In Silico Biology 5, 0050 (2005); ©2005, Bioinformation Systems e.V.  


I-superfamily conotoxins: sequence and structure analysis


Sukanta Mondal1, Ramachandran Vijayan2, Kannambath Shichina 2, Rajasekaran Mohan Babu3 and Suryanarayanarao Ramakumar1,3*



1 Department of Physics, Indian Institute of Science, Bangalore 560 012, India
2 Department of Botany, Bharathiar University, Coimbatore, India
3 Bioinformatics Centre, Indian Institute of Science, Bangalore 560 012, India




* Corresponding author

   Phone: +91-80-2293 2718, +91-80-2293 2469;
   Fax: +91-80-2360 2602, +91-80-2360 0551
   Email: ramak@physics.iisc.ernet.in




Edited by E. Wingender; received June 30, 2005; revised and accepted October 15, 2005; published October 30, 2005



Abstract

I-superfamily conotoxins have four-disulfide bonds with cysteine arrangement C-C-CC-CC-C-C, and they inhibit or modify ion channels of nerve cells. They have been characterized only recently and are relatively less well studied compared to other superfamily conotoxins. We have detected selective and sensitive sequence pattern for I-superfamily conotoxins. The availability of sequence pattern should be useful in protein family classification and functional annotation. We have built by homology modeling, a theoretical structural 3D model of ViTx from Conus virgo, a typical member of I-superfamily conotoxins. The modeling was based on the available 3D structure of Janus-atracotoxin-Hv1c of Janus-atracotoxin family whose members have been suggested as possible biopesticides. A study comparing the theoretically modeled structure of ViTx, with experimentally determined structures of other toxins, which share functional similarity with ViTx, reveals the crucial role of C-terminal region of ViTx in blocking therapeutically important voltage-gated potassium channels.

Keywords: I-superfamily conotoxins, protein sequence pattern, homology modeling, shaker K+ channel, inhibitor cystine knot



Introduction

Cone snail toxins, conotoxins, secreted by the venom gland of carnivorous molluscs are small peptides that target ion channels and G protein coupled receptors [Rajendra et al., 2004]. They provide promising application in the treatment among others of chronic pain, epilepsy and cardiovascular diseases. Conotoxins are classified into several superfamilies, namely A, M, O, P, S, T and I, based on characteristics such as disulfide connectivity, highly conserved N-terminal precursor sequence and similar mode of actions [Terlau and Olivera, 2004 ].

I-superfamily conotoxins contain four disulfide bonds and they inhibit or modify ion channel of nerve cells [Kauferstein et al., 2004]. They were characterized only recently and are relatively less well studied compared to other superfamily conotoxins. Hence in silico analysis of I-superfamily conotoxins is of value in order to gain understanding of their structural and functional features through a comparison with other better-characterized toxins.

In the present report, we have carried out bioinformatics analysis of I-superfamily conotoxins in two parts. In the first part, we have carried out a comparison of the available protein sequences of I-superfamily conotoxins to extract a characteristic sequence pattern, which is likely to be useful in the functional annotation of newly determined sequences. In the second part, we have chosen for analysis, a typical member of the I-superfamily conotoxins ViTx whose experimental 3D structure is currently not available. ViTx from Conus virgo inhibits the vertebrate K+ channels Kv1.1 and Kv1.3 but not Kv1.2 [Kauferstein et al., 2003]. ViTx has cysteine arrangement C-C-CC-CC-C-C which is similar to that in the biopesticide Janus-faced Atracotoxin (J-AcTx) family from Hadronyche versuta [Wang et al., 2000]. The similarity in cysteine arrangement has enabled us to build the 3D structure of ViTx through homology modeling based on the experimentally characterized structure of the J-AcTx-Hv1c. Further, we have compared the modeled 3D structure of ViTx with the known structures of functionally similar toxins. The study has suggested a crucial role for the C-terminal region of ViTx in blocking of pore-loop region of therapeutically important voltage-gated potassium channels.



Materials and methods


All the sequences (Table 1) for I-superfamily conotoxins were downloaded from Swiss-Prot [Bairoch et al., 2004], release 47.1.

