Address correspondence and reprint requests to Seong-Woo Jeong, Medical Building Rm 418, Ilsan-Dong 162, Wonju, Gangwon-do 220-701, Republic of Korea. E-mail: firstname.lastname@example.org
We investigated effects of Neuregulin 1 (NRG1) on the expression of nicotinic acetylcholine receptor (nAChR) in major pelvic ganglion (MPG) from adult rat. MPG neurons were found to express transcripts for type I and III NRG1s as well as α and β-type epidermal growth factor (EGF)-like domains. Of the four ErbB receptor isoforms, ErbB1, ErbB2, and ErbB3 were expressed in MPG neurons. Treating MPG with NRG1β significantly increased the transcript and protein level of the nAChR α3 and β4 subunits. Consistent with these molecular data, nicotinic currents (IACh) were significantly up-regulated in NRG1β-treated sympathetic and parasympathetic MPG neurons. In contrast, the type III NRG1 and the α form of the NRG1 failed to alter the IACh. Inhibition of the ErbB2 tyrosine kinase completely abolished the effects of NRG1β on the IACh. Stimulation of the ErbB receptors by NRG1β activated the phosphatidylinositol-3-kinase (PI3K) and mitogen-activated protein kinase (MAPK). Immunoblot analysis revealed that PI3K-mediated activation of Akt preceded Erk1/2 activation in NRG1β-treated MPG neurons. Furthermore, specific PI3K inhibitors abrogated the phosphorylation of Erk1/2, while inhibition of MEK did not prevent the phosphorylation of Akt. Taken together, these findings suggest that NRG1 up-regulates nAChR expression via the ErbB2/ErbB3-PI3K-MAPK signaling cascade and may be involved in maintaining the ACh-mediated synaptic transmission in adult autonomic ganglia.
Neuregulin 1 (NRG1) belongs to a family of growth and differentiation factors that contain an epidermal growth factor (EGF)-like domain, and activates the ErbB family of receptor tyrosine kinases (ErbB2, ErbB3, and ErbB4) and are crucial in the development and maintenance of the nervous system. Alternative splicing and use of cell-specific promoters generates multiple NRG1 isoforms that have different localization patterns and functions. These isoforms can be categorized into three groups (type I, II, and III) based on the NH2-terminal characteristics of their extracellular domain (Buonanno and Fischbach 2001; Falls 2003; Esper et al. 2006). NRG1 binds to either ErbB3 or ErbB4, which stimulates receptor dimerization forming either heterodimeric ErbB2/ErbB3 or homodimeric ErbB4/ErbB4 (Riese and Stern 1998). Receptor dimerization activates the intrinsic receptor tyrosine kinase of ErbB2 and ErbB4, which results in the phosphorylation of specific tyrosine residues within the cytoplasmic tail of the receptor. These events activate multiple downstream signaling pathways that produce various biological activities.
NRG1 was originally discovered as a regulator of the expression of the nicotinic acetylcholine receptor (nAChR) in skeletal muscles during development. Among the three isoforms, type I NRG1 (originally named ARIA, acetylcholine receptor-inducing activity) has been implicated in neuromuscular junction (NMJ) formation by stimulating muscular nAChR expression during development (Martinou et al. 1991; Corfas et al. 1993; Falls et al. 1993). The signaling events that mediate the actions of ARIA in muscles involve either the activation of mitogen-activated protein kinase (MAPK), phosphatidylinositol-3-kinase (PI3K), or cyclin-dependent kinase (CDK) pathway (Si et al. 1996; Tansey et al. 1996; Altiok et al. 1997; Si and Mei 1999; Lu et al. 2005).
