Secoisolariciresinol dehydrogenase: Mode of catalysis and stereospecificity of hydride transfer in Podophyllum peltatum

Article abstract of DOI:10.1039/b516563f

Secoisolariciresinol dehydrogenase (SDH) catalyzes the NAD⁺ dependent enantiospecific conversion of secoisolariciresinol into matairesinol. In Podophyllum species, (-)-matairesinol is metabolized into the antiviral compound, podophyllotoxin, wh

Full text of DOI:10.1039/b516563f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Secoisolariciresinol dehydrogenase: mode of catalysis and stereospecificity
of hydride transfer in Podophyllum peltatum
a
c
a
a
b
Syed G. A. Moinuddin, Buhyun Youn, Diana L. Bedgar, Michael A. Costa, Gregory L. Helms,
c
a
a
ChulHee Kang, Laurence B. Davin and Norman G. Lewis*
Received 23rd November 2005, Accepted 4th January 2006
First published as an Advance Article on the web 30th January 2006
DOI: 10.1039/b516563f
+
Secoisolariciresinol dehydrogenase (SDH) catalyzes the NAD dependent enantiospecific conversion of
secoisolariciresinol into matairesinol. In Podophyllum species, (−)-matairesinol is metabolized into the
antiviral compound, podophyllotoxin, which can be semi-synthetically converted into the anticancer
ꢀ
R
agents, etoposide, teniposide and Etopophos . Matairesinol is also a precursor of the cancer-
preventative “mammalian” lignan, enterolactone, formed in the gut following ingestion of, for example,
various high fiber dietary foods, as well as being an intermediate to numerous defense compounds in
vascular plants. This study investigated the mode of enantiospecific Podophyllum SDH catalysis, the
order of binding, and the stereospecificity of hydride abstraction/transfer from secoisolariciresinol to
+
153
167
171
NAD . SDH contains a highly conserved catalytic triad (Ser , Tyr and Lys ), whose activity was
abolished with site-directed mutagenesis of Tyr167Ala and Lys171Ala, whereas mutagenesis of
Ser153Ala only resulted in a much reduced catalytic activity. Isothermal titration calorimetry
+
measurements indicated that NAD binds first followed by the substrate, (−)-secoisolariciresinol.
Additionally, for hydride transfer, the incoming hydride abstracted from the substrate takes up the
pro-S position in the NADH formed. Taken together, a catalytic mechanism for the overall
enantiospecific conversion of (−)-secoisolariciresinol into (−)-matairesinol is proposed.
2
,27,28
Introduction
to protect against onset of breast, prostate and colon cancers;
enterolactone (9) is formed in the gut by action of intestinal flora
following ingestion of, for example, high fiber dietary foodstuffs.
Matairesinol (6) can also serve as an intermediate to various
heartwood-protecting substances, such as plicatic acid (8) and its
The lignans represent an abundant class of vascular plant natural
29
1
,2
products with important roles in human health protection,
3
,4
5–7
pharmacology and plant defense. Elucidation of the bio-
chemical pathways leading to their synthesis is resulting in
14,30–32
congeners, in western red cedar (Thuja plicata).
8–10
11–26
systematic disentanglement of the lignin
ways which can overlap in the same tissue types, e.g. during
heartwood formation. Of the various classes of lignans, entry
into the pathway leading to many of the 8–8 -linked metabolites,
and lignan
path-
In our earlier studies, secoisolariciresinol dehydrogenase
SDH†) was isolated from Forsythia intermedia, the encoding gene
cloned (as well as one from Podophyllum peltatum), with the fully
functional recombinant SDH so obtained characterized in terms
(
ꢀ
such as matairesinol (6), podophyllotoxin (7) and plicatic acid
18
of its basic kinetic parameters. Additionally, we solved crystal
(
8), occurs via dirigent-mediated stereoselective coupling of E-
coniferyl alcohol (1) to afford the furanofuran, (+)-pinoresinol
2a) (Fig. 1). The latter can then undergo sequential
structures of the apo-form and the corresponding binary/ternary
25
˚
complexes at 1.6 and 2.8/2.0 A resolution, respectively. P.
peltatum SDH exists as a homotetramer in both solution and
13,15,17,19,22,24
(
NADPH-dependent reduction to afford (−)-secoisolariciresinol
in the crystal lattice (D symmetry), but whose ternary crys-
2
(
4a) via catalysis by an enantiospecific pinoresinol-lariciresinol
+
tal complex (with NAD ) could only be obtained with the
product, (−)-matairesinol (6a), rather than the substrate, (−)-
secoisolariciresinol (4a). Based on homology comparisons with
other short-chain dehydrogenases, it was provisionally concluded
that SDH contained a conserved catalytic triad (Ser , Tyr and
11,12,14,20
reductase.
Of interest in this study is the next enantiospecific step
converting (−)-secoisolariciresinol (4a) into (−)-matairesinol (6a)
in the biochemical pathway to various lignans, such as to the
antiviral/anticancer agent (−)-podophyllotoxin (7) in Podophyl-
25
153
167
171 33
Lys ). In this study, we propose a mechanistic basis for this
17,18
lum species.
Matairesinol (6) is also a precursor of the so-
enantiospecific catalysis using both site-directed mutagenesis and
isothermal calorimetery titration (ITC), with NMR spectroscopic
analysis also being employed for determination of the stereospeci-
called “mammalian” lignan, enterolactone (9), which is believed
a
Institute of Biological Chemistry, Washington State University, Pullman,
+
ficity of hydride transfer to NAD .
WA, 99164-6340, USA. E-mail: lewisn@wsu.edu; Fax: +1 509 335 8206;
Tel: +1 509 335 2682
b
NMR Center, School of Molecular Biosciences, Washington State Univer-
sity, Pullman, WA, 99164-4660, USA
† Abbreviations: Pp, Podophyllum peltatum; PLR, pinoresinol–lariciresinol
reductase; SDH, secoisolariciresinol dehydrogenase; HMQC, heteronu-
clear multiple-quantum correlation.
c
School of Molecular Biosciences, Washington State University, Pullman,
WA, 99164-4660, USA
8
08 | Org. Biomol. Chem., 2006, 4, 808–816
This journal is © The Royal Society of Chemistry 2006

View Article Online
Fig. 1 Biochemical pathway to (−)-matairesinol (6a) in F. intermedia and P. peltatum and proposed conversions to (−)-podophyllotoxin (7) in
P. peltatum, plicatic acid (8) in T. plicata, and to enterolactone (9) by action of intestinal flora. Opposite enantiomers (structures not shown) are
depicted in the text as 2b to 6b.
