Synthesis, characterization, and PK/PD studies of a series of spirocyclic pyranochromene BACE1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2014.04.012

The development of 1,3,4,4a,5,10a-hexahydropyrano[3,4-b]chromene analogs as BACE1 inhibitors is described. Introduction of the spirocyclic pyranochromene scaffold yielded several advantages over previous generation cores, including increased potency, reduced efflux, and reduced CYP2D6 inhibition. Compound 13 (BACE1 IC₅₀ = 110 nM) demonstrated a reduction in CSF Aβ in wild type rats after a single dose.

Full text of DOI:10.1016/j.bmcl.2014.04.012

Bioorganic & Medicinal Chemistry Letters 24 (2014) 2477–2480
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis, characterization, and PK/PD studies of a series of
spirocyclic pyranochromene BACE1 inhibitors
a
a
a
b
b
Matthew Volgraf a, , Lina Chan , Malcolm P. Huestis , Hans E. Purkey , Michael Burkard , Mary Geck Do ,
Julie Harris b, Kevin W. Hunt b, Xingrong Liu a, Joseph P. Lyssikatos a, Sumeet Rana b, Allen A. Thomas b,
Guy P. A. Vigers b, Michael Siu a
⇑
a Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
b Array BioPharma, 3200 Walnut Street, Boulder, CO 80301, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The development of 1,3,4,4a,5,10a-hexahydropyrano[3,4-b]chromene analogs as BACE1 inhibitors is
described. Introduction of the spirocyclic pyranochromene scaffold yielded several advantages over pre-
vious generation cores, including increased potency, reduced efflux, and reduced CYP2D6 inhibition.
Compound 13 (BACE1 IC50 = 110 nM) demonstrated a reduction in CSF Ab in wild type rats after a single
dose.
Received 14 January 2014
Revised 1 April 2014
Accepted 4 April 2014
Available online 13 April 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
BACE1 inhibitors
Structure-based ligand design
Alzheimer’s disease
Alzheimer’s disease is characterized by the deposition of beta
amyloid (Ab) plaques and neurofibrillary tangles in the brain, lead-
ing to neurodegeneration, dementia, and ultimately death.1 The
amyloid hypothesis holds that the primary causative agent of the
disease is the Ab peptide, either as soluble oligomers or aggregated
fibers. As such, recent disease-modifying approaches have focused
on reducing or eliminating these Ab plaques.2 One approach
involves inhibiting the protease beta-site APP cleaving enzyme 1
(BACE1), which is responsible for the rate-limiting step during
the formation of Ab peptides.3
transpose the position of the A-ring oxygen to the d-position rela-
tive to the spirocyclic center, which molecular modeling suggested
might engage the N–H of a tryptophan residue (Trp76) in the
BACE1 binding site (Fig. 1). We envisioned that by maintaining
the same physicochemical property space as previous aminooxaz-
olines, while simultaneously improving BACE1 potency, we might
identify a compound suitable for benchmarking in our PK/PD assay
measuring the reduction of CSF Ab in wild type rats.
The presence of three contiguous chiral centers in the designs
made the synthesis and subsequent control of diastereoselectivity
a challenge within the d-oxa-pyranochromene series. Employing a
We have previously described spirocyclic BACE1 inhibitors
based on chroman4 and
c
-oxa-pyranochromene scaffolds5
readily available chiral building block (D-xylose) enabled the syn-
(Fig. 1). The most potent compounds within these series contained
spirocyclic acyl guanidines able to effectively engage the catalytic
aspartates of the protease, yet inhibitors featuring this warhead
invariably suffered from P-glycoprotein efflux ratios (P-gp ER) >3
(1, Fig. 1). It was found that the efflux ratios could be improved
by lowering the TPSA of the warhead using an aminooxazoline-
based warhead, albeit at the expense of BACE1 potency (2 and 3,
Fig. 1).
thesis of target compounds as single enantiomers (Scheme 1).
Towards this end, enantiomerically pure dihydropyran 46 was
reduced in the presence of triethylsilane and boron trifluoride
diethyl etherate to yield allylic acetate 5 in good yield. Following
deacetylation, nucleophilic aromatic substitution cleanly provided
ether 6. Reduction of the nitrile to the aldehyde and conversion to
the oxime provided key intermediate 7 which was primed for a
[3+2] cycloaddition across the unactivated olefin. Following
in situ formation of the nitrile oxide from the oxime using Koser’s
reagent,7 isoxazole tetracycle 8 was formed in excellent overall
yield. NAO bond cleavage8 and imine hydrolysis provided alcohol
9 which could be deoxygenated using a two-step elimination/
reduction protocol which also set the desired trans-ring junction
of the pyranochromenone 10. Tebbe olefination (?11) and
In an effort to improve the potency of the spirocyclic pyrano-
chromenes featuring aminooxazoline warheads, we sought to
⇑
Corresponding author. Tel.: +1 6504671305.
E-mail address: volgrafm@gene.com (M. Volgraf).
http://dx.doi.org/10.1016/j.bmcl.2014.04.012
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

2478
M. Volgraf et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2477–2480
"δ-Oxa-Pyranochromene"
Core:
"Chroman"
NH2
"γ-Oxa-Pyranochromene"
"Methyl-γ-Oxa-Pyranochromene"
NH2
NH2
NH2
O
O
N
N
N
O
N
N
N
N
R2
N
O
γ
O
Cl
Cl
O
Cl
A
B
O
C
β
δ
O
R1 = H, Me
R2 = Aryl
O
R1
O
O
NH
H
Me
1
2
3
Trp76
Goals:
BACE1 IC50
:
19 nM
140X
62
150 nM
510X
3.5
49 nM
1000X
8.4
Improve BACE1 potency by
transposing A-ring Oxygen to
engage an active siteTrp.
CatD Selectivity:
MDR1 Efflux Ratio:
cLogP / TPSA / CNS MPO:
3.9 / 80 / 4.1
2.0 / 78 / 5.4
2.3 / 78 / 5.3
Maintain desirable CNS
property space.
single enantiomer
single enantiomer
single enantiomer
Figure 1. BACE1 lead compounds and strategy for further optimization.
formation of the spriocyclic aminooxazoline using literature
conditions9,10 provided advanced intermediate 12 as a 1:1 mixture
of diastereomers about the spirocycle stereocenter. Suzuki
cross-coupling and separation of the diastereomers provided the
final targets 13–17 in enantiomerically pure form.
of the aminooxazoline warhead relative to the acyl guanidines,
resulting in accumulation of the more basic inhibitors in cellular
compartments where BACE1 is also localized.12 Meanwhile, selec-
tivity against cathepsin D (CatD), a key off-target implicated in
physiologically important roles, peaked at 180-fold.
The new cores demonstrated improved BACE1 enzyme potency
relative to the previously prepared spirocyclic pyranochromenes,
with the 3-chloro-5-fluorophenyl 17 affording an IC50 value of
62 nM (Table 1). Chloropyridine 13 also performed well in the
enzyme assay with an IC50 of 110 nM. However, the aminooxazo-
lines remained much less potent than the acyl guanidines previ-
ously synthesized on the program. Given the high likelihood of
P-gp efflux, acyl guanidines were not prepared with the new core.
A co-crystal structure of 13 with BACE1 confirmed our design
The in vitro DMPK profile of the new series, specifically P-gp
efflux ratios and predicted clearance in human liver microsomes
(HLMs), was critical in prioritizing new compounds for PD evalua-
tion (Table 1). The 3-chloro-5-fluorophenyl target 17 did not pos-
sess significant efflux (P-gp ER = 1.3), likely due to the increased
lipophilicity (cLogP = 3.3) and reduced polarity (TPSA = 66 Å) rela-
tive to the rest of the series. All other targets demonstrated P-gp
efflux (P-gp ERs >3), indicating an increased likelihood of impaired
brain exposure. Nevertheless, in vitro P-gp ERs were improved rel-
ative to the acyl guanidines previously described by our group.4,5
Meanwhile, HLMs for the new series were predicted to be stable
to moderately stable within the active stereochemical series (com-
pounds 13–17).
hypothesis, demonstrating
a unique water-bridged hydrogen
bonding network between the A-ring oxygen atom and Trp76
(Fig. 2A, PDB code: 4PZW). The tricyclic pyranochromene core
was oriented parallel to Tyr71, forming potentially favorable Van
der Waals interactions with the BACE1 ‘flap’ (Fig. 2B), while the
chloropyridine occupied the S3 pocket of the protease. The amino-
oxazoline engaged the catalytic aspartates (Asp32 and Asp228)
via hydrogen bonding and electrostatic interactions, similar to
other aminooxazoline-based BACE1 inhibitors reported in the
literature.11
During the course of our synthetic efforts, we also synthesized
the opposite A-ring trans stereoisomer of the pyranochromene
core. It was found that the opposite trans-ring junction greatly
impaired activity against BACE1, as observed in the comparison
of matched pairs 13 (IC50 = 110 nM) and 18 (IC50 = 1.0 lM). Crys-
tallography revealed the inability of this A-ring stereoisomer to
engage Trp76 via a water-mediated hydrogen bond (Fig. 2C, PDB
code: 4PZW) due to differences in A-ring conformation as well as
The cellular potencies were generally found to be more potent
than the enzyme IC50, presumably due to the increased basicity
OH
I
OTs
HO
1. Et3N/MeOH
Et3SiH,
1. DIBAL
N
2. NH2OH,
2. , Cs2CO3, DMF, 80 oC
BF3.OEt2
NC
O
Br
A
Br
DMF/MeOH, 0 oC
(88%)
OAc
OAc
NH4Cl
O
O
O
O
(77%)
(79%, 2 steps)
NC
A =
(93%, 2 steps)
OAc
O
H
H
4
5
Br
7
6
Ref. 6
(via D-Xylose)
F
O
O
O
N
OH
O
H
H
Burgess Reagent
2. Et3SiH, TFA
H
O
1.
Tebbe's Reagent,
THF, -78 oC
i. I2/AgOCN
H
H
Br
Br
Br
Br
Raney Ni
ii. NH4OH/THF
O
O
O
MeOH/H2O
(88%)
(87%)
(32%, 2 steps)
(76%)
O
O
11
O
H
H
8
10
9
NH2
NH2
O
O
ArB(OH)2,
Pd(PPh3)4, Na2CO3
H2O/Dioxane, 90 o
N
N
Br
Ar
C
O
O
O
O
H
H
12
13-17
mixture of diastereomers
Scheme 1. Synthesis of d-oxa-pyranochromene BACE1 inhibitors 13–17.

M. Volgraf et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2477–2480
2479
Table 1
Physicochemical and pharmacological properties of
c
-oxa-pyranochromene BACE1 inhibitors
NH2
NH2
NH2
O
O
O
N
N
N
R1
R1
R1
O
O
O
O
O
O
H
H
Me
Core:
A
B
C
c
Compound Core R1
cLogP/TPSA/CNS
BACE1 IC50
Cell IC50 APPWTd
M)
Enzyme/
cell
CatD/
BACE1
Papp
P-gp efflux
ratioe
HLMf (mL/
min/kg)
MPOb
(l
M)
(
l
(10À6 cm/s)
N
13
14
A
A
2.0/78/5.4
1.2/78/5.5
0.11
0.15
0.02
0.08
6.1
1.9
120
180
14
3.4
7.1
13
6
Cl
N
10
F
N
15
16
A
A
1.5/78/5.5
2.0/89/5.5
0.36
0.30
0.09
0.14
3.9
2.2
98
22
15
11
3.7
4.8
9
8
F
CN
F
17
A
3.3/66/5.1
0.06
1.0
0.11
0.6
56
11
1.3
10
Cl
Cl
Cl
N
N
18
19
B
2.0/78/5.4
2.3/78/5.3
0.13
ND
7.5
ND
100
0.5
4.7
10
12
15
15
Ca
16
4.8
a
b
c
d
e
f
Data for racemate.
CNS MPO score calculated using an experimentally derived pKa of 6.8 for compound 13.
Mean IC50 of at least two independent experiments.
Mean IC50 of at least two independent experiments for cellular Ab production.
