I. 0..., .r v .v.:‘. . .. s '3: .5: .z ‘. 1...: . 4.. ...12.:X:.-V. .. >1 {.22}. (-11. . 1. .. .5} : ‘63:. u. .: . o (- that 811...} .7 . 1’51: #1:.) u; t. 3.... .. fiat (- u .0 12.31. I nquIVoI ‘VI . IJ-Iv.lh§ vii v gilt..lr-||‘¢“¢. O! t A: I‘vrltob'il Iii. ll TH£$QL llllllllulll This is to certify that the dissertation entitled Effects of Lambert-Eaton Myasthenic Immunoglobulin on Synaptosomal Calcium Channels: Specificity, Serum Dependence and Subtype presented by Sandra Jeanne Hewett has been accepted towards fulfillment of the requirements for ph,D, Pharmacology/Toxicology . degree in “X Z ' ~ - lit/4% {5/ Major professor Datefléfwfojg /??;\ .MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 f a. LIBRARY ' Michigan State l University ‘ .J —_ PLACE IN RETURN BOX to remove this checkom from your record. TO AVOID FINES return on or before date due. ’féfi" DATE DUE DATE DUE DATE DUE # ‘1 J —l MSU Is An Affirmative Action/Equal Opportunity Institution cWMS—DJ EFFECTS OF LAMBERT-EATON MYASTHENIC IMMUNOGLOBULIN ON SYNAPTOSOMAL CALCIUM CHANNELS: SPECIFICITY, SERUM DEPENDENCE AND SUBTYPE BY Sandra Jeanne Hewett A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Pharmacology and Toxicology 1992 ABSTRACT EFFECTS OF LAMBERT-EATON MYASTHENIC IMMUNOGLOBULIN ON SYNAPTOSOMAL CALCIUM CHANNELS: SPECIFICITY, SERUM DEPENDENCE AND SUBTYPE BY Sandra Jeanne Hewett Lambert-Eaton Myasthenic Syndrome (LEMS) is a presynaptic disorder of neuromuscular transmission. The reduction in nerve evoked release of transmitter at the neuromuscular junction of patients with LEMS is thought to be caused by an autoantibody to motor nerve terminal Ca2+ channels. The overall goal of this dissertation was to test the hypothesis that a circulating autoantibody in the serum of patients with LEMS interferes with the function of nerve terminal Ca2+ channels associated with transmitter release. To test this a model system of isolated nerve terminals from mammalian CNS (synaptosomes) was used. Serum and plasma from several patients and IgG from a single patient with LEMS reduce 45Ca2+ influx into synaptosomes. The antibody appeared specific for Ca2+ channels as it was unable to disrupt 22Na+ uptake into or 86Rb" efflux from Na+ or K+ channels, respectively. Synaptosomal membrane integrity as well as synaptosomal membrane potential were unaltered following incubation with the LEMS autoantibody. In synaptosomes, the presence of human serum is important in mediating a functional change in Ca2+ channel activity by LEMS lgG. In the absence of human serum, LEMS lgG did not reduce uptake of 45Ca2+ into synaptosomes. Therefore, the role of the serum component complement in the pathogenesis of LEMS was assessed. Neither the membrane attack complex nor the alternative pathway of complement contributed to the disruption of Ca2+ channel function by LEMS lgG and serum in synaptosomes. In contrast, the 03 component of the complement cascade appeared to be important in mediating Ca2+ channel dysfunction. The pharmacological determination of the Ca2+ channel subtype affected by lgG from a patient with LEMS was next examined. The density of binding sites for [3H]-verapamil but not [3H]-nitrendipine nor [‘zsl]-ro-conotoxin was reduced by approximately 45% following incubation with LEMS lgG alone or LEMS lgG and serum. This suggests that the IgG molecule itself interacts at this site. Verapamil is presumed to act as an L-type Ca“ channel antagonist but recent evidence suggests that it may not be as selective as was previously thought. The studies in this dissertation were the first to examine specifically the ability of antibody from LEMS patients to interact with nerve terminal Ca”+ channels. This work has strengthened the hypothesis that Ca2+ channels are the antigenic target, provided new insight into the mechanism of Ca2+ channel dysfunction, and explored pharmacologically the subtype of Ca2+ channels in a nerve terminal preparation affected by the LEMS autoantibody. To my family, with all my love Especially my parents, Ann and Dennis Simas and my husband, Jim ACKNOWLEDGEMENTS I would like to thank my mentor, Dr. Vlfilliam Atchison, for providing me with the opportunity to grow and develop into an independent researcher. I am grateful for the excellent communication skills I have developed as a direct result of his guidance. He has been an invaluable role model demonstrating that persistence and patience invariably pay off. I’d also like to thank the members of my thesis committee, Drs. James Bennett, Kathy Brooks, Jim Galligan and Mary Rheuben for their guidance and flexibility. Special thanks to Drs. Jim Galligan, David Reinke and Mary Rheuben who were always willing to listen and lend advice. I am grateful to everyone in the laboratories of 3331, 346 and 347 (past and present), who are too numerable to list, for creating a most enjoyable working environment. I’d especially like to thank Annette McLane, Marintha Heil, Lynne Ireland, Drs. Jean Ray and Cindy Hoom for their friendship. They always seem to understand me and my moods as well as provide a shoulder when ever necessary. I am grateful for the excellent technical assistance of Michelle Welke, Paul Bertrand, Kevin Borden, Becky Heydecke and Annette McLane. I hope that their experiences in the laboratory were as worthwhile to them as their work has been to me. Special thanks to Gretchen Humphries for her technical, artistic and secretarial assistance. Most importantly, I would like to acknowledge my family. I would especially like to thank my mother and father, Ann and Dennis Simas, for always believing in me and teaching me that I could accomplish anything with hard work, discipline and spirit. I thank them for giving me the freedom to live, to grow and to leave even when I know it was difficult to let go. I’d like to thank John and Eileen Hewett for being the best second parents for which anyone could ever ask. Their support of both me and their son have helped us to accomplish lofty goals. Finally, I am forever grateful to my husband Jim. His strength, support, wit, companionship, understanding, and love have helped make this endeavor not only bearable but worthwhile. He encouraged me from beginning to end and always believed in me even when I had trouble believing in myself. Jim, for everything I thank you and I love you. vi PREFACE Some of the data contained in this dissertation have been published previously. Chapter Three appeared as Hewett and Atchison (1991). vii TABLE OF CONTENTS viii LIST OF TABLES ................................................................................................ xii LIST OF FIGURES ............................................................................................. xiii LIST OF ABBREVIATIONS ................................................................................. xvi CHAPTER ONE: INTRODUCTION A. General Introduction ................................. 2 B. Normal Neuromuscular Transmission ..................... 3 1. Synaptic transmission ............................ 3 2. Calcium channel subtypes ........................ 8 C. Disorders of Neuromuscular Transmrssion ................ 14 D. Lambert-5mm Myasthenic Syndrome: General Description ..... 15 1. Clinical Features ............................... 15 2. Electrophysiology .............................. 17 a. Clinical ................................. 17 b. Experimental ............................ 19 3. Biochemistry ................................. 20 4. Morphology .................................. 20 a. Morphometric analysis ..................... 20 b. Ultrastructural analysis ..................... 22 E. Lambert-Eaton Myasthenic Syndrome: Autoimmunity .......... 23 1. Clinical and experimental evidence .................. 23 2. The Antigenic Target ........................... 25 a. Calcium channel hypothesis ................. 25 b. Calcium channel subtype affected by LEMS lgG . . . 29 3. Pathophysiological Mechanisms ................... 30 a. Direct inhibition ........................... 30 b. Antigenic modulation ...................... 31 c. Complement-mediated destruction ............. 32 F. The Immune Response ............................... 36 1. Cellular interactions in the immune system ............ 37 2. Self- non-self recognition ......................... 40 a. Clonal Deletion ........................... 40 b. Clonal Anergy ........................... 41 3. Breakdown of self-tolerance ...................... 43 a. Decreased Suppression .................... 44 b. Modification of an autoantigen ................ 46 0. Self antigen plus aberrant self-encoding molecules . 46 d. Release of sequestered antigens .............. 47 e. Cross-reactivity ........................... 47 G. Autoimmune LEMSandtheroleotcancer ................. 49 CHAPTER TWO: SYNAPTOSOMES AND SYNAPTOSOMAL ION CHANNELS A General Introduction ................................ 53 B. Synaptosomes ..................................... 54 1. General Description ............................ 54 2. Synaptosomal ion channels ....................... 55 a. Sodium channels ......................... 55 b. Potassium channels ....................... 56 c. Calcium channels ......................... 57 i. Ca2+ channel subtypes ................. 60 C. Overall Aim ....................................... 62 1. Specific Aim 1 ................................ 64 2. Specific Aim 2 ................................ 64 ix 3. Specific Aim 3 ................................ 65 4. Specific Aim 4 ................................ 65 CHAPTER THREE: EFFECT OF SERUM AND PLASMA FROM PATIENTS WITH LAMBERT-EATON MYASTHENIC SYNDROME ON DEPOLARIZATION- INDUCED UPTAKE OF “Ca2+ INTO RAT CORTICAL SYNAPTOSOMES A. Summary ........................................ 68 B. Introduction ....................................... 69 C. Materials and Methods ............................... 70 D. Results .......................................... 75 E. Discussion ....................................... 78 CHAPTER FOUR: SPECIFICITY OF LAMBERT-EATON MYASTHENIC IMMUNOGLOBULIN FOR NERVE TERMINAL CALCIUM CHANNELS A. Summary ........................................ 87 B. Introduction ....................................... 88 C. Materials and Methods ............................... 90 D. Results .......................................... 99 E Discussion ...................................... 1 14 CHAPTER FIVE: DISRUPTION OF SYNAPTOSOMAL 032+ CHANNELS BY LAMBERT-EATON MYASTHENIC SYNDROME IMMUNOGLOBULIN: SERUM DEPENDENCY A. Summary ....................................... 122 B. Introduction ...................................... 123 C. Materials and Methods .............................. 124 D. Results ......................................... 126 E. Discussion ...................................... 136 CHAPTER SIX: PHARMACOLOGICAL DETERMINATION OF SYNAPTOSOMAL C CHANNEL SUBTYPE S AFFECTED BY LAMBERT-EATON MYASTHENIC SYNDROME IMMUNOGL ULIN A. Summary ....................................... 143 B. Introduction ...................................... 144 C. Materials and Methods .............................. 146 D. Results ......................................... 150 E Discussion ...................................... 157 F Note ........................................... 162 CHAPTER SEVEN: SUMMARY AND CONCLUSIONS A. Summary and Conclusions ........................... 164 LIST OF REFERENCES ................................... 173 LIST OF TABLES Table 1.1. Pharmacological and Biophysical Characteristics of Mammalian Ca2+ Channel Types ............................. 10 Table 4.1. Determination of Ouantal Content via Direct Method ....... 101 Table 6.1 [3H]Nitrendipine Binding Characteristics ................ 152 xii Figure 1.1. Figure 1.2. Figure 1.3. Figure 1.4. Figure 1.5. Figure 3.1. Figure 3.2. Figure 3.3. Figure 3.4. Figure 3.5. Figure 4.1. Figure 4.2. Figure 4.3. Figure 4.4. Figure 4.5. Figure 4.6. Figure 4.7. LIST OF FIGURES Neuromuscular Transmission ......................... 4 Electromyographic Characteristics of LEMS .............. 18 Complement .................................... 33 Immune Response ................................ 38 Pathogenesis of autoimmune disease .................. 45 Effect of sera from 4 patients with SCC on "‘SCaz+ uptake into rat forebrain synaptosomes ........................ 76 Effect of plasma and sera from 5 patients with LEMS on uptake of 45Ca2+ into rat forebrain synaptosomes ......... 77 Effect of serum from one patient with LEMS on “Ca2+ uptake into synaptosomes. .............................. 79 Effect of LEMS sera on synaptosomal viability. ............ 80 Effect of LEMS plasma on synaptosomal viability. .......... 81 Setup for recording under current clamp ................ 94 Effects of chronic treatment of mice with LEMS plasma on quantal content. ................................ 100 Frequency-dependent facilitation of neuromuscular transmission 102 Effect of LEMS lgG on uptake of “Ca2+ into rat forebrain synaptosomes ................................. 103 Effect of lgG from patients with small cell carcinoma ($00) on uptake of “Ca2+ into rat forebrain synaptosomes ........ 105 Effect of LEMS lgG on synaptosomal membrane integrity. . . . 106 Representative graph depicting typical fluorescent dye experiment ................................... 108 xiii Figure 4.8. Effect of LEMS lgG on synaptosomal membrane potentials . . 109 figure 4.9. Time course of 22Na+ uptake into rat forebrain synaptosomes. 110 Figure 4.10. Effect of LEMS lgG on uptake of ”Na“ into rat forebrain synaptosomes ................................. 1 1 1 figure 4.11. Time course of ast+ loading into synaptosomes ........ 112 figure 4.12. Effect of LEMS lgG on ”RE efflux from rat forebrain synaptosomes ................................. 1 13 Figure 5.1. Effect of LEMS lgG on uptake of 45Ca2+ into rat forebrain synaptosomes in the presence and absence of control human serum ................................. 127 figure 5.2. Effect of control human serum on “Ca2+ uptake into rat forebrain synaptosomes .......................... 128 figure 5.3. Effect of BSA on “Caz+ uptake into rat forebrain synaptosomes ................................. 129 figure 5.4. Effect of control and LEMS lgG and various concentrations of control human serum on 45Ca2+ uptake into rat forebrain synaptosomes ................................. 131 Figure 5.5. The effect of LEMS lgG in the presence of control or heat inactivated serum on uptake of “Ca2+ into rat synaptosomes ....... 132 figure 5.6. The effect of 05-deficient human serum on the LEMS lgG induced reduction of K”-stimulated 45Ca2+ uptake into rat synaptosomesl33 figure 5.7. The effect C3-deficient human serum on the LEMS IgG in the reduction in 1 sec K+-stimulated 45Ca2+ uptake into rat synaptosomes. ................................ 134 figure 5.8. The effect of control and LEMS lgG in the presence of Factor B-deficient human serum 45Ca2 uptake into rat synaptosomes.135 figure 6.1. Effect of serum on nitendipine binding to synaptosomes . . . . 151 figure 6.2. Comparative effects of Control and LEMS lgG on nitrendipine binding to synaptosomes. ............................. 153 figure 6.3. Comparative effects of control and LEMS lgG on Cng binding to rat forebrain synaptosomes. ..................... 155 figure 6.4. Comparative effects of control and LEMS lgG on verapamil binding to synaptosomes. ......................... 156 LIST OF ABBREVIATIONS ACh - acetylcholine AChE — acetylcholinesterase ANOVA - analysis of variance APC - alternative pathway of complement AZPs - active zone particles B"m - maximum number of binding sites BSA - bovine serum albumin BoTX - botulinum toxin BTX - batrachatoxin Ca2+ - calcium [Ca2+], - extracellular concentration of calcium ChAT - choline acetyltransferase CMAP - compound muscle action potential CNS - central nervous system Cng - m -conotoxin GVIA CVF - cobra venom factor DHP - dihydropyridine dis-02(5) 3,3’-diethyldicarbocyanine iodide DMSO - dimethyl sulfoxide DRG - dorsal root ganglion EGTA - ethyleneglycol bis-(B-aminoethyl ether)-N,N,N,N’-tetra acetic acid EPP - endplate potential EtOH - ethanol Fab - monovalent antigen binding fragment of IgG (papain-digested) F(ab)’2 - divalent antigen binding fragment of lgG (pepsin-digested) FIX - funnel web spider toxin HEPES - N-2-hydroxyethylpiperazine-N’2-ethanesulfonic acid lgG - immunoglobulin G lgs - total immunoglobulin K“ - potassium [K"], - extracellular concentration of potassium kDa- kilodalton 'Kd - equilibrium dissociation constant for ligand binding LDH - lactate dehydrogenase LEMS - Lambert-Eaton Myasthenic Syndrome MAC - membrane attack complex MEPP - miniature endplate potential MG - myasthenia gravis Mg2+ - magnesium MHC - major histocompatability complex mRNA - messenger ribonucleic acid msec - millisecond Na+ - sodium NaCN - sodium cyanide NiCI2 - nickel chloride NMJ - neuromuscular junction PAGE - polyacrylamide gel electrophoresis PKC - protein kinase C 800 - small cell carcimona of the lung SDS - sodiumdodecylsulfate sec - seconds SEM - standard error of the mean synaptosomes - isolated nerve terminals from rat forebrain TBA - tetrabutylammonium chloride TEA - tetraethylammonium chloride Th cells - helper T lymphocytes T, cells - suppressor T lymphocytes TTX - tetrodotoxin VER - veratridine 4-AP - 4-aminopyridine xviii CHAPTER ONE: INTRODUCTION A. General Introduction: Lambert-Eaton myasthenic syndrome (LEMS) is an autoimmune disorder characterized symptomatically by a generalized muscle weakness. The neuromuscular junction (NMJ) was implicated as the site of the abnormality in LEMS when no aberration could be found in the responses of the motor nerve or muscle to direct electrical stimulation. The presynaptic terminal, specifically, was implicated as the site of action of the LEMS autoantibody when analysis of the NMJ in biopsy specimens from patients with the syndrome demonstrated that the postsynaptic membrane responds normally to acetylcholine (ACh). Since a presynaptic defect could involve an abnormality in transmitter synthesis, storage, or release, a great deal of effort has been expended in order to determine which of these processes is affected. Experimental evidence indicates that both synthesis and storage of ACh are normal. Thus, the abnormality appears to be caused by a defect in neurotransmitter release. It has been suggested that membrane calcium (Ca2+) channels are the antigenic target for the LEMS autoantibody. The initial sections of Chapter One will introduce normal neuromuscular transmission. These sections provide a description of the components of the neuromuscular junction as well as the dependence of transmission on [Ca2+]. ions. Ca2+ channel subtypes believed to be involved in neurotransmitter release, in general, are discussed in detail. This is followed by a detailed discussion of the autoimmune, neuromuscular disorder, LEMS with attempts to delineate the stage at which normal functioning is altered by a pathological process. The final sections 3 of this chapter discuss the normal immune response focussing on the mechanisms of self- non-self recognition and how the breakdown of self-tolerance can lead to autoimmunity. B. Normal Neuromuscular Transmission: 1. Synaptlc transmlsslon The synapse formed between motor nerves and muscle cells is known as the neuromuscular junction (NMJ). The NMJ can be considered a transducing mechanism, in which electrical signals (nerve impulses) are transduced to chemicals signals (neurotransmitter release) and then back to electrical signals (muscle action potentials). The transmission of an impulse from the motor nerve to the muscle fiber involves a complex sequence of events. The structures and steps involved in neuromuscular transmission are shown in figure 1.1. The presynaptic part of the transducer consists of the nerve terminal. The terminal rests in a depression in the surface of the muscle fiber, the postsynaptic part of the transducer, and is separated from the muscle by the synaptic cleft (Bowden and Duchen, 1976). The nerve terminal contains large numbers of synaptic vesicles, each containing approximately 10,000 to 20,000 molecules of ACh (Potter, 1970; Kuffler and Yoshikami, 1975). Each vesicle represents one quantum of ACh (del Castillo and Katz, 1956; Martin, 1966; Hubbard, 1970). Transmitter release from the nerve terminal occurs both spontaneously and as a result of invasion of the axon terminal by an action potential. The spontaneous 5 - choline + acetate FIGURE 1.1. Neuromuscular Transmission Following stimulation of the nerve, an action potential will be propagated down the axon, leading to depolarization of the nerve terminal (1). Voltage-sensitive Ca2+ channels open and allow Ca2 ions to enter into the terminal (2). An increase in intraterminal Ca2+ triggers the fusion of synaptic vesicles with the plama membrane and the subsequent release of ACh into the synaptic cleft (3). ACh binds to nicotinic ACh receptors in the muscle membrane resulting in an excitatory postsynaptic potential (4). Diffusion of ACh away from the receptors and hydrolysis of ACh by AChE rapidly terminates the ACh-receptor combination (5). 