Results and discussion
TOP10 cells for protein expression and selected on carbenicillin,
with the resulting clones designated as S153A, Y167A and K171A,
respectively.
Site-directed mutagenesis of SDH catalytic triad
The mutagenized proteins (S153A, Y167A and K171A), as well
as the wild type SDH_Pp7, were then first individually purified
(circa 97% purity by silver staining) by anion exchange chromatog-
raphy and assayed for their capacity to convert secoisolariciresinol
(4) into either the intermediary lactol (5) or matairesinol (6).
SDH_Pp7 and the mutants, S153A, Y167A and K171A, were
thus individually incubated with (± )-secoisolariciresinols (4a/b)
To determine the effect on the catalytic activity of SDH, site-
1
53
167
directed mutagenesis of the proposed catalytic triad, Ser , Tyr
1
71
and Lys of SDH_Pp7, was first carried out to individually
mutate each of these residues into the hydrophobic non-reactive
amino acid Ala. Mutations at each position [i.e. 153 (S → A), 167
(
Y → A) and 171 (K → A)] (Table 1) were subsequently confirmed
+
by sequencing each clone both in the forward and reverse
directions (data not shown), as well as to ensure that no other
point mutations were introduced during the PCR reactions. The
three constructs were next individually transformed into E. coli
in the presence of NAD and the products separated by HPLC
as described in the Experimental section. As indicated in Table 2,
enzymatic activity was abolished when SDH was mutated at either
1
67
171
position Tyr or Lys . The S153A mutant, by contrast, displayed
153
167
171
Table 1 Primers used for site-directed mutagenesis of SDH amino acid residues Ser , Tyr and Lys to Ala
ꢀ
ꢀ
Mutation
5 Primer
3 Primer
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
S153 → A153
Y167 → A167
K171 → A171
5 GTATTCACTGCAGCTATTTCTTCCTTCACAGC 3
5 GCTGTGAAGGAAGAAATAGCTGCAGTGAATAC 3
ꢀ
ꢀ
5 GGTGTGTCGCATGTTGCCACCGCAACCAAG 3
5 CTTGGTTGCGGTGGCAACATGCGACACACC 3
ꢀ
ꢀ
ꢀ
ꢀ
5 GTTTACACCGCAACCGCGCATGCTGTCCTTGG 3
5 CCAAGGACAGCATGCGCGGTTGCGGTGTAAAC 3
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 808–816 | 809

View Article Online
a
Table 2 Conversion of (−)-secoisolariciresinol (4a) by wild type secoisolariciresinol dehydrogenase and three mutants (S153A, Y167A and K171A)
(−)-Lactol (5a) pmol min− lg prot
1
−1
(−)-Matairesinol (6a) pmol min−1 lg prot
−1
Protein
b
Wild type
S153A
Y167A
K171A
nd
3.5 ± 0.02
0.2 ± 0.01
nd
nd
nd
nd
nd
a
b
the (+)-antipode (4b) did not serve as substrate, nd, not detected.
very modest activity as evidenced by a relatively poor conversion
of (−)-secoisolariciresinol (4a) into (−)-lactol (5a); however, no
further conversion of 5a into (−)-matairesinol (6a) was observed.
Moreover, there was no further restoration of activity when
the mutated proteins were purified to apparent homogeneity
(
data not shown). These data were, therefore, consistent with
the proposed involvement of the catalytic triad for conversion of
−)-secoisolariciresinol (4a) into (−)-matairesinol (6a).
The mode of SDH-catalyzed hydride transfer to NAD from
either substrate (−)-secoisolariciresinol (4a) or the intermediary
−)-lactol (5a) was next investigated, and whether this involved
stereospecific hydride transfer to either the pro-R or pro-S
hydrogens) at C-4 of the NADH so formed.
(
+
(
(
2
2
Synthesis of [4R– H] and [4S– H]NADH
To experimentally differentiate between the above possibilities,
2
2
34
[
4R– H] and [4S– H]NADH were first enzymatically synthesized.
2
The [4R– H]-nicotinamide adenine dinucleotide (NADD) was
prepared by incubating NAD in the presence of ethanol-d
+
6
, horse
liver alcohol dehydrogenase and yeast aldehyde dehydrogenase,
1
Fig. 2 H NMR spectra of NADH with inset showing spectroscopic
regions for 4S and 4R protons at C-4 of the dihydropyridine ring. (A)
2
+
whereas [4S– H]NADH was prepared by incubating NAD with D-
glucose-1-d and Leuconostoc mesenteroides glucose-6-phosphate
2
Natural abundance NADH (Sigma, 0.47% EtOH). (B) [4S– H]NADH.
34
2
2
dehydrogenase (see Experimental section). The reduced nucleo-
tides so formed were then individually purified via a combination
of DEAE cellulose and Bio-Gel P-2 column chromatographic
steps, with fractions of interest containing NADD (A260 : A340
ratio <2.3) individually pooled and lyophilized.
(C) [4R– H]NADH. (D) Enzymatically synthesized [4R– H]NADH with
SDH_Pp7 following incubation with (± )-secoisolariciresinols (4a/b) and
2
+
[
4- H]NAD . X = Solvent (EtOH).
39
∼18–20 ppb due to the geminal deuterium isotope effect. The
The stereochemistry and isotopic purity of the two distinct
data so obtained were thus in good agreement with that of Arnold
2
36
37
38
forms of deuterated NADD [4R- and 4S– H] were individually
et al., Velonia et al. and Mostad and Glasfeld.
1
established by analysis of the H and heteronuclear multiple-
In the unlabelled NADH sample, the HMQC experiment shows
two sets of cross-peaks centered at 24.40 ppm in F1 which belong to
the diastereotopic protons that reflect the large geminal coupling
between the 4R and 4S protons. The 4S set is centered at 2.65 ppm
in F2, with the doublet signals appearing at 2.62 and 2.67 ppm
while the 4R set is centered at 2.77 ppm in F2 with the doublet
components appearing at 2.75 and 2.80 ppm in the 2D contour
plot (see Fig. 3A). By contrast, HMQC analysis of [4S– H]NADH
revealed loss of the pro-R proton cross-peak at d 2.62, F2 ( H)
(Fig. 3B) thereby establishing the deuterium atom to be at the 4S
position. In an analogous manner, analysis of the [4R– H]NADH
quantum correlation (HMQC) spectra, together with comparison
35–38
1
to undeuterated (natural abundance) NADH.
spectrum of unlabelled NADH (Sigma, 98% pure and containing
.47% EtOH) in D O (Fig. 2A and inset), the two C-4 protons
In the H NMR
0
2
in the dihydropyridine ring were clearly observed with resonances
for the 4S-proton centered at d 2.65 ppm (m, J4S − 4 18.3, J4S − 5
2
3.5, J4S − 6 0.8, J4S − 2 1.0 Hz) and for the 4R-proton centered at
1
d 2.77 ppm (m, J4R − 4 18.3, J4R − 5 2.8, J4R − 6 2.2, J4R − 2 0.8 Hz).
