Brimonidine prevents axonal and somatic degeneration of retinal ganglion cell neurons
© Lambert et al; licensee BioMed Central Ltd. 2011
Received: 26 October 2010
Accepted: 13 January 2011
Published: 13 January 2011
Skip to main content
© Lambert et al; licensee BioMed Central Ltd. 2011
Received: 26 October 2010
Accepted: 13 January 2011
Published: 13 January 2011
Brimonidine is a common drug for lowering ocular pressure and may directly protect retinal ganglion cells in glaucoma. The disease involves early loss of retinal ganglion cell transport to brain targets followed by axonal and somatic degeneration. We examined whether brimonidine preserves ganglion cell axonal transport and abates degeneration in rats with elevated ocular pressure induced by laser cauterization of the episcleral veins.
Ocular pressure was elevated unilaterally by 90% for a period of 8 weeks post- cauterization. During this time, brimonidine (1mg/kg/day) or vehicle (phosphate-buffered saline) was delivered systemically and continuously via subcutaneous pump. Animals received bilateral intravitreal injections of fluorescent cholera toxin subunit β (CTB) two days before sacrifice to assess anterograde transport. In retinas from the vehicle group, elevated pressure induced a 44% decrease in the fraction of ganglion cells with intact uptake of CTB and a 14-42% reduction in the number of immuno-labelled ganglion cell bodies, with the worst loss occurring nasally. Elevated pressure also caused a 33% loss of ganglion cell axons in vehicle optic nerves and a 70% decrease in CTB transport to the superior colliculus. Each of these components of ganglion cell degeneration was either prevented or significantly reduced in the brimonidine treatment group.
Continuous and systemic treatment with brimonidine by subcutaneous injection significantly improved retinal ganglion cell survival with exposure to elevated ocular pressure. This effect was most striking in the nasal region of the retina. Brimonidine treatment also preserved ganglion cell axon morphology, sampling density and total number in the optic nerve with elevated pressure. Consistent with improved outcome in the optic projection, brimonidine also significantly reduced the deficits in axonal transport to the superior colliculus associated with elevated ocular pressure. As transport deficits to and from retinal ganglion cell projection targets in the brain are relevant to the progression of glaucoma, the ability of brimonidine to preserve optic nerve axons and active transport suggests its neuroprotective effects are relevant not only at the cell body, but throughout the entire optic projection.
Glaucoma is a chronic disease that causes vision loss through the degeneration of retinal ganglion cell (RGC) neurons and their axons in the optic nerve [1, 2]. While age is an important risk factor, the only modifiable risk factor and sole target for clinical intervention is elevated intraocular pressure (IOP) or ocular hypertension (OHT) . Lowering IOP generally slows progression in glaucoma , but does not necessarily stop degeneration . Thus, there is great interest in identifying neuroprotective agents as potential therapies [5, 6].
Brimonidine (BMD; UK14304, Alphagan) is a non-selective α2-adrenergic receptor agonist currently used as a treatment to lower IOP in glaucoma. An effect of BMD is to decrease aqueous humor production by inhibition of adenylate cyclase inhibition, which lowers cAMP levels . The drug also increases uveoscleral outflow due to prostaglandin release and/or ciliary muscle relaxation . Independent of its IOP-lowering properties, BMD is neuroprotective for RGCs in various injury models [9–13]. Potential mechanisms underlying these effects include the inhibition of glutamate release, regulation of calcium influx in the inner retina, modulation of NMDA receptor signalling in RGCs, and upregulation of trophic factor expression [14–16]. Because of its dual action to lower IOP and perhaps protect against neuronal injury, BMD may hold promise in the treatment of glaucoma and other optic neuropathies .
Following acute ischemic injury, BMD preserves axonal transport along RGC axons from the retina to the superior colliculus . This is intriguing in the context of glaucoma, as recent evidence indicates axon-specific mechanisms play an early role in the pathology of the disease [19–21]. Here we investigated whether systemic treatment with BMD by continuous subcutaneous injection improves axon survival in the optic nerve and anterograde transport to the colliculus with acute OHT.
For all retinal quadrants, the density of SMI31+ RGCs in the control retinas peaked within 1-2 mm of the optic nerve head (3,400 to 5,400 RGCs/mm2) and decreased slightly (17 to 18%) at 4 mm out (Figure 4). In the superior retina of the vehicle group, OHT decreased RGC density by 15 to 17% across eccentricity compared to control eyes. RGC density in the same sector for the BMD OHT retinas was 14 - 24% higher than the vehicle OHT retinas; neither group was significant different than the corresponding control eye (p ≥ 0.2). In the inferior quadrant of vehicle retinas, OHT decreased RGC density 14 - 25% compared to control retinas with significant reductions of 25% and 23% observed at 1 and 2mm eccentric, respectively (p ≤ 0.05). For the BMD group in this quadrant, RGC density did not change with OHT compared to control (p ≥ 0.25). In the temporal quadrant for vehicle retinas, OHT decreased RGC density 14 - 27% compared to control; at 4mm eccentricity this decrease (27%) was significant (p = 0.04). For the BMD group in this quadrant, RGC density in OHT retinas was similar to control retinas at all eccentricities (p ≥ 0.2). The 31% improvement in RGC density compared to vehicle OHT retinas was significant at 1 mm from the optic disc (p = 0.05). Finally, RGC loss in vehicle OHT eyes was most dramatic in the nasal retina, where RGC density decreased at every location. Decreased density compared to control retinas ranged from 42% nearest the optic disc to 26% at 4 mm eccentricity (p ≤ 0.02). BMD treatment afforded partial rescue of RGCs in this quadrant, with significant increases of 30-42% at 1-3 mm eccentricity compared to vehicle OHT retinas (p ≤ 0.005). However, only at 1 mm was RGC density in the BMD OHT group comparable to the control group (p = 0.09).
In this study, we quantified how systemic delivery of BMD via subcutaneous osmotic pump protected RGCs challenged by elevated IOP in a rat model of OHT. In the absence of treatment, OHT for 8 weeks induced pathology at several levels. In the vehicle group, the worst OHT-related outcome was 70% depletion in transport of CTB to the superior colliculus (Figure 6). This deficit cannot be accounted for entirely by loss of active uptake of CTB by RGCs in the retina, which was reduced by only 44% (Figure 3). Depletion of transport was followed by a 33% loss of axons in the optic nerve (Figure 5) and finally by an average decrease in RGC somatic density of about 16% as measured by SMI31 labelling (Figure 4). In the retina, OHT was also associated with severely dystrophic axons, swollen RGC somas, and the accumulation of phosphorylated neurofilaments in dendrites (Figure 2).
These results support an OHT-induced progression in which a functional deficit precedes structural pathology. Loss of active uptake and transport of CTB was far worse than axon loss in the optic nerve, which in turn was worse than RGC somatic drop-out in the retina. This progression is similar to that described in the DBA2J mouse model of glaucoma [20, 21, 23]. Interestingly, RGC somatic loss as measured by SMI31 labelling was not uniform. Rather, it was worst in the nasal retina, where OHT decreased RGC density 26 to 42% across eccentricities (Figure 4). This spatial progression too is consistent with the sectorial pattern of RGC pathology observed in other glaucoma models [22, 23, 29, 30]. Though axonal transport to the superior colliculus was severely affected by OHT, our retinotopic maps also hint of sectorial progression (Figure 7; ).
Although there is some earlier evidence that BMD can preserve optic nerve axons and RGC axonal transport after acute injury [9, 31, 32], previous studies of BMD's effect in glaucoma models have focused on retinal outcomes [11, 12, 16, 33, 34]. Following episcleral vein cautery, BMD delivered via intraperitoneal injection prevented OHT-induced loss of RGC cell bodies labelled by retrograde transport of FluoroGold from the colliculus [12, 33]. In studies similar to ours in which OHT was induced by laser photocoagulation of episcleral and limbal veins for 3 weeks, BMD delivered by subcutaneous pump also improved the number of RGCs labeled retrogradely [11, 16, 34]. Our results are in agreement with these previous studies in that systemic BMD delivery increased RGC cell body survival (Figure 3 and Figure 4), even for our much longer period of OHT. However, given recent evidence that RGC somatic loss occurs late in disease progression , we examined axonal transport to the colliculus and axon survival in the nerve, both of which are challenged much earlier .
