Effects of Hydrogen Peroxide-Induced Oxidative Damage on Outflow Facility and Washout in Pig Eyes

Transcription

1 Investigative Ophthalmology & Visual Science, Vol. 32, No. 9, August 1991 Copyright Association for Research in Vision and Ophthalmology Effects of Hydrogen Peroxide-Induced Oxidative Damage on Outflow Facility and Washout in Pig Eyes Dovid 3. Yon,* Graham E. Trope,* C. Ross Erhier,j-1. Aravind Menon4 and Andrew Wakeham* Previous studies show that hydrogen peroxide (H 2 O 2 ) is present in the aqueous humor of many species and is capable of affecting outflow facility in animal model experiments. To study the hypothesis that oxidative damage to the outflow pathway may play a role in the pathogenesis of primary open-angle glaucoma, 3 mm H 2 O 2 with 20 mm 3-aminotriazole and 1 mm carmustine (BCNU) in Dulbecco's phosphate-buffered saline (DPBS) was perfused into enucleated pig eyes at constant pressure. Baseline and experimental perfusions were done at two different pressures (7.5 and 30 mm Hg) to study the effect of pressure on the response to oxidative damage. Outflow facility in the baseline experiments (with DPBS only) was observed to increase nonlinearly with time during the perfusions, but could be linearized if plotted as a function of the volume perfused. Thus, a term "volumetric washout" (W) was introduced and defined as the fractional rate of change of outflow facility with respect to the volume perfused. This quantity was found to be independent of pressure in the baseline studies. Perfusion of H 2 O 2 and inhibitors increased W at 7.5 mm Hg but decreased W at 30 mm Hg. These results indicate that oxidative damage increases outflow facility at normal pressure but decreases it at elevated pressure, suggesting that elevated pressure may increase the susceptibility of the outflow pathway to this form of insult. Invest Ophthalmol Vis Sci 32: ,1991 Aqueous humor is known to contain several active oxidative agents such as hydrogen peroxide (H 2 O 2 ) and superoxide anion. 1 " 3 It has been suggested that chronic oxidative insult induced by such agents can compromise trabecular meshwork (TM) function 4 " 6 and subsequently play a role in the pathogenesis of primary open-angle glaucoma. Calf TM has been shown to possess certain protective enzymes such as catalase, glutathione peroxidase, glutathione reductase, and glucose-6-phosphate dehydrogenase, 7 " 10 suggesting that this tissue is capable of detoxifying oxidative agents. However, it is not known whether dysfunction of this detoxification mechanism affects outflow facility, and thus it is interesting to evaluate the response of the outflow pathways to H 2 O 2. Others recently reported that the perfusion of 0.3 mm H 2 O 2 along with enzyme inhibitors into enucleated calf eyes decreased facility. 6 If intraocular pressure (IOP) could be demonstrated to affect From the Departments of ""Ophthalmology, fmechanical Engineering, and ^Medicine and Ophthalmology, University of Toronto, Canada. Supported by The Physicians Services Incorporated Foundation of Ontario and grant MA from the Medical Research Council of Canada. Submitted for publication November 28, 1990; accepted March 21, Reprint requests: Dr. G. Trope, Toronto General Hospital, 200 Elizabeth Street, #5 Eaton-306, Toronto, Ontario, Canada M5G 2C4. H 2 O 2 -induced changes in facility, this information could aid in our understanding of how oxidative damage to the TM affects facility since IOP itself is known to have a profound effect on facility. 