Triatoma infestans apyrases belong to the 5'-nucleotidase family.

Apyrases are nucleoside triphosphate-diphosphohydrolases (EC 3.6.1.5) present in a variety of organisms. The apyrase activity found in the saliva of hematophagous insects is correlated with the prevention of ADP-induced platelet aggregation of the host during blood sucking. Purification of apyrase activity from the saliva of the triatomine bug Triatoma infestans was achieved by affinity chromatography on oligo(dT)-cellulose and gel filtration chromatography. The isolated fraction includes five N-glycosylated polypeptides of 88, 82, 79, 68 and 67 kDa apparent molecular masses. The isolated apyrase mixture completely inhibited aggregation of human blood platelets. Labeling with the ATP substrate analogue 5'-p-fluorosulfonylbenzoyladenosine showed that the five species have ATP-binding characteristic of functional apyrases. Furthermore, tandem mass spectroscopy peptide sequencing showed that the five species share sequence similarities with the apyrase from Aedes aegypti and with 5'-nucleotidases from other species. The complete cDNA of the 79-kDa enzyme was cloned, and its sequence confirmed that it encodes for an apyrase belonging to the 5'-nucleotidase family. The gene multiplication leading to the unusual salivary apyrase diversity in T. infestans could represent an important mechanism amplifying the enzyme expression during the insect evolution to hematophagy, in addition to an escape from the host immune response, thus enhancing acquisition of a meal by this triatomine vector of Chagas' disease.

Hematophagy in triatomines (Hemiptera: Reduviidae) is associated with the presence of biochemical compounds in the salivary glands that are essential for obtaining blood meals. Indeed, most blood-feeding arthropods have salivary components with vasodilatory, anti-clotting, and anti-platelet aggregation activities that are capable of inhibiting hemostatic reactions of the host (1)(2)(3)(4).
Host platelet aggregation is considered to be an important hemostatic barrier against insect feeding, because it can stop the bleeding of small blood vessels regardless of other clotting factors (5,6). Rhodnius prolixus, a triatomine, neutralizes and overrides platelet aggregation induced by collagen, thrombin, thromboxane A 2 , and ADP (7)(8)(9)(10). Similarly, collagen-and thrombin-induced platelet aggregation is inhibited by, respectively, pallidipin and triabin, both of which are present in the saliva of Triatoma pallidipennis (11,12). However, the importance of ADP as a common mediator of platelet aggregation pathways is evidenced by the presence on the vascular endothelium surface of the CD39 apyrase, which limits platelet aggregation by hydrolyzing ADP, thus preventing thrombus formation (13). Thus, studying insect apyrases may lead to alternative strategies against the diseases they transmit, as well as new pharmaceutical tools for platelet aggregation-associated disorders. Apyrase removes inorganic phosphate from ATP and ADP, and thus prevents platelet aggregation (6,8,14). Apyrase activity has been characterized in the saliva of Anopheles stephensi, Aedes albopictus, and Aedes aegypti as 65-, 61-, and 68-kDa protein, respectively (15)(16)(17). Ae. aegypti and Anopheles gambiae apyrases are members of the 5Ј-nucleotidase family (17,18), whereas the gene coding for the 37.5-kDa apyrase of Cimex lectularius belongs to a novel protein family showing significant similarity to phlebotomine apyrases (19 -22) and to human and to rat apyrases (23,24).
Here, we report the purification and characterization of five salivary apyrases from Triatoma infestans, a vector of the agent of Chagas' disease, the protozoan Trypanosoma cruzi. The sequence of the gene encoding the 79-kDa apyrase confirmed that it belongs to the 5Ј-nucleotidase family apyrase.

EXPERIMENTAL PROCEDURES
Triatomines and Collection of Saliva-T. infestans were reared in an insectary maintained at 28 Ϯ 2°C, 70% Ϯ 5 relative humidity, with photoperiods of 12 h. Pipette tips were placed over the triatomines mouthparts and approximately 0.2-1 l of saliva was collected from each adult triatomine. Except for phenotyping, all experiments were performed with pooled saliva obtained from insects at 20 days following a blood meal. The saliva was filtered through an 0.22-m pore membrane and stored at Ϫ80°C until use.
