Nitric Oxide Enhancement of Cholinergic Amacrine Activity by Inhibition of Glycine Release

Transcription

1 1634 Investigative Ophthalmology & Visual Science, July 1997, Vol. 38, No. 8 TABLE i. Average Fibroblast Counts (POD 11) Site S/PTGFp S/PBSS Total 21* 26* 17 26* 24 17* * 16 All counts are averages of two slides from each animal. There were eight animals in each group. POD = ; TGF/3 = transforming growth factor /?; S/P = ; BSS =. * Statistical significance compared with similar site of BSS injected eyes (P < 0.05). from reaching convincing statistical significance. However, marked statistical significance was seen with the same number of subjects after whole blood was injected. In that study there was 100% clinical success (bleb leak seal), and a statistically significant increase in peribleb fibroblasts at all sites. Thus, we suspect the more probable conclusion is that TGF-/3 is one, not the sole, chemotactic agent in this particular healing cascade. It is probably but one of many trophic factors in whole blood that promote fibroblast migration and healing of bleb leak. It is unclear why none of the four TGF-/?-treated eyes that had persistent leakage demonstrated failed blebs and elevated IOPs similar to those seen in four of the control eyes. Perhaps the application of TGF- P resulted in sufficient leak closure to allow some continued preservation of the bleb. If these eyes had been observed for a longer time, some of them may have either healed the leak, preserving the bleb, or gone on to total bleb failure with elevated IOP. Continued work in this model will involve looking into whether more than one factor in the healing cascade is involved in sealing bleb leaks after injection of whole blood. Does plasma protein diffusion to the area of a leak, with subsequent cross-linking of factors, contribute to leak seal? Additionally, the role, if any, of red blood cells is unclear. The next step in our work involves separating blood into plasma and cellular components, obtaining other independent trophic factors, and proceeding with injections as before. Key Words antiproliferative effects, bleb leak, glaucoma surgery, rabbit, transforming growth factor-/? References 1. Smith MF, Magauran RG, BetchkelJ, Doyle JW. Treatment of postfiltration bleb leaks with autologous blood. Ophthalmology. 1995; 102: Doyle JW, Smith, MF, Garcia JA, Sherwood MB, Lau T. Injection of autologous blood for bleb leaks in New Zealand White rabbits. Invest Ophthalmol Vis Sci. 1996;37: Roberts AN, Apurn MB. The transforming growth factor /?s. In: Sporn MB, Roberts AB, eds. Handbook of Experimental Pharmacology: Peptide Groioth Factors and Their Receptors. Vol. 95. Berlin: Springer-Veiiag; 1990: Massague J. The transforming growth factor-b family. AnnuRev Cell Biol. 1990;6: Grant MB, Khaw PT, Schultz GS, Adams TL, Shimizu RW. Effects of epidermal growth factor, fibroblast growth factor, and transforming growth factor-b on corneal cell chemotaxis. Invest Ophthalmol Vis Sci. 1992;33: Brown GL, Curtsinger LJ, White M, et al. Acceleration of tensile strength of incisions treated with EGT and TGF-B. Ann Surg. 1988; 208: Ishikawa O, Yamakage A, LoRoy EC, Trojanowska M. Persistent effect of TGF-B on extracellular matrix gene expression in human dermal fibroblasts. Biochem BiophysRes Commun. 1990; 169: Khaw PT, Doyle JW, Sherwood MB, Smith MF, McGorray S. Effects of intraoperative 5-fluorouracil or mitomycin-c on glaucoma filtration surgery in the rabbit. Ophthalmology. 1993; 100: Olashaw NE, Okeefe EJ, Pledger WJ. Platelet-derived growth factor modulates epidermal growth factor receptors by a mechanism distinct from that of phorodol esters. Proc Natl Acad Sci USA. 1986;86: Bennett NT, Schultz GS. Growth factors and wound healing: Biochemical properties of growth factors and their receptors. Am f Surg. 1993; 165: Nitric Oxide Enhancement of Cholinergic Amacrine Activity by Inhibition of Glycine Release Michael Neal, Jo Cunningham, and Kim Matthews Purpose. To investigate a possible interaction between cholinergic and nitrergic amacrine cells in the rabbit retina. Methods. The activity of cholinergic amacrine cells was estimated by measuring the light-evoked release of [ 8 H]-acetylcholine (ACh) from the retina of rabbits anesthetized with urethane. An eyecup was prepared and filled with Krebs-Ringer bicarbonate solution, containing [ 3 H]-choline. After washing with fresh medium containing physostigmine, 0.5 ml of medium was placed in the eyecup. The medium was replaced every 5 minutes, and the radioactivity in the resultant samples was measured. In some experi-

