Glaucoma: A Disease of Early Cellular Senescence

Introduction

This review examines glaucoma as a disease of early cellular senescence. There is ample evidence for this position, for both the anterior and the posterior ocular segments. The existing evidence to support this view of glaucoma and avenues to treatment, both existing and new, will be examined in the light of glaucoma pathophysiology as a premature aging process. 

Age Is a Strong Risk Factor for Glaucoma

Of all the risk factors that have been reported for glaucoma, age is among the most consistent and strong. All epidemiological studies have concluded that age is an important risk factor for glaucoma. These include the Baltimore Eye Survey, ¹ the Beaver Dam Eye Study, ² the Blue Mountains Eye Study, ³ the Melbourne Visual Impairment Project, ⁴ the Rotterdam Study, ⁵ the Barbados Eye Study, ⁶ Proyecto VER, ⁷ and the Los Angeles Latino Eye Study. ⁸ In Caucasian patients, the prevalence begins to increase sharply after the age of 60, while in African American and Hispanic subjects, the prevalence begins to increase at an earlier age, starting around 40. Other risk factors, including intraocular pressure (IOP), family history, high myopia, low blood pressure, and others, are more variable and less consistent than age. While the relative risk of glaucoma increases with increasing IOP, patients with glaucoma can suffer visual loss associated with either high or low pressures, yet the majority of patients with high IOP never develop glaucoma. In some populations, most markedly Asian, glaucoma seems to progress at relatively low IOPs. ^9,10  

Coexistent with the age-related increase in the prevalence of glaucoma is the age-related decrease in anterior segment outflow facility. That is, the resistance to fluid flow across the trabecular meshwork (TM) increases with age in a linear fashion, and begins at a fairly young age. ¹¹ This decreased outflow facility is largely responsible for the elevated IOP that is encountered in Western populations with increasing age. Markers of cellular senescence are found in the TM of patients with primary open-angle glaucoma (POAG) to a much greater degree than in age-matched controls. It is thought that aging of these cells leads to their decreased function and therefore a consequent decreased outflow facility. ¹² It is hypothesized that aging failure of the normal, regulatory proteolytic systems in the TM may be responsible for the observed pathophysiological alterations of the outflow pathway that may contribute to glaucoma. ¹³  

Along with the decreased function of the outflow apparatus with age is a decreased population of retinal ganglion cells (RGCs) in the retina. The RGC is the pivotal cell in the retina that electrically couples the retina with the brain; it is the cell that is primarily damaged by glaucoma. Pathological studies have shown a steady attrition of RGCs with normal aging, starting at a young age, in the amount of approximately 5000 cells per year. ¹⁴ Given that the average RGC population is approximately 1 million in the human eye, and that an individual can lose approximately half of that population before significant visual dysfunction occurs, every person would normally develop glaucoma by the age of 100! Glaucomatous loss of RGCs can be viewed as a premature aging effect, causing the typical and unique optic neuropathy that clinicians diagnose as glaucoma, and the consequent characteristic patterns of visual field loss. Clinical studies have also shown a decrease with age in the thickness of the nerve fiber layer in normal individuals. ¹⁵ The nerve fiber layer carries the axons of the RGCs to the optic nerve, where they exit the laminar cribrosa and connect to their targets in the central nervous system. 

Biomechanical factors within the optic nerve head have been hypothesized to play a central role in RGC physiology, and contribute to the optic neuropathy of aging and glaucoma. ¹⁶ The posterior sclera of old monkeys is significantly stiffer than that of younger individuals and leads to lower strains with the higher stresses associated with elevated levels of IOP. This age-related stiffening of the sclera significantly influences the biomechanical properties of the optic nerve head and may contribute to age-related susceptibility to glaucomatous optic nerve damage. ¹⁷ Similar changes occur in humans, which may be an important underpinning of the age relationship with glaucoma. ¹⁸  

Research on the association between glaucoma and cerebrospinal fluid (CSF) pressure has recently been reported. ¹⁹ One hypothesis posits that the translaminar gradient of pressure is more important than the level of IOP. The IOP would represent the pressure anterior to the lamina, while CSF pressure is that which is posterior to the lamina. This hypothesis suggests that the magnitude of CSF pressure is just as important as IOP as a risk factor for glaucoma. The lower the CSF pressure, the greater the translaminar gradient at any given IOP and the greater the risk of glaucomatous damage. It has been shown that CSF pressure decreases with age. ²⁰ Thus, changes in CSF circulatory physiology in the aging individual may be important in the pathogenesis of some forms of POAG, particularly that type sometimes referred to as “normal tension glaucoma.” ²¹  

