Nimodipine Activates TrkB Neurotrophin Receptors and Induces Neuroplastic and Neuroprotective Signaling Events in the Mouse Hippocampus and Prefrontal Cortex

Abstract

The L-type calcium channel blocker nimodipine improves clinical outcome produced by delayed cortical ischemia or vasospasm associated with subarachnoid hemorrhage. While vasoactive mechanisms are strongly implicated in these therapeutic actions of nimodipine, we sought to test whether nimodipine might also regulate neurotrophic and neuroplastic signaling events associated with TrkB neurotrophin receptor activation. Adult male mice were acutely treated with vehicle or nimodipine (10 mg/kg, s.c., 1.5 h) after which the phosphorylation states of TrkB, cyclic-AMP response element binding protein (CREB), protein kinase B (Akt), extracellular regulated kinase (ERK), mammalian target of rapamycin (mTor) and p70S6 kinase (p70S6k) from prefrontal cortex and hippocampus were assessed. Nimodipine increased the phosphorylation of the TrkB catalytic domain and the phosphoslipase-Cγ1 (PLCγ1) domain, whereas phosphorylation of the TrkB Shc binding site remained unaltered. Nimodipine-induced TrkB phosphorylation was associated with increased phosphorylation levels of Akt and CREB in the prefrontal cortex and the hippocampus whereas phosphorylation of ERK, mTor and p70S6k remained unaltered. Nimodipine-induced TrkB signaling was not associated with changes in BDNF mRNA or protein levels. These nimodipine-induced changes on TrkB signaling mimic those produced by antidepressant drugs and thus propose common mechanisms and long-term functional consequences for the effects of these medications. This work provides a strong basis for investigating the role of TrkB-associated signaling underlying the neuroprotective and neuroplastic effects of nimodipine in translationally relevant animal models of brain trauma or compromised synaptic plasticity.

Introduction

Subarachnoid hemorrhage (SAH), i.e., bleeding into the subarachnoid space, most commonly occurs following cerebral aneurysm. Every year over 600 000 aneurysmal SAH cases occur worldwide (Feigin et al. 2009), requiring immediate surgical or endovascular intervention and intensive post-operative care. Yet, in a substantial number of cases SAH and its complications can lead to death or severe and persistent neurological and cognitive impairment (Sheldon et al. 2013; Macdonald 2014).

Delayed cerebral ischemia (DCI) caused by early brain injury, vasospasm (angiographic and micro circular), microthrombosis and cortical spreading ischemia are common and serious complications of SAH (Dorsch 2011; Macdonald 2014). The L-type calcium channel antagonist nimodipine effectively improves the outcome of SAH (Pickard et al. 1989; Laursen et al. 1988; Dorhout Mees et al. 2007). Originally the therapeutic actions of nimodipine were associated with its ability to produce vasodilatation via relaxing smooth muscle cells lining the blood vessels (Alborch et al. 1995; Triggle 2006). However, emerging evidence suggests that nimodipine improves clinical outcome even without notable vasoactive effects (Dreier et al. 2002; Ricci et al. 2002; Dreier et al. 2009; Pluta et al. 2009; Etminan et al. 2011; Sehba et al. 2012; Vergouwen et al. 2011; Woitzik et al. 2012). Thus, also non-vasoactive mechanisms may underlie the therapeutic actions of nimodipine.

L-type calcium channels are divided in four subgroups (CaV1.1–CaV1.4.) (Calin-Jageman and Lee 2008). The major isoforms CaV1.2 and CaV1.3 are expressed throughout the nervous system (Hell et al. 1993; Lipscombe et al. 2004). Activity-dependent calcium influxes through these channels regulate several important intracellular signaling pathways that modulate short- and long-term alterations in gene expression, synaptic plasticity and homeostasis in neurons (Bhat et al. 2012; Bading 2013; Park and Poo 2013; Frank 2014). However, excess activation of L-type calcium channels may lead to compromised plasticity, excitotoxicity or even neurodegeneration (Choi 1994; Mattson 2007). Indeed, the neuroprotective properties of nimodipine have been extensively studied and well-characterized in several mammalian models of injury and in human patients (Nuglisch et al. 1990; Regan and Choi 1994; Gepdiremen et al. 1997; Taya et al. 2000; Winkler et al. 2003; Liu et al. 2004; Dorhout Mees et al. 2007; Thomas et al. 2008).

