Cork suberin molecular structure: Stereochemistry of the C18 epoxy and vic-diol ω-hydroxyacids and α,ω-diacids analyzed by nmr

Article abstract of DOI:10.1021/jf400577k

Suberin is the biopolyester that protects the secondary tissues of plants against environmental variability and aggressions. Cork suberin is composed mostly of C₁₈ ω-hydroxyacids and α,ω-diacids, 9,10-substituted with an unsaturation, an epoxide ring, or a vic-diol group. Although determinant for suberin macromolecular structure, the stereochemistry of these monomers is poorly studied, sometimes with contradictory results. An NMR technique was used here to assign the configuration of the 9,10-epoxy and 9,10-diol groups in C₁₈ suberin acids, comparing the chemical shifts of diagnostic ¹H and ¹³C signals with the ones of model compounds, before and after conversion of the vic-diol group into benzylidene acetal derivatives. The relative configuration was proved to be cis in the C₁₈ 9,10-epoxy and threo in the C₁₈ 9,10-diol suberin acids. These monomers were present in suberin probably as racemic mixtures, as shown by polarimetry. The revealed stereochemistry allows the suberin macromolecule to be built as an ordered array of midchain kinked C₁₈ acids, reinforced by intramolecular hydrogen bonding.

Full text of DOI:10.1021/jf400577k

Article
pubs.acs.org/JAFC
Cork Suberin Molecular Structure: Stereochemistry of the C18 Epoxy
and vic-Diol ω‑Hydroxyacids and α,ω-Diacids Analyzed by NMR
Sara Santos,* Vanessa Cabral, and Jose Graca
̧
́
́
Centro de Estudos Florestais Instituto Superior de Agronomia, Universidade Tecnica de Lisboa, 1349-017 Lisboa, Portugal
S
* Supporting Information
ABSTRACT: Suberin is the biopolyester that protects the secondary tissues of plants against environmental variability and
aggressions. Cork suberin is composed mostly of C18 ω-hydroxyacids and α,ω-diacids, 9,10-substituted with an unsaturation, an
epoxide ring, or a vic-diol group. Although determinant for suberin macromolecular structure, the stereochemistry of these
monomers is poorly studied, sometimes with contradictory results. An NMR technique was used here to assign the configuration
of the 9,10-epoxy and 9,10-diol groups in C18 suberin acids, comparing the chemical shifts of diagnostic 1H and 13C signals with
the ones of model compounds, before and after conversion of the vic-diol group into benzylidene acetal derivatives. The relative
configuration was proved to be cis in the C18 9,10-epoxy and threo in the C18 9,10-diol suberin acids. These monomers were
present in suberin probably as racemic mixtures, as shown by polarimetry. The revealed stereochemistry allows the suberin
macromolecule to be built as an ordered array of midchain kinked C18 acids, reinforced by intramolecular hydrogen bonding.
KEYWORDS: suberin, cork, C18 9,10-epoxy suberin acids, C18 9,10-diol suberin acids, vic-diols stereochemistry determined by NMR,
benzylidene acetal derivatives
Although all suberins studied so far from different plant
species and tissues show the same basic composition pattern of
glycerol, ω-hydroxyacids, and α,ω-diacids, they can be quite
variable in the specific suberin acids of which they are made.
Suberin ω-hydroxyacids and α,ω-diacids typically have even-
numbered carbon chain lengths between C16 and C24.2,4 There
are suberins with significant amounts of C16 acids, while others
include C22 or longer acids as relevant monomers.2 However,
all analyzed suberins have C18 suberin acids in significant
amounts, meaning that they play a major role in its
macromolecular structure. These C18 suberin acids are specific
because they all have a midchain “functional” group, at carbons
C-9 and C-10: this can be a double-bond or an oxygen-
containing substituent group, either an epoxide ring or two
hydroxyl groups in vicinal positions (vic-diol). Because these
secondary substituents impart chirality to the carbons they are
attached to, these suberin acids can have different stereo-
chemical configurations.
Understanding the stereochemistry of the C18 suberin acids is
important for two main reasons: first, the spatial arrangement
of the C18 monomers within the polyester macromolecule is
dependent on their spatial configuration and conformation;3
second, because these suberin acids are potentially interesting
for a number of valuable industrial uses, their stereochemistry
will also determine their properties, reactivity, and possible
applications.1,5
INTRODUCTION
■
Suberin is a lipid macromolecule found in suberized plant cell
walls, which act as a frontier barrier toward environmental
variability and aggressions.1 Suberized cells comprise most of
the periderm, the tissue that envelops plant organs of secondary
growth, like tree trunks or potato tubers. Suberized cells also
develop as wound tissue in plants, after injury or leaf fall, and in
internal organs where water flow control is needed, like in the
Casparian bands of root endodermis.1 The epitome of
suberized plant tissues is cork, and the outer bark of the
cork-oak (Quercus suber L.), cork, is harvested in a renewable
and sustainable way with growth cycles of 9 years. As a material,
cork has a unique set of properties, which makes it used
worldwide in an enormous number of products, for some of
them without an alternative substitute.
The barrier and insulation properties of suberin, which are
vital for plant survival, derive ultimately from its molecular and
supramolecular structure in suberized cell walls. In spite of its
importance and the significant quantity of knowledge already
gathered [see comprehensive reviews in refs 2 and 3], many
doubts remain about how suberin is built up as a macro-
molecule.
Suberin is known to be a glycerolipid polyester: when the
ester bonds are broken, suberin depolymerization occurs,
releasing its constituent “monomers”. Besides glycerol, the
major suberin monomers are long-chain aliphatic acids,
collectively named as “suberin acids”, which typically account
for 80% or more of its depolymerization products.4 Within the
latter, two families of α,ω-bifunctional fatty acids are largely
dominant: ω-hydroxyacids and α,ω-diacids. The fact that ω-
hydroxyacids and α,ω-diacids have either hydroxyl or carboxylic
acid linking groups at both chain ends, allows the macro-
molecular polyester to grow.3
The focus of the present work was the stereochemistry of the
C18 suberin acids with oxygen-containing substituents, namely,
9,10-epoxide and 9,10-diol groups, found in cork suberin [the
Received: February 5, 2013
Revised: June 7, 2013
Accepted: June 24, 2013
Published: June 24, 2013
© 2013 American Chemical Society
7038
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
Figure 1. Structural representation of the stereochemistry of the C18 9,10-epoxy and C18 9,10-diol acids: possible relative (cis/trans and erythro/threo)
and absolute configurations.
stereochemical analysis of the C18 9-unsaturated has been
presented and discussed elsewhere6]. Cork suberin has
significant amounts of C18 9,10-epoxy and C18 9,10-diol suberin
acids, together representing ca. 44% of the suberin mass.4,7
They include the C18 epoxy-ω-hydroxyacid, 9,10-epoxy-18-
hydroxyoctadecanoic (from now on referred to as Hyd18Epox),
the C18 epoxy-α,ω-diacid, 9,10-epoxy-octadecane-1,18-dioic
(Di18Epox), the C18 diol-ω-hydroxyacid, 9,10,18-trihydroxy-
octadecanoic (Hyd18Diol), and the C18 diol-α,ω-diacid, 9,10-
dihydroxyoctadecane-1,18-dioic (Di18Diol). Each of these
epoxide and vic-diol suberin acids has two asymmetric carbons
and therefore can exist in the form of up to four different
stereoisomers (Figure 1). The suberin epoxyacids can have a cis
or trans relative configuration, each one with two possible
absolute configurations in the form of enantiomeric pairs: S,R
or R,S for the cis isomer and S,S or R,R for the trans isomer
(Figure 1). Because of the molecule symmetry, cis-Di18Epox is
a meso compound since the two enantiomers are identical
(Figure 1). The relative configuration of the C18 9,10-diol
suberin acids can be described using the Fischer projection, as
threo (with possible absolute configurations of R,R or S,S) or
erythro (R,S or S,R) (Figure 1). The erythro-Di18Diol is also
meso because the R,S and S,R forms are indistinguishable
(Figure 1).
Except for a few pioneering works dating back to the middle
of the last century, the stereochemistry of these C18 9,10-epoxy
and C18 9,10-diol acids in suberin have never been thoroughly
studied. The relative configuration of the Hyd18Diol (by then
also known as “phloionolic acid”) and Di18Diol (“phloionic
acid”) was first studied by Seoane and co-workers.8−11 By
comparing the melting points of erythro and threo vic-diols
synthesized through stereospecific reactions from unsaturated
acids of known stereochemistry, with the ones isolated from
cork, and following the empiric rule that erythro isomers have
7039
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
higher melting points than the corresponding threo isomers, the
threo configuration was proposed for both the Hyd18Diol and
Di18Diol suberin acids.9,11,12 On the contrary, in a more recent
publication, both threo and erythro Hyd18Diol and Di18Diol
suberin acids were identified in cork, although it was not
mentioned how the configuration was assigned.13 In this latter
work, cork suberin was depolymerized by methanolysis and the
solubilized suberin monomers analyzed by GC-MS, with the
assigned threo and erythro forms chromatographically resolved;
on average, for both vic-diols, the threo and erythro forms were
found in almost identical amounts.13
The only known study on the relative stereochemistry of the
C18 9,10-epoxy suberin acids was also from Seoane and co-
workers.14 These authors isolated the Hyd18Epox and
Di18Epox acids from cork suberin and hydrolyzed the epoxide
rings into vic-diols; the latter showed melting points similar to
the Hyd18Diol and Di18Diol suberin acids, to which these
same authors had previously assigned a threo configuration, as
discussed above; due to the stereospecific anti-hydroxylation
reaction used in the hydrolysis, a cis configuration was assigned
to the epoxide suberin acids.14
In order to get unambiguous and definite results on the
stereochemistry of the C18 9,10-epoxy and C18 9,10-diol suberin
acids from cork, in the present work we assessed this question
using NMR-based techniques. NMR has been successfully used
for the determination of both the relative and absolute
stereochemistry of organic compounds.15−17 The general
strategies include the use of chiral solvents or additives,
without covalent linking to the stereoisomer(s) under analysis;
the use of chiral derivatizing agents, with covalent linkage to the
substrates, sometimes using the two enantiomers of the CDA;
or the use of the nuclear Overhauser effect (NOE) through
space interactions between nuclei, which can show the atoms’
relative positions, particularly in rigid systems. In the case of the
two C18 9,10-epoxy and the two C18 9,10-diol suberin acids
under consideration here, their stereochemistry analysis by
NMR faced a number of situations: each compound has two
stereogenic carbons that can, in the case of the 9,10-diols, rotate
around the σ bond linking them therefore with many possible
spatial conformations; the chemical neighborhood on both
sides of the chiral centers are identical methylene chains up to
seven carbons away, giving symmetrical effects in many
approaches; there was no previous safe knowledge if one or
two of the possible relative configurations were present, and if
both, in what relative proportions, and neither if each of the
former includes one or two of its possible enantiomers (Figure
1).
The analysis of the stereochemical configuration of vicinal
secondary hydroxyls with symmetric alkyl substituents, like the
vic-diols we are dealing with here, was successively achieved by
NMR, after the esterification of both hydroxyls with the R and
S enantiomers of arylmethoxyacetic acids; the chemical shifts of
the protons linked to the chiral centers in the (R,R)- and (S,S)-
derivatives, namely, the sign of the δR-δS differences for the
same proton, allowed the unequivocal assignment of the
absolute configuration of those diols.16,18 This same approach
was essayed with the suberin acids vic-diols, but the methyne
protons linked to the hydroxyls, after esterification with R- and
S-methoxyphenylacetic acid (MPA), showed as unresolved
broad multiplets whose chemical shifts were identical in the
(S,S)- and (R,R)-MPA derivatives (data not shown).
compare the most diagnostic 1H and 13C chemical shifts of the
C
18 9,10-epoxy and C18 9,10-diol suberin acids with the ones of
model compounds with known relative stereochemistry. The
model compounds used included the C18 9,10-epoxy and C18
9,10-diol substituted octadecanoic (stearic) acids, structurally
identical to the suberin acids in what concerns the stereogenic
carbons and their vicinity (Figure 1). Identical C18 vic-diol
model compounds were used, but starting from C18 9-
unsaturated and C18 9,10-epoxy compounds with known
configuration and transforming them into vic-diols with
determined stereochemistry through stereospecific reactions,
to check the validity of the latter.
The NMR analysis was applied to the suberin acids and
model compounds as such but also carried out after converting
the vic-diols into benzylidene acetals (BzAc), a group which is
normally used to protect secondary vic-diols.19 The analysis of
the suberin epoxyacids stereochemistry was also done using this
same derivative, after conversion of the epoxides into vic-diols
through a stereospecific reaction. The advantage of the use of
BzAc derivatives is that the five-sided ring they form with the
vic-diol has restricted conformational mobility, thus keeping the
carbons and protons of interest for the NMR stereochemistry
analysis in relatively fixed positions.
The implications of the stereochemistry of the C18 midchain
oxygenated suberin acids for the macromolecular arrangement
of suberin are discussed, and a model is proposed based on
such results.
MATERIALS AND METHODS
■
Cork C18 9,10-Epoxy and C18 9,10-Diol Suberin Acids. The
two C18 9,10-epoxy and the two C18 9,10-diol suberin acids, ω-
hydroxyacids (identified by the prefix Hyd) and α,ω-diacids (Di), in
the form of methyl esters, were obtained from cork after a
methanolysis depolymerization reaction. These compounds were
isolated and purified through a process of successive extraction and
purification steps (method under patent submission) and obtained
with the following purities as determined by GC-MS analysis: 9-
epoxyoctadecanedioic acid, dimethyl ester (Di18Epox_Me), 95%; 9-
epoxy-18-hydroxyoctadecanoic acid, methyl ester (Hyd18Epox_Me),
97%; 9,10-dihydroxyoctadecanedioic acid, dimethyl ester (Di18-
Diol_Me), 99%; and 9,10,18-trihydroxyoctadecanoic acid, methyl
ester (Hyd18Diol_Me), 99%. The four C18 suberin acids were
characterized by mass spectrometry (EIMS), FTIR, and 1D NMR (1H
and 13C), and 2D NMR (COSY, HSQC, and HMBC) spectroscopy.
Model Compounds. C18 9,10-Epoxy and C18 9,10-diol monoacids
of known relative stereochemistry were used as model compounds and
purchased as pure standards: cis-9,10-epoxyoctadecanoic acid, 99%,
from Sigma-Aldrich (Lisboa, Portugal); rac-trans-9,10-epoxyoctadeca-
noic acid, 98%, and rac-threo-9,10-dihydroxyoctadecanoic acid, 98%,
from Santa Cruz Biotechnology (Heidelberg, Germany); and erythro-
9,10-dihydroxyoctadecanoic, 97%, from Larodan Fine Chemicals
(Malmoe, Sweden). The C18 9-unsaturated monoacids, cis-oleic acid
methyl ester and trans-elaidic acid methyl ester, 99%, were purchased
from Sigma-Aldrich. Two other compounds with a vic-diol group
substituted in an alkyl chain of known relative stereochemistry, were
used, namely, erythro-5,6-dodecanediol, 98%, and threo-5,6-dodeca-
nediol, 96%, purchased from TCI Europe (Zwijndrecht, Belgium). All
solvents were of HPLC grade and obtained from Merck (Lisboa,
Portugal).
GC-MS Analysis. Separation of cis/trans Isomeric Pairs. cis- and
trans-9,10-epoxyoctadecanoic acid mixtures were prepared for GC-MS
analysis from equimolar solutions of each of the two acids as follows:
1.0 mg (0.3 mmol) of cis-9,10-epoxyoctadecanoic acid was diluted in
130 μL of pyridine and 130 μL of N,O-bis(trimethylsilyl)-
trifluoroacetamide (BSTFA); an identical solution was prepared for
To allow an unambiguous assignment of the suberin acids’
relative configuration, the approach followed here was to
7040
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
trans-9,10-epoxyoctadecanoic acid; aliquots of 100 μL of each solution
were mixed in a vial to which 800 μL of pyridine was added.
