Secreted Variant of Nucleoside Diphosphate Kinase from the Intracellular Parasitic Nematode Trichinella spiralis

Abstract

The molecular components involved in the survival of the parasitic nematode Trichinella spiralis in an intracellular environment are poorly characterized. Here we demonstrate that infective larvae secrete a nucleoside diphosphate kinase when maintained in vitro. The secreted enzyme forms a phosphohistidine intermediate and shows broad specificity in that it readily accepts γ-phosphate from both ATP and GTP and donates it to all nucleoside and deoxynucleoside diphosphate acceptors tested. The enzyme was partially purified from culture medium by ATP affinity chromatography and identified as a 17-kDa protein by autophosphorylation and reactivity with an antibody to a plant-derived homologue. Secreted nucleoside diphosphate kinases have previously been identified only in prokaryotic organisms, all of them bacterial pathogens. The identification of a secreted variant of this enzyme from a multicellular eukaryote is very unusual and is suggestive of a role in modulating host cell function.

Nucleoside diphosphate kinases (NDPKs) play a key role in the maintenance of intracellular pools of deoxynucleoside triphosphates (dNTPs) and NTPs via the transfer of phosphate from an NTP donor to an NDP acceptor. In addition, certain variants of these enzymes are involved in a variety of cellular processes unrelated to their catalytic activity, such as differentiation, proliferation, and suppression of tumor metastasis (8). In particular, nm23-H2/NDPK B has been identified as a DNA-binding protein and transcriptional activator of the human c-myc gene, previously known as PuF (29, 31). NDPKs are typically intracellular enzymes, although recently an ectoenzyme has been detected on the surface of mammalian cells (19, 20), and NDPKs have been reported to be secreted by the prokaryotic pathogens Mycobacterium bovis, Pseudomonas aeruginosa, and Vibrio cholerae (32, 38, 39).

Trichinella spiralis is a ubiquitous nematode parasite of a wide variety of mammalian species, including humans, and is remarkable among multicellular parasites in adopting an intracellular habitat, both in the systemic phase of infection in skeletal muscle cells and in the enteral phase, in which it invades and migrates through mucosal epithelial cells (9). It is likely that secreted products are involved in survival and development of parasites in both environments. Consistent with this assumption, infective larvae possess a large organelle termed the stichosome, which is the major source of secreted proteins which can be recovered from in vitro culture of parasites (9).

We have previously demonstrated that T. spiralis infective larvae secrete serine/threonine protein kinases (3). During the course of these studies, it became apparent that phosphorylation of a protein with an estimated mass of 17 kDa was regulated independently of both exogenous substrates and the major endogenous parasite substrate for protein kinase activity, a doublet of 50 and 55 kDa. We hypothesized that this may result from the activity of an additional enzyme and demonstrate here that T. spiralis secretes an NDPK which is autophosphorylated as part of its catalytic mechanism.

MATERIALS AND METHODS

Parasites.

Infective larvae of T. spiralis were recovered from outbred rats 2 months after oral infection as previously described (3). Parasites were cultured in serum-free RPMI 1640 containing 0.25% glucose, 2 mM l-glutamine, penicillin (100 U ml⁻¹), streptomycin (100 μg ml⁻¹) and gentamicin (20 μg ml⁻¹) at 37°C in 5% CO₂ for up to 72 h with a daily change of medium. Pooled secreted products were cleared through 0.2-μm-pore-size filters, dialyzed against 25 mM HEPES (pH 7.5), concentrated by passage through Centricon 10 microconcentrators (Amicon), and assayed for protein content using the bicinchoninic acid microplate assay (Pierce).

NDPK phosphotransferase assay.

Phosphotransferase assays were conducted by incubating 1 μg of secreted proteins or 100 ng of a partially purified enzyme fraction at 37°C for 30 min in 25 mM HEPES (pH 7.5)–50 mM NaCl–10 mM MgCl₂–10 μM dithiothreitol (DTT) in a final volume of 10 μl with 1 μCi or [γ-³²P] ATP or [γ-³²P] GTP and a range of NDP acceptors at 10 μM. Control reactions were set up in the absence of parasite proteins or acceptor. Reactions were terminated with 1 μl of 500 mM EDTA, and 1 μl was resolved by thin-layer chromatography (TLC) on cellulose MN 300 polyethyleneimine-impregnated plates (Macherey-Nagel) developed with 0.75 M KH₂PO₄ (pH 3.65). Plates were dried and exposed to autoradiography.

