2-DE-based proteomic investigation of the saliva of the Amazonian triatomine vectors of Chagas disease: Rhodnius brethesi and Rhodnius robustus

Page 1

J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te / j p ro t

2-DE-based proteomic investigation of the saliva of theAmazonian triatomine vectors of Chagas disease: Rhodniusbrethesi and Rhodnius robustus

Camila M. Costaa, Marcelo V. Sousaa,⁎, Carlos André O. Ricarta, Jaime M. Santanaa,Antonio R.L. Teixeirab, Peter Roepstorffc, Sébastien Charneaua,d,⁎a Department of Cell Biology, Institute of Biology, University of Brasilia, Brasilia, Brazilb Faculty of Medicine, University of Brasilia, Brasilia, Brazilc Department of Biochemistry & Molecular Biology, University of Southern Denmark, Odense, Denmarkd Faculty of Ceilândia, University of Brasilia, Brasilia, Brazil

A R T I C L E I N F O

⁎ Corresponding authors. Laboratório de BioqBiológicas, Campus Darcy Ribeiro, Universida

E-mail addresses: [email protected] (M.V.

1874-3919/$ – see front matter © 2011 Elsevidoi:10.1016/j.jprot.2011.02.022

A B S T R A C T

Available online 27 February 2011

The triatomine bugs are obligatory haematophagous organisms that act as vectors ofChagas disease by transmitting the protozoan Trypanosoma cruzi. Their feeding success isstrongly related to salivary proteins that allow these insects to access blood bycounteracting host haemostatic mechanisms. Proteomic studies were performed on salivafrom the Amazonian triatomine bugs: Rhodnius brethesi and R. robustus, speciesepidemiologically relevant in the transmission of T. cruzi. Initially, salivary proteins wereseparated by two-dimensional gel electrophoresis (2-DE). The average number of spots ofthe R. brethesi and R. robustus saliva sampleswere 129 and 135, respectively. The 2-DE profileswere very similar between the two species. Identification of spots by peptide massfingerprinting afforded limited efficiency, since very few species-specific salivary proteinsequences are available in public sequence databases. Therefore, peptide fragmentationand de novo sequencing using a MALDI-TOF/TOF mass spectrometer were applied forsimilarity-driven identifications which generated very positive results. The data revealedmainly lipocalin-like proteins which promote blood feeding of these insects. Theredundancy of saliva sequence identification suggested multiple isoforms caused by geneduplication followed by gene modification and/or post-translational modifications. In thefirst experimental assay, these proteins were predominantly phosphorylated, suggestingfunctional phosphoregulation of the lipocalins.

© 2011 Elsevier B.V. All rights reserved.

Keywords:TriatomineSalivaProteomicsPhosphorylationMALDI-MS/MSChagas disease

1. Introduction

The triatomine bugs (Hemiptera: Reduviidae: Triatominae),commonly known as kissing bugs, are the natural vectors ofChagas disease, also known as American trypanosomiasis.

uímica e Química de Prde de Brasília, Brasília-DSousa), [email protected]

er B.V. All rights reserved

During blood feeding, they release faeces contaminated withTrypanosoma cruzi that infects the humans or vertebratereservoirs that are fed upon. More than a century after itsdiscovery and description by Carlos Chagas, this illness isstill considered a neglected disease by the WHO with

oteínas, Departamento de Biologia Celular, Instituto de CiênciasF, 70910-900, Brazil. Tel.: +55 61 3307 3094.(S. Charneau).

.

Page 2

1653J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

approximately 8 million people infected worldwide. Thedisease caused about 11,000 deaths throughout much ofMexico, Central America, and South America in 2008 [1].

The triatomine bugs are successful haematophagousarthropods. Such success is attributed to the presence ofpharmacologically active molecules in their saliva whichmaintain a fluid blood flow and generally make the bitepainless and asymptomatic. These substances, such as antic-oagulants, vasodilators, and platelet aggregation inhibitors,counteract vertebrate host haemostatic events and modulateboth the vertebrate host immune-response and inflammation.In addition, they also preserve the saliva integrity by workingas anti-microbial compounds.

Of the six tribes of Triatominae subfamily, the Rhodniiniand Triatomini are the two major tribes that are epidemio-logically more important tribes [2]. The domestic bug Triatomainfestans, of the Triatomini tribewas, until recently, consideredthe most relevant vector in the epidemiology of Chagasdisease in some ecosystems of South America, notably inBrazil [3,4]. Consequently, it is one of the best-studied species[5]. Recently, a proteomic investigation of T. infestans salivawas performed reporting more than 200 salivary secretedproteins by 2-DE, with a significant presence of isoforms [6].Moreover, the majority of the identified proteins are poten-tially engaged in platelet-aggregation inhibition, and belong tolipocalin and apyrase families [6]. Furthermore, considerabletranscriptomic analyses of salivary glands were performed forthree Triatoma species: T. brasiliensis [7] and T. infestans [8], bothof South America [9], and T. dimidiata [10] of North America [9].All Triatoma cDNA library insights showed that 55.0%, 89.9%and 93.8% of predicted salivary proteins, respectively ofT. infestans [8], T. dimidiata [10], and T. brasiliensis [7], werevariants of the lipocalin family. More recently, transcriptomicanalysis of salivary glands of Dipetalogaster maxima of theTriatomini tribe, an inhabitant of Mexico, revealed that 93% ofthe transcripts coding for putative secretory proteins corre-sponded to lipocalins as confirmed by proteomic analyses [11].

The most studied triatomine saliva is Rhodnius prolixus, aRhodniini tribe domestic bug. R. prolixus is a very importantvector in Latin America, particularly in northern countries ofSouth America – Venezuela and Colombia – and in some areasof Central America [5], principally savannas and some areas oftropical forests [12]. Extensive transcriptomic and proteomicanalyses of the salivary glands of these species permitted aremarkable expansion of the known lipocalin family [13],including the nitrophorins. These proteins are described asnitric oxide (NO)-binding heme lipocalins which primarilyfunction as vasodilators. Interestingly, no nitrophorin hassubsequently been identified in the saliva of Triatomini tribeindividuals.

Within the Rhodniini tribe groups, two sylvatic species ofthe genus Rhodnius, R. brethesi and R. robustus, cohabit in theAmazon rainforest that encompasses northern areas of Brazil,Venezuela and Colombia. In the last decade, a review [14]reported an increase in the number of human cases of Chagasdisease. This may be an indication that the illness is emergingin the Amazonian region. These Rhodnius bugs were evaluatedas being relevant in the vectorial transmission of T. cruzi and,as they predominantly occur in this endemic area, they couldbe participating in the proliferation of the protozoa [14].

All studies conducted on these species have mainlybeen restricted to biological and morphological aspects. Theystated that R. brethesi and R. robustus are morphologicallyand genetically very similar to R. prolixus, though the lasttwo are more distinct from the first [15,16]. In fact, the specieslevel identification of Rhodnius can be so difficult that, theprecise geographical distribution of some Rhodnius specieswould remain unclear [15,16]. This is particularly peculiar asR. prolixus is considered a major vector and is a domesticinsect, whereas R. robustus is a sylvatic insect. Besides, salivaryheme-containing protein (nitrophorins) profiles of both weredistinguished on a starch gel electrophoresis plate [17].

