Large-scale synthesis of SB-462795, a cathepsin K inhibitor: the RCM-based approaches

Article abstract of DOI:10.1016/j.tet.2009.06.022

Two stereoselective syntheses of SB-462795, a highly potent cathepsin K inhibitor, are described. Both routes feature a C5-C6 disconnection by ring closing metathesis to construct an azepane ring and are amenable to large-scale manufacturing.

Full text of DOI:10.1016/j.tet.2009.06.022

Tetrahedron 65 (2009) 6291–6303
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Large-scale synthesis of SB-462795, a cathepsin K inhibitor: the RCM-based
approaches
*
Huan Wang , Hayao Matsuhashi, Brian D. Doan, Steven N. Goodman, Xi Ouyang, William M. Clark, Jr.
Chemical Development, GlaxoSmithKline, 709 Swedeland Road, King of Prussia, PA 19406, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Two stereoselective syntheses of SB-462795, a highly potent cathepsin K inhibitor, are described. Both
routes feature a C5–C6 disconnection by ring closing metathesis to construct an azepane ring and are
amenable to large-scale manufacturing.
Received 2 March 2009
Received in revised form 1 June 2009
Accepted 2 June 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Available online 11 June 2009
Keywords:
RCM
Oxazolidinone cleavage
Curtius
Epoxide opening
1. Introduction
osteoarthritis.2 Its progression into clinical trials has necessitated
a synthetic route that is amenable to large-scale synthesis.
Cathepsin K is a cysteine protease of the papain super family that
plays a crucial role in bone degradation. Synthesized by osteoclast
cells, it dissolves the collagenous matrix after the bone has been de-
mineralized by acid and its inhibition therefore suppresses bone
destruction. Unlike bisphosphonate therapies that cause death of
osteoclasts, cathepsin K inhibition does not affect the osteoclastic
function of signaling for the replacement of resorbed bone by new
bone, thus preventing bone loss while allowing bone reformation to
continue. This characteristic of cathepsin K inhibitors provides
a crucial advantage over bisphosphonates, the current market
leader in osteoporosis treatment, and their development has
attracted significant attention.1 SB-462795 (1, Fig. 1) has recently
emerged as a potent cathepsin K inhibitor and demonstrated sig-
nificant therapeutic potential in the treatment of osteoporosis and
One of the main challenges in the synthesis of 1 exists in the
azepanone ring bearing two stereocenters at C4 and C7. While
several synthetic methods were successfully employed in the for-
mation of this ring during our investigation, a group of routes based
on a ring closing metathesis (RCM)3–5 disconnection of the C5–C6
bond proved to be very efficient in constructing this moiety. As
depicted in Scheme 1, the formation of the azepane ring 2 is
O
O
H
N
N
H
N
N
O
O
5
S
O
6
O
1
OH
OH
RHN
RHN
O
O
3
H
N
4
N
N
N
N
S
O
S
O
O
N
H
O
N
N
O
O
S
O
7
O
2
3
1, SB-462795
H
N
S
O O
N
Figure 1. Structure of SB-462795.
4
* Corresponding author. Tel.: þ1 610 270 5362; fax: þ1 610 270 4022.
E-mail address: huan.2.wang@gsk.com (H. Wang).
Scheme 1.
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.06.022

6292
H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
effected by RCM of a dienoaminoalcohol6 of the general structure 3,
in which the C7 stereocenter is derived from allyl sulfonamide 4.
Herein we wish to report our efforts in the development of
a scaleable route based on this RCM approach.
the stereochemistry of these two carbons offers a great deal of
potential in lowering the cost of the synthesis; and this possibility
was extensively explored. One of the RCM-based routes that adopts
this strategy is shown in Scheme 4.
O
2. Results and discussion
H
b
a
2.1. Establishing the C7 stereocenter: synthesis of chiral
allyl sulfonamide 4
N
N
S
O
N
S
N
O
O O
EtO
Br
OEt
O
4
13
Chiral allyl sulfonamide 4 was prepared from allyl chloride 5 in
four steps, as shown in Scheme 2. Displacement of the chloride
using potassium phthalimide provided racemic allylic phthalimide
6, which was resolved by Simulated Moving Bed (SMB) chromato-
graphy7 to afford chiral imide 7 in 48% yield and 98% ee. Depro-
tection using ethanolamine followed by HCl treatment gave chiral
amine 8 as the HCl salt in 82% yield. Coupling with 2-pyr-
idinesulfonyl chloride (9), prepared from 2-mercaptopyridine by
oxidation,8 afforded chiral sulfonamide 4 as a white crystalline solid
in excellent yield and enantiomeric purity on >700 kg scale.
N
OtBu
H
14
12
OH
OH
BocHN
BocHN
c
N
N
N
N
S
S
O
O
O
O
15, 2:2:1:1 mixture
16
OH
H2N
f, g, h
d, e
O
O
11
N
Cl
a
b
N
S
N
N
O
O
O
O
d
O
17
OH
5
6
7
N
H
O
O
NH2 HCl
c
H
N
18
S
N
Scheme 4. (a) Cs2CO3, 12, NMP; then H2SO4, 71%; (b) 14, sec-BuLi (2.2 equiv), ZnCl2
(1.1 equiv), THF, ꢀ20 ꢁC, 95%; (c) 2nd generation Grubbs catalyst (14 mol %), CH2Cl2, 95%;
(d) H2, RhCl(PPh3)3, HCl, 75%; (e) aq K2CO3, CH2Cl2, 70%; (f) 18, EDCI, HOOBt, NMM,
CH2Cl2; (g) DMSO, Ac2O; (h) Et3N, MeOH, 94% yield, 61% purity (by HPLC) over three steps.
O O
4
O
S
8
N
Cl
O
9
Scheme 2. (a) Potassium phthalimide, K2CO3, DMF, 83%; (b) SMB chromatography,
48%, 98% ee; (c) ethanolamine, EtOH, HCl, 82%; (d) 9, CH2Cl2, Et3N, 87%, 98% ee.
Alkylation of pyridinesulfonamide 4 with bromide 12 followed
by hydrolysis of the acetal gave aldehyde 13 in 71% yield over two
steps. The coupling between aldehyde 13 and the dianion of N-Boc-
2.2. Synthesis of the azepane moiety: carbon chain assembly
and RCM
10
allylamine mediated by ZnCl2 provided diene 15 in excellent
yield, with inconsequentially low diastereoselectivity at the newly-
formed stereogenic centers. Ring closure of the diastereomeric
mixture was accomplished using Grubbs’s second-generation cat-
alyst. Hydrogenation of the resultant C5–C6 double bond followed
by Boc removal gave a stereoisomeric mixture of aminoalcohols 17.
After coupling with side chain 18, the product mixture was directly
oxidized to the corresponding ketones and subjected to epimeri-
zation conditions. The desired product 1 was ultimately obtained,
however, the overall purity of the product was low and attempts to
purify it by crystallization were not successful.
2.2.1. Epimerizablility of the C4 center and a nonstereo-
selective route
It should be noted that with the C7 stereocenter established, the
C4 stereocenter of the azepanone is readily epimerizable and does
not have to be synthetically defined. In the case of the di-
astereomeric mixture bearing a 7R-methyl group (10, Scheme 3),
the cis diastereomer, which also happens to have the greater ca-
thepsin K inhibitory potency, is strongly favored over the trans
isomer in the epimerization equilibrium.5,9
The main challenge in the scale-up of this route stemmed from
its non-stereoselective nature, which led to mixtures of stereoiso-
mers for all of the later intermediates and made purifications vir-
tually impossible without chromatography. The difficulty in
purification eventually outweighed the potential economic benefit
of not setting the C3 and C4 stereochemistry, and it was concluded
that these two stereogenic centers would have to be carefully
controlled in a route suitable for large scale manufacturing.
This realization led to two stereoselective routes to the target
(Scheme 5), whereby the azepanone moiety is derived from either
of the two C3-epimeric aminoalcohols 19 and 23. The former would
arise from an Evans aldol condensation between aldehyde 13 and
crotonate imide 22, to afford the syn product 21. The stereochem-
istry at C4 would be preserved in a subsequent Curtius reaction to
install the amino group. The C3/C4 stereocenters in 23, however,
would be derived from epoxide 25, which in turn would be syn-
thesized using Sharpless epoxidation of divinyl carbinol (26), with
the C4 center inverted during the introduction of phthalimide via
a Mitsunobu reaction. In both cases, the azepane ring formation
would be effected using olefin metathesis.
O
O
H
N
N
H
O
O
N
N
S
O
10
O
O
O
H
S
ref. 5
N
N
H
O
O
N
N
S
O
S:R = 9:1
11
O
Scheme 3.
This fortuitous property of ketone 10 provides the opportunity
of constructing the carbon chain without defining the stereo-
chemistry at C3 and C4, but establishing the C4 stereocenter by
epimerization after the oxidation of the C3 carbinol. Not controlling

H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
6293
obtained in excellent yield, and the primary impurities were small
amounts (<2%) of unconsumed 27 and the anti diastereoisomer
(Scheme 6).
O
O
H
N
N
H
N
N
O
O
O
O
O
S
O
O
OH
1
O
O
CH2Cl2
0 °C
N
O
N
O
+
N
N
N
S
O
N
S
O
O
R
R
O
OH
OH
H2N
H2N
13
22, R=Bn
27, R=Ph
21, R=Bn
28, R=Ph
N
N
N
N
S
O
S
O
Enolization conditions
R
Yield/%
O
O
19
23
TiCl4 (1 equiv), sparteine (2 equiv)
TiCl4 (1 equiv), added to Hunig's base (1.1 equiv), NMP (2 equiv)
Bn
Ph
33
91
Scheme 6.
O
O
OH
OH
HO
N
Initial studies showed that the RCM reaction of 21 (or 28) required
very high catalyst loadings (Table 1).14 The reaction proceeded to
completion with 10 mol % of Hoveyda’s second generation catalyst
34,15 but lowerloadings(5%)led to incomplete reaction (entries 1 and
2). EvenlowerconversionwasobservedwiththeGrubbscatalysts(31,
33)16 and indenylidene catalyst 32.17 When Ti(Oi-Pr)4 was used as an
additive with catalysts 31 or 33,18 it led to significant decomposition
of the diene, not only by retro-aldol pathway but also cleavage of the
oxazolidinone moiety by 2-propoxide.
N
N
S
O
O
N
N
S
O
O
O
20
24
O
O
OH
O
N
N
N
H
N
O
+
O
O
N
N
N
S
S
O
O O
O
Table 1
Ph
RCM of oxazolidinone-dienes 21 and 28
21
4
25
O
O
OH
O
OH
O
RCM
catalyst
N
N
O
O
O
O
N
N
N
N
S
O
S
O
O
R
R
O
OH
O
O
N
+
S
N
21, R = Bn
28, R = Ph
29, R = Bn
30, R = Ph
O O
Ph
22
13
26
NMes
Ru
MesN
Cl
Cl
Ph
PCy3
Ru
PCy3
Scheme 5.
MesN
Cl
Cl
NMes
Ph
PCy3
Ru
PCy3
Cl
Cl
Cl
Cl
Ru
Ph
O
PCy3
2.2.2. Synthesis of aminoalcohol 19: the aldol–RCM route
The aldol–RCM disconnection via aminoalcohol 19 shown in
Scheme 5 was based on a synthesis previously developed at GSK for
the des-C7-methyl analog.11 At the outset, it was anticipated that
the current C7-methyl analog would also be accessible via this
strategy, however, it was soon discovered that the key aldol–RCM–
Curtius sequence posed significant challenges to efficient scale-up
(vide infra) and had to be completely retooled.
31
32
33
34
Entry
R
Catalyst
Loading/mol %
Conditions
Conv./%
1
2
3
4
5
Bn, Ph
Bn
Ph
Bn
Ph
34
34
31
32
33
10
5
5
5
5
Reflux, toluene
Reflux, toluene
DCM, 40 ꢁC
DCE, rt to 80 ꢁC
DCE, 80 ꢁC
100
60–79
6
7
24
The synthesis of aminoalcohol 19 began with an aldol reaction
between aldehyde 13 and crotonate imide 22.12 The modified
procedure by Crimmins was employed to mitigate the need for
boron triflate and cryogenic temperatures during the aldol cou-
pling.13 Thus, crotonate 22 was treated with TiCl4 to form a thick
slurry, to which sparteine was added to generate the titanium
enolate. Addition of aldehyde 13 gave the desired adduct 21 in
33% yield. The primary side product resulted from the di-
merization of oxazolidinone 22, presumably via a 1,4-addition to
the TiCl4-activated crotonate by the newly formed Ti enolate
during the addition of base. Although the insoluble nature of the
titanium complex of benzyl-substituted oxazolidinone 22 pre-
cluded the use of an inverse addition procedure, the complex
between phenyl-substituted 27 and TiCl4 was highly soluble in
CH2Cl2. Using the phenyl derivative 27 in lieu of 22 in the aldol
reaction thus allowed for an inverse-addition of its TiCl4 complex
to base for the enolate formation, which conveniently circum-
vented the dimerization side reaction. The desired product 28 was
The high catalyst loading necessary for the completion of the
reaction was detrimental to the scale-up of this process. The high
cost of the catalyst aside, it proved extremely difficult to remove all
residual ruthenium from the reaction. The product decomposed
when the reaction mixture was treated with ruthenium scavenger
P(CH2OH)4Cl/NaOH,19 presumably via a retro-aldol pathway. Taking
the crude product with a high ruthenium content forward was
impractical from a safety standpoint, as the ensuing chemical
transformations involved reagents or intermediates sensitive to
heavy metals (H2O2 for oxazolidinone removal and the acyl azide
intermediate in the Curtius reaction). The high catalyst loading in
RCM and the issues it entailed, therefore, rendered the aldol–RCM–
Curtius sequence originally used in the des-methyl analog virtually
inapplicable in this investigation, and a new method was needed to
convert the aldol adduct into aminoalcohol 19.

