Physicochemical and Biopharmaceutical Characterization of endo-2-(8-Methyl-8-azabicyclo[3.2.1]oct-3-yl)-2, 3-dihydro-1H-benz[e]isoindol-1-one (CR3124) a Novel Potent 5-HT3 Receptor Antagonist ANDREA CAPPELLI,1 VALTER TRAVAGLI,1 IACOPO ZANARDI,1 MAURIZIO ANZINI,1 GIANLUCA GIORGI,2 ALESSANDRO DONATI,3 MARIANNA AGGRAVI,3 MARIO CASOLARO,3 MASSIMO FRESTA,6 EUGENIO PACCAGNINI,4 FRANCESCO MAKOVEC,5 SALVATORE VOMERO1 1
Dipartimento Farmaco Chimico Tecnologico, Universita` degli Studi di Siena, Via A. Moro, 53100 Siena, Italy
2
Dipartimento di Chimica, Universita` di Siena, Via A. Moro, 53100 Siena, Italy
3
Dipartimento di Scienze Chimiche e dei Biosistemi, Universita` degli Studi di Siena, Via A. Moro, 53100 Siena, Italy
4
Dipartimento di Biologia Evolutiva, Universita` di Siena, Via A. Moro, 53100 Siena, Italy
5
Rottapharm SpA, via Valosa di Sopra 9, 20052 Monza, Italy
6
Dipartimento di Scienze Farmacobiologiche, Universita` ‘‘Magna Græcia’’ di Catanzaro, Complesso ‘‘Ninı` Barbieri’’, 88021 Roccelletta di Borgia, Catanzaro
Received 8 September 2005; revised 17 May 2006; accepted 18 May 2006 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20705
ABSTRACT: The physicochemical and biopharmaceutical properties, such as pKa, crystal habit, water solubility, logD, molecular structure and dynamics, and membrane permeability of CR3124 (endo-2-(8-methyl-8-azabicyclo[3.2.1]oct-3-yl)-2,3-dihydro-1Hbenz[e]isoindol-1-one, a novel potent 5-HT3 receptor antagonist) have been studied in order to obtain preformulation information. The study showed that CR3124 is a very rigid molecule in which conformational freedom due to the presence of a rotatable bond is restricted by the interaction between an activated hydrogen and the amide oxygen and the conformation of the tropane piperidine ring is regulated by the environment in such a manner as to optimize the intermolecular interactions with the solvent. This chameleon behavior appears to be capable of explaining the biopharmaceutical properties showed by CR3124, such as low wettability, relatively good solubility, and very high membrane permeability. ß 2006 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 95:2706–2721, 2006
Keywords: preformulation; solubility; permeability; crystal structure; NMR spectroscopy; serotonin; 5-HT3 receptor antagonist; molecular dynamics
INTRODUCTION Serotonin (5-hydroxytryptamine, 5-HT) is a neurotransmitter which plays an important role in Correspondence to: Andrea Cappelli (Telephone: þ39 0577 234320; Fax: þ39 0577 234333; E-mail:
[email protected]) Journal of Pharmaceutical Sciences, Vol. 95, 2706–2721 (2006) ß 2006 Wiley-Liss, Inc. and the American Pharmacists Association
2706
physiological functions and in many pathological conditions. Among the great variety of serotonin receptor subtypes, the 5-HT3 receptor is the only one belonging to the ligand-gated ion channel superfamily.1 The development of selective 5-HT3 receptor antagonists has received great attention in recent years and the therapeutic role of ondansetron (GR32634, 1, Fig. 1), granisetron (BRL 43694, 2), and tropisetron (ICS205-930, 3)
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
2707
the 5-HT3 receptor-dependent Bezold–Jarisch reflex in urethane-anesthetized rats. Moreover, 4 prevented scopolamine-induced amnesia in the mouse passive avoidance test, suggesting a potential usefulness in cognitive disorders.4 On the basis of these results, the biopharmaceutical properties of this potent 5-HT3 receptor antagonist have been studied in order to obtain useful preformulation data for the design and the development of appropriate delivery systems. In this study we describe the results of a study focused on the physicochemical and biopharmaceutical properties of compound 4 such as pKa, crystal habit, water solubility, log D, and membrane permeability. Some of these properties are discussed from the viewpoint of the molecular features of compound 4.5
Figure 1. 1–4.
Structure of 5-HT3 receptor antagonists
as antiemetic drugs has been clearly established. These drugs revolutionized anticancer chemotherapy, since they allow the use of cytotoxic treatment by blocking the nausea and vomiting stimulated by anticancer drugs and/or radiotherapy. They are also useful in the prevention of postoperative nausea and vomiting due to the anesthetics used in surgical procedures. The 5-HT3 receptor antagonists, alone or in combination with other antiemetic drugs, have become the agents of choice in controlling emesis because of the higher effectiveness and relatively lower adverse effect profile with respect to the usual antiemetic agents. The clinical efficacy as antiemetic agents and the safety profile of the various agents in this class are similar. The main site of action of these drugs appears to be the central 5-HT3 receptors, although inhibition of peripheral receptors may also play a role in the control of vomiting. Finally, the 5-HT3 receptor antagonists may also be useful in the treatment of pain, pruritus, fibromyalgia, gastrointestinal symptoms (e.g., irritable bowel syndrome), anxiety disorders, and alcohol dependency, but not enough clinical data are available to confirm their role in these disorders.2,3 CR3124 (endo-2-(8-methyl-8-azabicyclo[3.2.1] oct-3-yl)-2,3-dihydro-1H-benz[e]isoindol-1-one, 4) is a new chemical entity showing nanomolar affinity for the 5-HT3 receptor and potent antagonist properties on both 5-HT3 receptor-dependent [14C]guanidinium uptake in NG 108-15 cells and DOI 10.1002/jps
EXPERIMENTAL SECTION Materials Compound 4 was synthesized (gram scale) in our laboratories following a previously described procedure.4 Compound 4 was purified by recrystallization from ethyl acetate and used without further treatment (4-recrystallized: mp 174.5– 1758C, lit.4 174–1758C) or finely ground by means of a pestle and a mortar (4-crushed). The hydrochloride salt of 4 (4 HCl) was prepared by dissolving 0.31 g (1.0 mmol) of the free base in methanol (10 mL) and then adding 1.0 mL of concentrated HCl. The volatiles were removed under reduced pressure and the residue was dried and recrystallized from CHCl3 by slow evaporation to obtain X-ray quality crystals (0.33 g, yield 96%, mp > 3008C). 1H-NMR (200 MHz, CDCl3): 2.32–3.03 (m, 11H), 3.79 (m, 2H), 4.55 (br s, 3H), 7.46–7.67 (m, 3H), 7.91 (d, J ¼ 7.8, 1H), 8.00 (d, J ¼ 8.2, 1H), 9.13 (d, J ¼ 8.1, 1H), 12.24 (br s, 1H). Ethanol 96% v/v was purchased from Panreac (Barcelona, Spain). The other chemicals were purchased from Aldrich (Milwaukee, USA) and used without further purification. Buffers were prepared according to the European Pharmacopoeia6 and the pH values were measured by means of an Amel Model 337 pH-meter. Solution Properties Potentiometric and solution calorimetric measurements were carried out at 258C in aqueous media (0.15 M NaCl) following a previously
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2708
CAPPELLI ET AL.