Despite the relevant emergence of sylvatic R. brethesi andR. robustus vectoring, little is known about their salivarycompounds. With no sequenced or characterized proteinsavailable in sequence databases, proteomic characterizationbecomes more complicated. Conversely, R. prolixus saliva iswell documented [13,18]. Currently, the sequencing of itsgenome is in progress at the Washington University GenomeSequencing Center. Since these bugs belong to the samegenus, R. prolixus salivary data could permit exploration of thesaliva proteomes of R. brethesi and R. robustus by cross-speciesprotein searching.

Herein, this study investigated and characterized the salivaryproteomic diversity of two Amazonian triatomines, R. brethesiand R. robustus, using two-dimensional gel electrophoresis andMALDI–TOF/TOF mass spectrometry, and compared the resultswith the saliva of R. prolixus. We report the first 2-DE mapsRhodnius genus saliva. The present study shows that the salivaproteomesof the blood-feedingR. brethesi andR. robustus are verysimilar in that they both contain anti-haemostatic molecules,mainly R. prolixus lipocalin-homologous proteins, which aresubject to phosphorylation.

2. Materials and methods

2.1. Harvesting of triatomine bug saliva

Rhodnius brethesi and R. robustus colonies were maintained at28 °C and 70% relative humidity. Salivary glands were obtainedfrom adult individuals of each species and dissected at 7 to9 days after the blood meal. Glands were punctured, and intra-luminal fluids harvested by centrifugation (2000×g, 5 min, 5 °C).Protein samples from the saliva of both species were quantifiedusing the Plus One 2D Quant Kit (GE Healthcare, Uppsala,Sweden) according to manufacturer's instructions, withinparallel amino acid analysis in an amino acid analyzer (HitachiL8500, Tokyo, Japan) by means of ninhydrin post-columnderivatization. Saliva protein integrity was verified by visualiz-ing the profiles in a 12% SDS-PAGE silver-stained gel asdescribed by Blum et al. [19], with adaptation. The 1D gel wasrun on an SE 600 vertical slab gel electrophoresis unit (Hoefer,San Francisco, CA, USA) at 20 °C under denaturing conditions.

2.2. Two-dimensional gel electrophoresis

Eighty micrograms of saliva protein were solubilized anddenatured in 350 μL of 2-DE sample buffer (7 M urea, 2 Mthiourea, 2% v/v Triton X-100, 65mM DTT, 0.5% v/v Pharmalyte

Page 3

1654 J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

3–10, and 10% isopropanol). Briefly, samples were applied to18 cmpH3–10 IPGgel strips, andsubmitted to isoelectrofocusing(IEF) using an Ettan IPGphor3 Unit (GE Healthcare) as recom-mended by the manufacturer. A voltage of 30 V was applied tothestripsduring 6 h, followedby500 V for 1 h, 1000 V for 1 hand,8000 V for 4 h, with a maximum current of 50 μA per strip, asdescribed by Charneau et al. [6]. After focusing, the proteins onthe strips were reduced, alkylated and separated by 12% SDS-PAGE on a Protean II system (BioRad, Richmond, CA, USA) [19].After silver-staining, the gelswere either stored in 1%acetic acidor dried in 1:5:30 acetic acid/glycerol/methanol (v/v/v).

2.3. Image analysis

Silver-stained gels were scanned with an Image Scanner(PowerLook 1120; Amersham Biosciences, Uppsala, Sweden)at 300 dpi resolution. Digitized images were analyzed with theImageMaster™ 2D Platinum 6.0 software (GE Healthcare) tomatch and count protein spots.

2.4. Protein identification by mass spectrometry

2.4.1. In-gel digestionThe spots were excised, transferred to microtubes and in-geldigested with trypsin (Promega, Madison, USA) using anoptimized protocol [20,21]. The spots were washed with 50 μLof H2O, and dried with 90 μL of 100% acetonitrile. The proteinswere then subjected to reduction with 40 μL of 10mM DTT in100 mM NH4HCO3, at 56 °C for 30min, followed by alkylationwith40 μLof10.2 μg/μL iodacetamide in100 mMNH4HCO3 in thedark for 30min. After removal of the iodacetamide alkylationsolution, the spots were briefly washed with 100 μL of 50%acetonitrile (v/v), and dehydrated with 100 μL of 100% acetoni-trile for 15min. After removal of the acetonitrile, the spots weredried by vacuum centrifugation in a speed vac for 15min inorder to rehydrate themwith 20 μL of 50 mMNH4HCO3 contain-ing7 ng/μLof trypsinon ice for20min.Thenon-absorbedexcesstrypsin containing solution was removed and several micro-liters of NH4HCO3 were added in order to submerge the spotsprior to overnight incubation at 37 °C.

2.4.2. Peptide derivatization with 4-sulfophenyl isothiocyanate(SPITC)The N-terminal derivatization procedure with SPITC (Aldrich)was employed for de novo sequencing analyses. The SPITCsolution (10 μg/μL SPITC in 50 mM NH4HCO3, pH 8.6) wasalways freshly prepared as described by Zhang et al. [22]. Onevolume (about 7.5 μL) of SPITC solution was added to onevolume of each digest and the mixes were incubated at 55 °Cfor 30 min. The reactions were stopped by the addition of 1 μLof 5% TFA (v/v) [22].

2.4.3. Desalting and loading samples on the MALDI targetThe crude digests or the derivatized digests were desalted andconcentrated using a Poros R2 resin column of reversed-phasechromatography prepared in GELoader tips (Eppendorf) about0.5 cm long [23]. The columns were washed with 0.1% TFA (v/v)and theanalyteselutedon theMALDI targetusing 3 μLof 5 μg/μLCHCA (α-cyano-4-hydroxycinnamic acid, Sigma) in an aqueoussolution containing 70% acetonitrile (v/v), 0.1% TFA (v/v).

2.4.4. Mass spectrometryRecovered peptides were analyzed using MALDI–TOF andMALDI–TOF/TOF (Reflex IV e Autoflex II — Bruker Daltonics,and AB4800 — Applied Biosystems) mass spectrometersoperating in positive ion reflector mode. The spectra wereprocessed using the FlexAnalysis 2.4 and BioTools 3.0 soft-wares (Bruker Daltonics) or MoverZ Freeware Edition (http://bioinformatics.genomicsolutions.com/MoverZ.html). Proteinidentification was performed using three different strategies:PMF and MS/MS peptide fragmentation followed by MASCOTsearches (http://www.matrixscience.com) against the NCBI nrprotein database and de novo sequencing followed by MSBLAST sequence-similarity searches against the nr database(nrdb95) at http://genetics.bwh.harvard.edu/msblast/ [6,24].The mass spectrometers were calibrated using the Pepmix(Bruker Daltonics). Spectra containing autodigested trypsinpeptides were internally calibrated. MASCOT search para-meters were as follows: tryptic peptides with up to onemissedcleavage site; metazoa as taxonomic restriction; propionami-dation or carbamidomethylation of Cys residues as fixedmodification (depending if the alkylation was done withacrylamide or iodacetamide); and oxidation of methionine asvariable modification. The parameters of monoisotopicmass accuracy used were up to 100 ppm and 0.5 Da for PMFand MS/MS data, respectively. All precursor and fragmentions were single charged. Protein identifications by MS/MSpeptide fragmentation using MASCOT were considered confi-dent if hits were produced bymatching of at least threeMS/MSspectra with each peptide ion score higher than 20 or if hitswere produced bymatching of one or two spectra with at leastone peptide ion score of 50 or higher as established in [6,25]. ForMS BLAST searches, all SPITC-derivatized peptide sequencesobtained by de novo sequencing of all peptide precursors weremerged into a single MS BLAST query string before searches.High hits with a score above 55 were considered as establishedin [6,25].

