3022
B. Pelcman et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3017–3023
using the convenient reaction between 2,4-dioxovaleric acid ethyl
ester and hydrazine hydrate.39 Dimerization with SOCl2, followed
by the reaction with 2-chloro-4-fluoroaniline as before, gives the
the IC50 of thioureas, for example, 8, is low, the compounds could
not completely inhibit the enzyme (the Imax did not approach
100%). Phenylsulfonyl substituted compounds, for example, 9a,
have substantially lower activity and also do not inhibit the
enzyme completely. The sulfamide 9b as well as the 1N-benzyl
(10a) and 1N-methyl (10b) substituted pyrazoles lacks activity
completely. Finally, methylation of the nitrogen of the anilide in
the pyrazole 3-position (e.g., 7h) gave compounds devoid of inhibi-
tory activity.
In conclusion we have prepared several 1N-substituted pyra-
zole-3-carboxanilides that are good inhibitors of the human 15-
LOX-1. The best compounds are found among the ureas, in par-
ticular 7a and 7g, but they slowly decompose on standing in
solutions containing traces of water or other nucleophiles, which
makes them unsuitable for further drug development. However,
the 1N-unsubstituted pyrazole-3-carboxanilides 5 are active 15-
LOX-1 inhibitors and these were pushed forward and developed
as clinical candidates, which is discussed in detail in a following
Letter.30
desired
5-methylpyrazole-3-carboxanilide
5c.
5-tert-
Butylpyrazoles are obtained in a similar fashion.40
Both the 4- and 5-alkyl substituted pyrazole-3-carboxylic acids
are conveniently transformed to the anilides employing the dike-
topiperazine approach depicted above in Scheme 2. Again, treat-
ment with arylisocyanates gives the corresponding ureas. One
exception is the 5-tert-butylpyrazolecarboxanilide 5d, which on
treatment with hexylisocyanate expels the aniline portion of the
molecule to give the imidazopyrazoledione 12 (Scheme 4). Here
the bulky tert-butyl substituent forces the acylation to take place
on the 2-nitrogen of the pyrazole. Only minor amounts of 2-substi-
tuted pyrazole-3-carboxanilides could sometimes be isolated on
acylation of 5 with benzoyl chlorides and on alkylations.
Early on in the program we realized that chemical instability of
the pyrazole N-acyl bond might be an issue. Actually, some tar-
geted compounds could not be isolated, presumably due to hydrol-
ysis during aqueous workup. Judged by recording NMR spectra of
compounds upon standing in DMSO-d6, the order of stability
seemed to be alkyl ꢀ phenylsulfonyl > phenylurea P benzoyl. It
was also seen that alkylureas are more stable than arylureas and
that 5-substitution of the pyrazole core decreased stability. In con-
trast, N-sulfonyl substituted pyrazoles were stable DMSO-d6 but
were slowly methanolysed upon prolonged standing in MeOD
(Fig. 3). It must be emphasized that the isolated compounds were
stable enough to be evaluated as 15-LOX-1 inhibitors in our assay
systems. This is particularly important as also intermediates 5 pos-
sess considerably 15-LOX inhibitory activity.30
Before analyzing the results and the structure–activity relation-
ships (SAR) it is important to discuss the assays used and the inter-
pretation of the data in Table 1. In the beginning of the project
various assay conditions were evaluated in our primary assay:
enzyme or cell based assays, LA or AA as a substrate. IC50’s were
only determined for the most interesting compounds. As the work
progressed and we became more confident in how to run assays
and interpret data, we normally recorded the IC50’s for all com-
pounds both in an enzyme and a cellular assay using AA as a sub-
strate for both. It has also been pointed out that Imax, that is, the
extent of how much of the activity of 15-LOX-1 that can be inhib-
ited (where an Imax of 100% means complete inhibition of the enzy-
matic activity) varied for different compounds. Usually,
compounds with low IC50’s also had an Imax approaching 100%,
but for weaker inhibitors the Imax could be substantially lower.
Eventually, robust enzyme and cellular assays were developed,
which were also amenable to HTS.41 Thus, the data presented in
Table 1 must be interpreted with great care.
Acknowledgment
This work was financially supported by Biolipox AB.
References and notes
Table 1 shows a selection of compounds that illustrate the SAR
of 1N-substitution of the pyrazole-3-carboxanilides. The influence
of the substitution profile of the anilide portion on the SAR will be
discussed in a following Letter30 but 2-chloro-4-fluoroanilides is
one of the preferred. Using 1a as a benchmark one can deduce that
insertion of a methylene between the carbonyl group and the phe-
nyl group—benzoyl 1a versus phenacetyl 1b—completely abolishes
activity, but that the simple N-acetyl compound 1c show consider-
able inhibition. Arylcarbamates, here illustrated by 6, have an
activity comparable to the N-benzoyl substituted compounds, but
here there is a large difference depending on the substrate used
in the assay; when LA was used it showed only minute activity
whereas it showed good activity with AA. The discrepancy is hard
to explain but was not investigated. The best inhibitors were found
among the ureas, for example, 7a–c. Here the phenyl group can be
replaced with an alkyl group (7b), which may also carry function-
alities, for example, esters as in 7c. In contrast, the tertiary urea 7d
only showed low inhibitory activity. Although the absolute value of
16. Connor, D. T.; Roark, W. H.; Sorensen, R. J. WO 01/96299, 2001.
17. Barvian, N. C.; O’Brian, P. M.; Patt, W. C.; Picard, J. A.; Sliskovic, D. R. WO 01/
96298 A2, 2001.
18. Connor, D. T.; Roark, W. H.; Sexton, K.; Sorenson, R. J. WO 99/32433, 1999.
19. Connor, D. T.; Glease, S. S.; Purchase, T. S.; Roth, B. D.; Trivedi, B. K. WO 98/
06703, 1998.
21. Trivedi, B. K.; Roth, B. D.; Padia, J. K. WO 97/12615, 1997.
22. Cornicelli, J. A.; Padia, J. K.; Tait, B. D.; Trivedi, B. K. WO 97/12613, 1997.
25. Weinstein, D. S.; Langevine, C. M. WO 2006/034391, 2006.