described procedure.7 A TitraLab 90 titration system (from Radiometer Analytical, Lyon, France) was used for determination of the basicity constant by recording the potentiometric titration curves (mV vs. mL of titrant). The TitraLab 90 consists of three components: the TIM900, a powerful Titration Manager; the ABU901, a high-precision autoburet; and the SAM7, a convenient sample stand. Windows-based software (TimTalk9) was used to control the TIM900 Titration Manager. At least two titrations of compound 4, which is very soluble in the acidic region, were performed at different concentrations. Weighed amounts of solid 4 (45.7 mg and 30.4 mg) were dissolved in ca. 100 mL of 0.15 M NaCl containing a measured volume of standard HCl solution (0.1 M), under presaturated nitrogen. Titrations with standard 0.1 M NaOH showed a phase separation at pH ca. 8.6–8.9, that is, beyond the sharp inflection point. E0 calibration was performed before and after each titration. The concentration of hydrogen ions was calculated from electromotive force (emf) values (mV) by means of the formula [Hþ] ¼ exp(EE0)/ 25.693. The basicity constant, evaluated in the narrow pH range where solubility was maintained, was obtained by the Superquad program.8 Calorimetric titrations were performed under the same experimental conditions as the potentiometric ones using a Tronac solution calorimeter (model 1250). An aqueous solution (25 mL of 0.15 M NaCl), containing a weighed amount of 4 (33.1 mg) and a measured volume of standard 0.1 M HCl, was titrated with standard 0.1 M NaOH solution by means of a Gilmond buret at a constant rate of 0.0837 mL/min. The isothermal titration was automatically controlled by the renewed Thermal program previously reported.7 Calibration of the instrument and corrections of the titrant heats of dilution were made before the titration run. The enthalpy change, DH8 ¼ 39.3(8) KJ/mol, was computed with the Fith program.9
A suitable amount of compound 4 was dissolved in 10 mL of solvent or solvent mixture; the resulting mixture was heated, when necessary, until dissolution was complete. The solution was allowed to evaporate at room temperature and crystals were collected and submitted to X-ray diffraction studies. Single crystals of compound 4 (free base) and 4 HCl (hydrochloride) were used for X-ray data collection on a Siemens P4 four-circle diffractometer with graphite monochromated Mo-Ka ˚ ). The o/2y scan technique radiation (l ¼ 0.71073 A was used. The structures were solved by direct methods and refinement was carried out by fullmatrix anisotropic least-squares of F2 against all reflections. The hydrogen atoms were located on Fourier difference maps and included in the structure-factor calculations with a common isotropic temperature factor. Atomic scattering factors including f0 and f00 were taken from ref.14 Structure solution and refinement were carried out with the SHELX-97 package14 while molecular graphics were performed by the WinGX program.15 Compound 4: monoclinic; space group P21/c (n. ˚, 14); a ¼ 10.785(2), b ¼ 13.315(2), c ¼ 23.045(3) A 3 ˚ b ¼ 98.18(1), V ¼ 3275.6(9 )A , Z ¼ 8, Dc ¼ 1.24 g/ cm3. A total of 5758 unique reflections (Rint ¼ 0.027) was collected at 228C. The final refinement converged to R ¼ 0.056 and wR2 ¼ 0.099 for F2 > 2s(I). Minimum and maximum heights in the last Dr map were 0.19 and ˚ 3.4,5 0.17/A Full crystal data for 4 HCl are reported in Table 1. CCDC-272870 contains the supplementary crystallographic data for this study. These data can be obtained free of charge via www. ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (þ44) 1223-336-033; or e-mail: deposit@ccdc. cam.ac.uk). Differential Scanning Calorimetry
X-Ray Crystallography Recrystallization of compound 4 (free base) from different solvents was used for polymorph screening.10 – 13 The recrystallization solvents included ethanol– water (10:1), ethanol, ethyl acetate, diethyl ether, chloroform, n-hexane-dichloromethane (5:2), heptane-dichloromethane (5:2), methanol, carbon tetrachloride.
Differential scanning calorimetric (DSC) analysis was performed with a Perkin–Elmer DSC-7 instrument and PYRIS software (version5). Temperature axis and cell constant were calibrated with indium. A weighed sample (3.412 mg) of 4 in a pin-holed aluminium pan was heated at 108C/min over a temperature range of 100–3608C under nitrogen purge (20 mL/ min).16,17
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
described by Li et al.21 This method was selected because of wettability problems of the compound. The solubility was determined with the drug powder placed in a plexiglass donor cell separated from a thermostatted (22 0.58C) glass acceptor cell (SC-100B, nominal volume 3 mL) by a previously conditioned cellophane membrane (Medicell, London, UK, MWCO 12–14 kDa), as depicted in Figure 2. The donor side was loaded with 4-crushed in the absence of compression in order to favor formation of the saturated solution. Successively, 250 mL and 3 mL of water were added to the donor and acceptor compartments, respectively. The system was held together by a special holder and stirred magnetically (200 rpm). The concentration was determined by HPLC (Agilent 1100 LC/MSD; column: Agilent Zorbax 5 mm Eclipse XDB-C8, 4.6 mm 150 mm; mobile phase: methanol–water 80:20; UV detection at 223 nm). Equilibration was achieved in about 10 days.
Table 1. Crystal Data for 4 HCl Formula M Crystal size (mm) Crystal system Space group ˚ a/A ˚ b/A ˚ c/A b/8 ˚3 V/A Temperature/K Z F(000) Dc/gcm3 m(Mo-Ka)/mm Scan mode Scan range/8 Scan width/8 Scan speed/8/min Independent reflections Obs. reflections (I > 2s(I)) N. parameters refined R1 (I > 2s(I)) wR2 (I > 2s(I))
C20H22N2O HCl 342.85 0.5 0.4 0.2 Monoclinic P21/c (n. 14) 14.608(3) 7.1330(10) 17.429(3) 106.620(10) 1740.2(5) 293(2) 4 728 1.309 0.228 o/2y 2.4 y 25.0 1.16 3 3065 1128 244 0.0746 0.0968
Octanol/Water Partition Coefficient Determination
Scanning Electron Microscopy (SEM) Studies The material was mounted on aluminum holders by carbon conductive glue and coated with 20 nm gold in a Balzer’s MED 010 sputtering device. The samples were observed with a Philips XL20 scanning electron microscope operating at an accelerating voltage of 20 kV. Solubility Measurements An excess amount18–20 of compound 4 was added to vials containing 10 mL (or 50 mL) of water, phosphate buffer solution (pH 7.4), or ethanol at different dilutions (99.9, 96, 90, 80, 70% v/v). The samples were stirred at room temperature (22 0.58C) for a suitable time in order to achieve equilibrium. Furthermore, the solutions at decreasing EtOH concentration (60, 50, 40, 30, 25, 20, 16, 10, 5, 3.3, 1%) were prepared from the saturated EtOH (99.9%) solution of 4. Under these conditions, a white precipitate was observed. The samples were filtered with a 0.45 mm nylon membrane filter (Whatman, Maidstone, UK) and the absorbance of the filtrate measured by UV spectrophotometry (Perkin–Elmer Lambda 3B; lmax: 223 nm). All solubility determinations were performed in triplicate (CV% < 5). Aqueous solubility of 4-crushed (particle size 5– 20 mm) was determined by the dialysis method DOI 10.1002/jps
2709
The octanol/water partition coefficient of compound 4 was determined at various pH values by the shake-flask method.22,23 Octanol and aqueous buffer (pH values: 2.0, 3.0, 5.5, 6.0, 6.8, 7.4, 8.0, 9.0, 10.0, 11.0), prepared according to European Pharmacopoeia 5th Edition, were mutually saturated for 4 h and the phases were then separated. A solution (20 mL) of 4 in octanol (saturated with aqueous solution) was placed in a flask containing 20 mL of the saturated aqueous solution. The two phases were shaken (manually) for 30 min. After being allowed to stand for 24 h at room temperature (22.0 0.58C), the aqueous phase was run off (roctanol ¼ 0.8258/mL), centrifuged, and the concentration of 4 in the aqueous solution was determined by the above described HPLC method. Three replicates were performed for each pH value. The D-value corresponds to the quotient between aqueous and octanol concentration of the drug according Eq. 1.10 D¼
½Co Cw Cw
ð1Þ
where Co is the initial octanol concentration and Cw is the final aqueous concentration. In Vitro Intestinal Permeability Experiments The permeability studies were performed with Caco-2 monolayers as described in the literature.24
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2710
CAPPELLI ET AL.
Figure 2. Dialysis system adopted for the determination of aqueous solubility. (A) Acceptor compartment; (B) sampling port; (C) dialysis membrane; (D) Donor compartment; (E) magnetic bar location; (F) section plane; (G) section view.
Caco-2 cells were cultured in supplemented Dulbecco’s modified Eagle medium with 10% fetal bovine serum and seeded onto polycarbonate membranes for test compound transport experiments. Caco-2 cell membranes were grown by seeding on Snapwell supports incubated at 378C with 5% CO2/95% O2 and approximately 95% humidity for 15–21 days. Drug permeability experiments were performed at 378C at a final drug concentration of 10–100 mM in HBSS buffer (1% DMSO) in the apical chamber (donor side in the apical to basolateral (a–b) permeability study) or in the basolateral side in b–a permeability study. This concentration did not show any cytotoxic effect on Caco-2 cells. Samples were obtained from both chambers at regular time intervals and snap frozen on dry ice/methanol. Drug concentrations were determined by HPLCMS. [14C]Mannitol permeability was used to assess monolayer integrity and propanolol and vinblastine were evaluated as controls. Caco-2 apparent permeability values (Papp) were calculated by means of the following equation: Papp ¼
DQ 1 Dt AC0
ð2Þ
where DQ/Dt was the rate of appearance of the drug in the receiver chamber, C0 was the initial concentration of the drug in the donor chamber, and A was the surface area of the monolayer. All Papp values were standardized and reported as 106 cm/s. NMR Spectroscopy NMR experiments were performed with a Bruker DRX-600 AVANCE spectrometer, equipped with a xyz gradient unit, operating at 600.13 MHz for 1H
and with a Bruker AC200. Nuclear Overhauser enhancement spectroscopy (NOESY)25 and double quantum filtered Correlation Spectroscopy (dqfCOSY)26 spectra were acquired with 2048 complex points for 256 experiments with 6 s recycle, and a time proportional phase incrementation (TPPI) phase cycle.27 A 908 shifted squared-sine window function was used in both dimensions for every set of data and zero filling to 1024 points was applied along F1. For NOESY spectra, mixing times of 200 ms, 400 ms, and 800 ms were used. NMR data were processed with NMRpipe28 (version 3.3), and 2D spectra were analyzed with SPARKY.29
RESULTS Solution Properties The thermodynamic functions (pKa, DG0, DH0, and DS0) for protonation of the tropane nitrogen of 4 were determined in aqueous solution (0.15 M NaCl) at 258C. Titrations with standard 0.1 M NaOH showed phase separation at pH ca. 8.6–8.9 (i.e., beyond the sharp inflection point) and revealed a limited solubility of compound 4 and a pKa value of 9.11 (Tab. 2). Comparison of the measured pKa value for compound 4 with the corresponding values described for atropine (9.66)30 suggested that tropane nitrogen atom of 4 is slightly less available for protonation with respect to that of atropine, probably because of the differences in the conformational preferences of the tropane moiety (see below). This hypothesis is supported by the thermodynamic functions described in the literature for N-methylpiperidine and N-methylpyrrolidine (structural models of tropane moiety).31
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
2711
Table 2. The Thermodynamic Functions for the Protonation of the Tropane Nitrogen of Compound 4, Atropine, N-Methylpiperidine, and N-Methylpyrrolidine Compound 4 Atropine N-methylpiperidine N-methylpyrrolidine a
pKa
DG0 (kJ/mol)
DH0 (kJ/mol)
DS0 (J/mol/K)
9.11(7)a 9.66 10.08 10.33 10.46
51.9(3)a 55.3 57.4 59.1 59.7
39.3(8)a
42(3)a
39.5 41.5 37.9
60.0 59.0 73.2
Literature Bergstrom et al. 2003 (Ref 30) Cabani et al. 1971 (Ref. 31b) Blais et al. 1974 (Ref. 31a) Cabani et al. 1971 (Ref. 31b)
Values in parentheses are standard deviations of the last significant figure.
In fact, the entropic factor (DS8) relevant to the protonation of 4, which is related to the availability of the proton acceptor site, was the lowest among the considered compounds; on the other hand, DH8 values did not reflect the trend of pKa values.
Crystal Structure of 4 Hydrochloride The crystal structure of 4 HCl is depicted in Figure 3. The benzisoindolone moiety is close to planarity with N(2) and C(7) showing the largest deviations on the same side of the ring (0.030(5) ˚ , respectively). The dihedral angle and 0.029(6) A between the least-squares planes defined by the benzisoindolone and by the piperidine is equal to 72.6(2) in 4 HCl and it is close to the corresponding values of 79.268 and 83.668 found in the two molecules of the asymmetric unit of 4.4,5 The isoindolone is bound to the piperidine ring of the tropane at the bisectional position. The orientation of the bond N(2)C(30 ) causes the hydrogen at carbon C(30 ) to be in a cis position with respect to the pyrrolidone carbonyl. This allows the existence of a strong intramolecular interaction H(30 ). . .O(1) characterized by a distance equal to ˚ , which is quite close to the corresponding 2.403 A
˚ found in the two values of 2.410 and 2.399 A molecules of 4. This nonclassical hydrogen bonding interaction is also present in analogous benzamide derivatives.32,33 This interaction, together with the transannular steric repulsion of the tropane ethylene bridge with the pyrrolidone moiety and the presence of the additional hydrogen atom on the tropane nitrogen, is the driving force that causes the piperidine ring to adopt a half-chair conformation in 4 HCl, while it has a boat conformation in the crystal structure of 4 (Tab. 3). C(10) is bound to N(8) in the equatorial position. There is a strong interaction between the chloride ion and the hydrogen of the protonated ˚. N(80 ) with a H(80 ). . .Cl distance equal to 1.91(5) A Finally, the crystal packing is stabilized by van der Waals interactions.
