Melatonin, the MT2 melatonin receptor agonist IIK7 [N-butanoyl-2-(2-methoxy-6H-isoindolo[2,1-a]indol-11-yl)ethanamine], and the putative MT3 melatonin receptor agonist 5-MCA-NAT [5-methoxycarbonylamino-N-acetyltryptamine] have previously been shown to reduce intraocular pressure (IOP) in ocular normotensive rabbits. To gain a better understanding of the structure-activity relationship of compounds that activate MT2 and MT3 receptors mediating reductions in IOP, novel melatonin analogs with rationally varied substitutions were synthesized and tested for their effects on IOP in ocular normotensive rabbits (n = 160). All synthesized melatonin analogs reduced IOP. The best-effect lowering IOP was obtained with the analogs INS48848 [methyl-1-methylene-2,3,4,9-tetrahydro-1H-carbazol-6-ylcarbamate], INS48862 [methyl-2-bromo-3-(2-ethanamidoethyl)-1H-indol-5-ylcarbamate], and INS48852 [(E)-N-(2-(5-methoxy-1H-indol-3-yl)ethyl)-3-phenylprop-2-enamide]. These compounds produced dose-dependent decreases in IOP that were maximal at 0.1 mM (total dose of 0.259 μg for INS48848, 0.354 μg for INS48862, and 0.320 μg for INS48852) and 1 mM (total dose of 2.59 μg for INS48848, 3.54 μg for INS48862, and 3.20 μg for INS48852), with maximal reductions of 36.0 ± 4.0, 24.0 ± 1.5, and 30.0 ± 1.5% for INS48848, INS48862, and INS48852, respectively. Studies using melatonin receptor antagonists (luzindole, prazosin, and DH97 [N-pentanoyl-2-benzyltryptamine]) indicated that INS48862 and INS48852 activate preferentially a MT2 melatonin receptor and suggest that INS48848 may act mainly via a MT3 receptor. The most effective compounds were also well tolerated in a battery of standard ocular surface irritation studies. The implication of these findings to the design of novel drugs to treat ocular hypertension is discussed.
Glaucoma is a group of diseases characterized by retinal and optic neuropathy and progressive visual field loss. The most prevalent type, open angle glaucoma, is estimated to account for approximately 15% blindness worldwide (Thylefors and Négrel, 1994). The pathology of this type is secondary to elevated intraocular pressure (IOP), and reduction of IOP is the most common treatment modality.
In the normal eye, a balance of formation and outflow of aqueous humor regulates IOP, maintaining a mean IOP of approximately 16 mm Hg. In open angle glaucoma, normal aqueous humor outflow through the trabecular meshwork is impeded, and there is a consequential rise in IOP to values higher than 21 mm Hg (Schottenstein, 1996).
Circadian fluctuation of IOP is well established, and the relationship between melatonin and IOP has been explored in view of the involvement of pineal melatonin in the regulation of many circadian rhythms (Moore, 1997; Lewy, 1999). Despite contradictory findings, the most widely held conclusion is that decreased IOP correlates positively with increased intraocular melatonin levels (Rohde et al., 1985; Komaromy et al., 1998; Pointer, 1997). However, pharmacological characterization of melatonin receptors reveals three receptor subtypes (MT1, MT2, and MT3; Dubocovich, 1995) and that the relationship between individual receptor subtypes and IOP has not been evaluated comprehensively. The selective MT3 receptor ligand 5-methoxycarbonylamino-N-acetyltryptamine (5-MCA-NAT) has been shown to be a potent ocular hypotensive agent (Pintor et al., 2001, 2003; Serle et al., 2004), and the selective MT2 receptor agonist N-butanoyl-2-(2-methoxy-6H-isoindolo[2,1-a]indol-11-yl)ethanamine (IIK7; Sugden et al., 1999) also markedly decrease the IOP and was inhibited by selective MT2 receptor antagonists (Alarma-Estrany et al., 2008). Both 5-MCA-NAT and IIK7 offer themselves as starting points for new classes of drugs to lower IOP and for treating ocular hypertension and glaucoma.
With the 5-MCA-NAT and IIK7 findings as a starting point, we prepared a series of 16 compounds to evaluate the structure-activity relationships that determine the ability of these compounds to lower IOP. The compounds were tested for their ability to lower IOP in ocular normotensive rabbits, and molecular features associated with enhanced efficacy were determined. Furthermore, as a preliminary assessment of safety, ocular surface tolerance tests were carried out with the most efficacious compounds.
Materials and Methods
Compounds INS48476, INS48497, INS48834, and INS48887 were synthesized using a general procedure developed by Macor et al. (1993). In brief (Fig. 1A), an appropriately substituted 5-nitroindole was treated with oxalylchloride followed by ammonia and then reduced with borane to produce the tryptamine derivative. Acylation was effected with an appropriate anhydride or acid chloride. The nitro group then was reduced and acylated with an appropriate chloroformate. Compound INS48862 was formed by site-specific bromination of the parent compound using N-bromosuccinimide/acetic acid.
Compounds INS48803 and INS48862 were obtained by alkylation of precursor acyltryptamines with NaH/methyl iodide/dimethyl formamide and NaH/benzyl bromide/N,N-dimethylformamide, respectively, whereas compound INS48848 was obtained by hydrogenation and acylation of the parent nitrotetrahydrocarboline. Melatonin analogs (Fig. 1B) INS48836 and INS48852 were prepared by treatment of 5-methoxytryptamine with the appropriate acid chloride and triethylamine in dichloromethane, whereas INS48838, INS48853, and INS48882 were prepared by treatment of 5-methoxytryptamine with the appropriate anhydride and triethylamine in dichloromethane. Compounds INS48863 and INS48864 were formed by treatment of 5-methoxytryptamine with the corresponding sulfonyl chlorides and triethylamine in dichloromethane. Compound INS48879 was synthesized by treatment of 5-methoxytryptamine with diphenylphosphoryl chloride and triethylamine in dichloromethane. Computation of ADME (absorption, distribution, metabolism, and excretion) and molecular properties was carried out in QikProp (version 1.6; Schrodinger, New York, NY) and visualized using Spotfire 7 (Somerville, MA). All compounds exhibited satisfactory 1H NMR and high-resolution mass spectra.
Normotensive New Zealand white rabbits (160), weighing 3 ± 0.5 kg, were kept in individual cages with food and water ad libitum. They were maintained under a controlled 12-h/12-h light/dark cycle in the School of Optics' animal facilities (Universidad Complutense de Madrid Animal House). All of the procedures complied with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmology and Vision Research and also are in accordance with the European Communities Council Directive (86/609/EEC).
Formulation and Method of Administration.
All compounds were formulated in isotonic saline containing 1% dimethyl sulfoxide (DMSO) and tested at a final concentration of 0.648 μg for INS48848, 0.885 μg for INS48862, 0.801 μg for INS48852, 0.811 μg for INS48864, 1.056 μg for INS48879, 0.723 μg for INS48476, 0.885 μg for INS48793, 0.723 μg for INS48834, 0.813 μg for INS48497, 0.616 μg for INS48838, 0.651 μg for INS48853, 0.736 μg for INS48836, 1.142 μg for INS48882, 0.728 μg for INS48887, 0.705 μg for INS48863, and 0.758 μg for INS48803 (all compounds of 0.25 mM, giving a dose of 2.5 nmol in a volume of 10 μl). Compounds were applied unilaterally to the cornea at a fixed volume of 10 μl. The contralateral eye received the same volume of vehicle (1% DMSO v/v, 0.9% w/v saline). Because the application of the tonometer may produce discomfort in the rabbits, corneas were anesthetized by applying 10 μl of oxibuprocaine/tetracaine (1:4; Colicursi Double Anaesthetic; Alcon Cusi, Barcelona, Spain) before IOP was measured.
Experiments were performing following a blinded design; no indication was given to the experimenter with regard to the applied solution (agent or vehicle). IOP measurements were made using a Tonopen aplanation contact tonometer before and at several times after instillation of a compound. IOP was followed up to 5 h to study the time course of the effect. Afterward, the three most active compounds at decreasing IOP were tested again to study the time course of up to 8 h at a total dose of 0.259 μg for INS48848, 0.320 μg for INS48852, and 0.354 μg for INS48862 (0.1 mM final concentration, giving a dose of 1 nmol in a volume of 10 μl). To know which melatonin receptors were being activated, substances that produced hypotension were tested in the presence of N-pentanoyl-2-benzyltryptamine (DH97), prazosin, or luzindole. Ten microliters of these antagonists were added 30 min before the application of either melatonin analog at a dose of 100 μg (29.9 mM DH97, 23.82 mM prazosin, and 34.20 mM luzindole). Furthermore, dose-response experiments were carried out for the three most active compounds. On any given day, only a single dose was tested on a single animal, which was washed out at least 2 days between doses.
To examine ocular surface short-term tolerance, 50 μl (0.25 mM) of each formulated compound was instilled onto the cornea of both eyes in three animals, and observations were made using the Draize scale (Draize et al., 1944) immediately before treatment and at 2, 5, 15, and 60 min after treatment. The Draize scale is a long-established standard for assessing ocular surface tolerability and involves scoring six components of the eye's anterior segment: conjunctival chemosis, discharge and redness; corneal opacity and involvement; and iritis.
Structure-Activity Relationships and ADME Studies.
Table 1 displays the structure of each compound along with its maximal IOP-lowering effect during the 5-h period of examination. In general, compounds in which R4 was an alkoxycarbonylamino group were equally or slightly more efficacious than those in which R4 was a methoxy group, followed in efficacy by −OMe derivatives. INS48848, INS48852, and INS48862 showed some differences in quantitative calculation of ADME and molecular properties in QikProp. LogBB (which is a measure of the partitioning between brain tissues and the blood) of the selected efficacious compounds was relatively poor and clustered (range −1.3 to −0.8) relative to the broad set (range −1.5 to −0.4). A unique relationship was identified between the hydrophobic solvent-accessible surface area (FOSA), the electron affinity (EA), and the dipole moment of the three most efficacious compounds (Fig. 2). All three are set apart from the majority of the set with substantially higher EAs (mean “0.3 eV versus mean” 0 eV for the complete array), lower dipole moments (mean “4.7 versus mean” 6.5 for the complete array), and somewhat lower FOSA (mean “225 Å3 versus mean” 290 Å3 for the complete array).
Dose-Response and Time Course Experiments.
All of the synthesized compounds exerted a modulatory effect on IOP as can be seen in Table 1. INS48848, INS48862, and INS48852, which evoked the greatest reductions in IOP at a total dose of 0.648 μg for INS48848, 0.885 μg for INS48862, and 0.801 μg for INS48852 (0.25 mM, giving a dose of 2.5 nmol in a volume 10 μl), produced dose-dependent decreases in IOP, which were maximal at 0.1 mM (total dose of 0.259 μg for INS48848, 0.354 μg for INS48862, and 0.320 μg for INS48852), −1 mM (total dose of 2.59 μg for INS48848, 3.54 μg for INS48862, and 3.20 μg for INS48852) (1–10 nmol, 10 μl), with maximal reductions of 36.0 ± 4.0, 24.0 ± 1.5, and 30.0 ± 1.5% for INS48848, INS48862, and INS48852, respectively (n = 8; Fig. 3). The analysis of the curves permitted the calculation of the pD2 values (−logEC50) of 5.5 ± 0.2, 5.7 ± 0.3, and 5.5 ± 0.2 for INS48848, INS48862, and INS48852, respectively. These values are equivalent to doses of 8.19 ng for INS48848, 7.04 ng for INS48862, and 10.12 ng for INS48852.
The time course of the changes in IOP induced by the melatonin analogs was also examined. The three compounds, administered at a single dose of 0.259 μg for INS48848, 0.354 μg for INS48862, and 0.320 μg for INS48852 (0.1 mM, giving a dose of 1 nmol in a volume of 10 μl), evoked a reduction of the IOP, with a maximal effect observable between 30 min and 2 h after application (Fig. 4). In particular, and in agreement with the concentration-response studies, the melatonin analog INS48848 reduced IOP by 35% at 1 h, INS48852 produced a 31% reduction at 2 h, and INS48862 decreased IOP by 24% at 30 min. For all of these compounds, IOP returned to its normal level within 8 h of the instillation (n = 8).
Effects of Melatonin Receptor Antagonists.
Pretreatment with the nonspecific melatonin receptor antagonist luzindole (100 μg, 30 min) abolished the ocular hypotensive effect of INS48848 (0.259 μg) (1 nmol/10 μl) but did not inhibit the effect of the INS48862 or INS48852. Pretreatment with the MT3 binding site antagonist, prazosin, abolished the ocular hypotensive effect of INS48848 (0.259 μg) (1 nmol/10 μl) but did not inhibit the actions of INS48862 (0.354 μg) or INS48852 (0.320 μg) (both at 1 nmol/10 μl). The MT2 antagonist DH97 (100 μg equivalent to 29.9 mM) inhibited the effects of INS48862 and INS48852 but had not effected against INS48848 (Fig. 5). We extended the experiments with this MT2 melatonin receptor antagonist by assaying it at graded concentrations from 0 to 100 μg (Fig. 6). These experiments confirmed that INS48862 and INS48852 were antagonized by DH97 but not INS48848. These results indicate that INS48862 and INS48852 activate preferentially a MT2 melatonin receptor and suggest that INS48848 may act mainly via a MT3 receptor.
By the criteria, of conjunctival chemosis, discharge, and redness; corneal opacity and involvement; and iritis in the Draize scale, all of the compounds tested were well tolerated (results not shown).
In the present experimental work, we describe the synthesis of new melatonin analogs that reduce IOP in normal New Zealand White rabbits. Because there are substantial differences between human and rabbit ocular anatomy and physiology (Bito, 1984), the screening of the best compounds was based on maximal IOP-lowering effect rather than on the effect at a selected time point.
The outstanding feature associated with high efficacy appears to be structural rigidity and increased hydrophobicity on the eastern hemisphere of the molecule, as exemplified by INS48848, INS48852, and INS48862. In the case of compounds INS48879 and INS48864, the positive effect of added hydrophobicity may be mitigated by the higher charge density of the phosphoryl and sulfonyl groups compared with an acyl group.
Substitution of the N-acetyl group of the 3-position side chain with a charged functional group or a highly polarized group (INS48863, INS48882, and INS48887) almost abolished IOP-lowering activity, as did methylation of the R4-carbamate (INS48803), suggesting that hydrogen bonding ability at R4 may be a critical point for activity. However, quantitative calculation of ADME and molecular properties in QikProp revealed significant disparity between INS48848 and the other two most efficacious compounds (INS48852 and INS48862). Although it is tempting to speculate about differences in binding mode, based upon this information, it is also important to understand that because these compounds are administered onto the ocular surface, more specific ADME properties, such as Topliss' corneal permeability (Yoshida and Topliss, 1996), could be more relevant. Examination of calculated molecular properties was somewhat more enlightening. Thus, we observed a unique relationship between the FOSA, EA, and the dipole moment of the three most efficacious compounds.
Using a 0.1 mM concentration of INS48848 (total dose 0.259 μg), INS48852 (total dose 0.320 μg), and INS48862 (total dose 0.354 μg) to easily compare with concentrations of applied melatonin, 5-MCA-NAT, and IIK7 in recent studies (Alarma-Estrany et al., 2007, 2008, 2009), responses were similar to those obtained using 0.25 mM. Moreover, these compounds decreased IOP in a dose-dependent manner similar to melatonin, 5-MCA-NAT, and IIK7 (Pintor et al., 2001; Alarma-Estrany et al., 2008), confirming the efficiency of these melatonin analogs for decreasing IOP in a way similar to 5-MCA-NAT and IIK7.
The effects of INS48848 were completely blocked by prazosin, an antagonist of MT3 melatonin receptors (Paul et al., 1999; Pintor et al., 2003; Xia et al., 2008), and were potently inhibited by luzindole, a nonselective antagonist of melatonin receptors (Pintor et al., 2003). However, DH97, an MT2 receptor antagonist (Chen et al., 2005; Alarma-Estrany et al., 2007; Mendoza-Vargas et al., 2009), had little effect against INS48848. In sharp contrast, the results obtained for INS48862 and INS48852 were the opposite. Luzindole and prazosin had no significant effects against these two compounds, whereas DH97 blocked them completely. These results strongly suggest that the compound INS48848 could be acting through the MT3 melatonin receptors and that the compounds INS48862 and INS48852 could be acting preferentially through MT2 melatonin receptors.
Taking account that prazosin is also a α1-adrenoceptor antagonist, it should be considered that INS48848 could have been acting through adrenoceptors instead of melatonin receptors. However, this remains to be investigated in detail; 5-MCA-NAT, which is similarly antagonized by prazosin, is not antagonized by the α1-selective adrenoceptor antagonist corynanthine (Pintor et al., 2003).
Considering that the compounds described here can activate MT2 and MT3 receptors, an interesting approach could be to combine them to get stronger reductions in IOP. According to our results, a combination of INS48852 (which produces 30% IOP reduction via MT2 receptors) plus INS48848 (40% reduction via MT3) could be an interesting one. Nevertheless, it would be necessary to optimize the concentrations and instillation volumes of both compounds to get the most effective formulation.
Designing and developing new drugs to treat glaucoma requires consideration of many factors, including efficacy, specificity, bioavailability, safety, and toxicity. Our present study has begun to probe the viability of a new class of compounds as a source for IOP-lowering drugs. With these studies in rabbits, we begin to establish the relationship of varied molecular properties and substitution patterns with efficacy. Furthermore, we have determined that topical application of representative compounds from this class is well tolerated and does not cause short-term ocular surface irritation. In addition, we have recently found out that it is possible to dissolve these compounds in solvents that are acceptable for use in topical applications for humans (Andrés-Guerrero et al., 2009), rather than in DMSO or ethanol, which are not. Therefore, these novel drugs have clinical potential to treat ocular hypertension and glaucoma.
Participated in research design: Yerxa and Pintor.
Conducted experiments: Alarma-Estrany, Peral, Pelaez, Huete, and Plourde.
Contributed new reagents or analytic tools: Plourde.
Performed data analysis: Guzman-Aranguez and Pintor.
Wrote or contributed to the writing of the manuscript: Guzman-Aranguez and Pintor.
We thank Dr. Charles H. V. Hoyle for critical reading of this manuscript.
This work was supported by grants from Ministerio de Ciencia e Innovación [SAF2007-60835, SAF2010-16024]; RETIC Red de Patología Ocular del Envejecimiento, Calidad Visual y Calidad de Vida [RD07/0062/0004]; NEUROTRANS CM [S-SAL 0253-2006]; and BSCH-UCM [GR58/08].
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- intraocular pressure
- electron affinity
- hydrophobic solvent-accessible surface area
- dimethyl sulfoxide
- diphenyl (5-methoxy-1H-indol-3-yl)methylphosphoramidate
- 1-((5-methoxy-1H-indol-3-yl)methyl)-6-oxopiperidine-2-carboxylic acid
- Received December 16, 2010.
- Accepted March 1, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics