SAR of tricyclic sulfones as γ-secretase inhibitors

Article abstract of DOI:10.1007/s11426-011-4388-6

Novel tricyclic sulfones as γ-secretase inhibitors have been reported by this laboratory for the treatment of Alzheimers disease. Compounds in this series have comparable or better in vitro activities and in vivo efficacies than sulfonamide analogues reported previously by this laboratory. Based on the previously reported tricyclic sulfone scaffold, additional SAR studies of C ring were carried out. Various C-ring structures including cyclohexane, pyran, and piperidine were tolerated. Additionally, the 7- and 8- positions of the C-ring were identified as the best sites to introduce substituent for modulating the pharmacokinetic properties of compounds from this series.

Full text of DOI:10.1007/s11426-011-4388-6

SCIENCE CHINA
Chemistry
November 2011 Vol.54 No.11: 1688–1701
doi: 10.1007/s11426-011-4388-6
• ARTICLES •
· SPECIAL TOPIC · In Honor of the 80th Birthday of Professor Ronald Breslow
SAR of tricyclic sulfones as -secretase inhibitors‡
XU Ruo*, COLE David, ASBEROM Ted, BARA Tom, BENNETT Chad, BURNETT Duane A., CLADER John,
DOMALSKI Martin, GREENLEE William J., HYDE Lynn, JOSIEN Hubert, LI HongMei, McBRIAR Mark,
McKITTRICK Brian, PISSARNITSKI Dmitri, QIANG Li, RAJAGOPALAN Murali, SASIKUMAR
Thavalakulamgara, SU Jing, TANG HaiQun, WU Wen-Lian, ZHANG LiLi & ZHAO ZhiQiang
Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, New Jersey 07033, USA
Received June 30, 2011; accepted July 28, 2011; published online October 11, 2011
Novel tricyclic sulfones as -secretase inhibitors have been reported by this laboratory for the treatment of Alzheimer’s disease.
Compounds in this series have comparable or better in vitro activities and in vivo efficacies than sulfonamide analogues re-
ported previously by this laboratory. Based on the previously reported tricyclic sulfone scaffold, additional SAR studies of C
ring were carried out. Various C-ring structures including cyclohexane, pyran, and piperidine were tolerated. Additionally, the
7- and 8- positions of the C-ring were identified as the best sites to introduce substituent for modulating the pharmacokinetic
properties of compounds from this series.
tricyclic sulfones, -secretase, Alzheimer’s disease, inhibitors
generated from -amyloid precursor peptides (APP) by a
sequential proteolytic cleavage catalyzed by - and -
secretases, inhibitors of these secretases should reduce A
concentration in patients, which have thus been proposed as
therapeutic targets for AD [3]. As the result, there are many
ongoing secretase inhibitor programs in various research
institutions which have led to numerous literature reports
about the secretase inhibitors for the treatment of AD [4].
We have previously reported the design and synthesis of
tricyclic sulfones as -secretase inhibitors. In the earlier
report, A and B rings of the tircyclic core were optimized
first, and the C ring was then introduced to generate tricy-
clic sulfones [5]. These new compounds have completely
eliminated the P-450 enzyme inhibition and liver toxicity
problems of our earlier sulfonamide -secretase inhibitors
[6], and also have shown greatly reduced mechanism based
side effects compared to our previous -secretase inhibitors.
As a result, this new series became the focus of the
-secretase inhibitor program at this laboratory and addi-
tional SAR studies were carried out. The strategy was to
systematically study and optimize the size and type of the C
1 Introduction
Alzheimer’s disease (AD) is a debilitating neurodegenera-
tive disease that affects a large population world wide and is
not well treated by existing drug therapies. One of the main
characteristics of AD is the formation of amyloid plaques in
the patient’s brain, composed of mostly A40/42 amyloid
-peptides. Extensive studies of this disease in the past few
decades have led to a generally accepted theory regarding
the cause and progression of AD, commonly known as the
amyloid hypothesis. This hypothesis states that AD is
caused by the overproduction and/or reduced clearance of
A peptides, which in turn leads to the formation of -
amyloid plaques, and subsequent neuronal cell death [1].
Based largely on the amyloid hypothesis, lowering the A
concentration in the brains of patients is proposed as one
potential treatment strategy to prevent, slow down, or even
reverse the progression of AD [2]. Furthermore, since A is
*Corresponding author (email: ruo.xu@merck.com)
‡Dedicated to Prof. Ronald Breslow on the occasion of his 80th birthday
© Science China Press and Springer-Verlag Berlin Heidelberg 2011
chem.scichina.com www.springerlink.com

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1689
ring first while keeping the previously optimized A and B
rings constant, and then introduce various substitutions on
all positions of C ring to further optimize the resulting
compounds. Herein, we report the results of these studies.
point the thioether of 5 was oxidized to the sulfone and then
the double bond and ester were reduced by catalytic hydro-
genation and lithium borohydride respectively to give com-
pound 6. It should be noted if the thioether of 1 was oxi-
dized to the sulfone first, oxidation of the hydroxyl group of
the resulting sulfone led only to elimination of the sulfone
group. Standard aforementioned ring closure reactions gave
five membered C ring compound 7. On the other hand, the
synthesis of cyclobutyl C ring started from vinyl sulfone 2.
The Michael addition of the lithium salt of t-butyl acetate to
2 gave ester 8. Reduction of the ester to hydroxyl compound
9, followed by standard ring closure reactions gave com-
pound 10 with cyclobutane as the C ring. The synthesis of
the cyclohexane C ring analogue took only three steps from
a simple sulfone 11 as shown in Scheme 3. Compound 11
was alkylated with 6-bromohexene first, then the double
bond of resulting compound was oxidized with MCPBA to
give epoxide 12. This epoxide underwent non-selective ring
opening and SNAr ring closure reactions sequentially in one
pot to give two rather complex tricyclic compounds 13 and
14.
2 Chemistry
We have previously optimized the A and B rings, and de-
veloped two versatile intermediates 1 and 2 that bear all the
desired functionalities of the optimized AB system [5]. Rel-
atively straightforward and scalable synthetic routes were
also developed to produce large quantities of these interme-
diates. As the result, all subsequent studies commenced
from one of these intermediates. The first task was to syn-
thesize compounds with various C ring sizes and the syn-
thetic routes shown in Schemes 13 were developed for this
purpose. Oxidation of thioether 1 to the corresponding sul-
fone with MCPBA, followed by mesylation of the hydroxyl
group and ring closure with potassium t-butoxide gave
compound 3 with cyclopropyl as the C ring. Alternatively,
oxidation of the hydroxyl group of compound 1 with
Dess-Martin reagent gave aldehyde 4, which underwent
Wittig reaction to form -unsaturated ester 5. At this
Six membered C rings with hetero atoms, such as tetra-
hydropyran and piperidine were also synthesized (Scheme
4) starting again from the versatile intermediates 1 and
Scheme 1 Reagents and conditions: (a) MCPBA ; (b) MsCl, NEt3; (c) t-BuOK; (d) Dess Martin reagent; (e) Ph3P=CHCOOMe; (f) PtO2/C, H2; (g) LiBH4.
Scheme 2 Reagents and conditions: (a) LiCH2COOt-Bu; (b) LiBH4; (c) MsCl, NEt3; (d) t-BuOK.

1690
Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
Scheme 3 Reagents and conditions: (a) 6-bromohexene, NaH; (b) MCPBA; (c) t-BuOK.
2. The hydroxyl group of 1 was alkylated with TBS pro-
tected bromoethanol to give compound 15. The thioether of
15 was oxidized to sulfone with MCPBA and then the
TBS protecting group was removed with TBAF to give 16.
Standard ring closure procedure gave bis-pyran 17. Fur-
thermore, previously mentioned compound 4 underwent
reductive amination with aminoethanol, the resulting sec-
ondary amine was protected by a Boc group, and the thi-
oether was then oxidized with MCPBA yielding compound
18. Standard C ring closure procedure yielded 8-piperidine
C ring compound 19. The Boc group of 19 was removed,
followed by standard functional group manipulation to give
additional derivatives as exemplified by 20. On the other
hand, 3-amino propanol was added directly to 2, which
formed pure trans adduct 21. Since the secondary amine at
7-position was much hindered, no protection for this amine
was needed for subsequent ring closure reactions. Once the
C-ring was formed, the piperidine nitrogen could be further
modified, yielding compounds such as 23.
In order to further modify the overall profile of these
compounds, we systematically introduced substitutions on
the C ring starting again from fore mentioned advanced
intermediate 2. The typical synthetic scheme to introduce
substitutions on cyclohexane C ring is shown in Scheme 5.
We have previously reported that vinyl sulfone 2 acted as
the dienophile of Deals-Alder reaction en route to our early
lead compound 24 which has substitution at C-8 position [5].
In order to make C-7 substituted compounds, another syn-
thetic route was developed as shown in Scheme 5. Methyl
5-(benzyloxy)pentanoate was deprotonated by butyllithium
Scheme 4 Reagents and conditions: (a) BrCH2CH2OTBS, NaH; (b) MCPBA; (c) TBAF; (d) MsCl, NEt3; (e) t-BuOK; (f) NH2CH2CH2OH, NaBH4; (g)
Boc2O; (h) TFA; (i) CH2O, NaBH4; (j) 3-amino propanol; (k) Ac2O.

