Experience-Dependent Regulation of NMDA Receptor Subunit Composition and Phosphorylation in the Retina and Visual Cortex

Abstract

Purpose.: Experimental manipulation of experience during development can have profound effects on the functioning of the resulting circuits. N-methyl-d-aspartate glutamate receptor (NMDAR) activity is required for the establishment and refinement of neural circuits during development. In the present study, the authors addressed the issue of experience-dependent regulation of NMDARs by examining the effects of visual experience and deprivation on subunit composition and subunit phosphorylation of NMDAR in the retina and visual cortex.

Methods.: Total homogenates were prepared from retinas and visual cortices of 30-day-old (P30) Wistar rats, raised either in a normal 12-hour light/12-hour dark cycle (normal-reared [NR]) or in complete darkness from birth (dark-reared [DR]). Some of the DR animals were exposed to light for 6 hours at P30 (DR+6h). Immunoblotting was performed for the NMDAR subunits, NR2A and NR2B, and for the phosphorylated NR2B subunit protein at serine 1303 (pNR2B-Ser1303).

Results.: Dark rearing for 1 month decreased the NR2A/NR2B ratio and increased the level of phosphorylation of NR2B subunit at Ser1303 in the retina and visual cortex. Light exposure at P30 reversed the effects of visual deprivation on NMDAR composition and NR2B phosphorylation in both regions.

Conclusions.: These results indicated that NMDAR subunit composition and NR2B phosphorylation at Ser1303 is regulated bidirectionally by visual experience and deprivation in rat retina and visual cortex.

Neuronal activity is important for the formation and maturation of neural circuits in the central nervous system. Although the basic connections that define these circuits are genetically determined, refinement of these connections is strongly influenced by neuronal activity. Neuronal activity is driven by sensory experience. ^1,2 Experimental manipulation of experience during development can have profound effects on the functioning of the resultant circuits. ^3–6 The central nervous system modifies the properties of neural circuits according to the changing needs of the environment under mechanisms of synaptic plasticity and has been extensively studied in the visual system and especially in the visual cortex. ^7–11  

Excitatory synaptic transmission in the central nervous system is mediated by glutamate and is dependent on activation of the ionotropic glutamate receptors. A key property of the glutamatergic synapse is its plasticity. ¹² NMDARs mediate excitatory synaptic transmission of the glutamatergic synapse. NMDARs are heteromeric complexes consisting of an obligatory NR1 subunit, NR2 subunits that impart functional properties and are encoded by four distinct genes, and, more rarely, NR3 subunits. They have unique properties that classify them as key players in the development of neuronal networks and synapse formation. ¹³ NMDAR activity is required for the establishment and refinement of neural circuits during development by contributing to the formation and maturation of dendritic processes, dendritic spines, and synaptic connections themselves. ^14–16 The role of NMDARs has been well documented in forms of experience-dependent plasticity in the visual system. Blockade of NMDARs prevents ocularity changes in kitten visual cortex after reversed monocular deprivation. ¹⁷ In the visual cortex of light-deprived rats, the normal developmentally regulated shortening of NMDAR currents is delayed, ¹⁸ long-term potentiation is enhanced, and long-term depression is diminished over a range of stimulation frequencies; these effects can be reversed by light exposure. ¹⁹  

Molecular mechanisms underlying glutamatergic synapse plasticity involve dynamic and functional regulation of the NMDARs, such as trafficking, subunit composition, and posttranslational modifications. ^20–22 Visual experience and deprivation bidirectionally modify the NR2A and NR2B subunit composition of NMDARs, and these changes in turn modify the properties of synaptic plasticity in the visual cortex. ^23,24  

Direct phosphorylation of ionotropic glutamate receptors is a key mechanism regulating channel function and receptor localization at synapses. Many serine (Ser)/threonine phosphorylation sites have been identified in NMDA receptor subunits, which are substrates for several kinases, such as protein kinase C (PKC) and Ca/calmodulin-dependent protein kinase II (CaMKII). NR2B is phosphorylated at C-terminal Ser1303, Ser1323, and Ser1480 and several tyrosine sites. Ser1303 phosphorylation is driven by CaMKII, an abundant kinase in the postsynaptic density of glutamatergic synapses, or PKC. ^22,25–27  

