UDC: 547.993.02:595.443.8-114.5.008 COMPARATIVE ANALYSIS OF ARANEIDAE VENOMS AND TOXINS: CHEMICAL STRUCTURES AND ELECTROPHYSIOLOGICAL EFFECTS

© 2020 O. M. Klyuchko; Published by the Ivan Franko National University of Lviv on behalf of Біологічні Cтудії / Studia Biologica. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://www.budapestopenaccessinitiative.org and Creative Commons Attribution 4.0 License), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. UDC: 547.993.02:595.443.8-114.5.008

Investigations of Araneidae toxins are rather important because of the high effectiveness and specificity of these substances. Previously, it was suggested that toxin JSTX might be used as a universal "marker" of glutamate receptors in different representatives of fauna of the Earth. There is a significant similarity of glutamate receptors in different phylogenetically distant organisms: neuromuscular lobster junction, mammals' brain cells and others. Further, it was demonstrated that JSTX-3 cannot be used as "universal marker of glutamatergic synapses" for different species of fauna because of revealing complexity of glutamate CRC molecular structures.
Among these toxins, JSTX-3 has the simplest structure: its polyamine chain is the shortest and has no branching. The AR molecule has longer polyamine chain attached to it by 2 reactive amino groups. ARN-1 and ARN-2 have the polyamines with equal lengths to which amino groups are also attached. A distinctive characteristic of ARN-1 is the easily dissociating cationic group that contains the pentavalent nitrogen in the polyamine chain. According to the length of polyamine chain, all studied toxins can be arranged in following sequence: JSTX-3 <AR <ARN-1 = ARN-2.
In Figs. 2, 3 the blocking effect of integral venom JSTX-V and its main active component JSTX-3 on chemo-activated currents in rat hippocampal membrane is demonstra-ted. In Fig. 4, the blocking activity of integral venom AR-V on chemo-activated currents in neuronal membrane is shown. In Fig. 5, the blocking activity of the toxin argiopin (AR), the main active element from this venom is presented. In our experiments, all antagonists -venoms and toxins -acted in a reversible manner on glutamate receptors in hippocampal membranes, but the rates of their reversible effects were different in all cases (Fig. 6).  Рис. 6. Ступінь блокування каїнатактивованих струмів різними антагоністами (білі колонки) і ступінь відновлення амплітуд струмів після видалення антагоністів у розчині Рінґера (заштриховані колонки). Значення наведено у відсотках від амплітуди контрольної відповіді, значення якої прийнято за 100 % (верхня пряма пунктирна лінія). Діаграму виконано для 6 різних антагоністів глутаматних канало-рецепторних комплексів [27] In our experiments, both argiopin and AR-V blocked only the opened channel (Figs. 4, 5). The pair JSTX-3 and JSTX-V differed from them in their characteristics. JSTX-V and JSTX-3 acted both on opening and closure of Glu-and KK-activated channels (Figs. 2, 3). Any differences between antagonists were not detected in experiments for studying of potential-dependence of these toxins' actions. The action of all four antagonists to a greater or lesser extent depended on the holding membrane potential: the blocking properties of all of them become worth with membrane depolarization. Regarding the differences in the action between the pairs of venom-toxin from N. clavata and venom-toxin AR from A. lobata -they were visible in degrees of current blocking and their recovering during "washing", as well as in numerical values of parameters of blocking and "washing" (Fig. 6).
During the study of kinetics of electrical transmembrane currents blocking at the background of stationary KK-activated ionic currents, it was found that the pairs "venomtoxin" had similar kinetics of blocking (see Table)  . Thus, the blocking processes of CRC by both JSTX-V and JSTX-3 could be described by one exponent. The same processes of channel-receptor complex (CRC) blocking by AR-V and AR were approximated satisfactorily by two exponents in both cases (see Table)  . One exponent could also describe the process of AR-V and AR removing by "washing" in Ringer's solution. Qualitatively, general features of the action in "venom-toxin" pairs were reflected in the existence of one constant of the reaction of interaction of JSTX-V and JSTX-3 with CRC and two constants of the direct reaction of interaction between AR-V and AR. Quantitatively, the rate of interaction of antagonists with CRC can not be compared, since the concentrations of venoms and toxins were measured in different units. No differences were found in the specificity of antagonists' action within the pairs JSTX-V and JSTX-3; or AR-V and AR. In used conditions, they all suppressed the amplitudes of only Glu-, KK-, QL-activating currents. Consequently, in the composition of the venoms there were no antagonists of other receptors, namely GABA and glycine receptors.

Кінетичні характеристики блокування KK-активованих іонних струмів [27]
Antagonist Constant rate of blocking (direct reaction) Velocity of electrical current amplitude recovering A described similarity in characteristics of actions of venoms and toxins (their main active components) allow to conclude that the physiological properties of the venom JSTX-V and AR-V were determined by the toxins JSTX-3 and AR, respectively. However, the contribution of these toxins in overall effect although maximal, but does not exhaust all the properties of the venom. For example, the irreversibility of JSTX-V action (unlike JSTX-3), might be explained by the presence of other fractions in the venom, that are bound with the membrane irreversibly. During binding, such blockers could change irreversibly the properties of glutamate channel-receptor complex (gCRC) by themselves, or they could operate through different, more complex mechanisms. Via binding with the membrane, but not blocking the gCRC by itself, they could increase the stability of the complex gCRC -JSTX-3. Although the probability of second assumption is less, we can not refuse from it absolutely.
The results of our studies of venoms and toxins from A. lobata suggest the following mechanism. AR-V is "washed" much worse than AR. AR-V contains a number of fractions with toxic activity; AR is only one of them. Such fractions were capable to block KK-activated currents practically irreversibly. After studying of composition of venom AR-V, other active components like argiopinin 1 (ARN-1) and argiopinin 2 (ARN-2) have been isolated from it [28,29]. Both these substances blocked KK-activated currents, forming significantly more stable links on the membrane than AR. It was possible to see that argiopinin 1 acts less reversibly than other antagonists. Obviously, the greater degree of irreversibility of AR-V action in comparison with AR is determined by ARN-1 and similar toxins.
The pair JSTX-V and JSTX-3 the largest difference in these characteristics demonstrated. During the action of integral venom the amplitude of KK-activated response fell to a new steady state (34 % of control response) and no longer recovered. Then the JSTX-3 toxin was able to be removed partially with the recovery of the response amplitude (from 64 % to 39 % of control response). The same tendency but in less degree was observed in AR-V and AR pair. Thus, the degree of AR-V ability for removing is really lower (from 14 % to 32 % of control response) than for AR (from 20 % to 77 % of the control response).
We took into account the complexity of venoms composition. For example, any of the well-purified or synthetic toxins did not activate transmembrane ion currents by itself. If such activation occurred, the analysis showed the presence of glutamate in such poorly purified preparations. This fact makes us doubt in the data of other authors [63] who found that that Arthropoda toxins activate ion currents in the membranes by themselves. As a rule, the rough preparations or integral venoms were used in those experiments.
Other our results might be caused by complex composition of venoms. All venoms and toxins were kept frozen at -4 °С. During the first 3-4 months, the properties of those venoms remained unchanged. However, with long-term storage (more than 6-7 months), the properties of integral venoms be changed. For example, JSTX-V began to "wash off" after its removing, the KK-activated responses were partly restored. Similar observation was made by Japanese authors [47,48,57]. However, is impossible to answer unambiguously why is that happening. Possibly, venom preparations are less stable than individual toxins due to the mutual influence of venom components. With a prolonged storage, the molecules of irreversibly active substances can be decomposed more quickly. During dissolving, low molecular weight substances interact with the membrane. It is possible to explain that effect better after a more detailed study of Arthropoda venoms' and toxins' properties.
Comparative analysis of properties of different Araneidae toxins in a link with their chemical structures. The influences of all studied toxins on the glutamate receptor were characterized by a number of common features. They all blocked glutamate-(GLU), kainate-(KK), and quisqualate (QL)-activated currents in rat hippocampal membranes to various degrees and were washed differently. They did not act on the excitatory, glycine-, and GABA-activated currents in these neurons. Their blocking action depended on the transmembrane holding potential -it became worth with membrane depolarization. All these substances could block the open channels in gCRC. It is possible to assume that properties of these toxins are due to the fragments of molecules that are common to all toxins: phenolic or indole fragment coupled to asparagine.
Our experiments did not demonstrate any significant differences in the properties of toxins containing phenolic or indole fragments. The only slight difference is that the dissociation of AR derivatives that contained indole groups is slowed down slightly (see the  Table). However, the direct constant rates of toxins' interaction with AR derivatives did not differ significantly. Such pattern of properties of toxins, phenol-and indole-derivatives allows to make some conclusions (see below).
First, the mechanism of gCRC block by toxins should be based on the reaction of the interaction of aromatic toxin groups with membrane groups. Second, this reaction have to be common to indole and phenolic groups, so, the membrane should not "differentiate" them. Finally, probably these groups (or linked with asparagine) predetermine the effect of gCRC block and its main features (potential-dependence and others).
Influence of toxins on activated and non-activated receptor. The main difference in the action of toxins is that JSTX-3 interacts with gCRC, regardless of whether it is in the activated state or non-activated (Fig. 3). From other side, the main mechanism of action of AR, ARN-1, ARN-2 is the blocking of GLU-and QL-activated channels in the open state (Fig. 5). Similar results were obtained previously for AR [21,24]. What is the explanation in difference in the toxins action? The only structural difference between JSTX-3 and other ones is more simple molecular structure of the first one. It is possible that further minor complication of structure of A. lobata toxins' molecules (extension of their polyamine chains, adding of amino groups branched from the main chain), increases the selectivity of their interaction with gCRC in such a way that they lose the ability to bind to gCRC in the non-activated conformation. Thus, were simple JSTX-3 molecule is less "legible" and "do not distinguish" the receptor conformations that correspond to the activated and inactivated states. In both cases, JSTX-3 molecule blocks the ion currents almost completely. We would like to carry out similar experiments investigating more simple fragments of JSTX-3 molecule -2,4-dihydroxyphenylacetat-asparparin (DHPhA-Asp) and 2,4-dihydroxyphenylacetic acid (DHPhA) that can block gCRC [53].
An interesting feature is that the processes of AR and JSTX-3 removing by "washing" in Ringer's solutions (both in the presence and without agonists (GLU, KK)) were similar. According to our preliminary data, the presence of agonists (GLU, KK) was neces sary for the better "washing" of argiopin and some argyopinins in solutions. The processes of formation and dissociation of toxin-receptor complexes were better for these substances with the activation of gCRC. According to the literature data, another toxin -gCRC antagonist with similar characteristics is known: δ-philanthotoxin [4]. However, a significant difference in structure of its molecule (it has oxyphenol instead of 2,4-DHPhA or 2,4-DHPhA-Asp) and the absence of data on that issue does not allow us to make any conclusions about the relationship of its properties with the structure of its molecule.
According to data of some authors', the necessary condition for AR "washing" is the presence of agonist in washing solutions [48,24]. This contradicts to the results of our researches. However, that effect is easy to explain taking into account that in those works the roughly purified AR preparations were used. That effect could be caused by the impurities of argiopinins.
It should be noted that the obtained regularities of toxins' properties' correlation with their structure may be changed during the study of newly obtained antagonists with another chemical structures.
Degree of KK-activated currents blockage. None of studied toxins could block KKactivated currents completely, even at high concentrations (10 -4 mol/L). JSTX-3 suppressed the amplitudes of KK-activated currents the most effectively -up to 6 % of the initial values. The efficiency of KK-activated currents suppression decreased from left to right in the following sequences of toxins: JSTX-3 > AR > ARN-2 > ARN-1. As one can see, the degree of current suppression by the toxins may depend on the length of their polyami nes: the shorter the toxin, the more effectively it closes the ion channel. However, this finding should be supported by additional studies of other analogs with different lengths of polyami ne chain.
Irreversibility of toxins' action. For JSTX-3 analogues with different polyamine lengths, it was found that polyamine chain elongation caused the formation of more stable toxin-receptor complex [22,24]. However, AR demonstrated more revealed reversible action, although its length is only slightly longer than that of JSTX-3. Perhaps, this is due to the fact that two amino groups linked with AR polyamine chain can react easily, and the length of polyamine appeared to be functionally less important? To answer this question, we studied how the toxin structure is related to the rate of "washing" of other A. lobata toxins (see Table, Fig. 6).
We demonstrated that AR was in separate position among the other studied toxins from A. lobata. All other A. lobata toxins were "washed" much worse, regardless of their structures. However, if to study the sequence of analogs separately, it was shown that it was easier to wash the substances with longer polyamine fragment. This was true within the groups of argiopinins 1-5 and pseudoargiopinins 1-3 (see [29]). Our conclusion is opposite to the conclusion of other authors, that was made earlier for JSTX-3 analogues (see above). Thus, for AR and its analogues -the longer is the toxin molecule, the better it is "washed". The exception was ARN-1 (arginopinin 1) whose action is almost irreversible. However, unlike other toxins, its polyamine contains an easily dissociated cationic group with pentavalent nitrogen. This group probably contributes to the formation of more stable toxin-receptor complex.
Kinetic characteristics of toxins' blocking action. Based on the calculations' for the values of toxins' binding constants and constants of toxins' "washing" (see the Table), as well as on the basis of previous data obtained during the study of argiopinins 1-5 and pseudoargiopinins 1-3, the following conclusion was done. With the lengthening of the polyamine chain, the values of the first binding constants of the toxins with gCRC increased, and the constants of toxins "washing" increased too. Otherwise, the longer is the toxin molecule, the faster it binds to the receptor and breaks the links with gCRC more quickly during dissociation. This rule is not true for the molecules having long side chains, namely ARN-4, whose binding constant is higher than expected. Perhaps, this regularity should be transformed as follows: the longer and more branched is the toxin molecule, the faster it binds and breaks the bonds with gCRC.
Relationships between toxins structures and their electrophysiological charac te ristics as a base for novel methods of quantitative and qualitative analysis. All to xins studied in this work were isolated both from the natural source -venoms, and were syn thesized in laboratory conditions. The effect of synthetic analogues was completely iden tical to the action of natural toxins. This confirms the correctness of the decoded toxins' structures and the identity of the synthetic toxins with the corresponding natural analogs.
We suggest that the relationships between toxins structures and their electrophysiological properties described above can be used for the development of novel methods of qualitative and quantitative analysis [27,33,37,38]. Accordingly, this our suggestion can be defined as following algorithm.
1 -The necessary electrophysiological characteristics have to be registered in experiments. 2 -These data have to be accumulated and ordered in databases. 3 -The relationships between toxins structures and their electrophysiological properties have to be established and revealed (for example, in a graph form). 4 -Unknown chemical substance can be defined on the base of established relationships in comparison with known substances. It is necessary to emphasize that such methods have to be based obviously on good statistics and developed computer basis.

CONCLUSIONS
In present article we have analyzed the results of the studies of some Araneidae substances. The results of investigations of two spider species -Nephila clavata and Argiope lobata are presented. Some electrophysiological characteristics of these venoms and toxins -their main active components -have been observed. The results of electrophysiological properties studies for the pairs: venom from N. clavata with toxin JSTX-3, and venom from A. lobata with toxins argiopin AR, ARN-1, ARN-2 are presented in parallel. A described similarity in characteristics of actions of venoms and toxins (their main active components) allows concluding that the physiological properties of the venom JSTX-V and AR-V are determined by the toxins JSTX-3 and AR, respectively. However, although the contribution of these toxins in overall effect is significant, it does not cover all properties of the venom.
The similarity in blocking characteristics between each venoms and corresponding toxins is presented. In our comparative analysis of correlations between the electrophysiological effects of studied toxins and their chemical structures, following effects of these interactions were studied: 1 -toxins influence on the activated and inactivated receptors; 2 -degree of blocking of KK-activated currents by different studied antagonists; 3 -comparison of degrees of irreversibility of toxins action; 4 -differences in kinetic characteristics of toxins' blocking actions. The relationships between toxins' structures and their electrophysiological characteristics are described; they can be used for the development of novel methods of qualitative and quantitative analysis [27,33,37,38]. For these methods, the patents of Ukraine were obtained [30][31][32][33][34][35][36][37][38].