Stimulation of Tear Secretion by Topical Agents That Increase Cyclic Nucleotide Levels

Transcription

1 Investigative Ophthalmology & Visual Science, Vol. 31, No. 7, July 1990 Copyright Association for Research in Vision and Ophthalmology Stimulation of Tear Secretion by Topical Agents That Increase Cyclic Nucleotide Levels Jeffrey P. Gilbard,*t Scorr R. Rossi,* Kathleen Gray Heyda,* and Darlene A. Darrr*t The authors examined the effect of topical application of agents known to increase cyclic nucleotide levels on tear secretion by accessory lacrimal gland tissue in their rabbit model for keratoconjunctivitis sicca (KCS). Tear secretion was studied by changes in tear film osmolarity and tear volume caused by application of the agents relative to application of isotonic buffer solution alone. A decrease in tear film osmolarity or increase in tear volume was interpreted as an increase in tear secretion. Irritative stimulation was distinguished from pharmacologic stimulation by the prior use of topical proparacaine. The following agents significantly decreased tear film osmolarity and increased tear volume: vasoactive intestinal peptide (2 X 10" 8 to 2 X 10~ 6 M); three pro-opiomelanocortin fragments a-, /?-, and 7-melanocyte stimulating hormone at 10" 4, 10~ 3, and 10~ 3 M, respectively; the permeable cyclic adenosine monophosphate (camp) and cyclic guanosine monophosphate (cgmp) analogs 8-Br camp ( X 10 3 M) and 8-Br cgmp ( X 10 3 M); and the cyclic nucleotide phosphodiesterase inhibitor l-isobutyl-3-methyl xanthine ( X 10 3 M). Forskolin (2 X 10~ 4 M), which activates the catalytic subunits of adenyl cyclase, increased tear volume significantly. Secretin, adrenocorticotropic hormone, and pilocarpine were ineffective. The authors conclude that agents that increase either camp or cgmp levels pharmacologically stimulated tear secretion when applied topically to rabbit eyes with surgically induced KCS. Invest Ophthalmol Vis Sci 31: ,1990 To date, the mainstay of treatment for the spectrum of dry eye disorders has been the topical administration of artificial tear solutions. 1 Another approach has attempted to stimulate tear secretion by oral administration of bromhexine, but clinical studies have yielded mixed results, and the drug has not attained widespread use. 2 " 7 Tear film osmolarity is elevated in keratoconjunctivitis sicca (KCS) 8 ' 9 because of several mechanisms associated with decreased lacrimal gland secretion. First, tear film turnover declines with decreased tear secretion, giving evaporation more time to concentrate the preocular tear film. Second, as tear volume declines but interpalpebral surface area remains constant, tear film evaporation has a greater effect on tear From the Eye Research Institute,* and Department of Ophthalmology.f Harvard Medical School, Boston, Massachusetts. Supported in part by NEI grants EYO3373 (JPG) and EY06177 (DAD). The tear secretagogues described in this report have been issued U.S. patents 4,745,100, 4,753,945, and 4,868,154. Drs. Gilbard and Dartt are the inventors, and the Eye Research Institute of Retina Foundation is the assignee. Additional domestic and foreign patents are pending. The Eye Research Institute has a proprietary interest in this technology, and the authors will participate in its commercialization. Submitted for publication: May 15, 1989; accepted December 1, Reprint requests: Jeffrey P. Gilbard, MD, Eye Research Institute, 20 Staniford Street, Boston, MA film osmolarity. Finally, it is possible that lacrimal gland fluid osmolarity itself increases as the lacrimal secretory rate declines, independent of the effect of evaporation. We have previously reported that the osmolarity of lacrimal gland fluid collected directly from the cannulated glandular excretory ducts of rabbits 10 increased as flow rate decreased (as well as the converse). We postulated that for topically applied agents to be effective in stimulating tear secretion and, thus, in decreasing tear osmolarity, they would have to act on the accessory lacrimal gland tissue located within the conjunctiva. We recently described a rabbit model for KCS in which secretion from the accessory glands can be studied without contributions from the main lacrimal gland, nictitans gland, or Harderian gland." 12 Agents for testing in this model were chosen based on what is known about stimulation of main lacrimal gland secretion. Activation of two different cellular signal transduction pathways stimulates secretion of fluid and protein from the main lacrimal gland. 13 The first of these pathways involves increasing intracellular cyclic adenosine monophosphate (camp), which can be accomplished in several ways. Vasoactive intestinal peptide (VIP); porcine isoleucine-containing peptide (PHI) a member of the VIP family of peptides; a-melanocyte-stimulating hormone («-MSH); adrenocorticotropic hormone (ACTH); and /3-adren- 1381

