Conus striatus venom exhibits no significant effects on potassium currents in a Xenopus oocyte expression system. Michael Hughes rimental Neurobiology Spring 2001 Advisor: William Gilly Acknowledgements: 1 would like to give a special thanks to my advisor for this project, William Gilly, who has been a constant source of advice, support and encouragement. Without his efforts, this paper would have never been written. Second, I would like to thank all the instructors and students in Experimental Neurobiology, whose hard work and commitment has been inspiring throughout the quarter. Finally, I would like to thank Joseph Schulz, who has helped me more than I can ever describe. Abstract: Conus striatus is a predatory cone snail that subdues its prey with a wide array of neuroactive peptides. The milked venom of C. striatus has been shown to induce paralytic seizures when injected into fish and repetitive action potentials in frog sympathetic ganglion cells and neuromuscular junction (Joseph Schulz, personal communication). Recently, Craig et al. (1998) reported that the primary peptide constituent of the milked venom of C. striatus is a 4kD o-glycosylated peptide, named KA-conotoxin SIVA, and proposed that this peptide acts as a low affinity antagonist of Shaker potassium channels. In order to identify a specific, high affinity target of KA-conotoxin, experiments were carried out with Xenopus oocytes injected with mRNA encoding various voltage gated potassium channels. The pharmacological effects of milked venom on these channels were assayed using a conventional, two electrode voltage clamp method. None of the potassium channels tested (Xenopus Kvl.1, 1.2, 1.3 and Shaker B A6-46, A6-46 T449V) were significantly affected by milked venom at concentrations well above the minimum amount required to induce repetitive firing of frog neuromuscular junction. These data indicate that the high affinity target of KA-conotoxin is not one of the Shaker homologues tested. Introduction: The genus Conus includes more than 500 species of predatory snails. These snails have evolved a distinctive hunting strategy in which venom filled harpoons are injected into the tissue of nearby prey. This venom has attracted considerable interest, given the incredible diversity of toxins within the genus. For example, different species of cone snails produce venom specific for animals in five different phyla (Mclntosh et al. 2000). Since each snail produces between 50 and 100 unique peptides as constituents of their venom, the number of potentially different peptide toxins manufactured within the genus Conus has been conservatively estimated at 50,000 (Bingham 1998, Jacobson 2000). This great diversity of venom peptides corresponds to an even more impressive selectivity of individual conotoxins. Neuroactive peptides from cones snails have been shown bind specific ion channels or even subunits of channels. Thus, conotoxins have the potential to differentiate the ionic currents through closely related channels within a single tissue or cell (Terlau et al. 1996). Since much of the subtlety and sophistication of our nervous system stems from the differential expression of channel proteins, these toxins may be a powerful tool for neurobiologists and medical practitioners. As a result, researchers have spent considerable efforts to identify and characterize conotoxins with novel activities. In particular, the milked venom of Conus striatus elicits a peculiar pattern of repetitive action potentials in the frog neuromuscular junction (Joseph Schulz, personal communication). Although this phenomenon has been know for some time, there has been no satisfactory explanation of the mechanism by which C. striatus toxin exerts this influence. Several research groups have identified components of C. striatus venom that appear to alter the sodium currents in frog myelinated nerve fibers (Hahnin et al. 1990 and Gonoi et al. 1987 and Kobayashi et al. 1982). However, recent evidence indicates that the firing of repetitive action potentials may also be influenced by alterations in the behavior of potassium channels. Craig et al. report that the primary peptide constituent of C. striatus is a 4kD o-glycosylated peptide, named KA-conotoxin SIVA (1998). This peptide, when isolated with FPLC, has been shown to induce repetitive firing of action potentials in the frog neuromuscular junction similar to those induced by whole venom. Furthermore, Craig et al. demonstrate that when Shaker potassium channels are expressed in Xenopus oocytes, application of the purified peptide is sufficient to reduce the peak current elicited. Based on this experiment, Craig et al. suggest that the specific vertebrate target of C. striatus venom may be a Kvl type channel. In this study, 1 attempted build upon the work of Craig et al. and identify the specific, high affinity target of C. striatus toxin in frog using a conventional two electrode voltage clamp technique in conjunction with Xenopus oocytes expressing Kvl a channel subunits. Methods: Oocyte Extraction and Culture- As previously described, adult Xenopus frogs were anesthetized with 200 mg tricane dissolved in Ca+ free water and 1 or 2 ovarian lobes were surgically removed under conditions that induce hypothermia in frogs (Keller 1991). Immediately after surgery, frogs were retuned to Ca+ free water and observed for approximately an hour or until anesthesia had completely worn off. Lobes of oocytes were treated with 10mg collagenase A in 10ml Ca++ free OR2 (Keller 1991) for 1 hour at room temperature to remove the thin layer of sheath tissue. Äfter collagenase treatment, oocytes were manually disassociated with two pairs of forceps and subjected to additional collagenase treatments at the same concentration as necessary to completely remove sheath tissue. Stage 5 and 6 oocytes with uniformly pigmented animal and vegetable poles were segregated from the ovarian tissue and rinsed 5 times with OR2 and 5 additional times with ND96 (Kllet 1991). Oocytes were cultured with 1OmM antibiotic and 10mM sodium pyruvate. Cells were carried for up to one week. lon Channel Expression- Six different ion channels were expressed in Xenopus oocytes: frog Kvl.1, 1.2, 1.3, rat Kv4.1, ShakerA6-46 and ShakerA6-46 T449V. Concentrations of injected RNA were optimized for each channel type. In general, potassium currents of 2-16 uA were achieved by diluting between 0.5 and 18 ng RNA in 50nl RNAse free water and injecting the entire solution into the animal hemisphere of a stage 5 or 6 oocyte (Rosenthal 1996). Adequate expression was normally observed within 24 hours of injection. Voltage Clamp- 2 Glass microelectrodes between 0.7 and 2.0 MQ were filled with 3M KCl and used with a conventional voltage clamp (Model, Axon Instruments). The holding potential was et at -60mV. Ooctyes were subjected to voltage pulses lasting 25 ms and stepped at 10 mV increments between -50 and 40 mV. Äfter subtracting leakage and capacitive currents on line with a standard p/-4 technique, the remaining trace was taken to be the potassium current, Ig, due to heterologous ion channel expression. Typically, 300 ul of a 1:1000 dilution of C. striatus milked venom with known biological activity was added to the chamber (approximately 100 ul total volume) using a conventional perfusion setup. Ig was recorded upon establishment of a successful voltage clamp, after a control solution change of ND96, after the addition of toxin, and after 3-4 ml of ND96 was used to washout toxin (Craig 1998). Results: Kvl family channels demonstrate quick activation. Figure 1 illustrates a representative Ig trace recorded from a Xenopus oocyte expressing Xenopus Kv1.1 channels. Like all other Kvl channels tested, Kv1.1 demonstrates quick activation and reaches peak current levels in about 5ms, at +40mV. Furthermore, every Kvl clone tested demonstrated little or no inactivation over the course of a 25 ms pulse. Frog Kvl.1 and Kvl.3 channels are insensitive to a 1:1000 dilution of milked venom. As described in methods, families of ly traces at different voltages were recorded in the absence and presence of a 1:1000 dilution of C. striatus milked venom in ND96. Figure 2a and 2c are plots of the peak lk versus voltage for Kvl.1 and 1.3, respectively. These graphs demonstrate that for à representative cell, there is no significant effect of milked venom on either Kvl.1 or 1.3. In order to average the results of multiple experiments (n-3 Kvl.1, n-2 Kvl.3), potassium conductance (gr) was plotted versus membrane voltage. Conductance was calculated from the relation 1= gk (V-Vk), where V is equal to the membrane voltage and Vg is the reversal potential for Ig. Given the technical difficulties of determining reversal potential of Ig in a Xenopus oocyte, Vg was approximated with the Nernst equation. K) RT F Kl 120 Vg =-57.17 log Vr = 100mV Thus, a value of -100mV was used to calculate conductance, and data from