6294
H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
It was rationalized that the electron rich oxazolidinone and free
Fortunately, it was found that the oxazolidinone moiety in aldol
adduct 28 could be cleaved using hydrazine to give hydrazide 43
(Scheme 10).26 Treatment of hydrazide 43 with tert-butyl nitrite in
the presence of HCl readily generated acyl azide 44, which un-
derwent immediate Curtius rearrangement. The resultant iso-
cyanate was trapped by the neighboring C3 hydroxyl group to
provide carbamate 38 in excellent yield. It should be pointed out
that this aldol–Curtius sequence could serve as a general method
for the synthesis of chiral allylamine derivatives and 4-vinyl oxa-
zolidinones. This extension of the classic Evans aldol methodology
has been previously unexplored, presumably due to the high mi-
C3 hydroxyl group might coordinate to the ruthenium center and
inhibit catalyst turnover, thus leading to the high catalyst loading.
This hypothesis was supported by the observation that silylated
analog 35 underwent complete RCM with a much lower catalyst
loading (Scheme 7).20
O
O
OTES
N
OTES
N
O
O
a
N
N
O
O
N
N
S
S
Ph
Ph
O
O
O
O
gration propensity of the
b,
g-double bond.
35
36
O
O
Scheme 7. (a) 34 (2 mol %), reflux toluene, 4 h, 99% conversion.
OH
O
OH
N2H4
N
NH2NH
O
N
N
N
N
A reasonable way of circumventing the high catalyst loading,
therefore, was to carry out the oxazolidinone auxiliary removal and
Curtius reaction first, and perform the RCM reaction on carbamate
38 (Scheme 8). Carbamate 38 does not have a chelation pocket and
should allow for lower catalyst loading in the RCM. This new se-
quence also eliminates the safety concern of carrying heavy metals
into oxazolidinone removal step and the Curtius reaction. Fur-
thermore, RCM product 39 is expected to be more stable to base
and tolerate workup conditions for ruthenium removal.
98%
S
O
S
O
Ph
O
O
28
43
O
ONO
OH
60 °C
98%
N3
N
N
HCl, 60 °C
S
O
O
44
O
O
OH
OH
O
O
O
N
HO
O
HN
N
N
N
N
S
O
S
O
O
Ph
O
N
N
S
O
O
28
37
38
O
O
Scheme 10.
O
O
HN
HN
N
N
Olefin migration in both the starting material and the product
was a significant side reaction in the RCM of 38.27 The product (39)
underwent significant double bond migration (20–30%) when the
reaction was conducted in refluxing toluene while lower temper-
ature (60 ꢁC) suppressed this side reaction completely. Further-
more, when crude carbamate 38 from the Curtius reaction was used
without chromatographic purification, the RCM reaction was ex-
tremely capricious, often resulting in unconsumed starting material
while producing 40 and 42 as side products from olefin isomeri-
zation, especially at higher temperatures (Scheme 11). Although the
reaction proceeded in a number of solvents,28 no single solvent
provided the desired ring closing metathesis with a consistent and
predictable amount of the isomerization side products. For in-
stance, when the reaction was conducted in THF at 60 ꢁC, the result
ranged from 87% conversion with 2% isomerization to 37% con-
version with 60% isomerization. Addition of HOAc,29 styrene or
N
N
S
O
S
O
O
O
38
39
Scheme 8.
Therefore, aldol product 28 was hydrolyzed to give acid 37,21
which was subjected to diphenylphosphoryl azide (DPPA) and
base.22,23 The ensuing Curtius reaction of
b,g-unsaturated acid 37
occurred smoothly, however, isomerization product 40 was ob-
served along with desired product 38 (Scheme 9). Presumably, the
removal of oxazolidinone facilitated the migration of the double
bond into conjugation under the reaction conditions.24 Although
the amount of 40 could be reduced when hindered bases were
used, its formation was never completely suppressed.25 It became
clear that a non-basic method would be needed for the acyl azide
formation to prevent double bond migration.
O
O
O
O
OH
O
OH
O
LiOH
H2O2
O
HN
HN
DPPA
base
N
HO
O
+
N
N
N
N
S
O
90%
N
S
O
N
N
N
S
O
S
O
Ph
O
O
O
O
28
37
38
40
O
base
O
OH
O
Curtius
HN
N3
N
N
S
O
N
N
S
O
O
O
41
42
Scheme 9.

H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
6295
O
O
O
O
34 (1-2 mol%)
O
O
O
O
HN
HN
HN
HN
60 °C
+
+
N
N
N
N
N
N
N
N
S
O
S
O
S
O
S
O
O
O
O
O
38, crude
39
42
40
Scheme 11.
Cy3PO30 to the reaction mixture did not lead to appreciable im-
provement in the conversion or consistent suppression of olefin
migration.
O
O
O
N
a
b
O
Fortunately, it was found that isolating 38 as its HCl salt from the
Curtius reaction greatly improved its purity consistency and affor-
ded a much more robust RCM reaction, allowing for its scale-up.
The RCM reaction was performed in EtOAc at 60 ꢁC and the mi-
gration side products were observed to be <2%. Upon completion of
the reaction, the residual ruthenium was removed by extraction
using a basic solution of cysteine.31,32 Cyclic carbamate 39 was
isolated in 90% yield after crystallization and contained only small
amounts of residual ruthenium. Hydrogenation and deprotection of
the carbamate afforded aminoalcohol 19 as a white crystalline solid
in 91% yield. Scheme 12 outlines the finalized sequence for the
route to aminoalcohol 19.
OH
OH
26
46
25
Scheme 13. (a) Ti(Oi-Pr)4 10 mol %, (ꢀ)-DIPT 13 mol %, CH2Cl2, tert-butyl hydroper-
oxide, 1.8 equiv, 75% yield, 97% ee; or cumene hydroperoxide 1.1 equiv, 90% yield, 98%
ee; (b) DIAD, 1.3 equiv, PPh3 1.3 equiv, toluene, 85%.
precipitation of triphenylphosphine oxide and diisopropyl hydra-
zinedicarboxylate35 could purge >80% of these by-products, nu-
merous other impurities resulting from both the Sharpless
oxidation and the Mitsunobu reaction made the crystallization of 25
extremely difficult. Consequently, batch chromatographic purifica-
tions were necessary to afford 25 of suitable purity. While this se-
quence was successfully used to deliver >200 kg of epoxide 25 (in
60% yield) for early-phase supplies, the chromatographic operations
were a significant bottleneck in the overall synthetic route and
added significantly to the cost.
O
O
OH
O
O
O
a
N
O
N
+
O
N
N
S
O
N
N
S
O
O
Ph
Ph
27
O
In developing this route for use on commercial scale, it was
discovered that reaction of epoxide 25 with toluenesulfonate un-
der acidic conditions provided the more crystalline tosylate 47
(mp¼129.6 ꢁC).36 Tosylate 47 could be generated from crude ep-
oxide 25 and crystallized directly from the crude reaction mixtures
after the Sharpless and Mitsunobu steps in excellent purity and
yield. As shown in Table 2, optimized formation of the desired
tosylate 47 was achieved by using anhydrous toluenesulfonate
reagents (to minimize the formation of diol 48). Among the
tosylate nucleophiles screened, tetraalkylammonium tosylates
(entries 4 and 5) afforded exclusive epoxide opening from the
primary carbon, providing the desired product in excellent yield.
Epoxide 25 could then be regenerated from 47 in nearly quanti-
tative yield by treatment with DBU in toluene. Incorporation of
13
28
O
O
OH
O
b
HN
c
NH2NH
N
N
S
O
N
N
S
O
O
O
O
43
38
O
O
O
d
HN
e
HN
N
N
S
O
N
N
S
O
O
O
39
45
OH
H2N
tosylate 47 provided
a more scalable and cost-effective in-
f
termediate by eliminating the need for chromatographic purifi-
cation of epoxide 25. The process was readily scalable, allowing for
the preparation of 850 kg of tosylate 47 in 50% yield from divinyl
carbinol.
N
N
S
O
O
19
Generation of requisite RCM precursor 24 was accomplished by
coupling epoxide 25 with sulfonamide 4. Although this type of
coupling reaction is well precedented, initial attempts using in-
organic carbonates37 or Triton B failed to give the desired coupling
product 24. In most cases, the epoxide either did not react, or
simply decomposed during the reaction, presumably due to its
relative instability. Fortunately, it was found that this reaction could
be catalyzed by small amounts of organic bases in refluxing IPA
(Table 3).38,39 Among the bases screened, DBU and BTPP proved to
be the most efficient in catalyzing this coupling reaction. Good
yields were obtained when DBU was used as the catalyst, although
a main side pathway appeared to be slow reaction between DBU
and epoxide 25 to form an adduct. Thus, the yield was further
improved by w10% when the highly basic yet non-nucleophilic
tert-butylimino-tri(pyrrolidino)phosphorane (BTPP)40 was used in
lieu of DBU as the catalyst (Table 3).
Scheme 12. (a) TiCl4, Hunig’s base, NMP, 91%; (b) N2H4, quant.; (c) tert-butyl nitrite,
HCl, isopropanol, 50 ꢁC, 84%; (d) 34 (1 mol %), EtOAc, 60 ꢁC, then cysteine/NaOH, 90%,
Ru 148 ppm; (e) DME, Pd/C, H2; (f) aq NaOH, reflux, 91% over 2 steps, Ru 14 ppm.
2.2.3. Synthesis of aminoalcohol 23: the Sharpless–RCM route
Epoxide 25, the key intermediate in the synthesis of trans
aminoalcohol 23, was prepared from divinyl carbinol in two steps
(Scheme 13). An asymmetric Sharpless epoxidation using diiso-
propyl tartrate ligand and cumene hydroperoxide as the oxidant
provided epoxide 46 in 98% ee and 90% yield.33 Consistent with
literature observations,34 cumene hydroperoxide provided a higher
yield for epoxide 46 relative to tert-butyl hydroperoxide. Treatment
of 46 with phthalimide under Mitsunobu conditions provided the
desired product 25 in 85% yield.
Isolation of the low-melting phthalimidoepoxide 25 (mp¼
72.3 ꢁC), however, presented a significant challenge. Although co-

6296
H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
Table 2
Epoxide opening of 25 with toluenesulfonate
OH
OH
OTs
N
O
TsO
HO
+
HO
a
p-TsOH
(1.1 equiv)
N
O
N
N
O
O
O
O
O
+
O
O
salt additive
(1.05 equiv)
25
47
48
49
Entry
Salt additives
Solvent
Rxn temp/ꢁC
Diol 48/%b
Isomer 49/%b
Yield of 47/%
1
2
3
4
5
d
Toluene
CH2Cl2
CH2Cl2
Toluene
Toluene
60
40
40
60
60
15
13
6
2
2
15
4
2
<1
<1
50
64
70
97
96
2,6-Lutidinium OTs hydrate
Pyridinium OTs
NBu4OTs
NEt4OTs
a
Prepared from TsOH hydrate by azeotropic removal of H2O.
By area% in HPLC.
b
Table 3
Organic base-catalyzed epoxide opening of 25 with sulfonamide 4
O
O
N
SO2 OH
N
base (10 mol%)
IPA, reflux
H
N
O
+
N
N
S
N
O
O
O O
4
24
25
N
P
N
N
N
BTPP
Entry
Base
Yield/%
Entry
Base
Yield/%
1
2
3
Et3N
DABCO
Pyridine
<5
<5
30
4
5
6
DMAP
DBU
BTPP
63
73
85
Diene 24 proved to be an exceptional substrate for RCM. When it
is of high purity, complete conversion could be achieved with 0.1–
0.2 mol % of Hoveyda’s catalyst 34.41 Perhaps due to the steric bulk
of the allylic substituents, olefin migration was never observed,
even in cases where the diene was of low purity and the reaction
did not reach completion. Hoveyda’s catalyst 34 was found to be the
best catalyst for this RCM reaction. Grubbs 2nd generation catalyst
33 was also effective on small scale, however, the results were not
O
OH
c
b
N
O
O
O
N
N
OH
S
O
OH
OH
H2N
e
O
a
N
N
51
N
N
S
O
O
O
N
N
d
N
N
S
O
S
O
O
O
OH
O
H2N
23
24
50
N
N
S
O
O
52
Scheme 14. (a) 34, toluene, 96%; (b) 10% Pd/C, 21 dry wt %, H2 120 psi, 82%; (c) N2H4, 93%; (d) NH2OH, MeOH, 73%; (e) 10% Pd/C, 3 dry wt %, H2 60 psi, 98%.

