Directory UMM :Data Elmu:jurnal:B:Biological Psichatry:Vol48.Issue7.2000:
Associated Disturbances in Calcium Homeostasis and
G Protein–Mediated cAMP Signaling in
Bipolar I Disorder
Masoumeh Emamghoreishi, Peter P. Li, Lyanne Schlichter, Sagar V. Parikh,
Robert Cooke, and Jerry J. Warsh
Background: Evidence of extensive cross-talk between
calcium (Ca21)- and cAMP-mediated signaling systems
suggests that previously reported abnormalities in Ca21
homeostasis in bipolar I (BP-I) patients may be linked to
disturbances in the function of G proteins that mediate
cAMP signaling.
Methods: To test this hypothesis, the b-adrenergic agonist, isoproterenol, and the G protein activator, sodium
fluoride (NaF), were used to stimulate cAMP production
in B lymphoblasts from healthy and BP-I subjects phenotyped on basal intracellular calcium concentration
([Ca21]B). cAMP was measured by radioimmunoassay
and [Ca21]B by ratiometric fluorometry with fura-2.
Results: Isoproterenol- (10 mM) stimulated cAMP formation was lower in intact B lymphoblasts from BP-I patients
with high [Ca21]B ($ 2 SD above the mean concentration
of healthy subjects) compared with patients having normal
B lymphoblast [Ca21]B and with healthy subjects. Although basal and NaF-stimulated cAMP production was
greater in B lymphoblast membranes from male BP-I
patients with high versus normal [Ca21]B, there were no
differences in the percent stimulation. This suggests the
differences in NaF response resulted from higher basal
adenylyl cyclase activity.
Conclusions: These findings suggest that trait-dependent
disturbances in processes regulating b-adrenergic receptor sensitivity and G protein–mediated cAMP signaling
occur in conjunction with altered Ca21 homeostasis in
those BP-I patients with high B lymphoblast [Ca21]B.
Biol Psychiatry 2000;48:665– 673 © 2000 Society of Biological Psychiatry
From the Section of Biochemical Psychiatry (ME, PPL, JJW) and Mood Disorder
Program (SVP, RC, JJW), Centre for Addiction and Mental Health, Clarke Site;
Playfair Neuroscience Unit, The Toronto Hospital (LS); and the Departments of
Psychiatry (PPL, SVP, RC, JJW), Pharmacology (PPL, JJW), and Physiology
(LS) and the Institute for Medical Sciences (JJW), University of Toronto,
Toronto, Canada.
Address reprint requests to J.J. Warsh, M.D., Ph.D., University of Toronto, Centre
for Addiction and Mental Health, Clarke Division, 250 College Street, Toronto
Ontario M5T 1R8, Canada.
Received June 29, 1999; revised March 16, 2000; accepted March 20, 2000.
© 2000 Society of Biological Psychiatry
Key Words: Bipolar disorder, G proteins, signal transduction, cAMP, calcium homeostasis, B lymphoblasts
Introduction
E
vidence from studies of postmortem brain and peripheral blood cells implicate disturbances in G protein–
mediated cAMP signaling and Ca21 homeostasis in the
pathophysiology of bipolar affective disorder (BD; reviewed in Li et al 2000; Manji 1992; Warsh and Li 1996).
For instance, increased stimulatory G protein a subunit
(Gas) levels (Manji et al 1995; Young et al 1994), blunted
receptor- and G protein–stimulated cAMP accumulation
(Ebstein et al 1988; Mann et al 1985; Pandey et al 1979)
and increased basal intracellular calcium levels ([Ca21]B;
Dubovsky et al 1992b; Emamghoreishi et al 1997) have all
been observed in mononuclear leukocytes from patients
with BD. Moreover, some of these signaling disturbances
appear to be trait dependent, possibly reflecting underlying
molecular “vulnerability” factors in BD (Emamghoreishi
et al 1997; Warsh et al 1999). Although it has been
speculated that these signal transduction abnormalities are
linked through cross-talk mechanisms (Dubovsky et al
1992a; Warsh and Li 1996), to our knowledge there is no
empirical evidence for such a link.
Extensive cell-type specific cross-talk occurs between
G protein– coupled cAMP- and phospholipase C (PLC)–
linked second-messenger systems (Hill and Kendall 1989;
Liu and Simon 1996). Multiple adenylyl cyclase (AC)
subtypes provide the potential for Ca21 either to synergise
or attenuate cAMP signaling (Sunahara et al 1996). Thus,
Ca21/CaM directly activates AC isotypes I and VIII,
whereas Ca21 inhibits isotypes V and VI (Sunahara et al
1996). Ca21-activated protein kinase C (PKC) may also
increase or decrease cAMP levels by phosphorylating
b-adrenergic receptors (b-ARs; Bell and Brunton 1987),
Gsa, Gia (Katada et al 1985; Sugden and Klein 1988;
Yoshimasa et al 1987) or some AC subtypes (Cooper et al
1995). Recently, it has been shown that Ca21-activated
0006-3223/00/$20.00
PII S0006-3223(00)00884-2
666
M. Emamghoreishi et al
BIOL PSYCHIATRY
2000;48:665– 673
Table 1. Subject Characteristics
Age (years)
Gender (female/male)
Lifetime psychiatric
comorbidityb
B lymphoblast [Ca21]B
Healthy
subjects
BP-I
high [Ca21]B
BP-I
normal [Ca21]B
38.6 6 3.60a
7/4
NA
37.7 6 4.00
6/5
1—Panic disorder,
2—alcohol dependence
74.8 6 1.70c
39.8 6 4.27
3/8
5—Alcohol abuse or
dependence
57.5 6 1.03
53.2 6 1.75
a
Mean 6 SEM.
BP patients with a lifetime history of alcohol abuse or dependence had no recent history (.3 months) of substance abuse at
the time of study. One BP-I subject with normal [Ca21]B also had a past, but not recent, history of polysubstance abuse.
c
Tukey p ,.001, compared with healthy subjects and BP-I patients with normal [Ca21]B.
b
PKC suppresses protein kinase A (PKA)-mediated signaling in fibroblast cell lines (Dobbeling and Berchtold
1996), providing another site of interaction between these
signaling pathways.
cAMP-mediated processes also affect Ca21 signaling
by inhibiting Ca21 influx (Rasmussen 1986), desensitizing
inositol trisphosphate receptors (Supattapone et al 1988),
stimulating Ca21 removal from the cytosol by Ca21
ATPase (Helman et al 1986), and inhibiting PLCb2 (Liu
and Simon 1996). In other instances, G protein–mediated
cAMP signaling enhances that of Ca21, including PKAmediated phosphorylation of voltage-dependent Ca21
channels in excitable cells (Chad et al 1987), or the ion
gate that closes K1 channels in cells in general, thus
prolonging depolarization and enhancing Ca21 influx
(Kandel et al 1987). There is also evidence that in
non-excitable cells, receptor-activated Ca21 channels are
opened by an increase in the concentration of cytoplasmic
cAMP (Barritt 1999). Such mechanisms as noted above
would account for the types of interaction (i.e., stimulation/suppression) that occur between cAMP and phosphoinositide signaling pathways.
As disturbances in both G protein–mediated cAMP and
Ca21 signaling have been previously reported in BD,
cross-talk regulatory mechanisms are potentially important candidates to be considered in explaining abnormalities in these two signal transduction pathways. Thus, the
increased lymphocyte and platelet intracellular Ca21
([Ca21]i) reported in BD might result from the altered G
protein function previously identified in these patients, or
vice versa. In effect, defects in one signaling system may
decompensate or, alternatively, induce adaptive changes in
other signaling systems in an attempt to maintain
homeostasis.
To test this hypothesis, we used Epstein–Barr virus
(EBV)–immortalized B lymphoblasts derived from BD
patients, a cellular model that demonstrates trait-dependent abnormalities in Ca21 homeostasis (Emamghoreishi
et al 1997) and in b-AR sensitivity (Kay et al 1993) in this
disorder. These cell lines can be grown in continuous
culture removed from in vivo neurohormonal influences
and medications, which circumvents the potential of such
variables to hamper detection of signal transduction abnormalities and interactions between signaling cascades.
To screen for possible relationships between disturbed G
protein function and altered basal intracellular [Ca21]
([Ca21]B), isoproterenol and sodium fluoride (NaF)-stimulated cAMP formation were determined in B lymphoblasts from bipolar I (BP-I) patients stratified based on
[Ca21]B compared with healthy subjects. We report here
the coexistence of altered receptor- and G protein–mediated cAMP signaling and calcium homeostasis, supporting
the view that these abnormalities are both interrelated and
trait-dependent.
Methods and Materials
Subjects
Cell lines were selected from a subset of BP-I patients and
healthy subjects recruited and assessed as previously described
(Emamghoreishi et al 1997). Briefly, subjects were recruited
from the Mood Disorders inpatient and outpatient units of the
Clarke Institute of Psychiatry. After a clinical diagnosis of BP-I
disorder was made by a consulting psychiatrist, the diagnosis was
confirmed for research purposes using the Structured Clinical
Interview for DSM-IV Axis I Disorders (SCID I/P, version 2.0;
First et al 1995b) administered by a research psychiatrist or
trained psychiatric research assistant, complemented by review
of medical records. All interview and historical material was
further reviewed by a research psychiatrist to ensure the reliability of the diagnoses. Patients had no history of recent (.3
months) substance abuse. Healthy comparison subjects were
recruited through posted advertisements in the University of
Toronto and were assessed by a trained research assistant and a
research psychiatrist using the SCID-I/NP, version 2.0 (First et al
1995a). Healthy subjects had no past or family (first-degree
relatives) history of psychiatric disorder. All subjects were
physically healthy, had no past history of serious medical illness,
and had no personal or family history of diabetes or hypertension. The study was approved by the Human Subjects Review
Committee of the University of Toronto and, after providing a
complete description of the study, all subjects provided written
informed consent. Table 1 shows the demographics and [Ca21]B
Ca21 Homeostasis and cAMP Signaling in BD
of the BP-I patients and healthy subjects from whom cell lines
were selected for study.
B Lymphoblast Transformation and Intracellular
Calcium Measurements
Mononuclear leukocyte preparations were isolated by density
gradient separation through Ficoll-Hypaque and B lymphocytes
in these isolates were transformed by infection with EBV as
previously described (Emamghoreishi et al 1997). The transformed, immortalized B lymphoblasts were grown in RPMI1640 medium containing L-glutamine (2 mmol/L), pyruvate (1
mmol/L), 20% fetal bovine serum (FBS), 100 mg/mL streptomycin, 100 units/mL penicillin in an incubator (95% air/5% CO2 at
37°C) with passages every 3– 4 days. At 12–15 passages, aliquots
of cells were removed and [Ca21]B determined as described
below. The remaining cells were frozen in aliquots (107 cells/
mL) and stored over the vapour phase of liquid nitrogen until
they were regrown and assayed for receptor- and G protein–
stimulated cAMP formation.
[Ca21]B was measured in B lymphoblasts using ratiometric
fluorometry and the calcium sensitive dye, fura 2-acetoxymethylester, as described previously (Emamghoreishi et al 1997).
BP-I subjects were stratified based on [Ca21]B greater than
(high) or less than (normal) two standard deviations above the
mean [Ca21]B determined in cells from healthy subjects, to
classify patients into these two phenotypes.
Regeneration of B lymphoblast Cell Lines
B lymphoblasts were removed from the liquid nitrogen freezer
and immediately placed on dry ice. Cells were then thawed
quickly (,5 min) at 37°C and transferred gently into 25-cm2
culture flasks containing 5 mL pre-warmed (37°C) B lymphoblast medium and regrown as previously described (Emamghoreishi et al 1997). After nine passages from the time of regeneration
(total passages 21–24), cells were transferred to B lymphoblast
medium containing 10% FBS and incubated for three additional
passages. Finally, cells were transferred to B lymphoblast medium without FBS and incubated for 24 hours, then harvested
and divided in two aliquots. One aliquot (8 –10 3 107 cells) was
used immediately to assay isoproterenol-stimulated cAMP formation, and the other was stored at 270°C until used to assay G
protein-stimulated AC activity.
Determination of Basal and IsoproterenolStimulated cAMP Production in Intact
B Lymphoblasts
Isoproterenol was used to assess b-adrenergic receptor-stimulated cAMP formation in intact B lymphoblasts using a method
modified from Kay et al (1993). Cells were washed once in
buffer containing 130 mmol/L NaCl, 0.54 mmol/L KCl, 0.005
mmol/L CaCl2, 0.098 mmol/L MgCl2, 0.1% glucose and 15
mmol/L Tris.HCl pH 7.6, resuspended at 3 3 106 cells/mL, and
incubated in 0.9 mL aliquots (30 min, 37°C, 95% air/5% CO2).
The reactions were begun by adding 50 mL isobutyl-methylxanthine (IBMX, final concentration 0.5 mmol/L), 50 ml ascorbic
BIOL PSYCHIATRY
2000;48:665– 673
667
acid (final concentration 0.1 mmol/L) with or without isoproterenol, and incubating at 37°C for 10 min. Reactions were stopped
by boiling for 10 min, then centrifuging at 12,000 g for 10 min
at 4°C. The protein concentration in each pellet was determined
by the Lowry method (Lowry et al 1951). The supernatants,
which contained the cAMP, were evaporated to dryness under
vacuum, stored at 270°C until needed, then reconstituted with
100 mL assay buffer containing 50 mmol/L sodium acetate at pH
6.2, and 0.01% sodium azide. The cAMP content was measured
using an 125I-radioimmunoassay kit following the supplier’s
instructions (Mandel, Guelph, Canada). Maximally stimulating
and saturating concentrations of isoproterenol (1 and 10 mmol/L,
respectively) were determined in preliminary dose-response
studies, and subsequently used to assess B lymphoblast receptorstimulated cAMP production in the comparison groups.
Determination of Basal and NaF-Stimulated AC
Activity in B Lymphoblast Membranes
MEMBRANE PREPARATION. B lymphoblasts were thawed
on ice, resuspended in 20 volumes ice-cold buffer (10 mmol/L
Tris, pH 7.4, 1 mmol/L EGTA) and homogenized by hand using
a glass-teflon homogenizer (15 strokes). The homogenates were
centrifuged at 20,000 g (20 min, 4°C), then the pellets were
re-suspended and centrifuged twice more. Each pellet was
resuspended in 500 mL ice-cold TEMS buffer (50 mmol/L Tris,
pH 7.4, 1 mmol/L EGTA, 5 mmol/L MgCl2, 12% sucrose), and
protein concentrations were determined by the method of Bradford (Bradford 1976), using bovine serum albumin as a standard.
ASSAY OF NAF-STIMULATED AC ACTIVITY. Membrane
homogenates were diluted with TEMS buffer to give 1 mg
membrane protein/ml and the samples kept on ice until use.
Reaction mixtures (200 mL) contained 50 mmol/L Tris, pH 7.4,
1 mmol/L EGTA, 3.5 mmol/L MgCl2, 0.1 mmol/L IBMX, 1.5
mg/mL BSA, 0.5 mmol/L adenosine triphosphate (ATP), 25
mmol/L phosphocreatine, and 250 units/mL creatine phosphokinase. Membrane homogenates (50 mL) were incubated at 37°C
for 15 min in 100 mL of reaction mixture. The reactions were
initiated by adding 50 mL ATP (2 mmol/L) in the absence or
presence of NaF (2.5, 5, and 10 mmol/L), followed by gently
vortexing and incubation at 37°C for 10 min with gentle shaking.
The reaction was terminated by boiling the tubes for 10 min,
followed by cooling on ice and centrifuging at 12,000 g for 10
min at 4°C. The supernatants were then removed and stored at
270°C until used for cAMP measurements as described above.
Statistical Analysis
Differences in dependent variables (basal and stimulated cAMP
formation) were assessed by repeated-measures multivariate
analysis of variance (MANOVA). Concentration (basal and
various concentrations of isoproterenol or NaF) was the repeated
“within” factor and diagnosis was the “between” subject factor.
NaF-stimulated cAMP formation values were logarithmically
transformed to normalize the distribution of the data. Based on
previous studies, we tested directional hypotheses (i.e., of reduced responsivity to isoproterenol stimulation [Extein et al
668
M. Emamghoreishi et al
BIOL PSYCHIATRY
2000;48:665– 673
Table 2. Basal and Isoproterenol-Stimulated cAMP Production in B Lymphoblasts from Healthy
Subjects and Bipolar I Patients with High (.69.4 nmol/L) and Normal (,69.4 nmol/L) [Ca21]B
Bipolar I disorder
Basal
Isoproterenol (1 mmol/L)
% stimulation
Isoproterenol (10 mmol/L)
% stimulation
21
Healthy
High [Ca ]B
Normal [Ca21]B
9.24 6 1.28
16.9 6 2.55
82.9 6 12.5
16.6 6 3.00
79.4 6 19.4
10.7 6 1.71
16.8 6 4.24
52.3 6 15.7
14.7 6 3.83
33.5 6 15.2a,b
12.0 6 1.62
20.5 6 3.08
69.0 6 12.6
21.6 6 3.42
81.8 6 17.6
cAMP production (see Methods) is expressed as pmol/mg protein/10 min (mean 6 SEM, 11 subjects in each group). The %
stimulation over the basal level is shown.
a
Significantly different (t 5 2.0, p 5 .025) from stimulation by 1 mmol/L isoproterenol in the same patient group.
b
Significantly different from BP-I subjects with normal [Ca21]B (t 5 1.95, p 5 .03) and from healthy subjects (t 5 1.85,
p 5 .035).
1979; Pandey et al 1979] and increased NaF-stimulated cAMP
formation [Young et al 1993] in B lymphoblasts from BP-I
patients). Thus, one-tailed probability values of p ,.05 were
used to distinguish statistically significant differences. Statistical
analyses were performed using the SPSS (release 7.0) statistical
software package (SPSS Inc., Chicago).
Results
Basal and Isoproterenol-Stimulated cAMP
Production in Intact B Lymphoblasts
Repeated-measures analysis of variance (ANOVA) performed on the absolute values for cAMP formation
showed a significant effect of isoproterenol [F(2,60) 5
26.1, p ,.001], but no significant interaction between
diagnosis and isoproterenol [F(4,60) 5 1.0, p 5 .2] or
main effect of diagnosis [F(2,30) 5 0.71, p 5 .25].
When the isoproterenol-stimulated AC activity was expressed as the percent stimulation above basal, there was
a significant interaction between diagnosis and the percent
stimulation [F(4,60) 5 2.04, p 5 .05] and a trend
toward a significant main effect of diagnosis [F(2,30) 5
1.92, p 5 .08]. Post hoc comparison of cell means by
contrasts showed a significant reduction in the percent
stimulation above basal at 10 mmol/L isoproterenol in
BP-I patients with high [Ca21]B compared with normal
[Ca21]B, and compared with healthy subjects (59%, t 5
1.95, p 5 .03 and 58%, t 5 1.85, p 5 .035,
respectively; Table 2). The increase in absolute cAMP
levels was 17% lower than that achieved with 1 mmol/L
isoproterenol stimulation in B lymphoblasts from BP-I
patients with high [Ca21]B, but was not statistically
significant compared with the cAMP levels reached following stimulation of B lymphoblasts from healthy subjects with 10 mmol/L isoproterenol. Reduced percent
stimulation above basal was also observed with 1 mmol/L
isoproterenol in BP-I patients with high [Ca21]B compared
with healthy subjects, although this difference did not
reach statistical significance (37%, t 5 1.58, p 5 .065).
Additionally, the reduced isoproterenol-stimulated cAMP
production in BP-I subjects with high [Ca21]B was more
marked at 10 mmol/L and was significantly different from
that determined at 1 mmol/L isoproterenol (t 5 2.0, p 5
.025). In preliminary experiments using membrane preparations, isoproterenol-stimulation was very low and variable, as we have also observed for other tissues (Young et
al 1993). Thus, there was no further attempt to measure
isoproterenol-stimulated cAMP formation in membranes
from the subject samples studied.
Basal and NaF-Stimulated AC Activity in B
Lymphoblast Membranes
NaF increased cAMP production in a concentration-dependent manner, with maximal stimulation at 5 mmol/L
and a reduced response at 10 mmol/L NaF (Figure 1A). A
repeated-measures MANOVA test showed significant
main effects of NaF concentration [F(3,72) 5 23.1, p
,.001] and gender [F(1,24) 5 4.5, p 5 .02], but not
of diagnosis [F(2,24) 5 1.5, p 5 .12]. Moreover,
statistically significant interactions were found between
sex and NaF concentration [F(3,72) 5 6.5, p 5 .001],
and between diagnosis and NaF [F(6,72) 5 1.95, p 5
.04]. Simple effects were performed to examine the
interactions. Contrasts of cell means revealed that male
subjects were significantly different from female subjects
in the magnitude of the response to 2.5 mmol/L NaF
[F(1,24) 5 21, p ,.001] and in the reduced response
to 10 mmol/L versus 5 mmol/L NaF [F(1,24) 5 4.96,
p 5 .02) (Figure 1, C and D).