2 Reports 1635 merits the release of [ 3 H]-ACh and glycine was measured using isolated retinas. Results. Local application of the nitric oxide (NO) donors, S-nitroso-N-acetyl-DL-penicillamine and sodium nitroprusside strikingly enhanced the light-evoked release of [ 3 H]-ACh. In contrast, inhibition of nitric oxide synthase with L-nitromonomethylarginine (LNMMA) or N-nitro-L-arginine (LNA) greatly reduced the lightevoked release of [ 3 H]-ACh. In that the response of cholinergic amacrine cells is damped by an inhibitory feedback circuit involving glycinergic amacrine cells, the effect of strychnine on the inhibitory action of LNMMA was examined. Strychnine abolished the inhibitory effect of LNMMA on the light-evoked release of [ 3 H]-ACh, suggesting that endogenous NO normally has an inhibitory effect on glycinergic amacrine cells. This idea was supported by experiments using isolated retinas, in which sodium nitroprusside and S-nitroso-Nacetyl-DL-penicillamine inhibited the potassium-evoked release of glycine but enhanced the release of [ 8 H]- ACh. Conclusions. Endogenous NO is released in the retina and acts indirectly to facilitate the light-evoked response of cholinergic amacrine cells. Invest Ophthalmol Vis Sci. 1997; 38: A he cholinergic neurons in the rabbit retina are wellcharacterized, transiently responding amacrine cells. 1 They are symmetrically distributed in the amacrine and ganglion cell layers and ramify in the off and on sublaminae, respectively. Because of their distinctive morphology they are often referred to as starburst amacrine cells. 2 The cholinergic amacrine cells possess kainate receptors 1 and receive their excitatory input from cone bipolar cells, 2 which are thought to use glutamate as their transmitter. The activity of the (displaced) cholinergic amacrine cells in response to physiological stimulation with light can be estimated by measuring the release of [ 3 H]-acetylcholine (ACh). 1 The response of the cholinergic amacrine cells is damped by a negative feedback circuit involving glycinergic amacrine cells 3 (similar to Renshaw cells in the spinal cord). Nitric oxide synthase (NOS) has been described in many retinal cells, 4 but neuronal NOS in the rabbit seems to be located mainly in some amacrine cells. 5 ' 6 Clearly, the possibility exists that NO released from these NOS-positive amacrine cells could easily reach other cells synapsing in the inner plexiform layer and From the Division of Pharmacology, United Medical and Dental Schools of Guy's and St Thomas' Hospitals, London, United Kingdom. Submitted for publication December 3, 1996; revised January 29, 1996; accepted March 4, Proprietary interest category: N. Reprint requests: Michael Neal, Department of Pharmacology, United Medical and Dental Schools, St Thomas' Hospital, Lambeth Palace Road, London, SE1 7EH, United Kingdom. modulate their activity either by increasing cyclic guanosine monophosphate (cgmp) levels 7 " 9 or by other mechanisms. 4 In the current study we provide evidence that there is release of endogenous NO in the rabbit retina and that the NO enhances the lightevoked release of [ 3 H]-ACh indirectly by partially suppressing a glycinergic inhibitory feedback circuit. METHODS. Animal Preparation. The 65 male New Zealand White rabbits used in this study, (weight range, 2 to 2.5 kg), were cared for and treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The preparation of the animals has been described in detail. 1 Briefly, the rabbits were anesthetized with 1.5 g kg" 1 urethane, administered intraperitoneally, and an eyecup prepared which was filled for 30 minutes with Krebs Ringer bicarbonate solution containing 20 //Ci of [ 3 H]-Ch (0.5 fjm). After washing with fresh medium containing physostigmine (30 fjm), a syringe was used to place 0.5 ml of medium in the eyecup, with replacement of medium occurring at 5-minute intervals. The total radioactivity in the resulting samples was measured by liquid scintillation counting after the addition of scintillant (4 ml Ultima Gold; Packard, UK). The retina was stimulated for 5- minute periods by flickering light at 3 Hz (25% duty cycle, retinal illuminance 650 lux). [ 3 H]-Acetylcholine represented 95% to 100% of the increase of total radioactivity evoked by light flashes. 1 The viability of the eyecup preparation and the effects of drugs on the gross physiological response of the retina were assessed by electroretinography. The effects of light stimulation on the release of total radioactivity from the retina are expressed as the ratio of the peakevoked release of radioactivity to the average release of radioactivity in the two samples immediately preceding stimulation. The resulting ratio is referred to as the "light-evoked release of [ 3 H]-ACh." Mean values and standard error of the mean (SEM) were calculated and compared, using Student's Mests for paired and unpaired analyses. Significant differences were assumed at a level of P < Isolated Retina Preparation. Rabbits were killed by an overdose of 80 mg kg" 1 intravenous pentobarbital. The eyes were enucleated and the retinas dissected in Krebs-Ringer bicarbonate medium at 22 C. Hemiretinas were incubated with [ 3 H]choline (5 X 10~ 8 M) for 30 minutes and washed for 10 minutes. To measure release of [ 3 H]-ACh and glycine, tine retinas were placed in a small perspex chamber (vol 0.5 ml) containing Krebs-Ringer bicarbonate medium (350 //I; physostigmine 30 /xm) at room temperature and were gently agitated by a stream of 5% CO 2 in oxygen. At 10-minute intervals, the medium was drained off and replaced with fresh solution. The glycine in each of

3 1636 Investigative Ophthalmology & Visual Science, July 1997, Vol. 38, No. 8 the resulting 10-minute samples was measured by high-performance liquid chromatography, as described previously, 3 and the [ 3 H]-ACh measured by liquid scintillation counting. Potassium-evoked release of [ 3 H]-ACh and glycine was achieved by exposing the tissues for 10-minute periods to KC1 (50 mm). In each experiment, the retina was exposed twice to KC1 (Si and S 2 ) and the ratio of the evoked release S 2 :Si was calculated. The retinas were exposed to drugs during the second stimulation period (S 2 ) and the resulting S 2 :Si ratio compared with the ratio obtained in control experiments. In control experiments the spontaneous resting release of glycine was 264 ± 30 pmol*5 min" 1 and the potassium-evoked release (Si) was 597 ± 56 pmol-5 min" 1 (n = 8; P < 0.001). Materials. [ 3 H]choline (83 Ci/mmol, 1 mci/ml) was obtained from Amersham International (United Kingdom). Nitric oxide donors and NOS inhibitors were obtained from Sigma Chemical (St. Louis, MO). (a) _ « to m R) o ^ (1.56) (1.42) (2.34) immsnap (1.36) RESULTS. Effect of Nitric Oxide Donors on Light-Evoked [ 3 H]-Acetylcholine Release. In control experiments, using the rabbit eyecup preparation, stimulation of the retina with flickering light (3 Hz) increased the release of [ 3 H]-ACh by 1.6 ± 0.15 (n = 10) times the spontaneous resting release. Although there was a continuous decline in the rate of release of radioactivity, the light-evoked release of [ 3 H]-ACh (peak release-resting release) remained constant for at least 2.5 hours. When the retina was exposed to medium containing the NO donors S-nitroso-N-acetyl- DL-penicillamine (SNAP) or sodium nitroprusside (SNP), the light-evoked release of [ 3 H]-ACh was strikingly increased (Fig. 1). The increase in the lightevoked release of [ 3 H]-ACh produced by SNAP (1 mm) was abolished by the NO scavenger hemoglobin (25 /JM), which itself had no effect on either the spontaneous resting release or the evoked release of [ 3 H]- ACh (Fig. 1). Effect of Nitric Oxide Synthase Inhibitors on [ 3 H]-Acetylcholine Release. In contrast with the effect of NO donors, the NOS inhibitor L-nitromonomethylarginine (LNMMA; 1 mm) reduced the light-evoked release of [ 3 H]-ACh by more than 60% (Fig. 2), an effect that was abolished by arginine (200 fjm). N- nitro-l-arginine (LNA) had a similar inhibitory effect on the light-evoked release of [ 3 H]-ACh but was less potent (IC 50 for LNMMA and LNA being approximately 0.6 mm and 7 mm, respectively). The reduction in the light-evoked release of [ 3 H]-ACh by NOS inhibitors suggested that the release of endogenous NO normally facilitates the evoked release of [ 3 H]- ACh. However, it seems that NO does not act directly on the cholinergic amacrine cells because the effect of LNMMA on the light-evoked release of [ 3 H]-ACh was blocked by strychnine (Fig. 2), suggesting that NO B (10) D (4) E (3) FIGURE l. (a) Typical rabbit eyecup experiment showing that exposure of the retina to SNAP (1 mm) enhanced the lightevoked release of [ 3 H]-acetylcholine (ACh). Each histobar represents a 5-minute collection