Comparative genetic analysis of Arabidopsis purple acid phosphatases AtPAP10,AtPAP12,and AtPAP26 provides new insights into their roles in plant adaptation to phosphate deprivation

Liangsheng Wang,Shan Lu,Ye Zhang,Zheng Li,Xiaoqiu Du and Dong Liu

MOE Key Laboratory of Bioinformatics,Center for Plant Biology,School of Life Sciences,Tsinghua University,Beijing 100084,China.

INTRODUCTION

Phosphorus(P)is one of the most important yet least available of the nutrients required for plant growth and development.Inorganic phosphate(Pi)is the major form of phosphorus that plants acquire from soil through the Pi transporters located on their root surfaces(Bieleski 1973;Raghothama 2000).In most soils,however,the concentration of Pi is far below the level required for optimal plant growth(Vance et al.2003;Richardson 2009).To cope with Pi deficiency,plants have evolved a complex set of adaptive mechanisms that improve their acquisition and utilization of Pi.One of these mechanisms involves production and secretion of acid phosphatases(APases)(Yuan and Liu 2008;Tran et al.2010a).These enzymes possess a non-specific monophosphoricmonoester hydrolase(EC 3.1.3.2)activity that cleaves Pi from ester linkage sites under acidic conditions.The Pi-starvation induced(PSI)intracellular APases are likely involved in the remobilization and recycling of Pi from intracellular P monoesters and anhydrides of older tissues whereas extracellular or secreted APases are believed to scavenge Pi from organophosphate compounds in the external environment.In soils,most P exists in the form of organophosphate that cannot be taken up by roots(Vance et al.2003).Secretion of APases from plants may increase the availability of Pi for root absorption.

PSI APases have been biochemically purified and molecularly characterized from a variety of plant species,including those in white lupin(Lupinus albus;Ozawa et al.1995;Li and Tadano 1996;Miller et al.2001),tomato(Solanum lycopersicum;Bozzo et al.2002,2006),common bean(Phaseolus vulgaris;Liang et al.2010),tobacco(Nicotiana tabacum;Lung et al.2008),and Arabidopsis thaliana(Veljanovski et al.2006;Kuang et al.2009;Tran et al.2010b).Among these characterized APases is one unique class called purple acid phosphatases(PAPs),which appear purple or pink in water solution because they contain a bimetallic active centre(Olczak et al.2003).In Arabidopsis,there are 29 annotated PAP genes(AtPAPs)(Li et al.2002).Among them,transcripts of 28 have been detected in different plant organs,and at least 11 are up-regulated by Pi deficiency(del Pozo et al.1999;Haran et al.2000;Li et al.2002;Zhu et al.2005;Wang et al.2011).Veljanovski et al.(2006)directly purified a vacuolar APase from Arabidopsis suspension cells and seedlings,and identified it as AtPAP26.Recently,the same research group found that AtPAP12 and AtPAP26 are the two predominant isoforms of secreted APases in the liquid culture medium of Arabidopsis suspension cells or seedlings(Tran et al.2010b).Thus,AtPAP26 is targeted to both vacuoles and the culture medium;but whether AtPAP12 is also an intracellular APase remained elusive(Haran et al.2000).Using transgenic approach,Wang et al.(2009)demonstrated that overexpression of AtPAP15 in soybean improved Pi acquisition and increased plant growth in soil with phytate as the sole P source;however,the first genetic evidence that APase is essential for plant adaptation to Pi deprivation was only recently emerged.It was reported that the growth of the pap26 single mutant or the pap12/26 double mutant(Hurley et al.2010;Robinson et al.2012)was impaired under Pi deficiency.It is not known,however,whether the overexpression of the AtPAP26 or AtPAP12 gene enhances plant growth on a Pideficient medium containing organophosphate as the P source.

The PSI secreted APases can be assigned to two groups.One is released into the growth medium,and the other is tightly associated with the root surface after secretion.The latter root-associated APases have long been observed in many plant species(Mclachlan 1980;Boutin et al.1981;Silberbush et al.1981)but only one member of that group,AtPAP10,has been cloned(Wang et al.2011).Biochemical studies of the recombinant AtPAP10 proteins indicated that AtPAP10 has typical APase activity against a range of organophosphates that partially overlaps with the substrate ranges of AtPAP12 and AtPAP26.Although AtPAP10 is most closely related to AtPAP12 and AtPAP26 in the AtPAP family in Arabidopsis,it behaves quite differently.Instead of being an intracellular APase or released into the culture medium,AtPAP10 is predominantly associated with the root surface after secretion.Analyses of multiple mutant lines and overexpressing lines demonstrated that AtPAP10 is an essential component of plant adaptive responses to Pi limitation(Wang et al.2011;Wang and Liu 2012).

To systemically investigate the roles of AtPAPs in plant adaptation to Pi deprivation,we first tested the effect of mutation of each AtPAP gene on total intracellular APase activity.Our results indicated that only atpap12,atpap15,and atpap26 mutants showed a significant reduction in their intracellular APase activity.Because AtPAP10,AtPAP12,and AtPAP26 are three most closely related members in AtPAP family(Li et al.2002),we would like to determine their similarity and difference.Thus,we performed a detailed comparative genetic analysis of their APase activities and roles in plant adaption to Pi deprivation.Our results show that these three APases play distinct roles in plant adaptation to Pi deprivation.The growth phenotypes of the various atpap mutants,which we observed in this work,differ from what has been previously reported(Hurley et al.2010;Robinson et al.2012).We further compared the cellular Pi contents among the WT,various atpap mutants,and AtPAP-overexpressing lines,and found that Pi content was not correlated with growth phenotype.These results suggest that the APases may have roles in addition to enhancing internal Pi recycling or releasing Pi from external organophosphates for plant uptake.

RESULTS

Effects of AtPAP gene mutations on the activities of intracellular APases in Arabidopsis

To determine the relative contributions of each of the 29 AtPAP genes to the total intracellular APase activity in Arabidopsis,we first identified knockout or knockdown lines from T-DNA insertion lines.For each AtPAP gene,at least one homozygous T-DNA insertion line was obtained that was a null or a nearly null allele as demonstrated by RT PCR analysis of mRNA expression(Figure S1,Table S1).We then compared the total intracellular APase activity in shoot and root tissues among these T-DNA lines under both Pi-sufficient and-deficient conditions using pnitrophenyl phosphate(pNPP)as a substrate.In shoots of 14-day-old Arabidopsis seedlings grown on P+medium,total APase activity was 20%lower in the line pap12(GK-151C09)than in the WT,and was about 60%lower in the lines pap15(SAIL_529_D01)and pap26(SALK_152821)than in the WT(Figure 1A).Total APase activity in shoots of all T-DNA lines was higher under Pi deficiency than under Pi sufficiency;under Pi deficiency,however,total APase activity was 50%lower in pap12,25%lower in pap15,and 30%lower in pap26 than in the WT(Figure 1B).

Similarly,in the roots of seedlings grown on P+medium,APase activity was 20%lower in pap12 and about 60%lower in pap15 and pap26 than in the WT(Figure 1C).In the roots of Pistarved seedlings,APase activity was about 50%lower in pap12 and about 35%lower in pap15 and pap26 than in the WT(Figure 1D).In addition,APase activity in Pi-starved roots was reduced about 15%relative to the WT in the T-DNA insertion lines of AtPAP1,AtPAP2,AtPAP3,AtPAP4,and AtPAP5(Figure 1D).Taken together,the results indicate that AtPAP12,AtPAP15,and AtPAP26 are three major intracellular APases in Arabidopsis.

Because AtPAP10,AtPAP12,and AtPAP26 are three most closely related members in the Arabidopsis PAP family(Li et al.2002),we then focused on the comparative analyses of the activity of these three AtPAPs and their roles in plant adaptation to Pi deprivation.

In vivo intracellular APase activity of AtPAP10,AtPAP12,and AtPAP26 against different substrates

To further determine the relative contributions of AtPAP10,AtPAP12,and AtPAP26 to the total intracellular APase activity in Arabidopsis,we examined the effects of the mutations of these three genes on the intracellular APase activity against the following four organophosphate substrates:5-bromo-4-chloro-3-indolyl-phosphate(BCIP),ADP,ATP,and fructose-6-phosphate(Fru-6-P).The results obtained with the four substrates were relatively consistent with those obtained with pNPP as described in the previous section(Figure 2A).Under the P+condition and for all four substrates,knockout of the AtPAP26 gene resulted in the largest reduction of total APase activity in the shoots and roots;the reduction ranged from 50%–80%depending on the plant tissues examined.The reduction of APase activity in the shoots and roots was less in pap12 than in pap26.Under the P-condition,the reduction of APase activity in the shoots was similar for pap12 and pap26,ranging from 30 to 50%.In the roots,the APase activity was also lower for pap26 than for pap12 for three substrates but was reduced to a similar degree with BCIP.In none of cases,the mutant pap10(SALK_122362)showed an disenable reduction in the total intracellular APase activity,except in Pi-starved roots when BCIP was used as a substrate.

Figure 1.Intracellular APase activity(as determined with pNPP as the substrate)in 14-d-old Arabidopsis seedlings of the WT and the pap mutants grown on P+or P-medium(A)and(B)Intracellular APase activity in shoots.(C)and(D)Intracellular APase actvity in roots.The Pi conditions of the culture medium are indicated on the top of each plot.Values represent means with SE of six replicates.The experiment was repeated three times with similar results.Asterisks indicate significant differences compared to the WT(P＜0.05,t-test).

We also evaluated the APase activity in the shoots and roots of the double(pap10/12 and pap12/26)and triple(pap10/12/26)mutants under P+and P-conditions using pNPP as a substrate.In both the shoots and roots of P+and P-seedlings,there was a progressive reduction in total APase activity when the mutations of two or three PAP genes were combined into one line(Figure 2B,C).The effect of multiple PAP gene mutations was additive.In the pap10/12/26 triple mutant,however,the shoots and roots still retained 30%–40%of the APase activity.This remaining APase activity might be encoded by AtPAP15 and other APase genes.

Figure 2.Intracellular APase activity(as determined with various substrates)in the shoots and roots of 14-d-old Arabidopsis seedlings of the WT and pap mutants(including double and triple mutants)grown on P+or P-medium(A)Intracellular APase activity in the WT and in pap10,pap12,and pap26 mutants as determined with BCIP,ADP,ATP,and Fru-6-P as the substrates.(B)and(C)Intracellular APase activity in the WT and single,double,and triple pap mutants as determined with pNPP as the substrate.Values represent means with SE of six replicates.The experiment was repeated three times with similar results.A one-way ANOVA analysis was carried out for the whole dataset and post hoc comparisons were conducted using the SPSS Tukey HSD test at P＜0.05 level.Significant differences are indicated by different letters on the top of each column.

Secreted APase activity of AtPAP10,AtPAP12,and AtPAP26

Figure 3.Analysis of secreted APase activity in Arabidopsis seedlings of the WT and the pap mutants grown in or on P+or P-medium(A)APase activities in the medium as determined with pNPP as the substrate.(B)Root-associated APase activities as determined with pNPP as the substrate.(C)and(D)Detection of APase activities on the root surfaces as determined by staining with 0.01,0.08,or 0.16%BCIP.For(C)and(D),the seedlings were grown on a solid agar medium and were 8 days old when stained.The experiment in(A)and(B)were repeated three times with similar results;values represent means with SE of six replicates.A one-way ANOVA analysis was carried out for the whole dataset and post hoc comparisons were conducted using the SPSS Tukey HSD test at P＜0.05 level.Significant differences are indicated by different letters on the top of each column.

Secreted or extracellular APase activity includes that in the culture medium and that associated with the root surface.As reported previously(Wang et al.2011),the AtPAP10 mutation did not significantly affect the total APase activity in the culture medium(Figure 3A).When grown under the Pi-sufficient condition,APase activity was 10%lower in pap12 and 30%lower in pap26 than in the WT.Under Pi deficiency,however,the reduction of APase activity in the culture medium was greater for pap12 than for pap26(40%vs.25%).Reductions in APase activity in the medium caused by double and triple mutations were additive.We then evaluated the effect of AtPAP gene mutations on the root-associated APase activity(Figure 3B).Under the P sufficiency,total root-associated APase activity was 40%lower in pap10,25%lower in pap26,and not lower in pap12 than in the WT.Under Pi deficiency,although the total root-associated APase activity were all increased for all plants,root-associated APase activity was 40%lower in pap10 and 15%lower in pap12 and pap26 than in the WT.Like that in the culture media,reductions in rootassociated APase activity were also additive for double and triple mutants grown under both the P+and P-condition.Taken together,the results indicate that AtPAP12 and AtPAP26 exist both in the culture medium and with the root surface but that AtPAP10 is predominantly associated with the root surface.The remaining APase activity in the culture medium and the root surface of pap10/12/26 triple mutant indicated that APases other than AtPAP10,AtPAP12,and AtPAP26 also existed.

