Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Polychlorinated biphenyls (PCBs) are a class of persistent pollutants that are prevalent in the environment, and there is increasing evidence from both human epidemiological studies and animal models that developmental exposure to low levels of PCBs can result in subtle changes in behavior and cognition (see review by Brouwer et al., 1999). Because there is an absence of overt pathological alterations in the human as well as in animal models (Brouwer et al., 1999), it presently appears that subtle rather than gross macroscopic changes in human and animal nervous systems underlie the altered neurologic function and/or impaired cognition that occur following developmental PCB exposure. The cellular and molecular basis for PCB-induced developmental neurotoxicity is unclear; but in vitro, PCBs have been shown to disrupt Ca²⁺ homeostasis and processes involved in Ca²⁺-mediated signal transduction (reviewed in Tilson and Kodavanti, 1997).

Because Ca²⁺ signaling in developing and mature neurons can initiate and regulate a number of cellular responses, perturbations in temporal cellular Ca²⁺ signals may have important effects. Changes in intracellular Ca²⁺ concentration ([Ca²⁺]_i) can lead to subtle or profound changes in neuronal function by regulating diverse processes, such as cell survival and death, or changes in cellular phenotype and synaptic plasticity (Curtis and Finkbeiner, 1999). Thus, the impact of Ca²⁺ signals in developing cells is far-reaching. Ca²⁺ signals in neurons may be stimulated by many factors, and sources of these signals include influx through plasma membrane bound ion channels and release from intracellular stores operated by inositol 1,4,5-triphosphate (IP₃) receptors or ryanodine receptors (Berridge, 1998). In order to test the hypothesis that PCBs can affect Ca²⁺ signals in developing neurons, we have used an in vitro model system of developing neocortical cells that recapitulates many aspects of normal cortical neuron development, including transmitter pharmacology (Dichter, 1978; Inglefield and Shafer, 2000a). This model is appropriate to study the mechanisms of action of developmental neurotoxicants in view of the cognitive deficits, as well as functional changes in cortical (Altmann et al., 1998) and hippocampal (Gilbert and Crofton, 1999) long-term potentiation following PCB exposure. In previous studies, we reported that exposure to Aroclor 1254 (A1254), an environmentally relevant PCB mixture, results in temporal alterations in [Ca²⁺]_i and reductions in GABA_A receptor-mediated responses (Inglefield and Shafer, 2000a,b). The present study expands on our earlier work and investigates the initial mechanism of PCBs to perturb [Ca²⁺]_i, as well as the role of the initial mechanism of PCB action in the subsequent prolonged Ca²⁺ disturbances reported previously. Finally, because perturbations in temporal cellular Ca²⁺ signals may have important effects, two potential downstream consequences of altered Ca²⁺ signaling, cell viability, and transcription factor activation were examined in this model system.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Solutions. Fura-2 acetoxymethyl ester (fura-2-AM) and fura-2 free acid were obtained from Molecular Probes (Eugene, OR). Nifedipine, ionomycin, and EGTA were purchased from Sigma (St. Louis, MO). Thapsigargin, used to inhibit endoplasmic reticulum Ca²⁺-ATPases, was obtained from Sigma. IP₃ receptors were stimulated with carbachol (Sigma) and blocked with xestospongin C (Calbiochem, San Diego, CA). Ryanodine receptors were probed with caffeine (Sigma) and ryanodine (Research Biochemicals International, Natick MA). A1254 (technical grade purity; lot no. NTO1022) was obtained from UltraScientific (North Kingston, RI). 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. PCB congeners 2,2'-dichlorobiphenyl (DCB) (IUPAC PCB 4), 4,4'-DCB (PCB 15), 3,3',4,4',5-pentachlorobiphenyl (PCB 126), and 2,2',3,4,4',5'-hexachlorobiphenyl (PCB 138) (all 99% purity) were purchased from AccuStandard (New Haven, CT). All chemicals were diluted in HEPES buffer, and for those stock chemicals requiring dilution in dimethyl sulfoxide (DMSO), the final DMSO concentration was 0.1%.

Rat Neocortical Primary Cell Cultures. Rat neocortical cells were grown in primary culture as described previously (Inglefield and Shafer, 2000a). Cultures were prepared from 1-day-old Long-Evans rat pups (Charles River, Portage, MI) euthanized via approved protocols (U.S. Environmental Protection Agency's National Health and Environmental Effects Research Laboratory Animal Care and Usage Committee). Neocortices were dissected, minced, then trypsinized (0.25%) in a culture buffer solution containing 137 mM NaCl, 5 mM KCl, 0.17 mM NaH₂PO₄, 0.21 mM KH₂PO₄, 59 mM sucrose, 5 mM glucose, 100 IU/ml penicillin, and 0.1 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, Grand Island, NY), including 25 mM glucose, 2 mM glutamine, 100 IU/ml penicillin, 0.1 mg/ml streptomycin, and 10% horse serum]. The tissue was dissociated into single cells by trituration and gravity filtration through a 100-µm Nitex screen. Dissociated cells then were plated at a density of 3 × 10⁶ cells/well onto 25-mm round-glass coverslips that had been freshly coated with poly(L-lysine) and washed with distilled H₂O. Cells were grown in fresh cortical medium in a humidified incubator at 37°C with 5% CO₂/95% air (pH 7.4) for 3 days in vitro (DIV 1-3) 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 DIV 7, receiving fresh cortical medium upon removing cytosine arabinoside. All reagents were of the highest available grade from commercial sources. Cultures produced with these methods are enriched in neurons that have elaborate neurite processes, and culture wells at DIV 7 contain 70% neuron-specific enolase immunoreactive cells (neurons) that reside on a bed of glial fibrillary acidic protein-immunopositive glia (30%).

