Synthesis and evaluation of cis-hexahydropyrrolo[3,2-b]pyrrol-3-one peptidomimetic inhibitors of CAC1 cysteinyl proteinases

Article abstract of DOI:10.1016/j.bmc.2004.10.060

A stereoselective synthesis of functionalised cis-hexahydropyrrolo[3,2-b] pyrrol-3-ones has been developed through Fmoc and Cbz-protected intermediates 5 and 6. Building blocks 5 and 6 were prepared via the intramolecular cyclisation of anti-epoxide 17. T

Full text of DOI:10.1016/j.bmc.2004.10.060

Bioorganic & Medicinal Chemistry 13 (2005) 609–625
Synthesis and evaluation of cis-hexahydropyrrolo[3,2-b]pyrrol-3-
one peptidomimetic inhibitors of CAC1 cysteinyl proteinases
Martin Quibell,* Alex Benn,ꢀ Nick Flinn, Tracy Monk, Manoj Ramjee, Peter Ray,ꢁ
Yikang Wang and John Watts
Amura Therapeutics Limited, Incenta House, Horizon Park, Barton Road, Comberton, Cambridge CB37AJ, UK
Received 6 August 2004; accepted 29 October 2004
Available online 21 November 2004
Abstract—A stereoselective synthesis of functionalised cis-hexahydropyrrolo[3,2-b]pyrrol-3-ones has been developed through Fmoc
and Cbz-protected intermediates 5 and 6. Building blocks 5 and 6 were prepared via the intramolecular cyclisation of anti-epoxide
17. The intramolecular reaction occurred exclusively through the anti-epoxide to provide the 5,5-cis-fused bicycle, whereas the syn-
epoxide, which theoretically would provide the 5,5-trans-fused bicycle, remained unchanged. These experimental observations are
consistent with a key design element that we have introduced within this novel bicyclic ketone scaffold. Our bicyclic design strategy
provides chiral stability to the bridgehead stereocentre that is situated a to the ketone because the cis-fused geometry is both ther-
modynamically and kinetically stable. Building blocks 5 and 6 have been utilised in both solid phase and solution phase syntheses of
peptidomimetics 22, 36–40, which exhibit potent in vitro inhibition against a range of CAC1 cysteinyl proteinases. Compound 22, a
potent and selective inhibitor of human cathepsin K exhibited good primary DMPK properties along with promising activity in an
in vitro cell-based human osteoclast assay of bone resorption.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
targets3 now established for indications such as osteo-
porosis, multiple sclerosis, arthritis, atherosclerosis and
cancer as well as parasitic proteinases4 identified to
target infections such as malaria5a,b and Chagas dis-
ease.5c,d Consequently, significant resource is currently
devoted within the pharmaceutical industry towards
the development of CAC1 proteinases inhibitors that
are suitable for human administration.3,6 The most
widely examined inhibitor series to date are based
around covalent interaction of the proteinase active site
thiol with peptidomimetic nitriles,7a,b cyclic ketones,7c,d
ketoamides7e,f and nonpeptidic cyanamides,7g 2-carbo-
nitrile-pyrrolopyrimidines7h and 2-cyanopyrimidines.7i
Additionally, a noncovalent series of cathepsin K inhib-
itors based around arylaminoethylamides has received
recent attention.8 A number of compounds from these
initial series have now advanced into the clinic as inhib-
itors of cathepsin K with osteoporosis as the main ther-
apeutic indication.9 The preliminary findings have
clearly validated the efficacy of inhibitors of cathepsin
K in a clinical setting and removed a significant portion
of the risk associated with first-in-class studies. These
advances should invigorate the search for new inhibitor
series, against both cathepsin K and other CAC1 pro-
teinases, that are suitable for the clinic. In this context,
we have recently reported the design, synthesis and
The papain-like CAC1 family of cysteinyl proteinases
form a large group of widely expressed enzymes that
perform a range of biological functions throughout the
animal, plant, viral, bacterial and protozoal kingdoms.1
Until recently, it was thought that the major role of
mammalian CAC1 proteinases was a general nonspecific
lysosomal degradation of proteins. However, our under-
standing of the biology of human cysteinyl proteinase
function has developed significantly in the last few
years.2 It has now clearly been established that the phys-
iological roles of many mammalian CAC1 proteinases
fulfil specific functions concerning antigen presentation,
extracellular matrix turnover and processing events.3 As
such, pharmaceutical interest has increased significantly
over the past five years, with viable mammalian drug
Keywords: CAC1 proteinase inhibition; Bicyclic ketones; Anti-
resorptive.
*
Corresponding author. Tel.: +44 (0)1223 264211; fax: +44 (0)1223
265662; e-mail: martin.quibell@incenta.co.uk
ꢀ Present address: Siena Biotech S.p.A., Via Fiorentina, 1, 53100,
Siena, Italy.
ꢁ Present address: Organon Research Scotland, Newhouse, Lanark-
shire, Scotland ML1 5SH, UK.
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.10.060

610
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
inhibition kinetics of three novel series of peptidomi-
metic inhibitors of CAC1 proteinases, each based
Boc
N2
6a
N
H
Boc
'Conditions'
N
N
N
Fmoc
3a
around
constrained
ketone-containing
scaffolds
Fmoc
(1!3).7d,10
O
O
5 (3aS, 6aR)
7 (2S, 3R)
R2
O
6a O
3a
R1
6a
3a
O
O
2
3
R1
Scheme 1. Synthesis of (3aS,6aR)-5 via an N–H insertion reaction of
(3R)-tert-butoxycarbonylamino-(2S)-(2-diazoacetyl)pyrrolidine-1-carb-
oxylic acid 9H-fluoren-9-ylmethyl ester.
O
N
R
N
R
N
3a
R
NH
6a
R
N
H
O
O
O
O
O
O
O
1 (2R or S, 3R)
4 (3aR, 6aS)
3 (3aS, 6aR)
2 (3aR, 6aR)
Given the poor yield from this initial route, we sought
alternative methods and our retrosynthetic analysis fo-
cussed upon the 1,3-arrangement of the a-ring nitrogen
and b-ring ketone. This indicated that the intramolecu-
lar ring opening of an epoxide with a tethered amine
could provide the desired 5,5-bicyclic scaffold 4 (see
Scheme 2).14
A key design feature in each of these scaffolds was the
introduction of a moiety that stabilised the otherwise
chirally labile position situated a to the ketone, which
is a physiochemical characteristic that has hindered the
pre-clinical development of a number of ketone-based
inhibitor series.11 The stabilisation was achieved
through alkylation as in series 1 and 2 or by exploiting
the kinetic and thermodynamic stability of a cis-fused
5,5-bicycle as in series 3. Peptidomimetic analogues built
around scaffold 3 were found to exhibit potent inhibi-
tion against a range of CAC1 proteinases, where the
inhibition kinetics were characterised by markedly in-
creased enzyme–inhibitor association rates (kon) when
compared to equivalent analogues from series 1 and 2.
Encouraged by these findings, we have now extended
our molecular modelling studies for series 3, which fea-
tured inhibitors that bound exclusively in the nonprime
binding pockets of the proteinase.10 Our new modelling
studies have concluded that compounds based upon the
cis-hexahydropyrrolo[3,2-b]pyrrol-3-one scaffold 4 may
span the active site to provide binding interactions with-
in both the prime and nonprime sites. Herein we report
the building block preparation, solid and solution phase
syntheses and inhibition kinetics for a series of bicyclic
We envisaged that the amine epoxide 17 would be a key
intermediate in which only the anti-epoxide would cyc-
lise because potential cyclisation of the syn-epoxide
would be disfavoured as it would lead to the thermody-
namically less stable trans-5,5-bicycle. Examination of
this synthetic route commenced from the commercially
available 3,4-dehydroproline 8. Methylene homologa-
tion of 8 was achieved with retention of configuration
by Arndt–Eistert reaction followed by Wolff rearrange-
ment, providing protected pyrrolidineacetic acid ester
10 (overall 62% from acid 8).15 Ester 10 was reduced
by treatment with lithium borohydride in anhydrous
THF to give alcohol 11 (67%). Standard activation of
alcohol 11 as the mesylate 12 was followed by nucleo-
philic displacement with sodium azide in dimethylform-
amide (DMF) to give the azide 13 (79%). Azide 13
provided a key intermediate that could lead to bicyclic
analogues through two routes, both of which were suc-
cessfully implemented towards the bicyclic compound
20 (see Scheme 2).
peptidomimetic cis-hexahydropyrrolo[3,2-b]pyrrol-3-
one based analogues 4 that exhibit potent inhibition of
a range of CAC1 cysteinyl proteinases.
In the first route, azide 13 was reduced to amine 14
following the general methods described by Mandville
et al. using triphenylphosphine and water in THF16
and the crude amine converted to benzyloxycarbonyl
(Cbz) protected analogue 15 (overall 94% from azide
13) under standard Schotten–Baumann conditions.
Epoxidation of 15 with m-chloroperoxybenzoic acid
provided an approximately 1:3 mixture of syn- and
anti-epoxides 16 (95%), which co-eluted on analytical
HPLC and gave a broad figure of eight spot by TLC
analysis. We surmised that removal then re-addition of
the Cbz protection within epoxides 16 would provide a
new mixture in which the original anti-epoxide had cyc-
lised to give the Cbz-protected bicycle 6 along with the
unreacted syn-epoxide from mixture 16 (compound
18). Thus, hydrogenation of Cbz protected 16 over
10% palladium on carbon (Pd–C) proceeded smoothly
over 2.5h providing intermediate amine epoxide 17,
wherein anti-17 cyclised to give bicyclic intermediate
19, which was followed by re-protection with Cbz under
standard Schotten–Baumann conditions. Analysis of the
re-protected crude reaction mixture indeed revealed the
presence of a major new compound, with the same
molecular weight as 16, that was less mobile on TLC
2. Results and discussion
2.1. Chemistry
2.1.1. Synthesis of Fmoc and Cbz protected scaffolds 5
and 6. We have previously described the syntheses of
mono- and bicyclic ketone scaffolds 1!3 through a lith-
ium chloride/acetic acid promoted insertion reaction of
protected aminoacid a-diazomethylketones as the final
ring-closing synthetic step7d,10 and were hopeful that a
similar scheme could be followed towards cis-hexa-
hydropyrrolo[3,2-b]pyrrol-3-ones 4. We were initially
encouraged by literature precedence describing the
intramolecular N–H insertion reactions of amino-a-di-
azomethylketones in the preparation of pyrrolidinones,
which were promoted by reagents such as trifluoroacetic
acid (TFA)12a or rhodium(II) catalysis.12b However, we
found that adaptation of these methods with the corre-
sponding diazomethylketone 7 gave complex reaction
mixtures from which the urethane protected bicycle
5 was isolated in low yield following treatment of
7
with rhodium(II) acetate dimer in toluene
(Scheme 1).13

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
611
i
O
ii
COCHN2
iv
iii
N
Boc
N
Boc
N
Boc
N
Boc
N
Boc
COOH
OMe
OH
OMs
9
8
10
11
12
v
O
O
vi, vii
ix
viii
N
Boc
N
Boc
N
Boc
N
Boc
N3
NHCbz
NH2
NHCbz
Boc
16
17
15
13
vii
viii, ix, x
Boc
N
Boc
N
Boc
N
O
6a
3a
6a
6a
6a
3a
N
xi
ix, x
vii
N
Fmoc
N
N
N
N
H
3a
3a
NHCbz
O
OH
OH
Cbz
Fmoc
Boc
OH
20
5
6
18
19
'Via anti-epoxide'
'Via syn-epoxide'
Scheme 2. Synthesis of (3aS,6aR)-5 and 6. Reagents and conditions: (i) (a) isobutyl chloroformate, NMM, CH2Cl2, ꢁ15ꢁC; (b) etheral CH2N2,
ꢁ15ꢁC to rt; (ii) MeOH, THF, CF3CO2Ag, NMM, 0ꢁC to rt in dark; (iii) LiBH4, MeOH, THF; (iv) methanesulfonyl chloride, triethylamine, DCM;
(v) sodium azide, DMF; (vi) Ph3P, H2O, THF; (vii) 1.05equiv Cbz–Cl, 2.1equiv Na2CO3, 1,4-dioxane, water; (viii) m-chloroperoxybenzoic acid,
DCM, rt, o/n; (ix) Pd–C/H2, ethanol; (x) 1.05equiv Fmoc–Cl, 2.1equiv Na2CO3, 1,4-dioxane, water; (xi) Dess–Martin periodinane, DCM.
and eluted earlier by HPLC analysis (present at 67.2%
by UV analysis) when compared to the original mixture
16. The new compound was purified by silica chroma-
tography and confirmed by full analysis as the desired
previously noted that urethane protected 5,5-bicyclic ke-
tones exhibit an unusually broad HPLC elution profile,
which we have attributed to the presence of slowly con-
verting but resolvable rotamers about the urethane 3ꢁ
amide bond.10 We observed a similar HPLC profile here
with ketones in series 4.
5,5-bicyclic alcohol 6 (overall 61% from mixture 16)
22
D
(½aꢀ ꢁ109.2 (c 0.544, CHCl3)). Also recovered from this
reaction were the syn-epoxide 18 (20.3%) and a second
new compound (4.3%) with a molecular weight in-
creased by 2Da, that was poorly mobile on TLC, and
tentatively assigned as an uncyclised alcohol produced
by hydrogenation of the epoxide within 16. With bicy-
clic alcohol 6 in hand, hydrogenation followed by re-
protection with 9-fluorenylmethoxycarbonyl chlorofor-
Building blocks 5 and 6 provided the opportunity to
examine synthetic routes towards peptidomimetic inhib-
itors of CAC1 proteinases through both solid and solu-
tion phase syntheses. We initially chose to examine
routes towards compound 22 (see Scheme 3 and 4), since
it offered a direct comparison to the corresponding cath-
epsin K inhibitor from general series 3, (3aS,6aR)-4-tert-
butyl-N-[3-methyl-1S-(3-oxo-hexahydrofuro[3,2-b]pyr-
role-4-carbonyl)butyl]benzamide 23, a compound that
binds exclusively in the nonprime binding pockets of
cathepsin K for which we have previously described
the inhibition kinetics in detail.10 Starting from alcohol
6, we envisaged two solution-based routes towards
inhibitor 22, the former involving oxidation of the alco-
hol then protection of resulting ketone as a dimethyl-
ketal, with ketal hydrolysis as the final synthetic step
(Scheme 3) and the latter involving introduction of the
ketone through oxidation of a bicyclic alcohol as the
final synthetic step (Scheme 4).
mate (Fmoc–Cl) provided intermediate 20 (95%)
22
D
(½aꢀ ꢁ104.0 (c 0.2, CHCl3)). Having confirmed the fea-
sibility of this route, our repeat syntheses of compound
20 proceeded through direct Fmoc protection of inter-
mediate bicyclic amine 19 (along with isolation of the
corresponding Fmoc analogue of syn-epoxide 18).
In the second potentially shorter route, azide 13 was
treated with m-chloroperoxybenzoic acid to provide an
unquantified mixture of syn- and anti-epoxides 21. We
found that azido epoxide 21 was unstable, so the crude
mixture was immediately hydrogenated over 10% Pd–
C, essentially giving a mixture of amine epoxides 17
and 19. Treatment of this crude amine mixture with
Fmoc–Cl under standard Schotten–Baumann condi-
tions gave our second route to intermediate 20 (overall
27% from azide 13). In practice, we have generally used
the slightly longer first route for repeat syntheses since
the overall yield of 20 is higher (overall 51% from azide
13 via Cbz-protected 6), the synthetic steps are repro-
ducible on a multigram scale and epoxide 16 provides
a stable advanced intermediate. As the final step, alcohol
20 was smoothly oxidised by treatment with Dess–Mar-
Our initial target inhibitor 22 contained the N-benzoyl
(Bz) group as the potential prime-side binding substitu-
ent, whilst building block 6 contained the N-tert-butoxy-
carbonyl (Boc) group. Therefore, both of the syntheses
shown in Schemes 3 and 4 commenced with a common
transformation through acidolytic removal of the Boc
group with HCl followed by acylation with benzoic
anhydride to give intermediate 24 in essentially quanti-
tative yield. Then following Scheme 3, alcohol 24 was
oxidised by treatment with Dess–Martin periodinane
to ketone 25 (78%). Since the objective of this route
tin periodinane to provide target bicyclic ketone 5 (97%)
22
D
as a white solid (½aꢀ ꢁ140.0 (c 0.6, CHCl3)). We have

