Address correspondence and reprint requests to Dr Nigel L. Barnett, Vision, Touch & Hearing Research Centre, The University of Queensland, Brisbane 4072, Australia. E-mail: email@example.com
Neuronal and glial high-affinity transporters regulate extracellular glutamate concentration, thereby terminating synaptic transmission and preventing neuronal excitotoxicity. Glutamate transporter activity has been shown to be modulated by protein kinase C (PKC) in cell culture. This is the first study to demonstrate such modulation in situ, by following the fate of the non-metabolisable glutamate transporter substrate, d-aspartate. In the rat retina, pan-isoform PKC inhibition with chelerythrine suppressed glutamate uptake by GLAST (glutamate/aspartate transporter), the dominant excitatory amino acid transporter localized to the glial Müller cells. This effect was mimicked by rottlerin but not by Gö6976, suggesting the involvement of the PKCδ isoform, but not PKCα, β or γ. Western blotting and immunohistochemical labeling revealed that the suppression of glutamate transport was not due to a change in transporter expression. Inhibition of PKCδ selectively suppressed GLAST but not neuronal glutamate transporter activity. These data suggest that the targeting of specific glutamate transporters with isoform-specific modulators of PKC activity may have significant implications for the understanding of neurodegenerative conditions arising from compromised glutamate homeostasis, e.g. glaucoma and amyotrophic lateral sclerosis.
The regulation of extracellular glutamate levels within the central nervous system, including the retina, under physiological and pathophysiological conditions is a prerequisite for the prevention of excitotoxic neurodegeneration. Under normal conditions, the glial and neuronal high-affinity glutamate uptake systems (excitatory amino acid transporters, EAATs) allow the rapid removal of glutamate from the extracellular space, thus terminating the excitatory signal and reducing the possibility of excitotoxic neuronal damage. Failure or reversal of these transport systems leads to an elevation of extracellular glutamate and contributes to the development of ischaemic pathologies, e.g. amyotrophic lateral sclerosis and glaucoma (Rothstein et al. 1993; Dreyer et al. 1996; Osborne et al. 1999; Jabaudon et al. 2000; Rossi et al. 2000).
Dark Agouti rats (200–250 g) were used for all experiments. All animals were treated in accordance with the ethical guidelines of the Australian National Health & Medical Research Council and The University of Queensland.
Anti-d-aspartate, glutamate and GLAST antisera were raised by Dr David Pow, The University of Queensland, Australia. Anti-PKCα and PKCδ antisera were obtained from Santa Cruz Biotechnology, CA, USA. Anti-rabbit antiserum and streptavidin-biotinylated horseradish peroxidase complex were purchased from Amersham, NSW, Australia. All agents for electrophoresis were purchased from Invitrogen, Carlsbad, CA, USA. Epoxy resin (Araldite) for tissue embedding was bought from ProSciTech, Queensland, Australia. The inhibitor of calcium-dependent PKCs (α, β and γ) Gö6976, was supplied by Calbiochem, Darmstadt, Germany. The modulators of PKC activity, phorbol 12-myristate 13-acetate (PMA), 4α-phorbol-12,13-didecanoate, chelerythrine chloride and rottlerin, and all other reagents, were purchased from Sigma, NSW, Australia.