Controls for PCB Specificity. A1254 is a PCB mixture that consists primarily of ortho-substituted (>99%), as well as non-ortho-substituted (dioxin-like), PCB congeners. Depending on the system and responses under study, a structure-activity relationship for the differing Cl substitution patterns on the biphenyl ring has been demonstrated that distinguishes effects of ortho- and non-ortho substituted PCBs (reviewed in Tilson and Kodavanti, 1997). In this regard and to characterize the response, individual PCB congeners (5-10 µM) that are ortho-substituted (PCB 4 and PCB 138) or non-ortho-substituted (PCB 15, PCB 77, and PCB 126) were applied to the cortical cells, and [Ca²⁺]_i responses were monitored. There are no detectable levels of polychlorinated dibenzodioxins (PCDDs) and only very low abundance (0.0001%) dibenzofuran contaminants in the A1254 mixture (Kodavanti et al., 1999). Also, to control for the effects of DMSO on membrane integrity, 0.1% DMSO was the comparator (baseline) for the A1254 concentration-response studies.

Cytoplasmic Free [Ca²⁺] Concentration Measurements. [Ca²⁺]_i was measured with the Ca²⁺-sensitive fluorescent dye, fura-2-AM as described previously (Inglefield and Shafer, 2000a). Cells cultured on coverslips were incubated in cell-permeable fura 2-AM (5 µM) for 40 min at 30°C diluted in 2 ml of HEPES-buffered Hanks' balanced salt solution (referred to herein as HEPES buffer) consisting 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, and 10 mM HEPES (pH 7.4; 290-300 mOsm). Cells were washed twice with fresh HEPES buffer then equilibrated >30 min in the dark at room temperature to remove extracellular dye and to complete the de-esterification process (thereby converting the fura-2-AM to its Ca²⁺-sensitive form, fura-2). Coverslips containing fura-2-loaded cells were placed in a Leiden coverslip dish situated in a PDMI-2 microscope open incubator (23°C; Medical MicroSystems Corp., Greenvale, NY) that was mounted on the stage of a Nikon Diaphot inverted microscope with a Nikon Fluor40 objective (numerical aperture 1.3).

(1)

Cell Viability, Caspase 3 Activity, and Apoptosis. Cell viability was assessed by determining the ability of cortical cells to exclude Trypan Blue following exposure to A1254. Following a 24-h exposure to control (0.1% DMSO) or A1254-containing buffer (2, 10, or 20 µM), 0.4% Trypan blue was added to wells of cortical cultures that had been maintained in the incubator at 37°C (in 95% O₂/5% CO₂). Trypan blue excluding cells were counted using bright field microscopy (Olympus IMT2, 300×); in each of two identically treated wells, two randomly selected microscopic fields of at least 100 cells were examined for each concentration.

Cyclic AMP Responsive Element Binding Protein (CREB) Phosphorylation/Immunoblotting. DIV 6 cultures were washed free of serum by incubation in HEPES buffer with 0.1% bovine serum albumin (Sigma) for 2 h at 37°C before stimulation with the specified agent. The exposure was terminated by washing with ice-cold phosphate-buffered saline and addition of lysis buffer [1% Triton X-100, 20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 10% glycerol, 1 mM NaF, 1 mM Na₃VO₄, and 0.5% protease inhibitor cocktail (Calbiochem, San Diego, CA)]. Cells were removed from the wells, vortexed gently, and allowed to sit on ice for 10 min. The lysed cells were then centrifuged at 10,000g for 10 min at 4°C. An aliquot of the supernatant was taken for protein determination, and the remaining supernatant was added to an equal volume of sample buffer [62.5 mM Tris (pH 6.8), 25% glycerol, 2% sodium dodecyl sulfate, 0.01% bromophenol blue, and 5% -mercaptoethanol] and stored at 80°C.

Data Analysis. Data are presented as means ± S.E.M. For the Ca²⁺ responses, the percentage of responding cells in treatment groups are also noted. For comparisons of peak Ca²⁺ responses, statistical significance was ascertained using one- or two-way ANOVAs followed by suitable post hoc tests. To determine whether pharmacological pretreatments were effective at preventing A1254-induced responses, the proportion of cellular Ca²⁺ responses among groups in an experiment was compared using the ² test and Bonferroni-corrected p values for multiple comparisons. For the cell counts, a decrease in the cell density of control wells occurs with increasing DIV; thus, the experimental and statistical design took this into account, and data were analyzed with two-way ANOVA, followed by one-way ANOVA and Dunnett's t tests. Caspase 3 activity, TUNEL staining, and pCREB activation were assessed with one-way ANOVA and Dunnett's test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

A1254-Initiated Changes in Cytoplasmic Free [Ca²⁺]: Concentration Response. Figure 1A illustrates the typical response observed during 1 h of imaging of Ca²⁺ in a cortical cell exposed continuously to 10 µM (3 ppm) A1254. In the continued presence of A1254, a Ca²⁺ transient is followed 3 to 16 min later by disturbances in the basal Ca²⁺ level that often includes Ca²⁺ oscillations of ~200 to 700 nM in amplitude each (Fig. 1A). In a previous series of experiments, we have described the mode of action of PCBs on the latent Ca²⁺ oscillations in cortical neurons (Inglefield and Shafer, 2000b). The initial Ca²⁺ transient, the focus of this paper, comprises a rapid increase in [Ca²⁺]_i and a slow decay to basal levels (721 ± 98 s to return to 10% of peak). The latency to onset of the initial Ca²⁺ transient from the time of exposure to A1254 was always <30 s. In characterizing the initial Ca²⁺ transient, there was an A1254 concentration-dependent increase in the peak amplitude of [Ca²⁺]_i stimulated upon A1254 (1-25 µM, 0.3-7.5 ppm) exposure (Fig. 1B and Table 1). In addition to the effects on peak Ca²⁺ amplitude, there were also concentration-dependent increases in the proportion of cells responding (Table 1).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   A, effect of A1254 (10 µM) to stimulate intracellular Ca2+ in an intact P0 cortical cell maintained in culture for 5 days. Cells were loaded with the Ca2+-sensitive fluorescent indicator, fura-2, as described under Materials and Methods. The arrow indicates when the PCB mixture was bath-applied; the PCB mixture remained in the chamber for the duration of the experiment. The Ca2+ response obtained in the first 10 min (denoted by a box), when PCB addition occurred, is the focus of the data presented in B, as well as in Figs. 2 through 4 and Tables 1 through 3. B, representative examples of initial Ca2+ responses that occur following addition of A1254 from 1 to 25 µM. Each concentration was repeated in at least two different cultures with 27 to 60 cells per concentration.

Specificity of Ca²⁺ Responses. Addition of control buffer (0.1% DMSO) to cells from DIV 4 to 6 cortical cultures did not produce a [Ca²⁺]_i change (not shown) indicating that neither mechanical manipulation of the cells nor vehicle were responsible for the Ca²⁺ disturbance. Moreover, separate exposures of cells to two main classes of PCB congeners that are dibenzofuran- and dioxin-free (personal communication, Michael Bolger, AccuStandard) elicited distinct responses. Initial Ca²⁺ transients were consistently stimulated by ortho-substituted PCB congeners (PCB 4 and PCB 138), but such responses were rare with non-ortho-substituted PCB congeners (PCB 15, PCB 77, and PCB 126) (Fig. 2 and Table 2), which are aryl hydrocarbon receptor agonists like the dioxins. Both of the ortho-substituted congeners (PCB 4 and PCB 138; 10 µM) stimulated at least a 7-fold increase of [Ca²⁺]_i from baseline levels of 72 ± 5 nM in >70% of the cells. The characteristics of the initial Ca²⁺ transients stimulated by PCB 4 and PCB 138 were indistinguishable from that elicited by A1254 in that Ca²⁺ levels recovered toward basal levels over several minutes. Responses to the non-ortho-substituted PCBs occurred infrequently, although in those rare instances (in <10% of cells), the peak amplitude could be marked, reaching levels as high as 400 nM (Table 2).

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Effect of A1254 and selected PCB congeners on [Ca2+]i in developing cortical cells. Ortho-substituted PCB congeners 138 (2,2',3,4,4',5'-hexachlorobiphenyl) and 4 (2,2'-dichlorobiphenyl) at 10 µM selectively induced an initial Ca2+ transient with characteristics similar to that induced by the PCB mixture, A1254 (10 µM). Changes in intracellular Ca2+ were blunted or absent when non-ortho-substituted congeners 15 (4,4'-dichlorobiphenyl), 77 (3,3',4,4'-tetrachlorobiphenyl), or 126 (3,3',4,4',5-pentachlorobiphenyl) were applied (all at 10 µM, except PCB 126, which was added at 5 µM due to solubility issues). Arrows denote the time of PCB addition, and the PCBs remained in the bath for the rest of the experiment. # refers to the IUPAC number of the PCB congener used.

