Transactivation of EGF Receptors in Chicken Müller Cells by α2A-Adrenergic Receptors Stimulated by Brimonidine

Abstract

Purpose.: Alpha2-adrenergic receptor agonists are used in glaucoma treatment and have been shown to have some neuroprotective effects. We performed this study to test the hypothesis that epidermal growth factor receptors on chicken Müller cells are transactivated by α2-adrenergic receptors and we focused on the extracellular signal-activated kinases 1/2 (ERK) pathway.

Methods.: Embryonic chicken retina and cultures of primary Müller cells were stimulated by α2-adrenergic receptor agonist brimonidine. Immunostaining, quantitative RT-PCR, and Western blot techniques, in combination with Src, epidermal growth factor receptor kinase, and matrix metalloproteinase inhibitors, were used for analysis of the cellular responses.

Results.: Our results showed that Müller cells express α2A-adrenergic receptors in vivo and in vitro and that brimonidine triggered a robust and transient phosphorylation of ERK1/2. This ERK-response was Src-kinase dependent, associated with tyrosine phosphorylation of epidermal growth factor receptors (phospho-Y1068, Y1173) and was mediated by matrix metalloproteinase activity on the Müller cells.

Conclusions.: Müller cells express the α2A-adrenergic receptor, and brimonidine triggers both Src-kinase– and matrix metalloproteinase–mediated autocrine ligand-dependent activation of epidermal growth factor receptors on Müller cells. This response is consistent with transactivation of epidermal growth factor receptors by stimulation of α2-adrenergic receptors.

Introduction

Müller cells (MCs) have a broad range of functions, all of which are vital to the health of the retinal neurons. ¹ In addition, after acute injury MCs can reenter the cell cycle, dedifferentiate into a cycling population of retinal progenitor cells, and produce retinal neurons. ^2,3 This regenerative capacity of MCs varies among species and seems to be lower in warm-blooded animals. ⁴ Treating uninjured retina with exogenous factors, including FGFs, epidermal growth factor (EGF), or Wnts, triggers dedifferentiation and proliferation. ^5–7 The process is dependent on the activation of the mitogen-activated protein kinase (MAPK) pathway with phosphorylation of extracellular signal-activated kinases 1/2 (ERK1/2). ^7,8 Studies of species with a retina that has regenerative capacity have identified the EGF receptor (EGFR) as a key regulator in MC dedifferentiation and retina regeneration. ⁶  

Stimulation of α2-adrenergic receptors (α2-ADRs) has in experimental models been shown to be neuroprotective and to reduce the effects of retinal injury. ^9–12 The underlying mechanisms have been shown to include attenuation of excitotoxicity by modulation of N-methyl-D-aspartate receptor signaling in retinal ganglion cells or promotion of the trophic factor responses that contribute to increased neuronal survival. ^13–15 However, the molecular mechanisms behind the neuroprotection are still not fully understood. One of the immediate responses to α2-ADR stimulation is a robust ERK1/2 activation in MCs, ¹⁶ and this prompted us to further study the α2-ADR system in MCs. 

The α2-ADRs are G-protein–coupled receptors that signal by modulating the activities of adenylyl cyclase, phospholipase C, and phosphatidylinositide 3-kinase, as well as the MAPK pathway. There are three subtypes of α2-ADR (α2A, α2B, and α2C) with different tissue distribution, and pharmacological and signaling properties. ^17,18 The α2-ADRs are expressed in retina. ¹⁹ Commonly used α2-ADR agonists are dexmedetomidine and xylazine used for sedation and anesthesia in humans and animals, respectively. Another α2-ADR agonist is brimonidine (BMD), which is used in glaucoma treatment. ²⁰  

The α2-ADRs have been shown to transactivate EGFRs in transfected COS-7 cells. ^21,22 Transactivation can occur via the activation of the intracellular protein tyrosine kinase Src that causes a release of the membrane-bound EGFR ligand heparin-binding EGF (HB-EGF) by activating matrix metalloproteinase (MMPs) in an autocrine mode of action. Müller cells express EGFR, ²³ HB-EGF, ⁶ and α2-ADRs, but it is not known if EGFRs are engaged in the ERK1/2 MAPK response after stimulation of α2-ADR on MCs. 

