£]-Bungarotoxin (£]-Bgt), a presynaptic phospholipase A2 (PLA2) neurotoxin isolated from the venom of Bungarus multicinctus (Taiwan banded krait), consists of A chain and B chain, cross-linked by an interchain disulfide bond. A chain is structurally homologous with phospholipase A2 (PLA2) enzymes, while the sequence of B chain is homologous to the Kunitz-type protease inhibitor and dendrotoxin. In addition to PLA2 activity, £]-Bgt blocked the neurotransmission at the neuromuscular junction by selectively inhibiting certain voltage-sensitive potassium channels. The present studies investigated the B chain of £]-bungarotoxin and B chain homologous proteins in evolutionary relationship and cytotoxic mechanism. Eight A chain cDNAs and three B chain cDNAs have been cloned from B. multicinctus venom glands. Random combination of the A and the B chains should produce a number of £]-Bgt isotoxins. There are at least 16 isoforms of £]-Bgt were been isolated. Previous studies indicate that A and B chains are encoded separately by different genes, and the A and B chain genes do not originate from a common ancestor. These findings suggest that the intact £]-Bgt molecules should be derived from the pairing of A and B chains after their mRNAs are translated. And, our recent studies show that B chain genes and Naja naja atra chymotrypsin inhibitor (NACI) share the same genomic organization and high sequence identity. Alternatively, limited studies on the evolutionary divergence of B chain gene and its homologous have been reported. In the first part, the structural organization of the genes encoding B2, B4, B5 and B6 chains of £]-Bgt are reported. These genes shared virtually identical overall organization with three exons interrupted by two introns in similar positions. On the contrary, intron 1 of these genes had a similar size, a notable variation with the size of intron 2 was observed. It was found that two regions at the second intron of B1 and B2 chains were absent in that of B4, B5 and B6 chains. RT-PCR analyses indicated that Bungarus multicinctus venom gland, heart, liver and muscle expressed the RNA transcripts showing sequence similarity with the intronic segment being deleted in B4, B5 and B6 chain genes. This reflects that the ancestral gene of the intronic segment might insert in multiple loci of B. multicinctus genome. Comparative analyses of B chain genes showed that the protein-coding regions of the exons are more diverse than introns, except for in the signal peptide domain. These results suggest that intron insertions or deletions occur with the evolution of B chains, and that accelerated evolution may diversify the protein-coding sequence of B chain genes same as snake phospholipase A2, neurotoxin and cardiotoxin genes. The second part is to explore the functional contribution of the two subunits to the toxicity of £]-Bgt. £]-Bgt was found to induce apoptotic death of SK-N-SH cells via elevating intracellular Ca2+ and intracellular ROS production. Moreover, an activation of p38 MAPK was associated with the cytotoxicity of £]-Bgt. SB202190( p38 MAPK inhibitor), N-acetylcysteine (antioxidant reagent), 1,2-bis (2-amino-phenoxy) ethane-N,N,N,N-tetraacetic acid (BAPTA) (Ca2+ chelator) and the inhibitors of Ca2+ release from intracellular depots (ruthenium red and 2-aminoethoxydiphenyl borate) effectively attenuated the cytotoxicity of £]-Bgt. In sharp contrast to the inability of A chain, B chain was able to induce cytotoxic effects on SK-N-SH cells as £]-Bgt did. Abolishment of PLA2 activity did not significantly alter the cytotoxic activity of £]-Bgt. MK801 (an NMDA receptor antagonist), antibodies against NMDA receptor and 4-aminopyridine (a potassium channel blocker) markedly reduced the cytotoxic effects of £]-Bgt, B chain and catalytically inactivated £]-Bgt. Moreover, antibodies against NMDA receptor blocked the binding of rhodamine-labeled £]-Bgt to SK-N-SH cells. Taken together, our data indicate that B chain is a functional subunit responsible for the cytotoxicity of £]-Bgt, and suggest that the cytotoxicity of £]-Bgt is mediated by NMDA receptor and potassium conductance. Homologous proteins of the B chain, namely, protease inhibitor-like protein-1 (PILP-1), PILP-2, and PILP-3, were successfully cloned from the Bungarus multicinctus genome. The 3 cloned genes each comprised 3 exons and 2 introns, similar to the genes of the B chain of £]-bungarotoxin and NACI. Based on the Ka/Ks values of the gene sequences of PILPs, the B1 chain, and NACI, the PILPs may have undergone adaptive evolution. The cDNAs of the PILPs were recombined by PCR, and the recombinant proteins were successfully expressed and purified. PILPs induced cytotoxicity in U937 cells, but PILP-3 reduced viability only slightly. Induction of cell death by PILP-1 and PILP-2 was by apoptosis that occurred in a caspases-dependent manner. PILP-1 and PILP-2 activated p38 MAPK and downregulated ERK1/2; however, PILP-3 caused no such effects. PILP-1 increased the population of sub-G1 cells and caused mitochondrial damage. PILP-1-treated cells were observed to demonstrate loss of mitochondrial membrane potential, release of cytochrome C, and a decrease in the level of Bcl-2 and pro-caspase 9. Pretreatment of U937 cells with either SB202190 or a caspase 8 inhibitor significantly prevented the PILP-1- and PILP-2-induced degradation of caspase 8 or 3. PILP-1 and PILP-2 increased the level of TNFRII, which was suppressed by the p38 MAPK inhibitor. The transient transfection of PILP-1- and PILP-2-treated cells with pCMV-MEK1 efficiently increased their survival and phosphorylation of ERK1/2 but reduced the level of TNFRII. After PILP-1 and PILP-2 induced TNFRII, the excess TNF caused a significant reduction in cell viability. TNFRI levels were not significantly changed in PILP-1- and PILP-2-treated U937 cells. These results suggest that PILP-1 and PILP-2 induce apoptosis via an increase in the level of TNFRII and p38 MAPK activity and the suppression of ERK activity.