The activity of secreted APases on the root surface of Arabidopsis can also be visualized by applying an agar solution containing the APase substrate BCIP(Lloyd et al.2001).Cleavage of BCIP by APase produces a blue precipitate.As previously documented(Wang et al.2011),when 0.01%BCIP was applied,the roots of WT seedlings showed no blue staining on the P+medium(Figure 3C)but stained dark blue on the P-medium(Figure 3D).In contrast,the pap10 mutant grown on both the P+and P-medium did not stain blue.On the P+medium,none of the other 28 single pap mutants or the pap12/26 double mutants were stained blue(Figure 3C and data not shown).When the concentration of BCIP was increased from 0.01%to 0.08%and 0.16%,all the lines grown on the P+medium were stained light blue,except for the lines containing the pap10 mutation(Figure 3C).On the P-medium with 0.08%BCIP applied,the lines with the pap10 mutation were stained light blue(Figure 3D).When the concentration of BCIP was increased to 0.16%,the staining intensity was similar for the pap10 single mutant and the WT,whereas the staining intensity was lower for the pap10pap12 double mutant and the pap10pap12pap26 triple mutant than for the pap10 single mutant.In other words,the reduction in BCIP staining resulting from AtPAP12 and AtPAP26 mutations could only be detected in the plants carrying the AtPAP10 mutation and when the high concentration of BCIP was used.This seems to contradict the finding that AtPAP12 and AtPAP26 are two major secreted APases.Our explanation is that AtPAP12 and AtPAP26 have low affinity for BCIP,and thus their APase activities on the root surface do not contribute significantly to the total intensity of BCIP staining.The supporting evidence for this hypothesis is that the reduction in the root-associated APase activity in the pap12/26 double mutant was even greater than that in pap10 when a generic substrate pNPP was used in the APase assay(Figure 3B),but the blue staining of the double mutant was still as intense as the that of the WT(Figure 3D).

Effects of AtPAP10,AtPAP12,and AtPAP26 mutations on APase profiles

Each plant species has its own characteristic APase profile,which can be generated by in-gel assays.To investigate the effects of mutations of AtPAP10,AtPAP12,and AtPAP26 genes on the APase profile in Arabidopsis,we first examined the APase profiles of WT plants grown on P+and P-media.For the denaturing in-gel assay,total soluble proteins were extracted from shoots,roots,and media,respectively,and separated on an SDS-PAGE gel.After electrophoresis,the SDS was removed from the gel to allow the proteins to renature.The gel was then stained with APase substrates Fast Black K salt and β-naphthyl acid phosphate to display an APase profile(Trull et al.1998).When grown on both P+and P-medium,shoots of Arabidopsis seedlings contained five APase isoforms(A1 to A5)(Figure 4A)with activities of all isoforms,except A1,were greater under Pi deficiency than under Pi sufficiency.In roots,only three APase isoforms(A1-A3)were observed(Figure 4B).Appearance of the band above the isoform A2 was consistent in separate experiments.We speculated that this inconsistent band was the incompletely denatured A2 isoform because the change of its intensity under different Pi conditions was proportional to the A2 isoform.In contrast to A2 activity in shoots,A2 activity in roots was down-regulated by Pi starvation.In the culture medium,only A1,A2,and A3 isoforms were observed(Figure 4C).Different from that in shoots and roots,A2 activity in the medium was not affected by Pi starvation while A1 activity in the medium was induced by Pi starvation.For the non-denaturing in-gel assay,the same experimental procedure was used as in the denaturing in-gel assay except that SDS was omitted during electrophoresis.Under the non-denaturing condition,three APase isoforms(a1,a3,and a4)were detected in shoots and culture medium(Figure 4D,F)while four APase isoforms(a1,a2,a3,and a4)were detected in roots(Figure 4E).The activities of a1 were not affected by Pi starvation in shoots and roots but were enhanced in the P-culture medium(Figure 4D–F).While activity of a4 was increased by Pi deficiency in shoots,roots,and culture medium(Figure 4D–F),the activity of a3 was enhanced only in shoots and culture medium.

To determine the effects of AtPAP10,AtPAP12,and AtPAP26 mutations on the APase profiles,we compared the APase profiles among the WT,and single,double,and triple mutants of AtPAP10,AtPAP12,and AtPAP26 genes.In denaturing in-gel assays,the mutation of AtPAP10 caused only a slight reduction in A5 activity in shoots(Figure 4A)and had no obvious effect on the APase profiles of roots(Figure 4B)and culture medium(Figure 4C).In the shoots of the pap12 mutant,isoform A3 was eliminated,and the intensity of A4 and A5 was also reduced relative to the WT(Figure 4A).Similarly,in the roots and culture medium,knockout of the AtPAP12 gene eliminated A3(Figure 4B,C).A3 was also eliminated in the pap10/12 and pap12/26 double mutants and in the pap10/12/26 triple mutant under both P+and P-conditions.These results demonstrated that A3 contained only AtPAP12 proteins.In addition,in the Pistarved roots of the pap12 and pap12/26 lines,a faint band appeared above isoform A3(Figure 4B).This faint band was not present in the pap10/12 double mutant or the pap10/12/26 triple mutant,suggesting that it might be encoded by the AtPAP10 gene as a compensation for the loss of AtPAP12 and AtPAP26,but that inference requires confirmation by other experimental approaches.Furthermore,the intensity of A4 and A5 in shoots was much lower in the pap10/12 double mutant than in the pap12 single mutant(Figure 4A).This indicated that isoforms A4 and A5 also contained some AtPAP10 proteins.For the pap26 mutant,APase profiles for the shoots,roots,and culture medium did not differ from the WT under both P+and P-conditions(Figure 4A–C).Moreover,the addition of the pap26 mutation to the pap12 single mutant or the pap10/12 double mutant did not cause further change of the APase profiles.The remaining APase activity in the pap10/12 double mutant and pap10/12/26 triple mutants indicated the presence of other APase activities in the A4 and A5 isoforms.

In non-denaturing in-gel assays,the mutation of the AtPAP10 gene had no effect on the APase profiles of shoot,roots,and culture medium while knockout of AtPAP12 eliminated the activity of a4,and knockout of AtPAP26 eliminated the activity of a3(Figure 4D–F).In the pap12/26 double mutant and the pap10/12/26 triple mutant,no a3 or a4 activities were detected.These results suggested that the activity of AtPAP26 could not be revealed with a denaturing in-gel assay.In denaturing assays,AtPAP26 was probably irreversibly inactivated by SDS.

Figure 4.Analysis of APase profiles of 14-d-old Arabidopsis seedlings of the WT and the pap mutants grown on P+or P-medium(A)to(C)APase profiles of shoots(A),roots(B),and culture medium(C)as determined by denaturing in-gel assays.The molecular masses are indicated on the left.(D)to(F)APase profiles of shoots(D),roots(E),and culture medium(F)as determined by nondenaturing in-gel assays.The designation for each APase isoform in all experiments is indicated on the right.(+)and(-)indicate Pi sufficiency and Pi deficiency,respectively.