2 1382 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1990 Vol. 01 ergic agonists all bind to the stimulatory receptor subunit of adenylate cyclase. The resultant activation of adenylate cyclase produces camp, which in turn stimulates secretion. Cellular camp levels can also be increased by activating the catalytic subunits of adenylate cyclase (eg, by forskolin), by using permeable camp analogs (such as 8-bromo camp), or by preventing breakdown of camp by inhibiting campdependent phosphodiesterase activity (eg, by theophylline or l-isobutyl-3-methylxanthine [IBMX]). Cholinergic agonists stimulate main lacrimal gland secretion by a second pathway. They cause the breakdown of phosphatidylinositol bisphosphate into 1,4,5-inositol trisphosphate (1,4,5-IP 3 ) and diacylglycerol; 1,4,5-IP 3 then causes the release of intracellular Ca 2+, which stimulates secretion, whereas diacylglycerol activates protein kinase C to stimulate secretion. In the present study, we sought to determine whether activation of the camp-dependent pathway or of the cholinergic agonist-stimulated 1,4,5-IP 3 / Ca 2+ /diacylglycerol pathway stimulates accessory lacrimal gland secretion in a rabbit model for KCS. To determine whether stimulating the camp pathway was effective, we tested VIP and a-msh and related compounds of each (including secretin and glucagon), forskolin, 8-bromo camp, theophylline, and IBMX. Neither PHI nor /3-adrenergic agonists were tested because they were weak stimuli of main lacrimal gland secretion. To determine whether cholinergic activation of the 1,4,5-IP 3, Ca 2+, diacylglycerol pathway stimulated accessory lacrimal secretion, we also tested the cholinergic agonist pilocarpine. Although an increase in cgmp does not stimulate main lacrimal gland protein secretion, electrolyte and water secretion in the small intestine are affected. 14 We, therefore, used 8-bromo cyclic guanosine monophosphate (cgmp), a permeable cgmp analog, to determine whether activation of a cgmp-dependent pathway can stimulate accessory lacrimal gland secretion. Experimental Animals Materials and Methods Male and female New Zealand white rabbits weighing 2-3 kg were used. The rabbit model for KCS was created as previously described.'' All procedures using animals conformed to the ARVO Resolution on the Use of Animals in Research. Briefly, the lacrimal gland excretory duct and its orifice were closed with cauterization, and the nictitating membrane, nictitans gland, and Harderian gland were removed. This procedure resulted in an elevation of tear osmolarity by postoperative day 1. Tear samples approximately 0.3 /xl each were collected from the inferior marginal tear-strip, and tear osmolarity was measured as previously described The same person collected all tear samples and did not contact the ocular surface, lids, or lashes. Animals were restrained but not anesthetized during tear sampling. Osmolarity, measured in all operated eyes (n = 32) at least twice before rabbits were enrolled in the current protocol, averaged 317 ± 1.2 (standard error of the mean [SEM]) mosm/1 (normal range: mosm/1). Elevated tear film osmolarity indicated successful closure of the lacrimal gland excretory duct and was an inclusion requirement for all eyes studied. Materials All peptides (VIP; secretin; «-, 0-, and y-melanocyte-stimulating hormones [MSHs]; ACTH 1-24; and ACTH 4-10) were obtained from Peninsula Laboratories (Belmont, CA) with the exception of glucagon, which was obtained from Sigma (St. Louis, MO). Forskolin was obtained from Calbiochem-Behring (La Jolla, CA). Cyclic nucleotide analogs (8-bromoadenosine-3',5'-cyclic monophosphate [8-Br camp], 8-bromoguanosine-3',5'-cyclic monophosphate [8-Br cgmp]); phosphodiesterase inhibitors (3-isobutyl-lmethylxanthine [IBMX] and theophylline); and pilocarpine were obtained from Sigma. Atropine was obtained from Allergan Pharmaceuticals (Irvine, CA). Stock solutions of VIP and secretin were made in 0.1 N acetic acid, and stock solutions of glucagon in 1.6% glycerin. Stock solutions of forskolin were made in ethanol (100%). Stock solutions of the remaining peptides, cyclic nucleotide analogs, and phosphodiesterase inhibitors were made in an isotonic, buffered electrolyte solution (isotonic electrolyte solution). 17 The isotonic electrolyte solution contained the following millimolar concentrations: NaCl, 99; KC1, 24; NaHCO 3, 32; NaH 2 PO 4, 1.0; MgCl 2, 0.6; and CaCl 2, 0.8. The osmolarity of the solution was 302 mosm/1, and the ph was 7.4. Aliquots of peptide stock solutions were stored at -80 C until used. All solutions were diluted to the desired concentration by dilution with isotonic electrolyte solution. Vehicles in the control and experimental solutions were equivalent. Osmolarity Measurements In human KCS, tear film osmolarity decreased at 1 min after instillation of an isotonic drop but did not differ significantly from pretreatment values by 10 min. 18 Furthermore, tear film osmolarity decreases with increased tear secretion. 10 We postulated that if an agent stimulated tear secretion, isotonic solution with such an agent would decrease tear film osmolar-

3 No. 7 STIMULATION OF TEAR SECRETION DY TOPICAL AGENTS / Gilbord er ol 1383 ity more than isotonic solution alone. For this reason, we studied the effect of isotonic electrolyte solution with and without test compounds on tear secretion by examining its effect on tear osmolarity. Because pharmacologic agents were used in the present study, we expected a latency period before onset of action. We, thus, chose 5 min and 15 min after drop instillation as our time points for tear sample collection. All instilled drops were 10 ix\ and were administered with an Eppendorf pipette (Fisher Scientific, Medford, MA) to the superior bulbar conjunctiva by the same person. The superior bulbar conjunctiva was exposed by momentarily elevating the skin above the brow. The ocular surface, eyelids, and eyelashes were not touched. Instillation of any eye drop could induce reflex tearing for two reasons: routine sensory stimulation by placement of the drop and irritative stimulation from a particular agent in the drop. Routine sensory stimulation was controlled by vehicle controls and minimized by careful drop placement; a nontoxic isotonic electrolyte solution was used whenever possible 17 as a vehicle and control solution. Irritative stimulation by potential agents was ruled out by testing the highest doses of each agent in eyes pretreated with topical proparacaine. In the experiments with proparacaine, a 0.5% concentration was used, and ~30 ^1 was instilled 5 min before time zero tear sampling. At time zero a tear sample was taken for measurement of osmolarity. At 1 min the test drop was instilled. At 6 min and 16 min tear samples were taken for osmolarity measurements. During dose-response experiments, 5 min would elapse between the 16-min tear sample and the start of a new cycle initiated with the administration of proparacaine. For each cycle the test drop was alternated between isotonic electrolyte solution alone, which was given first, and isotonic electrolyte solution plus the test agent. For dose-response experiments, the concentration of the agent was increased with each alternate cycle. For tear osmolarity dose-response experiments in the absence of proparacaine, the above protocol was used with the same tear sampling and test drop administration time points. When agents were tested initially in dose-response experiments without proparacaine, experiments with proparacaine were performed in separate rabbits, with the use of the highest previously tested concentration of the agent. The result of this testing would indicate whether any observed effect without proparacaine had resulted from reflex tearing induced by irritation of the ocular surface. In an additional set of experiments, we studied the effect of 1.6 X 10"' M (4%) pilocarpine on tear osmolarity in the KCS rabbit model. Protocols were carried out both with and without proparacaine. Furthermore, in a separate experiment atropine (1.4 X 10~ 2 M, 1%) was administered 6 min before the instillation of pilocarpine (1.6 X 10"' M, 4%). Volume Measurements In the study of tear volume as an indication of tear secretion, the tear volume was measured during the period of the greatest decrease in tear osmolarity. At time zero, one drop of proparacaine was instilled in the eye to be tested. At 1 min 10 /nl of isotonic electrolyte solution was instilled, and between 5 and 6 min tear fluid was collected