Polymorph Screening In order to find possible polymorph species, we crystallized 4 from different solvents or mixtures of them (ethanol–water (10:1), ethanol, ethyl acetate, diethyl ether, chloroform, n-hexane– dichloromethane (5:2), heptane–dichloromethane (5:2), methanol, carbon tetrachloride). In all
Figure 3. Stereoscopic view of the crystal structure of 4 HCl. Ellipsoids enclose 50% probability. DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2712
CAPPELLI ET AL.
Table 3. Torsion and Puckering Parameters for 4 and 4 HCla 4b (mol. 1)
4b (mol. 2)
4 HCl
C(3 )-C(4 )-C(5 )-N(8 ) 19.01(1) N(80 )-C(10 )-C(20 )-C(30 ) 14.08(2) N(2)-C(30 )-C(40 )-C(50 ) 167.43(1) C(1)-N(2)-C(30 ).(N80 ) 158.61 130.76 C(1)-N(2)-C(30 )-C(20 ) C(20 )-C(10 )-N(80 )-C(10) 166.09 Ring C(10 )-C(20 )-C(30 )-C(40 )-C(50 )-N(80 ) ˚) Compound q2 (A 4 mol.1c 0.834 4 mol.2c 0.855 4 HCl 0.422
11.41(1) 16.76(1) 174.85(1) 7.81 127.51 167.07
52.29 49.87 107.38 154.30 140.23 159.74
˚) q3 (A 0.160 0.135 0.516
Y (8) 79.2 81.1 140.7
Torsion Angle (8) 0
0
0
0
F (8) 357.56 357.86 177.58
˚) Puckering amplitude (A 0.849 0.866 0.666
a
Puckering parameters according to reference 34. See references 4 and 5. c See reference 5. b
cases, we obtained crystals having the same color, shape, and dimensions, which suggest the same molecular arrangement in the crystal. To verify this preliminary observation, one or more crystals from each crystallization batch were submitted to X-ray diffraction for cell determination. About 20 different crystals were tested and all of them showed the same crystal cell and metric parameters. Differential Scanning Calorimetry Samples of 4 recrystallized from ethyl acetate were characterized by DSC over a temperature range of 100–3608C at a heating rate of 108C for the identification of potential solidstate changes. DSC analysis revealed a single endothermic fusion peak at 1758C (DH ¼ 95.1 J/g, 29.1 kJ/mol) thus confirming the absence of solidstate alteration in the observed temperature range (Fig. 4). SEM Studies Two samples of compound 4 (4-recrystallized and 4-crushed) were observed by scanning electron microscopy (SEM) after sputter coating with gold. The obtained image (Fig. 5) shows that the 4recrystallized sample appears to be formed by crystals showing similar shape, but different dimensions. However, most of the material is included in prisms larger than 20 mm. The treatment of the 4-recrystallized sample with a pestle-mortar system for 30 min was carried out in the attempt to simulate industrial grinding.
Figure 5 shows that this treatment produces both an homogenization and a considerable decrease in the particle size, which are now below 20 mm. As expected, the particles of 4 showing the smallest dimensions tend to form aggregates. Solubility Measurements Initially, the concentration data obtained at 24 h by means of shake-flask measurements starting from different amounts of 4-recrystallized or 4crushed showed significant variability. This trend of dissolution can be explained on the basis of wettability problems. In fact, in the case of small scale mortar and pestle comminution of 4 samples we obtain a decrease of the concentration in spite of an increase in the specific surface area (see Fig. 5). This is a typical phenomenon for a poorly wettable substance.35 The determination of a reliable water solubility value (0.16 0.05 mg/ mL) was, therefore, obtained by performing a large number of dissolution kinetic experiments at a longer time (1 week). Furthermore, when the phosphate buffer solution (pH 7.4) was used in the place of water, the solubility studies gave more precise results (2.25 0.10 mg/mL), from which the S0 value of 0.043 mg/mL (S0 at 228C ¼ 0.00014 M ) could be calculated. The ethanol solubility data of compound 4 (Fig. 6) show that the solubility of this compound reaches a maximum in 90% ethanol, while it tends to decrease at higher ethanol concentrations. This behavior appears to be similar to that shown by some other drug molecules (e.g., rofecoxib)36 and
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
2713
Figure 4. DSC curve of 4-recrystallized.
could be explained taking into consideration the complex interactions among dielectric constants, hydrogen bonds, and the solute–solvent system. Low ethanol concentrations produced a minimum from which the water solubility value 0.30 mg/mL could be extrapolated. Owing to the difficulty in the determination of water solubility by the traditional shake-flask methods, dialysis was used. A cellophane membrane was selected because of its greater permeability with respect to the system under investigation. The kinetic profile represented in Figure 7 shows a rapid passage of compound 4 into the acceptor compartment during the first 24 h, followed by a decrease in rate until the plateau was reached in about 10 days. In these conditions the equilibrium concentration, representing the solubility of the compound, was equal to ca. 0.15 mg/mL, which appears to be in agreement with the corresponding one obtained by the shake-flask solubility measurements. Apparent Octanol/Water Partition Coefficient Determination The logarithm of the apparent octanol/water partition coefficient (log D, Fig. 8) at the pH values tested ranges from 2 (pH 11) to 1 (pH 2) DOI 10.1002/jps
with a maximum of 2.42 at a pH value of 9, which corresponds to the pKa value determined by titration methods (pKa ¼ 9.11, see above).37 Interestingly, the log P value calculated by means of CS Chem Prop (CS Chem Office Pro 4.5) was in good agreement with the maximum experimental value. The software calculated a log P value of 2.33 0.47 following Crippen’s fragmentation38 and a log P value of 2.55 0.49 following Viswanadhan’s fragmentation.39 In Vitro Intestinal Permeability Experiments In the permeability studies performed with Caco2 monolayers, compound 4 showed a Papp value of 41.23 106 cm/s (a–b experiment). For comparison, propranolol showed a Papp value of 30.5 106 cm/s and vinblastine a Papp value of 0.08 106 cm/s in the same test. As propranolol is classified as a highly permeable model drug in the FDA permeability classification,40 compound 4 must be similarly considered a highly permeable chemical entity. In basolateral to apical (b–a) permeability experiments, compound 4 exhibited a Papp value of 36 106 cm/s; in comparison DLpropanolol showed a Papp value of 40.7 106 cm/ s and vinblastine (a P-gp substrate) a Papp value of 92.0 106 cm/s.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2714
CAPPELLI ET AL.
Figure 7. Water solubility of compound 4 as determined by means of dialysis.
Figure 5. SEM micrograph of two different samples of compound 4. (Top) 4-recrystallized, (bottom) 4crushed.
NMR Spectroscopy Studies The 1H-NMR spectra of 4 HCl obtained in different solvents, such as CDCl3, DMSO-d6, and D2O showed the presence of two species which
Figure 6. Solubility of 4 at various alcohol dilutions.
differ by the NCH3 (equatorial or axial) stereochemistry.41 The coexistence of these two species may have important implications from the point of view of both the interaction with the receptor and the hydrophilic surface of 4. A complete 1HNMR assignment of 1H-NMR signals of the major isomer of 4 HCl was obtained by the analysis of 2D-COSY and NOESY spectra obtained at 600 MHz (Tab. 4). Comparison of the NOESY connections found in the spectra of 4 HCl performed in the different solvents showed that orientation of the NCH3 group of 4 HCl major isomer is equatorial in all the three solvents used (see the cross-peaks between CH3 and 60 /70 d1 protons) and, generally, that the solvent properties do not appear to affect the conformational preferences of 4 HCl. This result is confirmed by the shape of the diagnostic signal attributable to the 30 proton, which appears as an invariable septet-like multiplet in the spectra obtained in the three different solvents. On the other hand, the signal attributable to the 30
Figure 8. Logarithm of the apparent octanol/water partition coefficient (log D) of 4 at various pH values.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
DOI 10.1002/jps
—
2.54 1.62 4.67 2.21 1.66 2.27 4.43 7.51 7.99 7.92 7.57 7.66 9.27
3.30
CDC3
0
0
5 4 7 6. 8 7. 9 8
20 /40 pa, 10 /50 , 30 20 /40 pe, 30 20 /40 pe, 20 /40 pa 60 /70 d2, 10 /50 60 /70 d1
0
2 /4 pe, 6 /7 d1
0
COSY Cross-Peaks 0
0
0
2 /4 pe, 2 /4 pa, 60 /70 d1, CH3 20 /40 pa, 10 /50 , 30 20 /40 pe, 10 /50 , 3 20 /40 pe 60 /70 d2, 10 /50 60 /70 d1 10 /50 20 /40 pa, 4 5. 3 4 7 6. 8 7. 9 8
0
NOESY Cross-Peaks
3.10 2.94 4.57 2.35 2.55 2.78 4.59 7.54 8.06 7.96 7.61 7.69 9.18 broad
3.84 2.70 2.40 4.56 2.29 2.13 2.66 4.71 7.73 8.19 8.09 7.63 7.70 9.08 10.48
3.91
0
0
0
0
COSY Cross-Peaks b
3.84 2 /4 pe, 2 /4 pa, 60 /70 d1, 60 /7’d2b 2.51 20 /40 pa, 10 /50 , 30 2.02 20 /40 pe, 10 /50 ,b 4.18 20 /40 pe, 20 /40 pa 2.29 60 /70 d2, 10 /50 1.92 60 /70 d1, 10 /50 b 2.73 3.91 6.91 5 7.49 4 7.60 7 7.36 6 7.38 9 8.28 8 — CH3
CDCl3 DMSO-d6 D2O
4-HCld (ppm)
H-NMR (600 MHz) Characteristics of 4 and of the Major Isomer of 4 HCla
4 d (ppm)
1
0
0
0
CDCl3 0
0
2 /4 pe, 2 /4 pa, 6 /7 d1, 60 /70 d2, CH3 20 /40 pa, 10 /50 , 30 20 /40 pe, 10 /50 , 30 , 3, 60 /70 d2 20 /40 pe, 20 /40 pa, 60 /70 d2, 10 /50 , CH3 3, 20 /40 pa, 60 /70 d1, 10 /50 10 /50 , 60 /70 d1 60 /70 d2, 20 /40 pa, 4 5, 3 4 7 6 9 8
0
0
0
0
DMSO-d6 0
0
2 /4 pe, 2 /4 pa, 6 /7 d1, 60 /70 d2, CH3 20 /40 pa, 10 /50 , 30 , 60 /70 d2 20 /40 pe, 10 /50 , 30 , 3, 60 /70 d2 20 /40 pe, 20 /40 pa 60 /70 d2, 10 /50 , CH3 3, 20 /40 pa, 60 /70 d1, 10 /50 10 /50 , 60 /70 d1, NH 60 /70 d2, 20 /40 pa, 4 5, 3 4 7 6 9 8 CH3
0
NOESY Cross-Peaks
0
D2O 0
2 /4 pe, 2 /40 pa, 60 /70 d1, 60 /70 d2, CH3 20 /40 pa, 10 /50 , 30 20 /40 pe, 10 /50 , 30 , 3, 60 /70 d2 20 /40 pe, 20 /40 pa 60 /70 d2, 10 /50 , CH3 3, 20 /40 pa, 60 /70 d1, 10 /50 10 /50 , 60 /70 d1 60 /70 d2, 20 /40 pa 5 4 7 6 9 8
0
The H-NMR spectra of 4-HCl showed the presence of two species which differ by the stereochemistry of NCH3. The 1H-NMR characteristics of the minor isomer is omitted for clarity. b These cross-peaks were detected only in the COSY spectra performed in DMSO-d6.
1
2 2 1 2 2 3 2 1 1 1 1 1 1 1
20 /40 pe 20 /40 pa 30 60 /70 d1 60 /70 d2 CH3 3 4 5 6 7 8 9 NH
a
2
1 /5
No.
0
0
H
Table 4.
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
2715
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2716
CAPPELLI ET AL.
proton of the minor isomer shows itself as an apparent triplet at 4.75 ppm (in CDCl3) or 4.27 ppm (in D2O) from which the J-value of about 10 Hz can be extracted, as observed in the case of the dimethyl quaternary derivative of 4.5This result suggests that when the methyl group occupies the axial position, the tropane moiety reaches maximum rigidity. When compared to compound 4 its hydrochloride salt (4 HCl) showed differences in the monodimensional 1H-NMR signals attributable to the tropane moiety and in COSY and NOESY connections. The most important difference between NOESY spectra of 4 and 4HCl is in the contacts of the methylene C-3 protons. The NOESY spectrum of 4 showed a cross-peak between methylene C-3 protons and the higher field couple of H-20 /H-40 (20 /40 pa protons), while that of the hydrochloride salt in the three different solvents showed that the C-3 protons were involved in dipolar coupling with 20 /40 pa and with protons 60 /70 d2 also. The latter contact is particularly important because it suggests that the C-3 and 60 /70 d2 protons, even located at different portions of the molecule, are in close proximity in the space because of conformational restraints. Compound 4 shows in principle three types of conformational freedom: (a) rotation around the single bond attachment of the tropane moiety to the fused pyrrolidone nucleus, (b) conformational
flexibility of the piperidine ring of the tropane moiety, which can adopt a conformation varying from the boat typical of the crystal structure of the free base (conformation 4B, Fig. 9) to the flattened chair (conformation 4FC) (the piperidine ring of tropane moiety of 4 cannot adopt the normal chair conformation because of the transannular steric interactions of the tropane ethylene bridge hydrogens with the pyrrolidone moiety), (c) pyramidal inversion of the tertiary amine nitrogen. In a previous study, we reported that the comparison between the experimental J-values and those calculated for three different conformations of the tropane moiety suggested that in solution the tropane piperidine ring of 4 adopts a conformation intermediate (conformation 4FB) between the boat (conformation 4B) and the flattened chair (conformation 4FC), or possibly it shows a rapid inversion between these two limit forms. Moreover, the cross-peak between the methylene C-3 and 20 /40 pa protons in the NOESY spectrum of 4 supported the existence, also in solution, of a conformational preference at the level of the rotation around the single bond attachment of the tropane moiety to the fused pyrrolidone nucleus, which is also shown by crystallography.5 On the other hand, X-ray crystallographic studies showed a flattened chair conformation of the tropane piperidine ring of 4-HCl. Indeed, in
Figure 9. Three conformations of compound 4. [Color figure can be seen in the online version of this article, available on the website, www.interscience.wiley.com.] JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
Table 5. Interprotonic Distances in 4 HCl
Protons
NOESY Distances in Cross-Peak NOE Solidstate Volume Distancea Conformationb
20 /40 pa-20 /40 pe 3-d2 3-20 /40 pa 30 -20 /40 pe 10 /50 -d1 10 /50 -d2
45.3 1.34 2.82 9.44 4.20 1.49
1.56 2.80 2.48 2.03 2.30 2.76
1.563 2.389/2.457 2.245/3.132 2.129/2.117 2.240/2.236 2.571/2.582
2717
tial usefulness in cognitive disorders.4 On the basis of these findings, the physicochemical and biopharmaceutical properties (such as pKa, crystal habit, water solubility, log D, and membrane permeability) of this compound have been studied in order to obtain preformulation information useful for the design and the development of suitable delivery systems. About 20 different crystals of 4, obtained from different solvents or solvent mixtures, were studied by single crystal X-ray diffraction and all of them showed the same crystal cell and metric parameters. Moreover, DSC analysis revealed a single endothermic fusion peak and strongly supported the absence of solidstate structural alteration in the observed temperature range. These results revealed the tendency of compound
a Interprotonic distances determined by means of ISPA approach starting from the geminal 20 /40 pa-20 /40 pe distance in the solidstate conformation. b Interprotonic distances measured in the crystallographic structure.
the hydrochloride salt, the additional hydrogen on the tropane nitrogen atom constrains the piperidine ring in the flattened chair conformation. The static structure typical of the solidstate was found to persist also in an organic solvent solution. 1HNMR studies performed in CDCl3 showed good agreement between the interprotonic distances measured in the crystallographic structure and those determined by means of Isolated Spin Pair Approximation (ISPA)42 approach from the volume of the cross-peaks obtained by means of NOESY experiments (Tab. 5). These results suggest that 4 hydrochloride in solution is characterized by a high degree of rigidity.
DISCUSSION Compound 4 is a potent 5-HT3 receptor antagonist both in vitro (in the 5-HT3 receptor-dependent [14C]guanidinium uptake in NG 108-15 cells) and in vivo (in the 5-HT3 receptor-dependent Bezold– Jarisch reflex in urethane-anesthetized rats) and prevented scopolamine-induced amnesia in the mouse passive avoidance test, suggesting potenDOI 10.1002/jps
Figure 10. Electrostatic potential (ESP) of 4 in the crystallographic boat conformation (4B, top) and the flattened chair conformation (4FC, bottom) mapped on the respective total electron density isosurface of 0.002 units (electrons/bohr3). The ESP values range from 0.08 (red) to 0.08 (blue) units. Green color denotes neutral isosurface (potential). Density functional calculations were performed at B3LYP/6-31G** level of theory using Gaussian03 package. [Color figure can be seen in the online version of this article, available on the website, www.interscience.wiley.com.]
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2718
CAPPELLI ET AL.
Figure 11. ESP of protonated 4 in the crystallographic conformation mapped on the respective total electron density isosurface of 0.002 units (electrons/ bohr3). The ESP values range from 0.08 (red) to 0.2 (blue) units. Green color denotes neutral isosurface (potential). Density functional calculations were performed at B3LYP/6-31G** level of theory using Gaussian03 package. [Color figure can be seen in the online version of this article, available on the website, www. interscience.wiley.com.]
4 to crystallize easily in a stable form. However, other studies on the effects of super-saturation are required in order to definitely rule out polymorphism problems for this compound. SEM studies revealed that 4-recrystallized samples appeared to be constituted by crystals showing similar shapes, but different dimensions and that most of the material was included in prisms larger than 20 mm. The particles of 4 showing the smallest dimensions tended to form aggregates. Among the techniques to increase the aqueous solubility of poor-soluble compounds, particle size reduction is one of primary choice.35 Marketed formulations of very slightly soluble substances show that particle size reduction and, hence, increased specific surface area are important means to achieve a satisfactory dissolution rate. However, for many compounds, decreasing the particle size may not lead to a significant or adequate enhancement in bioavailability. Moreover, the common way to reduce the particle size may result in particles with a broad size distribution, decreased crystallinity, poor flow properties, and static charges generation. Furthermore, physical instability of the size distribution is introduced when the size range is reduced to micrometer and submicrometer scale. The water solubility of 4 can be defined in the range from 0.15 mg/mL to 0.30 mg/mL, depending on the method used for the determination because
Figure 12. Simultaneous equilibria involving the different species/conformers of 4 in the transfer from the aqueous environment to the lipophilic one. Each equilibrium is affected by the properties of the single species/conformer with respect to those of the environment. For example, 4FCAH should be the preferred species in the aqueous environment at the physiological pH, while the rapidly interconverting 4B and 4FC are likely to be the major species in the lipophilic environment.
of wettability problems, while the solubility at pH 7.4 was 2.25 mg/mL (S0 at 22 8C ¼ 0.00014 M; log S0 ¼ 3.85 M). The projected dose of the oral administration dosage form of 4 (ca 5 mg each tablet or capsule) is in the same range as tropisetron (5 mg each capsule) and granisetron (1 or 2 mg each tablet). Thus, the solubility of 4 free base in 250 mL of aqueous media can be calculated
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
2719
Table 6. Biopharmaceutical Characteristics of Some Established Drugs and Compound 4a
Compound 4 (CR3124) Primaquine Atropine Desipramine
pKa
log P
9.11 2.42b 9.99, 3.74 2.72 9.66 1.64 10.08 3.79
log S0 (M)
S Class
3.85 2.77 1.61 3.81
h h h h
Caco-2 (a–b) (106 cm/s) a–b/b–a 41.23 176.74 19.50 101.17
1.14 2.17 1.10 0.97
Papp Class h h h h
a
The values reported in the table for primaquine, atropine, and desipramine are taken from Ref. 30 (Bergstrom, C.A.S.; Strafford, M.; Lazorova, M.; Avdeef, A.; Luthman, K.; Artursson, P. Absorption Classification of Oral Drugs Based on Molecular Surface Properties. J. Med. Chem. 2003, 46, 558–570). b The maximum log D value is reported as lipophilicity indicator of 4.
from the S0-value at the various pH values normally encountered in the gastrointestinal tract (1–7.5) and always largely exceeds the projected dose; solubility cannot be considered a limiting factor for the oral absorption of 4. Therefore, compound 4 can be classified, following the guidelines of the Biopharmaceutics Classification System (BCS),40 as a highly soluble drug substance. Finally, the basic nature of 4 allows its rapid dissolution in the acidic environment of the stomach. In the crystal conformation of 4 (free base), the tropane piperidine ring is locked in the boat conformation. This arrangement produces exposure of both a small hydrophilic surface and a large hydrophobic one (Fig. 10, conformation 4B), so that compound 4 shows crystal surfaces reluctant to interact with water (low wettability). On the other hand, rapid interconversion between the boat (conformation 4B) and the fattened chair conformation (conformation 4FC) occurs in lipophilic environments such as chloroform and the basic tropane nitrogen becomes more exposed to the solvent. The potentiometric titration studies revealed a pKa value of 9.11 for the tropane amine nitrogen of 4. This measured value is consistent with the distribution profile (log Dmax ¼ 2.42 at a pH value of ca. 9) and causes the protonation equilibrium to shift toward the protonated species at physiological pH. However, in these conditions compound 4 shows a log D value of 1.45 and a high membrane permeability (Papp value of 41.23 106 cm/s), which is presumably lowered by the pH decrease, as suggested by the log D profile shown in Figure 8, because of the increase of the protonated species. Both crystallographic and NMR studies revealed that, upon protonation, the conformation DOI 10.1002/jps
of the piperidine ring of the tropane moiety is locked in the half-chair conformation, in which compound 4 exposed the maximum hydrophilic surface area (Fig. 11). Therefore, compound 4 is a very rigid molecule in which (a) conformational freedom due to the presence of a rotatable bond is restricted by the interaction between an activated hydrogen and an amide oxygen and (b) the conformation of the tropane piperidine ring is regulated by the environment in such a manner as to optimize the intermolecular interactions with the solvent (large hydrophobic surface in the solidstate and in lipophilic environments such as membranes, large hydrophilic surface in the aqueous environment).43 This chameleon behavior44–48 (Fig. 12) appears to be capable of explaining the biopharmaceutical properties showed by 4, such as low wettability, relatively good solubility, and high membrane permeability. Finally, compound 4 shows biopharmaceutical features quite similar to those of well-known drugs, such as primaquine, atropine, and desipramine (Tab. 6) and can be classified, following the BCS approach,40 as a Class 1 (high solubility–high permeability) drug substance. These data can be used to expand the datasets and improve the general applicability of computational protocols focused on the theoretical biopharmaceutical classification of drug compounds.30a
ACKNOWLEDGMENTS Thanks are due to Rottapharm SpA and Italian MIUR for financial support. Prof. Stefania D’Agata D’Ottavi’s careful reading of the manuscript is also acknowledged.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
2720
CAPPELLI ET AL.
REFERENCES 1. Jackson MB, Yakel JL. 1995. The 5-HT3 receptor channel. Annu Rev Physiol 57:447–468. 2. Israili ZH. 2001. Clinical pharmacology of serotonin receptor type 3 (5-HT3) antagonists. Curr Med Chem Cent Nerv Syst Agents 1:171–199. 3. Silverstone PH, Greenshaw AJ. 1996. 5-HT3 receptor antagonists. Exp Opin Ther Patents 6:471– 481. 4. Cappelli A, Anzini M, Vomero S, Mennuni L, Makovec F, Doucet E, Hamon M, Menziani MC, De Benedetti PG, Giorgi G, Ghelardini C, Collina S. 2002. Novel potent 5-HT3 receptor ligands based on the pyrrolidone structure: Synthesis, biological evaluation, and computational rationalization of the ligand-receptor interaction modalities. Bioorg Med Chem 10:779–801. 5. Cappelli A, Gallelli A, Braile C, Anzini M, Vomero S, Mennuni L, Makovec F, Menziani MC, De Benedetti PG, Donati A, Giorgi G. 2002. Novel potent 5-HT3 receptor ligands based on the pyrrolidone structure. Effects of the quaternization of the basic nitrogen on the interaction with 5-HT3 receptor. Bioorg Med Chem 10:2681–2691. 6. European Pharmacopoeia. 2004. 5th Edition. Council of Europe. France: Strasbourg Cedex. pp 430– 435. 7. Casolaro M, Bottari S, Cappelli A, Mendichi R, Ito Y. 2004. Vinyl polymers based on L-histidine residues. Part 1. The thermodynamics of poly(ampholyte)s in the free and in the cross-linked gel form. Biomacromolecules 5:1325–1332. 8. Gans P, Sabatini A, Vacca A. 1985. SUPERQUAD: An improved general program for computation of formation constants from potentiometric data. J Chem Soc Dalton Trans 1195–1200. 9. Barbucci R, Casolaro M, Danzo N, Barone V, Ferruti P, Angeloni A. 1983. Effect of different shielding groups on the polyelectrolyte behavior of polyamines. Macromolecules 16:456–462. 10. Wells JI. 1988. Pharmaceutical preformulation. The physicochemical properties of drug substances. Chichester: Ellis Horwood Limited Publishers. 11. Huang LF, Tong WQ. 2004. Impact of solid state properties on developability assesment of drug candidates. Adv Drug Del Rev 56:321–334. 12. Snider DA, Addicks W, Owens W. 2004. Polymorphism in generic drug product. Adv Drug Del Rev 56:391–395. 13. Chawla G, Gupta P, Thilagavathi R, Chakraborti AK, Bansal AK. 2003. Characterization of solidstate forms of celecoxib. Eur J Pharm Sci 20, 305– 317. 14. Sheldrick G. 1997. SHELXL-97, A Program for Crystal Structure Refinement, University of Gottingen, Gottingen (Germany), Release 97-2.
15. Farrugia LJ, Win GX. 1999. Program for crystallography package. J Appl Cryst 32:837–838. 16. Chiou WL, Kyle LE. 1979. Differential thermal, solubility, and aging studies on various sources of digoxin and digitoxin powder: Biopharmaceutical implications. J Pharm Sci 68:1224–1229. 17. McGregor C, Saunders MH, Buckton G, Saklatvala RD. 2004. The use of high-speed differential scanning calorimetry (Hyper-dscTM) to study the thermal properties of carbamazepine polymorphs. Thermochim Acta 417:231–237. 18. Jain N, Yang G, Tabibi SE, Yalkowsky SH. 2001. Solubilization of NSC-639829. Int J Pharm 225:41– 47. 19. Ni N, Sanghavi T, Yalkowsky SH. 2002. Solubilization and preformulation of carbendazim. Int J Pharm 244:99–104. 20. Megrab NA, Williams AC, Barry BW. 1995. Oestradiol permeation across human skin, silastic and snake skin membranes: The effects of ethanol/ water co-solvent systems. Int J Pharm 116:101– 112. 21. Li CY, Zimmerman CL, Wiedmann TS. 1996. Solubilization of retinoids by bile salt-phospholipid aggregates. Pharm Res 13:907–913. 22. De Mello H, Echevarria A, Bernardino AM, CantoCavailhero M, Leon LL. 2004. Antileishmanial pyrazolopyridine derivates: Synthesis and structure-activity relationship analysis. J Med Chem 47:5427–5432. 23. OECD Guidelines for Testing of Chemicals, OECD: Paris, 1995, No. 107. 24. Artursson P. 1990. Epithelial transport of drugs in cell culture. I: A model for studying the passive diffusion of drugs over intestinal absorbtive (Caco2) cells. J Pharm Sci 79:476–482. 25. Macura S, Wu¨thrich K, Ernst RR. 1982. Separation and suppression of coherent transfer effects in twodimensional NOE and chemical exchange spectroscopy. J Magn Res 46:269–281. 26. Piantini U, Sorensen OW, Ernst RR. 1982. Multiple quantum filters for elucidating NMR coupling networks. J Am Chem Soc 104:6800–6801. 27. Marion D, Wu¨thrich K. 1983. Application of phase sensitive two-dimensional correlated spectroscopy (COSY) for measurements of 1H-1H Spin-Spin coupling constants in proteins. Biochem Biophys Res Com 113:967–974. 28. Delaglio F, Grzesiek S, Vuister G, Zhu G, Pfeifer J, Bax A. 1995. NMRPipe: A multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6:277–293. 29. Goddard TD, Kneller DG. 1999. SPARKY 3.76 Package, University of California, San Francisco, November 15. 30. Bergstrom CAS, Strafford M, Lazorova M, Avdeef A, Luthman K, Artursson P. 2003. Absorption
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
DOI 10.1002/jps
PHYSICOCHEMICAL AND BIOPHARMACEUTICAL CHARACTERIZATION OF CR3124
31.
32.
33.
34.
35.
36.
37.
38.
39.
classification of oral drugs based on molecular surface properties. J Med Chem 46:558–570. (a) Blais MJ, Enea O, Berthon G. 1974. Grandeurs Thermodynamiques de Protonation de la Pipe`ridine et de ses De`rive`s Alkylsubstitue`s. Thermochim Acta 8:433–438. (b) Cabani S, Conti G, Lepori L. 1971. Thermodynamic study on aqueous dilute solutions of organic compounds. Part. 1—Cyclic amines. Trans. Faraday Soc 67:1933–1942. (c) Christensen JJ. 1976. Handbook of Proton Ionization Heats and Related Thermodynamic Quantities. Ed. Wiley & Sons., Inc. Collin S, Moureau F, Quintero MG, Vercauteren DP, Evrard G, Durant F. 1995. Stereoelectronic requirements of benzamide 5HT3 antagonists. comparison with D2 antidopaminergic analogues. J Chem Soc Perkin Trans 2:77–84. Bradley G, Ward TJ, White JC, Coleman J, Taylor A, Rhodes KF. 1992. Novel antagonists of the 5-HT3 receptor. Synthesis and structure-activity relationships of (2-Alkoxybenzoyl)ureas. J Med Chem 35:1515–1520. Cremer D, Pople JA. 1975. General definition of ring puckering coordinates. J Amer Chem Soc 97:1354–1358. Perrut M, Jung J, Leboeuf F. 2005. Enhancement of dissolution rate of poorly-soluble active ingredients by supercritical fluid processes. Part I: Micronization of neat particles. Int J Pharm 288:3–10. Desai KGH, Kulkarni AR, Aminabhavi TM. 2003. Solubility of rofecoxib in the presence of methanol, ethanol, and sodium lauryl sulfate at (298.15, 303.15, and 308.15) K. J Chem Eng Data 48:942– 945. Avdeef A. 2003. Absorption and drug development (solubility, permeability, and charge state). Hoboken, New Jersey: John Wiley & Sons, Inc. Ghose AK, Crippen GM. 1987. Atomic physicochemical parameters for three-dimensional-directed quantitative structure-activity relationships. 2. Modeling dispersive and hydrophobic interactions. J Chem Inf Comp Sci 27:21–35. Viswanadhan VN, Ghose AK, Revankar GR, Robins RK. 1989. Atomic physicochemical parameters for three dimensional structure directed
DOI 10.1002/jps
40.
41.
42.
43.
44.
45.
46.
47.
48.
2721
quantitative relationships. 4. Additional parameters for hydrophobic and dispersive interactions and their application for an automated superposition of certain naturally occurring nucleoside antibiotics. J Chem Inf Comp Sci 29:163– 172. FDA guideline ‘‘ Waiver of in Vivo Bioavailability and Bioequivalence Studies for Immediate-Release Solid Oral Dosage Forms Based on a Biopharmaceutics Classification System ’’ URL: http:// www.fda.gov/cder/guidance/3618fnl.htm. Glaser R, Peng Q.-J, Perlin AS. 1988. Stereochemistry of the N-methyl group in salts of tropane alkaloids. J Org Chem 53:2172–2180. Wuthrich K. 1986. NMR of proteins and nucleic acids. New York: Wiley-Interscience Publication. Nosza´l B, Kraszni M. 2002. Conformer-specific partition coefficient: Theory and determination. J Phys Chem B 106:1066–1068. Carrupt P.-A, Testa B, Bechalany A, El Tayar N, Descas P, Perrissoud D. 1991. Morphine 6-glucuronide and morphine 3-glucuronide as molecular chameleons with unexpected lipophilicity. J Med Chem 34:1272–1275. El Tayar N, Mark AE, Vallat P, Brunne RM, Testa B, van Gunsterent WF. 1993. Solvent-dependent conformation and hydrogen-bonding capacity of cyclosporin A: Evidence from partition coefficients and molecular dynamics simulations. J Med Chem 36:3757–3764. Gaillard P, Carrupt PA, Testa B. 1994. The conformation-dependent lipophilicity of morphine glucuronides as calculated from the molecular lipophilicity potential. Bioorg Med Chem Lett 4: 737–742. Testa B, Kier LB, Carrupt PA. 1997. A systems approach to molecular structure, intermolecular recognition, and emergence-dissolvence in medicinal research. Med Res Rev 17:303–326. Avdeef A, Barrett DA, Shaw PN, Knaggs RD, Davis SS. 1996. Octanol-, chloroform-, and propylene glycol dipelargonat-water partitioning of morphine-6-glucuronide and other related opiates. J Med Chem 39:4377–4381.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 12, DECEMBER 2006
