Novel lead structures for p38 MAP kinase via FieldScreen virtual screening

Article abstract of DOI:10.1021/jm801399r

p38 MAP kinase has received considerable interest in the pharmaceutical industry and remains a valid and interesting target for the treatment of inflammation. To discover novel p38 inhibitors, we applied the ligand-based virtual screening technique, FieldScreen, to 1.2 million commercially available compounds. Fifty-eight diverse compounds were selected for biological analysis, using molecular field similarity to known inhibitors, while explicitly removing any structure that shared a scaffold with previously reported p38 inhibitors. Of these, 11 (19%) showed ≥20% inhibition of p38 at 10 μM. We chose to prepare analogues of two distinct chemical series resulting in a potential lead compound with pIC₅₀ of 6.4. Modeling of SAR using FieldAlign, a ligand alignment protocol, was used to rationalize the SAR of the series of thiadiazole based inhibitors.

Full text of DOI:10.1021/jm801399r

4200 J. Med. Chem. 2009, 52, 4200–4209
DOI: 10.1021/jm801399r
Novel Lead Structures for p38 MAP Kinase via FieldScreen Virtual Screening
Timothy J. Cheeseright,† Melanie Holm,‡ Frank Lehmann,‡ Sabine Luik,‡ Marcia Gottert,‡
James L. Melville,† and Stefan Laufer*,‡
†Cresset BioMolecular Discovery Ltd., BioPark Hertfordshire, Welwyn Garden City, Hertfordshire AL7 3AX, U.K., and ‡Department of
Pharmaceutical and Medicinal Chemistry, Institute of Pharmacy, Eberhard-Karls-University Tuebingen, Auf der Morgenstelle 8, D-72076
Tuebingen, Germany
Received November 6, 2008
p38 MAP kinase has received considerable interest in the pharmaceutical industry and remains a valid
and interesting target for the treatment of inflammation. To discover novel p38 inhibitors, we applied
the ligand-based virtual screening technique, FieldScreen, to 1.2 million commercially available
compounds. Fifty-eight diverse compounds were selected for biological analysis, using molecular field
similarity to known inhibitors, while explicitly removing any structure that shared a scaffold with
previously reported p38 inhibitors. Of these, 11 (19%) showed g20% inhibition of p38 at 10 μM. We
chose to prepare analogues of two distinct chemical series resulting in a potential lead compound with
pIC50 of 6.4. Modeling of SAR using FieldAlign, a ligand alignment protocol, was used to rationalize
the SAR of the series of thiadiazole based inhibitors.
Introduction
beyond early stage clinical trials. Hence, there remains a great
need for new lead structures for p38 MAPK.
Inflammation is a fundamental physiological process that is
essential for survival but at the same time is one of the major
causes of human morbidity and mortality. A large number of
diseases have their seeds in the complex process of inflamma-
tion and specifically in an overactive immune response, the
commonest of which include rheumatoid arthritis, psoriasis,
multiple sclerosis, and inflammatory bowel diseases.1
Virtual screening is now an established method for hit
finding within the pharmaceutical industry.5 p38 inhibitors
are popular choices for data sets for the retrospective valida-
tion of many virtual screening methods, including de novo
design,6 free energy calculations,7 ligand-based similarity
searching,8 docking,9 and QSAR.10 Additionally, molecular
dynamics11 and docking12 have been employed to rationalize
the binding mode of known inhibitors. However, there are
relatively few examples of the use of in silico methods for the
prospective screening of p38 inhibitors. A notable example is
provided by Jhoti and co-workers,13 who used docking to
construct target-specific fragment-based libraries, including
one focused toward p38. Recently, FieldScreen,14 a ligand-
based similarity searching method using the concept of mole-
cular fields,15 has proven itself valuable for the task of scaffold
hopping and virtual screening.16 Therefore, we sought to
apply this method for finding novel inhibitors of p38.
The mitogen-activated protein kinase (MAPKa) p38 is one
of the most extensively studied kinase target partly because of
its important role as a key enzyme in the production of pro-
inflammatory cytokines such as TNF-R and IL-1β. Of the
four different isoforms of the p38 family (R, β, γ, and δ), p38R
is ubiquitously expressed and is generally considered the most
important isoform in the inflammatory signal transduction
pathway and an appropriate target for anti-inflammatory
therapy.
The initial discovery in 1994 of p38 as the molecular target
for a novel class of cytokine suppressive inhibitors, exempli-
fied by the prototypical inhibitor 1 (SB203580),2 catalyzed the
search for potent, selective, efficacious, and safe p38 inhibi-
tors. These inhibitors show diversity not only in chemical
structure but also in how they interact with the protein.
To date, the well-known inhibitors can be divided into six
classes (Figure 1): (1) pyridinyl- and pyrimidinylimidazoles
and related structures, (2) bicyclic 6,6-heterocycles and
related structures, (3) N,N0-diarylureas and related structures,
(4) substituted benzamides, (5) diaryl ketones, and (6) indole
amides.3,4 However, few of these compounds have progressed
Design Process
As a similarity searching method, FieldScreen requires
an active ligand to use as a search query, ideally in a
bioactive conformation. We therefore reviewed the many
ligands that have been cocrystallized with p38. We desired
to mimic the closed form of the protein and hence immedi-
ately discounted ligands bound to alternative protein con-
formations. Of the remaining ligands, we gave preference to
highly active druglike ligands and to ligands that occupied
more of the active site. Using these criteria together with a
visual inspection of the available ligands (ensuring that
different chemotypes were chosen), we selected the ligands
present in PDB codes 1m7q, 1yw2, and 1ouk. Each of the
ligands present in these protein-ligand cocrystal structures
was extracted and examined.
*To whom correspondence should be addressed. Phone: 0049-7071-
2978788. Fax: 0049-7071-5961. E-mail: stefan.laufer@uni-tuebingen.
de.
aAbbreviations: MAPK, mitogen-activated protein kinase; SAR, structure-
activity relationship; FPP, field point pattern; ATF-2, activating transcription
factor-2.
r
pubs.acs.org/jmc
Published on Web 06/02/2009
2009 American Chemical Society

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4201
the highly polar piperidine was truncated to a methyl group to
emphasize the hinge region over the sugar and phosphate
binding pockets.
Taken together, the three chosen search queries present
a diverse set of field point patterns to the protein. As such, it
was envisaged that the three search queries would return
markedly different hit lists.
In the processing of the results of the FieldScreen experi-
ments it was decided to treat each of the three search queries
identically and therefore to progress a fixed number of results
from each hit list. Initially 250 results were taken from each
result list with the intention of selecting 50-100 molecules for
biological investigation. To aid the selection of molecules for
purchase, the three hit lists were combined. Only four com-
pounds were found to occur in more than one hit list. The
remaining 746 compounds were clustered based on their 2D
topologies. The resulting 96 clusters and 311 singletons were
visually inspected in 2D and as 3D alignments relative to the
original search query. In visually inspecting the 407 com-
pounds, we applied the following criteria: reject if a similar
scaffold was known as a p38 inhibitor; reject if the 3D
alignment against the search query did not interact with the
hinge region; reject if, in the opinion of our medicinal
chemists, the database molecule appeared undruglike (e.g.,
excessive numbers of divalent sulfur atoms or nitro groups).