By contrast, examination of the enzymatically generated forms of
NADD established that both were ∼91% deuterated at the C-4
2
1
position as determined by integration of the 1D H NMR spec-
(Fig. 3C) established absence of the pro-S cross-peak at d 2.75, F2
( H) thereby placing the deuterium atom at 4R.
2
1
trum. Furthermore, the 4R proton of [4S– H]NADH (Fig. 2B and
inset) gave a broad signal at d 2.75 ppm (J4R − 5 2.7 Hz), indicating
loss of the 18.3 Hz geminal coupling thereby establishing that the
deuterium atom resides at the 4S position, whereas the 4S proton
2
+
Synthesis of [4- H]NAD
2
2
of [4R– H]NADH (Fig. 3C and inset) appeared as a broad doublet
With the two distinct forms of NADD (4R- and 4S– H) unam-
biguously distinguished, the stereospecificity of hydride transfer
from the substrates (−)-secoisolariciresinol (4a)/(−)-lactol (5a) to
(
J
4S − 5 3.5 Hz) at d 2.63 ppm, again displaying loss of the 18.3 Hz
geminal coupling. The resonances also showed upfield shifts of
8
10 | Org. Biomol. Chem., 2006, 4, 808–816
This journal is © The Royal Society of Chemistry 2006

View Article Online
8
4
.0, J4H–6H 1.2, J4H–2H 1.2 Hz) for the pyridinium proton at C-
1
(Fig. 4A). By contrast, analysis of the H NMR spectrum of
2
+
the [4- H]NAD (Fig. 4B) revealed that the resonance at d 8.81
+
was ∼88% diminished, relative to unlabelled NAD (based on
1
H NMR integration) in accordance with C-4 being specifically
deuterated. This was further confirmed by the change in the
multiplicity at d 8.17 of the C-5 proton, i.e. from a dd (J5H–4H
8.0, J5H–6H 6.3 Hz, Fig. 4A) to a doublet (J5H–6H 6.3 Hz, Fig. 4B).
Generation of NADD during SDH catalysis
± )-Secoisolariciresinols (4a/b) were thus next incubated with a
(
partially purified SDH preparation (see Experimental section)
2
+
in the presence of [4- H]NAD . LC–APCI–MS analysis of the
resulting enzymatically generated matairesinol (6) (retention time
+
1
6.02 min) then gave a molecular ion [M ] at m/z 358.16
confirming its identity. Its optical purity was also established using
18
the procedure of Xia et al., thus confirming that only the (−)-
antipode (6a) was formed (data not shown).
The aqueous phase containing the SDH generated NADH was
Fig. 3 Expanded regions of the HMQC spectra of NADH (C-4 hydrogens
2
and C-4 carbon). (A) Natural abundance NADH. (B) [4S– H]NADH. (C)
2
2
2
next purified as previously described above for [4R– H] and [4S–
[
4R– H]NADH. (D) Enzymatically synthesized [4R– H]NADH following
2
1
2
+
H]NADH. The H NMR and HMQC analyses of the resulting
incubation with (± )-secoisolariciresinols (4a/b) and [4- H]NAD with
NADH generated during SDH catalysis revealed the presence of a
broad doublet at d 2.63 ppm (J4S − 5 3.0 Hz) again showing loss of
the 18.3 Hz geminal coupling, and absence of the pro-S crosspeak
at d 2.77 (Figs. 2D and 3D), i.e., establishing that the deuterium in
the enzymatically generated NADD was at the 4R position. Thus,
the hydride abstracted from either (−)-secoisolariciresinol (4a) or
SDH_Pp7.
+
2
+
NAD during SDH catalysis was investigated using [4- H]NAD
2
as cofactor. The latter was obtained by oxidizing [4R– H]NADH
with rabbit muscle a-glycerophosphate dehydrogenase in the
presence of dihydroxyacetone phosphate, with the enzymatically
formed [4- H]NAD purified further by gel permeation chro-
matography on Bio-Gel P-2 (see Experimental section).
Comparison of the H NMR-spectra of the resulting [4-
40
2
+
2
+
the (−)-lactol (5a) to [4- H]NAD occurred in such a manner
whereby hydride addition is at the B-face of the nicotinamide ring
1
+
2
of NAD giving [4R– H]NADD, and thus establishing SDH to be
2
+
+
H]NAD (Fig. 4B) with that of unlabelled NAD (Fig. 4A),
a B-type dehydrogenase. These data confirm the prediction made
from the previous X-ray crystallographic analysis, which sug-
25
was carried out in a manner analogous to that for NADH. The
+
unlabelled NAD displayed a signal at d 8.81 ppm (brdt, J4H–5H
gested [from analysis of the ternary complex with (−)-matairesinol
(
6a) rather than (−)-secoisolariciresinol (4a)] that the B-face of
the nicotinamide ring was open to the cleft with the C-4 of the
˚
nicotinamide ring being ∼5 A from the target hydroxyl group of
the substrate 4a. The current study now demonstrates that both
the nicotinamide and the substrate are in the proper orientation
for B-face specific hydride transfer to C-4 from the substrates [(−)-
secoisolariciresinol (4a)/lactol (5a)].
Isothermal titration calorimetry
41–43
Isothermal titration calorimetry (ITC)
was next employed
+
to study SDH_Pp7 in terms of its NAD cofactor-binding
characteristics and differential ligand affinity between the enan-
tiomeric substrates 4a and 4b. In every case, the calorimetric
data revealed that heat was released when enantiomerically pure
substrates 4a and 4b and cofactor were individually associated
with SDH_Pp7, indicating that those interactions had significant
enthalpic contributions in binding (Table 3). It also revealed
slightly unfavorable entropic contributions for each, possibly
indicating that the enzyme was slightly stabilized upon binding.
1
+
Fig. 4 Expansion of aromatic region of the H NMR spectra of NAD .