We found that systemic BMD treatment via subcutaneous delivery ameliorated the effects of OHT on each of our outcome measures. In the retina, BMD preserved RGC axonal and dendritic morphology (Figure 2), restored CTB uptake (Figure 3), and increased the number of SMI31+ RGCs, especially in the nasal retina where OHT-induced loss was greatest (Figure 4). In the optic nerve, BMD restored the axonal population to control levels as well (Figure 5). Finally, BMD improved anterograde axonal transport to the SC by 124% (Figure 6). As transport deficits to and from RGC targets are relevant to the progression of glaucoma [20, 21], the ability of BMD to preserve optic nerve axons and axonal transport suggest its neuroprotective effects are pertinent not only at RGC soma, but throughout the entire retinal projection.
BMD is a non-subtype-selective α2-adrenergic receptor agonist. All three α2 receptors subtypes are expressed in the retina, with expression of α2A and α2B receptors by RGCs and α2B receptors by glia [35, 36]. Activation of α2 receptors within the retina elicits many responses including reduced glial activation [11, 37], decreased oxidative stress [38, 39] and protection against apoptosis [40–42]. In addition, BMD may protect RGCs by regulating intracellular Ca2+ levels and glutamate availability within the retina [16, 43]. Excess glutamate can result in RGC apoptosis via overstimulation of NMDA receptors and subsequent increases in intracellular Ca2+. Stimulation of α2 receptors also results in the activation of the phosphatidylinositol 3 kinase (PI3K) pathway [44, 45]. Activation of the PI3K pathway promotes RGC survival after various injuries [46, 47], perhaps by altering gene expression [48–50], regulating protein activity [51, 52], or by affecting cellular metabolism [53, 54]. These and other responses make α2 agonists like BMD ideal neuroprotective agents within the retina.
A potential mechanism underling the neuroprotective effect of BMD on RGC axons involves the inactivation of glycogen synthase kinase-3 (GSK3) via the PI3K pathway. GSK3 is a constitutively active and ubiquitous kinase. In neurons, GSK3 phosphorylates microtubule-associated protein 1B, resulting in loss of stable microtubules . GSK3 also phosphorylates collapsin response mediator protein-2, which promotes microtubule assembly and links tubulin heterodimers to kinesin-1 to regulate protein transport to distal regions of axons [56–58]. Also, GSK3 activation is implicated in hyper-phosphorylation of axon cytoskeletal proteins, including neurofilaments [59, 60]. Interestingly, GSK3 can be inactivated by downstream kinases in the PI3K pathway [61–63], thus providing a possible link between α2 receptor activation and preservation of the axonal cytoskeleton. Activation of the PI3K pathway via treatment with BMD could counteract these changes, consequently preserving both axonal transport and structure. Since BMD applied either topically to the cornea or via systemic injection reaches the posterior segment in appreciable concentrations , its use could represent a viable intervention for combating early axon deficits in glaucoma.
Ocular hypertension in rats resulted in a substantial decline in RGC axonal transport to the superior colliculus, diminished axon survival in the optic nerve, and reduced RGC density in the retina, especially in the nasal quadrant. Systemic treatment with BMD significantly improved axonal transport and survival and either prevented or decreased loss of RGC density across retinal quadrants. As transport deficits to and from RGC targets are relevant to the progression of glaucoma, the ability of BMD to preserve optic nerve axons and axonal transport suggest its neuroprotective effects are pertinent not only at the cell body, but throughout the entire retinal projection as well.
Unilateral IOP elevation in male Sprague-Dawley rats (weight range, 350 - 400 g) was achieved by laser photocoagulation of episcleral and limbal veins as described previously . Rats were anesthetized with a mixture of ketamine (50 mg/kg), acepromazine (1 mg/kg), and xylazine (25 mg/kg) and two laser treatments were performed 1 week apart in order to achieve persistent IOP elevation. IOP was measured with a tonometer (TonoLab; Colonial Medical Supply, Franconia, NH). BMD (1 mg/kg/day) or vehicle phosphate-buffered saline (PBS) was administered systemically and continuously using an osmotic pump (Alzet; Durect, Cupertino, CA) inserted subcutaneously on the back of 16 animals per group. This concentration and mode of delivery was based on earlier results comparing 0.5 mg/kg/day vs. 1 mg/kg/day for efficacy in protecting RGC cell bodies during OHT . Treatment began at the time of initial IOP elevation and continued for 8 weeks. Age-matched naïve control rats (n = 16) were also included for comparison.
Forty-eight hours prior to sacrifice, rats were anesthetized with a mixture of ketamine (50 mg/kg) and xylazine (25 mg/kg) and eyes anesthetized locally using topical application of oxybuprocaine chlorhydrate . The hypertensive eye received an intravitreal injection (6 μl) of 0.5 mg Cholera toxin subunit β (CTB) conjugated to Alexa Fluor 594 (Molecular Probes, CA) while the control eye received a similar injection of CTB conjugated to Alexa Fluor 488 (Molecular Probes, CA) following earlier studies . After the 48 hour period, animals were deeply anesthetized with a lethal intraperitoneal injection of sodium pentobarbital (150 mg/kg) and perfused intracardially with warm (37°C) heparinized saline followed by 300 ml of Zamboni's fixative at 4°C. Brains were cryoprotected overnight in 30% sucrose/PBS and 50 μm coronal sections were taken through the midbrain and mounted on gelatin-coated slides. Retinas were dissected from the eye and vitreous removed after treatment with collagenase (400 units/ml) at 37°C for 15 minutes. Sections of optic nerve 2-3 mm proximal to the globe were isolated, post-fixed and prepared for embedding and semi-thin sectioning as previously described [66, 67].
RGCs were immuno-labelled using antibodies against phosphorylated neurofilament-heavy (SMI31, Sternberger Monoclonal Incorporated, Baltimore, MD) following our published protocol . Confocal images (0.101 mm2) were captured on an Olympus FV-1000 inverted confocal microscope along the midline of each retinal quadrant at 1 to 4 mm from the optic disc. The number of SMI31-positive and CTB-positive RGCs per image were counted and RGC density calculated as cells per mm2.
A 2- to 3-mm section of optic nerve proximal to the globe was isolated, post-fixed for 1 hour in 4% paraformaldehyde, and prepared for embedding and sem-ithin sectioning . Sections were stained with para-phenylenediamine (PPD) and then photographed using 100× oil-immersion and differential interference contrast optics. Photomicrographs of each optic nerve section were collected as a montage using an Olympus Provis AX70 microscope with motorized X-Y-Z stage and a digital CCD video camera. Each montage was contrast and edge-enhanced using the ImagePro software package (Media Cybernetics, CA). An additional routine was used to identify and count each axon in the montage for which a myelin sheath could be identified. We used this information to calculate the mean local axon density for each section of nerve. This was multiplied by the cross-sectional area of the nerve section to obtain an estimate of the total number of axons as described .
Alternate sections of the brain containing superior colliculus were photographed using a Spot-RT camera on an Olympus AX-70 upright microscope and intensity of label was quantified using ImagePro (Media Cybernetics, Bethesda, MD) as described previously . After normalizing with respect to background, intensity was recorded based on mediolateral location in the section. Intensity calculations from alternate sections were then combined to form a colorimetric representation of CTB signal across the retinotopic collicular map. For each colliculus, we determined the fraction of intact retinotopic map, defined as the percent area with CTB signal ≥70% maximum.
Unless otherwise indicated, all data are presented as the mean ± the standard error of the mean (SEM). SigmaPlot for Windows version 11.0 (Systat Software, Inc,; Chicago, IL) was used to calculate p values in comparing data using either ANOVA or t tests for data meeting criteria for normalcy or using non-parametric rank statistics for data failing normalcy.
cholera toxin subunit β
cyclic adenosine monophosphate
glycogen synthase kinase-3
phosphatidylinositol 3 kinase
retinal ganglion cells.
Funding provided by an Allergan, Inc. Discovery Research Grant (DJC), the American Health Assistance Foundation (DJC), the Melza M. and Frank Theodore Barr Foundation through the Glaucoma Research Foundation (DJC), NIH EY017427 (DJC), and an Unrestricted Grant from Research to Prevent Blindness to the Vanderbilt University School of Medicine Department of Ophthalmology and Visual Sciences. Imaging supported through the Vanderbilt University Medical Center Cell Imaging Shared Resource core facility (CTSA grant UL1 RR024975 from NCRR/NIH) and the Vanderbilt Vision Research Center (P30EY008126).
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.