NRG1 also plays critical roles in the development of autonomic ganglia. For example, NRG1 promotes cell motility as neural crest cells migrate to form sympathetic ganglia (Young et al. 2004; Britsch et al. 1998). As in the skeletal muscles, NRG1 has also been shown to aid in the formation of nicotinic synapses by increasing the expression of nAChR subunits in embryonic chick sympathetic neurons (Yang et al. 1998). Currently, however, it is unknown whether NRG1 regulates the expression of nAChR in autonomic neurons during adulthood. Furthermore, the functional role of NRG1 in parasympathetic neurons has not been reported. We subsequently raised the following two questions: (i) Are NRG1 and its receptors still expressed in autonomic ganglion neurons during mammalian adulthood? and (ii) Is NRG1 required for the regulation of nAChR expression in adult autonomic neurons regardless of cell types (i.e., sympathetic and parasympathetic)? To address these questions, MPG was used because they contain sympathetic and parasympathetic neurons in one ganglion capsule (Dail et al. 1975) and have been well studied (Zhu et al. 1995; Lee et al. 2002; Park et al. 2006a, b; Won et al. 2006). Finally, we investigated the signaling pathways that underlie NRG1-induced expression of nAChR in adult MPG neurons.
Materials and methods
Chemicals and antibodies
Recombinant human NRG1-α (amino acid residues 177–241), NRG1-β1 (amino acid residues 176–246), and type III NRG1 (SMDF) were purchased from R&D Systems (Minneapolis, MN, USA). Nerve growth factor (NGF), ciliary neurotrophic factor (CNTF), epidermal growth factor (EGF), and brain-derived neurotrophic factor (BDNF) were purchased from Sigma Chemical Co (St Louis, MO, USA). AG825, genistein, and U0126 were obtained from Calbiochem (Beeston, UK). LY294002 and wortmannin were purchased from A.G. Scientific, Inc. (San Diego, CA, USA). Rabbit polyclonal anti-nAChR-α3, anti-ErbB2 (Neu, C-18), anti-ErbB3 (C-17), and anti-ErbB4 (C-18) antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit polyclonal anti-nAChR-β4 antibody was purchased from Research Diagnostics (Flanders, NJ, USA). Rabbit polyclonal anti-Akt, anti-phospho Akt, anti-p44/42 MAPK (Erk1/2), and anti-phospho p44/42 MAPK (Erk1/2) antibodies were obtained from Cell Signaling Technology (Beverly, MA, USA). Rabbit monoclonal anti-β-actin antibody was purchased from Santa Cruz Biotechnology.
Preparation of MPG neurons
MPG neurons from adult male Sprague–Dawley rats (200–300 g) (OrientBio, Seongnam-Si, Korea) were enzymatically dissociated as previously described (Zhu et al. 1995; Won et al. 2006). Animal use and care procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Yonsei University Wonju College of Medicine (approval No. YWC-110330-1). For current recording, neurons were plated onto culture dishes (35 mm) that were coated with poly-l-lysine (Sigma) and maintained in humidified 95% air–5% CO2 incubator at 37°C. The current recording was made within 9 h after dissociation. For RT-PCR experiments, the dissociated cells were plated onto culture dishes with no poly-l-lysine. Twenty minutes after plating, non-attached neurons were harvested leaving the non-neuronal cells (e.g., glial cells) attached to the bottom of the dish.
