UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Selective tuning of L-type voltage-sensitive calcium channels to strong synaptic activity Liu, Zhi 2002

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

Item Metadata


831-ubc_2002-0156_fixed.pdf [ 3.68MB ]
JSON: 831-1.0090442.json
JSON-LD: 831-1.0090442-ld.json
RDF/XML (Pretty): 831-1.0090442-rdf.xml
RDF/JSON: 831-1.0090442-rdf.json
Turtle: 831-1.0090442-turtle.txt
N-Triples: 831-1.0090442-rdf-ntriples.txt
Original Record: 831-1.0090442-source.json
Full Text

Full Text

Selective tuning of L-type voltage-sensitive calcium channels to strong synaptic activity By Zhi Liu B.Sc. (Medicine), Hebei Medical College, CHINA, 1986  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE In THE FACULTY OF GRADUATE STUDIES (Neuroscience Graduate Program)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April 2002 © Zhi Liu, 2002  In  presenting  degree  at  this  the  thesis  in  partial  fulfilment  of  University  of  British  Columbia,  I agree  freely available for copying  of  department publication  this or of  reference and  thesis by  this  for  his thesis  study.  scholarly  or  her  for  I further  purposes  of  •/  The University of British Vancouver, Canada  Date  DE-6 (2/88)  that  be  It  gain shall not  permission.  Department  requirements  agree  may  representatives.  financial  the  / Columbia  Avr^l 2?. ^ X  that  the  by  understood be  an  advanced  Library shall  permission  granted  is  for  for  the that  allowed without  make  it  extensive  head  of  my  copying  or  my  written  Abstract , L-type voltage-sensitive Ca 2+ channels VSCCs are important in  mediating  activity-dependent gene expression that may contribute to neuronal plasticity. Evidence shows that the L-type VSCC has slow activation kinetics making it function as a filter'for action potentials and therefore respond more effectively to the long-lasting EPSP waveform. However, these results are derived from experiments  performed at room temperature and need to be tested  at  physiological temperature. Moreover, whether the square pulse analysis is legitimate for calcium channel behavior in the complex synaptic depolarization - waveform  remains unclear.  Using existing waveforms  modeled from  the  'NEURON' simulation program, we tested our hypothesis that at physiological temperature, slow activation kinetics underlie the L-type VSCCs' preference for EPSP type  waveforms  over  action  potential  type  waveforms.  Different  simulated waveforms were used as voltage clamp commands to HEK cells with heterologously expressed calcium channels, along with regular step pulses for basic channel property analysis. Our results show that activation kinetics are drastically accelerated compared to that obtained at room temperature such that all three channel types, L-, P/Q- and N-type channels are equally activated to full scale by action potentials. The L-type VSCCs mediate a distinctively larger total calcium influx in response to strong synaptic inputs in comparison with the P/Q- and N-type channels. Further analysis indicates that inactivation rather than activation characteristics underlies the distinctive Ltype calcium current response to strong synaptic inputs at  physiological  temperature. In summary, this inactivation-dominated behavior tunes the Ltype calcium channels to the strong synaptic inputs that may be important in triggering gene expression.  11  Table of Content Abstract  ii  List of Tables  v  List of Figures  v  Acknowledgements  vi  CHAPTER 1  INTRODUCTION  1  OVERVIEW OF CALCIUM CHANNEL  l  Calcium as a universal second messenger for neurons Categorization of calcium channels Structure of calcium channels Biophysical properties of calcium channels... Calcium channel regulation Function of calcium channels  1 1 4 7 8 9  CALCIUM CHANNELS AND GENE EXPRESSION  Gene expression in learning and memory L-type channels involved in gene expression Mechanisms underlying the selective tuning of L-type channels to relevant in gene transcription  9  9 11 activities  RESEARCH HYPOTHESIS PROJECT OUTLINE  CHAPTER 2  13 16  METHODS...  TRANSIENT TRANSFECTION OF HEK  17 293  CELLS  Appropriate medium condition DNA handling Transient calcium phosphate transfection ACUTE ISOLATION OF HIPPOCAMPAL CA1  NEURONS  Electrophysiology CHAPTER 3  12  RESULT  17  17 18 18 20  21 26  CHARACTERIZATION OF THE BIOPHYSICAL PROPERTIES OF CALCIUM CHANNELS AT ROOM TEMPERATURE WITH SQUARE PULSE  Activation Inactivation CALCIUM CHANNEL CHARACTERIZATION AT PHYSIOLOGICAL TEMPERATURE WITH SQUARE PULSE WAVEFORM ANALYSIS AT ROOM TEMPERATURE WAVEFORM ANALYSIS AT PHYSIOLOGICAL TEMPERATURE ANALYSIS OF ACTIVATION KINETICS AT PHYSIOLOGICAL TEMPERATURE INACTIVATION ANALYSIS AT PHYSIOLOGICAL TEMPERATURE  Inactivation revealed by spiking attenuation Channel recovery from spiking arrest Effect of inter-spike potential on inactivation Modulation of biophysical properties by p subunit CALCIUM CURRENT INDUCED BY HIPPOCAMPUS-DERIVED FIRING WAVEFORM CALCIUM CHANNELS I N ACUTELY ISOLATED HIPPOCAMPAL NEURONS  CHAPTER 4  26  26 28  DISCUSSION  29 51 53 56 58  59 60 61 62 64 64  67  in  APPROPRIATE EXPERIMENTAL CONDITIONS FOR CALCIUM CHANNEL INVESTIGATION  67  L-TYPE CALCIUM CHANNELS ARE TUNED TO STRONG SYNAPTIC ACTIVITIES ..:  72  ACTIVATION KINETICS ARE NOT ATTRIBUTABLE FOR DISTINCTION OF CALCIUM CHANNEL BEHAVIOR 76 INACTIVATION I S CRITICAL I N FUNCTIONAL DIFFERENTIATION OF CALCIUM CHANNELS  Residual calcium current Inactivation revealed from spiking attenuation Channel availability at the end of synaptic waveforms ; Inter-spike potential is important for voltage-dependent calcium channel inactivation /32A subunit reverses inactivation in the P/Q-type channels SUMMARY  ;  77  78 79 80 82 83 85  FUTURE DIRECTIONS  86  REFERENCE:  88  iv  List of Tables Table 1. Subunit composition and function of Ca 2+ cannel types Table 2. Basic biophysical factors revealed from step pulse, Table 3. Calcium current response to simulated waveforms at physiological temperature : Table 4. Parameters for calcium channels with p2A subunit  2 28 54 63  List of Figures Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure  1. 2. 3. 4.  1 : A schematic illustration of a typical calcium channel structure 6 A schematic outline of the data generation 23 Calcium currents in response to a 50 ms step depolarization 32 Current-voltage (IV) relationship of the L-, P/Q- and N-type calcium channels 33 5. An example showing the effect of surface potential on N-type calcium channel gating property 34 6. Effect of temperature on the response of calcium channels to mock APs with varied duration 35 7. Calcium currents in response to the repetitive firing and plateau waveform at room temperature 36 8. Example traces of calcium currents in response to different stimulus waveforms at physiological temperature 37 9. Group data showing calcium currents in response to different stimulation waveforms at physiological temperature ' 38 10. Calcium currents in response to theta burst firing waveform 39 1 1 . Calcium currents in response to one second step depolarization pulse 40 12. Paradoxical relation between rise time and relative AP peak amplitude at physiological temperature 41 13. Tail current constitutes the main body of the calcium current response to AP 42 14. Inactivated fraction of calcium current in response to the step pulse and the simulated waveform '. 43 15. Inactivation reflected by the amplitude of an end pulse in the simulated waveforms •: 44 16. Inactivation reflected by the end-pulse in calcium channels with different b subunits 45 17. Voltage-dependence of P/Q-type inactivation 46 18. Calcium current in response to the hippocampus-derived waveform 47 19. Example of temperature effect on activation rate and calcium peak currents 48 20. Calcium currents in hippocampal CA1 neurons 49 2 1 . Temperature effect on calcium current is reversible 50  v  Acknowledgements First  I  would  like to  express  my  sincere  gratitude  to  my  supervisor Dr. Timothy Hugh Murphy for his comprehensive support during the last two and half years. His guidance and help allowed me to overcome my research and academic difficulties. I could have  not  finished  and  this  project  in  this  period  without  his  patience  encouragement. I would also like to thank Ms. Jihong Ren for her hard work in simulating the synaptic waveforms, without which the project would not be possible.  I want to thank my committee members, Dr. Steven Kehl, Dr. James Mclarnon, Dr. Terry Snutch and Dr. Yutian Wang for their active suggestions for the project. During the project, I owed a great deal to Dr. Oliver Prange, the previous Ph.D student in our lab for his w a r m hearted help in many technical aspects during my experiments. I would like to thank our laboratory technician Jessica Yu, Lei Jiang and Tao Lou, without the help from whom my project would be much delayed. During the writing, I got tremendous help from my colleagues, Herman Brian Fernandes, Bo Li, Andy Shi, Catherine Gauthier-Campbell, Mannie Fan for revising my English. I will not forget the good time I spent in our laboratory with those 'night owls', Yo Otsu, Shawn Zhang and other buddies.  Finally, special thanks go to my wife, Irene Haiyan Li who has shared  every  piece of joy  and  distress  and  continues  delivering  encouragement, unconditionally. Thank you.  VI  Chapter 1  Introduction  Overview of calcium channel  Calcium as a universal  second messenger  for  neurons  Calcium is the essential signaling ion in almost any biological organism. Through the calcium channel, a protein complex that allows divalent ion flux upon a conformational change, calcium triggers a myriad of cellular process (Hille, 1992) such as muscle contraction, cell division, heart beat regulation, hormone and transmitter secretion as well as some forms of long term changes in brain structure (16). In the central nervous system, majority of crucial cellular processes, such as neuronal growth, programmed cell death, inter-neuron communication and neuronal plasticity, are dependent upon calcium influx via various calcium channels (3, 40, 58), which will become the focus of our study.  Categorization  of calcium  channels  Membrane proteins that are permeable to calcium can be either activated by membrane voltage change, by a ligand or a combination of both. The former is termed voltage-gated calcium channels and latter is termed ligand-gated calcium channels, mainly referring to NMDA receptors. Voltagegated (alternatively named voltage-sensitive or voltage-dependent) calcium channels can be further divided into high-voltage activated (HVA) calcium  l  channels and low-voltage activated (LVA) calcium channels. The LVA calcium channel type is solely comprised of T-type calcium channels while the HVA calcium channels include three major types, L-, P/Q- and N-type VSCCs. There is some recent innovation of calcium channel nomenclature according to their structural and functional similarities (27) as summarized in table 1. In this dissertation, traditional nomenclature for the voltage-gated calcium channels will be used.  Table 1 . Subunit composition and function of Ca Primary localizations  of CCIY subunits  Cav1.1  Ca2+ current type L  Skeletal muscle  Cav1.2  L  Cardiac muscle Endocrine cells Neurons  Ca2* channel  Cav1.3  L  Cav1.4 Cav2.1  L P/Q  Cav2.2  N  Endocrine cells Neurons Retina Nerve terminals Dendrites Nerve terminals Dendrites Cell bodies Dendrites Nerve terminals Cardiac muscle Skeletal muscle Neurons  Previous name  cannel types (13)  Specific blocker  Functions  <X1S  DHPs  Excitation-contraction coupling Calcium homeostasis Gene regulation  die  DHPs  Excitation-contraction coupling  co-Agatoxin  Hormone secretion Gene regulation Hormone secretion Gene regulation Tonic neurotransmitter release Neurotransmittler release  CC1B  oCTx-GVIA  Dendritic Ca2+ transients Neurotransmitter release  <X1E  None  Dendritic Ca2+ transients Ca2+-dependent action potentials  aiG  None  Neurotransmitter release Repetitive ring  CC1D  DHPs  (X1F  a-iA  Cav2.3  R  Cav3.1  T  Cav3.2  T  Cardiac muscle Neurons  aiH  None  Repetitive ring  Cav3.3  T  Neurons  an  None  Repetitive ring  Functional classification of HVA channels are primarily based on their pharmacological profiles as well as their biophysical properties. L-type calcium channels, originally characterized in skeletal muscle-derived calcium channels  was named from their large single channel conductance, ~25 pS (30, 70). The defining character of L-type channels is their sensitivity to dihydropyridines (DHPs  such  as  nifidipine  or  PN-200)  blockade. The  widely  accepted  characteristics of L-type calcium channels were obtained in comparison with Ntype channels since P/Q-type channels were discovered at a much later stage. The L-type calcium channels have relatively slow activation kinetics and exhibit little inactivation (14, 66). In contrast, N-type calcium channels (initially defined as Non-L HVA channels) was potently blocked by co-conotoxin GVIA, a peptide isolated from the venom of the cone snail (30, 70) and have a relatively fast activation and fast inactivation with higher activation threshold (~20 mV higher) (93). Subsequently another type of HVA calcium channels was discovered in cerebellar Purkinje neurons and so named P-channels. Pchannels are potently blocked by the spider toxin co-Agotoxin IVA (57). Soon after another calcium channel type was found with less potency from coAgotoxin IVA blockade and named Q-channel due to its similarity to P-type channels (97). The distinction between P- and Q- channels has been difficult to establish in many neurons. Therefore these channels are referred to as P/Q channels unless high-affinity block by co-Agatoxin IVA is unequivocal. The P/Qtype channels are also blocked in a nearly irreversible manner by co-conotoxin MVIIC (97). This toxin also blocks INHtype channels reversibly and spares other HVA and LVA channels. In neurons, there is an HVA component that is resistant to all known calcium channel blockers, named R-type channels. This R-type channels have a rapid inactivation profile and activate at a somewhat more negative threshold (97). The functional role of R-type channels remains enigmatic. Therefore, in neurons non-L-type calcium channels represents a 3  rather mixed cocktail of calcium channels, P-type, Q-type, N-type, R-type and T-type channels.  Structure  of calcium  channels  A complete calcium channel complex contains four subunits, a l , a.28, p and y subunit. The a l  subunit has a molecular weight of 190 kDa and  constitutes the pore-forming structure, responsible for the process of ionic permeation as reviewed by Catterall (13). It is comprised of about 2000 amino acid residues with sequence and structure similar to that of sodium channel a subunit (84). The amino acid sequence is organized in f o u r repeated domains (I to IV), each of which contains six transmembrane segments ( S I to S6). The four domains are spatially arranged such that the S5 and S6 form the pore lining of the channel while the S4 serves as the voltage sensor for activation with positively charged residues exposed to extra- or intracellular solution, providing the channel the gating properties (94). Expression of the a l subunit alone is sufficient to produce functional calcium channels, but with  low  expression level and abnormal kinetics and voltage-dependence of the calcium current (67). The a25 subunit is a disulfide-linked dimer with molecular mass of 170 kDa and is heavily glycosylated both on a2 and 8, with 3 0 % of its molecular mass composed of carbohydrates. Biosynthesis studies indicate that a28 subunit is an extracellular membrane protein attached to the  membrane  through disulfide linkage to the 8 subunit (34).  4  B subunit is an intracellular protein with molecular weight of 55 kDa. So far four different B subunits (pi-p4) (21) have been identified with distinct distribution  and  biophysical  profile.  In  situ  hybridization  and  immonohistochemistry study indicate in hippocampus, all p subunit types were observed with relatively low expression in p i and p2. p4 subunits only exist in a small portion of interneurons. p3 subunits are enriched in hippocampus and predominantly locate in mossy fibers and their terminal region in the CA3 stratum lucidum (20). Co-expression of a2S subunit and especially the p subunit enhances the level of expression and confers more normal gating properties (48, 76). More importantly, p subunit plays a crucial role channel inactivation process. In particular, plB and p3 accelerate the inactivation while p2 subunit features substantial slowing of inactivation (79). plB has a typical acceleration of calcium channel inactivation and is enriched in forebrain cortex, making it a commonly used auxiliary subunit in heterologous systems (52). y subunit is a transmembrane protein with molecular weight of 33 kDa (82). Co-expression or deprivation of a28 (42) and y subunits has little effect on channel functions (49).  5  domain:  Figure 1 . A schematic illustration of a typical calcium channel structure. A complete calcium complex contains a principal transmembrane a l subunit, which is the pore-forming subunit in association with a disulfide-linked a28 dimer, an intracellular p subunit, and a transmembrane y subunit. The amino acid sequence of a l subunit is organized in four repeated domains (I to IV), each of which contains six transmembrane segments (SI to S6). The four domains are spatially arranged such that the S5 and S6 form the pore lining of the channel while the S4 serves as the voltage sensor for activation with positively charged residues exposed to extra- or intracellular solution, providing the channel the gating properties.  