Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Despite a ban on production for more than 20 years, polychlorinated biphenyls (PCBs) continue to be prevalent contaminants in the environment, and pre- or early postnatal exposure to environmental levels of PCBs has been associated with altered neurological function and/or impaired cognition in humans (see Table 3 of review by Brouwer et al., 1999). Pre- and/or postnatal exposure to PCBs also results in altered behavior in animal models (Schantz et al., 1997; Bushnell and Rice, 1999). Altered neurological function (hypotonia and hyporeflexia at birth, delayed motor skills) and cognition after pre/postnatal exposure to PCBs occur in the absence of overt pathological alterations, suggesting that impairment of neurodevelopmental processes, rather than overt cytotoxicity, underlies the neurotoxic effects of developmental exposure to PCBs. The cellular mechanisms by which PCBs may alter neuronal development are not well understood, but are the subject of intense study.

Impaired Ca²⁺ homeostasis with elevations in resting [Ca²⁺]_i has been observed in many different cell types after acute in vitro exposure to micromolar concentrations of PCBs (Kodavanti et al., 1993; Voie and Fonnum, 1998; Bae et al., 1999; Fischer et al., 1999; Mundy et al., 1999), suggesting that PCBs alter Ca²⁺-dependent signal transduction. In addition to basal [Ca²⁺]_i levels, inositol phosphate hydrolysis, protein kinase C, and CaM kinase II have been a focus for PCB-induced alterations (Kodavanti et al., 1994; Brown et al., 1998; Fischer et al., 1999); these probably are events that are downstream from altered [Ca²⁺]_i (Shafer et al., 1996). Finally, other studies have demonstrated PCB effects on ryanodine-sensitive Ca²⁺ release channels (Wong et al., 1997) and voltage-sensitive Ca²⁺ channel (VSCC)-mediated changes in [Ca²⁺]_i (Inglefield and Shafer, 2000), both of which are also critically involved in Ca²⁺ signaling in neurons.

In the course of a recent investigation into the effects of PCBs on -aminobutyric acid (GABA)_A receptor-mediated Ca²⁺ and Cl responses in cultured cortical cells (Inglefield and Shafer, 2000), we observed a complex pattern of Ca²⁺ signaling induced by exposure to Aroclor 1254 (A1254), a commercial mixture of PCBs. In developing neurons, spatial and temporal patterns of Ca²⁺ signals are of greater importance for regulation of neuronal excitability, neurotransmitter release, gene transcription, and synaptic strength than is average cytoplasmic Ca²⁺ concentration (Gu and Spitzer, 1997). Recent evidence suggests Ca²⁺ oscillations affect the expression of distinct sets of genes (Dolmetsch et al., 1998). Thus, the observation of temporally coupled Ca²⁺ oscillations in response to A1254 may have important implications as a cellular mechanism of action of PCBs. We therefore have characterized the events induced by A1254 and investigated their pharmacology to elucidate the mechanism(s) underlying these Ca²⁺ signals. Overall, the results support the hypothesis that in vitro exposure to A1254 exerts disruptive effects on network Ca²⁺ signaling via multiple cellular targets in temporally distinct patterns and that this complexity should be taken into account when trying to address the consequences of PCBs on neural structures/processes in animals.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Solutions. The physiological solution consisted of 135 mM NaCl, 4.2 mM KCl, 1.5 mM CaCl₂, 0.5 mM MgCl₂, 0.34 mM Na₂PO₄, 0.44 mM KH₂PO₄, 10 mM glucose, 20 mM sucrose, 10 mM HEPES, pH 7.4, 290 to 300 mOsm. The physiological solution contained 0.5 mM Mg²⁺, so control cells from days in vitro (DIV) 4 to 7 cultures rarely exhibited a Ca²⁺ oscillation (Wang and Gruenstein, 1997). Stock solutions of A1254 (lot NTO1022, technical grade purity; UltraScientific, North Kingston, RI) were dissolved in anhydrous dimethyl sulfoxide (DMSO). Working concentrations were diluted 1:1000 to give a final [DMSO] of 0.1% in physiological solution. Molarity of the A1254 solutions was based on the average molecular weight of the congeners usually present in A1254, i.e., 326.4 g/mol. In experiments to examine the source of Ca²⁺ involved in the A1254-induced oscillations in intracellular Ca²⁺, the external Ca²⁺ was removed from the solution, and EGTA (1 mM) was included. GABA_A receptors were blocked with 50 µM bicuculline (BIC) (Sigma, St. Louis, MO). N-Methyl-D-aspartate (NMDA) receptors and/or -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid/kainate receptors were antagonized with 50 µM D-2-amino-5-phosphonopentanoic acid (d-AP5) (Research Biochemicals International, St. Louis, MO) with or without 10 µM 2,3-dihydroxy-6,7-dinitroquinoxaline (DNQX) (Research Biochemicals International), respectively. Nifedipine and EGTA were purchased from Sigma. All drugs were dissolved in physiological solution (and DMSO, when needed, at concentrations <0.1%) and applied by bath.

Rat Neocortical Primary Cell Cultures. Cortical cell cultures are a model system for the study of Ca²⁺ oscillations which are associated with a burst of action potentials (Robinson et al., 1993; Shen et al., 1996; Przewlocki et al., 1999). The procedure used for the preparation and maintenance of neocortical cultures prepared from newborn (less than 1-day-old) rat pups was as previously described (Inglefield and Shafer, 2000). The Long-Evans Rat (Charles River, Portage, MI) neocortices were dissected, minced, and then trypsinized (0.25%) in a culture solution containing 137 mM NaCl, 5 mM KCl, 0.17 mM NaH₂PO₄, 0.21 mM KH₂PO₄, 59 mM sucrose, 5 mM glucose, 100 I.U./ml penicillin, and 0.10 mg/ml streptomycin, pH 7.4. After a 5-min 0.016% DNase I digestion, the supernatant was removed, and the tissue resuspended in cortical medium [Dulbecco's modified Eagle's medium (no. 10313; Life Technologies, Inc., Grand Island, NY), including 10% horse serum, 25 mM glucose, 2 mM glutamine, penicillin/streptomycin (100 I.U. and 0.1 mg/ml, respectively), and 10 mM HEPES]. Dissociation of the tissue into single cells by trituration and gravity filtration through a 100-µm Nitex screen preceded plating of the cells onto poly(L-lysine)-coated 25-mm glass coverslips in six-well culture dishes at a density of 3 × 10⁶ cells/well (~3100 cells/mm² of the coverslip). Cells were incubated at 37°C in fresh cortical medium in a humidified incubator with 5% CO₂, 95% air for 3 DIV and were treated for the next 2 days with 5 µM -cytosine arabinoside (to limit replication of non-neuronal cells) in fresh cortical medium. Cultures were maintained for up to 8 DIV, receiving 50% fresh cortical medium every third day after removing cytosine arabinoside. All reagents were of highest available grade from commercial sources. This culture is highly enriched for neurons that reside on a bed of glia. The protocols used were approved by the U.S. Environmental Protection Agency's National Health and Environmental Effects Research Laboratory Institutional Animal Care and Use Committee.

Measurement of Fura-2 Fluorescence Changes. Coverslips were incubated in 2 ml of physiological solution containing cell-permeable fura-2/acetoxymethyl ester (fura-2/AM) (5 µM) purchased from Molecular Probes Inc. (Eugene, OR). Stock dye-loading solution was prepared by dissolving 50 µg of fura-2/AM in 11 ml of anhydrous DMSO containing 10% pluronic acid; this solution was dissolved in physiological solution to bring the final fura-2 concentration to 5 µM. Loading was for 30 min in the dark at 30°C, and then equilibrated >20 min at room temperature (23°C) in fresh physiological solution to remove unhydrolyzed dye and to allow for conversion of the fura-2/AM to its Ca²⁺-sensitive form, fura-2. After loading and washing, Ca²⁺ imaging experiments were performed as previously described (Inglefield and Shafer, 2000). Cytoplasmic Ca²⁺was measured at room temperature (23°C) with a DeltaScan dual excitation fluorescence imaging system from Photon Technology International (South Brunswick, NJ) and a Nikon Diaphot inverted microscope with Nikon Fluor40 objective (numerical aperture 1.3). A field of cells was illuminated alternately at 340 and 380 nm and emitted fluorescence (>510 nm) images collected on-line using a c2400 SIT video camera (Hamamatsu, Hamamatsu City, Japan) connected to a Photon Technology International image processor. The 340- and 380-nm images were typically obtained for a total of 1 s once every 10 s continuously for up to 1 h and stored on disk for subsequent processing. After baseline recording, A1254 exposure was initiated by addition of an equal volume containing 2× the final concentration. To verify there was no interaction of A1254 with fura-2, in separate studies we found the fura-2 fluorescence profile to be unchanged when A1254 was incubated with fura-2 free acid in a cell-free system (data not shown).

