OUP user menu

Hyperactive intracellular calcium dynamics in B lymphoblasts from patients with bipolar I disorder

Tatiana Perova, Michael J. Wasserman, Peter P. Li, Jerry J. Warsh
DOI: http://dx.doi.org/10.1017/S1461145707007973 185-196 First published online: 1 March 2008

Abstract

Substantial evidence implicates abnormalities of intracellular calcium (Ca2+) dynamics in the pathophysiology of bipolar disorder (BD). However, the precise mechanisms underlying such disturbances are poorly understood. To further elaborate the nature of altered intracellular Ca2+ signalling dynamics that occur in BD, we examined receptor- and store-operated Ca2+ responses in B lymphoblast cell lines (BLCLs), which have been found in earlier studies to ‘report’ BD-associated disturbances. Basal Ca2+ concentrations ([Ca2+]B), and lysophosphatidic acid (LPA)- and thapsigargin-stimulated Ca2+ responses were determined in BLCLs from 52 BD-I patients and 30 healthy comparison subjects using fura-2, and ratiometric fluorometry. ANOVA revealed a significant effect of diagnosis, but not gender, on [Ca2+]B (F1,63=4.4, p=0.04) and the rate of rise (F1,63=5.2, p=0.03) of LPA-stimulated Ca2+ responses in BLCLs from patients compared with those from healthy subjects. A significant gender×diagnosis interaction on the LPA-induced rate of rise (F1,63=4.6, p=0.03) was accounted for by a faster rate of rise (97%) in BLCLs from BD-I males compared with healthy males but not in those from female patients compared with healthy females. A gender×diagnosis interaction in thapsigargin-evoked Ca2+ influx (F1,61=3.8, p=0.05) resulted from a significantly higher peak [Ca2+]influx (24%) in BLCLs from female compared with male patients. The results suggest more rapid LPA-stimulated Ca2+ responses occur in BLCLs from BD-I patients compared with controls, which are probably mediated, in part, by canonical transient receptor potential type 3 (TRPC3)-like channels. Additionally, this study highlights sex-dependent differences that can occur in the pathophysiological disturbances involved in BD.

Key words
  • Bipolar I disorder
  • receptor-operated calcium entry
  • signal transduction
  • store-operated calcium entry

Introduction

Although the aetiopathology of bipolar disorder (BD) is poorly understood, substantial evidence points to the involvement of altered intracellular signal transduction and second-messenger signalling in the disorder (Li et al., 2000; Warsh and Li, 1996). In this regard, disturbances of intracellular calcium (Ca2+) signalling dynamics have emerged as being of considerable interest due to the vital role Ca2+ plays in many critical cellular functions that affect cellular resilience and neuroplasticity (Ghosh and Greenberg, 1995; Orrenius et al., 2003), disturbances of which have been implicated pathophysiologically in BD (Manji et al., 2000; Rajkowska, 2000) and are ameliorated by mood-stabilizer treatment (reviewed in Bachmann et al., 2005).

Intracellular Ca2+ homeostasis (ICH) is tightly regulated and involves many mechanisms. Calcium entry into the cell can occur via receptor-operated mechanisms, mediated by inositol trisphosphate (IP3) and ryanodine-induced release of Ca2+ from endoplasmic reticulum (ER) stores, and is followed by a sustained influx of extracellular Ca2+ through a process known as store-operated Ca2+ entry (SOCE). Transient receptor potential (TRP) channels, stromal interaction molecule 1 (Stim1) and Orai proteins have been implicated in these processes (reviewed in Ambudkar et al., 2006; Liao et al., 2007; Yuan et al., 2007). In addition, Ca2+ entry occurs through ligand- and voltage-gated ion channels. Plasma membrane and sarcoplasmic/ER Ca2+-ATPases pumps (PMCA and SERCA, respectively), Na+/Ca2+ exchangers, and sequestration and storage in ER and mitochondria also play critical roles in regulating ICH (Berridge et al., 2003).

Altered ICH in BD was first suggested by observations of elevated baseline intracellular Ca2+ levels ([Ca2+]B) and agonist-stimulated Ca2+ mobilization ([Ca2+]S) in platelets, lymphocytes and B lymphoblast cell lines (BLCLs) from BD patients (reviewed in Warsh et al., 2004). The specific mechanisms that account for ICH disturbances in BD are poorly understood but may involve alterations in one or more of the homeostatic processes that regulate intracellular Ca2+ fluxes, including receptor-operated Ca2+ entry (ROCE) (Berk et al., 1994; Dubovsky et al., 1992; Hough et al., 1999; Tan et al., 1995) and SOCE (Hough et al., 1999; Kato et al., 2003).

