Nucleotidase Cascades Are Catalyzed by Secreted Proteins of the Parasitic Nematode Trichinella spiralis

Extracellular nucleotides are signaling molecules whose receptor-mediatedeffects are involved in a variety of physiological responsesin mammalian tissues. An overwhelming body of data indicatethat inflammatory and other immune responses can be modulatedby the availability and local concentrations of nucleotidesvia nucleotide receptor signaling, but this is only just beginningto be investigated in the context of infectious disease. Evidenceis provided here that the parasitic nematode Trichinella spiraliscan catalyze the conversion and thus modulate both the availabilityand concentration of extracellular nucleotides by means of thefollowing secreted exoenzymes: apyrase, 5'-nucleotidase, andadenosine deaminase. These enzymes were characterized in termsof substrate specificity, kinetic behavior, pH, divalent cationpreferences, and response to a series of compounds. The secreted5'-nucleotidase was identified as a protein with an apparentmolecular mass of 67 kDa after N-terminal amino acid sequencingof the purified protein. The presence of adenosine deaminasewas confirmed in the secreted products by Western blotting withan antibody against a mammalian enzyme, as a protein with anapparent molecular mass of 38 kDa. These secreted proteins constitutean enzymatic cascade which catalyzes the degradation of extracellularnucleotides, with a potential physiological role in the regulationof purinergic signaling.

INTRODUCTION

Top

Introduction

Purinergic signaling relies on interactions between extracellularnucleotides and plasma membrane-bound receptors. This type ofsignaling therefore depends on nucleotide release, metabolismby enzymes acting extracellularly, and the presence of receptorswhich will selectively bind the resulting nucleotides and transducethe signal to the interior of the cell.

Receptors for purine and pyrimidine nucleotides exhibit a widetissue distribution including all hematopoietic cells (1, 16,29, 48), and numerous studies have demonstrated that extracellularnucleotides regulate a broad range of inflammatory and otherimmune responses (14, 15, 17, 39, 47). There is now clear evidencethat nucleotides are released extracellularly in a regulatedmanner. In addition, large amounts of nucleotides are releasedupon mechanical stimulation (such as stretching) of differentcell types, including epithelial and endothelial cells, andmassive release of nucleotides into the extracellular spacefollows cell damage resulting in the activation of purinergicreceptors (34, 36, 37, 39, 65). The latter are broadly dividedinto two categories, the P1 and the P2 receptors. P1 receptorsrespond specifically to adenosine (and some to inosine), whereasP2 receptors are responsive to ATP, UTP, ADP, and UDP (14, 48,39). P1 receptors have been classified on the basis of biochemicaland pharmacological properties into four G-protein coupled receptorsubclasses, designated A₁, A_2A, A_2B and A₃. P2 receptors aresubdivided into the G-protein coupled P2Y receptors and theligand-gated ion channel P2X receptors. To date, seven mammalianP2X receptors (P2X_1-7) and seven P2Y (P2Y_{1,2,4,6,11,12,13}) receptorshave been cloned (1, 9, 10, 16, 48).

The concentration of extracellular nucleotides, and the cellularresponses to these molecules, is regulated by the action ofa ubiquitous family of membrane-bound nucleotide-metabolizingenzymes, such as ecto-ATP/ADPases and ecto-nucleotidases. Ecto-proteinkinases and ecto-nucleoside diphosphate kinases (NDPKs) alsomodulate the availability of different nucleotides and are ofphysiological importance in determining the activation of variouspurinergic receptor subtypes and the nucleotide-mediated signalingevents (35, 37, 64). A large body of data indicate that inflammatoryresponses are regulated by the availability and/or local concentrationsof nucleotides via purine/pyrimidine receptor signaling, whichaffects events such as cytokine release, mast cell degranulation,expression of adhesion molecules on the endothelium, and plateletaggregation (16, 27, 30, 39, 47).