D-Aspartate uptake and immunohistochemistry
Rats were administered an overdose of sodium pentobarbital (200 mg/kg i.p) and the eyes were immediately enucleated. Isolated retinas were preincubated in vitro at 37°C in oxygenated Ames medium containing one of the following PKC modulators: the PKC activator, phorbol 12-myristate 13-acetate (PMA) 1 µm; the non-isoform-specific inhibitor, chelerythrine chloride 25 µm; the PKCδ-specific inhibitor, rottlerin 20 µm; or Gö6976 1 µm, an inhibitor of the classical calcium-dependent PKCs α, β and γ at this concentration. Control retinas were incubated with the inactive phorbol ester, 4α-phorbol-12,13-didecanoate (1 µm). After 45 min, d-aspartate (50 µm), a non-endogenous substrate of glutamate transporters that can be detected by immunohistochemistry (Gundersen et al. 1993), was added to the medium. Incubation with d-aspartate continued for a further 60 min. At least four retinas were exposed to each treatment regime. The retinas were then fixed with 2.5% glutaraldehyde in 0.1 m phosphate buffer (pH 7.2) for 16 h, dehydrated, embedded in epoxy resin and cut into semi-thin sections (0.5 µm). These sections were immunolabeled with specific antibodies for d-aspartate (1 : 30 000 dilution) (Pow and Barnett 1999) or glutamate (1 : 100 000) using standard techniques with previously characterized antisera (Pow and Crook 1993). They were then observed with a Zeiss Axioskop microscope and photographed with a Spot RT digital camera.
Semi-thin sections of retina obtained approximately 2 mm superior to the optic disk were stained with toluidine blue, viewed on a Zeiss Axioskop microscope and digitally imaged. The thickness of the inner plexiform layer, the inner nuclear layer, the outer nuclear layer and the total retinal thickness were analysed using Adobe Photoshop software. The retinas of four rats were measured and an unpaired t-test used to compare values obtained from PKC modulator-treated eyes with those from control eyes.
Following PKC modulation, isolated retinas were immersion fixed for 90 min in 4% paraformaldehyde in 0.1 m phosphate buffer, pH 7.4. The fixed retinas were embedded in 4% Agar and 40 µm-thick transverse sections cut on a Vibratome. Free-floating sections were labeled with antiserum raised in rabbit against the N-terminus of GLAST (Pow and Barnett 1999), diluted 1 : 100 000 in phosphate-buffered saline containing 0.2% Triton X-100 plus 1% bovine serum albumin. Labeling was detected using biotinylated donkey anti-rabbit antiserum diluted 1 : 300, followed by streptavidin-biotinylated horseradish peroxidase complex diluted 1 : 300. Diaminobenzidine was used as the chromogen.
Retinas were homogenized (400 µL/retina) in ice-cold buffer (500 mm Tris-HCl, 8% lithium lauryl sulfate, 0.2 mm 4-(2-aminoethyl)benzenesulfonyl fluoride hydrochloride (AEBSF), pH 8.5), centrifuged briefly and the pellet discarded. Following protein concentration determination, the samples were adjusted to 0.35 mg protein/mL with NuPAGE LDS Sample Buffer. Fifteen microliter retinal samples were separated by electrophoresis on a 4–12% Bis-Tris gel and the proteins transferred to a nitrocellulose membrane. The blots were probed with antibodies to PKCα (1 : 30 000), PKCδ (1 : 5000) or GLAST (1 : 20 000). The bands were detected using biotinylated donkey anti-rabbit antiserum diluted 1 : 2000, followed by streptavidin-biotinylated horseradish peroxidase complex diluted 1 : 1000. Diaminobenzidine was used as the chromogen. Under these conditions the GLAST antibody labeled a number of bands on the blot. However, preabsorbtion of the GLAST antibody with GLAST protein (10 µg/mL) for 1 h resulted in the loss of a single band of approximately 70 kDa (see Fig. 4). The blots were scanned and the labeling intensity of each band was analysed with NIH Image software. A paired t-test was used to compare the intensity of each band obtained from control retinas (n = 3) with respective bands obtained from retinas that had been incubated with chelerytherine (25 µm) (n = 3) for 45 min.
D-Aspartate uptake and immunohistochemistry
In the rat retina, exogenous d-aspartate is normally accumulated by the radial glial Müller cells. The photoreceptor inner segments also appear to be d-aspartate immunoreactive (Fig. 1a). Incubation with the inactive phorbol ester, 4α-phorbol-12,13-didecanoate (1 µm) does not affect the pattern of d-aspartate uptake. The inclusion of the PKC activator, phorbol 12-myristate 13-acetate (PMA, 1 µm) does not appear to alter the transport of the glutamate analog as reflected by d-aspartate immunoreactivity (Fig. 1b). However, the inhibition of PKC with the non-isoform-specific inhibitor, chelerythrine (25 µm), dramatically changes the distribution of d-aspartate uptake in the retina. d-Aspartate is exclusively accumulated by a population of rod and cone bipolar cells (Fig. 1c). There is no evidence of d-aspartate immunoreactivity in the Müller cells, suggesting that glial cell transport has been suppressed by the inhibition of PKC. Incubation of the retinas with rottlerin, a PKCδ isoform-specific inhibitor, elicits similar effects upon d-aspartate transport as chelerythrine. d-Aspartate immunoreactivity is predominantly localized to bipolar cells but weak labeling is also present in photoreceptors (Fig. 1d). There is no evidence of d-aspartate uptake by Müller cells following rottlerin treatment. Gö6976 (1 µm), a selective inhibitor of the calcium-dependent PKC isoforms α, β and γ at this concentration, does not affect the distribution of d-aspartate accumulation as compared with control incubated retinas (Fig. 1e).
GLAST immunoreactivity is present in the glial elements of the retina, particularly the Müller cells. Three distinct bands of GLAST immunolabeling are apparent in the inner plexiform layer, as well as intense labeling of the outer plexiform layer, corresponding to the lateral processes of the Müller cells. Furthermore, Müller cell processes ramify throughout the inner nuclear layer and ensheathe the bipolar cell perikarya. GLAST immunopositive radial Müller cell processes are seen throughout the outer nuclear layer (Fig. 2a). The stimulation of retinal PKC with PMA (1 µm) for 45 min does not alter the distribution or the expression of the GLAST glutamate transporter (Fig. 2b). Incubation with the non-isoform-specific PKC inhibitor chelerythrine (25 µm) (Fig. 2c), the PKCδ isoform-specific inhibitor rottlerin (20 µm) or the PKCα, β and γ isoform-selective inhibitor Gö6976 (1 µm) does not significantly alter the distribution or expression of GLAST in the rat retina.
In the normal retina, glutamate immunoreactivity is observed in bipolar cells and their processes in the inner plexiform layer, as well as the inner segments of the photoreceptors. The somata of ganglion cells are also strongly labeled for glutamate. No labeling is observed in the processes or somata of Müller cells (Fig. 3a). The distribution and intensity of glutamate labeling in the retina is not affected by a 45-min incubation with either the inactive phorbol ester, 4α-phorbol-12,13-didecanoate (1 µm), or the PKC activator PMA (1 µm) (Fig. 3b). In contrast, the inhibition of PKC with chelerythrine causes a significant change in the pattern of labeling. Glutamate immunoreactivity is lost from the photoreceptors and the ganglion cells but intense labeling is still present in a population of bipolar cells (Fig. 3c). The specific inhibition of PKCδ with rottlerin causes a similar change in the distribution of glutamate. In addition, weak labeling remains in the photoreceptors and ganglion cells (Fig. 3d).
The in vitro inhibition of retinal PKC with chelerythrine (25 µm) for 45 min does not alter significantly (p > 0.05) the total retinal content of PKCα, PKCδ or GLAST as assessed by western blotting (Fig. 4).
Morphometric analysis of isolated retinas treated with PKC modulators are summarised in Table 1. Statistical analysis was performed with an unpaired t-test to compare treated retinal values to corresponding control retinal values. There is no significant change in total retinal thickness following treatment with any of the PKC modulators. However, treatment with chelerythrine for 45 min results in significant (p = 0.04) swelling of the inner plexiform layer (IPL) and a significant (p = 0.05) thinning of the inner nuclear layer (INL). The thinning of the INL compared with control retina is indicative of neuronal cell death as seen in toluidene blue stained semi-thin retinal sections (Fig. 5a, control; Fig. 5c, chelerythrine). Inhibition of retinal PKC by rottlerin does not result in a statistically significant change in retinal thickness (Table 1) but some histological damage is evident (Fig. 5d). Swelling of Müller cell processes is observed within the outer nuclear layer (ONL) in addition to dying cells in the INL. The rottlerin induced histological damage is not as extensive as that caused by chelerythrine. Neither the activation of PKC by PMA nor the inhibition of PKC α, β and γ by Gö6976 affects retinal histology (Figs 5b and e, respectively).
Table 1. Morphometric analysis of retinal layer thickness following in vitro incubation with isoform-specific modulators of protein kinase C activity
PMA (1 µm)
Chelerythrine (25 µm)
Rottlerin (20 µm)
Gö6976 (1 µm)
Values are means ± SEM; *statistically significant (p ≤ 0.05); NS, not statistically significant (p >0.05); PMA, phorbol 12-myristate 13-acetate; IPL, inner plexiform layer; INL, inner nuclear layer; ONL, outer nuclear layer; Total, total retinal thickness.
This study made use of intact retinas to investigate how protein kinase C regulates glutamate transporter activity in situ. Previous cell culture investigations into the effects of PKC have revealed conflicting results. A study on C6 glioma cell-line cultures has demonstrated that EAAC-1 is activated (increased Vmax) by the phorbol ester PMA, an agent that stimulates PKC (Dowd and Robinson 1996). The activity of GLT-1 is reported to be stimulated in C6 cells by direct phosphorylation with PKC (Casado et al. 1993), no effect is reported by Tan et al. (1999) whereas an inhibition (increased Km) of glutamate uptake by GLT-1 is reported in Y79 human retinoblastoma cells (Ganel and Crosson 1998). Phorbol esters have been shown to inhibit GLAST activity in Xenopus oocytes (Conradt and Stoffel 1997) and to decrease the expression of the transporter protein in cultured chick retinal Müller cells (Gonzalez et al. 1999). Immunohistochemical techniques were employed to trace the uptake of d-aspartate (a non-metabolisable glutamate analog) following exposure of isolated retinas to modulators of PKC activity. The present results indicate that pan-isoform-specific PKC inhibition with chelerythrine decreases d-aspartate (and by inference, glutamate) uptake into Müller cells by GLAST, the dominant retinal transporter responsible for extracellular glutamate clearance (Rauen et al. 1998; Barnett and Pow 2000). This effect is mediated by PKCδ, as the specific inhibition with rottlerin of PKCδ the only reported PKC isoform present within Müller cells (Osborne et al. 1994), also blocks d-aspartate accumulation by Müller cells. Moreover, the inhibition of the neuronally localized PKCα or PKCβ isoforms (Wood et al. 1997) with Gö6976 does not suppress the GLAST mediated uptake of d-aspartate by Müller cells. Similarly, Lortet et al. (1999) have demonstrated that PKC inhibition decreases glutamate uptake by the high-affinity uptake transporter expressed by primary cortical neuronal cell cultures. Moreover, this inhibition is prevented by coexposure to the PKC activator, PMA. Conversely, it has been reported that PKC activation (as opposed to inhibition) suppresses glutamate uptake by GLAST in Xenopus oocytes and human embryonic kidney cells (Conradt and Stoffel 1997). The respective use of a transfected cell culture system and whole retinal tissue may explain the disparity between the reported effects of PKC modulation upon GLAST activity.