Release by A1254 of Intracellular Ca²⁺ Stores and Stimulation of Store-Operated Entry of Ca²⁺. PCB-induced disturbances in Ca²⁺ homeostasis have been reported to be due to both a mobilization of Ca²⁺ from an internal source(s) (Wong et al., 1997) and an influx of extracellular Ca²⁺ (Bae et al., 1999b; Mundy et al., 1999; Inglefield and Shafer, 2000b). Thus, we sought to determine the source of Ca²⁺ responsible for the initial Ca²⁺ transient induced by A1254. A rapid, initial Ca²⁺ transient was still present in Ca²⁺-free extracellular buffer solution (having an estimated total free Ca²⁺ concentration of ~20 nM), compared with Ca²⁺ replete solution. Although the peak Ca²⁺ amplitude was not significantly attenuated after the cells' exposure to a Ca²⁺-free buffer (Fig. 3 and Table 3), there was an effect to shorten the duration of Ca²⁺ responses stimulated by A1254 in Ca²⁺-free buffer (i.e., [Ca²⁺]_i exhibited a more rapid recovery to basal levels; Fig. 3A, inset). The Ca²⁺ responses stimulated by 20 µM A1254 in buffer containing the L-type Ca²⁺ channel blocker nifedipine (1 µM) also were not attenuated in terms of either the peak amplitude, percentage of cells responding (Table 3), or the decay to baseline (981 ± 187 s, N.S. p > 0.05 relative to A1254 + Ca²⁺). This indicates that the Ca²⁺ transient is largely due to release from intracellular Ca²⁺ stores, although Ca²⁺ influx from the extracellular solution contributes to and prolongs the decay. Interestingly, when Ca²⁺-containing buffer was reintroduced to cells exposed to A1254, the resting basal [Ca²⁺]_i became elevated rapidly in 59% of cells (Fig. 3B), suggestive of Ca²⁺ entry through plasma membrane-situated store-operated channels (SOCs) to refill A1254-activated intracellular Ca²⁺ stores (Fig. 3B). Figure 3C illustrates SOC-mediated Ca²⁺ entry stimulated with 1 mM of the muscarinic agonist, carbachol (79% of cells responded with store-operated Ca²⁺ entry); this has been shown to occur in developing neurons (Bouron, 2000). Overall, these results implicate the importance of intracellular Ca²⁺ pools in the peak initial Ca²⁺ transient stimulated by A1254 and demonstrate that a trigger of store-operated Ca²⁺ entry (also referred to as capacitative Ca²⁺ entry) from extracellular sources extends the period for return of intracellular Ca²⁺ to baseline.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Intracellular Ca2+ stores of developing cortical cells are activated by A1254, and this results in store-operated Ca2+ influx. A, relationship between PCB-stimulated release of internal Ca2+ and subsequent Ca2+ influx through transmembrane route(s). A1254 added to the DIV 4 to 6 cells without or with 1.5 mM Ca2+ in the extracellular buffer stimulated similar peak increases in the Ca2+ transient, but note the presence of a rapid recovery of intracellular Ca2+ to basal levels in Ca2+-free external buffer. The inset shows the significantly reduced time for [Ca2+]i to return to 10% of peak amplitude in the continued presence of A1254 in the Ca2+-free solution (n = 21-29 cells). *p < 0.05. Also different for the A1254-stimulated responses in Ca2+-free solution was the absence of Ca2+ responses previously referred to as a "type 3" (see Inglefield and Shafer, 2000b), which are rapid Ca2+ responses to A1254 that failed to return to basal levels when cells were incubated in a normal Ca2+-containing solution. Following [Ca2+]i stimulation by A1254 (20 µM; B) or carbachol (1 mM; C) and return to baseline in Ca2+-free solution, re-addition of Ca2+ (1.5 mM) produced a Ca2+ entry in 10 of 17 A1254-treated and in 11 of 14 carbachol (CCh)-treated cells.

Block of the A1254-Induced Initial Ca²⁺ Transient by Pharmacological Agents. Figure 4 shows the results of studies to determine the source of intracellular Ca²⁺ underlying the initial A1254-induced [Ca²⁺]_i transient. These mechanistic studies were performed with 20 µM A1254 since this concentration induced a nearly maximal response in the population of cortical cells. To determine the pool released by A1254, the effects of several pharmacological agents that alter Ca²⁺ release from intracellular Ca²⁺ stores were examined. The intracellular Ca²⁺ source for the majority of the A1254-induced initial Ca²⁺ transient was the endoplasmic reticulum because pretreatment with thapsigargin (10 µM for 10 min), a specific inhibitor of the endoplasmic reticulum Ca²⁺-ATPase pump, prevented detectable responses in 70% of the cells (Fig. 4). In those cells that did respond, the amplitude of the A1254 response was significantly attenuated (p < 0.05 following significant treatment effect in the one-way ANOVA) (Table 3). The same thapsigargin preincubation was sufficient to suppress completely the intracellular Ca²⁺ responses to stimulation by carbachol (1 mM) (response abolished in 16 of 18 cells). Further assessment of endoplasmic reticulum-activated mobilization of intracellular Ca²⁺ by A1254 was done using pharmacological agents active at IP₃ or ryanodine receptors. Both types of intracellular stores were functioning in these cells because applications of carbachol (1 mM), ryanodine (100 µM), or caffeine (20 mM) mobilized intracellular stores (Fig. 4). Although pretreatment with ryanodine (100 µM for 10 or 20 min) completely blocked the increase in [Ca²⁺]_i induced by caffeine, it was without effect on the A1254-induced increases in [Ca²⁺]_i (Fig. 4 and Table 3). In contrast, block of the Ca²⁺ transient occurred with the specific IP₃ receptor antagonist, xestospongin C (1 µM, 10 min preincubation). Only 11% of the 37 cells had a response to A1254 (p < 0.05 following significant treatment effect in the ² analysis); for these remaining cells, the peak amplitude of the response was significantly reduced (Fig. 4 and Table 3). Xestospongin C was a potent inhibitor of carbachol-stimulated [Ca²⁺]_i increases (Fig. 4). Similar to results with A1254, xestospongin C and thapsigargin also reduced Ca²⁺ transients initiated by 10 µM 2,2'-dichlorobiphenyl (PCB 4), although the percentage of responding cells was not attenuated (Table 3). As a final demonstration that A1254 was causing release of Ca²⁺ from an IP₃-sensitive Ca²⁺ pool, pretreatment of cells with 1 mM carbachol in Ca²⁺-free buffer (to prevent replenishment of the IP₃-sensitive stores as they were emptied) led to a complete inhibition of the Ca²⁺ transient when A1254 was subsequently applied (Fig. 4 and Table 3). Overall, these results indicated that the initial Ca²⁺ signals stimulated by ortho-substituted PCBs in A1254 in DIV 4 to 6 cortical cells were due to release of IP₃-sensitive Ca²⁺ stores. These findings also distinguish the mechanism for this early Ca²⁺ disturbance induced by PCBs from that of the latent Ca²⁺ disturbance/oscillations identified previously where Ca²⁺ entry through L-type channels and excitatory glutamate receptors occurred (Inglefield and Shafer, 2000b).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   A1254 induces release of Ca2+ from IP3-sensitive intracellular stores. The ability of A1254 to release Ca2+ from endoplasmic reticulum, IP3-, and ryanodine-sensitive stores was examined using specific pharmacologic antagonists. In the top row, thapsigargin (10 µM) was used to deplete the endoplasmic reticulum of Ca2+ before addition of 20 µM A1254 (left, top row) or the positive control, carbachol (CCh; 1 mM; center, top row). In the middle row, the IP3 receptor antagonist, xestospongin C (Xes C;1 µM) was used to block the release of IP3-sensitive Ca2+ stores from the endoplasmic reticulum before addition of 20 µM A1254 (left, middle row) or 1 mM carbachol (center, middle row). In the bottom row, ryanodine (100 µM) pretreatment was used to block the release of ryanodine-sensitive Ca2+ stores in the endoplasmic reticulum before addition of 20 µM A1254 (left, bottom row) or 20 mM caffeine (center, bottom row). The right panel in all three rows illustrates the response of cells to positive controls for each condition in the absence of antagonist (thapsigargin, xestospongin C, or ryanodine). The middle row, right panel also shows the effect of depleting carbachol-sensitive stores and preventing their refilling in Ca2+-free buffer before the addition of A1254. Thapsigargin and xestospongin C, but not ryanodine, treatments were effective at inhibiting the A1254-induced Ca2+ transient. Each panel illustrates the typical response to the indicated treatment(s) in an individual cell. All responses are taken from separate coverslips.

Association of the Initial Release of Intracellular Ca²⁺ Stores Caused by PCBs with Latent Ca²⁺ Disturbances. Previous work has shown a significant increase in basal Ca²⁺ levels arising over the course of a 1 h exposure to 10 or 20 µM A1254 [often with recurring Ca²⁺ oscillations (see Fig. 1A)] (Inglefield and Shafer, 2000a). Following the initial Ca²⁺ transient in the cortical cells with A1254, ~70% of cells later exhibited disturbances of Ca²⁺ homeostasis, i.e., increases in basal Ca²⁺. The dependence of latent Ca²⁺ disturbances on the initial Ca²⁺ transient was investigated by pretreating cells with thapsigargin (10 µM) or xestospongin C (1 µM) and assessing changes in basal [Ca²⁺]_i after a 30-min exposure to 20 µM A1254. As shown in Ca²⁺ traces (Fig. 5, A and B) and the bar graph (Fig. 5C), pretreatment with either thapsigargin or xestospongin before A1254 was effective at blocking the A1254-induced Ca²⁺ transient; both also significantly reduced the elevations in basal Ca²⁺ after a 0.5-h of A1254 exposure. These results suggested that the Ca²⁺ transient caused by A1254 is associated with later dysregulation of Ca²⁺ levels in the same cells.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Selected pretreatments (pre-tx) that attenuate the initial release of intracellular Ca2+ stores induced by A1254 also inhibit latent Ca2+ disturbances. Traces illustrate Ca2+ levels during the first 0.5 h in representative cells upon exposure to 20 µM A1254 alone (top) or with xestospongin C (Xesto.; 1 µM) (middle), which was added to the bath 10 min before the exposure. Values in the graph represent the increase in basal [Ca2+]i, measured 30 min into A1254 exposure. The pretreatments were 10 min, and the agents remained in the bathing buffer during the A1254 exposure. For the thapsigargin pretreatment, n = 24 cells. *p < 0.05 relative to no pretreatment. Thapsi., thapsigargin.