Trichinella spiralis is a ubiquitous nematode parasite of awide variety of mammalian species, including humans. T. spiralisinfective larvae invade the small intestine by migration throughthe mucosal epithelial cells (13). The effector mechanisms whichresult in parasite expulsion are not fully understood, but itis a generally held opinion that this is effected by a combinationof alterations in intestinal physiology and Th2-mediated inflammatoryresponses (18, 21). Infective larvae possess a large secretoryorganelle termed the stichosome, the source of secreted proteinswhich can be recovered from in vitro culture of parasites. Itis considered likely that these secreted products are involvedin the survival and development of parasites within host cells(13).

As infective larvae of T. spiralis cause extensive tissue damageduring migration through the intestinal epithelium, high concentrationsof nucleotides must be liberated with profound effects on inductionof local inflammation. We have previously demonstrated thatT. spiralis infective larvae possess membrane-bound ecto-proteinkinase and phosphatase activities and secrete serine/threonineprotein kinases as well as a nucleoside diphosphate kinase (3,20, 53). It is reported here that infective larvae additionallysecrete a repertoire of proteins which constitute an enzymaticcascade of nucleotide-metabolizing molecules, which would directlyaffect the availability and local concentrations of extracellularnucleotides. It is postulated that these enzymes could potentiallyregulate purine/pyrimidine receptor activation and the ensuinginflammatory response.

MATERIALS AND METHODS

Top

Materials and Methods

Parasites. Infective larvae of Trichinella spiralis were recovered fromoutbred rats 2 months after oral infection as previously described(3). Parasites were cultured in serum-free RPMI-1640 containing0.25% glucose, 2 mM L-glutamine, 100 U of penicillin ml^-1, 100µg of streptomycin ml^-1, and 20 µg of gentamicinml^-1 at 37°C, 5% CO₂ for 72 h with a daily change of medium.Secreted products were pooled, cleared through 0.2-µm-pore-sizefilters, dialyzed against 25 mM HEPES (pH 7.5), and concentratedby passage through Centricon 10 microconcentrators (Amicon).Protein content was determined by the BCA microplate assay (Pierce).

Enzyme assays. Enzymatic phosphate hydrolysis was followed by measuring therelease of inorganic phosphate from the appropriate substratesby the malachite green-phosphomolybdate reagent (4, 28). Releasedinorganic phosphate was measured by determining the absorbanceat 660 nm using a microtiter plate reader (Spectra Max 340PC;Molecular Devices) and quantified by comparison to a standardcurve for inorganic phosphate obtained with sodium dihydrogenphosphate. Reactions were carried out in a buffer containing25 mM HEPES (pH 7.5), 150 mM NaCl, 5 mM MgCl₂, and 2.5 mM dithiothreitol(DTT), unless otherwise stated, and initiated by the additionof appropriate concentrations of secreted proteins and nucleotidesubstrates. The plates were incubated at 37°C for 10 min(during which time the reaction was found to be linear) priorto the addition of the chromogenic reagent. Controls for nonenzymatichydrolysis of phosphate were always carried out in parallel.Determination of kinetic parameters for apyrase and 5'-nucleotidasewere performed in the above buffer in the presence of variousconcentrations of UDP or AMP, respectively. Curve fitting tothe data points was performed by GraphPad Prism software.

Adenosine deaminase activity in the secreted products of T.spiralis was measured in quartz cuvettes using a buffer containing25 mM HEPES (pH 7.5) and 150 mM NaCl, unless otherwise stated.The reactions were initiated by the addition of the appropriateamounts of adenosine. Either samples were scanned at definedtime intervals from 230 to 320 nm or the absorbance was measuredat 265 nm, using a Hewlett Packard 8453 spectrophotometer withthe cuvettes maintained at 37°C. One unit of adenosine deaminaseactivity is defined as the amount of enzyme required to catalyzethe deamination of 1 µmol of adenosine per min ( ${Delta}$ A = 8.6min^-1 ml^-1 at 265 nm) (2).

Protein purification, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), N-terminal sequencing, and Western blotting. The 5'-nucleotidase was purified from total parasite-secretedproteins by ATP-affinity chromatography following the procedurepreviously described (20). Following adsorption of secretedproducts and washing with KCl, the protein was eluted from theaffinity matrix by 2 mM ATP. Protein samples were resolved bySDS-12% PAGE. For sequencing, protein samples were transferredonto Sequi-Blot polyvinylidene difluoride membranes (Bio-Rad)following the manufacturer's instructions, and N-terminal sequencingwas carried out by Severn Biotech Ltd. (Worcestershire, UnitedKingdom). For Western blotting the proteins were transferredonto nitrocellulose membranes and overlaid with a 1:1,000 dilutionof a rabbit antibody to bovine adenosine deaminase (Alpha DiagnosticInternational, San Antonio, Tex.). Binding was determined bystandard procedures utilizing horseradish peroxidase-conjugatedsecondary antibodies and enhanced chemiluminescence (AmershamRPN 2209).

RESULTS

Top

Results

Infective larvae of T. spiralis migrate through intestinal epithelialcells causing extensive tissue damage, which would inevitablyresult in the release of high concentrations of nucleotides.The secreted products of T. spiralis were therefore examinedfor enzymatic activities characteristic of proteins involvedin the catabolism of extracellular adenine nucleotides, andthe data obtained are shown in Table 1. Enzymes which catalyzedthe hydrolysis of ADP and AMP were clearly present. Interestingly,only minimal activity towards magnesium ATP (but not sodiumATP) was detected under the conditions of the experiments. Allother available nucleoside 5'-triphosphates were used as substrates,i.e., UTP, CTP, GTP, ITP, and TTP, but no measurable activitywas detected (data not shown). The secreted products of T. spiralislarvae thus exhibit an apyrase (nucleoside 5'-diphosphate phosphohydrolase)and 5'-nucleotidase activities, but no significant activitydirected against nucleoside 5'-triphosphates. The assay usedmeasures release of inorganic phosphate, and thus, the apyraseactivity reported here is an overestimate, since AMP, the productof this enzyme's action, would subsequently be hydrolyzed bythe 5'- nucleotidase. Nevertheless, the 5'-nucleotidase exhibiteda high specific activity. The possibility that inorganic pyrophosphataseor other nonspecific phosphatases were present in the secretedproducts was investigated, but no activity was detected whensodium pyrophosphate (at 2 or 5 mM) or phosphate esters suchas glucose-6-phosphate (5 mM) and ß-glycerophosphate(5 mM) were used (data not shown).

View this table:
[in this window]
[in a new window]
TABLE 1. Enzymatic activities in T. spiralis secreted proteins involved in the extracellular metabolism of nucleotides^a

The substrate specificities of the apyrase and the 5'-nucleotidasewere examined, and the data obtained are shown in Fig. 1. The5'-nucleotidase is virtually monospecific for AMP, since allother nucleoside monophosphates tested were poor substrates(Fig. 1A). It was surprising, however, to observe that the apyraseshowed a distinct preference for UDP, with minor activitiesmeasurable for all other nucleoside diphosphates (Fig. 1B).The specific activity of the enzyme with UDP as the substratewas estimated to be 2,120 ± 52 nmol/min/mg of total secretedprotein.

View larger version (27K):
[in this window]
[in a new window]
FIG. 1. Substrate specificity of the 5'-nucleotidase (A) and the apyrase (B) in T. spiralis secreted products. The data are expressed as a percentage of the maximal activities obtained and are the means for five independent experiments carried out in triplicate.

The 5'-nucleotidase and the apyrase activities were furthercharacterized, using UDP as the substrate for the latter enzyme,and the data are shown in Fig. 2. In the absence of divalentcations, significant 5'-nucleotidase activity but minimal apyraseactivity were observed. Magnesium ions were activators of bothenzymes, although the activity of the 5'-nucleotidase was notabsolutely dependent on the presence of the cation. Calciumions could not substitute for magnesium and indeed inhibitedapyrase activity. Zinc was also inhibitory for both enzymaticactivities. Addition of the reductant DTT had a profound effecton apyrase activity and induced an additional 10% to 20% increasein 5'-nucleotidase activity. This would suggest the involvementof thiol groups in the active site of at least the former enzyme.Both enzymes were active in the alkaline pH region, exhibitinga broad optimum between the values of 7.0 and 8.5. A similarpH profile and similar ion and reductant requirements were obtainedwhen ADP instead of UDP was used as the substrate for the apyrase(data not shown).

View larger version (22K):
[in this window]
[in a new window]
FIG. 2. Ion and reductant requirements for maximal activities for the 5'-nucleotidase (white bars) and apyrase (black bars) in T. spiralis secreted products. Measurements were performed in 25 mM HEPES (pH 7.5), 150 mM NaCl (column 1) with the following additions: column 2, 5 mM MgCl₂; column 3, 5 mM CaCl₂; column 4, 5 mM ZnCl₂; column 5, 5 mM MgCl₂ plus 2.5 mM DTT. The data are expressed as a percentage of the maximal activities obtained for each enzyme and are the means for five independent experiments carried out in triplicate.

Sensitivity to various compounds was tested, and the data areshown in Table 2. Levamisole had no significant effect on theenzymes, thus confirming the absence of an alkaline phosphatasein the secreted products. Suramin, a purinergic receptor antagonist,inhibited both enzymes but had a more pronounced effect on the5'-nucleotidase activity. Sodium azide, which has been usedto differentiate apyrases from other nucleotide-metabolizingenzymes (8), had no significant effect on the activities ofthe enzymes studied here. N-ethylmaleimide inhibited both enzymes,in particular apyrase activity, reinforcing the suggestion thatthiol groups are involved in catalysis. Fluoride was inhibitoryto both enzymes. Adenosine 5'-[ ${alpha}$ ,ß-methylene]diphosphate(p[CH₂]pA) and concanavalin A have been used as specific inhibitorsof ecto-5'-nucleotidases (63). As seen in Table 2, p[CH₂]pAindeed inhibited the secreted 5'-nucleotidase activity, butit also had an inhibitory effect on the apyrase. ConcanavalinA at concentrations below 50 µg/ml had no significanteffect on either of the enzymatic activities but exhibited someinhibition of the 5'-nucleotidase activity at 100 µg/ml.

View this table:
[in this window]
[in a new window]
TABLE 2. Effect of various compounds on 5'-nucleotidase and apyrase activities^a

Kinetic parameters for the enzymes were derived, and these areshown in Fig. 3 for the 5'-nucleotidase and Fig. 4 for the apyrase.The rate of phosphate release was found to be linear withinthe first 15 min of the reaction for both enzymes. Initial velocitieswere plotted against AMP concentration for the 5'-nucleotidaseand correspond to sigmoidal kinetics (Fig. 3). Curve fittingof the data points to the Hill equation resulted in a V_max of2.517 ± 0.096 nmol/min, substrate concentration at half-maximalvelocity of 0.287 mM ± 0.02, and a Hill coefficient of2.31. Similarly, the kinetic parameters for UDP hydrolysis bythe apyrase were examined. Initial velocities plotted over arange of UDP concentrations again followed sigmoidal kinetics(Fig. 4), and curve fitting of the data to the Hill equationresulted in a V_max of 1.059 ± 0.042 nmol/min, substrateconcentration at half-maximal velocity of 0.102 mM ±0.03, and a Hill coefficient of 1.53.

View larger version (15K):
[in this window]
[in a new window]
FIG. 3. Substrate concentration dependence of the rate of AMP hydrolysis by the 5'-nucleotidase in T. spiralis secreted products. The values are averages for five experiments performed in triplicate. The data were fitted to the Hill equation, and curve fitting of the data resulted in an r² value of 0.998. The calculated kinetic parameters are described in the text.

View larger version (16K):
[in this window]
[in a new window]
FIG. 4. Substrate concentration dependence of the rate of UDP hydrolysis by the apyrase in T. spiralis secreted products. The values are averages for five experiments performed in triplicate. The data were fitted to the Hill equation, and curve fitting of the data resulted in an r² value of 0.995. Kinetic parameters are described in the text.

The product of the action of 5'-nucleotidase is adenosine, andI therefore examined whether the latter could be further metabolizedby the secreted products of T. spiralis larvae. Secreted proteinswere incubated in the presence of adenosine, and spectra wererecorded as described in Materials and Methods. A time-dependentshift of the absorbance maximum from 258 to 249 nm was observed(Fig. 5A), as was a decrease at 265 nm accompanied by an increaseat 241 nm, shown by the difference spectrum in Fig. 5B, characteristicof the conversion of adenosine to inosine and diagnostic ofadenosine deaminase activity (2). The decrease in absorbanceat 265 nm, which was linear for 30 min (data not shown), wasused to calculate the specific activity of the enzyme at 51.6± 3.2 (n = 5) mU/mg of total secreted parasite proteins.The presence of reductant (in the form of DTT) did not affectadenosine deaminase activity, and divalent cations were notessential. Although addition of magnesium ions up to a concentrationof 10 mM had no significant effect, 5 mM calcium or zinc wasinhibitory. Optimal activities were observed over a broad pHrange between the values of pH 6 to 9 (data not shown).

View larger version (25K):
[in this window]
[in a new window]
FIG. 5. Adenosine deaminase activity in T. spiralis secreted products. (A) UV spectra were recorded over a period of 1 h following the addition of 100 µM adenosine into the reaction medium. (B) Difference spectra of the data shown in panel A obtained by subtraction of each spectrum from that at time zero.

Secreted products of adult parasites recovered from the intestinesof rats 4 days postinfection were also examined for these enzymes,and levels of activities comparable to those found in larvalproducts were obtained (data not shown).

An attempt was made to correlate the observed enzymatic activitieswith specific proteins present in larval secreted products.We had previously adopted an affinity purification protocolin order to isolate a secreted nucleoside diphosphate kinaseand noted another protein present in the fraction which waseluted from the column with 2 mM ATP (20). The procedure wasrepeated, and a single protein with an apparent molecular massof 67 kDa was eluted with ATP (Fig. 6A). This fraction containedapproximately 0.006% of the total secreted products, which thereforeprecluded biochemical characterization. The protein was subjectedto N-terminal sequencing, however, which yielded the amino acidsequence XQLTLIHTND. A BLAST search on this sequence identifiedstrong similarities with 5'-nucleotidases from a diverse rangeof species.

View larger version (91K):
[in this window]
[in a new window]
FIG. 6. Identification of the 5'-nucleotidase and adenosine deaminase in T. spiralis secreted products. (A) SDS-PAGE following purification of the 5'-nucleotidase by ATP affinity chromatography. Lane 1, total secreted products; lane 2, protein fraction eluted with 2 mM ATP. (B) Western blot of secreted products reacted with an antibody to mammalian adenosine deaminase. The molecular masses of marker proteins (M) are shown in kilodaltons.

Total secreted proteins were probed with an antibody raisedagainst mammalian adenosine deaminase. Strong reactivity occurredto a single protein with an apparent molecular mass of 38 kDa(Fig. 6B), and it is therefore assumed that this representsthe secreted adenosine deaminase which was characterized biochemically(Fig. 5).

DISCUSSION

Top

Discussion

It is shown that T. spiralis infective larvae secrete proteinswhich constitute an enzymatic cascade for the metabolism ofextracellular nucleotides. To the best of my knowledge, secretionof these enzymes from a parasitic nematode has not been previouslyreported. Secreted products convert nucleoside 5'-diphosphatesto the corresponding monophosphates via the action of an apyrase.This enzyme catalyzed the hydrolysis of ADP and UDP, and underour conditions, UDP was the preferred substrate. We have beenunable as yet to correlate the activity with a particular protein.Significant hydrolysis of ATP or any other nucleoside triphosphatewas not observed. This is unusual, as the majority of the membersof the apyrase enzyme family catalyze the hydrolysis of bothnucleoside 5'-triphosphates and nucleoside 5'-diphosphates,although with different preferences for the type of nucleotide(64). The enzyme activity present in T. spiralis secreted productsappears to be more similar to a protein belonging to a nucleoside5'-diphosphate specific subclass, referred to as NTPDase5, whichhas recently been characterized and denoted CD39-L4 (45, 46).CD39-L4 is a human apyrase which, although it contains the fiveapyrase conserved regions shared by all members of the family(52), lacks the C-terminal hydrophobic domain. The enzyme issecreted, and although it efficiently hydrolyses ADP, it showsa strong preference for UDP as does the apyrase in T. spiralissecreted products reported here. CD39-L4 is expressed in macrophages(45, 46), and several suggestions have been put forward regardingits possible function. It was postulated that macrophages atsites of vascular injury secrete this enzyme, degrading ADPand thus attenuating platelet aggregation. Alternatively, ithas been suggested (45) that the preference for UDP could belinked to modification of the levels of extracellular UDP, whichspecifically activates the P2Y₆ receptor (9). The latter receptoris expressed on all leukocytes and in particular on infiltratingT cells present during some inflammatory states (54), althoughthe biological consequence of receptor engagement in T cellsis not understood. Interestingly, it has been shown that monocytesare stimulated by extracellular UDP acting via the P2Y₆ receptor,to produce interleukin 8 (IL-8), a chemoattractant for neutrophils.It was further demonstrated that IL-8 production was inhibitedwhen extracellular UDP was degraded by a heterologous apyrase.(59). The enzyme present in the secreted products of T. spiralislarvae hydrolyses ADP and UDP and could therefore prevent bothADP and UDP receptor activation. Thus, this secreted apyrasecould inactivate, via degradation of ADP, P2Y₁/P2Y₁₂ receptorsinvolved in platelet aggregation or could, via degradation ofUDP, inactivate the P2Y₆ receptor functions mediated by thisnucleoside on a variety of leukocytes.

T. spiralis larvae also secrete a 5'-nucleotidase enzyme whichconverts AMP to adenosine. This enzyme was identified as a proteinwith an apparent molecular mass of 67 kDa and the N-terminalamino acid sequence XQLTLIHTND. A BLAST search using this sequenceindicated high levels of similarity with mammalian and insect5'-nucleotidases. Enzymes in the 5'-nucleotidase family arecharacterized by five to seven conserved sequence regions (58),and indeed the N-terminal sequence of the T. spiralis proteincontains one of these regions, which is considered a 5'-nucleotidasesignature (LTLIHTND). A search of the T. spiralis EST databaseusing the N-terminal protein sequence indicated identity withan ORF from two overlapping cDNAs (GenBank accession numbersBG353625 and BG302003). This ORF shows homology to 5'-nucleotidasesfrom diverse sources, although the ESTs are partial sequencescorresponding to less than 20% of a predicted full-length protein.Nevertheless, the ORF predicted a 21-amino-acid hydrophobicsignal peptide immediately preceding the N-terminal sequenceof the mature protein.

