Stereoselective synthesis of a thiazolane amide using molecular recognition in the triazolyl-activated ester intermediate

Article abstract of DOI:10.1016/j.tetlet.2006.01.047

An amide derived from penicillin V and racemic (R/S)-2-aminobutanol was prepared with 83% de and shows significantly higher toxicity than the pure diastereomers prepared from homochiral 2-aminobutanol. This has been attributed to conformational changes in

Full text of DOI:10.1016/j.tetlet.2006.01.047

Tetrahedron Letters 47 (2006) 1737–1740
Stereoselective synthesis of a thiazolane amide using
molecular recognition in the triazolyl-activated ester intermediate
Peter Styringa,* and Sannie S. F. Chongb
aDepartment of Chemical and Process Engineering, The University of Sheffield, Mappin Street, Sheffield S1 3JD, UK
bDepartment of Chemistry, The University of Hull, Hull HU6 7RX, UK
Received 16 November 2005; revised 20 December 2005; accepted 12 January 2006
Abstract—An amide derived from penicillin V and racemic (R/S)-2-aminobutanol was prepared with 83% de and shows significantly
higher toxicity than the pure diastereomers prepared from homochiral 2-aminobutanol. This has been attributed to conformational
changes in the resolved product brought about through hydrogen-bonded self-assembly in the intermediate.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Efforts to combat the replication of the HIV1,2 virus
have led to the development of a number of inhibitors
of the HIV-1 protease enzyme based on readily available
penicillin G3,4 as the chiral building block. Derivatisa-
tion of penicillin G has resulted in a range of materials
that contain several amide motifs within the low mole-
cular weight structures.3 These are associated with the
N-substituted b-lactam ring, the carboxy function on the
thiazolane ring and also the amide formed on nucleo-
philic opening of the b-lactam ring using a primary
amine. Computer assisted molecular simulation and
crystallographic studies4,5 have shown these to be essen-
tial for binding with lysine and aspartic acid residues at
the active site of the HIV-1 protease enzyme.
of 2-aminobutanol as well as the racemic mixture with
surprising results. While both pure enantiomers react
as expected to give the homochiral amide product, reac-
tion of the acid with racemic 2-aminobutanol leads to the
isolation of an enantiomerically enriched amide, the
results of which will be reported here. As a comparison,
the same reactions were performed using the analogous
penicillin G intermediate, however, no resolution of
enantiomers was observed. A similar effect has been ob-
served by Birch and co-workers7 in the synthesis of pep-
tides using 1-hydroxybenzotriazole and DCC on racemic
mixtures of N-phthaloyl and N,N-dimethylamido pro-
tected amino acids. This has been shown to occur
through competitive reaction of the two enantiomers in
a kinetically controlled reaction. However, we will show
by comparison that our resolution is somewhat more
complex, involving a proposed self-organisation of the
ester intermediate, and that even with prolonged reaction
times the racemic amide is not formed to any great
extent, as would be expected if the resolution mechanism
was purely kinetically controlled.
In the course of our investigations into new HIV-1 pro-
tease inhibitors we have extended the penicillin-based
methodology to include the previously unstudied deriva-
tives of penicillin V. The final step in the synthetic proce-
dure is amide formation at the thiazolane carboxy group
of a benzylamine ring-opened penicillin V using TBTU
[2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium
tetrafluoroborate] as the coupling agent6 in the presence
of an organic amine base. TBTU gives the 1-oxybenzo-
triazolyl-activated ester that can be readily converted
to an amide by reaction with an appropriate amine. This
has been performed using each of the pure enantiomers
The synthetic procedures3 for both the penicillin G
(R1 = Ph) and V (R1 = PhO) derivatives were the same
and are summarised in Scheme 1. The amides of chiral
[5-(R) 5-(S) R1 = Ph, 6-(R) 6-(S) R1 = OPh] and racemic
2-aminobutanol (5-(R/S) R1 = Ph, and in theory 6-(R/
S) R1 = OPh) were obtained using a peptide-coupling
method employing TBTU as the activating agent with
DIPEA (N,N-diisopropylethylamine) as base.8 The tar-
get polyamides were isolated by flash chromatography
over silica gel using step gradient 1–10% MeOH in
Keywords: Self-assembly; Templated reaction; Resolution.
*
Corresponding author. Tel.: +44 (0) 114 222 7571; fax: +44 (0) 114
222 7501; e-mail: p.styring@sheffield.ac.uk
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.01.047

1738
P. Styring, S. S. F. Chong / Tetrahedron Letters 47 (2006) 1737–1740
Scheme 1. General scheme for the synthesis of the target amides.
DCM as an eluent. The yields reported are isolated
yields and the purity of the final products was confirmed
using H NMR, GC, IR and elemental analysis.
dised chain oxygen, we propose that there must be an
interaction between the aminoalcohol ‘substrate’ and
XV that leads to this stereoselectivity. In order to explain
1
this effect intermediate XV was constructed in CAChe
9,10
The carboxylate anions of 3 and 4 react with TBTU to
form an intermediate X with the elimination of tetra-
methylurea. This activates the acid carbonyl group
towards nucleophilic attack by an amine, as O-benzo-
triazole (ꢀOBt) is a good leaving group. The intermedi-
ate, obtained from penicillin V (XV) differs from that
obtained from penicillin G (XG) by a single oxygen atom
in the side chain [R1 = Ph (G) and PhO (V)]. Each peni-
cillin derivative was reacted with racemic (R/S)-2-
aminobutanol, and enantiomerically pure (R)- and (S)-
2-aminobutanol. The amides formed from penicillin G
with the aminoalcohols did not show any unusual fea-
Worksystem for Windows and the MM3
force field
implemented to generate a minimum energy structure.
Comprehensive conformational analysis calculations en-
sured that we arrived at the global minimum rather than
a local intermediate energy minimum. The model was
then studied in the presence of (R)-2-aminobutanol
and (S)-2-aminobutanol. By evoking the hydrogen bond
monitor within CAChe the inter-molecular interactions
between each chiral substrate and XV were probed. Both
(R)- and (S)-2-aminobutanol showed hydrogen-bonded
1
tures in the H NMR spectra using DMSO-d6 as a sol-
vent. The new amide proton, assigned as Hi (Scheme
1), appeared as a single proton doublet for the homochi-
ral derivatives and as two equivalent ‘half proton’ doub-
lets for the diastereomeric mixture. In the case of the
homochiral amides derived from penicillin V (6-(R)
and 6-(S)) the spectra observed showed single doublets
for Hi integrating to single protons at dH 7.54 and
7.76 ppm, respectively. The anomaly occurred when
intermediate XV was reacted with racemic (R/S)-2-
aminobutanol to give in theory 6-(R/S). Rather than
the two equivalent doublets expected for Hi, two non-
equivalent doublets were observed at dH 7.54 and
7.64 ppm with relative integrations of 91.5:8.5, respec-
tively. Therefore, it can be concluded that the (R)-amide
is forming stereoselectively relative to the (S)-amide. In
fact from the relative integration we arrive at a de for
the (R)-amide of 83%. As the only difference between
the penicillin G and V derivatives is the single sp3 hybri-
Figure 1. Nucleophilic attack of (R)-2-aminobutanol at the activated
ester of the self-assembled intermediate showing ‘bound’ (S)-2-
aminobutanol.

P. Styring, S. S. F. Chong / Tetrahedron Letters 47 (2006) 1737–1740
1739
O
R1
NH
H
Compound
LC50 (µg/ml) LC90 (µg/ml)
87 10 112
H
S
N
5-(R/S)
6-(R/S)
5-(R)
8
O
HN
R2
45
6
86 10
134 21
184 31
150 20
192 20
R2
78 10
102 20
101 12
136 22
R1
R1 = OPh
=
Ph
6-(R)
CONH
CONH
CONH
5-(R/S)
6-(R/S)
OH
OH
OH
5-(S)
R1
R1 = OPh
=
Ph
5-(R)
6-(R)
6-(S)
R1
R1 = OPh
=
Ph
5-(S)
6-(S)
Figure 2. Toxicology data for the amides derived from penicillin V and G.
interactions with XV, the amino group being associated
with the side chain amide of the penicillin, the phenyl
oxygen of the side chain and interestingly the nitrogens
of the triazole ring. Furthermore, it was noted that the
benzotriazole motif had reoriented through space, com-
pared to the free intermediate, to give the hydrogen
bonds. The role of the phenyl oxygen also became
apparent as a strong hydrogen bond was observed with
the hydroxyl motif of the chiral substrate. This is not
possible in the penicillin G derivative. Stereoselection
can be explained by the steric crowding when the roto-
mers obtained by varying the conformation of the ethyl
substituent at the chiral centre of the isomers of
2-aminobutanol associated with the intermediate XV
are considered.
benzotriazole group is now exposed towards nucleo-
philic attack. Therefore, the (R)-amide forms in excess.
Toxicological studies11–13 were performed using colonic
epithelial cells in Eagle’s medium supplemented by Foe-
tal Calf serum at 37 ꢁC.ꢀ All assays were performed in
triplicate. The 100% and 0% controls were used to con-
vert the other absorbances to a percentage growth
value. From these, the percentage inhibition of growth
was calculated and this was plotted against log10 drug
concentration. The concentrations giving 50% (LC50)
and 90% inhibition (LC90) were calculated and the val-
ues are shown in Figure 2. The data generally indicated
that a minimum concentration of 68 lg/ml was required
to impart a toxicological effect to 50% of the cultured
cell, or 104 lg/ml for 90% of the cultured cell. This gave
a positive implication stating that the above final com-
pounds were relatively ‘safe’ and should proceed to
the next stage of testing to determine their anti-viral
potential.
The only exception to this was the stereoenriched mix-
ture 6-(R/S) that had shown the anomaly in the 1H
NMR spectrum. In comparison, 6-(R/S) had much lower
LC50 and LC90 values. NMR showed that this mixture
ꢀ A colonic epithelial cell line HT29 was cultured in Eagle’s medium
supplemented with 10% v/v Foetal Calf serum. For the assay, cells
(ꢁ500 per well) were seeded into a 96-well multiwell micro titre plate
(Nunc) and 200 ll of medium was added. The test compounds were
added into the wells dissolved in DMSO; the maximum amount of
DMSO added to each well was 1% v/v. The plates were then
incubated at 37 ꢁC for 48 h in an atmosphere of 5% CO2 and O2 then
the medium in each well was removed and replaced with fresh
medium (100 ll) containing 0.1% MTT [3-(4,5-dimethyl-2-thiazolyl)-
2,5-diphenyl-2H-tetrazolium bromide; thiazole blue, Sigma]. The
plate was then replaced into the incubator and left for 2 h. After this
period, 100 ll of Sorensen’s medium (pH 10) containing DMSO (1%
v/v) was added to each well and the absorbance in each well read in a
plate reader at 570 nm. For each plate several controls were used,
which were cells grown in the absence of drug (100% control) and
cells incubated with 1% v/v DMSO (solvent control). One uninnoc-
ulated well was used to determine the zero value.
For (S)-2-aminobutanol, the substrate sits in the polar
‘cleft’ with the ethyl group pointing away form the bulk
of the molecule (Fig. 1). This may be considered to be a
favourable form. The (R)-2-aminobutanol on the other
hand fits less well with severe steric interactions between
the ethyl group on the substrate and the bulky benzo-
triazole, and the phenoxy and benzylamine groups on
XV. Therefore, binding of the (S)-isomer predominates
and (R)-2-aminobutanol is free in excess to react with
the activated ester group, which because of the confor-
mational changes brought about by the rotation of the

1740
P. Styring, S. S. F. Chong / Tetrahedron Letters 47 (2006) 1737–1740
A. V.; Cobley, K. N.; Orr, D. C.; Storer, R.; Weingarten,
G. G.; Weir, M. P. J. Med. Chem. 1992, 35, 3080;
Wonacott, A.; Cooke, R.; Hayes, F. R.; Hann, M. M.;
Jhoti, H.; McMeekin, P.; Mistry, A.; Murry-Rust, P.;
Singh, O. M. P.; Weir, M. P. J. Med. Chem. 1993, 36,
3113.
possessed a stereoexcess of 6-(R) and this suggested that
while being chemically identical to 6-(R) obtained from
the pure enantiomer of 2-aminobutanol (albeit with
some contamination from the other diastereoisomer), it
had a different conformational structure. It is proposed
that the conformation in the product is fixed by that in
the self-assembled intermediate, and is a structure with
low local minimum energy rather than the global mini-
mum formed using the homochiral reagent. The change
in conformation would then lead to a different bioactiv-
ity, most likely through the formation of a new polymor-
phic crystal form.14 This is reflected in the significantly
different result in the toxicity test. The difference is most
striking when comparisons are made between the prod-
ucts formed using the homochiral substrates and the
racemic substrates. If we compare the LC50 results for
5 we see that the value obtained for the racemic product
(87 10 lg/ml) is intermediate between those of the pure
(S)-(101 12 lg/ml) and pure (R)-(78 10 lg/ml) prod-
ucts. It is clear from the 1H NMR spectrum of 6 obtained
using the racemic substrate that it has a de of 83%, en-
riched with the (R)-substrate. Therefore, the LC50 results
for 6-(R/S) (45 6 lg/ml) should closely match 6-(R)
(102 20 lg/ml), which is clearly not the case.
6. Jones, J. Amino Acid and Peptide Synthesis; Oxford
University Press: New York, 1992; p 34.
7. Birch, D.; Hill, R. R.; Jeffs, G. E.; North, M. Chem.
Commun. 1999, 941.
8. In a typical procedure: To a solution of the free penicillic
V acid 4 (5.23 mmol) in DMF (50 cm3) were added
sequentially DIPEA (5.34 mmol), (R/S)-2-aminobutanol
(5.38 mmol) and TBTU (5.36 mmol). The resulting solu-
tion was stirred at room temperature overnight, before
partitioning between EtOAc (ca. 150 cm3) and water
(50 cm3). The organic phase was removed and the aqueous
phase was extracted with EtOAc (2 · 50 cm3). The com-
bined organic phases were washed with 2 N HCl (50 cm3),
saturated aqueous NaHCO3 (50 cm3) and brine (50 cm3)
before drying (MgSO4). The solvent was removed in
vacuo, and the resulting residue was purified by flash
column chromatography (silica gel, eluent system: 1–10%
MeOH in CH2Cl2 gradually increasing the polarity) to
22
afford 6 as a white solid. Yield = 23%; mp 85.1 ꢁC. ½aꢂD
+56.20 (c 0.01192, CDCl3). Elemental analysis: Found: C,
59.97; H, 7.03; N, 10.28%. Calcd for C27H36O5N4SÆ
0.5H2O (FW 537.66): C, 60.32; H, 6.93; N, 10.42%. IR
(KBr disc): mmax/cmꢀ1 3700–3140 (br, OH); 3080, 3040
(–CONH–); 2900–2800 (saturated C–H); 1650 (C@O
stretch, amide I band); 1530 (amide II); 1250 (C–O);
760, 700 (monosubstituted benzene ring).
Acknowledgements
The University of Hull is acknowledged for the provi-
sion of funding for a studentship to SSFC. Hull Analy-
tical Services are thanked for the provision of spectral
and analytical services.
1H NMR (400 MHz, DMSO): dH 0.83 (3H, t, J 7.6 and
7.3, CH2CH3); 1.14 (3H, s, terminal CH3); 1.33 (1H, m,
0
Hl ); 1.50 (3H, s, terminal CH3); 1.55 (1H, m, Hl); 3.27
0
(1H, m, Hj); 3.35 (1H, m, Hj ); 3.46 (1H, d, J 12.2, Ha);
3.66 (1H, m, Hk); 3.89 (1H, dd, J 12.2 and 8.3, Hh); 4.26,
0
4.28 (2H, dABq, J 13.6 and 2.9, Hg, Hg ); 4.40 (1H, t, J 8.3
References and notes
0
and 7.8, Hb); 4.54 (2H, ABq, J 14.9 and 14.7, Hf, Hf ); 4.64
(1H, t, J 5.4 and 5.1, OH); 4.96 (1H, t, J 8.0 and 7.8, Hc);
6.94 (3H, m, aromatic protons ortho/para to phenoxygen);
7.18–7.32 (7H, m, aromatic ring protons); 7.54 (0.915H, d,
J 8.6, Hi, diastereomer); 7.64 (0.085H, d, J 8.8, Hi,
diastereomer); 8.15 (1H, t, J 8.1 and 8.8, Hd); 8.57 (1H, t, J
6.1 and 5.8, He). MS (EI): m/z 528 [M]+; 497
[MꢀCH2CH3]+; 260, 231, 91 [PhCH2]+.
1. Gallo, R. C.; Montaigner, L. Sci. Am. 1988, 259, 40.
2. Coffin, J.; Hasse, A.; Levy, J. A.; Montagnier, L.;
Oroszlan, S.; Teich, N.; Temin, H.; Toyoshima, K.;
Varmus, H.; Vogt, P.; Weiss, R. Science (London) 1986,
232, 697.
3. Holmes, D. S.; Bethell, R. C.; Cammack, N.; Clemens, I.
R.; Kitchin, J.; McMeekin, P.; Mo, C. L.; Orr, D. C.;
Patel, B.; Paternoster, I. L.; Storer, R. J. Med. Chem.
1993, 36, 3129.
9. Allinger, N. L. J. Am. Chem. Soc. 1977, 99, 8127.
10. Yuk, Y. H.; Lii, J. H. J. Am. Chem. Soc. 1989, 111,
8551.
11. Toxicity tests were carried out by the Department of
Biological Sciences at the University of Hull.
12. Liu, Y. B.; Peterson, H.; Kimura, D. A.; Schubert, D.
J. Neurochem. 1997, 69, 581.
13. Adullah, M. A. Eur. J. Biochem. 1980, 112, 331.
14. Buttar, D.; Charlton, M. H.; Docherty, R.; Starbuck, J.
J. Chem. Soc., Perkin Trans. 2 1988, 763.
4. Humber, D. C.; Bamford, M. J.; Bethell, R. C.; Cammack,
N.; Cobley, K. N.; Evans, D. N.; Gray, N. M.; Hann, M.
M.; Orr, D. C.; Saunder, J.; Shenoy, B. E. V.; Storer, R.;
Weingarten, G. G.; Wyatt, P. G. J. Med. Chem. 1993, 36,
3120.
5. Daopin, S.; Piez, K. A.; Ogawa, Y.; Davies, D. R. Science
1992, 257, 396; Humber, D. C.; Cammack, N.; Coates, J.