J . Med. Chem. 1988,31, 2081-2086
2081
than compensates the polarity of the M3 minimum. Thus, if we postulate that P2 affinity and selectivity are based on specific hydrophobic interactions, a validation of this hypothesis cannot be found in MEP patterns. Clearly a variety of approaches are necessary for a relevant pharmacophore to be obtained. Experimental Section
STO-3G basis set. MEPs were obtained with a slightly modified version of DENPOT (QCPE 360) and were generated in planes parallel to the aromatic ring, viewing compounds from above and below the aromatic ring plane for each compound. All calculations were performed on the CDC CYBER 170/855 and CRAY 1s computers of the Federal Institute of Technology in Lausanne and a Norsk Data ND 560 computer of the University of Lausanne.
Molecular Electrostatic Computations. In order to reduce computing times the calculations were carried out on the primary amine analogues of all compounds. Although adrenoceptor ligands bind and act as in their protonated form, the MEP in thus study were calculated for the neutral molecules. This renders the MEP details much more visible while not altering the difference in electrostatic potential between local minima and maxima.20 Standard bond lengths and angles were used as input for the calculations. Wave functions and electronic densities were calculated with the ab initio Monstergauss 81 program as reported previously20 and with the Gaussian 82 program,51 using a minimal
Acknowledgment. We are indebted to the Swiss National Science Foundation for Grant 3.508-0.86.
(51) Pople, J. A,; Binkley, J. S.; Frisch, M. J.; DeFrees, D. J.; Raghavachari, K.; Whiteside, R. A.; Schelgel, H. B.; Fluder, E. M., University of Carnegie-Mellon, 1985.
Registry No. 1, 51-41-2; 2, 34391-04-3; 3, 20879-16-7; 4, 30236-31-8; 5, 116127-09-4; 6, 23846-71-1; 7, 59995-59-4; 8, 81102-77-4; 9, 26328-12-1; 10, 26328-11-0; 11, 47141-42-4; 12, 38104-34-6; 13, 42200-33-9; 14, 4199-09-1; 15, 116127-10-7; 16, 68107-82-4; 17, 95586-73-5; 18, 37936-65-5; 19, 91324-92-4; 20, 102151-02-0; 21, 93379-54-5; 22, 93221-48-8; 23, 99103-03-4; 24, 81024-42-2; 25, 102152-32-9; 26, 116076-61-0; 27, 72795-23-4; 28, 60979-28-4; 29, 54340-62-4; 30, 39832-48-9; 31, 26839-75-8; 32, 76596-57-1. Supplementary Material Available: Plots of the MEP of compounds 1, 3, 5-9, 11-13, 15, 16, 18, 29-32 in a plane parallel to, and 2 8, above, the plane of the aromatic ring (9 pages). Ordering information is given on any current masthead page.
Radioiodinated Benzodiazepines: Agents for Mapping Glial Tumors Marcian E. Van Dort, Brian J. Ciliax, David L. Gildersleeve, Philip S. Sherman, Karen C. Rosenspire, Anne B. Young, Larry Junck, and Donald M. Wieland* Departments of Internal Medicine and Neurology, University of Michigan Medical Center, A n n Arbor, Michigan 48109. Received March 3, 1988
Two isomeric iodinated analogues of the peripheral benzodiazepine binding site (PBS) ligand Ro5-4864 have been synthesized and labeled in high specific activity with iodine-125. Competitive binding assays conducted with the unlabeled analogues indicate high affinity for PBS. Tissue biodistribution studies in rats with these lZ5I-labeled ligands indicate high uptake of radioactivity in the adrenals, heart, and kidney-tissues known to have high concentrations of PBS. Preadministration of the potent PBS antagonist PK 11195 blocked in vivo uptake in adrenal tissue by over 75%, but to a lesser degree in other normal tissues. In vivo binding autoradiography in brain conducted in C6 glioma bearing rats showed dense, PBS-mediated accumulation of radioactivity in the tumor. Ligand 6 labeled with lZ3Imay have potential for scintigraphic localization of intracranial glioma.
Since their discovery in the late 19509, the 1,4-benzodiazepines have constituted a class of widely used anxiolytic and anticonvulsant drugs.' In the last several years, the presence of two pharmacologically distinct subclasses of benzodiazepine binding sites have been demonstrated. One class, the central benzodiazepine receptor, is localized in central neuronal tissue and represents the site at which benzodiazepineligands exert their anxiolytic In addition, binding sites for benzodiazepines have been identified in peripheral tissues including the adrenal cortex, nasal epithelium, kidney, and heart; these sites were subsequently termed the peripheral benzodiazepine binding sites (PBS).4y5 It has also been shown, however, (1) Sternbach, L. H. The Benzodiazepines: From Molecular Bi-
ology to Clinical Practice; Costa, E., Ed.; Raven: New York, 1983; pp 1-66. (2) Tallman, J. F.; Gallager, D. W. Annu. Reu. Neurosci. 1985,8, 21.