11 " 14 Thus our objective was to examine the effects of H 2 O 2 on facility and washout rate in pig eyes at various IOPs to determine whether pressure plays a role in either potentiating or inhibiting the effects of H 2 O 2. Since the pig eye has not been studied extensively in perfiisipn experiments, we also characterized the pig eye as an animal model of aqueous outflow by measuring the facility and washout rate in enucleated pig eyes at different IOPs. Materials and Methods Perfusion Protocol Paired eyes were carefully enucleated by one of us (AW) immediately after the pigs were slaughtered by exsanguination at a commercial slaughterhouse. The eyes were immersed in isotonic saline and packed on ice until the start of the perfusions approximately 2 hr later. Eyes were rejected if they had obvious anterior segment abnormalities or if they had sustained significant physical trauma. The pig eyes were perfused at constant pressure through a 23-gauge needle inserted into the posterior chamber and connected to an inlet reservoir; this reservoir was weighed with a specially designed balance and continuously recorded by computerized data ac- 2515

2 2516 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / August 1991 Vol. 32 quisition. The facility (C) during a given time interval At was then calculated as C(t) = AV(t)/IOPAt, where AV(t) is the volume perfused during At and IOP is the perfusion pressure. All experiments were done at room temperature (22 C) with Dulbecco's phosphate-buffered saline (DPBS; Media Services, University of Toronto, Toronto, Canada) and 5.5 mm glucose passed through a 0.22-^m GVWP Millipore filter (Millipore Corp., Bedford, MA) before use. Baseline experiments were done to characterize the pig-eye model by perfusing DPBS with glucose into the eyes through a 23-gauge (inlet) needle inserted into the posterior chamber. The eyes were perfused at either 7.5 or 30 mm Hg for approximately 6 hr. For the perfusions with H 2 O 2 at 7.5 and 30 mm Hg, a second 23-gauge (outlet) needle was inserted into the anterior chamber (AC) and connected to a reservoir 5 mm Hg lower than the level of the inlet reservoir to allow fluid in the AC to be exchanged. The line to the outlet needle was clamped off at all times except during the AC exchange. The DPBS with 5.5 mm glucose initially was perfused into paired eyes for approximately 2 hr. To do the AC exchange, the inlet lines were first clamped off, and the fluid in the inlet reservoir of one eye of each pair (the experimental eye) was then replaced with the solution containing H 2 O 2. The inlet and the outlet lines of both eyes then were undamped simultaneously to initiate theflowof the fluid from the inlet reservoir to the lower outlet reservoir. During the exchange, data collection was suspended. A total of 2 ml was flushed through the AC to ensure complete exchange; this usually required min. The effectiveness of the exchange procedure was verified by monitoring the entry of fluorescein into the AC. The experimental eye underwent AC exchange with a solution of DPBS and glucose to which 1 mm carmustine (BCNU; Bristol Laboratories of Canada, Belleville, Ontario Canada), 20 mm 3-aminotriazole (3-AT; Aldrich Chemical Co., Milwaukee, WI), and 3 mm H 2 O 2 (Fisher Scientific, Fairlawn, NJ) had been added. The addition of H 2 O 2 with BCNU and 3-AT did not affect the ph of the DPBS medium, but osmolality was increased slightly. The contralateral control eye also underwent an AC exchange with the DPBS and glucose perfusion vehicle only. At the completion of the exchange, the outlet lines were reclamped, and the eyes were perfused further with the same solutions introduced by AC exchange. Data Reduction In addition to the traditional plots of experimentally measured facility as a function of time (t), it was found useful to normalize C by a reference value C o for each eye: C = Co [1] This normalization procedure allows changes in C to be compared on a relative basis, thus reducing the effects of eye-to-eye variability on experimental response. For the baseline experiments, C o was taken to be the facility at the beginning of the perfusion (t = 0), estimated by doing a linear regression of the facility data from t = 30 min to t = 60 min and extrapolating back to t = 0. This procedure ignored data in the 0-30-min range to avoid the perfusion filling curve, during which the measured facility is erroneously high as a result of thefillingof the AC and the viscoelastic expansion of the globe. For the H 2 O 2 experiments, C o was taken to be the facility immediately before the AC exchange was done (at the end of the initial perfusion period). C o was estimated from a linear regression of the facility data in the initial perfusion period from t = 30 min up to the time of the AC exchange (t exc ). The normalized facility (norm) measurements for a given group of eyes (eg, all baseline eyes perfused at 7.5 mm Hg) were combined into "average normalized facilities" (C norm ) with respect to either time (t) or perfused volume (V). The average normalized facility (C n orm) f r a group of N eyes with normalized facilities Cnorm.1, C norm, 2...C norm>n at a given time (t) was defined as while at a given perfused volume (V), C norm was calculated as [2] = y c ( [3] Note that Equation 2 averages facilities measured at a common time during the perfusion, whereas Equation 3 averages facilities measured at a common perfused volume. For the baseline experiments, time and volume were referenced to the beginning of the perfusion. For the H 2 O 2 experiments, time and perfused volume were referenced to values at the point of the AC exchange. For reasons stated subsequently, perfusion data were analyzed in terms of "volumetric washout" (W) rather than the conventional definition of washout, ie, the rate of change of facility with respect to time. We define W to be the rate of change of normalized facility with respect to volume perfused, ie,

3 No. 9 EFFECTS OF H 2 O 2 ON OUTFLOW FACILITY IN PIG EYES / Yon er ol 2517 dv W was calculated for each eye by finding the slope of the linear regression of C norm with respect to volume perfused. For the baseline perfusions, values of W for all eyes at a given perfusion pressure were pooled, and the average volumetric washout at different pressures statistically compared by an unpaired, two-sided student t-test. For the H 2 O 2 perfusions, values of W for the control and experimental groups were compared by a paired, two-sided student t-test. [4] O Q 7.5 mmhg 30 mmhg T T>A r / Baseline Experiments Results Representative examples of facility data obtained during baseline perfusions of one eye at 7.5 mm Hg and one eye at 30 mm Hg are shown in Figure 1. Volumes perfused typically totaled 1 ml into eyes at 7.5 mm Hg and 4 ml at 30 mm Hg. In Figure 2, C norm is plotted as a function of time for 12 eyes at 7.5 mm Hg and 14 eyes at 30 mm Hg. The normalization procedure produced reference points f C nortn = 1 at t = 0 by definition. The average initial facility was ± ^1/min/mm Hg at 7.5 mm Hg and ± ^l/min/mm Hg at 30 mm Hg. Cnorm increased in a nonlinear, accelerating manner with respect to time at 30 mm Hg, but C norm and time appeared to be linearly related at 7.5 mm Hg. In Figure 3, C norm is plotted with respect to volume perfused for eyes at 30 mm Hg. It can be seen that the nonlinearity observed when C norm was plotted with time (min) Fig. 2. Plot of averaged, normalized facility C norm (±SEM) as a function of time for baseline perfusions of 12 eyes at 7.5 mm Hg and 14 eyes at 30 mm Hg. C norm varies in a nonlinear, accelerated manner with respect to time at 30 mm Hg, but not at 7.5 mm Hg. Average initial facility was ± jil/min/mm Hg at 7.5 mm Hg and ± Ml/min/mm Hg at 30 mm Hg. respect to time in Figure 2 now was largely eliminated and linear regression of the plot of C norm versus V yielded a correlation coefficient of r 2 = In Figure 4, C norm for eyes at both 7.5 and 30 mm Hg is plotted as a function of perfused volume. Only the first 0.8 ml of perfusion data at 30 mm Hg is shown for direct comparison because perfusions at 7.5 mm Hg typically totaled only 1 ml or less. By linear regression, W at 7.5 mm Hg (36.9%/ml, r " 7.5 mmhg 0.05" time (min) Fig. 1. Representative plot of facility data as a function of time for the baseline perfusion of one eye at 7.5 mm Hg and one eye at 30 mm Hg perfused volume V(ul) 5000 Fig. 3. Plot of averaged, normalized facility (C norm ) as a function of perfused volume (V) for baseline perfusions of 14 eyes at 30 mm Hg. Solid line: least squares linear regression of data (r 2 = 0.99). Compare with Figure 2.

4 2518 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / August 1991 Vol " 2.5" control^ H2O2 o " perfused volume V(ul) 7.5 mmhg 30 mmhg 1000 Fig. 4. Plot of averaged, normalized facility (C nortn ) as a function of perfused volume (V) for baseline perfusions of 12 eyes at 7.5 mm Hg and 14 eyes at 30 mm Hg. Only thefirst800 fil of data at 30 mm Hg are shown to facilitate direct comparison between the two pressures. Volumetric washout at 7.5 mm Hg (W = 36.9%/ml) and 30 mm Hg (W = 36.2%/ml) are not statistically different (P > 0.25). = 0.97) was not statistically different (P > 0.25) from W at 30 mm Hg (36.2%/ml, r 2 = 0.95). H 2 O 2 Perfusions Figure 5 details the results of H 2 O 2 perfusions into 11 pairs of eyes at 7.5 mm Hg. The volume of perfusate typically perfused into the eyes was 0.5 ml before and 2 ml after AC exchange. After introducing 3 mm H 2 O 2 with 1 mm BCNU and 20 m_m 3-AT into the experimental eye by AC exchange, C norm increased in 3.0 o.o AC exchange before AC exchange V a<ter AC exchange perfused volume (ul) H2O2 control 1600 Fig. 5. Plot of averaged, normalized facility (C norm ) as a function of perfused volume (V) for control and H 2 O 2 -perfused eyes at 7.5 mm Hg. Note that perfusion of H 2 O 2 increases normalized facility and volumetric washout relative to control eyes at 7.5 mm Hg. O 1.0" I o.oi AC exchange before AC exchange \ f after AC exchange volume perfused (ul) 4000 Fig. 6. Plot of averaged, normalized facility (C norm ) as a function of perfused volume (V) for control and H 2 O 2 -perfused eyes at 30 mm Hg. Note that H 2 O 2 does not initially affect volumetric washout at 30 mm Hg up to ~ 1850 n\; after that, it decreases volumetric washout relative to the control eyes. the experimental eyes relative to the control eyes. W in the experimental group of eyes (W expt = 68 ± 49%/ ml) was significantly greater (P = 0.01) than that in the control group (W ctrl = 39 ± 37%/ml) in the ^1 perfused volume range. Figure 6 details the results of H 2 O 2 perfusions into 14 pairs of eyes at 30 mm Hg. The volume of perfusate typically perfused into the eyes was 0.75 ml before and 3.5 ml after AC exchange. After AC exchange, C norm initially was slightly higher in the experimental eyes than in the control eyes. However, this situation later reversed as C norm in the experimental eyes stopped increasing at about ul Shortly thereafter, the experimental curve dipped below the control curve. W in the experimental group of eyes initially did not differ statistically from the control eyes in the ^1 perfused volume range (W expt = 44 ± 21%/ml, W ctrl = 43 ± 21%/ml, P = 0.82), but thereafter in the ^1 perfused volume range, W in the experimental group was significantly less than in the control group (W expt = 16 ± 18%/ml, W ctrl = 35 ± 25%/ml, P = 0.006). Discussion Characterization of the Pig Eye The pig eye was evaluated as a model of aqueous outflow for this study using perfusion techniques. The facility of pig eyes was greater at low pressure (C = jtl/min/mm Hg at 7.5 mm Hg) than at high pressure (C = j^l/min/mm Hg at 30 mm Hg), similar to that observed in enucleated human and bovine eyes. 11 " 14 At a presumed physiologic pressure of

5 No. 9 EFFECTS OF H 2 O 2 ON OUTFLOW FACILITY IN PIG EYES / Yon er ol mm Hg, the facility of human eyes (approximately 0.2 ^1/min/mm Hg) 1314 is much closer to that of pig eyes (0.156 /il/min/mm Hg) than bovine eyes (approximately 1.5 jiil/min/mm Hg). 14 It has long been recognized that facility continuously increases when nonhuman eyes are perfused with substitutes for aqueous humor. 15 " 18 This increase in facility previously was quantified as "washout," defined to be rate of change of facility with respect to time. 6 * 15 However, it is evident from Figure 2 that washout increased in eyes perfused at 30 mm Hg throughout the course of the perfusion, indicating that a single measurement of washout cannot characterize an entire perfusion. At 7.5 mm Hg, the relationship between normalized facility and time appears linear. Replotting the normalized facility data with respect to volume perfused (as first suggested by Johnson et al 19 ) has the surprising effect of linearizing the perfusion data at 30 mm Hg (Fig. 3). Equally important, the slope of normalized facility with respect to volume perfused (volumetric washout) was the same at 7.5 mm Hg as it was at 30 mm Hg (Fig. 4). These results suggest that the washout phenomenon is related directly to the volume of perfusate that passed through the outflow tissues (rather than the duration of the perfusion). 19 This suggests that washout is not directly a result of postmortem tissue degradation, which presumably correlates closely with time. These findings can be explained mathematically by analysing the relationships between C norm, volume perfused, and time. Assuming that C norm is related linearly to volume perfused (Figs. 3, 4), we can write c = i + wv [5] where V is the volume perfused, and W is the volumetric washout as previously defined in Equation 4. Taking the derivative of Equation 5 with respect to time (t) results in [6] dt dt \X dvv Substituting for the perfusion rate - = CP and redt placing C with C norm C 0 as previously defined in Equation 1 yields dc nc = C norm (WC 0 P) [7] dt which can then be integrated to obtain p _ e(wcop)t Equation 8 suggests that the nonlinearity observed in Figure 2 is exponential in nature. The apparent linearity of the data at 7.5 mm Hg [8] with respect to time and volume perfused is consistent with Equation 8 because for WC 0 Pt < 1, Equation 8 can be linearly approximated 20 as (WC 0 P)t [9] For the baseline perfusions at P = 7.5 mm Hg, C o = Atl/min/mm Hg, t = 400 min, and W = X 10" 3 /M1, the condition WC 0 Pt = 0.17 ^ 1 is satisfied. Thus, facility may appear to increase linearly with time at 7.5 mm Hg because the perfusion pressure is low and/or the duration of the perfusion is relatively short. Nonlinearity between facility and time may not have been previously observed in other perfusion studies because the experimental parameters met the condition WC 0 Pt<^ 1 for approximate linearity. The duration of the perfusion often may be the limiting factor since enucleated eyes generally are not perfused for more than 6 hr. In such cases, increased perfusion pressure (as in this study) reveals the nonlinearity at earlier times. Both W and C o are species-dependent factors that probably are responsible for the observed differences between the responses of bovine and pig eyes with regard to washout. Effects of H 2 O 2 on Facility and Washout Anderson et al 6 previously showed that perfusion of H 2 O 2 at 15 mm Hg decreases washout in calf eyes. In this study, we found that the effect of H 2 O 2 on the washout of pig eyes was pressure dependent. Our results with pig eyes perfused with H 2 O 2 at 30 mm Hg were consistent with the results obtained by Anderson et al 6 using calf eyes at 15 mm Hg. In both cases, there was little effect early on, but later H 2 O 2 caused a significant decrease in washout. Such a lag time in response to H 2 O 2 as measured by facility also was observed by Kahn et al, 5 who reported that pretreatment of calf eyes with BCNU/diamide and subsequent exposure to H 2 O 2 at 15 mm Hg did not produce an observed effect on facility until at least 2 hr after the commencement of H 2 O 2 exposure. However, we found that when a low perfusion pressure of 7.5 mm Hg was chosen, H 2 O 2 increased volumetric washout almost immediately and for the duration of the perfusion, unlike results at the higher pressure of 30 mm Hg. This suggests that the effects of oxidative damage and IOP on TM function somehow interact, and the nature or degree of this interaction is pressure dependent. If pressure had either increased or decreased facility at both 7.5 and 30 mm Hg, but to different degrees, we might speculate that pressure is primarily modulating a single factor affected by oxidative damage. However, we found that the effects of oxidative damage on facility were opposite at different pressures, suggesting

