Multitargeted Imidazoles: Potential Therapeutic Leads for Alzheimer's and Other Neurodegenerative Diseases

Article abstract of DOI:10.1021/acs.jmedchem.7b00475

Alzheimer's disease (AD) is a complex, multifactorial disease in which different neuropathological mechanisms are likely involved, including those associated with pathological tau and Aβ species as well as neuroinflammation. In this context, the development of single multitargeted therapeutics directed against two or more disease mechanisms could be advantageous. Starting from a series of 1,5-diarylimidazoles with microtubule (MT)-stabilizing activity and structural similarities with known NSAIDs, we conducted structure-Activity relationship studies that led to the identification of multitargeted prototypes with activities as MT-stabilizing agents and/or inhibitors of the cyclooxygenase (COX) and 5-lipoxygenase (5-LOX) pathways. Several examples are brain-penetrant and exhibit balanced multitargeted in vitro activity in the low μM range. As brain-penetrant MT-stabilizing agents have proven effective against tau-mediated neurodegeneration in animal models, and because COX-and 5-LOX-derived eicosanoids are thought to contribute to Aβ plaque deposition, these 1,5-diarylimidazoles provide tools to explore novel multitargeted strategies for AD and other neurodegenerative diseases.

Full text of DOI:10.1021/acs.jmedchem.7b00475

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Multi-targeted Imidazoles: Potential Therapeutic Leads
for Alzheimer’s and Other Neurodegenerative Diseases
Anne-Sophie Cornec, Ludovica Monti, Jane Kovalevich, Vishruti Makani, Michael
J James, Krishna G Vijayendran, Killian Oukoloff, Yuemang Yao, Virginia M.-
Y. Lee, John Q Trojanowski, Amos B Smith, III, Kurt R Brunden, and Carlo Ballatore
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 22 May 2017
Downloaded from http://pubs.acs.org on May 23, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 79
Journal of Medicinal Chemistry
1
Multiꢀtargeted Imidazoles: Potential Therapeutic Leads
for Alzheimer’s and Other Neurodegenerative Diseases
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
b
c
c
c
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
AnneꢀSophie Cornec, Ludovica Monti, Jane Kovalevich, Vishruti Makani, Michael J. James,
a
b
c
c
Krishna G. Vijayendran, Killian Oukoloff, Yuemang Yao, Virginia M.ꢀY. Lee, John Q.
c
a
c, *
b,*
Trojanowski, Amos B. Smith III, Kurt R. Brunden, and Carlo Ballatore
a
th
Department of Chemistry, School of Arts and Sciences, University of Pennsylvania, 231 South 34 St.,
b
Philadelphia, PA 19104ꢀ6323; Skaggs School of Pharmacy and Pharmaceutical Sciences, University of
c
California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093 ; Center for Neurodegenerative Disease
Research, Institute on Aging, University of Pennsylvania, 3600 Spruce Street, Philadelphia, PA 19104ꢀ
6323
*
Corresponding Authors.
RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required
according to the journal that you are submitting your paper to)
ABSTRACT: Alzheimer’s disease (AD) is a complex, multifactorial disease in which different
neuropathological mechanisms are likely involved, including those associated with pathological tau and
Aβ species, as well as neuroinflammation. In this context, the development of single multiꢀtargeted
therapeutics directed against two or more disease mechanisms could be advantageous. Starting from a
series of 1,5ꢀdiarylimidazoles with microtubule (MT)ꢀstabilizing activity and structural similarities with
known NSAIDs, we conducted structureꢀactivity relationship studies that led to the identification of
multiꢀtargeted prototypes with activities as MTꢀstabilizing agents and/or inhibitors of the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 79
cyclooxygenase (COX) and 5ꢀlipoxygenase (5ꢀLOX) pathways. Several examples are brainꢀpenetrant
and exhibit balanced multiꢀtargeted in vitro activity in the low µM range. As brainꢀpenetrant MTꢀ
stabilizing agents have proven effective against tauꢀmediated neurodegeneration in animal models, and
because COXꢀ and 5ꢀLOXꢀderived eicosanoids are thought to contribute to Aβ plaque deposition, these
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
,5ꢀdiarylimidazoles provide tools to explore novel multiꢀtargeted strategies for AD and other
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
neurodegenerative diseases.
KEYWORDS: polypharmacology, multiꢀtargeted ligands, 1,5ꢀdiarylimidazoles, microtubuleꢀstabilizing
agents, cyclooxygenase, lipoxygenase, Alzheimer’s disease, tauopathies, neuroinflammation.
Introduction. In Alzheimer’s disease (AD), aggregated tau proteins and Aβ peptides are thought to
participate in disease onset and progression. Moreover, other factors such as neuroinflammation and
oxidative stress are also believed to be intimately associated with the neurodegenerative process.
Although the precise relationship(s) linking these different neuropathological mechanisms remain the
focus of active research, it is clear that AD is a complex and multiꢀfactorial disease. Thus far, Aβ
amyloidosis has been the target of choice in AD drug discovery, as most drug candidates have been
designed to modulate Aβ homeostasis. However, in spite of generally strong preclinical data with such
therapeutic candidates, the outcomes of late stage AD clinical trials have thus far failed to demonstrate
clinical efficacy. These disappointing results raise the possibility that AD treatments may have to
interfere concurrently with more than one neuropathological mechanism to exert disease modifying
benefits. In this context, a single multiꢀtargeted drug may have distinctive advantages in terms of
1
efficacy and safety over drug combination therapies. As such, an increasing number of therapeutic
2
, 3
strategies that are based on polypharmacology have been proposed for AD.
ACS Paragon Plus Environment