Apyrase Purification-Saliva (800 g of protein) was applied to an oligo(dT) 12-18 cellulose column (0.7 ϫ 10 cm) pre-equilibrated with 25 mM Tris-HCl, pH 7.5. Bound proteins were eluted with isocratic steps of 10 ml of the same buffer containing 0.1, 0.5, and 1 M NaCl, respectively, at a flow rate of 0.5 ml/min. Fractions of 0.5 ml were collected on ice and tested for apyrase activity. The active fractions, which were eluted in 0.5 M NaCl, were pooled and concentrated to 50 l by ultrafiltration through a Centricon 30 (Amicon) at 4°C. The resulting protein solution (50 g) was loaded on a Superose 6 HR 10/30 gel filtration column (Amersham Biosciences), equilibrated with 25 mM Tris, 500 mM NaCl, pH 7.5, and eluted at a flow rate of 0.4 ml/min. The absorbency at 280 nm was recorded, and 0.4-ml fractions were tested for enzyme activity. The active fractions, pooled and concentrated as described above, this time to 0.5 ml, were stored at Ϫ80°C until use.
Deglycosylation-Proteins were deglycosylated with O-glycosidase (0.1 unit) and neuraminidase (0.5 unit) as described by the glycosidase procedures (Bio-Rad). For PNGase F 1 (0.1 unit) treatment, 2 g of purified saliva were boiled in 1% SDS and 2% 2-mercaptoethanol and diluted 10-fold in 100 mM phosphate containing 0.5% Nonidet P-40 prior to incubation with the enzyme for 12 h at 37°C. The mix was then subjected to 8% SDS-PAGE analysis.
FSBA Labeling-Purified apyrases were incubated at 37°C for 30 min with 1 mM FSBA (Sigma) solubilized in Me 2 SO. To assess labeling specificity, ATP (5 mM) was introduced in the reaction mix before the addition of FSBA. Mock control consisted of purified apyrases incubated with Me 2 SO only (10% final concentration). Then, proteins were submitted to 6% SDS-PAGE under reducing conditions, and immunoblot was performed with serum raised against FSBA in rabbit (a kind gift of Dr. A. R. Beaudoin). Immunoreactivity was detected by chemiluminescence following the manufacturer's instructions (Amersham Biosciences).
Platelet Aggregation Assay-From a fasting human donor who had not been exposed to any platelet aggregation-interfering drugs within the preceding 10 days, 5 milliliters of human blood was collected in sodium citrate 0.38%, and 2 ml of the platelet-rich plasma (PRP) was obtained by differential centrifugation (25). The PRP (400 l) was preincubated under stirring for 10 min at 37°C, and the aggregation was induced by the addition of ADP (5 M final concentration) and measured in an aggregometer (model 2020, NetLab, Sã o Paulo, Brazil) after incubation for 10 min. Different dilutions of crude or purified apyrase were added to the PRP at 4 min prior to the addition of the inducer.
Enzymatic Characterization of T. infestans Apyrase Activity-T. infestans salivary apyrase activity was determined essentially as described previously (26). Unless otherwise indicated, 30-l samples were incubated in microplates with 50 l of activity buffer (100 mM Tris, 10 mM KCl, 450 mM sucrose, 20 mM glucose, 2 mM MgCl 2 , 0.1 mM EDTA, pH 7.5) and 20 l of 5 mM ATP, ADP, or AMP at 37°C for 20 min. The released inorganic phosphate (P i ) was quantified by the Fiske and Subbarow method (27). Reaction was shown to be linear over a 30-min period. Each assay was run in triplicate. One unit of enzyme activity was defined as the amount of enzyme needed to release 1 mol of P i /min at 37°C. The activity of the purified apyrase mixture was determined at 25,30,37,42,56, and 65°C for 20 min, as described above. Likewise, the effect of pH on the apyrase activity was determined at 37°C using 50 mM buffering solutions of bis-Tris, pH 6 -7, Tris-HCl, pH 7-8, and borate, pH 9 -10. The effect of divalent cations on enzyme activity was determined by substituting the MgCl 2 present in the reaction mix with CaCl 2 , CuCl 2 , ZnCl 2 , MnCl 2 , or CoSO 4 at a final concentration of 1 mM. The effects of inhibitors on the apyrase activity were determined by preincubating purified apyrase at 37°C for 10 min with ouabain (1 mM), levamisole (1 mM), mycrocistin-LR (0.01 mM), okadaic acid (0.01 mM), sodium azide (20 mM), NaF (1 and 10 mM), vanadate (0.1 an 1 mM), 2-mercaptoethanol (5 mM), iodoacetamide (5 mM), or di(adenosine-5Ј)pentaphosphate (0.01 mM).
Mass Spectrometry-Following electrophoresis and Coomassie Blue staining, the 88-, 82-, and 79-kDa bands and the 68/67-kDa doublet were excised from the gel, washed, and analyzed in the manner described previously (28). In brief, the gel pieces were washed in 25 mM ammonium hydrogenocarbonate (NH 4 HCO 3 ), pH 8, then in 50% acetonitrile, 25 mM NH 4 HCO 3 , and finally with pure water, for 30 min each before complete dehydration in a vacuum centrifuge. The gel pieces were reswollen with a minimum amount of sequence grade modified porcine trypsin (Promega, Madison, WI) solution containing 0.5 g of protease (typically 10 l of an 0.05 g portion of trypsin/l solution in 25 mM NH 4 HCO 3 containing 10% acetonitrile). When necessary, NH 4 HCO 3 buffer was added until the gel piece achieved complete rehydration. Digestion occurred at 37°C for 3-5 h. The tryptic digest was extracted twice with a 50% acetonitrile, 25 mM NH 4 HCO 3 solution. The digest solution and the extracts were pooled, dried in a vacuum centrifuge, and desalted with ZipTip C18 (Millipore, Bedford, Ma) prior to a nanospray MS/MS analysis. A Q-Tof instrument (Micromass, Manchester, UK) was used with a Z-Spray ion source working in the nanospray mode. About 3-5 l of the desalted sample was introduced into a needle (medium sample needle, Protana Inc., Odense, Denmark) to run MS and MS/MS experiments. The capillary voltage was set to 1000 V and the sample cone to 50 V. Glufibrinopeptide was used to calibrate the instrument in the MS/MS mode. MS/MS spectra were transformed using MaxEnt3 (MassLynx, Micromass Ltd.), and amino acid sequences were analyzed using PepSeq (BioLynx, Micromass Ltd). Amino acid sequences were used in homology searches with BLAST or FASTA (www.ncbi.nlm.nih.gov/blast/).
PCR Amplification of apy79 cDNA-cDNA inserts of a salivary gland cDNA library constructed in ZAP bacteriophage (Stratagene) were PCR-amplified using the M13 reverse and T7 vector primers flanking the inserts as follows. After chelation of the MgCl 2 present in SM buffer (50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 8 mM MgSO 4 , 0.01% gelatin) by EDTA used for storage of the library phages, 10 l of the library boiled for 10 min was used as template in a 35-cycle standard PCR reaction. Amplification products ranging from 0.7 to 2.6 kb were then gel-purified using a QIAquick column (Qiagen).
To amplify the apy79 cDNA fragment, primers were derived from the 5Ј-nucleotidase conserved motif GNHEFD and the VYEEDDL/I peptide sequence of the 79-kDa apyrase. The degenerated primer set CONS1-s (GGNAAYCAYGARTTYGAY) and MS79-as (ADRTCRTCYTCYTC-RTA) was used. PCR amplification products were obtained from the amplified library with an annealing temperature of 45°C (8°C above calculated T m ) to ensure maximum specificity. The single band product of 1.2 kb was cloned into pGEM T-easy plasmid (Promega).
Library Screening-The 1.2-kb cDNA fragment was labeled using the Random Primer labeling kit (Invitrogen) and used to screen the salivary gland cDNA library by standard procedures ( ZAP manual), with washes done at 65°C in 0.1ϫ SSC, 0.1% SDS. Positive clones were submitted to in vivo excision ( ZAP manual) to rescue the plasmids from which the inserts were sequenced.
5Ј-RACE-To get the 5Ј missing region of the apy79 cDNA, we used the Stratagene GeneRacer kit that allows selection of complete mRNA by selective ligation of a RNA linker to their capped 5Ј end. We used as starting material 2 g of total RNA TRIzol extracted from the salivary glands of equal numbers of adult insects at 1, 2, 3, 4, 7, and 15 days after feeding. The 79GSP1-as primer (AGGTTGTGATCCTGTA) was used for reverse transcription, and the resulting cDNA was submitted to PCR with the 79GSP2-as primer (GCCAGAATGTCCCACGGCGAATA) and 5Ј end linker primer. Specificity of the 1-kb product obtained was confirmed by nested PCR using two other gene-specific primers and Southern blot with the 1.2-kb apy79 cDNA probe. This product was cloned into pCR 4-TOPO and sequenced.
Baculovirus Expression-Recombinant baculovirus was obtained using the Bac-to-Bac system and protocol (Invitrogen). The Apy79 coding region was cloned by standard molecular biology techniques into the pFastBac plasmid with native signal peptide or mellitin signal peptide (kind gift from S. Bibert). In addition, two His-tagged constructions were derived by adding six histidine residues at the C terminus. Sf9 insect cells were infected with the recombinant baculovirus, and supernatant aliquots were taken at 24, 48, 72, and 96 h after infection.
SDS-PAGE and Western Blot-Proteins were separated by 10% SDS-PAGE using the standard Laemmli procedure. After completion of the run, the proteins were transferred to a nitrocellulose membrane in a semidry system. Membranes were blocked by incubation in phosphatebuffered saline containing 5% dry skimmed milk and then probed with anti-apyrase serum raised in rabbit by immunization with purified 79-kDa recombinant apyrase. Secondary antibody was conjugated with horseradish peroxidase, and immunoblot was developed using diaminobenzidine.
Genomic Southern-Genomic DNA was extracted from 50 adult T. infestans insects, and the pooled DNA was probed by Southern blot with 32 P-labeled apy79 cDNA (29). Fifty g of genomic DNA was digested with XhoI, BamHI, or HindIII. This last enzyme promotes two cuts within the apy79 cDNA, generating a 468-bp internal fragment. The DNA fragments were separated in an 0.7% agarose gel and blotted onto a nitrocellulose membrane, which was probed with 25 ng of radiolabeled apy79 spanning positions 75-2449 of the deposited cDNA sequence (EcoRI digestion). The blotted nitrocellulose membrane was washed subsequently at 60°C in 2ϫ SSC (twice), 0.2ϫ SSC, and 0.1ϫ SSC, each containing 0.1% SDS. T. cruzi DNA was used as a negative control.

RESULTS
Purification and Characterization of Proteins with Apyrase Activity-Insect saliva from about 150 T. infestans reared in 1 The abbreviations used are: PNGase F, peptide N-glycosidase F; FSBA, 5Ј-p-fluorosulfonylbenzoyladenosine; PRP, platelet-rich plasma; MS/MS, tandem mass spectroscopy; RACE, rapid amplification of cDNA ends; bis- the laboratory was used to purify apyrase activity first described by Ribeiro et al. (3). An oligo(dT) 12-18 cellulose column was used for a first-step purification of the apyrase because this method has been successfully employed for purification of terminal deoxynucleotidyltransferases (30). A typical experiment showing chromatography of 800 g of saliva in an oligo(dT)-cellulose column is shown in Fig. 1A. About 10% of the total saliva proteins bound to the matrix and was eluted with 0.5 M NaCl. This fraction, which retained 84% of the salivary apyrase activity, was concentrated and loaded on a Superose 6 column. The apyrase activity was eluted as a large peak with an elution volume corresponding to a 400-kDa globular protein, as shown in Fig. 1B. This two-step chromatographic procedure led to an ϳ75-fold increase in apyrase specific activity with a yield of 35%.
SDS-PAGE analysis (Fig. 1C) showed that the 400-kDa apyrase fraction contained five proteins with molecular masses of 88, 82, and 79 kDa and a 68 -67 kDa doublet, implying that they are associated in a protein complex. 2 The presence of the 68 -67 kDa doublet is more clearly seen in SDS-PAGE with a lower polyacrylamide concentration (Fig. 2, lane 1). The purified sample was treated with neuraminidase, O-glycosidase, or PNGase F. Only PNGase F changed the migration pattern to 64-, 59-, 57-, and 54-kDa bands in SDS-PAGE (Fig. 2) showing that the various species are N-glycosylated and do not represent different glycosylation levels of one or more proteins.
The five apyrase species were specifically labeled by FSBA (Fig. 3), an ATP analogue that covalently binds its enzyme target and can be detected by anti-FSBA-specific antibodies (31). Antibody specificity was shown by the lack of reactivity against apyrases that were not incubated with FSBA (Fig. 3,  lane 1). The intensity of the signal was markedly reduced when the apyrases were preincubated with ATP (Fig. 3, lane 3). Therefore, the five proteins are likely to be functional species with respect to ATP binding.
The T. infestans apyrases inhibited ADP-induced human platelet aggregation (Fig. 4). This inhibition assay was run by incubating increasing concentrations of crude saliva with PRP followed by 5-min incubations with 5 M ADP. The experiment illustrated in Fig. 4 depicts 27% inhibition of platelet aggregation with 1 g of crude saliva and complete inhibition with 5 g and as low as 30 ng of purified apyrases. The apyrase activity of saliva spontaneously ejected by one insect completely abolished aggregation of platelets equivalent to 1 ml of human blood.
Characterization of the Apyrases-Enzymatic characterization was performed to assess the apyrase nature of the purified proteins. The purified mixture displayed the same optimal temperature (37°C) and optimal pH (pH 8.0) for the hydrolysis of ATP and ADP (maximum specific activities of 1500 and 1200 units/mg, respectively). Moreover, the two activities showed similar dependence upon divalent cations (1 mM final concentration) with a maximal hydrolytic activity against ATP and ADP obtained in the presence of 2 mM of Mg 2ϩ , Co 2ϩ , or Mn 2ϩ , whereas Ca 2ϩ , Cu 2ϩ , and Zn 2ϩ did not influence enzyme activity.
The enzymatic activity was ranked for nucleoside 5Ј-triphos- phates (ATP Ͼ CTP Ͼ GTP Ͼ UTP Ͼ ITP) and nucleoside 5Ј-diphosphates (ADP Ͼ IDP Ͼ CDP Ͼ GDP Ͼ UDP), which corresponds to a typical apyrase pattern. Consistent with these observations, the purified fraction did not hydrolyze nucleoside 5Ј-monophosphates, glycerol phosphate, glycose 6-phosphate, and UDP-galactose. A series of ATPases, ATPDases, alkaline phosphatase, protein phosphatase 1 and 2A, and adenylate cyclase inhibitors were tested as described under "Experimental Procedures." Both ATPase and ADPase activities were inhibited to similar levels by 1 mM vanadate and 1 mM NaF, indicating that the two enzymatic activities are unlikely to be held by different active sites. This was confirmed by a mixed substrate experiment (32). The apyrase activity against a mixture of ATP and ADP was close to the mean of activities determined for hydrolysis of each of the two substrates used separately (data not shown). Taken together, all of these observations are consistent with a single site for nucleoside 5Јdiphosphates and -triphosphates in each protein and exclude the possibility that the purified molecular species could correspond to a mixture of ATPases and ADPases.
Analysis of the Saliva Apyrases by Mass Spectrometry-The masses of peptides obtained by in-gel digestion of the 88-and 79-kDa proteins were compared with the in silico digestion of Swiss Protein and TrEMBL data base entries (33,34). The peptide mass maps from the 88-and 79-kDa proteins presented striking similarities with the Ae. aegypti apyrase map, and partial sequences were also conserved. To confirm these results, MS/MS peptide sequences were obtained from the 88-, 82-, 79-, and 68 -67-kDa proteins (Fig. 5). Because of their close migration, the 68-and 67-kDa species were excised together. Each of the four samples excised from the gel showed sequence similarities with 5Ј-nucleotidase family proteins (17,18,21,35,36). The VPVV(Q/K)A sequence, belonging to the highly conserved region of this protein family, was found in each polypeptide chain. In summary, the results presented in Fig. 5 reveal several similarities between the T. infestans saliva apyrases and members of the 5Ј-nucleotidase family belonging to species of four different orders. Although the five apyrase species display peptides with high similarities, differences among some residues suggest a diversity of genetic origin.
DNA Sequence of the 79-kDa Apyrase-We decided to clone the gene of the 79-kDa apyrase because this polypeptide is abundant in all individuals in the insectary. Degenerated oligonucleotides were derived from a peptide sequence obtained by mass spectrometry sequencing and a sequence motif conserved among the 5Ј-nucleotidase family. These oligonucleotides were used to PCR-amplify a unique cDNA fragment from a salivary gland cDNA library. This 1.2-kb fragment was used as a probe to screen the library, leading to the isolation of five clones. These clones presented cDNA inserts of different lengths but identical sequences, all truncated in their 5Ј extremity. The 5Ј-RACE strategy was successfully employed to complete the missing 5Ј end of the clone, leading to a 1-kb fragment showing a 440-bp overlap with the sequence from the longest cDNA clone. Deduced translation revealed a 390-bp 5Ј-untranslated region followed by an open reading frame corresponding to 557 amino acid residues (Fig. 6). The SignalP program (37) predicted a 23-residue signal peptide leading to a mature protein of 59.8 kDa. The sequences of nine peptides from the 79-kDa apyrase were found within the sequence of the predicted mature protein, displaying one mismatch over 84 residues.
Additionally, the cloned cDNA corresponding to the salivary apyrase was subjected to heterologous expression using the baculovirus expression system. Four constructions were tested, with or without His tag combined with native signal peptide or the mellitin signal peptide, known to favor protein secretion (38). Both signal peptides led to the same secretion of a protein that was detected in the supernatant of the insect Sf9 culture medium by a monospecific antibody raised in rabbit, which co-migrated with the 79-kDa T. infestans salivary apyrase (Fig.  7). Unfortunately, no apyrase activity could be detected, despite the presence in the culture medium of 0.4 M Zn 2ϩ and 7 mM Mg 2ϩ , which could be involved in the catalytic site (39,40).
When submitted to BLAST analysis, this protein sequence exhibited 40% identity with various 5Ј-nucleotidases and 32% identity and 50% similarity with the apyrase from Ae. aegypti. Moreover, the sequence corresponding to the mature protein was aligned with the sequence from the Escherichia coli periplasmic 5Ј-nucleotidase with the SAS (sequence annotated by structure) program (41). The T. infestans apyrase showed 25% identity with this 5Ј-nucleotidase, for which the structure is known (42), and the most important residues involved in metal coordination and substrate recognition were aligned (Fig. 6). These findings confirm that the 79-kDa apyrase from T. infestans belongs to the 5Ј-nucleotidase family. The genomic fragment corresponding to the coding sequence was amplified by PCR, and sequencing showed it to be intron-free.
Mapping the apy79 Gene in the Genome of T. infestans-Southern blot hybridizations were carried out on genomic DNA pooled from 50 adult T. infestans insects and hybridized with the 32 P-labeled apy79 cDNA. Genomic DNA digested with XhoI or BamHI, two sites that are absent within the apy79 cDNA sequence, produces two high molecular weight DNA fragments hybridizing with the probe (Fig. 8). Digestion with HindIII restriction enzyme, which is expected to cut within the apy79 gene, led to five bands: four high molecular weight species and the expected 468-bp fragment resulting from internal diges- tion. These results are consistent with the presence of two copies of the apy79 gene. DISCUSSION Triatomine vectors of Chagas' disease were shown to overcome host platelet aggregation by secreting apyrase activity in their saliva during feeding (3). In the present report, we show that pools of saliva from laboratory populations of T. infestans contain five glycosylated proteins of 88, 82, 79, 68 and 67 kDa associated with apyrase activity. The apyrase nature of these proteins is supported by peptide sequences similar to those of the 5Ј-nucleotidases. Moreover, they all specifically bound an ATP analogue, and the sequence of the gene encoding the 79-kDa apyrase displayed unambiguous similarities with the 5Ј-nucleotidase family. Furthermore, we showed that the purified apyrase mixture is a very potent inhibitor of human platelet aggregation.
The presence of five apyrases in T. infestans saliva is likely explained by gene multiplication and subsequent divergence. We considered that these features of the T. infestans salivary apyrases are unusual, as other arthropod saliva described thus far display a unique apyrase. Recently, two apyrase clusters have been reported in Ae. aegypti (43), but when we analyzed the sequences in detail, they appeared to consist of two parts of the already characterized apyrase gene (17). However, an apyrase and an apyrase-like gene were identified in An. gambiae (18) by functional genomics, but a conclusion as to whether they represent two different apyrases could not be drawn. In An. stephensi, three apyrase activities with distinguishable pIs were identified (15). Moreover, the recent finding of two clusters showing similarities with 5Ј-nucleotidases supports the hypothesis that different apyrase genes exist in this species (44). In this paper we present evidence that the apyrase activity found in T. infestans saliva is associated with five different proteins.