We performed this study to test the hypothesis that EGFRs on chicken MCs are transactivated by α2-ADRs. We first characterized the expression of α2-ADRs on chicken MCs and used the α2-ADR agonist BMD to study the signaling pathway in vivo and in primary MC cultures. Our results showed that MCs express the α2A-ADR and that BMD triggers both Src-kinase activity and MMP-mediated autocrine ligand-dependent activation of EGFRs on MCs, which is consistent with transactivation of EGFRs by stimulation of α2-ADRs. 

Materials and Methods

Animals and Intraocular Injection

Animal experiments were performed according to the guidelines given by ARVO and the local animal ethics committee in Uppsala. Fertilized White Leghorn chicken eggs were obtained from OVA Produktion AB (Västerås, Sweden) and incubated at 38°C in a humidified incubator (Maino, Naples, Italy). Intraocular injections were made in the dorsal quadrant of the eye using a Hamilton syringe with 26-gauge needle. Fertilized embryonic day (E) 18 eggs were opened at the blunt end and a small hole was made in the eggshell and chorioallantoic membranes. The head was pulled toward the membranes with a bent glass rod and injection was done through the amniotic membranes. Fifteen microliters (80 μg) of BMD tartrate (Supplementary Table S1) in sterile saline (0.15 M NaCl) solution was injected into the experimental (right) eyes (n > 5). As a control, saline solution was injected (n > 5). The injected eggs were sealed and incubated in a humidified incubator and analyzed after different time points. 

Müller Cell Cultures

Retinas from E14 embryos were dissected and immediately transferred to 1 mL Hank's Ca²⁺ and Mg²⁺ free balanced salt solution. Trypsin (final concentration 0.1%) was added and retinas were incubated at 37°C for 30 minutes. The trypsin was removed and the retinas were triturated using a Pasteur pipette in Dulbecco's modified eagle medium with 10% newborn calf serum, 2 mM glutamine, 100 U/mL penicillin, and 100 mg/mL streptomycin. The dissociated cells were transferred to a Falcon dish and cultured in a humidified atmosphere of 5% CO₂, at 37°C for up to 4 weeks. Media were changed three times a week. The cultures were ready to use when all neurons were gone and the cultures contained only MCs. The cells were then trypsinized and transferred to dishes containing coverslips (VWR, Radnor, PA, USA) and cultured for 2 days. The cell purity was assessed by immunocytochemistry. For drug treatment, serum-starved MCs were treated with BMD tartrate (100 μM), yohimbine (10, 20, 100 μM), EGF (0.5 μg/mL), or specific inhibitors: PP1 (5 μM), GM6001 (50 μM). As controls, cells were treated with vehicles (Supplementary Table S1). 

Immunohistochemistry and Cytochemistry

Eyes were dissected and fixed in 4% paraformaldehyde in PBS for 30 minutes at room temperature, incubated in 30% sucrose in PBS for 4 hours at 4°C, frozen in optimum cutting temperature freezing medium and sectioned (10 μm) in an orientation parallel to the center of the lens and through the optic nerve exit containing dorsal and ventral retina. The primary MCs on coverslips were fixed in 4% paraformaldehyde in PBS for 15 minutes, washed in PBS, and used for immunocytochemistry. The immunostaining was performed as previously described. ²⁴ Primary and secondary antibodies with working dilutions are listed in Supplementary Table S2. Microscopy was performed using a Zeiss Axioplan2 microscope (Jena, Germany) or LSM 510 confocal microscope (Jena, Germany). Micrographs were captured at a position 1.5 mm from the ciliary marginal zone in the dorsal position of the retina. 

Quantitative RT-PCR

Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA) and cDNA was reverse-transcribed from 1 μg DNase-treated RNA (MultiScribe RT; Applied Biosystems, Foster City, CA, USA) using random hexamer priming. The quantitative RT-PCR (qRT-PCR) analysis was performed using IQ SyBr Green Supermix (Bio-Rad, Hercules, CA, USA) with primers designed by using Primer Express v2.0 (Applied Biosystems). The amplification was checked for linearity and specificity. The mRNA levels were normalized to β-actin mRNA levels and the use of β-actin for normalization purposes has been validated. ²⁵ Primers and corresponding accession numbers of target sequences are listed in Supplementary Table S3. Expression levels were calculated from cycle threshold (Ct) and the 2^−ΔΔCt method. ²⁶  

Western Blot Analysis

Retinas were dissected and frozen on dry ice until use or the treated cells were scraped off the dish and homogenized in a lysis buffer. Lysis buffer preparation, protein concentration measurement, and Western blot analysis were performed according to the manufacturer's instruction and as previously described. ²⁷ Densitometry was performed using Image Lab 4.1 software (Bio-Rad). Primary and secondary antibodies with working dilutions are listed in Supplementary Table S2. 