5 release of one quantum of ACh leads to tiny depolarizations of the postsynaptic membrane known as miniature end-plate potentials (MEPPs) (Fatt and Katz, 1951; Boyd and Martin, 1956a; Liley, 1956). Stimulated release of many quanta of ACh leads to an excitatory postsynaptic potential which at the NMJ is termed an endplate potential (EPP) (Fat and Katz, 1959; Takeuchi, 1963). The number of vesicles released per nerve impulse is known as the quantal content of the EPP. Action potentials are propagated along the axon of the motor neuron by movement of Na+ and K+ into and out of the axon through their respective channels (Hodgkin and Huxley, 1952). Once the action potential reaches the axon terminal, it causes a depolarization of the terminal membrane (Brigant and Mallart, 1982; Mallart, 1985). As a result, voltage-dependent Ca2+ channels in the nerve terminal open allowing extracellular Caz+ ions to enter down their electrochemical gradient (Katz and Miledi, 1967a,b; Uinas et al., 1981). It is thought that these channels are clustered in specialized areas or active zones, where vesicles containing neurotransmitter are arranged in close apposition to the cell membrane for subsequent release. (Heuser et al., 1974, 1979; Ceccarelli and Hurlbut, 1980; Pumplin et al., 1981; Robitaille et al., 1990). ctat2+ entry produces the near simultaneous exocytotic release of large numbers of quanta of ACh from the active zone area. The presence of [Ca2 *1. is required for any form of depolarization to release ACh from nerve terminals. This requirement is not connected to the propagation 6 of action potentials into nerve terminals, for in conditions of Ca2+ deprivation, action potentials can still be recorded even though no ACh release occurs (Katz and Miledi, 1965; Dodge et al., 1969). The sequence of molecular events linking depolarization-dependent Ca2+ influx to the release of neurotransmitters from nerve terminals is unknown. However, interactions between Ca2+ and intracellular Ca2+- binding proteins are thought to trigger directly or indirectly the fusion of synaptic vesicles with the cell membrane, resulting in the release of neurotransmitter into the synaptic cleft (Augustine et al., 1987). One model suggests that activation of a Ca2+-binding protein in turn activates a protein kinase which phosphorylates a protein or proteins essential for the release of neurotransmitter (Krueger et al., 1977; DeLorenzo et al., 1979; Dekker et al., 1989a). Evidence for this centers on studies which demonstrate that the phosphorylation of synapsin l and B-50 (GAP-43) by Ca2*/calmodulin- dependent protein kinase II and Protein kinase C, respectively, correlate with neurotransmitter release. (Uinas et al., 1985; Robinson and Dunkley, 1985; Dekker et al., 1989a; Sihra et al., 1989; Dekker et al.,1990; Nichols et al., 1992). Furthermore, antibodies to 8-50 or injection of deposphorylated synapsin l inhibit release from the nerve terminal (Uinas et al., 1985; Dekker et al., 1989b). Synaptophysin, synaptotagmin, synexin, calpactin and other members of the calelectrin family possess a cytoplasmic Caz*-binding site and have also been implicated in the control of neurotransmitter release (Augustine et al., 1987; Smith and Augustine, 1988; Brose et al., 1992). Depolarization of nerve terminals is 7 accompanied by the activation of a calcium-stimulated protein phosphatase, as well. The physiological significance of the rapid dephosphorylation of the phosphoproteins, dephosphin, P96 and P139 (Robinson et al., 1987; Sihra et al., 1992) has yet to be determined. Once released, ACh crosses the synaptic cleft and binds to nicotinic ACh receptors located on the crests of the synaptic folds across from release sites (Landau, 1978; Ceccarelli and Hurlbut, 1980). The interaction between ACh and its receptor produces a conductance change in the postsynaptic membrane, markedly increasing the membranes permeability to Na+ and K+ (Takeuchi and Takeuchi, 1959; Takeuchi, 1963). This produces a local depolarization in the end- plate (Fatt and Katz, 1951; Boyd and Martin, 1956b). EPPs of sufficient magnitude spread from the endplate region to the surrounding muscle fiber membrane initiating an action potential. This leads to contraction of the muscle (Hubbard, 1973) Diffusion of ACh away from the receptors and hydrolysis of ACh by acetylcholinesterase (AChE) rapidly terminates the ACh-receptor combination. The choline molecules liberated by AChE action are actively taken up by nerve terminals for resynthesis into ACh (choline + acetyl CoA) by choline acetyltransferase (ChAT)(Collier and Macintosh, 1969; Potter, 1970; Kuher and Murrin, 1978). 2. Calclum channel subtypes Voltage-dependent Ca2+ channels occur in most neurons and in several other cell types. As stated above, influx of [Ca2+], through membrane ionic channels is the critical step for coupling the action potential to transmitter secretion (Katz and Miledi, 1965; 1967a,b; Dodge and Rahamimoff, 1969). Evidence indicates that more than one type of calcium channel exists (Nowycky, et al., 1985a,b; Perney, et al., 1986). This discovery has prompted questions as to which subtype(s) is (are) involved with the release of neurotransmitter. Wlth the exception of some invertebrate neurons, characterization of the type of channel that exists at the active zone is impossible as nerve terminals are too small for direct recording techniques. Therefore, electrophysiological and pharmacological characterizations of the various types of calcium channels were made in the cell some with the assumption that somal events model events occurring in the nerve terminal. Thus, the exact identity of the subtypes of calcium channels in the nerve terminal remains unknown. It is possible that the characteristics of calcium channels in the terminal may differ from those in the cell body (Leonard et al., 1987; Suszkiw et al., 1989; Charlton and Augustine, 1990). Three distinct channel subtypes have been described in a dorsal root ganglion (DRG) preparation (Nowycky et al., 1985a,b; Fox et al., 1987). The channel subtypes, T, L, and N are distinguishable by their differential kinetics of activation and inactivation, as well as their pharmacological sensitivity to agonists and antagonists. A fourth type has been defined in mammalian Purkinje cells 9 (guinea pig cerebellum) and in presynaptic terminals of the squid giant synapse (Uinas et al., 1989; Lin et al., 1990). This channel differs, both electrophysiologically and pharmacologically, from the previously defined T-, L- and N-type and is termed the P-type channel (for Purkinje)(lJinas et al., 1989; Lin et al., 1990). The biophysical and pharmacological profiles of the different Caz+ channel subtypes are summarized in Table 1.1. T-type (“transient”) channels activate at membrane potentials more positive than -70 mV, have very small unitary conductances, and inactivate quickly and completely during a maintained depolarization. L-type (“long-lasting") channels activate at postive test potentials ( > -10 mV), have a very large unitary conductance, and show little inactivation during a 200 msec depolarization . N-type (“neither T- nor L-") channels activate at positive test potentials ( > -20 mV), have a smaller unitary conductance than the L channel and decay with time constants ranging from 50-100 msec (Fox et al., 1987). Finally, P-type channels activate at relatively positive test potentials (-30 to -20 mV) and decay with a time constant of 650 msec (Uinas et al., 1989). Pharmacological differences exist between the Ca2+ channel subtypes also. The dihydropyridine (DH P) class of organic Caz+ channel agonists and antagonists such as Bay K 8644 and nitrendipine are presumed to interact with a site directly on or associated with L-type Ca2+ channels (Nowycky et al., 1985a) whereas N- and T- channels are unaffected by submicromolar concentrations of these drugs. The phenylalkylamine, verapamil and the benzothiazepine, diltiazem also appear 10 Table 1.1 Pharmacological and Biophysical Characteristics of Mammalian Ca2+ Channel Types T N L P Activation range >-70 >-20 >-10 -30--20 (mV) Inactivation fast moderate slow slow Kinetics r=m-50ms r=95ms r=>250ms r=650ms Divalent Cation Ni+ > ca2+ ca2+ > Ni2+ co“ > Ni2+ ca“, < 100uM sensitivity DHPs _ _ 'I' - ngTx _ + _ * _ PTX _ _ _ + Single Channel conductance (p8) 8 13 25 13 * Avian and amphibian L-type C“? are sensitive to Cng (Olivera et al., 1985-; Marqueze et al., 1988). Data for N-, L-, T-channels modified from Tsien et al., 1988. P-channel data compiled from Uinas et al., 1989 and Lin et al., 1990. 1 1 to act as an L-type antagonist but have separate binding sites from that of the DHPs (Miller, 1987; Glossmann and Striessnig, 1988). It is believed that three drug receptor sites on this one channel are linked to one another via heterotrophic allosteric coupling mechanisms (Hofmann et al.,1987; Kemp and Miller, 1987; Glossman and Striessnig, 1988) In mammalian neurons, N-type channels are blocked by o-conotoxin GVIA (Cng), a peptide purified from the venom of the marine snail Conus geographus (McCleskey et al., 1987; Kasai et al., 1987). Both N- and L-type channels in avian and amphibian preparations are blocked by Cng at sites which appear to be distinct from DHP binding sites (Olivera et al., 1985; Abe et al., 1986; Dooley et al., 1987; Marqueze et al., 1988). Cng lacks potency on the invertebrate nervous system, however (McCleskey et al., 1987). T) channels are DHP- and Cng- insensitive but can be blocked by amiloride and octanol (T sien et al., 1988; 1991). The T-., L-, and N-type of Ca2+ channels can be characterized by their sensitivity to inorganic ions as well. Inorganic heavy metal cations such as zinc, nickel, mercury, lead, copper, and cadmium block all three subtypes of channels (Hagiwara and Byerly, 1981 ; Augustine et al., 1987). There is considerable variation in the sensitivity of Ca2+ channels to these ions, however. For instance, in chick DRG neurons, cadmium blocks N- and L-type channels with a Kd of 10 uM but about ten times more cadmium is required to block T-type channels (Nowycky et al., 1985a; Fox et al., 1987). Conversely, T-type channels are blocked more readily by nickel than either N- or L-channels (T sien et al., 1988). ‘I 2 The peptide, o-Agatoxin and the polyamine, FI'X are potent Ca2+ channel blockers derived from the venom of the funnel-web spider Agelenopsis aperta (Adams et al., 1990; Uinas et al., 1989). Subtypes Ofm -Agatoxins are distinguished as type I, type II, and type III. All except type III toxins block Ca2+ channels in most insect motor nerve terminals and neuronal cell bodies at nanomolar concentrations (Adams et al., 1990; Bindokas et al., 1991). On the other hand, types II and Ill but not I toxins, block Cng-sensitive channels in chick brain synaptosomes (Adams et al., 1990; Venema et al., 1992). Evidence exists that Agatoxins blocks DHP- sensitive Ca2+ currents in rat sensory neurons (Scott et al., 1990; Mintz et al., 1991) and in cardiac muscle as well (Liebowitz et al., 1990; Mintz et al., 1991). P-type Ca2+ channels found in Purkinje cells and in the squid giant synapse (Uinas et al., 1989; Mintz et al., 1992) are blocked by o-Aga-lVa and FI'X respectively (Uinas et al., 1989; Mintz etal., 1992) but not by DHPs or Cng (Uinas et al., 1989; Charlton and Augustine, 1990). FI'X also reduces a Ca“ current observed following injection of total rat brain mRNA into Xenopus oocytes (Lin et al., 1990). This current is insensitive to DHP- agonists, antagonists, and Cng as well (Leonard et al., 1987). These results suggest that the channel type expressed from rat brain mRNA in the oocyte may be similar to the P-type channel. Despite considerable research, the channel type(s) involved in neurotransmitter release are still not clearly defined. Again, this is due to technical limitations of current techniques. However, results of some neurochemical (Miller,1987) and electrophysiological (Atchison and O’Leary, 1987; Hirning et al., 13 1988) studies suggest that neurotransmitter release is modulated by both the L- and N- type channel. At rat sympathetic neurons and neuromuscular junctions, L- type Ca2+ channels can have a modulatory role in neurotransmitter release (Perney et al., 1986; Atchison and O’Leary, 1987) but do not contribute to nerve-evoked release under normal conditions (Perney et al., 1986; Atchison, 1989). In contrast, L-type channels have been implicated in release of substance P from rat sensory neurons (Perney et al., 1986) and chick sensory neurons (Reno at al., 1987; Holz et al., 1988), growth hormone and prolactin release from rat anterior pituitary cells (Stojilkovic et al., 1988) and 3H-norepinephrine release from undifferentiated PC12 cells (Kongsamut and Miller, 1986). N-type channels have been reported to play a dominant role in mediating evoked-release of norepinephrine from rat sympathetic neurons (Perney et al., 1986; Himlng et al., 1988), differentiated PC12 cells (Kongsamut and Miller, 1986) and various areas of the central nervous system (Dooley et al., 1987). Data exist that suggest the P-type Ca2+ channel may also mediate Ca“ entry associated with transmitter release. As discussed previously, Ca2+ channels at the squid giant synapse have pharmacological and functional properties unlike those of T-, L-, or N-type Ca“ channels (Charlton and Augustine, 1990). They are DHP— and Cng-insensitive but partially sensitive to FI'X. Therefore, Ca2+ channels in the squid giant synapse appear similar to the P-type isolated from mammalian Purkinje cells (Uinas et al., 1989; Charlton and Augustine, 1990) and likely mediate release in these preparations. Likewise, Cng and DHPs were unable to block K‘- 1 4 evoked release of 3H-aspartate from rat hippocampal slices or release of ACh from murine motor nerve terminals either (Anderson and Harvey, 1987). Mammalian neuromuscular transmission was effectively abolished by FTX, however. FTX blocked presynaptic Ca2+ currents, nerve-evoked neurotransmitter release and muscle contraction from a murine nerve-muscle preparation (Uchitel at al., 1992). This suggests that channels which are neither N- or L-type but possibly P- are involved in neurotransmitter secretion from some mammalian nerve terminals. To date, there is no evidence indicating a role for T-type channels in neurosecretion. It should be noted that the characteristics attributed to Ca2+ channel subtypes do not always fit neatly into one of the above catagories. Thus, the original classification of Ca2+ channels into T-, L-, N-, and now P- subtypes is likely an oversimplification. It is possible that functional heterogeneity among channel subtypes as well as additional types of channels exist. This is supported by the isolation from rat brain of four different types of cDNA that code for putative Ca2+ channels (Snutch et al., 1990). Characterization of the primary structure suggest that several distinct voltage-gated Caz+ channel genes exist (Snutch at al., 1990; Mori et al., 1991; Snutch er al., 1991). Additionally, species and tissue differences in the type of Ca2+ channel involved in neurotransmitter release are evident (Brush et al., 1987; Nayler, 1988; Maggi et al., 1990). C. Disorders of Neuromuscular Transmission As the above review was meant to indicate, there are many steps in the process of synaptic transmission at the neuromuscular junction. Thus, there are 15 many ways in which the process can be disrupted. Disorders of neuromuscular transmission vary in etiology, frequency and clinical presentation. All produce muscle weakness through involvement of the neuromuscular junction, with a preference for certain muscle groups. These disorders may affect the neuromuscular junction presynaptically (the nerve terminal), postsynaptically (the muscle endplate) or both pro-and post-synaptically. This dissertation will focus on the presynaptic, neuromuscular disorder Lambert-Eaton Myasthenic Syndrome and will attempt to delineate the stage at which normal functioning is altered by a pathological process. 0. Lambert-Eaton Myasthenic Syndrome: General Description 1. Cllnlcal Features The principal symptoms of Lambert-Eaton Myasthenic Syndrome (LEMS) are weakness and easy fatigability of limb muscles, which temporarily improves in response to voluntary exercise. Tendon reflexes are reduced or absent at rest (Eaton and Lambert, 1957; Rocks of al., 1960; Lambert et al., 1961; M86 and McDen'not, 1962; Elmqvist and Lambert, 1968). Patients frequently complain of dry mouth, urinary hesitancy, constipation, sexual impotence, and impaired accomodation of the eye (Rubenstein et al., 1979; Mamdani et al., 1985; Heath et al., 1988; Khurana et al., 1988; Manji et al., 1990). These symptoms are indicative of a cholinergic dysautonomia and therefore demonstrate that the defect is not limited to striated muscle but can affect smooth muscle and glands as well. Drugs 16 that increase transmitter release, such as guanidine, 4-aminopyridine and 3,4- diaminopyridine give patients with LEMS symptomatic relief (Oh and Kim, 1973; Lundh, 1983; Newsom-Davis and Murray, 1984; McEnvoy, et al., 1989). Serious toxic effects, however, limit their use. AChE inhibitors have minimal benefit (Eaton and Lambert, 1957; O’Neill et al., 1988). LEMS occurs in two groups of patients. In both groups, male Caucasians (median age of 54 yrs) predominate (Elmqvist and Lambert, 1968; O’Neill et al., 1988). In one group, all have some type of malignancy, the most common (60- 7096) being small cell carcinoma of the lung (SCC)(Eaton and Lambert, 1957; Lambert et al., 1961; Wise and MacDermot, 1962; Elmqvist and Lambert, 1968; O’Neill et al., 1988). Symptoms of LEMS can precede the diagnosis of $00 by 5 months to 4 years (median 10 months) (O’Neill et al., 1988; Anderson et al.,1990) and sometimes improve when the tumor is either treated or removed (Ingram et al., 1985; Oh, 1989; Chalk et al., 1990). The second group of patients have no detectable malignancy but present with identical clinical signs. LEMS is generally accepted as being a paraneoplastic disease, a nervous system dysfunction in patients with cancer that is not caused by metastases (Vincent et al., 1989; Posner, 1991). Because the disease occurs in the absence of tumor as well as in its presence, it follows that the tumor favors the development of the condition but cannot in itself be the only cause (Lambert and Rocks, 1965; Chad and Recht, 1991). 17 2. Electrophyslology a. Cllnlcal Electromyography is diagnostic and conventionally consists of the recording of muscle action potentials generated following surface stimulation of nerves from living patients (Harvey and Masland, 1941; Eaton and Lambert, 1957). An electromyographic feature that characterizes LEMS is a significant reduction in the amplitude of the compound muscle action potential (CMAP). Low frequencies of nerve stimulation (1 per sec, 1H2) result in a progressive decrease in the amplitude of the CMAP while a marked increase in the response is noted when the nerve is stimulated at higher frequencies (10 to 50 Hz) (figure 1.2). An increase in response amplitude can also be demonstrated after brief periods of voluntary contraction of the muscle (Eaton and Lambert, 1957; Elmqvist and Lambert, 1968; Swift, 1981; Jablecki, 1984). Nerve conduction velocities in LEMS patients are within the range of normal values and remain normal at repetitive rates of stimulation (Eaton and Lambert, 1957; McOuillen and Johns, 1967; Elmqvist and Lambert, 1968). This suggests the defect in transmission is not due to failure of action potential generation or propagation. Direct stimulation of muscle fibers also reveals normal conduction velocities. Repetitive stimulation at increasing frequencies do not produce the typical changes 18 CONTROL LEMS Single Shock- CMAP Stimulation /\ (NERVE) _ A Low frequency repetitive stimuli I I i ii I II I High frequency repetitive stimuli l ..... I . . .. . (> 10 Hz) m | | I figure 1.2. Electromyographic Characteristics of LEMS This is a schematic representation of the EMG characteristics of LEMS. As you can see, there is a significant reduction in the CMAP. At low frequencies of stimulation there is a progressive decrease in the CMAP while a marked increase in the response is noted at higher frequencies of stimulation. 19 seen with indirect stimulation of the muscle via the nerve (McOuillen and Johns, 1967; Elmqvist and Lambert, 1968). Thus, the reduction in CMAP amplitude at low frequencies of stimulation and the increase in amplitude at high frequencies of stimulation do not arise from the muscle itself. b. Experimental To investigate the electrophysiological basis for the electromyographical observations, intracellular microelectrode recordings were made from biopsied intercostal muscle taken from patients with LEMS. The number of ACh-containing quanta released per nerve impulse (quantal content) was markedly reduced (Elmqvist and Lambert, 1968; Lambert and Elmqvist, 1971; Cull Candy et al., 1980). This suggests that in vivo the EPP would fail to achieve the threshold depolarization required for generating an action potential in the muscle membrane. As was seen electromyographically for the CMAP, repetitive stimulation of the nerve at 30 Hz increased quantal content. Spontaneous release of neurotransmitter was also recorded. MEPP amplitudes and frequencies were similar to the MEPPs recorded in normal tissue indicating resting quantal release of ACh was normal. (Elmqvist and Lambert, 1968; Lambert and Elmqvist, 1971). Resting, non-quantal release of ACh was decreased, however (Lang et al., 1984; Molenaar et al., 1987). 3. Blochemlstry Additionally, the sensitivity of LEMS muscle to ACh (Lambert and Elmqvist, 1971; Lindstrom and Lambert, 1978; Cull Candy et al., 1980) as well as the number of ACh receptors present are normal (Lindstrom and Lambert, 1978). Also, the reduced quantal content is not due to a decreased synthesis or storage of neurotransmitter because the ACh and ChAT content of the nerve terminal are normal (Molenaar et al., 1982; Molenaar et al., 1987). Thus, these potential factors which could have contributed to the neuromuscular weakness found in LEMS patients have been ruled out. The evidence so far strongly suggests a selective effect on the mechanism of ACh release. 4. Morphology a. Morphometric analysls Electron microscopic, morphometric analysis of biopsied endplate regions typically consists of determining the nerve terminal area, the number of synaptic vesicles per unit nerve terminal area, the mean synaptic vesicle diameter, the presynaptic membrane length and the postsynaptic membrane length per nerve terminal (Loud, 1962; Engel, 1970; Engel and Santa, 1971; Hesselmans et al., 1992). Engel and Santa (1971) reported that both methylene blue-stained intramuscular nerves and cholinesterase-stained endplates showed no definite abnormalities. Electron microscopy demonstrated that the mean nerve terminal area and the mean synaptic vesicle count per unit area were not significantly 21 different from control values. When the mean synaptic vesicle diameter was calculated for each biopsy, control and LEMS means did not differ significantly either (Engel and Santa, 1971). One abnormality noted, however, was an overdevelopment of the postsynaptic region. Analysis showed a significant increase in the mean area of clefts and folds per nerve terminal and in the mean postsynaptic to presynaptic membrane length ratio (Engel and Santa, 1971). Multiplication of endplate areas is a well known phenomenon in disorders of neuromuscular transmission such as botulinum toxin (BoTx) intoxication (Duchen, 1971), myasthenia gravis (MG; in some cases) (Coérs and Tellerman-Toppet, 1976) and congenital myasthenia with decreases in ACh receptors (Vincent et al., 1981). The increase in contact between muscle and nerve is likely to be compensatory to the deficiency in ACh release (LEMS and BoTx intoxication) or ACh receptors (myasthenia gravis). In contrast, Hesselmans and colleagues (1992) report that both postsynaptic areas and postsynaptic membrane lengths of NMJs are significantly less in LEMS than in controls (Hesselmans et al., 1992). This has also been observed in patients with MG and in the chronic phase of MG’s experimental model as well (Engel and Santa, 1971; Engel and Tsujihata, 1976). A factor which may have influenced the differences between Hesselmans’ and Engel’s studies is the duration of the disease before muscle biopsy was performed. In Hesselmans” study, the onset of symptoms was at most 1.5 years before biopsy. In the study by Engel and Santa (1971) duration of symptoms ranged from 3 to 11 years. Since newly-created 22 NMJs have small postsynaptic regions which in time increase, perhaps only short- terrn effects of ACh release blockade were studied by Hesselmans et al., (1992). b. Ultrastructural analysls Ultrastructural changes in the motor nerve terminal associated with LEMS were first reported by Fukunaga and coworkers (Fukunaga et al., 1982; 1983). Measurements of the density of the active zones and large intramembraneous active zone particles (AZPs) were made in freeze-fracture preparations of biopsied human intercostal muscle. Their results indicated that both the active zones and the AZPs were strikingly scarce and disorganized (Fukunaga et al., 1982). Prejunctional membranes taken from patients with LEMS had fewer active zones/u m2; fewer number of active zone particles /u m2; less particles/active zone; and a 3.6 fold increase in the number of large particles aggregated into clusters/u m2. All other membrane particles studied appeared normal. Since the average decrease in the AZPs per unit of membrane area was comparable to the average decrease in quantal content of the EPP in LEMS, Fukunaga and colleagues concluded that a reduction in AZPs in LEMS accounted for the diminished probability of ACh release. (Fukanaga er al., 1983). Active zones are topographically related to the sites of synaptic vesicle exocytosis, and thus are believed to be the sites at which neurotransmitter release occurs (Heuser et al., 1974; 1979). Thus, electron micrographic findings in LEMS are consistent with a presynaptic disorder of neuromuscular transmission as postulated from the results of microelectrode studies. 