Page 3 of 79
Journal of Medicinal Chemistry
Over the past several years, our laboratories and others have demonstrated that brainꢀpenetrant MTꢀ
1
2
stabilizing agents have profoundly beneficial effects in animal models of tau pathology due to the ability
of these compounds to normalize hyperdynamic axonal MTs, restore axonal transport, reduce the
burden of tau pathology, and prevent cognitive impairments and neuron loss associated with tau
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
ꢀ7
8
pathology. Furthermore, our recent work indicates that the binding of cyclooxygenase (COX)ꢀ and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
lipoxygenase (5ꢀLOX)ꢀderived eicosanoids to their cognate receptors in neurons results in increased
9ꢀ23
production of amyloid precursor protein (APP) and Aβ peptides, adding to a body of literature
implicating inflammatory eicosanoids as key contributing factors to Aβ plaque deposition in AD. Thus,
a stabilization of axonal MTs, combined with a concurrent suppression of eicosanoid production, may
be considered as a multiꢀtargeted therapeutic strategy that could attenuate both tauꢀ and Aβꢀmediated
2
4
25ꢀ31
neurodegeneration, as well as MT deficiencies and/or neuroinflammation,
in AD. Moreover, this
approach may also provide benefit in other neurodegenerative tauopathies.
The rational design of multiꢀtargeted compounds is challenging, especially when attempting to merge
multiple molecular frameworks and the corresponding underlying pharmacophores in a single small
32
molecule. However, a series of fungicidal diarylꢀpyrazoles (e.g., 1, Figure 1) and ꢀimidazoles (e.g., 2,
Figure 1) with known MTꢀstabilizing activity exhibit tricyclic structures similar to those found in known
3
3
nonꢀsteroidal antiꢀinflammatory drugs (NSAIDs), such as the potent COXꢀ1 inhibitor SC560 (3,
34
Figure 1) and to a lesser extent, the dual COX/5ꢀLOX inhibitor licofelone (4, Figure 1). Furthermore,
molecular docking studies indicated that selected representative examples of these MTꢀstabilizing
imidazoles could fit within the arachidonic acid binding site of COXꢀ1 with a predicted binding energy
that is comparable (i.e., within 2 kcal/mol) to that of 3 (see Figure 1 and Supporting Information). Thus,
although MTꢀstabilizing agents and NSAIDs are different classes of biologically active compounds that
interact with unrelated targets, the observed structural commonalities between these compounds
suggests that there may be significant degree of crosstalk between the different pharmacophores that
could be exploited to identify multiꢀtargeted ligands. Toward this end, the synthesis and subsequent
evaluation of several closely related 1,5ꢀdiarylimidazole congeners led to the identification of specific
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 79
substitution patterns that are required to achieve different activity profiles in vitro. In particular, in
addition to identifying compounds that act selectively as MTꢀstabilizers, and others that are selective
inhibitors of the COXꢀ and/or 5ꢀLOX pathways with no effect on MTs, these studies led to the
identification of a series of multiꢀtargeted molecules that at low µM concentration can concurrently
stabilize MTs and inhibit both COXꢀ and 5ꢀLOX derived eicosanoids in cellꢀbased assays. Furthermore,
several of these compounds were also confirmed to be brainꢀpenetrant, indicating that molecules of this
type may be considered as potential multiꢀtargeted prototype structures for AD and related tauopathies.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. (A) Representative examples of fungicidal diarylꢀpyrazoles (1) and ꢀimidazoles (2) with
reported MTꢀstabilizing activity, with known COXꢀ (3) and dual COX/5ꢀLOX (4) inhibitors; molecular
docking of 2 (B) and 3 (C) within the arachidonic acid binding site of COXꢀ1.
ACS Paragon Plus Environment