Analyses of growth phenotypes and Pi content of pap mutant lines

To assess the effect of AtPAP10,AtPAP12,and AtPAP26 mutations on plant growth and development,we first grew the mutant plants and the WT under normal growth conditions in agar plates.Shoot and root biomass of 16-day-old seedlings did not significantly differ among the WT and all single,double,and triple pap mutant lines(Figures 5A,S2A).The growth phenotypes were then compared when the WT and pap mutant lines were grown on P-agar medium or P-agar medium supplemented with different amounts of organophosphates(ADP or Fru-6-P).On P-medium,growth was significantly less for pap10(a 15%and 20%reduction in shoot and root biomass,respectively)and for pap12(a 10%reduction for both shoot and root biomass)than for the WT,whereas growth did not significantly differ between pap26 and the WT(Figure 5A).

Figure 5.Shoot and root fresh weights of 16-d-old seedlings of the WT and the pap mutants grown on P+medium and P-medium(A),or P-medium supplemented with ADP(B)and Fru-6-P(C)The amounts of ADP and Fru-6-P added to the P-medium are indicated on the top of the plots.The experiment were repeated three times with similar results.Values represent means with SE of three replicates.Each replicate contained 40 seedlings.A oneway ANOVA analysis was carried out for the whole dataset and post hoc comparisons were conducted using the SPSS Tukey HSD test at P＜0.05 level.Significant differences are indicated by different letters on the top of each column.

On P-medium supplemented with 50 or 150 μM ADP,shoot biomass was 15%lower for pap10 than for the WT but did not differ among the single mutants pap12 and pap26,the double mutant pap12pap26,and the WT(Figures 5B,S2B).The reduction of shootbiomass of the pap10/12 doublemutant and the pap10/12/26 triple mutant was similar to that of the pap10 single mutant.The root biomass was 25%and 30%lower for the single pap10 mutant than for the WT when grown on P-medium with 50μM and 150μM ADP,respectively,but the root biomasses for pap12 and pap26 single mutants were not reduced under the same conditions.The double or triple mutant containing the AtPAP10 mutation had similar reduction in root biomass as the pap10 single mutant.On P-medium with 50 μM ADP,however,the root biomass of the pap12/26 double mutant was reduced by 25%.

On P-medium supplemented with 50 μM Fru-6-P,shoot biomass was 25%lower and root biomass was 30%lower for the pap10 single mutant than for the WT but shoot and root biomasses did not differ among the pap12 and pap26 single mutants,the pap12/26 double mutant,and the WT(Figures 5C,S2C).The growth phenotypes of the pap10/12 double mutant and the pap10/12/26 triple mutant were similar to that of the pap10 single mutant.On P-medium with 150 μM Fru-6-P,however,the shoot and root biomasses were lower for all the mutant lines than for the WT,except that shoot growth did not differ between pap26 and the WT.

We then measured the cellular Pi contents of the WT and pap mutant lines grown under various Pi conditions.Pi contents in the shoots did not significantly differ between the WT and any of the pap mutant lines grown on P+,P-,or P-medium with 50 μM ADP(Figure S3).When grown on the P-medium supplemented with 150 μM ADP,50 μM Fru-6-P,or 150μM Fru-6-P,only the double and triple mutants showed a significant decrease of Pi contents in the shoots compared to the WT.Pi content in the roots was similar for all plants grown under the same Pi condition,except that the triple mutant contained less Pi than the WT on P-medium.These results indicate that Pi content is not correlated with growth phenotype of the WT and the pap mutants.

Overexpression of AtPAP10,AtPAP12,and AtPAP26 in transgenic plants

Previously,we reported that overexpression of the AtPAP10 gene improved plant growth on P-medium or P-medium supplemented with the organophosphates ADP,Fru-6-P,or phosphoenopyruvate(Wang et al.2011;Wang and Liu 2012).To test whether overexpression of AtPAP12 and AtPAP26 also increases growth under Pi deficiency,we generated transgenic Arabidopsis lines overexpressing AtPAP12 or AtPAP26.The genomic sequences of the AtPAP12 and AtPAP26 genes were fused to the CaMV 35S promoter and introduced into WT plants.More than 20 transgenic lines were generated for each construct.Four representative lines with a single T-DNA insertion for each construct were chosen and made homozygous for the transgene before further characterization.The overexpression of AtPAP12 and ATPAP26 mRNAs in these transgenic lines was confirmed by quantitative real-time PCR(Figure S4).In 35S:AtPAP10 lines,total intracellular APase activity in 14-day-old whole seedlings grown on P+and P-medium was only 10%greater than in the WT(Figure 6A);rootassociated activity however,was 150%–200%greater in four independent transgenic lines than in the WT(Figure 6B).This indicated that AtPAP10 produced in transgenic lines was mainly secreted and associated with the root surface.In the denaturing in-gel assay,the APase profile in shoots did not differ between the 35S:AtPAP10 line and the WT but the intensity of the A3 isoform was increased in the roots(Figure 6C).In addition,two new bands of high molecular weight appeared in the root profile,consistent with our previous published data(Wang et al.2011).In non-denaturing in-gel assays,the APase profile in shoots did not differ between the 35S:AtPAP10 line and the WT.In roots,however,one new band of high molecular weight was evident(Figure 6D).

For 35S:AtPAP12 and 35S:AtPAP26 lines,in contrast,total intracellular APase activity was 150%–300%greater than in the WT under both P+and P-conditions(Figure 6A).As was the case with the 35S:AtPAP10 lines,root-associated APase activity was much greater(150%–300%)in 35S:AtPAP12 and 35S:AtPAP26 lines than in the WT under both P+and P-conditions(Figure 6B).Ina denaturing in-gel assay using shoot tissue,the intensity ofA3 and A5 was significantly greater with the 35S:AtPAP12 o line than with the WT(Figure 6C).The roots of the At:AtPAP12 line also exhibited increased A3 activity and two new bands with high molecular mass(Figure 6C).In addition,both shoots and roots of the 35S:AtPAP12 line contained a new band below the A4 isoform,which might be the degradation products of AtPAP12.For the AtPAP26 overexpressing line,all APase isoforms were similar to those in the WT.A new band below A3,however,appeared in both shoots and roots(Figure 6C).This was probably due to the incomplete inactivation ofthe overproduced AtPAP26 by SDS.In the non-denaturing in-gel assay,the intensity of the a4 isoform in both shoots and roots of 35S:AtPAP12 lines and intensity of the a3 isoform was greatly enhanced,further confirming that a4 and a3 were indeed the products of the AtPAP12 and AtPAP26 genes(Figure 6D).

Analyses of growth phenotypes and Pi content of AtPAP overexpressing lines

AtPAP10,AtPAP12,and AtPAP26 overexpressing lines were grown under various conditions and their growth phenotypes were compared.Like AtPAP10(Wang et al.2011),under Pi-sufficient conditions,growth of 14-day-old seedlings of 35S:AtPAP12 and 35S:AtPAP26 lines did not differ from the WT when grown on agar medium(Figure 7A,B).Consistent with the previously published results(Wang et al.2011),the four AtPAP10 overexpressing lines grew better than the WT on P-medium and P-medium supplemented with 50 or 150 μM ADP(data not shown).Similarly,the four 35S:AtPAP12 lines gained 25%to 40%more shoot and root biomass than the WT on P-medium and P-medium supplemented with different amounts of ADP or Fru-6P,except in the case of root biomass on P-medium with 50 μM Fru-6-P(Figure 7A).Except for shoots on P-medium and roots on P-medium with 50 μM Fru-6-P,the four AtPAP26 overexpressing lines also produced 25%to 50%more shoot and root biomass than the WT(Figure 7B).

