Structure-activity relationships of small molecule inhibitors of RAGE-Aβ binding

Article abstract of DOI:10.1016/j.tet.2013.05.079

The Receptor for Advanced Glycation Endproducts ('RAGE') mediates transport of amyloid-β peptide (Aβ) into the brain, and is therefore an important target for the development of therapeutic agents for Alzheimer's disease. We describe structure-activity relationships for inhibition of RAGE-Aβ binding, derived from the analysis of a library of tertiary amides.

Full text of DOI:10.1016/j.tet.2013.05.079

Tetrahedron 69 (2013) 7653e7658
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Structureeactivity relationships of small molecule inhibitors
of RAGE-Ab binding
,
,d,
*
Nathan T. Ross a, Rashid Deane b c, Sheldon Perry b, Benjamin L. Miller a
a Department of Biochemistry and Biophysics, University of Rochester, Rochester, NY 14642, USA
b Center of Neurodegernerative and Vascular Brain Disorders, University of Rochester, Rochester, NY 14642, USA
c Center for Translational Neuromedicine, University of Rochester, Rochester, NY 14642, USA
d Department of Dermatology, University of Rochester, Rochester, NY 14642, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The Receptor for Advanced Glycation Endproducts (‘RAGE’) mediates transport of amyloid-
into the brain, and is therefore an important target for the development of therapeutic agents for Alz-
heimer’s disease. We describe structureeactivity relationships for inhibition of RAGE-Ab binding, derived
from the analysis of a library of tertiary amides.
b peptide (Ab)
Received 28 February 2013
Received in revised form 18 May 2013
Accepted 20 May 2013
Available online 24 May 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Receptor for advanced glycation
endproducts
Amyloid beta peptide
Focused library
Alzheimer’s disease
1. Introduction
range of ligands via its extracellular V domain. In addition to the
advanced glycation endproduct (AGE) proteins from which RAGE
gets its name, these include the S100 family of proteins, ampho-
Demographic changes and longer life spans have combined to
steadily increase the number of people afflicted with Alzheimer’s
disease, and this trend is predicted to continue.1 Given that current
estimates put the global prevalence of dementia at 24 million,2 the
development of new therapeutic strategies for Alzheimer’s disease
is therefore an urgent challenge to the chemistry community.
The amyloid hypothesis holds that Alzheimer’s disease
terin/HMGB1, and, the particular focus of this research, A
upregulated in the brain vasculature of Alzheimer’s disease,11 and
mediates transport of A
across the bloodebrain barrier.12
We recently reported the results of a high-throughput screen for
b. RAGE is
b
inhibitors of RAGE-A
b binding that yielded compounds 2e4 as
‘hits’.13 Analysis of the structural features of these compounds led
to the synthesis of a focused library of compounds designed based
on a common pharmacophore hypothesis. One compound selected
results from the accumulation of amyloid-
b peptide (Ab) in the
brain.3 Many targets have been explored as potential anti-Alz-
heimer’s therapeutic strategies,4 including
b
-
and
g-
secretase
from this library, 1, was found to inhibit RAGE-A
and dramatically decreased the brain accumulation of A
b
binding in vitro,
in
5
6
inhibitors, A
b
-targeted immunotherapy,7 and interference with the
b
formation of neuronal tangles consisting of aggregates of hyper-
phosphorylated tau protein.8 Recent work suggests that protein
kinase C may be an important player in Alzheimer’s pathology as
well.9 While all of these have produced exciting results in in vitro
assays and in some cases have progressed as far as human clinical
trials, none has yet yielded an FDA-approved drug.
a mouse model of Alzheimer’s disease. Since our initial report, 2-
aminopyrimidine derivatives14 and trisubstituted thiazoles15 with
the ability to disrupt RAGE-Ab binding have been reported. Here,
we provide a detailed discussion of the design, synthesis, and ac-
tivity of the full focused library, and an analysis of the structural
features underlying inhibition of RAGE-Ab binding.
An emerging target for anti-Alzheimer’s drug development is
the Receptor for Advanced Glycation Endproducts (RAGE). A
member of the immunoglobulin superfamily,10 RAGE binds a broad
2. Results and discussion
As described previously,13 a commercial diversity library of 5000
compounds (Comgenex) was assayed for the ability to disrupt
RAGE-Ab interaction using RAGE transfected CHO (Chinese hamster
* Corresponding author. Tel.: þ1 585 275 9805; fax: þ1 585 273 1346; e-mail
address: benjamin_miller@urmc.rochester.edu (B.L. Miller).
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.05.079

7654
N.T. Ross et al. / Tetrahedron 69 (2013) 7653e7658
Â
Ã
ovary) cells and I125-A
b. The three most active ‘hit’ compounds
identified from this screen have a central tertiary amide core, an
electron rich aromatic group, a hydrophobic group, and two of the
three have electron deficient aromatic rings (Fig. 1). Based on the
RAGE bound I125A
b
Â
Ã
Binding IndexA
¼
ꢀ ½AbTotal
ꢁ
b
I125
A
b
in medium
Subtracting the binding index obtained in the presence of
compound from the binding index in the absence of compound
similarity between these compounds
a focused library was
designed that could be readily assembled in parallel using com-
mercially available building blocks. The library contained building
blocks representing the three similar peripheral functionalities of
the previous hits: an electron rich aromatic group, a hydrophobic
group, and an electron deficient benzene. Building blocks were
chosen with an eye toward reducing molecular weight relative to
2e4, while providing structures with the common features seen in
these molecules. The 100-compound library (supplemental Fig. 1)
was assembled using Borch reductive amination16 followed by
parallel acetylation (Scheme 1).
provided a differential binding index; inhibition of RAGE-A
binding yields a differential binding index >0.
A broad range of compound activities were observed in the
primary screening assay, with five compounds from the library
performing as well or better than the best hits from the initial
library screen (Fig. 2).
b
Cl
Cl
Cl
Cl
Cl
Cl
O
N
O
N
O
N
O
N
6
1
5
O
Cl
O
Cl
N+
N+
1
-O
-O
N
O
N
O
8
7
Fig. 2. Top hits from the 100-compound focused library.
Fig. 1. Hit compounds from a high-throughput screen for modulators of the RAGE-A
interaction.
b
Since various portions of a molecule may contribute synergis-
tically to binding (or may antagonize one another), examination of
individual building blocks is an oversimplification, but still useful in
establishing an overall picture of the SAR. To that end, Figs. 3e5
show screening data sorted by building block class (hydrophobic
amine, electron rich aromatic, electron poor benzene). Ability to
i) NaBH3CN,
CH3CH2OH
R''
O
R
R
NH2
O
N
R'
H
O
ii)
R''
inhibit RAGE-Ab binding is reported on the y-axis as a differential
binding index, as described above.
R'
Cl
The strongest structureeactivity correlation was observed for
the electron rich aromatic group (R0 in Scheme 1; Fig. 3). When the
benzaldehyde building block was used, compounds were on aver-
age more than two times more active than the average performance
of all library members. All five of the compounds with the strongest
R: Hydrophobic Amines
R': Electron-Rich Aromatics
O
O
NH2
H
NH2
H
H
O
ability to inhibit RAGE-Ab binding incorporated this building block.
