In this work, we describe the discovery and characterization of a distinct branch of the conantokin family that is biochemically divergent from other conantokins. Conantokins-E and -P both have a large disulfide loop, with two γ-carboxyglutamate residues within the loop, a feature not found in any previously characterized member of the family. The two peptides, which are highly similar in amino acid sequence, are likely to exhibit similar biological activity. This novel group of conantokins is phylogenetically delineated; the two closely-related but distinctive species of fish hunting Conus that are the source of the peptides are divergent from the other fish hunting cone snail species (see ) from which conantokins have previously been characterized.
It had previously been shown for other Conus
peptide families that distinctive peptide toxins are present in C. purpurascens
and C. ermineus
. In the A superfamily, both Conus purpurascens
and Conus erminius
express the αA-conontoxins, a novel group of nicotinic antagonists (Hopkins et al., 1995
; Jacobsen et al., 1997
). The peptides in Conus purpurascens
and C. ermineus
venoms (e.g. αA-PIVA and αA-EIVA) although divergent in sequence, are more structurally similar to each other than to any other peptides in the large, well-characterized A-superfamily (Santos et al., 2004
), which is distributed across the entire genus. The work presented here shows that the conantokin family peptides from these Conus
species, predicted by the analysis of cDNA clones, are similarly distinctive when compared to other conantokins (see ).
We establish that despite the divergence in biochemical properties, conantokin-P is a bona fide conantokin; we have clearly demonstrated that like other conantokins, it is an antagonist of the NMDA receptor (although it more strongly discriminates against the NR2C and NR2D subtypes than do the other conantokins tested).
The presence of conantokins targeting NMDA receptors in three different clades of fish hunting cone snails was unexpected. It is easy to rationalize why venom components such as the α-conotoxins that inhibit the major postsynaptic receptor at the neuromuscular junction (McIntosh et al., 1994
; McIntosh et al., 1999
; Olivera, 1997
), ω- conotoxins that inhibit presynaptic calcium channels (Hillyard et al., 1992
; Terlau et al., 1996
) or μ-conotoxins that target voltage gated sodium channels (Cruz et al., 1985
; Terlau and Olivera, 2004
) responsible for action potentials are generally found in the venoms of Conus
species that prey on fish (Olivera and Cruz, 2001
) — these targets are essential molecular components for neuromuscular transmission in all vertebrates, and their inhibition leads to paralysis. Why peptides that antagonize the NMDA receptor are broadly found among fish hunting cone snails is far less apparent.
In contrast to mammalian systems, where no clear demonstration of functional NMDA receptors outside the central nervous system has been reported, there are peripheral glutamatergic circuits in fish that are presumably the relevant physiological targets of conantokins and other peptides found in Conus venoms that affect glutamate receptors. One possibility is the lateral line circuitry of fish, used for water movement detection. The broad distribution of peptides that target NMDA receptors strongly suggests that one of the adaptations of cone snails that has evolved for effectively capturing fish is to target this, or some other yet undefined glutamatergic circuitry in the peripheral nervous systems of teleost fish.
The acceleration in the rate of disulfide bond formation in Conantokin-P when Ca++
is present is reminiscent of what was observed for the spasmodic peptides from Conus textile
in the P-superfamily (Bulaj et al., 2003
). It was established in that instance that the observed acceleration in disulfide bond formation was due to the presence of two γ-carboxyglutamate residues in the spasmodic peptide from Conus textile
; the homologous peptide from Conus gloriamarus
, nearly identical in sequence but lacking the two Gla residues did not exhibit an accelerated rate of disulfide formation in the presence of Ca++
. The data we present in is consistent with a wider role of γ-carboxyglutamate residues in the oxidative folding of conopeptides.
We note that the likely order of occurrence of these two post-translational modifications is consistent with this suggestion: the γ-glutamyl carboxylase enzyme is present in the Conus
endoplasmic reticulum membrane, (Bandyopadhyay et al., 1998
; Bandyopadhyay et al., 2002
; Stnley et al., 1997
) and would therefore be the first post-translational modification acting on a newly synthesized polypeptide chain destined for secretion; in contrast, the disulfide isomerase is present in the lumen of the endoplasmic reticulum (Bulaj et al., 2003
; Bulaj and Olivera, 2008
; Buczek et al., 2004
). Thus, disulfide bond formation would be preceded by γ-carboxylation. Indeed, the synergistic interaction between the two post-translational events, γ-carboxylation of glutamate residues and the oxidation of Cys residues to form disulfide bonds has been suggested to be an evolutionarily ancient one that has been recapitulated in the Conus
peptide system because of the high density of disulfide bonds in many short conopeptides. How widely distributed the role of γ-carboxyglutamate in folding will prove to be, remains to be investigated.