July 2008
Volume 49, Issue 7
Free
Glaucoma  |   July 2008
Effect of Repeated IOP Challenge on Rat Retinal Function
Author Affiliations
  • Zheng He
    From the Department of Optometry and Vision Sciences, University of Melbourne, Parkville, Victoria, Australia.
  • Bang V. Bui
    From the Department of Optometry and Vision Sciences, University of Melbourne, Parkville, Victoria, Australia.
  • Algis J. Vingrys
    From the Department of Optometry and Vision Sciences, University of Melbourne, Parkville, Victoria, Australia.
Investigative Ophthalmology & Visual Science July 2008, Vol.49, 3026-3034. doi:10.1167/iovs.07-1628
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Zheng He, Bang V. Bui, Algis J. Vingrys; Effect of Repeated IOP Challenge on Rat Retinal Function. Invest. Ophthalmol. Vis. Sci. 2008;49(7):3026-3034. doi: 10.1167/iovs.07-1628.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. To characterize the effect of repeated brief intraocular pressure (IOP) elevations and the effect of IOP fluctuation on retinal function.

methods. The effects of one, two, and four episodes (70 mm Hg, 15 minutes) are compared by defining the time course of functional recovery after insults. The effect of IOP variation is considered by comparing a constant with a varying insult, keeping a common IOP-time integral (one 60-minute vs. two 30-minute vs. four 15-minute insults; 70 mm Hg). IOP elevation is induced by anterior chamber cannulation in anesthetized, dark-adapted rats (n = 5–7 per group). Electroretinograms are recorded every 6 minutes throughout each event. Recovery time course is modeled using a logistic function, and time for 50% recovery is compared by nonparametric bootstrap.

results. Electroretinographic recovery becomes progressively slower with more IOP episodes for bipolar cell and ganglion cell response (P < 0.05) but not for photoreceptor response (P > 0.05). With regard to IOP variation, bipolar cell recovery after four 15-minute insults is faster than it is after two 30-minute insults (P < 0.05), which is faster than after one 60-minute insult (P < 0.05). Ganglion cell recovery after varying (four 15-minute and two 30-minute) insults is faster than after a constant (one 60-minute) insult (P < 0.05). This improved recovery with varying IOP challenge is greater for bipolar cell than for ganglion cell responses (P < 0.05).

conclusions. Repeated IOP insults lead to cumulative dysfunction in the inner retina. For the conditions used in this study, IOP variation per se is not detrimental but appears to be beneficial.

