Assignment of absolute configuration of natural abundance deuterium signals associated with (R)- and (S)-enantioisotopomers in a fatty acid aligned in a chiral liquid crystal: Enantioselective synthesis and NMR analysis

Article abstract of DOI:10.1021/ja0617892

Previous experimental natural abundance deuterium (NAD) NMR results have shown an odd/even-related alternation in the (²H/¹H) ratio of the methylene groups of fatty acids (ChemBioChem 2001, 2, 425) and, by NAD NMR in CLC, a marked difference between enantiotopic deuterons for each methylenic site (Anal. Chem. 2004, 76, 2827). However, to date, the assignment of the absolute configuration for each deuterium has not been possible. To investigate further the origin of these effects, the assignment of NAD quadrupolar doublets observed in chiral oriented solvent is required. Here we describe the assignment of R- and S-isomers resulting from the isotopic substitution in positions 4 and 5 in the aliphatic chain of 1,1′- bis(thiophenyl)hexane 1 (BTPH) derived from natural linoleic acid of plant origin. This was achieved using an optimized synthetic strategy to obtain separately four regio- and stereoselectively deuterated enantiomers of BTPH. By reference to the deuterium spectra of these isotopically labeled reference compounds, we demonstrate that, on both 4 and 5 positions of BTPH, the isotopic enantiomers of S configuration are depleted relative to those of R configuration. This finding effectively explains the observed low ( ²H/¹H) ratio in NAD of some ethylenic sites of unsaturated fatty acids.

Full text of DOI:10.1021/ja0617892

Published on Web 08/08/2006
Assignment of Absolute Configuration of Natural Abundance
Deuterium Signals Associated with (R)- and
(S)-Enantioisotopomers in a Fatty Acid Aligned in a Chiral
Liquid Crystal: Enantioselective Synthesis and NMR Analysis
Vincent Baillif,† Richard J. Robins,† Isabelle Billault,*,† and Philippe Lesot*,‡
Contribution from the Laboratoire d’Analyse Isotopique et Electrochimique de Me´tabolismes,
CNRS UMR 6006, UniVersite´ de Nantes, BP 92208, 44322 Nantes, France, and Laboratoire de
Chimie Structurale Organique, CNRS UMR 8182, ICMMO, Baˆt. 410, UniVersite´ de Paris-Sud,
91405 Orsay Cedex, France
Received March 24, 2006; E-mail: isabelle.billault@univ-nantes.fr; philesot@icmo.u-psud.fr
Abstract: Previous experimental natural abundance deuterium (NAD) NMR results have shown an odd/
even-related alternation in the (2H/1H) ratio of the methylene groups of fatty acids (ChemBioChem 2001,
2, 425) and, by NAD NMR in CLC, a marked difference between enantiotopic deuterons for each methylenic
site (Anal. Chem. 2004, 76, 2827). However, to date, the assignment of the absolute configuration for
each deuterium has not been possible. To investigate further the origin of these effects, the assignment of
NAD quadrupolar doublets observed in chiral oriented solvent is required. Here we describe the assignment
of R- and S-isomers resulting from the isotopic substitution in positions 4 and 5 in the aliphatic chain of
1,1′-bis(thiophenyl)hexane 1 (BTPH) derived from natural linoleic acid of plant origin. This was achieved
using an optimized synthetic strategy to obtain separately four regio- and stereoselectively deuterated
enantiomers of BTPH. By reference to the deuterium spectra of these isotopically labeled reference
compounds, we demonstrate that, on both 4 and 5 positions of BTPH, the isotopic enantiomers of S
configuration are depleted relative to those of R configuration. This finding effectively explains the observed
low (2H/1H) ratio in NAD of some ethylenic sites of unsaturated fatty acids.
Introduction
However, deuterium NMR spectra of long-chain fatty acids
have too many overlapping signals to resolve fully all the
The diversity in the classes of fatty acids present in nature is
due to the large number of modifications of their chain
(introduction of a double bond or triple bond, conjugated double
bond, hydroxylation, epoxidation, etc.). Such reactions involve
small changes in the activity of enzymes of the same family
for which the enzymatic mechanisms are not completely
elucidated or are still unknown.1,2 The well-established technique
of quantitative 2H NMR has proved powerful for analyzing both
the natural deuterium distribution in bioproducts3-5 and kinetic
or physical isotope effects.6,7 This approach applied to long-
chain fatty acid biosynthesis has provided unique information
about potential enzyme mechanisms.8-12
hydrogen positions of interest: in particular, many of the
methylenic and ethylenic positions resonate at coincident fre-
quencies. To access a much larger number of unique deuterium
signals, we have developed a series of chemical modifications
of unsaturated methyl esters (methyl oleate and linoleate)8,9
isolated quantitatively from oils. This sample preparation,
followed by quantitative 2H NMR analysis, has provided detailed
data on the origin of hydrogens during elongation steps and
the mechanisms of primordial chain reactions, such as
desaturation,8-10 hydroxylation,11 and epoxidation.12
A notable feature of the (2H/1H) distribution observed in all
fatty acids so far analyzed is a true relative impoverishment at
one ethylenic site of each double bond.8,9 In yeast, a strong
† Universite´ de Nantes.
2
normal primary H-KIE has been shown to be associated with
‡ Universite´ de Paris-Sud.
(1) Hornung, E.; Pernstich, C.; Feussner, I. Eur. J. Biochem. 2002, 269, 4852.
(2) Behrouzian, B.; Buist, P. Phytochem. ReV. 2003, 2, 103.
