Eliminating the purple acid phosphatase AtPAP26 in Arabidopsis thaliana delays leaf senescence and impairs phosphorus remobilization


Author for correspondence:

William C. Plaxton

Tel: +1 613 533 6150

Email: plaxton@queensu.ca


  • Limitation of crop productivity by suboptimal phosphorus (P) nutrition is a widespread concern. Enhanced crop P-use efficiency could be achieved by improving P remobilization from senescing leaves to developing tissues and seeds. Transcriptomic studies indicate that hundreds of Arabidopsis thaliana genes are up-regulated during leaf senescence, including that encoding the purple acid phosphatase (PAP) AtPAP26 (At5g34850).
  • In this study, biochemical and functional genomic tools were integrated to test the hypothesis that AtPAP26 participates in P remobilization during leaf senescence.
  • An eightfold increase in acid phosphatase activity of senescing leaves was correlated with the accumulation of AtPAP26 transcripts and immunoreactive AtPAP26 polypeptides. Senescing leaves of an atpap26 T-DNA insertion mutant displayed a > 90% decrease in acid phosphatase activity, markedly impaired P remobilization efficiency and delayed senescence. This was paralleled by reduced seed total P concentrations and germination rates.
  • These results demonstrate that AtPAP26 loss of function causes dramatic effects that cannot be compensated for by any other PAP isozyme, even though Arabidopsis contains 29 different PAP genes. Our current and earlier studies establish that AtPAP26 not only helps to scavenge P from organic P sources when Arabidopsis is cultivated in inorganic orthophosphate (Pi)-deficient soils, but also has an important P remobilization function during leaf senescence.


Leaf senescence is a highly regulated developmental process that involves orderly changes in cellular physiology, biochemistry and gene expression leading to cell death (Buchanan-Wollaston et al., 2003; Lim et al., 2007). Senescing leaf cells experience sequential disorganization of cellular organelles and dramatic changes in metabolism that include loss of photosynthesis and catabolism of macromolecules that have been synthesized during the growth phase. Macronutrients such as nitrogen (N), potassium (K), sulphur (S) and phosphorus (P) are typically reallocated to young leaves and developing seeds. With the availability of its entire genome sequence and a host of related genomic tools, Arabidopsis thaliana has become an attractive plant model for studies of nutrient remobilization during senescence. Transcript profiling using microarrays indicated that at least 800 Arabidopsis genes are distinctively up-regulated during leaf senescence, whereas genes associated with photosynthesis and other anabolic processes are down-regulated (Buchanan-Wollaston et al., 2003; Gepstein et al., 2003; Lin & Wu, 2004; Lim et al., 2007). Senescence-associated genes probably play key roles in protein, lipid and nucleic acid degradation and nutrient recycling, amino acid transport and detoxification. However, despite its agronomic importance, the metabolic networks that mediate P remobilization from senescing leaves are poorly understood (Wang et al., 2010; Veneklaas et al., 2012). Maximizing the effectiveness of P remobilization from senescing tissues to younger leaves, and especially developing seeds, will probably make an important contribution to the creation of grain crops having enhanced P-use efficiency (PUE) (Veneklaas et al., 2012).

Leaf P reserves occur in various forms, including nucleic acids, phospholipids, phosphorylated metabolites and free inorganic orthophosphate (Pi). However, the nucleic acid pool, specifically ribosomal RNA, is the largest intracellular organic P pool, accounting for 40–60% of total organic P in a mature leaf (Veneklaas et al., 2012). DNA levels remain relatively constant as a leaf senesces, whereas RNA levels show a steady decrease, correlated with increased ribonuclease activity (Buchanan-Wollaston et al., 2003). In Arabidopsis, the S-like ribonuclease gene RNS2 is induced by senescence as well as Pi starvation (Taylor et al., 1993; Bariola et al., 1999). RNS2 localizes to the endoplasmic reticulum and cell vacuole where it mediates degradation of ribosomal RNA (Bariola et al., 1999; Hillwig et al., 2011). RNS2 orthologues from Antirrhinum and tobacco are also senescence- and Pi starvation-inducible (Lers et al., 2001; Liang et al., 2002). RNS2 induction during Arabidopsis leaf senescence is paralleled by increased AtPAP17 and AtPAP26 transcript levels (del Pozo et al., 1999; Gepstein et al., 2003). AtPAP17 and AtPAP26 are two of 29 purple acid phosphatase (PAP) isozymes encoded by the Arabidopsis genome (Tran et al., 2010b). PAPs catalyse Pi hydrolysis from a broad and overlapping range of phosphomonoesters with an acidic pH optimum and function in Pi production and recycling. Recent biochemical and functional genomic studies demonstrated that the dual-targeted AtPAP26 is the predominant intracellular (vacuolar) and secreted (rhizosphere and cell walls) PAP isozyme that functions in both vacuolar Pi recycling and extracellular Pi scavenging from organic P sources during nutritional Pi deprivation (Veljanovski et al., 2006; Hurley et al., 2010; Tran et al., 2010a; Robinson et al., 2012). Transcriptional induction of AtPAP26 in senescing Arabidopsis leaves (Gepstein et al., 2003) indicated that, similar to RNS2, this PAP isozyme might also be involved in mobilizing P during senescence. It was therefore of interest to establish whether AtPAP26 activities and polypeptide concentrations are up-regulated during senescence, as well as the degree to which P remobilization from senescing leaves of an atpap26 loss-of-function mutant (Hurley et al., 2010) might be compromised. The present study demonstrates that AtPAP26 is induced at both the transcript and protein level in order to remobilize P from senescing leaves.

Materials and Methods

Chemicals, plants and growth conditions

All biochemicals were purchased from Sigma-Aldrich Ltd, unless otherwise specified. For plant growth, Arabidopsis thaliana (L.) Heynh (Col-0 ecotype) and homozygous atpap26 T-DNA insertion mutant seeds (Hurley et al., 2010) were sown in a standard soil mixture (Sunshine Aggregate Plus Mix 1; SunGro, Vancouver, Canada) and stratified at 4°C for 3 d. Plants were cultivated in growth chambers (Model MTR30; Conviron, Winnipeg, MB, Canada) at 23°C (16 : 8 h photoperiod at 100 μmol m−2 s−1 photosynthetically active radiation) and fertilized biweekly by subirrigation with 4× diluted Hoagland's solution containing 2 mM Pi.

Induction of senescence

Prolonged darkness treatment of individual Arabidopsis leaves has been demonstrated to mimic natural senescence closely (Weaver et al., 1998; Gepstein et al., 2003; Keech et al., 2010). Thus, individual leaves of synchronously growing plants were covered with aluminium foil once leaves displayed serrated margins and visible trichomes (Weaver et al., 1998). These characteristics were seen c. 28 d after sowing. A maximum of two fully expanded leaves in the fifth or sixth positions were covered on individual plants. Leaves were determined to be senesced when they appeared to be 90–100% yellow. Brown/shrivelled leaves were classified as being fully senesced. Harvested leaves were immediately frozen in liquid N2 and stored at −80°C.

