AtDUR3 represents the major transporter for high-affinity urea transport across the plasma membrane of nitrogen-deficient Arabidopsis roots


(fax +49 711 4592 3295; e-mail


Despite the fact that urea is a ubiquitous nitrogen source in soils and the most widespread form of nitrogen fertilizer used in agricultural plant production, membrane transporters that might contribute to the uptake of urea in plant roots have so far been characterized only in heterologous systems. Two T-DNA insertion lines, atdur3-1 and atdur3-3, that showed impaired growth on urea as a sole nitrogen source were used to investigate a role of the H+/urea co-transporter AtDUR3 in nitrogen nutrition in Arabidopsis. In transgenic lines expressing AtDUR3-promoter:GFP constructs, promoter activity was upregulated under nitrogen deficiency and localized to the rhizodermis, including root hairs, as well as to the cortex in more basal root zones. Protein gel blot analysis of two-phase partitioned root membrane fractions and whole-mount immunolocalization in root hairs revealed the plasma membrane to be enriched in AtDUR3 protein. Expression of the AtDUR3 gene in nitrogen-deficient roots was repressed by ammonium and nitrate but induced after supply of urea. Higher accumulation of urea in roots of wild-type plants relative to atdur3-1 and atdur3-3 confirmed that urea was the substrate transported by AtDUR3. Influx of 15N-labeled urea in atdur3-1 and atdur3-3 showed a linear concentration dependency up to 200 μm external urea, whereas influx in wild-type roots followed saturation kinetics with an apparent Km of 4 μm. The results indicate that AtDUR3 is the major transporter for high-affinity urea uptake in Arabidopsis roots and suggest that the high substrate affinity of AtDUR3 reflects an adaptation to the low urea levels usually found in unfertilized soils.


Urea is excreted into the environment by a variety of organisms and represents a readily available nitrogen source in soils. In addition, urea has become the most frequently used form of nitrogen fertilizer worldwide due to its high nitrogen content and low production costs ( In most soils, urea is degraded rapidly to ammonium and CO2 by urease, a nickel-dependent enzyme that is synthesized and released by microorganisms (Watson et al., 1994). The half-life of urea in soils is short, ranging from several hours to a few days depending on the microbial activity of the soil. Hence, urea concentrations in soils are usually low and rarely exceed 70 μm in fertilized crop-planted soils (Gaudin et al., 1987). Considering this low urea concentration and the high urease activity in arable soils it has been assumed that plants take up most urea-derived nitrogen in the form of ammonium or as its nitrification product, nitrate (Marschner, 1995; Polacco and Holland, 1993).

For a long time urea was believed to enter plant cells via diffusion through plant membranes because of its low molecular weight and neutral character (Gallucci et al., 1971). A very limited number of short-term uptake studies indicated that urea uptake may be regulated by the plant, for example when ammonium and nitrate exerted adverse effects on urea uptake (Bradley et al., 1989). A major drawback of most in planta studies was a lack of verification of whether urea itself or its degradation product ammonium had been transported across plant membranes. The first reliable experimental evidence for protein-mediated urea uptake by plant cells was obtained by measurements of short-term influx of 14C-labeled urea in algal cells. This influx was inhibited by the ATPase inhibitor DCCD (N,N′-dicyclohexylcarbodiimide) and the protonophore carbonyl cyanide m-chlorophenylhydrazone (CCCP) and, therefore, appeared to be coupled to the proton gradient across the plasma membrane (Wilson et al., 1988). Since concentration-dependent uptake followed biphase or even multiphase kinetics, the authors suggested the combined action of a high- and a low-affinity urea transport system in planta. Kinetically and energetically, these transport systems are clearly set apart from a passive, diffusion-like transport mechanism (Wilson and Walker, 1988). More recently, CCCP-sensitive urea uptake has also been confirmed in Arabidopsis suspension cells (Liu et al., 2003a).

Heterologous expression of plant aquaporins in oocytes or in a yeast mutant defective in urea uptake showed that several members of the PIP, NIP or TIP subfamilies of aquaporins facilitate urea transport (Eckert et al., 1999; Gerbeau et al., 1999; Klebl et al., 2003; Liu et al., 2003b). Analogous to observations made with mammalian aquaporins (King et al., 2004), TIP2;1-mediated urea transport was pH insensitive and increased with a linear concentration dependency (Liu et al., 2003b). Plasma membrane-localized aquaporins were proposed as promising candidates for the low-affinity component of urea uptake by plant roots (Kojima et al., 2006). Nevertheless, it should be noted that for technical reasons, only micromolar concentrations of urea could be used to measure urea transport by aquaporins, so a critical analysis of high-affinity urea transport has not yet been possible.

Growth complementation of the urea uptake-defective yeast mutant dur3 on urea as a sole nitrogen source was conferred both by aquaporins and also by the Arabidopsis gene AtDUR3. Unlike growth complementation by aquaporins, however, AtDUR3-mediated urea transport was pH dependent (Liu et al., 2003a). Two-electrode voltage clamp analysis as well as accumulation of 14C-labeled urea in AtDUR3-expressing Xenopus laevis oocytes clearly indicated co-transport of urea with protons (Liu et al., 2003a). Similar to urea-induced currents, uptake of radio-labeled urea into AtDUR3-expressing oocytes was concentration dependent and saturated at approximately 50 μm urea (Liu et al., 2003a). Both transport assays allowed the calculation of a Km value of about 3 μm, indicating that AtDUR3 has the potential to mediate high-affinity transport of urea in Arabidopsis.