O
NH2
O
NH2
H
This result is particularly striking when compared with the average
inhibitory potency of compounds incorporating either the 4-
fluorobenzaldehyde or phenylacetaldehyde building blocks. Con-
versely, the indole containing electron rich aromatic was deleteri-
ous to binding; no compounds containing this functionality had
significant activity in the assay. Examination of structureeactivity
relationships for the hydrophobic amine (R in Scheme 1; Fig. 4) and
electron deficient benzene (R00 in Scheme 1; Fig. 5) groups suggests
that RAGE tolerates a significant amount of structural variation in
these portions of the molecule. However, chloro substitution of the
electron deficient benzene was generally more favorable than nitro
substitution. As can be seen from surface electrostatic potential
maps generated from density functional calculations at the B3LYP/
6-31G(d) level (GAMESS17) the most electron deficient aromatic
moieties participate in both ‘best’ (3-nitro-4-chloro) and ‘worst’
(3,5-dinitro) inhibitors (Fig. 6); it is possible that there is either
H
F
N
H
R'': Electron-Deficient Benzenes
O-
O
O
O
O
O
N+
Cl
Cl
Cl
Cl
Cl
O
Cl
Cl
Cl
Cl
N+
O-
N+
O-
N+
O-
O
O
O
Scheme 1. Synthetic scheme for library generation, and library building blocks.
Following synthesis, library members were purified to an aver-
age purity of >95% by reverse phase HPLC, and screened in RAGE
transfected CHO cells for their ability to block I125-A
peptide
binding. A binding index for A was determined in the presence
and absence of compound, where:
b
b

N.T. Ross et al. / Tetrahedron 69 (2013) 7653e7658
7655
O
H
O
O
O
H
H
H
H
O
N
F
H
0.06
0.05
0.04
0.03
0.02
0.01
0
1
5
9
13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 101 105
-0.01
-0.02
-0.03
Compound Number
Fig. 3. Focused library screening results sorted by electron rich aromatic group.
NH2
NH2
NH2
NH2
0.06
0.05
0.04
0.03
0.02
0.01
0
1
4
7
10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 100
-0.01
-0.02
-0.03
Compound Number
Fig. 4. Focused library screening results sorted by hydrophobic amine group.
a steric component to the selection in this case, or alternatively
there is insufficient structural diversity to generate a strong SAR.
In an attempt to further understand SAR, hierarchical clustering
of library compounds was conducted (Fig. 7; Tanimoto similarity
>85%), which revealed additional insights for favorable RAGE
binding. Compounds that contained both an indole electron rich
aromatic and a mono or di-nitro electron deficient benzene were
least favored for binding (cluster 1). Compounds containing mono
or di-chloro substituted electron deficient benzenes with an indole
electron rich aromatic were slightly less disfavored (cluster 3). The
next most disfavored cluster (cluster 10) contained compounds
with a naphthyl electron rich aromatic and a nitro containing
electron deficient benzene. As with indole containing electron rich
aromatic groups, naphthyl containing compounds performed bet-
ter when paired with chloro containing electron deficient ben-
zenes. Clusters with average differential binding indices near 0.01
all contained either naphthyl electron rich aromatics with chloro
substituted electron deficient benzenes or nitro containing electron

7656
N.T. Ross et al. / Tetrahedron 69 (2013) 7653e7658
O-
O
O
O
O
O
N+
Cl
Cl
O
Cl
Cl
Cl
Cl
Cl
Cl
Cl
N+
N+
N+
O-
O
O-
O
O-
O
0.06
0.05
0.04
0.03
0.02
0.01
0
1
4
7
10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 100 103
-0.01
-0.02
-0.03
Compound Number
Fig. 5. Focused library screening results sorted by electron deficient benzene group.
rich aromatics with any of benzyl, phenyl, or phenethyl electron
rich aromatic groups (clusters 5, 7, 9, 17, and 19). Improved com-
pound binding was observed for clusters that contained electron
deficient benzenes with 3-nitro-4-chloro substitution (e.g., clusters
2 and 14) as well as compounds with dichloro substitution (e.g.,
clusters 4, 13, and 16). The best group (cluster 11), contains 3
compounds each with a para-chloro electron deficient benzene and
a cyclohexyl hydrophobic amine (Fig. 8).