Nimodipine treatment reduces excessive calcium influx in pathological conditions, contributing to its neuroprotective properties (Zornow and Prough 1996; Kobayashi and Mori 1998), however we sought to test whether nimodipine would also regulate neurotrophic and neuroplastic signaling events associated with TrkB neurotrophin receptor activation. Activation of TrkB by its primary ligand BDNF (brain-derived neurotrophic factor) regulates multiple forms of neuronal plasticity. Upon ligand binding, specific intracellular tyrosine residues within TrkB are phosphorylated, which leads to the activation of intracellular signaling pathways implicated in neuronal survival (e.g., Akt, protein kinase B), neuronal differentiation (e.g., ERK, extracellular regulated kinase), synaptic plasticity (e.g., CREB, cAMP related element binding protein) and synapse formation (e.g., mTor, mammalian target of rapamycin; p70S6k, p70S6 kinase) (Huang and Reichardt 2001; Minichiello 2009; Gómez-Palacio-Schjetnan and Escobar 2013; Park and Poo 2013). Importantly, increased BDNF-TrkB signaling improves the clinical outcome in different brain pathologies (Nagahara and Tuszynski 2011). Furthermore, pharmacologically diverse antidepressant drugs induce TrkB signaling in the adult rodent hippocampus (HC) and prefrontal cortex (Saarelainen et al. 2003; Rantamäki et al. 2007). These effects may underlie antidepressant-induced neuroplastic effects that are beneficial against numerous nervous system conditions such as neuropsychiatric and neurodegenerative disorders, as well as stroke (Castrén and Rantamäki 2010; Chollet et al. 2011; Castrén et al. 2012; Castrén 2013; Castrén and Hen 2013). Here we investigated the acute effects of nimodipine on TrkB receptor phosphorylation and downstream signaling in experimental settings that closely follow conditions where antidepressant-induced rapid activation of TrkB signaling has been demonstrated.

Materials and Methods

Animals

Male adult C57BL/6 J mice (8 weeks of age, Harlan, The Netherlands) were used. All experiments were conducted according to the guidelines of the European Communities Council Directive (86/609/EEC) and were approved by the County Administrative board of Southern Finland. These guidelines comply with the guidelines established in Guide for the Care and Use of Laboratory Animals (Institute for Laboratory Animal Research, national Research council. Washington, DC: National Academy Press, 1996).

Drug Treatments and Tissue Sampling

For the drug treatments, animals were gently immobilized and injected subcutaneously (s.c.) with either nimodipine (10 mg/kg, Santa Cruz Biotechnology (SCB), CA, USA) or vehicle (0.5 % Tween-40 in saline) (N = 6 per group) into the neck pouch. All the animals were injected on the same day by the same researcher using 5 min interval between each animal. All the animals were sacrificed 90 min following injection, a lag period that corresponds to TrkB phosphorylation induced by pharmacologically diverse antidepressant drugs (Rantamäki et al. 2007). Mice were stunned with carbon dioxide and sacrificed with rapid cervical dislocation. Next, the HC and medial prefrontal cortex (mPFC) were isolated for extraction of protein and total RNA. Samples for total RNA were frozen immediately with dry ice and stored at −80 °C. For protein extractions, the HC and mPFC were collected on a cooled dish and homogenized in lysis buffer (137 mM NaCl, 20 mM Tris, 1 % NP-40, 10 % glycerol, 48 mM NaF, 2X Complete inhibitor mix (Roche Diagnostics, Hertforshire, UK), and 2 mM Na₃VO₄). After incubation on ice for 15 min, samples were centrifuged (16,100g, 15 min, +4 °C) and the supernatant collected for further analysis.

Western Blot

Sample protein concentrations were measured using a commercial kit (Bio-Rad DC protein assay). Next, 30/40 µg of total protein was separated in a SDS-PAGE under reducing conditions and transferred onto a PDVF membrane (300 mA for 1 h at 4 °C). The membranes were blocked with 3 % bovine serum albumin (1 h, room temperature) and incubated with the following primary antibodies: anti-p-TrkB^Y816 (rabbit polyclonal; 1:1000; kind gift from Dr. M. Chao, Skirball Institute, NY, USA), anti-p-TrkA^Y490/TrkB^Y515 (Y490)/TrkB (Y515) (rabbit polyclonal; 1:1000; #9141 cell signaling technology (CST), MA, USA), anti-Trk (rabbit polyclonal; 1:1000; sc-11, SCB), anti-p-CREB^S133 (rabbit polyclonal; 1:1000; #9191, CST), anti-CREB (rabbit monoclonal; 1:1000; #4820, CST), anti-p-Akt^T308 (rabbit monoclonal;1:1000; #4056, CST), anti-AKT (rabbit polyclonal; 1:1000; #9272, CST), anti-p-44/42 MAPK^T202/Y204 (mouse monoclonal; ERK1/2; 1:1000; #9106, CST), anti-p44/42 MAPK (rabbit polyclonal; ERK1/2; 1:1000; #9102, CST), anti-p-mTOR^S2481 (rabbit polyclonal; 1:1000; #2974, CST), anti-mTOR (rabbit polyclonal; 1:1000; #2972, CST), anti-p-p70S6 kinase^T421/S424 (rabbit polyclonal; 1:1000; #9204, CST), anti-p70S6 kinase (rabbit polyclonal; 1:1000; #9202, CST) and anti-GAPDH (rabbit polyclonal; 1:2000; sc-25778, SCB). After washing, membranes were incubated in HRP conjugated secondary antibody (1:10000, BIO-RAD; 1 h, room temperature) followed by visualization with an enhanced chemiluminescence kit (ECL+, Amersham Biosciences) and the detection of luminescence with Fuji LAS-3000 camera. Phospho-protein detection was always conducted first after which the filter was stripped and probed with the corresponding total protein antibody. GAPDH immunoblotting was used to control equal loading and for the quantification of TrkB protein expression.

