Rat models for glaucoma research

Rat models for glaucoma research

Primary open-angle glaucoma (POAG) is the most common form of glaucoma in the United States and Europe. This is a slowly progressive form of optic nerve damage and blindness that begins with loss of peripheral vision and is followed by gradual shrinkage of remaining central vision and, ultimately, disappearance of even central, sharp

Corresponding author. Tel.: +1 503 494 3038; Fax: +1 503 494 3075; E-mail: morrisoj@ohsu.edu

vision. Unfortunately, blindness is irreversible, as optic nerve fibers do not regenerate.

Clinically, glaucoma is recognized by characteristic ‘‘cupping’’ of the optic nerve head (ONH). Cupping results from loss of retinal ganglion cell (RGC) axons, combined with collapse and posterior bowing of their supporting connective tissue sheets, or lamina cribrosa. In many patients, these physical changes are most pronounced in the superior and inferior poles of the ONH, leading to vertical enlargement of the cup and eventual undermining of the neural rim beneath the edge of the sclera in these areas.

286

This regional glaucoma injury produces a specific pattern of visual field loss (Quigley and Green, 1979; Quigley et al., 1981, 1988, 1989; Kerrigan-Baumrind et al., 2000). The most characteristic field defect in glaucoma (the arcuate defect) arches either above or below fixation, following the path of the nerve fiber bundles that pass through the superior and inferior poles of the optic nerve (Sommer et al., 1991; Tuulonen and Airaksinen, 1991). In glaucomatous optic nerve cross sections, this results in an ‘‘hour-glass’’ configuration of damage, with the greatest atrophy in the superior and inferior regions (Quigley and Green, 1979).

The best explanation for this injury pattern appears to lie in the structure of the ONH. In humans and other species with relatively large eyes, structural support to the optic nerve bundles is provided by the lamina cribrosa, which consists multiple ‘‘plates’’ of connective tissue that span the scleral opening, with pores that allow the optic nerve fiber bundles to exit the eye (Hernandez, 1992; Morrison et al., 1994). In humans, the pores of the superior and inferior laminas appear larger than elsewhere, and their connective tissue beams are thinner and more sparse (Quigley et al., 1981; Radius, 1981). This suggests that the thinner lamina cribrosa in these regions provides less adequate support for nerve fibers, increasing their risk for glaucomatous damage. This also strongly suggests that the ONH is the likely site of early injury in glaucoma, along with observations of axonal transport obstruction within the lamina in human glaucoma and animal models (Quigley et al., 1979, 1981; Pease et al., 2000; Martin et al., 2003).

Vision loss in POAG is generally slow and progressive and may take decades to develop. Clinically, it has been observed that patients with greater nerve damage and field loss can suffer progressive visual loss at levels of intraocular pressure (IOP) that would be tolerated by the eyes with less damage (Grant and Burke, 1982). This suggests that there is something unique about the glaucomatous eye that renders the remaining optic nerve fibers more susceptible to IOP.

This progressive susceptibility most likely results from gradual changes occurring in the ONH,

within RGCs, or both. Understanding these changes will help explain why many patients continue to progress despite apparently successful pressure lowering and why pressure in others is never remarkably higher than the normal range. It may also lead to the development of specific treatments designed to reverse or stabilize these conditions and preserve the remaining nerve fibers.

Because we currently lack noninvasive methods for assaying cellular function in humans, it is not yet possible to study these possibilities directly in human glaucoma. Therefore, relevant animal models will remain essential in helping us understand the mechanisms of glaucomatous optic nerve damage.

Use of animal models for POAG

Most animal models of glaucoma employ experimental elevation of IOP. Although IOP is only one of several known glaucoma risk factors, large clinical trials have confirmed that aggressive IOP lowering is beneficial in a spectrum of open-angle glaucomas, including normal-tension glaucoma, ocular hypertension, and early and late glaucoma (Drance, 1999; Kass et al., 2002; Leske et al., 2003; Nouri-Mahdavi et al., 2004b). From this, it is reasonable to expect that models based on elevated IOP will be highly relevant to optic nerve damage in open-angle glaucoma.

Experimental models of pressure-induced optic nerve damage possess certain advantages over spontaneous models. First, unilateral pressure elevation leaves the fellow eye available as a control against the effects of inter-animal variability. Second, the more predictable onset of the pressure elevation makes it possible to determine sequential events of optic nerve and retina damage.

Anatomically, nonhuman primates should be the most relevant experimental glaucoma model for studying human diseases. However, these animals are expensive, making them impractical for cell biology studies and drug trials requiring a large number of animals. Without special training and experience, it can be difficult to monitor IOP frequently enough to develop a solid understanding of the pressure insult to which the eye is exposed.

For these reasons, a less-expensive, more manageable model of pressure-induced optic nerve damage is needed. We feel this need can be supplied by laboratory rats. A large body of knowledge on the cell biology of neuropathology based on rats already exists, thus providing an array of tools for studying pressure-induced optic nerve damage when produced in these animals.

This chapter will discuss the current status of rat models that can be used to study optic nerve damage in POAG. This will include methods for measuring IOP in rats, assessing damage, and a comparison of the major experimental methods used to produce elevated IOP. We will conclude with a summary of the additional advances needed to optimize the ability of these models to help us improve care of glaucoma patients.

Suitability of the rat for models of optic nerve damage in POAG

It should be noted that the rat ONH lacks a welldeveloped, collagenous lamina cribrosa (Morrison et al., 1995a). While this differs from the primate, it does not diminish the utility of the rat for understanding cellular mechanisms of axonal injury from elevated IOP and in glaucoma (Fig. 1).

Sparse connective tissue associated with blood vessels has been described in the rat ONH, lined with astrocytes and composed of extracellular matrix materials very similar to that of the primate lamina cribrosa (Morrison et al., 1995a). We have also found that the cellular response to elevated IOP in rats is very similar to that in human and nonhuman primate glaucomas (Hernandez et al., 1990, 2000; Morrison et al., 1990; Johnson et al., 1996). This includes disorganization of the normal columnar structure of astrocytes and aberrant deposition of collagen and laminin within spaces normally occupied by axon bundles.

The most useful feature that the rat eye offers for POAG research lies in the close association between astrocytes and axons within the ONH. Ultrastructurally, rat ONH astrocyte processes lie within axonal bundles, providing intimate contact with all axons, a situation that also exists in the primate (Morrison et al., 2005) (Fig. 2). This close

287

Fig. 1. Longitudinal section of the rat ONH, demonstrating close apposition of the ONH (*) with the superior sclera (S), while the inferior sclera is separated from the ONH by central artery (A) and vein (V). Arrowheads indicate the beginning of the myelinated optic nerve, posterior to the sclera ( 400).

Fig. 2. TEM of the rat ONH cross section showing extension of astrocyte (As) processes into axon bundles. Note close contact between processes and nearly all axons, which are cut in cross section ( 30,000).

association between axons and astrocytes, which rest on the connective tissue lamina and peripapillary sclera of the ONH, provides a potential link by which IOP-generated forces in the load-bearing tissues of the ONH can get translated into axonal damage (Burgoyne et al., 2005). In this way, the rat eye presents an opportunity to understand the

288

specific cellular mechanisms leading to glaucomatous optic nerve damage.

Methods for measuring IOP in rats

The handheld Tono-Pen tonometer was the first instrument used successfully to measure IOP in rats (Moore et al., 1993, 1995). This instrument, designed for use in humans, works when held either horizontally or vertically. Based on the MacKay–Marg principle, the Tono-Pen tip consists an annulus that lies flush with the tip of a central post that is connected to a strain gauge fixed in the body of the instrument. Contacting the tip with the cornea causes the post to move relative to the annulus. Internal processors analyze the waveform of this movement and display pressure readings that result from acceptable waveforms. The instrument then calculates and displays an average of several acceptable readings along with the percent standard deviation, which is a statistical measure of the repeatability of these measurements, but not a true reflection of accuracy. Thus, a consistent error in the measurement technique can produce the wrong pressure reading, despite an excellent standard deviation.

In our initial evaluation, pressures measured by the Tono-Pen in cannulated eyes connected to a pressure transducer correlated well with the actual IOP (Moore et al., 1993) (Fig. 3a). We found that

individual valid readings were associated with contact of the instrument tip with the cornea firm enough to move the eye slightly posterior. Invalid readings were found to result from overly hard contact with the cornea (producing an inaccurately high reading) and from too light a contact, which is generally recognized by a single-digit reading (resulting from tear film contact only) and no posterior eye movement. Other inaccurate readings are those that occur when the tip breaks contact with the cornea (‘‘off’’ readings). Because the Tono-Pen cannot identify these readings as inaccurate, the instrument-generated ‘‘averages’’ will also be unreliable. Therefore, it is best to note individual valid readings and calculate the mean of these readings (Moore et al., 1993). This reliability has been well documented by good correlations between pressure readings in eyes with elevated IOP and nerve damage, as well as many cellular responses in the ONH and the retina (Jia et al., 2000b; Johnson et al., 2000, 2006, 2007; Schlamp et al., 2001; Ahmed et al., 2004; Fortune et al., 2004; Morrison et al., 2005; Pang et al., 2005a).

The Tono-Pen, however, possesses some disadvantages. Because it is designed for use with the human eye, which has a larger corneal curvature, the tip and the central post are relatively large for use in rats. This means that the instrument must be held exactly perpendicular to the cornea and centered perfectly on its apex to obtain acceptable readings. This reduced tolerance for even slight