Distinctive Distribution of Secretory Phospholipases A2 in the Venoms of Afro-Asian Cobras (Subgenus: Naja, Afronaja, Boulengerina and Uraeus)

The protein abundances of phospholipases A2 in cobra venom proteomes appear to vary among cobra species. To determine the unique distribution of snake venom phospholipases A2 (svPLA2) in the cobras, the svPLA2 activities for 15 cobra species were examined with an acidimetric and a colorimetric assay, using egg yolk suspension and 4-nitro-3-octanoyloxy benzoic acid (NOBA) as the substrate. The colorimetric assay showed significant correlation between svPLA2 enzymatic activities with the svPLA2 protein abundances in venoms. High svPLA2 activities were observed in the venoms of Asiatic spitting cobras (Naja sputatrix, Naja sumatrana) and moderate activities in Asiatic non-spitters (Naja naja, Naja atra, Naja kaouthia), African spitters (subgenus Afronaja), and forest cobra (subgenus Boulengerina). African non-spitting cobras of subgenus Uraeus (Naja haje, Naja annulifera, Naja nivea, Naja senegalensis) showed exceptionally low svPLA2 enzymatic activities. The negligible PLA2 activity in Uraeus cobra venoms implies that PLA2 may not be ubiquitous in all snake venoms. The svPLA2 in cobra envenoming varies depending on the cobra species. This may potentially influence the efficacy of cobra antivenom in specific use for venom neutralization.


Introduction
Phospholipases A 2 (PLA 2 ) (EC 3.1.1.4) are enzymes that hydrolyze glycerophospholipids to lysophospholipids and fatty acids. The first snake venom PLA 2 enzymes were purified from the venoms of Naja naja and Naja tripudians as hemolysin due to their ability to lyse the phospholipid membranes of red blood cells [1]. Since then, various snake venom-derived PLA 2 (svPLA 2 ) have been characterized and shown to exist in virtually all venoms from the two major families of venomous snakes: Elapidae and Viperidae [2]. Homologous svPLA 2 are especially abundant and diverse in the Asiatic elapids, including cobras, coral snakes, kraits, and some sea snake species [3][4][5][6], implying that the enzyme plays an essential role in the function of the venom.

PLA2 Enzymatic Activities (Colorimetric Assay)
The PLA2 activities of the cobra venoms were measured using a colorimetric assay. The enzymatic activity of the venoms in hydrolyzing the non-micellar substrate (NOBA) over time is shown in Figure S2. High PLA2 activities were noted in the venoms of N. sputatrix (rate = 109.69 nmol/min/mg) and N. sumatrana (rate = 82.11 nmol/min/mg), followed by other species in the subgenera Naja (rate = 33.21-42.26 nmol/min/mg), Afronaja (rate = 45.15-53.82 nmol/min/g), and Boulengerina (rate = 48.03 nmol/min/mg). In comparison, cobra venoms of the Uraeus subgenus showed much lower PLA2 activities (rate = 7.12-13.52 nmol/min/mg) ( Figure 2).    Figure 3 shows the correlation between PLA 2 activities in the colorimetric assay and PLA 2 protein abundances in cobra venoms. The PLA 2 protein abundances of 12 cobra species were obtained from published studies that adopted a comparable quantitative approach, in which the protein abundances were estimated based on peak areas of reverse-phase high performance liquid chromatography (HPLC), followed by integration with the relative mass spectral intensity or relative gel band density of PLA 2 eluted (Table S1) [14,19,20,22,25,26,28,34]. The PLA 2 activities measured using the acidimetric assay showed a weak correlation with the PLA 2 abundances (coefficient of determination, R 2 = 0.01, p > 0.05). The colorimetric assay demonstrated a moderate to strong association between the PLA 2 activities and PLA 2 abundances of the cobra venoms studied (R 2 = 0.55, p < 0.01) ( Figure 3). The higher PLA 2 enzymatic activities were observed in the venoms of spitting cobras under the subgenus Naja (N. sputatrix and N. sumatrana) and subgenus Afronaja (African spitters), whose PLA 2 abundances were more than 20% of