PCB-95 Modulates the Calcium-Dependent Signaling Pathway Responsible for Activity-Dependent Dendritic Growth

1Dendritic growth is strongly influenced by neuronal activity as evidenced by the remarkable effect of experience on the development and refinement of synaptic connections, which not only patterns neural circuitry during development but also underlies associative learning. The effects of neuronal activity on dendritic growth are mediated primarily, if not exclusively, by changes in intracellular calcium ion (Ca2+) levels. We previously identified an NMDA (N-methyl-d-aspartate) receptor-mediated Ca2+-dependent signaling pathway that couples neuronal activity to dendritic arborization via sequential activation of Ca2+/calmodulin-dependent protein kinase kinase (CaMKK), Ca2+/calmodulin kinase-I (CaMKI), and extracellular signal-regulated kinase kinase (MEK/ERK) to enhance the cAMP response element binding protein (CREB)-mediated transcription of Wnt2 [wingless-type mouse mammary tumor virus (MMTV) integration site family member 2].

2Polychlorinated biphenyls (PCBs) alter neuronal connectivity by interfering with normal patterns of dendritic growth and plasticity. Non-dioxin-like (NDL) PCBs promote dendritic growth via ryanodine receptor (RyR)-dependent mechanisms; however, the downstream signaling events that mediate RyR-dependent effects of PCBs on dendritic arborization have yet to be described. In this study, we tested the hypothesis that the CaMKI–CREB–Wnt2 signaling pathway responsible for activity-dependent dendritic growth also couples NDL PCB-enhanced RyR activity to dendritic arborization. This hypothesis derives from observations that a) potentiation of RyR activity by NDL PCBs amplifies ionotropic glutamate receptor signaling and increases intracellular Ca2+ levels in neurons; b) Ca2+ signals generated by stimuli that promote dendritic growth, such as repetitive or prolonged depolarization or brain-derived neurotrophic factor (BDNF), are derived primarily from RyR-mediated Ca2+ release; and c) RyR-dependent Ca2+ signals activate CaM kinases, CREB, and transcription of genes encoding Ca2+-regulated proteins (;; W Li et al. 1998). Our findings demonstrate that in cultured rat hippocampal neurons, RyR activity, triggered by either NDL PCBs or neuronal activity, contributes to dynamic remodeling of dendritic architecture via CaMKI–CREB–Wnt2 signaling.

3A complete listing of reagents is provided in Supplemental Material, p. 3 (http://dx.doi.org/10.1289/ehp.1104833). Animals were treated humanely and with regard for alleviation of suffering according to protocols approved by the Institutional Animal Care and Use Committees of Oregon Health & Science University, University of California, Davis, and Washington State University, Pullman.

4Cell culture. Hippocampal neurons were dissociated from postnatal day–1 Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA) and cultured at high density (105 cells/cm2) in Neurobasal-A medium (Invitrogen, Carlsbad, CA) supplemented with B27 (Invitrogen) as described previously. To visualize dendritic arbors, cultures were transfected at 6 days in vitro (DIV) with the plasmid-encoding microtubule-associated-protein-2B MAP2B (which labels the somatodendritic domain) fused to enhanced green fluorescent protein (EGFP) using Lipofectamine-2000 (Invitrogen) according to the manufacturer’s protocol. A subset of cultures was simultaneously transfected with plasmids encoding dominant negative (dn) CaMKI (dnCaMKI), dnCREB (also referred to as ACREB), or Wnt inhibitory factor (Wif). PCBs or vehicle (DMSO at 1:1000 dilution) was added to the culture medium for 48 hr beginning at 7 DIV; in a subset of cultures, a CaMK kinase inhibitor (STO-609, 5 μM) or a MEK inhibitor (U0126, 10 μM) was also added to the medium during the same period.

