Effects of six calcium channel blockers on rat hippocampal cells grown in culture

Material Information

Effects of six calcium channel blockers on rat hippocampal cells grown in culture
Dohanyos, John
Place of Publication:
Denver, CO
University of Colorado Denver
Publication Date:
Physical Description:
100 leaves : ill. ; 29 cm.

Thesis/Dissertation Information

Master's ( Master of Arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Integrative Biology, CU Denver
Degree Disciplines:
Committee Chair:
Audesirk, Gerald
Committee Members:
Audesirk, Terry
Brockway, Alan


Subjects / Keywords:
Calcium -- Antagonists ( lcsh )
Rats -- Experiments ( lcsh )
Calcium -- Antagonists ( fast )
Rats -- Experiments ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Thesis (M.A.)--University of Colorado at Denver, 1994. Biology
Includes bibliographical references (leaves 89-100).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Biology.
General Note:
Department of Integrative Biology
Statement of Responsibility:
by John Dohanyos.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
32585905 ( OCLC )


This item is only available as the following downloads:

Full Text


EFFECTS OF SIX CALCIUM CHANNEL BWCKERS ON RAT HIPPOCAMPAL CELLS GROWN IN CULTURE by John Dohanyos B.A., College, 1972 A thesis submitted to the Faculty of the Graduate School of the University of Colorado at Denver in partial fulfillment of the requirements for the degree of of Arts Biology 1994 @


This thesis for the Master of Arts degree by John Dohanyos has been approved for the Graduate School by Gerald Audesirk Terry Audesirk 1;)../7/c;y I Date


Dohanyos, John (M.A., Biology) Effects of Six Calcium Channel Blockers on Rat Hippocampal Cells Grown in Culture Thesis directed by Associate Professor Gerald Audesirk ABSTRACT Four dihydropyridines, nifedipine, nicardipine, nimodipine, and isradipine and the phenylalklamine verapamil and the benzothiazapine diltiazem were added individually to the medium of cryopreserved, E18 rat hippocampal cells which were then grown in culture for two days. All of these compounds are known to be L-type calcium channel blockers. The effects of the calcium channel blockers on eight growth parameters were assessed after the two day period. Nicardipine and nimodipine reduced all parameters at low micromolar concentrations. Nifedipine seemed to have little effect except at close to 100 concentrations. Nifedipine may have increased axonal elongation. Isradipine reduced all parameters at concentrations of 25 and 50 However, isradipine enhanced all parameters at concentrations of 100 nM and 500 nM, giving isradipine a biphasic characteristic on the growth of rat hippocampal cells. Verapamil and diltiazem reduced growth parameters at concentrations of 25 and 50 JLM respectively. It is suggested that L-type calcium channel blockers reduce growth parameters by reducing calcium


influx into the neurons. Isradipine which may act as an agonist in nanomolar concentrations may increase intracellular calcium and thus stimulate growth in the neuron. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. Gerald Audesirk


This thesis is dedicated to Maureen Hogg who lost both sight and hearing at age fourteen due to an as yet undiagnosed syndrome.


CONTENTS Chapter 1. Introduction ............ .... -..................... 1 1.1 Calcium Channels . . . . . . . . . . . . . . . 2 1.2 Brief Explanation of Procedures ...................... 3 1.3 L-type Voltage-sensitive Calcium Channel Structure ............ 4 1.3.1 The at Subunit . . . . . . . . . ........... ... 6 1.4 Hypothesis for a Mechanism of Ca2+ Channel Opening .......... 6 1.5 Calcium Ion Binding Sites Within the Pore ................. 10 1.6 Dihydropyridine Binding Sites on the at Subunit .............. 11 1. 7 Dihydropyridine Structure . . . . . . . . . . . . . . 11 1. 8 Structure-Activity Relationships of Dihydmpyridines . . . . . 13 1. 8.1 Importance of the N(1)-H Group ...................... 15 1.8.2 Description of Nifedipine Binding to the S4 Helix ............ 16 1. 8.3 A Second Hypothesis for Dihydropyridine Binding . . . . . .17 1.9 Structure-Activity Relationships for Benzothiazipines and Phenylalkylamines . . . . . . . . . . . . . . . . . . . . . . . . 18 1.10 Phosphorylation and Direct Action of G Proteins on the L-type Calcium Channel ................................. -...... 21


1.11 Partition Coefficients as a Factor in Determining the Effects of Calcium Channel Blockers . . . . . . . . . . . . . . . . . .23 2. Neuritogenesis and Calcium Sensitive Events . . . . . . . . . . .25 2.1 Intracellular Calcium Flux Due to Contact with the Substrate . . . . .25 2.1.1 Effects of Polylysine ................................ 26 2.1.2 The Role of Integrins ................................ 27 2.1.3 N-cadherine\s and Immunoglobulins ........................ 29 2.2 Intracellular Calcium Levels and the Cytoskeleton ................ 32 2.2.2 Microtubles . . . . . . . . . . . . . . . . . . . 35 2.2.3 Neuroftlaments .................................... 36 2.3 Initiation ...................................... ... 37 2.4 Hypothesis on Initiation ............................ ... 39 3 .. Methods . . . . . . . . . . . . . . . . . . . . . 44 3.1 Cell Culture ........................................ 44 3.2.2 Assessment of Survival and Differentiation ................... 46 3.2 Statistical Analysis .......................... ...... .47 4. Results . . . . . . . . . . . . . . . . . . . . . 48 4.1 Cell Culture Experiments ............................... 48 4.1.1 Effects of the Dihydropyridines, the Phenylalkylamine, and the Benzothiazipine on Cell Culture .......................... 57 Survival ......................................... 57

PAGE 8 Initiation . . . . . . . . . . . . . . . . . . . . 60 Axon Cells as a Percentage ofGrowing Cells .................. 60 Axon :Length ..................................... 63 Numbers of Branches Per Axon . . . . . . . . . . . . . 65 Dendrite :Length .................................... 68 4.1.1. 7 Numbers of Dendrites Per Cell . . . . . . . . . . . . 68 Numbers of Branches Per Dendrite ........................ 70 5. Discussion ........................................ 71 5.1 Effects of Calcium Channel Blockers on Neurite Outgrowth ........... 71 5.1.1 Initiation ....................................... 71 5 .1.2 Isradipine . . . . . . . . . . . . . . . . . . . . 73 5 .1.3 Survival . . . . . . . . . . . . . . . . . . . . 77 5.1.4 Numbers of Cells with Axons Per Numbers of Cells Growing ......... 78 5.1.5 Elongation ofAxons and Dendrites .......... ............... 79 5 .1. 6 Branches on Axons and Dendrites . . . . . . . . . . . . . 80 5.2 Discussion of the Potencies of Calcium Channel Blockers . . . . . 81 5.3 Overall Effects of Blocking L-type Calcium Channels ............ 84 5.4 Conclusion ....................................... 87 Bibliography . . . . . . . . . . . . . . . . . . . . 89


ACKNOWLEDGEMENTS This thesis would not have been possible without the help and support of a number of laboratory assistants, namely, David Shugarts, Charlie Ferguson, Leigh Cabell, and Marcie Kein; and several librarians, including Rick Boeder, Lynn Schwalm, Michael Bennett, and Paul Worthington. Dr. Gerald Audesirk and Dr. Terry Audesirk provided technica1 advice and spent many hours editing my thesis and correcting exams for which lam thankful.. Dr. Alan Brockway also spent hours with the thesis and exams. Lisa Keller and her family provided moral support and real food in times of despair which gave me sustenance to continue with this project. ,-Rick Boeder provided countless numbers of cups of coffee and continued support through some trying times. My father and mother provided financial support which might otherwise have been spent on vacations in the Carribean. They were never really sharp as investors. My brother, Steve, was very tolerant of my midnight calls of de$pair during the beginnings of this project. I heartily thank all of the above for the. help and support. I would not have finished the thesis without you.


1. Introduction Recent research has suggested that optimal development of neurites, both axons and dendrites, from a neuron is dependent to some degree on intracellular calcium concentrations which are modulated, at least to some extent, by influx through voltage-sensitive calcium channels (Audesirk et al., 1989; Audesirk et al., 1990; Garyantes and Regehr, 1992; Mattson and Kater, 1987; Robson and Burgoyne, 1989; Rogers and Hendry, 1990; Suarez-Isla et al., 1984). Aspects of neurite development modulated by intracellular caicium include survival of the neurons, initiation of neurite growth from the embryonic cell, formation and elongation of the axon and dendrites, and growth cone movement which occurs at the ends of both axons and dendrites. Intracellular calcium concentrations which are either too high or too low will inhibit all or at least some of the parameters to some degree (Audesirk et al., 1990; Collins et al., 1991; Mattson and Kater, 1987). The purpose of these experiments was to study the effects of six different calcium channel blockers on a number of different aspects of neurite development from embryonic cells which have undergone their last mitotic division and are beginning to form neurites (the first protuberances from the cell body which have not been designated as an axon or It is hypothesized that blocking the influx of calcium through voltage sensitive calcium channels and thus, lowering or maintaining a low concentration of intracellular calcium, will adversely affect some or all of the aspects of neurite 1


development mentioned above. 1.1 Calcium Channels Calcium enters a cell via calcium channels, which are large proteins embedded in the plasma membrane. There are two major divisions of types of channels: those opened when a particular ligand binds to the channel (ligand-gated channels) thus changing its conformation and opening or modulating the channel, and voltage-gated channels which are .opened when a cell depolarlzes to a particular voltage which causes a change in the conformation of the channel leading to channel opening. When a calcium channel blocker binds to a voltage-sensitive calcium channel, the calcium channel blocker is considered a ligand, i.e., a molecule which binds to a channel and modulates the ion flow through the channel. Several types of voltage-sensitive calcium channels eXist, including: L-type which are slow to activate, requiring a large depolarization but remaining open for a relatively long period of time, and have a large unitary conductance. N-type and P type calcium channels both open at relatively high depolariztion and have a small conductance which is moderately rapidly inactivating. T -type calcium channels open at low depolarization levels and have a small conductance which is rapidly inactivating (Catterall and Striessnig, 1992; Kostyuk, 1989; Nowycky et al., 1985; 2


Spedding and Kenny, 1992). This thesis is primarily devoted to L-type calcium channels although reports exist implicating the types of calcium channel blockers used in these experiments having an effect on the other types of calcium channels as well (Takahashi and Akaike, 1990). 1.2 Brief Explanation of Procedures Embryonic day 18 (E-18) rat hippocampal cells were grown in culture to determine the effects of the calcium channel blockers on neurite development in culture. The individual calcium channel blockers were added to the medium of rat hippocampal cells which were then grown in culture for forty-eight hours. Parameters of cell culture growth were then assessed at the forty-eight hour time point. The experiments indicated that a relationship exists between intracellular calcium concentrations and neurite development, specifically, that (1) a lower intracellular calcium concentration will inhibit neurite development and (2) a slightly higher than normal intracellular calcium concentration may enhance some aspects of neurite development. The six compounds used in these experiments, nicardipine, nifedipine, nimodipine, isradipine, verapamil and diltiazem, are all specifically designated as Ltype (voltage-sensitive) calcium channel blockers. Nicardipine, nifedipine, 3


nimodipine and isradipine are all 1 ,4-dihyropyridines. Verapamil is a phenylalkylarnine. Diltiazem is abenzothiazepine. All six compounds bind to the at-subunit of the L-type calcium channel (Langs and Triggle, 1992; Triggle and Rarnpe, 1989), although the three groups bind to three different sites on the at subunit (Triggle et al., 1989). 1.3 L-type Voltage-sensitive Calcium Channel Structure The L-type voltage sensitive calcium channel has five subunits: the a1-subunit, a2-subunit, /3-subunit, -y-subunit, and <'>-subunit (see drawing on next page). The at subunit has a molecular weight of 212kDa (in rat brain) including its glycosylated C terminal region and is structurally very similar to the sodium channel at subunit (De Jongh et al., 1989; Starr et al., 1991). The calcium channel at subunit weighs approximately 175kDa or 190kDa (newer and better methods of determining molecular weights indicate that the 175kDa a1 subunit may actually weigh 190kDa [DeJongh et al., 1991]) without the glycolsylated C-terminal region mentioned above which contains at least three cAMP-dependent protein kinase phosphorylation sites (DeJongh et al, 1989; DeJongh et al., 1991). It is notable that both the 190kDa subunit and the. 212kDa subunit are capable of being phosphorylated by cAMP dependent protein kinase (DeJongh et al., 1989), protein kinase C, and calmodulin dependent protein kinase (Armstrong et al., 1991; Chang et al., 1991). There appear 4


Ca2+ I I I I 1 I J,, I f I J I I I I I I I I I I I Figure 1. Proposed orientation of 'L-type calcium channel subunits showing the a1 subunit with a pore. Phosphorylation sites marked with a 'P'. Used with permission from the N.Y. Acad. of Sciences, Catterall et al., 1989. to be at least four classes (A,B,C, and D) of a1 subunits in the mammalian brain and there are estimates that a minimum of eight different calcium channels may exist (Snutch et al; 1990; Starr et al., 1991). Genetic studies indicate that classes A and not code for proteins that are closely related to the dihydropyridine sensitive subunits that we will be discussing here and probably belong to one or more of the other types of calcium channels that are present in the mammalian brain (Starr et al., 1991). The genes of classes C and Dare most likely to code for the L-type calcium 5


charuiel ot1 subunit which is DHP sensitive (Snutch et al, 1990; Starr et al., 1991). One of the genes found in rat brain and belonging to the class C type calcium channel shows a 97% homology to a rabbit cardiac calcium channel which is known to be an L-type calcium channel (Snutch et al., 1990). 1.3.1. The a1 Subunit The a1 subunit contains the actual-channel or pore through 'Yhich.the calcium ion gains entry from the extracellular to the intracellular regions of the cell, while the other four subunits of the channel seem to be involved in modulation of channel activity (Gutierrez et al., 1991; Spedding and Kenny, 1992). The pore diameter, as determined using large organic ions in permeability experiments, is approximately 60 nanometers in diameter (Tsien et al., 1987). A calcium ion is approximately 20 nanometers in diameter so this channel size is ample (Tsien et al., 1987). 1.4 Hypothesis for a Mechanism of Ca:z+ Channel Opening The entire ot1 subunit of the L-type calcium channel consists of approximately 2005 amino acids which are divided into four domains (Catterall, 1988; Catterall and Striessnig, 1992; Langs and Triggle, 1992). Each domain has six transmembrane helices, labelled S1 thru S6, which are hydrophobic, with intermittent stretches of 6


hydrophilic segments which protrude into the cytoplasm or into the extracellular fluid (Catterall, 1988; Catterall and Striessnig, 1992; Langs and Triggle, 1992). There are, apparently, two hypotheses as to the placement of the S4 helix in each of the four domains of the channel. One theory states that the S4 transmembrane :helices surround the area immediately adjacent to the channel opening and all of the helices change position to open the channel(Catterall, 1988). The other hypothesis indicates that the S4 helix is surrounded by the other helices in the channel domains so that other helices are adjacent to the actual pore of the channel, but it is still the S4 helix that is responsible for opening the channel (Langs and Triggle, 1992). The voltage-gating and voltage sensing capabilities of the S4 helices are apparently due to the arrangemant of the positively charged amino acids in the S4 helix and the interaction of those amino acids with the surroundmg members (helices) of the channel (Catterall, 1988; Langsand Triggle, 1992). Every third amino acid of the S4 segment is charged; being either a lysine or an arginine (Catterall, 1988). When the cell membrane is in a polarized state, in which the cytoplasmic side (inside of the cell) of the plasma membrane is negatively charged as compared to the ex;tracellular side (outside of the cell) of the plasma membrane, the positively charged amino acids interact with other helices (Langs and Triggle, 1992). Therefore, the position of the helices is regulated by the negative potential of the cytoplasmic side of the plasma membrane (Catterall, 1988; Langs and Triggle, 1992). When the cell 7


depolarizes, i.e., as the cytoplasmic side of the plasma membrane becomes more positive, and the force which was holding the positively charged amino acids in place is lessened, the helix undergoes a spiraling of approximately 60 which moves the helix outward from the center by approximately 50 nanometers and upward by approximately 30 nanometers (see figure ) ( 1 d) space {posJ 1ve pore c ose helix in closed -?a postion helix pore open Figure 2. Top: pore closed { Bottom: helix rotates out and up with part of helix now in the extracellular region. 2) (Catterall, 1988; Langs and Triggle, 1992). This leaves negatively charged amino acids facing the center of the channel which will attract the positively charged ca).cium ions, according to the hypothesis in which the S4 helix surrounds the pore


(Catterall, 1988; Langs and Triggle, 1992). The second hypothesis would imply that the S4 helix would force a conformational change in the S3 helices which in this hypothesis surround the pore, thereby opening the pore. This movement of the helices does not occur in all four S4 segments simultaneously (Catterall, 1988). The four domains of the channel are arranged in a four cornered square around the pore and the helices rotate one after another in succession and open the pore (Catterall, 1988; Langs and Triggle, 1992). Thus, the channel is opened in stages. There are at least three states in which an L-type calcium channel can cycle: open, closed, and inactive. However, these states may operate by different mechanisms than the stages of opening and closing of the pore (Catterall, 1988). The dihydropyridines used in my experiments preferentially bind to. the inactive state of the channel preventing its cycling to the open state (Catterall and Striessnig, 1992; Tsien et al., 1987). such as verapamil seem to gain access to their binding sites on the calcium channel when the channel is in the open state (Catterall and Striessnig, 1992). The benzothiazapines such as diltiazem may or may not show a preference for a particular state of the calcium channel depending upon which benzothiazipine is used (Hering et al., 1993). 9


1.5 Calcium Ion Binding Sites Within the Pore Calcium ions are presumed to have two binding sites within the pore of the channel (Tsien et al., 1987). The pore is of a single-file type so that the binding sites occur one after another, i.e. one bindirig site is close to the extracellular surface and one is deeper into the channel towards the cytoplasmic side of the channel. The calcium ion is divalent and positively charged so it is likely that other divalent positively charged "ions, i.e., magnesium, would have some affinity for the calcium binding site. Charinel blocking by other divalent cations, i.e. cadmium, occurs when the ion binds at one of the binding sites within the channel preventing the calcium ion from entering (Tsien et al., 1987). Sodium, which is monovalent and positively charged, has a lower affiirlty for the binding sites by a ratio of 1000: 1 (calcium to sodium) (Langs and Triggle, 1992; Tsien et al., 1987). Another factor involved here is that the cations on the outer and the inner binding sites will repel each other (Tsien . et al., 1987). This force (repellirig) will cause a more rapid exit of the inner calcium ion into the cytoplasm which results in a more rapid current through the channel (Tsien et al., 1987). 10


1.6 Dihydropyridine Binding Sites on the a-1 Subunit of the L-type Calcium Channel There appear to be at least two hypotheses currently in the literature describing the proposed binding site for dihydropyridines to the calcium channel. Several authors propose that the binding site for the dihydropyridines lies within the structure of the S4 helix of the a1 subunit of the L-type calcium channel (Catterall, 1988; Langs et al., 1991; Langs and Triggle, 1992). The binding affmities of the dihydropyridines to the S4 helix as well as the the difference between an agonist or antagonistic function of the dihydropyridine is dependent upon the structure of the dihydropyridine (Catterall, 1988; Langs and Triggle, 1992). In the second hypotheis, dihydropyidines, benzothiazapines and phenylalkylamines all bind to a region within the pore of the calcium channel (Nakayama et al., 1991). 1. 7 Dihydropyridine Structure My experiments included four dihydropyridines: nicardipine, nifedipine, nimodipine and isradipine. The molecular structures are shown on the next two pages. Dihydropyridines consist of a 1,4-dihydropyridine ring with a nitrogen atom at the number one position and five carbons in order around the ring (see figure 3 below). The 1 ,4-dihydropyridine refers to the hydrogenation at the 1 and 4 positions, 11


CHEMICAL STRUCTURES OF CALCIUM CHANNEL BLOCKERS DffiYDROPYRIDINES .MeOOC Me Me NICARDIPINE NIMODIPINE Nicardipine and nimodipine structures used with permission from the New York Academy of Sciences and Dr. Triggle. Triggle et al., 1991. lla


CHEMICAL STRUCTURES OF CALCIUM CHANNEL BLOCKERS DffiYDROPYRIDINES NO 2 NIFEDIPINE ISRADIPINE Nifedipine structure used with permisssion from Biochemical and Biophysical Research Communications. Erne et al., 1984. Isradipine structure used with permission from Trends Pharmacol. Sci. Catterall, W.A. and Striessnig, J., 1992. 11b


i.e.,. .both positions have a hydrogen atom bonded to a primary carbon atom of the ring. The number four carbon atom (third from the nitrogen) has a phenyl ring bonded Dihydropyridine Ring C-5 C-6 Phenyl Ring characteristic of. all of Dihydropyridine Ring ...____ 'C-5. the dihydropyridines -,. used in these experiments but not C-4 C-3 Para Meta a universal Figure 3. Top: Dihydropyridine Ring Bottom: DHP Ring with phenyl t r a i t 0 f ring attached at C-4. dihydropyridines. Various substitutions on the aryl and the phenyl ring make up the multitude of different dihydropyridines. All of the dihydropyridine structures used for these experiments are reported to be L-type calcium channel blockers .under most conditions. However, isradipine has been reported to be an agonist when the a1 subunit is phosphorylated (Glossman et al., 1989; Hymel et al., 1988). See the figures on the next page for the similarity between isradipine, an antagonist, and 12


CHEN.UCALSTRUCTURES DlllYDROPYRIDINES \ ,o ISRADIPINE, a reported antagonist H SANDOZ 202791, a reported agonist Isradipine structure used with permission from Trends in Pharmacol. Sci. Catterall, W .A. and Striessnig, J., 1992. Sandoz 202791 structure used with permission from the European Journal of Pharmacology. Priego et al., 1993. 12a


Sandoz 202-791, an agonist. 1.8 Structure Activity Relationships of Dihydropyridines The activity of the dihydropyridine antagonists may be dependent upon the amount of puckering of the 1 ,4-dihydropyridine ring, which directly influences the positioning of the hydrogen bonding capabilities of the N(1)-H group (Fossheim et al., 1982; Langs etal., 1991). Normally, the 1,4-dihydropyridine ring is in a boat configuration with the phenyl ring in a flagpole configuration above the dihydropyridine ring (see figure 4) (Fossheim et al., 1982; Triggle, Langs and Janis, 1989). The more active antagonists have the 1,4-dihydropyridine ring more flattened than normal (Fossheim et al., 1982; Triggle, Langs and Janis, 1989). There are several reasons for the more planar configuration to occur; The substitutions in the phenyl ring cause the phenyl ring to rotate about the C-4 bond so that the phenyl ring will sit above and perpendicular to the 1,4-dihydropyridine ring in the flagpole configuration (Fossheim et al., 1982; Langs and Triggle, 1992; Triggle, Langs and, 1989). The substituents of the phenyl ring, especially in the ortho position (see figure 4), will cause torsion at the C4 bond. Note that the substituents in the ortho position of the phenyl ring can swing to a position over the dihydropyridine ring or they can swing out in front of the dihydropyridine ring to lie in front of the boat configuration of the dihydropyridine ring. The torsion at the C-4 bond wili cause the 13


entire 1,4-dihydropyridine ring to become strained and there will be torsion at the N-1 bond in proportion to the torsion on the C-4 bond (Fossheim et al., 1982). Substitutions, other than hydrogen, in the phenyl ring at the ortho position will cause molecular interactions between the phenyl ring and the 1 ,4-dihydropyridine ring, flattening the 1 ,4dihydropyridine ring resulting in a more active antagonist (Fossheim et al., 1982; R1 /port.@ Q SldtZ \ C=O trans-czstczr m -.-.. -_-(J) aryi ring 0 tors1on c-o J '\ @ CIS-tZshzr R2 starboard sid12 Figure 4. Top: boat configuration of DHP ring with phenyl group in flagpole configuration. Bottom: looking down on DHP ring showing torsion at C-4 bond. Reprinted by Permission. of. John Wiley & Sons, Inc. Triggle, Langs, and Janis 1989. 14


Triggle, Langs and Janis, 1989). Substutions at the 3 and 5 positions in the 1,4dihydropyridine ring also are important. The dihydropyridines with different structures (nonidentical) at the 3 and 5 positions are generally more active than compounds with the same structures at the-3 and 5 positions (Fossheim et al., 1982). 1.8.1 Importance of the N(1)-H Group of the Dihyropyridine Ring The N(1)-H group of the 1,4-dihydropyridine ring, using hydrogen bonding to other groups, is very important in deterrilining both the strength of antagonistic activity and in determining agonist or antagonistic activity (Fossheim et al., 1982; Langs et al., 1991; Langs and Triggle, 1992). Antagonists of the L-type calcium channel completely lose their activity without the N(1)-H group (Langs et al., 1991). The above discussion of planarity may be ultimately related to the positioning of the N(1)-H group in relation to the rest of the molecule as it binds to the S4 helix of the channel (Langs et al., 1991). The binding site of the channel may have multiple hydrogen binding sites forthe N(1)-H group and the binding to these sites may be important in both agonist and antagonist activity (Langs et al., 1991). Multiple binding sites for the N(l)-'"H group may be instrumental in agonist activity, whereas having only one binding site available to the N(1)-H group may indicate an antagonist activity (Langs and Triggle, 1992). The availability of binding sites for the N(1)-H group would depend upon the planarity of the boat configuration of the 15


dihydropyridine ring, i.e., if the ring were in a boat configuration (not flattened), there may be more binding sites available (indicating agonist activity) than if the ring were less boat-like and more planar (indicating antagonist activity). The binding regions of both the C-3 and the C-5 regions may also play a pivitol role in determining whether the compound is an agonist or an antagonist (Langs et al., 1991). Antagonists with an ester group at the C-3 position have a positive electrostatic potential whereas agonists may have a nitro group at this position, which causes a negative potential (Langs et al., 1991). Thus, the positioning of the dihydropyridine into the S4 helix is influenced by the location of the N(1)-H group in relation to the rest of the molecule i.e., the degree of puckering of the 1,4dihydropyridine ring, which in tum is influenced by the substituents of the phenyl ring especially, in .the ortho position, and the electrostatic potential of the C-3 group of the 1,4-dihydropyridine ring (Langs et al., 1991). 1.8.2. Description of Nifedipine Binding to the S4 Helix A possible scenario for the positioning of the 1,4-dihydropyridine (nifedipine) into the S4 helix is shown in figure 5 (adopted from Langs et al., 1991). The phenyl group is shown to fit against the backbone of the S4 helix, sliding into the groove between the two positively charged amino acids, namely two arginine molecules. 16


T h e t w 0 substituents of the R2 C-3 and C-5 groups /) 0 -)1 9E form hydrogen bonds with. the the N(1)-H group / 'b R1 \ N 'r. arginine groups and N of the 1,4-d"h d "d" Figure 5. Nifedipine fitting into the S4 1 Y ropyn me Helix. Used with permission from the Journal of Computer-Aided Molecular Design. group forms Langs et al., 1991. hydrogen bonds with the adjacent S2 helix (Langs et al., 1991). The binding of the antagonists to the L-type channel is presumed to occur when the channel has cycled to the inactive state and Triggle, 1992). 1.8.3. A Second Hypothesis for Dihydropyridine Binding There is another hypothesis for a possible binding site for the 1 ,4dihydropyridines. Several groups of researchers have hypothesized that the binding site for the dihydropyridines lies within the pore of the a1 subunit (Nakayama et al., 1991; Striessnig et al., 1991). According to this model, the extracellular amino acid 17


sequences of domain Ill between helices IllSS and IllS6 and domain IV between helices SS and S6 of the a1 subunit protrude into the pore (possibly these segments help form the channel) and it is to this segment of amino acids that the dihydropyridines bind (Nakayama et al., 1991; Striessnig et al., 1991). If this hypothesis proves to be true then the binding sites for the dihydropyridines, the benzothiazapines and the phenylalkylamines would be closer together with the dihydropyridines .and the benzothiazapines binding close to the extracellular side of the pore and the phenylalkylamines on the intracellular side of the pore (see below). And, of course, there may be two or more binding sites in the a1 subunit for dihydropyridines, one in the pore itself and the other in the S4 helix of domain ill. However, more research in this area will be needed to prove this point. 1 . 9 Structure-Activity Relationships for Benzothiazipines (Diltiazem) and Phenylalkylamines (V erapamil) Much less is known about the binding sites of verapamil and diltiazem. Computer molecular models have not been utilized to explain the binding of these two compounds. However, studies have been done of the potencies of verapamil and its derivatives (Mannhold et al., 1978). Mannhold et al. (1978) made substitutions at three different sites on the verapamil molecule. First, they made changes on the benzene ring, then they changed substituents at the asymmetric carbon atom, and 18


CHEMICAL STRUCTURES OF CALCIUM CHANNEL BLOCKERS PHENYLALYLAMINES VERAPAMIL BENZOTHIAZIPINES DILTIAZEM Verapamil and diltiazem structures used with penmss1on from Trends Pharmacol. Sci. Catteral, W.A. and Striessnig, J., 1992. 18a


finally, they made changes at the amino nitrogen. Changes in the substituents .of the benzene ring affected the potency of the drug .. Adding another -OCH3-group to the benzene ring to form 0600 increased the potency of verapamil; whereas additions to the size of verapamil by adding larger substituents to the carbon atom adjacent to the benzene ring decreased the potency of the drug (Mannhold et al., 1978). Substituents at the benzene ring affected potency but did not alter the general effect of the drugs. However, the characteristic drug action remained unchanged. Changingtheisopropyl substituent of the benzene ring to several other substituents altered the potency but not basic activity. Different substituents at the amino group also yielded appreciable differences in the potencies (Mannhold et al., 1978). Hescheler et al. (1982) tested the hypothesis that verapamil binds to the intracellular side of the calcium channel. They compared 0600,. a verapamil derivative, with 0890, an ionized form of 0600 which is membrane impermeable, by applying each of these compounds to_,the exterior of the cell first and then injecting them into the cell for another trial of their effectiveness. 0600 had similar effects whether applied externally or internally. 0890, however, had no effect if applied to the exterior the cell, but was effective if injected into the cell. They also noted that the time from the moment of application to the perception of the first effects of D600 was longer if the application was to the exterior of the cell rather than to the interior. They hypothesize that the binding site must be on the 19


intracellular side of the calcium channel. Therefore, verapamil must pass through the plasma membrane or through the channel itself to reach the binding site and block the calcium channel (Hescheler et al., 1982). This hypothesis does not necessarily contradict the hypothesis that the binding sites of the dihydropyridines are .adjacent and interfere with the binding sites of both verapamil and diltiazem and other calcium channel blockers. The fact that one calcium c:hannel blocker has the ability to disrupt the binding of another calcium channel blocker is not antagonistic to the hypothesis if the structure .and operation of the calcium channel is viewed simultaneously. Since the S4 helix reportedly rotates about its axis to open and close the channel, any binding to the helix at any point on its axis would stop rotation of the helix and thereby block the channel. An agonist m.lght bind to the S4 helix after it was open and hold the channel open. Furthermore, the binding of verapamil, for instance, to an intrachannel site of the pore may very well disrupt. the binding of a dihydropyridine, if for instance, the channel changed conformation due to the binding of the verapamil and thus would not allow the dihydropyridine into its normal binding position. And the same would hold true for other calcium channel blockers which could very easily bind to the helix at other JX?ints all along its axis or bind to regions within the pore of the channel. The result of this binding would, in fact, be an allosteric competition, but not necessarily a competition for exactly the same binding site. 20


The concept also holds true for diltiazem, which is a benzothiazipine. Several studies have been done to enhance the potency of diltiazem by .adding various substituents to its primary structure (Tanaka et al., 1992; Yanagisawa et al., 1992). Some.additions:proved to increase potency manyfold (Yanagisawa et al., 1992). The apparent binding site for diltiazem is to a region within the pore of the channel on the extracellular side (Hering et al., 1993). Using membrane impermeable analogs of diltiazem, namely .SQ32,428, a benzazapine, the researchers found. that extracellular application .of SQ32,428 blocked the calcium channel to 40% within 30 seconds, but intracellular application had no effect on the calcium channel currents (Hering et al., 1993). The potency of SQ32,428 and diltiazem is dependent on the presence of an amine group at the N(1) position and also on a hydrogen bond acceptor at position 4 of the aryl group (Hering et al., 1993). Binding of SQ32,428 was greatly decreased when divalent cations were included in the solution (Hering et al., 1993), a fact which might have a bearing on some of my results. 1.10 Phosphorylation and Direct Action of G Proteins on the L-type Calcium Channel There are at least two subunits of the L-type calcium channel that are capable of phosphorylation. Both the a1 subunit and the {3 subunit have phosphorylation sites (Catterall, 1988; Catterall et al., 1989; Nunoki et al., 1989). The 212kDa a1 subunit 21


has at least three more phosphorylation sites than the 190kDa a1 subunit (Starr et al., 1991). Cyclic AMP dependent protein kina8es (PKA), calcium calmodulin dependent kinases, and protein kinase C as well as several types of G proteins have been implicated in the phosphorylation and subsequential modulation of the L-type calcium channel (O'Callahan et al., 1988; Catterall, 1988; Hoffmann et al., 1987; Hosey and Lazdunski, 1988; Izumi et al., 1990; Nunold et 1989; Reuter, 1983). Phosphorylation of the L-type channel or the direct action of G proteins (the site of binding by G proteins is not yet known) can have various effects on channel properties, either activating the channel and increasing the current through it or inactivating the channel and .decreasing the current through it (O'Callahan et al., 1988; Hoffman et al., 1987; Kasai and Aosaki, .1989). An hypothesis has been proposed that calcium channel agonists such as Bay K 8644 act as agonists because they slow down dephosphorylation of the calcium channel thereby keeping it active longer, whereas antagonists such as nicardipine speed up dephosphorylation of the channel and thus inactivate it (Armstr

L-type calcium channels which were reconstituted into bilayers was the effect of isradipine (PN200-110) on phosphorylated channels (Hymel et al., 1988). Normally, isradipine is considered to be a potent antagonist of L-type calcium channels. However, Hymel et al. (1988) found that the L-type calcium channels which were reconstituted from skeletal muscle had to be phosphorylated (for example, by cAMP dependent protein kinase) to show activity (meaning that a current was flowing through the channel) and that this activity was increased and stabilized by the addition of 10/LM isradipine. D890 (an analog of verapamil) and cadmium ions (known to bl9ek calcium channels) acted to block the phosphorylated channels, and Bay K 8644, which is known to activate calcium channels, had no effect on unphosphorylated channels (Hymel et al., 1988). The unexpected agonist activity by isradipine may have a bearingon some of the results which I obtained in my experiments. 1.11 Partition Coefficients as a Factor in Determining the Effects of Calcium Channel Blockers. Partition coefficients seem to be important in the binding of dihydropyridines to the a1 subunit of the calcium channel (Mason et al., 1989; Mason et al., 1990). If the dihydropyridines approach their binding site after entering the membrane then the ability of the dihydropyridine to dissolve into the membrane would play a part in determining the effects of the dihydropyridine (Langs and Triggle, 1992; Mason et 23


al., 1989; Mason et al.,. 1990). Verapamil may have its binding site on the intracellular side of the calcium channel, .and thus may not be so dependent upon' a partition, coeficient. as it is on having the ability to cross a lipid bilayer rapidly (Hescheler etal.:, 1982;: Mannhold et al., 1978). (Note: Technically, a partition coefficient assumes that the solute dissolves without association with any other molecules into both solvents, in this case a lipid bilayer and the extracellular fluid. However, the lipid bilayer has several layers and it is possible that verapamil-may associate with a particular part of the membrane or another molecule within the membrane and be inclined to stay there rather than pass through the membrane completely.) Diltiazem has its binding site on the extracellular side of the calcium channel pore and does not have to cross the lipid membrane or dissolve into the to bind to the channel (Hering et al., 1993). Diltiazem would be independent of partition coefficients or membrane permeability. However, there are reports that diltiazem may have effects on the permeability of membranes in general (Caretta et al., 1991). The effectiveness of the various drugs tested here may depend, in part, upon the composition of the lipid bilayer of the particular cell type tested as well as the ability of the drug to pass through or dissolve into the lipid bilayer (Mason et al., 1989). 24