RNA isolation, reverse transcription, and PCR
Total RNA was extracted from either the dissociated MPG neurons or whole ganglia using the RNeasy plus Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's protocol. The quantity and quality of RNA were assessed by spectrophotometry at 260 nm. First-strand cDNA was synthesized from 1 μg of total RNA by incubating it with 50 U M-MLV (monkey murine leukemia virus) reverse transcriptase (Roche, Penzberg, Germany) at 42°C for 1 h. Parallel reactions without reverse transcriptase were performed to confirm the absence of genomic DNA amplification. Semiquantitative PCR was initiated with an initial denaturation step of 95°C for 5 min, followed by 35 cycles of 94°C for 30 s, 60°C for 40 s, and 72°C for 30 s. The PCR reaction was terminated by maintaining the temperature at 72°C for 7 min. PCR products were electrophoresed on a 1.5% agarose gel and visualized by ethidium bromide staining. The following primers were used: NRG1 type I (forward, 5′-ATCTTCGGCGAGATGTCTGA; reverse, 5′-TTGACAGGTCCTTCACCGT), type II (forward, 5′-ATAAAAGGAGGAGGGTCAGG; reverse, 5′-GGTCCCCAGTAGTAGCAGCA), type III (forward, 5′-TACAAGGTGACCATGCTGCT; reverse, 5′-TTTGACAGGTCCTTCACCGT), NRG1α (forward, 5′-ACGGTGAAGCCTGTCAAA; reverse, 5′-CCCATGAAAGTCCAAACCCA), NRG1β (forward, 5′-AGTGCGCGGAGAAGAAGAAA; reverse, 5′-GTTTACTGGTGATCGTTGCC), ErbB1 (forward, 5′-TATAGCTCCGATCCCACCAG; reverse, 5′-GTTGTCCAGGCTCATTTGGT), ErbB2 (forward, 5′-CCCATCAGAGTGATGTGTGG; reverse, 5′-TCATCTTCCAGCAGTGAACG), ErbB3 (forward, 5′-CGTCATGCCAGATACACACC; reverse, 5′-CTCCTCGTACCCTTGCTCAG), ErbB4 (forward, 5′-TGGAGGAAAGCCCTATGATG; reverse, 5′-CTGGGGGACCAAATATTCCT), and GAPDH (forward, 5′-TCATGACAACTTTGGCATCATGG; reverse, 5′-GTCTCCTGACTTCAA CAGCAAC).
Expression of the nAChR subunits was quantified by real-time PCR, as previously described (Huang et al. 2004). The following primers were used: nAChR-α3 (forward, 5′-ACTCCAAAAGCTGCAAGGAA; reverse, 5′-CTTGTGAGGTTGGCACTGAA), nAChR-β4 (forward, 5′-CTCCTGAACAAAACCCGGTA; reverse, 5′-CTGATGAGCTGGGACAGTGA), β-actin (forward, 5′-ATGGTGGGTATGGGTCAGAA; reverse, 5′-ACCAACTGGGACGATATGGA). β-actin gene was amplified as an internal reference. The specificity of each reaction was confirmed by analyzing the melting curve generated at the end of the PCR. The fold changes in gene expression were analyzed by the comparative 2−ΔΔCT method (Livak and Schmittgen 2001).
Western blot analysis
Whole ganglia were homogenized in ice-cold hypotonic buffer (10 mM Tris, pH 7.4) supplemented with a protease inhibitor mixture (Sigma), and then centrifuged at 16 100 g for 10 min. The precipitated samples were dissolved in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% MP-40, and protease inhibitors, pH 8.0), incubated for 40 min on ice, and then centrifuged at 13 200 rpm for 15 min to remove any insoluble material. Protein levels were quantified using protein assay solution (Bio-Rad, Hercules, CA, USA). After boiling in sodium dodecyl sulfate buffer for 5 min, 20 μg of sample was electrophoresed on an sodium dodecyl sulfate–polyacrylamide gel electrophoresis gel and transferred to a polyvinylidene difluoride membrane (BioTraceTM, Pall Co., East Hills, NY, USA). The membranes were blocked with 0.1% Tris-buffered saline/Tween-20 (TBST) plus 5% bovine serum albumin (BSA; Sigma) for 1 h at 20–22°C, and then incubated overnight at 4°C with primary antibodies against nAChR-α3 (H-100) (1 : 200), nAChR-β4 (1 : 200), ErbB2 (1 : 200), ErbB3 (1 : 200), ErbB4 (1 : 200), Akt (1 : 1000), phospho-Akt (1 : 1000), Erk1/2 (1 : 1000), phospho-Erk1/2 (1 : 1000), and, β-actin (1 : 2000). Membranes were washed several times with cold 0.1% TBST and incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at 20–22°C. After intensive washing, immunoreactive bands were visualized using the ECL western blot detection reagent (Amersham Biosciences, Piscataway, NJ, USA) on X-ray film.