H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
6297
formation. This crystallization also effectively purged the residual
ruthenium from the isolated product, giving low metal content
which was further lowered by downstream processing.
O
O
O
OH
a
b, c
N
O
Presumably due to the steric bulk of the phthalimide, RCM
product 50 exhibited very low reactivity in the subsequent hydro-
genation. Coupled with its low solubility, the hydrogenation of 50
required a high catalyst loading (21% w/w) and relatively high di-
lution (Scheme 14). Conversely, the hydrogenation of deprotected
aminoalcohol52 was much morefacile, requiring significantlylower
catalyst loading (w3% w/w) and hydrogen pressure. A plausible
explanation is that the hydrogenation of 52 is facilitated by co-
ordination of the C4 amino group to Pd, as it was observed that this
hydrogenation was significantly retarded by small amounts of acid.
Scheme 15 summarizes the synthetic route to aminoalcohol 23.
OH
26
46
25
O
OH
c
d
N
O
N
N
S
O
O
24
O
OH
OH
H2N
e
N
2.3. Synthesis of the side chain, coupling and oxidation
of C3 carbinol
N
N
S
O
O
f
N
N
S
O
O
O
50
52
Side chain 18 was prepared from known acid 5343 and
L-leucine
(Scheme 16). This sequence was used to produce >450 kg of 18 in
80% yield.
Undermostoftheconditions tested, thestereocenteronsidechain
18 was prone to partial racemization during the coupling with aze-
pane aminoalcohol23(Table 4). Asmuch as10%racemizationproduct
OH
H2N
N
N
S
O
O
23
Scheme 15. (a) Ti(Oi-Pr)4, 10 mol %, (ꢀ)-DIPT 13 mol %, CH2Cl2, cumene hydroperoxide
1.1 equiv, 90%, 98% ee; (b) DIAD, 1.3 equiv, PPh3, 1.3 equiv, tol, 85%; (c) 4, BTPP 10 mol %,
isopropanol, 73%; (d) 34 (0.5 mol %), toluene, 96%, residual Ru 359 ppm; (e) NH2OH,
MeOH, 73%; (f) 10% Pd/C, H2, 98%.
O
SOCl2, l-leucine
80%
COOH
OH
N
H
O
O
O
consistent upon scale-up and low conversion was often observed
even with pure diene 24.42 This RCM reaction was most conve-
niently performed in toluene, in which the product had very low
solubility and readily crystallized from the reaction mixture upon
53
18
Scheme 16.
Table 4
HOOBt-mediated amide coupling between 23 and 18
OH
H2N
O
EDCI, CH2Cl2, rt
+
OH
N
N
H
N
S
O
O
O
O
23
18
O
O
OH
OH
H
N
H
N
+
N
H
N
H
O
O
O
O
N
N
N
N
S
S
O
O
O
O
54
55
O
N
OH
N
N
HOOBt
Entry
Additive
Equiv
Racemization/%
1
2
3
4
5
6
7
d
d
10
32
25
0.8
1.2
5.3
0.4
DMAP
HOBt
HOOBt
HOOBt
HOOBt
HOOBt
1
1
1
0.01
0.001
0.01a
a
Reaction at 0 ꢁC.

6298
H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
54 was observed in the reaction mixture when the EDCI-mediated
coupling was carried out without any additive, and 32% racemization
occurred when DMAP wasused as an additive.44 The formation of this
diastereomer was not suppressed by the addition of HOBt (entries 1–
3).45 However, 3-hydroxy-1,2,3-benzotriazin-4(3H)-one (HOOBt)46
proved to be an effective additive in minimizing the racemization.
Only a very small amount of HOOBt was needed as the amount of
diastereomer 54 only increased slightly when the HOOBt loading was
lowered from 1 equivalent to 1 mol % (entries 4 and 5). Even when
only 0.1 mol % of HOOBt was used, the amount of racemization was
still significantly below thatobtained intheblank experiment. Finally,
the couplingreactionwascarried out with 1 mol % of HOOBtinCH2Cl2
at 0 ꢁC (entry 7). The coupling proceeded smoothly, providing the
desired product in 86–92% yield with >99% purity.
synthesized via trans aminoalcohol 23. The multi-kilo synthesis
also featured one of the first examples of large scale RCM in
a pharmaceutical setting.50
4. Experimental details
4.1. General
Unless otherwise indicated, all reactions were conducted in
glass-lined reactors under a nitrogen atmosphere. Solvents, re-
agents, allyl sulfonamide 4, phthalimidoepoxide 25, tosylate 47 and
side chain 18 were obtained from commercial sources and used
without further purification.
The oxidation of alcohol 55 to 1 has been extensively studied.47
After evaluating numerous conditions, it was found that this oxi-
dation was best accomplished using DMSO/Ac2O48 or NaOCl with
TEMPO as the catalyst.49 Both methods provided the desired ketone
product in good isolated yield (>85%), although the activated
DMSO conditions tended to be operationally more convenient and
more reproducible.
Although developed using trans aminoalcohol 23, this coupling–
oxidation sequence was also applicable to cis aminoalcohol 19. At
the same HOOBt loading, the coupling reaction with 19 was slightly
slower and gave higher amounts of racemization by-product than
that of 23, but 1 was obtained starting from either aminoalcohol in
comparable yield and purity profile, as shown in Scheme 17. Fur-
thermore, the process using 23 was successfully piloted on multi-
kilo scale, producing >200 kg of SB-462795 (1) which was suitable
for clinical studies.
4.2. N-[(1R)-1-methyl-2-propen-1-yl]-2-pyridinesulfon-
amide (4)
ES-MS: m/z 213 (MþHþ). 1H NMR
d ppm (CDCl3, 400.13 MHz):
8.72 (ddd, J¼4.7, 1.7, 0.9 Hz, 1H), 8.01 (dt, J¼7.9, 1.0 Hz, 1H), 7.90 (td,
J¼7.7, 1.8 Hz, 1H), 7.49 (ddd, J¼7.6, 4.8, 1.2 Hz, 1H), 5.64 (ddd, J¼17.2,
10.4, 6.0 Hz, 1H), 5.12 (br, 1H), 5.06 (dt, J¼17.2, 1.2 Hz, 1H), 4.92 (dt,
J¼10.4, 1.1 Hz, 1H), 4.03 (m, 1H), 1.22 (d, J¼6.8 Hz, 3H). 13C NMR
d
ppm (CDCl3, 100.61 MHz): 158.0, 149.9, 138.8, 137.9, 126.5, 122.2,
115.0, 52.2, 21.3. Anal. Calcd for C9H12N2O2S: C 50.92, H 5.70, N
13.20. Found: C 51.03, H 5.79, N 13.12.
4.3. N-(1-Benzofuran-2-ylcarbonyl)-L-leucine (18)
ES-MS: m/z 276 (MþHþ), 298 (MþNaþ), 339 (MþNaþþCH3CN).
1H NMR
d
ppm (CDCl3, 400.13 MHz): 7.68 (d, J¼7.9 Hz, 1H), 7.53 (m,
2H), 7.44 (ddd, J¼8.5, 7.2, 1.2 Hz, 1H), 7.31 (m, 1H), 7.01 (d, J¼8.3 Hz,
1H), 4.89 (m, 1H), 1.83 (m, 3H), 1.02 (dd, J¼6.1, 1.2 Hz, 6H). 13C NMR
OH
d
ppm (CDCl3, 100.61 MHz): 176.8, 159.0, 154.8, 147.7, 127.4, 127.2,
H2N
O
a, b
123.8, 122.8, 111.8, 111.5, 50.6, 41.4, 24.9, 22.8, 21.8. Anal. Calcd for
C15H17NO4: C 65.44, H 6.22, N 5.09. Found: C 65.55, H 6.24, N 5.04.
+
OH
N
N
H
N
S
O
O
O
O
4.4. 4,5-Anhydro-1,2,3-trideoxy-3-(1,3-dioxo-1,3-dihydro-2H-
isoindol-2-yl)-L-threo-pent-1-enitol (25)
19, 23
18
O
ES-MS: m/z: 230 (MþHþ), 252 (MþNaþ), 293 (MþNaþþCH3CN).
O
H
N
1H NMR
d ppm (CDCl3, 400.13 MHz): 7.82 (m, 2H), 7.70 (m, 2H), 6.14
N
H
(m, 1H), 5.29 (m, 2H), 4.46 (tm, J¼7.2 Hz, 1H), 3.68 (m, 1H), 2.92
O
O
N
N
S
O
(ddd, J¼6.3, 2.4, 2.1 Hz, 1H), 2.76 (m, 1H). 13C NMR
d ppm (CDCl3,
O
1
100.61 MHz): 167.6, 134.0, 131.7, 130.9, 123.3, 119.4, 56.2, 50.9, 46.8.
Anal. Calcd for C13H11NO3: C 68.12, H 4.84, N 6.11. Found: C 68.19, H
4.76, N 6.11.
Scheme 17. (a) EDCI, HOOBt, CH2Cl2, 0 ꢁC; (b) DMSO, Ac2O. From 19: 75% over 2 steps;
from 23: 74% over 2 steps.
3. Summary
4.5. 1,2,3-Trideoxy-3-(1,3-dioxo-1,3-dihydro-2H-isoindol-2-
yl)-5-O-[(4-methylphenyl)sulfonyl]-L-threo-pent-1-enitol (47)
In conclusion, two stereoselective synthetic routes to SB-
462795, a highly potent cathepsin K inhibitor, have been de-
veloped. In the first route, an aldol–Curtius–RCM sequence is used
to construct the key azepane moiety. It is highlighted by: (1)
a highly efficient aldol reaction involving a crotonate imide using
Crimmins modifications; (2) cleavage of the oxazolidinone using
hydrazine followed by a Curtius reaction to convert the resultant
ES-MS: m/z: 402 (MþHþ), 424 (MþNaþ). 1H NMR
d ppm (CDCl3,
400.13 MHz): 7.70–7.82 (m, 6H), 7.30 (d, J¼8.0 Hz, 2H), 6.12 (ddd,
J¼17.2, 10.2, 8.0 Hz, 1H), 5.24 (m, 2H), 4.89 (td, J¼7.8, 0.8 Hz, 1H),
4.44 (dt, J¼7.8, 4.9 Hz, 1H), 4.08 (d, J¼4.8 Hz, 2H), 3.25 (br, 1H), 2.40
(s, 3H). 13C NMR
d ppm (CDCl3, 100.61 MHz): 168.4, 145.0, 134.2,
132.2, 131.5, 131.1, 129.8, 127.9, 123.5, 120.4, 70.1, 68.6, 55.8, 21.6.
Anal. Calcd for C20H19NO6S: C 59.84, H 4.77, N 3.49. Found: C 59.91,
H 4.68, N 3.39.
b,g-unsaturated carbonyl intermediate into an allylamine de-
rivative without double bond migration; and (3) an efficient RCM
reaction that utilized low catalyst loading. The second route, uti-
lizing a Sharpless epoxidation–Mitsunobu sequence to establish
the C3/C4 stereocenters, is highlighted by: (1) a selective epoxide
opening using TsOH to give the monotosylate of a vicinal diol; (2)
an efficient epoxide opening reaction by sulfonamide catalyzed by
BTPP; and (3) a remarkably efficient RCM reaction requiring ex-
tremely low catalyst loading. Both routes are amenable to large-
scale manufacturing, and over 200 kg of SB-462795 has been
4.6. N-[(1R)-1-Methyl-2-propen-1-yl]-N-(2-oxoethyl)-2-
pyridinesulfonamide (13)
A mixture of 4 (1.0 kg, 4.71 mol), NMP (4 L) and Cs2CO3 (2.3 kg,
7.07 mol) was heated to 115 ꢁC under nitrogen. Once the temper-
ature was above 100 ꢁC, bromoacetaldehyde diethyl acetal 12
(1.393 kg, 7.07 mol) was charged via an addition funnel over