After visual inspection, 58 clusters were identified as po-
tential p38 inhibitors. To accelerate biological analysis, it was
decided to screen these clusters by choosing a representative
molecule from each cluster. The molecule that was closest to
the cluster center according to our 2D method was chosen for
ordering and biological analysis.
Figure 1. Overview of the different chemical classes of p38 inhibitors.
The ligand in PDB code 1m7q is a single digit nanomolar
inhibitor of p38.17 The ligand makes extensive contacts across
the hinge region and into the specificity pocket. However, it
also contains a piperazine moiety that is oriented toward the
sugar and phosphate binding pockets. This group signifi-
cantly changes the electrostatic field derived from the ligand,
and so it was decided to truncate this part of the ligand to
a simple dimethylamine (see Figure 2). This truncation pro-
motes virtual screening hits that interact with the hinge
binding region of p38 by increasing the proportion of the
field descriptors derived from this important region relative to
the size of whole field point pattern (FPP).
The ligand from PDB code 1yw2, like that from 1m7q,
binds to a protein conformation where the backbone amide
bond between Met109 and Gly110 is orientated to position the
NH close to the ligand and available for H-bonding. In this
respect both ligands present a strong negative field to the
protein and consequently give large negative field points close
tothe carbonylin1m7qandthe pyrimidine in 1yw2. However,
the ligand from 1yw2 lacks the explicit N-H of 1m7q,
presenting an aromatic hydrogen in its place giving a smaller
positive field in the hinge region. The major feature of the
1yw2 ligand is the strong negative field points associated with
the isoxalone moiety which interacts strongly with Lys53 in the
protein ligand cocrystal, a feature that could be important in
p38 inhibitors.
The final ligand that was chosen as a search query in our
virtual screening strategy was that from PDB code 1ouk. In
contrast to the other chosen inhibitors this ligand binds to a
subtly different closed conformation of the protein. In the
1ouk protein-ligand cocrystal the amide bond between
Met109 and Gly110 is oriented such that the carbonyl of
Met109 is directed toward and interacts with the ligand.
Consequently the field pattern of the ligand is more positive
(as would be expected from an aminopyrimidine) and pro-
vides a good complement to the other selected search queries.
In a similar manner to that described for the ligand of 1m7q,
Results
A selection of the compounds chosen for biological analysis
is shown in Table 1 together with the inhibition obtained in an
enzyme assay with ATF-2 as substrate and ATP as cosub-
strate and with the inhibitor at a concentration of 10 μM.18 Of
the 58compoundsanalyzed(the fulllist of compoundsis given
in the Supporting Information), 11 showed 20% or greater
inhibition of p38at10μM(a19%hitrate). Thiswasextremely
encouraging given that we had specifically excluded com-
pounds with obvious kinase binding motifs or that had been
reported as p38 inhibitors previously.
To ensure that these results could not have been achieved
using a simpler method, we used a 2D atom pairs similarity
metric to screen the complete database of 1.2 million com-
pounds against each of our three search queries.19 For each of
our search queries we constructed a results list consisting of
the similarity rankings for each database molecule to the
search query. By use of this method, the highest ranked com-
pound of our chosen 58 was at positions 6600 (1yw2 based
query), 20 400 (1m7q based query), and 26300 (1ouk based
query), well outside the top 250 we chose using FieldScreen.
We followed up the initial hits with a thorough patent and
literature search around each scaffold. On the basis of this
search and the initial activity that was found, we chose 3-(2-
chlorophenyl)-6-((4-methoxyphenoxy)methyl)[1,2,4]tria-
zolo[3,4-b][1,3,4]thiadiazole (c15) and 6-amino-1-benzyl-4-
(4-bromophenyl)-3-methyl-1,4-dihydropyrano[2,3-c]pyrazole-
5-carbonitrile (c38) for hit to lead investigations.
A small library of derivatives for the pyranopyrazole hit
(c38) was prepared starting from different substituted pyra-
zol-5(4H)-ones.20 Unfortunately this library showed only

4202 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Cheeseright et al.
Figure 2. Ligands from PDB codes 1m7q (top), 1yw2 (middle), 1ouk (bottom): ligand as present in PDB file showing selected protein residues
and H-bonds (green lines); ligand used as search query including field point patterns. Key to field point colors are as follows: blue = negative;
red=positive; orange=hydrophobic; yellow=surface.
moderate SAR across more than 20 analogues, and hence, our
attention turned to the thiadiazole series.
3-position. Incorporation of an o-substituted phenyl residue
gave significant and consistent increase in inhibition, suggesting
that a twisted conformation is strongly preferred in this region.
Structural change to m-, p-substitution or two residues in
o- and p-position gave a reduction in potency as did extending
from a phenyl to a benzyl substituent. In the same way, a
switch to a furyl residue or other five- and six-membered
heterocycles, respectively, resulted in loss of inhibition po-
tency. Substitution in the 6 position gave significant changes
in activity. Moving the para methoxy substituent to the meta
position led to a sharp drop in activity, whereas removal of the
p-methoxy substiuent of c15 led to a small increase in activity.
Swapping the ether linker to the equivalent alkane gave little
change in activity, whereas truncation of the p-methoxylphe-
noxymethyl substituent to a p-methoxybenzyl group led to
a gain in activity.
To further understand our SAR, we chose to carry out a
field-based alignment of the thiadiazole analogues to our
original search queries, using the program FieldAlign. In this
procedure the three search queries were loaded into FieldA-
lign as a single reference molecule. Next, each member of
the pyrazole series was conformationally explored and each
conformation of each molecule was aligned to the combined
We pursued SAR of c15 through a strategy of purchase and
synthesis. Where useful analogues were commercially avail-
able we purchased them (Figure 1 in Supporting Informa-
tion). However, not all the analogues that we wished to screen
were commercially available, and hence, we embarked on a
synthesis of key analogues. The synthesis of derivatives of c15
was achieved by preparing a range of triazole intermediates
that were reacted with carboxylic acids in refluxing phos-
phorus oxychloride.21 The triazole intermediates were pre-
pared directly from appropriate carboxylic acids by warming
a mixture with thiocarbohydrazide to the melting tempera-
ture.22 Where this approach was unsuccessful the triazole was
formed by reacting suitable carboxylic acids with hydrazine
and treating the resulting hydrazides with carbon disulfide
(Scheme 1) followed by a further reaction with hydrazine to
furnish the desired triazole intermediate.23
Thirty-two 3,6-disubstituted [1,2,4]triazolo[3,4-b][1,3,4]-
thiadiazoles were tested with inhibitions up to 83% (IC50 =
0.44 μM) (Table 2 and Supporting Information Figure 1) and
distinct SAR can be summarized as follows (Figure 3). Essen-
tial for inhibitory activity is a bulky lipophilic residue in the

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4203
Table 1. Inhibitors Discovered by Screening
a Inhibition at 10 μM. b Number of determinations was 3.
reference. For an alignment to score well, the conformation
under study must align well to each component of the
reference molecule in a single orientation. Using a multiple
molecule reference in this manner gives a better description of
the requirements of the protein for the binding of ligands.