Observations that thapsigargin (TG)-stimulated [Ca2+]S was enhanced in platelets (Hough et al., 1999) and BLCLs (Kato et al., 2003) from BD patients implicate altered SOCE in the BD pathophysiology. Recently, Akimoto et al. (2007) provided preliminary evidence suggesting that enhanced activity of protein kinase C, as observed in post-mortem brain and platelets from BD patients (Friedman et al., 1993; Wang et al., 1999; Wang and Friedman, 1996), may inhibit the activity of SOCE in platelets from BD patients. Moreover, chronic treatment with lithium and valproate attenuated receptor-mediated or store-depletion-induced Ca2+ mobilization in BLCLs from BD-I patients and healthy controls (Perova et al., 2005; Wasserman et al., 2004). Furthermore, chronic treatment with lithium reduced protein levels of canonical TRP type 3 (TRPC3) channel, a non-selective cation channel implicated in ROCE and SOCE (Groschner and Rosker, 2005), in BLCLs from BD-I patients (Andreopoulos et al., 2004). These new findings, taken together, provide additional support to the notion that altered SOCE and/or ROCE may be involved in the Ca2+ disturbances in BD.

The objectives of this study were to further elaborate the pathophysiological mechanisms underlying aberrant ICH in BD as expressed in BLCLs, a surrogate model which reports cellular disturbances associated with BD (Emamghoreishi et al., 1997; Iwamoto et al., 2004; Karege et al., 2004; Wasserman et al., 2004). To this end, two pharmacological probes were used in this study to scrutinize two specific functional modules involved in intracellular Ca2+ signalling. The first technique examined agonist-activated Ca2+ mobilization mechanisms using lysophosphatidic acid (LPA), which, at 100 µm, has been shown to provoke Ca2+ entry through a diacylgycerol (DAG)-dependent TRPC3-like channel in BLCLs (Roedding et al., 2006). The second technique employed the SERCA inhibitor, TG, to transiently deplete ER Ca2+ pools, thereby inducing SOCE (Putney et al., 2001). We report here significantly enhanced LPA-stimulated but not TG-evoked Ca2+ mobilization in BLCLs from BD-I patients compared with those from healthy subjects. These observations suggest altered Ca2+ homeostasis in BLCLs from BD-I patients is probably due to disturbances in DAG-gated TRPC3 channels that modulate the kinetics of Ca2+ mobilization.

Materials and methods

Subjects

Cell lines for this study were prepared from samples from a cohort of BD-I patients and healthy comparison subjects (age range 18–65 yr) who were recruited, as previously described (Emamghoreishi et al., 1997), between 1995 and 2006. Patients had a diagnosis of BD-I disorder according to DSM-IV criteria as established by the Structured Clinical Interview – Patient Edition (SCID-I/P; First et al., 1995b), and had no concurrent medical illness, diabetes or hypertension, no recent (>3 months) drug or alcohol abuse, and were competent to provide informed written consent. Details regarding patients' past history of the psychiatric disorder, including age of onset, number and types of hypomanic/manic or depressive episodes, presence or absence of psychotic symptoms, lifetime psychiatric comorbidity, and types and response to medications, as well as family history of psychiatric disorder were obtained from the SCID interview based on patient recall and, whenever possible, corollary medical records. Comparison subjects were physically healthy and had no past or current psychiatric illness as determined by the SCID-I – Non-Patient Edition (First et al., 1995a) and no family history of psychiatric disorders in first-degree relatives. The study was approved by the Human Subjects Review Board of the University of Toronto and all subjects provided written informed consent after a detailed explanation of the study.

BLCLs

Cell lines used for this study were either freshly transformed and expanded in culture (n=14; 7 from healthy subjects, 7 from BD-I patients), or selected from frozen stock (n=68; 23 from healthy subjects, 45 from BD-I patients) and regrown to the required cell numbers. Epstein–Barr virus-transformed BLCLs were prepared using standard techniques (Emamghoreishi et al., 1997). Cells were expanded in suspension culture in a 95% air/5% CO2 humidified atmosphere at 37°C in B cell media containing RPMI, 15–20% fetal bovine serum (FBS), 2 mmol/l l-glutamine, 1 mmol/l pyruvate, 100 µg/ml streptomycin and 100 U/ml penicillin. The medium was completely replaced every 2–3 d during expansion to the required cell number (≈6×107 cells). Cells were counted and viability determined by Trypan Blue exclusion (Emamghoreishi et al., 1997) at each replacement of medium. At 14–23 passages, cells were assayed for LPA and TG responses. Age- and sex-matched cell lines from BD-I patients and healthy controls were paired for each assay.

