Rapid glutamate receptor 2 trafficking during retinal degeneration

LIRD led to dramatic photoreceptors loss by post-light exposure day 7 (pLX7) (Figure 1A), demonstrating that LIRD is a "fast degenerating" animal model of retinal degeneration [32]. Consistent with our previous results [8, 12], light stress differentiated the mouse retina into survivor zones, where stressed photoreceptors and retinal neurons persisted, and light-damage zones, where rods and cones died (Figure 1B). Despite that DAPI staining showed a normal thickness of inner nuclear layer (INL) in both the survivor and light-damage zones and histological analysis of retina revealed normal lamination and topology at light onset (pLX0) through pLX1 (Figure 1A and 1B), protein levels of GluR2 showed immediate response to light stress. Specifically, protein levels of GluR2 increased immediately at pLX0, peaked at pLX7, and recovered to almost control levels by pLX30 (Figure 1C and 1D). GluR2 immunostaining of control tissues (postnatal day 80) revealed GluR2-immunoreactive puncta aggregating in the outer plexiform layer (OPL) with more GluR2-immunoreactive puncta found throughout the OFF and ON sublamina (Figure 1E).

GluR2 is assembled, with other GluR subunits, into iGluR channels in the postsynaptic membrane [33]. The C-terminus of GluR2 contains a PDZ domain that binds GluR2 to trafficking proteins including GRIP1, ABP, PICK1 [34‐36], and stargazin [37]. GluR2 is anchored at synaptic and intracellular membranes by ABP/GRIP but cycles between these membranes in association with PICK1 [38]. Consistent with a previous report [39], PSD-95 was most prominent in the rod spherules and cone pedicles of the OPL, weak PSD-95 labeling was also present in the inner plexiform layer (IPL) (Figure 2A). The retinal distributions of GluR2 trafficking proteins GRIP1, ABP, PICK1 and stargazin showed labeling in the OPL and IPL (Figure 2B-E). These data confirm the presence of GluR2-related PSD proteins such as ABP and GRIP1 in the retina as demonstrated by Gabriel et al [40]. Similar to findings in GluR2 protein expression, PSD-95, GRIP1, ABP, and stargazin expression were increased by light stress, levels of PSD-95, ABP, and GRIP1 peaked at pLX1 and then recovered, while the levels of stargazin were elevated from pLX1 to pLX30, the last time point in the experiment (Figure 2F and 2G). Protein levels of PKCα and PICK1 were unaltered by light exposure from pLX1 to pLX30 (Figure 2F and 2G). Western blotting images of PSD-95 protein levels revealed three adjacent bands (Figure 2F), potentially due to that the anti-PSD-95 IgG targets a common epitope in the homologous regions [39].

In control retinas, PKCα staining revealed fine, bushy dendritic processes of rod bipolar cells only in the OPL (Figure 3A), while rod bipolar cells in LIRD retinas exhibited dramatic neuritogenesis occurring by pLX7 in the survivor zone (data not shown) and becoming more obvious by pLX30. Rod bipolar cells in the LIRD retinas extended their dendrites from the OPL into outer nuclear layer (ONL) (Figure 3B and 3B', highlighted by arrows) in the survivor zone. Further, calbindin immunostaining revealed calbindin-positive horizontal cells undergoing similar neuritogenesis. While neurites of horizontal cells were confined in the OPL in the control retina (Figure 3C), horizontal cells in the LIRD retinas extended their neurites from the OPL into the ONL in the survivor zone (Figure 3D and 3D', highlighted by arrows). These findings are consitent with the formation of novel neurites found in human RP, AMD and animal models of retinal degenerations [5, 6, 13, 41‐43].

We correlated the findings of neuritogenesis here with alterations in GluR2 trafficking. Examination of expression patterns of GluR2 and its trafficking proteins in new neurites demonstrated that GluR2, PSD-95, stargazin, PICK1, GRIP1 and ABP were present in the OPL as well as in the ONL during LIRD (Figure 4A-F). GluR2 partially co-localized with PKCα in the new neurites (Figure 4A), suggesting that GluR2 is expressed in the new neurites of rod bipolar cells during retinal degenerations. Different from PKCα, PSD-95 co-localized well with GluR2 in the new neurites in the ONL (Figure 4B). GluR2 also co-localized well with PICK1 (Figure 4C), GRIP1 (Figure 4D), stargazin (Figure 4E), and ABP (Figure 4F) in the new neurites in the ONL. These data are consistent with the data proposing PSD-95, stargazin, PICK1, GRIP1 and ABP as anchoring of GluR2 at synaptic locations [44].