(3) (a) Martin, G.; Martin, M. Tetrahedron Lett. 1981, 22, 3525. (b) Martin,
G.; Zhang, B.-L.; Naulet, N.; Martin, M. J. Am. Chem. Soc. 1986, 108,
5116. (c) Martin, G. In Stable Isotopes in the Biosphere; Wada, E.,
Yoneyama, T., Minagawa, M., Ando, T., Fry, B., Eds.; Kyoto University
Press: Kyoto, Japan, 1995, pp 36-58.
desaturation13a,b while no primary KIE has been measured from
plant ∆9 desaturase.14 The impoverishment in deuterium at one
(9) Duan, J.-R.; Billault, I.; Mabon, F.; Robins, R. ChemBioChem 2002, 3,
752.
(10) Guiet, S.; Robins, R. J.; Lees, M.; Billault, I. Phytochemistry 2003, 64,
(4) Schmidt, H.-L.; Werner, R.; Eisenreich, W. Phytochem. ReV. 2003, 2, 61.
(5) Robins, R.; Billault, I.; Duan, J.-R.; Guiet, S.; Pionnier, S.; Zhang, B.-L.
Phytochem. ReV. 2003, 2, 87.
227.
(11) Billault, I.; Mantle, P. G.; Robins, R. J. J. Am. Chem. Soc. 2004, 126,
3250.
(6) Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117, 9357.
(7) Zhang, B.-L.; Jouitteau, C.; Pionnier, S.; Gentil, E. J. Phys. Org. Chem.
2002, 106, 2983.
(12) Billault, I.; Duan, J.-R.; Guiet, S.; Robins, R. J. J. Biol. Chem. 2005, 280,
17645.
(13) (a) Buist, P.; Behrouzian, B. J. Am. Chem. Soc. 1996, 118, 6295. (b) Buist,
P.; Behrouzian, B. J. Am. Chem. Soc. 1998, 120, 871.
(8) Billault, I.; Guiet, S.; Mabon, F.; Robins, R. ChemBioChem 2001, 2, 425.
9
11180
J. AM. CHEM. SOC. 2006, 128, 11180-11187
10.1021/ja0617892 CCC: $33.50 © 2006 American Chemical Society

2H Signals of (R)-/(S)-Enantioisotopomer in Fatty Acids
A R T I C L E S
of the 2H signals associated with the even and odd sites (notated
4 and 5 in the aliphatic chain), which exhibit the same chemical
shift in isotropic NAD NMR spectra. Thus, NAD NMR in CLC
of compound 1 was demonstrated as a feasible method to
evaluate quantitatively the natural 2H distribution on pro-R and
pro-S C-H(2H) directions at a methylenic site within a fatty
acid chain.15
Although this technique is able to overcome the limitations
associated with classical isotropic methods, the absolute as-
signment of (R)- and (S)-isotopic enantiomers from the Q-COSY
2D maps is not possible when only anisotropic observables
(dipolar couplings, quadrupolar splittings) extracted from spectra
recorded in chiral oriented media are used.20 This is due to the
fact that enantiomers are mirror images; hence, for a coherent
set of anisotropic data, the use of the S- and R-geometries in
the calculation of order parameters does not affect the final
results, thus preventing any attribution of a data set to the
corresponding enantiomer when a racemic mixture is used.20,21
As a consequence, only chemical strategies enable the
absolute configuration associated with an NMR signal to be
determined. In our case, we have opted to synthesize selectively
each monodeuterated enantiomer with the desired absolute
configuration to assign the R- and S-stereodescriptors for
quadrupolar doublets observed at either the odd or the even
methylene groups of the aliphatic chain of 1.15
Figure 1. Structure, notation, and numbering of compounds investigated.
ethylenic site can be explained by two possible mechanisms: a
secondary 2H-KIE and/or a large difference in the (2H/1H) ratio
at the pro-R and pro-S sites at the methylenic group in the
substrate molecule. A direct measurement at the pro-R and pro-S
sites at the methylenic group cannot, however, be obtained from
conventional 2H NMR spectra, because in achiral solvents these
deuterium atoms are equivalent and hence always resonate at
the same frequency.
We have recently demonstrated the potential of the natural
abundance deuterium (NAD) 2D NMR technique, using a poly-
peptide chiral liquid crystal (CLC) as a weakly ordering solvent,
to access individual deuterons in a methylenic site of a long-
chain fatty acid.15 Initially developed for analyzing monodeu-
terated enantiomers,16 this original approach using commercially
available polypeptides, such as poly-γ-benzyl-L-glutamate (PBLG)
or poly-ꢀ-carbobenzyloxy-L-lysine (PCBLL), has also proved
to be a very powerful tool for distinguishing both between
enantiomers of chiral molecules by virtue of their isotopic
substitution 2H/1H17 and between enantiotopic directions in the
case of prochiral molecules.18 Strictly, for related parent
molecules, the discrimination of the former is a consequence
of the discrimination of the latter, as they both involve the same
molecular recognition mechanisms.18b
During our previous investigation, the 1,1′-bis(thiophenyl)-
hexane 1 (BTPH) derived from the natural linoleic acid of
safflower (Carthamus tinctorius) was examined (Figure 1).15
From a biosynthetic point of view, compound 1 is one of two
products obtained by chemical modification of methyl linoleate;8
hence it contains the natural deuterium substitution patterns of
its parent molecule.8,9
The NAD Q-COSY 2D spectrum of 1 recorded in the PBLG/
CHCl3 solvent on a 9.4 T NMR spectrometer has clearly shown
a good spectral discrimination of all monodeuterated enantio-
isotopomers on the basis of differences in quadrupolar splittings,
∆νQ.15,16,19 In addition, this NMR tool allowed the separation
In the present article, we first describe the synthesis of
monodeuterated racemic compounds 2 and 3 and of the chiral
compounds 2a, 2b, 3a, and 3b required for the assignment of
2H signals in the CLC NMR spectrum. Key to the strategy
adopted is the use of an elegant lipase-catalyzed resolution as
a simple, efficient, and general method to introduce stereo-
selectively a deuterium atom into an aliphatic chain. Second,
the deuterium assignments obtained with compounds 2a, 2b,
3a, and 3b are compared with previous results obtained from 1
using NAD NMR in CLC and discussed in relation to the
2
observed natural distribution of H within fatty acids.