23 E. Lambert-5am Myasthenic Syndrome: Autoimmunity 1. Cllnlcal and experimental evidence Compelling evidence has indicated that the etiology and pathogenesis of LEMS is autoimmune. Clinically, beneficial effects of plasma exchange (Newsom- Davis and Murray, 1984) and immunosuppressive drugs (Streib and Rothner, 1981; Newsom-Davis and Murray, 1984) have been demonstrated. Furthermore, LEMS has frequently been shown to accompany one or more autoimmune disorders (Gutmann et al., 1972; Lennon et al., 1982; Paris at al., 1990). The autoimmune basis and involvement of a humoral factor was demonstrated with certainty when the electrophysiological features of the disorder were transferred passively to mice by injection of humoral substances from affected patients (Denys et al., 1972; Lang et al., 1981; Newsom-Davis et al., 1982; Lang et al., 1983a; Kim, 1985; Kim, 1986; Lambert and Lennon, 1988). Plasma (Newsom-Davis et al., 1982; Kim, 1985), crude immunoglobulin fractions (Lang et al., 1983a; Kim, 1986; Lambert and Lennon, 1988) and purified lgG (Denys et al., 1972; Kim, 1986) are all effective in transferring the disease with the production of characteristic in vitro electrOphysiologic abnormalities. Mice treated with lgG from patients exhibiting LEMS also had a significantly lower number of active zones and AZPs (Fukanaga et al., 1983; Nagel et al., 1988). Furthermore, immunolocalization of LEMS lgG was demonstrated in the region of presynaptic active zones of mice (Fukuoka et al., 1987). It has been proposed that the AZPs correspond to voltage-sensitive calcium 24 channels of the presynaptic membrane (Pumplin et al., 1981 ; Robitaille et al., 1990; Tarelli et al., 1991; Cohen et al., 1991). Indirect evidence to support this assumption is as follows: (1) presynaptic voltage clamp studies in the squid giant synapse demonstrated a very short latency between Ca2+ entry and postsynaptic response, suggesting that Ca2+ channels are located at the site of vesicle accumulation and neurotransmitter release (Uinas et al., 1976); (2) synaptic vesicle exocytosis captured by quick freezing occurs near the AZPs (Heuser et al., 1979); (3) AZPs continue to specify sites of exocytosis even when dispersed by a low Cay-medium (Ceccarelli et al., 1979); and (4) in the squid synapse, the maximum presynaptic Ca2+ current appears to be related to the number of AZPs (Pumplin et al., 1981). Direct evidence for this contention is the presynaptic localization of Cng- sensitive Ca2+ channels at the frog neuromuscular junction (Robitaille et al., 1990; Tarelli et al., 1991; Cohen et al., 1991). Cng- immunoreactivity revealed a pattern composed of regularly spaced bars running parallel to each other and perpendicularly to the major axis of the nerve terminal. This array corresponds to the ordered arrangement of active zones in this species. A correspondence was also found between the mean number of active zones and fluorescent bars present within the same length of axolemma (T arelli et al., 1991). If AZPs correspond to voltage—dependent Ca2+ channels, the reduced neurotransmitter release by nerve impulses in LEMS patients could be related to reduced entry of calcium into the nerve terminal. 25 2. The Antigenic Target a. Calclum channel hypothesls Autoantibodies that interfere with Ca2*-dependent neurotransmitter release may do so by three distinct mechanisms. The LEMS autoantibody could act on (1) Ca2+ entry through voltage-gated Ca2+ channels, (2) intraterminal Ca2+ metabolism or (3) the exocytotic release process. Results of several studies indicate that LEMS autoantibodies interact with voltage-dependent Ca2+ channels (Roberts et al., 1985; De Azipurua et al., 1988; Kim and Neher, 1988; Peers er al., 1990). Other studies have rendered the other possibilities unlikely (Lang et al., 1983; Kim, 1985; 1986; 1987; Wray et al., 1987; Lambert and Lennon, 1988). Except for the well-understood requirement of Ca2+ ions for prejunctional ACh release (Katz and Miledi, 1965; Dodge and Rahamimoff, 1969), specific molecular events leading to exocytosis are largely unknown. Because of this, it is difficult, if not impossible, to evaluate a possible dysfunction in the release mechanism itself. However, indirect effects on release mechanisms can be studied with pharmacological manipulation. Nerve evoked release of neurotransmitter from passively transferred mice was studied over a range of [Ca2+], concentrations (Kim, 1987; Wray et al., 1987). NMJs from LEMS lgG-treated mice showed the characteristic decrease in quantal content at both low and physiological Caz+ concentrations when compared to control lgG-treated mice. However, at unphysiologically high Ca2+ concentrations the number of quanta released by news stimulation in LEMS-treated NMJs reached the same plateau value as did 26 control-treated NMJs. This suggests that the number of available quanta was not different in passively transferred LEMS and the exocytotic release process itself was not impaired (Kim, 1987; Wray et al., 1987). The second possibility that intracellular mechanisms governing Ca2+ homeostasis may operate abnormally in LEMS is also not supported. An abnormality of this type would result in a decline in both nerve~evoked and spontaneous release of neurotransmitter. In human LEMS as well as passively transferred murine LEMS, no appreciable reduction in either MEPP frequency or amplitude is seen (Lang et al., 1983; Kim, 1985; 1986; Lennon and Lambert, 1988). With the exclusion of the latter two possibilities, and with the evidence presented thus far, the “calcium channel hypothesis” appears to be the most plausible explanation for the pathophysiology of LEMS. Numerous studies have helped to clarify the neuromuscular deficit associated with LEMS. Incubation of rat cortical synaptosomes with whole serum or high molecular weight serum fractions (> 10,000 daltons) from patients with LEMS, for as little as three minutes, decreased the amount of ijACh released during depolarization (Meyer et al., 1986). Although this study did not measure Ca2+ channel activity, it was important in that it indicated that acute exposure of synaptosomes to serum or serum fractions taken from LEMS patients could elicit responses qualitatively similar to those obtained in chronic passive transfer experiments. Also, antigens presumed to reside on human and rodent nerve terminals appear to share a common epitope recognized by the LEMS autoantibody. Likewise, Login et al., 27 (1987) exposed rat anterior pituitary cells to immunoglobulin isolated from a LEMS patient, and found that release of growth hormone and prolactin was significantly reduced by LEMS lgG compared with control lgG (Login et al., 1987). Again, no assessment of ion channel activity was made but secretion from these cells is Ca2+-dependent (Stojilkovic et al., 1988). Roberts et al., (1985) cultured human small cell carcinoma cells (800) in the presence and absence of LEMS IgG for seven days. Following the incubation period, 45Ca2+ uptake induced by 30 sec of K+ depolarization was measured. The cells treated with LEMS lgG exhibited a significant decrease in stimulated uptake compared with untreated cells or with cells grown in control lgG (Roberts et al., 1985). These studies were confirmed by De Azipurua et al., (1988a) and Lang et al., (1989b) who in addition demonstrated that PC12 cells were similarly affected (Lang et al., 1989a; 1989b). Kim and Neher (1988) applied lgG from LEMS patients to isolated bovine chromaffin cells from the adrenal medulla. Electrophysiological recordings indicated a 40% reduction in the current carried through calcium channels. Sustained inward Caz+ currents in undifferentiated mouse neuroblastoma x rat glioma hybrid cells (NG108-15) were reduced by LEMS lgG also (Peers et al., 1987). Taken altogether, these various results provide strong evidence for membrane calcium channels as a target site for the autoantibody in LEMS. However, recent evidence suggests that synaptotagmin may also be a LEMS antigen (Leveque et al., 1992). A 58 kDa antigen, identified as 28 synaptotagmin by a partial amino acid sequence, was detected by probing Western blots of partially purified calcium channels from rat brain with LEMS plasma and lgG. Synaptotagmin is a synaptic vesicle protein believed to be involved in docking synaptic vesicles to the plasma membrane at the active zone (Augustine, 1987; Leveque et al., 1992; Brose et al., 1992). As this protein was recognized in what was thought to be partially purified Ca2+ channels, Leveque and colleaques (1992) believe that synaptotagmin not only binds to the plasma a” channel as well. As such, membrane but is tightly associated with the C alterations in synaptotagmin function could indirectly alter Ca2+ channel activity (Leveque et al., 1992). However, several deficiencies in this study are evident. first, synaptotagmin is a putative transmembrane protein with the N-terminus oriented toward the interior of synaptic vesicles and the C-terminus at least partially inserted into the lipid bilayer (Perin et al., 1990). In order for the circulating antibody to bind, one must presume that this protein is exposed at the surface of the nerve terminal following exocytosis. Secondly, synaptotagmin and calcium channels may have a common epitope which can be recognized via SDS-PAGE followed by Western blot analysis but which is normally concealed. finally, the possibility that LEMS sera contains antibodies that only recognize calcium channel associated polypeptides in their undenatured form can not be excluded. 29 b. Calclum channel subtype affected by LEMS lgG The finding that LEMS lgG inhibits Caz+ entry in motor nerve terminals (Lang et al., 1987), 806 (Roberts et al., 1985; De Azipurua et al., 1988a; Lang et al., 1989a; 1989b), chromaffin (Kim and Neher, 1988), anterior pituitary (Login et al., 1987), NG108-15 (Peers et al., 1987) and PC12 cell lines (Lang et al., 1989a; 1989b) indicates that the antibodies recognize an epitope that is present on voltage-dependent Ca2+ channels from several sources. The use of these cell lines have led to assumptions concerning the subtype of Ca“ channel putatively affected by this antibody. LEMS lgG was shown to reduce Ca2+ currents with characteristics of L-type channels in NG108-15 and chromaffin cells (Kim and Neher, 1988; Peers et al., 1987). Secretion from anterior pituitary and undifferentiated PC12 cells is believed to be mediated by L-type channels also (Kongsamut and Miller, 1986; Login et al., 1987; Stoljilkovic et al., 1989). Data from studies utilizing 800 cell lines (Roberts et al., 1985; De Azipurua et al., 1988a; 1988b) suggest an effect on both L- and N-type channels, respectively. Finally, LEMS autoantibodies have been shown to immunoprecipitate 1’i‘I-Cng labeled voltage-gated Ca2+ channels (N-type) from a human neuronal cell line (Sher et al., 1989). The results from these studies would suggest that the L- and N- type Ca2+ channel are putative antigenic targets for the LEMS autoantibody. Although these studies have proven useful in determining the antigenic target in the cell some, the identity of Ca2+ channels in the nerve terminal remains unclear. What is clear, is 30 the types of Ca2+ channels at mammalian nerve terminals do not appear to be either N- or L-type and are possibly of the P-type (Anderson and Harvey, 1987; Sano et al., 1987; Charlton and Augustine, 1990; Uchitel et al., 1992). 3. Pathophysiologlcal Mechanisms The exact mechanism by which LEMS lgG exerts its pathological effects has not been fully established. In bovine adrenal chromaffin cells treated with LEMS lgG, kinetic and single channel studies indicate that the pathogenic autoantibodies neither modify the kinetics of Caz+ channel activation and inactivation nor single channel open time nor conductance (Kim and Neher, 1988). Also, compared with control serum, pooled LEMS sera from 5 patients reduced the binding of the Ca2+ channel antagonist 125M -conotoxin (Cng) by 22% following overnight incubation with the human neuronal cell line, lMR-32 (Sher et al., 1989). Thus, it appears as if the antibody-induced dysfunction of Ca“ channels occurs in an “all or none“ fashion. Each channel is either subject to a complete functional block or no blockade at all. Therefore, the pathogenic action of the LEMS autoantibody is characterized by a reduction in the number of functlonal Ca2+ channels. Functional alterations could be achieved by several mechanisms and may act in concert. a. Dlrect Inhlbltlon Bound, pathogenic autoantibodies could reduce the number of functional Ca2+ channels by direct inhibition. This does not appear to be the case in some 31 systems, however. The application of LEMS plasma to an in vitro organ culture for up to three hr had no effect on nerve-evoked quantal release (Kim et al., 1988). When mice were studied after various periods of IgG treatment, the t“ for the reduction in quantal content was at least 4-6 hr (Lambert and Lennon, 1988), or longer (Prior et al., 1985). However, It just may be that more than three hr of incubation is needed in order to gain access to, and inhibit the function of a2+ channels in some systems. For example, only a two hr prejunctional C incubation of chromaffin cells with LEMS antibodies resulted in significant reductions in whole cell Ca2+ currents (Kim and Neher, 1988) while incubation of synaptosomes with LEMS sera for as little as three min decreased the amount of [3H]-ACh released during depolarization (Meyer et al., 1986). b. Antlgenlc modulation In view of the morphologic alterations depicting the scarcity of AZPs, it would appear that the LEMS autoantibody eventually destroys or eliminates Ca2+ channels. Antibody-induced antigenic modulation could result in loss of channels by initiating accelerated channel degradation. Ca2+ channel crosslinkage by LEMS lgG could be an initial process required for final channel elimination. This process has been observed by Nagel et al., (1988) and Lang et al., (1987). These investigators observed that divalent F(ab)’2 (antigen-binding) fragments of LEMS autoantibodies which retain the antigen binding capacity of normal lgG effectively transferred the disease to mice, whereas monovalent Fab fragments failed to do so. This phenomenon was verified by measurements of AZP density (Nagel et al., 32 1988) and quantal content of the EPP (Lang et al., 1987). In the terminals treated with LEMS IgG or LEMS F(ab)’2, quantal content was reduced; monovalent LEMS Fab had no significant effect (Lang et al., 1987). There was also a significant decrease in the density of active zones and AZPs. In the AZPs remaining, interparticle distance was reduced (Nagel et al., 1988). This was considered to be an early event associated with disorganization and eventual destruction of active zones and AZPs (Fukuoka et al., 1987). c. Complement-mediated destruction Alternatively, loss of channels could be due to complement-mediated channel destruction. Activation of complement plays a central role in mediating postsynaptic membrane damage in animal models of the autoimmune neuromuscular disorder, myasthenia gravis (MG) as well as in the human disease state (Engel et al., 1977; Sahashi et al., 1978). It seems possible that complement activation could be important in LEMS as well. Complement is a complex series of proteins which form a triggered enzyme system in the plasma. Figure 1.3 provides a summary of the complement components and their interactions. Some of the complement components are designated by the letter 'C" followed by a number. The complement system may be activated by two major pathways (classical and alternative) that proceed in a sequential fashion. Activation of individual components of complement depends upon activation of a prior component or components in the sequence. Fragments of native complement components are either released into the fluid phase or bind (I) Classical Pathway Activation Signal j _ C4 cr——> c1 / >1 iP'" \ C2 products 1 CAT); (03 convertasc) 1 Alternative 3—9 C3-> 03a Pathway D Feedback Loop Amplification IP— C3b—-) C4b2e3b I CS -> 05a Membrane Attack Complex 9 C5b678 9n figure 1.3. Complement. Scheme of the reaction mechanisms following activation of the complement cascade. For detailed explanation please see text (pp. 32-35). 34 to other complement components or suitable surfaces such as immune complexes and cell membranes (Law and Levine, 1977; Capel et al., 1978; Takata et al., 1984). The classical pathway of complement activation is typically initiated by the union of antibody and antigen. The antigen can be either soluble or a constituent of a cellular or extracellular surface. Whether an antigen-antibody reaction activates complement depends upon the antibody molecule. Most immunoglobulins of the lgM or IgG (with the exception of lgG4) subclass activate the classical pathway (Loos, 1987). Antigen binds to Fab (antigen-binding) regions of immunoglobulins. Together, these molecules undergo a conformational change that permits binding and activation of the first component of complement, C1. Activated C1 cleaves C4 and C2, by limited proteolytic reactions. Cleavage of C4 results in the formation of a small peptide, 04a and a large fragment, C4b, which can bind either to the immune (antigen-antibody) complex or to the surface of the antigen. C2 is cleaved into a small peptide, 02b and a large fragment, 02a, which binds to C4b. This complex, C4b2a, (03 convertase) possesses new enzymatic activity that is capable of cleaving the third component of complement, 03. Proteolytlc cleavage of 03 yields a small fragment, 03a and a larger fl'Elgl'nent, 03b, which possesses the ability to bind to immune complexes, as well as to a wide variety of surfaces (Law and Levine, 1977; Capel et al., 1978; Takata 9‘ 31- . 1984). Once bound, 03b imparts a new specificity to C4b2a, enabling this 35 enzyme to function as a CS convertase. Bound 03b is also recognized by phagocytic cells (Hostetter etal., 1984), and functions as an important intermediate in the activation of the alternative pathway (Goetze, 1987). Cleavage of CS by CS convertases, constitutes the final proteolytic reaction in the classical pathway. A small fragment, 05a, is released into the fluid phase, but the larger fragment, CSb, binds to immune complexes and to a variety of surfaces. Whether bound or in the fluid phase, CSb forms a stable macromolecular complex with the complement components C6, C7, and C8. This complex induces orientation between two or more molecules of the terminal component 09. This leads to unfolding of 09 and insertion into the lipid bilayer and polymerization to a ”membrane attack complex" (MAC). MACs form transmembrane channels that permit bidirectional flow of ions and ultimately macromolecules producing cellular injury and/or lysis. (Porter and Reid, 1978; Bhakdl and Tranum-Jensen, 1978). The alternative pathway (APC) is a non-specific, natural defense system against pathogenic microorganisms (Joiner et al., 1984; Goetz et al., 1987). In addition, it is an amplification system for increasing 03b depositon and 03/05 convertase activity after complement activation through the classical pathway (Joiner et al., 1984). Once activated, the biological sequence of events of the APC does not differ from that of the classical pathway. Attempts to clarify the participation of complement in the pathogenesis of LEMS have yielded negative results. Passive transfer was attempted in mice which had received cobra venom factor (CVF) (Lennon and Lambert, 1988) and in mice 36 congenitally deficient in CS (Prior et al., 1985; Lennon and Lambert, 1988). CVF is a potent complement activator, which interacts with Factor B of the alternative pathway to form a stable C3 convertase (Grief: and Mi'iller-Eberhard, 1971). This leads to the consumption of plasma 03 and the later components of the complement cascade. The defect in neuromuscular transmission in mice receiving LEMS lgG was evident in mice of normal and deficient complement status (Prior et al., 1985; Lambert and Lennon, 1988). These results suggest that the effects of LEMS lgG may not be due to complement activation. This unaltered response in the absence of complement activity does not, however, mean that it is not involved when present. Most of the effects of lgG or the alternative pathway activators on cells are greatly enhanced in the presence of complement (Rother, 1987). F. The Immune Response As stated previously, the pathogenic alterations in neuromuscular transmission in LEMS are caused by an autoantibody circulating in the plasma of patients with the disorder. Alterations in the regulation of the normal immune response could lead to recognition of sell-components as foreign inducing autoimmunity. The next section will discuss the normal immune response with emphasis on mechanisms which disrupt regulatory pathways leading to faulty self- non-self recognition. Determinations of the mechanisms of autoimmunity in concert with determination of the antigenic target could have important therapeutic implications for LEMS patients or any patient with autoimmune disease. 37 1. Cellular Interactions In the Immune system The immune system can be considered simplistically the body's defense against foreign invaders. An intricate series of cellular events must occur in order for the body to mount an effective immune response. Typically, an immune response is generated by the coordinated interaction of lymphocytes and antigen- presenting cells (Fig 1.4)(see Miller et al., 1971). All of these cells derive from the bone marrow, from a pluripotent stem cell (Roit, 1988a). One class of lymphocytes consists of B lymphocytes, so called because maturation of these cells occurs in the bone marrow. B lymphocytes are the only cells capable of producing antibodies (Nossal et al., 1968; Jacobson et al., 1970). The second major class of lymphocytes are T lymphocytes, which arise in the bone marrow and then migrate to and mature in the thymus. It is in the thymus that the T lymphocyte undergoes a selection event in which 'leaming" takes place for recognition of self-encoding molecules known as the major histocompatibility complex (MHC). T lymphocytes are further subdivided into functionally distinct populations, the helper (T h) cell and the suppressor/cytotoxic T (T .) cell. T" cells promote immune responses to antigen whereas T, cells function to suppress such a response (Schwartz, 1987; Roit, 19883). Cells responsible for antigen presentation are usually of the phagocytic monocyte-macrophage lineage (Adler et al., 1966; Unanue and Askonas, 1968; Unanue and Cerottini, 1989). Other antigen-presenting cells include the Langerhans cells in the skin, dendritic cells of the lymph nodes and spleen, and * * antigen macrophage .cytokines ACtivated *— T, cell Clonal Expansion lCYIOKIDOS of T, cells B cell Plasma Cell Clone\’Y Y antibodies YYY Figure 1.4. Immune Response The immune response is initiated when antigen is processed and presented to a Th lymphocyte by a macrOphage. The Th cell becomes activated and releases a variety of soluble factors (cytokines) which act on both B and T lymphocytes to augment the immune response. Once triggered by antigen and Th cell cytokines, B cells proliferate forming a large clone of plasma cells which make antibody. 39 the interdigitating follicular cells of the thymus (Dalton and Bennett, 1992). Normally, the interaction between the Th lymphocyte and the antigen- presenting cell is the first event in an immune response. Helper T cells have an unusual specificity for antigens. They recognize only peptide antigens that are attached to MHC molecules expressed on the surfaces of antigen-presenting cells (Lamb and Feldman, 1984). The T cell does not recognize native antigen, however. Instead, the antigen-presenting cell goes through a series of events known as antigen processing, whereby native antigens are internalized, partially degraded and immunogenic fragments are returned to the cell surface (Unanue and Askonas, 1968; Unanue and Cerottini, 1989). In addition, the antigen-presenting cell releases protein hormones called cytokines to co-stimulate the Th cell. In response to stimulation, Th cells also secrete cytokines, whose function is to promote the proliferation and differentiation of the T cells themselves as well as other cells including 8 cells (Schwartz, 1987). Once triggered by antigen and Th cell cytokines, B cells undergo successive waves of proliferation to build up a large clone of plasma cells which make antibody that is antigen-specific (Jacobson et al., 1970). In addition, 8 cells are able to recognize some antigens without any T cell help. Certain antigens, such as bacterial lipopolysaccharides, have the ability to activate a proportion of the B cell population by their inherent mitogenic potential (Roit, 1988a). B cells have the ability to produce 2:10a different antibodies (Max, 1989). However, each B cell can produce only a single heavy and a single light chain, which interact to make an 40 antibody with a unique specificity (T onegawa, 1983). 2. Self- non-self recognltlon The key to effective immunity is the elimination of pathogens without concomitant host injury. The finely tuned interaction among lymphoid cells, cytokine molecules and foreign influences determines immune reactivity. This finely tuned modulation of the immune system is crucial since excessive activation can lead to allergies and autoimmune diseases and oversuppression may promote infectious diseases and cancer (Homo-Delarche et al., 1991). Curiously enough, autoreactive B and T cells are present normally (Dighiero et al., 1987; Rosenkrantz et al., 1987). The body has homeostatic mechanisms to prevent them from being triggered under normal circumstances. Immunologic tolerance to self in T and B lymphocytes can be induced by two basic mechanisms: (1) deletion, through cell death; and (2) functional inactivation (energy) without physical deletion. a. Clonal Deletion The clonal deletion theory of self-tolerance was first proposed by Bumet (1959) to occur during fetal or early postnatal life. Autoreactive clones of lymphocytes (both B and T) were believed to be eliminated during the stage in development immediately after the lymphocyte expresses its antigen receptor and before it becomes competent to respond positively to antigenic stimulation (Lederberg, 1959). In this way autoreactive clones are filtered out. The original theory was modified by Allison et al., (1971) and Weigle (1971) 41 who independently suggested that self-tolerance depends upon intrathymic elimination of self-reactive T lymphocytes while B cell responsiveness to self antigens remained intact. In this way, autoreactive T cells are unable to collaborate with B cells to generate autoantibodies. The models put forth to explain this were based on receptor affinity (for review see Sprent et al., 1988). T cells in the thymus which bind with high affinity to MHC molecules on the surface of intrathymic macrophage cells are eliminated. Those capable of reacting with MHC plus antigen but which only bind with poor affinity to self molecules alone, survive. The idea of clonal deletion has been now been verified for both T and B cells using transgenic mice (Kisielow et al., 1988; She et al., 1988; Nemazee and Bi'Jirki, 1989). However, clonal deletion may play a relatively minor role in B cell tolerance (Nemazee and Bt'irki, 1989; Goodnow et al., 1990). Because, autoreactive T and B cells can be recovered from normal humans and animals (Dighiero et al., 1987; Rosenkrantz et al., 1987), the process of clonal deletion must be incomplete. Other mechanisms must act to limit the potentially catastrophic immune response against self. b. Clonal Anergy I. Induced by Suppression Removal of anti-self effector lymphocytes may be attributable to elimination in some cases but may be due to the presence of peripheral blood regulatory lymphocytes in others. Self-tolerance based on the induction of T, cells has been postulated by Rose (1988). In this model, cells are functionally inactivated without 42 being physically deleted. Self-antigens are considered normally to induce mostly suppressor (rather than helper) T cells with the ability of suppressing that particular immune response (for review see Cruse and Lewis, 1992). This model allows for the existence of a reservoir of cells potentially capable of autoimmune behavior. Clonal energy has been demonstrated in the B cell repertoire as well (Weigle, 1973; Metcalf and Klinman, 1976; Nossal, 1989; Goodnow et al., 1990). Studies in tissue culture demonstrate that B lymphocytes exposed to antigens in the absence of T" lymphocytes can be induced into a state of nonresponsiveness. Subsequent challenge of the B cells with antigen in the presence of Th cells produces no response (Metcalf and Klinman, 1976). ii. Induced by Signal Failure Bretscher and Cohn (1970) suggested the need for two signals to activate a lymphocyte: one, antigen receptor occupation; the other, a co-stimulatory signal delivered by a second cell. In the case of B lymphocytes, the second signal comes from activated Th cells. In the case of Th lymphocytes, the second signal comes from antigen-presenting cells. Tolerance induction in this model was postulated to be delivery of signal one in the absence of signal two (Bretscher and Cohn, 1970). Inactivation of antigen-specific lymphocytes as a consequence of abortive antigen presentation may indeed play a role in self-tolerance. In healthy organisms, self-antigens may be presented without accessory signals (cytokines, MHC, adhesion molecules) necessary to induce clonal proliferation and differentiation. This would leave the T cell unchanged and possibly unable to respond later to an 43 adequate stimulus (Bretscher and Cohn, 1970; Lamb and Feldman, 1984; Parker and Eynon, 1991). In this way, immunocompetent, autoreactive T cells could be inactivated. Similierly, B lymphocytes whose receptors are cross-linked by certain multivalent antigens alone are rendered non-responsive, whereas simultaneous costimulatory signals by T cell cytokines lead to activation and differentiation (Nossal, 1983). 3. Breakdown of self-tolerance The state of self-tolerance may be an amalgam of many processes including (but not limited to) clonal deletion and specific suppressor cell networks. It lymphocyte/monocyte action is not limited by protective regulatory pathways then these cells will either directly or indirectly attack innocent bystander cells, disrupting tissue structure and leading to autoimmune disease. Direct attacks include the release of biologically active agents (enzymes, peroxides, cytotoxic cytokines etc.) from monocytes or cytotoxic lymphokines from T lymphocytes. Indirect attacks may be mediated through antibodies released from B cells and/or activation of the complement system (Grossman et al., 1991). Autoimmune diseases are chronic and in some cases severe. They are clearly multifactorial in origin with genetic, immune, endocrine and environmental elements contributing to their development (Rose, 1988; 1989; McGregor, 1990; Dalton and Bennett, 1992). A variety of factors have been identified that are both indirectly and directly responsible for the inappropriate activation of immune 44 effector cells (Bottazzo ef al., 1983; Cruse and Lewis, 1987; Sarvetnick et al., 1988; Sinhe et al., 1990; Grossman at al., 1991; Kronenberg, 1991). figure 1.5 summarizes schematically some of the initiators of the pathogenesis of autoimmune disease. These factors are often interrelated and can act in concert to give rise to pathogenic autoimmunity. a. Decreased Suppression Reversal of energy might form part of the spectrum of autoimmunity. Antigen specific T. cells responsible for blocking the effects of autoreactive T" cells could be rendered ineffecient and energy would be abrogated by allowing signals to reach the B lymphocyte. Distortion in the helper/suppressor cell ratio could produce an advantage to helper components (Krosgrud and Perkins, 1977; Tyan, 1977). Defects of T, cells have been described in systemic lupus erythemetosus, rheumatoid arthritis, MG, and autoimmune hemolytic anemia (Shoenfield and lsenberg, 1989a). In addition, aging alone has profound effects on thymic function and may alter the balance of helper/suppressor cells (Serge and Serge, 1987). b. Modlflcatlon of an autoantigen Self antigens may be altered by a variety of chemicals. Patients receiving chronic administration of the antihypertensive drug, a-methlydope, have a higher incidence of developing autoimmune hemolytic anemia. This may be attributable to a modification of the erythrocyte surface in such a way as to provide stimulation of B cells (Rose, 1988; Holt, 1988b; Shoenfeld and lsenberg, 1989b). The addition of a foreign determinant to native proteins may provide a mechanism for T cell 45 Decreased Self-antigen suppression + Altered self- In creased MHC l antlgen Molecular e elf-reactlve Seq ues gel-ed —> <-— rnImIcry Th cell antigen I I cytokines l IlgG autoantibodlesl figure 1.5. Pathogenesis of autoimmune disease. Autoimmunity may be initiated by a variety of mechanisms involving alterations in immune regulation, changes in auto-antigens or both. These factors are often interrelated and may act in concert to give rise to pathogenic autoimmunity. recognition. MG has been described in some patients on penicillamine (Bucknell et al., 1975; Czlonkowska, 1977) and a model of penicillamine-induced MG has been produced in animals (Bever et al., 1982; Bever and Asofsky, 1991). It is thought that penicillamine either causes a generalized alteration in immunoregulation (Bluestone and Goldberg, 1973; Epstein et al., 1979) and/or penicillamine becomes covalently linked to host proteins causing antigenic changes which lead to an immune response against the host protein, in this case, acetylcholine receptors (Dawkins et al., 1981; Bever et al., 1982). c. Self antigen plus aberrant self-encoding molecules In many autoimmune diseases there is an overexpression of MHC molecules on the target tissue (Shoenfield and lsenberg, 1989b). Because MHC molecules are known to be important in antigen-presentation, these observations led to the proposal that abnormal, excessive or ectopic self-encoding molecule expression leads to the presentation of autoantigen to autoreactive Th cells (Bottazzo et al., 1983). The basic principles of this hypothesis have been verified experimentally using the thyroid as a model. It was shown that interferon-y induced MHC expression in thyrocytes; that these cells could act as antigen-presenting cells; and that, in autoimmune disease, many of the infiltrating T cells specifically recognize these thyrocytes (Londei et al., 1984; 1985). 47 U. Release OI SOQUBSIBTBU antigens Anatomic alterations may lead to the exposure of antigens that are normally sequestered and concealed from the immune system. Because of this sequestration, individuals may not be immunologically tolerant to such antigens. Therefore, if such self antigens are released and interact with immunocompetent lymphocytes, specific immune responses may develop (Kronenberg, 1991). Examples of anatomically- sequestered antigens are intraocular proteins and sperm. 9. Cross-reactivity r. antigenic mimicry Many examples are known in which potential autoantigenic determinants are present on an exogenous cross-reacting antigen which provokes autoantibody formation (Shoenfield and lsenberg, 1989b). An antiserum raised against an antigen can cross-react with a partially related antigen which bears one or more similar determinants. The more determinants that are similar, the greater the cross- reactivity (Roit, 1988b). Infection by certain microorganisms may result in antigenic mimicry. Immune reactivity against microbial antigens may encompass the stimulation of Th lymphocytes capable of reacting with self. Alternatively, structural mimicry between selected antigens of microorganisms and mammalian tissue may result in cross- reactivity of the antibody response (Rose, 1989). Well known examples of the association of autoimmune disease with infection are the role of streptococcal or "_I l ' —1 48 infection with the development of rheumatic fever (Cruse and Lewis, 1987; Rolf, 1988b) and the onset of diabetes mellitus type 1 following mumps (Coxsackie) virus infection (King et al., 1983). Ii. aging Studies have also demonstrated a progressive decline in the immune response associated with aging. For example, there are several changes in effector functions of the T cell compartment (Krosgrud and Perkins, 1977; Tyan, 1977) and there is a consistent decrease in the amount and in the average affinity of antibody produced by B cells in aged animals (Price and Makinodan, 1972; Nordin and Makinodan, 1974; Callard et al., 1977). Paradoxically, these changes are accompanied by an increased incidence of autoantibodies (Goidl, 1987). Thus, the aging immune system is poorly regulated. The decline in antigen responsiveness associated with aging has been characterized by the production of antibody with decreased average affinity and decreased heterogeneity of the antibody affinity (Callard et al., 1977; Klinman, 1981). This decrease in heterogeneity in the aged is seen preferentially as the loss of medium- to high-affinity antibody. Thus, the expressed antibody repertoire is principally composed of low-affinity antibody (Doria et al., 1978; Goidl, 1987). The etiology of the increase in antibodies that recognize self-components could be explained then by inappropriate cross-reactive low-affinity antibodies. 49 G.AutoimmuneLEMSandtheroleofcancer It is now well established that self-reactivity is a common phenomenon, but the conversion from autoimmunity to autoimmune disease requires a variety of additional factors. The mechanisms of autoimmunity are topics for which theories and hypotheses outnumber facts. A major difficulty in defining the mechanisms of human autoimmune disease has been the inability to identify the antigens that initiate the autoimmune response. As a result the specific etiology of most autoimmune diseases, in general, and of LEMS, in particular, is unknown. Muscle weakness is often a feature of systemic malignancy, but is frequently dismissed as a non-specific complaint in a chronically debilitated patient (Gomm et al., 1990). A variety of non-metastatic neuromuscular disorders, including cerebellar degeneration, peripheral neuropathy and myopathy may be associated with malignant neoplasm, in particular with lung cancer (Brain at al., 1951; Hanson etal., 1954; Gomm etal., 1990) . As mentioned previously, an association of LEMS with SCC occurs in 60% of LEMS patients (Eaton and Lambert, 1957; Lambert et al., 1961; Wise and MacDermot, 1962; Elmqvist and Lambert, 1968; O'Neill et al., 1988). Early suggestions were made that 800 tumors, produced biologically active peptides which interferred with neuromuscular transmission (lshikawa et al., 1977). These studies, however, where unable to be duplicated. Further studies of the association of LEMS with neoplasms have proposed that an lgG autoantibody is produced in response to antigens on the surface of SCC tumor cells which cross-react with 50 Ca2+ channel determinants on the presynaptic motor nerve terminal (Newsom- Davis at al., 1982; Lennon et al., 1982; Vincent et al., 1989). In support of this hypothesis, oat (small) cells are thought to be of neuroectodermal origin and have many of the biochemical markers of neurons (Gadzar et al., 1981; Moody et al., 1981; Tapia et al., 1981). Electrophysiological recordings from, and radioactive “Ca2+ flux and Fura-2 measurements in small cells in tissue culture indicate that SCC cells have voltage-dependent Ca2+ channels (McCann et al., 1981; Roberts et al., 1985; De Aizpurua et al., 1988; Sher at al., 1990). Furthermore, 800 cells grown in LEMS lgG containing medium exhibited significantly less K*-stimuleted Ca2+ influx than control lgG-treated cells (Roberts et al., 1985; De Aizpurua et al., 1988). More recently, the immunocytochemicel characteristics of small cell carcinoma with and without associated LEMS was assessed by Morris et al., (1992). In LEMS-associated 800, there were numerous macrophages at the periphery of the tumor and many in the tumor mass. This contrasted with the non- LEMS cases in which there were fewer macrophages both in the periphery and in the tumor. Lymphocyte aggregates were also seen adjacent to the tumor in LEMS patients with fewer seen in non-LEMS tumors. Analysis of the lymphocyte populations using subset-specific monoclonal antibodies revealed that the majority of lymphocytes were B cells. Fewer T cells were seen and there was little difference in proportions of Th/T, cells in either LEMS or non-LEMS tumors. The marked macrophage and lymphocyte infiltration in LEMS-associated SCC raises 51 the possibility that it is indeed the tumor in LEMS that is provoking the immune response (Morris et al., 1992). Clinical evidence for neurological improvement after treatment or excision of the tumor would be consistent with this fact. Specific tumor therapy (resection and/or irradiation) may be followed by improvement or remission of the neurological deficit but this has not been generally noted (Ingram et al., 1984; Ferroir et al., 1989; Chalk et al., 1990; Oh, 1990). Furthermore, interpretation of the data from these studies can be difficult as many of the patients receiving specific tumor therapy are also receiving pharmacological treatment for LEMS. Therefore, it seems likely that the tumor may be the initial antigenic target, but this can not be stated with certainty. The clinical and electrophysiological features of LEMS do not show any consistent differences between those with underlying $00 and those who do not. Thus, the disease which is homogeneous in its clinical expression has more than one triggering factor. In about 60% of the cases, this factor appears to be 800 while in the remainder the factor(s) are unknown. One possibility is that LEMS patients without detectable cancer had a small occult tumor which was successfully eliminated by the immune response Mncent et al., 1989). CHAPTER TWO SYNAPTOSOMES AND SYNAPTOSOMAL ION CHANNELS 52 A. General IntdeJction: The exact identity of the antigenic target of the autoantibody which results in LEMS is unknown. Studies of the association of LEMS with $00 suggest that an antibody is produced in response to antigens on the surface of SCC tumor cells and that these antibodies cross-react with Ca2+ channel determinants on the presynaptic motor nerve terminal (Newsom-Davis et al., 1982; Lennon et al., 1982; Vincent et al., 1989). However, there is reason to believe that Ca2+ channels from other nerve terminals may also be susceptible to the the putative blocking actions of the LEMS autoantibody (Meyer et al., 1986). LEMS immunoglobulin (lg) appears to interact with voltage-dependent Ca2+ channels in a number of model cell systems (Roberts et al., 1985; Login et al., 1987; De Azipurua et al., 1988a, 1988b; Kim and Neher, 1988; Lang et al., 1989a; Peers et al., 1990). Assumptions concerning the subtype of Ca2+ channel affected by the LEMS autoantibody have been made based on the subtypes found in these various cell lines. However, the exact identity of the subtype(s) of Ca2+ channel affected in nerve terminals remains unknown. Again, it is not only possible but probable that characteristics of Ca2+ channels in the terminal differ from those in the cell body (Leonard et al., 1987; Suszkiw et al., 1989; Charlton and Augustine, 1990; Uchitel et al., 1992). Passive transfer of the electrophysiological characteristics of LEMS to rodents by chronic injection of lgG isolated from LEMS patients permits only indirect experimental investigation of the putative site of action of the LEMS autoantibody. This is due to the fact that motor nerve terminal Ca2+ channels are 54 not sufficieme accessible nor amenable to direct recording techniques. Since the low density of nerve terminals in skeletal muscle makes them impraticel to isolate, an alternative model of centrally isolated nerve terminals (synaptosomes) may be useful to examine this problem at a more molecular level. B.Syneptmomes 1. General Descriptlon Synaptosomal suspensions are heterogeneous preparations of intact, isolated nerve terminals. They are prepared from brain homogenates using density gradient centrifugation techniques (Gray and Whittaker, 1962). Synaptosomes may be formed by a process of pinching-off so that the cytoplasm is never at any time exposed to the isolation medium, or conversely by resealing of the fractured membrane after rupture of the axon with brief exposure of cytoplasm to isolation medium (Grey and Whittaker, 1962; Bradford, 1975). Whatever the mechanism, the results are completely sealed cytoplasmic sacs of 1-2 uM in diameter retaining all of the cell organelles visible in the intact terminal in situ, including mitochondria, endoplasmic reticulum and synaptic vesicles (Gray and Whittaker, 1962; Dodd et al., 1981). Also, many metabolic and synthetic pathways remain intact in synaptosomes. These include glycolysis, the Kreb’s cycle and neurotransmitter synthesis (de Belleroche and Bradford, 1973; Bradford, 1975). Synaptosomes maintain transmembrane ion gradients and therefore a membrane potential. By using a potentiometric, fluorescent dye, investigators estimate the presence of a 55 membrane potential in the -55 to -65 mV range (Blaustein and Goldring, 1975; Here and Atchison, 1992). High K’ concentrations and veratrum alkaloids, such as veratridine, depolarize the synaptosomal membrane and induce Na“ and Ca“ influx, K’ efflux, and Ca2”-dependent transmitter release (Bradford et al., 1973; Nachshen and Blaustein, 1980; Tamkun and Catterall, 1981; Draupeau and Blaustein, 1983; Bartschart and Blaustein, 1985; Nachshen, 1985). When synaptosomes are incubated in a glucose-containing medium at 37°C, they remain viable for up to four hours after isolation (Bradford, 1975). Incubation at lower temperatures or maintenance of synaptosomes on ice significantly increases their viability. Synaptosomes have been used widely in receptor binding assays (Turner and Goldin, 1985; Dunn, 1988; Marqueze at al., 1988; Suszkiw et al., 1989), and for studies of nerve terminal channel function (Nachshen and Blaustein, 1979, 1980; Bartschat and Blaustein, 1980; Tamkun and Catterall, 1981) and neurotransmitter synthesis and release (Bradford et al., 1973; Drapeau and Blaustein, 1983; Floor, 1983; Woodward et al., 1988). Results from such experiments have contributed significantly to knowledge of the neurochemical events that occur at the presynaptic nerve terminal. 