each cell was normalized to the maximum conductance of that cell. Figure 2b and 2d show the gg-V relations for Kvl.1 and 1.3. As expected, there is no significant difference in conductance profiles in the presence and absence of milked venom. Shaker clones are insensitive to milked venom. Given the insensitivity of frog Kvl channels to milked venom, further experiments were performed to test the effects of C. striatus milked venom on Shaker channels. Both clones tested, Shaker A6-46 and A6-46 T449V were completely insensitive to milked venom (n—4 and n-3, respectively). Concentrations as high as 1:100 elicited no appreciable attenuation of peak current levels, even when the oocyte was exposed to toxin for over 20 minutes (figure 3). Shaker channels are sensitive to Yellow Stuff. Given the insensitivity of Kvl channels to C. striatus milked venom, it became necessary to test the experimental manipulations to insure that there was no significant toxin effect. To this end, Yellow Stuff, an alternative toxin with known effects on Shaker channels was added to Xenopus oocytes expressing Shaker channels. In contrast to C. striatus milked venom, the addition of a 1:100 dilution of Yellow Stuff was shown to have a significant pharmacological effect on the peak currents of Shaker potassium channels (Kelly et al. in press). Figure 4 is a representative experiment in which peak current levels were shown to diminish by as much as 40 percent. This result is in partial agreement with the study by Kelly et al., in which Yellow Stuff was shown to diminish peak current levels and slow activation kinetics in cells expressing Shaker. Discussion: The results of the Yellow Stuff experiment indicate that the methods used in this study are sufficiently sensitive to observe a significant pharmacological effect. This procedural control demonstrates that the perfusion system, recording electrodes, and the ion channel expression system were adequate to observe significant changes in the behavior of Shaker potassium channel. Thus, Kvl.1, Kvl.3 and Shaker potassium channels Xenopus oocytes have been shown to be insensitive to C. striatus milked venom. These channels demonstrated no appreciable changes in potassium currents even at concentrations of milked venom 10 to 100 times greater than that required to induce repetitive firing of action potentials in frog neuromuscular junction. Moreover, given the absence of even a moderate alteration of conductance in the présence of toxin, we can conclude that the Kvl family most likely does not contain the specific, high affinity target of C. striatus venom. Although this statement cannot be proven to be absolutely correct until every Kvl channel has been tested for activity, the weight of evidence at this point suggests that the high affinity target is another type of voltage gated potassium channel, or perhaps a different type of channel altogether. These results disagree with the results of Craig et al. However, based upon the nearly absolute insensitivity of Shaker channels to the presence of milked venom at the concentrations tested, it seems possible that the decline of potassium currents in the presence of C. striatus peptide reported by Craig et al. is due to a non-specific, low affinity interaction seen at high peptide concentrations that has little to do with the physiological effects of this conotoxin on frog or fish neurons. Every Shaker clone tested in this study contained a deletion of amino acids 6 to 46. This deletion resulted in potassium currents that were insusceptible to fast, N-type inactivation. One of the clones, ShakerA6-46 T449V also did not display strong C-type inactivation. Since the reported effects of C. striatus venom dealt exclusively with the peak conductance, it is unlikely that the absence of inactivation has confounded these experiments. However, I cannot say with absolute certainty that C. striatus venom has no effect on true, wild-type Shaker channels with inactivation, i.e. the channel studied by Craig et al.. Since amino acids 6-46 are on the interior of the channel protein and thus unlikely to contact an extracellular toxin, the weight of available evidence suggests that the high affinity target of C. striatus milked venom is not a Kvl channel. Refrences: Bingham, J.B. (1998) "Novel Toxins from Conus." Ph.D. Thesis, University of Queensland. 155-184. Chesnut, T.J., Carpenter, D.O. and Strichartz, G.R. "Effects of venom from Conus striatus on the delayed rectifier potassium current of Molluscan Neurons. (1987) Toxicon 25, 267-278. Craig, A.G., Zafaralla, G., Cruz, L.J., Santos, A.D., Hillyard, D.R., Dykert, J., Rivier, J.E., Gray, W.R., Imperial, J., DelaCruz, R.G., Sporning, A., Terlau, H., West, P., Yoshikami, D., Olivera, B.M. “An O-glycosylated neuroexcitatory Conus peptide" (1998) Biochemistry 37, 16019-16025. Gonoi, T., Ohizumi, Y., Kobayashi, J., Nakamura, H., and Catterall, W.A. “Actions of a polypeptide toxin from the marine snail Conus striatus on voltage-sensitive sodium channels" (1987) Molecular Pharmacology 32, 691-698. Hahnin, R., Wang, G., Shapiro, B.I., Strichartz, G. “Alterations in sodium channel gating produced by the venom of the marine mollusk Conus striatus" (1991) Toxicon 29, 245- 259. Jacobson, R.B., Koch, E.D., Lange-Malecki, B., Stocker, M., Verhey, J., Van Wagoner, R.M., Vyazovkina, A., Olivera, B.M., and Terlau, H. "Single Amino Acid Substitutions in kConotoxin PVIIA disrupot interaction with the Shaker K+ channel" (2000) Journal of Biological Chemistry 275, 24639-24644. Keller, R. (1991) Methods in Cell Biology 36, 45-61. Kobayashi, J. Nakamura, H., Hirata, Y., Ohizumi, Y. "Isolation of a cardiotonic glycoprotein, striatoxin from the venom of the marine snail Conus striatus" (1982) Biochemical and Biophysical Research Communications 4, 1389-1395. Lirazan, M.B, Hooper, D., Corpuz, G.P., Ramilo, C.A., Brandyopadhyay, P., Cruz, L.J., Olivera, B.M. "The spasmodic peptide defines a new conotoxin superfamily" (2000) Biochemistry 39, 1583-1588. Mclntosh, J.M., Corpuz, G.O., Layer, R.T., Garrett, J.E., Wagstaff, J.D., Bulaj, G., Vyazovkina, A., Yoshikami, D., Cruz, L.J. and Olivera, B.M. "Isolation and characterization of a novel Conus peptide with apparent antinociceptive activity" (2000) Journal of Biological Chemisty 42, 32391-32397. Mclntosh, J.M., Santos, A.D., Olivera, B.M. “Conus peptides targeted to specific nicotinic acetycholine receptor subtypes" (1999) Annual Review of Biochemistry 68, 59-88. Rosenthal, J.C., Vickery, R.G. and Gilly, W.F. "Molecular identification of SqKvlA: A canidate for the delayed rectifier K channel in squid giant axon" (1996) Journal of General Physiology 108, 207-219. Terlau, H., Shon, K., Grilley, M., Stocker, M., Stuhmer, W. and Olivera, B.M. "Strategy for rapid immobilization of prey by a fish hunting marine snail" (1996) Nature 398, 148-151. Figure Legends: Figure 1: IV trace. A. Plot of current versus time in a typical voltage clamp experiment in Xenopus oocytes expressing frog KV1.1 channels. B. Schematic diagram of the voltage and duration of pulses applied to each cell. Figure 2: Toxin Response Curves in Frog KVI.1 and KVI.3. A. IV curve of a representative voltage clamp experiment in which a 1:1000 dilution of striatus milked venom was added to an oocyte expressing KVI.1. B. GV curve illustrating the averaged response of 3 independent experiments. C. IV curve of a representative voltage clamp experiment in which a 1:1000 dilution of striatus milked venom was added to an oocyte expressing KVI.3. B. GV curve illustrating the averaged response of 2 independent experiments. Figure 3: Toxin Response in Shaker potassium channels. A. IV curve of a representative voltage clamp experiment in which a 1:1000 dilution of striatus milked venom was added to an oocyte expressing Shaker d6-46 T449V. B. IV curve of a representative voltage clamp experiment in which a 1:100 dilution of striatus milked venom was added to an oocyte expressing Shaker d6- 46 over a 20 minute time course. Figure 4: IV curve of a Xenopus oocyte expressing Shaker d6-46 potassium channels in the presence and absence of Yellow Stuff. Figure 1: 5 UA -50 m 10 ms +40 m — Om s k -40 m 25 ms Figure 2: Kv 1.1 IV Curve Cell1 • Control 6 1:1000 Miked Venom Washout -100 -50 0 Membrane Voltage (mV) Kv1.3 IV Curve —15 • Control +1:1000 Milked Venom Washout -100 -50 Membrane Voltage (mV) Kv1.1 GV curve e Control 41:1000 Toxin 9.2 -50 0 -100 Membrane Voltage (mV) D. Kvi.3 GV Curve • Control 8 04 +1:1000 Milked Venom —0.2 Washout -100 Membrane Voltage (mV) Figure 3: Shaker449dV 1:1000 2.5 1.5 1 2 0.5 2 8 80 -0.5 mV -Holding Potential B. Shaker IV curve 2.5 2 1.5 0.5 O -60 -40 -20 20 Membrane Potential (mV) 100 40 60 120 • Control Control Pre-Pulse 1:1000 Toxin X Toxin after 15 min X Washout • Control +1:100 Milked Venom +20 minutes x Washout Figure 4: 25 20 15 10 -60 Shaker IV curve -40 -20 Membrane Voltage (mV) Control 1:100 Yellow Stuff Washout