Eur. J. Biochem. 262, 745±754 (1999) q FEBS 1999

Purification, structural characterization, cloning and immunocytochemicallocalization of chemoreception proteins from Schistocerca gregaria

Sergio Angeli1,2, Francesca Ceron2, Andrea Scaloni3, Maria Monti3, Gaia Monteforti1, Antonio Minnocci1,

Ruggero Petacchi1 and Paolo Pelosi2

1Scuola Superiore di Studi Universitari e di Perfezionamento `S. Anna', Pisa, Italy; 2Dipartimento di Chimica e Biotecnologie Agrarie,

University of Pisa, Pisa, Italy; 3Centro Internazionale Servizi di Spettrometria di Massa IABBAM, National Research Council, Napoli, Italy

Soluble low-molecular-mass protein isoforms were purified from chemosensory organs (antennae, tarsi and labrum)

of the desert locust Schistocerca gregaria. Five genes encoding proteins of this group were amplified by PCR from

cDNAs of tarsi and sequenced. Their expression products are polypeptide chains of 109 amino acids

showing 40±50% sequence identity with putative olfactory proteins from Drosophila melanogaster and Cactoblastis

cactorum. Direct structural investigation on isoforms purified from chemosensory organs revealed the presence in

the expression products of two of the genes cloned. Two additional protein isoforms were detected and their

molecular structure exhaustively characterized. MS analysis of all isoforms demonstrated that the four cysteine

residues conserved in the polypeptide chain were involved in disulfide bridges (Cys29±Cys38 and Cys57±Cys60)

and indicated the absence of any additional post-translational modifications. Immunocytochemistry experiments,

performed with rabbit antiserum raised against the protein isoform mixture, showed selective labelling of the outer

lymph in contact sensilla of tarsi, maxillary palps and antennae. Other types of sensilla were not labelled, nor were

the cuticle and dendrites of the sensory cells. No binding of radioactively labelled glucose or bicarbonate was

detected, in disagreement with the hypothesis that this class of proteins is involved in the CO2-sensing cascade. Our

experimental data suggest that the proteins described here could be involved in contact chemoreception in

Orthoptera.

Keywords: chemosensory proteins; contact sensilla; disulfide bridges; Schistocerca gregaria; sequence analysis.

Locusts and grasshoppers are major pests in agriculture. Theyhave a solitary and a gregarious phase, characterized by differentbehaviour and morphological features [1]. All crop damage iscaused by individuals in the gregarious phase.

In the species Schistocerca gregaria, the shift from thesolitary to the gregarious phase is preceded by an associatingphase, triggered by volatile aromatic compounds, such asguaiacol, veratrol and phenylacetonitrile [2]. In the gregariousphase, the same chemicals, and perhaps related structures,induce aggregation of great numbers of individuals. Under theseconditions this species becomes a plague and can destroy entirecrops. Therefore, it is evident that these insects rely on chemicalcommunication and that their populations could be controlled bythe use of the appropriate chemical stimuli.

Despite such pressing objectives, biochemical study of theolfactory system had been limited until recently to Lepidopteranspecies with large antennae. In the last few years, however,molecular biology techniques have made such research feasiblein small insects of wider interest, such as Drosophilamelanogaster. Membrane-bound olfactory receptor proteinshave not yet been detected in insects, but a number of putativeodorant-binding proteins (OBPs) were first characterized in

several Lepidopteran species [3±15]. More recently, suchinvestigations have been applied to representative species ofother orders of insects, such as Diptera [16±18], Hymenoptera[19], Hemiptera [20], Coleoptera [21], Blattoidea [22] andPhasmida [23±24]. A low degree of conservation duringevolution has prevented a clear classification of these proteinsexpressed in phylogenetically distant species, and identificationof subclasses is not yet as clear as in the case of Lepidoptera[25±29], where OBPs of different subclasses have been found tobe expressed in morphologically and functionally different typesof sensilla [30±33]. In other orders of insects, the fragmentarydata currently available do not yet allow such correlations.Moreover, chemosensory sensilla are present all over the bodyand have different chemoreception functions, such as olfaction,taste and perception of CO2.

In this paper we report the isolation of soluble low-molecular-mass proteins in antennae, tarsi and labrum of S. gregaria, theircomplete structural characterization by combined Edmandegradation/MS procedures and the cloning of five closelyrelated genes encoding these proteins in the tarsi. Polyclonalantibodies were used to immunolocalize these proteins incontact sensilla of tarsi, antennae and labrum.

MATERIALS AND METHODS

Animals

Gregarious phase S. gregaria (ForskaÊl) were bred undercrowded conditions in wooden cages (50 � 50 � 80 cm) in aventilated temperature-controlled room (30 ^ 1 8C day;

Correspondence to P. Pelosi, Dipartimento di Chimica e Biotecnologie

Agrarie, Via S. Michele 4, 56124 Pisa, Italy. Fax: + 39 050 574235,

Tel.: + 39 050 571564, E-mail: ppelosi@agr.unipi.it

Abbreviations: OBP, odorant-binding protein; CSP, chemosensory protein;

ESMS, electrospray MS; MALDIMS, matrix-assisted laser-desorption-

induced MS.

(Received 22 February 1999; accepted 22 March 1999)

746 S. Angeli et al. (Eur. J. Biochem. 262) q FEBS 1999

27 ^ 1 8C night), maintained on a 12 : 12 light : darknesscycle. The insects were fed with fresh wheat shoots and barleyflake.

Preparation of the extracts

Tissues were dissected from individuals killed with CO2 andused immediately or stored at 220 8C for few days. Crudeextracts were prepared by homogenization in 20 mm Tris/HClbuffer, pH 7.4 (Tris buffer), using a Polytron apparatus,followed by centrifugation at 10 000 g for 20 min. The clearsupernatant was used for electrophoretic analysis and chromato-graphic fractionation.

Purification of the proteins

In a typical experiment, the extract from 400 tarsi wasconcentrated to 0.5 mL and separated by gel filtration on aSuperose 12 column (1 � 30 cm) eluted with 50 mm ammo-nium bicarbonate, 1.0 mL fractions being collected. AfterSDS/PAGE (12% gels), tubes containing proteins of interestwere pooled and subjected to anion-exchange chromatographythrough a Mono Q HR 5/5 column. A linear gradient from 0 to0.5 m NaCl in 20 mm Tris/HCl buffer, pH 7.4, was used forelution. This procedure yielded a mixture of 14-kDa proteinisoforms, which was used for antibody production or furtherpurified by reversed-phase HPLC before structural investi-gations. Individual components of the family were isolated in ahomogeneous form by a chromatographic step on a Vydac C4

214TP54 column (250 � 4.6 mm; 5 mm; 30 nm pore size; TheSeparation Group, Hesparia, CA, USA). Proteins were dissolvedin 0.1% trifluoroacetic acid, loaded on to the column and elutedwith a linear gradient of from 5 to 70% acetonitrile in 0.1%trifluoroacetic acid over 30 min, at a flow rate of 1 mL´min21.

Gel electrophoresis

Electrophoresis in denaturing conditions (SDS/PAGE) wasperformed on 13% polyacrylamide gel, using a Bio-Rad Mini-Protean II apparatus, by the procedure of Laemmli [34]. NativePAGE was carried out under the same conditions, but in theabsence of SDS.

Small amounts of protein for N-terminal sequencing werepurified by continuous elution gel electrophoresis in nativeconditions, using the Mini Prep Cell apparatus (Bio-Rad) and10% cylindrical gels (0.7 cm internal diameter and 5 cmlength).

Reduction and alkylation of protein isoforms, enzymatichydrolysis and peptide separation

Before sequence analysis, protein samples were reduced andalkylated under the conditions described previously [35].Proteins were desalted on a Vydac C4 column as describedabove.

Native and carboxamidomethylated proteins were digestedwith endoprotease LysC and/or endoprotease AspN (Boehringer-Mannheim) in 0.4% ammonium bicarbonate, pH 6.5, at 37 8Cfor 18 h, using in both cases an enzyme/substrate ratio of 1 : 50(w/w).

Peptide mixtures were fractionated by reversed-phase HPLCon a Vydac C18 214TP52 column (250 � 2.1 mm; 5 mm, 30 nmpore size; The Separation Group), using a linear gradient of from5 to 60% acetonitrile in 0.1% trifluoroacetic acid over 60 min, atflow rate of 0.2 mL´min21. Individual components were

collected manually and dried down in a Speed-vac centrifuge(Savant).

MS analyses

Intact proteins or individual peptide fractions were submitted toelectrospray MS (ESMS) analysis, using a BIO-Q triplequadrupole mass spectrometer (Micromass, Manchester, UK).Samples were dissolved in 1% (v/v) acetic acid/50% (v/v)aqueous acetonitrile, and 2±10 mL injected into the massspectrometer at a flow rate of 10 mL´min21. The quadrupolewas scanned from m/z 500±1800 at 10 s´scan21 and the spectrawere acquired and elaborated using MASS-LYNX software.Calibration was performed using the multiply charged ions froma separate injection of myoglobin (molecular mass16 951.5 Da). All mass values are reported as average masses.

Matrix-assisted laser-desorption-ionization (MALDI) massspectra were recorded using a Voyager DE MALDI-TOF massspectrometer (Perkin-Elmer Perseptive Biosystem); a mixtureof analyte solution, a-cyano-4-hydroxycinnamic acid andbovine insulin was applied to the sample plate and dried invacuo. Mass calibration was performed using the molecular ionsfrom the bovine insulin at 5734.54 Da and the matrix at379.06 Da as internal standards. Raw data were analysed usingthe manufacturer's computer software and are reported asaverage masses. Assignments of the recorded mass valuesto individual peptides were performed on the basis of theirmolecular mass and confirmed by submitting the entirepeptide mixture to a single step of manual Edman degra-dation and re-running the MALDI spectrum of thetruncated peptide mixture as described previously [36].

Amino acid sequence analysis

Amino acid sequence was determined by direct Edmandegradation using a Perkin-Elmer-Applied Biosystems 477Apulsed-liquid protein sequencer equipped with a Perkin-Elmer-Applied Biosystems 120A HPLC apparatus for phenylthio-hydantoin-amino acid identification. Electroblotted sampleswere directly analysed from poly(vinylidene difluoride)membranes [37].

cDNA synthesis

Total RNA was extracted from the tarsi of a single mature maleor female adult S. gregaria, in its gregarious phase, using theTrizolTM reagent kit (Gibco±BRL), a modified guanidineisothiocyanate/phenol procedure. mRNA was affinity-purifiedusing the PolyA Tract mRNA isolation system (Promega).

The mRNA thus prepared was subjected to reverse tran-scription, using 200 U of Moloney murine leukaemia virusreverse transcriptase (Gibco BRL) and 0.5 mg of oligo(dT)12±18

(Sigma) in a 50-mL total volume. The mixture also contained0.5 mm each dNTP (Pharmacia Biotech, Uppsala, Sweden),75 mm KCl, 3 mm MgCl2, 10 mm dithiothreitol and0.1 mg´mL21 BSA in 50 mm Tris/HCl, pH 8.3. The reactionmixture was incubated at 37 8C for 60 min and the productswere directly used for PCR amplification or stored at 220 8C.

PCR

Aliquots of 1 mL of crude cDNA were amplified in a Bio-RadGene CyclerTM thermocycler, using 2.5 U of Thermus aquaticusDNA polymerase (Sigma), 1 mm each dNTP, 1 mm each PCR

q FEBS 1999 Chemoreception proteins in Schistocerca gregaria (Eur. J. Biochem. 262) 747

primer, 50 mm KCl, 2.5 mm MgCl2 and 0.1 mg´mL21 BSA in10 mm Tris/HCl, pH 8.3, containing 0.1% Triton X-100.

The 64-mer degenerate primer AGCTAAGCTTGARGAR-AARTAYACNAC, used at the 5 0 end encodes the first six aminoacids EEKYTT- of the mature protein (underlined) and ispreceded by a HindIII restriction site (in italics). The 3 0-endprimer was an oligo(dT)15, preceded by a BamHI restriction site(GGATCC). After a denaturing step at 95 8C for 5 min, thereaction was performed for 35 cycles (95 8C for 1 min, 48 8Cfor 1 min, 72 8C for 1 min), followed by a final step of 7 min at72 8C.

Cloning and sequencing

PCR products were analysed by electrophoresis on a 2%agarose gel and purified using microcolumns of theQIAquick PCR purification kit (Qiagen). The product wasthen treated with the restriction enzymes HindIII andBamHI (Boehringer-Mannheim), by incubating 1 mg of DNAwith 2 U of each enzyme in the appropriate buffer (total volume20 mL) for 1 h at 37 8C, and then purified again with the samekit.

The DNA so prepared was ligated into a pCRII vector(Invitrogen) that had been linearised with the restrictionenzymes HindIII and BamHI and dephosphorylated. Theplasmid (60 ng) and the insert were ligated by incubating in a1 : 5 molar ratio overnight at room temperature. The ligatedproduct was used to transform Epicurian Coli XL-1 Blue MRFSupercompetent Cells (Stratagene, La Jolla, CA, USA). Afterincubation in SOC medium for 1 h at 37 8C, the cells wereplated on Luria±Bertani broth/agar, containing ampicillin,isopropyl b-d-thiogalactoside and 5-bromo-4-chloroindol-3-ylb-d-galactoside. White colonies were then grown in liquidLuria±Bertani broth/ampicillin and analysed for the presence ofthe right insert by PCR, using the plasmid primers SP6 and T7,followed by 2% agarose gel electrophoresis.

Nucleotide sequences of both strands of the cDNA cloneswere determined from double-stranded plasmid DNA using thePerkin-Elmer-Applied Biosystems Dye Deoxy Terminator CycleSequencer Kit and a Perkin-Elmer-Applied Biosystems 310automated sequencer at the ENEA laboratory (Casaccia, Roma,Italy).

Preparation of antisera

Antisera were obtained by injecting an adult rabbit sub-cutaneously and intramuscularly with 150 mg of protein purifiedfrom the tarsi of adult females, followed by two additionalinjections of 100 mg after 18 and 30 days. The protein wasemulsified with an equal volume of Freund's complete adjuvantfor the first injection and incomplete adjuvant for furtherinjections. Animals were bled 10 days after the last injectionand the serum was used without further purification. Rabbitswere individually housed in large cages, at constant temperature,and all operations were performed according to ethical guide-lines in order to minimize pain and discomfort to animals.

Western-blot analysis

After electrophoretic separation under denaturing conditions,proteins were electroblotted on a nitrocellulose membrane, bythe procedure of Kyhse-Andersen [38]. After treatment with0.2% gelatin from porcine skin (Sigma)/0.05% Tween 20 inNaCl/Pi for 2 h, the membrane was incubated with the crudeantiserum against the protein at a dilution of 1 : 500 and then

with goat anti-(rabbit IgG)±horseradish peroxidase conjugate(dilution 1 : 1000). Immunoreacting bands were detected bytreatment with 4-chloro-1-naphthol.

Electron microscopy immunocytochemistry

Antennae, second tarsal segments of the fore legs and maxillarypalp tips were severed with a razor blade under a drop of fixative(4% paraformaldehyde and 2% glutaraldehyde in 0.1 mmcacodylate, pH 7.2). The samples were rinsed three times incacodylate buffer, dehydrated in an ethanol series and embeddedin LR White resin, by soaking in 2 : 1, 1 : 1 and 1 : 2 ethanol/LR White and leaving them in pure LR White at 4 8C overnight.The resin was polymerized at 60 8C in tightly closed gelatincapsules in an aluminium block. Ultrathin sections were cut witha diamond knife on an LKB 8800 Ultratome ultramicrotome andmounted on Formvar-coated single-hole grids.

For immunocytochemistry, the sections were floated on30 mL droplets of the following solutions: 2 � 10 min onNaCl/Pi containing 50 mm glycine; 2 � 10 min on PBG(NaCl/Pi containing 0.2% gelatin and 0.5% BSA); overnight onthe primary antiserum diluted with PBG (or in the controlpreimmune serum); 6 � 5 min on PBG; 90 min on the 1 : 20secondary antibody in PBG; 2 � 5 min on PBG; 2 � 5 min onNaCl/Pi/glycine; 2 � 5 min on NaCl/Pi. Silver amplificationwas performed on grids for 10 min, as described by Danscher[39]. The grids were then rinsed six times for 5 min in water,stained in aqueous 2% uranyl acetate for 10 min and finallyrinsed in water. Sections were blotted dry and observed in aJEOL 100 SX electron microscope at 80 kV.

The primary antiserum was used in dilutions of 1 : 30 000 fortarsi (pulvilli) and maxillary palps and 1 : 15 000 for antennae.The secondary antibody was anti-(rabbit IgG), coupled to 10-nmcolloidal gold (BioCell, Cardiff, UK). As a control the primaryantiserum was replaced by preimmune serum at the samedilution. The experiments were repeated on five females and fivemales.

RESULTS

Identification and purification of putative chemosensoryproteins (CSPs)

In a search for putative soluble proteins involved in chemo-reception, we screened extracts of the different parts of theS. gregaria body, focusing our attention on proteins selectivelyor preferentially expressed in sensory organs that were bothsmall (around 15 kDa) and acidic, as easily measured bydenaturing and native PAGE, respectively.

Figure 1A shows the electrophoretic patterns in denaturingconditions (SDS/PAGE) of crude extracts obtained from thedifferent parts of the body of female adult S. gregaria in theirgregarious phase. In parallel experiments male extracts wereanalysed with the same techniques, but no significant differ-ences were detected between sexes. Several bands of lowmolecular mass were found specifically in the extracts of tarsi,antennae and labrum. Electrophoretic analysis in native condi-tions, performed on the same extracts (not shown), indicated thepresence of fast-migrating bands only in the three sensoryorgans.

The proteins corresponding to the most intense bands werepurified by continuous elution electrophoresis under nativeconditions, further analysed by SDS/PAGE and electroblotted.N-terminal amino acid sequences of the excised bands wereobtained by automated Edman degradation.

748 S. Angeli et al. (Eur. J. Biochem. 262) q FEBS 1999

Three antennal proteins and one from the tarsi yielded thesame 32-residue N-terminal sequence, EEKYTTKYDNVNL-DEILANDRLLNKYVQXLLE-. Of the two proteins isolatedfrom the labrum, the first was a truncated form, lacking the firsteight amino acid residues at the N-terminus; the second differedfrom those of antennae and tarsi by two amino acid substitutions(4:Y/F; 6:T/D). These sequences exhibit significant homologywith a subclass of putative chemoreception proteins, referred toas OS-D-like from the first member discovered in the antennaeof D. melanogaster [16,17]. This family includes membersexpressed in species belonging to most orders of insects, fromPhasmids to Lepidoptera.

Furthermore, the proteins corresponding to the 14-kDa bandin the tarsi sample (Fig. 1A) were purified from the tarsi offemales by gel filtration on Superose-12 (Fig. 1B), followed byanion-exchange chromatography on Mono Q (not shown). Thepurified proteins migrated on a native gel as a single broad band.This method, applied to large quantities of extract, afforded asample of protein isoforms (see below), which were used for

extensive chemical characterization, polyclonal antibody prepa-ration and binding assays.

Gel filtration also indicated that these proteins under nativeconditions have a molecular mass of around 16 kDa. Thisestimate, compared with that measured by SDS/PAGE (14 kDa)and with the more accurate values obtained by ESMS orcalculated from their cDNA-derived sequences (see below),suggested their occurrence in a monomeric form, at least underthe conditions used. An isoelectric point of 5.1 was measured byisoelectric focusing in a 3.5±9.5 Ampholine gradient (data notshown).

Cloning and cDNA sequencing

Partial sequence information allowed the synthesis of adegenerate primer which, together with oligo(dT), was used toamplify by PCR the nucleotide sequences encoding the isolatedproteins. RNA was obtained from tarsi homogenate separatelyextracted from a single male or female; mRNA was purified andthe respective cDNA was synthesized from this sample, asdescribed in Materials and Methods. Amplification by PCRafforded a product of about 400±450 bp which was ligated intoa pCRII vector. After transformation of Escherichia colisupercompetent cells and plating, clones containing the insert,as analysed by PCR using the plasmid's primers SP6 and T7,were used to determine nucleotide sequences on the purifiedplasmids.

We thus obtained five different sequences, all encodingproteins of 109 amino acids. Given the high expression of theseproteins in chemosensory organs and as their specific function isstill unknown, we propose to give them the general namechemosensory protein (CSP). Nucleotide sequences have beendeposited in the GenBank Sequence Database with the accessionnumbers AF070961 to AF070965. Figure 2 reports the derivedamino acid sequences for the five clones (CSP-sg1 to CSP-sg5),together with two additional sequences (CSP-sg6 and CSP-sg7)determined by a combination of MS and Edman degradation, asdescribed below. Of the five derived amino acid sequences, thetwo obtained from the tarsi of females matched the first 32amino acids of the sequence directly determined by Edmandegradation of the protein and differed from each other by onlytwo residues (98% identity). The three sequences from the malesample were very similar (84±97% identity) and also resembledthose of the female (88±94% identity).

Structural characterization of the purified isoforms

The protein sample purified from the tarsi of female S. gregariawas directly subjected to analytical reversed-phase HPLCshowing the presence of four main components as shown inFig. 3. Each peak was collected manually and analysed byESMS. This analysis demonstrated the presence in each fractionof a single major polypeptide. The accurate molecular mass wasdetermined; minor isoforms accounting for less than 5% of thetotal protein were also detected (data not shown). Table 1summarizes the ESMS data obtained for the purified CSPisoforms, together with the theoretical masses calculated for thefive cDNA-derived sequences. Two of the experimental values,relative to components 2 and 3 (Fig. 3), were exactly fourmass units lower than those calculated for CSP-sg5 andCSP-sg4, respectively; the other values did not match anyof those calculated for the other sequences reported inFig. 2. These findings suggest that the four cysteineresidues present in the primary structure could be involvedin two intramolecular disulfide bridges. Determination of

Fig. 1. Electrophoretic patterns under denaturing conditions of crude

extracts obtained from different parts of the body of female S. gregaria

adults in their gregarious phase. (A) H, head; Th, thorax; Ab, abdomen; W,

wings; T, tarsi; A, antennae; L, labrum. A parallel analysis, performed on the

extracts from male individuals, did not show any significant differences.

Intense bands of apparent molecular mass <15 kDa, present in the extracts

of antennae, tarsi and labrum, suggested the presence of putative

chemoreception proteins in these organs. The three most intense bands of

low molecular mass in the antennal samples and two in the labrum sample

were purified by preparative electrophoresis and used for N-terminal

sequencing. The 14-kDa band of the tarsi was purified in milligram amounts

by gel filtration on Superose-12 (B) SDS/PAGE of fractions containing the

enriched protein, followed by anion-exchange chromatography on Mono Q

(not shown). This sample was utilized for structural characterization, binding

assays and preparation of polyclonal antobodies. Molecular-mass markers

(M) are, from the top: BSA (66 kDa), ovalbumin (45 kDa), carbonic

anhydrase (29 kDa), trypsin inhibitor (20 kDa) and a-lactalbumin (14 kDa).

q FEBS 1999 Chemoreception proteins in Schistocerca gregaria (Eur. J. Biochem. 262) 749

the accurate molecular masses of the native polypeptides,before and after alkylation with iodoacetamide, confirmedthe presence of two disulfide bridges in each isoform(results not shown).

To verify that components 2 and 3 correspond to CSP-sg5 andCSP-sg4, respectively, and to determine the complete sequenceof components 1 and 4, each fraction was digested withendoprotease AspN and the relative peptide mixtures wereanalysed by MALDIMS. The signals present in the peptide mapsobtained for components 2 and 3 allowed us to cover the entireamino acid sequence, thus demonstrating perfect identity oftheir primary structures with those reported for CSP-sg5 and

CSP-sg4, respectively (Table 2) and excluding, at the same time,the presence of additional post-translational modifications.Similarly the spectra of the remaining isoforms showed somemass values that did not match any of the calculated ones. Thecorresponding digests were then separated by reversed-phaseHPLC and the whole fractions analysed by MALDIMS. Thepeptides exhibiting MH+ at m/z 1251.3 (component 4) and m/z837.8, 1026.8, 1084.6, 1601.5, 1650.7, 3125.8, 3504.7,3528.5 (component 1) were subjected to automated Edmandegradation, leading to the amino acid sequence of CSP-sg6 andCSP-sg7, respectively, as reported in Fig. 2. Thus their proteinmolecular masses, calculated taking into account the amino acidsubstitutions observed, were 12 705.21 Da (isoform 6) and12 542.12 Da (isoform 7), respectively. These values agree withthose obtained by ESMS analysis for the intact components,12 702.01 ^ 0.94 Da and 12 539.42 ^ 0.67 Da, respectively,considering the presence of two disulfide bridges/molecule. Thetheoretical isoelectric points for these two proteins are 5.22 and5.88, respectively.

The arrangement of the two intramolecular S±S bonds wasestablished by analysis of the four native protein isoforms afterdigestion with endoprotease LysC. The reactions were per-formed at pH 6.5 to avoid scrambling of the disulfide bridges.The resulting peptide mixtures were then separated by reversed-phase HPLC and the whole fractions analysed by MALDIMS.All peptides were interpreted on the basis of the amino acidsequences reported in Fig. 2 and protease specificity. In the caseof the CSP-sg4 digest, two significant peaks showed thepresence of components with MH+ at m/z 1819.6, 1837.6 and2001.6 as reported in Fig. 4, which were associated with peptidefragments containing Cys residues. In fact, on the basis of their

Fig. 3. HPLC purification of CSP isoforms from the tarsi of S. gregaria.

The crude extract from females organs was purified by gel filtration and

anion-exchange chromatography and then fractionated on a reversed-phase

chromatography column as described in Materials and methods.

Fig. 2. Amino acid sequences of the seven CSP

isoforms obtained from the tarsi of S. gregaria,

compared with those of the same class reported

in other insect species. The first three sequences

were derived from cDNA isolated from males

(CSP-sg1, 2 and 3); the following two were

derived from cDNA obtained from females

(CSP-sg4 and 5) and confirmed by protein

chemistry procedures on a purified protein

sample; the remaining two (CSP-sg6 and 7) were

determined by MS/Edman degradation. CSP-ec1

is expressed in the antennae of E. calcarata, OS-D

in the antennae of D. melanogaster, CLP-1 in the

labial palpi of C. cactorum, EjB in the ejaculatory

bulb of D. melanogaster and p10 in the

regenerating legs of P. americana. Shaded

residues indicate differences between the seven

sequences of S. gregaria. Residues conserved in

most sequences are marked by a + sign.

Conserved cysteines are marked by an asterisk.

750 S. Angeli et al. (Eur. J. Biochem. 262) q FEBS 1999

molecular masses alone, these components were associated withpeptides (47±63), (47±59) + (60±63) and (26±43), demonstrat-ing the disulfide linkages Cys57±Cys60 and Cys29±Cys38.These findings were confirmed by reduction of an aliquot of

these components with dithiothreitol followed by re-examina-tion by MALDIMS, showing the presence of the reducedfragments. Furthermore, the remaining peptide was sub-jected to automated Edman degradation leading to theamino acid sequences expected; no phenylthiohydantoin-Cyswas observed at the cycles corresponding to Cys29 andCys57, while phenylthiohydantoin-cystine was detected atthe cycles corresponding to Cys38 and Cys60 [40]. Similarresults were obtained for the other three CSP isoforms.Table 3 summarizes the signals associated with the disulfide-containing peptides observed in the CSP-sg5, CSP-sg6 andCSP-sg7 digests.

Table 2. MALDIMS analysis of the reduced and carboxamidomethy-

lated (CAM) CSP isoforms purified from S. gregaria after digestion with

endoprotease AspN. CPS-sg4 to CPS-sg7 were eluted in HPLC fractions 3,

2, 4 and 1, respectively.

Observed MH+

CSP-sg4 CSP-sg 5 CSP-sg6 CSP-sg7 Peptide Modification

1617�.3 1601�.5 (1�±13)

2272�.9 2272�.9 (1�±19)

3213�.7 3125�.8 (14�±40) 2 CAM

1663�.9 1664�.9 1663�.9 1650�.7 (20�±32) CAM

1778�.7 1778�.7 (20�±33) CAM

2557�.9 2469�.8 (20�±40) 2 CAM

911�.6 911�.8 911�.6 837�.8 (33�±40) CAM

1085�.7 1142�.7 1085�.7 1084�.6 (41�±50)

3519�.4 3520�.2 3519�.4 3504�.7 (51�±80) 2 CAM

1221�.2 1251�.3 (81�±90)

3542�.8 3528�.5 (81�±109)

1054�.0 1054�.9 1054�.0 1026�.8 (102�±109)

Table 3. MALDIMS analysis of CSP proteolytic peptides containing disulfide bridges. The theoretical MH� values expected for the peptides in reduced

form are also indicated. CPS-sg4 to CPS-sg7 were eluted in HPLC fractions 3, 2, 4, 1, respectively.

Observed MH� Theor: MH�

CSP-sg4 CSP-sg5 CSP-sg6 CSP-sg7 reduced peptides Peptide S±S

2001�.6 2001�.9 2001�.6

1537�.3

2004�.1

1539.7

(26±43)

(29±43)

C29-C38

C29-C38

1819�.6 1819�.6 1822�.1 (47±63) C57-C60

1748�.7 1677�.0 1750�.9, 1678.9 (48±63) C57-C60

1837�.6 1837�.6 1347�.5, 493.5 (47±59) + (60±63) C57-C60

1766�.5 1694�.4 1276�.4, 493.5,

1218.4, 479.5

(48±59) + (60±63) C57-C60

Fig. 4. MALDI mass spectra of disulfide bridge-containing peptides

collected from narrow-bore HPLC analysis of the endoprotease LysC

digest of CSP-sg4 and eluted after 38 min (A) and 36 min (B). Signals in

the spectra correspond to the indicated disulfide-bridged peptides; the two

cysteine residues involved in the S±S bond are also reported.

Table 1. Measured and theoretical molecular-mass values and calcu-

lated isoelectric points of CSPs from S. gregaria. Samples used for ESMS

analysis were purified from tarsi of females as described in Materials and

methods. The isoelectric point measured with a sample purified from the

tarsi of both sexes was 5.1. Proteins CSP-sg4 to CSP-sg7 were eluted in

HPLC fractions 3, 2, 4 and 1, respectively.

Molecular mass (Da)

Protein Sex Experimental Theoretical Theoretical pI

CSP-sg1 M ± 12 587�.15 5�.52

CSP-sg2 M ± 12 546�.12 5�.22

CSP-sg3 M ± 12 704�.18 5�.22

CSP-sg4 F 12 671.68 �^ 1.57 12 675�.18 5�.22

CSP-sg5 F 12 728.37 �^ 0.61 12 732�.23 5�.23

CSP-sg6 F 12 702.01 �^ 0.94 ± ±

CSP-sg7 F 12 539.42 �^ 0.67 ± ±

q FEBS 1999 Chemoreception proteins in Schistocerca gregaria (Eur. J. Biochem. 262) 751

Preparation of antibodies and Western-blot analysis

To investigate the site of production of these proteins, polyclonalantibodies were raised against the sample purified from the tarsiof adults. The crude antiserum was employed in Western-blotanalysis of extracts from antennae, tarsi and palpi. In all thesamples, a strong cross-reactivity was associated only with the14-kDa protein. Other abundant proteins present in the extractswere not labelled (results not shown).

The tarsi extracts were also tested with the antiserum raisedagainst another protein of the same class (CSP-ec1; GenBankaccession number AF139196) purified from the tarsi ofEurycantha calcarata. Although the two proteins share nearly50% amino acid sequence, no cross-reactivity was observedduring these experiments (Fig. 5). A negative reaction was alsoobtained when the extracts of E. calcarata were treated withanti-(S. gregaria protein) serum.

Electron microscopy immunocytochemistry

The crude antiserum against the proteins isolated fromS. gregaria was used for immunocytochemical analysis ofsecond tarsal segments (as defined by Kendall [41]), maxillarypalp tips and antennae.

In sections of tarsal segments from both males and females,significant gold particle labelling was found exclusively in theundersurface, an area dense with contact chemoreception pegs(Fig. 6). Similarly within the sensilla, the lymph of the outerlumen and the sensillar sinus were heavily labelled, while thecuticle and the sensory dendrites were not.

On the tips of maxillary palps, labelling was present in theterminal sensilla (as defined by Blaney & Chapman [42]), againonly in the outer lumen and in sensillar sinus below the pegbase.

Similar results were obtained with antennae, where onlychaetic sensilla [43] were labelled in the same areas. In no casedid olfactory or coeloconic sensilla show any labelling (data notshown). No differences between sexes were observed.

DISCUSSION

In this work we have identified in Orthoptera soluble proteins ofthe chemosensory sensilla likely to be involved in chemicalcommunication. Among the several putative chemosensorycomponents visible in crude extracts of sensory organs ofS. gregaria, we have focused our attention on the most abundantamong the low-molecular-mass components. Structural analysisrevealed significant similarity with D. melanogaster OS-D[16,17] and proteins expressed in other orders of insects[14,19,22±24].

The primary structure of seven CSP isoforms fromS. gregaria was determined by a combined molecularbiology/protein chemistry approach. ESMS performed onsamples purified from tarsi of females was critical to verifythe sequences determined by molecular cloning, demonstratingat the same time the presence of a larger number of isoforms,possibly the result of a wider microdiversity and/or individualdifferences. Peptide mapping experiments on the nativecomponents agreed completely with the cDNA-derivedsequences and allowed the presence of two disulfide bonds inthe molecules (Cys57±Cys60 and Cys29±Cys38) to bedetermined.

The CSP sequences reported in this work, expressed in thetarsi of S. gregaria, are very similar to each other, with thepercentage of identical amino acids around 90% or higher.

Fig. 5. Partially purified sample of CSPs from the tarsi of S. gregaria

(lane 1) and crude tarsi extract of E. calcarata (lane 2) analysed by

SDS/PAGE (A) and Western blotting (B), using rabbit polyclonal

antiserum against the protein of S. gregaria. Although the proteins share

more than 50% of their residues between the two species, no immunocross-

reactivity is observed. Molecular-mass markers (M) are, from the top: BSA

(66 kDa), ovalbumin (45 kDa), carbonic anhydrase (29 kDa), trypsin

inhibitor (20 kDa) and a-lactalbumin (14 kDa).

Fig. 6. Immunocytochemical localization of CSPs in S. gregaria contact

chemoreceptor sensilla. (A) Female antenna. Longitudinal section through

the base of a sensillum cheticum labelled with CSP polyclonal antiserum

(1 : 15 000). The outer lumen lymph in the peg lumen and in the cavity

below the peg base is strongy labelled. OL, outer lumen lymph; IL, inner

lumen; C, cuticle. (B) Male and (C) female maxillary palp tip terminal

sensillum. Both longitudinal (1 : 30 000) and cross-section below the peg

base (1 : 30 000) after immunolabelling show high-density gold grains in

the outer lumen (OL) and in the cavity, but not in the dendrites (D) or in the

cuticle (C). (D) Male tarsus: cross-section below the peg base (1 : 30 000).

Also contact chemoreceptor sensilla on the undersurface of the tarsus show

selective labelling in the same parts as in the other contact sensilla.

752 S. Angeli et al. (Eur. J. Biochem. 262) q FEBS 1999

Expressed in single individuals (at least two in the female andthree in the male), they represent different genes which haveevolved towards differentiated functions. Differences betweenmale and female proteins, although limited to a few amino acids,could still support a role in sexual communication. The situationis quite different from that of Lepidopteran OBPs, in whichfewer similarities between sequences of the same species areobserved than between different species. This fact has allowed aclear segregation of Lepidopteran OBPs into classes [7,25±29],also suggesting that different functions are associated with eachclass.

The proteins of S. gregaria described here belong to a family,the first member of which was identified in the antennae ofD. melanogaster and named OS-D [16] or A-10 [17]. Classi-fication of this proteins as a `putative OBP' was not consideredby the authors, as it lacks the 6-cysteine motif of LepidopteranOBPs which is present, on the other hand, in other sequences ofDrosophila [16,17]. As several members of this family havebeen described in different orders of insects (see references citedabove), a role in olfaction has recently been proposed. Thecomplete known sequences are reported in Fig. 2, together withthose of S. gregaria, described here. These proteins arerelatively well conserved throughout evolution, from Phasmidsto Lepidoptera and Diptera, as compared with OBPs, in whichhardly any similarity is noticeable even between Lepidopteraand Diptera. It is also interesting to observe that similarity,measured as percentage of identical residues, is not a function ofphylogenetic distance, but is a more or less constant valuearound 40±50% between any two species or orders. This iseasily seen by looking at the shaded areas of Fig. 2. Somesegments of the sequences appear to be highly conserved, whiledifferences are confined to the remaining regions. In addition,the position of the disulfide bridges, connecting pairs ofcysteines conserved in all these proteins and separated by twoand eight amino acid residues, respectively, should not contributeto the stability of the polypeptide chain itself but only creates twosmall protruding loops which could be considered as a structuralpeculiarity of the CSP molecular architecture. These results,together with the absence of any other additional post-translational modifications (i.e. iron cluster), as shown by MS,ruled out the hypothesis suggested for Cactoblastis cactorumCLP-1 of a certain molecular similarity between this class ofproteins and some types of ferredoxin [14].

Most of the polypeptides reported in Fig. 2 are expressed inorgans rich in chemosensory sensilla, such as antennae, tarsi andpalpi, indicating a possible function in olfaction or taste. Twoexceptions are the protein identified in the ejaculatory bulb(EjB) of D. melanogaster (GenBank accession number U08281)and that purified from regenerating legs (p10) of Periplanetaamericana (Genbank accession number AF030340) [44].

The first protein is produced by the glands that synthesize thesexual pheromone in the male. Therefore, its similarity toproteins probably involved in chemoperception should not besurprising. In fact, in several vertebrates, very similar proteinsare present in the nasal mucosa [45±48] as well as in fluidsknown to be pheromone carriers, such as urine [49±51], saliva[52,53] and vaginal secretion [54]. In insects, similar cases arepoorly documented and limited so far to the mealworm beetleTenebrio molitor, which produces in the tubular accessory sexglands two proteins similar in their sequence to mothpheromone-binding proteins [54]. The physiological functionof the cockroach protein p10, with high similarity to the otherproteins of Fig. 2, poses puzzling questions. It has beensuggested by the authors that this protein may play a role inthe regeneration of legs during larval stages, being very

abundant during this process but nearly absent in the fullyregenerated leg, as well as in the adult. In contrast, we found ourproteins to be extremely concentrated in the tarsi of mature adultindividuals.

Additional information comes from the immunocyto-chemistry experiments using electron microscopy. The proteinsof S. gregaria described in this work appear to be very abundantin the outer lymph of contact sensilla in all chemoreceptionorgans investigated (antennae, tarsi and palpi) but are clearlyabsent from olfactory sensilla. Therefore, if these moleculesplay a role in chemoreception transduction by complexingchemical stimuli, we should search for specific ligands amongthose compounds that are perceived through direct contact. Thiscondition does not necessarily limit the choice to taste stimuli,as they are commonly classified in vertebrates. Apart from saltsand sugars, insects also use contact sensilla to perceive plantmetabolites and even pheromones. We have performed ligand-binding experiments using [3H]glucose and [14C]-labelledbicarbonate (data not shown), the latter suggested by thehypothesis, formulated for CLP-1 of C. cactorum [14], thatthese proteins could be involved in CO2 sensing. Theexperiments were performed in equilibrium dialysis conditions,using the protein at concentrations up to 100 mm, butappreciable quantities of bound ligand were never detected.On the basis of these negative results and the structuralconsiderations reported above, a role for these proteins as CO2

sensors seems unlikely, at least in S. gregaria.Any further hypothesis on the physiological function of the

CSPs can be formulated only after the identification of specificligands and the resolution of the three-dimensional structure ofthe protein. Both these projects require large amounts of sample,unobtainable from natural sources. The lack of post-translationalmodifications, apart from the two disulfide bridges, indicatesthat expression in heterologous systems may provide apolypeptide species identical with the native one. Moreover,this should provide easy access to pure isoforms that areotherwise difficult to purify in homogeneous form from thenatural mixture. Therefore, we are currently expressing theCSPs of S. gregaria in heterologous systems with the aim ofproducing large amounts of proteins for crystallization experi-ments and for ligand-binding assays.

REFERENCES

1. Chapman, R.F. & Joern, A. (1990) Biology of Grasshoppers. John Wiley

and Sons, New York.

2. Schmidt, G.H. (1997) Pheromones in Schistocerca gregaria (Forsk): the

present situation. In New Strategies in Locust Control (Krall, S.,

Peveling, R. & Badiallo, D., eds), pp. 313±321. BirkhaÈuser-Verlag,

Basel.

3. Vogt, R.G. & Riddiford, L.M. (1981) Pheromone binding and

inactivation by moth antennae. Nature (London) 293, 161±163.

4. Kaissling, K.E., Klein, U., De Kramer, J.J., Keil, T.A., Kanaujia, S.,

Hemberger, J. (1985). In Molecular Basis of Nerve Activity

(Changeux, J.P., Hucho, F., Maelicke, A. & Neuman, E., eds),

pp. 173±183. De Gruyter, Berlin.

5. Gyorgyi, T.K., Roby-Shemkovitz, A.J. & Lerner, M.R. (1988)

Characterization and cDNA cloning of the pheromone-binding protein

from the tobacco hornworm, Manduca sexta: a tissue-specific

developmentally regulated protein. Proc. Natl Acad. Sci. USA 85,

9851±9855.

6. Raming, K., Krieger, J. & Breer, H. (1990) Primary structure of a

pheromone-binding protein from Antheraea pernyi: homologies with

other ligand-carrying proteins. Comp. Physiol. B 160, 503±509.

7. Vogt, R.G., Prestwich, G.D. & Lerner, M.R. (1991) Odorant-binding-

q FEBS 1999 Chemoreception proteins in Schistocerca gregaria (Eur. J. Biochem. 262) 753

protein subfamilies associate with distinct classes of olfactory receptor

neurons in insects. J. Neurobiol. 22, 74±84.

8. Maida, R., Steinbrecht, R.A., Ziegelberger, G. & Pelosi, P. (1993) The

pheromone-binding protein of Bombyx mori: purification, character-

isation and immunocytochemical localisation. Insect Biochem. Mol.

Biol. 23, 243±253.

9. Du, G., Ng, C.S. & Prestwich, G.D. (1994) Odorant binding by a

pheromone binding protein-active site mapping by photoaffinity

labeling. Biochemistry 33, 4812±4819.

10. Pelosi, P. & Maida, R. (1995) Odorant binding proteins in insects.

Comp. Biochem. Physiol. 111B, 503±514.

11. Prestwich, G.D. (1996) Proteins that smell: pheromone recognition and

signal transduction. Bioorg. Med. Chem. 4, 505±513.

12. Nagnan-Le Meillour, P., Huet, J.C., Maibeche, M., Pernollet, J.C. &

Descoins, C. (1996) Purification and characterization of multiple

forms of odorant/Pheromone binding proteins in the antennae of

Mamestra brassicae (Noctuidae). Insect Biochem. Mol. Biol. 26,

59±67.

13. MaõÈbeÁche-Coisne, M., Sobrio, F., Delaunay, T., Lettere, M., Dubroca, J.,

Jacquin-Joly, E. & Nagnan-Le Meillour, P. (1997) Pheromone binding

proteins of the moth Mamestra brassicae: specificity of ligand

binding. Insect Biochem. Mol. Biol. 27, 213±221.

14. Maleszka, R. & Stange, G. (1997) Molecular cloning, by a novel

approach, of a cDNA encoding a putative olfactory protein in the

labial palps of the moth Cactoblastis cactorum. Gene 202, 39±43.

15. Pelosi, P. (1998) Odorant-binding proteins: structural aspects. Ann. N.Y.

Acad. Sci. 855, 281±293.

16. McKenna, M.P., Hekmat-Scafe, D.S., Gaines, P. & Carlson, J.R.

(1994) Putative Drosophila pheromone-binding proteins expressed

in a subregion of the olfactory system. J. Biol. Chem. 269,

16340±16347.

17. Pikielny, C.W., Hasan, G., Rouyer, F. & Rosbach, M. (1994)

Members of a family of Drosophila putative odorant-binding

proteins are expressed in different subsets of olfactory hairs.

Neuron 12, 35±49.

18. Ozaki, M., Morisaki, K., Idei, W., Ozaki, K. & Tokunaga, F. (1995) A

putative lipophilic stimulant carrier protein commonly found in the

taste and olfactory systems. Eur. J. Biochem. 230, 298±308.

19. Danty, E., Arnold, G., Huet, J.-C., Masson, C. & Pernollet, J.-C. (1998)

Separation, characterization and sexual heterogeneity of multiple

putative odorant-binding proteins in the honeybee Apis mellifera L.

(Hynenoptera: apidea). Chem. Senses 23, 83±91.

20. Dickens, J.C., Callahan, F.E., Wergin, W.P. & Erba, E.F. (1995)

Olfaction in a hemimetabolous insect: antennal-specific protein in

adult Lygus lineolaris (Heteroptera: Miridae). J. Insect Physiol. 41,

857±867.

21. Wojtasek, H., Hansson, B.S. & Leal, W.S. (1998) Attracted or repelled?

A matter of two neurons, one pheromone-binding protein and a chiral

center. Biochem. Biophys. Res. Commun. 250, 217±222.

22. Picimbon, J.F. & Leal, W.S. (1999) Olfactory soluble proteins of

cockroaches. Insect Biochem. Mol. Biol. in press.

23. Tuccini, A., Maida, R., Rovero, P., Mazza, M. & Pelosi, P. (1996)

Putative odorant-binding proteins in antennae and legs of

Carausius morosus (Insecta, Phasmatodea). Insect Biochem.

Mol. Biol. 26, 19±24.

24. Mameli, M., Tuccini, A., Mazza, M., Petacchi, R. & Pelosi, P. (1996)

Soluble proteins in chemosensory organs of Phasmids. Insect

Biochem. Mol. Biol. 26, 875±882.

25. Breer, H., Krieger, J. & Raming, K. (1990) A novel class of binding

proteins in the antennae of the silkmoth Antheraea pernyi. Insect

Biochem. 20, 735±740.

26. Vogt, R.G., Rybczynski, R. & Lerner, M.R. (1991) Molecular cloning

and sequencing of general odorant-binding proteins GOBP1 and

GOBP2 from the tobacco hawk moth Manduca sexta: comparison

with other insect OBPs and their signal peptides. J. Neurosci. 11,

2972±2984.

27. Krieger, J., Raming, K. & Breer, H. (1991) Cloning of genomic

and complementary DNA encoding insect pheromone binding

proteins: evidence for microdiversity. Biochim. Biophys. Acta

1088, 277±284.

28. Krieger, J., GaÈnssle, K., Raming, K. & Breer, H. (1993) Odorant

binding proteins of Heliothis virescens. Insect Biochem. Mol.

Biol. 23, 449±456.

29. Krieger, E.V., Nickisch, E., Mameli, M., Pelosi, P. & Breer, H. (1996)

Binding proteins from the antennae of Bombyx mori. Insect Biochem.

Mol. Biol. 26, 297±307.

30. Steinbrecht, R.A., Ozaki, M. & Ziegelberger, G. (1992) Immunocyto-

chemical localization of pheromone-binding protein in moth antennae.

Cell Tissue Res. 270, 287±302.

31. Steinbrecht, R.A., Laue, M. & Ziegelberger, G. (1995) Immunolocaliz-

ation of pheromone-binding protein and general odorant-binding

protein in olfactory sensilla of the silkmoths Antheraea and Bombyx.

Cell Tissue Res. 282, 203±217.

32. Laue, M., Steinbrecht, R.A. & Ziegelberger, G. (1994) Immunocyto-

chemical localization of general odorant-binding protein in olfactory

sensilla of the silkmoth Antheraea polyphemus. Naturwissenschaften

81, 178±180.

33. Steinbrecht, R.A. (1998) Odorant-binding proteins: functional aspects.

Ann. N.Y. Acad. Sci. 855, 323±332.

34. Laemmli, U.K. (1970) Cleavage of structural proteins during the

assembly of the head of bacteriophage T4. Nature (London) 227,

680±685.

35. OrruÁ, S., Scaloni, A., Giannattasio, M., Urech, K., Pucci, P. & Schaller,

G. (1997) Amino acid sequence, S±S bridge arrangement and

distribution in plant tissues of thionins from Viscum album. Biol.

Chem. 378, 989±996.

36. Scaloni, A., Di Martino, E., Miraglia, N., Pelagalli, A., Della Morte, R.,

Staiano, N. & Pucci, P. (1996) Amino acid sequence and molecular

modelling of glycoprotein IIb-IIIa and fibronectin receptor iso-

antagonists from Trimeresurus elegans venom. Biochem. J. 319,

775±782.

37. Palmieri, G., Di Palo, M., Scaloni, A., OrruÁ, S., Marino, G. & Sannia, G.

(1996) Glutamate 1-semialdeide aminotransferase from Sulfolobus

sulfataricus. Biochem. J. 320, 541±545.

38. Kyhse-Andersen, J. (1984) Electroblotting of multiple gels: a simple

apparatus without buffer tank for rapid transfer of proteins from

polyacrylamide to nitrocellulose. J. Biochem. Biophys. Methods 10,

203±209.

39. Danscher, G. (1981) Localization of gold in biological tissue. A

photochemical method for light and electronmicroscopy. Histo-

chemistry 71, 81±88.

40. Haniu, M., Acklin, C., Kenney, W.C. & Rohde, M.F. (1994) Direct

assignment of disulfide bonds by Edman degradation of selected

peptide fragments. Int. J. Peptide Protein Res. 43, 81±86.

41. Kendall, M.D. (1970) The anatomy of the tarsi of Schistocerca gregaria

Forskal. Z. Zellforsch. Mikrosk. Anat. 109, 112±137.

42. Blaney, W.M. & Chapman, R.F. (1969) The anatomy and

histology of the maxillary palps of Schistocerca gregaria. J. Zool.

157, 509±535.

43. Ochieng, S.A., Hallberg, E. & Hansson, B. (1998) Fine structure

and distribution of the antennal sensilla of the desert locust

Schistocerca gregaria (Orthoptera: acrididae). Cell. Tissue Res.

291, 525±536.

44. Nomura, A., Kawasaki, K., Kubo, T. & Natori, S. (1992) Purification

and localization of p10, a novel protein that increases in nymphal

regenerating legs of Periplaneta americana (American cockroach).

Int. J. Dev. Biol. 36, 391±398.

45. Pes, D. & Pelosi, P. (1995) Odorant-binding proteins of the mouse.

Comp. Biochem. Physiol. 112B, 471±479.

46. Pes, D., Mameli, M., Andreini, I., Krieger, J., Weber, M., Breer, H. &

Pelosi, P. (1998) Cloning and expression of odorant-binding proteins

Ia and Ib from mouse nasal tissue. Gene 212, 49±55.

47. Garibotti, M., Navarrini, A., Pisanelli, A.M. & Pelosi, P. (1997) Three

odorant-binding proteins from rabbit nasal mucosa. Chem. Senses 22,

383±390.

48. Ganni, M., Garibotti, M., Scaloni, A., Pucci, P. & Pelosi, P. (1997)

Microheterogeneity of odorant-binding proteins in the porcupine

754 S. Angeli et al. (Eur. J. Biochem. 262) q FEBS 1999

revealed by N-terminal sequencing and mass spectrometry. Comp.

Biochem. Physiol. 117B, 287±291.

49. Finlayson, J.S., Asofsky, R., Potter, M. & Runner, C.C. (1965)

Major urinary protein complex of normal mice: origin. Science

149, 981±982.

50. Dinh, B.L., Tremblay, A. & Dufour, D. (1965) Immunochemical study

of rat urinary proteins, their relation to serum and kidney proteins.

J. Immunol. 95, 574±582.

51. Cavaggioni, A., Findlay, J.B.C. & Tirindelli, R. (1990) Ligand binding

characteristics of homologous rat and mouse urinary proteins and

pyrazine binding protein of the calf. Comp. Biochem. Physiol. 96B,

513±520.

52. Shahan, K.M., Denaro, M., Gilmartin, M., Shi, Y. & Derman, E. (1987)

Expression of six mouse major urinary protein genes in the mammary,

parotid, sublingual, submaxillary and lachrymal glands and in the

liver. Mol. Cell. Biol. 7, 1947±1954.

53. Marchese, S., Pes, D., Scaloni, A., Carbone, V. & Pelosi, P. (1998)

Lipocalins of boar salivary glands binding odours and pheromones.

Eur. J. Biochem. 252, 563±568.

54. Henzel, W.J., Rodriguez, H., Singer, A.G., Stults, J.T., Macrides, F.,

Agosta, W.C. & Niall, H. (1988) The primary structure of aphrodisin.

J. Biol. Chem. 263, 16682±16687.

55. Paesen, G.C. & Happ, G.M. (1995) The B proteins secreted by the

tubular accessory sex glands of the male mealworm beetle, Tenebrio

molitor, have sequence similarity to moth pheromone-binding

proteins. Insect Biochem. Mol. Biol. 25, 401±408.