Elsevier

Toxicon

Volume 47, Issue 6, May 2006, Pages 650-663
Toxicon

Venom landscapes: Mining the complexity of spider venoms via a combined cDNA and mass spectrometric approach

https://doi.org/10.1016/j.toxicon.2006.01.018Get rights and content

Abstract

The complexity of Australian funnel-web spider venoms has been explored via the combined use of MALDI–TOF mass spectrometry coupled with chromatographic separation and the analysis of venom-gland cDNA libraries. The results show that these venoms are far more complex than previously realized. We show that the venoms of Australian funnel-web spiders contain many hundreds of peptides that follow a bimodal distribution, with about 75% of the peptides having a mass of 3000–5000 Da. The mass spectral data were validated by matching the experimentally observed masses with those predicted from peptide sequences derived from analysis of venom-gland cDNA libraries. We show that multiple isoforms of these peptides are found in small chromatographic windows, which suggests that the wide distribution of close molecular weights among the chromatographic fractions probably reflects a diversity of structures and physicochemical properties. The combination of all predicted and measured parameters permits the interpretation of three-dimensional ‘venom landscapes’ derived from LC-MALDI analysis. We propose that these venom landscapes might have predictive value for the discovery of various groups of pharmacologically distinct toxins in complex venoms.

Introduction

The research community and lay public have been interested for centuries in the secretions from venomous animals, mostly due to their noxious effects on human health (Nicholson and Graudins, 2002). Indeed, early work on snake, bee and scorpion venoms was motivated by the desire to understand the nature of these venoms and the reasons for their toxicity, and to develop efficient means of counteracting their deleterious effects. This work led to the isolation of many peptide and protein toxins as ‘active principles’, thus allowing biochemical and pharmacological characterization of these components. However, it is only in the past two or three decades that the potential of venom components as pharmacological tools and as potential leads for the development of new drugs and pesticides has begun to be realized. As a result, venoms and toxins have begun to generate much broader interest in both the scientific community and in the agrochemical and pharmaceutical industries. Peptide toxin research is nowadays less motivated by ecological, medical, and taxonomical considerations than by the enormous potential that these molecules represent as drug and pesticide leads, and as pharmacological agents.

Although there are about 39,000 described spider species, with an even greater number awaiting characterization, only four groups of spiders are capable of causing fatal human envenomations (Isbister and White, 2004, King, 2004). Thus, given the historical development of toxin research, it is not surprising that spider venoms have received less attention than venoms from other animals due to their minimal impact on human health. Early studies in the 1970s and 1980s focused almost exclusively on venom from widow spiders (Latrodectus spp.), recluse spiders (Loxosceles spp.), and Australian funnel-web spiders (genera Atrax and Hadronyche) due to their ability to inflict human fatalities. Venoms from other spiders were studied only sporadically until the seminal work of Adams and colleagues in the late 1980s revealed the wealth of pharmacology inherent in the different classes of agatoxins isolated from the American funnel-web spider Agelenopsis aperta (Adams, 2004). Since then, numerous spider venoms have been explored with the goal of elucidating their composition, but mostly as sources of novel ion channel and receptor ligands (Escoubas et al., 2000, Corzo and Escoubas, 2003). It has become increasingly apparent that spider venoms contain an extraordinary array of molecules with a dazzling variety of pharmacological properties (Grishin, 1999, Escoubas et al., 2000, Rash and Hodgson, 2002, Corzo and Escoubas, 2003, Sollod et al., 2005). Because of this complexity, spider venoms have become a valuable resource for the isolation of specific high-affinity ligands that can be used to study the role, localization, and regulation of cell components such as a variety of ion channel subtypes. Moreover, some of those molecules are being developed as therapeutic agents (Lewis and Garcia, 2003) or insecticides (Tedford et al., 2004).

Spider venoms represent an almost infinite ‘pharmacological space’ that we have barely begun to explore. Only ∼100 spider-venom peptides and proteins have been described to date; as an example, the 800 or so species of tarantulas (family Theraphosidae) have yielded less than 50 novel peptides so far (Escoubas and Rash, 2004). A very conservative estimate of ca. 100 peptides per venom only would translate into an enormous library of about 4 million molecules, and that is well below the actual numbers as we demonstrate in this report.

So why is so little known about spider-venom peptides? In addition to the late development of spider-venom research as outlined above, accessing this biological resource has always been problematic. In the 1950s and 1960s, scorpions could be collected in the thousands, enabling the extraction of gram quantities of venom for biochemical analysis. In contrast, the analytical resources available at that time were not appropriate for the exploration of milligram or microgram amounts of material. Spider venom can only be obtained by direct milking, electrical stimulation, or by extraction from the venom glands, and all three methods typically yield small amounts of venom. Thus, it is only with the advent of modern separation techniques (HPLC), automated and sensitive Edman sequencing, and more recently mass spectrometry that the full range of spider venoms and their contents has become amenable to biochemical exploration. We are now increasingly able to work with subnanomolar amounts of material originating from complex mixtures, and solve both separation and sequencing problems using submicrogram amounts of material, feats that were unthinkable only 10 years ago for most laboratories.

The rapid development of mass spectrometry and hyphenated LC–MS techniques has been crucial in this field. Toxinology is following in the footsteps of proteomics, which has motivated many of these recent technological improvements. Venom samples present essentially the same challenges as extracts of proteins from cells, organelles, or large macromolecular complexes. They are mixtures of hundreds of components, often of the same molecular size and isoelectric point, can show up to 10 orders of magnitude differences in protein expression, proteins are found in a multiplicity of closely related isoforms and contain post-translational modifications. All of which, when added to the complexity of the biological matrices and the small amounts of material available, often precludes easy separation of all components. Fortunately, a number of these problems have now been solved by both the implementation of hyphenated techniques such as LC–MS and two-dimensional LC–MS, as well as the ever-increasing sensitivity of mass spectrometers. In addition, modern tandem mass spectrometry instruments can now be used for the generation of sequences from peptides and proteins, and although they have not yet replaced Edman degradation for full sequencing, they will certainly be able to do so in the near future. In addition, beyond multidimensional chromatography, mass spectrometry affords a supplementary separative dimension due to its ability to ‘sort’ ions in the mass spectrometer and then analyze or fragment them individually. The net result of all these novel technological capacities for venom research is that today, a single drop of spider venom will yield an incredible amount of information and possibly a host of novel molecules.

However, full sequencing of all peptides directly from venoms remains an elusive target at the moment and therefore chromatography and mass spectrometry analysis have to be coupled with yet another weapon in the arsenal of modern biology: genomics analysis. Modern molecular biology techniques for the manipulation of RNA and DNA, and the advent of the polymerase chain reaction (PCR), now enable us to build a genomic library from a pair of venom glands and subsequently extract information on the components of the venom by sequencing the resultant cDNA library. Although the predicted venom composition may be somewhat removed from reality, as some genomic sequences may not be expressed, the combination of cDNA analysis and mass spectrometry of secretion products represents a formidable tool to analyze venom contents. The presence of sequences predicted from the cDNA analyses can be verified by mass spectrometry, which can also reveal the presence of possible post-translational modifications.

In this report, we demonstrate the application of this line of reasoning to two of the most potent spider venoms known to man: those from the Australian funnel-web spiders Atrax robustus and Hadronyche versuta (Araneae: Opisthothelae: Mygalomorphae: Hexathelidae). Successive analysis of the crude venoms and their chromatographic fractions by MALDI–TOF MS demonstrates the utility of chromatographic separation as well as mass spectrometry in unraveling the incredible complexity of these venoms. Parallel analysis of venom gland-derived cDNA sequences and mass spectrometric data validate the combined approach. We demonstrate that these venoms are much more complex than previously anticipated, with each venom comprising many hundreds of peptides.

Section snippets

Venom collection

Male Sydney funnel-web spiders (A. robustus) and female Blue Mountain funnel-web spiders (H. versuta) were collected from the Sydney metropolitan area and from the Blue Mountains to the west of Sydney, respectively. Crude venom was ‘milked’ by direct aspiration from the chelicerae of aggravated spiders. Venom was collected into silanized glass pipettes, eluted by rinsing the pipettes with 0.1% (v/v) trifluoroacetic acid (TFA), and immediately freeze-dried. Prior to separation by HPLC, the venom

Crude venom analysis

The HPLC chromatograms of venom from male A. robustus (Ar) (Fig. 1A) and female H. versuta (Hv) (Fig. 1B) were similar to those reported previously for these spiders (Wang et al., 1999, Alewood et al., 2003). The complexity of the crude venom appears similar to that of other mygalomorph spiders (Escoubas et al., 1997, Escoubas et al., 1998, Balaji et al., 2000). A total of 28 and 44 fractions covering the entire elution profile were manually collected for the male A. robustus and female H.

Discussion

We have analyzed the venom proteome of two species of Australian funnel-web spider using an LC-MALDI approach. The venoms were first fractionated using RP-HPLC then individual fractions were analyzed offline using MALDI–TOF. The results revealed that these venoms are much more complex than previously realized, with a total of 633 and 1018 peptide masses detected in the venom of male A. robustus and female H. versuta, respectively. In terms of the total number of peptide components, these venoms

Acknowledgements

We are indebted to Dr G.M. Nicholson for help in spider and venom collection and for a critical reading of this manuscript.

References (36)

Cited by (170)

  • CAP superfamily proteins from venomous animals: Who we are and what to do?

    2022, International Journal of Biological Macromolecules
  • Skeletal Muscle Toxicity Biomarkers

    2019, Biomarkers in Toxicology
View all citing articles on Scopus
View full text