Identification of Purple Acid Phosphatases in Chickpea and Potential Roles of CaPAP7 in Seed Phytate Accumulation

Purple acid phosphatases (PAPs) play important roles in phosphate (Pi) acquisition and utilization. These PAPs hydrolyze organic Phosphorus (P) containing compounds in rhizosphere as well as inside the plant cell. However, roles of PAPs in one of the most widely cultivated legumes, chickpea (Cicer arietnum L.), have not been unraveled so far. In the present study, we identified 25 putative PAPs in chickpea (CaPAPs) which possess functional PAP motifs and domains. Differential regulation of CaPAPs under different nutrient deficiencies revealed their roles under multiple nutrient stresses including Pi deficiency. Interestingly, most of the CaPAPs were prominently expressed in flowers and young pods indicating their roles in flower and seed development. Association mapping of SNPs underlying CaPAPs with seed traits revealed significant association of low Pi inducible CaPAP7 with seed weight and phytate content. Biochemical characterization of recombinant CaPAP7 established it to be a functional acid phosphatase with highest activity on most abundant organic-P substrate, phytate. Exogenous application of recombinant CaPAP7 enhanced biomass and Pi content of Arabidopsis seedlings supplemented with phytate as sole P source. Taken together, our results uncover the PAPs in chickpea and potential roles of CaPAP7 in seed phytate accumulation.

Chickpea is the world's third-most important cultivated legume crop. Around 6.2 mha of arable land in India is under chickpea cultivation. However, majority of chickpea growing areas are either marginal or sub-marginal lands with sub-optimal Pi levels. Thus, Pi deficiency is a critical constraint for chickpea production in India 29 . Amidst this scenario, organic-P mobilizing PAPs can be an effective resource to improve Pi utilization efficiency of chickpea. However, information on chickpea PAPs (CaPAPs) is missing so far. Therefore, in the present study, we first identified 25 novel CaPAPs and investigated their potential roles in Pi homeostasis. Many of the CaPAPs were found to be differentially regulated by Pi deficiency. However, unlike some of the past studies, CaPAPs did not show specific induction to Pi deficiency. This intrigued us to explore broader roles of CaPAPs in chickpea. Interestingly, majority of the CaPAPs were highly expressed in reproductive tissues. Further, association studies revealed strong association of low Pi inducible CaPAP7 with seed weight and phytate content. Therefore, in the present study, we investigated CaPAP7 for its possible roles in low Pi tolerance and regulation of seed weight/ phytate content.

Methods
Identification of PAPs in chickpea. The PAP sequences from 17 different organisms (Table S1) were identified from sequence databases like TAIR, TIGR, Plant GDB, Uniprot; using phrase "purple acid phosphatase". The organisms selected were Arabidopsis thaliana, Oryza sativa, Solanum lycospersicum, Medicago trunculata, Hordeum vulgare, Physcomitrella patens, Vitis vinifera, Caenorhabditis elegans, Homo sapiens, Dictyostelum discordeum, Vigna radiata, Rhizobium galegae, Phaseolus vulgaris, Glycine max, Zea mays, Solanum tuberosum and Chlamydomonas reinharbtii. The sequences so obtained were searched for the presence of metallophos domain using SMART tool (http://smart.embl-heidelberg.de/). Total sequences retrieved with metallophos domain were 348 (Text S1). The sequences with metallophos domain were selected, aligned and used to generate a Hidden Markov Model (HMM). The HMM profile was used to identify PAPs in chickpea using HMMER search in chickpea (Kabuli) protein database (http://ceg.icrisat.org/gt-bt/ICGGC/GenomeManuscript.htm). Sequences so obtained were then screened for the presence of metallophos domain and five conserved motifs (DXG/GDXXY/ GNH(D/E)/VXXH/GHXH), and finally 25 putative CaPAPs were identified. These PAPs were aligned with Arabidopsis PAPs (AtPAPs) using Clustal X2 and annotated in accordance to their homology with corresponding AtPAPs.
Expression analysis of CaPAPs in different plant tissues. We utilized the transcriptome data from CTDB (chickpea transcriptome database) (http://www.nipgr.res.in/ctdb.html) generated by Garg et al. 31 in order to identify the expression pattern of CaPAPs in different tissues (root, shoot, flower bud and young pod). Heat map was generated on RPM (number of unique reads mapped to each transcript per million) values using MeV 4.6.0 tool (http://mev.tm4.org/#/welcome).