Ca2+ mobilization assays

Agonist (LPA) stimulation

BLCLs (1×107/ml) were incubated with fura-2 AM (1 µm, Molecular Probes, Eugene, OR, USA), or for autofluorescence, DMSO (1 µm), in RPMI-1640 (30 min, 37°C). Cells were washed with RPMI-1640, resuspended in ice-cold Hepes buffered saline solution [Hepes-AG: 135 mm NaCl, 5 mm KCl, 1 mm CaCl2, 1 mm MgCl2, 10 mm glucose, 20 mm Hepes, 5 mm polyethylene glycol-8000 (PEG, Mallinckrodt, Point-Claire, QC, Canada), pH 7.4] at a density of 1.0×106 cells/ml, and then transferred to a cuvette and equilibrated (3 min, 37°C) in a temperature-controlled cuvette chamber of a PerkinElmer LS50B spectrofluorophotometer (Beaconsfield, UK). PEG was added to Hepes-AG to reduce aggregation of the cells upon stimulation of LPA and in turn the intra-assay variance of the induced Ca2+ response. The fura-2 AM loading time, sample handling conditions and the LPA (100 µm) stimulation of Ca2+ were optimized and performed as previously described (Wasserman et al., 2004).

TG stimulation

Cells (1–3×107/ml) were loaded with fura-2 AM or DMSO in RPMI-1640 (45 min, 37°C), washed with RPMI-1640 and then nominally Ca2+-free Hepes-AG, and resuspended at a density of 1×106 cells/ml and equilibrated as above. The determination of Ca2+ mobilization from ER and SOCE following stimulation with 200 nm TG in nominally Ca2+-free conditions and subsequent re-addition of 1 mm Ca2+ (final concentration) to the incubation medium, was performed as previously described (Wasserman et al., 2004).

Determinations were performed in triplicate aliquots of BLCLs from each subject and the mean values were used in all subsequent data analyses. Calibration of fura-2 fluorescence intensities and calculation of [Ca2+]i was performed using the Grynkiewicz equation as previously described (Emamghoreishi et al., 1997). The resulting [Ca2+]i response curves were smoothed using a ‘fine’ Lowess fitting algorithm (GraphPad Prism 4; GraphPad Software Inc., San Diego, CA, USA). Data used in subsequent statistical analyses were obtained using these smoothed curves.

Statistical analyses

Data are presented as means±s.d. Differences in the proportion of males to females in each diagnostic group were analysed using the χ2 test. After confirming homogeneity of variance (Levene's test), differences in dependent variables between subject groups were analysed by univariate analysis of variance (ANOVA) with diagnostic group and sex as factors, the latter of which has been shown to affect Ca2+ signalling parameters in BLCLs (Emamghoreishi et al., 2000; Yoon et al., 2001a,b). Interactions between factors for each dependent measure were examined using an analysis of simple effects (Gardner, 2001). Dependent measures determined in the LPA-stimulated Ca2+ mobilization procedure were analysed separately from those of the TG-stimulation protocol to avoid the reduced power to detect differences of small effect size, of an omnibus MANOVA followed by univariate ANOVAs, which would have excluded records with missing values due to procedural difficulties. Potential effects of categorical factors, including comorbid psychiatric diagnosis, mood state and type of mood-stabilizer treatment at the time of study, and family history of mood disorders were assessed statistically using MANOVA, followed by subsequent univariate ANOVAs to delineate the specific dependent variable(s) accounting for the significant effect(s). Tukey's HSD test was used for post-hoc comparisons. Statistical differences with two-tailed probability values of p⩽0.05 were taken as significant. Statistical analyses were performed using SPSS version 14.0 software (SPSS Inc., Chicago, IL, USA).

Results

Subjects

Table 1 summarizes the demographic characteristics of the BD-I patients and healthy comparison subjects from whom BLCLs were used in this study. While BLCL samples were prepared from 52 BD-I patients and 30 healthy subjects, the actual number of samples analysed for each parameter in each group was less than this total due to methodological difficulties resulting in sample loss. The mean age of the BD-I patients was not statistically different from that of the healthy subjects (t=0.86, p=0.39), nor was the proportion of females to males (χ2=0.05, p=0.83). Mean age at onset of the first episode of illness in patients was 22.2±14.6 yr. At the time of study, 52% of the patients were euthymic, 27% were depressed and the remainder were either hypomanic, manic, or in a mixed state. The most prevalent comorbid psychiatric disorder found in the patient group was alcohol abuse/dependence (33%). There was a positive family history of mood disorders in first-degree relatives in 62% of the patients. Among patients, 33% were receiving lithium monotherapy at the time of assessment.

View this table:
Table 1

Agonist-stimulated intracellular Ca2+ mobilization

A representative LPA-stimulated Ca2+ response is illustrated in Figure 1a. The parameters characterizing Ca2+ homeostatic mechanisms in the LPA stimulation paradigm were measured as previously described (Wasserman et al., 2004) and included [Ca2+]B, peak LPA-stimulated [Ca2+]S, the absolute difference (Δ[Ca2+]S) between LPA-stimulated [Ca2+]S and [Ca2+]B, and the slope of the initial rate of rise in LPA-stimulated [Ca2+]S (d[Ca2+]S/dt), which was estimated by determining the slope of best-fit linear regression line through the initial phase of the rise in [Ca2+]S during a 30-s epoch immediately following the addition of LPA. As the variance among comparison groups/factors was unequal for the d[Ca2+]S/dt, these data were normalized by log10 transformation for subsequent parametric statistical tests. ANOVA revealed a significant main effect of diagnosis on [Ca2+]B (F1,63=4.4, p=0.04) and d[Ca2+]S/dt (F1,63=5.2, p=0.03), but not peak [Ca2+]S (F1,63=1.0, p=0.3) or Δ[Ca2+]S (F1,63=0.5, p=0.5). These differences reflected modest but significantly higher mean [Ca2+]B (15%) and a markedly more rapid rate of rise in [Ca2+]S (40%) in BLCLs from BD-I patients compared with those from healthy subjects (Figure 2a, b, respectively).

Figure 1

Representative traces of the time-course (a) of the response of intracellular Ca2+ response to LPA (100 µm) stimulation and (b) of TG-induced SOCE, in a subject BLCL. Response parameters were measured as described in the Materials and methods section.

Figure 2

Hyperactive agonist (lysophosphatidic acid, LPA)-simulated Ca2+ mobilization measured in B lymphoblast cell lines (BLCLs) from BD-I patients. LPA (100 µm)-stimulated Ca2+ responses were determined in BLCLs from BD-I patients (n=42) and healthy subjects (n=25) using Fura-2 AM as described in the Materials and methods section. (a) BLCLs from BD-I patients had significantly higher basal intracellular Ca2+ concentration ([Ca2+]B) compared with those from healthy subjects (ANOVA: F1,63=4.4, p=0.04). (b) BLCLs from BD-I patients had significantly faster initial rate of rise of LPA-stimulated intracellular Ca2+ response (d[Ca2+]S/dt) compared with those from healthy subjects (ANOVA: F1,63=5.2, p=0.03). Means and error bars indicating±s.d. are shown (* p<0.05 vs. healthy subjects).

While no significant effect of gender was evident on any of the dependent Ca2+ measures, there was a significant interaction between gender and diagnosis on d[Ca2+]S/dt (F1,63=4.6, p=0.03). Analysis of simple effects revealed a significantly higher d[Ca2+]S/dt (97%) in BLCLs from BD-I males compared with healthy males (F1,63=10.2, p=0.002; Figure 3), but not in BLCLs from female BD-I patients compared with those from healthy females (F1,63=0.1, p=0.8; Figure 3). In contrast, BLCLs from female patients had moderately lower d[Ca2+]S/dt (−29%) compared with male BD-I patients but this difference fell short of statistical significance (F1,63=3.7, p=0.06).

Figure 3

Interaction between gender and diagnosis on lysophosphatidic acid (LPA)-induced rate of rise in intracellular Ca2+ levels (d[Ca2+]S/dt) in B lymphoblast cell lines (BLCLs). A significant effect of diagnosis in male subjects was due to a significantly faster d[Ca2+]S/dt in BLCLs from male BD-I patients (n=19) compared with healthy males (n=13) (ANOVA: F1,63=10.2, * p=0.002). There was no difference in d[Ca2+]S/dt between female BD-I patients (n=23) and healthy females (n=12). Means and error bars indicating±s.d are shown.

