Frederike Dijk

An immunocytochemical study on specific amacrine cell subpopulations in the rat retina after ischemia

Transient retinal ischemia leads to the loss of neurons in the inner retina. In an accompanying paper [F. Dijk, S. Van Leeuwen, W. Kamphuis, Differential effects of ischemia/reperfusion on amacrine cell subtype-specific transcript levels in the rat retina, Brain Res., 1026 (2004) 194-204] we present the results of a study on the effects of experimentally induced retinal ischemia on transcript levels of genes expressed by distinct subpopulations of amacrine cells. In response to 60-min ischemia, three different patterns of changes in transcript levels were found, indicating a differential vulnerability of amacrine subtypes: (i) a gradual decrease of transcript level without recovery (parvalbumin; PV); (ii) a gradual decrease, with varying rates and degrees, followed by partial recovery after 72 h of reperfusion (choline acetyltransferase (ChAT), calretinin (CR) and glycine transporter (Glyt1)); (iii) no significant changes (substance P (SP)). In order to verify whether the degree of cell loss can be predicted from the quantified alterations in gene expression level, immunocytochemical stainings were carried out. A 60-min ischemic period was administered to the rat eye by raising the intraocular pressure, followed by a reperfusion period lasting between 2 h and 4 weeks. Cryosections were immunostained for Glyt1, PV, ChAT, CR, and SP. Double-labelling with apoptosis marker TUNEL was used to demonstrate cell type-specific apoptosis. Following ischemia, the numbers of detected PV-, Glyt1, ChAT-, and CR-immunopositive somata showed a substantial, but differential, reduction at 1-4 weeks after ischemia. The total amount of immunoreactivity present in the inner plexiform layer (IPL) also decreased. The extent of alterations derived from immunocytochemical staining was greater than was anticipated from the decrease of transcript levels. Only for SP, no significant decrease in number of cells or in the intensity of immunoreactivity in IPL was observed, which is in agreement with the absence of significant changes in transcript levels. In conclusion, retinal ischemia/reperfusion differentially affects amacrine cell populations. Although both protein and mRNA levels are reduced, transcript levels are less attenuated. Caution must be applied in the use of real-time quantitative PCR (qPCR) screening as a tool to assess the cellular pattern of neurodegeneration in the retina.

Differential effects of ischemia/reperfusion on amacrine cell subtype-specific transcript levels in the rat retina

Transient retinal ischemia induces loss of retinal ganglion cells, supporting the hypothesis that ischemic conditions contribute to the induction and progression of glaucoma. However, after 60 min of ischemia, also amacrine cells are lost from the inner nuclear layer. The main goal was to determine the relative vulnerability of various amacrine subpopulations by measuring the levels of transcripts that are known to be specifically expressed by different amacrine subpopulations. A 60-min ischemic period was administered to the rat eye by raising the intraocular pressure, followed by a reperfusion period lasting between 2 h and 4 weeks. Total RNA was isolated from the whole retina and expression levels were assessed by real-time quantitative polymerase chain reaction (qPCR). Retinal ischemia/reperfusion has differential effects on the levels of the various transcripts. Three main patterns of changes were identified. (i) A gradual decrease of transcript level without recovery was observed for parvalbumin; this transcript is expressed by the glycinergic AII cells. (ii) A gradual reduction to different levels at 72 h of reperfusion followed by a partial or complete recovery (glycine transporter 1, glutamate decarboxylase, calretinin, and several other transcripts). The glycinergic amacrine cell markers recovered to 65-75% of the control level, while the main GABAergic markers had completely recovered at 4 weeks. (iii) No significant changes of transcript levels were found for markers of several smaller GABAergic subpopulations [including substance P (Tac1), somatostatin, and others]. Expression levels of photoreceptor-, horizontal cell-, and bipolar cell-specific transcripts were not altered. These patterns were confirmed by a cluster analysis of the data. Based on gene expression levels, it may be concluded that amacrine cells are vulnerable to ischemic insults and that the glycinergic amacrine cells are relatively more sensitive to ischemia than the GABAergic population. In particular, the extensive loss of the parvalbumin-containing AII amacrine cells, which serve in the rod pathway, may have functional implications for vision under scotopic conditions. In the accompanying paper [F. Dijk and W. Kamphuis, An immunocytochemical study on specific amacrine subpopulations in the rat retina after ischemia, Brain Res. (2004).], the results are evaluated at the protein level by immunostaining for a selection of the amacrine cell markers.

Expression of AMPA-type glutamate receptor subunit (GluR2) in ON-bipolar neurons in the rat retina.

The role of glutamate receptors (GluR) in the signal pathways of the retina is widely recognized. Photoreceptors make synaptic contact with functionally different classes of bipolar cells. The OFF‐type bipolar cells mediate light offset‐evoked responses and use ionotropic α‐amino‐3‐hydroxy‐5‐methyl‐isoxazole‐4‐propionate (AMPA)‐ or kainate‐type GluRs, whereas bipolars involved in the ON‐pathway use the metabotropic GluR6. This dichotomy predicts a defined expression pattern of AMPA‐type GluRs and mGluR6 in bipolar cell classes. This hypothesis was tested by performing immunocytochemical double labeling studies combining GluR‐specific antibodies with markers specific for the diverse bipolar cell populations in the rat retina. AMPA‐type receptors are composed of combinations of four types of subunits, GluR1–4. GluR1 is expressed by a few somata in the outer part of the inner nuclear layer (INL). Sparse colocalization with any of the bipolar markers used could be established. In contrast, GluR2 is expressed by many of the somata in the outer zone of the INL. At the transcript level, in situ hybridizations demonstrated abundant GluR2 expression over the complete width of the INL. In contrast to our expectations, approximately 70% of the somata labeled by the rod ON‐bipolar markers protein kinase C (PKC) or Goα, colocalized with GluR2. Approximately 90% of the OFF‐type bipolar cells, identified as recoverin‐positive, showed GluR2 immunoreactivity. At least 40% of the somata that were mGluR6‐immunoreactive, a both rod and cone ON‐type bipolar marker, were GluR2‐immunopositive. Ultrastructurally, examples were observed of GluR2 localization in bipolar processes with labeling outside the actual compartment associated with the synaptic complex of the rod terminal. No specific antibody was available against GluR3, but 74% of the PKC‐positive cells were GluR2/3‐positive. GluR4 did not show a somatic localization making double labeling impossible. On the basis of these results, we conclude that ionotropic GluRs are expressed by rod ON‐type bipolar cells (PKC‐ or Goα‐immunoreactive), and by cone ON‐ and OFF‐type bipolars based on a colocalization with nearly all of the present recoverin‐positive somata. Our observations show that the functional dichotomy in ON‐ and OFF‐type bipolars is not reflected in a matching expression pattern of ionotropic and metabotropic GluRs. This finding raises the intriguing possibility that the AMPA‐type GluRs are, in an as yet unclear manner, involved in the ON signaling pathways of rods and cones. J. Comp. Neurol. 455:172–186, 2003.

Ischemia-induced alterations of AMPA-type glutamate receptor subunit. Expression patterns in the rat retina--an immunocytochemical study.

This study investigates whether retinal ischemia/reperfusion leads to alterations in the expression of AMPA-type glutamate receptor (AMPAR) subunits GluR1-4. In ischemia-vulnerable hippocampal neurons, a subunit-specific downregulation of GluR2 precedes the actual neurodegeneration. Our purpose was to study whether retinal ischemia induces a similar downregulation of GluR2 preceding the loss of ganglion and amacrine cells. A 60-min ischemic period was followed by reperfusion lasting between 2 h and 7 days. Changes in the expression patterns of GluR1-4 were assessed using immunocytochemistry. In the same sections, alterations in cell density, thickness of retinal layers, and density of apoptotic cells were investigated. Two-hour post-ischemia, GluR1 immunoreactivity was nearly absent from the inner plexiform layer (IPL). Thereafter, labeling intensity recovered slowly and was close to control levels at 7 days, albeit in a thinner IPL. The decrease in GluR2/3 labeling intensity was most profound at 4 h. The recovery of GluR2/3 staining intensity was slow, and staining was still decreased at 7 days. GluR2 immunoreactivity was not attenuated after ischemia. GluR4 labeling showed a similar time course as observed for GluR1, but the decrease in immunoreactivity was less profound and the recovery was nearly complete. The immunostaining of PKCalpha, a rod bipolar cell marker, was unaffected at all reperfusion times. The reduction of GluR staining preceded both the typical thinning of the IPL and the peak of cell loss, but coincided with a significant swelling of the IPL. In conclusion, retinal ischemia/reperfusion leads to differential changes in the expression of the different AMPA-type GluR subunits, which may affect excitatory synaptic transmission in the inner retina. However, no evidence was found for a preferential loss of GluR2 immunoreactivity that could account for selective neurodegeneration of amacrine and ganglion cells after retinal ischemia.

Gene expression of AMPA-type glutamate receptor subunits in rod-type ON bipolar cells of rat retina.

The retinal rod bipolar cell type is involved in the sign‐inverting depolarizing ON‐type response to light. This response is mediated by the metabotropic glutamate receptor type 6 (mGluR6) expressed on the rod bipolar dendrites. In a previous immunocytochemical study, an unexpected colocalization was reported [W. Kamphuis et al. (2003) J. Comp. Neurol., 455, 172–186] of mGluR6 with the ionotropic AMPA‐type glutamate receptor subunit GluR2 in rod bipolar cells of rat retina. The aim of the present study was to investigate whether expression of both genes could be found at the single‐cell level. Two approaches were followed. (i) Retinal cells were isolated by enzymatic and mechanical treatment. Single cells with a bipolar morphology were harvested, subjected to multiplex PCR with protein kinase C (PKC)‐, mGluR6‐ and GluR1–4‐specific primers, followed by a real‐time quantitative PCR assay. Of 23 studied cells, 74% expressed PKC and 87% expressed mGluR6. Using the presence of both transcripts as the criterion for a rod bipolar cell signature (nu2003=u200315), 73% of these cells expressed GluR2, with a minor contribution of GluR1 (20%), GluR3 (7%), and GluR4 (20%). Quantification of the transcript levels demonstrated that mGluR6 and GluR2 genes are expressed at similar levels in rod ON‐type bipolar cells. (ii) Rod bipolar cells were identified in retinal sections by immunolabelling with a protein kinase C antibody and isolated using laser pressure catapulting (LPC). Quantitative PCR was employed to assess gene expression levels of reference genes, PKCα, mGluR6 and the GluR subunits. However, in samples from PKCα‐immunopositive somata no significant enrichment of PKCα transcript levels was observed when compared with control samples from immunonegative somata. We conclude that this approach lacks sufficient spatial specificity. In conclusion, the results show coexpression of mGluR6 and GluR2 in rod bipolar cells; this is in good agreement with the results of previous immunocytochemical studies. The functional implications of AMPA‐type glutamate receptors for ON‐type rod bipolar‐mediated signal transduction remains to be elucidated.