Stereoselective coupling of hemigossypol to form (+)-gossypol in moco cotton is mediated by a dirigent protein

Article abstract of DOI:10.1016/j.phytochem.2008.06.007

The terpenoid gossypol, a secondary metabolite found in the cotton plant, is synthesized by a free radical dimerization of hemigossypol. Gossypol exists as an atropisomeric mixture because of restricted rotation around the central binaphthyl bond. The dimerization of hemigossypol is regiospecific in cotton. In the case of some moco cotton, the dimerization also exhibits a high level of stereoselectivity. The mechanism that controls this stereoselective dimerization is poorly understood. In this paper, we demonstrate that a dirigent protein controls this stereoselective dimerization process. A partially purified protein preparation from cotton flower petals, which by itself is unable to convert hemigossypol to gossypol, converts hemigossypol with a 30% atropisomeric excess into (+)-gossypol when combined with an exogenous laccase, which by itself produces racemic gossypol.

Full text of DOI:10.1016/j.phytochem.2008.06.007

Phytochemistry 69 (2008) 3038–3042
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Stereoselective coupling of hemigossypol to form (+)-gossypol in moco cotton
is mediated by a dirigent protein
a
a
a
b
Jinggao Liu a, , Robert D. Stipanovic , Alois A. Bell , Lorraine S. Puckhaber , Clint W. Magill
*
a USDA-ARS, Southern Plains Agricultural Research Center, 2765 F&B Road, College Station, TX 77845, USA
b Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX 77843, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The terpenoid gossypol, a secondary metabolite found in the cotton plant, is synthesized by a free radical
dimerization of hemigossypol. Gossypol exists as an atropisomeric mixture because of restricted rotation
around the central binaphthyl bond. The dimerization of hemigossypol is regiospecific in cotton. In the
case of some moco cotton, the dimerization also exhibits a high level of stereoselectivity. The mechanism
that controls this stereoselective dimerization is poorly understood. In this paper, we demonstrate that a
dirigent protein controls this stereoselective dimerization process. A partially purified protein prepara-
tion from cotton flower petals, which by itself is unable to convert hemigossypol to gossypol, converts
hemigossypol with a 30% atropisomeric excess into (+)-gossypol when combined with an exogenous lac-
case, which by itself produces racemic gossypol.
Received 29 February 2008
Received in revised form 2 May 2008
Available online 17 July 2008
Keywords:
Dirigent protein
Moco cotton
Gossypium hirsutum
Malvaceae
(+)-Gossypol
Published by Elsevier Ltd.
Gossypol biosynthesis
Hemigossypol
Oxidative coupling
Peroxidase
Stereoselectivity
1. Introduction
only (ꢁ)-gossypol (2b) is toxic to animals. Toxicity toward insects
and pathogens is not affected by the (+)- to (ꢁ)-gossypol (2a/2b)
Cottonseed is composed of ꢀ22.5% of a high quality protein. The
estimate of US cottonseed production for 2007 is over 6.5 million
metric tons (National Agricultural Statistics Service, USDA, 2007).
This is sufficient protein to provide the total protein needs for over
71 million people for 1 year based on the USDA recommendation of
56.6 g of protein per day per person. However, cottonseed is under
utilized as a food/feed source because of the presence of a toxic
compound called gossypol (2). Gossypol (2) occurs in the seed, fo-
liage and roots of cotton; it imparts resistance against herbivorous
insects and pathogens, and is biosynthesized by the free radical
coupling of two molecules of hemigossypol (1) that yield both
atropisomers (2a/2b) (Fig. 1) (Benedict et al., 2006). The (+)-gossy-
pol (2a) has an absolute configuration of S and its stereo configura-
tion is shown in Fig. 1 (Jaroszewski et al., 1992). In most
commercial cottons (Gossypium hirsutum) grown in the U.S., the ra-
tio of (+)- to (ꢁ)-gossypol (2a/2b) is approximately 3:2 (Zhou and
Lin, 1988; Cass et al., 1991). However, this ratio can be as high as
98:2 in some moco cotton (G. hirsutum var. marie-galante) grown
in Brazil (Cass et al., 1991; Stipanovic et al., 2005). Significantly,
ratio (Stipanovic et al., 2006; Puckhaber et al., 2002). Therefore, if
the ratio could be shifted predominately to (+)-gossypol (2a), cot-
tonseed could be used as a feed for both ruminant and non-rumi-
nant animals.