Current recordings in the MPG neurons were obtained under the whole-cell-ruptured configuration of the patch-clamp technique using an EPC-10 amplifier and pulse/pulsefit (v8.50) software (HEKA Electronik, Lambrecht, Germany) as previously described (Jeong and Ikeda 1998; Park et al. 2006b). Cell types of the MPG neurons were recognized according to the previously established criteria: cell size as assessed by magnitude of the electrical capacitance (Zhu et al. 1995; Lee et al. 2002), responses to GABA and 5-HT (Park et al. 2006a, b; Won et al. 2006; Huang et al. 2011), presence or absence of the T-type Ca2+ channel-mediated anodal break rebound spike (Lee et al. 2002; Park et al. 2006b), and the firing patterns (i.e., tonic vs. phasic) in response to depolarizing current injection (Park et al. 2006b). To record the ACh-induced current (IACh), ACh was applied to single neurons via a gravity-fed fused silica capillary tube connected to an array of seven polyethylene tubes. All experiments were performed at room temperature (20–24°C).
Data analysis was performed using the IGOR data analysis package (Wave-Metrics, Lake Oswego, OR, USA) or GraphPad Prism (ver 4.0, GraphPad Software Inc, La Jolla, CA, USA). Data were presented as the mean ± SEM. The Student's t-test or one-way anova analyses with Fisher's PLSD were performed. A value of p < 0.05 was considered statistically significant.
Identification of the NRG1 and ErbB receptor isoforms expressed in the MPG
The expression of the NRG1 and ErbB receptor isoforms has not been previously reported for the autonomic neurons of the adult rat. Therefore, we examined the expression of NRG1 and its receptors in the adult rat MPG, which has been molecularly and electrophysiologically well defined (Zhu et al. 1995; Lee et al. 2002; Park et al. 2006a,b; Won et al. 2006). RT-PCR analysis using N-terminal sequence-specific primers revealed that the whole MPG expresses type I and type III NRG1 transcripts (Fig. 1a). In comparison with the whole ganglion, dissociated MPG neurons express type III NRG1 at a much lower level (data not shown). Unlike whole-brain tissue, which was used as a positive control, MPG neurons do not express type II NRG1. In addition, the α and β types of the EGF-like domains were found in MPG neurons and whole-brain tissue. The expression level of NRG1β appeared to be much higher than that of NRG1α. No PCR products were observed when the RT-PCR process was run without reverse transcriptase (RT), which indicates the absence of genomic DNA contamination (Fig. 1a and b). RT-PCR analysis also revealed that MPG neurons express transcripts encoding the ErbB1, ErbB2, and ErbB3 receptors; however, no ErbB4 receptor expression was found (Fig. 1b). Western blot analysis confirmed the presence of ErbB2 and ErbB3 proteins and the absence of ErbB4 protein in MPG (Fig. 1c). Interestingly, the ErbB3 protein expression level was higher in MPG when compared with whole-brain tissue, which strongly expressed the ErbB4 protein. Other autonomic ganglia (stellate, superior mesenteric, celiac, and intracardiac ganglia) exhibited the same expression patterns for the NRG1 and ErbB receptor isoforms as the MPG (data not shown).
NRG1β up-regulates the expression of the nAChR α3 and β4 subunits in MPG neurons
The nAChRs in MPG neurons were found to have α3β4* composition, which mediate fast excitatory synaptic transmissions (Park et al. 2006b). Therefore, we examined whether NRG1 was capable of up-regulating the expression of the nAChR subunits in MPG neurons. MPG neurons were treated with human recombinant NRG1β (5 nM) for 6 h. Real-time PCR analysis revealed that NRG1β significantly increased the expression of the nAChR α3 and β4 subunit transcripts (p < 0.001) (Fig. 2a). In contrast, treating MPG neurons with neurotrophic factors, such as NGF (100 ng/mL), CNTF (100 ng/mL), and BDNF (100 ng/mL) failed to up-regulate the expression of the nAChR subunits (Figure S1a and b). Western blot analysis was performed to examine the protein expression levels of the nAChR subunits. The protein expression levels of the nAChR subunits, which were normalized to β-actin, are summarized in Fig. 2b. Consistent with the real-time PCR data, NRG1β significantly increased the protein expression levels of the nAChR α3 (p < 0.001) and β4 (p < 0.01) subunits.
NRG1β treatment increases the nAChR current densities in sympathetic and parasympathetic MPG neurons
One of the unique features of the MPG is that sympathetic and parasympathetic neurons exist within the same ganglion capsule (Dail et al. 1975; Dail 1993). Therefore, we evaluated the nAChR channel activity in both types of MPG neurons after treating them with NRG1β (5 nM) for 6 h. The inward IACh was generated by 100 μM ACh in the MPG neurons held at −60 mV under the whole-cell-ruptured configuration of the patch-clamp techniques. Consistent with our PCR and western blot analysis results, NRG1β treatment significantly increased the IACh density in both types of MPG neurons (p < 0.001) (Fig. 3). On average, the IACh densities in the control and NRG1β-treated sympathetic MPG neurons were 452 ± 27 pA/pF(n = 19) and 596 ± 42 pA/pF(n = 19), respectively, while the IACh densities in the control and NRG1β-treated parasympathetic MPG neurons were 339 ± 22 pA/pF(n = 16) and 524 ± 34 pA/pF(n = 17), respectively.
Alternative splicing of the EGF-like domain produces NRG1α and NRG1β (Falls 2003). Unlike NRG1β, NRG1α (5 nM) failed to increase the IACh in MPG neurons (n = 6). A previous study showed that type III NRG1 increases nAChR expression in embryonic autonomic neurons (Yang et al. 1998). However, type III NRG1 treatment did not affect the IACh in sympathetic and parasympathetic MPG neurons (n = 7) from adult rats (Fig. 3). In addition, NGF, CNTF, and BDNF did not increase the IACh in sympathetic and parasympathetic MPG neurons (Figure S1c and d). This finding is consistent with the real-time PCR data showing that these trophic factors do not affect the expression of the nAChR subunits at the transcriptional level in MPG neurons. EGF (50 ng/mL), a ligand for ErbB1 (EGFR), also failed to affect the IACh in MPG neurons (Figure S1). Cycloheximide (10 μg/mL) prevented the up-regulation of the IACh in NRG1β-treated MPG neurons (n = 6) (Fig. 4). This finding indicates that the activity of NRG1 requires de novo protein synthesis. There was no increase in the IACh when NRG1β (5 nM) was acutely applied for 5 min during current recordings, which suggests the absence of non-genomic effects of NRG1β on nAChRs (data not shown). Furthermore, treating MPG neurons with NRG1β did not alter the voltage-gated T- and N-type calcium currents (data not shown).
NRG1 up-regulates nAChR subunit expression through the ErbB2/ErbB3-PI3K-MAPK signaling cascade in MPG neurons from adult rats
ErbB2 has no ligand-binding sites, but it does contain an intracellular kinase domain that is not found in ErbB3 (Riese and Stern 1998). Accordingly, we presumed that ErbB2 and ErbB3 form a heterodimer, which causes downstream signaling in MPG neurons. Consistent with this notion, treatment of MPG neurons with a selective ErbB2 tyrosine kinase inhibitor (AG825, 10 μM) and a non-specific tyrosine kinase inhibitor (genistein, 200 μM) completely abolish the NRG1β-induced increase in the IACh (Fig. 5). In the absence of NRG1β, AG825 and genistein only slightly decreased the IACh, which suggests that the ErbBRs are partially phosphorylated by the basal activity of the tyrosine kinases in the absence of ligands.
Phosphorylation of specific tyrosine residues in the cytoplasmic tail of the ErbB2/ErbB3 heterodimer complex recruits molecules from multiple signaling pathways, including the PI3K and MAPK pathways. Therefore, we examined the involvement of the PI3K pathway in the NRG1-induced up-regulation of the IACh using LY294002 and wortmannin, which are selective PI3K inhibitors. The specificity of LY294002 has been well established (Vlahos et al. 1994; Tansey et al. 1996). As shown in Fig. 6, treating MPG neurons with either LY294002 (20 μM) or wortmannin (50 nM) completely abolishes the NRG1-induced up-regulation of the IACh in sympathetic and parasympathetic MPG neurons. In the absence of NRG1β, the PI3K inhibitors had little effect on the IACh. Interestingly, U0126 (20 μM), a highly selective inhibitor of MEK1/2, also completely abrogated the NRG1-induced up-regulation of the IACh in sympathetic and parasympathetic MPG neurons (Fig. 7). Taken together, these data suggest that PI3K and MEK are critical for mediating the effect of NRG1β on nAChR expression.
PI3K acts upstream of MAPK in the NRG1-induced up-regulation of the IACh
It is unlikely that the PI3K and MAPK pathways are activated in parallel because each pathway is necessary for the NRG1-induced up-regulation of the IACh, Therefore, we presumed that there is crosstalk between the two pathways. Accordingly, the time dependency of the activation of each pathway was assessed by western blot analysis. Activation of PI3K and MEK results in the phosphorylation of Akt and ERK1/2, respectively. Figure 8 shows our immunoblot results using either the anti-pAkt or anti-pErk1/2 antibody on the lysate of NRG1β-treated MPG. The level of phospho-Akt (60 kDa) peaked 15 min after NRG1β treatment. The maximum levels of phospho-Erk1 (44 kDa) and phospho-Erk2 (42 kDa) were detected 30 min after NRG1β treatment. These data suggest that PI3K activation precedes MEK activation during NRG1/ErbB signaling in MPG neurons. Consistent with the electrophysiological data (Figs 67and), the PI3K inhibitors (LY294002 and wortmannin) and the MEK inhibitor (U0126) abrogated the NRG1-induced phosphorylation of Akt and ErK1/2, respectively (Fig. 9). LY294002 and wortmannin negated the NRG1-induced increase in the pAkt and pErk1/2 levels in MPG neurons. U0126, however, reduced the level of pErk1/2, but not pAkt in NRG1-treated MPG. Taken together, these data suggest that PI3K works upstream of MEK-ErK1/2 in the NRG1-induced up-regulation of the IACh.
It is well known that NGF is capable of activating both PI3K and MAPK pathways in neural tissues (Patapoutian and Reichardt 2001). Therefore, to compare with the NRG1-mediated signaling, we examined activation of the PI3K and MAPK pathways in NGF-treated MPG. Unlike NRG1, NGF induced the maximal phosphorylation of both Akt and Erk1/2 at 15 min after treatment (Figure S2). Furthermore, LY294002 and U0126 failed to inhibit NGF-induced phosphorylation of Erk1/2 and Akt, respectively, which suggest that there is no crosstalk between the PI3K and MEK pathways activated in parallel (Figure S3).
Autonomic ganglion neurons express nAChR α3, α4, α5, α7, β2, and β4 subunits (Mandelzys et al. 1994; Poth et al. 1997; Park et al. 2006b). Among the conceivable subunit combinations comprising the heteropentameric nAChRs, the α3β4* is considered to be the primary combination in autonomic ganglia including the superior cervical ganglion (SCG) and MPG (Covernton et al. 1994; Rust et al. 1994; Park et al. 2006b; Albuquerque et al. 2009). Currently, however, there is little published data on how the nAChR α3β4* subtype is regulated in adult autonomic neurons. In this study, we report for the first time that NRG1 up-regulates the expression of the nAChR α3 and β4 subunits in both sympathetic and parasympathetic MPG neurons from adult rats.
MPG was found to express transcripts encoding type I and type III NRG1. In comparison to the whole ganglion, dissociated MPG neurons express type III NRG1 at a much lower level. This finding is consistent with a previous study where expression of the type III NRG1 is shown to be significantly reduced in adult dorsal root ganglion neurons when compared with embryonic neurons (Syed and Kim 2010). Therefore, the primary sources of the transcript encoding the type III NRG1 may be the pre-synaptic terminal and/or non-neural cells (e.g. glial cells) in MPG. Previous studies indicate that pre-synaptic inputs are critical for controlling nAChR expression at the synapses of developing autonomic neurons (Rosenberg et al. 2002). In embryonic sympathetic neurons, the type III NRG1 functions as a pre-synaptic input-derived regulatory factor that increases nAChR expression (Yang et al. 1998). Of the two isoforms, however, only the type I NRG1 was found to increase nAChR expression in adult autonomic neurons. This finding supports the notion that regulation of functional nAChR expression in mature autonomic neurons is less dependent on pre-ganglionic interactions (Zhou et al. 1998). It is unclear, however, whether the type III NRG1 plays other functional roles in adult autonomic ganglia. Recent findings suggest that the type III NRG1 can participate in bidirectional juxtacrine signaling. For example, the type III NRG1 is involved in maintaining Schwann cells in the pre-synaptic axons of the somatic motor neurons innervating the diaphragm muscles (Wolpowitz et al. 2000). The type III NRG1 is also involved in the axonal targeting of extrasynaptic nAChR subunits such as α7 in sensory neurons (Hancock et al. 2008) and CNS neurons (Zhong et al. 2008). Thus, it would be interesting to determine if the type III NRG1–ErbB interaction results in backward signaling within the pre-synaptic region of autonomic neurons. Unlike the type III NRG1, which is bound to the pre-synaptic membrane, the type I NRG1 acts in a paracrine manner after proteolytic cleavage (Loeb and Fischbach 1995; Loeb et al. 1998). In a different series of experiments, we observed that cleaved type I NRG1 is released from MPG neurons by either the electrical stimulation of sympathetic or parasympathetic pre-ganglionic nerves (i.e., hypogastric nerves or pelvic nerves, respectively) or by high potassium-induced depolarization, which increases nAChR expression (Kim et al., paper in preparation). Taken together, the type I NRG1 appears to regulate nAChR expression in adult MPG neurons. Retrograde signals from target tissues are also essential for maintaining nAChR levels in CNS and PNS neurons (Rosenberg et al. 2002; Albuquerque et al. 2009). CNTF, a trophic factor expressed in muscle cells innervated by avian ciliary ganglion neurons, up-regulates the α3 nAChR protein level (Finn and Nishi 1996). NGF is a potent regulator of α3 and β4 transcription in PC12 cells (Rogers et al. 1992; Henderson et al. 1994; Hu et al. 1994). In this study, however, these target-derived neurotrophic factors (CNTF, NGF, and BDNF) did not alter the nAChR expression level in the MPG neurons from adult rats. A previous study showed that neurotrophic factors regulate the NRG1 expression in embryonic ventral spinal cord neurons (Loeb and Fischbach 1997). Therefore, it is possible that neurotrophic factors indirectly participate in regulating the nAChR expression level by controlling the expression level of NRG1 in autonomic neurons. In MPG neurons, we detected both NRG1α and NRG1β transcripts, which are splice variants of the EGF-like domain. NRG1β has a significantly greater affinity for receptors and consequently is more biologically active than the α isoform (Raabe et al. 1996; Jones et al. 1999). Consistent with these findings, NRG1β, but not NRG1α, was effective in up-regulating the nAChR current in sympathetic and parasympathetic MPG neurons.