6 2520 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / August 1991 Vol. 32 that there are at least two competing factors affected by pressure and oxidative damage. It would have been interesting to perfuse the eyes at 7.5 mm Hg for the same total volume as was perfused at 30 mm Hg to determine whether the decrease in washout observed later in the 30 mm Hg perfusions ultimately would have been observed in the 7.5 mm Hg perfusions. However, this would have required an extremely long perfusion time, raising questions about anterior segment tissue viability. In summary, we found in pig eyes that volume is a better experimental parameter than time for perfusion studies because facility varies linearly with respect to volume but nonlinearly with respect to time. In addition, the derivative of C norm with respect to volume (ie, volumetric washout) was independent of perfusion pressure. Thus, only a single measurement of W (independent of time and pressure) is required to characterize washout in all enucleated-eye perfusions with a given perfusion medium. Oxidative damage caused by H 2 O 2 with BCNU and 3-AT increased facility at normal pressure but decreased it at elevated pressure. We speculate that oxidative insult does not initiate the damage as seen in glaucoma. Rather, the susceptibility of the outflow pathway to oxidative insult may be increased after pressure has already been elevated by other pathogenic mechanisms, thus accelerating the disease process. Key words: oxidative damage, pig eye perfusion, glaucoma, washout, outflow facility References 1. Spector A and Garner WH: Hydrogen peroxide and human cataract. Exp Eye Res 33:673, Pirie A: Glutathione peroxidase in lens and a source of hydrogen peroxide in aqueous humor. Biochem J 96:244, Bhuyan KC and Bhuyan DG: Regulation of hydrogen peroxide in eye humors: Effect of 3-amino-lH-l,2,4-triazole on catalase and glutathione peroxidase of rabbit eye. Biochim Biophys Acta 497:641, Nguyen KPV, Chung ML, Anderson PJ, Johnson M, and Epstein DL: Hydrogen peroxide removal by the calf aqueous outflow pathway. Invest Ophthalmol Vis Sci 29:976, Kahn MG, Giblin FJ, and Epstein DL: Glutathione in calf trabecular meshwork and its relation to aqueous humor outflow facility. Invest Ophthalmol Vis Sci 24:1283, Anderson PJ, Karageuzian L, Chung M, and de Kater AW: Responses of calf TM to oxidative insult: Facility, enzyme loss, morphology. ARVO Abstracts. Invest Ophthalmol Vis Sci 30(Suppl):223, Scott DR, Karageuzian LN, Anderson PJ, and Epstein DL: Glutathione peroxidase of calf trabecular meshwork. Invest Ophthalmol Vis Sci 25:599, Nguyen KPV, Weiss H, Karageuzian LN, Anderson PJ, and Epstein DL: Glutathione reductase of calf trabecular meshwork. Invest Ophthalmol Vis Sci 26:887, Freedman S, Anderson PJ, and Epstein DL: Superoxide dismutase and catalase of calf trabecular meshwork. Invest Ophthalmol Vis Sci 26:1330, Nguyen KPV, Lee DA, Anderson PJ, and Epstein DL: Glucoses-phosphate dehydrogenase of calf trabecular meshwork. Invest Ophthalmol Vis Sci 27:992, Moses RA: The effect of intraocular pressure on resistance to outflow. Surv Ophthalmol 22:88, Ellingsen BA and Grant WM: The relationship of pressure and aqueous outflow in enucleated human eyes. Invest Ophthalmol 10:430, Brubaker RF: The effect of intraocular pressure on conventional outflow resistance in the enucleated human eye. Invest Ophthalmol 14:286, Hashimoto JM and Epstein DL: Influence of intraocular pressure on aqueous outflow facility in enucleated eyes of different mammals. Invest Ophthalmol Vis Sci 19:1483, Bassett-Chu S, Erickson-Lamy KA, and Epstein DL: Comparison of the facility-increasing effect of anterior chamber perfusion in eyes from different primate and subprimate species. ARVO Abstracts. Invest Ophthalmol Vis Sci 27(Suppl):353, Gaasterland DE, Pederson JE, and MacLellan HM: Perfusate effects upon resistance to aqueous humor outflow in the rhesus monkey eye. Invest Ophthalmol Vis Sci 17:391, Gaasterland DE, Pederson JE, MacLellan HM, and Reddy VN: Rhesus monkey aqueous humor composition and a primate ocular perfusate. Invest Ophthalmol Vis Sci 18:1139, Erickson KA and Kaufman PL: Comparative effects of three ocular perfusates on outflow facility in the cynomolgus monkey. Curr Eye Res 1:211, Johnson M, Kim A, Kamm RD, Epstein DL, and Gong H: Extracellular proteins and the "wash-out" effect. ARVO Abstracts. Invest Ophthalmol Vis Sci 31(Suppl):376, Selby SM, editor: Standard Mathematical Tables, 20th ed. CRC Press, Cleveland, OH, p. 456.