JOURNAL OF OCULAR PHARMACOLOGYAND THERAPEUTICSVolume 11, Number 3, 1995Mary Ann Liebert, Inc.

Sequential Bioactivationof Methoxime Analogs of

ß-Adrenergic Antagonists in the EyeNICHOLAS BODOR, LASZLO PROKAI, WHEI-MEI WU,

GABOR SOMOGYI, and HASSAN FARAG

Center for Drug Discovery, College of Pharmacy, University ofFlorida, Gainesville, Florida

ABSTRACT

Selected /S-adrenoreceptor antagonists are the antiglaucoma drugs of first choice inmost cases. However, a number of significant central nervous system (CNS),cardiovascular and respiratory side-effects following topical ocular installation of ß-blockers have also been reported. Site- and stereo-specific delivery to the eye of ß-blockers was achieved by application of a sequential bioactivation of the chemical deliverysystem (CDS) analogs of these drugs, which are converted to the active /?-blockersonly in the eye, thus systemic side effects are avoided. The corresponding ketoximeanalogs of the various /?-blockers were successfully applied and one of them, Alprenoxime,was tested in humans. The main problem with these compounds, however, is formulationstability: Alprenoxime has a tg0 of only 2-3 months. Thus, alternate structures weresearched. Methoxime analogs of selected /?-blockers were synthesized and their chemicalstability established at different pH's. Subsequently, their in vivo sequential enzymaticconversion was confirmed using HPLC. Comparative intraocular pressure (IOP) reducingactivity of the methoximes were then studied using normotensive rabbits. The methoximeanalogs showed significantly improved hydrolytic stability at pH ~ 7.0, the t90 is over 1 year.The in vivo sequential bioactivation, however, proceeds similarly to the oximes. Theintermediate ketones and the active /?-blockers can be detected in the various eyecompartments at different time intervals following topical administration of the methoximes.The mechanism of the eye-selective bioactivation and the results of the IOP reducingactivity studies will be discussed.

INTRODUCTION

/?-Adrenergic antagonists are still the drugs of choice for the treatment of glaucoma,based on their effect of reducing intraocular pressure (IOP). These drugs, however, were

developed for their systemic /^-blocking activity, and as such, their ophthalmic applicationis often accompanied by serious systemic side effects. Due to the fact that only a verysmall percentage (about 2%) of the topically applied drugs will reach the site of actionwithin the eye, and the rest will be absorbed systemically, it was found that eye drops of

Nagai Foundation, Tokyo-

Fellow 1994-1995

305

/?-blockers at 0.5% concentration cause reduction in heart rate, even in normal subjects (1).Similarly, and more importantly, hundreds of major respiratory events have been reportedwith the ophthalmic use of yff-blockers, (2-5) including a number of deaths. It is clear thatby simply eliminating the systemic side effects for the very same drugs, significant andsometimes dramatic improvement in the therapeutic indices could be achieved.

One of the most successful general approaches providing the potential for improvingtherapeutic indices of ophthalmic drugs involves the use of retrometabolic drug design (6).On one end of the "retrometabolic drug design loop" is the general concept of chemicaldelivery system (CDS). This approach uses CDS's, which are essentially inactive chemicalcompounds, as the drugs administered. However, by design, these CDS's have thespecific property to be differentially or exclusively activated by enzymes found at the siteof action. This activation does not take place in the rest of the body, thus true targeting(site-specificity) can be achieved.

The eye and, in particular, the iris-ciliary body is very rich in various enzymes whichgenerally are highly active (7). Thus, design of CDS's for ophthalmic use based on site-specific enzymes can be very rewarding. The first example was provided by the activationof diesters of adrenalone to epinephrine within the eye (8). This is a true site-specificchemical delivery system, where the ketone function is the targetor, which is being reduced,but only in the esterified adrenalone, to the alcohol necessary for the activity, followed byenzymatic cleavage of the protective ester functions. It was found, that the majority of thetopically applied CDS did not undergo activation to epinephrine in the rest of the body.This was the first successful site-specific ophthalmic application of the CDS concept. Sincethis involved reductive activation of /?-amino ketones to the corresponding/?-amino alcohols,a common pharmacophore both in ^-agonists and antagonists, we have considered thepossibility of extending this concept to the important class of yff-blockers. It was found,however, that the /?-amino ketones corresponding to /ff-blockers, such as propranolol, arevery difficult to synthesize, due to their chemical instability (9). On the other hand, theketones can successfully be stabilized in form of their oxime derivatives (9). Activation ofthese kinds of compounds, however, requires a potential enzymatic hydrolysis of theoximes, a process previously unknown. It was known that oximes do undergo hydrolyticconversion back to the ketone in aqueous solutions, but this process in general is far tooslow to produce the necessary ketone as a substrate for the reductase. Thus, it was hopedthat there are certain enzymes which could accelerate hydrolysis of the oxime. It was

found, indeed, that various /?-amino ketoxime analogs of /?-blockers do undergo facilehydrolysis within the eye to the corresponding ketones, which then are further reduced tothe active amino alcohol. It was further demonstrated that a sequence of enzymatichydrolysis-reduction produces, actually stereospecifically, the active S-(-isomer) (10). Atthe same time, even intravenous administration of large doses of the various ketoximes didnot produce any cardiovascular activity (11). Thus, a true site- and stereo-specific deliveryof the active yff-blockers to the eye was discovered, as shown in Scheme 1.

It was found, however, that even the oximes (R, = H) are not sufficiently stable inaqueous solution to provide the required shelf life stability in the formulation. In searchingfor more stable, but still enzymatically susceptible forms, the alkyl ethers of the oximeswere prepared. Specifically, various methoximes (R, = CH3) were synthesized and studied.Based on theoretically predicted mechanisms of the hydrolysis of the oxime, it wasconsidered that by eliminating the potential of the anion formation, or intramolecularhydrogen bonding between the oxime and the tertiary amine, one should be able toenhance chemical stability.

MATERIALS AND METHODSSynthesis

1-(lsopropylamino)-3-(1-naphthyloxy)-2-propanone hydrochloride. To a solution of1.0 g of propranolone oxime hydrochloride in 10 ml of hot water, 0.2 ml of cone HCI was

306

/o- RiN OIl II

Ar-

O-

CH2-

C-

CH2NHR2 -- Ar-

O-

CH2-

C-

CH2NHR2

OHT

Ar-

O-

CH2-

C-

CH2NHR2H

Ar - aryl; aryl-alkyl; heteroalkyl;

Ri - H or lower alkyl;,/ CH3 / LH3en.

R2 - branched alkyl; - CH. ; — C- CH3 ; etc...

»3CH3 \ CH,

Scheme 1. Sequencial bioactivation of ß-amino oximes.

added. The mixture was heated at 80°C for 30 minutes then cooled and left in therefrigerator overnight. The white solid was filtered and recrystallized from acetone-alcoholmixture (1:1) to give 0.7 g of crystalline propranolone hydrochloride, electrospray MS; m/z= 259 (M+H), m.p. 143-145°C.

1-(lsopropylamino)-3-(allylphenoxy)-2-propanone hydrochloride. To a solution of 1.0g of alprenolone oxime hydrochloride in 10 ml of hot water, 0.5 ml of cone HCI was added.The mixture was heated at 80°C for 10 minutes then cooled and left in the refrigeratorovernight. The white crystalline solid was filtered and dried to give 0.25 g of purealprenolone hydrochloride. The mother liquor was extracted with 25 ml of dichloromethane.The extract was dried with anhydrous sodium sulfate, concentrated to about half its volume,cooled in ice-acetone bath and 25 ml of anhydrous ether was then added to the extract.The solid separated was filtered and dried to give 0.55 g of pure product. The m.p. of thepure sample is 133-5°C (cf oxime hydrochloride 155-7°C), electrospray MS; m/z = 248(M+H).Solution Stability Studies

Solutions of propranolone hydrochloride were prepared by dissolving 2.94 mg (0.01mmol) of propranolone hydrochloride in 20 ml of 0.06 M phosphate buffer of pH's 6.5, 7.0and 7.4. The solutions were kept at room temperature or at 4°C, exposed to or away fromlight. Other solutions were prepared with the same concentrations of propranolonehydrochloride in buffer of pH 6.5 that contained 0.5 % propylene glycol (PG), 0.5 %polyethylene glycol (PEG), 0.025 mM 2-hydroxypropyl-yß-cyclodextrin (HPßCD), 0.05 mMHP/ïCD or 0.1 mM HP/ÎCD, respectively. These later solutions were kept in the refrigeratorat 4°C. All solutions were monitored for their concentrations of propranolone by HPLC andthe observed rates of its degradation were calculated.

The HPLC system used for propranolone stability consisted of a Spectra PhysicsSP8800 pump, Reodyne 7125 injector with 20 //I loop, a Spectra Physics SP8480 detector

307

operated at 450 nm and a Spectra Physics SP4270 integrator. Separation was done on a30 cm x 3.9 mm i.d. Phenomenex Bondclone C18 column with a mobile phase consistingof ACN, 0.1 M ammonium acetate, glacial acetic acid, triethylamine (15:35:1:0.01) at a flowrate of 1 ml/min.

Alprenolone oxime hydrochloride and alprenolone methoxime oxalate, respectively,were dissolved in 25 ml of isotonic phosphate vehicle (pH 4.5 to 6.5) by sonicating for 5min to obtain 35 mM solutions. The solutions were filtered, distributed in 3.0 ml portionsin borosilicate glass vials that protected against the change of pH (Fisher Scientific,Pittsburgh, PA). The vials were screw-capped with open-top phenolic closures using teflon-faced silicon rubber septa. The dead volume of the sample vials was purged with nitrogenbefore closing. The solutions were stored at room temperature (20 C), 37 °C, and 56 °C.During storage, 25-p\ aliquots were withdrawn from the vials and admixed to 500 p\aqueous bupranolol (2.5 mg/100 ml, internal standard) solution, and analyzed by high-performance liquid chromatography (12).

In Vivo Ocular Metabolism and Distribution

Adult New Zealand albino rabbits weighing 2.5-3.0 kg were used. Standard doses(100 p\) of 60 mM solution of alprenolone oxime hydrochloride or alprenolone methoximeoxalate in aqueous solution were administered topically in both eyes. Four animals wereincluded in a group. Each animal in a group were sacrificed at the same time afteradministration of the agent (15, 30, and 60 min, respectively, after the administration of theoxime, 15, 30, 45 and 90 min after the administration of the methoxime), and the eyes wereremoved surgically. Aqueous humor was obtained by making a single puncture at thelimbus using a 25 gauge x 5/8 in. needle attached to a 1 ml syringe. Then the cornea, theiris-ciliary body and the lens from both eyes were isolated. These tissues were

homogenized in a pH 9.8 alkaline buffer by a Tekmar STD tissuemizer to yield 20 % (w/w)homogenates. Aqueous humor was only diluted with the alkaline buffer before extraction.The samples were extracted twice with equal volumes of a méthylène chloride/diethyl ethermixture (1:4, v/v) by shaking for 15 min. The organic layer was then separated andreextracted with an equal volume of aqueous 0.01 M hydrochloric acid solution. Theaqueous layer was withdrawn, made alkaline again and the solvent extraction was

repeated. The organic layer was transferred into a deactivated glass vial, and the solventwas evaporated under a nitrogen stream. The residue was dissolved in 0.01 Mhydrochloric acid solution and analyzed by HPLC. Quantitation of alprenolol, alprenoloneoxime and alprenolone methoxime was done by a calibration curve obtained from samplesof known amount of compound added to tissue homogenates from untreated rabbit eyes.

The HPLC system consisted of a Spectra Physics (San Jose, CA) Model SP8800pump, a Rheodyne (Cotati, CA) Model 7125 injector with a 20 p\ loop, a Spectra PhysicsSP8480 detector operated at 280 nm, and a Spectra Physics SP4270 integrator. Separationwas done on a 7.5 cm x 4.6 mm i.d. 3-//m Supelcosil LC-8-DB column (Supelco, Bellefonte,PA) with mobile phase consisting of 30 % acetonitrile in 0.02 M monobasic potassiumphosphate (pH 3.0) buffer containing 0.01 % triethylamine. Flow rates were 1.1 ml/min forthe analysis, of tissue after administration of the oxime, and 1.5 ml/min for that of themethoxime. . Elution time for alprenolol was 6.2 and 4.3 min at 1.1 and 1.5 ml/min,respectively. Alprenolone E-oxime, alprenolone Z-oxime, and alprenolone methoxime(isomers not separated) eluted at 7.6, 8.9, and 10.2 min, respectively. Limit of quantitationfor each compound was around 100 ng/g tissue.

Ocular Pharmacology

Twelve healthy, normotensive, male New-Zealand Albino rabbits (about 3 kg of bodyweight) were used and kept in the individual cages with free access of food and water. The

308

animals were divided into several groups for crossover drug testing. Each formulation wastested on one group, and switched to another group for the following experiment, so everyanimal received all types of the treatment eventually. A Digilab Model 30Rpneumatonometer was used to determine the intraocular pressure (IOP). The IOPdetermination was performed on the unrestrained, conscious rabbits. One drop of localanesthetic, proparacaine hydrochloride (0.5%), was instilled in the eye immediately priorto each IOP measurement. Drug solutions (60 mM) were prepared freshly before eachexperiment, and water was used as the vehicle. After the IOP of both eyes were taken,drug solution (100 //I) was administered in one eye, and the fellow eye (instilled with water)was served as the control. IOP was determined in both eyes at appropriate time intervals.In order to perform the investigation on the healthy eyes, at least six days of recoveryperiod was allowed between two experiments. The experiments were carried out by thesame person using the same instrument under the same conditions. The IOP determinedprior and after the treatment was compared, and the activity of the compound wascalculated as the IOP difference between the treated eye and the control eye.

RESULTS AND DISCUSSION

In order to complete the picture, first the stability of the corresponding /?-aminoketones was revisited. Pure propranolone and alprenolone were synthesized. The stabilitystudies of propranolone are summarized on Table 1. It was found that propranolone,indeed, is highly unstable, having a t90 (the time within which 90% of the drug remainsintact) at pH 7.4 and room temperature of only about 13 hours. The stability can besomewhat improved by reducing the pH and the temperature. Further improvement wasachieved by using additives like a cyclodextrin (HP/ÏCD) or PEG. Nevertheless, the t90 evenin the most stable conditions, at 4°C at pH 6.5 and the presence of additives, cannot beimproved to more than about 17-18 days. The situation is not any better for alprenolone,the ketone corresponding to the oxime, which was found to be the most promisingcompound for reducing intraocular pressure, producing high activity following topical orintravitreal application. This compound, Alprenoxime, was actually taken to clinical trialsand has successfully passed Phase I studies, demonstrating complete lack ofcardiovascular side effects, even in humans, including isoprenaline induced tachycardia(13).

Problems with aqueous stability, however, emerged as shown in Table 2. The bestone can achieve is a t90 of 55 days in slightly acidic (pH 4.5) conditions. Although thiscould be sufficient for a freeze dried/reconstituted formulation, it was decided to search formore stable analogs.

The methoxime analog was prepared by a multi-step synthetic sequence, similar tothe corresponding oxime (14). It was isolated as the oxalate salt, which was obtained asa reproducible, equilibrium mixture of about 75% Z- and 25% E- isomer. No attempts weremade to develop the compound as a pure isomer of one or the other, since it was shownpreviously in the case of Alprenoxime that a pH and temperature dependent equilibrium isin solution around this isomer composition, and the two isomers have very similar activity.It was also found that, unlike in water where the isomerization is a very slow process, theequilibrium is established in an extremely short cime in most biological fluids, thus there isno point in using one or the other isomer. The oximes and methoximes are still inactiveand their activation is a rather similar process; there is, indeed, no justification to separatethe two isomers.

Next, the solution stability of the methoxime was studied and, as shown in Table 2,a dramatic improvement was observed. Actually, at more favorable pH's (around 6.5) androom temperature, the methoxime of alprenolone did not show any significant hydrolysisfor a year.

309

TABLE 1.

Stability of Propranolone in Solutions

No. System kobs (h"1) t90 (h)

1 pH 7.4,RT,light 7.84X10"3 13.4

2 pH 7.4, 4°C 2.71X10'3 38.8

3 pH 7.0,RT,light 3.89X10"3 27.0

4 pH 7.0,4°C 1.97X1--3 53.3

5 pH 6.5,RT,light 7.78X10"3 13.5

6 pH6.5,RT 4.78X10"3 21.9

7 pH 6.5,4°C 2.04X10"3 51.5

8 pH6.5,4°C,CD (1:0.5 MR) 8.40X10"4 125.0

9 pH6.5,4°C,CD(1:1 MR) 7.85X10"4 133.8

10 pH6.5,4°C,CD(1:2 MR) 5.24X10"4 200.5

11 pH 6.5,4°C,0.5% PG 6.04X10"4 174.4

12 pH 6.5,4°C,0.5% PEG 2.48X10'4 423.6

NOTES: 1- MR = Molar ratios (propranolone : CD- 2-hydroxy propyl-ß-cyclodextrin)

2- Corr.Coefficients not less than 0.9973- SD of Kobs for No. 8-12 are less than ± 0.22X10"4

It was demonstrated before that, after application of alprenolone oxime to the eye,the corresponding ketone and alprenolol can be detected in various eye compartments.4Thus, it was very important to determine if the similar enzymatic hydrolysis process thatproduces the bioreducible ketone will take place in the case of methoxime (15). The overallactivity of any of these ^-antagonists oxime or methoxime analogs depends on thecombination of the various rate processes described below, whereby the CDS oxime is firstpenetrating the cornea, and reaches various compartments within the eye. Enzymatichydrolysis then provides the ketone, which is further bioactivated to the corresponding ß-blocker, presumably in the iris-ciliary body. Accordingly, the rates corresponding to thetransport and the enzymatic hydrolysis will determine how much of the active componentwill ultimately be produced. An HPLC analysis of the iris-ciliary body of the rabbit eye 1 hrafter administration of the alprenolone methoxime (100 p\ of 60 mM solution) indicatesclearly the presence of the methoxime and relatively large concentrations of the alprenolol,

310

TABLE 2.

Solution Stability at Different Temperature and pH (60 mM Oxime, Isotonic Vehicle)

A. Alprenolone Oxime

20°C 37°C 56°CpH t90 (days) ty, (days) t% (days)

6.5 20±2 22±2 3±0.5

5.9 33+3 35±3 7±1

5.3 40+5 43±5 19±2

4.5 55+5 90±10 28±3

B. Alprenolone Methoxime

20°C 37°C 56°CpH t90 (days) tA (days) ty¡ (days)

6.5 >365 700±30 150±5

4.5 74±8 85±5 22+3

and also a small amount of the intermediate ketone (Figure 1). This relative distribution ofthe components is different than the one obtained when the oxime was administered, wheregenerally less alprenolol is found. The changes in the relative rate processes between theoxime and methoxime are further reflected by the time distribution of the major components(i.e., the initial CDS, alprenolone oxime or methoxime, and alprenolol) in various eyecompartments. First, the relative distribution of the CDS's, (alprenolone oxime andalprenolone methoxime) in the various eye compartments is compared on Figure 2a andb. While, as expected in both cases, the highest concentrations are in the cornea, whichis the major barrier, major differences in the concentrations in the iris-ciliary body are found.While relatively high concentrations of the oxime is in the iris-ciliary body, very smallmethoxime concentrations are found there. Accordingly, the situation is just the oppositewhen analyzing the active reduced product, alprenolol. As shown on Figures 3a and b, thealprenolol concentrations are significantly higher from the methoxime, than when the oximeis administered. The numerical values and the major differences mentioned are listed inTable 3.

Finally, the activity of the alprenolone methoxime was compared to that of the oximeand alprenolol. As shown on Figure 4, both the oxime and the methoxime are at least asactive as alprenolol in reducing intraocular pressure (IOP) in normotensive rabbits.Consistent with the finding on the distribution, the onset of action for the methoxime isfaster; it is producing almost the same pattern as alprenolol itself (16). On the other hand,it appears that the oxime is somewhat longer acting. The methoxime, however, shows very

311

min.

FIGURE 1. HPLC Analysis of the Iris-ciliary Body of Rabbit Eyes 1 Hr After theAdministration of 100 p\ of 60 mM Alprenolone Methoxime Solution. (A

-

Alprenolol, B-Alprenolone, C

-

Alprenolone Methoxime).

high activity still at three hours and then slowly returns to normal. As the relative enzymaticactivity in humans, as compared to rabbits in terms of oxime hydrolysis, is unknown, therelative value of oxime and methoxime in this respect cannot be predicted. It isencouraging, however, that the methoxime appears to parallel more the active yff-blocker,while at the same time provides a significant advantage in terms of formulation stability overthe oxime.

It has been interesting to see that the corresponding ketone, alprenolone, did notshow any activity in reducing IOP in normotensive rabbits, as shown in Figure 5. This isprobably due again to the significant changes in the transport-activation processes,

312

(a) 60

40 60Time (min)

120

120Time (min)

FIGURE 2. Concentration of Alprenolone Oxime (a) and Alprenolone Methoxime (b) inOcular Tissues of Rabbits After Topical Administration of 100 p\ of 60 mM Solutions. DataIndicate Mean ± S.E. (n=4).

313

(a) 60

50 H

40 -I=>wtn-30CDO)

20 H

10H

0

"* Cornea* Iris-ciliary body+ Aqueous humor Lens

20 40 60

Time (min)80 100 120

40 60Time (min)

120

FIGURE 3. Concentration of Alprenolol After Topical Administration of Solutions (100 //I,60 mM) of Alprenolone Oxime (a) and Alprenolone Methoxime (b). Data indicate Mean ±S.E. (n=4).

314

TABLE 3

AUC of Oxime/Methoxime Concentrations in Different Eye Compartments After Topical Administration of 100 p\ of theOximes/Methoxime in Equimolar Doses, 60 mM, (NZWR) (nmol/g x hr or nmol/ml x hr)

Compound Administered Cornea Aqueous Humor Iris-Ciliary Body Lens Vitreous Body

Propanolone Oxime

Betaxolone Oxime

Alprenolone Oxime

Alprenolone Methoxime

88.85+13.87

46.67+2.79

95.59±16.4

96.22+1.34

3.67+024

18.42+0.88

22.75+3.53

32.91+5.00

14.32+2.02

31.65+4.62

4697+12.63

5.96+0.34

7.66+044 9.83+061

6.47±1.03 3.78+0,48

21.61+3.21 2.21+0.25

AUC of 0-Adrenergic Antagonist Concentrations in Different Eye Compartments After Topical Administration of 100 p\ ofthe Oximes/Methoxime in Equimolar Doses, 60 mM, (NZWR) (nmol/g x hr or nmol/ml x hr)

Compound Administered Cornea Aqueous Humor Iris-Ciliary Body Lens Vitreous Body

Propanolone Oxime

Betaxolone Oxime

Alprenolone Oxime

Alprenclone Methoxime

864+1.52

5.37+0.87

8.38+1.71

15.46±1.00

202+029

13.89 + 1.67

4.43+1.82

17.74±1.17

10.03+ 1.26

24.77 + 1,16

19.45+3.07

80.61+4.78

8.23 + 0,25

1.68+0.44

1.55+0.23

6,86 + 0.50

1.99+1 44

1.85 + 0.17

AUC was calculated by the trapezoidal rule after randomly combining four time points from different animals. Shown aremean + SEM (n=4).

CDXEE

m

Q.O

c(S

JZO

Time, hrFIGURE 4. IOP Reducing Activity in Rabbits After Topical Administration of 100 //I of 60mM Compound Solution. Vehicle Control (A), Alprenolol (O), Alprenolone Oxime ( ),Alprenolone Methoxime (•). Data Indicate Mean ± S.E. (n=12).

315

TABLE 4.

Percent Change in IOP After Ocular Administration of 60 mM of the Compounds inRabbits.12

Time, Vehicle Betaxolol Betaxolone Betaxolone Betaxolonehr Control HCI Methoxime Methoxime Oxime

HCI Oxalate Maléate

0 0.00±0.00 0.00±0.00 O.OOiO.OO 0.00±0.00 0.00±0.00

0.5 -3.31±0.89 -10.17±1.50 -7.04±1.52 -5.33±1.02 -9.74±2.15

1 -0.95±1.28 -8.24±1.23 -5.83l1.92 -12.27±1.28**-14.76±2.22**

2 6.01±1.12 -7.01±2.03 -5.12±1.98 -12.81±2.44**-14.63±1.22**

3 2.59±0.86 -1.46±2.58 0.45±1.57 -6.24±1.32** -8.13±1.45**

4 1.63±1.02 3.77±3.01 6.32±2.78 2.61 ±1.67 -3.13±1.74

5 3.01 ±0.91 1.09±1.28 5.76±1.39 -0.38±0.85 -0.24±1.18

6 2.04±0.76 349±0.90 8.49±1.94 1.54±1.17 1.71±1.08

Maximum% reduced -4.45±0.73 -12.65±1.61* -10.62±1.89* -16.49±1.91 * -18.10±1.50*

Nine rabbits (eighteen eyes) were investigated in crossover experiments. Each drug wasadministered two times, at 0 and 10 min, bilaterally. H2O was used as vehicle.2Table entries are the mean ± SE mmHg of twelve rabbits.** p < 0.05 when compared to betaxolol HCI, at the same time point, using student t-test.* p < 0.005 when compared to control using student t-test.

probably not allowing high enough sustained alprenolol concentrations to be producedwithin the eye.

It was previously reported, that the oxime analog of one of the other importantophthalmic yff-blockers, betaxolol, shows dose dependent higher activity in reducing IOP innormotensive rabbits than betaxolol itself (16). We have recently restudied this importantcompound, as its activity was compared in a comprehensive cross-over study with thecorresponding betaxolol itself. As shown in Table 4, the oxime shows very good activityconsistently higher than betaxolol itself at the same molar concentration.

In conclusion, the successful application of the retrometabolic design approach fortargeting IOP-reducing /?-adrenergic antagonists to the eye was accomplished. Furthersignificant improvements are reported by using the corresponding methoxime analogs, as

opposed to the previous oximes, in terms of improved formulation stability. All data indicatethat this new class of compounds should be very important and much improved safeantiglaucoma agents.

316

10

Time, hr

FIGURE 5. Effect of Topical Administration of Alprenolone (100 //I, 60 mM Solution) onRabbit IOP. Vehicle Control (O), Alprenolone (•). Data Indicate Mean ± S.E. (n=12).

ACKNOWLEDGMENTS

Financial support by NIH (R01 EY09098) is gratefully acknowledged. We aregrateful to Ms. Julie Berger for helping prepare the manuscript.

REFERENCES

1. Mekki, Q., Warrington S. and Turner P. Ocular and cardiovascular effects of timololand pindolol eyedrops in normal volunteers. Br. J. Clin Pharmac. 16:632-633, 1984.

2. Schoene, R., Martin, T., Chavan, N. and French, D. Timolol induced bronchospasmin asthmatic bronchitis. J. Am. Med. Assoc. 245:1460-1461, 1984.

3. Schoene, R., Ward, R., Abuan, T. and Beasley, C. Effects of topical betaxolol,timolol and placebo on pulmonary function in asthmatic bronchitis. Am. J. Qphth.97:86-92, 1984.

4. Richard, R. and Tattersfield, A. Bronchial /?-adrenoreceptor blockade followingeyedrops of timolol and its isomer L-714 in normal subjects. Br. J. Clin. Pharmac.20:459-462, 1985.

5. Fraunfelder, F., and Barker, A. Respiratory effects of timolol, New Enql. J. Med.311:1441, 1985.

317

6. Bodor, N. New methods of drug targeting, in Trends in Medicinal Chemistry '90,Sarel, S., Mechoulam, R. and Agranat, I., eds. Blackwell Scientific Publications,Oxford, England, 1992, p. 35-44.

7. Shichi, H. and Nebert, D. Drug metabolism in ocular tissues, in ExtrahepaticMetabolism of Drugs and Other Foreign Compounds, Gram, T., ed. S.P. Medical andScientific Books, New York, 1980, p. 333-363.

8. Bodor, N. and Visor, G. Formation of adrenaline in the iris-ciliary body fromadrenalone diesters, Exp. Eye Res. 38:621-626, 1984.

9. Bodor, N., Elkoussi, A., Kano, M. and Nakamura, T. Improved delivery throughbiological membranes 26. Design, synthesis, and pharmacological activity of a novelchemical delivery system for /7-adrenergic blocking agents, J. Med. Chem. 31:100-106, 1988.

10. Bodor, N. and Prokai, L. Site- and stereospecific ocular drug delivery by sequentialenzymatic bioactivation, Pharm. Res. 7:723-725, 1990.

11. Bodor, N. and Elkoussi, A. Improved delivery through biological membranes LVI.Pharmacological evaluation of alprenoxime, a new potential glaucoma agent, Pharm.865,8:1389-1395, 1991.

12. Prokai, L., Simay, A., and Bodor, N. Reversed-phase high performance liquidchromatography of ketoxime analogues of /?-adrenergic blockers, J. of Chromatoq.541:469-473, 1991.

13. Howes, J. Progress Reports, Xenon Vision, Inc., Alachua, FL, 1991 (unpublished).

14. Simay, A., Prokai, L., and Bodor, N. Oxidation of aryloxyaminoalcohols with activateddimethylsulfoxide: A novel C-N oxidation, facilitated by neighboring group effect,Tetrahedron 45(13):4091-4102. 1989.

15. Prokai, L., Wu, W., Somogyi, G., and Bodor, N. Ocular delivery of the yff-adrenergicantagonist alprenolol by sequential bioactivation of its methoxime analog, J. Med.Chem. (in press), 1995.

16. Bodor, N. Designing safer ophthalmic drugs, in Proceedings of the Xth InternationalSymposium on Medicinal Chemistry, Trends in Med. Chem. 88, van der Goot, H.,Domany, G., Palios L., and Timmerman, H., eds. Elsevier Science Publishers, B.V.,Amsterdam, 1989, p. 145-164.

Received: April 5, 1995

Accepted for Publication: May 31, 1995

Reprint Requests: Nicholas Bodor, Ph.D.Center for Drug DiscoveryCollege of PharmacyUniversity of FloridaBox 100497 JHMHCGainesville, Florida 32610-0497U.S.A.

318