Orthologous AtDUR3 expressed sequence tag (EST) clones were found in other plant species but only a single DUR3 homolog has been identified in any plant species investigated to date. Besides aquaporins, no other urea transporters have been identified in plants (Kojima et al., 2006). A phylogenetic analysis of AtDUR3 and homologous amino acid sequences derived from other organisms revealed a relatively high similarity among DUR3 proteins from plants and yeast and indicated that all these sequences belong to the superfamily of sodium solute symporters (SSS; Liu et al., 2003a). Currently, the SSS family includes >100 members of prokaryotic and eukaryotic origin (Jung, 2002), and certain of the proteins have been described to transport sugars, amino acids, nucleosides, myo-inositols, vitamins, phenyl acetate, water and urea (Reizer et al., 1994; Saier, 2000; Turk and Wright, 1997). When the substrate specificity of AtDUR3 was tested, no current was observed in response to glucose, galactose, myo-inositol or proline. Only the presence of thiourea generated a current in AtDUR3-expressing oocytes that was weaker than that observed for urea (Liu et al., 2003a). Although this observation suggested a rather high specificity of AtDUR3 for urea among structurally unrelated substrates, the question remained open as to whether urea transport is the major function of AtDUR3 in planta.

The aim of the present study was to investigate and quantify the contribution made by AtDUR3 to urea uptake in Arabidopsis roots. For this purpose, growth on urea as the sole nitrogen source, urea concentrations in the root tissue and root uptake capacities for 15N-labeled urea were determined in wild-type plants and in two independent T-DNA insertion lines defective in AtDUR3 gene expression. Furthermore, promoter–reporter gene fusions were expressed in transgenic plants to localize AtDUR3 promoter activity in roots, gene expression studies under different nitrogen regimes were performed and immunological approaches were undertaken to determine membrane localization.


Disruption of AtDUR3 causes growth inhibition on urea when supplied as a sole nitrogen source

To investigate the contribution made by AtDUR3 to urea uptake in Arabidopsis roots, two independent T-DNA insertion lines in the Columbia-0 (Col-0) background were isolated from the insertion mutant collections of the Salk Institute Genomic Analysis Laboratory (Alonso et al., 2003). The T-DNAs were found to be inserted either 520 or 586 bp downstream of the start codon (Figure 1a), and the lines were named atdur3-1 and atdur3-3, respectively. Roots of hydroponically grown Col-0, atdur3-1 and atdur3-3 plants were then harvested for the extraction of total RNA and subsequent RNA gel blot analysis. Transcript levels of AtDUR3 were low in nitrogen-sufficient but strongly increased in nitrogen-deficient roots of wild-type plants, whereas no AtDUR3 mRNA was detected in roots of homozygous progenies from either T-DNA insertion line (Figure 1b).

Figure 1.

 Disruption of AtDUR3 leads to impaired growth under urea supply.
(a) Schematic representation of the positions of the T-DNA insertions in atdur3-1 and atdur3-3. The insertions in the AtDUR3 gene are located 520 and 586 bp, respectively, downstream of the putative translation start.
(b) Ribonucleic acid gel blot analysis of RNA from nitrogen-starved roots from wild-type Col-0, atdur3-1 and atdur3-3 plants using the coding sequence of AtDUR3 as a probe.
(c) Growth of the insertion lines atdur3-1 and atdur3-3 and their corresponding wild type (Col-0) on sterile half-strength MS medium supplied with 1 μm NiSO4 and 50 μminline image and different concentrations of ammonium nitrate or urea as a sole nitrogen source.

The two atdur3 insertion lines did not show any visible growth difference on soil or nutrient solution supplemented with nitrate or ammonium as the nitrogen source. When grown on sterile agar plates supplemented with urea as a sole nitrogen source, however, growth differences between Col-0 and atdur3 plants became apparent. While wild-type plants developed green leaves when supplied with 1 mm and even as little as 0.5 mm urea, both insertion lines became chlorotic and accumulated more anthocyanins than the wild type, this being another visible sign of nitrogen deficiency (Figure 1c). Shoot biomass production was little affected, if at all. With a supply of ammonium nitrate, shoot growth of all lines was better than with urea and no phenotypical differences were observed among the different lines (Figure 1c).

Plasma membrane localization of AtDUR3

The intracellular localization of AtDUR3 was then investigated employing protein gel blot analysis of membrane protein fractions from Arabidopsis roots. An antibody raised against 14 amino acids of the C-terminus of AtDUR3 detected a single band at approximately 55 kDa in the microsomal membrane fraction of Col-0 roots (Figure 2a). This was somewhat lower than the calculated molecular weight of 75 kDa for AtDUR3, but corresponded to the expected size with respect to the hydrophobicity of the protein. No signal was detected in microsomal membrane protein fractions from roots of the atdur3-1 and atdur3-3 insertion lines, confirming the specificity of the antibody. In wild-type roots accumulation of the AtDUR3 protein strongly increased during nitrogen deficiency (Figure 2b). Microsomal fractions from roots that were starved of nitrogen for a period of 3 days were then separated by two-phase partitioning into fractions enriched with plasma membranes or endosomal membranes (Larsson et al., 1987). In a subsequent protein gel blot analysis (Figure 2c), use of an antibody raised against the Arabidopsis plasma membrane H+-ATPase (AHA2) confirmed enrichment of plasma membrane proteins in the upper phase (U), while enrichment of endosomal membrane proteins in the lower fraction (L) was verified by detection of DET3, a subunit of the endosomal V-type H+-ATPase (Schumacher et al., 1999). A highly preferential enrichment of AtDUR3 in the upper phase of nitrogen-deficient protein fractions indicated plasma membrane localization of AtDUR3.

