Search

A novel microscope system is presented for observation of corneal cells in a living mouse. It enables tracking of individual cells in all layers of the cornea at various times, thus allowing the generation of time-lapse recordings. The system consists of three major components: an upright fluorescence microscope for visualization of corneal cells, a mouse-holding unit for immobilization of the animal and the eye, and a set of gimbals which permit observation of a wide area of corneal surface without refocusing. The same cells could be observed at different limes with the help of fiducial marks in the cornea, allowing their changes in position to be determined under natural and experimental conditions. This technique should prove useful in investigation of the cell movement in normal and diseased corneas, including the study of wound healing after an injury or surgery.

The author has done pretty much what he wanted to do throughout his professional life. Little harm resulted. A few findings may survive.

A bench specular microscope, modified to perform as a scanning microfluorometer, was used to make simple and precise measurements of the absorption of visible light by the isolated rabbit cornea. The tissue was scanned while identical FITC dextran solutions were perfused over both the surfaces, and the measured fluorescence levels in the solutions on the two sides were compared. The alteration in level corresponds to the difference between the absorption of the cornea and that of of an equal thickness of the solution, which was determined in a separate experiment. The absorption of light at 500 nm ranged from 0-8% with a mean of 3.3% (n = 20; S.E. = 0.52). Pigmented and non-pigmented rabbits did not show any clear difference in absorption. These results suggests that light absorption is not a major source of error in anterior segment fluorophotometry.

The rate of disappearance from the tear film of fluorescein and rhodamine dextran instilled together into the human eye was compared. In many cases fluorescein disappeared more rapidly, which was attributed to its penetration across the conjunctival surface. A corresponding mean fluorescein permeability across this surface of 2.5 x 10(-5) cm min-1 was calculated. This route of loss of fluorescein from the tears leads to an average overestimate of tear turnover of 25%.

An account is provided of developments in our understanding of the mechanism of corneal hydration control, particularly as regards the possibility of an active system for its regulation. Emphasis is given to issues that are contentious, such as the role of bicarbonate in the endothelial pump and the significance of water channels in both corneal limiting cell layers.

The hypothesis is advanced that the purpose of the eyeball movements during REM sleep is to stir the aqueous humor behind the closed lids and so avoid the risk that its stagnation could cause corneal anoxia. The relevance of the hypothesis to evolutionary biology and intensive care nursing is discussed.

A very small drop of fluorescein solution was placed at the bottom of the lower fornix in human volunteers and its appearance in the tear film was measured. This was quite variable, but its appearance time averaged about 4 min and the peak of fluorescence 8 min. The time was shortened by blinking. The appearance time was too fast to be accounted for by simple diffusion in an unmixed tear fluid. Under normal circumstances, the release of the dye from the fornix is faster than its loss from the conjunctival sac, so that it does not seem to be a controlling factor in this loss.

The reaction of the precorneal tear film of the human eye was determined non-invasively by instilling pyranine, a pH-sensitive fluorescent dye. The mean value was 7.83 (S.D. +/- 0.10) and it takes up this value immediately on opening the eye after the lids had been kept closed. The HCO3 system seems to be responsible for only a portion of the buffering power of the tear film. When a drop buffered to pH 6.4 with 0.075 M PO4 was instilled, the tears returned to their normal value in about 7 min, consistent with the washout times of solutes in the conjunctival sac. A pH of 7.3 was established in the tear fluid behind contact lenses either gas permeable or impermeable, probably as a result of their restricting the loss of CO2 from the eye.

Strips of rabbit stroma were prepared, and the force required to tear them apart along their length was determined. This amounted to an average of 10 g mm −1 width of tissue and it is independent of the depth of the plane in which the splitting of the cornea takes place. It could not be determined, however, whether the cohesive strength of the tissue is due to occasional collagen fibrils binding it together, interweaving of the lamellae or enmeshing of the collagen fibrils by ground substances. Other corneas were split by blunt dissection in vivo and allowed to recover for various periods of time, when the reformed strength of the split was measured. It was found to be negligible for about 5 days and rose quickly to 0·25–0·5 of the value of the untouched cornea. No increase in the force of adhesion of the split stroma was observed if an extract of corneal epithelium or a suspension of platelets was injected into the wound.

The permeability of the corneal endothelium and its aqueous-cornea distribution ratio were reevaluated in the rabbit eye. Both parameters were determined in an individual eye by applying the dye first by iontophoresis and then by intravitreal injection, which allows the influence of fluorescein glucuronide on the fluorophotometric measurements to be excluded. The corneal endothelial permeability coefficient was 5.13 +/- 1.64 X 10(-4) cm min-1, and the aqueous-cornea distribution ratio was 0.25 +/- 0.06 (mean +/- S.D., n = 11) on the average, and the former was considerably greater than the previous results, while the latter was considerably smaller.

Fluoresceinated dextran was injected into the vitreous cavity of rabbits and its concentration was followed in the vitreous and aqueous humors for a period of weeks by slit-lamp fluorometry. The concentrations in both fluids fell in parallel and in an exponential manner where the time constant was determined by the molecular weight of the dextran. The rate of movement of the dextran from the vitreous into the anterior chamber is controlled by passive diffusion and may be assumed to be relatively steady; accordingly, the faster the rate of flow in the aqueous humor the lower will be the concentration of dextran in it. Circadian variations in the aqueous flow were observed, a drop at night time being noted. These were small enough for changes due to systemic influences to be determined; in particular, it was found that general anesthesia with ketamine had no significant effect. The ratio of the concentration in the anterior chambers of the two eyes of one animal normally remains constant with time. Changes in the ratio can be used to determine the influence of unilateral interventions. This was illustrated by the use of adrenergic drops which led to a brief rise in flow rate and then a longer-lasting fall, and by X-ray irradiation with 1000 rads, which caused an immediate fall in flow rate interrupted for a time by a phase of increased flow. An absolute estimate of the aqueous flow can be derived from measurements of the total amount of dextran in the eye and its concentration in the anterior chamber. A definite value has not yet been established owing to ambiguities in determining the former quantity.

The time taken to cross the rabbit corneal epithelium and stroma was estimated for fluorescein (F), carboxyfluorescein (CF), rhodamine B (RB), and sulforhodamine B (SRB). Paired corneas were mounted in vitro; one was intact and the dye solution was kept in continuous contact with its epithelial surface; the epithelium was scraped from the other and the dye was applied as a pulse to the bare stroma. The time course of the dye appearing in a solution rapidly passing over the endothelial surface was determined by fluorometry. This rate of appearance was compared in the two cases and used to estimate the diffusional lag time introduced by the epithelium. For the very hydrophilic CF and SRB, the delay was too short to measure; this is compatible with the passage of these dyes taking place through the paracellular spaces. For the very lipophilic RB, the delay was about 2 min; this was rather too slow for it to be explained as being controlled entirely by diffusion in the cytosol. For the intermediate F, the delay was 5 min; it is suggested that this is a result of it partitioning between the spaces and the cytosol during its passage. The experiments also led to determinations of the permeability of the epithelial and endothelial layers to the dyes. In both cases lipophilicity was a strong determinant of penetration, but not the only one. The permeability of the endothelium to F was unchanged from its in vivo value in these experiments, but that of the epithelium was increased four-fold. The diffusion rate of the dyes across the stroma could also be determined. There was no clear relationship with molecular size or partition coefficient. The rate of diffusion of F across the tissue was about half that in its plane, as determined in previous experiments. This is possibly a result of the anisotropic structure of the tissue.

Regions of beef sclera free from perforations were mounted between two chambers and the movement of substances from one side to the other determined. All dyestuffs tested were virtually impermeant except for fluorescein and acid fuchsin. A variety of ions and solutes up to the size of serum albumin were also tested and were found to diffuse across about three times more slowly than they do in the rabbit corneal stroma. This rate of diffusion would slow the passage of locally applied drugs through the selera to an extent compatible with the delay in their action on the pupil.

The exchange of Na 22 and Cl 36 was measured between the blood and fluid contained in a basin formed above the eye by pulling up the lids, in rabbits. The permeability of the conjunctiva was approximately 5×10 −3 cm/hr to both ions in either direction. The electrical potential on the conjunctival surface was measured in man and the rabbit and found to be about −15 mV negative to the blood on the average, with values twice as great being recorded. The possible significance of this as representing a mechanism for absorption of the lacrimal fluid is discussed.

A fine saline-filled cannula connected to a manometer of very low volume displacement and leakage is introduced into an excised corneal stroma. A negative pressure is recorded equal to the swelling pressure of the tissue at all value of its hydration. When the cornea of a living rabbit is cannulated, either with a needle or with an indwelling tube, a negative pressure is developed which, allowing for the intraocular pressure, corresponds to the swelling pressure of the excised tissue. This shows that the stroma has a tendency to swell in vivo which must be countered by an active transport mechanism located in the limiting layers of the cornea. In the discussion it is shown that only a fraction of the imbibition pressure can arise from osmotic forces. The remainder probably arises from internal structural forces required to maintain the transparency of the tissue.