BDNF ELISA

Mature BDNF protein levels were assessed using ELISA method (as previously described Karpova et al. 2010). The assay shows no cross-reactivity with other neurotrophins (Karpova et al. 2010) and has been further validated using tissues obtained from BDNF^+/− mice (~50 % expression) and conditional BDNF^−/− mice (undetected). Briefly, transient acidified brain lysates, BDNF standards (7.8–1,000 pg/ml in Hanks; Promega), and POD-conjugated secondary BDNF antibody were transferred to pre-blocked (300 µl; Hanks buffer, 2 % BSA, 0.1 % Triton X-100, 2 h, RT) Maxisorb^® ELISA plates that were previously coated with the primary BDNF antibody. The following day, plates were washed with PBS-T and the POD substrate was added to the wells according to manufacturers instructions (BM Blue; Roche). The colorimetric reaction was stopped within 20 min with 1 M H₂SO₄ (50 µl), and absorbance was immediately measured at 490 nm. BDNF protein was normalized to total protein levels. The r ² for the standard curve was ≥0.99 in all experiments.

BDNF qPCR

We conducted RNA extraction followed by real-time quantitative PCR as previously described (Uutela et al. 2014). Total RNA was extracted from frozen tissues using QIAzol (Qiagen, Valencia, CA) and treated with DNaseI (Thermo Fisher Scientific Inc, Rockford, IL) according to the manufacturer´s instruction. cDNA were synthesized with 1 µg of total RNA using the Maxima First Strand cDNA Synthesis Kit (Thermo Fisher Scientific Inc, Rockford, IL). Real-time quantitative PCR was performed using the Maxima SYBR Green qPCR Master Mix (Thermo Fisher Scientific Inc, Rockford, IL) and the CFX96 Touch™ detection system (Bio-Rad, Hercules, CA). The primers described previously (Karpova et al. 2011) were used to amplify specific cDNA regions of transcripts: the coding region in the exon IX of the Bdnf gene for the total Bdnf mRNA (5′-GAAGGCTGCAGGGGCATAGACAAA-3′ and 5′-TACACAGGAAGTGTCTATCCTTATG-3′); the exon IV (5′-ACCGAAGTATGAAATAACCATAGTAAG-3′ and (5′-TGTTTACTTTGACAAGTAGTGACTGAA-3′), Gapdh (5′-GGTGAAGGTCGGTGTGAACGG-3′ and 5′-ATGTAGTTGAGGTCAATGAAGGG-3′) as a housekeeping control gene. Ct and quantitative values were calculated from each sample using CFX Manager™ software (Bio-Rad, Hercules, CA) and the quantitative values were normalized to the control Gapdh levels.

Data Handling and Statistical Methods

Protein bands were quantified using ImageJ program (NIH). Phospho-protein band intensities were divided by corresponding total protein band intensities. Final values were divided by the control group average and multiplied by 100. All the data are expressed as mean ±SEM (Standard Error of Mean; standard deviation divided by the square root of sample size) and as percentage of control. Statistical tests were performed using the two-tailed Student t test. The criterion for significance was set to p < 0.05.

