In vitro studies on a putative molecular mechanism of action of valproic acid in treatment of bipolar disorder

MISSING IMAGE

Material Information

Title:
In vitro studies on a putative molecular mechanism of action of valproic acid in treatment of bipolar disorder
Physical Description:
xi, 139 leaves : ill. ; 29 cm.
Language:
English
Creator:
Watterson, Jeannette M, 1974-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Valproic Acid -- therapeutic use   ( mesh )
Valproic Acid -- pharmacology   ( mesh )
Bipolar Disorder -- drug therapy   ( mesh )
Hippocampus -- drug effects   ( mesh )
Department of Pharmacology and Therapeutics thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Pharmacology and Therapeutics -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1999.
Bibliography:
Bibliography: leaves 124-138.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Jeannette M. Watterson.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 020241073
oclc - 51639014
System ID:
AA00022190:00001

Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
    List of Figures
        Page vi
        Page vii
        Page viii
    Abbreviations and definitions
        Page ix
    Abstract
        Page x
        Page xi
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
    Chapter 2. Materials and methods
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
    Chapter 3. Effect of VPA on MARCKS and GAP-43
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
    Chapter 4. Role of PKC in mechanism of action of VPA
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
    Chapter 5. Effect of VPA on cell viability, growth and morphology
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
    Chapter 6. Conclusions and future directions
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
    References
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
    Biographical sketch
        Page 139
        Page 140
        Page 141
        Page 142
Full Text
















IN VITRO STUDIES ON A PUTATIVE MOLECULAR MECHANISM OF ACTION
OF VALPROIC ACID IN TREATMENT OF BIPOLAR DISORDER










BY


JEANNETTE M. WATTERSON


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1999














ACKNOWLEDGMENTS


I wish to thank Dr. Edwin Meyer first and foremost, for he graciously accepted me
"as is" into his lab, thereby saving me from a decision I did not want to make. I am

grateful that he always saw and helped me to see the positive, and opened me up to new

ideas that I never thought to examine. I am very appreciative of Dr. Robert Lenox's

ability to stimulate me to always look for more, as well as to give more, and to never be

satisfied with less than 110%. 1 also credit him with redefining my conception of
"intensity" and "drive." I am indebted to Dr. David Watson for training me with patience

from start to finish, and for giving me direction and confidence at times when I had little.

I am grateful to my committee members Drs. Stephen Baker, Jeffrey Harrison, and Colin

Sumners for their guidance and approval, which I needed in order to maintain focus and
motivation. I am also thankful for the steadfast support and friendly atmosphere provided

by the department secretaries, the helpful technicians, and the many peers I have worked
with over the past four years, without whose help I could never have reached this point.

Finally, I would like to thank my parents, who instilled in me the ambition to aim high and

the drive to do my best.















TABLE OF CONTENTS


page

A C K N O W LE D G M E N T S ........................................................................................ ii

L IST O F T A B L E S ................................................................................. v

L IS T O F F IG U R E S ........................................................................................ ........ vi

ABBREVIATIONS AND DEFINITIONS ............................................................. ix

A B S T R A C T ............................................................................. ......................... .. x

CHAPTERS

I INTRODUCTION ..................................................... ................... 1

V P A ............................................................................................. 2
Bipolar Disorder .................... ...... 4
Mechanism of Action of Lithium in Treatment of Bipolar Disorder .................... 7
PKC and Its Substrates MARCKS and GAP-43 .......................................... 9
Developmental Effects of VPA .............................................................. 11
S u m m ary ........................................................................................ 1 1
S p ecifi c A im s ............................... ......................... ....................... 13

2 MATERIALS AND METHODS ................................... 20

C h e m ic a ls ....................................................................................... 2 0
H N 33 C ell C ulture ................ .......................................................... 20
P C 12 C ell C ulture ....................................................................... . 2 1
N eurite Q uantitation ........................................................................ 22
Protein Quantitation by Western Blotting .................................................. 22
Ribonuclease (Rnase) Protection Assay (RPA) ............................................ 23
P K C In hib ito rs ................................................................................. 24
P K C A ctivity ........................................... .. ................... . .. 24
Statistical A nalyses ......................................................................... 25











3 EFFECT OF VPA ON MARCKS AND GAP-43 ............................................ 27

Intro d u ctio n .................................................................................... 2 7
R e su lts ..................................................................................... . 3 7
D iscu ssion ............................................................................... . 53

4 ROLE OF PKC IN MECHANISM OF ACTION OF VPA .................................. 62

In tro d u ctio n .................................................................................... 6 2
R e su lts .......................................................................................... 6 7
D iscu ssion ............................................................................... . .. 78

5 EFFECT OF VPA ON CELL VIABILITY, GROWTH AND MORPHOLOGY .... 88

In tro d u ctio n .................................................................................... 8 8
R e su lts .......................................................................................... 9 1
D isc u ssio n ...................................................................................... 9 8

6 CONCLUSIONS AND FUTURE DIRECTIONS ............................................... 113

R E F E R E N C E S ................................................................ . ........................... ....... 124

BIOGRAPHICAL SKETCH .................................................................................. 139















LIST OF TABLES


Table pae

2.1. Properties of HN33 hybrid and parental cells ............................................ 21

2.2. A ntibodies for W estern blots ............................................................. 26

3.1. Addition of carbachol fails to potentiate the VPA-induced
MARCKS down-regulation in HN33 cells..... ......................... 43

3.2. Inositol supplementation fails to attenuate the VPA-induced
MARCKS down-regulation in HN33 cells ............................................. 44

3.3. Drugs, sources, and exposure conditions for HN33 cells ..................... 49

4.1. Antisense oligonucleotide sequences .................................................... 73

5.1. VPA induces an increase in HN33 cell doubling time ................................. 91

5.2. VPA has no effect on cell viability ....................................................... 91

6.1. Effects of two mood-stabilizing drugs, VPA and lithium,
as observed in HN33 cells following chronic exposure:
A su m m ary ................................................................................ 12 3















LIST OF FIGURES


page

1.1. Molecular effects of VPA in brain ......................................................... 17

1.2. Model for lithium regulation of MARCKS ............................................. 18

3.1. Comparison of MARCKS expression in mouse brain vs.
HN33 cells: Representative Western blot ......................................... 37

3.2. Concentration-dependent down-regulation of MARCKS
protein in HN33 cells following I d VPA exposure
a. R epresentative W estern blot ......................................................... 38
b Q u antitatio n ........................................................................... 3 8

3.3. Concentration-dependent down-regulation of MARCKS
protein in HN33 cells following 3 d VPA exposure
a. R epresentative W estern blot ......................................................... 39
b Q u antitatio n ....................... ............................................. ...... 3 9

3.4. Time course of VPA-induced MARCKS down-regulation
in H N 3 3 cells ..... .......................... .............................................. 4 0

3.5. Recovery of MARCKS following VPA-induced
long-term down-regulation
a. Q u antitatio n 4........................................................................... 4 1
b. Representative Western blot .................................................. 41

3.6. Effect of carbachol on the VPA-induced MARCKS down-
regulation in HN33 cells: Representative Western blot .......................... 43

3.7. Effect of myo-inositol on the VPA-induced MARCKS down-
regulation in HN33 cells: Representative Western blot ............................. 44

3.8. Additive effect of combined VPA and lithium exposure
on MARCKS protein
a Q u an titatio n ............................................................................ 4 5
b. Representative W estern blot ....................................................... 45









3.9. Effects of other psychotropic agents and VPA metabolites
on MARCKS protein
a. Q uantitatio n ............................................................................ 4 7
b. MARCKS expression following exposure to
two VPA metabolites: Representative Western blot ............................ 48
c. Structures of VPA and two tested metabolites .................................... 48

3.10. Time course of VPA-induced MARCKS mRNA alterations .......................... 50

3.11. Comparison of GAP-43 expression in rat brain vs.
HN33 cells: Representative Western blot ............................................. 51

3.12. Effect of chronic VPA exposure on GAP-43 expression
in HN33 cells: Representative Western blot ........................................ 51

3.13. Concentration-dependent increase in GAP-43 expression
in HN33 cells following VPA exposure ............................................. 52

3.14. Time course of VPA-induced GAP-43 increase in HN33 cells .............. 52

4.1. Effect of VPA exposure on PKC activity in HN33 cells .............................. 67

4.2. Effect of VPA on soluble PKC-ct expression in HN33 cells
a. Acute VPA Exposure: Representative Western blot ............................. 68
b. Chronic VPA Exposure: Representative Western blot .......................... 68
c Q u antitatio n ........................................................................... 6 8

4.3. Effect of VPA on PKC-6 expression in HN33 cells
a. Acute VPA Exposure: Representative Western blot ............................. 69
b. Chronic VPA Exposure: Representative Western blot .......................... 69
c Q u antitatio n ........................................................................... 69

4.4. Effect of VPA on PKC-& expression in HN33 cells
a. Acute VPA Exposure: Representative Western blot ............................. 70
b. Chronic VPA Exposure: Representative Western blot .......................... 70
c. Q u antitatio n ............................................................................ 70

4.5. Effect of VPA on PKC-C expression in HN33 cells
a. Acute VPA Exposure: Representative Western blot ............................. 72
b. Chronic VPA Exposure: Representative Western blot .......................... 72
c Q u antitatio n ............................................................................ 72

4.6. PKC-c antisense oligonucleotide design ................................ 73

4.7. Effect of PKC-a antisense oligonucleotides on PKC-ct
expression in HN33 cells: Representative Western blot .......................... 75










4.8. Effect of PKC inhibitor LY333531 on VPA-induced
down-regulation of MARCKS
a Q u antitatio n .7.......................................... ............................... 7 6
b. Representative Western blot ................................. 76


4.9. Effect of PKC inhibitor LY333531 on VPA-induced
GAP-43 increase
a. Q u antitatio n ........................................................................... 7 7
b. R epresentative W estern blot ........................................................ 77


5.1. Time-dependent VPA-induced neurite outgrowth
in HN33 cells: Representative photomicrographs
a. U ntreated, control H N 33 cells ...................................................... 92
b. 3 d VPA-exposed HN33 cells ...................................................... 92
c. 14 d VPA-exposed HN33 cells ................................................... 92
d Q u antitatio n ........................................................................... 9 2

5.2. Effect of PKC inhibitor LY333531 on
VPA-induced alterations in HN33 cells
a. E ff ect on cell density ................................................................. 94
b. Effect on neurite outgrowth ....................................................... 94

5.3. Concentration-dependent VPA-induced neurite outgrowth
in P C 12 cells ............................................................. . ....... . 95

5.4. Effect of PKC inhibitor BIM on VPA-induced
alterations in PC 12 cells
a. Effect of 0.6 mM VPA on PC 12 cell number ................................... 96
b. Effect of 1.8 mM VPA on PC12 cell number ..................................... 96
c. Effect of 0.6 mM VPA on PC12 cell neurite outgrowth ........................ 96
d. Effect of 1.8 mM VPA on PC 12 cell neurite outgrowth ........................ 96

5.5. Neuroprotective potential of VPA
a. Effect of VPA on cell number in
NGF-differentiated PC 12 cells .................................................. 97
b. Effect of VPA on neurite outgrowth in
NGF-differentiated PC 12 cells .................................................. 97














ABBREVIATIONS AND DEFINITIONS

BIM bisindolylmaleimide, an inhibitor of PKC

CMP-PA cytidine monophosphate-phosphatidic acid; also CDP-DAG

CNS central nervous system

DAG diacylglycerol, a stimulator of PKC

FDA U.S. Food and Drug Administration

GABA gamma-aminobutyric acid, an [inhibitory] amino acid neurotransmitter

GAP-43 43 kD growth-associated protein

HN33 immortalized hippocampal cell line derived from fusion of primary mouse
hippocampal and human neuroblastoma cells

IMPase myo-inositol I -monophosphatase

IP inositol [mono]phosphate

LY333531 a PKC inhibitor designed by Eli Lilly and Company

MARCKS myristoylated alanine-rich C kinase substrate

MBP myelin basic protein

NGF nerve growth factor

PC 12 rat pheochromocytoma (adrenal tumor) cell line

PI phosphoinositide

PKC protein kinase C

PDBu phorbol 12,13-dibutyrate

RPA RNase protection assay

SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis

VPA valproic acid or valproate














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IN VITRO STUDIES ON A PUTATIVE MOLECULAR MECHANISM OF ACTION
OF VALPROIC ACID IN TREATMENT OF BIPOLAR DISORDER

By

Jeannette M. Watterson

December 1999

Chair: Edwin M. Meyer
Major Department: Pharmacology and Therapeutics

Valproic acid (VPA; trade name Depakote) is a broad-spectrum anticonvulsant

which has of late been utilized as a treatment for bipolar disorder and has proven

efficacious in treating a variety of both acute manic and chronic bipolar states. Lithium is

currently the only other FDA-approved drug treatment for this disorder, and although it

has traditionally been the first-line treatment of choice, there is a recognized need for

alternatives, due to the low patient compliance rate resulting from troublesome side

effects, narrow therapeutic index, and less than optimal patient response.

Very little is known about how either drug exerts its mood-stabilizing properties,

and it has been the goal of this project to understand better the actions of VPA from a

mechanistic standpoint.

Initial experiments were conducted utilizing an immortalized hippocampal cell line,

the HN33 cells. Results indicate that at clinically relevant target serum concentrations,

chronic, but not acute, VPA reduces expression of the myristoylated alanine-rich C kinase

substrate (MARCKS), concomitant with increased GAP-43 expression. These effects are

associated with altered activity and expression of protein kinase C (PKC), an important









effector in the CNS, and the effect of VPA on MARCKS appears to be dependent, at least

in part, upon the activation/down-regulation of this enzyme. Furthermore, the observed

effects are accompanied by alterations in cellular morphology and proliferation rate, which
implies that VPA plays a major role in modulating neuronal plasticity.
Additional studies utilizing PC 12 cells showed similar VPA-induced

differentiation, with no evidence for neuroprotection following nerve growth factor-
induced differentiation of cells. These studies bring us closer to elucidating the basis for

the unique therapeutic profile of VPA, which includes both antiepileptic and

antimanic/mood-stabilizing clinical applications, as well as unfortunate teratogenic effects

(such as spina bifida, anencephaly, meningomyelocele, and encephalocele), and may
provide further insight into the pathophysiology underlying bipolar disorder.















CHAPTER 1
INTRODUCTION


VPA is a broad-spectrum anticonvulsant shown in the last two decades to be

efficacious in the treatment of bipolar disorder (manic-depressive illness). These two

actions can be distinguished from each other by their time-courses, with anticonvulsant

activity being observed much more quickly, on a scale of hours to days, in contrast to the

days to weeks treatment required for mood-stabilizing activity. The precise mechanisms

by which VPA exerts either its antiepileptic or antimanic effects are still unclear. VPA has

been found to affect Na+ flux, intracellular Ca++, and brain gamma hydroxybutyric acid

(GABA) levels, these actions probably relate to the anticonvulsant efficacy of this drug.

The basis for its mood-stabilizing efficacy remains a mystery. In addition to its therapeutic

actions, numerous studies have documented the deleterious effects of this drug on the

developing nervous system, especially the embryonic neural tube. It is the goal of this
project to elucidate further the mechanisms of action of VPA in brain, mainly through

studies of its effects on intracellular signaling in one particular cell model, in an effort to

understand better the underlying mechanism for this therapeutic profile. The specific aims

of this project are to: (1) characterize the therapeutically relevant time- and concentration-

dependent effects of VPA on PKC (activity and isozyme expression) and its substrates
(MARCKS and GAP-43), using the hippocampally derived HN33 cells as a model system

for neuronal effects of VPA; and (2) characterize the therapeutically relevant effects of
VPA on the viability, growth, and morphology of both HN33 cells and PC12 cells. Based

on preliminary evidence and previous reports, it is hypothesized that VPA will have a

significant effect on the expression of PKC substrates MARCKS and GAP-43, that PKC









will be integrally involved in any such effects, and that cells will undergo an alteration in

cell growth and morphology indicative of a transition to a more differentiated state.


VPA


VPA is a short-chain carboxylic acid that was first discovered to have antiepileptic

effects in 1963, and has been utilized as an anticonvulsant in the United States for over 20

years (Meunier el al., 1963; Ramsay, 1984). It first became available in the U.S. in 1978

as Depakene, the acid form of the drug. Five years later, the enteric-coated form,
Depakote, was introduced (Penry and Dean, 1989). This particular formulation, the most

popular form in use today, is a stable coordination compound containing equal

proportions of valproic acid and the sodium valproate salt (McElroy el al., 1989). VPA is

striking in its structural difference from other known anticonvulsants, yet it is well known

for its efficacy in treating a wide range of seizure disorders, giving it a broader therapeutic
profile for epilepsy treatment than perhaps any other anticonvulsant yet studied. VPA is

FDA-approved for treatment of simple and complex absence seizures, as well as mixed

seizures which include the absence type. Additionally, VPA is effective in treating such

primary generalized epilepsies as tonic-clonic, clonic, tonic, and myoclonic seizures,
infantile spasms, and photosensitive epilepsy, as well as secondary generalized seizures

and some partial seizures, including complex partial epilepsy (McElroy et al., 1989).

A number of studies have investigated the putative mechanism(s) through which

VPA exerts its anticonvulsant effects. No one signaling effect is likely to be responsible

for the entire panorama of the anticonvulsant efficacy of this drug, and a number of
molecular events appear to contribute to the net effect of reducing neuronal excitability.

The ability of VPA to increase neurotransmission of the inhibitory amino acid GABA has

been well documented. VPA may accomplish this through a variety of mechanisms by

increasing GABA levels through decreased degradation and increased synthesis and









release, or by increasing post-synaptic responsiveness to GABA through effects on the

GABA-receptor complex or altered K' conductance, effectively dampening neuronal

excitability (Chapman et al., 1982; Rimmer and Richards, 1985; McElroy el al., 1989;

Penry and Dean, 1989; Joffe, 1993). Because anticonvulsant effects are often observed

before total brain GABA levels become elevated, the VPA-induced increase in GABA

levels alone may not account for the drug's anticonvulsant effect.

VPA, like other anticonvulsants, is known to block sustained high frequency

repetitive firing of cultured neurons, due to its effect of slowing recovery of the sodium

channel following inactivation (Albus and Williamson, 1998). At high concentrations,

VPA has been shown to alter membrane potassium conductance, and relatively low

concentrations of the drug hyperpolarize neurons, both effects suggestive of direct effects

on neuronal membranes (Porter and Meldrum, 1995). Other studies have provided

evidence for reduced excitatory neurotransmission by the amino acid aspartate (Chapman

el al., 1982). Still other reports have suggested that VPA works through inhibition of

kindling, a model of epilepsy which holds that epileptic activity predisposes the system to

increased frequency of seizures triggered by lower thresholds over time (Porter and

Meldrum, 1995; McElroy el al., 1989).

In summary, the various mechanisms proposed encompass three major net effects

of VPA: (1) increased GABAergic synaptic inhibition post-synaptically; (2) reduction in

the excitatory synaptic transmission ultimately responsible for epileptic kindling; and (3)

reduction in repetitive firing via a direct effect on voltage-sensitive ion channels (Albus

and Williamson, 1998). The broad spectrum of anticonvulsant activity of VPA is probably

related to more than one of the molecular mechanisms demonstrated thus far. However, it

is unclear whether such anticonvulsive effects may also account for the antimanic efficacy

of VPA, and because little is known of the mood-stabilizing mechanism of VPA, much
remains to be explained with regard to what kinds of intracellular changes might underlie

this drug's mood-stabilizing effects.









Bipolar Disorder


In recent years, VPA has emerged as an effective alternative to lithium in the

treatment of acute mania in patients suffering from bipolar disorder (Gerner & Stanton,

1992; McElroy et al., 1992; Bowden el al., 1994), and is also being used for long-term

prophylactic management of this disorder. Bipolar disorder, also known as manic-

depressive illness, is a debilitating psychiatric condition for which there is no known cause
and only a limited range of treatment options available. It is estimated to affect about 1%

of the population in this country (Bowden et al., 1994). Patients afflicted with this

affective disorder experience frequent, recurring episodes of extreme alterations in mood.

Whereas the patient may appear normal between cycles, the episodes of mania and

depression that characterize this disorder can leave the patient functionally incapacitated.
The depressed phase involves motor activity alterations, impaired concentration, impaired

social and occupational functioning, general helplessness, and may result in changes in

sleep, appetite, weight, and energy level (Post, 1989). The manic episodes entail such

disruptive symptoms as hyperactivity, explosive temper, impaired judgement, insomnia,

disorganized behavior, hypersexuality, and grandiosity, any combination of which may

lead to alienation from family and friends, indebtedness, job loss, and other major life

problems (Bowden et al., 1994). If untreated, this disorder has a tendency to worsen,

with increasing severity, duration, and frequency of episodes over time (Post, 1989).

There is a high rate of suicide in patients suffering from this disorder, estimated at about

15% in untreated cases (Bowden el al., 1994).

Lithium has for over 40 years been the first-line drug therapy for both acute and

prophylactic management of this disorder, but due to its narrow therapeutic index and
many undesirable side effects (often resulting in compliance problems), the need for other

viable treatment options has become a major concern (Post, 1989; Bowden, 1996).

Although numerous treatments have been used experimentally in the past, including









antipsychotic drugs, calcium channel blockers (such as verapamil), high doses of thyroxine
(T4), serotonin precursors (such as L-tryptophan), and even electroconvulsive therapy

(Joffe, 1993), results have not been consistently positive, and none is used extensively as a
"mainstream" treatment at the present time. For example, the use of antidepressants is

generally not considered because patients may be brought out of their depressive episodes

only to be immediately submerged into the manic phase; further, certain drugs may

increase the frequency of cycling of mood swings (Post, 1989). The anticonvulsants

carbamazepine, lamotrigine, and gabapentin, as well as other pharmacological agents

(clozapine, clonazepam, verapamil), have shown promise in treating bipolar disorder, but

very few studies have been conducted and FDA approval has yet to be obtained for this

clinical indication (Bowden, 1996; Joffe, 1993; Hollister, 1995). Because lithium is the
approved first-line treatment, and has been used and studied extensively in the past, this

project will be focused in part toward understanding the mechanism of action of bipolar

drugs, based on what is known and has been reported experimentally for lithium.

VPA is emerging as a viable alternative or combinational drug treatment for

bipolar disorder, and in fact, VPA may even have a wider spectrum of efficacy than

lithium. Although VPA seems to be less effective in treating depression than in treating

the manic phase of the disorder, there is accumulating evidence of its use in atypical

(dysphoric/mixed) mania, rapid cycling, and secondary manias, which have poorer
prognoses and are generally less responsive to lithium than the usual forms of the disorder

(Calabrese & Delucchi, 1990; Calabrese et al., 1992, 1993; Bowden el al., 1994; Joffe,

1993; Brown, 1989). VPA has also been used to treat various conditions which are

comorbid with bipolar disorder, such as migraines, panic disorder, bulimia, attention-

deficit hyperactivity disorder (ADHD), and other brain-related illnesses, and VPA appears

to be quite effective in treating lithium non-responders (Bowden, 1995; Joffe, 1993;
Brown, 1989). The therapeutic index of VPA is considerably wider than that of lithium,

thereby allowing for greater freedom in dosage adjustment, with less risk of toxicity.









Furthermore, the side effect profile of VPA is less intimidating than that of lithium.

Whereas the side effects of lithium may be quite severe and disturbing to patients (thyroid,
renal, and CNS effects, dermatologic and cardiovascular complications), VPA typically

exhibits a different, more benign profile of side effects which are generally more

acceptable to patients (gastrointestinal upset, weight gain, hair thinning, increased hepatic

enzyme levels), and many of its bothersome side effects are alleviated over time or with a

decrease in drug dosage (Bowden, 1995; Penry and Dean, 1989). The exception to this is

the idiosyncratic, unpredictable and irreversible hepatic toxicity induced by VPA in

predisposed patients on rare occasion. While this condition may be fatal if not detected in

time, it affects only a very minimal portion of the population the highest incidence
reported was a rate of 0.01%. The majority of reported cases occur within the first three

months of VPA therapy, and because signs common to hepatic failure are clearly present,

patients can be monitored regularly for the first six months of therapy so as to reduce the
risk of fatality and prevent any long-term damage. Despite this risk, the side effect profile

of VPA is still considered to be safer and less bothersome than that of lithium.

Nonetheless, the mechanisms mediating the therapeutic properties of VPA in the

treatment of bipolar disorder are not clearly understood. Whereas the therapeutic actions

of VPA have been reportedly associated with potentiation of GABAergic transmission and
other neurotransmitter effects (Chapman el al., 1982; Post el al., 1992, Loscher, 1993,

Petty, 1995), these effects generally occur at rather high concentrations and may relate

more to the anticonvulsant properties of VPA than to its mood stabilizing effects see

Figure 1. 1 (Waldmeier, 1987, Motohashi, 1990). Further, as is the case for the time

course of lithium, there is a delay of several days in the onset of clinical antimanic action

of VPA, although loading strategies have demonstrated antimanic effects at times as early

as 3 d (McElroy el al., 1996). Thus, any mechanism postulated for the therapeutic action

of these drugs must take this delayed onset into account (Manji el al., 1995). Because
VPA and lithium are currently the only two drugs approved by the FDA for treatment of









acute mania, comparisons of the actions of these agents and efforts to elucidate the
mechanism(s) underlying the mood-stabilizing activity of both drugs is important to

understanding the pathophysiology of this illness. Further, this approach may lead to the

identification of new or existing drugs which may be useful in this application.


Mechanism of Action of Lithium in Treatment of Bipolar Disorder


Accumulating evidence from our laboratory and others strongly implicates
receptor-mediated PI signaling in the mechanism of action of lithium in the brain (Berridge

et al., 1982; Godfrey, 1989; Lenox and Watson, 1994). Numerous studies have

established lithium as an uncompetitive inhibitor of the enzyme inositol monophosphatase

(IMPase), and the resulting accumulation of intermediates in the pathway which are

responsible for additional downstream effects, including intracellular Ca++ release and

PKC activation (see Figure 1.2; Post et al., 1992). It was first suggested by Berridge and

colleagues that depletion of free inositol, resulting from the accumulation of intermediates
in the PI signaling pathway, is a major consequence of long-term lithium exposure

(Berridge et al., 1982; Berridge, 1989). This phenomenon would have greatest

implication in certain areas of the brain which are relatively more inositol-limited by virtue

of the blood-brain barrier as well as less capable of producing inositol de novo. Further,

because the PI signaling cascade is linked to G protein-coupled muscarinic receptor (in,

m3, m5) activation, areas of the brain undergoing the highest rate of receptor activation

may predictably be preferentially affected by lithium. Indeed, studies in our laboratory

have provided evidence for an inositol-driven attenuation of the effects of lithium in HN33

cells, as well as potentiation by a muscarinic receptor agonist (carbachol) and
prevention/reversal by a muscarinic receptor antagonist (atropine) (Watson and Lenox,

1996). Following exposure of HN33 cells to therapeutic concentrations of lithium (1

mM), MARCKS protein expression is significantly down-regulated (Lenox et al., 1996;









Watson and Lenox, 1996). This MARCKS reduction only occurs in the relative absence

of inositol, for supplementation of inositol even at concentrations as low as 5 4tM prevents

the lithium-induced down-regulation of MARCKS (Watson and Lenox, 1996; Watson et

al., 1998; see also Table 3.2). Similarly, the concomitant exposure of HN33 cells to both

lithium and the muscarinic agonist carbachol results in an increase in the level of

MARCKS protein down-regulation observed, and exposure of cells to the muscarinic

antagonist atropine reverses or prevents the lithium-induced reduction in MARCKS

expression (Watson and Lenox, 1996; Watson el al., 1998; see also Table 3.1). These

findings have provided further support for the PI signaling hypothesis.

Studies of the molecular mechanism of action of lithium in brain have focused

largely on its modification of PI signaling, particularly the lithium-induced accumulation of

intermediates such as DAG and the resulting stimulation of PKC. Such a PI-mediated

mechanism may account for the effects of lithium on PKC-regulated events, including

protein down-regulation/degradation and transcriptional and translational events. In

contrast, little evidence has been found to support a similar PI mechanism for VPA.

Vadnal and Parthasarathy (1995) showed that VPA had no effect, either stimulatory or

inhibitory, on IMPase partially purified from bovine brain. Further, lithium and

carbamazepine (a less commonly used antimanic agent) exhibited opposite effects on

IMPase, suggesting that the enzyme is not a common site of action for these two mood-

stabilizing medications. Dixon and Hokin (1997) reported that in cerebral cortical slices,

VPA caused no accumulation of inositol monophosphates, inositol bisphosphates, or

inositol 1,3,4-trisphosphate, all of which accumulated following lithium exposure. Based

on these observations, the depletion of inositol is not universally applicable as the basis for

the antimanic action of mood-stabilizing drugs.

Alternatively, it has become apparent that the action of chronic lithium in the brain,

which follows as a result of its interaction with the PI signaling pathway, is mediated

through subsequent regulation of PKC and the down-stream posttranslational modification









of select protein substrates (Lenox, 1987; Manji and Lenox, 1994; Watson and Lenox,
1996). Further, the clinical effects of lithium may be mediated, at least in part, through

these actions. Although VPA apparently does not share the property of PI signaling

inhibition (Vadnal and Parthasarathy, 1995; Dixon and Hokin, 1997), the question remains

as to whether it affects PKC, as previous studies have suggested (Chen el al., 1994;

O'Brien and Regan, 1998). It is a goal of this project to address further the potential role

of PKC in the action of VPA in the brain, especially as such effects might relate to the

mood-stabilizing properties of this drug.


PKC and its Substrates MARCKS and GAP-43


PKC exists as one of at least II structurally-related isozymes in mammals, many of

which are Cam-activated and/or DAG-dependent, and have been implicated in numerous

cellular responses associated with regulation of signaling and long-term events including

ion channel and gene regulation (Newton, 1995; Nishizuka, 1995). Numerous studies

have suggested a role for PKC in the long-term action of lithium (Lenox, 1987; Manji and

Lenox, 1994), and recent data have provided evidence that chronic VPA exposure alters

PKC activity as well as the expression of a and c isozymes in non-neuronal cells (Chen el

al., 1994; Manji el al., 1996b). Previous studies have demonstrated that chronic (but not

acute) lithium treatment of rats, resulting in clinically relevant brain concentrations,
produces a significant reduction in the PKC substrate MARCKS in the hippocampus,

which persists beyond treatment discontinuation (Lenox el al., 1992; Manji et al., 1996a).

This lithium-induced down-regulation of MARCKS has also been demonstrated in an
immortalized hippocampal cell line (Watson and Lenox, 1996), in which it was previously

demonstrated that phorbol esters, which directly activate PKC, down-regulate MARCKS
protein expression in a PKC-dependent manner (Watson el al., 1994).









MARCKS is a prominent and preferential substrate in the brain for PKC, which by
virtue of phosphorylation, regulates the cellular localization and activity of this protein.

MARCKS binds calmodulin in a calcium-dependent fashion and cross-links filamentous

actin, and has been implicated in cellular processes associated with cytoskeletal
restructuring and neuroplasticity, e.g., transmembrane signaling and neurotransmitter

release (Aderem, 1992; Blackshear, 1993). In fact, previous studies in mutant mice

lacking the MARCKS gene have shown that MARCKS expression is essential to normal

CNS development in animals (Stumpo el al., 1995). Insofar as MARCKS may also

represent a molecular target for mood stabilizers in the brain (Lenox and Watson, 1994,

Watson and Lenox, 1996), it should be of benefit to examine its potential regulation by

VPA as well as by other psychotropic agents.
The growth-associated protein GAP-43 is a second protein thought to be integral
to normal neuronal development and differentiation, as it is expressed at high levels during

nervous system development and is associated with neurite extension (Jap Tjoen San el

al., 1992; Aigner and Caroni, 1993; Strittmatter el al., 1995). GAP-43 shares a number

of properties with MARCKS, including ability to bind actin and calmodulin (albeit in a

Ca++-independent fashion), as well as regulation by PKC (reviewed in Benowitz and

Routtenberg, 1997). Inasmuch as previous studies have provided evidence for an inverse

regulation of MARCKS and GAP-43 in certain regions of rat brain (McNamara and

Lenox, 1998), it may be instructive to assess the relative levels of GAP-43 observed

concomitant with VPA-induced alterations in MARCKS expression. Investigation of

these two related but differentially regulated proteins will likely provide insight into the
molecular mechanism of this drug, as will examination of this drug's effects on PKC, the

primary enzyme responsible for the direct regulation of both of these proteins.









Developmental Effects of VPA


In addition to the therapeutic efficacy of VPA, this drug exhibits its share of

drawbacks, the most severe of which are its effects on the developing embryo. VPA has

been known for years to be a developmental teratogen. In recent years, women who are
pregnant or even of childbearing age are encouraged to supplement their diets with

adequate amounts of folic acid, for supplementation with this nutrient and its metabolites

has been associated with decreased frequency of neural tube defects (including spina

bifida, anencephaly, and encephalocele) in offspring of VPA-treated and nontreated
mothers (Smithells et al., 1981; Rhoads and Mills, 1986; Trotz el al., 1987; Nosel and

Klein, 1992; Ehlers et al., 1996). Numerous studies have investigated this phenomenon,

yet it remains unknown how VPA causes these developmental defects or how folate

supplements are able to prevent such effects. Previous work has demonstrated that VPA

inhibits cell proliferation through arrest of the cell cycle at mid G I phase (Martin and

Regan, 1991), as well as induces differentiation of neuroblastoma and glioma cells (Regan,

1985), which would suggest that VPA is interfering with development through

manipulation of the proliferation rate and differentiation state of cells central to nervous

system structure. Such effects on cell growth and morphology should be considered in

any putative model for the molecular mechanism of this drug, and this study will therefore
investigate the hypothesis that VPA exerts effects on both cell proliferation and cell

morphology which may account for its teratogenicity.


Summary


Experiments designed to elucidate the short- vs. long-term intracellular actions of
VPA will facilitate the development of a model which may explain, in part, the unique

therapeutic profile exhibited by this drug. Initial VPA studies have been modeled after









those conducted previously or in parallel for lithium, so that comparisons may be made

which might shed light on the basis for the common mood-stabilizing therapeutic action of

these two drugs. In summary, thus far, it has been shown that in the HN33 cell model,
lithium elicits a down-regulation in MARCKS primarily through its effect on PI signaling,

as evidenced by the carbachol-driven potentiation and either inositol- or atropine-driven

attenuation of effect (Watson and Lenox, 1996; Watson el al., 1998). This lithium-
induced MARCKS reduction occurs over a chronic time course, requiring days to weeks,

and the effect is greater in the soluble rather than membrane fraction of the cells (Lenox et

al., 1996; Watson and Lenox, 1996). A similar time course of MARCKS mRNA

reduction is also observed (Watson and Lenox, 1997). Further, chronic lithium

administration elicits a significant down-regulation of MARCKS protein in vivo, in a rat

model (Lenox el al., 1992). As for the effects of lithium on cell viability and related

parameters, no alteration in cell growth rate or morphology has been observed following

acute or chronic lithium exposure (Watson and Lenox, 1996), and no alteration in
expression of another prominent PKC substrate, GAP-43, was observed (Watson and

Lenox, 1997). Lithium exposure results in predictable changes in PKC activity (with

initial activation followed by chronic down-regulation), as well as alterations in expression

of the two PKC isoforms c and 6 (Watson and Lenox, 1997), similar to findings

previously reported by Manji el al. (1 996a,b).

Applications of research into the mechanism of action of VPA will be multifold. In
addition to providing a better understanding for the efficacy of VPA in treating bipolar

disorder, a clearer picture of the pathophysiology underlying this illness, and identification

of new or existing drug treatments, investigation of the molecular bases for these effects
may foster a better understanding of how VPA exerts its deleterious developmental effects

and the development of potential therapeutic interventions to prevent or correct such

defects.










Specific Aims


Characterize the Therapeutically Relevant Effects of VPA on PKC and Its

Substrates.

MARCKS and GAP-43 are known to be important in modulating central nervous

system (CNS) plasticity and intracellular signaling, by virtue of the regulatory enzyme

PKC. Using hippocampally derived HN33 cells, I will characterize both acute and chronic

VPA-induced alterations of PKC expression and activity, as well as on levels of expression

of MARCKS and GAP-43.



Effects of VPA on MARCKS protein expression. Previous studies showed that

chronic lithium exposure at therapeutic concentrations resulted in down-regulation of

MARCKS protein expression in HN33 cells, and involved alterations in PI signaling.

After (1) establishing the concentration-response relationship for the effects of VPA on

MARCKS protein, I will (2) establish the time course of action of VPA at therapeutic

concentration, as well as (3) the time course of system recovery after VPA withdrawal. In

addition, I will assess the importance of (4) inositol concentration and (5) muscarinic

receptor/Gq activation, in VPA-induced MARCKS effects. Furthermore, I will (6)

investigate the effects of combined VPA and lithium treatment. Effects will be examined

in both cytosolic and membrane fractions of cells. It is predicted that VPA will have

effects similar to those of lithium, i.e. significant down-regulation of MARCKS protein

expression following chronic (but not acute) exposure to concentrations of the drug within

a therapeutically relevant range, and that MARCKS will return to basal levels after drug

removal. In contrast to lithium, there is no evidence for VPA-induced modulation of the









PI signaling pathway, so no effects by inositol or carbachol (muscarinic receptor agonist)

are expected.

Effects of VPA on MARCKS mRNA. Preliminary studies have provided

evidence for a lithium-induced reduction in MARCKS mRNA. I will establish the time

course of therapeutically relevant VPA-induced effects on MARCKS mRNA levels. It is

predicted that VPA will elicit down-regulation of MARCKS mRNA is a fashion similar to

that of lithium, but following the time course of VPA-induced alterations in MARCKS

protein.

Effects of VPA on GAP-43 expression. I will establish a dose-response and time

course for therapeutically relevant VPA-induced effects on GAP-43 protein levels in both

cytosolic and membrane cellular fractions. Although there are no previous reports for

VPA- or lithium-induced alterations in GAP-43 expression, such effects are a likelihood

given preliminary evidence for VPA-induced morphological alterations, and may partially

account for the plastic changes (such as synaptic remodeling) essential during therapy for

seizure and mood disorders.

Effects of VPA on PKC isozyme expression. I will establish the time course, at

therapeutic concentration, of VPA-induced alterations in four PKC isozymes expressed in

HN33 cells. Based on previous findings with lithium and VPA, which provide preliminary

evidence for effects of each on PKC expression and activity, I predict that VPA will have

significant effects on both PKC activity and isozyme expression, especially if alterations in

MARCKS and GAP-43 are observed following chronic VPA exposure.

Effects of VPA on PKC activity. I will investigate the effects of therapeutic

concentrations of VPA on PKC activity (as measured by phosphorylation of a PKC-

specific substrate) over a broad time course from acute to chronic time points. VPA-









induced alterations in PKC activity are expected for the same reasons as described above

in D.

Role of PKC in mechanism of action of VPA. Through the use of PKC

inhibitors and antisense, which block PKC activity, I plan to further characterize A-E

above; specifically, MARCKS and GAP-43 alterations, as well as morphological effects

(see below), will be investigated following PKC inhibition via enzyme inhibitors and

isozyme-specific antisense, in order to further define the role of PKC in these parameters.



Characterize the Physiologically Relevant Effects of Chronic VPA on Viability of

HN33 Cells and PC12 Cells.



Effects of VPA on HN33 cell growth. I will investigate the effects of chronic

VPA exposure at therapeutic concentrations on cell viability through observation of cell

number and doubling time. The role of PKC in any observed effects will also be assessed.

Based on limited previous studies in other cell lines, it is predicted that VPA will produce

changes in cell growth which may help to explain, in part, the severe developmental effects

associated with prenatal VPA exposure.

Effects of VPA on HN33 cell morphology. I will also investigate the effects of

chronic, therapeutic VPA exposure on cell phenotype, looking specifically for

differentiated morphology and other markers indicative of altered state of maturation. The

role of PKC in any observed effects on cell differentiation will also be assessed. Despite

the lack of evidence for a lithium-induced alteration in cell morphology, VPA-induced

alterations in cell morphology are considered likely, given this drug's therapeutic profile

(anticonvulsant and mood-stabilizing, yet teratogenic). Observed alterations in MARCKS









and GAP-43 will further support this hypothesis, given their integral role in CNS

development and plasticity.

Effects of VPA on PC12 cell viability. Preliminary evidence for effects of VPA

on cell growth and morphology leads me to investigate more in depth the effects of VPA

on cell viability. Whereas the HN33 cell line is a sufficient model for the investigation of

intracellular signaling phenomena as outlined in the studies above, post-mitotic cell lines

may serve as better developmental models for investigation of such phenomena as

neuroprotection or drug toxicity, and these cell lines have been better characterized for

such investigations. Therefore, in order to investigate the potential of VPA to elicit either

toxic or neuroprotective effects, NGF-differentiated and non-differentiated PC 12 cells will

be utilized. In light of the well-established neuronal toxicity of VPA exhibited during

development, toxicity resulting from high levels of the drug would not be a surprise;

however, preliminary studies in our laboratory have shown no evidence for toxic effects of

this drug at therapeutic concentrations. Any cytoprotective effect of VPA has yet to be

established, and there is currently no evidence in support of such an effect.











anticonvulsant effects


VPA

III \ \

glutamate +" !Ii \ \
giulm~~ica \ \ butyric acid
g/utamioacid G / /
decarboxylase GAB V+I '


VPA


PKC


GAP-43


MARCKS


mood-stabilizing effects


Figure 1.1. Molecular Effects of VPA in Brain. Depicted are (1) at right, the effects of
VPA as previously established by other groups, and thought to be responsible for the
anticonvulsant efficacy of the drug, and (2) at left, the long-term effects hypothesized to
contribute to the mood-stabilizing efficacy of VPA. Of the short-term effects at right,
increased GABA neurotransmission and effects on intracellular calcium are generally
thought to contribute less than the effects of VPA on voltage-sensitive ion channels and the
associated decrease in excitatory neurotransmission and neuronal firing. The hypothesized
effects at left will be assessed in detail in the studies which follow.











Figure 1.2. Model for Lithium Regulation of MARCKS. (adapted from Lenox et al., 1996). Following agonist
occupancy of the receptor binding site, G-protein coupling mediates activation of the enzyme phospholipase C (PLC),
which catalyzes the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) into second messengers diacylglycerol
(DAG) and myo-inositol 1,4,5-triphosphate (Ins 1,4,5 P3). DAG is an endogenous activator of PKC, and Ins 1,4,5 P3
induces release of calcium from intracellular stores such as the endoplasmic reticulum. Lithium inhibits uncompetitively
the enzyme inositol monophosphatase (IMPase), thereby inhibiting the recycling of inositol. This results in accumulation
of metabolites in the DAG pathway during stimulation of receptor signaling. Activation of PKC* through both increased
calcium and DAG initiates posttranslational events such as the translocation of PKC to the membrane where it
phosphorylates the MARCKS protein, which then translocates to the cytosolic fraction. In the phosphorylated form,
MARCKS no longer binds calmodulin or cross-links actin, thereby altering compartmental intracellular calcium signaling
and affecting cytoskeletal restructuring. In addition, chronic lithium and receptor activation down-regulate MARCKS
protein expression, which may result from down-stream transcriptional/translational events.











PK Actin

04
MP-PA CaM ,
In _ositol ransrpin/Translation MARCKS
nslP ns E~ffector(s) mRNA
n P Ins 1,4,5 P3In 3 53
[.I~ns 4PJ .
Ca+Ca+Ca+DNA














CHAPTER 2
MATERIALS AND METHODS


Chemicals


VPA (2-propyl-pentanoic acid) and other chemicals were purchased from Sigma

(St. Louis, MO), unless otherwise noted. For some drug comparison studies, additional

VPA and VPA metabolites were obtained from Abbott Laboratories (Abbott Park, IL),

and fluoxetine and LY333531 (a PKC inhibitor) were provided by Eli Lilly and Company

(Indianapolis, IN). Tissue culture reagents were purchased through Gibco (Gaithersburg,

MD; horse serum, RPMI and DME media), Hyclone (Logan, UT, fetal bovine serum), and

Sigma (trypsin and gentamicin). Electrophoresis reagents were purchased from Bio-Rad

(Hercules, CA), and RPA reagents were purchased from Ambion (Austin, TX).


HN33 Cell Culture


The immortalized hippocampal cell line HN33.dw(9/21) was used to assess effects

of VPA on all parameters, including MARCKS protein and mRNA, GAP-43, PKC activity

and expression, etc. HN33 cells were kindly provided by Dr. Bruce Wainer (Emory

University, Atlanta, GA). These are derived from the fusion of primary neurons from the
hippocampus of postnatal d 21 mice with the N 18TG2 human neuroblastoma cell line, and

exhibit morphological, immunological, and electrophysiological parameters characteristic
of hippocampal neurons in culture (Lee et al., 1990). These properties are described in
Table 2.1. HN33 cells have been used in our laboratory as a model system to address the









mechanism of action of lithium, and we previously established the presence of a functional

second messenger (PI) signaling pathway in these cells (Watson and Lenox, 1997).

Furthermore, HN33 cells have recently been assessed and found to express at least two of

the neurofilament triplet proteins (Watterson, unpublished data), which supports their

consideration as phenotypically neuronal (Debus el al., 1983; Shaw, 1991). Cells were
grown at 37'C and 5% CO2 in Dulbecco's modified Eagle's medium (DMEM) containing

5% defined fetal bovine serum (FBS) and 1 mg/ml gentamicin, with or without the

addition of 0.6 mM VPA. The duration of exposures ranged from 0 to 14 d. VPA was

added to cultures at time of plating, and culture medium was changed or cells were

passaged every 3-4 d, with no replenishment of VPA between platings. Assays were

performed when cells reached 80-90% confluence.


Table 2.1 lists the properties present (+) or absent (-) in HN33 cells and the two parent
cell lines from which the HN33 were derived, primary mouse hippocampal neurons (21 d)
and N18TG2 neuroblastoma cells. This table was adapted from that reported in Lee et al.
(1990), in which the HN33 cell derivation and characteristics were first reported.


Table 2.1. Properties of HN33 Hybrid and Parental Cells

Tissue or Hybrid Neuritic Excitable NGF NGF

Cell Line GPI Isozvme Processes Membrane NF-M Protein mRNA

Hippocampus + + + + +

N18TG2 -

HN33 + + + + + +


PCI2 Cell Culture


PC 12 cells obtained from American Type Culture Collection (Rockville, MD) were

grown in RPMI 1640 containing 10% heat-inactivated horse serum, 5% FBS, and 0.5 mM
l-glutamine. Cultures were maintained at 37C and 5% CO2, and passaged every 3 d. To








assess VPA-induced differentiation, cells were exposed to 0.3 1.8 mM VPA for 2 d, and

then neurite outgrowth was quantitated as described below. The cytoprotective assay was

carried out following differentiation of cells by 7-d exposure to 100 ng/ml NGF, and
subsequent NGF removal and replacement with 0.3, 0.6, and 1.2 mM VPA for 2 d.

Viviana Puig is acknowledged for her extensive assistance in carrying out the technical

aspects of these experiments utilizing the PC 12 cells.


Neurite Quantitation


HN33 cells were cultured in the presence or absence of 0.6 mM VPA for 3-14 d,

and process outgrowth was assessed using phase-contrast microscopy. A neurite was

scored if its length was at least one respective cell body diameter. Treatments were

compared for percentage of cells with one or more neurites and two or more neurites, and
150-900 cells were scored per treatment group. PC 12 cells were similarly assessed, using

the same criteria, with the aid of NIH Image 1.47 imaging software. Dan O'Donnell,

Jeremy Grimes, Risha King, and Keith Templin are acknowledged for their participation in

quantitation of neurite outgrowth in the PC 12 cells.


Protein Quantitation by Western Blotting


Harvested cells were homogenized in a buffer containing 20 mM HEPES, pH 7.4,

2 mM EGTA, 1 mM PMSF, 2 mM DTE, and 10 .g/ml aprotinin. Cells were sonicated to

disrupt cell membranes, and the soluble and pellet fractions were separated by
centrifugation. The homogenate was centrifuged at 100,000 x g and the soluble fraction

collected. The pellet was resuspended in buffer containing 0.1% Triton X-100 and

solubilized for 30 min. Solubilized fractions were then centrifuged at 50,000 x g and the

supernatant containing the solubilized membrane protein collected. Samples were









adjusted by addition or dilution to 0.05 % Triton X-100. Equal amounts of soluble and
membrane cell protein (50 4g), as determined by the Bradford method, were separated by

SDS-PAGE and transferred to polyvinylidene fluoride membrane (Millipore Corp.,

Bedford, MA) in a Bio-Rad Trans-Blot electrophoresis apparatus at 100 V for 2 h using

Towbin's buffer (25 mM Tris, pH 8.3, 192 mM glycine, and 20% (v/v) methanol). All

antibodies were diluted in TS buffer (20 mM Tris, pH 7.5, and 0.5 M NaCI); see Table 2.2
for dilutions and incubation times. Detection of the immune complex was performed

using HRP-conjugated anti-rabbit or anti-mouse IgG, and the Pierce (Rockford, IL)

enhanced chemiluminescence system. Western blots were quantified using NIH Image

1.47 software for densitometric analysis. Results are expressed as the percentage of

protein present in drug-free controls grown in parallel and assayed on the same Western

blot.


Ribonuclease (RNase) Protection Assay (RPA)


The RPA was performed with the RPAII kit using biotinylated antisense

riboprobes (BIOTINscript and BrightStar system, Ambion). Biotinylated antisense

riboprobes from the Macs (mouse gene coding for MARCKS) and the cyclophilin genes

were generated and gel-purified. A fragment of Macs was cloned by RT-PCR into a TA

cloning vector (InVitrogen, Carlsbad, CA) and restricted at a unique XmaI site. In vitro

transcription using SP6 RNA polymerase and this construct as the template yields a 466 nt

riboprobe and protects a 386 bp fragment of MARCKS mRNA. The cyclophilin probe

(Ambion) protects a 103 bp fragment. For each sample, 15 jLg of RNA was incubated

with 1 ng of each probe at 44C overnight. Non-hybridized RNA was digested with a
mixture of RNases. Protected fragments were separated on a denaturing 5% PAGE and

then transferred onto a nylon membrane to allow detection. Resulting autoradiographs
were then quantified and the amount of Macs expression was normalized to the cyclophilin









internal control. Sharlynn Sweeney is acknowledged for her expertise in development of

the probes, as well as her technical skill in carrying out the RPAs and quantitating results.


PKC Inhibitors


HN33 cells were exposed to the PKC inhibitor LY333531 at 1.0 M (Eli Lilly and

Company) concomitantly with 1 mM VPA for 24 h. Alterations in VPA-induced

MARCKS down-regulation and GAP-43 up-regulation, as well as cell density and

morphology, were measured and compared between control groups and LY33353 1-

treated groups, each in the presence or absence of VPA. PC 12 cells were exposed to the

PKC inhibitor bisindolylmaleimide (BIM; Gibco, Gaithersburg, MD) at 200 nM for 48 h,

and alterations in cell density and morphology were quantitated and compared between

treatment groups.


PKC Activity


PKC activity was measured by the phosphorylation of myelin basic protein (MBP)

4-14 peptide (Upstate Biotechnology, Lake Placid, NY) and MARCKS peptide. Soluble

and membrane cell fractions were prepared as described above. Ten microliters of the cell

protein (1 4g/41) were added to 100 gI1 of a reaction mixture containing 100 pM MBP, 100
W CaC12, 20 mM HEPES (pH 7.5), 0.03% Triton X100, 10 mM MgCI2, 100 pWM ATP,

and 0.0228 p.Ci/J [y-32P]ATP, with or without 100 4.g/ml phosphatidylserine and 20 4g/ml

PMA. The mixture was then incubated at 300 C for 2 min. The reaction was stopped by

the addition of 50 RI of ice-cold 450 mM phosphoric acid. Following a 5 min incubation
on ice, 30 p.1 of each sample was spotted onto Whatman P81 filter paper. Filters were
washed (x4) with a mixture of 150 mM H3PO4 and 10 mM sodium pyrophosphate, rinsed

in ethanol and then acetone, then dried and radioactivity measured. Specific activity for









PKC was measured as nanomoles of 32P/min/mg protein, and expressed as the difference

between activity in the presence and absence of PMA/phosphatidylserine. Dr. David

Watson is acknowledged for his contribution to the development of the PKC activity assay

and his assistance in carrying out the experiments.


Statistical Analyses


Following consultation with a biostatistician, statistical differences between

treatments were analyzed by one-way ANOVA, and post hoc comparisons were

conducted using Fisher's PLSD test. In all cases, minimal acceptable level for significance
was p<0.05. All results are reported as mean percentage of an untreated control group

assayed in parallel, and error bars represent S.E.M.. Drs. Ronald Marks, Robert

McNamara, and Mark Lewis are acknowledged for their assistance in identifying valid

statistical methods for analyzing these data.










Table 2.2. Antibodies for Western Blots. This table lists the primary antibodies utilized in western immunoblots to detect
each of the proteins of interest. For each antibody, the following information is reported: source from which antibody was
obtained, dilution (in tris-saline, or TS, buffer) at which membrane was exposed to antibody, length of exposure of
membrane to antibody, and secondary HRP-conjugated antibody used for amplification and detection of primary antibody,
its source, dilution (in TS buffer + 0.2% Tween-20), and exposure time. The anti-MARCKS polyclonal primary antibody
was raised against a glutathione-S-transferase MARCKS fusion protein designed in our laboratory, as detailed in Watson
and Lenox (1996).


10 ALk


C, ,


I nr.


9o Ah


,~, rr~


n.1 Inr"hnt;nn
I t-k ource, UL U11 "%.,U a U"


;Iulhti;rn


Inn Ih~tlnnl


MARCKS (custom made)

GAP-43 Boehringer-
Mannheim

PKC-ct Transduction Labs

PKC-6 Santa Cruz

PKC-e Santa Cruz

PKC-C Transduction Labs


1:36,000

1:3,000


1:1,000

1: 750

!1,000

1: 1,000


overnight


anti-rabbit Bio-Rad


overnight anti-mouse


anti-mouse

anti-rabbit

anti-rabbit

anti-mouse


Bio-Rad


Bio-Rad

Bio-Rad

Bio-Rad

Bio-Rad


1:24,000

1:20,000


1: 10,000

1:20,000

1:20,000

1:20,000














CHAPTER 3
EFFECTS OF VPA ON MARCKS AND GAP-43


Introduction


VPA has traditionally been used for its antiepileptic properties, which were first

discovered in 1963 by Meunier et al., and encompass a broad range of seizure disorders.

Of late, VPA has been recognized as a suitable alternative to lithium, the traditional first-
line drug, for both the acute and prophylactic treatment of bipolar disorder. VPA and

lithium are currently the only two FDA-approved anti-manic drugs. Although lithium has

been studied more extensively and will serve as the model upon which experiments in this

study are based, the mechanism of anti-manic and mood-stabilizing action of this drug

remains unclear, as does the etiology of the psychiatric disorder targeted by these drugs.

Therefore, investigations into the mechanism(s) underlying the mood-stabilizing activity of

both drugs is an important area of study.


MARCKS


Studies conducted in our laboratory, using both animal and cell models, have

provided evidence for a significant down-regulation of MARCKS following chronic but

not acute exposure to therapeutically relevant concentrations of lithium (Lenox el al,

1992; Watson and Lenox, 1996). In immortalized hippocampal cells, Watson and Lenox
(1996) first demonstrated that long-term (3-10 d) exposure to 1-10 mM LiCl resulted in a

reversible, time- and concentration-dependent down-regulation of MARCKS protein.









Similarly, in vivo studies demonstrated that exposure of rats to chronic lithium (3-4

weeks), resulting in clinically relevant serum concentrations of 0.7-1.2 mEq/kg in brain,

also reduced MARCKS in the hippocampus (Lenox el al., 1992). Additionally, recent

data from our laboratory suggest a time course of MARCKS mRNA regulation by lithium
in HN33 cells similar to the time course of MARCKS protein down-regulation (with
reductions evident by 3 d and persisting beyond 10 d), thereby indicating alterations in

either transcription or post-transcriptional message stability (Watson and Lenox, 1997).

MARCKS expression is widely distributed, with highest levels normally expressed
in brain, spleen, testis, and lung, and lower expression in heart, kidney, liver, and muscle

tissues of mouse (Lobach el al., 1993; Swierczynski el al., 1996). In the brain, the

greatest expression is in the most plastic regions, including the amygdala, striatum,

olfactory bulb and cortex, and septum (Ouimet el al., 1990; McNamara and Lenox, 1997),

and MARCKS is localized to small dendritic branches and axon terminals of neurons and

both cytoplasmic and membranous components of glia (Katz el al., 1985; Patel and

Kligman, 1987; Ouimet el al., 1990; Rosen el al., 1990). MARCKS cross-links actin and

binds calmodulin in a Cal-dependent manner (these events are mutually exclusive; Sheu

et al., 1995; Allen and Aderem, 1995), and as a result, has been implicated in cell motility,

secretion and trafficking, and cell cycle regulation (for reviews, see Aderem, 1992 and

Blackshear, 1993). As is indicated by its name, the activity and expression of MARCKS

are regulated in part through phosphorylation by PKC, though PKC-independent

pathways may also play a role in its post-transcriptional regulation (Brooks et al., 1992).

The inhibition of PKC or mutation of the PKC phosphorylation sites of MARCKS results
in defects in cellular motility (Allen and Aderem, 1995).
MARCKS is a protein for which the precise function is unclear but which is

believed to be essential to normal CNS development, since mice lacking detectable mRNA

and protein, due to gene (Macs) knockout, died perinatally, either before or within just

hours after birth (Stumpo el al., 1995). Developmental defects included defects in midline









events such as neurulation, fusion of the cerebral hemispheres, formation of the great
forebrain commissures, and retinal and cortical lamination (Stumpo el al., 1995).

Ultrastructural examination revealed widespread neuronal ectopia, or abnormal neuronal

pathfinding, during cerebrocortical development (Blackshear et al., 1996). Whereas these
findings do not preclude the possibility of more subtle ultrastructural or immunological

defects in other non-neuronal tissues in which MARCKS is expressed, phenotypic

alterations were not observed outside the CNS, and the data overwhelmingly support the

notion that the mortality is primarily a result of the high frequency of midline defects in

brain resulting from the MARCKS-deficient phenotype. Interestingly, this knockout
phenotype was "rescued" by the expression of a supplemental human MARCKS gene,

MACS, in that the human transgene contained all of the elements necessary for normal

developmental expression of MARCKS, and was able to fully complement the MARCKS-

deficient phenotype of the knockout mice (Swierczynski et al., 1996).

The purpose of the following experiments is to investigate the effects of VPA on

MARCKS expression, since significant effects were reported following chronic lithium

exposure, as noted previously, and any common action of these two mood-stabilizing

agents should be instructive as to their mechanisms of action. If MARCKS alterations are

integral to the mood-stabilizing properties of these and other drugs, one would

hypothesize that VPA should down-regulate MARCKS in a manner similar to that of

lithium. Both concentration-response and time-course curves of VPA-induced effects will

be established. These parameters address the therapeutic relevance of observed effects,
since each drug exhibits its own clinical profile, including such characteristics as

therapeutic serum concentration and latency to response. Previous studies have reported

that the effects of lithium on MARCKS protein expression occur following chronic but not

acute exposure to concentrations within the reported therapeutic serum concentration
range of this drug, and that the observed down-regulation of MARCKS expression is

reversible upon drug withdrawal. These findings are consistent with the known clinical









profile of lithium. Similar results are hypothesized following VPA exposure, since this
drug exhibits similar clinical properties of delayed onset of antimanic drug action and

delayed return to baseline following drug removal, over a therapeutic concentration range

(50-125 p.g/ml). The long-term effects of VPA will be the focus of these studies, for they

may be important to the mood-stabilizing efficacy of this drug, whereas effects observed

more acutely are generally credited to the anticonvulsant properties of VPA.

As mentioned in Chapter 1, accumulating evidence implicates receptor-mediated

PI signaling in the mechanism of action of lithium in the brain (Berridge el al., 1982;

Godfrey, 1989; Lenox and Watson, 1994). Lithium is an uncompetitive inhibitor of the

enzyme IMPase, and as such, it leads to accumulation of intermediates in the pathway

responsible for additional downstream effects, including intracellular Ca++ release and

DAG-mediated PKC activation (refer to Figure 1.2; Post el al., 1992). It is hypothesized

that depletion of free inositol, resulting from the accumulation of intermediates due to
incomplete cycling of the PI signaling pathway, is a major consequence of long-term

lithium exposure (Berridge ei al., 1982; Berridge, 1989). Such a phenomenon would have
its greatest impact on certain areas of the brain which are relatively more inositol-limited,

for example by virtue of the blood-brain barrier or areas less capable of producing inositol
de novo. Further, because the PI signaling cascade is linked to muscarinic receptor/Gq

protein activation, areas of the brain undergoing the highest rate of receptor activation

should be preferentially affected by lithium. Indeed, previous studies in our laboratory
lend support to these hypotheses (Watson and Lenox, 1996). Firstly, in studies of HN33

cells exposed to lithium, inositol supplementation (at concentrations as low as 5 M) was

able to prevent the down-regulation of MARCKS induced by 1 mM lithium, thereby

supporting the inositol-depletion hypothesis. Secondly, addition of the muscarinic
receptor agonist carbachol (1 mM) increased the MARCKS down-regulation observed

(Watson and Lenox, 1996), which supports the notion that the action of lithium is

augmented in cells or tissue regions undergoing the highest rate of receptor activation.









Whereas the mechanism of action of lithium has often been attributed to this

modulation of PI signaling and its downstream effects, there is little evidence in support of

a similar role for VPA; in fact, studies have indicated that VPA is probably not working

through effects on PI signaling. In one study comparing the effects of lithium,

carbamazepine, or VPA on IMPase activity, lithium and carbamazepine were found to

decrease and increase IMPase activity, respectively, and VPA was shown to have no

effect, at concentrations ranging from 2 j.M to 5 mM (Vadnal and Parthasarathy, 1995).

In another study, 2 mM valproate failed to increase accumulation of inositol phosphates,

including the inositol monophosphates, inositol 1,3-bisphosphate, inositol 1,4-

bisphosphate, and inositol 1,3,4-trisphosphate, whereas lithium exposure resulted in

accumulation of all (Dixon and Hokin, 1997). In a study conducted in our laboratory,

VPA failed to result in accumulation of CMP-PA (also known as CDP-DAG; see Figure

1.2), a metabolite of DAG and a sensitive measure of PI signaling, whereas lithium

resulted in significant accumulation of CMP-PA in CHO-KI and HN33 cells (Watson el

al., 1998). These data suggest that inhibitory effects on IMPase may not be the common

site of action for mood-stabilizing agents. However, this does not dismiss the possibility

that such agents may act indirectly or on alternate locations to alter the PI signaling

cascade. Experiments discussed herein have been modeled after those described above for
lithium, and will serve to clarify further the relationship between inositol or carbachol and

VPA-induced effects. Based on data from previous studies showing a lack of influence by

VPA on the PI signaling cascade, it is hypothesized that neither inositol nor carbachol will

alter any observed effect of VPA on MARCKS protein expression.

As mentioned previously, attention will be given to the possible therapeutic

relevance of findings. Inasmuch as lithium and VPA are commonly administered in

combination to patients who fail to respond adequately to either drug alone, often
resulting in amelioration or attenuation of manic symptoms, one study has been designed

to test the hypothesis that there may be additive actions at the cellular level. Following









optimization of exposure conditions for each drug, cells will be exposed to the

combination of VPA and lithium, and effects on MARCKS protein assessed. A final study

involves exposing cells to a series of other psychotropic agents individually, in order to

assess the specificity of action of mood-stabilizing agents and serve as a control for other

studies conducted. If MARCKS mediates any of the mood-stabilizing properties of these

agents, one would hypothesize that a variety of drugs efficacious in the treatment of
bipolar disorder would elicit the MARCKS down-regulation observed following lithium

exposure.

Experiments described thus far have simply focused on alterations in MARCKS

protein expression following exposure to drug and manipulation of various other

conditions in vitro. However, little is known of how these changes come about, or how
MARCKS levels are regulated in vivo. Regulation of MARCKS expression is thought to

occur in several ways: (1) tissue-specific expression, (2) developmentally regulated

expression, (3) differentially regulated transcription rates, and (4) alterations in mRNA

levels in response to PKC (Blackshear, 1993; Lobach et al., 1993). VPA-induced effects

on PKC itself will be discussed in detail in Chapter 5. Accumulating evidence shows that

lithium and VPA have regulatory effects on various transcription factors, thereby

suggesting a role for these mood stabilizers in altering gene expression. Studies in

transfected cultured cells have indicated that VPA enhances expression of genes and

reporters which contain AP-1 elements in their regulatory domains (Simon et al., 1994;

Kuntz-Simon el al., 1995). Manji and colleagues showed further that VPA exposure

resulted in a time- and concentration-dependent increase in DNA binding activity of the

transcription factor AP- 1 in both rat C6 glioma and human neuroblastoma (SH-SY5Y)

cells; lithium reportedly induced similar effects (Chen el al., 1997).
It is unknown how MARCKS expression is regulated beyond transcription, or how

the lithium-induced down-regulation of MARCKS protein occurs. Expression of many
proteins is known to be regulated not only by transcriptionally, but also by post-









transcriptional mechanisms, so that changes in mRNA stability play a crucial role in the

control of GAP-43 expression, for example. Similar post-transcriptional mechanisms
might also be at work in the case of MARCKS, and it is the purpose of these experiments

to assess further this possibility. Preliminary studies in our laboratory have indicated that

following lithium exposure, MARCKS mRNA levels do decline along a time course

similar to that of the protein down-regulation (Watson and Lenox, 1997), and one would

expect the same for VPA-induced alterations in MARCKS mRNA: a down-regulation of

mRNA levels over time, at therapeutically relevant concentrations of VPA, and with a

delayed onset of effect and protracted recovery of expression upon VPA withdrawal.


GAP-43


The growth-associated protein, GAP-43, is another major PKC substrate in brain

which shares many similarities with MARCKS. Like MARCKS, GAP-43 binds both

calmodulin and filamentous actin, albeit in a Ca+-independent manner (reviewed in

Benowitz and Routtenberg, 1997), and may play a role in exocytosis (Dekker et al.,

1989). GAP-43, like MARCKS, appears to be integral to normal nervous system

development, because studies of mice lacking the protein as a result of antisense-mediated

gene knockout show a high frequency of early postnatal death, with few animals surviving

past three weeks of age and even fewer reaching maturity (Strittmatter el al., 1995).
Those mice that do survive exhibit abnormal neuronal pathfinding.

GAP-43 is a rapidly transported, presynaptic protein most highly expressed during

development and post-injury regeneration, and immunohistochemistry and in situ

hybridization studies have shown GAP-43 expression to be almost exclusively neuronal

(Benowitz and Routtenberg, 1997). Neurite outgrowth is generally accompanied by a

large increase in GAP-43 expression, and following differentiation and synaptogenesis,

GAP-43 levels decline sharply and remain low, except in the most plastic regions of the









brain, such as associative and limbic areas, where intense immunostaining persists in adults

throughout life. This selective expression is thought to confer the unique potential for

ongoing structural remodeling (Benowitz and Perrone-Bizzozero, 1991).

GAP-43 is associated with the cytoskeleton and is concentrated in the growth

cones of elongating axons and nerve terminals, thought to help stabilize growth cone
morphology and promote growth cone adhesion (Aigner and Caroni, 1995). Support for

this notion has been provided by a number of studies. Antisense-directed depletion of

GAP-43 in growth cones results in the absence of spreading, branching, and adhesion of
growth cones (and therefore, neurite outgrowth) in primary sensory neurons (Aigner and

Caroni, 1993). Whereas suppression of GAP-43 expression has been found to yield

adverse effects on axon outgrowth both in vivo and in vitro (Meiri et al., 1998; Benowitz

and Routtenberg, 1997; Strittmatter et at., 1995; Jap Tjoen San et al., 1992; Shea et al.,

1991), overexpression of the protein induced nerve sprouting in the adult nervous system

of transgenic mice (Aigner et al., 1995). Even in non-neuronal cells, expression of a

GAP-43 transgene has been found to induce extensive process outgrowth accompanied by
reorganization of the membrane cytoskeleton (Strittmatter el al., 1994; Yankner e al.,

1990; Zuber et al., 1989), yet, a mutant PC 12 cell line which does not express GAP-43

still exhibits neurite outgrowth (Baetge and Hammang, 1991).

Although studies are in disagreement as to the precise role of GAP-43 in these

events, it has been hypothesized that the almost exclusive presence of GAP-43 in neuronal

cells, and the persistence of GAP-43 expression in highly plastic limbic regions, confers

the ability to undergo the structural remodeling necessary for the functional plasticity of

the brain (Benowitz and Perrone-Bizzozero, 1991). Indeed, Skene and Willard (1989)
suggest that it is the induction of GAP expression that is necessary to post-injury
remodeling, and that the inability of many areas of the CNS to induce high levels of GAP

expression may in part underlie the failure of CNS axons to regenerate after axonal injury,

whereas areas of the peripheral nervous system retain the ability to up-regulate GAP-43 as









well as to regenerate axons after injury. However, the presence of GAP-43 is probably

not necessary to neurite outgrowth per se, but rather, plays an important role in regulating
precisely the various parameters involved in the overall event, such as navigation or

pathfinding (Aigner and Caroni, 1995). According to this "GAP hypothesis", nerve
growth is controlled in part by the expression of certain growth associated proteins, such

as GAP-43, which provide a mechanism for regulating neurite outgrowth and structural
plasticity in the nervous system (Skene and Willard, 1989).

Because of the vital role of GAP-43 in effecting plastic changes in brain,

experiments have been designed to address the effect(s) of VPA exposure on GAP-43

expression in HN33 cells. It may be that part of the efficacy of VPA in either bipolar

disorder or epilepsy, or both, lies in its ability to elicit structural modifications of plastic
regions exhibiting the pathology which leads to the manifestation of the particular

disorder. Indeed, Benowitz and Routtenberg (1997) pointed out that in brain of humans
and other primates, the highest levels of GAP-43 expression occur in regions known to be

involved with higher-level associative processes, lending support to the notion that GAP-

43 is important for maintaining plasticity.

Whereas in vilro studies in HN33 cells have shown no evidence for regulation of

GAP-43 by lithium, neither has lithium been shown to alter the morphology of these cells.

Preliminary studies have suggested that chronic VPA exposure induces a morphological

alteration in these cells which might be further characterized as differentiation (see

Chapter 5). Since neurite outgrowth and neuronal differentiation are known to be

associated with up-regulation of GAP-43 protein expression, it will be of interest to

investigate the effect of VPA on this protein.

The results of these studies will be instructive, for in addition to the development

of a model which may help explain the role of MARCKS and other plasticity proteins

(such as GAP-43) in the mechanism of action of both VPA and lithium, findings will

contribute to a better understanding of the pathophysiology involved in bipolar disorder.









The notion of synaptic remodeling as the underlying means for the drug- or therapy-
induced stabilization of mood disorders has been supported by some reports. One study

found increased levels of neural cell adhesion molecule (N-CAM)-immunoreactive

proteins in the cerebrospinal fluid of a population of bipolar patients (Poltorak el al,
1996). Abnormal function or turnover of N-CAM, which plays a role in synaptogenesis,

morphogenesis, and plasticity of the nervous system, may be suggestive of abnormal

synaptic plasticity in affected individuals. With regard to drug therapy for this disorder,

not only do lithium and VPA affect expression of proteins involved in neural plasticity and

cellular response (as previously discussed), but both drugs have also been shown to have

effects on transcriptional factors which regulate gene expression, and it is known that

long-term changes in neuronal synaptic function are dependent upon the selective

regulation of gene expression (Manji and Lenox, 1994). Lamotrigine, another drug used

in treatment of bipolar disorder, was recently shown to alter neuronal hyperexcitability via

inhibition of both Na' and Ca- channels, findings which were interpreted by the authors to

signify modification of synaptic plasticity (Xie and Hagan, 1998). Such evidence is

indirect, but overall suggestive that changes in synaptic plasticity may be a mechanism by

which anti-bipolar (anti-manic and mood-stabilizing) therapy facilitates long-term changes

in affect.









Results













1 2 1 2


Figure 3.1. MARCKS protein. Shown are Western
immunoblots comparing MARCKS protein expression in
(1) mouse brain and (2) HN33 cells. A shows cytosolic
MARCKS expression, and B shows membrane-
associated MARCKS.


Comparative MARCKS expression in HN33 cells (used in this study) vs. mouse

brain is shown in Figure 3.1 following enhanced chemiluminescent (ECL) detection.

Differences in apparent molecular weight may result from differential phosphorylation or

myristoylation of the protein. Note the more pronounced MARCKS protein expression in

mouse brain, even with comparable amounts of total loaded protein.



Effect of VPA Exposure on MARCKS Protein in HN33 Cells

VPA was administered at concentrations ranging from 0.001 to 1.0 mM, for I or 3

d (Figures 3.2 and 3.3, respectively). Representative Western blots are presented, along

with quantitation of mean data. Significant concentration- and time-dependent effects

were observed. Chronic (3 d) VPA exposure results in a concentration-dependent

reduction in MARCKS levels, with a more pronounced reduction in membrane than






38


soluble fraction of these cells. Significant down-regulation of MARCKS is observed in the

membrane fraction at concentrations as low as 0.06 mM (p<0.005), and in the soluble

fraction at 0.1 mM and higher concentrations (p<0.005). These concentrations correlate

well with the target therapeutic serum concentration range used clinically for treatment of

bipolar disorder.


a 2U 3 45 6
1 2 3 4 5 6 7


100-

80-

60-

40-

20 -

A


0.01 0.03 0.06 0.1 0.3 0.6 1


IVPAI, mM


Figure 3.2. Concentration-dependent down-regulation of MARCKS protein in
HN33 cells following I d VPA exposure. (a) Representative Western blot of
membrane-associated MARCKS expression following I d VPA exposure; mean data
are depicted graphically in 3.2b. Lanes 1-7 were treated for I d as follows:
untreated control, 0.01 mrM, 0.03 mM, 0. 1 mM, 0.3 mM, 0.6 mM, and 1.0 mM
VPA, respectively. (b) This figure shows quantitatively the MARCKS protein levels
in HN33 cells following I d exposure to varying concentrations of VPA. Data,
expressed as percent of MARCKS compared to untreated control cells grown in
parallel, were derived from Western immunoblots (e.g., Figure 3.2a) and are the
mean of at least four determinations. Error bars represent S.E.M.


T T T] Soluble
T T-T TTT T E] Membrane
T I i -I- 1

T I TT
T















1 2 3 4 5 6 7 8 9 10 11 12


0.001 0.01 0.03


0.06 0.1 0.3 0.6 1


IVPAI, mM




Figure 3.3. Concentration-dependent down-regulation of MARCKS protein in HN33
cells following 3 d VPA exposure. (a) Representative DAB Western blot of MARCKS
expression following 3 d VPA exposure. Lanes 1-6 show soluble MARCKS following
3 d treatment with 0, 0.03, 0.06, 0.1, 0.3, and 0.6 mM VPA, respectively, and lanes 7-
12 show membrane MARCKS following exposure to the same doses. (b) This figure
shows quantitative data for mean MARCKS levels following 3 d VPA exposure. Data,
expressed as percent of MARCKS compared to untreated control cells grown in
parallel, were derived from Western immunoblots and are the mean of at least four
determinations. Error bars represent S.E.M. *p<0.005


140-

120-

100-

80-

60-

40-

20-


TQ Soluble
I T Q Membrane


T T T
IT

11 T






40


Time Course of VPA-Induced MARCKS Down-Regulation in HN33 Cells

HN33 cells were exposed continuously to 0.6 mM VPA for periods ranging from 0

to 7 d. This concentration is equivalent to 100 ng/ml, which is within the reported

therapeutic range for the mood stabilizing action of VPA (50-125 ng/ml). Cells were

observed to undergo a significant down-regulation of MARCKS protein expression in

both fractions as early as 3 d, with no effect on MARCKS at I d (see Figure 3.4). The

MARCKS reduction in the membrane fraction is more pronounced than that in the

cytosol. Three d, 5 d, and 7 d values are significantly different (p<0.01) from I d

MARCKS values in both soluble and membrane fractions of these cells. These data are

consistent with the clinical profile of this drug, which exhibits a delay in onset of antimanic

action at target therapeutic concentrations.



100-
1 T Soluble ] Membrane

II

60
TI
O 40- T
20-I
S20

1 3 5

Time (Days)



Figure 3.4. Time course of VPA-induced MARCKS down-regulation in
HN33 cells. This figure illustrates the time-dependency of VPA-induced
MARCKS protein down-regulation. HN33 cells were exposed to 0.6 mM
VPA, a clinically relevant therapeutic concentration, for 1, 3, 5, or 7 d, and
MARCKS expression was assessed by quantitative immunoblot analysis.
Results, expressed as percent of untreated controls, represent the mean
S.E.M. of at least four determinations. *p<0.01










a
140 SolubleT
130 Soluble Withdrawal
120 0 Membrane
110 -0 Membrane Withdrawal
U 100 T
90
80 0
70 T
70
S50*
40
30-
S20-
10
0-
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Time (days)


b


1 2 3 4 5


Figure 3.5. Recovery of MARCKS following VPA-induced long-term down-
regulation. (a) This figure shows the time course of VPA-induced MARCKS down-
regulation in HN33 cells, as well as the recovery of MARCKS protein to basal levels
following drug withdrawal. HN33 cells were exposed to 0.6 mM VPA for 3 d, at
which time the drug was either removed or treatment continued. MARCKS levels
were assessed through Western immunoblot (Figure 3.5b is representative), and results
are expressed as percent of untreated control cultures grown in parallel. Results
represent mean S.E.M. of at least three determinations. *p<0.0001 (b) Shown is a
representative Western blot of MARCKS expression in the membrane fraction of HN33
cells following chronic VPA exposure and subsequent withdrawal. Lanes I and 3 show
control levels of MARCKS protein. Lane 2 shows MARCKS expression following 3 d
VPA exposure (compare to Lane 1), and Lane 4 shows MARCKS expression
following 6 d VPA exposure (compare to Lane 3). Lane 5 represents a VPA
withdrawal/recovery group, and shows MARCKS in cells exposed continuously for 3 d
to 0.6 mM VPA, and then for 3 d to media free of drug.


Recovery of MARCKS following VPA-lnduced Long-Term Down-Regulation

As shown in Figure 3.5a, HN33 cells exposed to 0.6 mM VPA for 0 to 14 d

exhibited a significant down-regulation of MARCKS protein expression in both fractions









(membrane > soluble) over time, as previously demonstrated in Figure 3.4. After 3 d of

continuous VPA exposure and subsequent drug washout and withdrawal, MARCKS

protein levels recovered with significant increase (p<0.0001) to control levels within 3-7 d

of culture in drug-free medium (Figure 3.5a). A representative Western blot is shown in

Figure 3.5b. These findings are consistent with clinical evidence for delayed onset of anti-

manic action of VPA, as well as prolonged anti-manic efficacy following drug

discontinuation. It is difficult to discern whether this experimental return of MARCKS to

baseline levels is the result of either a shift in the kinetics of MARCKS

production/degradation, or due to repopulation of the il vitro environment with

MARCKS-replete daughter cells.


Effects of Carbachol and myo-Inositol on the VPA-Induced Down-Regulation of

MARCKS

In order to examine the role of PI signaling in the mechanism of action of this

drug's effects on MARCKS, cells were exposed to VPA alone or in combination with

either carbachol (1 mM) or rnyo-inositol (up to 1 mM). In contrast to previous results

obtained with LiCl (Watson and Lenox, 1996), the VPA-induced down-regulation of

MARCKS protein expression was not altered by addition of carbachol (Table 3.1 and

Figure 3.6) to the culture medium. Similarly, supplementation of myo-inositol (Table 3.2

and Figure 3.7) to the culture medium at concentrations up to 1 mM did not modify the

VPA-induced down-regulation of MARCKS, whereas the LiCI-induced down-regulation

was prevented by the addition of as little as 5.0 iM rnyo-inositol. Representative Western

blots are presented in Figures 3.6 and 3.7. It is difficult to rule out completely the role of

PI signaling based on the data in Table 3.1 alone, due to the maximal level of MARCKS

down-regulation achieved in the case of VPA. However, the data in Table 3.2 are more

conclusive, as the levels of MARCKS reduction elicited by VPA and lithium are more

comparable.










Table 3.1. Addition of Carbachol Fails to Potentiate the VPA-Induced MARCKS Down-
Regulation in HN33 Cells. HN33 cells were exposed to 0.6 mM VPA for 3 d or 1 mM
lithium for 7 d, in the presence or absence of the muscarinic agonist carbachol. Results
are presented as percent of untreated control cultures grown in parallel, and represent the
mean S.E.M. of at least three determinations. *p<0.0001


MARCKS Protein, % Control (Mean S.EM.)


VPA-Exposed


LiCI-Exposed


[Carbachol] Soluble Membrane Soluble Membrane
0mM 41.86.6 17.08.5 81.6+4.5 90.23.1
lmm 46.57.5 10.82.0 66.6+6.1"* 68.75.9**


ens


1 2 3 4


Figure 3.6. Effect of carbachol on VPA-induced MARCKS down-
regulation. Shown is a representative Western blot of MARCKS
expression following VPA exposure with or without the concomitant
addition of carbachol. Lane I shows membrane-associated MARCKS
levels in untreated control HN33 cells, Lane 2 shows MARCKS in 3 d
VPA-treated cells, Lane 3 shows MARCKS in carbachol-treated cells,
and Lane 4 shows MARCKS in cells treated concomitantly with VPA
and carbachol. Mean data are reported in Table 3.1.










Table 3.2. Inositol Supplementation Fails to Attenuate the VPA-Induced MARCKS
Down-Regulation in HN33 Cells. HN33 cells were exposed to 0.6 mM VPA for 3 d or I
mM lithium (+1 mM carbachol) for 7 d, in the presence (or absence) of varying
concentrations of myo-inositol. Results are presented as percent of untreated control
cultures grown in parallel, and represent the mean S.E.M. of at least three
determinations (N.D. = not determined). *p<0.005


MARCKS Protein, % Control (Mean S.E.M.)


VPA-Exposed


LiCl/carbachol-Exposed


[myo-Inositol] Soluble Membrane Soluble Membrane
0 83.0 + 6.0 57.0 5.0 66.6 6.1 68.7 5.9
0.5 IiM N.D. N.D. 71.1 4.8 68.44.4
5 4M N.D. N.D. 97.9 3.6* 104.3 5.3*
40 tM 77.8 2.4 54.8 5.5 98.5 7.5* 103.6 8.2*
1 mM 78.3 9.4 51.3 10.8 N.D. N.D.


1 2 3 4



Figure 3.7. Effect of inositol on VPA-induced MARCKS down-
regulation. Shown is a representative Western blot of MARCKS
expression following VPA exposure with or without concomitant
addition of mnyo-inositol. Lanes 1-4 show membrane-associated
MARCKS in (1) control cells, (2) 3 d 0.6 mM VPA-treated cells,
(3) inositol-treated cells, and (4) combined VPA- and inositol-
treated cells. Mean data are reported in Table 3.2.






45


go-4 100.
a 0o
T -a Soluble
0 80- 1 Membrane
T

TT
W0 40.





VPA LiCl/Carbachol VPA + LiCl/Carbachol
Treatment


b A B


1 2 3 4 5 6 7 I 2 3 4 5 6 7


Figure 3.8. Additive effect of combined VPA and lithium exposure on MARCKS
protein expression. (a) HN33 cells were exposed to 0.6 mM VPA and/or 1 mM
lithium (+ 1 mM carbachol) in inositol-free DMEM for 3 d. MARCKS protein
levels, expressed as percent of untreated control, were assessed by Western
immunoblot analysis (see Figure 3.8b), and values represent mean S.E.M. of at
least three determinations. *p<0.01 (b) Representative Western blot of MARCKS
protein following 3 d exposure of cells to VPA and/or lithium. A shows cytosolic
MARCKS in HN33 cells after the following treatments: no drug (1); 0.3 mM VPA
(2); 0.6 mM VPA (3); 1 mM lithium (4); 1 mM lithium + 1 mM carbachol (5); 1
mM lithium + 1 mM carbachol + 0.3 mM VPA (6); and 1 mM lithium + 1 mM
carbachol + 0.6 mM VPA (7). B shows MARCKS in the membrane fraction of
HN33 cells following the same treatments for 3 d.

Additive Effect of Combined VPA and LiCI Exposure on MARCKS Protein

Expression in HN33 Cells

HN33 cells were exposed to 0.6 mM VPA and/or 1 mM LiCIA mM carbachol for

a period of 3 d, in order to assess the effects of the VPA-lithium combination on

MARCKS protein expression and distribution. Results, presented in Figure 3.8a, show

that a greater reduction of MARCKS protein was observed in both soluble and membrane









fractions of cells exposed to both VPA and lithium/carbachol than was produced by either

VPA or lithium/carbachol alone. The enhanced reduction in MARCKS protein produced

by the combination of VPA and lithium/carbachol at these concentrations appeared to be

additive. Furthermore, the reduction in MARCKS protein produced by the combined

treatment of VPA and lithium/carbachol was greater in the membrane fraction than in the

soluble fraction. A representative Western blot is shown in Figure 3.8b.


Effects of Other Psychotropic Agents on MARCKS Protein Expression

Eight structurally and functionally distinct psychotropic drugs (carbamazepine,

fluoxetine, imipramine, phenytoin, diazepam, haloperidol, morphine, and verapamil) were

administered to HN33 cells for 1, 3, or 7 d, and their effects on MARCKS protein

expression were assessed. Drugs were chosen to represent a wide range of

psychopharmacological actions. Minimal drug concentrations were based on the

therapeutically effective plasma concentration for each drug in humans, as outlined in
Table 3.3. Additionally, cells were exposed to concentrations up to 20 times the highest

reported therapeutic concentration so as not to overlook potential effects at supra-

therapeutic doses. At each concentration tested, cell viability, growth rate, and gross

morphological appearance of HN33 cells exposed to each of the psychotropic drugs were

not notably different than those of cells grown in the absence of drug. Data obtained

following 3 d exposure are shown in Figure 3.9a. For all of the drugs listed, no significant

dose effect on MARCKS protein expression was measured at any of the concentrations or

time points tested.
MARCKS protein levels in HN33 cells were also measured after exposure to each

of two VPA analogues or metabolites, hydroxyvalproic acid (HVPA, or 2-n-propyl-5-
hydroxy-pentanoic acid) and 2-n-propylglutaric acid (2-PGA), for 3 or 7 d (Table 3.3;
Figure 3.9ab) These VPA metabolites, the structures of which are depicted in Figure

3.9c, were provided by Abbott Laboratories (Abbott Park, IL), and have been found to









have little or no anticonvulsant efficacy in vivo (Chapman el aL, 1982; J. Sullivan,

personal communication). Cells were exposed to concentrations of 0.6, 1.5, and 3.0 mM.

Unlike VPA, which produced noticeable changes in cell morphology at 0.6 mM (see

Chapter 6), these analogs produced no such change at this or higher concentrations.

Following HVPA exposure, no significant reduction in MARCKS protein was evident in

the soluble fraction, even at concentrations 5-10 times higher than the minimal effective

concentration of VPA. However, in the membrane fraction, a significant reduction in

MARCKS protein was observed only in cells exposed to 3.0 mM HVPA (p<0.01 at 3 d;

p<0.05 at 7 d). At this highest concentration tested (3.0 mM; 5X the therapeutic level of

VPA), a 30-35% reduction in MARCKS protein was observed following 3-7 d exposure.

Following 2-PGA exposure, no significant reduction in MARCKS protein was apparent in

the soluble fraction, even at concentrations 5-10 times higher than the minimal effective

concentration of VPA A small but insignificant reduction in MARCKS protein in the

membrane fraction was observed following 3.0 mM exposure to 2-PGA for 3-7 d. A

representative Western blot is presented in Figure 3.9b.

0% 140
a I 4 T E] Soluble Membrane
l' 120, TTT
a TT
.10 T I
00T
T100 T T T
0- T F
i 80- .
o 60-

En 40-

1 20-


x Pk


Treatment










A B
X. ...iiiii~ iiiiiii~ iii i .. ...... . . . . . ..




1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9


C
HOOC

>OH COO OC


Hydroxyvalproic Acid *2-Propyl Glutaric Acid




COOH

*VaIproic Acid


Figure 3.9. Effects of other psychotropic agents on MARCKS protein expression.
(a) This figure shows the relative lack of effect of various other psychotropic agents
on MARCKS protein expression in HN33 cells. Cells were exposed to drugs other
than VPA or lithium (see Table 3.3 for details), as follows: 3.0 mM hydroxyvalproic
acid, 7 d; 3.0 mM 2-propylglutaric acid, 7 d; 100 p.M carbamazepine, 7 d; I pgg/ml
fluoxetine, 3 d; I pg/ml imipramine, 7 d; 20 pg/ml phenytoin, 3 d; I jig/ml
diazepam, 7 d; 20 ng/ml haloperidol, 7 d; 20 jtg/ml morphine, 7 d; 50 ng/ml
verapamil, 7 d. MARCKS levels were assessed as previously described; data
represent mean S.E.M. of at least three determinations. *p<0.05 (b) MARCKS
expression following exposure to VPA and two metabolites. Shown is a
representative Western blot of MARCKS protein in HN33 cells after 7 d exposure to
the following: no drug (1); 0.6 mM VPA purchased from Sigma Chemical Company
(2); 0.6 mM VPA provided by Abbott Laboratories (for comparison purposes; 3); 0.6
mM (4), 1.5 mM (5), and 3.0 mM (6) 2-propylglutaric acid; 0.6 mM (7), 1.5 mM (8),
and 3.0 mM (9) hydroxyvalproic acid. A shows cytosolic MARCKS expression, and
B shows membrane-associated MARCKS. (c) Structural comparisons of VPA and
metabolites. Shown are the chemical structures of VPA and two of its metabolites,
hydroxyvalproic acid, or 2-n-propyl-5-hydroxy-pentanoic acid, and 2-n-propylglutaric
acid.









Table 3.3. Drugs, Sources, and Exposure Conditions for HN33 Cells.
This table lists the drugs to which HN33 cells were exposed, their sources, target
circulating (plasma) concentrations in humans, concentrations tested in HN33 cells, and
vehicle in which drug was suspended. Drug concentrations tested encompassed a broad
range, including supratherapeutic concentrations, in order to account for discrepancies,
such as those presented by variation in circulating serum vs. brain concentrations of drug.


Drug / Source
Sodium Valproate
Sigma #P-4543
Abbott #A-44089.5


Target Plasma
Conc. in Humans
50-125 ig/ml
= 0.3-0.9 mM
(seizure control )ab


Concentrations Tested
0.01, 0.03, 0.06,
0.1, 0.3, 0.6, 1.0 mM


2-Propylglutaric Acid
Abbott #A-49999.0


Hydroxyvalproate, sodium salt
Abbott #A-49822.5


Carbamazepine
Sigma #C-4024


Fluoxetine-HCl
Lilly #110140L/F08083


Imipramine-HCI
Sigma #1-7379

Phenytoin, sodium salt
Sigma #D-0931

Diazepam
Sigma #D-0899



Haloperidol
Sigma #H- 1512

Morphine Sulfate
Sigma #M-8777

Verapamil-HCI
Sigma #V-4629

aHolford and Benet,
Meldrum, 1995.


8-12 g/ml
= 33-50 4M
(seizure control)b,c


varies


100-300 ng/ml
(antidepressant)a

10-20 .g/ml
(anticonvulsant)d,a

300-400 ng/ml
(anxiolytic)ab
>600 ng/ml
(seizure control)b

4-20 ng/ml
(anti-psychotic)b


5-20 ng/ml


0.6, 1.5, 3.0 mM


0.6, 1.5, 3.0 mM


10, 25, 100 M


50, 200, 1000 ng/ml


50, 200, 1000 ng/ml


1, 10, 20 gtg/ml


50, 200, 1000 ng/ml




5, 20, 100 ng/ml


5, 20, 100 ng/ml


(analgesic in non-tolerant patients)c


varies


50 ng/ml


1995; hBenet and Williams, 1990; cWeiner, 1996; dporter and


Vehicle
dH20


dH20


dH20


ETOH


dH20


dH20


ETOH


ETOH




dH20


dH20


ETOH






50


Effect of VPA on MARCKS mRNA

HN33 cells were exposed to 0.6 mM VPA for 0 to 14 d, and MARCKS mRNA

levels were measured by RPA as described in Materials and Methods. As shown in Figure

3.1 Oa, MARCKS mRNA levels decrease over time with continual VPA exposure

(p<0.05). A representative RPA is shown in Figure 3.10b.


100 -


1 3 5


7 10 14


Time (d)


40 40


:00i


MARCKS



Scyclophilin


$.db
Oo -O A- ....


A B C D E


F G H I J


Figure 3.10. Time course of VPA-induced MARCKS mRNA alterations. (a) This figure
shows the time course of VPA-induced alterations in MARCKS mRNA. HN33 cells
were exposed to 0.6 mM VPA for 0-14 d, and MARCKS mRNA levels were assessed by
RPA, with the aid of a cyclophilin internal control, as described in Materials and
Methods. Results for each treatment are calculated as a ratio of MARCKS to
cyclophilin, and then each value is expressed as a percentage ( S.E.M.) of untreated,
control samples prepared in parallel. *p<0.05 Shown in (b) is a representative RPA.
Upper band reflects MARCKS mRNA, and lower band reflects cyclophilin mRNA.
Treatments are as follows: A, control; B, I d VPA; C, 3 d VPA; D, control; E, 5 d
VPA; F, 7 d VPA; G, control; H, 10 d VPA; I, control; J, 14 d VPA.











Effect of VPA on GAP-43 Expression in HN33 Cells

GAP-43 expression in HN33 cells is compared to its expression in rat brain in

Figure 3.11. Following 3 d, 0.6 mM VPA exposure, GAP-43 levels in HN33 cells

increase, as shown in Figure 3.12. This increase is concentration-dependent, as shown in

Figure 3.13 (p
from 0.01 mM to 1.0 mM, and membrane-associated GAP-43 expression is shown (Figure

3.13).


A B



1 2 1


Figure 3.11. GAP-43. Shown is a Western
immunoblot comparing GAP-43 expression in (1)
soluble and (2) membrane fractions of(A) HN33
cells and (B) rat brain.








1 2


Figure 3.12. Effect of chronic VPA exposure on GAP-43 expression
in HN33 cells. Shown is a representative Western blot of GAP-43
expression in the membrane fraction of HN33 cells following VPA
exposure. Lane 1 shows GAP-43 expression in untreated control
cells, and Lane 2 shows GAP-43 in cells treated for 3 d with 0.6 mM
VPA.












.tVu -

300 -

200 -

100 -

0-


0.01 0.03 0.06 0.1 0.3 0.6


[VPA], mM

Figure 3.13. Concentration-dependent increase in GAP-43 expression in HN33
cells following VPA exposure. HN33 cells were exposed for 3 d to varying
concentrations of VPA, and membrane-associated GAP-43 expression was
assessed by Western immunoblot analysis. Results represent the mean of two
or three determinations, and are expressed as the percent of GAP-43 protein
expressed in untreated control cultures grown in parallel. *p

1 3 5 14
Time (d)

Figure 3.14. Time course of VPA-induced GAP-43 increase in HN33 cells. HN33
cells were exposed to 0.6 mM VPA for times ranging from I d to 14 d, and GAP-
43 levels were assessed by Western immunoblot analysis. Results reflect the mean
of at least three determinations, and are expressed as percent of GAP-43 in an
untreated control. *p








Time Course of VPA-Induced GAP-43 Increase in HN33 Cells

HN33 cells were exposed to 0.6 mM VPA for 0-14 d; results are shown in Figure

3.14. Up-regulation of GAP-43 was observed in membrane and soluble fractions of cells

exposed to VPA for 5 d and 14 d, respectively (p<0.05). Because of the experimental

protocol, which entailed growing cells to confluency and then diluting them every 3-4 d,

the data may not reflect the full potential for GAP-43 induction, since each splitting and

re-seeding procedure involved trypsinization and potential disruption of cellular processes.




Discussion


HN33 cells were utilized as the model for all studies, because their morphological,

immunological, and electrophysiological characteristics resemble those of hippocampal

neurons in culture. HN33 cells were chosen as the in vitro cell model in which to

investigate multiple steps in the PKC/GAP-43/MARCKS pathway, because of their

relative homogeneity and the presence of signal transduction pathways of interest. This is

in contrast to the use of other models, such as primary neuronal cultures, which are more

neurochemically heterogeneous in nature. In primary cultures, for example, it would be

difficult to assess whether PKC and other messenger systems were affected in the same

neuron. The homogeneity of HN33 cells makes signaling research and results less

confounding.


MARCKS


VPA exposure elicits both a time- and concentration-dependent reduction in

MARCKS protein expression. It has been well established that the clinical antimanic

response to VPA requires a lag period of days to weeks following drug administration









before onset of action, although loading strategies have proven useful in achieving faster

response (within 3 d) (Bowden el al., 1994; McElroy el al., 1989). Consistent with such
clinical observations, a significant reduction of MARCKS protein in both soluble and

membrane fractions of HN33 cells was not observed following acute (1 d) VPA

administration, except at levels above the maximal target therapeutic serum concentration
(1 mM). At 0.6 mM VPA, the down-regulation of MARCKS was greatest between d 0

and d 3, and little further reduction was observed upon continued exposure (5-10 d).

With regard to therapeutic concentrations, there have been reports of clinical response at
plasma concentrations of VPA as low as 45 g/ml (Bowden et al., 1994), which is

comparable to the observed effects of VPA on MARCKS down-regulation at

concentrations as low as 0.1 mM in this study. However, VPA is generally thought to be

most effective at therapeutic concentrations of 50-125 g/ml (0.3-0.75 mM), where the
most robust effect of VPA on MARCKS was observed in the HN33 cells. Following

VPA-induced MARCKS down-regulation and subsequent drug withdrawal, MARCKS

levels in these cells recovered to baseline within 3-7 d. This finding is consistent with the

third major parameter discussed with regard to clinical relevance, that of long-term effect

of VPA and delayed return of clinical response to baseline following treatment

discontinuation.

With regard to combined VPA/lithium exposure, a greater reduction of MARCKS
protein was observed in cells following exposure to the combination than was produced by

either VPA or lithium/carbachol alone, and the effects of the two drugs appear to be

approximately additive. Whereas VPA induced a mean reduction in MARCKS expression

of 26% in the soluble fraction and 50% in the membrane, and lithium/carbachol elicited
mean reductions of 14% and 33% in soluble and membrane, respectively, the combination

of VPA plus lithium/carbachol resulted in a 51% drop in soluble MARCKS and a 67%

drop in membrane-associated MARCKS. These data are in agreement with clinical studies

which show VPA/lithium combination therapy to have greater therapeutic efficacy for









treatment of bipolar disorder than either agent administered individually (Calabrese and
Delucchi, 1989; Hayes, 1989). In contrast to lithium, which exerts a preferential down-
regulation of MARCKS in the soluble fraction of HN33 cells (Watson and Lenox, 1996),

the effect of VPA on MARCKS was most pronounced in the membrane fraction of these

cells. While the significance of this finding is unclear, the differences observed in the

patterns of MARCKS down-regulation between these two drugs may be a function of

differences in the pathways of PKC activation, as later discussed. This is of particular
interest in light of the findings that concomitant exposure of cells to both VPA and lithium

appears to have an additive effect on MARCKS down-regulation at therapeutic

concentrations of both drugs, and such findings support the notion that the two drugs

operate through separate mechanisms.

In contrast to lithium, which appears to rely on relative depletion of myo-inositol

and receptor activation of PI signaling to optimize its effect on MARCKS in the

immortalized hippocampal cells, these manipulations had no effect on the VPA-induced

down-regulation of MARCKS in these cells. Addition of carbachol or inositol (even at

mM concentrations) to the culture medium had no effect on the VPA-induced reduction in
MARCKS, and these findings are consistent with the lack of effect of VPA on receptor-

coupled accumulation of CMP-PA in the CHO-KI cells (Watson el al., 1998). CMP-PA,

also known as CDP-DAG, is a byproduct of DAG metabolism and precursor for PI
reformation, and serves as a sensitive measure of cycling through the PI signaling cascade

(refer to Figure 1.2). Thus, in spite of the shared property of lithium and VPA in down-

regulating the expression of MARCKS, these data suggest that the mechanism through

which this occurs may involve different pathways. Nonetheless, there appears to be

evidence for a role of PKC regulation in the action of both lithium and VPA, for previous

studies have demonstrated that phorbol esters down-regulate MARCKS protein in
neuronally derived cell populations in a PKC-dependent manner (Watson el al., 1994).

This will be further discussed in later chapters.









Two structural analogs of VPA (see Figure 3.9c), which are significantly less

potent as anti-convulsants in comparison to VPA (Chapman et al., 1982), were studied.
Both analogs are known metabolites of VPA, although they are produced through minor

metabolic pathways and represent a minimal percentage of total brain plasma

concentration (Chapman ei al., 1982; Eadie, 1991 ). In previous studies, neither
hydroxyvalproic acid nor 2-propylglutaric acid was shown to be effective in preventing

pentylenetetrazol-induced seizures in mice (Chapman et al., 1982; J. Sullivan, personal

communication), and neither agent has been tested clinically for efficacy in the treatment

of bipolar disorder. In our studies, hydroxyvalproic acid produced a statistically

significant effect in down-regulating membrane-associated MARCKS, but at a

concentration well above that observed for VPA (3 .0 mM hydroxyvalproic acid vs. 0.1-

1.0 mM VPA). Comparatively, the 2-propylglutaric acid analog possessed little potency

in down-regulating MARCKS in either the membrane or cytosolic fraction. Future

analysis of structural differences among these VPA compounds may prove useful in

identifying structurally related fatty acids that may be more or less efficacious not only in

down-regulating MARCKS, but also potentially as mood stabilizers.
Exposure of immortalized hippocampal cells to structurally and functionally

diverse psychotropic agents, including carbamazepine, fluoxetine, imipramine, phenytoin,

diazepam, haloperidol, morphine, and verapamil, did not result in a down-regulation of

MARCKS, even at doses that were well above the therapeutic range. It is of particular

note that carbamazepine, another anti-convulsant used less frequently in the treatment of
manic patients, and verapamil, also used unconventionally and perhaps less efficacious

than the more established treatments (Janicak el al., 1998; Walton el al., 1996; Mathis et

al., 1988; Barton and Gitlin, 1987), did not have any effect on MARCKS expression.

Carbamazepine has been shown to have effects within the adenylyl cyclase cascade, a
property shared by lithium at higher concentrations (Manji el al., 1995; Mork et al.,
1992). However, carbamazepine has generally been shown to be less effective overall than









lithium or VPA as a mood stabilizer (Keck el al., 1992; Bowden, 1996), and verapamil, a

calcium channel blocker, has yet to be studied extensively for this application.
Following exposure to a therapeutic serum concentration (0.6 mM) of VPA for up

to 14 d, HN33 cells exhibited decreased MARCKS mRNA levels, along a time course

similar to that of observed MARCKS protein alterations. This finding suggests that the
down-regulation of MARCKS protein might be attributed, at least in part, to a reduction

of its precursor mRNA template. With regard to MARCKS regulation, such evidence
may support more than one hypothesis. The observed reduction in MARCKS protein may

be the result of either altered transcriptional rates, or post-transcriptional regulation, via

decreased mRNA stability, decreased translation, or increased degradation. Yet, the

observed alterations in mRNA expression support the notion that MARCKS is being

regulated at the level of mRNA synthesis or stability. Recent studies in our lab have

shown that lithium, while altering both MARCKS mRNA and protein levels, did not affect

mRNA stability (Watson and Lenox, 1997); this suggests that the lithium-induced

alteration in MARCKS mRNA is primarily attributable to effects on transcription of the

MARCKS gene.

These studies suggest that MARCKS may serve as a target in the brain for the

selective action of mood stabilizing drugs. Although the therapeutic action of these drugs

can probably not be attributed fully to their effects on MARCKS, it is plausible that the

role of MARCKS in modulating plasticity may account for some of the long-term

stabilizing actions exhibited by VPA and lithium in the treatment of bipolar disorder. This
notion is supported by the finding that other drugs were set apart by both their inability to

down-regulate MARCKS as well as their inefficacy in treating bipolar disorder. Further

support for the notion of selective action in the CNS has been provided by data from our

laboratory which shows that the expression of MARCKS protein in two separate murine

macrophage cell lines (IC-21 and MH-S) is unaltered following chronic exposure to

therapeutic levels of VPA (3-7 d at 0.6 mM, data not shown). In addition to effects of









VPA on MARCKS, such alterations in neuronal plasticity may result from actions on

other potential protein substrates, as discussed below.


GAP-43


It is clear from these studies that VPA has a time- and concentration-dependent

effect on GAP-43 expression in HN33 cells. Further, the concentration range over which

this increase in GAP-43 occurred was comparable to the concentration range over which

the down-regulation in MARCKS protein occurred, and this concentration range is well

within the target therapeutic serum concentration. There are several potential implications

of such an increase in GAP-43 expression, since GAP-43 is an early marker of post-

mitotic neurons, and is highly expressed during axonal growth and synapse formation

(Aigner et al., 1995). For example, increased GAP-43 expression might signify synaptic

regrowth and/or remodeling, cytoprotection, or even development of aberrant pathways.

Synaptic remodeling was discussed previously as a potential mechanism for long-term

mood stabilization. A novel finding observed in carrying out these studies was that

following chronic (>1 d) VPA exposure, HN33 cells appeared to become more adherent

to the tissue culture flasks in which they were grown. This is of interest in light of the

finding that chronic VPA exposure also elicited an increase in expression of GAP-43,

which is known to be important for promoting and maintaining growth cone adhesion

during neurite outgrowth (Aigner and Caroni, 1995). Later chapters will include further

studies on the growth and morphological changes induced by VPA.

As previously mentioned, preliminary studies have shown no evidence for lithium-

induced regulation of GAP-43 expression in HN33 cells, and this major difference
between VPA and lithium is intriguing. Since the two drugs share similar effects on

MARCKS, one might hypothesize that this effect provides a basis for their common ability

to treat bipolar disorder. Conversely, the difference in their ability to affect GAP-43 might









relate either to their differing efficacies in treating various types of bipolar disorders (i.e.
VPA is generally more efficacious in treating rapid-cycling bipolar disorder), or,

alternatively, to their very different pharmacological profiles aside from their mood-

stabilizing efficacy (i.e., VPA is also an anti-epileptic, whereas lithium is not).


Conclusions


With regard to alterations in protein expression and their necessity and/or
sufficiency in bringing about the desired end result (that of attaining successful mood-

stabilizing treatment), GAP-43 induction may or may not be either. There is little

evidence to suggest that increased GAP-43 is necessary for the successful amelioration of

bipolar symptoms, for although VPA elicited GAP-43 induction, lithium, the prototypic

mood-stabilizing drug, had no effect on GAP-43 expression in our in vitro cell model.

Nonetheless, this does not preclude the possibility that lithium may exert such an effect on

GAP-43 in other cell model systems or in brain tissue. Whether GAP-43 alterations alone
are sufficient for optimal bipolar therapy is another question entirely, which these data

alone cannot fully address. Many agents may lead to an increase in expression of GAP-43,

if by virtue of nothing other than altering cell growth and differentiation rates, but it does

not necessarily follow, based on what is yet known, that all of these agents would serve as

sufficient mood-stabilizing drugs.

Down-regulation of MARCKS protein may be more consistent with a mood-

stabilization mechanism. The observation that both VPA and lithium, to the exclusion of a

number of drugs used for various other psychotropic applications, elicited a significant

reduction in MARCKS, suggests that alterations in MARCKS may be involved in the
mechanism of action of these agents, and supports the identification of MARCKS as a

potential molecular target which may or may not contribute to the therapeutic effects of

drugs used in the treatment of bipolar disorder. However, the findings that









carbamazepine, another widely used mood-stabilizing drug therapy (although not FDA-

approved for this application), as well as some less commonly used mood-stabilizing

agents, including lamotrigine (data not shown) and verapamil, did not alter MARCKS

expression, point to the possibility that either MARCKS is not integral to mood-stabilizing

efficacy, or that the in vitro exposure conditions and model we employed were not optimal

to allow the down-regulation we may have observed following chronic exposure in vivo.

Another important question concerning the effects observed is that of sufficiency. The

identification of MARCKS as a potential target for the action of drugs used in the

treatment of bipolar disorder suggests that the down-regulation of MARCKS alone, by

any method, would be enough to yield successful amelioration of bipolar symptoms.

However, this diminishes the potential role of other mechanisms, such as the alterations in

PI signaling induced by lithium, or the effects of VPA on GAP-43 expression. It is

conceivable that the alterations in MARCKS expression observed in vitro would lead, in

vivo, to further downstream events important for altering plasticity or other parameters

involved in changing the course of the disorder. In other words, MARCKS may be just

one target of many which may ultimately result in the efficacy of this group of drugs, those

which exhibit anti-manic and/or mood-stabilizing action. This is more compatible with the

notion that other drugs may be able to positively affect the course of bipolar therapy.

Whereas VPA and lithium may operate, at least in part, through their effects on

MARCKS, the involvement of MARCKS in bipolar therapy is indirect, and other drugs

could theoretically elicit the same end result, yet through a separate pathway, perhaps

bypassing MARCKS altogether.

It is interesting that VPA led to opposite effects on two CNS proteins which are

both involved in cellular plasticity and intracellular signaling, and both regulated by PKC.
This is despite the fact that the VPA-induced alterations of these proteins occurred over a
very similar time course and dose range. These data support recent in vivo findings in our

laboratory which imply an apparent inverse regulation of MARCKS and GAP-43 in






61

certain highly plastic brain regions, such as the hippocampus, in that levels of GAP-43 are

high in some regions while MARCKS levels are low, and vice versa (McNamara and

Lenox, 1997). This would suggest that at certain developmental stages or in particular

brain regions, MARCKS and GAP-43 play alternative or opposing roles, even though they

can be regulated similarly. Such comparable pharmacodynamic properties suggest that

both proteins play important roles in the long-term therapeutic action of VPA.














CHAPTER 4
ROLE OF PKC IN MECHANISM OF ACTION OF VPA


Introduction


PKC is an 80-kD protein expressed at high concentrations in neuronal tissues, and

the observation that PKC is the major enzyme responsible for regulation of both GAP-43

and MARCKS make the effects of VPA on PKC an interesting area of investigation. PKC

is a serine/threonine kinase which, by virtue of its effects on intracellular signal

transduction, is thought to play a role in hormone and neurotransmitter secretion and

regulation of cell proliferation and differentiation, through mechanisms which are still not
fully understood (Fujise el al., 1994, Hofmann, 1997). PKC is actually comprised of a

large family of at least 11 mammalian isoforms, classified into three groups based on their

activation profiles. The conventional (cPKC) group includes c, 031/1311, and y, which are
regulated by Ca+, diacylglycerols (DAG), negatively charged phospholipids, and phorbol

esters (potent tumor promoters which mimic DAG). The novel (nPKC) group contains 6,

&, i" (L), and 0, and is characterized by activation requirements similar to those of the

cPKC's, but lack of Ca+' activation. The atypical (aPKC) members are C, t (k), and 4t(D),

and are all independent of Ca++ and DAG or phorbols, requiring only phospholipids for
activation (Fujise el al., 1994, Hofmann, 1997). The PKC isozymes differ in their primary

structure and genomic origin, so that they are not derived from a single mRNA, but rather
from a series of structurally related mRNAs coding for a family of enzymes with very
similar properties. In humans, the genes for the various isozymes are located on different









chromosomes, with the exception of the p1I and p11 isoforms, which are the result of

differential splicing (Stabel and Parker, 1991).
The primary structure of PKC contains four major regions, termed C1 through C4,

which exhibit a high degree of homology between isotypes. These conserved regions are

generally thought to represent areas involved with functions common to all PKC

isozymes, such as the kinase domain. Interspersed between the conserved regions are
variable regions VI through V5, which exhibit a lower degree of homology and probably

allow for properties specific to individual isoforms. The higher level structure of PKC is

comprised of catalytic and regulatory domains. The catalytic domain alone is referred to

as PKM, and has a molecular weight of 45-55 kDa. It is located in the C-terminal region

of the polypeptide, and confers its kinase activity. The regulatory region, located at the

N-terminal and with an approximate molecular weight of 32-36 kDa, contains

phospholipid-and phorbol-binding domains responsible for modulating alterations in the

activity of the enzyme. It is at the regulatory region that diacylglycerol (DAG) and

phorbol esters bind to transiently activate the enzyme, eventually resulting in an increased

rate of degradation (down-regulation). All members of the PKC superfamily carry in the

N-terminal region at least one zinc-finger structure (more commonly two), which is

thought to aid in protein-protein interactions. The N-terminal region also contains a

pseudosubstrate sequence, conferring the enzyme's autoinhibitory activity. This region

serves to regulate activity of the enzyme, and removal of this region results in constitutive

and activator-independent kinase activity. The pseudosubstrate domain resembles the

consensus phosphorylation sequence but contains no phosphorylatable residue; it controls

kinase activity by occupying the catalytic site and thereby blocking access of the substrate

(Stabel and Parker, 1991).

cPKC activation is dependent upon two major events: (1) formation of the ternary

enzyme+Ca2++phospholipid complex (the presence of the phospholipid further enhancing

the affinity of the enzyme for calcium), and (2) binding of DAG (or phorbol), leading to









the conformational changes which result in activation. Down-regulation of the enzyme

follows prolonged activation, and is accomplished through both increased proteolysis and

sequestration (association of the enzyme with cytoskeletal or nuclear components).

Autophosphorylation of the enzyme is not a requirement, but probably plays a role (Stabel

and Parker, 1991).
Data from our laboratory indicate that PKC activity is decreased over time

following chronic lithium exposure, as is expression of PKC isozymes a and e (Watson

and Lenox, 1997). These findings are in agreement with those reported by Manji and

colleagues, which reflected reductions in PKC activity and expression of a and s isoforms

following either lithium or VPA exposure, in either rats or C6 glioma cells (Manji et al.,

1993; Chen et al., 1994; Manji et al., 1996). Tissue distribution of PKC isozyme

expression varies, and the relative expression of each isoform may contribute to its unique
role in signal transduction. Individual PKC isozymes, upon activation, are differentially

compartmentalized, suggesting that they mediate distinct cellular functions. PKC-y
appears to be the only isoform unique to brain and spinal cord, with no expression

elsewhere; its precise role remains unknown. The same is true for most of the other

isoforms although there has been no dearth of studies into roles of the various PKC

isozymes, results have been inconclusive and in some cases contradictory (to be discussed

further in Discussion). Whereas PKC-ox and F have been implicated in tumor growth and

suppression of apoptosis, 8 may induce opposite effects (Hofmann, 1997). Both PKC-cL

and -5 are thought to be important in early brain development (Blackshear et al., 1996),
while PKC-C may play a role in proliferation, and several isoforms have reportedly been

implicated in neurite outgrowth and/or cell differentiation (Ways el al., 1994; Hundle et

al., 1995; Borgatti et al., 1996). We have previously established that the immortalized

hippocampal cell line HN33 expresses PKC isoforms a, 6, F, and C, and our laboratory has

demonstrated in vitro effects of lithium on PKC activity as well as on expression levels of
various PKC isozymes (Watson and Lenox, 1997). In light of previous reports









demonstrating VPA-induced alterations in PKC in a non-neuronal cell line, it will be of

interest to investigate the effects of VPA exposure on these parameters in HN33 cells,

especially given preliminary evidence from our laboratory for effects of VPA on

MARCKS and GAP-43 in these cells. It has been suggested that the colocalization of

some isoforms with substrate(s) may ensure preferential and rapid phosphorylation

following PKC activation. Based on evidence indicating a down-regulation of MARCKS

and increased expression of GAP-43 following chronic VPA exposure, it is predicted that

VPA will activate PKC, at least transiently, so that the downstream effects on MARCKS

and GAP-43 can occur as observed. If VPA does, indeed, activate PKC, both its activity

level and isoform expression should decrease over time, in the characteristic pattern of

down-regulation following long-term activation. This acute activation followed by

chronic down-regulation has been demonstrated previously in the HN33 cells with lithium,

phorbol esters, and retinoic acid.

These and other studies suggest that VPA has significant effects on both PKC and

its substrates, though the connection between such effects is unclear. Although one would

expect PKC alterations to be in large part responsible for any further downstream effects

on its substrates MARCKS and GAP-43, this direct cause-and-effect relationship has yet

to be established. The following studies will therefore help to establish whether or not

PKC is necessary to the downstream effects on MARCKS and GAP-43, and which

isoforms may be most important for each effect. After first establishing the ability of VPA

to effect alterations in both PKC activity and expression of various PKC isozymes in

HN33 cells, the next objective will be to address the role of PKC in altering MARCKS

and GAP-43 levels. VPA-induced alterations in PKC-specific activity will be measured

indirectly by phosphorylation of the substrate MBP, and associated alterations in PKC

isoform expression will be measured using Western immunoblotting techniques. PKC

inhibitors will then be used to assess whether PKC activity is necessary to elicit the

observed VPA-induced alterations in MARCKS and GAP-43, and isoform-specific






66


antisense oligonucleotides will be used to narrow down which isoforms are important for

each parameter. It is hypothesized that PKC activity will be necessary to any observed

effects by VPA on MARCKS or GAP-43, and that inhibition of the enzyme will block
such observed effects.









Results


Effect of VPA Exposure on PKC Activity in HN33 Cells

HN33 cells exposed to 0.6 mM VPA for 15 min 3 d were assayed for effects on

PKC activity, as measured by MBP phosphorylation (Figure 4.1). Acute VPA exposure

(15 min 4 h) resulted in a tendency toward increased membrane-associated activity and a

correlated decrease in PKC activity in the cytosolic fraction. This paradigm most closely

follows a translocation phenomenon, whereby cytosolic PKC is shuttled into the

membrane for activation. By the 3 d time point, activity in the membrane fractions is

down-regulated (p<0.05). Similar results were observed using MARCKS peptide as the

phosphorylation substrate (data not shown).

160
I 175
A 140 150 T T
1- 125
120 1(X, l

100 50
0 05 1 1.5 2 25 3 35 4 45


60-
E
4-"--- Soluble
40-
--- Membrane
20-
0 ... ... .. .. .. . .. .
g 0 4 8 12 16 20 24 28 32 3640 44 48 52 5660 64 68 72 76
Time (hours)

Figure 4.1. Effect of VPA exposure on PKC activity in HN33 cells. VPA exposure leads
to apparent PKC activation, followed by down-regulation. HN33 cells exposed to 0.6
mM VPA were processed and soluble and membrane fractions isolated as detailed in
Materials and Methods (Chapter 2). Assay was performed utilizing the PKC activators
PMA (phorbol), CaCl2, and phosphatidylserine. PKC activation was measured by the y-
32p phosphorylation (in pmol/min/mg) of the PKC substrate MBP. Mean activity of
control samples was 102 8.1 pmol/min/mg. The figure shows the overall pattern of
PKC activity from 15 min (acute) to 3 d (chronic) VPA exposure; the inset shows PKC
activation only at acute time points (up to 4 h). Results are mean S.E.M. of at least four
determinations, and are expressed as percent of activity in untreated control samples.
*p<0.05









Effect of VPA on PKC-a

PKC-t is observed entirely in the soluble fraction of HN33 cells. As shown in

Figure 4.2, exposure to 0.6 mM VPA results in an acute (15 min 12 h) increase in PKC-

a expression, followed by down-regulation over time (p<0.05).


a ONO me" AMb 4W#-00 m

1 2 3 4 5 6

160 ,


b 2m
1 2


TIME (hours)
Figure 4.2. Effect of VPA on PKC-ct expression in soluble fraction of HN33 cells. (a)
Representative Western immunoblot of soluble PKC-ax expression following acute (24 h
or less) VPA exposure @ 0.6 mM. Protein samples were loaded into lanes as follows:
Lane 1, untreated control; Lane 2, 15 min VPA; Lane 3, 1 h VPA; Lane 4, 4 h VPA; Lane
5, 12 h VPA; Lane 6, 24 h VPA. (b) Representative Western immunoblot of cytosolic
PKC-(x expression following chronic (3 d) VPA exposure @ 0.6 mM. Lane 1 was loaded
with protein from an untreated control, and Lane 2 was loaded with protein from a sample
exposed to VPA for 3 d. (c) HN33 cells were exposed to 0.6 mM VPA for 15 min 3 d,
and PKC-at levels (present only in the cytosolic fraction of these cells) were assessed by
Western immunoblot analysis. Results are expressed as percent of untreated control, and
are the mean S.E.M. of four determinations. *p<0.05


Effect of VPA on PKC-6

PKC-6 undergoes a significant increase in expression levels in both soluble

(p<0.001) and membrane (p<0.0001) fractions of cells during chronic VPA exposure (1 &

3 d), and as late as 3 d, there appears to be no down-regulation (Figure 4.3).















7 8 9 10


1 2 1 2


TIME (hours)


Figure 4.3. Effect of VPA on PKC-6 expression in HN33 cells. (a) Representative
Western blot of PKC-6 expression following acute (<24 h) VPA exposure @ 0.6 mM.
The blot at left (Lanes 1-6) shows cytosolic PKC in HN33 cells, while the blot at right
(Lanes 7-12) shows membrane-associated PKC. Protein samples were loaded into
lanes as follows: Lane 1, untreated control; Lane 2, 15 min VPA; Lane 3, 1 h VPA;
Lane 4, 4 h VPA; Lane 5, 12 h VPA; Lane 6, 24 h VPA; Lane 7, untreated control;
Lane 8, 15 min VPA; Lane 9, 1 h VPA; Lane 10, 2 h VPA, Lane 11, 4 h VPA; Lane
12, 12 h VPA. (b) Representative Western of PKC-8 expression following chronic (3
d) VPA exposure @ 0.6 mM. The blot at left shows cytosolic PKC in HN33 cells,
while the blot at right shows membrane-associated PKC. Lane 1 was loaded with
protein from an untreated control, and Lane 2 was loaded with protein from a sample
exposed to VPA for 3 d. (c) HN33 cells were exposed to 0.6 mM VPA for 15 min 3
d, and PKC-6 levels were assessed by Western immunoblot analysis. Results are
expressed as percent of untreated control, and are the mean S.E.M. of four
determinations. *p<0.001, soluble and *p<0.0001, membrane.


54


11 12


a
2 3










a di
1 2 3 4 5 6 1 2 3 4 5 6




b i UFII
I 2 I


TIME (hours)


Figure 4.4. Effect of VPA on PKC-e expression in HN33 cells. (a) Representative
Western immunoblot of PKC-6 expression following acute (24 h or less) VPA
exposure @ 0.6 mM. The six lanes at left show cytosolic (or soluble) PKC in HN33
cells, while the six lanes at right show membrane-associated PKC. Protein samples
were loaded into lanes as follows: Lane 1, untreated control; Lane 2, 15 min VPA;
Lane 3, 1 h VPA, Lane 4, 2 h VPA; Lane 5, 4 h VPA; Lane 6, 12 h VPA. (b)
Representative Western immunoblot of PKC-e expression following chronic (3 d)
VPA exposure @ 0.6 mM. The blot at left shows cytosolic (or soluble) PKC in
HN33 cells, while the blot at right shows membrane-associated PKC. Lane I was
loaded with protein from an untreated control, and Lane 2 was loaded with protein
from a sample exposed to VPA for 3 d. (c) Cells were exposed to 0.6 mM VPA for
15 min 3 d, and PKC-e levels were assessed by Western immunoblot analysis.
Results are expressed as percent of untreated control, and are the mean S.E.M. of
four determinations. *p<0.05









Effect of VPA on PKC-P
As shown in Figure 4.4, membrane-associated PKC-s increases acutely (15 min 4
h), and then returns to basal levels by the 3 d time point. Cytosolic PKC-c expression
remains close to baseline acutely, and then increases significantly following prolonged (1-3

d) VPA exposure (p<0.05). This pattern of expression reflects an apparent activation and

translocation of PKC-g, whereby PKC is activated, shuttled from cytosol into the

membrane (where the enzyme is active), and then down-regulated and returned from

membrane to cytosol following prolonged activation.


Effect of VPA on PKC-C
PKC-C shows little change in expression at early time points, and then is up-

regulated sometime between I d and 3 (p
regulation observed in the soluble fraction (Figure 4.5).


Effect of PKC-a Antisense Oligonucleotides on PKC-z Expression in HN33 Cells

Antisense oligonucleotides were designed to bind to various segments of the
mouse and/or human PKC-a mRNA sequence, and block its translation to protein. Figure

4.6 and Table 4.1 list the antisense oligonucleotides and show the regions of mouse PKC-

a gene they were designed to bind. All oligonucleotides were synthesized with
phosphorothioate backbones, in order to increase their stability in culture. A series of

experiments was then conducted to down-regulate PKC-aX expression in HN33 cells, first

using a single antisense oligonucleotide sequence, then using the combination of five

antisense oligonucleotides. A sense oligonucleotide was used as the negative control for

these experiments.












*MAN* M
1 2 3 4 5 6


1 2 1


1 2 3 4 5 6


2


0 10 20 30 40 50 60 70 80


TIME (hours)



Figure 4.5. Effect of VPA on PKC-C expression in HN33 cells. (a) Representative
Western immunoblots of PKC-C expression following acute (24 h or less) VPA
exposure @ 0.6 mM. The blot at left shows cytosolic (or soluble) PKC in HN33 cells,
while the blot at right shows membrane-associated PKC. Protein samples were loaded
into lanes as follows: Lane 1, untreated control; Lane 2, 15 min VPA; Lane 3, 1 h
VPA; Lane 4, 4 h VPA; Lane 5, 12 h VPA; Lane 6, 24 h VPA. (b) Representative
Western immunoblot of PKC-C expression following chronic (3 d) VPA exposure @
0.6 mM. The blot at left shows cytosolic (or soluble) PKC in HN33 cells, while the
blot at right shows membrane-associated PKC. Lane I was loaded with protein from
an untreated control, and Lane 2 was loaded with protein from a sample exposed to
VPA for 3 d. (c) HN33 cells were exposed to 0.6 mM VPA for 15 min 3 d, and
PKC- levels were assessed by Western immunoblot analysis. Results are expressed as
percent of untreated control, and are the mean S.E.M. of four determinations.
*p<0.0001, membrane










25-
24-


26- 27


-28


mouse PKC alpha gene


5 ta
sta rt


Figure 4.6. PKC-ci antisense oligo-nucleotide design. This figure depicts the
antisense oligonucleotides designed to bind to PKC-a mRNA and block its
translation. Shown is the representative PKC-a gene from mouse (as reported in
the NCBI database), and the five oligonucleotides (24, 25, 26, 27, 28) designed to
bind to the nucleotide sequence at the indicated, base-complementary regions.
Start codon is demarcated by the arrow; both oligonucleotides 24 and 25 span this
region. Not shown is #23, the sense oligonucleotide designed as a negative
control, which reads as the sense complement to antisense oligo 24. All
oligonucleotides are 20-base phosphorothioate linked, with the exception of #23
and #24, which are 15-base phosphorothioates, and all were cross-referenced
through GenBank to ensure unique sequence identity with PKC-a.




Table 4.1 lists the 5' to 3'nucleotide sequences of each PKC-a oligonucleotide
(ON) by number, and the species specificity of each.


Table 4.1. Antisense oligonucleotide sequences.

ON # Sequence of Nucleotides Species Specificity

23 ACCATGGCTGACGTT mouse & human

24 AACGTCAGCCATGGT mouse & human

25 CAGCCATGGTTCCCCCCAAC mouse

26 AGAAGGTAGGGCTTCCGTAT mouse & human

27 GTTACCAACAGGACCTAGCT mouse

28 TGAGTTCAGCATGACATTGT mouse









Initial experiments attempted to establish the optimal dose range for

oligonucleotide exposure and protein down-regulation. Oligonucleotide concentrations
ranging from 0.1 to 1.0 p.M (the standard concentrations reported in similar studies) had

no effect on cell viability or PKC-ct expression following 24 h exposure (data not shown).

Increasing oligonucleotide concentration to 10 VM and lengthening exposure time to 48 h

had no effect on PKC-ct expression (data not shown).

Next, methods were employed to increase oligonucleotide uptake and compensate

for potential oligonucleotide degradation. In keeping with previous reports of successful

antisense-mediated protein down-regulation, the cationic carrier lipid LipofectAINE

(Gibco BRL, Gaithersburg, MD) was used at 5 ptg/ml and 15 ptg/ml, and cells were serum-

deprived during oligonucleotide exposure. In addition, medium and its constituents were
replaced fresh every 12 h during exposure. PKC-ct expression remained at basal levels

following 24 or 48 h exposure to oligonucleotide under these conditions (data not shown).

The final approach utilized phorbol to down-regulate PKC expression (non-

selectively), then PKC-a antisense oligonucleotides were added following phorbol

removal, in an effort to selectively block PKC-a recovery. A representative Western is

shown in Figure 4.7. HN33 cells were exposed to 1.0 .M phorbol 12,13-dibutyrate
(PDBu) for 24 h, after which cells were washed with media to remove drug. Next, cells

were exposed to 1.0 p.M sense or antisense oligonucleotides (total of five antisense

oligonucleotides) directed to PKC-ot (Figure 4.6). Cells were exposed to oligonucleotides

for a total of 24 h, but antisense-treated cells showed no difference in PKC-a expression

than sense-treated or untreated control cells. As shown in Figure 4.7, there is clearly no

difference in PKC-ox expression between those PDBu-pre-treated cells which are

antisense-treated or untreated. PDBu exposure resulted in a reversible down-regulation of
PKC-a (Lane 2 vs. 1), and PKC-ct antisense failed to prevent the recovery of PKC-a

expression at rates comparable to control levels (Lanes 7&8 vs. 3-6). Whereas toxicity
commonly arises as a problem in antisense applications in cell and tissue culture, the HN33









cells showed no evidence of cytotoxicity following 24-48 h exposure to any of the
pharmacological agents at any of the concentrations tested.






1 2 3 4 5 6 7 8

Figure 4.7. Effect of PKC-ct antisense oligonucleotides on PKC-a expression in HN33
cells. Shown is a representative Western immunoblot of PKC-X expression in HN33 cells
following antisense treatment. Lane 1 holds an untreated control; Lane 2 shows PKC-ct
expression in cells treated with 1.0 pWM PDBu for 24 h; Lane 3 shows cells treated with
1.0 pM PDBu for 24 h, then no treatment for 24 h; Lane 4 shows cells treated with 1.0
jiM PDBu for 24 h, then LipofectAMINE for 24 h; Lane 5 shows cells treated with 1.0
4M PDBu for 24 h, then sense oligo #23 for 24 h; Lane 6 shows cells treated with 1.0 PM
PDBu for 24 h, then sense oligo #23 and LipofectAMINE for 24 h; Lane 7 shows cells
treated with 1.0 pM PDBu for 24 h, then antisense oligonucleotides #24, 25, 26, 27, and
28, each at 1.0 p.M, for 24 h; Lane 8 shows cells treated with 1.0 JIM PDBu for 24 h, then
antisense oligonucleotides #24, 25, 26, 27, and 28, plus LipofectAMINE, for 24 h.


Effect of PKC Inhibitor LY333531 on VPA-Induced Down-Regulation of MARCKS

LY333531 alone had no effect on MARCKS expression, but when combined with
VPA, LY333531 prevented the VPA-induced MARCKS down-regulation observed in the

membrane, but not soluble, fraction of HN33 cells (Figure 4.8). MARCKS expression in

both cytosol and membrane of VPA-treated cells was significantly different from that of

LY33353 1-treated cells (p<0.005, soluble; p<0.0001, membrane). Cytosolic MARCKS in

combined VPA/LY3 33531-treated cells was significantly different from that of

LY33353 1-treated cells (p<0.005), indicating that, even in the presence of the PKC

inhibitor LY333531, VPA was able to induce down-regulation of cytosolic (but not

membrane-associated) MARCKS expression. In contrast, LY333531 prevented the VPA-

induced down-regulation of MARCKS in the membrane fraction of HN33 cells.










120
110-
100-
90-

70
60
50-
40-
30-
20-
10-
0-


[] Soluble


0 Membrane


Control VPA LY333531 VPA + LY333531
Treatment


2 3 4 5 6 7 8


Figure 4.8. Effect of PKC inhibitor LY333531 on VPA-induced down-regulation
of MARCKS. (a) HN33 cells were exposed for 24 h to 1.0 mM VPA with or
without 1.0 W LY33353 1. MARCKS expression was quantitated by Western
blot analysis, and results (expressed as percent of untreated control) reflect the
mean S.E.M. of at least five determinations. *p<0.005, soluble; *p<0.0001,
membrane (b) Representative Western blot showing MARCKS expression in the
presence of VPA LY33353 1. Lanes 1-4 represent cytosolic MARCKS, and
were treated for 24 h as follows: control, untreated (1); 1.0 mM VPA (2); 1.0 JIM
LY333531 (3); 1.0 mM VPA + 1.0 p.M LY333531 (4). Lanes 5-8 show
membrane-associated MARCKS, and were treated for 24 h as described for 1-4
(respectively) above.

Effect of PKC Inhibitor LY333531 on VPA-Induced GAP-43 Induction

LY33353 1, alone or in combination with VPA, had no effect on GAP-43

expression in HN33 cells (Figure 4.9). GAP-43 levels in both soluble and membrane

fractions are significantly increased (p<0.05) in both VPA-treated groups, compared to







77


controls (untreated or LY33353 1-treated). Therefore, the PKC inhibitor LY333531 was

unable to prevent the VPA-induced GAP-43 increase, for even in the presence of

LY33353 1, VPA increased GAP-43 expression in both soluble and membrane fractions.

These data indicate that PKC inhibition alone is insufficient to prevent the GAP-43

increase observed in HN33 cells following 24 h 1.0 mM VPA exposure.


*


Control VPA


5] Soluble

0 Membrane


Treatment


OM Aw -"0Im


1 2 3 4 5 6 7 8




Figure 4.9. Effect of PKC inhibitor LY333531 on VPA-induced GAP-43 increase. (a)
HN33 cells were exposed for 24 h to 1.0 mM VPA with or without 1.0 M
LY33353 1. GAP-43 expression was quantitated by Western blot analysis, and results
(expressed as percent of untreated control) reflect the mean S.E.M. of three
determinations. *p<0.05 (b) Representative Western blot showing GAP-43 expression
in the presence of VPA LY33353 1. Lanes 1-4 represent membrane-associated GAP-
43, and were treated for 24 h as follows: control, untreated (1); 1.0 mM VPA (2); 1.0
p.M LY333531 (3); 1.0 mM VPA + 1.0 glM LY333531 (4). Lanes 5-8 show cytosolic
GAP-43, and were treated for 24 h as described for 1-4 (respectively) above.


*


240
220 -
200 -
180 -
160-
140 -
120 -
100 -
80-
60-
40-
20-
0-


LY333531 VPA + LY333531









Discussion


Inasmuch as PKC activation has been associated with regulation of expression of

both GAP-43 and MARCKS, the effects of VPA on PKC isozyme levels and PKC activity
were investigated more directly, in an effort to better assess the role of PKC in the

mechanism of action of VPA in the brain. Whereas lithium, the first-line anti-manic/mood-

stabilizer, is believed to operate in part through the PI signaling pathway, our group and

others have provided evidence against such a mechanism for VPA (Vadnal and
Parthasarathy, 1995; Lenox et al., 1996; Dixon and Hokin, 1997; Watson et al, 1998),

although both drugs have similar effects on MARCKS expression.

Our findings indicate that, like lithium, VPA induces alterations in expression of

PKC isoforms cc and (Watson and Lenox, 1997). However, the time course for VPA-

induced changes is earlier than that for lithium, and PKC-s expression actually increases

(acutely in membrane, chronically in cytosol), in contrast to the reduction previously
reported for both lithium and VPA (Chen et al., 1994; Manji et al., 1996; Watson and

Lenox, 1997). Chen et al. (1994) reported that chronic VPA exposure of C6 glioma cells

reduced the expression of PKC isozymes a and & in intact cells, a finding similar to that

observed following chronic lithium administration in the same cell model (Manji et al.,

1993). Subsequent studies revealed PKC-induced alterations in multiple components of
the 13-adrenergic receptor-coupled cAMP-generating system, including J31-AR, Gas, and

adenylyl cyclase, each of which is phosphorylated by PKC (Chen et al., 1996). The

significance of these changes is unclear, but together they suggest a major role for PKC in

the VPA- or lithium-induced molecular effects underlying their efficacy, and may be
instructive when other PKC-related effects are compared between the two drugs.

In this study, VPA exposure resulted in an apparent transient activation of PKC

(within 2 h), followed by a gradual reduction in activity with chronic exposure (4 h and

beyond). PKC activity was significantly down-regulated by the 72 h time point. Similar









results were reported following exposure of HN33 cells to the phorbol PMA, though

activation occurred within 15 minutes and down-regulation of PKC activity was evident

by I h (Watson et al., 1994). This finding is further in agreement with previous reports

which showed that chronic VPA or lithium exposure of C6 glioma cells significantly
reduced PKC activity in both membrane and soluble fractions (Chen et al., 1994; Manji et

al., 1996). The observed patterns of alterations in protein activity and expression suggest

that isoforms ox and F might, under these conditions, be involved in the phosphorylation

and regulation of MARCKS protein expression, since both VPA and lithium elicited

alterations in PKC-a and -e expression in HN33 cells, and the time course of MARCKS

alterations roughly corresponds to that of the changes observed in PKC isoform

expression. Indeed, previous studies have indicated the ability of these two isoforms to
phosphorylate MARCKS, whereas PKC-C is unable to phosphorylate MARCKS in vitro

(Fujise et al., 1994; Herget et al., 1995; Uberall et al., 1997). One could hypothesize a

process in which VPA, directly or indirectly, activates PKC, especially isoforms o and/or

&, at early time points (between 0 and 12 h). By 3 d, there is a significant down-regulation

of activity in the membrane and a comparable drop in the cytosol. Consistent with this
hypothesis are data which show an increase in PKC activity within 15 seconds following

phorbol exposure (Stabel and Parker, 1991), and a time course for MARCKS down-

regulation which closely follows the time course of PKC activation in HN33 cells exposed

to phorbol (Watson et al., 1994).

While VPA and lithium both affect PKC isozyme ax (and perhaps F), the two drugs
have very different effects on PKC-5. VPA induces a chronic up-regulation in PKC-6,

whereas lithium has little or no effect on PKC-5 expression (Watson and Lenox, 1997). In

comparing VPA and lithium, one might hypothesize that such differences in patterns of
PKC activation underlie the differences observed in downstream effects such as protein

expression or cell growth and morphology. For example, the up-regulation of PKC-5
resulting uniquely from VPA exposure is probably not important for the MARCKS down-









regulation observed, but might be involved in the GAP-43 induction observed following

VPA but not lithium exposure; in support of this hypothesis is evidence indicating that

PKC-8 and -& phosphorylate GAP-43 (Oehrlein et al., 1996). Indeed, in our study, the

GAP-43 increase occurred over a time course and of a magnitude similar to those of the

PKC-5 alterations observed. Conversely, PKC isozymes a and/or e may play a role in the

observed VPA- and lithium-induced down-regulation of MARCKS.

Reports are in disagreement as to the role of PKC in neuronal differentiation. A

number of studies have suggested that, rather than being a positive regulator of the

process, PKC may actually be an inhibitory influence, so that suppression of the enzyme

results in increased process outgrowth. For example, use of various PKC inhibitors, and

subsequent reductions in PKC activity, have been associated with increased neurite

outgrowth (Heikkila el al., 1989; Parodi et al., 1990; Tsuneishi, 1992; Wooten, 1992;

Jalava et al., 1993; Ekinci and Shea, 1997). Long-term down-regulation of PKC activity

with phorbols such as PMA has provided similar results (Tsuneishi, 1992; Carlson et al.,

1993; Jalava et al., 1993; Ekinci and Shea, 1997), with one study showing enhanced

neurite outgrowth following combined phorbol and PKC inhibitor treatment (Heikkila et

al., 1989). However, some of these same studies have further shown that down-

regulation of PKC activity alone is insufficient to elicit a differentiated phenotype (Ekinci

and Shea, 1997; Carlson et al., 1993; Wooten, 1992).

On the other hand, numerous reports have presented equally convincing evidence

for a positive role for PKC in regulating neuronal differentiation and a positive role in

maintaining the functionally active state (Coleman and Wooten, 1994; Cabell and

Audesirk, 1993; Parrow et al., 1992; Abraham et al., 1991; Tonini et al., 1991), though

this role is probably not unique to PKC. Previous data have implicated a number of

different PKC isoforms in cellular differentiation, including all four of the isotypes present

in HN33 cells: a, 5, F, and C (Wada et al., 1989; Leli et al., 1992, 1993; Ways et al.,

1994; Hundle et al., 1995; Borgatti et al., 1996). Indeed, the subcellular localization and









activation of PKC isozymes are regulated in an isoform-specific manner during

neurogenesis, suggesting they are involved in control of neural development and in
neuronal differentiation (Oehrlein et al., 1998).

The role of PKC in proliferation and differentiation remains unclear, for reports

differ as to which PKC isoform(s) is/are responsible for various aspects of cellular

differentiation. Some studies have attributed to PKC-at a role as the intermediate in

cellular differentiation, as a result of alterations in PKC-a mRNA and/or protein

expression observed concomitant with differentiation of PC 12 and other cell lines

(Borgatti et al., 1996; Parrow et al., 1995; Murray et al., 1993; Leli et al., 1993; Tonini et

al., 1991; Wada et al., 1989). However, these reports differ in their mechanistic

conclusions, for in three studies, PKC-a expression increased with differentiation

(Borgatti et al., 1996; Parrow et al., 1995; Murray et al., 1993), while in three other

studies, PKC-x and -, levels had decreased following comparable treatment and

neuritogenesis (Leli et al., 1993; Tonini el al., 1991; Wada et al., 1989). Leli's group

further showed that the intracellular delivery of PKC-a and -e isoform-specific antibodies

resulted in a morphologically differentiated phenotype in SH-SY5Y cells (Leli et al.,

1992).

While some groups observed no differentiation-associated alterations in isoforms

other than PKC-ox, a number of reports have indicated that in various cell lines, other

members of the PKC superfamily are involved. Borgatti's group, who showed that PKC-a

was up-regulated during PC 12 cell differentiation, found in the same study that PKC 6, s,

and were down-regulated (Borgatti et al., 1996). A number of studies have provided

evidence for a role of PKC-s in regulating neuronal differentiation, since neurite
outgrowth has been associated with altered PKC-6 levels (Ponzoni et al., 1993; Parrow et
al., 1995; Fagerstrom et al., 1996; Ekinci and Shea, 1997; Hundle et al., 1997; Zeidman et

al., 1999). In support of the notion that PKC-s may be involved in the process, as

suggested by Leli's and Wada's work (Leli et al., 1993; Wada et al., 1989), a separate









study showed that overexpression of PKC-s but not PKC-8 led to an increase in NGF-
induced neurite outgrowth in stably transfected PC 12 cell lines (Hundle et at., 1995). In

stark contrast to these findings, yet another group reported just the opposite in NIH 3T3

cells, in that PKC-8 overexpression induced a morphological change and reduced growth
rate consistent with differentiation that was not observed in cells overexpressing PKC-&

(Mischak et aL, 1993). Further, PKC-C has been similarly examined in transfected
leukemic cells, and results indicate that PKC-C overexpression is able to induce cellular

differentiation (Ways et al., 1994). Another study that utilized antisense techniques also

emphasized the importance of PKC-C in NGF-induced neuronal differentiation (Coleman

and Wooten, 1994).

All of these studies point to different PKC isozymes as mediators of similar effects.

These apparently contrasting results may merely be a function of the varying experimental

methods and conditions used from one study to the next, such as different cell lines,

exposure conditions, and other parameters. These findings may alternatively indicate that

(1) each isozyme plays its own, specialized role in the complex process of cellular
differentiation, and (2) alterations in the level of each isozyme may be compensated for by

alterations in the levels and actions of other isozymes. Further studies will need to be

conducted in order to answer these questions, to confirm more definitively the association

between each PKC isoform and the parameter in question, but these data strongly suggest

that both VPA and lithium may down-regulate MARCKS via a PKC-dependent

mechanism. On the other hand, the unique role of the PI signaling pathway in the action

of lithium, and the differential pattern of activation of the PKC isozymes by VPA and
lithium may indicate that there are multiple pathways to the down-regulation of MARCKS

expression, and may also confer different clinical therapeutic properties on these two
drugs. The findings further suggest that the action of VPA on MARCKS protein

expression could be cell and/or tissue specific, for it may be modulated, in part, by the

expression pattern of the PKC isozymes specific to the cell types. Indeed, Stabel and









Parker (1991) suggest that the distribution of MARCKS closely follows the distribution of
PKC, based on early studies of PKC activity showing wide disparity across regions (Kuo

et al., 1980; Minakuchi et al., 1981).
Having established an apparent causal relationship between VPA and PKC, the

purpose of the next experiments was to establish more precisely the necessity of PKC in
eliciting the obesrved VPA-induced effects on MARCKS and GAP-43. In order to better

establish the role of various PKC isoforms in the MARCKS down-regulation and GAP-43
increase observed, experiments were designed employing antisense technology. Through

this method, antisense oligonucleotides are designed and directed to specifically bind and

"knock down" the nucleic acid precursors of each individual PKC isozyme, resulting in

transient elimination of the protein in question. There are four main mechanisms by which

antisense oligonucleotides are hypothesized to work: (1) by inhibiting the necessary

processing and transport of RNA, (2) by interfering with binding of the ribosome to the

nucleotide sequence, so that translation is not initiated, (3) by inhibiting the completion of

translation after the process has begun, and (4) by binding to the sequence, thereby

characterizing it as abnormal and resulting in degradation by RNase H (Schlingensiepen et

al., 1997).
For this study, a number of trials were performed in an effort to down-regulate

PKC-a in the HN33 cells. One sense and five antisense oligonucleotides were prepared by

Gemini Biotech, Ltd. (The Woodlands, TX) through the DNA Synthesis Core of the UF

Interdisciplinary Center for Biotechnology Research (ICBR). Each oligonucleotide was

designed to bear unique identity to the PKC-ca genome of either mouse or human, as

reported in the National Center for Biotechnology Information database, and screened

through GenBank. Various regions of the cDNA sequence were targeted, including the

start codon (see Figure 4.6). In an effort to maximize stability and binding of

oligonucleotides, all were composed of 15 or 20 bases each, and had phosphorothioate
linkages in place of the endogenous phosphodiester backbone of DNA. Whereas toxicity









commonly arises as a problem in antisense applications in cell and tissue culture, the HN33

cells showed no evidence of cytotoxicity following I to 2 d exposure to I and 10 M
concentrations of PKC-ct oligonucleotides. In initial experiments, cells were exposed to

oligonucleotide(s) for 24-48 h, in an effort to establish a protocol for the consistent
reduction of PKC-a levels prior to VPA exposure. When numerous efforts failed to elicit

any reduction in PKC-ci expression, PKC was first down-regulated through 24 h, 1
PDBu (phorbol) administration prior to exposure of cell cultures to oligonucleotide.

Following thorough phorbol washout, test cells were exposed to all five oligonucleotides

concomitantly for 24 h. In order to increase uptake of oligonucleotides, cells were

deprived of serum nutrients and exposed to LipofectAMINE, an agent which serves as a

cationic carrier lipid to transport oligonucleotide into the cells. Additionally, media and

oligonucleotides were replaced after 12 h to compensate for any degradation which had

taken place. Despite these measures designed to optimize antisense-directed repression of
PKC-a levels following initial down-regulation by phorbol, the protein recovered in

antisense-treated cells at about the same rate as that of drug-free controls (see Figure 4.7).

Common explanations for the difficulty in achieving significant decreases in protein

expression include instability of the antisense oligonucleotide sequence, limited uptake of

the oligonucleotide, suboptimal binding of oligonucleotide to its target sequence, and half-

life of the particular protein. Despite efforts to account for and overcome these issues,

attempts to carry out further studies based on the antisense work were ultimately

abandoned, due to the persistent failure of PKC-ct antisense to down-regulate PKC-a

expression. Perhaps, in this study, the unpredictable tertiary conformation of the antisense

oligonucleotides was such that access and/or binding to the target mRNA sequence was

prevented, or perhaps the oligonucleotides were sequestered and/or degraded before they

even reached their destination. Because antisense technology is still in its developmental

stages, and oligonucleotide design and exposure conditions vary greatly among systems,









significant time and expense may often be invested before down-regulation of the protein

in question is optimized.
Whereas expression of each PKC isoform could not be down-regulated

individually, overall PKC activity inhibition was achieved through the use of the PKC

inhibitor LY333531 (Eli Lilly and Company, Indianapolis, IN), a direct competitor for

ATP binding (Jirousek et al., 1996). LY33353 1 is a PKC-P3-selective inhibitor (at low nm

concentrations) which also inhibits PKC isozymes oX, 6, and e at concentrations in the low

pM range (Ishii et al., 1996; Jirousek et al., 1996). Exposure of HN33 cells to 1.0 W

LY333531 for 24 h resulted in no change in either MARCKS or GAP-43 expression, nor

any evidence of toxicity. When combined with 1.0 mM VPA, LY333531 effectively

prevented the VPA-induced MARCKS down-regulation in the membrane fraction of

HN33 cells, but with no accompanying effect on the VPA-induced GAP-43 increase.
From these studies it is clear that PKC activation plays an integral role in the VPA-

induced MARCKS effects observed, as inhibition of PKC-directed phosphorylation of

MARCKS resulted in the inability of VPA to elicit down-regulation of membrane-

associated MARCKS. It is interesting that this LY33353 1-induced prevention of
MARCKS down-regulation by VPA occurred in membrane, the fraction of HN33 cells in

which VPA exerts its greatest effect. This finding is further consistent with the known

localization of PKC activity in the membrane, and also with a previous report showing that

phosphorylation of MARCKS is necessary for its dissociation and translocation from

membrane (Allen and Aderem, 1995). Whether PKC activation alone is sufficient to

reduce MARCKS has previously been investigated by our laboratory. It was shown that

phorbol esters, potent activators of PKC, yield a significant down-regulation of MARCKS
in the HN33 cells, and that the time course of phorbol-induced MARCKS alterations

follows closely behind the time course of phorbol-induced PKC activation and down-

regulation (Watson et al., 1994). These findings support the notion that activation (and

perhaps also subsequent down-regulation) of PKC, in and of itself, is sufficient to elicit a









reduction in MARCKS, though PKC activation may not be the only route through which

the MARCKS alterations may occur. Indeed, Brooks et al. (1992) reported evidence of

regulation of MARCKS through both PKC-dependent and independent pathways, with
platelet-derived growth factor (PDGF)-induced down-regulation of MARCKS through a

post-transcriptional mechanism.

As for GAP-43, PKC activation is probably not necessary to elicit the alterations in

GAP-43 observed. This is evidenced by the fact that PKC inhibition failed to block the
VPA-induced increase in GAP-43 in the HN33 cells. Even with the strong GAP-43

expression observed in the membrane of these cells, and the significant up-regulation of

this protein to several times basal levels, LY333531 caused no attenuation of the VPA-

induced GAP-43 increase observed. Assuming conditions were optimal for significant

PKC inhibition, we can conclude that PKC activation is not entirely responsible for the

VPA-induced GAP-43 increase observed, nor is PKC activation alone likely to be

sufficient for the effect, since lithium-induced alterations in PKC activity and isoform

expression did not lead to any change in GAP-43 expression.
The findings do little to address the issue of the involvement of PKC in bipolar

disorder and successful mood-stabilizing therapy. However, one can hypothesize that if

MARCKS dysregulation is integral to bipolar disorder, and PKC is the primary modulator

of MARCKS, then perhaps the enzyme is perturbed in some patients exhibiting symptoms

of the disorder. This would be especially applicable to cases in which the patient

demonstrates normal levels of MARCKS, for MARCKS, being downstream from PKC, is

regulated by PKC, and if PKC parameters are abnormal, then PKC-induced MARCKS

regulation will likely also be abnormal. Even if MARCKS itself is not involved in the
manifestation of the disorder, PKC may be a major player. Indeed, previous studies have

provided evidence for alterations in PKC-related signaling in patients with bipolar

disorder. Reports have shown that in both platelets and post-mortem brain samples of

individuals diagnosed with the disease, PKC translocation and activity levels were









significantly increased as compared to controls, as measured by phorbol- or serotonin-

stimulated histone phosphorylation (Friedman et al., 1993; Wang and Friedman, 1996). In

addition, PKC isoform expression was altered in comparison to controls, with higher
levels of PKC-ct, 6, and and lower PKC-e expression in bipolar patients (Wang and

Friedman, 1996). Further, two-week lithium treatment of subjects resulted in a reduction
in both PKC translocation and activity in platelets (Friedman et al., 1993). In a more

recent clinical study, the PKC-selective inhibitor tamoxifen (an estrogen receptor

antagonist widely used in the treatment of breast cancer) has been shown to be effective in

treating acute mania in a small sample of patients (Manji et al., 1999). Inasmuch as PKC

is believed to play a pivotal role in regulating neuronal signal transduction and in

modulating intracellular cross-talk between neurotransmitter systems (Nishizuka, 1992),

these studies are suggestive of the potential for affective (mood) dysfunction as a result of

alterations or dysfunction in PKC and associated neuronal signaling pathways. A

definitive association between PKC alteration and mood stabilization may facilitate the

development of a whole new line of agents useful in the treatment of bipolar disorder.














CHAPTER 5
EFFECT OF VPA ON CELL VIABILITY, GROWTH AND MORPHOLOGY


Introduction


While VPA has proven very useful in its clinical applications to seizures and

bipolar disorders, the detrimental effects of this drug on the developing embryo have been

well documented (Wiger et al., 1988; Gofflot el al., 1996; Menegola et al., 1996). High

frequencies of neural tube defects (NTDs) have been linked to prenatal VPA exposure

(Bjerkedal et al., 1982; Robert and Guibaud, 1982; Lindhout and Schmidt, 1986; Wiger et

al., 1988; Gofflot el al., 1996; Menegola et al., 1996). Such defects may include open

neural tube, spina bifida, anencephaly, meningomyelocele, and encephalocele, and are

thought to be caused by a number of different factors, such as environmental or

pharmacological teratogens, genetic anomalies, and various maternal predisposing factors

(Rhoads and Mills, 1986; Norman et al., 1995). In more recent years, folic acid and/or
methionine supplementation have been shown to reduce the frequency of NTDs, both in

the general population and in offspring of mothers undergoing anticonvulsant therapy

(Laurence et al., 1981; Smithells et al., 1981; Rhoads and Mills, 1986; Trotz el al., 1987;

Mulinare et al., 1988; Nosel and Klein, 1992; Ehlers el a., 1996). Therefore, women are

advised to discontinue VPA therapy or take folic acid supplements during and just before

pregnancy. The mechanism of this protective effect is not understood. Folate is known to

be a precursor of methionine, an important amino acid for cell proliferation, and vitamin
B 12 is an essential cofactor for methionine synthesis. In mothers of NTD-affected

offspring, circulating serum folate levels are low but not clinically deficient; the same is









true for B12 levels (Laurence et al., 1981; Rhoads and Mills, 1986; Kirke et al., 1993). In
addition, homocysteine, the direct precursor for methionine, is found at high levels in the

serum of some mothers in this population (Steegers-Theunissen et al., 1994; Mills et al.,
1995, 1996). This evidence suggests that, in at least a fraction of NTD cases, the mothers

of NTD-affiicted offspring exhibit a metabolic insufficiency, perhaps at the level of the
homocysteine-to-folate conversion step. This "metabolic defect" hypothesis is in contrast

to the simpler notion of a nutritional deficiency (of either folate or methionine), and is

consistent with the possibility that VPA therapy is one factor which may induce or

enhance this metabolic defect. It is of note that, in experimental studies, in vivo

supplementation of either nutrient to pregnant animals resulted in decreased frequencies of

NTDs in offspring, while in vitro studies did not reproduce the protective effect (Trotz et

al., 1987; Hansen and Grafiton, 1991; Nosel and Klein, 1992; Hansen et a., 1995; Ehlers

et al., 1996). Preliminary studies in our laboratory further support the notion that the

effect is of a complicated metabolic nature, for folate or methionine supplementation to

cell growth medium did not prevent the VPA-induced down-regulation of MARCKS

observed in HN33 cells (Watterson unpublished data).
In light of the deleterious developmental effects of this drug, the effects of VPA on

cell growth and differentiation merit in-depth investigation. Previous studies, while few in
number, have provided evidence for a dose-dependent VPA-induced inhibition of cell

proliferation via arrest at mid-GI of the cell cycle (Martin and Regan, 1991), as well as

VPA-induced differentiation of neuroblastoma and glioma cells (Regan, 1985; Cinatl et

al., 1996). The inhibition of cell proliferation, alterations in cellular interactions, and

altered differentiation have all been correlated with prenatal drug teratogenicity, especially
of the developing neural tube (Wilk et al., 1980; Mummery et al., 1984; Martin et al.,

1988). It is plausible that arrest of cell proliferation at a specific phase in the cell cycle,

accompanied by irreversible differentiation, may result in the effects observed to be

consistent with prenatal teratogenicity.