1

Data Analysis. Photon Technology International's ImageMaster software was used to acquire and process the images. Measurements from the soma of ~10 individual cells collected simultaneously were plotted individually. A typical field of fura-2-loaded cells is presented in Fig. 1. The solid arrows denote neuronal-appearing cell types that were included in the analysis, whereas the open arrows denote examples of cells with an astrocyte-appearing morphology that were not included in analysis. Data from at least two coverslips per condition, covering two or more culture dates were assessed. A1254 effects on fura-2 fluorescence were expressed as ratios of the 340 and 380 excitation images (F340/380). For the initial experiments, in situ calibration of [Ca²⁺]_i was carried out using the F340/380 and by obtaining the parameters for calculating intracellular free Ca²⁺ concentration as derived previously (Grynkiewicz et al., 1985). However, the majority of experiments did not require calibration of [Ca²⁺]_i because the goal was to characterize the frequency, pharmacological sensitivity, or other aspects of the Ca²⁺ oscillations; thus, data are presented herein as ratio signals. F340/380 ratios allow one to correct for baseline drift due to dye leakage or photo bleaching, which can be an issue in this length of experiment with UV illumination. Small differences in basal Ca²⁺ before exposure, reflected in the baseline fura-2 fluorescence ratio, were normalized to allow comparison across groups.

View larger version (188K): [in this window] [in a new window]   Fig. 1.   Representative image of a fura-2 fluorescence (340-nm excitation) microscopic field of a DIV 7 neocortical culture at 400× magnification. Cells with neuronal-type morphology (denoted by solid arrows) were included in the analysis, whereas cells with an astrocyte morphology (open arrows) were not. Scale bar, 10 µm.

Statistics. ANOVA followed by post hoc t tests for independent groups was used to compare effects of pharmacological treatments on A1254-induced oscillation frequency across DIV. For studies where DIV was not a factor, differences in the frequency of Ca²⁺ oscillation induced by A1254 before and after pharmacological treatments were compared using paired t tests. The Kolmogorov-Smirnov two-sample test was used to compare two distributions of accumulative frequency (probability) and to determine whether the amplitude distributions of two concentrations of A1254 differ with respect to location and dispersion.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Ca²⁺ Oscillations Induced by PCB Mixture A1254. The fluorescence imaging technique was used to identify a complex pattern of Ca²⁺-signaling events occurring in primary cell cultures of neonatal neocortex that was stimulated on acute exposure to A1254 (Fig. 2). A1254 (20 µM) caused an initial Ca²⁺ transient in 90% of cells. Within 3 to 15 min after the A1254-induced Ca²⁺ transient, there occurred nonperiodic Ca²⁺ oscillations in 70% of DIV 4 to 7 cells with net [Ca²⁺]_i amplitudes of ~200 to 600 nM each (from a baseline of ~100 nM) that typically required 15 to 60 s to decay to prespike levels. The focus of the present study is on the A1254-induced Ca²⁺ oscillations, whereas a detailed study of the mechanisms underlying the initial Ca²⁺ transient stimulated on immediate addition of A1254 is the focus of ongoing research (Inglefield et al., 2000). Because DIV 4 to 7 cortical cultures were studied in 0.5 mM Mg²⁺-containing solution, Ca²⁺ oscillations in control cells were infrequent, with only 6% of the control cells having >3 (to a maximum of six) oscillations during 1 h of imaging (data not shown). This is in contrast to the actively occurring oscillations when neonatal rat cortical neurons are maintained with zero or only nominal levels of Mg²⁺ in the solution (Robinson et al., 1993; Wang and Gruenstein, 1997) or when older cultures are used, e.g., cortical culture at DIV 8 and beyond (our unpublished data) when the interconnected neuronal networks are synaptically active. Synchrony, although occasionally observed, among the Ca²⁺ oscillations was not a constant feature among the cells in an imaged field. The effect of A1254 was not readily reversible by washout with A1254-free solution (data not shown). In addition, neither acute cytotoxicity nor moribund changes (cell swelling, neurite beading) occur after this 1 h of exposure to A1254 (Inglefield and Shafer, 2000); later time points (24 h) are currently being assessed.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Types, and concentration dependence, of Ca2+ responses observed in neocortical cells on addition of A1254. A, shown in order of successively greater disturbance (types I, II, and III) are representative 1-h traces of F340/380 ratio in fura-2-loaded neonatal cells maintained in culture 4 to 7 days. Arrows indicate addition of A1254, which remained in the bathing solution for remainder of experiment. The scale bar applies for all three responses shown. The criteria for a cell to exhibit a type I or II response was the presence of >3 oscillations in the latter 0.5 h of A1254 exposure; for a type II response, this was superimposed on an elevation in baseline Ca2+. B, increasing concentrations of A1254 (1-25 µM A1254) caused a rightward shift in the proportion of cells having type I, II, or III responses. Data are from two coverslips per concentration with the exception of the 25 µM A1254, which is from a single coverslip (because of the severity of response, 25 µM was not pursued in subsequent studies). NR, no response.

Multiple Types of Ca²⁺ Responses Observed in Developing Cortical Cells on Addition of Aroclor 1254. Careful examination of the responses of 134 individual cells to A1254 at several concentrations indicated that the responses could be classified into three types (Fig. 2A). Type I responses consisted of an initial transient, which returned to baseline, that was followed 3 to 15 min later by Ca²⁺ oscillations where the basal [Ca²⁺] did not rise appreciably. Type II responses were similar to type I responses, but were defined as distinct from type I because the basal Ca²⁺ level also became elevated in the course of the 1-h period. Type III responses were distinguished from the others by the delayed (or failure to) return to baseline levels after the initial transient as well as the absence of Ca²⁺ oscillations (Fig. 2A). With increasing A1254 concentration, there was a concentration-dependent shift in the percentage of cells having no response (NR; less than three oscillations in 30 min) to type I to type II to type III (Fig. 2B, i.e., more types II and III induced by 20 versus 2 µM). This progression of cellular responses from NR to type I, II, and III successively is consistent with progressively greater Ca²⁺ disturbances by A1254. A concentration of 20 µM A1254 was selected for the mechanistic studies (see below) because of the robust number of events in a 10-min recording period and because the highest concentration (25 µM) led to severe Ca²⁺ disturbances (i.e., 67% of 15 cells had a type III response) where no oscillations occurred. This is consistent with an inhibition of Ca²⁺ oscillations when baseline Ca²⁺ is elevated (Toescu et al., 1993). For 20 µM A1254, 26, 42, and 19% of the cells had type I, II, and III responses, respectively, with NR observed in 13% of cells.

Dependence of A1254 Concentration on Occurrence, Frequency, and Amplitude of Delayed Ca²⁺ Oscillations. We sought to characterize further this Ca²⁺-signaling pattern elicited by A1254 by examining the occurrence (percentage of responding cells), frequency, and amplitude of the Ca²⁺ oscillations. To ensure that the analyses were performed at a time of stable oscillations, analyses were performed on the last 30 min of the 1-h exposure. There was an association between the prevalence, frequency, and amplitude of Ca²⁺ oscillations and increasing concentrations of A1254 (Figs. 3 and 4). The percentage of cells exhibiting a Ca²⁺ disturbance (e.g., type I, II, or III response) increased as a function of increasing A1254 concentration with ~86% of 20 µM A1254-exposed cells exhibiting altered Ca²⁺ signaling (Fig. 3A). Assessment of the latter 30 min of the 1-h exposure revealed that higher A1254 concentrations were associated with an increased average oscillation frequency in the responding cells (Fig. 3B, also traces in Fig. 3C). The average frequency (min¹) of Ca²⁺ oscillations increased from 0.2 ± 0.07 in cells exposed to 1 µM A1254 to 0.9 ± 0.05 in cells exposed to 20 µM A1254. The traces in Fig. 3C portray the increased frequency as well as the amplitude of the events evoked with A1254 exposure. The mean amplitude as well as the total number of events detected increased in a concentration-dependent manner. For 10 and 20 µM A1254, amplitudes of Ca²⁺ oscillations were normally distributed with a mean ± S.E. of 0.836 ± 0.02 F340/380 (determined from 263 events recorded from 16 responding cells) for 10 µM and 0.955 ± 0.02 F340/380 (382 events, 26 responding cells) for 20 µM A1254 (Fig. 4). The dispersion and location of these distributions, compared with a Kolmogorov-Smirnov two-sample test, were significantly different (P < .05) (Fig. 4). Although not enough events were detected to observe a normal distribution, the average F340/380 amplitudes for 1 and 2 µM A1254 were 0.325 ± 0.014 (36 events from 5 cells responding) and 0.362 ± 0.024 (61 events from 10 cells responding), respectively.