Figure 2.

 AtDUR3 localizes to the plasma membrane.
(a) Protein gel blot analysis of microsomal membrane fractions from roots of wild-type (Col-0) or atdur3-1 and atdur3-3 plants cultured under nitrogen deficiency. For AtDUR3 protein detection a polyclonal antibody raised against the C-terminus of AtDUR3 was employed. An AtDUR3-specific signal was detected at approximately 55 kDa.
(b) Protein gel blot analysis of root membrane fractions from roots of wild-type (Col-0) plants cultured under a continuous supply of 2 mm ammonium nitrate (+N) or under nitrogen deficiency for 1, 2 or 3 days (−N).
(c) Microsomal fractions (M) from nitrogen-deficient roots (−N 3d) of wild-type (Col-0) plants were separated by aqueous two-phase partitioning into a plasma membrane-enriched upper phase (U) and an endosomal membrane-enriched lower phase (L). Protein gel blot analysis was conducted with antibodies against AtDUR3, an Arabidopsis plasma membrane H+-ATPase (AHA2) and a subunit of an endosomal H+-ATPase (DET3).
(d) Intracellular localization of the AtDUR3 protein in nitrogen-starved root hairs from Col-0 (a and c) or from atdur3-1 (b and d) plants. Whole-mount root samples were stained with an anti-AtDUR3 antiserum and a Cy3-conjugated red-fluorescing secondary antibody. Whole-mount images were taken by an inverted fluorescence microscope equipped with an apotome. The bar indicates 10 μm.

In a whole-mount immunohistochemical approach, fluorescence imaging assisted by an apotome was directed to root hairs, which are usually less vacuolated and densely filled with cytoplasm (Figure 2d-A). After whole-mount specimens of root hairs from nitrogen-starved roots of Col-0 plants were incubated with AtDUR3-specific antiserum and a Cy3-conjugated, secondary fluorescent antibody, red fluorescence was observed along the border of individual root hair cells (Figure 2d-C). Due to the absence of a large vacuole in the hair tip (Figure 2d-A), localization of AtDUR3 could be assigned to the plasma membrane. No AtDUR3 signal was detected in root hairs of the atdur3-1 insertion line (Fig. 2d-D). Taken together, both independent experimental approaches indicated that AtDUR3 accumulated in the plasma membrane of root cells of nitrogen-deficient Arabidopsis plants.

Localization of AtDUR3 promoter activity in outer root cells

Cell type-specific expression of AtDUR3 was investigated in Arabidopsis plants transformed with an AtDUR3 promoter:GFP construct containing 1046 bp of the 5’-upstream region of AtDUR3. Green fluorescent protein-dependent fluorescence in the roots of several independent transformants cultured under nitrogen-sufficient conditions was detected at a low level in the meristem of root apices and in the root cap as well as in basal root hair zones, but was absent from the younger root hair zone (Figure 3a,c,e). Green fluorescent protein-derived fluorescence strongly increased after plant culture under nitrogen deficiency (Figure 3b,d) when AtDUR3 transcripts and protein abundance also increased (Figures 1b, 2b). The nitrogen deficiency response of AtDUR3 promoter activity was confined to epidermal and cortical cells in the basal root hair zone but limited only to epidermal cells in the younger root hair zone (Figure 3b,d). In the root hair zone, promoter activity was also observed near the xylem in vasculature tissues, although at a comparatively low intensity (Figure 3b).

Figure 3.

 Localization of the promoter activity of AtDUR3.
Transgenic plants expressing an AtDUR3-promoter:GFP construct were grown on MGRL agar plates under nitrogen-sufficient (+N) or nitrogen-deficient conditions (−N). Images from root tips (e and f) or lower (c and d) and upper (a and b) root hair zones were taken by an apotome-equipped fluorescence microscope. Bars represent 50 μm. rh, ep, co and xy indicate root hair, epidermis, cortex and xylem, respectively.

Regulation of AtDUR3 gene expression by nitrogen availability

Changes in AtDUR3 gene expression in response to different nitrogen treatments were monitored in roots following RNA extraction from hydroponically grown Col-0 plants. While levels of AtDUR3 mRNA were high in nitrogen-deficient roots, they were rapidly repressed after resupply of ammonium or nitrate (Figure 4). Both forms of nitrogen, ammonium and nitrate, seemed to act with high efficiency even though residual mRNA levels were slightly higher after the ammonium treatment. Thus, upregulation of AtDUR3 gene expression under nitrogen deficiency most likely reflected a de-repression from ammonium, nitrate or other nitrogen sources, similar to that reported for the ammonium transporter genes AtAMT1;1 and AtAMT1;3 (Gazzarrini et al., 1999; Rawat et al., 1999).

Figure 4.