from the inferior marginal tear strip with a preweighed Weck-Cel sponge (Edward Week and Co, Research Triangle Park, NC). Volume was determined by the difference in sponge weight before and after tear fluid collection. After a 14-min interval, the sequence was repeated with the administration of proparacaine followed by isotonic electrolyte solution containing the highest dose of each experimental compound used for the osmolarity experiments. Data Analysis All data are expressed as mean ± SEM (n = number) and were analyzed with the use of the student t-test for paired or unpaired data. Results VIP and Related Peptides Because VIP was a potent camp-dependent stimulus of both fluid and protein secretion from the rat and rabbit main lacrimal gland, 1319 ' 20 we investigated the effects of VIP first. After prior instillation of proparacaine, VIP (2 X 10" 8 to 2 X 10" 6 M) lowered tear film osmolarity in a dose-dependent manner when compared with isotonic electrolyte solution (Fig. 1, top). We concluded that the decrease in tear film osmolarity after topical administration of VIP indicated an increase in tear secretion. We confirmed this by measuring the volume of tear secretion after instillation of VIP (2 X 10" 6 M) and of isotonic electrolyte alone (Fig. 1, bottom). At concentrations of 2 X 10~ 7, 2 X 10" 6, and 2 X 10~ 5 M, secretin decreased tear film osmolarity relative to isotonic electrolyte solution alone, but, unlike VIP, the effect of secretin was blocked by prior administration of proparacaine (data not shown). Furthermore, tear volume after instillation of 2 X 10" 5 M secretin did not differ significantly from that after instillation of isotonic electrolyte solution (5.0 ± 0.8 vs 5.5 ± 0.5 /x\, respectively; n = 3). Secre-

4 1384 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1990 Vol. 31 P<O.O2 <n-3) Butler i 2x10~ 6 M VIP Fig. 1. Effect of isotonic buffer solution with and without vasoactive intestinal peptide (VIP) on tear film osmolarity (top) and tear volume (bottom) after topical proparacaine. manner relative to the decrease seen with isotonic electrolyte solution alone. In separate experiments with proparacaine, /?- and 7-MSH at 10" 3 M produced greater decreases in tear osmolarity than did isotonic electrolyte solution alone. The effect of a-, (8-, and 7-MSH on tear volume was tested at the highest concentrations used during the osmolarity experiments. All three compounds increased tear volume significantly relative to the controls (Fig. 3, bottom). The ACTH fragment ACTH ^4 at concentrations of 10~ 7 to 10~ 4 M did not significantly decrease tear film osmolarity in the absence of proparacaine relative to controls. The ACTH fragment ACTH 4., 0 was tested at concentrations from 10~ 5 to 10~ 3 M. At a concentration of 10~ 3 M, ACTH decreased tear film osmolarity, but the effect was blocked by proparacaine. We concluded that the effect of ACTH4.10 resulted from irritative sensory stimulation. Forskolin Forskolin increases cellular camp levels by activating the catalytic subunits of adenyl cyclase and stimulates protein secretion from rat exorbital lacrimal glands in vitro. 22 ' 23 We were unable to test the effect of forskolin on tear osmolarity because it was tin appeared to increase tear secretion by irritative sensory stimulation, which was blocked when the ocular surface was anesthetized. Glucagon (2 X 10~ 7 to 2 X 10~ 5 M) decreased tear film osmolarity in a dose-dependent manner (Fig. 2, top). In separate experiments, the effect of 2 X 10~ 5 M was not blocked by prior administration of proparacaine. Glucagon (2 X 10~ 5 M) significantly increased tear volume compared with the control (Fig. 1, bottom). Pro-Opiomelanocortins Both a-msh and ACTH stimulate protein secretion from rat exorbital lacrimal glands in vitro. 21 We tested four pro-opiomelanocortin fragments: a-, /?-, and 7-MSH, and ACTH. a-msh at 1(T 4 M, 0-MSH at 10" 3 M, and 7-MSH at 10~ 3 M significantly decreased osmolarity relative to isotonic electrolyte solution alone (Fig. 3, top), fi- and 7-MSH did not lower osmolarity at concentrations of 10~ 5 and 10~ 4 M. In experiments performed with proparacaine, a-msh at concentrations of 10~ 6, 10~ 5, and 10~ 4 M decreased tear film osmolarity in a dose-dependent Buffer Glucagon 2x10~! M Fig. 2. Effect of isotonic buffer solution with and without glucagon on tear film osmolarity without prior application of proparacaine (top) and on tear volume after topical proparacaine (bottom).

