Simple and efficient syntheses of Boc- and Fmoc-protected 4(R)- and 4(S)-fluoroproline solely from 4(R)-hydroxyproline

Article abstract of DOI:10.1016/S0040-4020(02)01020-7

As building blocks of collagen model peptides, Boc- and Fmoc-protected 4(R)- and 4(S)-fluoroproline, which will be widely used in peptide synthesis including solid-phase strategy, were synthesized from the readily available 4(R)-hydroxyproline in higher y

Full text of DOI:10.1016/S0040-4020(02)01020-7

Tetrahedron 58 (2002) 8453–8459
Simple and efficient syntheses of Boc- and Fmoc-protected
4(R)- and 4(S)-fluoroproline solely from 4(R)-hydroxyproline
Masamitsu Doi,a Yoshinori Nishi,b Naruto Kiritoshi,a Tomoya Iwata,a Mika Nago,a
Hiroaki Nakano,b Susumu Uchiyama,b Takashi Nakazawa,c Tateaki Wakamiyad
and Yuji Kobayashib
,
*
aDepartment of Materials Science, Wakayama National College of Technology, 77 Noshima, Nada-cho, Gobo, Wakayama 644-0023, Japan
bGraduate School of Pharmaceutical Sciences, Osaka University, 1–6 Yamada-oka Suita, Osaka 565-0871, Japan
cDepartment of Chemistry, Faculty of Science, Nara Women’s University, Kitauoyanishi-machi, Nara 630-8506, Japan
dDepartment of Chemistry, Faculty of Science and Technology, Kinki University, 3-4-1 Kowakae, Higashi-Osaka, Osaka 577-8502, Japan
Received 28 June 2002; accepted 19 August 2002
Abstract—As building blocks of collagen model peptides, Boc- and Fmoc-protected 4(R)- and 4(S)-fluoroproline, which will be widely used
in peptide synthesis including solid-phase strategy, were synthesized from the readily available 4(R)-hydroxyproline in higher yield than with
conventional methods. To establish the stereospecificity of the Mitsunobu reaction and the subsequent fluorination that were presumed to
cause the inversion of configuration at the C-4 position of a proline derivative, the absolute configuration of one of the key products, Boc-
4(S)-fluoroproline, was determined by X-ray crystallography. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
In order to clarify what forces are controlling the self-
assembly of these chains and how a Hyp residue stabilizes
or destabilizes the triple structure, X-ray analyses were
carried out on crystals of these peptides.11,12 Although intra-
and inter-chain hydrogen bond formations were discussed
extensively both directly and through water molecules, the
contribution of Hyp residues were not well established.
Various polytripeptides have been synthesized as collagen
model peptides on the hypothesis that the unique amino acid
sequence of collagen, which has glycine in every third
position and approximately 300 repeats of X-Y-Gly where
X is often proline and Y is often 4(R)-hydroxyproline (4(R)-
Hyp), should result in its unique triple helical structure.1–3
Their physico-chemical properties have been intensively
investigated.
Recently Raines and co-workers synthesized (Pro-4(R)-
fPro-Gly)10, where 4(R)-fPro is 4(R)-fluoroproline, and
showed that it takes the most stable collagen mimic among
the all polytripeptides mentioned here.13,14 They explained
the increased stability by the stereoelectronic effects coming
from the electronegativity of fluorine atom which fixes the
pyrrolidine ring pucker and makes all three main-chain
torsion angles; v, w and c, favorable to fit the triple helix
formation.15 The inductive effects of hydroxyl group of Hyp
were analogous to those of fluorine atom in the typical
derivatives of fPro.13–15 The different angles correlated
with Cg pucker of the X-position and the Y-position. This
was also pointed out later by Zagari and co-workers with
X-ray analysis on (Pro-Pro-Gly)10 at higher resolution.
Furthermore, the preferential ring pucker was applied to the
case with 4(S)-Hyp analogues as well.16
For example, we synthesized a series of polytripeptide
(X-Y-Gly)n with defined number of n by a solid phase
polycondensation of tripeptides, X-Y-Gly.4,5 The results
showed that (Pro-Pro-Gly)10 and (Pro-4(R)-Hyp-Gly)10
have triple helical structures at lower temperature and
undergo thermal transition to single random coil states.6,7
The transition temperature of the latter was much higher
than that of the former.6–8 The additional thermal stabilities
provided by the existence of Hyp residues are consistent
with the co-relation found in naturally occurring collagen
molecules themselves between their Hyp contents and
thermal transition temperatures. However, (4(R)-Hyp-Pro-
Gly)10, (4(S)-Hyp-Pro-Gly)10 and (Pro-4(S)-Hyp-Gly)10 do
not form triple helices and exist as single coils.9,10
Despite these explanations we seem to solve the long-
unresolved questions regarding the factors to control the
thermal stability of the collagen-like triple helical structure.
To do this we would like to inspect them by systematic
thermodynamic experiments on the series of polytripeptides
Keywords: stereoselective synthesis; 4-fluoroproline; 4(R)-hydroxyproline;
Mitsunobu reaction; X-Ray analysis; collagen model peptides.
*
Corresponding author. Tel.: þ81-6-6879-8220; fax: þ81-6-6879-8224;
e-mail: yujik@protein.osaka-u.ac.jp
0040–4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(02)01020-7

8454
M. Doi et al. / Tetrahedron 58 (2002) 8453–8459
Scheme 1.
in solution. Thus various collagen models containing
diastereomers of fPro as well as those of Hyp need to be
prepared for further comprehensive thermodynamic studies.
For this purpose, solid phase peptide synthesis can provide
us with model peptides with an appropriate quantity and
quality if the requisite tert-butyloxycarbonyl (Boc) or
9-fluorenyl-methoxycarbonyl (Fmoc) derivative of fPro is
available in large amounts. A diastereomer of fPro has been
prepared by a stereospecific displacement of the hydroxyl
group of Hyp by fluorine; this includes the synthesis of 4(R)-
fPro from the N-protected-4(S)-Hyp derivatives,17–19 and
that of Boc-4(R)- and Boc-4(S)-fPro by the inversion of
configuration at the C-4 atoms of the corresponding N-Boc-
Hyp derivative.17,18,20 In particular, Demange et al.
employed the Mitsunobu reaction21 to obtain N-Boc-4(S)-
Hyp-OH from the 4(R) isomer in 58% yield, which
successfully managed to alleviate the difficulty in relatively
a lower availability of N-Boc-4(S)-Hyp-OH as the starting
material for the synthesis of 4(R)-fPro-OH.20 Perhaps this
also serves to enhance the practicality of alternative
methods that require the derivatives of N-Fmoc-4(S)-Hyp-
OH for the synthesis of Fmoc-4(R)-fPro-OH.22,23 However,
there still remain a few problems to solve: (1) The reaction
condition should be optimized so that the requisite 4(R)-fPro
derivative may be obtained more easily in an appreciably
improved overall yield. (2) Neither the Mitsunobu reaction
nor the fluorination has been guaranteed to undergo in a
manner of purely the SN2 reaction or even an enantio-
selective one with a certain extent of racemization, while the
optical purity of an amino acid is of particular concern for
use in solid-phase peptide synthesis.
diffraction analysis of Boc-4(S)-fPro-OH together with the
analysis of all the reaction products using a chiral column in
reversed phase high-performance liquid chromatography
(RP-HPLC) demonstrates unambiguously that the method
including the Mitsunobu reaction and the subsequent
fluorination proceeds stereospecifically as we had expected.
2. Results and discussion
2.1. Synthesis of 4(S)-Hyp derivatives from 4(R)-
Hyp-OH
In order to elaborate a simple and economical route for the
preparation of Boc- and Fmoc-protected 4(R)- or 4(S)-fPro
derivatives, we chose N-Boc-4(R)-Hyp-OH (1a) as the sole
starting material because it is much less expensive than the
4(S) isomer and is readily available. For the large-scale
preparation of 4(S)-fPro from the 4(R)-Hyp derivative, we
basically followed the method involving the Mitsunobu
reaction as reported by Demange et al.20 with some
modifications. After several trials with a variety of amino-
and carboxyl-protected 4(R)-Hyp derivatives, the most
satisfactory result was obtained in the one-pot synthesis of
N-Boc-4(S)-Hyp-OH (1b) from N-Boc-4(R)-Hyp-OMe (2a)
as shown in Scheme 1. The yield of 1b in this reaction was
78%, which amounted to roughly 20% better than that of a
conventional method20 starting from N-trityl-4(R)-Hyp-OH.
Obviously, this improvement in yield is due to the use of the
less acid-labile protective group of Boc rather than the
trityl group. By avoiding the need of exchanging these
N-protective groups, we could carry out the reaction in one
pot, which led to a substantially improved yield.
We suggest here a simpler and more efficient method for the
preparation of 4(S) and 4(R) diastereomers of Boc- and
Fmoc-fPro starting solely from 4(R)-Hyp-OH. The X-ray
It seems convenient to use the phenacyl (Pac) ester,

M. Doi et al. / Tetrahedron 58 (2002) 8453–8459
8455
The optical purity of 3a and 3b was checked by chiral HPLC
analysis. In principle, an ordinary RP-HPLC would suffice
for the analysis of diastereomers, but HPLC with a chiral
column enabled us to distinguish these isomers more clearly
so that even a minute content of racemization or
epimerization product could be discerned. As shown in
Fig. 1(A) and (B), each compound exhibited a single peak,
resolved well enough to discriminate one from another; the
retention times for 3a and 3b were 10.8 and 11.8 min,
respectively. This result indicates that the Mitsunobu
reaction leading to 1b has undergone with virtually
complete stereochemical inversion in a manner of the SN2
reaction, rendering the C-4 position of 3b the S-configur-
ation.
Figure 1. HPLC profiles of 3a (A), 3b (B), 4a (C) and 4b (D) at 208C at a
sample concentration of 0.2 mg/mL in ethanol. Column: Daicel CHIR-
ALPAK AS (4.6£250 mm). Eluent: 0–40% hexane in ethanol in 20 min.
Flow rate: 1.5 mL/min.
2.2. Conversion of the Hyp to fPro derivative
We obtained Boc-4(S)-fPro-OPac (4a) and Boc-4(R)-fPro-
OPac (4b) by the fluorination of 3a and 3b, respectively,
with morpholinosulphur trifluoride (Scheme 2). The
geometry of 4a was tentatively assigned to the 4(S)-
configuration by the assumption that the substitution of
the hydroxyl group with fluorine is a typical SN2 reaction;
this finally proved to be correct by allowing for the results of
the X-ray analysis of 5a as described below. Note that the
configuration at the C-4 atom is inverted during the
fluorination, by which 3b can also be converted to 4b in
the same way. The outcome of the reaction was, however,
not quite the same for each product: when CH2Cl2 was used
as a solvent, the yield of 4a from 3a was as high as 94%,
while that of 4b from 3b was 56%. In contrast, the yield of
4b from 3b was improved up to 79% by changing the
solvent from CH2Cl2 to tetrahydrofuran (THF). The
products, 4a and 4b, were analyzed by chiral HPLC with
respect to the optical purity. Similarly to the chromatograms
of 3a and 3b, there appeared a single peak for each sample
of 4a and 4b (Fig. 1C and D). The difference in retention
times between the peaks of these diastereomers was more
than 3 min, sufficient to rule out the possibility of
overlooking the contamination of one component with the
N-Boc-4(R)-Hyp-OPac (3a), as the starting material because
3a is convertible not only to Boc-4(S)-fPro-OPac (4a) but
also to Boc-4(R)-fPro-OPac (4b) through the fluorination of
N-Boc-4(S)-Hyp-OPac (3b). The merits of the Pac group are
that it is very stable to acidolytic conditions of Boc
deprotection and that it can be removed by the reduction
with Zn/AcOH in good yield (usually more than 90%),24
keeping both Boc and Fmoc groups intact. However, the
yield of 3b directly from the 4(R) isomer (3a) was
prohibitively low (35%) to be rated as practical. The steric
hindrance of the Pac group against the Mitsunobu reagents
and undesirable hydrolysis concomitant with the treatment
of reaction intermediate with NaOH could possibly reduce
the yield of 3b. Instead of taking this route, we therefore
chose the present one (Scheme 1) to obtain 1b, from which
the Pac ester (3b) was prepared by the standard procedure of
esterification. Although the direct conversion from 3a to 3b
was thus bypassed, the overall yield of 3b from 1a via 1b
was 64%, which still exceeded that attained through the
shorter pathway.
Scheme 2.

8456
M. Doi et al. / Tetrahedron 58 (2002) 8453–8459
epimerization products turned out to be negligible. Even
though there should arise a further need to purify these
products, a task of removing such a diastereomer contami-
nant in a small amount is much less laborious compared to
that of resolving a mixture of amino-acid enantiomers in
general. Therefore a small extent of epimerization at this
stage, if any, may not seriously degrade the efficacy of our
method as a whole.
2.4. Stereochemistry of the Mitsunobu reaction and the
fluorination
The Mitsunobu reaction of 2a to 1b appeared to proceed
with a complete inversion of the configuration as revealed
by the HPLC analysis of the reaction product. We also
found that 4a and 4b were produced by the SN2 reaction of
3a and 3b, respectively. There arose, however, a problem of
a much poorer performance of the reaction from 3b to 4b
than that from 3a to 4a under the same reaction conditions.
In unfavorable cases, poor yield can often be associated with
low stereoselectivity of the reaction.
Figure 2. ORTEP view of the X-ray crystal structure of 5a. Each atom is
drawn with 50%-probability thermal ellipsoid.
counterpart. These results indicate that 4a and 4b have been
obtained through an almost complete inversion of configur-
ation at the C-4 position of their respective precursors. In the
reaction of a derivative of 4(R)-O-tosyl-Hyp with fluoride
ion, it was reported that 17% of the fluorination product
retained the R-configuration at the C-4 atom.19 The
retention is considered to arise from the intra-molecular
participation of the ester carbonyl in the displacement
process. It is therefore remarkable that the reaction of 3a
with morpholinosulphur trifluoride appeared to proceed
without retention of configuration, while the ester carbonyl
group of 3a is in the same disposition to the hydroxyl group
as that of the 4(R)-O-tosyl-Hyp derivative.
Fluorination of 3b gave 4b in various yields ranging from 56
(in CH2Cl2) to 79% (in THF) depending on solvent
conditions. It is likely that the cause of variation in these
yields is attributable to a solvent-dependent conformational
change of 3b. When IR spectra of 3a and 3b were compared
in CH2Cl2 (Fig. 3(A) and (B)), there appeared noticeable
differences in the amide A region and the amide I region. In
the amide A region, 3a had one sharp absorption band at
3603 cm21, while 3b had one additional broad band around
3508 cm21 to a sharp one at 3602 cm21. In the amide I
region, 3a had two bands at 1759 and 1704 cm21, while 3b
had three at 1762, 1746 and 1702 cm21. We regard these
additional bands of 3b as the clue to indicate the presence of
intra-molecular hydrogen bond between the hydroxyl group
and the carbonyl group of the ester moiety. This is because
hydrogen bonding tends to alter the stretching frequencies
of bands due to both the proton donor and acceptor
2.3. Syntheses of Boc and Fmoc-fPro-OH and X-ray
diffraction analysis of 5a
By removing the Pac group with Zn/AcOH, we obtained
Boc-4(S)-fPro-OH (5a) and Boc-4(R)-fPro-OH (5b) in 90–
94% yields. A drawing of the crystal structure revealed by
X-ray diffraction analysis shows the geometry at the C-4
position of 5a to be the S-configuration (Fig. 2). Referring to
this configuration of 5a, we can now unambiguously
determine the geometry of every reaction product obtained
in the present method. This also confirms that both the
Mitsunobu reaction and the substitution of the hydroxyl
group with fluorine occur almost exclusively with inversion
of configuration at the C-4 position.
For the solid phase synthesis of collagen mimics based on
the Fmoc chemistry, each isomer of Fmoc-fPro-OH (7a and
7b) was also derived from the corresponding isomer of Boc-
fPro-OPac (4a and 4b). The procedure consists of: (1) Boc
removal in trifluoroacetic acid (TFA), (2) amino protection
by Fmoc-OSu, and (3) deprotection of the Pac group
(Scheme 2). To prevent a serious lowering of yield owing to
extremely high solubility of carboxyl-free fPro derivatives
in water, we deliberately avoided removing the Pac group
before acylation that reportedly has a likelihood of causing
racemization at the Ca position.25 Judged from the results of
HPLC analysis with a chiral column that can distinguish
between a pair of enantioners as well as diastereomers, it
was fortunate that contamination of 7a and 7b with their
Figure 3. FT-IR spectra of 3a (a) and 3b (b) at a sample concentration of
1.0 mM in CH2Cl2. Parts of the spectra of the amide A region and the amide
I region were shown in (A) and (B), respectively.

M. Doi et al. / Tetrahedron 58 (2002) 8453–8459
8457
groups.26,27 The relevant bands in the present case are those
with wavenumbers at 1746 cm21 (nCvO of the ester
carbonyl) and at 3508 cm21 (nO–H). If this hydrogen
bonding interferes with the attachment of hydroxyl oxygen
to the sulphur atom of morpholinosulphur trifluoride,
activation of the hydroxyl group into the better leaving
group and the subsequent nucleophilic attack of fluoride ion
become less likely to occur. Note that only the 4(S)-
configuration of 3b permits the intra-molecular hydrogen
bonding possible (see, for example, the structure shown in
Scheme 2). In THF, where the reaction with the fluorinating
reagent appeared to proceed without such an interference
and eventually gave 4b in considerably enhanced yield
(79%), which is roughly comparable with the case affording
4a (89%), the hydrogen bond in 3b could be very weak or
absent altogether.
FT-IR spectrometer (Shimadzu Co., Ltd, Kyoto, Japan).
Measurements were performed using a liquid cell (KBr,
5 mm-path length) containing samples at a concentration of
1
1.0 mM in CH2Cl2 at room temperature. H and 13C NMR
spectra were recorded on a Varian UNITY-plus 300
spectrometer (Varian Inc., Palo Alto, CA, USA). Chemical
shifts for 1H and 13C spectra are reported relative to
tetramethylsilane (TMS) at 0 ppm. Assignments were made
using two-dimensional correlation spectroscopy (COSY),
nuclear Overhauser enhancement spectroscopy (NOESY)
1
and H–13C hetero-scalar correlation spectroscopy (HET-
COR). Concentration of samples was approximately 30 mM
in CDCl3. Measurements of NOESY and HETCOR spectra
(data not shown) were indispensable for the assignment of
1H signals, especially of Cb and Cd protons, unambiguously.
In many cases, a couple of spin systems were located at a
particular site of the proline resonance by COSY cross-
peaks that were considered to represent a pair of trans- and
cis-conformers.28 The results of these assignments are
reported only for the major conformer (probably the trans
form) unless the signals of the minor component are clearly
recognizable. Molecular weights were determined by
electron-spray-ionization mass spectrometry (ESI-MS)
using an Applied Biosystems Marinere instrument
(Applied Biosystems Inc., Foster City, CA, USA) operating
in the mode that detects positive ions.
Because the yield of 4a is indistinctly lower in THF (89%)
than in CH2Cl2 (94%), it is unlikely that THF plays some
intrinsic role in facilitating the reaction. The improved
efficiency in fluorination to give 4b by using THF in place of
CH2Cl2 is thus regarded as a result of releasing the 4(R)-
Hyp derivative from forming an undesirable conformation
in which an intra-molecular hydrogen bond is assumed to
interfere with the reaction. Allowing for a remarkably high
yield of 4a achieved in CH2Cl2, we expect to be able to
improve the yield of 4b a little further by optimizing the
reaction conditions with respect to the choice of the
protective group as well as solvent.
3.1.1. N-Boc-4(R)-Hyp-OMe (2a). Boc-4(R)-Hyp-OH (1a;
mp 125,1268C, [a]2D2¼284.48 at c¼1.00 in CHCl3) was
prepared from 4(R)-Hyp-OH (1) and (Boc)2O (purchased
from Peptide Institute Inc. Osaka, Japan) by the standard
protocol using (Boc)2O.29 To a solution of 1a (15.0 g,
64.9 mmol) in diethyl ether (200 mL) was added an excess
of diazomethane in diethyl ether at 08C, and then the
reaction mixture was allowed to stand overnight. After
removing the solvent and diazomethane under reduced
pressure, the oily residue dissolved in ethyl acetate was
applied onto a silica gel column, and 1a was eluted with
ethyl acetate/hexane (1:2, v/v). Yield: 15.0 g (94.3%).
2.5. Syntheses of 4(R) and 4(S) isomers of Fmoc-fPro-OH
For the syntheses of Fmoc-fPro-OH (7a: 4(R) and 7b: 4(S)),
it would also be envisaged that an appropriate ester of
Fmoc-Hyp-OH might serve for direct fluorination as
suggested by other authors.22,23 Although our method does
have the option to utilize 1b as a precursor to N-Fmoc-4(R)-
Hyp-OH, the yield of 83%23 in the conversion of N-Fmoc-
4(S)-Hyp-OBzl (Bzl: benzyl) to Fmoc-4(R)-fPro-OBzl is
not impressive enough to let us prefer this reaction to the
corresponding pathway leading to 4b from 3b in 79% yield.
One of the main features of the present method is that it can
provide both 4(S) and 4(R) diastereomers of fPro derivatives
starting solely from 4(R)-Hyp-OH with a modest improve-
ment in overall yield for each compound, making them
more accessible to the studies of collagen.
3.1.2. N-Boc-4(S)-Hyp-OH (1b). To a solution of 2a
(7.04 g,
28.7 mmol),
triphenylphosphine
(15.1 g,
57.4 mmol), and formic acid (2.20 mL, 57.4 mmol) in
THF (75 mL) was added dropwise 40% diethyl azodi-
carboxylate (25.0 g, 57.4 mmol) in toluene at room
temperature under a nitrogen atmosphere. The reaction
mixture was stirred overnight, and then 1 M NaOH
(94.8 mL) was added to the solution. After stirring for 1 h
at room temperature, most of organic solvent was removed
in vacuo. The aqueous residue diluted by addition of H2O
(50 mL) was washed with ethyl acetate, and acidified with
1 M HCl, followed by extraction with ethyl acetate. The
extract was washed with water and saturated aqueous
NaHCO3, dried over Na2SO4, and concentrated in vacuo.
The residue was triturated with diethyl ether, and the
powdery product was collected by filtration. The thus-
obtained crude product was recrystallized from ethyl acetate
and hexane. Yield: 5.18 g (78.1%).
3. Experimental
3.1. General methods
The preparative column chromatography of synthetic
materials was carried out on a column (20£300 mm) of
Wakogelw C-300 (45,75 mm, Wako Pure Chemical
Industries, Ltd, Osaka, Japan). Analytical RP-HPLC was
performed with a CHIRALPAK AS (4.6£250 mm, Daicel,
Osaka, Japan) at 208C and a flow rate of 1.5 mL/min. A
solution containing 0.2 mg/mL in ethanol was loaded onto
the column and eluted with a linear gradient concentration
of hexane from 0% (0 min) to 40% (20 min) in ethanol.
Chromatograms were monitored by UV-detection at
220 nm. FT-IR spectra were recorded on a Shimadzu 8100
3.1.3. N-Boc-4(R)-Hyp-OPac (3a). To a solution of 1b
(5.75 g, 24.9 mmol) in methanol (100 mL) was added
Cs2CO3 (4.05 g, 12.5 mmol) in water (65 mL) at 08C, and
the mixture was concentrated in vacuo. The residue was

8458
M. Doi et al. / Tetrahedron 58 (2002) 8453–8459
dissolved in N,N-dimethylformamide (DMF) (125 mL), and
then phenacyl bromide (Pac-Br) (4.95 g, 24.9 mmol) was
added to the solution at 08C. After being stirred for 30 min,
the reaction mixture was filtered, and the filtrate was
concentrated in vacuo. The residue was extracted with ethyl
acetate, and the extract was washed with water and saturated
aqueous NaHCO3 (2£100 mL). The organic layer was dried
over Na2SO4, and then the solvent was evaporated to
dryness. The residue was triturated with diethyl ether, and
the powdery product was collected by filtration. The thus-
obtained crude product was recrystallized from ethyl acetate
and hexane. Yield: 7.66 g (88.0%).
minor isomer at 2.41), 1.44 (s, 9H, Boc; minor isomer at
1.49); 13C NMR (CDCl3, ppm) 177.12 (CO; minor isomer at
174.70), 155.53 (Boc-CO; minor isomer at 153.86), 91.91
(Cg, JC–F¼177 Hz; minor isomer at 91.19), 81.73 (Boc-C;
minor isomer at 80.97), 57.61 (Ca; minor isomer at 57.45),
53.57 (Cd, JC–F¼25 Hz; minor isomer at 52.98), 37.36 (Cb,
JC–F¼22 Hz; minor isomer at 35.65), 28.35 (Boc-CH3;
minor isomer at 28.26). Analysis calcd for C10H6O4NF: C,
51.50; H, 6.91; N, 6.01. Found: C, 51.36; H, 6.79; N, 5.97;
ESI-MS (m/z): 256.04 [MþNa]þ (calcd 256.10) and 272.00
[MþK]þ (calcd 272.07). Crystal data: orthorhombic,
˚
P212121, a¼9.654, b¼18.70, c¼6.455 A, Z¼4,
l
˚
(Mo Ka)¼0.71069 A, T¼296 K, 1562 unique reflections,
R¼0.094. Crystals contained one molecule in each
asymmetric unit.
3.1.4. N-Boc-4(S)-Hyp-OPac (3b). Esterification of 1b
(3.50 g, 15.1 mmol) with Pac-Br (3.00 g, 15.1 mmol) was
carried out in a similar manner as described above. The
crude product was purified by recrystallization from ethyl
acetate and hexane. Yield: 4.59 g (86.8%).
3.1.8. Boc-4(R)-fPro-OH (5b). In a fashion similar to the
preparation of 5a from 4a, 5b was prepared from 4b
(0.500 g, 1.42 mmol). The oily product of 5b was solidified
by triturating with hexane, and collected by filtration to give
amorphous powder. Yield: 0.300 g (90.6%). [a]2D2¼2628
(c¼1.00, CHCl3); 1H NMR (CDCl3) 7.29 (brs, 1H, COOH),
5.21 (m, JH–F¼53 Hz, 1H, CgH; minor isomer at 5.32), 4.54
(t, J¼8 Hz, 1H, CaH; minor isomer at 4.43), 3.90 (m,
JH–F¼21 Hz, 1H, CdH; minor isomer at 3.93), 3.52 (m,
J¼12 Hz, JH–F¼37 Hz, 1H, CdH0; minor isomer at 3.63),
2.58 (m, JH–F¼20 Hz, 1H, CbH; minor isomer at 2.65), 2.40
(m, JH–F¼40 Hz, 1H, CdH0; minor isomer at 2.17), 1.50 (s,
9H, Boc; minor isomer at 1.44); 13C NMR (CDCl3) 178.04
(CO; minor isomer at 174.01), 153.58 (Boc-CO; minor
isomer at 156.40), 91.21 (Cg, JC–F¼178 Hz; minor isomer
at 91.01), 82.36 (Boc-C; minor isomer at 81.15), 57.70 (Ca;
minor isomer at 57.48), 53.47 (Cd, JC–F¼23 Hz; minor
isomer at 52.95), 35.60 (Cb, JC–F¼22 Hz; minor isomer at
37.43), 28.29 (Boc-CH3; minor isomer at 28.18). Analysis
calcd for C10H6O4NF·0.1H2O: C, 51.10; H, 6.96; N, 5.96.
Found: C, 50.78; H, 6.61; N, 5.46; ESI-MS (m/z): 256.06
[MþNa]þ (calcd 256.10).
3.1.5. Boc-4(S)-fPro-OPac (4a). To a solution of 3a
(5.98 g, 17.1 mmol) in CH2Cl2 (80 mL) was added
morpholinosulphur trifluoride (15.0 g, 85.5 mmol) dropwise
over a 30 min period at 2788C under a nitrogen atmosphere,
and the mixture was then stirred for 48 h at room
temperature. The reaction mixture was concentrated in
vacuo, and to the residue was added water (50 mL). The
resulting oily product was extracted with ethyl acetate, and
the extract was washed with water and saturated aqueous
NaHCO3. The organic layer was dried over Na2SO4, and
then the solvent was evaporated to dryness. The crude
product was purified by silica-gel column chromatography
with ethyl acetate/hexane (1:2, v/v) eluents. Yield: 5.65 g
(94.0%).
3.1.6. Boc-4(R)-fPro-OPac (4b). Procedures for the
preparation of 4b were carried out in a similar manner as
those for 4a except that 3b (6.13 g, 17.5 mmol) was
dissolved in THF (80 mL) instead of CH2Cl2. Yield:
4.88 g (79.4%).
3.1.9. Fmoc-4(S)-fPro-OPac (6a). The solution of 4a
(4.40 g, 12.5 mmol) in 40 mL of TFA/CH2Cl2 (1:2, v/v)
was stirred at room temperature for 1 h, and then
concentrated in vacuo. The oily residue was triturated
with diethyl ether, and the powdery product was collected
by filtration and dried in vacuo. To the crude product was
added Fmoc-OSu (4.64 g, 13.8 mmol) and N-methyl-
morpholine (1.26 g, 12.5 mmol) in CH2Cl2 (50 mL) at
room temperature. The mixture was stirred for 6 h at room
temperature, and the solvent was removed in vacuo. The
oily residue was triturated with water, and the powdery
substance was collected by filtration, and washed with
ethanol. Yield: 5.41 g (91.2%). The thus-obtained 6a was
used for the next reaction without further purification.
3.1.7. Boc-4(S)-fPro-OH (5a). To a solution of 4a (0.800 g,
2.28 mmol) in acetic acid (20 mL) was added zinc dust
(7.44 g, 114 mmol), and the mixture was stirred overnight at
room temperature. The reaction mixture was filtered, and
the filtrate was concentrated in vacuo. The residue was
extracted with ethyl acetate, and the extract was washed
with water (2£30 mL) and 10% aqueous citric acid
(2£30 mL). The organic layer was dried over anhydrous
Na2SO4, and concentrated in vacuo. The thus-obtained
crude product was purified by silica-gel column chroma-
tography with ethyl acetate/hexane (1:2, v/v) eluents. The
oily product was solidified by trituration with diethyl
ether, and collected by filtration. Yield: 0.503 g (94.5%).
The thus-obtained 5a was dissolved in ethanol, and
colorless rectangular crystals for X-ray diffraction
analysis were prepared by evaporating the solvent
slowly at room temperature. [a]2D2¼2738 (c¼0.50,
3.1.10. Fmoc-4(R)-fPro-OPac (6b). Preparation of 6b from
4b (1.79 g, 5.10 mmol) was carried out in a similar manner
as described above. Yield: 2.22 g (91.9%).
1
CHCl3); H NMR (CDCl3, ppm) 7.83 (bs, 1H, COOH),
5.22 (m, JH–F¼53 Hz, 1H, CgH), 4.56 (d, J¼10 Hz, 1H,
CaH; minor isomer at 4.46), 3.77 (m, J¼13 Hz,
JH–F¼26 Hz, 1H, CdH; minor isomer at 3.85), 3.63 (m,
JH–F¼36 Hz, 1H, CdH0), 2.71 (m, JH–F¼21 Hz, 1H, CbH;
minor isomer at 2.55), 2.31 (m, JH–F¼41 Hz, 1H, CbH0;
3.1.11. Fmoc-4(S)-fPro-OH (7a). 7a was prepared from 6a
(3.80 g, 8.03 mmol) in a similar manner as described in the
preparation of 5a. Yield: 2.56 g (89.8%). [a]2D2¼2558
(c¼0.50, CHCl3); 1H NMR (CDCl3) 8.11 (brs, 1H,
COOH), 7.27–7.76 (m, 8H, Fmoc-ArH), 5.24 (m,

M. Doi et al. / Tetrahedron 58 (2002) 8453–8459
8459
JH–F¼53 Hz, 1H, CgH; minor isomer at 5.18), 4.63 (d,
J¼10 Hz, 1H, CaH; minor isomer at 4.37), 4.45 (d, 2H,
Fmoc-CH2; minor isomer at 4.44), 4.26 (t, J¼7 Hz, H,
Fmoc-CH; minor isomer at 4.16), 3.85 (dd, J¼13 Hz,
JH–F¼26 Hz, 1H, CdH), 3.66 (m, J¼13 Hz, JH–F¼35 Hz,
1H, CdH0; minor isomer at 3.69), 2.63 (dd, J¼16 Hz,
JH–F¼23 Hz, 1H, CbH; minor isomer at 2.57), 2.34 (m,
JH–F¼41 Hz, 1H, CbH0); 13C NMR (CDCl3) 175.22 (CO;
minor isomer at 176.09), 155.07 (Fmoc-CO; minor isomer
at 154.42), 91.96 (Cg, JC–F¼178 Hz; minor isomer at
90.90), 67.95 (Fmoc-CH2; minor isomer at 67.75), 57.66
(Ca, JC–F¼22 Hz; minor isomer at 57.16), 53.53 (Cd,
JC–F¼22 Hz; minor isomer at 53.22), 47.16 (Fmoc-CH),
36.26 (Cb; minor isomer at 37.55). Analysis calcd for
C20H18O4NF·H2O: C, 64.34; H, 5.41; N, 3.75. Found: C,
64.88; H, 5.35; N, 3.72; ESI-MS (m/z): 378.03 [MþNa]þ
(calcd 378.11) and 394.00 [MþK]þ (calcd 394.09).
ed. Pain, R. H., Ed.; Oxford University: Oxford, 2000; pp
330–351.
4. Sakakibara, S.; Kishida, Y.; Kikuchi, Y.; Sakai, R.; Kakiuchi,
K. Bull. Chem. Soc. Jpn 1968, 41, 1273.
5. Sakakibara, S.; Inouye, K.; Shudo, K.; Kishida, Y.; Kobayashi,
Y.; Prockop, D. J. Biochim. Biophys. Acta 1973, 303,
198–202.
6. Kobayashi, Y.; Sakai, R.; Kakiuchi, K.; Isemura, T.
Biopolymers 1970, 9, 415–425.
7. Berg, B. A.; Prockop, D. J. Biochem. Biophys. Res. Commun.
1973, 52, 115–120.
8. Uchiyama, S.; Kai, T.; Kajiyama, K.; Kobayashi, Y.;
Tomiyama, T. Chem. Phys. Lett. 1997, 281, 92–96.
9. Inouye, K.; Sakakibara, S.; Prockop, D. J. Biochim. Biophys.
Acta 1976, 420, 133–141.
10. Inouye, K.; Kobayashi, Y.; Kyogoku, Y.; Kishida, Y.;
Sakakibara, S.; Prockop, D. J. Arch. Biochem. Biophys.
1982, 219, 198–203.
3.1.12. Fmoc-4(R)-fPro-OH (7b). 7b was prepared from 6b
(2.00 g, 4.22 mmol) in a similar manner as described in the
preparation of 5b. Yield: 1.43 g (95.5%). [a]2D2¼2638
11. Bella, J.; Eaton, M.; Brodsky, B.; Berman, H. M. Science
1994, 266, 75–81.
12. Bella, J.; Brodsky, B.; Berman, H. M. Structure 1995, 3,
893–906.
1
(c¼1.0, CHCl3); H NMR (CDCl3) 7.92 (brs, 1H, COOH),
7.26–7.78 (m, 8H, Fmoc-ArH), 5.25 (m, JH–F¼53 Hz, 1H,
CgH; minor isomer at 5.20), 4.59 (t, J¼8 Hz, 1H, CaH;
minor isomer at 4.45), 4.47 (d, 2H, Fmoc-CH2; minor
isomer at 4.47), 4.28 (t, J¼7 Hz, H, Fmoc-CH; minor
isomer at 4.16), 3.77 (dd, J¼14 Hz, JH–F¼21 Hz, 1H, CdH),
3.62 (m, J¼13 Hz, JH–F¼36 Hz, 1H, CdH0), 2.65 (m,
JH–F¼20 Hz, 1H, CbH; minor isomer at 2.69), 2.34 (m,
JH–F¼39 Hz, 1H, CbH0; minor isomer at 2.17); 13C NMR
(CDCl3) 176.81 (CO; minor isomer at 175.59), 155.49
(Fmoc-CO; minor isomer at 154.57), 91.43 (Cg,
JC–F¼179 Hz; minor isomer at 90.79), 68.13 (Fmoc-CH2;
minor isomer at 67.99), 57.13 (Ca; minor isomer at 57.80),
53.22 (Cd, JC–F¼23 Hz; minor isomer at 53.62), 47.07
(Fmoc-CH; minor isomer at 47.12), 36.25 (Cb,
JC–F¼23 Hz; minor isomer at 37.68). Analysis calcd for
C20H18O4NF·0.1H2O: C, 67.25; H, 5.10; N, 3.92. Found: C,
66.81; H, 5.40; N, 3.49; ESI-MS (m/z): 378.03 [MþNa]þ
(calcd 378.11) and 394.00 [MþK]þ (calcd 394.09).
13. Holmgren, S. K.; Taylor, K. M.; Bretscher, L. E.; Raines, R. T.
Nature 1998, 392, 666–667.
14. Holmgren, S. K.; Bretscher, L. E.; Taylor, K. M.; Raines, R. T.
Chem. Biol. 1998, 6, 63–70.
15. Bretscher, L. E.; Jenkins, C. L.; Taylor, K. M.; DeRider, M. L.;
Raines, R. T. J. Am. Chem. Soc. 2001, 123, 777–778.
16. Vitagliano, L.; Berisio, R.; Mazzarella, L.; Zagari, A.
Biopolymers 2001, 58, 459–464.
17. Panasik, Jr. N.; Eberhardt, E. S.; Edison, A. S.; Powell, D. R.;
Raines, R. T. Int. J. Pept. Protein Res. 1994, 44, 262–269.
18. Eberhardt, E. S.; Panasik, Jr., N.; Raines, R. T. J. Am. Chem.
Soc. 1996, 118, 12261–12266.
19. Gottlieb, A. A.; Fujita, Y.; Udenfriend, S.; Witkop, B.
Biochemistry 1965, 4, 2507–2513.
20. Demange, L.; Menez, A.; Dugave, C. Tetrahedron Lett. 1998,
39, 1169–1172.
21. Mitsunobu, O. Synthesis 1981, 1–28.
22. Tran, T. T.; Patino, N.; Condom, R.; Frogier, T.; Guedj, R.
J. Fluorine Chem. 1997, 82, 125–130.
23. Malkar, N. B.; Lauer-Fields, J. L.; Borgia, J. A.; Fields, G. B.
Biochemistry 2002, 41, 6054–6064.
Acknowledgements
24. Takai, M.; Kurano, Y.; Kimura, T.; Sakakibara, S. In Peptide
Chemistry 1979. Yonehara, H., Ed.; Protein Research
Foundation: Osaka, 1980; p 187.
We thank Mr M. Takagaki, Wakayama Industrial Technol-
ogy Center, for the characterization of synthetic compounds
with mass spectroscopy.
25. Jones, J. The Chemical Synthesis of Peptides. Oxford
University: Oxford, 1994; pp 37–38.
26. Narita, M.; Doi, M.; Nakai, T. Bull. Chem. Soc. Jpn 1987, 60,
3255–3259.
References
27. Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric
Identification of Organic Compounds. 5th ed. Wiley: New
York, 1991; pp 95–96.
1. Bornstein, P.; Traub, W.; 3rd ed. The Proteins, Neurath, H.,
Hill, R. L., Eds.; Academic: New York, 1979; Vol. IV, pp
411–632.
28. Hondrelis, J.; Lonergan, G.; Voliotis, S.; Matsoukas, J.
Tetrahedron 1990, 46, 565–576.
29. Moroder, L.; Hallet, A.; Wu
¨nsch, E.; Keller, O.; Wersin, G.
2. Ramachandran, G. N. In Aspects of Protein Structure.
Ramachandran, G. N., Ed.; Academic: New York, 1963; p 39.
3. Baum, J.; Brodsky, B. In Mechanisms of Protein Folding. 2nd
Hoppe Seyler’s Z. Physiol. Chem. 1976, 357, 1651–1653.