period. Hatched histobars indicate stimulation with flickering light. The numbers above the hatched histobars indicate the peak release/average release of radioactivity in the two samples preceding stimulation (light-evoked release of [ 3 H]-ACh). (b) Summary of results showing that SNP also enhanced the lightevoked release of [ 3 H]-ACh and that hemoglobin (25 pim) abolished the effect of SNAP on [ 3 H]-ACh release. Results are expressed as light-evoked release of [ 3 H]-ACh in presence of drug/light-evoked release of [ 3 H]-ACh in control specimens X 100 (percentage of control light-evoked release). A, control; B, 1 mm SNP; C, 1 mm SNAP; D, 1 mm SNAP + 25 fjm hemoglobin; E, 25 ^tm hemoglobin. Numbers in parentheses are number of separate rabbit experiments. *P< 0.001, significandy different from control value. SNAP = S-nitroso-N-acetyl-DL-penicillamine; SNP = sodium nitroprusside. acts by reducing the activity of glycinergic amacrine cells that provide feedback inhibition (Fig. 3). Effect of Nitric Oxide Donors on the Potassium- Evoked Release of [ 3 H]-Acetylcholine and Glycine From Isolated Retinas. Because a light-evoked release

4 Reports 1637 (a) " Q. o ~ >, " g ra o (b) ^) ! I (2.09) I (2.31) (2.02) (197) (2.32) (1.64) (1.59) (2.49) I (2.11) (2.42) of glycine cannot be measured using the rabbit eyecup preparation, we examined the effects of NO donors on glycine and [ 3 H]-ACh release from isolated retinas. Sodium nitroprusside and SNAP almost doubled the potassium-evoked release of [ 3 H]-ACh but strikingly reduced the evoked release of endogenous glycine (Fig. 4). The increase in the potassium-evoked release of [ 3 H]-ACh caused by SNAP was blocked by strychnine, supporting the idea that NO increases the evoked release of [ 3 H]-ACh indirectly by inhibiting glycine release. Strychnine did not affect glycine release. Effect of Nitric Oxide Donors and Nitric Oxide Synthase Inhibitors on the Electroretinogram. When the retina was exposed to SNAP (1 mm), SNP (1 mm) for 5 minutes or to LNMMA (1 mm) and LNA (0.5 mm) for 20 minutes (n = 5 for each drug) there was no change in the amplitude of the electroretinogram b wave (not illustrated), indicating that their site of action was in the proximal retina and not in die photoreceptors or in the outer plexiform layer. Strychnine also had no significant effect on the results of the electroretinogram. (C) a 100: : 20. B C D E F G (3) (7) (5) (4) (5) (3) FIGURE 2. (a) Typical rabbit eyecup experiment showing that exposure of the retina to LNMMA inhibited the light-evoked release of [ 3 H]-ACh. (b) Typical rabbit eyecup experiment showing that strychnine blocks the action of LNMMA on the light-evoked release of [ 3 H]-ACh. (c) Summary of results showing that LNA also inhibited the light-evoked release of [ 3 H]-ACh and that strychnine (10 fj,m) and arginine (200 /xm) did not affect the lightevoked release of [ 3 H]-ACh but abolished the inhibitory action of LNMMA (1 mm). A, control; B, 5 mm LNA; C, 1 mm LNMMA; D, 1 mm LNMMA + 10 ^M strychnine; E, 1 mm LNMMA ^M L-arginine; F, 10 //M strychnine; G, 200 /JM L-arginine. Numbers in parentheses are the number of separate rabbit experiments. *P < 0.01, **P < 0.005, significantly different from control values. LNMMA = L-nitromonomethylarginine; LNA = N-nitro- L-arginine. DISCUSSION. The localization of neuronal NOS in the retina differs between species, the major difference being the presence of the enzyme in the outer retina (photoreceptors) and in Muller cells in nonmammalian species. 4 In the rabbit retina, neuronal NOS occurs predominantly in amacrine cells, 5 ' 6 which possess multiple, long, axon-like processes that ramif)' throughout the inner plexiform layer. It is believed that they are electrically coupled, and that wide-field stimulation of the retina presumably results in a relatively uniform release of NO. The target cells for NO, judged by light-induced increases in cgmp synthesis, include the cone-depolarizing bipolar cells 9 and a subset of ganglion cells. 8 After stimulation of the retina with light there is an increase in cgmp 4 in the bipolar cells that opens cgmp-gated cation channels causing depolarization and an increase in transmitter (glutamate) release. Because cone bipolar cells provide the input to the cholinergic amacrine cells, 2 changes in cgmp synthesis in these bipolar cells by NO donors and NOS inhibitors might be expected to account for the effects of these agents on the light-evoked release of [ 3 H]-ACh. However, our experiments suggest that NO does not enhance [ 3 H]-ACh release by a direct action on bipolar cells or on cholinergic amacrine cells but by inhibiting the release of glycine from the glycinergic amacrine cells that provide inhibitory feedback to the cholinergic neurones 3 (Fig. 4). It has been suggested that a diffuse release of NO from amacrine cells into the inner plexiform layer occurs, the rate of NO release being a function of overall retinal illumination. The NO so released would

