Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comparison of drug blockade of a neuronal calcium-activated potassium channel with cardiac repolarizing… Tong, Clement Tsz-Ming 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0639.pdf [ 2.27MB ]
JSON: 831-1.0087684.json
JSON-LD: 831-1.0087684-ld.json
RDF/XML (Pretty): 831-1.0087684-rdf.xml
RDF/JSON: 831-1.0087684-rdf.json
Turtle: 831-1.0087684-turtle.txt
N-Triples: 831-1.0087684-rdf-ntriples.txt
Original Record: 831-1.0087684-source.json
Full Text

Full Text

COMPARISON OF DRUG BLOCKADE OF A NEURONALCALCIUM-ACTIVATED POTASSIIJM CHANNEL WITH CARDIACREPOLARIZ1NG POTASSRJM CHANNELS BY POTENTIALCLASS ifi AGENTSbyCLEMENT TSZ-I’ilNG TONGB.Sc., The University ofBritish Columbia, 1992A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEIITHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF PHARMACOLOGY & THERAPEUTICSWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAOctober 1994© Clement Tsz-Ming Tong, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)______Department of,The University of British ColumbiaVancouver, CanadaDate (. . . IDE-6 (2/88)11ABSTRACTThe aim of the study was to examine the use of a repolarizing calciumdependent potassium current K(Ca) in neurons to describe the actions of a groupofnovel compounds with potential Class ifi actions on cardiac cells. This wasaccomplished by determination of the correlation between potency of the agentsto block single channel K(Ca) and potency to prolong effective refractory period(ERP) in heart. If a positive correlation could be established then elucidation ofmechanisms of drug actions on single channel K(Ca) could have utility in thedescription of drug actions on repolarizing K+ currents in heart. At present thelow unitary conductance of transient outward and delayed rectifier K+ channelsprecludes a mechanistic analysis of drug actions on cardiac cells.Initial experiments included the measurements of the single channelproperties of the K(Ca) using inside-out patches obtained from culturedhippocampal neurons. The channel conductances, with physiological (SK+ and140K+ across patches) and symmetrical (140 K+ across patches) were 110 pSand 170 pS respectively. A requirement of 4 jiM internal calcium was necessaryto maintain maximal channel activity with a threshold for K(Ca) activation at 0.7jiM. At low internal calcium concentration, depolarization increased theprobability of channel openings. The effect was found to be solely dependent onthe voltage-sensitive increase of the channel opening frequency; mean opentimes of the channel were not dependent on patch potential.The unitary K(Ca) is the microscopic basis for the macroscopicrepolarizing current Ic in hippocampal neurons. Ic is responsible for the laterepolarization phase and the early afterhyperpolarization (AHP) phaseassociated with the neuronal action potential. It was of interest to first determine111the effects on Ic of an agent, tedisamil, with known Class ifi activity in heart.The results showed the drug both prolonged the neuronal action potential andeliminated the subsequent AHP phase.The primary set of experiments involved the investigations of the effectsof 18 RSD novel compounds on unitary K(Ca). All the compounds, except forthree, were effective in exhibiting rapid transitions in the K(Ca) from theopening state to a non-conducting state in the inside-out patches. The mean opentimes of open events were reduced but the closed time durations and the channelamplitudes were not changed. The potency of the effect of the RSD compoundson the K(Ca) was determined as the concentration required to halve the meanopen time relative to control value. According to this index the compounds werecategorized into five different groups based on potency to decrease mean opentime of K(Ca). In addition the actions of five of the RSD compounds on K(Ca)were examined using outside-out patches excised from neurons.The potency of the RSD compounds on the neuronal K(Ca) wascompared with their potency in inhibiting repolarizing K+ currents in the ratwhole heart. The index for potency used in the whole heart experiments was theconcentrations of the compounds required to increase the effective refractoryperiod by 25%. Of the 18 compounds tested, 3 were found to be inactive (noobvious effect for concentration below 5OjiM) on properties ofK(Ca). Thesesame 3 agents also were ineffective in whole heart (in excess of 2OpM). For 13of the remaining 15 agents, a positive correlation was found (with a correlationcoefficient r of 0.71) between potency to block K(Ca) and potency to prolongERP in whole heart. However, with 2 agents there was no apparent correlationfor actions in the neurons and the heart. It was also established that drugs withpersistent effects on K(Ca) (likely due to prolonged bonding to membrane sites)were also long-lasting in whole heart experiment.TABLE OF CONTENTSI PAGEABSTRACTTABLE OF CONTENTS ivLIST OF TABLES viiLIST OF FIGURES viiiACKNOWLEDGEMENTS x1. INTRODUCTION1.1 The Calcium-activated potassium channels 11.1.1 Calcium-activated potassium conductances 11.1.2 Two types ofK(Ca) channels 11.1.3 Single channel properties of the K(Ca) High permeability High selectivity Gating kinetics 31.1.4 Pharmacological profile 51.1.5 Physiological functions of the K(Ca) Repolarization Afterhyperpolarization 61.1.6 Ca2 -activated channels in hippocampal neurons 71.2 Class III antiarrhythmic drugs 91.2.1 Outward K current in heart 91.2.2 The class ifi antiarrhythmics 101.2.3 Antiarrhythmic properties of tedisamil (KC8857) 12ivV1.2.4 Putative class ifi antiarrhythmics as potent K(Ca) blockers 131.2.5 The RSD compounds 132. METHOD2.1 Tissue culture preparation 152.2 Electrophysiology - patch-clamp recordings 152.2.1 Pipette preparation 152.2.2 Pipette mounting 162.2.3 Patching 172.2.4 Patch excision: inside-out and outside-out 182.2.5 Single channel recordings 182.2.6 Data analysis 202.3 Current clamp experiment on the hippocampal slice 212.4 Measurements of the ERP (effective refractory period) 213. RESULTS3.1 Single channel properties ofK(Ca) 233.1.1 Channel conductance 233.1.2 Calcium dependence 283.2.3 Voltage dependence 333.2 Macroscopic currents with tedisamil on hippocampal slices 363.3 Pharmacology ofK(Ca) 413.3.1 The RSD compounds- potential class ifi agents 413.3.2 Inside-out patch clamp experiments 443.3.2.1 The open time analysis 453.3.2.2 The closed time analysis 503.3.2.3 The amplitude analysis 50vi3.3.2.4 The open channel block 513.3.3 A comparison of the class ifi antiarrhythmics as potent 52K(Ca) blockers3.3.3.lAplotoft’ against /D] 533.3.3.2 Index of the potency of the compounds -MOT5O 573.3.3.3 Comparison ofMOT5O with putative class ifi 59antiarrhythnics3.3.3.4 Comparison oftheMOT5o and the ERP2S values 593.3.4 The wash-off’recovery times 723.3.5 Outside-out patches 794. DISCUSSION4.1 Single channel properties of CAl hippocampal K(Ca) 824.2 Correlation of the drug effects on K(Ca) and repolarizing 83currents4.2.1 A comparison of the recovery times 864.2.2 Inside-out vs. outside-out patches 875. CONCLUSIONS 88REFERENCES 89viiLIST OF TABLESTABLE Page I1 Onward blocking rate constants (ki) for five drugs 542 MOTso values of the RSD compounds 583 A comparison of the MOT50 and ERP2S values 734 A comparison of the recovery times 805 MOT50 values of five RSD compounds in 81outside-out patchesviiiLIST OF FIGURESIFIGTJRE Page I1 Typical unitary currents of K(Ca) measured in the 25cultured CAl hippocampal neurons2 Current (I) - voltage (V) plot 273 Unitary currents measured in a patch with no [Kj 30gradient4 I-Vplot 325 Unitary currents showing Ca2 dependence of K(Ca) 356 Unitary currents showing the voltage dependence of 38K(Ca)7 Voltage dependence ofK(Ca) 408 Effects of 5 jiM tedisaniil on the action potentials 43elicited in the hippocampal slice9 Effects of a RSD compound (971) on K(Ca) 4710 Effects of 4jiM 939 on the open times, closed times 49and amplitude distributions11 The plots of t’ against [D] 5612 - 14 Effects of 971, 986, 979 on the mean open time 61- 63LIST OF FIGURES (con’t)FIGURE Page I15 - 16 Effects of 984 and 987 (group 2: potent) on the mean 65 - 66open time17 - 18 Effects of 939 and 983 (group 3: intermediate) on the 68 - 69mean open time19 Effects of 973 (group 4: low potency) on the mean 71open time20 A comparison of the MOT50 and ERP25 7521 The correlation ofMOT50 and ERP25 77ixxACKNOWLEDGEMENTSI would like to express my deep gratitude for Dr. James G. Mclamon forgiving me the opportunity to work in his laboratory and to lead me through myadventurous graduate studies.I would also like to acknowledge the following for their help in makingthis thesis possible: Dr. M.J.A. Walker for his data on the whole heartexperiments; Dr. J. Church for his data on the current clamp experiments on thehippocampal slices; Dr. Baimbridge for his supply of neuronal cultures.Thanks also to the following people worked or working in Dr.McLamon’ s laboratory, who have been giving me tremendous amount of helpthrough times: Mr. Dale Sawyer, Mr. Huang Zhongxian, Cathy and Laura.Last but no way the least, I would like to give my deepest thanks to myLord and Saviour Jesus Christ. Without whom nothing could be accomplished.May ALL the glory be to Him and the Father and the Spirit, forever and ever,Amen.11. INTRODUCTON1.1 The Calcium-activated potassium channels1.1.1 Calcium-activated potassium conductancesInternal calcium was first demonstrated to be able to regulate potassiumflux across membranes of human erythrocytes by Gardos (1958) almost 35 yearsago. A more direct approach was given by Meech & Strumwasser (1970), whoobserved that a microinjection of intracellular calcium activated potassiumconductance inAplysia nerve cells and hyperpolarized the cell membrane. Based2+ . +upon these observations, a Ca -activated K conductance, GK(Ca), waspostulated (Meech, 1978). Since then, studies using the patch-clamp (Neher andSakmann eds., 1983) and reconstitution (Miller ed., 1986) techniques havedemonstrated several different types ofCa2-dependent K channels (Blatz andMagleby, 1987). Channel conductance, calcium sensitivity, voltage dependence,and pharmacological properties (i.e. antagonists studies) have been used todistinguish between these channels.1.1.2 Two types of K(Ca) channelsIn general, two major classes ofCa2-dependent K channels common tothe excitable cells have been catergorized. One group are the voltage-dependentK channels of large unit conductance (150-300 PS) (Marty, 1981; Blatz andMagleby, 1984). These are often referred to as the “Big (B)” or “Maxi” Kchannels, which are also widespread in non-excitable cells. The single channelproperties of these have been studied in detail (for example, Pallotta et aL, 1981;Moczydlowski and Latorre, 1983), thanks to the large conductance of thechannel which yields a high signal-noise ratio in the single channel recordings.2The second group ofK channels show little (e.g. Aplysia) or no voltagedependence (e.g olfactory neurons) and are of smaller conductance ( 80 pS)(Hermann and Erxleben, 1987; Maue and Donne, 1987). They are the “Small(S)” K channels, also known as the A[{P channels due to their predominantrole in afterhyperpolarization (see later) [Some authors subdivide this classfurther into the small K channels (conductance < 5OpS) and the intermediateK+ channels (50 - 1 5OpS)]. Throughout this paper the high conductance channelwill be referred to as the K(Ca) and the small conductance channel as the SK.1.1.3 Single channel properties of the K(Ca) High permeabilityThe K(Ca) possess a large conductance that is close to the limittheoretically expected for a pore (Rile, 1984). Depolarization enhances theprobability of channel opening, leading to an increased flow of current (Pallotaet al., 1981). The conductance also varies with the extracellular [K] in a nonlinear fashion. In excised patch from cultured rat muscle, the conductance, withsymmetrical 140 K across the membrane, is as high as 220 pS. The reduction of[.K]0tophysiological levels near 5mM results in a decrease in the singlechannel conductance to about 100 pS (Barrett et al., 1982; McLamon andWong, 1991). High selectivityThe K(Ca) also have a high cation selectivity. The situation is somewhatironical since a high selectivity implies strong interactions of the permeant ionwith the selectivity ifiter (Latorre and Miller, 1983). The channel is stronglyselective for K+ over Na, with a Na /K permeability ratio of less than 0.013(upper limit 0.03; Yellen, 1984). Thus K(Ca) are as selective as the delayedrectifier ofnerve and muscle but with a conductance lox to 50X bigger. TheK(Ca), like most other types ofK channels, have a similar ionic selectivitysequence (I3latz and Magleby, 1984; Gitter et al., 1987; Singer and Walsh,1987): Ti>K>Rb>>Cs, Na, Li’. The attempts to explain this highselectivity-conductance paradox have given birth to several hypotheses. One ofthem is the design of an appropriate selectivity ifiter for the K(Ca), based on theidea that if the entire length of the pore has the same narrowness as the filter partthe conductance would be low. Thus, a hypothetical structure has been proposedfor the K(Ca) in the SR ofmammalian skeletal muscle (Yellen, 1987). Hesuggested a pore having a wide entry and exit structure, with a very short andnarrow connecting tunnel, which acts as the selectivity region. The tunnel wouldnot exceed O.Snm in length, about one-tenth of the entire membrane thickness,and its cross-sectional area will be reduced to a minimum of around 0.2nm ascompared to the pore average of 0.5nm2A second hypothesis was derived fromtwo pieces of findings: that the K(Ca) are multi-ion channels, and they contain afixed negative charge in their vestibules (MacKinnon and Miller, 1988). At highpermeant ion concentrations, ion