Molecular evolution, functional variation, and proposed nomenclature of the gene family that includes sphingomyelinase D in sicariid spider venoms

Abstract

The venom enzyme sphingomyelinase D (SMase D) in the spider family Sicariidae (brown or fiddleback spiders [Loxosceles] and six-eyed sand spiders [Sicarius]) causes dermonecrosis in mammals. SMase D is in a gene family with multiple venom-expressed members that vary in functional specificity. We analyze molecular evolution of this family and variation in SMase D activity among crude venoms using a data set that represents the phylogenetic breadth of Loxosceles and Sicarius. We isolated a total of 190 nonredundant nucleotide sequences encoding 168 nonredundant amino acid sequences of SMase D homologs from 21 species. Bayesian phylogenies support two major clades that we name alpha and beta, within which we define seven and three subclades, respectively. Sequences in the alpha clade are exclusively from New World Loxosceles and Loxosceles rufescens and include published genes for which expression products have SMase D and dermonecrotic activity. The beta clade includes paralogs from New World Loxosceles that have no, or reduced, SMase D and no dermonecrotic activity and also paralogs from Sicarius and African Loxosceles of unknown activity. Gene duplications are frequent, consistent with a birth-and-death model, and there is evidence of purifying selection with episodic positive directional selection. Despite having venom-expressed SMase D homologs, venoms from New World Sicarius have reduced, or no, detectable SMase D activity, and Loxosceles in the Southern African spinulosa group have low SMase D activity. Sequence conservation mapping shows >98% conservation of proposed catalytic residues of the active site and around a plug motif at the opposite end of the TIM barrel, but alpha and beta clades differ in conservation of key residues surrounding the apparent substrate binding pocket. Based on these combined results, we propose an inclusive nomenclature for the gene family, renaming it SicTox, and discuss emerging patterns of functional diversification.

FIG. 1.—

Relationships of species from which SicTox genes and venoms were included in our analysis. This is a summary composite from analyses of Binford et al. (2008).

FIG. 2.—

Tree topology resulting from Bayesian analyses. Posterior probabilities are labeled above branches with a * indicating support >0.95. Labels below the branch indicate bootstrap support from NJ analyses with 1,000 bootstrap replicates with a * indicating >95%. The proposed SicTox nomenclature delineating major clades on the tree are indicated by labels next to branches. The terminal names indicate the SicTox nomenclature for terminal groups (summarized in supplemental table 1, Supplementary Material online). Colored branches distinguish species groups as delineated in tables 1 and 2. Triangles below branches illustrate branches that had ω > 1 in free-ratio codon likelihood analyses.

FIG. 3.—

One-dimensional gel electrophoresis separations of 5 μg crude venoms from species of (a) Loxosceles; (b) Sicarius and Drymusa. Gels (12.5% acrylamide) were stained with silver nitrate; (c) repeated separations of crude venoms from select species loaded with 3, 5, and 7 μg of total protein to illustrate the effect of uneven loading of total venom protein on visibility of bands. The region of origin of species is indicated. *Loxosceles rufescens is native to the Mediterranean.

FIG. 4.—

Fluorescence intensity (proportional to SMase D activity) resulting from assays of crude venom: (a) fluorescence measured from reactions that contained 0.5 or 1.0 μg of crude venom; (b) fluorescence from reactions that contained 10 or 50 μg of venom with and without sphingomyelin (SM) as the substrate. Some samples were not analyzed under all conditions because of low venom availability. Buffer was substituted for venom in blank samples. The region of origin is indicated. *Loxosceles rufescens is native to the Mediterranean.

FIG. 5.—

(a) Sequence conservation levels mapped onto the surface of Loxosceles laeta SMaseI (PDB ID 1XX1; chain A) for all spider sequences (left), the α clade (middle) and the β clade (right), for the top of the barrel containing the active site (top) and the bottom containing the conserved plug motif (bottom). Surface residues showing 60% conservation or less are shown in blue, whereas those showing 100% are shown in red. Intermediate levels of conservation show a range of color from blue to red, with 80% conservation appearing as white. Note that the levels of conservation observed in the α clade are higher than for the β clade. (b) Surface residues in and around the active site cleft showing >98% conservation (green, orange, and yellow groups) or clade-specific conservation (distal pocket residues in salmon). (c) Clade-specific sequence conservation patterns in the distal binding pocket. The residues shown for the α clade are >95% conserved within this group. Residue types in other clades that differ from the conserved residues in the α clade are underlined.