Gravitropism plays a critical role in plant growth and development, plant stability and acclimation to changes in water and nutrient availability. Ammonium (NH4+) is well known to have profound effects on root growth, but its impacts on gravitropism are poorly understood.
To determine which genes are essential for the maintenance of root gravitropism under NH4+ stress, we isolated and identified an NH4+-sensitive mutant, gsa-1 (gravitropism sensitive to ammonium-1), in Arabidopsis thaliana, using an agar plate root reorientation assay.
We found that, under NH4+ stress, gsa-1 displayed increased loss of root gravitropism. Gene cloning and sequencing revealed that gsa-1 contains a G to C transversion mutation at the highly conserved 5′-GT splice position of intron 10 of ARG1 (ALTERED RESPONSE TO GRAVITY1), known to participate in the transduction of the root gravity signal. Genetic complement tests established the locus of GSA-1/ARG1 and its role in resistance to NH4+ inhibition on root gravitropism. GSA-1/ARG1 is required for normal AUX1 expression and basipetal auxin transport in root apices. In addition, PIN-FORMED2 (PIN2) is proposed as a target in the reduction of root gravitropism under NH4+ stress, a response which can be antagonized by the GSA-1/ARG1-dependent pathway.
These results suggest that GSA-1/ARG1 protects root gravitropism in Arabidopsis thaliana under ammonium stress.
Gravitropism ensures the growth of plant organs along a specific vector relative to gravity, and dictates the upward growth of shoots and that of roots down into the soil. Root gravitropism plays an important role in determining the distribution of the root system in the soil, and is therefore critical to the physical anchorage of the plant, as well as to water and nutrient acquisition (Forde & Lorenzo, 2001; Perrin et al.,2005). Changes in orientation relative to gravity (gravistimulation) induce root bending towards the original growth direction, a process that can be conceptually divided into four successive steps: gravity perception, signal transduction, signal transmission and curvature response (Perrin et al., 2005). Columella cells containing sedimentable amyloplasts in the root cap are the principal gravity-perceptive sites in roots (Caspar & Pickard, 1989; Kiss et al., 1989; Blancaflor et al., 1998). The sedimentation of amyloplasts in the root cap has long been thought to trigger a signal transduction pathway that promotes the development of a lateral gradient of auxin, which is then transported to the elongation zone (EZ), where it leads to differential cellular elongation on opposite flanks of the corresponding EZ (Chen et al., 2002).
The formation and maintenance of auxin gradients are dependent on polar auxin transport, mediated by special transporters, including auxin influx carriers (in particular AUX1) and efflux facilitators (PIN-FORMED3 (PIN3) and homologous protein PIN2/AGR1 (AGRAVITROPIC1)/WAV6 (WAVY ROOTS 6)/EIR1 (ETHYLENE INSENSITIVE ROOT 1); Ge et al., 2010). For example, AUX1 encodes an auxin influx-mediating transmembrane protein, localized in the stele, the apical side of protophloem cells, columella, epidermis and lateral root cap tissues (Swarup et al., 2001). By contrast, PINs encode auxin efflux-mediating proteins, and polar PIN localization directs auxin flow (Wisniewska et al., 2006). A relocation of PIN3 within statocytes from a symmetrical distribution at the plasma membrane is believed to represent the initial step in the establishment of the lateral auxin gradient on gravistimulation (Friml et al., 2002; Ottenschläger et al., 2003; Harrison & Masson, 2008). Following this, the resulting lateral auxin gradient is shifted basipetally through the combined actions of AUX1 and PIN2 via lateral root cap and epidermal cells towards EZ (Friml, 2003; Swarup et al., 2005). AUX1 is present in the same cells as PIN3 and PIN2, and probably facilitates the uptake of auxin into the lateral root cap and the epidermal region, and PIN2 is believed to mediate its directional translocation towards EZ (Friml, 2003). However, differential cellular elongation on opposite flanks of the central EZ induced by the lateral auxin gradient may be responsible for only part of the gravitropic curvature (Chen et al., 2002). Recent findings have suggested the involvement of the transition zone (TZ), also known as the distal elongation zone (DEZ), as a secondary site/mechanism of gravity sensing for root gravitropism, which may be independent of the auxin gradient (Wolverton et al., 2002; Chavarría-Krauser et al., 2008; Baluška et al., 2010). Roots of pin3 mutants lack the bending response in the elongation region, but bending is not affected in TZ (Chavarría-Krauser et al., 2008; Baluška et al., 2010).
Given that the appropriate distribution of roots within the soil greatly affects plant survival, roots have evolved to sense adverse environmental cues and to modulate their growth direction through a variety of pathways. For example, moisture gradients and water stress can cause the desensitization of gravitropism in Arabidopsis by the degradation of amyloplasts in root columella cells, allowing roots to exhibit positive hydrotropism (Takahashi et al., 2003). Similarly, roots exposed to salt stress show inhibited gravitropism, which permits the active avoidance of stressed regions (Li & Zhang, 2008; Sun et al., 2008). In addition, the availability of phosphorus has been shown to regulate the root configuration of legumes by altering the growth angle of basal roots, so as to facilitate improved acquisition of phosphorus from the soil (Bonser et al., 1996; Liao et al., 2001). Similarly, reductions in external potassium have been observed to trigger agravitropic growth in Arabidopsis, so that roots grow away from potassium-impoverished regions, which may well represent a mechanism by which plants respond to mineral deficiencies in general (Vicente-Agullo et al., 2004). Furthermore, ammonium (NH4+), both a major nitrogen source and common toxicant (Kronzucker et al., 1997), frequently produces the inhibition of root growth and lateral root formation (Gerendás et al., 1997; Britto & Kronzucker, 2002; Qin et al., 2008; Li et al., 2010, 2011a,b, 2012, 2013; Kempinski et al., 2011), but has also been shown to affect the root gravitropism response (Zou et al., 2012).
Here, we report a novel Arabidopsis thaliana mutant, gsa-1 (gravitropism sensitive to ammonium-1), which displays reduced root gravitropism in response to NH4+ stress. Gene cloning shows gsa-1 to be allelic to ARG1 (ALTERED RESPONSE TO GRAVITY1), which is required for the establishment of the lateral auxin gradient across the root cap following gravistimulation (Boonsirichai et al., 2003; Harrison & Masson, 2008).Our results further demonstrate that the disruption of GSA-1/ARG1 can reduce basipetal auxin transport and the expression of AUX1 protein in root apices. Moreover, we demonstrate the involvement of PIN2 in the NH4+ regulation of the gravitropic response.
Materials and Methods
Plants and growth conditions
The plant material used in this work included the Columbia (Col-0) and Landsberg erecta (Ler) ecotypes of Arabidopsis thaliana (L.) Heynh, T-DNA-transformed mutants derived from Col-0, and the mutants arg1-3 (SALK_024542C), eir1-1 (Roman et al., 1995) and aux1-22 (Roman et al., 1995); the transgenic DR5::GUS (Ulmasov et al., 1997) was in the Col-0 background, proPIN2::PIN2-GFP in eir1-1 (Blilou et al., 2005) and proAUX1::AUX1-YFP in aux1-22 (Swarup et al., 2004). gsa-1 plants carrying the proAUX1::AUX1-YFP and proPIN2::PIN2-GFP constructs were derived from crosses between gsa-1 and the corresponding constructs of the transformed plants, respectively, and homozygous plants for both gsa-1 and the proAUX1::AUX1-YFP or proPIN2::PIN2-GFP insertion were used.
The double mutants gsa-1aux1-22 and gsa-1eir1-1, carrying both the gsa-1 and either the aux1-22 or eir1-1 mutation in the homozygous state, were obtained by crossing single mutants and selfing the corresponding F1 progeny. Plants homozygous for aux1-22 and eir1-1 were identified by germination on normal agar growth medium, with the addition of 0.1 μM 2,4-dichlorophenoxyacetic acid (2,4-D) and 1 μM 1-aminocyclopropane-1-carboxylate (ACC), respectively (Luschnig et al., 1998; Marchant et al., 1999). Several rosette leaves were excised from a single 2,4-D- or ACC-resistant F2 plant and prepared for total RNA extraction and reverse transcription-polymerase chain reaction (RT-PCR) amplification using the primers F2R2, as described below for the identification of gsa-1. The homozygous gsa-1 mutation displayed larger RT-PCR amplification products than did Col-0.
After being surface sterilized and cold treated at 4°C for 2 d, the seeds were sown on normal Arabidopsis thaliana growth medium, as described by Li et al. (2010). The nutrient medium contained 2 mM KH2PO4, 5 mM NaNO3, 2 mM MgSO4, 1 mM CaCl2, 0.1 mM Fe-EDTA, 50 μM H3BO3, 12 μM MnSO4, 1 μM ZnCl2, 1 μM CuSO4, 0.2 μM Na2MoO4, 1% (w/v) sucrose, 0.5 g l−1 MES and 0.8% (w/v) agar (adjusted to pH 5.7 with 1 M NaOH). The culture plates were placed vertically in a growth chamber at 23 ± 1°C under a light intensity of 100 μmol photons m−2 s−1, with a 16 h : 8 h light : dark cycle. Five-day-old seedlings germinated on normal growth medium, with relatively straight root tips and c. 1.5 cm in length, were selected for gravity stimulation experiments. The NH4+ treatment medium was prepared by the addition of varying concentrations of (NH4)2SO4 to the normal growth medium. Ion effects were analyzed using NH4Cl (with the same NH4+ concentration as in (NH4)2SO4), K2SO4 (with the same SO42− concentration) and KNO3 (with the same N concentration). For tests of general osmotic substitution of (NH4)2SO4, different concentrations of mannitol were used.
Five-day-old seedlings of similar size were transferred to new agar plates containing the appropriate treatments. Roots were placed vertically in rows, after recording the initial positions of the root tips, the plates were rotated by 90° and placed vertically for gravistimulation under different concentrations of NH4+ (0, 10, 20 and 30 mM (NH4)2SO4) in a cultivation chamber at time zero. Digital images of seedling growth were captured at regular, specified time points (as defined in the text) following gravistimulation with a Canon G7 (Canon Inc., Tokyo, Japan). The root elongation and tip angles from the vertical were determined as described by Sun et al. (2008). Root elongation refers to the length of the primary root after transfer to treatment solution, whereas the gravitropic angle refers to the angle of the root tip relative to the gravity vector.
Isolation of the gsa-1 mutant
A total of 5160 T-DNA insertional mutant lines in the Col-0 background, as described previously (Zuo et al., 2000; Zhang et al., 2005; Li et al., 2012), were used in gsa-1 isolation. Five-day-old seedlings with roots c. 1.5 cm in length were transferred onto new medium supplemented with 30 mM (NH4)2SO4, arranged in rows and then gravistimulated to screen for NH4+-sensitive phenotypes of root gravitropism. The gravitropic bending during the following 72 h was captured, and putative mutants with a reduced gravitropic response were selected and transferred to soil to grow to maturity and to self-fertilize. One of these putative mutants, called gsa-1, was selected for analysis. The homozygous M4 gsa-1 mutant was backcrossed against the parental wild-type (Col-0) another two times to remove the unlinked mutations caused by the mutagenesis.