H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
6299
approx. 2 h, and the mixture was heated at 115 ꢁC for a total of 17 h.
The slurry was cooled to 25 ꢁC and H2O (6 L) was added in portions
over approx. 10 min to control the slightly exothermic quench be-
low 35 ꢁC. The acetal intermediate was extracted into TBME/hep-
tane (1:2 ratio, 5 Lꢂ2), and the combined organic extracts were
washed with 1 M aq NaOH (2 L). The solution of the diethyl acetal
intermediate was distilled under vacuum (approx. 30 mmHg) to
remove w10 L of solvent. THF (2 Lꢂ2) was added and subsequently
distilled under vacuum to further drive off the extraction solvents.
THF (4 L) and 10% v/v H2SO4 (4 L) were added and the mixture
heated to 60 ꢁC for 2.5 h. The solution was then cooled to <35 ꢁC,
toluene (3 L) was charged, and the mixture distilled to remove 5–
6 L of solvent (mostly THF). The remaining layers were separated.
The aqueous layer was extracted with toluene (3 L). The combined
organic extracts were washed sequentially with satd aq NaHCO3
(1 L) and H2O (1 L), and more solvent (1.2 L) was removed by vac-
uum distillation while keeping the internal temperature at 37–
42 ꢁC.51 Toluene was added and removed by vacuum distillation in
2 L portions (total of 5 portions) until NMR analysis showed com-
plete removal of EtOH and THF, and the level of BrCH2CHO was
<1%. A total of 4.14 kg of toluene solution was obtained, which was
stored in a freezer. NMR and LC w/w analysis showed the solution
contained 20.6 wt % of 13 (3.35 mol, 71%, typically 70–80%). The
aldehyde was thermally unstable and was used as a toluene solu-
tion directly in the aldol reaction. An analytically pure sample could
be obtained by flash column chromatography. ES-MS: m/z: 255
(MþHþ), 277 (MþNaþ). ES-HRMS: calcd for [C11H14N2O3SþH]þ:
1.05 mol) was added. The resultant solution was added slowly to
a separate vessel charged with diisopropylethlyamine (195 mL,
141.2 g, 1.09 mol) in CH2Cl2 (2.3 L) while maintaining the temper-
ature at ꢀ5 ꢁC. NMP (199.2 g, 193.1 mL, 2.01 mol) was added and
the resulting dark purple mixture was stirred cold for 2 h. A solu-
tion of aldehyde 13 in 1:1 DCM/toluene (1.736 kg, 20.6% w/w,
1.41 mol) was charged and the resulting mixture was stirred for
90 min. The mixture was quenched by sequential addition of acetic
acid (300 mL), saturated aqueous Rochelle’s salt (400 mL), and 4 L
of 10% v/v aqueous HCl. The mixture was warmed to 20 ꢁC and the
layers separated. The organic layer was washed with 4 L of 10% v/v
aqueous HCl, then mixed vigorously with saturated aqueous
NaHSO3 at 40 ꢁC until all residual aldehyde 13 was consumed. The
layers were separated. The organic layer was concentrated under
vacuum and was used in further processing without any purifica-
tion (total weight: 522.7 g, product 28: 85% w/w, 0.915 mol, 91%
yield). An analytically pure sample could be obtained by flash col-
umn chromatography. ES-MS: m/z: 486 (MþHþ), 508 (MþNaþ). 1H
NMR
d
ppm (CDCl3, 400.13 MHz): 8.64 (dm, J¼4.7 Hz, 1H), 8.01 (d,
J¼7.9 Hz,1H), 7.91 (td, J¼7.7,1.5 Hz,1H), 7.49 (ddd, J¼7.6, 4.8, 0.7 Hz,
1H), 7.23–7.36 (m, 5H), 5.92 (ddd, J¼17.3,10.1, 8.9 Hz,1H), 5.79 (ddd,
J¼17.5, 10.3, 5.6 Hz, 1H), 5.45 (dd, J¼8.7, 3.6 Hz, 1H), 5.23 (m, 2H),
5.05 (m, 2H), 4.99 (br, 1H), 4.67 (m, 2H), 4.34 (m, 2H), 4.20 (dd,
J¼8.9, 3.7 Hz, 1H), 3.57 (dd, J¼15.1, 9.3 Hz, 1H), 3.40 (dd, J¼15.1,
2.5 Hz, 1H), 1.22 (d, J¼6.9 Hz, 3H). 13C NMR
d ppm (CDCl3,
100.61 MHz): 171.5, 158.2, 153.2, 149.4, 138.8, 138.4, 137.8, 131.5,
128.9, 128.5, 126.7, 125.6, 122.8, 120.4, 116.8, 69.9, 69.7, 57.5, 55.8,
50.8, 48.8, 17.0. Anal. Calcd for C24H27N3O6S: C 59.37, H 5.60, N 8.65.
Found: C 59.67, H 5.90, N 8.32.
255.0803; found: 255.0800. 1H NMR
d ppm (CDCl3, 400.13 MHz):
9.58 (t, J¼1.2 Hz, 1H), 8.63 (dm, J¼4.8 Hz, 1H), 7.88 (m, 2H), 7.47
(ddd, J¼6.8, 4.9, 1.6 Hz, 1H), 5.59 (ddd, J¼17.3, 10.6, 4.7 Hz, 1H), 5.04
(m, 2H), 4.67 (m, 1H), 3.95 (dd, J¼19.0, 1.3 Hz, 1H), 3.83 (dd, J¼19.0,
4.9. (4S)-3-{[(3S,4S,7R)-3-Hydroxy-7-methyl-1-(2-
pyridinylsulfonyl)-2,3,4,7-tetrahydro-1H-azepin-4-
yl]carbonyl}-4-phenyl-1,3-oxazolidin-2-one (30)
1.4 Hz, 1H), 1.11 (d, J¼6.9 Hz, 3H). 13C NMR
d ppm (CDCl3,
100.61 MHz): 199.5, 157.6, 150.0, 137.9, 136.4, 126.8, 122.1, 117.7,
54.6, 52.5, 16.6.
In a round bottom flask, aldol product 28 (233 mg, 0.480 mmol,
purified by flash column chromatography) was dissolved in tolu-
ene (2 mL, pre-degassed), and Hoveyda’s catalyst 34 (30.0 mg,
0.048 mmol, 10 mol %) was added. The resultant solution was again
degassed and heated to reflux. HPLC analysis of the reaction
mixture after 2.5 h showed that diene 28 was consumed. The
product was purified by flash column chromatography (170 mg,
0.372 mmol, 78%). ES-MS: m/z: 458 (MþHþ), 480 (MþNaþ), 937
(2MþNaþ). ES-HRMS: calcd for [C22H23N3O6SþH]þ: 458.1386;
4.7. (4S)-3-[(2E)-2-Butenoyl]-4-phenyl-1,3-oxazolidin-2-one (27)
The synthesis of 27 is based on procedure by Ho and Mathre.52
LiCl (0.299 kg, 7.05 mol) and Et3N (0.744 kg, 7.35 mol) were added to
(4S)-4-phenyl-1,3-oxazolidin-2-one (1.00 kg, 6.13 mol) in THF
(10 L). The mixture was stirred for 20 min at 23 ꢁC, then cooled to
15 ꢁC. Crotonic anhydride (0.964 kg, 6.25 mol) was added over
approx. 30 min while maintaining the internal temperature below
20 ꢁC. After the addition, the solution was stirred for 1 h at 23 ꢁC.
The reaction was quenched with the addition of H2O (5 L). The or-
ganic solvent was removed by vacuum distillation while maintain-
ing the internal temperature between 20–30 ꢁC. Once most of the
THF was removed, the oily product converted to a free-flowing solid
and was filtered. H2O (3 Lꢂ2) was charged to rinse the reactor and
the wet cake. The wet cake was then washed with TBME/heptane at
10 ꢁC (1:1 ratio, 3 Lꢂ2). The product was dried in a vacuum oven at
40 ꢁC to afford 27 (1.286 kg, 5.56 mol, 91%, HPLC purity 99.5%). ES-
MS: m/z: 232 (MþHþ), 254 (MþNaþ), 295 (MþNaþþCH3CN), 485
found: 458.1382. 1H NMR
d ppm (CDCl3, 400.13 MHz): 8.68 (d,
J¼4.3 Hz, 1H), 8.00 (d, J¼7.8 Hz, 1H), 7.90 (ddd, J¼7.7, 7.7, 1.3 Hz,
1H), 7.48 (m, 1H), 7.26–7.39 (m, 5H), 5.47 (m, 3H), 5.13 (m, 1H), 4.96
(m, 1H), 4.74 (t, J¼8.8 Hz, 1H), 4.60 (m, 1H), 4.24 (dd, J¼8.9, 3.9 Hz,
1H), 4.15 (dd, J¼15.5, 6.6 Hz, 1H), 3.29 (br, 1H), 3.23 (dd, J¼15.5,
7.3 Hz, 1H), 1.09 (d, J¼7.0 Hz, 3H). 13C NMR
d ppm (CDCl3,
100.61 MHz): 173.2, 158.1, 152.8, 149.9, 138.9, 138.1, 133.0, 129.0,
128.5, 126.6, 125.6, 122.4, 122.2, 70.4, 70.0, 57.6, 54.7, 47.5, 45.2,
18.2.
(2MþNaþ). 1H NMR
d
ppm (CDCl3, 400.13 MHz): 7.26–7.41 (m, 6H),
4.10. N-[(2S,3S)-3-(Hydrazinocarbonyl)-2-hydroxy-4-
penten-1-yl]-N-[(1R)-1-methyl-2-propen-1-yl]-2-
pyridinesulfonamide (43)
7.11 (ddd, J¼15.2,13.8, 6.8 Hz,1H), 5.50 (dd, J¼8.7, 3.8 Hz,1H), 4.71 (t,
J¼8.8 Hz,1H), 4.29 (dd, J¼8.9, 3.9 Hz,1H),1.95 (dd, J¼6.9,1.6 Hz, 3H).
13C NMR
d ppm (CDCl3,100.61 MHz): 164.3,153.6,147.1,139.0,129.0,
128.5, 125.8, 121.6, 69.8, 57.6, 18.4. Anal. Calcd for C13H13NO3: C
67.52, H 5.67, N 6.06. Found: C 67.24, H 5.62, N 5.93.
Aldol product 28 (436.7 g, 0.90 mol) was dissolved in THF
(1.25 L). The solution was cooled to 0 ꢁC and hydrazine hydrate
(55 mL, 56.8 g, 1.13 mol) was added dropwise. After the addition of
10 mL of hydrazine hydrate, the reaction mixture became cloudy.
Seed crystals of hydrazide 43 were added and the addition was
continued. After 90 min, TBME (2 L) was added and the mixture
was stirred at 0 ꢁC for 1 h. The resultant precipitate was collected
by filtration, washed with a 1:1 mixture of THF/TBME (2 L), air
dried for 1 h, then vacuum dried overnight at 50 ꢁC to give
4.8. N-((2S,3S)-2-Hydroxy-3-{[(4S)-2-oxo-4-phenyl-1,3-
oxazolidin-3-yl]carbonyl}-4-penten-1-yl)-N-[(1R)-1-methyl-2-
propen-1-yl]-2-pyridinesulfonamide (28)
Crotonate imide 27 (234.0 g, 1.01 mol) was dissolved in CH2Cl2
(2.3 L) at ꢀ10 ꢁC under nitrogen and titanium chloride (200 g,