5Organotypic hippocampal slices from postnatal day–5 rats were cultured for 3 days as described previously. At 3 DIV, slice cultures were biolistically transfected with plasmid-encoding tomato fluorescent protein (TFP) using the Helios gene gun (Bio-Rad, Hercules, CA) per the manufacturer’s directions. A subset of slice cultures was simultaneously transfected with siRNA (small interfering RNA) specific for RyR1 or RyR2. Slice cultures were exposed to vehicle, and PCBs were added to the culture medium during 4–6 DIV. A subset of cultures was also exposed to FLA365 [4-(2-aminopropyl)-3,5-dichloro-N,N-dimethylaniline] (10 μM), which was added to the culture medium during the same period.

6Dendritic morphology was quantified from digital images of green fluorescent protein–positive (GFP+) or TFP+ neurons using Image J version 1.44p with the Neuron J plug-in version 1.42 to trace neurons.

7Calcium imaging. Spontaneous and electrically evoked Ca2+ transients were measured in dissociated hippocampal neurons cultured on Greiner CELLSTAR® micro-clear wells (Sigma-Aldrich, St. Louis, MO). Cells were loaded with the Ca2+-sensitive dye Fluo-4 AM (5 µM; Invitrogen) at 37°C for 30 min in imaging buffer consisting of 140 mM sodium choride (NaCl), 5 mM potassium chloride (KCl), 2 mM magnesium chloride (MgCl2), 2 mM calcium chloride (CaCl2), 10 mM HEPES, and 10 mM glucose, at pH 7.4, and supplemented with 0.05% BSA (bovine serum albumin). Cultures were washed three times with imaging buffer and transferred to the stage of an inverted Olympus IX70 microscope (Olympus America, Center Valley, PA) equipped with a 60 × 1.25 numeric aperture objective. Fluo-4 was excited at 494 nm using a DeltaRam illuminator (Photon Technologies Int’l., Birmingham, NJ); fluorescence emission was captured at 510 nm. Full-frame images were captured with an Evolve® cooled charge coupled device camera (Photometrics, Tucson, AZ) at 30 frames/sec (fps) using EasyRatioPro software (Photon Technologies Int’l.). In a subset of experiments, cultures were exposed to PCB-95 (2,2´,3,5´6-pentachlorobiphenyl; 2, 20, or 200 nM) from 7–9 DIV before loading with Fluo-4. After baseline recording, cultures were sequentially stimulated with electrical bipolar field pulses (0.5 millisec) at 1, 2.5, 5, and 10 Hz for 10 sec with 50-sec interstimulus rest periods using platinum electrodes connected to a Master 8 stimulator (A.M.P.I, Jerusalem, Israel). After acquisition, regions of interest were drawn freehand to encompass soma and distal dendrites (separated from the soma by a length of > 2 times the soma diameter). Movies were replayed to quantitatively measure changes in Fluo-4 fluorescence within the regions of interest.

8At 7 DIV in separate experiments, spontaneous synchronized Ca2+ transients were measured from hippocampal neurons that had not been previously exposed to PCB-95 before loading with Fluo-A. PCB-95 (200 nM) or vehicle (DMSO at 1:1,000 dilution) in imaging buffer was acutely introduced into cultures by continuous flow perfusion (~ 1 mL/min). In a subset of cultures, ryanodine (500 µM; EMD Biosciences, Philadelphia, PA) was added to the culture medium 1 hr before loading the cells with Fluo-A to irreversibly block RyR channel activity, which was verified by a challenge with 4-chloro-m-cresol (4CmC, 100 µM; Sigma-Aldrich). Changes in cytoplasmic Ca2+ were continuously recorded at 30 fps. Ca2+ transients > 2 times the baseline amplitude were scored as oscillations. The number of oscillations was compared between vehicle and PCB-95–treated neurons. Transient amplitude was measured by normalizing peak change in Fluo-A fluorescence (ΔF) to the fluorescence baseline (F0) and presented as mean ΔF/F0 for each neuron included in the analysis. Statistical comparisons were made using neurons (n = 30) obtained from three separate dissections. Statistical analysis was performed using unpaired Student’s t test.