Retinal neurons can be rapidly compromised by acute elevation of intraocular pressure (IOP). 1 2 The number of insults is important in determining the neuronal consequences. Neural damage was greater in gerbil 3 4 5 and rat 6 7 brains exposed to repeated ischemia than to a single insult. In isolated rat retina, an ex vivo study 8 has shown that seven pressure insults (50 mm Hg × 1 minute, interval 1 minute) resulted in greater ganglion cell loss than a single episode (50 mm Hg × 1 minute). However, this has yet to be confirmed in the intact eye. 
Other attributes of repeated challenge, such as a greater degree of fluctuation, may also promote worse outcomes. In particular, fluctuation in oxygen breathing, 9 10 11 blood glucose, 12 13 intracranial pressure, 14 15 and IOP 16 17 are thought to promote greater neurologic deficits. In spite of these epidemiologic findings, no experimental evidence exists to support the theory that the fluctuation component of IOP insults promotes retinal dysfunction. On the contrary, a number of studies show that brief insults can protect tissue for subsequent insults, which is known as ischemic preconditioning. 18 19 Therefore, the effect of multiple IOP insults may reflect a balance between cumulative dysfunction and preconditioning. Moreover, this balance may not be the same for different retinal cell classes. Evidence for preconditioning has been identified in terms of improved bipolar cell function after repeated IOP insults, 20 whereas benefits for ganglion cells has been limited to anatomic findings. 21 22 The presence of an anatomic benefit for ganglion cells without functional support is likely to reflect the fact that in vivo ganglion cell function has, to date, been difficult to assess. Alternatively, there may be a real difference, in terms of the balance between detrimental and preconditioning effects of repeated insults, for bipolar and ganglion cells. To resolve this controversy, the effect of multiple insults must be assessed on various retinal cell classes in the same eye. 
Given these factors, the aims of this study were to assess the effect of the number of insults on retinal function and to consider the effect of IOP variation by comparing a constant with a varying insult of equivalent IOP-time integral (IOP × duration). Should IOP fluctuation independently compromise cellular integrity, a varying insult would be more detrimental than a sustained one. Alternatively, if preconditioning is dominant, the converse will be true. Finally, we assessed photoreceptor, bipolar, and ganglion cell recovery to establish whether all retinal neurons show similar responses to repeated IOP challenge. 
Materials and Methods
Animals
All procedures were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Approval was obtained from our departmental Animal Ethics Committee (A05003). Adult Long-Evans rats (age range, 8–12 weeks; weight range, 200–300 g) were housed in a 20°C environment (12-hour light/12-hour dark, 50 lux maximum) with food and water freely available. 
IOP Challenge
Dark-adapted (≥12 hours) rats were anesthetized by intramuscular injection of ketamine/xylazine (60:5 mg/kg; Troy Laboratories Pty Ltd., Smithfield, NSW, Australia). Maintenance doses of 50% the original were given at 50-minute intervals. Corneal analgesia and mydriasis was achieved with one drop of proxymetacaine (0.5%; Alcain; Alcon Laboratories, Frenchs Forest, NSW, Australia) and tropicamide (0.5%; Mydriacyl; Alcon Laboratories), respectively. A circulating water heat pad maintained body temperature at 37°C. After electrode placement, the anterior chamber of a randomly selected eye was cannulated using a 30-gauge needle connected through polyethylene tubing (inner diameter, 0.38 mm) to a pressure transducer (Transpac; Abbott Critical Care Systems, Sligo, Ireland), and Hanks balanced salt solution reservoir (JRH Biosciences, Lenexa, KA). IOP was controlled by manipulating the reservoir elevation to precalibrated heights, referenced to an aneroid sphygmomanometer (Livingstone, Rosebery, NSW, Australia). IOP was monitored throughout the experiment (Powerlab 8SP amplifier, Chart software v5.3; ADInstruments, Castle Hill, NSW, Australia). On cannulation, baseline IOP was set to 13 mm Hg, the average for control rats (n = 5) measured manometrically. This is consistent with levels reported in anesthetized rats. 23 All procedures were conducted under dim red light (λmax = 600 nm) to minimize light adaptation. One hour of complete darkness was allowed for stabilization and readaptation before IOP elevation and electroretinogram (ERG) recordings. 
Throughout this study, IOP was elevated from baseline (13 mm Hg) to 70 mm Hg. First, we investigated the effect of increasing the number of insults by comparing one (n = 6), two (n = 6), and four 15-minute insults (n = 5). Second, we considered the effect of IOP fluctuation by comparing a sustained (one 60-minute insult, n = 6) with a varying IOP elevation (two 30-minute insults, n = 7; and four 15-minute insults, n = 5), keeping a common IOP-time integral (4200 mm Hg/min–1). We adopted a fixed interval between insults (30 minutes in total), which was 30 minutes for the two 30-minute insults and 10 minutes for the four 15-minute insults (3 × 10 = 30 minutes). Data for single 15- and 60-minute insults have been reported previously 2 and are reanalyzed with permission from ARVO. 
Systolic blood pressure (BP) was monitored before, during, and after IOP elevation, using tail cuff sphygmomanometry (ML125; ADInstruments), which was calibrated against femoral artery cannulation with an accuracy of −0.2 mm Hg and a precision of ±11.4% (coefficient of variation). 
Electroretinography
As previously described, signals were recorded with custom-made silver-chloride electrodes before, during, and immediately after IOP elevation. 2 The active electrode was mounted in the center of the cornea, whereas a ring-shaped scleral inactive electrode was located near the equator of the eye. The ground electrode was a stainless steel (F-E2–30; Grass Telefactor, West Warwick, RI) needle inserted subcutaneously into the tail. Cornea hydration was maintained with 1.0% carboxymethylcellulose sodium (Celluvisc; Allergan, Irvine, CA). 
Stimuli were brief (1 ms) white flashes (Luxeon 5-W LEDs, 5500°K; Calgary, Alberta, Canada) delivered by way of a ganzfeld sphere (Photometric Solutions International, Huntingdale, Victoria, Australia). Luminous energy was calibrated using a photometer (IL1700; International Light Research, Peabody, MA) with a scotopic filter (Z-CIE) in place. After cannulation and readaptation (60 minutes), baseline responses were captured that consisted of a scotopic threshold response (STR) series (30 signals averaged, interstimulus interval 2 seconds, −4.95 log cd · s/m2) and a single bright flash response (1.01 cd · s/m2). During and after IOP insult, an STR followed by a bright flash was recorded every 6 minutes (up to 60 minutes for one 15-minute insult; 90 minutes for two 15-minute insults; 120 minutes for all other insults). Results from pilot studies showed that a 6-minute interval between the bright and the dim flash was adequate to allow complete STR recovery from the previous bright flash. These luminous energies therefore represent a balance between our desire to saturate the photoreceptor response 24 and to maximize the sampling rate. All signals were amplified (×1000), recorded with bandpass settings of 0.3 to 1000 Hz (−3 dB), and digitized at 4 kHz. 
Electroretinographic Analysis
Photoreceptor Response.
The leading edge of the photoreceptor (P3) a-wave can be modeled with a delayed Gaussian, 25 as described in our previous paper. 2 This model returns the maximum saturated amplitude (Rm P3; μV), which indicates the total number of nonspecific cationic channels closed by light activation, whereas the sensitivity (S; m2 · cd−1 · s−3) represents the amplification of all phototransduction stages. 26 P3 modeling was achieved by floating Rm P3 and S (with t d fixed to the average of control data, t d = 4.6 ms) and minimizing the sum-of-squares (SS) merit function for data points between t d and the a-wave minimum (or a maximum of 18 ms), using a software module (Excel Solver; Microsoft, Redmond, WA). 