Consequences of PCB Actions: I. Delayed Cytotoxicity That Is Not Apoptotic. In previous studies, exposure (1-4 h) to PCBs in the range of 20 µM (Inglefield and Shafer, 2000a) to 50 µM (reviewed in Tilson and Kodavanti, 1997) have not been reported to cause acute cytotoxicity, with one notable exception (Carpenter et al., 1997). However, data on the status of cell viability after prolonged in vitro exposure to PCBs are lacking, yet important, based on the ability of identified alterations in Ca²⁺ homeostasis in the endoplasmic reticulum to contribute to neuronal apoptosis and excitotoxicity (Mattson et al., 2000). The present study, conducted using Trypan blue staining after 24 h of A1254 exposure, found a decrease in cell viability by A1254 that was maturation-sensitive; cytotoxicity became prevalent as the A1254 concentration and in vitro age of the culture increased (Fig. 6A). Compared with DMSO-containing controls, concentrations of A1254 up to 20 µM (24 h) were not associated with significant cytotoxicity when assessed on DIV 4, and on DIV 6 only the highest concentration of A1254 caused moderate (approximately 21%), but significant cytotoxicity. By contrast, significant cytotoxicity was observed in DIV 7 cultures treated with the 10 and 20 µM A1254 for the previous 24 h (~20 and ~50% for these concentrations, respectively).

View larger version (71K): [in this window] [in a new window]   Fig. 6.   Concentration-dependent effects of subacute exposure to A1254 on cellular viability in mixed cortical culture: lack of apoptosis. A, the counts of Trypan blue excluding (viable) cells (from six different cultures) exposed to A1254 (2-20 µM, 24 h) on several DIV. Because a decrease in the cell density of control wells occurs with increasing DIV, the experimental and statistical design took this into account and data were analyzed with a two-way ANOVA, and the interaction of DIV and [A1254] was followed with one-way ANOVA and Dunnett's t tests *p < 0.05, relative to 0 µM A1254 for respective DIV). B, activity of caspase 3 in DIV 7 cells was assessed after 6 and 8 h of exposure to A1254 (0-20 µM) or the proapoptotic agent, staurosporine (StSp; 1 µM). Because there was no time-associated effect of the A1254-induced activity, data from these exposures were combined. C, differential labeling of TUNEL-stained DIV 7 neurons following 24 h exposure to A1254 (20 µM) or staurosporine (1 µM), also on DIV 7. Note TUNEL-negative cells among a single-positive cell following A1254 (20 µM) treatment, whereas the majority of StSp-treated cells are TUNEL-stained, indicative of apoptotic death. Bar = 50 µm. D, normal and apoptotic (TUNEL-positive) cells were scored in at least three separate coverslips from two different cultures by counting the cells in two distinct areas comprising >100 neurons. In marked contrast to ~40% of TUNEL-positive cells produced by staurosporine, the numbers of TUNEL-positive cells in the A1254-treated were equivalent to that in control cultures exposed to the diluent, 0.1% DMSO, also for 24 h. *p < 0.05 following one-way ANOVA and Dunnett's t test.

Consequences of PCB Actions: II. CREB Phosphorylation. Because cell viability is not compromised in the majority (80% with <20 µM A1254) of the population under the exposure conditions examined in these studies, ramifications other than cytotoxicity were also examined. CREB is an important transcription factor that is sensitive to Ca²⁺ signals (reviewed in Silva et al., 1998 and Curtis and Finkbeiner, 1999). Given the perturbations of Ca²⁺ homeostasis induced by A1254, studies were conducted to determine whether or not CREB phosphorylation (activation) was induced following A1254 exposure. Immunoblots using antibodies selective for the phosphorylated form of CREB (pCREB) and CREB (total expression) in DIV 6 cultures, demonstrated that pCREB was increased within 20 min of A1254 addition (Fig. 7A). Levels of pCREB reached maximal levels by 40 to 60 min of exposure to 10 µM A1254. A1254 did not induce de novo synthesis of CREB because the total amount of CREB did not increase as a result of the exposures (Fig. 7A). A1254 also led to concentration-dependent increases in pCREB activation, reaching a 2-fold induction with 10 and 20 µM (Fig. 7B). In comparison, glutamate (100 µM for 40 min) maximally stimulated pCREB ~6-fold, whereas the ortho-substituted PCB congener 4 (2,2'-DCB, 10 µM) stimulated pCREB 1.3 ± 0.1-fold (data not shown). In experiments conducted using DIV 4 cultures, A1254 induced CREB phosphorylation to levels similar to those observed using DIV 6 cultures (data not shown). These data demonstrate that CREB phosphorylation is not specifically linked to cytotoxicity because no cytotoxicity was observed in DIV 4 cultures.

View larger version (35K): [in this window] [in a new window]   Fig. 7.   Activation of CREB by A1254 in developing cortical cells. A, increased levels of phosphorylated CREB follow stimulation with A1254 (10 µM; 3 ppm) for the times shown (in min). CREB phosphorylation on serine-133 was assessed by Western blotting using antibodies selective for the phosphorylated form (top gel) or total CREB (bottom gel) from the same samples as described under Materials and Methods. B, quantitation of the selective increase in phosphorylation of serine-133 of CREB in response to different concentrations (2,10, and 20 µM) following 40-min exposure to A1254. For each A1254 concentration, values were normalized to the total amount of pCREB or CREB in the respective control. The pCREB/CREB ratio indicates a relative increase in phosphorylation of CREB normalized to that for control. n = 4 separate experiments. *p < 0.05 following one-way ANOVA and Dunnett's t test.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have identified several novel actions of individual ortho-substituted PCBs and a PCB mixture, A1254, in intact cells cultured from neonatal rat cerebral cortex. The results, including immediate Ca²⁺ release from IP₃-sensitive stores with subsequent activation of store-operated Ca²⁺ influx (also referred to as capacitative Ca²⁺ entry), further perturbation of Ca²⁺ homeostasis, and activation of a nuclear transcription factor (Fig. 8), may aid understanding the mode of cellular action of persistent, bioaccumulating toxicants such as PCBs. These findings are consistent with other known cellular actions of PCBs, wherein changes in intracellular Ca²⁺ homeostasis is a recurring finding in a variety of intact cell types (Carpenter et al., 1997; Voie and Fonnum, 1998; Bae et al., 1999b; Fischer et al., 1999). However, further consequences as in the status of cell viability after subacute exposure has received less attention. The concentrations of 10 µM and lower used here for the subacute exposure of cortical culture to A1254 are not unrealistic because rats dosed perinatally with A1254 (with 6 mg/kg via the dam) achieved levels in the frontal cortex on postnatal day 21 of 2.4 ppm (~7.2 µM) (Crofton et al., 2000). Previous in vitro investigations have used PCB concentrations higher than 20 µM. The demonstration of Ca²⁺ transients at PCB concentrations as low as 1 µM is among the lowest reported effects in intact cells. Future mechanistic studies exploring subacute exposure should take into account the present findings of decreased cell viability in the continuous presence of 20 µM PCBs.

View larger version (82K): [in this window] [in a new window]   Fig. 8.   Diagram of the temporal order of activation of the intracellular Ca2+ changes influenced by A1254 or ortho-substituted PCBs in cortical culture. A1254/PCBs release Ca2+ from IP3-releasable endoplasmic reticular (ER) stores (1). This results in Ca2+ entry via SOC channels (2) to refill the endoplasmic reticulum Ca2+ stores. These initial events appear to sensitize the postsynaptic cell to released excitatory amino acid neurotransmitters, resulting in latent perturbation of Ca2+ homeostasis (3). [Participation of the L-type VGCC and ionotropic glutamate receptor (GluR) was identified in Inglefield and Shafer (2000b).] Putative consequences of this heightened Ca2+ signaling include phosphorylation of nuclear CREB and inhibition of GABAA receptor responses (Inglefield and Shafer, 2000a, not shown). Whether PCBs directly activate IP3 production by a phospholipase C pathway or act on IP3-releasable Ca2+ stores themselves is not known (question marks). AP, action potential; P, phosphorylated.

A1254-Mediated Release of IP₃ Receptor-Linked Intracellular Ca²⁺ Stores and Interaction with IP Signaling. The initial Ca²⁺ transient induced by A1254 or ortho-substituted PCB congeners (PCB 4 and PCB 138) in cortical cells is analogous to that induced shortly after ortho-substituted PCB addition in granulocytes where a single transient and slow decay to baseline also is seen (Voie and Fonnum, 1998). However, the findings of an early, transient change in basal [Ca²⁺]_i is not universal in investigations with intact cells and exposure to ortho-substituted PCBs. The early-onset Ca²⁺ disturbances we observed for PCB 4 and PCB 138 in >70% of cortical cells occurred more rapidly than occurs in cultured cerebellar granule cells upon exposure to PCB 4 (reviewed in Tilson and Kodavanti, 1997; Mundy et al., 1999). In contrast to our finding that non-ortho-substituted PCBs (15, 77, and 126) failed to induce an initial Ca²⁺ transient, a rapid Ca²⁺ increase has been reported in intact hippocampal cells given dioxin (Hanneman et al., 1996), which is structurally similar to non-ortho-substituted PCBs. A number of factors could contribute to these differences, including different cell types used, maturity of the cells at time of testing, or different concentrations of the compound under examination.

On the Consequences of A1254-Induced Disturbances in Endoplasmic Reticulum Ca²⁺ Homeostasis. Both excitotoxicity (which can lead to necrosis) and apoptosis are possible outcomes of alterations in neuronal Ca²⁺ homeostasis. Certain proteins from central nervous system infections, such as Tat (a human immunodeficiency virus type-1 protein), cause disturbances in endoplasmic reticulum Ca²⁺ homeostasis that are considered central to later apoptotic cell death (New et al., 1997; Kruman et al., 1998; Haughey et al., 1999). Using calcium imaging of cortical neurons that had either a thapsigargin or xestospongin C pretreatment to attenuate the A1254-induced initial Ca²⁺ transient, we identified a role of the initial mechanism of PCB action in the subsequent prolonged Ca²⁺ disturbances. This is in agreement with the recognition of an association of the two phases of Ca²⁺ disturbance induced by Tat in human embryonic neurons maintained in culture (Haughey et al., 1999). These latent Ca²⁺ changes are mediated by heightened Ca²⁺ influx across the plasma membrane, as seen after stimulation with PCBs (Inglefield and Shafer, 2000b) and with Tat (Haughey et al., 1999). From the relationship identified between the Ca²⁺ mobilization from intracellular stores induced by A1254 (or Tat) and the facilitation of later Ca²⁺ dysregulation, it would appear that mobilization of intracellular Ca²⁺ stores by a "stressor" serves to change the "gain" of signaling proteins at the plasma membrane. It is also possible that the replenishing of depleted IP₃-sensitive Ca²⁺ stores contributes in some way to latent increases in basal [Ca²⁺]_i or to the activation of Ca²⁺-sensitive second messengers that alter the responsiveness of plasma membrane receptors. In contrast to the prevalent cell death that is produced by Tat in vitro, 24 h of A1254 exposure led to cell death only in a subset of neurons under specific conditions; survival of the youngest cells (DIV 3-4) across all A1254 concentrations tested was not different from control cultures, whereas cytotoxicity from A1254 occurred in a subset of neurons as the neurons matured (i.e., DIV 7). However, despite the noted similarities in the Ca²⁺ disturbing mechanisms that occur with Tat and PCBs, under the conditions tested here, we obtained no data to support an apoptotic cascade induced by A1254.

CREB Phosphorylation in Cortical Cells by A1254 and Implications. CREB phosphorylation serves as a convergence of Ca²⁺ signaling pathways in neurons and is believed important for neuronal development, as well as learning and memory processes (reviewed in Silva et al., 1998 and Curtis and Finkbeiner, 1999). These are processes negatively impacted by PCBs (Altmann et al., 1998; Niemi et al., 1998; Gilbert and Crofton, 1999). Given the ability of A1254 to cause immediate (present data) and prolonged (present data; Inglefield and Shafer, 2000a,b) perturbations of intracellular Ca²⁺ homeostasis in cortical neurons, we hypothesized that these perturbations may cause activation of this transcription factor. The increase of pCREB levels in samples exposed to A1254, at concentrations that did not induce apoptotic activity nor appreciable cell death at 24 h, suggested that the effect was not a cytotoxic response. In agreement with our findings with DIV 6 cells on activation of pCREB by PCBs, CREB was also phosphorylated to the same degree when cells were given A1254 on earlier DIV (data not shown), when there was no cytotoxicity. At present, it is not known whether this CREB phosphorylation by PCBs necessarily leads to gene transcription.

Conclusion. A release of IP₃ receptor-linked intracellular Ca²⁺ stores in rat cortical neuronal culture was found to precede latent Ca²⁺ disturbances involving L-type Ca²⁺ channels/glutamate receptors that arise in the continued presence of this toxicant (Bae et al., 1999b; Inglefield and Shafer, 2000b). It is clear from the present study that apoptosis is not the message conveyed by this period of induced Ca²⁺ signals, as opposed to the proapoptotic effects of a neurotoxin (Tat) that elicits very similar Ca²⁺ signals. Instead, alteration of a nuclear transcription factor in cerebral cortical cultures occurs with this PCB mixture. This may implicate phenotypic changes in developing nervous system tissue in vivo as the net effect of certain PCBs. As for PCB actions on signal transduction pathways leading to gene transcription, PCBs ultimately may regulate processes such as neurotransmitter phenotype or neuronal plasticity.