Trait association analysis of CaPAPs.
For association mapping, seed weight (100 seed weight in gm) and phytic acid content (in mg/100gm) data were retrieved from Kujur et al. 32 and Joshi-Shah and Reddy 33 , respectively. CaPAPs based SNP genotyping information (MAF 5%) as well as kinship matrix (K), PCA (P) data and population structure ancestry coefficient (Q matrix) generated from 92 chickpea accessions (association panel) were analysed using TASSEL v5.0, as described 32 . The potential SNP loci in the diverse coding and non-coding sequence components of CaPAPs revealing significant association with seed weight and phytic acid traits at a highest R 2 (degree of SNP marker-trait association) and lowest FDR adjusted P-values (threshold P ≥ 10 −6 ) were selected.

Plant growth condition and nutrient deficiency treatments. Chickpea seeds of accession ICC 4958
were used in all the experiments. For generating nutrient deficiency conditions, seedlings were raised in Hoagland media according to the conditions and protocol as described 34 . Briefly, plant growth conditions were maintained at 12/12 h photoperiod, 23/18 °C, 200-300 µM photons/m 2 /s photon density and ~70% relative humidity. Tissues were harvested after 7 and 15 days of treatment and unless stated otherwise experiments were performed in three biological replicates.
qRT-PCR based analysis of gene expression. Gene expression analysis was done as described, previously 34 . Briefly, root samples were taken at 7 days and 15 days after treatment for each nutrient deficiency. Tissues were frozen in liquid nitrogen and stored at −80 °C. The TRIzol method was used to extract total RNA from root. Prior to cDNA synthesis, DNaseI treatment was given to get rid of genomic DNA contamination from RNA. cDNA was synthesised using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to manufacturer's instructions.
The primers for quantitative real time PCR (qRT-PCR) were designed using PRIMER EXPRESS version 2.0 (PE Applied Biosystems TM , USA) with default parameters from coding region of the genes (Table S2). Each primer pair was analysed for its specificity using NCBI BLAST. The reaction was performed in Applied Biosystems 7500 Fast Real Time PCR. The relative expression was calculated using the ΔΔCt method. The elongation factor 1-α (AJ004960) was used as endogenous control and three biological replicates were considered for the experiment. To test the level of significance, Student's t-test was applied.
Cloning and subcellular localization of CaPAP7. The coding region of CaPAP7 was cloned in pSITE3CA vector to express it as N-terminal YFP fusion protein (YFP:CaPAP7). Plasmid carrying this construct was then transformed into onion epidermal cells as described 25 . Fluorescence was visualised under confocal microscope, AOBS TCS-SP2 (Leica, Germany).
Biochemical characterization of recombinant CaPAP7. For biochemical characterization, ORF sequence of CaPAP7 was amplified using gene-specific primers (Table S2). The PCR product of 1008 bp was cloned into pET28a vector and confirmed by sequencing. BL21 (DE3) pLysS cells were used for protein induction. The recombinant protein was induced with 0.3 mM IPTG at 16 °C for 12-16 h. The induced protein was purified using affinity chromatography (Ni-NTA). To confirm the identity of induced protein, coomassie-stained protein bands were digested with trypsin as per manufacturer's protocol and then identified by 4000Q TRAP LC/MS/MS.
The phosphatase activity assays were performed to study optimum temperature, pH, cofactor and substrate for CaPAP7 as described 25 . The optimum temperature for the PAP activity was measured over a temperature range of 25 °C to 60 °C in 50 mM sodium acetate buffer (pH 4.9), MgCl 2 (5 mM) using 10 mM para-nitrophenylphosphate (pNPP), as substrate for 30 min. To identify the optimum pH for CaPAP7 activity, a pH range of 4.0 to 8.0 was set using different buffers with varying pH and pNPP (10 mM) as substrate at 37 °C for 30 min. In order to determine the substrate specificity, CaPAP7 activity was assayed on different substrate namely, ATP, ADP, AMP, PPi, P-serine, P-threonine, glucose-6-P, fructose-6-P, pNPP, 2-deoxyriboseP and phytic acid at 37 °C for 30 min. In order to identify cofactors and inhibitors of CaPAP7, activity assays were performed with 5 mM of chloride salts of different metal ions and sodium salts of different cations as described earlier 25 .