ON-Bipolar Cell Response.
The postreceptoral P2 response has been shown to better describe the bipolar cell response than the raw b-wave. 25 The P3 is digitally subtracted from the raw ERG to expose the P2-OP (oscillatory potential) complex, which is then low-pass filtered (47 Hz, −3 dB) to remove the OPs. The P2 amplitude was measured at a fixed criterion time of 70 ms after the stimulus. 
Scotopic Threshold Response.
The STR has been shown to contain ganglion cell contributions in rats 27 and is more sensitive to IOP insults than other electroretinographic components. 1 2 As previous described, 1 2 we found that the negative STR component (nSTR) was the most sensitive component to acute IOP elevation (Fig. 1) . Thus, the STR was analyzed by measuring amplitude at a fixed criterion time of 220 ms after stimulus onset. 
Statistical Analysis
All ERG parameters were found to be normally distributed (Kolmogorov-Smirnov normality statistic) with equal variance (Barlett test, Prism v4.00; GraphPad Software Inc., San Diego, CA). One-way, repeated measures analysis of variance (RM ANOVA; Prism v4.00) was applied to test the null hypothesis that there was no cumulative effect of repeated IOP insults (α = 0.05). 
Functional recovery was assessed by recording the ERGs after IOP was returned to baseline (13 mm Hg). To compare recovery among different groups and to account for intergroup variability, ERG amplitudes were normalized (see Figs. 3D 4D 6D 7D ) to the average of the final three recordings. This procedure was based on several factors. First, the recovery of each ERG component always reached a stable level during the final three recordings (P > 0.05, one-way RM ANOVA). Second, retesting two animals 1 week after four 15-minute insults indicated that full functional recovery was achieved (Fig. 1 , bottom row). Therefore, the responses at the end of the acute challenge protocol reflected complete recovery within the limits of inter-animal variability. 
The time course of amplitude recovery was modeled using a logistic function. 2 The key parameter from this model is the time taken to achieve 50% amplitude (t 0.5, min). Given the normalization described previously, our model presumes 100% recovery. Parameter optimization was achieved by floating the slope “b” and “t 0.5” and minimizing the SS merit function using the solver module (Excel Solver; Microsoft). Parameter variability was established using a nonparametric bootstrap. 28 Group data at each time point are randomly sampled with replacement to derive a bootstrap data set. 29 Each bootstrap data set is then modeled with the logistic function, yielding a set of bootstrap parameters. By repeating this process (1000 times), a nonparametric distribution of the parameters provided the 2.5% and 97.5% confidence limits (95% CI). Therefore, t 0.5 for different IOP insults was compared with an α of 0.05. 
Results
To avoid confounding the effect of IOP on retinal function with any change in blood pressure (BP), it is important to establish that BP was stable throughout our experiments. The systolic BP (±SEM) was sampled before (108 ± 3 mm Hg, n = 20), during (109 ± 8 mm Hg, n = 9) and after (104 ± 4 mm Hg, n = 9) IOP elevation. Consistent with our previous study, 2 systolic BP was not altered over the time course of our experiments (F 2 = 0.17, P = 0.85). Moreover, the interaction term of systolic BP between the IOP treatment groups (constant vs. varying IOP insults) was not statistically significant (F 1,1 = 0.36; P = 0.56, two-way ANOVA). 
Single versus Multiple Insults
Figure 1shows representative data for repeated (four 15-minute) IOP insults. Waveforms were serially recorded to a bright (Fig. 1A ; 1.01 log cd · s/m2) and a dim (Fig. 1B ; −4.95 log cd · s/m2) flash. Before IOP elevation, response to the bright flash (Fig. 1A , thin traces) showed a typical a-wave trough (P3), followed by the dominant b-wave peak (P2). The STR (Fig. 1B , thin traces) waveform showed characteristic positivity at approximately 110 ms and negativity at approximately 220 ms. Baseline ERGs (thin traces) have been overlaid with waveforms recorded during and after IOP insults (thick traces). 
ERG waveforms became progressively attenuated with repeated insults and gradually recovered. In particular, the b-wave was reduced during the first insult and was possibly further abolished during the fourth insult. Similarly, the STR was progressively reduced with repeated insults. It is notable that the nSTR was completely suppressed during the first insult. After IOP normalization (example at Fig. 1 , third row, 18 minutes), a-wave recovery was faster than b-wave recovery, which in turn was faster than the nSTR. The delayed pSTR implicit time during recovery (Fig. 1B , third row, arrowhead) may arise from one of two possibilities. First, the early recovery of the “positivity” (Fig. 1B , third row) may be mediated by P2 recovery, which has a slower implicit time than pSTR. 30 The fact that the b-wave recovers more rapidly than the STR (right vs. left panels) lends credibility to this prospect. Second, a slower nSTR recovery relative to the pSTR may contribute to a relatively larger positivity and an apparent delay in implicit time. The bottom row of Figure 1shows that both the bright flash response and the STR have completely recovered 1 week after the four IOP insults. 
Figure 2shows mean photoreceptor amplitude (Rm P3 , ±SEM) over the time course of the one, two, and four 15-minute insults (Figs. 2A 2B 2C , respectively). The effect of repeated insults was considered by comparing Rm P3 during each of the four 15-minute insults (Fig. 2C , filled squares, 9 minutes after onset of each insult). Although Rm P3 declined from −398 ± 33 μV to −248 ± 56 μV during the first and fourth insults, respectively, this trend was not significant (F 3,12 = 1.74; P = 0.21). This negative outcome may reflect variability in our Rm P3 measures. After IOP normalization, Rm P3 recovery was rapid for all insults, with amplitude returning to within the 95% CI of its respective baseline levels immediately (1–2 minutes). This rapid recovery precludes logistic modeling, and a higher sampling rate (greater than every 6 minutes) is needed to reveal P3 recovery dynamics. Therefore, the P3 is not considered further. 
Figure 3shows the mean (±SEM) P2 amplitude in response to various insults. During each of the four 15-minute insults (Fig. 3C) , P2 amplitude was almost completely abolished. We were unable to find a significant cumulative effect for these consecutive insults (F 3,12 = 1.93; P = 0.18; Fig. 3C , filled squares). Therefore, the potential for a cumulative effect was further considered by comparing recovery after the single, two, and four 15-minute insults (Figs. 3Avs. 3Bvs. 3C ). The t 0.5 returned from the best fit logistic model (Fig. 3D)is summarized in Figure 3E , along with the bootstrap 95% CI. Although the t 0.5 for a single 15-minute insult (1.8 [95% CI: 0.6, 2.5] min) was not different from that of two 15-minute insults (1.8 [95% CI: 0.4, 3.5] min; P > 0.05), it was significantly faster than that observed after four 15-minute insults (3.3 [95% CI: 1.6, 5.7] min; P < 0.05). 
Recovery of ganglion cell response immediately after each of the four 15-minute insults (Fig. 4C , filled squares, 6-minute recovery) was not significantly different (F 3,12 = 0.54; P = 0.66), because the nSTR reduction was very severe in all cases. Slower nSTR recovery with repeated insults becomes evident when the normalized data are compared (Fig. 4D) . Recovery was significantly slower as the number of insults accumulated, with a gradient in t 0.5 (Fig. 4E)progressing from 15.1 (95% CI: 13.4, 16.6) for a single insult, 22.9 (95% CI: 20.8, 25.6) for two insults, to 46.4 (95% CI: 38.2, 56.8) minutes (bootstrap P < 0.05) for four insults. These findings are not surprising given that the total IOP-time integral increases with each additional insult. Therefore, the slower recovery reflects a greater insult integral. 
