Generation of an amlt1_nip1;2 double mutant line
Three independent T-DNA insertion mutants of NIP1;2, i.e., nip1;2-1 (SALK_126593), nip1;2-2 (SALK_147353) and nip1;2-3 (SALK_076128), displayed comparable hypersensitive phenotypes to Al stresses at pH 4.3 (Additional file 1: Fig. S1) [19]. To further study the functional and genetic relationships between NIP1;2 and ALMT1, a homozygous almt1_nip1;2 double mutant line was generated through a cross between almt1 (SALK_009629) and nip1;2-3 (hereafter nip1;2), followed by selection from the F2 population of mutant plants with homozygous almt1 and nip1;2 alleles.
Real-time reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analyses indicated that in the wild type (WT, Col-0), the expression of ALMT1 and NIP1;2 in the root was both induced by Al treatment although the levels of ALMT1 transcripts were about 4-fold higher than those of NIP1;2 (Fig. 1). Under the nip1;2 mutant background, the level of ALMT1 expression in the root was comparable with that in the WT (Fig. 1a), whereas NIP1;2 expression was greatly suppressed (Fig. 1b). In contrast, under the almt1 background, although the level of the NIP1;2 expression in the root were comparable with that in the WT, the Al-induced ALMT1 expression in the root was barely detectable (Fig. 1a). These results confirmed that both almt1 and nip1;2 are knockout (KO) mutants [19, 24] and the expression of ALMT1 and NIP1;2 is independent of the each other [35]. Under the almt1_nip1;2 background, the expression of ALMT1 and NIP1;2 in the root was both barely detectable (Fig. 1a, b), indicating that almt1_nip1;2 is a double KO mutant line.
Comparable sensitivity between almt1_nip1;2 and almt1 to Al stresses
To evaluate Al resistance of individual lines, relative root growth (RRG%) was calculated for 5-day-old plants of WT, nip1;2, almt1 and almt1_nip1;2 treated with a range of Al concentrations (0-50 mM) at pH 4.3 (Fig. 2). Root growth of both the almt1 and nip1;2 single mutants was more severely inhibited by Al than was the WT over the range of Al concentrations tested (Fig. 2). Nevertheless, root growth was more strongly inhibited in almt1 than in nip1;2 (Fig. 2). For instance, at Al concentration of 20 mM, root growth was inhibited by 82 and 69 % in almt1 and nip1;2, respectively (Fig. 2 and Additional file 1: Fig. S2).
In contrast, no significant differences in root growth were observed between the almt1_nip1;2 double mutant and the almt1 single mutant over the entire range of Al concentrations tested (Fig. 2). Therefore, the Al-resistant phenotype of almt1_nip1;2 resembled that of almt1, but not nip1;2.
Genetic analysis of the allelic effects of ALMT1 and NIP1;2 on Al resistance
To evaluate the effects of genotypic variations at the ALMT1 and NIP1;2 loci on Al resistance, a homozygous almt1 plant was crossed with a homozygous nip1;2 plant to generate heterozygous F1 plants. The dominant and recessive alleles of ALMT1 and NIP1;2 would be segregated among the F2 plants, resulting in nine distinct genotypic combinations (Table 1 and Additional file 1: Fig. S3).
Root growth was measured for F2 seedlings germinated and grown for 7 days in a hydroponic growth solution (pH 4.3) supplemented with 20 mM AlCl3. Based on their root growth, the F2 plants could be classified into three phenotypic groups (A, B and C) that showed significant differences in root growth (mm) under Al treatment (Table 1 and Additional file 1: Fig. S3). On the other hand, based on their allelic variations at the ALMT1 and NIP1;2 loci, the F2 population could be classified into nine distinct genotypic groups/combinations (Table 1 and Additional file 1: Fig. S3).