Gene expression profile of CaPAP7 in chickpea genotypes with contrasting seed weight. Four contrasting accessions for seed weight and phytate content were selected based on the information from previous reports 32,33 . In order to reveal the role of CaPAP7 in phytate remobilisation, we analysed the expression of CaPAP7 in young pod of these contrasting accessions. Tissues for young pod were collected in three replicates pooled from multiple plants, frozen in liquid nitrogen and thereafter stored at −80 °C. RNA was isolated using TRIzol and cDNA was prepared as described above. The qRT-PCR was done and relative expression was calculated using 2 −ΔCt method.

Results
PAPs form multigene family in chickpea. Using sequence homology approach and conserved domain analysis, we identified 25 putative PAPs in chickpea genome which were annotated according to their homology to Arabidopsis PAPs (Fig. 1A). Of these, complete metallophos domain was present in twenty-three while two (CaPAP23a and CaPAP23b) have partial domain. CaPAP23a had only pur_ac_phosphN terminal domain and CaPAP23b had only metallophos C domain which is essential for the phosphatase activity (Fig. 1B). The sequence analysis revealed that 21 CaPAPs have all the five blocks of conserved motifs (DXG/GDXXY/GNH(D/E)/VXXH/ GHXH) and seven metal ligating residues which are characteristics of known PAPs. In CaPAP16, 28 and 29 "Y" residue is replaced by "F" in the second block (GDXXY). Among the remaining PAPs, CaPAP16, 23b lack the fourth block (VXXH), CaPAP18a lacks fifth block (GHXH), whereas CaPAP23a has two blocks missing; third (GNH(D/E) and fifth (GHXH) ( Table S3). But both the sequences showed significant homology with known PAPs. So we considered them as putative PAPs.
Based on phylogenetic relationship between chickpea and Arabidopsis PAPs, CaPAPs were classified into three major (I, II and III) groups as per earlier classification 7 (Fig. 1A). The twenty-five CaPAPs were divided into seven subgroups (Ia-2, Ib-1, Ib-2, IIa, IIb, IIIa and IIIb) except subgroup Ia-1 where none of the CaPAPs was grouped. Further, phylogenetic analysis of CaPAPs with soybean, Arabidopsis and rice PAPs together suggested that CaPAPs are closer to dicots PAPs as only few of them were grouped with rice PAPs (Fig. S1).
Further, CaPAPs were found to be localised to different chromosomes, whereas five genes (CaPAP15a, 17a, 17b, 18c and 22) were present on scaffolds (scaffold 128, 88, 40, 6367 and 1710, respectively) ( Table S4). Analysis of gene structure of CaPAPs showed that number of exons varies from 2 (CaPAP9) to 12 (CaPAP1a, 1b, 27a and 27b) (Table S4). Molecular mass of identified PAPs varies from 17.33 kDa (CaPAP23a) to 73.91 kDa (CaPAP9). On prediction of secretory or non-secretory PAPs, we found the presence of signal peptide in 15 CaPAPs namely, CaPAP1a, 1b, 3, 7, 8, 9, 15a, 16, 18b, 18c, 21, 22, 24, 26 and 29, whereas no signal peptide was identified in remaining 10 PAPs i.e. CaPAP10, 15b, 17a, 18a, 20, 23a, 23b, 27a, 27b and 28. In order to correlate the nature of signal peptide, we predicted the subcellular localization of CaPAPs with the help of CELLO tool (http:// cello.life.nctu.edu.tw/). It turned out that most of the CaPAPs i.e. 22 out of 25 were either secretory or localised to lysosomes. CaPAP7 and 16 were predicted to be localised to cytoplasm while CaPAP28 to plasma membrane. Interestingly, CaPAP18b was predicted to be localised to nucleus and also has potential secretory nature (Table S5). Since, some of the PAPs have been reported to be localised in peroxisomes 30 , we determined presence of PTS1 (Peroxisome Targeted Signal) in CaPAPs. Interestingly, only low P responsive CaPAP17a showed presence of PTS in its C terminal region. Furthermore, glycosylation sites were present in all the CaPAPs varying in numbers 1 to 8 (Table S5).
As PAPs are largely reported as Phosphate Starvation Response (PSR) genes, we analysed CaPAPs promoters (3 kb upstream) for any putative cis-acting elements which can act as binding sites for transcription factors involved in Pi deficiency response. P1BS (PHR1 binding sites) elements are considered to be highly Pi responsive as AtPHR1 (master regulator of P deficiency response) binds to this region and regulate its downstream targets. Analysis of promoter regions of all chickpea PAPs for the presence of P1BS elements revealed that P1BS elements were present in the promoter region of 18 CaPAPs including CaPAP1a, 1b, 3, 8, 9, 10, 15a, 15b, 17a, 17b, 18c, 21, 22, 23a, 23b, 27b, 28 and 29 while we did not find P1BS elements in seven CaPAPs including CaPAP7, 16, 18a, 18b, 20, 26 and 27a (Table S5).
CaPAPs are differentially expressed under Pi deficiency. Since most of CaPAPs have P1BS elements in their promoter region, therefore, we studied their expression behaviour under Pi deficiency. Most of the CaPAPs (12 out of 25 genes which include CaPAP1a, 1b, 7, 8, 10, 15a, 17a, 20, 21, 23b, 27a and 28) showed significant upregulation except CaPAP18c which was downregulated after 7 days of deficiency (Fig. 2). CaPAP17b did not show detectable expression. The expression of ten genes (CaPAP3, 8, 10, 18a, 18c, 20, 22, 26, 27a and 29) found to be downregulated after 15 days of deficiency except two genes, CaPAP9 and CaPAP18b which were found upregulated. This indicated that most of the chickpea PAP genes are early Pi deficiency inducible, while CaPAP9b and CaPAP18b were late inducible to the Pi deficiency (Fig. 2). Moreover, four genes including CaPAP1b, 8,15 and 27a were found exclusively upregulated under Pi deficiency showing their Pi specific behaviour (Figs S2 and  S3). In addition to low Pi; CaPAPs also showed differential expression under low N, K, Fe and Zn indicating their roles in these deficiencies (Figs S2 and S3). CaPAPs showed preferential high expression in reproductive tissues. Spatial expression of CaPAPs in different tissues (root, shoot, mature leaf, flower bud and young pod) retrieved from CTDB database showed that most of the PAPs express in all the tissues but have exceptionally high expression in flower bud and young pod (Fig. 3). Further, except four CaPAPs, other root expressing ones did not express as high in roots as they did in reproductive tissues. Among all the tissues, CaPAPs showed least expression in shoot. This strong expression in flower bud and young pod indicates their important functions during reproductive development.
CaPAPs showed strong association with seed phytate content and seed weight in diverse chickpea accessions. Since most of the CaPAPs showed high expression in reproductive tissues (flowering bud and young pod), we performed candidate gene based association mapping in diverse chickpea accessions for phytate content and seed weight. For this, genotyping data of SNPs from diverse coding and non-coding  sequence components of CaPAPs along with phenotypic data (phytic acid and seed weight) of 92 accessions were collected. SNPs underlying CaPAPs were retrieved from 92 chickpea accessions and associated with seed weight data as described, previously 32 . Interestingly, we found strong association of a SNP (G/A) of Ca_05071 (CaPAP7) with seed weight (P value 1.8 × 10 −9 ) and seed phytate content (P value 1.4 × 10 −6 ). The proportion of phenotypic variation explained (PVE) for seed weight and phytate content by this SNP was found to be 25 and 19%, respectively (Table 1, Fig. 4). Additionally, another DRR (down-stream regulatory region) SNP, (A/C) in Ca_06712 (CaPAP26) was associated with seed weight trait (28% PVE with 2.4 × 10 −10 ) and phytate content (21% PVE with P value 1.8 × 10 −7 ). Closer analysis revealed that allele 'G' in URR of CaPAP7 mapped on chromosome 6 was associated with high seed weight and phytate, whereas allele ' A' is associated with low seed weight and low phytate content (Table 1). Similarly, allele ' A' in DRR of CaPAP26 mapped to chromosome 7 was associated with high seed weight and high phytate, whereas allele 'C' is associated with low seed weight and low phytate content.
CaPAP7 is differentially regulated in contrasting genotypes for seed weight and phytate content. Due to its induction in response to low Pi, conservation of all key residues and strong association with seed weight and seed phytate content traits, CaPAP7 was selected for further characterization. In order to validate association of URR SNP of CaPAP7 with seed weight and phytate content, we analysed the expression pattern of CaPAP7 in four accessions of chickpea varying in seed weight and phytate content (ICC16374, ICCV2, ICC12155 and ICC8151). Further, seed weight and phytate content were found strongly correlated with each other. Therefore, we analysed the expression pattern of CaPAP7 in young pod where phytate synthesis/accumulation is   expected to be high. CaPAP7 showed higher expression in accessions with low seed weight and phytate content. While its expression was lower in accessions with higher seed weight and phytate content (Fig. 5). Therefore, a negative correlation of CaPAP7 expression was obtained with the seed weight (r = −0.82) and seed phytate content (r = −0.99). This data further strengthen the important role of CaPAP7 in chickpea seed phosphate/phytate accumulation.
CaPAP7 is localized in cytoplasm and encodes a functional acid phosphatase. In-silico tools predicted CaPAP7 to be localised in the cytoplasm. To confirm this in vivo, we cloned CaPAP7 CDS in pSITE3CA vector to produce an YFP fused protein and analysed its localisation in onion epidermal cells. YFP fluorescence analysis confirmed its cytoplasmic localisation as predicted (Fig. S4). We next studied, whether CaPAP7 is a functional enzyme to further contemplate on its functions. Recombinant protein was produced in bacteria and purified. A band of expected size ~38 kDa was detected on SDS-page (Fig. S5). The identity of induced recombinant CaPAP7 was confirmed with LC-MS/MS (MASCOT score 120) (Fig. S6). CaPAP7 showed activity in the temperature range of 25-45 °C and pH range of 4-6 ( Fig. 6). Its high phosphatase activity in acidic pH confirmed it as APase. Further, activity assays with different divalent cations and anions revealed no specific cofactor preference of CaPAP7. However, high phosphate and carbonate concentration seemed to be slightly inhibitory for phosphatase activity of CaPAP7 (Fig. 6C). The enzyme activity assay showed that CaPAP7 has a broad substrate specificity hydrolysing different Pi containing compounds  33 ). Error bars represent SE (n = 2) (C) Average seed weight (100 seed weight in g) of contrasting chickpea accessions (Kujur et al. 32 ). Error bars represent SE (n = 3) CaPAP7 expression was in negative correlation with the seed weight (r = −0.82) and seed phytate content (r = −0.99). Different letters at the top of bar indicates different significant classes determined by oneway ANOVA followed by Duncan's multiple range test.
Exogenous application of recombinant CaPAP7 protein improved Arabidopsis growth on phytate containing media. As CaPAP7 has highest activity for phytic acid which is an abundant non-available form of P in soil. So to validate its phytase activity and its impact on hydrolysis of phytate in plant, we grew Arabidopsis seedlings on phytate containing media. As expected, seedlings showed ~50% reduction in root elongation on media without Pi (−P) and phytate (as P source) as compared to +P media (Fig. 7). To further test whether phytate can induce low Pi stress in our set-up, we visualized root APase activity using BCIP and found increased APase activity under −P and phytate supplemented conditions. However, when seedlings were grown on phytate media overlaid with purified CaPAP7 enzyme; root length increased by 50% as compared to seedling on phytate without enzyme. Seedlings supplemented with CaPAP7 also showed increased lateral roots length (~233%) and higher biomass (~11%) than ones without enzyme (Fig. 7). Pi accumulation per plant was also increased (~22%) in seedlings with the application of CaPAP7 in phytate media in comparison to phytate only media. All these results strongly suggest that CaPAP7 was able to hydrolyze phytate to release Pi for better plant growth (Fig. 7). To this end, we conclude that CaPAP7 encodes a functional enzyme which is able to hydrolyze phytate and potentially influences seed phytate accumulation, and thereby seed weight.