repulsion in the multiply-occupied poreincreases the rate ofK exit, and channel conductance becomes bigger than thatin a single-ion pore (Latorre, 1986). At low ion concentrations conductance isstill high because the negative charge potential created by the chargesconcentrates cations at the channel vestibules (Villarroel and Eisenman, 1987). Gating KineticsThe channel kinetics depend only on the Ca2 activity in the cytoplasm,not on extracellular Ca2. Several studies have demonstrated that Ca2 acts as aligand (Latorre et al., 1982; Moczydlowski and Latorre, 1983; McLarnon and4Sawyer, 1993) i.e. the relationship between the calcium concentration and thesteady-state open probability is a siginoidal function. In agreement with theseobservations it was found that a Ca2 -dependent biochemical pathway, e.g. thatinvolves calmodulin or protein kinase C, is usually not involved in the regulationof the K(Ca) opening. Gating kinetics ofK(Ca) requires the binding of severalCa2 ions to fully open a channel, and the number of ions required variesbetween different tissues. Generally, in the absence of other divalent cations, theHill coefficients that best describe the probability of opening vs [Ca2+]i varybetween 2 and 4, suggesting that at least 2-4 calcium ions have to be boundbefore the channel can open (Banett et al., 1982; Moczydlowski and Latorre,1983; McLamon and Sawyer, 1993). Generally, the channel activation occurs atconcentrations ofCa2 less than 1 jiM.The studies of the K(Ca) in the cultured rat myotubules have shown thatthe channel has more than one open state (McManus and Magleby, 1988). Fourdifferent modes ofkinetic activity were identified: normal, intermediate open,brief open and buzz. For the normal mode which covers about 96% of alltransitions, it was found that at least three to four open states and six to eightclosed states, including one or more infrequently adopted long-lived shut states,were present. Some divalent ions, including Mg2 and Ni2, appliedintracellularly, potentiates the Ca2 activation of the K(Ca). This result wasreflected as a dose-dependent increase of the channel open probability (inmuscle: Golowasch et aL, 1986; in nerve: McLarnon and Sawyer, 1993). It wassuggested that magnesium, by exposing more calcium binding sites which werepreviously “hidden,” increased the interaction between calcium and the bindingsites, as well as the coupling between the occupied sites and the opening activity(Golowasch et al., 1986). Several other factors are also thought to affect thisCa2 sensitivity. One of these is the type of lipid surrounding the channel, where5negatively charged lipids are found to increase the apparent Ca2+ sensitivity(Moczydlowski et al., 1985). Another factor is the neuronal development: inmature spinal neurons, raising the [Ca2+]j increased the probability of channelopening, such a Ca2+ sensitivity was lacking in young neurons (Blair andDionne, 1985).1.1.4 Pharmacological ProfileA pharmacological tool to discriminate between the two classes ofcalcium-activated potassium channels is to use their different sensitivities to theblocking agents charybdotoxin, TEA or apamin. The K(Ca) are blocked by bothexternal and internal TEA (tetraethylammonium) at different sites, with thesensitivity to [TEAJo about 15 times higher than that to [TEA]i (Blatz andMagleby, 1984; Yellen, 1984). The K(Ca) are also blocked by nanomolarconcentrations of charybdotoxin (CTX), a venom from the mideastern scorpionLeirus quinquestriatus (Miller et al., 1985). The SK, in contrast, are resistant toboth TEA and CTX blockade (with the exception in the Aplysia neurons;Hermann and Erxleben, 1987), but are blocked by nanomolar concentrations ofapamin, a peptide from bee venom (Romey and Lazdunski, 1984; Pemiefather etal., 1985). The K(Ca) are resistant to apamin blockade.1.1.5 Physiological functions of the K(Ca)Since their original discovery in molluscan neurons (Meech, 1978),Ca2tactiv ted K channels have been found widely distributed in cell membraneand play a variety ofphysical roles. Activation of potassium currents can lead tostabilization of the membrane potential; that is, these currents draw themembrane potential closer to the potassium equilibrium potential (Ej<) andfarther from the firing threshold. Uniquely, the K(Ca) provide a feedback control6of the voltage-dependent influx of calcium, and play important roles in regulatingsecretion (Peterson and Maruyama, 1984) and smooth muscle contraction(Singer and Walsh, 1984). K(Ca) also play a role in regulating the actionpotential frequency and duration (the repolarization phase) in neurons and otherexcitable cells. RepolarizationThe K(Ca) are proposed to play important roles in the repolarizing phaseof the action potential and in the control of the slow wave activity in somesmooth muscle cells (Singer and Walsh, 1984). Tn many smooth muscles, thecomplement of ionic currents that generate the electrophysiological responsesappears to be very similar. The K(Ca) (>200pS) are often the predominantcationic outward current, and generally coexist with smaller conductance SK(Tomita, 1988). K(Ca) repolarize smooth muscle cells after an increase in theintracellular Ca2 during, and (or) subsequent to, the upstroke of the actionpotential (Akbarali et al., 1990a). In tissues where action potentials are rare ornonexistent, such as the trachea (Kirkpatrick, 1975) and the esophagus (Akbaraliet al., 1990b), the intracellular Ca2+ that activates the K(Ca) current may comefrom the SR. A K(Ca) was also found to play a role in the action potentialrepolanzation in the rat vagal motoneurones (Sah and McLachlan, 1992) and inthe bullfrog sympathetic ganglion cells (Adams et al., 1982). AfterhyperpolarizationIn order to contract spontaneously mammalian myotubes have to generatespontaneous action potentials and an afterhyperpolanzation phase (AHP) is animportant component of the action potential. It is during the hyperpolarizationphase that the sodium channels which have inactivated become reactivated7again, thus allowing the generation of a new action potential. Over the pastdecade the importance ofCa2 - dependent A1{Ps in a number of cell types werewell documented and the coexistence of two calcium-activated potassiumcurrents have been reported in rat skeletal myotubules (Barrett et al., 1981),Aplysia neurons (Deitmer and Eckert, 1985), bullfrog sympthetic ganglion cells(Pennefather et al, 1985), GH3 cells (Ritchie, 1987), and rat sympatheticneurons (Smart, 1987). Generally, the large conductance, voltage dependent,and TEA sensitive K(Ca) contributes to the early phase of the AHP. The slowerCa2 - dependent K current (SK) is not affected by extracellular TEA but isblocked by apamin in a number of cells (e.g. bullfrog ganglion cells, Pennefatheret al. 1985; GTh cells, Ritchie 1987). The SK current mediates the slow AHPand the prolonged hyperpolarization phase. The accumulating evidence is thateach of the two Ca2 - activated K currents underlies a different component ofthe AHP. Since the K(Ca) current is voltage dependent it tends to turn-offrapidly at voltages close to resting potential and at physiological concentrationsofCa2. It is thus associated with a fast AHP. On the other hand, since the SKcurrent has little voltage-dependence, its decline may be more closely related toCa2 diffusion away from the membrane. It is active at lower Ca2concentrations and underlies the slower long-lasting AHP.1.1.6 Ca2+activ ted channels in bippocampal neuronsIn the CAl hippocampal neurons, the macroscopic current Ic, whichcorresponds to the activation of K(Ca), is thought to be the major repolarizingcurrent and responsible for the fast AHP (Storm, 1987). Using the single-electrode voltage clamp method on the CAl pyramidal cells in the rathippocampal slices, Storm was able to show that both the spike repolarizationand the fast AHP were sensitive to external K concentration and were inhibited82+ 2+ 2+ 2+by Ca -free medium or the divalent cations Co , Mn and Cd . The currentswere also sensitive to an inhibition by external TEA (0.5-1mM) and 3OnM ofcharybdotoxrn (CTX), but not noradrenaline. Thus CTX and TEA can increaseaction potential duration (Lancaster et al., 1986; Storm, 1987). A second slowercalcium-activated potassium current, thought to be mediated by the activation ofSK, was identified as the current responsible for the slow AHP adaptation (Lin)as well as the spike frequency (Lancaster and Adams, 1986; Lancaster et al.,1991). In contrast to the Ic, the current was blocked by noradrenaline andacetyicholine, but was voltage - and TEA - insensitive (Lancaster and Adams,1986). Noradrenaline can block the long-lasting Liii’ without affecting the spikeduration (or the fast A}{P) in hippocampal pyramidal cells (Madison and Nicoll,1982; Madison and Nicoll, 1984). As a result, the spike frequency adaptation,which normally occurs with depolarizing stimuli, is severely reduced. Thus, thenumber of spikes elicited by a depolarizing stimulus is greatly increased.. At agiven temperature, the TAlir’ is at least an order ofmagnitude slower than the Ic(Lancaster and Adams, 1986). Similarly in the bullfrog sympathetic ganglioncells (Pennefather et al., 1985), Ic activates rapidly by depolarization beyondabout -4OmV, and is responsible for the last two thirds of the spikerepolarization and the very first AHP (the fast AHP). As the membrane potentialreturns to rest and beyond, Ic rapidly turns off, although an elevated internalcalcium level still remains. However, when most of Ic turns off quickly after therepolarization due to voltage gating, some K(Ca) which are located close to theCa2 channels may still be activated due to the high internal Ca2 level(Lancaster and Nicoll, 1987). This may explain the possible role of the Ic as acomponent of the medium A[{P and early spike adaptation (Storm, 1990). TheLri is much better suited to produce prolonged AHP. Activated by Ca2 influxduring the action potential, the current activates slowly over a few seconds. The9deactivation is even slower, and is dependent on the size of the current and themembrane potential (Lancaster and Adams, 1986). The TAlil’ generates the slowbut prolonged AHP which follows spike bursts and single spikes, and helps tosustain further discharge by hyperpolarizing the cell. The Lin’ is thus crucial inthe negative feedback control of the discharge activity of the hippocampalneurons. Forming only a small fraction of the total Ca2 -activated potassiumconductance, the lAin’ does not contribute to the spike repolarization. However,it is responsible for the spike frequency adaptation in the repetitive firing typicalof the pyramidal neurones (Madison and Nicoll, 1984).Recently, at least six different potassium currents have been identified inthe CAl hippocampal pyramidal cells in slices. Apart from Ic and lAnp, there arealso L, b, Ix and Tv1, and the roles of these different potassium channels weremore clearly defined (Storm, 1990). The IA (fast transient current) both activatesand inactivates rapidly, and is responsible for the early spike repolarizationbefore Ic becomes predominant. ID (the delay current) and Tic (the delayedrectifier) are thought to participate in the spike repolarization. The Tivi (Mcurrent) is slowly activated by depolarizations beyond about -6OmV and doesnot inactivate. It reduces firing rate during spike frequency adaptation andcontributes to the medium A[{P, together with Ic.1.2 Class III antiarrhythmic drugs1.2.1 Outward IC current in heartThe duration of a cardiac action potential is mainly determined by itsplateau, which is in turn maintained by a fine balance of inward and outwardcurrents. The inward current is mediated through a Na channel, a Ca2 channel+ 2+and Na -Ca exchange. The outward current is camed by a number of10different K channels, a cr channel and the Na-K ATPase. Under normalphysiological conditions three outward K currents contribute to the net outwardcardiac current. The transient outward current (Ito) undergoes rapid voltagedependent activation and inactivation, thus is important for the earlyrepolanzation phase of the spike potential. In the rat ventricular cells, thiscurrent is the major repolarizing current. The delayed rectifier K+ current (1k) isthe predominant outward current throughout the entire plateau phase. It isactivated upon depolarization from about -50 mV to 0 mV (Noble, 1984) andtwo components (slow and fast) have been described in guinea pig atrial andventricular cells (Sanguinetti and Jurkiewicz, 1990). Tn other cells, such as therabbit and cat ventricular myocytes, the Tic has only the faster component. Theinward rectifier current (Tki) is important in establishing cell resting potential aswell as the final phase of the repolarization. The current is time-dependentlyactivated on hyperpolanzation and has an inward rectification - i.e. it passescurrent only at a membrane potential negative to -2OmV, as a result, it has verylimited role during the plateau phase of the cardiac action potential. The lid is2+also enhanced with mcreased extracellular K concentration.1.2.2 The class III antiarrhythmicsEven since quinidine was used therapeutically against cardiacarrhythmias, the antiarrhythmic significance of the changes in the refractoryperiod of the cardiac action potential had been recognised (Lewis et al., 1926).However, it was decades later before the effects of the drug were found to betwo-fold: quinidine prolongs the repolarization period and also reduces themaximum rate of depolarization (Singh and Nademanee, 1985). This gives riseto two classes of antiarrhythmic compounds: the class Ic agents (e.g. flecainide,encainide) which exhibit antiarrhythmic properties by delaying conduction, and11the class ifi agents (e.g. sotalol) which exhibit antiarrhythmic properties