Quantitative real-time polymerase chain reaction (qPCR) analysis

Total RNA was extracted using RNeasy Plant Mini Kit (Qiagen) according to the manufacturer's instructions. DNase-treated RNA was reverse-transcribed using SuperScript III Reverse Transcriptase (Life Technologies, Burlington, ON, Canada). Gene-specific primers (Supporting Information, Table S1) were designed using DNAMAN software (version 5.0) and are listed in Table S1. qPCR was performed using an Applied Biosystems 7500 Real-Time PCR System (Life Technologies) and GoTaq qPCR Master Mix (Promega). PCR conditions consisted of an initial step at 95°C for 5 min followed by 40 cycles of 95°C for 15 s, followed by 60°C for 15 s, and one cycle of 72°C for 34 s. For amplification product specificity, a melt curve was generated at the end of each run, and verified by agarose gel electrophoresis of PCR products. Relative expression was determined by normalizing the transcript abundances of AtPAP17, AtPAP26 and RNS2 with ACTIN2 transcripts. The generated data were analysed with Applied Biosystems 7500 software, version 2.0.1. Relative expression (fold change) was determined using the inline image method (Livak & Schmittgen, 2001). All qPCR experiments were repeated three times using cDNAs prepared from two independent samples of nonsenescing or senescing leaf tissue.

Protein extraction, acid phosphatase activity assays and determination of soluble protein concentration

Leaves were homogenized (1 : 2; w/v) in ice-cold extraction buffer (20 mM Na-acetate, pH 5.6, 1 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulphonyl fluoride, 5 mM thiourea, and 1% (w/v) insoluble poly(vinylpolypyrrolidone)) using a chilled mortar and pestle and a small spatula of white quartz sand (50 and 70 mesh). Homogenates were clarified by centrifugation for 10 min at 14 000 g and 4°C. APase activity was measured by coupling the hydrolysis of phosphoenolpyruvate to pyruvate to the lactate dehydrogenase reaction at 24°C and continuously monitoring NADH oxidation at 340 nm using a Molecular Devices Spectromax Plus Microplate spectrophotometer (Molecular Devices, Sunnyvale, CA, USA) (Veljanovski et al., 2006). Optimized assay conditions were as follows: 50 mM Na-acetate (pH 5.6), 5 mM phosphoenolpyruvate, 10 mM MgCl2, 0.2 mM NADH, and three units of rabbit muscle lactate dehydrogenase in a final volume of 0.2 ml. Assays were corrected for any background NADH oxidation by omitting phosphoenolpyruvate from the reaction mixture. All APase assays were linear with respect to time and concentration of enzyme assayed. One unit of activity is defined as the amount of enzyme resulting in the hydrolysis of 1 μmol of substrate min−1 at 24°C. Protein concentrations were determined using a modified Bradford assay with bovine γ-globulin as the standard (Tran et al., 2010a).

Protein electrophoresis, immunoblotting and in-gel acid phosphatase activity staining sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) followed by immunoblotting with anti-AtPAP26-IgG and chromogenic detection of antigenic polypeptides using an alkaline phosphatase-tagged secondary antibody were conducted as previously described (Veljanovski et al., 2006). SDS-PAGE sample buffer included 50 mM dithiothreitol, and samples were heated at 100°C for 3 min. All immunoblot results were replicated a minimum of three times, with representative results shown in the various figures.

APase activity was visualized following PAGE by in-gel activity staining. SDS-PAGE was performed as described in the previous paragraph except that the sample buffer lacked dithiothreitol and samples were not boiled. Following electrophoresis, the gel was incubated for 20 min at 25°C in 40 mM Tris-HCl (pH 9.0), 2 mM EDTA and 1% (w/v) casein, and then for 20 min in 100 mM Na-acetate (pH 5.3). APase activity staining bands were revealed by placing the gel in 100 mM Na-acetate (pH 5.3) containing 10 mM MgCl2, 0.02% (w/v) Fast Garnet GBC (2-methyl4-[(2- methylphenyl)azo]benzenediazoium) salt, and 0.02% (w/v) β-naphthyl-P.

Seed germination and total P assays

Col-0 and homozygous atpap26 mutant seeds were produced from mother plants grown in identical conditions and harvested at the same time before being stratified at 4°C for 3 d. Seeds were placed on moist filter paper and germination was scored as radicle emergence from the seed coat. Total P concentrations of ashed seeds and leaves were determined using a spectrophotometric Pi assay as previously described (Hurley et al., 2010).


All values are presented as means ± SE. Data were analysed using the one-tailed Student's t-test, and deemed significant if < 0.05.

Results and Discussion

Delayed leaf senescence of atpap26 mutant plants

Approximately 28 d after planting, leaves of Col-0 and atpap26 plants were fully expanded, had serrated margins, visible trichomes and showed no phenotypic differences or signs of yellowing. When individual leaves were wrapped with an aluminium foil sleeve to induce senesence (Weaver et al., 1998), atpap26 mutant plants displayed an obvious delay in leaf senescence (Fig. 1). After 3 d of dark treatment, senescence was initiated in Col-0 leaves as evidenced by their yellowing. This yellowing arises from chlorophyll degradation, as the chloroplast is the first organelle to be targeted during senescence (Buchanan-Wollaston et al., 2003). However, initial yellowing of atpap26 mutant leaves was not evident until after 6 d of darkness. Col-0 leaves were 90–100% yellow after 6 d of dark treatment, whereas it took the atpap26 mutant 9 d to become 90–100% yellow (Fig. 1). An analogous delay in the onset of leaf senescence was reported when the expression of senescence-associated genes encoding hydrolytic enzymes involved with macromolecule catabolism and nutrient recycling (e.g. RNS2, lipase) was disrupted (Thompson et al., 2000; He & Gan, 2002; Lers et al., 2006).

Figure 1.

Delayed senescence of atpap26 mutants relative to Col-0. Arabidopsis plants were cultivated as described in the Materials and Methods section, and individual leaves in the fifth or sixth position of 28-d-old plants were wrapped in foil sleeves for up to 9 d to induce senescence. Images shown are representative of at least 10 replicates each.

AtPAP26 is the major acid phosphatase up-regulated by senescing Arabidopsis leaves and plays an essential role in P remobilization

Previous studies documented the up-regulation of vacuolar and secreted AtPAP26 protein expression and activity during nutritional Pi deprivation of Arabidopsis suspension cells and seedlings (Veljanovski et al., 2006; Tran et al., 2010a). AtPAP26's possible additional involvement in P remoblization during leaf senescence was suggested by the results of Gepstein et al. (2003), who used RNA-gel blots to document the accumulation of AtPAP26 transcripts in senescing Arabidopsis leaves. Likewise, specific transcript profiling from a comprehensive microarray site for Arabidopsis (Botany Array Resource, http://bbc.botany.utoronto.ca/) indicated that a significant (> twofold) increase in AtPAP26 transcripts occurs in senescing leaves. This was corroborated by using qPCR with gene-specific primers, which confirmed that AtPAP26 transcript abundances were c. threefold higher in senescing relative to nonsenescing Col-0 leaves (Fig. 2). AtPAP17 and RNS2 served as positive controls (Fig. 2), as their induction during leaf senescence has been well documented (Taylor et al., 1993; del Pozo et al., 1999; Gepstein et al., 2003). Transcriptional induction of AtPAP17 in senescing Col-0 leaves was quite pronounced, as AtPAP17 transcript levels of control, nonsenescing, leaves are very low relative to those of AtPAP26 or RNS2 (del Pozo et al., 1999; Bariola et al., 1999; Veljanovski et al., 2006).

Figure 2.

Analysis of AtPAP26 (At5g34850) gene expression in nonsenescing and senescing wildtype (Col-0) leaves. Arabidopsis leaves were considered senesced when they became 90–100% yellow (6 d post-dark treament). Levels of mRNA were analysed by quantitative real-time polymerase chain reaction qPCR using gene-specific primers for AtPAP17, AtPAP26 and RNS2. The transcript abundance of each gene in nonsenescing control leaves was used as the calibrator. ACTIN2 was used as an internal control for normalization. All values represent means ± SE of n = 3 independent determinations using cDNAs prepared from two biological replicates.

Senescing Col-0 leaves exhibited an c.800% increase in APase-specific activity compared with nonsenescing controls, correlated with the accumulation of immunoreactive 55 kDa AtPAP26 polypeptides (Fig. 3a,b). When clarified extracts from Col-0 leaves were resolved by nondenaturing PAGE and subjected to in-gel APase activity staining, a 100 kDa band was observed that showed a pronounced increase in senescing leaves and that comigrated with the native vacuolar AtPAP26 homodimer isolated from Pi-deprived Arabidopsis (Fig. 3c). Immunoblotting with anti-AtPAP26-IgG confirmed the lack of immunoreactive AtPAP26 polypeptides in clarified leaf extracts of the atpap26 plants (Fig. 3b). This was paralleled by a massive (96%) decrease in the extractable APase activity of senescing leaves of atpap26 relative to Col-0 plants; and absence of the 100 kDa APase (AtPAP26) activity staining band following nondenaturing PAGE of leaf extracts (Fig. 3a,c). The nondenaturing gels also resolved a 35 kDa APase activity staining band that was significantly up-regulated in senescing leaves of both Col-0 and atpap26 plants, and which probably corresponds to AtPAP17, a low-molecular-weight PAP that is induced during leaf senescence or nutritional Pi deprivation (del Pozo et al., 1999; Gepstein et al., 2003). As AtPAP17 is a multifunctional enzyme that exhibits APase as well as alkaline peroxidase activity, it has been hypothesized to be involved in both P remobilization and the production of reactive oxygen species (ROS) during leaf senescence (del Pozo et al., 1999). The results in Fig. 3(a) also indicate that PAPs other than AtPAP26 were up-regulated during leaf senescence, as senescing atpap26 leaves exhibited significantly greater extractable APase activity than their counterpart nonsenescing leaves (Student's t-test, < 0.05). Nevertheless, P remobilization efficiency was seriously compromised in senescing atpap26 leaves as they could only remobilize c. 15% of their total P (Fig. 4a). By contrast, Col-0 leaves remobilized c. 70% of their total P during senescence, similar to the value of 78% reported by Himelblau & Amasino (2001). The soluble protein concentration of senescing Col-0 or atpap26 leaves was c. sevenfold lower than that of nonsenescing leaves (Fig. 3a); this agrees with previous studies documenting the large reduction in soluble protein concentrations that accompanies senescence (Buchanan-Wollaston et al., 2003; Lim et al., 2007).

Figure 3.

Influence of leaf senescence on extractable APase activity and AtPAP26 polypeptide concentrations of Col-0 and atpap26 mutant plants. Arabidopsis leaves were considered senesced when they became 90–100% yellow (6 and 9 d post-dark treament for Col-0 and atpap26 leaves, respectively). (a) APase activity and total soluble protein of clarified nonsenescing (NS) and senescing (S) leaf extracts (Col-0, closed bars; atpap26, open bars). All values represent means ± SE of duplicate determinations on n = 4 biological replicates; asterisks denote atpap26 values that are significantly different from Col-0 (P < 0.05). (b) Soluble proteins (2.5 μg per lane) of Col-0 and atpap26 leaves and vacuolar AtPAP26 (25 ng) purified from Pi-deprived Arabidopsis suspension cells (Veljanovski et al., 2006) were resolved by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and electroblotted onto a poly(vinylidene difluoride) membrane. Blots were probed with rabbit anti-AtPAP26-IgG and immunoreactive polypeptides were revealed using an alkaline phosphatase tagged secondary antibody and chromogenic detection (Veljanovski et al., 2006). (c) Soluble proteins from Col-0 and atpap26 leaves (20 μg per lane) as well as homogeneous vacuolar AtPAP26 (0.5 μg) purified from Pi-starved Arabidopsis suspension cells were resolved by nondenaturing PAGE and subjected to in-gel APase activity staining.

Figure 4.

Impact of AtPAP26 loss of function on Arabidopsis total leaf and mature seed phosphorus (P) concentrations, and seed germination rate and efficiency. (a) Total P concentration of nonsenescing (NS) and fully senesced (S) leaves and mature seeds were determined using bulked samples as described in the 'Materials and Methods' section. All values represent means ± SE of = 5 biological replicates; asterisks denote values that are significantly different from Col-0 (< 0.05). (b) Col-0 and atpap26 mutant seeds were germinated on moistened filter paper and germination scored once the radicle emerged from the seed coat. All values represent means ± SE of = 30 seeds.

Seed P concentrations and germination efficiency of atpap26 plants are reduced

Mature seeds of atpap26 mutant plants exhibited significantly less (16%) total P relative to Col-0 seeds (Fig. 4a). This supports the hypothesis that AtPAP26 is important in P remobilization from senescing leaves, as liberated nutrients are often allocated to developing seeds (Buchanan-Wollaston et al., 2003; Veneklaas et al., 2012). Seed P content is important, as rice, wheat and barley cultivars with higher seed P concentrations establish seedlings faster, and ultimately produce higher yields (Ros et al., 1997; Veneklaas et al., 2012; White & Veneklaas, 2012). For this reason, farmers often artificially increase germinating seed P concentrations by targeted early-season Pi fertilizer treatments (White & Veneklaas, 2012). The atpap26 mutant seeds also displayed an obvious delay in germination rate and efficiency relative to Col-0 seeds (Fig. 4b), which could be a result of their decreased P reserves.


The de novo synthesis of intracellular and secreted PAPs by roots or suspension cell cultures is a widely recognized and much researched adaptation of plants subjected to nutritional Pi deficiency (Tran et al., 2010b). Conversely, little is known about the functions and identities of specific PAP isozymes mediating P remobilization during leaf senescence. However, such an understanding may facilitate the development of effective biotechnological strategies for improving crop PUE (Veneklaas et al., 2012). The results of the current study demonstrate that AtPAP26 is the major contributor to the increased APase activity and associated P remobilization that occur during leaf senescence. Abolishing AtPAP26 expression disrupted P remobilization from senescing leaves, decreased the total P accumulated by seeds, and delayed leaf senescence and seed germination. These effects could not be compensated for by the action of AtPAP17 or any of the other 27 PAP isozymes encoded by the Arabidopsis genome (Tran et al., 2010b). AtPAP26 is also up-regulated and dual-targeted by Pi-deprived Arabidopsis to recycle Pi from nonessential P-monoesters in the cell vacuole, as well as to scavenge Pi from rhizosphere and cell wall-localized organic P compounds (Veljanovski et al., 2006; Hurley et al., 2010; Tran et al., 2010a; Robinson et al., 2012). Kinetic studies with purified native vacuolar and secreted AtPAP26 isoforms revealed that this PAP effectively hydrolyses Pi from a wide range of substrates with a high catalytic efficiency over a broad pH range (Veljanovski et al., 2006; Tran et al., 2010a). AtPAP26 thus joins AtPAP17 and RNS2 as examples of multifaceted hydrolases that appear to contribute to P-acquisition efficiency and PUE. A challenging yet intriguing aspect for future studies will be to delineate the respective signal transduction pathways that result in differential AtPAP26 expression and targeting during senescence versus nutritional Pi deprivation. It will also be of interest to establish the degree to which P remobilization during leaf senescence (and overall PUE) might be augmented in AtPAP26 overexpressing crop plants.


This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Queen's Research Chairs program to W.C.P. Ms. Vicki Knowles kindly assisted with non-denaturing PAGE and in-gel APase activity staining of Arabidopsis leaf extracts. This research arose from discussions at the workshop ‘Phosphorus, the Inside Story’, held at the School of Plant Biology, The University of Western of Australia (1–4 February 2011). W.C.P. is grateful to Profs Eric Venaklaas (University of Western Australia) and John Raven (University of Dundee, UK) for their kind invitation to participate in this informative workshop.