Discussion
Chickpea is one of the most important legume crops cultivated in the semiarid regions of the world. Chickpea being a leguminous crop has higher ability to mobilise the soil P 36 . Along with phosphate, legumes are known to restore other nutrients to the soil such as nitrogen through symbiotic N 2 fixation which makes them highly important for the sustainable agriculture 37 . Legumes demand more Pi than non nodulating nitrogen-fixing plants as it is critically important for legume-rhizobia symbiosis [38][39][40] . Therefore, Pi deficiency affects chickpea development in dual manner; limits both growth and nitrogen fixation. One of the adaptive strategies evolved in diverse plants to increase labile Pi pool is secretion of APases 41,42 . PAPs are APases with broad substrate specificity and are reported to hydrolyse various organic-P compounds in different plants (Arabidopsis, rice, soybean, kidney bean) 4,8,11,12 . Here we identified 25 novel PAPs in chickpea which form three main groups (I, II, III) and seven subgroups (Ia-2, Ib-1, Ib-2, IIa, IIb, IIIa and IIIb) as described for Arabidopsis 7 . This indicates similarity in their function and behaviour; except clade Ia-1 which includes only AtPAPs. CaPAPs exhibited more similarity to AtPAPs and GmPAPs than to OsPAPs, suggesting that their behaviour and function are similar to other dicots. While all CaPAPs had an Arabidopsis homolog, a lower number of PAPs were retrieved in chickpea as compared to other legumes and Arabidopsis 7, 23 . This could be due to either incomplete genome sequence in chickpea or peculiar genetic constitution of chickpea, and raise the possibility of identification of more PAPs in future.
PAPs are known to be induced by Pi deficiency to help plants utilize organic sources of P 13,17,25,43,44 . CaPAPs also had P1BS cis-elements in their promoter which is involved in low Pi inducibility. Although CaPAPs were induced under Pi deficiency; however, level of induction was not very high. Moreover, low Pi inducible CaPAP7 didn't possess any P1BS element. Further, a direct correlation wasn't seen between existence of P1BS element and low Pi inducibility of PAPs in other plants 28,45 indicating role of other unknown regulators. Notably, many of  46 . This indicates that P1BS elements are not obligatory for low Pi responsiveness of many such genes. Analysis of low Pi inducible CaPAP7 revealed presence of two TC elements in its promoter which might be involved in its low Pi dependent transcriptional regulation.
Most of the CaPAPs were also upregulated by nitrogen deficiency, which can be correlated with the large amount of Pi requirement during nitrogen fixation by the nitrogenase complex. Their differential expression in other nutrient deficiencies hinted towards other roles of CaPAPs. This deviation from their well-established roles in Pi deficiency is in quite agreement with other observations. For instance, GmPAP21 is involved in P metabolism in root nodules. AtPAP17 is involved in the metabolism of reactive oxygen species while AtPAP26 plays a role in senescence driven Pi remobilization. Similarly, GmPAP3 participates in ROS response while AtPAP2 plays key roles in carbon metabolism 16,[47][48][49] . In rice PAPs have also been reported to play role in grain filling 50 . Heat inducible PgPAP18 was also reported to play defensive role against several environmental stresses 51 . Some PAPs such as NtPAP12 are involved in root system modulation by degradation of xyloglucan oligosaccharides and cello-oligosaccharides in the cell walls 52 . These evidences suggest diverse roles of PAPs in plants. On such similar lines CaPAPs may also be playing other roles besides Pi deficiency response.
