Limits...
The venom-gland transcriptome of the eastern coral snake (Micrurus fulvius) reveals high venom complexity in the intragenomic evolution of venoms.

Margres MJ, Aronow K, Loyacano J, Rokyta DR - BMC Genomics (2013)

Bottom Line: New World coral snakes (Elapidae) are represented by three genera and over 120 species and subspecies that are capable of causing significant human morbidity and mortality, yet coral-snake venom composition is poorly understood in comparison to that of Old World elapids.Toxins exhibited high levels of heterozygosity relative to nontoxins, and overdominance may favor gene duplication leading to the fixation of advantageous alleles.Toxin gene duplication may be driven by heterozygote advantage, as the frequency of polymorphic toxin loci was significantly higher than that of nontoxins.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Biological Science, Florida State University, Tallahassee, FL 32306-4295, USA.

ABSTRACT

Background: Snake venom is shaped by the ecology and evolution of venomous species, and signals of positive selection in toxins have been consistently documented, reflecting the role of venoms as an ecologically critical phenotype. New World coral snakes (Elapidae) are represented by three genera and over 120 species and subspecies that are capable of causing significant human morbidity and mortality, yet coral-snake venom composition is poorly understood in comparison to that of Old World elapids. High-throughput sequencing is capable of identifying thousands of loci, while providing characterizations of expression patterns and the molecular evolutionary forces acting within the venom gland.

Results: We describe the de novo assembly and analysis of the venom-gland transcriptome of the eastern coral snake (Micrurus fulvius). We identified 1,950 nontoxin transcripts and 116 toxin transcripts. These transcripts accounted for 57.1% of the total reads, with toxins accounting for 45.8% of the total reads. Phospholipases A(2) and three-finger toxins dominated expression, accounting for 86.0% of the toxin reads. A total of 15 toxin families were identified, revealing venom complexity previously unknown from New World coral snakes. Toxins exhibited high levels of heterozygosity relative to nontoxins, and overdominance may favor gene duplication leading to the fixation of advantageous alleles. Phospholipase A(2) expression was uniformly distributed throughout the class while three-finger toxin expression was dominated by a handful of transcripts, and phylogenetic analyses indicate that toxin divergence may have occurred following speciation. Positive selection was detected in three of the four most diverse toxin classes, suggesting that venom diversification is driven by recurrent directional selection.

Conclusions: We describe the most complete characterization of an elapid venom gland to date. Toxin gene duplication may be driven by heterozygote advantage, as the frequency of polymorphic toxin loci was significantly higher than that of nontoxins. Diversification among toxins appeared to follow speciation reflecting species-specific adaptation, and this divergence may be directly related to dietary shifts and is suggestive of a coevolutionary arms race.

Show MeSH

Related in: MedlinePlus

The venom-gland transcriptome of Micrurus fulvius  was dominated by phospholipases A2 and three-finger toxins. Toxin gene expression was dominated by phospholipases A2 (PLA2s) and three-finger toxins (3FTxs). Full-length transcripts accounted for 57.1% of the total reads; toxin sequences accounted for 45.8% of the total reads. PLA2s and 3FTxs represent both the most abundant and most diverse toxin classes identified; 31 PLA2 clusters accounted for 64.9% of the toxin reads, and 15 3FTx clusters accounted for 21.1% of the toxin reads. Toxin sequences accounted for 10.4% more of the total reads in M. fulvius than in C. adamanteus, while nontoxins in C. adamanteus accounted for more than twice the total read percentage than in M. fulvius. The venom of M. fulvius was dominated by small neurotoxic components while the venom of C. adamanteus was dominated by larger hemorrhage-inducing proteins, suggesting that the transcriptional effort expended on toxins versus nontoxins may differ between venoms dominated by high-molecular weight components and venoms dominated by smaller proteins, with small-component venoms expressing toxins at much higher levels relative to nontoxin production.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3750283&req=5

Figure 2: The venom-gland transcriptome of Micrurus fulvius was dominated by phospholipases A2 and three-finger toxins. Toxin gene expression was dominated by phospholipases A2 (PLA2s) and three-finger toxins (3FTxs). Full-length transcripts accounted for 57.1% of the total reads; toxin sequences accounted for 45.8% of the total reads. PLA2s and 3FTxs represent both the most abundant and most diverse toxin classes identified; 31 PLA2 clusters accounted for 64.9% of the toxin reads, and 15 3FTx clusters accounted for 21.1% of the toxin reads. Toxin sequences accounted for 10.4% more of the total reads in M. fulvius than in C. adamanteus, while nontoxins in C. adamanteus accounted for more than twice the total read percentage than in M. fulvius. The venom of M. fulvius was dominated by small neurotoxic components while the venom of C. adamanteus was dominated by larger hemorrhage-inducing proteins, suggesting that the transcriptional effort expended on toxins versus nontoxins may differ between venoms dominated by high-molecular weight components and venoms dominated by smaller proteins, with small-component venoms expressing toxins at much higher levels relative to nontoxin production.

Mentions: The 1,950 nontoxin transcripts accounted for 11.3% of the reads. The 116 full-length toxin transcripts were grouped into 75 clusters (Figure 1B, Table 1) and accounted for 45.8% of the reads. In total, we accounted for 57.1% of the reads (Figure 2), comparable to the amount identified for C. adamanteus using a similar approach [16]. While the overall percentage of reads mapping to identified transcripts was similar for M. fulvius and C. adamanteus, toxin expression levels were quite different (Figure 2). The toxin transcripts identified in M. fulvius accounted for nearly half of the total sequencing reads (45.8%) while the toxin transcripts in C. adamanteus accounted for approximately one-third (35.4%) of the total reads (Figure 2) [16]. The numbers and abundances of nontoxin coding sequences were much lower in M. fulvius than in C. adamanteus, despite an increase in assembly effort (e.g., the addition of the reference-based assembly), as 1,950 nontoxin transcripts accounted for 11.3% of the total reads in M. fulvius while 2,879 nontoxin transcripts accounted for 27.5% of the total reads in C. adamanteus (Figure 2). The venom-gland transcriptome of C. adamanteus was characterized by large, hemorrhage-inducing toxins such as snake venom metalloproteinases (SVMPs), proteins that presumably require extensive downstream processing by nontoxin machinery prior to becoming mature, active toxins [16]. The vast majority of highly expressed nontoxin sequences identified in the transcriptome of C. adamanteus were involved with proteostasis (e.g., protein folding, degradation, and transport) [16], and the reduction in the expression levels of nontoxin transcripts in M. fulvius could potentially reflect a difference in the maintenance, production, and folding requirements of the venom components of each species. The venom of M. fulvius was dominated by three-finger toxins (3FTxs) and phospholipases A2 (PLA2s), relatively short toxins that may not require the degree of downstream processing needed by larger toxins to become functional. This suggests that venoms dominated by smaller proteins differ in the transcriptional effort expended on toxins relative to nontoxins in comparison to venoms characterized by high-molecular weight components, with small-component venoms expressing toxins at much higher levels relative to nontoxin production. The largely proteinaceous composition of venom makes it metabolically costly to produce [19], and a reduction in the machinery necessary for the production of functional toxic proteins may confer an energetic benefit to species expressing smaller peptides and enzymes. Simple, smaller toxins have a reduced mutational target relative to larger proteins as a function of sequence length, potentially reducing the ability to evolve effective counterdefenses to resistance development in frequently envenomed prey [6,7] and predators [1] where more complex venoms would be advantageous. However, as our hypotheses are based on a comparison between a single representative of each family, sequencing additional members of Viperidae and Elapidae are needed to test whether these putative differences in transcriptional effort are fixed or unique to M. fulvius and/or C.adamanteus.


The venom-gland transcriptome of the eastern coral snake (Micrurus fulvius) reveals high venom complexity in the intragenomic evolution of venoms.

Margres MJ, Aronow K, Loyacano J, Rokyta DR - BMC Genomics (2013)

The venom-gland transcriptome of Micrurus fulvius  was dominated by phospholipases A2 and three-finger toxins. Toxin gene expression was dominated by phospholipases A2 (PLA2s) and three-finger toxins (3FTxs). Full-length transcripts accounted for 57.1% of the total reads; toxin sequences accounted for 45.8% of the total reads. PLA2s and 3FTxs represent both the most abundant and most diverse toxin classes identified; 31 PLA2 clusters accounted for 64.9% of the toxin reads, and 15 3FTx clusters accounted for 21.1% of the toxin reads. Toxin sequences accounted for 10.4% more of the total reads in M. fulvius than in C. adamanteus, while nontoxins in C. adamanteus accounted for more than twice the total read percentage than in M. fulvius. The venom of M. fulvius was dominated by small neurotoxic components while the venom of C. adamanteus was dominated by larger hemorrhage-inducing proteins, suggesting that the transcriptional effort expended on toxins versus nontoxins may differ between venoms dominated by high-molecular weight components and venoms dominated by smaller proteins, with small-component venoms expressing toxins at much higher levels relative to nontoxin production.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3750283&req=5

Figure 2: The venom-gland transcriptome of Micrurus fulvius was dominated by phospholipases A2 and three-finger toxins. Toxin gene expression was dominated by phospholipases A2 (PLA2s) and three-finger toxins (3FTxs). Full-length transcripts accounted for 57.1% of the total reads; toxin sequences accounted for 45.8% of the total reads. PLA2s and 3FTxs represent both the most abundant and most diverse toxin classes identified; 31 PLA2 clusters accounted for 64.9% of the toxin reads, and 15 3FTx clusters accounted for 21.1% of the toxin reads. Toxin sequences accounted for 10.4% more of the total reads in M. fulvius than in C. adamanteus, while nontoxins in C. adamanteus accounted for more than twice the total read percentage than in M. fulvius. The venom of M. fulvius was dominated by small neurotoxic components while the venom of C. adamanteus was dominated by larger hemorrhage-inducing proteins, suggesting that the transcriptional effort expended on toxins versus nontoxins may differ between venoms dominated by high-molecular weight components and venoms dominated by smaller proteins, with small-component venoms expressing toxins at much higher levels relative to nontoxin production.
Mentions: The 1,950 nontoxin transcripts accounted for 11.3% of the reads. The 116 full-length toxin transcripts were grouped into 75 clusters (Figure 1B, Table 1) and accounted for 45.8% of the reads. In total, we accounted for 57.1% of the reads (Figure 2), comparable to the amount identified for C. adamanteus using a similar approach [16]. While the overall percentage of reads mapping to identified transcripts was similar for M. fulvius and C. adamanteus, toxin expression levels were quite different (Figure 2). The toxin transcripts identified in M. fulvius accounted for nearly half of the total sequencing reads (45.8%) while the toxin transcripts in C. adamanteus accounted for approximately one-third (35.4%) of the total reads (Figure 2) [16]. The numbers and abundances of nontoxin coding sequences were much lower in M. fulvius than in C. adamanteus, despite an increase in assembly effort (e.g., the addition of the reference-based assembly), as 1,950 nontoxin transcripts accounted for 11.3% of the total reads in M. fulvius while 2,879 nontoxin transcripts accounted for 27.5% of the total reads in C. adamanteus (Figure 2). The venom-gland transcriptome of C. adamanteus was characterized by large, hemorrhage-inducing toxins such as snake venom metalloproteinases (SVMPs), proteins that presumably require extensive downstream processing by nontoxin machinery prior to becoming mature, active toxins [16]. The vast majority of highly expressed nontoxin sequences identified in the transcriptome of C. adamanteus were involved with proteostasis (e.g., protein folding, degradation, and transport) [16], and the reduction in the expression levels of nontoxin transcripts in M. fulvius could potentially reflect a difference in the maintenance, production, and folding requirements of the venom components of each species. The venom of M. fulvius was dominated by three-finger toxins (3FTxs) and phospholipases A2 (PLA2s), relatively short toxins that may not require the degree of downstream processing needed by larger toxins to become functional. This suggests that venoms dominated by smaller proteins differ in the transcriptional effort expended on toxins relative to nontoxins in comparison to venoms characterized by high-molecular weight components, with small-component venoms expressing toxins at much higher levels relative to nontoxin production. The largely proteinaceous composition of venom makes it metabolically costly to produce [19], and a reduction in the machinery necessary for the production of functional toxic proteins may confer an energetic benefit to species expressing smaller peptides and enzymes. Simple, smaller toxins have a reduced mutational target relative to larger proteins as a function of sequence length, potentially reducing the ability to evolve effective counterdefenses to resistance development in frequently envenomed prey [6,7] and predators [1] where more complex venoms would be advantageous. However, as our hypotheses are based on a comparison between a single representative of each family, sequencing additional members of Viperidae and Elapidae are needed to test whether these putative differences in transcriptional effort are fixed or unique to M. fulvius and/or C.adamanteus.

Bottom Line: New World coral snakes (Elapidae) are represented by three genera and over 120 species and subspecies that are capable of causing significant human morbidity and mortality, yet coral-snake venom composition is poorly understood in comparison to that of Old World elapids.Toxins exhibited high levels of heterozygosity relative to nontoxins, and overdominance may favor gene duplication leading to the fixation of advantageous alleles.Toxin gene duplication may be driven by heterozygote advantage, as the frequency of polymorphic toxin loci was significantly higher than that of nontoxins.

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Biological Science, Florida State University, Tallahassee, FL 32306-4295, USA.

ABSTRACT

Background: Snake venom is shaped by the ecology and evolution of venomous species, and signals of positive selection in toxins have been consistently documented, reflecting the role of venoms as an ecologically critical phenotype. New World coral snakes (Elapidae) are represented by three genera and over 120 species and subspecies that are capable of causing significant human morbidity and mortality, yet coral-snake venom composition is poorly understood in comparison to that of Old World elapids. High-throughput sequencing is capable of identifying thousands of loci, while providing characterizations of expression patterns and the molecular evolutionary forces acting within the venom gland.

Results: We describe the de novo assembly and analysis of the venom-gland transcriptome of the eastern coral snake (Micrurus fulvius). We identified 1,950 nontoxin transcripts and 116 toxin transcripts. These transcripts accounted for 57.1% of the total reads, with toxins accounting for 45.8% of the total reads. Phospholipases A(2) and three-finger toxins dominated expression, accounting for 86.0% of the toxin reads. A total of 15 toxin families were identified, revealing venom complexity previously unknown from New World coral snakes. Toxins exhibited high levels of heterozygosity relative to nontoxins, and overdominance may favor gene duplication leading to the fixation of advantageous alleles. Phospholipase A(2) expression was uniformly distributed throughout the class while three-finger toxin expression was dominated by a handful of transcripts, and phylogenetic analyses indicate that toxin divergence may have occurred following speciation. Positive selection was detected in three of the four most diverse toxin classes, suggesting that venom diversification is driven by recurrent directional selection.

Conclusions: We describe the most complete characterization of an elapid venom gland to date. Toxin gene duplication may be driven by heterozygote advantage, as the frequency of polymorphic toxin loci was significantly higher than that of nontoxins. Diversification among toxins appeared to follow speciation reflecting species-specific adaptation, and this divergence may be directly related to dietary shifts and is suggestive of a coevolutionary arms race.

Show MeSH
Related in: MedlinePlus