Synthesis of new biotin derivatives

Article abstract of DOI:10.1016/S0040-4020(99)01022-4

It is now established that the formation of C-S bonds during the conversion of dethiobiotin into biotin involves intermediate carbon radicals. To examine the behavior of dethiobiotin analogs that could lead to allylic and α-epoxyradicals, we have synthesized 4,5-dehydrodethiobiotins and the corresponding epoxides and 5,6-dehydrodethiobiotins. All these compounds have been labelled with ¹⁴C on the carboxyl group.

Full text of DOI:10.1016/S0040-4020(99)01022-4

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 285–293
Synthesis of New Biotin Derivatives
†
*
´
Franc¸ois Moreau, Dominique Florentin and Andree Marquet
´
Laboratoire de Chimie Organique Biologique, Universite Paris VI, CNRS UMR 7613, 4 Place Jussieu, 75252 Paris Cedex 05, France
Received 3 June 1999; accepted 22 November 1999
Abstract—It is now established that the formation of C–S bonds during the conversion of dethiobiotin into biotin involves intermediate
carbon radicals. To examine the behavior of dethiobiotin analogs that could lead to allylic and a-epoxyradicals, we have synthesized
4,5-dehydrodethiobiotins and the corresponding epoxides and 5,6-dehydrodethiobiotins. All these compounds have been labelled with
14C on the carboxyl group. ᭧ 1999 Elsevier Science Ltd. All rights reserved.
Introduction
of analogs of DTB containing radical clocks, namely epoxy
and cyclopropyl groups a to the radical produced at C-6.4
The last step of biotin biosynthesis, namely the introduction
of sulfur into dethiobiotin (DTB) catalyzed by biotin synthase
has long raised a very puzzling problem (Scheme 1).
We have already described the synthesis of their ethylenic
precursors Z and E-4,5-dehydrodethiobiotin 2a and 2b.5
Both compounds are recognized by the enzyme and were
found to covalently bind to the protein when all the
cofactors necessary for the reaction are present. They are
the first mechanism-based inhibitors of biotin synthase.5
However, in the last five years, significant progress has been
made in the understanding of this complex mechanism. It is
now well established that biotin synthase, an [Fe–S] protein
Scheme 1.
is an S-adenosylmethionine (AdoMet) dependent enzyme.1
The reductive cleavage of AdoMet generates a deoxy-
adenosyl radical which is responsible for homolytic C–H
bond cleavage at positions 6 and 9 of DTB.2 We have also
recently shown that the sulfur source is very likely the
[Fe–S] center of the protein.3
This interesting result encouraged us to undertake the synth-
esis of the epoxy and cyclopropyl derivatives 3a, 3b and 4a,
4b.
We also targeted the 5,6-dehydrodethiobiotins 5a, 5b,
which could be of interest for the enzymatic studies
(Scheme 2).
Much more work is, however, necessary to understand in
detail this unprecedented mechanism. One of the
approaches that we are considering is to look at the behavior
Results and Discussion
Synthesis of dehydrodethiobiotins
Keywords: epoxides; labelling; phosphine ylids; biotin synthase.
* Corresponding author. Tel.: 33-1-44-27-5535; fax: 33-1-44-277150;
We report first a new synthesis of the intermediate alcohols
6a, 6b presenting several advantages over the previously
described one, which was based on the construction of the
imidazolidinone ring with pent-3-ynol as starting material.
e-mail: marquet@ccr.jussieu.fr
Present address: Institut de Recherche Jouveinal/Parke-Davis, Departe-
†
´
ment Biochimie-Enzymologie, 3-9 rue de la Loge, BP100, 94265
FRESNES Cedex, France.
0040–4020/00/$ - see front matter ᭧ 1999 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(99)01022-4

286
F. Moreau et al. / Tetrahedron 56 (2000) 285–293
Scheme 2. 2a and 7a, Z isomer; 2b and 7b, E isomer.
It was rather long with one non-stereospecific step leading
to a 90:10 mixture of stereoisomers, which were extremely
difficult to separate.5
The first step consisted of the lactone ring opening (Scheme 3).
Under acidic conditions, (ϩ)8 was recovered unchanged,
very likely because of an unfavorable equilibrium position.
Basic conditions led to an opened product 9. Due to the
possibility of epimerization at C7, we checked the stereo-
A more attractive starting material was the enantiomerically
pure lactone (ϩ)8, an intermediate in the Hoffmann–
La Roche total synthesis of biotin,6 which we have success-
fully used in the past for the synthesis of 9-mercaptodethio-
biotin.7 However, our first attempts to introduce an
unsaturated side chain failed because of the participation
of the hydroxyl group generated at C-9. It was therefore
reasonable to deoxygenate this position prior to any side-
chain construction.
3
chemistry of 9. The JH7–H8 value, which could be an
argument, is ambiguous: 4.5 Hz for trans DTB, 8 Hz for
cis DTB, 6.1 Hz for 9. The ambiguity was removed by
X-ray analysis of 10, obtained from 9 through a Barton-
type deoxygenation followed by LiAlH4 reduction. The
3D structure clearly establishes the trans relationship of
the two substituents (Fig. 1).
Scheme 3. (a) MeONa, MeOH, reflux; HCl, 69%; (b) 4-FPTCl, pyridine, CH2Cl2, RT, 94%; (c) Bu3SnH, AIBN, PhCH3, reflux; (d) LiAlH4, THF, 0ЊC, Hϩ,
steps c, d 47%.
Figure 1. X-Ray crystal structure determination of compound 10.

F. Moreau et al. / Tetrahedron 56 (2000) 285–293
287
Scheme 4. (a) LiAlH4, THF, 0ЊC, 98%; (b) 4-FPTCl, pyridine, CH2Cl2, Ϫ30ЊC, 49%; (c) Bu3SnH, AIBN, PhCH3, reflux, 57%; (d) MsCl, NEt3, CH2Cl2, RT,
80%; (e) KCN, 70ЊC, 89%; (f) DIBAH, PhCH3, Ϫ70ЊC, 70%; (g) (COCl)2, DMSO, NEt3, CH2Cl2, Ϫ78ЊC; (h) Ph3Pϩ(CH2)4OBn, BrϪ, nBuLi, THF, Ϫ60ЊC,
steps g, h: 54%, Z/Eˆ90:10; (i) Na, NH3, EtOH, 40%, HPLC; (j) Ph3PBr2, pyridine, CH3CN, RT; (k) KCN, DMSO, RT; (l) NaOH 1N, reflux, steps j, k, l: 41%.