byprolonging repolarization. The prototype of these latter class ifi agents, sotalol,prolongs repolanzation by inhibiting the delayed rectifier potassium current(with a smaller effect on the lid), and as a ,6 blocker, it slows heart rate andintranodal conduction, and prolongs the refractory period in the atrioventricularnode. The result is the prolongation of the action potential duration and theincrease in refractory periods in the atria, ventricles and His-Purkinje system(Singh, 1990). The agent which really defines the class ifi effect, however, isamiodarone. The major antiarrhythmic effect of ainiodarone is lengthening of theaction potential duration and an increase in the refractory period. This class ifieffect is important in pacemaker tissues; lengthening the action potentialduration will delay the onset of the next spontaneous diastolic depolarization,slowing down the voltage from reaching the threshold, and the cycle length oftachycardia will be prolonged. On the other hand, in nonpacemaker tissues, theincrease in the action potential duration and refractoriness will also slow downthe tachycardia. In both cases, the acceleration of tachycardia is slowed down,therefore reducing the chance of the deterioration of the tachycardia intofibrillation. Like sotalol, amiodarone also exerts major effects on the delayedrectifier current (Tic). Although beta blockade is not a major component ofaniiodarone’s actions (unlike sotalol), the inhibitory effects of the drug on heartare not specific. Thus, there is a need to search for “pure” class ifiantiarrhythmics - drugs which can selectively prolong repolarization (thuslengthening the action potential duration as well as the effective refractoryperiod) without slowing down conduction. Most of these compounds inhibit thedelayed rectffier (1k). UK 68798 (Gwilt et al., 1989), d-sotalol and E-4031(Sanguinetti & Jurkiewicz, 1990) were shown to inhibit Tic in the guinea pigmyocytes, and risotilide (WY 48986) specffically blocks Tic in the cat ventricular12myocytes but have no effects on the inward rectifier (Tici) (Folimer et aL, 1989).This class ifi effect is typically mediated by the inhibition of one or bothcomponents of the Tic, although the faster component is the primary target fordrugs For example, E-403 1 was shown to specificly block the fast component ofthe Tic in the guinea pig ventricular myocytes (Colatsky et al., 1990). The classIll actions of some other agents however are directed against the ha. Forinstance, RP 58866 and its enantiomer RP 62719 (terikalant) inhibit lid in ratmyocytes (Escande et al., 1989). These class ifi antiarrhythmios are attractingmajor attention since they seem to possess greater efficacy than the class I drugsin those experiments which are most representative of the clinical situation.Also, the class ifi antiarrhythmic drugs do not possess the negative inotropiceffects of all of the class I, II and TV antiarrhythmic agents, as mediated byalterations of intracellular Ca2.New “pure” class ifi agents such as sematilide,dofetilide (UK-68798) and E-403 1 have been developed and shown to beeffective in suppressing programmed stimulation-induced ventricular tachycardia(VT) and preventing the onset of ventricular fibrillation (VF) in the animalmodels (Katritsis and Camm, 1993). The drugs are being tested clinically and amajor concern is the risk ofproarrhytbmia (i.e. torsade de pointes) associatedwith these class ifi drugs. In the development ofnewer drugs, the balancebetween the proarrhythmic risk and the antifibrillatory actions will be a majordeterminant in evaluation of drug utility.1.2.3 Antiarrhythmicproperties oftedisamil (KC8857)Tedisamil, in a dose-dependent manner, lengthened the effectiverefractory period in intact rat heart, which prevented both ischaemia andelectrically-induced ventricular fibrillation (Walker and Beatch, 1988; Beatch etal., 1991). The ability of the drug to block the two predominant cardiac outward13K currents, the Tic and the transient outward K current (Ito), was examinedusing the voltage clamp technique. (Dukes and Morad, 1989) Tn a dose-dependent manner, tedisamil was found to block the time-dependent Tic in guineapig ventricular myocytes and the transient outward K+ current Ito in ratventricular myocytes (Dukes et aL, 1990). Dukes and co-workers (1990) alsodemonstrated that tedisamil was effective in blocking the same two currents (Itoand Tic) in mouse astrocytes. Tedisamil also inhibited the Ito current in singlesmooth muscle cells of the guinea-pig portal vein (Pfrunder and Kreye, 1992).1.2.4 Putative class III antiarrhythmics aspotent K(Ca) blockersUsing the patch-clamp technique, tedisamil was found to block K(Ca) insmooth muscle cells of the guinea-pig portal vein (Pfrunder and Kreye, 1991).The drug also blocked a K(Ca) in cultured mouse motoneurons (MeLarnon etal., 1992). Single channel patch clamp experiments in this laboratory haveshown that a number of class ifi antiarrhythmic agents including tedisaniil, UK-68,798 and risotilide inhibit K(Ca) in hippocampal CAl neurons (McLamon andWang, 1991) Tn addition the tedisamil block of a K(Ca) channel in mousemotoneurons has been documented (McLarnon et al., 1992). A comparison ofthe tedisamil block of Tic and Ito in cardiac cells with that ofK(Ca) inhippocampal neurons would suggest a considerable degree of correlationbetween the actions of the putative class ifi antiarrhythmics on repolarizing Kcurrents in the cardiac cells and those on the CAl neurons.1.2.5 The RSD compoundsSome RSD compounds appear to alter properties of intact rat heart in amanner similar to that of putative class ifi antiarrhythmics. For example, dose-dependent lengthening of the effective refractory period (ERP) is observed. An14important point of the present work was to examine the possibility that theseRSD compounds block K(Ca) in a similar fashion to other putative class ifidrugs such as tedisamil. The availability of a large number of these compoundsalso was useful in the examination of the correlation between the effects onprolonging cardiac action potentials with the inhibition of the CAl neuronalK(Ca).152. METHOD2.1 Tissue culture preparationThe procedures carried out to obtain Wistar rat hippocampal cultureswere generally based on the method employed by Banker and Cowan (1977).Briefly, the hippocampi of day 18 fetal rats were dissected out and dissociatedinto single cells by enzymatic (trypsinization) and mechanical (repeated pipettingthrough Pasteur pipette) treatments. Once counted, the cell suspension wasdiluted down to approximately 105 cells/cm2in Dulbecco’s Modified Eagle’sMedium (DMEM), and plated onto 18 mm laminin-coated coverslips, which hadbeen treated with poly-D-lysine to inhibit the proliferation ofnon-neuronal cells.These coverslips were then inserted into 6-well plates and incubated, with thegrowth side downwards, in DMEM and 5% CO2 at 37°C until use. Byexperience, it was best to perform patch clamp on hippocampal cultures withintwo weeks after their isolation. Thus, experiments were carried out on thecultured neurons 3-14 days after cells had been placed on coverslips, since theK(Ca) are only found to be expressed in hippocampal cell cultures that were atleast three days old, and patching becomes difficult when cells begin todeteriorate after about two weeks of isolation.2.2 Eletrophysiology - patch-clamp recordings2.2.1 Pipette preparationThe patch clamp electrodes were made from Coming #7052 borosilicateglass (A-M Systems, Washington), which has the advantage of reducing energyloss when voltage is being applied. The glass had a diameter of either 1.2 mm(0.68 mm inner diameter) or 1.65 mm (inner diameter of 1.2 mm). The16electrodes were prepared with a standard two-pull technique using a NarishigePP-83 vertical glass microelectrode puller. For each size of glass, fixed pullinglength and fixed settings for the two heaters were generally maintained tominimize the differences between different electrodes. The resultant electrodeshad tips with diameters of about 1-2 jim, corresponding to a pipette resistanceof 4-8 M2. These pipettes were then fire-polished on a homemade microphorgeconsisting of a glass-coated U-shaped platinum ifiament connected to a D.C.voltage supply. Under a light microscope (x 200 magnification), the pipette wasbrought close to the heated platinum ifiament and polished to create a smoothtip. An airstream was directed at the filament during polishing, controlling thetemperature of the ifiament as well as confining the polishing to the very tip ofthe glass pipette. The pipettes were then filled with the appropriate pipettesolutions. For inside-out patches, the pipette solution contained (in mM) NaCl140, KC1 5, CaC12 0.2 and 4-(2-hydroxyethyl)-l-piperazineethane sulfonic acid(HEPES) 10, at pH 7.3; for outside-out patches, it contained (in mlvi) NaC1 5,KC1 140, CaC12 0.2 and I{EPES 10, also at pH 7.3. The pipette was usuallyfilled only half way up or just enough for the reference wire to be immersed.2.2.2 Pipette MountingThe pipette was inserted into the pipette holder, which was connected tothe headstage amplifier - a current to voltage converter with a 50 G feedbackresistance (Axon model CV-3 with 1/100 gain) for the patch clamp amplifer(Axopatch). These were mounted onto an Optikon lab jack, which contained aNewport motor drive micromanipulator for fine upward and downwardmovement of the pipette and also another manipulator for coarse three-dimensional movements. After mounting onto the pipette holder, a small positivepressure was applied to the pipette through a rubber tubing to prevent the17building up of debris around the pipette tip when it was immersed into the bathsolution. The pipette was then lowered into the bath solution and positionedabove the cell chosen for patching.2.2.3 PatchingA coverslip ofhippocampal neurons was removed from the 6-well plateand placed in the Perspex circular recording chamber. For inside-out patches thebath solution used both before and during the patch excision contained (in mM):NaCl 140, KC1 5, CaCI2 0.2 and HEPES 10, at pH 7.3 (same as the pipettesolution). Immediately after an excised patch had been obtained, the bathsolution was changed to one that contained (in mM): NaC1 5, KC1 140, CaC120.2 and HEPES 10, also at pH 7.3. For outside-out patches, the high K solution(140 mM) was used for the pipette solution and the low K solution (5 mlvi) wasused as the bath solution both before and after the excision. In one experiment,the [Ca2]in bath was varied and controlled by using EGTA for measuring the[Ca24] dependency ofK(Ca). A 0.1SM KC1 agar-filled plastic bridge, connectedto the ground, was submerged into the bath solution and completed the circuit(when the pipette tip was immersed into the solution). The chamber, affixed tothe stage of a phase contrast microscope (Nikon), allowed the neurons to beviewed at a x300 magnification. At such a magnification both the viable cellsand the pipette were visible allowing the entire experiment to be monitoredvisually. After being lowered down to just above the chosen cell, a test pulse of0.2mV was applied to indicate the change in seal resistance together with anaudio signal which served the same purpose. The pipette was then slowlylowered onto the soma of the chosen cell using the Newport MPH-imicromanipulator. Once the pipette was touching the cell as indicated visually,electrically and audibly, a negative pressure was applied to the pipette through a18mouth suction tube which was connected to the pipette holder. This procedureusually resulted in the formation of a giga-ohm (> 10 G2) seal. The presence ofsuch a seal was confirmed with the disappearance of the 0.2 mV test pulse. Theoutput electrical signal was amplified with the Axopatch lB patch clampamplifier (Axon Instruments, Inc.), visible on a Kikusui 5020A oscilloscope.Now the electrical activity of the patch could be observed as rectangular pulseson the oscilloscope screen.2.2.4 Patch excision: Inside-out and Outside outAfter the formation of the cell-attached seals (the gigaohm seals), whichallowed a 10-fold reduction in background noise, the stage was set formanipulations to isolate membrane patches which could lead to one of the twodiffererent cell-free recording configurations. If the pipette was quicklywithdrawn with the microdriver, a patch of the membrane could be isolated inthe inside-out mode (cystolic side of the cell membrane exposed to the bathsolution). The mode was especially useful for measuring the effects of agonistsacting on receptors found on the cellular face of the cell membrane.Alternatively, instead of excising the patch at the cell-attached mode, a whole-cell patch could be produced by the disruption of the patch membrane with extrasuction. If the pipette was then pulled away from the whole-cell patch anoutside-out patch (with extracellular side of the cell membrane facing the bathsolution), was obtained.2.2.5 Single channel recordingsAfter the excision to the inside-out mode, channel activities in the patchgenerally required triggering by drawing the patch briefly out of the solutionwith exposure to air. It is presumed that the action would remove any vesicle19that had been fonned at the tip of the pipette during the excision of the patchfrom the cell membrane. The loss of channel activities during recording maysometimes be reversed using the same method along with the application ofpositive and/or negative potential. Since it is very easy to lose the patch duringthe take-out time, the procedure must be carried out very carefully and precisely.The procedure is usually not useful in the case of outside-out patches as theaction of taking the patch out into the air will in most cases end up in the loss ofthe patch.The unitary currents ofK(Ca) were