6300
H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
hydrazide 43 as a white solid (329.2 g, 0.93 mol, quant.). ES-MS:
solution over 1.5 h. The reaction mixture was heated at 60 ꢁC for
another 1.5 h and HPLC analysis of a reaction sample showed the
conversion was complete. Cysteine (13.16 g, 0.109 mol, 0.5 equiv)
and NaOH (19.2 g, 0.48 mol, 2.2 equiv) were dissolved in 400 mL of
H2O, and the resultant solution was slowly added to the reaction
mixture. A dark mixture formed. The biphasic mixture was stirred
at 60 ꢁC overnight and cooled to rt. The dark aqueous layer was
separated and discarded. The light-amber organic layer was
washed with aqueous NaOH solution (1 N, 400 mLꢂ2) and con-
centrated to w100 mL. Heptane (300 mL) was added and the re-
sultant suspension was stirred at rt overnight, filtered and dried at
45 ꢁC under vacuum to provide carbamate 39 (59.74 g, 0.193 mol,
89%, purity by HPLC: 98.8%). ES-MS: m/z: 310 (MþHþ), 332
m/z: 355 (MþHþ), 377 (MþNaþ). 1H NMR
d ppm (DMSO-d6,
400.13 MHz): 9.03 (s, 1H), 8.73 (dm, J¼4.7 Hz, 1H), 8.08 (td, J¼7.7,
1.7 Hz, 1H), 7.94 (d, J¼7.9 Hz, 1H), 7.67 (ddd, J¼7.6, 4.7, 1.0 Hz, 1H),
5.86 (m, 1H), 5.67 (m, 1H), 5.12 (m, 1H), 5.09 (m, 1H), 5.00 (m, 3H),
4.40 (m, 1H), 4.22 (s, 2H), 3.99 (m, 1H), 3.25 (dd, J¼15.0, 2.4 Hz,
1H), 3.10 (dd, J¼15.0, 9.2 Hz, 1H), 2.85 (dd, J¼8.6, 6.4 Hz, 1H), 1.18
(d, J¼6.9 Hz, 3H). 13C NMR
d ppm (DMSO-d6, 100.61 MHz): 170.6,
157.6, 150.2, 138.8, 138.4, 134.7, 127.2, 122.5, 118.0, 116.5, 70.1, 55.3,
53.6, 49.3, 17.7. Anal. Calcd for C15H22N4O4S: C 50.83, H 6.26, N
15.81. Found: C 50.94, H 6.09, N 15.66.
4.11. N-{[(4S,5R)-4-Ethenyl-2-oxo-1,3-oxazolidin-5-
yl]methyl}-N-[(1R)-1-methyl-2-propen-1-yl]-2-
pyridinesulfonamide (38)
(MþNaþ). 1H NMR
ppm (CDCl3, 400.13 MHz): 8.70 (dm, J¼4.6 Hz,
d
1H), 8.00 (d, J¼7.8 Hz, 1H), 7.92 (td, J¼7.7, 1.7 Hz, 1H), 7.52 (ddd,
J¼7.6, 4.7, 1.1 Hz, 1H), 6.72 (br, 1H), 5.30 (s, 2H), 5.17 (ddd, J¼10.8,
8.4, 6.1 Hz, 1H), 4.75 (m, 1H), 4.64 (dd, J¼8.4, 2.8 Hz, 1H), 4.21 (dd,
J¼14.6, 6.0 Hz, 1H), 3.37 (dd, J¼14.7, 10.9 Hz, 1H), 1.07 (d, J¼7.0 Hz,
To a mixture of hydrazide 43 (521.0 g, 1.47 mol) and 1.3 M HCl in
isopropanol (5.2 L) at 50 ꢁC was added a solution of tert-butyl ni-
trite (200 mL, 172.0 g, 1.67 mol) in IPA (1 L) over 2 h. The solution
was cooled and concentrated under vacuum. The resulting solid
was slurried in isopropyl acetate (1 L), collected by filtration,
washed with isopropyl acetate, and dried at ambient temperature
under vacuum to give the HCl salt of carbamate 38 as a beige solid
(460.3 g, 1.23 mol, 84%). ES-MS: m/z: 338 (MþHþ), 360 (MþNaþ).
3H). 13C NMR
d ppm (CDCl3, 100.61 MHz): 159.0, 157.8, 150.2, 138.2,
131.2, 127.0, 126.6, 122.4, 77.0, 55.1, 52.9, 42.7, 18.2. Anal. Calcd for
C13H15N3O4S: C 50.48, H 4.89, N 13.58. Found: C 50.05, H 4.79, N
13.25.
4.14. (3aR,6R,8aS)-6-Methyl-5-(2-pyridinylsulfonyl)-
octahydro-2H-[1,3]oxazolo[5,4-c]azepin-2-one (45)
1H NMR
d ppm (DMSO-d6, 400.13 MHz): 12.22 (br, 1H), 8.72 (dm,
J¼4.7 Hz, 1H), 8.07 (td, J¼7.8, 1.7 Hz, 1H), 7.95 (m, 2H), 7.66 (dd,
J¼7.6, 4.7 Hz, 1H), 5.83 (ddd, J¼17.1, 10.2, 6.9 Hz, 1H), 5.54 (ddd,
J¼17.1, 10.8, 5.3 Hz, 1H), 5.28 (m, 2H), 5.00 (m, 1H), 4.96 (d,
J¼10.2 Hz, 1H), 4.81 (m, 1H), 4.41 (m, 2H), 3.54 (dd, J¼15.7, 2.0 Hz,
1H), 3.20 (dd, J¼15.8, 9.9 Hz, 1H), 1.18 (d, J¼6.9 Hz, 3H). 13C NMR
RCM product 39 (39.8 g, 0.129 mol) was dissolved in 400 mL of
DME and the resultant mixture was divided in half. Pd/C catalyst
(10%, containing 65% H2O, 4 g) was added to each portion and the
resultant mixtures were hydrogenated at rt overnight, one portion
under 120 psi and the other under 60 psi of H2. No significant
difference was observed in the reaction profile by HPLC between
the two batches. The two portions were recombined, the Pd/C
catalyst was removed by filtration and the product was directly
used in the next step. An analytically pure sample could be
obtained by recrystallization from EtOH/TBME. ES-MS m/z: 312
d
ppm (DMSO-d6, 100.61 MHz): 157.9, 157.5, 150.4, 138.9, 137.5,
133.5,127.6,122.6,118.9,117.3, 77.6, 56.2, 55.3, 45.7,18.0. Anal. Calcd
for C15H20N3O4SCl: C 48.19, H 5.39, N 11.24. Found: C 48.07, 5.38,
11.05.
4.12. General procedure for RCM reaction using crude
carbamate diene 38
(MþHþ), 334 (MþNaþ). 1H NMR
d ppm (CDCl3, 400.13 MHz): 8.69
(dm, J¼4.7 Hz, 1H), 7.97 (d, J¼7.8 Hz, 1H), 7.91 (td, J¼7.7, 1.7 Hz,
1H), 7.50 (ddd, J¼7.5, 4.7, 1.2 Hz, 1H), 6.35 (br, 1H), 4.97 (ddd,
J¼11.2, 8.1, 5.5 Hz, 1H), 4.20 (m, 1H), 3.99–4.09 (m, 2H), 3.20 (dd,
J¼15.2, 11.2 Hz, 1H), 1.90 (m, 1H), 1.65–1.75 (m, 3H), 0.98 (d,
In a round bottom flask under N2, crude carbamate diene 38
(free base) from the Curtius reaction and Hoveyda’s catalyst 34 (1–
2 mol %) were dissolved in degassed solvent (w0.2 M) and the re-
sultant solution was re-degassed. The mixture was heated to 60 ꢁC
and the reaction mixture was analyzed using HPLC. Isomerization
product 42 was unstable and converted into 40 upon isolation by
flash column chromatography.
J¼6.7 Hz, 3H). 13C NMR
d ppm (CDCl3, 100.61 MHz): 159.1, 157.7,
150.0, 138.1, 126.8, 122.3, 76.8, 55.0, 52.4, 41.3, 30.0, 25.0, 17.1.
Anal. Calcd for C13H17N3O4S: C 50.15, H 5.50, N 13.50. Found: C
50.13, H 5.44, N 13.41.
4.12.1. N-[(4-Ethyl-2-oxo-2,3-dihydro-1,3-oxazol-5-yl)methyl]-N-
[(1R)-1-methyl-2-propen-1-yl]-2-pyridinesulfonamide (40)
ES-MS: m/z: 338 (MþHþ), 360 (MþNaþ). ES-HRMS: calcd for
4.15. (3R,4S,7R)-4-Amino-7-methyl-1-(2-pyridinylsulfonyl)-
hexahydro-1H-azepin-3-ol (19)
[C15H19N3O4SþH]þ: 338.1175; found: 338.1175. 1H NMR
d
ppm
The DME solution of the azepane carbamate 45 (0.129 mol) was
concentrated to w200 mL and aqueous NaOH (16.1 g, 0.402 mol,
3.1 equiv, in 400 mL of H2O) was added. The mixture was heated to
reflux for 2.5 h and HPLC analysis of a reaction sample showed the
conversion was complete. The organic solvent was removed by
distillation. The resultant suspension was cooled to rt overnight.
The solid product was collected by filtration, washed with EtOAc
(50 mL) and dried at 50 ꢁC under vacuum to give cis aminoalcohol
19 (33.68 g, 0.118 mol, 91%, purity by HPLC: 100%). Ru content:
(CDCl3, 400.13 MHz): 9.72 (s, 1H), 8.66 (d, J¼4.6 Hz, 1H), 7.84 (m,
2H), 7.44 (dd, J¼8.8, 4.6 Hz, 1H), 5.80 (m, 1H), 5.11–5.15 (m, 2H),
4.71 (m, 1H), 4.32 (d, J¼16.4 Hz, 1H), 4.12 (d, J¼16.4 Hz, 1H), 2.48 (q,
J¼7.6 Hz, 2H), 1.34 (d, J¼6.9 Hz, 3H), 1.18 (t, J¼7.6 Hz, 3H). 13C NMR
d
ppm (CDCl3, 100.61 MHz): 158.5, 156.8, 150.0, 137.8, 137.6, 131.1,
126.6, 126.4, 121.8, 116.8, 55.5, 37.5, 17.4, 16.6, 12.2.
4.13. (3aR,6R,8aS)-6-Methyl-5-(2-pyridinylsulfonyl)-
1,3a,4,5,6,8a-hexahydro-2H-[1,3]oxazolo[5,4-c]azepin-
2-one (39)
14 ppm. ES-MS m/z: 286 (MþHþ), 308 (MþNaþ). 1H NMR
d ppm
(DMSO-d6, 400.13 MHz): 8.73 (dm, J¼4.7 Hz, 1H), 8.07 (td, J¼7.7,
1.6 Hz, 1H), 7.95 (d, J¼7.9 Hz, 1H), 7.66 (ddd, J¼7.6, 4.7, 0.7 Hz, 1H),
4.94 (d, J¼2.7 Hz, 1H), 3.88 (ddd, J¼17.7, 12.1, 6.2 Hz, 1H), 3.50 (m,
1H), 3.39 (m, 1H), 3.19 (m, 1H), 2.98 (br, 1H), 1.87 (ddd, J¼14.2, 11.2,
11.0 Hz, 1H), 1.52 (m, 2H), 1.32 (br, 2H), 1.24 (m, 1H), 0.86 (d,
Carbamate diene 38 (HCl salt, 81.28 g, 0.217 mol) was suspended
in 800 mL of EtOAc and Hoveyda’s catalyst 34 (1.358 g, 2.16 mmol,
1 mol %) was dissolved in 400 mL of EtOAc. The entire system was
degassed and the EtOAc solution of the diene was heated to 60 ꢁC.
The EtOAc solution of the catalyst was slowly added to the diene
J¼6.4 Hz, 3H). 13C NMR
d ppm (DMSO-d6, 100.61 MHz): 158.6, 150.1,
138.7, 127.1, 121.9, 72.1, 53.2, 50.0, 42.8, 27.2, 26.0, 20.1. Anal. Calcd