KIF3A is a reported motor protein expressed at the basal body of the connecting cilium axoneme and at the synaptic ribbons of cones and rods [45]. Our data showed that KIF3A was expressed in the OPL and IPL of control retina (Figure 5A), thereby raising the possibility that KIF3A interacts with postsynaptic density proteins which are also present in the OPL and IPL. Protein expression of KIF3A during LIRD was relatively stable, with a slight decrease on pLX7 (Figure 5B). Interactions between KIF3A and GRIP1 and PSD-95 were demonstrated by immunoprecipitation analysis (Figure 5C). During LIRD, KIF3A was also expressed in the new neurites, where KIF3A co-localized well with PSD-95 (Figure 5D) and GRIP1 (Figure 5E), suggesting that KIF3A may be involved in guiding GluR2 and its trafficking protein to the new dendrites.

In contrast to the survivor zone, GluR2 expression disappeared by pLX30 in the light-damage zone, where only one layer of photoreceptors was left (Figure 6A). The expression of PSD-95 was also dramatically decreased in the light-damage zone (Figure 6B). Similar to GluR2, its trafficking proteins GRIP1, ABP, PICK1, and stargazin disappeared in the light-damage zone (data not shown). In parallel with the loss of GluR2 and its trafficking proteins, dendrites of rod bipolar cells and neurites of horizontal cells in the light-damage zone OPL were retracted and flattened (Figure 6C and 6D). We also observed a horizontal cell body located adjacent to the RPE at the distal border of the remnant ONL (Figure 6D, arrow).

The major subunits of AMPA receptors, GluR1 and GluR2, play have important roles in neuroplasticity [44]. Given that LIRD increased the levels of GluR2, we examined the effect of LIRD on GluR1. Phosphorylation of GluR1 is a major mechanism for controlling excitation. Since our light damage protocol spans the period of photoreceptor stress (increased extracellular glutamate) all the way to complete photoreceptor loss, we deemed it appropriate to screen events known in CNS to modulate excitation. Further, GluR2 binds to PICK1 after phosphorylation at serine 880 by PKCα, which induces dissociation of GluR2 from GRIP and the subsequent internalization of GluR2 by PICK1 for recycling or degradation [38]. Thus, we also measured the levels of phosphorylated GluR1 (Ser831) (pGluR1) and phosphorylated GluR2 (Ser880) (pGluR2). Our results revealed that light stress had no significant effect on the protein levels of GluR1 (Figure 7A and 7B). However, light stress rapidly decreased the levels of pGluR1, which showed an obvious decrease on pLX0 and reached a minimum on pLX7 but recovered by pLX30. Light stress also decreased the expression of pGluR2 by pLX0, extending its decrease by pLX30. Two calcium binding proteins, calretinin and calbindin, were also examined but showed no LIRD-induced change in protein expression (Figure 7A and 7B).

Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) senses cytosolic Ca²⁺ fluxes, which are largely mediated by glutamate-activated AMPA or NMDA receptors in CNS neurons [46]. CaMKII plays a central role in synaptic plasticity, both α- and β-CaMKII are neuron-specific and expressed in retinal neurons [47]. αCaMKII is activated by high Ca²⁺ levels, while βCaMKII is more sensitive to lower Ca²⁺ levels, and heteromeric CaMKII Ca²⁺ responsivity and activity depends on the α/β subunit ratio [48]. Decreases in neuronal activity decrease the CaMKII α/β ratio, likely by upregulating transcription of βCaMKII [49], which positively correlates with the increase in low calcium permeability GluR2 expression in our study. Our previous study revealed that CaMKII plays an important role in neuritogenesis during retinal degeneration [8]. Therefore, we studied the effects of AMPA receptors antagonist NBQX on CaMKII signaling and neuritogenesis. Our results showed that NBQX slightly mitigated neuritogenesis. Consistently, NBQX had no effect on the ratio of α/βCaMKII compared with vehicle group during LIRD (Figure 8A, B, and 8D-F). However, NMDA receptors antagonist D-AP5 obviously increased the ratio of α/βCaMKII and significantly accelerated neuritogenesis (Figure 8A, C and 8D-F). Because NBQX is neuroprotective [50] but does not block downregulation of GluR2 [51], our results suggest that NBQX affords neuroprotection by a direct block of GluR2-lacking, Ca²⁺-permeable AMPA receptors [31]. Due to that GluR1 levels were not altered during LIRD, the above results totally suggest that AMPA receptor GluR2 in particular drives CaMKII signaling associated with neuritogenesis during LIRD.

Age-matched female Balb/C albino mice (The Jackson Laboratory, Bar Harbor, ME) were maintained in dim light (20-40 lux) on a 12-12 cycle in normal phase (lights on 7:00-19:00) with ad libitum access to food and water. All experimental procedures were designed to minimize animal number and suffering, and were conducted with approval of the Institutional Animal Care Committees at the University of Utah in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

LIRD and subretinal injections were performed as described previously [8]. In brief, mice were placed into the light-damage chamber, where the visible light intensity ranged from 2,500 to 3,000 lux, for 24 h by excluding one normal night cycle. Following post-light exposure (pLX), all animals were returned to the dim cyclic light environment and maintained for 0, 1, 7, and 30 days (pLX0, pLX1, pLX7, and pLX30). Under ketamine-xylazine anesthesia, 5 μM NBQX (Sigma, St. Louis, MO, USA) and 50 μM D-AP5 (Sigma) were injected into the sub-retinal space (central region) of albino mice in 0.5 μl volumes using a 33-gauge micro syringe (Hamilton) 30 min before light exposure. A successful sub-retinal injection caused swelling of the retina. Control retinas were injected with vehicle (DMSO) (Sigma).

At the end of the experiment, eyes were rapidly harvested following decapitation under isoflurane anesthesia. For Western blotting analysis, retinas were dissected in Hank's balanced salt solution (HBSS, GIBCO, Carlsbad, CA, USA), then placed on ice in 1.5 ml tubes with RIPA buffer (RIPA lysis buffer kit, Santa Cruz, CA, USA) and homogenized using a sonicator (Fisher Scientific, Pittsburgh, PA, USA) three times for 12 sec each with 12 sec breaks between cycles. The samples were put on ice for 10 min, then centrifuged at 14,000 rpm at 4°C for 10 min. Supernatants were transferred to new tubes. The concentrations of protein in the samples were measured using the BCA assay (Pierce, Rockford, IL USA). For immunohistochemical analysis, whole eyes were removed and rinsed in HBSS, fixed in 4% paraformaldehyde (PFA) (Sigma) for 2 h at 4°C, and then washed with PBS (in g/L: NaCl 8, KCl 0.2, Na₂HPO₄ 1.44, KH₂PO₄ 0.24; pH 7.4) twice for 10 min each. Eyes were incubated in PBS with 20% sucrose for 4 h at 4°C. Serial 12 μm coronal sections were made with a cryostat microtome (Leica CM3050 S, Wetzlar, Germany) and collected on Superfrost/plus microscope slides (Fisher Scientific).

Retinas were lysed in RIPA buffer (RIPA lysis buffer kit, Santa Cruz) for 1 h at 4°C with gentle agitation. Lysates were immunoprecipitated for 1 h at 4°C using KIF3A antibody (1:1000) and then protein G agarose beads (Roche, Indianapolis, IN, USA) [8]. Samples were analyzed by Western blotting.

