Oxygen Distribution in the Macaque Retina

Transcription

1 Oxygen Distribution in the Macaque Retina Jameel Ahmed* Rod D. Braun,^ Robert Dunn, Jr. * and Robert A. Linsenmeier*X Purpose. Oxygen distribution was characterized in the macaque retina, which is more like the human retina than others studied previously. Methods. Profiles of oxygen tension (Po 2 ) as a function of distance were recorded in a parafoveal region about halfway between the fovea and optic disk, and from the fovea in one animal. A one-dimensional diffusion model was used to determine photoreceptor oxygen consumption (QO 2 ). Results. In the parafovea, the Po 2 decreased as the electrode was withdrawn from the choroid toward the inner retina, reaching a minimum value during dark adaptation of about 9 mmhg at about 70% retinal depth, and then increasing more proximally. Approximately 90% of the oxygen requirement of the photoreceptors was supplied by the choroidal circulation and 10% by the retinal circulation. In light adaptation, there was a monotonic Po 2 gradient from the choroid to the inner retina, indicating that all of the oxygen used by photoreceptors was supplied by the choroid. In the fovea, the choroid supplied almost all the oxygen in both dark and light adaptation, with a minor supply from the vitreous humor. Dark-adapted foveal oxygen consumption was lower than parafoveal oxygen consumption. Light reduced the oxygen consumption of the photoreceptors, in both regions studied, by 16-; Conclusions. The results show that oxygenation of the parafoveal monkey retina is similar to that previously observed in the cat area centralis. In the fovea, the oxygen distribution differs as expected considering the thinner retina and the absence of inner retinal neurons and retinal circulation. Invest Ophthalmol Vis Sci. 1993;34: Intensive measurements of oxygen distribution and photoreceptor oxygen consumption have been made in the retina of the cat 1 " 3 and miniature pig; 4 however, no intraretinal measurements, and only a small number of vitreal measurements, have been made in primates. In this paper, we report intraretinal measurements from two monkeys, and show that the cat retina From the Departments of *Biomedical and ~\Chemical Engineering and %Neurobiology and Physiology, Northwestern University, Evanston, Illinois. This work was supported by NEl grant EY05034 to RAL and a student fellowship (toj.a.) in memory of Herman Shane from the Fight for Sight Research Division of the National Society to Prevent Blindness. Submitted for publication: July 6, 1992; accepted November 10, Proprietary interest category: N. Reprint requests: Dr. Robert A. Linsenmeier, Biomedical Engineering Department, Northwestern University, 2145 Sheridan Road, Evanston, IL serves as a good model for oxygenation of the parafoveal primate retina. In one monkey, we were able to obtain measurements from the fovea. These measurements showed differences that were consistent with the decreased retinal thickness and the absence of retinal circulation in this region. MATERIALS AND METHODS Measurements were made from two monkeys using techniques that were similar to those used previously in studies on the cat. 25 One monkey was a cynomolgus macaque (M. fascicularis) weighing 3.3 kg and the other was a pig-tailed macaque (M. nemestrina) weighing 4.8 kg. We adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Animals were initially anesthetized with keta- 516 Investigative Ophthalmology & Visual Science, March 1993, Vol. 34, No. 3 Copyright Association for Research in Vision and Ophthalmology