When cAMP levels were expressed as percent change
above the basal level (Figure 1B), repeated-measures
MANOVA tests also showed a significant effect of the
NaF concentration [F(2,46) 5 33.9, p ,.001], an
interaction between gender and NaF concentration
[F(2,46) 5 4.0, p 5 .01], and a significant interaction
between NaF concentration and diagnosis [F(4,46) 5
2.1, p 5 .046]. Male subjects differed significantly in
Ca21 Homeostasis and cAMP Signaling in BD
BIOL PSYCHIATRY
2000;48:665– 673
669
Figure 1. Sodium fluoride (NaF)–stimulated adenylyl cyclase (AC) activity production: dose–response curves for B lymphoblast
membranes from bipolar I (BP-I) patients with high [Ca21]B (.69.4 nmol/L, E), normal [Ca21]B (,69.4 nmol/L, ‚), and their ageand gender-matched healthy controls (h). (A, B) Males and female subjects combined (n 5 10). (C–F) Results stratified by gender
(n 5 4 – 8 male and 2– 6 female subjects per subject comparison group). Repeated-measures multivariate analysis of variance
(MANOVA) tests revealed no significant differences in basal or NaF-stimulated AC activity among comparison groups (A and B). For
male subjects (C), repeated-measures MANOVA tests showed a significant main effect of diagnosis [F(2,14) 5 3.5, p 5 .028].
*Significant differences (p #.05) compared with corresponding NaF-stimulated response in the male BP-I group with normal
[Ca21]B.
the extent of reduction in response to 10 mmol/L compared to 5 mmol/L NaF [F(1,23) 5 6.2, p 5.01]
(Figure 1 E and F).
Figure 1 (C–F) shows NaF-stimulated cAMP production expressed in absolute values and as a percent above
basal cAMP production, in the comparison groups stratified by gender. Because the number of female BP-I
subjects with normal [Ca21]B was small, further statistical
analysis of data was performed only on male bipolar
subjects. Repeated-measures MANOVA showed significant effects of NaF concentration [F(3,42) 5 48.8, p
,.001], an interaction between NaF concentration and
diagnosis [F(6,42) 5 2.4, p 5 .02], and a significant
main effect of diagnosis [F(2,14) 5 3.5, p 5 .028;
Figure 1C]. Male BP-I patients with high [Ca21]B had
significantly higher absolute values of basal cAMP (142%
greater, p , .05) and NaF-stimulated cAMP production
(86% greater at 2.5 mmol/L NaF, p 5 .05; 94% at 5
mmol/L, p ,.05; 81% at 10 mmol/L, p ,.05) compared
with male BP-I patients with normal [Ca21]B. Their cAMP
values were also higher than healthy male subjects, but the
differences were not statistically significant (84% at 2.5
mmol/L, p 5 .09; 73% at 5 mmol/L, p 5 .08). In
contrast, analysis of the NaF-stimulated cAMP levels in B
670
BIOL PSYCHIATRY
2000;48:665– 673
lymphoblast membranes from male subjects expressed as
a percent increase above their respective basal activities
(Figure 1E) showed only a significant effect of NaF
concentration [F(2,28) 5 35.4, p ,.001].
Discussion
We have used B lymphoblasts from BP-I patients that
display high [Ca21]B to test whether alterations in calcium
homeostasis are related to disturbances in receptor-G
protein coupled cAMP formation in BP-I patients manifesting this cellular endophenotype. The principal findings
are reduced b-AR-stimulated cAMP formation in intact B
lymphoblasts, and elevated basal G protein/effector-dependent cAMP production in membranes of B lymphoblasts from BP-I patients with high [Ca21]B compared
with healthy subjects. Specifically, the b-AR response to
10 mmol/L isoproterenol was significantly lower in BP-I
patients with high [Ca21]B than in healthy subjects, and
although the response to 1 mmol/L isoproterenol was also
lower, it was not statistically significant. Male BP-I
patients with elevated B lymphoblast [Ca21]B also had
significantly higher basal and NaF-stimulated AC activity
compared with male BP-I subjects with normal [Ca21]B
and a trend towards higher NaF-stimulated AC activity
compared with healthy subjects. These findings suggest
that the alterations in intracellular calcium homeostasis in
B lymphoblasts from BP-I patients are indeed associated
with disturbances in receptor- and G protein– coupled AC
activity. Moreover, the co-expression of these disturbances in cells from BP-I subjects implies that this
spectrum of signal transduction abnormalities is trait
dependent.
Maximally stimulating and saturating concentrations of
isoproterenol were used to assess b-AR-stimulated cAMP
production in B lymphoblasts because of the limited
signal-to-noise ratio of the response and to ensure a
maximal level of cAMP stimulation was achieved in all
cases in the absence of full dose response studies on each
cell line. b-ARs in transformed B lymphoblasts show
qualitatively similar ligand binding characteristics and AC
activation to those in B lymphocytes, but lower density
and maximal enzyme activity (Ebstein et al 1985; Elliot
1987). Furthermore, because the functional interaction
between receptors and G proteins is strongly influenced by
their relative expression levels (Kenakin 1996), the small
responses observed here likely reflect the low density of
b-ARs in these cell lines. Dose-response studies were not
performed, and inferences about the efficacy of receptor-G
protein coupling (reflected in the EC50) and exact nature of
the altered responsivity will require future concentrationresponse characterization.
The small magnitude and inter-sample variability in the
M. Emamghoreishi et al
isoproterenol-stimulated cAMP production at 1 mmol/L
may have contributed to the lack of statistically significant
differences in B lymphoblast b-AR responses, between
BP-I patients with high [Ca21]B and either patients with
normal [Ca21]B or healthy subjects. Nevertheless, our
results highlight the possibility that factors regulating the
responsivity of the b-AR-G protein-AC complex are
differentially altered in BP-I patients who manifest the
endophenotype of high B lymphoblast [Ca21]B. Although
b-AR density was not determined in this study, previous
studies have shown reduced b-AR-stimulated AC activity
in lymphocytes (Mann et al 1985) and B lymphoblasts
(Kay et al 1993) from BD patients in the absence of
changes in b-AR density or affinity. Furthermore, we have
not found differences in Gas and Gai levels in B lymphoblasts from BP-I patients compared with their matched
healthy subjects (unpublished data). These observations
argue against the possibility that altered levels of b-AR
and of these G protein a subunits account for the reduced
b-AR-stimulated response in B lymphoblasts from BP-I
patients observed here. Other downstream processes that
regulate the response of the receptor-G protein-AC transduction apparatus should be considered, and it is notable
that Kay et al (1993) found less b-AR down-regulation in
response to isoproterenol in B lymphoblasts from BD
patients than in cells from control subjects. These observations suggested the process(es) regulating b-AR desensitization and/or down-regulation may be compromised in
BD. b-AR down-regulation and/or desensitization is mediated by receptor phosphorylation through PKA and/or
b-AR kinase (Sibley et al 1987). Therefore, the lower
b-AR responsiveness in BP-I patients with high [Ca21]B,
compared with healthy subjects and BP-I patients with
normal [Ca21]B, could similarly be related to alterations in
b-AR kinase or PKA activity.
Disturbances in cross-talk mechanisms could also be
involved. For example, PKC induces b-AR sequestration
and desensitization (Kassis et al 1985, Pitcher et al 1992;
Sibley et al 1987) and decreases b-AR affinity and
coupling to G protein in S49 lymphoma cells (Bell and
Brunton 1987). Ca21 may also affect PKA activity, either
directly or indirectly (Blumenthal et al 1986; Dobbeling
and Berchtold 1996). Thus, disturbances in Ca21 homeostasis in BD could readily affect b-AR-mediated
cAMP signaling at the receptor level through the intricate
framework of cross talk that exists between these signaling
systems.
NaF stimulation of cAMP formation was used to probe
the integrity of G protein function, as in preliminary
experiments NaF stimulated cAMP formation in B lymphoblasts more robustly than GTPgS. The latter GTP
analogue also acts directly on G protein a subunits
stimulating their coupling to effector enzymes (Yamanaka
Ca21 Homeostasis and cAMP Signaling in BD
et al 1986). Our NaF stimulation studies, although not
definitive, show some potentially important trends. Although there was no main effect of diagnosis in the
NaF-induced responses, there was a significant but unexpected interaction with gender, as well as an interaction
between increasing concentrations of NaF and the diagnosis. Male subjects differed from female subjects in their
concentration-response curves for NaF-stimulated AC activity. For this reason, and because of the small number of
female BP-I patients with normal [Ca21]B (n 5 3), only
male subjects were considered in further statistical analyses. Male BP-I patients with high [Ca21]B showed higher
basal (142%) and NaF-stimulated AC activity (81%–94%)
compared with BP-I patients with normal [Ca21]B. NaFstimulated AC activity was also higher (73%– 84%) in
male BP-I patients with high [Ca21]B compared with
healthy subjects, although these differences did not reach
statistical significance, possibly because of high variability
and small sample size; however, the greater NaF-stimulated AC activity in BP-I patients with high [Ca21]B likely
reflects higher basal AC activity in this group, as there
were no significant differences in percent stimulation over
basal levels among the comparison groups.
In contrast to the significantly higher basal cAMP
formation in the B lymphoblast membrane preparations
from male BP-I patients with high [Ca21]B compared to
other comparison groups, no difference was observed in
basal cAMP formation in intact B lymphoblasts. This
discrepancy is most readily explained by the different
cellular matrices and assay conditions used: the NaFstimulation employed a membrane preparation, whereas
intact cells were used in the isoproterenol studies. Basal
and stimulated responses in intact cells were assayed in the
presence of Ca21 (i.e., at physiological extracellular Ca21
concentrations), whereas the membrane assay used
EGTA-washed membranes, and was therefore Ca21 free.
The differences in basal cAMP formation in membranes
unmasked by removing Ca21 suggest altered regulation of
one or more Ca21-dependent processes that influence
cAMP formation. In effect, basal cAMP formation appears
to be “locked” at a higher baseline in membranes from BD
compared with healthy male subjects and is not further
upregulated in the presence of Ca21, as may occur in B
lymphoblast membranes from healthy subjects. Because
cAMP formation was measured following phosphodiesterase inhibition with IBMX, it is most likely that the posited
dysregulation of cAMP formation in B lymphoblast membranes from BP-I patients with high [Ca21]B involves the
AC subtypes expressed in these cells and/or factors modulating their activity. Multiple AC isotypes exist that differ
in their activation and inhibition by G protein a and bg
subunits, Ca21, PKC, and PKA (Sunahara et al 1996).
Therefore, higher basal AC activity in B lymphoblasts
BIOL PSYCHIATRY
2000;48:665– 673
671
from BP-I patients with high [Ca21]B than in the other
comparison groups might reflect differences in catalytic
activity or levels of AC isotype (Pieroni et al 1995),
particularly because G protein a subunit levels did not
differ; however, the specific AC isotypes expressed in
human B lymphoblasts have not been characterized.