The suppression of GLAST activity in response to PKC inhibition is revealed by a change in the pattern of d-aspartate uptake within the retina (Fig. 1). When GLAST is inhibited, the neuronally localized glutamate transporters accumulate d-aspartate. The suppression of GLAST activity by the specific inhibition of PKCδ which is expressed by retinal Müller cells (Osborne et al. 1994), allows photoreceptors and bipolar neurones to accumulate d-aspartate. Bipolar cells express two glutamate transporters: EAAT5 is localized to rod bipolar cells (Pow and Barnett 2000) whereas cone bipolar cells express GLT-1 (Rauen and Kanner 1994; Rauen et al. 1996). Figure 1 clearly shows that both rod and cone bipolar cells accumulate d-aspartate when GLAST activity is suppressed by PKCδ inhibition with rottlerin or chelerythrine. The intensity of d-aspartate labeling of these cells following exposure to chelerythrine (Fig. 1c) suggests that the transporters expressed by these cells may either be activated by, or unaffected by, PKC inhibition. There are no previous reports of PKC inhibition enhancing glutamate uptake by high-affinity transporters.
Although rottlerin is reported to be specific for the delta isoform of PKC (Gschwendt et al. 1994; Keenan et al. 1997), the lack of observable d-aspartate uptake by ganglion or amacrine cells (Fig. 1d), which are known to express glutamate transporters (Rauen and Kanner 1994; Rauen et al. 1996; Arriza et al. 1997; Pow and Barnett 2000; for a review see Pow 2001), may indicate that rottlerin is inhibiting transporters other than GLAST. We have previously shown that when GLAST activity is specifically suppressed with antisense oligonucleotides, ganglion cells accumulate d-aspartate in addition to the observed uptake by bipolar cells (Barnett and Pow 2000). Davies et al. (2000) have recently reported that rottlerin interacts with protein kinases other than PKC. We must therefore consider that rottlerin may be having effects other than the specific suppression of GLAST activity mediated by an inhibition of PKCδ. Furthermore, PKC has multiple roles in cellular signaling in the CNS, including the retina (Wood et al. 1997; Goekjian and Jirousek 1999). The delta isoform of PKC is, however, the only PKC isoform expressed by retinal Müller cells (Osborne et al. 1994).
Immunohistochemical labeling (Fig. 2) and western blot analysis (Fig. 4) of the glutamate transporters within the retina demonstrates that the inhibition of glutamate uptake is not due to the rapid down-regulation or degradation of transporter protein. This suggests that the suppression of GLAST mediated glutamate transport is due to a conformational change of the transporter protein or intracellular sequestration signaled via the modulation of PKC activity. Since PKC inhibition suppresses Müller cell glutamate transport, it suggests that the de-phosphorylation of GLAST inhibits the transport of glutamate. This conclusion seems plausible considering that the phorbol ester induced stimulation of GLT-1 is due to an increase in phosphorylation and a corresponding conformational change (Casado et al. 1993). Dowd and Robinson (1996) also reported a rapid increase in glutamate transport by EAAC-1 in response to the phorbol ester PMA, which is attributed to an increase in the catalytic rate of the glutamate transporter and/or an increase in the number of membrane-bound transporters removing glutamate from the extracellular space. Davis et al. (1998) showed that the effect of PMA is to increase the cell surface expression of EAAC-1 in C6 glioma cells.
Since PKC inhibition suppresses d-aspartate uptake by Müller cells, this suggested to us that PKC activation may enhance extracellular glutamate clearance in the retina. Activation of PKC has been found to stimulate the activity of glutamate transporters in glial cell cultures (Casado et al. 1991), the glial fraction of the rat forebrain (Daniels and Vickroy 1999), HeLa cells expressing GLT-1 (Casado et al. 1993) and C6 glioma cells expressing EAAC-1 (Casado et al. 1993; Dowd and Robinson 1996; Davis et al. 1998). However, we observed no increase in Müller cell d-aspartate immunoreactivity following the activation of PKC with PMA. Quantitative uptake studies must be performed to confirm this immunohistochemical observation. In a similar study, Lortet et al. (1999) could not demonstrate an increase in glutamate transport by PKC activation in cortical neuronal cultures, despite a chelerythrine induced reduction of glutamate transport in the same cells. It is suggested that a high baseline level of phosphorylation of the transporters, or the presence of transporter regulatory proteins, prevents a further increase in activity upon stimulation by PKC (Lortet et al. 1999). There is still some uncertainty as to the mechanism by which PKC modulates the function of glutamate transporters. PKC may directly phosphorylate the transporter protein to change either the substrate binding affinity or the rate at which the substrate is transported into the cell. Casado et al. (1993) demonstrated modulation of GLT-1 activity by direct phosphorylation with PKC. Mutation experiments determined that PKC acts directly on a PKC consensus site within the protein. It has been reported that PKC also modulates GLAST activity by direct phosphorylation, although not at any of the putative PKC sites within the protein (Conradt and Stoffel 1997). This led to the conclusion that PKC is acting at a non-consensus site. Recent evidence suggests that glutamate transporter activity may also be influenced by intracellular regulatory proteins. Thus, PKC may indirectly affect glutamate transporter activity by modulating these regulatory proteins. Disruption of the interaction between GLAST and an intracellular regulatory protein increases the rate of glutamate transport (Marie and Attwell 1999). Similarly, EAAC-1 mediated glutamate transport is regulated by the intracellular protein GTRAP3-18 (Lin et al. 2001).
Inhibition of the Müller cell transporter, GLAST, with chelerythrine or rottlerin also disrupts the pattern of glutamate immunolabeling. Immunohistochemical labeling of retinal sections for glutamate closely resembles the pattern of d-aspartate uptake following PKC inhibition (Fig. 3). Glutamate immunoreactivity is significantly decreased in the ganglion cells and in neuronal processes within the inner plexiform layer. The reduction in retinal glutamate levels following PKC inhibition is probably due to a disruption of the glial-neuronal glutamate-glutamine cycle (Barnett et al. 2000) as a consequence of impaired glutamate uptake by the Müller cells. The glutamate immunoreactivity retained by the bipolar cells is likely to reflect the direct reuptake of synaptically released glutamate by the glutamatergic bipolar neurones. This idea is supported by the pattern of d-aspartate uptake.
The retinal high-affinity glutamate transporters maintain low extracellular glutamate levels, thereby modulating synaptic transmission and protecting neurones from glutamate excitotoxicity (Newman and Reichenbach 1996; Rauen et al. 1998). Rothstein et al. (1996) reported that loss of either GLAST or GLT-1 activity within the spinal cord, striatum and hippocampus produces elevated extracellular glutamate levels and excitotoxic neuronal death. In the retina, no neuronal damage was observed in a GLAST knock-out mouse model (Harada et al. 1998) or following the specific inhibition of GLAST activity with antisense oligonucleotides (Barnett and Pow 2000). However, ganglion cell damage has been reported following GLAST inhibition (Vorwerk et al. 2000). Whilst Müller cell uptake by GLAST dominates retinal glutamate transport (Rauen et al. 1998; Barnett and Pow 2000), it is clear that the neuronally localized glutamate transporters can provide some protection for neurones following GLAST inhibition. The PKCδ inhibitor, rottlerin, selectively inhibits GLAST and prevents d-aspartate uptake by Müller cells, but does allow uptake by bipolar and photoreceptor neurones. Histological damage to the retina is noticeably less than that observed in the chelerytherine- treated retinas (Table 1) in which d-aspartate is only transported into a population of bipolar cells (Fig. 1). These data suggest that the targeting of specific glutamate transporters with isoform-specific modulators of PKC activity may have implications for the management of neurodegenerative conditions arising from compromised glutamate homeostasis, e.g. glaucoma and amyotrophic lateral sclerosis (Rothstein et al. 1993; Dreyer et al. 1996; Osborne et al. 1999).
This study was funded by the National Health and Medical Research Council (Australia). We thank Dr David Pow, The University of Queensland, for the kind donation of antibodies and Ms. Rowan Tweedale for her careful reading of the manuscript.