Next,we compared the Pi contents between the WT and overexpressing lines grown on different Pi media.Interestingly,although plant biomass significantly differed between the WT and overexpressing lines when grown on various P-media,the Pi contents in shoots and roots did not differ between the WT and overexpressing lines(Figure 8).These results indicate that the enhanced growth of overexpressing lines was not due to their enhanced accumulation of cellular Pi.

DISCUSSION

Figure 6.Continued

Previous studies have indicated that AtPAP26 is both a major intracellular and secreted APase(Veljanovski et al.2006;Hurley et al.2010;Tran et al.2010b).In this work,the observed significant reduction in the total intracellular APase activity(Figures 1,2)and the absence of a major APase isoform in the shoot,roots,and culture medium of the pap26 mutant(Figure 4D–F)further support the conclusion that AtPAP26 is both a major intracellular and secreted APase.Analytical gel infiltration indicated that the vacuole-targeted AtPAP26 exists as a homodimer composed of two 55-kD glycosylated subunits(Veljanovski et al.2006)and that secreted AtPAP26 is a monomer(Tran et al.2010b).Our non-denaturing in-gel assay,however,showed that the AtPAP26 isoforms isolated from shoots,roots,and culture medium have the same mobility on PAGE gel(Figure 4D–F).This suggested that AtPAP26 in all three locations existed in the same form,although their molecular mass could not be precisely determined by nondenaturing in-gel assay.Because AtPAP26 protein lacks a conserved cysteine residue that is predicted to participate in the formation of a disulphide bond,the formation of the AtPAP26 homodimer may result from the non-covalent association of two monomers.Perhaps the loose association of the two monomers could be disrupted during the extraction of proteins or electrophoresis under our experimental conditions.Another puzzling result was that the activity of the AtPAP26 isoform was not enhanced in roots in the current study,as shown by in-gel assay,but the protein level of AtPAP26 increased in previous studies as demonstrated by Western blot(Veljanovski et al.2006;Hurley et al.2010;Tran et al.2010b).Explaining this difference will require additional research.Furthermore,using the pap26 mutant and overexpressing lines,we demonstrated that AtPAP26 is also associated with the root surface after secretion(Figure 2B).

Figure 6.APase activities and APase profiles of 14-d-old seedlings of the WT and AtPAP overexpressing lines grown on P+or P-mediumA Total intracellular APase activity of the WT and four overexpressing lines of AtPAP10,AtPAP12,and AtPAP26.B Root-associated APase activity of the WT and four overexpressing lines of AtPAP10,AtPAP12,and AtPAP26.The experiment in(A)and(B)were repeated three times with similar results.Values represent means with SE of three replicates.Asterisks indicate significant differences compared to the WT(P＜0.05,t-test).In(A)and(B),the names of the construct of the overexpressing lines are indicated at the bottom of the plot.(C)APase profiles of shoots and roots of the WT and PAP overexpressing lines as determined by denaturing in-gel assay.(D)APase profiles of shoots and roots of the WT and AtPAP as determined by non-denaturing in-gel assay.In(C)and(D),the name of the construct of the overexpressing lines are indicated on the top of the chart.In(C),new APase bands that were not observed in the WT are indicated by arrows.In all panels,(+)and(-)mean Pi sufficiency and Pi deficiency,respectively.

Tran et al.(2010b)showed that AtPAP12 is a major secreted APase in the culture medium of Arabidopsis cell suspensions or seedlings.In our study,the elimination of a major APase isoform(Figure 4C–F)and the significant reduction of total APase activity in the culture medium of pap12(Figure 2A)provide further evidence for their conclusions.In addition,we found that AtPAP12,like AtPAP10 and AtPAP26,is associated with the root surface after its secretion when plants are grown under Pi deficiency(Figure 2B).The amount of root-associated AtPAP12 activity relative to the total root-associated APase activity of Pistarved plants is similar to that of AtPAP26(about 15%)but much lower than that of AtPAP10(40%).Furthermore,the significant reductionin total intracellular APase activity and the eliminationof a major APase isoform in the shoots and roots of pap12 demonstrated that AtPAP12 is also a major intracellular APase(Figures 1,2,4A–B,4D–E).AtPAP12 purified from the culture medium of Arabidopsis suspension cells or seedlings was shown to be a homodimer consisting of two 60-kD monomers(Tran et al.2010b).In this study,the molecular mass of the major AtPAP12isoform(A3)intheAPaseprofile ofshoots isabout 110kD(Figure 4A),close to that of a homodimer.Our denaturing and non-denaturing in-gel assays both showed that the major AtPAP12 isoform in the shoots,roots,and culture medium have similar mobility,indicating that they all existed as a dimer.The large differenceingel-mobilitybetweenAtPAP12and AtPAP26isoforms in our in-gel assay(Figure 4D–F)also support the notion that AtPAP12 is a dimer and that AtPAP26 is a monomer.In addition,some AtPAP12 proteins existed as an oligomer(bands A4 and A5,whose molecular masses were approximately 200 kD and＞250 kD,respectively).The oligomer form of AtPAP12,however,only represents a small fraction of the total amount of AtPAP12 proteins.This was probably why oligomer forms were not found during the direct biochemical purification of AtPAP12 from the culture medium(Tran et al.2010b).

Figure 7.Shoot and root fresh weights of 14-d-old seedlings of the WT and PAP overexpressing lines grown on P+medium,P-medium,or P-medium supplemented with 50 or 150 μM ADP or Fru-6-P(A)Values for the WT and four AtPAP12 overexpressing lines.(B)Values for the WT and four AtPAP26 overexpressing lines.The experiment were repeated three times with similar results.Values represent means with SE of three replicates.Each replicate represents 40 seedlings.Asterisks indicate significant differences compared to the WT(P＜0.05,t-test).

Figure 8.Cellular Pi content in shoots and roots of 14-d-old seedlings of AtPAP12(A)and AtPAP26(B)overexpression lines grown on medium of P+,P-and P-supplemented with different amount of ADP or Fru-6-PThe experiments were repeated for three times with similar results.A one-way ANOVA analysis was carried out for the whole dataset and post hoc comparisons were conducted using the SPSS Tukey HSD test at P＜0.05 level.No significant difference was found among the plants grown under a given Pi condition.