Analysis of the relationship between phenotypic and genotypic variations indicated that 1) the F2 plants in the phenotypic group A had at least one dominant wild-type allele at both the ALMT1 and NIP1;2 loci (Table 1 and Additional file 1: Fig. S3); 2) group B had homozygous almt1/almt1 alleles regardless of the status of the NIP1;2 alleles; and 3) group C had homozygous nip1;2/nip1;2 mutant alleles and at least one wild-type allele of ALMT1, i.e., almt1/ALMT1 or ALMT1/ALMT1.
In phenotypic group A, plants with at least one wild-type ALMT1 allele and one wild-type NIP1;2 allele had comparable root growth under Al treatment as those with a wild-type background, i.e., ALMT1/ALMT1NIP1;2/NIP1;2 (Table 1 and Additional file 1: Fig. S3). In contrast, homozygous almt1 and/or nip1;2 plants in phenotypic groups B and C were more sensitive to Al compared with the plants in group A (Table 1 and Additional file 1: Fig. S3). These results indicated that the wild-type alleles of ALMT1 and NIP1;2 were both completely dominant.
Although homozygous mutations of almt1 or nip1;2 caused significant root growth inhibition (Fig. 2 and Table 1), the effects of genotypic variation at one locus on the phenotypic expression of the other locus were quite different between ALMT1 and NIP1;2 (Table 1). For instance, under a homozygous almt1/almt1 background (group B, Table 1), root growth was solely determined by the homozygous almt1 mutant alleles regardless of the genotypic variation at the NIP1;2 locus (Table 1 and Additional file 1: Fig. S3). In contrast, under a homozygous nip1;2/nip1;2 background, the degrees of root growth inhibition by Al were strongly affected by the genotypic variation at the ALMT1 locus. For instance, plants with homozygous almt1/almt1 alleles, i.e., the almt1_nip1;2 double mutant plants, displayed greatly enhanced root-growth inhibition compared with those with one or two copies of the wild-type ALMT1 allele in group C (Table 1 and Additional file 1: Fig. S3). The fact that the homozygous almt1 mutation at the ALMT1 locus could mask/override the effects of genotypic variation at the NIP1;2 locus indicates that there exist interactions between the ALMT1 and NIP1;2 loci where ALMT1 is genetically epistatic to NIP1;2.
Additive effects of ALMT1 and MATE and epistatic relationship between ALMT1 and NIP1;2 in Al resistance
In Arabidopsis, the Al-activated and ALMT1-facilated malate exudation from the root-tip region plays a major role in Al resistance, while the Al-activated and MATE-mediated citrate release from more mature root regions plays a smaller but significant role [24-26]. Although both the almt1 and mate single mutants were more sensitive to a range of Al concentrations (0-50 mM) tested than was the WT, almt1 consistently displayed significantly stronger root-growth inhibition than did the mate mutant (Fig. 3a and Additional file 1: Fig. S2).
Compared with the almt1 and mate single mutants, the almt1_mate double mutant showed significantly more severe root-growth inhibition phenotypes over the entire range of Al concentrations tested (Fig. 3a and Additional file 1: Fig. S2). For instance, at 20 mM Al, root growth of almt1_mate was inhibited by 93%, whereas root growth of almt1 and mate by 82 and 72%, respectively (Fig. 3a). Thus, the effects of ALMT1 and MATE on Al resistance are additive, suggesting that ALMT1 and MATE function in different biochemical pathways, which is consistent with our previous observation that ALMT1 and MATE function independently in achieving overall Al resistance in Arabidopsis [25].
In contrast, the almt1-nip1;2 double mutant did not display stronger mutant phenotypes than did the almt1 single mutant (Fig. 2). Instead, root growth was comparable between almt1 and almt1_nip1;2 over the entire range of Al concentrations tested (Fig. 2).
ALMT1-mediated root exudation of malate is independent of the NIP1;2 function
To evaluate the effects of different genotypes on root organic acid exudation, Al-activated root exudation of malate and citrate was examined for WT, almt1, nip1;2 and almt1_nip1;2. Under the control condition (-Al), comparable basal levels of root exudation of malate and citrate were observed among individual lines (Fig. 4a, b). Al exposure triggered releases of large and comparable amounts of malate from the roots of WT and the nip1;2 mutant (Fig. 4a). In contrast, both almt1 and almt1_nip1;2 lacked detectable Al-activated root malate exudation (Fig. 4a). These results indicate that Al-activated malate exudation from the root is mainly facilitated by the ALMT1 malate transporter in Arabidopsis and the Al-activated and ALMT1-mediated root malate exudation is independent of the NIP1;2 function.
Compared with root malate exudation, Al exposure also triggered smaller, but significant, increases in citrate exudation from the root (Fig. 4b). In contrast, no significant differences were observed in the amounts of citrate in the root exudates from all lines examined upon Al exposure (Fig. 4b). These results indicate that the Al-activated and MATE-facilitated root citrate exudation is independent of the ALMT1 and NIP1;2 functions in Arabidopsis (Fig. 4b). Thus, the phenotypes of organic acid exudation of the almt1_nip1;2 double KO line resemble those of the almt1 mutant, but not the nip1;2 mutant.
ALMT1 functions upstream of NIP1;2 in the process of Al removal from the root cell wall
To test the relationship between ALMT1 and NIP1;2 in the processes of Al removal from the root cell wall, Al contents in the root cell wall and cell sap were measured for the WT, almt1, nip1;2 and almt1_nip1;2 plants treated with 50 mM AlCl3 at pH 4.3 for 2 d (Fig. 5). Compared with the WT plants, the almt1 and nip1;2 plants accumulated significantly higher and lower concentrations of Al in the root cell walls (Fig. 5a) and root cell sap (Fig. 5b), respectively. These results confirmed that both ALMT1-mediated malate releases and a functional NIP1;2 are required for Al removal from the root cell wall into the root cytosol [19]. However, compared with the nip1;2 mutant, the almt1 mutant also accumulated significantly higher concentrations of Al in the root cell wall (Fig. 5a) and lower concentrations of Al in the root cell sap (Fig. 5b). These results suggest that besides the ALMT1 and NIP1;2-dependent process, there exist ALMT1-depenedent but NIP1;2-indenpent processes for Al removal from the root cell wall in Arabidopsis.
The Al concentrations in the root cell wall (Fig. 5a) and root cell sap of the almt1_nip1;2 double mutant (Fig. 5b) were comparable with those of the almt1 single mutant, which were significantly different from those in the nip1;2 single mutant. These results indicate that ALMT1 is genetically epistatic to NIP1;2 in the biochemical pathway leading to Al removal from the root cell wall into the root symplasm in Arabidopsis.
Externally supplied malate partially restored NIP1;2-facilitated Al uptakes from the root cell wall in almt1 but not in almt1_nip1;2
To evaluate the effects of externally supplied malate on Al uptakes from the root cell wall for almt1, nip1;2 and almt1_nip1;2, plants were treated with 50 mM AlCl3 (pH 4.3) for 8 h, allowing Al to get into and be retained in the root cell walls (Fig. 6a) [19], followed by addition of 0 or 200 mM malate for another 8 h.
Between these two treatments, no statistically significant differences in Al contents in the root cell wall (Fig. 6a) and the root cell sap (Fig. 6b) were observed in the nip1;2 single mutant and the almt1_nip1;2 double mutant. In contrast, in almt1, compared with those under the Al treatment alone, external supplementation of malate after Al treatment led to significantly decreased Al concentrations in the root cell wall (Fig. 6a) and significantly increased concentrations in the root cell sap (Fig. 6b). These results indicate that even though the almt1 mutant has a functional NIP1;2 transporter [19], the presence of malate in the root cell wall is essential for NIP1;2-facilitated Al removals from the root cell wall. Thus, the ALMT1-mediated releases of malate to the root cell wall function in an earlier step in the NIP1;2-facilitated process for Al uptakes from the root cell wall to the root cytosol.