Gratifyingly a single, consistent binding mode was suggested
by these experiments (Figure 4). Our modeling suggested that
the R1 substituent binds to the hydrophobic pocket close to
Met109 and Gly110, the nitrogens of the diazole interact with
the hinge region, and the R2 substituent occupies the kinase
selectivity pocket (Figure 4). This binding mode is consistent
with the observed SAR.

4204 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Cheeseright et al.
Scheme 1. Synthesis of 3,6-Disubstituted Triazolo[3,4-b][1,3,4]
thiadiazoles (17-48)a
In a FieldPrint comparison, each pair of field points in a
molecule is recorded, along with the distance, type, and the
intensity of their interaction (represented as the product of the
size of the field points). The distances are binned (with a spacing
˚
of 1.5 A), and the most intense interaction of a given type and
distance range is stored. This produces a field “fingerprint” for
each molecule, which is used to compare two molecules without
requiring an alignment step. The top 40% scoring molecules
were carried through to FieldClique screening. FieldClique
comparisons identify sets of field points that are common to
both molecules via a clique perception algorithm. These field
points are then used to align the two molecules. The alignment is
scored by sampling the field of the query conformation at the
positions defined by the field points of the database conforma-
tion. This process is then reversed so that the field of the
database conformation is sampled by the field points of the
query conformation. The extents to which the field points of
one conformation sample a similar field in the other conforma-
tion are combined to produce the final similarity value for the
two conformations. The final stage in the FieldScreen process
was to optimize the alignment of the top scoring molecules from
FieldClique using a multidimensional simplex minimization of
the rigid conformations.15 This FieldSimplex routine gives the
most accurate measure of field similarity for any molecule
within the FieldScreen system. Additionally, an excluded vo-
lume can be defined surrounding a search query such that any
atom in any alignment that enters the excluded volume causes a
penalty to be added to the scoring function. The excluded
volume was defined using the conformation of the protein in
the appropriate protein-ligand crystal structure. The resulting
hit list was clustered with the group average linkage method,26
using topological atom pairs as descriptors,19,27 and a Dice
similarity coefficient threshold of 0.7 for merging two clusters.
Ligand preparation was performed by alignment and super-
position of the proteins in PDB codes 1m7q, 1ouk, and 1yw2
(manual alignment of equivalent residues, superposition of
equivalent C-R atoms using Accelrys DS Visualizer 1.6) fol-
lowed by correction of the bonding pattern of each of the
ligands. Analysis of each protein ligand crystal structure led to
the truncation of the ligands present in structures 1m7q and
1ouk as described above. Finally the modified ligands were
exported for use in FieldScreen in sdf format.
Molecular Alignment. Field based 3D molecular alignment
was carried out using FieldAlign, version 2.0.1,28 in an analo-
gous manner to that described for the FieldSimplex routine of
FieldScreen above. However, multiple prealigned single con-
formations were used as the reference in the alignment. The
reference was constructed by importing each of the ligands that
were prepared as FieldScreen search queries into a single
FieldAlign project. Molecules to be aligned (database mole-
cules) were either pasted directly from a relevant drawing
package or read in from an sdf file. Each database molecule
was conformationally explored within FieldAlign (keeping a
maximum of 100 conformations) before the addition of mole-
cular field points and alignment to the multimolecule template.
FieldAlign returns the best alignment based on the mean of the
scores to the individual members of the reference in a given
orientation.
a Reagents and conditions: (i) thiocarbohydrazide, “ball-tube oven”
(“Kugelrohr apparatus”), melting temp or (1) H2SO4, EtOH, 8 h, reflux,
(2) N2H4, ethanol, 6 h, reflux, (3) CS2, EtOH, 12 h, room temp, (4) N2H4,
MeOH, 6 h, room temp; (ii) R2-COOH, POCl3, 6 h, reflux.
Conclusion
p38 MAP kinase remains a valid and interesting target for
therapeutic intervention in inflammatory diseases. Using a
novel ligand based virtual screening method, we searched
commercially available compounds for novel actives. The
crude results list was found to be rich in chemotypes that were
known to be inhibitors of p38, validating the choice of search
queries and the search method. However, the known chemo-
types were now removed from the hit list before further
selection of compounds to purchase. From this new list we
chose just 58 compounds for biological analysis, representing
just 0.003% of the 1.2 million compounds originally searched.
Pleasingly, we found an excellent hit rate even among this
highly selected data set. Thissurprisingly positiveresult served
to justify our choice of a ligand based method in this project
where conventional practice would have employed a structure
based approach. Furthermore, there have been few examples
of molecular field based technologies being used in virtual
screening in either a retrospective or prospective manner. A
recent communication on the application of FieldScreen to a
retrospective data set suggested that it would perform poorly
for p38 kinase.14 Clearly the results of this application of
FieldScreen are above expectations.
Wechose to followup compounds foundto beactivein bio-
logical analysis with a dual strategy of purchase and synthesis.
Compound purchase was very attractive, as it provided rapid
SAR, but it proved difficult to get a full set of rationally
designed compounds. For this reason we also chose, and have
described here, the synthesis of small libraries of compounds
around two of the most active hits. Additionally, field based
alignment of analogues was used to facilitate the interpreta-
tion of our biological data and enabled us to propose a
binding mode for one of these series. We anticipate that the
binding models will greatly aid our future elaboration of the
3,6-disubstituted triazolo[3,4-b][1,3,4]thiadiazoles series.
Experimental Section
FieldScreen Virtual Screening. Field-based 3D similarity
searching was carried out using FieldScreen.14 In FieldScreen,
each molecule is described as a Field Point Pattern (FPP),15
using the local extrema of four molecular fields: positive elec-
trostatic, negative electrostatic, surface (van der Waals interac-
tions), and hydrophobic. The resulting FPPs are aligned and
scored to give a measure of similarity between a query molecule
and a database molecule in specific conformations. For the
query molecule, conformations were taken from the coordinates
of the X-ray structures. For the database molecules, 50 con-
formations were generated per molecule using the XedeX25
conformation hunter, and each conformation was scored sepa-
rately to the query molecule. The similarity of the best scoring
conformation for each database molecule was reported. The
comparison of field point patterns can be performed at three
levels of accuracy: FieldPrint, FieldClique, and FieldSimplex.