Although rosettes and inflorescences of gsa-1 were similar to those of Col-0 in morphology, its root displayed a defect in gravitropism when grown on agar plates for 3 d after germination (DAG) (Supporting Information Fig. S1). F1 and F2 progenies derived from the crosses of Col-0 × gsa-1 and Ler × gsa-1 were grown on normal growth medium for 3 d; agravitropic individuals were scored and compared statistically (χ2 test). Segregation of the root gravitropic phenotype of crossed F2 progeny seedlings at 5 DAG on agar medium supplied with 30 mM (NH4)2SO4 was also analyzed and compared statistically.
Mapping of gsa-1
Mapping populations were created by crossing the gsa-1 mutant (Col-0) with wild-type plants of Ler, and their F1 progeny were propagated by self-fertilization. F3 progenies from individual F2 plants were subjected to a gravitropism assay to determine the corresponding gravitropic genotype of the F2 parents. Homozygous F2 progeny mutant plants were selected and used to map gsa-1 by bulked segregation analysis. The protocol of mapping has been described previously (Li et al., 2012). DNA was extracted from each pool and subjected to PCR amplification with InDel (Insertion/Deletion) primers (Salathia et al., 2007). The candidate gene GSA-1 was linked with CER464751 (F: AAACCCCTTCCAGGATGAAC; R: ACGTTTTGAACCACCGCTAC), where a well-known agravitropic mutant, arg1, has been mapped previously (Sedbrook et al., 1999). Thus, the gsa-1 and arg1-3 mutants were crossed with each other for allelic testing.
Cloning of ARG1 genomic and cDNA sequences
DNA for PCR amplification was extracted according to the procedure of Weigel & Glazebrook (2002). Total RNA was extracted from roots of 7-d-old seedlings using the Trizol reagent (Invitrogen). One microgram of DNase-treated total RNA was used as a template for first-strand cDNA synthesis with reverse transcriptase M-MLV (TaKaRa Biochemicals, Dalian, China) and an oligo(dT) primer.
PCR and RT-PCR amplification were performed with Col-0 and gsa-1 genomic DNA and cDNA, respectively, using the primers F1/R1 (5′-CGAGAAGATGAGCGCGAAAAAGCTTGAA-3′/5′-TATATCATCAATCCACCATCACAAT-3′) (Harrison & Masson, 2008) and F2/R2 (5′-TCAACCAAGCTTCTTATCAGA-3′/5′-GCTTTGGCGCGGTTTCAAGA-3′), respectively, for the candidate gene, ARG1. PCR and RT-PCR amplification were performed in a 20-μl volume containing 10–100 ng μl−1 genomic DNA or cDNA (2 μl), 10 μM of each primer (0.4 μl), 2.5 mM deoxynucleoside triphosphates (dNTPs) (1.6 μl), 25 mM Mg2+ (1.2 μl), 10× PCR buffer (2.0 μl) and 0.2 units of Taq DNA polymerase (TaKaRa). The conditions for the PCR program used in the cloning of the ARG1 sequence were: 94°C for 5 min, followed by 30 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 3 min, and final extension at 72°C for 7 min. The following program was used for the RT-PCR amplification of ARG1: 94°C for 5 min, followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 2 min, and final extension at 72°C for 7 min. The PCR and RT-PCR products were cloned into a pMD®18-T vector and sequenced.
Complementation of gsa-1
To verify that the gravitropic defect of gsa-1 was associated with the mutations in ARG1, the ARG1 coding region, including the corresponding untranslated region (UTR) sequences, was amplified from Col-0 and gsa-1 cDNA, respectively, using the primer pairs 5′-CGGCTCTAGATTTTTTTCCTCGATCTTCTTCTTC-3′ and 5′-GGCGAGCTCGAAGCAAGTCTGGTATTATAATT-3′ with the XbaI and SacI restriction sites in italic, and cloned into the XbaI–SacI-digested pBI121 binary vector, under the control of the cauliflower mosaic virus (CaMV) 35S promoter and nopaline synthase (NOS) terminator sequences to create the transformation construct. These constructs were transformed into Agrobacterium, strain C58, and then introduced into the gsa-1 mutants by the floral dip method for the complementation assay (Clough & Bent, 1998), and the defect-carrying GSA-1 cDNA and pBI21 vector were transformed as a negative control. Transformants were identified by kanamycin resistance; eight, two and five independent homozygous and genetically stable transgenic lines were developed for wild-type GSA-1 cDNA/gsa-1, defective GSA-1 cDNA/gsa-1 and pBI21 vector/gsa-1, respectively. All of the wild-type GSA-1 cDNA/gsa-1 transgenic lines, but not the defective GSA-1 cDNA/gsa-1 and pBI21 vector/gsa-1 transgenic lines, complemented the gsa-1 phenotype, although only one representative from each is presented in this study.
Construction of proAtARG1:GUS fusion genes
To investigate the expression of GSA-1/ARG1, the promoter of AtGSA-1/AtARG1 was fused to a β-glucuronidase (GUS) reporter gene, and the recombinant was introduced into Col-0 to produce proAtARG1:GUS transgenic plants: DNA fragments covering the 5′ flanking regions, from −1405 to −1 of the ARG1 gene (At1 g68370), were amplified from Col-0 genomic DNA using the primers 5′-ACATAAGCTTTCACTCTCTTCCCCCTTCGTCCAA-3′ and 5′-GCGGCTCTAGACTTCTCGAAGAATTTGAAATCAC-3′, with the HindIII and XbaI sites in italic, and then inserted into a binary pBI121 vector, creating a recombinant transcription. The recombinant proAtARG1:GUS fusion gene was introduced into Col-0 plants.
Histochemical analyses of GUS gene enzyme activity of the transgenic plants containing the DR5::GUS reporter gene and the proAtARG1:GUS fusion construct were carried out according to Weigel & Glazebrook (2002). Histochemical staining, observation and measurement of amyloplasts in the columella cells of the root cap were performed as described by Takahashi et al. (2003). Images were obtained using an Olympus BX51 optical microscope equipped with differential interference contrast (DIC) for observation and an Olympus DP71 system for photography (Olympus Optical, Tokyo, Japan). The images shown are representative of at least 10 plants for each treatment, and the experiments were repeated at least twice. ProPIN2::PIN2-GFP and ProAUX1::AUX1-YFP were visualized using a Zeiss LSM710 confocal laser scanning microscope. The excitation wavelengths were 514 and 488 nm for yellow (YFP) and green (GFP) fluorescent proteins, respectively, and emission was detected at 520–550 nm. Images were taken under the same conditions and were representative of at least 10 individual plants from each treatment. Each experiment was repeated at least twice.
Basipetal 3[H]IAA transport
To measure basipetal auxin transport, 5-d-old seedlings were transferred to new agar plates with 0 or 30 mM (NH4)2SO4 for 1 d, and then the root tips were incubated with 1% agar blocks containing 100 nM 3H-labelled IAA (American Radiolabeled Chemical, St Louis, MO, USA) for 5 h in the dark. The apical 2 mm of the roots were discarded, and the apical 5-mm sections of the remaining roots were excised for radioactivity counting, as described by Li et al. (2011a), and following the procedure by Lewis & Muday (2009). The results are pooled from three experiments of 10 seedlings per treatment.
Data were analyzed statistically using SPSS version 13.0 (SPSS, Chicago, IL, USA). One-way ANOVA with an LSD (least-significance difference) test was used for the analysis of differences in root growth and gravitropism following NH4+ treatments. Sigma Plot 13.0 was used for the generation of graphs and Photoshop for photocomposition.
Isolation of gsa-1 and genetic analysis
Our time-dependence results show that 30 mM (NH4)2SO4 inhibits significantly the gravitropic bending of Col-0 at shorter exposure times of < 24 h. However, at prolonged exposure times (longer than 24 h), no significant difference was found in root gravitropic angles between treatments and controls (Zou et al., 2012). On the basis of these findings, and using the experimental system shown in Fig. 1(a), we transferred 5-d-old seedlings, screened 5160 T-DNA insertion mutants and isolated a mutant with distinct NH4+ sensitivity in the root gravitropic response, which we refer to as gsa-1. In contrast with Col-0, the differences in root gravitropic angle between NH4+-treated plants and gsa-1 controls did not diminish with time, and 30 mM (NH4)2SO4 inhibited significantly the root gravitropic response of gsa-1 at each time point tested (Fig. 1b).
gsa-1 displayed reduced root gravitropic curvature dynamics relative to those of Col-0 under NH4+-free control conditions (i.e. on normal growth medium, Fig. 1b), indicating that gsa-1 represents a root-sensing or gravitropism-associated mutant. In observations of the process of seed germination, gsa-1 was seen to grow in random directions and without the regular wavy root growth habit relative to Col-0. These differences were more apparent in 3-d-old seedlings (Fig. S1a,b). The F1 seedlings generated from gsa-1 crossed with wild-type Col-0 displayed a similar gravitropism response to the wild-type, and self-fertilized F2 seedlings revealed a 3 : 1 segregation ratio of wild-type/gsa-1 (Table 1; Fig. S1c). Therefore, the gsa-1 mutant was caused by a single recessive nuclear mutation.
Table 1. Genetic analysis of Arabidopsis thaliana mutant gsa-1
Wild-type phenotype or mutant phenotype was determined for seedlings at 3 d after germination (DAG).
The calculated value was based on the expected ratio of three wild-type seedlings to one mutant seedling (P >0.05).
Col-0 × gsa-1
Characteristics of root gravitropism in Col-0 and gsa-1 in response to ammonium
To better understand the gravitropic response of gsa-1 roots to NH4+, we monitored root elongation and gravitropic curvature over time in Col-0 and gsa-1 in response to different concentrations of NH4+ (10, 20 and 30 mM (NH4)2SO4). In the short term (< 12 h), all concentrations of (NH4)2SO4 applied reduced significantly root gravitropism, but the effect decreased in Col-0 with extended (24 h and beyond) treatment (Fig. 2a). After 48 h of gravistimulation, 10 mM (NH4)2SO4 even increased the maximum curvature of treated roots (displaying smaller gravitropic angle) in contrast with controls. Roots treated with 20 and 30 mM (NH4)2SO4 displayed no significant difference in root gravitropic angles relative to those of controls at 48/72 h. After gravistimulation for 48 h, 10 mM (NH4)2SO4 had no significant impact on the root gravitropic angle in gsa-1, whereas 20 and 30 mM (NH4)2SO4 inhibited significantly gsa-1 root gravitropic bending, an effect not seen in Col-0 (Fig. 2b). These results show that the root gravitropic response of gsa-1 is more sensitive to NH4+ relative to that of Col-0. Nevertheless, the tracking of root growth over time at all concentrations of (NH4)2SO4 showed the effect of NH4+ on root elongation to be similar between Col-0 and gsa-1 (Fig. 2c,d). After a 3-d treatment with 10, 20 and 30 mM (NH4)2SO4, the mean root elongation values of Col-0 and gsa-1 were reduced to 71.5%, 47.4% and 30.1%, and 67.9%, 45.6% and 24.7%, respectively (Fig. 2c,d). The lack of correlation between the effects of NH4+ on root gravitropic angle and root elongation illustrates that the influence of NH4+ on the gravitropic response of gsa-1 is not a secondary effect of its inhibition of root elongation.
Given that NH4+ applications were in the form of (NH4)2SO4, it was necessary to ascertain whether the observed effects were indeed attributable to the NH4+ ion, and not the counterion (SO42−), highly concentrated N or the osmotic strength of the applied salts. To address this issue, NH4Cl applications with identical concentrations of NH4+, K2SO4 applications with identical concentrations of SO42−, KNO3 applications with identical concentrations of N and isosmotic applications of mannitol were substituted for (NH4)2SO4 in the treatment medium. gsa-1 roots displayed largely indistinguishable gravitropic kinetics under NH4Cl as under (NH4)2SO4 (Figs S2a, 2a), whereas both K2SO4 and KNO3 at the indicated concentrations had little impact on gsa-1 root gravitropism (Fig. S2b,c). Further, mannitol produced no inhibitory effect on root gravitropism at the concentrations applied (Fig. S2d). Thus, the action of (NH4)2SO4 on gsa-1 root gravitropism could be attributed to NH4+.
Identification of GSA-1
As T-DNA insertion loci isolated by thermal asymmetric interlaced-PCR (TAIL-PCR) were not associated with the NH4+ response phenotype, a map-based cloning strategy was pursued to isolate the mutated gene. With preliminary positional cloning analysis, gsa-1 showed linkage to CER464751 (at base pair 26 628 510 of chromosome 1), where a well-known mutant, arg1 (altered response to gravity1), displaying a similar gravitropism phenotype to that of gsa-1, has been identified. Genetic analysis by crossing gsa-1 and arg1-3 containing a T-DNA disruption of the ARG1 gene demonstrated that gsa-1 failed to complement arg1-3 (Fig. S3), indicating that the two mutants are allelic to each other. The genomic sequence displayed a point mutation of G to C at base pair 2274 of AtARG1 (the universally conserved 5′ GT splice site of intron 10 of AtARG1) (Fig. 3a), thus abolishing the GU splice site consensus, leading to the retention of base pair 172 in intron 10. Semi-quantitative RT-PCR expression analysis of total mRNA extracted from 7-d-old seedlings of the indicated genotype, using the ARG1 primers F1/R1, as described by Harrison & Masson (2008), detected a larger AtARG1 cDNA product in gsa-1 relative to Col-0, which was further different from the null allele arg1-3 (Fig. 3b). Primers of F2/R2 across intron 10 of ARG1, as shown in Fig. 3(a), were further used to amplify the genomic DNA and mRNA isolated from Col-0 and gsa-1, respectively. PCR and RT-PCR products of gsa-1 had the same sizes as the Col-0 genomic DNA amplification products, which were larger than the RT-PCR product of Col-0 (Fig. 3c). Corresponding sequencing confirmed the mutation of G to C at the 5′ conserved GT splice site of intron 10. It was inferred that the disruption of GU caused a frameshift and premature stop codons in the C-terminus of the gsa-1 mutant, and thus destroyed the predicted coiled-coil region of ARG1 (Fig. 3a). Transformed gsa-1 plants expressing wild-type AtARG1 cDNA under the control of the CaMV 35S promoter oriented in a similar manner to Col-0, whereas those carrying the defective AtARG1 cDNA or pBI121 construct could not rescue the agravitropic phenotype of gsa-1 (Figs S4, 3d). To further determine the role of GSA-1/ARG1 in NH4+-induced root agravitropism, we examined the gravitropic dynamics in Col-0, gsa-1, T3 generation homozygous lines of transgenic Arabidopsis thaliana wild-type GSA-1 cDNA/gsa-1, defective GSA-1 cDNA/gsa-1 and pBI21 vector/gsa-1. Transgenic lines of wild-type GSA-1 cDNA/gsa-1 displayed gravitropic curvature kinetics similar to those of Col-0 under NH4+ stress, whereas transgenic lines of defective GSA-1 cDNA/gsa-1 and pBI21 vector/gsa-1 were similar to gsa-1 (Fig. 3e). These results indicate that the agravitropism phenotype of gsa-1 is indeed caused by the mutation of GSA-1/ARG1, and that GSA-1/ARG1 plays an important role in the resistance to NH4+-induced reduction of root gravitropism.
gsa-1 displays similar amyloplast degradation kinetics to Col-0 under ammonium stress
As reported previously by Harrison & Masson (2008), the mutation of GSA-1/ARG1 results in larger amyloplast accumulation in roots relative to Col-0, verified by starch staining (Fig. 4a,b), probably by increasing cells with columella identity in the root cap (Harrison & Masson, 2008). As amyloplast degradation is well known to be involved in a variety of environmental stress responses (Takahashi et al., 2003), we proceeded to test whether the sensitivity of gsa-1 root gravitropism to NH4+ was associated with accelerated degradation of amyloplasts under NH4+ stress. At shorter NH4+ exposure times (< 24 h), no rapid degradation of amyloplasts was observed in the columella cells of Col-0. However, starch was gradually degraded with increasing treatment times and NH4+ concentrations, such as a 72-h treatment with 20 mM (NH4)2SO4 or > 48-h treatment with 30 mM (NH4)2SO4 (Figs 4c, S5). Interestingly, the gsa-1 mutant showed a similar trend in amyloplast degradation as that seen in Col-0 under NH4+ stress (Figs 4d, S5). These results show that NH4+ treatments affect the levels of starch in columella cells, but that an altered degradation rate of amyloplasts is not involved in the NH4+-induced disruption of root gravitropism.
Ammonium increases the expression of GSA-1/ARG1
The expression of GSA-1/ARG1 was studied with the GUS reporter gene under the control of the 1405-bp GSA-1/ARG1 promoter (ProARG1:GUS). Histochemical analysis revealed GUS staining throughout the entire seedling, including the young leaves, vascular tissues of cotyledons, shoot apical meristem, hypocotyl and roots (Fig. 5a). The expression of ProARG1:GUS increased with seedling age (day 0–day 3) under control conditions, and GUS staining in the root tip was further intensified with prolonged NH4+ treatment (Fig. 5b). These results indicate that the expression of the GSA-1/ARG1 gene is regulated by seedling age and is increasingly induced by extended exposure to elevated NH4+.
GSA-1/ARG1 regulates basipetal auxin transport in roots
DR5::GUS staining showed that, under vertical growth conditions, gsa-1 seedlings accumulated higher levels of auxin in the expanded domain of the root cap relative to the wild-type, and there was no detectable lateral redistribution of auxin on gravistimulation (Figs 6, S6). In the presence of NH4+, the development of the lateral auxin gradient was both delayed and prolonged in Col-0 (Fig. 6), whereas epidermal staining was evident further away from the root tip in gsa-1 after 24 h (Fig. 6).
Basipetal auxin transport of gsa-1 was assessed by direct measurement with radiolabeled 3[H]IAA. The mutation of ARG1 reduced basipetal auxin transport by 38.96% (P <0.05). However, there was no further reduction in basipetal auxin transport in gsa-1 roots with NH4+ treatment (Fig. 7). In the wild-type, treatment with 30 mM (NH4)2SO4 for 24 h decreased basipetal auxin transport by 28.68% (P >0.05). These results show that GSA-1/ARG1 is required for basipetal auxin transport in roots.
GSA-1/ARG1 regulates the expression of AUX1, but not PIN2, at the protein level
The appropriate expression and location of the auxin transporters AUX1 and PIN2 in lateral root caps and epidermal cells are required for normal basipetal auxin transport in roots (Friml, 2003). However, the expression and localization pattern of proAUX1::AUX1-YFP in lateral root caps of Col-0 were not affected under 30 mM (NH4)2SO4 stress on gravistimulation (Fig. 8a). By contrast, the fluorescence intensity of proPIN2::PIN2-GFP in whole-root apices of Col-0 decreased gradually with time, implying a relation of PIN2 during NH4+ stress (Fig. 8b).
To examine the distribution and levels of AUX1 and PIN2 proteins in the gsa-1 background, the proAUX1::AUX1-YFP and proPIN2::PIN2-GFP constructs were introduced into the gsa-1 mutant. The expression of proAUX1::AUX1-YFP resulted in a nearly undetectable AUX1-YFP fluorescence signal in the gsa-1 root apices under the same conditions as Col-0. After extending the master gain of the YFP channel from 679 to 1017, a normal localization of the YFP signal (similar to that of Col-0) was seen in phloem tissue, columella cells, epidermal cells and the lateral root cap of the mutant primary roots (Fig. 9a). These results indicate that the AUX1 protein is distributed normally in the mutant, except for reduced expression in root apices. Nevertheless, NH4+ treatments did not affect the localization and expression intensity of AUX1-YFP in mutant root apices any further on gravistimulation, although an overall reduced root length was observed (Fig. 9b). By contrast, the mutation of GSA-1/ARG1 had no effect on the expression and localization of PIN2 in root apices relative to Col-0 (Fig. 9c). Ammonium treatments decreased dramatically the PIN2-GFP fluorescence signal intensity in gsa-1 root apices, as was also seen in Col-0 (Fig. 9c), implying that the expression and localization pattern of PIN2 in root apices are independent of GSA-1/ARG1.
eir1-1 and aux1-22 enhance the gravitropic defect of gsa-1
To further explore the roles of AUX1 and PIN2 in NH4+-induced gsa-1 root gravitropism, we constructed double mutants of gsa-1aux1-22 and gsa-1eir1-1. Both aux1-22 and eir1-1 exacerbated the phenotype of gsa-1 for seedlings germinated on normal growth medium, making the double mutants gsa-1aux1-22 and gsa-1eir1-1 grow in random directions (Fig. 10a). Root gravitropic angles relative to the gravity vector were measured and displayed for 5-d-old seedlings, with the system shown in Fig. 10(b). It was found that the root gravitropic angles of Col-0, gsa-1, aux1-22 and eir1-1 were distributed within 60° (−15° to 45°), 210° (−105° to 105°), 360° and 150° (−45° to 105°), respectively. Both aux1-22 and eir1-1 showed a wider distribution of gsa-1 root gravitropic angles, with the double mutants gsa-1aux1-22 and gsa-1eir1-1 within 360° (Fig. 10c).
The root gravitropic kinetics of these mutants under NH4+ stress were also explored (Fig. 11). In contrast with the NH4+ effects on Col-0 and gsa-1 root gravitropism, NH4+ partially restored the root agravitropism of aux1-22. In addition, gsa-1aux1-22 roots had the same tendency of gravitropic kinetics as roots of the aux1-22 mutant line under NH4+ treatments. Ammonium also partially restored the agravitropism of eir1-1 and gsa-1eir1-1, but the gravitropic kinetics of gsa-1eir1-1 roots under NH4+ stress were not consistent with that scored for eir1-1 roots.
Much recent interest has centered around the mechanisms governing the root gravitropic response, in particular under environmental stress (Chavarría-Krauser et al., 2008; Li & Zhang, 2008; Sun et al., 2008; Rigas et al., 2012). With respect to the important environmental stress of excess NH4+ (Britto & Kronzucker, 2002), however, very little is known about the specific targets and pathways that lead to impaired gravitropism (Zou et al., 2012). To gain an insight into the mechanisms of the effects of NH4+ on root gravitropism, we employed a molecular genetics approach, based on a mutant screen for altered response to NH4+. Strongly enhanced NH4+ sensitivity of root gravitropism was found in gsa-1; this defect was rescued by transformation using wild-type GSA-1cDNA constructs, and wild-type GSA-1cDNA/gsa-1 lines displayed similar root gravitropic dynamics to wild-type Col-0 under NH4+ stress (Fig. 3d,e). These results suggest that GSA-1/ARG1 plays an important role in the NH4+-induced impairment of root gravitropism. It defines a novel genetic locus essential for NH4+ tolerance in Arabidopsis.
By examining the new mutant, we identified and characterized a new mutant allele in the Arabidopsis ARG1 gene. Distinguished from the T-DNA insertional null mutant arg1-3, the agravitropism phenotype of the gsa-1 allele was caused by a G to C mutation at the universally conserved 5′-GT dinucleotide splicing motif of intron 10, giving rise to aberrantly spliced mRNA (Fig. 3a–c). Previously, a mutation has been described at the aux1-22 allele, caused by a T to A mutation at the 5′-splice site of intron 5, thus abolishing the GT splice site consensus (Marchant & Bennett, 1998). With gsa-1, the mutant had a similar phenotype to that of the null mutant arg1-3. It is plausible that the larger mRNAs in gsa-1 were removed to prevent the accumulation of potentially harmful proteins. As a result, gsa-1 also presented as a null mutant at the protein level. The selective degradation of nonsense-containing transcripts has been reported in yeast, in the nematode Caenorhabiditis elegans and in plants (Marchant & Bennett, 1998). Further, the C-terminal coiled-coil region might be essential for the function of GSA-1/ARG1, or the function of GSA-1/ARG1 in root gravitropism may require a joint action of the N-terminal J domain and the C-terminal coiled-coil region, as a consequence of which the disruption of the C-terminal coiled-coil region would lead to a loss of function in GSA-1/ARG1 (Boonsirichai et al., 2003). Therefore, the identification of new mutant alleles of the ARG1 gene of Arabidopsis thaliana provides a useful system to examine novel mutations affecting mRNA stability and pre-mRNA splicing or functional protein stability.
We also explored the function of GSA-1/ARG1 in the regulation of basipetal auxin transport, and the expression of AUX1 at the protein level. The ARG1 mutation reduced basipetal auxin transport by 38.96% (Fig. 7), which is in good agreement with the greater amount of auxin accumulated in an expanded domain of the root apex relative to the wild-type (Boonsirichai et al., 2003; Harrison & Masson, 2008; Fig. 6). These findings suggest that the reduction in basipetally transported auxin results in auxin accumulation in the root cap. A similar phenomenon has also been described in the agravitropic mutant pin2/eir1/wav6/agr1, where a reduction in basipetally transported auxin was compensated for increased auxin accumulation in root tips, as indicated by DR5::GUS, DR5::GFP and IAA2:uidA, and by direct chemical measurement of the apical IAA content (Rashotte et al., 2000; Ottenschläger et al., 2003; Shin et al., 2005; Swarup et al., 2005).
ARG1 and PIN3 are engaged in the same gravity signal transduction pathway in root statocytes, and ARG1 is required for PIN3 relocalization and the asymmetrical redistribution of auxin (Harrison & Masson, 2008). The arg1 mutant shows a more severe root gravitropic defect relative to that of pin3 (Harrison & Masson, 2008), which may be associated with other established roles of GSA-1/ARG1, such as basipetal auxin transport and the expression of AUX1 protein (Figs 7, 9a). The specific AUX1 expression in the lateral root cap and in epidermal cells is required for the establishment of auxin gradients and root curvature following gravistimulation (Bainbridge et al., 2008). AUX1 probably functions by facilitating the uptake of auxin into the lateral root cap and the epidermal region, and PIN2 by mediating its directional translocation towards EZ in the gravitropic response (Friml, 2003). This interpretation is supported by direct basipetal auxin transport measurements and the higher accumulation of auxin in the root cap (Figs 6, 7). However, the mechanisms by which GSA-1/ARG1 regulates AUX1 expression have yet to be elucidated.
The double mutants aux1-22gsa-1 and eir1-1gsa-1 exacerbated the phenotype of gsa-1, indicating that the disruption of either aux1-22 or eir1-1 in the gsa-1 background may enhance the gravitropic defect of gsa-1 (Figs 10a,c, 11). Although it was found that the expression and localization of AUX1 were not affected, the PIN2 protein was decreased gradually under NH4+ stress (Figs 8, 9). This suggests a tight relationship for PIN2 in the NH4+-induced inhibition of root gravitropism. In addition to participating in AUX1-mediated basipetal auxin transport, PIN2 has been shown to be involved in the gravitropic response, and the pin2 mutant lacks the classic TZ bending response (Chavarría-Krauser et al., 2008; Baluška et al., 2010). This effect on TZ in relation to PIN2 was seen as a general stress target on account of the pronounced sensitivity to multiple adverse environmental stresses, such as cold, dark, salt and aluminum toxicity (Baluška et al., 2010). For example, under salt stress, PIN2 was targeted to endomembrane compartments and selectively degraded to disassemble the PIN2-based gravisensitive network and to allow roots to deviate from the gravity vector and avoid salt-enriched areas of soil (Li & Zhang, 2008; Sun et al., 2008). PIN2 is also degraded and basipetal auxin transport is inhibited if roots are kept in the dark (Laxmi et al., 2008; Wan et al., 2012), which is the natural root environment (Yokawa et al., 2011, 2013).
Two distinct bending regions have been reported in the literature, corresponding to EZ and TZ, and these drive the differential growth of root gravitropism (Verbelen et al., 2006; Baluška et al., 2010). The auxin concentration gradient-dependent EZ bending and the auxin concentration gradient-independent (but PIN2-related) TZ/DEZ bending are partially uncoupled from each other, and it has been shown that the mutation of PIN3 affects the bending of EZ, but not TZ/DEZ curvature (Chavarría-Krauser et al., 2008). We speculate that, similar to the pin3 mutant, the gsa-1 mutant lacks the auxin concentration gradient-dependent EZ curvature, whilst possessing normal TZ/DEZ bending. First, this is supported by the distinctly positive gravitropic response of gsa-1 in the absence of lateral auxin redistribution following gravistimulation (Figs 1-3, 6). Second, NH4+ decreased significantly the root gravitropism of gsa-1, whilst not further reducing basipetal auxin transport (Fig. 7).
The analysis of the root gravitropism kinetics of gsa-1, aux1-22, eir1-1, gsa-1aux1-22 and gsa-1eir1-1 under NH4+ stress (Fig. 10d) supports the model in which AUX1 and GSA-1/ARG1 participate in the same pathway, whereas PIN2 and GSA-1/ARG1 function independently. Lateral auxin gradient-dependent EZ bending, which involves PIN3, regulates the lateral auxin gradient across the root cap, and AUX1- and PIN2-mediated basipetal auxin transport into EZ. Therefore, we speculate that GSA-1/ARG1 functions in this process by direct or indirect regulation of PIN3 and AUX1, consistent with the observed increased GSA-1/ARG1 expression, the delayed lateral auxin gradient and the prolonged root gravitropic response of Col-0 under NH4+ stress (Fig. 6).
In summary, our molecular genetic and physiological results identify a gene locus essential for NH4+ tolerance in the specific context of root gravitropism. The gsa-1 mutant isolated displays distinct root gravitropic characteristics relative to those of the Col-0 wild-type under NH4+ stress. The effects of NH4+ on root gravitropism were independent of the classic amyloplast-involving gravity-sensing pathway. In addition to PIN3 (Harrison & Masson, 2008), GSA-1/ARG1 can be shown to be further required for appropriate AUX1 expression and basipetal auxin transport in root apices. Moreover, the expression of PIN2 was dramatically reduced, independent of GSA-1/ARG1 function, in root apices under NH4+ stress. Taken together, we propose that, during NH4+ stress, GSA-1/ARG1 expression is increased and is essential for the establishment of the PIN3-mediated lateral auxin gradient across the root cap. AUX1-related basipetal auxin transport on gravistimulation can partially antagonize the reduction in PIN2-mediated root gravitropism in the wild-type. However, the NH4+-induced inhibition of PIN2 could also cause the reduction of the root gravitropic response via disruption of the GSA-1/ARG1-dependent pathway. Thus, PIN2 in TZ emerges as a general stress target during multiple adverse environmental stresses, including NH4+ stress. Our working model provides new insights into the acclimation of root gravitropism to environmental stress.
We thank Professors Malcolm Bennett (University of Nottingham), Ben Scheres (Utrecht University) and Tom Guilfoyle (University of Missouri) for providing the transgenic lines of Arabidopsis, and the Arabidopsis Biological Resource Center for the mutant seeds. We are grateful to Dr Jianru Zuo (Institute of Genetics and Developmental Biology, Chinese Academy of Sciences) for kind provision of Arabidopsis seeds mutagenized with T-DNA transformation, Feifei Sun (Institute of Genetics and Developmental Biology, Chinese Academy of Sciences) for assistance with amyloplast staining and Dr Guangjie Li (Institute of Soil Science, Chinese Academy of Sciences) for assistance with DR5::GUS staining. We also thank other members of our team and members of the laboratory of Professor Yanhua Su (Institute of Soil Science, Chinese Academy of Sciences) for their assistance. This work was supported by the National Natural Science Foundation of China (31200189 and 30771285), the Chinese Academy Sciences Innovation Program (ISSASIP1103) and the Natural Sciences and Engineering Research Council of Canada (NSERC, Discovery Grant 217277-2009).