NDPK autophosphorylation and protein phosphorylation assays.

For NDPK autophosphorylation assays, parasite secreted proteins (3 to 4 μg) were incubated in 25 mM HEPES (pH 7.5), 140 mM NaCl, 1 mM MgCl₂, 0.8 mM CaCl₂, 5 mM KCl, and 5 mM EDTA in the presence of 10 μCi of [γ-³²P] ATP at 37°C for 30 min in a final volume of 10 μl. For protein phosphorylation assays, 2 μg of proteins was incubated in 25 mM HEPES (pH 7.5)–50 mM NaCl–10 mM MgCl₂–10 mM DTT in the presence of 10 μCi of [γ-³²P] ATP for 30 min at 37°C. Reactions were terminated by addition of Laemmli sample buffer, resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on a 15 or 20% gel, and exposed to autoradiography.

ATP-agarose chromatography.

Cyanogen bromide-activated ATP agarose (linked through C-8; nine-atom spacer; Sigma) was washed and equilibrated with 25 mM HEPES-KOH (pH 7.5)–2 mM MgCl₂–25 mM KCl–0.1 mM EDTA–0.05 mM DTT. Secreted parasite proteins were loaded onto the column, and the column was washed with the above buffer. Bound proteins were eluted with 0.5 M KCl in the above buffer, followed by elution with 2 mM ATP (pH adjusted to 7.5). Fractions thus obtained were concentrated by passage through Centricon 10 microconcentrators (Amicon) and extensively washed with 25 mM HEPES (pH 7.5).

Western blotting.

Protein samples were resolved by SDS-PAGE on a 15% gel, transferred to nitrocellulose membranes, and overlaid with a 1:400 dilution of a rabbit polyclonal antibody to NDK-P1 from Pisum sativum, a kind gift of Paul A. Millner (12). Binding was determined by standard procedures utilizing horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence (Amersham RPN 2209).

Acid/alkali stability of phosphorylated NDPK residues.

Samples in which the NDPK was autophosphorylated as described above were resolved by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes. Radioactivity was localized by autoradiography, and the areas on the membrane corresponding to NDPK were isolated and treated as described elsewhere (5). Briefly, pieces of PVDF membrane were incubated for 2 h at 45°C in 200 μl of the appropriate buffer containing 5% methanol, and radioactivity released was measured by scintillation counting. The following buffers were used: 50 mM KCl-HCl (pH 1), 50 mM glycine-HCl (pH 3), 0.1 M Tris-HCl (pH 7), 50 mM KCl-NaOH (pH 12), and 1 M KOH (pH 14). The membranes were then subjected to a further 2-h incubation at either pH 1 or pH 14 as indicated.

RESULTS

In previous studies, we had observed a phosphoprotein with an estimated mass of 17 kDa in the products secreted by infective larvae of T. Spiralis. We hypothesized that this might be the result of an autocatalytic phosphorylation event catalyzed by an NDPK enzyme, although this would be unusual in that NDPKs are typically intracellular enzymes. We therefore screened for an activity in secreted products which transferred the terminal γ-phosphate from radiolabeled NTP donors to NDP acceptors; as shown in Fig. 1, the requisite phosphotransferase activity was indeed present in parasite secreted products. The enzyme was capable of utilizing both ATP (Fig. 1A) and GTP (Fig. 1B) as donors, and all NDPs as acceptors. Furthermore, the enzyme could efficiently transfer phosphate to dNDP acceptors (data not shown).

FIG. 1.

NDPK activity in secreted products of infective larvae. (A) Transfer of γ-³²P from ATP to UDP (lane 1), CDP (lane 2), and GDP (lane 3). (B) Transfer of γ-³²P from GTP to UDP (lane 1), CDP (lane 2), and ADP (lane 3). Positions of migration of NTPs in TLC are indicated, as are the positions of liberated radiolabeled phosphate (P_i) and the origin (O).

In all NDPKs investigated, autophosphorylation of an active-site histidine is an intermediate in the catalytic mechanism. The parasite secreted NDPK was therefore identified via reaction with [γ-³²P] ATP under conditions which have been previously defined to optimize autophosphorylation (4). The results are presented in Fig. 2, which illustrate the phosphorylation of a 17-kDa protein (lane 1), which was completely abolished by the inclusion of 10 μM NDP acceptor in the reaction buffer (lane 2). Although, as stated above, NDPKs are known to phosphorylate on an active-site histidine residue, there have been numerous reports of additional serine (1, 7, 15, 22, 26) as well as aspartate and glutamate phosphorylation (36). In particular, phosphoserine formation has been linked to a variety of cellular processes involving NDPK activity other than the transfer of terminal phosphate between nucleosides. As shown in Fig. 3, we observed that 80 to 90% of the radioactivity was acid labile and alkali stable, indicative of histidine phosphorylation. Furthermore, alkali-stable radioactivity was released by subsequent acid treatment. It has been reported that addition of cyclic AMP in the reaction medium inhibits serine phosphorylation of human NDPK/nm23-H1 (22). Addition of excess cyclic AMP in the reaction medium resulted in no significant changes in the profile that we observed (data not shown). We did, however, reproducibly observe 5 to 10% acid-resistant phosphorylation and therefore carried out phosphoamino acid analysis after acid hydrolysis of the phosphorylated NDPK, which revealed no phosphorylated serine residues (data not shown).

FIG. 2.

Identification of the secreted NDPK as a 17-kDa protein by autophosphorylation. Lanes: 1, reaction performed in the absence of an NDP acceptor; 2, reaction performed in the presence of 10 μM GDP. Reaction products were resolved by SDS-PAGE (20% gel) and exposed to autoradiography. Autophosphorylated NDPK is indicated by an asterisk.

FIG. 3.

pH stability of autophosphorylated secreted NDPK. Treatment of NDPK immobilized on PVDF membranes was carried out as described in Materials and Methods. Values show percentage of total radioactivity released after incubation for 2 h in solutions of different pH (white bars) or after a subsequent incubation at pH 1 (hatched bars) or pH 14 (black bars). Values are averages of three independent experiments (standard deviations are indicated).

We proceeded to purify the NDPK by ATP-agarose chromatography. Figure 4 shows the profiles of total secreted products (lane 1), unbound proteins (lane 2), and those eluted by 0.5 M KCl (lane 3) and subsequently by 2 mM ATP (lane 4). Proteins with apparent masses of 70, 58, 30, and 17 kDa were eluted in the KCl fraction, and a single protein of 70 kDa was eluted by ATP. Western blotting and reaction with a polyclonal antibody to NDPK from P. sativum demonstrated reactivity to the 17-kDa protein in total secreted products and the KCl-eluted fraction, conclusively identifying this protein as a parasite-secreted NDPK. Antibody binding was occasionally observed to a protein with apparent molecular mass of 70 kDa in the latter fraction, possibly indicative of a multimeric association of the enzyme, although there was no reactivity to the 70-kDa protein eluted with ATP (lane 4).

FIG. 4.

Partial purification of the enzyme by ATP affinity chromatography. Samples from the various fractions were stained with Coomassie blue (A) or transferred onto nitrocellulose and reacted with an antibody to NDPK from P. sativum (B). Lanes 1, total secreted proteins; 2, unbound proteins; 3, proteins eluted with 0.5 M KCl; 4, proteins eluted with 2 mM ATP. Proteins were resolved by SDS-PAGE (15% gel). The molecular masses of marker proteins are shown in kilodaltons.

Given that we had previously identified serine/threonine protein kinase activity in T. spiralis secreted products (3) and that NDPKs have been demonstrated to phosphorylate other proteins with which they form close association (10, 11), we sought to discriminate between these activities to determine whether they indeed were catalyzed by two distinct enzymes secreted by these organisms. Figure 5A demonstrates that NDPK activity (assayed by transfer of phosphate to CDP) was present in both total secreted products and the KCl-eluted fraction of the ATP column but absent from in the flow through from the same column. These fractions were then assayed under conditions optimal for protein phosphorylation. Figure 5B shows that the total secreted products phosphorylated a triplet of proteins between 50 and 60 kDa and a protein of 17 kDa (lane 1). Phosphorylation of the 17-kDa protein alone was observed in the KCl-eluted fraction, whereas phosphorylation of the 50- to 60-kDa triplet alone was obtained with the flowthrough. These data therefore demonstrate that both NDPK and protein kinase activities are present in parasite secreted products and that they may be effectively separated by ATP-agarose chromatography under the conditions described.

FIG. 5.

ATP affinity chromatography separates NDPK and protein kinase activities. (A) Phosphotransferase assay with [γ-³²P] ATP donor and CDP acceptor resolved by TLC. The migration of NTPs is indicated. (B) Phosphorylation assays performed under conditions described in Materials and Methods. Proteins were resolved by SDS-PAGE (15% gel), and the molecular masses of marker proteins are shown on the right. Lanes: 1, total secreted proteins; 2, unbound proteins; 3, proteins eluted with 0.5 M KCl.

A number of secreted T. spiralis proteins have been shown to possess a novel family of tri- and tetra-antennary N-glycans capped by unusual tyvelose residues, and it has been demonstrated that tyvelose-specific monoclonal antibodies block invasion of epithelial cells by infective larvae in vitro (2). We therefore examined whether any of the proteins which bound to the ATP column were modified in this manner via Western blotting with a tyvelose-specific monoclonal antibody designated 18H, provided by Judith A. Appleton (23). We observed no binding to any of these proteins (data not shown), and found the expected profile of reactivity against other secreted products, and therefore conclude that the ATP-binding proteins described here are not modified by N-glycans incorporating tyvelose residues. In addition, we carried out deglycosylation reactions on the autophosphorylated form of the NDPK with N-glycanase, the results of which suggested that this enzyme is not glycosylated (data not shown).

DISCUSSION

The data presented here show that infective larvae of T. spiralis secrete NDPK. This was demonstrated by phosphotransferase activity, autophosphorylation, and reactivity with an antibody to NDPK from a plant source (P. sativum). Both of the latter procedures identified a protein of 17 kDa in secreted products, and inhibition of autophosphorylation by GDP identified this as an NDPK rather than a substrate for a protein kinase.

It was further observed that NDPK and protein kinase activities could be separated by ATP affinity chromatography. The identities of the other nucleotide-binding proteins at 30, 58, and 70 kDa are not known. NDPK and protein kinase activities could also be effectively distinguished by manipulation of the Mg²⁺ concentration. Thus, when EDTA was used to generate low levels of Mg²⁺, NDPK was the only protein phosphorylated in total secreted products (Fig. 2). It has been previously shown that in Candida albicans, NDPK autophosphorylation occurs with Mg²⁺ in the nanomolar range, and analogous to our findings, under optimal conditions it was the only protein phosphorylated in crude extracts (4).

We tested for the formation of the high-energy phosphoenzyme intermediate and showed that the autophosphorylated enzyme donates all of the phosphate to GDP (Fig. 2) and that almost all of the radioactivity is acid labile (Fig. 3). We therefore conclude that in this secreted variant of NDPK only the high-energy phosphoenzyme is formed. The low levels of acid-resistant radioactivity could be due to nonenzymatic transphosphorylation (5), and release of radioactivity at pH 7 can be accounted for by the low thermal stability of histidine-associated phosphate at neutral pH as previously reported (5, 6, 21). Secreted NDPK from T. spiralis copurified, under our conditions, with other proteins (Fig. 4). The identities of these proteins are unknown, but NDPKs from other sources have also been shown to copurify with a number of other proteins (10, 28).

NDPKs are ubiquitous enzymes which, in keeping with their role in the maintenance of intracellular nucleotide pools, in eukaryotes are found in the nucleus, the mitochondria and chloroplasts, and the cytosol (8). Different isoforms of NDPK show discrete patterns of expression in different tissues and during differentiation, although generally they are considered intracellular enzymes. An ectoenzyme was recently found to be associated with the surface of a human astrocytoma cell line (20), and ecto-NDPKs were subsequently described for a variety of other cell lines (19). It was suggested that this extracellular transphosphorylating activity might play a role in modulating adenine and uridine nucleotides in order to influence cellular functions via the P2 receptor class of signaling proteins (19, 20).

Only three examples of NDPK secretion, all from bacterial pathogens, appear to have been reported in the literature. M. bovis (and M. smegmatis) secrete both NDPK and ATPase (38). Extrinsic ATP acting through P2Z receptors on macrophages has been shown to induce both cell death by apoptosis and killing of resident mycobacteria (18). As secreted products from mycobacteria prevent ATP-induced macrophage apoptosis, it was suggested that depletion of extracellular ATP by these enzymes may act to promote survival of mycobacterium-infected cells (38). This postulate assumes that nucleotide-utilizing enzymes secreted by mycobacteria resident in phagosomes gain egress to the external environment. Moreover, the potential role of an NDPK in this process is less clear than that provided by an ATPase, as in addition to ATP depletion, the phosphotransferase activity of NDPK would generate ATP from other extracellular NDPs.

The other organisms for which NDPK secretion has been reported are P. aeruginosa (39) and V. cholerae (32). In both cases, NDPK was one of multiple nucleotide-utilizing enzymes secreted, including ATPase, 5′-nucleotidase, and adenylate kinase. Rather than protecting cells from ATP-induced apoptosis, the secreted products from these organisms are cytotoxic for macrophages and mast cells. It was suggested that the dichotomy in postulated functions for these enzymes lay on the one hand in the intracellular habitat of mycobacteria and on the other in the extracellular environment of P. aeruginosa and V. cholera, in which leukocytes present a potential threat rather than a requirement for survival (32, 39). Although it is unclear how these contrasting functions might be regulated, it is of interest that in the case of P. aeruginosa, NDPK secretion was observed only from virulent mucoid strains isolated from cystic fibrosis patients, not from avirulent strains (39). A specific role had previously been proposed for intracellular NDPK in the provision of GTP for synthesis of the exopolysaccharide alginate, associated with the transition to mucoidy (33).

To the best of our knowledge, the current data provide the first documented example of NDPK secretion by a eukaryotic organism, although significantly this is another infectious agent. A possible function for a T. spiralis secreted NDPK might lie in regulation of host cell proliferation and differentiation. Six isoforms of NDPK in humans, termed nm23-H1 to -H6, have been described (24, 25, 30, 34). One of these (nm23-H2) has been shown to act as a positive regulator of c-myc transcription (31), whereas nm23-H1 is a potential negative regulator of growth factor genes (8), and nm23-H1, -H2, and -H3 have all been implicated in arrest of differentiation in different cell types (17, 27, 35). Infective larvae of T. spiralis are isolated from nurse cells, a modified compartment of skeletal muscle with hypertrophic nuclei and endoplasmic reticulum (9). Cell cycle reentry and arrest in apparent G₂/M phase is a feature of these cells, as are extensive alterations in gene expression which result in the loss of characteristics associated with differentiated muscle cells (16). A number of parasite secreted proteins have been localized in host cell nuclei, although their identities and functions are unknown, and the relative contributions from host and parasite in cell cycle reentry and altered gene expression are unclear (37). One could potentially envisage a role for the T. spiralis secreted NDPK in participating in the alterations in gene expression associated with intramuscular development of the parasite.

Alternatively, a secreted NDPK could have a role in the subsequent intestinal phase of the life cycle, as it is possible that under the culture conditions used to maintain the parasites in vitro, they acquire certain characteristics of more advanced parasitic stages. Given the crucial involvement of mast cells in expulsion of T. spiralis from the gut (13, 14), it is interesting that secreted products of P. aeruginosa and V. cholerae show ATP-dependent cytotoxicity toward mast cells (32, 39), although this has not been specifically linked to NDPK per se. We therefore intend to investigate the potential involvement of the T. spiralis secreted NDPK in directed cytotoxicity against a variety of cell types. We are also in the process of cloning genes, localizing the enzymes, and determining their dynamics of expression throughout the life cycle in order to elucidate the roles of this multifunctional protein.