2 Oxygen Distribution in the Macaque Retina 517 mine (22 mg/kg) plus acepromazine (0.15 mg/kg) IM, and anesthesia was maintained during surgery with sodium thiamylal (5 wt%). Long-term anesthesia was provided by urethane ( mg/kg loading dose and mg/hr thereafter). After surgery, animals were paralyzed with pancuronium bromide ( mg/kg/hr) and artificially respirated. The eye was attached to a stabilizing ring, and hypodermic needles were inserted into the eye to carry the oxygen microelectrode, an Ag/AgCl vitreal reference electrode for the oxygen measurement, and an Ag/AgCl vitreal electrode for recording the electroretinogram (ERG). An Ag/AgCl reference electrode for the ERG was sewn into tissue outside the orbit. Arterial blood parameters were measured and controlled by adjustments of the respirator and composition of the inspired air. During data collection, P a O 2 was in the range of 94 to 120 mmhg; P a CO 2 was in the range of mmhg and ph a was between 7.37 and Oxygen tension (Po 2 ) was measured with doublebarreled oxygen microelectrodes constructed and calibrated as previously described. 6 Microelectrode penetrations of the retina were made in both animals between the fovea and the optic disk (parafoveal), using changes in the local ERG recorded by the second barrel of the electrode as an index of penetration. In M. nemestrina, recordings were also made in the fovea. Zero percent depth was taken to be the point at which the retina was first touched, indicated by a transient deflection in the signal from the voltage barrel, and 100% depth was indicated by the transepithelial potential across the retinal pigment epithelium. The data shown and the analysis of oxygen consumption are based on continuous withdrawals of the electrode from the choroid to the vitreous in l-/*m steps at 2 jtim/sec. Electrode penetrations were always made under dark-adapted conditions. Electrode withdrawals were done either in dark adaptation or when the retina was adapted to a steady light sufficient to saturate the b-wave of the vitreal ERG (7.9 log equivalent quanta (555 nm)/deg 2 -sec). All distances have been corrected for a penetration angle that was assumed to be 45 deg with respect to the retinal surface, and are therefore approximately radial distances. For both animals, we used the special three-layer model of oxygen consumption developed for the cat retina and described previously. 3>5>7 In this model, profiles of Po 2 as a function of distance in the outer half of the retina are fitted to a one-dimensional diffusion equation in which oxygen consumption (Q) is confined to a single small region between two nonconsuming regions. For the foveal data, in which there is no retinal circulation, the entire thickness of the retina was modelled. Because the model yields values for Q/Dk, values for D (diffusivity of oxygen in the retina) and k (oxygen solubility) must be assumed to arrive at Q. We used previously determined values for D of 1.97 X 10" 5 cm 2 /sec and for k of 2.4 X 10" 5 ml O 2 /ml tissue-mmhg. 3 The only values reported here that were obtained from the model are for: (1) photoreceptor QO 2 ; (2) the percentage of photoreceptor QO 2 that was obtained from the retinal circulation; and (3) the size of the consuming layer. Values are given as mean ± SD. RESULTS Figure 1 shows examples of profiles of Po 2 as a function of depth from the parafoveal retina of both mon- A B Percent retinal depth FIGURE l. Profiles of oxygen tension as a function of depth obtained during electrode withdrawals from the parafoveal monkey retina. (A) Profiles from the light- and dark-adapted retina of M. fascicularis. (B) Profiles from M. nemestrina. Dots indicate dark-adapted profiles.