The finding of higher basal AC activity in male BP-I
patients with high [Ca21]B may be important pathophysiologically. In addition to the factors discussed above, the
levels and activity of phosphodiesterase isozymes (Houslay and Milligan 1997) and the extent and type of cross
talk (Cooper et al 1995) also contribute to determining
basal cAMP levels. If basal cAMP is near the threshold for
activating PKA, a small increase in cAMP either from
increased AC activity or reduced phosphodiesterase activity will have a large and rapid effect (Houslay and
Milligan 1997). In contrast, when basal AC activity is low,
increased AC activity may be the predominant means of
triggering PKA responses. Thus, the predicted physiological consequences of modulating PDE and/or AC activity
can differ.
In summary, the reduced isoproterenol stimulation and
elevated basal AC activity in B lymphoblasts from BP-I
patients with high [Ca21]B suggests a link between disturbances in calcium homeostasis and processes regulating
b-AR sensitivity and G protein-coupled AC functionality
in a subgroup of patients with BD. Further replications are
warranted, however, to ensure the replicability of these
preliminary findings given the small sample size. Which
signaling system (Ca21 or cAMP) is first affected, and to
what extent the changes in the linked transduction cascade
are secondary or adaptive responses, remain to be elucidated. Regardless, these disturbances in signal transduction systems, which are expressed in transformed B
lymphoblasts from BP-I patients, appear to represent
trait-dependent abnormalities that distinguish a pathophysiologically distinct subgroup of BP-I disorder.
Supported by Medical Council of Canada Grant No. MT12500 (JJW).
The contributions of the following individuals are greatly appreciated:
Dr. Jane Mitchell for her helpful advice about the G protein stimulation
assays; Arvind Kamble for assistance with B lymphoblast transformation, growth of cell lines and cAMP assays; Liliana Mihic and Veronica
Hoang for assistance in subject recruitment and interview; and Kathy
Spegg for advice regarding the statistical analyses.
References
Barritt GJ (1999): Receptor-activated Ca21 inflow in animal
cells: A variety of pathways tailored to meet different
intracellular Ca21 signaling requirements. Biochem J 337:
153–169.
Bell JD, Brunton LL (1987): Multiple effects of phorbol esters on
672
BIOL PSYCHIATRY
2000;48:665– 673
hormone-sensitive adenylate cyclase activity in S49 lymphoma cells. Am J Physiology 252:E783–E789.
Blumenthal DK, Takio K, Hansen RS, Krebs EG (1986):
Dephosphorylation of cAMP-dependent protein kinase regulatory subunit (type II) by calmodulin-dependent protein
phosphatase. Determinants of substrate specificity. J Biol
Chem 261:8140 – 8145.
Bradford M (1976): A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal Biochem 72:248 –252.
Chad T, Kalman D, Armstrong D (1987): The role of cAMPdependent phosphorylation in the maintenance and modulation of voltage-dependent calcium channels. Soc Gen Physiol
Ser 42:167–186.
Cooper DMF, Mons N, Karpen JW (1995): Adenylyl cyclases
and the interaction between calcium and cAMP signaling.
Nature 347:421– 424.
Dobbeling U, Berchtold MW (1996) Down-regulation of the
protein kinase A pathway by activators of protein kinase C
and intracellular Ca21 in fibroblasts cells. FEBS Lett 391:
131–133.
Dubovsky SL, Murphy J, Christiano J, Lee C (1992a): The calcium
second messenger system in bipolar disorders: Data supporting
new research directions. J Neuropsychiatry 4:3–14.
Dubovsky SL, Murphy J, Thomas M, Rademacher BD (1992b):
Abnormal intracellular calcium ion concentration in platelets
and lymphocytes of bipolar patients. Am J Psychiatry 149:
118 –120.
Ebstein RP, Lerer B, Shapira B, Shamesh Z, Moscovich D,
Kindler S (1988): cAMP second messenger signal amplification in depression. Br J Psychiatry 152:665– 669.
Ebstein RP, Steinitz M, Mintzer J, Lipshitz I, Stessman J (1985):
b-Adrenergic-stimulated adenylate cyclase activity in normal
and EBV-transformed lymphocytes. Experientia 41:1552–1554.
Elliott JM (1987): Characterization of [125I]-iodocyanopindolol
binding to cultured human lymphoblast cells. Br J Pharmacol
92:734P.
Emamghoreishi M, Schlichter L, Li PP, Parikh S, Sen J, Kamble
A, Warsh JJ (1997): High intracellular calcium concentrations
in transformed lymphoblasts from subjects with bipolar I
disorder. Am J Psychiatry 154:976 –982.
Extein I, Tallman J, Smith CC, Goodwin FK (1979): Changes in
lymphocyte b-adrenergic receptors in depression and mania.
Psychiatry Res 1:191–197.
First MB, Spitzer RL, Robert L, Gibbon M, Williams JBW
(1995a): Structured Clinical Interview for DSM-IV Axis I
Disorders—Non-patient Edition (SCID-I/NP, Version 2.0).
New York: New York State Psychiatric Institute, Biometric
Research Department.
First MB, Spitzer RL, Robert L, Gibbon M, Williams JBW
(1995b): Structured Clinical Interview for DSM-IV Axis I
Disorders—Patient Edition (SCID-I/P, Version 2.0). New
York: New York State Psychiatric Institute, Biometric Research Department.
Helman J, Kuyatt BL, Takuma T, Seligmann B, Baum BJ (1986):
ATP-dependent calcium transport in rat parotid basolateral
membrane vesicles. Modulation by agents which elevate
cyclic AMP. J Biol Chem 261:8919 – 8923.
Hill SJ, Kendall DA (1989): Cross-talk between different recep-
M. Emamghoreishi et al
tor-effector systems in the mammalian CNS. Cell Signal
1:135–141.
Houslay MD, Milligan G (1997): Tailoring cAMP-signaling
responses through isoform multiplicity. Trends Biochem Sci
22:217–224.
Kandel ER, Castellucci VF, Goelet P, Schacher S (1987):
Cell-biological interrelationship between short-term and
long-term memory. In: Kandel ER, editor. Molecular Neurobiology in Neurology and Psychiatry. New York: Raven,
111–132.
Kassis S, Zaremba T, Patel J, Fishman PH (1985): Phorbol esters
and b-adrenergic agonists mediate desensitization of adenylate cyclase in rat glioma C6 cells by distinct mechanisms.
J Biol Chem 260:8911–9817.
Katada T, Gilman AG, Watanabe Y, Bauer S, Jakobs KH (1985):
Protein kinase C phosphorylates the inhibitory guaninenucleotide binding regulatory component and apparently
suppresses its function in hormonal inhibition of adenylate
cyclase. Eur J Biochem 151:431– 437.
Kay G, Sargeant M, McGuffin P, Whatley S, Marchbanks R,
Baldwin D, et al (1993): The lymphoblast b-adrenergic
receptor in bipolar depressed patients: Characterization and
down-regulation. J Affec Disord 27:163–172.
Kenakin T (1996): The classification of seven transmembrane
receptors in recombinant expression systems. Pharmacol Rev
48:413– 463.
Li PP, Andreopoulos A, Warsh JJ (2000): Signal transduction
abnormalities in bipolar affective disorder. In: Reith MEA,
editor. Cerebral Signal Transduction: From First to Fourth
Messengers. Totowa, NJ: Humana, 283–309.
Liu M, Simon MI (1996): Regulation by cAMP-dependent
protein kinase of a G protein-mediated phospholipase C.
Nature 382:83– 87.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951): Protein
measurement with the folin phenol reagent. J Biol Chem
193:265–275.
Manji HK (1992): G-proteins—implications for psychiatry. Am J
Psychiatry 149:746 –760.
Manji HK, Chen G, Shimon H, Hsiao JK, Potter WZ, Belmaker
RH (1995): Guanine nucleotide binding proteins in bipolar
affective disorder: Effects of long-term lithium treatment.
Arch Gen Psychiatry 52:135–144.
Mann JJ, Brown RP, Halper JP, Sweeney JA, Kocsis JH, Stokes
PE, Bilezikian JP (1985): Reduced sensitivity of lymphocyte
b-adrenergic receptors in patients with endogenous depression and psychomotor agitation. N Engl J Med 313:715–720.
Pandey GN, Dysken MW, Garver DL, Davis J (1979): Betaadrenergic receptor function in affective illness. Am J Psychiatry 136:675– 678.
Pieroni JP, Harry A, Chen J, Jacobowitz O, Magnusson RP,
Iyengar R (1995): Distinct characteristics of the basal activities of adenylyl cyclases 2 and 6. J Biol Chem 270:21368 –
21373.
Pitcher J, Lohse MJ, Codina J, Caron MG, Lefkowitz RJ (1992):
Desensitization of the isolated b2-adrenergic receptor by
b-adrenergic receptor kinase, cAMP-dependent protein kinase and protein kinase C occurs via distinct molecular
mechanisms. Biochemistry 31:3193–3197.
Rasmussen H (1986): The calcium messenger system. N Engl
Med 314:1094 –1101.
Ca21 Homeostasis and cAMP Signaling in BD
Sibley DR, Benovic JL, Caron MG, Lefkowitz RJ (1987):
Molecular mechanisms of b-adrenergic receptor desensitization. Adv Exp Med Biol 221:253–273.
Sugden D, Klein DC (1988): Activators of protein kinase C act
at a postreceptor site to amplify cAMP production in rat
pinealocytes. J Neurochem 50:149 –155.
Sunahara RK, Dessauer CW, Gilman AG (1996): Complexity
and diversity of mammalian adenylyl cyclases. Annu Rev
Pharmacol Toxicol 36:461– 480.
Supattapone S, Danoff SK, Theibert A, Joseph SK, Steiner J,
Snyder SH (1988): Cyclic AMP-dependent phosphorylation
of a brain inositol trisphosphate receptor decreases its release
of calcium. Proc Natl Acad Sci U S A 85:8747– 8750.
Warsh JJ, Emamghoreishi E, Li PP (1999): Elevated B lymphoblast
resting intracellular calcium is a trait-dependent abnormality
specific to bipolar-I disorder. Bipolar Disord 1(suppl 1):57.
BIOL PSYCHIATRY
2000;48:665– 673
673
Warsh JJ, Li PP (1996): Second messenger systems and mood
disorders. Curr Opin Psychiatry 9:23–29.
Yamanaka G, Eckstein F, Stryer L (1986): Interaction of retinal
transducin with guanine triphosphate analogues: Specificity
of the g-phosphate binding region. Biochemistry 25:6149 –
6153.
Yoshimasa T, Sibley DR, Bouvier M, Lefkowitz RJ, Caron MG
(1987): Cross-talk between cellular signalling pathways suggested by phorbol ester induced adenylate cyclase phosphorylation. Nature 327:67–70.
Young LT, Li PP, Kamble A, Siu KP, Warsh JJ (1994):
Increased mononuclear leukocyte G proteins in bipolar but
not unipolar depression. Am J Psychiatry 151:594 –596.
Young LT, Li PP, Kish SJ, Siu KP, Kamble A, Hornykiewicz O,
Warsh JJ (1993): Cerebral cortex Gas protein levels and
forskolin-stimulated cAMP formation are increased in bipolar
affective disorder. J Neurochem 61:890 – 898.
G Protein–Mediated cAMP Signaling in
Bipolar I Disorder
Masoumeh Emamghoreishi, Peter P. Li, Lyanne Schlichter, Sagar V. Parikh,
Robert Cooke, and Jerry J. Warsh
Background: Evidence of extensive cross-talk between
calcium (Ca21)- and cAMP-mediated signaling systems
suggests that previously reported abnormalities in Ca21
homeostasis in bipolar I (BP-I) patients may be linked to
disturbances in the function of G proteins that mediate
cAMP signaling.
Methods: To test this hypothesis, the b-adrenergic agonist, isoproterenol, and the G protein activator, sodium
fluoride (NaF), were used to stimulate cAMP production
in B lymphoblasts from healthy and BP-I subjects phenotyped on basal intracellular calcium concentration
([Ca21]B). cAMP was measured by radioimmunoassay
and [Ca21]B by ratiometric fluorometry with fura-2.
Results: Isoproterenol- (10 mM) stimulated cAMP formation was lower in intact B lymphoblasts from BP-I patients
with high [Ca21]B ($ 2 SD above the mean concentration
of healthy subjects) compared with patients having normal
B lymphoblast [Ca21]B and with healthy subjects. Although basal and NaF-stimulated cAMP production was
greater in B lymphoblast membranes from male BP-I
patients with high versus normal [Ca21]B, there were no
differences in the percent stimulation. This suggests the
differences in NaF response resulted from higher basal
adenylyl cyclase activity.
Conclusions: These findings suggest that trait-dependent
disturbances in processes regulating b-adrenergic receptor sensitivity and G protein–mediated cAMP signaling
occur in conjunction with altered Ca21 homeostasis in
those BP-I patients with high B lymphoblast [Ca21]B.
Biol Psychiatry 2000;48:665– 673 © 2000 Society of Biological Psychiatry
From the Section of Biochemical Psychiatry (ME, PPL, JJW) and Mood Disorder
Program (SVP, RC, JJW), Centre for Addiction and Mental Health, Clarke Site;
Playfair Neuroscience Unit, The Toronto Hospital (LS); and the Departments of
Psychiatry (PPL, SVP, RC, JJW), Pharmacology (PPL, JJW), and Physiology
(LS) and the Institute for Medical Sciences (JJW), University of Toronto,
Toronto, Canada.
Address reprint requests to J.J. Warsh, M.D., Ph.D., University of Toronto, Centre
for Addiction and Mental Health, Clarke Division, 250 College Street, Toronto
Ontario M5T 1R8, Canada.
Received June 29, 1999; revised March 16, 2000; accepted March 20, 2000.
© 2000 Society of Biological Psychiatry
Key Words: Bipolar disorder, G proteins, signal transduction, cAMP, calcium homeostasis, B lymphoblasts
Introduction
E
vidence from studies of postmortem brain and peripheral blood cells implicate disturbances in G protein–
mediated cAMP signaling and Ca21 homeostasis in the
pathophysiology of bipolar affective disorder (BD; reviewed in Li et al 2000; Manji 1992; Warsh and Li 1996).
For instance, increased stimulatory G protein a subunit
(Gas) levels (Manji et al 1995; Young et al 1994), blunted
receptor- and G protein–stimulated cAMP accumulation
(Ebstein et al 1988; Mann et al 1985; Pandey et al 1979)
and increased basal intracellular calcium levels ([Ca21]B;
Dubovsky et al 1992b; Emamghoreishi et al 1997) have all
been observed in mononuclear leukocytes from patients
with BD. Moreover, some of these signaling disturbances
appear to be trait dependent, possibly reflecting underlying
molecular “vulnerability” factors in BD (Emamghoreishi
et al 1997; Warsh et al 1999). Although it has been
speculated that these signal transduction abnormalities are
linked through cross-talk mechanisms (Dubovsky et al
1992a; Warsh and Li 1996), to our knowledge there is no
empirical evidence for such a link.
Extensive cell-type specific cross-talk occurs between
G protein– coupled cAMP- and phospholipase C (PLC)–
linked second-messenger systems (Hill and Kendall 1989;
Liu and Simon 1996). Multiple adenylyl cyclase (AC)
subtypes provide the potential for Ca21 either to synergise
or attenuate cAMP signaling (Sunahara et al 1996). Thus,
Ca21/CaM directly activates AC isotypes I and VIII,
whereas Ca21 inhibits isotypes V and VI (Sunahara et al
1996). Ca21-activated protein kinase C (PKC) may also
increase or decrease cAMP levels by phosphorylating
b-adrenergic receptors (b-ARs; Bell and Brunton 1987),
Gsa, Gia (Katada et al 1985; Sugden and Klein 1988;
Yoshimasa et al 1987) or some AC subtypes (Cooper et al
1995). Recently, it has been shown that Ca21-activated
0006-3223/00/$20.00
PII S0006-3223(00)00884-2
666
M. Emamghoreishi et al
BIOL PSYCHIATRY
2000;48:665– 673
Table 1. Subject Characteristics
Age (years)
Gender (female/male)
Lifetime psychiatric
comorbidityb
B lymphoblast [Ca21]B
Healthy
subjects
BP-I
high [Ca21]B
BP-I
normal [Ca21]B
38.6 6 3.60a
7/4
NA
37.7 6 4.00
6/5
1—Panic disorder,
2—alcohol dependence
74.8 6 1.70c
39.8 6 4.27
3/8
5—Alcohol abuse or
dependence
57.5 6 1.03
53.2 6 1.75
a
Mean 6 SEM.
BP patients with a lifetime history of alcohol abuse or dependence had no recent history (.3 months) of substance abuse at
the time of study. One BP-I subject with normal [Ca21]B also had a past, but not recent, history of polysubstance abuse.
c
Tukey p ,.001, compared with healthy subjects and BP-I patients with normal [Ca21]B.
b
PKC suppresses protein kinase A (PKA)-mediated signaling in fibroblast cell lines (Dobbeling and Berchtold
1996), providing another site of interaction between these
signaling pathways.
cAMP-mediated processes also affect Ca21 signaling
by inhibiting Ca21 influx (Rasmussen 1986), desensitizing
inositol trisphosphate receptors (Supattapone et al 1988),
stimulating Ca21 removal from the cytosol by Ca21
ATPase (Helman et al 1986), and inhibiting PLCb2 (Liu
and Simon 1996). In other instances, G protein–mediated
cAMP signaling enhances that of Ca21, including PKAmediated phosphorylation of voltage-dependent Ca21
channels in excitable cells (Chad et al 1987), or the ion
gate that closes K1 channels in cells in general, thus
prolonging depolarization and enhancing Ca21 influx
(Kandel et al 1987). There is also evidence that in
non-excitable cells, receptor-activated Ca21 channels are
opened by an increase in the concentration of cytoplasmic
cAMP (Barritt 1999). Such mechanisms as noted above
would account for the types of interaction (i.e., stimulation/suppression) that occur between cAMP and phosphoinositide signaling pathways.
As disturbances in both G protein–mediated cAMP and
Ca21 signaling have been previously reported in BD,
cross-talk regulatory mechanisms are potentially important candidates to be considered in explaining abnormalities in these two signal transduction pathways. Thus, the
increased lymphocyte and platelet intracellular Ca21
([Ca21]i) reported in BD might result from the altered G
protein function previously identified in these patients, or
vice versa. In effect, defects in one signaling system may
decompensate or, alternatively, induce adaptive changes in
other signaling systems in an attempt to maintain
homeostasis.
To test this hypothesis, we used Epstein–Barr virus
(EBV)–immortalized B lymphoblasts derived from BD
patients, a cellular model that demonstrates trait-dependent abnormalities in Ca21 homeostasis (Emamghoreishi
et al 1997) and in b-AR sensitivity (Kay et al 1993) in this
disorder. These cell lines can be grown in continuous
culture removed from in vivo neurohormonal influences
and medications, which circumvents the potential of such
variables to hamper detection of signal transduction abnormalities and interactions between signaling cascades.
To screen for possible relationships between disturbed G
protein function and altered basal intracellular [Ca21]
([Ca21]B), isoproterenol and sodium fluoride (NaF)-stimulated cAMP formation were determined in B lymphoblasts from bipolar I (BP-I) patients stratified based on
[Ca21]B compared with healthy subjects. We report here
the coexistence of altered receptor- and G protein–mediated cAMP signaling and calcium homeostasis, supporting
the view that these abnormalities are both interrelated and
trait-dependent.
Methods and Materials
Subjects
Cell lines were selected from a subset of BP-I patients and
healthy subjects recruited and assessed as previously described
(Emamghoreishi et al 1997). Briefly, subjects were recruited
from the Mood Disorders inpatient and outpatient units of the
Clarke Institute of Psychiatry. After a clinical diagnosis of BP-I
disorder was made by a consulting psychiatrist, the diagnosis was
confirmed for research purposes using the Structured Clinical
Interview for DSM-IV Axis I Disorders (SCID I/P, version 2.0;
First et al 1995b) administered by a research psychiatrist or
trained psychiatric research assistant, complemented by review
of medical records. All interview and historical material was
further reviewed by a research psychiatrist to ensure the reliability of the diagnoses. Patients had no history of recent (.3
months) substance abuse. Healthy comparison subjects were
recruited through posted advertisements in the University of
Toronto and were assessed by a trained research assistant and a
research psychiatrist using the SCID-I/NP, version 2.0 (First et al
1995a). Healthy subjects had no past or family (first-degree
relatives) history of psychiatric disorder. All subjects were
physically healthy, had no past history of serious medical illness,
and had no personal or family history of diabetes or hypertension. The study was approved by the Human Subjects Review
Committee of the University of Toronto and, after providing a
complete description of the study, all subjects provided written
informed consent. Table 1 shows the demographics and [Ca21]B
Ca21 Homeostasis and cAMP Signaling in BD
of the BP-I patients and healthy subjects from whom cell lines
were selected for study.
B Lymphoblast Transformation and Intracellular
Calcium Measurements
Mononuclear leukocyte preparations were isolated by density
gradient separation through Ficoll-Hypaque and B lymphocytes
in these isolates were transformed by infection with EBV as
previously described (Emamghoreishi et al 1997). The transformed, immortalized B lymphoblasts were grown in RPMI1640 medium containing L-glutamine (2 mmol/L), pyruvate (1
mmol/L), 20% fetal bovine serum (FBS), 100 mg/mL streptomycin, 100 units/mL penicillin in an incubator (95% air/5% CO2 at
37°C) with passages every 3– 4 days. At 12–15 passages, aliquots
of cells were removed and [Ca21]B determined as described
below. The remaining cells were frozen in aliquots (107 cells/
mL) and stored over the vapour phase of liquid nitrogen until
they were regrown and assayed for receptor- and G protein–
stimulated cAMP formation.
