The relationship between neurotrophic factors and CaMKII in the death and survival of retinal ganglion cells

Introduction

‘‘All of the present studies suggest that neuroprotective therapy will probably become the treatment of choice in the near future for glaucomatous optic

Corresponding author. Tel.: 502-852-1474; Fax: 502-852-3028; E-mail: nigelcooper@louisville.edu

neuropathy’’ (Yamamoto, 2001). Neurotrophic factors have the potential to aid recovery from diseases and disorders of the nervous system, including the neural retina and its optic nerve. For this potential to be fully realized, we first need to thoroughly establish the cellular networks, including the genomic and environmental parameters, which regulate such factors. Specifically, we need to map the signal transduction pathways,

522

which stimulate and/or inhibit or otherwise modify the cell’s regulatory machinery for neurotrophic factors. This review attempts to summarize some of the evidence for the pivotal involvement of Ca2+ and calmodulin-dependent protein kinase II (CaMKII) in cell death and survival pathways in the retina. In particular, we review the evidence for the signal transduction pathway leading to the transcription, translation, secretion, and action of brain-derived neurotrophic factor (BDNF) from the retinal ganglion cells (RGCs). These cells are particularly vulnerable in retinal diseases such as retinal ischemia and glaucoma.

Glaucoma and the RGCs

While glaucoma is the leading cause of blindness in the world, it cannot be simply defined, due to its multiple etiologies and its several underlying genetic and environmental causes. However, we can speak of a class of eye diseases in which optic nerve damage is a most frequently observed feature. In addition to optic nerve damage and the consequential visual field defects, increased intraocular pressure (IOP), is often cited as a third correlate in the most common form of the disease, although this may be a coincidental observation or aggravating condition because normal-tension glaucoma results in the same blinding condition without the elevated IOP. The incidence of glaucoma increases with increasing age (Armaly et al., 1980), and other known risk factors including family history, ethnicity, myopia, and diabetes (Leske, 1983).

Once nerve damage occurs there is a functional discontinuity between the retina and the higher order visual processing centers with a consequential degradation and/or loss of vision. This loss is permanent because of the well-established dogmas;

(1) the axons of central nervous system neurons in mammals do not re-grow; (2) axonal section results in the death of the cell bodies in the central nervous system of mammals.

Unfortunately, glaucoma is an insidious disease in which visual loss occurs incrementally and patients do not complain of problems until much of the optic nerve damage has already occurred. The exact causes of optic nerve damage from glaucoma are not fully understood, but leading postulates include mechanical compression and/or

decreased blood flow to the optic nerve. As to the death of the RGCs, one related hypothesis implicates the loss of neurotrophic factor, due to a block in its retrograde transport from higher visual centers, along the optic nerve to the cell somata of RGCs. Morphological changes of the optic nerve head are some of the earliest macroscopic signs of trouble in the clinic, although it is possible that such signs are not only relatively late indicators of trouble in the visual system, but also distal to the most likely primary causes of the problem. While there are those who strongly advocate a recreation of the condition seen in the clinic in the development of animal and cellular models of this disease, such advocacy may actually impede the exploration and discovery derived from other interesting avenues of research.

Observation of altered structure in the optic nerve head, it is frequently argued, precedes observations of the malaise and/or death of the RGCs. Axonopathy is not unique to glaucoma and perhaps we can learn something from studies and models of other diseases. For example, in a cell culture model for amyotrophic lateral sclerosis, excitotoxicity is known to lead to an accumulation of cytoskeletal proteins in the axons of cultured spinal neurons (King et al., 2007). This may be worthy of further investigation in RGCs. It is most important to recall the dogma, which contends that the cell somata of neurons are the genetic and trophic centers which nourish and support the other parts of the cell including, in the case of glaucoma, the RGC axons present in the optic nerve. Because of this fact, it would seem plausible to examine all models and experimental paradigms which tell us something about the prodeath and pro-survival machinery within these genetic and trophic centers of the RGCs. Thus, the science should look far beyond the account and treatment of symptoms to the underlying and fundamental molecular mechanisms. With this approach, we may be able to determine if the cellular machinery can be manipulated to advantage, to inhibit cell death, to promote cell survival, and even to promote axonal re-growth.

The ultimate and permanent loss of vision in the glaucomatous patient is due to the death of the RGCs. These cells die by apoptosis (Quigley, 1995; Kerrigan et al., 1997) leading some investigators to

propose anti-apoptotic therapies for the treatment of glaucoma (Nickells, 1999; Tatton, 1999). Elevated levels of ocular glutamate have been proposed as being responsible for the glaucomatous death of RGCs (Vorverk et al., 1999) and several glutamate receptor antagonists have been tested as potential antiglaucoma agents (Hare et al., 2001; Chaudhary et al., 1998; Pang et al., 1999). Although elevated vitreous glutamate levels have been reported in glaucoma patients and in animal models of glaucoma (Dreyer et al., 1996) other studies were unable to confirm these initial findings (Honkanen et al., 2003).

Are other retinal cells affected in glaucoma?

It is well established that glaucoma describes a condition which impacts the life of the RGCs, and therefore, this condition may be considered as part of a larger class of diseases/disorders in which these particular cells are affected. Although it initially appeared that only RGCs were destroyed in the glaucomatous retina, other retinal cells may also become damaged. Several studies over the last decade or so indicate that there may also be losses among the amacrine cells in the glaucomatoushuman, the glaucomatous non-human primate, and in other animal models (Table 1), although the significance of these data with respect to the human condition is less well established.

Kielczewski et al. (2005) used translimbal laser treatment to raise IOP and found that rats with increased IOP did not have significantly reduced numbers of amacrine cells when compared to control eyes. This is in contradiction to many other examples

523

where amacrine cell loss occurs as a result of elevated IOP, and this apparent contradiction may be indicative of the manner in which the elevated IOP is produced. However, the former study did note significant reductions in GABA and glycine containing amacrine cells in eyes when another well established model of glaucoma was used, in which the optic nerve had been transected. Hood et al. (1999) used intravitreal injections of N-methyl-D- aspartate (NMDA) and tetrodotoxin (TTX) in monkeys to examine the altered waveforms of the multifocal electroretinogram (mfERG). These authors noted changes in the ERG waveform contributed by the inner retina (ganglion and amacrine cells) of the treated monkey eyes, which are consistent with records from patients with glaucoma and diabetes. In addition to the RGCs, Diijik et al. (2004) used cell specific transcript to assess cell types lost following acutely elevated IOP, these authors noted that among the amacrine cells, the glycinergic AII amacrine cells were the most affected in this animal model. The authors, May and Mittag (2004), used immunohistochemistry to document the loss of neuronal nitric oxide synthase (nNOS) positive retinal amacrine cells in the DBA/ 2NNia murine model of angle-closure glaucoma. It should be noted that the possibility of cell death in the displaced amacrine cells of the RGC layer, has not yet been well documented because of the lack of specific markers for such cells. It seems likely that the shared vulnerabilities of the RGCs and the amacrine cells in several of these studies reflect some shared phenotype, some common property or properties in their cell membranes and/or protein compositions. The presence of both ionotropic glutamate receptors (Laabich and Cooper, 1999) and CaMKII (Laabich