We hypothesized that the different apyrases could derive from a single gene by alternative splicing as it occurs for other apyrases (45). However, this possibility was discarded following comparison of the 79-kDa apyrase cDNA and genomic DNA showing the absence of an intron. These observations lead to the conclusion that the apyrases are encoded by at least two independent genes. Moreover, because peptides sequences obtained from different protein bands differed in few amino acids residues, it is likely that the various apyrases do not represent different post-translational modifications of a single protein core.
Because the 79-and 68 -67-kDa proteins are present in the saliva of each triatomine in our colony, whereas the 88-and 82-kDa apyrases are absent in some individuals, 3 the 88-and 82-kDa apyrases may correspond to a gene different from the 79-and 68 -67-kDa apyrases. As the 79-and 68 -67-kDa species are present in all individuals but display differences in amino acid sequences, they should be encoded by different loci. Combining these elements leads to the conclusion that at least three apyrase loci exist in T. infestans, whereas it is not possible to determine the relation between the 68-and 67-kDa apyrases. However, it might be that these two latter proteins originate from the same gene product with different post-translational modifications including limited proteolysis. Southern hybridization with the apy79 cDNA as a probe showing two bands with the T. infestans genomic DNA digests of XhoI and BamHI and five bands with the HindIII is consistent with the presence of two copies of the apy79 gene. This observed gene duplication supports the hypothesis that the different apyrases arose from gene duplication and subsequent divergence.
Differences in the sequence of the vasodilator maxadilan among populations of the sandfly are thought to be related to antigenic variations (46). We propose that the gene multiplication shown by T. infestans could be an important mechanism of enzyme amplification, in addition to a mechanism of escape from the host immune response. Therefore, the apparent redundancy of multiple apyrases may represent a T. infestans mechanism to enhance meal acquisition, developed during evolution to hematophagy. Cloning of the other apyrase genes would provide stimulating evolutionary data by comparing the different genes and assessing the gene duplication hypothesis. Furthermore, it would be interesting to examine whether these genetic features are also present in the field populations. If different apyrase genetic patterns were found in different geographical locations, salivary apyrases could be used, together with other polymorphic loci, as a molecular marker for studying the settlement history of T. infestans in the American continent.
Convergent evolution is the prevailing model for adaptation to blood feeding, because species that diverged before emergence of hematophagy independently developed the same molecular tools. Indeed, peptide and cDNA sequencing show that T. infestans apyrases belong to the same family as those of the mosquitoes Ae. aegypti, An. gambiae, and An. stephensi (17,18,44) but differ from the apyrase from the bedbug C. lectularius (20) to which it is phylogenetically closer. Enzymatic characterization showed that T. infestans apyrases hydrolyze nucleosides 5Ј-triphosphates and 5Ј-diphosphates with preference for ATP and ADP, characteristic of this enzyme family. No phosphohydrolase activity on AMP or other nucleoside 5Ј-monophosphates was observed, in contrast to ancestral 5Ј-nucleotidases, which are able to hydrolyze nucleoside 5Јmonophosphates (39). Thus ADP degradation by apyrases leads to AMP production and not adenosine. Both AMP and adenosine are thought to benefit blood feeding (47), but the evolution of apyrase from 5Ј-nucleotidase to favor AMP production may reflect a more important anti-hemostatic effect of AMP.
Bacterial 5Ј-nucleotidases hydrolyze nucleoside 5Ј-tri-, di-, and monophosphates and can be membrane-anchored or soluble. On the other hand, vertebrate 5Ј-nucleotidases only hydrolyze nucleoside 5Ј-monophosphates and are often bound to membrane by a glycosylphosphatidylinositol anchor (39). In contrast, T. infestans 79 kDa apyrase is a truly soluble protein, as it lacks the C-terminal hydrophilic region and the conserved Ser residue present in the vertebrate glycosylphosphatidylinositol-anchored 5Ј-nucleotidase. This situation is also found in salivary 5Ј-nucleotidases from mosquito, sandfly, and tick (17,21,48).