AtCAT6, a sink-tissue-localized transporter for essential amino acids in Arabidopsis


  • Ulrich Z. Hammes,

    Corresponding author
      *(fax +1 314 587 1521; e-mail or fax +49 9131 852 8751; e-mail †Present address: Molecular Plant Physiology, Friedrich-Alexander University Erlangen-Nürnberg, Staudtstrasse 5, 91058 Erlangen, Germany.
    Search for more papers by this author
  • Erik Nielsen,

    1. Donald Danforth Plant Science Center, 975 North Warson Road, St Louis, MO 63122, USA
    Search for more papers by this author
  • Loren A. Honaas,

    1. Donald Danforth Plant Science Center, 975 North Warson Road, St Louis, MO 63122, USA
    Search for more papers by this author
  • Christopher G. Taylor,

    1. Donald Danforth Plant Science Center, 975 North Warson Road, St Louis, MO 63122, USA
    Search for more papers by this author
  • Daniel P. Schachtman

    Corresponding author
      *(fax +1 314 587 1521; e-mail or fax +49 9131 852 8751; e-mail
    Search for more papers by this author

*(fax +1 314 587 1521; e-mail or fax +49 9131 852 8751; e-mail

†Present address: Molecular Plant Physiology, Friedrich-Alexander University Erlangen-Nürnberg, Staudtstrasse 5, 91058 Erlangen, Germany.


Amino acids represent the major form of reduced nitrogen that is transported in plants. Amino acid transporters in plants often show tissue-specific expression patterns and are used by plants to transport these metabolites from source to sink during development and under changing environmental conditions. We identified one amino acid transporter, AtCAT6, which is expressed in sink tissues such as lateral root primordia, flowers and seeds. Additionally AtCAT6 was induced during infestation of roots by the plant-parasitic root-knot nematode, Meloidogyne incognita. Quantitative reverse-transcriptase PCR revealed nematode inducibility throughout the duration of nematode infestation and in nematode-induced feeding sites. Promoter analyses confirmed expression in endogenous sink tissues and nematode-induced feeding sites. In Xenopus oocytes, AtCAT6 mediated electrogenic transport of proteinogenic as well as non-proteinogenic amino acids with moderate affinity. AtCAT6 transported large, neutral and cationic amino acids in preference to other amino acids. Knockout mutants of this transporter failed to grow on medium containing l-glutamine as the sole nitrogen source. Our data suggest that AtCAT6 plays a role in supplying amino acids to sink tissues of plants and nematode-induced feeding structures.


Amino acids represent the major form of transported organic, reduced nitrogen in most plant species. All proteinogenic amino acids are found in both phloem and xylem with concentrations being higher in the phloem. The amino acids aspartate, glutamate and their respective amides asparagine and glutamine are the predominant amino acids in phloem and xylem (Coruzzi, 2003). Amino acids passively follow the phloem bulk flow, the direction of which is dictated by the major osmolyte, generally sucrose (Münch, 1930). Therefore sink tissues must efficiently take up amino acids, especially those which are less abundant, to meet the demands of tissue.

Amino acid concentrations in the vasculature as well as in specific tissues are dynamic and change in response to abiotic stimuli such as light or drought and in response to seasonal changes (Cyr et al., 1990; Lam et al., 1995; Newton et al., 1986; Salleh and Owen, 1984). Amino acid concentrations also vary in response to pathogens and parasites such as aphids and nematodes (Krauthausen and Wyss, 1982; Salleh and Owen, 1984; Sandstrom et al., 2000). Sedentary plant-parasitic nematodes, such as cyst and root-knot nematodes induce highly specialized feeding sites in their host plants. These nematodes derive their supply of essential amino acids by withdrawing them from their host through highly specialized feeding structures which function as strong terminal sinks (Jones and Northcote, 1972; McClure, 1977). Nematode-induced sink tissues consist of cellular structures that share similarities with specialized transfer cells in which transport processes are highly upregulated (Bird, 1961; Jones and Northcote, 1972; Offler et al., 2003). Transfer cells or similar cellular structures that contribute to the amino acid supply of sink tissues are typically found in seed coats or the tapetum (Lee and Tegeder, 2004; Tegeder et al., 2000) and in plant tissues which are involved in nutrient flow during parasitic and symbiotic interactions of plants.

In order to respond to numerous physiological needs and environmental changes, plants have evolved several amino acid transporters with distinct but overlapping expression patterns and physiological properties (Lalonde et al., 2004). The majority of putative amino acid transporter ‘genes fall into two superfamilies: the amino acid transporter (ATF1) superfamily (41 members) and the amino acid-polyamine-choline (APC) superfamily (14 members) (Wipf et al., 2002). Most of the amino acid transporters from plants that have been characterized belong to the ATF1 superfamily, with the amino acid permease (AAP) family being the best studied subfamily among them (Boorer and Fischer, 1997; Boorer et al., 1996; Fischer et al., 1995, 2002; Okumoto et al., 2002). APC transporters from plants are only poorly understood. In plants APC transporters of the l-type amino acid transporter (LAT) subfamily (five members) have not been characterized and only a few members of the cationic amino acid transporters (CAT) family (nine members) have been studied (Frommer et al., 1995; Su et al., 2004). The CAT transporters contain 11 to 14 putative membrane-spanning domains and localize to the plasma membrane or the vacuole. The amino acid transporters from the ATF and APC families in plants are proton-coupled high-affinity transporters (Lalonde et al., 2004). Except for the proline transporters, all amino acid transporters display a broad substrate specificity (Rentsch et al., 1996).

In this study we describe the isolation and characterization of AtCAT6, a sink-specific amino acid transporter, the expression of which is induced by nematodes.