Biological Testing. According to the developed method,18 test
compounds were tested in concentrations ranging from 10-5 to
10-8 M. Pyridinylimidazole 1 was used as a reference compound,
and the optimized ATP concentration at which the test was
performed was 100 μM. Briefly the assay involves the immobili-
zation of the kinase substrate ATF-2 (activating transcription
factor-2) on microtiter plates, addition of the kinase reaction
mixture, and measurement of substrate phosphorylation using a
two-step antigen-antibody reaction in which the primary anti-
body binds to the double (Thr69 and Thr71) phosphorylated
ATF-2 and acts as antigen for the secondary antibody. Secondary
antibody is conjugated with alkaline phosphatase which is able,

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4205
Table 2. Biological Activity of Synthesized 3,6-Disubstituted Triazolo[3,4-b][1,3,4]thiadiazolesc
a Inhibition at 10 μM. b Number of determinations was 3. c The asterisk (/) indicates purchased 3,6-disubstituted [1,2,4]triazolo[3,4-b][1,3,4]-
thiadiazoles.

4206 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Cheeseright et al.
in the last step, to dephosphorylate 4-nitrophenolphosphate
disodium salt (4-NPP), 4-nitrophenol being the detected species
by an ELISA reader at 405 nm.
Merck silica gel 60 (0.015-0.040 mm) or RP18 columns. The
progress of the reactions was monitored by thin-layer chroma-
tography (TLC) performed with Merck silica gel 60 F-245
plates. Where necessary, reactions were carried out in a nitrogen
Chemistry. Infrared spectra were recorded on a “Perkin-Elmer
Spectrum One” infrared spectrophotometer. 1H (200 MHz,
digital resolution 0.3768 Hz) and 13C (50 MHz, digital resolu-
tion 1.1299 Hz) NMR data were recorded on a Bruker AC 200.
The data are reported as follows: chemical shift in ppm from
Me4Si as external standard, multiplicity, and coupling constant
(Hz). GC-MS was performed on a HP6890 series system. EI
mass spectra were recorded on a Varian MAT 311A (70 eV) and
FD mass spectra on a MAT-95 (Finnigan). LC/MS (ESI) was
performed on a Thermo Finnigan Surveryer system with Ther-
mo Finnigan TSQ Quantum triple quadrupole mass spectro-
meter. HPLC was carried out on Merck Hitachi L-6200A
intelligent pump, Merck Hitachi AS-2000A autosampler, and
Merck Hitachi L-4250 UV-vis detector. HRMS (EI) (electron
impact high resolution mass spectrometry) was executed at the
Abteilung fur Massenspektrometrie des Instituts fur Organische
Chemie der Universitat Tubingen. For clarity, only the highest
measured signal is given for FD mass spectra. Buchi melting
point B-545 apparatus was used, and melting points are un-
corrected. Where appropriate, column chromatography was
performed for crude precursors with Merck silica gel 60
(0.063-0.200 mm). Column chromatography for test com-
pounds was performed using a La-Flash system (VWR) with
˚
atmosphere using 4 A molecular sieves. All reagents and sol-
vents were obtained from commercial sources and used as
received. Reagents were purchased from Sigma-Aldrich Che-
mie, Steinheim, Germany; from VWR, Darmstadt, Germany;
or from Acros, Geel, Belgium. The β-keto esters, 1H-pyrazol-5
(4H)-ones, and dihydropyrano[2,3-c]pyrazoles (1-30) were re-
cently published.20
General Method A for 5-Substituted [1,2,4]Triazoles. Equi-
molar amounts of the appropriate acid and thiocarbohydrazide
were carefully heated up to the melting point (160-180 °C) in a
“ball-tube oven” (“Kugelrohr apparatus”) under slow rotation
until the liquefied material got solid again (20-30 min). The
solidified smelter was dissolved in the required amount of
ethanol and optionally up-concentrated until the product pre-
cipitated.
4-Amino-5-phenyl-4H-1,2,4-triazole-3-thiol (7). Yield: 61%.
1H NMR (200 MHz, DMSO-d6): δ [ppm] = 5.80 (s, 2H, -NH2),
7.51-8.17 (m, 5H, 20,30,40,50,60-H), 13.93 (s, 1H, -SH). GC-
MS (EI): 10.16 min, m/z [%] = 193.0 (12), 192.0 ([M*], 100),
121.0 (20), 104.0 (23), 77.0 (18). IR: 3072 (w), 2832 (w), 2666 (w),
2555 (w), 1680 (s), 1601 (m), 1583 (m), 1535 (w), 1497 (w), 1453
(m), 1421 (m), 1324 (m), 1289 (s), 1186 (m), 1128 (m), 1101 (w),
1073 (m), 1027 (m), 1000 (w), 932 (s), 805 (m), 704 (s), 684 (m),
667 (s). Mp: 227.1 °C.
4-Amino-5-(2-fluorophenyl)-4H-1,2,4-triazole-3-thiol (8). Yield:
58%. 1H NMR (200 MHz, DMSO-d6): δ [ppm] = 5.64 (s, 2H,
-NH2), 7.51-7.99 (m, 5H, 20,30,40,50,60-H), 14.03 (s, 1H, -SH).
GC-MS (EI): 11.93 min, m/z [%] = 211.0 (12), 210.0 ([M*],
100), 139.0 (28), 122.0 (28), 121.0 (25), 95.0 (16), 60.0 (11). IR:
3310 (w), 1625 (w), 1451 (m), 1318 (w), 1237 (m), 1055 (w), 1003
(w), 950 (m), 822 (m), 765 (s), 726 (m), 681 (m), 629 (m), 602 (m),
535 (m), 476 (w), 440 (w). Mp: 181.6 °C.
4-Amino-5-benzyl-4H-1,2,4-triazole-3-thiol (9). Yield: 67%.
1H NMR(200 MHz, DMSO-d6):δ [ppm]=4.02(s, 2H, -CH2-),
5.80 (s, 2H, -NH2), 7.32-7.23 (m, 5H, 20,30,40,50,60-H), 13.53
Figure 3. Analysis of SAR for thiadiazole series.
Figure 4. Proposed binding mode of the thiadiazole series, obtained by 3D molecular alignment of compound c15 to the three prealigned
ligands of PDB codes 1m7q, 1ouk, and 1yw2. Shown are selected residues from 1m7q and compound c15 (magenta) and in (a) the native ligand
of 1m7q (cyan), (c) modeled orientation of compound c15 in the active site of p38 from PDB code 1m7q relative to the native ligand (d) showing
potential H-bond interactions to the hinge region.15 (a) and (b) show a solvent accessible surface of protein 1m7q as calculated in DS Visualizer
2.0 using a probe radius of 1.4.24

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4207
(s, 1H, -SH). GC-MS (EI): 13.35 min, m/z [%] = 207.1 (13),
206.1 ([M*], 100), 190.0 (17), 117.1 (10), 93.1 (11), 91.1 (36).