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Concentration dependence of oscillations in intracellular Ca2+ elicited by A1254 exposure in DIV 4 to 7 neocortical cells. The percentage of cells (occurrence) having Ca2+ oscillations, as well as the amplitude and frequency of Ca2+ oscillations, were concentration dependent (1-20 µM). Fura-2 ratio images were acquired 1 per 10 s, and the spontaneous events were interactively detected for the latter 30 of the 60-min exposure, a time when Ca2+ events were the most regular. A, occurrence of cells with 3 Ca2+ oscillations in the latter 0.5 h (e.g., type I and II responses) increased with A1254 concentration. Parentheses indicate total number of cells examined for the different concentrations and represent measurements from two coverslips from different culture dates. B, average frequency also increased with A1254 concentration. Frequency was calculated by dividing the total number of events detected in each responding cell by the time window (30 min). There was a significant overall effect of A1254 concentration (one-way ANOVA), with both 10 and 20 µM A1254 having oscillation frequencies significantly greater than control (P < .05, Tukey's post hoc test). n = 5 to 26 responding cells, depending on A1254 concentration. C, representative Ca2+ traces covering the last half of the 60-min exposure show that the pattern of Ca2+ signals changed with increasing A1254 concentration. Note the increases in oscillation frequency and amplitude.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Distribution histograms plotting the number of events versus amplitude of the fura-2 340/380 ratio (bin width = 0.2 F340/380). The detected events in all cells exposed to 10 or 20 µM A1254 covering t = 30 to 60 min is presented. Amplitude of F340/380 events, corresponding to an increase in [Ca2+]i (Grynkiewicz et al., 1985), was the increase above the immediate 20 s preceding baseline. Both 10 and 20 µM A1254 had histograms with normal distributions. Comparison of these histograms with a Kolmogorov-Smirnov two-sample test showed that the two differed significantly with respect to location and dispersion. In situations where cells had a "type II" response (i.e., where an oscillation occurred at a time of elevated basal Ca2+), the fura-2 ratio amplitude includes the elevation in baseline level.

Mechanisms Underlying A1254-Induced Ca²⁺ Oscillations. As an initial step to determine mechanisms underlying A1254-induced Ca²⁺ oscillations, the dependence of oscillations on intra- or extracellular Ca²⁺ was investigated. When cells were exposed to A1254 in a nominally Ca²⁺-free physiological solution, Aroclor 1254-induced Ca²⁺ oscillations were absent (0 of 12 cells oscillated in the presence of 0 Ca²⁺ and A1254) (Fig. 5A). Because L-type VSCCs have been demonstrated to be one major route of Ca²⁺ influx underlying spontaneous Ca²⁺ oscillations in cortical neurons (Wang and Gruenstein, 1997), subsequent experiments examined effects of a low concentration of an L-type VSCC antagonist, nifedipine, on A1254-induced Ca²⁺ oscillations. Application of nifedipine (1 µM) to the cells quelled A1254-induced oscillations, significantly decreasing the frequency from 0.88 ± 0.17 min¹ to 0.15 ± 0.07 min¹ (n = 6; Fig. 5B). Finally, spontaneous Ca²⁺ oscillations in cortical cell cultures can arise from trans-synaptic activity and are blocked by the Na⁺ channel antagonist tetrodotoxin (Shen et al., 1996; Przewlocki et al., 1999). Addition of tetrodotoxin (2 µM) inhibited the intracellular Ca²⁺ oscillations that had been induced by 1 h of A1254 (Fig. 5C), suggesting PCB exposure had induced excitatory synaptic activity.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Synaptic activity underlies the Ca2+ oscillations induced by A1254 exposure. Representative traces of F340/380 of individual neocortical cells (A and C) or a simultaneously imaged group of three cells (B) pretreated with A1254 (20 µM) and then exposed to different conditions. The trace in A represents the response in 12 of 12 cells where Ca2+ was removed from the sample chamber solution and replaced with 1 mM EGTA 5 min before A1254 addition. B, bath application of the L-type selective Ca2+ channel blocker nifedipine (1 µM) completely prevented oscillations evoked by A1254 in 12 of 13 cells tested. C, A1254-induced Ca2+ oscillations were prevented in the presence of tetrodotoxin (2 µM) in 14 of 14 cells.

