Design of pyridopyrazine-1,6-dione γ-secretase modulators that align potency, MDR efflux ratio, and metabolic stability

Article abstract of DOI:10.1021/acsmedchemlett.5b00070

Herein we describe the design and synthesis of a series of pyridopyrazine-1,6-dione γ-secretase modulators (GSMs) for Alzheimer's disease (AD) that achieve good alignment of potency, metabolic stability, and low MDR efflux ratios, while also maintaining favorable physicochemical properties. Specifically, incorporation of fluorine enabled design of metabolically less liable lipophilic alkyl substituents to increase potency without compromising the sp³-character. The lead compound 21 (PF-06442609) displayed a favorable rodent pharmacokinetic profile, and robust reductions of brain Aβ42 and Aβ40 were observed in a guinea pig time-course experiment.

Full text of DOI:10.1021/acsmedchemlett.5b00070

Letter
pubs.acs.org/acsmedchemlett
Design of Pyridopyrazine-1,6-dione γ‑Secretase Modulators that
Align Potency, MDR Efflux Ratio, and Metabolic Stability
Martin Pettersson,*,† Douglas S. Johnson,† John M. Humphrey,‡ Todd W. Butler,‡
Christopher W. am Ende,‡ Benjamin A. Fish,‡,§ Michael E. Green,† Gregory W. Kauffman,‡
Patrick B. Mullins,‡ Christopher J. O’Donnell,‡ Antonia F. Stepan,† Cory M. Stiff,‡
Chakrapani Subramanyam,‡ Tuan P. Tran,‡ Beth Cooper Vetelino,‡ Eddie Yang,‡ Longfei Xie,‡
Kelly R. Bales,† Leslie R. Pustilnik,‡ Stefanus J. Steyn,† Kathleen M. Wood,† and Patrick R. Verhoest†
†Pfizer Worldwide Research & Development, 610 Main Street, Cambridge, Massachusetts 02139, United States
‡Pfizer Worldwide Research & Development, Eastern Point Road, Groton, Connecticut 06340, United States
S
* Supporting Information
ABSTRACT: Herein we describe the design and synthesis of a series
of pyridopyrazine-1,6-dione γ-secretase modulators (GSMs) for
Alzheimer’s disease (AD) that achieve good alignment of potency,
metabolic stability, and low MDR efflux ratios, while also maintaining
favorable physicochemical properties. Specifically, incorporation of
fluorine enabled design of metabolically less liable lipophilic alkyl
substituents to increase potency without compromising the sp3-
character. The lead compound 21 (PF-06442609) displayed a favorable rodent pharmacokinetic profile, and robust reductions of
brain Aβ42 and Aβ40 were observed in a guinea pig time-course experiment.
KEYWORDS: Alzheimer’s disease, gamma secretase modulators, fluorine, lipophilic metabolism efficiency, LipMetE
lzheimer’s disease (AD) is a progressive neurodegener-
Aative disorder that is characterized by gradual loss of
memory, impaired speech and motor functions, and is
ultimately fatal. An estimated 5.2 million people currently
suffer from AD in the US alone, and it is a major unmet medical
need given the lack of disease modifying therapies to slow the
progression of the disease.1 Accumulation of brain amyloid
plaques is a characteristic feature of AD pathology. These
plaques consist primarily of the neurotoxic peptide amyloid β42
(Aβ42), which is formed via sequential cleavage of the amyloid
precursor protein (APP) by β-secretase (BACE) and γ-
secretase.2 Development of compounds that inhibit or
modulate these enzymes has therefore become a major focus
in the search for disease modifying drugs.
