Peptides 31 (2010) 384–393

Identification, by molecular cloning, of a novel type of I2-superfamilyconotoxin precursor and two novel I2-conotoxins from the worm-huntersnail Conus spurius from the Gulf of Mexico

Roberto Zamora-Bustillos, Manuel B. Aguilar *, Andres Falcon

Laboratorio de Neurofarmacologıa Marina, Departamento de Neurobiologıa Celular y Molecular, Instituto de Neurobiologıa, Universidad Nacional Autonoma de Mexico,

Campus Juriquilla, Queretaro 76230, Mexico

A R T I C L E I N F O

Article history:

Received 2 September 2009

Received in revised form 3 October 2009

Accepted 6 October 2009

Available online 15 October 2009

Keywords:

Conoidea

Conidae

Conus spurius

I-conotoxin

cDNA cloning

Conotoxin precursor

A B S T R A C T

cDNA was prepared from the venom duct of a single Conus spurius specimen collected near the coast of

Campeche, Mexico. From it, PCR products were generated aiming to clone I-conotoxin precursors. Thirty

clones were sequenced and predicted to encode ten distinct precursors: seven of I2-conotoxins and three

of I2-like-conotoxins. These precursors contain three different, mature toxins, sr11a, sr11b and sr11c, of

which two are novel and one (sr11a) has been previously purified and characterized from the venom of

this species. The precursors include a 26- (I2) or 23- residue signal peptide (I2-like), a 31-residue ‘‘pro’’

region (I2-like), and a 32-residue mature toxin region (I2 and I2-like). In addition, all the precursors have a

13-residue ‘‘post’’ region which contains a g-carboxylation recognition sequence that directs the g-

carboxylation of Glu-9 and Glu-10 of toxin sr11a and, possibly, Glu-13 of toxin sr11b and Glu-9 of toxin

sr11c. This is the first time that a ‘‘post’’ region has been found in precursors of I-conotoxins that also

contain a ‘‘pro’’ region. The ‘‘post’’ peptide is enzymatically processed to yield the amidated mature toxin

sr11a, which implies that g-carboxylation occurs before amidation. Phylogenetic analysis at the whole

precursor level indicates that the I2-like-conotoxins of C. spurius are more related to I2-conotoxins than

to I1- and I3-conotoxins from other species, and that they might represent a new subgroup of the I2-

superfamily. The three I-conotoxins from C. spurius have charge differences at seven to nine positions,

suggesting that they might have different molecular target types or subtypes.

� 2009 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Peptides

journa l homepage: www.e lsev ier .com/ locate /pept ides

1. Introduction

Cone snails or cones (superfamily Conoidea, family Conidae,genus Conus) are marine venomous predators found in tropicalhabitats. They produce a large number of small, highly structuredpeptides (‘‘conopeptides’’ if they have zero or one disulfide bridge,or ‘‘conotoxins’’ if they have two or more disulfide bonds). Thesepeptides are essential for prey capture, and they participate in self-defense and competitive interactions [52].

During Conus evolution, conotoxins and conopeptides havebeen optimized to bind ion channels and neurotransmitterreceptors and transporters with very high affinity and selectivity,and most of them can distinguish between closely related targetsubtypes [43]. Consequently, conotoxins are useful pharmacolo-gical tools for characterizing ion channels and receptor subtypes in

* Corresponding author at: Lab. B-01, Instituto de Neurobiologıa, Campus UNAM-

UAQ Juriquilla, Boulevard Juriquilla 3001, Juriquilla, Qro. 76230, Mexico.

Tel.: +52 442 238 1043; fax: +52 442 238 1043.

E-mail address: maguilar@servidor.unam.mx (M.B. Aguilar).

0196-9781/$ – see front matter � 2009 Elsevier Inc. All rights reserved.

doi:10.1016/j.peptides.2009.10.005

neuroscience research, and they also have considerable potentialfor developing new drugs [22,34].

It has been calculated that the 500–700 Conus species mayexpress 50,000–70,000 distinct conotoxins [42]. The precursors ofconotoxins generally comprise 60–90 amino acids, and they havethree well-defined regions: the N-terminal signal sequence, anintervening ‘‘pro’’ region, and the C-terminal, mature toxin region[23,45]. Based mainly on the sequences of the signal peptides,these highly diverse toxins are classified into several genesuperfamilies (A, C, D, I, J, L, M, O, P, S, T, and V)[24,26,29,36,42,47,48]. Usually, the members of each superfamilyshare a highly conserved signal sequence and a particulararrangement of Cys residues in their mature toxins [42,45,52].Each superfamily can include several pharmacological families,which are defined by the molecular targets and how the toxinsaffect them [27,38,52].

The I-superfamily of conotoxins was defined by five excitatorypeptides purified from the venom of C. radiatus and similarsequences cloned from the venom duct of this species [25]. Thissuperfamily also includes toxins k-BtX from C. betulinus [15] andViTx from C. virgo [30], which target K+ channels, and precursors

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393 385

cloned from the venom ducts of C. vexillum, C. miles, C. striatus, C.

capitaneus, C. imperialis, and C. textile [3,31]. The arrangement ofthe eight Cys residues within the I-conotoxins (C–C–CC–CC–C–C,where ‘‘–’’ denotes one or more non-Cys residue) has been assignedframework # 11 or XI [44]. The I-superfamily has been subdividedinto gene superfamilies I1 and I2 based largely on the signal peptidesequence and the presence or absence of the canonical pro-peptideregion in the precursor [4]. Recently, five I1-conotoxin-likeprecursors with a novel signal sequence have been proposed todefine the I3-superfamily [55].

Here we describe the sequences of novel I-conopeptideprecursors from the venom duct of the worm-hunting cone snailConus spurius, which were identified by cDNA cloning based onconserved elements in the 50-untranslated regions (UTRs) ofpreviously reported I-conotoxins.

2. Materials and methods

2.1. Biological material

One specimen of C. spurius was collected near Sand Island, Stateof Campeche, in the Gulf of Mexico in May, 2008. The venom ductwas dissected from the living snail, immediately added to RNAlater(Qiagen, Hilden, Germany), incubated overnight in the reagent at6 8C, and then stored at �70 8C.

2.2. Isolation of total RNA

Total RNA from a half of the venom duct from a single C. spurius

snail (about 10 mg) was isolated and purified using an SV TotalRNA Isolation System Kit (Promega, Madison WI) according to theinstruction manual.

2.3. cDNA cloning and sequencing

Approximately 200 ng of total RNA from the venom duct wasused to generate cDNA using Superscript II reverse transcriptase(Invitrogen, Carlsbad CA) with a 30-RACE Adapter Primer: 50-GGCCACGCGTCGACTAGTAC-(dT)17-30 (Sigma-Genosys, The Wood-lands TX). For PCR amplification of cDNAs encoding I-superfamilyconotoxins, forward primer ISF-1 (50-AGAGAAGTGACGGAGAT-CAA-30) and reverse adapter primer ISR-1 (50-GGCCACGCGTCGAC-TAGTAC-30) were employed. Primer ISF-1 was designed on thebasis of conserved elements in 50-untranslated regions (UTRs) ofknown I-superfamily conotoxins, whereas ISR-1 is a shorterversion of the 30-RACE Adapter Primer without the poly dT tail.PCR amplification was carried out under the following conditions:a total volume of 50 ml contained 2 ml of cDNA, 1� colorlessreaction buffer, 2 mM MgCl2, each primer at 0.2 mM, 0.24 mMdNTP Mix, and 0.04 U of GoTaq1 DNA Polymerase (Promega). ThePCR amplifications were performed on a Thermal cycler 2720(Applied Biosystems, Foster City, CA) programmed for 3 min at95 8C for initial denaturation, followed by 35 cycles of 95 8C for30 s, 58 8C for 30 s, 72 8C for 30 s, and a final step of 72 8C for 5 min.The PCR products were analyzed by electrophoresis on 1.5% low-melting point agarose gels (Promega). A prominent DNA band of�500 bp was excised from the gel and purified using the PureLinkQuick Gel Extraction Kit (Invitrogen). The PCR products werecloned into the pGEM-T-Easy vector (Promega) via TA cloning, andtransformed into electro-competent E. coli cells XL1-Blue (Strata-gene, La Jolla CA). Thirty clones were randomly selected for cDNApurification and sequencing. Plasmids were purified employing theWizard Plus SV Minipreps DNA Purification Systems (Promega),and they were sequenced in both directions with M13 primers atthe Laboratorio Nacional de Genomica para la Biodiversidad,CINVESTAV-Irapuato (Irapuato, Gto., Mexico), using the dideoxy

chain termination method on a 3730xl DNA Analyzer (AppliedBiosystems).

2.4. Sequence analysis

The nucleotide sequences yielded by the automated sequencerwere first edited by removing vector and/or adapter sequences.The open reading frames (ORFs) corresponding to the precursors ofthe cloned cDNA sequences were predicted with the ORF Finderserver [40]. Signal peptides were predicted, using the neuralnetwork method and hidden Markov models, with the SignalP 3.0server [12,14]. Physicochemical parameters such as theoreticalisoelectric point [2] and grand average of hydropathicity [32] werecalculated by means of the ProtParam server [19].

For sequence comparison and phylogenetic analysis of pre-cursors and mature toxins, we relied on the ConoServer database[8,29]. The search for protein I1-, I2- and I3-superfamily precursorsyielded six, 29, and seven entries, respectively (August 25, 2009).For our analysis, we removed six entries: three lacking the signalpeptide, and three containing Cys arrangements distinct from thatof the I-conotoxins. The search for protein I1-, I2- and I3-superfamily wild type toxins yielded 32, 29 and seven entries,respectively; we removed three entries containing a Cys arrange-ment distinct from that of the I-conotoxins. We kept the namesannotated in this database.

The ClustalW2 [33] server at EMBL-EBI [13] was employed formultiple amino acid sequence alignment and calculation ofpercent sequence identities, using the default settings. Multiplesequence alignments were entered into MEGA 4.0 [51], and thephylogenetic analyses were performed with the neighbor-joiningmethod [49] using the Jones–Taylor–Thornton (JTT) matrix [28]and pair-wise deletion of gaps. Bootstrap values [16] wereestimated from 2000 replicates (95% confidence) [20] with arandom seed.

For the phylogenetic analysis, the mature toxin regions of theI2-like-precursors from C. spurius were considered to start at the C-side of the last basic residue of the ‘‘pro’’ region preceding the firstCys residue of the C-end of the precursor; the mature regions of theI2-precursors from this species were considered to start at the C-side of the predicted signal peptides, which is in agreement withthe chemically determined amino acid sequences of toxins sr11a[1] and sr11b (unpublished results); in both cases, Gly and theother residues of the ‘‘post’’ region were removed from oursequence dataset.

Consensus sequences were identified by the graphical method‘‘Sequence logos’’ [50], as implemented in the WebLogo server andemploying the default essential (non-graphic) settings [11].

2.5. Nomenclature

The ten distinct precursor sequences obtained in this work arenamed as Sr11.1 to Sr11.10, where ‘‘Sr’’ refers to the species C.

spurius, ‘‘11’’ specifies the type of Cys framework (C–C–CC–CC–C–C), and ‘‘1’’ to ‘‘10’’ are numbers arbitrarily assigned to the differentprecursor sequences. In order to save space, only the sequencescorresponding to the ORFs are displayed.

3. Results

3.1. Identification of cDNA clones encoding I-conotoxin precursors

PCR amplification of cDNA from the venom duct from C. spurius

with primers F1 and R1 (based on conserved sequences in the 50-UTRs of I-conotoxins and the adapter sequence of a 30-RACEAdapter Primer, respectively) yielded one prominent PCR productband with �500 bp, which was amplified in E. coli cells. With this

Fig. 1. (A) Nucleotide sequences and deduced amino acid sequences of the open reading frames of the most frequent PCR products corresponding to I-superfamily cDNAs from

Conus spurius. The names of the encoded mature toxins are in parentheses. The predicted ‘‘pro’’ regions are underlined, the mature toxins are shaded, and the ‘‘post’’ regions

are doubly underlined. The Cys residues of the mature toxins are in bold type. (B) Comparison of the ten I-conotoxin pre-pro-peptide sequences found in C. spurius. The figures

in parentheses are the numbers of clones obtained. The ‘‘pro’’ peptides are underlined, the predicted mature toxins are shaded, and the ‘‘post’’ peptides are doubly underlined.