Constant versus Varying Insult
Figure 5compares ERG waveforms during recovery from a single 60-minute and four 15-minute insults. Baseline waveforms (thin traces) are overlaid with those recorded at 1, 12, and 30 minutes after IOP restoration (thick traces). It is clear that function recovered at different rates after varying and constant insults. Specifically, b-wave recovery (Fig. 5A)was faster after four 15-minute insults than after a single 60-minute insult. The difference is less apparent for nSTR recovery (Fig. 5B) , with only a slightly faster nSTR recovery after four 15-minute versus a single-60 minutes insult (see bottom row). The general impression from the representative waveforms (Fig. 5)is that repeated insults produce faster recovery than a sustained insult. 
Figure 6A 6B 6Cshows that P2 amplitude was quickly and completely suppressed during all IOP insults. Differences between insults become apparent when P2 recovery is normalized. Figure 6Dshows that the fastest P2 recovery occurs after the four 15-minute insults (squares, solid thin curve) and the slowest after the single 60-minute insult (unfilled circles, gray curve). This is confirmed in Figure 6E , which shows that t 0.5 is significantly faster for the four 15-minute insults (3.3 [95% CI: 1.6, 5.7] min), followed by the two 30-minute insults (8.6 [95% CI: 4.5, 12.1] min), and finally the single 60-minute insult (16.7 [95% CI: 12.9, 17.2] min; P < 0.05). 
Figure 7shows that nSTR amplitude (mean ± SEM) recovers at similar rates after restoration of IOP for the various insults. The amplitudes during recovery were normalized and modeled in Figure 7D . The t 0.5 (Fig. 7E)was slightly longer for the single 60-minute insult (55.6 [95% CI: 49.4, 58.0] min) than for the two 30-minute (46.5 [95% CI: 41.2, 52.3] min) and the four 15-minute (46.4 [95% CI: 38.2, 56.8] min) insults. However, no significant difference (P > 0.05) was found between the later two treatments. 
Discussion
Accumulation of Repeated Challenge
This study considers the role that multiple, brief IOP insults can have on retinal neurons. The effect of one, two, and four 15-minute IOP insults on photoreceptor amplitude (Rm P3 ) was not severe enough to differentiate the insults from each other (Fig. 2) . In contrast, the postreceptor components (P2 and nSTR) were better indicators of multiple challenge, consistent with a greater susceptibility of the inner retina to IOP elevation compared with the outer retina. 1 2 31 32  
Functional recovery after IOP insults becomes slower as the number of insults increases (Figs. 3 4) . This outcome provides in vivo evidence that repeated IOP challenge temporarily exacerbates neuronal dysfunction with recovery. Anatomic damage has been reported after repeated pressure insult in an ex vivo model. 8 In contrast, Kim et al. 33 did not find any cumulative b-wave dysfunction after three IOP episodes. This negative outcome may arise because the b-wave (P2) shows less cumulative compromise compared with our STR measure. Figure 8Ashows that functional recovery becomes further delayed with more insults. However, this effect is less pronounced for the P2 (unfilled) than the nSTR (filled; P < 0.05). Alternatively, Kim et al. 33 used a longer (1 hour) interval between IOP challenges, which allowed full b-wave recovery before subsequent insults. Our short intervals (10 minutes) provided only partial b-wave recovery before the next challenge (Fig. 3C) . Short intervals may place greater stress on cellular reserves needed for recovery, thereby fostering cumulative dysfunction. Consistent with this possibility, our data for two insults (Fig. 3B) , which had a longer interval and allowed full recovery, showed no cumulative effect (Fig. 3A)
It is unclear whether the IOP-related dysfunction observed in this study occurred through ischemic or mechanical processes. It is thought that both factors may occur simultaneously at high IOP because studies 34 35 have shown that blood vessels and axoplasmic flow are compromised by IOP-induced deformation of the lamina cribrosa. According to the ischemic theory, ocular perfusion becomes compromised when IOP reaches a level that disrupts normal retinal blood flow, producing an ischemic challenge. 36 On the other hand, high pressure may act directly on retinal neurons. Stretch-activated channels on ganglion cells have been shown to open in response to mechanical force (e.g., IOP), triggering potassium efflux. 37 Moreover, mechanical stretching of neuronal membranes can activate N-methyl d-aspartate receptors, 38 which, in turn, increases intracellular calcium. Although calcium influx can lead to neuronal apoptosis in longstanding or severe IOP elevation, 39 we did not find any sustained functional loss (Fig. 1 , bottom row). We postulate that exposure to four 15-minute insults is too brief to induce apoptosis. Specifically, any changes in intracellular potassium and calcium may be too low to induce apoptosis but still high enough to dampen cellular excitability and glutamate release. Such a mechanism has been recently described in cultured midbrain dopaminergic neurons after 5 minutes of ischemia. 40 This mechanism will minimize the possibility for glutamate excitotoxicity at the expense of neurotransmission, thus manifesting as a transient functional deficit. 
Effect of IOP Variation
Figure 8Ademonstrates that neuronal dysfunction accumulates after repeated exposure, which is more evident for ganglion than bipolar cell responses. What is not apparent is whether it is the IOP-time integral or the insult variation that produces this effect. This is considered in Figure 8B , which shows that a prolonged single 60-minute insult is more detrimental than the two 30-minute and four 15-minute insults. This provides evidence that insult variation per se is not detrimental to retinal neurons for the conditions used in our experiment. To our knowledge, this is the first study to separate the effect of variation from the IOP-time integral because we kept total duration and IOP magnitude constant. 
We found that repeated insults compromised neuronal integrity in an additive manner as the IOP-time integral increased (Fig. 8A) . However, the variation component of the insult was not detrimental (Fig. 8B) . The latter finding contradicts the theory that IOP variation is an independent risk factor for ganglion cell damage. 16 17 41 However, caution is needed when comparing our findings with those of clinical studies because our study investigated acute insults at a relatively high IOP. 
P2 and nSTR amplitudes recovered more rapidly after the varying (four 15-minute) insult than after the constant (single 60-minutes) challenge, indicating that IOP variation can be beneficial. P2 recovery was faster after four 15-minute insults than after two 30-minute insults even though the total interval between insults was the same (3 × 10 = 30 minutes). This is consistent with data showing that preconditioning is best achieved with brief (5–8 minutes) and sublethal episodes of ischemia. 20 42  
The most interesting outcome of our work was the selective vulnerability and the limited benefit of preconditioning to ganglion cell function. In comparison to bipolar cells, ganglion cell function was affected more by repeated insults (Fig. 8A ; P < 0.05), and benefited less from preconditioning (Fig. 8B ; P < 0.05). 
In summary, this study demonstrates that neuronal dysfunction accumulates with the number of IOP insults. Insult variation per se is not detrimental but produces a capacity for preconditioning that is more evident in bipolar cells than ganglion cell responses. Thus, ganglion cells are likely to be more susceptible to repeated IOP elevation. 
 
