ACTA OPHTHALMOLOGICA VOL. 5 3 1975
The University Eye Department (Head: T h . L. Thomassen, M.D.) and the Institute o\ Pathology, Electron Microscopic Laboratory (Head: T . Hovig, M.D.), Rikshospitalet, Oslo, Norway
PERMEABILITY OF RABBIT CORNEAL EPITHELIUM TO HORSERADISH PEROXIDASE AFTER THE INFLUENCE OF BENZALKONIUM CHLORIDE BY
ASBJ0RN M. T0NJUM
A tight barrier against permeation of horseradish peroxidase into the corneal epithelium exists at the corneal surface adjacent to the tear film. The present light and transmission electron microscopic study reveals that the cationic surfactant, benzalkonium chloride, which is commonly added to eye drops as a preservative, breaks down this barrier. Lysis of the cell membranes was demonstrated, resulting in a leakage of the tracer into and underneath the superficial cells. The grade of cellular destruction caused by benzalkonium chloride was dependent upon the concentration and exposure time of the drug.
K e y words: permeability - cornea - epithelium - horseradish peroxidase - cationic surfactant - benzalkonium - rabbit.
In a previous paper Tsnjum (1974) demonstrated that horseradish peroxidase, having a molecular weight of about 40,000, did not enter the normal rabbit epithelium from the anterior side. A tight barrier was present between the most superficial cells adjacent to the tear film. A lesion involving only the superficial cell layer of the epithelium might, therefore, tend to increase the permeability of water, solutes, and even macromolecules. ~~~
~
~~~
Received January 22, 1975.
335
Asbjern M . Tsnlum
Fig. 1.
Light micrographs of rabbit corneal epithelium embedded in Epon and stained with toluidine blue. a) Epithelium exposed to horseradish peroxidase in vitro for 10 min. b) Epithelium exposed to 0.02 O/o benzalkonium chloride for 5 min and to horseradish peroxidase for 10 min in vitro. Disintegration of superficial cells and peroxidase reaction products inside the cells at several layers. c) Epithelium exposed to BA 0.01 'O/o for 2 min and horseradish peroxidase for 10 min. Disintegration of the most superficial cell layer with peroxidase reaction products inside these cells. x 470.
The cationic surfactant, benzalkonium chloride (BA), which is commonly used as a preservative in eye drops, has been shown to increase the action of drugs applied topically (O'Brien & Swan 1942, Swan 1944) and the permeability of fluorescein across the cornea in vitro (Green & Tenjum 1971). I t has also been demonstrated that BA enhances the penetration of prednisolone phosphate into the eye in vitro (Green & Downs 1974). Another cationic surfactant, cetyl pyridinium chloride, has been shown to increase the permeability of isolated frog skin (Webb 1965). T h e purpose of the present paper was to investigate whether benzalkonium chloride would influence the permeability of the corneal epithelium to the protein molecule, horseradish peroxidase.
Materials and Methods Healthy, adult albino rabbits weighing 3-4.5 kg were used. Their eyes were found to be normal prior to experimentation. The in vitro studies were done in a similar manner as described by Tenjum (1974). The excised corneas were clamped between two lucite chambers and BA 336
Benznlkonium and Corneal Permeability
a t a concentration of 0.02 O/O in Krebs-Ringer solution was placed in the epithelial side chamber for 5 min, and 0.01 ' / o BA for 1 or 2 min. After these exposure periods the BA was removed and this chamber was washed with KrebsRinger solution. T h e epithelial side chamber was then filled with Krebs-Ringer solution to which was added horseradish peroxidase Sigma type I1 (PO) at a concentration of 2.5 mgiml and left for 10 min. T h e same procedures were done in parallel experiments, the only difference being the exclusion of BA. T h e endothelial side chamber was filled with Krebs-Ringer solution throughout the experiment. T h e in vivo experiments were done by applying BA a t a concentration of 0.02 o / o in Krebs-Ringer solution topically into a rabbit eye, one drop 3 or 10 times with 30 sec intervals. Thereafter one drop of PO in Krebs-Ringer solution a t a concentration of 2.5 mgiml was applied every half min for 10 min.
F i g . 2.
Normal rabbit epithelium exposed to horseradish peroxidase for 10 min in vitro. No peroxidase reaction products underneath the surface. Stained with alcaline lead. MV: microvilli. IS: intercellular spaces. Numerous dark glycogen granules. x 24,000. 337
Asbjsrn M . Tsnjum
F i g . 3. Rabbit corneal epithelium exposed to benzalkonium chloride 0.02 O i o in uitro for 5 min and horseradish peroxidase for 10 min. Disintegration of several cell layers with peroxidase reaction products inside the superficial and the wing cells. Stained with alcaline lead. x 18,000.
