lastly, as we previously reported, AtOPT3 was expressed in developing seeds and embryos (Stacey et al. 2002b). Further examination of developing seeds showed that increased AtOPT3 expression in vascular tissues of funiculi and siliques also occurred around the onset of the heart stage of embryogenesis (data not shown).

Fig. 5
figure 5

a–j Histochemical localization of AtOPT8 and AtOPT7 promoter–GUS expression during seed development. Whole siliques or excised seeds were stained for GUS and subsequently cleared to visualize AtOPT8 (a–e) or AtOPT7 (f–j) expression at various stages of seed development. GUS-staining patterns were observed in siliques immediately following anthesis (a, f) and in seeds at the preglobular (b), globular (c, g, h), heart (d, i), and torpedo (e, j) stages of seed development. Scale bars=300 μm (a, e, f, g), 60 μm (b–d, g–i)

Full size image

Corrected annotation for the AtOPT3 genomic sequence

We recently described an embryo lethal phenotype for opt3 plants (Stacey et. al. 2002b). AtOPT3 showed the highest sequence similarity among the nine AtOPTs to BjGT1, a recently characterized glutathione (GSH) transporter in Brassica juncea (Bogs et al. 2003). Bogs et al. (2003) also suggested that the predicted protein derived from the annotated AtOPT3 gene (At4g16370) displays an erroneous C-terminal sequence, and proposed a sequencing error in the At4g16370 genome sequence. This conclusion was supported by comparison of BjGT1 and AtOPT3 with a full-length Col-O cDNA sequence (AY054590) derived from the AtOPT3 gene (Bogs et al. 2003). As a first step for further studies on possible functions of AtOPT3, we re-isolated an AtOPT3 cDNA from Col-O. Our cDNA sequence was identical to that of the AY054790 nucleotide sequence except for a single nucleotide difference at position 133 (A in AtOPT3 and G in AY054790), translating into an amino acid difference at position 45 (T in AtOPT3 and A in AY054790). However, our cDNA sequence corresponded to the nucleotide and predicted amino acid sequence to the annotated At4g16370 at this position and to the amino acid sequence of BjGT1 (Bogs et al. 2003). Our cDNA sequence also supported the previous suggestion of a sequencing error at the 3′ end of At4g16370. An additional G in At4g16370 at position 12378 causes a frameshift in the annotated amino acid sequence and a premature stop codon 93 base pairs upstream of the actual stop codon.

Previous data showed that S. cerevisiae expressing AtOPT3 was unable to mediate the uptake of glutathione ([H3 ]-GSH) or of several peptides that were obtained with the truncated Ler AtOPT3 cDNA based on the annotated At4g16370 sequence (Koh et al. 2002). Therefore, we re-evaluated the glutathione (GSH) and peptide transport capacities of AtOPT3 by expressing the full-length Col-O cDNA in Xenopus laevis oocytes or S. cerevisiae. As control, we also used ScOPT1p (Hauser et al. 2000), previously shown to be a GSH transporter in S. cerevisiae (Bourbouloux et al. 2000). Oocytes expressing ScOPT1 exhibited GSH-evoked inward current when substrate was added in the bath solution at pH 5.0. However, oocytes expressing AtOPT3 did not show measurable changes in the presence of GSH in the bath solution (data not shown). Likewise, no evidence for GSH or peptide (YGGFL, KLGL and KLLLG) uptake was found in yeast cells expressing AtOPT3 (data not shown). It is possible that AtOPT3 functions in peptide efflux rather than uptake. However, transporters in this class are generally reversible. The direction of substrate movement would then depend on the electrochemical gradient, which in uptake experiments favors substrate influx. Additional experiments are required to establish true substrates and the transport mechanism of AtOPT3.

Expression of AtOPT3 is induced by limiting iron

Gene expression of other OPT family members, for example, the maize YS1 and the Arabidopsis YS-like proteins (YSLs), was shown to be regulated by iron availability in the growth medium (Curie et al. 2001; DiDonato et al. 2004). Therefore, the expression of AtOPT-GUS fusions was also examined in seedlings grown under iron-sufficient and iron-deficient conditions. As shown in Fig. 6a, AtOPT3 expression was induced two- to fourfold under iron-deficient conditions. AtOPT3 expression was increased in rosette leaves and root tissues under limiting iron (Fig. 6b). Increased AtOPT3 expression in leaves was observed mostly in younger leaves (sixth to tenth leaf) and was not limited to vascular tissues. In roots, increased AtOPT3 expression under iron-deficient conditions was observed predominantly in vascular tissues. No detectable AtOPT3 expression was observed in root hairs and root tips under both iron-sufficient or iron-deficient conditions (Fig. 6b). Histochemical staining showed no effect on the expression of AtOPT1, 4, 6, 7, and 8 under limiting-iron availability (data not shown) and this was confirmed by GUS assays as shown in Fig. 6a.

Fig. 6
figure 6

a Quantitative GUS activity assays of transgenic Arabidopsis seedlings expressing various AtOPT promoter–GUS fusions using MUG as substrate. GUS activity in T2 whole seedlings was measured after growing plants for 5 days in iron-sufficient (50 μM Fe(III)–EDTA) or iron-deficient medium. For each AtOPT promoter–GUS fusion, three or four independent lines were assayed (indicated by numbers on x-axis). Values represent average relative ratio of GUS activity in iron-deficient plants to the GUS activity of iron-sufficient plants (fold induction). For each transgenic line, GUS activity was measured in six to eight replicates performed over three trials. Standard error bars are shown. b GUS-stained plants expressing AtOPT3 promoter-GUS grown for five days in iron-sufficient (+FE) or iron-deficient (−Fe) medium. Scale bar=2 mm (b, upper panels), 0.5 mm (b, lower panels)

Full size image

Examination of the abiotic stress microarray dataset from the AtGenExpress project (http://www.weigelworld.org/resources/microarray/AtGenExpress) showed no induction of AtOPT expression under the various stress conditions examined, except for a significant induction of AtOPT2 under salt stress. We found no effect of salt stress (150 mM NaCl) on the expression AtOPTs, including AtOPT2 data not shown. It is possible that the AtOPT2 promoter fusion that we used lacked important regulatory regions responsive to salt stress.

Discussion

AtOPTs may function in nitrogen mobilization during seed germination and senescence

One proposed function for peptide transport in plants is the mobilization of peptides derived from proteolyzed storage proteins during germination (Higgins and Payne 1981, 1982; reviewed by Waterworth et al. 2001). For example, physiological and biochemical analyses of storage protein degradation in germinating cereal seeds showed that storage proteins in the endosperm are hydrolyzed to form a reservoir of small peptides (two to six residues) and amino acids, which are subsequently transported across the scutellum to the growing embryo (Higgins and Payne 1981). To examine if the AtOPTs are involved in nitrogen mobilization during germination and subsequent early seedling growth, the expression of AtOPTs was examined to see if it coincided temporally and spatially with storage protein degradation during these early stages of plant growth. Germination is regarded to be complete when the radicle breaks through the seed coat after which the onset of seedling growth begins (Bewley 1997). Proteolysis of storage proteins prior to radicle emergence was reported in barley (Higgins and Payne 1981), vetch, garden beans and rape seeds (Schlereth et al. 2001; Tiedemann et al., 2000; reviewed by Müntz et al. 2001). Consistent with this timing, AtOPT1, AtOPT3, AtOPT6 and AtOPT7 showed detectable GUS staining in embryonic cotyledons prior to radicle emergence. Likewise, the prominent expression of AtOPTs in prevascular tissues of germinating seeds, and subsequently in vascular tissues of postgerminative seedlings, is of significance with regard to their potential role in nitrogen mobilization during the germination process. In germinating vetch and garden bean seeds, the prevascular strands are among the first tissues where protein mobilization is initiated during germination (Schlereth et al. 2001; Tiedemann et al. 2000). Local proteolysis is initiated in these tissues presumably to support growth and differentiation prior to the formation of functional vascular bundles (Schlereth et al. 2001; Tiedemann et al. 2000). Bulk storage protein degradation in the cotyledons, where most stored nutrients are located, occurs during the postgerminative period (Müntz et al. 2001). Thus, in germinating Arabidopsis seeds, most of the AtOPTs are likely involved in the transport of peptides derived from storage protein degradation in cells undergoing growth and differentiation and, subsequently, in the mobilization of peptides from the cotyledons to the embryonic axis during postgerminative growth. It is curious, however, that no detectable AtOPT expression was seen in root tissues, which also have protein reserves that are mobilized during germination (Tiedemann et al. 2000).

Senescence in plants is characterized by an ordered degradation of cell constituents and functions mainly for mobilization and recycling of nutrients (reviewed by Hörtensteiner and Feller 2002; Buchanan-Wollaston et al. 2003). Significant levels of peptides were reported in phloem and xylem exudates and peptide transport was suggested to be important for bulk movement and redistribution of organic nitrogen during senescence (Higgins and Payne 1980; Higgins and Payne 1982). Ribulose-1,5-bisphosphate carboxylase (Rubisco) accounts for at least 50% of total soluble protein in a fully expanded leaf and was suggested to be an important nitrogen storage component that could be rapidly mobilized during leaf senescence (Mae et al. 1983). Arabidopsis leaves have a very short lifetime. Senescence starts as soon as full expansion is reached, as indicated by diminished levels of Rubisco large subunit as soon as leaf expansion is complete (Stessmann et al. 2002; Buchanan-Wollaston et al. 2003). The fact that none of the AtOPTs were expressed in the root epidermis or root hairs, argues against a role for these transporters in peptide acquisition from the growth medium. Rather, their expression patterns support the idea that AtOPTs mediate the long-distance mobilization of endogenous peptides, possibly derived from proteolysis in senescing older leaves. This hypothesis is consistent with the prominent expression of AtOPTs in the vascular tissue of source tissues, especially in cotyledons and in older leaves.

In addition to the AtOPTs, peptide transporters belonging to the PTR family were also proposed to be important for the mobilization of nitrogen during germination and in the long-distance transport of peptides to growing tissues. These include the barley HvPTR1 (West et al. 1998; Waterworth et al. 2000), the Faba bean VfPTR1 and VfPTR2 (Miranda et al. 2003) and the Arabidopsis AtPTR1 (Dietrich et al. 2004).

Differential expression of AtOPTs during flower and seed development indicates distinct roles for specific AtOPTs

Unlike vegetative growth where only minor differences in tissue-specific expression among the various AtOPTs were observed, major differences in AtOPT expression were observed during flower and seed development (Table 1). AtOPT1, 3 and 8 were strongly expressed in anthers and pollen grains, whereas AtOPT4,6 and 7 were not. AtOPT1, 3 and 8 were expressed late in pollen development, starting at stage 12 of anther development right before anther dehiscence and the opening of floral buds (Smyth et al. 1990; Sanders et al. 1999). These data appear to rule out an important role for these AtOPTs in the early events of anther development. Rather, the timing of AtOPT expression suggests a role in anther dehiscence and/or during pollination. AtOPT1 and AtOPT6 are the only peptide transporters shown so far to be expressed in pollen tubes and ovules, respectively. AtOPT1 is likely involved in the transport of peptides from the surrounding maternal transmitting tract tissues for use in nourishing pollen tube growth. Likewise, AtOPT6 may also have a nutritional function during ovule development.

Table 1 Tissue-specific expression of AtOPT promoter–GUS fusions during flower and seed development
Full size table

Transgenic plants expressing promoter-GUS fusions to AtOPT2, AtOPT5 and AtOPT9 showed little to no expression (data not shown) and hence, detailed expression analyses for these genes were not pursued. Massively parallel signature sequencing (MPSS) data also indicated very low expression levels for AtOPT2 and AtOPT9. However, examination of the developmental microarray dataset (Schmid et al. 2005) showed high expression of AtOPT2 and AtOPT9 in young flowers (at stages 9 and 10 of flower development), as well as in stamen and pollen grains. AtOPT5 also showed significant expression in floral organs of fertilized and unfertilized flowers. Unlike AtOPT2 and AtOPT9, however, AtOPT5 showed a significant expression in developing seeds, as well as in vegetative tissues of seedlings and adult plants. The specificity of AtOPT2 and AtOPT9 expression in flowers could indicate an important function in flower development. Although, in general, our GUS expression data fit well with the AtGenExpress data (Schmid et al. 2005), the use of the promoter–GUS fusions allowed a more detailed examination, involving a wider variety of conditions and developmental stages, than practical with DNA microarray technology.

Immediately following fertilization up to the globular stage, AtOPT8 is expressed in integument and endosperm tissues of developing seeds. At around the start of heart stage, AtOPT8 expression was observed only in seed funiculi. A similar expression pattern was observed for AtOPT3 in developing seeds, except that, unlike AtOPT8, it was expressed in developing embryos (Stacey et al. 2002b). This pattern of expression indicates that AtOPT3 and AtOPT8 are likely involved in the apoplastic transport of peptides between integument cells and the symplastically isolated endosperm during early embryogenesis. At the heart stage of embryogenesis, the start of endosperm cellularization, rapid embryo growth and deposition of storage proteins in embryonic cotyledons create a high demand for nitrogen assimilates. Moreover, at the heart stage of embryo development, a specialized maternal tissue, the chalazal proliferative tissue, is formed in close proximity to the vascular tissue of the funiculus and was proposed to facilitate solute transfer across the plasmalemma between the maternal plant and the apoplastic endosperm (Olsen 2004; Brown et al. 1999). The timing of AtOPT3 and AtOPT8 expression, as well as that of other AtOPTs, in the vascular tissues of funiculi starting at around the onset of the heart stage of embryogenesis is highly suggestive of their functional role in the vascular transport of nitrogen assimilates from the maternal plant to the developing seeds. Vascular transport of peptides and other assimilates during later stages of embryogenesis could be a more efficient way to meet higher nitrogen demand during these stages. Consistent with this notion, the Arabidopsis amino acid (AAP2) transporter was also strongly induced at the heart stage of embryo development in the vascular strands of siliques and funiculi (Hirner et al. 1998). It is very likely that AtOPT5 is also involved in nitrogen mobilization during seed development, as indicated by its relatively high expression in young siliques and developing seeds (AtGenExpress developmental microarray data; Schmid et al. 2005).

Non-protein-derived peptides as substrates of AtOPTs

The OPT family includes the maize Yellow-Stripe1 (YS1) protein involved in the uptake of Fe3+ when complexed with mugeneic acids (MAs) (Yen et al. 2001; Curie et al. 2001), peptide-like compounds synthesized from three molecules of S-adenosyl methionine. Arabidopsis and other dicots do not produce MAs but synthesize a structurally similar compound, nicotianamine (NA), the biosynthetic precursor of MAs. YS1-like proteins (YSLs) have been characterized in Arabidopsis (AtYSL2; DiDonato et al. 2004) and rice (OsYSL2; Koike et al. 2004). AtYSL2 and OsYSL2 were shown to transport metal–NA complexes and their expression was influenced by the availability of iron in the growth medium. On the basis of their in planta expression patterns, both proteins were proposed to transport internal metal–NA complexes, rather than metal acquisition from the soil, as is the case with ZmYS1. The YSL subfamily of OPT transporters is phylogenetically distinct from the AtOPT transporters (Yen et al. 2001). However, Wintz et al. (2003) showed that AtOPT3 expression was also influenced by the availability of iron in the growth medium and that AtOPT3 mediated the transport of copper, manganese and iron when expressed in yeast. Consistent with these earlier findings, we found that AtOPT3 promoter–GUS expression was inducible by iron limitation. Limiting iron conditions had no effect on the expression of AtOPT1, 4, 6, 7 and 8. The preferential expression of AtOPT3 in vascular tissues of major plant organs and its lack of expression in root hairs suggest that AtOPT3, like the Arabidopsis and rice YSLs, is likely involved in internal metal–chelate transport rather than in metal acquisition from the external environment. If this hypothesis is correct, then the expression of AtOPT3 in the embryo and the embryo-lethal phenotype of an opt3 mutation argue for a major role for AtOPT3-mediated metal transport or efflux of metal chelators during embryo development. Other members of the OPT family also mediate the uptake of non-protein-derived peptides such as glutathione (GSH and GSSG) and glutathione derivatives (Bourbouloux et al. 2000; Bogs et al. 2003; Zhang et al. 2004; Cagnac et al. 2004). Expression of AtOPT6, but not AtOPT7, has recently been shown to be inducible by ABA and the sulfonylurea herbicide primisulfuron (Cagnac et al. 2004). It has been suggested that this transporter may be involved in detoxification and/or stress resistance, including metal stress. It remains to be tested if the expression of other AtOPTs is also affected by primisulfuron.

In summary, careful analysis of the expression patterns of AtOPT promoter–GUS fusions during various stages of Arabidopsis growth and development has led to testable hypotheses regarding transporter function. In general, AtOPT expression in vascular tissues, especially in source tissues such as cotyledons and older leaves, suggests a role in long distance movement of peptides resulting from proteolytic breakdown of storage protein. This role is also reflected in the germinating seed where AtOPTs likely mobilize protein reserves for new growth. The lack of AtOPT expression in the root epidermis and root hairs argues against a role in mediating peptide uptake from the medium. AtOPT1 expression in pollen tubes suggests a role in transport of peptides from the maternal tissue during fertilization. AtOPT3 and AtOPT8 showed a developmentally regulated pattern of expression in developing seeds. The data suggest a role for these transporters in aiding embryo nutrition from endosperm reserves until the heart stage. Subsequently, AtOPT expression switches to the vascular tissue of the funiculus, concomitant with the development of the funiculus-mediated pathway for maternal nutrition of the developing seed. Finally, although AtOPT3 is divergent in sequence from members of the YSL subfamily of OPT transporters, its expression is enhanced under limiting iron and studies in yeast suggest a role in metal uptake. Therefore, this transporter is likely involved in long-distance movement of metal–chelate complexes in the plant. This role appears especially critical, given the embryo-lethal phenotype of an opt3 mutant.