 Regulation of AtDUR3 gene expression in Arabidopsis roots in dependence on the nitrogen supply.
Ribonucleic acid gel blot analysis of total root RNA from hydroponically grown 6-week-old Arabidopsis plants using the coding sequence of AtDUR3 as a probe. Plants were cultured under continuous supply of 2 mm ammonium nitrate (+N) or under nitrogen deficiency for 4 days (−N). Nitrogen-deficient plants were resupplied with 2 mminline image or inline image or with 1 mm urea in the presence of the urease inhibitor phenylphosphorodiamidate (PPD) for up to 24 h. Ethidium bromide-stained RNA (EtBr) served as a loading control.

Resupply of urea to nitrogen-deficient plants led to a dramatic upregulation of levels of AtDUR3 mRNA in roots (Figure 4), which exceeded even the transcript levels present in nitrogen-deficient plants. Thus, AtDUR3 gene expression was also substrate inducible, similar to the transcriptional regulation of the high-affinity nitrate transporter AtNRT2;1 by nitrate (Lejay et al., 1999).

Role of AtDUR3 in the accumulation of urea in roots

Urea is highly sensitive to enzymatic degradation by urease, which is among the most persistent enzymes in nature and almost ubiquitously expressed by most organisms (Polacco and Holland, 1993). Influx of urea into the roots of nitrogen-deficient wild-type plants was, therefore, determined in the presence of the urease inhibitor phenylphosphorodiamidate (PPD). Neither urea influx and translocation to shoots nor AtDUR3 gene expression or urea concentrations in the roots of nitrogen-sufficient or nitrogen-deficient plants significantly differed in the presence or absence of PPD under our growth conditions (Figure S1). Moreover, an influx analysis of 15N-labeled urea was performed with Arabidopsis plants grown in a sterile medium in Magenta boxes. A comparison with influx analyses conducted under non-axenic growth conditions provided no indication of urea degradation prior to short-term urea uptake from nutrient solutions (data not shown).

In order to independently verify whether urea is taken up as an intact molecule or else in the form of its degradation product (ammonium), the concentration of urea in the root tissue of plants subjected to different nitrogen treatments was determined. To avoid any possible degradation of urea, nutrient solutions were supplemented with PPD. When Col-0 plants and the two insertion lines atdur3-1 and atdur3-3 were cultured with a continuous supply of 2 mm ammonium nitrate, root urea concentrations were around 25 μmol g−1 dry weight (DW) and did not differ significantly between lines (Figure 5). Under conditions of nitrogen deficiency, urea levels in all lines decreased by approximately 80%. After resupply of 50 or 100 μm urea for 3 h to nitrogen-deficient plants, urea accumulated only in wild-type roots and reached three- to fivefold higher levels than in the two insertion lines (Figure 5). Thus, functional expression of AtDUR3 enhanced the accumulation of intact urea molecules, suggesting the acquisition of externally fed urea by AtDUR3 in planta. However, after supply of 1 mm urea, accumulation of urea in both atdur3 insertion lines increased to a level corresponding to 60% of the wild type, suggesting that other, low-affinity, transport systems then contributed to urea uptake.

Figure 5.

 Urea accumulation in roots of wild-type plants and atdur3 insertion lines in response to urea resupply.
Plants were cultured hydroponically for 6 weeks under continuous supply of 2 mm ammonium nitrate (+N). After a nitrogen starvation treatment for 4 days (−N), plants were resupplied with different concentrations of urea for 3 h in the presence of the urease inhibitor phenylphosphorodiamidate (PPD). Urea concentrations were determined in freeze-dried root samples.

High-affinity urea uptake by AtDUR3 in Arabidopsis roots

For a concentration-dependent influx analysis of 15N-labeled urea in roots, the two insertion lines atdur3-1 and atdur3-3 were grown together with their wild type for 38 days on nutrient solution containing 2 mm ammonium nitrate before they were subjected to nitrogen deficiency for another 4 days. In a concentration range of 3–200 μm, urea influx in Col-0 roots approached saturation at approximately 17 μm (Figure 6a). In contrast, urea influx into roots of atdur3-1 and atdur3-3 plants was between 70 and 90% lower than in the wild type and identical in both insertion lines, showing no saturable kinetics within this concentration range.

Figure 6.

 Disruption of AtDUR3 decreases the capacity for high-affinity uptake of 15N-labeled urea.
Concentration-dependent influx of 15N-labeled urea into roots of Col-0, atdur3-1 and atdur3-3 plants. Plants were cultured in nutrient solution containing 2 mm ammonium nitrate for 38 days before transfer to nitrogen deficiency for 4 days.
(a) Urea influx into wild-type and atdur3-1 or atdur3-3 plants at an external supply of 3–200 μm urea.
(b) Urea influx into roots of wild-type and atdur3-1 plants at an external supply of 200–1200 μm urea. Symbols indicate means ± SD, n = 10.

A low-affinity uptake analysis conducted with the wild-type Col-0 showed that urea influx continued to increase with an almost linear concentration dependency between 200 and 1200 μm (Figure 6b). Urea influx in atdur3-1 increased with increasing external supply at a similar rate as in the wild type. At 1.2 mm urea, uptake rates in atdur3-1 were only 20–30% lower than in the wild type. Thus, kinetic analysis revealed a dominant role for AtDUR3 only in high-affinity urea uptake in Arabidopsis roots and allowed calculation of an apparent affinity constant (Km) for urea transport by AtDUR3 of 4.0 μm.