generally recorded with the patchclamp amplifier (Axopatch, Axon Instruments), digitized at sampling rates of 5kHz, with a low-pass filter set at 2kHz. The low-pass filter removed the highfrequency component of the background electrical noise of the signal allowingthe single channel current to be resolved. Since the reference ground was thebath, the extracellular side of the channel had zero potential. Thus, whenpositive K ions entered the pipette through the K(Ca) in an inside-out patch(140 K in bath, 5 K in pipette), a downward deflection from the baseline wasproduced. Conversely, the same K(Ca) currents were recorded as an upward.deflection from the baseline in an outside-out patch (5 K in bath, 140 K inbath). For the pharmacological experiments, all the events were obtained withthe patch held at 0 mV, where the driving force for the K currents wasgenerated by the concentration gradient ofK across the patch. Steps ofdepolarizations and hyperpolarizations were also applied to determine thevoltage dependence of the channel.The patch clamp data were recorded onto an IBM-compatible PC (486DX33) using the programme pCLAMP V.5.0 (Axon Instruments) for off-lineanalysis. For backup recording an Instrutech VR-10 digital data recorder wasconnected to a Panasonic Omrnvision Hi-Fi VHS, model PV-4760-K. When the20channel open times were very short (e.g. rapid channel ifickerings producedduring a channel inhibition by an K(Ca) blocker), the currents were digitized at ahigher sampling rate of 20 kHz, with the low-pass filter set at 5 kHz. When thefrequency of openings was low, an Al 2020 event detector could also be used tostart recording only when an event of the proper amplitude had occurred.2.2.6 Data AnalysisOff-line analysis of the data were performed with pCLAMP V.5 (AxonInstruments). Three major bits of information about the single channel data wereobtained: the amplitudes of the single channel events, the durations of theevents, and the length of time between two consecutive events. Afterdetermination of the baseline (closed channel level) and the threshold (estimatedamplitude of the events), the computer programme would detect the openingsusing a half-amplitude threshold criterion. Namely, the initiation and terminationof an event (opening and closing of a channel) were defined by the crossings ofthe half threshold level by the current trace. The half-amplitude thresholdmethod is a common procedure used in the collection of single channel events.The sampling rate of 5 k}{z allowed resolution of channel openings with opendurations at 400 jis or greater. The minimum resolvable closing duration was setat 200 j.ts.The amplitude, open time and interval duration between two consecutiveopens of the selected single channel events were saved on a separate record andwere used to generate mean values and histograms, using the subprogrammepSTAT V.5.0 (Axon Instruments). Diagrams of the distributions of theamplitude, open intervals and closed intervals could be produced by thissubprogranime. Plots of this analysis were output on a HP plotter.212.3 Current clamp experiment on the hippocampal sliceRat hippocampal slices (400 ji11) were prepared according to standardtechniques (Church and McLerman, 1989) and placed in a recording chamber at34.5°C at the interface between a humidified atmosphere (95% 02: 5% C02)and control artificial cerebrospinal fluid (ACSF) which contained (mM): NaC1125.5, KC1 3, NaHCO3 21, NaThPO4 1.5, MgSO4 1.5, CaCh 2, D-glucose 10(pH 7.4 after equilibration with 95% 02: 5% C02). Intracellular recordings weremeasured using current clamp. The current clamp pulses were 5 ms in durationand increased in 0.1 jiA steps from rest. The hippocampal action potentials wereelicited first in control. Tedisamil was added to the perfusing solution and 30minutes were allowed for the drug to act on the slice. With this procedure theconcentration of tedisaniil at the active site was not known and had to beestimated.2.4 Measurements of the ERP (effective refractory period)The ERP measurements belong to part 2 of the four standard screens usedto examine the RSD compounds, where screen I includes the measurement ofthe effects of the administration of cumulative iv (intravascular) doses of thedrugs on the heart rate, blood pressure and ECG of the intact animal. MaleSprague-Dawley rats weighing from 150 to 350g were used, and wereanaesthetised with sodium pentobarbital (60 mg/kg i.p.). The right carotid arterywas cannulated for the measurement of the blood pressure. Stimulatingelectrodes, placed 1-2 mm apart, were implanted in the left ventricle toaccomplish electrical stimulation. The ECG was recorded with a special chestlead configuration (Penz et aL, 1992) where two electrodes were used. Thesuperior and the lower electrodes were placed 0.5 cm from the midline of thetrachea, at the level of the right clavicle and that of the 9th and 10th ribs22respectively. ERP, as part of the ECG, was obtained while the electricalstimulation was applied. After obtaining the stable control values (2 identicalconsecutive readings or the average of 3 similar ones), the RSD drugs wereadministered as iv infusions through the cannulated right jugular vein over aperiod of 5 minutes. Starting at a volume of about 0.05m1!kg/5 mm. the infusionrate is doubled as the experiment proceeds but not to exceed lOmlJkg/5 mm. TheECG was recorded on a Grass polygraph (model 7D) at a chart speed of 100mm/sec. With screen I of the examination giving a general picture of thecardiovascular actions of the RSD compounds, the screen II experiments allowsan assessments of the actions of the drugs (if effective) on the myocardial ionic(Na and K). Drugs known to affect the sodium and/or the potassium channelshave been shown to produce clear profile of actions in these screen II tests. Withour interests in the potassium channel blocking ability of the RSD drugs, theERP value, which is a good representative of the repolarization activity of theanimal heart, was measured for a number ofRSD compounds.233. RESULTS3.1 Single channel properties of K(Ca)3.1.1 Channel conductanceUnitary currents were recorded primarily from inside-out patches of thecultured rat CAl hippocampal neurons. With a physiological K gradient of 140mM K internal and 5 mM K external across these patches, a channel withconductance of 80 - 110 pS could be obtained. Typical unitary activity is shownin Fig. 1, where (by convention) upward deflection indicates the outward flow ofpositive K ions. Starting from -40 mV the holding potential of the patch wasincreased with each depolarizing step of 2OmV, up to 60 mV. The variation inchannel activity is also show in Fig. 1, using representative portions of traces ateach indicated voltage. As the patch was depolarized from - 40 mV to 60 mV,the amplitude of the opens increased accordingly. The current-voltage relation ofthis K(Ca) is expressed by the I/V plot shown in Fig.2. The conductance of thechannel, given by the equation Tm = C.Vm + Vo (where Im is the patch current, Cthe conductance, Vm the patch potential, and Vo the zero-current potential), isdetermined by the straight portion of the slope of the curve (-20 mV to +60 mV).The conductance is found to be 1 lOpS, which is close to the value measured fora high conductance K(Ca) in cultured rat muscle (Barrett, et a!., 1982). The useof non-symmetrical K across the patch results in a significant non-linearity inthe curve (as predicted from the Goldman equation), and prevents an estimate ofthe zero current potential from the curve.24Fig. 1 Typical unitaiy currents ofK(Ca) measured in the cultured CA]hippocampal neurons.By convention, the channel openings are the upward rectangulardeflections from the baseline. Currents were recorded from an inside-out patchwith 140 mM K in the bath solution and 5 mM K in the pipette solution.Starting from the top trace recorded at a Vm of -40 mV, the traces shownrepresent depolarizing steps of 20 mV, to + 60 mV. The amplitude of thechannel activity increased with depolarization.25Fig. 1-20±40+60SpAlOOms26Fig. 2 Current (7) - voltage (19 plotI-V plot for K(Ca) openings recorded in an excised inside-out patch with140 mM K in the bath solution and 5mM K in the pipette solution. The curveis a visual fit to the data, and the linear portion gave a slope of 110 pS.Z28Most measurements of K(Ca) channels have used symmetrical 140 mM Kacross patches since the increase in extracellular [Kj can increase unitaryconductance. The results of changing the external solution to the high 140 mMK is shown in Fig. 3. With symmetrical K concentration across the patch thezero-current (reversal) potential was 0 mV. Thus the only driving force for thecurrent flow was the holding potential, which was changed (increments of 20mV) from -40 mV to +60 mV. The direction of the current flow changes as theholding potential varies from above the zero-current potential to below it: anegative holding potential causes an influx ofK and a positive one drives anefflux ofK. The amplitude of the current.also varies with the holding potentialwhich increases as the absolute value of the holding potential increases. Tracesshowing the variation of the channel activity with the holding potential (varyingfrom -40 mV to +60 mV with a depolarizing step of 20 mV) are shown in Fig.3.The current-voltage plot is given in Fig. 4. The slope conductance for this I-Vcurve was found to be approximatelyl7OpS.3.1.2 Calcium dependenceIn our experiments we have demonstrated the dependence of the K(Ca) oninternal calcium by varying the calcium concentration of the bath solution.Throughout our experiments with the inside-out patches, a concentration ofinternal calcium (0.2mM) was used to measure the K(Ca) channel activities. Itwas found that decreasing the calcium concentration from 0.2 mM to about 10j.iM had little or no effect on the patch activity and decreasing intracellularcalcium to 0.07MM totally abolished the openings. In order to investigate the29Fig. 3 Unitary currents measured in a patch with no [K+] gradientCurrents were recorded from an inside-out patch with 140 mM K in boththe bath and the pipette solutions. Starting from a Vm = -40 mV and movingdown in 20 mV depolarizing steps to + 60 mV, traces of channel activity areshown. For a Vm <0 mV, the channel openings are downward deflections fromthe baseline; for a Vm> 0 mV, the openings are upward deflections. Theamplitude of the channel activity increases as I VmI increases (i.e. away fromzero-current potential).30Fig. 3TvNo÷20+40+60SpAlOOms31Fig. 4 I - VplotI - V plot for K(Ca) openings recorded in an inside-out patch with bothbath and pipette solutions containing 140 mM Kt The curve is a linear fit to thedata, with a slope of 170 pS.32Fig. 4I (pA)I •I__20 40 60V(mV)33calcium-dependence of the channel activity, a number of bath solutionscontaining different concentrations of calcium were prepared. The proceduresused EGTA added to a stock solution of Ca2 and calculations using a computerprogram outlined in Fabiato and Fabiato (1979). By increasing theconcentrations of free calcium gradually from the minimum of 0.07 jiM to highervalues, an estimate for the threshold of internal calcium concentration requiredfor channel activity was made. This value was near 2.2 jM, which was theminimum concentration at which channel activity was evident. Upon the returnof internal Ca2 above 4 jiM the channel activity was maximal as judged by nofurther increase in the Popen. The variation of the channel activity with respect todifferent [Ca2]is represented by the traces ofunitary activity in Fig. 5. Similardependence of the CAl hippocampal K(Ca) on internal Ca2 has beendemonstrated in this laboratory previously and a sigmoidal relationship betweenthe channel open probability and [Ca21jhas been obtained (McLarnon andSawyer, 1993). The internal calcium-induced variation in the hippocampalK(Ca) open probability was primarily due to changes in mean open time of thechannel and a slower closed time component.3.1.3 Voltage dependenceIn experiments to examine the voltage dependency of Popen, the patchpotential was changed between 0 mV and +20 mV. These studies used acalcium concentration of 0.7 jiM Ca2to keep Popen low. The Popen (openprobability) was measured as the probability of the channel in the open state (inpercentage) and was obtained as a product of the mean open time and the34Fig. 5 Unitary currents showing Cc? dependence ofK(Ca)Currents recorded from an excised inside-out patch showing the effects ofchanging the internal Ca2 concentration on K(Ca). The Ca concentrationdecreases down the traces from a control of 0.20 mM to 3.80 jM [0.2 mMEGTA] (with normal channel activities) and to 0.07 j.iM [0.4 mM EGTA] (totalabolishment of openings). The fourth trace shows the increase of Ca2 to 0.15jiM [0.3 mM EGTA], with still no openings. An increase to 0.30 jiM [0.25 mMEGTAJ induces some channel activity but of a smaller amplitude, and anincrease back to 3.8 jiM results in the return of the normal channel activities.35Fig. 53.8OM Ca2+ JLV1flUP0.07jiM Ca20.15jiM Ca20.30jiM Ca23.80pM Ca2+ fJfL%jJJ(SpAlOOms36frequency of opens, divided by the overall time of measurement (lOs). Theresults for one patch gave Popen of 0.57 ± 0.10 % (at 0 mV) and Popen of 0.91 ±0.18 % (at +20 mV). These measurements were based on analysis of 10 separateapplications of the two potentials. The increase in Popen was then studied todetermine ifpotential altered the mean open time, the frequency of openings, orboth properties. It was found that mean open time was not significantly changed(2.0 ± 0.12 ms at 0 mV to 2.07 ± 0.09 ms at +20 mV). The increase in openfrequency was significant (from 27.6 ± 4.5 per 10 seconds at 0 mV to 44.4 ± 7.0per 10 seconds at +20 mV). The results showed that the mean open time was notaffected by the change in voltage. The increase in open probability from 0 to+20 mV, obtained at a low calcium concentration of 0.7 i.IM, was solely due toan increase in the frequency of channel openings. An illustration of thedifference in channel activities at 0 mV and +20 mV is shown in Fig. 6, with theincrease in the number of openings in the traces recorded at +20 mV. Acomparision of the changes in the number of opens, mean open time and openprobability, as affected by the voltage change, is given in Fig. 7.3.2 Macroscopic currents with tedisamil on hippocampal slicesThe large conductance K(Ca) is the unitary basis of the macroscopic Icdescribed in hippocampal neurons which is important in the later two thirds ofthe spike repolarization and the fast and the middle afterhyperpolarizationphases. It was therefore of interest to determine the possible effects ofantiarrhythmic drugs on this hippocampal Ic using current clamp technique onthe hippocampal slice preparation. Tedisaniil, a possible class ifi antiarrhythmic37Fig. 6 Unitary currents showing the voltage dependence ofK(Ca)Traces ofunitary currents obtained from excised inside-out patches arechosen to represent the actual situation of the channel activities. Changingbetween 0 mV to +20 mV consecutively, an increase in the number of channelopenings is observed at + 20 mV. Notice also that the channel openings arelarger at + 20 mV due to the enhanced driving force.38Fig. 6o mV+20 mV+20 mVSpAlOOms39Fig. 7 Voltage dependence ofK(Ca)A comparision of the A) frequency of channel openings, B) mean opentimes, and C) Open probability between channel activities recorded at 0 mV and+20 mV from an inside-out patch. Significant differences exist betweenactivities recorded at the two Vm in terms of the frequency of openings andopen probability (P < 0.05; one-tailed t-test), while no such differences betweenthe mean open time values. The increase in Vm increases the open probabilityby increasing the frequency of openings but not the mean open times.40Fig. 7ABC* P<O.05Mean # of events*605550454035N30252015105020 mVMean open times2.521.5ms10.50ZeromV 20 mVMean probabilities1.2 *1.110.9— 0.80.6o mV41agent and known to block K(Ca) in the hippocampal neurons (McLamon andWang, 1991), was used in this study. Tedisamil (at 5 pM) lengthened actionpotential duration as recorded from the hippocampal slice by prolonging the laterphase of the action potential repolarization (Fig. 8). In addition theafterhyperpolarization (AHP) phase was reduced. This result suggests thattedisamil blocked the K(Ca)-mediated Ic in the hippocampal neurons, causing aprolongation in the repolanzation phase of the action potential in thehippocampal slice. These results would indicate that putative class ifiantiarrhythmic agents could block repolarizing K channels in excitablemembrane. Thus the measurements of drug actions, at the single channel levelon neuronal K currents, may be useful in the characterization of mechanisms ofactions of antiarrhythmic agents. At present, similar studies on cardiac Kchannels are severly limited by the low unitary conductance of repolarizing Kchannels in cardiac cells such as Ito (transient outward K current) and Tic(delayed rectifier K current). A critical point however was to determine ifpotency for drug block of K(Ca) in hippocampal neurons was correlated withdrug potency to alter properties in whole heart. This point served as a focus forthe subsequent experiments3.3 Pharmacology of K(Ca)3.3.1 The RSD compounds - potential class III agentsThe class ifi antiarrhythmic compounds (the K blockers such as UK68798, tedisamil and risotilide) have been demonstrated to be effective inblocking the K(Ca) in a dose-dependent manner (McLarnon and Wang, 1991).42Fig. 8 Effects of 5 pM tedisamil on the action potentials elicited in thehippocampal sliceThe traces show the intracellular recordings of a hippocampal slice usingcurrent clamp. Current clamp pulses were of 5 ms duration and increased in thesteps of 0.1 pA from rest. An action potential first elicited in the slice is shownas the control (bottom trace). The top trace shows the addition of tedisamil (5pM) into perfusing solution after 30 minutes of waiting time. The prolongationof the late 2/3 of the repolarizing phase and an abolishment of the earlyafterhyperpolarization are evident.43Fig. 8(Control)1lOmVims44The action of some of the newly developed RSD compounds in diminishing theK(Ca) is the subject of investigation in this section. These drugs act on K andNa channels in cardiac cells with the potency for K and Na inhibition differentand variable among the RSD compounds. A total of 18 RSD compounds wereused and they are listed, in an ascending numerical order, as follows: 921, 935,939, 942, 949, 952, 956, 959, 968, 969, 971, 973, 974, 979, 983, 984, 986 and987.3.3.2 Inside-out patch clamp experimentsThe RSD compounds were tested for their actions in inhibiting K(Ca).After measurements of channel properties in control solutions the compoundswere applied to the bath solution at an initial concentration of 1 jiM. If channelunitary properties such as open times or opening frequency were altered bycompounds applied at 1 pM then a concentration range between 1 - 10 pM wasgenerally studied. If a compound showed little or no effect on properties ofK(Ca) at a concentraton of 1 jiM, the concentration was then increased to highervalues (i.e. 10, 20, 50 pIvi), to a maximum value of 100 jiM. The blockage ofK(Ca) was evident from the increased flickering (channel transitions) of theunitary activity from open to closed (blocked) states. On the basis of singlechannel records it was not possible to distinguish between closed or blockedstates. However, bursts of channel closures, not present in control records, weretaken as evidence for drug block. In most cases wash-off of drug was possibleyielding values for channel kinetics close to that found prior to drug application.The RSD compound RSD 971was the most potent blocker of K(Ca); atypical action of this agent is shown in Fig. 9. The top trace indicates the channel45activity in control. With addition of the compound, the flickeriness of thechannel increased with a decrease in the time of the openings (i.e. shorteropenings) as the concentration of the drug increased from 1 iM to 4pM. Thebottom trace showed the recovery of the long openings as the compound waswashed away by re-perfusing the bath with the control solution. Themeasurement of these channel activities can yield some useful analysis of thedata: the unitary amplitude of the channel, the closed and open duration and thefrequency of channel events. The open time analysisTypical open time distributions are shown for another potent K(Ca)blocker (RSD 939) in Fig. 10 A. The distributions were fit with singleexponential curves as indicated on the figures. After comparing the goodness offit, it was found that the single component (compared with fits using multi-components) gave the best fit; this point was determined by comparing errors inthe fits using different numbers of components. The arithmetic mean open timewas also useful in estimating this corrected to. The two measurements (fit vsmean value) were identical except for a factor of the system resolution time.(Colquhoun and Sigworth, 1983). For most of the to measurements in ourexperiments, the corrected to was smaller than the arithmetic to by a consistenttime (in the range of 0.2 - 0.6 ms). The single exponential fit indicated that thereis only one channel opening state of the hippocampal neuronal K(Ca) which wasalso confirmed by the amplitude distributions (see 9 Effects ofa RSD compound (971) on K(Ca)Unitary currents obtained from an inside-out patch at 0 mV are shown tobe affected by the administration ofRSD 971(concentrations as shown). Thecompound was the most potent of the 18 compounds tested in the investigation.A dose-dependent decrease of the mean open times (as well as the increase offlickering) of the channel activities is evident. A return to the control solutionresulted in kinetic behavior close to that found prior to compound application.47Fig. 9ControlljtM 9714 jiM 971Recovery____5pAlOOms48Fig. 10 Effects of4u1vI 939 on the open times, closed times and amplitudedistributionsHistograms of the open times (A), closed times (B) and amplitudes (C)distributions were constructed to show the effects of 4iM RSD 939. The opentime distributions were fitted by a single exponential component and the closedtime distributions by two components. For the control solution: the mean opentime was 14.07±0.98 ms (A), the closed time had components 0.86±0.06 ms and4. 13±1.25 ms (B). For 4jjM 939: the mean open time was 6.84±0.77 ms, withthe closed times components 0.77±0.03 ms and 4.79±0.83 ms. The results,which indicate reduction in the mean open time with the compound and nosignificant changes in both the amplitude and closed time distributions, suggeststhat 939 blocks K(Ca) in an uncompetitive manner.Fig. 10ABC0z0z.00zControlz04pM 93949—14.07—’— 4 sOpo dorzti*is (mi) Opeo dur.doas (ms)-0.86 ms0.7cIosd durtIos (es)— 4.79 sClosed dwatioas (me)—o (pA)jeAmplitude (pA)50In Fig. 10 A, compound RSD 939 (4 jiM) was shown to decrease the t(roughly halving it), reflecting a significant reduction in the mean open times ofthe channel. Most other RSD compounds showed a similar effect in decreasingthe mean open times but with varying potency. An investigation and comparisonof this effect of the compounds is given in the later section (3.3). The closed time analysisTypical closed time distributions are shown in Fig. 10 B. The closed timedistributions could be best fit with a two-component exponential function. Thetwo components were related to the single channel records discussed previously:on this basis the fast component (tj) represents the closures within a singleburst, largely due to the block of the channel by the compound. The slowcomponent (t), on the other hand, represents closures between individualbursts. The presence of 939 had little or no effects on both tef and tC (Fig. 10B). Other RSD compounds have been found to act in a similar manner and hadno significant effects at any concentrations on the closed time distributions(either component) of the K(Ca). The amplitude analysisTypical amplitude histograms are shown in Fig. 10 C (both recorded atzero mV). The single peak distribution indicates that i) there is only one type ofchannel activity in the patch, and ii) no channel subconductance state waspresent. The amplitude showed little variation between patches (within 0.5 pA)and amplitudes of unitary currents were not altered with any of the RSDcompounds at any concentration studied. This result indicated that the51compounds did not exhibit very fast channel block since in such cases unitarycurrent amplitudes can be reduced by the compound. In this case the full currentamplitudes cannot be resolved due to limitations in amplifier band width. The open channel blockThe decrease in the mean open time, with very little change in theamplitude and closed time duration distributions, are typical characteristics of anopen channel block. An open channel block model has been useful in describingthe actions of putative class ifi antiarrhythmics in blocking K(Ca) in CAlhippocampal neurons including KC 8851 (McLarnon, 1990) and UK 68798,tedisamil and risotilide (McLarnon and Wang, 1991). Tn addition tedisamilblocked K(Ca) in motoneurons (MeLarnon et al., 1992). Fig. 10 represents thegeneral situation of the actions of the RSD compounds tested on the three majoraspects of the channel kinetics (open and closed times and amplitude). The openchannel block model can be represented by the scheme shown below:ki k2== Bi±ik-i k-2where ki is the onward rate constant from the closed state to the openstate and k-i represent the transition from open to closed state. k2 is the onwardrate constant for the association of the drug, and k-2 is the off (unblocking) rate52constant. The dissociation rate constant kD is defined as k-2/k2. The drug-induceddose-dependent decrease in the mean open times of K(Ca) can be expressed as alinear function of the concentration of the blocker, in the form of the rate ofdecay (i.e, the reciprocal of the mean open time, t-’). The equation is given by:t-1 =k2[D]+k-1A plot of t-’ vs [D] can thus be used to determine the blocking rateconstant k2. If such a plot is not fit by a linear relation, then the data wouldsuggest inapplicability of an open channel block model to describe the actions ofthe RSD compounds. Otherwise, the effects on the mean open times can be usedas an index of the potency of the RSD compounds.3.3.3 A comparison of the class III antiarrhythmics as potent K(Ca)blockersAs noted above the RSD compounds examined, like other putative classifi antiarrhythmics, block K(Ca) with evident increase in channel transitionsfrom the open to a non-conducting state. Thus, measurements of mean open timeprovides an index for drug block of K(Ca). In particular, I have used theconcentration which caused mean open time in control to be halved as ameasurement of the potency of the RSD compounds in blocking K(Ca). Thisvalue is denoted as MOT50. The quantitative assessment of channel block hasutilized the specific open channel block scheme already described. According tothis model, a plot of the inverse of mean open time in presence of the compound53against the compound concentration should be linear. The slope of the graph isthe onward blocking rate constant. A plot of r- against [D]The k2 value (blocking rate constant) gives an estimate of the blockingpotency of the drugs. The k2 can be determined by plotting ‘r-1 (calculated as theinverse of the mean open time) against the concentration of the agonist ([D]).The slope of the fitted straight line gives k2, while