Maturation-Dependent Changes in Postsynaptic Receptors Mediating A1254-Induced Ca²⁺ Oscillations. Because block of oscillations by tetrodotoxin suggested a role for synaptic activity in the A1254-induced oscillations and because excitatory amino acid receptors are initiators of Ca²⁺ oscillations (Wang and Gruenstein, 1997), we used antagonists selective to postsynaptic receptors to pharmacologically isolate the oscillatory activity. Depending on the stage of development, GABA and glutamate have important excitatory actions on synaptic transmission in developing neurons (Cherubini et al., 1991; Yuste and Katz, 1991; Wang et al., 1994; Obrietan and van den Pol, 1995; Leinekugel et al., 1997). Therefore, antagonism of receptors on these major ligand-gated ion channels was assessed to confirm that A1254-evoked Ca²⁺ oscillations were indeed synaptically mediated. As an initial assessment of the involvement of excitatory amino acid receptors, a mixture of antagonists to glutamate receptors (50 µM d-AP5 and 10 µM DNQX) and GABA_A receptors (50 µM BIC) was added to the bath solution before A1254 (20 µM) exposure. In 11 of 11 DIV 5 cells tested, this reversibly abolished A1254-induced Ca²⁺ oscillations (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   A1254-induced Ca2+ oscillations were blocked by antagonists to excitatory transmitters, GABA and glutamate. A, representative traces of A1254-induced Ca2+ oscillations in cultured neocortical cells before and after the addition of the GABAA antagonist BIC (50 µM) (top and middle) and/or glutamate receptor antagonists d-AP5 (50 µM) and DNQX (10 µM) (middle and bottom) to the cultures. B, summary data for the effects of different antagonists of postsynaptic receptors () on A1254-induced Ca2+ oscillations in cells of different DIV. Data are the mean ± S.E. for 13 to 20 cells per DIV from at least two different culture dates. At this duration of exposure, maximal oscillation amplitudes induced by A1254 20 µM (before antagonist addition, ) varied slightly for the cells maintained in culture 4 to 7 days, with a range of 0.8 to 2.2 min1. Note that the frequencies are more variable and greater in some cases versus those presented in Fig. 3, probably owing to the much shorter sampling time (8-10 min) herein. The block of oscillations by BIC was effective only on DIVs when GABA is excitatory in these cells (<DIV 6) (Inglefield and Shafer, 2000).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Using the Ca²⁺ indicator fura-2, the effects of the commercial PCB mixture A1254 on Ca²⁺ signaling in neocortical cells were examined. The novel effect of PCBs to exceed a threshold and cause increased periods of Ca²⁺ oscillations through the mechanisms of action identified in this study is distinct from the widely studied signal of a slow, tonic increase in basal Ca²⁺ from PCB exposure (Kodavanti et al., 1993; Voie and Fonnum, 1998; Fischer et al., 1999; Mundy et al., 1999). In cultures of neonatal rat cortical neurons, there is a correlation of spontaneous Ca²⁺ oscillations with spontaneous synaptic potentials (Robinson et al., 1993; Shen et al., 1996); thus, changes in Ca²⁺ oscillation frequency can be an indication of altered interconnected network activity. Ca²⁺ oscillations are also found in more complex, intact preparations secondary to spontaneous electrical activity, such as with giant depolarizing potentials in developing hippocampal neurons in slices (Yuste and Katz, 1991; Leinekugel et al., 1997).

The Ca²⁺ response patterns as well as the frequency and amplitude of Ca²⁺ oscillations elicited by A1254 were examined and determined to be concentration dependent. With respect to types of Ca²⁺ responses observed during an hour of exposure to A1254 (as classified in Fig. 2), increasing A1254 concentration induced a graded response with increasing proportions of cells recruited from nonactive (NR, i.e., <3 oscillations in a 0.5-h period) to recurring Ca²⁺ oscillations (types I and II), and finally, to a severe Ca²⁺ disturbance (type III) that precluded Ca²⁺ oscillatory behavior. This concentration relationship was also seen for the percentage of cells in the field recruited to undergo oscillations as well as for frequency and amplitude of the Ca²⁺ oscillations. It seems likely that with increasing oscillation frequency the response strains the Ca²⁺ normalizing mechanisms so that basal Ca²⁺ cannot be maintained; the shift to types II and III (where basal Ca²⁺ is not maintained) with increasing A1254 concentration corroborates the elevated basal [Ca²⁺]_i seen previously after 1 h of A1254 exposure (Inglefield and Shafer, 2000). Because the contribution of specific pathways to Ca²⁺ spike firing has been widely studied recently (Shen et al., 1996; Wang and Gruenstein, 1997; Przewlocki et al., 1999; Mermelstein et al., 2000), cellular mechanisms of action for the A1254-induced Ca²⁺oscillations could be readily explored.