Heterologous expression in yeast and oocytes allowed identification of AtDUR3 as a high-affinity urea transporter in Arabidopsis (Liu et al., 2003a). Belonging to the SSS superfamily, AtDUR3-mediated urea transport depends on co-transport with protons. Of the broad range of neutral and charged solutes that have been reported as substrates for SSS-type transporters in different organisms (Reizer et al., 1994), heterologous expression studies with AtDUR3 permitted the testing of only a few (Liu et al., 2003a). Furthermore, these studies could not exclude the possibility that urea permeation might reflect a physiologically irrelevant transport activity of AtDUR3. In other words, the physiological nature of this unique transport system as to its significance in plant nitrogen nutrition has been poorly understood so far. The present study set out to investigate the physiological role of AtDUR3 in urea transport in planta and shows that AtDUR3 does indeed represent the major transporter for high-affinity urea uptake in Arabidopsis roots.

AtDUR3 acts as a nitrogen-regulated urea transporter at the root plasma membrane

A general feature of membrane transporters that fulfill a function in nutrient uptake by roots is their transcriptional upregulation under limiting supply of the corresponding nutrient (Ahn et al., 2004; Gazzarrini et al., 1999; Rausch and Bucher, 2002; Takahashi et al., 2000). In the case of the high-affinity nitrate and ammonium transporter genes in Arabidopsis roots, NRT2;1 and AMT1;1, transcriptional upregulation under nitrogen deficiency reflected a de-repression most likely due to decreasing root concentrations of glutamine or other reduced nitrogen forms (Nazoa et al., 2003; Rawat et al., 1999; Vidmar et al., 2000). Corresponding with such a nitrogen-dependent regulation, transcript levels of AtDUR3 increased in nitrogen-deficient roots and were strongly repressed after resupply of ammonium or nitrate (Figure 4). Since nitrate-dependent repression of AtDUR3 mRNA levels was even stronger than that of ammonium, it is possible that nitrate per se is a signal for transcriptional repression in addition to a repression by its downstream metabolites (Wang et al., 2000, 2001). Moreover, levels of AtDUR3 transcript were strongly induced after resupply of urea to nitrogen-deficient roots (Figure 4). Substrate induction was even stronger than the preceding de-repressive effect by nitrogen deficiency in root cells. This suggests that AtDUR3 is not exclusively a component of the nitrogen deficiency stress response in Arabidopsis but also represents a substrate-inducible transport system, similar to NRT2;1 for nitrate (Lejay et al., 1999; Zhuo et al., 1999).

Two independent experimental approaches indicated that the AtDUR3 protein resides predominantly in the plasma membrane of root cells. First, protein gel blot analysis employing a specific antibody raised against the C-terminus of AtDUR3 documented a several-fold higher abundance of the protein in a plasma membrane-enriched protein fraction from nitrogen-deficient roots relative to a protein fraction depleted of plasma membrane proteins (Figure 2c). The fact that a small portion of AtDUR3 protein was also detected in endosomal membrane fractions was most likely due to an incomplete separation of the two fractions. In addition, a minor fraction of AtDUR3 may have resided in endosomal compartments, reflecting, for example, proteins that were trafficking to or from the plasma membrane (Takano et al., 2005). Secondly, whole-mount immunohistochemistry allowed tracing of AtDUR3-dependent fluorescence along the border of root hair cells (Figure 2d). Since these cells were densely filled with cytoplasm, the observed fluorescence could be assigned to the plasma membrane. Also in this approach, only a minor portion of the labeled protein appeared to be localized inside the cells. Taken together, the highly predominant localization of AtDUR3 in the root plasma membrane strongly supports a role for the protein in substrate exchange between root cells and the external medium.

A role for AtDUR3 in urea uptake from the external medium was indicated by a phenotypical analysis of two T-DNA insertion lines, in which no AtDUR3 mRNA or protein could be detected (Figures 1b and 2a). The absence of any visual symptoms or growth defects when ammonium nitrate was supplied as the sole nitrogen source suggested that gene disruptions in AtDUR3 did not interfere with the acquisition of ammonium or nitrate in general. When grown on agar medium supplied with 500–1000 μm urea as a sole nitrogen source, both insertion lines atdur3-1 and atdur3-3 exhibited chlorotic leaves and appeared to accumulate increased levels of anthocyanins (Figure 1c). These two symptoms are highly indicative of nitrogen deficiency (Marschner, 1995; Steyn et al., 2002) and suggest that AtDUR3 also contributed to urea uptake in the millimolar concentration range. In fact, concentration-dependent uptake studies confirmed that urea influx in the atdur3 mutant was still significantly lower than in the wild type in a concentration range from 0.5 to 1.2 mm urea, although the contribution of AtDUR3 to urea influx did not change in absolute terms (Figure 6b). As there are no close homologs to AtDUR3 within the SSS superfamily of Arabidopsis (Kojima et al., 2006), it is likely that AtDUR3 performs a unique physiological function in urea-based nitrogen nutrition.

Urea concentrations in the roots of hydroponically grown wild-type plants decreased approximately fivefold under conditions of nitrogen deficiency (Figure 5). Although our approach to determine urea cannot rule out the interference of other ureides, this observation supports the notion that internal accumulation of urea depends on provision of nitrogen to the roots and that root urea pools are also broken down under nitrogen deficiency during vegetative growth (Walker et al., 1985). As nitrogen-dependent changes in internal urea pools were similar in wild-type and insertion lines, the loss of AtDUR3 function did not substantially impair the internal utilization of urea. However, slightly lower urea concentrations in the two insertion lines after a period of nitrogen starvation pointed to the possibility that AtDUR3 contributes to the retrieval of urea lost by efflux, e.g., via urea-transporting aquaporins in the plasma membrane (Eckert et al., 1999; Gerbeau et al., 1999). The strong increase in urea concentrations in roots after resupply of urea to nitrogen-deficient wild-type plants relative to the insertion lines (Figure 5) clearly indicated that AtDUR3 mediated a significant accumulation of urea when supplied in the micromolar concentration range; at millimolar concentrations of external urea, additional transport processes must have contributed to urea uptake.

Transient storage of urea inside the cell will most likely depend on urea transport into the vacuole, because urease is a cytoplasmic enzyme (Witte et al., 2002). Several-fold higher urea transport capacities across the tonoplast membrane than across the plasma membrane have been reported for tobacco and wheat, and the mercury sensitivity of this transport process pointed to an involvement of aquaporins (Tyerman et al., 1999), most likely represented by homologs of the TIP subfamily of aquaporins (Liu et al., 2003b). With regard to the size of the vacuole relative to the cytoplasm, such a high vacuolar loading capacity could then create an intracellular sink that withdraws a large amount of urea from urease-mediated degradation and might finally increase the driving force for AtDUR3-dependent and -independent urea transport across the plasma membrane.

Physiological role of AtDUR3-mediated urea transport in Arabidopsis roots

Transgenic plants expressing an AtDUR3-promoter:GFP fusion construct showed high promoter activity in rhizodermal cells including root hairs as well as in cortical cells of mature root hair zones (Figure 3). Promoter activity became more restricted to outer root cells in younger root zones but was absent in the root apex. This nitrogen deficiency-enhanced fluorescence was supported by increasing mRNA levels of AtDUR3 in nitrogen-deficient roots (Figure 4; Liu et al., 2003a). Such a cell-type-specific expression pattern was reminiscent of that found for high-affinity ammonium and nitrate transporters which also contribute to import of nitrogen from the external medium (Guo et al., 2001; Loquéet al., 2006; Nazoa et al., 2003). Interestingly, AtDUR3 promoter activity was also observed in the stele of nitrogen-deficient roots (Figure 3b,d), most likely reflecting developing xylem vessels or underlying xylem parenchyma cells. Indeed, earlier studies reported that urea was translocated in the xylem sap when urea was supplied as a dominant nitrogen form or when nickel deficiency prevented urease from degrading urea prior to translocation (Gerendás et al., 1998; Hine and Sprent, 1988). Thus, xylem-associated AtDUR3 expression might reflect an involvement of the transporter in xylem loading or in retrieval of urea to xylem parenchyma cells.

A quantitative determination of the uptake capacity of AtDUR3 was obtained by influx studies using 15N-labeled urea. Concentration-dependent influx of urea in atdur3-1 and atdur3-3 showed a linear concentration dependency (Figure 6a), which is typical for channel-mediated urea transport, as has been demonstrated for AtTIP2;1 when expressed in oocytes (Liu et al., 2003b). In contrast, urea influx in wild-type plants was steeply elevated relative to the increase in external urea supply at concentrations below 50 μm, reaching up to 10 times the activity of atdur3. Between 50 and 200 μm urea wild-type plants showed no further significant increase in urea influx (Figure 6a). Taking into account the fact that urea influx in wild-type and atdur3-1 plants showed the same concentration-dependent increase at higher urea supply (Figure 6b), this study demonstrated that AtDUR3 is the major high-affinity urea transporter in Arabidopsis roots. Subtracting the urea influx in atdur3-1 from that in wild-type plants allowed calculation of a substrate affinity of AtDUR3 of 4 μm. This value accords well with the Km value of 3 μm as determined by 14C-urea transport assays and electrophysiological studies in AtDUR3-expressing Xenopus oocytes (Liu et al., 2003a). Considering the average urea concentrations of below 70 μm in natural or agricultural soils (Cho et al., 1996; Gaudin et al., 1987; Mitamura et al., 2000a,b), this particularly high substrate affinity is very likely to allow saturation of AtDUR3-mediated substrate transport into root cells in most soils. The low Km of AtDUR3 may be seen as an adaptation to the ubiquitous occurrence of microbial ureases. These enzymes usually have an affinity for their substrate in the millimolar range (Dalal, 1985) and so are unable to completely deplete soil urea. We propose a role for AtDUR3 as a unique transport system in Arabidopsis, allowing the direct use of urea, a limited but highly valuable nitrogen source in soils.

Experimental procedures

Isolation of T-DNA insertion lines and growth test on agar plates

The T-DNA insertion lines for AtDUR3 SALK_042649 (atdur3-1) and SALK_036318 (atdur3-3) were obtained from the Salk Institute collections, selfed and selected for homozygosity of the T-DNA insertion by PCR using primers for the left border of the insert (5′-GCGTGGACCGCTTGCTG-3′) and for the AtDUR3 gene (5′-GGAAGAAACGTTAAGACAGGA-3′). The PCR products from the T-DNA insertion lines were cloned and sequenced to confirm the positions of the insertions. Homozygous lines were analyzed by RNA gel blots.

