cis-6-oxo-hexahydro-2-oxa-1,4-diazapentalene and cis-6-oxo- hexahydropyrrolo[3,2-c]pyrazole based scaffolds: Design rationale, synthesis and cysteinyl proteinase inhibition

Article abstract of DOI:10.1016/j.bmcl.2005.01.022

The 5,5-bicycles cis-6-oxo-hexahydro-2-oxa-1,4-diazapentalene 3 and cis-6-oxo-hexahydropyrrolo[3,2-c]pyrazole 4 were designed as rotationally restricted templates towards the preparation of inhibitors of CAC1 cysteinyl proteinases. The design strategy was

Full text of DOI:10.1016/j.bmcl.2005.01.022

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1327–1331
cis-6-Oxo-hexahydro-2-oxa-1,4-diazapentalene and
cis-6-oxo-hexahydropyrrolo[3,2-c]pyrazole based scaffolds:
design rationale, synthesis and cysteinyl proteinase inhibition
Yikang Wang, Alex Benn, Nick Flinn, Tracy Monk, Manoj Ramjee,
John Watts and Martin Quibell*
Amura Therapeutics Limited, Incenta House, Horizon Park, Barton Road, Comberton,
Cambridge CB3 7AJ, UK
Received 11 October 2004; revised 7 January 2005; accepted 12 January 2005
Abstract—The 5,5-bicycles cis-6-oxo-hexahydro-2-oxa-1,4-diazapentalene 3 and cis-6-oxo-hexahydropyrrolo[3,2-c]pyrazole 4 were
designed as rotationally restricted templates towards the preparation of inhibitors of CAC1 cysteinyl proteinases. The design strat-
egy was exemplified through the solution and solid phase preparation of potent inhibitors of human cathepsin K and may poten-
tially be applied to inhibitors of other CAC1 proteinases.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Inhibitors of the CAC1 class of proteolytic enzyme1
have the potential to treat a wide range of indications
such as osteoporosis, multiple sclerosis, rheumatoid
and osteoarthritis, atherosclerosis and cancer as well
as parasitic infections such as malaria and Chagas dis-
ease.2–4 Consequently, a significant resource is currently
devoted within the pharmaceutical industry towards the
development of CAC1 proteinase inhibitors,5 which has
recently provided the first clinical candidates with
cathepsin K as the target and osteoporosis as the main
therapeutic indication. In the search for new protease
inhibitors, we have described the design, synthesis and
inhibition kinetics for bicyclic peptidomimetics 16 and
25 that are constrained scaffolds providing potent and
selective inhibitors of CAC1 cysteinyl proteinases. A
key design feature of 1 and 2 was stabilisation of the
otherwise chirally labile position situated a to the ke-
tone,7 by exploiting the kinetic and thermodynamic sta-
bility of a cis-fused 5,5-bicycle. We have now extended
the design process and herein report the rationale, syn-
thesis and inhibition kinetics for the new heterobicycles
cis-6-oxo-hexahydro-2-oxa-1,4-diazapentalene 3 and cis-
6-oxo-hexahydropyrrolo[3,2-c]pyrazole 4.
We have previously reported our molecular modelling
studies based on N-acylated leucine analogues of scaf-
folds 1 and 2 with cathepsin K.5,6 These studies pre-
dicted that the leucine carbonyl could form
a
conserved hydrogen bond with the proteinase backbone
N–H of glycine66. However, within the predicted bioac-
6
tive conformation of the si tetrahedral intermediate,
this required the carbonyl group to occupy approxi-
mately the same plane as the 5,5-bicyclic framework
with a tertiary amide rotational angle (x) of >140ꢁ
(i.e., akin to structure 1a rather than 1b). In general,
the rotational freedom of peptide bonds is restricted
due to the partial double bond character of the CO–
NH secondary amide, which results in a high-energy
barrier for cis ! trans isomerisation. However numer-
ous studies have shown that N-acylated prolines (gen-
eral structure 5a) are unique amongst aminoacids in
that they exist as readily interconvertable mixtures of
cis and trans isomers about the CO–N tertiary amide
bond.8 We considered the possibility that the rotational
freedom present in N-acylprolines may be mirrored in
scaffolds 1 and 2 which also contain the tertiary amide
bond. Therefore, a range of design strategies were exam-
ined with the potential to restrict rotational freedom
about the CO–N bond within 1 and 2 and confine the
Keywords: Cysteinyl proteinase inhibitor; Heterobicyclic ketone.
*
Corresponding author. Tel.: +44 1223 264211; fax: +44 1223
265662; e-mail: martin.quibell@incenta.co.uk
Current address: Siena Biotech S. p.A., Via Fiorentina, 1 53100 Siena,
Italy.
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.01.022

1328
Y. Wang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1327–1331
O
O
δ
3a
δ
3a
Y
N
X
N
δ
6a
6a
R2
R1
R1
R1
R1
N
N
R2
O
O
O
N
N
N
R3
6a
6a
3a
3a
ω
ω
R1
O
O
O
O
O
O
O
O
O
5a Y = CH2
b Y = O
c Y = NH
3 (3aS, 6aS) X = O
4 (3aR, 6aS) X = NH
1b (3aS, 6aR)
2 (3aR, 6aS)
1a (3aS, 6aR)
o
o
ω
ω
= 0
= 180
x-bond angle close to the predicted bioactive conforma-
tion.9 Initially we considered introduction of steric bulk
to the d-position of the bicyclic scaffold (see scaffolds 1
and 2) to spatially occlude rotation about the CO–N
bond. However, this was dismissed because we had pre-
viously shown that the d-methylene protons and the a-
CAC1 proteinases for compounds 15 and 32 derived
from scaffolds 3 and 4, respectively.
Synthesis of the cis-hexahydropyrrolo[3,2-b]pyrrol-3-
one scaffold 2 has previously been described in detail
and utilises an intramolecular cyclisation of a tethered
amine epoxide to control stereochemistry during forma-
tion of the bicyclic framework.5 We envisaged that an
analogous strategy would provide compounds derived
from bicyclic scaffolds 3 and 4 (Schemes 1 and 2).12
˚
proton of the P2 leucine residue are only 2.5 A apart
in the predicted bioactive conformation.6 Thus, even
simple d-methylation could have a deleterious effect by
inducing a rotation in the psi (w) angle of the P2 leucine
within the inhibitor into a sub-optimal binding confor-
mation. Alternatively, we were intrigued by the litera-
ture precedence for heteroatom substitution of the
d-methylene as a method for restricting rotational free-
dom in N-acylated prolines.10,11 Literature clearly dem-
onstrates that N-acetyl-5-oxo-proline (5b, R1 = CH3,
R3 = OH) exists almost exclusively trans (x ꢀ 180ꢁ) at
neutral pH whereas N-acetyl-proline (5a, R1 = CH3,
R3 = OH) forms a 1:1 cis/trans mixture.11 A simplistic
explanation for the predominant trans conformation
of 5b is that the carbonyl oxygen and 5-oxo-oxygen
would be eclipsed in the cis-conformation, which
through electron/electron repulsion provides a high en-
ergy barrier to rotation. Theoretically, a similar conclu-
sion may be drawn for N-acetyl-5-aza-proline (5c,
R1 = CH3, R3 = OH), although experimentally this has
not been shown to date. We were hopeful that the ob-
served stabilisation of the trans-amide in N-acetyl-5-het-
ero-prolines may provide a corresponding effect when
incorporated into the new cis-6-oxo-hexahydro-2-oxa-
Synthesis of protected ketone 6, a building block that
was suitable for solid phase syntheses of analogues of
scaffold 3,5,12 commenced from the available Boc–dehy-
droproline 7. Conversion of acid 7 to methyl ester 8
(88%) was followed by reduction to alcohol 9 (92%).
Conversion of hydroxyl 9 to mesylate 10 (100%), pro-
vided a crude oil that was used without further purifica-
tion. Deprotonation of benzyl N-hydroxycarbamate
(Cbz-NHOH) provided the anion for nucleophilic dis-
placement of mesylate 10, which after heating at 65 ꢁC
overnight and following silica gel purification gave the
desired aminooxymethyl intermediate 11 (26%)13 and
recovered mesylate (47%). Epoxidation of alkene 11
provided 12 as a mixture of syn-12a and anti-12b epox-
ides (62%) that co-eluted upon silica chromatography
and exhibited complex proton NMR spectra. We envis-
aged that within epoxide mixture 12 only anti-12b could
cyclise because potential cyclisation of the syn-epoxide
12a would lead to the thermodynamically disfavoured
trans-5,5-bicycle.5 Indeed, treatment of mixture 12 with
potassium carbonate in acetonitrile, followed by purifi-
cation over silica gel, led to the recovery of syn-epoxide
1,4-diazapentalene
3
and cis-6-oxo-hexahydropyr-
rolo[3,2-c]pyrazole 4 scaffolds. In this context, we have
now examined the effects on inhibitor potency against
ii
iv
iii
OH
O
OMs
COR4
N
Boc
N
Boc
N
Boc
N
Boc
NHCbz
7 R4 = OH
10
9
11
i
8 R4 = OCH3
v
Boc
N
Boc
N
Boc
N
O
3a
6a
3a
3a
vi
vii, viii
ix
O
O
N
Fmoc
O
O
N
N
Boc
N
NHCbz
6a
6a
Cbz
OH
OH
O
Fmoc
13
12a syn
12b anti
6
14
'Via anti-epoxide'
Scheme 1. Synthesis of (3aS,6aS)-6. Reagents and conditions: (i) ethereal CH2N2, ꢁ15 ꢁC—rt; (ii) LiBH4, MeOH, THF; (iii) methanesulfonyl
chloride, pyridine, DCM; (iv) Cbz-NH-OH, NaH, THF, 65 ꢁC; (v) m-chloroperoxybenzoic acid, DCM; (vi) potassium carbonate, CH3CN; (vii)
Pd–C, H2, ethanol; (viii) 1.05 equiv Fmoc–Cl, 2.1 equiv Na2CO3, 1,4-dioxane, water; (ix) Dess–Martin periodinane, DCM.

Y. Wang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1327–1331
1329
R6
i
v
vii
Boc
N
R5
R7
N
Cbz
N
N
N
Cbz
23 R7 = H
COOCH3
N
Cbz
N
NHAlloc
vi
Cbz
H
17 R5 = OH
18 R5 = OMs
19 R5 = N3
20 R5 = NH2
21 R6 = H
16
ii
viii
22 R6 = Boc
24 R7 = Alloc-Leu
25 R7 = H-Leu
iii
vii
ix
iv
26 R7 = 4-But-PhCO-Leu
x
R8
N
O
R9
O
COPh
3a
3a
xiv
O
xi
Boc
N
N
N
Boc
N
R7
N
N
N
Cbz
6a
6a
N
N
4-But-PhCO-Leu
H
OH
H
O
But
28 R8 = Cbz
29 R8 = H
31 R9 = Boc
27a syn
27b anti
xii
xv
32 R9 = H
xiii
30 R8 = PhCO
Scheme 2. Synthesis of (3aR,6aS)-N-[(1S)-1-(4-benzoyl-6-oxo-hexahydro-pyrrolo[3,2-c]pyrazole-1-carbonyl)-3-methyl-butyl]-4-tert-butylbenzamide
32. Reagents and conditions: (i) LiBH4, MeOH, THF; (ii) methanesulfonyl chloride, triethylamine, DCM; (iii) sodium azide, DMF, 110 ꢁC; (iv)
Ph3P/H2O, 1,4-dioxane, 50 ꢁC; (v) 3-phenyloxaziridine-2-carboxylic acid allyl ester, DCM; (vi) (Boc)2O, triethylamine/MeOH, 60 ꢁC; (vii) Pd(PPh3)4,
PhSiH3, DCM; (viii) Alloc-Leu-F, DMF; (ix) 4-tert-butylbenzoic acid, HBTU, HOBT, NMM, DMF; (x) m-chloroperoxybenzoic acid, DCM; (xi)
potassium carbonate, CH3CN, 60 ꢁC; (xii) Pd–C, H2, ethanol; (xiii) (PhCO)2O, DMF; (xiv) Dess–Martin periodinane, DCM; (xv) TFA.
12a (25%), together with a new compound that was less
mobile on TLC and identified as the cis-heterobicycle 13
(50%). Removal of Cbz protection was followed by re-
protection with 9-fluorenylmethoxycarbonyl chlorofor-
mate (Fmoc–Cl), providing bicyclic alcohol 14 (75%).
As the final step in Scheme 1, alcohol 14 was smoothly
oxidised by Dess–Martin periodinane to provide target
bicyclic ketone 6 (80%) as a white solid.14 With ketone
6 in hand, a solid phase synthesis, analogous to that
previously described for bicycles 1 and 2,5,6,12 was
undertaken to give (3aS,6aS)-N-[(1S)-1-(4-benzoyl-6-
oxo-hexahydro-2-oxa-1,4-diaza-pentalene-1-carbonyl)-
3-methyl-butyl]-4-dimethylamino benzamide 15 (Table
1).15
version to mesylate 18 (95%) provided a crude oil that
was used without further purification. Surprisingly, all
attempts to directly prepare protected hydrazide 21
through displacement of mesylate 18 with allyl-
oxycarbonylhydrazide16 unexpectedly failed. Numerous
solvents and reaction conditions were examined, but in
each case either no reaction or an intractable mixture
was obtained. Therefore, we explored other strategies
towards hydrazide 21 and successfully adapted the elec-
trophilic N-amination methods of Niederer et al., used
to prepare protected hydrazino acids from the corre-
sponding aminoacids.17 Thus, treatment of mesylate 18
with sodium azide in DMF gave azidomethyl intermedi-
ate 19 (72%), which was reduced to amine 20 (74%) fol-
lowing the general methods of Mandville et al.18
Treatment of amine 20 with N-Alloc-3-phenyloxazir-
idine17 in DCM afforded the desired hydrazide 21 as a
pale yellow oil (36%), which was N-Boc substituted to
give the tri-orthogonal protected 22 (79%). Unfortunately,
Next we undertook the synthesis of inhibitor 32, an ana-
logue of scaffold 4, entirely in solution (Scheme 2). Syn-
thesis commenced from Cbz-dehydroproline methyl
ester 16, which was reduced to alcohol 17 (90%). Con-
Table 1. Preliminary inhibitory activities (Ksis, nM) for 5,5-bicyclic inhibitors 15 and 32 and the equivalent d-methylene bicycles derived from scaffold
25,21 against CAC1 proteinases (mean of n = 3 determinations). Assay conditions are as detailed previously12
O
3a
O
X
N
N
6a
N
H
O
O
R
Compound
Cat. K
Cat. L
Cat. S
Cat. B
Cruz.
CPB
15 (X = O, R = Me2N)
32 (X = NH, R = But)
3321 (X = CH2, R = Me2N)
345 (X = CH2, R = But)
40.1 16.7
3.5 0.1
5.5 2.0
>4300
370 80
>3000
>6000
2500 1500
93.2 31.7
600 350
173 86
1000 600
49.5 23.6
1170 540
691 350
434 142
>3000
>4500
119 12
>10,000
>10,000
1800 630
>3500
10.1 6.7

1330
Y. Wang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1327–1331
the analogous epoxidation and base catalysed ring-
closure described earlier (11 ! 13) was not an option
for intermediate 22 due to the complicating presence
of the Alloc protecting group. Thus, removal of Alloc
protection following the general conditions described
by Dessolin et al.19 provided hydrazide 23 (97%) that
was treated with the acyl fluoride of Alloc-leucine5 to
obtain acylated intermediate 24 (63%). Removal of
Alloc protection19 (96%) was followed by acylation of
amine 25 with 4-tert-butylbenzoic acid using standard
uronium activation chemistries to give substituted
hydrazide 26 (84%). Epoxidation of 26 now proceeded
smoothly to give a mixture of syn-27a and anti-27b
(62%). In an analogous manner to that described earlier
for epoxide mixture 12, we envisaged that only anti-27b
would cyclise. Thus, treatment of epoxide mixture 27
with potassium carbonate in acetonitrile, followed by
purification over silica gel, gave a new compound that
was less mobile on TLC and was identified as the cis-
heterobicycle 28 (42%). The Cbz group was then re-
placed with a benzoyl group to obtain the Boc-protected
alcohol 30 (two steps 68%). Subsequent oxidation of 30
with Dess–Martin periodinane provided N-Boc pro-
tected bicyclic ketone 31 (81%),20a which was treated
with trifluoroacetic acid to give the final fully deprotec-
ted inhibitor 32 (37%).20b
cycles 3 and 4 may provide improved potency against
CAC1 proteinases. Synthetic routes were successfully
devised and example heterobicyclic inhibitors 15 and
32 prepared. Compound 32 was designed with appropri-
ate binding moieties for cathepsin K inhibition,5 but
shows a significant increase in potency against all
CAC1 proteinases examined. This suggests that the
heterobicyclic framework defined in scaffold 4 may in-
deed be rotationally (conformationally) restricted when
compared to the equivalent methylene scaffold 2, but
still able to access the bioactive conformational space.
Therefore scaffold 4 has the potential to give high po-
tency inhibitors of CAC1 proteinases.
Acknowledgements
The authors wish to thank Cambridge University Chem-
istry Department for compound analyses 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. Handbook
of Proteolytic Enzymes; Academic: New York, 1998.
2. Lecaille, F.; Kaleta, J.; Bro¨mme, D. Chem. Rev. 2002, 102,
4459.
3. Bro¨mme, D.; Kaleta, J. Curr. Pharm. Des. 2002, 8, 1639.
4. Sajid, M.; McKerrow, J. H. Mol. Biochem. Parasitol.
2002, 120, 1.
5. For example, see Quibell, M.; Benn, A.; Flinn, N.; Monk,
T.; Ramjee, M.; Ray, P.; Wang, Y.; Watts, J. Bioorg. Med.
Chem. 2005, 13, 609, and references cited therein.
6. Quibell, M.; Benn, A.; Flinn, N.; Monk, T.; Ramjee, M.;
Wang, Y.; Watts, J. Bioorg. Med. Chem. 2004, 12(21),
5689.
Bicyclic inhibitors 15 and 32 were screened against
cathepsins K, L, S and B as well as the parasitic protein-
ases cruzain and CPB.12 Preliminary steady-state inhibi-
tion 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 3321 and 34.5
7. Fenwick, A. E.; Gribble, A. D.; Ife, R. J.; Stevens, N.;
Witherington, J. Bioorg. Med. Chem. Lett. 2001, 11,
199.
8. For example, see: Fischer, S.; Dunbrack, R. L., Jr.;
Karplus, M. J. Am. Chem. Soc. 1994, 116, 11931.
9. In-house molecular modelling was performed as previ-
ously described in detail using WebLab ViewerPro (http://
www.accelrys.com): see, Watts, J.; Benn, A.; Flinn, N.;
Monk, T.; Ramjee, M.; Ray, P.; Wang, Y.; Quibell, M.
Bioorg. Med. Chem. 2004, 12, 2903.
10. Shireman, B. T.; Miller, M. J.; Jonas, M.; Wiest, O. J.
Org. Chem. 2001, 66, 6046.
11. Galardy, R. E.; Alger, J. R.; Liakopoulou-Kyriakides, M.
Int. J. Pept. Protein Res. 1982, 19, 123.
12. Full experimental protocols for Schemes 1 and 2 are
described in: Quibell, M.; Ray, P. C.; Watts, J. P. WO
Patent 04007501. Within this patent see syntheses for 6 (pp
416–423); 15 (p 415); 22 (pp 426–432); 32 (pp 433–437) and
enzymatic methods (pp 531–543).
Table 1 shows that the new heterobicyclic inhibitors de-
rived from scaffolds 3 and 4 provide low nanomolar
inhibitors of human cathepsin K and have the potential,
when substituted with appropriate binding elements, to
inhibit other CAC1 proteinases.5 For the cathepsin K
inhibitors disclosed, the d-methylene (33) to d-oxygen
(15) modification has given an approximately 8-fold loss
in potency. However, the d-methylene (34) to d-nitrogen
(32) modification has given an across the board increase
in potency ranging from approximately 3-fold for
cathepsin K to 14-fold for Leishmania mexicana CPB
and greater than 80-fold for cathepsin B. The inhibition
kinetics for inhibitor 32 are fully reversible against each
proteinase. One possible inference for this universal in-
crease in potency is that the design process detailed here-
in
for
restricting
rotational
(and
therefore
conformational) freedom about the CO–N tertiary
amide bond has, in the case of inhibitor 32 been success-
ful. However, inferences of binding modes based solely
upon changes in potency are an over simplification.
The introduction of the d-heteroatom into scaffold 2
may induce numerous effects such as changes in ring
geometry and puckering of the bicyclic framework,
any of which may influence presentation of the electro-
philic ketone to the active site thiol and alter potency.
13. Data for intermediate 11: TLC (single spot, Rf = 0.35,
EtOAc–hexane 2:3). Anal. Calcd for C18H24N2O5: C,
62.05; H, 6.94; N, 8.04, found C, 62.18; H, 7.05; N, 7.90.
Exact mass calcd for C18H24N2O5 (MNa+): 371.1583,
22
found 371.1590 (d +1.83 ppm); ½aꢂD ꢁ42.4 (c 0.663,
CHCl3).
14. Data for ketone 6: TLC (single spot, Rf = 0.30, EtOAc–
heptane 2:3). Exact mass calcd for C25H26N2O6 (MNa+):
22
473.1689, found 473.1690 (+0.24 ppm); ½aꢂD ꢁ92.4 (c
0.224, CHCl3). dH (300 K, 500 MHz, CDCl3) mixture of
rotamers major:minor 1.5:1, 1.48 (s, C(CH3)3, 5.4H), 1.50
(s, C(CH3)3, 3.6H), 3.49–3.58 (m, BocNCHCH2, 1H),
In summary, a restricted rotation design process applied
to cis-5,5-bicyclic scaffold 2 indicated that new heterobi-

Y. Wang et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1327–1331
1331
3.78–3.92 (m, BocNCH2, 2H), 4.13 (d, J = 9.5 Hz,
BocNCHCH2, 0.4H), 4.20–4.29 (m, Fmoc–CH and
BocNCHCH2, 1.6H), 4.46–4.52 (m, Fmoc–CH2, 1H),
4.60–4.74 (m, Fmoc–CH2, FmocNCH, BocNCH, 2.4H),
4.83 (dd, J = 7.5 and 4.3 Hz, BocNCH, 0.6H), 7.29–7.78
(aromatic, 8H); dC (300 K, 125 MHz, CDCl3) 28.38, 28.31
(C(CH3)3), 46.96, 47.05 (Fmoc–CH), 52.40, 52.93
(BocNCH2), 61.95 (BocNCH), 64.48, 65.31 (FmocNCH),
68.59, 68.76 (Fmoc–CH2), 77.17, 77.31 (BocNCHCH2),
81.61 (C(CH3)3), 120.02, 125.11, 125.35, 127.21, 127.28,
127.98 (Fmoc aromatic CH), 141.29, 141.33, 143.04,
143.12 (Fmoc quaternary), 153.09, 154.00, 157.64 (Boc
and Fmoc C@O), 204.85, 205.44 (C@O).
17. Niederer, D. A.; Kapron, J. T.; Vederas, J. C. Tetrahedron
Lett. 1993, 34(43), 6859.
18. Mandville, G.; Ahmar, M.; Bloch, R. J. Org. Chem. 1996,
61, 1122.
19. Dessolin, M.; Guillerez, M.-G.; Thieriet, N.; Guibe, F.;
Loffet, A. Tetrahedron Lett. 1995, 36(32), 5741.
20. (a) Data for ketone 31: TLC (single spot, Rf = 0.65,
EtOAc–heptane 3:2). Exact mass calcd for C34H44N4O6
(MNa+): 627.3153, found 627.3167 (+2.26 ppm); (b) Data
for inhibitor 32: TLC (single spot, Rf = 0.26, EtOAc–
heptane 3:1); analytical RP-HPLC tR = 18.10–19.70 min
(broad major peak > 95% by UV 215 nm, Phenomenex
˚
Jupiter C4, 5 lm, 300 A, 250 · 4.6 mm, acetonitrile–water
15. Data for inhibitor 15: analytical RP-HPLC tR = 13.18 min
(major peak > 95% by UV 215 nm, Phenomenex Jupiter
gradient). Exact mass calcd for C29H36N4O4 (MNa+):
22
527.2629, found 527.2619 (ꢁ1.78 ppm); ½aꢂD ꢁ168.2 (c
˚
C4, 5 lm, 300 A, 250 · 4.6 mm, acetonitrile–water gradi-
0.022, CHCl3).
ent). Exact mass calcd for C27H32N4O5 (MNa+): 515.2265,
found 515.2285 (+3.96 ppm).
21. Prepared as detailed in Ref. 5. Data for inhibitor 33:
TLC (single spot, Rf = 0.25, EtOAc–heptane 9:1). Exact
mass calcd for C28H34N4O4 (MNa+): 513.2478, found
16. Prepared following, Macleay, R. E.; Lange, H. C. U.S.
Patent 5,338,853, 1994.
22
513.2492 (+2.72 ppm); ½aꢂD ꢁ38.4 (c 0.503, CHCl3).