The Cys residues of the predicted mature toxins are in bold type. The asterisk before the N-terminal Gly residues of the ‘‘post’’ regions indicates probable amidation of the

preceding amino acid. Identical residues (*), conserved substitutions (:), and semi-conserved substitutions (.) among all the precursors are shown along the bottom. Refer to

GeneBank accession numbers GU013531, GU013532, GU013533, GU013534, GU013535, GU013536, GU013537, GU013538, GU013539, and GU013540.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393386

strategy, we obtained 30 cDNAs coding for whole precursors of I-conotoxins. After deducing the encoded amino acid sequences(Fig. 1A), ten distinct precursors were identified: Seven precursors(Sr11.1–Sr11.7) have the structure: signal peptide-mature toxin-‘‘post’’ region, characteristic of the I2-conotoxins, whereas threeprecursors (Sr11.8–Sr11.10) have the organization: signal peptide-‘‘pro’’ region-mature toxin-‘‘post’’ region, typical of the I1- and I3-conotoxins, except for the presence of the ‘‘post’’ region (Fig. 1B).These precursors contain three different mature toxins: sr11a(clones Sr11.1 and Sr11.2), sr11b (clones Sr11.3–Sr11.7), and sr11c(clones Sr11.8–Sr11.10) (Fig. 1A). Given that precursors Sr11.8–Sr11.10 share two features with I2-conotoxins (similar signalsequence and the ‘‘post’’ region) but only one (the ‘‘pro’’ region)with the I1- and I3-conotoxins (Fig. 2), we will tentatively refer tothem as the I2-like precursors from C. spurius.

In the case of the I2-precursors, the signal peptides, the maturetoxins, and the ‘‘post’’ regions all have 26, 32, and 13 residues,respectively. In the case of the I2-like-precursors the signal

peptides, the ‘‘pro’’ regions, the mature toxins, and the ‘‘post’’regions all have 23, 31, 32 and 13 residues, respectively. Themature toxins of all the precursors have the expected Cys patternthat characterizes the I-conotoxins (C–C–CC–CC–C–C) [25](Fig. 1B). In all cases, the ‘‘post’’ regions are followed by singleTGA stop codons (Fig. 1A).

Generation of the mature toxins sr11a and sr11b from the I2-precursors requires proteolytic cleavage at the C-side of the Alaresidue in the sequence-TNA located at the C-terminal end of thesignal peptide, whereas in the case of the I2-like-precursors,biosynthesis of the mature toxin sr11c involves proteolyticcleavage at the C-side of the Arg residue in the sequence-LSRlocated at the C-terminal end of the ‘‘pro’’ region of the precursors(Fig. 1B). According to their cDNA sequences, the three distinctmature toxins (sr11a, sr11b, and sr11c) derived from all theseprecursors should have an amidated Pro at their C-termini dueto standard processing of the Gly residue at the C-end of themature toxin regions [18,54] from -FGPG to -FGP-nh2 (clones

Fig. 2. Comparison of the three most frequent I-conotoxin precursors from C. spurius with the other known I-conotoxin precursors. Identical residues (*), conserved

substitutions (:), and semi-conserved substitutions (.) among all the precursors are shown along the bottom. Residues in the multiple sequence alignment (by ClustalW2)

identical to the consensus sequence (the most frequent residues in the ‘‘sequence logo’’) (by WebLogo) are highlighted in black background. The figures to the right are the

number of amino acid residues that comprise each precursor. The names of the precursors are those annotated in the ConoServer database [28]. The first one or two letters

refer to the Conus species. Piscivorous species: Sx, C. striolatus; S, C. striatus; R, C. radiatus. Molluscivorous species: Ep, C. episcopatus; Tx, C. textile. Vermivorous species: Sr, C.

spurius; Be, Bt, C. betulinus; Lt, C. litteratus; Eb, C. eburneus; Im, C. imperialis; Em, C. emaciatus; Mi, C. miles; Cp, C. capitaneus; Vx, C. vexillum; Vi, C. virgo (ViTx) or C. vitulinus

(Vi11.3 and Vi11.7); Ar, C. arenatus; Ca, C. caracteristicus; Pu, C. pulicarius.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393 387

Sr11.1–Sr11.7) and -LVPG to -LVP-nh2 (clones Sr11.8–Sr11.10)(Fig. 1B). This processing has been confirmed in the case of toxinsr11a [1], and it implies that the removal of the ‘‘post’’ region,which contains a g-carboxylation recognition site (g-CRS) [3], is aprerequisite for amidation to occur.

It should be noted that there are two precursors (differing atpositions four and nine of the signal peptide) encoding toxin sr11a,whereas there are five precursors (differing at positions four, nineand 21 of the signal peptide, and at position four of the ‘‘post’’region) that code for toxin sr11b, and three precursors (differing atpositions 15 of the ‘‘pro’’ region, and at position five of the ‘‘post’’region) that encode toxin sr11c (Fig. 1B). The origin of thesedifferences is not clear; given that all the clones were obtainedfrom a single specimen, they may be due to allelic variation or oneevent of gene duplication and mutation in the case of the twoprecursors of sr11a; in the case of sr11c, two events of geneduplication and mutation, or one gene duplication and allelic

variation might explain the presence of three precursors; the fiveprecursors of sr11b might imply up to four events of geneduplication and mutation, or two gene duplications and allelicvariations.

Since all the clones were obtained with the primer pair F1-R1(based on conserved elements in the I-conotoxin 50-UTRs that arevery close to the initiation codon, and the adapter sequence of theuniversal 30-RACE Adapter Primer, respectively), we have noadditional information about the 50-UTRs of these precursors;however, we found short (�160 bp) 30-UTRs followed by poly-Atails of variable length (16–92 bp) (data not shown).