Page 5 of 79
Journal of Medicinal Chemistry
Results. Chemistry: As part of these studies, a library comprising 81 imidazoles has been synthesized,
1
2
including 36 known compounds (i.e., 2, 25, 29–31, 35, 36, 38, 43–46, 51, 54, 55, 57, 58, 61, 62, 64, 65,
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
5
67, 68, 70, 71, 73, 74, 83–86, 88, 92, 94, 95, and 97, Table 1), most with reported antiꢀfungal or antiꢀ
microbial activity, as well as 45 new structures (i.e., 26–28, 32–34, 37, 39–42, 47–50, 52, 53, 56, 59, 60,
6
3, 66, 69, 72, 75–82, 87, 89–91, 93, 96, 98–104, Table 1). In general, the synthesis of the 1,5ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
6
diarylimidazoles was conducted following the van Leusen imidazole synthesis. Thus, after
condensation of the appropriate arylꢀamine and benzaldehyde, the resulting Schiff base (5–24, Scheme
1) was cyclized in the presence of toluenesulfonylmethyl isocyanide (TosMIC) to form the 1,5ꢀ
diarylimidazole (25–28, 32, 33, 37–43, 47–53, Scheme 1). Further derivatizations of the imidazole ring
included the introduction of small alkyl groups and/or halogens in position 2 and/or 4 (29–31, 34–36,
44–46, in Scheme 1, and 54–104 in Scheme 2). Imidazole derivatives bearing a methyl substituent at C4
were accessed by employing the appropriately C substituted TosMIC reagent during the cyclization step
(
29, 35, 44), while alkylated derivatives at C2 (e.g., 34, 86) were obtained by treating the appropriate
1,5ꢀdiarylimidazole with LDA followed by addition of the desired alkyliodide. Finally, halogenation of
the imidazole ring was accomplished upon treatment with Nꢀchlorosuccinimide. When the halogenation
was carried out on imidazoles that were not substituted at C2 and C4, the reaction typically proceeded
with the generation of separable mixtures of monoꢀ and diꢀchlorinated imidazoles.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Reagents and reaction conditions: (a) DeanꢀStark apparatus, toluene, 110 °C, 4 days, 99%;
(b) 4 Å molecular sieves, Et O, 30 °C, 48 h, 99%; (c) appropriate TosMIC reagent, K CO , DMF/1,2ꢀ
2
2
3
dimethoxyethane, 100 °C, 16 h, 7–90%; (d) Nꢀchlorosuccinimide, CHCl , 60 °C, 16 h, 43–71%; (e)
3
LDA, methyliodide, THF, –20 °C to r.t., 1.5 h, 29–56%.
ACS Paragon Plus Environment

Page 7 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Reagents and reaction conditions: (a) Nꢀchlorosuccinimide, CHCl , 60 °C, 16 h, 9–77%
3
for monoꢀhalogenation at C4, 3–18% for monoꢀhalogenation at C2, and 2–20% for diꢀhalogenation at
C2/C4; (b) LDA, methyliodide, THF, –20 °C to r.t., 1.5 h, 2–48%; (c) NaOMe, MeOH, THF, r.t., 19 h,
3
3ꢀ97%; (d) LDA, ethyliodide, THF, –20 °C to r.t., 1.5 h, 35%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 79
Structureꢀactivity relationships: The MTꢀstabilizing activity of test compounds was determined in
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
7
QBI293 cells by monitoring compoundꢀinduced elevation in acetylated αꢀtubulin (AcTub), which is a
3
8
known marker of stable MTs. Furthermore, as the MTꢀstabilizing imidazoles are believed to act on
35
MTs in a manner similar to that described for the triazolopyrimidines, and considering that certain
3
9
triazolopyrimidines have been found to induce a proteasomeꢀdependent degradation of αꢀtubulin, the
effect of compound treatment on total αꢀtubulin was also monitored. Evaluation of compound inhibition
of the biosynthesis of COXꢀ and 5ꢀLOXꢀderived eicosanoids was conducted in a modified rat basophilic
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
0
leukemia (RBLꢀ1) cell assay that has previously been described for the evaluation of 5ꢀLOX and
which we found can be used to assess both COXꢀderived prostaglandins (PGs) and 5ꢀLOXꢀderived
leukotrienes (LTs) formed in the presence or absence of test compounds.
All compounds were initially screened at 1 and 10 µM concentration in the QBI MT assay and at
1
0 µM in the RBLꢀ1 assay (Table 1). In this initial RBLꢀ1 analysis, 5ꢀLOXꢀderived LTB and COXꢀ
4
derived PGD and PGE , which coꢀeluted during LC/MS/MS, were quantified. Selected compounds
2
2
(Table 2) exhibiting evidence of multiꢀtargeted activity underwent further confirmatory studies in the
RBLꢀ1 assay, which included concentrationꢀresponse testing, and further LC/MS/MS analyses using a
refined protocol to individually detect and quantitate PGD and PGE , as well as both LTB and LTC .
2
2
4
4
As the formation of each of these eicosanoids depends on separate enzymatic steps in the COX or 5ꢀ
41
LOX pathways, concurrent analysis of all four of these eicosanoids provides a convenient and reliable
method to evaluate the overall effect of test compounds on COX and 5ꢀLOX pathways in a cellular
milieu. Finally, several compounds in Table 2 were also assessed for their relative brainꢀtoꢀplasma
exposure levels after administration to mice.
MTꢀstabilizing activity: Evaluation of compounds in the QBI MT assay revealed that out of 71 tested
compounds, 48 examples exhibited activity on MTs as revealed by statistically significant changes in
AcTub levels. Among MTꢀactive compounds, 7 examples (i.e., 2, 58, 65, 71, 85, 86, and 97) were found
to be active at 1 µM compound concentration whereas the remaining 41 compounds were active only at
1
0 µM. Consistent with the notion that the MTꢀstabilizing 1,5ꢀdiarylimidazoles may be acting on MTs
ACS Paragon Plus Environment

Page 9 of 79
Journal of Medicinal Chemistry
3
5
in a similar manner as the triazolopyrimidines, the data presented in Table 1 reveal that like the latter
molecules, these imidazoles can be broadly divided into two subsets: one group of 15 entries, which
cause a significant reduction in total αꢀtubulin levels at doses typically required to induce an elevation in
AcTub (i.e., 2, 30, 34, 58, 62, 63, 65, 66, 71, 73, 85, 94–97); and a second group of 33 compounds that
produce an elevation in AcTub at 1 or 10 ꢁM without decreasing αꢀtubulin level (29, 31, 35, 36, 44–46,
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
4, 55–57, 59, 60, 64, 67–70, 72, 74, 78, 82–84, 86–88, 90–93, 102, 104). Also of interest, a comparison
of the activities of compounds in the AcTub assay at 1 and 10 µM suggest that selected entries, such as
compounds 2, 58, and 71, may exhibit a bellꢀshaped concentrationꢀresponse relationship, and it is
possible that additional examples would show a similar concentrationꢀresponse if tested at higher
concentrations. This phenomenon was observed in the case of the triazolopyrimidines that reduce αꢀ
3
9
tubulin level, such as 108. Irrespective of whether or not test compounds cause a reduction in total αꢀ
tubulin, the AcTub SAR emerging from the data shown in Table 1 appear to be in general agreement
with SAR from prior studies in which MTꢀstabilizing 1,5ꢀdiarylimidazoles were evaluated for antiꢀ
3
5
fungal activity. Thus, our results indicate that ring A is preferably a 4ꢀchloroꢀphenyl, although
examples bearing other substituents in the paraꢀposition, such as a trifluoromethoxy (83), are also
active. Additional substitutions in the A ring, as in compound 77 and 78 in which a fluorine is present in
the orthoꢀposition, appear to impact negatively MT interactions of the compound (cf., 77 with 87, and
7
8 with 84). With respect to the B ring, the degree and pattern of fluorination appear to be generally
important factors that determine the ability of these imidazoles to produce a significant elevation in the
marker of stable MTs in QBI293 cells. Whereas derivatives with either one or no fluorination in the B
ring (i.e., 50, 51, 75, 76, 98–101) were found to be essentially devoid of MT activity in our assay
conditions, other diꢀ and triꢀsubstituted congeners (e.g., 54, 68, 84) produced a significant increase in
AcTub. Furthermore, a comparison between the inactive difluorinated derivative 80 and the active diꢀ
and triꢀfluorinated compounds (i.e., 54, 68, and 84) suggests that the presence of two fluorine atoms in
the orthoꢀpositions of ring B is required for MT activity. Finally, with respect to the imidazole ring (ring
C), our results indicate that 1,5ꢀdiarylimidazoles that are not substituted in positions 2 and 4 (e.g., 25,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 79
2
6, 42, 43, 48, 49, 52 and 53) do not cause a detectable increase in AcTub, whereas tetrasubstituted
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
imidazoles, including C2/C4 dihalogenated (e.g., 88) and dimethylated (e.g., 46) derivatives, as well as
analogues bearing both a small alkylꢀ and a chloroꢀsubstituent in either C2 or C4 positions (e.g., 85 and
4
5), are generally more active than corresponding derivatives bearing a single substitution in either
position 2 or 4 (cf., 87, 88, and 44, 46). These results also suggest that MT activity does not strictly
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
5
require a halogen in position 4 of the imidazole ring, as was required for optimal antiꢀfungal activity,
since congeners halogenated at C2 appear to retain comparable MT activity (cf., 54, 56 and 84, 87).
ACS Paragon Plus Environment

Page 11 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
Table 1.
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
PGD /E %
LTB %
2
2
4
Relative
Cpd #
2
X
Y
Ring A
Ring B
Relative Change
Inhibition
Inhibition
c
Change in αꢀ
a
b
in AcTub
d
at 10 µM
at 10 µM
Tubulin
× 7.1 (1 µM)**
× 0.6 (1 µM)**
Me
Cl
39
44
88
24
25
71
54
×
2.9 (10 µM)** × 0.3 (10 µM)**
25
26
29
30
H
H
H
H
NS
NS
NS
NS
NS
65
H
Me
Me
59
× 3.5 (10 µM)**
Cl
<10
× 9.2 (10 µM)** × 0.5 (10 µM)**
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
31
34
35
36
42
43
44
45
Me
Me
H
Me
H
<10
49
51
45
74
58
55
46
44
42
× 14.2 (10 µM)**
NS
× 4.7 (10 µM)** × 0.9 (10 µM)**
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
Me
Me
H
74
× 3.4 (10 µM)**
× 14.3 (10 µM)**
NS
× 1.2 (1 µM)**
Cl
H
<10
38
NS
NS
× 1.2 (1 µM)*
NS
H
H
15
NS
H
Me
Me
72
× 1.6 (10 µM)**
× 4.1 (10 µM)**
Cl
<10
NS
ACS Paragon Plus Environment

Page 13 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
4
4
4
6
7
8
9
Me
H
Me
H
<10
31
40
52
84
78
× 8.3 (10 µM)**
× 1.2 (1 µM)*
NS
NS
NS
× 1.2 (10 µM)*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
H
H
66
NS
NS
H
H
<10
5
0
1
H
H
H
H
40
94
16
NS
NS
NS
NS
5
<10
52
53
54
H
H
H
H
H
19
24
91
<10
<10
46
NS
NS
NS
NS
NS
Cl
× 3.3 (10 µM)**
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
55
56
57
58
59
60
62
63
Me
Cl
Cl
Cl
H
Cl
H
<10
34
<10
61
48
38
86
75
78
69
× 5.7 (10µM)**
× 3.1 (10 µM)**
NS
NS
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
Cl
Cl
Cl
H
<10
12
× 4.2 (10 µM)**
× 3.3 (1 µM)**
NS
× 0.6 (1 µM)**
×
2.5 (10 µM)** × 0.3 (10 µM)**
93
× 4.1 (10 µM)**
NS
NS
Cl
H
85
× 3.4 (10 µM)**
Cl
H
79
× 2.2 (10 µM)** × 0.8 (10 µM)**
× 0.8 (1 µM)**
× 2.3 (10 µM)**
Cl
<10
×
0.9 (10 µM)**
ACS Paragon Plus Environment

Page 15 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
64
65
66
67
68
69
70
71
H
Me
Cl
H
Cl
Cl
H
71
34
63
84
88
78
69
24
66
75
× 2.4 (10 µM)**
× 1.7 (1 µM)**
NS
× 0.6 (10 µM)**
×
9.7 (10 µM)**
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
<10
54
× 2.2 (10 µM)** × 0.8 (10 µM)**
Cl
Cl
H
× 2.1 (10 µM)**
× 2.0 (10 µM)**
× 2.4 (10 µM)**
NS
NS
NS
H
82
Cl
H
26
Cl
Cl
74
× 1.9 (10 µM)**
× 3.9 (1 µM)**
NS
× 0.8 (1 µM)**
Me
<10
×
2.4 (10 µM)** × 0.7 (10 µM)**
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
7
7
7
2
3
4
Cl
H
H
Cl
Cl
32
65
58
<10
26
× 3.2 (10 µM)**
NS
× 1.9 (10 µM)** × 0.8 (10 µM)**
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
Me
<10
× 4.4 (10 µM)**
NS
7
5
6
Cl
H
H
87
85
57
59
NS
NS
NS
NS
7
Cl
77
78
79
80
Cl
H
H
Cl
H
<10
36
49
14
96
69
NS
NS
NS
NS
NS
× 2.6 (10 µM)**
Cl
H
76
NS
NS
Cl
46
ACS Paragon Plus Environment

Page 17 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
81
82
83
84
85
86
87
88
H
Cl
Cl
H
Cl
H
46
19
19
94
45
30
50
17
78
61
81
76
87
81
92
90
NS
NS
× 2.2 (10 µM)**
× 6.1 (10 µM)**
NS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Cl
Cl
Cl
Cl
H
NS
× 4.0 (10 µM)**
× 2.5 (1 µM)*
NS
Me
Et
Cl
Cl
× 0.8 (10 µM)**
×
15.7 (10 µM)**
× 1.8 (1 µM)**
NS
NS
NS
×
5.0 (10 µM)**
× 4.8 (10 µM)**
Cl
× 7.7 (10 µM)**
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
89
90
91
92
93
94
95
96
H
Cl
Cl
H
Cl
H
54
58
85
80
56
75
73
63
91
68
NS
NS
NS
NS
NS
NS
× 3.4 (10 µM)**
× 6.3 (10 µM)**
× 3.6 (10 µM)**
× 5.7 (10 µM)**
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
Cl
Cl
H
40
90
Cl
Cl
H
42
Cl
Cl
H
<10
48
× 18.2 (10 µM)** × 0.7 (10 µM)**
× 4.0 (10 µM)** × 0.8 (10 µM)**
× 8.2 (10 µM)** × 0.6 (10 µM)**
Cl
<10
ACS Paragon Plus Environment

Page 19 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
× 5.1 (1 µM)**
× 0.9 (1 µM)**
9
7
Cl
Cl
17
88
×
10.1 (10 µM)** × 0.4 (10 µM)**
9
8
9
H
Cl
Cl
H
Cl
H
94
75
14
95
71
52
NS
NS
NS
NS
NS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
9
× 1.2 (10 µM)**
1
1
00
01
Cl
Cl
<10
<10
NS
NS
1
1
1
02
03
04
H
Cl
Cl
Cl
H
44
28
66
56
38
21
× 3.5 (10 µM)**
NS
NS
NS
NS
Cl
× 3.8 (10 µM)**
Ibuprofen (105)
95
<10
NS
NS
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
3
§
§
79
<10
NS
NS
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
4
2
MK866 (106)
§
§
<
10
70
ND
ND
NS
Triazolopyrimidine
(107)
∧
× 2.5 (1 µM) ,
ND
ND
ND
ND
∧
×
4.9 (10 µM)
Triazolopyrimidine
∧
× 6.0 (1 µM) ,
× 0.7 (1 µM),
× 0.4 (10 µM)
(
108)
∧
× 0.5 (10 µM)
a
Inhibition of COX pathway was determined by monitoring via LC/MS/MS analyses the combined production of COXꢀderived PGD and PGE in RBLꢀ1
2
2
b
upon stimulation with arachidonic acid in the presence or absence of test compounds, with all samples run in at least triplicate; Inhibition of 5ꢀLOX pathway
was determined by monitoring via LC/MS/MS analyses the production of 5ꢀLOXꢀderived LTB in RBLꢀ1 upon stimulation with arachidonic acid in the
4
c
presence or absence of test compounds, with all samples run in at least triplicate; MTꢀstabilizing activity in QBIꢀ293 cells was determined by monitoring via
d
ELISA the relative changes in acetylated αꢀtubulin levels in response to 4ꢀhour treatment with test compounds, with all samples run in at least triplicate;
ACS Paragon Plus Environment

Page 21 of 79
Journal of Medicinal Chemistry
1
§
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
Relative changes in total αꢀtubulin, as determined by tubulin ELISA, in response to compound treatment, with all samples run in at least triplicate; compound
∧
39
tested at 10 nM; previously published data; *P < 0.05; **P < 0.01 compared with vehicle (DMSO)–treated controls as determined using an twoꢀtailed
unpaired tꢀtest; NS: not significant; ND: not determined.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 79
Inhibition of prostaglandin biosynthetic pathway: Evaluation of test compounds as potential inhibitors
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of COXꢀdependent formation of PGD /PGE in RBLꢀ1 cells revealed that optimal inhibition of PG
2
2
synthesis is typically achieved when ring A is substituted in the paraꢀposition with either a chloride
e.g., 54) or methoxy group (e.g., 59). With respect to ring B, comparison of 54, 68, 76, 79, 84, and 98
(
clearly suggests that the presence of fluorine substituents is not detrimental, nor necessary, for inhibition
of PG production. Evaluation of the effect of different substitutions in the imidazole ring reveals that,
with the possible exception of compound 104, which shows a >50% inhibition in the PGD /PGE assay,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
tetrasubstituted compounds (e.g., 45, 46, 85, 88) are generally devoid of inhibitory activity against PG
synthesis. Conversely, 1,5ꢀdiarylimidazole congeners with one substitution at either C2 or C4 are
generally active, with the presence of a substituent at C4 being possibly preferred over the C2 position
(cf., 62 with 63, 64 with 66, 84 with 87, and 92 with 93). Among 1,5ꢀdiarylimidazoles that are not
substituted at C2 and C4, although these compounds appear to be generally less active than the
corresponding compounds bearing a substituent at C2 or C4, selected examples, such as 51, have been
identified that produce a near complete inhibition of the combined PGD /PGE synthesis in the RBLꢀ1
2
2
assay when tested at 10 µM.
Inhibition of leukotriene biosynthetic pathway: Finally, evaluation of compound inhibitory activity in
the LTB assay revealed that whereas a number of substitutions in the paraꢀposition of ring A appeared
4
to be well tolerated, the presence of a 2ꢀchloroꢀpyridine (see 73 and 74) prevented inhibition of LTB4
synthesis. In addition, similarly to what was observed for the SAR of PGD /PGE synthesis inhibition,
2
2
the fluorination of the B ring did not impact activity, as both fluorinated (e.g., 59, 80 and 84) and nonꢀ
fluorinated (e.g., 98) congeners exhibit comparable inhibitory activity in the LTB assay. However, in
4
contrast to what was observed for inhibitory activity against PG synthesis, the choice of substituents at
C2 and/or C4 of the imidazole ring did not play a critical role as examples of diꢀ (e.g., 47), triꢀ (e.g., 89
and 90) and tetraꢀsubstituted (e.g., 91) imidazoles were found to exhibit comparable suppression of
LTB synthesis in RBLꢀ1 cells at 10 µM.
4
ACS Paragon Plus Environment

Page 23 of 79
Journal of Medicinal Chemistry
Thus, in light of the observed elements of SAR for the three biological targets, the particular
1
2
substitution pattern decorating the 1,5ꢀdiarylimidazole scaffold can result in compounds with different
activity profiles, including multiꢀtargeted activity on MTs and/or eicosanoid biosynthesis (Figure 2).
Selected representative examples of multiꢀtargeted compounds were evaluated further to determine
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the IC values in the RBLꢀ1 assays and brain penetration (see Figure 3 and Table 2). Moreover, the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
compounds in Table 2 were also examined with improved LC/MS/MS methodologies for their ability
to inhibit the individual COXꢀderived products PGE and PGD , as well as a second 5ꢀLOXꢀderived
2
2
product, LTC (see Supplemental Figure 1, Supporting Information). Finally, two example
4
compounds underwent full concentrationꢀresponse analyses in the AcTub assay (see Supplemental
Figure 2, Supporting Information). Collectively, these studies confirmed that all Table 2 compounds
effectively inhibited both COX and 5ꢀLOX biosynthetic pathways with multiple compounds
exhibiting balanced multiꢀtargeted activity in the low µM range. Finally, consistent with the relatively
small size of these molecules, these compounds appeared to be generally brainꢀpenetrant.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Summary of SARs (top) and representative structures exemplifying the different activity
profiles exhibited by the 1,5ꢀdiarylimidazoles (bottom).
ACS Paragon Plus Environment

Page 25 of 79
Journal of Medicinal Chemistry
Table 2. Further evaluation of test compounds with multiꢀtargeted activity as MTꢀstabilizing agents
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and/or inhibitors of COX and 5ꢀLOX pathways.
PGD /E
LTB4
Inhibition
Acetylꢀ
Tubulin
Assay
2
2
Brain Plasma
Cpd# Structure Inhibition
IC (µM)
B/P
(
nM)
(nM)
IC (µM)
50
50
×
3.5
10 µM)
3.4
10 µM)
^
^
6
00
500
1.4
± 0.3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
9
12.1
4.7
13.4
23.7
9.7
± 200
± 300
(
(
(
(
×
^
^
4
±
00
200
1.9
± 0.2
35
200
± 100
×
4.1
5
6
9
0
2.4
ND
ND
ND
10 µM)
×
3.4
^
^
210
210
0.9
9.8
2.2
20.8
9.5
± 90
± 80
10 µM)
× 2.4
± 0.1
#
#
400
190
2.1
64
±
200
± 90
± 0.5
(10 µM)
2.1
×
#
#
#
240
260
0.9
6
7
6.5
2.9
7.8
2.4
± 100
± 90
± 0.2
(
10 µM)
× 2.0
#
#
700
300
2.8
± 0.8
68
±
500
± 200
(10 µM)
1.9
×
#
#
700
240
3.0
± 1.0
7
0
4.6
6.0
± 200
± 50
(
10 µM)
× 4.0
^
^
700
400
1.9
± 0.2
84
2.4
13.8
10.7
±
400
± 200
(10 µM)
4.8
10 µM)
3.6
10 µM)
×
#
#
900
480
1.9
± 0.4
8
7
15.0
± 300
± 60
(
(
×
^
^
300
110
2.5 ±
0.1
92
1.8
11.7
±
100
± 60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 79
^
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
200
^
3.3
± 0.1
Inactive
Inactive
60 ± 30
7
9
2.2
1.0
2.5
± 80
^
200
^
2.4
± 0.1
80 ± 20
98
12.5
±
60
All concentrationꢀresponse analyses were conducted at multiple concentrations with triplicate samples per
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
#
^
concentration. Study conducted after a 5 mg/kg i.p. injection of test compound; Study conducted after a 2
mg/kg i.p. injection of test compound with drug concentrations adjusted proportionally upward to allow
comparison to 5 mg/kg doses; ND: not determined. +/ꢀ errors represent standard deviations; all analyses were
conducted with groups of 3 mice
Figure 3. Representative doseꢀresponse curves in the RBLꢀ1 cell assays. Samples were run in
triplicate at each concentration, with the error bars representing standard error of the mean.
Discussion. Although selected cholinesterase inhibitors and a single NMDA receptor antagonist are
approved for the treatment of AD, these provide only a modest and temporary symptomatic
improvement in AD patients and thus far there have been no examples of diseaseꢀmodifying drug
4
3
candidates for AD with demonstrated efficacy in Phase III clinical trials. This situation may be due, at
ACS Paragon Plus Environment

Page 27 of 79
Journal of Medicinal Chemistry
least in part, to the fact that several neuropathological mechanisms are likely to provide potentially
alternative or redundant pathways to drive/sustain the neurodegenerative process in AD. NSAIDs, for
example, once attracted considerable attention as potential candidates for AD treatment due to
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
44ꢀ48
promising results in a series of preclinical studies in animal models with Aβ plaque pathology,
coupled with compelling retrospective epidemiological data indicating that individuals who were on
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
5, 49
sustained regimens of COXꢀinhibiting NSAIDs had reduced AD risk.
different COX inhibitors, however, yielded mostly disappointing results.
A series of clinical trials with
45, 50ꢀ52
Although there may be
multiple reasons for the lack of efficacy in these clinical trials, including issues with the particular
candidate compounds used in the studies and/or trials that were conducted in patient populations with
disease that was too advanced to respond to this therapeutic approach, or who had dementia that was not
due to underlying AD pathology, another possible mechanistic explanation may be that even if
substantial COX inhibition is achieved within the AD brain, this may not prevent the detrimental effects
of inflammatory LTs. Indeed, different studies have now shown that these 5ꢀLOX products are likely to
8
, 18, 20, 21, 23
contribute to Aβ plaque pathology,
and there is evidence that blocking COX metabolism can
53ꢀ55
lead to shunting of arachidonic acid to the 5ꢀLOX pathway.
Furthermore, it may be also argued that
any potential benefit of lowering Aβ production and plaque pathology alone may be limited if tau
pathology persists. Indeed, a recent retrospective analysis of several lateꢀstage AD clinical trials,
including the NSAIDs studies, concluded that candidate therapeutics affecting one target alone may not
43
be likely to impact significantly disease progression, suggesting that the evaluation and development
5
6
of multiꢀtargeted approaches for AD would be desirable. However, although different examples of
multiꢀtargeted constructs have been proposed as experimental AD treatments, there have been only few
5
7
reports of molecules that exhibit polypharmacology against targets that are likely involved in tau and
Aβ neuropathology.
Prior studies suggested that axonal transport deficits associated with a loss of MTs and/or altered MTꢀ
dynamics in the axon of neurons comprise a key neuropathological step of tauopathies and possibly
24
other neurodegenerative diseases. As part of our ongoing efforts to develop brainꢀpenetrant MTꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 79
stabilizing agents as potential treatments for these diseases, we identified selected heterocyclic
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
58
compounds that are not derived from natural products, as exemplified here by triazolopyrimidine 107
3
7, 39
(Table 1), which exhibit a promising combination of MTꢀstabilizing activity and brain penetration.
Among compound classes that are structurally and functionally related to the triazolopyrimidines are a
35
series of 1,5ꢀdiarylimidazoles. Like the triazolopyrimidines, these imidazoles are found to promote
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
MTꢀpolymerization in cell free assays and exhibit relatively potent and broadꢀspectrum fungicidal
35
activity. Furthermore, considering that these tricyclic compounds exhibit structural similarities with
known NSAIDs, they may be considered as possible starting points in the development of small
molecules that could concurrently stabilize MTs and lower eicosanoid production. Given that MT
2
4, 59
dynamics are believed to be affected in AD and other neurodegenerative disorders,
and considering
the involvement of both COXꢀ and 5ꢀLOX derived eicosanoids in AD and likely other
neurodegenerative diseases, a normalization of axonal MT dynamics, combined with a suppression of
eicosanoid production, may provide a potential multiꢀtargeted strategy to mitigate concurrent disease
mechanisms in AD and other neurodegenerative conditions.
Our multiꢀtargeted structureꢀactivity relationship (SAR) studies began with the observation that
3
5
selected examples from the previously reported fungicidal imidazoles, such as 2, exhibit evidence of
activity both as MTꢀstabilizing agents and inhibitors of eicosanoid biosynthesis in vitro, suggesting a
potentially significant overlap between the pharmacophore of these MTꢀstabilizing imidazoles and that
3
3
of NSAIDs, such as 3 (COXꢀ1 inhibitor ). Based on these initial findings, and taking into consideration
3
5
prior SAR studies that defined a set of preferred substitutions on the imidazole ring that are generally
important for fungicidal activity, and thus presumably for MTꢀstabilizing activity, we focused on
evaluating whether changes in the substitution patterns of 1,5ꢀdiarylimidazoles may result in compounds
with improved and balanced multiꢀtargeted activities. In particular, given that optimal broadꢀspectrum
fungicidal activity is typically achieved with tetrasubstituted imidazoles (e.g., 2) and other related
congeners that comprise a 4ꢀchlorophenyl group at N1, a chlorine atom in position 4, and a diꢀ or triꢀ
ACS Paragon Plus Environment

Page 29 of 79
Journal of Medicinal Chemistry
fluorinated phenyl group in position 5, we focused primarily on evaluating (a) monoꢀ and diꢀ
substitutions in the A ring; (b) chloroꢀ and/or alkylꢀsubstitutions in positions 2 and/or 4 of the imidazole
ring; and (c) the effect of varying the degree and pattern of fluorination in the B ring. As shown in Table
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
, these studies demonstrated that relatively small changes in the substitution pattern of 1,5ꢀ
diarylimidazoles can have profound effect in the activity profile of these molecules and ultimately led to
the identification of several compounds with MTꢀstabilizing activity and inhibitory activity against the
COX and/or 5ꢀLOX pathways (see Table 2). Furthermore, with respect to the MTꢀstabilizing activity,
our studies suggest that like the triazolopyrimidines, MTꢀstabilizing 1,5ꢀdiarylimidazoles are likely to
affect MT structures in different ways depending on the particular substitution pattern. In the case of
triazolopyrimidines, it was suggested that congeners bearing an alkoxy side chain in the fluorinated
phenyl ring (e.g., 108) typically produce an unusual cellular phenotype that is characterized by a bellꢀ
shaped concentrationꢀresponse relationship in the AcTub assay and proteasomeꢀdependent degradation
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
9
of αꢀtubulin. Interestingly, a similar type of observation is seen with the 1,5ꢀdiarylimidazoles, as all
MTꢀactive compounds bearing an alkoxy substituent in ring B (i.e., 2, 58, 95–97, Table 1) produce a
loss in αꢀtubulin at compound concentrations that are required to observe an increase in AcTub.
However, our results with the 1,5ꢀdiarylimidazoles clearly suggest that the substitution pattern of ring B
may not be the only determining factor in whether αꢀtubulin is affected, as other substitutions in the A
and C ring appear to play a role (e.g., compare 29 with 30, and 54 with 73, Table 1). As was noted for
3
9
the triazolopyrimidines, compounds found to cause a reduction in total αꢀtubulin may not be desirable
for neurodegenerative diseases as this effect would be expected to ultimately compromise rather than
restore axonal MT function. Conversely, certain multiꢀtargeted congeners, including examples listed in
Table 2 that do not appear to reduce total αꢀtubulin levels at concentrations that increase AcTub, may be
more promising candidates for tauopathy treatment. Further studies, however, will be necessary to
confirm that these compounds can produce beneficial effects in an animal model of tau pathology, as
60
observed previously with MTꢀstabilizing natural products like paclitaxel, epothilone D, and
4
ꢀ7
dictyostatin.
ACS Paragon Plus Environment