H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
6301
for C12H19N3O3S: C 50.51, H 6.71, N 14.73. Found: C 50.05, H 6.71, N
14.44.
4.18. 2-[(3S,4S,7R)-3-Hydroxy-7-methyl-1-(2-pyridinyl-
sulfonyl)-2,3,4,7-tetrahydro-1H-azepin-4-yl]-1H-isoindole-
1,3(2H)-dione (50)
4.16. 1,3,4,5-Tetradeoxy-3-(1,3-dioxo-1,3-dihydro-2H-
isoindol-2-yl)-1-[[(1R)-1-methyl-2-propen-1-yl](2-
A solution of diene 24 (80.0 kg, 181 mol) in toluene (880 L) was
degassed. Hoveyda’s catalyst 34 (600 g, 0.5 mol %) was added and
the mixture was degassed once again. The reaction was then heated
to 110 ꢁC and kept for 5 h at this temperature. Upon cooling to
60 ꢁC, 80% tetrakis(hydroxylmethyl)phosphonium chloride (5.8 L)
and sodium bicarbonate (2.7 kg) were added. The resulting mixture
was stirred at 60 ꢁC for 13 h, and cooled to 40 ꢁC. Water (112 L) and
cyclohexane (880 L) were added before further cooling to 20 ꢁC.
The slurry was filtered with centrifuge, washed with water
(112 Lꢂ2) and IPA (56 Lꢂ2), then dried under vacuum at 35 ꢁC to
afford 50 (71.8 kg, 174 mol, 96%) as a light gray solid. ES-MS: m/z:
pyridinylsulfonyl)amino]-
from epoxide 25
L-threo-pent-4-enitol (24)
To a mixture of allyl sulfonamide 4 (75.0 kg, 353 mol) and ep-
oxide 25 (83.4 kg, 364 mol, 1.03 equiv) in 2-propanol (339 L) was
added tert-butylimino-tri(pyrrolidino)phosphorane (BTPP) (11 kg,
35.2 mol,10 mol %) at room temperature. The mixture was heated at
gentle reflux for 8 h before being cooled to 35 ꢁC. Aqueous citric acid
solution (10%, 165 L) was charged to the mixture followed by slow
addition of water (488 L) over 30 min, while maintaining the tem-
perature between 35–40 ꢁC. Seed crystals (300 g, 0.4 mol %) were
added. The resulting suspension was stirred overnight, cooled to
room temperature, and stirred for an additional 1 h at this temper-
ature. The slurry was filtered with centrifuge, washed with cold 50%
aqueous IPA (150 L), and dried under vacuum at 35 ꢁC to afford diene
24 (113 kg, 256 mol, 73%) as a white crystalline solid. ES-MS: m/z:
414 (MþHþ), 436 (MþNaþ), 849 (2MþNaþ). 1H NMR
d (CDCl3,
400.13 MHz): 8.86 (dm, J¼4.8 Hz, 1H), 8.09 (d, J¼7.8 Hz, 1H), 8.01
(td, J¼7.7, 1.7 Hz, 1H), 7.81 (m, 2H), 7.70 (m, 2H), 7.64 (ddd, J¼7.6,
4.8, 1.2 Hz, 1H), 5.58 (ddd, J¼11.0, 5.1, 2.9 Hz, 1H), 5.22 (m, 2H), 4.84
(br, 1H), 4.68 (m, 1H), 4.46 (dd, J¼9.7, 4.4 Hz, 1H), 4.29 (d, J¼16.2 Hz,
1H), 3.91 (dd, J¼16.2, 4.6 Hz, 1H), 1.36 (d, J¼6.9 Hz, 3H). 13C NMR
442 (MþHþ), 464 (MþNaþ), 905 (2MþNaþ). 1H NMR
d
(CDCl3,
d ppm (CDCl3, 100.61 MHz): 168.0, 157.2, 150.6, 138.6, 133.8, 131.6,
400.13 MHz): 8.61 (dm, J¼4.8 Hz, 1H), 8.01 (dm, J¼7.9 Hz, 1H), 7.92
(td, J¼7.7,1.7 Hz,1H), 7.81 (m, 2H), 7.69 (m, 2H), 7.48 (ddd, J¼7.6, 4.8,
1.0 Hz,1H), 6.28 (ddd, J¼17.2,10.1, 8.7 Hz,1H), 5.80 (ddd, J¼17.4,10.6,
5.0 Hz, 1H), 5.37 (d, J¼17.2 Hz, 1H), 5.28 (d, J¼10.2 Hz, 1H), 5.10 (m,
2H), 4.87 (br, 1H), 4.67 (t, J¼9.0 Hz, 1H), 4.59 (td, J¼9.2, 2.6 Hz, 1H),
4.39 (m, 1H), 3.55 (dd, J¼15.1, 2.6 Hz, 1H), 3.44 (dd, 1H, J¼15.2,
131.0, 126.8, 126.7, 123.0, 122.3, 71.5, 54.8, 52.8, 48.3, 19.6. Anal.
Calcd for C20H19N3O5S: C 58.10, H 4.63, N 10.16. Found: C 58.01, H
4.59, N 10.05.
4.19. 2-[(3S,4S,7R)-3-Hydroxy-7-methyl-1-(2-pyridinyl-
sulfonyl)hexahydro-1H-azepin-4-yl]-1H-isoindole-1,3(2H)-
dione (51)
9.2 Hz),1.28 (d, J¼6.9 Hz, 3H).13C NMR
d (CDCl3,100.61 MHz): 168.1,
158.0,149.5,138.5,137.3,133.8,132.2,131.8,126.8,123.2,122.9,120.8,
117.2, 68.7, 57.6, 55.7, 48.9,17.6. Anal. Calcd for C22H23N3O5S: C 59.86,
H 5.25, N 9.52. Found: C 59.80, H 5.26, N 9.49.
To a Hastelloy pressure reactor were charged alkene 50 (28.0 kg,
67.7 mol), 10% Pd/C (PMC1625C, 16.1 kg, 63% wet, 20 wt %) and THF
(420 L). The mixture was stirred at 35 ꢁC under hydrogen pressure
(120 psi) for 4 h. Upon cooling, the mixture was filtered through
a pad of Celite, and the reactor and the pad were rinsed with
dichloromethane (364 L). The combined filtrate was concentrated
under atmospheric pressure to w140 L. Acetone (280 L) was
charged, and the distillation was continued until the volume of the
mixture was w140 L. IPA (28 L) was added. The slurry was heated to
reflux for 30 min, cooled to 0 ꢁC, and filtered with centrifuge. The
wet cake was washed with IPA (56 L), and dried under vacuum at
35 ꢁC to afford 51 (23.1 kg, 55.6 mol, 82%) as an off-white crystalline
solid. ES-MS: m/z: 416 (MþHþ), 438 (MþNaþ), 853 (2MþNaþ). 1H
4.17. 1,3,4,5-Tetradeoxy-3-(1,3-dioxo-1,3-dihydro-2H-
isoindol-2-yl)-1-[[(1R)-1-methyl-2-propen-1-yl](2-
pyridinylsulfonyl)amino]-
from tosylate 47
L-threo-pent-4-enitol (24)
A slurry of tosylate 47 (8.5 kg, 21.2 mol) and toluene (34 L) was
heated to 56 ꢁC and DBU (3.3 kg, 21.7 mol) was added via pres-
sure vessel to maintain the temperature between 55–65 ꢁC. The
starting material dissolved during the addition and the mixture
was aged at this temperature for 60 min. Upon completion of the
reaction, the mixture was cooled to 25 ꢁC and 10% w/w aq citric
acid (0.42 L water with 45 g citric acid monohydrate) was added.
After stirring for 10 min, water (25.5 L) was added, the mixture
stirred and the layers allowed to separate. The organic layer was
then washed twice with water (8.5 Lꢂ2). The solution of crude 25
was then transferred to a second reactor via an inline filter to
remove a small layer of film, and the transfer line then rinsed
with toluene (8.5 L). To this second reactor was charged 4 (4.5 kg,
21.2 mol) and the solvent partially removed by vacuum distilla-
tion to reach 15 L total volume (36.0 L toluene removed). 2-
Propanol (25.5 L) was charged and the system vacuum distilled to
achieve 15 L total volume (25.5 L removed). 2-Propanol (15.0 L)
was added followed by BTPP (0.7 kg, 2.12 mol), and the mixture
heated to 80 ꢁC over w30 min. After aging for 3.5 h, the reaction
was complete and heptane was added (17.0 L) while maintaining
the temperature ꢃ60 ꢁC. The temperature was then adjusted to
53 ꢁC and seeds of 24 (4.5 g in 2:1 heptane/2-propanol) were
charged to the reactor. The slurry was held at 52 ꢁC for 60 min,
then cooled to w10 ꢁC at 0.5 ꢁC/min. After holding for 30 min, the
slurry was isolated by centrifugal filtration, washed with cold
heptane/2-propanol (2:1 v/v, 8.5 Lꢂ2). The product was dried
under vacuum at 45 ꢁC to afford 24 (7.25 kg, 16.4 mol, 78%, HPLC
purity >99.9%).
NMR
d
(CDCl3, 400.13 MHz): 8.87 (dm, J¼4.7 Hz, 1H), 8.06 (d,
J¼7.8 Hz, 1H), 7.98 (td, J¼7.7, 1.7 Hz, 1H), 7.80 (m, 2H), 7.70 (m, 2H),
7.60 (ddd, J¼7.6, 4.8, 1.1 Hz, 1H), 4.90 (br s, 1H), 4.34 (m, 1H), 4.24
(m, 3H), 3.75 (dd, J¼16.6, 3.6 Hz, 1H), 2.48 (m, 1H), 1.57 (m, 3H), 1.23
(d, J¼6.9 Hz, 3H). 13C NMR
d (CDCl3,100.61 MHz): 168.2,157.7,150.2,
138.6, 133.8, 131.9, 126.8, 123.1, 122.4, 72.0, 57.9, 51.2, 46.4, 34.6,
25.3, 17.6. Anal. Calcd for C20H21N3O5S: C 57.82, H 5.09, N 10.11.
Found: C 57.92, H 5.03, N 10.08.
4.20. (3S,4S,7R)-4-Amino-7-methyl-1-(2-pyridinylsulfonyl)-
hexahydro-1H-azepin-3-ol (23) from 51
A solution of phthalimide 51 (32.0 kg, 77.2 mol) in 95% EtOH
(320 L) was heated to w55 ꢁC, and hydrazine monohydrate (9.3 L,
193 mol) was added. The reaction mixture was stirred at 75 ꢁC for
1 h before being cooled to 35 ꢁC. The mixture was diluted with
CH2Cl2 (256 L). The resulting slurry was filtered with centrifuge and
the solid was washed with CH2Cl2 (128 L). The combined filtrate
was concentrated to the minimum stir volume of the vessel. Water
(96 L) was added. The mixture was concentrated under vacuum to
w60 L, extracted with CH2Cl2 (320 Lþ96 L) to give aminoalcohol 23
as a CH2Cl2 solution (quantitative yield assumed). This solution was
used in the next coupling step without further purification. A pure