Expression analysis of the AtCAT6 gene

AtCAT6 is a member of the APC family of amino acid transporters with a molecular weight of about 65 kDa and contains 11–14 trans-membrane domains (Schwacke et al., 2003; Su et al., 2004). The expression of this gene was found to be induced during infestation of Arabidopsis roots with the root-knot nematode, Meloidogyne incognita. Real-time RT-PCR was carried out to determine AtCAT6 transcript abundance in response to nematode infestation and in various organs of Arabidopsis (Figure 1). In roots of infested plants, AtCAT6 expression levels were higher 1, 2 and 4 weeks after infestation than in control plants of the same age. The difference between expression levels in infested and uninfested plants increased over time (Figure 1a). The expression levels in manually dissected root-knots 2 weeks after infestation were more than threefold higher than in the rest of the infested root, indicating that the increase of AtCAT6 expression observed in whole roots is mainly due to elevated expression in root-knots (Figure 1b).

Figure 1.

 Expression of AtCAT6 during nematode infestation and in various tissues determined by quantitative real-time RT-PCR.
(a) Transcript abundance in roots 1, 2 and 4 weeks after nematode infestation (open bars) and control plants at the same age (closed bars).
(b) Transcript abundance in manually excised knots and in ‘knot-free’ root sections of infested plants 2 weeks after infestation.
(c) Transcript abundance in various tissues of soil-grown wild-type (Col-0) plants. Young flowers were not yet opened, old flowers were opened. Young siliques were <0.5 cm, old siliques were >0.5 cm.

In soil-grown, uninfested wild-type plants expression levels were highest in fully opened and mature flowers (Figure 1c). Expression levels in the mature flowers were two times higher than those in young, unopened flower buds. AtCAT6 is also expressed in siliques throughout development. The lowest levels of expression were detected in leaves and stems, and expression was barely detectable in roots.

Expression of AtCAT6 in sink tissues

To obtain more detailed information about the tissue specificity of AtCAT6 expression, we conducted a reporter gene study in which the AtCAT6 promoter was used to drive β-glucuronidase (GUS) expression. A total of 30 individual lines containing the PAtCAT6:GUS constructs were screened for β-glucuronidase activity. The staining pattern was consistent in 27 independent transgenic lines. The GUS expression patterns described below are consistent with patterns measured by real-time RT-PCR from various plant tissues and with data in two expression databases: AMPL ( and GENEVESTIGATOR ( (Ward, 2001; Zimmermann et al., 2004). These results indicate that all the elements necessary for proper expression are contained in the 1817 bp promoter fragment and that AtCAT6 is active in sink tissues.

In nematode-infested roots GUS staining was visible in root-knots (Figure 2a). In thin sections through root-knots, the promoter activity was strongest in nematode-induced giant cells (Figure 2b, arrows). In uninfested, soil-grown plants strong GUS staining was observed in flowers, predominantly in anthers, particularly in pollen, and to a lesser extent in petals (Figure 2c). During silique development staining was observed in the vasculature of the silique, the funiculus and the developing seeds (Figure 2d). One day after germination weak staining was observed in the entire seedling with the most intense staining in the root tip (Figure 2e). Two days after germination weak staining was observed in the cotyledons and strong staining in the root tips of the seedlings (Figure 2f). Staining was absent from the hypocotyl. Staining was also observed in the endosperm layer (Figure 2f, arrowhead), indicating a possible role for AtCAT6 in providing amino acids to the seed and redistribution of amino acids within the developing seedling. In roots staining preceded the formation of the lateral root meristem and was visible in the primordia of lateral roots and in the root tips of the lateral roots (Figure 2g–i). A specific staining pattern was not observed in the stems or in fully expanded leaves. After an extended period of incubation in X-Gluc, faint staining was observed in the parenchyma of young leaves and stems. In all cases staining was absent from the vasculature of those tissues (data not shown). Taken together these results show that AtCAT6 expression mainly localizes to plant tissues that are thought to be amino acid sink tissues.

Figure 2.

AtCAT6 promoter activity in PAtCAT6:uidA promoter–reporter gene plants visualized by GUS staining.
(a) Root-knot approximately 2 weeks after infestation. Knots are frequently associated with lateral roots (arrow) and in some occasions adventitious lateral roots are formed (arrowhead) which show intense staining.
(b) Section through a root-knot 2 weeks after infestation at a similar developmental stage as shown in (a), the giant cells are stained (arrows).
In flowers (c) promoter activity is detected in petals and anthers.
In siliques (d) staining is visible in the vasculature, the parenchyma and the developing seeds.
(e) Seedling 1 day after germination and (f) seedling 2 days after germination; seed and aleurone layer (arrowhead) are stained.
Throughout root development (g–i) promoter activity is detectable exclusively in the root tip: (g) lateral root meristem induction, (h) lateral root meristem formation and (i) root tip.

Functional characterization of AtCAT6 in Xenopus laevis oocytes

To examine the transport characteristics of AtCAT6, the Xenopus laevis oocyte expression system was used. Oocytes injected with AtCAT6 cRNA accumulated significantly (P < 0.01) more radiolabelled [35S]-l-methionine than water-injected control oocytes. AtCAT6-expressing oocytes accumulated 30.1 ± 4.6 pmol l-methionine per oocyte more than water-injected oocytes after a 15-min incubation (n = 5 oocytes; [H+] = 3.2 μm). This showed that AtCAT6 is a functional amino acid transporter in Xenopus oocytes. To further investigate the biophysical properties of this transporter, oocytes expressing AtCAT6 were bathed in 10 mml-alanine and recorded using a two-electrode voltage clamp system (Figure 3). l-alanine induced inward currents in oocytes expressing AtCAT6, indicating that amino acid transport through AtCAT6 is electrogenic. The substrate-induced inward currents were fully reversible and returned to baseline levels after removal of l-alanine from the oocyte bathing solution. Recordings from a representative experiment with an AtCAT6-expressing oocyte in the absence (Figure 3a) and in the presence (Figure 3b) of l-alanine illustrate the substrate-induced currents. Currents were induced in AtCAT6-expressing oocytes while no currents were induced upon the introduction of l-alanine to the bath solution containing the water-injected oocytes (Figure 3c,d). Oocytes injected with AtCAT6 displayed a background current in the absence of substrate (Figure 3a). This current was probably due to the movement of protons since it was strongly dependent on the pH of the bath solution and decreased at more alkaline pH values (not shown). These recordings were obtained by stepping the membrane potential from the holding potential, −40 mV, to ranges between −160 and 40 mV in 20-mV increments (Figure 3e). The current traces consisted of an initial capacitive transient, which relaxed to a steady-state level about 50 msec after the onset of the voltage pulse. At membrane voltages more negative than −80 mV, the currents displayed some time-dependent features and reached steady-state levels 150 msec after the onset of the voltage pulse. Steady-state alanine-dependent currents recorded between −160 and 40 mV displayed a supralinear dependence on voltage, did not saturate at negative potentials and did not reverse at positive potentials (Figure 3f). The same response was observed for all neutral amino acids. In contrast steady-state currents obtained for both positively and negatively charged amino acids saturated between −120 and −160 mV (not shown).

Figure 3.

 The addition of l-alanine to bath solutions induces currents in oocytes expressing AtCAT6 but not in water-injected oocytes.
Current traces were obtained from an AtCAT6 cRNA-injected oocyte (a, b) and a water-injected oocyte (c, d) before (a, c) and after (b, d) the addition of 10 mml-alanine to the bath solution ([H+] = 3.2 μm).
(e) The voltage pulse protocol was from 40 to −160 mV in 20 mV increments.
(f) l-alanine-induced currents were obtained for single oocytes by subtracting the steady-state currents in the absence of substrate from the l-alanine-induced currents, plotted as a function of Vm: l-alanine 50 mm (bsl00001), 20 mm (bsl00084), 10 mm (bsl00072), 2 mm (bsl00043) and 0.5 mm (•).

Substrate selectivity

To investigate the substrate specificity of AtCAT6, bath solutions containing oocytes were perfused with various substrates at a concentration of 2 mm at [H+] = 3.2 μm (Figure 4). With the exception of l-proline, all proteinogenic amino acids induced steady-state currents in oocytes expressing AtCAT6 at a membrane voltage of −160 mV. The addition of amino acids to bath solutions containing water-injected oocytes did not induce currents greater than 5 nA (not shown). AtCAT6 transported large and uncharged amino acids in preference to other amino acids (Figure 4). The non-proteinogenic amino acid l-norvaline induced currents similar to proteinogenic amino acids with the same properties, indicating that the large non-polar side chain contributes to substrate recognition. Compared with amino acids with large and uncharged side groups, amino acids with polar side chains, positively charged side chains or small side chains induced smaller currents in oocytes expressing AtCAT6. The anionic amino acids l-glutamate and l-aspartate induced less current than their respective amides, indicating that their negatively charged side chains are responsible for their poor substrate recognition. Very small amino acids such as l-glycine and γ-amino butyric acid (GABA) were barely transported by AtCAT6. The small dipeptide l-alanylalanine did not induce currents in oocytes injected with AtCAT6 cRNA, demonstrating that this peptide, and perhaps others, are not transported by AtCAT6. d-alanine induced significantly (P < 0.01) smaller currents in AtCAT6-expressing oocytes as compared to l-alanine, indicating that AtCAT6 exhibits stereo-specific substrate recognition. In the context of the pathogen inducibility of AtCAT6, it is noteworthy that at physiological pH, some of the best substrates for this transporter, as determined by the magnitude of induced currents, are six out of the eight essential amino acids for animals (l-leucine, l-phenylalanine, l-methionine, l-isoleucine, l-tryptophan, l-valine).

Figure 4.

 Substrate selectivity of AtCAT6.
Steady-state currents induced by various substrates (2 mm) were recorded at −160 mV and [H+] = 3.2 μm. Substrate-induced currents from water-injected oocytes were typically <5 nA. Values are mean ± SE from n = 9 oocytes. All amino acids with the exception of d-alanine are l-amino acids. Proteinogenic amino acids are abbreviated in three letter code: Nor, norvaline; Orn, ornithine; Cit, citrulline, GABA, γ-amino butyric acid; AlaAla, l-alanylalanine.

Kinetic properties

A more detailed kinetic analysis was performed using l-alanine, l-lysine, l-glutamate and l-methionine as representative members of groups of substrates that have different charges and induced varying amounts of currents in AtCAT6-expressing oocytes. Steady-state, substrate-dependent currents were measured at [H+] = 3.2 μm during applications of substrate at concentrations between 1 μm and 10–100 mm. Current amplitude measured at a membrane voltage of −160 mV was plotted against substrate concentration (Figure 1) and curves were fitted using Equation 1 (see Experimental procedures, Kinetic analysis section). At −160 mV K0.5 values were lowest for l-lysine followed by l-methionine, which were in a comparable range (K0.5 = 0.156 ± 0.01 mm and 0.275 ± 0.03 mm, respectively) whereas the inline image (0.06 ± 0.004 nA) for l-lysine was only about half of the inline image (0.156 ± 0.01 nA) for l-methionine. The inline image (0.134 ± 0.01 nA) for l-alanine was similar to that for l-methionine, whereas the K0.5 (5.5 ± 0.6 mm) for l-alanine was an order of magnitude higher than that for l-methionine and l-lysine. The K0.5 (45.2 ± 3.2 mm) and inline image (0.3 ± 0.04 nA) values for l-glutamate were higher than for any other substrate. These values are only an approximation, since substrate saturation could not be reached because oocytes became damaged at l-glutamate concentrations higher than 100 mm. Based on substrate specificity studies and the kinetic analysis, the substrates that may be transported under physiological conditions are large and/or positively charged amino acids, most of which are essential amino acids for animal nutrition.

inline image

Coupling mechanism

In order to determine whether AtCAT6-induced currents in oocytes are modulated by pH or Na+ ions, the electrogenic transport of amino acids in AtCAT6-expressing oocytes was measured using 10 mm alanine (Figure 6a). When the membrane potential of AtCAT6-expressing oocytes was hyperpolarized to −160 mV at [H+] = 3.2 μm, the currents induced by l-alanine in the absence of sodium were about 85% (57 ± 2 nA) of the currents in the presence of sodium (68 ± 2 nA) in n = 5 oocytes. These same oocytes were then used to record the currents induced by 50 mm alanine at [H+] = 10 μm (pH 5.0) and [H+] = 0.01 μm (pH 8.0) in the presence of sodium (Figure 6b). We used 50 mm alanine in order to more easily detect currents at low [H+]. At [H+] = 0.01 μm the currents were only about 30% (62 ± 7 nA) of the currents induced at [H+] = 10 μm (180 ± 14 nA). The larger currents in AtCAT6-expressing oocytes at more acidic pH are consistent with coupling of proton and amino acid transport through AtCAT6.

Figure 6.

 Determination of the coupling ion for electrogenic alanine transport through AtCAT6.
(a) Currents induced by 10 mm alanine in the presence and absence of sodium and (b) currents induced by 50 mml-alanine at [H+] = 10 μm (pH 5.0) and [H+] = 0.01 μm (pH 8.0). Steady-state substrate-induced currents were recorded at −160 mV. Substrate-induced currents from water-injected oocytes were typically <5 nA. Values are mean ± SE from n = 5 oocytes.

Localization of the AtCAT6 protein

Functional expression of AtCAT6 in Xenopus oocytes indicated that the protein is localized to the plasma membrane in oocytes. In order to investigate the localization of the AtCAT6 protein in planta, yellow fluorescent protein (YFP) was fused to the N-terminus of the AtCAT6 sequence. To ensure that the N-terminal tag did not interfere with AtCAT6 function, the fusion protein was expressed in oocytes (Figure 7). Fluorescence was detected exclusively in the plasma membrane of oocytes expressing the fusion protein (Figure 7a) but was absent in oocytes expressing AtCAT6 without the YFP tag. The current–voltage (I/V) relationships of oocytes expressing YFP-AtCAT6 (Figure 7b) and AtCAT6-expressing oocytes (Figure 7c) bathed in l-alanine were indistinguishable (n = 5 oocytes), which provided further evidence that the presence of the N-terminal YFP did not interfere with the function of AtCAT6.

Figure 7.

 Localization of the YFP-AtCAT6 fusion protein.
In Xenopus oocytes YFP-AtCAT6 localized to the plasma membrane of Xenopus oocytes (a), confocal image (single section). Current/voltage (I/V) relationships of l-alanine induced currents were undistinguishable in oocytes expressing (b) YFP-AtCAT6 or (c) AtCAT6. Values are mean ± SE from n = 6 oocytes.
In Arabidopsis root tips the protein localized the plasma membrane and additional endo-membrane structures.
(d) Light microscope image of an Arabidopsis root tip.
(e) Confocal image (single section) of the same root shown in (d). YFP-AtCAT6 localized to the plasma membrane and the ER. The ER is identifiable by a ring-like structure surrounding the nuclei (arrowheads, in e) and the absence of fluorescence in the centre of the nuclei. Localization was also visible in an additional endo-membrane structure, probably the Golgi apparatus.
Part (f) is an overlay of (d) and (e).

While YFP-AtCAT6 localized to plasma membranes in oocytes, we wanted to determine the localization of this transporter in plant cells. To do this, confocal imaging techniques were employed to analyse the subcellular distribution of YFP-AtCAT6 stably expressed under the control of the 35S promoter. In 12 independently transformed lines, AtCAT6 was localized primarily in the plasma membranes (Figure 7). In some cells, AtCAT6 fluorescence was observed in membranes surrounding the nucleus, which would be consistent with an endoplasmic reticulum (ER) localization (Figure 7e, arrowheads). In many of these cells, additional structures, similar in shape and size to the Golgi apparatus, were seen (Figure 7e, arrows). As AtCAT6 must traverse these compartments en route to the plasma membrane, the localization of YFP-AtCAT6 to these compartments probably results from overexpression of the protein.

Physiological characterization of AtCAT6 using Arabidopsis insertion mutants

In order to investigate the role of AtCAT6 in planta, we characterized two independent lines, SALK_067045 and SALK_076909, carrying a T-DNA insertion in the AtCAT6 locus. Both insertions are located in an exon. Homozygous knockout alleles were called cat6-1 (SALK_067045) and cat6-2 (SALK_076909). The absence of AtCAT6 transcript in both lines was verified by quantitative real-time RT-PCR and standard RT-PCR (not shown). Based on the results of our PAtCAT6:GUS experiments that indicated AtCAT6 was highly expressed in seeds and newly germinated seedlings, we tested the germination and growth of Arabidopsis seedlings on MS medium containing 2.5 mm or 5 mm of a single amino acid as the sole nitrogen source. Under these conditions only l-aspartate, l-asparagine, l-glutamate, l-glutamine, l-alanine, l-GABA and l-proline supported germination of the three lines with no differences in germination frequency between knockout alleles and wild type. After germination both cat6-1 and cat6-2 ceased to grow on l-glutamine as the sole nitrogen source whereas wild-type plants continued to grow (Figure 8). As growth stopped in both cat6 mutants, we also observed a high level of anthocyanin accumulation, as indicated by the purple colour of the leaves. Similar results were obtained when plants were grown on 5 mml-glutamine as the sole nitrogen source (data not shown).

Figure 8.

 Growth analysis of two independent knockout alleles of AtCAT6 in comparison to wild-type plants.
Wild-type plants and cat6-1 and cat6-2 plants were grown on medium containing 2.5 mml-glutamine as the sole nitrogen source (left panel) or MS medium (right panel) for 10 days. When grown under these conditions cat6 mutant plants died whereas wild-type plants continued to grow.

Since AtCAT6 is expressed in developing silique and in seeds, we quantified the content of free amino acids in mature seeds to determine whether the cat6 knockouts had any detectable differences in amino acid content as compared to the wild-type seeds. All plants were carefully grown at the same time and using a randomized design. High-performance liquid chromatography profiles of amino acids extracted from seeds were the same in wild type as in both cat6 knockout alleles. Although a phenotype for the cat6 alleles was observed in seedlings, no change in seed amino acid content was detected (not shown).

In order to investigate the role of AtCAT6 during nematode infestation both knockout alleles were inoculated with M. incognita. No difference between wild-type and mutant plants was observed in knot formation, knot development or the support of nematode development (not shown).


AtCAT6 is expressed in terminal sink tissues

Sink tissues in plants are important for plant survival through the establishment of storage tissues such as roots and seeds. These sink tissues often represent the edible portions of the plant which are also a source of essential human nutrients. Sink tissues depend on the translocation of minerals and carbon from other parts of the plant. In order to ensure the supply of these nutrients, specific long-distance and cellular transport mechanisms are needed to direct and load these nutrients into the sink tissues. In this work we identified and characterized an amino acid transporter, AtCAT6, from Arabidopsis which is specifically expressed in sink tissues including developing seeds, lateral roots and nematode-induced feeding structures. In contrast to the expression of AtCAT6 in several sink tissues, other amino acid transporters that are expressed in sink tissues showed either a very specific expression pattern exclusively in one tissue type (Lee and Tegeder, 2004; Rentsch et al., 1996; Schwacke et al., 1999; Tegeder et al., 2000) or expression in both sink and source tissues (Fischer et al., 1995; Okumoto et al., 2002). AtCAT6 appears to be more broadly expressed in plants, but specifically in sink tissues.

In roots infested by M. incognita, a root-knot nematode, feeding sites are established in which one of the most striking features is large giant cells. These cells are thought to function as sink tissues from which the nematode obtains essential nutrients from the host plant. We have shown that expression of AtCAT6 is upregulated by root-knot nematodes in giant cells, and that AtCAT6 is an amino acid transporter protein. These observations, plus the determination that AtCAT6 is localized at the plasma membrane, are all consistent with the transport of amino acids into the cytoplasm of these giant cells to possibly contribute to the nutrition of the nematode. Giant cells are thought to be symplastically isolated from the surrounding tissue (Huang and Maggenti, 1969; Jones, 1980), and therefore nutrients must be taken up across the giant cell plasma membrane. Consistent with this concept, a recent characterization of transporter gene expression in Arabidopsis roots infested with root-knot nematodes showed that at least 50 putative membrane transporters are regulated during the course of the establishment of nematode feeding sites (Hammes et al., 2005). In a microarray study of plants infested by the beet cyst nematode, Heterodera schachtii, expression analysis also revealed that certain transporters are upregulated (Puthoff et al., 2003). The gene encoding the sucrose transporter AtSUC2 was induced in roots infested with the cyst nematode (Juergensen et al., 2003), but was not found to be expressed in syncytia. In roots infected with the cyst nematode the syncytia were shown to be symplastically connected to the phloem which led to the diffusion of reporter gene product into syncytia (Hoth et al., 2005). The situation is different in case of AtCAT6, because the AtCAT6 promoter is not active in the phloem and therefore the reporter gene product would not have entered the feeding site by diffusion. Thus expression of the reporter gene product in giant cells is more likely to be due to promoter activity in giant cells. However, we cannot rule out that symplastic connections exist in the giant cells induced by the root-knot nematode. Symplastic connections would reduce the need for carrier-mediated transport across the plasma membrane except where carriers are needed for retrieval. AtCAT6 may contribute to the supply of essential amino acids for nematode feeding sites, but is not essential for the establishment of feeding sites since feeding sites can be established in plants where AtCAT6 is knocked out. This finding supports the idea that there is redundancy in amino acid transporter expression or compensatory expression of amino acid transporters in the cat6 knockouts if the nutrient loading into those structures depends on an apoplastic step.

The substrate specificity of AtCAT6

AtCAT6 is one member of a nine-member gene family in plants (Wipf et al., 2002). Previous studies indicated that two members of the AtCAT family, AtCAT1 and AtCAT5, are likely to be proton-driven high-affinity transporters that transport mainly cationic amino acids (Frommer et al., 1995; Su et al., 2004). Our characterization of AtCAT6 shows that this transporter has a high affinity for cationic amino acids and is also likely to be energized by protons. Studies on AtCAT1 and AtCAT5 used a yeast system for functional characterization; whereas we were able to characterize the function of AtCAT6 using the Xenopus oocyte heterologous expression system (Dreyer et al., 1999). Characterization in oocytes allowed for the analysis of the transport activity of AtCAT6 with the full complement of proteinogenic as well as a few non-proteinogenic amino acids.

Based on our results, we found some potential similarities and differences between the function of AtCAT6 and the previously characterized AtCAT1 and AtCAT5. For example, we found that AtCAT6 has a lower affinity for positively charged amino acids than AtCAT1 and AtCAT5. AtCAT1-mediated l-histidine transport with a K0.5 of 35 μm, AtCAT5-mediated l-arginine transport with a K0.5 of 12 μm and AtCAT6-mediated l-lysine transport with a K0.5 of 128 μm. While it is possible that AtCAT6 constitutes an amino acid transport system of lower affinity compared with the other CATs, it is also possible that the differences in affinities arise from the use of different heterologous expression systems (i.e. yeast versus oocytes; Dreyer et al., 1999). For example, the K0.5 of the amino acid transporter AtAAP1 for proline in a yeast expression system was approximately 60 μm (Frommer et al., 1993) whereas the K0.5 for proline in Xenopus oocytes was determined to be 3 mm at a membrane voltage of −150 mV (Boorer et al., 1996). In the yeast system membrane potential is not usually controlled or measured during uptake experiments, and in some cases the affinity of transporters varies with membrane potential (Boorer et al., 1996; Chandran et al., 2003; Yao et al., 2000). Therefore membrane potential may account for observed differences between yeast and oocytes. The K0.5 of AtCAT6 decreases with the membrane voltage and was 38 ± 4 μm at a membrane voltage of −40 mV (data not shown) which is a high membrane potential for yeast, but one that has been previously recorded (Kotyk, 1994). Our results correspond well with results (Frommer et al., 1995; Su et al., 2004) showing that the AtCATs transport cationic amino acids with moderate to high affinity.

In this study we found that large neutral amino acids induced larger currents than cationic amino acids in oocytes expressing AtCAT6. Therefore AtCAT6 may differ in substrate specificity from AtCAT1 and AtCAT5 because in previous studies they have been shown to transport the cationic amino acids l-arginine and l-histidine (Frommer et al., 1995; Su et al., 2004). However, direct comparisons between certain functional characteristics of AtCAT1 or AtCAT5 and AtCAT6 cannot be made because large neutral amino acids were not tested in previous studies, except for valine which efficiently reduced histidine transport through AtCAT1 (Frommer et al., 1995). We tested all amino acids at the same concentration (2 mm) using oocytes. Several amino acids were also tested across a range of concentrations in oocytes expressing AtCAT6. l-lysine was tested as a representative positively charged amino acid and l-methionine as a representative large, neutral amino acid. Although AtCAT6 has a higher affinity for l-lysine, the maximum currents were greater for large and neutral amino acids such as l-methionine. The relative transport efficiencies (i.e. relation of K0.5 to inline image) for transport of l-lysine and l-methionine are 2.1 and 1.7, respectively, thus making both amino acids equally good substrates. It will be interesting to determine whether large, neutral amino acids are transported by other members of the CAT family in plants or whether this is a unique feature of AtCAT6. It is possible that anionic amino acids are transported in their uncharged form. Further experiments using those substrates at higher pH values will be required to address that issue. To determine which amino acids are transported by AtCAT6 in planta, it will be necessary to know the relative concentrations of amino acids as well as the pH in the apoplast of the tissues in which AtCAT6 is expressed.

High-affinity transport of cationic amino acids has also been reported to be mediated by AtLHT1, a member of the lysine/histidine transporter (LHT) family (Chen and Bush, 1997). A recent paper also reported high-affinity transport of neutral and anionic amino acids and little transport of cationic amino acids by AtLHT2 (Lee and Tegeder, 2004). Full characterization of each amino acid transporter is required to draw conclusions about the functional characteristics of the entire family. In contrast to the CAT and LHT families, much more is known about the substrate specificity for the AAP family in plants. AtAAP1 to AtAAP6 were all investigated in detail in the Xenopus expression system (Boorer and Fischer, 1997; Boorer et al., 1996; Fischer et al., 2002). AtAAP1 to AtAAP5 exhibit a moderate or low affinity for their amino acid substrates, a characteristic they share with AtCAT6. Only AtAAP6 displays affinities for some amino acids in the lower micromolar range (Fischer et al., 2002). AtCAT6 exhibits a higher affinity for l-lysine than any AAP. The highest affinity for l-lysine was determined in AtAAP5 (K0.5 = 500 μm; Boorer and Fischer, 1997). Similar to AtCAT6, AtAAP5 also has a high affinity for l-methionine; however, other large and neutral amino acids were not transported well by AtAAP5. In conclusion, the transport characteristics found for AtCAT6 are unique to what has been reported for other plant amino acid transporters in that only this protein mediates the transport of cationic and large, neutral amino acids with high affinity and in the physiological range of the concentration of these substrates (Zhu et al., 2005).

A physiological role for AtCAT6

Knockout mutants of AtCAT6 did not grow on medium containing 2.5 mm or 5 mm of l-glutamine as the sole nitrogen source. This phenotype may be due to the impaired uptake of l-glutamine and hence the mutants died because of nitrogen deficiency. Uptake of l-glutamine is consistent with our studies showing that oocytes expressing AtCAT6 transport glutamine and that yeast expressing AtCAT6 is more sensitive to the toxic l-glutamine analogue (S)-2-amino-6-diazo-5-oxo-l-norleucine (DON) than the control (Su et al., 2004). However, AtCAT6 is not expressed in regions of the root that are involved in the net uptake of nutrients. The expression of AtCAT6 was found mainly in root tips, which are not involved in translocation of nutrients to the rest of the plant. This favours a model in which l-glutamine would be taken up by another unknown amino acid transporter, and the cat6 knockouts are more likely to be impaired in the translocation and/or retrieval of those amino acids. Further studies will be required to establish the precise role of AtCAT6 in the transport of amino acids into sink tissues.

In summary, the transport properties of AtCAT6 partially overlap with a number of other amino acid transporters of plants but may differ in that only AtCAT6 has been shown to mediate both high-affinity transport of both cationic and large, neutral amino acids. These amino acids are essential for animal nutrition and must be obtained from plants or other sources by heterotrophic organisms. When the agronomically important pathogen M. incognita invades plant roots, a feeding site is induced along with the expression of AtCAT6. Invasion by this pathogen triggers giant cells to form which function as sinks. Specific transport processes are altered (Hammes et al., 2005), including the induction of AtCAT6 which may be involved in transporting nutrients required for nematode development. The increased expression of AtCAT6 and, perhaps, other amino acid transporters, in seeds represent a potential target for engineering seeds with increased essential amino acids such as l-lysine and l-methionine that are often limited in a plant-based diet (Schaafsma, 2000).

Experimental procedures

DNA cloning

The AtCAT6 (At5g04770) open reading frame was amplified by RT-PCR using Pfu-polymerase (Stratagene, La Jolla, CA, USA) and the following primers: ATGGAGGTCCAAAGCAGCAGC and TCACACTTCAATTAGCTCTTTCATG. The PCR product was phosphorylated, cloned into the SmaI site of pOO2 (Ludewig et al., 2002) and sequenced to ensure that no changes were introduced by PCR.

The YFP-AtCAT6 fusion protein was constructed by amplifying the AtCAT6 open reading frame by RT-PCR using Pfu-polymerase (Stratagene) and the following primers: CACCATGGAGGTCCAAAGCAGCAGC and TCACACTTCAATTAGCTCTTTCATG. The PCR product was introduced into pENTR D/Topo (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. The insert was sequenced to ensure that no changes were introduced by PCR. The resulting entry clone was introduced into the destination plasmid pEARLYGATE N-YFP (Earley et al., 2006) to yield pEARLYGATE YFP-AtCAT6, which allows the expression of fusion proteins with YFP in the N-terminal position under control of the CaMV 35S promoter.

To express the YFP-AtCAT6 fusion protein in Xenopus oocytes the YFP-AtCAT6 cassette was released from pEARLYGATE YFP AtCAT6 as a XhoI fragment. This cassette was cloned into pOO2 and cRNA was obtained as described below.

The 1817 bp AtCAT6 promoter fragment was amplified by PCR from genomic DNA using Pfu-polymerase (Stratagene) and the following primers: GTTTCATAGATGGTTAGATGAG and TTTCCTCACCGGCCAGATTCG. The resulting PCR product was phosphorylated, cloned as a transcriptional fusion in front of the uidA gene, separated by the super-ubiquitin intron (CGT, unpublished data) and sequenced to ensure that no changes were introduced by PCR.

Quantitative PCR

Primers flanking a 172 bp region of the 3′ region of AtCAT6 and two internal standards (UBQ 10 At4g05320 and elF4 At3g60240) were designed using PrimerSelect (Lasergene; DNASTAR, Madison, WI, USA). Primer sequences are available upon request. Real-time PCR was performed by using an iCycler (Bio-Rad, Hercules, CA, USA) and SyBr green I (Molecular Probes, Eugene, OR, USA) detection. All reactions were carried out in triplicates. The reactions contained 20 nm Fluorescein (Bio-Rad) as a passive reference dye. Polymerase chain reaction was carried out using the heat-activated Hot Master Taq Polymerase (Eppendorf, Westbury, NY, USA) spiked with 1 μl of a 1:100 000 dilution of SYBR green I. The number of targeted messages in each sample was determined by relating the iCycler results for the gene of interest to a standard curve for each individual gene. Expression levels were normalized to both UBQ10 and elF4 to ensure the similarity of relative expression levels.

Nematode culture and sample collection

Wild-type Arabidopsis Col-0 seeds were sterilized and germinated on agar plates (2% sucrose, 0.3% Gamborg's B-5, 0.6% Gelrite, pH 6.1, 8-h light/16-h dark 23°C). After 7 days, five seedlings were transferred onto plates and grown vertically for 3 weeks under the same conditions. Three weeks after germination approximately 1000 root-knot nematode (M. incognita) eggs collected from nematodes in sterile culture were added to half of the plates. A set of uninfested plates was kept as a control group. Infested plants had at least 10 galls per plant. Total root material was harvested from both infested and non-infested plants at 1, 2 and 4 weeks after inoculation, frozen in liquid nitrogen and RNA from both tissues was isolated as described below. To investigate the transcript abundance in root knots compared with infested, but knot-free, root material, knots of similar size and comparable developmental stage were excised from infested roots 2 weeks after infestation with a scalpel under a stereomicroscope. The remaining tissue was separated and the two samples were then frozen in liquid nitrogen. Ribonucleic acid from both tissues was isolated independently using the Qiagen RNA Isolation Kit (Qiagen, Valencia, CA, USA) following the manufacturer's instructions.

Staining for β -glucuronidase activity

For histochemical localization of β-glucuronidase (GUS) activity, plant material was infiltrated with 1 mg ml−1 X-Gluc in staining buffer (100 mm sodium phosphate, pH 7.0, 10 mm EDTA, 10 mm K4[Fe2(CN)6], 10 mm K3[Fe2(CN)6], 1% Triton X-100, 10% DMSO). Staining was performed for 4 h at 37°C or, where specifically stated, overnight. After staining was completed the plants were transferred to 70% (v.v) ethanol. For sections, plants were not treated with ethanol, fixed at 4°C overnight in 3% v.v glutaraldehyde, 0.03% Triton X-100, 100 mm Tris pH 7.2, and subsequently embedded in 3% (w/v) agarose and sectioned on the same day to 75 μm on a Vibratome model 1500 (Vibratome, St. Louis, MO, USA).

Preparation of oocytes

Stage V or VI oocytes were surgically removed from Xenopus laevis females and manually separated into clusters of 10 oocytes, then defoliculated by incubation in 5 ml Barth's solution (88 mm NaCl, 1 mm KCl, 0.33 mm Ca(NO3)2, 0.41 mm CaCl2, 0.82 mm MgSO4, 2.4 mm NaHCO3, 10 mm HEPES, pH 7.6. with NaOH) containing 50 mg collagenase for 1.5–2 h at 23°C. Oocytes were then washed five times with 1 mg ml−1 BSA in Barth's solution. Subsequently oocytes were incubated in 100 mm KHPO4 pH 7.5 containing 0.2% (w/v) BSA for an hour and washed again five times with 1 mg ml−1 BSA in Barth's solution. All experiments were performed at least twice on oocytes from different frogs.

In vitro transcription and injection into oocytes

Complementary RNA was synthesized using the AMBION SP6 mMessage mMachine kit (Ambion, Austin, TX, USA) following the manufacturer's instructions. Oocytes were injected with 50 nl of cRNA (25 ng) with a manual oocyte microinjection pipette (Drummond Scientific, Broomall, PA, USA) on the day after removal from the frog and stored in Barth's solution plus 0.1 mg ml−1 gentamycin at 15°C. The buffer was changed every day. Recordings were made 6–7 days after injection. Pipettes for injection and recording were pulled using a model P-87 puller (Sutter Instrument Co., Novato, CA, USA).

Electrophysiological methods

Recording pipettes were pulled from 1.5 mm outer diameter thin-wall borosilicate glass (KIMAX-51, Vineland, NJ, USA). Pipettes were filled with 3 m KCl and connected to the head stage with a silver chloride wire. Pipette resistance was 0.7–1.5 MΩ in the bath solution. Two-electrode voltage clamping (TEVC) was done using a Dagan TEVC-200A amplifier (Dagan Corp., Minneapolis, MN, USA). Currents were recorded via an LM12 interface (Dagan Corp.) using Clampex 5.5.1 (Axon Instruments Inc., Union City, CA, USA).

Unless specifically stated otherwise, oocytes were bathed in Na+ Ringer (115 mm NaCl, 2.5 mm KCl, 1.8 mm CaCl2, 1 mm NaHCO3, 1 mm MgCl2, 10 mm HEPES) with continuous perfusion at about 1 ml min−1. For studies in the absence of sodium, NaCl was replaced by choline chloride.

The oocyte plasma membrane was held at −40 mV, and membrane currents were measured after stepping from the holding potential (Vh) to test potentials (Vm) between −160 and 40 mV in 20 mV increments. Each voltage pulse was applied for 500 msec. The currents were filtered at 500 Hz. Steady-state currents were obtained by calculating the average amplitude of steady-state currents between 350 and 400 msec using Axograph V2.0 (Axon Instruments). Steady-state currents in the absence of substrate were measured before and after supplying the substrate. The two data sets were averaged and subtracted from the data obtained in the presence of substrate to determine the substrate-induced currents. After every exposure oocytes were washed in substrate-free solution until the currents returned to base-line levels.

The substrate specificity of AtCAT6 was investigated by measuring steady-state substrate-dependent currents at a membrane voltage of −160 mV with 2 mm of various organic substrates (Sigma, St Louis, MO, USA) at [H+] = 3.2 μm.

Uptake of radiolabelled substrate

Uptake experiments with radiolabelled methionine were performed 6 days after injection. Oocytes were incubated in 2 ml Na+ Ringer solution at pH 5.5, containing 1 μml-[35S]-methionine (Amersham, Piscataway, NJ, USA) for 15 min at room temperature. The final concentration of l-methionine was 2 mm. To stop the uptake, oocytes were washed gently five times with 2 ml of ice-cold Na+ Ringer solution (pH 5.5) containing 2 mm of the unlabelled substrate. Oocytes were subsequently lysed in 2% SDS for 30 min, the scintillation mixture was added, and the sample was counted.

Kinetic analysis

For kinetic analysis the substrate-induced steady-state currents (i) at each test potential were fitted to Equation 1 using Sigma Plot V8.0 (Systat, Point Richmont, CA, USA):


where [S]0 is the amino acid concentration, inline image is the substrate induced maximum current, inline image is the apparent K0.5 of the substrate (S0 giving half the inline image). To normalize currents across experiments and different batches of oocytes, the currents induced by 10 mm alanine were measured for each oocyte at the beginning and the end of each of the recordings. Substrate-induced currents at −160 mV were normalized to the average values induced by 10 mm alanine at −160 mV (n = 10; I = 0.076 ± 0.015 μA). Oocytes were discarded when the currents at the end of the experiment deviated by >5% from those at the beginning.

Plant growth conditions

Two independent insertion mutants in the AtCAT6 locus (SALK_067045 and SALK_076909) were obtained from the SALK institute (La Jolla, CA, USA) (Alonso et al., 2003). Homozygous individuals were identified by PCR as described ( The absence of AtCAT6 transcript was confirmed by real-time PCR. Prior to growth assays both mutant alleles and the corresponding wild type (Col-0; CS60000) were grown together under the same conditions to generate wild-type and knockout seed for further experiments.

Germination/growth assays were performed under a 16-h light/8-h dark cycle at 22°C on nitrogen-free MS medium (Sigma) with sucrose, supplemented with a defined concentration of a single amino acid as sole nitrogen source. Germination/growth assays were scored 10 days after imbibition.


The authors wish to thank Doris Rentsch and Ruth Stadler for discussions. R. Howard Berg and Ruth Stadler are gratefully acknowledged for their help with sections and microscopy. We thank Susanne Schrack for excellent technical assistance. This work was supported by National Science Foundation grant 0344265 to EN, DS and CT and Deutsche Forschungsgemeinschaft (DFG) Schwerpunkt Dynamik und Regulation des pflanzlichen Membrantransportes SPP1108 grants HA 3468/1-1 and HA 3468/2-1 to UZH.