Synthesis and NMR Results
Synthesis of Compounds. The synthesis of 2a/2b and 3a/
3b, described in Scheme 1, is based on the preparation and
lipase-catalyzed resolution of the racemic alcohols 8 and 9.
Detailed schemes are reported in the Supporting Information
(see Scheme SI1). Both alcohols were synthesized in four steps
using the same approach. The bromo- precursors possess a
terminal double bond to introduce the phenyldithioacetal func-
tion via an aldehyde. The chain length of both starting materials,
a bromoderivative and an aldehyde, defines the carbon number
of the final compound as well as the final position of the
deuterium atom in the chain, giving rise to 8 and 9.
From these key intermediates 8 and 9, the deuterium atom
was introduced in two steps in 60% yield by first a tosylation
(TsCl, pyridine) followed by a nucleophilic substitution in the
presence of LiAl2H4 leading to racemic 2 and 3.22
The enzymatic resolution of the racemic secondary alcohols
8 and 9 was carried out by stereoselective lipase-catalyzed
(14) Behrouzian, B.; Buist, P. J. Chem. Soc., Chem. Commun. 2001, 401.
(15) Lesot, P.; Aroulanda, C.; Billault, I. Anal. Chem. 2004, 76, 2827.
(16) Canet, I.; Courtieu, J.; Meddour, A.; Loewenstein, A.; Pe´chine´, J.-M. J.
Am. Chem. Soc. 1995, 117, 6520.
(20) Aroulanda, C.; Lesot, P.; Merlet, D.; Courtieu, J. J. Phys. Chem. A 2003,
107, 10911.
(21) Lesot, P.; Merlet, D.; Courtieu, J.; Emsley J. W.; Rantala, T. T.; Jokisaari,
J. J. Phys. Chem. A 1997, 101, 5719.
(22) (a) Brown, H. C.; Yoon, N.-M. J. Am. Chem. Soc. 1966, 88, 1464. (b)
Krishnamurthy, S.; Brown, H. C. J. Org. Chem. 1980, 45, 849. (c) Kabalka,
G. W.; Varma, M.; Varma, R. S. J. Org. Chem. 1986, 51, 2386.
(17) Meddour, A.; Canet, I.; Loewenstein, A.; Pe´chine´, J.-M.; Courtieu, J. J.
Am. Chem. Soc. 1994, 116, 9652.
(18) (a) Aroulanda, C.; Merlet, D.; Courtieu, J.; Lesot, P. J. Am. Chem. Soc.
2001, 123, 12059. (b) Lesot, P.; Lafon, O.; Courtieu, J.; Berdague´, P.
Chem.sEur. J. 2004, 10, 3741.
(19) Lesot, P.; Sarfati, M.; Courtieu, J. Chem.sEur. J. 2003, 9, 1724.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11181

A R T I C L E S
Baillif et al.
Scheme 1. Synthesis of Racemic Compounds 2 and 3 and Chiral Compounds 2a,2b and 3a,3ba
a Reagents and conditions: (a) Mg, Et2O, propionaldehyde; (b) Mg, Et2O, acetaldehyde; (c) OsO4, N-methylmorpholine oxide, CHCl3/H2O; (d) NaIO4,
MeOH; (e) PhSH, AlCl3, CH2Cl2; (f) TsCl, pyr.; (g) LiAl2H4, Et2O; (h) 1-Lipase, vinyl acetate, hexane, 2-NaOH, MeOH.
Table 1. ee % Obtained after Lipase-Catalyzed (PS-Amano)
Resolution of Alcohols 8 and 9
the diamagnetic effect of the phenyl ring, protons in the shielding
cone of this ring exhibit a H chemical shift, which is shifted
upfield relative to those not in this cone. Thus, on the basis of
1H NMR chemical shift analysis, the arrangement of the terminal
methyl of the alcohol can be as shown in Figure 2.
1
ee %
8
9
ee % residual alcohol (8a or 9a)
(isolated yield)
ee % alcohol after acetate hydrolysis (8b or 9b)
90
60
(43%)
98
(43%)a
(47%)
95
(40%)a
For example, the methyl proton in 14b that is cis-configured
relative to the phenyl ring of the Mosher’s acid moiety displays
a weaker chemical shift than that in 14a. Also, the CH(SPh)2
signal of 14a shows a smaller chemical shift than the same signal
in 14b. Thus, alcohols 8a and 8b are, respectively, S- and
R-configured. A similar procedure applied to compound 15
indicates that alcohols 9a and 9b are S- and R-configured,
respectively.
(isolated yield)
a Isolated yield of acetylated product.
acetylation. Enantio-enriched compounds 2a, 2b or 3a, 3b were
obtained respectively from racemic alcohols 8 and 9 after
enzymatic resolution and reduction (Scheme 1). The ee % of
the alcohols 8a, 8b, 9a, and 9b was measured by chiral HPLC
(Table 1) and confirmed by 2H-{1H} NMR in CLC on 2a, 2b,
3a, and 3b (see below).