Comparisons have been made between glaucoma and other age-related neurodegenerations, most notably Alzheimer's disease (AD). Comparisons between these two diseases regarding their pathogenesis and course could provide insight with respect to causative mechanisms. ²² A loss of ganglion cell axons from the optic nerve in AD, similar to that seen in glaucoma, has been reported. ²³ An increased prevalence of glaucoma occurs in AD patients: 26%, compared to 5% in an age-matched non-AD control group. ²⁴ Substantial similarities between rodent models of AD and glaucoma with respect to retinal damage have been reported. ²⁵ An approved treatment of AD with a glutamate receptor blocker, memantine, was the subject of a recent clinical trial of neuroprotective treatment for glaucoma. Although the results of the trial were not encouraging (Memantine Study, Allergan, Inc., unpublished data, 2008), it is still unclear whether or not this drug, or drugs like this, would benefit a subset of patients with progressive POAG. 

Although much of this review is focused on POAG, the most common form of glaucoma in Western countries, it is important not to exclude the substantial societal load posed by the existence of angle closure glaucoma. This is clearly an age-related disease, and tends to affect Asian populations to a greater extent than Western populations. The most important cause for primary angle closure glaucoma is related to anterior chamber narrowing and closure of the anterior chamber angle and to the aging increase in size of the crystalline lens. ²⁶  

Rate of Disease Versus Longevity

Figure 1 represents a conceptual schematic of the relationship between visual function and longevity in various scenarios of chronic glaucoma. It is apparent that a faster rate of disease will cross a threshold to a level of visual disability earlier in one's lifetime. A corollary of this would be that if the rate of disease is slow, some patients may do quite well without treatment, and will live their entire lives free from visual disability. There is a large body of clinical trial data demonstrating that early intervention is more effective at slowing the rate of progression than late intervention. ^27,28 When this is applied to the schematic with respect to “fast progressors,” one can see that early intervention has a chance of keeping patients with a fast-paced disease from becoming visually disabled, even with long longevity. This becomes an important concept given the aging population and the imminent explosion of the proportion of older individuals in our population. ²⁹  

Figure 1

 

A schematic representation of the relationship between visual function and longevity in glaucoma. Early intervention in fast progressors is required to prevent visual disability.

Estimates of the rate at which individuals deteriorate help direct treatment to the right patients at the right time. This also allows us to avoid the artificial dichotomy of glaucoma patients being either “stable” or “worse”; in reality, all patients are getting worse, but at different rates. We should also accept that the functional and structural scales with which rates are measured are not synchronous, nor are they linear. A considerable body of clinical research has been devoted to the most appropriate way to measure the rates of visual field loss in glaucoma. These approaches are largely confounded by (1) the high level of intertest variability inherent in psychophysical measurements; (2) the limitations of the sensitivity and specificity of the technique; (3) a low signal-to-noise ratio; (4) the requirement to perform multiple tests to reduce the noise and make the signal detectable; (5) the requirement for confirmatory tests; (6) the lack of a “gold standard”; and (7) the slow course of the disease, which may progress over years or decades. ³⁰  

Regressions of visual field thresholds over time, in either a pointwise, cluster, or global fashion, have been used. The most frequently applied thus far, and the simplest, is a linear model. Evidence suggests, however, that a linear model may not be the best for all patients. ³¹ A recent approach uses a pointwise exponential regression and isolates fast and slow components of visual field decay in glaucoma. ³² Such an approach can be used to identify patients who are progressing very quickly, and can be used to predict patterns of future visual field loss within appropriate confidence intervals. This approach has been shown to be effective across a wide range of disease severity and can identify “rapid progressors” for appropriately aggressive treatment. ³¹ Figure 2 shows an example of this approach applied to a patient having many visual fields over a period of nine years. While the superior paracentral area of the visual field is getting worse very quickly (at the rate of approximately 30% per year), the remainder of the field is quite stable and shows essentially no damage. This is a patient who should be classified as a fast progressor, given the dramatic loss in near-central vision with great impact on quality of life. Such focal but important worsening of visual function can be missed by inspection of global indices such as MD or VFI alone because of their inherent lack of sensitivity to localized defects. ³³ A search for baseline prognostic factors that can predict which patients are at highest risk for rapid visual deterioration is in progress. Preliminary results suggest that age, and measures of underlying damage such as the severity of visual field loss and the vertical cup-to-disc ratio, are significant and important prognostic factors (Lee J, Caprioli J, Coleman A, et al., unpublished data, 2008). 

Figure 2

 

An example of the differential behavior of the fast and slow components in a series of visual fields from an eye of a patient with treated primary open-angle glaucoma. The test locations shown in red represent a cluster of points that decay at a faster rate than the rest of the visual field. Decay rates of the fast and slow clusters are shown, respectively, on the right.

Cellular Senescence Contributes to Retinal Ganglion Cell Death

Figure 3 summarizes many of the genetic defects heretofore reported to be important in the pathogenesis of glaucoma. Most of these genetic defects are encountered in the congenital and developmental forms of glaucoma that are relatively rare. So far, no single gene or group of genes can be held responsible for a significant proportion of patients with the adult form of POAG. It is this author's opinion that a host of genetic variants are involved in the pathogenesis of glaucoma, not to mention posttranscriptional alterations and environmental interactions. Glaucoma most likely represents a group of diseases with diverse molecular mechanisms of pathogenesis, which have in common a final common pathway of the typical optic nerve damage and consequent characteristic patterns of visual field loss that we call glaucoma, and which may progress to blindness in the most severe cases. Because of the diverse nature of the causes of glaucoma and the heterogeneity of the disease, it appears that both IOP-dependent and IOP-independent processes are at work. Different patients may have varying contributions from these two broad categories of insults, on a spectrum that may vary from completely IOP-dependent to completely IOP-independent disease. 

Figure 3

 

This table summarizes many of the genetic defects so far reported to be associated with primary and developmental glaucoma.

Figure 4 is the conceptual organization of some of the leading theories of the pathogenesis of RGC damage in glaucoma. While a comprehensive review of these pathogenic theories is not possible within the scope of this review, I would like to focus on the ubiquitin–proteasome system (UPS) and chaperone system (CS) in aging. The UPS is the main intracellular pathway that regulates protein turnover and is essential for cellular homeostasis. This system allows the cell to regulate its protein expression and respond to changing physiologic conditions caused by disease. The major theory regarding the causes of cellular aging invokes the concept that the UPS becomes slowly and progressively dysfunctional with age. Cellular protein “junk” thereby accumulates in the cell, causing a protein constipation that can no longer be processed by the aging UPS, and ultimately leads to cellular dysfunction and death. ^34,35 Stress proteins, also known as protein chaperones or heat shock proteins, play a similar role in helping proteins fold into their appropriate, functional conformations, thereby also preventing the accumulation of dysfunctional proteins within the cell. Heat shock factors are known to stimulate the expression of these stress proteins in RGCs. ³⁶ Figure 5 shows an overexpression of heat shock factor 2 in aging rats compared to younger rats, likely a consequence of a decreased constitutive expression of heat shock proteins. Anti-heat shock protein antibodies have been described in the sera of patients with POAG compared to controls. ³⁷ A dysfunctional CS associated with aging can lead to increased vulnerability of RGCs to a variety of insults, including those that lead to glaucoma. 

Figure 4

 

A conceptual organization of some of the leading theories to explain the molecular mechanisms of retinal ganglion cell death in glaucoma. Both necrotic and apoptotic pathways to cell death are shown.

Figure 5

 

Immunohistochemistry of heat shock factor 2 in retinal ganglion cells of aging rats. Elevated levels of heat shock factor in older rats, right photomicrograph, compared to younger rats, left, are consistent with decreasing levels of constitutive chaperone proteins. HSF, heat shock factor; GFAP, glial fibrillary acidic protein. GFAP is used as a housekeeping protein marker.

The only known maneuver that extends life in animal models of longevity is calorie deprivation. This is a case in which a little starvation improves health and longevity. It has been shown in a rat model that calorie restriction, as mimicked by the administration of 2-deoxy-D glucose, protects RGCs against ischemia and excitotoxicity, ³⁸ two possible mechanisms for RGC death from glaucoma. 

Stem Cells Present Opportunities and Challenges to Prevent and Repair Optic Nerve Damage

Stem cells offer the advantages of having unlimited self-renewal and of being pluripotent; that is, they can form cells and tissues from any of the three embryonic germ cell layers. There are three types of stem cells: embryonic, induced pluripotent, and somatic (adult). Advances in stem cell technology have provided us with the means and opportunity to replace damaged or diseased cells with normal functioning cells. This approach could be used to replace abnormal, diseased, or missing cells in either the anterior or posterior segment of the eye. Aging changes in the TM lead to progressive dysfunction with age. Normal TM cells secrete matrix metalloproteinases, phagocytose debris, and provide fluid transport into Schlemm's canal. Decreased TM cellularity is associated with decreased outflow facility and increased IOP. It has been observed that laser trabeculoplasty increases TM cell division, ³⁹ and that most cell division occurs in the anterior TM at the insertion zone near Schwalbe's line. It has been suggested that this insertion zone may contain somatic stem cells. ⁴⁰ TM contains a cell population that exhibits markers of pluripotent stem cells. ⁴¹ It has been shown that TGF-β2 induces premature senescence-associated TM changes and activates a senescence-related signal transduction pathway. ⁴² Blocking or reducing TGF-β2 levels in TM could help prevent the aging process in the meshwork, and thereby aid in the treatment of open-angle glaucoma. 

Another avenue for stem cell–related treatment of glaucoma is the delivery of neurotrophic factors to RGCs at risk. For instance, brain-derived neurotrophic factor (BDNF) is 14-kDa protein that is essential for RGC development and survival. Retrograde transport of BDNF from the thalamus to RGCs is interrupted in animal models of increased IOP. Delivery of BDNF can prevent ganglion cell death under these circumstances. ⁴³ Stem cells injected into the vitreous and manipulated to secrete BDNF or other survival factors may be used to deliver these factors to the specific target tissue in therapeutic concentrations. 

The final frontier of stem cell therapy resides in the theoretical ability for RGCs that are dysfunctional or lost to be replaced by new ones. Such an approach lags considerably behind similar approaches to replace photoreceptors by injecting stem cells under the retina. ⁴⁴ The challenges to stem cell RGC replacement include (1) cell integration into the ganglion cell layer; (2) the successful formation of local synapses within the retina; (3) the extension of axons through the lamina cribrosa to targets in the brain; and (4) the formation of correct, topographically oriented connections to the central nervous system (in humans, mostly in the lateral geniculate nucleus). Preliminary work has shown that human Müller cells can differentiate into RGCs with appropriate molecular manipulations, including notch inhibition and fibroblast growth factor supplementation. ⁴⁵ However, such cells have thus far not been shown to make functional connections to their central targets. 

The associated risks of stem cell therapy should be noted. These include teratoma formation, immune rejection, viral integration into the human genome, and ethical concerns. 

How to Slow or Reverse Age-Related Damage: Approaches to Saving Sight in Glaucoma

Treatments in three general arenas can be described as effective treatments for glaucoma: neuroprotection, neurorescue, and neuroregeneration. Neuroprotection, through IOP reduction or other treatments, is directed to prevent damage to RGCs. Neurorescue provides treatment to improve the function of dysfunctional RGCs, that is, those that are damaged but are still present and are recoverable. Neuroregeneration refers to the replacement of RGCs with functional neurons that make appropriate synapses to their correct central nervous system targets and can restore visual function to those who have already lost vision from glaucoma. 

Animal models have been used to demonstrate a relationship between IOP and glaucomatous damage to RGCs. After several weeks of moderately elevated IOP in the rat, significant loss of RGCs can be demonstrated in a dose-dependent fashion (a product of time and IOP level). ⁴⁶ While the revelation of exact molecular mechanisms of damage remains elusive, presumably due to the complexity of the underlying cellular pathophysiology, models like this have been used to demonstrate new approaches to neuroprotection and neurorescue, and have begun to elucidate the complex challenges presented by neuroregeneration. Since information about exact molecular mechanisms of damage is lacking, and since glaucoma may represent a diverse group of diseases with different molecular mechanisms, one realistic approach would be to treat the disease on a regulatory (mechanism-nonspecific) basis. I have discussed the central role of the ubiquitin–proteasome and chaperone systems in aging. One appealing approach may be to improve the function of the increasingly dysfunctional protein regulatory systems (UPS and CS) in the aging organism. Heat shock protein induction with heat, zinc, and certain drugs is an effective approach toward reducing the vulnerability of RGCs to IOP-induced damage in the rat model. Drugs currently approved for other indications and nutraceuticals should be considered as possible effective means of treatment through endogenous neuroprotective regulatory pathways. 

Our goals in the treatment of glaucoma are to slow, stop, or repair damage to the RGCs caused by the disease. Heretofore, it was a general clinical dictum that visual loss from glaucoma cannot be reversed. Recent work from the laboratory and the clinic suggests that this may not be so. The IOP and age dependence of RGC dysfunction has been studied in the DBA/2J mouse model of glaucoma. RGC susceptibility to IOP elevation increases with age, and RGC dysfunction in older mice can be improved with IOP reduction. ⁴⁷ Recent clinical evidence suggests that robust IOP reduction in humans with glaucoma can improve visual sensitivities in the portion of the visual field most affected by glaucoma. Figure 6 shows a summary of the behavior of the fast and slow components of visual field decay after glaucoma surgery and robust IOP reduction. While there is significant slowing of the fast component of VF decay after surgery, it is also apparent that there is a significant improvement of visual sensitivities at these locations after surgery. This improvement is typical, is not uncommon, and is significant in amplitude and appears to be sustained. Such observations from the laboratory and the clinic are consistent with the concept that a population of RGCs that has become dysfunctional from glaucoma, but is not yet dead, may respond to treatment and that its function can be restored. That is, there is a group of ganglion cells that may be sick but not dead, and can be revived. This repair may be considered neurorescue and may already be available in the form of treatment by robust IOP reduction. 

Figure 6

 

The average behavior of the slow and fast components of the visual fields of 83 eyes of 83 patients who underwent trabeculectomy. The black arrow indicates the time of trabeculectomy. The blue and yellow curves indicate the rates of decay of visual sensitivities before and after surgery. Fits are made with an exponential model. The reduction in the fast rate after surgery is statistically significant. The small decrease of sensitivity in the slow component is thought to be caused by media changes after surgery.

The concept of neurorescue is relatively new as it pertains to glaucoma treatment. There are hints that robust IOP reduction through surgery, or perhaps through new drug delivery systems, may offer promise to rescue RGCs that may be sick but not yet dead from glaucoma. The delivery of growth factors or survival factors, either through the injection of stem cells tuned to deliver these factors or through novel, long-term drug delivery systems, also offers hope for new effective treatments. 

The reduction of IOP is the only proven treatment that provides neuroprotection against glaucoma. The evidence for this comes from numerous high-quality clinical trials and a large body of clinical evidence; it is strong and irrefutable. Recent evidence, both basic and clinical, suggests that alpha agonism may provide neuroprotection beyond that which can be attributed to its IOP-lowering properties. ^48,49 Confirmatory studies, however, are lacking, and alternative hypotheses can also explain the results. ⁵⁰ Future approaches would depend in part upon our ability to identify specific molecular mechanisms as triggers for RGC death. In the meantime, mechanism-nonspecific approaches through the manipulation of regulatory pathways offer promise. Examples may include the regulation of the UPS and CS related to their important roles in the age-related dysfunction of cells. 

Neuroregeneration, the final potential treatment realm for glaucoma, has been an elusive goal. There is evidence that retinal ganglion cell bodies exist in the retina long after their axons, and therefore their function, have vanished from the optic nerve. ⁵¹ Appropriate triggers for these cells to regrow their axons along existing tracts and make appropriate synaptic connections to their central nervous system targets should be identified. Pluripotent stem cells have the ability to replace RGCs in the retina. Preliminary work with Müller cell differentiation holds promise. Formidable challenges exist; these include cellular integration, formation of synapses in the retina, extension of axons through the lamina to the correct targets in the brain, and the formation of topographically oriented connections. An additional hurdle may also exist: Target cells may actually disappear in the brain after the cessation of signaling from peripheral neurons. ⁵²  

Summary of Key Needs and Opportunities

•  
  Methods to safely provide robust decreases of IOP require further development. Such approaches may be achieved through new forms of surgery or through drug delivery by novel intraocular sustained-release devices.
•  
  Delivery of growth factors through new drug delivery platforms or by injection of stem cells into the vitreous may provide neurorescue of RGCs. Development of the basic techniques followed by clinical trials is required to establish the efficacy of the approach.
•  
  The search for neuroprotective drugs should continue. Until the exact mechanisms of molecular damage are known, the pursuit of mechanism-nonspecific pathways, through the manipulation of regulatory pathways, may be productive. The manipulation of the aging UPS and CP, which result in protein dysfunction, is a particularly attractive area of research. Drugs that are already available, approved for other indications, and nutraceuticals should be considered as immediate and readily available targets for drug development.
•  
  The challenges of neuroregeneration seem formidable. Early work has shown that Müller cells can differentiate into functioning RGCs. Work should begin on finding solutions to the considerable problems related to re-creating central neuronal connections. The impact of even limited success in this area would be great, and could be readily extended to other neurodegenerative diseases.

Acknowledgments

Disclosure: J. Caprioli, None