Results

Nimodipine Rapidly Increases the Phosphorylation TrkB in the Prefrontal Cortex and Hippocampus

Autophosphorylation of the TrkB autocatalytic domain (TrkB^Y705/6) is the initial and essential post-translational modification inducing the kinase activity of TrkB (Segal et al. 1996). Nimodipine treatment readily induced phospho-TrkB^Y705/6 levels in the prefrontal cortex and HC 90 min post-injection (Fig. 1a). However, the phosphorylation state of the Shc binding domain (TrkB^Y515) remained unaltered after the treatment (Fig. 1b). By contrast, the phosphorylation level of the PLCγl-binding tyrosine within TrkB (TrkB^Y816) was significantly increased by nimodipine (Fig. 1c). Intriguingly, these nimodipine-induced changes on TrkB phosphorylation closely resemble those produced by antidepressant drugs (Saarelainen et al. 2003; Rantamäki et al. 2007). Furthermore, and similarly with antidepressant drugs (Rantamäki et al. 2011), the levels of the BDNF protein and total mRNA remained unaltered after nimodipine administration in the mPFC and HC (Fig. 2).

Fig. 1.

Nimodipine activates TrkB in the mouse brain. The effect of acute nimodipine treatment (10 mg/kg, s.c., 1.5 h) on the phosphorylation of TrkB catalytic domain (a), phosphorylation of TrkB Shc binding site (b) and phosphorylation of TrkB PLCγ1 binding site (c). Phospho-TrkB levels are normalized to total Trk protein levels that remained unaltered by the treatment (d). Control group is set to 100 % and all the data are expressed as mean ±standard error of mean (SEM). N = 6 per group. ** <0.01, *<0.05, Student t test. This experiment has been conducted in two set of animals with essentially similar results. CNTR control, NIMO nimodipine, mPFC medial prefrontal cortex, HC hippocampus, pTrkB ^Y705/6-phosphorylated TrkB autocatalytic domain, pTrkB ^Y515-phosphorylated Shc binding domain, pTrkB ^Y816-phosphorylated PLCγ1 binding site, GAPDH glyceraldehyde 3-phosphate dehydrogenase

Fig. 2.

The BDNF levels after acute nimodipine administration. Total Bdnf mRNA (a) and mature BDNF protein (b) levels remain unaltered in the hippocampus and prefrontal cortex after acute nimodipine treatment. Messenger-RNA and protein data are normalized against GAPDH mRNA and total protein, respectively. Control group is set to 100 % and all the data are expressed as mean ±standard error of mean (SEM). N = 6 per group. Student t test was used for statistical analyses. CNTR control, NIMO nimodipine, mPFC medial prefrontal cortex, HC hippocampus

Nimodipine Induces Neuroprotective and Neuroplastic Signaling in the Prefrontal Cortex and Hippocampus

The phosphorylation of TrkB^Y515 serve as a docking site for Shc adaptor proteins that further regulate the phosphorylation and activity of downstream signaling molecules implicated in neuronal apoptosis/survival (Akt) and neuronal differentiation (ERK1/2) (Huang and Reichardt 2001). In line with the unaltered levels of phospho-TrkB^Y515, phosphorylation levels of ERK1/2 in the HC and the prefrontal cortex were indistinguishable in the control and nimodipine-treated animals (Fig. 3a). However, nimodipine treatment readily increased the phosphorylation of Akt in the mouse prefrontal cortex and a trend was also seen in the HC (Fig. 3b). Apart from regulating neuronal apoptosis/survival, Akt has also been linked with the activation of mTor-p70S6k pathway that is implicated in dendritic spine formation and morphology (Kumar et al. 2005). However, the phosphorylation levels of mTor and its downstream target kinase remained unaltered after acute nimodipine administration (data not shown).

Fig. 3.

Nimodipine induces neuroprotective (Akt) and neuroplastic (CREB) signaling cascades in the mouse brain. Phosphorylation levels of ERK1/2 remained unaltered after acute nimodipine treatment (10 mg/kg, s.c., 1.5 h) (a); acute nimodipine treatment increases the phosphorylation of Akt and CREB in the medial prefrontal cortex (b, c). Phospho-protein levels are normalized to respective total protein levels that remained unaltered by the treatment. Control group is set to 100 % and all the data are expressed as mean ±standard error of mean (SEM). N = 6 per group. *<0.05, Student t test. CNTR control, NIMO nimodipine, mPFC medial prefrontal cortex, HC hippocampus, pERK1/2 ^T202/Y204-phosphorylated extracellular regulated kinase, pAkt ^T308-phosphorylated protein kinase B, pCREB ^S133-phosphorylated cAMP related element binding protein

Phosphorylation of TrkB^Y816 leads to the activation PLCγ1 that subsequently increases intracellular calcium mobilization via IP₃ signaling (Huang and Reichardt 2001). This signaling event has been tightly associated with the phosphorylation and activation of CREB, a transcription factor that critically regulates the transcription of genes implicated in synaptic plasticity (e.g., long-term potentiation) (Minichiello 2009). Indeed, and in line with observed changes in phospho-TrkB^Y816 levels, the phosphorylation levels of CREB were significantly increased after nimodipine in both brain areas investigated (Fig. 3c).

Discussion

Nimodipine improves the clinical outcome after aneurysmal SAH (Laursen et al. 1988; Pickard et al. 1989; Dorhout Mees et al. 2007). Pathophysiology of SAH is a complex combination of early brain injury, vasospasm, secondary injuries, and cortical spreading ischemia (Macdonald 2014). Although the therapeutic actions of nimodipine have been considered to be largely mediated by its vasoactive properties, clinical data indicate that also other mechanisms may be involved (Etminan et al. 2011). Indeed, nimodipine possess broad-ranging neuroprotective properties by reducing neuronal and glial apoptosis, increasing fibrinolysis of microthrombosis and inhibiting cortical spreading ischemia (Lazarewicz et al. 1990; Dreier et al. 2002; 2009; Pluta et al. 2009; Etminan et al. 2011; Vergouwen et al. 2011; Hashioka et al. 2012; Sehba et al. 2012; Woitzik et al. 2012). Nimodipine has also been reported to induce synaptogenesis after chronic administration (de Jong et al. 1992).

Given that BDNF and its receptor TrkB regulate neuroprotection and neuroplastic changes in the healthy brain and under pathological conditions, we sought to examine whether nimodipine might also regulate TrkB receptor signaling. Indeed, we show that systemic administration of nimodipine induces the autophosphorylation of TrkB and the activation of downstream signaling implicated in neuronal survival (Akt) and plasticity (CREB). Remarkably, the nimodipine-induced changes on TrkB phosphorylation closely resemble those previously seen after antidepressant drug treatment (Rantamäki et al. 2007; Di Lieto et al. 2012). Both nimodipine and antidepressant drugs specifically induce the phosphorylation of TrkB autocatalytic domain and the PLCγ1 phosphorylation status of the Shc site unaltered. Moreover, the levels of total Bdnf mRNA and mature BDNF protein remained unaltered by acute nimodipine treatment, suggesting that nimodipine does not induce rapid changes in BDNF synthesis. Indeed, although a single-antidepressant drug treatment is sufficient to activate TrkB (Rantamäki et al. 2007; 2011), the levels of BDNF mRNA and protein are increased only after weeks of treatment (Nibuya et al. 1995).

Our findings indicate that common molecular mechanisms mediate TrkB activation after acute nimodipine and antidepressant drug treatments. Interestingly, conventional antidepressant drugs, such as fluoxetine, block L-type calcium channels, and suppress intracellular calcium spikes (Deák et al. 2000; Kim et al. 2013). Moreover, L-type calcium channel antagonists facilitate antidepressant effects of conventional medication, and even show independent antidepressant effects in rodents that appear to be dependent on the CaV1.2 channel (Mogilnicka et al. 1987; Czyrak et al. 1989; 1990; Cohen et al. 1997; Dubovsky et al. 2001; Taragano et al. 2001; Sinnegger-Brauns et al. 2004; Taragano et al. 2005). Furthermore, although L-type calcium channel blockers have been shown to block activity-dependent BDNF synthesis, our recent data support a BDNF-independent mechanism underlying rapid antidepressant-induced TrkB activation (Zafra et al. 1990; Poulsen et al. 2004; Rantamäki et al. 2011). In contrast to antidepressant drugs however, nimodipine activates Akt, a major survival-promoting factor. The mechanisms behind the differential effects of nimodipine and antidepressants on Akt remain unknown, however these findings may suggest stronger neuroprotective properties for nimodipine than for antidepressant drugs (Zhao et al. 2006).

In conclusion, this molecular study provide good basis to investigate the role of TrkB behind the neuroprotective effects of nimodipine in translationally relevant animal models of brain trauma. Current findings also suggest that nimodipine may activate synaptic plasticity in a manner reminiscent to that induced by antidepressant drugs. Our previous findings demonstrate that long-term antidepressant treatment reactivates developmental-type of plasticity mechanisms in the adult brain, which allows the remodeling of synaptic connectivity if combined with appropriate rehabilitation (Maya Vetencourt et al. 2008; Karpova et al. 2011). Indeed, sustained antidepressant treatment after ischemic stroke improves motor function when combined with physiotherapy (Chollet et al. 2011). Whether the therapeutic effects of nimodipine, in SAH and other nervous system conditions that benefit from induced plasticity and neuroprotection, can be facilitated by more prolonged administration and active rehabilitation remains to be studied in animal models.