120
J. Med. Chem. 1996, 39, 120-125
Biarylcarboxylic Acids and -amides: Inhibition of Phosphodiesterase Type IV versus [3H]Rolipram Binding Activity and Their Relationship to Emetic Behavior in the Ferret Allen J. Duplantier,* Michael S. Biggers, Robert J. Chambers, John B. Cheng, Kelvin Cooper, David B. Damon, James F. Eggler, Kenneth G. Kraus, Anthony Marfat,* Hiroko Masamune, Joann S. Pillar, John T. Shirley, John P. Umland, and John W. Watson Central Research Division, Pfizer Inc, Groton, Connecticut 06340 Received July 13, 1995X
In addition to having desirable inhibitory effects on inflammation, anaphylaxis, and smooth muscle contraction, PDE-IV inhibitors also produce undesirable side effects including nausea and vomiting. In general, compounds that inhibit PDE-IV also potently displace [3H]rolipram from a high-affinity binding site in rat cortex.1,2 While this binding site has not been identified, it has been proposed to be an allosteric binding site on the PDE-IV enzyme.3 Preliminary studies have suggested that the emetic potency of PDE-IV inhibitors is correlated with affinity for the brain rolipram binding site rather than potency at inhibiting PDE-IV enzyme activity. Efforts to eliminate the emetic potential of PDE-IV inhibitors were directed toward developing compounds with decreased [3H]rolipram binding affinity while retaining PDE-IV potency. Thus, a novel series of 4-(3-alkoxy-4-methoxyphenyl)benzoic acids and their corresponding carboxamides were prepared and evaluated for their PDE-IV inhibitory and rolipram binding site properties. Modification of the catechol ether moiety led to phenylbutoxy and phenylpentoxy analogues that provided the desired actitivity profile. Specifically, 4-[3-(5-phenylpentoxy)-4methoxyphenyl]-2-methylbenzoic acid, 18, was found to exhibit potent PDE-IV inhibitory activity (IC50 0.41 µM) and possessed 400 times weaker activity than rolipram for the [3H]rolipram binding site. In vivo, compound 18 was efficacious in the guinea pig aerosolized antigen induced airway obstruction assay (ED50 8.8 mg/kg, po) and demonstrated a significant reduction in emetic side effects (ferret, 20% emesis at 30 mg/kg, po). Introduction Inhibition of phosphodiesterase type IV (PDE-IV) represents a promising therapeutic target for the treatment of various inflammatory diseases such as arthritis and asthma.4-7 However, the currently known PDEIV inhibitors, such as the prototypical agent rolipram, are hampered by nausea and emetic side effects limiting their therapeutic potential.8 While it is possible that PDE inhibitors have peripheral emetic actions, there is current evidence demonstrating an effect on the area postrema.9 Therefore it is reasonable to assume that keeping a PDE-IV inhibitor out of the brain could reduce its emetic potential. To reduce this side effect liability, we reasoned that incorporation of an arenecarboxylic acid would reduce penetration into the brain and, thus, reduce activity at the emetic center in the central nervous system (CNS). Compound 4 was designed on the basis of a previously reported series of catechol ethers with potent PDE-IV inhibitory activity10 and synthesized as an initial test of this hypothesis. Although 4 displayed potent PDE-IV inhibitory activity, it also showed a similar emetic liability to rolipram. As a follow-up hypothesis, we considered that the affinity of a compound for the [3H]rolipram binding site might be a factor in determining its emetic potency. In fact, [3H]rolipram binding activity has been correlated with emetic side effects in a study in dogs.11 Furthermore, a reduction in the ratio of the concentrations required for inhibition of PDE-IV to [3H]rolipram binding activity to mouse brain was previously achieved by catechol ether modifications to a series of rolipram-related imiX
Abstract published in Advance ACS Abstracts, December 1, 1995.
0022-2623/96/1839-0120$12.00/0
dazolidinones (1).12 Following this reasoning, 4 was modified to reduce its affinity for the [3H]rolipram binding site, leading to compound 18 which displayed a greatly reduced emetic liability. The SAR of this novel series of arenecarboxylic acid and -carboxamide catechol ethers leading to compound 18 is reported herein.