Figure 1.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms in response to the four 15-minute IOP insults. Thin traces: baseline ERG before insult. Thick traces: responses during and after insults. First and second rows recorded 9 minutes after onset of the first and fourth insult, respectively. Third and fourth rows recorded 18 minutes and 1 week after insults, respectively. Vertical line: pSTR implicit time for baseline ERG. Arrowheads: pSTR implicit time for IOP-treated waveforms.
Figure 1.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms in response to the four 15-minute IOP insults. Thin traces: baseline ERG before insult. Thick traces: responses during and after insults. First and second rows recorded 9 minutes after onset of the first and fourth insult, respectively. Third and fourth rows recorded 18 minutes and 1 week after insults, respectively. Vertical line: pSTR implicit time for baseline ERG. Arrowheads: pSTR implicit time for IOP-treated waveforms.
Figure 2.
 
Effect of repeated insults on photoreceptor response (Rm P3 , mean ± SEM). The effects of one, two, and four 15-minute insults are shown (A, unfilled circles, n = 6; B, filled circles, n = 6; C, squares, n = 5). Horizontal dashed lines: mean baseline level. Shaded area: duration of IOP elevation. Filled squares: Data recorded 9 minutes after onset of each insult.
Figure 2.
 
Effect of repeated insults on photoreceptor response (Rm P3 , mean ± SEM). The effects of one, two, and four 15-minute insults are shown (A, unfilled circles, n = 6; B, filled circles, n = 6; C, squares, n = 5). Horizontal dashed lines: mean baseline level. Shaded area: duration of IOP elevation. Filled squares: Data recorded 9 minutes after onset of each insult.
Figure 3.
 
Effect of repeated insults on ON-bipolar cell response (P2, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 9 minutes after onset of each IOP insult. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final baseline. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 3.
 
Effect of repeated insults on ON-bipolar cell response (P2, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 9 minutes after onset of each IOP insult. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final baseline. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 4.
 
Effect of repeated insults on ganglion cell response (nSTR, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 7 minutes after each IOP normalization. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 4.
 
Effect of repeated insults on ganglion cell response (nSTR, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 7 minutes after each IOP normalization. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 5.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms after single 60-minute and four 15-minute insults. Thin traces: baseline ERG before insult. Thick traces: responses after insult.
Figure 5.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms after single 60-minute and four 15-minute insults. Thin traces: baseline ERG before insult. Thick traces: responses after insult.
Figure 6.
 
Effect of insult variation on ON-bipolar cell response (P2, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 6.
 
Effect of insult variation on ON-bipolar cell response (P2, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 7.
 
Effect of insult variation on ganglion cell response (nSTR, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 7.
 
Effect of insult variation on ganglion cell response (nSTR, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 8.
 
Relative t 0.5 (±95% CI) for ON-bipolar cell (P2, unfilled symbols) and ganglion cell responses (nSTR, filled symbols) after multiple insults, normalized to t 0.5 returned after a single insult. (A) Comparison of normalized data between single and multiple insults (Figs. 3Evs. 4E ). (B) Comparison of normalized data between constant and varying insults (Figs. 6Evs. 7E ).
Figure 8.
 