(3) Tallman, J. F.; Paul, S. M.; Skolnick, P.; Gallager, D. W. Science (Washington, D.C.) 1982, 207, 274. (4) DeSouza, E. B.; Anholt, R. R. H.; Murphy, K. M. M.; Snyder, S. H.; Kuhar, M. J. Endocrinology 1985, 116, 567. (5) Davies, L. P.; Huston, V. Eur. J . Pharmacol. 1981, 73, 209. 0022-2623/88/1831-2081$01.50/0
Table I. Radiosvnthetic Data isolated radiochemical ligand yield, % sp. act., Ci/mmol 117-128 ( N = 3) 5 18-31 (N = 3) 103-145 ( N = 3) 21-43 (N = 6) 6
radiochemical purity, %
>98 >98
that these sites though present mainly in the periphery are also located in the olfactory bulb as well as glial cells in brain.6s7 The pharmacological distinction between these two types of binding sites can be demonstrated by the use of highly selective ligands. The clinically potent anticonvulsant clonazepam binds with nanomolar affinity to the central benzodiazepine receptor but with less than micromolar affinity to PBS.s Conversely Ro5-4864, which exhibits no anxiolytic activity, has nanomolar affinity for the peripheral binding site but a 1000-fold lower affinity (6) Anholt, R. R. H.; Murphy, K. M. M.; Mack, G. E.; Snyder, S. H. J . Neurosci. 1984, 4, 593. (7) Schoemaker, H. M.; Morelli, M.; Deshmukh, P.; Yamamura, H. I. Brain Res. 1982, 248, 396. (8) Wang, J. K. T.; Taniguchi, T.; Spector, S. Mol. Pharmacol. 1984, 25, 349. 0 1988 American Chemical Society
Van Dort et al.
2082 Journal of Medicinal Chemistry, 1988, Vol. 31, No. 11 Scheme I
for the central receptor.8 Moreover, benzodiazepine binding to PBS is unaffected by the presence of y-aminobutyric acid unlike the case of central receptor^.^!^ Recent studies have indicated the presence of high concentrations of PBS in human glial tumors.'O Successful delineation of intracranial (26 gliomas in rats has been achieved by in vivo binding autoradiography using the peripheral binding site specific ligand [3H]Ro5-4864.11 This observation indicated that an analogue of Ro5-4864 radiolabeled with a y-emitting radionuclide could have clinical utility for scintigraphic localization of intracranial human glioma. We report here the synthesis of two 1251labeled analogues of Ro5-4864, their in vitro affinity for PBS, and preliminary brain autoradiographic studies conducted in c6 glioma bearing rats. Chemistry The novel approach of Sugasawa and co-workers12was used with minor modifications to prepare the halogenated 2-aminobenzophenone precursors (Scheme I). Condensation of the aniline-boron trichloride complex with the appropriate benzonitrile analogue in the presence of AlC13 gave the corresponding ketimine which afforded the 2aminobenzophenone analogue on acid hydrolysis. The latter underwent smooth cyclization with glycine ethyl ester hydrochloride in pyridine to give the corresponding l,.i-benzodiazepine analogue by the procedure of Sternbach et al.13 N-Methylation was carried out in excellent yields by treatment of the corresponding sodium salt with methyl iodide.I3 Radioiodide labeling was performed by slight modification of the ammonium sulfate catalyzed solid-phase exchange technique described previously.'* The results are summarized in Table I. Exchanges were conducted in 3-mL septum-sealed multidose vials vented with a Marangos, P. J.; Patel, J.; Boulenger, J. P.; Clark-Rosenberg, R. Mol. Pharmacol. 1982, 22, 26. Starosta-Rubinstein,S.;Ciliax, B. J.; Penney, J. B.; McKeever, P.; Young, A. B. Proc. Natl. Acad. Sci. U.S.A. 1987,84, 891. Ciliax, B. J.; Young, A. B., unpublished findings. Sugasawa, T.; Toyoda, T.; Adachi, M.; Sasakura, K. J. Am. Chem. SOC. 1978, 100, 4842. Sternbach, L. H.; Fryer, R. I.; Metlesics, W.; Reeder, E.; Sach, G.; Saucy, G.; Stempel, A. J. Org. Chem. 1962, 27, 3788. (a) Mangner, T. J.; Wu, J. L.; Wieland, D. M. J. Org. Chem. 1982,47, 1484. (b) Otto, C. A.; Lee, H.; Mangner, T. J.; Wieland, D. M. Appl. Radiat. Zsot. 1986, 37, 205.
