Rapid expansion of the protein disulfide isomerase gene family facilitates the folding of venom peptides

Abstract

Formation of correct disulfide bonds in the endoplasmic reticulum is a crucial step for folding proteins destined for secretion. Protein disulfide isomerases (PDIs) play a central role in this process. We report a previously unidentified, hypervariable family of PDIs that represents the most diverse gene family of oxidoreductases described in a single genus to date. These enzymes are highly expressed specifically in the venom glands of predatory cone snails, animals that synthesize a remarkably diverse set of cysteine-rich peptide toxins (conotoxins). Enzymes in this PDI family, termed conotoxin-specific PDIs, significantly and differentially accelerate the kinetics of disulfide-bond formation of several conotoxins. Our results are consistent with a unique biological scenario associated with protein folding: The diversification of a family of foldases can be correlated with the rapid evolution of an unprecedented diversity of disulfide-rich structural domains expressed by venomous marine snails in the superfamily Conoidea.

Keywords: cone snail venom; conotoxins; gene expansion; peptide folding; protein disulfide isomerase.

Fig. 1.

Identification of a previously unidentified PDI sequence (csPDI) in the venom gland of C. geographus (GenBank accession no. KT874567). Multiple sequence alignment with canonical PDI from the same species (GenBank accession no. KT874559) and three additional species [C. betulinus (ADZ76593), C. marmoreus (ABF48564), and C. eburneus (ADZ76591)] identifies regions of divergence between csPDI and PDIs (white, 100% identity; light gray, 99–80% identity; dark gray, 80–60% identity; black: <60% identity). The alignment was performed in Geneious by using the Blosum62 similarity option for coloring (Version 8.1.3). csPDI and PDI sequences share 65–66% identity. Thioredoxin domain organization is depicted above the sequences and was predicted by using known boundaries for human PDI (31). Signal sequences (gray bar above sequence) were predicted by using InterProScan (32). The C terminus containing ER-retention motifs is also depicted with a gray bar. Active site CGHC motifs are boxed.

Fig. 2.

Phylogenetic analysis of full-length PDI and csPDI protein sequences supports the presence of two gene families originating from an ancestral gene duplication event (black arrow, posterior probability: 1). Diversity and genetic variance for the csPDI family are apparent by more than one distinct csPDI sequence per species and long branch lengths, respectively. Two groups (the primitive and complex groups) resolved within the csPDI branch and correlate with molecular phylogenetics analyses (14). Amino acids of the active site motif are provided for sequences with divergent active sites. Arrows indicate C. geographus sequences selected for subsequent functional characterization. Bayesian tree was constructed using MrBayes (Version 3.2.2; ref. 33) with two runs each of four Markov chains sampling every 200 generations. The log-likelihood score stabilized after 1,100,000 generations. The consensus tree was calculated after omitting the first 25% of the samples as burn-in.

Fig. 3.

Analysis of the venom gland proteome of C. geographus shows high abundances for PDI and csPDIs as determined by 2D gel electrophoresis and subsequent mass spectrometric protein identification. Data deposited in the original study (15) were revisited and examined for mass spectrometric peptide hits that matched PDI and csPDI sequences obtained in the present study. Gel spots identified as PDI and csPDIs are depicted and correlate with predicted molecular weights (MW) and isoelectric points (pI). The number of total and unique peptide matches obtained for PDI and different members of the csPDI family are provided (score > 99 using Protein Pilot; Version 3.0; AB SCIEX). Sequences and position of matched peptides onto the full-length sequences are provided in SI Appendix, Fig. S5. Reproduced from ref. , copyright the American Society for Biochemistry and Molecular Biology.

Fig. 4.

(A and B) Sequence variation in cone snail PDIs (A) and csPDIs (B) mapped onto a representation of the crystal structure of full-length human PDI (Protein Data Bank ID code 4EKZ). Multiple sequence alignments of PDIs and csPDIs were used to assign a variation score for each position in the alignment. This score was then converted to a red–white color range, where darker color indicates higher sequence variation. Heavy atoms of active-site cysteines are depicted as space-filling models (gray, C; yellow, S). The four thioredoxin-like domains are indicated, and arrows point to the +2 position C-terminal of the CXXC motifs in the a and a′ domains of the csPDIs, which shows sequence variation only in this group of enzymes and not in the PDIs (see text for details). (C) Residues of the hydrophobic patch of the b′ domain are shown as stick models. Sequence variation is apparent in all but two of these residues (arrows).

Fig. 5.

Oxidative folding of conotoxin substrates in the presence of PDI and two members of the csPDI family from C. geographus. Sequences of the three conotoxin substrates tested are shown with their names, molecular targets, and disulfide connectivities. Amino acids: Z, pyroglutamate; O, hydroxyproline; *C-terminal amidation. (A and B) Folding assays for ω-GVIA (A) and μ-SmIIIA (B) were carried out at room temperature in the absence and presence of 2 μM enzyme in 100 mM Tris⋅HCl (pH 7.5), 1 mM EDTA, 0.4 mM reduced glutathione, and 0.2 mM oxidized glutathione. (C) Folding of δ-PVIA was performed at 4 °C and in the presence of 1% Tween-20. Reactions were initiated by adding 20 μM reduced toxin, quenched at different time points with formic acid (final 10% vol/vol), and analyzed by reverse-phase chromatography on a C₁₈ column. Chromatograms of reactions without enzyme (black), with PDI (green), and with csPDI_GH/GH (blue) are shown in A1, B1, and C1, respectively. The area under the curve was determined for the native, fully folded toxin and plotted against the reaction time (A2, B2, and C2). Half-time for the appearance of folded toxins (95% confidence values) was calculated in Prism (Version 6.0e; GraphPad) and is shown in A3, B3, and C3. Reactions that were significantly different from no-enzyme controls are indicated. *P < 0.01 (unpaired t test with Welch's correction).