(B to A)/(A to B) value derived for compounds (1
lM) evaluated in LLC-PK1 cells transfected with human MDR1 (P-gp).
Compounds (1 M) incubated with human liver microsomes for 20 min at 37 °C in the presence of NADPH.
l
ND = not determined.
potent stereoisomer did not inhibit CYP2D6 while the other 3 ster-
eoisomers prepared all displayed potent CYP2D6 inhibition of
<1 lM (Table 2). Taken together, this data suggests that CYP2D6
inhibition within this subseries is dependent on both core and
spirocenter stereochemistry. This observation coincides with the
well studied difference in CYP2D6 inhibition by the stereoisomers
quinidine and quinine, illustrating the delicate interplay between
stereochemistry and the 2D6 active site.13
We sought to improve the BACE1 enzyme potency of the d-oxa-
pyranochromene core via methyl substitution strategies. In the
related c-oxa-pyranochromene series, methyl substitution in the
b-position relative to the spirocyclic center improved BACE1
potency and CatD selectivity (2?3, Fig. 1). Preparation of the
matched pair 19 yielded a dramatic loss in potency (>16 lM). This
loss in potency may be explained by the stereochemical preference
of the pyranochromene A-ring observed between the two series
(e.g. 13: 110 nM vs 18: 1.0 lM). In each case, the methyl group
would occupy an opposite plane relative to the core in the active
site.
Figure 2. (A) Crystal structure of 13 bound to BACE1 active site (PDB code: 4PZW).
(B) Surface rendering of BACE1 active site with 13 bound. (C) Crystal structure of 18
bound to BACE1 active site (PDB code: 4PZX). (D) Surface rendering of BACE1 active
site with 18 bound.
We examined 3-chloropyridine 13 in a PD study comparing CSF
Ab in wild type rats with free brain concentrations of the inhibitor
(Fig. 3). Analysis of the free brain/free plasma ratios of 13 at the 3
and 5 h timepoints after a 60 mg/kg dose were 0.45 and 0.29,
respectively. Thus, despite predicted issues of P-gp efflux, brain/
plasma ratios observed in vivo were acceptable. A 40% reduction
in CSF Ab was achieved after 3 h, with free brain concentrations
of 13 reaching 250 nM, roughly 11-fold the cellular IC50 against
a less favorable Van der Waals contacts with Tyr71 (Fig. 2D),
accounting for the 10-fold loss in potency.
With these diastereomers in hand, we were surprised to dis-
cover a dramatic stereochemical dependence on CYP2D6 inhibition
within the pyranochromene series. Serendipitously, the most

2480
M. Volgraf et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2477–2480
Table 2
Stereochemical dependence on CYP2D6 inhibition of
c-oxa-pyranochromene BACE1 inhibitors
2500
0.3
Acknowledgments
CSF Aβ40 (pg/mL)
Brain Free Drug (μM)
The authors would like to thank Yanzhou Liu, Yutao Jiang, Chris
Hamman, Mengling Wong and Michael Hayes for analytical and
purification support and Melis Coraggio and Young G. Shin for bio-
analytical support.
0.25
0.2
2000
1500
1000
500
0
-32%
Supplementary data
0.15
0.1
-40%
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.
04.012. These data include MOL files and InChiKeys of the most
important compounds described in this article.
0.05
0
References and notes
Vehicle, 3 h 13 (60 mg/kg),
3 h
Vehicle, 5 h 13 (60 mg/kg),
5h
1. Kakob-Roetne, R.; Jacobsen, H. Angew. Chem., Int. Ed. 2009, 49, 3030.
2. Karran, E.; Mercken, M. M.; De Strooper, B. Nat. Rev. Drug Discov. 2011, 10, 698.
3. De Strooper, B.; Vassar, R.; Golde, T. Nat. Rev. Neurol. 2010, 6, 99.
4. Hunt, K. W.; Cook, A. W.; Watts, R. J.; Clark, C. T.; Vigers, G.; Smith, D.; Metcalf,
A. T.; Gunawardana, I. W.; Burkard, M.; Cox, A. A.; Geck Do, M. K.; Dutcher, D.;
Thomas, A. A.; Rana, S.; Kallan, N. C.; DeLisle, R. K.; Rizzi, J. P.; Regal, K.;
Sammond, D.; Groneberg, R.; Siu, M.; Purkey, H.; Lyssikatos, J. P.; Marlow, A.;
Liu, X.; Tang, T. P. J. Med. Chem. 2013, 56, 3379.
5. Thomas, A. A.; Hunt, K. W.; Volgraf, M.; Watts, R. J.; Liu, X.; Vigers, G. P. A.;
Smith, D.; Sammond, D.; Tang, T. P.; Rhodes, S. P.; Metcalf, A. T.; Brown, K. P.;
Otten, J. N.; Burkard, M.; Cox, A. A.; Geck Do, M. K.; Dutcher, D.; Rana, S.;
DeLisle, R. K.; Regal, K.; Wright, A. D.; Groneberg, R.; Scearce-Levie, K.; Siu, M.;
Purkey, H. E.; Lyssikatos, J. P.; Gunawardana, I. W. J. Med. Chem. 2014, 57, 878.
6. Matsuo, K.; Shindo, M. Tetrahedron 2011, 67, 971.
Figure 3. Rat pharmacodynamics of 13 (Ab1–40 lowering).
BACE1. The 5 h timepoint achieved 32% reduction with free brain
concentrations roughly 10-fold over the cellular IC50
.
In summary, structure-based design was used to optimize the
pyranochromene core of a series of BACE1 inhibitors. By transpos-
ing the position of an oxygen atom on the core to target a water-
bridged hydrogen bond with an active site tryptophan, potency
was improved while also maintaining physicochemical properties
akin to marketed CNS drugs.14 It was found that the relative config-
uration of the bicyclic ring system and the spirocyclic aminooxaz-
oline was important for both BACE1 potency and CYP2D6
inhibition. Free brain concentrations of chloropyridine 13 in a
PD-assay in wild type rat were in excess of the cellular IC50 by
greater than 10-fold and achieved a maximal Ab reduction in CSF
of 40% after 3 h. Despite achieving acceptable observed free
brain/free plasma ratios, the observed Ab reduction was not
sufficient to progress 13 beyond rat PD. Further work within the
aminooxazoline series focused on continuing to improve enzyme
potency and reducing P-gp efflux. The results of these efforts will
be disclosed in subsequent publications.
7. Raihan, M. J.; Kavala, V.; Kuo, C.-W.; Raju, R.; Yao, C.-F. Green Chem. 2010, 12,
1090.
8. Kozikowski, A. P.; Adamczyk, M. Tetrahedron Lett. 1982, 23, 3123.
9. Wittekind, R. R.; Rosenau, J. D.; Poos, G. I. J. Org. Chem. 1961, 26, 444.
10. Hassner, A.; Lorber, M. E.; Heathcock, C. H. J. Org. Chem. 1967, 32, 540.
11. Huang, H.; La, D. S.; Cheng, A. C.; Whittington, D. A.; Patel, V. F.; Chen, K.;
Dineen, T. A.; Epstein, O.; Graceffa, R.; Hickman, D.; Kiang, T.-H.; Louie, S.; Luo,
Y.; Wahl, R. C.; Wen, P. H.; Wood, S.; Fremeau, R. T., Jr. J. Med. Chem. 2012, 55,
9156.
12. Hilpert, H.; Guba, W.; Woltering, T. J.; Wostl, W.; Pinard, E.; Mauser, H.;
Mayweg, A. V.; Rogers-Evans, M.; Humm, R.; Krummenacher, D.; Muser, T.;
Schnider, C.; Jacobsen, H.; Ozmen, L.; Bergadano, A.; Banner, D. W.;
Hochstrasser, R.; Kuglstatter, A.; David-Pierson, P.; Fischer, H.; Polara, A.;
Narquizian, R. J. Med. Chem. 2013, 56, 3980.
13. Hutzler, J. M.; Walker, G. S.; Wienkers, L. C. Chem. Res. Toxicol. 2003, 16, 450.
14. Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A. ACS Chem. Neurosci. 2010, 1,
435.