Benzalkoniicm and Corneal Permeability
The corneas were then removed from the chambers or excised from the animals, which were killed with an overdose of sodium pentobarbital. The corneas were quickly washed in Krebs-Ringer solution, and prefixed in 2&010 glutaraldehyde. The tissues were then incubated with diaminobenzidine (DAB) and hydrogen peroxide, postfixed in osmium tetroxide, dehydrated in ethanol, and embedded in Epon 812. Semithin sections of about 1 pin in thickness for light microscopy were cut with an LKB Ultratome and examined either stained or unstained with toluidine blue. T h e ultrathin sections were either left uncontrasted or were contrasted with alcaline lead and/or uranyl acetate, and electron micrographs were taken with a Siemens Elmiskop 1 A.
Observations The in vitro experiments
a) Exposure of 0.02 010 BA to the corneal surface for 5 min. There was a gross
disintegration of the cells in the superficial epithelium, with numerous defects of the cell membrane a t the surface (Figs. 1 and 4). The cytoplasma had leaked out of the cells, whereas PO was demonstrated inside. The intercellular spaces were widened and the cell membranes were stretched corresponding to the desmosomes, probably due to influx of fluid. Accumulations of circular structures were found, particularly at the site near the nucleus (Fig. 3). In the wing cell layer the architecture of the cells was fairly well preserved, with only slight intercellular oedema. However, large amounts of peroxidase reaction products (PORP) were present within the cytoplasma and the nuclei. Between the wellpreserved deep wing cells and the basal cells there were traces of PORP, but no gross intercellular oedema. b) Exposure of 0.01 010 BA to the corneal surface for 2 min. The ultramicroscopic changes were, in principle, equal to those of the previous experiment, but were limited to the most superficial cell layer. This was, however, completely disorganized, with defects of the anterior cell surface, presence of PORP inside these cells, and lysis of the intracellular organelles including the nuclei (Fig. 5 ) . In the deeper layers PORP were demonstrated in the intercellular spaces (Fig. 6). c) Exposure o f 0.01 O f 0 BA to the corneal surface for I min. The structural changes found in these experiments were moderate. Even some cells of the corneas not treated with BA appeared to be dead, and contained PORP. In the present set of experiments some more cells appeared to be dead, and others had relatively few microvilli (Fig. 7 ) . 339
Asbjern M. Tenjum
Fig. 4. Rabbit corneal epithelium exposed to benzalkonium chloride 0.02 O/o for 5 min and horseradish peroxidase for 10 min in vitro. Large amounts of peroxidase reaction products inside the superficial and wing cells and between the deep wing cells. Intercellular oedema (IS) with stretching of the cell membranes at the numerous desmosomes (D). Defects of the superficial cell membranes (arrows). Stained with alcaline lead. x 27,000.
The in vivo experiments
a) When 0.02 O/o BA was applied as eye drops 10 times, i.e., over a period of 5 min, severe damage of the two superficial layers of the epithelium was de340
Benzalkonium and Corneal Permeability
monstrated. Relatively small amounts of PORP were seen within the cells and in the intercellular spaces of the deeper layers (Fig. 8). b) When 0.020/0 BA was applied 3 times, i.e., 1 min exposure, no definite structural changes ascribed to BA could be detected.
Comments Benzalkonium chloride is commonly used as a preservative in eye drops at concentrations known to range from 0.0040/0 to 0.020/0. The molecules vary in size, as the alkyl chain may consist of 8 to 18 carbon atoms. The present study showed that benzalkonium chloride in concentrations used
Fig. 5. Rabbit corneal epithelium exposed to benzalkonium 0.01 O/o for 2 min and horseradish peroxidase for 10 min in vitro. Damage of the superficial cell layer. Alcaline lead. x 20,000.
34 1 Acta ophthal. 53, 3
23
Asbjern M . Tenjiim
in eye drops, after prolonged exposure to the eye, produces a disintegration of the cellular architecture and function. The degree of the changes was dependent upon the concentration and exposure time of the BA.
Fig. 6. Rabbit corneal epithelium exposed to benzalkonium chloride 0.01 O/o for 2 min and horseradish peroxidase for 10 min in vitro. Uncontrasted. Peroxidase reaction products within the superficial cells and between the deeper cells (arrows). x 18.000.
342
Benzalkonium and Corneal Permeability
Fig. 7. Rabbit corneal epithelium exposed to benzalkonium chloride 0.01 O / o for 1 min and horseradish peroxidase for 10 min in nitro. No gross lysis of cell membranes or organelles. JC: tight intercellular junctional complex. The left cell has a straight surface with few microvilli. The right one appears undamaged. Alcaline lead. x 20,000.
343
Asbjsrn M . Tsnjum
Fig. 8. Rabbit corneal epithelium after topical application of benzalkonium chloride 0.02 O/o 10 times with 30 sec intervals and horseradish peroxidase 20 times in v i m Gross lysis of the superficial layers. Alcaline lead. x 18,000.
344
Benzalkonitcm and Corneal Permeability
I t is difficult to compare the effect of BA in the in vivo and in vitro experiments, but it appeared that the cornea in vivo was more resistent to the destruction caused by BA than in vitro, even taking into account the dilution of the drug which takes place in vim. A possible explanation for this might be the protective effect of the tear film in vivo. The in v i t r o experiments showed that exposure to 0.010/0 BA for 2 min damaged cells of the entire surface. Exposure for 1 min left some cells relatively unharmed, whereas others were affected. Exposure to BA 0.01 o/o for 1 min in vitro, therefore, seems to approach the lethal toxicity of BA to corneal epithelial cells. That all cells are not affected to the same degree may be due to different vulnerabilities to BA in different stages of the life cycle of corneal epithelial cells, which have a high turnover rate.
Fig. 9 . Rabbit corneal epithelium alter topical application of benzalkonium chloride 0.02 O/o 3 times and horseradish peroxidase 20 times, in vivo. No gross damage of the ultrastructure. The microvilli seems less prominent than in epithelium untreated with benzalkonium. Alcaline lead. x 24,000.
345
Asbjorn M . Tsnjum
Even when only the superficial cell layer is damaged, the barrier function of the epithelium is broken down, and PORP could be found in the intercellular spaces of the deep cell layers (Fig. 6). T h e PORP staining might have been more intensive with a longer exposure time for PO. Another problem is visualizing PO in the deeper layers when the tissues are incubated en bloc because of the poor penetration of diaminobenzidine to these tissues. In the superficial cells, isolated or aggregated circular structures, or vesicles, were present. They may reflect the general tendency of membranes to form globules when being disintegrated, or they may be swollen intracellular organelles. T he mechanism by which BA breaks down the permeability barrier appears to be the lytic effect upon the cellular membrane, and the intracellular membranes as well, allowing fluid and solutes to enter the cells. Eventually the disintegration reaches the stage when the cell contents leak out, leaving a shell consisting of a defect cell membrane. It is reasonable to assume that long term clinical use of BA enchances the turnover rate of the corneal epithelial cells. It might also be assumed that BA penetrates into the deeper corneal tissues, affecting other cells such as the keratocytes and endothelial cells. On the other hand, in some clinical conditions it might be an advantage, at least for a limited period of time, to enhance the penetration of a drug into the eye, particularly the water soluble drugs. It is also important to be aware of BA activity when comparing the effects of different drugs upon the eye.
Acknowledgement Financial support from the Norwegian Research Council for Science and Humanities, and from Aase and Knut Tenjum, Tensberg, is gratefully acknowledged.
References Green, K. & Downs, S. J. (1974) Prednisolone phosphate penetration into and through the cornea. Invest. Ophthal. 13, 316-319. Green, K. & Tonjum, A. M. (1971) Influence of various agents on corneal permeability. Amer. J . Ophtlial. 72, 897-905. O’Brien, C. S. & Swan, K. C. (1942) Carbaminoylcholine chloride in the treatment of glaucoma simplex. Arch. Ophthal. 27, 253-263.
346
Benzalkonium and Corneal Permeability Swan, K. C. (1944) Reactivity of ocular tissues to wetting agents. Amer. /. Oplithal. 27, 1 1 18-1 122. Tenjum, A . M. (1974) Permeability of horseradish peroxidase in the rabbit corneal epithelium. Acta oplitlznl. (Kbli.) 52, 650-658. Webb, G. D. (1965) The effects of surfactants on the potential, short-circuit current, and ion fluxes across the isolated frog skin. A d a physiol. scarid. 63, 377-384.
Author’s nddress: Asbjern M. Tsnjum, M.D., University Eye Department, Rikshospitalet, Oslo 1 , Norway.
34 7
The University Eye Department (Head: T h . L. Thomassen, M.D.) and the Institute o\ Pathology, Electron Microscopic Laboratory (Head: T . Hovig, M.D.), Rikshospitalet, Oslo, Norway
PERMEABILITY OF RABBIT CORNEAL EPITHELIUM TO HORSERADISH PEROXIDASE AFTER THE INFLUENCE OF BENZALKONIUM CHLORIDE BY
ASBJ0RN M. T0NJUM
A tight barrier against permeation of horseradish peroxidase into the corneal epithelium exists at the corneal surface adjacent to the tear film. The present light and transmission electron microscopic study reveals that the cationic surfactant, benzalkonium chloride, which is commonly added to eye drops as a preservative, breaks down this barrier. Lysis of the cell membranes was demonstrated, resulting in a leakage of the tracer into and underneath the superficial cells. The grade of cellular destruction caused by benzalkonium chloride was dependent upon the concentration and exposure time of the drug.
K e y words: permeability - cornea - epithelium - horseradish peroxidase - cationic surfactant - benzalkonium - rabbit.
In a previous paper Tsnjum (1974) demonstrated that horseradish peroxidase, having a molecular weight of about 40,000, did not enter the normal rabbit epithelium from the anterior side. A tight barrier was present between the most superficial cells adjacent to the tear film. A lesion involving only the superficial cell layer of the epithelium might, therefore, tend to increase the permeability of water, solutes, and even macromolecules. ~~~
~
~~~
Received January 22, 1975.
335
Asbjern M . Tsnlum
Fig. 1.
Light micrographs of rabbit corneal epithelium embedded in Epon and stained with toluidine blue. a) Epithelium exposed to horseradish peroxidase in vitro for 10 min. b) Epithelium exposed to 0.02 O/o benzalkonium chloride for 5 min and to horseradish peroxidase for 10 min in vitro. Disintegration of superficial cells and peroxidase reaction products inside the cells at several layers. c) Epithelium exposed to BA 0.01 'O/o for 2 min and horseradish peroxidase for 10 min. Disintegration of the most superficial cell layer with peroxidase reaction products inside these cells. x 470.
The cationic surfactant, benzalkonium chloride (BA), which is commonly used as a preservative in eye drops, has been shown to increase the action of drugs applied topically (O'Brien & Swan 1942, Swan 1944) and the permeability of fluorescein across the cornea in vitro (Green & Tenjum 1971). I t has also been demonstrated that BA enhances the penetration of prednisolone phosphate into the eye in vitro (Green & Downs 1974). Another cationic surfactant, cetyl pyridinium chloride, has been shown to increase the permeability of isolated frog skin (Webb 1965). T h e purpose of the present paper was to investigate whether benzalkonium chloride would influence the permeability of the corneal epithelium to the protein molecule, horseradish peroxidase.
Materials and Methods Healthy, adult albino rabbits weighing 3-4.5 kg were used. Their eyes were found to be normal prior to experimentation. The in vitro studies were done in a similar manner as described by Tenjum (1974). The excised corneas were clamped between two lucite chambers and BA 336
Benznlkonium and Corneal Permeability
a t a concentration of 0.02 O/O in Krebs-Ringer solution was placed in the epithelial side chamber for 5 min, and 0.01 ' / o BA for 1 or 2 min. After these exposure periods the BA was removed and this chamber was washed with KrebsRinger solution. T h e epithelial side chamber was then filled with Krebs-Ringer solution to which was added horseradish peroxidase Sigma type I1 (PO) at a concentration of 2.5 mgiml and left for 10 min. T h e same procedures were done in parallel experiments, the only difference being the exclusion of BA. T h e endothelial side chamber was filled with Krebs-Ringer solution throughout the experiment. T h e in vivo experiments were done by applying BA a t a concentration of 0.02 o / o in Krebs-Ringer solution topically into a rabbit eye, one drop 3 or 10 times with 30 sec intervals. Thereafter one drop of PO in Krebs-Ringer solution a t a concentration of 2.5 mgiml was applied every half min for 10 min.
F i g . 2.
Normal rabbit epithelium exposed to horseradish peroxidase for 10 min in vitro. No peroxidase reaction products underneath the surface. Stained with alcaline lead. MV: microvilli. IS: intercellular spaces. Numerous dark glycogen granules. x 24,000. 337
Asbjsrn M . Tsnjum
F i g . 3. Rabbit corneal epithelium exposed to benzalkonium chloride 0.02 O i o in uitro for 5 min and horseradish peroxidase for 10 min. Disintegration of several cell layers with peroxidase reaction products inside the superficial and the wing cells. Stained with alcaline lead. x 18,000.
Benzalkoniicm and Corneal Permeability
The corneas were then removed from the chambers or excised from the animals, which were killed with an overdose of sodium pentobarbital. The corneas were quickly washed in Krebs-Ringer solution, and prefixed in 2&010 glutaraldehyde. The tissues were then incubated with diaminobenzidine (DAB) and hydrogen peroxide, postfixed in osmium tetroxide, dehydrated in ethanol, and embedded in Epon 812. Semithin sections of about 1 pin in thickness for light microscopy were cut with an LKB Ultratome and examined either stained or unstained with toluidine blue. T h e ultrathin sections were either left uncontrasted or were contrasted with alcaline lead and/or uranyl acetate, and electron micrographs were taken with a Siemens Elmiskop 1 A.
Observations The in vitro experiments
a) Exposure of 0.02 010 BA to the corneal surface for 5 min. There was a gross
disintegration of the cells in the superficial epithelium, with numerous defects of the cell membrane a t the surface (Figs. 1 and 4). The cytoplasma had leaked out of the cells, whereas PO was demonstrated inside. The intercellular spaces were widened and the cell membranes were stretched corresponding to the desmosomes, probably due to influx of fluid. Accumulations of circular structures were found, particularly at the site near the nucleus (Fig. 3). In the wing cell layer the architecture of the cells was fairly well preserved, with only slight intercellular oedema. However, large amounts of peroxidase reaction products (PORP) were present within the cytoplasma and the nuclei. Between the wellpreserved deep wing cells and the basal cells there were traces of PORP, but no gross intercellular oedema. b) Exposure of 0.01 010 BA to the corneal surface for 2 min. The ultramicroscopic changes were, in principle, equal to those of the previous experiment, but were limited to the most superficial cell layer. This was, however, completely disorganized, with defects of the anterior cell surface, presence of PORP inside these cells, and lysis of the intracellular organelles including the nuclei (Fig. 5 ) . In the deeper layers PORP were demonstrated in the intercellular spaces (Fig. 6). c) Exposure o f 0.01 O f 0 BA to the corneal surface for I min. The structural changes found in these experiments were moderate. Even some cells of the corneas not treated with BA appeared to be dead, and contained PORP. In the present set of experiments some more cells appeared to be dead, and others had relatively few microvilli (Fig. 7 ) . 339
Asbjern M. Tenjum
Fig. 4. Rabbit corneal epithelium exposed to benzalkonium chloride 0.02 O/o for 5 min and horseradish peroxidase for 10 min in vitro. Large amounts of peroxidase reaction products inside the superficial and wing cells and between the deep wing cells. Intercellular oedema (IS) with stretching of the cell membranes at the numerous desmosomes (D). Defects of the superficial cell membranes (arrows). Stained with alcaline lead. x 27,000.
The in vivo experiments
a) When 0.02 O/o BA was applied as eye drops 10 times, i.e., over a period of 5 min, severe damage of the two superficial layers of the epithelium was de340
Benzalkonium and Corneal Permeability
monstrated. Relatively small amounts of PORP were seen within the cells and in the intercellular spaces of the deeper layers (Fig. 8). b) When 0.020/0 BA was applied 3 times, i.e., 1 min exposure, no definite structural changes ascribed to BA could be detected.
Comments Benzalkonium chloride is commonly used as a preservative in eye drops at concentrations known to range from 0.0040/0 to 0.020/0. The molecules vary in size, as the alkyl chain may consist of 8 to 18 carbon atoms. The present study showed that benzalkonium chloride in concentrations used
Fig. 5. Rabbit corneal epithelium exposed to benzalkonium 0.01 O/o for 2 min and horseradish peroxidase for 10 min in vitro. Damage of the superficial cell layer. Alcaline lead. x 20,000.
34 1 Acta ophthal. 53, 3
23
Asbjern M . Tenjiim
in eye drops, after prolonged exposure to the eye, produces a disintegration of the cellular architecture and function. The degree of the changes was dependent upon the concentration and exposure time of the BA.
Fig. 6. Rabbit corneal epithelium exposed to benzalkonium chloride 0.01 O/o for 2 min and horseradish peroxidase for 10 min in vitro. Uncontrasted. Peroxidase reaction products within the superficial cells and between the deeper cells (arrows). x 18.000.
342
Benzalkonium and Corneal Permeability
Fig. 7. Rabbit corneal epithelium exposed to benzalkonium chloride 0.01 O / o for 1 min and horseradish peroxidase for 10 min in nitro. No gross lysis of cell membranes or organelles. JC: tight intercellular junctional complex. The left cell has a straight surface with few microvilli. The right one appears undamaged. Alcaline lead. x 20,000.
343
Asbjsrn M . Tsnjum
Fig. 8. Rabbit corneal epithelium after topical application of benzalkonium chloride 0.02 O/o 10 times with 30 sec intervals and horseradish peroxidase 20 times in v i m Gross lysis of the superficial layers. Alcaline lead. x 18,000.
344
Benzalkonitcm and Corneal Permeability
I t is difficult to compare the effect of BA in the in vivo and in vitro experiments, but it appeared that the cornea in vivo was more resistent to the destruction caused by BA than in vitro, even taking into account the dilution of the drug which takes place in vim. A possible explanation for this might be the protective effect of the tear film in vivo. The in v i t r o experiments showed that exposure to 0.010/0 BA for 2 min damaged cells of the entire surface. Exposure for 1 min left some cells relatively unharmed, whereas others were affected. Exposure to BA 0.01 o/o for 1 min in vitro, therefore, seems to approach the lethal toxicity of BA to corneal epithelial cells. That all cells are not affected to the same degree may be due to different vulnerabilities to BA in different stages of the life cycle of corneal epithelial cells, which have a high turnover rate.
Fig. 9 . Rabbit corneal epithelium alter topical application of benzalkonium chloride 0.02 O/o 3 times and horseradish peroxidase 20 times, in vivo. No gross damage of the ultrastructure. The microvilli seems less prominent than in epithelium untreated with benzalkonium. Alcaline lead. x 24,000.
345
Asbjorn M . Tsnjum
Even when only the superficial cell layer is damaged, the barrier function of the epithelium is broken down, and PORP could be found in the intercellular spaces of the deep cell layers (Fig. 6). T h e PORP staining might have been more intensive with a longer exposure time for PO. Another problem is visualizing PO in the deeper layers when the tissues are incubated en bloc because of the poor penetration of diaminobenzidine to these tissues. In the superficial cells, isolated or aggregated circular structures, or vesicles, were present. They may reflect the general tendency of membranes to form globules when being disintegrated, or they may be swollen intracellular organelles. T he mechanism by which BA breaks down the permeability barrier appears to be the lytic effect upon the cellular membrane, and the intracellular membranes as well, allowing fluid and solutes to enter the cells. Eventually the disintegration reaches the stage when the cell contents leak out, leaving a shell consisting of a defect cell membrane. It is reasonable to assume that long term clinical use of BA enchances the turnover rate of the corneal epithelial cells. It might also be assumed that BA penetrates into the deeper corneal tissues, affecting other cells such as the keratocytes and endothelial cells. On the other hand, in some clinical conditions it might be an advantage, at least for a limited period of time, to enhance the penetration of a drug into the eye, particularly the water soluble drugs. It is also important to be aware of BA activity when comparing the effects of different drugs upon the eye.
Acknowledgement Financial support from the Norwegian Research Council for Science and Humanities, and from Aase and Knut Tenjum, Tensberg, is gratefully acknowledged.
References Green, K. & Downs, S. J. (1974) Prednisolone phosphate penetration into and through the cornea. Invest. Ophthal. 13, 316-319. Green, K. & Tonjum, A. M. (1971) Influence of various agents on corneal permeability. Amer. J . Ophtlial. 72, 897-905. O’Brien, C. S. & Swan, K. C. (1942) Carbaminoylcholine chloride in the treatment of glaucoma simplex. Arch. Ophthal. 27, 253-263.
346
Benzalkonium and Corneal Permeability Swan, K. C. (1944) Reactivity of ocular tissues to wetting agents. Amer. /. Oplithal. 27, 1 1 18-1 122. Tenjum, A . M. (1974) Permeability of horseradish peroxidase in the rabbit corneal epithelium. Acta oplitlznl. (Kbli.) 52, 650-658. Webb, G. D. (1965) The effects of surfactants on the potential, short-circuit current, and ion fluxes across the isolated frog skin. A d a physiol. scarid. 63, 377-384.
Author’s nddress: Asbjern M. Tsnjum, M.D., University Eye Department, Rikshospitalet, Oslo 1 , Norway.
34 7