3.2. Sequence characterization of the I-conotoxin precursors

The signal peptides of the I-conotoxin precursors of C. spurius

are highly conserved, with the eight distinct signal peptides having18 identical residues (the signal peptides of the three I2-like

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393388

precursors are identical) (Fig. 1B). However, this region is longer inthe I2-precursors (26 residues) than in the I2-like- (23 residues),and clear differences are observed at the C-terminus of this region,starting at position 21: -(V/F)VLTNA (I2-) vs. -SPG (I2-like-)(Fig. 1B); there are other differences at positions four and nine,but in these cases the residues present in the three I2-like-precursors (Arg and Gly, respectively) also occur in four (Arg) andthree (Gly) I2-precursors (Fig. 1B). The ‘‘pro’’ regions of the three I2-like-precursors all have 31 residues, and they are almost absolutelyconserved; one of them differs from the others only at position 15(Fig. 1B). In both the I2- and the I2-like-precursors, the ‘‘post’’regions comprise 13 amino acid residues, of which eight areconserved (Fig. 1B).

The three different mature I-conotoxins described in this workhave the same number of amino acids between the Cys residues,including three amino acids between the two pairs of adjacent Cysresidues (Fig. 1B); this latter feature has been found in all I2- and I3-conotoxins and in most group-B I1-conotoxins described thus far[3,6,15,25,30,31,55]. However, compared with toxin sr11a purifiedfrom the venom of C. spurius [1], the novel toxins sr11b and sr11chave many amino acid substitutions and only 13 (including eightCys) residues are shared by the three, 32-residue mature toxins(Fig. 1B).

The Cys codons of the mature toxins from C. spurius are highlyconserved: in the thirty clones coding for the ten distinctprecursors, Cys-1, Cys-3, Cys-5, and Cys-7 are encoded by theTGC triplet, whereas Cys-4, Cys-6 and Cys-8 are coded by the TGTcodon; interestingly Cys-2 is encoded by TGC in the precursors ofsr11a (Sr11.1–Sr11.2) and sr11c (Sr11.8–Sr11-10), but it is codedby TGT in those of sr11b (Sr11.3–Sr11.7) (Fig. 1A).

4. Discussion

In this work, we found ten different precursors encoding threedistinct mature I-conotoxins from the vermivorous snail C. spurius

collected off the coast of Campeche, Mexico. Two of these toxinsare novel (sr11b and sr11c), whereas one (sr11a) was purified andchemically sequenced from the venom of specimens collected offthe coast of Yucatan, Mexico [1]. Given that we conducted thisstudy from a single Conus spurius specimen, we cannot concludethat all the individuals belonging to this species express theparticular conotoxins we found.

The I-conotoxins have been subdivided into the gene super-families I1 and I2 based on the sequence of the signal peptide, thepresence (I1) or absence (I2) of the ‘‘pro’’ region, and the absence(I1) or presence (I2) of the ‘‘post’’ region [4,39]; five I1-likeprecursors with a novel signal sequence have been recentlyproposed to define the I3-superfamily [55]. Here, we found sevenprecursors with the typical structure of I2-conotoxins, but we alsoidentified three precursors having the organization that char-acterizes the I1- and the I3-superfamilies except that they alsocontain a ‘‘post’’ region, which had only been found in the I2

superfamily (Figs. 1B and 2). Thus, C. spurius is the first species inwhich ‘‘post’’ peptides have been identified in I1/I3-like precursors,which, as explained in Section 3.1, we have tentatively named as I2-like precursors.

In order to gain insight into the relationship between the I2 andthe I2-like precursors of C. spurius, and between the latter and theI1-, I2-, and I3-superfamilies, we conducted a phylogenetic analysisof the precursors reported to date. For this analysis, we onlyincluded the most frequent precursors from C. spurius, Sr11.1,Sr11.4, and Sr11.9 (Fig. 1), each encoding one of the three distinctmature toxins, sr11a, sr11b, and sr11c, respectively. The overallmean p-distance (calculated by MEGA 4.0) of the 39-sequenceClustalW2 multiple alignment (Fig. 2) is 0.564 (pair-wise deletion),which is within the acceptable range (<0.8) for obtaining a reliable

phylogenetic tree [21]; the corresponding value for the multiplealignment of the 68 mature toxins is 0.609. The analysis indicates:

(1) The separation of I-conotoxins into three well-supportedbranches, I1, I2, and I3, as previously proposed by Buczeket al. [4], Mondal et al. [39], and Yuan et al. [55]. This was clearlyobserved at the whole precursor (Fig. 3A) and signal peptidelevels (data not shown), but not at the level of the mature toxinregion where, for example, I1 and I3 conotoxins occur togetherin one branch defined with 35% probability (Fig. 3B). Thisfinding was not unexpected, given the well-known hypervaria-bility of mature conotoxins [52].

(2) At the whole precursor (Fig. 3A) and signal peptide levels (notshown), the three representative precursors from C. spurius

found in this work are included within the I2 branch. The I2

precursors (Sr11.1 and Sr11.4) are more related to each otherthan to the I2-like precursor (Sr11.9) (Fig. 3A). Thus, despitehaving a ‘‘pro’’ region, the I2-like precursors from C. spurius aremore related to the I2 precursors than to either the I1- or the I3-conotoxins. Indeed, we noticed that the ‘‘pro’’ regions of the I2-like precursors are clearly distinct from those of the I1 and I3

precursors; they are considerably longer (31 residues) thanthose of the I1 (14–18 residues) and the I3 (23 residues)precursors (Fig. 2). In addition, the ‘‘pro’’ regions of the I2-likeprecursors share very few amino acid residues with those of theI1 and I3 precursors (Fig. 2). Therefore, we propose thatprecursors Sr11.8 to Sr11.10 might represent a novel class of I2-conotoxins, which, following the nomenclature of I-conotoxinsproposed by Buczek et al. [4], could be named group D.However, it should be noted that Liu et al. have recentlyreported three I2-conotoxins (Eb11.3, Lt11.1, and Im11.3) withnumbers of inter-Cys residues in some loops that clearly differfrom those present in groups A, B, and C, and that mightrepresent another structural group of I-conotoxins [35],distinct from the one we proposed here; indeed, two of thesetoxins (Eb11.3 and Lt11.1) are grouped together in a I2 sub-branch at both the precursor (Fig. 3A) and mature toxin(Fig. 3B) levels.

(3) The difference between I1-conotoxins belonging to groups Aand B at the whole precursor (Fig. 3A), signal peptide (data notshown), and mature region (Fig. 3B) levels is not absolute. Inthe first case, group-B precursor R11.3 is grouped together withgroup-A precursor RXIC with 96% probability (Fig. 3A);regarding the mature toxins, for example, three group-B toxins(RXID, R11.16, and M11.5) are included in one branch(supported by 87% probability) that includes mainly group-Atoxins (Fig. 3B).

(4) At the mature region level, there are five toxins that areproduced by two or three different species: M11.1a/R11.5 by C.

magus and C. radiatus, M11.5/R11.1 by C. magus and C. radiatus,Fi11.1a/R11.11 by C. figulinus and C. radiatus, BeTx/Sx11.2 by C.

betulinus and C. striolatus, and Eb11.7/Em11.3/Im11.10 by C.

eburneus, C. emaciatus and C. imperialis (Fig. 3B). This is aninteresting finding, because it has been thought that essentiallyno molecular overlap or redundancy exists in the sets ofconotoxins of different Conus species [7,46]. However, theunusual occurrence of this phenomenon has been pointed outfor the a- [31,41], the T-1- [56], and the I2-conotoxins [35].

I-conotoxins, in general, have been purified or cloned from thethree major feeding types of cone snails: piscivorous [4,6,25,31],molluscivorous [3,4,6], and vermivorous [1,4,6,15,30,31,35,53,55].This fact was pointed out by Kauferstein et al., who also proposedthat these toxins may be relatively ancient and widely distributedamong Conus species [31]. Indeed, some piscivorous (C. striatus),molluscivorous (C. episcopatus), and vermivorous (C. figulinus, C.

Fig. 3. Evolutionary relationships of the I-conotoxins of C. spurius (Sr11.1, Sr11.4, and Sr11.9) with the I1-, I2-, and I3-conotoxins. The phylogenetic trees were generated by

MEGA4 from ClustalW2 multiple sequence alignments of the whole precursors (A) and mature regions (B). The probable evolutionary histories were inferred using the

Neighbor-Joining method. The bootstrap consensus trees from 2000 replicates are taken to represent the evolutionary histories of the sequences analyzed. In each panel, the

percentages of replicate trees in which the associated sequences clustered together in the bootstrap test are shown next to the branches, and the superfamilies to which the

sequences belong are indicated to the right. For the abbreviations of the names of the species see legend to Fig. 2; in addition, M, C. magus; Fi, C. figulinus; Au, C. aulicus; L, C.

lynceus; Rg, C. regius.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393 389

Table 1Physicochemical properties and sequence comparison of the three I-conotoxins from C. spurius.

aPositive charges (+) at N-termini and side chains are indicated in the same lines of the sequences and below them, respectively; single (�) and double (=) negative charges of

side chains are indicated below the sequences; positions at which differences in charge (opposite charges, or charged vs. uncharged side chain) are highlighted in gray

background. g, gamma-carboxy-Glu; &, amidated C-terminus. Identical residues (*), conserved substitutions (:), and semi-conserved substitutions (.), as defined by

ClustalW2, are shown at the middle of each pair-wise alignment.bCalculated by means of ProtParam [19]: GRAVY, Grand average of hydropathicity [32]; pI [2]. Approximate values, because posttranslational modifications of residues are

not taken into account by ProtParam.cManually calculated using standard values for the pKa’s of the NH2–, –COOH, and side chain moieties of the standard amino acids [37].dPercent sequence identity (by ClustalW2) of each pair of toxin sequences.eConservative substitutions have been included in the differences.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393390

betulinus) species have been found to synthesize both I1- and I2-conotoxins [4,6,31], strongly suggesting that they have animportant role in prey capture, defense against predators, and/or competitive interactions.

The outstanding sequence diversity of I-conotoxins was noticedwhen this superfamily was first described, and the possibility thatthese peptides have diverse molecular targets was also suggested[25]. This remarkable variety at both the intra- and inter-specieslevels and the accompanying structural differences could makethese toxins valuable tools to study structure-function relation-ships. At the molecular level, I-conotoxins have been shown toaffect two types of molecular targets: Na+ and K+ channels. Toxin k-BtX increases the probability of opening of BK channels [15],whereas toxin ViTx blocks Kv1.1 and Kv1.3 channels [30]; toxin i-RXIA acts as an agonist of Nav1.2, Nav1.6, and Nav1.7 channels[5,17].

It has been suggested that toxin sr11a, encoded by clones Sr11.1and Sr11.2, might target BK channels due to its similarities withtoxin k-BtX; however, this needs to be confirmed by electro-physiological experiments [1]. The novel toxins sr11b and sr11cshare some physicochemical features with toxin sr11a: they arehydrophilic and acidic peptides, negatively charged or neutral atphysiological pH (Table 1). However, despite this generalresemblance, compared with toxin sr11a, they have 15 and 16amino acid substitutions that result in charge differences at 8 and 9positions, respectively (Table 1). Thus, it is likely that the three I-conotoxins from C. spurius have different subtypes (or even types)of molecular targets.

In order to gain insight into the probable molecular targets oftoxins sr11b and sr11c, we compared them with the known matureI-conotoxins (Fig. 4). This analysis revealed that: (1) despite thedifferences mentioned above, the toxins from C. spurius have moresequence identity (by ClustalW2) to each other than to themajority of the other I-conotoxins; (2) there are some toxins (all of

them belonging to the I2-superfamily) from other species that havesequence identities with toxins sr11a, sr11b, and sr11c (high-lighted in gray background) comparable to those among the latterthemselves (highlighted in bold face); unfortunately, only toxinsBeTX and TxXI have been purified from venoms, and the moleculartarget has been identified just in the case of toxin BeTX (BKchannels [15]); and (3) the sequence identities of toxins sr11a,sr11b, and sr11c with the only other toxin, the I1-conotoxin i-RXIA,whose molecular target has been recently identified (Nav1.2,Nav1.6, and Nav1.7 channels [5,17]) are very low (25%, 28%, and28%, respectively), and thus it is unlikely that any of them targetsone or more of these channels.