total venom proteins. The non-spitting Asiatic Naja cobras (N. kaouthia, N. naja and N. atra) had intermediate PLA 2 abundances (12-14%) with moderate PLA 2 enzymatic activities per unit venom. The venom of N. haje (African non-spitting cobra under the Uraeus subgenus), however, showed very low PLA 2 enzymatic activity that was in line with its low PLA 2 content (4%). statistical significance (p < 0.05) is indicated by different lower-case letters at the top of the bar. Bars without any common lowercase letter denote values that were significantly different (p < 0.05). Figure 3 shows the correlation between PLA2 activities in the colorimetric assay and PLA2 protein abundances in cobra venoms. The PLA2 protein abundances of 12 cobra species were obtained from published studies that adopted a comparable quantitative approach, in which the protein abundances were estimated based on peak areas of reverse-phase high performance liquid chromatography (HPLC), followed by integration with the relative mass spectral intensity or relative gel band density of PLA2 eluted (Table S1) [14,19,20,22,25,26,28,34]. The PLA2 activities measured using the acidimetric assay showed a weak correlation with the PLA2 abundances (coefficient of determination, R 2 = 0.01, p > 0.05). The colorimetric assay demonstrated a moderate to strong association between the PLA2 activities and PLA2 abundances of the cobra venoms studied (R 2 = 0.55, p < 0.01) ( Figure 3). The higher PLA2 enzymatic activities were observed in the venoms of spitting cobras under the subgenus Naja (N. sputatrix and N. sumatrana) and subgenus Afronaja (African spitters), whose PLA2 abundances were more than 20% of total venom proteins. The non-spitting Asiatic Naja cobras (N. kaouthia, N. naja and N. atra) had intermediate PLA2 abundances (12-14%) with moderate PLA2 enzymatic activities per unit venom. The venom of N. haje (African non-spitting cobra under the Uraeus subgenus), however, showed very low PLA2 enzymatic activity that was in line with its low PLA2 content (4%). Enzymatic measurement using colorimetric assay for 12 cobra venoms. The relative abundance of PLA2 (% by total venom proteins) in 12 cobra venoms were adopted from published literature for subgenus Naja: N. sputatrix [14], N. naja [19], Naja atra [20], N. kaouthia [22], Naja sumatrana [34]; subgenus Afronaja: N. katiensis, N. mossambica, N. pallida, N. nubiae, and N. nigricollis [26]; subgenus Boulengerina: N. melanoleuca [25]; and subgenus Uraeus: N. haje [28]. Abbreviations:  Figure 3 depicts the correlation between PLA2 activities exhibited by the whole venoms and relative abundances of PLA2 in the venoms. The PLA2 activities were measured using whole venoms and are expressed in nmol/min/mg venom proteins as previously established [29,35]. This allowed the determination of correlation between the PLA2 activities (measured per unit mass of venom) and the relative abundances of PLA2 in the cobra venoms. The higher the PLA2 relative abundance (% of total venom proteins) in a venom, the higher the snake venom PLA2 activity. The PLA2 activity of snake venom can also be expressed as "PLA2 specific activity" [36], where the PLA2 activity of a whole venom is normalized (divided) by the amount of PLA2 in the venom. This measurement is suitable for characterizing the enzymatic activity of purified or isolated PLA2, expressed in the unit of Enzymatic measurement using colorimetric assay for 12 cobra venoms. The relative abundance of PLA 2 (% by total venom proteins) in 12 cobra venoms were adopted from published literature for subgenus Naja: N. sputatrix [14], N. naja [19], Naja atra [20], N. kaouthia [22], Naja sumatrana [34]; subgenus Afronaja: N. katiensis, N. mossambica, N. pallida, N. nubiae, and N. nigricollis [26]; subgenus Boulengerina: N. melanoleuca [25]; and subgenus Uraeus: N. haje [28]. Abbreviations:  Figure 3 depicts the correlation between PLA 2 activities exhibited by the whole venoms and relative abundances of PLA 2 in the venoms. The PLA 2 activities were measured using whole venoms and are expressed in nmol/min/mg venom proteins as previously established [29,35]. This allowed the determination of correlation between the PLA 2 activities (measured per unit mass of venom) and the relative abundances of PLA 2 in the cobra venoms. The higher the PLA 2 relative abundance (% of total venom proteins) in a venom, the higher the snake venom PLA 2 activity. The PLA 2 activity of snake venom can also be expressed as "PLA 2 specific activity" [36], where the PLA 2 activity of a whole venom is normalized (divided) by the amount of PLA 2 in the venom. This measurement is suitable for characterizing the enzymatic activity of purified or isolated PLA 2 , expressed in the unit of nmol/min/mg PLA 2 isolated [36]. In this study, the PLA 2 specific activity is included along with the relevant parameters used in calculating the values ( Table 1). The findings showed that the cobra venoms tested could be generally classified into four groups: (1) Asiatic spitting cobras

Phylogenetics of Cobras in Relation to Venom PLA 2 Activities
The relative PLA 2 enzymatic activities of the 15 cobra venoms (by acidimetric and colorimetric methods) are related to the phylogeny of cobras in Figure 4. The venoms of Asian spitting cobras (N. sputatrix and N. sumatrana) exhibited the highest PLA 2 enzymatic activity tested with the colorimetric method, whereas the highest PLA 2 activity determined by the acidimetric method was observed in the African forest cobra N. melanoleuca venom. The venoms of other cobra species within the subgenera Naja (Asian cobras) and Afronaja (African spitting cobras) showed moderate to high levels of PLA 2 activities in both enzymatic assays. The venoms of cobras within the Uraeus subgenus, representing a monophyletic group of non-spitting African cobras separated from the African forest cobras more recently, showed exceptionally low PLA 2 activities in both acidimetric and colorimetric assays.  . Phylogenic tree relating the venom phospholipase A2 activities of 15 cobra species by subgenera. Blue: acidimetric assay; red: colorimetic assay. The venom PLA2 activity is expressed in a ratio relative to the highest activity detected by acidimetric or colorimetric assay in this study (1.0 implies the highest activity). More intense color indicates higher PLA2 activity. The phylogenetic tree was redrawn with adaptation from phylogenetics of cobras [32]. Naja sumatrana is depicted here as a sister taxon of Naja sputatrix. Spitting cobras are marked with the snake symbol next to their species name.

Discussion
Phospholipase A2 (phosphatidylcholine 2-acylhydrolase) catalyzes the hydrolysis of phosphatidylcholine at the sn-2 ester bond to produce lysophospholipid and free fatty acids. We tested the activities of svPLA2 based on two different types of PLA2 assays [37]: an acidimetric assay that measured the release of proton from fatty acids during the hydrolysis of phosphate ester bond, and a colorimetric assay that measured the amount of chromogenic 4-nitro-3-hydroxybenzoic acid released by the cleavage of PLA2 at the ester bond between the octanol group of NOBA [38]. The activities tested on NOBA showed a better correlation with the PLA2 contents in the cobra venoms. The egg-yolk-based acidimetric assay probably contained less specific substrates (phospholipids, triglycerides) that could be targets of other lytic enzymes in the venoms, or atmospheric carbon dioxide could have interfered with the assay by dissolving into the suspension during the stirring process. On the whole, the enzymatic rates measured by both assays support that PLA2 enzymatic activities vary according to the svPLA2 composition in different cobra species. The enzymatic svPLA2 distribution is unique following a clustering trend among the four subgenera, and notably the remarkable lack of svPLA2 within the Uraeus subgenus.
The enzymatic activities measured in this study for the venoms of African spitting cobras (N. nubiae, N. nigricollis, N. mossambica, N. pallida, N. katiensis) and those from Asia (N. sumatrana, N. sputatrix) were high, consistent with the high abundances of svPLA2 reported previously in these venoms [14,26,34]. Some svPLA2 of Afronaja cobras were shown to be cytotoxic and to have dermonecrotic and myonecrotic activities [39,40]. Some African spitting cobra venoms possess coagulotoxic activity, which were inhibited by the use of phospholipase A2 inhibitor LY315920 in vitro [41]. In envenoming, bites from the African spitting cobras (Afronaja) are commonly associated with . Phylogenic tree relating the venom phospholipase A 2 activities of 15 cobra species by subgenera. Blue: acidimetric assay; red: colorimetic assay. The venom PLA 2 activity is expressed in a ratio relative to the highest activity detected by acidimetric or colorimetric assay in this study (1.0 implies the highest activity). More intense color indicates higher PLA 2 activity. The phylogenetic tree was redrawn with adaptation from phylogenetics of cobras [32]. Naja sumatrana is depicted here as a sister taxon of Naja sputatrix. Spitting cobras are marked with the snake symbol next to their species name.

Discussion
Phospholipase A 2 (phosphatidylcholine 2-acylhydrolase) catalyzes the hydrolysis of phosphatidylcholine at the sn-2 ester bond to produce lysophospholipid and free fatty acids. We tested the activities of svPLA 2 based on two different types of PLA 2 assays [37]: an acidimetric assay that measured the release of proton from fatty acids during the hydrolysis of phosphate ester bond, and a colorimetric assay that measured the amount of chromogenic 4-nitro-3-hydroxybenzoic acid released by the cleavage of PLA 2 at the ester bond between the octanol group of NOBA [38]. The activities tested on NOBA showed a better correlation with the PLA 2 contents in the cobra venoms. The egg-yolk-based acidimetric assay probably contained less specific substrates (phospholipids, triglycerides) that could be targets of other lytic enzymes in the venoms, or atmospheric carbon dioxide could have interfered with the assay by dissolving into the suspension during the stirring process. On the whole, the enzymatic rates measured by both assays support that PLA 2 enzymatic activities vary according to the svPLA 2 composition in different cobra species. The enzymatic svPLA 2 distribution is unique following a clustering trend among the four subgenera, and notably the remarkable lack of svPLA 2 within the Uraeus subgenus.
The enzymatic activities measured in this study for the venoms of African spitting cobras (N. nubiae, N. nigricollis, N. mossambica, N. pallida, N. katiensis) and those from Asia (N. sumatrana, N. sputatrix) were high, consistent with the high abundances of svPLA 2 reported previously in these venoms [14,26,34]. Some svPLA 2 of Afronaja cobras were shown to be cytotoxic and to have dermonecrotic and myonecrotic activities [39,40]. Some African spitting cobra venoms possess coagulotoxic activity, which were inhibited by the use of phospholipase A 2 inhibitor LY315920 in vitro [41]. In envenoming, bites from the African spitting cobras (Afronaja) are commonly associated with local tissue damages [42,43], which could be attributed to the svPLA 2 and cytotoxins present abundantly in these venoms [26]. Previous studies also indicated that cobra svPLA 2 enzymes worked synergistically with cytotoxins (cardiotoxins) to enhance venom toxicity [44,45] and the combination are probably responsible for causing venom ophthalmia (venom-induced conjunctivitis, chemosis, corneal erosions). Although the Asiatic non-spitters (N. naja, N. kaouthia, N. atra) also showed remarkably high PLA 2 activities (correlated with the composition); their svPLA 2 are mainly of acidic subtypes that lack lethal activity [46,47]. Two acidic PLA 2 subtypes from Indian N. kaouthia venom were reported previously to exhibit anticoagulant activity [48]; this coagulopathic or hemotoxic effect, however, has not been commonly reported in clinical cobra envenoming. Similarly, the African forest cobra (N. melanoleuca, subgenus Boulengerina) is a non-spitting cobra species whose venom exhibited a strong PLA 2 activity in this study; however, its svPLA 2 had been shown to play no crucial role in the toxicity of the venom [25]. The pathophysiological role of these apparently non-toxic svPLA 2 of cobras remain to be further elucidated, although these enzymes probably have more important ecological roles for the adaptation of the cobras to different niches. The lack of svPLA 2 in the African non-spitting cobras of the subgenus Uraeus is a unique venom phenotype unveiled in this study. The finding implies that the svPLA 2 is probably the least medically significant in the envenoming by this group of African cobras.
Snake venom PLA 2 s are isoenzyme products of multiple genes, and are further divided into distinct groups based on the differences in the number of disulfide bonds and the presence/absence of an N-terminal heptapeptide [49]. The svPLA 2 of elapid snakes (including cobras) belongs to Group Ia PLA 2 isoenzymes, which are homologous with the mammalian pancreatic PLA 2 (Group Ib PLA 2 ). Lynch [50] concluded that in Group I PLA 2 enzymes, gene duplication and diversification occurred after speciation. This implies that the sequence homology and antigenicity of cobra svPLA 2 are probably divergent between the different species of an individual subgenus. Hence, an antibody used against a PLA 2 subtype of a specific cobra species may reveal variable immunoreactivity between different cobra venoms. From a practical point of view, the potentially diverse svPLA 2 antigenicity and varying svPLA 2 protein abundance among the different cobras pose challenges for antivenom production and usage in some regions. The phenomenon may variably affect the neutralizing efficacy of antivenoms that are produced and used for specific treatment against the envenoming by different cobra species in Asia and Africa [51,52].
From the phylogenetic perspective, the moderate-to-high enzymatic activity of svPLA 2 is common in the Afronaja, Naja and Boulengerina lineages. Within the Asiatic Naja subgenus, the high PLA 2 enzymatic activities along with the emergence of basic and neutral svPLA 2 which are lethal, and the ability to 'spit' (to be exact, spray) venom, represents a more recently derived venom phenotype and defense trait unique to some of the Asiatic spitting cobras (N. sputatrix and N. sumatrana in this study). Although the Chinese/Formosan cobra (N. atra) and an unrecorded subpopulation of the Thai monocled cobra (N. kaouthia) were anecdotally reported to spit/spray venom on rare occasions, these two species are not considered accomplished spitters in this study due to their lack of formally documented specialized dental adaptations. The Asiatic spitting cobras, however, are known to be less accurate spitters compared with their African counterparts of the Afronaja subgenus that might have evolved a better defensive trait of venom spraying [53,54]. Some spitting cobras, e.g. N. sumatrana and N. sputatrix, produce lethal basic and/or neutral svPLA 2 related to the pathophysiology of systemic envenoming and venom ophthalmia [2,3,14,55].
The subgenus Uraeus broadly encompasses non-spitting cobras of the N. haje complex living in the open areas of Africa. The extremely low svPLA 2 activities and the negligible PLA 2 enzyme content are unique venom phenotypes in the Uraeus cobras, reflecting a less critical role of svPLA 2 in envenomation. The negligible svPLA 2 content could be correlated with weak cytotoxicity and low dermonecrotic activity of the venoms. These venoms are generally more neurotoxic among the African cobras. The loss of svPLA 2 functions in the lineage probably followed a decelerated mode of evolution or pseudogenization of the svPLA 2 as the cobras diverged from the common ancestor shared by their closest kin, the forest cobra of Boulengerina, which retained or evolved a venom with a high svPLA 2 enzymatic activity. The underlying cause, mechanism, and implication of the evolutionary event await further investigation.

Conclusions
The present study demonstrated the correlation of svPLA 2 enzymatic activities with the enzyme protein abundances in Afro-Asian cobras of different subgenus. The dominant presence of enzymatically active svPLA 2 is in line with the emergence of venom-spitting (spraying) behavior, once in the African Naja (Afronaja) spp., and once in the more derived Asiatic spitters of Naja (Naja) spp. The African non-spitting cobras of the subgenera Boulengerina and Uraeus diverged with a distinctive svPLA 2 distribution: the forest-dwelling species (Boulengerina) continued to use a venom rich in svPLA 2 , whereas the open-land species (Uraeus) adapted to a venom that has little or negligible svPLA 2 . The lack of svPLA 2 in the venom phenotype of African non-spitters from the Uraeus subgenus is hence striking. This provides an alternative view on the commonly perceived ubiquitous presence of svPLA 2 in cobra venoms, and implies that the significance of svPLA 2 in cobra envenoming varies among different species.

PLA 2 Assay (Acidimetric Method)
Phospholipase A 2 activities of the cobra venoms were determined by the acidimetric method as described by Tan and Tan [56]. The egg yolk substrate was prepared in a suspension constituted of chicken egg yolk, 18 mM calcium chloride, and 8.1 mM sodium deoxycholate in a 1:1:1 ratio. The substrate suspension pH was adjusted to 8.0 using sodium hydroxide. To ensure good mixing, the substrate suspension was continuously stirred at room temperature. One hundred microliters of venom solution (containing 10 µg venom) was added to 5 mL of the substrate suspension. The rate of pH decrease was recorded using a pH meter. A decrease of 1 pH unit of the egg yolk suspension corresponded to 133 µmol of fatty acids released. The enzyme activity is expressed as µmoles of fatty acids released/min/mg. The values are expressed as means ± S.E.M. of triplicates.

PLA 2 Assay (Colorimetric Method)
Phospholipase A 2 activities of the cobra venoms were assayed according to the colorimetric method as described by Holzer and Mackessy [35] and modified for use in a 96-well plate [57] using the synthetic chromogenic substrate (NOBA). Briefly, the standard assay mixture contained 200 µL of buffer (10 mM Tris-HCl, 10 mM CaCl 2 , and 100 mM NaCl, pH 8.0), 20 µL of substrate (3 mM), 20 µL of water, 20 µL of PLA 2 , and a final volume of 260 µL. Sample (10 µg in 20 µL) was then added to the mixture and incubated at 37 • C for 40 min, with the reading of absorbance at kinetic interval of 5 min over a period of 60 min at 425 nm using Tecan i-control™ infinite M1000Pro microplate reader (Männedorf, Switzerland). The enzyme activity, expressed as the initial velocity of the reaction, was determined by calculating the increase in absorbance after 20 min at 425 nm. The enzyme activity is expressed as mean ± S.E.M of triplicates.

Statistical Analyses
The PLA 2 activities of both acidimetric and colorimetric assays are expressed as means ± S.E.M. (standard error of mean) of triplicates. The statistical differences between the venom samples in individual assay were analyzed by one-way analysis of variance (ANOVA) (p < 0.05) followed by Tukey's post hoc test using SPSS version 20 software (IBM, Armonk, NY, USA, 2016). Lower-case letters are labelled at the top of the bars to indicate if there is a significant difference between the values. Sharing of any common lower-case letters between bars indicates there was no significant difference (p > 0.05) between the values charted, whereas bars that do not share any common lowercase letters have values that are significantly different from one another (p < 0.05). The relationship between the PLA 2 activity and the relative abundance of PLA 2 in each venom sample was also studied by linear regression analysis, using Graphpad Prism 6 software (San Diego, CA, USA). The strength of the association was interpreted by the coefficient of determination (R 2 ) and the cut-off value at p < 0.01 indicates a highly significant regression between the PLA 2 activity and the PLA 2 protein abundance.
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6651/11/2/116/s1, Figure S1: Time-dependent pH changes in acidimetric assay for the venoms of four subgenera of cobra: (A) Naja, (B) Afronaja, (C) Boulengerina, and (D) Uraeus. Hydrolysis of phospholipids by phospholipase A 2 released fatty acids that reduced the suspension pH in a time-dependent manner. Figure S2: Time-dependent absorbance changes in colorimetric assay for the venoms of four subgenera of cobra: (A) Naja, (B) Afronaja, (C) Boulengerina, and (D) Uraeus. Changes in absorbance were due to the hydrolysis of the synthetic chromogenic substrate (NOBA), corresponding to the enzymatic activity of phospholipases A 2 in the venoms. Table S1: Relative abundances of snake venom phospholipase A 2 of 12 cobra species (Genus: Naja).