The cis stereochemistry on the imidazolidinone ring being a
priori crucial for enzyme recognition, we decided to sacrifice
the chirality and to fully reduce the lactone into the cis diol 11
(Scheme 4).
the case of 14, the strong brown color that appeared after the
addition of the second butyl lithium equivalent indicated
that the betaine ylide was formed but the composition did
not change for unknown reasons. Either they did not equi-
librate or the proportions correspond to the equilibrium
mixture. 15 was debenzylated and its major component,
the Z isomer 7a, was then purified by crystallization. The
E isomer was obtained by HPLC purification of the residue.
The major product 7a was finally converted into Z-5,6-dehy-
drodethiobiotin 5a through the reaction sequence already
reported.5
A Barton-type deoxygenation8 gave the key intermediate
12. 12 was transformed in three steps, mesylation, cyanation
and DIBAH reduction into the aldehyde 13 which was
converted into the two isomeric 4,5-dehydrodethiobiotins
2 as already described.5
The regioisomeric 5,6-dehydrodethiobiotin 5a was also
obtained from 12. 12 was oxidized to the aldehyde 14
which was submitted to a Wittig reaction. The best yield
of 15 was obtained in THF at Ϫ70ЊC. The yield was lower
(Ϸ30%) in THF or THF/HMPT at 0ЊC. Under all con-
ditions, a Z/E ratio of 90:10 was observed. Surprisingly,
this ratio did not change when we used the Schlosser modi-
fication9 under the conditions used for the homologous
aldehyde 13.5 In this latter case, the Schlosser reaction
gave, as expected, a higher proportion of the E isomer.5 In
Synthesis of the epoxides 3a and 3b
The Z-4,5-dehydrodethiobiotin 2a was readily epoxidized
with mCPBA into a 50:50 mixture of the cis epoxides 3a
(Scheme 5). Under the same conditions, a very slow
reaction was observed with the trans isomer 2b and if larger
amounts of mCPBA were used, only degradation products
were obtained.
Scheme 5. (a) mCPBA, CH2Cl2, RT, 90%; (b) H2O2, MTO, H2O, RT, 74%.

288
F. Moreau et al. / Tetrahedron 56 (2000) 285–293
Scheme 6.
Reaction with dimethyloxirane, a more powerful reagent10
also more suitable for the oxidation of small amounts of
radioactive compounds, was also investigated. With 2b a
complete degradation was observed and with 2a the yield
in epoxides 3a was lower that with mCPBA.
Preparation of the [14C] labeled compounds
The synthesis of [14C]2a and [14C]2b has already been
described.5 [14C]5a was obtained from 7a following the
same procedure.
Finally, the best results were obtained with H2O2 in the
presence of methyltrioxorhenium,11 which allowed the
epoxidation of 2b in 74% yield leading also to a 50:50
mixture of the two epoxides 3b.
The epoxides [14C]3a and [14C]3b were obtained by oxi-
dation of [14C]2a and [14C]2b, respectively, with H2O2 in
the presence of MTO. [14C]3a1 and [14C]3a2 were separated
by HPLC. [14C]3b1 and [14C]3b2 could not be separated and
were recovered as a mixture.
The cis epoxides 3a1 and 3a2 were separated by HPLC.
Interestingly, the NMR spectrum of each of them, recorded
in methanol and in DMSO as well, revealed the presence of
two forms in equilibrium. The first one, unique just after
dissolution of the dried compound is slowly transformed
into a second one until a stable 60:40 composition is
obtained. The phenomenon is reversible since after drying
this mixture again, only the first form is observed just after
dissolution. Relying on the strong deshielding of protons e
and especially f in that form (Scheme 6) we assume that we
are observing an equilibrium between an hydrogen-bonded
‘closed’ form (C) and an ‘opened’ form (O).
Preliminary biological evaluation
The ethylenic compounds 2a, 2b, 5a and the epoxides 3a1,
3a2 and 3b1ϩ3b2 were evaluated under different conditions
with purified E. coli biotin synthase, in the in vitro system
described by our group containing besides the enzyme,
S-Adenosyl Methionine, an electron transfer system
(NADPH, flavodoxin, flavodoxin reductase) and different
cofactors.2a
They were first tested as substrates: the supernatants of the
incubation media were assayed for their growth promoting
activity of Lactobacillus plantarum, a biotin auxotroph
commonly used for biotin microbiological determination.14
The epoxides 3a and 3b did not promote the growth of L.
plantarum. On the other hand, the ethylenic compounds did,
with the order of efficiency 2aϾ5aq2b.
The trans epoxides 3b1 and 3b2 could not be separated by
HPLC. The NMR spectrum of the mixture shows that each
of them is present under only one form in which d Hf is
around 4.5 ppm. They probably exist exclusively in the (C)
form.
Attempts to synthesize the cyclopropyl derivatives 4 were
performed on the Z protected derivative because the solu-
bility of the deprotected form 2a was incompatible with the
solvents required for the cyclopropanation (Fig. 2).
TLC was then performed on these supernatants. Revelation
of the plates with L. plantarum showed that the compounds
which promoted the growth of the microorganism had the
same RF as biotin.
CH2I2/Zn12 at reflux of dimethoxyethane (DME) or
dichloroethane (DCE) with sonication or ClCH2I/Et2Zn13
in DCE or benzene at room temperature left the starting
material unchanged. Steric hindrance by the N-benzyl
group is probably responsible for this lack of reactivity.
The analogs may have been transformed into biotin analogs
with unsaturated side-chains, supporting the growth of
L. plantarum.
When the radioactive compounds were incubated under the
same conditions and the supernatant analyzed by TLC, no
new radioactive compounds could be detected. Due to the
very low conversion yield, the amount of radioactivity
expected for these compounds lies in the range of the noise.
The in vitro system that we are using is presently not
catalytic. With the natural substrate, DTB, less than one
mole of biotin is produced per mole of enzyme. With the
analogs, which are worse substrates, we expect very tiny
amounts and we cannot consider determination of the
Figure 2. Protected 2a.

F. Moreau et al. / Tetrahedron 56 (2000) 285–293
289
structures of the products. This should become possible
once a more efficient in vitro system is available.
JGemˆ15.2 Hz, 2NCHHPh); 4.32 (dd, 1H, Jˆ4.1 Hz,
JGemˆ11.7 Hz, CHH); 4.54 (dd, 1H, Jˆ3.6 Hz,
JGemˆ11.7 Hz, CHH); 6.99–7.03 (m, 2H, 2 CHCO);
7.07–7.14 (m, 2H, 2CHCF); 7.23–7.36 (m, 10H, 2
NCH2Ph). 13C NMR: d 42.26–46.61 (2NCH2Ph); 52.51
(COOCH3); 54.27 (CN); 56.56 (CN); 71.62 (CH2); 116.26
Experimental
2
3
(d, JC–Fˆ24.42 Hz, CCF); 123.20 (d, JC–Fˆ8.13 Hz,
CCCF); 127.57–128.67; 135.92; 136.39 (2CH2Ph); 149.07
(OCCCCF); 159.08 (CyO); 160.62 (d, JC–Fˆ246 Hz, CF),
170.16 (COOCH3); 194.54 (CyS). Anal. calcd for
C27H25FN2O5S: C, 63.76; H, 4.97; N, 5.51. Found C,
63.59; H, 4.98; N, 5.53.
All reactions using nonaqueous reagents were run under a
dry argon atmosphere. Organic layers were dried on mag-
nesium sulfate (MgSO4). Flash chromatographies were
performed on Kieselgel 60 Merck (230–400 mesh). Reac-
tion progress was monitored by analytical TLC on silica gel
60 F-254 from Merck. Visualization of TLC was done by
phosphomolybdic acid or by UV light or paradimethyl-
1
amino cinnamaldehyde for the amido groups. H and 13C
The FTP derivative (4.25 g, 8.37 mmol), AIBN (687 mg,
4.18 mmol) and tributyltin hydride (Bu3SnH (3.6 mL,
13.4 mmol) were dissolved in anhydrous toluene (85 mL)
under argon. The solution was deoxygenated with argon
during 40 min and heated under reflux for 1.5 h with contin-
uous bubbling of argon. Bu3SnH (0.9 mL, 3.35 mmol) and
AIBN (172 mg, 1.04 mmol) were added and reflux was
continued for 30 min. After evaporation and purification
on a flash silica gel column (C6H12:CH3COOC2H5 9:1 to
7:3) 2.24 g of dehydroxylated compound were obtained
(containing stannous impurities) and dissolved at 0ЊC in
anhydrous THF (15 mL), 1M LiAlH4 in THF (3.6 mL,
3.6 mmol) was added. After 10 min stirring, the mixture
was hydrolyzed with 1N HCl and the product extracted by
CHCl3. The organic phase was dried and concentrated under
vacuum. A flash silica gel column (C6H12:CH3COOC2H5
7:3 to 2:8) yielded 10 (1.23 g, 47% from FTP derivative)
as a white powder which crystallized from C5H12:CH2Cl2.
NMR spectra were recorded on a Bruker AR X 400 (400 and
100 MHz, respectively). Solutions in chloroform-d (CDCl3)
were used unless otherwise stated, coupling constants have
been obtained by spin decoupling. Mass spectra (CI) were
carried out on a Nermag R10-10C spectrometer using NH3
as reactant gas. Melting points were measured on a Kofler
apparatus. Elemental analyses were performed at the
´
Service Regional de Microanalyse (SIAR-Jussieu). All
chemicals were purchased from Aldrich or Acros. [14C]
KCN was from NEN. Analytical HPLC was performed
with a Perkin Elmer pump, a PE 785 A absorbance detector
and a PE Model 1022 recorder. Counting was performed
with a LKB 1214 Rackbeta scintillation counter, using
Aquasol-2 (Packard) or Optiphase ‘Hisafe’ (Wallac) as scin-
tillation liquid. Radioactive TLC plates were scanned with
an automatic TLC linear analyzer Berthold LB 2821.
1
mp 111–113ЊC. H NMR: d 1.6 (d, 3H, Jˆ6.1 Hz, CH3);
trans-1,3-Dibenzyl-5-hydroxymethyl-4-methoxycarbonyl-
imidazolidin-2-one 9. At room temperature, 8 (15 g,
47 mmol) and then sodium methoxide (0,38 g, 7 mmol)
were successively dissolved in anhydrous methanol
(30 mL). The mixture was refluxed for 4 h. After neutrali-
zation with 12N HCl, evaporation and purification on a
silica gel column (C6H12:CH3CO2C2H5, 9:1 to 7:3), 9 was
obtained as a white solid (11.5 g, 69%) which crystallized
from C6H12:CH3CO2C2H5, mp: 105–107ЊC. Starting
1.78–1.82 (m, 1H, –OH); 2.96–3.00 (m, 1H, CHN);
3.32–3.37 (m, 1H, Jˆ7.5 Hz, CHN); 3.30–3.42 (m, 1H,
Jˆ11.7 Hz, CHHOH); 3.54–3.59 (m, 1H, Jˆ11.7 Hz,
CHHOH); 4.14; 4.33; 4.62; 4.75 (4d, 4H, JGemˆ15.3 Hz;
2NCHHPh); 7.22–7.35 (m, 10H, arom.). 13C NMR: d
18.42 (CH3); 45.56; 46.59 (2NCH2Ph); 50.51 (CHN);
60.52 (CH2OH); 62.4 (CHN); 127.29–128.77; 137.26–
137.70 (arom.); 160.76 (CyO). Anal. calcd for
C19H22N2O2: C, 73.56; H, 7.09; N, 9.03. Found: C, 73.46;
H, 7.22; N, 9.09.
1
material 8 was also recovered 2.7 g (25%). H NMR: d
1.83 (s, 1H, OH); 3.43–3.48 (m, 1H, CHH); 3.50–3.53
(m, 1H, CHN); 3.58–3.62 (m, 1H, CHH); 3.60 (s, 3H,
COOCH3); 3.91 (d, 1H, Jˆ6.1 Hz, CHN); 4.22; 4.32;
4.65; 4.93 (4d, 4H, JGemˆ15.2 Hz, 2NCHHPh); 7.23–7.35
(m, 10H, arom.). 13C NMR: d(46.49 et 46.89 (2NCH2Ph);
52.35 (COOCH3); 56.52 (CHN); 57.97 (CHN); 60.90 (CH2);
127.57–137.16 (arom.); 159.95 (CyO); 170.91 (COOCH3).
Anal. calcd for C20H22N2O4: C, 67.82; H, 6.21; N, 7.91.
Found: C, 67.70; H, 6.25; N, 7.91.
cis-1,3-Dibenzyl-4,5-dihydroxymethyl-imidazolidin-2-
one 11. 8 (40 g, 124 mmol) was dissolved in THF (250 mL)
and 1 M LiAlH4 in THF (93 mL, 93 mmol) was added at
0ЊC. The mixture was stirred 30 min at 0ЊC and hydrolyzed
with 1N HCl (300 mL). The product was extracted with
CH2Cl2 ꢀ3×100 mL† and the organic layers washed with
brine, dried (MgSO4) and evaporated under vacuum.
Crystallization from CH2Cl2 gave 11 (39.1 g, 98%) as
1
white crystals. mp: 131–133ЊC. H NMR: d 3.49–3.56
trans-1,3-Dibenzyl-4-hydroxymethyl-5-methyl-imidazo-
lidin-2-one 10. 9 (7.68 g, 21.7 mmol), 4-fluorophenylchloro-
thionoformate (4-FPTCl) (3.35 mL, 23.9 mmol) and
pyridine (2.28 mL, 28.2 mmol) were dissolved in anhydrous
CH2Cl2 (20 mL). After 15 min stirring at room temperature
CH2Cl2 (300 mL) was added. The organic solvents were
washed with HCl ꢀ2×20 mL† then with saturated NaCl
(10 mL) and evaporated under vacuum. After crystallization
(C5H12:CH3COOC2H5) the FTP derivative was obtained as a
white solid (10.4 g, 94%). mp 108–110ЊC. 1H NMR: d 3.65
(s, 3H, COOCH3); 3.80–3.83 (m, 1H, CHN); 3.85 (d, 1H,
Jˆ5.6 Hz, CHN); 4.18; 4.24; 4.92; 5.08 (4d, 4H,
(m, 2H, 2CHN); 3.73–3.87 (m, 6H, 2CH2OH); 4.15; 4.91
(2d, 4H, JGemˆ15.2 Hz, 2NCHHPh); 7.28–7.37 (m, 10H,
arom.). 13C NMR: d 45.60 (2NCH2Ph); 56.45 (2CHN);
57.71 (2CH2OH); 127.59; 127.90; 128.75; 137.06 (arom.)
161.40 (CyO). Anal. calcd for C19H22N2O3: C, 69.92; H,
6.79; N, 8.58. Found: C, 69.75; H, 6.78; N, 8.49.
cis-1,3-Dibenzyl-4-hydroxymethyl-5-methyl imidazolidin-
2-one 12. In a solution of 11 (11.8 g, 34 mmol) in anhydrous
CH2Cl2 (200 mL) were added anhydrous pyridine (4.23 mL,
52.4 mmol) and then at Ϫ30ЊC, under vigorous stirring and

290
F. Moreau et al. / Tetrahedron 56 (2000) 285–293
dropwise, 4-FPTCl (5 g, 26 mmol) dissolved in anhydrous
CH2Cl2 (100 mL). After 30 min stirring at Ϫ30ЊC, solvents
were evaporated under vacuum and the residue purified on a
flash silica gel column (C6H12:CH3COOC2H5 1:0 to 1:1 and
then ethanol) the FTP derivative was obtained as a white
solid (8.03 g, 49%) which crystallized from CH2Cl2. mp:
Jˆ6.1 Hz, CH3); 2.43–2.54 (m, 2H, CH2CN); 3.58–3.68
(m, 2H, 2CHN); 4.12: 4.24; 4.88; 4.91 (4d, 4H,
JGemˆ15.2 Hz, 2NCHHPh); 7.29–7.40 (m, 10H, arom.).
13C NMR: d 12.57 (CH3); 16.50 (CH2CN); 45.27; 45.88
(2NCH2Ph); 51.18; 53.10 (2CHN); 116.92 (CN); 127.51;
128.81; 136.57; 136.75 (arom.); 159.87 (CyO). Anal.
calcd for C20H21N3O: C, 75.21; H, 6.63; N, 13.15. Found:
C, 75.04; H, 6.68; N, 13.18.
1
104–106ЊC H NMR: d 2.51 (s, 1H, OH); 3.59–3.63 (m,
1H, Jˆ9.1 Hz, CHN): 3.76–3.84 (2dd, 2H, JGemˆ12.7 Hz,
Jˆ2.5 Hz, Jˆ5.1 Hz, CHHOH); 3.87–3.92 (m, 1H, CHN);
4.77; 4.91 (2dd, 2H, JGemˆ11.7 Hz, Jˆ6.1 Hz, Jˆ4.6 Hz,
CHHOCS); 4.30; 4.31; 4.80; 4.96 (4d, 4H, JGemˆ15.7 Hz,
2NCHHPh); 7.01–7.05 (m, 2H, 2CHCO); 7.08–7.14 (m,
2H, 2CHCF); 7.28–7.40 (m, 10H, 2NCH2Ph). 13C
NMR: d 46.10 and 46.57 (2NCH2Ph); 53.68 (CHN);
56.31 (CHN); 56.80 (CH2OH); 71.60 (CH2OCS);
The cyano derivative (1.15 g, 3.60 mmol) was dissolved in
anhydrous toluene (15 mL) and at Ϫ70ЊC, 1.5 M diisobutyl-
aluminium hydride (DIBAH) in toluene (6 mL, 9 mmol)
was added. After 30 min stirring at Ϫ70ЊC, the mixture
was hydrolyzed by 1N HCl (0.5 mL) and was allowed to
warm to room temperature, then 1N HCl (75 mL) was
added. The product was extracted with CH2Cl2. The organic
phase was washed with brine, dried and evaporated under
vacuum. Purification performed on a flash silica gel column
(C6H12:CH3COOC2H5, 7:3 to 1:1) yielded 13 as a colorless
2
3
116.28 (d, JC–Fˆ22.4 Hz, CCF); 123.29 (d, JC–F
8 Hz, CCCF); 127.51–128.77; 136.96; 137.19
(2CH2Ph); 149.09 (O–C arom.); 160.66 (d,
ˆ
JC–Fˆ246.2 Hz, C–F); 161.00 (CyO); 194.54 (CyS).
Anal. calcd for C26H25FN2O4S: C, 64.99; H, 5.24; N,
5.83. Found: C, 64.85; H, 5.32; N, 5.77.
1
oil (800 mg, 70%). H NMR: d 1.00 (d, 3H, Jˆ6.6 Hz,
CH3); 2,64 (ddd, 1H, Jˆ8.2 Hz, JGemˆ18.3 Hz, Jˆ0.6 Hz,
CHHCHO); 2.70 (ddd, 1H, Jˆ5.1 Hz, Jˆ18.3 Hz,
JGemˆ0.6 Hz, CHHCHO); 3.64 (dq, 1H, Jˆ7.6 Hz,
Jˆ6.6 Hz, CHN); 3.88–3.94 (m, 1H, CHN); 4.05; 4.23;
4.71; 4.91 (4d, 4H, JGemˆ15.2, 2NCHHPh); 7.27–7.37 (m,
10H, arom.); 9.66 (s, 1H, CHO). 13C NMR: d 12.61 (CH3);
42.17 (CH2CHO); 45.12; 46.00 (2NCH2Ph); 51.03 (CHN);
52.13 (CHN); 127.35–128.62; 137.18; 137.27 (arom.);
160.33 (CyO); 199.42 (CHO).
FTP derivative (4.21 g, 8.77 mmol), AIBN (720 mg,
4.38 mmol) and freshly distilled Bu3SnH (3.77 mL,
14 mmol) were dissolved in anhydrous toluene (150 mL).
Argon was bubbled in the mixture during 45 min. After
heating with reflux for 1 h with continuous bubbling of
argon, toluene was evaporated under vacuum and the
residue was purified on flash silica gel column
(C6H12:CH3COOC2H5 1:0 to 4:6). 12 (1.51 g, 57%) was
obtained as a colorless oil. 1H NMR: d(1.21 (d, 3H,
Jˆ6.6 Hz, CH3); 2.02 (s, 1H, OH); 3.39–3.43 (m, 1H,
CHN); 3.55–3.62 (m, 1H, Jˆ8.7 Hz, CHN); 3.69; 3.74
(2dd, 2 H, JGemˆ12.2 Hz, Jˆ4.1 Hz, Jˆ5.1 Hz, CHHOH);
4.09; 4.30; 4.78; 4.86 (4d, 4H, JGemˆ15.2 Hz, 2NCHHPh);
7.27–7.37 (m, 10H, arom.). 13C NMR: d 12.70 (CH3); 45–
25; 46.30 (2NCH2Ph); 51.02 (CHN); 57.34 (CHN); 59.39
(CH2OH); 127.31; 127.43; 127.84; 128.00; 128.52; 128.64;
137.28; 137.84 (arom.); 161.12 (CyO). Anal. calcd
C19H22N2O2: C, 73.52; H, 7.14; N, 9.02. Found: C, 73.44;
H, 7.19; N, 9.07.
cis-1,3-Dibenzyl-4-formyl-5-methyl imidazolidin-2-one 14.
DMSO (273 mL, 385 mmol) was dissolved in anhydrous
CH2Cl2 (10 mL) at Ϫ78ЊC and oxalylchloride (169 mL,
1.93 mmol) was added dropwise. After 10 min stirring at
Ϫ78ЊC, 12 (500 mg, 1.61 mmol) in CH2Cl2 (5 mL) was
added slowly. The mixture was stirred 30 min, then NEt3
(673 mL, 4.84 mmol) was added and after 40 min stirring at
Ϫ78ЊC, H2O (2 mL). The mixture was extracted with
CH2Cl2 ꢀ4×15 mL†: The organic phase was washed with
brine, dried, evaporated under vacuum. The crude aldehyde
14 was obtained as a colorless oil (496 mg) and was not
1
purified. H NMR: d 1.15 (d, 3H, Jˆ6.1 Hz, CH3); 3.70–
cis-1,3-Dibenzyl-5-methyl-4-(2-oxyethyl) imidazolidin-2-
one 13. 12 (1.97 g, 6.35 mmol) was dissolved into anhy-
drous CH2Cl2 (20 mL), mesyl chloride (0.64 g, 8.2 mmol)
and NEt3 (1.32 mL; 9.52 mmol) were added. After stirring
15 min at room temperature, 1N HCl (10 mL) was added,
the product was extracted with CH2Cl2. The organic phase
was washed with brine, dried (MgSO4) and evaporated.
Purification was performed on a flash silica gel column
(C6H12:CH3COOC2H5, 7:3 to 1:1) giving the mesylate
3.77 (m, 2H, Jˆ9.1 Hz, 2CHN); 4.11; 4.31; 4.82; 4.88 (4d,
4H JGemˆ15.2 Hz, 2NCHHPh); 7.26–7.37 (m, 10, arom.);
9.51 (d, 1H, Jˆ4.1 Hz, CHO). 13C NMR: d 14.07 (CH3);
45.47; 47.09 (2NCH2Ph); 50.87; 64.20 (2CHN); 127.57–
128.75; 136.37; 136.47 (arom.); 160.15 (CyO); 200.59
(CHO).
cis-4-[(Z and E)-5-Hydroxypent-1-enyl]-5-methyl imida-
zolidin-2-one 7a, 7b. 1.6N n-butyllithium in hexane
(5.63 mL, 9 mmol) was added at 0ЊC to a solution of
Ph3Pϩ (CH2)4OBn, BrϪ15 (4.8 g, 9.75 mmol, dried for
30 min at 80ЊC under vacuum) in anhydrous THF
(15 mL). The brick red solution was stirred 10 min at 0ЊC
and cooled at Ϫ60ЊC. 14 (2 g, 6.5 mmol) dissolved in anhy-
drous THF (15 mL) was then added, followed after 25 min
stirring by 1.6N n-butyllithium (4.38 mL, 7 mmol). The
mixture was allowed to warm to Ϫ30ЊC, kept at Ϫ30ЊC
for 30 min and then cooled at Ϫ60ЊC, treated with absolute
ethanol (550 mL, 9.5 mmol) with strong agitation, then
warmed to room temperature. t-BuOK (1.43 g, 12.7 mmol)
was added and after 2 h stirring at room temperature,
(2 g, 80%) as
a
white solid recrystallized from
CH2Cl2:C5H12, mp: 124–126ЊC. To a solution of the
mesylate (1.74 g, 4.48 mmol) in anhydrous DMSO
(10 mL) was added KCN (583 mg, 8.96 mmol). After heat-
ing at 70ЊC during 20 h, DMSO was evaporated under
vacuum at 50ЊC. The crude product was dissolved in
CH2Cl2 (300 mL) and CH2Cl2 was washed successively
with water (15 mL) and brine (15 mL) then dried and evapo-
rated under vacuum. Purification on a flash silica gel column
(C6H12:CH3COOC2H5, 8:2 to 1:1) yielded the cyano deriva-
tive as a white solid (1.28 g, 80%) recrystallized from
1
CH2Cl2:C5H12. mp 125–127ЊC. H NMR: d 1.24 (d, 3H,

F. Moreau et al. / Tetrahedron 56 (2000) 285–293
291
CH2Cl2 (200 mL) was added. The organic phase was
successively washed with 1N HCl (60 mL) and brine and
dried. The solvents were evaporated under vacuum and the
residue purified on silica gel column (C6H12:CH3COOC2H5,
1:0 to 1:1) yielding 15 (1.59 g, 54%) as a mixture of Z and E
purified on a C18 reverse phase column (CH3CN:H2O;
1:9). After evaporation 5a was obtained (69 mg, 41%
from 7a) as a white solid crystallized from H2O.
mp: 145–147ЊC. 1H NMR (CD3OD): d 1.07 (d, 3H,
Jˆ6.6 Hz, CH3); 1.64–1.71 (m, 2H, CH2CH2COOH);
2.07–2.22 (m, 2H, CH2(CH2)2COOH); 2.30 (t, 2H,
Jˆ7.1 Hz, CH2COOH); 3.85–3.92 (m, 1H, Jˆ8.3 Hz,
CHN); 4.57–4.62 (m, 1H, CHN); 5.44–5.49 (m, 1H,
CHyCH–CH2); 5.63 (dt, 1H, Jˆ11.2 Hz, Jˆ7.6 Hz,
CHyCHCH2). 13C NMR (CD3OD): d 16.67 (CH3);
25.85 (CH2CH2COOH); 27.71 (CH2(CH2)2COOH);
34.12 (CH2COOH); 53.10 (CHN); 54.32 (CHN);
128.19 (CHyCH); 134.35 (CHyCH); 166.14 (CyO);
177.28 (COOH). Anal. calcd for C10H16N2O3: C,
56.59; H, 7.60; N, 13.20. Found: C, 56.66; H, 7.46;
N, 13.24.
1
isomers (9:1 from H NMR).
In a three-necked round-bottom flask, at Ϫ78ЊC, NH3
(70 mL) was condensed and absolute ethanol (3 mL) was
added. The solution was allowed to warm to Ϫ30ЊC. Then
15 (1.59 g, 3.6 mmol) in THF (8 mL) was added, followed
by small sodium pieces (850 mg, 31.5 mat), waiting for
decolorization between the two additions. The addition
was stopped when the blue color persisted for 1 min. NH3
was then evaporated at room temperature. After addition of
water and then 12N HCl up to neutralization, the solvents
were evaporated under vacuum at 50ЊC. The residue was
purified by C18 reverse phase chromatography
(CH3CN:H2O; 0:100 to 20:80) yielding 7 (260 mg, 40%)
as a colorless oil (mixture of Z and E isomers). 7a was
obtained as white crystals by crystallization from
MeOH:CH3CN and 7b as an oil after HPLC of the residue
(Lichrosorb C18 reverse phase column, H2O:TFA, 0.1%;
CH3CN, 92:8).
cis-4-(5-Carboxy-cis-2,3-epoxy-pentyl)-5-methyl imidazo-
lidin-2-one 3a. 2.50 mL of a 70% metachloroperbenzoic
acid solution in CH2Cl2 (244 mg, 0.99 mmol) was added
to 2a (70 mg, 0.33 mmol) at Ϫ70ЊC. After 3 h stirring,
CH2Cl2 was evaporated, the residue dissolved in H2O
(10 mL) and purified on a C18 silica gel reverse phase
column (CH2CN:H2O; 1:9). 3a (67 mg, 90%) was obtained
as a mixture of two diastereoisomers, which were separated
by HPLC (Lichrosorb C18, H2O, TFA 0.1%:CH3CN,
97.6:2.4).
7a: mp: 115–117ЊC. 1H NMR (CD3OD): 1.07 (d, 3H,
Jˆ6.1 Hz, CH3); 1.56–1.63 (m, 2H, CH2CH2OH); 2.11–
2.23 (m, 2H, CH2CH2CH2OH); 3.55 (t, 2H, Jˆ6.6 Hz,
CH2OH); 3.86–3.92 (m, 1H, Jˆ8.2 Hz, CHN); 4.60–4.64
(m, 1H, CHN); 5.42–5.47 (m, 1H, Jˆ10.7 Hz,
CHCHyCH); 5.65 (dt, 1H, Jˆ10.7 Hz, Jˆ7.6 Hz,
3a1(C): white solid, mp: 217–219ЊC. 1H NMR (CD3OD: d
1.13 (d, 3H, Jˆ6.6 Hz, CH3); 1.56–1.75 (m, 2H, NCHCH2);
2.08–2.17 (m, 1H, CHHCH2COOH); 2.25–2.34 (m, 1H,
CHHCH2COOH); 2.53–2.58 (m, 2H, CHN); 3.65–3.71
(m, 1H, CH2CHO); 3.85–3.92 (dq, 1H, Jˆ8.2 Hz,
Jˆ6.6 Hz, CHN); 3.94–4.00 (m, 1H, CHN); 4.46–4.50
(dt, 1H, Jˆ4.1 Hz, Jˆ7.6 Hz, OCHCH2). 13C NMR
(CD3OD): d 15.77 (CH3); 24.74 (CH2CH2COOH); 29.40
(CH2COOH); 33.67 (NCHCH2); 52.75 (CHN); 55.78
(CHN): 72.93 (CH2CO): 84.51 (OCCH2); 165.95 (CyO);
180.19 (COOH). Anal. calcd for C10H16N2O4: C, 52.62; H,
7.06; N, 12.27. Found: C, 52.58; H, 7.04; N, 12.26. I.C.
(NH3) MHϩˆ229.
3a1(O): 1H NMR (CD3OD): d 1.13 (d, 3H, Jˆ6.6 Hz, CH3);
1.57–1.88 (m, 4H, NCH–CH2, CH2CH2COOH); 2.39–2.58
(m, 2H, CH2COOH); 3.39–3.43 (m, 1H, O–CH–CH2);
3.48–3.54 (m, 1H, CH2CHO); 3.68–3.93 (m, 1H, CHN);
3.95–4.00 (m, 1H, CHN). 13C NMR (CD3OD): d 15.92
(CH3); 29.20 (CH2CH2COOH); ϩ31.39 (CH2COOH)
34.26 (CH2–CO); 53.09 (CHN); 54.36 (CHN); 72.03
(CH2CO); 74.78 (OCCH2); 166.04 (CyO); 180.26
(COOH).
CHyCHCH2). 13C NMR:
d
16.67 (CH3); 24.8
(CH2CH2CH2OH); 33.33 (CH2CH2OH); 53.10 (CHN);
54.32 (CHN); 62.04 (CH2OH); 127.74 (CHCHyCH);
134.72 (CHyCHCH2); 166.16 (CyO). Anal. calcd for
C9H16N2O2: C, 58.67; H, 8.75; N, 15.20. Found: C, 58.78;
H, 8.77; N, 15.16.
1
7b: H NMR (CD3OD): d 1.06 (d, 3H, Jˆ6.1 Hz, CH3);
1.58–1.64 (m, 2H, CH2CH2OH); 2.12–2.17 (m, 2H,
CH2CH2CH2OH); 3.55 (t, 2H, Jˆ6.6 Hz, CH2OH);
3.84–3.91 (m, 1H, Jˆ8.2 Hz, CHN); 4.17–4.21 (m,
1H, CHN); 5.49 (dd, 1H, Jˆ15.3 Hz, Jˆ8.1 Hz,
CHCHyCH); 5.71 (dt, 1H, Jˆ15.3 Hz, Jˆ6.6 Hz,
CHyCHCH2). 13C NMR:
d
16.87 (CH3); 29.57
(CH2CH2CH2OH); 33.06 (CH2CH2OH); 53.31 (CHN);
59.89 (CHN); 62.22 (CH2OH); 127.80 (CHCHyCH);
134.92 (CHyCHCH2); 165.98 (CyO).
cis-4-(Z-5-Carboxypent-1-enyl)-5-methyl imidazolidin-
2-one 5a. At 0ЊC, 7a (143 mg, 0.78 mmol) was dissolved
in anhydrous acetonitrile (3 mL) then anhydrous pyridine
(250 mL, 3.10 mmol) and triphenylphosphine bromide
(1.15 g, 2.73 mmol) were added. After stirring 1 h at room
temperature, quick filtration on silica gel and evaporation of
the filtrate yielded crude compound 16a (decomposition at
room temperature and on silica gel, conservation at Ϫ20ЊC),
which was directly dissolved in DMSO (3 mL). KCN
(152 mg, 2.34 mmol) was added; after sonication at room
temperature during 2 h, DMSO was evaporated under
vacuum at 50ЊC. The residue (nitrile 17a) was dissolved
in 1N NaOH (6 mL) and heated under reflux 90 min. Then
12N HCl was added to obtain pHϷ1: The mixture was
3a2(C): 1H NMR (CD3OD): d 1.12 (d, 3H, Jˆ6.6 Hz, CH3);
1.68–1.80 (m, 2H, NCHCH2); 2.08–2.18 (m, 1H,
CHHCH2COOH); 2.23–2.33 (m, 1H, CHHCH2COOH);
2.52–2.57 (m, 2H, CH2COOH); 3.66–3.70 (m, 1H,
CH2CHO); 3.82–3.88 (m, 1H, CHN); 3.92 (dt, 1H,
Jˆ8.2 Hz, CHN); 4.51 (m, 1H, Jˆ3.6 Hz, Jˆ7.1 Hz,
OCHCH2. 13C NMR (CD3OD): d 15.78 (CH3); 24.75
(CH2CH2COOH); 29.40 (CH2COOH); 33.69 (NCHCH2);
52.76 (CHN); 57.79 (CHN); 72.94 (CH2CO); 84.51
(OCCH2); 165.96 (CyO); 180.19 (COOH).

292
F. Moreau et al. / Tetrahedron 56 (2000) 285–293
3a2(O): 1H NMR (CD3OD); d 1.11 (d, 3H, Jˆ6.1 Hz, CH3);
1.59–1.87 (m, 4H, CH2CHO, CH2CH2COOH); 2.38–2.52
(m, 2H, CH2COOH); 3.41 (dt, 1H, Jˆ3.6 Hz, Jˆ10.2 Hz,
OCHCH2); 3.52 (dt, 1H, Jˆ3.6 Hz, Jˆ9.7 Hz, CH2CHO);
3.81–3.97 (m, 1H, 2CHN).
synthase, 50–500 mM substrates. 2 mM NADPH,
0.2 mM S-Adenosyl-l-methionine, 0.5 mM l-cysteine,
5 mM dithiothreitol, 2 mM 1,6-fructose biphosphate,
10 mM KCl, 40 mM Tris–HCl buffer pHˆ8: Mixture B
consisted of 5 mM flavodoxin, 1 M flavodoxin reductase,
1 mM Fe(NH4)2(SO4)2 in H2O. The final volume of the
assay was 100 mL. The reaction was stopped by precipi-
tation with trichloroacetic acid. Precipitated proteins were
removed by centrifugation and the amount of biotin formed
was determined in the supernatant by the paper disc plate
method using L. plantarum.15
cis-4-(5-Carboxy-trans-2,3-epoxy-pentyl)-5-methylimidazo-
lidin-2-one 3b. To a solution of 2b (20 mg, 0.094 mmol) in
H2O (2 mL) were added methyltrioxorhenium (MTO, 7 mg,
0.028 mmol) and a 35% H2O2 solution into H2O (16 mL,
1.88 mmol). After 15 min sonication and 3 h stirring at
room temperature, H2O2 (16 mL, 1.88 mmol) and MTO
(7 mg, 0.028 mmol) were added once more. The mixture
was stirred for 30 min and immediately laid down on a
C18 silica gel reverse phase column (CH3CN:H2O; 5:95)
yielding 3b (16 mg, 76%) as a colorless oily (50:50) mixture
Analytical TLC. Supernatants of the assays with the differ-
ent analogs were chromatographed on TLC (silica gel 60
Merk 5748, 0.2 mm, eluent: CHCl3; CH3OH; CH3COOH,
90:10:5). After migration and drying, the plate was trans-
ferred on an agar plate cultivated with L. plantarum. After
one night at 37ЊC, the RF value of the spot, which appeared
on each lane, was compared to that of pure biotin taken as
reference.
1
of two diastereoisomers. H NMR (CD3OD): d 1.11; 1.13
(2d, 3H, Jˆ6.1 Hz, CH3); 1.48–1.56; 1.58–1.68; 1.72–1.78
(3m, 2H, NCHCH2); 2.14–2.27 (m, 2H, CH2CH2COOH);
2.52–2.57 (m, 2H, Jˆ7.6 Hz, CH2COOH); 3.82–4.01 (m,
3H, Jˆ8.1 Hz, Jˆ7.6 Hz, 2CHN, CH2CHO); 4.45 (dt, 1H,
Jˆ4.1 Hz, Jˆ7.6 Hz, OCHCH2). 13C NMR (CD3OD): d
15.78; 15.96 (CH3); 22.81; 22.97 (CH2CH2COOH); 29.36
(CH2COOH); 32.37; 33.63 (NCHCH2); 52.70; 52.86
(CHN); 53.91; 56.08 (CHN); 69.87; 71.97 (CH2CHO);
84.61; 84.87 (OCHCH2); 165.92; 166.38 (CyO); 180.09
(COOH). I.C. (NH3) MHϩˆ229.
cis-4-(Z-5[14C]-Carboxypent-1-enyl)-5-methyl imidazoli-
din-2-one [14C]5a. [14C]KCN (SA 46 Ci molϪ1) into H2O
(300 mL, 300 mCi, 6.51 mmol) was diluted with CH3OH
(1 mL) and evaporated under vacuum, anhydrous DMSO
(100 mL) and 16a (1.1 mg, 5.21 mmol) were added. After
stirring for 2 h at 40ЊC, DMSO was evaporated under
vacuum at 50ЊC, 1N NaOH (0.5 mL) was added, following
by heating under reflux for 2 h, addition of 12N HCl
(150 mL) and purification by HPLC (Lichrospher C8, H2O,
TFA, 0.1%: CH3CN, 87:13) yielded [14C]5a (27.6 mCi,
0.6 mmol).
Acknowledgements
Lactone 8 and methyltrioxorhenium were generous gifts of
the Hoffmann–La Roche Society (Basel, Nutley) and H.
´
Rudler from Universite Paris VI, respectively. We express
our thanks to Dr J. Waissermann (Laboratoire de Chimie des
´
´
Metaux de Transition, Universite Paris VI) for carrying out
X-ray analysis. This work has been supported in part by the
´
´
IFSBM (Institut de Formation Superieure Biomedicale,
Villejuif, France).
References
1. Marquet, A.; Florentin, D.; Ploux, O.; Tse Sum Bui, B. J. Phys.
Org. Chem. 1998, 11, 1–7 and references cited therein.
2. (a) Guianvarch, D.; Florentin, D.; Tse Sum Bui, B.; Nunzi F.;
Marquet, A. Biochem. Biophys. Res. Commun. 1997, 236, 402–
406. (b) Escalettes, F.; Florentin, D.; Tse Sum Bui, B.; Lesage, D.;
Marquet, A. J. Am. Chem. Soc. 1999, 121, 3571–3578.
cis-4-(5-[14C]-Carboxy-cis-2,3-epoxy-pentyl)-5-methyl
imidazolidin-2-one [14C]3a. To [14C]2a5 (SA: 53 Ci molϪ1
)
into H2O (1 mL, 0.322 mmol, 17.1 mCi) MTO (traces) and a
0.35% H2O2 solution (286 mL) were added. After 30 min
stirring at room temperature, MTO and H2O2 (67 mL)
were added once more. The mixture was stirred for 1 h at
room temperature and purified by HPLC (Lichrosopher C8,
H2O, TFA 0.1%:CH3CN, 95:5) giving [14C]3a1 (110 nmol,
5.84 mCi) and [14C]3a2 (107 nmol, 5.65 mCi).
3. Tse Sum Bui, B.; Florentin, D.; Fournier, F.; Ploux, O.;
´
Mejean, A.; Marquet, A. FEBS Lett. 1998, 440, 226–230.
4. (a) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13,
´
317–323. (b) Eydoux, F.; Chlenov, M. A.; Reglier, M. Bioorg.
Med. Chem. Lett. 1995, 5, 941–944 and references cited
therein.
5. Jestin, E.; Moreau, F.; Florentin, D.; Marquet, A. Bioorg. Med.
Chem. 1996, 4, 1065–1075.
6. Gerecke, M.; Zimmermann, J. P.; Aschwanden, W. Helv. Chim.
Acta 1970, 53, 991–999.
7. Even, L.; Florentin, D.; Marquet, A. Bull. Soc. Chim. Fr. 1990,
127, 758–768.
8. Barton, D. H. R.; Jaszberenyi, J. C. Tetrahedron Lett. 1989, 30,
2619–2622.
cis-4-(5-[14C]-Carboxy-trans-2,3-epoxy-pentyl)-5-methyl
imidazolidin-2-one [14C]3b. A solution of [14C]2b5 (SA:
53 Ci molϪ1) into H2O (1.37 mL, 0.629 mmol, 34 mCi)
was treated as above for 3a yielding after HPLC purification
(H2O, TFA 0.1%:CH3CN, 96:4) 3b (89 nmol, 4.8 mCi) as a
mixture of epoxides [14C]3b1 and [14C]3b2.
Biological experiments
9. Schlosser, M.; Christmann, K. F.; Piskala, A. Chem. Ber. 1970,
103, 2814–2820.
Assays with biotin synthase of E. coli. Two mixtures were
separately prepared and left under argon for 1 h at room
temperature after which they were mixed and incubated
for 3 h at 37ЊC. Mixture A consisted of 6.3 mM pure biotin
10. Adam, W.; Bialas, J.; Hadjiarapoglou, L. P. Chem. Ber. 1991,
124, 2377.
11. (a) Herrmann, W. A.; Fischer, R. W.; Rauch, M. U.; Scherer,
W. J. Mol. Catal. 1994, 86, 243–266. (b) Rudler, H.; Ribeiro

F. Moreau et al. / Tetrahedron 56 (2000) 285–293
293
Gregorio, J.; Demse, B.; Bregeault, J. M.; Deloffre, A. J. Mol.
Catal. A: Chem. 1998, 133, 255–265.
14. Izumi, Y.; Osakai, M.; Yamada, H. J. Microbiol. Meth. 1987,
6, 237–246.
12. Repic, O.; Vogt, S. Tetrahedron Lett. 1982, 23, 2729–2732.
13. Denmark, S. E.; Edwards, J. P. J. Org. Chem. 1991, 56,
6974–6981.
15. Cannon, J. R.; Chow, P. W.; Metcalf, B. W.; Power, A. J.
Tetrahedron Lett. 1970, 4, 325–328.