For the growth test (Figure 1c), modified half-strength Murashige and Skoog (MS) medium (Murashige and Skoog, 1962) without nitrogen was supplemented with 1 μm NiSO4 and 50 μm KNO3. Plates containing no further nitrogen served as a negative control. Either 500 and 5000 μm NH4NO3 or 50, 500, 1000 and 5000 μm urea were added as nitrogen sources. Columbia-0 and atdur3 insertion lines were cultured for 3 weeks in a growth chamber (Percival, under a 10-h/22°C light and 14-h/19°C dark regime.

Promoter:GFP analysis

For the AtDUR3-promoter:GFP fusion, 1046 bp of the genomic region of AtDUR3 upstream of the translation initiation site was amplified by PCR using Pfu Turbo DNA polymerase (Stratagene, and the DNA sequence was verified. The gene-specific primers DUR3pro5 (5′-AAAAGCTTAAGGTAAAGAAAGGATACTTGTA-3′) and DUR3pro3 (5′-AAACCATGGTTCCTCTTCTTCTTTACGTTTT-3′) were used to generate a HindIII restriction site at the 5′-end, and a NcoI restriction site at the 3′-end, respectively. The HindIII–NcoI fragment of the AtDUR3 promoter fragment was used to replace the CaMV 35S promoter in CaMV35S–SGFP–TYG–nos (Chiu et al., 1996) and directly fused with sGFP. Then the AtDUR3 promoter–sGFP–nos gene cassette was subcloned into pGreen0029 (Hellens et al., 2000) as a HindIII–EcoRI fragment. Agrobacterium tumefaciens (strain GV3101:pMP90)-mediated Arabidopsis transformation was described previously (Loquéet al., 2006). Nitrogen-deficient medium for microscopic observation of promoter:GFP plants (Figure 3) was prepared by replacing nitrate in MGRL medium (Fujiwara et al., 1992) with an equivalent of chloride salts (Inaba et al., 1994). Fluorescence of GFP in transgenic plants was observed under an inverted fluorescence microscope equipped with an apotome (Zeiss Axiovert 200 M,

Whole-mount immunohistochemistry

The procedure for whole-mount preparations was modified from Lauber et al. (1997). Fourteen-day-old Col-0 and atdur3-1 roots were fixed in 4% paraformaldehyde dissolved in microtubule-stabilizing buffer [MTSB; 50 mm 1,4-piperazinediethanesulfonic acid (PIPES), 5 mm EGTA, 5 mm MgSO4, pH 6.9–7.0] at room temperature for 30 min and washed 3 times in phosphate-buffered saline buffer (PBS; 0.02 m sodium phosphate buffer with 0.15 m sodium chloride, pH 7.4) and two times in ultra-pure water. Roots were placed onto silane-coated microscopic slides (Histo Bond, Marienfield, with a drop of water. Coverslips were removed after dipping the slides into liquid nitrogen. After drying for 1 h, specimens were rehydrated in PBS for 10 min. Cell walls were partially digested with 2% driselase (Sigma, for 30 min. The plasma membrane was permeabilized with 0.4% Nonident P40 in 10% DMSO–PBS for 1 h. Non-specific interactions of antiserum were blocked with 1% BSA in PBS overnight, and antisera were diluted in 3% BSA in PBS and incubated overnight. Primary antiserum raised against AtDUR3 was used at the dilution of 1:500. Cy3-conjugated anti-rabbit-IgG was employed as a secondary antibody (Dianova; at a dilution of 1:200. Apotome scanning of specimens was performed with an inverted fluorescence microscope (Zeiss;, equipped with appropriate filters for Cy3, using the AxioVision software (version 4.5; Zeiss).

Hydroponic plant culture

Arabidopsis thaliana seeds were germinated in the dark for 4 days and cultured on rockwool moistened with tap water. After 1 week, tap water was replaced by half-strength nutrient solution containing 1 mm KH2PO4, 1 mm MgSO4, 250 μm K2SO4, 250 μm CaCl2, 100 μm Na–Fe–EDTA, 50 μm KCl, 30 μm H3BO3, 5 μm MnSO4, 1 μm ZnSO4, 1 μm CuSO4, and 1 μm NaMoO4, pH adjusted to 5.8 by KOH. Nitrogen was supplied as 2 mm NH4NO3. To suppress exogenous degradation of urea by urease liberated from decaying root cells, urea was supplied with 75 μg l−1 of the urease inhibitor phenylphosphorodiamidate (PPD; Martens and Bremner, 1984; Pedrazzini et al., 1987). The nutrient solution was renewed once a week during the first 3 weeks, twice in the fourth week and every 3 days for the following weeks. Plants were grown hydroponically under non-sterile conditions in a growth cabinet under the following conditions: 10 h/14 h light/dark; light intensity 280 μmol m−2 sec−1; temperature 22°C/18°C and 60% humidity.