Relative t 0.5 (±95% CI) for ON-bipolar cell (P2, unfilled symbols) and ganglion cell responses (nSTR, filled symbols) after multiple insults, normalized to t 0.5 returned after a single insult. (A) Comparison of normalized data between single and multiple insults (Figs. 3Evs. 4E ). (B) Comparison of normalized data between constant and varying insults (Figs. 6Evs. 7E ).
BuiBV, EdmundsB, CioffiGA, FortuneB. The gradient of retinal functional changes during acute intraocular pressure elevation. Invest Ophthalmol Vis Sci. 2005;46:202–213. [CrossRef] [PubMed]
HeZ, BuiBV, VingrysAJ. The rate of functional recovery from acute IOP elevation. Invest Ophthalmol Vis Sci. 2006;47:4872–4880. [CrossRef] [PubMed]
TomidaS, NowakTS, Jr, VassK, LohrJM, KlatzoI. Experimental model for repetitive ischemic attacks in the gerbil: the cumulative effect of repeated ischemic insults. J Cereb Blood Flow Metab. 1987;7:773–782. [CrossRef] [PubMed]
VassK, TomidaS, HossmannKA, NowakTS, Jr, KlatzoI. Microvascular disturbances and edema formation after repetitive ischemia of gerbil brain. Acta Neuropathol (Berl). 1988;75:288–294. [CrossRef]
KatoH, LiuY, ArakiT, KogureK. Temporal profile of the effects of pretreatment with brief cerebral ischemia on the neuronal damage following secondary ischemic insult in the gerbil: cumulative damage and protective effects. Brain Res. 1991;553:238–242. [CrossRef] [PubMed]
NakanoS, KatoH, KogureK. Neuronal damage in the rat hippocampus in a new model of repeated reversible transient cerebral ischemia. Brain Res. 1989;490:178–180. [CrossRef] [PubMed]
KimuraR, ShiinoA, MatsudaM, et al. Acute focal cerebral ischemia in rats studied by diffusion-weighted magnetic resonance imaging—an experimental study. Surg Neurol. 1996;46:67–74. [CrossRef] [PubMed]
RestaV, NovelliE, VozziG, et al. Acute retinal ganglion cell injury caused by intraocular pressure spikes is mediated by endogenous extracellular ATP. Eur J Neurosci. 2007;25:2741–2754. [CrossRef] [PubMed]
SaitoY, OmotoT, ChoY, et al. The progression of retinopathy of prematurity and fluctuation in blood gas tension. Graefes Arch Clin Exp Ophthalmol. 1993;231:151–156. [CrossRef] [PubMed]
CunninghamS, FleckBW, EltonRA, McIntoshN. Transcutaneous oxygen levels in retinopathy of prematurity. Lancet. 1995;346:1464–1465. [CrossRef] [PubMed]
YorkJR, LandersS, KirbyRS, ArbogastPG, PennJS. Arterial oxygen fluctuation and retinopathy of prematurity in very-low-birth-weight infants. J Perinatol. 2004;24:82–87. [CrossRef] [PubMed]
LiW, LiuX, YanoffM, CohenS, YeX. Cultured retinal capillary pericytes die by apoptosis after an abrupt fluctuation from high to low glucose levels: a comparative study with retinal capillary endothelial cells. Diabetologia. 1996;39:537–547. [CrossRef] [PubMed]
Del PratoS. In search of normoglycaemia in diabetes: controlling postprandial glucose. Int J Obes Relat Metab Disord. 2002;26(suppl 3)S9–S17. [CrossRef] [PubMed]
IenceanSM. A new classification and a synergetical pattern in intracranial hypertension. Med Hypotheses. 2002;58:159–163. [CrossRef] [PubMed]
WostynP. Normal-tension glaucoma and Alzheimer’s disease: hypothesis of a possible common underlying risk factor. Med Hypotheses. 2006;67:1255–1256. [CrossRef] [PubMed]
AsraniS, ZeimerR, WilenskyJ, et al. Large diurnal fluctuations in intraocular pressure are an independent risk factor in patients with glaucoma. J Glaucoma. 2000;9:134–142. [CrossRef] [PubMed]
Nouri-MahdaviK, HoffmanD, ColemanAL, et al. Predictive factors for glaucomatous visual field progression in the Advanced Glaucoma Intervention Study. Ophthalmology. 2004;111:1627–1635. [CrossRef] [PubMed]
RothS. Endogenous neuroprotection in the retina. Brain Res Bull. 2004;62:461–466. [CrossRef] [PubMed]
SteigerHJ, HanggiD. Ischaemic preconditioning of the brain, mechanisms and applications. Acta Neurochir (Wien). 2007;149:1–10. [CrossRef] [PubMed]
RothS, LiB, RosenbaumPS, et al. Preconditioning provides complete protection against retinal ischemic injury in rats. Invest Ophthalmol Vis Sci. 1998;39:777–785. [PubMed]
CaprioliJ, KitanoS, MorganJE. Hyperthermia and hypoxia increase tolerance of retinal ganglion cells to anoxia and excitotoxicity. Invest Ophthalmol Vis Sci. 1996;37:2376–2381. [PubMed]
EttaicheM, HeurteauxC, BlondeauN, et al. ATP-sensitive potassium channels (K(ATP)) in retina: a key role for delayed ischemic tolerance. Brain Res. 2001;890:118–129. [CrossRef] [PubMed]
CohanBE, BohrDF. Goldmann applanation tonometry in the conscious rat. Invest Ophthalmol Vis Sci. 2001;42:340–342. [PubMed]
WeymouthAE, VingrysAJ. Rodent electroretinography: methods for extraction and interpretation of rod and cone responses. Prog Retin Eye Res. 2008;27:1–44. [CrossRef] [PubMed]
HoodDC, BirchDG. A quantitative measure of the electrical activity of human rod photoreceptors using electroretinography. Vis Neurosci. 1990;5:379–387. [CrossRef] [PubMed]
LambTD, PughEN, Jr. A quantitative account of the activation steps involved in phototransduction in amphibian photoreceptors. J Physiol. 1992;449:719–758. [CrossRef] [PubMed]
BuiBV, FortuneB. Ganglion cell contributions to the rat full-field electroretinogram. J Physiol. 2004;555:153–173. [CrossRef] [PubMed]
EffronB, GongG. A leisurely look at the bootstrap, the jackknife, and cross-validation. Am Statistician. 1983;37:36–48.
BuiBV, WeisingerHS, SinclairAJ, VingrysAJ. Comparison of guinea pig electroretinograms measured with bipolar corneal and unipolar intravitreal electrodes. Doc Ophthalmol. 1998;95:15–34. [CrossRef] [PubMed]
NaarendorpF, SatoY, CajdricA, HubbardNP. Absolute and relative sensitivity of the scotopic system of rat: electroretinography and behavior. Vis Neurosci. 2001;18:641–656. [CrossRef] [PubMed]
FrishmanLJ, ShenFF, DuL, et al. The scotopic electroretinogram of macaque after retinal ganglion cell loss from experimental glaucoma. Invest Ophthalmol Vis Sci. 1996;37:125–141. [PubMed]
FortuneB, BuiBV, MorrisonJC, et al. Selective ganglion cell functional loss in rats with experimental glaucoma. Invest Ophthalmol Vis Sci. 2004;45:1854–1862. [CrossRef] [PubMed]
KimSY, NayakMS, KitaM, MarmorMF. Electroretinogram recovery in the rabbit after repetitive short-term ischemia in light and dark. Invest Ophthalmol Vis Sci. 1994;35:664–668. [PubMed]
Van BuskirkEM, CioffiGA. Predicted outcome from hypotensive therapy for glaucomatous optic neuropathy. Am J Ophthalmol. 1993;116:636–640. [CrossRef] [PubMed]
BurgoyneCF, DownsJC, BellezzaAJ, SuhJK, HartRT. The optic nerve head as a biomechanical structure: a new paradigm for understanding the role of IOP-related stress and strain in the pathophysiology of glaucomatous optic nerve head damage. Prog Retin Eye Res. 2005;24:39–73. [CrossRef] [PubMed]
FlammerJ, OrgulS, CostaVP, et al. The impact of ocular blood flow in glaucoma. Prog Retin Eye Res. 2002;21:359–393. [CrossRef] [PubMed]
KalapesiFB, TanJC, CoroneoMT. Stretch-activated channels: a mini-review: are stretch-activated channels an ocular barometer?. Clin Exp Ophthalmol. 2005;33:210–217. [CrossRef]
KlodaA, LuaL, HallR, AdamsDJ, MartinacB. Liposome reconstitution and modulation of recombinant N-methyl-D-aspartate receptor channels by membrane stretch. Proc Natl Acad Sci U S A. 2007;104:1540–1545. [CrossRef] [PubMed]
OsborneNN, UgarteM, ChaoM, et al. Neuroprotection in relation to retinal ischemia and relevance to glaucoma. Surv Ophthalmol. 1999;43(suppl 1)S102–S128. [CrossRef] [PubMed]
SinghV, CarmanM, RoeperJ, BonciA. Brief ischemia causes long-term depression in midbrain dopamine neurons. Eur J Neurosci. 2007;26:1489–1499. [CrossRef] [PubMed]
BergeaB, BodinL, SvedberghB. Impact of intraocular pressure regulation on visual fields in open-angle glaucoma. Ophthalmology. 1999;106:997–1004. [CrossRef] [PubMed]
ZhangC, RosenbaumDM, ShaikhAR, et al. Ischemic preconditioning attenuates apoptotic cell death in the rat retina. Invest Ophthalmol Vis Sci. 2002;43:3059–3066. [PubMed]
Figure 1.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms in response to the four 15-minute IOP insults. Thin traces: baseline ERG before insult. Thick traces: responses during and after insults. First and second rows recorded 9 minutes after onset of the first and fourth insult, respectively. Third and fourth rows recorded 18 minutes and 1 week after insults, respectively. Vertical line: pSTR implicit time for baseline ERG. Arrowheads: pSTR implicit time for IOP-treated waveforms.
Figure 1.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms in response to the four 15-minute IOP insults. Thin traces: baseline ERG before insult. Thick traces: responses during and after insults. First and second rows recorded 9 minutes after onset of the first and fourth insult, respectively. Third and fourth rows recorded 18 minutes and 1 week after insults, respectively. Vertical line: pSTR implicit time for baseline ERG. Arrowheads: pSTR implicit time for IOP-treated waveforms.
Figure 2.
 
Effect of repeated insults on photoreceptor response (Rm P3 , mean ± SEM). The effects of one, two, and four 15-minute insults are shown (A, unfilled circles, n = 6; B, filled circles, n = 6; C, squares, n = 5). Horizontal dashed lines: mean baseline level. Shaded area: duration of IOP elevation. Filled squares: Data recorded 9 minutes after onset of each insult.
Figure 2.
 
Effect of repeated insults on photoreceptor response (Rm P3 , mean ± SEM). The effects of one, two, and four 15-minute insults are shown (A, unfilled circles, n = 6; B, filled circles, n = 6; C, squares, n = 5). Horizontal dashed lines: mean baseline level. Shaded area: duration of IOP elevation. Filled squares: Data recorded 9 minutes after onset of each insult.
Figure 3.
 
Effect of repeated insults on ON-bipolar cell response (P2, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 9 minutes after onset of each IOP insult. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final baseline. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 3.
 
Effect of repeated insults on ON-bipolar cell response (P2, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 9 minutes after onset of each IOP insult. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final baseline. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 4.
 
Effect of repeated insults on ganglion cell response (nSTR, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 7 minutes after each IOP normalization. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 4.
 
Effect of repeated insults on ganglion cell response (nSTR, mean ± SEM). (A) Single 15-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 15-minute insults (filled circles, n = 6). (C) Four 15-minute insults (squares, n = 5). Filled squares: data recorded 7 minutes after each IOP normalization. (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single, two, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 15-minute insult. Symbols in (D, E) are defined in (AC).
Figure 5.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms after single 60-minute and four 15-minute insults. Thin traces: baseline ERG before insult. Thick traces: responses after insult.
Figure 5.
 
Representative bright flash (A, 1.01 log cd · s/m2) and STR (B, −4.95 log cd · s/m2) waveforms after single 60-minute and four 15-minute insults. Thin traces: baseline ERG before insult. Thick traces: responses after insult.
Figure 6.
 
Effect of insult variation on ON-bipolar cell response (P2, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 6.
 
Effect of insult variation on ON-bipolar cell response (P2, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 7.
 
Effect of insult variation on ganglion cell response (nSTR, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 7.
 
Effect of insult variation on ganglion cell response (nSTR, mean ± SEM). (A) Single 60-minute insult (unfilled circles, n = 6; replotted from He et al. 2 with permission from ARVO). (B) Two 30-minute insults (filled circles, n = 7). (C) Four 15-minute insults (squares, n = 5). (D) Relative recovery normalized to final baseline. Gray, dotted, and thin solid curves: logistic models for single 60-, two 30-, and four 15-minute insults, respectively. Shaded area: 95% CI for mean final recovery. (E) t 0.5 (±95% CI). Shaded area: 95% CI for single 60-minute insult. Symbols in (D, E) are defined in (AC).
Figure 8.
 
Relative t 0.5 (±95% CI) for ON-bipolar cell (P2, unfilled symbols) and ganglion cell responses (nSTR, filled symbols) after multiple insults, normalized to t 0.5 returned after a single insult. (A) Comparison of normalized data between single and multiple insults (Figs. 3Evs. 4E ). (B) Comparison of normalized data between constant and varying insults (Figs. 6Evs. 7E ).
Figure 8.
 
Relative t 0.5 (±95% CI) for ON-bipolar cell (P2, unfilled symbols) and ganglion cell responses (nSTR, filled symbols) after multiple insults, normalized to t 0.5 returned after a single insult. (A) Comparison of normalized data between single and multiple insults (Figs. 3Evs. 4E ). (B) Comparison of normalized data between constant and varying insults (Figs. 6Evs. 7E ).
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×