The function of the retina in the perfused eye

The data show that the enucleated eye of the cat can be maintained in apparently physiologically functioning condition by appropriate arterial perfusion. Under appropriate conditions, photically evoked electrical mass responses can be recorded from various parts of the isolated, perfused eye for 8 to 10 hours. ERGs as well as responses from axonal bundles of the optic nerve exhibit shapes, amplitudes and time courses comparable to their counterparts in vivo. Homeostasis of the perfusion ensures the stability of these light-evoked electrical responses. Transient changes in biophysical parameters of the perfusate rapidly induce marked, although reversible, changes in the amplitudes of b-waves of the ERGs. Increases or decreases in the flow rate of the perfusate induce parallel increases or decreases in the amplitudes of the b-waves as well as of the optic nerve responses. Similar alterations in the oxygen concentration of the perfusate induce similar and proportional changes in the amplitudes of the b-waves. It is concluded, that low flow rates of hemoglobin-free perfusate induce hypoxia; consequently, acceleration of the flow can compensate for hypoxia in a certain range. Previous studies on the effects of and recovery after transient hypoxia in mammalian retina are in concordance with the present data. Progressive decrease of temperature induces gradual and reversible reductions in the amplitudes of the b-waves and increases their latencies and peak-times. It is suggested, that initial hypothermia, which occurs during the period of cannulation, reduces the deliterious effects of the coincident unavoidable hypoxia on retinal neuronal elements. Since light-evoked electrical responses can be maintained for many hours in these preparations and since movements of cardiovascular and respiratory origin, invariably present to varying extent in the in vivo experiments, are eliminated, this preparation is suitable for intracellular recordings from neuronal elements of the retina. Potentials were recorded from cells in various layers of the retina of the cat; intracellular recordings from horizontal cells (S-potentials) are described in detail. Spectral analysis of S-potentials allowed to distinguish between three types according to their inputs: a mixed, rod-cone type, which was most frequently encountered, a pure cone- and a pure rod-type. Light- and electronmicroscopic investigation of the retina after perfusion revealed that (1) the extent of cellular damage depends on the flow rate of the perfusate; (2) little cellular damage is observed if medium flow rates, which maintain physiologic responsiveness of the isolated eye to light, were applied for two hours; (3) high flow rates applied for two hours, or medium flow rates applied for 7 hours appear to induce cystic changes in the pigment epithelium, but only minor changes in the cells of the inner nuclear layer.