Table 11. Affinity of Ligands for Central Receptor and Peripheral Benzodiazepine Binding Site
R2
K;,O nM f SD R2 PBS, nM CBR,nM C1 7.3 f 2.3 >loo0 5 C1 13.3 f 4.8 >1000 6 c1 I 13.6 f 4.5 >1000 " N = 3 for each ligand per assay; PBS is peripheral benzodiazepine binding site; CBR is central benzodiazepine receptor.
ligand Ro5-4864
R1 C1 I
disposable 10-cm3 plastic syringe which served as the distillate condenser. During the heating process, charcoal and aqueous sodium thiosulfate traps were connected in series with the condenser to trap any volatile radioiodine that might be released. Radiochemical purity was assessed by radio-TLC and by reversed-phase high-performance liquid chromatography (radio-HPLC) using systems A and B. In all cases the retention times of the labeled materials were in agreement with that of the unlabeled ligands. The chemical purity of the ligands following the isotope-exchange procedures was assessed by an indirect approach. The exact exchange procedure and workup described above was repeated for each ligand with Na1271 instead of Na1251,and the reaction products were analyzed by normal phase HPLC (system A). These studies demonstrated that the labeling conditions resulted in ligands with greater than 95% chemical purity. Results and Discussion The affinity of the ligands for PBS was evaluated in vitro by their ability to displace [3H]flunitrazepam (5 nM) binding in rat brain and kidney tissue sections. These results are summarized in Table 11. As seen from this data, both 5 and 6 display high affinity for PBS with Ki values of 13.3 and 13.6 nM, respectively. The Ki for Ro5-4864 under the same conditions was determined to be 7.3 nM. Thus, substitution of iodine for chlorine at
Radioiodinated Benzodiazepines
Journal of Medicinal Chemistry, 1988, Vol. 31, No. 11 2083
Table 111. Concentration of Radioactivity at Various Time Intervals after Intravenous Administration of lz5I-Labeled brain time ligand adrenal blood heart kidney thyroide 5 min 5 4.85 f 0.54 0.75 f 0.06 0.20 f 0.01 1.15 f 0.06 1.10 f 0.03 1.50 f 0.43 3.27 f 0.58 1.00 f 0.10 0.22 f 0.01 2.18 f 0.11 6 7.10 f 1.85 1.78 f 0.13 0.74 f 0.06 30 min 5 3.63 f 0.53 0.46 f 0.03 0.15 f 0.02 0.77 f 0.10 4.42 f 0.48 0.60 f 0.05 0.11 f 0.04 1.13 f 0.10 1.05 f 0.12 6 7.85 f 0.32 2.25 f 0.14 0.41 f 0.01 0.12 f 0.00 2h 5 1.84 f 0.22 0.63 f 0.04 0.58 f 0.03 5.57 f 0.74 0.51 f 0.03 4.26 f 0.34 6 2.38 f 0.25 0.35 f 0.02 0.10 f 0.01 0.50 f 0.05 (0.49 f 0.07) 6h 5 1.30 f 0.10 0.46 f 0.04 0.12 f 0.02 0.57 f 0.04 17.28 f 4.90d 0.59 f 0.04 0.14 f 0.04 0.21 f 0.03 11.85 f 1.88d 6 0.60 f 0.05 0.05 f 0.01 0.19 f 0.02 0.25 f 0.07 0.25 f 0.08 17.20 f 2.65 24 h 5 0.42 f 0.16 0.05 f 0.02 0.21 f 0.06 0.02 f 0.01 0.01 f 0.00 0.01 f 0.00 6 0.05 f 0.01 0.03 f 0.01 10.85 f 5.05 'Concentrations, normalized to a 200-g rat, are percent of injected dose/gram of tissue (wet weight). bMale Sprague-Dawley rats ( N = 5 for each ligand per time point) weighing 160-265 g were used. 'Value in parentheses was obtained with rats on oral KI solution. dThe mean concentration of radioactivity in the thyroid of two rats administered NalZ5I(20 pCi) was 360.5% dose/g, suggesting that