In addition, this comparison revealed some interesting generalpatterns when we determined the consensus sequences of thetoxins according to their superfamily (not shown), and identifiedthe corresponding consensus residues within each sequence(Fig. 4): (1) Gly residues at positions 10 and 33 (of the multiplealignment) are present in most I1-, I2-, and I3-conotoxins, and thusthey might have a very important structural role; (2) there are twomain types of I1-conotoxins (also apparent in Fig. 3B) thatcorrespond to groups A (24 sequences) and B (seven sequences)defined by Buczek et al. [4,6]; and (3) three main types of I2-conotoxins (also evident in Fig. 3B) can be distinguished, and one ofthem comprises two of the three I2-conotoxins suggested by Liuet al. that might represent another structural group of I-conotoxins[35] (Fig. 3A and B). Furthermore, this analysis made patentresidues that are almost exclusive of some groups, for example, Glyresidues at positions eight (I3-conotoxins), 21 (I3- and group-B I1-conotoxins), and 45 (I3- and group-A I1-conotoxins). Finally, thiscomparison clearly shows the existence of four and two subgroupswithin group-B I1- and the second group of I2-conotoxins,respectively, and the identities, similarities (conservative sub-stitutions), and differences among the toxins included in each ofthem.

Fig. 4. Comparison of the three mature I-conotoxins found in C. spurius (bold type) with the other known mature I-conotoxins. The original output order (aligned) of the

ClustalW2 multiple sequence alignment was partially modified to group the toxins according to their superfamilies; in the initial alignment sub-groups of the superfamilies

appeared in the order I1, I3, I2, I1, and I2; the sequences belonging to these sub-groups correspond (with three exceptions) to the main branches indicated to the right of Fig. 3B.

On the right are the number of amino acid residues that comprise each toxin (in parentheses) and its percent sequence identity (by ClustalW2) with the toxins from C. spurius.

The figures at the top are the positions along the alignment. Identical residues (*) among all the toxins are shown at the bottom. For each sequence, residues that are

coincident with the consensus amino acids for the corresponding superfamily are highlighted (Cys residues, in gray background; other residues, in black background). The

names of the toxins are those annotated in the ConoServer database [28]. For the abbreviations of the names of the species see legends of Figs. 2 and 3.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393 391

Fig. 5. Comparison of the mature toxin and ‘‘post’’ regions of the I2-conotoxins from

C. spurius with the known I2-conotoxins that contain g-carboxy-Glu residues (g).

The name of each C. spurius clone is followed by the name of the corresponding

mature toxin in parentheses. The other toxins are referred to by the name of the

mature toxin: Gla-TxXI, from C. textile; k-BtX, from C. betulinus. The g-carboxylation

recognition site [3] is shown below the ClustalW2 alignment, and the consensus

amino acids are in bold type (B, basic amino acid; X, any amino acid; J, hydrophobic

amino acid). Residues in the ‘‘post’’ regions that are coincident with the consensus

amino acids are highlighted in black background, whereas identical or similar

residues in most of the ‘‘post’’ regions are highlighted in gray background. Glu

residues that are potential substrates for g-carboxylation are in bold type and

doubly underlined.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393392

The precursors identified in this study contain a 13-residue‘‘post’’ region that includes a g-carboxylation recognition site(g-CRS) very similar to those present in the precursors ofI2-conotoxins k-BtX [15] and Gla-TxXI [3], which have three andone g-carboxy-Glu residues (Gla), respectively, and amidated C-termini (Fig. 5). A consensus sequence (B–X–X–J–X–X–X–X–B,where B, X and J are basic, any, and hydrophobic amino acidresidues, respectively) has been identified from the alignment ofthe ‘‘pro’’ and ‘‘post’’ regions of the precursors of Gla-containingConus peptides [3] (Fig. 5). The ‘‘post’’ region is predicted to beprocessed to yield the corresponding amidated mature toxins. Thepresence of Gla residues at positions 9 and 10 has beendemonstrated in toxin sr11a, which has also been shown to havean amidated C-terminus [1]. Thus, we conclude that in these threeI2-conotoxins the carboxylation of Glu residues occurs before theamidation of the C-terminal residue.

The presence of g-CRS within the ‘‘post’’ regions of theprecursors of toxins sr11b and sr11c suggests that they mighthave Gla residues at positions 13, and 4 and 9, respectively (Fig. 5).It should be pointed out that, given that toxins k-BtX and Gla-TxXIhave Gla residues at position 4, it would be expected that Glu-4 oftoxin sr11a would have the same post-translational modification(Fig. 5); the absence of g-carboxylation at this residue might be dueto incomplete processing of some Sr11.1 precursors, or to thethree-dimensional structure of the toxin, that might make thisresidue inaccessible to the g-glutamyl carboxylase; notice thatsr11a has two additional Glu residues at positions 23 and 25 thatalso are not g-carboxylated.

Interestingly, we found that most I-conotoxin precursors havethe same position-specific Cys codon usage as the precursors ofsr11a and sr11c (Cys-1, TGC; Cys-2, TGC; Cys-3, TGC; Cys-4, TGT;Cys-5, TGC; Cys-6, TGT; Cys-7, TGC; Cys-8, TGT). However, thereare several exceptions: (1) in I2-conotoxin BeTX, Cys-2 [15]; (2) inI3-conotoxin Ca11.1, Cys-5 and Cys-8 [55]; (3) in the 17 I1-precursors from C. radiatus, Cys-7 and Cys-8, in R11.3, Cys-5, and inRXIC, RXIB, R11.17, RXIA, R11.7 and RXID, Cys-1 [25,29]; (4) in I2-conotoxin Eb11.3, Cys-1, Cys-4, and Cys-5 [35]; (5) in I2-conotoxinLt11.1, Cys-8 [35] (data not shown). This non-absolute position-specific codon conservation also occurs for the Cys residues in themature region of other conotoxin superfamilies, but its origin andimplications are still unknown [9,10].

In summary, this work revealed the primary structures of twonovel I2-conotoxins and one new type of I2-conotoxin precursorfrom C. spurius. Although the molecular targets of these toxins areunknown, one of them (sr11a) causes stiffening of body, limbs, andtail when injected intracranially into mice [1]; due to the charge

differences with sr11a, the novel toxins sr11b and sr11c probablyhave distinct biological activities and types or subtypes ofmolecular targets. In addition, these findings provide insights intothe evolution of this class of conotoxins, and they will be useful forsynthesizing these peptides in order to identify their moleculartargets.

Acknowledgements

This investigation was supported by Grant: 79929 (to M.B.A.)from Consejo Nacional de Ciencia y Tecnologıa, Mexico (CON-ACYT). R.Z.B. was supported by a post-doctoral fellowship from theDireccion General de Asuntos del Personal Academico, UniversidadNacional Autonoma de Mexico. We gratefully acknowledge Dr.Sudhir Kumar for suggesting the parameters for the phylogeneticanalysis, and Dr. Dorothy D. Pless for revising the manuscript. Wealso thank Lic. Pilar Galarza for retrieving bibliographic references,and Dr. Anaid Antaramian Salas and M. en C. Adriana GonzalezGallardo from the Proteogenomics Unit for providing LB mediumwith antibiotics in Petri dishes and access to the thermal cycler andthe NanoDrop spectrophotometer.

References

[1] Aguilar MB, Lopez-Vera E, Heimer de la Cotera EP, Falcon A, Olivera BM, MailloM. I-conotoxins in vermivorous species of the West Atlantic: peptide sr11afrom Conus spurius. Peptides 2007;28:18–23.

[2] Bjellqvist B, Basse B, Olsen E, Celis JE. Reference points for comparisons of two-dimensional maps of proteins from different human cell types defined in a pHscale where isoelectric points correlate with polypeptide compositions. Elec-trophoresis 1994;15:529–39.

[3] Brown MA, Begley GS, Czerwiec E, Stenberg LM, Jacobs M, Kalume DE, et al.Precursors of novel Gla-containing conotoxins contain a carboxy-terminalrecognition site that directs g-carboxylation. Biochemistry 2005;44:9150–9.

[4] Buczek O, Yoshikami D, Watkins M, Bulaj G, Jimenez EC, Olivera BM. Char-acterization of D-amino-acid-containing excitatory conotoxins and redefini-tion of the I-conotoxin superfamily. FEBS J 2005;272:4178–88.

[5] Buczek O, Wei D, Babon JJ, Yang X, Fiedler B, Chen P, et al. Structure and sodiumchannel activity of an excitatory I1-superfamily conotoxin. Biochemistry2007;46:9929–40.

[6] Buczek O, Jimenez EC, Yoshikami D, Imperial JS, Watkins M, Morrison A, et al.I1-superfamily conotoxins and prediction of single D-amino acid occurrence.Toxicon 2008;51:218–29.

[7] Bulaj G, Olivera BM. Folding of conotoxins: formation of the native disulfidebridges during chemical synthesis and biosynthesis of Conus peptides. Anti-oxid Redox Signal 2008;10:141–55.

[8] ConoServer home page: http://research1t.imb.uq.edu.au/conoserver/.[9] Conticello SG, Pilpel Y, Glusman G, Fainzilber M. Position-specific codon

conservation in hypervariable gene families. Trends Genet 2000;16:57–9.[10] Conticello SG, Gilad Y, Avidan N, Ben-Asher E, Levy Z, Fainzilber M. Mechan-

isms for evolving hypervariability: the case of conopeptides. Mol Biol Evol2001;18:120–31.

[11] Crooks GE, Hon G, Chandonia JM, Brenner SE. WebLogo: a sequence logogenerator. Genome Res 2004;14:1188–90.

[12] Dyrløv Bendtsen J, Nielsen H, von Heijne G, Brunak S. Improved prediction ofsignal peptides: SignalP 3.0. J Mol Biol 2004;340:783–95.

[13] European Molecular Biology Laboratory-European Bioinformatics Institutehome page: http://www.ebi.ac.uk/Tools/clustalw2/index.html.

[14] Expert Protein Analysis System home page: http://www.cbs.dtu.dk/services/SignalP/.

[15] Fan C-H, Chen X-K, Zhang C, Wang L-X, Duan K-L, He L-L, et al. A novelconotoxin from Conus betulinus, k-BtX, unique in cysteine pattern and infunction as a specific BK channel modulator. J Biol Chem 2003;278:12624–33.

[16] Felsenstein J. Confidence limits on phylogenies: an approach using the boot-strap. Evolution 1985;39:783–91.

[17] Fiedler B, Zhang MM, Buczek O, Azam L, Bulaj G, Norton RS, et al. Specificity,affinity and efficacy of iota-conotoxin RXIA, an agonist of voltage-gatedsodium channels Nav1.2, 1.6 and 1.7. Biochem Pharmacol 2008;75:2334–44.

[18] Fricker LD. Neuropeptide-processing enzymes: applications for drug discov-ery. AAPS J 2005;7:E449–55.

[19] Gasteiger E, Hoogland C, Gattiker A, Duvaud S, Wilkins MR, Appel RD, et al.Protein identification and analysis tools on the ExPASy server. In: Walker JM,editor. The proteomics protocols handbook. Totowa, NJ: Humana Press; 2005.p. 571–607.

[20] Hall BG. Phylogenetic trees made easy. A how-to manual, 2nd ed., Sunderland:Sinauer Associates, Inc.; 2004. p. 50.

[21] Hall BG. Phylogenetic trees made easy. A how-to manual, 3rd ed., Sunderland:Sinauer Associates, Inc.; 2008. p. 55–6.

R. Zamora-Bustillos et al. / Peptides 31 (2010) 384–393 393

[22] Han TS, Teichert RW, Olivera BM, Bulaj G. Conus venoms—a rich source ofpeptide-based therapeutics. Curr Pharm Des 2008;14:2462–79.

[23] Hillyard DR, Olivera BM, Woodward S, Corpuz GP, Gray WR, Ramilo CA, et al. Amolluscivorous Conus toxin: conserved frameworks in conotoxins. Biochem-istry 1989;28:358–61.

[24] Imperial JS, Bansal PS, Alewood PF, Daly NL, Craik DJ, Sporning A, et al. A novelconotoxin inhibitor of Kv1.6 channel and nAChR subtypes defines a newsuperfamily of conotoxins. Biochemistry 2006;45:8331–40.

[25] Jimenez EC, Shetty RP, Lirazan M, Rivier J, Walker C, Abogadie FC, et al. Novelexcitatory Conus peptides define a new conotoxin superfamily. J Neurochem2003;85:610–21.

[26] Jimenez EC, Olivera BM, Teichert RW. aC-conotoxin PrXA: a new familyof nicotinic acetylcholine receptor antagonists. Biochemistry 2007;46:8717–24.

[27] Jones RM, Bulaj G. Conotoxins new vistas for peptide therapeutics. Curr PharmDes 2000;6:1249–85.

[28] Jones DT, Taylor WR, Thornton JM. The rapid generation of mutation datamatrices from protein sequences. Comput Appl Biosci 1992;8:275–82.

[29] Kaas Q, Westermann JC, Halai R, Wang CK, Craik DJ. ConoServer, a database forconopeptide sequences and structures. Bioinformatics 2008;24:445–6.

[30] Kauferstein S, Huys I, Lamthanh H, Stocklin R, Sotto F, Menez A, et al. A novelconotoxin inhibiting vertebrate voltage-sensitive potassium channels. Tox-icon 2003;42:43–52.

[31] Kauferstein S, Huys I, Kuch U, Melaun C, Tytgat J, Mebs D. Novel conopeptidesof the I-superfamily occur in several clades of cone snails. Toxicon2004;44:539–48.

[32] Kyte J, Doolittle RF. A simple method for displaying the hydropathic characterof a protein. J Mol Biol 1982;157:105–32.

[33] Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H,et al. ClustalW2 and ClustalX version 2. Bioinformatics 2007;23:2947–8.

[34] Lewis RJ, Garcia ML. Therapeutic potential of venom peptides. Nat Rev DrugDiscov 2003;2:790–802.

[35] Liu Z, Xu N, Hu J, Zhao C, Yu Z, Dai Q. Identification of novel I-superfamilyconopeptides from several clades of Conus species found in the South ChinaSea. Peptides 2009. doi: 10.1016/j.peptides.2009.06.036.

[36] Loughnan M, Nicke A, Jones A, Schroeder CI, Nevin ST, Adams DJ, et al.Identification of a novel class of nicotinic receptor antagonists: dimericconotoxins VxXIIA, VxXIIB, and VxXIIC from Conus vexillum. J Biol Chem2006;281:24745–5.

[37] Mathews CK, van Holde KE, Ahern KG. Biochemistry, 3rd ed., San Francisco,CA: Benjamin/Cummings, an Imprint of Addison Wesley Longman; 1999. p.128.

[38] McIntosh JM, Jones RM. Cone venom—from accidental stings to deliberateinjection. Toxicon 2001;39:1447–51.

[39] Mondal S, Babu RM, Bhavna R, Ramakumar S. I-conotoxin superfamily revis-ited. J Pept Sci 2006;12:679–85.

[40] National Center for Biotechnological Information home page: http://www.ncbi.nlm.nih.gov/gorf/gorf.html.

[41] Nicke A, Loughnan ML, Millard EL, Alewood PF, Adams DJ, Daly NL, et al.Isolation, structure, and activity of GID, a novel a4/7-conotoxin with anextended N-terminal sequence. J Biol Chem 2003;278:3137–44.

[42] Norton RS, Olivera BM. Conotoxins down under. Toxicon 2006;48:780–98.[43] Olivera BM. E.E. Just Lecture, 1996. Conus venom peptides, receptor and ion

channel targets, and drug design: 50 million years of neuropharmacology. MolBiol Cell 1997;8:2101–9.

[44] Olivera BM. Conus venom peptides: reflections from the biology of clades andspecies. Annu Rev Ecol Syst 2002;33:25–47.

[45] Olivera BM. Conus peptides: biodiversity-based discovery and exogenomics. JBiol Chem 2006;281:31173–7.

[46] Olivera BM, Teichert RW. Diversity of the neurotoxic Conus peptides: a modelfor concerted pharmacological discovery. Mol Interv 2007;7:251–60.

[47] Peng C, Tang S, Pi C, Liu J, Wang F, Wang L, et al. Discovery of a novel class ofconotoxin from Conus litteratus, lt14a, with a unique cysteine pattern. Peptides2006;27:2174–81.

[48] Peng C, Liu L, Shao X, Chi C, Wang C. Identification of a novel class of conotoxinsdefined as V-conotoxins with a unique cysteine pattern and signal peptidesequence. Peptides 2008;2:985–91.

[49] Saitou N, Nei M. The neighbor-joining method: a new method for reconstruct-ing phylogenetic trees. Mol Biol Evol 1987;4:406–25.

[50] Schneider TD, Stephens RM. Sequence logos: a new way to display consensussequences. Nucleic Acids Res 1990;18:6097–100.

[51] Tamura K, Dudley J, Nei M, Kumar S. MEGA4: molecular evolutionary geneticsanalysis (MEGA) software version 4.0. Mol Biol Evol 2007;24:1596–9.

[52] Terlau H, Olivera BM. Conus venoms: a rich source of novel ion channel-targeted peptides. Physiol Rev 2004;84:41–68.

[53] Vianna Braga MC, Konno K, Portaro FC, de Freitas JC, Yamane T, Olivera BM,et al. Mass spectrometric and high performance liquid chromatographyprofiling of the venom of the Brazilian vermivorous mollusk Conus regius:feeding behavior and identification of one novel conotoxin. Toxicon2005;45:113–22.

[54] Wu J, Dai L, Lan Z, Chi C. Genomic organization of three neurotoxins active onsmall conductance Ca2+-activated potassium channels from the scorpionButhus martensi Karsch. FEBS Lett 1999;452:360–4.

[55] Yuan DD, Liu L, Shao XX, Peng C, Chi CW, Guo ZY. New conotoxins define thenovel I3-superfamily. Peptides 2009;30:861–5.

[56] Zamora-Bustillos R, Aguilar MB, Falcon A, Heimer de la Cotera EP. Identifica-tion, by RT-PCR, of four novel T-1-superfamily conotoxins from the vermivor-ous snail Conus spurius from the Gulf of Mexico. Peptides 2009;30:1396–404.