Chemistry All of the biaryl carboxylic acids reported herein were prepared from 5-bromoguaiacol as presented in Scheme 1. Etherification of 5-bromoguaiacol with the appropriate alcohol under Mitsunobu conditions13 gave the catechol diethers 2a-h in 80-90% yield. Catechol diether 2k was synthesized from 5-bromoguaiacol by acylation with 4-phenylbutyryl chloride followed by treatment with [bis(cyclopentadienyl)titanium](µ-chloro)(µ-methylidene)dimethylaluminum (Tebbe’s reagent)14 to give enol ether 2j (Scheme 2). Subsequent subjection of 2j to a Simmons-Smith reaction15 readily provided 2k. Catechol diethers 2a-h and 2k were then converted to their corresponding arylzinc chlorides by treatment with n-butyllithium and zinc chloride (Scheme 1) and were coupled in situ to the desired aryl iodide or bromide by palladium catalysis.16 The resulting biaryl © 1996 American Chemical Society
Biarylcarboxylic Acids and -amides
Scheme 1a
Journal of Medicinal Chemistry, 1996, Vol. 39, No. 1 121
either ip or po.20 Both emetic (vomiting and retching) as well as prodromal activity (gagging, mouth scratching, and salivation) were scored. Structure-Activity Relationships
a Reagents and conditions: (i) R1OH, DEAD, PPh , THF, 25 3 °C; (ii) n-butyllithium, THF, -78 °C; ZnCl2, then warm to 25 °C; Pd(PPh3)4, aryl iodide or bromide; (iii) NaOH, EtOH, reflux; (iv) SOCl2; DMF (cat.), NH3(g)/CH2Cl2 or NH4OH/THF.
Scheme 2a
a Reagents and conditions: (i) 4-phenylbutyryl chloride, 4DMAP, pyridine, 0 °C; (ii) Tebbe reagent, pyridine, 1:3 THF/ toluene; (iii) CH2I2, Zn-Cu couple, I2, ether, reflux.
esters (e.g. 3) were then saponified to provide carboxylic acids 4-12 and 18. Selected carboxylic acids were subsequently converted to their corresponding amides by successive treatment with thionyl chloride and either anhydrous ammonia or ammonium hydroxide. Biology Over the last three decades PDEs have evolved into seven major families (PDE-I, -II, -III, -IV, -V, -VI, and -VII) according to their substrate sensitivity, Ca2+/ calmodulin requirement, and inhibitor selectivity.17 PDE-IV, a cAMP-specific and Ca2+-independent enzyme, was shown to be a key isozyme that controls the hydrolysis of cAMP in mast cells, basophils, eosinophils, monocytes, and lymphocytes.17 As a result, inhibitors of PDE-IV block the activation of various mediators from these cells and may represent a new class of antiinflammatory drugs. Recently, four human cDNA isoforms of PDE-IV (PDE-IV-A, -B, -C, and -D) were identified, and all four were expressed in human lung.18 PDE-IV was isolated from human lung homogenates by differential centrifugation and ion-exchange chromatography.19 Human lung PDE-IV cannot be stimulated by CaCl2 (1 µM to 100 mM) and is activated by MnCl2 and MgCl2 (EC50 ∼10 µM). The double reciprocal plot of the results of enzyme kinetic experiments reveal a Km of 1.0 ( 0.1 µM and a Vmax ) 0.51 ( 0.1 µM/(min/ mg) protein for lung PDE-IV (n ) 4). Except for rolipram (IC50 ) 3.45 µM), vinpocetine (PDE-I inhibitor), SKF-94120 (PDE-III inhibitor), zaprinast (PDE-V inhibitor), and cGMP are inactive at 100 µM. [3H]Rolipram binding activity was evaluated using whole mouse brain.2,12 The antianaphylactic activity of PDE-IV inhibitors was determined by the ability to block the airway obstructive response of anti-OA-IgG1 passively sensitized guinea pigs challenged with aerosolized ovalbumen. Emetic activity was determined in ferrets dosed
As mentioned in the Introduction, biarylcarboxylic acid 4 was designed to have minimal CNS exposure. We reasoned that the carboxylic acid functionality would be polar enough to prohibit any compound migration into the brain. Furthermore, the PDE-IV inhibitory activity (IC50 0.52 µM) of compound 4 was enhanced compared to that of rolipram (Table 1), making 4 an exciting compound for testing our hypothesis. However, compound 4 caused emesis in a ferret emesis model (100% prodromal or emetic behavior at 10 mg/kg, ip), and this result led us to the redirected strategy of preparing potent PDE-IV inhibitors with reduced [3H]rolipram binding activity. We chose the cyclopentoxy group and the carboxylic acid functionality of compound 4 for SAR development. The corresponding methyl ester 3 did not inhibit PDEIV, and this result suggested a hydrogen-bonding interaction between the carboxylic acid moiety in 4 and the PDE IV isozyme. The m-carboxylic acid 5 also inhibited PDE-IV, but was 3 times less potent than 4. Thus, the p-carboxylic acid 4 was chosen for further SAR studies. Replacement of the cyclopentyl group with either an (S)- or (R)-2-exo-norbornyl group10 (6 and 7, respectively) had no effect on rolipram binding, and to our surprise, the two enantiomers had equipotent PDE-IV inhibition. In efforts to reduce the affinity to the [3H]rolipram binding site, we next applied some of the catechol ether side chain modifications previously disclosed for compound 1.12 In parallel to the SAR derived for 1, the 2-indanyl analogue 8 was 6 times more potent than 4 against the PDE-IV enzyme and displayed ∼4 times less [3H]rolipram binding activity. Moreover, the 4-phenylbutyl (9) and the (S)- and (R)-5-phenylpent-2-yl analogues (10 and 11, respectively) were 10 times weaker than 4 for [3H]rolipram binding activity. The 10 times difference in PDE-IV potency between compound 10 and its enantiomer, 11, revealed for the first time enantiospecificity at the catechol ether moiety. Interestingly, the achiral cyclopropyl analogue 12 was equipotent to 10 for both PDE-IV inhibition and rolipram binding activity. The COOH group was the next structural moiety that we modified. We believed that replacement of the COOH group with CONH2 would still allow for the required hydrogen bonding at this site to the PDE-IV enzyme and should therefore provide analogues with potent PDE-IV inhibition. Thus, replacement of the carboxylic acid functionality of the acid analogue 8 with a carboxamide group provided compound 13, displaying slightly weaker PDE-IV activity and slightly more potent [3H]rolipram binding affinity. The 4-phenylbutoxy analogue 14 was ∼60 times less potent than its acid homologue as a PDE-IV inhibitor, but most interestingly 14 showed very low affinity for the [3H]rolipram binding site. More striking was the dramatic difference in PDE-IV potency between the 5-phenylpent-2-yl enantiomers, 15 and 16. The cyclopropyl analogue 17 showed more potent PDE-IV activity (IC50 0.67 µM) compared to the 4-phenylbutoxy analogue 14. This
122
Journal of Medicinal Chemistry, 1996, Vol. 39, No. 1
Duplantier et al.