612
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
Bz
Boc
Bz
Bz
Bz
3a
6a
6a
N
3a
N
3a
6a
6a
N
N
i, ii
N
iii
iv
v
N
N
N
6a
MeO
26
3a
N
N
H
3a
Cbz
O
OH
Cbz
OMe
Cbz
Cbz
OH
OMe
MeO
24
25
6
27
vi ,v
3a
Bz
N
Bz
Bz
3a
6a
N
vii
3a
6a
N
viii
N
O
6a
O
N
RHN
N
OMe
MeO
N
H
N
H
OMe
O
MeO
O
O
O
28. R = Cbz
29. R = H
30
22
Scheme 3. Solution synthesis of (3aR,6aS)-22 via ketal. Reagents and conditions: (i) 4N HCl in 1,4-dioxane, 30min, rt; (ii) benzoic anhydride,
NMM, DMF, rt, 1h; (iii) Dess–Martin periodinane, DCM; (iv) trimethylorthoformate, anhydrous MeOH, cat. p-TsOH, under Ar, 65ꢁC; (v) Pd–C/
H2, ethanol, methanol; (vi) 1equiv Cbz-Leu-F, DMF, rt; (vii) 1equiv 4-tert-butylbenzoic acid, HBTU, HOBt, NMM, DMF, rt; (viii) 95% TFA/5%
water, rt.
Bz
Bz
Bz
ketone 22 from ketal 30, a carboxylic acid side product
was formed by hydrolysis of the 3ꢁ amide bond between
the leucinyl carbonyl and the bicyclic scaffold. The same
side product was also observed during the acidolytic re-
lease of compounds from the solid phase (see later) and
appears to be particularly prevalent for compound 22 in
comparison to the other analogues we have prepared
(see later). However, treatment of ketal 30 with TFA
and water for 3.5h, followed by work-up and silica chro-
matography gave ketone 22 (24%) as a white solid.
3a
6a
3a
6a
6a
3a
N
N
N
i
ii
N
N
N
H
CbzHN
OH
Cbz
OH
OH
O
24
31
32
i
Bz
Bz
3a
3a
6a
N
N
iii
iv
O
22
N
6a
N
H2N
N
H
Next we examined the alternative synthesis of com-
pound 22 as shown in Scheme 4. Catalytic hydrogena-
tion of the Cbz group from alcohol 24 was complete
after 8.5h when performed in ethanol, giving amine
31. Acylation of amine 31 with Cbz-leucine-F in DMF
at ambient temperature for 75min gave alcohol 32
(68%) following purification. However, in direct con-
trast to aminoketal 27, we found that aminoalcohol 31
could also be quantitatively acylated with Cbz-leucine
succinyl ester in DMF at ambient temperature. Hydro-
genation of the Cbz group from alcohol 32 was slow
when performed in ethanol, requiring multiple additions
of catalyst and 24h for completion, giving amine 33 in
quantitative yield. Acylation of amine 33 with 4-tert-
OH
OH
O
33
O
34
Scheme 4. Solution synthesis of (3aR,6aS)-22 via alcohol. Reagents
and conditions: (i) Pd–C/H2, ethanol/methanol; (ii) 1equiv Cbz-Leu-F,
DMF, rt; (iii) 1equiv 4-tert-butylbenzoic acid, HBTU, HOBt, NMM,
DMF, rt; (iv) Dess–Martin periodinane, DCM, 2equiv TFA.
was to protect the ketone functionality for the remain-
der of the synthesis, we chose to convert 25 to the
dimethylketal 26 (77%), which was obtained by reaction
with trimethylorthoformate in methanol and p-toluene-
sulfonic acid catalysis. Catalytic hydrogenation of the
Cbz protection from ketal 26 proceeded to completion
within 90min when performed in methanol, providing
free amine 27 in essentially quantitative yield. Facile
acylation of amine 27 was achieved by reaction with
the pre-formed acyl fluoride of Cbz-leucine. The acyl-
ation reaction was essentially complete within 1h at
ambient temperature which, after work-up and silica
chromatography, gave intermediate 28 (90%). Catalytic
hydrogenation of the Cbz protection from ketal 28 pro-
ceeded to 95% completion within 6h when carefully
performed in methanol, providing free amine 29 in
essentially quantitative yield. Acylation of amine 29 with
4-tert-butylbenzoic acid through standard uronium acti-
vation chemistries17 gave ketal 30 (78%). The final step
in Scheme 3, acid hydrolysis of the dimethylketal to ke-
tone, proved difficult. We observed that under the
strong acidolytic cleavage conditions required to release
butylbenzoic acid as detailed earlier gave alcohol 34
22
D
nal oxidation of alcohol 34 was achieved by treatment
(68%) as a white solid (½aꢀ ꢁ84.5 (c 0.084, CHCl3)). Fi-
with Dess–Martin periodinane in dichloromethane
22
D
(DCM) to give ketone 22 (74%) (½aꢀ ꢁ82.0 (c 0.49,
CHCl3)).
2.2. Solid phase synthesis
Building block 5, containing the orthogonal Fmoc and
Boc protecting groups, was suitable for the rapid gener-
ation of CAC1 proteinase inhibitor analogues by solid
phase methods. Our synthetic strategy was based upon
reversible anchorage of the ketone functionality of
building block 5 via a hydrazide linker bond using the
general multipin techniques that we have previously des-
cribed in detail (Scheme 5).7d,10,19

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
613
O
Boc
O
R2
R1
N
N
iii, iv, iii, v,
N
i, ii
R3
N
H
5
O
N
Fmoc
vi, vii, viii, ix
O
N
O
N
H
N
H
22, 36-40
35
COR
a. R = OH
b. R = NH-SP
Scheme 5. Synthesis and use of supported linker construct 35b towards full length inhibitors 22, 36–40. Reagents and conditions: (i) trans-
4{[(hydrazinocarbonyl) amino]methyl}cyclohexanecarboxylic acid trifluoroacetate,18 EtOH, H2O, NaOAc, reflux; (ii) 3equiv 35a, 3equiv HBTU,
3equiv HOBt, 6equiv NMM, H2N-SOLID PHASE, DMF; (iii) 20% piperidine/DMF (v/v), 30min; (iv) 20equiv Fmoc–NHCHR2–COOH, 20equiv
HBTU, 20equiv HOBt, 40equiv NMM, DMF, rt, o/n, then repeat with fresh reagents for 6h; (v) 10equiv R3–COOH, 10equiv HBTU, 10equiv
HOBt, 20equiv NMM, DMF, rt o/n; (vi) TFA/DCM (35:65, v/v), 30min; (vii) NMM/DMF (2:98, v/v), 10min; (viii) 20equiv benzoic anhydride,
10equiv NMM, DMF, o/n; (ix) TFA/H2O, (95:5, v/v), 2h.
Final compounds 22 and 36–40 were prepared from lin-
ker-construct 35b by initially building the nonprime
binding elements (R3CONHCHR2CO) through a series
of sequential washing and coupling reaction steps
involving removal of Fmoc, double coupling of an acti-
vated Fmoc–NHCHR2–COOH to the secondary amine
of the bicyclic scaffold, removal of Fmoc and coupling
of an activated R3–COOH. Subsequent treatment with
35% TFA in DCM gave quantitative removal of the
Boc group, exploiting the greater acid lability of Boc
in comparison to that of the hydrazone linker.20 Follow-
ing neutralisation of the multipin-bound TFA salt,
treatment with benzoic anhydride in DMF provided
quantitative acylation of the bicyclic secondary amine.
Finally, cleavage of the hydrazone linker was achieved
with 95% TFA/H2O, followed by semi-preparative
HPLC purification to provide inhibitors 22, 36–40. In
general the quality of crude inhibitors from this solid
phase scheme was excellent as judged by analytical
HPLC at >90%. In each crude sample we observed a
side product consisting of the carboxylic acid R3CON-
HCHR2–COOH that derives from cleavage of the terti-
ary amide bond between the P2 aminoacid carbonyl and
the bicyclic scaffold. Within crude samples 36–40, this
carboxylic acid contaminant was present at 1–10% and
for crude sample 22 was present at 19%.21
that we have described previously.10 For example, ana-
logue 22 was a potent (Ksis ¼ 10:1nM) and selective
(>300-fold vs S and L) inhibitor of cathepsin K, which
represented a 9-fold increase in activity when compared
to the equivalent bicycle from series 3 (compound 23).10
Additionally, analogue 37 was a potent (Ksis ¼ 29:3nM)
and selective (>26-fold vs K and 9-fold vs L) inhibitor of
bovine cathepsin S with significantly improved potency
compared to that of the parent monocycle standard
(Ksis ¼ 220nM).22 Analogue 38 was
a
potent
(Ksis ¼ 78:3nM) and selective (>50-fold vs K and S)
inhibitor of human cathepsin L with equivalent potency
to that of the parent monocycle standard
(Ksis ¼ 70nM).22 Analogue 39 was a potent mixed inhib-
itor of cathepsins K (Ksis ¼ 112nM), L (Kiss ¼ 81:7nM)
and the parasitic proteinase CPB (Ksis ¼ 64:8nM), whilst
analogue 40 was a potent (Ksis ¼ 300nM) inhibitor of
cruzain. The across-the-board increase in potency could
be due to the introduction of a binding group within
general series 4 that enables inhibitors to bind in both
nonprime and prime-side binding pockets of the protein-
ases. However, the explanation is likely to be more com-
plex and additionally involve factors such as ring strain
and stereoelectronic effects of the ꢂoxygenꢃ to ꢂacyl-nitro-
genꢃ substitution.23a–c
In our previous studies concerning the inhibition kinet-
ics of analogues from series 310, we investigated the indi-
vidual association (kon) and dissociation (koff) rate
constants for bicyclic ketone inhibitors of human cath-
epsin K. We have repeated these studies for the potent
cathepsin K analogues from series 4, compounds 22
and 36, along with corresponding analogues from series
310, compounds 23 and 42 and the monocyclic ketone
analogues10, compounds 41 and 43 (Table 2). All of
these are potent inhibitors of human cathepsin K with
steady state inhibition constants ranging from 4.9 to
87.4nM. However, for the examples in Table 2, it is
clear that the bicyclic analogues (22, 23, 36 and 42) ex-
hibit significantly slower dissociation rates when com-
pared with the corresponding monocyclic analogues
(41 and 43). Conversely, in general the monocyclic ana-
logues exhibit faster association rates when compared
with the corresponding bicyclic analogues (compare
inhibitor 41 with inhibitors 22 and 23; inhibitor 43 with
42). The exception to these findings is inhibitor 36,
which combines a relatively fast association rate with a
2.3. Enzyme inhibition studies
Bicyclic inhibitors 22, 36–40 were screened against cath-
epsins K, L, S and B as well as the parasitic proteinases
cruzain and CPB.13 Preliminary steady-state inhibition
constants (Ksis) are shown in Table 1 (mean of n = 3
determinations). The substituents detailed in Table 1
were chosen to provide a direct comparison with our
previously detailed bicyclic inhibitors based around
scaffold 310 and are based upon binding groups that pro-
vide potent inhibitors when combined with the unsubsti-
tuted monocyclic scaffold described by GSK.22
Table 1 clearly shows that with the appropriate combi-
nation of substituents suitable for a particular protein-
ase, compounds based around general series 4 provide
potent and selective inhibitors of a range of CAC1 pro-
teinases. In each example, the cis-hexahydropyrrolo[3,2-
b]pyrrol-3-ones 4 detailed herein are more potent than
the corresponding tetrahydrofuro[3,2-b]pyrrol-3-ones 3

614
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
Table 1. Preliminary inhibitory activities (Ksis, nM) for 5,5-bicyclic inhibitors 22, 36–40 against CAC1 proteinases (mean of n = 3 determinations)
No
Structure
Cat. K
Cat. L
>3500
Cat. S
>4500
Cat. B
Cruz.
CPB
O
N
O
N
O
Ph
O
N
22
10.1 6.7
>10,000
173 86
691 350
N
H
O
O
O
Ph
N
36
37
38
39
4.9 0.7
766 78
>35,000
112 15
221 87
1242 270
29.3 1.8
>3500
>4000
>1500
>3000
>3500
224 11
1838 333
562 219
354 56
91 17
1337 182
>3000
N
H
O
N
S
N
O
O
N
Ph
O
257
5
N
N
H
O
O
HO
78.3 8.5
O
N
Ph
N
Br
N
H
O
O
O
O
N
Ph
81.7 31.0
1016 33
64.8 9.4
N
H2N
N
H
O
O
HO
O
40
O
1325 312
>3000
>12,000
>10,000
300 10
1565 184
N
Ph
N
N
H
O
O
Assay conditions are as detailed previously.13
relatively slow dissociation rate and is therefore a very
potent inhibitor of human cathepsin K with Ki[koff/kon]
at 1.6nM.
(Table 3). The basic physiochemical properties of inhib-
itors 22 and 36 are presented in Table 4. Analogues 22
and 36 have a moderate number of hydrogen bond do-
nors and acceptors,25 medium polar surface area,26
moderate molecular weight and moderate rotational
freedom,25 which are molecular properties that conform
to major predictors of desirable drug-like features. Fi-
nally, in vitro assessment of inhibitors 22 and 36 against
the major cytochrome P450 isozymes 1A2, 2B6, 2D6,
3A4 and 2C19 showed no significant inhibition at
10lM.27 Considering these findings and the results from
Tables 3 and 4, we concluded that analogues 22 and 36
exhibited a reasonably good range of properties suitable
for functional assessment of cathepsin K inhibition.
2.4. In vitro stability and physiochemical properties
We further evaluated the potent cathepsin K inhibitors
22 and 36 through a range of stability studies, physio-
chemical property determinations and in vitro second-
ary assays (Tables 3 and 4) and for comparison
included our previously reported data for the corre-
sponding analogues from series 3, compounds 23 and
42.10 These studies aimed to establish the basic physical
properties of inhibitors 22 and 36 and to determine their
suitability for functional assessment of cathepsin K inhi-
bition.24 Inhibitors 22 and 36 exhibited good stability at
acid and neutral pH and in human plasma with
t1/2 > 24h. At high pH, both 22 and 36 were reasonably
stable, however compound 22 was clearly degraded
more rapidly. Indeed, we have generally observed that
analogues from both bicyclic series 3 and 4 that contain
the 4-methylpiperazin-1-ylbenzamide group to be more
stable under virtually all conditions when compared to
the corresponding 4-tert-butylbenzamide analogues
2.5. Osteoclast resorption assay
Analogue 22 was evaluated in a human osteoclast
resorption assay,28 to determine the general ability of
cathepsin K inhibitors from series 4 to inhibit bone
resorption. Briefly, human osteoclasts were cultured on
bovine bone slices and allowed to differentiate and re-
sorb bone. The C-terminal telopeptides (CTX) of type
I collagen released into the culture media were then

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
615
Table 2. Association (kon) and dissociation rate constants (koff) for 5,5-bicyclic inhibitors 22 and 36 against human cathepsin K along with the
corresponding ꢂoxygenꢃ bicycles from series 3, compounds 23 and 42 and the corresponding ꢂmonocyclesꢃ, compounds 41 and 43 for comparison10
No
Structure
kon (Mꢁ1 sꢁ1) (· 105)
koff (sꢁ1) (· 10ꢁ3
)
Ki(koff/kon) (M) (· 10ꢁ9
)
Ksis (M) (· 10ꢁ9
)
O
O
N
Ph
N
22
1.4 1.2
1.7 0.6
12.1
10.1 6.7
N
H
O
O
O
O
2310
4110
36
4.9 3.9
12.0 2.0
46 0.4
5.3 4.6
>10
7.5 1.7
40.3 21.5
7.3 0.6
15.2
33.6
1.6
87.4 0.8
41.0 2.1
4.9 0.7
8.7 0.4
38.9 2.7
N
N
H
O
O
O
O
N
H
N
N
H
O
O
O
O
O
O
Ph
N
N
H
N
N
O
O
O
N
4210
5.5 0.7
10.4
N
H
O
O
O
N
N
N
N
O
H
N
4310
18.7 14.6
––
N
H
O
Table 3. Stability studies for bicyclic analogues 22, 23 and 36, 42
Table 4. Physiochemical properties for bicyclic analogues 22, 23 and
36, 42
PBS
(pH7.4
Acid
(ꢂpH1.5
Base
(pH10.5
Human
plasma
t1/2 (h)d
HLM
t1/2 (h)e
Sol.a LogD7.4 No
No
H-bond H-bond
donors acceptors bonds
No
Polar
b
)
)
)
t1/2 (h)a
t1/2 (h)b
t1/2 (h)c
rotatable surface
area
22
72.2
28.9
42.5
23.2
64.9
16.9
6.4
n.d.
38.6
27.9
24.4
4.5
3.4
2.8
5.0
7.5
2
˚
(A )
2310
36
110.2
36.5
30.2
16.3
22
M/H 2.84
1
1
1
1
4
4
6
6
7
6
7
6
86.8
75.7
93.3
82.2
4210
2310
36
M
H
H
1.36
1.68
0.04
For each analysis, aliquots at appropriate times were quantified by
HPLC–MS, using single ion monitoring and the ion intensity data
converted to a t1/2 for loss of parent analogue.
4210
a Aqueous solubility was assessed by measuring turbidity of solutions
of compound prepared in PBS (10mM; pH7.4) at 200, 100, 50 and
25lM, by light scattering at 650nm. Compounds were assigned as
having high (H, >100lM), medium (M, 50–100lM) or low (L, <50
lM) solubility.
a Compounds were incubated at 10lM in PBS (10mM; pH7.4) at
37ꢁC.
b Compounds were incubated at 10lM in 0.1M HCl/CH3CN (80:20)
at 37ꢁC.
c Compounds were incubated at 10lM in potassium phosphate
(10mM; pH10.5) at 37ꢁC.
b Partitioning of the compounds between n-octanol and PBS (10mM;
pH7.4) was assessed using a miniaturised shake-flask method,
employing HPLC–UV analysis.
d Compounds were incubated at 10lM in human plasma at 37ꢁC and
after protein precipitation with CH3CN, extracted aliquots were
analysed by HPLC–MS, using single ion monitoring.
e Compounds were incubated at 50lM with human liver microsomes
(0.5mg/mL of microsomal protein, final concentration) in potassium
phosphate (50mM; pH7.4) at 37ꢁC and the reaction was initiated
with NADPH (1mM final concentration). Quenching was achieved
by protein precipitation with CH3CN and the extracted aliquots were
analysed by HPLC, employing UV detection.
quantified as an index of bone resorption. Compound 22
and a positive control, E-64 (a potent inhibitor of cath-
epsin K), were added into the cell cultures after the dif-
ferentiation period and their effects on the resorbing
activity of mature osteoclasts were determined (Fig. 1).
As shown in Figure 1, compound 22 was a potent

616
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
ROMIL Ltd, UK at SpS or Hi-Dry grade unless other-
wise stated. 1H NMR and 13C NMR were obtained on a
1
Bruker DPX400 (400MHz H frequency and 100MHz
13C frequency; QXI probe) or Bruker Avance
500MHz (TXI probe with ATM) in the solvents indi-
cated. Chemical shifts are expressed in parts per million
(d) and are referenced to residual signals of the solvent.
Coupling constants (J) are expressed inHz. All analyti-
cal HPLC were obtained on Phenomenex Jupiter C4,
˚
5lm, 300A, 250 · 4.6mm, using mixtures of solvent A
(0.1% aq TFA) and solvent B (90% acetonitrile
(CH3CN)/10% solvent A) on automated Agilent systems
with 215 and/or 254nm UV detection. Unless otherwise
stated a gradient of 10–90% B in A over 25min at
1.5mL/min was performed for full analytical HPLC.
HPLC–MS analysis was performed on an Agilent 1100
series LC/MSD, using automated Agilent HPLC sys-
tems, with a gradient of 10–90% B in A over 10min
˚
on Phenomenex Luna C8, 5lm, 300A, 50 · 2.0mm at
0.6mL/min. Semi-preparative HPLC purification was
Figure 1. Inhibition profile for compound 22 as judged by human
osteoclast bone resorption assay. Bone resorption activity was mon-
itored by determining collagen fragments in culture medium (CTX)
using CrossLapsꢂ. Asterisks indicate values that are statistically
significantly different from baseline (with a p-value of less than 0.05).
One asterisk (*) indicates a p-value between 0.05 and 0.01 and two
asterisks (**) a p-value between 0.01 and 0.001.
˚
performed on Phenomenex Jupiter C4, 5lm, 300A,
250 · 10mm, using a gradient of 10–90% B in A over
25min at 4mL/min on automated Agilent systems with
215 and/or 254nm UV detection. Flash column purifica-
tion was performed on silica gel 60 (Merck 9385). Poly-
amide multipins (10lmol loadings, SPMDINOF, see
www.mimotopes.com) were used for the solid phase syn-
thesis. Biochemical protocols together with enzyme
assays were carried out as previously described.13 Sub-
strates utilising fluorescence resonance energy transfer
methodology (i.e., FRET-based substrates) were synthe-
sised using standard solid phase Fmoc chemistry meth-
ods,29 and employed Abz (2-aminobenzoyl) as the
fluorescence donor and 3-nitrotyrosine [Tyr(NO2)] as
the fluorescence quencher.30
inhibitor of bone resorption in this cell-based assay,
exhibiting 75% inhibition at a concentration of 100nM.
3. Conclusions
We have shown that 5,5-bicyclic series 4 provides a con-
formationally constrained scaffold towards the design of
potent and selective inhibitors of a range of therapeuti-
cally attractive mammalian and parasitic CAC1 cystein-
yl proteinase targets. The inhibitors may readily be
prepared either on the solid phase for rapid analogue
generation, or in solution for large scale-up applications,
following simple high yield syntheses of key building
blocks 5 and 6. Preliminary pharmacokinetic analysis
of the representative cathepsin K inhibitors 22 and 36,
showed a reasonably good range of physiochemical
and stability properties in both chemical and biological
media, providing a sound basis for further pre-clinical
evaluation. Compound 22, a potent in vitro inhibitor
of human cathepsin K with Ksis ¼ 10:1nM, was assessed
in a functional human osteoclast assay of cathepsin K
inhibition. In this cell-based assay, compound 22 exhib-
ited 75% inhibition of the release of C-terminal telopep-
tides (CTX) of type I collagen when tested at 100nM.
These preliminary data provide a promising platform
for the development of potent analogues from series 4
towards a cathepsin K inhibitor for the treatment of
osteoporosis.
4.1.1. (S)-2-(2-Diazoacetyl)-2,5-dihydropyrrole-1-carb-
oxylic acid tert-butyl ester (9). (S)-2,5-Dihydropyrrole-
1,2-dicarboxylic acid 1-tert-butyl ester
8 (1.066g,
5mmol) was dissolved with stirring in anhydrous
DCM (40mL). The reaction was flushed with nitrogen
and cooled to ꢁ15ꢁC. Isobutyl chloroformate
(0.713mL, 5.5mmol) in anhydrous DCM (5mL) and
N-methylmorpholine (1.099mL, 10mmol, NMM) in
anhydrous DCM (5mL) were added simultaneously in
1mL aliquots over 50min. The mixture was stirred for
2.5h at ꢁ15ꢁC then etheral diazomethane [ꢂ15mmol
generated from addition of diazald (4.7g) in diethyl
ether (75mL) onto sodium hydroxide (5.25g) in water
(7.5mL)/ethanol (15mL) at 60ꢁC] was added to the acti-
vated amino acid solution. The mixture was allowed to
warm to ambient temperature and stirred for 2.5h. A
few drops of acetic acid were cautiously added to the
mixture, followed by DCM (40mL). The etheral layers
were washed with aqueous saturated sodium hydrogen
carbonate solution (NaHCO3) (2 · 40mL), dried
(Na2SO4) and the solvents removed in vacuo to leave
a yellow residue (1.4g). Flash chromatography of the
residue over silica (35g) eluting with ethyl acetate
4. Experimental
4.1. General procedures
(EtOAc)–heptane 3:7 gave diazomethylketone
9
(1.024g, 86%). TLC (single spot, Rf = 0.45, EtOAc–hep-
tane 1:1), analytical HPLC tR = 11.54min; HPLC–MS
497.2 [2M+Na]+; 1H NMR (500MHz, CDCl3 at
Standard vacuum techniques were used in the handling
of air sensitive materials. Solvents were purchased from