4-Amino-5-(2-chlorobenzyl)-4H-1,2,4-triazole-3-thiol (10). Yield:
66%. 1H NMR (200 MHz, DMSO-d6): δ [ppm] = 4.13 (s, 2H,
-CH2-), 5.59 (s, 2H, -NH2), 7.20-7.55 (m, 5H, 20,30,40,50,60-
H), 13.53 (s, 1H, -SH). GC-MS (EI): 13.04 min, m/z [%] =
242.1 (32), 241.1 (10), 240.1 ([M*], 79), 205.1 (54), 188.1 (63),
127.0 (40), 126.0 (11), 125.0 (100), 116.1 (48), 89.1 (82), 63.1 (43),
60.0 (71), 59.0 (49). IR: 3147 (w), 1594 (w), 1475 (m), 1419 (m),
1306 (m), 1054 (m), 1037 (m), 1012 (m), 979 (m), 838 (w), 802
(w), 813 (s), 751 (s), 682 (m), 662 (m), 641 (m), 554 (m), 464 (w),
445 (w). Mp: 204.8 °C.
4-Amino-5-phenethyl-4H-1,2,4-triazole-3-thiol (11). Yield:
63%. 1H NMR (200 MHz, DMSO-d6): δ [ppm] = 2.94 (s, 4H,
-CH2-CH2-), 5.61 (s, 2H, -NH2), 7.13-7.35 (m, 5H,
40,50,60,70,80-H), 13.44 (s, 1H, -SH). GC-MS (EI): 12.75 min,
m/z [%] = 221.0 (14), 220.0 ([M*], 100), 91.0 (100), 65.1 (16),
60.0 (10). IR: 3140 (w), 2933 (w), 1636 (w), 1571 (m), 1481 (m),
1420 (m), 1309 (m), 1146 (w), 1090 (w), 1045 (m), 971 (m), 748
(s), 699 (s), 655 (m), 580 (m), 531 (m), 498 (m), 473 (m). Mp:
227.9 °C.
4-Amino-5-(pyridin-4-yl)-4H-1,2,4-triazole-3-thiol (16). Yield:
72%. 1H NMR (200 MHz, DMSO-d6): δ [ppm] = 5.86 (s, 2H,
3
-NH2), 8.00-8.03 (d, 2H, J = 6.13 Hz, 20,60-H), 8.74-8.77
(d, 2H, 3J = 6.13 Hz, 30,50-H), 13.89 (s, 1H, -SH). IR: 3158 (w),
2450 (w), 1812 (w), 1606 (m), 1572 (w), 1517 (w), 1449 (w), 1415
(w), 1315 (m), 1217 (w), 1086 (w), 1064 (w), 1038 (w), 1004 (m),
941 (s), 824 (s), 736 (w), 709 (w), 688 (m), 6001 (s), 528 (m), 498
(m). Mp: 235.4 °C.
General Method for 3,6-Disubstituted [1,2,4]Triazolo[3,4-b]
[1,3,4]thiadiazoles. A mixture of the selected triazole (1.00 equiv),
the corresponding carboxylic acid (2.00 equiv), and phosphorus
oxychloride (22.0 equiv) was refluxed for about 6 h. The reaction
mixture was poured on crushed ice. The solid thus separated was
filtered and treated with NaOH solution (aqueous 10%) to
remove the unreacted material. The solid residue was filtered,
washed with water, and crystallized from ethanol.
6-((3-Methoxyphenoxy)methyl)-3-phenyl[1,2,4]triazolo[3,4-b]
[1,3,4]thiadiazole (17). Yield: 74%. 1H NMR (200 MHz, DMSO-
d6): δ [ppm]=3.76 (s, 3H, -CH3), 5.62 (s, 2H, -CH2-), 6.72-
7.21 (m, 4H, 400,500,600,800-H), 7.64 -8.25 (m, 5H, 20,30,40,50,60-H).
IR: 1589 (w), 1478 (m), 1466 (m), 1376 (w), 1229 (m), 1170 (w),
1079(w), 1055 (w), 1014(w), 968 (m), 827(w), 801(w), 751 (s), 728
(m), 690 (m), 663 (m), 554 (w), 507 (m), 464 (m), 434 (w). Mp:
136.7 °C. HPLC purity: 99.9% (tR = 6.77 min).
6-(Phenoxymethyl)-3-phenyl[1,2,4]triazolo[3,4-b][1,3,4]thiadia-
zole(18). Yield:70%. 1H NMR (200 MHz, DMSO-d6):δ [ppm]=
5.43 (s, 2H, -CH2-), 7.35-7.62 (m, 5H, 400,500,600,700,800-H), 8.32
-8.39 (m, 5H, 20,30,40,50,60-H). IR: 2980 (w), 1587 (w), 1468 (m),
1364 (w), 1294 (w), 1238 (m), 1171 (m), 1079 (w), 1053 (m),
1016 (w), 967 (m), 823 (w), 771 (w), 756 (s), 684 (s), 511 (m). Mp:
146.0 °C. HPLC purity: 99.4% (tR=7.55 min). HRMS: calcd for
C16H12N4OS [M•+] 308.0732, found 308.0749.
4-Amino-5-((4-methoxyphenoxy)methyl)-4H-1,2,4-triazole-3-
thiol (12). Yield: 62%. 1H NMR (200 MHz, DMSO-d6):
δ [ppm] = 3.70 (s, 3H, -CH3), 5.04 (s, 2H, -CH2-), 5.69 (s, 2H,
-NH2), 6.84-6.89 (d, 2H, 3J = 9.16 Hz, 20,60-H), 6.96-7.01 (d,
3
2H, J = 9.16 Hz, 30,50-H),13.79 (s, 1H, -SH). GC-MS (EI):
14.35 min, m/z [%] = 252.0 ([M*], 26), 124.0 (100), 123.0 (52),
109.0 (30), 95.0 (22), 81.0 (12). IR: 3142 (w), 2031 (w), 1580 (w),
1503 (m), 1462 (w), 1286 (w), 1223 (m), 1082 (w), 1031 (m), 982
(w), 812 (m), 735 (s), 665 (m), 607 (w), 514 (m), 482 (m). Mp:
176.8 °C.
General Method B for 5-Substituted [1,2,4]Triazoles. Carbon
disulfide (3.3 mL, 5.5 mmol) was added dropwise to a solution of
the suitable hydrazide (3.6 mmol) in ethanol (70 mL) containing
potassium hydroxide (3.0 g, 5.5 mmol). The mixture was stirred
at room temperature overnight, then cooled with an ice bath
and diluted with diethyl ether. The precipitate was filtered,
washed with diethyl ether, and dried. The hydrazinecarbo-
dithioates were obtained in quantitative yields and used without
further purification.
Hydrazine hydrate (2.4 mmol) was added to a solution of the
potassium salt (1.6 mmol) suspended in water (2 mL), and the
mixture was refluxed under stirring for 6 h. To the cooled
reaction mixture, water (80 mL) was added and the solution
acidified with concentrated hydrochloric acid. The precipitate
was filtered with water and dried.
3-(2-Fluorophenyl)-6-(phenoxymethyl)[1,2,4]triazolo[3,4-b]
[1,3,4]thiadiazole (19). Yield: 71%. 1H NMR (200 MHz,
DMSO-d6): δ [ppm] = 5.58 (s, 2H, -CH2-O-), 7.58-8.09
(m, 9H, 20,30,40,50,400,500,600,700,800-H). IR: 1589 (w), 1467 (m),
1376 (w), 1230 (m), 1122 (w), 1079 (w), 1055 (w), 1014 (w), 969
(m), 827 (w), 800 (w), 751 (s), 728 (m), 690 (m), 663 (m), 555 (w),
507 (m), 464 (m), 434 (m). Mp: 140.1 °C. HPLC purity: 99.9%
(tR = 7.08 min). HRMS: calcd for C17H13ClN4O2S [M•+
]
326.0637, found 326.0622.
6-(2-Chlorophenyl)-3-((4-methoxyphenoxy)methyl)[1,2,4]tria-
zolo[3,4-b][1,3,4]thiadiazole (20). Yield: 74%. 1H NMR (200
MHz, DMSO-d6): δ [ppm]=3.70 (s, 3H, -OCH3), 5.50 (s, 2H,
-CH2-), 6.85-6.90 (d, 2H, 3J = 9.22 Hz, 40,80-H), 7.03-7.07 (d,
2H, 3J=9.22 Hz, 50,70-H), 7.54-8.04 (m, 4H, 300,400,500,600-H). IR:
2980 (w), 1594 (w), 1502 (m), 1447 (m), 1356 (w), 1304 (w), 1231
(s), 1106 (m), 1019 (m), 820 (s), 761 (m), 718 (s). Mp: 145.9 °C.
HPLC purity: 99.9% (tR=7.01 min).
3-(2-Chlorophenyl)-6-(phenoxymethyl)[1,2,4]triazolo[3,4-b]
[1,3,4]thiadiazole (21). Yield: 77%. 1H NMR (200 MHz,
DMSO-d6): δ [ppm] = 5.56 (s, 2H, -CH2-), 6.99-8.03 (m,
9H, 30,40,50,60,400,500, 600,700,800-H). IR: 2916 (w), 1590 (w), 1488
(w), 1467 (m), 1365 (w), 1229 (m), 1170 (w), 1081 (w), 1051 (m),
970 (m), 817 (w), 725 (s), 690 (m), 511 (w). Mp: 127.4 °C. HPLC
purity: 99.9% (tR=7.13 min). HRMS: calcd for C16H11ClN4OS
[M•+] 342.0342, found 342.0335.
4-Amino-5-(4-fluorophenyl)-4H-1,2,4-triazole-3-thiol (13).
1
Yield: 75%. H NMR (200 MHz, DMSO-d6): δ [ppm] = 5.78
(s, 2H, -NH2), 7.33-8.12 (m, 4H, 20,30,50,60-H), 13.89 (s, 1H,
-SH). GC-MS (EI): 11.92 min, m/z [%] = 211.0 (12), 210.0
([M*] 100), 139.0 (28), 122.0 (28), 121.0 (25), 95.0 (16), 60.0 (11).
4-Amino-5-(2-chlorophenyl)-4H-1,2,4-triazole-3-thiol (14).
Yield: 77%. 1H NMR (200 MHz, DMSO-d6): δ [ppm] = 5.59
(s, 2H, -NH2), 7.54-7.68 (m, 4H, 30,40,50,60-H), 13.99 (s, 1H,
-SH). GC-MS (EI): 12.38 min, m/z [%] = 227.0 (12), 226.0
([M*], 100), 158.0 (11), 154.9 (23), 140.0 (10), 138.0 (30), 102.0
(32), 75.0 (16), 60.0 (19). IR: 3083 (w), 1629 (w), 1556 (w), 1488
(m), 1460 (m), 1327 (m), 1128 (w), 1080 (w), 1047 (m), 997 (m),
938 (s), 766 (s), 733 (m), 759 (s), 602 (s), 534 (w), 463 (m), 417 (w).
Mp: 161.8 °C.
3-(2-Chlorophenyl)-6-phenethyl[1,2,4]triazolo[3,4-b][1,3,4]thia-
1
diazole (22). Yield: 72%. H NMR (200 MHz, DMSO-d6): δ
[ppm] = 3.06 (t, 2H, 3J=7.58 Hz, 100-H), 3.39 (t, 2H, 3J=7.96
Hz, 200-H), 7.15-7.76 (m, 9H, 30,40,50,60,400,500,600,700,800-H). GC-
MS (EI): 13.43 min, m/z [%] = 342.10 (38), 341.10 (21), 340.10
([M*] (96), 139.00 (10), 137.00 (27), 115.10 (10), 102.00 (15), 91.10
(100). IR: 3001 (w), 1526 (w), 1496 (w), 1453 (s), 1137 (w), 1229
(m), 1106 (w), 1078 (w), 1052 (w), 1005 (w), 967 (m), 845 (w), 754
(s), 732 (m), 720 (s), 700 (s), 657 (m), 639 (m), 534(m), 480(m), 456
(w). Mp: 160.1 °C. HPLC purity: 99.9%(tR = 7.33 min). HRMS:
calcd for C17H13ClN4S [M3 +] 340.0549, found 340.0530.
4-Amino-5-(2,4-dichlorophenyl)-4H-1,2,4-triazole-3-thiol (15).
1
Yield: 81%. H NMR (200 MHz, DMSO-d6): δ [ppm] = 5.78
(s, 2H, -NH2), 7.60-7.88 (m, 3H, 30,50,60-H), 14.06 (s, 1H,
-SH). GC-MS (EI): 13.55 min, m/z [%] = 263.9 (15), 261.9
(73), 261.0 (13), 259.9 ([M*], 100), 190.9 (20), 188.9 (30), 173.9
(18), 172.9 (19), 171.9 (28), 170.9 (27), 136.0 (22), 100.0 (17), 60.0
(20). IR: 3093 (w), 1602 (w), 1563 (w), 1487 (m), 1374 (w), 1334
(m), 1147 (w), 1100 (m), 1045 (w), 949 (m), 864 (m), 843 (m), 813
(s), 734 (m), 624 (m), 569 (w), 525 (w), 479 (m), 436 (w). Mp:
213.6 °C.
3-(2-Chlorophenyl)-6-((4-methoxyphenoxy)methyl)[1,2,4]tria-
zolo[3,4-b][1,3,4]thiadiazole (23). Yield: 75%. 1H NMR

4208 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14
Cheeseright et al.
(200 MHz, DMSO-d6): δ [ppm] = 3.70 (s, 3H, -CH3), 5.49 (s,
2H, -CH2-), 7.15-7.76 (m, 9H, 30,40,50,60,400,500,600,700,800-H).
IR: 2837 (w), 1596 (w), 1555 (w), 1506 (s), 1465 (m), 1444 (m),
1352 (w), 1308 (w), 1234 (s), 1184 (m), 1106 (w), 1052 (m), 1032
(m), 961 (m), 827 (m), 770 (s), 728 (m), 678 (m), 657 (m), 597 (s),
518 (w), 465 (m), 426 (w). Mp: 145.9 °C. HPLC purity: 99.9%
References
(1) O’Neill, L. A. J. Targeting signal transduction as a strategy to
treat inflammatory diseases. Nat. Rev. Drug Discovery 2006, 5,
549–563.
(2) Lee, J. C. ; Laydon, J. T. ; McDonnell, P. C. ; Gallagher, T. F. ;
Kumar, S. ; Green, D. ; McNulty, D. ; Blumenthal, M. J. ; Heyes,
J. R. A protein kinase involved in the regulation of inflammatory
cytokine biosynthesis. Nature 1994, 372, 739–746.
(3) Wagner, G. ; Laufer, S. Small molecular anti-cytokine agents.
Curr. Med. Res. Rev. 2006, 26, 1–62.
(4) Wrobelski, S. T. ; Doweyko, A. M. Structural comparison of p38
inhibitor-protein complexes: a review of recent p38 inhibitors
unique binding interactions. Curr. Top. Med. Chem. 2005, 5,
1005–1016.
(5) Alvarez, J.; Shoichet, B. Virtual Screening in Drug Discovery; CRC
Press: Boca Raton, FL, 2005.
(6) Pierce, A. C. ; Rao, G. ; Bemis, G. W. BREED: generating novel
inhibitors through hybridization of known ligands. Application
to CDK2, P38, and HIV Pprotease. J. Med. Chem. 2004, 47, 2768–
2775.
(7) Pearlman, D. A. Evaluating the molecular mechanics Poisson-
Boltzmann surface area free energy method using a conge-
neric series of ligands to p38 MAP kinase. J. Med. Chem. 2005,
48, 7796–7807.
(8) Zhang, Q. ; Muegge, I. Scaffold hopping through virtual screening
using 2D and 3D similarity descriptors: ranking, voting, and
consensus scoring. J. Med. Chem. 2006, 49, 1536–1548.
(9) Lee, H. S. ; Choi, J. ; Kufareva, I. ; Abagyan, R. ; Filikov, A. ; Yang,
Y. ; Yoon, S. Optimization of high throughput virtual screening by
combining shape-matching and docking methods. J. Chem. Inf.
Model. 2008, 48, 489–497.
(10) Xiao, Z. ; Varma, S. ; Xiao, Y. D. ; Tropsha, A. Modeling of p38
mitogen-activated protein kinase inhibitors using the Catalyst
HypoGen and k-nearest neighbor QSAR methods. J. Mol. Gra-
phics Modell. 2004, 23, 129–138.
(11) Ottosen, E. R. ; Sorensen, M. D. ; Bjoerkling, F. ; Skak-Nielsen, T. ;
Fjording, M. S. ; Aaes, H. ; Binderup, L. Synthesis and structure-
activity relationship of aminobenzophenones. A novel class of p38
MAP kinase inhibitors with high antiinflammatory activity.
J. Med. Chem. 2003, 46, 5651–5662.
(12) Hauser, D. R. J. ; Scior, T. ; Domeyer, D. M. ; Kammerer, B. ;
Laufer, S. A. Synthesis, biological testing, and binding mode
prediction of 6,9-diarylpurin-8-ones as p38 MAP kinase inhibitors.
J. Med. Chem. 2007, 50, 2060–2066.
(tR = 7.01 min). HRMS: calcd for C17H13ClN4O2S [M•+
]
372.0448, found 372.0417.
3-(2-Chlorophenyl)-6-(4-methoxybenzyl)[1,2,4]triazolo[3,4-
b][1,3,4]thiadiazole (24). Yield: 82%. 1H NMR (200 MHz,
DMSO-d6): δ [ppm] = 3.74 (s, 3H, -OCH3), 4.37 (s, 2H,
-CH2-), 6.91-6.95 (d, 2H, 3J = 8.72 Hz, 400,600-H), 7.30-7.34
(d, 2H, 3J = 8.72 Hz, 300,700-H), 7.53-7.82 (m, 4H, 30,40,50,60-H).
IR: 2995 (w), 1606 (w), 1585 (m), 1503 (m), 1453 (m), 1296 (w),
1238 (s), 1182 (m), 1122 (m), 1108 (m), 1027 (m), 1009 (m), 857
(m), 822 (m), 809 (m), 756 (s), 732 (m), 721 (s), 687 (w), 666 (m),
641 (m), 548 (m), 525 (m), 457 (m), 437 (m). Mp: 158.1 °C. HPLC
purity: 99.9% (tR = 7.26 min). HRMS: calcd for C17H13ClN4OS
[M•+] 356.0498, found 356.0464.
3-(4-Fluorophenyl)-6-phenethyl[1,2, 4]triazolo[3,4-b][1,3,4] thia-
1
diazole (25). Yield: 73%. H NMR (200 MHz, DMSO-d6): δ
[ppm] = 3.12 (t, 2H, 3J=7.56 Hz, 100-H), 3.45 (t, 2H, 3J=7.77
Hz, 200-H), 7.18-7.54 (m, 7H, 30,50,400,500,600,700,800-H), 8.12-8.31
(m, 2H, 20, 60-H). GC-MS(EI): 12.10min, m/z [%]=325.10 (22),
324.10 ([M*], 100), 121.00 (35), 91.10 (71). IR: 2980 (w), 1604 (w),
1478 (m), 1452 (s), 1221 (m), 1163 (m), 1015 (m), 970 (m), 859 (s),
814 (m), 746 (m), 719 (m), 695 (s), 632 (m), 584 (m), 521 (s), 481
(m). Mp: 165.9 °C. HPLC purity: 99.0% (tR=7.82 min). HRMS:
calcd for C17H13FN4S [M•+] 324.08445, found 324.0811.
3-(2,4-Dichlorophenyl)-6-phenethyl[1,2,4]triazolo[3,4-b][1,3,4]
thiadiazole (26). Yield: 75%. 1H NMR (200 MHz, DMSO-d6): δ
[ppm]=3.15 (t, 2H, 3J=7.52 Hz, 100-H), 3.29 (t, 2H, 3J=7.71
Hz, 200-H), 7.18-7.39 (m, 8H, 30,50,60,400,500,600,700,800-H). IR:
1603 (w), 1494 (w), 1470 (m), 1441 (s), 1194 (m), 1064 (w),
1016 (w), 991 (w), 977 (m), 839 (m), 750 (m), 698 (s), 687 (m), 646
(w), 567 (w), 524 (m), 493 (m). Mp: 145.5 °C. HPLC purity:
99.9% (tR = 8.26 min). HRMS: calcd for C17H12Cl2N4S [M•+
]
374.01595, found 374.0142.
3-(2,4-Dichlorophenyl)-6-((4-methoxyphenoxy)methyl)[1,2,4]tria-
zolo[3,4-b][1,3,4]thiadiazole (27). Yield: 73%. 1H NMR (200 MHz,
DMSO-d6): δ [ppm] = 3.70 (s, 3H, -OCH3), 5.49 (s, 2H, -CH2-
O-), 6.86-6.91 (d, 2H, 3J=9.22 Hz, 500,700-H), 7.02-7.07 (d, 2H,
3J=9.22 Hz, 400,800-H), 7.67-7.96 (m, 3H, 30,50,60-H). IR: 1597 (w),
1503 (m), 1467 (m), 1238 (s), 1190 (m), 1104 (m), 1036 (m), 967 (m),
818 (s), 766 (w), 725 (w), 671 (m), 609 (m), 565 (w), 523 (m). Mp:
161.8 °C. HPLC purity: 99.9% (tR=7.90 min). HRMS: calcd for
C17H12Cl2N4O2S [M•+] 406.0058, found 406.0074.
(13) Hartshorn, M. J. ; Murray, C. W. ; Cleasby, A. ; Frederickson, M. ;
Tickle, I. J. ; Jhoti, H. Fragment-based lead discovery using X-ray
crystallography. J. Med. Chem. 2005, 48, 403–413.
(14) Cheeseright, T. J. ; Mackey, M. D. ; Melville, J. L. ; Vinter, A.
FieldScreen: virtual screening using molecular fields. Application
to the DUD dataset. J. Chem. Inf. Model. 2008, 48, 2108–2117.
(15) Cheeseright, T. ; Mackey, M. ; Rose, S. ; Vinter, A. Molecular field
extrema as descriptors of biological activity: definition and valida-
tion. J. Chem. Inf. Model. 2006, 46, 665–676.
(16) Low, C. M. R. ; Buck, I. M. ; Cooke, T. ; Cushnir, J. R. ; Kalindjian,
S. B. ; Kotecha, A. ; Pether, M. J. ; Shankley, N. P. ; Vinter, J. G. ;
Wright, L. Scaffold hopping with molecular field points: identifi-
cation of a cholecystokinin-2 (CCK2) receptor pharmacophore
and its use in the design of a prototypical series of pyrrole-
and imidazole-based CCK2 antagonists. J. Med. Chem. 2005, 48,
6790–6802.
(17) Stelmach, J. E. ; Liu, L. ; Patel, S. B. ; Pivnichny, J. V. ; Scapin, G. ;
Singh, S. ; Hop, C. E. C. A. ; Wang, Z. ; Strauss, J. R. ; Cameron,
P. M. ; Nichols, E. A. ; O’Keefe, S. J. ; O’Neill, E. A. ; Schmatz, D.
M. ; Schwartz, C. D. ; Thompson, C. M. ; Zaller, D. M. ; Doherty,
J. B. Design and synthesis of potent, orally bioavailable dihydro-
quinazolinone inhibitors of p38 MAP kinase. Bioorg. Med. Chem.
Lett. 2003, 13, 277–280.
6-Phenethyl-3-(pyridin-4-yl)[1,2,4]triazolo[3,4-b][1,3,4]thia-
diazole (28). Yield: 62%. 1H NMR (200 MHz, DMSO-d6):
δ [ppm] = 3.06 (t, 2H, 3J = 7.52 Hz, 100-H), 3.39 (t, 2H, 3J =
7.39 Hz, 200-H), 7.08-7.41 (m, 5H, 400,500,600,700,800-H), 7.64-
7.91 (m, 4H, 20,30,50,60-H). IR: 1596 (w), 1553 (w), 1539 (w),
1496 (w), 1452 (s), 1101 (m), 1078 (w), 1051 (w), 1000 (w), 966
(m), 859 (m), 818 (s), 796 (m), 700 (m), 647 (w), 561 (m), 539
(m), 511 (w), 481 (m), 462 (m). Mp: 206.3 °C. HPLC purity:
98.5% (tR = 6.55 min). HRMS: calcd for C16H13N5S [M•+
]
307.0892, found 307.0906.
Acknowledgment. The research was realized with EU fi-
nancial support, part of the FP6 project “Macrocept”, and
thanks are given to all project team members of “Macrocept”.
Thanks to are also given to Dr. Schachtele and Dr. Totzke
(ProQinase GmbH) for testing all triazoles for their inhibition
activity against a panel of protein kinases and for generating
the selectivity profiles.
(18) Laufer, S. ; Thuma, S. ; Peifer, C. ; Greim, C. ; Herweh, Y. ;
Albrecht, A. ; Dehner, F. An immunosorbent, nonradioactive p38
MAP kinase assay comparable to standard radioactive liquid-
phase assays. Anal. Biochem. 2005, 344, 135–137.
(19) Carhart, R. E. ; Smith, D. H. ; Venkataraghavan, R. Atom pairs as
molecular features in structure-activity studies: definition and
applications. J. Chem. Inf. Comput. Sci. 1985, 25, 64–73.
(20) Lehmann, F. ; Holm, M. ; Laufer, S. Three-component combina-
torial synthesis of novel dihydropyrano[2,3-c]pyrazoles. J. Comb.
Chem. 2008, 10, 364–367.
Supporting Information Available: Structures of purchased
compounds and HRMS, HPLC, and inhibition data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(21) Malhotra, S. ; Manher, V. ; Chadha, V. K. Bridgehead nitrogen
heterocycles: syntheses of 3,6-disubstituted-s-triazolo[3,4-b][1,3,4]
thiadiazoles, and related systems. Indian J. Heterocycl. Chem. 2003,
12, 257–262.

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 14 4209
(22) Invidiata, F. P. ; Furno, G. ; Lampronti, I. ; Simoni, D. 1,2,4-
Triazoles. Improved synthesis of 5-substituted 4-amino-3-mercap-
to-(4H)-1,2,4-triazoles and a facile route to 3,6-disubstituted 1,2,4-
triazolo[3,4-b][1,3,4]thiadiazoles. J. Heterocycl. Chem. 1997, 34,
1255–1258.
(25) http://www.cresset-bmd.com/xedex.shtml (accessed Oct 3, 2008).
(26) Brown, R. D. ; Martin, Y. C. Use of structure-activity data to
compare structure-based clustering methods and descriptors for
use in compound selection. J. Chem. Inf. Comput. Sci. 1996, 36,
572–584.
(23) Reid, J. R. ; Heindel, N. D. Improved syntheses of 5-substituted-4-
amino-3-mercapto-(4H)-1,2,4-triazoles. J. Heterocycl. Chem.
1976, 13, 925–926.
(27) Sheridan, R. P. ; Hunt, P. ; Culberson, J. C. Molecular transfor-
mations as a way of finding and exploiting consistent local QSAR.
J. Chem. Inf. Model. 2006, 46, 180–192.
(24) http://www.accelrys.com/ (accessed Oct 3, 2008).
(28) http://www.cresset-bmd.com/fieldalign.shtml (accessed Oct 3, 2008).