In the retina, the NR2A and NR2B subunits have been localized in the inner plexiform layer, ^28–30 and NR2A and NR2B mRNAs are found in virtually every ganglion cell and some amacrine cells. ³¹  

It has been shown that afferent activity regulates dendritic remodeling and stratification of retinal ganglion cells. ^32–34 Tian and Copenhagen ³⁵ have reported that dark rearing blocks both the maturation loss of ON-OFF responsive retinal ganglion cells and the stratification of dendrites in mouse retina. Under the hypothesis that NMDARs might be involved in experience-dependent plasticity of the retina, we examined the effects of manipulating visual experience on subunit composition and subunit phosphorylation of NMDARs. We studied changes in NR2A and NR2B subunit expression and in the phosphorylation of the NR2B subunit at Ser1303 in the rat retina and in the visual cortex with the use of immunoblot analysis. Our results show that the alterations in molecular characteristics of NMDARs observed in the retina and visual cortex are bidirectionally modified by visual experience and dark rearing. 

Methods

Animals and Visual Manipulation

All animal experiments were carried out in accordance with the European Communities Council Directive (86/609/EEC) guidelines for the care and use of laboratory animals and with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Pregnant female rats (Wistar) were housed singly in cages maintained in a 12-hour light/12-hour dark cycle. Two to three days after giving birth, cages of mother and pups were left for normal rearing, 12-hour light/12-hour dark cycle (normal-reared [NR]), or a photon-free room for rearing in complete darkness (dark-reared [DR]). Both DR and NR animal groups were raised until 30 days of age (postnatal day [P] 30). The animals in the dark were cared for with the use of infrared vision goggles under dim infrared light. At P21, NR and DR pups were weaned and transferred into cages of three to four animals. To study the effect of light exposure, some of the 30-day-old DR animals were taken out to the light-dark cycle environment for 30 minutes, 6 hours, or 3 days. Both male and female rats were used, and preparation of homogenates from NR animals was interleaved with those from DR animals. 

Homogenate Preparation from Retinal and Visual Cortex Tissues

Animals were deeply anesthetized with diethyl ether and decapitated. All procedures were performed so as to maintain the tissue fractions at 4°C. Retinas and visual cortices were rapidly removed in ice-cold dissection buffer (ACSF solution: 117 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 25 mM NaHCO₃, 1.2 mM NaH₂PO₄, and 11 mM glucose at pH 7.3–7.4 and were equilibrated with 95% O₂/5% CO₂). Both retinas from the same animal were homogenized in 200 μL ice-cold lyses buffer (2% SDS, 1 mM PMSF, protease inhibitor cocktail [30 μL for 100 mg tissue, Sigma; P2714], phosphatase inhibitor cocktail 1 [1 μL/100 mL lysis buffer, Sigma; P2850], and phosphatase inhibitor cocktail 2 (1 μL/100 mL lysis buffer, Sigma; P5726) by sonication and were stored in aliquots at −80°C. The same procedure was followed for the visual cortices. Protein concentrations were determined with the bicinchoninic acid assay (BCA; Pierce, Rockford, IL). 

Quantitative Immunoblotting and Statistical Analysis

Equal amounts of homogenate proteins, within the linear range of standard curves, were resolved on 7% (wt/vol) sodium dodecyl sulfate-polyacrylamide gels and electrotransferred to 0.45 μm nitrocellulose membranes (Amersham Pharmacia Biotech, Piscataway, NJ). The membranes were blocked in blocking buffer (Tris-buffered saline containing 5% nonfat milk and 0.005% Tween 20) at room temperature for 1 hour and then immunoblotted for NMDAR subunit-specific polyclonal antibodies against NR2A (1:1000; Millipore, Billerica, MA) and phosphoNR2BSer1303 (1:1000; Millipore) overnight at 4°C. After several washes with blocking buffer, blots were incubated with peroxidase-conjugated secondary anti–rabbit IgG (1:3000; Boehringer Mannheim, Indianapolis, IN) for 90 minutes at room temperature. Immunolabeled protein bands were detected using an enhanced chemiluminescence (ECL) system (Amersham Pharmacia Biotech). The membranes could then be stripped for reprobing. Identical stripping conditions were used (Re-blot Plus Mild Solution; Millipore), and membranes were probed with anti–NR2B polyclonal antibody (1:750; Millipore). Molecular weights were determined by comparison with prestained protein molecular weight marker standards (Prestained SDS-PAGE Standards; Bio-Rad, Hercules, CA). Blots were reprobed with anti–tubulin mouse monoclonal antibody (1:20,000; Sigma) rinsed with TBS-Tween and incubated with anti–mouse antibody. Tubulin served as a gel-loading control. 

The ECL-exposed films (Biomax Lightfilm; Kodak, Rochester, NY) were scanned and quantified (Quantity One software; Bio-Rad). The signal of each sample on a blot was normalized to the average signal from NR or DR samples, respectively, to obtain the percentage of average NR or percentage of average DR values, which were compared across different experimental groups using one-way ANOVA. 

Results

Effect of Visual Deprivation on NMDA Receptor Subunit Expression in Rat Retina and Visual Cortex

To determine the effects of sensory experience on NMDAR subunit composition of retina and visual cortex, total homogenates were prepared from retinas and cortex of 30-day-old Wistar rats, raised either in a normal 12-hour light/12-hour dark cycle (NR) or in complete darkness from birth (DR). Immunoblotting was performed for each of the NMDAR subunit proteins, NR2A and NR2B. 

In the rat retina, statistical analysis of the data revealed a significant effect of dark rearing on NR2A protein levels (NR = 100% ± 7.85% of NR, n = 11; DR = 60.11% ± 3.45% of NR, n = 9; P < 0.05; Fig. 1A). Our results have shown that the level of NR2A protein from the retinas of DR animals is reduced by 40% compared with NR controls. However, dark rearing did not affect NR2B protein levels in retina (NR = 100% ± 3.70% of NR, n = 24; DR = 102.25% ± 5.02% of NR, n = 22; P > 0.05; Fig. 1C). 

Figure 1.

 

Effect of visual deprivation on NMDA receptor subunit expression in rat retina and visual cortex. NR2A and NR2B subunit expression levels of the NMDA receptor studied by immunoblotting in the retina (A, C) and visual cortex (B, D) of NR and DR rats. In the retina, dark rearing reduced NR2A protein levels by 40% (A), whereas NR2B protein levels were not affected (C). In the visual cortex, dark rearing reduced NR2A protein levels by 32% (B), whereas NR2B protein levels were not affected (D). Results are expressed as mean ± SEM (n = 9–24) and representative Western blot analyses;*P < 0.05.

In the rat visual cortex, statistical analysis of the data revealed a significant effect of dark rearing on NR2A levels (NR = 100% ± 6.83% of NR, n = 12; DR = 68.24% ± 5.76% of NR, n = 12; P < 0.05; Fig. 1B). Our results have shown that the level of NR2A protein from the visual cortex of DR animals is reduced by 32% compared with NR controls. Dark rearing did not affect NR2B protein levels in the visual cortex, similar to the retina (NR = 100% ± 5.85% of NR, n = 14; DR = 97.36% ± 4.71% of NR, n = 13; P > 0.05; Fig. 1D). 

Calculating the ratio of NR2A to NR2B, it was evident that dark rearing decreased the NR2A/NR2B ratio to 40% or 32% of NR control values in the retina or visual cortex of P30 rats, respectively. 

Effect of Visual Deprivation on the Phosphorylation of NMDA Receptor NR2B Subunit in Rat Retina and Visual Cortex

To examine whether visual experience has a further effect on NR2B subunit phosphorylation, Western blot assays were performed for total NR2B and phosphorylated NR2B subunit protein at Ser1303 (pNR2B-Ser1303) in total homogenates of retina and visual cortex of both NR and DR 30-day-old rats. 

In the rat retina, statistical analysis of the data revealed a significant effect of dark rearing on pNR2B-Ser1303 levels (NR = 100% ± 5.79% of NR, n = 19; DR = 123.58% ± 7.32% of NR, n = 18; P < 0.05; Fig. 2A). Similarly, a significant effect of visual experience was found on pNR2B-Ser1303 levels in the rat visual cortex (NR = 100% ± 3.63% of NR, n = 14; DR = 126.15% ± 6.75% of NR, n = 9; P < 0.05; Fig. 2B). Our results have thus shown that the level of phosphorylation of the NR2B subunit at Ser1303 is increased by 24% and 26% compared with NR controls in the retina and the visual cortex of DR animals, respectively. 

Figure 2.

 

Effect of visual deprivation on the phosphorylation of NMDA receptor NR2B subunit in rat retina and visual cortex. Expression levels of the phosphorylated NR2B at the Ser1303 residue (pNR2B-Ser1303) studied by immunoblotting in the retina (A) and visual cortex (B) of NR and DR rats. In the retina, dark rearing increased pNR2B-Ser1303 protein levels by 24% (A). In the visual cortex, dark rearing increased pNR2B-Ser1303 protein levels by 26% (B). Results are expressed as mean ± SEM (n = 9–19) and representative Western blot analyses; *P < 0.05.

Light Exposure Reverses the Effect of Visual Deprivation on NMDAR Composition and Phosphorylation in Rat Retina and Visual Cortex

It has been demonstrated that NMDAR subunit composition of the visual cortex of DR animals can be restored to the value characteristic of age-matched NR animals exposed to light. ^23,24 In the present study, we examined whether this bidirectional experience-dependent regulation of NMDARs subunit composition could be induced in the rat retina as in the visual cortex. Given that we have shown that NR2A protein levels, but not NR2B, are affected by dark rearing in both retina and visual cortex, we performed Western blot analysis for the NR2A subunit. To address this question, 30-day-old DR animals were exposed to light for 30 minutes, 6 hours, and 3 days. Exposure to light for 30 minutes did not change the NR2A levels, and exposure for 3 days and for 6 hours had a similar effect (data not shown). 

DR animals exposed to light for 6 hours (DR+6h) showed a robust increase of NR2A levels in the rat retina that completely reversed the DR effect (NR = 100% ± 6.83% of NR, n = 11; DR = 60.11% ± 3.45% of NR, DR + 6 = 97.23% ± 5.22%. n = 9; P < 0.05; Fig. 3A). 

Figure 3.

 

Effect of light exposure after dark rearing on NR2A subunit expression and NR2B subunit phosphorylation in rat retina and visual cortex. NR2A subunit expression levels of the NMDA receptor (A, B) and expression levels of the phosphorylated NR2B at the Ser1303 residue (pNR2B-Ser1303) (C, D) studied by immunoblotting in the retina (A, C) and visual cortex (B, D) of NR and DR rats. The decrease in NR2A expression levels observed in DR animals was reversed by 6-hour light exposure in both retina (A) and visual cortex (B) to NR values. Increases in pNR2B-Ser1303 expression levels observed in DR animals were reversed by 6-hour light exposure in both retina (C) and visual cortex (D) to NR values. Results are expressed as mean ± SEM (n = 9–18) and representative Western blot analyses; *P < 0.05.

Our results revealed a significant increase of NR2A levels in the visual cortex of DR+6 animals, which completely reversed the dark-rearing effect, as in the retina (NR = 100% ± 7.83% of DR, n = 15; DR = 68.24% ± 5.76% of NR; DR + 6 = 93.31% ± 6.71%, n = 18; P < 0.05; Fig. 3B). 

To explore further the experience-dependent regulation of NMDA receptors, we examined whether light exposure could have an effect on NR2B subunit phosphorylation. We showed that the phosphorylation of NR2B is affected by dark rearing in both retina and visual cortex (Fig. 2); we then performed Western blot analysis for the phosphorylated NR2B subunit protein at Ser1303 (pNR2B-Ser1303). Light exposure for 30 minutes did not change the pNR2B-Ser1303 levels, and exposure for 3 days and for 6 hours had a similar effect (data not shown). 

Exposure of DR rats to 6 hours of light caused a significant decrease of pNR2B-ser1303 levels in the rat retina, which restored the levels to the age-matched NR value (NR = 100% ± 5.79% of DR, n = 15; DR = 123.58% ± 7.32% of NR; DR + 6 = 101.63% ± 3.61%, n = 18, P < 0.05; Fig. 3C). 

As in the retina, a significant decrease was found for pNR2B-Ser1303 levels in the visual cortex of the same group of rats (DR+6), which completely reversed the dark-rearing effect (NR = 100% ± 3.63% of NR, n = 14; DR = 126.15% ± 6.75% of NR; DR+6 = 99.37% ± 6.54%, n = 9; P < 0.05; Fig. 3D) 

Discussion

The neural retina is part of the central nervous system and, as such, contains the same kinds of synaptic machinery that are present in the rest of the central nervous system. ³⁶ To examine whether the retina shares analogous developmental mechanisms with the rest of the visual system, we examined the effects of manipulating visual experience in the retina and the visual cortex. 

It is well established that NMDARs are involved in some forms of experience-dependent plasticity in the visual cortex. ^18,19 In the present study, we examined the effects of 1 month of dark rearing on NMDARs of the rat retina and visual cortex. We chose an extensive period of dark rearing that covers the period from birth to the peak of the critical period reported for the effects of monocular deprivation on cortical reorganization in the rat. ³⁷ Furthermore, this dark-rearing protocol was selected based on findings indicating that NMDA-mediated excitatory postsynaptic currents (EPSCs) in rat visual cortical cells ^18,38 and retinal ganglion cells ³⁹ progressively decrease during the first postnatal month and reach adult levels at P30. Finally, it was based on findings indicating that visual deprivation for the same period alters the dendritic maturation properties of the inner retina. ^40,41  

In the present study, animals that were dark reared showed a reduced ratio of NR2A/NR2B protein in the retina and visual cortex. This change is exclusively attributed to the reduced expression of the NR2A subunit levels because the NR2B protein levels were not affected in either region. Our results indicate that NMDAR subunit composition is regulated by visual experience in the rat retina, as in the visual cortex. Our findings concerning the visual cortex are consistent with previous studies using immunoblotting, which showed that the same protocol of visual deprivation reduces the ratio of NR2A/NR2B protein. ²³ Furthermore, our data showing a reduced NR2A/NR2B protein ratio in the DR retina provide evidence that the same pattern of light alterations is followed by the rat retina. At the mRNA level, Guenther et al., ³⁹ using RT-PCR, have also reported that the ratio of the percentage of retinal ganglion cells expressing NR2A to the ones expressing NR2B was lower in DR animals than in NR animals. 

For the visual cortex, it has been suggested by others ^23,24 that the NR2A/NR2B subunit ratio is a key factor in the regulation of the kinetics of NMDAR. Previous studies have shown that animals dark reared for 1 month compared with NR animals are characterized by a delay of the developmental shortening of NMDAR EPSC duration and larger current amplitudes both in rat retina and visual cortex. ^18,38,39 Because it is known that NR2A-containing receptors have faster kinetics, ^24,42,43 we can suggest that the change observed in NMDA receptor subunit composition in the retina of rats dark reared for the first postnatal month could be responsible for the enhanced NMDAR currents. 

In rat visual cortex it has been demonstrated that the experience-dependent regulation of NMDA receptor subunit composition during postnatal development is bidirectional. ^23,24 We reasoned that this experience-dependent effect might be the same in the retina. Animals that were dark reared for the first month were exposed to light for several time intervals. We observed that 6 hours of normal visual experience was enough to restore the NR2A/NR2B ratio to the value characteristic of age-matched NR animals in rat retina similar to visual cortex. This observation seems to correlate with the restoration of the NMDA current recorded in retinal ganglion cells of DR animals that were returned to normal light/dark cycle for 5 days. ³⁹ Our results thus indicate that light exposure reverses the effect of visual deprivation on NMDAR composition in the rat retina and that the retina and visual cortex share the same tendency as far as the change of NMDAR subunit composition under the effect of manipulating visual experience. 

Our data showed that the expression level of NR2B subunit phosphorylation at Ser1303 in animals that were dark reared was significantly higher than in age-matched NR animals and therefore demonstrate for the first time that NR2B subunit phosphorylation is regulated by visual experience in the rat retina and visual cortex. Furthermore, 6 hours of light exposure were enough to restore phosphorylated NR2B subunit levels to the values of age-matched NR animals. Our results thus indicate that the experience-dependent regulation of NR2B phosphorylation at Ser1303 during postnatal development is bidirectional in both retina and visual cortex. 

Phosphorylation of synthetic peptides has indicated that the Ser1303 site on NR2B is a PKC substrate in vitro and that PKC can directly phosphorylate Ser1303, leading to enhanced currents through NMDA receptor channels. ⁴⁴ In addition to PKC, CaMKII-α enhances the desensitization of NR2B-containing NMDA receptors. ⁴⁵ It has also been reported that CaMKII phosphorylation of NR2B Ser1303 promotes the dissociation of CaMKII-α from NR2B ^46,47 and that CaMKII is increased by dark rearing in the retina. ⁴⁸ By taking into account these reports, we can suggest that the higher expression level of the phosphorylated NR2B subunit in animals that were dark reared is consequently followed by the dissociation of preformed CaMKII-NR2B complexes, which may result in lower desensitization of NR2B currents and, thus, enhanced NMDAR currents. Therefore, the enhanced NMDAR currents observed by others ^18,38,39 in DR animals for the first postnatal month could be attributed to the increased phosphorylation of NR2B at Ser1303 by either PKC or CaMKII. This provides a novel molecular mechanism for bidirectional, experience-dependent synaptic changes in the retina and visual cortex. 

In conclusion, we propose that translational (subunit expression) and posttranslational (phosphorylation) modifications of NMDA receptor subunits might directly account for the functional changes of the NMDARs that take place during visual experience manipulation. However, we cannot exclude that posttranslational modifications may indirectly—through trafficking of the receptors—play a role in the mechanism of glutamate synapse plasticity. 

Several studies have established a role for NMDAR-mediated activity in dendrite development. ^14,49–52 Furthermore, Wong et al. ⁵³ have demonstrated that dendritic motility is dependent on NMDA and non-NMDA receptors in developing chick retina. ⁵³ A recent study ⁵⁴ has shown that the manipulation of NMDAR composition directly by exogenous expression or knockdown of NR2A or NR2B subunits regulates the local dendritic arbor architecture and affects dendritic branch clustering of tectal neurons in Xenopus tadpoles and that exogenous NR2B expression restricts dendritic growth and branch clustering in response to light input. Considering that shifting NMDAR subunit composition can have an effect on structural neuronal plasticity, we reasoned that the light-deprived modification in the NR2A/NR2B ratio of retinal NMDA receptors observed in the present study could account for the slowing of developmental segregation of retinal ganglion cell dendrites into ON and OFF synaptic pathways in dark-reared mice. ^40,41  

Footnotes

 Supported by the European Social Fund, Operational Program for Educational and Vocational Training II, Program PYTHAGORAS ΙΙ, and Polembros Shipping Limited.

Footnotes

 Disclosure: M. Giannakopoulos, None; E.D. Kouvelas, None; A. Mitsacos, None