9Quantitative polymerase chain reaction (qPCR). Total RNA was isolated from dissociated hippocampal neuron cultures (9 DIV) using Trizol (Invitrogen) according to manufacturer’s instructions. Levels of Wnt2 mRNA were quantified by qPCR and normalized to GAPDH (glyceraldehyde 3-phosphate dehydrogenase gene) mRNA levels in the same sample. Primer sequences and a more detailed description are provided in Supplemental Material, p. 4 (http://dx.doi.org/10.1289/ehp.1104833).

10PCB-95 alters Ca2+ signals in cultured hippocampal neurons. We used real-time Ca2+ imaging techniques to examine how subchronic (48 hr) or acute exposure to PCB-95 influences Ca2+ signaling behaviors in cultured hippocampal neurons. Exposure to PCB-95 during 7–9 DIV caused a concentration-dependent attenuation of the amplitude of electrically evoked Ca2+ transients (Figure 1A,B). This observation is consistent with previous findings that prolonged RyR activation by PCB-95 can deplete RyR-sensitive Ca2+ stores. Evidence that the GABAA (type A γ-aminobutyric acid) receptor antagonist bicuculline (BIC) triggers dendritic growth via transient rather than prolonged activation of Ca2+-dependent signaling pathways suggests that PCB-95 may similarly trigger dendritic growth via amplification of Ca2+ signals on a shorter time scale. We therefore examined Ca2+ oscillatory behavior in PCB-naïve hippocampal neurons loaded with the Ca2+ indicator Fluo-A at 7 DIV immediately before focal applications of vehicle or PCB-95. Although vehicle alone had no influence (Figure 2A), acute application of PCB-95 at 200 nM, a concentration that promotes robust dendritic growth, significantly enhanced the amplitude and oscillations of Ca2+ transients in both soma and distal dendrites of cultured rat hippocampal neurons (Figure 2B, top traces). Amplification of Ca2+ oscillations by PCB-95 was dependent on RyR activity as oscillations were completely prevented by pre-incubation of neurons with ryanodine (Figure 2B, bottom traces) under conditions previously shown to completely block gating of RyRs (500 µM, 1 hr). To verify that ryanodine pretreatment in fact blocked all RyR channel activity in hippocampal neurons, a challenge of 4CmC [previously shown to activate RyR1 and RyR2, but not RyR3, channel activity], was performed at the end of each recording. Ryanodine pretreatment resulted in a near complete block of 4CmC responses (Figure 2C).

11CaMKI–CREB–Wnt2 signaling mediates PCB-95–induced dendritic growth. Previous reports demonstrated that treatment of cultured hippocampal neurons with the CaMK kinase inhibitor STO-609 (2–5 μM) inhibits CaMKI but not CaMKII. STO-609 (5 μM) selectively blocked PCB-95 effects on dendritic length (Figure 3A) and branching (Figure 3B) without altering basal dendritic growth. Expression of a previously characterized dnCaMKI, also suppressed the stimulatory effects of PCB-95 on dendritic length and branching with no effect on basal parameters (Figure 3A,B). To determine which CaMKI isoforms may be involved in PCB-95–induced dendritic growth, previously characterized small hairpin RNA (shRNA) constructs that suppress individual CaMKI isoforms were expressed in hippocampal neurons. Expression of CaMKIα shRNA significantly reduced and CaMKIγ shRNA blocked the stimulatory effects of PCB-95 on dendrites; in contrast, shRNAs for β and δ CaMKI had no effect (Figure 3C).

12To probe downstream effectors of CaMKI, we first determined whether PCB-95–induced dendritic growth requires MEK/ERK signaling. Western blot analyses of phosphorylated ERK (pERK) levels did not consistently detect significantly increased pERK at 5–30 min after acute exposure of 7-DIV hippocampal cultures to PCB-95 (data not shown). However, the MEK inhibitor U0126 (10 μM) completely blocked the stimulatory effects of PCB-95 on dendritic length and branching in the absence of any effect on basal dendritic growth (Figure 3D,E).

13Western blot analyses revealed that, as previously reported, BIC increased phosphorylation of the active site (Ser133) of CREB (Figure 4A,B). PCB-95 similarly elicited rapid (< 5 min) and sustained (> 30 min) CREB phosphorylation. Expression of either dnCREB or CREB-specific shRNA blocked the effects of PCB-95 on dendritic length and branching but had no effect on dendritic growth in vehicle control cultures (Figure 4C,D). We previously showed that expression of either dnCREB or shCREB (short hairpin CREB) does not induce nuclear pyknosis, increase levels of cleaved caspase-3, or significantly alter MAP2B-EGFP expression or dendritic morphology in hippocampal neurons grown in the presence of the serum-free supplement, B27.

14Activity-dependent CREB activation stimulates dendritic growth in part via increased Wnt2 expression. Stimulation of hippocampal neurons with PCB-95 triggered a doubling of Wnt2 mRNA and this effect was blocked by treatment with the selective RyR antagonist FLA365 (Figure 4E). Expression of the Wnt inhibitor Wif, a secreted protein that binds Wnt and prevents it from activating its receptor Frizzled, completely suppressed PCB-induced dendritic growth without altering basal dendritic growth (Figure 4F,G).

15BIC-induced dendritic growth requires RyR activity. These data raised the question of whether RyR activity is required for activity-dependent dendritic growth in the absence of PCBs. To address this question we used our previously described model of activity-dependent dendritic growth in which the GABAA receptor antagonist BIC promotes dendritic growth in cultured hippocampal neurons via activation of NMDA receptors. Consistent with our previous studies, BIC (20 μM) enhanced dendritic growth in cultures of dissociated hippocampal neurons (Figure 5A). BIC-induced dendritic growth was completely blocked by the RyR channel blocker FLA365 (Figure 5A). Expression of siRNA specific for either RyR1 or RyR2 [for characterization of siRNA specificity, see Supplemental Material, p. 4 and Figure S1 (http://dx.doi.org/10.1289/ehp.1104833)] completely suppressed BIC stimulation of dendritic length and branching but had no effect on basal dendritic growth (Figure 5B–D). Similarly, BIC enhanced dendritic arborization of pyramidal neurons in hippocampal slice cultures, and this effect was blocked by FLA365 and by siRNA knockdown of RyR1 or RyR2 (Figure 5E).

16The major finding of this study is that in cultured rat hippocampal neurons, RyR activity, triggered by either NDL PCBs or neuronal activity, contributes to dynamic remodeling of dendritic architecture via Ca2+-dependent activation of the CaMKI–CREB–Wnt2 signaling pathway. We previously demonstrated that activity-dependent dendritic growth in hippocampal neurons is mediated by sequential activation of NMDA receptors, CaMKK, CaMKI, Ras, MEK/ERK, and CREB-dependent transcription of Wnt2. In this study, we confirmed that PCB-95 increases spontaneous Ca2+ oscillations in hippocampal neurons in a manner blocked by ryanodine, and we identified these same signaling molecules as downstream effectors that couple RyR activation to dendritic growth. Previous studies in cultured neurons demonstrated that the commercial PCB mixture Aroclor 1254 (A1254) and RyR activity increased neuronal levels of pCREB. We confirmed that PCB-95 activated CREB in hippocampal neurons as determined by Western blot analyses of pCREB expression, and we extended this finding to demonstrate that PCB-95 also up-regulated expression of the CREB-responsive gene Wnt2, as determined by qPCR. More importantly, we established a crucial role for not only CREB and Wnt2, but also CaMKK–CaMKI and MEK/ERK signaling in RyR-dependent dendritic growth using multiple experimental approaches including pharmacological inhibition and expression of dominant-interfering or si/shRNA constructs specific for each of these signaling molecules. Consistent with our earlier studies of activity-dependent dendritic growth in cultured hippocampal neurons, shRNAs against both α and γ CaMKI blocked the stimulatory effects of PCB-95 on dendrites, whereas shRNAs for β and δ CaMKI had no effect.

17These data suggest a model in which PCB-95 potentiation of RyR activity promotes dendritic growth by triggering the same Ca2+-dependent signaling pathway linked to activity-dependent dendritic growth in hippocampal neurons. We propose that PCB-95 interacts with RyR channels to increase the amplitude and frequency of intracellular Ca2+ oscillations, which then trigger sequential activation of CaMKK, CaMKI, and MEK/ERK, resulting in the CREB-activated transcription of Wnt2 (Figure 6). Wnt2, a secreted glycoprotein, binds to the Frizzled family of receptors to activate the scaffold protein Dishevelled, which in turn stimulates dendritic growth via effects on the actin and microtubule cytoskeleton mediated by β-catenin and/or activation of the RhoGTPases Rac and/or JNK [members of the rho family of guanosine triphosphate (GTP) hydrolases]. In our model, PCB-95 triggers this signaling pathway downstream of the NMDA receptor. However, RyRs are functionally coupled to NMDA receptors, which considered in the context of our previous studies demonstrating that inhibition of NMDA receptors blocks activity-dependent dendritic growth, suggested that RyR activity is required for not only PCB-induced dendritic growth, but also activity-dependent dendritic growth. In support of this hypothesis, we observed that the selective RyR antagonist, FLA365, as well as siRNA knockdown of either RyR1 or RyR2, completely blocked BIC-induced dendritic growth in both dissociated cultures of hippocampal neurons and hippocampal slice cultures.

18Our observations add to the emerging experimental evidence supporting a central role for RyRs in neuronal excitability and use-dependent synaptic plasticity. Functional aspects of neuroplasticity, including long-term potentiation, long-term depression (ST; ) and associative memory are altered by experimental manipulation of RyR activity by ligands such as the RyR accessory FK506-binding protein (FKBP12) that directly modulate RyR or by agents that deregulate RyR by disrupting the RyR/FKBP12 complex. Spatial learning is tightly correlated with selective up-regulation of RyR expression. Targeted deletion of RyR3 in mice causes impairments in social behavior and deficits in contextual fear conditioning but improves spatial learning in the Morris water maze task, whereas selective knockdown of RyR2 and RyR3 impairs avoidance memory processes. Previous studies have implicated a local release of Ca2+ from intracellular stores in the maintenance of dendrites, the regulation of the motility of dendritic filopodia, and the increased size of dendritic spines. More recently, inhibitory concentrations of ryanodine were reported to block BDNF-enhanced spine formation in primary hippocampal neurons. However, our findings are the first to demonstrate that RyR activity is sufficient to promote dendritic growth and is necessary for activity-dependent dendritic growth.

19Defective neuronal connectivity is a common pathological feature in most neurodevelopmental disorders, including autism spectrum disorders (ASD), which suggests that NDL PCBs are environmental risk factors for ASD. There is experimental evidence that developmental exposure to PCB-95 elicits some aspects of ASD, including an imbalance between excitation and inhibition in the auditory cortex of weanling rats and altered social behaviors in rats. However, it seems more likely that NDL PCBs interact with ASD-linked genes that converge on common signaling pathways to increase ASD risk and/or severity. Emerging evidence suggests that defective neuronal connectivity associated with ASD is mediated in part by defects in neuronal Ca2+ signaling. A number of genes that encode proteins that have a primary role of generating intracellular Ca2+ signals or are themselves tightly regulated by local fluctuations in Ca2+ concentrations, including CREB and Wnt2, have been linked to ASD. More compelling are recent studies of a gain-of-function missense mutation in the L-type Ca2+ channel CaV1.2 that causes Timothy syndrome, which has a 60% rate of co-morbidity with autism, making it one of the most penetrant monogenic forms of ASD. Neurons differentiated from induced pluripotent stem cells derived from patients with Timothy syndrome revealed increased Ca2+ oscillations and up-regulated expression of genes linked to Ca2+-dependent regulation of CREB, including CaMK. Such studies clearly establish these Ca2+-signaling molecules as possible convergence points for genetic variants linked to ASD risk and environmental factors such as NDL PCBs.