5 1638 Investigative Ophthalmology & Visual Science, July 1997, Vol. 38, No. 8 act as a gain control on information transfer through on-bipolar cells. 6 Our results support the idea that NO is released by stimulation of the retina with light and show that NO increases the gain between bipolar cells and cholinergic amacrine ceils. However, the mechanism involved in the enhanced light-evoked release of ACh is indirect, resulting from partial suppression of the glycinergic feedback onto the cholinergic amacrine cells (Fig. 4) or bipolar cells. 3 A consequence of CO *** _L con 1mM 500\iNl 500 im SNAP SNP SNAP + -lo^m strychnine FIGURE 4. Diagrammatic representation of inhibitory feedback loop in the inner retina. The cone-depolarizing ("on") bipolar cells (DPBC) release an excitatory amino acid (glutamate?) that activates kainate receptors 1 (K) on the cholinergic amacrine cells (ACh). Release of [ 3 H]-acetylcholine (ACh) from these cells activates muscarinic receptors (M) on the glycinergic amacrine cells (shaded) which in turn release glycine (Gly) probably onto the cholinergic amacrine cells, which are known to possess glycine receptors. 10 Activation of these strychnine-sensitive glycine receptors partially inhibits [ 3 H]-ACh release. 2 Our present results show that nitric oxide modulates the activity of cholinergic amacrine cells indirectly by reducing glycinergic inhibitory feedback, thus enhancing the light-evoked release of [ 3 H]-ACh. CO CO ** T * * JL this more complicated mechanism is that NO is able to enhance the light-evoked response of cholinergic amacrine cells without increasing the spontaneous resting release of ACh, an effect that might be expected if NO merely increased glutamate release from bipolar cells. Key Words acetylcholine (ACh) release, cholinergic amacrine cells, nitric oxide 0.0 1mM SNP SNAP FIGURE 3. Release of [ 3 H]-acetylcholine (ACh) and glycine from isolated rabbit retinas exposed to KCl (50 mm). The ordinate S 2 :Si shows the release as a ratio of the release evoked by exposing the retina to KCl on two occasions (Si and S2). In control specimens, no drug was applied to the tissue. The effect of drugs was examined by exposing die retina to drug during the second period of potassium depolarization (S2). (a) The potassium-evoked release of [ 3 H]- ACh was strikingly increased by SNP and SNAP, the effect of SNAP being abolished by strychnine, (b) The potassiumevoked release of glycine was inhibited by SNP and SNAP. *P < 0.05, **P < 0.03, ***P < 0.01 significandy different from controls. Each result is the mean ± SEM of 3 to 7 experiments. SNAP = S-nitroso-N-acetyl-DL-penicillamine; SNP = sodium nitroprusside. References 1. Cunningham JR, Neal MJ. Effect of excitatory amino acids and analogues on [ 3 H]-acetylcholine release from amacrine cells of the rabbit retina. / Physiol. 1985; 366: Famiglietti EV. Synaptic organisation of starburst amacrine cells in rabbit retina: Analysis of serial thin sections by electron microscopy and graphic reconstruction./ Comp Neurol. 1991;309: Neal MJ, Cunningham JR. Baclofen enhancement of acetylcholine release from amacrine cells in the rabbit retina by reduction of glycinergic inhibition. JPhysiol. 1995;482: Goldstein IM, Ostwald P, Roth S. Nitric oxide: A review of its role in retinal function and disease. Vision Res. 1996; 36: Perez MTR, Larsson B, Aim P, Andersson KE, Ehinger B. Localisation of neuronal nitric oxide synthase

6 Reports 1639 immunoreactivity in rat and rabbit retinas. Exp Brain Res. 1995; 104: KoistinahoJ, Sagar M. NADPH-diaphorase reactive neurones in the retina. Progress in Retinal and Eye Research. 1995; 15: Schiells R, Falk G. Retinal on-bipolar cells contain a nitric oxide-sensitive guanylate cyclase. Neuroreport. 1992; 3: Ahmad I, Leinders-Zufall T, Kocsis JD, Shepherd GM, Zufall F, Barnstable CJ. Retinal ganglion cells express a cgmp-gated cation conductance activatable by nitric oxide donors. Neuron. 1994; 12: Koistinaho J, Swanson RA, de Vente J, Sagar SM. NADPH-diaphorase (nitric oxide synthase) reactive amacrine cells of rabbit retina: putative target cells and stimulation by light. Neuroscience. 1993; 57: Zhou ZJ, Fain GL. Neurotransmitter receptors of starburst amacrine cells in rabbit retinal slices. J Neurosci. 1995; 15: Effect of Varying the Mitomycin-C Treatment Area in Glaucoma Filtration Surgery in the Rabbit M. Francesca Cordeiro, Peter H. Constable, Robert A. Alexander, Shomi S. Bhattacharya, and Peng T. Khaw Purpose. To investigate the effect of varying the treatment area of subconjunctival mitomycin-c (MMC) using an adapted rabbit model of filtration surgery. Methods. Twenty-four New Zealand White rabbits underwent filtration surgery, with random allocation to one of three treatments: 5-minute subconjunctival applications of MMC (0.4 mg/ml) with either a large (8 X 10 mm) or small (4X2 mm) sponge or no treatment (control). Drainage was achieved by placing an intravenous cannula through a scleral tunnel into the anterior chamber. Rabbits were examined at set intervals for up to 30 days after surgery. Measurements of appearance, size, height, and vascularity of blebs and of intraocular pressure and anterior chamber depth were made by a masked observer. Histologic analysis of eyes was performed at 3, 14, and 30 days. Results. Statistical analysis showed a significant difference in bleb survival among all groups (log rank P = , with 100% survival with large areas of MMC treatment). Comparison between large and small treatment area groups revealed significant differences in bleb survival (log rank P= ), bleb area (betweensubject analysis, P = 0.009), and bleb height (betweensubject analysis, P = 0.005). These differences were seen clinically, with large areas of MMC treatment producing From the Wound Healing Group and Glaucoma Unit, the Department of Pathology, and the Department of Molecular Genetics, Institute of Ophthalmology and Moorfields Eye Hospital, London, England. Supported in part by grant from the Wellcome Trust (M1? C) and by the Frost Trust (PHC). Submitted for publication October 22, 1996; revised February 28, 1997; accepted March 4, Proprietary interest category: N. Reprint requests: Peng T. Khaiu, Wound Healing Group and Glaucoma Unit, Pathology Department, Institute of Ophthalmology, Bath Street, London, EC1V 9EL, United Kingdom. diffuse and elevated blebs, small areas of treatment producing thin-walled and localized blebs with scarring at 21 days, and no treatment resulting in comparatively vascularized and scarred blebs before 14 days. Histologic analysis revealed clear differences among groups, with an increase in subconjunctival cellularity and scar tissue in eyes with failed blebs. Conclusions. The size of the area of subconjunctival MMC treatment significantly affects surgical outcome. Histologic features mirror differences observed clinically. Alteration of the size of the MMC treatment area may provide an alternative and more controllable approach to modulating the wound-healing response after drainage surgery and, more important in the clinical context, to modifying bleb morphology. Invest Ophthalmol Vis Sci. 1997;38: JVlitomycin-C (MMC) has been highly effective as an adjuvant treatment for preventing postoperative scarring in patients at high risk of failure from conventional glaucoma filtration surgery. Its success in these cases is owing to its effect on reducing subconjunctival scar formation at the bleb site, 1 " 3 primarily by inhibitory and cytotoxic effects on such local cells as fibroblasts. Unfortunately, its use is associated with several complications, in particular the production of thin, avascular cystic blebs with the attendant risks of persistent hypotony or endophthalmitis. 1 Studies designed to investigate how to limit the potent and toxic effects of MMC have been directed mostly at varying the concentration of MMC. 4 However, we have shown previously that the effect of MMC delivered on a sponge is very focal. 5 To date, the surface area of MMC application has been largely ignored as a possible variable in influencing the treatment outcome. We describe a prospective, randomized, masked-observer study in which we investigated the effect of varying the area of conjunctiva treated with MMC on the surgical outcome, using a new model of filtration surgery in the rabbit. This model was designed specifically to investigate the healing response