Cornea Endofhelial Rose Bengal Photosensitizafion

Transcription

1 Cornea Endofhelial Rose Bengal Photosensitizafion Effect on Permeability, Sodium Flux, and Ultrastructure David S. Hull,* Keirh Green,*f and Lisa Laughter* Rabbit corneal endothelial cells demonstrated an increase of endothelial membrane permeability to inulin following perfusion with 5 X 10~ 6 M rose bengal and exposure to 1050 /txw/cm 2 incandescent light for either 1 min or 5 min when compared to corneas similarly perfused with rose bengal but not exposed to light. Endothelial permeability to dextran was unaltered following the photosensitization reaction. Rose bengal perfusion in the absence of light caused an increase in the unidirectional sodium fluxes (J stroma endothelium and J endothelium stroma); however, a net sodium flux was still present and directed from the stroma to the endothelium. Exposure of rose bengal perfused endothelial cells to incandescent light resulted in a further increase in the unidirectional passive sodium flux in both directions and a reduction in the active component of the sodium flux from stroma to endothelium. This caused a reduction in the net active transport of sodium across the endothelium following the photosensitization reaction. Transmission electron microscopy showed no alteration of endothelial cell ultrastructural integrity following rose bengal perfusion in the absence of light. However, rose bengal perfusion accompanied by exposure to 1050 /iw/cm 2 incandescent light caused swelling of endoplasmic reticulum and mitochondria. The data indicate that rose bengal photosensitization has an adverse effect on endothelial cell energy production as well as transport and barrier function. Invest Ophthalmol Vis Sci 25: , 1984 Corneal endothelial cells perfused in vitro with 5 X 10~ 6 M rose bengal and exposed to 25 watt incandescent light (1050 jiw/cm 2 ) undergo a wavelength, and oxygen-dependent alteration of physiologic function that results in corneal swelling. 1 " 3 The corneal swelling rate is dependent upon both the rose bengal concentration as well as the duration of light exposure. 1 Additional work has shown that net endothelial bicarbonate flux, which may be linked with the extrusion of fluid from the corneal stroma, 4 5 was reduced following the photosensitization reaction. 2 The photodynamic effect is, in part, secondary to functional alteration of the cells, caused by hydrogen peroxide produced during the dismutation reaction of superoxide free radical that is catalyzed by superoxide dismutase. 1 The photodynamically induced alteration of cornea endothelial cell function has been demonstrated to have no relationship to cornea endothelial intracellular total, or percent oxidized, glutathione. 6 From the Departments of Ophthalmology* and Physiology,! Medical College of Georgia, Augusta, Georgia. Supported in part by research grant EY (Dr. Hull) and EY (Dr. Green) from the National Eye Institute and in part by a research grant from the Georgia Lions Lighthouse Foundation, Inc., and in part by a Research to Prevent Blindness, Departmental Research Award. Submitted for publication: January 17, Reprint requests: David S. Hull, MD, Department of Ophthalmology, Medical College of Georgia, Augusta, GA It was the purpose of this study to determine the effect of the rose bengal photosensitization reaction on cornea endothelial cell ultrastructure and sodium fluxes as well as endothelial membrane permeability to inulin and dextran.. Materials and Methods Adult albino rabbits were killed with an overdose of sodium pentobarbital. The corneal epithelium was removed gently with a knife and the eyes were enucleated along with the lids and conjunctiva. The de-epithelialized corneas were mounted on specular microscope mounting rings and placed between two water-jacketed flux chambers at 37 C. 5 " 7 ' 8 The chambers were filled with Krebs-bicarbonate Ringer at ph 7.3 and 307 ± 2 mosm that during preparation was bubbled with a 97% O 2-3% CO 2 mixture. The solution also contained 0.3 mm reduced glutathione and 0.5 mm adenosine. The chamber facing the de-epithelialized corneal surface contained a teflon stirring bar and was stirred at approximately 400 rpm, using an externally-mounted horseshoe magnet attached to a small motor. Rose Bengal Photosensitization Paired corneas were used with the experimental cornea being perfused with 5 X 10" 6 M rose bengal across the endothelial surface and exposed to a 25 watt in- 455

