RESULTS AND DISCUSSION: The toxin transcripts showed high redundancy (41-82% of the total mRNA expression) and comprised 23 gene families expressed in NK-M and NK-T, respectively (22 gene families were co-expressed). Among the venom genes, three-finger toxins (3FTxs) predominated in the expression, with multiple sequences noted. Comparative analysis and selection study revealed that 3FTxs are genetically conserved between the geographical specimens whilst demonstrating distinct differential expression patterns, implying gene up-regulation for selected principal toxins, or alternatively, enhanced transcript degradation or lack of transcription of certain traits. One of the striking features that elucidates the inter-geographical venom variation is the up-regulation of α-neurotoxins (constitutes ∼80.0% of toxin's fragments per kilobase of exon model per million mapped reads (FPKM)), particularly the long-chain α-elapitoxin-Nk2a (48.3%) in NK-T but only 1.7% was noted in NK-M. Instead, short neurotoxin isoforms were up-regulated in NK-M (46.4%). Another distinct transcriptional pattern observed is the exclusively and abundantly expressed cytotoxin CTX-3 in NK-T. The findings suggested correlation with the geographical variation in proteome and toxicity of the venom, and support the call for optimising antivenom production and use in the region. Besides, the current study uncovered full and partial sequences of numerous toxin genes from N. kaouthia which have not been reported hitherto; these include N. kaouthia-specific l-amino acid oxidase (LAAO), snake venom serine protease (SVSP), cystatin, acetylcholinesterase (AChE), hyaluronidase (HYA), waprin, phospholipase B (PLB), aminopeptidase (AP), neprilysin, etc. Taken together, the findings further enrich the snake toxin database and provide deeper insights into the genetic diversity of cobra venom toxins.

BIOLOGICAL SIGNIFICANCE: Laticauda colubrina (yellow-lipped sea krait) is a widely distributed, semi-aquatic venomous snake species. The venom proteome at the level of protein family is unsophisticated and consistent with its restricted prey selection. Nonetheless, the subproteomic findings revealed geographical variability of the venom for this widely distributed species. In contrast to two previous reports, the results for the Balinese L. colubrina venom showed that LNTX Neurotoxin a and Neurotoxin b were co-existent while the PLA2 lethal subtype (PLA-II) was undetected by means of LCMS/MS and by in vivo assay. This is an observable trait of L. colubrina considered divergent from specimens previously studied for the Philippines and the Solomon Islands. The stark geographical variation might be reflective of trophic adaptation following evolutionary arms race between the snake and the prey (eels) in different localities. The preferred trait would likely propagate and remain significant within the geographical population, since the strong behaviour of site fidelity in the species would have minimized gene flow between distant populations. Meanwhile, the in vivo neutralization study verified that the efficacy of the heterologous Sea Snake Antivenom (Australian product) is attributable to the cross-neutralization of SNTX and LNTX, two principal lethal toxins that made up the bulk of L. colubrina venom proteins. The findings also implied that L. colubrina, though could be evolutionarily more related to the terrestrial elapids, has evolved a much streamlined, neurotoxin- and PLA2-predominated venom arsenal, with major antigenicity shared among the true sea snakes and the Australo-Papuan elapids. The findings enrich our current understanding of the complexity of L. colubrina venom and the neutralizing spectrum of antivenom against the principal toxins from this unique elapid lineage.