MPG neurons express the ErbB1 (EGFR), ErbB2, and ErbB3 receptors, but not the ErbB4 receptor. This expression pattern for the ErbB receptors was commonly found in other types of autonomic neurons (see 'Results'). Among the receptors expressed in MPG neurons, only ErbB3 is known to bind to NRG1 (Riese and Stern 1998). Several studies have shown that ErbB3 can form a heterodimer with either EGFR or ErbB2 (an orphan receptor), when it binds NRG1 (Riese and Stern 1998). In our study, a selective ErbB2 tyrosine kinase inhibitor (AG825) completely blocked the NRG1-induced up-regulation of nAChR expression. A previous study has shown that activation of the EGFR by EGF is capable of transmodulating ErbB2 (Riese and Stern 1998). However, EGF failed to alter nAChR expression in our experiments, which suggests that the EGFR or the EGFR/ErbB2 complex is not involved in the NRG1-induced up-regulation of nAChR expression in MPG neurons. Overall, we concluded that ErbB2 is a dimerization partner of ErbB3 that mediates NRG1 signaling in the MPG neurons of adult rats.
Even though the effects of NRG1 on nAChR expression have been studied in different types of neurons (Yang et al. 1998; Liu et al. 2001; Hancock et al. 2008; Zhong et al. 2008), the signaling mechanisms underlying the function of NRG1 have not been extensively investigated. The activation of ErbB2/ErbB3 by NRG1 is capable of directly or indirectly stimulating multiple enzymes (PI3K, MAPK, Phospholipase C-γ, cyclin-dependent kinases, etc.) to mediate a variety of biological responses (Citri et al. 2003). Of the EGF family of receptors, only ErbB3 has the ability to couple to PI3K (Fedi et al. 1994; Soltoff et al. 1994). In our experiments, treating MPG neurons with NRG1 resulted in the activation of the PI3K pathway, which supports the involvement of ErbB3 in NRG1-induced up-regulation of nAChR expression. The involvement of the PI3K pathway in stimulating or inhibiting nAChR expression has been reported in the Sol 8 muscle cell line and chick embryo myoblasts, respectively (Tansey et al. 1996; Altiok et al. 1997). In MPG neurons, the PI3K pathway is found to be stimulatory. In addition to the PI3K pathway, activation of ErbB2/ErbB3 by NRG1 also activates the MEK/MAPK (Erk1/2) pathway, which mediates the up-regulation of nAChR expression in MPG neurons. The roles of the PI3K and MEK/MAPK pathways in the up-regulation of nAChR expression appear to be similar to their role in muscle cells (Tansey et al. 1996; Altiok et al. 1997). However, our results show that activation of both PI3K and MAPK is necessary for increasing nAChR expression in MPG neurons, which suggest that there is crosstalk between two pathways. The NRG1 signaling in MPG neurons is quite different from that in muscle cells (Tansey et al. 1996) and from NGF signaling in the same neurons (see 'Results') in which the enzymes are activated in parallel. As depicted in Fig. 10, PI3K and MAPK are activated sequentially when ErbB receptors are stimulated by NRG1. In this study, several lines of evidence support that PI3K operates upstream of MEK–MAPK. First, the peak phosphorylation level of Akt preceded that of Erk1/2. Second, specific PI3K inhibitors (LY294002 and wortmannin) prevented the phosphorylation of Erk1/2. Third, the inhibition of MEK using U0126 failed to inhibit the phosphorylation of Akt. Furthermore, the PI3K requirement for activating the MEK–MAPK pathway has been previously reported (King et al. 1997; Aksamitiene et al. 2011).
The synaptic strength in the autonomic ganglion is determined by several factors, including the quantal content, the number of post-synaptic nAChRs, and the geometry of the neuronal cells (Vernino et al. 2009). Therefore, our findings indicate that NRG1, especially type I, may be critical for maintaining the synaptic strength at mature cholinergic synapses by regulating the expression of nAChR in autonomic neurons. Increasing evidence indicates that the down-regulation of nAChRs in autonomic neurons is associated with autonomic dysfunctions, such as neurogenic impotence (Huang et al. 2011). In this regard, NRG1 may be of therapeutic value.
This study was supported by a grant from the Myung-Sun Kim Memorial Foundation (2008). All authors approved the final version of the manuscript and have no conflicts of interest to declare.