Proteomic Analysis Provides Insights on Venom Processing in Conus textile

INTRODUCTION

Interests in the biochemistry of animal venoms started decades ago. ^1-3 Over the years, the venom components have been isolated, characterized and used as biochemical tools to probe the mammalian neuromuscular system. One that has caught the attention of biologists, neuroscientists and the pharmaceutical industry is the venom of Conus species of marine gastropods, each of which contain a potent cocktail of 100-200 neuroactive peptides.^4-7 It is estimated that 50,000 – 100,000 different peptides exist in the venom of the 500-700 species of the genus Conus. Although only a few percent of Conus peptides have been sequenced and studied, they are already well established and widely used as molecular tools in neuroscience with several in clinical trials and one has been approved as a drug against pain.⁸

Sequencing of cDNA from the Conus textile venom duct⁹ revealed that each peptide is encoded by a distinct mRNA with a prepropeptide organization. Subsequent studies on many conotoxins from different species indicated that the prepropeptides of conotoxins belonging to the same superfamily have a highly conserved signal sequence (pre region), a relatively conserved pro-region, and a highly variable C-terminal toxin-encoding region. Within the hypervariable toxin region are highly conserved cysteine residues corresponding to disulfide patterns characteristic of particular gene superfamilies, such as Osuperfamily, M-superfamily, A-superfamily.^10-13 A superfamily may consist of several families, each family targeting a specific ion channel and/or receptor.

Initial sequencing efforts involved chromatographic methods for the isolation of individual Conus peptides and Edman degradation for amino acid sequencing of conotoxins. Advances in molecular biology, particularly the expressed sequence tag strategy coupled with automated nucleic acid sequencing have facilitated the use of cDNA libraries of Conus venom ducts in determining conopeptide sequences.^14-15However, many of the peptides predicted from conotoxin gene transcripts have not yet been isolated and their degrees of post-translational modification remain uncertain.

Mass spectrometry-based sequencing offers a rapid approach for the elucidation of conotoxin structures. It requires relatively small amount of sample compared to chemical sequencing via Edman degradation and more importantly, samples usually need not be in a purified form. A recent development in mass spectrometry, Makarov's Orbitrap is capable of detecting ions with high resolution by their motion in the spindle-shaped electrode¹⁶ and its high mass accuracy and tandem mass spectrometry capability can provide unambiguous identification of peptides and proteins.^17-18 High resolution and high accuracy mass measurement should enable more rapid exploration of the Conus venom peptide library and at the same time identify post-translational modifications.

Proteomic profiling of venom from different portions of the C. textile duct may provide insights on the overall envenomation strategy used by Conus species. For example, a puzzling observation of Endean and co-workers^1,19 in Conus species is the higher potency of venom from the posterior (proximal) end of the duct that is attached to the venom bulb compared to the venom extracted from the anterior (distal) portion near the proboscis. More recently, Marshall et al. (2002) observed a variation in epithelial composition in the proximal and distal portions of the duct and suggested that this may be responsible for the variation in conotoxin expression along the venom duct.²⁰

To address the localization issue of conotoxin expression in the venom duct, venom samples were obtained from the proximal, central and distal portions of the tubular duct of Conus textile and analyzed by using a LTQ-Orbitrap tandem mass spectrometer to identify the toxins present. Database search was performed using SEQUEST²¹ for unmodified peptides and the ProLuCID search engine²² for identification of post-translationally modified peptides. The post-search summary program DTAselect was used to assemble the results using a reverse database approach.²³ Conotoxins identified in the three duct sections were quantified using spectral counting²⁴ and the number of spectra matching each peptide/protein was taken as a measure of relative concentrations.^24-25

In view of previous observations on Conus biology, the proteomic data on conotoxin distribution along the duct were examined to gain insights on the envenomation process. The different levels of post-translational modification found in many peptides indicate a much greater diversity of peptides in Conus venom than can be gleaned from the available molecular biology data.