2 456 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / April 1984 Vol. 25 candescent light at 5 cm (1050 jiw/cm 2 ) for either 1 min or 5 min. These conditions previously have been shown to induce corneal swelling. 1 " 36 Controls were perfused with an identical concentration of rose bengal across the endothelial surface, but were not exposed to light. A minimum of six experimental and six control corneas were used. Ringer dual-labeled with 1.12 jtci/ml dextran (carboxyl" 14, mol wt 60,000-90,000, New England Nuclear; Boston, MA) and 1.07 /ici/ml inulin ( 3 H, mol wt 5000, Amersham; Arlington Heights, IL) was used to replace the solution bathing the endothelial surface of corneas following the rose bengal sensitization. After addition of the solution containing the radioisotopes, the tissues were allowed to equilibrate for 1 hr. At the end of the equilibration period, the chamber facing the de-epithelialized surface or "cold" chamber (1.2 ml volume) was flushed with 5 ml of normal Ringer that was discarded. A further 5 ml flush was collected immediately in a tared vial, weighed, and sampled, providing a background or residual count. At 30-min intervals for 3 hr, the cold chamber was flushed with 5 ml of normal Ringer; each flush was collected in a tared vial, weighed, and sampled. After the final collection for the cold chambers, each "hot" chamber (1.5 ml volume) was also flushed with 5 ml of normal Ringer; the flush collected in a tared vial, weighed, and sampled, providing the experimental hot count. The Nalgene filmware counting system was used with 5 ml of Aquasol to 100 fi\ of sample. A Searle Isocap 300 scintillation counter was used for dual-labeled counting. Counts were corrected for the 10.5% 14 C spillover into the 3 H channel and the 0.5% 3 H spillover into the I4 C channel. The endothelial permeability was calculated using the following formula: K increase in counts on unlabeled side trans =. concentration of counts on labeled side X area of membrane X time This formula has the dimensions of cm/sec. 7 ' 910 Sodium fluxes were determined as previously described for bicarbonate. 2-5 Paired corneas were used, one for the stroma to aqueous flux, the other for the aqueous to stroma flux. 22 Na-labeled Ringer was used to replace the solution bathing the endothelial surface of one cornea and the de-epithelialized surface of the other cornea of each pair. After addition of the isotope, the tissues were allowed to equilibrate for 1 hr. Collections were made at 30-min intervals for 3 hr as with the dual labeled group. The samples were counted on a Tracor gamma counter. Unidirectional sodium fluxes from stroma to endothelium (J stroma endothelium) and from endothelium to stroma (J endothelium stroma) were computed and net flux (J net) was expressed as the algebraic sum of the two. In all cases, the mean of the data at 30-min intervals was computed and the standard error of the mean determined. A comparison of experimentals and controls was made with the t-test. A P level of 0.05 was utilized to determine statistical significance. In preparation for transmission electron microscopy, experimental and control corneas were mounted in the specular microscope and perfused with rose bengal as previously described. 1 Experimental corneas were exposed for 2 min to a 25 watt incandescent light at 5 cm. A water filled petri dish between the light source and the cornea was used to absorb heat. Following perfusion, corneas were fixed in 2.67% phosphate buffered glutaraldehyde, postfixed in 2% osmium tetroxide, dehydrated through graded alcohols and propylene oxide, and then placed in Spurr resin. Following overnight infiltration with Spurr resin, they were embedded in freshly prepared Spurr resin and the tissue blocks polymerized at 70 C. Tissues were then sectioned, stained, and examined. Inulin Results Rose bengal perfusion in the absence of light showed that rose bengal itself caused a decrease in endothelial inulin permeability (Fig. 1). The exposure of rose bengal to light for 1 min and 5 min resulted in a 30% and 17% increase, respectively, in endothelial inulin permeability when compared to corneas similarly perfused, but not exposed to light; P < Exposure to rose bengal in the absence of light for 5 min caused an increase in permeability relative to 1 min exposure; P < Dextran Endothelial dextran permeability was not altered by exposure of rose bengal perfused corneas to incandescent light for 1 min or 5 minutes; P > 0.05 (Fig. 1). Sodium The flux of sodium from the stroma to the endothelium (J stroma endothelium) was increased 20% and 91%, respectively, in 1 min and 5 min rose bengal dark perfused control corneas compared with corneas perfused in the absence of rose bengal (Fig. 2). Neither 1 min of light exposure nor 5 min of light exposure caused any further alteration of (J stroma endothelium). The flux of sodium from the endothelium to the stroma (J endothelium stroma) was increased 17% and 81%, respectively, in 1 min and 5 min rose bengal

3 No. 4 CORNEA ENDOTHELIAL PHOTOSENSITIZATION / Hull er ol. 457 Endothelial Permeability 5 x 10~ 6 M Rose Bengal <0 U 0) (/) E o no rose bengal rose bengal-dark control for V rose bengal-1' light rose bengal-dark control for 5' rose bengal- 5' light p <0.05 Inulin o Dextran o (Effective Molecular Radius = 14A) (Effective Molecular Radius = 38A) Fig. 1. Cornea endothelial permeability to inulin and dextran following perfusion with 5 X 1O~ 6 M rose bengal. Controls were not exposed to light whereas experimentals were exposed to a 25 watt incandescent light at 5 cm (1050 ^W/cm 2 ) for either I min or 5 min. dark perfused control corneas compared with corneas perfused in the absence of rose bengal (Fig. 2). Exposure of corneas to either 1 min of light or 5 min of light resulted in a further 28% increase in (J endothelium stroma) compared with corneas similarly perfused but not exposed to light; P < 0.05 (Fig. 2). The net flux of sodium (J net = J stroma endothelium - J endothelium stroma) was nearly abolished Sodium Flux 5 x 10" 6 M Rose Bengal E3 no rose bengal 6S rose bengal-dark control for 1' E3 rose bengal-1' light rose bengal-dark control for 5' ED rose bengal- 5' light * p < 0.05 / lenflotheliuml I J Stroma f Net = yjendaihtliumj ^ Jsnomi f Fig. 2. Cornea endothelial sodium flux following perfusion with 5 X 10 6 M rose bengal. Controls were not exposed to light, whereas experimentals were exposed to a 25 watt incandescent light at 5 cm (1050 fiw/cm 2 ) for either 1 min or 5 min.

4 458 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / April 1984 Vol. 25 Fig. 3. Cornea endothelial cells following perfusion with 5 X 10 * M rose bengal without exposure to light. The apical cell membrane, nuclei, intercellular space, mitochondria, endoplasmic reticulum, cytoplasmic homogeneity, and the apex of the intercellular junction are normal. (Original magnification XI2,546). by exposure of rose bengal perfused corneas to incandescent light for either 1 min or 5 min; P < 0.05 (Fig. 2). Infrastructure Perfusion ofcorneal endothelial cells with 5 X 10" 6 M rose bengal in the absence of light did not alter endothelial cell ultrastructural integrity (Fig. 3). Perfusion with a similar concentration of rose bengal and exposure to 1050 ^W/cm 2 incandescent light for 2 min resulted in swelling of endoplasmic reticulum and mitochondria (Fig. 4). Discussion Previous work has demonstrated that corneal endothelial cell function is altered following exposure of rose bengal perfused corneas to incandescent light. 1 " 3 The present work indicates that endothelial permeability to inulin is increased following the photosensitization reaction. The permeability to the larger dextran molecule was unaltered by the photosensitization reaction. This would indicate that the size of the photodynamically-induced lesion in the endothelial membrane is large enough to allow enhanced penetration of inulin (effective molecular radius, 14 nm but not large enough to influence dextran permeability (effective molecular radius, 38 nm). In this experiment, the flux of sodium from stroma to endothelium (J stroma endothelium) of corneas perfused without rose bengal was about 20% active (J net/j stroma endothelium) and about 80% passive. Perfusion of corneas with rose bengal in the dark caused a marked increase of (J stroma endothelium) that reflects an increase in passive permeability. Exposure of rose bengal perfused corneas to light caused a reduction of the active component of (J stroma endothelium). This effect, coupled with a further increase in the passive component of (J stroma endothelium), as well as a 28% increase in the totally passive (J endothelium

5 No. 4 CORNEA ENDOTHEUAL PHOTOSENSITIZATION / Hull er ol. 459 Fig. 4. Cornea endothelial cells following perfusion with 5 X 1O~ 6 M rose bengal and a 2 min exposure to 25 W incandescent light that resulted in a 34/^m increase in corneal thickness. The apical cell membrane is intact. The endoplasmic reticulum and mitochondria are swollen. (Original magnification X 17,425). 'ft * «V stroma) resulted in a near-total abolishment of net trans-endothelial sodium flux following the photo reaction. A previous study evaluated the effects of rose bengal photosensitization on endothelial cell bicarbonate movement. The terminology used to describe fluxes in that study is slightly different from that used in the present study; however, rose bengal photosensitization was found to increase bicarbonate permeability and reduce active bicarbonate flux. Perfusion of corneal endothelial cells with 5 X 1CT 6 M rose bengal in the absense of light caused no alteration of endothelial cell ultrastructural integrity. However, perfusion with a similar concentration of rose bengal and exposure to incandescent light for 2 min resulted in swelling of the endoplasmic reticulum and mitochondria. It appears from the information found in this study, coupled with information found in previous studies, that rose bengal photosensitization of the rabbit cornea endothelium has an adverse effect on cell physiology and anatomy. 1 " 36 This adverse effect results in a compromise of endothelial cell energy production, water movement, and ionic transport as well as overall membrane barrier function. It is currently not known whether the rose bengal model represents an acceleration of the effects of a lifetime of light exposure of an individual cornea. However, the progressive endothelial cell loss noted with increasing age is provoking, and continued studies relating light exposure to endothelial cell function seem warranted. Key words: rose bengal, photosensitization, cornea, endothelium, sodium, inulin, dextran

6 460 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / April 1984 Vol. 25 References 1. Hull DS, Strickland EC, and Green K: Photodynamically induced alteration of cornea endothelial cell function. Invest Ophthalmol VisSci 18:1226, Hull DS, Green K, Csukas S, and Livingston V: Photodynamic alteration of cornea endothelium: relation to bicarbonate fluxes and oxygen concentration. Biochim Biophys Acta 640:231, Hull DS, Csukas S, and Green K: Rose bengal induced corneal swelling: relation to inciting wavelength. Curr Eye Res 1:487, Mayes K.R and Hodson S: Local osmotic coupling to the active transendothelial bicarbonate flux in the rabbit cornea. Biochim Biophys Acta 514:286, Hull DS, Green K, Boyd M, and Wynn HR: Corneal endothelium bicarbonate transport and the effect of carbonic anhydrase inhibitors on endothelial permeability, fluxes and corneal thickness. Invest Ophthalmol VisSci 16:883, Hull DS, Csukas S, and Green K: Corneal endothelial glutathione after photodynamic change. Invest Ophthalmol Vis Sci 22:405, Kim JH, Green K, Martinez M, and Paton D: Solute permeability of the corneal endothelium and Descemet's membrane. Exp Eye Res 12:231, Hull DS, Green K, and Bowman K: Corneal endothelial permeability following storage in moist chamber or MK medium. Acta Ophthalmol 57:999, Maffly RH, Hays RM, Lamdin E, and Leaf A: The effect of neurohypophyseal hormones on the permeability of the toad bladder to urea. J Clin Invest 39:630, Hull DS, Farkas S, Green K, Laughter L, Elijah RD, and Bowman K: Radial keratotomy: effect on cornea and aqueous humor physiology in the rabbit. Arch Ophthalmol 101:479, 1983.