2.5. Direct revelation of phosphorylated proteins

Following 2-DE, spots corresponding to phosphoproteins weredyed with Pro-Q Diamond reagent (Invitrogen, Carlsbad, CA,USA) according to the manufacturer's protocol and then silver-stained.Thisproduct interactswithphosphorylated threonines,tyrosines and serines forming a fluorescent complex. Gelimages were digitalized by a Typhoon 9410 PhosphorImagerscanner (AmershamBiosciences). The excitation source and theemission filter were programmed to work at 532 nm and560 nm, respectively, in order to access the product'smaximumfluorescence. Gels were silver-stained for subsequent analysisof these spots. This procedure does not interfere with thesubsequent mass spectrometry investigation.

3. Results

3.1. Gel electrophoresis profiles ofR. brethesi andR. robustussalivary secreted proteins

Firstly, the salivary secreted proteins of the two bugs ofRhodnius genus were analyzed by SDS-PAGE to evaluate the

Page 4

1655J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

integrity of the proteins post-harvest (Fig. 1). The 1-DE analysisshowed that both profiles were indistinguishable with fewbands displayed. The protein bandswere observed in themassrange from 14 to 30 kDa which is compatible with themolecular masses of the predicted proteins encoded by fulllength cDNA sequences from R. prolixus salivary glands [13]and some R. prolixus characterized proteins such as nitrophor-ins (NP) [26–28], biogenic amine-binding protein (ABP) [29] andR. prolixus aggregation inhibitor 1 (RPAI-1) [30].

Comparative 2-DE analysis of the two Rhodnius speciesusing isoelectric focusing within a wide pH range (3-10) and12% SDS-PAGE in second dimension was performed using theprotocol previously optimized for salivary proteins [6]. Upon80 μg loading, the 2-DE profiles of the two saliva types wereextremely similar (Fig. 2). Several repetitions were performedand the protein spot profiles obtained displayed reproduciblenumbers of protein spots, their relative positions and inten-sities. Although there are some exclusive spots in eachsalivary 2-DE profile, most were common to both species.The average numbers of silver-stained detected spots were129 for R. brethesi (Fig. 2A) versus 135 spots for R. robustus(Fig. 2B).

3.2. Protein identification

As there was a high degree of similarity between both saliva 2-DE profiles including number, position and intensity of thespots, and between their biological traits of life and thedifficulty to differentiate the geographical localization of someRhodnius species [15,16], we considered that their proteincompositions were similar. 56 well-focused landmarks repre-senting all of the spot clusters scattered over both 2-DE gelswere cut from the R. brethesi saliva gels (Fig. 2A). After in-geltrypsin digestion, the protein sequences were analyzed by

Fig. 1 – 1-DE profile of triatomine salivary extracts of theRhodnius genus. Laneswith 2 μg of salivary secreted proteinsof R. brethesi (1) and R. robustus (2) were separated by 12%SDS-PAGE under reducing conditions and silver-stained.

Fig. 2 – Wide pH range 2-DE maps of salivary proteins ofRhodnius genus triatomines. Gels of R. brethesiwith 129 spots(A) and R. robustus with 135 spots (B), containing 80 μg ofsalivary protein. 2-DE was performed under denaturingconditions using 3–10 linear IPG strips in IEF and 12%SDS-PAGE. The gels were silver-stained. Spot numberscorrespond to the identified proteins.

MALDI–TOF/TOF mass spectrometry. The resulting massspectra were used to search for the protein identities againstthe NCBInr database using the MASCOT software.

The search was initiated by peptide mass fingerprinting(PMF) method (Table 1). As expected, due to a lack of genomicdata for these species, this conventional database searchingapproach delivered limited identification efficiency. This isbecause PMF relies on exact correlations of the mass-to-

Page 5

Table 1 – Identification of R. brethesi salivary proteins by peptide mass fingerprinting (PMF).

Spot a Identity Accessionnumber(NCBI) c

Score d Cov.(%) e

Mass(kDa)

pI

[Rhodnius prolixus] b Exp. andTheo. f

Exp. andTheo. g

R. brethesi salivary amine binding proteins3 ABP gi|27968047 112>72 49 30 24.7 6.9 6.0

R. brethesi salivary nitrophorins6 NP-1A precursor gi|33518667 104>73 49 29 24.7 6.8 6.210 Complex NP-1 with Histamine gi|3212578 84>73 30 27 20.7 6.9 6.415 NP-4 precursor gi|3219833 94>74 34 24 22.4 7.0 6.817 NP-2 (Prolixin-S) gi|8569632 169>77 68 24 20.3 6.6 6.1

R. brethesi salivary aggregation inhibitors25 AI-5 precursor gi|33518673 75>74 34 20 20.8 4.7 6.0

a Spot number indicated in Fig. 2A.b Results obtained by MASCOT software search in the NCBI nr database.c Accession number in the NCBI protein database. All accession numbers refer to sequences from R. prolixus.d Probability-based Mowse score of MASCOT software that evaluates if the peptides subjected to search are the same as those found in thedatabase originated by in silico digestion of a known protein. The numbers on the right of the symbol “>” represent the minimum scorestatistically significant as calculated by the software (p<0.05).e Coverage is the percentage of predicted protein sequence covered by matched peptides via MASCOT.f Experimental and theoretical molecular masses of proteins, visualized in the gels and calculated from amino acid sequences, respectively.g Experimental and theoretical isoelectric points of proteins, visualized in the gels and calculated from amino acid sequences, respectively.

1656 J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

charge ratios of digested peptides to corresponding databaseentries, which generally requires the availability of genomicsequences. Only 6 proteins were identified by PMF withsignificant hits and all corresponded to salivary proteins ofR. prolixus (Table 1).

Subsequently, for the non-identified proteins, the trypticpeptides corresponding to the most intense peaks of the PMFspectra were fragmented in the mass spectrometer. Thus, 20proteins, homologous to R. prolixus salivary proteins, wereidentified by the conventional tandem mass spectrometry(MS/MS) based on peptide fragmentation and exact correla-tions of mass over charge values of precursors and fragmentsusing MASCOT. Two borderline identifications (spots 7 and 24)provided a MASCOT Mowse score inferior to 50. Since theymatched to R. prolixus salivary protein sequences they wereadded to the list of identified proteins (Table 2).