MATERIALS AND METHODS

RESULTS

DISCUSSION

Peptide profiling of Conus venom has been used as a method to unravel the complexity and examine the intraspecific variation of conopeptides.^37,38 In this study, a proteomics approach was used to identify conotoxins in the C. textile venom, to determine their levels of expression in different sections of the venom duct and to determine post-translational modifications. This study takes advantage of high mass resolution and high mass accuracy with the use of a linear ion trap (LTQ)-Orbitrap hybrid mass spectrometer¹⁶ and the computational tools SEQUEST/ProLuCID to identify 31 conotoxin sequences. In addition, 25 PTM variants of 15 conotoxins with masses ranging from 1000-3900 Da (see Tables 1, 2 and 5) were also detected. (Among these were 7 peptides in C. textile that matched to sequences described in US Patents for unspecified Conus species and one peptide matched to a sequence from C. litteratus (LeDr243). It is possible that LeDr243 is another example of peptides like conopressins that perform regulatory functions and occur widely in different species of cone snails.^39-42Although conopressins may perform other functions in the land snail Lymnea⁴³ and Aplysia⁴⁴, their presence in Conus venom suggests an important role in the envenomation strategy of Conus.

Using a different extraction procedure, bigger polypeptides have been identified. The number of conotoxins reported in this study is less than the total number of C. textile peptides in the Conoserver database⁴⁵ and they represent only a subset of the venom proteome at a given time and set of conditions. In a study on pooled Conus virgo venom, more components were detected using direct infusion of crude venom extract compared to LC-MS analysis.⁴⁶ Thus, the chromatographic step performed in this study could contribute to the decreased number of peptides identified compared to what is in the database.

By using spectral count as a measure of relative abundance, most of the identified conotoxins were observed to be differentially expressed in various parts of the tubular duct apparatus of the specimen. This is consistent with the chromatographic profile (Fig.1) obtained for the proximal, central and distal portions of the venom duct. Although other metabolites such as arachidonic acid and low molecular weight amines^47,48 may contribute to the observed differences in chromatograms, the main components of Conus venoms are peptides and proteins.

The sensitivity and resolution of LTQ-Orbitrap provide a powerful tool for looking at venom from single specimens and offer the possibility for future analysis of intra-species variations. The data presented here was obtained using only ~10% of the extract from each of the three duct sections. Recently, the use of LTQ-Orbitrap in determining the sequences of peptides from 7% of the venom of a single C. textile specimen was reported.³⁴ Of course, the results obtained from a single snail cannot be used to give a general statement regarding the behavior of other individuals. However, the data when considered in light of findings from different approaches provide insights, and presents insights to plan for future work.

Some conotoxins were found throughout the specimen's venom duct but many were observed only in the proximal-central or central-distal regions, and a couple of peptides belonging to the O-superfamily were restricted to the central region. Our data on the specific location of Tx3c, Tx3a, Tx3.5 are consistent with a previous report on the expression patterns of specific gene superfamilies in particular regions of the venom duct of C. textile.⁴⁹ The data presented in this paper support previous findings that certain portions of the venom duct specialize in the production of a specific set of venom peptides. This regional specialization of gene expression, processing and secretion of conotoxins in the venom duct was demonstrated for the first time by Garrett and co-authors⁴⁹ using a combination of molecular biology and LC-MS techniques. The biosynthesis of certain conotoxins is probably associated with the specific type of epithelial cells found in different sections of the venom duct.²⁰ The biological observations of Endean and co-workers^1,19 and our own observations indicate greater toxicity of venom from the proximal portion compared to the distal portion. This could only be true if there is a difference of toxin levels in the two regions.

Though it is not fully clear how the variation in conotoxin expression is taking place, it is possible that cone snails may have adopted differential expression of peptides to alter venom content and generate different types of toxin cabals depending on the environmental situation. A relevant example is the ability of captive C. geographus to adapt its predatory behavior depending on the nature of prey and hunting environment. The geography cone utilizes different toxin cabals to hunt a single prey or to hunt a school of fish.⁵⁰

In this study on C. textile, three major superfamilies were observed in the venom. The O-superfamily to which King Kong 2 belongs has the highest spectral count followed by M-superfamily and T-superfamily. Along with the other peptides in Table 2, all of these conotoxins work in concert to affect the various molecular targets of victims. Conus textile is a mollusk hunter that preys on other snails in their habitat. The availability of these food resources for C. textile may be dictated by oscillating water currents, daily tidal fluctuations and other changes in the overall coral reef system that affect the availability of prey and presence of marine competitors. The capability of C. textile to adapt to its niche in this fluctuating environment apparently depends on the evolution of its many toxins that are involved in prey immobilization, defense and competition with marine organisms. Most probably, the venom composition can vary depending on the snail's physiological state and environmental conditions.

It is tempting to speculate that even if an arsenal of peptide-encoding genes exist, only a given set of peptides are expressed or processed for a particular envenomation event. This intricate behavioral repertoire of Conus, which is associated with its shifting environment, may be the primary factor that drives a differential expression of venom components.⁵⁰ This “selective expression” strategy in C. textile could result in a number of neurotoxic peptide cocktails that may successfully allow them to shift depending on prey availability or offer protection from a vast number of changing predators or competitors.

As in other cone snails, the venom apparatus of C. textile (Figure 3) includes a venom bulb, a long convoluted venom duct (stretched to show the proximal, central and distal sections), a radula sac that may contain up to 30 harpoons at different stages of formation and a proboscis. In Conus catus, as the cone prepares to sting, the radular tooth has been observed to be first positioned in the mid section of the proboscis then propelled by a high-speed ballistic mechanism to hit the prey.⁵¹ The snail's venom bulb is a muscular organ with a thick wall and a lumen filled with clear liquid. Attached to the posterior end of the duct, the bulb is believed to be involved in the transport and delivery of the potent venom components. It may be able to control the volume of venom injected and determine whether the components injected come only from the distal section or includes components from the central and proximal sections. Cone snails generally inject prey a few times as suggested by the presence of several harpoons in the regurgitate of fishbones and scales. The composition of venom injected might vary with successive injections. Delivering an appropriate amount and a proper combination of conotoxins at each injection may be a key determinant in producing the intended physiological response to subdue the prey. It is also noteworthy to take into account the diversity of families and superfamilies of identified conotoxins with each member targeting a specific ion channel or receptor. As stated by Jones et al. (2000)⁵² the combinatorial strategy of Conus venom presents a perfect scheme for the successful ecological adaptation and envenomation of prey and competitors by cone snails.

Another possible level of control and source of diversity/versatility is the post-translational modification of peptides (PTM) in the Conus venom. The spectral counts of both unmodified and posttranslationally modified conotoxins given in Table 4 show the quantitative variation in extent of processing. The varying levels of modification for each conotoxin could represent underprocessed venom component or deliberately produced variants to increase conotoxin diversity. It would be interesting to find out if cone snails have the ability to regulate the degree of posttranslational processing so that a larger conotoxin pool will be accessible to a combinatorial strategy for envenomation. The PTM variants of conotoxins could offer additional diversity to the myriad of constituents present in Conus venom and might provide additional leverage for the ecological adaptation of these snails. However, further studies should be conducted on the exact role of the PTM variants since the sample analyzed was obtained on venom squeezed out of duct segments and not from milked venom.

One interesting set of peptides found in the present study is shown in Figure 3. Here the previously reported Tx3a precursor³³ are shown along with Tx3a^11,34, Peptide 3 (Tx3a analogue with a hydroxylated valine and Gly instead of amide group at the C-terminus) Peptide 4 (Tx3a containing a hydroxylated valine and an amidated C-terminus) and Peptides 1 and 2 which have additional 10 residues at the N-terminus of Tx3a. The arrows on the Tx3a precursor mark the putative cleavage sites for generating the smaller peptides. Are these alternative cleavage sites in the processing of the Tx3a precursor to generate two conotoxins or is Site 1 cleaved as a result of non-specific proteolytic digestion? However, it is interesting to note that Peptide 2 resulting from cleavage at Site 1 has the PTM pyroglutamate at the N-terminus, which tends to stabilize peptides.

The specificity of posttranslational modifications observed in this study has been found in many Conus peptides.⁵³ One PTM variant of the Textile convulsant peptide (Table 5) that was identified in this study contain five proline residues but only the proline at position 12 underwent hydroxylation. This is consistent with a recent study on TxMEKL-P2 where only Trp-19 was brominated although Trp is found also at position 18.³⁴ What governs this PTM selectivity at present remains unclear at present. A study on Conantokin-G precursor demonstrated that the propeptide region has a potential function of being a recognition sequence for the Conus venom enzyme γ-glutamyl carboxylase.⁵⁵ Recognition signals in the pro-region may specify the amino acid residues to be modified in a conopeptide sequence.⁵³

Proteomics provide insights on the biochemical events involved in the modification process. A typical example of such is the C-terminal amidation of several conotoxins of C. textile. ProLuCID identified C-terminal amidation of several conotoxins (see Table 5) while SEQUEST detected their C-terminal acid counterparts (see Table I). Our data are consistent with the enzymatic conversion of glycine to a C-terminal amide group. The data on the two PTM variants of TxIIIC containing their pro-regions confirm that modification such as hydroxylation and C-terminal amidation can take place before cleavage of the precursor region of the conotoxin. Previous studies have reported TxIIIC to contain two PTM's.^11,34 The current work validates the previous findings and also identified the modifications in both the precursor and mature peptide.

Posttranslational modification is known to perform several roles in conopeptides. In some cases such as γ-carboxylation in conantokin-G, modification is extremely important for activity⁵⁶, in other cases posttranslational modification could improve the functional efficiency of the conopeptides.^53,54,57,58 Studies on tx9a and gm8a peptides⁵⁷ showed that an alternative role to be considered is the facilitation of correct folding of conotoxins particularly when modification occurs early in the secretory process.⁵³ Perhaps another possible role is to stabilize the toxin from enzymatic degradation, as in the racemization of l- to d-tryptophan in contryphans⁶⁰ and cylization of N-terminal glutamate to pyroglutamate^57,61,62 in C. geographus and C.striatus. According to a recent review by Bingham and co-authors⁶³, PTMs have evolved to provide maximal targeting function from a single translational gene product and greater response to evolutionary changes and environmental demands.

The study has provided a number of insights regarding venom processing, However, the results raised a number of questions that could be answered only by comparing the composition of crude venom from the duct with milked venom. The venom actually injected by cone snails has been reported to contain fewer components than that extracted from the duct and ~50% of the component masses found in milked venom of C. consors were not detected in the venom obtained by dissection⁶⁴. It has been suggested that some of the component masses found only in the milked venom could be produced by the salivary gland⁶⁵ and by the epithelial zones between the venom duct and pharynx²⁰. However, it is certainly possible that a significant amount of the component masses distinct to the milked venom could be PTM variants that are produced as the venom is further processed from the time the cone snail detects and hunts its prey until the venom passes from the duct to the proboscis and harpoon. This might involve enzymes produced in the venom duct as well as in the salivary gland and epithelial regions of the pharynx.

CONCLUSION

This study takes advantage of developments in mass spectrometry-based proteomics to gain insights on the envenomation strategy used by C. textile in its predatory behavior. Analysis of the venom using the LTQ-Orbitrap and database searches revealed qualitative and quantitative differences in conotoxin components in the various sections of the venom duct of C. textile. Our findings are consistent with previous reports on the diversity of conotoxins. The data in this study suggest that PTM variants can greatly enrich the peptide pool for the combinatorial strategy used in envenomation. The specialization of conotoxin biosynthesis in different sections of the venom duct, the control of the level of post-translational modification may contribute to the production of toxin cabals appropriate to a particular predatory or defense situation.