In an effort to determine if other molecular features could be
correlated to compound activity, differential binding index values
were compared against common ‘drug-like’ properties. The only
significant correlation observed, while still quite weak, was be-
tween differential binding index and compound molecular weight
(R2¼0.156; Supplemental Fig. 3), while all other properties com-
pared were not significantly correlated (R2<0.1) (rotatable bond
count, amide count, atom count, H-bond donors, H-bond acceptors,
polar surface area, and clogp).
A key goal in the library design was to reduce molecular weight,
while retaining inhibitory potency. The average molecular weight
of the three best commercial compounds tested in the original li-
brary screen was 521 Da. The average molecular weight of the five
library members with best RAGE-Ab inhibitory activity was 362 Da.
The most active of these compounds weighs just 327 Da. As de-
scribed elsewhere,13 compound 1 showed a 52-fold increase in
brain uptake in vivo relative to 4, likely due at least in part to its
reduced molecular weight. Compound 1 also inhibits RAGE-
mediated influx of circulating Ab40 and Ab42 into the brain of
aged APPsw/0 mice, and normalizes their cognitive performance.
3. Conclusion
We have described the synthesis and analysis of a 100-
compound focused library targeting inhibition of RAGE-Ab bind-
ing. Results indicate that RAGE tolerates substantial structural
variability within the context of the overall pharmacophore
model; of the structures tested, those bearing an indole or nitro
Fig. 6. Electrostatic potential maps of electron deficient benzene derivatives included
in the focused library, visualized with MacMolPlt.18 ‘Best’ and ‘Worst’ indicate the
average inhibitory potency of library members incorporating these building blocks.

N.T. Ross et al. / Tetrahedron 69 (2013) 7653e7658
7657
Fig. 7. Compounds clustered in groups with >85% structural similarity, bars colored by average differential binding index per cluster. Hierarchical clustering with Tanimoto
similarities >85% reveals chemical features deleterious to effective RAGE binding.
deficient acid halides to produce the final tertiary amide library
with the three desired peripheral groups. A representative syn-
thesis is described below.
4.1. N-Benzyl-4-chloro-N-cyclohexylbenzamide (1)
To generate Compound 1, cyclohexylamine (0.5 M in MeOH) was
combined with benzaldehyde (0.5 M in MeOH) and stirred for 3 h at
64 ꢂC. The mixture was cooled to rt, followed by two additions of
sodium cyanoborohydride (0.5 M in EtOH), each followed by stir-
ring at rt for 30 m. The mixture was then heated to 64 ꢂC for 6 h. The
reaction mixture was worked-up with water and extracted three
times with CH2Cl2. The organic fractions were pooled, dried with
magnesium sulfate, and reduced in vacuo. The secondary amine
was then solubilized in dry CH2Cl2 and combined with 4-chloro-
benzoylchloride. Equivalents were based upon the assumption that
the secondary amine was formed in 100% yield. Dimethylamino-
pyridine (0.1 equiv) was solubilized in dry CH2Cl2 and added di-
rectly to the stirring solution of secondary amine, followed by
addition of diisopropylethylamine (1.1 equiv). The reaction was
then capped, purged with nitrogen gas, and stirred at rt overnight.
The reaction mixture was then reduced in vacuo to an oil, which
was resolubilized and purified using reverse phase preparative
Fig. 8. Representative compounds from hierarchical clustering based upon Tanimoto
HPLC (isocratic elution: 75% acetonitrile, 25% water). Following
preparative HPLC, compound purity was determined using reverse
phase analytical HPLC. Compounds were purified to an average
purity of greater than 95% (supplementary Table 1). For Compound
1: IR (thin film from CDCl3): 2935, 2858, 2246, 1624, 1495, 1418,
similarities of >85% revealing features associated with binding.
substituted electron deficient benzenes were worse than the par-
ent compounds, while phenyl electron rich aromatics and chloro
substituted electron deficient benzenes were preferred. Given the
broad range of targets bound by RAGE in vivo this promiscuity is
perhaps unsurprising, and structural studies will be essential in
order to understand the mode(s) of compound binding. While the
in vivo activity of compound 1 is encouraging, significant work
remains to produce a variant with better bioavailability. Efforts
along those lines are in progress in our laboratories.
1091, 908, 838, 734 cmꢃ1
;
1H NMR (400 MHz, CDCl3)
d
7.52 (4H),
7.36 (4H), 7.20 (2H), 4.55 (1H), 3.65 (1H), 1.80 (4H), 1.47 (4H), 1.06
(2H); 13C NMR (75 MHz, CDCl3)
171.2, 142.5, 139.0, 135.2, 128.7,
d
128.4, 127.7, 126.8, 77.17, 59.4, 59.4, 44.6, 32.0, 30.8, 25.7, 25.1, 25.1,
9.29; HRMS m/z calculated for (MþH); 328.1390, found: 328.1477.
4.2. Screening strategy
4. Experimental section
The 125I-A
reported.13 First, RAGE-CHO cells were incubated at 4 ꢂC for 3 h with
125I-A
40 (5 nM) in the absence or presence of library members at
a concentration of 10 M. At the end of the incubation period the
immobilized cells were washed with the cold non-radioactive
b40 binding assay in RAGE-CHO cells was performed as
To generate the 100-compound focused library, Borch reductive
amination16 was used to combine hydrophobic primary amines
with electron rich aromatic aldehydes to generate secondary
amines. These secondary amines were then combined with electron
b
m

7658
N.T. Ross et al. / Tetrahedron 69 (2013) 7653e7658
Sacchi, K. L.; Seubert, P. A.; Shopp, G. M.; Thorsett, E. D.; Tung, J. S.; Wu, J.; Yang,
medium to remove 125I-A
b
40, and lysed in a solution containing 1%
S.; Yin, C. T.; Schenk, D. B.; May, P. C.; Altstiel, L. D.; Bender, M. H.; Boggs, L. N.;
Britton, T. C.; Clemens, J. C.; Czilli, D. L.; Dieckman-McGinty, D. K.; Droste, J. J.;
Fuson, K. S.; Gitter, B. D.; Hyslop, P. A.; Johnstone, D. M.; Li, W. Y.; Little, S. P.;
Mabry, T. E.; Miller, F. D.; Audia, J. E. J. Neurochem. 2001, 76, 173e181.
7. Schenk, D.; Barbour, R.; Dunn, W.; Gordon, G.; Grajeda, H.; Guido, T.; Hu, K.;
Huang, J.; Johnson-Wood, K.; Khan, K.; Kholodenko, D.; Lee, M.; Liao, Z.; Lie-
berburg, I.; Motter, R.; Mutter, L.; Soriano, F.; Shopp, G.; Vasquez, N.; Vandevert,
C.; Walker, S.; Wogulis, M.; Yednock, T.; Games, D.; Seubert, P. Nature 1999, 400,
173e177.
NP-40 in 0.1 M NaCl at 37 ꢂC for 15 min. The radioactivity was de-
termined using 1470 Wallac Wizard (PerkinElmer, Meriden, CT)
gamma counter. The fraction of 125I-A
b40 that was bound to the cell-
surface RAGE was determined as previously reported.13
Acknowledgements
8. Wischik, C.; Bentham, P.; Wischik, D.; Seng, K. Alzheimer’s Dementia 2008, 4,
T167.
9. Khan, T. K.; Nelson, T. J.; Verma, V. A.; Wender, P. A.; Alkon, D. L. Neurobiol. Dis.
This work was supported by the ISOA/ADDF. We thank members
of the former Center of Neurodegenerative and Vascular Brain
Disorders (Department of Neurosurgery), University of Rochester,
for their help.
2009, 34, 332e339.
10. Neeper, M.; Schmidt, A. M.; Brett, J.; Yan, S. D.; Wang, F.; Pan, Y. C.; Elliston, K.;
Stern, D.; Shaw, A. J. Biol. Chem. 1992, 267, 14998e15004.
11. Yan, S. D.; Chen, X.; Fu, J.; Chen, M.; Zhu, H.; Roher, A.; Slattery, T.; Zhao, L.;
Nagashima, M.; Morser, J.; Migheli, A.; Nawroth, P.; Stern, D.; Schmidt, A. M.
Nature 1996, 382, 685e691.
Supplementary data
12. Deane, R.; Yan, S. D.; Submamaryan, R. K.; LaRue, B.; Jovanovic, S.; Hogg, E.;
Welch, D.; Manness, L.; Lin, C.; Yu, J.; Zhu, H.; Ghiso, J.; Frangione, B.; Stern, A.;
Schmidt, A. M.; Armstrong, D. L.; Arnold, B.; Liliensiek, B.; Nawroth, P.; Hofman,
F.; Kindy, M.; Stern, D.; Zlokovic, B. Nature Med. 2003, 9, 907e913.
13. Deane, R.; Singh, I.; Sagare, A. P.; Bell, R. D.; Ross, N. T.; LaRue, B.; Love, R.; Perry,
S.; Paquette, N.; Deane, R. J.; Thiyagarajan, M.; Zarcone, T.; Fritz, G.; Friedman,
A. E.; Miller, B. L.; Zlokovic, B. V. J. Clin. Invest. 2012, 122, 1377e1392.
14. Han, Y. T.; Choi, G.-I.; Son, D.; Kim, N.-J.; Yun, H.; Lee, S.; Chang, D. J.; Hong, H.-S.;
Kim, H.; Ha, H.-J.; Kim, Y.-H.; Park, H.-J.; Lee, J.; Suh, Y.-G. J. Med. Chem. 2012, 55,
9120e9155.
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.tet.2013.05.079.
References and notes
1. Sosa-Ortiz, A. L.; Acosta-Castillo, I.; Prince, M. J. Arch. Med. Res. 2012, 43,
600e608.
2. Mayeux, R.; Stern, Y. Cold Spring Harbor Perspect. Med. 2012, 2 a006239.
3. Walsh, D. M.; Selkoe, D. J. J. Neurochem. 2007, 101, 1172e1184.
4. Citron, M. Nat. Rev. Drug Discovery 2010, 9, 387e398.
15. Lee, Y. S.; Kim, H.; Kim, Y.-H.; Roh, E. J.; Han, H.; Shin, K. J. Bioorg. Med. Chem.
Lett. 2012, 22, 7555e7561.
16. Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am. Chem. Soc. 1971, 93, 2897e2904.
17. GAMESS v22 Feb 2006 (R5) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert,
S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S.;
Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem. 1993, 14,
1347e1363.
5. Sankaranarayanan, S.; Price, E. A.; Wu, G.; Crouthamel, M. C.; Shi, X. P.; Tu-
gusheva, K.; Tyler, K. X.; Kahana, J.; Ellis, J.; Jin, L.; Steele, T.; Stachel, S.; Coburn,
C.; Simon, A. J. J. Pharmacol. Exp. Ther. 2008, 324, 957e969.
6. Dovey, H. F.; John, V.; Anderson, J. P.; Chen, L. Z.; de Saint Andrieu, P.; Fang, L. Y.;
Freedman, S. B.; Folmer, B.; Goldbach, E.; Holsztynska, E. J.; Hu, K. L.; Johnson-
Wood, K. L.; Kennedy, S. L.; Kholodenko, D.; Knops, J. E.; Latimer, L. H.; Lee, M.;
Liao, Z.; Lieberburg, I. M.; Motter, R. N.; Mutter, L. C.; Nietz, J.; Quinn, K. P.;
18. MacMolPlt v7.2.1 Bode, B. M.; Gordon, M. S. J. Mol. Graphics Modeling 1998, 16,
133e138.