To contemplate on these roles, we studied expression patterns of CaPAPs in different tissues. Interestingly, they showed least expression in root. Majority of them showed high expression in flower buds and young pods suggesting their important roles in reproductive development. Incidentally, several AtPAPs are also reported to be expressed during reproductive development 9 . High expression of PAPs in Arabidopsis pollens has been correlated with increased hydrolysis of phytate and germination of pollen 11 . Function of these enzymes therefore, could also be intracellular; in remobilizing Pi towards seed as seed development requires more P in comparision to normal cellular processes. Seeds also accumulate more P than other organs 53 and increasing seed P content has been suggested to improve seedling vigour 54 . Interestingly, among 25 identified CaPAPs, two PAPs, CaPAP7 and CaPAP26 showed strong association with both seed weight and phytate content. Chickpea seeds are high in phytate content, and seed size directly correlates with the phytate content of the seed 33 . It is noteworthy here that CaPAP26 is a close homologue of AtPAP26 which is known to be involved in Pi remobilization and senescence 48 . CaPAP7 showed high low Pi inducibility, localizes to cytoplasm and is of non-secretory nature. Recently, the Arabidopsis homologue of CaPAP7 (AtPAP7) was found to localise in peroxisomes 30 . However, we did not find any PTS in CaPAP7. Further, particle bombardment assay revealed no co-localisation of CaPAP7 with peroxisome marker in onion epidermal cells (data not shown). Therefore, CaPAP7 may not be a true homolog of AtPAP7. Further studies revealed that CaPAP7 to be enzymatically active at pH range (4-6) with optima at 4.5, validating its function as acid phosphatase. Intriguingly, while CaPAP7 showed broad substrate specificity, highest activity was observed on phytate, concluding CaPAP7 as PAP with phytase activity. PAPs are known to be glycosylated which affect their stability, targeting and enzyme kinetics 2 . Therefore, a high phytase activity of CaPAP7 needs further investigation using plant purified native CaPAP7. Nevertheless, PAPs with phytase activity do exist in plants 8,11 . Few purple acid phosphatases (PAPs) have also been reported to possess phytase activity in soybean, Arabidopsis, rice and tobacco 8,9,11,12,19 . Recently, OsHAD1, an APase has also been reported to possess both acid phosphatase and phytase activity 55 . Further, improvement of Arabidopsis growth on phytate, when supplemented with CaPAP7 enzyme validated CaPAP7 efficacy to release Pi from phytate by its hydrolysis. Phytases from different sources have also been earlier reported to enhance the plant growth by increasing Pi acquisition and biomass accumulation 10,[56][57][58] .
In plants, P used by vegetative tissues is ultimately mobilized to developing seeds, where nearly 75% of it is stored in the form of phytic acid 59 . During germination and seedling growth, this phytic acid is broken down by the phytases for utilization by seedlings 60 . Phytate content of chickpea seed is quite high which again corresponds to the need of these phytases during seed development and seed germination. Also, expression of few CaPAPs was quite high in mature leaf tissue, where their role can be correlated with the Pi recycling from old tissues to the new ones. Phytate binds with other minerals like Fe, Mg, Zn and therefore, reduces their bioavailability 61,62 . High consumption of phytate rich food is considered as one reason for mineral deficiency in humans in developing world 63 . Further, almost 50% of applied P fertilizers are removed from farmers field in the form of phytate in seeds every year 59 . Monogasteric animals cannot digest phytate which is excreted out and led to the eutrophication of water bodies. It is further shown that phytate accumulation is not absolutely necessary for seed functions. Rather, plants store excess of Pi in the form of phytate which seeds use during germination 53 . Therefore, low phytate food crops are a necessity and such crops can be engineered with new emerging resources.
Different approaches have been used for reducing seed phytate content by targeting phytate biosynthesis enzymes 63 , its transporter in tissue-specific manner 64 or by blocking the translocation of Pi to developing seed 65 . One of the problems coupled with lowering phytate content by reducing its biosynthesis in seed is corresponding increase in Pi which inhibits starch biosynthesis 66 . Consequently, low starch accumulation leads to reduced seed weight and crop yield. However, lowering of seed total P by reducing Pi transporter activity does not lead to any significant yield loss while reducing the seed phytate by 50% in rice 64 . Therefore, while reducing phytate content may affect seed fitness in the wild, it may not be a problem for cultivated crops in fertilized farmers' fields 63 . Here, CaPAP7 emerged as potential new genetic/genomic regulator for seed phytate content in chickpea. CaPAP7, besides phytate, can also hydrolyse other organic Pi compounds abundant in cell cytoplasm. It showed high expression in tissues with high phytate content i.e. flower bud (pollens) and young pod, its expression levels correlate negatively with seed phytate and weight. Therefore, it is very likely that CaPAP7 mediated phytate hydrolysis in these tissues results in high soluble Pi content which inhibits starch biosynthesis and ultimately leads to low phytate and seed weight. How precisely CaPAP7 regulates this process, needs further investigations. Nevertheless, our data revealed that CaPAPs have functional diversification in chickpea while some of them may be involved in Pi acquisition, CaPAP7 may play important roles in seed phytate accumulation and seed yield.