Synthesis of sterically hindered 3-(azolyl)pyridines

Sterically hindered 2,4-disubstituted 3-(1,2,4-oxadiazol-3-yl)-, 3-(imidazol-2-yl)- and 3-(thiazol-2-yl)pyridines were synthesized from the corresponding nicotinonitriles via amidoximes, amidines and thioamides, respectively. N -Alkyl- and N -arylamidines were prepared directly from nicotinonitriles using microwave technology. The series of 3-(azolyl)pyridines form a combinatorial library of heterocyclic derivatives of nicotinic acid and were examined by prognostic software PASS.


Introduction
One of the most promising strategies for searching for new pharmaceuticals and improving the properties of known ones is a replacement of metabolically unstable functional groups with bioisosteric five-membered heterocyclic rings. 1,2The best-known azole that is used as a metabolism-resistant surrogate for the carboxylate group is 5-substituted tetrazole, since they both possess comparable acidity and size. 3Nicotinic acid and its derivatives can be very often found among the carboxylic acids investigated with respect to such replacements. 2,4Thus, 3-(5tetrazolyl)pyridines were reported as a new class of lipolysis inhibitors 5 and bioisosteres of arecoline. 2t the same time, other azoles can also serve as the isosteric equivalents for labile functional groups in the derivatives of nicotinic acid.Thus, 3-(1,3,4-oxadiazol-2-yl)pyridines show antimicrobial and antifungal activity 6 and were tested as nonpeptidic inhibitors of human neutrophil elastase. 7The corresponding derivatives of 1,2,4-oxadiazole displayed high affinity and efficacy as muscarinic agonists. 8An antimycobacterial 9 as well as an anthelmintic 10 activity of pyridines substituted with 1,2,4-oxadiazoles as isosteres of nicotinic acid was described.3-(Imidazol-2yl)pyridines are known as inhibitors of xanthine oxidase, 11 nonsteroidal antiinflammatory agents, 12 potent and selective NPY5 receptor antagonists 13 as well as being tested as KDR kinase inhibitors. 14-(Thiazol-2-yl)pyridines are known as inhibitors of superoxide production by human neutrophils, 15 selective cyclooxygenase-2 inhibitors 16 and effective inhibitors of human cytochrome P-450 (CYP2A6), the major nicotine metabolizing enzyme; 17 they also displayed antifungal and antibacterial 18 as well as antitubercular activity; 19 besides, these compounds were used as building block for synthesis of berninamycinic acid 20 and heterocyclic core of micrococcin P1, 21 preparation of the central heterocyclic skeleton of antibiotic A10255, 22 macrocyclic antibiotic GE 2270 A 23 and other various thiostrepton-type macrocyclic antibiotics.24 On the other hand, it was found that pyridine derivatives bearing a sterically hindered tetrazole unit at position 3 show some very interesting pharmacological properties.25 In view of above, it would be interesting to create a combinatorial library consisting of sterically hindered 3-(azolyl)pyridines 2-6 as derivatives of nicotinic acid (Figure 1).Such combinatorial library would be also of interest since only two examples of 2,4,5-trisubstituted 3-(azolyl)pyridines [namely, 3-(1,2,4-oxadiazol-3-yl)pyridines, structure 4] were found in the literature.26 It should be noted that these two compounds were identified as mixtures (purity 60%) prepared with low yields (25%).