[Ca21]B was measured in B lymphoblasts using ratiometric
fluorometry and the calcium sensitive dye, fura 2-acetoxymethylester, as described previously (Emamghoreishi et al 1997).
BP-I subjects were stratified based on [Ca21]B greater than
(high) or less than (normal) two standard deviations above the
mean [Ca21]B determined in cells from healthy subjects, to
classify patients into these two phenotypes.
Regeneration of B lymphoblast Cell Lines
B lymphoblasts were removed from the liquid nitrogen freezer
and immediately placed on dry ice. Cells were then thawed
quickly (,5 min) at 37°C and transferred gently into 25-cm2
culture flasks containing 5 mL pre-warmed (37°C) B lymphoblast medium and regrown as previously described (Emamghoreishi et al 1997). After nine passages from the time of regeneration
(total passages 21–24), cells were transferred to B lymphoblast
medium containing 10% FBS and incubated for three additional
passages. Finally, cells were transferred to B lymphoblast medium without FBS and incubated for 24 hours, then harvested
and divided in two aliquots. One aliquot (8 –10 3 107 cells) was
used immediately to assay isoproterenol-stimulated cAMP formation, and the other was stored at 270°C until used to assay G
protein-stimulated AC activity.
Determination of Basal and IsoproterenolStimulated cAMP Production in Intact
B Lymphoblasts
Isoproterenol was used to assess b-adrenergic receptor-stimulated cAMP formation in intact B lymphoblasts using a method
modified from Kay et al (1993). Cells were washed once in
buffer containing 130 mmol/L NaCl, 0.54 mmol/L KCl, 0.005
mmol/L CaCl2, 0.098 mmol/L MgCl2, 0.1% glucose and 15
mmol/L Tris.HCl pH 7.6, resuspended at 3 3 106 cells/mL, and
incubated in 0.9 mL aliquots (30 min, 37°C, 95% air/5% CO2).
The reactions were begun by adding 50 mL isobutyl-methylxanthine (IBMX, final concentration 0.5 mmol/L), 50 ml ascorbic
BIOL PSYCHIATRY
2000;48:665– 673
667
acid (final concentration 0.1 mmol/L) with or without isoproterenol, and incubating at 37°C for 10 min. Reactions were stopped
by boiling for 10 min, then centrifuging at 12,000 g for 10 min
at 4°C. The protein concentration in each pellet was determined
by the Lowry method (Lowry et al 1951). The supernatants,
which contained the cAMP, were evaporated to dryness under
vacuum, stored at 270°C until needed, then reconstituted with
100 mL assay buffer containing 50 mmol/L sodium acetate at pH
6.2, and 0.01% sodium azide. The cAMP content was measured
using an 125I-radioimmunoassay kit following the supplier’s
instructions (Mandel, Guelph, Canada). Maximally stimulating
and saturating concentrations of isoproterenol (1 and 10 mmol/L,
respectively) were determined in preliminary dose-response
studies, and subsequently used to assess B lymphoblast receptorstimulated cAMP production in the comparison groups.
Determination of Basal and NaF-Stimulated AC
Activity in B Lymphoblast Membranes
MEMBRANE PREPARATION. B lymphoblasts were thawed
on ice, resuspended in 20 volumes ice-cold buffer (10 mmol/L
Tris, pH 7.4, 1 mmol/L EGTA) and homogenized by hand using
a glass-teflon homogenizer (15 strokes). The homogenates were
centrifuged at 20,000 g (20 min, 4°C), then the pellets were
re-suspended and centrifuged twice more. Each pellet was
resuspended in 500 mL ice-cold TEMS buffer (50 mmol/L Tris,
pH 7.4, 1 mmol/L EGTA, 5 mmol/L MgCl2, 12% sucrose), and
protein concentrations were determined by the method of Bradford (Bradford 1976), using bovine serum albumin as a standard.
ASSAY OF NAF-STIMULATED AC ACTIVITY. Membrane
homogenates were diluted with TEMS buffer to give 1 mg
membrane protein/ml and the samples kept on ice until use.
Reaction mixtures (200 mL) contained 50 mmol/L Tris, pH 7.4,
1 mmol/L EGTA, 3.5 mmol/L MgCl2, 0.1 mmol/L IBMX, 1.5
mg/mL BSA, 0.5 mmol/L adenosine triphosphate (ATP), 25
mmol/L phosphocreatine, and 250 units/mL creatine phosphokinase. Membrane homogenates (50 mL) were incubated at 37°C
for 15 min in 100 mL of reaction mixture. The reactions were
initiated by adding 50 mL ATP (2 mmol/L) in the absence or
presence of NaF (2.5, 5, and 10 mmol/L), followed by gently
vortexing and incubation at 37°C for 10 min with gentle shaking.
The reaction was terminated by boiling the tubes for 10 min,
followed by cooling on ice and centrifuging at 12,000 g for 10
min at 4°C. The supernatants were then removed and stored at
270°C until used for cAMP measurements as described above.
Statistical Analysis
Differences in dependent variables (basal and stimulated cAMP
formation) were assessed by repeated-measures multivariate
analysis of variance (MANOVA). Concentration (basal and
various concentrations of isoproterenol or NaF) was the repeated
“within” factor and diagnosis was the “between” subject factor.
NaF-stimulated cAMP formation values were logarithmically
transformed to normalize the distribution of the data. Based on
previous studies, we tested directional hypotheses (i.e., of reduced responsivity to isoproterenol stimulation [Extein et al
668
M. Emamghoreishi et al
BIOL PSYCHIATRY
2000;48:665– 673
Table 2. Basal and Isoproterenol-Stimulated cAMP Production in B Lymphoblasts from Healthy
Subjects and Bipolar I Patients with High (.69.4 nmol/L) and Normal (,69.4 nmol/L) [Ca21]B
Bipolar I disorder
Basal
Isoproterenol (1 mmol/L)
% stimulation
Isoproterenol (10 mmol/L)
% stimulation
21
Healthy
High [Ca ]B
Normal [Ca21]B
9.24 6 1.28
16.9 6 2.55
82.9 6 12.5
16.6 6 3.00
79.4 6 19.4
10.7 6 1.71
16.8 6 4.24
52.3 6 15.7
14.7 6 3.83
33.5 6 15.2a,b
12.0 6 1.62
20.5 6 3.08
69.0 6 12.6
21.6 6 3.42
81.8 6 17.6
cAMP production (see Methods) is expressed as pmol/mg protein/10 min (mean 6 SEM, 11 subjects in each group). The %
stimulation over the basal level is shown.
a
Significantly different (t 5 2.0, p 5 .025) from stimulation by 1 mmol/L isoproterenol in the same patient group.
b
Significantly different from BP-I subjects with normal [Ca21]B (t 5 1.95, p 5 .03) and from healthy subjects (t 5 1.85,
p 5 .035).
1979; Pandey et al 1979] and increased NaF-stimulated cAMP
formation [Young et al 1993] in B lymphoblasts from BP-I
patients). Thus, one-tailed probability values of p ,.05 were
used to distinguish statistically significant differences. Statistical
analyses were performed using the SPSS (release 7.0) statistical
software package (SPSS Inc., Chicago).
Results
Basal and Isoproterenol-Stimulated cAMP
Production in Intact B Lymphoblasts
Repeated-measures analysis of variance (ANOVA) performed on the absolute values for cAMP formation
showed a significant effect of isoproterenol [F(2,60) 5
26.1, p ,.001], but no significant interaction between
diagnosis and isoproterenol [F(4,60) 5 1.0, p 5 .2] or
main effect of diagnosis [F(2,30) 5 0.71, p 5 .25].
When the isoproterenol-stimulated AC activity was expressed as the percent stimulation above basal, there was
a significant interaction between diagnosis and the percent
stimulation [F(4,60) 5 2.04, p 5 .05] and a trend
toward a significant main effect of diagnosis [F(2,30) 5
1.92, p 5 .08]. Post hoc comparison of cell means by
contrasts showed a significant reduction in the percent
stimulation above basal at 10 mmol/L isoproterenol in
BP-I patients with high [Ca21]B compared with normal
[Ca21]B, and compared with healthy subjects (59%, t 5
1.95, p 5 .03 and 58%, t 5 1.85, p 5 .035,
respectively; Table 2). The increase in absolute cAMP
levels was 17% lower than that achieved with 1 mmol/L
isoproterenol stimulation in B lymphoblasts from BP-I
patients with high [Ca21]B, but was not statistically
significant compared with the cAMP levels reached following stimulation of B lymphoblasts from healthy subjects with 10 mmol/L isoproterenol. Reduced percent
stimulation above basal was also observed with 1 mmol/L
isoproterenol in BP-I patients with high [Ca21]B compared
with healthy subjects, although this difference did not
reach statistical significance (37%, t 5 1.58, p 5 .065).
Additionally, the reduced isoproterenol-stimulated cAMP
production in BP-I subjects with high [Ca21]B was more
marked at 10 mmol/L and was significantly different from
that determined at 1 mmol/L isoproterenol (t 5 2.0, p 5
.025). In preliminary experiments using membrane preparations, isoproterenol-stimulation was very low and variable, as we have also observed for other tissues (Young et
al 1993). Thus, there was no further attempt to measure
isoproterenol-stimulated cAMP formation in membranes
from the subject samples studied.
Basal and NaF-Stimulated AC Activity in B
Lymphoblast Membranes
NaF increased cAMP production in a concentration-dependent manner, with maximal stimulation at 5 mmol/L
and a reduced response at 10 mmol/L NaF (Figure 1A). A
repeated-measures MANOVA test showed significant
main effects of NaF concentration [F(3,72) 5 23.1, p
,.001] and gender [F(1,24) 5 4.5, p 5 .02], but not
of diagnosis [F(2,24) 5 1.5, p 5 .12]. Moreover,
statistically significant interactions were found between
sex and NaF concentration [F(3,72) 5 6.5, p 5 .001],
and between diagnosis and NaF [F(6,72) 5 1.95, p 5
.04]. Simple effects were performed to examine the
interactions. Contrasts of cell means revealed that male
subjects were significantly different from female subjects
in the magnitude of the response to 2.5 mmol/L NaF
[F(1,24) 5 21, p ,.001] and in the reduced response
to 10 mmol/L versus 5 mmol/L NaF [F(1,24) 5 4.96,
p 5 .02) (Figure 1, C and D).
When cAMP levels were expressed as percent change
above the basal level (Figure 1B), repeated-measures
MANOVA tests also showed a significant effect of the
NaF concentration [F(2,46) 5 33.9, p ,.001], an
interaction between gender and NaF concentration
[F(2,46) 5 4.0, p 5 .01], and a significant interaction
between NaF concentration and diagnosis [F(4,46) 5
2.1, p 5 .046]. Male subjects differed significantly in
Ca21 Homeostasis and cAMP Signaling in BD
BIOL PSYCHIATRY
2000;48:665– 673
669
Figure 1. Sodium fluoride (NaF)–stimulated adenylyl cyclase (AC) activity production: dose–response curves for B lymphoblast
membranes from bipolar I (BP-I) patients with high [Ca21]B (.69.4 nmol/L, E), normal [Ca21]B (,69.4 nmol/L, ‚), and their ageand gender-matched healthy controls (h). (A, B) Males and female subjects combined (n 5 10). (C–F) Results stratified by gender
(n 5 4 – 8 male and 2– 6 female subjects per subject comparison group). Repeated-measures multivariate analysis of variance
(MANOVA) tests revealed no significant differences in basal or NaF-stimulated AC activity among comparison groups (A and B). For
male subjects (C), repeated-measures MANOVA tests showed a significant main effect of diagnosis [F(2,14) 5 3.5, p 5 .028].
*Significant differences (p #.05) compared with corresponding NaF-stimulated response in the male BP-I group with normal
[Ca21]B.
the extent of reduction in response to 10 mmol/L compared to 5 mmol/L NaF [F(1,23) 5 6.2, p 5.01]
(Figure 1 E and F).
Figure 1 (C–F) shows NaF-stimulated cAMP production expressed in absolute values and as a percent above
basal cAMP production, in the comparison groups stratified by gender. Because the number of female BP-I
subjects with normal [Ca21]B was small, further statistical
analysis of data was performed only on male bipolar
subjects. Repeated-measures MANOVA showed significant effects of NaF concentration [F(3,42) 5 48.8, p
,.001], an interaction between NaF concentration and
diagnosis [F(6,42) 5 2.4, p 5 .02], and a significant
main effect of diagnosis [F(2,14) 5 3.5, p 5 .028;
Figure 1C]. Male BP-I patients with high [Ca21]B had
significantly higher absolute values of basal cAMP (142%
greater, p , .05) and NaF-stimulated cAMP production
(86% greater at 2.5 mmol/L NaF, p 5 .05; 94% at 5
mmol/L, p ,.05; 81% at 10 mmol/L, p ,.05) compared
with male BP-I patients with normal [Ca21]B. Their cAMP
values were also higher than healthy male subjects, but the
differences were not statistically significant (84% at 2.5
mmol/L, p 5 .09; 73% at 5 mmol/L, p 5 .08). In
contrast, analysis of the NaF-stimulated cAMP levels in B
670
BIOL PSYCHIATRY
2000;48:665– 673
lymphoblast membranes from male subjects expressed as
a percent increase above their respective basal activities
(Figure 1E) showed only a significant effect of NaF
concentration [F(2,28) 5 35.4, p ,.001].
Discussion
We have used B lymphoblasts from BP-I patients that
display high [Ca21]B to test whether alterations in calcium
homeostasis are related to disturbances in receptor-G
protein coupled cAMP formation in BP-I patients manifesting this cellular endophenotype. The principal findings
are reduced b-AR-stimulated cAMP formation in intact B
lymphoblasts, and elevated basal G protein/effector-dependent cAMP production in membranes of B lymphoblasts from BP-I patients with high [Ca21]B compared
with healthy subjects. Specifically, the b-AR response to
10 mmol/L isoproterenol was significantly lower in BP-I
patients with high [Ca21]B than in healthy subjects, and
although the response to 1 mmol/L isoproterenol was also
lower, it was not statistically significant. Male BP-I
patients with elevated B lymphoblast [Ca21]B also had
significantly higher basal and NaF-stimulated AC activity
compared with male BP-I subjects with normal [Ca21]B
and a trend towards higher NaF-stimulated AC activity
compared with healthy subjects. These findings suggest
that the alterations in intracellular calcium homeostasis in
B lymphoblasts from BP-I patients are indeed associated
with disturbances in receptor- and G protein– coupled AC
activity. Moreover, the co-expression of these disturbances in cells from BP-I subjects implies that this
spectrum of signal transduction abnormalities is trait
dependent.
Maximally stimulating and saturating concentrations of
isoproterenol were used to assess b-AR-stimulated cAMP
production in B lymphoblasts because of the limited
signal-to-noise ratio of the response and to ensure a
maximal level of cAMP stimulation was achieved in all
cases in the absence of full dose response studies on each
cell line. b-ARs in transformed B lymphoblasts show
qualitatively similar ligand binding characteristics and AC
activation to those in B lymphocytes, but lower density
and maximal enzyme activity (Ebstein et al 1985; Elliot
1987). Furthermore, because the functional interaction
between receptors and G proteins is strongly influenced by
their relative expression levels (Kenakin 1996), the small
responses observed here likely reflect the low density of
b-ARs in these cell lines. Dose-response studies were not
performed, and inferences about the efficacy of receptor-G
protein coupling (reflected in the EC50) and exact nature of
the altered responsivity will require future concentrationresponse characterization.
The small magnitude and inter-sample variability in the
M. Emamghoreishi et al
isoproterenol-stimulated cAMP production at 1 mmol/L
may have contributed to the lack of statistically significant
differences in B lymphoblast b-AR responses, between
BP-I patients with high [Ca21]B and either patients with
normal [Ca21]B or healthy subjects. Nevertheless, our
results highlight the possibility that factors regulating the
responsivity of the b-AR-G protein-AC complex are
differentially altered in BP-I patients who manifest the
endophenotype of high B lymphoblast [Ca21]B. Although
b-AR density was not determined in this study, previous
studies have shown reduced b-AR-stimulated AC activity
in lymphocytes (Mann et al 1985) and B lymphoblasts
(Kay et al 1993) from BD patients in the absence of
changes in b-AR density or affinity. Furthermore, we have
not found differences in Gas and Gai levels in B lymphoblasts from BP-I patients compared with their matched
healthy subjects (unpublished data). These observations
argue against the possibility that altered levels of b-AR
and of these G protein a subunits account for the reduced
b-AR-stimulated response in B lymphoblasts from BP-I
patients observed here. Other downstream processes that
regulate the response of the receptor-G protein-AC transduction apparatus should be considered, and it is notable
that Kay et al (1993) found less b-AR down-regulation in
response to isoproterenol in B lymphoblasts from BD
patients than in cells from control subjects. These observations suggested the process(es) regulating b-AR desensitization and/or down-regulation may be compromised in
BD. b-AR down-regulation and/or desensitization is mediated by receptor phosphorylation through PKA and/or
b-AR kinase (Sibley et al 1987). Therefore, the lower
b-AR responsiveness in BP-I patients with high [Ca21]B,
compared with healthy subjects and BP-I patients with
normal [Ca21]B, could similarly be related to alterations in
b-AR kinase or PKA activity.
Disturbances in cross-talk mechanisms could also be
involved. For example, PKC induces b-AR sequestration
and desensitization (Kassis et al 1985, Pitcher et al 1992;
Sibley et al 1987) and decreases b-AR affinity and
coupling to G protein in S49 lymphoma cells (Bell and
Brunton 1987). Ca21 may also affect PKA activity, either
directly or indirectly (Blumenthal et al 1986; Dobbeling
and Berchtold 1996). Thus, disturbances in Ca21 homeostasis in BD could readily affect b-AR-mediated
cAMP signaling at the receptor level through the intricate
framework of cross talk that exists between these signaling
systems.
NaF stimulation of cAMP formation was used to probe
the integrity of G protein function, as in preliminary
experiments NaF stimulated cAMP formation in B lymphoblasts more robustly than GTPgS. The latter GTP
analogue also acts directly on G protein a subunits
stimulating their coupling to effector enzymes (Yamanaka
Ca21 Homeostasis and cAMP Signaling in BD
et al 1986). Our NaF stimulation studies, although not
definitive, show some potentially important trends. Although there was no main effect of diagnosis in the
NaF-induced responses, there was a significant but unexpected interaction with gender, as well as an interaction
between increasing concentrations of NaF and the diagnosis. Male subjects differed from female subjects in their
concentration-response curves for NaF-stimulated AC activity. For this reason, and because of the small number of
female BP-I patients with normal [Ca21]B (n 5 3), only
male subjects were considered in further statistical analyses. Male BP-I patients with high [Ca21]B showed higher
basal (142%) and NaF-stimulated AC activity (81%–94%)
compared with BP-I patients with normal [Ca21]B. NaFstimulated AC activity was also higher (73%– 84%) in
male BP-I patients with high [Ca21]B compared with
healthy subjects, although these differences did not reach
statistical significance, possibly because of high variability
and small sample size; however, the greater NaF-stimulated AC activity in BP-I patients with high [Ca21]B likely
reflects higher basal AC activity in this group, as there
were no significant differences in percent stimulation over
basal levels among the comparison groups.
In contrast to the significantly higher basal cAMP
formation in the B lymphoblast membrane preparations
from male BP-I patients with high [Ca21]B compared to
other comparison groups, no difference was observed in
basal cAMP formation in intact B lymphoblasts. This
discrepancy is most readily explained by the different
cellular matrices and assay conditions used: the NaFstimulation employed a membrane preparation, whereas
intact cells were used in the isoproterenol studies. Basal
and stimulated responses in intact cells were assayed in the
presence of Ca21 (i.e., at physiological extracellular Ca21
concentrations), whereas the membrane assay used
EGTA-washed membranes, and was therefore Ca21 free.
The differences in basal cAMP formation in membranes
unmasked by removing Ca21 suggest altered regulation of
one or more Ca21-dependent processes that influence
cAMP formation. In effect, basal cAMP formation appears
to be “locked” at a higher baseline in membranes from BD
compared with healthy male subjects and is not further
upregulated in the presence of Ca21, as may occur in B
lymphoblast membranes from healthy subjects. Because
cAMP formation was measured following phosphodiesterase inhibition with IBMX, it is most likely that the posited
dysregulation of cAMP formation in B lymphoblast membranes from BP-I patients with high [Ca21]B involves the
AC subtypes expressed in these cells and/or factors modulating their activity. Multiple AC isotypes exist that differ
in their activation and inhibition by G protein a and bg
subunits, Ca21, PKC, and PKA (Sunahara et al 1996).
Therefore, higher basal AC activity in B lymphoblasts
BIOL PSYCHIATRY
2000;48:665– 673
671
from BP-I patients with high [Ca21]B than in the other
comparison groups might reflect differences in catalytic
activity or levels of AC isotype (Pieroni et al 1995),
particularly because G protein a subunit levels did not
differ; however, the specific AC isotypes expressed in
human B lymphoblasts have not been characterized.
The finding of higher basal AC activity in male BP-I
patients with high [Ca21]B may be important pathophysiologically. In addition to the factors discussed above, the
levels and activity of phosphodiesterase isozymes (Houslay and Milligan 1997) and the extent and type of cross
talk (Cooper et al 1995) also contribute to determining
basal cAMP levels. If basal cAMP is near the threshold for
activating PKA, a small increase in cAMP either from
increased AC activity or reduced phosphodiesterase activity will have a large and rapid effect (Houslay and
Milligan 1997). In contrast, when basal AC activity is low,
increased AC activity may be the predominant means of
triggering PKA responses. Thus, the predicted physiological consequences of modulating PDE and/or AC activity
can differ.
In summary, the reduced isoproterenol stimulation and
elevated basal AC activity in B lymphoblasts from BP-I
patients with high [Ca21]B suggests a link between disturbances in calcium homeostasis and processes regulating
b-AR sensitivity and G protein-coupled AC functionality
in a subgroup of patients with BD. Further replications are
warranted, however, to ensure the replicability of these
preliminary findings given the small sample size. Which
signaling system (Ca21 or cAMP) is first affected, and to
what extent the changes in the linked transduction cascade
are secondary or adaptive responses, remain to be elucidated. Regardless, these disturbances in signal transduction systems, which are expressed in transformed B
lymphoblasts from BP-I patients, appear to represent
trait-dependent abnormalities that distinguish a pathophysiologically distinct subgroup of BP-I disorder.
Supported by Medical Council of Canada Grant No. MT12500 (JJW).
The contributions of the following individuals are greatly appreciated:
Dr. Jane Mitchell for her helpful advice about the G protein stimulation
assays; Arvind Kamble for assistance with B lymphoblast transformation, growth of cell lines and cAMP assays; Liliana Mihic and Veronica
Hoang for assistance in subject recruitment and interview; and Kathy
Spegg for advice regarding the statistical analyses.
References
Barritt GJ (1999): Receptor-activated Ca21 inflow in animal
cells: A variety of pathways tailored to meet different
intracellular Ca21 signaling requirements. Biochem J 337:
153–169.
Bell JD, Brunton LL (1987): Multiple effects of phorbol esters on
672
BIOL PSYCHIATRY
2000;48:665– 673
hormone-sensitive adenylate cyclase activity in S49 lymphoma cells. Am J Physiology 252:E783–E789.
Blumenthal DK, Takio K, Hansen RS, Krebs EG (1986):
Dephosphorylation of cAMP-dependent protein kinase regulatory subunit (type II) by calmodulin-dependent protein
phosphatase. Determinants of substrate specificity. J Biol
Chem 261:8140 – 8145.
Bradford M (1976): A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal Biochem 72:248 –252.
Chad T, Kalman D, Armstrong D (1987): The role of cAMPdependent phosphorylation in the maintenance and modulation of voltage-dependent calcium channels. Soc Gen Physiol
Ser 42:167–186.
Cooper DMF, Mons N, Karpen JW (1995): Adenylyl cyclases
and the interaction between calcium and cAMP signaling.
Nature 347:421– 424.
Dobbeling U, Berchtold MW (1996) Down-regulation of the
protein kinase A pathway by activators of protein kinase C
and intracellular Ca21 in fibroblasts cells. FEBS Lett 391:
131–133.
Dubovsky SL, Murphy J, Christiano J, Lee C (1992a): The calcium
second messenger system in bipolar disorders: Data supporting
new research directions. J Neuropsychiatry 4:3–14.
Dubovsky SL, Murphy J, Thomas M, Rademacher BD (1992b):
Abnormal intracellular calcium ion concentration in platelets
and lymphocytes of bipolar patients. Am J Psychiatry 149:
118 –120.
Ebstein RP, Lerer B, Shapira B, Shamesh Z, Moscovich D,
Kindler S (1988): cAMP second messenger signal amplification in depression. Br J Psychiatry 152:665– 669.
Ebstein RP, Steinitz M, Mintzer J, Lipshitz I, Stessman J (1985):
b-Adrenergic-stimulated adenylate cyclase activity in normal
and EBV-transformed lymphocytes. Experientia 41:1552–1554.
Elliott JM (1987): Characterization of [125I]-iodocyanopindolol
binding to cultured human lymphoblast cells. Br J Pharmacol
92:734P.
Emamghoreishi M, Schlichter L, Li PP, Parikh S, Sen J, Kamble
A, Warsh JJ (1997): High intracellular calcium concentrations
in transformed lymphoblasts from subjects with bipolar I
disorder. Am J Psychiatry 154:976 –982.
Extein I, Tallman J, Smith CC, Goodwin FK (1979): Changes in
lymphocyte b-adrenergic receptors in depression and mania.
Psychiatry Res 1:191–197.
First MB, Spitzer RL, Robert L, Gibbon M, Williams JBW
(1995a): Structured Clinical Interview for DSM-IV Axis I
Disorders—Non-patient Edition (SCID-I/NP, Version 2.0).
New York: New York State Psychiatric Institute, Biometric
Research Department.
First MB, Spitzer RL, Robert L, Gibbon M, Williams JBW
(1995b): Structured Clinical Interview for DSM-IV Axis I
Disorders—Patient Edition (SCID-I/P, Version 2.0). New
York: New York State Psychiatric Institute, Biometric Research Department.
Helman J, Kuyatt BL, Takuma T, Seligmann B, Baum BJ (1986):
ATP-dependent calcium transport in rat parotid basolateral
membrane vesicles. Modulation by agents which elevate
cyclic AMP. J Biol Chem 261:8919 – 8923.
Hill SJ, Kendall DA (1989): Cross-talk between different recep-
M. Emamghoreishi et al
tor-effector systems in the mammalian CNS. Cell Signal
1:135–141.
Houslay MD, Milligan G (1997): Tailoring cAMP-signaling
responses through isoform multiplicity. Trends Biochem Sci
22:217–224.
Kandel ER, Castellucci VF, Goelet P, Schacher S (1987):
Cell-biological interrelationship between short-term and
long-term memory. In: Kandel ER, editor. Molecular Neurobiology in Neurology and Psychiatry. New York: Raven,
111–132.
Kassis S, Zaremba T, Patel J, Fishman PH (1985): Phorbol esters
and b-adrenergic agonists mediate desensitization of adenylate cyclase in rat glioma C6 cells by distinct mechanisms.
J Biol Chem 260:8911–9817.
Katada T, Gilman AG, Watanabe Y, Bauer S, Jakobs KH (1985):
Protein kinase C phosphorylates the inhibitory guaninenucleotide binding regulatory component and apparently
suppresses its function in hormonal inhibition of adenylate
cyclase. Eur J Biochem 151:431– 437.
Kay G, Sargeant M, McGuffin P, Whatley S, Marchbanks R,
Baldwin D, et al (1993): The lymphoblast b-adrenergic
receptor in bipolar depressed patients: Characterization and
down-regulation. J Affec Disord 27:163–172.
Kenakin T (1996): The classification of seven transmembrane
receptors in recombinant expression systems. Pharmacol Rev
48:413– 463.
Li PP, Andreopoulos A, Warsh JJ (2000): Signal transduction
abnormalities in bipolar affective disorder. In: Reith MEA,
editor. Cerebral Signal Transduction: From First to Fourth
Messengers. Totowa, NJ: Humana, 283–309.
Liu M, Simon MI (1996): Regulation by cAMP-dependent
protein kinase of a G protein-mediated phospholipase C.
Nature 382:83– 87.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951): Protein
measurement with the folin phenol reagent. J Biol Chem
193:265–275.
Manji HK (1992): G-proteins—implications for psychiatry. Am J
Psychiatry 149:746 –760.
Manji HK, Chen G, Shimon H, Hsiao JK, Potter WZ, Belmaker
RH (1995): Guanine nucleotide binding proteins in bipolar
affective disorder: Effects of long-term lithium treatment.
Arch Gen Psychiatry 52:135–144.
Mann JJ, Brown RP, Halper JP, Sweeney JA, Kocsis JH, Stokes
PE, Bilezikian JP (1985): Reduced sensitivity of lymphocyte
b-adrenergic receptors in patients with endogenous depression and psychomotor agitation. N Engl J Med 313:715–720.
Pandey GN, Dysken MW, Garver DL, Davis J (1979): Betaadrenergic receptor function in affective illness. Am J Psychiatry 136:675– 678.
Pieroni JP, Harry A, Chen J, Jacobowitz O, Magnusson RP,
Iyengar R (1995): Distinct characteristics of the basal activities of adenylyl cyclases 2 and 6. J Biol Chem 270:21368 –
21373.
Pitcher J, Lohse MJ, Codina J, Caron MG, Lefkowitz RJ (1992):
Desensitization of the isolated b2-adrenergic receptor by
b-adrenergic receptor kinase, cAMP-dependent protein kinase and protein kinase C occurs via distinct molecular
mechanisms. Biochemistry 31:3193–3197.
Rasmussen H (1986): The calcium messenger system. N Engl
Med 314:1094 –1101.
Ca21 Homeostasis and cAMP Signaling in BD
Sibley DR, Benovic JL, Caron MG, Lefkowitz RJ (1987):
Molecular mechanisms of b-adrenergic receptor desensitization. Adv Exp Med Biol 221:253–273.
Sugden D, Klein DC (1988): Activators of protein kinase C act
at a postreceptor site to amplify cAMP production in rat
pinealocytes. J Neurochem 50:149 –155.
Sunahara RK, Dessauer CW, Gilman AG (1996): Complexity
and diversity of mammalian adenylyl cyclases. Annu Rev
Pharmacol Toxicol 36:461– 480.
Supattapone S, Danoff SK, Theibert A, Joseph SK, Steiner J,
Snyder SH (1988): Cyclic AMP-dependent phosphorylation
of a brain inositol trisphosphate receptor decreases its release
of calcium. Proc Natl Acad Sci U S A 85:8747– 8750.
Warsh JJ, Emamghoreishi E, Li PP (1999): Elevated B lymphoblast
resting intracellular calcium is a trait-dependent abnormality
specific to bipolar-I disorder. Bipolar Disord 1(suppl 1):57.
BIOL PSYCHIATRY
2000;48:665– 673
673
Warsh JJ, Li PP (1996): Second messenger systems and mood
disorders. Curr Opin Psychiatry 9:23–29.
Yamanaka G, Eckstein F, Stryer L (1986): Interaction of retinal
transducin with guanine triphosphate analogues: Specificity
of the g-phosphate binding region. Biochemistry 25:6149 –
6153.
Yoshimasa T, Sibley DR, Bouvier M, Lefkowitz RJ, Caron MG
(1987): Cross-talk between cellular signalling pathways suggested by phorbol ester induced adenylate cyclase phosphorylation. Nature 327:67–70.
Young LT, Li PP, Kamble A, Siu KP, Warsh JJ (1994):
Increased mononuclear leukocyte G proteins in bipolar but
not unipolar depression. Am J Psychiatry 151:594 –596.
Young LT, Li PP, Kish SJ, Siu KP, Kamble A, Hornykiewicz O,
Warsh JJ (1993): Cerebral cortex Gas protein levels and
forskolin-stimulated cAMP formation are increased in bipolar
affective disorder. J Neurochem 61:890 – 898.