Different and overlapping functions of Arabidopsis LHT6 and AAP1 transporters in root amino acid uptake

Summary Plants use soil amino acids as a nitrogen source. Plasma membrane-localized transport proteins are essential for the import of a broad spectrum of amino acids into the root cells.


Introduction
Nitrogen (N) is needed by the plant in relatively high amounts to ensure sufficient growth, development, and reproduction. In agricultural systems, high biomass and seed production is often achieved through the supply of the inorganic N forms nitrate and ammonium, which can both be taken up by the root cells through the activity of plasma membrane transport proteins (Huang et al., 1999;Kaiser et al., 2002;Sohlenkamp et al., 2002;Orsel et al., 2007). However, in organic agriculture or other cropping systems that rely on degradation of organic matter such as manure for plant N nutrition, amino acids may represent an essential N source (Khan, 1971;Senwo and Tabatabai, 1998). Similarly, soils of some ecosystems contain significant amounts of amino acids, which may be the predominant N forms acquired by the root (Lipson et al., 2001;Jones and Kielland, 2002;Näsholm et al., 2009). While the preference of plants for inorganic (i.e. nitrate and ammonium) versus organic (i.e. amino acids) N forms has not been fully resolved, and might depend on the plant species (Streeter et al., 2000;Näsholm and Persson, 2001;Thornton, 2001;Persson and Näsholm, 2002;Biernath et al., 2008;Jamtgard et al., 2008;El-Naggar et al., 2009;Ge et al., 2009), soil amino acid concentrations (Näsholm and Persson, 2001), and soil pH (Kielland, 1994;Jones and Kielland, 2002), it has been shown that amino acids are an important N source for plants even in the presence of nitrate and ammonium (Aslam et al., 2001;Henry and Jeffereies, 2003;Gioseffi et al., 2012;Vinall et al., 2012).
Root epidermal cells and root hairs comprise the majority of root surface area and are thought to be the main sites for uptake of nutrients, including amino acids, from the soil (Dittmer, 1937;Itoh and Barber, 1983;Lazof et al., 1992Lazof et al., , 1996Gassmann and Schroeder, 1994;Gahoonia and Nielsen, 1998). Following entry into these root cells, amino acids move symplasmically to the xylem, which delivers the N to the shoot (Miller, 1985;Rentsch et al., 2007;Tegeder and Rentsch, 2010). In addition, root-cap cells located at the root tip might import some organic soil N, and/or amino acids may move apoplasmically in the cell wall space of the root cortex until they reach the Casparian strip of the root endodermis where they must be imported into the symplast for partitioning to the vasculature (Miller, 1985;Barlow, 2002;Walch-Liu et al., 2006). Uptake of amino acids into root hairs, epidermis, or along the apoplasmic route into the cortex and endodermis requires transport proteins mediating passage of the organic N across the plasma membrane (Tegeder, 2014).
Free amino acid concentrations in soils range from 0 to 150 μM, and while many of the protein amino acids are found, glutamate, aspartate, glutamine, asparagine, glycine, serine, and alanine are often predominant (Senwo and Tabatabai, 1998;Raab et al., 1999;Schmidt and Stewart, 1999;Weigelt et al., 2005). The type of transporter that is functioning in root uptake would depend on which amino acids are present in the respective rhizosphere and their concentrations. In Arabidopsis, to date three transport proteins have been demonstrated to affect root uptake of amino acids, and these are the amino acid permeases AAP1  and AAP5 (Svennerstam et al., , 2011, and the lysine-histidine-type transporter LHT1 (Hirner et al., 2006;Svennerstam et al., 2007Svennerstam et al., , 2008Svennerstam et al., , 2011. In addition, the compatible solute transporter ProT2 is involved in proline acquisition (Lehmann et al., 2011). Of these amino acid transporters, only AAP1 has been shown to be localized to the root epidermis and root hairs (and root tip), where uptake of large amounts of nutrients might occur (Dittmer, 1937;Itoh and Barber, 1983;Lazof et al., 1992Lazof et al., , 1996Gassmann and Schroeder, 1994;Gahoonia and Nielsen 1998), and to function in import of neutral and acidic amino acids . Transport studies using mutants suggest that AAP1 might operate in soil N uptake at high amino acid concentrations Svennerstam et al., 2011). In contrast, LHT1 and AAP5 seem to be involved in acquisition of neutral and acidic amino acids, and basic amino acids, respectively, at soil solution levels below 50 μM (Svennerstam et al., , 2011. However, in young, developing Arabidopsis plants, LHT1 function in amino N uptake seems to be restricted to the root tips, as indicated by promoter-β-glucuronidase (GUS) studies (Hirner et al., 2006), and in seedlings it is expressed throughout the root with the exception of root tips (Liu et al., 2010). Localization of AAP5 within the root has not been resolved yet, although transcripts may be present in the root cortex and endodermis (Brady et al., 2007;Gifford et al., 2008).
Here, it is hypothesized that besides AAP1, which is involved in import of glutamate and neutral amino acids into the root hairs and epidermis when present at relatively high soil concentrations Svennerstam et al., 2011), at least one additional transporter is functioning in the same cell types to access the broad spectrum of soil amino acids present at lower, naturally occurring concentrations during Arabidopsis development. Candidates might belong to the LHT family that is predicted to contain high affinity amino acid transporters for neutral and acidic amino acids (Chen and Bush, 1997;Lee and Tegeder, 2004;Hirner et al., 2006). It was found that LHT6 (At3g01760) is strongly expressed in roots and this transporter was examined further with respect to its function in amino acid root uptake using localization analyses, as well as growth and transport studies with lht6 mutants. In addition, aap1 and lht6/aap1 double mutants were examined to analyse the potential cooperation of LHT6 and AAP1 in amino acid uptake.