2. Synaptosomal Ion channels a. Sodium channels Influx of Na+ through voltage-dependent channels is involved in action potential generation and propagation. A variety of toxins from both plants and 56 animals specifically interact with voltage-dependent Na” channels in excitable membranes. Saxitoxin (STX) and tetrodotoxin (TTX) specifically block these channels while batrachatoxin (BTX) and veratridine (VER) persistently activate voltage-sensitive Na+ channels. The latter two toxins will also depolarize excitable cells by increasing Na+ permeability (for review see Catterall, 1980). The use of these neurotoxins as specific pharmacological probes has made it possible to establish biochemical methods to study Na+ channel properties in excitable tissues and in synaptosomes. Synaptosomal Na+ channels were indirectly demonstrated by VER- stimulated 45032" uptake (Blaustein, 1975; Nachshen and Blaustein, 1979), VER- stimulated neurotransmitter release (Blaustein, 1975; Cotman at al., 1976) and VER- and BTX- induced synaptosomal depolarization (Krueger and Blaustein, 1980). Direct evidence for the presence of Na+ channels in synaptosomes was demonstrated by Krueger and Blaustein (1980) and Tamkun and Catterall (1981), who reported VER-stimulated uptake of 22Na+ into synaptosomes. This uptake was TTX-sensitive. The use of these toxins has enabled these investigators to conclude that Na+ channels in synaptosomes retain functional properties that are identical with those in intact neurons (Krueger at al., 1980; Tamkun and Catterall, 1981). b. Potassium channels Potassium channels located in the presynaptic nerve terminal control excitability and neurotransmitter release and thus have the potential to be key regulators of synaptic transmission. Previous work has shown that synaptosomes 57 contain a variety of K“ channels whose function can be studied with “RN effiux methods (Bartschat and Blaustein, 1985a,b). The various channels can be distinguished on the basis of their kinetic and pharmacological characteristics (Bartschat and Blaustein, 1985a,b; Benishen et al., 1988) and correspond to known conductances determined in cell systems by electrophysiological techniques. Rat brain synaptosomes contain at least four distinct types of K” efflux mechanisms: (1) a resting K+ permeability, seen with 5 mM K+ solutions; (2) a fast, inactivating, voltage-dependent K+ channel sensitive to block by tetraethylammonium (TEA), tetrabutylammonium (TBA) and 4-aminopyridine (4-AP); (3) a slowly inactivating, voltage-dependent K+ channel which is less sensitive to TEA, TBA and 4-AP; and (4) a Ca2*-activated K+ channel (Bartschat and Blaustein, 19853). Rb" passes through these various types of channels as well as K+ itself. It is accumulated in the cell by the Na”/K* ATPase and is released under depolarizing conditions (Bartschart and Blaustein, 1985a,b). Since a“Rb“ has a longer half-life (days) than ‘ZK" (hours), it is a more practical choice for assessing K“ channel activity in synaptosomes or intact cells. c. Calclum channels Ca2+ plays an essential role in the control of a wide range of cellular functions. Normally, cells maintain an [Ca2+], in the range of 100 nM while the [Ca2+]. is typically in the low millimolar range. A transient rise in the [Ca2+], can 58 trigger many biological responses, including excitation-contraction coupling and neurotransmitter release. One way in which [Ca2+], can be rapidly increased is by the opening of voltage-sensitive Ca2+ channels in the cell membrane, allowing Ca2+ to flow down its electrochemical gradient. Ca2+ channels can be classified into various subtypes differing dramatically in the mechanism of inward current decay during a maintained depolarization and in their sensitivity to pharrnacologicel agonists and antagonists (see Chapter One). Radiotracer flux studies, using 45Ca2+, have been useful in demonstrating Ca2+ influx into synaptosomes. Voltage-dependent uptake of Caz+ into synaptosomes consists of two distinct phases (Nachshen and Blaustein, 1980; Suszkiw and O’Leary, 1983; Nachshen, 1985). The 'fast" phase of influx occurs during the first second of depolarization. It is sensitive to block by lanthanum (< 1 it M) and nickel (40 u M) and inactivates during the the first 2-5 see of K"-induced depolarization (Nachshen and Blaustein, 1980; Nachshen, 1985). The second or "slow" phase of Ca2+ entry remains active for 20-90 see after depolarization and is insensitive to block by lanthanum (Nachshen and Blaustein, 1980). The “slow” phase of influx is believed to be mediated primarily by a reversed Nai/Ca2+ exchanger with a residual slowly inactivating Ca2+ channel component (Blaustein and Oborn, 1975; Nachshen et al., 1986). Removal of Na+, largely reduces 'slow' influx (Turner and Goldin, 1985; Suszkiw et al., 1986; Shafer and Atchison, 1989). The "fast" phase of influx is presumably mediated by voltage- dependent Ca2+ channels (Nachshen and Blaustein, 1980; Suszkiw and O'Leary, 59 1983). Uptake in this phase is independent of external Na“ (Shafer and Atchison, 1989) and is associated with neurotransmitter release (Drapeau and Blaustein, 1983; fioor, 1983; Suszkiw and O’Leary, 1983; Nachshen, 1985; Leslie et al., 1985). Much like Ca2+ channels in other cell systems, voltage-dependent Ca2+ channels in synaptosomes undergo both voltage- and Ca2+-dependent inactivation (Suszkiw and O’Leary, 1983; Suszkiw et al., 1986; 1989) and are sensitive to block by inorganic, multivalent metal cations (Nachshen, 1984). The most potent blockers are lanthanum, cadmium and lead. The block is rapid in onset and is not attributed to changes in either the membrane or surface potential. Inhibition of Caz+ flux by these cations is relieved by increasing [Ca2+], suggesting that the inhibitory ions compete with Ca2+ and directly block passage of Ca2+ ions through the Ca2+ channel (Nachshen, 1984). The antibiotics neomycin and polymixin B block synaptosomal Ca2+ channels in a manner that is reversed by increasing [Ca2+], also (Atchison et al., 1988). In addition to inorganic metals, the organic metals methylmercury (Atchison et al., 1986; Shafer and Atchison, 1989; Hewett and Atchison, 1992) and ethylmercury (Hewett and Atchison, 1992) block 45Ca2+ influx into synaptosomes in a manner that is irreversible by increasing [Ca2+]. (Atchison et al., 1986; Hewett and Atchison, 1992). It appears that both charge and Iipophilicity are important in determining the blocking characteristics of mercurials, as only positively charged mercurials blocked synaptosomal Ca2+ channels and only organic mercurials do 60 so in a voltage-dependent manner (Hewett and Atchison, 1992). finally, the organic Ca2+ channel blocker, verapamil also decreases 4‘r’Ca2“ influx into synaptosomes (Nachshen and Blaustein, 1979; Akerman and Nicholls, 1981; Norris et al., 1983; Nachshen, 1985; Carvalho et al., 1986). Single-channel recordings of synaptosomal Ca2+ channels in lipid bilayers confirm that lanthanum, cadmium and verapamil are potent blockers of synaptosomal Ca2+ channels (Nelson, 1985). i. Ca2+ channel subtypes Synaptosomes are likely to contain more than one subtype of Ca2+ channel but the types present have yet to be identified clearly. Block of Ca2+ channels by verapamil suggests that L-type channels may be involved in Ca2+ influx. This is a subject of much controversy as DHPs, also L-type antagonists, do not have consistent effects on 4"’Ca2+ influx into synaptosomal preparations. DHP antagonists bind to synaptosomal preparations with high affinity (Boles et al., 1984; Greenberg et al., 1984; Turner and Goldin, 1985; Wei and Chang, 1985; Caivalho et al., 1986; Suszkiw et al., 1986; Dunn, 1988). However, a number of investigators have found evidence for both DHP-sensitive and -insensitive pathways mediating synaptosomal Ca2+ influx and neurosecretion. Middlemiss and Spedding (1985) reported that submicromolar concentrations of the DHP agonist, Bay K 8644, could augment K*-evoked neurotransmitter release from cortical brain slices. The augmented release was blocked by the DHP antagonist, nifedipine (Middlemiss and Spedding, 1985). 61 Woodward and Leslie (1986) and White and Bradford (1986) independently obtained similar results in synaptosomes (Woodward and Leslie, 1986; White and Bradford, 1986; Woodward et al., 1988a). From the above studies, it appears that L-type Ca2+ channels in the CNS may modulate neurotransmitter release but do not contribute under normal circumstances. As you may recall, this was the conclusion drawn from studies using rat sympathetic neurons as well as rat NMJs (Perney at al., 1986; Atchison, 1989). In contrast, various investigators report direct block of synaptosomal K*- stimulated Ca2+ influx and neurotransmitter release by nifedipine, nimodipine and nitrendipine (Turner and Goldin, 1985; White and Bradford, 1986; Dunn, 1988). Still, others report no effect of DHPs on K"-stimulated Ca2+ uptake into synaptosomes (Nachshen and Blaustein, 1979; Daniell et al., 1983; Wei and Chang, 1985; Suszkiw at al., 1986; Reynolds 91 al., 1986; Vickroy et al., 1992). Thus, the 4.. relationship of DHP receptors to synaptosomal Ca2 channel function is still uncertain. Synaptosomes also contain specific high-affinity receptors for the N-type channel antagonist, Cng. (Abe et al., 1986; Marqueze et al., 1988; Wagner at al., 1988; Scheer, 1990). Polymixin B and several aminogycoside antibiotics potently inhibit binding of radiolabeled Cng to guinea-pig cerebral cortex membranes (Knaus et al., 1987) and to rat forebrain synaptosomes (Wagner et al., 1988). Rapid “Caz+ flux measurements into rat brain synaptosomes demonstrates only partial antagonism (a 20-40%) of voltage-dependent Ca2+ channels by Cng, 62 however (Reynolds et al., 1986; Suszkiw et al., 1987; Woodward et al., 1988b; Suszkiw et al., 1989). Thus, N-type channels may mediate only a proportion of Ca2+ uptake and neurotransmitter release from synaptosomes. Because depolarization-induced Caz+ entry into rat brain synaptosomes is largely resistant to both Cng and DHPs, it appears that synaptosomes contain other high-voltage-activated channels which may be neither N- nor L-type (Lundy et al.,1992). In addition, Suszkiw and coworkers (1989) demonstrated that it was not possible to clearly classify the Ca2+ channels into the T-, L- or N-type based on their activation and inactivation kinetics (Suszkiw et al., 1989). In some respects, Ca2+ influx into synaptosomes has characteristics similar to currents mediated by P-type channels (Leonard et al., 1987; Uinas et al., 1989; Suszkiw et al., 1989). Likewise, the P-type Ca2+ channel blockers, m-agatoxin IVA and FTX, are potent blockers of K*-induced “Ca2+ uptake into rat brain synaptosomes (Mintz et al., 1992; Uchitel at al., 1992). These findings suggest that P-type Ca2+ channels may be involved in transmitter release from mammalian synaptosomes, but this requires verification. C. Overall Aim It is reasonably clear that LEMS is due to autoimmune attack directed against the nerve terminal in affected patients. However the exact identity of the target remains speculative. In large part this is due to the inacceslbility of the putative target site, the voltage-dependent Ca2+ channel in the presynaptic motor 63 nerve terminal. The overall objective of this dissertation was to test the hypothesis that a circulating autoantibody in the plasma of patients with LEMS interferes with nerve terminal Ca2+ channels associated with neurotranmitter release. To test this, the model system of isolated nerve terminals from mammalian CNS was used. Several lines of pharmacological evidence suggest close similarity between Ca“ channels in synaptosomes and those at motor nerve terminals. Cng and DHPs are unable to block release of ACh from murine motor nerve terminals (Anderson and Harvey, 1987; Atchison, 1989; Protti, 1991). Rapid “Ca"+ flux measurements in rat brain synaptosomes have shown that the voltage-dependent Caz+ channels that mediate transmitter release are, in most but not all cases, DHP insensitive (Nachshen and Blaustein, 1979; Turner and Goldin, 1985; Suszkiw et al., 1986; Woodward and Leslie, 1986) and at best are only partially antagonized by Cng (Reynolds, 1986; Suszkiw et al., 1989) as well. However, mammalian neuromuscular transmission and voltagedependent Ca2+ channels in synap- tosomes are effectively blocked by FI'X (Uchitel et al., 1992). FTX blocked presynaptic Ca2+ currents, nerve-evoked transmitter release and muscle contraction from a murine nerve-muscle preparation as well as K*-stimulated 45Ca2+ influx into rat brain synaptosomes (Uchitel, 1992). Thus, synaptosomes may be an useful model system in which to study the potential interaction of the autoantibody in LEMS with Ca2+ channels involved in neurotransmitter release. 1. Specific Alm 1 Incubation of rat cortical synaptosomes with serum or high molecular weight serum fractions from patients with LEMS decreased the amount of [3Hj-ACh released during depolarization (Mayer et al., 1986). This study was important in that it indicated that acute exposure of synaptosomes to the serum or serum fractions from LEMS patients could elicit responses qualitatively similar to those obtained in chronic passive experiments in mice. It did not measure Ca2+ channel activity, however. Therefore, studies were performed to determine whether acute application of plasma and serum from patients with LEMS could reduce depolarization-dependent uptake of Ca2+ into rat forebrain synaptosomes. The results from these studies are presented in Chapter Three. 2. Speclflc Alm 2 Chronic injection of mice with plasma, crude lgs or purified lgG from patients with LEMS was effective in transferring both the electrophysiological and morphological characteristics of the disease (Denys et al., 1972; Lang et al., 1981; Newsom-Davis eteI., 1982; Fukunaga etal., 1983; Ifim, 1985; 1986; Lennon and Lambert, 1988; Nagel et al., 1988). This led to the conclusion that the abnormality in LEMS was due to the production of lgG autoantibodies. The objective of the studies in Chapter Four was to test whether acute application of IgG isolated from a patient with LEMS could disrupt synaptosomal Ca2+ channels, and if so, whether other ion channels in nerve terminals were similarly affected. In addition, studies 65 were performed to determine if non-specific membrane damage contributed to the pathogenesis of LEMS. 3. Speclflc Aim 3 Activation of the serum component complement plays a central role in mediating postsynaptic membrane damage in animal models of the autoimmune neuromuscular disorder MG, as well as in the human disease state (Engel at al., 1977; Sahashi et al., 1978). Attempts to clarify the participation of complement in the pathogenesis of LEMS have so far yielded negative results (Prior et al., 1985; Lennon and Lambert, 1988). As it is possible that complement activation may be important in LEMS, studies were undertaken to assess the role of serum in the pathogenesis of Ca2+ channel dysfunction induced by LEMS lgG in isolated nerve terminals from mammalian CNS. The results from these studies are presented in Chapter Five. 4. Speclflc Alm 4 LEMS immunoglobulin (lg) appears to interact with voltage-dependent Ca2+ channels in a number of model cell systems (Roberts et al., 1985; Login et al., 1987; Peers et al., 1987; De Azipurua et al., 1988a, 1988b; Kim and Neher, 1988; Lang et al., 1989a). Although, these studies have proven useful in the characterization of the subtype of Ca2+ channel affected by the LEMS autoantibody in the cell some, the exact identity of the subtype(s) of Ca2+ channel affected in 66 nerve terminals remains unknown. Therefore, attempts to characterize pharmacologically the subtype of Ca2+ channel in synaptosomes that is affected by the LEMS autoantibody are presented in Chapter Six. CHAPTER THREE EFFECT OF SERUM AND PLASMA FROM PATIENTS WITH LAMBERT- EATON MYASTHENIC SYNDROME ON DEPOLARIZATION-INDUCED UPTAKE or 45cit2+ INTO RAT CORTICAL SYNAPTOSOMES 67 A. Summary The reduction in nerve-evoked release of transmitter at the neuromuscular junction of patients with LEMS is thought to be caused by a circulating autoantibody to calcium channels of presynaptic motor nerve terminals. Studies were undertaken to determine whether acute application of plasma and serum from patients with LEMS or SCC would reduce depolarization-dependent uptake of “scat+ into isolated nerve terminals of the CNS. Net potassium-stimulated influx was reduced by sera and plasma from five patients with LEMS but not by sera from four patients with 900. Lactate dehydrogenase (LDH) release from synaptosomes incubated with plasma or serum from patients with LEMS was not increased over control indicating the reduction in “Ca2+ uptake was not a result of disruption of synaptosomal membrane integrity. These results are significant because: (1) they demonstrate that acute exposure to a circulating factor in sore/plasma from a patient with LEMS is sufficient to inhibit Ca2+ channel activity in isolated nerve terminals, as opposed to chronic regimens used in other models for the disease; (2) they indicate that the existence of SCC alone in some patients is insufficient to trigger a LEMS-like autoimmune response; (3) they suggest that Ca2+ channels of nerve terminals secreting different transmitters may share common epitopes recognized by the LEMS autoantibody; and (4) they suggest that synaptosomes will be useful in determining the neurochemical site and specificity of the LEMS autoantibody. B. Introduction LEMS is a presynaptic disorder of neuromuscular transmission frequently associated with 800 (Eaton and Lambert, 1957; Wise and MacDermot, 1962; Elmqvist and Lambert, 1968). The disease is characterized by a reduction in Ca2+- dependent quantal release of ACh in response to nerve stimuli (Lambert and Elmqvist, 1971; Cull-Candy et al., 1980). There is no evidence of a defect in ACh synthesis nor of decreased sensitivity of the postsynaptic membrane to ACh (Lambert and Elmqvist, 1971; Lindstrom and Lambert, 1978; Cull-Candy et al., 1980). Muscle fiber excitability and nerve conduction appear normal (Pascuzzi and Kim, 1978). Thus, LEMS is presumed to result from defective transmitter release. Evidence for an autoimmune origin of LEMS exists (Lang et al., 1981; Lennon at al., 1982; Newsom-Davis and Murray, 1984). The ability to transfer the human neuromuscular defect to mice by injections of either plasma or immunoglobulin derived from LEMS patients (Denys et al., 1982; Newsom-Davis etel., 1982; Lang etel., 1983a; Lang etel., 1984; Kim, 1985; Kim, 1986; Lambert and Lennon, 1988) supports the hypothesis that circulating autoantibodies to some presynaptic component are present in patient serum and mediate the pathogenesis of this disease. Evidence suggests that these autoantibodies interact with voltage- dependent Ca2+ channels since they have the ability to alter Caz+ flux in several cell lines of diverse origin (Roberts et al., 1985; Login et al.,1987; Peers et al., 1987; Kim and Neher, 1988; De Azipurua et al., 1988). These studies have proven useful in the characterization of the LEMS autoantibody's antigenic target in the cell 70 some. However, studies which specifically examine the ability of the LEMS autoantibody to interact with Caz+ channels in nerve terminals are lacking. This is important as it is possible that the type of Ca2+ channel in the terminal may differ from that in the cell body (Suszkiw et al., 1989; Charlton and Augustine, 1990; Uchitel et al., 1992). In the present investigation, we sought to determine whether acute application of plasma and serum from LEMS patients could disrupt Ca2+ channel activity in presynaptic nerve terminals from the CNS. Synaptosomes maintain the morphological and functional properties of intact news terminals and several studies suggest that Ca2+ channels in the CNS resemble those in the periphery (Anderson and Harvey, 1987; Nachshen and Blaustein, 1979; Reynolds et al., 1986; Suszkiw etel., 1986;1989; Atchison, 1989; Protti etel., 1991; Uchitel etel., 1992). LEMS patients do not, however, normally exhibit CNS symptomology because lgG is a high molecular weight protein (1:150 kDa) that does not readily pass the blood brain barrier. If the autoantibody in the sera of patients with LEMS could gain access to the calcium channels of the CNS by incubation in vitro, it might be possible to induce a passive transfer of the nerve terminal dysfunction to central neurons. C. Materials and Methods 1. Anlmals Male Sprague-Dawley rats (Harlan; 175-2009) were maintained on a 12 hr light—dark cycle in a temperature- and humidity-controlled room for at least 3 days 71 prior to use. Standard laboratory rodent chow and water were provided ed Iibitum. 2. Serum/Plasma Samples Sara and plasma samples from patients diagnosed as having LEMS were kindly provided by the following persons: Andrew Massey, M.D., Dept of Neurology, University of Kentucky Medical Center, Lexington,KY; Mark Glasberg, M.D., Henry Ford Hospital, Dept of Neurology, Detroit, MI; David Johnson, MD, Vanderbilt University, Nashville, TN; Donald Sanders, MD, Division of Neurology, Duke University Medical Center, Durham, NC; Daniel Watson, M.D., Dept of Neurology, Johns Hopkins School of Med, Baltimore, MD; Eva Feldman, M.D., Dept. of Neurology, University of Michigan, Ann Arbor, MI and Daniel Drachman M.D., Dept of Neurology, John Hopkins School of Med., Baltimore, MD. Plasma and serum samples from patients with 800 were provided by David Johnson, MD, Vanderbilt University, Nashville, TN and purchased from the Cooperative Tissue Network, Columbus, OH. Control plasma from patients with no history of neurological or immunological disease was purchased from the American Red Cross, Lansing MI. 3. Preparation of Synaptosomes Synaptosomes were prepared from whole brain (minus the brainstem and cerebellum) of male Spraque-Dawley rats according to a modified method of Gray and Whittaker (1962). Forebrains were homogenized in 10 ml of 0.32 M unbuffered sucrose and centrifuged at 1000xg, 4°C, for 10 min. The supernatant was 72 decanted and the pellet discarded. The supernatant was next centrifuged at 10,000xg for 20 min and the resulting pellet resuspended in fresh 0.32 M unbuffered sucrose. The synaptosomal fraction was further purified by layering onto a sucrose density gradient consisting of 0.8 and 1.2 M unbuffered sucrose, and centrifuged at 69,000xg for 2 hr. The synaptosome-enriched fraction was subsequently removed from the interface between the 0.8 and the 1.2 M layers of sucrose, and diluted 1 :1 (v/v) with a physiological saline buffer. The synaptosomal suspension was then washed free of sucrose by centrifugation at 10,000xg for 10 min and resuspended in the physiological saline to an approximate concentration of 15-25mg protein/ml as determined by the method of Lowry et al., (1951). Following isolation, equal aliquots of the synaptosomal suspension were incubated at 37°C for one hr, while shaking, in a pure O2 atmosphere with varying concentrations of either plasma or sera (expressed as " % of total incubation volume”) from patients with no known neurological or immunological illness (control), SCC in which no neurological involvement was observed, or LEMS. The proper volume of physiologic saline buffer was added to maintain a constant reaction volume. 4. Solutions The solutions used in the 4"’Ca2+ uptake experiments contained the following: physiological saline buffer (mM): NaCl, 145; KCI, 5; MgClz, 1 ; d-glucose, 10; and HEPES, 10; depolarizing K: solutions (mM): NaCI, 72.5; KCI, 77.5; MgCIz, 1; CaClz, 0.04; d-glucose, 10; and HEPES, 10; and quench solution (mM): KCI, 5; 73 MgClz, 2; EGTA, 1; d—glucose, 10; HEPES, 10; and N-methyl-d-glucamine, 145. Non-depolarizing K“ solutions were identical to depolarizing K“ solutions except that [K"], was 5 mM, and 72.5 mM choline chloride was added to maintain iso- osmolarity. The LDH reaction mixture contained (mM): KzHPO,, 41.2; KH2P04, 6.8; Na+ pyruvate, 0.6; NADH, 0.2 ; and NaHCOa, 1.1. 5. Measurement of “Ca2+ Uptake Uptake of “Ca“ was measured as described previously (Atchison et al., 1986). fifty ill of the synaptosomal suspension was added to 50 ill of either depolarizing or non-depolarizing K*-buffer containing 1 llCi of “Ca2+ (New England Nuclear, Boston, MA) and uptake was measured for 10 sec. The uptake of 4ice“ associated with transmitter release occurs within the first 2 sec and accounts for 80% of the uptake measured over a 10 sec period (Nachshen and Blaustein, 1980; Drapeau and Blaustein, 1983). Therefore, these data accurately reflect the uptake of “"‘Ca2+ associated with voltage-dependent Ca”+ channels in the synaptosomes as reported previously (Nachshen and Blaustein, 1980; Drapeau and Blaustein, 1983; Suszkiw et al., 1986; Suszkiw et al., 1989; Shafer and Atchison, 1989). The reaction was stopped by the addition of 2 ml of cold quenching solution followed by suction-filtration of the sample through 0.45 itM HVLP Millipore filters (Millipore, Bedford, MA). Radioactivity remaining on the filters in low [K‘], solution was subtracted from that obtained in high [K‘]. solutions as depolarization- 74 independent background. Experiments were replicated five to seven times. Values obtained for any particular experiment were the average of three replicates. Readings of dpm were corrected back to the original volume of synaptosomes filtered and expressed as nmole “Cazv 119 protein (Lowry et al., 1951). 6. Measurement of Lactate Dehydrogenase (LDH) Activity LDH release was quantified by the method of Bergmeyer and Bernt (1974). Synaptosomes were incubated with varying concentrations of control and LEMS lgG as described above. Immediately after the incubation period, synaptosomes were pelleted by centrifugation. The supernatant was decanted into a separate test tube. A 30 ul aliquot of supernatant was added to cuvettes containing 3 ml of the reaction mixture. LDH activity was quantified by the rate of oxidation of nicotinamide adenine dinucleotide (NADH), which was followed spectrophotometrically at 340 nM for 3 min. Activity was expressed as the percentage of LDH released into the supernatant. Total LDH activity (100 %) was determined by lysing an equal volume of synaptosomes in 0.1 % (v/v) Triton-X-100 followed by sonication. 7. Statlslcal Analysls Data were analyzed via a randomized complete block analysis of variance followed by Dunnett’s t-test or by Student’s t-test for paired samples. Statistical significance was assessed at p < 0.05. 75 D. Results 1. Uptake of “Ca” Many LEMS patients, including patients used in this study, have SCC. Uptake of “Ca2+ was determined following incubation of synaptosomes with SCC serum. Serum from four patients with SCC in which no neurological involvement was observed produced no substantial reduction in depolarization-dependent uptake of “Ca2+ into rat forebrain synaptosomes (fig. 3.1). The effect of both control and LEMS patient sera on “Ca2+ uptake into synaptosomes during 10 sec K*-stimulation is depicted in figure 3.2. Concentrations of 6 and 12% (v/v) control sera and plasma, had little effect on depolarization-induced uptake of “Ca2+ into synaptosomes when compared to uptake into synaptosomes with no added sera/plasma. However, incubation with 24% (v/v) control plasma/sera significantly reduced uptake compared to incubations with no sera/plasma (fig. 3.2). Conversely, incubation of synaptosomes with plasma or sera from five LEMS patients at concentrations of 6 and 12% of the total incubation volume resulted in significant reductions in net K*-stimulated influx (fig 3.2). Inhibition of 4"’Ca'“ uptake ranged from 0 to 20% (10.6 t 4.2%) and 14 to 30% (20.4 t 2.6%) at concentrations of 6 and 12%, (v/v) respectively. At a concentration of 24% (v/v), LEMS sera/plasma reduced influx by 0 to 61.8% (22.4 1 10.8%), but this was not significantly less than its collective paired control. Incubation of synaptosomes with serum from a sixth patient with LEMS reduced uptake of “Ca” by 20.2 t 18.1%, 76 4. —-con —scc s " ' 0 iii. a 5% —"-e- i. " “-3 ' At T ° Dazp—‘r— . 8 —-—-_ I}: O A D 83:“ i' ‘2 5 o I I I In 0 6 12 24 SERUM CONCENTRATION (% of total lncubatlon volume) Figure 3.1. Effect of sera from 4 patients with SCC and no evidence of neurological involvement on “Ca2+ uptake into rat forebrain synaptosomes during 10 sec of depolarization with 42.5 mM KCI. Prior to uptake, synaptosomes were incubated for 1 hr with SCC or control serum. Uptake in 5 mM KCI was subtracted from that in 42.5 mM KCI as depolarization-independent background. Each symbol represents the mean of at least 3 experiments per patient. Paired experiments are denoted by identical symbols in the control and SCC group. The horizontal bar represents the mean of all 4 patients tested. 77 A6” ——con —LEMS 3 v £5 .0 V V 5% 1 , _. e v 33.3". .a 8b. a s o t — "“0 8; El U E — :2 '3' 15 ° El E 5 O 0 i l 1 #4 0 6 12 24 SERA/PLASMA CONCENTRATION (% of total Incubation volume) Figure 3.2. Effect of plasma and sera from 5 patients with LEMS on uptake of 45Ca2+ into rat forebrain synaptosomes during a 10 sec depolarization. Synaptosomes were incubated for 1 hr with either LEMS serum or plasma or control serum or plasma, respectively. Uptake was measured as described in figure 1.1. Each symbol represents the mean of at least three experiments per patient. Paired experiments are denoted by identical symbols in both the control and LEMS-treated group. The horizontal bar denotes the mean of all 5 patients tested. ‘ Indicates values which are significantly less than its corresponding paired control value (p < 0.05). b Indicates a value that is significantly less than uptake into synaptosomes that received no serum or plasma (p < 0.05). 78 52.2 2 18.4% and 43.8 1 9.4% at 10, 30 and 50% of the total incubation volume, respectively (fig 3.3). 2. Synaptosomal Membrane Integrity To rule out the possibility that the effect of LEMS sera/plasma was a result of disruption of the synaptosomal membrane integrity, synaptosomal viability was assessed by measuring lactate dehydrogenase (LDH) released into the extrasynaptosomal fluid following incubation with plasma and sore. Neither incubation with LEMS serum nor LEMS plasma increased LDH release compared to incubation with control serum or plasma (figs. 3.4, 3.5). E. Discussion The aim of the present study was to test whether acute application of plasma and serum from LEMS patients could disrupt synaptosomal Ca2+ channel activity. Our results indicate that serum and plasma from patients with LEMS act on presynaptic nerve terminal Ca2+ channels in the CNS to reduce depolarization- induced uptake of 45Ca“. These deficits are not due to a decrease in synaptosomal viability as LDH release was not increased over that released by control sera/plasma. These studies, to my knowledge, are the first to examine specifically the ability of a circulating humoral factor in serum from patients with SCC and LEMS to interact with nerve terminal Ca2+ channels. This is important as the putative site of action for the LEMS autoantibody is believed to be Ca2+ channels on motor nerve terminals. 79 I: con LEMS A figure 3. 3. Effect of serum from one patient with LEMS on 45Ca2+ uptake into synaptosomes. Synaptosomes were incubated with serum for 1 hr and uptake measured as described in Fig. 1.1. The asterisk, , indicates uptake that is significantly smaller than its respective control (p < 0. 05). O) ]——r |——I |__. 3W “Ca” UPTAKE (nmoI/pg protein E-OS) o) k\\\\\\\\\\\\l—” CD 30 5 SERUM CONCENTRATION (% total incubation volume) 0 1OO - _ —A- CONTROL a fi 3 —o-- LEMS O . =3 5° I :I: 32 T I 1.--...“ d T/ l l l l J 0 1O 20 30 40 50 0 SERUM CONCENTRATION ( % of total Incubation volume ) figure 3.4. Effect of LEMS sera on synaptosomal viability. LDH activity was measured from the supernatant of synaptosomes treated for 1 hr with either LEMS or control sera. LDH activity was quantified by the rate of oxidation of NADH at 340 nm. The quantity of enzyme is expressed as the percent release of the total enzyme content of the synaptosomes, which was determined by lysis of an equal volume of synaptosomes with 0.1% (v/v) Triton-X-100 followed by sonication. Values are the mean 1 SEM of 3 experiments. 81 100 ' -A- CONTROL Ill ‘3 m 13 ’ —o- LEMS < 8 m I- —l o I.“ 3 50 ' o: . i iii 0 I -' 2 ’ OT”~~T»”’A;> ‘~ 3 s r‘” 1 o l l l 1 I O 10 2O 30 40 50 PLASMA CONCENTRATION (% of total Incubation volume) Figure 3.5. Effect of LEMS plasma on synaptosomal viability. LDH activity was measured from the supernatant of synaptosomes treated for 1 hr with either LEMS plasma, control plasma or physiological saline. LDH activity was quantified by the rate of oxidation of NADH at 340 nm. The quantity of LDH is expressed as the percent over release in physiological saline only. All values are the mean 1 SEM of 5 experiments. 82 Serum from SCC patients lacking neurological symptoms did not reduce depolarization-dependent uptake of “Ca2+ into synaptosomes. This study controlled for possible effects associated with neoplastic disease, since approximately 60% of patients with LEMS also have SCC of the lung (Eaton and Lambert, 1957; Wise and MacDermot, 1962; Elmqvist and Lambert, 1968). Oat cells (small cells) possess several neuronal properties and are possibly of neural origin (Gadzar et al., 1981; Moody at al., 1981; Tapia et al., 1981). They are electrically excitable, and action potentials in cat cells may be calcium channel- mediated as they are inhibited by Co2+ and Mn“, known calcium channel blockers (McCann et al.,1981; Roberts et al., 1985). It has been suggested that LEMS may arise as an immune response to SCC and that cholinergic neurons may have similar epitopes leading to cross-reactivity between tumor and neuron (Newsom- Davis, 1982; Lennon et al., 1982; Vincent et al., 1989). LEMS, however, also exists in the non-neoplastic form and autoantibodies antagonistic to SCC Caz+ channel activity were found in patients with LEMS independent of cancer status (Roberts et al., 1985; De Azipurua et al., 1988). Since sera from SCC patients failed to inhibit calcium flux at any concentration tested, this suggests that the existence of SCC neoplastic disease is insufficient to trigger an autoimmune response in all patients. These data also demonstrate that interpretation of data derived from LEMS patients with SCC would not be complicated by the presence of carcinoma. Incubation of synaptosomes with control plasma/sera at a concentration of 24% (v/v) significantly reduced “Ca2+ uptake compared to uptake in 83 synaptosomes which were incubated without sera/ plasma. This may be explained simply by increased serum binding of 45Ca2+ thereby effectively decreasing the concentration of radioactivity available for uptake by synaptosomes. This needs to be tested further. It is evident from the variability in the data that the circulating factor found in different LEMS patients sera/plasma is not equieffective in disrupting “Ca2+ influx into synaptosomes. This is most likely due to differences in antibody titer between patients. Nevertheless, a modest yet significant decrease in uptake mediated largely by voltage—dependent Ca2+ channels was observed in synaptosomes treated with 6 and 12% sera or plasma from five LEMS patients as compared to their paired controls. A striking decrease in synaptosomal 45Ca2+ uptake was observed in one patient studied in isolation. Neither incubation with LEMS plasma nor LEMS serum increased LDH release compared to incubation with control serum or plasma. Thus, it is unlikely that the decrease in calcium uptake is due to decreased synaptosomal viability, but rather suggests block or disruption of presynaptic voltage-dependent calcium channels in the nerve terminal. It is well established that nerve-evoked quantal release of neurotransmitter is dependent upon the influx of extracellular Caz+ through membrane ionic channels (Katz and Miledi, 1965; Katz and Miledi, 1967a; Dodge 91 al., 1969). In the absence of 0a“, the nerve action potential invades the nerve terminal but fails to evoke transmitter release (Katz and Miledi, 1965; Dodge and Rahamimoff, 1967). 84 Transmission at many synapses has a steep dependence on extracellular Ca2+ concentration (Dodge and Rahamimoff, 1967; Katz and Miledi, 1970; Drapeau and Blaustein, 1983; Smith et al., 1985; Zucker and Fogelson, 1986). These observations have led to the hypothesis that Ca2+ ions act in a cooperative, exponential manner at intracellular sites to trigger secretion of transmitter quanta. The power index appears to vary considerably at different synapses and it has been reported that transmitter release following a nerve impulse may increase with as much as the fourth or even higher power of [Ca2+]. (Dodge and Rahamimoff, 1967; Katz and Miledi, 1970; Zucker and Fogelson, 1986). This power relationship may be of functional importance in controlling transmitter release. Changes in the magnitude of presynaptic Ca2+ entry could have an amplifying effect on the amount of neurotransmitter secreted (Smith et al., 1985). Thus, if transmitter secretion at the neuromuscular junction and/or in synaptosomes follows this power relationship then modest changes in Ca2+ influx, such as those in this study, could conceivably result in large changes in neurotransmitter secretion. Rat cortical synaptosomes are neurochemically heterogeneous with only 10% percent of the suspension thought to be cholinergic (Gray and Whittaker, 1962). Incubation with LEMS sera/plasma resulted in decreases of “Ca2+ uptake which ranged from 0 to 60%. These decreases suggest that the autoantibody must interact with nerve terminal Ca2+ channels which secrete different neurotransmitters. These data also provide further evidence for the similarity between somatic motor nerve terminal and central nerve terminal Ca2+ channels. 85 Reduction in 10 sec depolarization-dependent “Ca2+ uptake into synaptosomes was evident following 1 hr of incubation of plasma/sera. Thus, acute exposure to LEMS autoantibody was sufficient to induce a change in activity associated with Ca2+ channels. This is in contrast to other models of the disease which have utilized chronic exposure. Using this acute, in vitro system may allow for more efficient use of the isolated immunoglobulin fraction from patients sera, in subsequent studies. This is desirable as patient sera is often difficult to acquire as the incidence of the disorder is low. In conclusion, incubation of rat cortical synaptosomes with sera/plasma from patients with LEMS but not SCC reduced depolarization-dependent uptake of 4"’Ca'“ This was not due to decreased synaptosomal viability. The results from this study are significant as they demonstrate that acute rather than chronic exposure to LEMS sera/plasma is sufficient to inhibit Ca2+ channel activity in isolated nerve terminals; that data obtained from patients with LEMS with and without SCC can be analyzed identically; and that Ca“ channels of nerve terminals secreting different transmitters may share common epitopes recognized by the LEMS autoantibody. Thus, the synaptosomal preparation may a useful model in which to study, in more detail, the potential interaction of the autoantibody in LEMS with calcium channels involved in transmitter release. CHAPTER FOUR SPECIFICITY OF LAMBERT-EATON MYASTHENIC SYNDROME IMMUOGLOBULIN FOR NERVE TERMINAL CALCIUM CHANNELS 86 87 A. Summary Data from Chapter Three provide evidence which suggests that synapto- somal Ca2+ channel function is altered by incubation with serum /plasma from patients with LEMS. The goal of the present study was to examine the specificity of LEMS autoantibodies for nerve terminal Caz+ channels compared to other voltage-sensitive ion channels in nerve terminals, and to determine if non-specific membrane damage contributed to the pathogenesis of LEMS. The ion channel specificity of LEMS autoantibody was assessed by comparing the ability of acute application of lgG isolated from the plasma of a patient with LEMS to reduce depo- larization-dependent uptake of “Ca“ and ”Na” into or efflux of “RV from rat forebrain synaptosomes. The clinical diagnosis of LEMS was confirmed electro- physiologically by treatment of mice for 30 days with plasma (1.5 mI/day) taken from this patient. Characteristic reduction of quantal content elicited at 1 Hz and facilitation at 20 Hz was observed in mice treated with LEMS plasma compared to those treated with control plasma. One sec, K*-stimulated 45Ca2+ uptake was inhibited 36.5 1 14.5% and 44.5 1 9.8% by acute application of 2 and 4 mg/ml LEMS lgG, respectively; lgG from patients with SCC had no effect on 45Ca2+ entry. The same concentrations of LEMS lgG affected neither voltage-dependent uptake of 22Na* into veratridine-depolarized synaptosomes nor a“Rb" efflux from K"- depolarized synaptosomes. LDH release from synaptosomes incubated with varying concentrations LEMS IgG was not increased over control indicating the reduction in 45Ca2+ uptake was not a result of disruption of synaptosomal 88 membrane integrity. Likewise, normal activation of Ca2+ channels was not pre- vented by IgG-induced alterations in synaptosomal membrane potentials as the fluorescence of a membrane potential sensitive carbocyanine dye (diS-C2(5)) was unaltered following acute application of 2 mg/ml LEMS or control lgG. These results indicate that immunoglobulins isolated from a patient with LEMS alter presynaptic nerve terminal Ca2+ channel activity. These deficits are not due to decreases in synaptosomal viability nor to lgG-induced alterations in the synaptosomal membrane potential but appear to be a direct block or modulation of voltage-dependent Ca2+ channels. The antibody from this patient appears to be specific for nerve terminal Ca2+ channels as it did not affect flow through voltage- dependent Na+ or K+ channels. B. Infmducfion The electrophysiological features of LEMS include reduced quantal content at low frequencies of nerve stimulation with marked facilitation at high frequencies of stimulation (Elmqvist and Lambert, 1968; Lambert and Elmqvist, 1971; Cull- Candy et al., 1980). Morphologic data suggest that AZPs represent the target of the pathogenic autoantibodies in LEMS (Fukunaga et al., 1982; 1983; Fukuoka et al., 1987). Mice injected with either plasma or immunoglobulin derived from LEMS patients (Kim, 1985:1986; Lang et al., 1983) reproduce both the electTOphy- siological and morphological features of LEMS. It has been proposed that AZPs correspond to voltage-sensitive Ca2+ 89 channels of the presynaptic membrane (Pumplin etel., 1981; Robitaille at al., 1990; Tarelli at al., 1991; Cohen et al., 1991). Thus, the reduced transmitter release in LEMS could be related to a reduced entry of Ca2+ into the nerve terminal. In support of this contention, LEMS autoantibodies appear to interact with voltage- dependent Ca2+ channels in several cell lines of diverse origin (Roberts et al., 1985; Login et al., 1987; De Azipurua etel., 1988a, 1988b; Kim and Neher, 1988; Lang et al., 1989a; Peers et al., 1990). Although these studies have proven useful, several important deficiencies still remain. first, the identity of Ca2+ channels in motor nerve terminals remains unclear. Ca2+ channel subtypes in nerve terminals or populations of nerve terminals may differ from those in the cell body (Suszkiw et al., 1986; Charlton and Augustine, 1990). Second, the function and number of ion channels could be affected by damage to the membrane or membrane depolarization. Third, no assessment has been made of the Ca2+ channel specificity of the LEMS autoantibody as opposed to other voltage-gated channels present in nerve terminal. An antibody directed at an epitope on voltage-dependent Ca2+ channels might also affect other voltage-dependent ion channels if the lgG recognizes portions of the channel proteins common to all ion channels. Data from the previous chapter indicates that serum and plasma from patients with LEMS reduces depolarization-dependent uptake of “Ca2+ into synaptosomes (Hewett and Atchison, 1991). As an lgG autoantibody is presumed to mediate the pathogenesis of LEMS, the present study was designed to answer the following three questions: does lgG isolated from a patient with LEMS reduce 90 depolarization-dependent “Ca2+ into nerve terminal Ca2+ channels?; if so, does the lgG cause Caz+ channel dysfunction secondary to nonspecific membrane damage or alterations of membrane potential?; and are other voltage-dependent ion channels in the terminal also affected? C. Materials and Methods 1. Materials Dimethyl sulfoxide (DMSO) was obtained from EM Science (Gibbstown, NJ). Tetraethylammonium chloride (TEA), tetrabutylammonium chloride (TBA), tetrodotoxin (TTX), bovine serum albumin (BSA) cyclophosphamide, oligomycin, nickel chloride (NiCIz), ethyleneglycol-bisw-aminoethylether) N,N,N',N’-tetraacetic acid (EGTA), and ouabain were purchased from Sigma Chemical Co. (St. Louis, MO.) Sodium cyanide (NaCN) was obtained from Mallinckrodt Chemical Works (St. Louis, MO). Triton-X-100 was obtained from Research Products Intemetional Corp. (Elk Grove, IL). The radiolabels 4“Caz" (15.50 mCi/mg), at’Rb“ (1-10 mCi/mg), and ”Na“ (800-1000 mCi/mg) were purchased from New England Nuclear (Boston, MA). 3,3’-Diethyldicarbocyanine iodide [dis-02(5)] was obtained from Molecular Probes (Eugene, OR). Disease-free control human plasma was purchased from the American Red Cross (Lansing, MI). Control human serum was donated by a healthy volunteer. Plasma from patients with the clinical and electromyographic characteristics of LEMS was provided by Drs. Mark Glasberg, Eva Feldman, Andrew Massey, and James Albers. Serum from patients with SCC was provided by Dr. David Johnson. 91 2. Solutlons For electrophysiological measurements of neuromuscular function, solutions used for both bathing and recording contained (mM): NaCl, 137.5; KCI, 2.5; MgCIz, 1; CaClz, 2; d—glucose, 11; and HEPES, 14. The solutions used in the “Ca2+ uptake experiments were identical to those described in Chapter Three. For experiments using ”Na“, incubation buffer contained (mM): KCI, S; MgClz, 1; choline chloride, 145; d-glucose, 10; HEPES, 10; veratridine (VER), 0.15; ouabain, 0.1; BSA, 1 mg/ml; 1 TTX, 0.001. Uptake solutions were identical to the incubation buffer except that 2 mM NaCl and ”Na” (2 irCi/ml) were added and the ouabain concentration was increased to 1 mM. Cold quench solution was identical to that described in Chapter Three. Veratridine was used in lieu of K*-depolar- ization because voltage-dependent Na+ channels inactivate rapidly in response to normal depolarizing stimuli (T amkun and Catterall, 1981). The wash buffer used in the E"’Rb‘ efflux experiments contained (mM): NaCl, 145; KCI, 5; MgCIz, 2; RbCl, 0.1; d-glucose, 10; and HEPES, 10. Nondepolarizing efflwr buffer was identical to the wash buffer. In the depolarizing efflux buffer, NaCl was replaced by KCI for a final [K‘], of 150 mM. Cold stop solution contained (mM); TEA, 70; TBA, 10; NiCIz, 10; RbCl, 5; MgCIz, 1; d-glucose, 10; HEPES, 10; and N-methyI-d-glucamine, 50. Stock solutions of dis-02(5) for fluorescence measurements were dissolved in DMSO. Working solution were made daily by dilution of the stock into physiologi- cal saline buffer (final concentration, 0.5 uM). The final concentration of DMSO 92 after addition of dye to synaptosomes was 0.05% (v/v). 3. Immunoglobulln Isolation Control, SCC and LEMS sera and/or plasma were stored at -20°C until used. After thawing, fibrin was removed from the plasma samples by clotting which was facilitated by the addition of 1 M CaClz. Total immunoglobulin (Ig’s) were isolated by serial precipitations of serum in 30% and 50% saturated ammonium sulfate at 4°C as described by Heide and Schwick (1978). Each precipitate was pelleted by centrifugation at 1000xg for 15 min and redissolved in 0.85% (w/v) saline. The final precipitate was redissolved to 1/2 the original volume and dialyzed at 4°C against saline using Spectrapor 2 dialysis tubing (15-25000 m.w. cutoff, Spectrum Medical lnd., Los Angelese, CA) until the sulfate ions were no longer detectable (at 72 hr) as discerned by the barium chloride test (Garvey et al., 1977). fiocculent material formed during dialysis was removed by centrifugation at 1000xg for 15 min at 4°C. Ig’s were quantitated by bicinchoninic acid protein assay (Pierce Chemical, Rockford, IL). The lgG concentration was presumed to be 80% of the total protein (Kimbal, 1986). The LEMS lgG used in the synaptosomal studies was isolated from a patient with SCC. We show here that this patients’ plasma caused the characteristic reduction of evoked release of ACh at the neuromuscular junction (Patient 3; fig. 4.2). 93 4. Measurement of Ouantal Content (m) Plasma was used from three patients with typical clinical and electromyographic features of LEMS. One LEMS patient (Patient 3) has SCC. Plasma fractions (1.5 ml) from these patients and from disease-free human controls were injected daily intraperitoneally into Swiss Webster female mice (15-20 g) (Kim, 1985). In order to induce Immune tolerance to the foreign protein, animals were given cyclophosphamide, 300 mg /kg Up), one day following the first plasma injection (Lang et al., 1983). On day 30, the diaphragm with phrenic news intact was isolated and was perfused continuously with a low K” (2.5 mM) saline solu- tion continually aerated with 100% 02. Muscle fibers were transected approximately 4 mm on either side of the main intramuscular nerve branch to prevent contractions due to stimulation of the phrenic nerve (Barstad and Lilleheil, 1968; Glavinovic, 1979). Preparations were pinned out in a Sylgerd-coated plexiglass chamber and perfused with low [K*] saline. All experiments were performed at room temperature (~25°C) and replicated in a minimum of four animals. The setup for recording under current clamp is depicted schematically in figure 4.1 . Intracellular microelectrode recordings of synaptic potentials were made using conventional techniques and borosilicate glass microelectrodes (1.0 mm, WPI, New Haven, CT) having impedences of 15-30 meg it when filled with 3 M KCI. Prior to nerve stimulation, an average MEPP amplitude was obtained by recording 5 min of MEPPs on magnetic tape using an FM instrumentation tape recorder. EPPs were elicited by single shock rectangular voltage stimuli of 0.07 msec figure 4.1. Setup for recording under current clamp. Hemidiaphragm preparations with phrenic nerve intact were removed from control- and LEMS-treated mice. Preparations were transected to prevent contraction following stimulation of the nerve. Muscles were pinned down in a Sylgerd-coated plexiglas chamber and perfused with Low K“ buffer. End-plate regions were impaled with a KCL-filled microelectrode and EPPs and MEPPs recorded. figure modified from that contained in the dissertation of John M. Spitsbergen 95 duration applied to the phrenic nerve at a frequency of 1, 20, or 40 Hz (200 stimuli) allowing 5 min of rest between pulse protocols for redistribution of Ca2+. Ouantal content was determined off-line from the amplitudes and variances of the EPPs (variance method) and by dividing the average amplitude of EPPs by the average MEPP amplitude (mean value method). EPP amplitudes were standardized to a membrane potential of -40 mV (Katz and Theseleff, 1967). Only those EPPs and MEPPs with rise time s 1 msec were used for subsequent analysis. 5. Preparatlon of Synaptosomes Synaptosomal preparation was identical to that previously described in Chapter Three. In all experiments except “RN efflux, synaptosomes were incubated at 4°C under 100% 02 for 60 min with 10% (v/v) control human serum and varying concentrations of IgG from patients with no known neurological or immunological illness (control), SCC or LEMS. The rationale for using control human serum will be discussed in Chapter five. The proper volume of physiologic saline buffer was added to maintain a constant incubation volume. Prior to uptake measurements, synaptosomes were allowed to equilibrate for 30 min at 37°C. 6. Measurement of “Ca“ Uptake Uptake of “Ca2+ was measured as described in detail in Chapter Three. Readings of dpm were corrected back to to the original volume of synaptosomes filtered and expressed as nmole “Ca“ /u g protein or as % of disease-free control. 96 7. Measurement of 22m“ Uptake The time course of uptake of 22Na“ into rat forebrain synaptosomes was measured by the addition of 50 ul of synaptosomes to 50 ul of incubation buffer containing 150 uM VER with or without TTX (1 uM) for varying periods of time. VER is used to activate synaptosomal Na+ channels as these channels inactivate rapidly in response to K*-depolarization (Kreuger and Blaustein, 1980; Tankum and Catterall, 1981). NaCl was replaced isotonically with choline chloride in all solutions. Ouabain was present during both the incubation and uptake periods to prevent 22Na+ extrusion by blocking the Na”/K” ATPase. Following this incubation, a 100 pl of uptake solution containing 22Na+ (2 uCi/ml) was added. After 5 sec of incuba- tion, 2 ml of cold quench solution was added followed by suction-filtration. Uptake of ”Na" due to activation of voltage-dependent Na+ channels was determined as the difference of uptake in the presence and absence of TTX. Readings of dpm were corrected back to the original volume of synaptosomes filtered and expressed as nmole 22Na“/ug protein. 22Na‘ uptake following incubation with control and LEMS lgG (2 and 4 mg/ml) was measured as described above except that the synaptosomal suspension was added to the incubation buffer for 10 min only prior to the addition of uptake solution. 8. Measurement of l"’Flb“ Loadlng and Efflux Loading of synaptosomes with ”Rb“ and efflux under depolarizing conditions were done using a modification of the method of Bartschat and 97 Blaustein (1985a). Synaptosomes were first incubated for 45 min at 4°C with the following: (1) ”Rb“ (20 uCi/ml) alone; (2) a"Rb“ plus 10% (v/v) control human serum; (3) ”Rb", 10% (v/v) control human serum and control lgG (4 mg/ml); or (4) 13‘int)“, 10% (v/v) control human serum and LEMS IgG (4 mg/ml). Prior to the initiation of m5Rb+ release, synaptosomes were incubated for an additional 45 min at 37°C. It is during this time that the synaptosomes became loaded with the radiolabel. Next, 100 pl of labeled synaptosomes were placed on a filtration apparatus containing 2 ml of a wash solution and suction applied. Synaptosomes were washed in this manner with a total of 10 ml of washing solution to remove any extracellular “RV. Following the final wash, the filter was removed to a second filtration apparatus and 2 ml of depolarizing (77.5 mM) or resting (5 mM) K+ solution was added. Five sec later, 2 ml of stop solution was added and suction applied. The filtrate was collected directly. Radioactivity in the filtrate and that remaining on the filter was estimated by gamma counting. Efflux in low [K‘]. was subtracted from that in high [K], solutions as depolarization-independent back— ground. In order to control for possible differences in ”RV loading, efflux is expressed as a % of total. 9. Measurement of Lactate Dehydrogenase (LDH) Actlvlty LDH release was quantified by the method of Bergmeyer and Bernt (1974), as described in detail in Chapter Three. Activity was expressed as the percentage of LDH released into the supernatant. Total LDH activity (100%) was determined by lysing an equal volume of synaptosomes in 0.1% (v/v) Triton-X-100. 98 1o. Fluorescence Measurements Measurements of dis-02(5) fluorescence as an index of synaptosomal membrane potential were made as described by Hare and Atchison (1992). Briefly, a 50 ul aliquot of prewarmed synaptosomal suspension was added to 2 ml physiologic saline buffer containing 0.5 uM dis-02(5) and maintained at 37°C for 10 min. Determination of diS-Cz(S)-mediated fluorescence in synaptosomes was made using a SPEX fluorolog spectrofluorometer (SPEX Industries, Edison, NJ) at excitation and emission wavelengths of 649 and 682 nm respectively. In order to measure plasma membrane potential fluorescence independent of mitochondrial fluorescence, NaCN (2 mM) and oligomycin (4 ug/ml) were added to the suspension to collapse the mitochondrial membrane potential(Heinonen er al., 1985; Hare and Atchison, 1992). Once baseline fluorescence was judged to be stable, KCI (40 mM) was added to depolarize the plasma membrane. Data are expressed as the percent change from baseline fluorescence. 11. Statlstlcal analysls Differences in quantal content in animals receiving control or LEMS plasma were analyzed by one way analysis of variance (ANOVA) followed by Dunnett’s t- test. Uptake of 4"’Ca“ and 22Na+ were analyzed using Student’s t-test for paired samples. Efflux of “RV and fluorescence measurements were analyzed by a completely blocked ANOVA. The LDH experiments were analyzed by split-plot design ANOVA (2X5). Difference were considered significant at p < 0.05. D. Results 1. Quantal Content To verify experimentally the clinical diagnosis of LEMS, and thus correlate potential actions on synaptosomal Ca2+ channels with clinical neuromuscular effects, the ability of LEMS patient plasma to reduce quantal content when given chronically was examined. Quantal content of nerve-evoked EPPs was reduced significantly in mice receiving chronic injection of plasma from three patients clinically diagnosed with LEMS. Mean quantal content as determined by the variance method was reduced from s: 70 in diaphragms from control mice to 28-19 in diaphragms from mice receiving LEMS plasma (Fig. 4.2). This represents a reduction of approximately 60 to 70%. Similar percent reductions were observed using the mean value method (Table 4.1). MEPP amplitude was unaffected by treatment with LEMS plasma (Table 4.1). Cyclophosphamide alone had no effect on quantal content (results not shown). Facilitation of quantal content at 20 Hz was observed in diaphragms from mice treated with LEMS plasma from all three patients. Quantal content in mice treated with plasma from two of three patients facilitated at 40 Hz. Facilitation was not observed in mice treated with control plasma (Fig. 4.3). 2. “Ca” Uptake The effect of LEMS lgG from Patient 3 on uptake of “Ca2+ during 1 sec of K*-depolarization into rat forebrain synaptosomes is depicted in Figure 4.4. Depolarization-induced uptake was inhibited 36.5 2 14.5%, 44.5 i 9.8% and 34.6 100 vvvvvvv 100 '— z In '— z 8 _, so I5 * * * z < D O Control LEMST LEM82 LElvlss TREATMENT Figure 4.2. Effects of chronic treatment of mice with LEMS plasma on quantal content. Mice were treated for 30 days with 1.5 ml/day 0p.) of plasma from control or one of 3 different patients with the clinical diagnosis of LEMS. Neuromuscular transmission was examined using conventional intracellular recordings of MEPPs and EPPS elicited at 1 Hz from isolated hemidiaphragm preparations. Contractions evoked by phrenic nerve stimulation were prevented by transsection of the myofibers. Quantal content (m) was determined from the amplitudes and variances of the EPPs. At least 200 EPPs were used in the calculation of m. Values are the mean 2 SEM of at least four animals. Asterisks indicate values of quantal content which are significantly lower than control (p < 0.05). 101 Table 4.1 Determination of Quantal Content via Direct Method l m MEPP (mV) Control 9.75 z 2.2 0.353 t 0.05 Patient 1 3.92 2 1.7‘ 0.335 a 0.14 Patient 2 5.78 a 1.2‘ 0.328 .+. 0.04 Patient 3 3.33 .+. 0.51' 0.350 a 0.06 Mice were treated for 30 days with 1.5 ml/day 0p.) of plasma from control or one of 3 different patients with the clinical diagnosis of LEMS. Neuromuscular transmission was examined using conventional intracellular recordings of MEPPs and EPPS elicited at 1 Hz from isolated hemidiaphragm preparations. Quantal content (m) was determined by dividing the average amplitude of EPPs by the average amplitude of the MEPPs. Values are the mean 1 SEM of at least four animals. Asterisks indicate values of quantal content which are significantly lower than control (p < 0.05). 102 - Control. LEMS1[--] LEMS 2|: LEMS 3 € aoo~ 5 C o .. o zoor ax° v nF--__ 5 too- "" V 2 ’/ ': 0 d o g -1oo- 20 Hz 40 Hz STIMULATIONFREQUENCY Figure 4.3. Frequency-dependent facilitation of neuromuscular transmission in mice treated chronically with LEMS plasma. Treatments and recording procedures were as described in Figure 4.2. EPPs were elicited at the indicated frequencies of stimulation. Facilitation is defined as an increase in quantal content compared to that obtained at 1 Hz. 103 125 ' N'— ‘5 Ca” UPTAKE of control) \I 01 O .xmfi °° “N— "S 2 4 [lgG] (mg/ml synaptosomal suspension) Figure 4.4. Effect of LEMS lgG on uptake of “Ca2+ into rat forebrain synapto- somes. Synaptosomes were incubated for 1 hr at 4°C with varying concentrations of LEMS or control lgG and 10 96 control human serum. Prior to initiation of uptake, synaptosomal suspensions were equilibrated at 37°C for 30 min. One sec uptake in 5 mM K+ was subtracted from that in 42.5 mM K+ as depolarization- independent background. “Ca2+ uptake is expressed as 96 of control. All values are the mean 1- SEM of 7 experiments with each individual experiment being the average of 3 measurements. Asterisks indicate values which are significantly lower than their respective paired control (p < 0.05). 104 x 19.8% by 2, 4 and 8 mg/ml LEMS lgG respectively, as compared to synaptosomes incubated with equivalent concentrations of control lgG. Depolariza- tion-independent uptake was unaffected (results not shown). A number of LEMS patients, including the patient used in the previous and proceeding studies have 300. Uptake of “Ca“ was determined following incubation with SCC lgG pooled from a number of patients in which no neurological involvement was observed to control for possible effects associated with neoplastic disease. SCC lgG (2 and 4 mg/ml) produced no reduction in depolarization-dependent uptake of “Ca2+ (Fig. 4.5). 3. Synaptosomal Membrane lntegrlty To rule out the possibility that the effect of LEMS lgG was a result of nonspecific disruption of the synaptosomal membrane, LDH release following incubation with varying concentrations of LEMS and control IgG was measured. LDH released into the extrasynaptosomal fluid following incubation with LEMS lgG was not increased over that which occurred with control lgG (Fig. 4.6). Thus, it is + unlikely that decreased Ca2 uptake is due to disruption of synaptosomal membrane integrity. 4. Potentiometrlc Dye Fluorescence Alterations in the synaptosomal membrane potential by lgG could prevent normal activation of the Ca2+ channels. We tested for this possibility using the fluorescent, potentiometric carbocyanine dye dis-Cz(5). NaCN (2 mM) and 105 '3‘ con sec 4 i " Ca ‘* UPTAKE (nmollug proteln 505) .° 8 W “m 0 2 I G (ITIMM syn-small WWO") Figure 4.5. Effect of lgG from patients with small cell carcinoma (SCC) on uptake of “Ca“ into rat forebrain synaptosomes during 1 sec of depolarization by 41.25 mM KCI. Synaptosomes were incubated for 1 hr at 4°C with 10% control human serum and control or SCC lgG. Prior to initiation of uptake, synaptosomal suspensions were equilibrated for 30 min at 37°C. Uptake in 5 mM K+ was subtracted from that in 42.5 mM K” as depolarization-independent background. Data are expressed as nmol 45Ca/lig protein. All values are the mean 1 SEM of 4 experiments with each individual experiment being the average of 3 measurements. 106 100 r -0— CONTROL "I“ LEMS 50 LDH RELEASE (% of total) 0 1 2 4 8 (mg/ml synaptosomal Mien) Figure 4.6. Effect of LEMS lgG on synaptosomal membrane integrity. LDH release was measured from the extrasynaptosomal fluid following treatment for 90 min with 10 % control human serum and varying concentrations of LEMS or control lgG. LDH activity was quantified by the rate of NADH oxidation at 340 nM. The quantity of the enzyme is expressed as the percent of release of the total enzyme content of the terminals, which was determined by lysis of equal volume of synaptosomes with 0.1% triton-X-100 followed by sonication. Values are expressed as the mean i SEM of 4 experiments. 107 oligomycin (4 lug/ml) were first added to the suspension to collapse the mitochondrial membrane potential (Fig. 4.7). LEMS lgG caused no alterations in dye fluorescence following addition of KCI (40 mM) (Fig. 4.8). 5. ”Na“ Uptake In order to determine the length of time the synaptosomes should be incubated with VER, a time course of ”Na" uptake was performed. The reults are depicted in Figure 4.9. Five sec ”Na" uptake into synaptosomes is maximal following 10 min incubation with incubation buffer containing 150 AM VER. These results are similar to those of Tankum and Catterall (1981). Incubation of synaptosomes with concentrations of LEMS lgG which significantly altered uptake of “Ca2+ during K*-evoked depolarization had no effect on depolarization-induced uptake of 22Na+ into VER-treated synaptosomes (Fig. 4.10). e. “30* Loading and Efflux Treatment of synaptosomes with 10 % (v/v) control human serum or serum and either control or LEMS lgG (4 mg/ml) did not alter loading of a"Rb“ into rat forebrain synaptosomes (Fig. 4.11). Likewise, incubation of synaptosomes with concentrations of LEMS lgG which significantly altered uptake of 45Ca2+ during K”- evoked depolarization had no effect on “RE efflux from K*-depolarized synapto- somes (Fig. 4.12). 108 E 1.50 r 3 E _ 3‘ 2 71‘ 3 r: -s f £5 0.75 f t 2 5 K’ 3 v . l 8 cu- ; ollgo E 0.00 ‘ ' ‘ ' 0 200 400 Time (sec) Figure 4.7. Representative graph depicting typical fluorescent dye experiment. An aliquot of untreated synaptosomal suspension was added to 2 ml of physiologic saline buffer containing 0.5 uM diS-C2(5) and maintained at 37°C for 10 min. In order to measure fluorescence derived from the plasma membrane independent of mitochondrial fluorescence, NaCN (2 mM) and oligomycin (4 pg /ml) were added to the synaptosomal suspension to collapse the mitochondrial membrane potential. Once baseline was judged to be stable (visibly), KCI was added to depolarize the plasma membrane. Data expressed in counts per second. 109 100 FLUORESCENCE (% charge from baseline} 01 O O PLASMA MEMBRANE Figure 4.8. Effect of LEMS lgG on synaptosomal membrane potentials. Synaptosomes were incubated with 2 mg/ml of control or LEMS lgG, or no lgG for 90 min. Following incubation, synaptosomes were combined with a fluorescent, potentiometric dye, dis-C2(5). Initially, NaCN (2 mM) and oligomycin (4 ug/ml) were added to collapse the mitochondrial membrane potential. Addition of 40 mM KCI depolarized the plasma membrane potential. Data are expressed as % change in baseline fluorescence. Values are the mean 2 SEM of 3 experiments. 110 12l 22 Na * LPTAKE (W90 m 3'00) O) ‘o ‘ s 10‘ 20 Time (min) Figure 4.9. Time course of 22Na+ uptake into rat forebrain synaptosomes. Synaptosomes were added to incubation buffer which contained no added Na‘, 150 MM VER, 100 0M 1 TTX 1 uM) for varying periods of time. After incubation, uptake solution containing Na+ was added and allowed to mix for 5 sec subsequent to addition of cold quenching solution. Uptake is expressed as nmol 22Na"/ug protein. Values are the mean 2 SEM of 4 experiments. 111 CZICON -LEMS 2"’ Na * UPTAKE W m m .p. . |-' |—_. “—1 0 im 0 2 [I96] (mg/ml synaptosomal suspension) Figure 4.10. Effect of LEMS lgG on uptake of 22Na+ into rat forebrain synapto- somes depolarized with veratridine. Synaptosomes were incubated with lgG and serum as in Fig. 4.4 after which they were combined with veratridine (150 uM) for 10 min in a solution in which the NaCl had been replaced with choline chloride. Synaptosomes were then incubated with 22N3“ for 5 sec. Uptake of 22Na+ due to activation of voltage-dependent Na channels was determined as the difference of uptake in the presence and absence of tetrodotoxin (TTX; 1 14M). Values are the mean 1 SEM of 4 separate experiments. 112 fi [:1 0011 tails l "Na’ UPTAKE (nmollug protein E-OB) r. W 0 2 4 ['99] (I'IIII synaptosomal suspension) figure 4.1 1. Time course of “RV loading into synaptosomes. Synaptosomes were treated for 1 hr as described in methods. Prior to addition of radiolabel, synaptosomes were allowed to equilibrate at 37°C for 30 min. ”RH (20 uCi/ml) was added to the synaptosomal suspension and loading allowed to proceed for varying periods of time. After the specified time, 2 ml of cold quench solution was added to terminate loading. Uptake is expressed as nmol m3Rb“/ug protein. 113 E Con lgG o l s ‘ t‘ 39"“ “- 15 ' Serum 0 None only LEMS+lgG 98 ' __, Serum § 10 ' ‘ 'O'O'OV “E l % 335:3: LU . / + 5 / .0 / ‘I z 3 ° 2 3 Treatment figue 4.12. Effect of LEMS lgG on ”Rb“ efflux from rat forebrain synaptosomes. Following incubation and removal of residual extracellular ”RN, efflux of 1”Pb“ was measured by addition of depolarizing and nondepolarizing K+ solutions to the filter well. After 5 sec, cold quenching solution was added followed by suction filtration. Radioactivity in the filtrate and that remaining on the filter was determined. Efflux in low [K‘], solutions was subtracted from that in high [K‘]. solutions as depolarization-independent background. Effiux is expressed as % of total. Values are the mean i SEM of 4 separate experiments. 114 E. Discussion The aim of the present study was to test explicitly whether LEMS autoantibody affects nerve terminal Ca2+ channels and whether this action was direct and/or specific. In particular, we sought to determine if effects of LEMS lgG on nerve terminal function were related to a generalized membrane disturbance, and if not, whether the lgG recognized other voltage-dependent ion channels in axon terminals. This is important as nonspecific membrane alterations and/or alteration in function of other ion channels could result indirectly in Ca2+ channel dysfunction. Our results indicate that lgG from a patient with LEMS acts on presynaptic nerve terminal Ca2+ channels in the CNS to reduce depolarization- dependent uptake of 45Ca2+. The antibody from this patient appears to be specific for Ca2+ channels as it is unable to disrupt ion flow through voltage-dependent Na“ or K+ channels. These deficits are not due to decreases in synaptosomal viability nor to lgG-induced alterations in the synaptosomal membrane potential but appear to be a direct block of or alteration in voltage-dependent Ca2+ channel function. Passive transfer experiments demonstrated that the plasma from the three patients studied produced the electrophysiological characteristics associated with LEMS. Quantal content of the EPP was decreased when the nerve was stimulated at low frequencies of stimulation, but was facilitated at high frequencies of stimu- lation. MEPP amplitude was unaffected. These transmission characteristics are qualitatively similar whether the disease is transferred passively by plasma, crude 115 lg’s or purified lgG (Newsom-Davis et al., 1982; Lang et al., 1983; Kim, 1985; 1986; Lennon and Lambert, 1988). Passive transfer experiments were useful in determining that a presynaptic abnormality existed in LEMS. However, they represent an indirect measurement of nerve terminal Ca2+ channel activity as the postsynaptic response measured results from many processes ongoing in the presynaptic nerve terminal. Likewise, Meyer et al. (1986) and Login et al.(1987) showed that serum and serum fractions from patients with LEMS reduced ACh release from synaptosomes and growth hormone and prolactin release from rat anterior pituitary cells respectively. However, no measurement of ion channel activity was made (Meyer et al., 1986; Login et al., 1987). Still, others have shown that LEMS lgG reduced the function of somal Ca2+ channels in human soc (Roberts et al., 1985; De Azipurua et al., 1988a,b), rodent neuroblastoma X glioma hybrid (Peers et al., 1987) , bovine adrenal chromaffin (Kim and Neher, 1988), and rat pheochro- mocytoma (PC12) cells (Lang et al., 19893,b). It is quite possible, however, that the type of Ca2+ channel in the terminal may differ from that in the cell body (Nachshen and Blaustein, 1979; Suszkiw et al., 1989; Charlton and Augustine, 1990; Utchitel et al., 1992). Also, the binding of 01-0ng to somal Ca”+ channels has been used as an index of localization of LEMS autoantibodies (Sher et al., 1989), although «.1me does not block mammalian neuromuscular transmission (Anderson and Harvey, 1987; Protti et al., 1991) whereas the LEMS autoantibody clearly does (Lambert and Elmqvist, 1971; Lang et al., 1983; Kim, 1985:1986). 116 We demonstrated previously that serum and plasma from patients with LEMS reduced depolarization-dependent uptake of “Ca“ (10 sec) into synaptosomes (Hewett and Atchison, 1991; Chapter Three). In the present study the interaction of isolated LEMS lgG with nerve terminal Ca2+ channels thought to be involved in neurotransmitter release was examined directly. Reduction in depolarization-dependent uptake of 45Ca2+ during 1 sec was evident following 90 min incubation with LEMS lgG. This 'fast' phase of uptake (1 sec) into synapto- somes has been consistently correlated with neurotransmitter release from synaptosomes (Drapeau and Blaustein, 1983; fioor, 1983; Leslie et al., 1985). lgG from SCC patients lacking neurological symptoms did not reduce depolarization-dependent uptake of “Ca” into synaptosomes. Thus, the presence of a tumor alone is not sufficient to result in nerve terminal Ca2+ channel dysfunction. This is not unexpected, as many patients with SCC do not have the manifestation of LEMS. Nonetheless, interpretation of data obtained by using lgG from a LEMS patient with SCC is not complicated by actions of lg’s in the plasma of patients with SCC. + The decrease in “Ca2 uptake into synaptosomes was not a result of disruption of synaptosomal membrane intergrity. LDH is a cytosolic enzyme that can be released into the extrasynaptosomal fluid if the plasma membrane is per- meabilized. Its release following incubation with LEMS lgG was not increased over that of paired controls, so it is unlikely that the decrease in Ca2+ uptake was due to nonspecific membrane damage. It remains possible that the LEMS lgG causes 117 membrane damage in only a fraction of the terminals and the release of LDH from the whole population may not be sufficiently sensitive to detect this. However, in view of the lack of discrimination of the LEMS lgG for L- and N-type channels in other systems, and the magnitude of decreases seen with both serum/plasma (Hewett and Atchison, 1991) and lgG (present study) in synaptosomes, it seems unlikely that a specific population of terminals would be targeted. lgG-induced alterations in the synaptosomal membrane potential could reduce “Ca2+ uptake indirectly. lf lgG depolarized the synaptosomal membrane, a greater fraction of Ca2+ channels might be inactivated and unable to respond to subsequent K*-depolarization. Alternatively, hyperpolarization of the synaptosomal membrane would decrease uptake of Ca2+ by reducing the probability of opening of channels for a given depolarizing pulse of K“. Changes in the fluorescence of carbocyanine dyes as an indicator of alterations in membrane potentials have been studied in a variety of cell types and in synaptosomes (Blaustein and Goldring, 1975; Heinonen et al., 1985; Hare and Atchison, 1992). Multiple dye compart- mentation is avoided by elimination of the mitochondrial membrane potential by depolarization (Heinonen et al., 1985; Hare and Atchison, 1992). Neither plasma membrane depolarization nor hyperpolarization is involved in the reduction of 4‘r’Cafl uptake since LEMS lgG incubation caused no alterations in dye fluorescence following addition of KCI (40 mM). Thus, LEMS lgG induces a direct functional disruption of nerve terminal Ca2+ channels rather than indirectly by altering the membrane potential. 113 To determine if the LEMS antibody was specific for Ca2+ channels, we examined its potential to interact with other voltage-dependent channels. If the antibody were not specific for Caz+ channels, it may indicate the lgG could recognize a portion of the channel molecule common to all ion channels. This possibility had not been previously tested directly. Synaptosomes contain voltage-dependent Na+ channels. Maximum uptake of 22Na+ into synaptosomes occurs following 10 min incubation with 150 uM VER. VER is used to activate synaptosomal Na+ channels as these channels inactivate rapidly in response to K*—depolarization (Kreuger and Blaustein, 1980; Tankum and Catterall, 1981). There is considerable homology between the a subunit of the voltage- dependent Na+ channel and or1 subunit of the dihydropyridine (DHP)-sensitive (L- type) voltage-dependent Caz+ channel (Ellis et al., 1988). LEMS lgG affects L-type Ca2+ channels in several cell lines (Login et al., 1987; Peers et al., 1987; Kim and Neher, 1988). Since the 0 subunits contain the binding site for pharmacological antagonists, are the pore forming subunits and are indispensable for the function of the channeI(Catterall, 1988), they are potential candidates as antigenic determinants for autoantibodies. Therefore, LEMS antibodies directed against the a, subunit of the Ca2+ channel might cross-react with the a subunit of Na“ channels as well. Incubation of synaptosomes with concentrations of LEMS lgG that significantly altered uptake of “Ca2+ into K"-depolarized synaptosomes had no effect on uptake of ”Na“ into veratridine-depolarized synaptosomes. This is 1 19 consistent with the work of Pascuzzi and Kim (1988) in which functional in vivo measurements of nerve conduction velocity made from animals chronically treated with LEMS lgG demonstrated that conduction was normal. In some cells, Ca2+ and K“ channels may be structurally similar (Hume, 1985; Nerbonne et al., 1987). Also, some patients with MG, another autoimmune, neuromuscular disease, have antibodies against a p-bungarotoxin binding protein (Lu et al., 1991). This protein has been demonstrated to belong to a class of voltage-gated K+ channels (Rehn and Betz, 1984). If patients with LEMS have a similar antibody, alterations in K+ channel conductance could indirectly alter Ca2+ channel conductance. Treatment of synaptosomes with serum or serum and lgG did not alter the loading of “RN into synaptosomes by the Na“/K+ ATPase. Likewise, concentrations of LEMS lgG that significantly altered uptake of 4‘SCaz+ into K"—depolarized synaptosomes had no effect on efflux of “RN following K‘- depolarization. This provides further evidence that the LEMS autoantibody recognizes an epitope which is specific for voltage-dependent Ca2+ channels. In summary, LEMS IgG from Patient 3 acts on central nerve terminal Ca2+ channels to reduce depolarization-dependent uptake of 45Ca“. The antibody appears specific for Ca2+ channels as it is unable to disrupt ion flow through voltage-dependent Na+ or K“ channels. These deficits are not due to indirect effects on membrane integrity or synaptosomal membrane potential but appear to result from a direct block of or alteration in Ca2+ channel function. The preferential action of the LEMS antibody for nerve terminal Ca2+ channels suggests that there 120 t is no common determinant or epitope on different voltage—gated channel proteins to which the antibody can bind. We cannot exclude, however, the possibility that some patients with LEMS may synthesize antibodies with different specificities. These data also suggest antigenic similarities exist between somatic motor nerve terminal and central nerve terminal Ca2+ channels. These findings are important as the existence of autoantibodies which are specific for a component of the neurosecretory process may serve as molecular tools for studying the steps involved in quantal transmitter release both in the peripheral and central nervous system. CHAPTER FIVE DISRUPTION OF SYNAPTOSOMAL 662+ CHANNELS BY LAMBERT-EATON MYATHENIC SYNDROME IMMUNOGLOBULIN: SERUM DEPENDENCY 121 1 22 A. Summary Chapter Four demonstrated that in the presence of control human serum, a2+ into immunoglobulins isolated from a patient with LEMS reduced flux of C isolated nerve terminals during depolarization. The objective of the present study was to determine the role of serum in reducing uptake of “Ca2+ into rat brain synaptosomes by LEMS lgG. Depolarization-dependent uptake of “0a“ through voltage-gated Ca2+ channels was determined using synaptosomes incubated with control (disease-free) and LEMS lgG with or without control human serum. In the absence of human serum, LEMS lgG did not reduce uptake of "“Ca2+ into synaptosomes. However, in the presence of control human serum (10% of total incubation volume), “Ca2+ uptake was reduced significantly by LEMS lgG (2 and 4 mg /ml), but not by lgG from disease-free patients or by 10% (v/v) control human serum alone. This concentration of serum was found to be optimal; higher concentrations produced significant reductions in Caz+ uptake, whereas at lower concentrations the serum / lgG combination was ineffective. The depressant effect of serum alone on 45C3“ uptake was mimicked by bovine serum albumin (BSA) suggesting that deficits in “Ca2+ uptake produced by serum were the result of increased protein binding of the radiolabel. Heat-inactivating the serum abolished its ability to interact with LEMS lgG to depress 45Ca2+ uptake. This suggested a role for complement in this effect. To test whether the membrane attack complex (MAC) or the alternative pathway of complement (APO) contributed to the alteration in Ca” channel function, “Ca2+ uptake into synaptosomes was measured 123 following incubation with LEMS IgG and C5- or Factor B- deficient serum, respectively. LEMS lgG reduced K*-stimulated uptake in both cases. Thus, neither the MAC nor the APC contributed to disruption of Ca2+ channel function by LEMS lgG in synaptosomes. In contrast, incubation of synaptosomes with LEMS IgG and CS-deficient serum prevented the reduction in “Ca2+ uptake suggesting that the C3 component of the complement pathway contributes to Ca2+ channel dysfunction in synaptosomes following incubation with serum and LEMS lgG. B. Introduction Autoantibodies, such as those in LEMS, that interfere with the release process may either interrupt Ca2+ flux into the terminal or some intraterminal event critical to transmitter release. Results from several studies indicate that LEMS autoantibodies interact with voltage-dependent Ca2+ channels in cell lines of diverse origin (Peers et al., 1987; De Azipurua, 1988a; Kim and Neher, 1988). LEMS serum also reduces depolarization-dependent Ca2+ entry into and ACh release from isolated nerve terminals of rat brain (Meyer et al., 1986; Hewett and Atchison, 1991). Data in Chapter Four demonstrate that in the presence of control human serum, lgG isolated from a patient with LEMS reduced Ca2+ flux into synaptosomes as well. Activation of the serum component complement plays a central role in mediating postsynaptic membrane damage in animal models of the autoimmune neuromuscular disorder MG, as well as in the human disease state (Engel et al., 1977; Sahashi et al., 1978). Attempts to clarify the participation of complement in 124 the pathogenesis of LEMS has so far yielded negative results (Prior er al., 1987; Lennon and Lambert, 1988). As it is still possible that complement activation may be important in LEMS, the present study was undertaken to assess the role of serum in the pathogenesis of Ca2+ channel dysfunction induced by LEMS lgG in isolated nerve terminals from mammalian CNS. C. Materials and Methods 1.Materlals N-methyI-d-glucamine, 05-, C3-and Factor B-deficient human sera were purchased from Sigma Chemical Co. (St. Louis, MO). HVLP (0.45 0M pore size; 25 mm diameter) filters were purchased from Millipore (Bedford, MA). HEPES was purchased from United States Biochemical Corp. (Cleveland, OH). The radiolabel “Ca" (15-50 mCi/mg) was purchased from New England Nuclear (Boston, MA). Disease-free control human plasma was purchased from the American Red Cross (Lansing, MI). Control human serum was donated by a healthy volunteer. Plasma from a patient with the clinical and electromyographic features of LEMS was donated by Mark Glasberg, M.D., Henry Ford Hospital, Detroit, MI. 2. Immunoglobulln Isolation Immunoglobulln from control and LEMS patients was isolated and quantitated as described in detail in Chapter Four. The LEMS lgG used in these studies was identical to that used in Chapter Four as well. 125 3.Preparatlon of Synaptosomes Synaptosomes were prepared according to a modification of the method of Gray and Whittaker (1962), as described in detail in Chapter Three. Equal aliquots of the synaptosomal suspension were incubated at 4°C under 100% 02 for 60 min with either no serum or 10% (v/v) control, 05-, 03- or Factor B-deficient serum and varying concentrations of lgG from patients with no known neurological or immunological illness (control) or LEMS. Alternatively, one concentration of lgG and varying concentrations of human serum were added to the suspension to test effects of serum on 4"’Ca2+ uptake into synaptosomes. Prior to uptake measure- ments, synaptosomes were allowed to equilibrate for 30 min at 37°C. 4. Solutions and Measurement of “Ca“ Uptake The solutions for and the measurement of “Ca“ are described in detail in Chapter Three. Uptake is expressed as nmole 45Ca2+/ttg protein or as a percentage of disease-free control. 5. Statistical Analysis Data were analyzed by randomized complete block analysis of variance followed by Dunnett’s t-test or Students t-test for paired samples. Differences were considered significant at p < 0.05. 126 D. Results 1. LEMS lgG In the Presence and Absence of Control Human Serum Uptake of 4”Ca” into synaptosomes during 1 sec of K+-depolarization was unaffected by 90 min incubation with LEMS lgG alone (fig. 5.1). However, in the presence of control human serum (10% of the total incubation volume) 2 and 4 mg/ml LEMS lgG inhibited uptake by 28.3 t 10.0% and 39.1 t 21.3%, respectively, as compared to their paired controls incubated with equivalent concentrations of control (disease-free) lgG (fig. 5.1). This concentration of serum (10 %; v/v) is the concentration used in standard tissue culture procedures. 2. Control Human Serum Since the presence of serum with LEMS lgG is important in reducing “Ca“ uptake, the effect of control (disease-free) human serum on uptake of “Ca“ into synaptosomes was examined (fig. 5.2). Synaptosomes were incubated with 2 mg/ml IgG from disease-free patients and varying concentrations of control (disease-free) human serum. A concentration-dependent suppression of uptake is evident, with uptake at concentrations 2 15 % (v/v) being significantly lower than those in the absence of serum. No statistically significant reduction in uptake was observed with a serum concentration of 10 % (v/v). Similar results were obtained when bovine serum albumin was substituted for serum (fig. 5.3). Synaptosomes were incubated with 2 mg/ml control or LEMS lgG and varying concentrations of disease-free human serum to determine the optimal concentration of serum needed in the reaction mixture. Again, suppression of 12 ‘1 No Control serum ‘ 1 g ‘5 . —Jfi * _i_. * t’g’ °°f 2:3 .0. / % 0' A // 2 [LEMSIgG] 4 (molml synaptosomal suspenslon) figure 5.1. Effect of LEMS lgG in the presence and absence of control human serum on uptake of 45Ca2+ into rat forebrain synaptosomes. Synaptosomes were incubated for 60 min at 4°C with 2 and 4 mg/ml LEMS or control lgG 1- control human serum. Prior to the initiation of uptake, synaptosomal suspensions were equilibrated at 37°C for 30 min. Uptake in 5 mM K” was subtracted from that in 42.5 mM K+ as depolarization-independent background. 45Ca2+ uptake is expressed as % of paired control, control being synaptosomes incubated with equivalent concentrations of control lgG. All values are the mean 1 SEM of 5 experiments. Asterisks indicate values that are significantly lower than their respective paired control (p < 0.05). 128 1 5 P g; X III ‘::{E 10 . , ‘* v v ‘ i- 4 ‘%~9 ‘3? ”as 5%” ,.- l . 9 9 , :3 Q”! wuw' ”3% iVVV E: ‘Vfl’ 0%0' Qflfll fifi’ 5 r so. 1941 poor 040 :3 cos flfifii fififi "fit 13 c} (oat flfifii one: ””5 90-.940 -oos too sot 1404 too« 449 E5 - .avi flfifli §V5t ”V5 3 so. 900 on, too ~n ‘Qfl’ 0%”! fifi? QR? ”MP ififlfi 29%5 «Q9 0 A A A . .A.A. ’A’A’A‘ .A.A.A‘ ’A’A’ 0 5 10 15 20 25 PERCENT SERUM figure 5.2. Effect of control human serum on “"’Ca2+ uptake into rat forebrain synaptosomes during 1 sec depolarization by 41.25 mM KCI. Synaptosomes were incubated with 2 mg/ml control lgG and varying concentrations of control human serum for 60 min at 4°C. Prior to the initiation of uptake, synaptosomes were allowed to equilibrate for 30 min at 37°C. Uptake in 5 mM K+ was subtracted from uptake in 42.5 mM K“ as depolarization-independent background. 45Ca2+ uptake is expressed as nmol/pg protein. All values are the mean : SEM of 3 experiments. Asterisks indicate values that are significantly lower than uptake into synaptosomes which had been incubated with no serum. 129 * a 10 15 20 25 ERCENT BSA ""Ca‘a UPTAKE s\\\\\\\\\\\\l~ W W” (nmollug protein E-OS) o W ' '0‘” figure 5.3. Effect of BSA on “Ca2+ uptake into rat forebrain synaptosomes during 1 sec of depolarization. A BSA solution containing the approximate protein concentration of serum was made in deionized water. Synaptosomes were incubated with 2 mg/ml control lgG and varying concentrations of BSA as in figure 5.2. All values are the mean 1 SEM of 5 experiments. Asterisks indicate values that are significantly lower than uptake into synaptosomes which had been incubated with no BSA (p < 0.05). 130 uptake by serum alone is evident (fig. 5.4). “Ca2+ uptake was reduced significantly only by the combination of 2mg/ml LEMS lgG and 10% (v/v) serum when compared to its paired control (2 mg/ml control lgG and 10% serum). 3. LEMS lgG and Complement Control human serum was incubated for 45 min at 57°C to inactivate complement. This prevents the reduction in uptake of “Ca2+ into the synap- tosomes by LEMS lgG (2 and 4 mg /ml) and control human serum (10%) (fig. 5.5). Two crucial components of the complement cascade are 03 and Cs. Removal of CS prevents the assembly of the membrane attack complex (MAC). The ability of LEMS lgG and control human serum to reduce “Ca2+ uptake into synaptosomes was unaltered when control human serum was replaced with C5- deficient human serum (Fig. 5.6). Conversely, removal of the C3 component of complement from control human serum prevented the action of the LEMS lgG (fig. 5.7). Incubation of synaptosomes with 2 and 4 mg/ml of LEMS lgG and C3- deflcient serum (10 %; v/v) abolished the reduction in uptake produced by the same concentrations of LEMS lgG in the presence of control human serum or OS- deficient serum. To determine if the alternative pathway of complement (APC) played a role in mediating the dysfunction of synaptosomal Ca2+ channels by LEMS lgG and control human serum, synaptosomes were incubated with 2 and 4 mg/ml control and LEMS lgG in the presence of either control human serum or Factor B-deficient serum (fig. 5.8). Factor B is essential for activation of the APO. In the presence of 131 A '1— CONTROL lgG * --o-- LEMS igc “Ca '* UPTAKE (nmol/pg protsln E-06) ° 0 b 1.0 I 20 PERCENT SERUM Figure 5.4. Effect of control and LEMS £962 and various concentrations of control human serum on 1 sec K*-stimulated Caz“ uptake into rat forebrain synapto- somes. Synaptosomes were incubated with 2 mg/ml control and LEMS IgG and varying concentrations of control human serum as described in Figure 5. 2. All values are the mean + SEM of 5 experiments. ‘;", indicates values that are significantly different from their paired control. (b), indicates values that are significantly different from uptake into synaptosomes that were incubated with no serum. (p < 0.05) 132 1:1 Control m Heated seruri serLI'n 120- LU _ K: