|Home | About | Journals | Submit | Contact Us | Français|
We tested the hypothesis that glaucoma disrupts electrophysiological conduction properties and axon function in optic nerve as a function of intraocular pressure (IOP) levels and age in the DBA/2J mouse model of glaucoma. The amplitude and the integral of electrical signals evoked along the axons decreased considerably by 6 months of age as a function of increasing IOP levels. At young ages, raised IOP was directly associated with increased vulnerability to metabolic challenge. Changes in the physiological function of the optic nerves were accentuated with aging, leading to loss of compound action potential in an entire population of fibers—small, slow conducting axons. This loss was accompanied with loss of small fiber axon counts and declining metabolic reserve by demonstrating IOP-dependent ATP decrease in mouse optic nerves. These data shed light on a novel potential mechanism of glaucoma pathology whereby increased IOP and declining metabolic capacity lead to axon liability and eventually dysfunction and loss.
Intraocular pressure (IOP) increase, a major risk factor for glaucoma development (Flanagan, 1998; Friedman et al., 2004), is a result of pigment dispersion that blocks aqueous humor outflow in eyes from the DBA/2J mouse glaucoma model (John et al., 1998; Anderson et al., 2002; Libby et al., 2005b). The resultant IOP increase has been correlated with neurodegenerative changes such as axon loss (John et al., 1998; Libby et al., 2005b; Inman et al., 2006). The mechanism of vision loss in glaucoma is not understood, but evidence such as transport blockade of tracers or cargo at the optic nerve head (ONH) (Johansson, 1986, 1988; Quigley et al., 2000), retention of intraretinal retinal ganglion cell (RGC) axons concomitant with axon loss in the optic nerve (Soto et al., 2008), a retrograde course of degeneration as assessed through axon quantification (Schlamp et al., 2006), and maintenance of RGC somata while retrograde label is lost (Buckingham et al., 2008) demonstrates that RGC axons are a critical site of early pathological change. Bax knockout mice in which degeneration proceeds in the axon while RGC somata survive increased IOP (Libby et al., 2005a) further illustrates the compartmentalization of degenerative changes that can occur. These data indicate a primary sensitivity of the optic nerve to glaucomatous insult and argue strongly for further analysis of potential mechanisms that underpin axonal vulnerability.
There are a number of optic nerve qualities that may contribute to a unique sensitivity to injury in glaucoma, a disease whose chronic nature implicates age in the pathophysiology of the disease. Studies of optic nerve in the context of white matter injury demonstrate age is a determining factor in ischemic injury response. In contrast to young optic nerve, severity of injury from the same ischemic insult was greater in optic nerve from older mice (≥12 months) (Baltan et al., 2008). Increased vulnerability of aging white matter to ischemia has clear implications for glaucoma because age is a major risk factor. In addition, the vascular dysregulation that has been correlated with increased progression of visual field defect in glaucoma patients (Tokunaga et al., 2004; Grieshaber and Flammer, 2005; Resch et al., 2009) and underlies the impaired RGC activity in animal models (Grehn and Prost, 1983; Siliprandi et al., 1988) likely leads to ischemic damage.
To characterize the onset and mechanisms of optic nerve functional decline in the DBA/2J mouse glaucoma model, we utilized an in vitro optic nerve preparation that allows quantitative evaluation of injury using electrophysiology, immunochemical and biochemical methods. Our results demonstrate an IOP-dependent increased vulnerability to oxygen-glucose deprivation (OGD) in the young optic nerve (ON) of DBA/2J mice. We also observed an interaction of IOP and aging that contributed to prominent loss of slow conducting fibers in old mice with high IOP. Increased IOP hindered the compound action potential (CAP) area recovery after OGD in young ONs, whereas recovery in old ONs was independent of IOP. Consistent with this, increased IOP compromised ATP levels in young and old ON, but age had a greater impact on ON ATP levels in older mice. Some of these results were previously presented in abstract form (Baltan et al., 2009; Inman et al., 2009).
These experiments adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. DBA/2J and C57Bl/6 mice were originally obtained from Jackson Laboratories (Bar Harbor, ME) and then bred and housed in a specific pathogen-free barrier facility at Harborview Medical Center, Seattle, WA. New breeders were added to the DBA/2J colony every four months to counteract any potential genetic drift. Mice were maintained in a 12h light-dark cycle with standard rodent chow available ad libitum. The University of Washington Institutional Animal Care and Use Committee approved of all experimental procedures.
Intraocular pressure was measured in lightly anesthetized (Avertin, 1.3% tribromoethanol, 0.8% tert-amyl alcohol) mice using the TonoLab rebound tonometer from Tiolat Oy (Helsinki, Finland), calibrated for mice. Proparacaine (0.5% proparacaine hydrochloride ophthalmic solution, Bausch&Lomb, Tampa, FL) was applied to the eye before IOP measurement. Twenty IOP measurements were taken and averaged for each eye, each month between two and 10 months of age. IOP does not increase equally in both eyes in the DBA/2J mice. IOP is a continuous variable and mice were chosen for their IOP extremes within their respective age groups. For example, from all mice at 10 months of age, we chose the ones with the highest and the lowest IOP. Instrument averages (automatic to the device after every sixth reading) and readings beyond one standard deviation (indicative of an incorrect reading) were omitted. IOP readings shown in Figures 1 and and66 are the final averaged measure taken from that respective eye.
ATP extraction was done using a method first described for brain and modified for optic nerve tissue (Khan, 2003). The nerve was cut in 2-3 small pieces using a small needle while in 75μl 10% HClO4, then homogenized 5 times by sonication for 1 sec and centrifuged at 4500 rpm for 10 min at 40°C. The supernatant was collected then neutralized with 30 μl 2.5 M KOH and centrifuged at 14,000 rpm for 10 min at 40°C. The precipitate was removed and the supernatant was kept on ice for further ATP measurement. Total cellular ATP concentrations were measured using the ATP bioluminescence assay kit (Roche, Pleasanton, CA), based on the ATP dependence of luciferase catalyzed oxidation of luciferin. Samples were diluted and mixed with the luciferase reagent then absorbance measured at 560 nm. Blank values were subtracted from the raw data and ATP concentrations were calculated from a log-log plot of the standard curve data and normalized by the protein concentration. The values are expressed as μmole ATP/ mg protein.
Mice were transcardially perfused with 4% paraformaldehyde in 0.1M phosphate buffer, pH 7.4 (PFA). Optic nerves were dissected free then post-fixed for 24h in PFA before rinsing in cacodylate buffer and then further fixation in 1% osmium (OsO4) in sodium cacodylate buffer for 30 minutes. Nerves were then rinsed in cacodylate buffer, dehydrated in graded alcohol solutions, and embedded in Epon-Araldite at 60°C for 72 hours. For axon quantification, we used an Olympus Provis AX70 microscope equipped with a motorized X-Y-Z stage, a digital video camera, and 100x oil-immersion, differential interference contrast optics. We collected photomicrographs from 1-2 μm cross-sections of each nerve en montage so that the entire surface area of the nerve was represented. Each micrograph was contrast and edge-enhanced using macro-routines written using the ImagePro software package (Media Cybernetics, CA). An additional routine was used to identify and count each axon for which a single, intact myelin sheath could be identified and to calculate the internal cross-sectional area of each identified axon. Ten mouse optic nerves (ONs) each at low and high IOP were quantified.
The experimenter was blinded to the IOP levels of the mice during the preparation of the optic nerves and physiological recordings. Optic nerves were removed from DBA/2J of C57Bl/6 mice following CO2 narcosis and decapitation. Preparation of nerves, recording technique and ischemic insult proceeded as previously described (Baltan et al., 2008). After gently removing the dural sheath, optic nerves were placed in an interface perfusion chamber (Medical Systems, Greenvale, NY). Both nerves from each animal were used simultaneously on identical setups to keep animal numbers to a minimum. Mouse optic nerves (ONs) were superfused with artificial cerebrospinal fluid (ACSF) containing the following (in mmol/L): 124 NaCl, 3.0 KCl, 2.0 CaCl2, 2.0 MgCl2, 1.25 NaH2PO4, 23 NaHCO3, and 10 glucose. The perfusion chamber was continuously aerated by a humidified gas mixture of 95% O2/5% CO2. All experiments were performed at 37°C.
Suction electrodes back-filled with ACSF were used for stimulation (Isostim 520; WPI, Sarasota, FL) and recording compound action potential (CAP). The recording electrode was connected to an Axoclamp 2A amplifier, and the signal was amplified 50 times, filtered at 30 kHz, and acquired at 20-30 kHz. Stimulus pulse (30-50 μs duration) strength was adjusted to evoke the maximum CAP possible, and then increased an additional 25% for supramaximal stimulation. The ONs were equilibrated for at least 15 min in the chamber in normal ACSF before experiments. During experiments, the supramaximal CAP was elicited every 30s.
Oxygen glucose deprivation (OGD) was induced by switching to glucose-free ACSF (replaced with equimolar sucrose to maintain osmolarity) and a gas mixture containing 95% N2/5% CO2. To ensure no oxygen was inadvertently delivered by the superfusate, ACSF was always bubbled with 95% N2/5% CO2 Ransom and Philbin, 1992). OGD was applied for 45 or 60 min to determine the effects on axonal conduction and recovery. After OGD, control ACSF and O2 were restored and CAPs were recorded for up to 6h.
The suction electrode recording configuration at supramaximal stimulation allows all axons in optic nerve to be activated and the resultant action potential from each axon to be recorded. This method has the advantage of enabling stable recording of three-peaked (P1, P2 and P3) characteristic shape of CAPs (See Figure 2A). These peaks reflect three sub-groups of axon populations based on their conduction velocity (Geddes 1972) and the area under the CAP reflecting the number of contributing axons (Cummins et al., 1979). Supramaximally evoked CAPs were used to assess the characteristic three-peaked shape, the maximum CAP amplitude (P2 amplitude), the CAP area and the latency. The CAP latency was measured as the time between the onset of the stimulus artifact to the first peak as previously described (Nashmi et al., 1997). The CAP amplitude, the CAP area and the latency were calculated for each animal after averaging thirty traces from baseline conditions. The stimulus artifact was isolated after addition of 1 μM TTX in the bath at the end of the experiments to accurately determine latency measurements.
K+ channel blocker 4-aminopyridine (Tocris, 100 μM) dissolved in ACSF was applied while recording baseline CAPs to determine whether fast acting K+ channels contributed to noted decrease in the CAP area, and the CAP amplitude with increased IOP in the young or with aging in the older mice as result of dysmyelination.
Optic nerve function was monitored quantitatively as the area under the supramaximal CAP. The CAP area is proportional to the total number of excited axons and represents a convenient and reliable means of monitoring optic nerve axon function (Cummins et al., 1979; Stys et al., 1991). CAP recovery was measured by determining residual CAP area, normalized to control CAP area, 4–6 h after the conclusion of OGD. Data were normalized by setting the mean of initial baseline values (measured over 15 min) to a value of 1.0. Results from several nerves were pooled, averaged, and plotted against time. All data are presented as mean ±SEM. In time-course plots, SEM bars are only shown every 3 min to make the illustration less cluttered; n indicates number of optic nerves used in experiments (Table 2).
Figure 1A shows a box plot of IOP from all DBA/2J mice (6- and 10-month-old) used in the electrophysiological experiments; the low and high IOP groups were significantly different (t-test, p<0.001). Differential IOP allows us to distinguish between changes in pathology attributable to IOP, and thus glaucomatous changes, as opposed to aging in these mice. By calculating the average difference in IOP for low and high IOP groups at 6 and 10 months, we were able to estimate differences in the magnitude of IOP exposure (Table 1). For low versus high IOP groups at 6 months, the high IOP mice have experienced an average of 99.7 mmHg-days of additional intraocular pressure. At 10 months, the difference in IOP exposure between low and high IOP was 528 mmHg-days.
Optic nerves from DBA/2J mice at 6 weeks “control,” at 6 months “young” and at 10 months of age “old” were used to evoke supramaximal compound action potentials (CAPs). Six-week-old DBA/2J mice were used as control mice for the “young” group, because they have the DBA/2J background but do not exhibit signs of glaucomatous changes. Lacking old mice with a DBA/2J background that do not have glaucomatous changes, we used C57BL/6 mice at 9-12 months of age as controls. IOP for C57BL/6 mice ranges between 11.2 ± 3 mmHg (Zhou et al., 2005) and ~13mmHg (Savinova et al., 2001), but is significantly lower than IOP in DBA/2J mice (see Figure 1A). Hence the C57BL/6 mouse serves as a mouse with normal IOP, a control for old DBA/2J mice with either high or low pressure.
In the young mice, the max CAP amplitude was only reduced in the high IOP group (one-way ANOVA, p < 0.05, Figure 1B, Table 2). In the old group, the effect of IOP levels on the max CAP amplitude became more prominent. There was major reduction of the max CAP amplitude in both the low IOP (one-way ANOVA, p < 0.01) and high IOP (one-way ANOVA, p < 0.001) groups compared to old control ONs.
While the max CAP amplitude, measured as the amplitude of the second peak (P2) represents the majority of axons with medium conduction velocity in optic nerve (see below), the CAP area (the area under the entire CAP amplitude curve) provides a reliable measure of the size of the axon population contributing to the CAP (Cummins et al., 1979; Stys et al., 1991). Changes in the CAP area indicate loss of the individual conduction potential contribution from axons in the ON. As observed with max CAP amplitude, the CAP area showed depression with age and IOP levels (Figure 1C). In the young mice, the CAP area was reduced in ONs from the high IOP group compared to the low IOP group (one-way ANOVA, p < 0.001) and control group (one-way ANOVA, p < 0.001). Among old ONs, IOP levels dictated the extent of the CAP area decline. The CAP area was reduced considerably in low IOP (one-way ANOVA, p < 0.05) and high IOP groups (one-way ANOVA, p < 0.01) compared to old control ONs.
Compound action potentials recorded from optic nerve in suction electrode configuration exhibit three peaks (P1, P2, P3); each peak corresponds in general to axons conducting at speeds that are fast (P1), medium (P2) or slow (P3) (Figure 2A). This characteristic of axon conduction is well preserved with age (Baltan et al., 2008) (Figure 2A, lower panel). A change in characteristics of CAPs such as loss of a peak may indicate loss of a group or subset of axons contributing to the CAPs (Geddes 1972). To test how IOP affects physiology, we recorded baseline CAPs from young and old ONs with low and high IOP levels. Figure 2A demonstrates representative CAPs from each of the experimental groups. Young ONs showed modest decline in peak number with increasing IOP (Table 2 and Figure 2A and B). In control mice, all 7 ONs tested exhibited three-peaked CAPs. Figure 2B illustrates this finding with overlapping of P1 (black square), P2 (red circle) and P3 (green triangle) symbols for all 7 CAPs. In this age group, 15 ONs were evaluated with low IOP levels, and the CAPs were evoked in 14, whereas P3 was lost in only one nerve (Figure 2B, note P3 green triangle showing 13 P3s out of a possible 14). Among young high IOP ONs, the CAPs were evoked in 6 out of 7 ONs and P3 was lost only in two CAPs. Aging had no effect on whether the CAPs had three peaks; however, IOP levels gravely affected the CAP shape and presence of peaks in old ONs. Among old low IOP ONs, the CAPs were evoked in 6 out of 8 ONs and P3 was missing in 2 of these ONs. In the high IOP group, although the CAPs were evoked in all nerves tested, only 3 nerves exhibited P2 and none presented with P3 (Figure 2B, note green triangle for P3 at zero).
Our results suggested that the degree of IOP exposure dictated the shape of the CAPs even in the young group. In the old high IOP group, loss of P3 in all ONs and P2 in 3 out of 6 mice suggested that an entire class of slow conducting fibers and a major portion of medium conducting fibers were dysfunctional.
To determine whether glaucomatous changes were associated with conduction delay along optic nerve, the CAP latency, a measure of speed of axon responsiveness to stimulation was assessed. Latency was measured as the time from the end of stimulus artifact to the peak amplitude of P1, P2 and P3. IOP levels did not alter latency to P1, P2 or P3 in the young ONs (Table 2 and Figure 3A) while IOP changes in the old ONs slightly but significantly prolonged latency to P1 between control and high IOP group (Figure 3A, one-way ANOVA, p=0.001). P3 was missing completely in the old high IOP group. Figure 3B shows no differences in the CAP duration across the control and experimental groups. The absence of conduction delay between young and old control groups together with latency changes in old ON with increased IOP implied an IOP-dependent disturbance in optic nerve conduction as a contributing factor in chronic glaucoma.
Application of 4-aminopyrolidine (4-AP), a voltage-sensitive K+ channel blocker, improves abnormal electrophysiological properties, including attenuation of the CAP amplitude and conduction velocity due to demyelination or dysmyelination (Blight, 1989; Nashmi et al., 2000). To rule out whether demyelination contributes to glaucoma-related changes in electrophysiological properties of ONs (Figure 3A), we applied 4-AP to the bath while recording from young ONs with low or high IOP. We recorded baseline CAPs for 30 min and then applied 4-AP and monitored the CAPs for another 30 min. There was no change in the CAP amplitude or the CAP area with 4-AP application (Figure 3C), suggesting that the exposure of K+ channels due to myelination abnormalities do not contribute to electrophysiological CAP changes observed in DBA/2J ONs. Figure 3C shows sample CAP traces from a young low (left) and high IOP (right) DBA/2J ON before and after 4-AP bath application. Application of 4-AP had no effect on control ON baseline responses (data not shown). If the decrease in conduction velocity is not due to demyelination, as suggested by the 4-AP data, then axon integrity or energy availability may be contributing to CAP changes.
Histopathological confirmation of small caliber axon dropout was obtained by quantifying myelinated axon area across plastic embedded ON cross sections taken 100 to 300 microns distal to the optic nerve head. Figure 3D shows a histogram of axon area distribution, comparing ONs with low and high IOP from 5 to 8 months of age. The percentage of total axons with cross-sectional area between 0.4 and 0.8 μm2 decreased in the ONs from mice with high IOP (Figure 3D). These areas correspond to axons with diameters between 0.36 and 0.56 μm, meeting the criteria of small and medium fibers in ONs (Honjin et al., 1977).
To identify metabolic challenges imposed by glaucomatous changes in the optic nerve, we investigated the extent of functional recovery in axons damaged in response to oxygen (60 min) or oxygen-glucose deprivation (OGD, for 45 or 60 min). The ability of axons to use glucose to maintain their function in the absence of oxygen only (anoxia) was identical in ONs from low and high IOP at young ages (data not shown). In contrast, the effects of OGD on axon excitability in ONs obtained from young and old DBA/2J mice with low or high IOP changed significantly. Under normoxic conditions, axon function (quantified as the area under the CAP evoked by supramaximal stimulation) remained stable over many hours at 37°C in all groups. After 1 hr of displaying stable control CAPs, OGD was imposed and resulted in a progressive loss of the CAP area. The extent of irreversible injury detected after exposure to OGD varied as a function of OGD duration, age and IOP levels (Figures 4C, D, E, F). For example, after 45 min of OGD in young ONs, the CAP area recovered to 42.6 ± 4.3 % (n = 7) in control while the CAP area recovered significantly less in low IOP (16.05 ± 3.5%, n=4, p<0.01, one-way ANOVA) and high IOP (22.1 ± 6.3%, n=4, p<0.05, one-way ANOVA) ONs. Similarly, the extent of irreversible axon injury after 60 min OGD was also highly dependent on IOP levels in young ONs (Figure 4A and E). The CAP area recovery in young low IOP and high IOP groups (Figure 4E) was progressively reduced compared to control ONs (35. 6 ± 1.6 %, n = 7 for control, 16. 9 ± 3.1 %, n = 14 for low IOP, one-way ANOVA, p < 0.001 and 9.5 ± 2.2 %, n = 6 for high IOP, one-way ANOVA, p < 0.001). These data suggest that IOP exposure exacerbated OGD-induced axon injury in young ONs. In agreement with this, a scatter plot of the CAP area recovery after 60 min OGD in young ONs showed that IOP reasonably predicted CAP recovery in young DBA/2J mice with high IOP (r2=0.63; Figure 5) but not for DBA/2J mice with low IOP (data not shown). Mice with the highest IOP showed the least recovery after OGD (Figure 4E and and55).
IOP levels did not have a significant effect on the OGD-induced axon injury in old ONs during either 45 min or 60 min OGD. The similar extent of the CAP area recovery in old ONs with low and high IOP compared to control after 45 min OGD suggested that IOP levels had little effect on axon function recovery at this level of OGD (Figure 4D). After 60 min of OGD, regardless of IOP levels, axon function showed lower recovery in ONs from DBA/2J mice compared to control (Figure 4F). The CAP area in ONs from the old control group recovered to 13.6 ± 1.1 % (n = 18) of control (Table 2 and Figure 4B and F), while ONs from old low and old high IOP mice recovered modestly to 3.3 ± 2.4 % ( n = 6, one-way ANOVA, p < 0.01) and 8.7 ± 3.2 % ( n = 6, one-way ANOVA, p > 0.05) of control, respectively (Figure 4B and F). In contrast with the young ONs at 60 min OGD, there was no significant difference between CAP recovery in old low and old high IOP ONs after 60 min OGD, suggesting that IOP levels did not determine the extent of recovery in old DBA/2J ONs. These results suggested that duration of OGD was the determining factor in axon function recovery in old mice.
It should be noted that the CAP area recovery remained stable after OGD among young ONs with low or high IOP, without any significant change over 6 hr of recording in each age group (Figure 4A). Curiously, in the old group with low or high IOP, axon function showed a delayed loss of excitability which became prominent 90 min after the end of OGD (Figure 4B).
The Na+/K+ ATPase maintains the ionic gradients that enable action potential conduction, consuming nearly half of the available ATP in the axon (Erecinska and Silver, 1994). ATP levels were measured in individual optic nerves from 6 and 10-month old mice with low and high IOP. Figure 6A shows the IOP distribution in these mice, with significant differences in IOP across the two age groups, but also between low and high within each age group. ATP levels decreased significantly in ONs from mice with increased IOP. Mice with low IOP had significantly higher ATP levels when compared to high IOP in both age groups, but the effect was especially pronounced in young ONs, suggesting that some mechanism associated with aging also readily depletes ATP levels (Figure 6B). In order to quantify the contribution of aging to ON ATP levels, we measured ATP levels in the two control groups. There was a decrease in ATP levels with aging in the control groups (32.7±12.9%; t-test, ns, p=0.18), but it was exceeded by the decreases that resulted from high IOP as displayed in Figure 6B. There was a 73.5±5.0% decrease in ATP levels in young low vs. young high IOP (t-test, p<0.05), and a 49.3±6.0% decrease in the old low vs. old high IOP group (t-test, p<0.05). Between young and old low IOP ONs, the ATP decrease was 70±6.5% (t-test, p<0.05). Between young and old high IOP ONs, the ATP decrease was 42.5±6.8% (t-test, p<0.05). The magnitude of the ATP decrease with aging in the low IOP group is more than double what would have been expected due to aging alone (70% versus 32.7%). This may be explained by the length of time of IOP exposure (10 months) and the fact that membership in the “low” IOP group at 10 months is due to relative IOP measured at that age. In the high IOP group, aging is responsible for the bulk of the ATP decrease (32.7% of a 42.5% decrease) in ON, but increased IOP still contributes to the decrease.
We tested the hypothesis that glaucoma disrupts optic nerve conduction properties as a function of IOP levels and age in this first electrophysiological characterization of the DBA/2J mice. We found that the amplitude and integral of electrical signals evoked in RGC axons decreased considerably by age 6 months as a function of increasing IOP. We observed that at young ages, raised IOP is directly associated with increased vulnerability to metabolic challenge. Changes in physiological nerve function were accentuated with aging, leading to CAP loss in an entire population of ON fibers—small, slow conducting axons. We confirmed this loss with axon counts and implicated declining metabolic reserve by demonstrating IOP-dependent ATP decrease in ON. These data shed light on a novel potential mechanism of glaucoma pathology whereby increased IOP and declining metabolic capacity lead to axon liability, eventual dysfunction and loss.
The entire RGC axon population contributes to optic nerve function, so the CAP area provides a measure of total axon function. Variations in amplitude and the conduction velocity of axons dictate the three-peaked shape of CAPs. Despite the fact that RGC loss is minimal at 6 months in the DBA/2J model (Buckingham et al., 2008), axon electrophysiological properties were drastically altered. The prominent decrease in the CAP area at 6 months with high IOP demonstrated loss of axon function at an age when RGC somas remained intact. Axon function in this age group was characterized by decreases in the maximum CAP amplitude (P2 amplitude, Figure 1B) and by the beginning of P3 loss in some ONs, particularly in high IOP mice (Figure 2B). These changes were accentuated with age, confirming further axon loss. The onset of P3 loss at 6 months became very prominent by 10 months: CAPs from 4 out of 8 mice lost P3 with low IOP and all mice lost P3 with high IOP. In addition to P3 in the high IOP group, P2 was lost in 3 of 6 mice. Decreases in the CAP amplitude by six months are the first evidence that axon dysfunction occurs upstream of axonal transport deficits. These findings demonstrated that slow conduction fibers are the first to display functional decline in glaucoma, although with age and increased IOP, medium-sized fibers also decline. Such alterations in the shape and conduction properties of CAPs are sufficient to cause interruption of visual information transfer in glaucoma.
Physiological loss of small and medium caliber axons means absence of visual information from subclasses of RGCs. RGC morphological analyses showed a strong correlation between soma diameter and axon diameter across RGC subtypes (Huxlin and Goodchild, 1997; Coombs et al., 2006). This, combined with highly detailed studies of rodent RGC subtypes (Huxlin and Goodchild, 1997; Combs et al., 2000; Sun et al., 2002; Badea and Nathans, 2004; Kong et al., 2005; Volgyi et al., 2009) provides insight into which RGCs undergo conduction failure with increased IOP in the DBA/2J mice. The RGCs with the smallest axons (as measured 100 to 300 microns closest to the soma) have been classified as clusters M1 and M11 (Coombs et al., 2006) or groups RGB2 and RGC1 (Huxlin and Goodchild, 1997). The M1 RGCs are monostratified, (Coombs et al., 2006) similar to classes B2 and B4 (Sun et al., 2002) and Cluster 1 (Badea and Nathans, 2004), while the M11 cluster is bistratified, similar to class B2 (Sun et al., 2002) and Badea & Nathans Cluster 4 (Coombs et al., 2006). These cells have small somas, small proximal axons and smaller dendritic fields compared to other RGCs. Most of these small axon bearing RGCs have dendritic arbors stratified in the middle (~50%) (Badea and Nathans, 2004; Kong et al., 2005; Coombs et al., 2006; Volgyi et al., 2009) to outer (~35%) (Sun et al., 2002) IPL. The M1 and M11 cluster cells (Coombs et al., 2006) have axon diameters that are roughly 0.65 μm, which corresponds to axon areas of ~0.33 μm2, a size within the range of axons that decrease in DBA/2J ON with high IOP (0.25 to 0.8 μm2) (Figure 3D).
Loss of P2, indicative of medium fiber conduction decline, occurred with high IOP in old ONs (Figure 2B). This fiber caliber corresponds to the majority of RGC subtypes, limiting our ability to implicate specific cell types in these physiological changes.
In humans, evidence suggests preferential loss of large RGC axons. The optic nerve areas most susceptible to axon loss were superior and inferior central optic nerve, and within these regions, a greater proportion of larger axons were lost (Quigley et al., 1988). A comparison of the proportion of surviving fibers in the most and least damaged nerves from laser trabeculotomy-induced glaucoma in monkey showed few large fibers in the most damaged nerves, but also greater numbers of small fibers when compared to the least damaged (Quigley et al., 1987). In the perifoveal region in retina, there was no preferential RGC loss (Quigley et al., 1989). Differences in modeling may account for the disparate observations regarding axon populations lost in glaucoma, but the simplest explanation might lie in morphology versus physiology. Axons may no longer conduct action potentials yet remain in the optic nerve. For example, slow conducting fibers ceased to function prior to larger, faster fibers after optic nerve stimulation in cats whose IOP equaled mean arterial blood pressure (Grehn and Prost, 1983). Low perfusion pressures could halt all RGC axon conduction (Grehn and Prost, 1983). In addition, there is evidence that some RGCs are more physiologically sensitive to increased IOP (Zhou et al., 1994).
By demonstrating physiological dropout of slow conducting axons, our data argues against mechanisms of glaucoma that would predict global loss of axons. Physical infringement on axons within the optic nerve head (ONH) is one such hypothesis, predicting axon compromise across all classes of conduction. Our data shows small fibers are preferentially affected with glaucoma. As the disease progresses, medium size fibers are recruited by the injury process. This specificity elminates indiscriminate pathological mechanisms. A slightly larger percentage of the smallest RGC axons reside more centrally in the ON (Honjin et al., 1977), but it is unclear if loss of these specific axons would account for loss of P3 in the CAP. With only 1.2 percent of the ON unmyelinated (Honjin et al., 1977), P3 includes small, lightly myelinated fibers and unmyelinated axons. Unmyelinated fibers have higher energy requirements given the enormous levels of ATP required to sustain action potentials without saltatory conduction. Axons with high ATP requirements (Bristow et al., 2002) have many more mitochondria per unit length, but these axons would also be preferentially targeted by low ATP conditions. The unique finding of the sensitivity of small caliber axons suggests that high metabolic demand is a key predictor of axon vulnerability to raised IOP.
Axons need a constant supply of oxygen and glucose to maintain moment-to-moment function. The rate of axon function loss is closely correlated to the depletion of energy substrates and ATP (Swanson et al., 1989; Wender et al., 2000). Furthermore, OGD mainly disrupts oligodendrocytes and axons (Tekkok et al., 2007) indicating that myelinated axon function is intimately correlated to energy status in ONs. Increased IOP exposure determined the CAP recovery after 60 min of OGD in young ONs (Figure 4C and D), but recovery was determined primarily by age in old ONs (Figure 4F). Despite the apparent lack of significant impact of IOP on CAP recovery, we did observe a significant effect of IOP on ATP levels in the DBA/2J ON.
Many aspects of ON function change with aging, and we measured ATP levels in ONs to determine if IOP would have an effect on energy availability in the ON. Young and old control ONs showed that aging accounts for ~33 % of ATP level decreases. Quantification of ATP levels in young and old DBA/2J ONs showed a considerable reduction in ATP levels with increased IOP (Figure 6). At 6 months of age in the DBA/2J, 100% of mice have at most mild axon degeneration (2% or fewer axons lost) in the ON. By 10 months of age, a minority of DBA/2J mice (roughly 30%) have greater than 50% axon loss (Anderson et al., 2005; Libby et al., 2005b). A ON has roughly 50,000 axons. If an aged cohort had 80 percent axon loss in 30 percent of the mice, that is an average decrease of 28 percent (14,000 axons) overall. With this most extreme scenario, the number of axons lost on average in the 10 month DBA/2J would not account for the observed decreases in ATP levels. Additionally, the significant differences in ATP levels between low and high IOP groups reinforces the effect that glaucomatous changes exert on ON energy availability. These data, combined with the OGD results, demonstrate that a key, early pathological feature in glaucoma is the presence of a lower bioenergetic reserve in optic nerve that correlates with their exposure to raised IOP. Both high IOP and age contributed to lower ATP in ON in a fashion similar to the CAP area and the CAP amplitude findings. Besides having a significant, negative impact on axon transport, low ATP indicates fundamental problems with mitochondrial function. Mitochondrial dysfunction has been implicated in several mechanisms of neurodegeneration, including glaucoma (Tezel, 2006) and optic neuropathy (Yu Wai Man et al., 2005). These data argue for further studies on axon physiology and bioenergetics and suggest that interceding before damage to metabolic machinery or boosting metabolic substrates in the optic nerve will likely have therapeutic value.
This work was supported by the American Heart Association National Scientist Development Grant [to S.B.]; the Glaucoma Research Foundation and The Melza M. and Frank Theodore Barr Foundation [to P.J.H and D.J.C]; and the National Institutes of Health [NS35533 and NS056031 to R.S.M.].