Table 1: I-superfamily conotoxins and J-atracotoxins family members.
Superfamily/family Species Swiss-Prot entries
I-superfamily conotoxin C. miles P69498 (27-60)
C. striatus P69499 (27-61)
C. vexillum P69500 (27-60), P69501 (27-60)
C. betulinum Q9U3Z3 (27-57)
C. imperialis P69495 (27-60), P69496 (27-60), P69497 (27-60)
C. capitaneus P69494 (27-61)
C. virgo Q7YZS9 (27-60)
C. radiatus Q7Z0A5 (1-44), Q7Z0A3 (1-44), Q7Z0A1 (1-42), Q7Z098 (1-45), Q7Z091 (1-41), Q7Z0A0 (1-43), Q7Z092 (1-46), Q7Z093 (1-46), Q7Z0A6 (1-42), Q7Z099 (1-45), Q7Z090 (1-40), Q7Z094 (1-46), Q7M4K5 (1-37), Q7Z0A4 (1-44), Q7Z0A2 (1-42), Q7Z097 (1-45), Q7Z096 (1-43), Q7Z095 (1-46)
J-AcTx H. versuta P82227 (1-36), P82226 (1-36), P82228 (1-37)
Starting and ending positions of a matured toxin are mentioned in parentheses with corresponding entries.


Sequence pattern analysis

Of the available 28 I-superfamily conotoxin sequences, redundant sequences were removed to obtain 19 sequences. Two sequences having identity of greater than 95% were considered redundant, and lesser length sequence was excluded from this analysis. Multiple sequence alignment (MSA) was undertaken with the help of the EBI CLUSTALW [Higgins et al., 1994] with default parameters. The sequence logo was constructed with the help of the software Weblogo available online at http://weblogo.berkeley.edu [Schneider et al., 1990]. Amino acid composition for this superfamily was calculated using locally developed perl script.

By using the MSA of I-superfamily conotoxins, we have defined sequence pattern for this superfamily. Sensitivity, selectivity, specificity and correlation of the defined sequence pattern were verified against Swiss-Prot (release 47.1) by ScanProsite [Gattiker et al., 2002] option in ExPASy server. We have calculated sensitivity, selectivity, specificity and correlation, as follows,

where, TP, True Positives (correctly recognized positives); TN, True Negatives, (correctly recognized negatives); FP, False Positives (negatives recognized as positives); and FN, False Negatives (positives recognized as negatives). TN is calculated as the total number of sequences on the Swiss-Prot (181821 entries) less the sum of TP and FN.The similar approach was followed for sequence analysis of J-AcTx family.


Homology modeling of ViTx from Conus virgo

Three-dimensional structural model for I-superfamily conotoxin ViTx (Swiss-Prot databank Accession number: Q7YZS9) was constructed using SDPMOD [Kong et al., 2004], a web server specially designed for comparative modeling of small disulfide bonded proteins. For homology modeling, we used 3D structure of the insecticidal neurotoxin J-AcTx-Hv1c from Hadronyche versuta (Protein Data Bank [Berman et al., 2000] code 1DL0) as a template with cysteine arrangement similar to that of ViTx [Wang et al., 2000]. Statistical evaluation of theoretical 3D model of ViTx was carried out using PROCHECK [Laskowski et al., 1996]. In addition to this, 3D structural superimpositions between a pair of structures were done by using the program ALIGN [Cohen, 1997].



Results and discussion


I-superfamily conotoxins

The novel conopeptides belonging to I-superfamily were found in five or six major clades of cone snails and basal phylogenetic position of Conus imperialis suggest that these toxins are comparatively ancient and more universally distributed among several clades of cone snails, than other groups of apparently derived clade specific conotoxins [Kauferstein et al., 2004]. The aim of our study was to characterize the sequence and structural properties of less explored I-superfamily conotoxin through in silico approach.


Sequence pattern

Sequence information of these toxins (Table 1) was collected from Swiss-Prot database (release 47.1, 181821 entries). Although a high degree of conservation is seen among precursor region of I-superfamily conotoxins, genetic analysis exhibited high divergence in intercysteine region pointing not yet known physiological activities [Kauferstein et al., 2004]. Fig. 1 depicts the MSA and Fig. 2 the weblogo of the toxins where it may be seen that only cysteine residues were conserved and they have the arrangement C-C-CC-CC-C-C. The eight-conserved cysteines provide a structural scaffold with four disulfide bonds. Characterization of I-superfamily conotoxin closes the primarily unfilled gap defined previously in conotoxins superfamily with two, three and five disulfide bonds [Jimenez et al., 2003].



Figure 1: MSA of 28 sequences belonging to I-superfamily conotoxins. The eight conserved cysteines are indicated by * at the top of the alignment. Negatively charged (D and E), positively charged (K and R), polar (N, Q, S and T), aliphatic (I, L, M and V) and aromatic (F, W and Y) residues are shown in red, blue, green, coral and cyan color respectively. Cysteine residues are shown in yellow color. The other amino acids are presented in misty rose color.


Figure 2: Weblogo representation for 19 non-redundant sequences of I-superfamily conotoxin. The nine members excluded from the dataset (Table 1) due to redundancy are P69498, P69501, Q7YZS9, Q7Z0A3, Q7Z098, Q7Z0A6, Q7Z0A4, Q7Z0A2, Q7Z095. The conserved cysteines are represented in yellow color.


We are able to define sequence pattern for this superfamily based on MSA. The defined sequence pattern for I-superfamily is C-{C}(6)-C-{C}(5)-C-C-{C}(1,3)-C-C-{C}(2,4)-C-{C}(3,10)-C where regular expressions are according to the PROSITE format. The arrangement of the cysteine residues and the lengths of inter-cysteine loops characterize the pattern (Fig. 3). This sequence pattern was used for scanning against Swiss-Prot by the help of ScanProsite, yielded true positive hits only. Since no false positive and false negative are detected for the pattern, the sensitivity, selectivity, specificity and correlation are unity. We have deposited the defined sequence pattern in the PROSITE database and it can be accessed with PROSITE accession number PS60019 under PDOC60004 documentation.



Figure 3: Schematic representation of the arrangement of eight conserved cysteines.


Amino acid composition

We have carried out detailed amino acids composition analysis for full length and loop-wise of this superfamily members, whose information might elucidate on the functional mechanism of I-superfamily conotoxins. The study revealed the most frequent occurrence of cysteines in sequences which are involved in disulfide bridge and these bridges can play a crucial role in structural stability of these toxins. Apart from cysteines, hydrophobic residues (I, L, V, M, F, Y, W, A, C, H and T) constitute a major part with 52% and charged residues (R, K, D and E) with 16% (Fig. 4). Eight conserved cysteines in these toxins were numbered I-VIII from N-terminus to C-terminus. The position between Cys I and Cys II were represented as "loop 1", Cys II and Cys III were as "loop 2", Cys IV and Cys V as "loop 3", Cys VI and Cys VII as "loop 4", Cys VII and Cys VIII as "loop 5" (Fig. 3). Positively charged residues (R and K), especially lysine, were found to be dominating in loop 1. Negatively charged residues (D and E) and polar residues (N, Q, S and T) present in loop 2 with dominant ratio. In loop 5 proline was observed predominantly (Fig. 4). Leucine is largely present in loop 4 and 5 while glycine residues prevail in loop 1, 4,and 5. Histidine is mainly present in the 2nd loop. In addition to it, charged residues are observed before Cys I and hydrophobic residues as well as charged residues were present after Cys VIII (Fig. 1).



Figure 4: Amino acid composition for 19 non-redundant sequences of I-superfamily conotoxins. (A) For full length protein and (B) For individual loop. Charged residues (D, E, K and R), negatively charged (D and E), positively charged (K and R), polar (N, Q, S and T), aliphatic (I, L, M and V), aromatic (F, W and Y) and other residues are represented as CH, CHN, CHP, POL, ALI, ARO and OTH respectively.


The toxins from snails, spider, scorpions act as structurally robust neuropharmacological ligands with selectivity for blocking various ion channels involved in inflammation, neuropathic pain, traumatic brain injury, focal cerebral ischemia, spinal chord injury [Mclntosh and Jones, 2001] etc. Studies focusing on charged and hydrophobic residues have been carried out to highlight the importance of electrostatic and hydrophobic interactions in pore-loop region of voltage-gated ion channels blocking. A similar study may be expected to throw light on the functional mechanism of I-superfamily conotoxins since its functions have not fully characterized.


Homology modeling of ViTx - a voltage gated K+ channel blocker conotoxin

Template search

ViTx, a novel conotoxin from Conus virgo (Swiss-Prot databank accession number Q7YZS9) is cross-linked with four disulfide bridges. The mature part of the ViTx from Conus virgo comprises amino acids ranging from 27 to 61 [Kauferstein et al., 2003]. Since 3D structure of ViTx is not available, molecular modeling acts as powerful tool to explore at the structural level, its role in voltage-gated ion channel blocking mechanism. SDPMOD, a web server specially meant for modeling small, disulfide rich proteins, was used for modeling ViTx. The cysteine arrangement of ViTx is similar to J-AcTx-Hv1c from Hadronyche versuta (PDB code: 1DL0) (Fig. 5) [Wang et al., 2000], which was used as template for modeling. Spider toxin J-AcTx-Hv1c belongs to Janus Atracotoxin family (Table 1). The manual option in SDPMOD server was used for generating the 3D model for ViTx.



Figure 5: Cartoon representation of J-AcTx-Hv1c (PDB code: 1DL0). All the eight conserved cysteines are represented as ball-and-sticks. The inhibitor cystine knot motif (ICK motif) (Cys3-Cys17, Cys10-Cys22 and Cys16-Cys32) and vicinal disulfide bridge (Cys13-Cys14), disulfide-directed beta hairpin (DDH motif) are highlighted.


Conus peptides typically have two to four disulfide cross-links between cysteine residues with characteristic arrangement in the primary sequence of the peptides. Conotoxins having identical cysteine pattern may sometimes differ in the disulfide connectivity. For example, alpha- and chi-conotoxins have same cysteine arrangement (CC-C-C) but the Cys connectivity for alpha-conotoxin is CysI-CysIII and CysII-CysIV (PDB code: 1B45 [Favreau et al., 1999]) while for chi-conotoxins it is CysI-CysIV and CysII-CysIII (PDB code: 1IEO [Sharpe et al., 2001]), where, four conserved Cys of this family were numbered I to IV from N-terminus to C-terminus. Similarly, mu and mini-M conotoxins have identical Cys arrangement (CC-C-C-C-C) but differ in Cys connectivity [McDougal and Poulter, 2004]. The theoretical 3D model of ViTx has been generated assuming that the disulfide connectivity in ViTx is the same as that in the template, though they are varying in the inter-cysteine loop lengths.

J-AcTx, a family of insecticidal neurotoxins isolated from the venom of Blue Mountains funnel-web spider Hadronyche versuta, plays a crucial role in the insect specific biopesticide technology. J-AcTxs have proved to be insect specific peptide toxins suitable for engineering into plants and viruses, making them valuable for design of novel insecticides [Wang et al., 2000]. Biopesticide J-AcTx (CX6CX2CCXCCX4CX9-10C) (Fig. 6) and I-superfamily conotoxins (CX6CX5CCX1-3CCX2-4CX3-10C) have similar cysteine arrangement C-C-CC-CC-C-C, but differ in lengths of inter-cysteine loops.



Figure 6: MSA of J-AcTx family. Color code as mentioned in Fig. 1.


We have defined sequence pattern for J-AcTx family from MSA (Fig. 6). The defined sequence pattern for J-AcTx family as C-{C}(6)-C-X(2)-C-C-X-C-C-X(4)-C-X(9,10)-C, where regular expressions are according to the PROSITE format. The defined pattern was used for scanning against Swiss-Prot by the help of ScanProsite, yielded true positive hits only. Since no false positive and false negative are seen for the pattern, the sensitivity, selectivity, specificity and correlation are unity, on the Swiss-Prot based on the above mentioned equations (see Materials and methods). We have deposited the defined sequence pattern in the PROSITE database and it can be accessed with PROSITE accession number PS60020 under PDOC60020 documentation.


3D structural model of ViTx

The 3D structural model of ViTx has been generated as described in the section "Materials and methods". The statistical analysis of Ramachandran plot for theoretical model of ViTx (Fig. 7) showed 84% residues in the most favoured regions, 12% residues in additional allowed regions, and 4% residues in generously allowed regions (Fig. 8). The PROCHECK analysis for bond angles, bond length, side chain torsions showed no major deviations from the corresponding allowed values. We have deposited the theoretical model of ViTx in the Protein Data Bank and it can be accessed with the code 1ZJS.



Figure 7: Cartoon representation of theoretical model of ViTx (PDB code: 1ZJS). All the eight conserved cysteines are represented as ball-and-sticks. The model has been generated assuming that the disulfide connectivity in ViTx is the same as that in the template. The ICK motif (Cys3-Cys22, Cys10-Cys25 and Cys21-Cys29) and vicinal disulfide bridge (Cys16-Cys17) are seen.


Figure 8: Ramachandran plot for theoretical 3D model of ViTx (Protein Data Bank code 1ZJS).


The structural superimposition between theoretical 3D model of ViTx and template J-AcTx-Hv1c showed similarity in the over all fold (Fig. 9).



Figure 9: Superimposed structures of ViTx and 1DL0 (C-alpha atoms only) after least-square superimposition using program ALIGN.


Comparative analysis between ViTx and functionally similar animal toxins

Recently Huang et al., 2005, suggested a general voltage-gated ion channel pore-blocking mechanism by their corresponding antagonists such as animal toxins. Blockers are pre-aligned to their target, pore loop region of voltage-gated ion channel by long-range electrostatic interactions. Then, a critical "induced fit" process (reorientation and conformational adjustment) takes place to achieve multiple specific interactions, by short-range interactions such as hydrophobic contacts and hydrogen bonding [Huang et al., 2005].

In ViTx, positively charged residues are present in two clusters, one (Arg2) towards N-terminal region and other three (Arg26, Arg32, Lys35) towards C-terminal region. These positively charged residues might help to give the initial approach of ViTx to the pore-loop region of voltage-gated Kv1-type channel (Fig. 10). The lysine residue (Lys35) is located at the C-terminus of ViTx and it is rather unlikely that it is involved in binding to the pore-loop region of voltage-gated Kv1-type channel [Kauferstein et al., 2003].



Figure 10: Electrostatic representation of theoretical structural models of K+ channel blocker ViTx and pore-loop (P-loop) region of the voltage-gated K+ channel Kv1.1 from Human Homo sapiens (Swiss-Prot Accession no. Q09470). The P-loop region of Kv1.1 from Human was constructed based on the crystallographic structure of KcsA (PDB code: 1BL8 [Doyle et al., 1998]) by structural homology. The molecular modeling calculations were carried out in Homology module in Insight II package (Accelrys, SanDiego, USA) on Silicon Graphics Workstation. After the optimization of side chains, four P-loop regions complex were subject to energy minimization with Steepest Descent and Conjugate Gradient algorithms to optimize the stereochemistry of the P-loop regions model using GROMOS96 implemented in Swiss-Pdb Viewer. PROCHECK was applied for a statistical evaluation of the final model. The P-loop region of Kv1.1 including structural features of the selectivity filter exhibits similar folds as that of KcsA. The C-terminal region of ViTx contains positively charged residues represented in blue color, which is complementary to the negatively charged pore loop region of Kv1.1, which is represented in red color.


In order to identify the residues important in channel blocking, we compared the theoretical model of ViTx with other functionally similar well-characterized animal toxins, such as Kappa conotoxin PVIIA (PDB code: 1KCP [Savarin et al., 1998]) (Fig. 11A), Kaliotoxin (KTX) (PDB code: 2KTX [Gairi et al., 1997] (Fig. 12A), and Agitoxin 2 (AgTx2) (PDB code: 1AGT [Krezel et al., 1995] (Fig. 13A) from scorpions. A similar cysteine arrangement (C-C-C-C-C-C) and high sequence and structural identity between AgTx2 and KTX have been observed (details not shown here). Though ViTx, AgTx2 and KTX have similar function, they differ in their folds. In spite of existence of these toxins with different folds, they are able to recognize various Kv channel subtypes. Accordingly, it has been demonstrated that these peptides may have some key molecular determinants in common [Gasparini et al., 2004; Mouhat et al., 2004; Mouhat et al., 2005].

Structural comparative studies were done for ViTx with potassium channel blocker Kappa conotoxin PVIIA (PDB code: 1KCP [Savarin et al., 1998]) (Fig. 11A) with cysteine arrangement C-C-CC-C-C from fish hunting cone Conus purpurascens (Purple cone) which is the first conotoxin shown to target voltage–gated potassium channels and it was also found to be effective in blocking Shaker potassium channels with high affinity [Shon et al., 1998]. The result revealed no structural equivalent position of the key residues (Lys7, Phe9 and Phe23) of the Kappa conotoxin PVIIA [Savarin et al., 1998] with similar type of residues of ViTx for ion-channel pore blocking mechanism (Fig. 11B).



Figure 11: (A) Cartoon representation of 1KCP. (B) Structure based sequence alignment of 1ZJS (ViTx) and 1KCP.


KTX, which is purified from venom of scorpion Androctonus mauretanicus and a cluster of ten amino acids positioning from 26 to 35 are highly conserved and can play a prominent role in toxin-channel interactions [Romi et al., 1993]. A structural equivalence was observed in Arg26, Val28, and Cys29 of ViTx with Lys27, Met29, and Cys33 respectively of KTX in C-terminal region (Fig. 12).



Figure 12: (A) Cartoon representation of 2KTX. (B) Structure based sequence alignment of 1ZJS and 2KTX.


Agitoxin2 isolated from the venom of scorpion Leirus quinquestriatus are the best characterized channel blocking toxins and it is a strong inhibitor of Shaker related potassium channels [Eriksson and Roux, 2002]. The structural superimposition between ViTx with AgTx2 showed structural equivalence between Arg26 of ViTx with Lys27 of AgTx2 in the C-terminal region [Krezel et al., 1995] (Fig. 13).



Figure 13: (A) Cartoon representation of 1AGT. (B) Structure based sequence alignment of 1ZJS and 1AGT.


Structurally and phylogenetically unrelated toxins that interact with voltage-activated K+ channels usually share a dyad motif composed of a lysine and a hydrophobic amino acid residue (Tyr, Phe or Leu) known as "functional dyad" [Mouhat et al., 2004; Mouhat et al., 2005]. The only C-terminus lysine (Lys35) residue of ViTx is not structurally equivalent with functionally important lysine of other toxins. This result suggests that Lys35 is unlikely involved in binding to a channel structure and this observation has also been suggested by Kauferstein et al., 2003. The Arg26 residue of ViTx could be structurally superimposed on the lysine residue, which is a part of the functional dyad of other toxins. Thus, the above comparative studies suggest that the C-terminal residues, especially Arg26 of ViTx, might provide crucial interaction in channel blocking mechanism.

Therefore, we were interested to check the role of C-terminal region of ViTx, especially Arg26, in blocking the pore-loop region of Vertebrate Kv1 channel. We studied the interaction of vertebrate Kv1 channel from human with ViTx through in silico docking.



Figure 14: Energetically favourable two different orientations of the vertebrate channel blocker ViTx from Conus virgo, docked with theoretically modeled P-loop region of Voltage gated K+ channel (Kv1.1) from Human. Top and bottom panels represent top view and side view of the complex structure respectively. Affinity module of Insight II was used for docking ViTx to Kv1.1 by identifying low energy orientations of the ViTx within the Kv1.1. The C-terminus residue Lys35 is represented in blue color. All arginines of ViTx are shown in pink color. In both orientations of ViTx in Kv1.1, Arg2 is away from selectivity filter. Arg32 and Arg26 residues of ViTx may play a crucial role in interactions with the selectivity filter of P-loop region of Voltage-gated potassium channels.


The Lys35 of ViTx is away from the selectivity filter and it has no role in blocking the potassium ion movement. Moreover Arg2 is also away from the selectivity filter (Fig. 14). The residues in the C-terminal region especially Arg26 and Arg32 of ViTx are near the selectivity filter. The carbonyl oxygen atoms of residues in selectivity filter are essential for dehydration process of K+ ions [Doyle et al., 1998]. The hydrogen bond between Arg26 and Arg32 of ViTx with carbonyls of the residues in the selectivity filter disturbs the dehydration process and prevents the entry of potassium ions. In the docked model, ring of positively charged residues of ViTx anchor to the residues of the K+ channels pore loops although their side chains do not penetrate the channel pore.

There are at least two different mechanisms proposed for blocking the voltage gated potassium channels by corresponding blockers (toxins). One mechanism involves positively charged lysine and hydrophobic residue (Tyr or Phe), act as a dyad, and occludes the potassium channel pore [Mouhat et al., 2004; Mouhat et al., 2005]. The second mechanism involves clustering of positively charged residues on one surface and it act as anchor blocking the pore as a lid [Verdier et al., 2005]. Our in silico docking studies suggest that ViTx may block by means of the second type of mechanism as observed in kappa-M conotoxins RIIIK complexed with TSha1 K+ channel [Verdier et al., 2005].



Conclusions

In this study, we have carried out a detailed bioinformatics analysis of the sequence and structural features of I-superfamily conotoxins in order to gain understanding of the functional aspects of the recently characterized conotoxins superfamily. The selective and sensitive sequence pattern defined by us for this superfamily is likely to be useful in the classification and annotation of conotoxin superfamilies. Comparative studies between theoretical model of ViTx (developed by us) and functionally similar toxins propose the importance of C-terminal region in blocking the voltage gated potassium channels. Our study has drawn attention to the functionally important region, in particular of ViTx, a typical member of the I-superfamily conotoxins, and provides a pointer to experimental work to validate the observations made here based on the in silico analysis.



Acknowledgements

Facilities at the Bioinformatics Centre of Excellence funded by the Department of Biotechnology (DBT), India, were used and are gratefully acknowledged. S. Ramakumar thanks International Business Machines (IBM) for a CAS fellowship grant. We are grateful for the constructive comments offered during the review process. We thank Ms. Bhavna Rajasekaran for useful discussions.




References