5 No. 7 STIMULATION OF TEAR SECRETION BY TOPICAL AGENTS / Gilbord er ol a-msh 10~ 4 M (n=4) /J-MSH 1(f 3 M (n=5) y-msh 10' 3 M (n=6) Cyclic Nucleotide Phosphodiesterase Inhibitors Both camp and cgmp phosphodiesterases are inhibited by IBMX, although IBMX inhibits the former more effectively. In experiments without proparacaine, IBMX (0.3 X 10" 3 M to 3.0 X 10" 3 M) decreased tear osmolarity in a dose-dependent manner (Fig. 7, top). In separate experiments, the effect of 3 X 10" 3 M IBMX was not blocked by the prior administration of proparacaine. The effect on tear secretion was confirmed by tear volume measurements (Fig. 7, bottom). Theophylline, a less effective cyclic nucleotide phosphodiesterase inhibitor than IBMX, also decreased tear osmolarity with and without proparacaine, but a relatively high concentration was required (6 X 10" 2 M). Buffer a-msh 10" 4 M Fig. 3. Effect of isotonic buffer solution with and without a-, j8-, and 7-melanocyte-stimulating hormone (MSH) on tear film osmolarity without topical proparacaine (top) and tear volume after topical proparacaine (bottom). Each MSH was tested in separate experiments. dissolved in solution with ethanol, which interferes with the freezing-point depression method used to measure osmolarity. However, topical administration of 2 X 10~ 4 M forskolin significantly increased tear volume relative to the vehicle alone (Fig. 4). Cyclic Nucleotide Analogs camp analogs, but not cgmp analogs, have stimulated lacrimal gland protein secretion from rat exorbital lacrimal glands in vitro. 23 In experiments without proparacaine, 8-Br camp (0.3 X 10" 3 M to 3.0 X 10" 3 M), a permeable camp analog, decreased tear osmolarity in a dose-dependent manner (Fig. 5, top). In separate experiments, the effect of 3.0 X 10~ 3 M was not blocked by the prior instillation of proparacaine. The effect on tear secretion was confirmed by tear volume measurements (Fig. 5, bottom). A permeable cgmp analog, 8-Br cgmp also produced a dose-dependent decrease in tear osmolarity (Fig. 6, top). In separate experiments the effect of 10 X 10~ 3 M was not blocked by proparacaine. The effect on tear secretion was again confirmed by tear volume measurements (Fig. 6, bottom). Pilocarpine In unanesthetized eyes, 1.6 X 10"' M (4%) pilocarpine, a cholinergic agonist, was more effective than isotonic electrolyte solution in decreasing tear film osmolarity, but the effect was blocked by prior administration of topical proparacaine (Fig. 8). The cholinergic antagonist atropine at 1.4 X 10~ 2 M (1%) (a concentration that blocked pilocarpine-induced LU _l O > 15 T *p<.01 (n=3) 10-- Buffer Forskolin 2x10" 4 M Fig. 4. Effect of isotonic buffer solution with and without forskolin on tear volume after topical proparacaine.