Separation of erythro/threo Isomeric Pairs. Solutions for GC-MS
analysis of equimolar mixtures of threo- and erythro-9,10-dihydroxy-
octadecanoic acid, and threo- and erythro-5,6-dodecanediol were
identically prepared as described above.
Analysis of Suberin Acids. Solutions of each of the suberin acids, 9-
epoxyoctadecanedioic acid, dimethyl ester (Di18Epox_Me), 9-epoxy-
18-hydroxyoctadecanoic acid, methyl ester (Hyd18Epox_Me), 9,10-
dihydroxyoctadecanedioic acid, dimethyl ester (Di18Diol_Me), and
9,10,18-trihydroxyoctadecanoic acid, methyl ester (Hyd18Diol_Me)
were prepared for GC-MS analysis by dissolving 1.0 mg of each
suberin acid in the TMS-derivatizing solution of 130 μL of pyridine
and 130 μL of BSTFA.
GC-MS Conditions. All of the above solutions were injected in a
7890A gas chromatographer coupled to a 5975C mass spectrometer
detector (Agilent Technologies, USA) in the following chromato-
graphic conditions: the column was DB5-MS (60 m, internal diameter
0.25 mm, film thickness 0.25 μm); the oven program was with an
initial temperature of 200 °C, followed by a temperature increase to
250 °C at 8 °C/min, and then to 300 °C at 3 °C/min; the final
temperature was kept for 15 min; injections were made in splitless
mode, with an injector temperature of 300 °C. Mass spectrometer
conditions were electron ionization, 70 eV; source temperature, 230
°C; and quadrupole temperature, 150 °C; the transfer line temperature
was kept at 310 °C.
Hydroxylation of the C18 9-Unsaturated Monoacids into C18
9,10-Diols. vic-Diols of determined stereochemistry were obtained
through a stereospecific syn-hydroxylation of the C18 9-unsaturated
monoacid model compounds, namely, cis-oleic acid methyl ester
(converted to the erythro-9,10-diol) and trans-elaidic acid methyl ester
(converted to the threo-9,10-diol) by permanganate oxidation
following a procedure adapted from Bhushan et al.20
Hydrolysis of the C18 9,10-Epoxyacids into C18 9,10-Diols. A
stereospecific anti-hydroxylation converting the epoxides into vic-diols,
namely, the cis-9,10-epoxides into threo-9,10-diols and the trans-9,10-
epoxides into erythro-9,10-diols, was applied to the C18 9,10-epoxide
model compounds (cis- and trans-9,10-epoxyoctadecanoic acids) and
to the C18 9,10-epoxy suberin acids (Hyd18Epox_Me and Di18-
Epox_Me). The anti-hydroxylation reaction was carried out by an
acid-catalyzed hydrolysis, described as follows: 50.1 mg (0.2 mmol) of
cis-9,10-epoxyoctadecanoic acid was dissolved in 5 mL of tetrahy-
drofuran and 5 mL of water, and 500 μL of sulphuric acid was added
dropwise. The mixture reacted at room temperature for 72 h with
stirring, and after the reaction completion, 20 mL of water and 20 mL
of dichloromethane were added and the organic phase recovered. After
solvent removal, 42 mg of threo-9,10-dihydroxyoctadecanoic acid in a
purity of 72% was recovered from the organic phase and further
purified in 0.5 mm TLC silica plates with hexane/isopropyl alcohol/
formic acid 70:30:1. The band with Rf = 0.8 was collected, extracted
with methanol/dichloromethane 1:1, giving 12.0 mg of threo-9,10-
dihydroxyoctadecanoic acid methyl ester with a purity of 90%. The
same procedure was applied to 10.0 mg (0.03 mmol) of trans-9,10-
epoxyoctadecanoic acid with a final recovery of 10.0 mg of erythro-
9,10-dihydroxyoctadecanoic acid (88% pure); 17.8 mg (0.05 mmol) of
Hyd18Epox_Me, resulting in 16.0 mg of recovered organic phase with
Hyd18Diol in 98% purity, which was methylated afterward as
described below; and 62.8 mg (0.2 mmol) of Di18Epox_Me resulting
in 44.7 mg of Di18Diol in 97% purity, also further methylated.
Methylation of the Free Acids. Some of the purchased model
compounds, as well as their synthesized derivatives, when in the form
of free acids, were methylated (to avoid signal overlapping with
diagnostic signals of interest in the NMR spectra) in methanol with
1% sulphuric acid, for 3 h, in reflux at 80 °C.
Derivatization of vic-Diol Groups into Benzylidene Acetals
(BzAc). All vic-diol compounds were converted to the corresponding
BzAc derivatives, including the C18 9,10-diol suberin acids
(Di18Diol_Me and Hyd18Diol_Me); the equivalent C18 9,10-diols
obtained from the anti-hydroxylation of the C18 9,10-epoxy suberin
acids (Di18Epox_Me and Hyd18Epox_Me); the purchased erythro-
and threo-C18 9,10-dihydroxyoctadecanoic acids; the latter two
obtained from (i) the syn-hydroxylation of the model compounds
cis-oleic acid methyl ester and trans-elaidic acid methyl ester and (ii)
from the anti-hydroxylation of the model compounds trans- and cis-C18
9,10-epoxyoctadecanoic acids; and the model compounds C12 erythro-
5,6-dodecanediol and C12 threo-5,6-dodecanediol.
Cetyltrimethylammonium Permanganate (CTAP) Preparation.
KMnO4 (5.0 g (31.7 mmol)) was dissolved in 160 mL of water. In
another flask, 12.6 g (34.7 mmol) of cetyltrimethylammonium
bromide (CTAB) was dissolved in 155 mL of water during 1 h with
stirring. This CTAB solution was added slowly and dropwise to the
KMnO4 solution for 1 h, with strong stirring. The mixture was allowed
to further react for 30 min and the resulting violet precipitate
recovered by filtration on a G3-porosity glass filter and washed with
water. This residue (CTAP) was dried on a vacuum oven over P4O10.
Dry CTAP (10.2 g) was ground to a fine powder and kept in a
refrigerator.
BzAc Derivatization Reaction. To prepare the BzAc derivatives, a
procedure adapted from McElhanon et al.21 was used. To a known
quantity of each vic-diol compound, benzaldehyde dimethyl acetal
(α,α-dimethoxytoluene) and p-toluenesulfonic acid monohydrate were
added in molar proportions of 100× and 0.1×, respectively.
Dimethylformamide was added up to a total volume of 3 mL/0.1
mmol of the vic-diol compound. A few beads of molecular sieve were
added, and the solution mixture was allowed to react in an oil bath at
100 °C for 24 h, in reflux with stirring. The reaction mixture was then
partitioned in 20 mL of water/20 mL of dichloromethane, the organic
phase recovered, and analyzed by GC-MS to control the reaction yield.
Purification Steps. The vic-diol BzAc derivatives were isolated and
purified from the corresponding organic phases by MPLC (in the same
system described above) and, when necessary, also by TLC to a
minimum of 98% purity. Di18Diol_Me BzAcs (one from the suberin
acid Di18Diol_Me and the other from the hydrolysis of the suberin
acid Di18Epox_Me) were purified in hexane/ethyl acetate 9:1;
Hyd18Diol_Me BzAcs (one from the suberin acid Hyd18Diol_Me
and the other from the hydrolysis of the suberin acid Hyd18-
Epox_Me) were purified in chloroform/ethyl acetate 9:1; threo-9,10-
dihydroxyoctadecanoic acid_Me BzAcs (one from the model
compound threo-9,10-dihydroxyoctadecanoic acid_Me, a second
from the hydroxylation of cis-oleic acid methyl ester, and a third
from the hydrolysis of the cis-9,10-epoxyoctadecanoic acid) were
purified in chloroform/ethyl acetate 7:3; erythro-9,10-dihydroxyocta-
decanoic acid_Me BzAcs (one from the model compound erythro-
9,10-dihydroxyoctadecanoic acid_Me, a second from the hydroxyla-
Hydroxylation Reaction. cis-Oleic acid methyl ester (1.0 g (3.4
mmol)) was dissolved in 10 mL of dichloromethane, to which a
solution of 6.8 g (16.9 mmol) of CTAP dissolved in 102 mL of
dichloromethane was added dropwise, and the mixture allowed to
react for 100 h at room temperature. The reaction mixture was filtered
on a G3-porosity glass filter, 50 mL of dichloromethane added to the
filtered solution, and the organic solution washed with 3 × 150 mL of
water. After dichloromethane removal in a rotary evaporator, erythro-
9,10-dihydroxyoctadecanoic acid methyl ester was recovered in a
reaction yield of 37%. The same procedure was followed to hydrolyze
trans-elaidic acid methyl ester: 0.9 g (3.0 mmol) was dissolved in 9 mL
of dichloromethane, and 6.1 g (15.2 mmol) of CTAP dissolved in 92
mL of dichloromethane was added. The recovered organic phase
included threo-9,10-dihydroxyoctadecanoic acid methyl ester in a
reaction yield of 90%.
Purification of the C18 9,10-Diols Obtained from the Hydroxyl-
ation of the C18 9-Unsaturated Monoacids. Each of the C18 9,10-
diols obtained was purified by medium pressure liquid chromatog-
raphy (MPLC) as follows: the solid material recovered from the
organic phases was dissolved in chloroform/ethyl acetate 7:3 and
applied to a VersaFlash silica column, 40 × 75 mm, and eluted with the
same solvent at a flow rate of 50 mL/min. From the mixture enriched
in erythro-9,10-dihydroxyoctadecanoic acid methyl ester, a fraction
with 82.0 mg of the latter was recovered in a purity of 99%. From the
mixture enriched in threo-9,10-dihydroxyoctadecanoic acid methyl
ester, 131.5 mg of the latter was recovered in a purity of 99%.
7041
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
tion of trans-elaidic acid methyl ester, and a third from the hydrolysis
of the model compound trans-9,10-epoxyoctadecanoic acid) were
purified in chloroform/ethyl acetate 7:3 (the latter two BzAc
derivatives were lost in the purification steps and were not used for
the NMR analysis); threo-5,6-dodecanediol BzAc and erythro-5,6-
dodecanediol BzAc were purified in hexane/ethyl acetate 9:1.
DSC Analysis. For melting point determination, the suberin acids
Hyd18Epox_Me, Di18Epox_Me, and Di18Diol_Me were analyzed by
differential scanning calorimetry (DSC). The thermograms were
acquired on a Maia DSC 200 F3 (Netzsch, Germany) in a temperature
range from −50 °C to +90 °C, with a heating rate of 20 °C/min. The
suberin acid samples, previously dried in a vacuum oven, were weighed
in aluminum pans (5.0 mg of Hyd18Epox_Me, 5.8 mg of
Di18Epox_Me, and 9.6 mg of Di18Diol_Me), and cooled to −50
°C with liquid nitrogen. DSC was calibrated with an expanded
uncertainty of measurement error <0.5 °C.
FTIR Analysis. FTIR spectra were acquired from the four suberin
acids and from the 9,10-epoxy and the vic-diol model compounds,
either in ATR or transmission mode, depending on its physical state at
room temperature, in an Alpha-P spectrometer (Bruker Optik,
Germany). ATR spectra were obtained from liquid and semisolid
samples in a diamond cell with a pressure clamp; transmission spectra
were obtained from solid samples in KBr pellets after grinding to a fine
powder (1.5 mg of sample with 200 mg of KBr). Thirty-two scans
were acquired per spectrum with a resolution of 4 cm−1 in a
wavenumber range from 4000 cm−1 to 400 cm−1.
determined by DSC, respectively, as 40.7 °C and 57.4 kJ/mol
for Hyd18Epox_Me and 21.7 °C and 46.2 kJ/mol for
Di18Epox_Me. Literature values for the melting point of
Hyd18Epox_Me is 38 °C, and Di18Epox_Me is reported as an
oil at room temperature.14
A first approach to check if these epoxyacids were present in
cork suberin both as cis- and trans-isomers or only in one of the
two possible relative configurations was made by GC-MS
analysis, comparing with model compounds. The latter, C18 cis-
9,10-epoxyoctadecanoic acid and C18 trans-9,10-epoxyoctade-
canoic acid, both as methyl esters, were mixed and baseline
separated in the chromatographic conditions described in
Materials and Methods. Each of the two suberin C18 9,10-
epoxyacids was then analyzed in the same chromatographic
conditions, including solution concentration and injection
volume, but only one peak was recognizable in the chromato-
grams. This gave evidence that the two suberin epoxyacids,
Di18Epox and Hyd18Epox, were present in cork suberin
probably in only one of the forms, either cis or trans.
FTIR and Raman Analyses. The cis or trans configuration of
epoxide rings can be tentatively assessed by vibrational
spectroscopy, namely, FTIR and Raman. The diagnostic
bands used for this purpose derive from the epoxide ring
vibrations: in epoxides substituted in alkyl chains, typically, for
cis-isomers this band is observed in the range of 785−865 cm−1
and for trans-isomers at 860−950 cm−1.23 In the cis and trans
model compounds analyzed here, these bands were con-
spicuous: in the C18 cis-epoxyacid, it was found at 847 cm−1 and
in the C18 trans-epoxyacid at 877 cm−1 (see Supporting
Information, Figure SM1). With respect to the two suberin
epoxyacids, a relatively strong band at 844 cm−1 was observed
in Hyd18Epox_Me, and a medium to weak band at 837 cm−1
was present in Di18Epox_Me (Supporting Information, Figure
SM2). The former can probably be assignable to a cis-epoxide
ring vibration, but the latter, due to its weak intensity, was
ambiguous. However, in both suberin acids (and also in the two
model compounds) there were a number of weak intensity
bands in the window assignable to the ring vibrations of cis- and
trans-epoxides, and therefore, no definite conclusion can be
drawn about their configuration based only on the FTIR
analysis.
Raman Analysis. Raman spectra were obtained from the cork
suberin 9,10-epoxyacids, Di18Epox_Me and Hyd18Epox_Me, and
from the 9,10-diol cork suberin acid, Di18Diol_Me, in a system
comprising a double monochromator Spex 1403 (Horiba, Japan), with
an argon ion laser line at 514.5 nm, model 2016 (Spectra-Physics,
USA), and a R928 photomultiplier detector (Hamamatsu Photonics,
Japan). The spectra were acquired at room temperature in 90°
geometry, with a spectral resolution of 4 cm−1.
NMR Analysis. NMR spectra, including the 1D 1H and 13C,
together with 2D correlation experiences (COSY, HSQC, and
HMBC), were obtained from the suberin acids, model compounds,
synthesized vic-diols, and all BzAc derivatives. The NMR spectra were
acquired on an Avance II+ 600 spectrometer (Bruker Biospin,
Germany), with frequency resonances of 600.13 MHz for protons and
150.96 MHz for carbons, equipped with a cryoprobe and pulse
gradient units, able to produce magnetic field pulsed gradients in the z-
direction of 56.0 G/cm. All samples were solubilized in deuterated
chloroform with 0.03% of TMS, with an approximate concentration of
5 mg/500 μL, and the spectra acquired at a temperature of 300 K. The
TMS signal (0.00 ppm) was used to calibrate the chemical shifts in 1H
spectra, and chloroform (77.00 ppm) was used as reference in 13C
spectra. The NMR spectra were processed in MestReNova, version
8.0.1 (Mestrelab Research, Spain).
Polarimetry Analysis. The specific rotation of the suberin acids,
Hyd18Epox_Me, Di18Epox_Me, Hyd18Diol_Me, and Di18Diol_Me
was measured in a Perkin-Elmer 241 MC Polarimeter, equipped with a
sodium lamp (589 nm) and a 1 mL quartz cell with 1 dm of length.
Sample concentration was between 5 and 10 mg/mL in dichloro-
methane and the measurements acquired at room temperature (ca. 25
°C).
The Raman spectra analysis was also inconclusive: both
Hyd18Epox_Me and Di18Epox_Me showed absorption bands
in the window of the cis- and trans-epoxide ring vibrations;
however, some of these bands could also be assignable to
skeletal vibrations of the alkyl chains, which are stronger in
Raman than in FTIR spectra.
NMR Analysis. The two suberin acids, Di18Epox_Me and
Hyd18Epox_Me, and the two model compounds of known
relative configuration, C18 cis-epoxyacid and C18 trans-
1
epoxyacid, were analyzed by NMR including H, 13C, and 2D
correlation techniques, namely, COSY, HSQC, and HMBC.
The chemical shifts of the protons and carbons relevant for the
discussion of the cis or trans configuration, namely, of the
epoxide methines (C-9 and C-10) and of the adjacent
methylenes (C-8 and C-11), are presented in Table 1. A first
approach for the assignment of the cis or trans configuration
would be the coupling constants between the two epoxide
methine protons. However, in the proton spectra of all analyzed
epoxides these protons showed up as a broad unresolved
multiplet, as would be expected due to the similarity of the alkyl
substituents on the epoxide ring. A second approach would be
the values of the chemical shifts of protons and carbons directly
RESULTS AND DISCUSSION
■
Spectroscopic Analysis and Stereochemistry of the
C18 9,10-Epoxy Suberin Acids. GC-MS and DSC Analyses.
The two C18 9,10-epoxy suberin acids were obtained from cork
as methyl esters, after suberin depolymerization by meth-
anolysis, with purities, as determined by GC-MS, of 95% for
Di18Epox_Me and 96% for Hyd18Epox_Me. The EIMS of
methyl esters and TMS derivatives of suberin acids have been
previously presented and discussed.22 The calculated Kovats
indexes were for Di18Epox_Me 2355 and for Hyd18Epox_Me
2369. Their melting points and fusion enthalpies were
7042
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
Table 1. Chemical Shifts (ppm) of 1H and 13C Methines and
Adjacent Methylenes in C18 9,10-Epoxyacids from Cork
Suberin and Model Compounds
(Table 1). Differences of this order of magnitude were also
found in the equivalent methylene carbons in several other C18
epoxyacids between cis- and trans-isomers.26 This higher
shielding observed in the C-8 and C-11 carbons in cis-isomers
might be attributed to the so-called “gamma-effect”, an
empirical observation in which the chemical shift of a carbon
is affected by a spatially close substituent located three bonds
away.27 In cis-epoxides “the substituent group” is the methylene
group on the opposite side of the epoxide ring, gamma-
positioned to the carbon under consideration; this spatial
proximity and the van der Waals interactions believed to be
behind the “gamma-effect” only exist in cis-isomers (Figure 2).
Theoretical calculations taking into account bond length
compression or stretching, or angular distortions, imparted by
the steric effects of the substituent groups in the chemical shift,
were able to predict to some extent the comparative shielding
of gamma-carbons in similar alkyl-substituted cis-epoxides,
when compared to trans-epoxides.28
epoxide methines (C-9
and C-10)
adjacent methylenes
(C-8 and C-11)
1H
13C
1H
13C
C18 9,10-epoxy suberin acids
cis-Di18Epox_Me
2.89
2.89
57.07
1.49
1.50
27.74
cis-Hyd18Epox_Me
57.16 and
57.19
27.78 and
27.79
C18 9,10-epoxy model compounds
cis-9,10-
2.91
57.23 and
57.28
1.50
1.51
27.77 and
27.81
epoxyoctadecanoic acid
trans-9,10-
2.66
58.90 and
58.95
32.07 and
32.11
epoxyoctadecanoic acid
δcis‑epox−δtrans‑epox
+0.25 −1.67 and
−1.67
−0.01 −4.30 and
−4.30
linked to the epoxide group, due to different shielding in the cis
and trans environments. In fact, a significant difference was
found between the chemical shifts of the methine epoxide
protons, in the cis and trans model compounds analyzed, with a
ΔHδcis‑δtrans of +0.25 ppm (Table 1). This comparative shielding
of the trans-epoxide methine protons was also observed in a
number of other cis/trans isomers of C18 epoxy fatty acids, with
the epoxide group in different chain positions.24,25 This
systematic upfield effect might be imputable to the fact that
the methine protons in trans-isomers are spatially closer to the
methylene protons located on the other side of the epoxide
ring, in comparison to the cis-isomers, where such proximity
does not exist (Figure 2). In this way, having in mind the
The chemical shifts of the epoxide methine carbons (C-9 and
C-10) and of the epoxide vicinal carbons (C-8 and C-11) both
in Hyd18Epox_Me and Di18Epox_Me were very similar to
those of the cis model compound (Table 1), thus reinforcing
the hypothesis that the suberin epoxyacids had a cis
configuration.
Spectroscopic Analysis and Stereochemistry of the
C18 9,10-Diol Suberin Acids. GC-MS and DSC Analyses.
The C18 9,10-diol cork suberin acids were extracted from cork
suberin and purified up to 95%, in the case of Di18Diol_Me,
and 97% in the case of Hyd18Diol_Me, as controlled by GC-
MS. The calculated Kovats indexes were, respectively, 2522 and
2528. The melting point of Di18Diol_Me, measured by DSC,
was 75.0 °C, with a fusion enthalpy of 42.5 kJ/mol. To check if
the vic-diol suberin acids were present in cork both in the
erythro and threo forms or only on one of them, the separation
of the model compounds C18 threo-9,10-dihydroxyoctadecanoic
acid and C18 erythro-9,10-dihydroxyoctadecanoic acid, which
are structurally very close to the C18 9,10-diol suberin acids, was
essayed by GC. A 1:1 mixture of the two latter model
compounds was successfully separated by GC with baseline
resolution, with a difference in the retention times of 12 s (data
not shown). Each of the C18 9,10-diol suberin acids, analyzed
exactly in the same GC conditions, afforded a single peak,
suggesting that only one of the two possible relative
configurations, either threo or erythro, was found in cork
suberin.
NMR Analysis. The C18 9,10-diol suberin acids were
analyzed by NMR, together with two pairs of threo and erythro
model compounds, namely, C18 threo- and erythro-9,10-
dihydroxyoctadecanoic acids and C12 threo- and erythro-5,6-
dodecanediols. The chemical shifts of protons and carbons of
the C-9 and C-10 (C-5 and C-6 in the case of C12 diols)
methines, where the hydroxyl groups are linked, which are
relevant for the configuration discussion, are presented in Table
2. The differences in these chemical shifts between the threo
and erythro model compounds are presented in Table 3.
A systematic difference was observed between the chemical
shifts of the erythro- and threo-methine protons in the vic-diol
model compounds, with a ΔHδerythro‑δthreo of +0.19 ppm (Table
3). To understand this difference, we have to consider the
possible conformations of the vic-diol group due to the rotation
around the σ bond that links the two hydroxyl-substituted
carbons and therefore the different magnetic environments in
each case (Figure 3). A study carried out in 2,3-butanediol, a
Figure 2. NMR chemical shift effects in cis- and trans-epoxides. (A) In
trans-epoxyacids, shielding of the epoxide methine protons (marked)
by the opposite methylenes. (B) In cis-epoxyacids, shielding of the
adjacent methylene carbons (one marked) due to the γ-effect of the
“substituent” group three carbons away (the opposite adjacent
methylene).
structural similarity with the model compounds analyzed, the
chemical shifts of the methine epoxide protons of suberin
epoxyacids could be seen as indicative of a cis configuration:
2.89 ppm both for Di18Epox_Me and Hyd18Epox_Me, a value
very close to 2.91 ppm observed in the cis model compound
and farther apart from the 2.66 ppm of the corresponding trans
isomer (Table 1).