3 518 Investigative Ophthalmology & Visual Science, March 1993, Vol. 34, No. 3 keys in light and dark adaptation. In general, these were similar to those recorded in the cat. The choroidal Po 2 was 53.4 ± 5.4 mmhg (n = 10 profiles) in one animal and 83.6 ± 6.5 mmhg (n = 19) in the other. A gradient of Po 2 was observed as the electrode was withdrawn from the choroid, with the slope being greater when the retina was dark-adapted. The direction of the gradients indicated that the choroidal circulation supplied the oxygen for the entire outer half of the retina in the light. During dark-adapted conditions, however, there was a flux of oxygen from the retinal circulation to the outer retina in 11 of the 12 profiles. The diffusion model allowed us to calculate 3 that 8.9 ± 6.8% (n = 12) of the oxygen consumed by the photo receptors came from the retinal circulation. The minimum Po 2 observed in the outer retina during dark adaptation was higher than that in the cat retina, in which it is typically almost zero, even during normoxia. 2 The point of minimum Po 2 occurred at 70 ± 3% retinal depth (n = 5) in M. nemestrina and averaged 6.9 ± 2.8 mmhg at this point. The point of minimum Po 2 was more proximal in M. fascicularis, which suggests that the microdrive was not pulling the electrode as smoothly during this experiment; however, the minimum Po 2, 8.8 ± 3.5 mmhg, was similar. At corresponding retinal depths during light adaptation, the Po 2 was considerably higher in both animals, 21.4 ± 2.9 mmhg (n = 9). This clearly indicates that oxygen consumption decreases in light adaptation, as has been reported for other vertebrates. 2 In the inner retina, the Po 2 was lower than at the choroid and showed a heterogeneity that depended on proximity to blood vessels (Fig. 1). The average values of Po 2 appeared to be similar to those obtained in the cat retina. 3 Because no location in the cat retina corresponds to the fovea in terms of retinal thickness or domination by cones, we wanted to record from this area in the primate. The electrode was positioned in the avascular part of the fovea under ophthalmoscopic observation, with the local ERG being used to confirm its position. In the fovea, the receptor potential has a greater influence on the shape of the local ERG than in other regions. 8 Examples of the dark-adapted local ERG in response to 4-sec flashes of diffuse light are shown in Figure 2. Figure 2A shows responses in the parafovea at two retinal depths, while Figure 2B shows responses from the fovea at similar depths. At shallow depths, the parafoveal response shows a negative-going b-wave followed by the positive-going c-wave, while the foveal response has a positive-going sustained receptor potential in addition to the c-wave, and there is no distinct b-wave. Deeper in the retina, the receptor potential plateau is of larger amplitude in the foveal ERG. The retina was thinner in the fovea, 172 ± 24 /im (n = 7) in foveal penetrations, compared with 224 ±21 /xm (n = 12) in parafoveal penetrations. Choroidal Po 2 was approximately the same as in the parafoveal region, and there was a similar drop in Po 2 as the electrode was withdrawn. Figure 3 shows oxygen profiles from the fovea and the light-evoked change in Po 2 recorded when the electrode was positioned at 35% retinal depth during a penetration. Like the parafoveal profiles, the lightadapted profiles had a higher minimum, which, along with the increase in Po 2 resulting from stimulation by light, indicates that cone oxygen use decreases in the light. The profiles in both light and dark adaptation showed a distinct bend, corresponding to a region of high oxygen consumption, and were then linear most A) Parafoveal retinal depth B) Foveal retinal depth Parafoveal 647. retinal depth Foveal 7A7. retinal depth I 4 sec Illumination 2mv 4 sec Illumination FIGURE 2. Intraretinal electroretinograms from parafoveal and foveal retina of M. nemestrina in response to 4-sec flashes of diffuse white light at 7.9 log equivalent quanta (555nm)/deg 2 -sec.

4 Oxygen Distribution in the Macaque Retina r ^ X 60 E E Q_ ?Hel jr Dark adapted ^WHW^ Percent Retinal Depth FIGURE 3. Profiles of oxygen tension as a function of depth in the foveal retina of M. nemestrina. Inset: Po 2 change due to 30-sec Hash of diffuse white light at 7.9 log equivalent quanta (555nm)/deg 2 -sec; recorded from 35% retinal depth; baseline = 8 mmhg. of the way toward the vitreous. This linearity indicates that, in this region, oxygen consumption is very low, and the lack of any local peaks suggests that oxygen diffusion from vascularized regions around the fovea is minimal. In the most proximal part of the retina, most of the profiles showed another small increase, indicating that some oxygen diffused into the fovea from the vitreous, but this appears to contribute very little to foveal oxygen use. Although the basic aspects of oxygen supply and use are apparent from the shapes of the profiles, we used diffusion modelling to learn more about relative oxygen consumption in light and dark and in foveal vs parafoveal regions. Profiles were fitted by the model used previously for the cat. Two measures of oxygen consumption are given, based on the results of the modelling. The first, Q* v, is the corrected average oxygen consumption per unit volume of tissue over the outer half of the retina. Q* v is given by Q* = QJ X (L2 - Ll)/L, where: Q2* is the oxygen consumption per unit volume of the consuming layer, corrected for mechanical deformation of the retina; 5 LI and L2 are the boundaries of the consuming layer; and L is the total thickness of the outer avascular retina. This is explained more fully by Haugh et al. 7 Q* v is the consumption of the photoreceptor layers in the fovea and parafovea, but may slightly underestimate relative foveal oxygen consumption, because the total retinal thickness (approximately 172 /im) could be analyzed in the fovea, while only the outer retina (approximately 112 pm) was modelled in the parafoveal profiles. A second measure of consumption, the total oxygen consumption of the photoreceptors per unit area of retinal surface, Qf ol = Qf X (L2 - LI) (1) was used to allow us comparison of foveal and parafoveal profiles. Figure 4 shows values for consumption obtained from the model. Average outer retinal consumption in the parafovea (Fig. 4A) was lower in the light than in the dark for both monkeys. In comparison with the cat, dark-adapted consumption appeared to be lower and the percent change with light was smaller. The consuming region began about 50 fim from the choroid (ie, LI = 50), and it occupied 20-30% of the outer half of the retina. From Figure 1 of Schein's anatomical study on the macaque fovea, 9 it appears that the inner segment region of the photoreceptors, unconnected for shrinkage, begins about 40 nm from the choroid, placing the consuming region slightly proximal to the inner segment region, a characteristic that has also been found in the cat. This may result partially from the shrinkage in anatomic sections and partially from a slight distortion in the profiles. Figure 4B shows values for total consumption from the parafovea and fovea of M. nemestrina. From this graph, we can see that dark-adapted oxygen consumption is lower in the fovea, and that the effect of illumination is less than in the parafovea. DISCUSSION The measurements reported here show that the choroid is the major source of oxygen for primate photoreceptors, as it is for photoreceptors in other vertebrates. The choroidal Po 2 in the two experiments differed by approximately 30 mmhg. Reasons for this difference are not known, but this level of variability in choroidal Po 2 is also found among cats. We do not believe that this is due to a species difference between the animals. In dark adaptation, only about 10% of the oxygen for parafoveal photoreceptors is derived from the retinal circulation. The maximum (ie, darkadapted) oxygen consumption of the outer half of the parafoveal primate retina appears to be approximately 75% of the oxygen consumption of the cat area centralis; however, because we have data from only two monkeys, this value must be considered somewhat preliminary. The conclusion that consumption is lower is partially based on the values obtained from modelling, where we can compare the average values for Q^v of

5 520 Investigative Ophthalmology 8c Visual Science, March 1993, Vol. 34, No. 3 E 5 I 0) 3 n.5 4 s 3 O 2 6 r A Dark-adapted Light adapted o A 6 c E I 5 0) 3 n - 4 E \ CM O I 2 B n=6 t l Dark-adapted I I Light adapted M.fasc. M. nem Cat Parafoveal M. nemestrina FIGURE 4. (A) Average outer retinal oxygen consumption per unit volume (Qjf v ), in light and dark adaptation, for both monkeys and for the cat. (B) Total outer retinal oxygen consumption per unit area (Q;!^) in parafovea and fovea of M. nemestrina. Cat data from Linsenmeier and Braun; s error bars indicate standard deviation; n is the number of profiles. Foveal 3.1 and 3.8 ml O 2 /100 g tissue-min in the two monkeys to the average value for cat of ml O2/100 g tissue-min; 37 and partially based on the observation that the minimum Po 2 in dark adaptation is higher in the monkey retina than in the cat retina. A lower oxygen use in the outer half of the parafoveal primate retina is not unexpected, given the lower total receptor density in the primate. In M. nemestrina, 2 mm from the fovea, there are about 15,000 cones/mm 210 and, in the similar region in the cat retina, there are only about 7,000 cones/mm 2." The difference in rod density is in the opposite direction, however, and is much larger. Our oxygen measurements in cat have been made in a region in which rod density is 300, ,000/inm 2, whereas the primate measurements are from a region in which rod density is only about 100,000/mm 2. In view of these numbers, it is interesting that the difference in oxygen consumption between the cat and the monkey is as small as it is. In both the cat and the monkey, light reduces oxygen consumption, but the percentage change at comparable levels of illumination appears to be smaller in the primate than in the cat. 3 This may be, in part, because of the larger number of cones in the monkey and our failure to reduce cone oxygen consumption maximally with light. Despite some differences, it appears that the cat retina provides a relatively good model for oxygen distribution and consumption in the parafoveal retina of primates and, presumably, humans. From the foveal profiles and light-evoked changes in Po 2, we can report for the first time that cone oxygen use is also reduced by light. We cannot exclude the possibility that larger changes might have been observed if the illumination had been greater. It also appears that dark-adapted foveal oxygen consumption per unit area of tissue is somewhat lower than parafoveal consumption. Estimates of peak foveal cone density are in the range of 100, ,000/mm 2, 10 but the steepness of the cone density distribution makes it impossible to know the cone density near our electrode and, therefore, the consumption of an individual cone. We can suggest that the lower oxygen consumption and the slightly higher minimum Po 2 in the distal retina may make the monkey retina (foveal and parafoveal) somewhat less susceptible to hypoxia than the cat retina. Because of poor regulation in the choroidal circulation during hypoxia, however, we would expect the minimum Po 2 in the distal retina to reach zero at an arterial Po 2 of approximately mmhg. At lower arterial Po 2, we would expect to find decreases in primate photoreceptor oxygen consumption, as we have observed in the cat. 3 We have no reason to suspect a particular sensitivity of the fovea to hypoxia and, in fact, the still lower oxygen use there would be protective, perhaps contributing to the sparing of central vision during hypoxia. 12 Key Words macaque, oxygen, oxygen consumption, retina, retinal metabolism

6 Oxygen Distribution in the Macaque Retina 521 References 1. Alder VA, Cringle SJ and Constable IJ. The retinal oxygen profile in cats. Invest Ophthalmol Vis Sci. 1983;24: Linsenmeier RA. Effects of light and darkness on oxygen distribution and consumption in the cat retina. J Gen Physiol. J986;88: Linsenmeier RA, Braun RD. Oxygen distribution and consumption in the cat retina during normoxia and hypoxemia. / Gen Physiol. 1992;99: Pournaras CJ, Riva CE, Tsacopoulos M, Strommer K. Diffusion of O 2 in the retina of anesthetized miniature pigs in normoxia and hyperoxia. Exp Eye Res. 1989;49: Yancey CM, Linsenmeier RA. Oxygen distribution and consumption in the cat retina at increased intraocular pressure. Invest Ophthalmol Vis Sci. 1989;30: Linsenmeier RA, Yancey CM. Improved fabrication of double-barreled recessed cathode O 2 microelectrodes. J Appl Physiol. 1987;63: Haugh LM, Linsenmeier RA, Goldstick TK. Mathematical models of the spatial distribution of retinal oxygen tension and consumption including changes upon illumination. Ann Bioined Eng. 1990; 18: Valeton JM, van Norren D. Fractional recording and component analysis of primate LERG: separation of photoreceptor and other retinal potentials. Vision Res. 1982;22: Schein SJ. Anatomy of macaque fovea and spatial densities of neurons in foveal representation. J Covip Neurol. 1988;269: Packer O, Hendrickson AE, Curcio CA. Photoreceptor topography of the retina in the adult pigtail macaque (Macaca nemestrina). J Covip Neurol. 1989; 288: Steinberg RH, Reid M, Lacy PL. The distribution of rods and cones in the retina of the cat (Eelis domesticus).j Comp Neurol. 1973; 148: Ernest JT, Krill AE. The effect of hypoxia on visual function: psychophysical studies. Invest Ophthal Vis Sci. 1971;10: