ABSTRACT

A double-blind randomized clinical study was conducted todetermine whether nicardipine hydrochloride was a usefultreatment for dry eye.We examined its effect on the tear film,ocular surface and ocular comfort. Nicardipine hydrochloride,3-isobutyl-1-methylxanthine and pilocarpine hydrochloridewere dissolved in an artificial tear vehicle and applied topicallyto one eye of 12 subjects on separate days. Ocular physiol-ogy, ocular comfort and tear volume were assessed.The trialwas repeated with nicardipine in an aqueous gel vehicle.Tearswere collected and assessed for protein concentration andprotein profile, using electrophoresis and mass spectrometry.Nicardipine induced conjunctival redness and symptoms ofdryness and irritation. There was no change in total tearprotein concentration or volume. An increase in a 68 kDaprotein was observed, this was probably due to conjunctivalvessel dilation and leakage of albumin. The adverse sympto-matology and increased conjunctival redness experiencedwith nicardipine make it an undesirable treatment for dry eye.

Key words: dry eye, lacrimal, nicardipine hydrochloride,phosphodiesterase, tears.

INTRODUCTION

The repeated instillation of artificial tears is currently themost common treatment for dry eye conditions. A poten-tially more effective therapy, however, may be the pharma-cological stimulation of tear secretion. Stimulation of thelacrimal gland can increase the volume of tears, and can alsochange the protein composition of tears.1 Selectively chang-ing tear protein composition may be beneficial for dry eyebecause tear film proteins have recently been shown to con-tribute to low tear surface tension, and to increase tearwetting times in vitro.2,3 In particular, the tear protein,

lipocalin, was identified as important because of its ability tobind lipids.4

Tear proteins are either serum derived (albumin), consti-tutively secreted (sIgA), or under regulated secretion (lacto-ferrin, lipocalin, lysozyme). Constitutively secreted proteinsare released according to their rate of synthesis.5 During regulated protein secretion, second messenger compoundswithin the acinar cells cause fusion of stored secretoryprotein vesicles with the apical membrane, and the sub-sequent release of protein into the lacrimal lumen. Acinarcell receptors are activated by a variety of neuroreceptoragonists and hormones.6 An important second messenger fortear protein secretion in the rat is cyclic adenosinemonophosphate (cAMP).7,8

The presence of phosphodiesterase (PDE) inhibitorsleads to an increase in cAMP by inhibiting its breakdown by phosphodiesterase. Application of a non-selective PDEinhibitor, 1-isobutyl-3-methyl xanthene (IBMX), to theocular surface decreased the osmolarity of tears in a rabbitdry eye model9 and caused a decrease in tear osmolarity androse bengal staining in a clinical trial of 11 human dry eyepatients.10 Whether or not there was a subsequent change intear protein composition was not examined.

In this study we compared the effect on tear proteinsecretion of the PDE inhibitor nicardipine hydrochloride(HCl) with IBMX and pilocarpine HCl in humans.Pilocarpine was chosen as it is known to affect tear secretionin humans. Nicardipine was chosen for the study becausepreviously it has been demonstrated that topical applicationof this PDE inhibitor to rabbit eyes causes an increase insecretion of specific proteins.11

METHODS

Informed consent was obtained from all subjects in accor-dance with the Declaration of Helsinki. The protocol wasapproved by an institutional ethics committee. All subjects

Clinical and Experimental Ophthalmology (2000) 28, 208–211

Lens and Cornea

Modulation of tear film protein secretion with phosphodiesterase inhibitorsVictoria Evans BOptom,1 Mark D P Willcox PhD2 and Thomas J Millar PhD1

1Cooperative Research Centre for Eye Research and Technology, University of New South Wales, Sydney and 2School of Science,University of Western Sydney Nepean, Sydney, New South Wales, Australia

■ Correspondence: Victoria Evans, Cooperative Research Centre for Eye Research and Technology, University of New South Wales, Sydney, NSW 2052,

Australia. Email: v.evans@cclru.unsw.edu.au

Phosphodiesterase inhibitors and tears 209

were healthy with no ocular disease. Subjects ranged in agefrom 21 to 37 years and were an equal mix of males andfemales.

Part 1: Nicardipine HCl in aqueous solutioncompared with IBMX and pilocarpine HCl

A sterile vehicle solution of isotonic saline with 1.4%polyvinyl alcohol at pH 7.0 was prepared. Solutions ofnicardipine HCl (3 mmol/L), IBMX (3 mol/L) and pilo-carpine HCl (6 mmol/L) were dissolved in the vehicle.Nicardipine HCl and IBMX were purchased from Sigma(Sydney, Australia) and pilocarpine HCl from Allergan(Sydney, Australia).

This was a double-blind trial of 12 subjects. The threeagents were tested in random order on separate days, andthe placebo vehicle solution was applied to the contralateraleye. The test and control eye were randomized (right orleft) at the beginning of the trial and then remained consis-tent throughout the studies. Two 20 µL drops of the agentand vehicle were instilled in the test and control eyes,respectively. Drops were instilled after the baseline checkand then observations were made at 10 min, 1 h and 4 h.

Part 2: Nicardipine HCl in a gel formulation

Nicardipine HCl (3 mmol/L) was dissolved in a carbomergel (Viscotears; Ciba Vision Ophthalmics, Sydney,Australia). This was a double-blind trial of three subjects.Two 20 µL drops of the agent and vehicle were instilled inthe test and control eye, respectively. Drops were instilledafter the baseline check (always before noon) and thenobservations were made at 10 min, 1 h and 3 h. A seconddose was given after 3 h and observations were repeated at10 min, 1 h and 3 h.

Variables assessed

The following variables were assessed to monitor the ocularresponse to the agents. Visual acuity and pupil size weremeasured at normal room illumination. Descriptors of ocularsymptoms (such as pain, itchiness, scratchiness and stinging)were recorded. Tear volume measurements were made usingthe Zone-quick test (Showa/Menicon, Adelaide, Australia).Tears were collected (in parts 1 and 2 at each time point) bytouching a polished glass capillary tube to the tear prism atthe outer corner of the eye. Care was taken not to inducereflex tear secretion (tear flow rate > 3 µL/min). Sampleswere stored in plastic Eppendorf tubes and frozen at –86°C.Following tear collection, bulbar and limbal conjunctivalredness and palpebral conjunctival redness were gradedusing a 0–4 grading scale, with 0 being no redness and 4 being severe redness.

Total tear protein

The total tear protein concentration of each sample (fromparts 1 and 2 and at each time point) was estimated using the

Pierce Bicinchoninic Acid (BCA) Protein Assay Kit withbovine serum albumin as a standard (Lab Supply, Sydney,Australia).

Sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS-PAGE)

In order to separate proteins on the basis of molecularweight, SDS-PAGE was performed on the tear samplesunder reducing conditions according to the method ofLaemmli.12 The gels were precast 10% – 20% gradient, TrisHCl-buffered acrylamide with a 4% stacker (Bio-Rad,Sydney, Australia). Kaleidoscope prestained standards (Bio-Rad, Sydney, Australia) were used as molecular weightmarkers. Electrophoresis was performed at 80 V for 30 min,then at 100 V until completion of the run. Gels were stainedwith colloidal Coomassie Blue G250 or transferred onto anitrocellulose membrane (pore size 0.2 µm; Bio-Rad,Sydney, Australia) for western blotting.

Western blot analysis

In order to identify individual proteins, the gels were trans-ferred onto a nitrocellulose membrane overnight (30 V, 4°C,Mini-Trans Blot; Bio-Rad, Sydney, Australia). The mem-brane was blocked with 3% skim milk in Tris-buffered saline,pH 7.5. The polyclonal antibodies used were sheep anti-human lipocalin (donated by Dr Bernhard Redl, Innsbruck,Austria) and goat antihuman albumin (Calbiochem-Nova,Sydney, Australia) at 1 : 5000 dilution. The horseradish per-oxidase conjugated secondary antibodies were rabbitantigoat immunoglobulin (DAKO, Sydney, Australia) andrabbit antisheep immunoglobulin (Calbiochem-Nova,Sydney, Australia) at 2 : 5000 dilution. For detection of pro-teins, 3,3′-diaminobenzine (DAB) and 4-chloro-1-naphthol(4CN) (Sigma, Sydney, Australia) were used.

Matrix assisted laser desorption ionization-timeof flight (MALDI-TOF)

A second method of identifying individual proteins wasMALDI-TOF mass spectrometry.13 This was performed onlyfor tears from part 2 of the study. Protein bands were excisedfrom the SDS-PAGE gel and a trypsin digest was performed.The extracted peptides underwent MALDI-TOF. The resulting peptide masses were searched against SWISS-PROT and TrEMBL databases using the PeptIdent searchtool to gain a sequence identity for the protein.14,15

Two-dimensional polyacrylamide gel electrophoresis (2-D PAGE)

To better separate the protein species present in the tearsamples, they were first separated with pH range 4–7 iso-electric pH gradient strips (IPG strips and Multiphor II;Amersham Pharmacia Biotech, Sydney, Australia) and thenwith 16.5% Tris-Tricine precast gels (Bio-Rad, Sydney,Australia). The gels were stained with silver diamine or col-loidal Coomassie Blue G250.

RESULTS

Part 1: Nicardipine in aqueous solution compared with IBMX and pilocarpine

No changes in ocular physiology or tear function wereobserved with IBMX. Pilocarpine caused mitosis but therewas no change in tear volume or tear collection rate.Nicardipine HCl caused an increase in bulbar and limbalconjunctival redness after instillation. There was no changein total protein. Two-dimensional electrophoresis of the tearsamples did not reveal a difference in the amount or isoformexpression of any protein.

Part 2: Nicardipine in a gel formulation

Nicardipine HCl, when applied as a gel, caused a change inprotein profile over time on SDS-PAGE. There was anincrease in a 68 kDa band after drop instillation. There wasalso variation in a 27 kDa and a 32 kDa band. MALDI-TOFmass spectrometry identified seven of the bands andalbumin identified by this method but its presence was con-firmed by western blot (Fig. 1). Lipocalin was identified bywestern blot but its apparent concentration did not changeduring the studies. The repeated dosing caused increases in

redness and in specific protein bands (Fig. 1). There were nosignificant differences in protein profiles between parts 1and 2 of the study.

Symptomatology

While pilocarpine HCl and IBMX did not induce anyadverse symptoms, nicardipine HCl was not well tolerated.The number of patients reporting symptoms was higher forthe test eye than for the control (Fig. 2). The symptomsreported included numbness, tired/heavy eyes, irritation,itchiness, stinging, soreness and dryness.

DISCUSSION

Given the promising results seen with these agents in therabbit model and in lacrimal tissue culture, we expected tosee an alteration in the tear protein profile with nicardipineHCl. However, no change in total protein or tear volumewas observed either in the aqueous or gel formulations. Norwere there any changes in the protein profiles between theaqueous and gel formulations. It is probable, although notproven by the current study, that the drugs in gel formula-tion remained in the eye longer. The increase in bulbar andlimbal conjunctival redness observed with nicardipine HClwas consistent with its vasodilatory properties. The increasein the 68 kDa band over time with nicardipine HCl is mostlikely due to conjunctival vessel dilation and leakage ofalbumin. The increase in the 32 kDa and the 27 kDa bandmay be an effect of nicardipine HCl, however, the effect ofthese proteins on wetting is unknown. The redness and

210 Evans et al.

Figure 1. Optical density profile of tear samples from a nicardipine HCl treated eye (gel formulation) on a sodiumdodecylsulfate–polyacrylamide gel electrophoresis 10–20% acryl-amide gel stained with colloidal Coomassie Blue G250. Molecularweights of each peak were estimated from MW standards and areindicated on the graph. Note the increase in the 68 kDa peak from the baseline to 3 h and the increase in the 32 kDa and the27 kDa peak. Peaks were identified using matrix assisted laser desorption ionization-time of flight and western blot as follows:93 kDa, lactoferrin; 68 kDa, albumin; 57 kDa, ADHAP synthase;49 kDa, zinc-α2-glycoprotein; 40 kDa, unknown; 32 kDa, un-known; 27 kDa, unknown; 21 kDa, lipocalin; 17 kDa, lysozyme; 9 kDa, lacryglobin.

—, baseline; , 10 min; , 1 h; , 3 h.

Figure 2. The percentage of subjects reporting symptoms for thetest or control eye after instillation of nicardipine HCl (gel formu-lation) at the baseline, 10 min, 1 h and 3 h. The symptoms reportedincluded numbness, tired/heavy eyes, irritation, itchiness, stinging,soreness and dryness. Note the increase in symptoms in the test eye h compared with the control eye j at 1 h and 3 h after dropinstillation.

Phosphodiesterase inhibitors and tears 211

adverse symptomatology experienced with nicardipine HClmake it an undesirable agent for dry eye treatment.

None of the agents caused a consistent increase in tearvolume. Nicardipine HCl did not cause a change in proteinprofile when in aqueous formulation, but it did cause achange when applied twice a day as a gel. This highlightsproblems with the dosing regimen and drug penetration.Topical application of these drugs was chosen for the studyas this was the least invasive and the simplest deliverymethod; one which could easily be translated into a treat-ment regimen. It seems, however, that repeated dosing ofthese drugs may be required for the agents to penetrate theconjunctiva and reach the lacrimal gland. In terms of conve-nience for the dry eye sufferer, repeated instillation oftopical drops is not an improvement over repeated instilla-tion of artificial tears. Possibly, a slow-release device such asLacrisert (Sigma, Sydney, Australia) would be more effectivefor drug delivery of lacrimal stimulants such as IBMX, butnot for nicardipine HCl due to its adverse effect on comfortand the redness experienced.

ACKNOWLEDGEMENTS

The authors wish to thank Dr Bernhard Redl (Innsbruck,Austria) for his generous gift of human lipocalin antibodiesand Cassandra Vockler and Dr Peter Hains for their

assistance with 2-D gels and MALDI-TOF analysis. Thiswork was supported in part by a grant from the AustralianFederal Government under the Cooperative ResearchCentres Programme. VE was funded by a University ofWestern Sydney Nepean/CRCERT postgraduate award.This research has been facilitated by access to the AustralianProteome Analysis Facility established under the AustralianMajor National Research Facilities Program.

REFERENCES

1. Fullard RJ et al. Invest. Ophthal. Vis. Sci. 1991; 32: 2290–2301.2. Schoenwald RD et al. Adv. Exp. Med. Bio. 1998; 438: 391–400. 3. Nagyova B et al. Curr. Eye Res. 1999; 19 4–11. 4. Glasgow BJ et al. Curr. Eye Res. 1995; 14: 363–72.5. Burgess TL et al. Ann. Rev. Cell Biol. 1987; 3: 243–93.6. Dartt DA et al. Adv. Exp. Med. Bio. 1998; 438: 113–21.7. Jahn R et al. Eur. J. Biochem. 1982; 126: 623–9.8. Dartt DA et al. Am. J. Physiol. 1984; 247: G502–9.9. Gilbard JP et al. Invest. Ophthal. Vis. Sci. 1990; 31: 1381–8.

10. Gilbard JP et al. Arch. Ophthalmol. 1991; 109: 672–6.11. Koutavas H. PhD Thesis, University of New South Wales,

Australia, 1998.12. Laemmli UK et al. Nature 1970; 227: 680–85.13. Yates JR III. J. Mass Spectrom. 1998; 33: 1–19.14. Molloy MP et al. Electrophoresis 1997; 18: 2811–15.15. Cottrell JS. Pept. Res. 1994; 7: 115–24.