Abstract Our previous study showed that KChIP3 (Kv Channel Interacting Protein 3) probably was a physiological targete protein of beta-bungarotoxin (beta-Bgt) as evidenced by yeast two-hybrid system. Thus, extensive efforts are carried out to explore the molecular interaction between KChIP3 and beta-Bgt in the present study. KChIPs are potassium (Kv) channel-interacting proteins that bind to the 1~90 amino acid of N-terminus of Kv4 alpha-subunits and regulate the ion current density, shift the voltage dependence of activation and speed their recovery from inactivation. beta-Bgt, a presynaptic neurotoxin purified from Bungarus multicinctus venom, consists of A chain and B chain which cross-linked by an interchain disulfide bond. The results of pull-down assay revealed that, in contrast to other KChIP proteins, KChIP3 bound with beta-Bgt. Moreover, it was found that the B chain of beta-Bgt was a functional subunit in the binding with KChIP3, and the binding of KChIP3 to beta-Bgt showed a Ca2+-dependent manner. Removal of the third and the fourth EF-hand regions of KChIP3 abolished its interaction with beta-Bgt. Noticeably, the binding of beta-Bgt with KChIP3 did not influence the interaction between KChIP3 and Kv4. In the meantime, rat brain KChIP3 could be isolated using a beta-Bgt-Sepharose column. These observations suggest that KChIP3 is an intra-cellular target recognized by beta-Bgt. Accidently, it was found that direct protein-protein interaction between Taiwan cobra cardiotoxin3 (CTX3) and potassium channel-interacting proteins (KChIPs) was investigated. It was found that KChIPs bound with CTX3, in which KChIP1 and CTX3 formed a 1:1 complex as evidenced by the results of chemical crosslinking. Pull-down assay revealed that the intact EF-hand 3 and 4 of KChIP1 was critical for CTX3-binding. Whereas, all mutated KChIP3 were able to bind with CTX3. In contrast to the interaction between KChIP1 and KvN, the binding of CTX3 to KChIP1 showed a Ca2+-independent manner. Fluorescence measurement revealed that CTX3 affected the binding of ANS to Ca2+-bound KChIP1, but not Ca2+-free KChIP1. Alternatively, KChIP1 simultaneously bound with KvN and CTX3, and the interaction between KChIP1 and KvN was enhanced by CTX3. In terms of the fact that KChIPs regulate the electrophysiological properties of Kv K+ channel, the potentiality of beta-bgt and CTX for this biomedical application could be considered.

Programmed cell death is a widespread phenomenon in the developing nervous system. During early development, neurons are initially produced in excess and up to 70% of them are eliminated in later stages of development, during a period of synapse formation with their targets. However, the mechanisms that initiate the death of neurons are not clear. In the avian ciliary ganglion, neurons go through the period of target-dependent cell loss between E8 and E14; however, almost all neurons in the ganglion are prevented from dying by the chronic in ovo treatment with α7-nAChRs specific antagonists, α- bungarotoxin or MLA. Since α7-nAChRs are implicated in the cell death of ciliary ganglion neurons, I tested whether the activation of these receptors directly on the ciliary ganglion neurons facilitates cell death by inducing large increases in intracellular Ca2+. I found that the ciliary ganglion neurons are heterogeneous with respect to their surface α7-nAChR density and, as a result, activation of these receptors by nicotine leads to large increases in [Ca2+]i in some neurons but not in others. Furthermore, immature E8 neurons exhibit slower rates of Ca2+ decay after nicotine stimulation than E13 neurons, suggesting that E8 neurons do not clear [Ca2+]i efficiently and could be more susceptible to Ca2+ overload. Expressing the αbtx that is tethered to the cell membrane via the glycosylphosphatidylinositol anchor (GPIαbtx) in the ciliary ganglion neurons inhibits the increases in [Ca2+]i induced by nicotine through α7-nAChRs specifically. This cellautonomous inhibition of α7-nAChRs prevents cell death of ciliary and choroid neurons. For this to happen, GPIαbtx must be expressed in neurons; the expression of this construct in the surrounding non-neural tissue does not prevent neuronal loss in the ciliary ganglion. Later in development, α7-nAChRs are prevented from inducing cell death by the chicken PSCA molecule that is significantly upregulated in the ciliary ganglion between E8 and E15. The chicken PSCA is neuronal specific molecule that belongs to the Ly-6/neurotoxin superfamily that includes αbtx and lynx1 and compared to other tissues, it is highly expressed in the ciliary ganglion. The expression of the PSCA mRNA in tissues correlates with the expression of α7-nAChR mRNA, suggesting that PSCA modulates the signaling via these receptors. In fact, overexpressing the PSCA in the ciliary ganglion neurons prevents nicotine-induced increases in [Ca2+]i through α7- nAChRs. Misexpressing the PSCA in E8 ciliary ganglion prevents choroid but not ciliary neurons from dying. Therefore afferent inputs can induce cell death by activation of α7- nAChRs in the developing ciliary ganglion by increasing the [Ca2+]i over the threshold for cell death. Upregulation of endogenous prototoxins, such as PSCA, opposes the large increases in [Ca2+]i via α7-nAChRs and prevents these channels from facilitating cell death after the final numbers of neurons have been established. These results indicate that the control of cell death is more complex than originally proposed by the neurotrophic hypothesis and present the mechanism by which cell death in the developing ciliary ganglion is regulated, thus, further highlighting the importance of non-traditional roles of α7-nAChRs during the development of the nervous system.

Our previous studies showed that recombinant £]-bungarotoxin B chain exhibited membrane-damaging activity. Given that membrane-damaging activity is crucial for bactericidal effect of antibacterial peptide, the causal relationship between membrane-damaging activity and antibacterial action of B chain was performed in this study. £]-bungarotoxin B chain exhibited a growth inhibition on Escherichia coli (Gram-negative bacteria), but marginally displayed bactericidal effect on Staphylococcus aureus (Gram-positive bacteria). Destabilization of lipopolysaccharide (LPS) layer and inhibition of lipoteichoic acid (LTA) biosynthesis on cell wall increased bactericidal effect of B chain on E. coli and S. aureus. B chain induced leakage and fusion of bacterial membrane-mimicking liposomes. Compared with LPS, LTA notably suppressed membrane-damaging activity and fusogenicity of B chain. B chain showed similar binding affinity with LPS and LTA. Circular dichroism measurement revealed that LPS- and LTA-binding differently induced conformational change of B chain. Taken together, our data indicate that antibacterial action of B chain is related to its ability to induce membrane permeability and fusogenicity, and suggest that LTA- and LPS- induced conformational change of B chain affect membrane-damaging activity, fusogenicity and antibacterial activity of B chain.

Thesis (Ph.D.)--Brown University, 2008. / Vita. Advisor : Edward Hawrot. Includes bibliographical references.

beta-Bungarotoxin (beta-Bgt), the presynaptic neurotoxin purified from the venom of Bungarus multicinctus, consists of the A chain and the B chain, cross-linked by an interchain disulfide bond. In this study, two novel beta-Bgt isotoxins were purified from Bungarus multicinctus venom by the combinations of ion-exchange chromatography and reverse phase HPLC. Amino acid sequencing, peptide mapping and mass analyses revealed that they probably contained the same A chain, but their B chain differed. Consequently, the discrepancies in their biological activity and fine structure reflected the role of B chain in intact of beta-Bgt. In Yeast-Two-Hybrid system, a potassium channel binding protein was identified to interact with the B chain of beta-Bgt. Although the recombinant potassium channel binding protein functionally bound with Ca2+, but it could not prove to bind with BM12 and BM13 as revealed by in vitro cross-linking assay. The A chain genes including A1 chain, A2 chain and A8 chain genes were amplified by PCR reaction. Their nucleotide sequences shared up to 97.5% identity. Alignment of the determined A chain genes with A chain cDNAs revealed that the A1 chain gene was organized with four exon and three intron, while A2 chain gene comprised three exons and two introns. When A2 chain is expressed, the region corresponds to the first exon of A1 chain gene is skipped instead of inclusion of intronic sequence adjacent to the second intron. The resulting A2 chain mRNA encoded a 25 residues signal peptide, which different from A1 chain mRNA with a 27 residues signal peptide. Comparative analyses on phospholipase A2 genes and cDNAs suggest that this is the first report on skipping of exon changes the signal peptide sequence of snake venom proteins.

beta-Bungarotoxin (beta-BuTx), the presynaptic neurotoxin purified from the venom of Bungarus multicinctus, consists of two dissimilar polypeptide subunits. A phospholipase A2 subunit named A chain, and a non-phospholipase A2 subunits named B chain. The A chain and B chain are covalently linked by one disulfide bridge. Although it has been widely accepted that the toxic effect of beta-BuTx is attributed to the disturbance of presynaptic transmitter release, however the inhibition of transmitter release by beta-BuTx is still obscure. Here we investigate the mechanism that mediates facilitation of transmitter release at the neuromuscular junction induced by beta-BuTx, using Xenopus nerve-muscle coculture. Application of beta-BuTx and isotoxins BM12, BM13 led to a marked increase in the frequency of spontaneous synaptic currents (SSCs) after a short period (12~18 min) of latency. The synaptic potentiation induced by these toxins was abolished when Ca2+ in the medium is substituted by Ba2+ (a potent phospholipase A2 inhibitor). Application of PLP-BM12 and PLP-BM13, which have been chemical-modification to lose their PLA2 activity from BM12 and BM13, failed to potentiate the transmitter release.

The NAD glycohydrolase (NADase) from Bungarus fasciatus venom was purified over 1000-fold to electrophoretic homogeneity through a 3-step procedure which included affinity chromatography on Cibacron Blue agarose. The enzyme exhibited a broad pH profile with the optimum range between 7-8. Studies on the substrate specificity of B. fasciatus venom NADase demonstrated that alterations in the purine ring were less pronounced then alterations in the pyridinium moiety of NAD. Product inhibition studies indicated nicotinamide to be a noncompetitive inhibitor with a K_i = 1.4 mM and ADP-ribose to be a competitive inhibitor with a K_i =0.4 mM. The purified enzyme was inactivated by both 2,4-pentane-dione and Woodward's Reagent K suggesting the involvement of a lysine and carboxyl group in the catalytic process. In contrast to other known NADases, the snake venom enzyme did not self-inactivate. The purified B. fasciatus venom NADase catalyzed a transglycosidation reaction (ADP-ribose transfer) with a number of acceptor molecules. The functioning of a variety of substituted pyridine bases as acceptor molecules was demonstrated through the formation of the corresponding NAD analogs. The enzyme also catalyzed the transfer of ADP-ribose to aliphatic alcohols (methanol to hexanol, inclusive) and a positive chainlength effect was observed in the functioning of these acceptors. Kinetic studies of transglycosidation reactions were consistent with the partitioning of an enzyme-ADP-ribose intermediate between water and nucleophilic acceptors as has been proposed in earlier studies of mammalian NADases. The partitioning of this intermediate between water and pyridine bases can be correlated with the basicity of the ring nitrogen of the pyridine derivative. The K_i of pyridine bases in the hydrolytic reaction did not equate to the K_m of these bases in the pyridine base exchange reaction suggesting two forms of the NADase with varying affinity for the pyridine bases. This implys the pyridine base exchange reaction to be more complicated than originally proposed. / Ph. D.

Calcium (Ca²⁺) influx through Caᵥ1.2 channels mediates cardiac excitation-contraction coupling, tunes cardiac action potential duration and excitability, and regulates cardiomyocytes’ (CM) gene expression. Mechanisms regulating the sub-cellular localization, trafficking, and dynamics of surface Caᵥ1.2 in ventricular CMs are poorly understood though these are critical determinants of cardiac function. To gain new insights into Caᵥ1.2 organization, dynamics, and regulation at the CM surface we generated transgenic mice expressing an αMHC controlled cardiac-specific, dihydropyridine (DHP)- resistant α₁_ᴄ construct, tagged at the N-terminus with FLAG and HA epitopes, at the C- terminus with YFP, a 13-residue bungarotoxin binding site (BBS) inserted into in the third extracellular loop of domain II, and mutations that prevent cleavage of the C-terminus. We found robust inducible expression of DHP-resistant FLAG-HA-BBS-α₁_ᴄ-YFP in the heart that targeted to dyadic junctions, generated nisoldipine-resistant Ca²⁺ currents, supported cardiac excitation-contraction coupling, and was normally up-regulated by β-adrenergic activation with isoproterenol. Incubating transgenic CMs with AlexaFluor₆₄₇-conjugated α- bungarotoxin (BTX₆₄₇) enabled selective labeling of surface BBS-tagged Caᵥ1.2 channels. We used total internal fluorescence (TIRF) microscopy to investigate the spatiotemporal organization and dynamics of surface Caᵥ1.2 channels. Similar to endogenous Caᵥ1.2, transgenic α1C-YFP forms clusters with exponentially distributed sizes at the cell surface. A flow cytometry-based optical pulse-chase assay revealed surface Caᵥ1.2 channels in adult cardiomyocytes fully turn over within two hours. Application of angiotensin II (Ang II) decreased transgenic Caᵥ1.2 surface density and this effect was blocked by the selective Ang II receptor type I (AT1R) blocker losartan. Application of losartan by itself increased Caᵥ1.2 surface density, suggesting the potential presence of constitutively active Ang II receptors in adult CMs. Our results provide new insights into spatiotemporal organization, dynamics, and regulation of Caᵥ1.2 channels in adult CMs and introduce an approach that can be widely applied to elucidate spatiotemporal dynamics of cardiac ion channels and membrane proteins.

Six new structural parameters £rB, £pB, £rC, £pC, £rS, and £pS are proposed to enhance the side chain actions in protein structures. Programs for calculating these new parameters based on phi and psi torsion angles vector algebra calculation method are established. A bivariate model with von Mises marginal distributions are applied to establish models of phi and psi in protein class Ophiophagus hannah neurotoxins and alpha-bungarotoxins respectively. 11 global structural parameters include phi and psi torsion angles, bond lengths of C-N, C-O, C£\ -C, and N-C£\, and bond angles of C-N-C£\, C£\-C-N, C£\-C-O, N-C£\-C, and O-C-N are considered to classify long alpha-neurotoxins by Ward's cluster method and LIBSVM program package. Those global structural parameters of loop II Trp residues of alpha-neurotoxins are discussed.