6302
H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
sample can be obtained by recrystallization from water in 85% yield.
(200 g, 1.23 mol, 1.1 mol %), and the reaction mixture was cooled to
0 ꢁC. EDCI (24.0 kg, 126 mol) was added, and the reaction mixture
was stirred at this temperature for 5 h. Upon the completion of the
reaction, it was quenched with 3% aqueous sodium bicarbonate
(200 L). The organic layer was washed with water (195 L) and
concentrated under vacuum to the minimum stir volume of the
vessel. The residue was diluted with toluene (390 L), and the
mixture was concentrated to ca. 290 L. Toluene (98 L) was added.
The mixture was heated to 90 ꢁC and clear dissolution was ob-
served. The mixture was cooled to 58 ꢁC, seeded (10 g, 0.03 wt %),
and held at this temperature for 30 min. TBME (325 L) was charged
over 30 min at 52 ꢁC. The mixture was cooled to 15 ꢁC over 1 h, and
filtered with centrifuge. The solid product was washed with TBME
(195 L), and dried under vacuum at 45 ꢁC to afford coupling product
55 (50.7 kg, 93.4 mol, 82% over two steps) as a white crystalline
ES-MS: m/z: 286 (MþHþ), 571 (2MþHþ), 593 (2MþNaþ). 1H NMR
d
(DMSO-d6, 400.13 MHz): 8.69 (ddd, J¼4.7, 1.6, 0.8 Hz, 1H), 8.04 (td,
J¼7.7, 1.8 Hz, 1H), 7.89 (d, J¼7.9 Hz, 1H), 7.62 (ddd, J¼7.6, 4.7, 1.1 Hz,
1H), 4.72 (br, 1H), 4.00 (ddd, J¼14.6, 12.7, 6.4 Hz, 1H), 3.47 (m, 2H),
3.23 (m, 1H), 2.59 (ddd, J¼7.4, 5.2, 2.8 Hz, 1H), 1.32–1.68 (m, 6H),
1.03 (d, J¼6.4 Hz, 3H). 13C NMR
d (DMSO-d6, 100.61 MHz): 158.6,
149.6,138.4, 126.8, 122.0, 73.9, 54.4, 53.4, 46.0, 28.4, 26.7, 20.2. Anal.
Calcd for C12H19N3O3S: C 50.51, H 6.71, N 14.73. Found: C 50.51, H
6.84, N 14.67.
4.21. (3S,4S,7R)-4-Amino-7-methyl-1-(2-pyridinylsulfonyl)-
2,3,4,7-tetrahydro-1H-azepin-3-ol (52)
Phthalimide 50 (2.4 kg, 5.8 mol) was added to a solution of
hydroxyamine hydrochloride (0.9 kg, 13 mol, 2.2 equiv) in metha-
nol (19.5 L). A methanol solution of NaOMe (25% w/w, 6.4 kg,
30 mol, 5.2 equiv) was slowly added while maintaining the tem-
perature of the reaction mixture between 22 and 27 ꢁC. The re-
action mixture was stirred at 28–30 ꢁC for 1.5 h. HPLC analysis of
the reaction mixture indicated that the starting material was
consumed. Water (9.6 L) was added and the mixture was distilled
to w10 L. The aqueous mixture was extracted by CH2Cl2
(24 Lþ7.2 L), and the combined organic layer was distilled to w5 L.
Isopropanol (4.8 L) was added and the mixture was concentrated
to w5 L. A second portion of isopropanol (2.4 L) was added and the
mixture was concentrated to w6 L. TBME (5 L) was charged to the
resultant suspension. The mixture was cooled to 0 ꢁC and stirred at
0 ꢁC for 1 h. The product was collected by centrifuge filtration. The
wet cake was washed with TBME (5 L) and dried at 35–45 ꢁC under
vacuum. The product was obtained as a white crystalline solid
(1.2 kg, 4.2 mol, 73%). ES-MS: m/z: 284 (MþHþ), 306 (MþNaþ), 567
solid. ES-MS: m/z: 543 (MþHþ), 565 (MþNaþ). 1H NMR
d (CDCl3,
300.13 MHz): 8.69 (d, J¼4.6 Hz, 1H), 8.05 (dt, J¼7.8, 1.0 Hz, 1H), 7.93
(td, J¼7.8, 1.6 Hz, 1H), 7.66 (ddd, J¼7.8, 1.2, 0.7 Hz, 1H), 7.53 (tm,
J¼3.7 Hz, 1H), 7.50 (m, 1H), 7.48 (s, 1H), 7.44 (ddd, J¼8.4, 7.1, 1.3 Hz,
1H), 7.31 (ddd, J¼7.9, 7.0, 1.1 Hz, 1H), 7.10 (br, 1H), 6.80 (br, 1H), 4.66
(m, 1H), 4.19 (dd, J¼16.3, 3.0 Hz, 1H), 3.93 (m, 2H), 3.74 (m, 1H), 3.46
(d, J¼16.2 Hz, 1H), 1.66–1.87 (m, 6H), 1.40 (m, 1H), 0.97–1.05 (m,
9H). 13C NMR
d (CDCl3,100.61 MHz): 171.5,159.0,158.0,154.8,149.6,
147.8, 138.2, 127.2, 127.1, 126.8, 123.7, 122.9, 122.6, 111.8, 110.9, 72.5,
54.8, 52.5, 51.8, 44.6, 41.0, 30.7, 24.7, 24.4, 22.8, 22.0, 18.6. Anal.
Calcd for C27H34N4O6S: C 59.76, H 6.32, N 10.32. Found: C 59.45, H
6.39, N 10.09.
4.24. SB-462795 (1)
To a solution of coupling product 55 (50.7 kg, 93.4 mol) in DMSO
(507 L) was added acetic anhydride (26.5 L, 280 mol, 3 equiv) at
room temperature, and the resulting solutionwas stirred at 50–55 ꢁC
for 3 h. Upon the completion of the reaction, the mixture was cooled
to room temperature, diluted with ethyl acetate (507 L), and washed
with water (304 Lꢂ3). The organic layer was concentrated under
vacuum to w150 L, warmed to 55 ꢁC, and filtered through an in-line
filter. The reactor and the filter were rinsed with ethyl acetate (51 L),
and the combined product solution was diluted with TBME (507 L).
The mixture was heated to 55 ꢁC. Seeds of 1 (50 g, 0.1 wt %) were
added as a TBME slurry. The mixture was stirred at this temperature
for 1 h, diluted further with warm TBME (507 L), then cooled to
ꢀ10 ꢁC at a rate of 0.5 ꢁC/min. It was filtered with centrifuge, washed
with cold TBME (253 L), and the solid was spun dry under nitrogen to
give 1/TBME complex (69.8 kg). The TBME solvate was slurried in
heptane (760 L), and the mixture was heated at 80 ꢁC for 3 h before
cooling to 20 ꢁC over 2 h. The solid was isolated with centrifuge,
washed with heptane (760 L), and dried under vacuum at 40 ꢁC to
afford SB-462795 (43.1 kg, 79.7 mol, 85%) as a white crystalline solid.
(2MþHþ), 589 (2MþNaþ). 1H NMR
d (DMSO-d6, 400.13 MHz): 8.73
(dm, J¼4.7 Hz, 1H), 8.08 (td, J¼7.8, 1.6 Hz, 1H), 7.94 (d, J¼8.0 Hz,
1H), 7.66 (ddd, J¼7.6, 4.7, 0.5 Hz, 1H), 5.31 (m, 2H), 5.14 (br, 1H),
4.67 (q, J¼6.9 Hz, 1H), 3.64 (dd, J¼14.6, 4.7 Hz, 1H), 3.50 (dd,
J¼14.6, 5.7 Hz, 1H), 3.30–3.40 (m, 2H), 1.53 (br, 2H), 1.01 (d,
J¼7.1 Hz, 3H). 13C NMR
d (CDCl3, 100.61 MHz): 157.6, 150.1, 138.7,
133.4, 130.8, 127.1, 122.1, 73.7, 53.3, 52.3, 48.5, 18.0. Anal. Calcd for
C12H17N3O3S: C 50.87, H 6.05, N 14.83. Found: C 50.85, H 5.79, N
14.66.
4.22. (3S,4S,7R)-4-Amino-7-methyl-1-(2-pyridinylsulfonyl)-
hexahydro-1H-azepin-3-ol (23) from 52
To a Hastelloy pressure reactor were charged alkene 52 (1.1 kg,
3.9 mol), 10% Pd/C (JM10R391, 66 g, 50% wet, 3 wt %) and industrial
methylated spirits (IMS, 11 L, 94% ethanol, 5% methanol, 1% water).
The mixture was stirred at 33–37 ꢁC under hydrogen pressure
(60 psi) for 3.5 h. HPLC analysis of the reaction mixture indicated
that the starting material was consumed. CH2Cl2 (4.4 kg) was
charged and the mixture was stirred at 30 ꢁC for 0.5 h. The mixture
was filtered through a pad of Celite, and the reactor and the pad
were rinsed with CH2Cl2 (4.4 kg). The filtrates were combined
(total weight: 17.6 kg, containing 23: 6.2% w/w, w1.1 kg, 3.8 mol,
98%).
ES-MS: m/z: 541 (MþHþ), 563 (MþNaþ). 1H NMR
d (CDCl3,
300.13 MHz): 8.71 (m, 1H), 7.99 (d, J¼7.9 Hz, 1H), 7.92 (tm, J¼7.7 Hz,
1H), 7.66 (dm, J¼7.8 Hz,1H), 7.50 (m, 3H), 7.42 (tm, J¼7.7 Hz,1H), 7.28
(m,1H), 7.15 (br,1H), 7.02 (br,1H), 5.12 (m,1H), 4.77 (m, 2H), 4.41 (m,
1H), 3.86 (dd, J¼19.4, 2.2 Hz,1H), 2.15 (m, 2H),1.74 (m, 3H),1.45–1.62
(m, 2H), 0.94–1.00 (m, 9H). 13C NMR
d (CDCl3, 100.61 MHz): 206.2,
170.7, 158.5, 157.5, 154.7, 150.2, 148.0, 138.1, 127.3, 127.0, 126.9, 123.6,
122.6, 122.2, 111.8, 110.8, 57.1, 52.1, 52.0, 51.4, 41.8, 33.2, 27.0, 24.7,
22.8, 22.0,15.4. Anal. Calcd for C27H32N4O6S: C 59.98, H 5.97, N 10.36.
Found: C 59.86, H 6.03, N 10.27.
4.23. N-[(1S)-1-({[(3S,4S,7R)-3-Hydroxy-7-methyl-1-
(2-pyridinylsulfonyl)hexahydro-1H-azepin-4-yl]-
amino}carbonyl)-3-methylbutyl]-1-benzofuran-
2-carboxamide (55)
Acknowledgements
The CH2Cl2 solution of aminoalcohol 23 (w114 mol) was dis-
tilled to ca. 330 L under atmospheric pressure, and cooled to 20 ꢁC.
To the resultant mixture were added side chain 18 (31.4 kg,
114 mol) and 3-hydroxy-1,2,3-benzotriazin-4(3H)-one (HOOBt)
The authors thank Neil Badham, Jessica Bailey, Qunying Dai, Ann
Eldridge, John Rougas and Qiaogong Su for their effort in laboratory
investigations and pilot campaigns, Dr. Alan Freyer for assistance in
NMR spectroscopy, and Dr. Lianming Wu for HRMS analysis.

H. Wang et al. / Tetrahedron 65 (2009) 6291–6303
6303
24. Aldol adducts bearing the oxazolidinone moiety have low kinetic acidity due to
allylic strain effect, see: Evans, D. A.; Ennis, M. D.; Le, T. J. Am. Chem. Soc. 1984,
106, 1154.
25. A number of bases, including Et3N, Hunig’s base, NMM, pyridine, DMAP and NaH,
were tested for this reaction. It was found that the ratio of 38/40 varied between 1:
5 and 8:1, with Hunig’s base in toluene providing the highest amount of 38.
26. Bock, M. G.; DiPardo, R. M.; Evans, B. E.; Rittle, K. E.; Boger, J. S.; Freidinger, R. M.;
Veber, D. F. J. Chem. Soc., Chem. Commun. 1985, 109.
Supplementary data
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.tet.2009.06.022.
References and notes
27. For olefin isomerization/migration as
a side reaction to metathesis, see:
Schmidt, B. Eur. J. Org. Chem. 2004, 1865 and references cited therein.
28. Solvents screened include: dichloromethane, dichloroethane, chloroform, tol-
uene, trifluorotoluene, ethyl acetate, isopropyl acetate, 2-methyl THF, THF,
DME, methyl ethyl ketone.
29. Hong, S. H.; Sanders, D. P.; Lee, C. W.;Grubbs, R. H. J. Am. Chem. Soc. 2005,127,17160.
30. Bourgeois, D.; Pancrazi, A.; Nolan, S. P.; Prunet, J. J. Organomet. Chem. 2002,
643–644, 247.
1. Grabowska, U. B.; Chambers, T. J.; Shiroo, M. Curr. Opin. Drug Discov. Dev. 2005,
8, 619.
2. (a) Kumar, S.; Dare, L.; Vasko-Moser, J. A.; James, I. E.; Blake, S. M.; Rickard, D. J.;
Hwang, S.-M.; Tomaszek, T.; Yamashita, D. S.; Marquis, R. W.; Oh, H.; Jeong, J. U.;
Veber, D. F.; Gowen, M.; Lark, M. W.; Stroup, G. Bone 2007, 40, 122; (b) Cum-
mings, M. D.; Marquis, R. W.; Ru, Y.; Thompson, S. K.; Veber, D. F.; Yamashita, D.
S. PCT Int. Appl. WO 2001070232 A1, 2001.
31. Several methods were tested for the removal of residual ruthenium. Adsorptive
methods using silica gel or alumina typically removed <50% of the residual
ruthenium while P(CH2OH)4Cl/base or cysteine/base gave superior results
(>80% Ru removal). The latter was further favored as the reaction between base
and P(CH2OH)4Cl produced formaldehyde which added to the carbamate ni-
trogen. Small amounts of residual P reagents also had a more adverse effect in
the subsequent hydrogenation.
32. Cysteine has been demonstrated to be an effective scavenger for palladium: (a)
Villa, M.; Cannata, V.; Rosi, A.; Allegrini, P. PCT Int. Appl. WO 9851646 A1, 1998;
(b) Fan, Y.; Saenz, J. E.; Shi, B.; Srirangam, J. K.; Yu, S. U.S. Patent Appl. Publ. US
2,006,091,067 A1, 2006.
33. (a) Schreiber, S. L.; Schreiber, T. S.; Smith, D. B. J. Am. Chem. Soc. 1987, 109, 1525;
(b) Smith, D. B.; Wang, Z.; Schreiber, S. L. Tetrahedron 1990, 46, 4793; (c) Jager,
V.; Schroter, D.; Koppenhoefer, B. Tetrahedron 1991, 47, 2195; (d) Jager, V.; Stahl,
U.; Hummer, W. Synthesis 1991, 776.
34. Romero, A.; Wong, C.-H. J. Org. Chem. 2000, 65, 8264.
35. Anderson, N. G.; Lust, D. A.; Colapret, K. A.; Simpson, J. H.; Malley, M. F.;
Gougoutas, J. Z. J. Org. Chem. 1996, 61, 7955.
3. For general reviews on ring-closing metathesis: (a) Hoveyda, A. H.; Zhugralin,
A. R. Nature 2007, 450, 243; (b) Schrock, R. R. Angew. Chem., Int. Ed. 2006, 45,
3748; (c) Grubbs, R. H. Angew. Chem., Int. Ed. 2006, 45, 3760; (d) Grubbs, R. H.
Handbook of Metathesis; Wiley: Weinheim, 2003 and references cited therein.
4. For application of ring-closing metathesis in the synthesis of nitrogen-con-
taining heterocycles, see: Deiters, A.; Martin, S. F. Chem. Rev. 2004, 104, 2199.
5. For an earlier synthesis of 1 based on C3–C4 disconnection using RCM, see:
Yamashita, D. S.; Marquis, R. W.; Xie, R.; Nidamarthy, S. D.; Oh, H.-J.; Jeong, J. U.;
Erhard, K. F.; Ward, K. W.; Roethke, T. J.; Smith, B. R.; Cheng, H.-Y.; Geng, X.; Lin,
F.; Offen, P. H.; Wang, B.; Nevins, N.; Head, M. S.; Haltiwanger, R. C.; Narducci
Sarjeant, A. A.; Liable-Sands, L. M.; Zhao, B.; Smith, W. W.; Janson, C. A.; Gao, E.;
Tomaszek, T.; McQueney, M.; James, I. E.; Gress, C. J.; Zembryki, D. L.; Lark, M.
W.; Veber, D. F. J. Med. Chem. 2006, 49, 1597.
6. When C3 is in the oxidation of
a ketone, it readily loses the pyridine-
sulfonamide group upon treatment of strong bases via deprotonation of C2
followed by elimination of SO2 and pyridine. It is therefore necessary to
maintain a hydroxyl oxidation state for C3 for the early steps in the retro-
synthetic analysis.
36. (a) Ogawa, K.; Ohta, S.; Okamoto, M. Synthesis 1987, 281; (b) Nishitani, K.;
Harada, Y.; Nakamura, Y.; Yokoo, K.; Ymamakawa, K. Tetrahedron Lett. 1994, 35,
7809; (c) Dai, L.; Lou, B.; Zhang, Y. J. Am. Chem. Soc. 1988, 110, 5195.
37. (a) Albanese, D.; Landini, D.; Lupi, V.; Penso, M. Adv. Synth. Catal. 2002, 3, 299;
(b) Albanese, D.; Landini, D.; Penso, M.; Petricci, S. Tetrahedron 1999, 55, 6387;
(c) Albanese, D.; Landini, D.; Penso, M. Tetrahedron 1997, 53, 4787; (d) Albanese,
D.; Landini, D.; Penso, M. Chem. Commun. 1999, 2095.
38. A solvent study indicated that the reaction occurred in various alcoholic solvents
but not in aprotic solvents such as acetonitrile, THF, ethyl acetate or toluene.
39. Epoxide ring opening by toluenesulfonamide is known to be catalyzed by
pyridine, see: (a) Baker, B. R.; Querry, M. V.; Kadish, A. F. J. Org. Chem. 1950, 15,
402; (b) Magerlein, B. J.; Weisblat, D. I. J. Am. Chem. Soc. 1954, 76, 1702.
40. (a) Schwesinger, R.; Schlemper, H. Angew. Chem., Int. Ed. Engl. 1987, 26, 1167; (b)
Schwesinger, R.; Willaredt, J.; Schlemper, H.; Keller, M.; Schmitt, D.; Fritz, H.
Chem. Ber. 1994, 127, 2435 and references cited therein.
41. For a detailed account of the effect of impurities as well as development of this
RCM reaction on large scale, see: Wang, H.; Goodman, S. N.; Dai, Q.; Stockdale,
G. W.; Clark, W. M. Org. Process Res. Dev. 2008, 12, 226.
42. A detailed account of development of this RCM reaction using alternative
catalysts will be published separately.
43. Fuson, R. C.; Kneisley, J. W.; Kaiser, E. W. Org. Synth. 1944, 24, 33.
44. For DMAP-induced racemization during peptide coupling, see: Gamet, J. P.;
Jacquier, R.; Verducci, J. Tetrahedron 1984, 40, 1995.
7. (a) Broughton, D. B.; Clarence, G. G. U.S. Patent 2,985,589, 1961. (b) Balannec, B.;
Hotier, G. In Preparative and Production Scale Chromatography; Ganetsos, G.,
Barker, P. E., Eds.; Marcel Dekker: New York, NY, 1993; p 301; (c) Nicoud, R. M.;
Majors, R. E. LC-GC 2000, 18, 680.
8. Corey, E. J.; Posner, G. H.; Atkinson, R. F.; Wingard, A. K.; Halloran, D. J.; Radzik,
D. M.; Nash, J. J. J. Org. Chem. 1989, 54, 389.
9. Marquis, R. W.; Ru, Y.; LoCastro, S. M.; Zeng, J.; Yamashita, D. S.; Oh, H.-J.; Er-
hard, K. F.; Davis, L. D.; Tomaszek, T. A.; Tew, D.; Salyers, K.; Proksch, J.; Ward,
K.; Smith, B.; Levy, M.; Cummings, M. D.; Haltiwanger, R. C.; Trescher, G.; Wang,
B.; Hemling, M. E.; Quinn, C. J.; Cheng, H.-Y.; Lin, F.; Smith, W. W.; Janson, C. A.;
Zhao, B.; McQueney, M. S.; D’Alessio, K.; Lee, C.-P.; Marzulli, A.; Dodds, R. A.;
Blake, S.; Hwang, S.-M.; James, I. E.; Gress, C. J.; Bradley, B. R.; Lark, M. W.;
Gowen, M.; Veber, D. F. J. Med. Chem. 2001, 44, 1380.
10. Resek, J. E.; Beak, P. Tetrahedron Lett. 1993, 34, 3043.
11. Trout, R. E. L.; Marquis, R. W. Tetrahedron Lett. 2005, 46, 2799.
12. For aldol addition reactions of chiral crotonate imides, see: Evans, D. A.; Sjog-
ren, E. B.; Bartroli, J.; Dow, R. L. Tetrahedron Lett. 1986, 27, 4957.
13. (a) Crimmins, M. T.; King, B. W.; Tabet, E. A.; Chaudhary, K. J. Org. Chem. 2001,
66, 894; (b) Crimmins, M. T.; She, J. Synlett 2004, 1371.
14. In the case of the less hindered des-C7-methyl substrate, the RCM was effected
by 1st generation Grubbs catalyst (catalyst loading unspecified), see Ref. 11.
15. Garber, S. B.; Kingsbury, J. S.; Gray, B. L.; Hoveyda, A. H. J. Am. Chem. Soc. 2000,
122, 8168.
45. Konig, W.; Geiger, R. Chem. Ber. 1970, 103, 788.
46. (a) Konig, W.; Geiger, R. Chem. Ber. 1970, 103, 2024; (b) Konig, W.; Geiger, R.
Chem. Ber. 1970, 103, 2034.
47. (a) A detailed account of the oxidation of 55 to 1 will be published separately.
(b) Oxidation of a secondary carbinol with the restrictions placed on process
chemistry (non-toxic, lowcost, solvent selection, etc.) is extremelychallenging. For
a general review of large-scale oxidation methods, see: Caron, S.; Dugger, R. W.;
Ruggeri, S. G.; Ragan, J. A.; Ripin, D. H. B. Chem. Rev. 2006, 106, 2943.
48. Albright, J. D.; Goldman, L. J. Am. Chem. Soc. 1965, 87, 4214.
16. (a) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H. Angew. Chem., Int. Ed.
Engl. 1995, 34, 2039; (b) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc.
1996, 118, 100; (c) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953; (d) Fu¨rstner, A.; Ackermann, L.; Gabor, B.; Goddard, R.; Lehmann, C. W.;
Mynott, R.; Stelzer, F.; Thiel, O. R. Chem.dEur. J. 2001, 7, 3236.
17. Fu¨rstner, A.; Guth, O.; Du¨ ffels, A.; Seidel, G.; Liebl, M.; Gabor, B.; Mynott, R.
Chem.dEur. J. 2001, 7, 4811.
18. (a) Fu¨rstner, A.; Langemann, K. J. Am. Chem. Soc. 1997, 119, 9130; (b) Yang, Q.;
Xiao, W.-J.; Yu, Z. Org. Lett. 2005, 7, 871.
49. Anelli, P. L.; Biffi, C.; Montanari, F.; Quici, S. J. Org. Chem. 1987, 52, 2559.
50. (a) Yee, N. K.; Farina, V.; Houpis, I. N.; Haddad, N.; Frutos, R. P.; Gallou, F.;
Wang, X.; Wei, X.; Simpson, R. D.; Feng, X.; Fuchs, V.; Xu, Y.; Tan, J.; Zhang,
L.; Xu, J.; Smith-Keenan, L. L.; Vitous, J.; Ridges, M. D.; Spinelli, E. M.;
Johnson, M. J. Org. Chem. 2006, 71, 7133; (b) Nicola, T.; Brenner, M.;
Donsbach, K.; Kreye, P. Org. Process Res. Dev. 2005, 9, 513.
51. Safety concerns over the handling of aldehyde 13 restricts that the compound be
handled as a solution, and that the temperature during distillation remain ꢄ50 ꢁC.
52. Ho, G.-J.; Mathre, D. J. J. Org. Chem. 1995, 60, 2271.
19. (a) Maynard, H. D.; Grubbs, R. H. Tetrahedron Lett. 1999, 40, 4137; (b) Ferguson,
M. L.; O’Leary, D. J.; Grubbs, R. H. Org. Synth. 2003, 80, 85.
20. For a similar observation in cyclopentene formation, see: Crimmins, M. T.;
Tabet, E. A. J. Org. Chem. 2001, 66, 4012.
21. Evans, D. A.; Britton, T. C.; Ellman, J. A. Tetrahedron Lett. 1987, 28, 6141.
22. Shioiri, T.; Ninomiya, K.; Yamada, S. J. Am. Chem. Soc. 1972, 94, 6203.
23. For general reviews on the Curtius reaction, see: (a) Smith, P. A. S. Org. React.
1946, 3, 337; (b) Shioiri, T. In Comprehensive Organic Synthesis; Trost, B. M.,
Flemming, I., Eds.; Pergamon: Oxford, 1991; Vol. 6, p 795.