A significant difference between the cis and trans model
compounds was also observed in the 13C NMR spectra, namely,
in the carbons of the methylenes adjacent to the epoxide ring
(C-8 and C-11). In fact, the cis-methylene carbons were
comparatively more shielded with a ΔCδcis‑δtrans of −4.30 ppm
7043
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
1
Table 2. Chemical Shifts (ppm) of H and 13C Methines in vic-Diols of Cork Suberin Acids (C-9 and C-10) and Model
Compounds (C-9 and C-10 or C-5 and C-6), with the vic-Diol as a Free Group and as a Benzylidene Acetal (BzAc) Derivative
f
free vic-diol group
benzylidene acetal (BzAc) derivative
1H
13C
1H
13C
HBzAc
CBzAc
C18 9,10-diol suberin acids
threo-Di18Diol_Me
3.40
3.41
74.46
3.75
3.75
81.52 and 82.80
81.52 and 81.54
82.80 and 82.83
81.52 and 82.80
81.52 and 81.53
82.80 and 82.82
5.86
5.86
102.59
102.60
threo-Hyd18Diol_Me
74.46 and 74.48
a
threo-Di18Diol_Me
3.40
3.40
74.48
3.75
3.75
5.86
5.86
102.59
102.59
b
threo-Hyd18Diol_Me
74.47 and 74.49
C18 9,10-diol model compounds
threo-9,10-dihydroxyoctadecanoic acid
3.38
3.60
3.41
3.60
74.35 and 74.40
74.71 and 74.76
74.48 and 74.54
74.63 and 74.70
3.74
4.11
3.75
81.55 and 81.57
82.82 and 82.85
78.74 and 78.77
79.20 and 79.28
81.53 and 81.56
82.80 and 82.86
5.86
5.77
5.86
102.56
103.05
102.58
erythro-9,10-dihydroxyoctadecanoic acid methyl ester
c
threo-9,10-dihydroxyoctadecanoic acid methyl ester
d
erythro-9,10-dihydroxyoctadecanoic acid methyl ester
e
threo-9,10-dihydroxyoctadecanoic acid
3.75
81.53 and 81.56
82.80 and 82.85
5.86
102.58
C12 5,6-diol model compounds
threo-5,6-dodecanediol
3.41
3.60
74.51 and 74.53
3.76
4.12
81.55 and 81.57
82.84 and 82.85
79.28
5.87
5.77
102.57
103.08
eryhtro-5,6-dodecanediol
74.68 and 74.70
a
b
Obtained from the hydrolysis of the cork suberin acid C18 cis-Di18Epox_Me. Obtained from the hydrolysis of the cork suberin acid C18 cis-
c
d
Hyd18Epox_Me. Obtained from the syn-hydroxylation of the model compound C18 trans-elaidic acid methyl ester. Obtained from the syn-
e
hydroxylation of the model compound C18 cis-oleic acid methyl ester. Obtained from the hydrolysis of the model compound C18 cis-9,10-
f
epoxyoctadecanoic acid. HBzAc and CBzAc are the H and 13C chemical shifts of the acetal methine in the benzylidene group.
1
1
Table 3. Chemical Shift Differences (Δppm) in H and 13C
of vic-Diol Methines (from Suberin Acids and Model
Compounds) with the vic-Diol Group in the Free Form and
a
as the Benzylidene Acetal (BzAc) Derivative
ΔδEf‑Tf
ΔδEBzAc‑TBzAc ΔδTBzAc‑Tf ΔδEBzAc‑Ef
1H
Hyd18Diol_Me (cork
suberin)
+0.34
+0.35
Di18Diol_Me (cork
suberin)
5,6-dodecanediol
+0.19
+0.19
+0.36
+0.36
+0.35
+0.34
+0.52
+0.51
9,10-dihydroxy-
octadecanoic
acid_Me
13C
Figure 3. 1H NMR multiplets of the vic-diol methine protons (H1 and
H2) of erythro (A) and threo (B) model compounds (5,6-
dodecanodiol) with the Newman projection conformations that
allow hydrogen bonding between the hydroxyl groups (only R,S-
erythro and R,R-threo shown).
Hyd18Diol_Me (cork
suberin)
+7.07 and
+8.34
Di18Diol_Me (cork
suberin)
+7.06 and
+8.34
5,6-dodecanediol
+0.17 and −3.57 and
+0.17 −2.28
+7.05 and +4.58 and
+8.32 +4.60
9,10-dihydroxy-
octadecanoic
acid_Me
+0.22 and −3.59 and
+7.07 and +4.48 and
+8.29 +4.05
hydroxyl groups. These same conformational preferences can
also be valid for the vic-diols in the present study, although they
have much longer methylene chains attached to the hydroxyl-
substituted chiral carbons. These conformations that approx-
imate the vicinal hydroxyls (but avoid the eclipsed forms that
would be sterically unfavorable) are shown in Figure 3, namely,
gauche+ and gauche− in erythro, and gauche+ and anti in threo
isomers. If these are the preferred conformations in the vic-diols
we are dealing with here, they can explain the observed
chemical shift differences between the erythro- and threo-
isomers (Figure 3) because methine protons have different
neighborhoods: in the erythro case, only one of the methine
+0.23 −2.79
a
Ef, Erythro with free vic-diol group; Tf, Threo with free vic-diol group;
EBzAc, Erythro with vic-diol group in the form of benzylidene acetal
derivative; TBzAc, Threo with vic-diol group in the form of
benzylidene acetal derivative.
much simpler vic-diol but with identical stereochemistry,
29
1
analyzed the dominant conformers by H NMR. In low
polarity solvents such as CDCl3, both the erythro (meso-2,3-
butanediol) and threo ( -2,3-butanediol) isomers preferred the
conformations that allowed hydrogen bonding between the
7044
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
protons is spatially close to the opposite methylene group (H1
to R2 or H2 to R1, Figure 3A); but in the threo case, both
methine protons are close to their opposite methylene groups
(H1 to R2 and H2 to R1, Figure 3B). The spatially close
opposite methylenes shield the methine protons, an effect
stronger in threo isomers, thus explaining the difference of
+0.19 ppm observed in erythro methine protons in model
compounds (Table 3).
These assumed preferred conformations for vic-diols just
discussed would also explain the comparative shielding
observed in the carbons of the adjacent methylenes in erythro-
versus threo-isomers, a difference of ca. −2 ppm, as measured in
model compounds. A more significant “gamma-effect”, similar
to the one described above in cis-epoxides, exists in erythro-
isomers due to the spatial proximity between R groups (R1 to
R2), observed in both favored conformations; in threo-isomers,
this proximity is only observed in one of the conformations
(gauche+, Figure 3).
Comparing the chemical shifts of the vic-diol methine
protons of Hyd18Diol_Me and Di18Diol_Me suberin acids
with the ones of model compounds, we noticed that they are
very similar to those of the threo-isomers (Table 2). However,
because there is no direct proof of the preferred conformations
around the vic-diol group in suberin acids and model
compounds, some doubt can remain on the use of these
chemical shift comparisons as proof of the relative config-
uration. To circumvent this situation, we reassessed the NMR
analysis after converting the vic-diol groups into a cyclic
structure, thus strongly diminishing the conformational
variability as discussed below.
NMR Analysis of the vic-Diol Benzylidene Acetal
(BzAc) Derivatives. All vic-diols, including the suberin acids,
Hyd18Diol and Di18Diol, and the two pairs of erythro and threo
model compounds, C18 9,10-diol monoacids and C12 5,6-
dodecanediols, were derivatized as benzylidene acetals (BzAc),
producing a “1,3-dioxolane” five-membered ring (Figure 4).
Also, the suberin C18 9,10-epoxyacids, Hyd18Epox and
Di18Epox, and the C18 cis-9,10-epoxyacid model compound
were hydrolyzed into C18 9,10-diols and BzAc derivatized in the
same way. The stereospecificity of the epoxide hydrolysis was
used to confirm the cis or trans relative configuration of the
suberin epoxyacids, as discussed below. Moreover, as a control,
erythro- and threo-C18 9,10-diol monoacids, identical to the
above model compounds, were obtained by the syn-
hydroxylation of cis-oleic acid methyl ester and trans-elaidic
acid methyl ester, respectively, and their BzAc derivatives also
prepared. The NMR results of these BzAc derivatives with
respect to the proton and carbon chemical shifts of the C-9 and
C-10 (or C-5 and C-6) methines of the derivatized diols, as well
as of the BzAc acetal methine (see Figure 4), are presented in
Table 2.
Proton Spectra. Differences between the chemical shifts of
the methine protons of erythro- and threo-vic-diol BzAc
derivatives were conspicuous in model compounds, as shown
in Table 3. As a rule, these methine protons in the erythro-BzAc
derivatives had a chemical shift of ca. 4.12 ppm and in the threo-
BzAc derivatives of ca. 3.75 ppm (Table 2), a difference of
ΔHδerythroBzAc‑δthreoBzAc of ca. +0.36 ppm (Table 3). Because this
difference is significant and the chemical shift values are
constant, this seems to be, by itself, a strong diagnostic feature
to assign an erythro or threo configuration in these vic-diol
structures.
The higher shielding in the threo isomers can, as in the case
of the epoxide methine protons, again be explained by the
“substituent” effect of the opposite methylene group, which is
spatially closer to the methine proton under consideration
(Figure 4). Because of the restricted conformational mobility of
the dioxolane ring, this effect would be stronger than the one
observed in the underivatized vic-diols, being in fact almost
double (+0.19 ppm in the underivatized vic-diols and +0.36
ppm in the BzAc derivatives, see Table 3). The effect of the
phenyl group of the BzAc derivative in the methine chemical
shifts of the derivatized vic-diols is harder to evaluate. Besides
some conformational variability of the dioxolane ring, and the
free rotation of the phenyl group around the acetal carbon, the
BzAc derivates can have actually two different configurations.
When the BzAc derivative is formed from the vic-diol and the
precursor benzaldehyde dimethyl acetal, the acetal carbon of
the latter becomes asymmetric, originating two new possible
stereoisomers, the (R)-BzAc and the (S)-BzAc, thus locating
the phenyl group in two possible opposite sides (Figure 4).
Because in theory both reaction positions are identically
probable, both enantiomers should exist in similar proportions,
making the interpretation of the phenyl effect in the chemical
shifts more complicated.
Another difference between the BzAc derivatives of erythro-
and threo-vic-diols was observed in the BzAc structure itself,
namely, in the acetal methine proton. In fact, in the model
compounds analyzed, this proton had a chemical shift of 5.77
ppm in the erythro-isomers and 5.86 ppm in the corresponding
threo-isomers (Table 2). This systematic difference of −0.09
ppm, showing that this proton in the erythro-BzAc derivatives is
more shielded and the consistency of the chemical shift values
in the model compounds analyzed, suggests that the chemical
shift of the acetal proton in the BzAc derivatives of alkyl
substituted vic-diols can also be a reliable way to assign the
erythro or threo configuration.
13C Carbon Spectra. The chemical shifts of the 13C carbons
in BzAc derivatives and their differences between the erythro-
and threo-isomers, as observed in model compounds, also
showed to be diagnostic of the relative configuration of vic-diols
(Tables 2 and 3). To compare the 13C chemical shifts, we have
to consider that in the free vic-diol form there are different
signals for the C-9 and C-10 carbons (or C-5 and C-6 in case of
dodecanediols) because neither the suberin acids nor the model
Figure 4. Benzylidene acetal derivatives of vic-diols shown in the
envelope conformation. R1 and R2 are as described in Figure 1 (plus a
CH2 for each (CH2)n group). [The quirality of the acetal carbon is not
represented, depending on the R1 and R2 sustituents.]
7045
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
compounds are completely symmetric in relation to the vic-diol
group. The exception is the Di18Diol_Me from suberin, which
due to its complete symmetry only shows one signal for both
C-9 and C-10 carbons. Comparing the 13C chemical shifts of
the erythro and threo forms of the underivatized model vic-diols,
ΔCδerythro‑δthreo, a small average difference of ca. +0.2 ppm was
observed (Table 3). This difference is probably too small for a
safe assignment of the vic-diol configuration. However, much
bigger differences were found when this comparison was made
in the respective BzAc derivatives, either −2.8 or −3.6 ppm,
depending on the carbon chemical shift considered (Table 3).
After the BzAc derivatization, the vic-diol methine carbons
were significantly deshielded compared to its previous free
hydroxyl underivatized forms, in an order of magnitude of ca.
+8 ppm for the threo model compounds and ca. +4 ppm for the
corresponding erythro (Table 3). This showed that the
deshielding effect of the BzAc group in these carbons was
stronger in the threo isomers. Nevertheless, the most diagnostic
feature to assign the erythro or threo configuration for these vic-
diols using 13C NMR was simply the values of their chemical
shifts: in the vic-diol BzAcs of erythro model compounds, the
vic-diol carbons were in the window of 78.7 to 79.3 ppm, and in
the threo equivalents, they were from 81.5 to 82.9 ppm (Table
2). Both ranges are short and unambiguously apart, making the
configuration assignment safe. In both C18 9,10-diol suberin
acids, the chemical shifts of their vic-diol methine carbons as
BzAc derivatives were in the threo range; identically, both C18
9,10-epoxy suberin acids hydrolyzed into the corresponding C18
9,10-diols had values corresponding to a threo relative
configuration.
here for the C18 cork suberin acids as cis-9,10-epoxy and threo-
9,10-diol.
Although practically nothing is known about how the suberin
acids are three-dimensionally organized within the suberin
macromolecule, at least a partial ordered arrangement is
possible. As seen at the ultrastructural level, suberized cell walls
show a composite lamellar structure, with alternate dark and
translucent contrast, the latter with a regular thickness of about
3 nm.30 An ordered packing of the C18 suberin acids was
proposed to be the structural basis of these translucent
lamellae.3 Studies made by 13C solid state NMR have shown
that a part of the methylene chain in suberins was motionally
restricted;31,32 also a WISE NMR study in potato wound
periderm found the most diagnostic suberin carbons in the
vicinity of low mobility protons.33 In both cases, the observed
rigidity can be justified by an ordered “crystalline” phase.
Some fats can be regarded as a comparable situation to that
of suberin, in the sense of being based on glycerol esterified to
C18 midchain modified fatty acids. This is the case of triolein,
the triacylglycerol of C18 cis-9-oleic acid, which in the solid
phase packs the alkyl chains in an ordered manner, kinked at
the double bond,34 as favored by its cis configuration. In
suberin, both the C18 9,10-epoxide and the C18 9-unsaturated
acids6 have a cis configuration, which will identically favor a
similar bent molecular arrangement of their alkyl chains.
Finally, if we bend in the same way the C18 threo-vic-diol
suberin acids, they project their two hydroxyl groups to
opposite directions, minimizing the steric hindrance between
them. We should also bear in mind that these C18 midchain
modified suberin acids must play a major role in cork suberin
structure since they account for about 50% of all its long-chain
monomers.7 If all of these C18 suberin acids, with the midchain
unsaturation, epoxide, and vic-diol group assume this kinked
conformation, they could be conveniently packed in an ordered
parallel way (see the Table of Contents Graphic in the Web
edition). This packing would favor strong intramolecular
bonding, namely, hydrogen bonding between epoxides and
vic-diol groups. Ultimately, this molecular arrangement would
allow an ordered macromolecular structure for suberin and
justify the organized lamellar ultrastructure observed in cork
cell walls.
Stereochemistry of the C18 Suberin Acids and the
Macromolecular Structure of Suberin. The results
1
discussed above showed that the H and 13C chemical shifts
of the methine protons and carbons of vic-diols substituted in
alkyl chains derivatized as benzilidene acetals can unambigu-
ously be used to assign their erythro or threo configuration. The
comparison of these chemical shifts in the model compounds
with the ones of cork suberin acids, Hyd18Diol_Me and
Di18Diol_Me (Table 2), proved that both exist in only one
relative configuration, which is threo. In the same way,
1
comparing the H and 13C epoxide methine and methylene
NMR chemical shifts of model compounds, with the ones of
C18 9,10-epoxy suberin acids, Hyd18Epox_Me and Di18-
Epox_Me, the two latter compounds were shown to be
concordant with a cis configuration. To confirm this
assumption, both cork suberin and model epoxyacids were
hydrolyzed by a stereospecific anti-hydroxylation, where cis-
epoxides gave rise to threo-vic-diols, and trans-epoxides
originated erythro-vic-diols. The chemical shifts of the hydro-
lyzed epoxyacids, either as free vic-diols or in the form of BzAc
derivatives, showed them to be threo, definitely proving that the
suberin epoxyacids had a cis configuration. The optical activity
of these cis-epoxide and threo-vic-diol suberin acids measured by
polarimetry was zero, showing that they probably exist as
racemic mixtures of their corresponding pairs of enantiomers.
The stereochemistry of the C18 9,10-epoxy and C18 9,10-diol
suberin acids in cork is a priori conditioned by the
stereoselectivity of their biosynthetic pathways. Biosynthesis
studies had shown that cis-oleic acid could be the precursor of
C18 9,10-epoxy and C18 9,10-diol suberin acids,2 following the
stereochemically favored routes of oxidation of the cis-double
bond to cis-epoxide and hydroxylation of the latter to threo-diol.
This is in agreement with the configuration that was proved
ASSOCIATED CONTENT
* Supporting Information
FTIR (ATR) spectra of model C18 9,10-epoxyacids and C18
cork suberin 9,10-epoxyacids. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: sarasantos@isa.utl.pt.
Funding
This work was supported by the Portuguese Technology and
Science Foundation (FCT) Project PTDC/QUI/70589/2006
and Grant SFRH/BD/31033/2006 and is part of the activities
of Forest Research Centre (Centro de Estudos Florestais).
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We are grateful for the help of Eurico Cabrita, Angelo
Figueiredo, and Maria Manuel Marques (Universidade Nova de
■
̂
7046
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047

Journal of Agricultural and Food Chemistry
Article
(24) Gunstone, F. D.; Jacobsberg, F. R. Fatty-acids. 35: Preparation
and properties of complete series of methyl epoxyoctadecanoates.
Chem. Phys. Lipids 1972, 9, 26−34.
(25) Gunstone, F. D.; Schuler, H. R. Fatty-acids. 46. PMR spectra of
several epoxyoctadecenoic, epoxyoctadecynoic, and diepoxyoctadeca-
noic esters. Chem. Phys. Lipids 1975, 15, 189−197.
́
Lisboa), Luıs Santos (Instituto Superior Tec
́
and Jose Rodrigues (Instituto de Investigaca̧ o Cientıfica
̃
́
nico, TULisbon),
́
Tropical). We acknowledge the Portuguese National NMR
network (RNRMN), funded by FCT, for the NMR analysis.
(26) Bascetta, E.; Gunstone, F. D. C-13 Chemical shifts of long-chain
epoxides, alcohols and hydroperoxides. Chem. Phys. Lipids 1985, 36,
253−261.
REFERENCES
■
(1) Kolattukudy, P. E. Bio-polyester membranes of plants − cutin
and suberin. Science 1980, 208, 990−1000.
(2) Kolattukudy, P. Suberin from Plants. In Biopolymers: Polyesters I;
(27) Gunther, H. NMR Spectroscopy; John Wiley & Sons: New York,
̈
2001.
Doi, Y., Steinbuchel, A., Eds.; Wiley-VCH: Weinheim, Germany, 2002;
̈
(28) Kleinpeter, E.; Seidl, P. R. The gamma- and the delta-effects in
(13)C NMR spectroscopy in terms of nuclear chemical shielding
(NCS) analysis. J. Phys. Org. Chem. 2004, 17, 680−685.
pp 41−73.
(3) Graca̧ , J.; Santos, S. Suberin: a biopolyester of plants’ skin.
1
Macromol. Biosci. 2007, 7, 128−135.
(29) Gallwey, F. B.; Hawkes, J. E.; Haycock, P.; Lewis, D. H NMR
(4) Holloway, P. J. Some variations in the composition of suberin
from the cork layers of higher-plants. Phytochemistry 1983, 22, 495−
502.
spectra and conformations of propane-1,2-diol, meso- and racemic
butane-2,3-diols, and some alditols in non-aqueous media. J. Chem.
Soc., Perkin Trans. 2 1990, 1979−1985.
(5) Sanz, V.; Seoane, E. Synthesis of ambrettolide from phloionolic
acid. J. Chem. Soc., Perkin Trans. 1 1982, 8, 1837−1839.
(6) Santos, S.; Graca̧ , J. Stereochemistry of C18 monounsaturated
cork suberin acids determined by spectroscopic techniques including
1H NMR multiplet analysis of olefinic protons. Phytochem. Anal. 2013,
in press.
(30) Sitte, P. Der feinbau verkorten zellwande. Mikroskopie 1955, 10,
̈
178−200.
(31) Stark, R. E.; Garbow, J. R. Nuclear magnetic resonance
relaxation studies of plant polyester dynamics 2. Suberized potato cell
wall. Macromolecules 1992, 25, 149−154.
(32) Gil, A. M.; Lopes, M.; Rocha, J.; Neto, C. P. A C-13 solid state
nuclear magnetic resonance spectroscopic study of cork cell wall
structure: The effect of suberin removal. Int. J. Biol. Macromol. 1997,
20, 293−305.
(7) Graca
̧ , J.; Pereira, H. Methanolysis of bark suberins: Analysis of
glycerol and acid monomers. Phytochem. Anal. 2000, 11, 45−51.
́
́
(8) Seoane, E.; Ribas, I. Sobre la quımica del corcho. VII. Los acidos
(33) Yan, B.; Stark, R. E.; Wise, A. NMR approach to heterogeneous
biopolymer mixtures: Dynamics and domains in wounded potato
tissues. Macromolecules 1998, 31, 2600−2605.
(34) Larsson, K.; Quinn, P.; Sato, K.; Tiberg, F. Lipids: Structure,
Physical Properties and Functionality; The Oily Press: Bridgwater,
England, 2006.
floionico y floionolico. An. R. Soc. Esp. Fis. Quim., Ser. B 1951, 47, 61−
66.
(9) Seoane, E.; Ribas, I.; Fandino, G. Constitution of phloionolic and
isophloionolic acids. Chem. Ind. 1957, 16, 490−491.
(10) Seoane, E.; Ribas, I.; Fandino, G. Quımica del corcho. XIV:
́
́ ́ ́ ́
Constitucion y estereoquimica de los acidos floionolico e isofloiono-
lico. An. R. Soc. Esp. Fis. Quim., Ser. B 1959, 55, 839−846.
(11) Seoane, E.; Arno, M. Total synthesis and stereochemistry of
phloionolic acids. Anal. Quim. 1977, 73, 1336−1339.
́
(12) Dupont, G.; Dulou, R.; Cohen, J. Structure sterique de l’acide
phloionique. Bull. Soc. Chim. Fr. 1956, 5, 819−824.
(13) Conde, E.; Garcia-Vallejo, M. C.; Cadahia, E. Variability of
suberin composition of reproduction cork from Quercus suber
throughout industrial processing. Holzforschung 1999, 53, 56−62.
(14) Seoane, E.; Serra, M. C.; Agullo, C. Two new epoxy-acids from
the cork of Quercus suber. Chem. Ind. 1977, 15, 662−663.
(15) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
John Wiley & Sons: New York, 1994.
(16) Seco, J. M.; Quinoa, E.; Riguera, R. The assignment of absolute
configuration by NMR. Chem. Rev. 2004, 104, 17−117.
(17) Wenzel, T. J. Discrimination of Chiral Compounds Using NMR
Spectroscopy; John Wiley & Sons: Hoboken, NJ, 2007.
(18) Freire, F.; Seco, J. M.; Quinoa, E.; Riguera, R. Determining the
1
absolute stereochemistry of secondary/secondary diols by H NMR:
Basis and applications. J. Org. Chem. 2005, 70, 3778−3790.
(19) Kocienski, P. J. Protecting Groups; Thieme: New York, 1994.
(20) Bhushan, V.; Rathore, R.; Chandrasekaran, S. A simple and mild
method for the cis-hydroxylation of alkenes with cetyltrimethylammo-
nium permanganate. Synthesis 1984, 5, 431−433.
(21) McElhanon, J. R.; Wu, M. J.; Escobar, M.; Chaudhry, U.; Hu, C.
L.; McGrath, D. V. Asymmetric synthesis of a series of chiral AB(2)
monomers for dendrimer construction. J. Org. Chem. 1997, 62, 908−
915.
(22) Eglinton, G.; Hunneman, D. H.; McCormic., A. Gas
chromatographic-mass spectrometric studies of long chain hydroxy
acids 3. Mass spectra of methyl esters trimethylsilyl ethers of aliphatic
hydroxy acids. A facile method of double bond location. Org. Mass
Spectrom. 1968, 1, 593−611.
(23) Socrates, G. Infrared and Raman Characteristic Group
Frequencies: Tables and Charts, 3rd ed.; John Wiley & Sons:
Chichester, England, 2001.
7047
dx.doi.org/10.1021/jf400577k | J. Agric. Food Chem. 2013, 61, 7038−7047