6  Biophysical  properties  of calcium  channels  HVA calcium channels activate rapidly upon depolarization and are closed by hyperpolarization, a process called depolarization. Once activated, the HVA calcium channels in a few milliseconds become inactivated manifested by a gradually deteriorated calcium current during the depolarization. Upon repolarization, a calcium current can be observed, termed 'tail current'. Tail current is generated due to three reasons. First, the channels are open. Upon repolarization (membrane potential returning from depolarized level back to hyperpolarized potential), the channels remain open for about 1 millisecond (40). Second, the full driving force for calcium influx is restored upon repolarization. Third, the channel inactivation is not saturated. Since the depolarization period is often variable, the tail current is predominantly affected by inactivation status of the channels (46). To study the activation kinetics, a maximal tail current is usually generated with a series of short pulses within which the calcium channel inactivation has not developed. Different HVA calcium channels have distinct biophysical kinetics. In particular, L-type calcium channels feature slow activation and little inactivation in comparison with N-type calcium channels in according to  experiments  performed at room temperature (14, 66). There inactivation  are two forms and the  other  of inactivation. One is is voltage-dependent  calcium-dependent  inactivation.  Calcium-  dependent inactivation has been thoroughly characterized for the L-type channel, via which calcium entry decays in a calcium-dependent fashion (31). Calcium entry in response to a depolarization induces formation of bound-form 7  of CaM, which interacts with calcium  channel a l subunit at IQ motif in the  cytoplasmic tail, leading to partial closure of the channel pore (44, 98). In HVA calcium  channels,  L-type  inactivation  is mainly  governed  by calcium-  dependent inactivation while N- and P/Q-type channel type of inactivation is predominantly voltage-dependent (31, 36).  Calcium channel  regulation  HVA calcium channels can be regulated. A general rule is that L-type calcium channels have a relatively low level of basal activity, which can be readily upregulated by a variety of mechanisms (23). Typically, L-type channel conductance is enhanced by a prepulse depolarization reminiscent of the situation when NMDA receptors deliver the EPSP to somatic L-type calcium channels in response to a repeated presynaptic stimulation (8). When serving to trigger transmitter release at some specialized synapses, L-type calcium channels can also be inhibited by activation of GABA_B-like receptors (38, 39). In contrast, N- and P/Q-type calcium channels normally have a high open probability upon depolarization, which can be downregulated by a variety of mechanisms (45). Briefly, the hormone or transmitter induced activation of Gprotein-coupled receptors results in the reduction of calcium peak current. G protein Py subunit interacts and binds to the intracellular loop between domain I and I I of calcium channel a l subunit. According to the model proposed by Yue et al (65), the G-protein Py complex preferentially binds to the deeply closed channel state (reluctant to open). A large depolarization prepulse can dissociate the Py subunits from the calcium channel, making the transition of calcium channel from 'reluctant' state to 'willing' to open state. 8  Function  of calcium  channels  In this section focus will be on calcium channels in central nervous system. There is abundant evidence showing that both N- and P/Q-type calcium channels are localized at nerve terminals (91, 92). Despite the difference in biophysical profile between P/Q- and N-type calcium channels (19), it is widely believed that N- and P/Q-type mediate the calcium influx into neuronal terminal required for transmitter release. In contrast, L-type calcium channels' were found localized primarily in cell bodies and proximal dendrites (2, 90). Consistent with this localization, Ltype calcium channels have a dominant role in calcium influx into the cell bodies of hippocampal CA3 neurons (26). Then the question rises that what neuronal functions might be regulated by calcium influx into cell body and dendrites via L-type calcium channels? A large body of evidence indicates that L-type calcium channels have a crucial role in regulation of gene transcription (58). The following section will address the role of L-type calcium channel in calcium-dependent regulation of gene transcription.  Calcium channels and gene expression  Gene expression in learning and memory Mechanisms of learning and memory have been investigated for centuries. At the cellular level, long-term potentiation (LTP) has been proposed as a model for learning and memory (7). Upon a strong presynaptic 9  stimulation, operationally a one-second long 100 Hz pulse, the amplitude of the excitatory postsynaptic response can be enhanced for hours, or even for days or weeks, providing a viable substrate for induction and storage of memory. Likewise, low frequency stimulation leads to persistent weakening of synaptic strength, also known as long-term depression (LTD) (5). Along with functional changes at synapses, a series of structural modification follows to consolidate the postsynaptic response.  In response to a brief tetanus, the  postsynaptic neuron exhibits an early phase LTP that lasts only for hours and a late phase LTP that lasts for days (35, 6 1 , 85). Early phase LTP involves only AMPA receptor modifications (phosphorylation and/or upregulation), while in late phase LTP, new protein synthesis occurs following induction of gene expression in response to elevated calcium levels in the nucleus (60). It is this late phase LTP during which the neuron is provided with the necessary structural substrates for memory maintenance. Early evidence has suggested that light deprivation in mice caused a reversible reduction in the number of dendritic spines (32, 87, 88). Similarly, an increase in spine density is observed after visual stimulation (63). Other sophisticated environmental manipulations, such as social isolation have been shown to alter spine morphology (18). Further investigation shows that mice with inactivated immediate-early genes (IEGs), such as c-fos and FosB are profoundly deficient in some forms of neuronal adaptive responses (11, 28, 89). All of these findings provide support for the involvement of protein synthesis in the emergence of new structure or cellular machinery required for long-term neuronal plasticity.  10  L-type  channels  involved  in gene  expression  Tremendous effort has been devoted to examine the signaling pathways that link synaptic activity to nuclear gene expression. There is a strong consensus that calcium is essential for activity-dependent gene expression. In Aplysia for instance, microdomains of postsynaptic Ca2+ near the plasma membrane can activate type I adenylyl cyclase to produce cyclic AMP, which can either diffuse into, the nucleus or activate PKA, causing the latter to translocate into the nucleus (4). Downstream from calcium influx, MAP kinase can be activated and translocated into the nucleus during the induction of LTP in Aplysia (53). Murphy et al have shown that L-type channel mediated calcium influx induces expression of immediate early genes, such as c-fos, jun-B, zif268, and fos-B (58). In mammalian neurons, there are two lines of somewhat contradictory evidence regarding how calcium signal can be relayed into the nucleus. Deisseroth et al showed that Ca2+ influx through N-Methyl-DAspartate (NMDA) receptors or L-type VSCCs might activate co-localized calmodulin (CaM) and that the activated CaM translocates into nucleus, resulting in nuclear translocation of CaM and phosphorylation of the cAMP responsive element-binding protein (CREB) (22). In contrast, Hardingham and co-workers argue that, via diffusion mechanism, calcium enters the nucleus, which possesses all the necessary machinaries including CaM to trigger gene expression (37). Murphy et al also showed that a significant nuclear calcium transient could be observed with a delay of ~60 ms after the cytoplasmic calcium transient following calcium influx via the L-type calcium channels (59). This finding is consistent with a dependence of nuclear calcium elevation on  n  diffusion from a cytoplasmic source. Although evidence suggests that P/Q-type channels are involved in the synthesis of a presynaptic protein, syntaxin-lA (81), L-type calcium is clearly play a dominant role in providing the calcium needed for gene expression in neuronal plasticity.  Mechanisms to activities  underlying relevant  the selective  in gene  tuning of L-type  channels  transcription  Even though it is well established that L-type calcium channels are involved in gene expression regulation (58, 73), there is little evidence to explain why the L-type calcium channel is preferred over other HVA calcium channel types in activating gene expression.  Nakazawa et al demonstrated  that nuclear calcium elevation was largely reduced by application of L-type channel blocker, PN200-110. Addition of a single action potential (AP) on top of a large EPSP did not significantly affect the nuclear calcium transient and therefore suggests a preference of L-type channels for EPSP over AP component  in eliciting calcium  influx  (59). Consistently, Tsien's  group  provided evidence that via the L-type calcium channels, the EPSP elicits a significantly higher level of CREB phosphorylation (phosphorylated CREB initiates gene transcription in nucleus) than the APs do (55). In their study, the selective response to the EPSP was attributed to the relatively negative activation threshold and slower activation  kinetics of L-type  channels.  However, it seems untenable to make the comparison of calcium influx contribution between AP and EPSP without careful consideration of the integrated depolarization given that the integrated size of AP is incomparably  12  smaller than EPSP. Nonetheless, some channel properties can only be revealed when a train of APs is present (43). However, the arguments are made from evidence obtained from experiments performed mostly at room temperature. It  is highly  possible that at physiological temperature  calcium  channels  conduct calcium influx in a rather different fashion compared to that obtained at room temperature. Unfortunately, evidence from physiological temperature experiments has been intermittent. It has been shown that the activation and inactivation kinetics of calcium channels are both greatly accelerated upon temperature increase to physiological temperature level (50, 72). Further characterization of calcium channels at more physiological condition becomes indispensable for determination of the role of L-type calcium channels in gene expression.  Research hypothesis Previously, our laboratory has developed a computer simulation model to address the role of L-type calcium channels in gene expression (71). In the model, it was indicated that the L-type calcium channels prefer the longlasting plateau depolarization waveform over the repetitive firing waveform in response to tetanus stimuli. This waveform preference was manifested by larger calcium influx and was attributed to the slower activation kinetics of Ltype channels. A well acknowledged advantage for applying  physiological  waveform is that these physiological waveforms can reveal some channel properties  that  are  not  manifested  with  square  pulse  waveforms  (44).  Therefore, we decided to test whether in real cells the conclusion derived from the simulation holds. ;  13  With the physiological waveforms, we optimized our experiments with more physiological conditions. These conditions were largely neglected in previous studies (43, 44, 55, 59, 79) and are manifested in following aspects. 1) A tendency of performing experiment at room temperature for explanation of mammalian  system. This can be overcome  by  using a  temperature  coefficient Q 10 , a temperature coefficient defined as the change in the rate of a process as a result of increasing the temperature by 10°C. Application of Q i 0 is based on the assumption that the process of interest is linear. Over a long range of temperature, few of biological process display linear development as shown from our experiment (see fig. 19). 2) Using barium as charge carrier for  studying  calcium  behavior.  Since  barium  does  not  initiate  calcium-  dependent inactivation, it is used in some experiments to isolate the voltagedependent inactivation. However, it is often used as a signal booster for pure cosmetic reason since barium readily yields a 3 folds larger signal as calcium does (33, 77). The benefit is apparent in single channel recording in which signal detection is sometimes difficult.  However, barium can change the  inactivation profile and therefore distort the calcium response because it largely removes the calcium-dependent component of inactivation in whole cell recording mode (75, 78), making conclusion somewhat unreliable. 3) Using high concentration of charge carrier (barium or calcium). Another common practice to boost the signal is using high concentration of the permeant ion in the extracellular milieu, typically 20 or 40 mM barium or calcium. The concern of using high concentration of barium or calcium is the screening effect that is present when a divalent cation is in the extracellular milieu. The screening effect (or named surface potential of divalent ion) of a high concentration  14  calcium shifts the voltage-dependence of both activation and inactivation (40), thereby distorting the response to depolarization waveforms. To test the simulation results, the condition under which experiments are performed  is critical for producing a realistic  result. Therefore, our  experiments were performed in more physiological condition with following factors in addition to our physiological waveforms, physiological temperature of  37°C,  using  calcium  as  charge  carrier  and  applying  physiological  concentration (2 mM) of calcium in recording bath solution. Our ultimate goal is to test, under these physiological conditions, 1) whether L-type calcium channels prefer long-lasting plateau depolarization waveform over repetitive firing waveform; 2) whether the waveform results can unravel any novel behavior  of the calcium  channels.  3) to examine the  basic  biophysical  properties of HVA calcium channels.  15  Project outline Most of our experiments were carried out with human embryonic kidney cells, a cell line called HEK 293 cells, which carry only one type of transiently transfected  HVA calcium channels with varied subunit composition.  HVA  calcium channels were L-type, P/Q-type and N-type channels. N-type channels were from a stably expressed in HEK tsa201 cells. In these cells, following examination were performed. 1) Basic biophysical properties such as activation and inactivation kinetics, IV plot and temperature dependence of activation etc. 2) Among four major waveforms, plateau, repetitive firing, strong theta and weak  theta,  how  does  a particular  channel  type  respond  to  these  waveforms? What is the preference priority? 3) p2A subunit was used in lieu of p l B in some of the experiments to test the contribution of voltage-dependent inactivation to responses of a particular calcium channel type to waveforms. 4) To test the results obtained from HEK cells, some tests were performed in acutely  isolated  hippocampal  CA1  pyramidal  neurons.  Plateau  and  repetitive waveforms were applied. 5) Modifications of waveforms were made for further waveform analysis.  16  Chapter 2  Methods  Transient transfection of HEK 293 cells  Appropriate  medium  condition  The HEK cells were grown to 85% confluence in modified minimum essential medium (MEM) supplemented with 1-10% fetal bovine serum, 2 mM L-glutamine and 1 % penicillin/streptomycin (MEM + ). The pH of MEM+ medium was no higher than 7.4 when the pH of MEM medium was adjusted at 7.0. The pH in which cells were kept was proven to be critical for cells' health. The common practice is to make the MEM medium pH adjusted at 7.4. However, during the period for homogenization and filtration at room environment without the buffering partner for NaHC03, pH sneakily climbed up due to escape of C02, resulting in excessive NaOH left in the medium. The pH from this medium ranged from 7.7 to 8.5 as observed by keeping the medium without cell in the 5% C02 incubator for overnight. Bubbling the MEM medium with 5% C02 and 95% 02 during preparation had proven to be not helpful for this problem. The best solution in my practice was to make the medium exposure to air as short as possible. The MEM+ medium pH was close to 7.4 when the MEM medium pH is made at 7.0 initially. MEM medium with 7.0 pH value made the HEK cells much more healthier with smoother surface, more irregular shape and more processes after addition of supplements, as compared with cells in MEM pH adjusted at 7.4 at initial preparation. The new  17  solution reduced the cell death after calcium phosphate transfection  and  boosted transfection efficiency, both dramatically.  DNA  handling  DNAs were dissolved in water or l x TE solution (pH 8.0) and aliquoted in 500 (j.I volume vials with no more than 50 ^g DNA in each vial. They were routinely stored in - 2 0 °C freezer and thawed when used. Repeated thawing was proven to be less destructive to DNA than storage at 4 °C. DNAs that were stored at 4 °C for more than 2 weeks were not used.  Transient  calcium  phosphate  transfection  HEK 293 cells used for transient calcium phosphate transfection were regularly maintained in 37 °C incubator containing 5% C02 and split every two or three days via trypsin EDTA with 8 5 % confluence before passaging. For calcium phosphate transfection, cells were split 24 hours before transfection into a 10 cm petridish with desired number of cover slips lined at the bottom of the plate. Transfection only proceeded if most of the cells on the cover slips were healthy with irregular shape instead of rounded. Otherwise cells could suffer from poor health condition and accordingly poor transfection efficiency. The L-type and P/Q-type calcium channels were transiently transfected with calcium phosphate method. For each channel type, there were three subunits incorporated, a l subunit, a25 subunit and p i B subunit. a l C and ctlA encode Ltype and P/Q-type calcium channel respectively. For some experiments, p2A substituted for p i B subunit.  18  The DNAs were prepared for transfection in 10 cm petridish containing 4 |ig a l subunit, 4 |ug a25 subunit and 4 ng p i subunit. In our experiments, DNA in excess of 15 \xg in a 10 cm petridish tended to be toxic producing excessive cell death. The three DNAs were first placed in a .1.5 \x\ vial with a tenth volume of 3 M sodium acetate added, which needed to be well mixed via pipetting. Then 3x volume of 100% ethanol was added into the  DNA-  containing vial and was mixed well with DNA water solution by finger-tapping the vial until a white DNA cloud is clearly visible. After a short spin-down on a desktop centrifuge, the DNA-containing vial was filled up with 100% ethanol and undergoes further spin-down at 14,000 rpm in 4 °C environment. After this centrifuging, remaining procedures were performed all in sterilized hood. Once transferred into the hood, the fluid was carefully dumped as not to lose the DNA pellet. The DNA-containing vial needed to be sit up-side down until DNA pellet was dried. Transfection agents, 450 \x\ double-distilled water, 50 p.! 2.5 m CaCI2 and 2x BES were added into the DNA-containing vial with the sequence  as  mentioned. A thorough  mixture  followed  each  addition  of  transfection agent via finger-tapping or inverting. The DNA-containing vial sat for 20 minutes for formation of DNA-calcium phosphate complex, during which cell plates for transfection were transferred to a 3% C02 incubator for the same period so as to soften the cell membrane for easy penetration by calcium phosphate. At the end of the DNA-calcium period, DNA transfection solution was added into  phosphate  formation  10 ml warmed  MEM+  medium, which then was used for substitution for the medium in the cell plate. The optimal incubation time was 6 hours, longer than which seems toxic for the cells in our practice. At the end of transfection, the cover slips were first  19  rinsed in PBS solution and then transferred into a 6 well plate containing 2 ml warm MEM+ medium. The transfected cells were kept in 37 °C 5 % C02 incubator for 24 hours before being transferred into a 29 °C 5 % C02 incubator for up to 4 days. The best period in our practice for recording was from 3 rd to 5 th days after termination of transfection. To  monitor the transfection  efficiency,  each  batch  of  transfection  included a plate of cells transfected with green fluorescent protein (GFP) in substitution for a l subunit with the rest DNA species the same as in the regular calcium channel protein plates. Cells were not used for recording if the transfection efficiency was, lower than 3 0 % as indicated by GFP expression with fluorescence microscope examination. cDNAs encoding calcium channels were not co-transfected with GFP because in our preparation calcium channel expression level was greatly compromised by GFP co-expression. In the absence of GFP co-expression, irregularity of cell shape and cell process appeared to be good indicators of a sufficient calcium channel expression.  Acute isolation of hippocampal CA1 neurons Two to three weeks old wistar rat pups were chosen for acute isolation of hippocampal C M neurons. The animal was decapitated with a guillotine after anesthetization with halothane. The brain was exposed and dissected out within 1 minute, during which the brain was kept moistured with 1-4 °C 1 0 0 % oxygenated PIPES solution (in mM: 120 NaCI, 5 KCI, 1 CaCI 2/ 1 MgCI 2 , 20 PIPES,  and  25  glucose,  pH7.0,  300  mOsm).  Immediately  after  this,  hippocampi were isolated into freezing cold PIPES solution bubbled with 100% oxygen. The hippocampi were placed on a cold plastic platform under a 20  dissection microscope. Using a scalpel blade (FisherScientific,  08-916-5A,  No.10), the hippocampi were sliced manually into slices no thicker than 1 m m . In. each  hippocampus  hippocampus  were  only three  slices that  used. The total  are  period from  in the  middle  decapitation  to  of  the  end  of  hippocampus slicing was restrained in 15 minutes. The slices were incubated in 10 ml PIPES solution bubbled with 1 0 0 % oxygen at room temperature for 1 hour. At the end of one hour incubation, all the slices were transferred into the dissociation bottle containing 5 ml PIPES solution with 3 mg/ml protease (type XIV; Sigma) and incubated for 10 minutes at 37 °C bubbled with 100% oxygen. Thus enzymatically treated slices could be used for up to 10 hours for electrophysiology  experiment.  When needed, CA1 region was dissected out and gently triturated with a series of Pasteur pipettes with decreasing bore size of (in mm) 1, 0.8, 0.6, 0.4 and 0.2. Bubbles must be avoided. Cells were transferred onto the cover slip in the recording chamber without solution inside and sat for 3 minutes for cell attachment to cover slip surface.  Electrophysiology Voltage clamp recording were performed for all cell types at both room temperature (21.5-23°C) and physiological temperature (35-37.5°C) on the stage  of an inverted  microscope  (Aviovert  Carl Zeiss, Thornburg,  NY).  Borosilicate glass pipettes (Warner instrument Corp., GC150TF-7.5, 1.5 mm OD) were pulled using a Narashige PP-83 two steps vertical puller. Electrodes were routinely coated with sylgard (Dow Corning) to reduce the electrode pipette induced fast transient and followed by fire polishing with Narashige 21  Microforge. Series resistance of the electrodes were typically smaller than 3 MQ after routine 7 5 % compensation. The extracellular recording solution for HEK cells contained i n ' m M : 2 CaCI 2/ 1 MgCI 2 , 145 TEAMeS0 3 , and 10 HEPES (pH 7.38 adjusted with TEA-OH) with osmolarity adjusted to 300 mOsm. 20 mM CaCI2 and 125 mM TEAMeS03 were used as indicated for some of the Ntype channel experiments. The intracellular recording solution for HEK cells contained in mM: 120 CsMeS03, 5 CsCI, 10 EGTA, 1 MgCI2, 4 MgATP, 0.3 Na3GTP, and 10 HEPES (pH 7.3, CsOH) with osmolarity adjusted to 285 mOsm. Neurons were first perfused in the bath solution containing in mM: 140 NaCI, 5.4 KCI, 1 MgCI 2 , 1.8 CaCI2, 10 NaHEPES and 10 glucose (pH 7.38, NaOH)  with  formation,  osmolarity  the  normal  adjusted bath  to  solution  300 was  mOsm.  Upon  switched  to  membrane the  seal  extracellular  recording solution containing in mM: 1.8 CaCI2, 1 MgCI 2 , 10 HEPES, 140 TEACI, 5 CsCI, 10 glucose and 0.0003 TTX with osmolarity adjusted to 300 mOsm (pH 7.38, with Tris Aminomethane). Intracellular solution for neurons contained in mM:  190 NMD-Glutamine-CI, 10 EGTA, 12 phosphocreatine, 4 MgCI 2 , 3  Na2ATP, 0.2 Na3GTP and 40 HEPES with osmolarity adjusted to 280 mOsm (pH 7.30, with NMDG).  22  Currents were sampled at 10 kHz, filtered at 5 kHz, and acquired/analyzed using pCLAMP software and the Axopatch 200A amplifier (Axon Instruments, Foster City, CA, USA). During experiment, a cover slip with cells was placed in the recording chamber of 300 \i\ in volume that was perfused constantly at 1 ml per minute rate from a gravity driven system. With the two suction technique, typically a 3 GQ seal resistance was obtained after the membrane was ruptured. The leak current ranges from -5 to -50 pA. Command pulses were generated by a 12-bit digital-to-analog converter controlled by Clampex 8.0  (Axon  instrument).  The  junction  potentials  were  calculated  and  compensated for HEK cells (-14 mV) and for neurons ( + 9 mV) respectively. Leak subtraction was routinely performed with standard P/4 protocol for all recordings. The holding potential was - 7 0 mV for both HEK cells and neurons unless otherwise indicated. The step depolarization was set to 0 mV for the Land P/Q-type channels, +30 and +20 mV for the N-type channels with 20 mM and 2 mM calcium in the bath solution respectively. To overcome the rundown effect on calcium current comparison, only a short session of command waveform protocols with randomized sequence was delivered starting always with a 1 second step pulse. The total time period of waveform session was controlled in less than 2 minutes. The integrated calcium currents in response to simulated waveforms were normalized with a step response less than 20 seconds before the simulated waveform. Since the step pulse from - 7 0 mV to 0 mV was used for normalization for the L- and P/Q-type channel types and normalizing signal has to be standardized for all channel types, the peak amplitude of the normalizing step response was corrected with a factor obtained from the grouped IV data to represent the real peak step response  24  for each particular channel type. At physiological temperature, the factor is 0.98 and 0.93 for the L- and P/Q-type channels respectively. The two-tailed t test for paired and unpaired data was performed as appropriate to evaluate the statistical significance of differences between two group means. Values of P<0.05 were considered to indicate significance. Group data are expressed as means + SE. n refers to number of cells recorded from.  25  Chapter 3  Result  All the results were obtained from HEK 293 cells unless otherwise indicated  Characterization of the biophysical properties of calcium channels at room temperature with square pulse  Activation Activation kinetics were represented by 10-90% rise time, which was measured as the period from 10% to 90% of the peak calcium current amplitude, as well as half activation voltage, termed as V _ l / 2 or V_half. V_l/2 is the step depolarization voltage level at which calcium current peak amplitude is half as large as the maximal current peak when a current-voltage (IV) relationship is depicted with a series of step pulses with increasing depolarization level. An IV curve measures how easy or difficult a channel can be opened. In HEK cells, activation kinetics were measured at both room temperature and physiological temperature as shown in table2. The L-type channels showed a slightly longer activation rise time than the  N-type  channels (3.1+0.2 ms, n = 12 for L-type and 2.5+0.1, n = 17 ms for N-type channels, P<0.05). However, the P/Q-type channels exhibited a much longer rise time even than the L-type channels (6.3+0.07, n = 17; P<0.001). For the N-type channels, we also performed experiments with 20 mM calcium in bath solution, to examine the screening effect of positive charged divalent cation on  26  channel gating that will be discussed later. The speed of activation was not significantly affected by increased level of bathing calcium (2.43±0.29 ms, n = 5 for 2 mM calcium and 2.64±0.09 for 20 mM calcium). In contrast, the channel gating was markedly shifted by 20 mM bathing calcium as reflected from I-V relationship. I-V relation curve was plotted with a series of step depolarization from - 5 0 to +80 mV ( + 110 mV for N-type channels). The duration of I-V steps were set to a period that was marginally long enough to fully activate the channels so that a full tail current can be obtained at the same time. Fig. 3  show  an  example  of  activation  shift  by  increasing  bathing  calcium  concentration from 2 mM to 20 mM in the same cell at room temperature. The averaged IV peak shifted from 20 mV with 2 mM calcium to 35 mV with 20 mM calcium. The V _ l / 2 was obtained from activation curve that was plotted with channel conductance formulated as Ica/(Vm-E Ca ) from I-V curve. The I-V relation of the L-type calcium channels showed a more depolarized activation threshold as reflected in V _ l / 2 (-10.2+3.1 mV n = 5), compared to N-type channels with 2 mM calcium concentration in the bath (6.2±1.5 mV, n = 2; P<0.05). Again, P/Q-type showed a similar gating profile similar to the L-type channels as revealed from V _ l / 2 (12.6±0.9 mV, n = 9 ) .  27  Table 2. Basic biophysical factors revealed from step pulse L 22°C  P/Q N L  36°C  P/Q N  Rise time 3.2±0.2, n=12 6.8±0.6, n=20 2.5±0.1, n=17 0.55±0.04, n=19 0.37±0.03, n=17 0.47±0.05, n=10  Tau-1 29.6±2.0, n=8  Tau-1 % 45.3±2.3, n=8  Tau-2 298.5+17.7, n=7  Tau-2% 40.8±2.6, n=8  Tau-1+2 86.1 ±1.1, n=8  30.4+2.6, n=10 18.5+2.0, n=8 30.3+3.1, n=12 23.2±2.9, n=13  89.7+1.9, n=10 92.5+1.0, n=8 95.6+1.0, n=12 96.8±0.7, n=13  255.2+32.7, n==9 58.8+4.5, n=9 12.9+0.9, n=17 16.2±1.5, n=23 11.3±0.7, n=13  59.3±2.7, n=10 74.0±2.8, n=8 65.2±3.2, n=12 73.7±3.2, n=13  356.1 ±39.1, n=11 97.8+16.6, n=8 46.3+3.0, n=12 43.4±2.5, n=10  Residual* 14.9±1.1, n=7 24.9±4.0, n=9 12.9±2:0, n=12 6.2±1.0, n=11 2.7+0.4, n=14 22.±0.6, n=9  V 1/2 -10.2+3.1, n=5 -12.6+0.9, n=9 6.20+1.5, n=2 -20.0±1.3, n=8 -23.2±1.2, n=16 5.5±2.6, n=6  *: the ratio of the calcium current level at the end of 1 second step response versus the same step response peak.  Inactivation To examine inactivation, the inactivation time constant and steady state level were measured. Inactivation kinetics were measured by curve fitting with equation:  f(t)=iAte-^  +C  1= 1 At room temperature, the inactivation process was measured by taking the full length of one second step pulse time course for curve fitting. Interestingly, the P/Q-type channels showed only a slow inactivation component that fits a single exponential decay while both the L- and N-type channels had a double exponential decay, a fast and a slow inactivation process as shown in table 2. Another aspect of inactivation is the steady state level or residual current as measured as the ratio of the calcium current level at the end of the one second pulse depolarization versus the peak amplitude of the same step  28  response. Again, the L-type calcium channels showed a similar residual current to the N-type channels while the P/Q-type channels inactivated to a much less level compared to the other two channel types (see fig. 2). Therefore, room temperature data suggest that L-type calcium channels share much of the inactivation feature common to N-type channels and little to P/Qtype channels despite the fact that P/Q-type and N-type calcium channels have similar function in presynaptic transmitter release.  Calcium channel characterization at physiological temperature with square pulse Elevation in temperature to 35 - 37.5 °C drastically accelerated both activation and inactivation for all three HVA channel types (see fig. 3). The 1090% rise time shortened by about 10 times for all three channel types (see table 2) with little difference among these channel types. The activation time course of the three channel types (all within 0.6 ms in their rise time) suggests a similar ability for responding to APs. It also indicates the possibility that activation kinetics may not be critical in their distinct roles in cellular function. In addition to acceleration of activation time course, the channel opening also became easier for both L-type and P/Q-type calcium channels but not N-type channels as reflected in half activation voltage. The V_l/2 of both L- and P/Q-type channels- shifts to more negative potentials by about 10 mV (P<0.001 for both L- and P/Q-channel types). In our inactivation analysis, the first 200 ms was arbitrarily chosen for inactivation process curve fitting since more than 200 ms often resulted in  29  unpredictable  numbers  of  exponential  component.  Within  200  ms,  the  inactivation process was well fit for all three channel types with double exponential.  In  addition  to  the  acceleration  of  their  inactivation,  physiological temperature also uncovered a fast component for  the  P/Q-type  inactivation that was not seen at room temperature. Therefore, all three channel types carried two components, a fast component and a slow one with total inactivation fraction 9 3 % , 9 6 % and 9 7 % for the L-, P/Q- and N-type channels respectively.  The leftover from these inactivated channels matched  very well with actually measured residual current fraction ( 6 % , 3% and 2 % for the L-, P/Q- and N-type channels respectively, see table 2). The residual current of the L-type channels (6.1±1.0%, n = l l ) from these curve fitting was significantly larger than the P/Q-type (2.7±0.4%, n = 1 4 ; P<0.01) and the Ntype channels (1.8+0.2%, n = 12; P<0.001). This somewhat agrees with the previously proposed notion that the L-type channels have less inactivation compared  to  other  types.  However  the  extent  of  inactivation  changes  dramatically upon temperature increase to physiological temperature. For each particular channel type studied, the residual non-inactivated fraction was less than 1 0 % at 36°C. Whether this small though significantly different remaining channels can make a significant difference in the calcium entry in response to physiological depolarization waveforms  is yet to be clarified and will be  resolved at a later stage the result session. No difference in residual current fraction between the P/Q- and N-type channels was observed. Our data with step pulse indicate that both activation and inactivation was greatly accelerated at physiological temperature. From technical perspective, 1 second pulse appears sufficient for the calcium channels to reach the steady  30  state at physiological temperature and the first 200 ms segment of the 1 s pules is appropriate for predicting the residual current fraction with two exponential fit.  31  22 °C  36 °C  500 pA 10 ms  P/Q  200 pA 10 ms  300 pA 10 ms  Figure 3. Calcium currents in response to a 5 0 ms step depolarization. Calcium currents were recorded from either HEK 293 cells transiently transfected with L-type or P/Q-type calcium channel cDNA, or HEK tsa201A cells stably expressing N-type calcium channel protein. Holding potential is -70 mV for all channel types. The standard step pulse depolarization level is 0 mV for L- and P/Q-type cha.nnels and 20 (for physiological temperature experiments) or 30 mV (for room temperature experiments) for N-type channels.  32  Voltage (mV)  Figure 4. Current-voltage ( I V ) relationship of the L- P/Q- and N-type calcium channels. The IV relationship of the L- and P/Q-type channels is slightly more hyperpolarized at physiological temperature than that at room temperature but not that of the N-type channels. Note the IV curve of the N-type channels is more depolarized than other two channel types.  33  Voltage (mV) 80 ftO  — • — : 20 mM calcium —•—  : 2 mM calcium  1.0 J* CO  d)  CL  N 0.5  2 mM calcium  20 mM calcium  CO  E  o.o-40  0 40 Voltage (mV)  80  Figure 5. An e x a m p l e showing t h e effect of surface potential on N-type calcium channel gating property. The upper panel shows the switching bathing solution from 2 mM calcium solution to 20 mM calcium solution shifts the IV relation to a more depolarized level, indicating that the channels are more reluctant to open under higher concentration of divalent ion in the bath solution. The lower panel shows the activation curves plotted with calculated channel conductance from the IV plot.  34  Step duration (ms)  Figure 6. Effect of temperature on the response of calcium channels to mock APs with varied duration. A, example traces from the L-type calcium channels showing calcium currents in response to a series of step pulses with varied duration from 0.1 -3 ms with 0.2 ms increment to mimic AP with varied duration. B, Group data indicate that at room temperature, the amplitude of AP response is dependent up AP duration over a wide range. In contrast at physiological temperature, the AP responses of all channel types saturate within 0.7 ms, suggesting that all channel types are equally capable of responding to APs in full scale.  35  120-1 X 3  100 : Plateau  3  O  8 •o  80-  : Repetitive  60 -  N CO  40  E 20  L-type  P/Q-type  N-type  Figure 7. Calcium currents in response to t h e repetitive firing and plateau w a v e f o r m at room t e m p e r a t u r e . The values shown are I_ca++ integrals normalized to the peak of a preceding step pulse. The L-type channels show a preference for the plateau waveform to the repetitive firing waveform while no difference observed for the P/Q- and N-type channels between two waveforms. However, bigger differences are observed.between the L-type and other two channel types, revealing a novel behavior of L-type calcium channel.  36  Step  Repetitive  . Plateau  OmV -70 mV J  T P/Q  N  r r  200 ms UN"" 05nA 200 ms  V  b  nA  200 ms  Figure 8. Example traces of calcium currents in response t o different stimulus w a v e f o r m s at physiological t e m p e r a t u r e . The topmost panel shows the computer simulated waveforms, below which are calcium current traces recorded from HEK293 cells transfected with various calcium channel cDNAs. The different firing pattern between the plateau waveform and the repetitive firing waveform results from 30-fold larger synaptic conductance for the model neuron producing the plateau waveform than the one producing the repetitive firing waveform. The large synaptic conductance as shown in plateau waveform leads to larger depolarization that subsequently results in sodium channel inactivation and activation of the calcium-dependent potassium channels, both of which serve to diminish the APs. 37  \  * * *  100 i  1  X 3  c  80  E ^  hss'  BZ 22  Plateau Repetitive  Jfellllll^  60  CO  o  1 40 CO  E  20-  P/Q  N  Figure 9. Group data showing calcium currents in response to different stimulation waveforms at physiological temperature. The L-type calcium channels, when compared to the P/Q-type calcium channels, have a at least 3-fold larger calcium influx in response to both waveforms (P<0.001). It suggests that L-type VSCCs are selectively activated in response to intense physiologically-relevant waveforms. Both the L-type and P/Q-type calcium channels show preference for the plateau waveform over the repetitive waveform, suggesting that waveform preference is not specific to L-type channels.  38  Strong theta  VVeiak theta | 30  J  ffTTTin^  MVKJVJV  |2nA 400'  f'ffy^f^TlTV' | 400 pA  P/Q  f  400  i| p p r i" r" | 1 nA .  N  400  B 80 N or m ali ze d  ca  60  P=0.00003 -  T  : Strong PZ\ :Weak  40 -i  P=0.1  lei m in  0  m  m P/Q  N  Figure 1 0 . Calcium currents in response to theta burst firing waveform. A, Example traces show the response to strong theta waveform and the weak theta waveform. B, a significant difference in calcium current in response to the strong theta waveform is observed between the L-type calcium channels and other two channel types. No difference is observed in calcium current in response to the weak theta waveform between any pair of three channel types, suggesting that L-type calcium channels are associated with only the strong synaptic inputs but not the weak ones.  39  B OmV  100  -70 m V J  x 3  80-  E 200 ms  L P/Q N  I  3  0 60 re o  1 40 (0  E  20-  L P/Q N Figure 1 1 . Calcium currents in response to one second step depolarization pulse. Compared to the large difference in calcium influx in response to the complex depolarization waveforms, the differences in step responses are moderate with a marginal difference only between the L- and N-type channels (P=0.048).  40  B *** 0.6  n  1 ^ -.  **  I  o 0)  o o>  1  re o Q.  o E 0.4  ^9 o^  =i=f f  T  </> J*  E,  *** 1  0.2  i  Q_ < "D © N  1.0 •  re  05-  'T ,.-... .*  iE _ o z  I  o  •sv  '  •  *  •  .  n .  L  P/Q  N  P/Q  N  Figure 12. Paradoxical relation between rise time and relative AP peak amplitude at physiological temperature. A, P/Q-type calcium channels have fastest activation kinetics compared to Land N-type channels as revealed by 10-90% rise time. B, Unexpectedly, P/Qtype calcium channels have smallest response to APs in spite of its fastest activation kinetics. This paradox raises the question that whether activation kinetics is the determinant that shapes the calcium influx in response to APs.  41  +25 mV  J  -70 mV  Simulated  Figure 13. Tail current constitutes the main body of the calcium current response to AP. The left panel shows a L-type calcium current evoked by a 2 ms square pulse, demonstrating a full tail current. Note the depolarization-induced calcium current is only 2 nA. The right panel shows a calcium current response to an AP in the same recording trace. Two responses are 200 ms apart. The smaller amplitude of AP response is due to the depolarized EPSP before the AP.  42  B 8  ,«_,  ##*  1  50  i  ***  ^5  § •-  ,  +J  O (0 L_  *•  40  8»4-  ro  3 "O W d)  c o  T 2-  § 20  _L ' '.'.','  0) 10  01  a:  :•:-:•: n.  L  P/Q  N  L  P/Q  N  Figure 14. Inactivated fraction of calcium current in response to the step pulse and the simulated waveform. A, the residual fraction is the percentage of the residual current versus the peak of the step response. The steady state current of the L-type calcium channels is larger than those of the P/Q- and N-type channels (P<0.001 for both comparisons). B, the difference in inactivation revealed by residual fraction of the spiking among different HVA channel types appears larger in the response to the repetitive firing waveform, suggesting that the complex synaptic waveform response confers more physiological relevance than square pulse. It also provides a viable explanation for the distinctively large calcium current in response to the strong synaptic activity mediated by the L-type calcium channels. The inset shows a repetitive firing response.  43  Repetitive  Plateau ammwmnw  /irtAMA%wW  300 pA 200 ms  P/Q  III ill linn in NTT',,  Y~ 500 pA 200 ms  N  prr  'IH"I"""'  "  mm  200 pA 200 ms  Figure 15. Inactivation reflected by the amplitude of an end pulse in the simulated waveforms. An artificial step pulse with 5 ms duration and +25 depolarization level intended to examine the channels available at the end of one second waveform. Different calcium channel types exhibits distinct inactivation behavior revealed by the end-pulse response inconsistent with the AP responses, as shown most apparent in the N-type channels. These data suggestthat different channel inactivation behavior may be attributed to their distinct recovery from inactivation as can be hinted by the glitch labeled by the asterisks.  44  A  L-P2A  'igpjpr1'*^  K M M M i  300 pA 200 ms  P/Q-P2A ^YYYY*f^  200 ms Repetitive Plateau  B 1.0 (0  *~l  0.8  TO O Q.  0.6  c •l  0.4  re o a.  0.2 0.0 P1b  P2a  (31b  P2a  P/Q Figure 16. Inactivation reflected by the end-pulse in calcium channels with different b subunits. A, example traces of the L- and the P/Q-type calcium channels co-transfected with b2A subunits. B, group data show that the inactivation in the P/Q-type channels are mostly reversed by b subunit substitution while that in the L-type channels is also attenuated. 45  A  P/Q  -70 mV  200 ms /Q  -110 mV w  100 pA 200 ms  200 ms  Figure 17. Voltage-dependence of P/Q-type inactivation. A, digitally removal of the EPSP component of the P/Q-channels largely attenuates the inactivation at -70 mV holding potential (middle). Lower panel shows hyperpolarization to -110 mV completely abolishes the inactivation revealed by spiking reduction. B, the L-type calcium channels recover spiking arrest (by digitally removal of 4 APs) while the inactivation of P/Q-type channels proceeds after the gap. These data suggest a unique inactivation profile of P/Q-type calcium channels, voltage-dependent and slow recovery from inactivation compared to other channel types.  46  -70 mV  800 pA 200 ms  200 ms  Figure 18. Calcium current in response to the hippocampus-derived waveform. Calcium current of the L-type (middle) and the P/Q-type calcium channels (lower) in response to hippocampus-derived depolarization waveform (upper). Group data indicate similar difference between the L-type and the P/Q-type calcium current in response the hippocampus-derived waveform, compared to the simulated repetitive firing waveform.  47  ^umptf.mwur  E4.0 0)  .13.0 •*—f  5 2.01 1000pA  1.0  20 ms  0.0 '22  26 30 34 TemperatureXP)  38  26  38  < 9 -'2 TO  * 1 22  30  34  Temperature^)  Figure 19. Example of temperature effect on activation rate and calcium peak currents. The right panel shows an example cell (an acutely isolated hippocampal CAl neuron) undergoing temperature change. As temperature increases from room temperature, activation process accelerates as seen in shortening 10-90% rise time (upper right) and the macroscopic current increases (lower right).  48  Repetitive firing  Plateau  "jpp— 1 nA 200 ms  All channel types miiHfniiim*inm**«>**fl','"r  L-component  0.5 nA 200 ms  Figure 20. Calcium currents in hippocampal CAl neurons. Calcium currents in hippocampal CAl neurons in response to the repetitive firing (top left) and the plateau (top right) waveforms. 'All channel types' denotes the currents are obtained in absence of any channel blocker. The 'Lcomponent' is obtained in presence of 1 \xM co-conotoxin GIVA (for N-type channel blockade), 1 fiM co-conotoxin MVIIC and 200 nM co-agatoxin IVA (for the P/Q-channel blockade).  49  0 < £Z  \  ^  -2 -  C  0  /  <D v_ i_  O +  -4 -  ©: 22°C before 36°C (D: 36°C CD: 22°C after 36°C  CN  05  O  -6 "  1 ® -8 0  i  i  i  i  \  i  i  i  i  50  i  100  Time (ms) Figure 2 1 . Temperature effect on calcium current is reversible. An example of a L-type calcium current showing that the acceleration of activation kinetics and the augmentation of current amplitude can be reversed by re-perfusing the cell with room temperature bath solution.  50  Waveform analysis at room temperature With the advantage of computer simulation of neuron activity from our laboratory, our project started with an attempt to test the synaptic waveform preference of L-type calcium channels. Evidence suggests that L-type calcium channels prefer the long-lasting EPSP component of depolarization waveform to waveforms that mainly contain APs (55, 59). To test this hypothesis, two major  waveforms  were  used,  repetitive  firing  waveform  and  plateau  waveform. In modeling these waveforms, two neurons were included with all cellular machinaries, glutamate receptors, different kinds of membrane ionic channels and calcium buffers for example, that are necessary for generating neuronal activity such as spike accommodation, LTP etc. In our two neuron model, a classical LTP-inducing stimulation protocol, a one second 100 Hz buzz, was applied to the presynaptic neuron. To simulate the different synaptic response intensity, total synaptic conductance was manipulated in the model to represent the amount of response from glutamate receptors, both AMPA and NMDA receptors. For the two major waveforms, the plateau waveform was generated from a postsynaptic neuron with 30 times synaptic conductance as large as the one for the repetitive firing waveform. At room temperature, the L-type calcium channels showed a unique preference of large long-lasting plateau waveform with a more than two folds larger total calcium current over that in response to the repetitive firing waveform depolarization (P<0.01) (see fig. 7). The square step pulse induced calcium current far exceeded even the plateau- waveform response by about  51  two folds (P<0.001). In contrast, both P/Q-type and N-type calcium channels showed no preference for either of the two simulated waveforms. These data are in agreement with previous reports that the L-type calcium channels favor the long-lasting large depolarization for the calcium influx, comparing to repetitive firing depolarization waveform (55, 59). Compared to the plateau waveform, the L-type calcium channels mediated only less than two fold larger calcium current in response to the 1 second step depolarization. The P/Q- and N-type channels mediated more than 20 folds larger calcium current than the plateau waveform response. In the case of the P/Q-type channels, the step depolarization induced calcium current even exceeded that of the  L-type  channels by 6 0 % (see table 2), suggesting that P/Q- and N-type VSCCs are more voltage-dependent compared to L-type VSCCs. In comparison of the total calcium current among the three HVA calcium channels, the L-type calcium channels was found to mediate  surprisingly  larger current in response to the simulated waveforms compared to other two channel types. This provides a possible explanation as to why L-type calcium channels are adapted to plasmic calcium activity considering the relatively large cytoplasmic volume. Another observation came from P/Q-type channels' response to step pulse. In contrast to simulated waveform response, the P/Qtype channels mediated a much larger total calcium current in response to step pulse depolarization. Combined with the observation of  inactivation  profile, this large total calcium current may well be due to the large residual current from less inactivation. Consistent with previous findings, our data obtained at room temperature indicate that in response to the  complex  depolarization waveforms, the L-type VSCCs exhibited a strong preference for  52  the plateau waveform over the repetitive firing waveform while the other two HVA calcium channels did not show any preference for either of the two waveforms.  Waveform analysis at physiological temperature In  addition  experiments, waveform  we  and  to also weak  the  waveforms  modeled theta  two  used  at  additional  waveform  for  the  room  waveforms, physiological  temperature strong  theta  temperature  experiments. To better present the waveform preference by calcium channels, the repetitive waveform was slightly modified such that the APs were spread throughout  the  one  second  depolarization  period  with  slightly  lower  depolarization level (-50 mV) (the EPSP level for the other repetitive firing waveform is - 4 5 mV) to amplify the difference in calcium influx between repetitive firing and plateau waveform. This new repetitive firing waveform was tested at room temperature and showed no difference from the other repetitive waveform in its total calcium current. All the experiments at the physiological temperature were performed with this new repetitive waveform and no distinction will be mentioned hereafter.  53  Table 3. Calcium current response to simulated w a v e f o r m s at physiological t e m p e r a t u r e 36 °C L  P/Q N  Repetiti ve (22°C) 45.5±7.0 , n = 10 21.6+2.7 , n= l l 2.1+0.5, n=8 -  Repetiti Plateau ' (22°C) ve 57.0±2.7, n = 15 16.5+1.1, n = 17 10.7+1.9, n=8  Plateau  strong theta  weak theta  Step : (22°C)  Step  107.8±7. 94.6±6. 62.7±6.2 13.70+2. 92.0+12. 92.0+12. 9, n = 10 6, n = 15 3, n=8 2, n = 9 , n=8 2, n=9 ,23.9+7.4 32.6+2. 17.6+2.1 9.0+1.6, 74.6+13. 74.6+13. n=4 '••'-., n = 10 2, n = 14 , n=6 3, n=8 3, n=8 5,1+1.4, 14.3+3. 13.4+2.3 7.5+1.3, 62.0+8.4, 62.0+8.4 6, n = 8 , n = 12 n = 12 : ;.;:-n=8 , n=8 •V..ri= 8  The difference in the repetitive firing and plateau waveforms observed at the room temperature remained significant at physiological temperature. However, this difference was not unique to L-type channels anymore as at room temperature. Like the L-type channels, the P/Q-type calcium channels also exhibited a preference for the plateau waveform at 36 °C. Nonetheless, the difference in the total calcium currents between the repetitive firing and plateau waveform was relatively small, only less than one fold. This somewhat reduces the importance of the preference of L-type channels for long-lasting large EPSP containing waveform. Interestingly, the L-type calcium channels mediated much larger total calcium current in response to the simulated depolarization waveforms compared to the P/Q- and N-type channels. In response to the repetitive firing waveform for instance, the L-type channels mediated 3.5 and 5 folds total calcium current as large as the P/Q- and N-type channels mediated total calcium current respectively.  In contrast to the  apparently large L-type calcium current in response to complex  synaptic  waveforms, the total L-type calcium current in response to step depolarization showed no difference from the P/Q-type channels and a marginally larger, current than N-type channels (P=0.048) as shown in table 3 and fig. 1 1 . This  54  outstanding large calcium influx induced by physiological  depolarization  patterns via the L-type channels has not been reported before and is especially attractive in that it provides insight into the cellular functions of calcium channels, which will be discussed later. It also suggests that step depolarization pulse is limited in its capacity for studying calcium channel behaviour in physiological condition. It is possible that our waveforms are not representative of physiological patterns in live tissues despite the fact that all conceivable physiological parameters have been incorporated into the model. To test this possibility, Dr. Murphy simulated two additional waveforms with the 'NEURON' program, a strong theta and a weak theta waveform that were generated in response to the theta-like presynaptic stimulation with the postsynaptic conductance resembling the one in the plateau and the repetitive firing waveforms respectively. Theta rhythm is believed to be dominant in the hippocampus during walking and rapid-eye-movement sleep and therefore represents a significant portion of physiological firing pattern in live animals. Our result indicated that the L-type calcium channels appear to be superior to the P/Qand N-type channels in their ability of responding to the strong theta depolarization waveform as to the repetitive firing and the plateau waveforms (see table 3 and fig. 10). As for the response to the weak theta waveform, no difference was observed between L-type and P/Q-type channels and only a moderately smaller calcium current via N-type  channels was observed  compared to L-type channels (P=0.017) as shown in table 3. Comparison of the total calcium current in response to these depolarization patterns with decreasing synaptic conductance suggests that the L-type calcium channels 55  are associated with strong synaptic inputs, therefore providing a strong support for the important role of the L-type channels in the activity-dependent gene expression. The next imminent task is to identify the biophysical property that contribute the L-type calcium channels' superiority in mediating the strong synaptic input induced calcium influx. Previously, it has been reported that the L-type calcium channels are associated with long-lasting large EPSP better than APs (55, 59). The preference of L-type calcium channels for EPSP component was attributed- to their slower activation kinetics. However, our results unraveled a drastic difference in both activation and  inactivation  kinetics of calcium channels between room temperature and physiological temperature. Therefore, it is possible that a different biophysical aspect may be manifested in channel behavior at physiological temperature not evident at room temperature.  Analysis of activation kinetics at physiological temperature As mentioned in the previous section, activation rise time shortened to about one tenth as experimental temperature increased to 36°C. The L-type channel rise time was no difference from the N-type channel rise time larger than the P/Q-type channels by about 5 0 % (see table 2). Previous studies have suggested that the L-type VSCCs respond preferentially to a large long-lasting depolarization and do not respond well to brief stimuli such as single APs (55, 58).  However, these studies were limited in that they were performed on a  . 56  heterogeneous population of channels at room temperature. The studies also failed to  describe  how  L-type  VSCCs would  be activated  by  bursts  of  APs/EPSPs associated with synaptic plasticity (12, 15, 86). To further address this issue voltage steps of different duration were used to simulate APs of limiting duration. This was tested with a series of step pulse with varied duration from 0.1 - 3 ms with 0.2 ms increment to mimic the APs with different widths. The peak amplitude of the calcium current was measured for curve  fitting. The test  was  performed  at  both  room  and  physiological  temperature. At room temperature, the step widths of < 1 ms resulted in < 5 0 % of maximal channel activation for L-, P/Q- and N-type channels (half maximal activation occurred at 1.48 ms n = 4 , 1.40 ms n = 4 and 1.39 ms n = 17 for L-, P/Q- and N-type channels at 22 °C, respectively). However at .36 °C simulated AP widths of as low as 0.6 ms resulted in greater than 9 0 % of maximal channel activation for all three channel types studied ( 5 0 % of maximal activation occurred at 0.45 ms n = 3, 0.44 ms n = 8, and 0.48 ms n = 5 for L, P/Q, and N-type channels at 36 °C). This result indicates that at 36 °C L, P/Q, and N are all effectively activated by single AP stimuli (0.4 - 0.7 ms half width) and that large EPSPs are not necessarily required. Thus, at reduced temperature (22 °C) it is possible that conditions that alter AP duration such as the modulation of potassium channels would change both the degree and duration of channel activation (Sabatini and Regehr 1997). To verify the results obtained from duration variation, amplitude of the APs in the repetitive firing waveform and plateau waveform were examined with the peak of the preceding step repulse as the normalizing signal. The AP relative amplitude was similar between the L-type and the N-type calcium  57  channels with  1.38±0.04  (n = 28) for the L-type  channels and  1.17±0.05  (n = 29) for the N-type channels. The P/Q-type channels have even smaller AP response (0.87±0.05, n = 20; P<0.01 in comparison with both the L- and Ntype channels). Looking closer into the detail of the AP response, we found that the main body, of AP response was composed of tail current (see fig. 13). These results delivered two folds of important messages. First, activation kinetics are not the predominant factor that shapes the calcium influx in response to APs. Instead, the inactivation is most likely the main force that affects the size of AP response. Secondly, the L-type calcium channels are at least equally capable of responding to AP in full scale as the P/Q- and N-type channels, if not more. In summary, our results indicated that  activation  kinetics in the physiological temperature range has little impact on shaping the calcium influx in response to APs.  Inactivation analysis at physiological temperature We next examined inactivation for a possible candidate  mechanism  underlying the differential calcium influx in response to synaptic waveforms via various calcium channels. As revealed from residual calcium current, the Ltype channels showed less inactivation compared to the P/Q- and channels.  This  suggests  that  the  inactivation  may  be  critical  N-type for  the  distinctively large calcium entry via the L-type channels. Again, this needs to be corroborated with waveform analysis.  58  Inactivation  revealed  by spiking  attenuation  In the repetitive firing waveform, the responses to the first and the last APs were  compared  temperature. inactivation  The  for  L-type  (57.2±1.9%,  the  proportion  calcium n=9)  of  inactivation .at  channels  compared  showed to  the  physiological  considerably P/Q-type  less  channels  (90.1+0.8%, n = 8; P<0.001) and the N-type channels (94.6±1.6%,  n=8;  P<0.001) (See fig. 14). Interestingly, the amount of L-type inactivation from this spiking waveform was drastically less than that from step pulse as revealed by residual calcium current (94%) as shown in table 2. It led to a untouched  possibility  that  the  inactivation  characteristics  makes  L-type  calcium channels unique in mediating a large calcium influx that plays the critical role in triggering gene transcription. To further confirm the observation from AP comparison, we further tested the channel availability at the end of the repetitive firing waveform since the last AP amplitude does not reflect the full channel availability due to its narrow width (1.2 ms half width for the last AP in the repetitive firing waveform). The last AP in the repetitive firing waveform was replaced with a square pulse with a 5 ms duration and a +25 mV depolarization level. With this paradigm, the L- and P/Q-type channels showed slightly less inactivation (45.0±4.2%, n = 8 and 86.2±2.4%, n = 7 for the L- and P/Q-type  channels  respectively) as compared to the original waveform to a similar extent. In contrast, the N-type inactivation revealed by this 5 ms step pulse was 5 1 . 8 % (42.8+8.2%, n = 4 ; P<0.001) less than that in the original waveform. The surprisingly large difference in the N-type inactivation observed during 5 ms  59  pulse and the normal AP is possibly due to a quick recovery from inactivation (will be discussed in detail later).  Channel recovery  from spiking  arrest  Since our 5 ms end pulse experiment suggested a differential recovery among different calcium channel types, it is necessary to know how much recovery can be attained after high frequency spiking arrest. Previously, Patil et al (64) showed that  that  P/Q  and  N-type  channels  can  become  preferentially inactivated during the interval between successive stimuli. This form of inactivation was termed 'intermediate closed-state inactivation' and might help explain why the P/Q channels were selectively inactivated by the complex stimuli that contain relatively brief periods of depolarization when compared to sustained step depolarization. A similar phenomenon also occurs with Na+ channels and contributes to activity-dependent  alterations in  dendritic spiking (17, 56). In the case of our VSCC data, the depolarized AP peaks in combination with more moderate EPSPs may be ideal for producing selective inactivation of P/Q and N-type Ca2+ channels. To test this idea we produced simulated waveforms in which 4 APs were removed to determine whether inactivation of P/Q-type channels would proceed in the absence of AP stimulation. Our results indicate that the spiking arrest had little effect on the inactivation of subsequent AP-evoked P/Q-type currents suggesting that the inactivation process once begun could proceed in the absence of stimulation (Fig. 2C). To quantify this effect we measured the average current attributed to the 3 spikes preceding the gap and 3 after the gap. In the case of the P/Q channel a significant reduction in the current amplitude associated with the 60  spikes was observed (13.3±1.3%, n = 5; P<0.05); while removal of the APs led to an increase in the L-type VSCC response (7.2+2.7%, n = 9 ; P<0.05).  Effect of inter-spike  potential  on  In the intermediate close-state  inactivation inactivation theory, the amount  inactivation of P/Q- and N-type channels is voltage-dependent  of  (44). To  determine voltage-dependence of inactivation in our waveforms, the EPSP depolarization level of the repetitive firing waveform was varied by digitally removing the EPSP component without changing the spatial distribution and the depolarization level of the APs, resulting in a waveform termed AP-train-70 (-70 denotes the inter-spike voltage level). The amount of inactivation in waveform is measured as the reduction percentage of the last AP response compared to the first AP response. Our results indicated that there was only moderate inactivation attenuation for the L- and N-type channels ( 5 5 . 4 % to 4 1 . 6 % , n = 10 and 9 7 . 0 % to 8 7 . 6 % , n = 5 respectively) by EPSP removal. In contrast, the P/Q-type channels experienced a drastic inactivation attenuation ( 8 8 . 9 % to 4 7 . 8 % , n=9) by EPSP removal. Since the P/Q-type  channels  seemed more responsive to the inter-spike potential manipulation, the interspike potential was changed to - 1 1 0 mV to amplify the effect. As predicted, the inactivation as revealed by the relative size of the last AP was mostly abolished to 16.5±4.8% (n = 5). In summary, all three HVA channels were nearly equally activated at physiological temperature. Therefore the activation kinetics are not likely to be responsible for the differential response to synaptic depolarization waveforms among the three HVA channel types. Instead, inactivation profile  differs 61  among different channel types and appears to play a critical role in shaping the calcium entry in response to synaptic input. In addition to the extent of inactivation, the recovery rate is as well shown to be important in this regard.  Modulation  of biophysical  properties  by [3 subunit  Since the L-type and P/Q-type calcium channels presented drastic contrast in their inactivation profile, we investigated the mechanisms that may contribute to their, distinct inactivation features. There has been a consensus that inactivation of calcium channel is modulated by p subunit type (78, 80). Among various types of p subunit, p2A is known for the substantial slowing of inactivation. This is especially important for characterization of P/Q-type calcium channels because P/Q-type channels are mainly presynaptic calcium channels and undergo much voltage-dependent modulation from AP activity (83).  62  /  Table 4. P a r a m e t e r s for calcium channels w i t h p2A subunit Biophysical properties Rise time Residual* L  .  ' 0.43±0.0  6, n = 5 0.47±0.0  16+2.0,  . pulse 95±5,  n-5 32±3.0,  n=5 80±4,  Waveform responses Repetitive Plateau K . . 80±20, 156±29,  \. *. ,.L ,. Step theta theta \__ 60+10, 17.5±6.3 253±12,  n=9 n-9 n-4 , n-4 n-5 188±30, 576±77, 282±21, 26.0±16. 517±42,  '^ 3, n = 6 n=5 n=6 n=6 n=6 n=4 3, n = 2 n=4 * : denotes the percentage of calcium current at the end of 1 second depolarization to 0 mV in comparison with the peak of the same response.  To investigate the role of B subunit in modulation of inactivation in our waveform, p i B subunit in our transient transfection system was replaced with p2A subunit for both the L- and P/Q-type calcium channels. Our results showed that the steady state inactivation level revealed by residual current ratio of P/Q-type channels was markedly attenuated by p2A substitution, with a more than 10 times reduction in steady state inactivation (residual current ratio increases from 2.7+0.4%, n=8 to 32.5±2.6%, n = 5; P=9.2xl0" 1 3 ). For Ltype channels, the residual current ratio increased by only less than two fold, from 6.2±1.0% (n = l l ) to 16.0±2.0 (n = 5, P=1.8xl0" 4 ) as shown in table 4. In contrast, the activation kinetics were not affected by p subunit substitution for either of channel types studied with the p2A subunit. As predicted from inactivation attenuation by p subunit substitution, the inactivation as revealed by the last AP relative amplitude in our simulated waveforms was greatly reversed in the P/Q-type channels but not the L-type channels. The relative amplitude of the last AP response in the repetitive firing waveform in the P/Qtype channels increased from 9.8±0.8% ( n = 8 ) with the p i B subunit to 73±3% (n = 6, P = 4 . 1 x l 0 " u ) with the p2A subunit. No difference was observed for the L-type channels between p i B and p2A subunit (with p i B : 42.7+1.9%, n = 9 ;  63  with p2A: 46.9+3.5%, n=9; P=0.33). Our data indicated that p2A reversed the P/Q-type inactivation and exerted little effect on the L-type inactivation, suggesting that p2A subunit only has an effect on the voltage-dependent inactivation, possibly via elimination of voltage-dependent inactivation.  Calcium current induced by hippocampus-derived firing waveform To validate our simulated waveforms, a repetitive firing waveform obtained from hippocampal slice (experiment performed at 33°C) was applied to the L-type and P/Q-type calcium channels in HEK cell expression system. Consistent to that obtained from the HEK cells, at physiological temperature, the L-type calcium channels mediated much larger total calcium current (113.8±10.2, n = 10) than the P/Q-channels (50.1+8.2, n=4; P<0.01), which justified the application of the simulated waveforms with our approach.  Calcium channels in acutely isolated hippocampal neurons Since all the efforts was dedicated to explore calcium channel behavior in the central neurons, results obtained from HEK cells have to be challenged by neurons. Therefore, we went on to perform experiments on the acutely isolated hippocampal CA1 pyramidal neurons. Hippocampus is the structure in medial temporal lobe that is involved in formation of declarative memory and has been under intensive study. CA1 region is the center of, focus since it undergoes considerable long-term 64  potentiation (LTP) that is believed to be a critical substrate of memory after the discovery of LTP (7). The advantage of using hippocampal neurons is that result reflects real physiological nature of calcium channels with all the modulatory machinaries present for an adaptive calcium influx. The disadvantage exists in that the unidentified calcium channel type named R-type easily contaminates  the  calcium channel behavior observed since it lacks effective antagonist. In hippocampus, there is a mixture of several calcium channel types including L-type, P/Q-type, N-type and the unidentified R-type channels (9092). For our purpose of studying the L-type calcium channels, usage of potent P/Q- and N-type calcium channel blockers allows us to estimate the L-type channel mediated current. These blockers include one potent N-type channels blocker, co-conotoxin GVIA ( 1 JIM), two P/Q-type channel blockers co-agatoxin IVA (200 nM) and co-conotoxin MVIIC (l-\iM).  Our data showed that the L-type  current constitutes predominant fraction (about 60%) of total CA1 neuron calcium current. Only a small amount of calcium current remained  after  blockade with the toxins mentioned above and the L-type channel potent antagonist nifedipine. It suggests that our experiments with toxins for P/Qand N-type channels will closely estimate the behavior of the L-type calcium channels. In this section, all the data are obtained in the presence of a cocktail of co-conotoxin GVIA ( 1 \xM), co-agatoxin IVA (200 nM) and co-conotoxin MVIIC (l^iM). Our results indicated that activation kinetics of neuronal L-type channels resembled that observed irv.HEK cells with 1 0 - 9 0 % rise time 0.40±0.04 ms ( n = 6 , P<0.05). The waveform results showed that as in HEK cells, the L-type 65  component of hippocampal CAl neuron calcium current yielded a total calcium current 57.2±7.3 (normalized to the peak of the preceding step response), n=7 and 85.7±22.0, n = 7 in response to the repetitive firing and plateau waveform respectively. No difference was obtained when compared to that obtained in HEK 293 cells.  66  Chapter 4  Discussion  Appropriate experimental conditions for calcium channel investigation It has been well established that ionic channel kinetics are affected by experimental  condition  such  as  temperature  (41)  and  divalent  ion  concentration. Among these experimental conditions, temperature is of special importance because many of the biological processes, such as ion channel function activation or inactivation and enzyme activity and so forth, are believed to  be temperature-dependent.  Unfortunately, there  is meager  evidence for calcium channel behavior at physiological temperature (1> 47, 72). Our knowledge about calcium channels is largely based on experiments performed at room temperature. To overcome the temperature gap, a compensation strategy for data analysis, the temperature coefficient called Qi 0 was developed. Qi 0 denotes the change in the rate of a biological process as a result of increasing the temperature by 10°C. Utilization of this coefficient is based on the assumption that the biological process has a linear dependence on temperature. For a narrow and defined temperature range, Q10 value does afford a satisfactory approximation. Yet over a wide range as from 22°C to 36°C, it is rather possible that there are several processes involved in channel function, each operating optimally only over a portion of the full temperature range. Unless the reaction rates for each process are identical, their collective effect will be nonlinear. Our results indicate that both activation kinetics and  67  calcium current amplitude change over a wide temperature  range in a  nonlinear fashion (see fig. 19). The nonlinear temperature-dependence of channel properties has been reported previously and attributed to changes in the fluidity of membrane lipids that, in t u r n , influence the degree and rate of channel opening (29). The transition between gel and liquid-crystal phases in biological  membrane  is  well  known  (54).  This  phase  transition  would  presumably influence the intramembranous portion of embedded ion channels. In the case of calcium channels, it is the a l subunit that is entirely located within the membrane where it would be expected to be vulnerable to this thermotropic transition. Nonetheless, it is a l subunit that undergoes voltagedependent  activation. Some  more recent studies showed that  temperature accelerated both activation and inactivation  increased  rate of calcium  channels (47, 72). However, the studies were not specific to a particular channel type and the characterization was rather incomplete. Another common practice that causes misunderstanding about calcium channel property is the use of high concentration of extracellular calcium. High concentration of either barium or calcium as charge carrier becomes attractive due to following reasons. Sometimes calcium signal can be quite small even undetectable. For transiently expressed calcium channels, the portion of cells with decent signal is often quite small even though the transfection efficiency is very high as indicated from p-gal staining. One possibility is that the calcium channels may not be well expressed in the cell membrane  or  become  internalized depending on the health condition of the cells (6). Under these conditions, co-expression with some transfection indicators such as green fluorescent protein (GFP) or p-gal does not yield satisfactory match with signal  68  pick-up during the recording of calcium currents. High concentration of calcium or barium can augment the signal size to well above detectable level, boost the signal-noise ratio by many folds (our data, not shown). However, high concentration of divalent ions exerts a surface charge on cell membrane, thus shifting the channel activation to the right as shown in fig. 5. As a consequence, a depolarization pattern as shown in our simulated waveforms would yield less calcium influx in presence of high concentration of calcium, which would certainly  make all the calcium-dependent  cellular  events  unpredictable. In our experiments, all the calcium responses to waveforms are normalized to a control signal that shortly precedes the waveform recording by about 20 seconds. Since all the calcium signals are expected to be affected by the activation shift, whether the activation shift would exert equal effect on the calcium currents in response to the different synaptic waveforms remains uncertain. Our results indicate that normal calcium concentration yields larger calcium currents in response to the repetitive firing and plateau waveform. Since both activation and inactivation can both be affected by extracellular calcium concentration, the mechanism that underlies the difference in our results between normal and high concentration divalent ion is yet to be determined.  Our  data  suggests  that  at  macroscopic  level,  normal  concentration is capable of yielding sufficient calcium signal and usage of high calcium concentration should be avoided due to its apparent effect on calcium channel gating. Barium is often used as a substitute for calcium since using barium can easily generate a calcium current 3 fold larger than that with calcium as the permeant ion (33, 77). A legitimate reason for using barium is to isolate the 69  voltage-dependent  inactivation  since barium  has little  calcium-dependent  inactivation but is of higher permeability for the calcium channels (66, 95). With barium as a charge carrier, the activation rate is slower while inactivation is almost completely abolished at room temperature (66). Even for study in which inactivation does not seem to be important, considerable caution should be taken given that our understanding about calcium channels is far from thorough. For example, there is a strong consensus that the size of the calcium current in response to the AP is affected by activation kinetics. This holds true when the experiments are performed at room temperature since the response to the APs (about 0.8 ms half width from our results) is filtered by channel's slow activation rate (with a rise time range of 3.18±0.19 ms from our results). However, when the experiments are performed at physiological temperature, all the activation rise time are within 0.7 ms. All the HVA channel types become capable of responding to APs in a full scale. Looking into the AP response closer we found that the AP is composed mainly of the tail current, the size of which is largely modulated by the inactivation status  (40).  Utilization of barium as permeant ion would cause some unpredictable effects and therefore distort the normal calcium current in response to synaptic waveforms. Finally, the characterization of calcium channel using square-shaped step pulse also appears problematic. Yue et al showed that the inactivation in high frequency AP firing is different from that in square pulse depolarization when the total depolarization area and summed duration are equal between the AP train waveform and the step depolarization waveform (64). Similarly,  70  novel property of calcium channel may emerge if challenged by more complex firing pattern. Taking these together, experimental temperature, charge carrier specie, concentration of charge carrier and the application of physiological waveform are necessary for characterization of calcium channels. Our experiment utilized all the concerned elements that make the condition for channel analysis physiological. One exception is the genetic difference between the HEK cells and neurons. Neurons possess intracellular proteins that interact with calcium channels such as G-protein (96) etc while HEK may have different sets of plasmic and membrane interacting proteins. To date, the difficulty in analyzing calcium channel remains standing due to lack of effective antagonist for Rtype calcium channels. Two strategies have been developed. One is to approximate the interested channel type with rest of the known channel types blocked, with the R-type channels unblocked. The other is to use the blocker of the  interested channel type and analyze the channel  by  generating  difference signal between the signal without blocker and the signal with the blocker. The latter is not reliable to some extent in that current rundown or other cellular activity deterioration occurs during the drug effecting, especially for experiments performed at physiological temperature due to more rapid current  rundown. In our experiments, approximation  for L-type  calcium  channel characterization was performed with blockers for P/Q- and N-type calcium channels to reproduce the major result in hippocampal CA1 neurons. The experiments intended for investigation of mechanisms for our major results were performed with the HEK cells. To summarize, our experiment  71  conditions were optimized to reproduce the physiological condition in live neurons for the calcium channel characterization.  L-type calcium channels are tuned to strong synaptic activities In this discussion chapter, unless otherwise pointed out, all the data presented  refer  to  those  obtained  from  experiments  at  physiological  temperature and with 2 mM bathing calcium. Our results demonstrated that the L-type calcium channels are tuned to strong synaptic activities. In another word, the L-type calcium channels respond with distinctively larger calcium influx in response to the strong synaptic inputs in comparison with other two channel types but not in response to the weak synaptic waveforms. This activity tuning of L-type calcium channels is manifested in two folds. First, the L-type calcium channels prefer the long-lasting EPSP-like depolarization waveform over the AP firing waveform. Secondly, the L-type calcium channels mediate much larger calcium influx in response to strong synaptic activities compared to other two HVA calcium channel types, the P/Q- and the N-type channels. Among the four simulated waveforms, the plateau, the repetitive firing, the strong theta and the weak theta waveform, the first three waveforms generated 3 folds or more calcium currents via the L-type channels than the ones via the P/Q- and the Ntype channels.  The weak theta waveform yielded no difference in calcium  current among the L-, the P/Q- and the N-type calcium channels.  72  Our initial attempt was to take advantage of the  depolarization  waveforms that were previously simulated with the 'NEURON' program in our laboratory and demonstrate whether the simulation results hold in real calcium current recording. This simulation results showed that the L-type calcium channels are able to discriminate normal synaptic activities from those linked to plasticity and gene expression. This discrimination ability comes from the Ltype "channels' preference for the long-lived synaptic depolarization over the AP train waveform. The waveform preference is indeed attractive due to the following causes. First, the EPSP-like waveform in general represents the collective effect of the strong synaptic inputs that depolarize the membrane potential high enough to inactivate the fast inactivating-sodium channels thereby diminishing the APs (40). It is the strong synaptic inputs that are relevant in initiation of long-term changes in neuronal function and structure. Secondly, L-type calcium channels have been shown to play an important role in triggering gene expression (58, 73). The preference of L-type calcium channels for waveforms that represent strong synaptic inputs may reflect the unique role of the L-type calcium channel in gene expression initiation. However, caution needs to be taken for comparing the effect of EPSPlike depolarization waveform and AP train-like waveform. It would be obviously untenable to compare an EPSP with a single AP for their effects due to the large difference in the total integrated sizes. What about two APs? And three? What is the standard to make a legitimate comparison? The comparison is hard to make due to a number of reasons. First, what is the reasonable depolarization level for the EPSP for this comparison given that the AP depolarization level is naturally restricted to about 0 to +30 mV? Another 73  factor is even harder to standardize, the AP firing frequency. The comparison is plausible only when the firing frequency serves as a variable as in the study by Patil et al for analyzing the intermediate closed state inactivation (64). However, the problem becomes apparent when the comparison is made for the analysis of calcium influx where the firing frequency is not a variable as in the case of the study .performed by Mermelstain et al (55). In their study, frequencies at 5 Hz for 180 seconds or 50 Hz for 18 seconds were delivered to cultured hippocampal neurons to elicit calcium transient. The comparisons were made with these stimulation frequencies between with and without EPSP using glutamate receptor blockade. Extrapolation for higher frequencies is difficult to make since the inactivation profile may well differ among different firing frequencies. Our data also show that the P/Q-type channels also manifest a preference for the plateau waveform at physiological temperature. Therefore, the L-type calcium channel's preference for the strong synaptic activities, even if fully confirmed, is not distinctive. However, we do not argue against the possibility that an optimal although not ideal protocol can be attained for a particular purpose at-a particular experimental condition. Unfortunately, this comparison was made in our results without a standardized criterion between calcium channels' preference for waveforms. A very recent report indicated that somatic APs are capable of activating the signaling cascades that are involved in late-phase LTP (24), suggesting that the APs suffice to induce the intracellular calcium elevation that is needed for gene expression and what matters is the total amount of calcium that enters the cytosol. To conclude, the possibility still exists that L-type calcium channels are specialized at responding to the long-lived EPSP-like depolarization 74  patterns. However, to make a solid conclusion for waveform preference, a stricter standard is yet to be carried out. Our major finding is that L-type calcium channels are selectively tuned to the strong synaptic activities but not the weak ones. In response to the strong synaptic waveforms (the repetitive firing, the plateau and the strong theta waveforms), the L-type calcium channels mediated much larger calcium current than those mediated via the P/Q- and the N-type  channels.  In  contrast, the weak theta waveform yielded no difference in calcium current among the three channel types studied. The difference in the responses to the strong synaptic waveforms is quite similar among the three channel types studied despite the large difference in their overall synaptic conductance as mentioned in Method section. It is surprising that this phenomenon has never been reported before given that the single channel conductance is larger in the brain-derived L-type channels than the P/Q- and N-type channels'. We think our observation with the macroscopic current is not a simple expansion of what occurs at single channel level. First, the difference in single channel conductance is much smaller among different HVA calcium channel types. Second, the distinction in among different channel types at macroscopic level can only be observed when the channels are subjected to the strong synaptic inputs, suggesting that the macroscopic behavior of calcium channels can not be fully predicted by single channel characteristics. If the comparison of the preference for waveforms is intended to provide some mechanistic link between L-type calcium channels and the gene expression, the superiority of L-type channels in response to the strong synaptic waveforms over other HVA channel types would be more relevant. I t  75  suggests that only the L-type calcium channels on the soma can provide sufficient amount of calcium to reach nucleus and elicit gene expression that is needed for the long-term modification of neuronal responsiveness.  Activation kinetics are not attributable for distinction of calcium channel behavior There has been a general agreement that the L-type calcium channels have slow activation kinetics with little inactivation, which makes it an ideal filter for APs (14, 66). These attributes endow the L-type channels with a preference for the long-lasting EPSP waveforms (55). However, the conclusion was drawn from experiments performed at room temperature and mostly with high concentration of barium as permeant ion. As discussed earlier, these nonphysiological factors distort our understanding of the properties of calcium channels. Our results were obtained with conditions that corrected all the inappropriate experimental conditions and showed that the activation of L-type channels was no longer slow, and the inactivation of L-type channels was no longer insignificant. Instead, all HVA channel types activated with a similar rate and inactivate to a similar extent with the L-type channels' steady state level slightly lower than the other two HVA calcium channel types (see table 2). Activation comparison indicated that the L-type calcium channels' rise time was slightly longer than the P/Q- and N-type channels. Yet comparison of their response to a series of mock APs with varied duration showed that all HVA channel types were equally capable of responding to APs with full capacity,  76  suggesting that activation kinetics can not serve to distinguish the different behaviors of calcium channels. Comparison of normalized calcium current in response to AP in the simulated  waveforms  indicated that  the  normalized  AP response  was  significantly smaller in the P/Q-type than that in the L- and N-type channels. Given that the L-type activation was slightly slower than the P/Q- and N-type channels, this observation argues against the proposal that calcium current response to AP is mainly shaped by activation kinetics. Carefully looking into the calcium current in response to an AP, we found that the AP current response was composed of mainly the tail current, which is shaped by inactivation status of the calcium channels. This suggests that at physiological temperature, even the APs are not formed  by activation  kinetics as  traditionally thought to be. This is consistent with the previous report that the L-type calcium channels exhibited prolonged activity upon depolarization (68). Therefore, activation kinetics can not be held accountable for the large difference in their total calcium current in response to the strong synaptic depolarizations.  Inactivation is critical in functional differentiation of calcium channels Our results indicated that inactivation plays a crucial role in shaping the calcium currents in response to synaptic depolarization patterns.. It can be examined from aspects as indicated below.  77  Residual  calcium  current  Our results indicated that the inactivation revealed from residual current level appears larger in the L-type calcium channels than those in the P/Q- and N-type channels. At room temperature, the residual levels measured at the end of 1 second depolarization are 15%, 25% and 13% for the L-, P/Q- and Ntype calcium channels respectively. This is inconsistent to the prevalent notion that L-type channels have little inactivation. One of the reason is that the statement  about the characteristics  of L-type  channels  was  made  in  comparison with N-type channels since the P- and Q-type calcium channels . were discovered at a much later stage (51, 57, 74, 97). This may be another commonly neglected aspect in our understanding of calcium channels and may have some interesting implication in that the considerable difference in inactivation profile (at room temperature) occurs in two channel types that perform  similar  cellular  tasks,  i.e.  neurotransmitter  release.  Once  at  physiological temperature, the inactivation was drastically accelerated (6%, 3% and 2% for the L-, P/Q- and N-type channels respectively). The resemblance of inactivation extent between the P/Q- and N-type channels is consistent to their functional similarity but presents a contrast to the L-type inactivation at room temperature. As the numbers shown, the inactivation level of the P/Q-type channels resembles that of the N-type channels. This is consistent with their functional similarity yet presents a contrast to what was observed at room temperature with respect to their inactivation comparison between the L- and P/Q-type channels. It suggests that the channel behavior at physiological temperature is not a simple amplification of what was  78  observed at room temperature and therefore it is not appropriate to apply Q i0 to  a  particular  temperature-sensitive  process  in  a  temperature.  At  physiological temperature, it is highly possible that there are different sets of enzyme activated for calcium channel modulation.  Inactivation  revealed from spiking  attenuation  Inactivation revealed from comparison in the amplitude of the last AP to the first AP in the repetitive firing waveform indicated that inactivation of the L-type calcium channels was much less than the P/Q- and N-type channels. In response to the 1 second pulse at physiological temperature as shown above, the difference in the proportion of inactivation between the L- and .P/Q-type channels is only 3%, while in response to the repetitive firing waveform the difference in the spike attenuation between the two channel types becomes 33%. Certainly the inactivation revealed from step depolarization can not fully account for the inactivation difference revealed from the complex synaptic waveform. Up to date, there are two kinds of inactivation that are responsible for calcium current decay. One is calcium-dependent channel inactivation and the other is voltage-dependent inactivation. Analysis of the contribution of the two types of inactivation requires consideration of the channel type and depolarization  potential.  The  L-type  inactivation  is  believed  to  be  predominantly calcium-dependent while the P/Q- and N-type inactivation are largely voltage-dependent (9, 31, 36). At negative potentials the calcium current decay is dominated by calcium dependent inactivation since voltagedependent inactivation does not occur until potential reaches more depolarized level ( 3 1 , 62). In contrast, at more positive potentials the continuing calcium 79  current decay would be comprised of largely calcium-dependent inactivation because voltage-dependent inactivation has already been completed. The time course of calcium-dependent inactivation and the separation of the inactivation components have been reported (10, 44, 69). However, there may well be drastic difference in proportion of two types of inactivation for results obtained at  physiological  temperature  since  their  results  were  obtained  from  experiments at room temperature. And no evidence has been reported with regard to the relative proportion of voltage-dependent inactivation versus calcium-dependent inactivation, which is crucial in understanding the calcium channel activities.  Channel availability Our  data  at the end of synaptic  showed that  the  waveforms  L-type calcium  channels  inactivated  considerably less than the P/Q- and N-type channels in synaptic waveforminduced responses as discussed above. The channel status at the end of the simulated firing waveforms remained unknown. Therefore, a 5 ms step pulse with +25 mV depolarization level was placed digitally at the end of the repetitive firing and the plateau waveforms with an attempt to fully open the channels remaining available and test the inactivation status. The 5 ms endpulse unraveled intriguing inactivation behaviour that was different from those manifested from the original APs. In the case of the L- and P/Q-type channels, they both present a consistent end-pulse response with their AP responses, in which the L-type channels respond with a large end-pulse response and the P/Q-type responds with current no larger than those in AP responses. In contrast, the N-type channels exhibited an interestingly large end-pulse 80  response as compared to the AP response, in which the end-pulse response exceeds the AP response by more than 100 folds as shown from the example trace in fig. 15. For the N-type channels, there seems to be an interesting voltage-dependent activity going on that is affecting the channel availability at the end of the firing. It may be partly because at the end of the repetitive firing, the N-type channels entered an voltage-dependent inactivation state in which the channels may be fully re-opened by a higher depolarization pulse. The difference in the amplitude between the end-pulse response and the first AP response may be due to the calcium-dependent component of its total inactivation. This is yet to be tested experimentally. Another alternative explanation is the rate of channel recovery. L-type and N-type channels may represent calcium channels that have rapid channel recovery rate as shown from our data. Lack of either small inactivation or rapid recovery may lead to drastic reduction in calcium influx as seen in the P/Q- and N-type channels. This partly helps to provide some mechanistic explanation of the differential response to the synaptic waveforms among the three calcium channel types studied. However, the precise mechanism for the channel inactivation recovery postulation has yet to be explored. In the repetitive firing waveform, the inactivation observed from reduction of AP response at the end of the waveform is possibly due to the spiking  during  the  waveform,  given  that  the  inactivation  from  step  depolarization is much smaller in either of the three channel types. Removal of a cluster of APs diminished the inactivation and induced a recovery from inactivation in the L-type calcium channels as shown in fig. 17. For the P/Qtype channels, inactivation proceeded after the firing gap, supporting the 81  proposal that the L-type channels have a faster recovery from inactivation than P/Q-type channels. The apparent separation of inactivation behavior manifested by the AP and the end-pulse suggests that the inactivation development and the recovery from inactivation may be governed by distinct mechanisms.  Inter-spike channel  potential  is important  for voltage-dependent  calcium  inactivation  The 5 ms end-pulse data implies that there is a considerable portion of voltage-dependent component in the inactivation, especially for the P/Q- and N-type channels. The L- and P/Q-type calcium channels were chosen for further investigation of the voltage-dependence of the inactivation manifested in the repetitive waveform response. As shown in fig. 17, removal of the EPSP component caused opposite changes in calcium current, a decrease for the Ltype channels and an increase for the P/Q-type channels in response to the repetitive firing waveform. A similar increase is observed for the N-type channels. Although dramatically reduced compared to the group with the EPSP component intact, the difference in the calcium current in response to the repetitive firing'waveform remained larger for the L-type channels compared to the P/Q- and N-type channels. It suggests that the EPSP component or inter-spike potential contribute a great deal to the privileged calcium, influx of the L-type calcium channels in response to the strong synaptic depolarization patterns. A recent report demonstrated that a significant portion of voltagedependent inactivation, although not as large as the calcium-dependent  82  inactivation component, was observed with a fast and a slow component (75). This explains partly why the EPSP removal did not enhance the calcium current in response to the repetitive firing waveform for the L-type calcium channels as for the P/Q- and N-type channels. On the other hand, the calcium current reduction upon EPSP removal in the P/Q- and N-type channels suggests that P/Q- and N-type channels mainly undergo voltage-dependent inactivation with possibly only a negligible proportion of calcium-dependent inactivation. For the P/Q-type channels, examination of inactivation revealed by reduction of AP response showed that inactivation was greatly reduced (by about half) with -70 mV inter-spike potential. The inactivation for the P/Qtype channels was almost completely abolished (reduced by about 85%) when the inter-spike potential further lowered to -110 mV, further suggesting that the voltage-dependent inactivation constitutes the major portion of calcium current decay for the P/Q-type channels. It also provides us a basis for a quantitative estimation of the proportion of calcium-dependent and voltagedependent inactivation. We expect the N-type channels to behave similarly to P/Q-type channels. However, this work has yet to be accomplished.  p2A subunit reverses inactivation in the P/Q-type  channels  Consistent with previous findings that calcium channel with p2A has a substantial slowing effect on calcium channel inactivation (25, 79), our experiments showed that p2A greatly reduced inactivation observed in both step pulse response and the waveform response. Our results further suggest that p2A subunit mainly removes the voltage-dependent component of inactivation without interference of calcium-dependent  inactivation. This 83  postulation is extrapolated from observations in which p2A did not appreciably affect the inactivation revealed by AP response in the L-type channels while almost completely reversed the inactivation in the P/Q-type channels. Taking together the previous studies and our results, p2A subunit appears to largely remove the voltage-dependent component of inactivation.  84  Summary Combining all our results, we propose a characteristic set of properties exhibited by the L-type calcium channel. First, L-type VSCCs are selectively tuned to the strong synaptic input to mediate the calcium influx that is responsible for triggering gene expression, which provides the structural basis for neuronal plasticity and hence memory formation. In contrast, P/Q- and Ntype channels have evolved probably only for providing calcium influx that is required for transmitter release, which is moderate given the spatial limitation at nerve terminals. Secondly, at physiological temperature, activation kinetics do not contribute significantly to any known calcium channel  behaviour.  Thirdly, inactivation characteristics play a crucial role in calcium  channel  behaviours. Fourthly, combined with previous studies, our results yield an approximation of proportion of inactivation type for a given calcium channel type. For L-type channels, there is approximately a 7 0 % calcium-dependent inactivation component with about 3 0 % voltage-dependent inactivation. For P/Q-type channels, there is about 8 5 % voltage-dependent inactivation and 1 5 % calcium-dependent inactivation. For N-type channels, the proportion for voltage-dependent inactivation versus calcium-dependent inactivation is about 9 5 % compared to 5%. inactivation  Finally, P/Q-type calcium channels exhibit a unique  profile different  from  N-type channels and  L-type  channels,  although they have similar functions to N-type channels.  85  Future directions Our results present a new profile for the HVA calcium channel behavior at  physiological  conditions. The * physiological  waveforms  are capable  of  displaying real calcium channel behavior as shown from our data. However, the characterization of these channel types needs to be refined and measured in a quantitative manner. Based on our findings and previous studies, I propose following studies to be carried out in the future. 1) Comparison of our findings using the physiological conditions with previous studies with  non-physiological  conditions  indicates that  calcium  channels need to be characterized in physiological conditions. The crucial parameters for physiological conditions includes physiological temperature, appropriate permeant ion, physiological concentration of the permeant ion. Under these conditions, L-, P/Q- N- and T-type calcium channels will be investigated  with  waveforms  composed  of  single  or  multiple  square  depolarization pulse for analysis of several important of calcium  channel  behavior. These investigations will include quantitation of the portion of calcium-dependent and voltage-dependent inactivation, re-examination of the intermediate close-state inactivation as proposed by Yue group (64), reevaluation of prepulse facilitation. Experiments will include calcium channel modulation by p subunit. 2) Work by Plummer and Hess (69) show that with single channel recording, there seem to be two groups of N-type calcium channels with distinct inactivation profile, one with fast inactivation with about 40 ms time course and the other with slow inactivation with about 1 second time course. A  86  recent report show that the voltage-dependent inactivation in the L-type calcium channels exhibit two inactivation components, a fast component and a slow component (75). Changing the slow component affects the time course of the fast component, suggesting the two components are intrinsically coupled. Taking their results together with our data, I hypothesize that for each calcium channel type there are two modes of voltage-dependent inactivation that are inter-convertible,  a fast  inactivation  and slow  inactivation  components  depending upon the activation status of the channels. The inter-conversion of two inactivation modes serve distinctive functional aspect of the channels. We plan to use single channel recording technique to test this hypothesis in heterologous system expressing either L, P/.Q- or N-type calcium channels. The prerequisite for mode switch will be closely examined.  87  Reference:  1. Acerbo, P., and M. Nobile. Temperature dependence of multiple high voltage activated Ca2+ channels in chick sensory neurones. Eur Biophys J 23: 189-95, 1994. 2. Ahlijanian, M. K., R. E. Westenbroek, and W. A. Catterall. Subunit structure and localization of dihydropyridine-sensitive calcium channels in mammalian brain, spinal cord, and retina. Neuron 4 : 819-32., 1990. 3. Augustine, G. J., M. P. Charlton, and S. J. Smith. Calcium action in synaptic transmitter release. Annu Rev Neurosci 10: 633-93, 1987. 4. Bacskai, B. J., B. Hochner, M. Mahaut-Smith, S. R. Adams, B. K. Kaang, E. R. Kandel, and R. Y. Tsien. Spatially resolved dynamics of cAMP and protein kinase A subunits in Aplysia sensory neurons. Science 260: 222-6, 1993. 5. Bear, M. F., and R. C. Malenka. Synaptic plasticity: LTP and LTD. Curr Opin Neurobiol 4 : 389-99., 1994. 6. Berdan, R. C , J. C. Easaw, and R. Wang. Alterations in membrane potential after axotomy at different distances from the soma of an identified neuron and the effect of depolarization on neurite outgrowth and calcium channel expression. J Neurophysiol 69: 151-64., 1993. 7. Bliss, T. V., and T. Lomo. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J Physiol 232: 331-56., 1973. . 8. Bourinet, E., P. Charnet, W. J. Tomlinson, A. Stea, T. P. Snutch, and J. Nargeot. Voltage-dependent facilitation of a neuronal alpha 1C L-type calcium channel. EmboJ 13: 5032-9, 1994. 9. Branchaw, J. L., M. I. Banks, and M. B. Jackson. Ca2+- and voltage-dependent inactivation of Ca2+ channels in nerve terminals of the neurohypophysis. J Neurosci 17: 5772-81'., 1997. 10. Brehm, P., and R. Eckert. Calcium entry leads to inactivation of calcium channel in Paramecium. Science 202: 1203-6., 1978. 11. Brown, J. R., H. Ye, R. T. Bronson, P. Dikkes, and M. E. Greenberg. A defect in nurturing in mice lacking the immediate early gene fosB. Cell 86: 297-309, 1996. 12. Buzsaki, G. Theta oscillations in the hippocampus. Neuron 33: 325-40., 2002. 13. Catterall, W. A. Structure and regulation of voltage-gated Ca2+ channels. Annu Rev Cell Dev Biol 16: 521-55, 2000. 14. Charnet, P., E. Bourinet, S. J. Dubel, T. P. Snutch, and J. Nargeot. Calcium currents recorded from a neuronal alpha 1C L-type calcium channel in Xenopus oocytes. FEBS Lett 344: 87-90, 1994. 15. Christie, B. R., D. Stellwagen, and W. C. Abraham. Reduction of the threshold for long-term potentiation by prior theta- frequency synaptic activity. Hippocampus 5: 52-9, 1995. 16. Clapham, D. E. Calcium signaling. Cell 80: 259-68., 1995. 17. Colbert, C. M., J. C. Magee, D. A. Hoffman, and D. Johnston. Slow recovery from inactivation of Na+ channels underlies the activity- dependent attenuation of dendritic action potentials in hippocampal CA1 pyramidal neurons. J Neurosci 17: 6 5 1 2 - 2 1 . , 1997. 18. Connor, J. R., and M. C. Diamond. A comparison of dendritic spine number and type on pyramidal neurons of the visual cortex of old adult rats from social or isolated environments. J Comp Neurol 210: 99-106., 1982.  88  19. Currie, K. P., and A. P. Fox. Comparison of N- and P/Q-type voltage-gated calcium channel current inhibition. J Neurosci 17: 4570-9, 1997. 20. Day, N. C , S. G. Volsen, A. L. McCormack, P. J. Craig, W. Smith, R. E. Beattie, P. J. Shaw, S. B. Ellis, M. M. Harpold, and P. G. Ince. The expression of voltagedependent calcium channel beta subunits in human hippocampus. Brain Res Mol Brain Res 60: 259-69. bin/cas/tree/store/bresm/cas_sub/browse/browse.cgi?year=1998&volume=60&i ssue=2&aid=71930, 1998. 21. De Waard, M., M. Pragnell, and K. P. Campbell. Ca2+ channel regulation by a conserved beta subunit domain. Neuron 13: 495-503., 1994. 22. Deisseroth, K., H. Bito, and R. W. Tsien. Signaling from synapse to nucleus: postsynaptic CREB phosphorylation during multiple forms of hippocampal synaptic plasticity. Neuron 16: 8 9 - 1 0 1 , 1996. 23. Dolphin, A. C. L-type calcium channel modulation. Advances in Second messenger and phosphoprotein research 33: 153-177, 1999. 24. Dudek, S. M., and R. D. Fields. Somatic action potentials are sufficient for latephase LTP-related cell signaling. Proc Natl Acad Sci USA 99: 3962-7., 2002. 25. Ellinor, P. T., J. F. Zhang, A. D. Randall, M. Zhou, T. L. Schwarz, R. W. Tsien, and W. A. Home. Functional expression of a rapidly inactivating neuronal calcium channel. Nature 363: 455-8., 1993. 26. Elliott, E. M., A. T. Malouf, and W. A. Catterall. Role of calcium channel subtypes in calcium transients in hippocampal CA3 neurons. J Neurosci 15: 6433-44., 1995. 27. Ertel, E. A., K. P. Campbell, M. M. Harpold, F. Hofmann, Y. Mori, E. PerezReyes, A. Schwartz, T. P. Snutch, T. Tanabe, L. Birnbaumer, R. W. Tsien, and W. A. Catterall. Nomenclature of voltage-gated calcium channels [letter]. Neuron 25: 533-5, 2000. 28. Finkbeiner, S., and M. E. Greenberg. Ca2+ channel-regulated neuronal gene expression. J Neurobiol 37: 1 7 1 - 8 9 , 1 9 9 8 . 29. Fischbach, G. D., and Y. Lass. A transition temperature for acetylcholine channel conductance in chick myoballs. J Physiol 280: 527-36., 1978. 30. Fox, A. P., M. C. Nowycky, and R. W. Tsien. Single-channel recordings of three types of calcium channels in chick sensory neurones. J Physiol 394: 173-200., 1987. 31. Giannattasio, B., S. W. Jones, and A. Scarpa. Calcium currents in the A7r5 smooth muscle-derived cell line. Calcium- dependent and voltage-dependent inactivation. J Gen Physiol 98: 987-1003., 1991. 32. Globus, A., and A. B. Scheibel. The effect of visual deprivation on cortical neurons: a Golgi study. Exp Neurol 19: 331-45., 1967. 33. Guia, A., M. D. Stern, E. G. Lakatta, and I. R. Josephson. Ion concentrationdependence of rat cardiac unitary L-type calcium channel conductance. Biophys J 80: 2742-50., 2 0 0 1 . 34. Gurnett, C. A., M. De Waard, and K. P. Campbell. Dual function of the voltagedependent Ca2+ channel alpha 2 delta subunit in current stimulation and subunit interaction. Neuron 16: 431-40., 1996. 35. Gustafsson, B., and H. Wigstrom. Physiological mechanisms underlying longterm, potentiation. Trends Neurosci 1 1 : 156-62., 1988. 36. Gutnick, M. J., H. D. Lux, D. Swandulla, and H. Zucker. Voltage-dependent and calcium-dependent inactivation of calcium channel current in identified snail neurones. J Physiol (Lond) 412: 197-220, 1989. 37. Hardingham, G. E., S. Chawla, C. M. Johnson, and H. Bading. Distinct functions of nuclear and cytoplasmic calcium in the control of gene expression. Nature 385: 260-5,1997. 38. Heidelberger, R., and G. Matthews. Calcium influx and calcium current in single synaptic terminals of goldfish retinal bipolar neurons. J Physiol 447: 235-56., 1992.  89  39. Heidelberger, R., and G. Matthews. Inhibition of calcium influx and calcium current by gamma-aminobutyric acid in single synaptic terminals. Proc Natl Acad Sci U SA 88: 7135-9., 1 9 9 1 . 40. HiMe, B. Ionic channels of excitable membranes. Sinauer Asociates Inc. Publishers 2nd edition, 1994. 41. Hodgkin, A. L , and B. Katz. The effect of temperature on the electrical activity of the giant axon of the squid. Journal of physiology 109: 240-249, 1949. 42. Hofmann, F., L. Lacinova, and N. Klugbauer. Voltage-dependent calcium channels: from structure to function. Rev Physiol Biochem Pharmacol 139: 33-87, 1999. 43. Impey, S., M. Mark, E. C. Villacres, S. Poser, C. Chavkin, and D. R. Storm. Induction of CRE-mediated gene expression by stimuli that generate long- lasting LTP in area CA1 of the hippocampus. Neuron 16: 973-82, 1996. 44. Imredy, J. P., and D. T. Yue. Mechanism of Ca(2+)-sensitive inactivation of Ltype Ca2+ channels. Neuron 12: 1301-18, 1994. 45. Jones, S. W. Overview of voltage-dependent calcium channels. J Bioenerg Biomembr 30: 299-312, 1998. 46. Jung, H. Y., N. P. Staff, and N. Spruston. Action potential bursting in subicular pyramidal neurons is driven by a calcium tail current. J Neurosci 2 1 : 3 3 1 2 - 2 1 . , 2 0 0 1 . 47. Kenyon, J. L , and H. R. Goff. Temperature dependencies of Ca2+ current, Ca(2+)-activated CI- current and Ca2+ transients in sensory neurones. Cell Calcium 24: 35-48., 1998. 48. Lacerda, A. E., H. S. Kim, P. Ruth, E. Perez-Reyes, V. Flockerzi, F. Hofmann, L. Birnbaumer, and A. M. Brown. Normalization of current kinetics by interaction between the alpha 1 and beta subunits of the skeletal muscle dihydropyridine-sensitive Ca2+ channel. Nature 352: 527-30., 1991. 49. Letts, V. A., R. Felix, G. H. Biddlecome, J. Arikkath, C. L. Mahaffey, A. Valenzuela, F. S. Bartlett, 2nd, Y. Mori, K. P, Campbell, and W. N. Frankel. The mouse stargazer gene encodes a neuronal Ca2+-channel gamma subunit. Nat Genet 19: 3407., 1998. 50. Li, G. R., and S. Nattel. Properties of human atrial ICa at physiological temperatures and relevance to action potential. Am J Physiol 272: H227-35., 1997. 51. Llinas, R., M. Sugimori, J. W. Lin, and B. Cherksey. Blocking and isolation of a calcium channel from neurons in mammals and cephalopods utilizing a toxin fraction (FTX) from funnel-web spider poison. Proc Natl Acad Sci U S A 86: 1689-93., 1989. 52. Ludwig, A., V. Flockerzi, and F. Hofmann. Regional expression and cellular localization of the a l p h a l and beta subunit of high voltage-activated calcium channels in rat brain. J Neurosci 17: 1339-49., 1997. 53. Martin, K. C , D. Michael, J. C. Rose, M. Barad, A. Casadio, H. Zhu, and E. R. Kandel. MAP kinase translocates into the nucleus of the presynaptic cell and is required for long-term facilitation in Aplysia. Neuron 18: 899-912, 1997. 54. Melchior, D. L., and J. M. Steim. Thermotropic transitions in biomembranes. Annu Rev Biophys Bioeng 5: 205-38, 1976. 55. Mermelstein, P. G., H. Bito, K. Deisseroth, and R. W. Tsien. Critical dependence of cAMP response element-binding protein phosphorylation oh L-type calcium channels supports a selective response to EPSPs in preference to action potentials. J Neurosci 20: 266-73, 2000. 56. Mickus, T., H. Y. Jung, and N. Spruston. Slow sodium channel inactivation in CA1 pyramidal cells. Ann N Y Acad Sci 868: 9 7 - 1 0 1 , 1999. 57. Mintz, I. M., M. E. Adams, and B. P. Bean. P-type calcium channels in rat central and peripheral neurons. Neuron 9: 85-95., 1992. 58. Murphy, T. H., P. F. Worley,. and J. M. Baraban. L-type voltage-sensitive calcium channels mediate synaptic activation of immediate early genes. Neuron 7: 625-35, 1 9 9 1 .  90  59. Nakazawa, H., and T. H. Murphy. Activation of nuclear calcium dynamics by synaptic stimulation in cultured cortical neurons. J Neurochem 73: 1075-83, 1999. 60. Nguyen, P. V., T. Abel, and E. R. Kandel. Requirement of a critical period of transcription for induction of a late phase of LTP. Science 265: 1104-7, 1994. 61. Nicoll, R. A., J. A. Kauer, and R. C. Malenka. The current excitement in longterm potentiation. Neuron 1 : 97-103., 1988. 62. Obejero-Paz, C. A., S. W. Jones, and A. Scarpa. Calcium currents in the A7r5 smooth muscle-derived cell line. Increase in current and selective removal of voltagedependent inactivation by intracellular trypsin. J Gen Physiol 98: 1127-40., 1 9 9 1 . 63. Parnavelas, J. G., A. Globus, and P. Kaups. Continuous illumination from birth affects spine density of neurons in the visual cortex of the rat. Exp Neurol 4 0 : 742-7., 1973. 64. Patil, P. G., D. L. Brody, and D. T. Yue. Preferential closed-state inactivation of neuronal calcium channels. Neuron 20: 1027-38, 1998. 65. Patil, P. G., M. de Leon, R. R. Reed, S. Dubel, T. P. Snutch, and D. T. Yue. Elementary events underlying voltage-dependent G-protein inhibition of N-type calcium channels. Biophys J 7 1 : 2 5 0 9 - 2 1 . , 1996. 66. Perez-Garcia, M. T., T. J. Kamp, and E. Marban. Functional properties of cardiac L-type calcium channels transiently expressed in HEK293 cells. Roles of alpha 1 and beta subunits. J Gen Physiol 105: 289-305., 1995. 67. Perez-Reyes, E., H. S. Kim, A. E. Lacerda, W. Home, X. Y. Wei, D. Rampe, K. P. Campbell, A. M. Brown, and L. Birnbaumer. Induction of calcium currents by the expression of the alpha l-subunit of the dihydropyridine receptor from skeletal muscle. Nature 340: 233-6., 1989. 68. Pietrobon, D., and P. Hess. Novel mechanism of voltage-dependent gating in Ltype calcium channels. Nature 346: 651-5., 1990. 69. Plummer, M. R., and P. Hess. Reversible uncoupling of inactivation in N-type calcium channels. Nature 3 5 1 : 657-9., 1 9 9 1 . 70. Plummer, M. R., D. E. Logothetis, and P. Hess. Elementary properties and pharmacological sensitivities of calcium channels in mammalian peripheral neurons. Neuron 2: 1453-63., 1989. 71. Ren, J., Z. Liu, S. Wang, and T. Murphy. Modeling nuclear calcium transients mediated by synaptically activated L-type calcium channels. Neuroscience conference abstract, 2000. 72. Rosen, A. D. Temperature modulation of calcium channel function in GH3 cells. Am J Physiol 2 7 1 : C863-8., 1996. 73. Rosen, L. B., D. D. Ginty, and M. E. Greenberg. Calcium regulation of gene expression. Adv Second Messenger Phosphoprotein Res 30: 225-53, 1995. 74. Sather, W. A., T. Tanabe, J. F. Zhang, Y. Mori, M. E. Adams, and R. W. Tsien. Distinctive biophysical and pharmacological properties of class A (BI) calcium channel alpha 1 subunits. Neuron 1 1 : 291-303., 1993. 75. Shi, C , and N. M. Soldatov. Molecular determinant of voltage dependent slow inactivation of Ca2+ channel. J Biol Chem 18: 18, 2 0 0 1 . 76. Singer, D., M. Biel, I. Lotan, V. Flockerzi, F. Hofmann, and N. Dascal. The roles of the subunits in the function of the calcium channel. Science 253: 1553-7., 1991. 77. Soldatov, N. M., M. Oz, K. A. O'Brien, D. R. Abernethy, and M. Morad. Molecular determinants of L-type Ca2+ channel inactivation. Segment exchange analysis of the carboxyl-terminal cytoplasmic motif encoded by exons 40-42 of the human alphalC subunit gene. J Biol Chem 273: 957-63., 1998. 78. Stea, A., S. J. Dubel, M. Pragnell, J. P. Leonard, K. P. Campbell, and T. P. Snutch. A beta-subunit normalizes the electrophysiological properties of a cloned Ntype Ca2+ channel alpha l-subunit. Neuropharmacology 32: 1103-16, 1993. 79. Stea, A., W. J. Tomlinson, T. W. Soong, E. Bourinet, S. J. Dubel, S. R. Vincent, and T. P. Snutch. Localization and functional properties of a rat brain alpha 1A calcium  91  channel reflect similarities to neuronal Q- and P-type channels. Proc Natl Acad Sci U 5 A 9 1 : 10576-80, 1994. '' ._ 80. S t o t z , S . C , and G. W. Zamponi. Structural determinants of fast inactivation of high voltage-activated Ca(2+) channels. Trends Neurosci 24: 1 7 6 - 8 1 . , 2 0 0 1 . 81. Sutton, K. G., J. E. McRory, H. Guthrie, T. H. Murphy, and T. P. Snutch. P/Qtype calcium channels mediate the activity-dependent feedback of" syntaxin-lA. Nature 4 0 1 : 800-4, 1999. 82. Takahashi, M., M. J. Seagar, J. F. Jones, B. F. Reber, and W. A. Catterall. Subunit structure of dihydropyridine-sensitive calcium channels from skeletal muscle. Proc Natl Acad Sci U S A 84: 5478-82., 1987. 83. Takahashi, T., and A. Momiyama. Different types of calcium channels mediate central synaptic transmission. Nature 366: 156-8., 1993. 84. Tanabe, T., H. Takeshima, A. Mikami, V. Flockerzi, H. Takahashi, K. Kangawa, M. Kojima, H. Matsuo, T. Hirose, and S. Numa. Primary structure of the receptor for calcium channel blockers from skeletal muscle. Nature 328: 313-8., 1987. 85. Teyler, T. J., and P. DiScenna. Long-term potentiation. Annu Rev Neurosci 10: 1 3 1 - 6 1 , 1987. 86. Thomas, M. J., A. M. Watabe, T. D. Moody, M. Makhinson, and T. J. O'Dell. Postsynaptic complex spike bursting enables the induction of LTP by theta frequency synaptic stimulation. J Neurosci 18: 7118-26, 1998. 87. Valverde, F. Apical dendritic spines of the visual cortex and light deprivation in the mouse. Exp Brain Res 3: 337-52, 1967. 88. Valverde, F. Rate and extent of recovery from dark rearing in the visual cortex of the mouse. Brain Res 33: 1-11., 1971. 89. Watanabe, Y., R. S. Johnson, L. S. Butler, D. K. Binder, B. M. Spiegelman, V. E. Papaioannou, and J. O. McNamara. Null mutation of c-fos impairs structural and functional plasticities in the kindling model of epilepsy. J Neurosci 16: 3827-36, 1996. 90. Westenbroek, R. E., M. K. Ahlijanian, and W. A. Catterall. Clustering of L-type Ca2+ channels at the base of major dendrites in hippocampal pyramidal neurons. Nature 347: 281-4, 1990. 91. Westenbroek, R. E., J. W. Hell, C. Warner, S. J. Dubel, T. P. Snutch, and W. A. Catterall. Biochemical properties and subcellular distribution of an N-type calcium channel alpha 1 subunit. Neuron 9: 1099-115., 1992. 92. Westenbroek, R. E., T. Sakurai, E. M. Elliott, J. W. Hell, T. V. Starr, T. P. Snutch, and W. A. Catterall. Immunochemical identification and subcellular distribution of the alpha 1A subunits of brain calcium channels. J Neurosci 15: 6403-18, 1995. 93. Witcher, D. R., M. De Waard, and K. P. Campbell. Characterization of the purified N-type Ca2+ channel and the cation sensitivity of omega-conotoxin GVIA binding. Neuropharmacology 32: 1127-39., 1993. 94. Yang, N., and R. Horn. Evidence for voltage-dependent S4 movement in sodium channels. Neuron 15: 213-8., 1995. 95. Yue, D. T., P. H. Backx, and J. P. Imredy. Calcium-sensitive inactivation in the gating of single calcium channels. Science 250: 1735-8, 1990. 96. Zamponi, G. W., and T. P. Snutch. Modulation of voltage-dependent calcium channels by G proteins. Curr Opin Neurobiol 8: 351-6., 1998. 97. Zhang, J. F., A. D. Randall, P. T. Ellinor, W. A. Home, W. A. Sather, T. Tanabe, T. L. Schwarz, and R. W. Tsien. Distinctive pharmacology and kinetics of cloned neuronal Ca2+ channels and their possible counterparts in mammalian CNS neurons. Neuropharmacology 32: 1075-88., 1993. 98. Zuhlke, R. D., G. S. Pitt, K. Deisseroth, R. W. Tsien, and H. Reuter. Calmodulin supports both inactivation and facilitation of L-type calcium channels [see comments]. Nature 399: 159-62, 1999. -  92  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items