the fitted line can beextrapolated to intercept the y-axis at the point yielding k-i . Five drugs werechosen for the analysis of the linearity of the relationship between mean opentime and compound concentration and for calculating k2. As noted above thecheck of linearity is essentially a check for applicability of the blocking model.The results are summarized in Table 1 and the plots are shown in Fig. 11. Thelinear relationship between the compound concentration and the inverse of meanopen times justifies the use of the open block scheme to describe the blockingaction of the RSD compounds. It is also the basis to determine the potency ofthe compounds in terms of concentration required to exert a given change in themean open time (in this case, the mean open time to be one-half of the controlvalue). The k-i values, which are independent of the presence of drugs, werefound to be similar for the five RSD compounds used. This result suggests thatthe kinetics of channel closing were similar in control from patch to patch.An analysis for the value of k-2 (the concentration-independent off-rateconstant from the blocked state to the open state) was done as follows. Recordsof unitary events were scanned and regions of rapid closures were analysed fortime durations of closed state transitions. This analysis assumed such closures54Table 1. Onward blocking rate constants (k2)forfive drugsThe r1, calculated as the inverse of the mean open time in thepresence of drugs, is related to k2 by the equation c-1 = k2 [Dl + k-i.The k2 is defined as the slope of the graph of t-1 against [D](concentration ofRSD compound), and the five slopes are shownin Fig. 11.RSD n ID] (pM) t (sec 1) k2 (1O M1 sect)971 4 0 65.89±8.48 7.601 154.6±12.174 369.9 ± 19.10986 4 0 79.50±8.38 7.141 143.8±26.684 365.2±72.20984 3 0 61.83±4.19 3.601 113.8±7.704 205.9±7.96939 4 0 98.40 ± 12.54 2.351 132.0 ± 15.884 192.6± 18.1010 363.5 ± 12.55983 3 0 67.60 ± 2.70 0.801 84.40 ± 4.954 96.80±9.0410 133.9±3.1855Fig. 11 The plots of -i against [D]The mean open time and concentration relationship of five RSDcompounds are shown. The RSD compounds acted by causing a dose-dependentdecrease of the mean open time, which can be expressed by the open channelblock scheme v-i = k2 [D] + k-i. The plots oft-i vs [D] are thus linear. For dataplease refer to Table 1.“0Fig.11f’(sec‘)[7O9•4A998j400350A300250200A150Ax100x[DJ(jIM)57were channel blocking sojoums. This procedure yielded only an estimate for k-2(inverse of the times) since it was difficult to differentiate between blocked andclosed state transitions. For RSD 971 the off-rate constant was estimated to beapproximately 1/1000 sec-1,which would give a kD value near 0.1 jiM. Index ofthe potency ofthe compounds -MOT50The MOT50 values of 15 RSD compounds are listed in Table 2. The nvalues indicate number of experiments with given drugs and was at least n=3 forall agents with the exception of RSD 952 (with n=2). A MOT50 value wasobtained for 15 out of 18 compounds with the remaining 3 agents having MOT50values higher than the maximum concentration used in these experiments (set at0.1mM). A system of subdividing the RSD compounds according to theirpotency for blocking K(Ca) allowed grouping them into 4 concentration ranges.These ranges were based on concentrations of the compounds to cause meanopen time to be half of control values. These were as follows:1) Highly potent (0.5 jiM- 1.5 pM) - 971, 959, 986, 979, 9212) Potent (1.6pM - 2.5 pM) - 949, 984, 987, 9693) Intermediate (2.6 pM - 5 pM) - 968, 974, 9394)Low potency (6 pM -20 pM) - 952, 983, 9735)Noteffective - 935,942,95658Table 2. MOT5O values of(lie RSD compounds.The MOT50 values are defined as the concentration (in M) of thecompound required to reduce the mean open time to 50% of its original value.The values are catorgorized into 4 group according to the potency of thecompounds (1 - highly potent; 2 - potent; 3 - intermediate; 4 - low potency). 15compounds are listed below, with the exception ofRSD compounds 935, 942and 956, which were ineffective at a concentration as high as 100 1.iM.1 23 4RSD n MOTso(971 4 0.78±0.12959 4 1.03 ±0.16986 4 1.25±0.17979 6 1.30±0.16921 5 1.40±0.17RSD n MOT5O(968 3 3.60±0.31974 5 4.20 ± 0.46939 4 4.38 ± 0.24RSD n MOTso(949 5 1.80±0.12984 4 1.88±0.24987 4 2.00 ± 0.35969 5 2.34±0.14RSD n MOT50(952 2 6.50±0.50983 4 9.00±0.41973 3 15.30±1.4559The effects of the RSD drugs on open time durations (n= 1) are shown inFig. 12 - Fig. 19 as mean open time distributions fitted with a single exponentialcurve along with the value given on each graph. Compounds chosen fromeach of the first four groups of potency are shown: from Fig. 12 to Fig. 14, drugs971, 986 and 979; from Fig. 15 to Fig. 16, drugs 984 and 987; from Fig. 17 toFig. 18, drugs 939 and 983; in Fig. 19, drug 973. Open time distributions of thecontrol and two concentrations of the compound are shown in each figure withconcentrations being 1 pM and 4 pM for most of the compounds. Higherconcentrations of the less potent drugs are shown (4 pM and 10 pM of 983, 10pM and 20 jiM of 973). Comparison ofMOT5o with putative class III antiarrhythmicsThe MOT5O values determined for RSD compounds can be compared withother class ifi antiarrhythmics (McLarnon and Wang, 1991). It was found thatboth the sulfonamide compound UK 68798 (MOT50 = O.4p.M) and thedthydrochloride derivative tedisamil (MOT50 = 1pM) belong to group 1, withUK 68798 even more potent than RSD 971(MOT50 = 0.78pM). Risotilide (7.5pM) was low in potency compared to these RSD compounds. RP 62719, abenzopyran compound (personal communication from this laboratory) with anMOT50 value of 3.5 pM, was intermediate in its potency. Comparison ofthe MOT50 and the ERP25 valuesA comparison was made between the inhibiting actions of the RSDcompounds on the neuronal K(Ca) and their actions on repolarizing K currentsin whole heart. The potency of the RSD compounds in inhibiting the repolarizingcurrents in the whole heart was determined by measuring the concentration ofthe compounds required to increase the effective refractory period (ERP) by60Fig. 12 - Fig. 14 Effects of971, 986, 979 on the mean open timeThree RSD drugs of group 1 (very potent) are effective in decreasing themean open time, shown here (n=1) as histograms of open time distributions,fitted with a single-exponential curve.Fig. 12 -A) Control, with mean open time = 12.97±0.45 msB) 1 p.M 971, with mean open time = 6.58±0.46 msC) 4pM 971, with mean open time = 2.76±0.19 msFig. 13 -A) Control, with mean open time = 13.36±0.7 1 msB) lj.iM 986, with mean open time = 8.78±0.62 msC) 4pM 986, with mean open time = 5.56±0.34 msFig. 14-A) Control, with mean open time = 12.06±0.99 msB) 1pM 979, with mean open time = 6.09±0.31 msC) 4pM 979, with mean open time = 3.77±0.20 ms614èVCIa.EzCC.a,CCzFig. 12CCC)ECz12.97 ms0Open durations (ma)AControlB1jiM97lC4 pM 971100= 6.58 ma10Open durations (ma)40120602.76 insOpen durations (ins)0 0. C 0OCNumberofeventsNumber of eventsO 0 0NumberofeventsC0 0 0. C •1 C e I0‘C C 0 0.C 0I C’a’,63Fig. 14 AControlB1 jM979C4jiM979C0aEzz110C0‘ssa0ECz90CC)CI-0C0sot= 12.06 ms40Open durations (ms)t,=6.O9ms24Open durations (ins)45t3.77ms5025Open durations (nis)64Fig. 15- Fig. 16 Effects of984 and 987 (group 2: potent) on the mean opentime (n=1)Fig. 15 -A) Control, with mean open time = 12.87±0.76 msB) lp.M 984, with mean open time = 7.63±0.27 msC) 41iM 984, with mean open time = 4.45±0.2 1 msFig. 16 -A) Control, with mean open time = 11.69±0.46 msB) lj.tM 987, with mean open time = 7.84±0.34 msC) 4pM 987, with mean open time = 4.49±0.26 ms70 65CC,CC,8CzCC,0‘70IC)8CzC)C)1CC,.C 358Cz= 12.87 ms50Open durations (Ens)110AFig. 15ControlB1pM984C4jiM984100= 7.63 ms0 30Open durations (ma)60= 4.45 EnSOpen dur*tions (ma)66Fig. 16 AControl9045EzB1 jiM987C4 M987C55ECzCC.C.0IC.EC;=IL69nis110Open durations (ms)to = 7.84 nisOpen durations (ins)r0=4.49 nis20Open durations (ins)67Fig. 17- Fig. 18 Effects of939 and 983 (group 3: intermediate) on the meanopen time (n=])Fig. 17 -A) Control, with mean open time = 14.69±0.91 msB) ijiM 939, with mean open time = 12.43±0.60 msC) 4pM 939, with mean open time 7.13±0.14 msFig. 18 -A) Control, with mean open time = 15.09±1.05 msB) 4pM 983, with mean open time = 10.58±0.43 msC) 10pM 983, with mean open time = 7.99±0.42 ms7068Fig. 17AControl100B1jiM939C4jiM939 C) 1051C)E1000C)z= 14.69 ins90Open durations (ins)50= 12.43 ins0Open durations (ins)t0=7.13 ins30Open durations (ins)6069Fig.18 A0Control o40Sz 15.09 maz0 50 —B90Open durations (ma)1000C)4-41iM983 4sC).0S0z ;1O.58ms—10 so110Open durations (ma)100C.150C)4-10 1iM 983 eSSIC).0S0 t. 7.99 ma0 40Open durations (ma)70Fig. 19 Effects of973 (group 4: low potency) on the mean open time(n=1)A) Control, with mean open time = 11.69±1.55 msB) 10pM 973, withmean open time = 6.05±0.38 msC) 2OjiM 973, with mean open time = 4.47±0.14 ms71‘0AFig. 19C.45Controlt.11.69msCzso0 40Open duxations (ins)140BC.7QlOjiM 973 E ;=6.OSmsCz3’0 15Open durations (ins)140C.11CU7o2OpMCz0 1*Open durations (ms7225% per kg weight of the animal per minute. These concentrations, representedby ERP25 were obtained from experiments performed in Dr. M.J.A. Wailcer’slaboratory. Since the actions of the compounds reach a steady state in the wholerat experiment, we can compare the ERP25 values with the MOTso values. Atable summarizing the two sets of values are given in Table 3, and the graphshowing the correlation between these data are given in Fig. 20. Fig. 20 includesall of the 18 compounds tested in both laboratories. The data suggests theexistence of a general correlation between the actions of the compounds, exceptfor the two outliers RSD 979 and 971. Using simple linear correlation theory itwas found that with the outliers, the MOTso compares poorly with the ERP25and gave a correlation coefficient (r) of 0.04. By excluding the two outliers, thecorrelation was greatly improved with r = 0.71 (thus yielding correlation index,r2, of 0.504). See Fig. 21. The correlation was statistically significant withp<ü. in committing a type I error. All three RSD compounds (935, 942, 956)which were found to be ineffective in inhibiting the K(Ca) (with a MOT50 > 0.1mM) were also not potent in inhibiting the repolarizing K+ currents in theisolated heart (with ERP2S 20 pM), and are shown on the right hand corner ofFig. 20. The general conclusion was that the effects of the drugs to block therepolarizing K(Ca) in neurons showed a positive correlation with drug effects toprolong ERP in whole heart.3.3.4. The wash-off/recovery timesOne interesting area of the study was to determine the wash-off7recoverytimes for different compounds. Although the inside-out patch clamp set-up maybe a poor representation of the actual physiological situation, the ability of the73Table 3. A comparison ofthe MOT5O and ERP25 valuesThe MOT5O values (in M) and the ERP25 values (in j.iMlkg/min.)of 15 RSD compounds are listed below. The MOT50 values are obtainedas described in Table 1. The ERP25 values were calculated from theECG traces obtained in the screen II test of the RSD drugs on the whole ratheart.RSD MOT5O (i.LM) ERP25(pM/kg/mm.)971 0.78 8.40959 1.03 1.30986 1.25 0.30979 1.30 18.0921 1.40 0.50949 1.80 1.30984 1.88 1.20987 2.00 0.90969 2.34 4.40968 3.60 3.50974 4.20 4.40939 4.38 2.30952 6.50 3.60983 9.00 1.50973 15.3 6.8074Fig. 20 A comparison ofthe MOT50 and ERP25The graph ofMOT50 vs ERP2S is shown for the 18 RSD compoundstested. Two outliers, RSD 971 and 979, could be seen at the bottom of thegraph. The estimated positions of the three RSD compounds (935, 942, 956)indicated on the righthand corner of the plot were found to be ineffective in bothinhibiting the K(Ca) (with MOT50 > 0.1 mlvi) and the repolarizing K currents inthe isolated heart (with ERP25 20 iMJkg/min.).InFig.20MOT50(jiM)1201:30252015I10-.I.I.•I.0-IIII05101520E11P25(ji.MIkg/min)4076Fig. 21 The correlation ofMOT5o and ERP25The simple linear correlation plot of MOT50 vs ERP2S is shown. The threenon-responsive RSD compounds (935, 942, 956) as well as the two outliers(971, 979) were excluded. With all of the 15 compounds correlated, acorrelation coefficient (r) of 0.04 was obtained. By removing the two outliers, rbecame statistically significant (P<O.05) with a value of 0.71.Fig.21MOT5O(LM)2016+12-+8+++4-++-H+III012345678ERP25(jiM/kg/mm)78compounds to remain at sites associated with the channel may still gives clues tothe actual behavior of the drugs in vivo. Apart from the ability of the compoundto stay on the membrane (the magnitude of its off-rate constant), the position ofthe patch relative to the fluid inlet can also significantly affect the time taken bythe reperfusing control solution to remove the compounds. If the patch lies closeto the path of the flow of the solution wash-off may be more rapid; otherwise thesolution will have to first fill up the bath before the compounds can be washedoff. Eight drugs were observed for their washing-off phenomena including 959,968, 969, 971, 983, 984, 986 and 987. The index used for the determination ofthe wash-off time was as follows. With the average flow rate of 1 ml in every 17seconds, a 50% recovery of the control mean open time within 20 seconds wasconsidered to be fast. A 50% recovery within 30 - 90 seconds was considered asmedium, and a 50 % recovery that came after 2 minutes was considered to beslow. The results showed that 983, 984 and 987 can be washed off easily (fastrecovery). 968, 969, 971 and 986 were considered to give intennediaterecovery, and 959 was found to be persistent against being washed off (slowrecovery). The wash-off/recovery time of these compounds were compared totheir wash-off times found in the isolated heart experiments in Dr. M.J.A.Walker’s laboratory. A recovery of the ERP was considered to be fast when itoccurred within 30 seconds. Between 1 and 5 minutes it was rendered slow andextremely slow for a period of more than 5 minutes. The results showed that983, 984, 987, 968 and 969 all gave fast recovery, and 971 and 986 were slowin recovery and 959 being extremely slow. A comparison of the recoveryphenomena of these compounds in the two set-ups indicate a certain degree ofcorrelation in the compound affinity. Further experiments however, will be79required for a more detailed comparison of the recovery times. The comparisonof these times are shown in Table Outside-out patchesThe actions of five RSD compounds were studied by adding thecompounds to the outside of the membrane in the outside-out patches. Longopenings were observed in the control solution. Upon the application of thecompounds rapid flickering transitions to the non-conducting state andcharacteristic of open channel block were observed. The potency of thecompounds, expressed in MOT50 , are given in Table 5. It was found that two ofthe compounds, RSD 974 and 979, has a MOTso value similar to that found withinside-out patches. The other three compounds, 949, 959 and 969, all showed areduction in potency in the outside-out patches. At present it is impossible todifferentiate between an internal and external site of action for these compounds.However, the similarity of the actions of the compounds in both the inside-outand outside-out patches suggest that the same site may be involved in both typesof patches.80Table 4. A comparison ofthe recovery timesIn the inside-out patch experiments, the recovery times wereclassified as fast (within 20 s), intermediate (between 30 - 90 s) andslow (more than 2 minutes). In the isolated heart experiments, therecovery times were categorized as fast (within 30 s), slow (between1 and 5 minutes), and extremely slow (more than 5 minutes).Recovery mode Inside-out patch Isolated heart (lid-)(.K(Ca))A) Fast 983 983984 984987 987968969B) Intermediatefor patch 968clamp experiments and 969slowfor the isolated heart 971 971experiments 986 986C) Slowforpatch clamp 959 959experiments and extremelyslowfor the isolated heartexperiments81Table 5. MOT5O values offive RSD compounds in outside-outpatchesThe MOT5O values are defined as the concentration (in jiM)of the compound required to reduce the mean open time to 50% ofits original value. The MOT50 values found in the inside-out patchesfor the same compounds are given also for comparison.MOT50 valuesfor MOTSO valuesforPSD compounds n outside-outpatch inside-outpatch(iM) (p949 2 5.85 1.80959 2 5.10 1.03969 3 6.23± 1.50 2.34 ±0.14974 2 4.50 4.20979 3 1.10±0.20 1.30±0.16824. DISCUSSIONThe essential question forming the basis for the present studies concernedthe applicability of a repolarizing K channel in neurons (K(Ca)) to serve as atarget for putative class ifi antiarrhythmic compounds. If this point could beshown then K(Ca), which is readily isolated and active in both inside-out andoutside-out patches, could serve a role as a plausible model for drug interactionswith K currents in cardiac cells. Previous studies have indeed shown that anumber ofputative class ifi antiarrhythniic compounds, with a diversity onstructures, blocked K(Ca) at concentrations similar to values which alteredproperties in whole heart preparations. In order to test the hypothesis that K(Ca)could serve as a surrogate for repolarizing K in ventricular cells, anexperimental program was carried out in a systematic fashion. First, a knownagent with class ifi actions, tedisamil, was applied to hippocampal slices. Theexperiment was designed to measure drug effects on Ic which is the macroscopicrepolarizing current in hippocampal neurons and is built on K(Ca). Secondly,unitary properties of K(Ca) were established by recording from inside-out andoutside-out patches. Finally, the critical experiments were carried out to measureactions of 18 RSD compounds on K(Ca) and results correlated with drug actionsin whole heart.4.1 Single channel properties of CAl hippocampal K(Ca)The single channel properties of the K(Ca) were examined. Underphysiological conditions of K+ (140 mM intracellular and 5 mM extracellular),the channel conductance was found to be 110 pS. By putting symmetrical K(140 mM) across the patches, a higher value of conductance of 170 pS was83obtained. The dependence of the CAl K(Ca) on internal Ca2 can be expressedby a sigmoidal dose-response curve with channel open probability againstinternal Ca2 concentration (McLarnon and Sawyer, 1993). An internal Ca2concentration of 4 j.tM was required for maximal channel activity as judged byno further increase in the Popen with increased Ca2. At a low internal Ca2concentration of 0.7jiM, the Popen was increased from 0.57 % to 0.91 % by adepolarization step of 20 mV. The increase in Popen was found to be the soleresult of a voltage-sensitive increase of the opening frequency and the meanopen times of the unitary current remained unaffected by the depolarization.Current clamp experiments were carried out to examine the effect oftedisamil on the elicited action potentials in the hippocampal slices. Tedisamil (5iiM) prolonged the action potential by prolonging the repolarization phase andabolishing the afterhyperpolarization phase (Fig. 8). The result has demonstratedthat class ifi agents like tedisamil could be useful in blocking repolarizing Kcurrents in excitable membranes and the measurements of drug actions at thesingle channel level may be useful in the characterization of the actions ofantiarrhythmic agents on repolarizing K currents.4.2 Correlation of the drug effects on K(Ca) and repolarizing+K currentsPrevious findings have appeared to suggest the similarity between the inhibitoryeffects on the repolarizing K currents in cardiac cells with drug actions on therat K(Ca). In order to test this possibility, a number of potential class ifi agentswere tested separately for their actions on the neuronal K(Ca) and comparedwith actions on whole heart. A total of 18 RSD compounds were used in these84experiments. All of the compounds were examined for their inhibitory actions onK(Ca) in the cultured rat CAl neurons using inside-out patches; in a few casesthe effects on outside-out patches were also examined. In the presence of mostof the compounds the unitary activity showed increased transitions from theopening state to a non-conducting state. In the presence of the compounds themean open time of the channel decreased with the closed time duration and theamplitude of the current remaining unchanged. This result was consistent in allof the compounds tested which showed significant inhibitory effects on theK(Ca); 3 of the compounds, RSD 935, 942 and 956 showed no significantinhibitory effects on the K(Ca) at concentrations as high as 0.1 mM. The datasuggested open channel blockade as the mechanism of action for these RSDcompounds on the CAl K(Ca). The open channel blockade scheme relates themean open time of the channel to the concentration of the drug by the followingequation: r1= k2 [D] + k-i. The plot of t-’ against [DI thus should give a straightline with a slope equal to the onward (blocking) rate constant k2. The ‘c-i vs [Djplots for 5 RSD compounds are shown in Fig. 11. The linearity of the datasuggests that these RSD compounds inhibit K(Ca) consistent with an openchannel blockade model. The results were consistent with the identification ofthe mechanism of action on the CAl K(Ca) for a number of other class ifiagents as open channel block (KC 8851 - McLamon, 1990; UK 68798,tedisamil and risotilide - McLarnon and Wang, 1991).The index of the potency of the RSD compounds was chosen as theconcentration of the compound which caused mean open time in control to behalved. This value is denoted as MOT50. By using this index, a group of 18compounds can be subdivided into 5 categories according to their potency ininhibiting the K(Ca) (see RESULTS). These data were then used to comparewith the potencies of these compounds in inhibiting the repolarizing iC currents85in rat whole heart. The index of potency for these compounds in blocking theoutward K currents was the concentration of the compound per kg weight ofanimal per minute required to increase the effective refractory period (ERP) by25% and represented by ERP25. The correlation of the MOT50 and the ERP25 isgiven in Fig. 20. Initial inspection of the available data seems to suggest nocorrelation between the responses of neuronal K(Ca) and of cardiac repolarizingiC channels to these Class III agents. Compounds highly effective in prolongingthe ERP in whole rat (e.g. 971 and 979) are not found to be comparativelyeffective in the neuronal situation, which seems to confirm that no correlationexists. The present data only allow a correlation within the range of roughly 0 -16 tM in MOT50 and around 0 - 8 iM!kg!min in ERP25, far too narrow for agood correlation system. On the other hand, however, the plot does show ageneral correlation between the two sets of values without the two outliers (RSD979 and 971). By excluding these outliers the simple correlation yields acorrelation coefficient of 0.71 which is statistically significant (P<0.05) althoughonly a coefficient of 0.04 is obtained by including them. By using theChauvenet’s criterion, we can legitimately reject any data points of ERP25 whichexceed 3.89 + 7.07 = 10.98 tM/kg/min. in value (with a Chauvenet’s Criterionvalue of 8%, any values beyond ±1.55 standard deviations of the mean in anormal distribution are rejected). In this way we can exclude 979, which givesus a correlation coefficient of 0.375 with the remaining 14 agents, which isstatistically significant.Several points should be considered in the results. First the ERP values fordrug effects in whole animals may be subject to errors. For example, thepharmacokinetic behavior of the drugs may lead to drug potencies in wholeanimal which do not reflect actions only on repolarizing K currents. Secondly,the two sets of data do correlate with the two outliers removed. Furthermore,86analysis of both ERP25 and MOT50 values show agreement in the identificationof RSD drugs which are not potent (see Fig. 20). Thus, unitary K(Ca) can serveas a reasonable model for drug actions in cardiac cells. Since single channelstudies on the cardiac K+ channels are limited by the low unitary conductance ofrepolarizing K+ channels (such as Ito and 1k), the K(Ca) in neurons have utilityas a screening device for putative class III agents.4.2.1. A comparison of the recovery timesTn addition to the drug potency, the K(Ca) experiments can also be usefulin determining the offset time for drugs. Eight RSD compounds were tested forthe recovery time in the inside-out patches. A 50 % recovery of the control meanopen time within first 20 seconds of reperfusion was considered as fast,intermediate recovery was defined as 50% recovery between 30-90 seconds, andslow recovery, if it occurred, after 2 minutes. The same RSD compounds weretested for the recovery period in the whole heart set-up (see description inRESULTS) and the comparison of the two sets of recovery time are given intable 4. The comparison shows a general agreement between the two sets ofrecovery times. All of 983, 984 and 987 were found to be fast in recovery (i.e.washed off quickly) in both set-ups, and 959 was persistent in both singlechannel and whole heart experiments. Two of the intermediate compounds in theinside-out patch clamp experiments, 968 and 969, were found to wash-offrapidly in the isolated heart experiment. Although a detailed comparisonbetween the recovery times was not possible, valuable information can beobtained from the K(Ca) experiments. Most evidently was 959 which clearly,relative to other RSD compounds, was persistent in action. Little recovery fromdrug actions was observed even after reperfusion of control solution for long87periods. This compound, in heart, has been found to be essentially irreversible.Thus K(Ca) would also seem to be useful in the assessment of drug wash-off.4.2.2. Inside-out vs. outside-out patchesOutside-out experiments (n=2) were earned out on five RSD compounds(949, 959, 969, 974, 979). A comparison of the MOT50 values obtained frominside-out patches and outside-out patches showed that 974 and 979 had similarpotencies in inhibiting the K(Ca). This result was consistent with that obtainedwith some other class III agents (McLarnon and Wang, 1991). However 949,959 and 969 were found to be somewhat less potent when applied to the outsideof the patch membrane. The low number of patches studied in the outside-outconfiguration (n=2 or 3) may have contributed to the measured differences inpotency. The use of inside-out vs outside-out patches is theoretically very usefulin determining whether the active site is internal or external. Unfortunately in oursystem a rapid perfusion system which can apply drugs in the order of ms wasnot used. Thus, it was not possible to differentiate between an internal orexternal active site. Previous work with tedisamil on the cardiac Ito also foundthat tedisamil blocked this cardiac current with both internal and externalapplication (Dukes and Morad, 1989).885. CONCLUSIONSA detailed investigation of the structure-activity relationship for RSDcompound was not possible at this stage since drug structures were notpublished. The establishment of a correlation between MOT50 and ERP25 isimportant since the patch clamp technique allows the measurement of drugactions on single channel K(Ca) activity. This allows a rapid method forexamination of potential Class III compounds and also gives importantimfonnation on sites of action and wash-off phenomena. At present similarexperiments on unitary currents in cardiac cells are not feasible, in part due tothe low unitary conductance of repolarizing K+ channels in heart. Thecorrelation between drug actions on K(Ca) and ERP values in heart wouldsuggest former data have relevance to a description of drug mechanisms ofaction. The investigation suggests the utility of more detailed single channelexamination of these potential class III agents in the future. The procedure mayalso provide a rapid screening device in the development of putative class Illantiarrhythmic compounds.Some caution should be exercised in the application of drug effects onK(Ca) in neurons to describe drug actions in heart. The primary repolanzing Kchannel in rat myocytes is Ito and although this current is modulated by calciumit is not dependent on calcium for activation as is K(Ca) in neurons. Indeed, theanalog of Ito in neurons is most likely to be IA (the transient rapid repolarizingK current) and it is well known that the pharmacological profile of IA andK(Ca) are different. It would seem evident from the present work, however, thatconsiderable homology occurs for repolarizing channels in different tissues.89ReferencesAdams PR, Constanti A, Brown DA, Clark RB (1982): Intracellular Ca2+ activates afast voltage-sensitive K+ current in vertebrate sympathetic neurones. Nature 296:746-749.Akbarali H, Nakajima T, Wyse DG, et Giles W (1990): Ca2+-activated K+ currentsin smooth muscle. Can. I Physiol. Pharmacol. 