RNA gel blot analysis

Total RNA was isolated by phenol-guanidine extraction followed by lithium chloride precipitation according to Logeman et al. (1987) or by extraction with TRIzol (Invitrogen, following the manufacturer’s protocol. Total RNA (10–20 μg per lane) was separated by electrophoresis on 3-(N-morpholino) propanesulfonic acid (MOPS)-formaldehyde agarose gels, blotted onto Hybond-N+ nylon membranes (Amersham, and cross-linked to the membrane by incubation at 80°C for 2 h. The coding sequence of AtDUR3 was used as a probe for hybridization to total RNA. Hybridization to a randomly primed 32P-radiolabeled probe was performed at 42°C in 50% (v/v) formamide, 1% (w/v) sarcosyl, 5× SSC and 100 μg ml−1 yeast t-RNA. Membranes were washed at 42°C once in 2× SSC, 0.1% (w/v) SDS for 40 min and once in 0.2× SSC, 0.1% (w/v) SDS for 40 min. Ethidium bromide-stained gels were used as the RNA loading control.

15N influx analysis

Urea influx measurements in plant roots were conducted after rinsing the roots in 1 mm CaSO4 solution for 1 min, followed by incubation for 10 min in nutrient solution containing different concentrations of 15N-labeled urea (95 at.%15N) as the sole nitrogen source. After a final rinse in 1 mm CaSO4 solution, roots and shoots were separated and stored at −70°C before freeze-drying. Each sample was ground and 1.5–2.5 mg sample powder was used for 15N determination by mass spectrometry (Finnigan; Values obtained for concentration-dependent urea influxes up to 200 μm urea were directly fitted to the Michaelis–Menten equation. Uptake experiments were repeated two or three times, and representative results are shown.

Preparation of microsomal and plasma membrane fractions

Fresh root tissue was ground in a buffer containing 250 mm 2-amino-2-(hydroxymethyl)-1,3-propanediol (TRIS)–HCl (pH 8.5), 330 mm sucrose, 25 mm EDTA, 5 mmβ-mercaptoethanol (β-ME), 2 mm dithiothreitol (DTT) and 1 mm phenylmethylsulfonyl fluoride (PMSF). Homogenates were centrifuged at 10 000 g for 15 min. Supernatants were filtered through nylon mesh (58 μm) and centrifuged at 100 000 g for 30 min to pellet microsomal membrane fractions. The pellet was resuspended in conservation buffer (5 mm Bis-TRIS propane-MES, pH 6.5, 250 mm sorbitol, 20% w/v glycerol, 1 mm DTT, 2 mm PMSF) and gently homogenized in a potter, as previously described in Loquéet al. (2006). Plasma membrane fractions were prepared by aqueous two-phase partitioning based on Larsson et al. (1987). A microsomal pellet was resuspended in microsomal buffer (5 mm KH2PO4, pH 7.8, 330 mm sucrose) and added to dextran-polyethylene glycol buffer (6.4% dextran T-500, 6.4% PEG 3350, 5 mm KH2PO4, 3 mm KCl and 330 mm sucrose). The two phases were mixed and centrifuged at 1500 g for 5 min. Upper and lower phases were collected and re-partitioned twice with fresh washing buffer (5 mm KH2PO4, pH 7.8, 330 mm sucrose, 1 mm PMSF). The upper and lower phases were diluted with washing buffer and centrifuged at 100 000 g for 60 min to pellet the membranes, respectively. The pellet was re-suspended in conservation buffer and gently homogenized in a potter. Protein concentrations were determined using a Bradford protein assay kit (Bio-Rad, using BSA as a standard.

Protein gel blot analysis

Polyclonal antibodies were raised against an oligopeptide representing the C-terminal 14 amino acids of AtDUR3 (n-LLELEKTKKNDEEG-c). The antiserum was affinity-purified using a nitrocellulose membrane for peptide binding as described in Ludewig et al. (2003). Sodium dodecyl sulfate-PAGE and protein gel blot analysis were performed as described previously (Loquéet al., 2006). Antiserum raised against AtDUR3 was diluted in blocking solution at 1:5000 for Figure 2(a) and 1:10 000 for Figure 2(b,c), respectively. The dilution of the antibodies against AHA2 and DET3 has been described in Yuan et al. (2007).

Determination of urea concentrations

Root urea concentrations were determined based on a colorimetric reaction used by Kyllingsbæk (1975). Approximately 50 mg of freeze-dried plant tissues was milled and suspended in 1 ml of cold 10 mm formic acid. After centrifugation at 16000 g and 4°C for 15 min, 30 μl of the supernatant were incubated with 1 ml of a color development reagent (4.6 mm diacetylmonoxime, 1.28 mm thiosemicarbazide, 6.6% H2SO4, 14.6 μm ferric chloride hexahydrate and 0.006% orthophosphoric acid) at 99°C for 15 min, and then cooled down at 4°C for 5 min. Absorbance at 540 nm was measured with a photometer. The ureides allantoin, ornithine, arginine, and uric acid did not interfere with urea determinations, but other ureides were not tested.


We thank Bernhard Bauer, Elke Dachtler and Susanne Reiner, University of Hohenheim, for their skillful technical support, as well as Dr Hideki Takahashi, RIKEN Plant Science Center, Japan, and Dr Christina Morris, Adelaide, for critically reading the manuscript. We are grateful to Dr Karin Schumacher, ZMBP Tuebingen, for kindly providing the AHA2 and DET3 antibodies. This study was supported by a grant from the Deutsche Forschungsgemeinschaft DFG, Bonn, Germany, to NvW (WI1728/2).