Results and Discussion
Two syntheses of 1,2,4-oxadiazoles 12 from nitriles 7 are widely known. 29The first can be accomplished via the imidates 8, 30 and the second approach is the condensation of amidoximes 10 with carboxylic acid derivatives (anhydrides, chlorides) followed by final ring closure (Scheme 1).The acylated intermediates 9, 11 were mentioned as isolable compounds for both reaction routes. 29It was insisted in many related publications that the both imidates 8 and amidoximes 10 can be easily prepared using facile procedures 9,10,30,31 including conditions of parallel synthesis.8d,26,32 It should be noted that aromatic nitriles described in many of these papers had no substituents alongside of cyano group.However, all our attempts to prepare the imidates 8 from nicotinonitriles 1 failed.It was not surprising to us, since many reactions of nitrile groups are very susceptible to steric hindrance (for example, the Pinner reaction). 33Further, the same result was obtained when we attempted to synthesize the amidoximes 10 from nicotinonitriles 1 using some conventional procedures (interaction of nitriles with NH 2 OH • HCl in the presence of various bases).Indeed, a dependence of this reaction on sterical circumstances was also reported.31d Now, we have found that the sterically hindered nicotinonitriles 1 can be converted into desired amidoximes 13 by long-continued heating with 50% aqueous solution of hydroxylamine (Scheme 2).The resultant amidoximes 13 were in many cases contaminated by corresponding amides 14 yields of which increased along with growth of volume of the substituent R 1 at position 4 (Table 1).Finally, amide 14e turned out the sole reaction product for 4-tert-butylsubstituted nicotinonitrile 1e.It was reported previously that such amide formation is a result of initial attack by the oxygen atom of hydroxylamine to cyano group but not due to hydrolysis of nitrile or amidoxime. 34We found that the "amidoxime: amide" ratio did not depend on ratios of starting reagents as well as temperature and reaction time.Unfortunately, our attempts to improve the yields of amidoximes 13 as well as to reduce the reaction time using microwave conditions turned out also to be inefficient.The target amidoximes 13a-d were easily separated from the amides 14 with column chromatography or crystallization.A treatment of the amidoximes 13a-d with aromatic and aliphatic acyl chlorides gave rise to O-acylated intermediates 15 (Scheme 3) one of them was isolated and purified for identification.The other O-acylamidoximes 15 were used as crude materials for further cyclization without an additional purification.The cyclization was carried out by action of tetra-n-butylammonium fluoride (TBAF) according to literature procedure. 35Besides, the product 16f was obtained in reaction of starting amidoxime 13c with excess of trifluoroacetic anhydride as acylating and dehydrating reagent (i.e., without isolation of intermediate 15).The synthesis of sterically hindered 3-(imidazol-2-yl)pyridines 19a,b and 22 utilized the condensation of amidines 17a,b with α-haloketones 18, 20 (Schemes 4, 5).However, the conversion of nicotinonitriles 1 into amidines 17 as an initial step of this reaction presented severe difficulty.Most convenient synthesis of amidines from nitriles consist of two steps where a conversion of nitriles into imidates (the Pinner reaction) followed by reaction with amines yields the desired amidines. 36Unfortunately, this approach turned out entirely unsuitable for sterically congested nicotinonitriles 1 (vide supra).After several unsuccessful efforts to reveal the conditions for direct reaction of the nicotinonitriles 1 with amines using anhydrous aluminium chloride, 37 lithium bis(trimethylsilyl)amide 38 and some other reagents, 12,39 trimethylaluminum was employed for this reaction. 40However, very low yields of the amidines were achieved under the conditions described. 40Therefore, we modified these conditions and found that the target amidines 17 can be successfully prepared from the nicotinonitriles 1 and amines in the presence of trimethylaluminum using microwave technology (MW) (Scheme 4).The reactions of the amidines 17a,b with 3-chlorobutan-2-one 18 were also carried out under microwave irradiation yielding directly the desired 3-(imidazol-2-yl)pyridines 19a,b (Scheme 4).At the same time, simple refluxing of mixture of amidine 17a with 3-bromo-1,1,1-trifluoroacetone 20 gave rise to hydroxy derivative 21 that was easily dehydrated with trifluoroacetic anhydride (Scheme 5).Interestingly, the compound 21 existed as a mixture of two stereoisomers.The 1 H NMR spectrum of this product 21 was a superposition of two pictures in ratio 57:43 where a noncoincidence of chemical shifts was observed only for the resonances of methyl and methylene groups (see Experimental Section).It is obviously connected with difficulty of rotation at bond "pyridine-imidazoline" since further dehydration gave rise to the sole product 22.The structure of the final 3-(imidazol-2-yl)pyridine 22 was confirmed by the assignment of 1 H and 13 C NMR spectra involving 1 H- 13 C HSQC and HMBC experiments.The position of trifluoromethyl group at position 4 of imidazole ring was established by the cross-peaks in the 2D-1 H- 13 C-HMBC spectra between 5-H in imidazole nucleus and 1-C in N-phenyl group as well as by interaction NOE in 2D-NOESY spectra between 5-H and ortho-protons of phenyl (Figure 2).Thioamides 23a-i required for synthesis of the sterically hindered 3-(thiazol-2-yl)pyridines 25am were prepared by reactions of starting nicotinonitriles 1a-i with O,O-diethyldithiophosphoric acid (Scheme 6, Table 3).The latter and similar acids are known as effective reagents for conversion of various nitriles into thioamides in high yields under conditions sufficiently mild to preserve intact a wide range of functional groups. 41Some conventional methods (e.g., using ammonium sulfide 42 ) for the transformation of nitriles into thioamides were initially tested but turned out insufficient.It should be noted that the complete conversion of nitriles 1a-i into amidines 23a-i was not achieved in all the experiments even by long-term heating of reaction mixtures (up to 100 h).Some amount of a starting nitrile 1 remained and can be recovered in many cases (Table 3).Moreover, the conversion decreased if the temperature of reaction mixture was elevated to 100-120 o C. It can be explained by thermal decomposition of thioamides 23 to starting nitriles 1 and H 2 S. 43 It is known also that a heating of thioamides with pyridine caused an analogous decomposition; 44 it is not improbable that a similar effect can have the starting cyanopyridines 1.Unfortunately, our attempts to increase the conversion and improve the yields of the desired thioamides 23a-i using microwave technology turned out unsuccessful.The series of sterically hindered 3-(thiazol-2-yl)pyridines 25a-m was synthesized by wellknown reactions 45 of thioamides 23a-i with α-haloketones 18, 20, 24a-i (Scheme 8).The reactions were carried out in anhydrous ethanol to form hydrohalogenides of the target 3-(thiazol-2yl)pyridines 25a-m (Table 4).At the same time, intermediate alcohols 26a-f were isolated in a number of cases (Scheme 8, Table 4).These alcohols 26a-f were easily dehydrated by trifluoroacetic anhydride according to known procedures. 46

Conclusions
In summary, we have developed procedures to convert a series of readily available sterically hindered nicotinonitriles into the corresponding 2,4-disubstituted 3-(1,2,4-oxadiazol-3-yl)-, 3-(thiazol-2-yl)-and 3-(imidazol-2-yl)pyridines.These compounds together with previously described 3-(5-tetrazolyl)pyridines 27 and 3-(1,3,4-oxadiazol-2-yl)pyridines 27 could constitute a basis for a potential combinatorial library.This entire library containing more than 200 compounds was tested by the computer software PASS (Prediction of Activity Spectra for Substances). 47This program illustrates the predicted activity spectrum of a compound as probability of activity (P a ) and probability of inactivity (P i ).For example, it was predicted by the PASS that 3-(1,2,4-oxadiazol-3yl)pyridines 16 can possess an antidepressant property, and anxiolytic activity for five of the tested six examples with P a more than 70%.Our further investigations will be directed to the synthesis of the compounds with the potential activity predicted, and then to a pharmacological evaluation of the latter.

Experimental Section
General Procedures.Starting nicotinonitriles 1a-i were prepared according to literature procedure. 48All the acid chlorides, α-haloketones and other reagents are commercially available.The reactions were monitored by TLC (aluminium sheets, silica gel 60 F 254 , Merck).Merck Kieselgel 60 (230-400 mesh) was used for a column chromatography. 1 H NMR spectra were recorded on a Bruker Avance DRX 400 (400.13MHz) spectrometer equipped with a 5 mm inverse multinuclear gradient probehead in DMSO-d 6 or CDCl 3 , and the chemical shifts (δ) are given in ppm relative to the signal for TMS as internal standard. 13C NMR spectrum was recorded on the same instrument at 100.61 MHz using DMSO-d 6 as solvent.The assignments of signals in  24) and CEM Discover microwave labstations (both operating at 2450 MHz under continuous internal temperature control) were used for the experimental and scale-up reactions.

2-Methyl-N,4-diphenylpyridine-3-carboximidamide hydrochloride (17a).
2M solution of trimethylaluminum in toluene (10 mL, 20 mmol) was slowly added dropwise to solution of aniline (1.4 g, 15 mmol) in anhydrous dichloroethane (10 mL) under stirring at 0 o C in inert atmosphere (argon).The mixture was stirred at room temperature for 30 min then solution of 2-methyl-4phenylnicotinonitrile 1d (1.94 g, 10 mmol) in anhydrous dichloroethane (10 mL) was added under stirring.The reaction mixture formed was subjected to microwave irradiation at 130 o C for 3 h then cooled to room temperature and carefully treated with methanol (100 mL).The jelly-like precipitate was filtered off through Celite and washed with methanol.The filtrate was concentrated under reduced pressure, and the residue was chromatographed (silica gel, chloroform/methanol in gradient 0 to 30%) to give the viscous unstable oil that was identified as hydrochloride.For that, the oily compound was dissolved in anhydrous dioxane (15 mL), and 4M solution of HCl in dioxane (7.5 mL, 3 eq.) was added under stirring.The precipitate formed was filtered off, washed with anhydrous ether and dried in vacuum at room temperature to afford the titled compound 17a (

Table 1 .
Conditions and yields of amidoximes 13 and amides 14

Table 3 .
Conditions and yields of thioamides 23a-i

pyridines 25a-m
a Yields from the starting thioamides 23.