TG-sensitive ER Ca2+ release and parameters of SOCE

A representative TG-induced Ca2+ response is illustrated in Figure 1b. Parameters measured to characterize the TG-stimulated SOCE-mediated Ca2+ responses were as previously described (Wasserman et al., 2004) and included the basal [Ca2+] ([Ca2+]B1), the peak TG-sensitive [Ca2+]i accumulation ([Ca2+]TG), and the baseline plateau ([Ca2+]B2), which in this Ca2+ add-back protocol were measured in a nominally Ca2+-free condition. After re-introduction of Ca2+ to the extracellular medium, the peak of Ca2+ influx ([Ca2+]influx), the difference between [Ca2+]influx and [Ca2+]B2 (Δ[Ca2+]influx), and the slope of the initial phase of rise in [Ca2+]influx (d[Ca2+]influx/dt), which was estimated during the 30-s epoch immediately following the addition of Ca2+ to the extracellular medium, were measured. ANOVA revealed no significant effect of diagnosis or gender on [Ca2+]B1 (3%), [Ca2+]TG (−8%), [Ca2+]B2 (5%), [Ca2+]influx (4%), Δ[Ca2+]influx (5%), or d[Ca2+]influx/dt (13%) in TG-evoked Ca2+ mobilization (F1,61 values=0.02–1.6, p values ⩾0.22). Intriguingly, there was a significant interaction between gender and diagnosis in the peak store-operated [Ca2+]influx (F1,61=3.8, p=0.05). Analysis of simple effects revealed a significantly higher peak [Ca2+]influx (24%) in BLCLs from female BD-I patients compared with those from male BD-I patients (F1,61=5.4, p=0.02; Figure 4) but not in BLCLs from healthy females compared with those from healthy males (F1,61=0.5, p=0.5; Figure 4). Although BLCLs from female BD-I patients exhibited higher peak [Ca2+]influx (20%) compared with those from healthy females, this difference did not reach significance (F1,61=3.1, p=0.08).

Figure 4

Interaction between gender and diagnosis on thapsigargin (TG)-evoked peak store-operated [Ca2+]influx in B lymphoblast cell lines (BLCLs). TG stimulation was performed in a nominally Ca2+-free medium for 4 min followed by re-addition of 1 mm Ca2+ to the medium to assess SOCE. A significant effect of gender in BD-I patients was due to a significantly higher peak [Ca2+]influx in BLCLs from female BD-I patients (n=20) compared with male BD-I patients (n=22) (ANOVA: F1,61=5.4, * p=0.02). There was no significant difference in peak [Ca2+]influx between male and female healthy subjects. Means and error bars indicating±s.d. are shown.

Relationship between Ca2+ mobilization measures and clinical characteristics of BD patients

To explore potential relationships between measured parameters of Ca2+ homeostasis and clinical characteristics of patients, the data were analysed with patient samples stratified based on their respective state of illness at the time of study (euthymic, depressed, or manic), medications (naive, lithium monotherapy, lithium and neuroleptics, valproate monotherapy or valproate and neuroleptics), psychiatric comorbidity (none, panic disorder, alcohol abuse/dependence), familial history of mood disorders, presence of psychosis, or age at onset of BD. MANOVA of the parameters of LPA-stimulated and TG-induced Ca2+ mobilization revealed no significant effect of the state of illness (F6,84=0.6, p=0.7 and F10,74=1.1, p=0.4), mood-stabilizer medication (F6,74=0.4, p=0.9 and F10,60=1.1, p=0.4), presence of a psychiatric comorbidity (F6,58=0.7, p=0.6 and F10,54=1.3, p=0.3), presence of psychosis (F3,42=0.1, p=0.9 and F5,37=0.9, p=0.5) or age at onset (F3,43=0.2, p=0.9 and F5,37=0.8, p=0.6) in BLCLs from patients stratified based on each of these characteristics.

Interestingly, there was a significant effect of family history of mood disorders on the parameters of LPA-stimulated Ca2+ mobilization in BLCLs from patients compared with those from healthy subjects (F6,74=2.4, p=0.04). Subsequent univariate analysis revealed that BLCLs from BD-I patients with a positive family history of mood disorders had significantly higher [Ca2+]B (35%) compared with those from healthy subjects (F2,38=3.5, p=0.04; Tukey p=0.03; Figure 5a). Although patients with a negative family history of mood disorders had lower [Ca2+]B (11%) compared with those with a positive history and higher [Ca2+]B (21%) compared with those from healthy subjects, the differences were not statistically significant (Tukey p=0.6 and p=0.4, respectively). In addition, BLCLs from patients with a negative family history of mood disorders showed significantly higher d[Ca2+]S/dt (58%) compared with those from healthy subjects (F2,38=3.2, p=0.05; Tukey p=0.04) but were only non-significantly higher (26%) than those from patients with a positive family history (Tukey p=0.2; Figure 5b). The d[Ca2+]S/dt was also faster (17%) in BLCLs from patients with a positive family history of mood disorders compared with those from healthy subjects, but this difference was not statistically significant (Tukey p=0.7).

Figure 5

Basal intracellular Ca2+ concentration ([Ca2+]B) and the initial rate of rise in lysophosphatidic acid (LPA)-stimulated [Ca2+]S (d[Ca2+]S/dt) in B lymphoblast cell lines (BLCLs) from BD-I patients with or without a family history of mood disorders (MD). (a) BLCLs from BD-I patients with a positive history of MD in first-degree relatives (n=15) had significantly higher [Ca2+]B compared with those from healthy subjects (n=17) (ANOVA: F2,38=3.5, p=0.04; Tukey p=0.03). (b) BLCLs from BD-I patients with a negative history of MD in first-degree relatives (n=9) had significantly higher d[Ca2+]S/dt following LPA stimulation compared with those from healthy subjects (n=17) (F2,38=3.2, p=0.05; Tukey p=0.04). Means and error bars indicating ± s.d. are shown (* p<0.05 vs. healthy subjects).

Discussion

The principal finding of this study is that of significantly more rapid Ca2+ influx kinetics reflected in the faster rate of rise of [Ca2+]S following LPA stimulation in BLCLs from BD-I patients compared with those from healthy subjects, in addition to replicating higher [Ca2+]B levels in BLCLs from BD-I patients as previously reported (Emamghoreishi et al., 1997). In contrast, the lack of differences between BD-I patients and healthy controls in the TG-provoked Ca2+ flux parameters in BLCLs, which probes both TG-sensitive ER Ca2+ pools and SOCE mechanisms, suggests the disturbance(s) that accounts for the enhanced Ca2+ mobilization in BD-I patients may not involve store depletion-mediated Ca2+ entry, at least as reflected by the TG-perturbation protocol used.

The intracellular Ca2+ response provoked by LPA has for some time been accredited to G protein-coupled receptor (GPCR)-activated phospholipase C (PLC)/IP3-dependent pathway, which occurs at low (<10 µm) concentrations (Fukushima et al., 2002; Rosskopf et al., 1998). At the concentration (100 µm) used in the present study, however, LPA probably gates Ca2+ influx predominantly through a TRPC3-like channel (Roedding et al., 2006), acting at a DAG-sensitive site similar to the DAG analogue, 1-oleoyl-2-acetyl-sn-glycerol (OAG), with which LPA shares structural homology and which gates Ca2+ entry through the TRPC3/6/7 channel subfamily (Hofmann et al., 1999). Only the TRPC3 subtype of the latter TRPC subfamily is expressed in BLCLs, as well as the non-DAG-sensitive TRPC1 and TRPC5 subtypes (Roedding et al., 2006). Given this predominant mode of action on TRPC3 of the higher LPA concentration used here, the differential rate of rise in LPA-stimulated Ca2+ response in BLCLs from BD-I patients implicates functional alterations in TRPC3 channels that affect the kinetics of Ca2+ influx.

While we used BLCLs that were freshly transformed or regrown from frozen stock, the reported differences in [Ca2+]B and the rate of Ca2+ influx were observed in both sets of BLCLs (data not shown). Such observations indicate that enhanced Ca2+ responses in BLCLs from BD-I patients are not due to differences consequent to freezing and passaging of the cell lines, at least within the range of passages in this study.

The factors affecting the rate of influx through TRPC3 channels are still poorly understood but may involve processes that influence the plasma membrane levels of these channels or channel gating. TRPC3 channels exhibit multiple modes of regulation determined by the tetrameric subunit composition, vesicular trafficking mechanisms, conformation-coupling with ER membrane IP3 receptors, and other protein–protein interactions such as with PLCγ, src, the STIM1–Orai1 complex, and calmodulin that constrain and shape the Ca2+ signalling microdomains (Ambudkar et al., 2006; Liao et al., 2007; Yuan et al., 2007). The finding that reduced expression and number of open channels of another TRPC subtype, TRPC1, diminishes the rate of rise in Ca2+ in response to stimulation (Cai et al., 2006), supports the notion that alterations in the expression levels or the number of open TRPC3 channels might account for the differences observed in BLCLs from BD-I patients. As no differences were found in the mRNA or protein levels of TRPC3 in BLCLs from BD-I patients compared with those from healthy controls (Andreopoulos et al., 2004), the basis for the disturbed LPA-activated response may lie in an alteration of one or more of the other aforementioned processes which regulate TRPC3 gating and function.

The possibility that alterations in the G protein-coupled LPA receptor-stimulated Ca2+ mobilization cascade, which would also be activated by LPA, contribute to the differentially enhanced rate of rise of [Ca2+]S response cannot be ruled out. Indeed, higher agonist (platelet-activating factor, thrombin, serotonin)-stimulated [Ca2+]S, acting through a GPCR pathway has been reported in platelets and lymphocytes from BD patients (reviewed in Warsh et al., 2004). However, it was the maximal responses and not the rates of activation of the responses that showed differences in BD patients compared with controls in such preparations. Finally, the extent to which the changes in Ca2+ flux and the underlying abnormalities reflect direct consequences of the pathophysiology of the disorder, or compensatory changes that result from alterations in one or more of the other intracellular signalling cascades implicated in BD interacting with Ca2+ signalling, remains to be elucidated.

The results of this study also suggest that potential gender differences occur in intracellular Ca2+ mobilization processes in BLCLs from BD-I patients, since the kinetics of LPA-stimulated Ca2+ mobilization and TG-evoked SOCE were differentially altered in BLCLs from male and female patients. Such findings are not unexpected given that earlier studies in BLCL preparations reported sex-dependent differences in the mRNA expression of myo-inositol monophosphatase 2 (Yoon et al., 2001a), and basal and sodium fluoride-stimulated adenylyl cyclase activity (Emamghoreishi et al., 2000) as well as mRNA expression of Disrupted-in-Schizophrenia-1 (DISC1; Maeda et al., 2006). Thus, the present findings reinforce the notion that gender-dependent factors may operate to modify BD-associated disturbances of ICH.

Earlier observations of higher TG-induced [Ca2+]S in platelets and lymphocytes from BD patients compared with healthy subjects (Hough et al., 1999) suggested that altered TG-stimulated SOCE would also be found in BLCLs from BD-I patients. However, the current study did not reveal diagnostic differences among the parameters measured in the TG-evoked store depletion protocol. Several factors could explain such a discrepancy. First, the inability to detect changes similar to those previously reported by Hough et al. (1999) may reflect cell type-specific differences in the SOCE mechanisms. However, the ‘Ca2+ add-back’ protocol (Putney et al., 2001) used in the current study more effectively discriminates SOCE-mediated Ca2+ influx from other mechanisms that would have been operative under the conditions (i.e. incubation with TG in the presence of extracellular Ca2+) used in the earlier study (Hough et al., 1999). Second, there is a possibility that the confounding effects of state-related factors, circulating hormones and medications may have affected Ca2+ responses in platelet and lymphocyte preparations, which were isolated freshly from patients' blood samples, resulting in the differences reported. In contrast, BLCLs used here were isolated from such carry-over effects, given that cells are passaged and expanded in culture ex vivo for numerous generations. Third, potential gender differences in the TG-stimulated [Ca2+]influx, as noted above, may have confounded the detection of altered TG-stimulated parameters in BLCLs of BD-patients. In this regard, it is interesting to note that a preponderance of female patients (23/30) was studied by Hough et al. (1999).

It should be noted the [Ca2+]B in the TG Ca2+ add-back protocol was determined in the absence of extracellular Ca2+ as necessitated by the ‘Ca2+ add-back’ technique to observe the SOCE component, whereas [Ca2+]B in the case of LPA stimulation was determined at physiological extracellular [Ca2+]. The elevated [Ca2+]B in BLCLs from BD patients found in the latter condition (Figure 2a) is dependent upon extracellular Ca2+ and may reflect the constitutive activity of certain plasma membrane Ca2+ permeable channels, such as TRPC channels, and/or Ca2+ pumps.

The heterogeneity of BD is well acknowledged and certain clinical characteristics reflecting putative endophenotypes of BD may distinguish genetically more homogeneous subgroups of this complex trait disorder (Gottesman and Gould, 2003; MacQueen et al., 2005; Payne et al., 2005). The clinical subphenotypes examined here include such characteristics as age at onset (Bellivier et al., 2003; Leboyer et al., 2005), lifetime history of certain psychiatric comorbidities (MacKinnon et al., 2002), and BD with psychosis (Potash et al., 2003), among others. Of particular note, a familial history of mood disorders appears to be associated with altered [Ca2+]B and the rate of rise of [Ca2+]s following LPA stimulation. In this regard, BD-I patients with a positive familial history of mood disorders showed significantly higher [Ca2+]B, consonant with findings of significantly elevated [Ca2+]B in BD-I patients with a positive family history of mood disorder in a larger sample cohort (Warsh et al., unpublished data). In contrast, significantly higher d[Ca2+]S/dt was evident in BD-I patients with a negative family history of mood disorders compared with healthy controls. On the one hand, the possibility can not be discounted that the above differential associations with positive and negative family history of mood disorders may reflect an artifact attributable to small sample size or extraneous variables. On the other hand, these differential associations may reflect different pathogenetic mechanisms that influence the cellular and molecular processes involved in Ca2+ responses activated by LPA. While the exact mechanisms are yet to be elucidated, these relationships to family history of mood disorders hint at the possibility that abnormal ICH may be heritable and a pathophysiological phenotype of BD.

The lack of differences in BLCL Ca2+ mobilization parameters in patients receiving mood-stabilizer treatment compared to those who were not, was not unexpected, as BLCLs were expanded extensively in culture and thus isolated from in-vivo medication effects. The sample size may have been too small, however, to detect potential differences in Ca2+ responses among subjects on the different mood-stabilizer treatments.

Finally, potential limitations of the cell model used in this study must be considered. The BLCLs are derived from the transformation of B lymphocytes and are not of neural lineage. Substantive evidence, however, supports that BLCLs express or report disease-associated abnormalities relevant to the pathophysiology of a number of neuropsychiatric disorders including BD. In examples drawn from studies of Huntington's disease (Panov et al., 2002; Sawa et al., 1999) and Alzheimer's disease (Ibarreta et al., 1997) with established neurodegenerative pathology, the molecular pathophysiology that disrupts ICH in neurons (Mattson, 2007) is evident to some extent in BLCLs from affected patients. BLCLs have been shown to report alterations in ICH (Andreopoulos et al., 2004; Emamghoreishi et al., 1997; Kato et al., 2003; Washizuka et al., 2005; Wasserman et al., 2004) and the expression of examined candidate genes in BD (Maeda et al., 2006; Sun et al., 2004; Yoon et al., 2001b). To be sure, BLCLs can not express the array of changes as would be expected of the affected neurons in brain from patients with these disorders. An additional caveat is that the viral transformation could distort the nature of the signalling disturbances that are reflected in this model from that which might be expressed in primary cell lines. This does not diminish the potential investigative utility of this preparation, however, as long as the findings show disease specificity and, arguably, are hereditable. It should also be noted that the pharmacological analysis used only provides indirect evidence implicating enhanced TRPC3 functionality in BLCLs from BD-I patients: measurement of TRPC3 function as characterized electrophysiologically is necessary to confirm directly the inferences drawn. Finally, replication of the findings in a larger sample suitably powered subject to MANOVA with multiple factors is warranted to confirm the abnormalities attributed to TRPC3 implicated in this study.

In summary, our study indicates hyperactive LPA-stimulated Ca2+ responses occur in BLCLs from BD-I patients and provides further support to the notion that aberrant intracellular Ca2+ dynamics is an important aspect of BD pathophysiology. Given the complex nature of LPA-stimulated Ca2+ responses in BLCLs (Roedding et al., 2006), such aberrant Ca2+ dynamics may be due to alterations at the level of DAG-sensitive TRPC3 channels. Findings of recent genetic studies showing significant association between BD and at least two types of non-voltage- dependent Ca2+-permeable channels, including the melastatin TRP type 2 (TRPM2) channel (McQuillin et al., 2006; Xu et al., 2006), the expression of which is altered in BLCLs from BD-I patients (Yoon et al., 2001b), and the nucleotide-gated purinergic P2X7 receptor cation channel (Barden et al., 2006) provide substance to the possible pathogenetic mechanisms that initiate the cascade of disturbances which disrupt ICH and play out in the pathophysiology of BD.

Acknowledgements

This work was supported by the Canadian Institutes of Health Research (grant MOP12851). The technical assistance provided by Kin Po Siu and Karen Woo is greatly appreciated. Bronwen Hughes assisted in the recruitment of subjects whose B lymphoblasts were used in this study. Michael J. Wasserman has become affiliated with H.I.G. Ventures, Miami, FL after completion of his experimental contributions to this work.

Statement of Interest

None.

References

View Abstract