Optimization of tocotrienols as antiproliferative and antimigratory leads

Article abstract of DOI:10.1016/j.ejmech.2012.11.012

The vitamin E family members γ- and δ-tocotrienols (2 and 3, respectively) are known natural products with documented anticancer activities. Redox-silent structural modifications, such as esterification, etherification and carbamoylation, of 2 and 3 significantly enhanced their anticancer activities. However, hit-to-lead optimization of tocotrienols and their analogs was yet to be reported at the outset of the project described herein. Subjecting the chroman ring of 2 and 3 to the electrophilic substitution reactions, namely, Mannich and Lederer-Manasse procedures, afforded 42 new products. These included the 3,4-dihydro-1,3-oxazines 3-29 and 35-44, Mannich bases 30-31, and the hydroxymethyl analogs 32-34. Of these, the δ-tocotrienol analogs 8, 11, 18, 24, 25, 27, and 40 inhibited the proliferation of the highly metastatic +SA mammary epithelial cancer cell line, with IC₅₀ values in the nanomolar (nM) range. In NCI's 60 human tumor cell line panel, 8, 17, 38, and 40 showed antiproliferative activity, with nM GI₅₀ values. The δ-tocotrienol analogs 10 and 38 inhibited the migration of the highly metastatic human breast cancer cell line MDA-MB-231 with IC₅₀ values of 1.3 and 1.5 μM, respectively, in the wound-healing assay. A dose of 0.5 mg/day for 14 days of one of the active analogs, 30, significantly slowed the growth of +SA mammary tumors in the syngeneic BALB/c mouse model, compared to the vehicle- and the parent γ-tocotrienol-treated control groups. Electrophilic substitution reactions promoted tocotrienols to lead level and can enable their future use to control metastatic breast malignancies.

Full text of DOI:10.1016/j.ejmech.2012.11.012

European Journal of Medicinal Chemistry 59 (2013) 329e341
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Optimization of tocotrienols as antiproliferative and antimigratory leads
Fathy A. Behery, Mohamed R. Akl, Suryatheja Ananthula, Parash Parajuli, Paul W. Sylvester,
*
Khalid A. El Sayed
Department of Basic Pharmaceutical Sciences, College of Pharmacy, University of Louisiana at Monroe, 1800 Bienville Drive, Monroe, LA 71201, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 9 July 2012
Received in revised form
20 October 2012
Accepted 10 November 2012
Available online 23 November 2012
The vitamin E family members g- and d-tocotrienols (2 and 3, respectively) are known natural products
with documented anticancer activities. Redox-silent structural modifications, such as esterification,
etherification and carbamoylation, of 2 and 3 significantly enhanced their anticancer activities. However,
hit-to-lead optimization of tocotrienols and their analogs was yet to be reported at the outset of the
project described herein. Subjecting the chroman ring of 2 and 3 to the electrophilic substitution reac-
tions, namely, Mannich and LederereManasse procedures, afforded 42 new products. These included the
3,4-dihydro-1,3-oxazines 3e29 and 35e44, Mannich bases 30e31, and the hydroxymethyl analogs 32
Keywords:
e34. Of these, the d-tocotrienol analogs 8, 11, 18, 24, 25, 27, and 40 inhibited the proliferation of the
Antimigratory
Antiproliferative
Breast cancer
In vitro
In vivo
1,3-Oxazine
Mannich
highly metastatic þSA mammary epithelial cancer cell line, with IC50 values in the nanomolar (nM)
range. In NCI’s 60 human tumor cell line panel, 8, 17, 38, and 40 showed antiproliferative activity, with
nM GI50 values. The
human breast cancer cell line MDA-MB-231 with IC50 values of 1.3 and 1.5
d
-tocotrienol analogs 10 and 38 inhibited the migration of the highly metastatic
M, respectively, in the
m
wound-healing assay. A dose of 0.5 mg/day for 14 days of one of the active analogs, 30, significantly
slowed the growth of þSA mammary tumors in the syngeneic BALB/c mouse model, compared to the
LederereManasse
Tocotrienol
vehicle- and the parent
g-tocotrienol-treated control groups. Electrophilic substitution reactions
promoted tocotrienols to lead level and can enable their future use to control metastatic breast
malignancies.
Ó 2012 Elsevier Masson SAS. All rights reserved.
1. Introduction
activities may be attributed to their unique modulatory effects on
various intracellular signal transduction and gene expression pathways
The lipid-soluble vitamin E members are important antioxi-
dants and essential nutrients for normal growth and development
in animals [1]. Vitamin E is the generic name of tocopherols and
tocotrienols that have the same general structure of methyl-
substituted tocols. The term ‘tocol’ trivially refers to 2-methyl-2-
(4,8,12-trimethyltridecyl)chroman-6-ol [2]. Tocotrienols, as the
[4,5].
Structurally, tocotrienol skeleton can be divided into three
domains correlated with biological activities. Domain I (phytyl side
chain) docks the compound to the lipophilic membranes. Domain II
(chroman ring) is responsible for vitamin E effects on different
signaling enzymes. Finally, domain III (phenolic OH group) is
responsible for the redox activity [8].
name implies, have three double bonds
in the tridecyl side chain, while tocopherols have a fully saturated
phytyl side chain. The naturally occurring tocotrienol isomers ( -,
-, -, and -) differ in the number and position of the methyl
groups on the 6-chromanol head [3].
D
3,7,11 with E-configuration
Different structural modifications of tocotrienols enhanced the
in vitro anticancer activity. Domain III esterification, carbamoyla-
tion, and etherification resulted in products with improved anti-
cancer activities [7,9e13]. The optimum length of domain I for
proapoptotic activity was 2e3 isoprenyl units [13]. No reports for
modifications of tocotrienols’ domain II even though these modi-
fications are expected to greatly affect the anticancer activity. This
can be concluded based on the difference in the proapoptotic
potency between tocotrienol isomers which differ in the number
and position of the methyl groups on the chroman ring [13].
Phenols with unsubstituted o- or p-positions are sufficiently
active to undergo electrophilic substitution reactions such as
Mannich and LederereManasse reactions [14]. Mannich reaction is
a
b
g
d
Tocotrienols have better antioxidant activity than tocopherols [4,5].
In addition, tocotrienols exhibit a wide range of redox-silent biological
activities in humans, independent of their radical-scavenging abilities
[6]. This includes antiproliferative, antimigratory, anti-angiogenic,
cholesterol-lowering, and neuroprotective activities [4,7]. These
* Corresponding author. Tel.: þ1 318 342 1725; fax: þ1 318 342 1737.
E-mail address: elsayed@ulm.edu (K.A. El Sayed).
0223-5234/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2012.11.012

330
F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
extensively studied with phenols and naphthols, however only few
publications dealt with tocopherols, not tocotrienols, as a reactive
scaffold for this reaction to produce the corresponding Mannich
base [15].
In Mannich reaction, the condensation of phenol, formaldehyde,
and primary amine, is expected to produce multiple products, such
as 3,4-dihydro-1,3-2H-benzoxazines, o-amino-methylphenols, or
N,N-disubstituted amines. Phenol and the primary amine substit-
uents along with the ratio of the reactants can determine the course
of this reaction and subsequently the nature of the reaction prod-
ucts [16,17].
Mannich reaction products have been of great importance to
pharmaceutical chemistry [18,19]. N-substituted-3,4-dihydro-2H-
1,3-oxazines condensed with aromatic rings has various phar-
macological effects such as antifungal and antitumor activities
[20e24]. This study is the first report for the modifications of
tocotrienols’ domain II by electrophilic substitution reactions
(Mannich and LederereManasse reactions) and the assessment of
the effect of such modifications on the antiproliferative and
antimigratory activities, in vitro and in vivo.
corresponding Mannich base (30 and 31, respectively) with the
concomitant loss of one molecule of formaldehyde (Scheme 1c).
In LederereManasse reaction [27], refluxing toluene solution of
1 with paraformaldehyde in presence of boric and acetic acids
produced 32, the 5-hydroxylmethyl analog of 1. The hydrox-
ymethylation of 2, took place at either C-5 or C-7, producing 33 and
34, respectively (Scheme 1d).
2.1.2. Structure elucidation
The HREIMS of 3 suggested the molecular formula C31H48NO2
with nine degrees of unsaturation which inferred a new ring
formation. The 1H-NMR spectrum of 3 was identical to that of 1 [28],
with the replacement of the aromatic proton signal of H-5 with three
new singlet proton signals at dH 2.58, 3.76, and 4.67. In HMQC, The
proton methyl singlet H3-500
(dH 2.58) was correlated to the new
methyl carbon at dC 40.4. The other two proton signals, dH 3.76 (H2-
400) and 4.67 (H2-200), were correlated to the methylene carbons at dC
50.5 and 83.0, respectively. In HMBC, the methyl H3-500 showed 3J-
HMBC correlations to carbon signals dC 50.5 (C-400) and 83.0 (C-200).
Proton singlet H2-400 showed 3J-HMBC correlations tothe quaternary
carbons at dC 145.2 (C-6) and 115.0 (C-9), in addition to a 2J-HMBC
correlation to C-200. The methylene singlet H2-200 showed showed 3J-
HMBC correlations to C-6, C-400, and C-500. Although the reaction
conditions and the molar ratio of the reactants would suggest the
formation of a Mannich base [12,14], this spectral data indicatethat 2
is not a Mannich base but (R)-2,5,6,8-tetramethyl-8-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno-
[5,6-e] [1,3]oxazine. Extensive analyses of 1H, 13C (Tables 1e3,
and S1eS8) and 2D NMR data along with the HRESIMS analysis
confirmed the 3,4-dihydro-2H-1,3-oxazine ring formation in 4e
29 and 35e44 in a similar fashion used with 3. The oxazine
ring is characterized by the two proton singlets, H-200 and H-400
at the range dH 4.30e4.88 and 3.76e3.95, respectively. C-200 and
C-400 appears in the range of dC 79.9e83.4 and 44.8e50.5,
respectively.
2. Results and discussions
2.1. Chemistry
Both
g- and d-tocotrienols (1 and 2, respectively) were isolated
from a tocotrienol-rich fraction of palm oil using normal phase
vacuum liquid chromatography [7]. Forty-two new tocotrienol
analogs were prepared by electrophilic substitutions of the chro-
man ring.
2.1.1. Synthesis of tocotrienol analogs
In a Mannich-type reaction, the condensation of equimolar
amounts of tocotrienol, primary amine, and formaldehyde
produced the 3,4-dihydro-1,3-oxazine analogs 3e29, 35e44.
However, using equimolar ratios of the reactants is expected to
produce the Mannich bases [15,25]. In a primary amine, one
hydrogen can be substituted to give the Mannich base, or the two
hydrogen atoms can be substituted to give either the 3,4-dihydro-
1,3-oxazine or bis-substituted tertiary amine [18,26]. In addition,
the nature of the phenol and its substituents may play a major role
in directing the reaction, regardless of the molar ratio of the reac-
tants [17]. Both 1 and 2 were employed and they underwent the
reaction successfully. Because 2 has two nonsubstituted o-positions
(C-5 and C-7), two positional isomers were produced. The C-5
substituted product was predominant due to the higher reactivity
of C-5 than C-7 site toward the electrophilic substitution reactions
[3]. In addition, it is well established that the reaction takes place in
d-Tocotrienol (2) reacted with 2-aminoethanol and HCHO,
affording the isomers 10 (major) and 11 (minor). Both showed the
characteristic two proton singlets and two methylenes of the oxa-
zine ring. The spectroscopic identification of 10 as 2-((R)-6,8-
dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-3,7,11-trienyl)-9,10-
dihydrochromeno[5,6-e] [1,3]oxazin-2(1H,3H,8H)-yl)ethanol was
achieved in a similar manner used for 3. The aromatic proton
singlet H-5 (dH 6.39, Table 3) in 11 showed 3J-HMBC correla-
tions to the quaternary aromatic carbons C-7 and C-10 (dC 123.4
and 146.0, respectively, Table 1). The proton singlet H2-400
(dH
3.90) showed 3J-HMBC correlations to the quaternary aromatic
carbons at dC 146.4 (C-6) and 120.5 (C-8), confirming the oxa-
zine ring closure at C-7. Therefore 11 was proven to be 2-((R)-
5,7-dimethyl-7-((3E,7E)-4,8,12-trimethyl-trideca-3,7,11-trienyl)-
8,9-dihydrochromeno[7,6-e] [1,3]oxazin-3(2H,4H,7H)-yl)ethanol.
The HRESIMS of the Mannich base analogs 30 or 31 suggested
molecular formula of C39H58N O4 or C37H54NO3, respectively, 12
mass units less than their parents 23 or 26. This suggested the 1,3-
oxazine hydrolytic ring opening with the loss of a formaldehyde
molecule. This was confirmed by NMR data (Tables 1 and 2). The
1H- and 13C-NMR data for 30 and 31 were identical to those of their
3,4-dihydro-1,3-oxazine parents 23 and 26, respectively, with the
loss of the methylene proton and the carbon signals H2/C-200. The
the o-position
the hydroxyl group in a
a
to the ring fusion in binuclear phenols that have
-position to the ring fusion [18]. Different
b
aliphatic, alicyclic, and aralkyl primary amines were used with
variable alkyl chain lengths and diverse terminal functional groups.
The reaction did not proceed with secondary amines. The reaction
proceeded toward the 1,3-oxazine formation regardless of the
molar ratio of the reactants and their mixing order. The reaction
was carried out either by mixing the primary amine, formaldehyde,
and tocotrienol or condensing the amine and formaldehyde first
and then the product was allowed to react with tocotrienol. Either
dioxane or ethanol was used and occasionally the reaction was
allowed to proceed under solvent-less conditions. Both aqueous
formaldehyde and paraformaldehyde were satisfactorily used
(Scheme 1a, b).
proton singlet H2-100 dH 3.90) showed 3J-HMBC correlations to the
(
quaternary aromatic carbons at dC 147.7 (C-6) and 116.2 (C-8),
confirming the oxazine ring opening and the attachment of the
aminomethylene C-100 to C-5. Therefore compounds 30 and 31
were confirmed to be (R)-5-((3,4-dimethoxyphenethyl-amino)
methyl)-2,7,8-trimethyl-2-((3E,7E)-4,8,12-trimethyltrideca-3,7,11-
trienyl)chroman-6-ol and (R)-5-((4-hydroxyphenethylamino)
The Mannich bases were prepared from the corresponding 3,4-
dihydro-1,3-oxazine analogs. 23 and 26 were shaken for 48 h at
room temperature in 75% methanol solution to give the

F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
331
R`
HN
R`
N
5
HO
O
O
R
+
c
a or b
N
O
C16H27
R
O
C16H27
R`
7
O
C16H27
R`
R
R`
R`
R`
30 e
9'
5'
1'
11 CH2CH2OH
3
CH3 CH3
21 CH3 (CH2)5COOH
31
f
25
4
22
e
H
CH3
H
(CH2)5COOH
5
23
CH3 CH2CH=CH2
CH3
e
e
f
6
CH3
CH3
H
a
b
b
24 H
26
7
CH3
27 H
28
8
f
9
CH3 CH2CH2OH
CH3
29
g
g
10
H
CH2CH2OH
H
12 CH3 (CH2)3OH
35 CH3 (CH2)4OH
13
36
H
(CH2)3OH
H
(CH2)4OH
37 CH3 (CH2)6OH
38
2
O
14 CH3
c
c
d
4
15
H
H
(CH2)6OH
CH3 (CH2)8OH
40 H (CH2)8OH
41
16
39
CH3
8
5
17 CH3 (CH2)5OH
18
H
(CH2)5OH
CH3 (CH2)10OH
HO
19 CH3 (CH2)4COOH 42 H
(CH2)10OH
R
20
43
H
(CH2)4COOH
CH3 (CH2)12OH
R
44
H
(CH2)12OH
HO
HO
1 CH3
2
H
HO
HO
d
+
R
O
C16H27
O
C16H27
R
32 CH3
34
33
H
13''
11''
N
O
OH
H
O
O
O
N
O2N
7''
7''
9''
5''
9''
7''
7''
8''
5''
5''
5''
5''
5''
5''
b
c
d
e
g
f
a
Scheme 1. Semisynthetic preparation of tocotrienol analogs. (a) 37% aq. CH2O þ RNH2/dioxane, stirring, rt, 1 h, reflux, overnight. (b) (CH2O)n þ RNH2, stirring, rt, overnight. (c)
200 mL 25% aq. MeOH, stirring, 48 h (d) (CH2O)n þ H3BO3 þ CH3COOH/toluene, reflux, overnight.
methyl)-2,7,8-trimethyl-2-((3E,7E)-4,8,12-trimethyltri-deca-
3,7,11-trienyl)chroman-6-ol, respectively.
2.2. Biological evaluations and structureeactivity relationship
The HRESIMS of 32 suggested a molecular formula of C29H44O3.
2.2.1. In vitro activities
The hydroxymethylene proton singlet H2-100
(dH 4.85, Table 3)
2.2.1.1. Antiproliferative activity against the highly metastatic þ SA
showed 3J-HMBC correlations to the quaternary aromatic carbons
at dC 147.4 (C-6, Table 2) and 118.8 (C-9), in addition to a 2J-HMBC
correlation to the quaternary aromatic carbon at dC 115.2 (C-5). This
confirmed the hydroxymethylation at C-5. Accordingly, 32 was
confirmed to be (R)-5-(hydroxymethyl)-2,7,8-trimethyl-2-((3E,7E)-
4,8,12-trimethyltri-deca-3,7,11-trienyl)chroman-6-ol. The HRESIMS
of 33 and 34 suggested the same molecular formula C28H42O3.
Compound 33 was confirmed to be (R)-5-(hydroxymethyl)-2,8-
dimethyl-2-((3E,7E)-4,8,12-trimethyltrideca-3,7,11-trienyl)chroman-
6-ol, in a similar fashion to 32. C-7 hydroxymethylation in case
of 34 was confirmed through the 3J-HMBC correlations of
mouse mammary epithelial cancer cells. The parent tocotrienols 1
and
metastatic þSA mammary epithelial cell line at IC50 values of 4.0
and 3.0 M, respectively [29]. The antiproliferative activities of 3e
2 effectively inhibited the proliferation of the highly
m
44 against the abovementioned cells were evaluated by MTT assay
and the IC50 values were calculated (Fig. 1, Table 4). Thirty out of 42
prepared compounds have shown IC50 values less than 3
mM, with
better activity than -tocotrienol (2). Of these, seven compounds
d
showed nanomolar IC50 values for the first time in the tocotrienol
family (Table 4).
Generally, the oxazine analogs of 2 (
IC50 values than the corresponding analogs of 1 (g
d
-isomer) showed lower
hydroxymethylene proton singlet H2-100
(dH 4.87, Table 3) to the
-isomer).
quaternary aromatic carbons at dC 148.8 (C-6, Table 2) and 122.3
(C-8), the proton signal at was correlated to the carbon signal at
dC 124.1 (C-7). Therefore, compound 34 was confirmed to be (R)-5-
(hydroxymethyl)-2,7,8-trimethyl-2-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)chroman-6-ol.
This pattern was consistent along most of the prepared oxazines
except those with terminal CH3CH2O (24e15), COOH (19e22), or
NO2 (28e29) groups. However, this pattern was reversed in the
ManasseeLederer-based analogs 32e34. The
g- isomer analog 32
showed lower IC50 value the than the -isomers 33 and 34.
d

332
F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
Table 1
Table 2
13C-NMR data of 3, 10, 11, 23, and 26.a
13C-NMR data of 30e34.a
dC, mult.
dC, mult.
3
10
11
23
26
Position
30
31
32
33
34
2
74.5, C
74.7, C
75.6, C
74.5, C
74.6, C
2
74.1, C
74.2, C
74.4, C
74.5, C
75.5, C
2-CH3
3
4
5
6
7
23.9, CH3
31.1, CH2
18.9, CH2
114.7, C
145.2, C
124.0, C
23.9, CH3
31.0, CH2
19.1, CH2
115.2, C
146.3, C
116.4, CH
e
24.2, CH3
31.3, CH2
22.3, CH2
113.7, CH
146.4, C
123.4, C
e
23.8, CH3
31.1, CH2
19.0, CH2
115.0, C
145.2, C
124.0, C
23.9, CH3
31.1, CH2
18.9, CH2
114.8, C
145.2, C
124.0, C
2-CH3
3
4
5
6
7
23.7, CH3
31.6, CH2
20.4, CH2
115.5, C
147.7, C
123.0, C
23.7, CH3
31.6, CH2
20.4, CH2
116.2, C
149.0, C
125.0, C
23.8, CH3
31.5, CH2
19.9, CH2
115.2, C
147.4, C
125.9, C
23.8, CH3
31.4, CH2
19.9, CH2
118.3, C
148.3, C
116.8, CH
e
24.1, CH3
31.3, CH2
22.3, CH2
113.7, CH
148.8, C
124.1, C
e
7-CH3
8
11.4, CH3
123.0, C
11.5, CH3
123.1, C
11.5, CH3
123.1, C
7-CH3
8
11.8, CH3
124.9, C
11.9, CH3
123.1, C
11.5, CH3
123.1, C
126.0, C
15.9, CH3
117.8, C
146.0, C
39.6, CH2
22.2, CH2
124.5, CH
135.3, C
16.1, CH3
39.8, CH2
26.8, CH2
124.3, CH
135.1, C
16.0, CH3
39.8, CH2
26.7, CH2
124.2, CH
131.4, C
17.8, CH3
25.8, CH3
81.9, CH2
47.9, CH2
53.9, CH2
59.2, CH2
120.5, C
10.5, CH3
116.9, C
146.0, C
39.9, CH2
22.4, CH2
124.5, CH
135.3, C
16.1, CH3
39.8, CH2
26.8, CH2
124.3, CH
135.1, C
16.0, CH3
39.8, CH2
26.7, CH2
124.2, CH
131.4, C
17.8, CH3
25.8, CH3
81.6, CH2
48.7, CH2
53.9, CH2
58.9, CH2
127.3, C
15.9, CH3
119.9, C
145.5, C
39.5, CH2
22.2, CH2
124.5, CH
135.3, C
16.2, CH3
39.8, CH2
26.8, CH2
124.3, CH
135.1, C
16.1, CH3
39.8, CH2
26.7, CH2
124.3, CH
131.4, C
17.8, CH3
25.8, CH3
59.6, CH2
122.3, C
11.2, CH3
121.3, C
145.3, C
39.8, CH2
22.5, CH2
124.5, CH
135.2, C
16.0, CH3
39.8, CH2
26.8, CH2
124.3, CH
135.1, C
15.9, CH3
39.8, CH2
26.7, CH2
124.2, CH
131.4, C
17.8, CH3
25.8, CH3
60.6, CH2
8-CH3
9
11.7, CH3
115.0, C
144.6, C
11.7, CH3
115.0, C
145.2, C
11.7, CH3
115.0, C
145.1, C
8-CH3
9
12.0, CH3
116.2, C
144.2, C
12.0, CH3
116.2, C
144.3, C
12.0, CH3
118.8, C
144.9, C
10
10
10
39.8, CH2
22.3, CH2
124.5, CH
135.2, C
16.1, CH3
39.8, CH2
26.8, CH2
124.4, CH
135.0, C
11.7, CH3
39.8, CH2
26.7, CH2
124.3, CH
131.3, C
17.8, CH3
25.8, CH3
83.0, CH2
50.5, CH2
40.4, CH3
39.8, CH2
22.3, CH2
124.5, CH
135.2, C
16.1, CH3
39.8, CH2
26.8, CH2
124.4, CH
135.1, C
16.0, CH3
39.8, CH2
26.7, CH2
124.3, CH
131.4, C
17.8, CH3
25.8, CH3
81.7, CH2
48.8, CH2
53.8, CH2
34.8, CH2
132.7, C
39.2, CH2
22.3, CH2
124.5, CH
135.2, C
16.1, CH3
39.2, CH2
26.8, CH2
124.4, CH
135.1, C
16.0, CH3
39.2, CH2
26.7, CH2
124.3, CH
132.1, C
17.8, CH3
25.8, CH3
81.5, CH2
48.8, CH2
53.9, CH2
34.0, CH2
131.3, C
10
39.8, CH2
22.3, CH2
124.4, CH
135.2, C
16.1, CH3
39.8, CH2
26.8, CH2
124.3, CH
135.0, C
16.0, CH3
39.8, CH2
26.7, CH2
124.3, CH
131.4, C
17.8, CH3
25.8, CH3
47.6, CH2
50.0, CH2
35.6, CH2
131.8, C
39.7, CH2
22.3, CH2
124.5, CH
135.2, C
16.1, CH3
39.7, CH2
26.8, CH2
124.4, CH
135.1, C
16.0, CH3
39.7, CH2
26.7, CH2
124.3, CH
131.4, C
17.8, CH3
25.8, CH3
47.4, CH2
49.9, CH2
34.9, CH2
130.8, C
39.5, CH2
22.3, CH2
124.5, CH
135.2, C
16.0, CH3
39.8, CH2
26.8, CH2
124.4, CH
135.1, C
15.9, CH3
39.8, CH2
26.7, CH2
124.3, CH
131.3, C
17.5, CH3
25.8, CH3
60.1, CH2
20
20
30
30
40
40
40-CH3
50
40-CH3
50
60
60
70
70
80
80
80-CH3
90
80-CH3
90
100
100
110
110
120
120
120a-CH3
120b-CH3
200
12’a-CH3
12’b-CH3
100
400
300
500
400
600
500
700
600
111.9, CH
149.3, C
149.1, C
111.4, CH
120.7, CH
56.0, CH3
55.9, CH3
129.9, CH
115.6, CH
154.6, C
115.6, CH
129.9, CH
800
112.0, CH
148.9, C
147.5, C
111.3, CH
120.6, CH
56.0, CH3
56.0, CH3
129.8, CH
115.3, CH
154.1, C
115.3, CH
129.8, CH
700
900
800
1000
1100
1200
1300
1400
900
1000
1100
1200
a
In CDCl3, 100 MHz. Carbon multiplicities were determined by APT experiments,
C, quaternary; CH, methine; CH2, methylene; CH3, methyl carbons.
a
In CDCl3, 100 MHz. Carbon multiplicities were determined by APT experiments,
C, quaternary; CH, methine; CH2, methylene; CH3, methyl carbons.
4-carbons (36), 6-carbons (38), and 8-carbons (40). However,
increasing the length of the N-300 substituent to 12-carbons, led to
reducing the activity and increasing the mean60 value. This may
indicate that the optimal length of the N-300 substituent for best
Extending the length of the N-300 substituent (up to 8-carbons)
proved to enhance the activity. In addition, a terminal OH group at
N-300 substituent was preferred for the activity over the COOH (20),
NH (16), or NO2 (28, Fig. 2). Masking the OH by etherification did
not affect the activity (23, 24, 15) except in 14, whose activity was
significantly decreased (Table 4).
activity was 8-carbons. Comparison of the mean60 values for
-isomers substituents indicate that the oxazine analogs of 2
-isomer) were more cytotoxic or growth inhibitors than the cor-
responding 1 analogs ( -isomer, Fig. 3). Representing the data in the
g- and
d
(d
g
2.2.1.2. The antiproliferative and cytotoxic activities against the
National Cancer Institute (NCI) 60 human cell lines panel. The
growth inhibitory activity of twenty compounds, eighteen oxazines
(8, 13, 17, 18, 23, 24, 26, 27, and 35e44) and two Mannich bases (30
and 31), was evaluated against the National Cancer Institute’s panel
of 60 human cell lines (NCI-60) at a single dose of 10 mM (Fig. SI1).
The growth inhibition activity of these compounds, except 23, 26,
30, and 31, have been evaluated at five different doses (100, 10, 0.1,
form of mean graphs, in which the percent growth of each cell line is
subtracted from the mean60, enables the detection of the differen-
tial growth inhibition activity of the compound on different tumor
tissues and cell lines. Tocotrienol analogs showed selective growth
inhibition of blood and colon malignant cell lines (Fig. 4).
Compounds 8, 17, 38, and 40 inhibited the growth of leukemia and
colon cancer cell lines at nanomolar GI50 values (Fig. SI2). This
selectivity pattern to leukemia and colon cell lines was further
confirmed when IC50 values of the growth inhibitory activity of some
of the tocotrienol analogs were determined by MTT assay against the
human breast cancer cell lines MDA-MB-231 and MCF7 (Table 5).
0.01, 0.001
growth of 50% of the cell, were calculated (Fig. SI2). Compounds 13,
36, 38, and 40 were lethal to the cells at 10 M dose, showing
mM). The GI50 values, the concentrations that inhibit the
m
negative mean percent growth values across all 60 cell lines
(mean60), while 1 was inactive (Fig. 3). It was obvious from the
mean60 values that the oxazine ring was essential pharmacophore
required for the activity since the Mannich bases 30 and 31, without
the oxazine ring, were showing the same activity as 1. The mean60
was decreasing by increasing the length of the N-300 substituent from
2.2.1.3. Antiproliferative activity against the non-tumorigenic
MCF10A epithelial cell line. Compound 1, along with ten toco-
trienol analogs which showed the high potencies against the highly
metastatic þSA cell line, were then tested against the non-
tumorigenic MCF10A human mammary epithelial cells in order to

F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
333
Table 3
1H-NMR data of 3, 10, 11, 23, 26, 30e34.a
dH, mult., J
Position
3
10
11
23
26
30
31
32
33
34
2-CH3
3
4
5
1.24, s
1.78, m
2.44, m
1.25, s
1.67, m
2.45, m
1.25, s
1.75, m
2.69, m
6.39, s
1.24, s
1.75, m
2.43, m
1.24, s
1.77, m
2.42, m
1.20, s
1.74, m
2.56, t, 6.9
1.21, s
1.74, m
2.57, t, 6.4
1.23,s
1.77, m
2.62, m
1.23, s
1.74, m
2.60, m
1.24, s
1.74, m
2.68, m
6.44, s
7
6.49, s
6.51, s
7-CH3
8-CH3
10
2.10, s
2.10, s
1.54, m
2.07, m
5.13, m
1.59, s
1.97, m
2.05, m
5.13, m
1.59, s
1.97, m
2.05, m
5.13, m
1.59, s
1.67, s
2.10, s
2.09, s
1.55, m
2.12, m
5.11, m
1.59, s
1.97, m
2.05, m
5.11, m
1.59, s
1.97, m
2.05, m
5.11, m
1.58, s
1.67, s
2.10, s
2.09, s
1.55, m
2.14, m
5.10, m
1.60, s
1.97, m
2.05, m
5.10, m
1.60, s
1.97, m
2.05, m
5.10, m
1.58, s
1.68, s
2.13, s
2.09, s
1.52, m
2.10, m
5.08, m
1.57, s
1.94, m
2.03, m
5.08, m
1.57, s
1.94, m
2.03, m
5.08, m
1.56, s
1.65, s
3.89, s
2.13, s
2.09, s
1.52, m
2.11, m
5.08, m
1.57, s
1.96, m
2.04, m
5.08, m
1.57, s
1.96, m
2.04, m
5.08, m
1.58, s
1.66, s
3.89, s
2.15, s
2.11, s
1.52, m
2.12, m
5.11, m
1.59, s
1.96, m
2.05, m
5.11, m
1.59, s
1.96, m
2.05, m
5.11, m
1.58, s
1.67, s
4.85, s
2.12, s
1.98, s
2.10, s
1.52, m
2.10, m
5.10, m
1.58, s
1.98, m
2.05, m
5.10, m
1.58, s
1.98, m
2.05, m
5.10, m
1.58, s
1.67, s
4.79, s
2.10, s
1.53, m
2.09, m
5.08, m
1.58, s
1.96, m
2.04, m
5.08, m
1.58, s
1.96, m
2.04, m
5.08, m
1.58, s
1.67, s
4.87, s
1.55, m
2.14, m
5.09, m
1.59, s
1.96, m
2.05, m
5.09, m
1.59, s
1.96, m
2.05, m
5.09, m
1.58, s
1.52, m
2.11, m
5.10, m
1.59, s
1.96, m
2.04, m
5.10, m
1.59, s
1.96, m
2.04, m
5.10, m
1.58, s
20
30
40-CH3
50
60
70
80-CH3
90
100
110
12’a-CH3
12’b-CH3
100
1.67, s
1.67, s
200
4.67, s
4.74, s
4.75, s
4.78, s
4.78, s
300
2.91, t, 7.3
2.77, t, 6.4
2.88, t, 7.1
2.74, t, 6.9
400
3.76, s
2.58, s
3.84, s
2.95, t, 5.1
3.67, t, 5.2
3.90, s
2.95, t, 5.1
3.68, t, 4.8
3.85, s
2.99, t, 6.22
2.85, t, 6.42
3.84, s
2.98, t, 7.7
2.82, t, 7.7
500
600
6.69, d, 1.8
7.01, d, 8.2
6.74, d, 8.7
700
800
6.74e6.77, m
7.06, d, 8.4
6.71, d, 8.4
900
6.78, d, 7.8
6.74, d, 8.7
7.01, d, 8.2
1000
1100
1200
1300
1400
6.72, dd, 1.8, 8.2
3.86, s or 3.84, s
3.86, s or 3.84, s
6.74e6.77, m
6.74e6.77, m
3.86, s
6.71, d, 8.4
7.06, d, 8.4
3.84, s
a
In CDCl3, J in Hz. 400 MHz.
investigate their selectivity for tumor cells over normal cells. All
tested tocotrienol analogs, except 23, showed higher selectivity
for þ SA cells than for MCF10A (Table 7). The selectivity indices of
tocotrienol analogs were higher than the parent compound 1
(selectivity index ¼ 1.1, Table 7).
melanoma, prostate, as well as breast cancer cell lines [7,9,30e32].
The wound-healing assay (WHA) is a simple method for the study
of in vitro cell migration [7]. The antimigratory activities of the
prepared compounds were evaluated against the highly metastatic
MDA-MB-231 human breast cancer cells (Fig. 5). The concentra-
tions required to inhibit the migration of 50% of the cells across the
wound, IC50 values, were calculated (Table 6). The parent toco-
trienols 1 and 2 showed limited activity against this aggressive cell
2.2.1.4. Antimigratory activity against the highly metastatic MDA-
MB-231 breast cancer cell line. Tocotrienols were reported to
inhibit the metastasis and invasion of gastric adenocarcinoma,
line with IC50 values more than 40 mM. The 1,3-oxazine analogs are
more active than both the Mannich bases and the hydroxymethyl
derivatives as antimigratory agents. A terminal OH group, alcoholic
4
15
2.5
2.0
1.5
1.0
0.5
17
18
26
Table 4
Antiproliferative activity of tocotrienol analogs against the highly metastatic þSA
mouse mammary epithelial cancer cell line.a
*
Compound
IC50
(
mM)
Compound
IC50
1.7
0.70
1.8
11.9
5.0
5.1
1.8
0.87
0.84
1.5
(
m
M)
Compound
IC50 (mM)
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2.2
1.3
11.6
2.0
8.7
0.8
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
2.3
2.6
>20
6.6
2.2
1.8
2.1
1.3
1.4
0.79
4.4
3.6
>5.0
2.0
*
*
*
*
*
1.7
4.2
*
*
0.96
2.5
1.8
9.3
1.0
4.6
*
*
0.91
24.6
29.1
1.7
0
1
2
Treatment Concentration (µM)
Fig. 1. Effect of tocotrienol analogs 4, 15, 17, 18, and 26 on the viability of þSA
mammary tumor cells. P < 0.05 as respect to each compound respective vehicle
a
The IC50 values are calculated by non-linear regression fitting using GraphPad
treatment control (0
mM).
Prism 5.04.

334
F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
1-treated control groups. In contrast, þSA mammary tumors in
animals receiving 0.5 mg/day of compound 30 grew at a signifi-
cantly slower rate, as compared to tumors in the vehicle control-
and 1-treated groups (Fig. 6). Body weights did not significantly
differ among the different treatment groups at any time during the
14-day treatment experimental period.
This remarkable improvement in tocotrienol activity both
in vitro and in vivo by the introduction of the new carbonecarbon
bond through Mannich-based reaction opens new horizons for
future hit-to-lead development of tocotrienol analogs as potential
drugs to control metastatic breast cancer.
25
20
15
10
5
3. Conclusions
New 42 tocotrienol 3,4-dihydro-1,3-oxazine, Mannich base and
hydroxymethyl analogs were prepared. The 3,4-dihydro-1,3-
oxazine ring together with N-300 substituent were important
pharmacophores for the antiproliferative and antimigratory activ-
ities of the newly prepared tocotrienol analogs. In addition, the
degree of methylation of the chroman ring has shown to affect the
activities in a significant way. Modifications of domain II of toco-
trienols by electrophilic substitution reactions can selectively
improve not only the in vitro antiproliferative and antimigratory
activities with minimal toxicity to normal cells but also the in vivo
tumor growth inhibition activities of these compounds. This study
shows the potential of tocotrienol as scaffold candidates for use to
design future drugs for the control of metastatic breast cancer.
26
17
23
16
20
28
Compound Number
Fig. 2. The effect of the terminal group at N-300 substituent on the viability of þSA
mammary tumor cells. Free OH (17 and 26) or OCH3 (23)
(20) > NO2 (28).
> NH (16) > COOH
or phenolic, is important for the activity. The oxazine analogs of 2
(
d
-isomer) were more active than the corresponding analogs of 1
-isomer) as antimigratory agents. The most active compounds
M,
(g
were 10, 38, 36, and 27 with IC50 values of 1.3, 1.5, 2.2, and 2.4
m
respectively (Table 6). Compounds 10, 38, and 36 possess linear
aliphatic ethanol, hexanol, or butanol side chains attached to N-300.
This clearly shows the preference of a terminal free hydroxy func-
tionality at 2e6 carbons distance from N-300. The only analog
among the most active compounds with aromatic (tyramine) side
chain (27) can still be envisioned as a 6-carbons chain separating
the free phenol hydroxy and N-300. Therefore, these compounds can
be considered as potential hits appropriate for optimization as
antimetastatic agents.
4. Materials and methods
4.1. Chemistry
All reagents and chemicals were purchased from SigmaeAldrich
Chemical Co. Tocotrienol-rich fraction 50 g (Palm TRF 70%, low in
tocopherol from First Tech International Ltd., Hong Kong) was
fractionated on Si gel 60 using n-hexane-ethyl acetate (gradient
elution). Optical rotations were measured on a Rudolph Research
Analytical Autopol III polarimeter. IR spectra were recorded on
a Varian 800 FT-IR spectrophotometer. The 1H- and 13C-NMR
spectra were recorded in CDCl3, using TMS as an internal standard,
on a JEOL Eclipse-400 NMR spectrometer, operating at 400 MHz for
1H and 100 MHz for 13C. The HREIMS experiments were conducted
at Louisiana State University on a 6200-TOF LCMS (Agilent)
equipped with multimode source (mixed source that can ionized
the compounds alternatively by ESI and APCI). The ESIMS was
conducted using 3200 Q-trap LC/MS/MS system (Applied Bio-
systems, Foster City, CA) using Analyst version 1.4.1 software (MDS
Sciex; Toronto, Canada). The analytes were ionized using electro-
spray ionization (ESI) interface operated in the positive mode. The
analysis was conducted using Q1 scan and mass scan range was m/z
50e500 (0.15 s/scan). TLC analysis was carried on precoated Si gel
2.2.2. In vivo activity
The effects of daily iv treatment with 0 or 0.5 mg of the parent
compound 1 or 30 dissolved in 0.1 mL propylene glycol on þSA
mammary tumor growth in syngeneic BALB/c mice is shown in
Fig. 6. Palpable þSA mammary tumors started to appear in all
groups approximately 40 days following tumor cell injection into
the #4 mammary gland fat pad. Measurable tumors continued to
grow throughout the 14-day treatment period in the vehicle- and
120
100
80
60
40
60 F254 500
mm TLC plates (EMD Chemicals), using variable
20
proportions of n-hexaneeethyl acetate and ethyl acetatee
methanolewater as a mobile phase. Vanillin in concentrated
H2SO4 (1% w/v) was used as visualizing reagent. For column chro-
0
-20
-40
-60
-80
matography, Si gel 60 (Natland, 63e200 mm) was used.
4.2. General procedure for the preparation of 3,4-dihydro-2H-1,3-
oxazines (3e29 and 35e44)
1
8
13 17 18 23 24 26 27 30 31 35 36 37 38 39 40 44
Compound
4.2.1. Method A
To a dioxane [15] solution (5 mL) of tocotrienol (4 mmol), the
amine (20 mmol) was added. Mixture was cooled in an ice bath and
37% formaldehyde (21 mmol) was added drop wise while stirring. It
was then stirred at room temperature for 1 h and then refluxed
Fig. 3. The mean percent growth activity of the compounds 1, 8, 13, 17, 18, 23, 24, 26,
27, 30, 31, 35e40, 44 across all the 60 cell lines (mean60). The negative values indicate
cytotoxicity.

F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
335
Fig. 4. The selectivity pattern of tocotrienol analogs on the cell lines belonging to different tissues. The selectivity is calculated by subtracting the mean growth percent of each
tissue from the mean60. A positive value indicates selectivity.
overnight. Reaction mixture was concentrated under vacuum and
the yellow residue obtained was dissolved in ethyl acetate (20 mL),
washed several times with saturated NaCl solution, dried over
anhydrous MgSO4, and concentrated in vacuum. The residue ob-
tained was subjected to column chromatography using normal
phase silica gel as stationary phase and gradient ethyl acetate/n-
hexane system as mobile phase (Scheme 1a).
4.2.3. (R)-2,5,6,8-tetramethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno[5,6-e] [1,3]oxazine
(3)
Yellow viscous oil, ½a D25
þ1.7 (c 0.016, CH2Cl2); IR (CH2Cl2)nmax
ꢀ
3599, 2975, 2926, 1605 cmꢁ1; 1H- and 13C-NMR: see Tables 1 and 3;
HREIMS m/z 466.3664, [M þ H]þ (calcd for C31H48NO2, 466.3680).
4.2.4. (R)-2,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-3,7,11-
4.2.2. Method B
trienyl)-1,2,3,8,9,10-hexahydrochromeno[5,6-e] [1,3]oxazine (4)
A mixture of tocotrienol [25] (5.0 mmol), paraformaldehyde
(5.0 mmol) and amines (5.0 mmol) was stirred and left overnight at
rt. The reaction was carried out under solvent-less conditions (few
drops of dioxane or ethanol may be added). The residue was
purified by flash chromatography directly from the reaction
mixture without any work-up (Scheme 1b).
Yellow viscous oil, ½a D25
þ4.7 (c 0.024, CH2Cl2); IR (CH2Cl2)nmax
ꢀ
3660, 22858, 1618, 1476, 1345, 1099, 942 cmꢁ1 1H- and 13C-NMR:
;
see Tables S1 and S5; HREIMS m/z 452.6907, [M þ H]þ (calcd for
C30H46NO2, 452.6917).
4.2.5. (R)-2-allyl-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno[5,6-e] [1,3]oxazine (5)
Yellow viscous oil, ½a D25
ꢀ þ0.1 (c 0.03, CH2Cl2); IR (CH2Cl2)nmax
Table 5
3676, 2923, 1606 cmꢁ1 1H- and 13C-NMR: see Tables S1 and S5;
;
Antiproliferative activity (IC50, mM) of tocotrienol analogs against the highly meta-
HREIMS m/z 492.3843, [M þ H]þ (calcd for C33H50NO2,492.3836).
static MDA-MB-231 and MCF7 human cancer cell lines and the non-tumorigenic
MCF10A epithelial cell line.a
4.2.6. (R)-2-benzyl-5,6,8-trimethyl-8-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno
[5,6-e] [1,3]oxazine (6)
Compound
MCF7
MDA-MB-231
1
6
7.1
>20.0
6.0
>20.0
>20.0
>20.0
>20.0
9.1
>20.0
10.6
18.2
>20.0
16.3
>20.0
>20.0
>20.0
>20.0
19.4
4.6
>20.0
>20.0
9.4
27.3
>5.0
9.6
>10.0
0.8
>50.0
9.0
>50.0
>5.0
>5.0
13.8
>5.0
>10.0
>5.0
>20.0
11.7
Yellow viscous oil, ½a D25
ꢀ þ0.6 (c 0.04, CH2Cl2); IR (CH2Cl2)nmax
10
12
13
16
17
18
23
26
27
35
36
37
38
39
40
41
42
43
44
3600, 2927, 2850, 1605 cmꢁ1; 1H- and 13C-NMR: see Tables S1 and
S5; HREIMS m/z 542.3980, [M þ H]þ (calcd for C37H52NO2,
542.3993).
4.2.7. (R)-2-(1-benzylpiperidin-4-yl)-5,6,8-trimethyl-8-((3E,7E)-
4,8,12-trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-
hexahydrochromeno[5,6-e] [1,3]oxazine (7)
Yellow viscous oil, ½a D25
ꢀ þ0.2 (c 0.047, CH2Cl2); IR (CH2Cl2)nmax
3599, 2928, 1605, 1378 cmꢁ1; 1H- and 13C-NMR: see Tables S1 and
S5; HREIMS m/z 625.4733, [M þ H]þ (calcd for C42H61N2O2,
625.4728).
15.7
12.4
>50.0
>50.0
4.2.8. (R)-2-(1-benzylpiperidin-4-yl)-6,8-dimethyl-8-((3E,7E)-
4,8,12-trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-
hexahydrochromeno[5,6-e] [1,3]oxazine (8)
Yellow viscous oil, ½a D25
ꢀ
þ0.2 (c 0.083, CH2Cl2); IR (CH2Cl2)nmax
a
The IC50 values are calculated by non-linear regression fitting using GraphPad
Prism 5.04.
3684, 2927, 1605, 1474 cmꢁ1 1H- and 13C-NMR: see Tables S1 and
;

336
F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
120
100
80
60
40
20
0
12
27
35
39
40
42
Compound Concentration (µM)
Fig. 5. Doseeresponse antimigratory activity of 12, 27, 35, 39, 40, and 42 against the highly metastatic MDA-MB-231 human breast cancer cells in wound-healing assay.
S5; HREIMS m/z 611.4581, [M þ H]þ (calcd for C41H59N2O2,
and 3; HREIMS m/z 482.3620, [M þ H]þ (calcd for C31H48NO3,
611.4571).
482.3629).
4.2.9. 2-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)ethanol (9)
4.2.12. 3-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)propan-1-ol (12)
Yellow viscous oil, ½a D25
ꢀ
þ66.0 (c 0.0008, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ1.7 (c 0.03, CH2Cl2); IR (CH2Cl2)nmax
3602, 2850,1605,1091 cmꢁ1; 1H- and 13C-NMR: see Tables S1 and S5;
3599,1647,1289,1242 cmꢁ1; 1H- and 13C-NMR: see Tables S1 and S5;
HREIMS m/z 496.3769, [M þ H]þ (calcd for C32H50NO3, 496.3785).
HREIMS m/z 510.3949, [M þ H]þ (calcd for C33H52NO3, 510.3942).
4.2.10. 2-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)ethanol (10)
4.2.13. 3-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)propan-1-ol (13)
Yellow viscous oil, ½a D25
ꢀ
þ0.2 (c 0.06, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ0.2 (c 0.05, CH2Cl2); IR (CH2Cl2)nmax
3599, 2928, 2855, 1606, 1473 cmꢁ1; 1H- and 13C-NMR: see Tables 1
and 3; HREIMS m/z 482.3632, [M þ H]þ (calcd for C31H48NO3,
482.3629).
3599, 2929, 1606 cmꢁ1 1H- and 13C-NMR: see Tables S1 and S5;
;
HREIMS m/z 496.3783, [M þ H]þ (calcd for C32H50NO3, 496.3785).
4.2.14. (R)-2-(3,3-diethoxypropyl)-5,6,8-trimethyl-8-((3E,7E)-
4,8,12-trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-
hexahydrochromeno[5,6-e] [1,3]oxazine (14)
4.2.11. 2-((R)-5,7-dimethyl-7-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-8,9-dihydrochromeno[7,6-e] [1,3]oxazin-
3(2H,4H,7H)-yl)ethanol (11)
Yellow viscous oil, ½a D25
ꢀ þ21.5 (c 0.0011, CH2Cl2); IR (CH2Cl2)nmax
a 25
Yellow viscous oil, ½ ꢀD þ24.0 (c 0.018, CH2Cl2); IR (CH2Cl2)nmax
3599, 2922, 2853, 1606, 1374, 1067 cmꢁ1 1H- and 13C-NMR: see
;
3599, 2996, 1606, 1292, 1241 cmꢁ1; 1H- and 13C-NMR: see Tables 1
Tables S2 and S5; HREIMS m/z 582.4539, [M þ H]þ (calcd for
C37H60NO4, 582.4517).
Table 6
The IC50 values of the most active antimigratory tocotrienol analogs against the
highly metastatic human breast cancer cell line MDA-MB-231 in wound-healing
assay.a
Table 7
Antiproliferative activity (IC50
, mM) of selected active tocotrienol analogs against the
normal human non-tumorigenic MCF10A mammary epithelial cell line.a
Compound
IC50
(
m
M)
Compound
MCF10A
Selectivity indexb
1
2
>40.0
>40.0
1.3
8.1
6.0
4.5
2.4
10.7
2.2
1.5
1
4.4
6.5
8.9
4.6
0.7
8.5
1.9
4.2
2.5
2.0
1.7
1.10
3.61
5.24
6.57
0.39
5.67
2.09
2.34
1.92
1.43
2.15
13
17
18
23
26
27
36
38
39
40
10
12
13
26
27
35
36
38
39
40
42
4.7
4.9
7.2
a
The IC50 values are calculated by non-linear regression fitting using GraphPad
Prism 5.04.
a
b
The IC50 values are calculated by non-linear regression fitting using GraphPad
Values are calculated by dividing the IC50 value of each compound against
MCF10A normal cells over its IC50 value against þSA tumor cells.
Prism 5.04. The experiment was conducted in triplicate.

F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
337
4.2.20. 5-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
Vehicle
2(1H,3H,8H)-yl)pentanoic acid (20)
Yellow viscous oil, ½a D25
ꢀ þ0.2 (c 0.05, CH2Cl2); IR (CH2Cl2)nmax
3
γT
30
3676, 3599, 2927, 2855, 1748, 1606, 1473 cmꢁ1 1H- and 13C-NMR:
;
see Tables S2 and S6; HREIMS m/z 538.3873, [M þ H]þ (calcd for
C34H52NO4, 538.3891).
4.2.21. 6-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)hexanoic acid (21)
*
*
Yellow viscous oil, ½a D25
ꢀ þ0.5 (c 0.063, CH2Cl2); IR (CH2Cl2)nmax
*
3684, 3600, 2930, 2856, 1748, 1606, 1458, 1093 cmꢁ1; 1H- and 13C-
NMR: see Tables S2 and S6; HREIMS m/z 566.4198, [M þ H]þ (calcd
for C36H56NO4, 566.4204).
1
3
5
7
9
11
13
Days
4.2.22. 6-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)hexanoic acid (22)
Fig. 6. Effect of daily iv treatment of 0.5 mg of 1 (
g
T3) and 30 on þSA mammary tumor
growth in intact female syngeneic BALB/c mice throughout a 14-day experimental
period. Treatments were initiated around 40 days after 1ꢂ 106 þSA mammary tumor
cells were injected into the #4 mammary gland fat pad, when tumors became palpable
at approximately 3e5 mm in diameter, and continued for the next 14 days. Data points
indicate the average tumor volume (cm3 ꢃ SEM) for 14e15 mice/treatment group.
*P < 0.05, as compared with the vehicle-treated control group.
Yellow viscous oil, ½a D25
ꢀ þ0.2 (c 0.05, CH2Cl2); IR (CH2Cl2)nmax
3676, 3598, 2922, 2855, 1748, 1606, 1473, 1096 cmꢁ1 1H- and 13C-
;
NMR: see Tables S3 and S6; HREIMS m/z 552.4043, [M þ H]þ (calcd
for C35H54NO4, 552.4047).
4.2.23. (R)-2-(3,4-dimethoxyphenethyl)-5,6,8-trimethyl-8-
((3E,7E)-4,8,12-trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-
hexahydrochromeno[5,6-e] [1,3]oxazine (23)
4.2.15. (R)-2-(3,3-diethoxypropyl)-6,8-dimethyl-8-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno
[5,6-e] [1,3]oxazine (15)
Yellow viscous oil, ½a D25
ꢀ þ0.1 (c 0.036, CH2Cl2); IR (CH2Cl2)nmax
3683, 3600, 2921, 2855, 1606, 1516, 1085 cmꢁ1; 1H- and 13C-NMR: see
Tables 1 and 3; HREIMS m/z 616.4366, [M þ H]þ (calcd for C40H58NO4,
616.4360).
Yellow viscous oil, ½a D25
ꢀ þ67.3 (c 0.0012, CH2Cl2); IR (CH2Cl2)nmax
3599, 2922, 2854, 1606, 1374, 1055 cmꢁ1 1H- and 13C-NMR: see
;
Tables S2 and S5; HREIMS m/z 568.4353, [M þ H]þ (calcd for
C36H58NO4, 568.4360).
4.2.24. (R)-2-(3,4-dimethoxyphenethyl)-6,8-dimethyl-8-((3E,7E)-
4,8,12-trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-
hexahydrochromeno[5,6-e] [1,3]oxazine (24)
4.2.16. (R)-5,6,8-trimethyl-2-(piperidin-4-ylmethyl)-8-((3E,7E)-
4,8,12-trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-
hexahydrochromeno[5,6-e] [1,3]oxazine (16)
Yellow viscous oil, ½a D25
ꢀ þ0.3 (c 0.083, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ0.6 (c 0.02, CH2Cl2); IR (CH2Cl2)nmax
3682, 3600, 2917, 2854, 1606, 1516, 1029 cmꢁ1 1H- and 13C-NMR:
;
3666, 2922, 2854, 1606, 1521 cmꢁ1; 1H- and 13C-NMR: see Tables S2
and S6; HREIMS m/z 549.4392, [M þ H]þ (calcd for C36H57N2O2,
549.4420).
see Tables S3 and S6; HREIMS m/z 602.4192, [M þ H]þ (calcd for
C39H56NO4, 602.4204).
4.2.25. (R)-3-(3,4-dimethoxyphenethyl)-5,7-dimethyl-7-((3E,7E)-
4,8,12-trimethyltrideca-3,7,11-trienyl)-2,3,4,7,8,9-
hexahydrochromeno[7,6-e] [1,3]oxazine (25)
4.2.17. 5-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)pentan-1-ol (17)
Yellow viscous oil, ½a D25
ꢀ þ0.3 (c 0.013, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ18.4 (c 0.0012, CH2Cl2); IR (CH2Cl2)nmax
3676, 2922, 2854, 1606, 1521, 1068, 953 cmꢁ1 1H- and 13C-NMR:
;
3682, 3600, 2929, 2855, 1606, 1378, 1236, 1096 cmꢁ1; 1H- and 13C-
NMR: see Tables S2 and S6; HREIMS m/z 538.4253, [M þ H]þ (calcd
for C35H56NO3, 538.4255).
see Tables S3 and S6; HREIMS m/z 602.4204, [M þ H]þ (calcd for
C39H56NO4, 602.4204).
4.2.26. 4-(2-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]
oxazin-2(1H,3H,8H)-yl)ethyl)phenol (26)
4.2.18. 5-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)pentan-1-ol (18)
Yellow viscous oil, ½a D25
ꢀ þ0.6 (c 0.026, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ0.2 (c 0.046, CH2Cl2); IR (CH2Cl2)nmax
3676, 3656, 2922, 2855, 1606, 1516, 1083 cmꢁ1 1H- and 13C-NMR:
;
3679, 3600, 2929, 2855, 1606, 1473 cmꢁ1 1H- and 13C-NMR: see
;
see Tables 1 and 3; HREIMS m/z 572.4090, [M þ H]þ (calcd for
Tables S2 and S6; HREIMS m/z 524.4109, [M þ H]þ (calcd for
C38H54NO3, 572.4098).
C34H54NO3, 524.4098).
4.2.27. 4-(2-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)ethyl)phenol (27)
4.2.19. 5-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)pentanoic acid (19)
Yellow viscous oil, ½a D25
ꢀ þ2.0 (c 0.033, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ0.2 (c 0.04, CH2Cl2); IR (CH2Cl2)nmax
3688, 3599, 2932, 2850, 1606, 1516, 1172 cmꢁ1 1H- and 13C-NMR:
;
3683, 3601, 2927, 2854, 1750, 1606, 1465 cmꢁ1 1H- and 13C-NMR:
;
see Tables S3 and S7; HREIMS m/z 558.3925, [M þ H]þ (calcd for
see Tables S2 and S6; HREIMS m/z 552.4033, [M þ H]þ (calcd for
C37H52NO3, 558.3942).
C35H54NO4, 552.4047).

338
F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
4.2.28. (R)-5,6,8-trimethyl-2-(3-nitrobenzyl)-8-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno
[5,6-e] [1,3]oxazine (28)
4.2.36. 10-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)decan-1-ol (41)
Yellow viscous oil, ½a D25
ꢀ
þ22.5 (c 0.00067, CH2Cl2); IR (CH2Cl2)
Yellow viscous oil, ½a D25
ꢀ þ10.0 (c 0.0008, CH2Cl2); IR (CH2Cl2)nmax
nmax 3656, 2922, 2854, 1606, 1531, 1083 cmꢁ1
;
1H- and 13C-NMR:
3615, 2928, 1463, 1378, 1098, 950, 846 cmꢁ1; 1H- and 13C-NMR: see
Tables S4 and S8; HREIMS m/z 608.5064, [M þ H]þ (calcd for
C40H66NO3, 608.5037).
see Tables S3 and S7; HREIMS m/z 587.3826, [M þ H]þ (calcd for
C37H51N2O4, 587.3843).
4.2.37. 10-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)decan-1-ol (42)
4.2.29. (R)-6,8-dimethyl-2-(3-nitrobenzyl)-8-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)-1,2,3,8,9,10-hexahydrochromeno
[5,6-e] [1,3]oxazine (29)
Yellow viscous oil, ½a D25
ꢀ þ7.7 (c 0.0056, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ2.8 (c 0.013, CH2Cl2); IR (CH2Cl2)nmax
3613, 2927, 1683, 1472, 1378, 1046, 860 cmꢁ1; 1H- and 13C-NMR: see
Tables S4 and S8; HREIMS m/z 594.4893, [M þ H]þ (calcd for
C39H64NO3, 594.4881).
3676, 3599, 2922, 2854, 1606, 1540, 1374, 1084 cmꢁ1; 1H- and 13C-
NMR: see Tables S3 and S7; HREIMS m/z 573.3669, [M þ H]þ (calcd
for C36H49N2O4, 573.3687).
4.2.38. 12-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)dodecan-1-ol (43)
4.2.30. 4-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)butan-1-ol (35)
Yellow viscous oil, ½a D25
ꢀ þ0.5 (c 0.00213, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
ꢀ þ 8.4 (c 0.00166, CH2Cl2); IR (CH2Cl2)nmax
3614, 3427, 2928, 1463, 1379, 1098, 950, cmꢁ1 1H- and 13C-NMR:
;
3434, 2926,1641,1464,1377,1094, 954 cmꢁ1; 1H- and 13C-NMR: see
Tables S3 and S7; HREIMS m/z 524.4095, [M þ H]þ (calcd for
C34H54NO3, 524.4098).
see Tables S4 and S8; HREIMS m/z 636.5361, [M þ H]þ (calcd for
C42H70NO3, 636.5350).
4.2.39. 12-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)dodecan-1-ol (44)
4.2.31. 4-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)butan-1-ol (36)
Yellow viscous oil, ½a D25
ꢀ þ7.7 (c 0.00246, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
þ 10.6 (c 0.01133, CH2Cl2); IR (CH2Cl2)
ꢀ
3616, 2909, 1473, 1379, 1073, 1043, 940, 859 cmꢁ1 1H- and 13C-
;
nmax 3210, 2923, 1473, 1379, 1058, 859 cmꢁ1; 1H- and 13C-NMR: see
Tables S3 and S7; HREIMS m/z 510.3936, [M þ H]þ (calcd for
C33H52NO3, 510.3941).
NMR: see Tables S4 and S8; HREIMS m/z 622.5181, [M þ H]þ (calcd
for C41H68NO3, 622.5194).
4.3. General procedure for the preparation of the Mannich bases
(30 and 31)
4.2.32. 6-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)hexan-1-ol (37)
Each of compounds 23 and 26 was individually dissolved in 75%
aqueous methanol and left for 48 h at room temperature with
shaking (125 rpm). The solution was concentrated in vacuum and
then extracted with dichloromethane (3 ꢂ 100 mL). The organic
solvent was then evaporated under vacuum and the residue puri-
fied using column chromatography. The yield of hydrolysis was 15
and 20%, respectively (Scheme 1c).
Yellow viscous oil, ½a D25
ꢀ þ8.0 (c 0.0114, CH2Cl2); IR (CH2Cl2)nmax
3617, 2924, 1465, 1378, 1095, 950 cmꢁ1 1H- and 13C-NMR: see
;
Tables S4 and S7; HREIMS m/z 552.4410, [M þ H]þ (calcd for
C36H58NO3, 552.4411).
4.2.33. 6-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)hexan-1-ol (38)
4.3.1. (R)-5-((3,4-dimethoxyphenethylamino)methyl)-2,7,8-
trimethyl-2-((3E,7E)-4,8,12-trimethyltrideca-3,7,11-trienyl)
chroman-6-ol (30)
Yellow viscous oil, ½a D25
ꢀ þ12.6 (c 0.019, CH2Cl2); IR (CH2Cl2)nmax
3616, 2916, 1473, 1379, 1070, 1046, 939, 859 cmꢁ1 1H- and 13C-
;
Yellow viscous oil, ½a D25
ꢀ þ6.0 (c 0.0013, CH2Cl2); IR (CH2Cl2)nmax
NMR: see Tables S4 and S8; HREIMS m/z 538.4252, [M þ H]þ (calcd
3623, 3158, 2923, 1591, 1514, 1381, 1094, 855 cmꢁ1 1H- and 13C-
;
for C35H56NO3, 538.4254).
NMR: see Tables 2 and 3; HREIMS m/z 604.4371, [M þ H]þ (calcd for
C39H58N O4, 604.4360).
4.2.34. 8-((R)-5,6,8-trimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)octan-1-ol (39)
4.3.2. (R)-5-((4-hydroxyphenethylamino)methyl)-2,7,8-trimethyl-
2-((3E,7E)-4,8,12-trimethyltrideca-3,7,11-trienyl)chroman-6-ol (31)
Yellow viscous oil, ½a D25
ꢀ þ7.5 (c 0.013, CH2Cl2); IR (CH2Cl2)nmax
Yellow viscous oil, ½a D25
þ6.7 (c 0.0127, CH2Cl2); IR (CH2Cl2)nmax
ꢀ
3617, 2922, 1465, 1378, 1096, 950, 845 cmꢁ1; 1H- and 13C-NMR: see
Tables S4 and S8; HREIMS m/z 580.4718, [M þ H]þ (calcd for
C38H62NO3, 580.4724).
3582, 3322, 2917, 1612, 1513, 1379, 1173, 1095, 913 cmꢁ1 1H- and
;
13C-NMR: see Tables 2 and 3; HREIMS m/z 560.4108, [M þ H]þ
(calcd for C37H54NO3, 560.4098).
4.2.35. 8-((R)-6,8-dimethyl-8-((3E,7E)-4,8,12-trimethyltrideca-
3,7,11-trienyl)-9,10-dihydrochromeno[5,6-e] [1,3]oxazin-
2(1H,3H,8H)-yl)octan-1-ol (40)
4.4. General procedure for the preparation of 32e34
To a toluene solution [27] of each of tocotrienols 1 and 2
(25 mmol, 5 mL), acetic acid (0.5 mL), boric acid (40 mmol), and
paraformaldehyde (2.4 g) were added. The mixture was refluxed
overnight at rt. The reaction was stopped by cooling and washing
with cold water, followed by vigorous shaking with a Na2CO3
Yellow viscous oil, ½a D25
ꢀ þ7.7 (c 0.03223, CH2Cl2); IR (CH2Cl2)nmax
3616, 2909, 1473, 1379, 1073, 1043, 940, 859 cmꢁ1 1H- and 13C-
;
NMR: see Tables S4 and S8; HREIMS m/z 566.4561, [M þ H]þ (calcd
for C37H60NO3, 566.4567).

F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
339
aqueous solution (5% W/V) for 0.5 h, to decompose boric acid
complex. Finally, the mixture was washed with water, dried over
anhydrous Na2SO4, and concentrated under vacuum. The residue
was subjected to column chromatography using normal phase Si
gel 60 as a stationary phase using gradient n-hexane/ethyl acetate
system as a mobile phase (Scheme 1d).
prepared by plating various concentrations of cells, as determined
by hemocytometer, at the start of each experiment [35]. Non-linear
regression analysis of treatment effects on viable cell number was
used to determine the 50% growth inhibition concentration (IC50
for individual treatments using GraphPad Prism 5.04.
)
4.5.1.2. MDA-MB-231 and MCF7 human breast cancer cell lines.
The antiproliferative effects of tocotrienol analogs on the highly
metastatic MDA-MB-231 and MCF7 human cancer cell lines are
evaluated as previously described with some modifications [9].
MDA-MB-231 and MCF7 cells were plated at a density of 8 ꢂ 103
cells/well and 104 cell/well in 96-well culture plates, respectively.
The compounds were fed in serum-free media containing 5% and 1%
fetal bovine serum for MDA-MB-231 and MCF7 cells, respectively.
4.4.1. (R)-5-(hydroxymethyl)-2,7,8-trimethyl-2-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)chroman-6-ol (32)
Yellow viscous oil, ½a D25
ꢀ þ3.0 (c 0.05, CH2Cl2); IR (CH2Cl2)nmax
3590, 3385, 2927, 1620, 1470, 890 cmꢁ1 1H- and 13C-NMR: see
;
Tables 2 and 3; EIMS m/z ¼ 423.1.
4.4.2. (R)-5-(hydroxymethyl)-2,8-dimethyl-2-((3E,7E)-4,8,12-
trimethyltrideca-3,7,11-trienyl)chroman-6-ol (33)
Yellow viscous oil, ½a D25
ꢀ
þ2.6 (c 0.048, CH2Cl2); IR (CH2Cl2)nmax
4.5.1.3. MCF10A normal human non-tumorigenic mammary epithe-
lial cells. MCF10A (ATCC cat # CRL-10317) cells were maintained in
serum-free defined medium consisting of Dulbecco’s modified
Eagle’s medium (DMEM)/F12 containing 5% horse serum, 1% peni-
3595, 3393, 2922, 1607, 1469, 1378, 860 cmꢁ1
;
1H- and 13C-NMR:
see Tables 2 and 3; HREIMS m/z 425.3055, [M ꢁ H]þ (calcd for
C28H41O3, 425.3061).
cillin/streptomycin, 0.5
toxin, 10 g/mL insulin, and 20 ng/mL epidermal growth factor
(rhEGF). For subculturing, cells were rinsed twice with sterile Ca2þ
mg/mL hydrocortisone, 100 ng/mL cholera
4.4.3. (R)-7-(hydroxymethyl)-2,8-dimethyl-2-((3E,7E)-4,8,12-
m
trimethyltrideca-3,7,11-trienyl)chroman-6-ol (34)
-
Yellow viscous oil, ½a D25
ꢀ
þ0.9 (c 0.0066, CH2Cl2); IR (CH2Cl2)nmax
and Mg2þ-free phosphate buffered saline (PBS) and incubated in
0.25% trypsin containing 0.025% EDTA in PBS for 5 min at 37 ꢄC. The
released cells were centrifuged, resuspended in fresh media and
counted using hemocytometer. For testing, MCF10A cells were
plated at a density of 10 ꢂ 104 cells/well. A stock solution of the
compounds was prepared in DMSO. DMSO was added to all treat-
ment media such that the final DMSO concentration was the same
in all treatment groups within a given experiment and was always
less than 0.1%. The compounds were fed in serum-free media.
MCF10A mammary epithelial cell’s viable number was determined
by MTT colorimetric assay as previously described [9,34,35].
3588, 3393, 2924, 1467, 1100, 985 cmꢁ1
;
1H- and 13C-NMR: see
Tables 2 and 3; HREIMS m/z 425.3074, [M ꢁ H]þ (calcd for
C28H41O3, 425.3061).
4.5. Biological assays
4.5.1. In vivo activities
4.5.1.1. þSA cell line. All materials were purchased from Sigma
Chemical Company (St. Louis, MO) unless otherwise stated. The
highly malignant þSA mouse mammary epithelial cell lines were
serially passaged at subconfluent cell density. The þSA mammary
tumor cell line was derived from an adenocarcinoma that devel-
oped spontaneously in a BALB/c female mouse [33]. Cell culture
conditions have been previously described in detail [34].
Briefly, þSA cells were maintained in serum-free defined medium
consisting of Dulbecco’s modified Eagle’s medium (DMEM)/F12
containing 5 mg/mL bovine serum albumin (BSA), 10 mg/mL
transferrin, 100 U/mL soybean trypsin inhibitor, 100 U/mL penicillin
G, 0.1 mg/mL streptomycin, 10 mg/mL insulin, and 10 ng/mL
epidermal growth factor. For subculturing, cells were rinsed twice
with sterile Ca2þ and Mg2þ-free phosphate buffered saline (PBS)
and incubated in 0.05% trypsin containing 0.025% EDTA in PBS for
5 min at 37 ꢄC. The released cells were centrifuged, resuspended in
fresh media and counted using hemocytometer. A stock solution of
the compounds was prepared in DMSO. Once dissolved, this solu-
tion was added to a small volume of sterile 10% BSA in water and
incubated overnight at 37 ꢄC. This solution conjugated to BSA was
4.5.1.4. NCI-60 human tumor cell line panel. The human tumor cell
lines [36e38] of the cancer screening panel are grown in RPMI 1640
medium containing 5% fetal bovine serum and 2 mM
For a typical screening experiment, cells are inoculated into 96-well
microtiter plates in 100 L at plating densities ranging from 5000 to
L-glutamine.
m
40,000 cells/well depending on the doubling time of individual cell
lines. After cell inoculation, the microtiter plates are incubated at
37 ꢄC, 5% CO2, 95% air and 100% relative humidity for 24 h prior to
addition of experimental drugs. After 24 h, two plates of each cell
line are fixed in situ with TCA, to represent a measurement of the
cell population for each cell line at the time of the addition of the
drug (T0). Experimental drugs are solubilized in dimethyl sulfoxide
at 400-fold the desired final maximum test concentration and
stored frozen prior to use. At the time of drug addition, an aliquot of
frozen concentrate is thawed and diluted to twice the desired final
maximum test concentration with complete medium containing
used to prepare various concentrations (0e40
m
M) of tocotrienol
50
made to provide a total of five concentrations plus control. Aliquot
of 100 L of this drug dilution is added to the appropriate microtiter
wells already containing 100 L of medium, resulting in the
mg/mL gentamicin. Additional four, 10-fold serial dilutions are
analogs. DMSO was added to all treatment media such that the final
DMSO concentration was the same in all treatment groups within
a given experiment and was always less than 0.1%.
m
m
þSA mammary epithelial cell viable number was determined by
the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl tetrazolium bromide
(MTT) colorimetric assay as described previously [34,35]. Briefly, on
the assay day, treatment medium was replaced with fresh control
medium containing 0.42 mg/mL MTT, and the cells in 24-well
plates were incubated at 37 ꢄC for 4 h. The medium was
removed, and the MTT crystals were dissolved in isopropyl alcohol
(1 mL/well). The optical density of each sample was read at 570 nm
required final drug concentrations. Following drug addition, the
plates are incubated for an additional 48 h at 37 ꢄC, 5% CO2, 95% air,
and 100% relative humidity. For adherent cells, the assay is termi-
nated by the addition of cold TCA. Cells are fixed in situ by the
gentle addition of 50 mL of cold 50% (w/v) TCA (final concentration,
10% TCA) and incubated for 60 min at 4 ꢄC. The supernatant is
discarded, and the plates are washed five times with tap water and
air dried. Sulforhodamine B (SRB) solution (100 mL) at 0.4% (w/v) in
on
a
microplate reader (SpectraCount, Packard BioScience
1% acetic acid is added to each well, and plates are incubated for
10 min at room temperature. After staining, unbound dye is
removed by washing five times with 1% acetic acid and the plates
Company), against a blank prepared from cell-free cultures. The
number of cells/well was calculated against a standard curve

340
F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
are air dried. Bound stain is subsequently solubilized with 10 mM
trizma base, and the absorbance is read on an automated plate
reader at a wavelength of 515 nm. For suspension cells, the meth-
odology is the same except that the assay is terminated by fixing
approximately 0.5 mm in diameter. Tumor bearing mice were then
divided into the following treatment groups: (1) vehicle-treated
controls; (2) 0.5 mg of compound 1 in 0.1 mL of propylene glycol/
DMSO (1:1) vehicle; and (3) 0.5 mg of compound 30 in 0.1 mL of
propylene glycol/DMSO (1:1) vehicle. Treatments were adminis-
tered daily by tail vein injection throughout the entire 14 day
experimental period. Tumor size and body weighs were recorded
every other day. Average tumor diameter for each palpable tumor
was determined using the mean of the two largest perpendicular
diameters as measured with vernier calipers as previously
described [33,39e41]. Tumor volume was determined using the
equation:
settled cells at the bottom of the wells by gently adding 50 mL of 80%
TCA (final concentration, 16% TCA). GI50 is the concentration of test
drug where 100 ꢂ (T ꢁ T0)/(C ꢁ T0) ¼ 50. Where T is the optical
density of the test well after a 48-h period of exposure to test drug,
T0 is the optical density at time zero, and C is the control optical
density. The GI50 is a measure of the growth inhibitory power of the
test agent.
4.5.1.5. Wound-healing assay. The highly metastatic human [7]
MDA-MB-231 breast cancer cells were cultured in RPMI 1640
Volume ¼ Length ðcmÞ ꢂ Width2 ðcmÞ=2 or V ¼ L ꢂ W2=2
medium containing 10 mM HEPES, 4 mM
L-glutamine, 10% fetal
Statistical differences between treatment groups were deter-
mined using analysis of variance, followed by Duncan’s multiple
range test. A difference of p < 0.05 was considered to be significant,
as compared to the vehicle-treated control group.
bovine serum, penicillin (100 IU/mL), and streptomycin (50
m
g/mL),
and grown in a 5% CO2 atmosphere at 37 ꢄC. Cells were plated onto
sterile 24-well and allowed to recover for a confluent cell mono-
layer formed in each well (>95% confluence). Wounds were then
inflicted to each cell monolayer using a sterile 200 mL pipette tip.
Acknowledgments
Media were removed, cells were washed twice with PBS, and then
fresh media containing test compounds were added to each well.
Test compounds were prepared in DMSO at five non-toxic
concentrations, each in triplicate using DMSO as negative control.
The incubation was carried out for 24 h under low serum-content
conditions (0.5%), after which media was removed and cells were
fixed and stained using Diff Quick staining (Dade Behring Diag-
nostics, Aguada, Puerto Rico). The number of cells migrated on the
scratched wound were counted under the microscope in three or
more randomly selected fields (magnification: 40ꢂ). Final results
are expressed as mean ꢃ SEM per 40ꢂ field.
This publication was supported in part by First Tech Interna-
tional Ltd. (Hong Kong). The Developmental Therapeutics Program
e National Cancer Institute (NCI), Bethesda Maryland, is acknowl-
edged for the growth inhibitory screening assays of some
compounds.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.ejmech.2012.11.012.
4.5.2. In vitro activity
References
4.5.2.1. Tumor implantation into mammary fat pads of syngeneic
BALB/c mice. The malignant þSA mammary epithelial cell line was
derived from an adenocarcinoma that developed spontaneously in
a female BALB/c mouse [33,39]. These cells can be grown on plastic
in culture and display the ability to grow in soft agarose [33,39]. Re-
implantation of þSA mammary tumor cells back into the mammary
fat pad of syngeneic mice results in the rapid develop and growth of
metastatic tumors [33,39]. Prior to experimentation, þSA cells were
serially passaged at subconfluent cell density. For subculturing,
cells are rinsed with calcium/magnesium-free phosphate buffered
saline (PBS) and incubated in 0.05% trypsin containing 0.025% EDTA
in PBS for 5 min at 37ꢄ. The released cells are then diluted in DMEM/
F12 media and pelleted by centrifugation. Cell pellets are then
resuspended in control culture media, counted by hemocytometer,
and then plated at the desired density in culture plates. Cells were
then fed fresh media every other day. On the day of tumor
implantation, þSA cells grown in culture are isolated with trypsin,
washed, counted and diluted to the desired concentration in fresh
culture medium. Female BALB/c mice, 3e4 months old were
purchased from Harlan SpragueeDawley (Indianapolis, IN) and
housed in plastic cages in a temperature-regulated (24 ꢃ 0.5ꢄ) and
light-controlled (14 h light/10 h dark) room and allowed standard
laboratory mouse chow and water ad libitum. At the time of
experimentation, animals were anesthetized with an ip injection of
ketamine/xylazine (10 mg ketamine: 1 mg xylazine/mL saline,
Henry Schein, Inc, Melville, NY) at a dose of 0.1 mL/10 g bw. A small
incision was then made in the skin along the midline of the
abdomen, and 1 ꢂ 106 cells/0.1 mL culture medium was injected
into the #4 mammary gland fat pad (1 tumor/mouse). The incision
was then closed, animals allowed to recover, and then returned to
their treatment group cages. Approximately 40 days after trans-
plantation, tumors became palpable at an average size of
[1] W.G. Burton, Proc. Nutr. Soc. 53 (1994) 251e262.
[2] R. Eitenmiller, J. Lee, Vitamin E: Food Chemistry, Composition, and Analysis,
Marcel Dekker Inc., New York, Basil, 2004.
[3] A. Kamal-Eldin, L. Appelqvist, Lipids 31 (1996) 671e701.
[4] L. Packer, S.U. Weber, G. Rimbach, J. Nutr. 131 (2001) 369Se373S.
[5] C.K. Sen, S. Khanna, S. Roy, Life Sci. 78 (2006) 2088e2098.
[6] J.P. Kamat, T.P.A. Devasagayam, Neurosci. Lett. 195 (1995) 179e182.
[7] F.A. Behery, A.Y. Elnagar, M.R. Akl, V.B. Wali, B. Abuasal, A. Kaddoumi,
P.W. Sylvester, K.A. El Sayed, Bioorg. Med. Chem. 18 (2010) 8066e8075.
[8] J. Neuzil, K. Kågedal, L. Andera, C. Weber, U.T. Brunk, Apoptosis 7 (2002) 177e185.
[9] A.Y. Elnagar, V.B. Wali, P.W. Sylvester, K.A. El Sayed, Bioorg. Med. Chem. 18
(2010) 755e768.
[10] J. Neuzil, M. Tomasetti, Y. Zhao, L.F. Dong, M. Birringer, X.F. Wang, P. Low,
K. Wu, B.A. Salvatore, S. Ralph, J. Mol. Pharmacol. 71 (2007) 1185e1199.
[11] Y. Yano, H. Satoh, K. Fukumoto, I. Kumadaki, T. Ichikawa, K. Yamada,
K. Hagiwara, T. Yano, Int. J. Cancer 115 (2005) 839e846.
[12] N. Akaho, J. Takata, T. Fukushima, K. Matsunaga, A. Hattori, R. Hidaka, K. Fukui,
M. Yoshida, T. Fujioka, Y. Karube, K. Imai, Drug Metab. Dispos. 35 (2007) 1502e1510.
[13] M.Birringer,J.H. EyTina,B.A.Salvatore,J.Neuzil,Br. J. Cancer88(2003)1948e1955.
[14] F.F. Blicke, Organic Reactions, John Wiley and Sons, Inc., New York, 1942.
[15] T. Nakamura, S. Kijima, Chem. Pharm. Bull. 19 (1971) 2318e2324.
[16] W.J. Burke, J. Am. Chem. Soc. 71 (1949) 609e612.
[17] W.J. Burke, R.P. Smith, C. Weatherbee, J. Am. Chem. Soc. 74 (1952) 602e605.
[18] M. Tramontini, Synthesis 12 (1973) 703e775.
[19] M. Tramontini, L. Angiolini, Tetrahedron 46 (1990) 1791e1837.
[20] D. Shi, S. Rong, G. Dou, M. Wang, J. Comb. Chem. 12 (2010) 25e30.
[21] J.B. Chylinska, T. Urbanski, M. Mordarski, J. Med. Chem. 6 (1963) 484e487.
[22] L. Benameur, Z. Bouaziz, P. Nebois, M. Bartoli, M. Boitard, H. Fillion, Chem.
Pharm. Bull. 44 (1996) 605e608.
[23] S. Wang, Y. Li, Y. Liu, A. Lu, Q. You, Bioorg. Med. Chem. Lett. 18 (2008) 4095e4097.
[24] B.P. Mathew, A. Kumar, S. Sharma, P.K. Shukla, M. Nath, Eur. J. Med. Chem. 45
(2010) 1502e1507.
[25] L. Cappannini, C. Cimarelli, S. Giuli, G. Palmieri, M. Petrini, Tetrahedron
Asymm. 18 (2007) 1022e1029.
[26] W.J. Burke, C.W. Stephens, J. Am. Chem. Soc. 74 (1952) 1518e1520.
[27] T. Nakamura, S. Kijima, Chem. Pharm. Bull. 20 (1972) 1681e1686.
[28] S. Ohnmacht, R. West, R. Simionescu, J. Atkinson, Magn. Reson. Chem. 46
(2008) 287e294.
[29] B.S. McIntyre, K.P. Briski, A. Gabor, P.W. Sylvester, Proc. Soc. Exp. Biol. Med.
224 (2000) 292e301.

F.A. Behery et al. / European Journal of Medicinal Chemistry 59 (2013) 329e341
341
[30] H.K. Liu, Q. Wang, Y. Li, W.G. Sun, J.R. Liu, Y.M. Yang, W.L. Xu, X.R. Sun,
B.Q. Chen, J. Nutr. Biochem. 21 (2010) 206e213.
[31] P.N. Chang, W.N. Yap, D.T. Lee, M.T. Ling, Y.C. Wong, Y.L. Yap, Nutr. Cancer 61
(2009) 357e366.
[32] W.N. Yap, P.N. Chang, H.Y. Han, D.T.W. Lee, M.T. Ling, Y.C. Wong, Y.L. Yap, Br. J.
Cancer 99 (2008) 1832e1841.
[33] L.W. Anderson, K.G. Danielson, H.L. Hosick, In Vitro 15 (1979) 841e843.
[34] V.B. Wali, P.W. Sylvester, Lipids 42 (2007) 1113e1123.
[35] P.W. Sylvester, H.P. Birkenfeld, H.L. Hosick, K.P. Briski, Exp. Cell. Res. 214
(1994) 145e153.
[36] M.R. Boyd, K.D. Paull, L.R. Rubinstein, in: F.A. Vleriote, T.H. Corbett, L.H. Baker
(Eds.), Cytotoxic Anticancer Drugs: Models and Concepts for Drug Discovery
and Development, Kluwer Academic, Hingham, MA, 1992, pp. 11e34.
[37] M.R. Grever, S.A. Sehepartz, B.A. Chabners, Semin. Oncol. 19 (1992) 622e638.
[38] A. Monks, D. Schudiero, P. Skehan, R. Shoemaker, K. Paull, D. Vistica, C. Hose,
J. Langley, P. Cronise, A. Vaigro-Wolff, M. Gray-Goodrich, H. Campbell, J. Mayo,
M. Boyd, J. Natl. Cancer Inst. 83 (1991) 757e766.
[39] K.G. Danielson, L.W. Anderson, H.L. Hosick, Cancer Res. 40 (1980) 1812e1819.
[40] P.W. Sylvester, M. Russell, M.M. Ip, C. Ip, Cancer Res. 46 (1986) 757e762.
[41] P.W. Sylvester, C. Ip, M.M. Ip, Cancer Res. 46 (1986) 763e769.