(
+
+
2
+
This effect is especially noticeable for NAD , as previously
indicated by the significant reduction of B-values of one loop
A) Natural abundance NAD (Sigma). (B) [4- H]NAD . Resonances
denoted by C H, C H, C H and C H are from protons at carbons 2,
, 5 and 6 of the pyridine ring, respectively, whereas AC H and AC H are
those at carbons 2 and 8 of the adenine (A) ring. RC1ꢀ H and AC1ꢀ H refer
to the anomeric carbons of the ribose rings in NAD , respectively.
2
4
5
6
+
constituting the binding pocket upon formation of the NAD
4
2
8
25
binary complex. As shown in Fig. 5 and Table 3, both (−)- and
(+)-secoisolariciresinols 4a and 4b showed comparable binding
+
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 808–816 | 811

View Article Online
1
53
171
167
+
Table 3 SDH_Pp7 binding parameters determined by isothermal titra-
residues, i.e. of Ser , Lys and Tyr . Binding of NAD (shown
tion calorimetry
2
+
as [4- H]NAD ) then releases the bound water molecules in an
entropically favorable manner, with the phenolic and quaternary
ammonium groups at Tyr and Lys hydrogen bonded to the 2
and 3 hydroxyl groups of the ribosyl ring, respectively, thereby
fixing the position of the cofactor during catalysis (Fig. 6B).
Binding of NAD to Lys in this way favors deprotonation of
the phenolic Tyr group, thus lowering its pK , with the resulting
phenolate anion being further stabilized by hydrogen bonding
to the Ser hydroxyl group. Following substrate binding, the
Tyr phenolate group then serves as a general base in substrate
1
−1
Compound(s)
K
d
/lM
DH/kcal mol− DS/cal mol−1 degree
)
1
67
171
ꢀ
ꢀ
4
4
a
b
44.5 ± 3.4 −19.4 ± 0.8
50.6 ± 3.7 −22.9 ± 1.0
5.5 ± 0.3 −28.6 ± 0.8
2.3 ± 0.1 −12.2 ± 0.1
4.5 ± 0.7 −25.7 ± 0.7
−45.1
−57.0
−76.5
−15.2
−43.3
+
+
NAD
+
171
NAD /4a
NAD /4b
1
67
+
a
1
53
+
affinities to SDH_Pp7 in the absence of NAD . However, when
167
the latter is present, there is a 10–20 fold decrease in K values, with
d
deprotonation and hence facilitates hydride transfer during SDH
catalysis as demonstrated in this study.
a significant difference (∼2 fold) in binding of the (−)-antipode 4a
25
(
K
d
= 2.3 lM) over that of the (+)-enantiomer 4b (K
d
+
= 4.5 lM).
Deprotonation of the bound (−)-secoisolariciresinol (4a) is
next followed by concomitant intramolecular cyclization/hydride
transfer (Fig. 6C), to afford the intermediary lactol (−)-5a, i.e.
This is indicative of a significant involvement of NAD to facilitate
binding interactions of the substrate, even though only 4a is
processed catalytically. Taken together, we interpret the ITC data
2
+
where hydride addition occurs at the B face of [4- H]NAD to
+
as indicating that NAD binds first, followed by the substrate, with
2
produce [4R– H]NADH (Fig. 6D). The latter is then released
from the active site since the Tyr phenolic group no longer
holds the resulting neutral NADH. In an analogous manner, the
catalytic conversion of (−)-4a, and NADH leaving last.
167
subsequent conversion of (−)-lactol (5a) to (−)-matairesinol (6a)
2
+
involves binding of a second molecule of [4- H]NAD as before,
with reiteration of the catalytic process (Figs. 6E and 6F), thereby
2
generating a second molecule of [4R– H]NADH and the final
product, (−)-matairesinol (6a). Thus, the overall SDH-mechanism
envisaged is concerted.
Conclusion
SDH belongs to the large, functionally heterogeneous, SDR
protein family, with the latter being of increasing scientific interest
since several are correlated with hormone-dependent cancers,
44
obesity, diabetes, and infectious diseases. In our studies, analysis
of SDH X-ray crystal structure, site-directed mutagenesis and
NMR spectroscopic data have resulted in delineation of the
catalytic mechanism including that of the role of the conserved
residues. Site-directed mutagenesis of two of the catalytic triad
1
67
171
residues (Tyr and Lys ) completely abolished catalytic activity,
this being presumably explicable on the basis of their direct roles
+
in catalysis and in binding of the ribose unit within NAD by
1
67
171
153
153
Tyr /Lys . While replacement of Ser by Ala did not result
in complete abolition of SDH activity, it nevertheless plays an
important role in catalysis since only modest amounts of lactol (5a)
were formed but not (−)-matairesinol (6a) with the S153A mutant.
Lastly, our cumulative sequence data, detailed 3-D structures
and knowledge of the catalytic mechanism of the participating
enzymes in the lignan biosynthetic pathway are systematically
providing a comprehensive understanding of their physiological
roles in vascular plants. This will also facilitate the opportunity
to initiate metabolic engineering approaches that should provide
a facile source of these beneficial lignans in foodstuffs or of
antiviral/anticancer lignans, such as podophyllotoxin.
Fig. 5 Heat of injection experimentally determined during titration of
potential substrates and cofactor in the presence of SDH_Pp7. (A) With
+
(
−)- and (+)-secoisolariciresinols (4a and 4b). (B) With NAD alone,
together with (−)- and (+)-secoisolariciresinols (4a and 4b). Solid lines
represent the least square fits of the data using a one-site binding model.
Catalytic mechanism of SDH
25
Based on the predictions from X-ray crystal structural analysis,
Experimental
and the results from the current study, a catalytic mechanism for
SDH can now be proposed (Fig. 6). Beginning with the apo-
SDH_Pp7 structure (Fig. 6A), several water molecules form a
hydrogen bonded network with the hydroxyl, quaternary ammo-
nium and phenolic groups of the highly conserved catalytic triad
Materials
Horse liver alcohol dehydrogenase, yeast aldehyde dehydrogenase,
Leuconostoc mesenteroides glucose-6-phosphate dehydrogenase
8
12 | Org. Biomol. Chem., 2006, 4, 808–816
This journal is © The Royal Society of Chemistry 2006

View Article Online
Fig. 6 Proposed catalytic mechanism of SDH_Pp7.
