Abstract The shapes and frequencies of action potentials of Doriopsilla albopunctata neuronal soma were studied in artificial sea water (ASW), and ASW solutions that were Na -free, Caz-free with Co¬, or with tetraethylammonium jons (TEA). Action potentials were also measured in ASW solutions containing 67 micromolar forskolin, 1 micromolar chlorophenylthiol CAMP (CPT CAMP), or 50 nanomolar tetradecanoyl-phorbol-acetate (TPA). The Nat-free, Ca2-free Co2*, and TEA solutions were used to study action potentials missing Na’, Caz*, and K' currents respectively. Na’-free solution caused action potentials to decrease in amplitude and increase in width. Action potentials became slight undulations or sine wave like. Ca2'-free solution removed repetitive firing. Potentials also lost amplitude and gained width. TEA removed the undershoot of hyperpolarization as well as increased the width of action potentials. he addition of both TEA and Ca2-free Co2 solutions to cells produced and two stage effect where the voltage fluctuated between a resting potential and a potential 40 mV above rest. Forskolin and CPT CAMP were used to observe the putative effects of CAMP on action potentials. Forskolin increased action potential width and seemed to decrease threshold voltage. CPT CAMP had no observable effects. TPA was used to determine the putative effects of diazoglycerol. It seemed to decrease threshold voltage in Na free solution. These results suggest that action potential shape is determined by varjous current components and that second messengers may affect action potential shape by blocking or enhancing certain currents. Introduction It has been hupothesized that second messengers plag a role in changing rates of ion currents in neurons thus changing action potential shape resulting in different amounts of transmitter released from sunaptic terminals. Klein and Kandel have suggested a model where CAMP levels may control sensitization in Aplysia gill withdrawal response. Accordinq to this model, serotonin binding to receptors activates adenulate cuclase leading to an increase in cAMP levels in the presynaptic terminal of the sensoru neuron. CAMP activates a cAMP-dependent protein kinase which phosphorglates a voltage sensitive K* channel, inactivating it. The subsequent decrease in K' current prolongs the time of depolarization during action potentials, allowing more calcium jons to enter with the result that more transmitter is released from the terminal (1). Castellucci et al. (2) and Kaczmarek et al. (3) have injected the catalutic subunit of bovine cAMP-dependent protein kinase into Aplusia sensory neurons and bag cells respectivelg. Action potential height and width were incressed. Results seem to indicate an increase in Ca¬ current and a decrease in K' current were responsible for this effect. TPA (12-0-tetradecanoul-phorbol-13 -acetate), a tumor promoting phorbol ester, has also been reported to increase action potential height in Aplusia bag cells. Presumably, it acts by mimicing the effects of diszoglucerol which activates protein kinase C. Injection of purified protein kinase C also increases action potential height by increasing Ca¬ current (4). Ca channels may be phosphorglated by this kinase thus bringing more channels from an inactive to an active state. In this paper, the effects of forskolin, chlorophenglthiol cAMP, and TPA on action potential shape in the mollusc Doriopsilla albopunctata were studied. Forskolin is a membrane soluble diterpene which stimulates the increase of cAMP levels by activating adenglate cyclase. It also seems to have the effect of increasing the slow inward Na' current in molluscan neurons which is not medisted by CAMP (5). CPT CAMP is a nonhgdolyzable membrane permeable analog of cAMP. To try to determine what types of currents were affected bu each of the chemicals, action potentials were observed in solutions where Na ions were removed, Ca ions were removed, or tetraethulammonium ions (TEA) was added. This allowed the observation of action potentials with certain currents blocked. Materials and Hethods Doriopsilla albopunctata were obtained from Tom Otis who dove foi them in the Monterey Bay area. They were kept in a tank through which filtered sea water was continuouslg run through. The sea water was put through a degassifier which consisted of a tube filled with plastic beads in order to prevent gas from coming out of solution inside the animals due to the difference in temperature between the sea and the tank. Preparation of Brains Doriopsilla brains were dissected by making two incisions from the gill to each rhinophore using scissors. The flap of skin was pulled up exposing the brain which could be easilg identified because of its orange color, two black eues that were attached to it, and because of its position between and slightlu posterior to the rhinophores. The brain and surrounding muscle tissue were removed using fine scissors and pinned to a petri dish with a silgard base. They were then rinsed in an artificial sea water solution (ASW) (see Table 1). Excess tissue such as salivarg gland tissue was removed using fine forceps. Next, the brain was treated with a small amount of dispase, covered with a small piece of kimwipe to preserve moisture and incubated for 1 hour at 4 degrees centigrade. he dispsse was rinsed off and the brain was incubsted for at lesst 2 hours in artificial sea water also at 4 degrees centigrade. This procedure loosened the epineurial sheath surrounding the ganglions. The sheath was later torn off using fine forceps. Intracellular Recordings Recordings were done on either the three large cells of the pedal ganglions or the large posterior cells in the pleural ganglions. Iwo electrodes filled with 3 molar KCI with resistances of between 2 and 15 megachms were inserted into the cells using manipulators. The electodes were connected to amplifiers. One electrode was used to regulate current, while the second was used to record voltage. Current and voltage were monitored using an oscilloscope, and a computer was used to program current pulses. During the recording procedure, the brain was pinned in a perfusion dish on a cooling plate at about 15 degrees centigrade. The perfusion dish was designed so that the bathing solution could be changed while keeping the brain totally immersed with minimal damage to the cells. The insertion of the electrodes was done under a light microscope at a magnification of 500 times. Bathing Solutions Na'-free, Ca-free Co-*, and TEA solutions are listed in Table 1. Half Ca2-free Co*, half TEA solution was made by mixing equal volumes of Ca2-free Co* and TEA solutions. TPA was made in a 50 micromolar stock solution in dimethul sulfoxide (DHSO) and added to the cells so that the final concentration was 50 nanomolar. Forskolin was stored as a 10 millimolar stock in DHSO and used at a final concentration of 67 micromolar. Stock CPT CAMP was 1 millimolar in distilled water. It was used at a final concentration of 1 micromolar. Solutions were added to the cells by perfusion. Chemicals were added directly into the bathing solution and wash out by perfusion. Results Figure 1. shows control action potentials in normal artificial sea water. Cells had resting potentials between -20 and -50 mV. The peaks of action potentials were between 130 and 150 mV. Cells fired repetitivelu when stimulated by long current injections. In Na -free solution, the resting potentials consistently became more negative to approximately -70 mV. During the replacement of ASW with Na -free, spike frequency decreased, spike height decreased, and spike width incressed. The rate of voltage increase of the rising phase of the action potential decreased and the rate of voltage decrease of the falling phase decreased. When perfusion was completed, repetitive firing was either abolished after the first action potential or verg slight undulations followed the first action potential. Figure 2. shows action potentjals in Na’-free solution. The upper traces show an action potential during perfusion, and the lower traces show an action potential after perfusion was complete. These effects were all reversible by washing with ASW. The effects of Ca+-free Co* solution is shown in Figure 3. Spike height was decreased and spike width increased although not as much as in Na'-free solution. Repetitive firing was decreased with a long current pulse causing two spikes, the second much shorter and wider than the first, and then a flat line. TEA caused a major incresse in the width of action potentials. Action potential width at half maximal height changed from control values of approximatelu 12 mS to approximately 180 mS. Resting potential became more positive by approximately 5 mV, and peak action potential height decressed slightlg. The huperpolarization at the end of a spike was abolished (see Figure 4.). When the TEA solution and Ca--free Co* solution were mixed, an interesting effect was observed in besting cells, shown in Figure 5. Following a spike, a cell never huperpolarized back to resting potential but huperpolarized to a voltage approximately 40 mV obove rest. A verg weck ad f. dimdfld second spike was obser veu 10n0weu ug à sever di seconu fiat intei vai during which the voltage was slowlg but consistently decreasing. After reaching a certain voltage, slight action potential like spikes began appearing growing in size (both action potential peak voltage was rising and lowest voltage was becoming more negative). Äfter these potentials resched a certain height, voltage returned to resting potential. This process was repeated in a beater like fashion. At resting potential, a definite membrane resistance could be measured but during the time that the cell was at the 40 mV above the resting stage, membrane resistance, as messured by injecting a huperpolarizing current, decreased to the extent that a 1 nA injected current had an immeasurable effect on voltage. The addition of ASW to the TEA/Ca-free Co 2 solution decreased the amount of time spent in the higher voltage state. It could also abolish the higher voltage state totally and insteed, make the cell fire repetitivelg for a short interval followed bu an interval where the cell did not fire spontaneously similar to a bursting pacemaker cell. Forskolin increased resting potential (ie. from -22 to -12.9 mV) and in some cases may have increased action potential frequency. Action potential width at half meximal height was increased approximatelg 300 percent. Observations were made where low current pulses (ie. 1 nA) did not change voltage in control (ASW) cells while higher current pulses (ie. 5 né) produced action potentials and repetitive firing if maintained. Atter cells were treated with forskolin, the lower voltage produced action potentials as well as repetitive firing if maintained while the higher current pulses produced a weaker action potential followed by inhibition of firing. This effect was observed in all four trials but the results mag be misleadinq since in all cases the same current pulses were used regardless of resting potential height. Since forskolin raised action potential levels, injecting equivalent currents mag not have resulted in same absolute voltages compared to controls. Only in one case was the cell huperpolarized to its previous resting potential before being tested. This cell acted in the same way as the others. Forskolin consistently increased the width at half maximal height of action potentials: from 10 mS to 30 mS. It consistently caused a décrease in height of action potentials: from 85 mY from peak to bottom of huperpolarization to 65 mV, and in one out of four trials, it noticeably increased action potential frequency during a stimulus. CPI CAMP was added in concentrations of 1 micromolar. At this concentration, no effects were observed measuring action potential frequencu, height, or width. Resistance of the cell also did not change. Four experiments were done with CPT CAMP. All gielded the same results. Adding TPA at a concentration of 50 nH to cells in ASW produced no noticeable effect on action potential frequency and no consisted etfects on action potential height or width. Adding TPA to cells in Na'-free sea water stimulated action potential production over control levels at low voltages. Cells in Na’-free sea water had resting potentials of approximatelu -70 mV. At this potential, pulses of up to 5 nA produced no action potentials in control cells. Cells treated with TPA however, produced low frequency firing as a result of these pulses. When cells were depolarized to higher resting potentials, ie. the resting potential of the cell before addition of Na’-free solution (appoximatelg -50 mV), control cells and TPA treated cells both produced similar numbers and frequencies of action potentials (see figure 8.). Discussion It has been observed that in molluscan cell soma, the major currents involved during action potentials are carried by Na“, Ca2*, and K* ions (6). Hat and Ca2t currents are necessarg for the rising phase as Na“ and Laz- jons enter the cell and Kt ions leaving the cell account for the subsequent huperpolarization during the potential. The experiments in which Nar¬ free. Ca2t-free, and TEA sea water were used to observe action potentials in the selective absence of these ions seems to verify this. In all cases, action potential shape was changed drastically. In Na“-free solution, the sine wave like undulations indicate that Ca2* current has a slower activation and a slower deactivation than Hat current. The repetitions of the undulations seems to indicate that in these cells, Ca¬* current may be necessary for repetitive firing. In contrast, with the Cazt free solution, Nat current turned on rapidly and deactivated rapidlg. The difference in the time it takes from the peak of the action potential to the repolarization is different. Since K* current should be the same in both cases, this difference can be accounted for by the rapid deectivation of the Nat channels. In Ca2t-free solution, there are one or two sharp action potentials followed by a flat line. This seems to indicate that Ha“ current bu itself cannot maintain repetitive firing. A verg simplified model of an action potential would be that the Na* current plays a major role in the quick upward and downward slopes of the spikes while Caz* current plags a major role in the repetition of the spikes. The summation of the voltages during a spike in Na“-free and a spike in Ca2t-free solutions does not result in the shape of a "normal“ action potential in ASW. This indicates that another factor is necessarg for achieving action potential shape. This factor mag be the negative or positive feedback that the currents experience as a result of the change in potential. As the voltage increases, the probability or a Cazt or Na channel being open also increases. Ädding the spikes observed in Na“-free and Ca2t-free solutions increases the voltage which in turn changes Na and Ca2t currents. To obtain a normal action potential shape by the addition of Nat and Ca2* currents, these currents should be measured at voltages ranging from resting potential to maximum spike height. As the Nat-free curves and the Ca2t-free curves are added to form a higher voltage, the Nat and Ca2* curves for the new voltage should be used to continue Tetraethul ammonium ions block K* current. The addition of IEA decreased the rate of hyperpolarization during an action potential as would be expected if K* current was hindered. The slowed huperpolarization that remained was probebly caused by the inactivation of Nat and Ca2t current. The undershoot at the end of an action potential was abolished. This also was expected since this is caused by increased Ktcurrent. The slight lowering of the peak of the action potential was not expected and does not seem to be a result of the direct action of the K* ions. It can be explained indirectly by the result that adding TEA reduces some resting Kt current resulting in an increased resting potential (a 5 my increase in resting potential was observed when cells were in TEA solution). This increased resting potential mag prevent some of the Na channels from going to the closed state from the inactive state (Na¬ channels need a huperpolarized voltage to go the the closed state from the inactive state). Having more Nat channels in the inactive state would prevent an action potential from acheiving maximum height. Also, the absence of the huperpolarization mag have the same etfect. The addition of the combination of TEA solution and Ca2*-frée Coz* solution produced an interesting effect in beating cells. The cell seems to be in an unstable equilibrium between two states: the resting state and a state approximatelu 40 mV above the resting state. The resting state can be explained bu the permeability of the membrane to potassium ions. Although TEA blocks some Kt channels, mang K* channels necessarg for the maintenance of the resting potential are not effected by TEA. Also, in the combination solution used in this case, the TEA has been diluted bg Lazr¬ free and ASW solutions. The state in which the cell is at a potential 40 my above resting can be explained by the opening of Ca2* channels. This is unexpected since the presence of Ca2t-free Co would be expected to block Ca2t channels. However, the fact that the resistance of the cell at the 40 my above resting state is verg low compared to the resistance at resting shows that this effect cannot be caused solely by a decrease in K conductance. An increase in Nat conductance would explain this phenomenon except that voltage gated Nat channels would be expected to inactivate instead of staging open so long. Therefore, this second state seems to be explained best by a state achieved by the equilibrium of the conductances of both Kt channels and Ca2* channels, both of which have been decreased by TEA and Ca2+-free Co. Since the cell originally was a beater, resting potential is inherentig unstable. Slow inward current or B current probablg depolarizes the cell past threshold voltage. The first two spikes can be attributed to Na“ current and inactivation. However, since voltage dependent K* current is inhibited to an extent by TEA, the voltage does not return to resting potential but reaches this second state where the equilibrium is dependent on a balance between Ca2t and Kt current. At this point, Na“ current is blocked since the Nat channels are trapped in the inactive state by the depolarization. This second state is not stable since the voltage has been observed to decrease steadily. This decrease can be explained by factors such as the Nat/Kt pump, slow Ca2* current inactivation, and an increase in Ca2t dependent K* current as Ca2* concentration rises in the cells. As voltage decreases, more and more Na“ channels are transferred from the inactive to the closed state. Since voltage is above threshold at this stage, these Nat channels open and close causing fluctuations that increase in amplitude as voltage decreases. During this time, Cazt current continues to inactivate until Kt current overwhelms it and brings it back to resting potential. B current brings the voltage back to threshold again causing the cycle to repeat. The time spent in the second state is longer than the 300 ms that the cell is above restinq potential when it is in a TEA solution. This indicates that some other factor besides the decrease in K* current and the inactivation of Ca2t current is involved in the cell remaining in the second state. It may be that Co2t has an effect of leaving some Ca¬ channels open longer or blocking a different subset of K* channels than TEA does, although both of these hypotheses are speculation. Forskolin was expected to behave in a wag similar to TEA. It was expected to lead to increases in cAMP dependent protein kinase which would phosphorglate voltage dependent K* channels thus blocking them, leading to a result similar to that observed when TEA is added. The increase in resting potential observed when forskolin was added is similar to the effect observed in TEA solution and mag be the result of decreased Kt conductance. The increase in action potential width also agrees with the possibilitu that K* channels were blocked. The decrease in injected current needed to produce cell spiking mag imply that forskolin lowers the threshold voltage of cells. However, it mag also be an artifact resulting from the incresse in resting potential of the cell when exposed to the drug. Hore experiments must be done to confirm which of the two is correct. In the literature, there is no mention of forskolin affecting threshold voltages of cells. Since forskolin has been observed to have effects other than those mediated bu CAMP, such as changes in A current or transient K* current, chlorophenulthiol CAMP, a membrane permeable cAMP analog, was used to tru to duplicate its effects and separate which effects were CAMP mediated and which were not. 1 micromolar concentrations were used and no changes in spike height, width, or frequency were observed. It is possible that changing CAMP levels has no effect on action potential shape or frequency in Doriopsilla although this seems verg unlikelg, since cAHF is such a ubiquitous second messenger and has been shown to affect spike shape in Aplusia neurons. The more likelg possibilitg is that 1 nH concentrations of CPT CAMP are too low to cause an effect. Higher concentrations should be used for further experimentation. 14 The effects of TPA in Nat-free sea water indicate that TPA lowers the threshold voltage of a cell. Spiking occurs in the presence of TPA at voltages where in its absence, it does not occur. However, at voltages where spiking occurs in both, frequency remains unchanged. In ASW, observations were only made at voltages where spiking occurred in both cases. Threshold was not measured so further experiments will have to be done to see if the lowering of threshold voltages occurs in ASW as well as Nat-free sea water. An incresse in spike height and width such as has been observed in Aplusia neurons was not observed. A possible explanation for the decreased threshold voltage observed in Nat-free solution with IPA is that TPA activates protein kinase C which phosphorglates Caz channels bringing them from the inactive to closed state. This increases the number of Ca2t channels that may be opened. Although the probability of ang one channel being open at a particular voltage mag not change, since there are more channels available, more of them will be open at a particular voltage thus lowering threshold, similar to the idea that the threshold voltage is lower at the spike initiating zone because there is a higher concentration of Nat channels there. Conclusions Although the results of this experiment are far from conclusive, it seems likelu, from the current literature, and from some results here, that second messengers do have an effect on action potential shape, frequencg, or threshold. Neurons do not just passivelg transmit action potentials. Il this were the case, learning and memorg would not occur. Neurons must integrate and modifu the information that is transmitted to them. Second messengers, among other things, seem to plag a role in this moditication. Using second messengers, models for simple forms of learning, such as the Klein and Kandel model for Aplusia sensitization, can be made. Therefore, the effects that second messengers have on neurons should be fürther studied. Acknowledgements would like to thank Dr. Stuart Thompson for his helpful advice and ideas and for teaching me the techniques involved in Doriopsilla dissection and intracellular recording. Also, I would like to thank Drs Baxter, Epel, Swezeu, and Dennu, the postdoctoral fellows and graduate student in the Thompson lab, and Tiffanu Tom for their advice and concern, and Tom Utis for the Dorids. Figure Legends Figure 1. Action potentials in ASW. Upper trace is injected current; lower trace is voltage. Figure 2. Action potentials in Na free sea water. The upper two traces are input current and voltage measured during perfusion to replace ASW with Na free sea water. The lower traces are input current and voltage measured after perfusion was complete. Figure 3. Effects of Ca free Co sea water on action potentials. The upper two sets of current and voltage traces show input current and action potentials in ASW. The lower sets show input current and action potentials in Ca free Co solution. Figure 4. Action potential in TEA containing sea water. Upper trace is current; lower trace is voltage. Figure 5. Action potentials in a mixture of Ca free Co sea water and sea water containing TEA. The upper two voltage traces show the two different states that the voltage fluctuated between: resting voltage, and 40 mV above resting. The lowest voltage trace shows how diluting the Ca free Co and TEA mixture with ASW results in a burster like activity of the cell. Figure 6. Effect of forskolin on action potentials. The upper two traces are current and voltage in ASW. The lower two are current and voltage in ASW with forskolin added. Figure 7. The effect of TPA in Na free sea water with resting voltage at - 70 mV. The upper two traces are current and voltage in Na free solution without TPA. The lower two traces are current and voltage in Na free solution with TPA. Figure 8. The effect of TPA in Na free sea water with resting voltage depolarized to -50 mV. Upper two traces are in Na free solution - TPA; lower two are in Na free solution +TPA. 18 gure 1. 19 Figure 2. 20 Figure 3. Figure 4. Figure 5. 23 Figure 6. 24 Figure 7. Figure 8. 26 Nacl KCI Caci rigc12 Hepes Tris CoCl2 TEA-CI Table 1. Composition of bathing solutions in millimolars A5W Na free La free C0 TEA 470 370 470 10 10 10 10 50 50 10 470 100 References 1. Kandel, E. R., Schwartz, J. H., Principles of Neuroscience Elsevier Second Edition (1985). 2. Castellucci, V. F., Kandel, E. R., Schwartz, J. H., Wilson, F. D., Nairn, A. C., and Greengard, P., Proc. Natl. Acad. Sci. USA. 77, 7492-7496 (1980). 3. Kaczmarek, L. K., Jennings, K. R., Strumwasser, F., Nairn, A. C., Walter, U., Wilson, F. D., and Greengard, P., Proc. Natl. Acad. Sci. USA 77, 7487¬ 7491 (1980). 4. DeRiemer, S. A., Strong, J. A., Albert, K. A., Greengard, P., and Kaczmarek, L.K., Nature 313, 313-316 (1985). 5. Coombs, J., Thompson, S. H., J. Neurosci. 7, 443-452 (1987). 6. Adams, D. J., Smith, S. J., Thompson, S. H., Ann. Rev. Neurosci. 3, 141- 167 (1980).