The assignment of absolute configuration to the alcohols 8a
(S), 8b (R), 9a (S) and 9b (R) was performed by 1H NMR after
classic derivatization by Mosher’s acid chloride (see Scheme
SI2).23 The racemic alcohol 8 was esterified by (S)-MTPA-Cl
to give the Mosher’s esters mixture 14a/14b, and the racemic
alcohol 9 was esterified by (R)-MTPA-Cl to give the Mosher’s
esters mixture 15a/15b (Figure 2). The proton NMR spectra of
diastereoisomeric compounds 14a/14b and 15a/15b in each
mixture were compared. The difference in chemical shifts for
-CH(SPh)2 and -CH2CH3 signals between both diastereoiso-
mers were measured, and the ∆δ values are given in Table 2.
The absolute configuration at C4 for 14a or 14b, and at C5
for 15a or 15b, was assigned using the model of the preferential
conformation (Figure 2).23 This predicts that, on the basis of
Residual alcohols 8a and 9a obtained by lipase-catalyzed
resolution were submitted to derivatization with Mosher’s
chloride, providing the corresponding diastereoisomer. Their
associated 1H NMR spectra were compared with the reference
spectra obtained in racemic series, thus leading to an attribution
of their absolute configurations. Thus, alcohols 8a and 9a,
S-configured, are not substrates of the lipase PS “Amano” since
this enzyme preferentially reacts with the R-isomers 8b or 9b
to give the acetylated products. These results are in full
agreement with the known selectivity for the resolution of a
secondary alcohol using lipase PS “Amano”.24
Deuterated compounds 2a(R), 2b(S), 3a(R), and 3b(S) were
obtained, respectively, from alcohols 8a(S), 8b(R), 9a(S), and
9b(R) in two steps following a similar scheme to that used for
the preparation of racemic 2 and 3 (Scheme 1). As the
deuteration was performed by nucleophilic substitution, a strict
(23) (a) Dale, A. D.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512. (b) Barth,
G.; Voelter, W.; Mosher, H. S.; Bunnenberg, E.; Djerassi, C. J. Am. Chem.
Soc. 1970, 92, 875. (c) Seco, J. M.; Quinoa, E.; Riguera, R. Tetrahedron:
Asymmetry 2001, 12, 2915. (d) Seco, J. M.; Quinoa, E.; Riguera, R. Chem.
ReV. 2004, 104, 17.
(24) (a) Kaslauskas, R. J.; Weissfloch, A. N. E.; Rappaport, A. T.; Cuccia, L.
A. J. Org. Chem. 1991, 56, 2656. (b) Cygler, M.; Grochulski, P.;
Kaslauskas, R. J.; Schrag, J. D.; Bouthillier, F.; Rubin, B.; Serreqi, A. N.;
Gupta, A. K. J. Am. Chem. Soc. 1994, 116, 3180.
9
11182 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

2H Signals of (R)-/(S)-Enantioisotopomer in Fatty Acids
A R T I C L E S
Figure 2. Models explaining the variation in proton chemical shifts of the methyl and methylene groups.
Table 2. 1H NMR Chemical Shift and Absolute Configuration of
fact that order parameters are dependent on various external
parameters, such as temperature or the sample composition, the
comparison of the two sets of data (isotopically normal and
enriched compound) is only valid if the spectra are recorded
under the same experimental conditions. In our previous study,
we prepared the NMR sample with 14.3% w/w of polymer and
28.6% w/w of solute (see Table SI1 in Supporting Information).
To overcome the intrinsically low sensitivity of NAD NMR
spectroscopy, we used a large amount of isotopically normal
solute (200 mg).15 To keep the same w/w relative proportions
for samples with deuterated solutes, we have therefore prepared
a mixture of 1 isotopically normal (∼190 mg) and deuterated
compounds 2 or 3 (∼10 mg). As the orientational behaviors
for the two kinds of solute (see above) are identical, we can
assume a priori that we are comparing “identical” samples. It
could be argued that the compensation of deuterated solute by
a larger amount of chloroform to keep the % w/w of polymer
identical should also provide “identical” samples. In fact, this
option is not valid, since we have observed experimentally a
significant increase in quadrupolar splittings (around 30%) of
deuterated BPTH (2 and 3) compared with values measured on
the NAD NMR spectrum of 1 (see Tables SI2 and SI3 in the
Supporting Information). This suggests that molecules of BTPH
generate a larger orientational disorder inside the PBLG phase
compared to chloroform, thus reducing the magnitude of the
order parameters for the solute (as well as for the cosolvent).
From this, we clearly show that the effects of solute and
cosolvent on the liquid crystalline properties of the phase are
not identical. These differences can be attributed to differences
in dielectric constant, polarity, and molecular size of the BTPH
compared with chloroform, as well as the fact that BTPH does
not dissolve the PBLG fibers.
C4 for 14a/14b and of C5 for 15a/15b
δ(ppm)
δ(ppm)
14a
14b
∆δ
15a
15b
∆δ
CH(SPh)2
CH2CH3
4.35
0.92
4.39
0.79
-0.04 4.37
+0.13 1.26
4.30
1.34
(+)5-S (-)5-R
9a 9b
+0.07
-0.08
configuration (+)4-S (-)4-R
alcohol n° 8a 8b
inversion of configuration was expected.25 This was seen to have
occurred.
Study by 2H NMR in Chiral Liquid Crystal. Before
examining the 2H spectra of deuterated isotopomers 2a, 2b, 3a,
and 3b recorded in oriented solvents, the NAD NMR signals
associated with deuterons 4,4′ and 5,5′ of 1 dissolved in the
achiral phase PBG/CHCl3 and chiral phase PBLG/CHCl3 must
be described (see Figure 3).
