Recombinant mouse PAP has pH-dependent ectonucleotidase activity and acts through A(1)-adenosine receptors to mediate antinociception

Abstract

Prostatic acid phosphatase (PAP) is expressed in nociceptive neurons and functions as an ectonucleotidase. When injected intraspinally, the secretory isoforms of human and bovine PAP protein have potent and long-lasting antinociceptive effects that are dependent on A(1)-adenosine receptor (A(1)R) activation. In this study, we purified the secretory isoform of mouse (m)PAP using the baculovirus expression system to determine if recombinant mPAP also had antinociceptive properties. We found that mPAP dephosphorylated AMP, and to a much lesser extent, ADP at neutral pH (pH 7.0). In contrast, mPAP dephosphorylated all purine nucleotides (AMP, ADP, ATP) at an acidic pH (pH 5.6). The transmembrane isoform of mPAP had similar pH-dependent ectonucleotidase activity. A single intraspinal injection of mPAP protein had long-lasting (three day) antinociceptive properties, including antihyperalgesic and antiallodynic effects in the Complete Freund's Adjuvant (CFA) inflammatory pain model. These antinociceptive effects were transiently blocked by the A(1)R antagonist 8-cyclopentyl-1, 3-dipropylxanthine (CPX), suggesting mPAP dephosphorylates nucleotides to adenosine to mediate antinociception just like human and bovine PAP. Our studies indicate that PAP has species-conserved antinociceptive effects and has pH-dependent ectonucleotidase activity. The ability to metabolize nucleotides in a pH-dependent manner could be relevant to conditions like inflammation where tissue acidosis and nucleotide release occur. Lastly, our studies demonstrate that recombinant PAP protein can be used to treat chronic pain in animal models.

Figure 1. Purification of recombinant mPAP.

(A) A thrombin cleavage site (Tr) followed by hexahistidine tag (H₆) and stop codon (*) were added to the C-terminus of the secretory isoform of mPAP. SP = signal peptide of mPAP. Map is not drawn to scale. (B) Amino acid sequence at the junction between the catalytic domain and Tr-H₆ tag. Arrow marks thrombin cleavage site. Asterisk marks stop codon. (C) GelCode blue-stained SDS-PAGE gel and (D) western blot of purified recombinant mPAP protein (1 µg and 5 µg, respectively). The western blot was probed with an anti-hexahistidine antibody.

Figure 2. Inhibition of mPAP by L-(+)-tartrate.

The indicated concentrations of L-(+)-tartrate were added to reactions (n = 3 per concentration) containing mPAP (1 U/mL), 100 mM sodium acetate, pH 5.6 and the fluorescent acid phosphatase substrate DiFMUP. Relative fluorescence units (RFU). All data are presented as means±s.e.m. Prism 5.0 (GraphPad Software, Inc) was used to generate curve.

Figure 3. mPAP dephosphorylates purine nucleotides in a pH-dependent manner.

Plot of initial velocity at the indicated concentrations of AMP, ADP and ATP at (A) pH 7.0 and (B) pH 5.6. Reactions (n = 3 per point) were stopped after 3 min. Inorganic phosphate was measured using malachite green. All data are presented as means±s.e.m. Error bars are obscured due to their small size.

Figure 4. TM-PAP dephosphorylates extracellular purine nucleotides in a pH-dependent manner.

HEK 293 cells were transfected with expression vectors containing (A–F) mouse TM-PAP or (G–L) the fluorescent protein Venus as a control. Cells were then histochemically stained using AMP, ADP or ATP (each 6 mM) as substrate at pH 7.0 or pH 5.6. Cells were not permeabilized with detergent. Scale bar (bottom right panel), 50 µm for all panels.

Figure 5. Dose-dependent antinociceptive effects of intrathecal mPAP.

(A) Effects of increasing amounts of mPAP on paw withdrawal latency to a radiant heat source. (B) Paw withdrawal threshold to a semi-flexible tip mounted on an electronic von Frey apparatus. (A, B) BL = Baseline. Injection (i.t.) volume was 5 µL. n = 8 wild-type mice were used per dose. There were significant differences over time between mice injected with heat-inactivated (0 U) mPAP and mice injected with active (1 U or 2 U) mPAP (Repeated measure two-way ANOVA; P<0.0001 for each dose). Post-hoc paired t-tests were used to compare responses at each time point between mice injected with active mPAP to mice injected with heat-inactivated mPAP (** P<0.005; *** P<0.0005). For the heat-inactivated mPAP control, the protein concentration was equivalent to the maximum 2 U dose of mPAP (1.1 mg/mL). All data are presented as means±s.e.m.

Figure 6. The antinociceptive effects of mPAP can be transiently inhibited with a selective A₁R antagonist.

Wild-type mice were tested for (A) noxious thermal and (B) mechanical sensitivity before (baseline, BL) and following injection of CFA (CFA-arrow) into one hindpaw. The non-inflamed hindpaw served as control. All mice were injected with active mPAP (mPAP-arrow; 2 U, i.t.). Two days later, half the mice were injected with vehicle (CPX/V-arrow, circles; intraperitoneal (i.p.); 1 hr before behavioral measurements) while the other half were injected with 8-cyclopentyl-1, 3-dipropylxanthine (CPX/V-arrow, squares; 1 mg/kg i.p.; 1 hr before behavioral measurements). There were significant differences over time between mice injected with vehicle and mice injected with CPX (Repeated measure two-way ANOVA; P<0.01). Post-hoc paired t-tests were used to compare responses at each time point between vehicle (n = 10) and CPX-injected mice (n = 10); same paw comparisons. *** P<0.0005. All data are presented as means±s.e.m.