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
617
300K): d 1.41–1.51 (m, C(CH3)3, 9H), 4.11–4.35 (m,
BocNCH2, 2H), 4.86–5.05 (m, BocNCH, 1H), 5.25–
5.50 (m, CHN2, 1H), 5.70–5.80 (m, olefinic CH, 1H)
and 5.88–6.03 (m, olefinic CH, 1H); 13C NMR
5.67–5.84 (m, 2 · olefinic CH, 2H); 13C NMR
(125MHz, CDCl3 at 300K): d 28.4 (C(CH3)3), 37.4
and 38.7 (CH2CH2OH), 53.45 and 53.6 (NCH2), 59.2
and 59.6 (OCH2), 61.2 and 61.9 (BocNCH), 79.9 and
80.1 (OC(CH3)3), 124.4 and 125.3 (olefinic CH), 130.3
and 131.1 (olefinic CH), 154.4 and 156.0 (NCO2).
(125MHz, CDCl3 at 300K):
d
28.3 and 28.4
(C(CH3)3), 51.8 and 52.3 (CHN2), 53.65 and 54.0
(BocNCH2), 71.5 and 72.3 (BocNCH), 80.6 and 80.9
(OC(CH3)3), 126.1 and 126.3 (olefinic CH), 128.35 and
128.5 (olefinic CH), 153.7 and 154.15 (NCO2), 192.7
and 193.4 (COCHN2).
4.1.4.
(R)-2-(2-Methanesulfonylethyl)-2,5-dihydropyr-
role-1-carboxylic acid tert-butyl ester (12). Triethylamine
(2.35mL, 16.9mmol) was added dropwise to a stirred
solution of compound 11 (2.33g, 10.9mmol) in DCM
(20mL) at 0ꢁC over 2min followed by methanesulfonyl
chloride (1.27mL, 16.4mmol) over 4min. The mixture
was stirred for 1h at 0ꢁC then washed with water
(170mL) and brine (170mL), dried (Na2SO4) and the
solvents removed in vacuo to leave the mesylate 12
(3.42g), which was used without further purification.
HPLC–MS 236.0 [M+2HꢁBu]+, 314.1 [M+Na]+, 605.1
[2M+Na]+.
4.1.2. (R)-2-Methoxycarbonylmethyl-2,5-dihydropyrrole-
1-carboxylic acid tert-butyl ester (10). Compound 9
(912mg, 3.85mmol) was dissolved in tetrahydrofuran
(14mL) and methanol (1.6mL) then cooled to 0ꢁC. A
solution of silver trifluoroacetate (94mg) in NMM
(1.06mL) was added, and the mixture allowed to warm
to ambient temperature over 6h in the dark. Methanol
(40mL) was added, followed by 10% aqueous citric acid
solution (100mL). The majority of the organic solvents
were removed in vacuo then the aqueous phase ex-
tracted with EtOAc (3 · 40mL). The combined organic
layers were washed with brine (40mL), dried (Na2SO4)
and evaporated in vacuo to give a residue (1.35g). Flash
chromatography of the residue over silica (200g) eluting
with EtOAc–hexane 3:17 gave methyl ester 10 as a col-
ourless oil (670mg, 72%). TLC (single spot, Rf = 0.25,
EtOAc–hexane 1:4), analytical HPLC tR = 15.03min;
HPLC–MS 505.3 [2M+Na]+; 1H NMR (500MHz,
CDCl3 at 300K): d 1.44–1.53 (m, C(CH3)3, 9H), 2.37–
2.55 (m, CH2CO2Me, 1H), 2.90–4.00 (m, CH2CO2Me,
1H), 3.63–3.70 (m, OCH3, 3H), 3.97–4.26 (m,
BocNCH2, 2H), 4.70–4.90 (m, BocNCH, 1H), 5.74–
5.89 (m, 2 · olefinic CH, 2H); 13C NMR (125MHz,
CDCl3 at 300K): d 28.2, 28.3 and 28.5 (C(CH3)3),
39.4 and 38.4 (CH2CO2Me), 51.5 and 51.6 (OCH3),
53.3 and 53.5 (BocNCH2), 60.7 and 60.9 (BocNCH),
79.6 and 80.0 (OC(CH3)3), 126.0 and 126.1 (olefinic
CH), 129.3 and 129.5 (olefinic CH), 153.9 (NCO2),
171.5 and 171.7 (CO2Me).
4.1.5.
(R)-2-(2-Azidoethyl)-2,5-dihydropyrrole-1-carb-
oxylic acid tert-butyl ester (13). Sodium azide (3.55g,
54.7mmol) was added to a stirred solution of mesylate
12 in DMF (45mL) under an atmosphere of argon.
The mixture was stirred at 60ꢁC for 1.5h then the
majority of solvents were removed by distillation in
vacuo and the residue partitioned between water
(200mL) and EtOAc (200mL). The EtOAc layer was
washed with brine (150mL), dried (Na2SO4) and the sol-
vents removed in vacuo to leave a residue (2.49g), which
was purified by flash chromatography over silica eluting
with EtOAc–heptane mixtures 0:100–10:90 to give azide
13 as a colourless oil (2.05g, 79%). TLC (single spot,
Rf = 0.45, EtOAc–hexane 3:7), analytical HPLC
tR = 15.91min; HPLC–MS 139.1 [M+2HꢁBoc]+, 183.1
[M+2HꢁBu]+, 499.2 [2M+Na]+. Anal. Calcd for
C11H18N4O2: C, 55.44; H, 7.61; N, 23.51, found C,
55.37; H, 7.59; N, 23.45. Exact mass calcd for
C11H18N4O2 (MNa+): 261.1327, found 261.1320
(ꢁ2.97ppm); 1H NMR (500MHz, CDCl3 at 300K):
d
1.40–1.50 (m, C(CH3)3, 9H), 1.90–2.10 (m,
4.1.3. (ꢁ)-(R)-2-(2-Hydroxyethyl)-2,5-dihydropyrrole-1-
carboxylic acid tert-butyl ester (11). Methanol
(0.27mL, 6.7mmol) was added dropwise to a stirred sus-
pension of lithium borohydride (146mg, 6.6mmol) in
tetrahydrofuran (3.5mL) under an atmosphere of argon
over 4min, followed by a solution of compound 10
(0.8g, 3.3mmol) in tetrahydrofuran (8mL) over
15min. The mixture was stirred for 1h then poured into
water (25mL). The product was extracted into DCM
(3 · 20mL), dried (Na2SO4), and the solvents removed
in vacuo. The residue was purified by flash chromato-
graphy over silica eluting with EtOAc–heptane mixtures
0:100–25:75 to give alcohol 11 as a colourless oil (0.48g,
67%). TLC (single spot, Rf = 0.35, EtOAc–hexane 1:1),
analytical HPLC tR = 12.16min; HPLC–MS 236.1
[M+Na]+, 449.3 [2M+Na]+. Exact mass calcd for
NCHCH22H), 3.17–3.33 (m, CH2N3, 2H), 3.96–4.27
(m, BocNCH2, 2H), 4.53–4.68 (m, BocNCH, 1H),
5.66–5.86 (m, 2 · olefinic CH, 2H); 13C NMR
(125MHz, CDCl3 at 300K):
d
28.3 and 28.5
(C(CH3)3), 32.5 and 33.0 (CH2CH2N3), 47.5 and 47.9
(CH2N3), 53.6 and 53.8 (BocNCH2), 62.0 and 62.3
(BocNCH), 79.55 and 79.9 (OC(CH3)3), 125.6 and
126.1 (olefinic CH), 128.9 and 129.4 (olefinic CH),
154.2 and 154.3 (NCO2).
4.1.6. (ꢁ)-(R)-2-(2-Benzyloxycarbonylaminoethyl)-2,5-
dihydropyrrole-1-carboxylic acid tert-butyl ester (15).
Water (1.9mL, 106mmol) was added to a stirred solu-
tion of azide 13 (2.65g, 11.17mmol) and triphenylphos-
phine (4.35g, 16.6mmol) in tetrahydrofuran (330mL)
under an atmosphere of argon. The solution was stirred
at 45ꢁC for 7.5h then at ambient temperature for 14h. A
5.0mL aliquot was removed for analysis, then the
remainder of the solution was concentrated in vacuo
to obtain amine 14 as an oily residue. The residue was
dissolved in 1,4-dioxane (35mL) with stirring, ice-cooled
and a solution of sodium carbonate (2.45g, 23.1mmol)
C11H19NO3 (MNa+): 236.1257, found 236.1264
22
(d + 2.95ppm); ½aꢀ ꢁ127 (c 1.0, CHCl3); 1H NMR
D
(500MHz, CDCl3 at 300K):
d 1.42–1.55 (br s,
C(CH3)3, 9H and NCHCH2, 1H), 1.84–1.95 (m,
NCHCH2, 1H), 3.60–3.72 (m, CH2OH, 2H), 3.93–4.28
(m, BocNCH2, 2H), 4.53–4.78 (m, BocNCH, 1H),

618
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
in water (35mL) was added. Benzyl chloroformate
(2.18g, 1.824mL, 12.8mmol) in 1,4-dioxane (10mL)
was then added dropwise over 30min and the mixture
stirred for an additional 30min before adding water
(250mL). The aqueous phase was extracted with DCM
(2 · 250mL) and the combined organic layers were dried
(Na2SO4), filtered and reduced in vacuo to leave a clear
mobile oil (10.2g). Flash chromatography over silica,
eluting with EtOAc–heptane mixtures gave compound
15 (3.58g, 94%) as a mobile colourless oil. TLC
(Rf = 0.30, EtOAc–heptane 1:1), analytical HPLC single
main peak, tR = 17.39min; HPLC–MS 247.1 [M+2Hꢁ
66.7 (OCH2Ph), 80.3 and 80.7 (OC(CH3)3), 128.0,
128.1, 128.2, 128.4, 128.5 (5 · aromatic CH), 136.7
(OCH2C), 155.1, 155.9, 156.3 and 156.6 (NHCO2 and
NCO2).
4.1.8. 2R-(2-Aminoethyl)-6-oxa-3-azabicyclo[3.1.0]hex-
ane-3-carboxylic acid tert-butyl ester (17) and
(3S,3aS,6aR)-3-hydroxyhexahydropyrrolo[3,2-b]pyrrole-
1-carboxylic acid tert-butyl ester (19). Epoxide mixture
16 (3.57g, 9.86mmol) was dissolved in ethanol
(60mL), cooled to 0ꢁC and 10% Pd–C (0.40g) added.
The mixture was stirred, then evacuated and flushed
with hydrogen. The mixture was allowed to warm to
ambient temperature and after 2.5h filtered through
Celite. The filter cake was washed with ethanol
(3 · 60mL) and the combined organic filtrates
reduced in vacuo to provide crude amines 17 and 19
(ꢂ2.4g). HPLC–MS 173.1 [M+2HꢁBu]+, 229.1
[M+H]+.
Boc]+, 291.1 [M+2HꢁBu]+, 347.1 [M+H]+, 369.1
22
[M+Na]+, 715.2 [2M+Na]+; ½aꢀ ꢁ74.9 (c 0.334,
D
CHCl3). Anal. Calcd for C19H26N2O4: C, 65.87; H,
7.56; N, 8.09, found C, 65.79; H, 7.53; N, 7.97. Exact
mass calcd for C19H26N2O4 (MNa+): 369.1790, found
1
369.1803 (+3.37ppm); H NMR (500MHz, CDCl3 at
300K): d 1.45 (br s, C(CH3)3, 9H), 1.60–1.95 (m,
BocNCHCH2, 2H), 3.00–3.44 (m, CH2NH, 2H), 3.90–
4.29 (m, BocNCH2, 2H), 4.45–4.81 (m, BocNCH, 1H),
5.01–5.16 (m, OCH2Ph, 2H), 5.50–5.83 (m, 2 · olefinic
CH, 2H) and 7.25–7.38 (m, C6H5 and NH, 6H); 13C
NMR (125MHz, CDCl3 at 300K): d 28.4 (C(CH3)3),
34.4, 34.6 (CH2CH2NH), 37.2, 37.6 (CH2NH), 53.6,
53.7 (BocNCH2), 61.7, 62.1 (BocNCH), 66.4, 66.6
(OCH2Ph), 79.6, 79.9 (OC(CH3)3), 125.2, 125.9, 127.0,
127.6, 127.9, 128.0, 128.4, 129.5, 130.2 (5 · aromatic
CH and 2 · olefinic CH), 154.3, 155.0, 156.2, 156.5
(NHCO2 and NCO2).
4.1.9.
(ꢁ)-(3S,3aS,6aR)-3-Hydroxyhexahydropyrrolo-
[3,2-b]pyrrole-1,4-dicarboxylic acid 4-benzyl ester 1-tert-
butyl ester (6). Crude amines 17 and 19 (ꢂ2.4g) were dis-
solved in 1,4-dioxane (30mL) with stirring, ice-cooled
and a solution of sodium carbonate (2.19g, 20.7mmol)
in water (25mL) was added. Benzyl chloroformate
(1.63mL, 11.4mmol) in 1,4-dioxane (15mL) was then
added dropwise over 30min and the mixture stirred
for a further 30min. The mixture was then reduced in
vacuo by approximately 2/3 volume to leave a mobile
pulp. Water (200mL) was added and the aqueous phase
extracted with DCM (2 · 100mL). The combined or-
ganic layers were dried (Na2SO4), filtered and reduced
in vacuo to leave a clear mobile oil (3.96g). Flash chro-
matography over silica, eluting with EtOAc–heptane
mixtures gave bicycle 6 (2.16g, 61%) as an opaque
gum. TLC (Rf = 0.15, EtOAc–heptane 1:1), analytical
HPLC single main peak, tR = 17.15min; HPLC–MS
263.1 [M+2HꢁBoc]+, 307.1 [M+2HꢁBu]+, 363.1
4.1.7. (2R)-2-(2-Benzyloxycarbonylaminoethyl)-6-oxa-3-
aza-bicyclo[3.1.0]hexane-3-carboxylic acid tert-butyl
ester (16). Compound 15 (3.57g, 10.3mmol) was dis-
solved in anhydrous DCM (60mL) with stirring and
meta-chloroperoxybenzoic acid (27.3g, 65% reagent,
103mmol) added. The mixture was stirred at ambient
temperature under argon for 16h. DCM (400mL) was
added and the organic phase washed with 10% aqueous
w/v solution of sodium hydroxide (2 · 400mL), then
dried (Na2SO4), filtered and reduced in vacuo to leave
a clear oil (ꢂ5g). Flash chromatography over silica,
eluting with EtOAc–heptane mixtures gave epoxides 16
[M+H]+,
385.1
[M+Na]+,
747.2
[2M+Na]+;
22
½aꢀ ꢁ109.2 (c 0.544, CHCl3). Anal. Calcd for
D
C19H26N2O5: C, 62.97; H, 7.23; N, 7.73, found C,
62.82; H, 7.39; N, 7.57. Exact mass calcd for
C19H26N2O5 (MNa+): 385.1739, found 385.1725
1
(3.57g, 95%) as
a
mobile colourless oil. TLC
(ꢁ2.15ppm); H NMR (400MHz, CD3CN at 348K): d
(Rf = 0.35 (minor) and 0.40 (major) (mixture of syn-
and anti-epoxides), EtOAc–heptane 2:1), analytical
HPLC single main peak, tR = 17.74min; HPLC–MS
263.1 [M+2HꢁBoc]+, 307.1 [M+2HꢁBu]+, 363.1
[M+H]+, 385.1 [M+Na]+, 747.2 [2M+Na]+. Anal. Calcd
for C19H26N2O5: C, 62.97; H, 7.23; N, 7.73, found C,
62.93; H, 7.22; N, 7.61. Exact mass calcd for
C19H26N2O5 (MNa+): 385.1739, found 385.1725
1.46 (s, C(CH3)3, 9H), 1.90–2.00 (m obscured by
NMR solvent peaks, BocNCHCH2, 1H), 2.14 (dd,
J = 6.15, 13.15Hz, BocNCHCH2, 1H), 3.07–3.20
(m, OH + CbzNCH2, 2H), 3.24 (dd, J = 3.80, 12.20Hz,
BocNCH2, 1H), 3.51 (d, J = 12.20Hz, BocNCH2,
1H), 3.68 (ddd, J = 1.70, 8.80, 10.60Hz, CbzNCH2,
1H), 4.10 (br d, J = 5.80Hz, CbzNCH, 1H), 4.27 (br
s, CH OH, 1H), 4.40–4.46 (m, BocNCH, 1H), 5.12
(d, J = 12.7Hz, OCH2Ph, 1H), 5.16 (d, J = 12.70Hz,
OCH2Ph, 1H) and 7.42–7.29 (aromatics, 5H);
1
(ꢁ3.82ppm); H NMR (500MHz, CDCl3 at 300K): d
1.32–1.62 (m, C(CH3)3 and CH2CH2NH, 10H), 1.67–
2.00 (m, CH2CH2NH, 1H), 2.90–4.21 (m, CH2NH,
BocNCHCH, BocNCH2CH, 7H), 4.70–5.17 (m,
OCH2Ph, 2H), 5.78–6.05 (m, NH, 1H) and 7.27–7.37
(aromatics, 5H); 13C NMR (125MHz, CDCl3 at
300K): d 28.1, 28.3, 28.35 and 28.4 (C(CH3)3), 30.8
and 31.2 (CH2CH2NH), 37.4 and 37.7 (CH2NH),
46.15 and 46.6 (BocNCH2), 53.9, 54.2, 54.9 and 55.8
(2 · epoxide CH), 58.1 and 58.2 (BocNCH), 66.5 and
13C NMR (100MHz, CDCl3 at 300K):
d 28.5
(C(CH3)3), 29.7, 30.4, 31.2, 31.9, 32.0 (BocNCHC2),
45.5, 45.7 (CbzNC2), 53.1, 53.4, 53.5 (BocNC2), 60.1,
61.2 (BocNCH), 67.2, 67.6, 68.2, 68.4, 69.0 (OCH2-
Ph + Cbz-NCH), 72.7, 73.3, 73.4 (CHOH), 79.9,
80.1 (OC(CH3)3), 127.9, 128.0, 128.2, 128.3, 128.5,
128.6, 136.3, 136.4 (aromatics), 154.1, 154.2, 155.2
(2 · NCO2).