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1691
Scheme 5 Reagents and conditions: (a) Methyl 5-(benzyloxy)pentanoate, LDA, 78 °C to rt; (b) Pd(OH)2, H2; (c) MsCl, NEt3; (d) t-BuOK; (e) LiBH4; (f)
mesyl chloride, NEt3; (g) NaN3; (h) PPh3, H2O; (i) Tf2O, NEt3; (j) 3-methyltetrahydro-2H-pyran-2-one, LDA,78 °C to rt, 31%; (k) (R)-4-methyltetrahy-
dro-2H-pyran-2-one, LDA,78 °C to rt, 88%; (l) Mg(OMe)2; (m) LAH.
first, the resulting anion was added to 2 in a Michael reac-
tion fashion to yield 25 which was a mixture of diastere-
omers at this point. The benzyl protecting group was re-
moved and resulting alcohol underwent standard ring clo-
sure reactions to yield compound 26. We were pleasantly
surprised that the product had pure cis configuration be-
tween sulfone and methyl ester group, which was the result
of methyl ester group isomerizing to its more thermody-
namically stable form upon treatment with a strong base.
The ester of 26 can be transformed to a variety of groups
using similar procedure as mentioned before. The direct
alkylation of 25 or 26 failed to generate 7,7-disubstituted
compounds probably due to the steric hindrance of this po-
sition. So the second substitution at 7-position had to be
introduced prior to Michael addition reaction in order to
generate 7,7-di substituted compounds. Starting with methyl
5-(benzyloxy)-2-methylpentanoate, compounds 29, 30 and
31 were synthesized using the similar procedure as that of
28. The difference was that the methyl ester could no longer
isomerize. The desired pure cis compound was separated
from the cis/trans mixture by column chromatography.
7,8-Disubstituted compounds were also made using a simi-
lar procedure. Pure enantiomeric starting material (R)-4-
methyltetrahydro-2H-pyran-2-one was added to 2 to give
compound 32. The lactone ring of 32 was opened with
magnesium methoxide, and the resulting alcohol went
through standard ring closure procedure to form compound
33. The methyl ester could again be isomerized, although
not as readily as 26, to the cis configuration between sul-
fone and methyl ester. Column chromatography could be
used to separate the diastereomers of compound 34.
Two typical routes to introduce functionalities on the pi-
peridine C ring compounds are shown in Scheme 6 and they
both started with the Michael addition of a nucleophile to
vinyl sulfone 2. (S)-1-Azido-4-(benzyloxy)butan-2-amine,
which can be easily synthesized from corresponding amino
acid, was added to 2 to give 36. The benzyl protecting
group was removed with boron trichloride and the product

1692
Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
Scheme 6 Reagents and conditions: (a) (S)-1-azido-4-(benzyloxy)butan-2-amine, DIPEA, 57%; (b) BCl3; (c) MsCl, NEt3; (d) t-BuOK, 52% for three steps;
(e) PPh3, H2O; (f) MsCl, NEt3; (g) methyl 2-((2-(benzyloxy)ethyl)(tert-butoxycarbonyl)amino)acetate, LDA, 78 oC to rt; (h) NaOMe; (i) BH3SMe3; (j)
Dess-Martin reagent; (k) NH3, Ti(O-i-Pr)4, NaBH4; (l) MsCl, NEt3; then HCl; (m) CH2O, NaBH(OAc)3.
Table 1 Membrane A40 IC50 data for various tricyclic compounds
underwent standard ring closure reactions to form tricyclic
37. The azide was reduced to amine and the usual deriva-
tives were made as before. On the other hand, methyl 2-((2-
(benzyloxy)ethyl)(tert-butoxycarbonyl)amino) acetate could
be added to 2 in Michael reaction fashion to form com-
pound 39. Benzyl removal, ring closure and functional
group manipulation yielded compounds 40, 41 and 42. It
should be noted that the aforementioned Michael addition,
and then ring closure reaction sequence was very robust and
could be used to introduce many other substituted C rings.
Mem. A40
Compounds a)
X
Y
Z
IC50 (nM) b)
3100
303
3 Result and discussion
3
10





CH2
CH2
O

CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
NH
We have determined previously that a tricyclic system could
give rise to very potent compounds in both in vitro and in
vivo assays of gamma secretase inhibition. The goal of cur-
rent study was to develop the SAR of C ring and further
improve the overall profile of resulting compounds. The
first task was to determine the optimal size and type of the
newly introduced C ring. Thus compounds with different C
rings were synthesized while keeping the previously opti-
mized A and B rings constant. The membrane A40 IC50
data of these compounds are listed in Table 1. The results
showed unequivocally that the six-member C ring is optimal.
The activity increased drastically when the size of all car-
bon C-ring increased from 3 (3100 nM) to 6 (20 nM).
7-Member C-ring compounds were also synthesized and
they were not as active as 6 member C ring compounds [7].
Furthermore, the data also showed both cyclohexane and
7
113
13
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
22
17
22
() 17 c)
() 17 c)
20a
20b
22
O
10000
9.0
O
NH
NCH3
CH2
CH2
780
132
81
23
NAc
47
a) All compounds are racemic unless otherwise indicated. b) Values are
means of two experiments. c) Pure enantiomer.
pyran C rings are well tolerated. A piperidine C ring was
also tolerated although to a lesser extent. Bis pyran 17 was
resolved to its pure enantiomers, and results showed that the
activity resided mostly in () enantiomer as shown for pre-

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1693
viously published compounds [5]. As the result, whenever it
is possible, the pure enantiomers were obtained via chiral
HPLC separation and were used for more detailed biologi-
cal studies.
the resulting compounds have lower single digit nanomolar
activities in both membrane and whole cell assays. On the
other hand, the in vivo efficacies of the compounds de-
pended on both the in vitro IC50s and in vivo PK profiles.
Among all the compounds tested, the triflamide 28b had the
best rat PK and brain penetration profile. As the result, it
had the highest percentage inhibition although it did not
have the lowest IC50. This compound was studied in more
detail in our animal studies and the results showed that it
inhibited plasma and brain A40 of CRND8 mouse by 98
and 69 percent respectively when dosed orally at 30 mg/kg.
Additionally, the peripheral properties of these compounds,
such as CYP inhibition and PXR induction, are generally
acceptable. The detailed in vivo studies of these compounds
were carried out and will be reported in due time.
Since both 7- and 8-positions are well tolerated, we de-
cided to make di-substituted compounds at these two posi-
tions. Large libraries of compounds with di-substitutions at
7- and/or 8-positions including amides, sulfonamides, urea,
carbonamides, sulfones and heterocycles were synthesized
and evaluated and a few of representative compounds are
listed in Table 2. As the data for 3035 show, membrane
A40 IC50s of 7, 8-disubstituted compounds were generally
lower than mono-substituted analogs. Although 7,7-di-
substituted compounds could yield very active compounds
in terms of IC50, further studies showed that they offered no
advantage in terms of overall profile of the compounds.
Compounds such as 13 and 17 were potent enough to
warrant further in vivo studies. Unfortunately, subsequent
studies revealed that they have a very poor pharmacokinetic
profile, most likely due to their high logP values. Because
of their poor PK, they also had limited efficacies in our in
vivo CRND8 mouse studies as shown by the data in Table 2.
Thus further modification of this core was needed to im-
prove their overall profile, especially to reduce their logP
value. Since the 6 member C ring was determined to be the
optimal ring size, we concentrated our effort on 6-mem-
bered C ring compounds. We have demonstrated previously
that substitutions at the C-8 position on the tricyclic system
were well tolerated and the compounds with polar substitu-
tions at 8-position had excellent in vitro and in vivo activi-
ties as represented by compound 24 [5]. A complete SAR
studies around the C-ring was then carried out and the in
vitro and in vivo data for representative compounds are
shown in Table 2. When the substitutions were introduced
to 7-position, the resulting compounds were among the
most potent in this series as indicated by compound 2428e.
When appropriate functional groups, such as in compounds
28c, 28d and 28e, were introduced at this position and the
compounds were resolved to their pure enantiomeric form,
Table 2 In vitro IC50 and rat in vivo data for various tricyclic compounds
% inh. of A40 CRND8 mouse e)
Rat AUC0-6h (h.nM)
Mem A40
Whole Cell A40 Whole Cell A42
Compounds a)
IC50 (nM) b)
IC50 (nM) b)
IC50 (nM) b)
brain/plasma
plasma, brain
13
() 17 c)
20a
20b
22
22
9
43
10
27
8
19, 32


20, 12
780
132
81
81
58
70
108
27
21
15
2
61
44
47
46
15
8
0, 
23
47
4, 
90, 42
24 c)
26
27
9254, 0.3
21
0, 
27 c)
28b c)
28c
28d
28e
30 d)
31
15
14
6
1868, 1.68
43, 
185, 
540
98, 69
73, 34
49, 3
2.5
1.6
3.1
19
2
4
2
4
4
6, 21
78
26
0
550
128
271
172
290
92
33 c)
34 c)
35 c
37 c)
38 c)
40
364
80
41
33
18
42
122
a) All compounds are racemic unless otherwise indicated. b) Values are means of two experiments. c) Pure enantiomer. d) Mixture of diastereomers. e)
Dosed at 30 mg/kg po.

1694
Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
Compounds with a piperidine C ring were also studied in
detail and a few representative compounds are listed in Ta-
ble 2. As the data for compounds 3742 shown, these com-
pounds are generally less active than corresponding cyclo-
hexane compounds. Although selected compounds in
di-substituted series should show improved activity via en-
antiomeric resolution, the majority of work was done on
mono-substituted compounds, such as 24 and 28b due to
their easy synthetic accessibility.
This report concentrated on p-chlorophenyl at the aryl
sulfone portion of the compounds since it was shown to be
one of the best groups in various studies. Other groups at
this position, especially the p-trifluorophenyl group were
also studied. Since the synthesis and SAR trend of these
compounds were almost the mirror image of results pre-
sented here, they are not included in this paper. Detailed
biological studies were carried out to differentiate the com-
pounds with p-chlorophenyl or p-trifluorophenyl as aryl
sulfone and the results will be reported in due course.
or M + Na peak by LCMS. In vitro and in vivo data given
throughout the text and in Tables 1 and 2 were collected
using previously described method [8].
7b-(4-Chlorophenylsulfonyl)-4,7-difluoro-1,1a,2,7b-tetra-
hydrocyclopropa [c]chromene (3)
1 (2 g, 5.8 mmol) was dissolved in 100 mL DCM and
MCPBA (77%, 3.9 g, 17.5 mmol) was added. The reaction
was stirred at room temperature for 3 h. 8 g NaHSO3 in
50 mL water were added to quench excess MCPBA. The
organic layer was separated, washed with 1 N NaOH solu-
tion, brine, dried over Na2SO4 and concentrated. The prod-
uct was purified by column using EtOAc/hexane as the el-
uent (gradient from 0/100 to 75/25 in 40 min). Yield: 1.6 g,
75%. Part of the product (0.3 g, 0.76 mmole) was dissolved
in 50 mL DCM and MsCl (0.43 g, 3.8 mmol) and NEt3
(0.38 g, 3.8 mmol) were added. The mixture was stirred at
room temperature for 2 h. KOt-Bu (1 M in THF, 5 mL) was
added and the reaction was stirred at room temperature for
1 h. 50 mL water was added. The organic layer was washed
with brine (50 mL), dried over Na2SO4 and concentrated.
The product was purified by column using EtOAc/hexane as
eluent (gradient from 0/100 to 25/75 in 40 min). Yield: 101
mg, 37%. 1H NMR (400 MHz, CDCl3)  7.53 (d, J = 8.8 Hz,
2 H), 7.39 (d, J = 8.8 Hz, 2 H), 6.99 (m, 1 H), 6.67 (m, 1 H),
4.84 (dd, J = 11.7 and 8.1 Hz, 1 H), 3.33 (dd, J = 11.7 and
8.1 Hz, 1 H), 2.57 (m, 1 H), 2.44(dd, J = 11.7 and 8.1 Hz, 1
H), 1.34 (t, J = 5.9 Hz, 1 H).
4 Conclusion
Detailed C ring SAR studies of tricyclic sulfone -secretase
inhibitors were carried out. Compounds in this new series
have comparable or better in vitro activities and in vivo ef-
ficacies than previously reported sulfonamide compounds.
The 6-membered ring was identified as the optimal size for
C ring. Pyran and cyclohexane were the best while piperi-
dine was also tolerated as C ring. Additionally, 7- and 8-
positions were identified as sweet spots on the tricyclic core
to introduce polar groups for the fine tuning the final com-
pounds. The best compounds in this series had excellent in
vitro activities and in vivo efficacies in addition to their
clean peripheral profiles. Further studies also revealed that
they have reduced mechanism based Notch side effects that
plagued previous sulfonamide -secretase inhibitors.
Trans-4-(4-chlorophenylthio)-5,8-difluorochroman-3-carba-
ldehyde (4)
1 (2.8 g, 8.8 mmol) was dissolved in 15 mL DCM and
Dess-Martin reagent (4.1 g, 9.7 mmol) was added. The re-
action was stirred at room temperature for 3 h. 40 mL
EtOAc and 30 mL saturated Na2S2O3 solution were added
and the organic layer was washed with saturated NaHCO3
solution, dried over Na2SO4 and concentrated. The residue
(2.8 g, quant.) was used in next step without further purifi-
cation. 1H NMR (400 MHz, CDCl3) 9.70 (s, 1 H), 7.45 (d,
J = 8.8 Hz, 2 H), 7.34 (d, J = 8.8 Hz, 2 H), 6.97 (m, 1 H),
6.62 (m, 1 H), 4.874.95 (m, 2 H), 4.74 (dd, J = 11.7 and
2.9 Hz, 1 H), 2.83 (m, 1 H).
5 Experimental
General
Silica gel chromatography was performed using pre packed
silica gel cartridges (Biotage or Isco) on ISCO Com-
biFlash™ or Analogix intelliflash 280 system. NMR spectra
were obtained on a Varian XL-400 spectrometer in CDCl3
unless otherwise indicated and are reported as ppm down-
field from Me4Si. Purity was checked via LCMS analysis,
performed on an Applied Biosystems API-100 mass spec-
trometer and Shimadzu SCL-10A LC column: Alltech plat-
inum C18, 3 micron, 33 mm × 7 mm ID; gradient flow: 0
min-10% CH3CN, 5 min-95% CH3CN, 7 min-95% CH3CN,
7.5 min-10% CH3CN, 9 min-stop. All tested compounds
were more than 95% pure by this HPLC method unless oth-
erwise noted and all compounds also showed correct M + 1
Trans-methyl (3-(4-(4-chlorophenylthio)-5,8-difluorochroman-
3-yl))acrylate (5)
4 (0.43 g, 1.2 mmol) was dissolved in 20 mL THF and
methyl (triphenylphosphoranylidene) acetate (0.77 g, 2.3
mmol) was added. The reaction was stirred at room temper-
ature for 3 h. 30 mL EtOAc was added. The organic layer
was washed with water, dried over Na2SO4 and concentrat-
ed. The product was purified by column chromatography
1
(EtOAc/hexane from 0/100 to 25/75 in 45 min). H NMR
(400 MHz, CDCl3) 7.44 (d, J = 8.8 Hz, 2 H), 7.34 (d, J =
8.8 Hz, 2 H), 7.00 (m, 1 H), 6.79 (m, 1H), 6.60 (m, 1 H),
5.86 (dd, J = 16.1 and 1.5 Hz, 1 H), 4.76 (dd, J = 11.0 and

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1695
2.2 Hz, 1 H), 4.374.43 (m, 2 H), 2.81 (m, 1 H).
butoxy-2-oxoethyl)lithium in 10 mL THF at 78oC was
added via cannula. The reaction was stirred at 78oC for 1.5
h. The reaction was quenched with 100 mL EtOAc and 100
mL of water. The organic layer was separated, washed with
brine (2 × 50 mL), dried over Na2SO4 and concentrated. The
residue was purified by column chromatography (Hex-
ane/EtOAc from 100/0 25/75 to 75/25 in 50 minutes). Yield:
Trans-3-(-4-(4-chlorophenylsulfonyl)-5,8-difluorochroman-
3-yl)propan-1-ol (6)
5 (0.21 g, 0.53 mmol) was dissolved in 10 mL DCM and
MCPBA (77%, 0.5 g) was added. The reaction was stirred
at room temperature for 3 h. 1 g Na2S2O3 in 50 mL water
and 100 mL EtOAc were added. The organic layer was sep-
arated, washed with 1 N NaOH solution (50 mL), brine (50
mL), dried over Na2SO4 and concentrated. The residue was
dissolved in 20 mL EtOAc and 2 drops of acetic acid was
added. PtO2/C (0.1 g) was added and hydrogen was intro-
duced via a balloon. The reaction was stirred at room tem-
perature overnight. The catalyst was filtered and residue
was dissolved in 10 mL THF and LiAlH4 (1 M in THF, 0.9
mL) was added. The mixture was stirred at room tempera-
ture for 0.5 h. 50 mL EtOAc was added and the organic
layer was washed with 1 N HCl solution (2 × 50 mL), dried
over Na2SO4 and concentrated. The product was purified by
column (EtOAc/hexane from 0/100 to 50/50 in 35 min).
1
330 mg, 51%. H NMR (400 MHz, CDCl3)  7.74 (d, J =
8.8 Hz, 2 H), 7.49 (d, J = 8.1 Hz, 2H), 7.00 (m, 1H), 6.42
(m, 1 H), 4.89 (dd, J = 11.7 and 2.9 Hz, 1 H), 4.48 (s, 1 H),
4.29 (d, J = 11.7 Hz, 1 H), 3.22 (m, 1 H), 2.34 (dd, J = 16.8
and 6.6 Hz, 1 H), 2.10 (dd, J = 11.7 and 8.7 Hz, 1 H).
8b-(4-Chlorophenylsulfonyl)-5,8-difluoro-2,2a,3,8b-tetrah-
ydro-1H-cyclobuta[c]chromene (10)
10 was synthesized from 9 using the similar procedure as
that of compound 7. 1H NMR (400 MHz, CDCl3)  7.54 (d,
J = 8.1 Hz, 2 H), 7.42 (d, J = 8.7 Hz, 2 H), 7.01 (m, 1H),
6.48 (m, 1 H), 4.21 (dd, J = 11.7 and 5.1 Hz, 1 H), 4.02 (dd,
J = 11.7 and 4.4 Hz, 1 H), 3.68 (m, 1H), 3.17 (m, 1 H),
2.242.41 (m, 2H), 1.89 (m, 1 H).
1
Yield: 60 mg, 29% for three steps. H NMR (400 MHz,
CDCl3) 7.71 (d, J = 8.8 Hz, 2 H), 7.51 (d, J = 8.8 Hz, 2 H),
7.02 (m, 1H), 6.39 (m, 1 H), 4.96 (dd, J = 11.7 and 2.9 Hz,
1 H), 4.414.46 (m, 2 H), 3.523.67 (m, 2 H), 2.84 (m, 1H),
1.301.70 (m, 5 H).
(5-(4-Chlorophenylsulfonyl)-5-(2,3,6-trifluorophenyl)pentyl)-
oxirane (12)
11 (1.0 g, 3.1 mmol) and 5-bromopentene (1.5 g, 6.3
mmol) were dissolved in 20 mL THF and t-BuOK (1M in
THF, 6.2 mL) was added. The reaction was stirred at room
temperature overnight. 50 mL water and 50 mL EtOAc
were added. The organic layer was washed with brine (50
mL), dried over Na2SO4 and concentrated. The product,
2-(1-(4-chlorophenylsulfonyl)hept-6-enyl)-1,3,4-trifluorobe-
nzene, was purified by column using EtOAc/Hexane as el-
uent (gradient from 0/100 to 25/75 in 40 min). The purified
product (0.65 g, 1.6 mmol) and MCPBA (77%, 0.71g, 3.2
mmol) were stirred in 50 DCM for 5 h. 2 g Na2SO3 in 100
mL water were added to quench excess MCPBA. The or-
ganic layer was separated, washed with 1 N NaOH solution
(50 mL), brine (50 mL), dried over Na2SO4 and concentrat-
ed. The product was purified by column using EtOAc/
hexane as eluent (gradient from 0/100 to 40/60 in 40 min).
(0.52 g, 77%) 1H NMR (400 MHz, CDCl3) 7.63 (dd, J =
8.8 and 1.5 Hz, 2 H), 7.44 (d, J = 8.1 Hz, 2 H), 7.14 (m, 1
H), 6.78 (m, 1 H), 4.60 (m, 1 H), 2.83 (m, 1H), 2.71 (m,
1H), 2.392.50 (m, 3 H), 1.351.70 (m, 6H).
9b-(4-Chlorophenylsulfonyl)-6,9-difluoro-1,2,3,3a,4,9b-
hexahydrocyclopenta[c]chromene (7)
6 (48 mg, 0.12 mmol) and MsCl (16.3 mg, 0.14 mmol)
were dissolved in 5 mL DCM. NEt3 (24 mg, 0.24 mmol)
was added. The mixture was stirred at room temperature for
10 min. 50 mL water and 50 mL EtOAc were added. The
organic layer was washed with water (50 mL), brine (50
mL), dried over Na2SO4 and concentrated. The residue was
dissolved in 10 mL THF and KOt-Bu (1 M in THF, 0.3 mL)
was added. The mixture was stirred at room temperature for
30 min. 50 mL water and 50 mL EtOAc were added. The
organic layer was washed with brine (50 mL), dried over
Na2SO4 and concentrated. The product was purified by
column (EtOAc/hexane from 0/100 to 25/75 in 45 min).
1
Yield: 30 mg, 65% for two steps. H NMR (400 MHz,
CDCl3)  7.51 (d, J = 8.1 Hz, 2 H), 7.41 (d, J = 8.8 Hz, 2 H),
6.99 (m, 1 H), 6.45 (m, 1 H), 4.58 (dd, J = 11.7 and 5.1 Hz,
1 H), 3.87 (dd, J = 11.7 and 5.8 Hz, 1 H), 3.27 (m, 1 H),
2.86 (m, 1 H), 2.41 (m, 1 H), 2.12 (m, 1 H), 1.95 (m, 1 H),
1.61 (m, 2 H).
(6aR,10aS)-10a-(4-chlorophenylsulfonyl)-1,4-difluoro-6a,7,
8,9,10,10a-hexahydro-6H-benzo[c]chromene (13) and 7(S)-
[(4-chlorophenyl)sulfonyl]-8,11-difluoro-2,3,4,5,6,7-hexahy
dro-2(S),7-methano-1-benzoonin (14)
12 (0.52 g, 1.24 mmol) was dissolved in 25 mL THF and
t-BuOK (1 M in THF, 3.72 mL) was added. The reaction
was heated to 50 °C for 4 h. 50 mL water and 50 mL EtOAc
were added. The organic layer was washed with brine
(50 mL), dried over Na2SO4 and concentrated. The product
t-Butyl-2-((3S,4R)-4-(4-chlorophenylsulfonyl)-5,8-difluoro-
chroman-3-yl)acetate (8)
2 (0.49 g, 1.4 mmol) was dissolved in 10 mL dry THF
and cooled to 78°C. (2-tert-Butoxy-2-oxoethyl)lithium
(208 mg, 1.70 mmol) in 10 mL of THF was cooled to
78°C and added to the solution of 23 via cannula. The re-
action was stirred at 78°C for 1 h. TLC showed starting
material was still present so an additional 203 mg (2-tert-

1696
Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
was purified by column using EtOAc/hexane as eluent (gra-
dient from 0/100 to 25/75 in 40 minutes). Two compounds
were isolated.
was added. The mixture was stirred at room temperature for
2 h. 50 mL water and 50 mL EtOAc were added. The or-
ganic layer was washed with 1 N HCl solution (2 × 50 mL),
brine (50 mL), dried over Na2SO4 and concentrated. The
product (26 mg, 0.052 mmol) was dissolved in 5 mL THF
and KOt-Bu (1 M in THF, 0.5 mL) was added. The mixture
was stirred at room temperature overnight. 50 mL water and
50 mL EtOAc were added. The organic layer was washed
with brine (50 mL), dried over Na2SO4 and concentrated.
The product was purified by column chromatography
(EtOAc/hexane from 0/100 to 50/50 in 45 min). Yield: 17
mg, 81%. 1H NMR (CDCl3 400 MHz) 7.62 (d, J = 8.8 Hz,
2 H), 7.51 (d, J = 8.8 Hz, 2 H), 7.10 (m, 1 H), 6.44 (m, 1 H),
5.16 (dd, J = 8.8 and 2.9 Hz, 1 H), 4.14 (d, J = 12.4 Hz, 1
H), 3.82-3.97 (m, 2 H), 3.32 (t, J = 11.7 Hz, 1 H), 3.12 (t, J
= 11.7 Hz, 1 H), 2.89 (m, 1H), 2.56 (d, J = 13.1 Hz, 1 H),
2.33 (m, 1 H).
Compound 13 (27 mg, 5.4%) 1H NMR (400 MHz,
CDCl3)  7.59 (d, J = 8.1 Hz, 2 H), 7.48 (d, J = 8.8 Hz, 2 H),
7.06 (m, 1 H), 6.39 (m, 1 H), 5.22 (dd, J = 11.7 and 2.9 Hz,
1 H), 4.14 (d, J = 11.7 Hz, 1 H), 2.6 (m, 2H), 0.91.9 (m, 7
H).
Compound 14 (112 mg, 23%) 1H NMR (400 MHz,
CDCl3) 7.53 (d, J = 8.1 Hz, 2 H), 7.39 (d, J = 8.1 Hz, 2 H),
7.01 (m, 1H), 6.60 (m, 1 H), 4.85 (m, 1 H), 3.12 (m, 1 H),
2.80 (m, 1 H), 1.12.4 (m, 8 H).
tert-Butyl-{2-[trans-4-(4-chloro-phenylsulfanyl)-5,8-difluoro-
chroman-3-ylmethoxy]-ethoxy}-dimethyl-silane (15)
1 (0.4 g, 1.2 mmol), (2-bromoethoxy)-tert-butyl-dime-
thylsilane (0.56 g, 2.3 mmol) and NaH (60% in oil, 0.5 g)
were stirred in 25 mL THF at 50 °C for 5 h. 100 mL water
and 100 mL EtOAc were added. The organic layer was
washed with brine (100 mL), dried over Na2SO4 and con-
centrated. The product was purified by column chromatog-
raphy (EtOAc/hexane from 0/100 to 10/90 in 45 min). Yield:
The racemic mixture can be separated into two pure en-
antiomers using chiral OJ (size: 15 x 250 mm) column with
pure ethanol as the solvent. The first fraction is compound
17(), [] = 138.4 degree (c =1.000 in DCM). The second
fraction is compound 17(+), [] = + 137.2 degree (c =1.000
in DCM).
1
85 mg, 15%. H NMR (CDCl3 400 MHz) δ 7.43 (d, J = 8.8
Hz, 2 H), 7.30 (d, J = 8.8 Hz, 2 H), 6.96 (m, 1 H), 6.57 (m,
1 H), 4.60 (dd, J = 8.8 and 2.2 Hz, 1 H), 4.51 (br, 1 H), 4.38
(td, J = 11.0 and 1.5 Hz, 1 H), 3.68 (t, J = 4.4 Hz, 2 H),
3.283.50 (m, 4H), 2.35 (m, 1 H), 0.86 (s, 9H), 0.01 (s, 6
H).
tert-Butyl (4-(4-chlorophenylsulfonyl)-5,8-difluorochroman-
3-yl)methyl(2-hydroxyethyl)carbamate (18)
4 (2.2 g, 5.9 mmol) and ethanolamine (1.1 g, 18 mmol)
were dissolved in 20 mL THF. The reaction was stirred at
room temperature overnight. NaBH4 (2 g, 52 mmol) and
MeOH (10 mL) were added and the reaction was stirred for
3 h. 100 mL water and 100 mL EtOAc were added. The
organic layer was washed with water (100 mL), dried over
2-[trans-4-(4-Chloro-benzenesulfonyl)-5,8-difluoro-chroman-
3-ylmethoxy]-ethanol (16)
15 (85 mg, 0.17 mmol) was dissolved in 25 mL DCM
and MCPBA (77%, 0.11 g, 0.51 mmol) was added. The
reaction was stirred at room temperature for 3 h. 2 g
Na2S2O3 in 50 mL water and 100 mL EtOAc were added.
The organic layer was washed with 1 N NaOH solution (50
mL), brine (50 mL), dried over Na2SO4 and concentrated.
The product (75 mg, 0.14 mmol) was dissolved in 20 mL
THF and TBAF (0.2 g) was added. The reaction was stirred
at room temperature for 3 h. 50 mL water and 50 mL
EtOAc were added. The organic layer was washed with
brine (50 mL), dried over Na2SO4 and concentrated. The
product was purified by column chromatography (EtOAc/
hexane from 0/100 to 50/50 in 35 min). Yield: 50 mg, 70%
for two steps. 1H NMR (CDCl3 400 MHz)  7.73 (d, J = 8.1
Hz, 2 H), 7.51 (d, J = 8.1 Hz, 2 H), 7.03 (m, 1 H), 6.41 (m,
1 H), 4.90 (dd, J = 8.8 and 2.9 Hz, 1 H), 4.57 (s, 1 H), 4.39
(td, J = 11.7 and 1.5 Hz, 1 H), 3.68(br, 2 H), 3.413.54 (m,
3 H), 2.30 (t, J = 9.5 Hz, 1 H), 3.15 (m, 1 H).
Na2
SO and concentrated. The product was purified by
4
column chromatography (EtOAc/hexane from 25/75 to
100/0 in 45 min). The desire product (0.40 g, 18%) was
dissolved in 20 mL DCM and Boc2O (0.24 g, 1.2 mmol)
was added. The reaction was stirred at room temperature for
3 h. MCPBA (77%, 0.8 g, 3.6 mmol) was then added and
the reaction was stirred at room temperature for 2 h. 2 g
Na2S2O3 in 50 mL water was added to quench the reaction
and 100 mL EtOAc was added to extract the product. The
organic layer was washed with 1 N NaOH solution (50 mL),
brine (50 mL), dried over Na2SO4 and concentrated. The
product was purified by column chromatography (EtOAc/
hexane from 0/100 to 50/50 in 45 min). Yield: 0.44 g, 85%.
1H NMR (CDCl3 400 MHz)  7.71 (d, J = 8.1 Hz, 2 H),
7.49 (d, J = 8.8 Hz, 2 H), 7.02 (m, 1 H), 6.40 (m, 1 H), 4.90
(d, J = 11.7 Hz, 1 H), 4.40 (m, 1 H), 4.28 (d, J = 11.7 Hz, 1
H), 3.7 (m, 2 H), 3.103.40 (m, 5 H), 1.26 (s, 9 H).
Trans-10b-(4-Chloro-benzenesulfonyl)-7,10-difluoro-1,4a,
5,10b-tetrahydro-2H,4H-pyrano[3,4-c]chromene (17)
16 (25 mg, 0.06 mmol) and MsCl (68 mg, 0.59 mmol)
were dissolved in 5 mL DCM and NEt3 (59 mg, 59 mmol)
tert-Butyl-10b-(4-chlorophenylsulfonyl)-7,10-difluoro-4,4a,5,
10b-tetrahydro-1H-chromeno[3,4-c]pyridine-3(2H)-carbox
ylate (19)
Compound 19 is synthesized from 18 using the same

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1697
method as that of 17. 1H NMR (CDCl3 400 MHz)  7.60 (d,
J = 8.1 Hz, 2 H), 7.50 (d, J = 8.8 Hz, 2 H), 7.09 (m, 1 H),
6.43 (m, 1 H), 5.20 (dd, J = 11.7 and 2.2 Hz, 1 H), 4.23 (d, J
= 12.4 Hz, 1 H), 4.1 (m, 2 H), 2.302.82 (m, 4 H), 2.11 (m,
1 H), 1.44 (s, 9 H).
hexahydro-1H-chromeno[3,4-b]pyridine (22)
21 (290 mg, 0.69 mmol) was dissolved in 50 mL of di-
chloromethane. MsCl (64 L, 0.83 mmol) and NEt3 (117
L, 0.83 mmol) were added respectively and stirred at room
temperature overnight. The solution was quenched with
water (40 mL) and dichloromethane (40 mL). The layers
were separated and the aqueous layer was washed with
DCM. The combined organics were dried over Na2SO4 and
concentrated under vacuum. The product was used without
(4aR,10bS)-10b-(4-Chlorophenylsulfonyl)-7,10-difluoro-2,3,
4,4a,5,10b-hexahydro-1H-chromeno[3,4-c]pyridine (20a)
19 (0.21 g, 0.42 mmol) was dissolved in 20 mL DCM
and TFA (5 mL) was added. The mixture was stirred at
room temperature for 1 h. 100 mL saturated Na2CO3 solu-
tion and 100 mL EtOAc were added. The organic layer was
washed with saturated Na2CO3 solution (50 mL), dried over
Na2SO4 and concentrated. The residue was pure product.
Yield, 0.16 g, 100%. 1H NMR (CDCl3 400 MHz)  7.61 (d,
J = 8.8 Hz, 2 H), 7.49 (d, J = 8.8 Hz, 2 H), 7.08 (m, 1 H),
6.42 (m, 1 H), 5.18 (dd, J = 11.7 and 2.2 Hz, 1 H), 4.15 (d, J
= 11.7 Hz, 1 H), 2.984.11 (m, 2 H), 2.582.73 (m, 3 H),
2.35 (td, J = 12.4 and 1.5 Hz, 1 H), 2.08 (m, 1 H).
1
further purification. Yield: 0.28 g, 81%. H NMR (CDCl3
400 MHz) δ 7.71 (d, J = 8.8 Hz, 2 H), 7.50 (d, J = 8.8 Hz, 2
H), 7.01 (td, J = 10.3, 5.1 Hz, 1 H), 6.41 (td, 8.8, 3.7 Hz, 1
H), 4.80 (dd, J = 11.8, 2.2 Hz, 1 H), 4.53 (s, 1H), 4.48 (d, J
= 11.8 Hz, 1 H), 4.22 (t, 5.9 Hz, 2 H), 3.75 (s, 1 H), 2.93 (s,
3 H), 2.822.67 (m, 2 H), 1.79 (m, 2 H).
The product from above reaction (0.28 g, 0.57 mmol)
was dissolved in 25 mL of THF then 1 M t-BuOK solution
(2.4 mL, 2.4 mmol) was added. The reaction was stirred at
room temperature for 3.5 h. The reaction was quenched
with water (50 mL) and washed with 50 mL EtOAc. The
organic layer was dried over Na2SO4 and concentrated. The
compound was then purified by prep TLC (EtOAc/hexane =
(4aR,10bS)-10b-(4-Chlorophenylsulfonyl)-7,10-difluoro-3-
methyl-2,3,4,4a,5,10b-hexahydro-1H-chromeno[3,4-c]pyri-
dine (20b)
1
50/50). Yield: 0.14g, 60%. H NMR (CDCl3 400 MHz)
20a (10 mg), CH3I and NaH (60% in oil, 10 mg) were
refluxed in 5 mL DCM for 5 h. 50 mL water and 50 mL
EtOAc were added. The organic layer was washed with
brine (100 mL), dried over Na2SO4 and concentrated. The
product was purified by column (EtOAc/hexane from 0/100
7.60 (d, J = 8.1 Hz, 2H), 7.49 (d, J = 8.8 Hz, 2H), 7.08 (td,
J = 9.5, 4.4 Hz, 1H), 6.476.40 (m, 1H), 5.21(dd, J = 11.8,
2.2 Hz, 1 H), 4.33 (d, J = 11.8 Hz, 1 H), 3.70 (s, 1 H), 3.00
(d, J = 13.2 Hz, 1 H), 2.76 (td, J = 12.4 Hz, 2.9 Hz, 1 H),
2.68 (d, J = 13.2 Hz, 1 H), 2.17(tt, J = 13.2, 2.9 Hz 1 H),
1.66-1.44 (m, 2 H), 1.211.07(m, 1 H).
1
to 100/0 in 45 min). Yield; 1.8 mg, 17%. H NMR (CDCl3
400 MHz)  7.64 (d, J = 8.8 Hz, 2 H), 7.50 (d, J = 8.8 Hz, 2
H), 7.09 (m, 1H), 6.43 (m, 1 H), 5.17 (dd, J = 11.7 and 2.9
Hz, 1 H), 4.16 (d, J = 11.7 Hz, 1 H), 2.74-2.98 (m, 2H),
2.64 (m, 1 H), 2.26 (m, 2 H), 2.18 (s, 3 H), 1.97 (t, J = 13.2
Hz, 1 H), 1.67 (m, 1 H).
1-[4a-(4-Chloro-benzenesulfonyl)-5,8-difluoro-2,3,4,4a,10,10a-
hexahydro-9-oxa-1-aza-phenanthren-1-yl]-ethanone (23)
The compound was made using standard amide coupling
1
reaction. H NMR (CDCl3 400 MHz) δ 7.61 (d, J = 8.8
Hz, 2 H), 7.43 (d, J = 8.8 Hz, 2 H), 7.05 (td, J = 9.5, 4.4 Hz,
1 H), 6.606.51 (m, 1 H), 5.40 (bs, 1 H), 4.50 (dd, J = 11.8,
3.7 Hz, 1 H), 4.20 (dd, J = 11.8, 5.9 Hz, 1 H), 3.62 (bs, 1 H),
2.90 (bs, 1 H), 2.702.60 (m, 1 H), 2.502.40 (m, 1 H),
2.222.03 (m, 4 H), 1.521.41 (m, 1 H).
3-(4-(4-Chlorophenylsulfonyl)-5, 8-difluorochroman-3-ylamino)-
propan-1-ol (21)
2 (0.84 g, 1.94 mmol) was dissolved in 20 mL THF and
3-aminopropanol (1.7 g, 22.6 mmol) was added. The reac-
tion was stirred at room temperature overnight. The reaction
was quenched with 100 mL saturated Na2CO3 solution and
100 mL EtOAc was added. The organic layer was washed
with water (100 mL), brine (100 mL), dried over Na2SO4
and concentrated. The product was purified by column us-
ing EtOAc/hexane as the eluent (gradient from 0/100 to
Methyl-2-(4-(4-chlorophenylsulfonyl)-5,8-difluorochroman-
3-yl)-5-hydroxypentanoate (25)
Diisopropylamine (6.16 g, 61 mmol) was dissolved in
100 mL of dry THF and the reaction was cooled to 0 °C.
BuLi (2.5M in THF, 24.0 mL) was added slowly and the
reaction was stirred at 0 °C for 10 min then cooled to
78 °C. Methyl 5-(benzyloxy)pentanoate (11.5 g, 51.8
mmol) in 50 mL dry THF was added via cannula. The reac-
tion was stirred for 55 minutes at 78 °C. 2 (6.94 g, 20.2
mmol) in 100 mL dry THF (pre-cooled to 78 °C) was
added via cannula over 10 min. The reaction was stirred at
78 °C for 1 h. 100 mL of water was added at 78 °C to
quench the reaction. The reaction was allowed to warm up
to room temperature overnight. 600 mL EtOAc was added.
1
25/75 in 50 min). Yield: 0.61 g, 75%. H NMR (CDCl3
400 MHz)  7.69 (d, J = 8.8 Hz, 2 H), 7.48 (d, J = 8.1 Hz, 2
H), 6.99 (td, J = 9.5, 5.1 Hz, 1 H), 6.38 (td, J = 9.5, 3.7 HZ,
1 H), 4.80 (dd, J = 12.5, 2.9 Hz, 1 H), 4.53 (s, 1 H), 4.43
(d, J = 11.8 Hz, 1 H), 3.78 (s, 1 H), 3.61 (t, J = 5.9 Hz,
2 H), 2.862.71 (m, 2H), 2.25 (bs, 1 H), 1.57 (m, J = 5.9
Hz, 2 H).
10b-(4-Chlorophenylsulfonyl)-7,10-difluoro-2,3,4,4a,5,10b-

1698
Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
The organic layer was washed with brine (2 x 400 mL),
dried over Na2SO4 and concentrated. The residue (13.5 g)
was dissolved in 200 mL EtOAc. Pd(OH)2 (20% on carbon,
2.5 g) was added then the system was purged with hydrogen.
The reaction was stirred at room temperature for 3.5 h. The
reaction solution was filtered through a celite cake and
concentrated. The product was purified by column using
hexane/EtOAc as the eluent (gradient from 100/0 to 40/60
in 60 min). Yield: 7.80 g, 69%, as a mixture of
diastereomers. 1H NMR (CDCl3 400 MHz)  7.717.80 (m,
2 H), 7.507.56 (m, 2 H), 7.007.09 (m, 1 H), 6.50 (m, 0.6
H), 6.43 (m, 0.4 H), 4.90 (m, 1 H), 4.52 (m, 1 H), 4.27 (m, 1
H), 3.543.66 (m, 6 H), 3.11 (m, 0.4 H), 2.95 (m, 0.6 H),
1.321.96 (m, 2 H).
1 H), 4.57 (d, J = 11.7 Hz, 1 H), 2.96 (m, 1 H), 2.80 (m, 1
H), 2.56 (m, 2 H), 1.88 (td, J = 13.2 and 2.8 Hz, 1 H), 1.75
(m, 2 H), 1.001.50 (m, 3 H).
N-((10a-(4-Chlorophenylsulfonyl)-1,4-difluoro-6a,7,8,9,10,
10a-hexahydro-6H-benzo[c]chromen-7-yl)methyl)-1,1,1-trif-
luoromethanesulfonamide (28b)
1H NMR (CDCl3 400 MHz) δ 7.62 (d, J = 8.1 Hz, 2 H),
7.51 (d, J = 8.1 Hz, 2 H), 7.09 (m, 1 H), 6.46 (m, 1 H), 5.20
(d, J = 13.2 Hz, 1 H), 5.08 (bs, 1 H), 4.52 (d, J = 13.2 Hz, 1
H), 3.413.58 (m, 2 H), 2.57 (m, 2 H), 1.601.92 (m, 4 H),
1.31(m, 1 H), 1.00 (m, 1 H).
N-((10a-(4-Chlorophenylsulfonyl)-1,4-difluoro-6a,7,8,9,10,
10a-hexahydro-6H-benzo[c]chromen-7-yl)methyl)acetamide
(28c)
(6aR,7R,10aS)-methyl-10a-(4-chlorophenylsulfonyl)-1,4-
difluoro-6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromene-
7-carboxylate (26)
1H NMR (CDCl3 400 MHz) δ 7.59 (d, J = 8.8 Hz, 2 H),
7.47 (d, J = 8.8 Hz, 2 H), 7.05 (m, 1 H), 6.43 (m, 1 H), 5.85
(t, J = 5.9 Hz, 1 H ), 5.12 (dd, J = 12.4 and 2.2 Hz, 1 H),
4.68 (d, J = 12.4 Hz, 1 H), 3.61 (m, 1 H), 3.26 (m, 1 H),
2.58 (d, J = 12.4Hz, 1 H), 2.38 (d, J = 11Hz, 1 H), 1.97 (s, 3
H), 1.82 (m, 1 H), 1.631.74 (m, 2 H), 1.54(m, 1 H), 1.24
(m, 1 H), 1.05 (m, 1 H).
26 was synthesized using the same procedure as that for
compound 17. However, during ring closure step upon
treatment with strong base, the methyl carboxlate group
completely isomerized to more stable trans isomer. 1H
NMR (CDCl3 400 MHz)  7.61 (d, J = 8.1 Hz, 2 H), 7.49
(d, J = 8.8 Hz, 2 H ), 7.08 (m, 1 H), 6.63 (m, 1 H), 5.19 (dd,
J = 12.4 and 2.9 Hz, 1 H), 4.20 (d, J = 11.7 Hz, 1 H), 3.73
(s, 3 H), 2.98 (M, 1 H), 2.64 (m, 1 H), 2.45 (m, 1 H), 1.93
(m, 2 H), 1.77 (m, 1 H), 1.54 (m, 1 H), 1.11 (m, 1 H).
N-(((6aR,7R,10aS)-10a-(4-Chlorophenylsulfonyl)-1,4-difluoro-
6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromen-7-yl)-
methyl)ethanesulfonamide (28d)
1H NMR (CDCl3 400 MHz) δ 7.61 (d, J = 8.8 Hz, 2 H),
7.49 (d, J = 8.8 Hz, 2 H), 7.05 (m, 1 H), 6.43 (m, 1 H), 5.13
(dd, J = 12.4 and 2.2 Hz, 1 H), 4.96 (t, J = 6.8 Hz, 1 H),
4.59 (d, J = 12.4 Hz, 1 H), 3.39 (m, 1 H), 3.21 (m, 1 H),
3.06 (q, J = 7.2 Hz, 2 H), 2.58 (m, 2 H), 1.86 (m, 1 H),
1.341.74 (m, 7 H), 1.05 (m, 1 H).
((6aR,7R,10aS)-10a-(4-Chlorophenylsulfonyl)-1,4-difluoro-
6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromen-7-yl)meth-
anol (27)
26 (0.88 g, 1.93 mmol) was dissolved in 25 mL dry THF
and excess NaBH4 (1.0 g) was added. The reaction was
stirred at room temperature overnight. 1 N HCl solution
(100 mL) and 50 mL EtOAc were added. The organic layer
was washed with brine (50 mL), dried over Na2SO4 and
concentrated. The residue was purified by column using
EtOAc/hexane as eluent (gradient from 0/100 to 50/50 in 35
min) to yield desired product (0.51 g, 62%). 1H NMR
(CDCl3 400 MHz)  7.62 (d, J = 8.4 Hz, 2 H), 7.49 (d, J =
8.8 Hz, 2 H), 7.05 (m, 1 H), 6.43 (m, 1 H), 5.17 (dd, J =
12.4 and 2.4 Hz, 1 H), 4.63 (d, J = 12.4 Hz, 1 H), 3.88 (m, 1
H), 3.70 (m, 1 H), 2.70 (m, 1 H), 2.52 (m, 1 H), 1.541.95
(m, 5 H), 1.11 (m, 1 H).
N-(((6aR,7R,10aS)-10a-(4-Chlorophenylsulfonyl)-1,4-
difluoro-6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromen-7
-yl)-methyl)cyclopropanesulfonamide (28e)
1H NMR (CDCl3 400 MHz) δ 7.62 (d, J = 8.8 Hz, 2 H),
7.49 (d, J = 8.8 Hz, 2 H), 7.06 (m, 1 H), 6.44 (m, 1 H), 5.15
(dd, J = 12.4 and 2.2Hz, 1 H), 4.84 (t, J = 6.8 Hz, 1 H), 4.60
(d, J = 12.4 Hz, 1 H), 3.40 (m, 1 H), 3.28 (m, 1 H), 2.58 (m,
2 H), 2.42 (m, 1 H), 1.541.91 (m, 4 H), 0.98-1.42 (m, 6 H).
Methyl-5-(benzyloxy)-2-(4-(4-chlorophenylsulfonyl)-5,8-
difluorochroman-3-yl)-2-methylpentanoate (29)
Diisopropylamine (2.0 g, 20 mmol) was dissolved in 100
mL dry THF and the reaction was cooled to 0 °C. BuLi (2.5
M in toluene, 8.0 mL, 20 mmol) was added and the reaction
was stirred at 0 °C for 10 min. The reaction was cooled to
78 °C and 3-methyltetrahydro-2H-pyran-2-one (2.3 g, 20
mmol) in 30 mL dry THF was added. The reaction was
stirred for 1 h at 78 °C. 2 (3.4 g, 10 mmol) in 70 mL THF
(pre-cooled to 78 °C) was added. The reaction was taken
out of dry ice bath and stirred for 1 h. 200 mL brine and 200
((6aR,7R,10aS)-10a-(4-Chlorophenylsulfonyl)-1,4-difluoro-
6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromen-7-yl)meth-
anamine (28a)
28a was synthesized using same procedure as that for 24.
The compound 27 was separated using chiral OD® column
and the active enantiomer was carried forward to give enan-
1
tiomerically pure final compounds. H NMR (CDCl3 400
MHz)  7.61 (d, J = 8.8 Hz, 2 H), 7.48 (d, J = 8.8 Hz, 2 H),
7.05 (m, 1 H), 6.41 (m, 1 H), 5.15 (dd, J = 12.4 and 2.9 Hz,

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1699
mL EtOAc was added. The organic layer was washed with
brine (100 mL), dried over Na2SO4 and concentrated. The
product was purified by column using EtOAc/hexane as
eluent (gradient from 0/100 to 50/50 in 45 minutes). Yield:
isomers of compound 33 and 34 are diastereomers thus
could be separated with normal column chromatography. As
the result, they were separated and only the active dia-
stereomers are listed here.
1
1.4 g, 31%. H NMR (CDCl3 400 MHz)  7.767.85 (m, 2
(4R)-3-(4-(4-Chlorophenylsulfonyl)-5,8-difluorochroman-3-
H), 7.527.56 (m, 2 H), 7.027.10 (m, 1 H), 6.486.59 (m,
1 H), 4.804.88 (m, 1 H), 4.624.76 (m, 1 H), 4.0050 (m,
2 H), 3.28 (m, 1 H), 1.302.00 (m, 8 H).
yl)-4-methyltetrahydro-2H-pyran-2-one (32)
1
This was a mixture of diatereomers. H NMR (CDCl3
400 MHz)  7.74 (m, 0.7 H), 7.447.56 (m, 3.3 H),
6.987.07 (m, 1 H), 6.456.53 (m, 1 H), 4.8593 (m, 1 H),
4.244.56 (m, 3.3 H), 4.01 (m, 0.3 H), 3.38 (m, 0.7 H), 3.15
(m, 0.3 H), 2.70 (m, 0.7 H), 2.28 (m, 0.3 H), 1.902.15 (m,
2 H), 1.601.70 (m, 1 H), 1.151.28 (m, 4 H).
The product from above step (1.4 g, 3.1 mmole) was
dissolved in 100 mL MeOH and Mg(OMe)2 (6%10% in
MeOH, 6.7 mL) was added. The reaction was stirred at
room temperature for 2 h. The reaction was then concen-
trated to about 20 mL. 200 mL EtOAc and 200 mL 1N HCl
solution were added. The organic layer was washed with
brine (2 × 200 mL), dried over Na2SO4 and concentrated.
The product was purified by column (EtOAc/hexane from
(6aR,7R,8R,10aS)-Methyl-10a-(4-chlorophenylsulfonyl)-1,4-
difluoro-8-methyl-6a,7,8,9,10,10a-hexahydro-6H-benzo[c]-
chromene-7-carboxylate (33)
1
0/100 to 60/40 in 45 min). Yield: 1 g, 66%. H NMR
This was the desired pure diastereomer that was separat-
ed from the reaction mixture. H NMR (CDCl3 400 MHz)
(CDCl3 400 MHz)  7.687.76 (m, 2 H), 7.487.54 (m, 2
H), 6.987.05 (m, 1 H), 6.3849 (m, 1 H), 4.79.86 (m,
1 H), 4.374.66 (m, 2 H), 3.543.67 (m, 1 H), 3.53 (s, 1.5
H), 3.41 (s, 1.5 H), 3.043.16 (m, 1 H), 1.301.88 (m, 4 H),
1.03 (m, 3 H).
Starting from 29, compounds 30 and 31 were synthesized
using the same procedure as that of compounds 26 and 27.
The only difference is that the cis and trans isomers
between hydroxymethyl and sulfone of 31 were separated
by column chromatagraph.
1
 7.63 (d, J = 8.8 Hz, 2 H), 7.49 (d, J = 8.8Hz, 2 H), 7.07 (m,
1 H), 6.43 (m, 1 H), 5.19 (dd, J = 12.4 and 2.8 Hz, 1 H),
4.33 (d, J = 12.4 Hz, 1 H), 3.71 (s, 3 H), 3.14 (d, J = 12.4
Hz, 1 H), 2.63 (dd, J = 12 and 4.8 Hz, 1 H), 2.40 (m, 1 H),
1.34.60 (m, 2 H), 0.96 (d, J = 7.2 Hz, 3 H).
((6aR,7R,8R,10aS)-10a-(4-Chlorophenylsulfonyl)-1,4-difluo-
ro-8-methyl-6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chrom-
en-7-yl)methanol (34)
This was the desired pure diastereomer. 1H NMR (CDCl3
400 MHz)  7.63 (d, J = 8.4 Hz, 2 H), 7.49 (d, J = 8.4 Hz, 2
H), 7.06 (m, 1 H), 6.45 (m, 1 H), 5.17 (dd, J = 12.4 and 2.4
Hz, 1 H), 4.55 (d, J = 12.0 Hz, 1 H), 3.88 (dd, J = 10.8 and
4.8 Hz, 1 H), 3.74 (dd, J = 10.8 and 4.8 Hz, 1 H), 2.79 (d, J
= 12.0 Hz, 1 H), 2.30 (m, 1 H), 2.19 (m, 1 H), 2.04 (m, 1 H),
1.74 (m, 1 H), 1.50 (m, 1 H), 1.34 (m, 1 H), 1.02 (d, J = 6.8
Hz, 3 H).
Methyl-10a-(4-chlorophenylsulfonyl)-1,4-difluoro-7-methyl-
6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromene-7-carbox-
ylate (30)
The product was a mixture of cis and trans isomer
1
between methyl ester and sulfone. H NMR (CDCl3 400
MHz)  7.447.54 (m, 4 H), 6.9910 (m, 1 H), 6.336.41
(m, 1 H), 5.135.25 (m, 1 H), 4.83 (d, J = 12.4 Hz, 0.5 H),
4.32 (d, J = 12.4 Hz, 0.5 H), 3.74 (s, 1.5 H), 3.37 (s, 1.5 H),
2.64 (m, 1 H), 1.252.10 (m, 6 H), 1.42 (s, 1.5 H), 1.03 (s,
1.5 H).
N-(((6aR,7R,8R,10aS)-10a-(4-Chlorophenylsulfonyl)-1,4-
difluoro-8-methyl-6a,7,8,9,10,10a-hexahydro-6H-benzo[c]-
chromen-7-yl)methyl)-1,1,1-trifluoromethanesulfonamide (35)
1H NMR (CDCl3 400 MHz)  7.62 (d, J = 8.4 Hz, 2 H),
7.48 (d, J = 8.4 Hz, 2 H), 7.07 (m, 1 H), 6.46 (m, 1 H), 5.18
(d, J = 12.8 Hz, 1 H), 4.40 (d, J = 12.8 Hz, 1 H), 3.52 (dd, J
= 13.2 and 4.8 Hz, 1 H), 3.22 (dd, J = 13.2 and 4.8 Hz, 1 H),
2.68 (d, J = 11.6 Hz, 1 H), 2.30 (m, 1 H), 2.16 (m, 1 H),
2.05 (m, 1 H), 1.72 (m, 1 H), 1.55 (m, 1 H), 1.29 (m, 1 H),
1.00 (d, J = 6.8 Hz, 3 H).
Using the same procedures as described above and
starting from pure enantiomer (S)-1-azido-4-(benzyloxy)
butan-2-amine, compounds 3638 were synthesized. Since
the starting material was a chiral compound, the resulting
diastereomers 37 could be separated with normal column
chromatography. As the result, they were separated and
only the active diastereomers were taken forward.
((6aR,7R,10aS)-10a-(4-chlorophenylsulfonyl)-1,4-difluoro-
7-methyl-6a,7,8,9,10,10a-hexahydro-6H-benzo[c]chromen-
7-yl)methanol (31)
Cis and trans isomers were separated by column chro-
matography and trans compound is studied in more detail
1
because it is the more active isomer. H NMR (CDCl3 400
MHz)  7.4452 (m, 4 H), 7.07 (m, 1 H), 6.39 (m, 1 H),
5.18 (dd, J = 12.4 and 4.8 Hz, 1 H), 4.74 (d, J = 12.4 Hz, 1
H), 3.45 (d, J = 11.6 Hz, 1 H), 3.05 d, J = 11.6 hz, 1 H),
2.57-2.64 (m, 2 H), 1.882.02 (m, 2 H), 1.58 (m, 1 H),
1.201.28 (m, 2 H), 1.19 (m, 3 H).
Using the same procedures as described above and start-
ing from optically pure (R)-4-methyltetrahydro-2H-pyran-
2-one, compounds 3235 were synthesized. Since the start-
ing material was a chiral compound, the different stereo

1700
Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
N-((S)-1-Azido-4-(benzyloxy)butan-2-yl)-4-(4-chlorophenyls-
3.00 (m, 1 H), 2.65 (m, 1 H), 2.33 (m, 1 H), 1.26 (2, 9 H).
ulfonyl)-5,8-difluorochroman-3-amine (36)
1
This was a mixture of diastereomers. H NMR (CDCl3
N-(((4R,4aR,10bS)-10b-(4-Chlorophenylsulfonyl)-7,10-diflu-
oro-2,3,4,4a,5,10b-hexahydro-1H-chromeno[3,4-c]pyridin-
4-yl)methyl)methanesulfonamide (41)
400 MHz)  7.74 (d, J = 8.4 Hz, 1.35 H), 7.64 (d, J = 8.4
Hz, 0.65 H), 7.50 (d, J = 7.6 Hz, 1.35 H), 7.43 (d, J = 7.6
Hz, 0.65 H), 7.197.37 (m, 5 H), 6.947.03 (m, 1 H),
6.376.44 (m, 1 H), 4.92 (dd, J = 12.0 and 2.8 Hz, 0.35 H),
4.79 (dd, J = 12.0 and 2.8 Hz, 0.65 H), 4.204.62 (m, 3 H),
3.90 (m, 1 H), 2.903.58 (m, 5 H), 1.441.76 (m, 2 H).
1H NMR (CDCl3 400 MHz)  7.63 (d, J = 8.4 Hz, 2 H),
7.51 (d, J = 8.4 Hz, 2 H), 7.09 (m, 1 H), 6.47 (m, 1 H), 5.15
(dd, J = 12.4 and 2.4 Hz, 1 H), 4.49 (d, J = 12.4 Hz, 1 H),
3.33-3.45 (m, 2 H), 2.99-3.06 (m, 1 H), 3.00 (s, 3 H), 2.76
(m, 1 H), 2.582.66 (m, 2 H), 2.40 (m, 1 H), 2.07 (m, 1 H).
(3S,4aR,10bR)-3-(Azidomethyl)-10b-(4-chlorophenylsul-
fonyl)-7,10-difluoro-2,3,4,4a,5,10b-hexahydro-1H-chromeno-
[3,4-b]pyridine (37)
This was the more potent enantiomer that was isolated
from the reaction. 1H NMR (CDCl3 400 MHz)  7.59 (d, J
= 8.8 Hz, 2 H), 7.49 (d, J = 8.8 Hz, 2 H), 7.07 (m, 1 H),
6.43 (m, 1 H), 5.17 (dd, J = 12.0 and 2.4 Hz, 1 H), 4.29 (d, J
= 12.0 Hz, 1 H), 3.86 (bs, 1 H), 3.69 (m, 1 H), 3.26 (m, 1 H),
3.08 (m, 1 H), 3.41 (m, 2 H), 1.401.80 (m, 2 H).
N-(((4R,4aR,10bS)-10b-(4-chlorophenylsulfonyl)-7,10-diflu-
oro-3-methyl-2,3,4,4a,5,10b-hexahydro-1H-chromeno[3,4-
c]pyridin-4-yl)methyl)methanesulfonamide (42)
1H NMR (CDCl3 400 MHz)  7.66 (d, J = 8.4 Hz, 2 H),
7.52 (d, J = 8.4 Hz, 2 H), 7.10 (m, 1 H), 6.49 (m, 1 H), 5.17
(dd, J = 13.2 and 2.8 Hz, 1 H), 4.51 (d, J = 13.2 Hz, 1 H),
3.44 (s, 2 H), 3.00 (s, 3 H), 2.782.84 (m, 2 H), 2.62 (m, 1
H), 2.18 (s, 3 H), 1.942.18 (m, 3 H).
N-(((3S,4aR,10bR)-10b-(4-Chlorophenylsulfonyl)-7,10-
difluoro-2,3,4,4a,5,10b-hexahydro-1H-chromeno[3,4-b]pyridin-
3-yl)methyl)methanesulfonamide (38)
We thank Jesse Wong, Jianshe Kong, Mark Liang and Tao Meng for scal-
ing up of intermediates; Rebecca Osterman, Andy Evans, T. M. Chan and
Alexei Buevich for determining and confirming stereochemistry of the
compounds via NMR studies.
This was a pure enantiomer. 1H NMR (CDCl3 400 MHz)
 7.57 (d, J = 8.4 Hz, 2 H), 7.48 (d, J = 8.4 Hz, 2 H), 7.07
(m, 1 H), 6.43 (m, 1 H), 5.15 (dd, J = 12.0 and 2.8 Hz, 1 H),
5.06 (d, J = 12.0 Hz, 1 H), 4.29 (d, J = 11.6 Hz, 1 H), 3.78
(bs, 1 H), 3.41 (m, 1 H), 3.10 (m, 1 H), 3.03 (m, 1 H), 2.99
(s, 3 H), 1.60 (m, 1 H), 1.45 (m, 1 H).
1
(a) Hardy JA, selkoe DJ. The amyloid hypothesis of Alzheimer’s dis-
ease: Progress and problems on the road to therapeutics. Science,
2002, 297: 353356; (b) Selkoe DJ. The molecular pathology of
Alzheimer's disease. Neuron, 1991, 6: 487498; (c) Hardy JA,
Higgins GA. Alzheimer's disease: The amyloid cascade hypothesis.
Science, 1992, 256: 184185
Using the same procedures as described above and
starting from methyl 2-((2-(benzyloxy)ethyl)(tert-butoxy-
carbonyl)amino)acetate, compounds 3942 were synthesized.
The compounds were racemic mixtures of cis isomer
between hydroxylmethyl and sulfone.
2
(a) Wu W, Zhang L. -Secretase inhibitors for the treatment of Alz-
heimer's disease. Drug Develop Res, 2009, 70: 94100; (b) Ngu-
yen J, Yamani A, Kiso Y. Views on amyloid hypothesis and secretase
inhibitors for treating Alzheimer's disease: Progress and problems.
Curr Pharmaceut Des, 2006, 12: 42954312
3
4
Harrison T, Churcher I, Beher D. -Secreatase as a target for drug in-
tervention in Alzheimer’s disease. Curr Opin Drug Discov, 2004, 7:
709719
Methyl-2-((2-(benzyloxy)ethyl)(tert-butoxycarbonyl)amino)-
2-(4-(4-chlorophenylsulfonyl)-5,8-difluorochroman-3-yl)ac-
etate (39)
(a) Neitzel ML, Aubele DL, Marugg JL, Jagodzinski JJ, Konradi AW,
Pleiss MA, Szoke B, Zmolek W, Goldbach E, Quinn KP, Sauer JM,
Brigham EF, Wallace W, Bova MP, Hemphill S, Basi G.
Amino-caprolactam γ-secretase inhibitors showing potential for the
treatment of Alzheimer's disease. Bioorg Med Chem Lett, 2011,
37153720; (b) Brodney MA, Auperin DD, Becker SL, Bronk BS,
Brown TM, Coffman KJ, Finley JE, Hicks CD, Karmilowicz MJ,
Lanz TA, Liston D, Liu X, Martin BA, Nelson RB, Nolan CE, Obors-
ki CE, Parker CP, Richter KE, Pozdnyakov N, Sahagan BG,
Schachter JB, Sokolowski SA, Tate B, Van Deusen JW, Wood DE,
Wood KM. Diamide amino-imidazoles: A novel series of γ-secretase
inhibitors for the treatment of Alzheimer's disease. Bioorg Med Chem
Lett, 2011, 26312638; (c) Hamblett CL, Shah S, Heidebrecht R, Jr
Munoz B. -Secretase inhibition: an overview of development of in-
hibitors for the treatment of Alzheimer's disease. Meth Princ Med
Chem, 2010, 45: 353390; (d) Nguyen J-T, Yamani A, Kiso Y.
Views on amyloid hypothesis and secretase inhibitors for treating
Alzheimer's disease: Progress and problems. Curr Phar Des, 2006,
12: 42954312
1
This was a mixture of diastereomers. H NMR (CDCl3
400 MHz)  7.88 (m, 1 H), 7.63 (m, 1 H), 7.51 (m, 1 H),
7.45 (m, 1 H), 7.157.37 (m, 5 H), 6.9507 (m, 1 H),
6.476.52 (m, 1 H), 4.605.10 (m, 2 H), 4.48 (m, 2 H),
4.294.40 (m, 2 H), 4.004.15 (m, 2 H), 3.403.70 (m, 5 H).
1.43 (m, 9 H),1.361.45 (m, 2 H).
(4R,4aR,10bS)-tert-Butyl-10b-(4-chlorophenylsulfonyl)-7,10-
difluoro-4-(hydroxymethyl)-4,4a,5,10b-tetrahydro-1H-chro-
meno[3,4-c]pyridine-3(2H)-carboxylate (40)
This was a racemic mixture. However, it had cis
configuration between hydroxymethyl and sulfone groups.
1H NMR (CDCl3 400 MHz)  7.60 (d, J = 8.4 Hz, 2 H),
7.47 (d, J = 8.4 Hz, 2 H), 7.01 (m, 1 H), 6.40 (m, 1 H), 5.03
(dd, J = 12.0 and 2.4 Hz, 1 H), 4.35 (d, J = 12.0 Hz, 1 H),
4.00 (bs, 1 H), 3.79 (m, 2 H), 3.43 (m, 1 H), 3.18 (m, 1 H),
5
(a) Xu R, Cole D, Asberom T, Bara T, Bennett C, Burnett D, Clader J,
Domalski M, Greenlee W, Hyde L, Josien H, Li H, McBriar M,

Xu R, et al. Sci China Chem November (2011) Vol.54 No.11
1701
McKittrick B, Pissarnitski D, Qiang L, Rajagopalan M, Sasikumar T,
Su J, Tang H, Wu W, Zhang L, Zhao Z. Design and synthesis of tricy-
clic sulfones as -secretase inhibitors with greatly reduced Notch tox-
icity. Bioorg Med Chem Lett, 2010, 25912596; (b) Su J, Tang H,
McKittrick BA, Xu R, Clader JW, Greenlee WJ, Hyde L, Zhang L..
Synthesis and SAR study of tricyclic sulfones as γ-secretase inhibitors:
C-6 and C-8 positions. Bioorg Med Chem Lett, 2010, 34473451; (c)
Li H, Xu R, Cole D, Clader JW, Greenlee WJ, Nomeir AA, Song L,
Zhang L. Design, synthesis, and structure-activity relationship studies
of N-arylsulfonyl morpholines as γ-secretase inhibitors. Bioorg Med
Chem Lett, 2010, 20: 66066609. (d) Sasikumar TK, Qiang L, Burnett
DA, Cole D, Xu R, Li H, Greenlee WJ, Clader J, Zhang L, Hyde L.
Tricyclic sulfones as orally active γ-secretase inhibitors: Synthesis and
structure-activity relationship studies. Bioorg Med Chem Lettt, 2010,
12: 36323635
(a) Josien H, Bara T, Rajagopalan M, Asberom T, Clader JW, Fa-
vreau L, Greenlee WJ, Hyde LA, Nomeir AA, Parker EM, Pissar-
nitski DA, Song L, Wong GT, Zhang L, Zhang Q, Zhao Z. Small
conformationally restricted piperidine N-arylsulfonamides as orally
active γ-secretase inhibitors. Bioorg Med Chem Lett, 2007, 17:
53305335; (b) McBriar MD, Clader JW, Chu I, Del Vecchio RA,
Favreau L, Greenlee WJ, Hyde LA, Nomeir AA, Parker EM, Pis-
sar-nitski DA, Song L, Zhang L, Zhao Z. Discovery of amide and
het-eroaryl isosteres as carbamate replacements in a series of orally
active γ-secretase inhibitors. Bioorg Med Chem Lett, 2008, 18:
215219; (c) Li H, Asberom T, Bara TA, Clader JW, Greenlee WJ,
Josien HB, McBriar MD, Nomeir A, Pissarnitski DA, Rajagopalan
M, Xu R, Zhao Z, Song L, Zhang L. Discovery of 2,4,6-trisubsti-
tuted N-arylsulfonyl piperidines as γ-secretase inhibitors. Bioorg
Med Chem Lett, 2007, 17: 62906294; (d) Asberom T, Zhao Z, Bara
TA, Clader JW, Greenlee WJ, Hyde LA, Josien HB, Li W, McPhail
AT, Nomeir AA, Parker EM, Rajagopalan M, Song L, Wong GT,
Zhang L, Zhang Q, Pissarnitski DA. Discovery of γ-secretase inhib-
itors efficacious in a transgenic animal model of Alzheimer’s disease.
Bioorg Med Chem Lett, 2007, 17: 511516
7
8
Unpublished internal data
(a) Zhang L, Song L, Terracina G, Liu Y, Pramanik B, Parker E. Bio-
chemical characterization of the -secretase activity that produces - am-
yloid peptides. Biochemistry, 2001, 40: 50495055; (b) Hyde LA,
McHugh NA, Chen J, Zhang Q, Manfra D, Nomeir AA, Josien H, Ba-
ra T, Clader JW, Zhang L, Parker EM, Higgins GA. Studies to inves-
tigate the in vivo therapeutic window of the -secretase inhibitor
N2-[(2S)-2-(3,5-difluorophenyl)-2-hydroxyethanoyl]-N1-[(7S)-5-methyl-
6-oxo-6,7-dihydro-5H-dibenzo[b,d]azepin-7-yl]-L-alaninamide (LY411,
575) in the CRND8 mouse. J Pharm Exp Therap, 2006, 319:
11331143
6