Gossypolhemiquinone, a dimeric sesquiterpenoid identified in cotton (Gossypium)

Article abstract of DOI:10.1016/j.phytochem.2015.12.009

The report that the cotton leaf perforator, Bucculatrix thurberiella, is one of the few insect herbivores to attack Gossypium thurberi prompted an investigation of the terpenoids present in the leaves of this wild species of cotton. Members of Gossypium produce subepidermal pigment glands in their leaves that contain the dimeric sesquiterpenoid gossypol as well as other biosynthetically related terpenoids. In addition to gossypol, a previously unknown dimeric sesquiterpenoid, gossypolhemiquinone (GHQ), was identified in trace amounts in G. thurberi, a member of the D genome. Other members of the D genome of Gossypium were subsequently found to contain this compound, but GHQ was not detected in commercial cotton cultivars. When fed to Helicoverpa zea in an artificial diet, GHQ delayed days-to-pupation, reduced pupal weights, and survival to adulthood to a lesser or equal extent than gossypol in comparison to the control diet. However, GHQ had a synergistic effect on survival and days-to-pupation when combined with gossypol at the highest dosage tested (0.18%; 15.5:84.5 GHQ:gossypol). Because gossypol exhibits anti-cancer activity, GHQ was also evaluated for its anti-cancer activity against the National Cancer Institute's 60-Human Tumor Cell Line Screen. Significant inhibitory activity against most of these cell lines was not observed, but the results may offer some promise against leukemia cancer cell lines.

Full text of DOI:10.1016/j.phytochem.2015.12.009

Phytochemistry 122 (2016) 165–171
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Gossypolhemiquinone, a dimeric sesquiterpenoid identified
in cotton (Gossypium)
a
a
a
a
Robert Stipanovic a, , Lorraine Puckhaber , James Frelichowski Jr. , Jesus Esquivel , John Westbrook ,
⇑
Mike O’Neil a, Alois Bell a, Michael Dowd b, Kater Hake c, Sara Duke a
a USDA, Agricultural Research Service, Southern Plains Agricultural Research Center, College Station, TX 77845, United States
b USDA, Agricultural Research Service, Southern Regional Research Center, New Orleans, LA 70124, United States
c Cotton Incorporated, 6399 Weston Parkway, Cary, NC 27513, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The report that the cotton leaf perforator, Bucculatrix thurberiella, is one of the few insect herbivores to
attack Gossypium thurberi prompted an investigation of the terpenoids present in the leaves of this wild
species of cotton. Members of Gossypium produce subepidermal pigment glands in their leaves that con-
tain the dimeric sesquiterpenoid gossypol as well as other biosynthetically related terpenoids. In addition
to gossypol, a previously unknown dimeric sesquiterpenoid, gossypolhemiquinone (GHQ), was identified
in trace amounts in G. thurberi, a member of the D genome. Other members of the D genome of Gossypium
were subsequently found to contain this compound, but GHQ was not detected in commercial cotton cul-
tivars. When fed to Helicoverpa zea in an artificial diet, GHQ delayed days-to-pupation, reduced pupal
weights, and survival to adulthood to a lesser or equal extent than gossypol in comparison to the control
diet. However, GHQ had a synergistic effect on survival and days-to-pupation when combined with
gossypol at the highest dosage tested (0.18%; 15.5:84.5 GHQ:gossypol). Because gossypol exhibits anti-
cancer activity, GHQ was also evaluated for its anti-cancer activity against the National Cancer
Institute’s 60-Human Tumor Cell Line Screen. Significant inhibitory activity against most of these cell
lines was not observed, but the results may offer some promise against leukemia cancer cell lines.
Published by Elsevier Ltd.
Received 14 September 2015
Received in revised form 4 December 2015
Accepted 14 December 2015
Available online 22 December 2015
Keywords:
Gossypium thurberi
Malvaceae
Insect herbivory resistance
Di-sesquiterpenoid
Helicoverpa zea
Gossypol
Gossypolhemiquinone
1. Introduction
latter is exceptional in that it is the only species that produces the
terpenoid raimondal (8) (Fig. 1). One species, Gossypium thurberi,
Members of the genus Gossypium produce pigment glands in the
foliage that contain a mixture of terpenoid aldehydes such as
gossypol (1) (Fig. 1). Gossypol (1) and biosynthetically related
compounds such as heliocides H1–H4 (3–6) and hemigossypolone
(7) (Fig. 1) suppress the growth of some herbivorous insect pests
(Lukefahr and Martin, 1966; Stipanovic et al., 1977). Based on tax-
onomy, as well as cytological and cross pollination studies, cotton
breeders have conveniently given alphanumeric designations to
Gossypium species where those sharing the first letter have some
degree of interfertility (Percival et al., 1999). An early TLC study of
foliar extracts from Gossypium species found that most members
of the D genome produce only gossypol (1) in the foliage pigment
glands (Stipanovic et al., 1977). Most members of the D genome
inhabit mainly dry regions of the western portions of Mexico, with
one species found in the Galapagos (Gossypium klotzschianum
Andersson) and one in Peru (Gossypium raimondii Ulbrich); the
extends north as far as central Arizona. It is in Arizona that Karban
studied G. thurberi Todaro and found that the cotton leaf perforator
(Bucculatrix thurberiella Busek) is one of the few insect herbivores
that feed on this plant (Karban, 1993). This report prompted the
current investigation of the terpenoid aldehydes present in the
leaves of this plant which resulted in the identification of a
previously unknown minor component in addition to gossypol
(1). The structure of this compound, which we call gossypol-
hemiquinone (9, GHQ) (Fig. 1), was established based on synthesis
and subsequent detailed 1-D and 2-D NMR analyses, and high res-
olution MS. Phillips and Hedin (1990) had previously identified
the related compound gossypolone (2, GQ) (Fig. 1) as a component
of the pigment glands in flower buds in some cotton cultivars.
In the current study, Helicoverpa zea (Boddie) was selected as a
test-insect to determine if GHQ (9) might augment resistance to
insect herbivores. Its effect on the growth and development of this
insect was determined by adding 0.06%, 0.12% and 0.18% GHQ (9)
to artificial diets. Mortality, days to pupation, and pupal weight
were determined for larvae fed the GHQ (9) diets and compared
⇑
Corresponding author.
E-mail address: bob.stipanovic@ars.usda.gov (R. Stipanovic).
http://dx.doi.org/10.1016/j.phytochem.2015.12.009
0031-9422/Published by Elsevier Ltd.

166
R. Stipanovic et al. / Phytochemistry 122 (2016) 165–171
O
O
O
O
CH OH
H
C
CH
O
OH
O
O
C
H
HO
HO
HO
HO
OH
OH
OH
OH
O
2 Gossypolone
1 Gossypol
O
O
CH
O
CH
O
R1
R4
HO
HO
R1
R2
HO
HO
R2
R3
O
O
5 Heliocide H2
R2=H; R1= H2C
3 Heliocide H1
R , R =H; R =CH ; R =
H2C
1
3
4
2
3
4 Heliocide H4
6 Heliocide H3
R1, R2=H; R3=CH3; R4=
H C
H2C
R1=H; R2=
2
O
O
O
OH
CH OH
CH
O
HC
OH
HO
HO
OCH3
HO
HO
O
HO
CH HO
HO
HO
O
8 Raimondal
7 Hemigossypolone
O
7
O
14
26
1a (+)-Gossypol
CH
OH
O
HC
24
HO
HO
22 OH
2
17
9
1
4
16
19
23
20
8
5
3
18
27
6
21
OH
10
25
15
28
O
11
12
29
13
30
9 Gossypolhemiquinone
Fig. 1. Terpenoids identified in Gossypium foliage.
to that from larvae raised on a diet containing gossypol (1) at the
same concentrations or a control diet containing no terpenoids.
Diets with a mixture of GHQ (9) and gossypol (1) (15.5:84.5) at
the same concentrations as GHQ (9) were also studied.
group ortho to the aldehyde have a propensity to lose a molecule of
water during mass spectrometry. A small peak (0.4%) was observed
at m/z 533, but a substantially larger peak was observed at m/z 555
(10%). The latter peak was suspected to be due to [M] (532 amu)
+ Na (23 amu), while the m/z 533 peak was tagged as the parent
[M]+1 peak; this corresponds to the [M]+1 molecular weight of
several gossypol-like terpenoids including gossypolhemiquinone
(GHQ) (9). To further characterize this metabolite, the compound
was purified by semi-prep HPLC. The purified metabolite was sub-
ject to EI/MS analysis using a direct exposure probe. When only
masses above 400 amu were considered, the compound provided
a base peak at m/z 514, a fragment ion at m/z 499 (62.9%) and an
apparent molecular ion at m/z 532 (2.2%). This molecular weight
agreed with the ready loss of H2O to provide the base peak and a
subsequent loss of a CH3 group. High resolution ESI/MS provided
an ion at m/z 533.1814 with a molecular formula of C30H29O9
(Calcd. C30H29O9: 533.181158, accuracy 0.45 ppm).
Gossypol (1) shows a wide range of biological activity including
anticancer (Shelly et al., 2000; Oliver et al., 2005), spermicidal (Kim
et al., 1984), antiamoebic (Gonzalez-Garza and Said-Fernandez,
1988), antibacterial (Yildirim-Aksoy et al., 2004) and antifungal
(Mellon et al., 2011). Since both gossypol (1) and GQ (2) show anti-
cancer activity, GHQ (9) was tested for its anticancer activity in the
National Cancer Institute’s 60-Human Tumor Cell Line Screen.
Gossypol (1) exists as stable enantiomers at room temp. due to
restricted rotation around the central binaphthyl bond
(Jaroszewski et al., 1992). The anti-cancer activity of (+)- and
(À)-gossypol (1) differ, with (À)-gossypol being more active
(Loberg et al., 2007). Since the enantiomers of gossypol (1) are
stable at room temperature, but those of GQ (2) racemize at room
temp. (Dao et al., 2004), the optical stability of (+)-GHQ was
investigated.
2.2. Synthesis of ( )-gossypolhemiquinone (GHQ) (9) and (+)-
gossypolhemiquinone (2)
2. Results and discussion
To establish the identity of the unknown, GHQ (9) was synthe-
sized [along with GQ (2)] using a modification of the FeCl3 oxida-
tion of gossypol (1) developed by Hass and Shirley (1965). The
final synthetic product was >97% pure (HPLC) [m.p. 158–160 °C;
2.1. Identification of gossypolhemiquinone (GHQ) (9)
Extracts from the young leaves of G. thurberi showed the pres-
ence of gossypol (1) and an unidentified compound. The extract
was subjected to LC/MS analysis and the unidentified compound
provided a base peak of m/z 515. Gossypol-type compounds with
a hydroxyl group peri to the aldehyde group and a second hydroxyl
UV: (95% EtOH) kmax (log e) 234, (3.81), 274 (3.57), 372 (3.08)]
and the structure was established to be GHQ (9) based on 1-D
and 2-D NMR analyses [chemical shifts are reported in Tables 1
and 2 (HMBC proton–carbon couplings are illustrated in the online

R. Stipanovic et al. / Phytochemistry 122 (2016) 165–171
167
Table 1
using TLC plates, only gossypol (1) was identified in this tissue
(Stipanovic et al., 1977). Thus, terpenoids such as the sesquiter-
penoid hemigossypolone (7) and the sesterterpenoids heliocides
H1, H2, H3 and H4 (3–6) (Fig. 1) that are found in commercial
Upland cottons (Gossypium hirsutum), as well as other members
of Gossypium, were not found. To extend this investigation, several
Gossypium accessions of the D genome available in the USDA Cot-
ton Germplasm collection were grown in the greenhouse; leaves
were collected, freeze dried, and ground. The ground tissue was
extracted and immediately subjected to HPLC analysis; the results
are shown in Table 3. As found previously, other sesquiterpenoids
and sesterterpenoids normally in G. hirsutum were not detected in
G. thurberi nor in most other members of the D genome. However,
GHQ (9) was present in small amounts in these cottons. As previ-
ously reported (Stipanovic et al., 1977), within the D genome only
the leaves of Gossypium gossypioides (Ulbrich) contained
hemigossypolone (7) and heliocides (3–6); the leaves of other
members of the D genome did not contain these terpenoids. In
addition, the leaves of G. raimondii contained only gossypol (1)
and the sesquiterpenoid, raimondal (8) (Fig. 1). Four commercial
cultivars of G. hirsutum cotton also were analyzed (Table 4). GHQ
(9) was not detected in these plants, but, as expected, gossypol
(1), hemigossypolone (7) and heliocides H1–H4 (3–6) were
detected.
13C chemical shifts for gossypolhemiquinone (GHQ) (9) compared to the dimeric
sesquiterpenoids gossypol (Goss) (1) and gossypolone (GQ) (2) (for specific carbon
assignments refer to Fig. 1).
C#
Goss (1)
GHQ (9)
C#
GQ (2)
GHQ (9)
d
d
d
d
1
2
3
4
5
6
7
8
150.4
115.8
133.7
118.1
134.1
143.4
156.0
111.7
114.6
129.6
27.8
150.2
115.6
131.6
117.9
133.8
143.4
156.0
111.8
117.3
129.4
27.8
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
186.5
147.2b
138.1b
184.4
141.5
149.4
152.4
115.9
126.7
127.6
198.5
14.8
186.7a
149.0
140.3
186.8a
141.4
149.4
152.3
115.6
127.2
127.5
197.9
14.6
9
10
11
12
13
14
15
20.2
20.2
199.3
20.3
20.1
20.3
199.3
20.8
28.7
19.7
19.6
28.8
20.0
19.7
a
Assignments may be interchanged.
Assignments are based on that in GHQ; in the HMBC experiment, no coupling
b
was shown between the protons on C-27 with either C-17 or C-18.
Table 2
1H chemical shifts for gossypolhemiquinone (GHQ) (9) compared to the dimeric
sesquiterpenoids gossypol (Goss) (1) and gossypolone (GQ) (2) (for carbon assign-
ments refer to Fig. 1).
2.4. Effect of GHQ (9) on growth and development of H. zea larvae
C#
Type
Goss (1)
d
GHQ (9)
d
C#
Type
GQ (2)
d
GHQ (9)
d
Since gossypol (1) and related terpenoids are important compo-
nents in the plant’s arsenal of terpenoids that protect it from her-
bivorous insects, it seemed appropriate to test the toxicity of GHQ
(9) to insects. Thus, racemic gossypol (1) and racemic GHQ (9)
were individually incorporated into a soybean artificial diet using
the non-nutritive additive, Alphacel, at concentrations of 0.00%
(control), 0.06%, 0.12% and 0.18%. To determine if GHQ (9) has a
synergistic effect with gossypol (1), diets were also prepared with
a total concentration of terpenoids of 0.06%, 0.12% and 0.18%, and a
15.5:84.5 ratio of GHQ (9):gossypol (1). One-day-old H. zea larvae
were placed in plastic cups containing the various diets described
above. After pupation, the pupae were allowed to harden for one
day and then weighed. Days-to-pupation, pupal weights and per-
cent survival to adulthood were recorded.
The mean number days-to-pupation increased as the concentra-
tion of gossypol (1) and GHQ (9) increased compared to larvae fed
the control diet (Fig. 2). However, the difference in days-to-pupa-
tion between the larvae fed the gossypol (1) and GHQ (9) were
not significantly different. When the larvae were fed the 15.5:84.5
mixtures of GHQ (9) and gossypol (1), an extension in
days-to-pupation was not observed at the 0.06% and 0.12% concen-
trations. However, at the 0.18% concentration, the days-to-pupation
were significantly longer compared to all other diets.
The mean pupal weights (Fig. 3) were significantly less for diets
containing 0.12% and 0.18% gossypol (1), GHQ (9), or the 15.5:84.5
mixture of the two terpenoids compared to those fed the control
diet. However, pupal weights for larvae fed gossypol (1) were sig-
nificantly less than those fed an equivalent concentration of GHQ
(9). No synergistic effect in reducing pupal weight was observed
with the mixture.
The percent survival to adulthood of larvae fed 0.06% gossypol
(1) or the 0.06% 15.5:84.5 GHQ (9):gossypol (1) mixture was
greater than the control (Fig. 4). This may be due to a hormetic
effect (i.e., a biphasic dose–response exhibited by some toxic sub-
stances when fed at low concentrations) that has been demon-
strated in other gossypol (1) feeding studies (Stipanovic et al.,
1986; Celorio-Mancera et al., 2011). Survival of larvae fed 0.06%
or 0.12% GHQ (9) or the 15.5:84.5 mixtures were not affected. At
0.18%, gossypol (1) and GHQ (9) reduced survival with gossypol
1
4
6
7
11
12
13
14
15
C1-OH
Ar-H
C6-OH
C7-OH
CHMe2
CH3
CH3
HC@O
Ar-CH3
5.89
7.77
6.39
15.10
3.88
1.53
1.53
11.1
2.13
6.60a
7.65
21
22
26
27
28
29
30
C21-OH
C22-OH
HC@O
Ar-CH3
CHMe2
CH3
6.59
13.01
10.58
2.04
4.12
1.42
6.39a
12.97
10.50
1.97
4.05
1.39
6.37a
15.05
3.82
1.48
1.53
10.90
2.19
CH3
1.44
1.42
a
Assignments for GHQ (9) C1-OH, C6-OH and C21-OH may be interchanged.
Supplementary material section)] and direct probe EI/MS. The
identity of the unknown from G. thurberi was then confirmed to
be GHQ (9) based on HPLC retention times [10.83 min for GHQ
(9) compared to 9.54 min for GQ (2) and 15.66 min for gossypol
(1)] and UV spectra (shown in Supplementary material section)
that matched the synthesized compound.
(+)-GHQ (9) was prepared from (+)-gossypol (1a) (99% ee). The
synthesized (+)-GHQ (9) had a specific rotation of [
20
a]
+36.4° (c
D
0.278, EtOH) and was established to be 82% enantiomeric excess
via HPLC analysis using a Phenomenex Lux Cellulose column and
an iPrOH:hexane mobile phase. HPLC analysis showed that in a
solution of iPrOH:hexane (ꢀ1:1), the GHQ (9) slowly racemized
at room temperature. Over 5 days, solutions that were 65% ee were
reduced to 12% ee. The ee of an aliquot of this solution held at
À20 °C did not significantly racemize after 5 days. The solid sam-
ples also could be stored at À20 °C without detectable racemiza-
tion. Others have shown that GQ (2) readily undergoes
racemization at room temperature (Dao et al., 2004). Thus, GHQ
(9) appears to be intermediate between gossypol (1) and GQ (2)
in its propensity to racemize.
2.3. GHQ (9) and gossypol (1) in the D genome of Gossypium and in
commercial cottons
G. thurberi is a member of the D genome of Gossypium. In a pre-
vious investigation of terpenoids in the young leaves of G. thurberi

168
R. Stipanovic et al. / Phytochemistry 122 (2016) 165–171
Table 3
Concentration of gossypol (Goss) (1) and gossypolhemiquinone (GHQ) (9) in freeze dried leaves of select D genome Gossypium greenhouse grown plants.
Gossypium species
(Accession #)
Genome
#
GHQ (9)
g/mg
Goss (1)
g/mg
Gossypium species
Genome
#
GHQ (9)
g/mg
Goss (1)
g/mg
l
l
l
l
G. thurberi (#6)
G. thurberi (#21)
G. thurberi (#22)
D1
0.17
0.12
0.13
0.18
0.12
0.15
0.10
0.15
0.10
0.10
0.11
0.12
0.13
0.10
0.13
0.17
0.15
10.12
13.46
19.94
9.34
5.72
12.87
1.97
6.34
3.61
2.16
4.14
2.86
9.56
1.87
8.50
12.02
7.19
G. raimondii (#4)a
D5
ndb
nd
nd
nd
nd
0.76
0.73
0.71
0.52
0.24
0.28
0.24
6.59
22.39
15.33
23.52
14.73
23.81
7.72
9.35
6.10
5.24
23.92
c
G. raimondii (#20)
G. raimondii (#26)d
G. gossypioides (#1)e
G. gossypioides (#3)f
G. gossypioides (#7)g
G. gossypioides (#8)h
G. lobatum (#4)
G. trilobum (#1)
G. trilobum (#4)
G. trilobum (#5)
G. trilobum (#8)
G. trilobum (#9)
G. laxum (#4)
G. armourianum (#1–6)
G. armourianum (#1–7)
G. harknessii (#2–2)
G. davidsonii (#d-2)
G. davidsonii (#d-3)
G. davidsonii (#d-4)
G. davidsonii (#d-21)
G. davidsonii (#d-23)
G. davidsonii (#d-26)
G. davidsonii (#d-30)
G. davidsonii (#d-32)
G. klotzschianum (#k-57)
G. aridum (#12)
D2
D6
D2–2
D3
nd
nd
D7
D8
0.11
0.25
0.14
0.33
0.18
0.18
0.15
0.10
0.12
0.13
0.13
D9
D10
D3
D4
G. turneri (#2)
G. turneri (#7)
G. turneri (#8)
G. schwendimanii (#1)
G. aridum (#13)
D11
a
b
c
Raimondal (8) 8.72
nd = not detected (limit of detection 0.05
Raimondal (8) 5.20
Raimondal (8) 5.82
Hemigossypolone (HGQ) (7) 6.46
lg/mg.
l
g/mg).
l
g/mg.
g/mg.
d
e
f
l
l
g/mg, H1 + H3 (3, 6) 10.50
lg/mg, H2 (5) 5.43 lg/mg, H4 (4) 5.76 lg/mg; HPLC method did not separate heliocide H1 (3) from H3 (6).
HGQ (7) 0.34
HGQ (7) 2.28
HGQ (7) 0.41
lg/mg, H1 + H3 (3, 6) 5.52
lg/mg, H1 + H3 (3, 6) 7.24
lg/mg, H1 + H3 (3, 6) 7.06
l
l
l
g/mg, H2 (5) 4.10
g/mg, H2 (5) 4.27
g/mg, H2 (5) 3.19
l
l
l
g/mg, H4 (4) 2.69
g/mg, H4 (4) 3.90
g/mg, H4 (4) 3.93
lg/mg.
lg/mg.
lg/mg.
g
h
Table 4
400
300
200
100
0
Concentration (lg/mg) of gossypol (Goss) (1), hemigossypolone (HGQ) (7), heliocides
H1–4 (3–6) and gossypolhemiquinone (GHQ) (9) in freeze dried leaves of commercial
cultivars (the HPLC method did not separate heliocide H1 (3) from heliocide H3 (6);
gossypolone (2) was not detected.
Cultivar
HGQ (7) GHQ (9) Goss (1) H1 + H3 H2 (5) H4 (4)
(3, 6)
Coker-312
FiberMax 958
FiberMax 966
1.18
0.49
1.27
nda
nd
nd
nd
0.36
0.13
0.27
1.92
0.44
0.25
0.43
0.60
1.10
0.40
0.44
0.76
0.06
0.07
0.16
0.17
Stoneville 474 0.56
a
nd = not detected (limit of detection 0.05 lg/mg).
Fig. 3. Mean (with standard error bars) pupal weights for Helicoverpa zea one-day-
old larvae fed various concentrations of gossypol (1) or gossypolhemiquinone
(GHQ) (9) or a 15.5:84.5 mixture of GHQ (9):gossypol (1) in an artificial diet.
40
30
20
10
0
94.3%
94.3%
100%
80%
60%
40%
20%
0%
91.4%
88.6%
88.6%
85.7%
82.9%
57.1%
40.0%
28.6%
Fig. 2. Mean number (with standard error bars) of days-to-pupation for Helicoverpa
zea one-day-old larvae fed various concentrations of gossypol (1) or gossypol-
hemiquinone (GHQ) (9) or a 15.5:84.5 mixture of GHQ (9):gossypol (1) in an
artificial diet.
Fig. 4. Percent survival for Helicoverpa zea larvae fed various concentrations of
gossypol (1), gossypolhemiquinone (GHQ) (9) or
(9):gossypol (1) in an artificial diet.
a 15.5:84.5 mixture of GHQ
2.5. Effect of GHQ (9) on cancer cell lines
(1) showing mortality >50% compared to larvae fed the control
diet. However, the survival to adulthood of larvae fed the 0.18%
15.5:84.5 GHQ (9):gossypol (1) diet was less than one-third those
fed the control diet indicating a potential synergistic effect.
Both gossypol (1) and GQ (2) have been extensively tested for
anti-cancer activity. (À)-Gossypol (termed AT-101) is currently in

R. Stipanovic et al. / Phytochemistry 122 (2016) 165–171
169
clinical trials as an anticancer agent (Oliver et al., 2005; Ready
et al., 2011). GQ (2) shows some anticancer activity but appears
to be less active than gossypol (1) (Gilbert et al., 1995). Since
GHQ (9) possesses structural features of both molecules, it seemed
reasonable to investigate its anticancer activity. The compound
was submitted to the National Cancer Institute 60-Human Tumor
Cell Line Screen. Experimental details for the Tumor Cell Line
Screen can be found online (http://dtp.nci.nih.gov/branches/btb/iv-
clsp.html). Results are shown in Supplementary section. No potent
activity was noted against slow-growing tumors; however, the
results offer some promise against leukemia cancer cell lines. Addi-
tional screening for apoptosis, angiogenesis, cell invasion and
migration may be indicated.
Varian 500-MS Ion Trap Mass Spectrometer (Walnut Creek, CA).
That is, samples were injected on the Agilent instrument then a
splitter sent 81% of the column eluent to the DAD detector and
19% to the Varian MS. The MS was operated in the positive polarity
full scan electrospray ionization (ESI) mode with spray chamber
50 °C, nebulizer gas N2, pressure 35 psi, drying gas 350 °C, drying
gas pressure 10 psi, needle voltage 4200 V, spray shield 600 V, cap-
illary 80 V, and RF loading 100%. The product ion start mass was
100 m/z and end mass was 600 m/z.
Direct exposure probe MS were acquired on a Thermo Electron
DSQ in positive ion EI mode (70 EV, source 200 °C, scan rate
300 amu/s, scan 50–600 amu).
1H NMR, 13C NMR, 1H–1H COSY, HMBC and HSCQ spectra were
acquired on a Bruker Avance III 500 instrument (Billerica, MA, USA)
equipped with a cryoprobe operating at 500 MHz for 1H and 125
MHZ for 13C. Spectra were determined in CDCl3, which was used
as an internal standard (1H: 7.24 d; 13C: 72.0 d). One and two
dimensional 1H and 13C NMR spectra (COSY, HMBC, and HSQC)
were used to assign specific proton and carbon assignments. The
optical rotation for GHQ (9) synthesized from (+)-gossypol (1a)
was determined using a Perkin Elmer Model 241 polarimeter uti-
lizing a 1 dm temperature controlled microsample holder. UV
spectra were recorded on a Hewlett Packard ultraviolet–visible
spectrometer Model 8456.
3. Conclusion
In these artificial insect feeding studies, GHQ (9) was no more
effective or was less effective than gossypol (1) at comparable con-
centrations. However, at the highest concentration tested (0.18%),
the 15.5:84.5 GHQ (9):gossypol (1) mixture demonstrated a signif-
icant increase in days-to-pupation, as well as a reduced survival
rate compared to gossypol (1) or the control. Since GHQ (9) is pre-
sent at very low concentrations in G. thurberi, it is probably not
responsible for the plant’s resistance to herbivorous insects. How-
ever, if the concentration of GHQ (9) could be increased, then a
synergistic interaction as observed in this diet study could provide
cotton plants with enhanced resistance to herbivorous insects. This
may be an achievable goal, since the conversion of gossypol (1) to
GHQ (9) may involve a single enzyme, possibly an oxidation via a
P450 enzyme (see Supplementary material section for proposed
mechanism). A molecular marker assisted breeding program direc-
ted at a P450 gene might facilitate incorporation of GHQ (9) at levels
high enough to increase resistance to herbivorous insects in cotton.
Since GHQ (9) is not responsible for the apparent resistance
reported by Karban, then the high concentration of gossypol (1)
in the leaves may be a major contributor to this resistance. How-
ever, increasing the gossypol (1) content in leaves usually leads
to an increase in levels of gossypol (1) in the seed. For example,
in a study of 28 accessions of G. thurberi, the mean concentration
4.2. Plant material
Leaves of accessions listed in Table 3 were collected from green-
house grown plants; these accessions are part of the U.S. National
Cotton Germplasm Collection (accessible through http://www.ars-
grin.gov). Leaves from commercial cultivars (Table 4) also were
taken from greenhouse grown plants. After collection, the leaves
were frozen and then freeze dried; the freeze dried leaves were
ground to a fine powder in an agate mortar and pestle. Ground
samples were stored at 2 °C until used. Leaves from G. thurberi
plants grown in the field also contained GHQ (9).
4.3. HPLC isolation of GHQ (9) from G. thurberi
Isolation of GHQ (9) was initially achieved from a G. thurberi
extract [10 g tissue in CH3CN (10 mL)] on the Agilent HPLC instru-
ment using a Scientific Glass Engineering ProteCol-GP-C18-125
(4.6 Â 250 mm) column at 23–24 °C and an isocratic mobile phase
of 3:1 CH3CN:H2O with 0.1% HCO2H. The flow rate was 1.25 mL/
min and the run time was 20 min. The eluent was monitored at
272 10 nm (referenced to 550 50 nm) and spectra of detected
of gossypol (1) in seed ranged from 17.02
lg/mg to 3.10 lg/mg
with a mean of 13.77 g/mg (Stipanovic et al., 2005). In the current
l
study, the concentrations of gossypol (1) in the seed of G. thurberi
D1–6, D1–21 and D1–22 (greenhouse grown plants) were
21.86 lg/mg, 21.08 lg/mg, and 25.02 lg/mg, respectively. In
contrast, a study of gossypol (1) concentrations in the seed of 14
commercial and experimental G. hirsutum cottons grown at five
different locations from the Lower Rio Grande Valley to the high
peaks were stored over 210–600 nm. Multiple injections of 50 lL
were carried out and the peak of interest was manually collected
from the eluent from the DAD. The collected fractions were
extracted three times with Et2O after the addition of an equal vol-
ume of H2O. Back washing with H2O (2Â), drying over Na2SO4 and
final evaporation provided the compound of interest (>95% pure).
plains of Texas varied from 1.03 lg/mg to 0.41 lg/mg (Stipanovic
et al., 1988). Unfortunately, cottonseed with high levels of gossypol
(1) is unsuitable as a feed for cattle and dairy cows, which is the
major market for this important byproduct of cotton production.
Therefore, a phytochemical such as GHQ (9) that augments the
activity of gossypol (1) may offer an alternative strategy to increase
resistance to herbivorous insects.
4.4. Preparation of GHQ (9) and (+)-GHQ
A modification of the method described by Hass and Shirley
(1965) was used to prepare GHQ (9). Specifically, gossypol acetic
acid (400 mg) was dissolved in acetone (48 mL) and glacial AcOH
(32 mL) in a round bottom flask equipped with a stirring bar. While
the solution was stirred at room temperature, a freshly prepared
aqueous solution of 10% FeCl3Á6H2O (25 mL) was added over four
minutes. The solution was stirred for a total of 35 min. Dilute
H2SO4 (20%, 20 mL) then was added followed by H2O (25 mL).
The solution (dark black) was extracted with Et2O (1 Â 100 mL
and 3 Â 50 mL). The combined organic layer was washed 1Â with
saturated brine, dried over Na2SO4 (anhydrous) and evaporated to
4. Experimental
4.1. General
HPLC isolations, quantitative analyses and enantiomeric ratio
determinations were performed on an Agilent Technologies HPLC
instrument (Waldbronn, Germany) equipped with a 1200 solvent
degasser, 1200 quaternary pump, 1100 autosampler, and 1100
diode array detector (DAD). The LC/MS chromatography was car-
ried out with the Agilent HPLC instrument in conjunction with a

170
R. Stipanovic et al. / Phytochemistry 122 (2016) 165–171
dryness. The dried material was stored overnight at 2 °C. The pro-
duct was purified by low pressure chromatography on a dry-
packed silica gel (Mallinckrodt CC-4; 48 g) column as follows.
The crude material was dried onto silica gel (2.5 g) and placed on
the top of the column. The column was developed with a gradient
of acetone:cyclohexane beginning with 1% acetone and increasing
in 1% increments to 7% acetone (210 mL solvent per increment),
and finally to 10% acetone (200 mL). Beginning at 6% acetone, frac-
tions were collected and checked for GHQ (9) purity via HPLC as
described above in Section 4.3 except 1 lL injections were used.
Individual fractions with GHQ (9) purity >97% were allowed to
evaporate slowly in a fume hood. Crystals began to form after
about two days [m.p. 158–160 °C; UV: (95% EtOH) kmax (log e)
234 (3.81). 274 (3.57), 372 (3.08)] (spectra shown in Supplemen-
tary material section). After about 5 days, the mother liquor was
removed from tubes in which crystals formed and the crystals
were washed with cyclohexane and allowed to dry at room
temperature.
(+)-GHQ was prepared from (+)-gossypol (1a) (>98% ee) as
detailed above. It was found that (+)-GHQ slowly racemizes in
solution at room temperature over a period of days, such that after
5 days the enantiomeric excess of (+)-GHQ was <1%. Thus, the syn-
thetic (+)-GHQ was allowed to crystalize over a period of only two
days.
amounts of terpenoids were dissolved in acetone (12 mL) and
quantitatively added to Alphacel (5.04 g). Hexane (10 mL) was
added, and the suspensions were dried under vacuum at room
temperature on a rotoevaporator. Acetone (10 mL) was added, fol-
lowed by hexane (25 mL). The suspension was evaporated to dry-
ness as before. Gossypol (1) was added as the acetic acid
complex, but the final concentrations of gossypol (1) and GHQ
(9) were 0.06%, 0.12% and 0.18% of the final product. To account
for the AcOH in the gossypol (1) samples, an equivalent amount
of AcOH was added with the GHQ (9). The Alphacel samples were
stored at À20 °C and then placed under vacuum for 20 h. The
control diet was prepared as described above including all solvents
and Alphacel but without any added terpenoids. The Alphacel sam-
ples (5.04 g) were mixed with a diet-premix of Instant Soybean-
Wheat Germ Insect Diet, Stonefly Industries, Inc., Bryan, TX
(42 g). Next, a dilute vinegar solution [H2O and 5% vinegar (Albert-
son’s, Boise, ID; white vinegar 0.2% volume) without 0.1% formalin]
was prepared and added to the Alphacel and diet-premix mixtures
to yield a total wet weight of 168 g per diet.
H. zea larvae were obtained from laboratory reared eggs
(Mississippi State University). Insects were maintained at a pho-
toperiod of 14:10 light:dark hours at 27 °C (relative humidity
50–70%) in an insect-rearing room at the College Station, TX labo-
ratory. A one-day-old bollworm larva was placed in a 22-mL plastic
cup with 4–5 g of diet. Thirty larvae were used per treatment. Ten
days after initiation of the experiment, the cups were inspected
daily. After pupation, pupae were allowed to harden for 1 day,
weighed, and returned to their respective cups; pupae were
inspected daily until adult emergence. Days-to-pupation, pupal
weight, and number of survivors reaching adulthood were
recorded.
4.5. HPLC determination of the GHQ (9) enantiomeric ratio
The enantiomeric ratio of the isolated and the synthesized
GHQ (9) was determined via HPLC using the Agilent Technologies
instrument
with
a
Phenomenex
Lux-Cellulose-4–5 lm
(4.6 Â 150 mm) column at 23–24 °C and an isocratic mobile phase
of 4:1 hexane:iPrOH with 0.1% CF3CO2H. The flowrate was 1.5 mL/
min and the run time was 10 min. The eluent was monitored at
254 10 nm (referenced to 550 50 nm) and spectra were stored
over 240–600 nm. Under these conditions, (+)-GHQ (9) elutes at
2.2 min and (À)-GHQ elutes at 3.8 min.
4.8. Anti-cancer activity
GHQ (9) was submitted to the National Cancer Institute’s Devel-
opmental Therapeutics Program (DPI) in the small molecule In
Vitro Cell Line Screening Project. This assay utilizes 60 different
human tumor cell lines representing leukemia, melanoma and can-
cers of the lung, colon, brain, ovary, breast, prostate, and kidney.
The object is to identify compounds that exhibit selective growth
inhibition or cell killing of selected tumor cell lines.
4.6. Extraction and HPLC quantitation of terpenoids in D genome leaf
tissue
Freeze dried leaf powder was accurately weighed
(100 mg 2 mg) into
a test tube and CH3CN:H2O with 0.1%
H3PO4 was added (10 ml, 4:1, v/v). The tube was sonicated for
5 min, vortexed momentarily and then centrifuged for 5 min at
3000 rpm. A portion of the resulting clear supernatant was trans-
ferred to a vial for quantitative HPLC analysis. For each sample,
the total time from the beginning of the extraction until HPLC
injection was <15 min. The HPLC analysis was performed using
the same method as that for the isolation of GHQ (9) (Section 4.2)
Acknowledgements
We thank Howard Williams of Texas A&M University for pro-
viding NMR spectra, and Stephanie Sullivan and Christine Soriaga
for excellent technical assistance. We thank Cotton Inc. for partial
support for this work. Kathy Knighten of Mississippi State Univer-
sity provided H. zea eggs from which larvae were reared for the
insect feeding study.
except that a 20
(3–6) were present, the run time was 30 min instead of 20 min.
Standard g versus peak area curves obtained from pure ter-
lL injection was used; when the H heliocides
l
penoids were employed to calculate the concentration of the com-
pounds in the leaf tissue from the peak areas in the chromatogram.
Appendix A. Supplementary data
The limit of detection of GHQ (9) was 0.05 lg/mg. Each cotton line
was analyzed once.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.phytochem.2015.
12.009.
4.7. Insect feeding studies
References
Insect feeding studies were conducted as previously described
(Stipanovic et al., 2014). Thus, Alphacel (ICN Nutritional Biochem-
icals, Cleveland, OH) a non-nutritive polysaccharide was used as a
carrier in all diets. The control diet contained Alphacel alone, while
terpenoid diets contained three different concentrations of either
gossypol (1), GHQ (9) or a 15.5:84.5 ratio of GHQ (9):gossypol (1)
mixed with Alphacel. The diets were prepared as follows. Weighed
Celorio-Mancera, M.P., Ahn, S.J., Vogel, H., Heckel, D.G., 2011. Transcriptional
responses underlying the hormetic and detrimental effects of the plant
secondary metabolite gossypol on the generalist herbivore Helicoverpa
armigera. BMC Genomics 12, 575–590.
Dao, V.-T., Dowd, M.K., Martin, M.-T., Gaspard, C., Mayer, M., Michelot, R.J., 2004.
Cytotoxicity of enantiomers of gossypol Schiff’s bases and optical stability of
gossypolone. Eur. J. Med. Chem. 39, 619–624.

R. Stipanovic et al. / Phytochemistry 122 (2016) 165–171
171
Gilbert, N.E., O’Reilly, J.E., Chang, C.J.G., Lin, Y.C., Brueggemeier, R.W., 1995.
Antiproliferative activity of gossypol and gossypolone on human breast
cancer cells. Life Sci. 57, 61–67.
Gonzalez-Garza, M.T., Said-Fernandez, S., 1988. Entamoeba histolytica: potent
in vitro antiamoebic effect of gossypol. Exp. Parasitol. 66, 253–255.
Hass, R.A., Shirley, D.A., 1965. The oxidation of gossypol. II. Formation of
gossypolone with ferric chloride. J. Org. Chem. 30, 411–413.
Jaroszewski, J.W., Strom-Hansen, T., Hansen, S.H., Thastrupo, O., Kofod, H., 1992. On
the botanical distribution of chiral forms of gossypol. Planta Med. 58, 454–458.
Karban, R., 1993. Costs and benefits of induced resistance and plant density for a
native shrub, Gossypium thurberi. Ecology 74, 9–19.
Kim, I.C., Waller, D.P., Marcelle, G.B., 1984. Comparative in vitro spermicidal effects
of ( )-gossypol, (+)-gossypol, (À)-gossypol, and gossypolone. Contraception 30,
253–259.
Loberg, R.D., McGregor, N., Ying, C., Sargent, E., Pienta, K.J., 2007. In vivo evaluation
of AT-101 (R-(À)-gossypol acetic acid) in androgen-independent growth of
VCaP prostate cancer cells in combination with surgical castration. Neoplasia 9,
1030–1037.
Lukefahr, M.J., Martin, D.F., 1966. Cotton-plant pigments as a source of resistance to
the bollworm and tobacco budworm. J. Econ. Entomol. 59, 176–179.
Mellon, J.E., Zelaya, C.A., Dowd, M.K., 2011. Inhibitory effects of gossypol-related
compounds on growth of Aspergillus flavus. Lett. Appl. Microbiol. 52, 406–412.
Oliver, C.L., Miranda, M.B., Shangary, S., Land, S., Wang, S., Johnson, D.E., 2005. (À)-
Gossypol acts directly on the mitochondria to overcome Bcl-2- and Bcl-XL-
mediated apoptosis resistance. Mol. Cancer Ther. 4, 23–31.
Ready, N., Karaseva, H.A., Orlov, S.V., Luft, A.V., Popovych, O., Holmlund, J.T., Wood,
B.A., Leopold, L., 2011. Double blind, placebo-controlled, randomized phase 2
study of the proapoptotic agent AT-101 plus docetaxel, in second-line non-
small cell lung cancer. J. Thorac. Oncol. 6, 781–785.
Shelly, M.D., Hartley, L., Groundwater, P.W., Fish, R.G., 2000. Structure-activity
studies on gossypol in tumor cell lines. Anticancer Drugs 11, 209–216.
Stipanovic, R.D., Altman, D.W., Begin, D.L., Greenblatt, G.A., Benedict, J.H., 1988.
Terpenoid aldehydes in upland cotton: analysis by aniline and HPLC methods. J.
Agric. Food Chem. 88, 509–515.
Stipanovic, R.D., Bell, A.A., Lukefahr, M.J., 1977. Natural insecticides from cotton
(Gossypium). In: Hedin, P.A. (Ed.), Host Plant Resistance to Pests, Symposium
Series #62. American Chemical Society, Washington, DC, pp. 197–214.
Stipanovic, R., Esquivel, J., Westbrook, J., Puckhaber, L., O’Neil, M., Bell, A., Duke, S.,
Dowd, M., Lopez Jr., J.D., Hake, K., 2014. The effect of gossypolone on the
growth and development of Helicoverpa zea (Boddie). Southwest Entomol. 39,
1–7.
Stipanovic, R.D., Puckhaber, L.S., Bell, A.A., Percival, A.E., Jacobs, J., 2005. Occurrence
of (+)- and (À)-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., Williams, H.J., Smith, L., 1986. Cotton terpenoid inhibition of
Heliothis virescens development. In: Green, M.A., Hedin, P.A. (Eds.), Natural
Resistance of Plants to Pests, Symposium Series #296. American Chemical
Society, Washington, DC, pp. 79–94.
Yildirim-Aksoy, P., Lim, C., Dowd, M.K., Wan, P.J., Klesius, P.H., Shoemaker, C., 2004.
In vitro inhibitory effect of gossypol from gossypol-acetic acid, and (+)- and (À)-
isomers of gossypol on the growth of Edwardsiella ictaluri. J. Appl. Microbiol. 97,
87–92.
Phillips, V.A., Hedin, P.A., 1990. Spectral techniques for structural analysis of cotton
terpenoid aldehydes gossypol and gossypolone. J. Agric. Food Chem. 38, 525–
528.
Percival, A.E., Wendel, J.E., McStewart, J.M., 1999. Taxonomy and germplasm
resources. In: Smith, C., Cothren, J.T. (Eds.), Cotton: Origin, History, Technology,
and Production. John Wiley and Sons Inc, New York, pp. 33–63.