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
619
4.1.10.
(ꢁ)-(3S,3aS,6aR)-3-Hydroxyhexahydropyr-
(185mg). Crude 21 was dissolved in ethanol (6.8mL)
and cooled to 0ꢁC. 10% Pd–C (84mg) was added to
the mixture and the atmosphere purged with hydrogen
gas. The mixture was stirred overnight under a hydrogen
atmosphere at ambient temperature, filtered over Celite
and the filter cake washed with excess EtOAc. The
filtrate was concentrated in vacuo, and the residue sus-
pended in a solution of sodium carbonate (193mg,
1.82mmol) in water (4mL). 1,4-Dioxane (2mL) was
added and the mixture cooled to 0ꢁC, then a solution
of 9-fluorenylmethyl chloroformate (198mg, 0.77mmol)
in 1,4-dioxane (2mL) added in small portions over
40min. The mixture was then allowed to warm to ambi-
ent temperature over 40min. Water (40mL) was added
and the product extracted into DCM (3 · 40mL). The
combined organic layers were dried (Na2SO4) and
evaporated in vacuo to afford a residue (335mg). Flash
chromatography of the residue over silica gel (35g) elut-
ing with EtOAc–heptane 1:4 followed by 1:1 gave bicy-
cle 20 (90mg, 27%). TLC (single spot, Rf = 0.25,
EtOAc–heptane 1:1), analytical HPLC tR = 18.35min;
HPLC–MS 451.2 [M+H]+, 473.2 [M+Na]+, 923.4
[2M+Na]+.
rolo[3,2-b]pyrrole-1,4-dicarboxylic acid 1-tert-butyl ester
4-(9H-fluoren-9-ylmethyl) ester (20). Bicycle 6 (0.54g,
1.5mmol) was dissolved in ethanol (10mL), cooled to
0ꢁC and 10% Pd–C (0.055g) added. The mixture was
stirred, then evacuated and flushed with hydrogen. The
mixture was warmed to ambient temperature and after
2.5h filtered through Celite. The filter cake was washed
with ethanol (3 · 10mL) and the combined filtrates re-
duced in vacuo to provide the crude amine (ꢂ0.36g).
HPLC–MS 173.1 [M+2HꢁBu]+, 229.1 [M+H]+. The
crude amine was dissolved in 1,4-dioxane (15mL) with
stirring, ice-cooled and a solution of sodium carbonate
(0.33g, 3.15mmol) in water (15mL) was added. 9-
Fluorenylmethyl chloroformate (0.463g, 1.79mmol) in
1,4-dioxane (10mL) was added dropwise over 30min
and the mixture stirred for a further 30min. Water
(200mL) was then added and the aqueous phase ex-
tracted with EtOAc (2 · 100mL). The combined organic
layers were dried (Na2SO4), filtered and reduced
in vacuo to leave a clear mobile oil (1.02g). Flash
chromatography over silica, eluting with EtOAc–hep-
tane mixtures gave bicycle 20 (0.64g, 95%) as a fine
white crystalline solid. TLC (Rf = 0.35, EtOAc–heptane
2:1), analytical HPLC single main peak, tR = 19.98 min;
4.1.12.
(ꢁ)-(3aS,6aR)-3-Oxo-hexahydropyrrolo[3,2-
HPLC–MS 395.1 [M+2HꢁBu]+, 451.1 [M+H]+, 473.1
b]pyrrole-1,4-dicarboxylic acid 1-tert-butyl ester 4-(9H-
fluoren-9-ylmethyl) ester (5). Bicyclic alcohol 20 (0.495g,
1.10mmol) was dissolved in anhydrous DCM (18mL)
with stirring under argon. Dess–Martin periodinane
(0.962g, 2.27mmol) was added and the mixture stirred
for 4h. The mixture was concentrated in vacuo and
the residue purified by flash chromatography over silica,
eluting with EtOAc–heptane mixtures to give bicyclic
ketone 5 (0.480g, 97%) as a white crystalline solid.
TLC (Rf = 0.40, EtOAc–heptane 1:1), analytical HPLC
single broad main peak, tR = 20.27–21.79min; HPLC–
22
D
[M+Na]+, 923.2 [2M+Na]+; ½aꢀ ꢁ104.0 (c 0.2, CHCl3).
Anal. Calcd for C26H30N2O5: C, 69.31; H, 6.71; N, 6.22,
found C, 69.11; H, 7.06; N, 5.84. Exact mass calcd for
C26H30N2O5 (MNa+): 473.2052, found 473.2053
1
(+0.06ppm); H NMR (400MHz, CD3CN at 348K): d
1.46 (s, C(CH3)3, 9H), 1.75–1.90 (m, BocNCHCH2,
1H), 2.05–2.13 (m, BocNCHCH2, 1H), 3.02 (m,
FmocNCH2, 1H), 3.08–3.20 (m, BocNCH2, 1H), 3.46
(m, BocNCH2, 1H), 3.46–3.60 (m, FmocNCH2, 1H),
3.90–4.15 (m, FmocNCH and CHOH, 2H), 4.28 (t,
J = 6.10Hz, FmocCH, 1H), 4.34–4.40 (m, BocNCH,
1H), 4.49 (d, J = 6.10Hz, FmocCH2, 2H), 7.31–7.45
(m, Fmoc aromatics, 4H), 7.65 (d, J = 7.35Hz, Fmoc
aromatics, 2H), 7.83 (d, J = 7.50Hz, Fmoc aromatics,
2H); 13C NMR (100MHz, CDCl3 at 300K): d 28.45
(C(CH3)3), 30.2, 31.2, 32.0 (BocNCHC2), 44.8, 45.3,
45.6 (FmocNC2), 47.3, 47.4 (FmocCH), 52.8, 53.1,
53.4, 53.5 (BocNCH2), 60.1, 60.8 (BocNCH), 65.9,
66.2, 67.3 (FmocCH2), 67.85, 68.4, 68.9 (FmocNCH),
72.5, 72.9, 73.3, 73.6 (CHOH), 79.95 (OC(CH3)3),
119.8, 120.0, 124.6, 124.9, 125.0, 127.0, 127.4, 127.8
(Fmoc CH aromatics), 141.3, 141.5, 143.7, 143.8, 144.1
(Fmoc quaternary aromatics), 154.0, 154.3, 155.0,
155.2 (2 · NCO2).
MS 393.1 [M+2HꢁBu]+, 449.1 [M+H]+, 471.1
22
D
[M+Na]+, 919.2 [2M+Na]+; ½aꢀ ꢁ140.0 (c 0.6, CHCl3).
Microanalysis indicated the presence of residual solvent
following drying in vacuo and re-drying at elevated tem-
perature lead to partial thermal decomposition. Exact
mass calcd for C26H28N2O5 (MNa+): 471.1896, found
1
471.1903 (+1.44ppm); H NMR (500MHz, CD3CN at
348K): d 1.51 (s, C(CH3)3, 9H), 1.90–2.00 (m obscured
by CD3CN, BocNCHCH2, 1H), 2.16–2.21 (m,
BocNCHCH2, 1H), 3.25–3.32 (dt, J = 7.20, 7.30Hz,
FmocNCH2, 1H), 3.52–3.58 (br, FmocNCH2, 1H),
3.72 (d, J = 19.00Hz, BocNCH2, 1H), 3.92 (d,
J = 19.10Hz, BocNCH2, 1H), 4.28 (t, J = 6.60Hz,
FmocCH, 1H), 4.42 (m, FmocCH2, 2H), 4.52 (d,
J = 8.35Hz, FmocNCH, 1H), 4.69 (m, BocNCH, 1H),
7.35 (t, J = 7.35Hz, Fmoc H-2 and H-7), 7.43 (t,
J = 7.45Hz, Fmoc H-3 and H-6), 7.68 (d, J = 7.45Hz,
Fmoc H-1 and H-8), 7.83 (d, J = 7.55Hz, Fmoc H-4
and H-5); 13C NMR (125MHz CDCl3 at 300K): d
28.36 (C(CH3)3), 30.50, 30.93, 31.20 (BocNCHC2),
45.68 (FmocNC2), 47.20 (FmocCH), 51.71, 52.22
(BocNC2), 57.58, 58.64 (BocNCH), 63.03, 63.57
(FmocNCH), 67.70, 68.06 (FmocCH2), 81.10
(OC(CH3)3), 119.94, 124.99, 125.15, 125.29, 127.05,
127.55, 127.71, 127.85 (Fmoc CH aromatics), 143.69,
143.92, 144.23 (Fmoc quaternary aromatics), 153.99,
154.74, 155.04 (2 · NCO2), 206.33, 206.59 (C@O).
4.1.11. Alternative preparation of (3S,3aS,6aR)-3-
hydroxyhexahydropyrrolo[3,2-b]pyrrole-1,4-dicarboxylic
acid 1-tert-butyl ester 4-(9H-fluoren-9-ylmethyl) ester
(20). meta-Chloroperoxybenzoic acid (864mg, 57–86%)
was added to
a solution of azide 13 (175mg,
0.735mmol) in anhydrous DCM (4mL). The mixture
was stirred at ambient temperature for 7h then satu-
rated aqueous NaHCO3 (40mL) and DCM (60mL)
were added. The phases were mixed and separated and
the organic phase washed with 10% aqueous sodium
hydroxide solution (40mL), dried (Na2SO4) and evapo-
rated in vacuo to afford crude azido epoxide 21