Results

α2-Adrenergic Receptors in E18 Chicken Retina

Expression of α2-Adrenergic Receptors.

We analyzed the expression of α2-ADRs in E18 chick retinas by using immunohistochemistry. To test the anti-α2-ADR antibodies, the immunoreactivity (IR) in adult rat retina was studied. The patterns were consistent with published data (not shown). ¹⁹ In chicken retina, α2A-ADR IR was found in the ganglion cell layer, in the vitreal part of inner nuclear layer, in the outer plexiform layer, at the outer limiting membrane, and in the inner and outer segments of photoreceptors (Fig. 1A). We co-labeled with either Sox2, a transcription factor expressed in chick MCs (Fig. 1B) ⁷ or with the 2M6 antigen, which stains MCs and their processes from the ganglion cell layer to the outer limiting membrane (Figs. 1C, 1D). ²⁸ We found α2A-ADR positive (+) cell bodies in the inner nuclear layer that co-labeled for 2M6 and Sox2 (Fig. 1E). In addition, there was a clear overlap of α2A-ADR IR and 2M6+ MC processes at the outer limiting membrane (Fig. 1E). We detected α2B-ADR IR in the photoreceptor outer segments (Figs. 1F–H) and the α2B-ADR IR overlapped with that of visinin (Figs. 1I–K), a protein expressed by photoreceptors. ²⁹ The α2A and α2B-ADR IR were seen in the entire E18 retina. In contrast to α2A-ADR, no staining of α2B-ADR was observed in 2M6+ or Sox2+ cells (Fig. 1G). We did not find any antibody for α2C-ADR that gave a consistent IR pattern in chicken retina. 

Figure 1

 

Expression of α2-ADRs in E18 chicken retina. Fluorescence micrographs of immunohistochemistry for (A) α2A-ADR, (B) Sox2, and (C) 2M6. (D) Cell nuclei labeled with DAPI (4′,6-diamidino-2-phenylindole) (blue) and (E) merged micrograph (A–D). Sox2, 2M6, and DAPI show that α2A-ADR IR was localized to MC bodies (arrows) and processes at the outer limiting membrane (arrowheads). (F) α2B-ADR (arrow) and (G) merged micrograph for α2B-ADR, Sox2, 2M6, and DAPI showing absence of α2B-ADR IR in MCs. (H) Negative control with secondary antibodies only. (I) Alpah2B-ADR IR (green) overlapped with (J) visinin IR (red) present in photoreceptors and (K) merged micrograph (I, J). Scale bar in (H) is 35 μm, valid also for (A–G), and scale bar in (K) is 15 μm, valid also for (I, J). OLM, outer limiting membrane; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer.

Stimulation of α2-ADRs Leads to Activation of ERK1/2 Mapkases.

We used BMD to stimulate α2-ADRs and analyzed the phosphorylation of ERK1/2 at 2 hours, 6 hours, and 24 hours after injection (Fig. 2A). Immunohistochemistry for phospho-ERK1/2 (P-ERK) in combination with 2M6 was used to identify MCs that were stimulated by BMD (P-ERK, 2M6 double-positive cells; Figs. 2B, 2C). The P-ERK IR was mainly located in the vitreal endfeet in the nerve fiber layer and on 2M6+ somata in the inner nuclear layer (Figs. 2B, 2C). The P-ERK IR was high in 2M6+ cells at 2 hours after BMD treatment (Fig. 2B), lower at 6 hours (Fig. 2C), and back to the normal levels by 24 hours (Fig. 2D). In control retinas injected with vehicle, we did not observe any increased levels of P-ERK IR (Fig. 2E). To verify and quantify the P-ERK immunohistochemistry results, Western blot analysis was performed (Fig. 2F). The P-ERK levels were increased 2-fold at 2 hours after BMD treatment and had decreased by 6 hours and were back to the control levels by 24 hours (Fig. 2G). 