There are several examples of PCB-induced intracellular Ca²⁺ disturbances where the source of cellular Ca²⁺ increase has been discerned as being either extra- or intracellular. For example, withdrawal of extracellular Ca²⁺ from cultured cerebellar granule cells (Mundy et al., 1999) or immune cells (Voie and Fonnum, 1998) abolishes elevations of basal [Ca²⁺]_i due to pure ortho-substituted PCB congeners, the primary components of A1254. However, ortho-substituted PCB congeners also cause increased activation of ryanodine-sensitive Ca²⁺ release channels, situated in the membrane of the endoplasmic reticulum (Wong et al., 1997), suggesting that release of intracellular Ca²⁺ stores may contribute to PCB-induced elevations in basal [Ca²⁺]_i. Experiments conducted to discriminate between these different pools of Ca²⁺ for A1254-mediated oscillations in the developing cortical cells found that the A1254-induced oscillatory activity was not observed after treatment of cells with A1254 in the absence of extracellular Ca²⁺ (Fig. 5A), demonstrating the importance of extracellular Ca²⁺ for this effect. In addition, the block of A1254-induced Ca²⁺ oscillations by addition to the bath of nifedipine (1 µM) indicated that Ca²⁺ entry occurred primarily through L-type VSCCs. This result is consistent with the role of postsynaptically situated L-type VSCCs in mediating spontaneous Ca²⁺ oscillations in cortical neurons (Wang and Gruenstein, 1997; Przewlocki et al., 1999). In other non-neuronal preparations, it has also been shown that Ca²⁺ disturbances by Aroclor PCB mixtures are inhibited by high concentrations of L-type VSCC antagonists (Bae et al., 1999; Fischer et al., 1999); however, the concentrations of such modulators are not selective for VSCCs (Skeen et al., 1994; Hargreaves et al., 1996). Inhibitors of the N-, P/Q-, and T-type VSCCs may also participate in the Ca²⁺ responses because they are partial antagonists for Ca²⁺ oscillations (Wang and Gruenstein, 1997). Although at present we cannot rule out a release of Ca²⁺ from intracellular stores, e.g., the ryanodine receptor, as a Ca²⁺ channel participating directly in the A1254-stimulated oscillations, Wang and Gruenstein (1997) reported that mobilization of Ca²⁺ from ryanodine-sensitive intracellular stores is not instrumental in Ca²⁺ oscillations that operate through a mechanism of VSCC activation.

Overall, the sensitivity of the A1254-induced oscillations to tetrodotoxin and the dependence of Ca²⁺ oscillations in cortical neurons on synaptic activity (Robinson et al., 1993; Shen et al., 1996) suggested that excitatory synaptic input was required to generate PCB-induced Ca²⁺ oscillations. Additional support for this hypothesis was provided by experiments with postsynaptic receptor antagonists. Antagonism of ionotropic glutamate receptors with a combination of d-AP5 and DNQX was effective at blocking Ca²⁺ oscillations mediated by the A1254 on DIV 6 and 7 [no data were collected for the earlier days in vitro but the NMDA antagonist probably would be effective then also (Leinekugel et al., 1997)]. In keeping with this tenet that Ca²⁺ influx through L-type VSCCs occurs secondary to NMDA receptor activation, Ca²⁺ imaging of oscillations in hippocampal neurons also identified an essential role for NMDA receptors, whereas -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors played a lesser role in the generation of Ca²⁺ oscillations (Przewlocki et al., 1999). It is interesting to note that in the presence of the receptor antagonists used herein, A1254 did not increase the basal [Ca²⁺]_i (and, in fact, there was a "dip" to below A1254-induced baseline Ca²⁺ levels upon the antagonist-mediated cessation of Ca²⁺ oscillations; Fig. 6A). This effect of a slight fall in baseline Ca²⁺ values was also reported when blockers of membrane receptors were added to quell the Ca²⁺ oscillations mediated by glutamatergic synaptic transmission in rat hippocampal cultures (Shen et al., 1996).

The results with receptor antagonists also indicated that in addition to ionotropic glutamate receptor activation, GABA_A receptor activation participated prominently in the Ca²⁺ oscillations in cells from the earliest days in vitro (before DIV 6). Besides the well known membrane depolarization induced by activation of glutamate receptors, GABA, acting on GABA_A receptors, also exerts excitatory actions on developing neurons, i.e., it activates a Cl efflux that in turn stimulates voltage-sensitive Ca²⁺ channels and thereby elevates [Ca²⁺]_i (Cherubini et al., 1991; Yuste and Katz, 1991; Segal, 1993; Wang et al., 1994; Owens et al., 1996). Shortly thereafter, GABA_A receptor stimulation in intact cells matures to a hyperpolarizing effect due to influx of Cl, and this does not activate VSCCs (Obrietan and van den Pol, 1995; Owens et al., 1996; Inglefield and Schwartz-Bloom, 1997). A similar developmental profile of GABA_A receptor responses occurs in the cortical preparation used in the present experiments, wherein activation of GABA_A receptors before DIV 6 stimulates Cl efflux and Ca²⁺ entry via L-type VSCCs, but after DIV 6 stimulates only Cl influx (Inglefield and Shafer, 2000). The contribution of GABA_A receptors to A1254-elicited Ca²⁺ oscillations varied at two different stages of maturation in vitro, being substantial (>60% effective) on DIV 4 and DIV 5; but insignificant on DIV 6 and later (<10% effective). Thus, the involvement of GABA_A receptors in A1254-induced Ca²⁺ oscillations parallels the developmental profile of GABA_A receptor responses in this culture.

In light of the finding that the actions of A1254 were blocked by both glutamate and GABA_A receptor antagonists, it would appear that A1254 excited the network independent of direct effects on these particular neurotransmitter receptors. Very recently, effects of ortho-substituted PCBs on presynaptic machinery (e.g., enhanced neurotransmitter release and reduced reuptake) have been identified (Heck and Wiegand, 1999; Mariussen et al., 1999) and such enhanced transmitter levels, either from increased release or reduced reuptake/metabolism, is consistent with the fact that A1254-induced Ca²⁺ oscillations are blocked after the inhibition of synaptic activity with tetrodotoxin. Thus, a tentative model for the initiation of Ca²⁺ oscillations by A1254 can be postulated (one that ties together results from other laboratories that have investigated presynaptic actions of ortho-substituted PCBs): a recurring depolarization of the cell membrane by A1254, secondary to elevated levels of glutamate (or GABA, when it is excitatory), activates voltage-operated calcium channels to elevate [Ca²⁺]_i. This results in Ca²⁺ oscillations in the soma that are generated by the synaptic activity of the cortical cell network. Alternatively, other mechanisms cannot presently be ruled out, such as the block of K⁺ channels and/or activation of Na⁺ channels by A1254 or actions involving Ca²⁺ induced Ca²⁺ release. However, because postsynaptic receptor antagonists were effective in decreasing oscillation amplitude and frequency, it is unlikely that the increased excitability by A1254 would be mediated solely via an effect on one of these other molecular targets.

Finally, these results in a cortical cell culture suggest that a novel mechanism, changes in the operation of central nervous system cellular networks, may underlie the actions of PCBs to alter neurodevelopmental processes. Patterns of Ca²⁺ signals mediated by surface receptors such as VSCCs and ligand-gated ion channels encode cellular genetic responses (Dolmetsch et al., 1998; Rajadhyaksha et al., 1999; Mermelstein et al., 2000). In the developing nervous system, spatial and temporal Ca²⁺ signaling are critical to a large number of processes, namely, neuronal migration/survival (Behar et al., 1997; Lauder et al., 1998), growth factor synthesis and release (Berninger et al., 1995), and growth of developing neurites (Obrietan and van den Pol, 1996; for review, see Gu and Spitzer, 1997). In addition, Ca²⁺ oscillations at frequencies higher than those observed herein result in delayed neurotoxicity (McLeod et al., 1998). Thus, because Ca²⁺ oscillations convey selective signals to the cell nucleus (Dolmetsch et al., 1998; Rajadhyaksha et al., 1999; Mermelstein et al., 2000), future studies are needed to assess the link between Ca²⁺ oscillations in developing neocortical cells induced by PCBs and the activation of downstream genetic events.