Due to quadrupolar interaction that is no longer averaged to
zero as in an isotropic medium, the NMR signal of a deuterium
atom (spin I ) 1) in oriented phases is characterized by a
quadrupolar doublet, whose splitting between both components
is noted ∆νQ. In the achiral medium, where no enantiodiscrimi-
nation is possible,26 only two doublets centered on the same
chemical shift are observed for the methylene groups 4 and 5
(Figure 3a). The large difference in peak intensity enables both
doublets to be assigned, based on the well-characterized
differences in the 2H/1H ratio at even- and odd-numbered
positions in the fatty acid aliphatic chain. In contrast, in the
chiral oriented matrix where enantioselective interactions exist,
a doubling of quadrupolar doublets appears due to the discrimi-
nation of isotopic enantiomers (Figure 3b), and for both
methylene groups (4 and 5), the peak intensity (and the peak
area) is different. However, as emphasized in the Introduction,
it is not possible to assign the absolute configuration for each
quadrupolar doublet. We can now compare the various deuter-
ated isotopomers (2a, 2b, 3a, and 3b) associated with methylenic
groups 4 and 5 with the NAD NMR spectrum of 1. Due to the
Figure 4 shows the six 2H-{1H} 1D NMR spectra associated
with BTPH selectively deuterated on the methylene group 4 or
5 in racemic and enantio-enriched series. Spectral data of each
spectrum as well as data obtained for isotopically normal
compound 1 are described in the Supporting Information (Tables
SI2 and SI3).
The assignment of each quadrupolar doublet in methylene
groups 4 and 5 of (()-2 and (()-3 is made by a direct
comparison of the spectra with those recorded with enantio-
(25) Streitwieser, A. J., Jr.; Walsh, T. D.; Wolfe, J. R., Jr. J. Am. Chem. Soc.
1965, 87, 3682.
(26) Canlet, C.; Merlet, D.; Lesot, P.; Meddour, A.; Loewenstein, A.; Courtieu,
J. Tetrahedron: Asymmetry 2000, 11, 1911.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11183

A R T I C L E S
Baillif et al.
Figure 3. Quadrupolar doublets associated with deuterons 4,4′ and 5,5′ (sum of columns) extracted from the NAD Q-COSY 2D map of 1 dissolved in (a)
the PBG/CHCl3 achiral phase and (b) the PBLG/CHCl3 chiral phase (see ref 15 for experimental details). The quadrupolar splitting, ∆νQ, between both
components is equal to (3/2)e2 × QD × qC-D × SC-D where QD is the deuterium quadrupole moment, qC-D is the electric field gradient (EFG) along the
C-2H bond, and SC-D is the order parameter of the C-2H bond.
Discussion
In previous work, we have demonstrated that NAD 2D NMR
spectroscopy in chiral polypeptide liquid crystals (PBLG) can
be exploited to analyze the isotopic (2H/1H) ratio at pro-R and
pro-S sites in a methylene position in an aliphatic chain in a
natural fatty acid, via the derivative 1,1′-bis(thiophenyl)hexane.15
The initial experimental NMR results identified the same odd/
even variation in the 2H/1H ratio as that observed in the achiral
medium8,9 but, additionally, distinguished between the enan-
tiotopic deuterons for each stereogenic site. We have now
demonstrated that it is possible to assign the absolute config-
uration for each deuterium signal to each enantiotopic C-H(2H)
direction for all accessible stereogenic sites. To reach this goal,
we have developed an optimized synthetic strategy to obtain
separately four regio- and stereoselective deuterated enantiomers
in positions 4 and 5 in the aliphatic chain of BTPH and used
these to assign unambiguously the NAD NMR signals in the
PBLG phase.
Fatty acid biosynthesis is dominated by a succession of
enantioselective reactions in the FAS complex.27 In all organ-
isms, the origin of hydrogens in the aliphatic chain of fatty acids
is different between odd and even methylene sites and also
between pro-R and pro-S sites.28 At even methylene sites,
hydrogens are provided by acetate and water, while both
hydrogens at odd methylene sites are introduced by two
reduction steps, which involve a keto and an enoyl reductase
operating with the cofactor NAD(P)H. Moreover, each organism
displays a different stereochemistry for the introduction of
hydrogens from acetate or water and for the reduction steps;
Figure 4. 61.4 MHz 2H-{1H} 1D NMR spectrum of BTPH deuterated
thus the natural distribution of deuterium at pro-R and pro-S
should change in fatty acids isolated from different organisms
(mammals, plants, bacteria, algae, fungi). In contrast to mam-
malian and bacterial sources, the stereochemical sequence of
steps performed by the FAS is still not well described in plants,
algae, and fungi.29
on the methylene group 4 (left column) or 5 (right column) recorded at
300 K in the PBLG/CHCl3 phase. (a and d) Racemic sample corresponding
to 2 and 3, respectively; (b and e) R-enriched sample corresponding to 2a
and 3a, respectively; (c and f) S-enriched sample made up of 2b and 3b,
respectively. The 1D spectra of racemic and enantio-enriched mixtures were
recorded by adding 2700 and 4000 scans, respectively. For all spectra, no
filtering window was applied. The spectra are plotted with the same
frequency scale. The signal labeled with an asterisk corresponds to the
shielded component of the quadrupolar doublet of chloroform.
After elongation, important modifications of the fatty acid
chain occur, such as desaturation, hydroxylation, or epoxidation.
enriched compounds, 2a, 2b and 3a, 3b. For both methylene
positions, the comparison indicates that the inner and outer
quadrupolar doublets correspond to the R- and S-enantiomers,
respectively (Figure 4). As a consequence, the depletion
observed in the CLC NMR spectra of 1 corresponds to the
S-enantio-isotopomers. In other words, the pro-S site in both 4
and 5 prostereogenic methylene groups possesses the smaller
isotopic (2H/1H) ratio.
(27) Ohlrogge, J.; Browse, J. Plant Cell. 1995, 7, 957.
(28) (a) Saito, K.; Kawaguchi, A.; Nozoe, S.; Seyama, Y.; Okuda, S. Biochem.
Biophys. Res. Commun. 1982, 108, 995. (b) Sedgwick, B.; Cornforth, J.;
French, S.; Gray, R.; Kelstrup, E.; Willadsen, P. Eur. J. Biochem. 1977,
75, 481. (c) Sedgwick, B.; Cornforth, J. Eur. J. Biochem. 1977, 75, 465.
(e) Sedgwick, B.; Morris, C.; French, S. J. Chem. Soc., Chem. Commun.
1978, 193.
(29) (a) Sedgwick, B.; Morris C. J. Chem. Soc., Chem. Commun. 1980, 96. (b)
Saito, K.; Kawaguchi, A.; Seyama, Y.; Yamakawa, T.; Okuda, S. J.
Biochem. 1981, 90, 1697.
9
11184 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

2H Signals of (R)-/(S)-Enantioisotopomer in Fatty Acids
A R T I C L E S
natural abundance between the (2H/1H) ratios for the n and n +
1 ethylenic hydrogens at the positions of unsaturation.
The next question concerning fatty acid biosynthesis in plants
is to establish which hydrogen (pro-R or pro-S) comes from
water and which from acetate in the even methylene sites and
which comes from the reduction steps by a keto and an enoyl
reductase in the odd sites. As we demonstrate here, we have at
our disposal a very specific analytical method with which to
tackle this question. At present, the (2H/1H) ratios in fatty acids
isolated from Fusarium lateritium, an oleaginous fungus,
2
cultured in controlled media (enriched in H in water or in
glucose) are under investigation.
Figure 5. Comparison of the relative values of the (2H/1H) ratios in BTPH
with the pro-R and pro-S positions of the original methyl linoleate.
Experimental Section
Once again, these reactions are stereoselective. In the case of
desaturation, it has been reported that both pro-R hydrogens
are eliminated, giving rise to a cis-configured double bond.30
Previously, we have measured the (2H/1H) ratio at even and
odd sites of unsaturated fatty acids isolated from plants. In all
cases, odd ethylenic sites were strongly depleted while even
sites were not. Initially, we interpreted this observation as the
result of a strong secondary KIE on the odd site during the
desaturation step, considering the existence of a large difference
in the deuterium distributions between pro-R and pro-S sites in
the substrate molecule improbable, as both hydrogens are
introduced by two reduction steps with the same pool of cofactor
(NAD(P)H).
General Methods. Lithium aluminum deuteride, LiAl2H4 (98% 2H),
(R)- and (S)-R-methoxy-R-trifluoromethyl-phenyl-acetyl chloride (98-
99%) were purchased from Sigma-Aldrich; Lipases PS, AY, and GC20
were purchased from Amano. 1H and 13C NMR spectra were recorded
with a Bruker DRX 500 or DPX 400 spectrometer. Chemical shifts
are given in ppm from TMS. Flash chromatography was performed on
silica gel Normasil (40-60 µm, Prolabo). Reactions were followed by
gas chromatography on an HP-5 capillary column (30 m × 0.32 mm,
film thickness 0.52 µm); carrier gas, He; flow, 1.2 mL min-1; split,
1:40; FID temp, 280 °C; thermal gradient 100 °C initially for 1 min,
ramped at 10 °C min-1 to 240 °C for 1 min, ramped at 10 °C min-1 to
280 for 1 min, then ramped at 10 °C min-1 to 290, and held at 290 for
10 min. All solvents were distilled before use. Reactions were performed
at room temperature (20 °C). Enantiomeric excess (ee %) was
determined by HPLC on an HP series 1100 apparatus; UV detector
operating at 254 nm; Column: Chiracel AD-H (250 mm × 4.6 mm)
elution with hexane/2-propanol (95/5) at 0.5 mL min-1 for product 8;
Chiracel OD-H (250 mm × 4.6 mm) elution with hexane/2-propanol
(98/2) at 0.5 mL min-1 for product 9.
However, an initial analysis by CLC NMR indicated that this
second explanation is justified, one hydrogen of the methylene
group being richer in 2H than the other (see Table 3 in ref 15).
In the present study, we have extended this analysis and shown
2
that it is the pro-R hydrogens that are enriched in H and the
1,1′-Dithiophenyl-4-2H-hexane (2). To a solution of the racemic
tosylate 10 (140 mg, 0.3 mmol) in anhydrous Et2O (3 mL) was added
LiAl2H4 (37.8 mg, 0.9 mmol), and the reaction mixture was stirred for
2 h at room temperature. The reaction was cooled and then quenched
by the addition of several drops of water. The mixture was diluted
with Et2O and washed with H2O. The aqueous layer was extracted with
Et2O, and the combined organic phases were dried (MgSO4), filtered,
and evaporated. The crude residue was purified by flash chromatography
(petroleum ether) and gave 2 (66 mg, 72%) as a colorless oil. 1H NMR
(400 MHz, CDCl3) δ 0.89 (t, J ) 7 Hz, 3H), 1.29 (m, 3H), 1.62 (m,
2H), 1.87 (qd, J ) 7 Hz, 2H), 4.82 (t, J ) 6.7 Hz, 1H), 7.25-7.5 (m,
10H); 13C NMR (100 MHz, CDCl3) δ 14.1, 22.5, 26.8, 31.03 (t, J )
19 Hz), 36.0, 57.7, 127.8, 129.0, 132.9, 134.6. GC-MS (EI): 303,
194, 123, 109, 123, 84, 56. EI-MS: calcd ) 303.1220; found )
303.1299.
pro-S that are, relatively, depleted. Hence, as the stereochemistry
of desaturation exclusively eliminates the pro-R hydrogens
during double-bond formation,2,31 it can be expected that the
retained hydrogens will be impoverished relative to the mean
of the methylenic groups. The relative values of the (2H/1H)
ratios in BTPH can now be compared with the pro-R and pro-S
positions of the original methyl linoleate (Figure 5). For this
comparison, the (2H/1H) ratios of pro-S and pro-R of the 2, 3,
4, and 5 positions of BTPH are used, these being equivalent,
respectively, to the 14, 15, 16, and 17 positions of methyl
linoleate (see Figure 6).
From Figures 5 and 6, it can be seen that the predicted values
of 2H in sites 8 to 14 of methyl linoleate are in good agreement
(within 10-15 ppm) with the experimentally determined values.8
Notably, positions 9 and 13 are strongly impoverished relative
to the 10 and 12 positions. Thus, the apparent relative
impoverishment at the odd ethylenic positions is explained by
the ∆(2H/1H) for the pro-S and pro-R enantiotopic sites in the
methylenic groups from which they are derived.
(R)-1,1′-Dithiophenyl-4-2H-hexane (2a) and (S)-1,1′-Dithiophenyl-
4-2H-hexane (2b). These compounds were prepared from 10a-(S) and
10b-(R) following the same protocol as that for 2 from 10. Deuterated
compounds 2a-(R) (130 mg), from 10a-(S), and 2b-(S) (120 mg), from
1
10b-(R), were obtained in 66% and 64% yield, respectively. Their H
NMR and 13C NMR spectra were in complete agreement with spectra
obtained from the racemic mixture 2.
1,1′-Dithiophenyl-5-2H-hexane (3). To a solution of racemic tosylate
11 (118 mg, 0.25 mmol) in anhydrous Et2O (3 mL) was added LiAl2H4
(21 mg, 0.5 mmol), and the reaction mixture was stirred for 3 h at
room temperature. The reaction was cooled and then quenched by the
addition of several drops of water. The mixture was diluted with Et2O
and washed with H2O. The aqueous layer was extracted with Et2O,
and the combined organic phases were dried (MgSO4), filtered, and
evaporated. The crude residue was purified by flash chromatography
(petroleum ether) and gave 2 (69 mg, 90%) as a colorless oil. 1H NMR
(400 MHz, CDCl3): δ 0.88 (d, J ) 6.6 Hz, 3H), 1.27 (m, 3H), 1.62
Conclusion
In this work, we have shown for the first time that the (2H/
1H) ratio of pro-S enantiotopic sites in methylenic groups (even
and odd) within a fatty acid isolated from a plant (Carthamus
tinctorius) is strongly depleted relative to the pro-R. This
disparity is sufficient to explain the differences observed at
(30) McInnes, A.; Walker, J.; Wright, J. Tetrahedron 1983, 39, 3515.
(31) Behrouzian, B.; Buist, P. Curr. Opin. Chem. Biol. 2002, 6, 577.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11185

A R T I C L E S
Baillif et al.
Figure 6. Estimated (2H/1H) ratios from the height of doublet peaks from CLC 2H NMR spectra of 1 (ref 15). (2H/1H)pro-S at odd sites is the mean of
(2H/1H) values found from sites 3S and 5S, (2H/1H)pro-R at odd sites is the mean of (2H/1H) values found from sites 3R and 5R, (2H/1H)pro-S at even sites
is the mean (2H/1H) values found from sites 2S and 4S, and (2H/1H)pro-R at even sites is the mean (2H/1H) values found from sites 2R and 4R.
(qt, J ) 7.5 Hz, 2H), 1.87 (qd, J ) 7 Hz, 2H), 4.42 (t, J ) 6.7 Hz,
1H), 7.25-7.5 (m, 10H). 13C NMR (100 MHz, CDCl3): δ 14.2, 22.4
(t, J ) 19.07 Hz), 27.1, 31.5, 36.2, 58.9, 128.0, 129.2, 133.0, 134.8.
GC-MS (EI): 303, 194, 123, 109, 123, 84, 56. EI-MS: calcd )
303.1220; found ) 303.1201.
1H NMR spectrum was in complete agreement with the 1H NMR
spectrum of the racemic alcohol 8. The ee % of 8b measured by chiral
HPLC was 98%. [R]D ) -4.5 (4.9 × 10-3, CHCl3, ee % ) 98%).
20
(+)-(S)-1,1′-Dithiophenyl-5-hydroxy-hexane (9a) and (R)-1,1′-
Dithiophenyl-5-acetyl-hexane (13). To a solution of the racemic
alcohol 9 (330 mg, 1.0 mmol) in hexane (15 mL) were added the lipase
PS Amano (0.6 g) and vinyl acetate (78 µL, 0.84 mmol). The mixture
was stirred at room temperature. After 24 h, the conversion [%] of 9
to acetylated compound 13 reached about 50 (measured by GC). The
reaction mixture was filtered, and the solvent was evaporated under a
vacuum. The crude residue was purified by flash chromatography to
give alcohol 9a (major configuration S) (0.41 mmol, 47%) and 13
(R)-1,1′-Dithiophenyl-5-2H-hexane (3a) and (S)-1,1′-Dithiophenyl-
4-2H-hexane (3b). These compounds were prepared from 11a-(S) and
11b-(R) following the same protocol for 3 from 11. Deuterated
compounds 3a-(R) (150 mg), from 11a-(S), and 3b-(S) (110 mg), from
1
11b-(R), were obtained in 69% and 73% yield, respectively. Their H
NMR and 13C NMR spectra were in complete agreement with spectra
obtained from the racemic mixture 3.
1
(major configuration R) (0.48 mmol, 40%) as colorless oils. The H
Enzymatic Resolution. (+)-(S)-1,1′-Dithiophenyl-4-hydroxy-hex-
ane (8a) and (R)-1,1′-Dithiophenyl-4-acetyl-hexane (12). To a solution
of the racemic alcohol 8 (576 mg, 1.81 mmol) in hexane (30 mL) were
added the lipase PS Amano (1 g) and vinyl acetate (157 µL, 1.7 mmol).
The mixture was stirred at room temperature. After 24 h, the conversion
[%] of 8 to acetylated compound 12 reached about 50 (measured by
GC). The mixture reaction was filtered, and the solvent was evaporated
under a vacuum. The crude residue was purified by flash chromatog-
raphy to give the residual alcohol 8a (major configuration S) (275 mg,
0.86 mmol, 47%) and 12 (major configuration R) (283 mg, 0.786 mmol,
1
NMR spectrum of 9a was in complete agreement with the H NMR
spectrum of the racemic alcohol 9. The ee % of 9a measured by chiral
HPLC was 60% to 94.5%. [R]D20 ) +3.5 (5.8 × 10-3, CHCl3, ee % )
94.5%).
1
NMR data of 13: H NMR (500 MHz, CDCl3): δ 1.21 (d, J ) 6.2
Hz, 3H), 1.43-1.71 (m, 4H), 1.87 (m, 2H), 2.03 (s, 3H), 4.40 (t, J )
6.6 Hz, 1H), 4.88 (sp, J ) 6.3 Hz, 1H), 7.25-7.5 (m, 10H). 13C NMR
(125 MHz, CDCl3): δ 20.1, 22.1, 23.7, 36.0, 36.5, 59.1, 71.4, 128.5,
129.7, 133.5, 134.9, 171.4.
1
44%) as colorless oils. The H NMR spectrum of 8a agreed with the
(-)-(R)-1,1′-dithiophenyl-5-hydroxy-hexane (9b). To a solution
of 13 (150 mg, 0.416 mmol) in MeOH (5 mL) was added NaOH (80
mg, 2.08 mmol), and the mixture was stirred at room temperature. After
3.5 h, the solvent was evaporated, and the crude residue was diluted in
CH2Cl2. The organic phase was washed with water, and the aqueous
phases were extracted with CH2Cl2. The combined organic phases were
dried (MgSO4), filtered, and evaporated to afford 9b-(R) (108 mg, 82%).
The 1H NMR spectrum was in complete agreement with the 1H NMR
spectrum of the racemic alcohol 9. The ee % of 9b measured by chiral
1H NMR spectrum of the racemic alcohol 8. The ee % of 8a measured
by chiral HPLC was 90% to 98%. [R]D20 ) +3.2 (5.4 × 10-3, CHCl3,
ee % ) 97%).
NMR data of 12: 1H NMR (500 MHz, CDCl3) δ 0.87 (t, J ) 7.4
Hz, 3H), 1.54 (qt, J ) 7.2 Hz, 2H), 1.83-1.89 (m, 4H), 1.97 (s, 3H),
4.42 (t, J ) 6.1 Hz, 1H), 4.78 (m, 1H), 7.2-7.5 (m, 10H); 13C NMR
(125 MHz, CDCl3) δ 9.7, 21.3, 27.2, 30.9, 31.5, 58.3, 74.7, 128.03,
129.1, 133.0, 133.1, 134.1, 134.2, 171.0.
20
HPLC was 95% to 96%. [R]D ) -3.8 (6.4 × 10-3, CHCl3, ee % )
(-)-(R)-1,1′-Dithiophenyl-4-hydroxy-hexane (8b). To a solution
of 12 (283 mg, 0.786 mmol) in MeOH (10 mL) was added NaOH
(160 mg, 4 mmol). After the solution was stirred overnight at room
temperature, solvent was evaporated and the crude residue was diluted
with CH2Cl2 and then washed with water. The aqueous layer was
extracted twice with CH2Cl2, and the combined organic phases were
dried (MgSO4), filtered, and evaporated. The crude residue was filtered
on silica gel (CH2Cl2) to give the alcohol 8b-(R) (244 mg, 98%). The
96%).
NMR in Oriented Solvent. The various samples were composed
of around 100 mg of PBLG (Sigma) with MW ∼171 300 g/mol (DP
) 782), ∼200 mg of solute mixture, and ∼700 mg of dry CHCl3. The
exact composition for each oriented NMR sample is given in Table
SI1 of the Supporting Information. The mixture components were
directly weighed into 5-mm NMR tubes. Then they were homogenized
9
11186 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

2H Signals of (R)-/(S)-Enantioisotopomer in Fatty Acids
A R T I C L E S
by centrifugation (see ref 15) until an optically homogeneous birefrin-
and D. Maume (LABERCA, EVN, Nantes) for recording IR,
SM, and HR-SM spectra, respectively. This work was financed
in part by a grant from the French Ministry of Education and
Research.
2
gent phase was obtained. The H-{1H} 1D NMR experiments were
performed at 9.4 T on a Bruker DRX 400 high-resolution spectrometer
equipped with a selective 2H probe operating at 61.4 MHz for deuterium
and a standard variable temperature unit. Sample temperature was
carefully controlled at 300 K. The protons were broadband decoupled
using the WALTZ-16 composite pulse sequence. For other details on
the method, see refs 15, 16, and 19.
Supporting Information Available: Details of synthesis.
2
Experimental data concerning H NMR in the chiral liquid-
crystalline phase. This material is available free of charge via
the Internet at http://pubs.acs.org.
Acknowledgment. We are grateful to Prof. C. Rabiller for
advice and discussion. The authors thank M.-J. Bertrand for ee
% measurements by chiral HPLC and D. Legoff, G. Nourrisson,
JA0617892
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11187