Vertebrate 5'-nucleotidase enzymes have been subdivided intovarious classes depending on their cellular origin, substratespecificities, pH optima, and cation requirements (63). Membrane-boundforms of 5'-nucleotidases are known to be surface-located ecto-enzymeslinked to the lipid bilayer via a glycosyl phosphatidylinositol(GPI) anchor. The soluble enzymes fall into three classes: twoare truly cytosolic, with K_m values in the millimolar range,and are subdivided into AMPases and IMPases depending on thesubstrate preference. Another soluble form has also been identifiedwhich results from cleavage of the GPI-linked ecto-enzyme byphosphatidylinositol-specific phospholipase C (63, 64) and hasproperties identical to those of the membrane-anchored enzyme.A different type of soluble 5'-nucleotidase with propertiesdistinct from those of the above-mentioned classes has beenfound in human seminal plasma (44), and a 5'-nucleotidase, additionallyexhibiting phosphodiesterase activity, has been reported tooccur in the salivary glands of the hematophagous sandfly Lutzomyialongipalis (7, 50). The kinetic properties, substrate specificity,and sensitivity to various compounds of the T. spiralis 5'-nucleotidaseindicate that this enzyme is distinct from previously reportedmembers of this enzyme family. Concanavalin A and p[CH₂]pA arespecific and potent inhibitors of ecto-5'-nucleotidases (65)and the soluble forms resulting from cleavage of the GPI anchorand have been used to distinguish these forms from the cytosolicvariants. Interestingly, although the enzyme was inhibited byp[CH₂]pA, it was relatively resistant to concanavalin A, againsuggesting that it is not a typical member of its class. Althoughthe possibility that the T. spiralis enzyme is recovered insecreted products following phospholipase cleavage of a membrane-boundform cannot be categorically excluded, this is unlikely notonly due to its properties but also because it is secreted ona daily basis at a constant rate over the term of culture.

The product of the enzymatic action of 5'-nucleotidases is adenosine,which can potentially be further metabolized by the enzyme adenosinedeaminase. The presence of this protein in the T. spiralis-secretedproducts was demonstrated by biochemical analysis and confirmedby Western blotting with an antiserum to a bovine enzyme. Adenosinedeaminases catalyze the irreversible hydrolytic deaminationof adenosine to inosine and ammonia. This enzyme class has abroad phylogenetic distribution, with amino acid sequences beinghighly conserved from bacteria to mammals (11). The propertiesof the T. spiralis adenosine deaminase appear to be similarto those from heterologous sources. Adenosine deaminase is typicallya cytosolic enzyme present in all mammalian tissues, althoughin recent years it has also been found in the salivary glandsof hematophagous arthropods (6, 38, 51). As activity was foundto be reduced following a blood meal, it has been suggestedthat the enzyme in these insects is secreted.

Recently extracellular adenosine deaminases have also been shownto exist on the surfaces of hematopoietic cells, and a surface-boundvariant on lymphocytes is known to bind to the T-cell activationmarker CD26 (19). It has been proposed that this protects Tcells from an adenosine-mediated inhibition of proliferationand regulates the ability of CD26 to promote cytokine production.

Adenosine and inosine, the products of the 5'-nucleotidase andadenosine deaminase, respectively, are well-known regulatorsof inflammation and have been shown to have immunomodulatoryeffects (12, 39). Adenosine not only inhibits platelet aggregationbut suppresses the production of inflammatory cytokines, suchas IL-1, IL-12, tumor necrosis factor alpha, and macrophageinflammatory protein 1 ${alpha}$ , by monocytes and macrophages and enhancesIL-10 production (25, 30, 31, 40). Inosine has also been shownto suppress inflammatory cytokine synthesis (26, 41). The roleof ADP in platelet aggregation is well characterized, whereasthat of extracellular ATP is more complex and can have bothproinflammatory and anti-inflammatory effects. Thus, althoughATP can trigger mast cell chemotaxis and the oxidative burstin neutrophils (33, 42), it is also known to suppress IL-12and TNF- ${alpha}$ release from macrophages but to enhance release ofIL-10 (24).

Mucosal mast cells have been demonstrated to be important effectorcells against several intestinal nematodes, notably T. spiralis(22, 23). The mechanisms whereby mast cells contribute to expulsionare unclear but may include local production of cytokines andother inflammatory mediators or secretion of mucosal mast cell-specificserine proteases. Deletion of the mouse mast cell protease-1gene is associated with significantly delayed expulsion of T.spiralis, demonstrating a role for a defined mast cell productin immunity to enteric nematodes (32). Interestingly, it isunclear how mast cells are activated during nematode infection,as degranulation occurs in the absence of the classical IgE/Fc ${epsilon}$ R1interaction (21), but this has also been reported to be modulatedby P1 receptor activation (39, 55).

Most blood-sucking arthropods contain large amounts of nucleotide-metabolizingenzymes in their saliva, such as ATPase, apyrase, 5'-nucleotidase,and adenosine deaminase (5, 6, 7, 50, 51, 56, 57), as nucleotidesare released in high concentrations from injured cells duringfeeding, and the repertoire of enzymes is predicted to act soas to modulate host hemostasis, inflammation, and immunity (49).Bacterial pathogens such as Mycobacterium bovis, Pseudomonasaeruginosa, Burkholderia cepacia, and Vibrio cholerae also secretea range of nucleotide-metabolizing enzymes, including ATPase,5'-nucleotidase, adenylate kinase, and NDPK (43, 60, 61, 62).It has been suggested that these enzymes interfere with purinergicsignaling on mast cells and macrophages through alterationsin the availability of agonist molecules.

The absence of an ATPase activity in T. spiralis secreted productswas initially surprising, as this would indicate that extracellularATP is not degraded by proteins in the secreted products. Althoughthis may well be the case, it is worth noting, as we have previouslyshown, that T. spiralis larvae secrete serine/threonine kinasesand a nucleoside diphosphate kinase (3, 20). These enzymes wouldutilize ATP, and importantly, NDPK would convert ATP to ADPor UTP to UDP in the presence of suitable nucleoside phosphateacceptors.

At this stage we can only speculate on the role of these enzymessecreted by T. spiralis. As they are secreted rather than membrane-boundenzymes it is unlikely that they have a function in nucleosidesalvaging. Nevertheless, they could modulate both the availabilityand local concentrations of nucleotides and would thereforebe expected to control purine/pyrimidine receptor activationwith concomitant regulatory effects on inflammatory responsesoccurring during host cell invasion. A model diagram of thecascade catalyzed by the secreted proteins showing the potentialfate of extracellular nucleotides and some of the cells likelyto be affected is shown in Fig. 7.

View larger version (35K):
[in this window]
[in a new window]
FIG. 7. A simplified model diagram of the enzymatic cascade catalyzed by the secreted proteins of T. spiralis. Extracellular nucleotides can be either available for receptor binding or metabolized by the enzymes identified. The type, local concentration, and availability of nucleotides may thus be modulated by the parasite. For simplicity, only receptors on hematopoietic cells without subtype-specific distribution are indicated, and cell membrane-bound ecto-enzymes involved in nucleotide degradation are not included. Details are described in the text.

The fact that bacterial pathogens and hematophagous insects,the latter inducing extensive host tissue damage, also secretesimilar enzymes suggests that this may be a conserved but asyet underappreciated feature of many organisms with importantconsequences for pathogenicity. Different types of purinergicreceptors may regulate different facets of immune and inflammatoryresponses, and it is possible that the parasite can modulatethese to its own advantage. We are currently investigating theseprocesses.

ACKNOWLEDGMENTS

This work was supported by the Wellcome Trust through a ResearchLeave Award.

I am indebted to Murray Selkirk for his continuous support andencouragement.

FOOTNOTES

* Mailing address: Department of Biological Sciences, Imperial College of Science, Technology, and Medicine, London SW7 2AY, United Kingdom. Phone: 44 20 7594 5209. Fax: 44 20 7594 5207. E-mail: k.gounaris{at}ic.ac.uk.

Editor: W. A. Petri, Jr.

REFERENCES

Top