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 808–816 | 813

View Article Online
and rabbit muscle a-glycerophosphate dehydrogenase were pur-
chased from Sigma-Aldrich, as well as ethanol-d (99.5 atom
D), D-glucose-1-d (97 atom %D), b-NAD (sodium salt), NADH
98% pure, disodium salt), DEAE cellulose and dihydroxyacetone
phosphate (dilithium salt). D O (99.96 atom% D) was obtained
from Cambridge Isotope Lab, Inc., MA, whereas Bio-Gel P-2
45–90 lm) was purchased from Bio-Rad Laboratories. All other
linear prediction from 512 real points to 1024 real points followed
by apodization with a Gaussian function of 0.033 s time constant,
zero filled to 4 K complex points and Fourier transformed to give
6
+
%
(
−
1
−1
digital resolution of 4.6 Hz pt in F2 and 9.2 Hz pt in F1.
2
Synthesis of (−)-and (+)-secoisolariciresinols 4a and 4b
(
Racemic (± )-pinoresinols (2a/2b) were synthesized as described
solvents and chemicals used were reagent or HPLC grade.
18
in Xia et al. by oxidative coupling of E-coniferyl alcohol (1)
with FeCl
3
. The racemate 2a/2b was then resolved using a
Analytical methods
48
Chiralcel OD (Chiral Technologies, Inc.) column to give the
(
+)-(2a, 21.1 mg) and (−)-(2b 24.0 mg) antipodes, respectively.
UV spectra were recorded on a Perkin-Elmer Lambda 20 UV–VIS
instrument. HPLC analyses were carried out using an Alliance
Enantiomerically pure (+)- and (−)-pinoresinols (2a and 2b)
were next individually converted to (+)-(3a, 7.8 mg) and (−)-(3b,
2
2
695 HPLC system (Waters, Milford, MA) with detection at
80 nm. Reversed-phase HPLC used a NovaPak C18 column
8
(
.8 mg) lariciresinols together with (−)-(4a, 3.0 mg) and (+)-
4b, 4.0 mg) secoisolariciresinols, respectively, as described in
(
150 × 3.9 mm, Waters) with the following elution conditions
at a flow rate of 1 cm min : linear gradients of CH CN: 3%
O from 10 : 90 to 35 : 65 between 0 and
5 min, then to 95 : 5 in 5 min with this composition held for
49
Katayama et al.
3
−1
3
(v/v) AcOH in H
2
+
Synthesis of deuterated NADH and NAD
1
2
an additional 2 min, and finally returned back to 10 : 90 in
[4R– H]NADH. The synthesis and purification of [4R–
2
34
2
min, this being held for 15 min. Chiral HPLC separations of (+)-
H]NADH was carried out as described by Viola et al. and
50
and (−)-secoisolariciresinols (4a) and (4b) employed a Chiralcel
Anderson and Lin, respectively, with the following modifications:
Ethanol-d (2.8 mM) was incubated with NAD (5.6 mM), horse
liver alcohol dehydrogenase (50 units), and yeast aldehyde dehy-
drogenase (100 units), with the reaction monitored by following
the increase in absorbance at 340 nm. After completion of the
+
OD (Chiral Technologies, Inc.) column (250 × 4.6 mm) eluted
6
3
−1 17,45
with hexanes : EtOH (7 : 3) at a flow rate of 0.5 cm min .
Liquid chromatography-mass spectrometry with an atmospheric
pressure chemical ionization interface system (LC–APCI–MS)
was performed using a Finnigan MAT (San Jose, CA, USA) ion
trap mass spectrometer equipped with a Finnigan electrospray
interface, with LC separations being carried out on an Alliance
Instrument as described above for HPLC analyses. A fast-protein
liquid chromatography (FPLC, Amersham Pharmacia Biotech)
and a BioCad (Applied Biosystems) systems were used for
purification of reduced nucleotides and SDH, respectively.
3
reaction (3 h), carbon tetrachloride (1 cm ) was added to the
reaction mixture, with the whole vortexed and centrifuged (290g,
3
5 min). The aqueous phase (∼10 cm ) containing the reduced
nucleotide was next filtered (0.45 lm) and applied to a DEAE
cellulose column (1.6 × 40 cm) previously equilibrated in 10 mM
Tris-base buffer (pH 9.2). After washing the column with the same
3
2
buffer (35 cm ), [4R– H]NADH was next eluted (and separated
from unreacted NAD ) with a linear gradient of NaCl (from 0
+
to 0.4 M) in 10 mM Tris-base buffer (pH 9.2) over 150 min, at a
NMR spectroscopy
3
−1
flow rate of 1 cm min and with detection at 280 nm. Fractions
1
2
One dimensional H and two dimensional HMQC spectra were
containing [4R– H]NADH and having a A : A340 ratio of less
2
60
collected on a Varian Inova 500 MHz spectrometer operating
than 2.4 were individually pooled, desalted on a Bio-Gel P-2
1
13
at 499.85 MHz for H and 125.67 MHz for C, respectively.
Proton spectra obtained in D O were referenced to the TSP
sodium-3-trimethylsilyl-(2,2,3,3-d )propionate) proton signal at
.0 ppm, whereas the C dimension of the HMQC spectra was
column (2.6 × 40 cm) equilibrated in distilled H O (pH 9.0) at a
2
3
−1
2
flow rate of 0.8 cm min , then combined, lyophilized and stored
◦
2
(
0
4
at −80 C. The isotopic purity of [4R– H]NADH was determined
1
3
1
by H NMR and HMQC spectroscopic analyses to be ∼89%.
referenced indirectly to the TSP signal via the method described
2
2
[
4S– H]NADH. The synthesis of [4S– H]NADH was carried
out as described by Viola et al. by incubating D-glucose-1-d
13.5 mM, 10 mg), NAD (12.1 mM, 34 mg) and Leuconostoc
46
by Wishart et al. J values for the dihydropyridine ring of NADH
34
were calculated by extensive homonuclear spin decoupling and
+
(
47
simulation using the MestRe-C program. Gradient enhanced
phase-sensitive HMQC spectra were obtained using the standard
Varian pulse sequence, with HMQC spectra collected with sweep
mesenteroides glucose-6-phosphate dehydrogenase (42 units), but
with the following modifications: the reaction progress was
monitored by measuring the increase in absorbance at 340 nm
and was completed by 28 h, at which point the 260 to 340 nm
1
widths (acquisition times) of 4,668 Hz (219 ms) in t2 ( H) and
1
8854 Hz (6.8 ms, 128 × 2 hypercomplex increments or 27.2 ms for
3
ratio was 2.3. Carbon tetrachloride (1 cm ) was then added to
13
5
12 × 2 hypercomplex increments) in t1 ( C). Data were processed
the reaction mixture, with the whole vortexed and centrifuged
in F2 by applying a Gaussian function with a 0.101 s time constant
prior to Fourier transformation. The F1 processed data utilized
a linear prediction of the original 128 real points, in the case of
data sets acquired for 6.8 ms in t1, to 256 points, apodizing with a
Gaussian function with a 0.011 s time constant followed by zero
filling to 2 K complex points, followed by Fourier transformation
(
290g, 5 min). After reduction of the aqueous phase volume
to two thirds of its original volume under reduced pressure at
◦
2
3
5 C, [4S– H]NADH was purified as described above for [4R–
2
1
H]NADH. The isotopic purity was again determined by H NMR
and HMQC spectroscopic analyses to be ∼91%.
−
1
−1
2
+
leading to digital resolution of 4.6 Hz pt in F2 and 18.4 Hz pt
in F1. Data sets collected for 27.2 ms in t1 were first extended by
[4- H]NAD . This was enzymatically synthesized as described
40
2
by Oppenheimer by incubating [4R– H]NADH (5 mg, 0.48 mM),
8
14 | Org. Biomol. Chem., 2006, 4, 808–816
This journal is © The Royal Society of Chemistry 2006

View Article Online
dihydroxyacetone phosphate (25.5 mg; 9.3 mM) and rabbit muscle
a-glycerophosphate dehydrogenase (460 units) at 30 C for 8 to
153 (S → A), 167 (Y → A) and 171 (K → A), respectively,
were confirmed by sequencing using pTrcHis2 forward and
reverse primers (Invitrogen) flanking the gene. Both strands were
completely sequenced using an automated DNA sequencer with
◦
1
0 h, at which time solid trichloroacetic acid was added. After
centrifugation (290g, 10 min), the supernatant was lyophilized,
3
40
TM
with the resulting residue reconstituted in H
solution was next applied to a Bio-Gel P-2 column (1.6 ×
00 cm) equilibrated in distilled water (pH 7.5) at a flow
2
O (1 cm ). This
Big-Dye terminator technology to ensure that there were no
other mutation(s) in the open reading frame as a result of
PCR. The three constructs were individually transformed into
the E. coli strain TOP 10 (Invitrogen) for protein expression, and
1
3
−1
2
+
rate of 0.3 cm min in order to separate [4- H]NAD from
unreacted [4R– H]NADH. Fractions that absorbed at 260 nm
were pooled and lyophilized. The isotopic purity of [4- H]NAD
was determined by H NMR spectroscopic analysis to be ∼88%.
2
−3
selected on LB plates containing 100 lg cm carbenicillin. The
2
+
resulting clones were designated as S153A, Y167A and K171A,
respectively. Individual mutant lines (S153A, Y167A and K171A)
were next grown overnight in LB medium, with the heterologously
expressed proteins purified and assayed for (−)-matairesinol (6a)
formation as described above for wild type SDH_Pp7.
1
Expression and partial purification of SDH
Podophyllum peltatum SDH_Pp7, cloned into an Invitrogen
ꢀ
R
pTrcHis2-TOPO TA vector, was transformed and expressed into
SDH assays
18
TOP10 E. coli as previously described. The cell pellet containing
SDH_Pp7 was suspended in BugBuster (10 cm ), lysozyme
10 kilounits mm , 1 mm ) and benzonase (25 units mm ,
0 mm ). After shaking at room temperature for 10 min, the
slurry was centrifuged (8,000g, 10 min). The supernatant was
next subjected to (NH SO precipitation, with the proteins
precipitating between 20 to 60% (NH SO reconstituted in buffer
3
Matairesinol (6) formation. Each assay consisted of (± )-
3
+
−
3
3
−3
secoisolariciresinols (4a/b, 5 mM, 10 mm ), NAD (10 mM,
(
1
3
3
10 mm ), the purified SDH_Pp7 or mutant, S153A, Y167A and
K171A (1 to 4 lg) and Tris-HCl buffer (50 mM, pH 8.8) to a
final volume of 200 mm . After incubation for 10 min at room
temperature, the reaction was stopped by addition of glacial
AcOH (10 mm ). An aliquot (80 mm ) was next subjected to
3
4
)
2
4
4
)
2
4
3
3
A (50 mM Tris-HCl, pH 7.5), and further desalted over PD-10
columns (Amersham Pharmacia Biotech). The desalted fraction
was then applied to a DEAE cellulose column (2.5 × 27 cm)
reversed-phase HPLC to separate 4 from the products 5 and 6
(see Analytical methods section).
3
−1
equilibrated with buffer A at a flow rate of 1 cm min . After
2
[
4- H]NADH formation. To assay mixtures consisting of (± )-
3
washing the column with buffer A (5 cm ), SDH_Pp7 was eluted
as follows: linear NaCl gradients in buffer A first from 0 to 1 M
in 100 cm and then from 1 to 2 M in 20 cm with this final NaCl
concentration held for another 10 cm . The SDH_Pp7 containing
fractions, eluted at ∼0.2 M NaCl were combined, concentrated
and desalted into Buffer A by ultrafiltration (Amicon cell)
using a 10 kDa cut-off membrane (Millipore). The concentrated
SDH_Pp7 was next applied to a POROS HQ anion exchange
3
secoisolariciresinols (4a/b) (5 mM in MeOH, 0.61 cm ), 50 mM
3
2
+
Tris-HCl buffer (pH 8.8, 9.29 cm ) and [4- H]NAD [4.2 mg in
3
3
3
50 mM Tris-HCl buffer (pH 8.8), 2 cm ] was added the partially
3
3
purified SDH_Pp7 as described above (0.305 cm , 884.5 lg). After
incubation at room temperature with shaking for 10 min, the
reaction mixture was quenched with EtOAc (6 cm ). To establish
the integrity of the enzymatic reaction, the EtOAc solubles were
3
dried (Na
2
SO ) and evaporated to dryness in vacuo. The EtOAc-
4
(
7
4.6 × 100 mm) column equilibrated in buffer A at a flow rate of
3
derived residue was reconstituted in MeOH (2 cm ) with an aliquot
3
−1
3
cm min . After washing the column with buffer A (3.3 cm ),
3
(
20 mm ) subjected to reversed-phase HPLC analysis with the
proteins were next eluted with NaCl linear gradients first from 0
to 0.3 M in buffer A (50 cm ) and then to 1 M (8.3 cm ), with this
final concentration further held for 3.3 cm . Fractions containing
SDH_Pp7 eluting between 0.13 and 0.17 M NaCl were combined,
desalted and subjected to POROS HQ chromatography under the
same conditions as above, with fractions containing SDH_Pp7
combined and directly used for assays.
enzymatically formed matairesinol (6a) identified by LC–APCI–
MS analysis as described in the Analytical methods section.
The chirality of the enzymatically formed matairesinol (6) was
determined as previously described.
was subjected to centrifugation (1,200g, 10 min), filtered (0.45 lm),
with the enzymatically formed [4- H]NADH purified as des-
cribed above. Fractions containing [4- H]NADH were pooled,
lyophilized and subjected to H NMR and HMQC spectroscopic
3
3
3
17,45
Finally, the aqueous phase
2
2
1
Site-directed mutagenesis of SDH_Pp7
analyses.
Primers (Table 1) were designed and synthesized (Invitrogen,
Carlsbad, USA) to individually convert Ser , Tyr and Lys
Isothermal titration calorimetry (ITC)
1
53
167
171
4
1–43
◦
in Podophyllum peltatum SDH_Pp7 into the hydrophobic non-
reactive alanine. Site specific mutations were performed using
a QuikChange XL site-directed mutagenesis kit (Stratagene, La
Jolla, USA) following the manufacturer’s instructions with PCR
Isothermal titration calorimetry (ITC)
was performed at 25 C
using a VP-ITC instrument (MicroCal, Northampton, MA).
Enantiomerically pure substrates 4a and 4b were first individually
dissolved in MeOH (0.3 cm , 48 mM), with buffer B (20 mM
Tris, pH 8.0, 1 mM DTT, 1 mM EDTA, 10% glycerol) next added
such as the MeOH content was 1% and the final concentration of
either 4a or 4b was 0.48 mM. NAD (0.48 mM) was dissolved in
buffer A containing 1% MeOH. Purified SDH_Pp7 was dialyzed
extensively against buffer B in an Amicon stirred cell at 4 C, after
3
◦
conditions modified as follows: initial denaturation at 95 C for
◦
◦
1
6
min, 18 cycles at 95 C for 1 min, 60 C for 1 min and
◦
+
8 C for 7 min, followed by Dpn I digestion and transformation
into XL10 Gold ultracompetent E. coli cells. Transformants
were selected on LB plates containing 100 lg cm carbenicillin.
ꢀ
R
25
−
3
◦
Positive clones containing a single point mutation at positions
which MeOH was added to give a 1% final concentration. The
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 808–816 | 815

View Article Online
final enzyme concentration (30 lM of monomer) was determined
13 L. B. Davin, H.-B. Wang, A. L. Crowell, D. L. Bedgar, D. M. Martin,
51
S. Sarkanen and N. G. Lewis, Science, 1997, 275, 362–366.
by the Bradford method using BSA as standard. Both the enzyme
and substrate/cofactor solutions were individually degassed in
vacuo for 5 min before ITC. Titration experiments were performed
1
4 M. Fujita, D. R. Gang, L. B. Davin and N. G. Lewis, J. Biol. Chem.,
1
999, 274, 618–627.
15 D. R. Gang, M. A. Costa, M. Fujita, A. T. Dinkova-Kostova, H.-B.
Wang, V. Burlat, W. Martin, S. Sarkanen, L. B. Davin and N. G. Lewis,
Chem. Biol., 1999, 6, 143–151.
3
as follows: aliquots (10 mm ) of the ligand solution (4a, 4b or
NAD) were injected into the reaction cell containing SDH_Pp7
1
6 N. G. Lewis and L. B. Davin, in Comprehensive Natural Products
Chemistry, ed. Sir D. H. R. Barton, K. Nakanishi and O. Meth-Cohn,
Elsevier, Oxford, 1999, vol. 1, p. 639-712.
3
(
2 cm ) with stirring set at 300 rpm. Twenty nine injections for each
assay condition were performed with an equilibration interval of
1
1
7 Z.-Q. Xia, M. A. Costa, J. Proctor, L. B. Davin and N. G. Lewis,
Phytochemistry, 2000, 55, 537–549.
8 Z.-Q. Xia, M. A. Costa, H. C. P e´ lissier, L. B. Davin and N. G. Lewis,
J. Biol. Chem., 2001, 276, 12614–12623.
4
00 s between each.
Binary complex titrations were individually carried out by
+
adding 4a, 4b or NAD solution to the SDH_Pp7 solution in
the reaction cell. Ternary complex titrations were carried out by
mixing SDH_Pp7 and NAD in a 1 : 1 ratio (30 lM each) before
19 S. C. Halls and N. G. Lewis, Biochemistry, 2002, 41, 9455–9461.
20 T. Min, H. Kasahara, D. L. Bedgar, B. Youn, P. K. Lawrence, D. R.
Gang, S. C. Halls, H. Park, J. L. Hilsenbeck, L. B. Davin, N. G. Lewis
and C. Kang, J. Biol. Chem., 2003, 278, 50714–50723.
21 M.-H. Cho, S. G.A. Moinuddin, G. L. Helms, S. Hishiyama, D.
Eichinger, L. B. Davin and N. G. Lewis, Proc. Natl. Acad. Sci. U. S. A.,
+
adding the solution to the reaction cell, with titrations individually
performed using solutions of enantiomerically pure 4a or 4b. Heats
of dilution were determined by titration of each ligand individually
in Buffer A.
The experimental data were fitted to an n-equivalent binding
site model using the nonlinear least-squares regression from
the Origin software package (OriginLab Corp, Northampton,
MA). Since the number of binding sites, n, converged to values
between 0.9 and 1.1 in the initial regression analyses, a single-site
binding model (1.0) was used for the final analyses. This yielded
2
003, 100, 10641–10646.
2
2
2 L. B. Davin and N. G. Lewis, Phytochem. Rev., 2003, 2, 257–288.
3 S. G. A. Moinuddin, S. Hishiyama, M.-H. Cho, L. B. Davin and N. G.
Lewis, Org. Biomol. Chem., 2003, 1, 2307–2313.
4 S. C. Halls, L. B. Davin, D. M. Kramer and N. G. Lewis, Biochemistry,
2
2
2
004, 43, 2587–2595.
5 B. Youn, S. G. A. Moinuddin, L. B. Davin, N. G. Lewis and C. Kang,
J. Biol. Chem., 2005, 280, 12917–12926.
26 L. B. Davin and N. G. Lewis, Curr. Opin. Biotechnol., 2005, 16, 398–406.
2
2
2
3
7 H. Adlercreutz, Bailliere’s Clin. Endocrinol. Metab., 1998, 12, 605–
23.
8 A. Stark and Z. Madar, J. Pediatr. Endocrinol. Metab., 2002, 15, 561–
72.
9 S. P. Borriello, K. D. R. Setchell, M. Axelson and A. M. Lawson,
J. Appl. Bacteriol., 1985, 58, 37–43.
the following thermodynamic parameters: binding constant (K ),
d
6
binding enthalpy (DH) and the entropy change (DS), respectively.
5
Acknowledgements
0 J. A. F. Gardner, G. M. Barton and H. MacLean, Can. J. Chem., 1959,
37, 1703–1709.
This research was supported in part by the National Institutes
of Health (GM66173), the United States Department of Agricul-
ture (99-35103-8037), the National Science Foundation (MCB-
31 J. A. F. Gardner, B. F. MacDonald and H. MacLean, Can. J. Chem.,
1
960, 38, 2387–2394.
3
2 J. A. F. Gardner, E. P. Swan, S. A. Sutherland and H. MacLean,
0
417291), the G. Thomas and Anita Hargrove Center for Plant
Can. J. Chem., 1966, 44, 52–58.
33 W. L. Duax, J. F. Griffin and D. Ghosh, Curr. Opin. Struct. Biol., 1996,
Genomic Research, McIntire-Stennis, and the Murdock Chari-
table Trust. We thank D.J. Pouchnik for DNA sequencing and
G. Munske for assistance with isothermal titration calorimetry;
P.K. Lawrence also provided preliminary technical assistance in
the mutagenesis experiments.
6
, 813–823.
4 R. E. Viola, P. F. Cook and W. W. Cleland, Anal. Biochem., 1979, 96,
34–340.
3
3
3
3
5 N. J. Oppenheimer, L. J. Arnold and N. O. Kaplan, Proc. Natl. Acad.
Sci., U. S. A, 1971, 68, 3200–3205.
6 L. J. Arnold, Jr., K.-s. You, W. S. Allison and N. O. Kaplan,
Biochemistry, 1976, 15, 4844–4849.
References
37 K. Velonia, I. Tsigos, V. Bouriotis and I. Smonou, Bioorg. Med. Chem.
Lett., 1999, 9, 65–68.
38 S. B. Mostad and A. Glasfeld, J. Chem. Educ., 1993, 70, 504–506.
39 P. E. Hansen, Prog. Nucl. Magn. Reson. Spectrosc., 1988, 20, 207–255.
40 N. J. Oppenheimer, J. Biol. Chem., 1986, 261, 12209–12212.
41 T. Wiseman, S. Williston, J. F. Brandts and L. N. Lin, Anal. Biochem.,
1989, 179, 131–137.
1
2
H. Adlercreutz, J. Steroid Biochem. Mol. Biol., 2003, 83, 113–118.
F. Boccardo, G. Lunardi, P. Guglielmini, M. Parodi, R. Murialdo, G.
Schettini and A. Rubagotti, Eur. J. Cancer, 2004, 40, 84–89.
P. Meresse, E. Dechaux, C. Monneret and E. Bertounesque, Curr. Med.
Chem., 2004, 11, 2443–2466.
3
4
5
T. F. Imbert, Biochimie, 1998, 80, 207–222.
42 I. Jelesarov and H. R. Bosshard, J. Mol. Recognit., 1999, 12, 3–18.
43 J. E. Ladbury, BioTechniques, 2004, 37, 885–887.
44 U. Oppermann, C. Filling, M. Hult, N. Shafqat, X. Wu, M. Lindh, J.
Shafqat, E. Nordling, Y. Kallberg, B. Persson and H. J o¨ rnvall, Chem.-
Biol. Interact., 2003, 143–144, 247–253.
R. M. Moraes, F. E. Dayan and C. Canel, Stud. Nat. Prod. Chem.,
2
002, 26, 149–182.
6
J. A. Gonzalez, A. Estevez-Braun, R. Estevez-Reyes, I. L. Bazzocchi, L.
Moujir, I. A. Jimenez, A. G. Ravelo and A. G. Gonzalez, Experientia,
1
995, 51, 35–39.
45 T. Umezawa, L. B. Davin and N. G. Lewis, J. Biol. Chem., 1991, 266,
7
8
C.-B. Bernard, J. T. Arnason, B. J. R. Philog e` ne, J. Lam and T. Waddell,
Phytochemistry, 1989, 28, 1373–1377.
N. G. Lewis, L. B. Davin and S. Sarkanen, in Comprehensive Natural
Products Chemistry, ed. Sir D. H. R. Barton, K. Nakanishi and
O. Meth-Cohn, Elsevier, Oxford, 1999, vol. 3, p. 617–745.
N. G. Lewis, Curr. Opin. Plant Biol., 1999, 2, 153–162.
10210–10217.
46 D. S. Wishart, C. G. Bigam, J. Yao, F. Abildgaard, H. J. Dyson, E.
Oldfield, J. L. Markley and B. D. Sykes, J. Biomol. NMR, 1995, 6,
135–140.
47 J. C. Cobas and F. J. Sardina, Concepts Magn. Reson., Part A, 2003,
9
19A, 80–96.
1
1
1
0 L. B. Davin and N. G. Lewis, Curr. Opin. Biotechnol., 2005, 16, 407–
4
48 L. B. Davin, D. L. Bedgar, T. Katayama and N. G. Lewis, Phytochem-
istry, 1992, 31, 3869–3874.
15.
1 A. Chu, A. Dinkova, L. B. Davin, D. L. Bedgar and N. G. Lewis, J. Biol.
49 T. Katayama, L. B. Davin, A. Chu and N. G. Lewis, Phytochemistry,
1993, 33, 581–591.
50 J. A. Anderson and B.-K. Lin, Phytochemistry, 1993, 32, 811–812.
51 M. M. Bradford, Anal. Biochem., 1976, 72, 248–254.
Chem., 1993, 268, 27026–27033.
2 A. T. Dinkova-Kostova, D. R. Gang, L. B. Davin, D. L. Bedgar, A.
Chu and N. G. Lewis, J. Biol. Chem., 1996, 271, 29473–29482.
8
16 | Org. Biomol. Chem., 2006, 4, 808–816
This journal is © The Royal Society of Chemistry 2006