Stereoselective free radical coupling reactions are a poorly
understood biochemical process involved in plant secondary
metabolite biosynthesis, as is the case for (+)-gossypol (2a) forma-
tion. Perhaps the best studied biological bimolecular free radical
coupling reaction is the regio- and stereoselective dimerization of
E-coniferyl alcohol to either (+)- or (ꢁ)-pinoresinol. Lewis’ group
(Davin et al., 1997) elegantly demonstrated the intervention of
what they call a dirigent protein. The dirigent protein, isolated
from Forsythia intermedia, by itself has no catalytic activity to con-
vert E-coniferyl alcohol to pinoresinols. However, when the diri-
gent protein is combined with a laccase from F. intermedia, only
(+)-pinoresinol is formed. If the dirigent protein is missing from
this reaction, the result is formation of a racemic mixture of prod-
ucts. Boiling deactivates the protein (Hall and Lewis, 2002). Davin
and Lewis (2000) proposed that the dirigent protein captures the
E-coniferyl alcohol-derived free radical and directs the stereoselec-
tive coupling. Homologous dirigent DNA sequences have been
detected in both gymnosperms and dicotyledonous angiosperms
* Corresponding author. Tel.: +1 979 260 9233; fax: +1 979 260 9319.
E-mail address: jinggao.liu@ars.usda.gov (J. Liu).
0031-9422/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.phytochem.2008.06.007

J. Liu et al. / Phytochemistry 69 (2008) 3038–3042
3039
because they contained a higher % of (+)-gossypol (2a) than seeds,
and are relatively free of tannins and other interfering factors that
can rapidly degrade proteins.
O
CH OH
OH
OH
HO
HO
In the present investigation, three assays to measure the forma-
tion of (+)- and (ꢁ)-gossypol (2a) and (2b) from hemigossypol (1)
were investigated using various co-substrates. The protein prepa-
ration with oxygen as co-substrate is referred to below as the
‘‘oxidase assay”, the preparation with hydrogen peroxide as co-
substrate is referred to below as the ‘‘peroxidase assay”, and the
preparation with oxygen as co-substrate and supplemented with
an excess of exogenous laccase is referred to below as the ‘‘dirigent
assay”. In each case, controls were also studied that includes: (1) a
no protein preparation; (2) a boiled protein preparation; (3) a no
hemigossypol (1) preparation; and (4) a 0 reaction time prepara-
tion. After 10–60 min, each reaction was extracted with EtOAc/
hexane (1:1, v/v) and the organic phases concentrated to dryness.
OH HC
O
O
CH OH
( )-Gossypol
HO
HO
O
O
HC
OH
CH
HO
HO
OH
Hemigossypol
OH
HO
(+)-Gossypol
The residues were derivatized with D-alaninol and the derivatives
Fig. 1. Biosynthesis of gossypol by free radical coupling of hemigossypol generated
by peroxidase in the presence or absence of a cotton dirigent protein.
of (+)- and (ꢁ)-gossypol (2a) and (2b) were qualitatively and quan-
titatively analyzed using HPLC (Kim et al., 1996). A series of exper-
iments were designed to determine if the active agent was a
stereoselective enzyme or a dirigent type protein.
(Gang et al., 1999). In 2000, Davin and Lewis showed that a protein
in flaxseed produces the corresponding (ꢁ)-antipode. Thus, non-
catalytic proteins can control the regio- and stereoselective dimer-
ization of free-radical species. Recently, Zhu et al. (2007) obtained
two dirigent-like cDNA clones, Gbd 1 and Gbd 2 from cotton (Gos-
sypium barbadense). The two genes shared identities of 82% and
were induced after infection by a fungal pathogen Verticillium dah-
liae. Phylogenetic analysis of dirigent and dirigent-like proteins
(Ralph et al., 2007) indicated that the two G. barbadense genes be-
longed to subfamily DIR-b/d, one of the five subfamilies whose bio-
chemical function are yet not established. Genes encoding dirigent
proteins for the regio- and stereoselective coupling of E-coniferyl
alcohol belonged to subfamily DIR-a.
Only one such dirigent protein has been characterized function-
ally to show that the protein guided the regio- and stereoselective
coupling of the lignan and lignin pathway substrate E-coniferyl
alcohol. Dirigent proteins that guided regio- and stereoselective
coupling of other substrates have not been demonstrated, even
though in vivo formation of regio- and stereoselectively coupled
products of phenolic substrates are ubiquitous in nature. Herein,
we show that a (+)-gossypol-forming dirigent protein mediates a
peroxidase catalyzed reaction to steer the stereoselective coupling
of a terpenoid pathway substrate, hemigossypol (1), to (+)-gossy-
pol (2a).
Dimerization of hemigossypol (1) with exogenous peroxidase,
laccase, or ammonium persulfate gave a racemic mixture of gossy-
pol [49.7 0.3%, 49.4 0.5%, and 51.1 1.6% of (+)-gossypol (2a),
respectively]. In contrast, when hemigossypol (1) was added to a
crude enzyme preparation of moco flower petals using oxygen as
co-substrate (oxidase assay), (+)-gossypol (2a) was preferentially
formed [74% (+)-gossypol (2a), 2.2 nmol total gossypol/h/g tissue
(Table 1)]. When H2O2 was added to the crude flower petal prepa-
ration (peroxidase assay), the rate of total gossypol (2a) formation
by the naturally occurring peroxidase in the crude protein prepara-
tion [11.4 nmol total gossypol/h/g tissue] was about five-fold high-
er than in the oxidase assay. Here again, (+)-gossypol (2a) was
preferentially formed (79%) (Table 1). When exogenous laccase
was added to the same protein preparation followed by hemi-
gossypol (1) (Table 1, dirigent assay) using oxygen as co-substrate,
the % of (+)-gossypol (2a) (78.0%) was slightly higher compared to
that formed in the oxidase assay and the total nmol of gossypol/h/g
tissue was ca. 10 times that observed in the oxidase assay (Table 1,
oxidase assay). When the crude preparation was boiled, and then
exogenous laccase was added, only racemic gossypol (2a/2b) was
formed (Table 1, control assay).
In the dirigent assay, a large excess of laccase [more than 10-
fold excess of oxidase activity compared to that in the crude en-
zyme preparation (oxidase assay)] is present; as shown above,
exogenous laccase alone produced racemic gossypol (2a/2b). If a
stereoselective oxidase (laccase or peroxidase) was responsible
for controlling the (+)- to (ꢁ)-gossypol ratio (2a/2b), one would ex-
pect the (+)- to (ꢁ)-ratio to decrease when the exogenous laccase
was added, this was because the added laccase would be compet-
ing for the same substrate, and relatively more racemic gossypol
(2a/2b) would be produced. Furthermore, if a stereoselective oxi-
dase were involved, based on the total nmol of gossypol (2) pro-
duced per hour per gram fresh tissue in the oxidase assay and
dirigent assay and the (+)-gossypol (2a) percentage in the oxidase
assay as shown in Table 1, only 52.3% of (+)-gossypol (2a) would be
2. Results and discussion
2.1. Enzymatic and non-enzymatic formation of (+)- and (-)-gossypol
(2a) and (2b) from hemigossypol (1)
We have previously demonstrated (Benedict et al., 2006) that
embryo extracts of the G. hirsutum var. marie-galante (moco cotton)
preferentially converted hemigossypol (1) to (+)-gossypol (2a)
[72% (+)-gossypol (2a) vs. 28% (ꢁ)-gossypol (2b)]. In this study,
we used flower petal tissue of moco cotton as the source material
Table 1
Total and (+)-gossypol formation from hemigossypol by the crude protein preparation’s oxidase activity, peroxidase activity, and dirigent activity
Assay type
Oxidase/co-substrate
Substrate
Enzyme
% (+)-Gossypol
Total nmol gossypol/h/g tissue
Oxidase
Peroxidase
Dirigent
Control
O2
H2O2
Laccase O2
Laccase O2
+
+
+
+
HG
HG
HG
HG
+
+
+
+
Crude enzyme
Crude enzyme
Crude enzyme
Boiled crude enzyme
?
?
?
?
74.2 0.4
79.4 1.1
77.9 0.3
50.7 0.8
2.2 0.3
11.4 0.9
22.3 2.4
19.6 2.5
Aliquots of 240 lL crude enzyme preparation were used in each assay. Assays were done in triplicate.

3040
J. Liu et al. / Phytochemistry 69 (2008) 3038–3042
tained in the crude enzyme preparation. The rate of total
gossypol (2a) formation increased ca. 30% when the amount of pro-
tein added in the dirigent assay was increased to ca. 23 g. The rate
remained almost constant when the added amount of protein in
the assay increased further. A slight increase (ca. 10%) in the rate
of total gossypol formation is expected due to the presence of
endogenous oxidases in the crude protein preparation.
11
10
9
90
80
70
60
50
40
%(+)-Gossypol Formed
Total Gossypol Formed
l
8
7
6
Racemic
2.2. Preferential formation of (+)-gossypol (2a) from hemigossypol (1)
using a partially purified enzyme preparation
5
4
3
To further explore the biochemical mechanism for the (+)-gos-
sypol (2a) formation, we subjected the crude protein preparation
to chromatography on a CM-Sepharose FF column. The protein
concentration of the individual fractions collected from the column
was determined (Fig. 3). Selected fractions were assayed using the
dirigent assay (i.e., enzyme preparation and exogenous laccase)
and the percent (+)-gossypol (2a) formed in the assays were deter-
mined. Excess (+)-gossypol (2a) was produced by fractions 28–34.
Fraction 32 was selected for further study because it showed the
highest activity. An aliquot of the original crude enzyme prepara-
tion that also gave approximately 65% (+)-gossypol (2a) in a diri-
gent assay (Table 2, assay #1) was also selected. In Table 2, we
compared the ability of these two preparations to form (+)-gossy-
pol (2a) in the dirigent assay, peroxidase assay, and oxidase assay.
In the peroxidase and oxidase assays (Table 2, assays #7 and 8,
respectively), the no enzyme control experiments formed about
30 and 40 pmol of gossypol (2), respectively. Thus, values close
to these numbers indicate that no or very little gossypol (2) was
produced. In the peroxidase assay (Table 2, assay #2), which in-
cluded the crude enzyme, hemigossypol (1) and H2O2, over 65
times more total gossypol (2) was produced as compared to the
purified fraction (Table 2, assay #5). The purified fraction produced
less gossypol (2) in the peroxidase assay (Table 2, assay #5) than in
the assay with no enzyme added (Table 2, assay #7). This demon-
strates that the column successfully separated the protein that
controls the (+)- to (ꢁ)-gossypol (2a/2b) ratio from the endogenous
peroxidase that is present in the crude preparation. Similarly,
when hemigossypol (1) was added alone to the crude enzyme
preparation with oxygen as co-substrate (Table 2, assay #3),
646 pmol gossypol/assay/h was produced; yet no gossypol (2)
was produced in the oxidase assay with Fraction 32 (Table 2, assay
#6). However, when Fraction 32 was mixed with laccase (Table 2,
assay #4), 9743 pmol of gossypol (2) were produced of which 65%
was (+)-gossypol (2a). Thus, the protein in Fraction 32 was incapa-
ble of converting hemigossypol (1) into gossypol (2a) by itself, but
required the addition of an oxidative enzyme such as laccase.
When laccase was added, gossypol (2a) was readily produced with
a (+)- to (ꢁ)-gossypol (2a/2b) ratio of 13:7. These results demon-
strate that the stereoselective formation of (+)-gossypol (2a) in
moco cotton was mediated by a dirigent protein.
2
100
0
20
40
60
80
Amount of Protein Added (μg)
Fig. 2. Dependence of % (+)-gossypol formation and total gossypol formation on the
amount of crude dirigent protein preparation added to the dirigent assay mixture
containing hemigossypol and exogenous laccase. Experiments were done in
triplicate.
formed in the dirigent assay. As seen, the opposite was observed
and the percent (+)-gossypol (2a) increased slightly. These results
argue against the involvement of a stereoselective oxidase, and
they are consistent with the operation of a dirigent type protein.
The (+)-gossypol (2a) percentage increased steadily from race-
mic (ca. 50%) to 80% in the dirigent assay, when the amount of pro-
tein preparation of the flower petal added to the assay increased
from 0 to 93 lg (Fig. 2), consistent with a mechanism involving a
dirigent protein. Again, all assays contained at least a 10-fold ex-
cess of oxidase activity from exogenous laccase than that con-
80
75
70
65
60
55
50
45
40
8
7
6
5
4
3
2
1
0
% (+)-Gossypol Formed
Protein Conc.
Racemic
10
15
20
25
30
35
Fraction Number
Fig. 3. Protein concentration and % (+)-gossypol formation in fractions purified by a
CM-Sepharose FF column with hemigossypol and laccase in the dirigent assay.
Aliquots of 240
in triplicate.
lL column fractions were used in each assay and assays were done
Table 2
Comparison of CM-Sepharose FF column Fraction 32 and the crude enzyme preparation for ability to form (+)-gossypol as determined by their respective dirigent protein activity,
peroxidase activity, and oxidase activity
Assay # and type
Oxidase/co-substrate
Substrate
Enzyme
% (+)-Gossypol
Total pmol gossypol/assay/h
1. Dirigent
2. Peroxidase
3. Oxidase
4. Dirigent
5. Peroxidase
6. Oxidase
Laccase O2
H2O2
O2
Laccase O2
H2O2
O2
+
+
+
+
+
+
+
+
HG
HG
HG
HG
HG
HG
HG
HG
+
+
+
+
+
+
Crude enzyme
Crude enzyme
Crude enzyme
Fraction #32
Fraction #32
Fraction #32
65.4 0.6
67.1 1.4
64.0 1.1
65.3 0.9
68.4 5.8
59.7 6.3
57.3 4.5
54.9 2.3
6812 1374
1696 246
646 35
9743 1829
26 12
34 17
7. Peroxidase control
8. Oxidase control
H2O2
O2
30
9
40 13
Aliquots of 23
triplicate.
lL of crude enzyme preparation were used in assays #1, 2, and 3. Aliquots of 240 lL of Fraction 32 were used in assays #4, 5, and 6. Assays were done in

J. Liu et al. / Phytochemistry 69 (2008) 3038–3042
3041
to the procedures of Laemmli (1970). Protein concentration was
determined by the method of Bradford (1977).
80
70
60
50
40
20
15
10
5
% (+)-Gossypol Formed
in Dirigent Assay
Total Gossypol Formed
in Peroxidase Assay
3.2. Plant material
Plants of G. hirsutum var. marie-galante (moco cotton) were
grown and maintained in a greenhouse in 2 gal pots. Unopened
flower petal tissues were collected on the day of blooming and fro-
zen in liquid N2 and stored at ꢁ80° C.
Racemic
3.2.1. Dirigent assay
0
90
This assay is designed to detect the presence of dirigent protein
even with the enzyme preparations containing no free radical gen-
erating ability. Thus, crude enzyme preparations and fractions col-
0
10
20
30
40
50
60
70
80
Fraction Number
lected during purification were assayed in a total volume of 300
0.1 M KPi buffer, pH 5.5 containing 104 M hemigossypol (1)
(20 L, 1.56 mM HG dissolved in ethanol), Coriolus versicolor lac-
case (8.3 mU), and the enzyme preparations (up to 100 g pro-
teins). The assay mixtures were incubated for 10 min at 30° C.
Controls were run either without C. versicolor laccase, without
HG, without enzyme preparation, or with boiled (10 min) enzyme
preparation. The reaction products and substrate were extracted
lL
Fig. 4. Dirigent activity and peroxidase activity in fractions from Q-Sepharose FF
column. Dirigent activity is expressed as % (+)-gossypol formed in dirigent assay
and peroxidase activity is expressed as total gossypol formed in peroxidase assay.
l
l
Aliquots of 240
in triplicate.
lL column fractions were used in each assay and assays were done
l
2.3. Native molecular mass and substrate specificity of the (+)-
gossypol-forming dirigent protein
three times with 160
phases were concentrated to dryness. The residues were deriva-
tized with -alaninol and the resulting diastereomeric adducts
lL EtOAc/hexane (1:1, v/v) and the organic
We also used a Q-Sepharose FF column to fractionate the diri-
gent protein. The separation between the peroxidase activity and
dirigent activity peaks was much less than that achieved by CM-
Sepharose FF column (Fig. 4), but the overall recovery of dirigent
activity was at least 20-fold higher. That is, twelve 5 mL fractions
(Fractions #67–78, Fig. 4) from the Q-Sepharose FF column pro-
vided >60% (+)-gossypol (2a) dirigent activity, while only two to
three fractions from the CM-Sepharose FF column gave >60% (+)-
gossypol (2a) dirigent activity. Pooled dirigent active fractions
from the Q-Sepharose FF column was used to determine the native
molecular mass of the dirigent protein on a calibrated gel filtration
column. A major peak and a minor peak corresponding to native
molecular mass of 77.2 and 122.5 kDa were observed, reflecting
the tendency of dirigent proteins to form aggregates. The (+)-gos-
sypol-forming dirigent protein obtained from pooled active frac-
tions of Q-Sepharose FF column converted a structurally related
substrate present in some cotton species, hemigossypol 6-methyl
ether, to (+)-gossypol 6,60-dimethyl ether, but it did so less effi-
ciently [81.7 0.5% vs. 62.7 0.8% (+)-gossypol (2a) from hemi-
gossypol (1) and hemigossypol 6-methyl ether, respectively].
Very low levels of gossypol 6-methyl ether are present in the seed
of some cotton cultivars, but are not observed in moco cottonseed
(Stipanovic et al., 2005).
D
formed from (+) and (ꢁ)-gossypol (2a) and (2b) were qualitatively
and quantitatively analyzed on a HPLC by the method of Kim et al.
(1996).
3.2.2. Oxidase assay
The oxidative coupling of hemigossypol (1) for the formation of
gossypol (2) by the enzyme preparation’s endogenous laccase type
oxidase activity was assayed in a total volume of 300
buffer, pH 5.5 containing 104 M hemigossypol (1), the enzyme
preparations (up to 100 g proteins), and oxygen from the air.
lL 0.1 M KPi
l
l
The assay mixtures were incubated for 30 min at 30° C. Controls
were run either without HG, without enzyme preparation, or with
a boiled (10 min) enzyme preparation. The reaction products and
substrate were analyzed as described in the above paragraph.
3.2.3. Peroxidase assay
The oxidative coupling of hemigossypol (1) for the formation of
gossypol (2) by the enzyme preparation’s endogenous peroxidase
activity were assayed in a total volume of 300
pH 5.5 containing 104 M hemigossypol (1), 5 mM hydrogen per-
oxide and the enzyme preparations (up to 100 g proteins). The as-
lL 0.1 M KPi buffer,
l
l
say mixture was incubated for 30 min at 30° C. Controls were run
either without hydrogen peroxide, without HG, without enzyme
preparation, or with boiled (10 min) enzyme preparation. The reac-
tion products and substrate were analyzed as described in the
above paragraph.
2.4. Concluding remarks
These experiments established that a dirigent protein controls
the stereoselective coupling of hemigossypol (1) free radicals gen-
erated by cotton peroxidases for the preferential formation of (+)-
gossypol (2a) in moco cotton. We have also partially purified and
characterized the dirigent protein. Our current knowledge con-
cerning the properties of the dirigent protein will enable us to de-
sign new approaches to purify this protein to homogeneity, obtain
partial protein sequence information, and clone the encoding gene.
3.2.4. Gossypol formation from hemigossypol by horseradish
peroxidase, C. versicolor laccase, and ammonium persulfate
Dimerization of hemigossypol (1) with exogenous peroxidase,
laccase or ammonium persulfate were carried out with horseradish
peroxidase (0.13 U, 100 ng) as in the ‘‘peroxidase assay” without
enzyme preparations, with C. versicolor laccase in the ‘‘dirigent
assay” without enzyme preparations, and with ammonium persul-
fate (100 mM) in the ‘‘oxidase assay” without enzyme prepara-
tions, respectively.
3. Experimental
3.1. Chemicals and general experimental procedures
3.3. Protein extraction and purification
Hemigossypol (1) and hemigossypol 6-methyl ether were ob-
tained from cotton stele extracts by the procedure of Bell (1967).
SDS–PAGE of the protein preparations was carried out according
All extractions and purifications were conducted at 4° C. Storage
of protein preparations was at ꢁ20° C with 10% glycerol added.

3042
J. Liu et al. / Phytochemistry 69 (2008) 3038–3042
Buffer exchange and concentration were carried out either with
dialysis tubes (10 kDa cut-off) or Centricon-10 concentrators
(Amicon).
the same buffer at a flow rate of 1 mL/min. Fractions of 1 mL were
collected and assayed for dirigent activity. A major dirigent activity
eluted at a retention time of 116 min and a minor dirigent activity
eluted at 108 min; these retention times were used to determine
the molecular mass of the dirigent protein.
3.3.1. Protein extraction
The frozen tissue was ground to a fine powder in liquid N2 in a
mortar. Cold acetone (ꢁ20° C) was added to the ground powder
(5 mL/g tissue) and the mixture was stirred for 10 min. After cen-
trifugation at 40,000g for 10 min, the supernatant was decanted
and discarded. The pellet was suspended in the cold acetone and
the above procedure was repeated three more times. The resulting
pale pellet was suspended in 0.1 M KPi buffer, pH 5.5 containing
1.5 M NaCl (5 mL/g tissue) and stirred for 1 h. The mixture was
centrifuged at 12,000g for 10 min. The supernatant was buffer ex-
changed into 10 mM KPi buffer, pH 5.5, stored at ꢁ20° C after addi-
tion of 10% glycerol, and used as a crude enzyme preparation. The
crude enzyme preparation was very stable under these storage
conditions and no appreciable loss of dirigent activity and peroxi-
dase activity was observed up to 1 year of storage.
Acknowledgements
We thank Ms. Lacy Maberry and Ms. Chelsi Davis for excellent
technical assistance.
References
Bell, A.A., 1967. Formation of gossypol in infected or chemically irritated tissue of
Gossypium spp.. Phytopathology 57, 759–764.
Benedict, C.R., Liu, J., Stipanovic, R.D., 2006. The peroxidative coupling of
hemigossypol to gossypol in cottonseed extracts. Phytochemistry 67, 356–361.
Bradford, N.M., 1977.
A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein–dye
binding. Anal. Biochem. 72, 248–254.
Cass, Q.B., Tiritan, E., Matlin, S.A., Freire, E.C., 1991. Gossypol enantiomer ratios in
cotton seeds. Phytochemistry 30, 2655–2657.
Davin, L.B., Lewis, N.G., 2000. Update on lignins and lignans. Dirigent proteins and
dirigent sites explain the mystery of specificity of radical precursor coupling in
lignan and lignin biosynthesis. Plant Physiol. 123, 453–462.
Davin, L.B., Wang, H.B., Crowell, A.L., Bedgar, D.L., Martin, D.M., Sarkanen, S., Lewis,
N.G., 1997. Stereoselective bimolecular phenoxy radical coupling by an
auxiliary (dirigent) protein without an active center. Science 275, 362–366.
Gang, D.R., Costa, M.A., Fujita, M., Dinkova-Kostova, A.T., Wang, H-B., Burlat, V.,
Martin, W., Sarkanen, S., Davin, L.B., Lewis, N.G., 1999. Regiochemical control of
monolignol radical coupling: a new paradigm for lignin biosynthesis. Chem.
Biol. 6, 143–151.
Hall, S.C., Lewis, N.G., 2002. Secondary and quaternary structures of the (+)-
pinoresinol-forming dirigent protein. Biochemistry 41, 9455–9461.
Jaroszewski, J.W., Strøm-Hansen, T., Hanson, L.L., 1992. Optical stability of gossypol.
Chirality 4, 216–221.
Kim, H.L., Calhoun, M.C., Stipanovic, R.D., 1996. Accumulation of gossypol
enantiomers in ovine tissues. Comp. Biochem. Physiol. 113B, 417–420.
Laemmli, U.K., 1970. Cleavage of structural proteins during assembly of the head of
bacteriophage T4. Nature 227, 680–685.
3.3.2. Purification on CM-Sepharose FF column
The crude enzyme preparation (from 40 g flower tissue) was
buffer exchanged into 40 mM Mes buffer, pH 5.0 and loaded on
to a 1.5 ꢂ 18.0-cm CM-Sepharose FF column which had been equil-
ibrated with 40 mM Mes buffer, pH 5.0. The column was eluted
first with 25 mL 40 mM Mes buffer, pH 5.0, followed by a linear
gradient of 75 mL 40 mM Mes buffer, pH 5.0 and 75 mL 40 mM
Mes buffer, pH 5.0 containing 1 M NaCl at a flow rate of 1 mL/
min. Five-milliliter fractions were collected and assayed for ‘‘diri-
gent”, ‘‘peroxidase”, ‘‘oxidase” activities, and protein content.
3.3.3. Purification on Q-Sepharose FF column
The crude enzyme preparation (from 40 g flower tissue) was
also fractionated on a 1.5 ꢂ 18.0-cm Q-Sepharose FF column after
buffer exchanging into 10 mM KPi, pH 8.0 buffer. The fractionation
was identical as with CM-Sepharose FF column purification except
the 40 mM Mes, pH 5.0 component was replaced by 10 mM KPi, pH
8.0.
Puckhaber, L.S., Dowd, M.K., Stipanovic, R.D., Howell, C.R., 2002. Toxicity of (+)- and
(ꢁ)-gossypol to the plant pathogen, Rhizoctonia solani. J. Agric. Food Chem. 50,
7017–7021.
Ralph, S.G., Jancsik, S., Bohlmann, J., 2007. Dirigent proteins in conifer defense. II:
extended gene discovery, phylogeny, and constitutive and stress-induced gene
expression in spruce (Picea spp.). Phytochemistry 68, 1975–1991.
Stipanovic, R.D., Puckhaber, L.S., Bell, A.A., Percival, A.E., 2005. Occurrence of (+)-
and (ꢁ)-gossypol in seed of wild species of cotton and in Gossypium hirsutum
var marie-galante (Watt) Hutchinson. J. Agric. Food Chem. 53, 6266–6271.
Stipanovic, R.D., Lopez Jr., J.D., Dowd, M.K., Puckhaber, L.S., Duke, S.E., 2006. Effect of
racemic, and (+)- and (ꢁ)-gossypol on the survival and development of
Helicoverpa zea larvae. J. Chem. Ecol. 32, 959–968.
3.4. Molecular mass determination of the native dirigent protein
The dirigent active fractions from the Q-Sepharose FF column
were pooled, buffer exchanged, and concentrated to 0.5 mL in
10 mM KPi, pH 5.5 containing 0.2 M NaCl. This sample was loaded
onto a 190 mL Ultrogel AcA34 gel filtration column previously cal-
ibrated with six known molecular mass standards and eluted with
Zhou, R.H., Lin, X.D., 1988. Isolation of (ꢁ)-gossypol from natural plant.
Contraception 37, 239–245.
Zhu, L., Zhang, X., Tu, L., Zeng, F., Nie, Y., Guo, X., 2007. Isolation and characterization
of two novel dirigent-like genes highly induced in cotton (Gossypium barbadense
and G. hirsutum) after infection by Verticillium dahliae. J. Plant Pathol. 89, 41–45.