Because AtPAP10,AtPAP12,and AtPAP26 have similar and also different biochemical behaviours(Veljanovski et al.2006;Tran et al.2010b;Wang et al.2011,and this work),we wanted to determine whether they function differently during plant responses to Pi deprivation.Previously,it has been reported that the growth of pap26 single mutant or pap12/26 double mutant was reduced under Pi deficiency condition(Hurley et al.2010;Robinson et al.2012).But no results have been published on the growth phenotypes of pap10/12/26triple mutant as well as AtPAP12-or AtPAP26-overexpressing lines.Under Pisufficient conditions on agar plates,plant morphologies were similar among the WT,all mutants,and all overexpressing lines.This implicates that AtPAP10,AtPAP12,and AtPAP26 may not play direct roles in plant growth and development under normal growth condition.It also suggests that the functions of these three AtPAPs(and other APases)may overlap in regulating plant growth and development.Thus,even when all of three were mutated,detrimental effects were notapparent.Finally,because the APase activity of AtPAP26,and probably also of AtPAP10 and AtPA12,is strongly inhibited by Pi(in the presence of 5 mM Pi,the activity of AtPAP26 is reduced to 30%of the control)(Veljanovski et al.2006),these three APase may not function well in shoots and roots under Pi sufficiency(The Pi concentration in the plant cells is in the range of 1–10mM under normal growth condition).

Plant growth was reduced by the mutation of AtPAP10 on all types of P-media tested,whether supplemented with organophosphates or not.The difference in growth between atpap10(also the AtPAP10 overexpressing line)and the WT even on P-medium was probably due to the residual organophosphates present within the agar used in our experiments.Another possibility is that AtPAP10 might help recycle Pi esters that leaked from the plants during growth on P-medium.In contrast,mutation of AtPAP12 reduced plant growth only on P-medium or P-medium with 150 μM Fru-6-P,and mutation of AtPAP26 had little inhibitory effect,except on root biomass on P-medium with 150 μM Fru-6-P.These results differ from those of Hurley et al.(2010)and Robinson et al.(2012)who reported a significant reduction in the growth of pap12 and pap26 single mutants,and pap12/26 double mutant grown on P-medium in agar plates.This discrepancy may be due to the agars used in growth assay,which might have contained different amounts and types of residual organophosphates.Our results also indicated a lack of correlation between the growth phenotype and level of the intracellular APase activity.On the majority of Pi-deficient media tested(Figures 5,S2),the growth of pap12 and pap26 mutants,which have significantly reduced intracellular APase activities,was much less inhibited than that of the pap10 mutant,which has no detectable reduction of intracellular APase activity.One explanation is that our measurement of shoot and root biomass was performed on 14-or 16-day-old seedlings.At this early developmental stage,the intracellular APases may not affect plant growth and development because intracellular APases are thought to function in the mobilizing and recycling of Pi from old tissues to young,growing tissues.

Interestingly,while the knockout of AtPAP12 and AtPAP26 only moderately or slightly inhibited plant growth,overexpression of these two genes caused the plants to grow better under various P-conditions.One explanation is that ADP and Fru-6-P may not be the effective substrates for AtPAP12 and AtPAP26 in vivo.If that is correct,loss of AtPAP12 or AtPAP26 would not significantly affect plant growth on P-medium supplemented with ADP or Fru-6-P.Another possibility is that in the pap12/26 double mutant,the remaining AtPAP10 activity is sufficient to compensate for the loss of AtPAP12 and AtPAP26 in scavenging Pi from ADP or Fru-6-P.The AtPAP12 and AtPAP26 overexpressing lines,however,exhibit a large increase in both intracellular APase activity and secreted APase activity.When overproduced,the secreted AtPAP12 or AtPAP26 APase may be able to release sufficient Pi from ADP or Fru-6-P,even if these two compounds are not their optimal substrates.This explanation is supported by the fact that BCIP stains the root surfaces of AtPAP12 and AtPAP26 overexpressing lines as intensely as the root surfaces of AtPAP10 overexpressing lines(data not shown),although BCIP is not an effective substrate for AtPAP12 or AtPAP26(Figure 3C,D).

Pi starvation-induced APases are generally believed to function in internal Pi recycling or to increase the availability of Pi for plant uptake by releasing Pi from external organophosphates.When we compared the Pi contents in the shoots and roots among the WT,the pap mutants,and AtPAP overexpressing lines grown under various Pi conditions,however,we did not find a correlation between the Pi content and plant growth phenotype(Figures 5,7,8 and S3).In particular,for all the overexpressing lines grown under the six Pi conditions tested,the Pi contents in their shoots and roots were essentially the same as those of the WT(Figure 8).In addition,although the shoot fresh weight was about 3-to 4-times higher for both the WT and overexpressing lines when they were grown on P-medium supplemented with 50 μM ADP or Fru-6-P rather than on P-medium(Figure 7),the Pi contents in all these plants were also similar(Figure 8).These results suggest that an increased Pi level is not a prerequisite for the improved growth under Pi-deficient conditions.

One possible explanation for why the overexpressing lines do not accumulate more Pi than the other lines is that the extra Pi taken up may be quickly converted to organophosphates such as phospholipids and RNAs,which could not be reflected in our analysis.This hypothesis can be tested by comparing the total amount of phospholipids or RNAs between the WT and overexpressing lines.Another possibility is that the Pi released in the rhizosphere by the overproduced APases may trigger a signalling cascade that alters the plant responses to Pi deprivation.This hypothesis is supported by research with the Arabidopsis PHO1 gene,which is involved in the translocation of Pi from roots to shoots(Poirier et al.1991;Hamburger et al.2002).A pho1 null mutant has reduced Pi level in the shoots and displays a stunted shoot growth.However,Rouached et al.(2011)reported that Arabidopsis plants with reduced PHO1 expression have shoot growth similar to that of the plants grown under Pi sufficiency,even though their leaves are strongly Pi deficient.The authors hypothesized that the Pi starvation-induced inhibition of shoot growth might not be a direct consequence of Pi deficiency but could be a result of extensive reprogramming of gene expression triggered by Pi deprivation.In PHO1 underexpressors,translocation of a Pi deficiency signal from root to shoot might be impaired,resulting in a suppression of reprogramming of gene expression.A similar scenario may also occur in the AtPAP-overexpressing lines,i.e.,the extra Pi released from organophosphates in the rhizosphere may be sufficient to suppress the plant Pi starvation responses,including the reduced shoot growth.

As a final explanation for why the overexpressing lines do not accumulate more Pi than the other lines,the overexpression of these three AtPAPs may alter cell wall properties by modifying some cell wall-associated enzymes involved in cell wall loosing.In fact,these three AtPAPs have been reported as cell wallassociated APases(Bayer et al.2006;Robinson et al.2012)and can use phosphorylated-serine and-tyrosine as their substrates(Tran et al.2010b;Wang et al.2011).The increased activity of these APases on the cell walls may promote cell expansion under Pi deprivation and thus enhance plant growth.

MATERIALS AND METHODS

Plant material and growth conditions

Arabidopsis thaliana ecotype for all plants used in this study were Columbia(Col-0).Surface-sterilized seeds were placed in plates containing solidified full-strength MS medium with 1%sucrose and 0.55%(W/V)agar(Sigma,Catalog No.A1269).The pH of the medium was adjusted to 5.8.This Pi-sufficient medium is referred as P+medium.For the Pi-deficient medium(referred as P-),KH2PO4in the P+medium was replaced by K2SO4.After seeds were stratified at 4°C for 2 days,the plates were placed horizontally in a growth room with a 16-h light/8-h dark photoperiod(100 μmol·m-2·s-1)at 23 °C.For examination of root surface APase by BCIP staining,culture plates were placed vertically in the same growth room,and the concentration of agar used in culture media was increased to 1.2%.

Identification of homozygous T-DNA insertion lines

T-DNA insertion lines were obtained from The Arabidopsis Biological Resource Center.Genomic DNA was extracted from each T-DNA line and analysed by PCR using the primers specific for each T-DNA insertion.The mRNA expression of each AtPAP gene in the T-DNA insertion lines was analyzed by a semiquantitative RT-PCR method,and the sequences of the primers are listed in Table S3.

Quantitative analysis of total APase activities

Total proteins were extracted from shoots and roots or collected from liquid culture media as described(Trull and Deikman 1998).Quantitative analysis of total APase activities was performed according to Wang et al.(2011).

Quantitative analysis of root-associated APase activities

The root-associated APase activity was measured according to Boutin et al.(1981)and McLachlan(1980)with some modifications.The roots were excised from two 9-day-old seedlings,and the roots lengths were measured.The excised roots were thoroughly rinsed with distilled water three times,then transferred to a 1.5-mL Eppendorf tube containing 650 μL of reaction buffer(10 mM MgCl2,50 mM NaAc,pH 4.9)and 50 μL of 1.0 mg/mL pNPP;the tubes with roots were then incubated at 37°C for 1 h before the reaction was terminated by addition of 100 μL of 2%SDS.Absorbance was measured spectrophotometrically at 410 nm,and root-associated APase activity was expressed as A410/cm root.

In-gel assays of APase profiles

The protein extraction and in-gel assay for APase profiles were performed essentially as described(Trull and Deikman 1998).The proteins extracted from shoots and roots of Arabidopsis seedlings or collected from liquid culture media were lyophilized at-50°C.The lyophilized samples were resuspended in 200 μL of ice-cold extraction buffer.The proteins were quantified and separated on a 10%non-reducing SDSPAGE gel at 4°C.The gels were gently washed in cold,distilled water(10 min per wash and a total of six washes)to remove the SDS.The gels were then gently shaken in a buffer containing 50 mM Na-acetate(pH 4.9)and 10 mM MgCl2(two times,15 min each time)to allow the proteins to renature.After equilibrium with the buffer,gels were stained for APase activity in a buffer containing 0.5 mg/L Fast Black K salt and 0.3 mg/L β-naphthyl acid phosphate at 37°C for 1–3 h.

Construction of plant transformation vectors and plant transformation

The generation of AtPAP10 overexpressing lines has been described by Wang et al.(2011).To obtain AtPAP12 and AtPAP26 overexpressing lines,we cloned genomic sequences of AtPAP12 and AtPAP26 from WT Arabidopsis seedlings by PCR.The AtPAP12 gene was amplified using primers 5’-GAGGATCCAAAGGTTGATTAAAAATAAAATAA-3’and 5’-TCGAGCTCGATTAGTGAAATGGTATATTAACTTG-3’,and the AtPAP26 gene was amplified using primers 5’-GGGGATCCAGCGATATCGTGGAAGT-3’and 5’-GGGAGCTCGTTTGAAACCTCGCAAAGAG-3’.During amplification,BamHI and SacI restriction sites were added to the 5’and 3’ends of the PCR fragments.The amplified PCR products were first cloned to vector pGM-T and sequenced,and they were then inserted into the plant expression vector pZH01 after the CaMV 35S promoter using BamHI and SacI sites.Both gene constructs were transferred into Agrobacterium tumefaciens strain GV3101 and transformed into WT Arabidopsis plants by the floral dip method(Clough and Bent 1998).The resultant transgenic plants were first selected on MS medium containing 50 μM hygromycin and then confirmed by PCR analysis.

Quantitative real-time PCR analysis

Quantitative real-time RT-PCR was performed as described(Wang et al.2011).The genes and the primers used for detection of their mRNA expression are listed in Supplemental Table S2 and S3.

Quantitative analysis of cellular Pi content

The cellular Pi contents were determined using the method described by Ames(1966).

ACKNOWLEDGEMENTS

We thank Dr.Daowen Wang of the Institute of Genetics and Developmental Biology,Chinese Academy of Sciences,for helpful discussions.We also thank the Arabidopsis Biological Resource Center for providing the T-DNA insertional lines.This work was supported by the National Natural Science Foundation of China(31370290 to D.L.and 30971554 to X.D.),and the Ministry of Agriculture of China(2014ZX0800932B to D.L.).

Ames BN(1966)Assay of inorganic phosphate,total phosphate and phosphatases.Method Enzymol 8:115–118

Bayer EM,Bottrill AR,Walshaw J,Vigouroux M,Naldrett MJ,Thomas CL,Maule AJ(2006)Arabidopsis cell wall proteome defined using multidimensional protein identification technology.Proteomics 6:301–311

Bieleski RL(1973)Phosphate pools,phosphate transport and phosphate availability.Annu Rev Plant Physiol 141:988–999

Boutin JP,Provot M,Roux L(1981)Effect of cycloheximide and renewal of phosphorus supply on surface acid-phosphatase-activity of phosphorus deficient Tomato roots.Physiol Plant 51:353–360

Bozzo GG,Dunn EL,Plaxton WC(2006)Differential synthesis of phosphate-limitation inducible purple acid phosphatase isozymes in tomato(Lycopersicon esculentum)suspension cells and seedlings.Plant Cell Environ 29:303–313

Bozzo GG,Raghothama KG,Plaxton WC(2002)Purification and characterization of two secreted purple acid phosphatase isozymes from phosphate-starved tomato(Lycopersicon esculentum)cell cultures.Eur J Biochem 269:6278–6286

Clough SJ,Bent AF(1998)Floral dip:A simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana.Plant J 16:735–743

del Pozo JC,Allona I,Rubio V,Leyva A,de la Pena A,Aragoncillo C,Paz-Ares J(1999)A type 5 acid phosphatase gene from Arabidopsis thaliana is induced by phosphate limitation and by some other types of phosphate mobilising/oxidative stress conditions.Plant J 19:579–589

Hamburger D,Rezzonico E,MacDonald-Comber Peté tot J,Somerville C,Poirier Y(2002)Identification and characterization of the Arabidopsis PHO1 gene involved in phosphate loading to the xylem.Plant Cell 14:889–902

Haran S,Logendra S,Seskar M,Bratanova M,Raskin I(2000)Characterization of Arabidopsis acid phosphatase promoter and regulation of acid phosphatase expression.Plant Physiol 124:615–626

Hurley BA,Tran HT,Marty NJ,Park J,Snedden WA,Mullen RT,Plaxton WC(2010)The dual-targeted purple acid phosphatase isozyme AtPAP26 is essential for efficient acclimation of arabidopsis to nutritional phosphate deprivation.Plant Physiol 153:1112–1122

Kuang R,Chan K,Yeung E,Lim BL(2009)Molecular and biochemical characterization of AtPAP15,a purple acid phosphatase with phytase activity,in Arabidopsis.Plant Physiol 151:199–209

Li M,Tadano T(1996)Comparison of characteristics of acid phosphatases secreted from roots of lupin and tomato.Soil Sci Plant Nutr 42:753–763

Li D,Zhu H,Liu K,Liu X,Leggewie G,Udvardi M,Wang DW(2002)Purple acid Phosphatases of Arabidopsis thaliana-Comparative analysis and differential regulation by phosphate deprivation.J Biol Chem 277:27772–27781

Liang C,Tian J,Lam H,Lim BL,Yan X,Liao H(2010)Biochemical and molecular characterization of PvPAP3,a novel purple acid phosphatase isolated from common bean enhancing extracellular ATP utilization.Plant Physiol 152:854–865

Lloyd JC,Zakhleniuk OV,Raines CA(2001)Identification of mutants in phosphorus metabolism.Annals Applied Biol 138:111–115

Lung SC,Leung A,Kuang R,Wang Y,Leung P,Lim BL(2008)Phytase activity in tobacco(Nicotiana tabacum)root exudates is exhibited by a purple acid phosphatase.Phytochemistry 69:365–373

Mclachlan KD(1980)Acid phosphatase activity of intact roots and phosphorus nutrition in plants.1.Assay conditions and phosphatase activity.Aust J Agr Res 31:429–440

Miller SS,Liu J,Allan DL,Menzhuber CJ,Fedorova M,Vance CP(2001)Molecular control of acid phosphatase secretion into the rhizosphere of proteoid roots from phosphorus-stressed white lupin.Plant Physiol 127:594–606

Olczak M,Morawiecka B,Watorek W(2003)Plant purple acid phosphatases—genes,structures and biological function.Acta Biochim Pol 50:1245–1256

Ozawa K,Osaki M,Matsui H,Honma M,Tadano T(1995)Purification and properties of acid phosphatase secreted from Lupin roots under phosphorus-deficiency conditions.Soil Sci Plant Nutr 41:461–469

Poirier Y,Thoma S,Somerville C,Schiefelbein J(1991)A mutant of Arabidopsis deficient in xylem loading of phosphate.Plant Physiol 97:1087–1093

Raghothama KG(2000)Phosphate transport and signaling.Curr Opin Plant Biol 3:182–187

Richardson AE(2009)Regulating the phosphorus nutrition of plants:Molecular biology meeting agronomic needs.Plant Soil 322:17–24

Robinson WD,Park J,Tran HT,Del Vecchio HA,Ying S,Zins JL,Patel K,McKnight TD,Plaxton WC(2012)The secreted purple acid phosphataseisozymes AtPAP12 and AtPAP26 play a pivotal role in extracellular phosphate scavenging by Arabidopsis thaliana.J Exp Bot 63:6531–6542

Silberbush M,Shomerilan A,Waisel Y(1981)Root surface phosphataseactivity in ecotypes of aegilops-peregrina.Physiol Plant 53:501–504

Tran HT,Hurley BA,Plaxton WC(2010a)Feeding hungry plants:The role of purple acid phosphatases in phosphate nutrition.Plant Sci 179:14–27.

Tran HT,Qian WQ,Hurley BA,She YM,Wang DW,Plaxton WC(2010b)Biochemical and molecular characterization of AtPAP12 and AtPAP26:the predominant purple acid phosphatase isozymes secreted by phosphate-starved Arabidopsis thaliana.Plant Cell Envir 33:1789–1803

Trull MC,Deikman D(1998)An Arabidopsis mutant missing one acid phosphatase isoform.Planta 206:544–550

Vance CP,Uhde-Stone C,Allan DL(2003)Phosphorus acquisition and use:Critical adaptations by plants for securing a nonrenewable resource.New Phytol 157:423–447

Veljanovski V,Vanderbeld B,Knowles VL,Snedden WA,Plaxton WC(2006)Biochemical and molecular characterization of AtPAP26,a vacuole purple acid phosphatase up-regulated in phosphatedeprived Arabidopsis suspension cells and seedlings.Plant Physiol 142:1282–1293

Wang X,Wang Y,Tian J,Lim BL,Yan X,Liao H(2009)Overexpressing AtPAP15 enhances phosphorus efficiency in soybean.Plant Physiol 151:233–240

Wang L,Li Z,Qian W,Guo W,Gao X,Huang L,Wang H,Zhu H,Wu J,Wang D,Liu D(2011)The Arabidopsis purple acid phosphatase AtPAP10 is predominantly associated with the root surface and plays an important role in plant tolerance to phosphate limitation.Plant Physiol 157:1283–1299

Wang L,Liu D(2012)Arabidopsis purple acid phosphatase 10 is a component of plant adaptive mechanism to phosphate limitation.Plant Signal Behav 7:306–310

Yuan H,Liu D(2008)Signaling components involved in plant responses to phosphate limitation.J Integr Plant Biol 50:849–859

Zhu H,Qian W,Lu X,Li D,Liu X,Liu K,Wang D(2005)Expression patterns of purple acid phosphatase genes in Arabidopsis organs and functional analysis of AtPAP23 predominantly transcribed in flower.Plant Mol Biol 59:581–594

SUPPORTING INFORMATION

Additional supporting information can be found in the online version of this article:

Figure S1.RT-PCR analyses of AtPAP gene expression in various T-DNA insertion lines

Figure S2.Growth phenotype of 14-day-old seedlings of the WT and pap mutants that were grown on P+and various P-media

Figure S3.Cellular Pi content in shoots and roots of WT,pap10,pap12,pap26,pap10/12,pap12/26 and pap10/12/26 grown on medium of p+,p-and p-supplemented with different amount of ADP or Fru-6-P

Figure S4.Relative expression of the AtPAP10,AtPAP12,and AtPAP26 genes in the overexpressing transgenic lines

Table S1.Positions of T-DNA insertion in each allele of the AtPAP genes

Table S2.Sequences of the primers used for expression analysis of AtPAP genes by quantitative real-time RT-PCR

Table S3.Sequences of the primers used for expression analysis of AtPAP genes