Samples of entire group were pooled. Protein samples (~20 μg protein) were combined with NuPAGE LDS sample buffer (4×) (Invitrogen, Carlsbad, CA, USA) and NuPAGE sample reducing agent (10×) (Invitrogen), then boiled for 10 min at 95°C. Western blotting analysis was carried out using NuPAGE 4-12% Bis-Tris gels (Invitrogen) at 200 V for 40 min. Gels were electro-blotted onto PVDF membrane (Millipore, Bedford, MA, USA) for 1 h at 25 V using a wet electro-blotting system (XCell SureLock Mini-Cell, Invitrogen). Blots were blocked for 1 h in PBS with 0.1% Triton-x100, pH 7.4 (PBST) containing 5% non-fat dry milk (NFDM). Blots then were incubated overnight at 4°C in primary antibodies diluted in 5% NFDM-PBST solution (GluR2, 1:1000, Chemicon; pGluR2, 1:1000, Millipore; protein kinase C α (PKCα), 1:5000, Sigma; GluR1, 1:2000, Upstate; pGluR1, 1:1000, Millipore; PSD-95, 1:500, Chemicon; protein interacting with C kinase 1 (PICK1), 1:400, Santa Cruz; GRIP1, 1:1000, Millipore; stargazin, 1:500, Chemicon; AMPA receptor binding protein (ABP), 1:400, Santa Cruz; KIF3A, 1:1000, Covance; calretinin, 1:2000, Abcam; calbindin, 1:2000, Abcam; αCaMKII 1:1000, Upstate; βCaMKII: 1:1000, Invitrogen). Blots were washed three times for 10 min in PBST, incubated for 2 h in secondary antibodies (IgG-HRP, Santa Cruz; 1:5000 in 5% NFDM-PBST) followed by three more washes of 10 min in PBST. Immunostaining was revealed by the SuperSignal West Dura Extended Duration Substrate kit (Thermo Scientific, Waltham, MA, USA), and scanned using the Quantity One imaging system (Bio-Rad). Densitometry for each band was measured using ImageJ (U.S. National Institutes of Health, Bethesda, MD, USA). β-actin (1:5000, Sigma) was used as a loading control.

Cryosections were washed 2 × 10 min in PBS, then blocked with blocking buffer for 30 min. Cryosections were incubated overnight at 4°C in primary antibodies diluted in blocking buffer (GluR2, 1:500, Chemicon or 1:200, Santa Cruz; PKCα, 1:2000; PSD-95, 1:500; PICK1, 1:200; GRIP1, 1:500; stargazin, 1:500; ABP, 1:100; KIF3A, 1:500; calbindin, 1:1000). These primary antibodies were from different species in every case and they did not cross-react on Western Blots. After washing 3 × 10 min in PBS, sections were incubated for 45 min at room temperature in secondary antibodies (cy3-, 488- or 647-conjugated IgG, Invitrogen) diluted 1:1000 in blocking buffer. For double or triple staining, sections were sequentially incubated with primary antibodies and secondary antibodies as above. After incubation with antibodies, sections were washed 3 × 10 min in PBS, then treated with 10 μM DAPI (Invitrogen) for 5 min at room temperature. Sections from all groups were processed simultaneously to reduce staining artifacts or intensity differences. Negative controls were performed by omission of the primary antibodies.

Fluorescent images were acquired with an Olympus FV1000 laser-scanning confocal microscope (Olympus, Tokyo, Japan). Settings were chosen so that pixel intensities for the brightest samples were just below saturation, except when the processes of retinal neurons had to be clearly determined, in which case signals from certain areas (soma of the retinal neurons) were saturated in order to obtain a clear perimeter of the neurites. Optical slice units were 0.5 μm. Neuritogenesis was analyzed as described [8]. In brief, neurite length was measured from the point of emergence at the cell body to the tip of each segment. One section was selected from each animal, and 20 longest dendrites from 20 rod bipolar cells in a specific region (ventral mid-peripheral region, 120 μm in length) were measured and calculated as the mean. Quantification of the morphological parameters was carried out using ImageJ by investigators blinded to experimental conditions.

Data of neuritogenesis were expressed as mean ± SD and analyzed with SPSS 12.0 (SPSS Inc.). Statistical comparisons were made using Bonferroni tests and analysis of variance (ANOVA), P < 0.05 was defined as the level of significance. Protein levels represented pooled data of entire groups and were expressed as means only. To study the rationality of pooling samples in our experiment, we actually set up 3 batches of animals. There were 3 mice per group in each batch (i.e. 9 mice in total for each group). We pooled the 3 mice's retinas from each group in each batch and ran the western blotting for protein analysis. It was found that the biological variations for the 9 mice were low and GluR2 increase, for example, was significant during early LIRD (Additional file 1 Figure S1).