Table 1. PDE Inhibition of Biarylcarboxylic Acids and -amides Prepared from 2
entry
R1
R2
R3
formula
rolipram 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
cyclopentyl cyclopentyl cyclopentyl (S)-(+)-2-exo-norbornyl (R)-(-)-2-exo-norbornyl 2-indanyl 4-phenylbutyl (i) (ii) (iii) 2-indanyl 4-phenylbutyl (i) (ii) (iii) 5-phenylpentyl
CO2Me CO2H H CO2H CO2H CO2H CO2H CO2H CO2H CO2H CONH2 CONH2 CONH2 CONH2 CONH2 CO2H
H H CO2H H H H H H H H H H H H H CH3
C20H22O4 C19H20O4 C19H20O4 C21H22O4 C21H22O4 C23H20O4 C24H24O4 C25H26O4 C25H26O4 C26H26O4 C23H21NO3 C24H25NO3 C25H27NO3 C25H27NO3 C26H27NO3 C26H28O4
analysisa mp (°C) C, H C, H C, H C, H C, H C, Hb C, Hc C, H C, H C, H C, H, Nd C, H, N C, H, N C, H, Ne C, H
131-2 230-2 149-50 230-2 234-6 244-7 178-9 159-60 159-60 141-2 237-9 178-80 182-4 183-4 154-5 127-8
PDE-IV IC50 (µM) or % inh, conc ( SEM (n)f
rolipram bind. IC50(µM) or % inh, conc ( SEM (n)f
3.45 ( 0.34 (7) 25.1 ( 3.9%, 100 µM (3) 0.52 ( 0.18 (3) 1.64 ( 0.81 (2) 0.11 ( 0.03 (2) 0.13 ( 0.04 (2) 0.081 ( 0.006 (3) 0.30 ( 0.04 (2) 0.14 ( 0.01 (3) 1.42 ( 0.15 (5) 0.13 ( 0.05 (2) 0.16 ( 0.11 (2) 20.3 ( 15.0 (4) 1.94 ( 1.36 (3) 24 ( 10%, 100 µM (3) 0.67 ( 0.17 (3) 0.41 ( 0.02 (2)
0.004 ( 0.0002 (111) 0.13 ( 0.04 (3) 0.009 ( 0.0005 (3) 0.015 ( 0.002 (3) 0.003 ( 0.0005 (3) 0.003 ( 0.0004 (3) 0.015 ( 0.004 (3) 0.323 ( 0.061 (3) 0.111 ( 0.035 (5) 0.623 ( 0.050 (3) 0.129 ( 0.013 (3) 0.005 ( 0.002 (3) 26 ( 3%, 10 µM (3) 33 ( 5%, 10 µM (3) 41 ( 1%, 100 µM (3) 1.90 ( 0.75 (3) 1.68 ( 0.005 (2)
a Unless indicated, all values were within 0.4%. b C: calcd, 76.65; found, 75.93. c C: calcd, 76.56; found, 76.06; H: calcd, 6.43; found, 5.92. d N: calcd, 3.90; found, 3.27. e C: calcd, 77.78; found, 74.61. f Number in parentheses denotes number of times tested.
Table 2. In Vivo Results for Selected PDE-IV Inhibitors compound
PDE-IV IC50, µM
rolipram binding IC50 or % inh, conc
AAIAO assay ED50 mg/kg, po
ferret emesis (drug plasma conc)e
rolipram 15 17 18
3.45 1.94 0.67 0.41
0.004 µM 33%, 10 µM 1.90 µM 1.68 µM
0.6a 1.1a 26 ( 20% inh at 10 mg/kg 8.8a
2/10b at 0.03 mg/kg, poc (