γ-Secretase inhibitors (GSIs) have progressed to phase II and
phase III clinical trials and demonstrated robust reduction of
Aβ42 in human cerebrospinal fluid (CSF).3 However, several
adverse events were observed, such as gastrointestinal toxicity,
increased occurrence of skin cancer, and negative effects on
cognition.4,5 These findings may in part be attributed to
inhibition of Notch signaling, which is critical for cell
differentiation.6 It has also been hypothesized that accumu-
lation of the C99 BACE cleavage product of APP could be a
contributing factor to the cognitive decline.7 In contrast to
GSIs, γ-secretase modulators (GSMs) do not inhibit the
proteolytic activity of γ-secretase, but rather shift the cleavage
site of APP to reduce formation of Aβ42 and Aβ40, in favor of
the shorter, less toxic species Aβ38 and/or Aβ37.8,9 Develop-
ment of GSMs has therefore emerged as a promising approach
for the treatment of AD. γ-Secretase is an intramembrane-
cleaving aspartyl protease and is composed of at least four
subunits; presenilin, nicastrin, Aph1, and Pen2.10 Through the
use of photoaffinity probes, we and others have demonstrated
that several classes of GSMs bind to allosteric sites on
presenilin that are distinct from the binding sites of GSIs.11,12
Examination of the patent literature points to significant
challenges in identifying potent GSMs within favorable CNS
physicochemical property space.13 We have previously
disclosed an amide series (generically represented by 2) of
GSMs using E2012 (1) as a scaffold (Scheme 1).14−16
However, efforts to align potency (IC50 < 100 nM) with
adequate physicochemical properties and acceptable absorp-
tion, distribution, metabolism, and excretion (ADME)
attributes were unfruitful in this series. Furthermore, stronger
inhibition of cytochrome P450 enzymes (CYP450) was noted
with more potent analogues. Our focus was therefore shifted
toward designing novel, conformationally restricted cores with
increased polarity.17,18 As previously published, merging the
hydroxy-substituted bicyclic core of 3 with pyridyl amide 2
afforded a novel pyridopyrazine-1,6-dione heterocyclic scaffold
(i.e., 4 and 5; Scheme 1) that maintained the key polar
interactions required for potency but was devoid of liabilities
associated with the phenol.17
Received: February 11, 2015
Accepted: March 27, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acsmedchemlett.5b00070
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Scheme 1. Evolution of the Pyridopyrazine-1,6-dione Series
Table 1. Activity of 9−14
Development of robust, parallel enabled chemistry allowed
for rapid optimization to afford pyridone 4,17,19 which has
moderate Aβ42-lowering activity in CHO APP cells (Aβ42 IC50
= 101 nM) and favorable physicochemical properties: cLogP =
3.3,20 central nervous system multiple parameter optimization
score (CNS MPO) = 4.6,21 and LipE = 3.9.22 Compound 4
displayed dose-dependent reduction of brain Aβ42 in a guinea
pig model, but additional gains in potency and in vivo efficacy
were ultimately needed. Ongoing parallel chemistry efforts
delivered naphthyl analogue 5, which displayed excellent in
vitro activity while maintaining good properties (Aβ42 IC50
=
15 nM; cLogP = 2.85). However, despite excellent human liver
microsomal stability, this compound exhibited high turnover in
rat liver microsomes (HLM CLint,app = 8.1 mL/min/kg; RLM
CLint,app = 469 mL/min/kg), which complicates in vivo efficacy
and safety studies in rodents.23,24 Herein, we describe our
efforts to optimize the pyridopyrazine-1,6-dione series to
combine potency under 10 nM with good ADME profile and
favorable physicochemical properties. The strategic use of
fluorine played a key role in achieving these objectives.
a
Aβ42 IC50 values were obtained in a whole cell assay using CHO
APPwt cells. Aβ42 IC50 values are the geometric mean of at least two
Through parallel medicinal chemistry using a previously
developed protocol (Scheme 2), we sought to identify
b
experiments. MDR efflux ratio using a MDR1/MDCK assay utilizing
MDCK cells transfected with the gene that encodes human P-
glycoprotein.26 Human liver microsome-derived scaled intrinsic
c
Scheme 2. Synthesis of 9−13 and 15−21
clearance.23
lowering activity by replacing a methyl group with a
trifluoromethyl substituent in this region of the molecule.17
The resulting analogue 11 achieved both design objectives;
not only was the Aβ42 IC50 improved from 302 to 18 nM, but
the MDR ER was reduced from 4.9 to 2.5 while maintaining a
cLogP value below 3.0. Similarly, library hit 12 incorporating a
difluoromethyl-substituted isoxazole exhibited a promising in
vitro profile. Using the same tactic of introducing a
trifluoromethyl group, potency was improved from 72 to 17
nM (compound 13), and the MDR ER was reduced from 3.7 to
2.1 while maintaining good stability in human liver microsomes
(CLint,app = 20.0 mL/min/kg). Finally, we explored a small set
of targets designed specifically as naphthyl replacements. These
efforts resulted in benzothiazole 14, which had potency
approaching that of 5 while maintaining good ADME and
properties (Aβ42 IC50 = 25 nM, MDR ER = 2.1, and cLogP =
3.0).
heterocyclic replacements for the naphthyl group of 5.17,25
Several interesting compounds emerged such as thiazole 9 and
isothiazole 10, which maintained moderate Aβ42-lowering
activity (Table 1). However, consistent with previous
observations, increased polarity in the right-hand section of
the molecule came at the expense of high MDR efflux ratio
(MDR ER).26 To address this issue, we explored replacing the
methyl group of 9 with an electron-withdrawing trifluoromethyl
group to reduce the electron density of the heterocycle. This
modification also had the potential to improve potency since
we had previously observed a dramatic improvement in Aβ42-
The lead compounds were then evaluated in vivo to examine
their pharmacokinetic (PK) profile. However, compounds 13
and 14 were found to have inadequate oral exposure when
dosed in rat at 5 mg/kg even though RLM stability had been
B
DOI: 10.1021/acsmedchemlett.5b00070
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
improved relative to 5 (5, 13, and 14 had RLM CLint,app = 469,
89.5, and 87.0 mL/min/kg, respectively). One concern with
these compounds was the presence of four aromatic rings
resulting in reduced sp3-character.27 The design strategy was
therefore shifted to improve this parameter while maintaining
good alignment of potency, physicochemical properties, and
ADME profile. Re-examination of the library data revealed
compounds with simple alkyl substituents in the ortho-position
that maintained reasonable activity and that could serve as
starting points for further optimization. Two of these included
t-butyl analogue 15 and cyclobutyl compound 16, which had
Aβ42 IC50 values of 128 and 73 nM, respectively (Table 2).
Metabolic stability was poor, which may be expected given that
the aliphatic substituents are susceptible to oxidative metabo-
lism.
structure−activity relationship (SAR) observations.17 Toward
this end, difluoro-substitution was installed on the cyclobutyl
group of 16, as well as an additional fluorine in the para-
position of the phenol since we have previously shown that
blocking this metabolic soft spot can lead to an improvement in
both HLM and RLM stability.17 The resultant compound 17
did indeed have significantly better metabolic stability (HLM
CLint,app = 17.3 vs 93.8 mL/min/kg for 16), but in vitro Aβ42-
lowering activity was only marginally improved from 73 to 59
nM.
To increase potency and metabolic stability of 15, one of the
methyl groups of the t-butyl substituent was replaced with a
trifluoromethyl group to afford 18. The resultant 2,2-
dimethyltrifluoroethyl moiety has previously been reported in
the literature as a t-butyl replacement with improved metabolic
stability.29 As in the case of 17, a fluorine atom was also
introduced in the para-position of the right-hand aromatic ring
to block this metabolically labile site. The resulting analogue 18
represented a significant advancement in that Aβ42-lowering
activity was enhanced from 128 to 30 nM, and HLM stability
was improved from 95.7 to 23.7 mL/min/kg while also
maintaining low MDR ER and acceptable lipophilicity. The
synthetic route to this custom phenol allowed ready access to
the corresponding monomethyl analogues (i.e., 19 and 20; see
Supporting Information). Interestingly, removing one of the
methyl groups of 18 followed by separation of enantiomers
resulted in a further gain in potency and LipE. Compound 19
had an Aβ42 IC50 value of 20 nM, whereas the enantiomer 20
was >4-fold less active. Metabolic stability was improved
further, which is consistent with a reduction in cLogP. As
shown in Table 2, fluorine appears to play a key role in
improving the potency of 15 to ultimately deliver 19; however,
in the absence of an X-ray structure of γ-secretase, the nature of
the specific molecular interactions remains unknown.
We hypothesized that clearance could be improved by
introducing fluorine atoms to block potential sites of oxidative
metabolism. Furthermore, incorporation of fluorine into this
region would likely improve potency based on previous
Table 2. Activity of 15−20
While compounds 11, 13, and 14 (Table 1) relied on a
fourth aromatic ring to achieve good potency, incorporation of
fluorine allowed successful design of potent GSMs such as 19
(Table 2) that have reduced aromatic ring-count and increased
sp3-character.27 Specifically, incorporation of fluorine enabled
the creation of lipophilic aryl-substituents with improved
metabolic stability. Lipophilic metabolism efficiency (LipMetE)
is a useful parameter to gauge metabolic stability at a given
LogD (eq 1).28 Higher LipMetE is preferable because it
indicates that metabolic stability can be achieved within a wider
LogD range. This is particularly relevant when pursuing an
intramembrane target like γ-secretase, which may require higher
lipophilicity to achieve sufficient potency and in vivo efficacy.
As shown in Table 2, optimization of the ortho-alkyl substituent
and incorporation of fluorine at specific sites led to significant
improvements in metabolic stability while keeping lipophilicity
relatively unchanged. This is reflected in an increase in
LipMetE as exemplified by compounds 15 and 18, which
have LipMetE values of 0.9 and 1.8, respectively.28
LipMetE = LogD − Log10(CLint,u
)
(1)
7.4
While optimizing the terminal aryl ring we also explored
introduction of conformational control elements on the
methylene linker to orient the aryl ether into the putative
hydrophobic pocket of the binding site. A methyl scan revealed
that potency could be improved about 3-fold by introducing an
S-methyl substituent as shown in 21 (Table 3; the
corresponding R-enantiomer was 6-fold less active than 21,
not shown). Lead GSM 21 (PF-06442609)30 represented a
a
Aβ42 IC50 values were obtained in a whole cell assay using CHO
APPwt cells. Aβ42 IC50 values are the geometric mean of at least two
experiments. MDR efflux ratio using a MDR1/MDCK assay utilizing
b
MDCK cells transfected with the gene that encodes human P-
glycoprotein.26 Human liver microsome-derived scaled intrinsic
c
clearance.23 Lipophilic efficiency.22 Lipophilic metabolism effi-
d
e
ciency.28
C
DOI: 10.1021/acsmedchemlett.5b00070
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 3. Pharmacokinetic Parameters
In Vitro Potency/Selectivity
Aβ42 IC50
6 nM
NICD IC50
>10 μM
Physicochemical Properties
cLogP/SFLogD
3.1/3.4
4.8/2.0
4.0
LipE/LipMetE
CNS MPO
a
solubility (pH 6.5)
101 μM
Figure 1. Guinea pig time course study.
In Vitro ADME
HLM CLint,app
RLM CLint,app
RRCK Papp,A→B
MDR ER
12.7 mL/min/kg
69.6 mL/min/kg
16.0 × 10−6 cm/s
1.4
Aβ38 was increased 46% in accordance with the typical profile
of a GSM. The corresponding unbound brain and plasma
exposures at the 3 h time-point were 225 45 and 680 152
nM, respectively (see Supporting Information for further details
on efficacy and exposure data).
Rat PK
b
B/P
0.8
While the initial synthesis of 21 (Scheme 2, using 7) was
adequate for producing milligram quantities for in vitro
screening, it was not suitable for large scale synthesis to
support in vivo studies. Steric bulk imparted by the 1,1,1-
trifluoropropan-2-yl substituent resulted in a sluggish chloride
displacement reaction that was accompanied by formation of a
side product resulting from elimination of 7 rather than
alkylation. This prompted our investigation into an alternative
disconnection giving rise to amine 27 and lactone 29 (Scheme
3). This strategy offered a more convergent approach, allowing
for C−O bond formation between phenol 25 and the readily
available (S)-N-Boc-alaninol. The final step in generating 21
was envisioned as a lactone-to-lactam conversion utilizing this
chiral amine.
Our synthesis of the requisite phenol 25 began with Suzuki
coupling of commercially available boronic acid 22 with 2-
bromo-3,3,3-trifluoropropene to afford the styrene derivative
23. Demethylation of the methoxy group using BBr3 afforded
phenol 24, and asymmetric catalytic hydrogenation with
([RuCl(p-cymene)(S)-Segphos)]Cl) delivered the chiral phe-
nol 25 in excellent ee (96% ee; 72% yield).34 Alkylation of 25
with (S)-N-Boc-alaninol methanesulfonate, followed by Boc
deprotection and isolation of the amine as a solid fumarate salt
completed the synthesis of amine 27.
The highly crystalline lactone 29 was prepared from the
known pyridone carboxylic acid 2817 via bis-alkylation with 1,2-
dibromoethane. The amide 30 could then be obtained through
simple amidation with 27 in hot methanol. However, to achieve
acceptable conversion it was necessary to use superstoichio-
metric quantities (>3 equiv) of the precious amine 27, which
was undesirable for large scale synthesis. Instead, the amidation
was more efficiently carried out using the Lewis acidic DABAL-
Me3 reagent to afford 30 in 72% yield. Lactam formation to
generate the final target 21 was subsequently achieved in a one
pot procedure through in situ generation of a mixed anhydride
with trifluoroacetic anhydride followed by intramolecular
alkylation of the amide by treatment with DBU.
bc
,
C
b,u/Cp,u
0.4
d
CL
34.1 mL/min/kg
1.15 h
d
T1/2
e
F
51%
Kinetic solubility was measured at Analiza, Inc.35 Determined from
10 mg/kg oral dose, 1 h time point. Plasma and brain free fractions of
21 in rat were 2.7% and 1.3%, respectively. Determined from 1 mg/
kg intravenous dose. Calculated using the exposure from 1 mg/kg
a
b
c
d
e
intravenous dose and 5 mg/kg oral dose.
major milestone in that single-digit potency (Aβ42 IC50 = 6
nM) was successfully aligned with excellent microsomal
stability, good passive permeability,31 and low MDR efflux
ratio (HLM CLint,app = 12.7 mL/min/kg; RRCK Papp,A→B = 16.0
× 10−6 cm/s;31 MDR ER = 1.4) while maintaining favorable
physicochemical properties (cLogP = 3.1; SFLogD = 3.4;32
CNS MPO = 4.0). This is further highlighted by the fact that
21 has one of the highest LipE and LipMetE values of the
compounds explored in this series (LipE = 4.8; LipMetE = 2.0).
In addition to good HLM stability, 21 also had significantly
improved stability in rat liver microsomes as compared to initial
lead 5 (RLM CLint,app = 69.6 vs 469 mL/min/kg for 21 and 5,
respectively), which contributed to acceptable rat pharmacoki-
netic parameters (CL = 34.1 mL/min/kg; F = 51%; Table 3).
In addition, 21 achieved good brain penetration in rat as
indicated by a brain to plasma ratio (B/P) of 0.8 and an
unbound brain to unbound plasma ratio (Cb,u/Cp,u) of 0.4.
Finally, GSM 21 exhibited only moderate hERG activity (IC50
= 9.2 μM),33 and it was devoid of Notch inhibition (NICD IC50
> 50 μM),17 which is consistent with the desired profile of a
GSM.
In vivo efficacy of 21 was evaluated in a guinea pig time-
course experiment. Compound 21 was administered orally at a
dose of 60 mg/kg, and Aβ42, Aβ40, Aβ38, and Aβ-total were
measured along with compound exposure at time-points
ranging from 1 to 24 h. Sustained reductions of both Aβ42
and Aβ40 were achieved, while Aβ-total remained relatively
unchanged (Figure 1). A 41% reduction of brain Aβ42 was
observed at the 3 h time-point for the 60 mg/kg dose, whereas
We have described the design strategies and synthetic efforts
leading to advanced GSM 21 using pyridopyrazine-1,6-dione
leads 4 and 5 as starting points. Strategic use of fluorine played
D
DOI: 10.1021/acsmedchemlett.5b00070
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
a
Scheme 3. Synthesis of 21
a
Reagents and conditions: (a) 2-bromo-3,3,3-trifluoroprop-1-ene, (PPh3)2PdCl2, K2CO3, THF, H2O, rt, 70%; (b) BBr3, CH2Cl2, −78 °C to rt, 74%;
(c) 75 psi H2, [RuCl(p-cymene)(S)-Segphos)]Cl, EtOH, 50 °C, 72%, 96% ee; (d) (S)-N-(t-butoxycarbonyl)alaninol methanesulfonate, Cs2CO3,
DMF, 60 °C; (e) TFA, CH2Cl2, rt, then fumaric acid in MeOH, 78% (2 steps from 26); (f) 1,2-dibromoethane, Cs2CO3, DMF, 90 °C, 76%; (g) 27,
DABAL-Me3, THF, 70 °C, 72%; (h) TFAA (3.5 equiv), DBU (8 equiv), CH3CN, 0 °C to rt, 85%.
a key role in aligning potency and ADME properties such as
reducing the MDR efflux ratio for GSMs incorporating a
heterocyclic D-ring. Specific introduction of fluorine also
enabled creation of metabolically less liable lipophilic aryl
substituents leading to superior in vitro potency and HLM
stability while maintaining good physicochemical properties
and increased sp3-character. This is reflected in improved LipE
relative to 4: compounds 4 and 21 have whole-cell LipE values
of 3.9 and 4.8, respectively. LipMetE was utilized to gauge
progress with respect to metabolic stability by normalizing for
changes in lipophilicity. Compound 21 had a LipMetE value of
2.0, which was the highest in the series and compares favorably
to the initial t-butyl substituted library hit 15 (LipMetE = 0.9).
Finally, 21 demonstrated good rodent brain penetration and
oral bioavailability as well as robust in vivo Aβ42- and Aβ40-
lowering in guinea pig. Additional efforts focused on further
improving in vivo efficacy and reducing the efficacious plasma
concentration will be disclosed in a future publication.
REFERENCES
■
(1) Thies, W.; Bleiler, L. Alzheimer’s disease facts and figures.
Alzheimers Dement. 2013, 9, 208−245.
(2) Karran, E.; Mercken, M.; De Strooper, B. The amyloid cascade
hypothesis for Alzheimer’s disease: an appraisal for the development of
therapeutics. Nat. Rev. Drug Discovery 2011, 10, 698−712.
(3) Bateman, R. J.; Siemers, E. R.; Mawuenyega, K. G.; Wen, G.;
Browning, K. R.; Sigurdson, W. C.; Yarasheski, K. E.; Friedrich, S. W.;
Demattos, R. B.; May, P. C.; Paul, S. M.; Holtzman, D. M. A γ-
secretase inhibitor decreases amyloid-β production in the central
nervous system. Ann. Neurol. 2009, 66, 48−54.
(4) De Strooper, B. Lessons from a failed γ-secretase Alzheimer trial.
Cell 2014, 159, 721−726.
(5) Coric, V.; van Dyck, C. H.; Salloway, S.; Andreasen, N.; Brody,
M.; Richter, R. W.; Soininen, H.; Thein, S.; Shiovitz, T.; Pilcher, G.;
Colby, S.; Rollin, L.; Dockens, R.; Pachai, C.; Portelius, E.;
Andreasson, U.; Blennow, K.; Soares, H.; Albright, C.; Feldman, H.
H.; Berman, R. M. Safety and tolerability of the γ-secretase inhibitor
avagacestat in a phase 2 study of mild to moderate Alzheimer disease.
Arch. Neurol. 2012, 69, 1430−1440.
(6) Haapasalo, A.; Kovacs, D. M. The many substrates of presenilin/
γ-secretase. J. Alzheimer’s Dis. 2011, 25, 3−28.
(7) Mitani, Y.; Yarimizu, J.; Saita, K.; Uchino, H.; Akashiba, H.;
Shitaka, Y.; Ni, K.; Matsuoka, N. Differential effects between γ-
secretase inhibitors and modulators on cognitive function in amyloid
precursor protein-transgenic and nontransgenic mice. J. Neurosci.
2012, 32, 2037−2050.
(8) Weggen, S.; Eriksen, J. L.; Das, P.; Sagi, S. A.; Wang, R.; Pietrzik,
C. U.; Findlay, K. A.; Smith, T. E.; Murphy, M. P.; Bulter, T.; Kang, D.
E.; Marquez-Sterling, N.; Golde, T. E.; Koo, E. H. A subset of NSAIDs
lower amyloidogenic Aβ42 independently of cyclooxygenase activity.
Nature 2001, 414, 212−216.
(9) Kounnas, M. Z.; Danks, A. M.; Cheng, S.; Tyree, C.; Ackerman,
E.; Zhang, X.; Ahn, K.; Nguyen, P.; Comer, D.; Mao, L.; Yu, C.;
Pleynet, D.; Digregorio, P. J.; Velicelebi, G.; Stauderman, K. A.;
Comer, W. T.; Mobley, W. C.; Li, Y. M.; Sisodia, S. S.; Tanzi, R. E.;
Wagner, S. L. Modulation of γ-secretase reduces β-amyloid deposition
in a transgenic mouse model of Alzheimer’s disease. Neuron 2010, 67,
769−780.
(10) De Strooper, B.; Iwatsubo, T.; Wolfe, M. S. Presenilins and γ-
Secretase: Structure, Function, and Role in Alzheimer Disease. Cold
Spring Harb. Perspect. Med. 2012, 2, a006304.
(11) Pozdnyakov, N.; Murrey, H. E.; Crump, C. J.; Pettersson, M.;
Ballard, T. E.; am Ende, C. W.; Ahn, K.; Li, Y.-M.; Bales, K. R.;
Johnson, D. S. γ-Secretase modulator (GSM) photoaffinity probes
reveal distinct allosteric binding sites on presenilin. J. Biol. Chem. 2013,
288, 9710−9720.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental data for compound synthesis and character-
ization, assay protocols, in vivo efficacy, and exposure data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Tel: (617) 395-0705. E-mail: martin.pettersson@pfizer.com.
Present Address
§1723 Dove Cottage Drive, Charlotte, North Carolina 28226,
United States.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Stacey Becker, Emily Miller, Michael Marconi, Emily
Sylvain, and Karin Wallace for their contributions to the in vivo
studies. We also thank the Pfizer ADME technology group for
generating the in vitro pharmacokinetic data to support the
SAR efforts.
E
DOI: 10.1021/acsmedchemlett.5b00070
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
(12) Crump, C. J.; Johnson, D. S.; Li, Y.-M. Development and
mechanism of γ-secretase modulators for Alzheimer’s disease.
Biochemistry 2013, 52, 3197−3216 and references therein.
(13) Pettersson, M.; Stepan, A. F.; Kauffman, G. W.; Johnson, D. S.
Novel γ-secretase modulators for the treatment of Alzheimer’s disease:
a review focusing on patents from 2010 to 2012. Expert. Opin. Ther.
Pat. 2013, 23, 1349−1366.
(14) Pettersson, M.; Johnson, D. S.; Subramanyam, C.; Bales, K. R.;
am Ende, C. W.; Fish, B. A.; Green, M. E.; Kauffman, G. W.; Lira, R.;
Mullins, P. B.; Navaratnam, T.; Sakya, S. M.; Stiff, C. M.; Tran, T. P.;
Vetelino, B. C.; Xie, L.; Zhang, L.; Pustilnik, L. R.; Wood, K. M.;
O’Donnell, C. J. Design and synthesis of dihydrobenzofuran amides as
orally bioavailable, centrally active γ-secretase modulators. Bioorg. Med.
Chem. Lett. 2012, 22, 2906−2911.
(15) For related approaches, see Hall, A.; Patel, T. R. γ-Secretase
modulators: current status and future directions. Prog. Med. Chem.
2014, 53, 101−145 and references therein.
(16) Gijsen, H. J.; Mercken, M. γ-Secretase modulators: can we
combine potency with safety? Int. J. Alzheimer’s Dis. 2012, 2012,
295207.
(27) Lovering, F.; Bikker, J.; Humblet, C. Escape from flatland:
increasing saturation as an approach to improving clinical success. J.
Med. Chem. 2009, 52, 6752−6756.
(28) LipMetE was calculated using measured shake flask LogD32 and
unbound intrinsic clearance (CLint,u) derived from the intrinsic
apparent clearance (CLint,app) and calculated human liver microsomal
binding values. For additional details see Stepan, A. F.; Kauffman, G.
W.; Keefer, C. E.; Verhoest, P. R.; Edwards, M. Evaluating the
differences in cycloalkyl ether metabolism using the design parameter
″lipophilic metabolism efficiency″ (LipMetE) and a matched
molecular pairs analysis. J. Med. Chem. 2013, 56, 6985−6990.
(29) Tanaka, H.; Shishido, Y. Synthesis of aromatic compounds
containing a 1,1-dialkyl-2-trifluoromethyl group, a bioisostere of the
tert-alkyl moiety. Bioorg. Med. Chem. Lett. 2007, 17, 6079−6085.
(30) Compound 21, also known as PF-06442609, is now
commercially available from Sigma Aldrich; catalog #PZ0258.
(31) Callegari, E.; Malhotra, B.; Bungay, P. J.; Webster, R.; Fenner, K.
S.; Kempshall, S.; LaPerle, J. L.; Michel, M. C.; Kay, G. G. A
comprehensive non-clinical evaluation of the CNS penetration
potential of antimuscarinic agents for the treatment of overactive
bladder. Br. J. Clin. Pharmacol. 2011, 72, 235−246.
(17) Pettersson, M.; Johnson, D. S.; Subramanyam, C.; Bales, K. R.;
am Ende, C. W.; Fish, B. A.; Green, M. E.; Kauffman, G. W.; Mullins,
P. B.; Navaratnam, T.; Sakya, S. M.; Stiff, C. M.; Tran, T. P.; Xie, L.;
Zhang, L.; Pustilnik, L. R.; Vetelino, B. C.; Wood, K. M.; Pozdnyakov,
N.; Verhoest, P. R.; O’Donnell, C. J. Design, synthesis, and
pharmacological evaluation of a novel series of pyridopyrazine-1,6-
dione γ-secretase modulators. J. Med. Chem. 2014, 57, 1046−1062.
(18) Pettersson, M.; Johnson, D. S.; Humphrey, J. M.; am Ende, C.
W.; Evrard, E.; Efremov, I.; Kauffman, G. W.; Stepan, A. F.; Stiff, C.
M.; Xie, L.; Bales, K. R.; Hajos-Korcsok, E.; Murrey, H. E.; Pustilnik, L.
R.; Steyn, S. J.; Wood, K. M.; Verhoest, P. R. Discovery of indole-
derived pyridopyrazine-1,6-dione γ-secretase modulators that target
presenilin. Bioorg. Med. Chem. Lett. 2015, 25, 908−913.
(32) Shake flask LogD. Assay method adapted from published
protocol: Hay, T.; Jones, R.; Beaumont, K.; Kemp, M. Modulation of
the partition coefficient between octanol and buffer at pH 7.4 and pKa
to achieve the optimum balance of blood clearance and volume of
distribution for a series of tetrahydropyran histamine type 3 receptor
antagonists. Drug Metab. Dispos. 2009, 37, 1864−1870.
(33) Kutchinsky, J.; Friis, S.; Asmild, M.; Taboryski, R.; Pedersen, S.;
Vestergaard, R. K.; Jacobsen, R. B.; Krzywkowski, K.; Schroder, R. L.;
Ljungstrom, T.; Helix, N.; Sorensen, C. B.; Bech, M.; Willumsen, N. J.
Characterization of potassium channel modulators with QPatch
automated patch-clamp technology: system characteristics and
performance. Assay Drug Dev. Technol. 2003, 1, 685−693.
(34) RuCl(p-cymene)(R)-dm-segphos)Cl, CAS [944451−30−3].
(35) Solubility was measured by Analiza, Inc., Cleveland, OH
(http://analiza.com/physchem/logd.html#qf).
(19) Tran, T. P.; Mullins, P. B.; am Ende, C. W.; Pettersson, M.
Synthesis of pyridopyrazine-1,6-diones from 6-hydroxypicolinic acids
via a one-pot coupling/cyclization reaction. Org. Lett. 2013, 15, 642−
645.
(20) Values for cLogP were calculated using the BIOBYTE (www.
biobyte.com) program cLogP, version 4.3.
(21) Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A. Moving
beyond rules: the development of a central nervous system
multiparameter optimization (CNS MPO) approach to enable
alignment of druglike properties. ACS Chem. Neurosci. 2010, 1,
435−449.
(22) LipE = −log(EC50) − LogD where EC50 is in molar units and
LogD was experimentally determined using shake flask LogD.32 For
additional detail, see Ryckmans, T.; Edwards, M. P.; Horne, V. A.;
Correia, A. M.; Owen, D. R.; Thompson, L. R.; Tran, I.; Tutt, M. F.;
Young, T. Rapid assessment of a novel series of selective CB(2)
agonists using parallel synthesis protocols: a lipophilic efficiency
(LipE) analysis. Bioorg. Med. Chem. Lett. 2009, 19, 4406−4409.
(23) Assay method adapted from published protocols: Riley, R. J.;
McGinnity, D. F.; Austin, R. P. A unified model for predicting human
hepatic, metabolic clearance from in vitro intrinsic clearance data in
hepatocytes and microsomes. Drug. Metab. Dispos. 2005, 33, 1304−
1311.
(24) Obach, R. S. Prediction of human clearance of twenty-nine
drugs from hepatic microsomal intrinsic clearance data: An
examination of in vitro half-life approach and nonspecific binding to
microsomes. Drug. Metab. Dispos. 1999, 27, 1350−1359.
(25) Compound 14 was not prepared via the alkylation route in
Scheme 2, but rather using the HATU-mediated coupling/cyclization
as previously published: See Pfizer WO 2012/131539.
(26) Feng, B.; Mills, J. B.; Davidson, R. E.; Mireles, R. J.; Janiszewski,
J. S.; Troutman, M. D.; de Morais, S. M. In vitro P-glycoprotein assays
to predict the in vivo interactions of P-glycoprotein with drugs in the
central nervous system. Drug. Metab. Dispos. 2008, 36, 268−275.
F
DOI: 10.1021/acsmedchemlett.5b00070
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX