A glycoform of the purple acid phosphatase AtPAP26 interacts with a GNA-apple domain lectin (AtGAL1) in cell walls of phosphate-starved Arabidopsis thaliana

Abstract

Purple acid phosphatases (PAPs) function in the acquisition and recycling of inorganic phosphate (Pi), a crucial but environmentally-limiting macronutrient for plant growth. Among 29 predicted PAPs of the model plant Arabidopsis thaliana, AtPAP26 (At5g34850) functions as the principal intra- and extracellular PAP isozyme that scavenges Pi during Pi-deprivation or leaf senescence. A pair of secreted, 55 kDa AtPAP26 ‘glycoforms’ (AtPAP26-S1 and AtPAP26-S2) were resolved during lectin-affinity chromatography of cell wall extracts from Pi-deprived Arabidopsis suspension cells and purified. High resolution liquid chromatography-tandem mass spectrometry (LC-MS/MS) demonstrated that their glycans were comparable at Asn103, whereas Asn365 and Asn422 glycosylation sites were extensively modified in AtPAP26-S2 by the addition of up to seven terminal mannose residues (to N-acetylglucosamine) to form high mannose glycans. A 55 kDa protein that co-purified with AtPAP26-S2 was identified by LC-MS/MS as Arabidopsis GNA and apple domain-containing lectin-1 (AtGAL1). AtGAL1 (At1g78850) belongs to Arabidopsis’ Galanthus nivalis agglutinin (GNA) lectin family, whose GNA domain binds high mannose N-glycans. AtGAL1 cross-reactivity with anti-AtGAL1-IgG was markedly attenuated when the lectin was incubated in the presence of a thiol-reducing reagent (consistent with three predicted disulfide bonds in AtGAL1’s apple domain). Reciprocal far western immunodot blotting demonstrated a specific interaction between purified AtGAL1 and AtPAP26-S2, but not AtPAP26-S1. Analytical gel filtration FPLC indicated that AtGAL1 and AtPAP26-S2 associate to form a 112 kDa heterodimer. Bimolecular fluorescence complementation assays established that like AtPAP26, AtGAL1 is also targeted to lytic vacuoles of Pi-starved Arabidopsis cells, and that both proteins interact in vivo. AtGAL1 pre-incubation significantly enhanced the acid phosphatase activity and thermal stability of AtPAP26-S2, but not AtPAP26-S1. Interestingly, LC-MS/MS also revealed that purified AtGAL1 was bisphosphorylated at Tyr38 and Thr39. Secreted AtGAL1 polypeptides were upregulated to a far greater extent than AtGAL1 transcripts during Pi deprivation, indicating post-transcriptional control of AtGAL1 expression. I hypothesize that AtGAL1 plays a key role during Pi deprivation through its interaction with high mannose glycans of AtPAP26-S2, and consequent positive impact on AtPAP26-S2 phosphatase activity and stability. The current study appears to provide the first definitive evidence for involvement of glycoforms, lectins, or a phosphotyrosylated protein in plant Pi starvation responses.