Tocotrienamines and tocopheramines: Reactions with radicals and metal ions

Article abstract of DOI:10.1016/j.bmc.2011.08.050

The antioxidant activity of vitamin E (VE) homologs a, c and d-tocotrienamines (4b-6b), never studied before, and a, c and d-tocopheramines (4a-7a) was investigated by means of different total antioxidant capacity (TAC) tests. In all the test model systems, compounds 4a-7a and 4b-6b showed similar or higher TAC values than the parental vitamin E forms and their physiological metabolites. α-Homologs of VE amines showed markedly higher activity than the VE congeners in the TEAC test, which is tailored for liposoluble antioxidants, while c-homologs of the amine analogs showed higher activity in the FRAP tests. Kinetics analysis of the reaction with DPPH showed higher second order rate k for 4a than for α-tocopherol (1a). α-Tocopherolquinone 1f was the common main oxidation product for both 1a and α-tocopheramine (4a) exposed to ferric ions or DPPH, and the implied oxidative deamination of 4a was accompanied by a nitration reaction of phenolic substrates that were added to the reaction medium. Possible mechanisms of these reactions were studied.

Full text of DOI:10.1016/j.bmc.2011.08.050

Bioorganic & Medicinal Chemistry 19 (2011) 6483–6491
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Tocotrienamines and tocopheramines: Reactions with radicals and metal ions
Francesco Galli a, , Francesco Mazzini b, Luca Bamonti a, Lars Gille c, Stefan Böhmdorfer d, Marta Piroddi a,
⇑
Thomas Netscher e, Frank J. Kelly f, Thomas Rosenau d
a Department of Internal Medicine, Laboratory of Clinical and Nutritional Biochemistry, University of Perugia, Perugia 06126, Italy
b Department of Chemistry and Industrial Chemistry, University of Pisa, Pisa 56126, Italy
c Institute of Pharmacology and Toxicology, Department of Biomedical Sciences, University of Veterinary Medicine Vienna, Vienna A-1210, Austria
d Department of Chemistry, University of Natural Resources and Life Sciences, Vienna A-1190, Austria
e Research and Development, DSM Nutritional Products, Basel PO Box 2676, CH-4002, Switzerland
f MRC-HPA Centre for Environmental and Health, King’s College London, London SE1 9NH, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 July 2011
Revised 22 August 2011
Accepted 23 August 2011
Available online 28 August 2011
The antioxidant activity of vitamin E (VE) homologs a, c and d-tocotrienamines (4b–6b), never studied
before, and
capacity (TAC) tests. In all the test model systems, compounds 4a–7a and 4b–6b showed similar or higher
TAC values than the parental vitamin E forms and their physiological metabolites. -Homologs of VE
amines showed markedly higher activity than the VE congeners in the TEAC test, which is tailored for lip-
osoluble antioxidants, while -homologs of the amine analogs showed higher activity in the FRAP tests.
Kinetics analysis of the reaction with DPPHÅ showed higher second order rate k for 4a than for
-tocoph-
-tocopher-
a, c and d-tocopheramines (4a–7a) was investigated by means of different total antioxidant
a
c
Keywords:
Vitamin E
Tocopherol
Tocopheramine
Nitric oxide
Nitration
a
erol (1a). a-Tocopherolquinone 1f was the common main oxidation product for both 1a and a
amine (4a) exposed to ferric ions or DPPHÅ, and the implied oxidative deamination of 4a was
accompanied by a nitration reaction of phenolic substrates that were added to the reaction medium. Pos-
sible mechanisms of these reactions were studied.
Free radicals
Transition metals
Oxidation
Ó 2011 Elsevier Ltd. All rights reserved.
Antioxidants
1. Introduction
d-tocopheramine was compared in different model systems,6
showing a superior capability of tocopheramines in the quenching
Vitamin E (VE) is a term that encompasses a group of fat soluble
vitamers (four tocopherols and four tocotrienols), derived from the
structure of 6-chromanol,1 which exhibit the biological activity of
of singlet oxygen, while tocopherols were more reactive towards
phenoxyl radicals.
Recently, we reported a very simple route for the preparation of
all the VE amines in stereoisomerically pure form, to be used as pre-
cursors of VE amide analogs.7 According to this original procedure,
a
-tocopherol 1a (Fig. 1). In 1942, Smith et al. reported that the
amino analog of racemic 1a, racemic -tocopheramine 4a, has
approximately the same vitamin and antioxidant properties as 1a
itself.2 Later on, also (all-rac)-N-methyl-b- and
-tocopheramine
a
a-, c- and d-tocotrienamine (4b, 5b, 6b, respectively, Fig. 1) have
c
been synthesized for the first time. These VE analogs have been
recently demonstrated to show in vitro and in vivo anticancer
activity.8,9
were reported to show vitamin activity close to 1a.3,4 Using an
isotope dilution method, it was proved that their vitamin activity
was not due to a conversion of the amine analogs into 1a, and that
those analogs shared the same metabolic pathway with the
corresponding VE forms, providing carboxyethyl hydroxychromans
(CEHCs) and tocopherolquinone as metabolites.5 More recently,
With the present study, we comparatively investigated the anti-
oxidant activity of all these amine analogs. Based on the recent
work by Muller et al.10 three electron-transfer assays were used
to comparatively evaluate antioxidant capacity of VE and VE amine
compounds, namely the Trolox equivalent antioxidant capacity
(TEAC) assay, the 2,2-diphenyl-1-picrylhydrazyl (DPPHÅ) assay
and the ferric ion reducing antioxidant parameter (FRAP). In addi-
tion, we carried out a preliminary study on the oxidation of 4a that
disclosed an interesting nitrating activity of electron rich aromatic
amines under oxidative conditions.
the antioxidant behavior of
a- to d-tocopherol and a- to
⇑
Corresponding author. Address: Department of Internal Medicine, Laboratory of
Clinical and Nutritional Biochemistry, University of Perugia, Via del Giochetto,
06126 Perugia, Italy. Tel.: +39 075 585 7445; fax: +39 075 585 7441.
E-mail address: f.galli@unipg.it (F. Galli).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.08.050

6484
F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
Figure 1. Structures of vitamin E compounds assayed and a-tocopherolquinone.
Although the three tests provided different reducing activities,
which is expected because of chemical differences of the assay sys-
tems,10,14 all VE amines investigated showed similar or higher
antioxidant activity than their tocopherol and tocotrienol
counterparts.
In the copper-mediated ascorbic acid oxidation test, a fairly
concentration-dependent antioxidant effect (lower oxidation rate
of ascorbic acid) was demonstrated for all the tocopheramine
derivatives following the order: 6a > 5a > 4a. Baseline oxidation
rate for the copper-induced ascorbic acid depletion test (calculated
using the vehicle ethanol as sample) was 11.4 0.9 nM sꢀ1. A con-
siderable decrease of the oxidation rate was observed for VE
2. Results and discussion
2.1. Antioxidant tests
The TEAC assay is based on the reduction of the radical
cation from 2,20-azinobis-(3-ethylbenzothiazoline)-6-sulfonic acid
(ABTSÅ+). This is a mixed electron-transfer and hydrogen atom-
transfer assay that has been described to be more suitable than
other TAC probes in assessing fat soluble antioxidants.11 The
results of this assay expressed as Trolox equivalents (Fig. 2, upper
panel) showed that VE amines have higher antioxidant power than
all the other VE compounds assessed in this study. Among the
amines when applied at concentrations P10
was significant vs 1a only at the final concentration of 100
with percentage of reduction of 19.5 8.3, 35.0 4.9 and
44.7 3.2 for the homologs 4a, 5a and 6a, respectively (p <0.05
for 5a and 6a).
Altogether these results demonstrate that VE amines can be a
valuable alternative to natural forms of VE when used as antioxi-
dants in different model systems based on radical- or metal-depen-
dent oxidation reactions. Although this evidence was suggested in
earlier studies [see6 and references therein], here we show for the
first time a semi-quantitative comparison of antioxidant power of
all the series of VE amine analogs, including tocotrienamines, never
investigated before.
l
M, and this effect
amine compounds, antioxidant data of the
a
-homolog was higher
l
M,
when compared to
c
- and d-homologs, in agreement with previous
results,10 indicating that the degree of methylation as well as the
character of the side chain made little difference in determining
the antioxidant capacity of tocopherols and tocotrienols in this
assay.
a
The FRAP test is a pure electron-transfer assay that determines
the capability of antioxidant compounds to reduce ferric ions that
are detected by the formation of a complex with the probe
di-2,4,6-tripyridyl-s-triazine (TPTZ).12,13 Differences among the
assayed compounds were negligible in this test except for the
c-homolog of VE amines that showed appreciably higher reduction
capacity (Fig. 2, middle panel).
Reduction of DPPHÅ, a stable organic nitrogen radical, (Fig. 2, low-
er panel) showed that antioxidant activity of both VE and VE amines
paralleled the degree of methylation of the chroman ring with the
2.2. Stopped-flow analysis of DPPHÅ reaction with 1a and 4a
order of magnitude
a > c > d, independently of structural features
Besides steady-state tests, the reaction kinetics was investi-
gated by stopped flow spectroscopy (Fig. 3). The reaction of DPPHÅ
with tocopheramines/tocotrienamines and 1a was studied in
ethanol. By means of the kinetic isolation technique (10- to 20-fold
of the side chain and without differences between VE and VE amines.
This parallels the expected electron donating capability of chroma-
nols even if hydrogen transfer could be also involved in this test.14

F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
6485
identical rate constants. Related to the
a
-congeners of VE amines,
90
TEAC
respectively, the hypomethylated derivatives showed decreased
rate constants, but still higher rates than 1a. This in line with the
observation of Mukai et al.16 that the rate constants for tocopherol
congeners in the reaction with an aroxyl radical increase with the
degree of methylation.
*
50
40
30
20
10
0
*
*
*
*
*
2.3. Iron- and DPPHÅ-mediated deamination of VE amines
2
d
2
2
2
2
T3 T3 T3 H
TH TH TH
ci
H
-
-
-
d
-
-
-
a
d
a
g
NH
At first sight, those similarities in the antioxidant behavior of VE
and VE amines may suggest that they share the same mechanism
of action. Indeed, their structures are very closely related, and they
form the same oxidation end product, the tocopherolquinone.2
DPPHÅ-mediated and ferric ion-mediated oxidation of 1a and 4a
were investigated in order to verify whether these vitamers form
gDPPH
TN
-
C
3NH
TN T
-
d
a
-
CEHCCEHC
90
80
70
60
50
40
30
20
10
0
T3NTH
O
T
-
-
-
a
-
-
g
g
d
g
*
a
*
*
*
a-tocopherolquinone (1f, Fig. 1) as oxidation product in both reac-
*
tion systems.
2
2
2
T3
d
2
i
T3 T3 T3 H
TH TH TH
c
H
H2 H
-
-
-
d
-
-
-
a
d
NH
a
g
2.3.1. Reaction with DPPHÅ
gFRAP
a
TN
-
C
TN TN
-
CEHC
-
T3N
-
-
a
-
g
d
d
TO
-
Tocopherolquinone 1f was produced during the DPPHÅ-medi-
ated oxidation of 1a or 4a (1:1 molar ratio, Fig. 4), in similar yields.
GC–MS analysis was used to confirm the conversion of 4a to 1f in
this reaction (not shown) and in the Fe-induced oxidation reaction
(Fig. 7, right panel). In order to detect possible reaction products of
DPPHÅ, the mixtures of DPPHÅ (0.5 mM) with 1a/4a (0.5 mM) were
subjected to HPLC analysis after 20 min reaction (Fig. 5). While the
formation of nitrated DPPHÅ or DPPH-H was not directly observed,
the HPLC analysis revealed a product (retention time 4 min), which
was only observed after stoichiometric reduction of DPPHÅ with 4a
but not with 1a. A LC/MS analysis gave MS peaks at 228 and 263 m/
z. So far these data suggest a breakdown product of DPPHÅ (such as
a nitrated diphenyl hydrazine derivative or trinitroaniline), which
arises only in the presence of 4a.
90
d
g
*
a
70
60
50
40
30
20
10
0
*
*
*
2b 3b
a
c
6b 1c 2c
3
1a 2a 3a 1b
4a
5a
5b
6
4b
1g
Figure 2. Steady-state electron-transfer capacity of VE amines and their corre-
sponding VE forms. Compounds 1a–6a, 1b–6b, 1c–2c and 1g were investigated
with three total antioxidant capacity tests (from top to bottom left panels): the
trolox equivalent antioxidant capacity (TEAC), 2,2-diphenyl-1-picrylhydrazyl
(DPPHÅ) bleaching and the ferric reducing ability (FRAP) test. Further details on
⁄
these analyzes methods are reported in the text. p <0.05 with respect to the
corresponding form in the tocopherol or tocotrienol series; 1c–2c and 1g were
This raised the question which intermediates of tocopheramine
oxidation could be observed. Attempts to detect and identify ami-
nyl radicals of 4a at room temperature failed so far. However, it
was possible to show that upon photooxidation of the N-methyl-
compared with tocopherols.
excess of the antioxidant over DPPHÅ) directly the second order rate
constants (L molꢀ1 sꢀ1) were obtained (Fig. 3).
ated a-tocopheramine model compound 1d (Fig. 1) in the presence
of oxygen, a fairly stable radical could be detected by ESR spectros-
copy (Fig. 6 left panel). Simulation of the spectrum based on a hyp-
othetic nitroxyl radical agrees with the experimental spectrum.
The splitting pattern for this nitroxyl radical of 1d shows similari-
ties to the spectrum obtained by Murphy et al.17 during oxidation
The rate constant obtained for 1a was 280 22 L molꢀ1 sꢀ1. In
contrast tocopheramines/tocotrienamines exhibited much higher
rate constants, such as 4287 1371 L molꢀ1 sꢀ1 for 4a. The ob-
tained rate constants for 1a were in the range of data
(245 15 L molꢀ1 sꢀ1) obtained in a previous study.15 For toco-
pheramines and tocotrienamines so far no rate constants with
DPPHÅ were available. The permethylated 4a and 4b showed nearly
of N-methyl-c-tocopheramine.
Since time resolution of ESR spectroscopy was too low to detect
similar intermediates also for 4a during oxidation with DPPHÅ and
other oxidants, possible intermediates were studied by optical
stopped flow spectroscopy. Figure 6 (right panel) shows UV/vis
spectroscopy profiles obtained between 300 and 650 nm of the
products formed during the reaction between DPPHÅ radical and
1a and 4a in acetonitrile. The results are consistent with the forma-
tion of the chromanoxyl radical from 1a (kmax at 428 nm; Fig. 6,
right upper panel) in agreement with previous findings.15 During
the reaction of 4a with DPPHÅ a reaction intermediate of 4a with
kmax at 474 nm (Fig. 6, right lower panel) was observed in the time
windows of 100–200 ms after mixing. Although UV/vis absorption
is not suitable for proper identification of the intermediate, it is
interesting in this context that nitroxyl radicals of other com-
pounds have an absorption maximum between 400 and
500 nm.18 These data suggest that intermediates of tocopheram-
ines with higher oxidation states of nitrogen are possibly involved
in their oxidation by DPPHÅ and other oxidants.
10000
1000
100
Figure 3. Second order rate constants for the reaction of 1a and selected
tocopheramines/tocotrienamines with DPPHÅ. The rate constants were obtained
from the decay of the DPPHÅ absorption at 518 nm in ethanol in the presence of an
at least 10-fold excess of the antioxidant over DPPHÅ. Data represent mean
values SD of at least five experiments.
2.3.2. Reaction with ferric iron
In the case of 1a, the ferric ion-catalyzed reaction proceeded to
completion very rapidly producing 1f in the presence of 1:1 molar

6486
F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
Figure 4. HPLC-ECD analysis of the reaction between FeCl3 (top)/DPPHÅ (bottom) and 1a and 4a. (right) HPLC-ECD analysis of 1a and 4a oxidation products formed during the
reaction with DPPHÅ. Peak identity was confirmed using authentic external standards and by GC–MS analysis that was carried out as described in the text.
DPPH-H
50000
40000
30000
20000
10000
0
0.25
0.20
0.15
0.10
0.05
0.00
DPPH
4a
DPPH
DPPH + -TNH2
1a
DPPH + -TOH
P 4 min
2
3
4
5
6
7
8
9
10
11
Retention Time [min]
Figure 5. HPLC analysis of the DPPHÅ related reaction products obtained from the reaction of DPPHÅÅ with 1a/4a. (left) Chromatogram obtained by UV detection at 304 nm of
DPPHÅ (starting compound), DPPHÅ + 1a and DPPHÅ + 4a. In addition, the major reduction product (DPPH-H) and a unknown DPPHÅ reaction product (P 4 min) are indicated.
(right) Peak area of the reaction product eluting at 4 min in different reagents and the reaction mixtures of DPPHÅ + 1a and DPPHÅ + 4a. Data represent mean values SD of five
experiments.
ratio (or higher, i.e., 1:4 and 1:40) between the vitamer and the
oxidant (Figs. 4 and 7). Compound 1f was also obtained in the oxi-
dation of 4a, but only with a partial conversion at 1:40 4a to FeCl3
molar ratio (approx. 45% conversion as estimated by HPLC-ECD,
Fig. 4, bottom left panel), while a molar ratio of 1:4 was not suffi-
cient to form 1f from 4a. This yield of 1f lower than that obtained
with 1a is in apparent contrast with FRAP data (Fig. 2, lower panel)
where 4a and 1a produced the same extent of one-electron reduc-
tion. This can be possibly explained by the different reaction con-
ditions of these two experiments. While 1f formation (Fig. 7) was
assessed in unbuffered ethanol (pH 6.3), the reduction of Fe3+
–
TPTZ complex in the FRAP test was carried out in acetate buffer
300 mM, pH 3.6, with a 1:500 molar ratio between vitamin E com-
pounds and ferric ions.12,13

F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
6487
Figure 6. Intermediates observed during reaction between DPPHÅÅ and tocopheramines. (left) ESR spectrum of the photooxidation product of 1d. The simulated ESR spectrum
is based on the structure of the nitroxyl radical intermediate of 1d. The coupling constants (Gauss) used for the simulation are listed adjacent to the respective atoms. (right)
Stopped flow UV/vis spectra obtained during the reaction of DPPHÅ with 1a (upper) and 4a (lower). The spectra show the absorption maximum of DPPHÅ at 518/514 nm, which
is rapidly decaying in this reaction. In the wavelength region below in the case of 1a the chromanoxyl radical (kmax 428 nm) and in the case of 4a an intermediate with kmax of
474 nm is detected.
2.4. Phenol substrate nitration in the iron-mediated
deamination of 4a
with results in the following order of magnitude: phenol > 2,
6-dimethylphenol > 2,6-di-tert-butylphenol (Table 1).
On the other hand, if the phenol group was blocked, as in
anisole and 2,6-dimethoxybenzene, no nitration product was
observed. From these data, it seems that the nitration reaction is
a radical process, in which the formation of an intermediate phe-
noxyl radical is critically involved. If the phenoxyl radical cannot
be formed, like in anisole, or its formation is difficult because of
steric hindrance, the reaction does not occur or proceeds to a lesser
extent, respectively. Traces of nitrated phenol (i.e., o- and p-nitro-
phenol, 2,6-dimethyl-4-nitrophenol) were observed in the reaction
on methoxy analogs employing a 100:1:1 ratio, indicating that an
O-demethylation reaction occurred and the resulting free phenol
was then nitrated.
Since both metal- and radical-dependent oxidation reactions
lead to tocopheramine deamination to form 1f, we preliminarily
investigated the fate of the nitrogen species released into the reac-
tion medium, starting from the hypothesis that the oxidation of
the amino group induced the release of a radical species which could
correspond or evolve to NOx species. The analogy with the nitro rad-
ical scavenging ability of tyrosine19 suggested the use of phenolic
compounds in order to trap possible tocopheramine derived radicals
or NOx species during the FeCl3 induced oxidation of 4a. Under 1:10
and 1:100 4a to FeCl3 molar ratio (in 50:50 v/v H2O/EtOH), phenol,
2,6-dimethylphenol and 2,6-di-tert-butylphenol were all found to
form the corresponding nitration products. In this series of sub-
strates, increasing steric hindrance in proximity to the phenol group
reduced the conversion rate into corresponding nitration products
We did not find any nitroso-, azo-, azoxy- or nitro-tocopheryl
analogs originating from direct oxidation of the 6-amino group in
4a.

6488
F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
TQ peak
500000
450000
400000
350000
300000
250000
200000
150000
100000
50000
0
1a + Fe3+
(1/4 mol/mol)
3500000
3000000
2500000
2000000
1500000
1000000
500000
1a + Fe3+
(1/40 mol/mol)
α-tocopheramine (α-TNH2)
23000
21000
19000
17000
15000
13000
11000
9000
2HN
206.9
429.3
O
C16H33
164.0
1a
280.9
132.9
7000
5000
3000
72.9
248.8
340.9
388.8
478.9
1000
m/z-->
50 100 150 200 250 300 350 400 450
-50000
0
2
4
6
8
10
12
14
16
18
400000
350000
300000
250000
200000
150000
100000
50000
0
600000
550000
500000
450000
400000
350000
300000
250000
200000
150000
100000
50000
4a + Fe3+
α-tocopheryl quinone (α-TQ)
(1/4 mol/mol)
221.0
OH
60000
55000
50000
45000
40000
35000
30000
25000
20000
15000
10000
5000
C
16H33
O
178.0
O
4a
57.0
91.0134.9
4a + Fe3+
(1/40 mol/mol)
280.9
251.1 326.9372.2
428.3
m/z-->
40 80 120 160 200 240 280 320 360 400 440
0
4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Time (min)
-50000
0
2
4
6
8
10
12
14
16
18
Time (min)
Figure 7. HPLC-ECD (left) and GC–MS (right) analysis of the oxidation products formed during the ferric ion-catalyzed reaction of 1a and 4a. 1a and 4a were exposed to a
molar excess of FeCl3 (1:40 or 1:4 mol/mol) in ethanol and immediately assessed by HPLC-ECD or GC–MS analysis as described in detail in the text.
Table 1
the corresponding nitration product 5-nitro-salicyclic acid. When
applying an 1:1 4a to salicylic acid ratio, 8% of the salicylic acid
was nitrated so that the two outcomes are consistent. In the case
of 2a and its model compound 2d, the yield of the corresponding
nitration products was 13% and 3%, respectively, in the case of
1:1 4a to phenol ratio.
Nitration of phenolic substrates during the ferric ion-catalyzed oxidation of 4a
Compound
Nitration producta (%)
Nitration productb (%)
PhOH
PhOMe
HOC6H3(Me)2-(2,6)
MeOC6H3(Me)2-(2,6)
HOC6H3(t-Bu)2-(2,6)
12
0
4
0
0.2
26
2
9
Traces
0.4
3. Conclusions
Data were mean values of triplicate experiments.
a
FeCl3/4a/phenolic 10:1:1.
FeCl3/4a/phenolic 100:1:1.
Tocopheramines and, for the first time, tocotrienamines were
comparatively assessed for their antioxidant function using three
TAC tests, namely TEAC, DPPHÅ bleaching and FRAP test. In all these
single-electron tests, VE amines showed similar or higher antioxi-
dant capacity than the parental tocopherols, tocotrienols and chro-
b
In order to investigate if the observed reaction was specific for
4a or shared with other aromatic amines, aniline and 4-aminoani-
sole were used in place of 4a in the presence of phenol. In the case
of aniline, no detectable nitration product was observed. On the
other hand, 4-aminoanisole, a model compound possessing higher
similarity to 4a, gave 2% and 6% conversion into nitrophenol
employing FeCl3 to 4-aminoanisole to phenol 10:1:1 and 100:1:1
molar ratios, respectively.
Those results suggested that, in the presence of phenoxyl radi-
cals generated under oxidative conditions, aromatic primary
amines bearing electron donating groups can transfer the nitrogen
to another phenolic substrate, the end product being a nitro deriv-
ative. The proposed formation of radical intermediates is currently
under further investigation by ESR and stopped flow techniques.
The nitration reaction was also preliminarily investigated in
apolar conditions using 2a, its analog 6-hydroxy-2,2,7,8-tetrame-
thylchroman 2d (Fig. 1) and salicylic acid as phenolic substrates.
Oxidation reactions were carried out in hexane using activated
Ag2O as oxidizing agent. In all these cases nitration products were
detectable, the best results obtained by the exposure of 4a to Ag2O
in n-hexane with a fivefold molar excess of the oxidant species. For
salicyclic acid, these conditions produced a 33% of conversion into
manol metabolites.
a-Tocopheramine and a-tocotrienamine
showed markedly higher activity than the congeners in the TEAC
test, which is tailored for liposoluble antioxidants and is a mixed
electron-transfer and hydrogen atom-transfer assay, while gamma
forms of the amine analogs showed high activity in the FRAP tests
which is a pure electron-transfer test. Kinetics analysis of DPPHÅ
reduction, showed higher second order rate k for VE amines than
1a. Deamination of tocopheramines occurred during either ferric
ions or DPPHÅ reactions with the formation of tocopherolquinone
as oxidation product. At the same time, present phenolics were ni-
trated, proving that the amino group was released and transformed
into nitrating species that in turn were trapped by the phenol. The
nitration of phenolic substrates depended on the presence of a free
phenolic hydroxyl group, and the steric hindrance of this phenolic
hydroxyl group influenced the nitration yield. The data suggest a
radical mechanism for the deamination of VE amine compounds,
involving phenoxyl radicals in the nitration mechanism.
This study clearly demonstrates that VE amines are synthetic
analogs with superior antioxidant activity in the VE family of com-
pounds. Although the detailed antioxidant mechanism of those
compounds remains undisclosed, we here identified that in both

F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
6489
metal- and radical-catalyzed reactions, VE amines exert nitrating
activity against phenolic substrates. This is a new finding that de-
serves further investigation to ascertain its biological relevance
and a possible role in the antioxidant function of VE amines.
These results in this study may help to better understand vita-
min E activity as well as the anticancer and anti-inflammatory
properties of these VE analogs.
Helium gas was used as carrier gas at constant flow of 1.2 ml/
min. Injector and transfer line temperature were 250 °C and
290 °C, respectively. A split–splitless inlet set to pulsed splitless
mode was used for injection. Starting oven temperature was set
at 50 °C and compound separation was obtained with a linear
temperature ramp of 50 °C/min from 50 °C to 320 °C; this final
temperature was held for 5 min. MS source temperature was
set at 230 °C and Electron Ionization (EI) was operated at constant
voltage with the electro multiplier set at 1200 V and 4 min sol-
vent delay. Total ion current was performed and the collected
data were analyzed with Enhanced Chemstation software (Agilent
Technologies, Inc.) that was supported with NIST library for MS
spectra analysis.
4. Experimental section
4.1. Chemicals and general experimental
Synthesis and characterization of VE amines were described in
Ref. 7. All the test compounds were stored as powders or oils in
sealed amber vials under nitrogen at ꢀ20 °C until use. Stock solu-
tions were prepared by dissolving the VE and VE analogs in abso-
lute ethanol, divided in aliquots and maintained under the same
storage conditions. Before each experimental session, compound
purity and actual concentration of stock solutions were verified
by UV spectroscopy and HPLC analysis (see below). (2R,40R,80R)-
4.4. Antioxidant activity assays
Comparative analysis of the antioxidant activity of VE com-
pounds was performed according to the approach recently de-
scribed by Muller et al.10 Three tests based on single electron or
H-atom transfer were used to assess antioxidant activity of VE
and VE amine compounds, namely TEAC, DPPHÅ and FRAP. Assay
details are described in Ref. 10. To adapt the FRAP assay to the
investigation of lipophilic antioxidants, minor changes to the origi-
nal method of Benzie and Strain12 were introduced according to
the protocol developed in Ref. 13.
The antioxidant capacity of tocopheramines was also assessed
in comparison with 1a using the copper-induced ascorbic acid
depletion assay. Briefly, all exposures were performed in triplicate
in UV 96 well flat-bottomed plates (Greiner bio-one) at a final vol-
a-, c- and d-tocopherol, salicylic acid, diphenyl picryl hydrazyl rad-
ical (DPPHÅ) and 2,20-azino-bis(3-ethylbenzthiazoline-6-sulphonic
acid) (ABTS) were from Sigma Chemicals (Milan, Italy). CEHC
metabolites were prepared as described in Ref. 20. Metabolite 1g
was kindly provided by Dr. Marc Birringer, Department of Human
Nutrition, Institute of Nutrition, Friedrich Schiller University, Jena,
Germany. Other reagents and solvents were commercially avail-
able and used without further purification. Reagent-grade solvents
were used for all extractions and workup procedures. n-Hexane,
diethyl ether, ethyl acetate and petroleum ether used in chroma-
tography were distilled before use. TLC was performed using Merck
silica get 60 F254 pre-coated plates. Flash chromatography was
ume of 200
a concentrated stock of ascorbate (2 mM, pH adjusted to 7 by
NaOH) to obtain a starting ascorbate concentrations of 200 M in
each well that contained 160 l of VE suspension and 20 l of Che-
lex-100 resin treated water containing 2 M Cu+ ions that were
ll. Exposures were initiated by the addition of 20 ll of
l
l
l
performed using Baker silica gel (40 lm particle size). All products
l
were purified to homogeneity and verified by TLC/GC analysis.
Melting points, determined on a Kofler-type micro hot stage with
Reichert-Biovar microscope, are uncorrected. NMR spectra were
recorded at 300.13 MHz for 1H and at 75.47 MHz for 13C NMR in
CDCl3 if not otherwise stated. Chemical shifts, relative to TMS as
internal standard, are given in d values, coupling constants in Hz.
13C resonances were assigned by means of APT, HMQC and HMBC
spectra. Elemental analyzes were performed at the Microanalytical
Laboratory of the Institute of Physical Chemistry at the University
of Vienna.
preliminarily verified to produce 1st order oxidation kinetics of
ascorbate. Immediately prior to the addition of ascorbate to each
assay well, the plate was pre-incubated for 10 min at 37 °C in a
plate reader (Spectra Max 190), and during the exposure the plate
was maintained at this temperature. After addition of ascorbate,
the concentration remaining in each well was monitored every
2 min for
a period of 2 h by measuring the absorbance at
265 nm. The initial rate of ascorbic acid depletion was measured
in triplicate experiments and DTPA was used as metal chelator to
verify the specificity of the analysis.
4.2. Statistics
4.5. Stopped flow reaction of 1a, tocopheramines and
tocotrienamines with diphenyl picryl hydrazyl radical (DPPHÅ)
Data were presented as mean SD of three series of experi-
ments run in triplicate. Differences between TAC data were as-
sessed by Student Newman–Keuls post-hoc ANOVA and a level of
p <0.05 was accepted as significant.
Stock solutions of antioxidants (ca. 2 mM), DPPHÅ (200
lM)
were prepared in ethanol (HPLC-grade). The concentrations of
the individual antioxidant stock solutions were determined via
their extinction coefficients reported in Ref. 7. This concentration
divided by two provides the starting concentration of the antioxi-
dant after mixing (CAntioxidant).
4.3. Vitamin E analysis
Analysis of VE compounds was performed by HPLC-ECD (elec-
trochemical detector) and GC–MS as described before.21,22 HPLC
analysis was carried out under isocratic conditions on a MetaSil
For stopped flow experiments a stopped flow mixing accessory
(RX 2000, Applied Photophysics) was used in conjunction with a
diode array photometer (Multispec 1501, Shimadzu). The two
syringes were loaded with 2 ml each of the DPPHÅ solution and
the antioxidant solution. After initiation of the data acquisition
at the photometer for the range of 350–600 nm in 0.1 s intervals
the mixing was initiated by the pneumatic drive accessory. The
time point t = 0 s was obtained from the spike of the appearance
of the DPPHÅ absorption. From the 3D scans the time traces of
518 nm minus 600 nm were extracted for kinetic analysis of the
DPPHÅ decay. The second order rate constant (k) was directly ob-
tained by nonlinear regression according to the following formula:
AQ 3l, C18 150 ꢁ 4.6 mm column (Varian Inc.) using 10 mM so-
dium perchlorate in 98:2 MeOH/H2O as mobile phase. The ECD
detector was composed of a glassy carbon cell with working elec-
trode set at 750 mV oxidation potential against a AgCl reference
electrode. For redox-silent compounds, UV detection was oper-
ated using a Jasco UV975 detector module set at 210 nm and
mounted in series before the ECD detector. GC–MS analysis was
performed on a 7890A GC system equipped with a 5975C VL tri-
ple-axis mass spectrometry detector (Agilent Technologies, Inc.).

6490
F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
Antioxidantꢁðtimeꢀtimeoffset
OD ¼ ODtotal decay ꢁ eꢀkꢁc
Þ þ ODoffset
4.7.1. Methyl-(N,2,2,5,7,8-pentamethylchroman-6-yl)-amine
(1d, NHMe–PMC)
Data were collected in at least three different experimental ses-
sions. To detect intermediates of antioxidants formed during the
reaction with DPPHÅ, stock solutions of 1a (10 mM), 4a (10 mM),
TLC: Rf = 0.19 (n-hexane/ethyl acetate 5:1 (v/v)); 1H NMR
3
(CDCl3) d: 1.30 (s, 6H, H-2a), 1.78 (t, JHH = 6.8 Hz, H-3), 2.08 (s,
3H, H-7a), 2.13 (s, 3H, H-8b), 2.17 (s, 3H, H-5a), 2.59 (t, 2H,
3JHH = 6.8 Hz, H-4), 2.80 (s, 6H, N–CH3); 13C NMR (CDCl3) d: 12.0
(C-8b), 14.2 (C-5a), 15.1 (C-7a), 21.2 (C-4), 27.0 (C-2a), 33.0 (C-3),
43.0 (N-CH3), 72.6 (C-2), 100.0 (C-4a), 117.0 (C-8), 133.3 (C-7),
135.0 (C-5), 141.4 (C-6), 188.0 (C-8b); EI-MS m/z 233 (100%), 177
(95%), 178 (41%), 149 (36%), 134 (19%), 234 (16%), 148 (13%); Anal.
Calcd for C16H25ON: C, 77.68; H, 10.19; N, 5.66. Found: C, 77.68; H,
10.32; N, 5.60.
DPPHÅ (200
lM) in acetonitrile (HPLC-grade) were used in the de-
scribed rapid mixing procedure.
4.6. HPLC analysis of reaction products of DPPHÅ with 1a and 4a
Aerobic solutions of DPPHÅ (0.5 mM) were mixed with 1a
(0.5 mM) or 4a (0.5 mM) and incubated for 20 min at room tem-
perature. The solvent of the mixture was evaporated in a stream
of argon and the solid residue was dissolved in the HPLC eluent.
HPLC analysis was performed on a Waters LC-1 Module. The col-
4.7.2. Dimethyl-(N,N,2,2,5,7,8-pentamethylchroman-6-yl)-
amine (1e, NMe2–PMC)
TLC: Rf = 0.72 (n-hexane/ethyl acetate 5:1 (v/v)); 1H NMR
umn (Hibar, LiChrospher 100, RP-18 (5
with acetonitrile/H2O (70/30 v/v) at a rate of 0.8 ml/min and
l. The DPPHÅ-related compounds
lm), 250-4) was eluted
3
(CDCl3) d: 1.30 (s, 6H, H-2a), 1.80 (t, JHH = 6.8 Hz, H-3), 2.12 (s,
25 °C using inject volumes of 10
l
3H, H-8b), 2.19 (s, 3H, H-7a), 2.24 (s, 3H, H-5a), 2.63 (t,
3JHH = 6.8 Hz, H-4), 2.65 (s, 3H, N–CH3), 2.81 (br, 1H, NH); 13C
NMR (CDCl3) d: 12.4 (C-8b), 13.6 (C-5a), 14.5 (C-7a), 21.7 (C-4),
27.2 (C-2a), 33.4 (C-3), 37.0 (N-CH3), 72.9 (C-2), 117.5 (C-4a),
122.9 (C-8), 127.4 (C-5), 129.3 (C-7), 139.1 (C-6), 148.4 (C-8a);
EI-MS m/z 247 (100%), 192 (43%), 148 (30%), 191 (27%), 176
(19%), 248 (17%), 190 (14%), 232 (13%), 177 (10%); Anal. Calcd for
were detected by an UV-detector at 304 nm. Peaks of DPPHÅ, DPPH-
H were identified by standard compounds. The DPPHÅ reaction
product at a retention time of 4 min was collected from 10 HPLC
runs and then subjected to LC/MS. LC/MS was done on a Dionex
Ultimate 3000 system with a dual micro-flow pump, autosampler,
temperature controlled column compartment and UV-DAD. The
flow was split before the MS, so that only 200 ll/min entered the
C15H23ON: C, 77.21; H, 9.93; N, 6.00. Found: C, 77.14; H, 10.00;
ESI source. MS was performed on an Agilent MSD 6320 XCT ion
trap in negative mode ESI (capillary voltage: 3500 V, dry tempera-
ture: 300 °C, dry gas flow: 10 l/min, nebulizer pressure: 25 psi).
The scan range was set to 300–500 m/z and five scans were
averaged for one mass spectrum (Smart ICC target: 20,000, maxi-
mum accumulation time: 500 ms, smart parameter target mass:
400 m/z).
N, 5.92.
4.8. Nitration of phenolic substrates
Nitration of phenolics (2a, 2d, salicylic acid) upon oxidation of
4a was preliminarily investigated using FeCl3 as the oxidant
(10:1 and 100:1 molar ratio with VE compounds) in 50:50 v/v
H2O/EtOH. Nitration studies in hexane were carried out with
Ag2O as the oxidant, using either salicylic acid, 2a or its analog
2d as the phenol derivatives and 4a as the aniline derivatives. Toc-
opheramine and oxidants were simultaneously introduced in the
reaction tube. In the case of 2a and 2d, a 4:1-ratio was also used,
4.7. Detection of intermediates during the oxidation of 1d by
ESR spectroscopy
The
a-tocopheramine model compound 2,2,5,7,8-pentame-
and the c-chromanol substrates were introduced 5 min later than
thylchroman-6-amine (NH2–PMC, PMC = pentamethylchromane)
was prepared according to,7 starting from 2,2,5,7,8-pentame-
thylchroman-6-ol instead of 1a. To obtain the N-methylated deriv-
atives, NH2–PMC (223 mg, 1.02 mmol) was dissolved in DMSO
(2.5 ml) under an argon atmosphere. NaOH (300 mg, 7.5 mmol)
and CH3I (216 mg, 1.53 mmol, 1.53 equiv) were added consecu-
tively, and the mixture was stirred at room temperature. After
5 h the reaction was quenched with water, and the mixture was
extracted three times with dichloromethane. The organic extract
was washed four times with deionized water and once with brine.
The organic extract was dried over MgSO4, filtered and evaporated
in vacuo. The residue (204 mg, yellow oil) was purified by column
chromatography (5 g silica gel, n-hexane/ethyl acetate 5:1), pro-
viding N,N-dimethyl-(2,2,5,7,8-pentamethylchroman-6-yl)-amine
1e (NMe2–PMC) in 37% yield (94 mg) as a yellow oil, and N-
methyl-(2,2,5,7,8-pentamethylchroman-6-yl)-amine 1d in 36%
yield (85 mg) as a colorless oil.
the oxidant to minimize their direct oxidation. The structure of the
nitration products was confirmed by UV and NMR spectroscopy.
4.8.1. 5-Nitro-c
-tocopherol24
Red oil. TLC: Rf = 0.65 (n-hexane/diethyl ether, v/v = 9:1); UV
(EtOH): kmax (nm) = 265, 315, 418; 1H NMR (CDCl3) d: 1.67 (t, 2H,
3J = 6.9 Hz, H-3), 2.11 (s, 3H, H-7a/8b), 2.14 (s, 3H, H-7a/8b), 2.96
(t, 2H, 3J = 6.8 Hz, H-4), 10.65 (s, 1H, –OH); 13C NMR (CDCl3) d:
11.87 (C-7a), 13.05 (C-8b), 21.83 (C-4), 23.99 (C-2a), 31.14 (C-3),
75.02 (C-2), 113.12 (C-4a), 125.43 (C-7), 134.22 (C-5), 137.12
(C-8), 144.03 (C-8a), 148.00 (C-6), isoprenoid side chain: 19.76
(C-4a0), 19.84 (C-8a0), 21.09 (C-20), 22.67 (C-130), 22.74 (C-12a0),
24.50 (C-60), 24.65 (C-100), 28.00 (C-120), 32.64 (C-80), 32.75
(C-40), 37.33 (C-70), 37.46 (C-50), 37.47 (C-90), 37.52 (C-30), 39.32
(C-110), 39.69 (C-10); Anal. Calcd for C28H47O4N: C, 72.84; H,
10.26; N, 3.03. Found: C, 72.78; H, 10.42; N, 2.94.
A
solution of 1d (50 mM) in CH2Cl2 was prepared and
transferred to a quartz flat cell, which was immersed in liquid
nitrogen. Then the flat cell was rapidly transferred to the
standard resonator of the ESR instrument (Brucker EMX) and
irradiated using a high pressure mercury lamp. The ESR spectra
were recorded with the following parameters: microwave
frequency 9.78 GHz, modulation frequency 100 kHz, modulation
amplitude 2 G, time constant 41 ms, center field 3491 G, sweep
80 G, sweep time 10 s, receiver gain 9 ꢁ 104, scans 1.
Simulation of the spectra was performed using the program
WINSIM.23
4.8.2. 6-Hydroxy-5-nitro-2,2,7,8-tetramethylchroman
Reddish-colored crystals, mp: 79–81 °C; TLC: Rf = 0.50 (n-hex-
ane/ethyl acetate, v/v = 30:1); UV (EtOH): kmax (nm) = 265, 315,
418; IR (KBr): 2977, 2929, 1595, 1532, 1444, 1276, 1173; 1H
NMR (CDCl3) d: 1.24 (s, 6H, H-2a), 1.67 (t, 2H, 3J = 6.8 Hz, H-3),
2.11 (s, 3H, H-7a/8b), 2.15 (s, 3H, H-7a/8b), 2.95 (t, 2H,
3J = 6.8 Hz, H-4), 10.60 (s, 1H, –OH); 13C NMR (CDCl3) d: 11.92
(C-7a), 13.18 (C-8b), 21.87 (C-4), 26.74 (C-2a), 32.31 (C-3), 73.22
(C-2), 113.04 (C-4a), 125.16 (C-7), 133.98 (C-5), 136.90 (C-8),
145.18 (C-8a), 148.22 (C-6); MS (ESI Q-TOF) m/z: calcd for

F. Galli et al. / Bioorg. Med. Chem. 19 (2011) 6483–6491
6491
C13H17O4N (M+H)+ 252.1230. Found: 252.1208. Anal. Calcd for
13H17O4N: C, 62.14; H, 6.82; N, 5.57. Found: C, 62.40; H, 6.73;
N, 5.43.
2. Smith, L. I.; Renfrow, W. B.; Opie, J. W. J. Am. Chem. Soc. 1942, 64, 1082.
3. Bieri, J. G.; Prival, E. L. Biochemistry 1967, 6, 2153.
C
4. Schwieter, U.; Tamm, R.; Weiser, H.; Wiss, O. Helv. Chim. Acta 1966, 49, 2297.
5. Gloor, U.; Würsch, J.; Schwieter, U.; Wiss, O. Helv. Chim. Acta 1966, 49, 2303.
6. Itoh, S.; Nagaoka, S.; Mukai, K.; Ikesu, S.; Kaneko, Y. Lipids 1994, 29, 799.
7. Mazzini, F.; Netscher, T.; Salvadori, P. Eur. J. Org. Chem. 2009, 2063.
8. Turanek, J.; Wang, X. F.; Knotigova, P.; Koudelka, S.; Dong, L. F.; Vrublova, E.;
Mahdavian, E.; Prochazka, L.; Sangsura, S.; Vacek, A.; Salvatore, B. A.; Neuzil, J.
Toxicol. Appl. Pharmacol. 2009, 237, 249.
9. Mazzini, F.; Betti, M.; Canonico, B.; Netscher, T.; Luchetti, F.; Papa, S.; Galli, F.
ChemMedChem 2010, 5, 540.
10. Muller, L.; Theile, K.; Bohm, V. Mol. Nutr. Food Res. 2010, 54, 731.
11. Pellegrini, N.; Serafini, M.; Colombi, B.; Del Rio, D.; Salvatore, S.; Bianchi, M.;
Brighenti, F. J. Nutr. 2003, 133, 2812.
4.8.3. 5-Nitro-salicylic acid (2-hydroxy-5-nitrobenzoic acid)
Mp: 227–229 °C; TLC: Rf = 0.44 (toluene/ethyl acetate,
v/v = 8:2); 1H NMR (DMSO-d6) d: 7.14 (d, 1H, 3J = 9.4 Hz, H-3),
8.32 (dd, 1H, 3J = 9.4 Hz, 4J = 1.8 Hz, H-4), 8.55 (d, 1H, 4J = 1.8 Hz,
H-6); 13C NMR (DMSO-d6) d: 117.3 (C-3), 119.4 (C-1), 123.0 (C-6),
124.9 (C-4), 129.6 (C-5), 162.7 (C-2), 165.2 (COOH). Anal. Calcd
for C7H5O5N: C, 45.91; H, 2.75; N, 7.65. Found: C, 45.86; H, 2.91;
N, 7.66.
12. Benzie, I. F.; Strain, J. J. Anal. Biochem. 1996, 239, 70.
13. Floridi, A.; Piroddi, M.; Pilolli, F.; Matsumoto, Y.; Aritomi, M.; Galli, F. Acta
Biomater. 2009, 5, 2974.
14. Huang, D.; Ou, B.; Prior, R. L. J. Agric. Food Chem. 2005, 53, 1841.
15. Gregor, W.; Grabner, G.; Adelwöhrer, C.; Rosenau, T.; Gille, L. J. Org. Chem. 2005,
70, 3472.
16. Mukai, K.; Tokunaga, A.; Itoh, S.; Kanesaki, Y.; Ohara, K.; Nagaoka, S.; Abe, K. J.
Phys. Chem. B 2007, 111, 652.
17. Murphy, P. A.; Lin, J. S.; Olcott, H. S.; Windle, J. J. Lipids 1976, 11, 296.
18. Nakahara, K.; Iwasa, S.; Iriyama, J.; Morioka, Y.; Suguro, M.; Satoh, M.; Cairns, E.
J. Electrochimica Acta 2006, 52, 921.
19. Bartesaghi, S.; Ferrer-Sueta, G.; Peluffo, G.; Valez, V.; Zhang, H.; Kalyanaraman,
B.; Radi, R. Amino acids 2007, 32, 501.
20. Mazzini, F.; Galli, F.; Salvatori, P. Eur. J. Org. Chem. 2006, 5588.
21. Galli, F.; Polidori, M. C.; Stahl, W.; Mecocci, P.; Kelly, F. J. Vitam. Horm. 2007, 76,
263.
22. Galli, F.; Lee, R.; Dunster, C.; Kelly, F. J. Free Radic. Biol. Med. 2002, 32, 333.
23. Duling, D. R. J. Magn. Reson. 1994, 104, 105.
Acknowledgments
This study was part of the research program sponsored by the
‘Fondazione per la ricerca sulla Fibrosi Cistica (FFC)’, Grant No.
FFC#13/2008. Compound analysis and synthesis were sponsored
in part with the grant of the ‘Ministero dell0Università e della Ric-
erca’ within the grant program ‘Programmi di ricerca scientifica di
rilevante interesse nazionale (PRIN)’. Prot. No. 20078TC4E5.
Preliminary results in this study were presented on May 2010
as part of the dissertation of Miss. Adriana Tanzarella at the school
of ‘Scienze dell’Alimentazione
e
della Nutrizione Umana’,
University of Perugia. The financial support of the Austrian Chris-
tian Doppler Research Society (laboratory ‘Advanced cellulose
chemistry and analytics’) is gratefully acknowledged.
24. Patel, A.; Liebner, F.; Netscher, T.; Mereiter, K.; Rosenau, T. J. Org. Chem. 2007,
72, 6504.
References and notes
1. Rosenau, T. In The Encyclopedia of Vitamin E; Preedy, V. R., Watson, R. R., Eds.;
CAB International, 2007; p 69.