Figure 2

 

The α2-ADR agonist, BMD, stimulates ERK1/2 phosphorylation in E18 chicken retina. Intraocular injections of 80 μg BMD or control (vehicle). (A) Experimental outline. (B–E) Fluorescence micrographs of P-ERK and 2M6 immunohistochemistry were performed on (B) 2 hours, (C) 6 hours, (D) 24 hours, and (E) control 2 hours after BMD treatment. Panels also show separation of red and green channels. (F) Representative Western blot analysis of P-ERK in retina 2 hours, 6 hours, and 24 hours after BMD treatment. (G) Bar graph with densitometry of P-ERK levels normalized by β-actin levels. Bar graph is mean ± SEM, n = 3 (*P < 0.01, ***P < 0.0001) analyzed by one-way ANOVA and Tukey's post hoc test. Scale bar in (E) is 40 μm, valid also for (B–D).

α2-ADRs in Primary MCs In Vitro

Expression of α2-ADRs.

Primary MCs were grown for 4 weeks in culture. The purity of the primary MC culture was determined and checked by the fraction of 2M6+ cells in the culture and it was more than 95% in the cultures when used (Figs. 3A–F). The α2A-ADR antibody labeled the primary cells diffusely over the plasma membrane (Fig. 3A) and the IR clearly overlapped with that of 2M6 (Fig. 3C). Alpha2B-ADR IR could not be detected in the cultures (Figs. 3D, 3F). Quantitative RT-PCR analysis was performed to analyze the relative mRNA levels of all three α2-ADR subtypes in primary MCs. We observed high levels of α2A-ADR mRNA and close to background levels of α2B-ADR and α2C-ADR mRNA (Fig. 3G). 

Figure 3

 

Expression of α2-ADRs in primary chick MC culture. Fluorescence micrographs of immunocytochemistry for (A) α2A-ADR, (B, E) 2M6, (C) merged micrographs (A, B), (D) α2B-ADR, and (F) merged micrographs (D, E). (G) Quantitative RT-PCR analysis of α2-ADRs in primary MC culture showing high levels of α2A-ADR mRNA and background levels of α2C-ADR and α2B-ADR mRNA. Bar graph is mean ± SEM, n > 5. Scale bar in (F) is 50 μm, valid also for (A–E).

Stimulation of α2-ADRs Leads to Activation of ERK1/2 Mapkases.

We first studied if stimulation of α2-ADRs by BMD in the cultured primary MCs led to activation of ERK1/2 by P-ERK1/2 immunohistochemistry and Western blot analysis. To maintain low basal levels of P-ERK1/2, the cultured MCs were serum-starved for 4 hours, then treated with BMD or vehicle and analyzed after different time points (Fig. 4A). We observed bright P-ERK IR in the primary MC culture at 10 minutes and lower levels at 30 minutes (Figs. 4B–J). The 2M6 IR verified the MC phenotype in the cells (Figs. 4G, 4J). Western blot analyses were performed to quantify the P-ERK levels (Fig. 4K). P-ERK levels were normalized by β-actin levels (Fig. 4L) or by total ERK levels (Supplementary Fig. S1). Densitometric analysis showed that P-ERK levels increased within 5 minutes after BMD treatment with 3-fold increase in peak levels at 10 minutes, with a gradual decrease to lower levels by 30 minutes (Fig. 4L; Supplementary Fig. S1). 

Figure 4

 

Brimonidine induces P-ERK1/2 in cultured primary chick MCs. (A) Experimental outline. Serum-starved primary MCs treated with 100 μM BMD, and cells were analyzed after 0, 5, 10, 15 and 30 minutes. (B–J) Fluorescence micrographs showing immunocytochemistry for (B, E, H) P-ERK and (C, F, I) 2M6 at (B–D) 0 minutes, (E–G) 10 minutes, and (H–J) 30 minutes after BMD treatment. (K) Western blot analysis of P-ERK in BMD-treated primary MCs at 0, 5, 10, 15, and 30 minutes after treatment. (L) Bar graph with densitometry of P-ERK levels normalized by β-actin levels. Normalization to total ERK showed similar results (Supplementary Fig. S1). Bar graph is mean ± SEM, n = 3 (**P < 0.001, ***P < 0.0001) analyzed by one-way ANOVA and Tukey's post hoc test. Scale bar in (J) is 30 μm, valid also for (B–I).

Blocking of α2-ADRs Attenuated ERK1/2 Activation in Primary MCs

To assess whether the activation of ERK1/2 by BMD in MCs was specific to α2-ADR signaling, we added the selective α2-ADR blocker yohimbine (Fig. 5A). Immunocytochemistry showed that the P-ERK IR increase was abolished with yohimbine (Figs. 5B–J). Cells treated only with yohimbine did not alter the basal levels of P-ERK IR (Figs. 5K–M). Western blot analyses verified the effects (Fig. 5N). Densitometric analysis showed that P-ERK levels were reduced in a concentration-dependent fashion with 100 μM yohimbine reducing P-ERK to control levels (Fig. 5O). 

Figure 5

 

Blocking of BMD-induced P-ERK1/2 in primary MCs in culture by yohimbine treatment. (A) Experimental outline. Serum-starved primary MCs pretreated with 10, 20, and 100 μM α2-ADR antagonist yohimbine or vehicle (control) for 20 minutes followed by treatment with 100 μM BMD or vehicle for 10 minutes and then analyzed. Fluorescence micrographs showing immunocytochemistry for (B, E, H, K) P-ERK and (C, F, I, L) 2M6 on (B–D) control, (E–G) BMD, (H–J) BMD+100 μM yohimbine, and (K–M) 100 μM yohimbine-treated primary MCs. (N) Western blot analysis of P-ERK levels in MCs. (O) Bar graph with densitometry of P-ERK levels normalized by β-actin levels. Bar graph is mean ± SEM, n = 3 (*P < 0.01, ***P < 0.0001) analyzed by one-way ANOVA and Tukey's post hoc test. Scale bar in (M) is 30 μm, valid also for (B–L).

Brimonidine Triggers Phosphorylation of Tyrosine Residues in the EGFRs

To study if α2-ADRs transactivate EGFRs in MCs, we analyzed the phosphorylation of the EGFR at tyrosine residues 1173 and 1068 of the EGFR after BMD stimulation (Fig. 6A). Cells were treated with EGF as a positive control. Immunocytochemistry with antibodies to P-EGFR (Y1173) and (Y1068) produced distinct IR in MCs after BMD stimulation (Figs. 6F–I). Epidermal growth factor treatment of the MC cultures produced a similar pattern (Figs. 6J–M). Control cells did not have any increased levels of P-EGFR IR (Figs. 6B–E). In parallel, Western blot analysis was performed to determine the P-EGFR levels after BMD treatment at different time points (Fig. 6N). The result showed that the P-EGFR (Y1173) levels doubled within 5 minutes after BMD treatment and gradually decreased to basal levels by 60 minutes (Fig. 6O). We were not able to get Y1068 antibody to work in the Western blot analysis. The levels of total EGFR protein were unchanged at the different time points after BMD stimulation (Fig. 6N). 

Figure 6

 

Brimonidine stimulates the phosphorylation of EGFR in primary MCs. (A) Experimental outline. Serum-starved primary MCs were treated with 100 μM BMD or control (vehicle), 0.5 μg/mL EGF and cells were analyzed after 0, 5, 10, 15, 30, and 60 minutes. (B–M) Immunocytochemistry was performed on harvested cells, 10 minutes after the treatment with BMD, EGF, or vehicle. Fluorescence micrographs showing (B–E) control, and (F–I) BMD- and (J–M) EGF-treated cells, labeled for phospho-Y1173 or -Y1068 of the EGFR and for Sox2. (N) Western blot analysis of pEGFR (Y1173) and total EGFR levels in MCs. Protein extracts obtained from BMD-treated cells at 0, 5, 10, 15, 30, and 60 minutes after treatment. (O) Bar graph with densitometry of pEGFR (Y1173) and EGFR levels. Bar graphs are mean ± SEM, n = 3 (*P < 0.01, ***P < 0.0001) analyzed by one-way ANOVA and Tukey's post hoc test. Scale bar in (M) is 50 μm, valid also for (B–L).

Src-Kinase is Required for the Brimonidine-Induced ERK1/2 Response in MCs

Src is a nonreceptor tyrosine kinase that is known to play a role in G-protein coupled receptor and as such in α2-ADR signal transduction. Studies also have shown that transactivation of EGFR by G-protein–coupled receptors in different systems is dependent of Src. ^30,31 To study whether Src was involved in the BMD-induced ERK1/2 response in MCs, we studied the effect of PP1 and PP2, two potent Src kinase inhibitors ³⁰ on BMD-induced ERK1/2 and EGFR activation in the cultured primary MCs. Cells were treated with BMD or EGF after pretreatment with PP1 (Fig. 7A). PP1 inhibited the BMD-induced but not the EGF-induced P-ERK IR as demonstrated by P-ERK, 2M6 double-labeled cells (Figs. 7B–J, N–S). Cells treated with only PP1 did not alter the background levels of P-ERK (Figs. 7K–M). Western blot analysis showed that P-ERK levels remained on control levels when PP1 was added before BMD treatment (Figs. 7T, 7U). There was no reduction of P-ERK levels after PP1 and EGF treatment (Figs. 7V, 7W). PP2 gave similar results as PP1 (Figs. 7T', 7U'; Supplementary Fig. S2). The results confirmed that Src-kinases are involved in BMD-induced, but not in the EGF-induced ERK1/2 activation in MCs. 

Figure 7

 

Effects of Src kinase inhibitors (PP1, PP2) on BMD- or EGF-induced P-ERK1/2 or P-EGFR (Y1173) in primary Müller cells. (A) Experimental outline. Serum-starved primary MCs pretreated with 5 μM PP1, 5 μM PP2, or control (vehicle) for 20 minutes followed by treatment with 100 μM BMD or 0.5 μg/mL EGF or vehicle for 10 minutes and analysis. Fluorescence micrographs showing immunocytochemistry for (B, E, H, K, N, Q) P-ERK and (C, F, I, L, O, R) 2M6 on (B–D) vehicle, (E, F) BMD, (H–J) BMD+PP1, (K–M) PP1, (N–P) EGF, and (Q–S) EGF+PP1-treated primary MCs. (T, T', V) Western blot analysis of P-ERK in MCs and (X) Western blot analysis of P-EGFR (Y1173) and EGFR levels in MCs. (U, U', W) Bar graphs with densitometry of P-ERK levels normalized by β-actin levels. Normalization to total ERK showed similar results (Supplementary Fig. S2). (Y) Bar graph with densitometry of P-EGFR (Y1173) and EGFR levels. Bar graphs are mean ± SEM, n = 3 (***P < 0.0001) analyzed by one-way ANOVA and Tukey's post hoc test. Scale bar in (S) is 30 μm, valid also for (B–R).

To test whether Src inhibitor affected the phosphorylation of EGFR after BMD treatment in MCs, we studied the effect of PP2 on BMD-induced P-EGFR (Y1173) in primary MCs. Western blot analysis showed that the P-EGFR (Y1173) levels were inhibited by PP2 (Figs. 7X, 7Y). This result indicated that BMD-induced EGFR phosphorylation in MCs involves Src kinase in contrast to EGF-induced EGFR phosphorylation. 

MMP-Mediated Ligand-Dependent Activation of ERK1/2 in MCs

In addition to Src, it has been shown that transactivation of EGFRs by G-protein–coupled receptors involves MMP activity, ³² as introduced earlier in this article. To assess if ERK1/2 activation in the MCs is dependent on the MMP activity, we studied the effect of the broad-spectrum MMP inhibitor GM6001³³ on BMD-induced ERK1/2 activation. We also studied if GM6001 had any effect on EGF-induced ERK1/2 activation. Primary MCs were treated with BMD or EGF after pretreatment with GM6001 (Fig. 8A). Serum starvation gave low background levels of P-ERK and BMD produced bright P-ERK IR (Figs. 8B–G). We found that pretreatment with 50 μM GM6001 inhibited the BMD-induced ERK1/2 activation, as demonstrated by P-ERK, 2M6 double-positive cells (Figs. 8H–J). GM6001 did not change the basal level of pERK1/2 IR (Figs. 8K–M). GM6001 treatment did not reduce P-ERK IR elicited after added EGF (Figs. 8N–S). Western blot analysis confirmed the inhibitory effect of GM6001 on BMD-induced but not on EGF-induced ERK1/2 activation in primary MCs (Figs. 8T, 8V). The result showed that P-ERK levels were significantly lower after GM6001 treatment before BMD treatment (Fig. 8U), but not before EGF treatment (Fig. 8W). The results are consistent with the involvement of MMP activity in the BMD-induced P-ERK activation. 

Figure 8

 

Effect of MMP inhibitor, GM6001, and EGFR kinase inhibitor, AG1478, on BMD- or EGF-induced P-ERK1/2 in primary MCs. (A) Experimental outline. Serum-starved primary MCs pretreated with 50 μM GM6001 or control (vehicle) for 30 minutes followed by treatment with 100 μM BMD, 0.5 μg/mL EGF, or vehicle for 10 minutes and analysis. Fluorescence micrographs showing immunocytochemistry for (B, E, H, K, N, Q) P-ERK and (C, F, I, L, O, R) 2M6 on (B–D) vehicle, (E, F) BMD, (H–J) BMD + GM6001, (K–M) GM6001, (N–P) EGF, and (Q–S) EGF+GM6001-treated primary MCs. (T, V, X) Western blot analyses of P-ERK in MCs treated with (T) BMD + GM6001, (V) EGF + GM6001, and (X) BMD + AG1478. (U, W, Y) Bar graphs with densitometry of P-ERK levels normalized by β-actin levels. Normalization to total ERK showed similar results (Supplementary Fig. S3). Bar graphs are mean ± SEM, n = 3 (***P < 0.0001) analyzed by one-way ANOVA and Tukey's post hoc test. Scale bar in (S) is 30 μm, valid also for (B–R).

Our results indicate that BMD triggers an Src-dependent activation of a ligand-mediated phosphorylation of the EGFR and a P-ERK response that is mediated by MMP activity. The BMD-induced ERK1/2 activation required EGFR activation and we studied the effect of AG1478, a potent EGFR kinase inhibitor ³⁴ on BMD-induced ERK activation in MCs. Western blot analysis showed that inhibition of the EGFR by AG1478 reduced the BMD-induced ERK activation in MCs (Figs. 8X, 8Y; Supplementary Fig. S3). The result is consistent with that ERK1/2 activation by BMD in part requires EGFR activation in MCs. 

One of the endogenous EGFR ligands that may mediate such a response is HB-EGF. The release of HB-EGF from its membrane-anchored precursor, pro-HB-EGF is under the control of proteolytic activity by MMPs. We confirmed the expression of HB-EGF in chicken retina and in the primary MC culture by qRT-PCR analysis (Fig. 9). The amplification levels of HB-EGF mRNA were in the same range as those of Sox2 mRNA. 

Figure 9

 

Expression of HB-EGF mRNA in chicken retina and in primary chick MC culture. Quantitative RT-PCR analysis of HB-EGF and Sox2 mRNA in E18 normal chick retina and primary MC culture. Bar graph showing the relative mRNA levels in relation to the levels of β-actin mRNA. Bar graph is mean ± SEM, n > 5.

Discussion

In this study, we have characterized the expression of α2-ADRs in E18 chick embryo retina and in cultured primary MCs. Our results show that α2A-ADRs are expressed by chicken MCs and that α2A-ADR agonists trigger a robust MAPK response with phosphorylation of ERK1/2 in the MCs, both in vivo and in vitro. The response is mediated by Src-kinase and involves both ligand-dependent transactivation of EGFRs on the MCs and Src-dependent EGFR ligand-independent ERK1/2 activation. 

The data show that α2A-ADRs are localized to the cell body of E18 chick MCs and on processes at the outer limiting membrane (Fig. 1A). Alpha2A-ADR+ cells were also seen in the ganglion cell layer. The pattern in the E18 chick retina was similar to published data in the rat, monkey, and human retina. ^19,35 We did not detect α2B-ADRs in MCs in the chick retina or in primary MC cultures (Figs. 1F, 2D). Alpha2B-ADRs were seen only as a distinct immunoreactivity in photoreceptor outer segments. In rat, α2B-ADRs were present in all layers of the retina. ¹⁹ The pattern of α2B-ADRs across species seems to be less well conserved than that of α2A-ADRs. Quantitative RT-PCR analysis of cultured primary MCs confirmed that α2A-ADR mRNA was highly expressed and the α2B- and α2C-ADR mRNAs were low (Figs. 2G, 3G). 

We used the α2A-ADR agonist BMD to stimulate MCs. Brimonidine is used in glaucoma treatment to reduce IOP, ²⁰ but the specific mechanisms are not clear. Brimonidine stimulation resulted in a robust activation of MAPK/ERKs in MCs. The results are consistent with earlier studies showing ERK1/2 activation in rat MCs after systemic administration of the α2-ADR agonist xylazine. ¹⁶ The specificity of the ERK1/2 activation by BMD was shown by the ability of the α2-ADR-antagonist yohimbine, to completely block the transient increase in P-ERK1/2 (Figs. 5H–J, 5N, 5O). 

Several studies have shown, in different cell types, that activation of ERK1/2 by G-protein–coupled receptors requires or involves transactivation of the EGFR. ²¹ Our results show that BMD stimulation of MCs transactivates EGFR and these results are consistent with dexmedetomidine-induced transactivation of EGFR in astrocytes and intact brain tissues. ^36,37 Cytosolic Src-kinases have been implicated in α2-ADR–induced transactivation of EGFR ³⁰ and based on the results from the Src inhibitor PP2, our study shows that BMD-induced transactivation of EGFR in MCs is strictly dependent on Src-kinase activity (Figs. 7X, 7Y). MAPK-activation in PC12 cells by epinephrine requires Src but has both an EGFR-dependent and an independent component. ³² Our results support this observation. The Src blocker PP1 completely abolished the BMD-induced ERK1/2 activation (Figs. 7H–J, 7T, 7U), but could not block ERK1/2 activation by added EGF, showing that ligand-activated EGFR signaling in MCs did not require Src (Figs. 7Q–S, 7V, 7W). The BMD stimulation of MCs resulted in phosphorylation of Y1068 and Y1173 in the intracellular domain of the EGFR (Figs. 6B–J). These tyrosine residues are major autophosphorylation sites that allow interaction of adaptor proteins Grb2 and Shc with the receptor and that mediate Ras-activated MAPK signaling, including ERK1/2. ³⁸ The autophosphorylation is indicative of a ligand-mediated EGFR activation. In line with these data, our results show that the EGFR inhibitor AG1478 could reduce the BMD-induced ERK1/2 activation in the MCs (Figs. 8X, 8Y). 

Initially, the α2-ADR–induced transactivation of EGFR was thought to be mediated by a direct effect by Src kinases. However, MMP, including the members of the membrane-bound ADAM (A disintegrin and metalloproteinase) family, have been shown to serve key roles in G-protein–coupled receptor-induced transactivation of EGFR signaling. ^21,39 Their catalytic activity is responsible for the proteolytic release of growth factors, such as the membrane-bound EGFR ligands HB-EGF or amphiregulin, which activate EGFRs in an auto- or paracrine mode of action. ³² Src-kinases contribute to MMP activation by a direct interacting with proline-rich Src-homology (SH3) domains on the cytosolic portion of the MMP proteins. ⁴⁰ The BMD-induced ERK1/2 activation in cultured primary MCs was significantly abrogated by GM6001, an MMP inhibitor, indicating a ligand-dependent mechanism for transactivation of EGFR in MCs. However, GM6001 was unable to completely block the BMD-induced ERK1/2 activation (Figs. 8H–J, 8T, 8U), showing that the EGFR/ligand-independent activation of ERK1/2 via Src-kinase, as already discussed, along with ligand-dependent EGFR activation, is present in the BMD-stimulated MCs. 

Our cultures of primary MCs express HB-EGF (Fig. 9) but we have not shown that it is HB-EGF that mediates the ligand-depending EGFR transactivation after BMD stimulation. HB-EGF is required for injury-induced MC proliferation and dedifferentiation into progenitor cells and is capable of inducing proliferation in the uninjured retina along with Wnt/β-catenin signaling. ⁶ Sustained ERK1/2 activation in chick MCs after excitotoxic retinal damage or after growth factor treatment, leads to proliferation and de-differentiation, ^3,7,8 and activation of ERK1/2 in MCs plays a protective role against excitotoxic insults. ⁴¹ Our results open up for the possibility to modulate the MC response after injury but it remains to be studied if the transient ERK1/2 activation in MCs elicited by BMD stimulation is involved in any of the neuroprotective or regenerative effects seen by BMD at various retinal injuries. 

In conclusion, our results show that chick MCs express α2A-ADR and that stimulation by BMD leads to Src-mediated ERK1/2 signaling, which includes ligand-dependent transactivation of the EGFR. 

Supplementary Materials

pdf S1 

Acknowledgments

We thank Nobuyuki Takei and Hiroyuki Nawa, Brain Research Institute, Niigata, for suggesting the MMP pathway. 

Supported by the Swedish Research Council (Vetenskapsrådet) M 12187, Ögonfonden, and Stiftelsen Kronprinsessan Margaretas Arbetsnämnd för synskadade. 

Disclosure: M. Harun-Or-Rashid, None; N. Lindqvist, None; F. Hallböök, None