68:1489-1494.Banker and Cowan (1977): Rat hippocampal neurons in dispersed cell culture. BrainResearch 126: 397-425.Barret iN, Barret EF, Dribin LB (1981): Calcium-dependent slow potassiumconductance in rat skeletal myotubes. Dev. Biol. 82: 25 8-66.Barret N, Magleby KL, Pallota BS (1982): Properties of single calcium-activatedpotassium channels in cultured rat muscle. I PhysioL (Lond.) 331: 22 1-30.Beatch GN, Abraham S, MacLeod BA, Yoshida NR, Walker MJA (1991):Antiarrhytbmic properties of tedisamil (KC8857), a putative transient outward K+current blocker. Br. I Pharmacol. 102: 13-18.Blair AL, Dionne VE (1985): Developmental acquisition of Ca2+ sensitivity by K+channels in spinal neurons. Nature 315: 329-31.Blatz AL, Magleby KL (1984): Ion conductance and selectivity of single calcium-activated potassium channels in cultured rat muscle. 3. gen. Physiol. 84:1-23.Blatz AL, Magleby KL (1987): Calcium-activated potassium channels. Trends inNeurosci. 10:463-7.Colatsky TJ, Folimer CH (1989): K+ channel blockers and activators in cardiacarrhythmias. Cardiovasc. Drug Rev. 7: 199-209.Colatsky TJ, Folimer CR, Bird LB (1989): Cardiac electrophysiologic effects of WY-48,986, a novel Class ifi antiarrhythmic agent I Molec. Cell Cardiol. 21: S20.Colatsky TJ, Foilmer CR, Starmer CF (1990): Channel specificity in antiarrhythniicdrug action. Mechanism of potassium channel block and its role in suppressing andaggravating cardiac arrhythinias. Circulation 82: 2235-42.90Colquhoun D, Sigworth FJ (1983): Statistical analysis and fitting of single-channelrecords, in Single-channel Recording, Sakmann B and Neher E, eds.: 19 1-263.Plenum Press, New York.Deitmer JW, Eckert R (1985): Two component of Ca-dependent potassium current inidentified neurones ofAplysia calfornica. Pfluegers Arch 403: 353-359.Dukes ID, Cleeman L, Morad M (1990): Tedisamil blocks the transient and delayedrectifier K+ currents in mammalian cardiac and glial cells. JPharmacol. Exp. Ther.254: 560-569.Dukes ID, Morad M (1989): Tedisainil inactivztes transient outward K+ current in ratventricular myocytes. Am. I Physiol. 257: H1746-H1749.Escande D, Mestre M, Cavero I, Brugada 3, Kirchhof C (1989): RP 58866 and itsactive enantiomer PP 62719 (terkalant): Blockers of the inward rectifier K+ currentacting as pure class ifi antiarrhythmic agents. I Cardiovasc. Pharmacol. 20 (suppl2): S106-S113.Fabiato A, Fabiato F (1979): Calculator programs for computing the compostion ofsolutions containing multiple metals and ligands used for expenments in skinnedmuscle cells. I Physiol. 75: 463-505.Folimer CH, Poczobutt MT, Colatsky TJ (1989): Selective block of delayedrectification (1k) in feline ventricular myocytes by WY-48,986, a novel class Illantiarrhythmic agent. I Mol. Cell Cardiol. 21: S 185.Gardos G (1958): The function of calcium in the potassium permeability of humanerythrocytes. Biochim. Biophys. Acta 30:653-4.(litter AR, Beyenbach KW, Chadwick CW, Gross P, Minuth WW, Fromter E (1987):High-conductance K+ channels in apical membranes of principal cells cultured fromrabbit renal cortical collecting duct anlagen. Pflugers Arch. 408:282-90.Golowasch J, Kirwood A, Miller C (1986): Ailosteric effects of Mg2+ on the gatingof Ca2+-activated K+ channels from mammalian skeletal muscle. J. Exp. Biol. 124:5-13.Gorman ALF, Thomas MV (1978): Changes in the intracellular concentration of freecalcium ions in a pace-maker neurone, measured with the metallocbromic indicatordye arsenazo ifi. I Physiol. (Lond) 275: 357-76.91Gustafsson B, Zangger P (1978): Effect of repetitive activation on theafterhyperpolarization in dorsal spino-cerebellar tract neurones. I Physiol. 275: 303-19.Gwilt M, Dairymple HW, Burges RA, Blackburn KJ, Arrowmith JE, Cross PE,Higgins AJ (1989): UK 68,798 is a novel potent and selective Class III antiarrhythmicdrug. I Mo!. Cell. Cardiol. 21: Si 1Gwilt M, Dairymple HW, Burges RA, Blackburn KJ, Dickinson RP, Cross PE,Higgins AJ (1991): Electrophysiologic properties of UK 66914, a novel class ifiantiarrhytbmic agent. J. Cardiovasc. Pharmacol. 17: 376-3 85.Hermann A, Erxleben C (1987): Charybdotoxin selectively blocks small Ca-activatedK channels in Aplysia Neurons. 3. Gen. Physiol. 90:27-47.mile, B (1984): Ionic Channels ofExcitable Membranes. pp. 184-8. Suderland,Mass: Sinauer. 42’lpp.Hugues M, Romey G, Duval D, Vincent JP, Lazdunski M (1982): Apamin as aselective blocker of the Ca-dependent K channel in neuroblastoma cells. Voltage-clamp and biochemical characterization. Proc. natn. Acad. Sd. USA 79:1308-12.Kass RS, Arena JP, Walsh KB (1990): Measurement and block of potassium channelcurrents in the heart: importance of channel type. Drug Dev. Res. 19: 115-27.Katritsis D, Camm AJ (1993): New class III antiarrhythmic drugs. European HeartJournal 64 (suppl. 4): 93-99Kirkpatrick CT (1975): Excitation and contraction in bovine tracheal smooth muscle.J. Physiol. (London), 244: 263Lancaster B, Adams PR (1984): Single electrode voltage clamp of the slow AHPcurrent in rat hippocampal pyramidal cells. Soc. Neurosci. Abstr. 10: 257.3.Lancaster B, Adams PR (1986): Calcium-dependent current generating theafterhyperpolarization of hippocampal neurons. I Neurophysiol. 55: 1268-1282.Lancaster B, Madison DV, Nicoll RA (1986): Charybdotoxin selectively blocks a fastCa-dependent afterhyperpolarization (AHP) in hippocampal pyramidal cells.Neurosci. Abstr. 12: 560.Lancaster B, Nicoll RA, Perkel DJ (1991): Calcium activates two types of potassiumchannels in rat hippocampal neurons in culture. I Neuroscience 11: 23-30.92Latorre R,. Miller C (1983): Conductance and selectivity in potassium channels. IMembr. BioL 71:11-30.Latorre R, Vegara C, Hidalgo C (1982): Reconstitution in planar lipid layers of aCa2+-dependent K+ channel from transverse tubule membranes isolated from rabbitskeletal muscle. Proc. Nat!. Acad. Sd. USA 79:805-9.Latorre R (1986): The large calcium-activated potassium channel. In Ion ChannelReconstitiution, ed. C. Miller, pp.43 1-67. New York: Plenum. 5T7pp.Lewis T, Drury AN, ffiesu CC, Wedd AM (1921): Observations relating to the actionof quinidine upon the dog’s heart: with special reference to its action on clinicalfibrillation of the amides. Heart 9: 55-86.MacKinnon R, Miller C (1988): Trimethyloxomum modification of the highconductance Ca2+-activated K+ channel. Biophys. 1 53: 260a.Madison DV, Nicoll RA (1982): Noradrenaline blocks accommodation of pyramidalcell discharge in the hippocampus. Nature 299: 63 6-8.Madison DV, Nicoll RA (1984): Control of repetitive discharge of rat pyramidalneurones in vitro. I. Physiol. 354: 319-31.Martin CL, Chinn K (1992): Contribution of delayed rectifer and inward rectifier torepolarization of the action potential: Pharmacolgical separation. I Cardiovasc.PharmacoL 19: 830-837.Marty A (1981): Calcium-dependent channels with large unitary conductance inchromaffin cell membranes. Nature 291: 497-500.Maue RA, Dionne VE (1987): Patch-clamp studies of iolated mouse olfactoryreceptor neurons. I Gen. Physiol. 90:95-125.McLarnon JG, Wang XP (1991): Actions of cardiac drugs on a calcium-dependentpotassium channel in hippocampal neurons. Mole. Pharmacology 39: 540-546.McLarnon JG, Sawyer D, Michikawa M, Kim SU (1992): Tedisamil blocks acalcium-dependent potassium channel in cultured motoneurons. Neuroscience Letters144: 185-188.93McLamon JG, Sawyer D (1993): Effects of divalent cations on the activation of acalcium-dependent potassium channel in hippocampal neurons. Pflugers Arch 424: 1-8.McManus OB, Magleby KL (1988): Kinetic states and modes of single large-conductance cakium-activated potassium channels in cultured rat skeletal muscle. .JPhysiol. (Lond.) 402: 79-120.Meech RW, Strumwasser F (1970): Intracellular calcium injection activatespotassium conductance in Aplysia nerve cells. Fed Proc. 29:834aMeech RW (1978): Calcium-dependent potassium activation in nervous tissues.Annu. Rev. Biophys. Bloeng. 7:1-18.Miller C, Moczydlowski E, Latorre R, Philipps M (1985): Charybdotoxin, a proteininhibitor of single Ca2+-activated K+ channels from mamalian skeletal muscle.Nature 313:316-8.Miller C, ed. (1986): Ion Channel Reconstitution. New York: Plenum. 577 pp.Moczydlowski E, Latorre R (1983): Gating kinetics of Ca2+-activated potassiumchannels from rat muscle incorporated into planar lipid bilayers: evidence for twovoltage-dependent Ca2+ binding reactions. I Gen. Physiol. 82:511-42.Moczydlowski E, Alvarez 0, Vergara C, Latorre R (1985): Effect of phospholipidsurface charge on the conductance and gating of a Ca2+ - activated K+ channel inplanar lipid bilayers. I Membr. Biol. 83: 273-82.Neher E, Salcmann B, eds. (1983): Single-channel Recording. New York: Plenum.503 pp.Noble D (1984): The surprising heart: A review of recent progress in cardiacelectrophysiology. J. Physiol. 353: 1-50.Pallotta BS, Magleby KL, Barrett JN (1981): Single channel recordings of a Ca2+activated K+ current in rat muscle cell culture. Nature 293: 47 1-4.Pennefather P, Lancaster B, Adams PR, Nicoll RA (1985): Two distinct Ca-dependent K currents in bullfrog sympathetic ganglion cells. Proc. nain. Acad. Sd.U.S.A. 82:3040-44.94Penz WP, Pugsley MK, Hsieh MZ, Walker MJA (1992): A new ECG measure (RSh)for detecting possible sodium channel blockade in vivo in rats. I Pharmacol.Methods. 27: 5 1-58.Peterson OH, Maruyama Y (1984): Calcium-activated potassium channels and theirrole in secretion. Nature 307: 693-6.Pfrunder D, Kreye VAW (1992): Tedisamil inhibits the delayed rectifier K+ currentin single smooth muscle cells of the guinea-pig portal vein. Pflugers Arch 421: 22-25.Pfrunder D, Kreye VAW (1991): Tedisamil blocks single large-conductance Ca2+ -activated K+ channels in membrane patches from smooth muscle cells of the guinea-pig portal vein. Pflugers Arch 418: 308-312.Reeves R, Farley 3, Rudy B (1986): cAMP dependent protein kinase opens several Kchannels from mammalian brain. Soc. Neurosci. Absir. 13: 1343.Ritchie AK (1987): Two distinct calcium-activated potassium currents in a ratanterior pituitary cell line. I Physiol. 385: 59 1-609.Romey G, Lazdunski M (1984): The coexistence in rat muscle cells of two distinctclasses of Ca2+-dependent K+ channels with different pharmacological propertiesand different physiological functions. Biochem. biophys. Res. Commun. 118: 669-74.Sah P. McLachlan EM (1992): Potassium currents contributing to action potentialrepolarization and afterhyperpolarization in rat vagal motoneurons. I Neurophysiol.68(5): 1834-41.Sanguinetti MC, Jurkiewicz NK (1990): Two components of cardiac delayed rectifierK+ current: differential sensitivity to block by class ifi antiarrhytbmic agents. J. Gen.Physiol. 96: 194-214.Singer 33, Walsh TV (1984): Large conductance Ca activated K channels in smoothmuscle cell membrane. Biophys. J. 45:68-70.Singer 33, Walsh TV (1987): Characterization of calcium-activated potassiumchannels in single smooth muscle cells using the patch-clamp technique. PfluegersArch. 408: 98-111.Singh BN, Nademanee K (1985): Control of arrhythmias by selective lengthening ofcariac repolarization: theoretical considerations and clinical observations. AlviHeart1 109: 421-30.95Singh BN (1990): Advantages of beta blockers versus antiarrhythmic drugs andcalcium antagonists in secondary prevention in survivors of myocardial infarction.Am. J. Cardio. 66: 9-20.Smart TG (1987): Single calcium-activated potassium channels recorded fromcultured rat sympathetic neurones. I physiol. 389: 337-360.Storm iF (1987): Action potential repolarization and a fast after-hyperpolarization inrat hippocampal pyramidal cells. J. Physiol. 365: 733-59.Storm IF (1990): Potassium currents in hippocampal pyramidal cells. Prog. brain res.83: 161-87.Tomita T (1988): Ion channels in smooth muscle studied with patch clamp methods.Jpn. J. Physiol. 38: 1-18.281.Villarroel A, Eisenman 0(1987): Surface charge in a barrier model can explain thelow concentration I-V behavior of the Ca2+-activated K+ channel. Biophys. J. 51:546a.Walker MJA, Beatch GN (1988): Electrically induced arrhythmias in the rat. Proc.West. Pharmacol. Soc. 31: 167-170.Walsh JV, Singer JJ (1983): Ca2+ - activated K+ channels in vertebrate smoothmuscle cells. Cell Calcium 4: 321-330.Yellen G (1984): Ionic permeation and blockade in Ca2+-activated K+ channels ofbovine chromaffin cells. I Gen. Physiol. 84:157-86.Yellen G (1987): Permeation in potassium channels: implications for channelstructure. Annu. Rev. Biophys. Chem. 16: 227-46.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items