Finally, when the stringent MASCOT database searches nolonger led to successful results, database mining wasperformed by de novo interpretation of unassigned spectraand MS BLAST sequence-similarity searches with deduced

Notes to Table 2⁎ Tolerant search to error. In this case, since the identification was inmodifications were tested and residue substitutions were considered.a Spot number indicated in Fig. 2A.b Results obtained by MASCOT software search in the NCBI nr database.c Accession number in the NCBI protein database. All accession numbersd Probability-based Mowse score of MASCOT software that evaluates ifdatabase originated by in silico digestion of a known protein. The numstatistically significant as calculated by the software (p<0.05).e Coverage is the percentage of predicted protein sequence covered by mf Experimental and theoretical molecular masses of proteins, visualizedg Experimental and theoretical isoelectric points of proteins, visualized inh Peptide sequences identified via MASCOT following the experimental pewith peptide ion score below 20 was considered [6,25].i Indicates that the probability-based Mowse score of MASCOT is below 5

sequence candidates. Moreover, in order to enhance the denovo sequencing interpretation of spectra, the SPITC deriva-tization procedure was employed. Hence, 10 additionalproteins were identified, all corresponding to R. prolixussalivary proteins (Table 3). Thirty-six of the 56 R. brethesiprotein spots analyzed were identified by successive PMF,MS/MS analysis using MASCOT and de novo sequencingusing MS BLAST. All identifications were significant only forproteinswith a high degree of sequence similarity to R. prolixussequences available in the NCBI database.

Based on the same rationale, protein identification wasperformed for seven spots found to be exclusively expressedin the R. robustus salivary 2-DE gels (Fig. 2B). Peptidefragmentation using MASCOT identified three proteins thatwere homologous to R. prolixus salivary secreted proteins(Table 4).

Finally, 61% of the analyzed spots of both species wereidentified mainly by peptide fragmentation using MASCOTand de novo sequencing using MS BLAST, rather than PMF.The results suggest that salivary protein sequences of

valid, the search was repeated with wider parameters. All possible

refer to sequences from R. prolixus.the peptides subjected to search are the same as those found in thebers on the right of the symbol “>” represent the minimum score

atched peptides via MASCOT.in the gels and calculated from amino acid sequences, respectively.the gels and calculated from amino acid sequences, respectively.ptide masses after parental ion fragmentation. No matched spectrum

0 but matched to R. prolixus salivary protein sequences [6,25].

Page 6

1657J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

R. brethesi, R. robustus, and R. prolixus are homologous, but arenot identical. Moreover, we observed that tandem massspectrometry is undeniably the most adequate proteinidentification strategy when working with species missingrich genomic data. Nevertheless, in spite of spectra obtainedfrom SPITC-derivatized peptides sequenced de novo andsubmitted to search via MS BLAST, some spots were notidentified, for example the annotated R. robustus spot ‘NI’ inFig. 2B (Fig. 3).

Table 2 – Identification of R. brethesi salivary proteins by MS/MS

Spot a Identity Accessionnumber(NCBI) c

Score d Cov(%) e

[Rhodnius prolixus] b

R. brethesi salivary amine binding proteins2 ABP gi|27968047 136>38 10

4 ABP gi|27968047 110>39 10R. brethesi salivary nitrophorins7 NP-3 gi|3319210 i 49>40 6

285>39 128 NP-1 precursor gi|3219825

NP-4 precursor gi|3219833

9 NP-3 precursor gi|3219827 71>36 1042>35 6

12 NP-3 precursor gi|3219827 79>32 618 NP-2 precursor (Prolixin-S) gi|3219826 546>38 32

19 NP-2 precursor (Prolixin-S) gi|3219826 126>39 8R. brethesi salivary triabin-like lipocalin proteins11 TLLP-4 precursor gi|33518683 110>38 14

23 TLLP-2 precursor gi|33518679 57>37 524 TLLP 2 precursor gi|33518679 i 36>32 5

R. brethesi salivary aggregation inhibitors1 AI-7 precursor gi|33518711 257>39 38

26 AI-5 precursor gi|33518673 129>38 10

30 AI-7 precursor gi|33518711 51>37 7

31⁎

31 RPAI-2 gi|1572727 87>39 832 RPAI-2 gi|1572727 82>38 833 AI-7 precursor gi|33518711 67=67 734 AI-4 precursor gi|33518671 243>34 17

35 AI-7 precursor gi|33518711 53>48 13R. brethesi salivary proteins with unknown function36 Precursor of MYS-1 gi|33518705 61>50 10

3.3. Functional classification of the identified secretedsalivary proteins of the Rhodnius genus

The 39 identified macromolecules were related by sequencesimilarity to previously deposited R. prolixus proteins in theNCBI database and described in the first annotated catalogueof triatomine salivary proteins [13,18]. With the exception ofone protein that was found to be a R. prolixus MYS-1 with nosignificant similarities to known proteins, corresponding to

peptide fragmentation using MASCOT.

. Mass

(kDa)

pI Peptide ion (m/z) andsequence h

Exp. andTheo. f

Exp. andTheo. g

(peptide ion score>20)

30 24.7 6.3 6.0 1194.6 — TYSYDISFAK1336.7 — LSGLFDATTLGNK

30 24.7 7.2 6.0 1523.8 – IKEVFSNYNPNAK

26 22.1 6.2 6.1 1488.8 — DLGDLYTVLSHQK27 6.4 6.8 1135.5 — EALYHYDPK

23.0 1248.6 — DLGDLYAVLNR22.7 1376.7 — LKEALYHYDPK

1547.8 — GNKDLGDLYAVLNR27 22.1 9.8 6.1 2492.0— DSYTLTVLEADDSSALVHICLR

1488.7 — DLGDLYTVLSHQK26 22.1 7.1 6.1 1488.9 — DLGDLYTVLSHQK24 22.6 7.2 5.8 1115.5 — TSFYNIGEGK

1502.7 — DLGDLYTVLTHQK1734.9 — SAVTQAGLQLSQFVGTK1903.9 — EGSKDLGDLYTVLTHQK2491.2— NSYTLTVLEADDSSALVHICLR

23 22.6 7.1 5.8 1734.9 — SAVTQAGLQLSQFVGTK

26 18.9 6.4 9.4 1085.5 — SVVQYGYNR1565.8 — TILETDGNSALLYR

21 19.9 5.2 5.9 1213.6 — SADDYFILNR21 19.9 5.5 5.9 1213.8 — SADDYFILNR

35 17.3 9.7 9.0 1077.5 — GQWHLTHAK1464.7 — AVSDFNFDKFFK1931.9 — CGSHEGPTKDNYLVGQR2388.2— FTSYVSVLATDYDNYVLVYR

20 20.8 5.3 6.0 1056.5 — GNWQVTHSK1107.6 — IGAMFPICGK1123.6 — IGAMFPICGK

19 17.2 9.7 9.0 1464.7 — AVSDFNFDKFFK⁎1492.7 — AVSDFNFDKFFK(+ 28 Da , K9)⁎1932.8 — GSHEGPTKDNYLVGQR(+ 175 Da, G N-term)⁎2392.1— FTSYVSVLATDYDNYVLVYR(+ 2 Da, S3 and S6)

19 19.4 8.6 8.2 1717.7 — YGSSAVEDNFLVFNR18 19.4 8.3 8.2 1717.8 — YGSSAVEDNFLVFNR18 16.8 9.7 9.0 1491.6 — ATDYDNYVLVYR17 18.7 7.5 6.7 1038.5 — DNFLLFNR

1272.7 — LNQLELTSLNK1529.7 — FFTGDWYLTHSR

17 16.8 9.7 9.0 2388.0— FTSYVSVLATDYDNYVLVYR

17 16.4 9.0 8.5 1924.7 — LDEEEIDNCEDGPGYR

Page 7

1658 J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

3% of the total, all proteins belonged to only four proteingroups of the lipocalin family. Nitrophorins (NPs) constituted38%, aggregation inhibitor proteins (AIs) comprised 38%, 13%were triabin-like lipocalin proteins (TLLP), and 8% were aminebinding proteins (ABPs) (Table 5). Themost abundant lipocalingroup is constituted of heme-containing NPs, which carrynitric oxide (NO) from gland lumen into the lesion. The NOcauses vasodilatation and inhibits local platelet aggrega-tion [26,31]. After the release of NO, it is suggested that NPmembers can also sequester histamine [32,33]. In addition,NP2 interferes with coagulation factors and thereby preventsclotting [34,35]. AIs are related by sequence similarity to theR. prolixus Platelet Aggregation Inhibitor 1 (RPAI-1), an ADP-binding protein that inhibits ADP-induced platelet aggregation[30,36]. Triabin-like lipocalins are uncharacterized. However,as they share sequence homology with triabin, a Triatomathrombin inhibitor [37,38], they may have the same role inpreventing platelet aggregation by inhibiting thrombin [13].Other proteins were found to match with ABP anti-plateletaggregators. These proteins bind to serotonin and epinephrinesecreted by dense granules of platelets. This contributes to theactivation and aggregation of platelets to induce the formationof thrombus [29]. ABPs binding norepinephrine would preventvasoconstriction in order to equilibrate the function of NPsthat induce vasoconstriction by releasing nitric oxide [18,29].

Moreover, the same NCBI database-deposited proteins wereidentified in different spots on the 2-DE gels (Fig. 2A andB): NP-1(gi|3219825) in spots 13, 14 and 37; NP-2 precursor (gi|3219826) inspots 18 and 19; NP-3 precursor (gi|3219827) in spots 7, 9 and12; RPAI-2 (gi|1572727) in spots 31 and 32; AI-4 precursor(gi|74841089) in spots 28 and 29; AI-7 precursor (gi|33518711) inspots 1, 30, 33 and 35; TLLP-2 precursor (gi|33518679) in spots 22,23 and24; andABP (gi|27968047) in spots 2, 3 and 4. These resultsalso revealed that several proteins were present in multipleisoforms.Although lipocalin familieshaveamulti-genepattern,it is possible that post-translational modifications (PTMs)might also be a source of lipocalin heterogeneity for a proteinsequence, as previously proposed in the case of T. infestanssalivary proteome [6], where the five T. infestans apyrase formswere found to be N-glycosylated [39].

3.4. 2-DE mapping of the salivary phosphoproteome

A phosphate group adds a negative charge to selected aminoacid residues. A single phosphorylation decreases the pI valueof a peptide by 0.1 and increases mass by 80 Da [40].Consequently, in a 2-DE gel, the different degrees of phos-phorylation between spot isoforms can be discerned, mainlyby pI, rather than mass modification.

As several redundant protein identifications revealed thepresence of the same protein in different spots in closeproximity, direct 2-DE gel detection was performed to checkfor possible phosphorylations.

We used a 2-DE gel-based approach to gain insight into theputative phosphoproteins. Sequentially, the 2-DE gels weretreated with phosphoprotein-specific Pro-Q Diamond dye, andphosphorylated protein spots visualized by scanning andcomparison with the total number of silver-stained proteins.Interestingly, a considerable proportion of spots were found tocontain putative phosphorylation sites for both species (Fig. 4A

andB). In comparisonwith the silver-stained gels (Fig. 2AandB),the dyed spots corresponded to several identified lipocalinsindicating that these lipocalins are potentially phosphorylated.Of the 43 R. brethesi phosphorylation-dyed spots, 23 spots wereidentified: 3, 5–10, 12–15, 17, 22–26, 29, 31, and 33–36. Nine spotsappeared to be phosphorylated exclusively in the R. brethesisaliva compared to R. robustus, including eight identified spots(8, 12, 13, 23, 29, 31, 33 and 36). The 34 remaining spots werefound in both types of saliva.

As several 2-DE spots were dyed by Pro-Q Diamond, an insilico analysis was performed in order to predict phosphory-lation sites in the protein sequences, using the followingserver: http://www.cbs.dtu.dk/services/NetPhos/ [41]. Numer-ous predicted phosphorylation sites were found in the 15R. prolixus sequences identified from phosphorylated spots(Table 6). The following neighbouring phosphorylation-dyedspots matched the same protein sequences, but with distinctexperimental pI (Tables 1–3), and consequently appeared to bedifferentially phosphorylated isoform groups: 8 and 15; 8, 13and 14; 9 and 12; 22, 23 and 24; 25 and 26; and 34 and 35.

4. Discussion

In the present study, the application of a proteomic approach,combining 2-DE and MALDI–TOF/TOF mass spectrometryanalysis, enabled the establishment of representative 2-DEsalivary protein maps for R. brethesi and R. robustus secretedproteins.

The high degree of similarity between the 2-DE salivaryprofiles of these Rhodnius species was impressive, in agree-mentwith the high degree of similarity of their biological traits[15,16]. Under the same experimental conditions, the averagenumber of silver-stained spots detected in both cases wasapproximately 130 (Fig. 2) versus 200 found in T. infestanssaliva [6]. This difference could be correlated to the fact thatboth Rhodnius species and T. infestans have, respectively, 2 and3 salivary gland pairs which, consequently, interfere withproteome complexity. The majority of the saliva secretedproteins of both Rhodnius bugs were visualized in the 14 to30 kDamass range in both 1-DE (Fig. 1) and 2-DE gels (Fig. 2), aspreviously observed for Rhodnius species [13] and Triatomaspecies [6–8].

Despite the absence of R. brethesi and R. robustus secretedsalivary protein sequences available in theNCBI nr database, allidentificationswere solelyobtainedbycross-species searchesofavailable protein sequences emerging from the extensive R.prolixus cDNA library [13]. Nevertheless, protein identificationbased on conventional database searching by PMF wasattempted with little success (Table 1). The identification was,on the other hand, successfully performed by MS/MS usingMASCOT (Tables 2 and 4) and completed by de novo sequencing(Table 3). All identifications were genera-specific and fluid-specific; under no circumstances were proteins found to besignificantly identified by cross-species comparison with theavailable sequences of species of genera other than Rhodniussuch as Triatoma [7,8]. It appears that the similarity between thesaliva content of R. brethesi, R. robustus and R. prolixus observedhere corroborates their morphological and genetic similarities

Page 8

Table 3 – Identification of R. brethesi salivary proteins by de novo sequencing and MS BLAST.

Spot a Identity Accessionnumber(NCBI) c

TotalScore d

Matched/submitted

characters e

Mass(kDa)

pI Peptide ion (m/z) andsequence h

[Rhodnius prolixus] b Exp. andTheo. f

Exp. andtheo g

R. brethesi salivary nitrophorins5 NP-1A precursor gi|33518667 197 31/55 29 24.7 6.2 6.2 1472 — VSGFEGNDGQYSGR

1600 — NVLVNQDGTKLDER1823 — S…(1426)LSGLFDATTLSSSK⁎1335 — LLAQQFSEYTR

13 NP-1 precursor gi|3219825 146 21/36 25 23 6.4 6.8 1533 — GNNDLGDLYAVLNR944 — FSDFLSTK⁎1543 — TPHVWNLFLFNR

14 148 19/19 26 7.4 1258 — DLGDLYAVLNR944 — FSDFLSTK

R. brethesi salivary triabin-like lipocalin proteins20 TLLP-1 precursor gi|74841086 153 21/25 23 20.5 9.7 9.9 1427 — FFSGTWFVTHAK

1832 — S…QAELQSENLNTR22 TLLP-2 precursor gi|33518679 163 24/29 21 19.9 4.8 5.9 1583 — FTVASADNNAAVLYK

1635 — SATFSTDDYFVLNRR. brethesi salivary aggregation inhibitors16 AI-6 gi|74841087 68 10/13 24 18.8 5.9 10 1559 — SQDHEDNVFLLSR21 69 10/11 22 8.4 1460 — S…HEENVFVLSR27 RPAI-2 gi|74844443 204 28/66 20 19.4 7.0 9.2 1826 — ANQFFTGDWYLTHAR

2219 — SSVAYGTSTTPEDNFLLFNR1332 — LSQLDLNTSWR⁎2347 — SSFWGTSTTPEDNFLLFNQR

28 AI-4 precursor gi|74841089 120 17/52 20 18.7 6.7 7.5 798 — YALLYR2216 — SSAVYGQSTTPLDNFLVFNR⁎1433 — YQVSPDLHD…R⁎1826 — AMTSTTLPDNFLVFNR

29 140 20/34 20 8.4 1026 — LFSGFDANR1987 — SSGLYGSALQDNFLVFNR⁎897 — HPTLHHR

Bold residues matched either precisely or conservatively the sequence found and underlined residues did not match.Some sequenced peptides (*) did not match available sequences in the NCBI database.a Spot number indicated in Fig. 2A.b Results obtained by MASCOT software search in nrdb95 database using the MS BLAST tool [24].c Accession number in the NCBI protein database. All accession numbers refer to sequences from R. prolixus.d Probability-based Mowse score of MASCOT software that evaluates the if the peptides subjected to search are the same found in the databaseoriginated by in silico digestion of a known protein (p<0.05).e Amino acid number ratio of predicted protein sequence covered bymatched query sequence and the total number of amino acids submitted tothe search.f Experimental and theoretical molecular masses of proteins, visualized on the gels and calculated from amino acid sequences, respectively.g Experimental and theoretical isoelectric points of proteins, visualized on the gels and calculated from amino acid sequences, respectively.h Sequenced peptides submitted to the search (NCBI database).

1659J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

[15,16]. Moreover, since these triatomine bugs inhabit the samebiomes and are therefore exposed to the same selectivepressures, it is probable that their habitat, prey and abioticfactors could have influenced the development of similar salivacompounds. The Rhodnius saliva proteomesmainly contain fourprotein groups of the lipocalin family: nitrophorins, aggregationinhibitor proteins, triabin-like lipocalin proteins and aminebinding proteins. All of which are highly directed to respond tothe haemostatic events of their prey, thus promoting the bug'sblood-feeding. However, to illustrate the lesser salivary pro-teome complexity of Rhodnius, non-lipocalin salivary proteins ofTriatoma species, such as trialysin, triatox, inositol phosphatase,serine protease and apyrase, and Triatoma lipocalin proteins,such as triafestin and pallidipin [6–8,10], appeared to be absent.

Nonetheless, the apparent redundancy of anti-haemostaticfunctions of different proteins observed from the same or

different families could represent an evolutionary adaptationof the insects to ensure that they are successful blood-suckers.Rhodnius saliva displays several proteins of a same familyderived from a high degree of sequence-similarity caused bygene duplication followed by gene modifications, such asmutation, deletion and insertion, which results in a multi-gene protein family. For example, the R. prolixus NP sequenceshave a high degree of similarity among each other, and arerelated to functional diversification with different efficienciesof binding and release of ligands [13,18,33]. Furthermore,PTMs, such as N-glycosylation, might also be a source ofprotein heterogeneity in triatomine saliva [6,39]. Importantgene modifications and PTMs could explain why severalproteins were not identified. This is the first time thattriatomine saliva lipocalin phosphorylation has been sug-gested and is based on three facts. Firstly, several redundant

Page 9

Table 4 – Identification of R. robustus salivary proteins by MS/MS peptide fragmentation using MASCOT.

Spot a Identity Accessionnumber(NCBI) c

Scored

Cov.(%) e

Mass(kDa)

pI Peptide ion (m/z)and sequence h

[Rhodnius prolixus] b Exp.and

Theo. f

Exp.and

Theo. g

R. robustus salivary nitrophorins37 NP-1 precursor gi|3219825 107>35 9 19 23.0 7.0 6.8 1135.5 — EALYHYDPK

NP-4 precursor gi|3219833 22.7 1248.6 — DLGDLYAVLNR38 Complex of mutant L133v of NP-4 with NO gi|114793424 187>39 23 18 20.4 7.1 6.3 1376.7 — LKEALYHYDPK

1533.7 — GNKDLGDVYAVLNR2059.0 — TQDTFYDVSELQVESLGK

R. robustus salivary aggregation inhibitors39 AI-4 precursor gi|33518671 240>36 18 16 18.7 7.5 6.7 1010.5 — IMSGFDANR

1272.7 — LNQLELTSLNK1529.7 — FFTGDWYLTHSR

a Spot number indicated in Fig. 2B.b Results obtained by MASCOT software search in the NCBInr database.c Accession number in the NCBI protein database.d Probability-based Mowse score of MASCOT software that evaluates if the peptides subjected to search are the same as those found in thedatabase originated by in silico digestion of a known protein. The numbers on the right of the symbol “>” represent the minimum scorestatistically significant as calculated by the software (p<0.05).e Coverage is the percentage of predicted protein sequence covered by matched peptides via MASCOT.f Experimental and theoretical molecular masses of proteins, visualized on the gels and calculated from amino acid sequences, respectively.g Experimental and theoretical isoelectric points of proteins, visualized on the gels and calculated from amino acid sequences, respectively.h Peptide sequences identified via MASCOT following the experimental peptide masses after fragmentation.

Fig. 3 – Spectra of masses generated from the processing of the spot NI of R. robustus salivary 2-DE gel. PMF spectrum of thedigested spot annotated ‘NI’ of Fig. 2B (A), spectra of the fragmented SPITC-derivatized peptides of m/z 1240.6 (B) and 1417.6(C) corresponding to the peaks indicated on the PMF spectrum (A). The amino acid residues annotated (B–C) were deduced bymanual de novo sequencing.

1660 J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

Page 10

Table 5 – Biological process categories of the Rhodniusidentified proteins.

Proteinfamily

Totalnumber ofidentifiedspots

Identifiedspots perfamily (%)

Proposedanti-haemostatic

function

Nitrophorin 15 38 Vasodilatation inhibitorand platelet inhibitor,anti-histaminic reactionand coagulation inhibitor

Aggregationinhibitor

15 38 Platelet inhibitor

Triabin-likelipocalinprotein

5 13 Coagulation inhibitor

Amine bindingprotein

3 8 Platelet inhibitorand vasodilatationinhibitor

MYS-1 1 3 Unknown

Table 6 – Putative phosphorylation sites of the proteinsequences corresponding to Rhodnius phosphorylation-dyedspots.

Pro-QDiamond

dyedidentified

spot

Accession number(NCBI)

Total numberof predicted

phosphorylationsites

[R. prolixus] Ser Thr Tyr

3 gi|27968047 7 2 617 gi|8569632 6 3 525 and 26 gi|33518673 4 0 234 and 35 gi|33518671 6 4 410 gi|3212578 4 2 622, 23 and 24 gi|33518679 12 5 329 gi|74841089 6 4 431 gi|1572727 4 3 333 gi|33518711 3 3 436 gi|33518705 3 3 25 and 6 gi|33518667 5 3 87 gi|3319210 6 6 78 and 15 gi|3219833 3 1 68, 13 and 14 gi|3219825 5 2 69 and 12 gi|3219827 6 6 8

1661J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

protein identifications were obtained from spots with thesamemolecular mass and different pI. Secondly, it was shownthat human tear proteins are post-transcriptionally modified[42] and, particularly, a lipocalin is the predominant phos-phoprotein within this fluid [43]. This tear lipocalin appearedseparated into 3 different phosphorylated isoforms by directdetection in 2-DE gel by Pro-Q Diamond dye and fourphosphorylation sites were identified by LC–MS/MS [43].Third, post-translational modifications regulate the function-al activity of proteins [43]. Protein phosphorylation is well-recognized as a key PTM which modulates protein activity bychanging the 3D structure [44]. The biological relevance of thephosphorylation of triatomine proteins remains to be ex-plored. In addition, it will be important to confirm thephosphorylation sites by LC–MS/MS. Moreover, it is importantto note that direct glycostaining detection in 2-DE gels hasrecently shown a glycosylated secreted lipocalin in tears [45],

Fig. 4 – Wide pH range 2-DE maps of saliva phosphoproteins of tr(A) and R. robustus (B) containing 80 μg of saliva protein were pestrips in IEF and 12% SDS-PAGE. The gels were treated with the

whichmay provide assistance in the future characterization ofsaliva proteins.

In conclusion, the present work demonstrates that thesaliva proteomes of the blood-feeding Amazonian insects – R.brethesi and R. robustus – are homologous to R. prolixus salivaryproteome in agreement with their biological similarities andrich in lipocalins. Furthermore, the study presents the firstexperimental evidence of phosphorylated triatomine salivasecreted proteins. These novel data will facilitate futuretriatomine saliva proteomic studies. Future research into thePTMs of triatomine saliva proteins will lead to a betterunderstanding of their functional diversity based on similarprotein structures.

iatomines of the Rhodnius genus. 2-DE gels of both R. brethesirformed under denaturing conditions using 3–10 linear IPGphosphoprotein-specific Pro-Q Diamond dye.

Page 11

1662 J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

Acknowledgements

We thankNunoDomingues for his technical assistance in aminoacid analyses. This work was supported by Brazilian Grants fromFAPDF (PRONEX Program), FINATEC, DPP/UnB, CAPES, CNPq, andFINEP (CT-INFRA program for equipment acquisition).

R E F E R E N C E S

[1] WHO. Chagas disease: control and elimination. ExecutiveBoard; 2008.

[2] de Paula AS, Diotaiuti L, Schofield CJ. Testing the sister-grouprelationshipof theRhodniini andTriatomini (Insecta:Hemiptera:Reduviidae: Triatominae). Mol Phylogenet Evol 2005;35:712–8.

[3] Pereira MH, Gontijo NF, Guarneri AA, Sant'anna MR, DiotaiutiL. Competitive displacement in Triatominae: the Triatomainfestans success. Trends Parasitol 2006;22:516–20.

[4] Massad E. The elimination of Chagas' disease from Brazil.Epidemiol Infect 2008;136:1153–64.

[5] Dias JC, Silveira AC, Schofield CJ. The impact of Chagasdisease control in Latin America: a review. Mem Inst OswaldoCruz 2002;97:603–12.

[6] Charneau S, Junqueira M, Costa CM, Pires DL, Fernandes ES,Bussacos AC, et al. The saliva proteome of the blood-feedinginsect Triatoma infestans is rich in platelet-aggregation inhibitors.Int J Mass Spectrom 2007;268:265–76.

[7] Santos A, Ribeiro JM, Lehane MJ, Gontijo NF, Veloso AB,Sant'Anna MR, et al. The sialotranscriptome of theblood-sucking bug Triatoma brasiliensis (Hemiptera,Triatominae). Insect Biochem Mol Biol 2007;37:702–12.

[8] Assumpção TC, Francischetti IM, Andersen JF, Schwarz A,Santana JM, Ribeiro JM. An insight into the sialome of theblood-sucking bug Triatoma infestans, a vector of Chagas'disease. Insect Biochem Mol Biol 2008;38:213–32.

[9] Schofield CJ, Galvao C. Classification, evolution, and speciesgroups within the Triatominae. Acta Trop 2009;110:88–100.

[10] Kato H, Jochim RC, Gomez EA, Sakoda R, Iwata H, ValenzuelaJG, et al. A repertoire of the dominant transcripts from thesalivary glands of the blood-sucking bug, Triatoma dimidiata,a vector of Chagas disease. Infect Genet Evol 2010;10:184–91.

[11] Assumpcao TC, Charneau S, Santiago PB, Francischetti IM, MengZ, Araujo CN, et al. Insight into the salivary transcriptome andproteomeofDipetalogastermaxima. J ProteomeRes2011;10:669–79.

[12] Abad-Franch F, Monteiro FA. Biogeography and evolution ofAmazonian triatomines (Heteroptera: Reduviidae):implications for Chagas disease surveillance in humidforest ecoregions. Mem Inst Oswaldo Cruz 2007;102(Suppl 1):57–70.

[13] Ribeiro JM, Andersen J, Silva-Neto MA, PhamVM, Garfield MK,Valenzuela JG. Exploring the sialome of the blood-sucking bugRhodnius prolixus. Insect Biochem Mol Biol 2004;34:61–79.

[14] Coura JR, Junqueira AC, Fernandes O, Valente SA, Miles MA.Emerging Chagas disease in Amazonian Brazil. TrendsParasitol 2002;18:171–6.

[15] Monteiro FA, Wesson DM, Dotson EM, Schofield CJ, Beard CB.Phylogeny and molecular taxonomy of the Rhodniini derivedfrom mitochondrial and nuclear DNA sequences. Am J TropMed Hyg 2000;62:460–5.

[16] Monteiro FA, Lazoski C, Noireau F, Sole-Cava AM. Allozymerelationships among ten species of Rhodniini, showingparaphyly of Rhodnius including Psammolestes. Med VetEntomol 2002;16:83–90.

[17] Soares RP, Gontijo NF, Romanha AJ, Diotaiuti L, Pereira MH.Salivary heme proteins distinguish Rhodnius prolixus from

Rhodnius robustus (Hemiptera: Reduviidae: Triatominae). ActaTrop 1998;71:285–91.

[18] Andersen JF, Gudderra NP, Francischetti IM, Ribeiro JM. Therole of salivary lipocalins in blood feeding by Rhodnius prolixus.Arch Insect Biochem Physiol 2005;58:97–105.

[19] Blum H, Beier H, Gross HJ. Improved silver staining ofplant-proteins, RNA and DNA in polyacrylamide gels.Electrophoresis 1987;8:93–9.

[20] Shevchenko A,WilmM, VormO, MannM.Mass spectrometricsequencing of proteins silver-stained polyacrylamide gels.Anal Chem 1996;68:850–8.

[21] Gharahdaghi F, Weinberg CR, Meagher DA, Imai BS, MischeSM. Mass spectrometric identification of proteins fromsilver-stained polyacrylamide gel: a method for the removalof silver ions to enhance sensitivity. Electrophoresis 1999;20:601–5.

[22] Zhang X, Rogowska-Wrzesinska A, Roepstorff P. On-targetsample preparation of 4-sulfophenyl isothiocyanate-derivatizedpeptides using AnchorChip Targets. J Mass Spectrom 2008;43:346–59.

[23] Gobom J, Nordhoff E, Mirgorodskaya E, Ekman R, RoepstorffP. Sample purification and preparation technique based onnano-scale reversed-phase columns for the sensitiveanalysis of complex peptide mixtures by matrix-assistedlaser desorption/ionization mass spectrometry. J MassSpectrom 1999;34:105–16.

[24] ShevchenkoA, Sunyaev S, LobodaA, Bork P, EnsW, StandingKG.Charting the proteomes of organisms with unsequencedgenomes by MALDI-quadrupole time-of-flight massspectrometry and BLAST homology searching. Anal Chem2001;73:1917–26.

[25] Balbuena TS, Silveira V, Junqueira M, Dias LL, Santa-CatarinaC, Shevchenko A, et al. Changes in the 2-DE protein profileduring zygotic embryogenesis in the Brazilian Pine (Araucariaangustifolia). J Proteomics 2009;72:337–52.

[26] Champagne DE, Nussenzveig RH, Ribeiro JM. Purification, partialcharacterization, and cloning of nitric oxide-carrying hemeproteins (nitrophorins) from salivary glands of the blood-suckinginsect Rhodnius prolixus. J Biol Chem 1995;270:8691–5.

[27] Sun J, Yamaguchi M, YudaM,Miura K, Takeya H, Hirai M, et al.Purification, characterization and cDNA cloning of a novelanticoagulant of the intrinsic pathway, (prolixin-S) fromsalivary glands of the blood sucking bug, Rhodnius prolixus.Thromb Haemost 1996;75:573–7.

[28] Sun J, Yuda M, Miura K, Chinzei Y. Characterization andcDNA cloning of a hemoprotein in the salivary glands of theblood-sucking insect, Rhodnius prolixus. Insect Biochem MolBiol 1998;28:191–200.

[29] Andersen JF, Francischetti IM, Valenzuela JG, Schuck P,Ribeiro JM. Inhibition of hemostasis by a high affinity biogenicamine-binding protein from the saliva of a blood-feedinginsect. J Biol Chem 2003;278:4611–7.

[30] Francischetti IM, Ribeiro JM, Champagne D, Andersen J.Purification, cloning, expression, and mechanism of action ofa novel platelet aggregation inhibitor from the salivary glandof the blood-sucking bug, Rhodnius prolixus. J Biol Chem2000;275:12639–50.

[31] Ribeiro JM, Hazzard JM, Nussenzveig RH, Champagne DE,Walker FA. Reversible binding of nitric oxide by a salivaryheme protein from a bloodsucking insect. Science 1993;260:539–41.

[32] Ribeiro JM, Walker FA. High affinity histamine-binding andantihistaminic activity of the salivary nitric oxide-carryingheme protein (nitrophorin) of Rhodnius prolixus. J Exp Med1994;180:2251–7.

[33] Andersen JF, Ding XD, Balfour C, Shokhireva TK, ChampagneDE, Walker FA, et al. Kinetics and equilibria in ligand bindingby nitrophorins 1–4: evidence for stabilization of a nitric

Page 12

1663J O U R N A L O F P R O T E O M I C S 7 4 ( 2 0 1 1 ) 1 6 5 2 – 1 6 6 3

oxide-ferriheme complex through a ligand-inducedconformational trap. Biochemistry 2000;39:10118–31.

[34] Ribeiro JM, Schneider M, Guimaraes JA. Purification andcharacterization of prolixin S (nitrophorin 2), the salivaryanticoagulant of the blood-sucking bug Rhodnius prolixus.Biochem J 1995;308(Pt 1):243–9.

[35] Zhang Y, Ribeiro JM, Guimaraes JA, Walsh PN. Nitrophorin-2:a novel mixed-type reversible specific inhibitor of theintrinsic factor-X activating complex. Biochemistry 1998;37:10681–90.

[36] Francischetti IM, Andersen JF, Ribeiro JM. Biochemicaland functional characterization of recombinantRhodnius prolixus platelet aggregation inhibitor 1 asa novel lipocalin with high affinity for adenosinediphosphate and other adenine nucleotides.Biochemistry 2002;41:3810–8.

[37] Noeske-Jungblut C, Haendler B, Donner P, Alagon A, Possani L,Schleuning WD. Triabin, a highly potent exosite inhibitor ofthrombin. J Biol Chem 1995;270:28629–34.

[38] Fuentes-Prior P, Noeske-Jungblut C, Donner P, SchleuningWD, Huber R, Bode W. Structure of the thrombin complexwith triabin, a lipocalin-like exosite-binding inhibitor derivedfrom a triatomine bug. Proc Natl Acad Sci USA 1997;94:11845–50.

[39] Faudry E, Lozzi SP, Santana JM, D'Souza-Ault M, Kieffer S,Felix CR, et al. Triatoma infestans apyrases belong to the5′-nucleotidase family. J Biol Chem 2004;279:19607–13.

[40] Larsen MR, Thingholm TE, Jensen ON, Roepstorff P, JorgensenTJ. Highly selective enrichment of phosphorylated peptidesfrom peptide mixtures using titanium dioxide microcolumns.Mol Cell Proteomics 2005;4:873–86.

[41] BlomN,Gammeltoft S, BrunakS.Sequenceandstructure-basedprediction of eukaryotic protein phosphorylation sites. J MolBiol 1999;294:1351–62.

[42] Li N, Wang N, Zheng J, Liu XM, Lever OW, Erickson PM, et al.Characterization of human tear proteome using multipleproteomic analysis techniques. J Proteome Res 2005;4:2052–61.

[43] Zhao Z, Liu J,Wasinger VC,Malouf T, Nguyen-Khuong T,WalshB, et al. Tear lipocalin is the predominant phosphoprotein inhuman tear fluid. Exp Eye Res 2010;90:344–9.

[44] Mann M, Ong SE, Gronborg M, Steen H, Jensen ON, Pandey A.Analysis of protein phosphorylation using mass spectrometry:deciphering the phosphoproteome. Trends Biotechnol 2002;20:261–8.

[45] You J, Fitzgerald A, Cozzi PJ, Zhao Z, Graham P, Russell PJ, et al.Post-translation modification of proteins in tears.Electrophoresis 2010;31:1853–61.