2. Neuritogenesis and Calcium Sensitive Events My experiments have utilized cryopreserved E18 rat hippocampal cells, as described in. the Methods' section. At this stage of developement, the neurons have passed their last mitotic division and have begun to develop an axon and dendrites. At E18 the cells still appear as round or slightly irregular round masses (processes which may have started before dissociation were sheared off leaving the cells round again). After two days in culture healthy cells will show considerable development of a single long axon and numerous dendrites of varying lengths. Initiation is the process .in which the cell body, which is basically round, changes to a flattened, irregular shape and extends a neurite with a growth cone of filipodia and lamellipodia to begin the process of neurite development (Bray, 1992; Hall,1992). (Note that lamellipodia are flattened extensions of the growth cone and filipodia are single finger-like protrusions from the lamellipodia). The following discussion will provide the necessary background information to understand the importance of calcium in the developmental process of neuronal. growth. 2.1 Intracellular Calcium Flux Due to Contact with the Substrate 2.1.1 Effects of Polylysine The first calcium sensitive event which may occur is the actual attachment of the cell to the substrate (polylysine, in my case) in the cell culture dish. Polylysine is 25


a positively charged (cation) peptide. (a long chain of amino acids) which is used in cell culture as a coating on plastic culture dishes to which the cells can attach and grow (Cannella and Ross, 1987). All of the experiments described in this thesis incorporated the use of polylysine as a coating on plastic culture dishes. Recently, polylysine has also been used to coat glass fibers in experiments to assess the migration of neural granule cells (Fishman and Hatten, 1993). In these experiments, it was found that .polylysine does not promote migration; instead, the polylysine promoted a stationary neuron which produced neurites (Fishman and Hatten, 1993). Astroglial membranes, fibronectin and laminin all promoted at least some migration, whereas collagen did not provide a substrate for migration (Fishman and Hatten, 1993). Collagen, polylysine, laminin, fibronectin, and polyethyleneimine (a recent addition) have all been used in culturing various types of cells (Cannella and Ross, 1987; Lein et al., 1992; Lelong et al., 1992). Polylysine may affect several systems (besides the possibility of attaching to adhesion molecules such as .the integrins) in a cell which in turn may affect the process of initiation and the growth of neurites in the neuron (FujitaY arnaguchi et al., 1989; Mittag et al., 1993; Morrison et al., 1989; Needham et al., 1988). Endothelial cells are stimulated by polylysine to produce prostacyclines and cytoplasmic purines (Needham et al., 1988). Calmodulin, while not normally phosphorylated by casein kinase-2 (CK2), is phosphorylated by CK2 when polylysine 26


is present (Hardy et al., 1994; Meggio et al., 1992; Mittag et al., 1993; Sarno et al., 1993). Polylysine is also involved in the activation of other enzymes such as tyrosine kinases, inositol phospholipid kinases and phosphatases (Hardy et al., 1994). Finally, poly lysine stimulates the insulin receptor which activates its tyrosine kinase, starting a .phosphorylation cascade (Li et .1992; Morrison et al., 1989). 2.1.2 The Role of lntegrins Since the binding of a neuron to its substrate may be a calcium sensitive event mediated by an influx of calcium through. calcium channels, it is important to examine the process of attachment. As discussed above with polylysine, this process of attachment may be the primary event which stimulates initiation. And these events which may be by attachment might be sensitive to calcium channel blockers. The attachment of a cell to the substrate is normally mediated by the integrins which are one of the three classes of attachment molecules which span the plasma membrane of the cell: integrins (a particular type of glycoprotein), cadherins (also a glycoprotein), and those belonging to the immunoglobulin superfamily such as NCAM and fasciculin II (Hall, 1992; Jacobson, 1991; Kandel et al., 1991). The cadherins and the immunoglobulin type of proteins will normally bind to homologous 27


molecules (i.e., L1 will bind to another L1) displayed by other cells and the integrins will bind to one or more extracellular matrix molecules of the substrate upon which the cell is growing (Kandel et al., 1991). Integrins will attach to fibronectin and laminin, both o{ which are extracellular matrix molecules naturally secreted by various cells. (Kandel et al., 1991). The integrin attachment may mediate intracellular events requiring a rise in calcium (Kandel et al., 1991). Neutrophils express oscillations in intracellular calcium levels when CD 11 b and CD 18 integrin molecules are stimulated (Ng-:Sikorskiet al., 1991). The calcium iricrease is due, at least in part, to extracellular calcium entering the cell and is not sensitive to pertussis toxin, indicating that G-proteins may not be involved in this particular event (pertussis toxin blocks the effects of G proteins)(Ng-Sikorski et al., 1991). In contrast, N1 (Neuroblastoma) eells undergo a marked hyperpolarization (indicating a more negative potential within the cell instead of a more positive or depolarized state) when they are in contact with fibronectin and vitronectin, which returns to normal levels within an hour (Arcangeli et al., 1993). N7 cells (another neuroblastoma line) adhered to these two substrates but did not initiate neurites as did the N1 cells. Pertussus toxin abolished both the hyperpolarization and the neurite outgrowth in the N1 cells indicating that a G-protein is involved with the 28


hyperpolarizition and consequent neurite initiation in this cell line (Arcangeli et al., 1993). This initiation ofneurites was inhibited by cesium, which blocks potassium channels (Arcangeli et al., 1993). Therefore, in N1 neuroblastoma cells, hyperpolarization mediated by a G-protein activated potassium channel is essential for neurite initiation when the cells are plated on fibronectin or vibronectin because these substrates bind to integrins on the N1 cells (Arcangeli et al., 1993). The reactionof a particular cell type to the particular substrate is relevent to the intracellular mediated response. There appears to be great variation in these reactions. 2.1.3 N-cadherins and .Immunoglobulins N-cadherins and the immunoglobulins are the other two types of adhesion molecules, and these normally mediate attachment of one cell to another cell .. However they are also involved with the initiation of neurites in some cases (Bixby and Jhabvala, 1990). They are important to this discussion as the cells used in my experiments may bind to polylysine with these adhesion molecules (personal hypothesis). Attachment by N-cadherins and NCAMs can cause a rise in intracellular calcium in PC12 cells (Doherty et al., 1991). Normally, PC12 cells require integrin receptor activation and NGF for neuritogenesis. However, when PC12 cells are grown on 3T3 cells transfected with NCAM and/or N-cadherin, neuritogenesis occurs with a concurrent rise in intracellular calcium levels (Doherty et al., 1991). This rise 29


in intracellular calcium was linked to cell differentiation and neurite outgrowth and could be blocked by verapamil, diltiazem and nifedipine, all of which block L-type calcium channels (Doherty et al., 1991). There was an additional inhibition of neurite initiation in these cells with the addition of conotoxin, which blocks N-type c3Icium channels, indicating that some of the calcium entering the cell is through N channels (Doherty et al., 1991). This outgrowth could also be inhibited by pertussus toxin, indicating the involvment of a G-protein (Doherty et al., 1991) .. Antibodies against NCAM and L1 (L1 also belongs to the immunoglobulin family) adhesion molecules in PC12 cells also elicited a rise in intracellular calcium (Schuch et al., 1989). This rise in calcium was blocked by pertussus toxin and by verapamil (2 I'M) and diltiazem ( 4 I'M) but not by nicardipine ( 4 I'M) or nifedipine (4 I'M) (Schuch et al., 1989). Higher concentrations of nicardipine and nifedipine not used, but this lack of sensitivity of the L-type calcium channel to nicardipine and nifedipine may have relevance to my experiments. The response of L1 and NCAM also evoked a decrease in levels of inositol! ,4-bisphosphate (IP:z) and inositol 1,4,5,-trisphosphate (IP3), and this event was also pertussus toxin sensitive (Schuch et al., 1989). The proposed mechanism suggests that activation of NCAM or Ll activates an inhibitory G-protein which inhibits phospholipase C and this leads to reduced levels of both IP2 and IP3 Both diacylglycerol and protein kinase C levels would also be reduced. PKC reduction may lessen phosphorylation ofL-type calcium 30


channels and this may lead to activation of the calcium channels in this particular cell type thus allowing an influx of calcium (Schuch et al., 1989). In addition to the implication of L-type channels in neurite initiation by activation of L1 and NCAM, Saffell et al. (1992) used high potassium (40 mM) evoked depolarization to open calcium channels which mimicked the response elicited by the activation of L1 and NCAM. this response was inhibited by L-type and N-type c3Icium channel blockers, diltiazem (10 pM) and conotoxin (025 pM) respectively (Saffell et al., 1992). The kinase .inhibitor K-252b also inhibited the response indicating that the rismg calcium concentration may be activating one or more protein kinases (Saffell et al., .1992). The substrate upon which cells are grown can determine the second messenger pathways used in initiation. Protein kinase C inhibitors elicited opposite effects in chick ciliary ganglion cells grown onfibronectin or collagen, which elicit an integrin mediated response, compared to cells grown on a substrate containing L1 or N cadherin which interact with L1 and N-cadherin molecules on another cell membrane (Bixby and Jhabvala, 1990). PKC inhibitors block the growth on fibronectin or collagen, but promote growth on cells which are grown on L1 and N-cadherin (Bixby and Jhabvala, 1990). Although these researchers did not determine intracellular calcium levels, other experiments (see above) have considered that a drop in PKC 31


may activate L-type calcium channels via a lessened phosphorylation of the channels, the hypothesis being that in this particular cell type (PC12 cells) phosphorylation of the L-type calcium channel deactivates the channel (Schuch et al., 1989). However, research in our laboratory, found that in rat hippocampal cells, which I used for my experiments, PKC inhibitors decreased initiation (Cabell and Audesirk, 1993). Therefore, the response of a particular cell type to a particular substrate may be entirely different from a response elicited in a different cell type on a different substrate. 2.2 Intracellular Calcium Levels and the Cytoskeleton The second major calcium sensitive response is concerned with the polymerization and depolymerization of the cytoskeleton of the cell. The cytoskeleton of a cell consists of three major types of proteins; microtubules, intermediate fibers (called neurofilaments in a neuron) and microfllaments. The smallest fibers of the cytoskeleton are the microfilaments. These are 6 to 8 nm in diameter and are composed of the protein actin. Actin exists in the cytoplasm as a monomer coupled to a molecule of adenosine triphosphate (ATP). In this state the actin is a globular protein called G actin. When three of the monomers join, they are a nucleus ofF-actin (filamentous actin). Each monomer is now coupled to ADP, as the formation has cost each actin monomer one phosphate group. The resultant 32


trimer is thermodynamically unstable, but the nucleus favors polymerization rather than depolymerization (Rawn, 1989). The actin molecule is polar, appearing in electron micrographs as having a barbed end and a pointed end (Mitchison and Kirschner,1988; Rawn, 1989). During polymerization, G actin-A TP monomers attach to the barbed end, also called the plus end. During depolymerization, ADP-G-actin monomers are removed from the pointed end, also called the minus end. Both of these reactions are dependent upon the critical concentration of free monomers in solution around the filament. The critical concentration is defined as that concentration above which monomers are added to the barbed end and below which monomers are subracted from the pointed end (Rawn, 1989). In equilibrium, monomers are added to the plus end as fast as monomers are subtracted from the minus end. The process of both adding monomers to the barbed end and subtracting the monomers from the pointed end is termed treadmilling. The filament length remains constant, only the monomers change. The actual microfilament involved with this phenomenon consists of two filaments of actin wound in a helix, so that the additions and subtractions of the monomers takes place on each of the two strands (Kandel et al., 1991). This dynamic actin filament, simultaneously elongating and shortening, is important to the growth of the neurite as well as movement of the growth cone. Actin polymerization is most probably the method used to facilitate forward movement of the growth cone (Mitchison and 33


Kirschner, 1988). Active actin filaments, via the act of treadmilling, are the. crux of the leading of growth cones (Mitchison and Kirschner, 1988), which may be the first anatomical entity formed. during neurite initiation. The actin filaments are ACTIN POLYMERIZATION IN THE GROWTH CONE ACTIH MONOMERS ADD TO PLUS OR BARBED EHD MOHUNERS SUBTRIICT FRUN MINUS EHII Figure 6. Polymerization of actin at the end of a growth cone. See text. perpendicular to the leading edge of the growth cone. The barbed ends of the filament are close to the edge and the pointed end of the filament is at the neck of the growth cone where the cone meets the neurite shaft. There are many filaments of actin, in parallel, in the growth cone. Each filament is losing monomers from its pointed end and gaining monomers at the barbed end. Light microscopy reveals the appearance of retrograde motion while the growth cone proceeds forwards. The constant addition of monomers at the leading edge with the constant loss of 34


monomers at the neck makes the filament appear to be moving backward while in fact it is only changing monomers. The monomers fonn a strong, albeit transient, link with: the substrate via a linking protein, talin, and an integral protein of the membrane, an. integrin, which attaches to the substratum via external binding sites. Myosin I has been localized on the leading edge of motile cells with Myosin TI predominantly at the tail end (Bray, 1992). However, neurons seem to have both types of myosin located at the leading edge of groWth cones (Miller et al., 1992). The leading edge of the growth cone is pushed forward by the polymerization of actin filaments. As .the monomers constantly move backwards forming attachment points to. the substratum talin and integrins) for leverage, the leading edge of the actin filament gains monmers and pushes the leading edge forward (Luna and Hitt, 1992; Mitchison and Iq.rschner, 1988). In the rounded, embryonic state, a hippocampal neuron contains a meshwork of actin filaments just under the plasma membrane (Bennett and Weeds, 1986; Luna and Hitt, 1992). The actin filaments are attached to the plasma membrane by membrane binding proteins, such as talin, and these membrane binding proteins, in tum, are secured to the membrane with membrane-spanning anchoring proteins, such as the integrins (Bennett and Weeds, 1986; Luna and Hitt, 1992). 35


2.2.2 Microtubules As soon as the growth cone is formed, with actin at its leading edge, microtubules composed of tubulin must begin to form to provide the structure of the neurite shaft as it leaves the cell body. The microtubles, which form the basic structure of the axon or dendrite, are composed oftubulin dimers (Hall, 1992; Rawn, 1989). The tubulin molecule exists in two forms, the alpha form and the beta form. Together, they form the tubulin dimer (Hall, 1992; Rawn, 1989). Each dimer is bound to a molecule of GTP or GDP. Tubulin dimers join end to end to form a protofilament and thirteen protofilaments arranged in a cylinder form a microtubule (Rawn, 1989). The resulting microtubule is polar, with the GTP-dimers adding to the plus end and the GDP-dimers depolymerizing at the minus end (Rawn, 1989). . The dimer-GTP end is more stable (less likely to depolymerize) than the dimer-GDP end (more likely to depolymerize), so the dimer-GTP forms a depolmerizing cap (Bray, 1992). Rapidly growing microtubules contain large quantities of dimer-GTP and are less inclined to depolymerize but slowly growing microtubles contain more dimer-GDP and can depolymerize rapidly. This event is known as dynamic instability and may play a role in the shape changes of a cell during the early stages of initiation (Bray, 1992). New neurite outgrowth contains microtubules arranged with .the plus ends facing the growth cone (Hall, 1992). Dendrites will eventually have both plus and negative microtubules facing the growth cone, but axons remain 36


unipolar (Hall, 1992). Polymerization of microtubules can occur at the neck of the growth cone or the microtubules can be assembled and transported in the axon (Hall, 1992). Intracellular calcium concentrations play a role in microtubule fonnation and stabilization (Bennett and Weeds, 1986; Jacobson, 199i). High calcium cOncentrations are known to reduce neurite elongation (Audesirk et al., 1990; Mattson and Kater, 1987). Lowering the intracellular calcium concentration, then, may promote neurite elongation. The exact mechanism. remains unknown. .. 2.2.3 Neuroftlaments Neuro:filaments, which are classified as intennediate filaments (approximately 10 nm in diameter) fonn a support system for the neuronal cytoskeleton, but the components of the filaments are not actively adding and subtracting units. these neurofllaments are a stabilizing factor for the cytoskeleton. 2.3 Initiation Initiation of a new growth cone from the embryonic cell body may employ similar mechanisms. In order to restructure the cellular cytoskeleton to allow the fonnation of a growth cone, a large increase in calcium concentration may be 37


necessary .. "Hot spots" of high calcium concentrations have been observed in growth cones of NiE-115 cells (Silver et al, 1990). These "hot spots" may indicate a clustering ofcalcium channels, which when activated by normal depolarization of the neuron, produce a large increase in local calcium levels which serve as the first step in growth cone formation or movement (Silver et al., 1990). Calcium influx through a cluster of calcium channels may increase the intracellular calcium concentrations by as much. as 1 p.M (Silver et al., 1990). These large increases in intracellular calcium serve several purposes. First, the calcium ion. triggers events leading to cytoskeletal disruption and membrane "softening", and second, the active development of a growth cone (Rehder and Kater, 1992). Actin filaments (microfilaments) would be severed and microtubules destabilized to allow for restructuring (Forscher, 1989; Liu and Storm, 1990; Rehder and Kater, 1992). Several proteins and mechanisms are involved. Gelsolin severs actin filaments and is calcium dependent (Bray, 1992; Forscher, 1989). Gelsolin has two binding sites for G-actin, one binding site for F-actin, and one binding site for calcium (Bray, 1992). When calcium is available and binds to gelsolin, all three binding sites for the actin are active. However, in the absence of calcium, only the F-actin site is active (Bray, 1992). When calcium concentrations rise above 100 nM to 1 #'M, gelsolin birids to the sides of the actin fllaments and disrupts the bonding of actin monomers, thus severing the fllament (Bray, 1992; Rawn, 1989). Gelsolin 38


also binds to the barbed or plus end of the actin filament and so prevents the addition of more monomers 1992). In contrast to the mechanism of severing actin gelsolin has the ability to bind with two other monomers of actin and form a nucleus for the start of new actin filimants (Bray, 1992). This latter quality of gelsolin would be especially important in restructuring events in a growth cone which might be changing direction, in a transected axon which is fonning a new growth cone, and possibly in the development of a neurite or growth cone during initiation (Forscher, 1989; Rehder and Kater, 1992). When calcium concentrations return to normal, actin polymerization would be underway and gelsolin would become inactive due to a lack of calcium. This mechanism is complicated by several other events which would be occurring simultaneously in the cell and are described briefly below. 2.4 Hypothesis on Initiation The process of initiation is a complex event and a complete scenerio is not possible at this time, but a sketch is necessary to understand the importance of calcium channels in the process. The following chain of events, based on the process suggested by Rehder. and Kater (1992) may be possible. There would be several stages: (1) attachment of the cell to a substrate; (2) a rising calcium concentration. triggering events to disrupt the cytoskeleton; (3) a gradual decrease in calcium 39


concentration with concurrent stabilization of the cytoskeleton. (1) The first event is attachment of the cell to a substrate. This attachment may activate one or more G proteins. (2.A) G proteins may interact directly with calcium channels and open them (Brown et al., 1989; Dolphin and Scott, 1989). (2.B) Other G proteins may be activated which, in tum, activate phospholipase c. (3.A) Calcium enters the. cell through the activated calcium channels. The calcium now binds to several proteins: gelsolin, thus activating the gelsolin and beginning the process of severing existing actin microfiliments; calmodulin, which activates protein kinases; and, protein kinase C, which is also activated by high concentrations of calcium. Calmodulin may be bound to neuromodulin close to the membrane and thus available for early calcium entry through the calcium channels. Neuromodulin, also known as growth-associated protein or GAP-43, is a protein associated with growth cones. One of its characteristics is its binding affinity for calmodulin. It is phosphorylated by protein kinase C and when phosphorylated does not bind as well to calmodulin, and thus releases the calmodulin (Liu and Storm, 40


1990). The increasing activity of PKC, as described below, would generate more and more activated PKC, thus releasing more and more calmodulin from neuromodulin and the calmodulin would be activated immediately by the ever growing calcium concentration. Phospholipase C normally hydrolyzes IP2 (phosphatidylinositol 4,5bisphosphate) to produce IP3 and DAG. However, the IP2 may be bound to a protein. called profilin (eight IP2 molecules to one profilin molecule) (Forscher, 1989). Profilin also binds to actin monomers forming a profilin-actin complex which slows down actin polymerization (Forscher, 1989). IP2 competes with actin for the binding site on profilin so in times of high IP2 concentration actin is freed from the profilin actin complex and can particiipate in forming actin filaments. But high concentrations activated phospholipase C, which would occur through continued activation of G proteins, would eventually cause the hydrolysis of IP2 to occur, thus freeing the profilin from the IP2 and allowing the profilin to bind to actin and thus slow actin polymerization. As long as the profilin remained bound to the actin, actin polymerization would be slowed. The IP2 hydrolyzed to diacylglycerol (DAG) and inositol 1,4,5-triphosphate (IP3), would activate PKC via DAG and thus phosphorylate Ca2+ channels and open them to allow more calcium into the cell and the IP3 would activate calcium channels in the endoplasmic reticulum and thus release intracellular stores of calcium (Forscher, 1989). (Schuch et al. (1989) report that in 41


PC12 cells lowering PKC levels may activate calcium channels which is the opposite of this hypothesis for rat hippocampal cells). As a feedback control, the activated PKC would also be increasing the polymerization of actin fllaments (Forscher, 1989). These events would leave the cell fluid but adoptable while cell shape was changing. Since the initiation process involves a flattening of the cell, more and more cell surface area would be contacting the substrate and, conceivably, this would cause more and more activation of phospholipase C through a G protein mediated event. Also active during this period of rising Ca2+ concentration would be gelsolin which would be severing actin filaments and maintaining the cytoplasm in a fluid state. (3.C) Phospholipase C hydrolyzes IP2 which breaks down into IP3 and DAG. The IP3 may activate the release of calcium from iritracellular stores and the DAG activates protein kinase C (Saffell et al., 1992). Protein kinase C may phosphorylate neuromodulin which would free calmodulin. The phosphorylated neuromodulin inhibits phosphotidylinositol 4-phosphate kinase (Forscher, 1989). This kinase phosphorylates PIP (phosphatidylinositol) thus making IP2 which serves as the substrate for phospholipase C. Therefore, the phosphrylated neuromodulin may begin to slow the process by reducing the avalability of IP2 This, in tum reduces the availability of IP3 and DAG. (4.A) Restructuring and stabilization of the cytoskeleton would begin with a 42


gradual lowering of the calcium concentration. The decreased levels of DAG would lead to reduced levels of PKC and thus the calcium channels would gradually become dephosphorylated and inactive and leading to a lowering of intracellular calcium concentrations. If calcium concentrations. did not return to a normal level, the above events would continue and the cytoskeleton would remain unstable. As neurite formation occurs and growth cones on the neurites become the only anatomical parts of the cell in motion, the signal first triggered by attachment would localized to the growth cone. It is not lmown why this signal stops in the soma, but continues in the growth cone. Deactivation of G proteins may contribute to closing the calcium channels, except during their normal opening during action potential depolarization. As the channels close for longer periods of time, the calcium concentration would subside, as ATPase would be pumping calcium out of the cell. Gelsolin would cease its action as a severing tool of actin microfiliments would now bind to actin monomers forming nuclei upon which polymerization of actin microfiliments would occur. Microfiliments would become more stable. Calmodulin would bind to the . neuromodulin which has, presumably, been dephosphorylated by a phosphatase. Profilin would be binding to IP2 and, thus, allow the actin monomers to polymerize. 43


3. Methods 3.1. Cell Culture Embryonic rat hippocampal neurons were isolated and cryopreserved by a modification of the methods of Banker and Cowan (1977) and Mattson and Kater (1988) (Audesirk et al., 1991, 1993; Cabell and Audesirk, 1993; Kern et al., 1993). Briefly, hippocampi were dissected from E18 rat embryos obtained from timed pregnant Sprague-Dawley rats. Hippocampi were collected in Ca2+-Mg2+ free Hanks Balanced Salt Solution (HBSS), incubated in HBSS containing 2 mg/ml trypsin for 15 minutes, then incubated in 2 mg/ml trypsin inhibitor for 5 minutes. Cells were dissociated by trituration with a fire polished pipette, then cryopreserved as described by Mattson and Kater (1988). Cells were frozen in a mixture of Eagle's minimum essential medium (MEM) buffered with 10 mM NaHCQ3 and 25 mM HEPES, supplemented with 2 mM L-glutamine, 2% glucose, 1 mM sodium pyruvate, 10% fetal bovine serum, 8% dimethyl sufox.ide, and 15mM KCL (final concentration, 20mM K+). Cells were frozen at -70C. in 2 m1 freezing tubes (.25ml per tube, approximately 2 million cells) in styrofoam containers one inch thick. Hippocampal neurons were cultured by rapid thawing in a 37C water bath and immediate plating onto poly-D-lysine (MW 500,000) coated, 35 mm grided 44


dishes at a density of 200,000 cells/dish in 2 ml of culture medium. The medium consisted of Eagle's Minimum Essential Medium (MEM) with 25 mM HEPES (pH 7 .3), supplemented with 10 mM NaHC03 0.1% glucose (fmal concentration, 0.2% ), 2% fetal bovine serum, 1 mM sodium pyruvate, and 15 mM KCl (final concentration 20 mM KCl). Cells were cultured in medium alone. or in medium plus the desired concentration of calcium channel blocker. Specifically, stock solutions of the various calcium channel blockers were made (in DMSO) and aliquots of the stock solutions were added to the media to the desired concentration on the day the experiments were set up. Because the calcium channel blockers were dissolved in DMSO, all media, including control, contained 0.1% DMSO. Previous experiments showed that this concentration of DMSO does not alter any parameters of neurite development. The concentrations of the various calcium channel blockers chosen for these experiments varied considerably in their abilities to affect rat hippocampal outgrowth. Initially, concentrations which previous research in this lab and other labs had proven to be effective were tried (Audesirk et al., 1990). However, rat hippocampal cells proved to differ in their susceptibility to the calcium channel blockers therefore, I tested a range of concentrations until effects were seen. 45


After plating, the cells were incubated at 37C in a humidified, 5% C02 atmosphere for 2-4 hours to allow for attachment. The medium containing DMSO from the freezing medium was then suctioned from the plate and fresh, warmed medium was added to the plate. Attached living cells were counted in four 250X microscope fields, and the cultures returned to the incubator for 2 days; Living neurons (defined by trypan blue exclusion) were round, firmly attached to the substrate, and phase-bright (newly plated cells and cells that did not initiate neurites), or flattened, variably bright or dark (apparently depending on the degree of flattening) with no obvious vacuolization of the cytoplasm, and usually bearing processes. In cultures maintained for 48 hours, there were very few non-neuronal cells, and these could be easily distinguished from neurons morphologically and by immunostaining (Caceres, et al., 1986, and Kern, unpublished results). 3.1.2 Assessment of Survival and Differentiation After 48 hours, survival and neurite development were measured using a digitizing tablet and morphometric software (Sigmascan: Jandel Corp.). The following parameters were then calculated: (a) survival (the average number of cells alive in the same four 250X microscope fields that were examined at 4 hours); (b) percent neurite initiation (the percentage of living neurons that had produced at least one process); (c) percent axon cells (the percentage of living cells that produced an 46


axon, defined as a at least twice as long _as any ()ther process on a given cell and whose presence is used to define a pyramidal-like cell; Mattson and Kater, 1988); (d) mean axon length; (e) _mean bt:anches per axon; (t) mean number of dendrites per pyramidal-like cell; (g) mean dendrite length; and (h) mean number of per dendrite .. that measurements (d) through (h) requiTe distinction between axons and dendrites and were made only on cells that met the criterion for a pyramidal-:like cell (one process at least twice the length of any other process). 3.4 Statistical Analysis For morphology experiments, culture dishes were coded so that all measurements were performed "blind .. Data (before conversion to percent of controls; see Results below) were evaluated by analysis of variance followed by the Student-Newman Keuls test for multiple comparisons among experimental means or the Dunnett's test for a comparison of several experimental conditions with a single control.. 47


4. Results 4.1 Cell Culture Experiments The cell culture experiments were assessed and for eight parameters which included: (1) the numbers of neurons which survived, designated as those which were actively growing as as. ha4. not initiated neurites but were round, bright and attached to the substrate under phase-contrast microscope; (2) the numbers of neurons which had initiated neurites;_ (3) the percent of growing neurons which had developed a.Xons; (4) mean axon length; (5) number of branches per axon; (6) mean dendrite length; (7) number ofbranches per dendrite; (8) number of dendrites per axon cell; The highest concentrations of nifedipirie (100 and diltiazem (100 were tried briefly (n = 2). (See tables for nifedifine and diltiazem). The results of these experimentS were dramatic with so few cells growing that the experiments were stopped. Nifedipine (100 I-'M) caused initiation to drop to 43% of the controls and diltiazem (100 caused initiation to drop to 30% of the controls. Although neurites were present, very few of them were measurable and very few cells had axons. 48


.$:'\0 CONC I ----N %IN.IT SEM I I CONTR 53.44 I SEM 3.24 I I 1 Jl.M 7 47.33 I SEM 2.73 I 3 Jl.M 6 44.38 SEM 1.40 5 JJ.M 8 37.84 SEM 3.02 10 JJ.M 4 5.66 SEM 2.29 Table 1 NICARDIPINE CUMULATIVE DATA FOR CULTURE EXPERIMENTS #AXON. %AXON AXON AXON #BRNC %SURV CELLS. CELLS LENGTH BRNC /AXON /GROW LENGTH 62.85 8.75 125.5 22.85 0.63 4.56 0.93 4.63 7.57 3.09 0.11 74.64 7.71 ;!0".57 130.6 35.37 0.47 6.49 1.30 4.79 12.66 7.07 0.12 63.01 6.33 37.44 121.8 74.17 0 .23 2.19 0.76 2.03 9.98 23.69 o.o:s 51.06 4.13 27.41 105.8 18.38 0.31 2.80 1.03 7.69 10.18 3.7 0.15 48.78 --------------10.83 ------------------DE NOR #DEND #BRNC LENGTH /CELL /DEND 2.47 0.03 1.43 0.17 0.02 18.41 2.22 0.01 2.42 0.27 0.01 14.13 2.21 0.02 1.22 0.2 0.02 17.07 1.47 0 1.29 0.23 0 -----------------CONC = concentration of compound; SEM = standard error of the means; N = of dishes counted for data; %INIT = percentage of living which initiated neurites; %SURV = number of cells surviving after twenty-four-hours; #AXON CELLS. = the number of cells with axons; %AXON CELLS/GROW = percentage of growing cells having axons; AXON LENGTH = mean length of axons; AXON BRNC LENGTH = mean length of axon branches; #BRNC/AXON = numbers of branches per axon; DENDR LENGTH = mean length of dendrites; #DEND/CELL = numbers of dendrites per cell; #BRNC/DEND = numbers of branches per dendrite. ---= unmeasurable.


LJ1 0 CONC ----N %INIT SEM CONTR 4 67.63 SEM 5.54 10 IJ.M 4 52.74 SEM 2.47 15 IJ.M 2 36.29 SEM 10.58 20 IJ.M 4 28.29 SEM 8.16 30 IJ.M 2 4.29 SEM 1. 59 Table 2 NIMODIPINE CUMULATivE DATA FOR CULTURE EXPERIMENTS #AXON %AXON AXON .AXON #BRNC %SURV CELLS CELLS LENGTH BRNC /AXON /GROW LENGTH 54.58 11.00 52.04 138.4 28.58 0.67 2.05 2.86 20.74 19.0.5 4.8 0.07 53.70 11.25 60.03 138.6 22.19 0.30 5.55 1.31 6.01 0.72 1.16 0.14 63.14. 6.50 54.44 130.0 25.00 0.41 8.48 1.50 1.11 6.0 0.00 0.21 53.72 7.50 75.32 102.1 34.00 0.03 9.25 1.50 4.42 17.24 --0.03 29.74 0.50 -----------9.05 0.50 --------DENDR #DEND #BRNC LENGTH /CELL /DEND 16.75 3.11 0.04 1.57 0.53 0.02 15.40 2.88 0.01 0.76 0.07 0.01 12.94 2.49 0 2.52 0.11 0 13.39 1.76 0 0.23 a .17 0 ---------------CONC = concentration of compound; SEM = standard error of means; N = number of dishes counted for data; %INIT = percentage of living cells which initiated neurites; %SURV = number of cells surviving after twenty-four hours; #AXON CELLS = the number of cells with axons; %AXON CELLS/GROW = percentage of growing cells having axons; AXON LENGTH = mean length of axons; AXON BRNC LENGTH = mean length of axon #BRNC/AXON = numbers of branches per axon; DENDR LENGTH = mean length of dendrites; #DEND/CELL = numbers of dendrites per cell; #BRNC/DENb = numbers of branches per dendrite; ---= unmeasurable.


lJ1 ..._. CONC ----N %!NIT SEM CONTR 12 62.99 SEM 3.48 5 J.LM 3 65.83 SEM 2.92 20 J.LM 9 69.57 SEM 3.77 30 p.M 3 56.82 SEM 11.25 40 p.M 8 62.56 2.83 70 p.M 8 61.53 3.56 100 p.M 2 27.42 1. 61 Table 3: NIFEDIPINE CUMULATIVE DATA FOR CULTURE EXPERIMENTS #AXON %AXON AXON AXON #BRNC %SURV CELLS CELLS LENGTH BRNC /AXON /GROW LENGTH 65.57 12.25 51.62 134.3 29.5 0 .78 2 94 1.22 4.11 7.43 3.87 0.09 60.31 12.33 52.41 118.6 30.54 0.5 4.93 1.2 5.67 10.85 9.5 0.21 66.85 11.89 48.65 177.8 37.44 0.67 7.96 1.14 3.45 8.83 .3.42 0.10 62.23 10.67 56.07 134.6 66.52 0,5 10.42 2.73 4.72 0.89 32.5 0.11 60.35 10.75 51.47 188.5 28.78 0.95 3.11 1.24 5.45 8.55 2.31 0.18 60.51 10.25 47 . 9 144.6 25.99 0.76 3.45 0.96 4.61 22.28 3.13 0.13 51.82 2.00 ----------------5.75 2.00 ----------------DENDR #DEND #BRNC LENGTH /CELL /DEND 16.16 2.89 0.07 0.74 0.2 0.02 18.48 2.56 0.03 1.9 0.45 0.02 17.93 2.9 0.02 0.9 0.21 0.01 15.84 2.71 0 2.43 0.17 0 15.30 3.84 0.05 0.63 0.45 0.01 15.06 3.26 0.02 0.99 .030 0.01 ----------------------CONC = concentration of compound; SEM = standard error of the means; N = number of dishes counted for data; %!NIT = percentage of living cells which initiated neurites; %SURV = number of cells surviving after twenty-four hours; #AXON CELLS = the number of cells with axons; %AXON CELLS/GROW = percentage of growing cells having axons; AXON LENGTH = mean length of axons; AXON BRNC LENGTH= mean length of axon branches; #BRNC/AXON =numbers of branches per axon; DENDR LENGTH = mean length of dendri.tes; #DEND/CELL = numbers of dendrites per cell; #BRNC/DEND = of branches per dendrite; ---= unmeasurable.


lJ1 N -CONC ----N %INIT SEM. I CONTR 22 56.33 I SEM 1.77 100 nM 9 65.53 SEM 2.87 500 nM 14 64.53 SEM 3.77 1 J1M 7 63.25 SEM 2.21 5 J1M 6 54.66 SEM 4.96 25 J1M 7 50.96 SEM SO IJ.M 5 2.48 SEM 0.63 Table 4: ISRADIPINE CUMULATIVE DATA FOR CULTURE EXPERIMENTS #AXON %AXON AXON AXON #BRNC %SURV CELLS CELLS ,LENGTH B R N C /AxON /GROW LENGTH 65.49 8.91 44.43 145.1 21.51 1. 71 2. 92 0.64 2.79 5. 63 2.13 0.15 71.06 11.22 51.92. 178.3 20.24 1.81 4.66 0.91 3.14 6.71 1.63 0.31 77.71 11.00 54.01 173.4 24.56 1.85 3.6 0.83 3 . 75 7.47 1.61 0.15 80.68 12.57 58.85 170.2 20.59 2.05 4.92 0.87 3.42 10.3 2.02 0.31 59.42 9.00 49.68 162.0 22.99 1. 61 6.6 1.98 9.45 13.8 2.01 0.27 58.08 11.14 63.50 169.9 25.46 1.13 6.21 1.26 1. 89 12.83 2.51 0 .17 37.35 0.20 ----------------4.91 0.20 ------------------DENDR #DEND #BRNC LENGTH /CELL /DEND ; 16.86 3.56 0.14 0.71 0.19 0.02 16.27 3.66 0.14 0.94 0.16 0.02 16.61 : 4.06 0.19 0.73 0.19 0.02 16.86 3.93 0.16 .0.92 0.37 0.03 17.51 4.19 0.13 1.06 0.44 0.03 14.28 3.16 0.06 0.81 0.28 0.02 -----------------------CONC = concentration of compound; SEM = standard error of the means; N = number of dishes counted for data; %INIT = percentage of cells which initiated neurites; %SURV = number of cells surviving after twenty-four hours; #AXON CELLS = the number of cells with axons; %AXON CELLS/GROW = percentage of growing cells having axons; AXON LENGTH = mean length of axons; AXON BRNC LENGTH = mean length of axon branches; #BRNC/AXON = numbers of branches per axon; DENDR LENGTH = mean length of dendrites; #DEND/CELL = numbers of dendrites per cell; #BRNC/DEND numbers of branches per dendrite; ---= unmeasurable.


lJl w I CONC ----N %!NIT SEM CONTR 4 61.56 SEM 4.53 5 J.I.M 4 66.69 SEM 6.19 25 J.I.M 6 51.84 SEM 4.35 50 J.I.M 4 22.02 SEM 5.33 Table 5 VERAPAMIL CUMULATIVE DATA FOR CULTURE EXPERIMENTS ---#AXON %AXON AXON ,AXON #BRNC %SURV CELLS CELLS LENGTH BRNC /AXON /GROW LENGTH 58.45 10.00 46.55 125.9 22.63 1.57 3.79 0.41 8.90 2.82 2.95 0.09 58.62 10.50 43.75 136.5 21.98 1.73 6.52 1.55 4.73 5.77 0.77 0.09 47.38 7.17 36.85 109.9 19.88 0.49 7.31 1.80 7.53 5.78 4.39 0.15 35.37 --------------------4.38 ----------------DENDR #DEND #BRNC LENGTH /CELL /DEND 16.92 4.60 0.18 1.21 '0.24 0.05 17.77 3.53 0.10 1.3 0.29 0.06 14.88. 2.28 0.04 1.18 0.35 0.01 -------------------CONC = concentration of compound; SEM = standard error of the means; N = number of dishes counted for data; %!NIT = percentage of living cells which' initiated neurites; %SURV = number of cells surviving after twenty-four hours; #AXON CELLS = the number of cells with axons; %AXON CELLS/GROW = percentage of growing cells having axons; AXON LENGTH = mean length of axons; AXON BRNC LEGNTH = mean length of axon branches; #BRNC/AXON = numbers of branches per axon; DENDR LENGTH = mean length of :dendrites; #DEND/CELL = numbers of dendrites per cell; #BRNC/DEND = numbers of branches per dendrite; ---= unmeasurable.


U1 .c-CONC N --%!NIT SEM CONTRL 6 56.34 SEM 3.92 5 Jl.M 6 54.81 SEM 6.89 25 JJ.M 6 60.01 SEM 4.02 SO Jl.M 8 44.26 .SEM 3.98 100 JJ.M 2 17.38 SEM 12.03 Table 6 DILTIAZEM CUMULATivE DATA FOR CULTURE EXPERIMENTS #AXON %AXON AXON AXON #BRNC %SURV CELLS CELLS LENGTH BRNC /AXON /GROW LENGTH 61.11 7.17 35.72 128.9 17.0 1.34 9.27 1;54 5.95 1.59 0.34 59.22 8.33 45.99 134.2 15.77 1.66 8.16 1.28 3.47 16.22 1.14 0.3 53.03 10.17 50.05 168.6 20.15 1.25 8.6 0.98 4.45 16.04 3.13 0.33 45.91 4.38 31.25 133.8 23.51 1.38 4.97 0.32 3.06 8.92 4.41 0.23 30.8 0.5 -----------4.04 ---------------DENDR #DEND #BRNC LENGTH /CELL /,PEND 17.56 4.54 0.02 1.48 0.44 0.04 17.03 3.75 0.14 1.96 0.33 0.03 16.62 3.00 0.10 0.9 0.26 0.02 18.69 2.63 0.17 1.33 0.26 0.08 ------------------CONC = concentration of compound; SEM = standard error of the means; N = number of dishes counted for data; %INIT = percentage of living cells which initiated neurites; %SURV = number of cells surviving after twenty-four hours; #AXON CELLS = the number of cells with axons; %AXON CELLS/GROW = percentage of growing cells having axons; AXON LENGTH = mean length of axons; AXON BRNC LENGTH = mean length of axon branches; #BRNC/AXON = numbers of branches per axon; DENDR LENGTH = mean length of 'dendrites; #DBND/CELL = numbers of dendrites per cell; #BRNC/DEND = numbers of branches per dendrite; ---= unmeasurable.


l.J1 l.J1 ---CALCIUM CHANNEL BLOCKER NICARDIPINE NIMODIPINE NIFEDIPINE ISRADIPINE VERAPAMIL DILTIAZEM Table 7: CUMULATIVE DATA FOR CULTURE EXPERIMENTS % of controls ------% % %AXON AXON #BRNC DENDR CONC INIT SURV /GROW LENGTH /AXON LENGTH 1/LM 88. 57 118.7 86.06 104.00 74.02 92.34 31LM 83.05 100.2 79.41 97.05 36.02 66.00 SILM 70.81 81.25 58.15 84.25 48.97 52.00 10 ILM 10.59 77.63 0.00 ------10 ILM 77.98 98.39 115.36 100;11 -44.32 91.96 20 ILM 41.83 68.18 144.74 73.73 4.12 79.92 20 ILM 110.44 101.9 94.26 132.39 86.12 110.97 40 ILM 99.31 92.03 99.72 140.36 121-.56 94.69 70 ILM 97.68 93.25 92.80 142.69 97.67 93.19 100 nM 116.33 108.5 116.9 122.89 105.43 96.46 500nM 114.55 118.7 121.6 119.54 107.81 98.48 11LM 112.28 123.2 132.5 117.30 119.71 99.97 SILM 97.04 90.73 111.8 111.68 94.28 103.83 25/LM 90 46 88.69 142.9 117.14 65.71 84.67 SOILM 4.40 57.03 ----------SILM 108.34 100.3 93.97 108.43 110.34 105.02 25 ILM 81.91 77.76 62.64 87.16 30.11 94.52 50 ILM 35.77 60.51 0.00 ------. --SILM 97.29 96.89 128.76 104.10 123.53 96.97 25/LM 106.51 86.77 140.11 130.85 92.99 94.66 SO ILM 78.56 75.12 87.48 103.84 102.99 106.43 -----#DEND #BRNCH /CELL /DENDR 89.68 23.69 89.42 86.90 59.47 0.00 -----92.57 24.35 56.48 0.00 100.62 22.11 132.93 62.61 112.98 22.01 102.57 99.31 114.01 136.83 110.14 114.34 117.52 94.82 88.65 41.57 -----76.71 56.92 44.37 17.79 ------82.68 67.95 66.00 47.22 58.03 84.83


V1 Q'\ I Table 8: TRENDS IN DATA FOR CULTURE EXPERIMENTS % of controls: NC =within 5% of controls; single arrow = not significant; double arrow = significant CALCIUM % % %AXON AXON #BRNC DENDR #DEND CHANNEL CONC INIT SURV /GROW LENGTH /AXON LENGTI! /CELL BLOCKER 111M t NC ' 3JLM NC NC ' NICARDIPINE 5JLM H H ' H 10 JLM H -------10 11M NC t NC H ' NIMODIPINE 20 ILM H t H H 20 11M t NC tt t NC NIFEDIPINE 40 11M NC NC t t t t 70 JLM NC ' t t NC t 100 nM t t tt t t t t NC NC 500nM t t t t t t t t NC t lJLM t t t t t t NC t ISRADIPINE 5J1M NC t t NC t 25JLM ' tt t ' 50 JLM H H -------5JLM t NC t t NC VERAPAMIL 251J.M ' ' H H 50 ILM H H --------5J1M NC NC t NC t NC DILTIAZEM 251J.M t t t ' H 50 !J.M ' NC NC t H .. #BRNCH /DENDR ' H H NC t t ' -' ' '


The concentrations of calcium channel blockers used in these experiments to elicit effects may seem unusually high when compared to concentrations used with other cell types or during electrophysiological experiments. At least one other group of researchers, Zhang et al. (1993), found that high concentrations of some calcium channel blockers were necessary to reduce rising intracellular calcium concentrations due to high potassium mediated calcium influx. Zhang et al. (1993) used rat cerebellar granule cells for their experiments. (See the discussion section for details). 4.1.1 Effects of the Dihydropyridines, the Phenylalylamine, and the Benzothiazipine on Cell Culture Survival Survival decreased with virtually every calcium channel blocker. tested at the higher concentrations. (See. tables and graphs on survival). The effects of both isradipine at 1 JLM, 500 nM, and 100 nM, and nicardipine at 1 JLM were to increase survival. The 500 nM and 1 JLM concentrations of isradipine were statisically significant (p < 0.05). At the higher concentrations, nicardipine (5 and 10 nimodipine (20 JLM), nifedipine (100 JLM)(n = 2), isradipine (25 JLM and 50 verapamil (25 JLM and 50 ILM) and diltiazem (50 JLM and 100 [n = 2]) decreased . survival, which was statistically significant for isradipine at 50 and verapamil at 57


......:l > 0:: rJ) SURVIVAL AS A -PERCENT OF CONTROLS 120 100 80 60 ... v--------------------\7 .... ..... -20 0 20 CONCENTRATION ISRADIPINE \7 NIFEDIPINE .., VERAPAMIL 58 40 60 80 uM o DILTIAZEM NICARDIPINE !:::, NIMODIPINE


z 0 1--1 E--t

50 pM (p < 0.05). Nifedipine at concentrations below 100 pM had very little effect on survival. .1.1.2 Initiation Initiation was severely reduced by all three classes of calcium channel blockers tested, particularly at the highest concentrations. Nicardipine (both 5 pM and 10 pM), nimodipine (20 pM) and isradipine (50 pM) significantly reduced initiation at the highest concentrations tested (p < 0.05). (See tables for nimodipine and isradipine. Also, see the graph for initiation). Isradipine actually increased initiation in concentrations of 1 pM and below and these increases were significant at 500 nM and 100 nM (p = 0.05). Nifedipine decreased initiation (to 43% of the controls) at the 100 pM concentration (n = 2) but had very little effect on initiation at lower concentrations. Verapamil significantly decreased initiation at 50 pM (p < 0.05). Diltiazem reduced initation at the 50 pM concentration but not significantly. At 100 pM (n = 2), diltiazem reduced initiation to 31% of the controls. (See table for diltiazem.) Axon Cells as a Percentage of Growing Cells This parameter measured the the numbers of cells which had axons as a percentage of the numbers of cells which were growing, i.e., had neurites. Axons 60




AXON LENGTH AS A PERCENT OF CONTROLS ::r:: 100 CJ 50 z C:i:l z 0 >< 0 ... \l"ti" ....................... .,.._.r.- ....... . -,..__ .............. o "' .,_ \_"'\ ... ....... -20 0 20 CONCENTRATION ISRADIPINE \7 NIFEDIPINE ..-VERAPAMIL 40 60 80 62 uM o DILTIAZEM NICARDIPINE ::, NIMODIPINE


were at least two times as long as the longest dendrite. This parameter was difficult to interpret as it seemed inconsistent. Nicardipine and verapamil caused gradl:Jal decreases in the percent of axon cells per growing cells as concentrations increased. Nicardipine at 5 ,uM greatly reduced the percentage of axonal cells and at 10 ,uM reduced axonal cells to zero. Nifedipine had little effect except at 100 ,uM (n = 2) which revealed a marked decrease (to 48% of the controls). (See the table for nifedipine.) Nimodipine (10 and 20 ,uM), isradipine (all except 50 and diltiazem .(5 and 25 ,uM) caused an increase in the percentage of axon cells per growing cells. lsradipine, at 25 ,uM and 1 caused statistically significant (p < 0.05) increases in the percentage of axon cells per growing cells. Diltiazem decreased this parameter at .the higher concentrations of 50 and 100 (n = 2). (See tables and graph for the percent axon cells per growing cells.) Axon Length All of the calcium channel blockers reduced axon length at the highest concentrations tested. (See graph for axon length.) However, several of them caused an increase in axonal length at lower cocentrations. Both nicardipine and nimodipine gradually decreased axonal length as concentrations increased. Nifedipine had statistically significant increases in axonal length at concentrations of 20 40 and 70 (p < 0.05). lsradipine increased axonal length at concentrations of 100 63


z 0 >< CiJ 0.. (/) CiJ ::r:: u z (:0 0 (/) CiJ (:0 z 140 120 100 80 60 40 20 0 -20 -20 BRANCHES PER AXON AS A .PERCENT OF CONTROLS 0 ... ,"V-. .... '-..... ,'',' ... -----v ......... .. \\ li T._ 0. 20 ... CONCENTRATION - 40 60 uM o DILTIAZEM 80 ISRADIPINE v NIFEDIPINE ..-VERAPAMIL NICARDIPINE 6 NIMODIPINE 64


nM, 500 nM, 1 #-'M, 5 I-'M and 25 I-'M. These increases were statistically significant at the 100 nM and 500 nM concentrations (p < 0.05). At the highest concentration of isradipine (50 I-'M) there were no axonal cells which indicates a severe effect on axonal length. Verapamil had little effect at 5 I-'M and 25 I-'M and at 50 I-'M there were no axonal cells, again indicating a severe effect on axonal length. Diltiazem had little effect at 5 I-'M and at 50 I-'M but increased axonal length at 25 I-'M. Numbers of Branches Per Axon All of the calcium channel blokers reduced this parameter in the highest concentrations tested including nifedipine at 100 I-'M (to zero [n = 2]) and diltiazem at 100 I-'M (also to zero [n = 2]). (See tables for nifedipine and diltiazem.) Nicardipine and nimodipine decreased the numbers of branches per axon at all concentrations. (See graph for branches per axon.) These were significant (p < 0.05) for nimodipine at both 10 I-'M and 20 I-'M. Nifedipine decreased the numbers of branches per axon at 20 I-'M and increased this parameter at 70 #-tm, neither of which was significant. Isradipine increased the numbers of branches per axon at 100 nM, 500 nM, and 1 pM and decreased this parameter at both 5 pM and 25 pM, with none of the above being significant. Verapamil significantly decreased the numbers of branches per axon at 25 pM (p < 0.05). Diltiazem had little effect on this parameter except at the 100 pM concentration mentioned above. (See table for




....J ....J iJ :.J iJ /) iJ :::::) 2:; iJ :::::) I.,. :) n iJ !:1 ::l 2:; 140 120 100 80 60 40 20 0 -20 -20 NUMBERS OF DENDRITES PER CELL AS A PERCENT OF CONTROLS .. \ :.... '\--o-- ' "'.\ .... 0 20. 40 CONCENTRATION uM 60 80 ISRADIPINE v NIFEDIPINE -.. VERAPAMIL o DILTIAZEM NICARDIPINE NIMODIPINE 6T


diltiazem.) Dendrite Length The highest concentrations tested for each of these compounds, including nifedipine at 100 (n = 2) and diltiazem at 100 (n = 2), decreased the length of the dendrites. (See tables for nifedipine and diltiazem.) Nicardipine and nimodipine decreased dendrite length a8 concentrations. increased but none. of the changes were statistically significant. Nifedipine, isradipine, verapamil and diltiazem had little effect on dendrite length at lower concentrations. (See graph for dendrite length.) Numbers of Dendrites Per Cell With the exception of nifedipine, all of the calcium channel blockers tested reduced the numbers of dendrites per axon cell. (See graph for dendrites per cell.) Nicardipine caused a significant drop in the _numbers of dendrites per cell at 5 (p < 0.05) and at 10 there were no dendrites. Nimodipine significantly decreased the numbers of dendrites per cell at 20 (p < 0.05). Nifedipine increased the numbers of dendrites at 40 uM and 70 uM, but not signifieantly. Isradipine increased the numbers of dendrites at 100 nM, 500 nM, and 5 and 68


-ec:: NUMBERS OF BRANCHES PER DENDRITE AS A PERCENT OF CONTROLS Q 1 50 .--------.----.-----.-------.----'-----. z w Q ec:: w 0,.; 1 00 rJ) C:il :c u 2 co 0 50 /0 ... .n /. ; ... ':,'," _/; ...... __ . ...... ... , ...... rJ) 0 ',,., .. 6 ... ;::q ::J z :_20 0 20 40 60 80 CONCENTRATION uM ISRADIPINE o DILTIAZEM v NIFEDIPINE NICARDIPINE T VERAP AMIL 6 NIMODIPINE 69


decreased this parameter at 25 I'M. Both verapamil (25 I'M) and diltiazem (25 I'M and 50 I'M) caused a significant drop in the numbers of dendrites per cell (p < 0.05). Numbers of Branches Per Dendrite With exception of lsradipirie at 500 nM and li'M, all of the calcium channel blockers decreased the numbers of branches per dendrite. (See graph for branches per dendrite.) This decrease was stastically significant for nifedipine at 20 I'M and 70 JLM and isradipine at 25 I'M. Isradipine increased this parameter at 500 nM and 1 JLM but not significantly. 70


5. Discusion 5.1 Effects of Calcium Channel Blockers on Neurite Outgrowth Initiation The original hypothesis was that intracellular calcium concentrations were important in regulating the process of initiating neurites from a neuron. It also included the premise that an influx of calcium through L-type calcium channels was directly involved in regulating the intracellular calcium concentrations and thus, the initiation of neurites. Using six different, known L-type calcium channel blockers, the initiation of neurites in rat hippocampal cells was analyzed. Nicardipine and nimodipine were potent on rat hippocampal cells. Nicardipine at 5 p.M and 10 p.M and nimodipine at 10 p.M and 20 p.M reduced initiation. These results indicate that as intracellular calcium concentrations are reduced by curtailing the influx of calcium through L-type calcium channels, the initiation of neurites is also reduced. Isradipine may be an L-type calcium channel agonist at lower concentrations and an L-type calcium channel antagonist at higher concentrations (see next section for 71


a discussion of this phenomenon). Initiation was increased significantly by isradipine at both 100.nM and 500 nM concentrations and initiation was still above the controls at 1 /LM. At 5 /LM, the initiation was just slightly below the controls. These results suggest that higher than normal intracellular calcium concentrations would increase initiation. At higher concentrations of isradipine (25 ILM and 50 ILM) initiation was reduced and was 4% of the controls at 50 /LM. The higher concentrations had the same effects on 'initiation as did the other calcium channel blockers tested in this thesis, i.e., they reduced initiation. These results suggest that isradipine is acting in a biphasic manner: as an agonist at concentrations of 5 ILM and below and as an antagonist at concentrations above 5 I'M. This phenomenon is discussed in some detail below (see section 5.1.2). Nifedipine had very little effect on initiation except at 100 /LM concentrations in which initiation was reduced (n = 2, see table for nifedipine). Several other studies also indicate that nifedipine has little effect on some types of cells except at the 100 ILM concentration (Zhang et al., 1993; Regan and Choi, 1994). Verapamil began to affect initiation at 25 ILM and significantly affected initiation at 50 /LM. Diltiazem affected initiation at 50 ILM but not at lower concentrations. 72


Initiation may begin with an intracellular calcium concentration dependent event (possibly activation of a calcium dependent kinase(s), i.e., protein kinase c [Cabell and Audesirk, 1993]) mediated by an influx of calcium through L-type calcium channels, which is triggered by attachment to a particular substrate by one or more of the several cell membrane attachment molecules, i.e., integrins, CAMs, and N cadherin (see chapter 2). Blocking this influx of calcium through L-type calcium channels would inhibit the ability of the cell to respond to the intracellular signaling induced by the attachment process. Alternately, if a cell responded to the intracellular signal induced by the attachment process with a slightly higher than normal calcium concentration produced by influx through L-type calcium channels, then the initiation process might be enhanced. lsradipine, in nanomolar concentrations, may be acting in this fashion, 5.1.2 Isradipine Some dihydropyridines exist as stereoisomers called enantiomers (Williams et al., 1985; Hering et al., 1989; Priego et al., 1993). Enantiomers are chemical structures which are mirror images of each other and are not superimposable. These enantiomers are labelled with a ( +) or a (-) indicating which way the pure enantiomet rotates light: a (+}indicating a rotation of light in the clockwise direction and a(-) indicating a rotation in the counterclockwise direction. Bay K 8644, which 73


is known for its agonist abilities on the L-type calcium channel consists of two enantiomers which have opposite effects on the L-type calcium channel (Priego et al., 1993). The(+) enantiomer is a weak L-type calcium channel antagonist and the(-) enantiomer is a very strong L-type calcium channel agonist (Priego et al., 1993; Franckowiak et al., 1985). Bay K 8644 seems to be ubiquitous in its agonist abilities towards the L-type calcium channel, but other enantiomers are specific for tissue types (Priego et al, 1993). Sandoz 202791, which has a structure very similar to isradipine (see figure on next page) consists of a ( +) enantiomer which is a weak L-type calcium channel agonist and a (-) enantiomer which is a strong Ltype calcium channel antagonist (Hering et al., 1989). Normally, Sandoz 202791 is Fig. 7 Isradipine: showing the exchange considered to be an agonist of L-type of groups of the 3 1 position and 51 position of the dihydropyridine ring. structure used with permission from Trends in Pharmacol. Sci. Catterall, W.A. and Striessnig, J., 1992. 74


. CHENnCALSTRUCTURES DIHYDROPYRIDINES c-o-cH, I 0 ..... ISRADIPINE, a reported antagonist SANDOZ 202791, a reported agonist Isradipine structure used with permission from Trends in Pharmacol. Sci. Catterall, W.A. and Striessnig, J., 1992. Sandoz 202791 structure used with permission from the European Journal of Pharmacology. Priego et al., 1993. 74a


calcium channels. Isradipine also consists of two stereoisomers in a racemic mixture (personal communications from Research Biochemicals, Inc.). The enantiomers of isradipine exist by exchanging the groups at the three position and the five position of the dihydropyridine ring (see figure 7). The enantiomers of Sandoz 202791 exist in the same fashion. The similarity of the .chemical structures of isradipine and Sandoz 202791 suggests the possibility that they might share characteristic properties with the L-type calcium channel. The fact that isradipine has been characterized as having predominantly antagonistic qualities to the L-type calcium channel may indicate that one of the enantiomers is a strong antagonist while the other enantiomer is a weak agonist. In rat hippocampal cells, low concentrations of isradipine may favor binding of the weaker agonist enantiomer. PCA 50941, another dihydropyridine, actually consists of four enantiomers since there are diester substiutions instead of monoester substiutions at the 3' and the 5' positions of the 1,4-dihydropyridine ring (Priego et al., 1993)(see figure next page). This compound has unusual biphasic activity on pig coronary artery contractions. It is an agonist to L-type calcium channels at concentrations below 1 JLM and an antagonist to the L-type calcium channel at concentrations above 1 JLM (Priego et al., 1993). This biphasic activity causes contractions in the pig coronary artery at lower concentrations (below 1 JLM) and relaxation of the arteries at higher concentrations (above 1 JLM) (Priego et al., 1993). (Note: an agonist to an L-type calcium channel 75


CHEN.UCALSTRUCTURES DlliYDROPYRIDINES BAY K 8644, a reported agonist 0 .. 0 '0 0 H PCA 50941, a reported agonist Bay K 8644 and PCA 50941 structures used with permission from the European Journal of Pharmacology and Dr. Priego. Priego et al., 1993. 75a


will increase intracellular calcium concentrations in the muscle cell thus activating the calcium dependent myosin, troponin unit and causing contractions of the muscle, i.e., vasoconstiction. The opposite is true for antagonists which cause vasorelaxation, i.e., vasodilation). In the rat portal vein, PCA 50941 reduced contractions, thus acting as an antagonist to the L-type calcium channel (Priego et al., 1993). Bay K 8644 exhibited purely characteristics (acting as an agonist) in both of these instances (Priego et al., 1993). PCA 50941 also produced biphasic activity in rat aorta causing vasodilation in concentrations below 1 #'M (acting as an agonist) and reversing that in concentrations above 1 #'M (Priego et al., 1993). Again, Bay K 8644 was purely a vasoconstrictor in this instance. In rat portal vein PCA 50941 reduced contractions :(acted as an antagonist) while Bay K 8644 acted as an agonist and increased contractions (Priego et al., 1993). In short, PCA 50941 exhibited some biphasic activity and some agonist activity that was tissue specific. Binding affinities of the different enantiomers of a particular calcium channel agonist/antagonist will vary (Hering et al., 1989; Williams et al., 1985). This may be the cause of the biphasic patterns of activity seen when using a racemic mixture of a dihydropyridine (Williams et al., 1985; Priego et al., 1993; Hering et al., 1989). Hering et al. (1989) suggest an interesting model in which the L-type calcium channel antagonists/agonists change the rate constants between the open and closed states of the L-type calcium channel. However, there appear to be many hypothesis, too 76


numerous to detail here (Hering et al., 1989). The 100 nM, 500 nM, and 1 JLM concentrations of isradipine also seemed to enhance the other .cell growth parameters that were measured in these experiments. These included survival, the percent of axon cells per growing cells, the axon length, the numbers of branches per axon, the numbers of dendrites per cell, and the numbers of branches per dendrite. Dendrite length seemed to be unaffected by isradipine at these low concentrations. These data suggest that a higher than normal intracellular calcium concentration which is mediated by an influx of calcium through L-type calcium channels may enhance many growth parameters of neurons (Rogers and Hendry, 1990). An extremely high influx of calcium through L-type calcium channels greatly increasing intracellular calcium concentrations may inhibit neurite outgrowth (Robson and Burgoyne, 1989; Baret al., 1993; Kobayashi et al., 1992). s.t.3. Survival In the lowest concentrations, all of the compounds used either produced no change or, in the case of isradipine increased the survival, of the neurons. In the higher concentrations, all of the calcium channel blockers reduced survival. These results are in accord with other published research which indicates that potassium mediated depolarization of neurons which, presumably, opens the voltage-dependent 77


L-type calcium channels, greatly enhances survivability of neurons (Collins et al., 1991). Isradipine, which may be acting as. an agonist at the nanomolar concentrations, follows this hypothesis in enhancing survival. The reduction of survival as the concentrations of the L-type calcium channel blockers increase, thus decreasing intracellular calcium, is also compatible with this hypothesis. 5.1.4 Numbers of Cells with Axons Per Numbers of Cells Growing This parameter appeared inconsistent iii my results. lsradipine, at all concentrations except the highest tested (50 JLM) at which there were no axonal cells, enhanced this Diltiazem (5 JLM and 25 JLM) and nimodipine seem to enhance this parameter as well. Nicardipine, verapamil and decreased this parameter at the highest concentrations tested. Nifedipine did not seem to have much effect. As discussed in the Chapter 2 calcium may play a role in a number of growth parameters. One hypothesis to account for an increase in the number of axonal cells is that of inhibiting a calcium dependent enzyme such as gelsolin which normally may be activated by calcium. The reduction of calcium .would inhibit this enzyme and so promote microtubule elongation by preventing microtubule depolymerization. This would enhance the formation of an axon from the neurite which had been committed to becoming an axon. Reduction of this parameter may be due to reduced axonal elongation so that no one neurite becomes long enough to meet the criteria of the 78


laboratory of being an axon, i.e., twice as long as the other neurites. Another possibility is that there is a failure of one of the neurites to become committed to becoming an axon. 5.1.5 Elongation of Axons and Dendrites Calcium channel blockers are known to affect the elongation of neurites in some cell types (Audesirk et al., 1990; Mattson and Kater, 1987). The mechanisms underlying elongation seems to be uncertain but the involvement of calcium channels in some way is plausible (Mattson and Kater, 1987). Axonal length may increase as both organic and inorganic L-type calcium channel blockers are used in culture. Cabell and Audesirk (1993), reduced axonal length using a calcium-calmodulin dependent protein kinase n inhibitor (KN-62) indicating the involvement of protein kinase IT. Indeed, research on tau and MAP proteins indicate that phosphorylation of these proteins makes them longer and stiffer (Diez-Guerra and Avila, 1993; Friedrich and Aszodi, 1991). So, inhibiting kinases could prevent or hinder elongation. However, blocking calcium channels would not seem to be involved with activating kinases since some of them are dependent upon calcium for activation, i.e. Ca2+ -calmodulin dependent kinase. Instead, the elongation seen with calcium channel blockers may involve the deactivation of a degradation enzyme which may be present and somewhat active throughout the growth process of axons, i.e., gelsolin or another 79


enzyme yet to be identified. Blocking calcium channels would prevent activation of this enzyme by calcium and would induce elongation. Dendrite length went down somewhat with both nicardipine and nimodipine but was unaffected by the other calcium channel blockers used in these experiments. This parameter remains elusive as Cabell and Audesirk (1993) were unable to affect dendrite length with any kinase inhibitors or phorbal esters. 5.1.6 Numbers of Branches on Axons and Dendrites The branching .of both and dendrites increased with 500 nM and 1 isradipine concentrations. Most other parameters of growth were also enchanced by these concentrations of isradipine making conclusive remarks difficult as this may be a result of increased overall growth. However, in higher concentrations isradipine decreased both of these parameters and the possibility exists that blocking L-type calcium channels may have a role in branching. As mentioned above in_ relation to elongation, phosphorylation seems to play a part in lengthening and stiffening certain MAPs and certain kinases are calcium dependent so that blocking calcium channels would prevent acivation of the kinases and thus, hinder phosphorylation (Diez-Guerra and Avila, 1993; Friedrich and Aszodi, 1991). This process ofphosphorylation .. may also be the key to branching as certain MAPs have an arm which protrudes to the 80


side of the molecule and which, if stiffened, may push the microtuble bundles away from each other and so create a branch to the side upon which microtubule assembly could occur (Friedrich and Aszodi, 1991). It is interesting that nifedipine, which had very little effect on most other parameters of growth, significantly reduced dendrite branching at both 20 JLM and 70 JLM and it also significantly increased axonal length while axon branching remained essentially unchanged. Is it possible that nifedipine inhibits a particular kinase in dendrites? Various kinase inhibitors decrease dendrite branching. KN-62, a calcium-calmdulin dependent kinase inhibitor is very effective (Cabell and Audesirk, 1993). The other calcium channel blockers also reduced dendrite and axon branching at higher concentrations, but here again, other growth parameters were also affected. 5.2 A DiScussion of the Potencies of the Calcium Channel Blockers The potencies of the various compounds may vary for several reasons. The ability of the compound to either enter the membrane to approach the S4 helix of the channel from a lateral aspect (the dihydropyridines), or pass entirely through the membrane to approach the channel from the intracellular aspect (verapamil), may be one determination of their potencies (although this would not affect diltiazem). 81


Partition coefficients would be an indication of this parameter, but this could vary from cell type to cell type depending upon the constituents of the plasma membranes (Mason et al., 1989). Nifedipine was potent at concentrations of 1 I'M to 50 I'M with chick neurons, but seemed to have little effect on rat hippocampal cells in concentrations up to 100 I'M in my experiments (Audesirk et al., 1990). Experiments performed with this particular cell type, i.e., rat hippocampal cells, may not have a bearing on experiments done with other cell types. Zhang et al. (1993), seemed to see similar effects (similar to my experimental effects) on intracellular calcium concentrations using rat cerebellar cells and nicardipine, nifedipine, verapamil and diltiazem. They used 50 mM KCl to depolarize the rat cerebellar cells which made intracellular calcium concentrations rise to 800 nM. The calcium channel blockers were used to prevent this rise in intracellular calcium concentrations. They found that 10 I'M nicardipine inhibited the rise in intracellular calcium by 80% and 100 I'M nifedipine inhibited the rise by only 43%. Both verapamil at 50 I'M and.diltiazem at 100 /LM concentrations inhibited the rise by approximately 50%. My experiments seem to complement these results with nicardipine showing significant effects at 5uM and lOuM; nifedipine showing very little change untillOO /LM concentrations were reached (n = 2)(data shown in table on nifedipine); verapamil had significant effects at 50 I'M; and diltiazem had some effect at 50 ILM but greatly reduced all paramaters at 100 ILM (n = 2)(data 82


shown in table on diltiazem. See results section). Initiation was dramatically reduced in all cases when a critical concentration for each of the calcium channel blockers was reached. The four dihydropyridines had an order of potency as nicardipine > nimodipine > isradipine > nifedipine. Verapamil slightly more potent than diltiazem. Isradipine, which was tested in nanomolar concentrations seemed to enhance growth at those lower concentrations. No other cell type tested and reported in the literature seems to react to isradipine in this way. With a 5 concentration, nicardipine reduced initiation, axon length, numbers of processes,' dendrite length, and branching. Nimodipme had similar results at 15 isradipine at 50 and nifedipine at 100 Electrophysiological stUdies using patch-clamping techniques are an indication of the potencies of various compounds, but when cells are clamped to a particular depolarized potential, the various channel blockers appear more potent than they seem to be in vivo (O'Dell and Alger, 1991; Cohen and McCarthy, 1987; Bohle, 1992). O'Dell and Alger found that 10 nifedipine and 20 nimodipine significantly reduced L-channel conductance in rat hippocampal neurons when these cells were stepped from -80 mV to -40 mV. Cohen and McCarthy (1987) used nimodipine to block calcium channels in pituitary cells at concentrations as low as 10 They showed that at those low concentrations block of the channel took as much as 150 mS 83


when the cell was stepped from a resting state of -90 m V to -20 m V. Their experiments indicate both a time and voltage dependent effect with nimodipine. The time differential may be. due to the amount of time necessary for a low concentration of the drug to enter the plasma membrane and approach the appropriate binding site on S4 helix or other region of the channel. Once the nimodipine reaches the channel there is a 90% blockage of the calcium channel. It is possible that a critical concentration of the drug must be reached before blockage of the channel occurs. It has been estimated that the actual concentration of the drug in the plasma membrane is many times the concentration of the drug in solution outside of the cell (Mason et al., 1989). 5.3 Overall Effects of Blocking L-type Calcium Channels Several possibilities exist for the inhibition of neurite outgrowth due to the blockage of L-type calcium channels. Nerve growth factor (NGF) has been used to stimulate sympathetic superior cervical ganglion cells taken from 2 to 4 day old rat pups (Rogers and Hendry, 1990). The neurite outgrowth was attenuated by various calcium channel blockers at. differing concentrations including cadinium at an IC50 of 48 JLM, Cobalt at an IC50 of 129 JLM, diltiazem at an IC50 of 17 JLM, and nifedipine at an IC50 of 3 JLM. Rogers and Hendry (1990) also used Bay K 8644 (a calcium channel agonist) in their experiments. The Bay K 8644, in nanomolar 84

PAGE 100

concentrations, enhanced NGF-induced neurite outgrowth by as much as forty percent, indicating that a higher concentration of intracellular calcium improved neurite outgrowth or, at least, the greater calcium ion flux through L-type calcium channels, improved the neurite outgrowth (see Rasmussen (1986) for a discussion of influx and efflux of calcium ions). This increase in calcium concentration and/or calcium ion flux would have a ceiling as several authors report that high concentrations of potassium which should depolarize the cells and open calcium channels, severely hamper neurite outgrowth (Bar et al., 1993; Kobayashi et al., 1992; Robson and Burgoyne, 1989). However, these experiments with 50 mM potassium seem to the intracellular calcium levels far beyond what would be considered physiologically sound for the neuron cell body, i.e. to 800 nM and above (see Zhang et al., 1993). It may be that slightly higher concentrations of intracellular calcium than normal, perhaps 150 nM to 200 nM, will increase neurite outgrowth as was demonstrated by Rogers and Henry with the use of Bay K 8644, but large increases in calcium concentration or influx through L-type calcium channels will greatly decrease the ability of the cell to produce neurite outgrowth. This may explain the apparent increase in neurite outgrowth that my experiments with isradipinme produced in nanomolar concentrations. Bay K 8644 may act as an agonist or an antagonist to the calcium channel depending upon the isomer which has the predominent effect at the particular concentration used (Rogers and Hendry, 1990). This same phenomonon may occur with isradipine depending upon the 85

PAGE 101

concentration used (see above, section 5.1.2). Another possibility for the apparent enhancement of neurite growth observed with the lower concentrations of isradipine may have been due to isradipine binding to membrane bound calmodulin or unidentified kinases. Some studies suggest that calcium antagonists such as those used in these experiments may bind to calmodulin and may affect the activation of protein kinases which are dependent on calmodulin. However, the overexpression of calcium-calmodulin dependent protein kinase ll has been shown to promote neurite outgrowth in NG108-15 and Neuro2A cells (Goshima et al., 1993). Daly et al. (1983) found that one micromolar concentrations of diltiazem, verapamil, and nifedipine failed to inhibit or activate a calcium-calmodulin dependent phosphodiesterase. Other studies indicate that there are definite binding sites on calmodulin for the calcium antagonists such as those used in my experiments but that the activity of calmodulin may not be affected (Schaeffer et al., 1991; Tanaka et al., 1983). 5.4 Conclusion The hypothesis that fluxes in intracellular calcium concentrations mediated by Ltype calcium channels affect neurite outgrowth and especially initiation in embryonic 86

PAGE 102

neurons was supported by this thesis. The four dihydropyridines, nifedipine, nicardipine, nimodipine, and isradipine, which are all known to be L-type calcium channel blockers, affected neurite outgrowth .in the highest concnetrations tested. Nifedipine was the least effective of the dihydropyridines. Nlcardipine and nimodipine were very close in the concentrations which were effective on neurite outgrowth. The unexpeCted finding that isradipine could enhance growth parameters at lower concentrations shed new light on the capabilities of that particular compound. At higher concentrations isradipine was as effctive as nicardipine and nimodipine on neurite outgrowth. Verapamil and diltiazem were originally included in the experiments to verify the results gathered from the use of dyhydropyridines. If the dihydropyridines blocked L-type calcium channels and had a particular effect on neurite outgrowth then other L-type calcium channel blockers should have a similar effect on neurite outgrowth. Both veraparnil and diltiazem affected neurite outgrowth in rat hippocampal cells in much the same fashion as the dihydropyridines although they were not as potent as nicardipine, nimodipine and. isradipine. The study of calcium channel blockers and their effects on neurite outgrowth is a complex but challenging subject which deserves more study. The dyhydropyridines were particularly interesting and there are dozens of them currently in various stages 87

PAGE 103

of testing throughout the world. Some of them might have unusual effects on rat hippocampal cells grown in culture. There are also a number of other L-type calcium channel blockers available which might prove interesting. 88

PAGE 104

Bibliography Arcangeli, A., Bechetti, A., Mannini, A., Mugnai, G., De Filippi, P., Tarone, G., Del Bene, R., Barletta, E., Wanke, E., and Olivotto, M. (1993) Integrin-Mediated Neurite Outgrowth in Neuroblastoma Cells Depends on the Activation of Potassium Channels. The Journal of Cell Biology 122:1131-1143. Armstrong, D.L., Rossier, M.F., Shcherbatko, A.D., and White, R.E. (1991) Enzymatic Gating of Voltage-Activated Calcium Channels. Annals of the New York Academy of Sciences 635:26-34. Audesirk, G., Shugarts, D., Nelson, G., and PrZekwas, J. (1989) Organic and Inorganic Lead Inhibit. Neurite Growth in Vertebrate and Irivertebrate Neurons in Culture. In Vitro Cellular and Developmental Biology 25:1121-1128. Audesirk, G., Audesirk, T., Ferguson, C., Lomme, M., Shugarts, Rosack, P., Caracciolo, P., Gisi, T., and Nichols, P .. (1990) L-type Calcium Channels May Regulate Neurite Initiationin Cultured Chick Embryo Brain Neurons and N1E-115 Neuroblastoma Cells. Developmental Brain Research Audesirk, T., Audesirk, G., Ferguson, C., and Shugarts, D. (1991) Effects of Inorganic Lead on Differentiation and Growth of Cultured Hippocampal and Neuroblastoma Cells. Neurotoxicology 12:529-538. Banker, G.A. and Cowan, W.M. (1977) Rat Hippocampal Neurons in Dispersed Cell Culture. Brain Research 126:397-425. Bar, P.R., Renkema, G.H.,. Veraart; C.M., Hoi, E.M .. and Gispen, W.H. (1993) Nimodipine Protects Cultured Spinal Cord Neurones from Depolarization-induced Inhibition of Neurite Outgrowth. Cell Calcium 14:293-299. Bennett, J. and Weeds, A. (1986) Calcium and the Cytoskeleton. British Medical Joumal42(4):385-390. Bixby, J.L. and Jhabvala, P. (1990) Extracellular Matrix Molecules and Cell Adhesion Molecules Induce Neurites through Different Mechanisms. The Journal of Cell Biology 111:2725-2731. BOhle, Th. (1992) The Effect of the Benzothiazepine Diltiazem on Force and Ca2+ Current in Isolated Frog Skeletal Muscle Fibres. Journal of Physiology 445:303-318. 89

PAGE 105

Bray, Dennis. (1992) Cell Movements. Garland Publishing, Inc. New York and London. 1992. Brown, A.M., Yatani, A., Imoto, Y., Codina, I., Mattera, R., and Birnbaumer, L. (1989) Direct G-Protein Regulation of Ca2+ Channels. Annals of the New York Academy of Sciences 560:373-386. Cabell, L. and Audesirk, G. (1993) Effects of Selective Inhibition of Protein Kinase, and Calcium-Calmodulin-Dependent Protein Kinase on Neurite Development in Cultured Rat Hippocampal Neurons. International Journal of Developmental Neuroscience 10:357-368. Cannella, M.S., Ross, R.A. (1987) Influence of Substratum on the Retrograde Response of the Rat Superior Cervical Ganglion in vitro. Experimental Neurology 95:652-660. Caretta, A., Sorbi, R.T., Stein, P.J., Tirindelli, R. (1991) Diltiazem at High Concentration Increases the Ionic Permeability of Biological Membranes. Journal of Membrane Biology 122:203-213. Catterall, W.A. (1988) Structure and Function of Ion Channels. Science 242:50-61. Catterall, W.A., Seagar, M.J., Takahashi, M., and Nunold, K. (1989) Molecular Properties of Dihydropyridine-Sensitive Calcium Channels. Annals of the New York Academy of Sciences 560:1-12. Catterall, W.A. and Striessnig, .J., (1992) Receptor Sites for Ca2+ Channel Antagonists. Trends in Phannaceudical Sciences 13:256-262. Chang, C.F., Gutierrez, L.M., Mundina-Weilerimann, C., and Hosey, M.M. (1991) Dihydropyridine:.sensitive Calcium Channels from Skeletal Muscle IT. Functional Effects of Differential. Phosphorylation of Channel Subunits. The Journal of Biological Chemistry 266: 16395-16400. Cohen, C.J. and McCarthy, R.T. (1987) Nimodipine Block of Calcium Channels in Rat Anterior Pituitary Cells. Journal of Physiology 387:195-225. Collins, F. and Lile, J.D. (1989) The Role of Dihydropyridine-sensitive Voltage gated Calcium Channels in Potassium-mediated Neuronal Survival. Brain Research 502:99-108. 90

PAGE 106

Collins, F., Schmidt, M.F., Guthrie, P.B., and Kater, S.B. (1991) Sustained IIicreases in Intracellular Calcium Promotes Neuronal Survival. The Journal of Neuroscience 11:2582-2587. Daly, M.J., Perry, S., and Nayler, W.G. (1983) Calcium Antagonists and Calmodulin: Effect of Verapamil, Nifedipine, and Diltiazem. European Journal of Pharmacology 90:103-108. Deeg, H.J., Erickson, K., Storb, R., and Sullivan, K.M. (1992) Photoinactivation of Lymphohemopoietic Cells: Studies in Transfusion Medicine and Bone Marrow Transplantation. Blood Cells 18:151-162. De Jongh, K.S., Merrick, D.K. and Catterall, W.A. (1989) Subunits of Purified Calcium Channels: A 212-kDa Form of a-1 and Partial Amino Acid Sequence of a Phosphorylation Site of an Independent {j Subunit. Proceedings of the National Academy of Science 86:8585-8589. De Jongh, K.S., Warner, C., Colvin, A.A., and Catterall, W. (1991) Characterization of the Two Size Forms of the a-1 Subunit of Skeletal Muscle L-type Calcium Channels. Proceedings ofthe National Academy of Science 88:10778-10782. Diez-Guerra, F.J. and Avila, J. (1993) MAP2 Phosphorylation Parallels Dendrite Arborization in Hippocampal Neurones in Culture. NeuroReport 4:419-422. Doherty, P., Ashton, S.V., Moore, S.E., and Walsh, F.S. (1991) Morphoregulatory Activities of NCAM and N-Cadherin Can Be Accounted for by G Protein-dependent Activation of L-and N-type Neuronal Calcium Channels. Cell67:21-33. Doherty, P., Singh, A., Rimon, G., BolSover, S.R., and Walsh, F.S. (1993) Thy-1 Antibody-triggered Neurite Outgrowth Requires an Influx of Calcium into Neurons via N-and L-type Calcium channels. The Journal of Cell Biology 122:181-189. Dolphin, A.C. and Scott, R.ll. (1989) Modulation of Ca2+ Channel Currents in Sensory Neurons by Pertussis Toxin-Sensitive G-Proteins. Annals of the New York Academy of Sciences 560:387-390. Erne, P., Burgisser, E., Buhler, F.R., Dubach, B., Kuhnis, H.,. Meier, M., and Rogg, H. (1984) Enhancement of Calcium Influx in Human Platelets by CGP 28392, a Novel Dihydropyridine. Biochemical and Biophysical Research Communications 118(3):842-847. 91

PAGE 107

Fishman, R.B. and Hatten, M.E. (1993) Multiple receptor Systems Promote CNS Neural Migration. The Journal of Neuroscience 13(8):3485-3495. Forscher, P. (1989) Calcium and Polyphosphoinositide Control of Cytoskeletal Dynamics. Trends in Neuroscience 12(11):468-474. Fossheim, R., Svarteng, K., Mostad, A., Romming, C., Shefter, E., and Triggle, (1982) Crystal Structures and Pharmacological Activity of Calcium Channel 4-nitro-, and 2,4-dinitrophenyl)-1 ,4-dihydropyridine. Journal of Medicinal Chemistry 25:126-131. Franckowiak, G., Bechem; M., Schramm, M., and Thomas, G. (1985) The Optical Isomers of the 1,4-DihydropyridineBay K 8644 Show Opposite Effects on Ca2+ Channels. European Journal of Pharmacology 114:223-226. Friedrich, P. and Aszodi, A. (1991) MAP2: a Sensitive Cross-linker and Adjustable Spacer in Dendrite Architecture. FEBS 295(1,2,3):5-9. FujitaYamaguchi, Y., Kathuria, S.; Xu, Q.-Y.,.McDonald, J.M., Nakano, H., and Kamata, T. (1989) In Vitro 'l'yrosihe Phosphorylation Studies on RAS Proteins and Calmodulin Suggest That Polylysine-Like Basic Peptides or Domains May Be Involved in Interactions Between Insulin Receptor Kinase and Its Substrate. Proceedings of the National Academy of Sciences 86:7306-7310. Garyantes, T.K. and Regehr, W.G. (1992) Electrical Activity Increases Growth Cone Calcium but Fails to Inhibit Neurite Outgrowth from Rat Sympathetic Neurons. The Journal of Neuroscience 12:96-103. Glassman, H., Striessnig, J., Knaus, H.G., Muller, J., Grassegger, A., H.D., Marrer, S., Hymel, L., and Schindler, H.G. (1989) Structure of Calcium Channels. Annals of the New York Academy of Sciences 560: 198-214. Goshima, Y., Ohsak:o, S., and Yamauchi, T. (1993) Overexpression of Ca2+/Calmodulin-dependent Protein Kinase IT in Neuro2a and NG108-15 Neuroblastoma Cell Lines Promotes Neurite Outgrowth and Growth Cone Motility. The Journal of Neuroscience 13(2):559-567. Gutierrez, L.M., Brawley, R.M., and Hosey, M.M. (1991) Dihydropyridine-sensitive Calcium Channels from Skeletal Muscle. The Journal of Biological Chemistry 266:16387-16394. 92

PAGE 108

Hall, Z.W. (1992) An Introduction to Molecular Neurobiology. Sinauer Associates, Inc., Sunderland, Massachusetts. Hardy, R.W., McDonald, J.M., Remsen, E.E., d'Avignon, D.A., and Sacks, D.B. (1994) The Interaction of Calmodulin and Polylysine as Studied by 1H NMR Spectroscopy and Sedimentation Equilibriaum Centrifugation. Biochemical and Biophysical Research Communications 198(1):309-317. Hering, S., Kleppish, E.N., Timin, E.N.,Bodewei, R. (1989) Characterization of the Calcium Channel State Transitions Ind11ced by the Enantiomers of the 1,4dihydropyridine Sandoz 202791 in Neonatal Rat Heart Cells. Pjlugers Archives 414:690-700. Hering, S., Savchenko, A., Strobing, C., Lakitsch, M., Striessnig, J. (1993) Extracellular Localization of the Benzothiazepine Binding Domain of LType Ca2+ Channels. Molecular Pharmacology 43:820-826. Hescheler, J., Pelzer, D., Trube, G., and Trautwein, W; (1982) Does the Organic Calcium Channel Blocker D600 Act from Inside or Outside on the Cardiac Cell Membrane? Pjlugers Archives 393:287-291. Hof, D. and Fox, J.M. (1983) Changes of Ultraviolet Sensitivity of Voltage-clamped Sodium Channels During Their Potential-Induced Conductailce Cycle. Journal of Membrane Biology 71:31-37. Hoffmann, :F., Nastainczyk, W., Rohrkasten, A., Schneider, T., and Sieber, M. (1987) Regulation of the L-type Calcium Channel. Trends in Pharmaceudical Science 8:393-398. Hosey, M;M. and Lazdunski, M. (1988) Calcium Channels: Molecular Pharmacology, Structure and Regulation. Journal of Membrane Biology 104:81-105. Hymel, L., Striessnig, J., Glossman, H., and Schindler, H. (1988) Purified Skeletal Muscle 1 ,4-:dihydropyridine Receptor Forms Phosphorylation-dependent Oligomeric Calcium Channels in Planar Bilayers. Proceeding QS the National Academy of Science 85:4290-4294. Ikeda, S., Oka, J.I., and Nagao, T. (1991) Effects of Four Diltiazem Stereoisomers on Binding of d-cis-rH]diltiazem and ( + )-[3H]PN200-11 0 to RabbitT -tubule Calcium Channels. European Journal of Pharmacology -Molecular Pharmacolgy Section 208:199-205. 93

PAGE 109

Ishida, A.T., Bindokas, V.P., Nuccitelli, R. (1991) Calcium Ion Levels in Resting and Depolarized Goldfish Retinal Ganglion Cell Somata and Growth Cones. Journal of Neurophysiolgy 65:968-979. Izumi, S., Stojikovic, S.S., lida, T., Krsmanovic, L.Z., Omeljaniuk, R.I., Catt, K.J. (1990) Role of Voltage-sensitive Calcium Channels in [Ca2+]i and Secretory Responses to Activators of Protein Kinase C in Pituitary Gonadotophs. Biochemical and Biophysical Research Communications 170:359-367. Jacobson, M. (1991) Developmental Neurobiology. Plenum Press, New York and London. Kandel, E.R., Schwartz, J.H., and Jessen, T.M. (1991) Principles of Neural Science. Elsevier Press, New York, Amsterdam, London, and Tokyo. Kasai, H., Aosaki, T. (1989) Modulation of Ca-Channel Current by an Adenosine Analog Mediated by a GTP-binding Protem in Chick Sensory Neurons. Pfliigers Archives 414:145-149. Kater, S.B. and Mills, L.R. (1991) Regulation of Growth Cone Behavior by Calcium. The Journal of Neuroscience 11:891-899. Kay, R.E. and Bean, R.C. (1970) Effcts of Radiation on Artificial Lipid Membranes. Advances in Biological and Medicinial Physics. 13:235-253. Kern, M., Audesirk, T., and Audesirk, G. (1993) Effects oflnorganic Lead on the Differentiation and Growth of Cortical Neurons in Culture. Neurotoxicology 14(2/3):319-326. Kim, B.K., Mitsui, M., Karaki, H. (1992)The Long-term Inhibitory Effect of a Ca2+ Channel Blocker, Nisodipine, on Cytosoljc Ca2+ and Contraction in Vascular Smooth Muscle. European Journal of Pharmacology 223:157-162. Kirchhofer, D., Grzesiak, J., and Pierschbacher, M.D. (1991) Calcium as a Potential Physiological Regulator of Integrin-mediated Cell Adhesion. The Journal of Biological Chemistry 266:4471-4477. Kobayashi, T., Yamauchi, R., and Murata, S. (1992) Effect of Ca2+ Antagonists on High-K+ Evoked Increase in [Ca2+]i in Rat Cerebral Synaptosomes and Hippocampal Neurons. Japanese Journal of Pharmacology 58:417-425. 94

PAGE 110

Kostyuk, P.G. (1989) Diversity of Calcium Ion Channels in Cellular Membranes. Neuroscience 28:253-261. Langs, D.A., Kwon, Y.W., Strong, P.D., and Triggle, D.J. (1991) Molecular Level Model for the Agonist/Antagonist Selectivity of the 1,4-Dihydropyridine Calcium Channel Receptor. Journal of Computor-Aided Molecular Design 5:95-106. Langs, D.A. and Triggle, D.J. (1992) Structural Motifs for Ion Channels .in Membranes. In The Structure of Biological Membranes. Philip Yeagle, editor. CRC Press. Boca Raton, Ann Arbor, London. pp. 721-780._ Lein, P.J., Banker, G.A. and Higgins, D. (1992) Laminin Selectively Enhances Axonal Growth and Accelerates the Development of Polarity by Hippocampal Neurons in Culture. Developmental Brain Research 69:191-197. Lelong, I.H., Petegnief, V., and Rebel, G. (1992) Neuronal Cells Mature Faster on Polyethylenenimine Coated Plates Than on Polylysine Coated Plates. Journal of Neuroscience Research 32:562-568. Li, S.-L., Yan, Paz, I.B., and FujitaYamaguchi, Y. (1992) Huinan Insulin Receptor (j-Subunit Transmembrane/Cytoplasmic Domain Expressed in a Baclovirus Expression System: Purification, Characteri.zcition, and Polylysine Effects on the Protein Kinase Activity. Biochemistry 31:12455-12462. Liu, Y. and Storm, D.R. (1990) Regulation of Free Calmodulin Levels by Neuromodulin: Neuron Growth and Regeneration. Trends in Pharmaceudical Sciences 11: 107-111. Luna, E.J. and Hitt, A.L. (1992) Cytoskeleton-Plasma Membrane Interactions. Science 258:955-964. Mannhold, R., Steiner, R., Haas, W., and Kaufmann, R. (1978) Investigations on the Structure-activity Relationships of Verapamil. Naunyn-Schmiedeberg 's Archives of Pharmacology 302:217-226. Mason, R.P., Gonye, G.E., Chester, D.W., and Herbette, L.G. (1989) Partitioning and Location of Bay K 8644, 1 ,4-dihydropyridine Calcium Channel Agonist, in Model and Biological Membranes. Biophysical Journal 55:769-778. Mason, R.P., Rhodes, D.G., and Herbette, L.G. (1990) Reevaluation of Dissociation Constants for 1,4-Dihydropyridines (DHP) Based on Membrane Drug 95

PAGE 111

Concentration. Biophysical Journal 57:302a. Mattson, M.P. and Kater, S.B. (1987) Calcium Regulation of Neurite Elongation and Growth Cone Motility. The Journal of Neuroscience 7:4034-4043. Mattson, M.P. and Kater, S.B. (1988) Isolated Hippocampal Neurons in Cryopreserved Long-term Cultures: Development ofNeuroarchitecture and Sensitivity to NMDA. International Journal of Developmental Neuroscience. 6:439-452. Matus, A. (1987) Putting Together the Neuronal Cytoskeleton. Trends in Neuroscience 10:186-188. Meggio, F., Boldtree, B., Marin, 0., Marchior, F., Perich, J.W., lssinger, 0.-G., and Pinna, L.A. (1992) The Effect of Polylysine on Casein-kinase-2 Activity Is Influenced by Both the Structure and of the Protein/Peptide Substrates and the Subunit Composition of the Enzyme. European Journal ofBiochemistry 205:939945. Miller, M., Bower, E., Levitt, P., Li, D., and Chantler, P.O. (1992) Myosin II Distribution in Neurons is Consistent with a role in Growth Cone Motility but Not Synaptic Vesicle Neuron 8:25-44. Mitchison, T. and Kirschner, M; (1988) Cytoskeletal Dynamics and Nerve Growth. Neuron 1:761-772. Mittag, T.W., Yoshimura, N., Tormay, A. and Podos, S.M. (1993) Polylysine Stimulated Protein Phosphorylation in Rabbit Ciliary Processes: Casein Kinase Activities. Experimental Eye Research 56(2):143-150. Morrison, B.D., Feltz, S.M., and Pessin, J.E. (1989) Polylysine Specifically Activates the Insulin-dependent Insulin Receptor Protein Kinase. The Journal of Biological Chemistry 264(17):994-1001. Nakayama, H., Tak:i, M., Striessnig, J. Glossman, H., Catterall, W.A., and Kanaoka, Y. (1991) Identification of 1,4-Dihydropyridine Binding Regions Within the a1 Subunit of Skeletal Muscle Ca2+ Channels by Photoaffinity Labeling with Diazipine. Proceedings.ofthe National Academy of Science 88:9203-9207. Needham, L., Hellewell, P.G., Williams, T.J., and Gordon, J.L. (1988) Endothelial Functional Responses and Increased Vascular Permeability Induced by Polycations. Laboratory Investigation 59(4):538-548. 96

PAGE 112

Ng-Sikorski, I:, Anderson, R., Patarroyo, M., and Andersson, T. (1991) Calcium Signaling Capacity of the CD11b/CD/18 Integrin on Human Neurtrophils. Experimental Cell Research 195:504-508. Nowycky, M.C., Fox, A.P., and Tsien, R.W. (1985) Three Types of Neuronal Calcium Channel With Different Calcium Agoirlst Sensitivity. Nature 316:440-443. Nunold, K., Florio, V., and Catterall, W .A. (1989) Activation of Purified Calcium Channels by Stoichiometric Protein Phosphorylation. Procedings of the National Academy of Science 86:6816-6820. O'Callahan, C.M., Ptasienski, J., and Hosey, M.M. (1988) Phosphorylation of the 165-kDa Dihydropyridine/Phenylalkylamine Receptor from the Skeletal MusCle by Protein Kinase C. The Journal of Biological Chemistry 263:17342-17349. O'Dell, T.J. and Alger, B.E. (1991) Single Calcium Channels in Rat and Guinea-pig Hippocampal Neurons. Journal of Physiology 436:739-767. Ohtsuka, M., Yokota, M., Kodama, I., Yamada, and Shibata, S. (1989) New Generation Dihydropyridine Calcium Entry Blockers: In Search of Greater Selectivity For One Tissue Subtype. General Pharmacology. 20:539-556. Oxford, G.S., Pooler, J.P. (1975) Ultraviolet Photoalteiation of Ion Channels in Voltage-clamped Lobster Giant Axons. Journal of Membrane Biology 20:13-30. Priego, J., Gonzalez-Morales, M.A., Cillero, F.J., Villarroya, M., Sunkel, C., Casa Juana, M.F., Lopez, M.G., Artalejo, C.R., Pascual, R., and Garcia, A.G. (1993) European Journal of Pharmacology 243:25-34. Rampe, D. and Triggle, D.J. (1990) New Ligands for L-type Ca2+ Channels. Trends in Pharmaceudical Science 11: 112-115. Rasmussen, H. (1986) The Calcium Messenger System. New England Journal of Medicine 314(17):1094-1101. Rasmussen, H. (1986) The Calcium Messenger System. New England Journal of Medicine 314(18):1164-1170. Rawn, J.D. (1989) Biochemistry. Neil Patterson Publishers. Carolina Biological Supply Company, Burlington, North Carolina. 97

PAGE 113

Regan, R.F. and Choi, D.W. (1994) The Effect of NMDA, AMPA/Kainate, and Calcium Channel Antagonists on Traumatic Cortical Neuronal Injury in Culture. Brain Research 633:236-242. Rehder, V. and Kater, S.B. (1992) Regulation of Neuronal Growth Cone Filopodia by Intracellular Calcium. The Journal of Neuroscience 12:3175-3186. Reuter, H. (1983) Calcium Channel Modulation by Neurotransmitters, Enzymes and Drugs. Nature 301:569-574. Robson, S.J. and Burgoyne, R.D. (1989) L-type Calcium Channels in the Neurite Outgrowth from Rat Dorsal Root Ganglion Neurons in Culture. Neuroscience Letters 104:110-114. Rogers, M. and Hendry, I. (1990) Involvement of Dihydropyridine-Sensitive Calcium Channels in Nerve Growth Factor-dependent Neurite Outgrowth by Sympathetic Neurons. Journal of Neuroscience Research 26:447-454. Saffell, J., Walsh, F.S., and Doherty, P., (1992) Direct Activation of Second Messenger Pathways Mimics Cell Adhesion Molecule-dependent Neurite Outgrowth. The Journal of Cell Biology 118:663-670. Sarno, S., Marin, 0., Meggio, F., and Pinna, L. (1993) Polyamines as Negative Regulators of Casein Kinase-2: The Phosphorylation of Calmodulin Triggered by Polylysine and by the a[66-68] Peptide Is Prevented by Spermine. Biochemical and Biophysical Research Communications 194(1):83-90. Schaeffer, P., Lugnier, C., and Stoclet, J.-C. (1991) Interactions of Calmodulin Antagonists with Calcium Binding Sites. European Journal of Phannacology-Molecular Phannacology Section 206:325-332. Schuch, U., Lohse, M.J., and Schachner, M. (1989) Neural Cell Adhesion Molecules Influence Second Messenger Systems. Neuron 3:13-20. Schwarz, W. and Fox, J.M. (1977) ffitraviolet-induced Alterations of the Sodium Inactivation in Myelinated Nerve Fibers. Journal of Membrane Biology 36:297-310. Silver, R.A., Lamb, A.G., and.Bolsover, S.R. (1989) Elevated Cytosolic Calcium in the Growth Cone Inhibits Neurite Elongation in Neuroblastoma Cells: Correlation of Behavioral States with Cytosolic Calcium Concentration. The Journal of Neuroscience 9:4007-4020. 98

PAGE 114

Silver, R.H., Lamb, A. G., and Boldsover, S.R. (1990) Calcium Hotspots Caused by L-Channe1 Clustering Promote Morphological Changes in Neuronal Growth Cones. Nature 343:751-754. Snutch, T.P., Leonard, J.P., Gilbert, M.M., Lester, H.A., and Davidson, N. (1990) Rat Brain Expresses a Heterogeneous Family of Calcium Channels. Proceedings of the National Academy of Sciences 87:3391-3395. Spedding, M. and Kenny, B. (1992) Voltage,..dependent Calcium Channels: Structure and Drug-binding Sites. Biochemical Society Transactions 20:147-153. Spielberg, H., June, C.H., .. Blair, O.C., Nystrom-Rosander, C., Cereb, N., and Deeg, H.J. (1991) UV Irradiation of Lymphocytes Triggers an Increase in Intracellular Ca2+ and Prevents Lectin-stimulated Ca2+ Mobilization: Evidence for l1Vand Nifedipine-sensitive Ca2+ Channels. Experimental Hematology 19:742,.748. Starr, T.V.B., Prystay, W., and Snutch, T.P. (1991) Primary Structure of a Calcium Channel That Is Highly Expressed in. the Rat Cerebellum. Proceedings of the National Academy of Science 88:5621-5625. Striessnig, J., Glossman, H., and Catterall, W.A. (1990) Identification of a PhenylalkylamineBinding Region Within the a-1 Subunit ofSkeletal Muscle Calcium Channels. Proceedings of the National Academy of Science 87:9108-9112. Striessnig, J., Murphy, B.J., and Cattetall, W.A. (1991) Dihydropyndine Receptor of L-type Calcium Channels: Identification of Binding Domains for rHJ( + )-PN200110 and [3]Azidopine Within the a-1 Subunit. Proceedings of the National Academy of Science 88: Suarez-ISla, B.A., Pelto, D.J., Thompson, J.M., and Rapoport, S.I. (1984) Blockers of Calcium Permeability Inhibit Neurite Extension and Formation of Neuromuscular Synapses in Cell Culture. Developmental Brain Research 14:263270. Takahashi; K. and A,kaike, N. (1990) Nicergoline Inhibits T-type cal+ Channels in Rat Isolated Hippocampal CAl Pyramidal Neurones. British Journal of Pharmacology 100:705-710. Tanaka, T., Inoue, H., Date, T., Okamura, K., Aoe, K., Takeda, M., Kugita, H., Murata, S., Yamaguchi, T., Kikkawa, K., Nakajima, S., and Nagao, T. (1992) Synthesis of the Optically Active trans-Isomers of Diltiazem and Their 99

PAGE 115

Cardiovascvular Effects and Ca-Antagonistic Activity. Chem. Pharmacological Data. 40:1476-1480. Tanaka, T., Ohmura, T., and Hidaka, H. (1983) Calmodulin Antagonists' Binding Sites on Calmodulin Pharmacology 26:249-257. Triggle, D.J.,'Langs, D.A., Janis, R.A. (1989) Ca2+ Channel Ligands: Structure Function Relationships of the 1,4-Dihydropyridines. Medicinal Research Reviews 9(2):123-180. Figure 4 of Introduction reprinted by permission by John Wiley & Sons, Inc. Copyright 1984 by Academic Press. Triggle, D.J. and Rampe, D. (1989) 1,4-Dihydropyridine Activators and Antagonists: Structural and Functional Distinctions. Trends in Pharmaceudical. Science 10:507-511. Triggle, D.J., HaWthorn, M., Golpalakrishnan, M., Minarini, A., Avery, S., Rutledge, A., Bangalore, R., and Zheng, W. (1991) Synthetic Organic Ligands Active at Voltage-Gated Catcium Channels. Annals of 'he New York Academy of Sciences 635: 123-138. Tsien, R.W., Hess, P., McCleskey; E.W., and Rosenberg, R.L. (1987) Calcium Channels: Mechanism of Selectivity, Pemieation, and Block. Annual Review of Biophysics and Biophysical Chemistry Williams, J.S., Grupp, I.L., Vaghy, P.L., Dumont, L .. Schwartz, A., Yatani, A., Hamilton, S., and Brown, A.M. (1985) Profile of the Oppositely Acting Enantiomers of the Dihydropyridine 202-791 in Cardiac Preparations: Receptor Binding, Electrophysiological, and Pharmacological Studies. Biochemical and Biophysical Research Communications: 131(1): Yanagisawa, H., Fujimoto, K., Shimoji, Y., Kanazaki, T., Mizutari, K., Nishino, H., Shiga, H., and Koike, H. (1992) Synthesis and Antihypertensive Activity of 3-Acetoxy-2, 3-dihydro-5-[dimethyl( amino )ethyl]-2-( 4-methoxyphenyl)-1 ,5benzothiazepln-4(5H)-one (Diltiazem) Derivatives Having Substituents at the 8 Position. Chem. Pharm. Bulletin 40:2055-2061. Zhang, X.Q., Watanabe, Y., Ohnishi, M., Baba, T., and Shibuya, T. (1993) Comparative Studies on the Inhibitory Effects of Calcium Antagonists on Cytosolic Calcium Levels Increased by High-Potassium or Glutamate in Cultured Rat Cerebellar Granule Cells. Japanese Journal of Pharmacology 62:411-414. 100