Plant materials and growth conditions
Arabidopsis (ecotype Columbia) wild-type and mutant plants were grown in 0.785 cm 3 pots containing Sungro professional growing mix LC1 (Seba Beach, AB, Canada) consisting of peat (70-80%), perlite, and domestic limestone (20-30%). Plant growth conditions were set to 16 h light at 150-200 μmol photons m -2 s -1 , 40% humidity, and day and night temperatures of 20 °C and 16 °C, respectively. For in vitro studies, Arabidopsis seeds were rinsed in 70% (w/w) ethanol followed by sterilization for 5 min in a solution containing 0.5% SDS (w/v) and 2% NaOCl (w/v). The sterile seeds were stratified in water for 3 d at 4 °C, and then transferred to Petri dishes with solid growth medium (see below). Seeds and growing plants were cultured in environmentally controlled chambers at 16 h light, 125-130 μmol photons m -2 s -1 , 50% humidity, and a constant temperature of 20 °C.

Identification of homozygous T-DNA insertion lines
The LHT6 T-DNA insertion line (SALK_049092.50.70) was obtained from the Arabidopsis Biological Resource Center (Ohio State University, Columbus, OH, USA). This line was screened for homozygous T-DNA insertion by PCR using the primers: T-DNA-specific 5′-TGGTTCACGTAGTGGGCCATCG-3′ and LHT6-specific 5′-CTTAAGTGCACTGGGTGAAATGG-3′ and 5′-CAGCACAAGCCCAAAAATGATGC-3′. In previous studies on the AAP1 transporter, AAP1 T-DNA insertion lines in the Wassilewskija ecotype background were analysed . To be able to compare lht6 and aap1 mutants and for double mutant production, a homozygous AAP1 T-DNA insertion line of ecotype Columbia (SAIL_871_C03) was identified by PCR using T-DNA-specific 5′-TTCATAACCAATTCTCGATACAC-3′, and AAP1-specific 5′-ATGGTCGAGAGAAGCGTACC-3′ and 5′-GATGCAAAACAGGACTGTCG-3′ primers. The PCR products obtained by using gene-specific and T-DNA left border (LB) primers were cloned into the vector pGEM ® -T Easy (Promega, Madison, WI, USA) and sequenced to determine the exact location of the T-DNA insertions in LHT1 and AAP1, respectively.

Construct preparation
The LHT6 promoter and cDNA were isolated and each cloned into pGEM ® -T Easy as described in Foster et al. (2008). To build an LHT6 promoter-LHT6 cDNA construct for complementation of the Arabidopsis lht6 mutant, the LHT6 cDNA was excised with NotI, blunted, and cloned into the blunted SpeI site of the pGEM ® -T Easy vector carrying the LHT6 promoter (Foster et al., 2008). The LHT6 promoter-LHT6 cDNA cassette was then cut with NcoI (blunted)/PstI and transferred into the SmaI/PstI site of the binary vector pTKAN derived from pPZP212 (Hajdukiewicz et al., 1994). The vector was kindly provided by Karin Schumacher, ZMBP, Tübingen, Germany. The LHT6 promoter-LHT6 cDNA in pTKAN was transferred into Agrobacterium tumefaciens strain GV3101 (pMP90) (Holsters et al., 1980;Koncz and Schell, 1986) and used for transformation of the lht6 mutant line via the floral dip method (Clough and Bent, 1998).

Uptake of 14 C-labelled amino acids
Uptake studies with amino acids at concentrations of 150 μM and 2 mM were performed with plants grown horizontally on solid fullstrength MS medium (Murashige and Skoog, 1962;pH 5.5-5.7) with myo-inositol (100 mg l -1 ), sucrose (10 g l -1 ), 2-(N-morpholino)ethane sulphonic acid (MES; 0.5 g l -1 ), and agar (8 g l -1 ) as described in Lee et al. (2007). After 6 d of growth, when the cotyledons/leaves of the seedlings have no contact with the media surface, channels were cut into the solid media and filled with radiolabelled [ 14 C] amino acids (Moravek Biochemicals) at 2 μCi along with 5 ml of non-labelled amino acids diluted in 2.5 mM MES buffer (pH 5.5-5.7) at final concentrations of 150 μM and 2 mM, respectively. At least 15 single seedlings were collected after 48 h of feeding, rinsed four times in water, and placed in vials with 2 ml of scintillation solution containing sample solubilizer (ScintiSafe™ Plus 50% Cocktail, Fisher Chemical). Radioactivity was measured for single seedlings using a Packard Tri-Carb series 1500 liquid scintillation analyzer (Downers Grove, IL, USA). As no differences in seedling dry weight were detected (see Fig. 3A), the total counts per minute and seedling were normalized to the wild type (100%). The results are presented as means ±standard deviation (SD). One-way analysis of variance (ANOVA) was used to determine statistical significance using SigmaPlot 8.0. Each experiment was repeated at least three times. To justify normalization, it was confirmed in three independent experiments that there were no differences in growth between the wild type and mutants by determining fresh and dry weights of whole seedlings by measuring six pools each consisting of eight seedlings.
Uptake of amino acids at concentrations of 30 μM was performed with plants grown vertically for 2 weeks on half-strength glycine-free solid MS medium (pH 5.7), myo-inositol (50 mg l -1 ), sucrose (5 g l -1 ), MES (0.25 g l -1 ), and agar (8 g l -1 ). Glycine was removed to exclude potential regulatory effects on uptake of amino acids offered at low concentrations. Plants of each genotype were transferred to plates that were 10 mm high and 3.5 cm in diameter, and the roots were submerged in 4 ml of uptake solution (halfstrength liquid MS without N at pH 5.7) containing 14 C-labelled amino acids at 0.28 μCi and non-labelled amino acids at concentrations of 30 μM. Plants were removed after 3 h, rinsed four times in water, and patted dry. Roots and shoots were separated and their fresh weight was analysed for at least five pools of each eight plants. After drying overnight at 60 °C, the root and shoot dry weights were measured for each pool and the radioactivity of the different tissues was determined using liquid scintillation counting (see above). Since no differences in root and shoot dry weight was detected (see Fig. 3B), the total counts per minute could be related to the dry weight of roots or whole plants and were then normalized to the wild type (100%). At least three independently grown sets of plants were analysed.

LHT6 is highly expressed in Arabidopsis roots and seedlings
Previous research demonstrated that LHT6 is expressed in buds and flowers and localized to the plasma membrane (Foster et al., 2008). To determine if the transporter is also present in roots and other organs, reverse transcription-PCR was performed using RNA from different Arabidopsis organs of 6-week-old plants and from 6-day-old seedlings. The expression profile showed high levels of LHT6 transcripts in roots and seedlings, while relatively weak expression was detected in stems, rosette leaves, buds, and flowers (Fig. 1A).
To examine tissue-specific localization of LHT6, histochemical analysis of LHT6 promoter-GUS lines was performed. GUS staining was found in the taproot of 6-day-old seedlings ( Fig. 2A, D) and in lateral roots of 2-and 3-weekold plants (Fig. 2B, C), which is consistent with the high levels of LHT6 RNA expression in these organs. Within roots, LHT6 is localized in root hairs (Fig. 2D-H) as well as in other root cells involved in uptake including epidermal, cortex, and endodermis cells of seedlings and growing plants (Fig. 2F, G). GUS staining was generally not found in the root tips of 2-and 3-week-old plants (Fig. 2H).

LHT6 expression is knocked out in the mutant
To analyse LHT6 function in planta, a homozygous Arabidopsis LHT6 T-DNA insertion line (lht6; SALK_049092.50.70) was isolated containing a T-DNA in the third out of seven exons of the LHT6 gene (Fig. 1B). Further, lht6/LHT6 complementation lines were produced by transforming the lht6 plants with an LHT6 promoter-LHT6 cDNA construct. LHT6 transcript levels were analysed in 2-week-old wild-type, lht6, and lht6/LHT6 plants by RT-PCR. The results showed similar levels of LHT6 expression in wild-type and lht6/LHT6 plants, whereas no transcripts were detected in the lht6 mutant (Fig. 1C). This suggests that LHT6 function is knocked out in the mutant, while transporter expression and function is restored in the complementation line.

Growth analyses of mutants
To analyse if the mutations in the LHT6 or AAP1 transporter affect root or shoot growth of mutants versus the wild type when cultured on media, root, shoot, and whole-plant fresh and dry weights of 6-day-old seedlings as well as of 2-and 3-week-old plants were determined. The results demonstrate that when knocking the transporters out or down, growth of lht6, aap1, and lht6/aap1 double mutants is not affected (Fig. 3). This result will also help to determine if the observed differences in amino acid uptake are in fact due to changes in transport activities rather than to alterations in growth or biomass. It is noteworthy that in some cases (see Figs 4 and 7A), biomass and uptake studies were performed with different sets of plants, and there might be the possibility of variation.

Uptake of alanine, aspartate, and glutamate offered at high concentrations is decreased in lht6 plants
To resolve if LHT6 is functioning in amino acid acquisition at high concentrations, growth and uptake studies were performed with lht6, wild-type, and lht6/LHT6 plants (cf. Lee et al., 2007). While no differences were observed when plants were grown on media supplemented with single amino acids, uptake studies with 14 C-labelled amino acids (2 mM) revealed changes in amino acid acquisition between mutant and the wild type (Fig. 4). First, uptake of [ 14 C]glutamate was analysed since this amino acid only slightly, if at all, inhibits growth of Arabidopsis plants even when it is offered at very high concentrations . In addition, three lht6/LHT6 lines were tested to determine if in all lines LHT6 function is fully restored and if they perform like the wild type (Fig. 4A). The results show that uptake of glutamate was significantly decreased by ~19% in the lht6 mutant compared with the wild-type and lht6/ LHT6 plants. No difference was found in glutamate root uptake between the wild type and the complementation lines (Fig. 4A). These data support that the observed reduction in glutamate uptake for the lht6 mutant is in fact due to the knockout of LHT6 and that LHT6 is involved in glutamate uptake when offered in high amounts (Fig. 4). Uptake of aspartate, and neutral and basic 14 C-labelled amino acids was then analysed. The results demonstrate that in addition to glutamate, aspartate and alanine uptake are also reduced in the lht6 mutant versus the wild type or complementation line by 17% and 32%, respectively (Fig. 4B). This shows that at high amino acid concentration LHT6 transports acidic amino acids and alanine.

Growth of aap1 and lht6/aap1 plants is affected on low levels of toxic amino acid analogues
To resolve if LHT6 plays a role in amino acid uptake at ecological concentrations, wild-type and lht6 plants were analysed for their ability to germinate and grow on media containing low levels of amino acid analogues that are toxic to wild-type plants. Specifically, toxic analogues for glutamine (DON), isoleucine (OMT), phenylalanine (FPA), proline (AZC), glutamate (MSX), and arginine (CAN) were used (Figs 5 and 6). In addition, growth of the aap1 mutant and lht6/aap1 double mutant was analysed. AAP1 has previously been shown to play a role in uptake of amino acids by the root, at least when amino acids are offered at high concentrations Svennerstam et al., 2011). The lht6 mutant showed no differences in growth compared with the wild type on any of the toxic analogues tested (Figs 5 and  6). However, the aap1 and lht6/aap1 plants grew better than wild-type and lht6 plants on media with toxic analogues for glutamate, isoleucine, proline, and glutamine (Figs 5 and 6). Dependent on the amino acid analogue, fresh weights were 50-190% higher in aap1 and lht6/aap1 compared with wild-type or lht6 plants (Fig. 6B, D, E, G). This suggests that less of the organic N compounds were taken up due to the knockout of AAP1 expression. Additionally, the lht6/aap1 plants grow ~50% better than the single mutants and wild type on the phenylalanine analogue (Figs 5 and 6F). This points to some function of both LHT6 and AAP1 in phenylalanine transport that is only seen in the double mutant. Neither the single nor the double mutants showed a change in growth on the arginine analogue, indicating that the basic amino acid is not a substrate for LHT6 or AAP1 (Figs 5 and 6C). Together, these results suggest a role for AAP1 in the uptake of glutamate and neutral amino acids, when they are present at low Results are shown for one experiment, but three independent experiments were performed. Error bars depict the standard deviation (± SD). Asterisks indicate significant differences when using one-way analysis of variance (ANOVA; P-values ≤0.05).

Fig. 4.
Uptake of 14 C-labelled amino acids by 6-day-old lht6, lht6/ LHT6, and wild-type (WT) seedlings exposed to 2 mM amino acid concentrations. (A) Uptake studies using [ 14 C]glutamate and plants from the WT, lht6 mutant, and three lht6/LHT6 complementation lines. (B) Uptake studies using acidic, basic, and neutral amino acids. Since no differences in fresh or dry weights were detected for lht6, lht6/LHT6, and WT seedlings (see Fig. 3A), the total counts per minute and per seedling could be normalized to the WT, which was set to 100%. Error bars depict the standard deviation (± SD). Asterisks indicate significant differences when using one-way analysis of variance (ANOVA; P-values ≤0.05). Results are shown for one experiment and are a mean of a minimum of 15 replicates (n≥15). Three independent experiments were performed.
concentrations in the growth media. The lack of a growth phenotype for lht6 mutants might suggest that (i) the analogues/amino acids tested are not substrates for LHT6 and/or (ii) other root amino acid transporters (e.g. AAP1 and LHT1) compensate for LHT6 function. However, while growth studies on toxic analogues are useful to draw conclusions on transporter function in uptake of a specific physiological substrate, they need also be interpreted with care, since inhibitory growth effects might be dependent on the specific analogue rather than its concentration. As shown in this study, for example, MSX and DON are already toxic for the wild type at a low level (1 μM) while OMT or FPA inhibit growth at 5 μM or 70 μM, respectively (Figs 5 and 6). In addition, the affinity of a transporter for the amino acid versus its analogue might differ (cf. Mourad and King, 1995). Therefore, to resolve LHT6 and AAP1 function in amino acid import into Arabidopsis roots further, uptake studies were performed at low amino acid concentrations.

Uptake of neutral and acidic amino acids offered at biologically relevant concentrations is reduced in lht6, aap1, and lht6/aap1 plants
Uptake studies were performed to analyse the function of the amino acid transporters in organic N import into roots at ecological concentrations. Free amino acid concentrations found in the soil solution vary greatly and might range from 0 to 150 μM (Chapin et al., 1993;Raab et al., 1996Raab et al., , 1999. In previous work, it was demonstrated that AAP1 is involved in root uptake of 14 C-labelled alanine and glutamine at concentrations of 150 μM . Here, it was tested if LHT6 might have a similar or diverse function. Uptake studies were performed with 6-day-old seedlings that developed a taproot with root hairs (see Fig. 2A, D-F). The results show a decrease in both [ 14 C]alanine and [ 14 C]glutamate acquisition in lht6 compared with wild-type or lht6/LHT6 seedlings of 23% and 16%, respectively (Fig. 7A). Further, uptake studies using 14 C-labelled neutral, acidic, and basic amino acids at concentrations of 30 μM were performed with 2-week-old lht6, aap1, and lht6/aap1 plants that had developed lateral roots (see Fig. 2B, G, H). Uptake was measured as total counts per minute and related to root dry weight (Fig. 7B) and whole plant dry weight ( Supplementary  Fig. S1 available at JXB online) and they were then normalized to the wild type (100%). The lht6 plants showed a significant reduction between 20% and 27% in [ 14 C]alanine, [ 14 C] glutamine, [ 14 C]aspartate, and [ 14 C]glutamate root uptake, respectively, but no change in acquisition of basic and other neutral amino acids compared with the wild type (Fig. 7B). For aap1 and lht6/aap1 plants, a significant decrease was observed in root uptake of alanine, glutamine, proline, serine, and glutamate that, dependent on the amino acid, ranged C]alanine at 150 μM concentrations by 6-day-old wild-type (WT), lht6, and lht6/LHT6 seedlings. Studies were performed with at least 15 seedlings (n≥15). As no differences were detected in seedling dry weight (see Fig. 3A), radioactivity was measured for single seedlings, and the total counts per minute and seedling were normalized to the WT type, which was set to 100%. (B) Uptake of 14 C-labelled acidic, basic, and neutral amino acids at 30 μM concentrations by roots of roots of 2-week-old WT, lht6, aap1, and lht6/aap1 plants. Five pools of plants (n=5) were analysed each containing eight plants. Since no differences in root and shoot dry weight were detected (see Fig. 3B, C), the total counts per minute could be related to dry weight of roots and were then normalized to the WT (100%). Results are shown for one experiment, but three independent experiments were performed. Error bars depict the standard deviation (± SD). Asterisks indicate significant differences when using oneway analysis of variance (ANOVA; P-values ≤ 0.05). Fig. 6. Fresh weight of wild-type (WT), lht6, aap1, and lht6/aap1 plants grown on media with toxic amino acid analogues (see Fig. 5). Plants were grown for 3 weeks on half-strength MS medium containing low concentrations of the amino acid analogue inhibiting growth of WT plants. Analogues for glutamate (1 μM N-methyl sulfoximine; MSX), arginine (50 μM canavanine; CAN), glutamine (1 μM 2-amino-6-diazo-5-oxo-l-norleucine; DON), isoleucine (5 μM O-menthy-l-threonine; OMT), phenylalanine (70 μM fluorophenylalanine; FPA), and proline (70 μM azetidine-2-carboxylate; AZC) were tested. Fresh and dry weights were determined by measuring six pools (n=6) each containing eight plants. Results are shown for one experiment, but three independent experiments were performed. Error bars depict the standard deviation (± SD). Asterisks indicate significant differences when using one-way analysis of variance (ANOVA; P-values ≤0.05). from 10% to 43%. In addition, the double mutant displayed reduced uptake of phenylalanine. During the uptake studies, some of the labelled amino acids (4-28% dependent on the amino acid) are already translocated in the xylem to the transpiring leaves. Therefore, whole seedling uptake was also determined ( Supplementary Fig. S1). The results were generally consistent with root uptake data, showing differences for alanine, glutamine, proline, serine, and glutamate (compare Fig. 7B with Supplementary Fig. S1).

LHT6 functions in uptake of amino acids from the environment
Five Arabidopsis transporters have recently been identified that might play a role in uptake of a broad spectrum of amino acids from the soil, AAP1, LHT1, AAP5, CAT6, and CAT8 (for a review, see Tegeder, 2012). LHT1 (Hirner et al., 2006), CAT6 (Hammes et al., 2006), and CAT8 (Yang et al., 2010) seem to function in amino acid import into the root tips of developing Arabidopsis plants, while the location of AAP5 function has not been resolved. Only AAP1 was found in root cells that are predicted to play a major role in nutrient uptake from the soil, including epidermis cells and root hairs (Dittmer, 1937;Itoh and Barber, 1983;Lazof et al., 1992Lazof et al., , 1996Gassmann and Schroeder, 1994;Gahoonia and Nielsen 1998;Lee et al., 2007). In this study, LHT6 was shown to be expressed in the same root cell types as AAP1, suggesting a role for LHT6 in N acquisition from the environment (cf. Lee and Tegeder, 2004 and Fig. 2; see Fig. 7B). LHT6 is a member of the LHT-like family containing high affinity transport systems for acidic and neutral amino acids (Lee and Tegeder, 2004;Hirner et al., 2006;Svennerstam et al., 2007Svennerstam et al., , 2011. In fact, feeding experiments with labelled amino acids demonstrated that LHT6 imports amino acids into the root, specifically acidic amino acids, alanine and glutamine, at low concentrations, but also acidic amino acids and alanine when offered at high amounts (see Figs 4 and 7; Supplementary S1 at JXB online).

LHT6 and AAP1 display different as well as overlapping functions
Previous uptake studies with aap1 mutants suggested that AAP1 is involved in glutamate and neutral amino acid import when available at high concentrations Svennerstam et al., 2011). However, growth studies using toxic amino acid analogues at very low concentrations showed that aap1 compared with wild-type plants grew better on analogues for glutamate, glutamine, isoleucine, and proline (Figs 5 and 6). This suggests that less of the toxic analogues were taken up by the mutant, and that AAP1 might import glutamate and selective neutral amino acids into root cells when present at low levels in the medium. Uptake studies corroborated these predictions (Fig. 7B) and demonstrated AAP1-mediated import of alanine, glutamine, proline, serine, and glutamate into the root at naturally occurring soil concentrations. However, basic amino acids (arginine) seem not to be significant substrates for AAP1 (and LHT6) at these low concentrations. This is, at least in part, in contrast to recent studies by Svennerstam et al. (2011) who performed similar experiments with a different aap1 knockout mutant with T-DNA insertion (GABI-KAT 135G05; see http://www. gabi-Kat.de, last accessed 20 June 2014) and using a low level of glutamine, alanine, aspartate, glutamate, arginine, and lysine. They could not observe any significant differences in uptake of alanine, glutamine, and glutamate between aap1 mutants and the wild type, while uptake of arginine and lysine seemed increased. Future experiments with the GABI-KAT aap1 mutant (Svennerstam et al. 2011), such as growth studies on toxic amino acid analogues, might help to resolve this inconsistency. The observed phenotype associated with the growth and uptake studies for the aap1 and lht6/aap1 double mutants but not for the lht6 mutant also indicates that the transport function of AAP1 at low levels of glutamate, glutamine, alanine, isoleucine, proline, and serine is not compensated for by LHT6.
In contrast, the lht6 mutant compared with the wild type showed no differences in either growth or uptake of the majority of neutral amino acids tested (Figs 4-7), suggesting that LHT6 might simply not transport these organic N forms. However, it cannot be excluded that AAP1 and/or LHT1 are compensating for LHT6 function when the transporter is knocked out. On the other hand, alanine and amides, specifically glutamine, are transported by LHT6 as uptake is decreased when this amino acid is offered in low amounts. Both lht6 and aap1 seem to mediate some uptake of phenylalanine at very low concentrations as the double mutant shows a significant decrease in phenylalanine acquisition (Fig. 7).
In the case of acidic amino acids, both transporters import glutamate into the root cells, when supplied at ecologically relevant concentrations, as indicated by a significantly reduced uptake of the charged amino acid in both single lht6 and aap1 mutants and double mutant (see Fig. 7; Lee et al., 2007). However, aspartate seems to be only transported by LHT6. Previous studies using heterologous expression analysis led to the conclusion that members of the LHT family are high affinity transport systems for aspartate (Lee et al., 2004;Hirner et al., 2006), which is consistent with the observation that aspartate uptake in the lht6 mutant was reduced by 25% when the amino acid was supplied at 30 μM. This is further in agreement with the observed reduction in aspartate acquisition in the double mutant, as well as with the lack of AAP1 function in aspartate transport when the transporter was biochemically analysed in yeast, Xenopus oocytes, and plants ( Fig. 7; Fischer et al., 1995Fischer et al., , 2002Lee et al., 2004;Hirner et al., 2006;Svennerstam et al., 2008Svennerstam et al., , 2011. Taken together, the growth and uptake studies support that when amino acids are present at low concentrations (i) LHT6 takes up acidic amino acids, alanine and glutamine, into the root cells; and (ii) AAP1 transports neutral amino acids and glutamate. However, additional amino acids might be substrates for the respective transporters and/or the lack of LHT6 or AAP1 expression might be counter-balanced by other amino acid transporters present in root hairs and epidermal cells.

Importance of root amino acid uptake systems and their regulation
Studies in maize (and other plant species) analysing uptake via root tips versus upper root parts support that the majority of amino acids are taken up via the root epidermis and hairs (Dittmer, 1937;Itoh and Barber, 1983;Lazof et al., 1992Lazof et al., , 1996Gassmann and Schroeder, 1994;Gahoonia and Nielsen 1998), and LHT6 and AAP1 are localized to the plasma membrane of these cell types ( Fig. 2; Lee et al., 2007;Foster et al., 2008). Thus, surprisingly, in lht6 and aap1 single or double mutants only a 10-43% reduction in uptake was observed for those amino acids that are substrates for the LHT6 and AAP1 transporters (Fig. 7). This might be due to the presence of other, as yet unidentified transporters that are also expressed in similar cells, such as members of the CAT family (Hammes et al., 2006;Yang et al., 2010). In addition, it might well be that in Arabidopsis both the symplasmic and the apoplasmic transport routes are of high relevance for amino acid uptake into root cells. The apoplasmic route would require amino acid importers in the cortex and endodermis cells to circumvent the apoplasmic blockade (i.e. Casparian strip) and to move the organic N to the vascular bundle. At least in root nodules, such cortex-and endodermis-localized organic N transporters have recently been found, and their importance for shoot N supply has been demonstrated (Collier and Tegeder, 2012). LHT6 is not only expressed in the root epidermis and root hairs, but also in the cortex and endodermis, supporting a function in both the symplasmic and apoplasmic transport pathway. These transport pathways might further involve other transporters including AAP5 for basic amino acids, as aap5 mutants show a severe reduction in arginine and lysine uptake (Svennerstam et al., , 2011, and root cell sorting experiments with young seedlings indicate expression of AAP5 in the root cortex and endodermis (Brady et al., 2007;Gifford et al., 2008). However, localization of AAP5 function in roots of seedlings and especially in developing plants still remains to be resolved. In addition, uptake of amino acids by the root tip may play a more important role than previously assumed, since mutants of LHT1 show a reduction in uptake of neutral and acidic amino acids by up to 80% and the expression of LHT1 in developing plants seems to be constrained to the tip of lateral roots (Hirner et al., 2006;Svennerstam et al., 2007Svennerstam et al., , 2008Svennerstam et al., , 2011. On the other hand, the strong reduction in root amino acid acquisition in lht1 plants might also hint at an additional, regulatory role for LHT1 in amino-N uptake. In the present studies, uptake experiments were performed in 2-week-old plants when LHT6 and AAP1 expression in the wild type is restricted to the root (see Fig. 2; Lee et al., 2007) while uptake studies with lht1 mutants were assayed when wild-type LHT1 expression is high in mesophyll cells (cf. Hirner et al., 2006;Svennerstam et al., 2007Svennerstam et al., , 2011. Further, compared with wild-type plants, lht1 mutants accumulate amino acids in the leaf cell wall. These high apoplasmic levels might negatively regulate root uptake of neutral and acidic amino acids, finally leading to a severe reduction in amino acid acquisition in the lht1 plants (Hirner et al., 2006;Forsum et al. 2008;Svennerstam et al., 2007Svennerstam et al., , 2008Svennerstam et al., , 2011. A similar regulation has been shown for nitrate and ammonium uptake, where high amino acid levels down-regulate N uptake (Wang et al., 2012;Nacry et al., 2013). In line with this is also the observation that N uptake from the soil seems to be positively regulated when amino acid import into leaf cells is increased Zhang et al., 2010). This further suggests that, while amino acid transporter function in root cells is generally important for organic N acquisition from the soil, their regulation by shoot-localized transporters such as LHT1 (Hirner et al., 2006;Tan et al., 2010;Zhang et al., 2010) or by shoot-derived signals might be essential to govern the amounts of amino acids that are taken up by the root.
To exclude potential shoot effects on root amino acid uptake, Hirner et al. (2006) re-expressed the LHT1 transporter in the shoot of lht1 plants by using a leaf-specific promoter. When cultured on high (5 mM) aspartate, the root-specific lht1 mutants showed the same growth phenotype as lht1 plants. Therefore, it was concluded that LHT1 function in roots rather than in mesophyll cells is responsible for aspartate uptake. Unfortunately, uptake experiments with aspartate or other amino acids were not carried out to support this conclusion further. These would be especially interesting as LHT6 is also involved in root aspartate import at both low and high concentrations (see Figs 4 and 7), and one might expect that loss of LHT1 function in roots is, at least in part, complemented by LHT6. On the other hand, LHT6 seems not to be expressed in root tips (Fig. 2H), and aspartate uptake might mainly occur via the tip (see above). Clearly, further studies will need to be performed to dissect the role of root-localized transporters in amino acid acquisition versus the regulatory function of shoot-expressed transporters and their contribution to root N uptake.

Supplementary data
Supplementary data are available at JXB online. Figure S1. Uptake of 14 C-labelled acidic, basic, and neutral amino acids at 30 μM concentrations by whole (root and shoot) 2-week-old wild-type, lht6, aap1, and lht6/aap1 plants.