620
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
4.1.13. (3aR,6S,6aS)-4-Benzoyl-6-hydroxyhexahydropyr-
rolo[3,2-b]pyrrole-1-carboxylic acid benzyl ester (24). A
solution of HCl in 1,4-dioxane (4.0M, 11mL, 44mmol)
was added to compound 6 (450mg, 1.24mmol). The
solution was stirred for 65min whereupon a white sus-
pension formed. The solvents were removed in vacuo
and the residue azeotroped with diethyl ether
(3 · 15mL) and then DMF (10mL) and benzoic anhy-
dride (295mg, 1.31mmol) added. The solution was
placed under an atmosphere of argon then NMM
(0.29mL, 2.6mmol) was added to the solution dropwise
whilst stirring over 0.5min. The mixture was stirred for
1.75h then the solvents were removed in vacuo. The res-
idue was dissolved in EtOAc (100mL) then washed with
saturated aqueous NaHCO3 (100mL), pH3 HCl
(100mL) then brine (100mL). The organic layer was
dried (Na2SO4) and evaporated in vacuo to afford com-
pound 24 as a pale yellow gum (465mg), which was used
without further purification. Analytical HPLC
tR = 14.97min; HPLC–MS 367.1 [M+H]+, 733.1
[2M+H]+. Microanalysis indicated the presence of resid-
ual solvent following drying in vacuo and re-drying at
elevated temperature lead to partial thermal decomposi-
tion. Exact mass calcd for C21H22N2O4 (MNa+):
389.1477, found 389.1476 (ꢁ0.40ppm); 1H NMR
(500MHz, CDCl3 at 300K): d approximately 3:1 mix-
ture of rotamers, 2.10–2.21 (m, BzNCHCH2, 1H),
2.24–2.36 (m, BzNCHCH2, 1H), 3.20–3.35 (m,
CbzNCH2, 1H), 3.35–3.66 (m, BzNCH2, 2H), 3.74–
3.80 (m, CbzNCH2, 1H), 4.16–4.20 (m, CbzNCH, 1H),
4.38–4.42 (br s, CHOH, 1H), 4.94–5.04 (m, BzNCH,
1H), 5.08–5.22 (m, OCH2Ph, 2H), 7.30–7.52 (aromatic
CH, 10H); 13C NMR (125MHz, CDCl3 at 300K): d
31.02, 30.40 (BzNCHCH2), 45.73, 45.86 (CbzNCH2),
56.43, 56.74 (BzNCH2), 60.58, 61.49 (BzNCH), 66.68,
67.33 (OCH2Ph), 66.92, 67.76 (CbzNCH), 73.01, 73.86
(CHOH), 126.77, 127.48, 127.99, 128.23, 128.28,
128.45, 128.56, 128.87, 130.02, 130.38 and 134.53 (CH
aromatics), 136.16, 136.23, 136.33 (aromatic quater-
nary), 154.26, 154.97 (CbzC@O), 170.06, 171.19
(BzC@O).
130.48, 130.83 (CH aromatics), 135.07, 136.14 (quater-
nary aromatics), 154.54, 155.03 (CbzCO2), 170.58
(BzCO), 204–207 (br, C@O).
4.1.15. (3aR,6aS)-4-Benzoyl-6,6-dimethoxyhexahydro-
pyrrolo[3,2-b]pyrrole-1-carboxylic acid benzyl ester (26).
Ketone 25 (0.60g, 1.65mmol) was dissolved in methanol
(10mL) with stirring. Trimethylorthoformate (1.8mL,
16.5mmol) was added followed by para-toluenesulfonic
acid (40mg) and the mixture heated under argon at
65ꢁC for 16h. The mixture was reduced in vacuo to
leave a dark oil (0.8g). Flash chromatography over sil-
ica, eluting with EtOAc–heptane mixtures gave ketal
26 (0.52g, 77%) as a fine white crystalline solid. TLC
(Rf = 0.40, EtOAc–heptane 3:1), analytical HPLC single
main peak, tR = 18.22min; HPLC–MS 411.1 [M+H]+,
433.1 [M+Na]+, 843.1 [2M+Na]+. Exact mass calcd for
C23H26N2O5 (MNa+): 433.1739, found 433.1727
1
(ꢁ2.94ppm); H NMR (500MHz, CDCl3 at 300K): d
1.96–2.07, 2.15–2.22 (m, BzNCHCH2, 2H), 3.04–3.42
(m, 2 · OCH3, 6H), 3.25 (m, CbzNCH2, 1H), 3.4 (m,
BzNCH2, 1H), 3.58–3.67 (m, BzNCH2, 1H), 3.96–4.07
(m, CbzNCH2, 1H), 4.35–4.58 (m, CbzNCH, 1H),
4.98–5.26 (BzNCH + OCH2Ph, 3H), 7.28–7.49 (aromat-
ics, 10H); 13C NMR (125MHz, CDCl3 at 300K): d
32.27, 32.59 (BzNCHCH2), 46.74 (CbzNCH2), 49.36,
51.10, 51.59 (2 · OCH3), 54.59, 56.08 (BzNCH2),
60.77, 61.08 (BzNCH), 62.47 (CbzNCH), 67.28
(OCH2Ph), 106.76, 107.02 (C(OCH3)2), 126.84, 127.35,
127.90, 128.06, 128.39, 130.05, 130.38 (CH aromatics),
135.91, 136.48 (quaternary aromatics), 155.44
(CbzCO2), 169.54 (BzCO).
4.1.16. (3aS,6aR)-(3,3-Dimethoxyhexahydropyrrolo[3,2-
b]pyrrol-1-yl)phenylmethanone (27). Methanol (15mL)
was cautiously added dropwise to a stirred mixture of
ketal 26 (0.48g, 1.17mmol) and 10% Pd–C (100mg) at
0ꢁC under an atmosphere of argon over 10min. The
argon was then replaced by an atmosphere of hydrogen
and stirring continued at ambient temperature for
85min before replacing the hydrogen with argon and
adding ethanol (30mL). The mixture was filtered under
reduced pressure through Celite and the filter cake
washed with methanol (25mL) then ethanol (70mL).
Solvents were removed from the filtrate in vacuo to ob-
tain amine 27 as a colourless oil (340mg), which was
used without further purification. HPLC–MS 277.1
[M+H]+, 553.2 [2M+H]+.
4.1.14.
(3aR,6aS)-4-Benzoyl-6-oxo-hexahydropyrrolo-
[3,2-b]pyrrole-1-carboxylic acid benzyl ester (25). Crude
alcohol 24 (0.78g, 2.13mmol) was dissolved in DCM
(20mL) with stirring under argon. Dess–Martin period-
inane (1.804g, 4.26mmol) was added and the mixture
stirred for 16h. The mixture was concentrated in vacuo
and the residue purified by flash chromatography over
silica, eluting with EtOAc–heptane mixtures to give
bicyclic ketone 25 (0.61g, 78%) as an off-white gum.
TLC (Rf = 0.25, EtOAc–heptane 3:1), analytical HPLC
single main peak, tR = 14.65–16.30 min; HPLC–MS
365.1 [M+H]+, 383.1 [M+H+H2O]+. Microanalysis indi-
cated the presence of residual solvent following drying in
vacuo and re-drying at elevated temperature lead to
partial thermal decomposition. Exact mass calcd for
C21H20N2O4 (MNa+): 387.1321, found 387.1324
(+0.76ppm); 13C NMR (125MHz, CDCl3 at 300K): d
30.41, 30.89, 31.23 (BzNCHCH2), 45.75 (CbzNCH2),
54.55, 63.04 (BzNCH + CbzNCH), 57.91, 58.45, 58.99,
59.73 (BzNCH2), 67.60, 68.07 (OCH2Ph), 127.00,
127.38, 127.48, 127.98, 128.11, 128.48, 128.62, 128.74,
4.1.17. (ꢁ)-(3aR,6aS)-[(1S)-1-(4-Benzoyl-6,6-dimethoxy-
hexahydropyrrolo[3,2-b]pyrrole-1-carbonyl)-3-methylbut-
yl]carbamic acid benzyl ester (28). (i) Preparation of Cbz-
L-leucine fluoride. Cbz-L-leucine (1.115g, 4.2mmol) was
dissolved in DCM (50mL) with stirring under argon.
(Diethylamino)sulfur trifluoride (DAST, 792 lL,
6.0mmol) was added and the mixture stirred for
40min. Ice-cooled water (200mL) was added to the mix-
ture and the organic layer separated, dried (Na2SO4)
and reduced in vacuo to a mobile tan oil (1.14g). An
analytical sample, pre-treated with 10% pyridine in
methanol for 15min gave HPLC–MS 266.1 [M+H]+
(acid, <5%), 280.1 [M+H]+, 302.1 [M+Na]+, 581.1
[2M+Na]+ (methyl ester, ꢂ95%).

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
621
(ii) Crude amine 27 (ꢂ1.17mmol) was dissolved in anhy-
drous DMF (5mL) with stirring. Cbz-L-leucine fluoride
(0.33g, 1.23mmol) was added and the mixture stirred
under argon for 1h. The mixture was reduced in vacuo
to a semi-mobile dark oil (ꢂ1.0g). Flash chromatography
over silica, eluting with EtOAc–heptane mixtures gave
compound 28 (0.55g, 90%) as an off-white crystalline
solid. TLC (Rf = 0.35, EtOAc–heptane 3:1), analytical
HPLC single main peak, tR = 19.40min; HPLC–MS
0.994mmol) was added to a solution of 2-(1H-benzo-
triazole-1-yl)-1,1,3,3-tetramethyluronium hexafluoro-
phosphate (HBTU) (189mg, 0.497mmol), 1-hydro-
xybenzotriazole monohydrate (76mg, 0.497mmol) and
4-(tert-butyl)benzoic acid (88mg, 0.497mmol) in DMF
(12.5mL). The solution was stood for 5min then added
to amine 29 (0.497mmol). The mixture was stirred
at ambient temperature for 1h 50min then the solvents
were removed in vacuo. The residue was dissolved in
DCM (60mL) then washed with pH3 HCl (40mL), sat-
urated aqueous NaHCO3 (40mL) and brine (40mL),
then dried (Na2SO4) and the solvents removed in vacuo.
The residue was purified by flash chromatography over
silica eluting with EtOAc–heptane mixtures 0:100–50:50
to give compound 30 as a white solid (230mg), which
contained approximately 5% of compound 28. The lat-
ter compound was removed by dissolving the mixture
(220mg) in methanol (11mL) then adding to a stirred
suspension of 10% Pd–C (70mg) in ethanol (11mL)
under an atmosphere of argon at 0ꢁC. The argon was
then replaced by hydrogen and the mixture stirred at
ambient temperature for 80min, then water (11mL)
was added and the mixture filtered through Celite. The
filter cake was washed with ethanol (200mL) then the
filtrate concentrated in vacuo to give an oily solid 28
(203mg, 78%). TLC (single spot, Rf = 0.65, EtOAc–hep-
524.1 [M+H]+, 546.1 [M+Na]+, 1069.2 [2M+Na]+;
22
D
½aꢀ ꢁ114.5 (c 0.083, CHCl3). Anal. Calcd for C29-
H37N3O6: C, 66.52; H, 7.12; N, 8.02, found C, 66.26; H,
7.30; N, 7.86. Exact mass calcd for C29H37N3O6
(MNa+): 546.2580, found 546.2584 (+0.67ppm); 1H
NMR (500MHz, CDCl3 at 300K): d 0.92–1.04 (m, 2 ·
Leu dCH3, 6H), 1.45–1.55 (m, Leu bCH2, 2H), 1.73–
1.84 (m, Leu cCH, 1H), 1.92–1.99, 2.10–2.16, 2.22–2.30
(m, 4:6:10, BzNCHCH2, 2H), 2.94, 3.19, 3.23 and 3.40
(s, C(OCH3)2, 6H), 3.14–3.38 (m, 1 · BzNCH2 and
1 · CbzLeuNCH2, 2H), 3.60–3.68 (d, 4:6, J = 11.55,
10.90Hz, 1 · BzNCH2, 1H), 4.03–4.10, 4.11–4.18 (m,
4:6, 1 · CbzLeuNCH2, 1H), 4.33 (d, J = 6.30Hz, Cbz-
LeuNCH, 0.4H), 4.5–4.65 (m, Leu aCH, 0.6H), 4.82
(d, J = 6.45Hz, CbzLeuNCH, 0.6H), 4.87–4.93 (m,
Leu aCH, 0.4H), 5.0–5.14 (m, BzNCH + OCH2Ph,
3H), 5.42 (d, J = 8.35Hz, LeuNH, 0.6H), 5.57 (d,
J = 8.95Hz, LeuNH, 0.4H), 7.3–7.5 (aromatics, 10H);
13C NMR (125MHz, CDCl3 at 300K): d 21.99, 22.20,
22.67, 23.06, 23.67 (2 · Leu dCH3), 24.37, 24.57 (Leu
cCH), 31.58, 31.86, 33.26 (BzNCHCH2), 42.78 (Leu
bCH2), 44.12, 45.79, 47.05 (CbzLeuNCH2), 49.31, 49.99
(1 · OCH3), 51.18, 51.27, 51.30, 51.47 (1 · OCH3 + Leu
aCH), 55.55, 57.03 (BzNCH2), 59.69, 61.32 (BzNCH),
60.30, 61.04 (CbzLeuNCH), 66.39, 66.88 (OCH2Ph),
106.27, 107.11 (C(OCH3)2), 126.84, 127.32, 127.42,
127.87, 127.94, 128.09, 128.43, 128.48, 130.18, 130.43,
130.50 (CH aromatics), 135.68, 135.81, 136.28, 136.73
(quaternary aromatics), 155.58, 156.21 (CbzCO2),
169.56, 169.62 (BzCO), 172.35, 173.36 (Leu C@O).
tane 9:1), analytical HPLC tR = 21.41min; HPLC–MS
22
D
550.2 [M+H]+; ½aꢀ ꢁ80.3 (c 0.615, CHCl3). Anal. Calcd
for C32H43N3O5: C, 69.92; H, 7.88; N, 7.64, found C,
69.52; H, 8.12; N, 7.40. Exact mass calcd for
C32H43N3O5 (MH+): 550.3281, found 550.3284
(+0.55ppm).
4.1.20. (ꢁ)-(3aR,6aS)-N-[(1S)-1-(4-Benzoyl-6-oxo-hexa-
hydropyrrolo[3,2-b]pyrrole-1-carbonyl)-3-methylbutyl]-4-
tert-butylbenzamide (22). Ketal 30 (0.19g, 0.345mmol)
was dissolved in ice-cooled TFA/water (95:5 v/v,
10mL) with stirring. The ice bath was removed and
the mixture stirred at ambient temperature for 3.5h.
The mixture was then reduced in vacuo and evaporated
from diethyl ether (2 · 10mL) to give a semi-mobile tan
gum (0.3g). Flash chromatography over silica, eluting
with EtOAc–heptane mixtures gave ketone 22 (0.042g,
24%) as a white solid. TLC (Rf = 0.30, EtOAc–heptane
9:1), analytical HPLC single broad main peak,
tR = 19.43–21.37min; HPLC–MS 504.1 [M+H]+. Anal.
Calcd for C30H37N3O4Æ0.5TFA: C, 66.45; H, 6.75; N,
7.50, found C, 66.04; H, 7.19; N, 7.24. Exact mass calcd
for C30H37N3O4 (MNa+): 526.2682, found 526.2677
(ꢁ0.96ppm).
4.1.18. (2S,3aR,6aS)-2-Amino-1-(4-benzoyl-6,6-dimeth-
oxyhexahydropyrrolo[3,2-b]pyrrol-1-yl)-4-methylpentan-
1-one (29). Methanol (15mL) was cautiously added
dropwise to 10% Pd–C (75mg) at 0ꢁC under an atmos-
phere of argon over 10min whilst stirring followed by a
solution of compound 28 (0.52g, 0.99mmol) in metha-
nol (15mL). The argon was then replaced by an atmos-
phere of hydrogen and stirring continued at ambient
temperature for 5.5h. A suspension of 10% Pd–C
(15mg) in methanol (1mL) was added and stirring con-
tinued for 2.25h. The hydrogen was replaced by argon
then ethanol (100mL) was added before filtering the
mixture through Celite. The filter cake was washed with
ethanol (100mL) then the filtrate was concentrated in
vacuo to give amine 29 as a white solid, TLC (single
spot, Rf = 0.05, EtOAc–heptane 9:1), HPLC–MS 390.2
[M+H]+, 801.2 [2M+Na]+. The amine contained
approximately 5% of starting material 28 and was used
without further purification.
4.1.21. (3S,3aS,6aR)-(3-Hydroxyhexahydropyrrolo[3,2-
b]pyrrol-1-yl)phenylmethanone (31). Ethanol (5mL) was
added cautiously dropwise to 10% Pd–C (50mg) at
0ꢁC under an atmosphere of argon over 10min whilst
stirring followed by a solution of compound 24
(465mg, 1.27mmol) in ethanol (10mL). The argon was
then replaced by an atmosphere of hydrogen and stir-
ring continued at ambient temperature for 4.5h. The
hydrogen was then replaced by argon and 10% Pd–C
(20mg) was added at 0ꢁC. The argon was then replaced
with hydrogen and stirring was continued for 4h. The
hydrogen was replaced by argon then the mixture was
4.1.19. (ꢁ)-(3aR,6aS)-N-[(1S)-1-(4-Benzoyl-6,6-dimeth-
oxyhexahydropyrrolo[3,2-b]pyrrole-1-carbonyl)-3-methyl-
butyl]-4-tert-butylbenzamide (30). NMM (0.109mL,

622
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
filtered through Celite. The filter cake was washed with
ethanol (75mL) then the filtrate concentrated in vacuo
to obtain amine 31 as a colourless oil (309mg), which
was used without further purification. HPLC–MS
233.1 [M+H]+, 465.1 [2M+H]+.
which was used without further purification. HPLC–MS
346.2 [M+H]+, 713.3 [2M+Na]+.
4.1.24. (ꢁ)-(3aR,6aS)-N-[(1S)-1-((6S)-4-Benzoyl-6-hydr-
oxyhexahydropyrrolo[3,2-b]pyrrole-1-carbonyl)-3-meth-
ylbutyl]-4-tert-butylbenzamide (34). NMM (0.17mL,
1.55mmol) was added to a solution of HBTU (293mg,
4.1.22.
(ꢁ)-(3aR,6aS)-[(1S)-1-((6S)-4-Benzoyl-6-hydr-
oxyhexahydropyrrolo[3,2-b]pyrrole-1-carbonyl)-3-methyl-
butyl]carbamic acid benzyl ester (32). Cbz-leucine-F
(350mg, 1.31mmol) was dissolved in DMF (5mL) then
added to amine 31 (304mg, 1.24mmol) under an atmos-
phere of argon. The solution was stirred for 75min then
the solvents removed in vacuo. The residue was purified
by flash chromatography over silica eluting with
EtOAc–heptane 0:100–80:20 to give compound (32) as
0.77mmol),
1-hydroxybenzotriazole
monohydrate
(118mg, 0.77mmol) and 4-tert-butylbenzoic acid
(138mg, 0.77mmol) in DMF (7.5mL). The solution
was stood for 5min then added to amine 33 (0.77mmol).
The mixture was stirred at ambient temperature for 1h
and 5min then the solvents removed in vacuo (water
bath temperature <28ꢁC). The residue was dissolved
in DCM (75mL) then washed with pH3 HCl
(60mL), saturated aqueous NaHCO3 (60mL) and brine
(60mL), dried (Na2SO4) and the solvents removed
in vacuo. The residue (512mg) was purified by flash
chromatography over silica eluting with EtOAc–heptane
0:100–85:15 to give alcohol 34 as a white solid (263mg,
68%). TLC (single spot, Rf = 0.15, EtOAc–heptane 9:1),
a white solid (402mg, 68%). TLC (single spot,
Rf = 0.10, EtOAc–heptane 65:35), analytical HPLC
tR = 16.80min; HPLC–MS 480.2 [M+H]+, 981.3
22
[2M+Na]+; ½aꢀ ꢁ131.1 (c 0.479, CHCl3). Exact mass
D
calcd for C27H33N3O5 (MNa+): 502.2318, found
1
502.2311 (ꢁ1.44ppm); H NMR (500MHz, CDCl3 at
300K): d mixture of rotamers, tentative assignment of
proton 1.2–2.4 (m, 2 · Leu dCH3, Leu bCH2, Leu
cCH, BzNCHCH2, 11H), 3.3–4.0 (m, BzNCH2, CbzLeu-
NCH2, 4H), 4.2–5.0 (BzNCH, CbzLeuNCH, CHOH,
Leu aCH, 4H), 5.0–5.1 (OCH2Ph, 2H), 5.47 (d,
J = 8.30Hz, NH, 1H), 7.4–7.6 (aromatic, 10H); 13C
NMR (125MHz, CDCl3 at 300K): d 21.73, 21.89 and
23.22, 23.36 (2 · Leu dCH3), 24.59, 24.67 (Leu cCH),
31.86 (BzNCHCH2), 42.02, 42.22 (Leu bCH2), 46.52
(CbzLeuNCH2), 50.94, 51.02 (Leu aCH), 56.58
(BzNCH2), 59.72 (BzNCH), 67.00 (OCH2Ph), 67.98
(CbzLeuNCH), 75.25 (CHOH), 127.34, 128.02, 128.18,
128.28, 128.36, 128.52, 130.34 (aromatic CH), 136.09,
136.18 (aromatic quaternary), 156.18 (NHC@O),
170.08 (PhC@O), 172.32 (CH2NC@O).
analytical HPLC tR = 19.34min; HPLC–MS 506.2
22
D
[M+H]+; ½aꢀ ꢁ84.5 (c 0.084, CHCl3). Microanalysis
indicated the presence of residual solvent following dry-
ing in vacuo and re-drying at elevated temperature lead
to partial thermal decomposition. Exact mass calcd for
C30H39N3O4 (MNa+): 528.2838, found 528.2818
(ꢁ3.89ppm).
4.1.25. (3aR,6aS)-N-[(1S)-1-(4-Benzoyl-6-oxo-hexahydro-
pyrrolo[3,2-b]pyrrole-1-carbonyl)-3-methylbutyl]-4-tert-
butylbenzamide (22). A solution of alcohol 34 (174mg,
0.345mmol) in DCM (10mL) was added to Dess–
Martin periodinane (292mg, 0.689mmol) under an
atmosphere of argon whilst stirring over 2.5min. The
mixture was stirred for 3min then TFA (53lL,
0.689mmol) was added. The mixture was stirred for
14h then solvents removed in vacuo. The residue was
purified by flash chromatography over silica eluting with
EtOAc–heptane 0:100–55:45 to give ketone 22 as a white
solid (128mg, 74%). TLC (single spot, Rf = 0.30,
EtOAc–heptane 9:1), analytical HPLC broad peak
tR = 19.2–20.6min; HPLC–MS single broad UV peak,
4.1.23. (2S,3aR,6aS)-2-Amino-1-((6S)-4-benzoyl-6-hydr-
oxyhexahydropyrrolo[3,2-b]pyrrol-1-yl)-4-methylpentan-
1-one (33). Ethanol (15mL) was added cautiously to a
stirred mixture of compound 32 (370mg, 0.77mmol)
and 10% Pd–C (50 mg) at 0ꢁC under an atmosphere
of argon. The argon was replaced by an atmosphere of
hydrogen then stirring continued at ambient tempera-
ture for 1.75h. The hydrogen was replaced by argon
then the mixture was cooled to 0ꢁC before adding a fur-
ther portion of 10% Pd–C (20mg). The argon was
replaced by hydrogen then stirring continued at ambient
temperature for 5.25h. The hydrogen was replaced by
argon then the mixture was cooled to 0ꢁC before adding
a further portion of 10% Pd–C (20mg). The argon was
replaced by hydrogen then stirring continued at ambient
temperature for 14h. The hydrogen was replaced by
argon then the mixture was cooled to 0ꢁC before adding
a further portion of 10% Pd–C (10mg). The argon was
replaced by hydrogen then stirring continued at ambient
temperature for 2h. The hydrogen was replaced by
argon then the mixture was diluted with ethanol
(60mL) and filtered through Celite. The filter cake was
washed with ethanol (40mL) then the filtrate concen-
trated in vacuo. The residue was azeotroped with EtOAc
(35mL) to give amine 33 as an oily white solid (270mg),
22
1
504.1 [M+H]+; ½aꢀ ꢁ82.0 (c 0.49, CHCl3); H NMR
D
(500MHz, CDCl3 at 300K): d Tentative assignment of
peaks due to presence of rotamers 0.95 (d, J = 6.50Hz,
Leu dCH3, 3H), 1.01 (d, J = 6.20Hz, Leu dCH3, 3H),
1.31 (s, C(CH3)3, 9H), 1.58–1.81 (m, Leu bCH2 and
Leu cCH, 3H), 1.85–2.73 (m, BzNCHCH2, 2H), 3.55–
3.69 (m, BzNCHCH2CH2, 1H), 3.85–5.20 (m,
BzNCHCH2CH2, BzNCH2C(@O)CH and Leu aCH,
6H), 6.70–6.89 (m, NH, 1H), 7.40–7.52 (m, COC6H5
and
CHCHCC(CH3)3,
7H),
7.65–7.76
(m,
CHCHCC(CH3)3, 2H); 13C NMR (125MHz, CDCl3 at
300K): d 22.06, 23.28 (2 · Leu dCH3), 24.82 (Leu
cCH), 31.11, 31.14 (C(CH3)3), 31.67, 31.86 (BzNCHC2),
34.89, 34.92 (C(CH3)3), 42.43 (Leu bCH2), 46.10
(BzNCHCH2CH2), 48.93 (Leu aCH), 60.2 (BzNCH2),
61.0 (BzNCH or BzNCH2C(@O)CH), 68.2 (BzNCH
or BzNCH2C(@O)CH), 125.41, 125.49, 125.54, 126.91,
126.97, 127.11, 127.46, 128.33, 128.79, 130.84 (CH aro-
matics), 130.70, 131.16, 135.0 (quaternary aromatics),

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
623
155.35 (CC(CH3)3), 167.07, 170.67, 172.61 (3 · NC@O),
205.0 (C@O).
1113.6 [2M+Na]+. Exact mass calcd for C31H39N5O4
(MH+): 546.3075, found 546.3090 (+2.72ppm).
4.2. Solid phase chemistry
4.3.3. (3aR,6aS)-Thiophene-3-carboxylic acid [(1S)-2-(4-
benzoyl-6-oxo-hexahydropyrrolo[3,2-b]pyrrol-1-yl)-1-cyclo-
hexylmethyl-2-oxo-ethyl]amide (37). HPLC tR = 16.4–
17.2min (95%), HPLC–MS 494.2 [M+H]+, 1009.4
[2M+Na]+. Exact mass calcd for C27H31N3O4S
(MH+): 494.2108, found 494.2090 (ꢁ3.64ppm).
4.2.1. Preparation of (3aS,6aR)-3-oxo-hexahydropyr-
rolo[3,2-b]pyrrole-1,4-dicarboxylic acid 1-tert-butyl ester
4-(9H-fluoren-9-ylmethyl) ester––linker construct (35a).
Building block 5 (680mg, 1.52mmol, 1equiv) was
dissolved in a mixture of ethanol (42.5mL) and
water (6.1mL) containing sodium acetate trihydrate
(620mg, 4.55mmol, 3equiv). 4-{[(Hydrazinocarbon-
yl)amino]methyl}cyclohexane carboxylic acid trifluoro-
acetate (1.0g, 3.04mmol, 2equiv)18 was added and the
mixture was heated for 90min at 86ꢁC then allowed to
cool to ambient temperature and diluted with chloro-
form (350mL). The chloroform layer was washed with
dilute aqueous HCl (pH3, 2 · 350mL), brine (250mL),
dried (Na2SO4) and evaporated in vacuo to give 35a as
a colourless gum (1160mg). Analytical HPLC gave
two main peaks at tR = 19.62 and 21.30min (mixture
of E- and Z-isomers); HPLC–MS (main UV peaks with
tR = 5.46 and 6.24min 646.3 [M+H]+, 1291.6 [M+Na]+).
Crude 35 was used directly for construct loading.
4.3.4. (3aR,6aS)-N-[(1S)-2-(4-Benzoyl-6-oxo-hexahydro-
pyrrolo[3,2-b]pyrrol-1-yl)-1-(4-hydroxybenzyl)-2-oxo-eth-
yl]-3-bromobenzamide (38). HPLC tR = 13.21min (95%),
HPLC–MS 576.1/578.1 [M+H]+. Exact mass calcd for
C29H26N3O5Br (MNa+): 598.0948, found 598.0957
(+1.44ppm).
4.3.5. (3aR,6aS)-3-Aminomethyl-N-[(1S)-1-(4-benzoyl-6-
oxo-hexahydropyrrolo[3,2-b]pyrrole-1-carbonyl)-3-meth-
ylbutyl]benzamide (39). HPLC tR = 7.42min (95%);
HPLC–MS 477.2 [M+H]+, 495.2 [M+H+18]+, 975.3
[2M+Na]+. Exact mass calcd for C27H32N4O4 (MH+):
477.2496, found 477.2515 (+3.93ppm).
4.3.6. (3aR,6aS)-N-[(1S)-2-(4-Benzoyl-6-oxo-hexahydro-
pyrrolo[3,2-b]pyrrol-1-yl)-1-(4-hydroxybenzyl)-2-oxo-eth-
yl]-4-tert-butylbenzamide (40). HPLC tR = 16.0–17.1min
(95%); HPLC–MS 554.3 [M+H]+, 1129.5 [2M+Na]+.
Exact mass calcd for C33H35N3O5 (MH+): 554.2649,
found 554.2645 (ꢁ0.74ppm).
4.3. Solid phase protocols
Example inhibitors (22, 36–40) were prepared from con-
struct 35a by solid phase assembly techniques utilising
multipins and standard Fmoc chemistry protocols.19,29
Polyamide crowns with 10lmol loading (SPMDINOF)
were used for scale-up synthesis and purification of se-
lected examples. Construct 35a, 3equiv with respect to
solid phase surface loading, were coupled overnight
onto the 10lmol crowns using standard 3equiv HBTU,
3equiv HOBt and 6equiv NMM pre-activation (5min)
in a minimum volume of DMF to provide loaded con-
struct 35b. Construct 35b was utilised in standard
rounds of washing, Fmoc deprotection and coupling,
followed by acidolytic cleavage to give crude inhibitors
22, 36–40.21 Examples were purified by semi-preparative
HPLC (see general methods) and appropriate fractions
combined and lyophilised into pre-tared glass vials.
Purified analogues were then weighed and a volume of
dimethylsulfoxide added as appropriate to give 10mM
stock solutions used for general storage and inhibition
assays.
4.4. Assays for cysteinyl proteinase activity
Assay protocols were based on literature precedent1 and
modified as required to suit local assay protocols and
have been described previously.13,31
4.5. Measurement of inhibitor on-rates and off-rates
The observed rates of reaction for the association of
compound with enzyme (kon) and for the dissociation
of compound from enzyme (koff) were analysed as previ-
ously described.32 Recombinant human cathepsin K was
assayed in 100mM sodium acetate; pH5.5; 1mM
EDTA; 10mM L-cysteine, 0.05% Tween 20 employing
1.5lM (equal to KaMpp) Z-Leu-Arg-AMC as the sub-
strate.33 For measurements of the association rates,
assays were carried out by addition of various concen-
trations of inhibitor to assay buffer containing substrate
and initiated by the addition of enzyme. For the meas-
urements of dissociation rates, pre-incubated enzyme
plus inhibitor were diluted at least 20-fold into assay
buffer containing substrate. During the course of the
assay less than 10% of the substrate was consumed
and the observed rates corrected for substrate kinetics.
Each purified analogue was analysed giving the follow-
ing characterisation data:
4.3.1. (3aR,6aS)-N-[(1S)-1-(4-Benzoyl-6-oxo-hexahydro-
pyrrolo[3,2-b]pyrrole-1-carbonyl)-3-methylbutyl]-4-tert-
butylbenzamide (22). HPLC tR = 19.43–21.37min (main
peak with 96% of total UV absorbance at 215nm),
HPLC–MS 504.3 [M+H]+, 1029.5 [2M+Na]+. Exact
mass calcd for C30H37N3O4 (MNa+): 526.2682, found
526.2677 (ꢁ0.96ppm).
4.6. Measurement of bone resorption activity
Bone resorption assays were carried out as a service by
Pharmatest Services Ltd, Ita¨inen Pitka¨katu 4, 20520
Turku, Finland. Bone resorption was studied using a
model where human osteoclast precursor cells derived
from peripheral blood were cultured on bovine bone
4.3.2. (3aR,6aS)-N-[1-(4-Benzoyl-6-oxo-hexahydropyr-
rolo[3,2-b]pyrrole-1-carbonyl)-3-methylbutyl]-4-(4-meth-
ylpiperazin-1-yl)benzamide (36). HPLC tR = 12.61min
(97%); HPLC–MS 546.3 [M+H]+, 564.3 [M+H+18]+,

624
M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
Macchia, W.; Zhu, L.; Capparelli, M. P.; Goldstein, R.;
Wigg, A. M.; Doughty, J. R.; Bohacek, R. S.; Knap, A. K.
J. Med. Chem. 2001, 44(26), 4524; (c) Marquis, R. W.; Ru,
Y.; LoCastro, S. M.; Zeng, J.; Yamashita, D. S.; Oh,
H.-J.; Erhard, 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.; McQue-
ney, M. S.; DꢃAlessio, 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; (d) Watts, J.; Benn, A.; Flinn,
N.; Monk, T.; Ramjee, M.; Ray, P.; Wang, Y.; Quibell, M.
Bioorg. Med. Chem. 2004, 12(11), 2903; (e) Catalano, J.
G.; Deaton, D. N.; Long, S. T.; McFadyen, R. B.; Miller,
L. R.; Payne, J. A.; Wells-Knecht, K. J.; Wright, L. L.
Bioorg. Med. Chem. Lett. 2004, 14, 719; (f) Tavares, F. X.;
Boncek, V.; Deaton, D. N.; Hassell, A. M.; Long, S. T.;
Miller, A. B.; Payne, A. A.; Miller, L. R.; Shewchuk, L.
M.; Wells-Knecht, K.; Willard, D. H., Jr.; Wright, L. L.;
Zhou, H.-Q. J. Med. Chem. 2004, 47, 588; (g) Falgueyret,
J.-P.; Oballa, R. M.; Okamoto, O.; Wesolowski, G.;
Aubin, Y.; Rydzewski, R. M.; Prasit, P.; Riendeau, D.;
Rodan, S. B.; Percival, M. D. J. Med. Chem. 2001, 44, 94;
(h) Betschart, C.; Hayakawa, K.; Irie, O.; Sakaki, J.;
Iwasaki, G. Patent WO 03/020721; (i) Altmann, E.;
Hayakawa, K.; Iwasaki, G. Patent WO 03/020278.
8. (a) Altmann, E.; Green, J.; Tintelnot-Blomley, M. Bioorg.
Med. Chem. Lett. 2003, 13, 1997; (b) Kim, T.-S.; Hague,
A. B.; Lee, T. I.; Lian, B.; Tegley, C. M.; Wang, X.;
Burgess, T. L.; Qian, Y.-X.; Ross, S.; Tagari, P.; Lin,
C.-H.; Mayeda, C.; Dao, J.; Jordan, S.; Mohr, C.;
Cheetham, J.; Viswanadhan, V.; Tasker, A. S. Bioorg.
Med. Chem. Lett. 2004, 14, 87.
slices for 9days and allowed to differentiate into bone-
resorbing osteoclasts, and the formed mature osteoclasts
were then allowed to resorb bone. After the culture per-
iod, collagen fragments (CTX) released from the bone
slices during the culture period were determined using
CrossLapsꢂ for culture assay (Nordic Bioscience, Her-
lev, Denmark). A baseline group without test compound
was included providing a negative control. Additionally,
a positive control E64, an inhibitor of cathepsin en-
zymes and osteoclastic bone resorption, was used to
demonstrate that the test system was able to detect inhi-
bition of bone resorption. Compound 22 was added to
the test culture media at 10lM, 1lM and 100nM
(n = 6 replicates).
Acknowledgements
The authors wish to thank Dr. Chris Urch for his help in
editing this manuscript and Mr. Alan Dickerson,
Mr. Paul Skelton and Mr. Brian Crysell of Cambridge
University, Chemistry Department for elemental
analysis, high resolution mass spectrometry and
NMR analysis and Mr. Mark Sleeman of University
of Oxford for optical rotation data.
References and notes
1. Barrett, A. J.; Rawlings, N. D.; Woessner, J. F. In
Handbook of Proteolytic Enzymes; Academic: New York,
1998.
2. (a) Wolters, P. J.; Chapman, H. A. Respir. Res. 2000, 1(3),
170; (b) Shi, G.-P.; Bryant, R. A. R.; Riese, R.; Verhelst,
S.; Driessen, C.; Li, Z.; Bromme, D.; Ploegh, H. L.;
Chapman, H. A. J. Exp. Med. 2000, 191(7), 1177; (c)
Villadangos, J. A.; Bryant, R. A. R.; Deussing, J.;
Driessen, C.; Lennon-Dumenil, A.-M.; Riese, R. J.; Roth,
W.; Saftig, P.; Shi, G.-P.; Chapman, H. A.; Peters, C.;
Ploegh, H. L. Immunol. Rev. 1999, 172, 109.
3. (a) Lecaille, F.; Kaleta, J.; Bro¨mme, D. Chem. Rev. 2002,
102, 4459; (b) Bro¨mme, D.; Kaleta, J. Curr. Pharm. Des.
2002, 8, 1639.
4. Sajid, M.; McKerrow, J. H. Mol. Biochem. Parasitol.
2002, 120, 1–22.
9. In mid 2002 GlaxoSmithKline announced the advance-
ment of SB-462795, a cyclic peptidomimetic ketone
inhibitor of cathepsin K, into Phase I clinical trials for
osteoporosis. In autumn 2002, Novartis announced the
advancement of AAE-581, an inhibitor of cathepsin K,
into Phase I clinical trials progressing to phase IIb by
autumn 2003.
10. Quibell, M.; Benn, A.; Flinn, N.; Monk, T.; Ramjee, M.;
Wang, Y.; Watts, J. Bioorg. Med. Chem. 2004, 12(21),
5689.
11. Fenwick, A. E.; Gribble, A. D.; Ife, R. J.; Stevens, N.;
Witherington, J. Bioorg. Med. Chem. Lett. 2001, 11,
199.
5. (a) Rosenthal, P. J.; Sijwali, P. S.; Singh, A.; Shenai, B. R.
Curr. Pharm. Des. 2002, 8, 1659; (b) Lee, B. J.; Singh, A.;
Chiang, P.; Kemp, S. J.; Goldman, E. A.; Weinhouse, M.
I.; Vlasuk, G. P.; Rosenthal, P. J. Antimicrob. Agents
Chemother. 2003, 47(12), 3810; (c) Urbina, J. A. Curr.
Pharm. Des. 2002, 8, 287; (d) Urbina, J. A. Expert Opin.
Ther. Pat. 2003, 13(5), 661.
6. (a) Veber, D. F.; Thompson, S. K. Curr. Opin. Drug
Discov. Dev. 2000, 3, 362–369; (b) Hernandez, A. A.;
Roush, W. R. Curr. Opin. Chem. Biol. 2002, 6, 459; (c)
Kim, W.; Kang, K. Expert Opin. Ther. Patents 2002,
12(3), 419; (d) Leung-Toung, R.; Li, W.; Tam, T. F.;
Karimian, K. Curr. Med. Chem. 2002, 9, 979.
7. For example, see: (a) Robichaud, J.; Oballa, R.; Prasit, P.;
Falgueyret, J.-P.; Percival, M. D.; Wesolowski, G.;
Rodan, S. B.; Kimmel, D.; Johnson, C.; Bryant, C.;
Venkatraman, H.; Setti, E.; Mendonca, R.; Palmer, J. T..
J. Med. Chem. 2003, 46, 3709; (b) Greenspan, P. D.;
Clark, K. L.; Tommasi, R. A.; Cowen, S. D.; McQuire, L.
W.; Farley, D. L.; van Duzer, J. H.; Goldberg, R. L.;
Zhou, H.; Du, Z.; Fitt, J. J.; Coppa, D. E.; Fang, Z.;
12. (a) Yang, H.; Jurkauskas, V.; Mackintosh, N.; Mogren,
T.; Stephenson, C. R. J.; Foster, K.; Brown, W.; Roberts,
E. Can. J. Chem. 2000, 78, 800; (b) Moyer, M. P.;
Feldman, P. L.; Rapoport, H. J. Org. Chem. 1985, 50,
5223.
13. See: Quibell, M.; Ray, P. C.; Watts, J. P. Patent WO 04/
007501.
14. Synthesis of the cis-5,5-bicycle (3aS,6aR)-1-benzyl-hexa-
hydrocyclopenta[b]pyrrol-6-one
has
recently
been
reported via a similar intramolecular cyclisation of a
tethered amine, see: Hu, Q.-Y.; Rege, P. D.; Corey, E. J.
J. Am. Chem. Soc. 2004, 19, 5984,
H
H
1. NBS, pentane
Swern [O]
NHBn
N
N
2. CuBr / CH2Cl2
3. LiOH, H2O, DME
HO
H
H
H
Ph
O
Ph
15. For the homologation of Cbz-proline to Cbz-2-pyrroli-
dineacetic acid see: Cassal, J. M.; Fuerst, A.; Meier, W.
Helv. Chim. Acta 1976, 59(6), 1917.

M. Quibell et al. / Bioorg. Med. Chem. 13 (2005) 609–625
625
16. Mandville, G.; Ahmar, M.; Bloch, R. J. Org. Chem. 1996,
61, 1122.
cult to ascertain due to the chiral instability of the
monocyclic ketone system.11
17. Knorr, R.; Trzeciak, A.; Bannwarth, W.; Gillessen, D.
Tetrahedron Lett. 1989, 30(15), 1927.
18. Murphy, A. M.; Dagnino, R.; Vallar, P. L.; Trippe, A. J.;
Sherman, S. L.; Lumpkin, R. H.; Tamura, S. Y.; Webb, T.
R. J. Am. Chem. Soc. 1992, 114, 3156.
19. For a thorough description of general multipin techniques
see: Grabowska, U.; Rizzo, A.; Farnell, K.; Quibell, M.
J. Comb. Chem. 2000, 2(5), 475.
23. (a) Conroy, J. L.; Sanders, T. C.; Seto, C. T. J. Am. Chem.
Soc. 1997, 119, 4285; (b) Conroy, J. L.; Seto, C. T. J. Org.
Chem. 1998, 63, 2367; (c) Marquis, R. W.; Ru, Y.; Zeng,
J.; Lee Trout, R. E.; LoCastro, S. M.; Gribble, A. D.;
Witherington, J.; Fenwick, A. E.; Garnier, B.; Tomaszek,
T.; Tew, D.; Hemling, M. E.; Quinn, C. J.; Smith, W. W.;
Zhao, B.; McQueney, M. S.; Janson, C. A.; DꢃAlessio, K.;
Veber, D. F. J. Med. Chem. 2001, 44, 725.
20. An examination of the acid lability of the Boc group
from linker-construct 35b was performed by quantitative
measurement of the Fmoc chromophore following
acidolytic treatment.19 After treatment with 35% TFA in
DCM for 30min, >95% of the Fmoc chromophore
remained. Quantitative removal of Boc under these
conditions was confirmed by coupling the acid treated
construct with benzoic anhydride, followed by cleavage
and quantitative assessment of the corresponding benzo-
ylated analogue.
24. Stroup, G. B.; Lark, M. W.; Veber, D. F.; Bhattacharyya,
A.; Blake, S.; Dare, L. C.; Erhard, K. F.; Hoffman, S. J.;
James, I. E.; Marquis, R. W.; Ru, Y.; Vasko-Moser, J. A.;
Smith, B. R.; Tomaszek, T.; Gowen, M. J. Bone Miner.
Res. 2001, 16(10), 1739.
25. Veber, D. F.; Johnson, S. R.; Cheng, H.-Y.; Smith, B. R.;
Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45,
2615.
26. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P.
J. Adv. Drug Deliv. Rev. 1997, 23, 3.
21. We have now prepared in excess of 1000 analogues
around general series
27. Riley, R. J. Curr. Opin. Drug Discov. Dev. 2001, 4(1), 45.
28. James, I. E.; Dodds, R. A.; Olivera, D. L.; Nuttall, M. E.;
Gowen, M. J. Bone Miner. Res. 1996, 11, 1453.
29. Atherton, E.; Sheppard, R. C. In Solid Phase Peptide
Synthesis: A Practical Approach; IRL: Oxford, UK, 1989.
30. Meldal, M.; Breddam, K. Anal. Biochem. 1991, 195, 141.
31. Cornish-Bowden, A. Fundamentals of Enzyme Kinetics;
Portland: London, 1995.
4
and following acidolytic
cleavage have noted that in many analogues the car-
boxylic acid R3CONHCHR2–COOH side product is
virtually undetectable, whilst in others it represents
up to 20% material by UV analysis (unpublished
results).
22. The corresponding analogues based around the monocy-
clic 3-amidotetrahydrofuran-4-one scaffold described by
GSK (Ref. 11) have been described previously, see
Quibell, M.; Taylor, S. Patent WO 0069855. However,
accurate inhibition kinetics for these analogues are diffi-
32. (a) Morrison, J. F.; Walsh, C. T. Adv. Enzymol. Relat.
Areas Mol. Biol. 1988, 61, 201; (b) Tian, W. X.; Tsou, C.
L. Biochemistry 1982, 21(5), 1028.
33. Bossard, M. J. J. Biol. Chem. 1996, 271(21), 12517.