6 1386 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / July 1990 Vol Br camp 0.3x10" 3 M 8-Br CAMP 1x10" 3 M 8-Br camp 3x10" 3 M 315 T IBMX 0.3x10~ 3 M O- buffer - 8-Br camp n=3 *- p< * 6 Time (min) 300 Buffer 8-Br camp 3x10" 3 M Fig. 5. Effect of isotonic buffer solution with and without 8-bromoadenosine-3',5'-cyclic monophosphate (8-Br camp) on tear osmolarity without topical proparacaine (top) and tear volume after topical proparacaine (bottom). 8-Br cgmp 1x10" 3 M 8-Br cgmp 3x10' 3 M 8-Br cgmp 10x10" 3 M Buffer IBMX 3x10" 3 M Fig. 7. Effect of isotonic buffer solution with and without 3-isobutyl- 1-methylxanthine (IBMX) on tear osmolarity without topical proparacaine (top) and tear volume after topical proparacaine (bottom). 310^ O- buffer - 8-Br cgmp n=3 *- p<0.05 pupillary meiosis) did not block pilocarpine-induced secretion in unanesthetized eyes (data not shown) n' T * 6 Buffer 8-Br cgmp 10x10" 3 M Fig. 6. Effect of isotonic buffer solution with and without 8-bromoguanosine-3',5'-cyclic monophosphate (8-Br cgmp) on tear osmolarity without topical proparacaine (top) and tear volume after topical proparacaine (bottom). Discussion The aqueous portion of the tear film is formed primarily by the secretion of the main and accessory lacrimal glands. Secretory mechanisms in the main lacrimal gland have been studied extensively, and the evidence indicates that fluid and protein secretion is stimulated by increasing intracellular levels of camp or Ca " 23 In the rat exorbital lacrimal gland, 8-Br cgmp did not stimulate protein secretion in vitro, suggesting that cgmp does not play a role in protein secretion in this gland. 23 Data on cgmp and fluid secretion from the main gland are not available. In the current study we were able to investigate for the first time tear secretion from the accessory lacrimal gland tissue and to demonstrate that accessory lacrimal gland secretion can be stimulated by the topical application of agents that increase cyclic nucleotide levels. In contrast to main lacrimal gland tissue in which only camp appears to play a role, both camp and cgmp increased accessory gland fluid secretion. This is analogous to secretion in the small

7 No. 7 STIMULATION OF TEAR SECRETION DY TOPICAL AGENTS / Gilbord er ol O Buller Pilocarpine (1.6x10" 1 M) drop instilled # p<0.05 proparacaine did not induce tear secretion by irritative sensory stimulation. We clearly did not stimulate fluid secretion from the main lacrimal gland in our KCS rabbits. Furthermore, in three normal rabbit eyes, topically applied VIP (2 X 10~ 6 M) did not stimulate fluid secretion from cannulated lacrimal gland excretory ducts (unpublished results). An increase in cyclic nucleotides is not known to affect fluid transport at the corneal epithelium-tear interface. We can also eliminate a change in the conjunctival vascular permeability as a possibility because the conjunctival vessels are fenestrated and "leaky," and cyclic nucleotides are not known to affect blood vessel permeability. We cannot eliminate the possibility that we are changing the rate of conjunctival epithelial fluid transport, but we consider this unlikely. Patients with KCS frequently can have normal Schirmer test results. 9 ' 25 Increased tear film osmolarity in these patients, 9 nonetheless, suggests inadequate rates of lacrimal fluid secretion under normal circumstances. Placement of Schirmer strips in these patients seems to provide sensory stimulation that is adequate to increase lacrimal gland fluid secretion rates. This observation with regard to Schirmer testing in patients with dry eyes indicates the clinical promise of therapeutic approaches seeking to stimulate accessory lacrimal gland tissue through pharmacologic means. Key words: accessory lacrimal glands, cyclic AMP, cyclic GMP, tear osmolarity, tear secretion Time (min.) Fig. 8. Effect of isotonic buffer solution with and without pilocarpine on tear film osmolarity without {top) and with (bottom) prior application of topical proparacaine. intestine in which both camp and cgmp stimulate fluid secretion. 14 Furthermore, the accessory lacrimal glands are not under cholinergic control 24 as the main lacrimal gland appears to be, given the failure of the muscarinic agonist pilocarpine to stimulate tear secretion in the presence of proparacaine and the failure of the muscarinic antagonist atropine to block the effect of pilocarpine in the unanesthetized eye. Secretion from the accessory glands appears to be under neural control because sensory reflex stimulation of these glands by the irritant pilocarpine was blocked by topical local anesthetic. Did we really stimulate tear secretion by stimulating accessory lacrimal gland secretion by camp-dependent compounds? Because the effect of pilocarpine on tear film osmolarity was blocked by proparacaine, we conclude that the agents not blocked by 16 References 1. Gilbard JP: Topical therapy for dry eyes. Trans Ophthalmol SocUK 104:484, Frost-Larsen K, Isager H, and Manthorpe R: Sjogren's syndrome treated with bromhexine: A randomised clinical study. BrMed J 1:1579, Mackie I and Seal DV: Sjogren's syndrome, bromhexine, and tear secretion [letter]. Br Med J 2:638, Tapper-Jones LM, Aldred MJ, Cadogan SJ, et al: Sjogren's syndrome treated with bromhexine: A reassessment. Br Med J 280:1356, Avisar R, Savir H, Machtey I, Ovaknin L, Shaked P, Menache R, and Allalouf D: Clinical trial of bromhexine in Sjogren's syndrome. Ann Ophthalmol 13:971, Scharf JM, Obedeanu N, Meshulam T, et al: Influence of bromhexine on tear lysozyme level in keratoconjunctivitis sicca. Am J Ophthalmol 92:21, Prause JU, Frost-Larsen K, Hoj L, Isager H, and Manthorpe R: Lacrimal and salivary secretion in Sjogren's syndrome: The effect of systemic treatment with bromhexine. Acta Ophthalmol 62:489, Gilbard JP, Farris RL, and Santamaria J: Osmolarity of tear microvolumes in keratoconjunctivitis sicca. Arch Ophthalmol 96:677, Gilbard JP and Farris RL: Tear osmolarity and ocular surface disease in keratoconjunctivitis sicca. Arch Ophthalmol 97:1642, Gilbard JP and Dartt DA: Changes in rabbit lacrimal gland

8 1388 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1990 Vol. 01 fluid osmolarity with flow rate. Invest Ophthalmol Vis Sci 23:804, Gilbard JP, Rossi SR, and Gray KL: A new rabbit model for keratoconjunctivitis sicca. Invest Ophthalmol Vis Sci 28:225, Gilbard JP, Rossi SR, Gray KL, Hanninen LA, and Kenyon KR: Tear film osmolarity and ocular surface disease in two rabbit models for keratoconjunctivitis sicca. Invest Ophthalmol Vis Sci 29:374, Dartt DA, Shulman M, Gray KL, Rossi SR, Matkin C, and Gilbard JP: Stimulation of rabbit lacrimal gland secretion with biologically active peptides. Am J Physiol 254 (Gastrointest Liver Physiol 17):G300, Field M: Secretion of electrolytes and water by mammalian small intestine. In Physiology of the Gastrointestinal Tract, Johnson LR, editor. New York, Raven Press, 1981, pp Gilbard JP and Farris RL: Ocular surface drying and tear film osmolarity in thyroid eye disease. Acta Ophthalmol 61:108, Gilbard JP, Gray KL, and Rossi SR: Improved technique for storage of tear microvolumes. Invest Ophthalmol Vis Sci 28:401, Gilbard JP, Rossi SR, Gray Heyda K: Ophthalmic solutions, the ocular surface, and a unique therapeutic artificial tear formulation. Am J Ophthalmol 107:348, Gilbard JP and Kenyon KR: Tear diluents in the treatment of keratoconjunctivitis sicca. Ophthalmology 92:646, Dartt DA, Baker AK, Vaillant C, and Rose PE: Vasoactive intestinal polypeptide stimulation of protein secretion from rat lacrimal gland acini. Am J Physiol 247 (Gastrointest Liver Physiol 10):G502, Stolze HH and Sommer HJ: Influence of secretagogues on volume and protein pattern in rabbit lacrimal fluid. Curr Eye Res 4:489, Jahn R, Padel U, Porsch PH, and Soling HD: Adrenocorticotropic hormone and alpha-melanocyte-stimulating hormone induce secretion and protein phosphorylation in the rat lacrimal gland by activation of a camp-dependent pathway. Eur J Biochem 126:623, Mauduit P, Herman G, and Rossignol B: Forskolin as a tool to study the (8-adrenergic receptor-elicited, labeled protein secretion in rat lacrimal gland. FEBS Lett 153:21, Dartt DA, Donowitz M, Joshi VJ, Mathieu RS, and Sharp GW: Cyclic nucleotide-dependent enzyme secretion in the rat lacrimal gland. J Physiol 352:375, Botelho SY, Hisada M, and Fuenmayor N: Functional innervation of the lacrimal gland in the cat. Arch Ophthalmol 76:581, Van Bijsterveld OP: Diagnostic tests in the sicca syndrome. Arch Ophthalmol 82:10, 1969.