Food Chemistry 126 (2011) 1629–1635

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Astaxanthin is responsible for antiglycoxidative properties of microalga

Chlorella zofingiensis
Zheng Sun a, Jin Liu a, Xiaohui Zeng a, Jieqiong Huangfu a, Yue Jiang b, Mingfu Wang a,⇑, Feng Chen a,⇑
a
School of Biological Sciences, The University of Hong Kong, Pokfulam Road, Hong Kong, PR China
b
Department of Biology, Hong Kong Baptist University, Kowloon Tong, Hong Kong, PR China

a r t i c l e i n f o a b s t r a c t

Article history: The antiglycoxidative properties of microalga Chlorella zofingiensis were investigated for the first time in
Received 14 September 2010 this study. Algal extracts containing different contents of astaxanthin were prepared. Through the com-
Received in revised form 13 November 2010 parison, it was shown that the extract rich in astaxanthin exhibited higher antioxidant abilities as well as
Accepted 5 December 2010
stronger antiglycative capacities, including the inhibition of advanced glycation endproducts (AGEs) for-
Available online 10 December 2010
mation, glucose autoxidation as well as glycation-induced protein oxidation. The extract was further frac-
tionated using TLC. Among all fractions obtained, the fraction of astaxanthin in diester form was found to
Keywords:
contain the strongest inhibitory effects on the glycation cascade. Its tentative structure was subsequently
Astaxanthin
Chlorella zofingiensis
identified by LC–MS analysis. These results clearly ascertained the antiglycoxidative properties of asta-
Diabetic complications xanthin derived from C. zofingiensis and supported the possibility of using natural antioxidants as glyca-
Glycoxidation tion inhibitors. The microalga C. zofingiensis, therefore, might be the beneficial food and preventive agent
choice for diabetic patients.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction dehyde, which subsequently participates in the AGEs formation

Non-enzymatic glycation, also referred to as the Maillard reac-
tion, is a process in which reducing sugars react spontaneously
with amino groups of proteins. It occurs not only in foods but also
in various biological systems, and leads to the formation of ad-
vanced glycation endproducts (AGEs). AGEs are a group of complex
and heterogeneous molecules, comprising various products with
different structures and characters (Ahmad & Ahmad, 2006). Once
formed in living organisms, they can generate cross-links between
key molecules, cause their structural modification and functional
impairment by altering enzymatic activity, decreasing ligand bind-
ing, changing protein half-life and reducing the immunogenicity
(Vlassara & Palace, 2002). Therefore non-enzymatic glycation and
the accumulation of tissue AGEs are believed to act as major path-
ogenic processes in many human diseases especially diabetes and
its complications, such as retinopathy, cataract, neuropathy and
nephropathy (Nathan, 1993).
Free radicals and oxidative steps are known to get involved in
the glycation process which is called glycoxidation, and they are
closely associated with the development of diabetic complications.
During glycoxidation, glucose reduces molecular oxygen to gener-
ate superoxide radicals and converts itself to a dicarbonyl ketoal-
⇑ Corresponding authors. Tel.: +852 22990309; fax: +852 22990311 (F. Chen),
Tel.: +852 22990338; fax: +852 22990340 (M. Wang).
E-mail addresses: mfwang@hkusua.hku.hk (M. Wang), sfchen@hkusua.hku.hk
(F. Chen).
(Hunt, Bottoms, & Mitchinson, 1993). Once formed, the superoxide
radicals can be converted to the highly reactive hydroxyl radicals
via the Fenton reaction, which will induce further oxidative protein
degradation (Wolff & Dean, 1987). Furthermore, many lines of
evidence indicate that oxidative stress is the trigger that drives
various biochemical pathways associated with hyperglycaemia-in-
duced cell damage (Brownlee, 2001). Thus, regarding the key roles
of oxidative stress in the development of diabetic complications,
the discovery of antioxidants with inhibitory effects on the glyca-
tion cascade in recent years has received much attention. Such re-
search is believed to offer a promising therapeutic and preventive
approach (Ardstani & Yazdanparast, 2007). In fact, there is a large
body of evidence that compounds with combined antioxidant
and antiglycative properties are more effective in treating diabetes
(Duraisamy et al., 2003; Xi et al., 2008).
We are interested in screening and development of antioxidants
containing significant antiglycative capacities from natural sources
because compared to synthetic agents, natural products have few-
er side effects and are much safer for human consumption. In mic-
roalgae, a wide range of antioxidants can be produced such as
carotenoids, polyunsaturated fatty acids and polysaccharides
(Chen, 1996). The green microalga Chlorella zofingiensis is known
as a natural source of astaxanthin, a red ketocarotenoid that pos-
sesses potent anti-oxidative capacity. The antioxidant activity of
astaxanthin is as high as 10 times more than other carotenoids
such as zeaxanthin, lutein, canthaxanthin and b-carotene, and
100 times more than a-tocopherol (Miki, 1991). Although

0308-8146/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2010.12.043
1630 Z. Sun et al. / Food Chemistry 126 (2011) 1629–1635
C. zofingiensis does not produce as much astaxanthin as another
microalga Haematococcus pluvialis which accumulates the highest
astaxanthin content in nature, this strain has many other advanta-
ges such as its fast growth and low sensitivity to contamination or
unfavourable environments (Ip & Chen, 2005). These properties
make C. zofingiensis a very suitable host for mass production of
astaxanthin. However little is known about its antiglycative activ-
ity. Therefore, the aim of the present study was to investigate the
effects of C. zofingiensis, a natural producer of astaxanthin in the
glycoxidation process.
2. Materials and methods
2.1. Reagents and chemicals
SDS–PAGE reagents were purchased from Bio-Rad Laboratories
(Hercules, CA, USA). Unless otherwise stated, all other reagents
were purchased from Sigma Chemical Co. (St. Louis, MO, USA).
All analytical and HPLC grade solvents used were obtained from
BDH Laboratory Supplies (Pool, UK).
2.2. Algal strain and culture conditions
The green microalga C. zofingiensis (ATCC 30412) was obtained
from the American Type Culture Collection (Rockville, MD, USA).
The algal cells were first grown in Kuhl medium (Kuhl, 1962) at
25 °C under continuous illumination of 25 lmol photon m2 s1.
After 4 days, approximately 10% (v/v) exponentially growing algal
cells were inoculated into the medium containing 30 g L1 of glu-
cose. Cells were cultured in 250 mL Erlenmeyer flasks each con-
taining 100 mL of medium and maintained at 25 °C with orbital
shaking at 130 rpm in darkness.
2.3. Preparation of extracts
Cells were harvested at the 2nd and 14th day, which exhibited
the colour of green and red, respectively. After being freeze-dried
for 24 h, the algal biomass (0.1 g) was extracted with ethyl acetate
(8 mL) at room temperature. The tube containing the extracts was
centrifuged at 4500 g for 10 min and the supernatant was recov-
ered. The extraction was repeated and the two supernatants were
combined. Samples were purged to dryness using nitrogen and
stored at 0 °C before use.
2.4. Trolox equivalent antioxidant capacity (TEAC) assay
The TEAC assay was performed according to Re et al. (1999)
with minor modification. Briefly, a 2,20 -azinobis [3-ethylbenzo-
thiazoline-6-sulphonate] radical anion (ABTS+) solution was pre-
pared using 7 mM ABTS and 2.45 mM potassium persulfate. The
solution was kept in the dark for 16 h before use. After being di-
luted with PBS, the ABTS.+ solution gave an absorbance of
0.700 ± 0.050 at 734 nm. For measuring antioxidant capacity,
10 lL of samples were mixed with 990 lL of ABTS+ solution, to give
absorbance values 20–80% of that of the blank. The absorbance of
mixture was measure at 734 nm after 6 min. Trolox solutions were
used to create a standard curve. The results were expressed as mg
Trolox g1 dried extract.
2.5. 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical inhibition assay
The DPPH radical scavenging assay was performed according to
Blois (1958) with slight modification. Briefly, 1 mL of algal extracts
was mixed with 1 mL of DPPH (0.1 mM) in ethanol solution. After
30 min incubation at room temperature, the decrease in absorbance
was monitored at 517 nm. The percentage of inhibition was calcu-
lated by: %inhibition = [1  (absorbance of the solution with
extracts/absorbance of the solution without extracts)]  100%.
2.6. In vitro glycation of bovine serum albumin
The incubation of bovine serum albumin (BSA) was performed
as described by Sun et al. (2010). BSA (20 mg mL1) was incubated
with glucose (800 mM) and NaN3 (0.2 mg mL1) in phosphate buf-
fer (0.2 M, pH 7.4). All incubations were carried out at 37 °C for
2 weeks in the absence and presence of inhibitors. Aminoguanidine
(AG) solution (1 mM) was used as the positive control.
2.7. Detection of glycated proteins by fluorescence measurement
The measurement of fluorescent intensity of AGEs was per-
formed using a Hitachi F-2500 fluorescent spectrometer (Hitachi
Corporation, Tokyo, Japan). The presence of AGEs was character-
ised by a typical fluorescence with excitation and emission max-
ima at 330 and 410 nm, respectively. Percent inhibition of AGE
formation by each extract was calculated using the following equa-
tion, %inhibition = [1  (fluorescence of the solution with inhibi-
tors/fluorescence of the solution without inhibitors)]  100%.
2.8. Detection of glycated proteins by SDS–PAGE
The formation of AGE products was also detected using SDS–
PAGE. Briefly, samples were separated on a 12% polyacrylamide
gel with 10 lg of protein loaded in each lane. The gels were stained
with coomassie brilliant blue and photographed using the Bio-Rad
Gel documentation system (Bio-Rad, Hercules, CA, USA). The inten-
sity of protein bands was calculated using the Quantity one soft-
ware (Bio-Rad, Hercules, CA, USA).
2.9. Detection of hydroxyl radicals from sugar autoxidation
The detection of hydroxyl radicals induced by sugar autoxida-
tion was carried out as described by Hunt, Dean, and Wolff
(1988). 1 mM sodium benzoate, 500 mM glucose and 0.1 mM
CuSO4 were dissolved in 100 mM potassium phosphate buffer.
The mixture was incubated at 37 °C for 4 days in the absence and
presence of inhibitors. The decrease in benzoate hydroxylation re-
sulted from the scavenging of hydroxyl radical production was
measured by the fluorescence intensity (excitation maxima:
308 nm; emission maxima: 410 nm).
2.10. Detection of protein carbonyl formation
The content of protein carbonyl products generated from gly-
coxidation process was determined by the method of Ardstani
and Yazdanparast (2007). Briefly, 1 mg glycated BSA/ original BSA
was reacted with 1 mL of 10 mM 2,4-dinitrophenylhydrazine
(DNPH) in 2 M HCl. After incubation at room temperature for
30 min, 1 mL of cold trichloroacetic acid (10%, w/v) was added to
the mixture and centrifuged at 3000g for 10 min. The collected
protein pellet was washed three times with ethanol/ethyl acetate
(1:1 v/v) and re-dissolved in 1 mL of guanidine hydrochloride
(6 M, pH 2.3). The absorbance of the sample was measured spec-
trophotometrically at 370 nm. The molar extinction coefficient of
DNPH is e = 2.2  104 cm1 M1.
2.11. Determination of thiol groups
Thiol groups were measured according to Ellman’s assay (Ellman,
1959). Briefly, glycated BSA/ original BSA (3.5 mg mL1) was reacted
with 5,50 -dithio-bis-(2-nitrobenzoic acid) (DTNB) (2.5 mM). The
 Z. Sun et al. / Food Chemistry 126 (2011) 1629–1635
reaction was allowed to stand at room temperature for 15 min,
yielding a yellow-coluored product known as 2-nitro-5-thiobenzoic
acid (TNB). The absorbance was measured at 410 nm. A standard
curve was prepared using L-cysteine hydrochloride monohydrate.
The concentration of free sulfhydryl is calculated using the equation
c = A/bE, whereas c = concentration (M), A = absorbance, b = path
length in centimeters and E = 14,150 M1 cm1, the molar extinction
coefficient of TNB.
2.12. Thin layer chromatography analysis
The fractionation of the extract obtained from the 14th day was
carried out by thin layer chromatography (TLC). The extract was
evaporated under nitrogen gas and the residues were suspended
in 50 lL of acetone. The stationary phase was silica gel 60 plates
(Merck) whereas the mobile phase was ethyl acetate: hexane
(40:60 v/v). Samples were loaded on TLC plates and separated into
individual pigment classes. After the identification by HPLC analy-
sis, each fraction was re-dissolved in 50 lL of acetone, of which
30 lL was used in following assays.
2.13. HPLC analysis
To identify different fractions obtained from TLC analysis, the cor-
responding spots were scraped off the plate, extracted with acetone
and analysed using the HPLC system, as described by Baroli, Do, Ya-
mane, and Niyogi (2003) with minor modification. The HPLC system
was equipped with a Waters 2695 separations module and a Waters
2996 photodiode array detector. After the filtration through a 0.22-
lm Millipore organic membrane, 20 lL of samples were analysed by
HPLC on a Waters Spherisorb 5 lm ODS2 4.6  50 mm analytical
column (Waters, Milford, MA, USA). Flow rate was 1.2 mL/min. Sol-
vent A was acetonitrile: methanol:0.1 M Tris–HCl, pH 8.0 (84:2:14).
Solvent B was methanol: ethyl acetate (68:32). A linear gradient was
performed from 100% solvent A to 100% solvent B in 15 min, followed
by 20 min of solvent B. The absorption spectra were 300–700 nm.
Peaks were measured at 450 nm. The fractions were identified and
quantified by using standard curves.
2.14. Liquid chromatography-mass spectrometry (LC–MS) analysis
To get a clear understanding of the chemical structure of asta-
xanthin diester, the fraction of astaxanthin diester obtained from
TLC was analysed on a LC–MS/MS instrument equipped with an
electrospray ionization (ESI) source interfaced to a QTRAP mass
spectrometer (Applied Biosystems, Foster City, CA, USA). Liquid
chromatography was run on a separation model (Agilent 1100;
Agilent Technologies, Santa Clara, CA, USA) with a degasser, a qua-
ternary pump and a thermostatted autosampler. Separation was
conducted on a Sunfire™ C18 column. The mobile phases were
water added with 0.05% formic acid (solvent A) and acetonitrile
(solvent B), and the flow rate of mobile phase was set at
0.2 mL min1. The elution started with 90% A (water of 0.05% for-
mic acid) for 10 min, then linear gradient to 80% B (acetonitrile)
in 30 min and finally kept at 80% B till 50 min. The post running
time was 20 min. The MS parameters were as follows: (experiment
I) positive ion mode, spray voltage 3.5 kV, temperature 350 °C, scan
range 50–1500 Da, DP 5 V; (experiment II) positive ion mode, spray
voltage 4.5 kV, temperature 450 °C, scan range 50–500 Da, DP 60 V.
2.15. Statistical analysis
Experimental results were obtained as mean value ± SD (n = 3).
Statistical analyses were performed using the SPSS statistical pack-
age (SPSS Inc., Chicago, IL, USA). Paired-samples T-test was applied.
The statistical significances were achieved when p < 0.05.
1631
3. Results and discussion
Due to the significant roles of glycation in the development of
diabetic complications, there has been an increasing interest in
the investigation of AGE formation inhibitors in recent years.
Although some pharmacological compounds showed strong inhib-
itory effects such as the nucleophilic hydrazine compound amino-
guanidine (AG), the first AGE inhibitor explored in clinical trials,
their side effects and high toxicity for diabetic patients are serious
concerns for people, which hinders the ultimate approval for their
commercial production (Thornalley, 2003). It is now generally ac-
cepted that using natural products such as plant extracts and puri-
fied constituents would be beneficial in the treatment of diabetes.
In this study, the green microalga C. zofingiensis was evaluated for
its effects on the glycoxidation process. This algal strain has been
highlighted by a number of studies for its high accumulation of
astaxanthin when grown with glucose as sole carbon and energy
source (Ip & Chen, 2005; Sun, Wang, Li, Huang, & Chen, 2008).
The alga accumulates primary carotenoids such as lutein and b-
carotene to protect the cells from oxidative damage. However, if
the amount of primary carotenoids is not enough, secondary
carotenoids (i.e. astaxanthin, canthaxanthin and adonixanthin) will
be generated to diminish the excessive oxidative stress, whereas
astaxanthin acts as the major secondary carotenoids and over
90% of the astaxanthin is in the form of mono- and di-esters (Bar,
Rise, Vishkautsan, & Arad, 1995). It was shown that under hetero-
trophic conditions, the colour of C. zofingiensis gradually changed
from green to red, indicating the accumulation of astaxanthin
within algal cells. The contents of total carotenoids and astaxan-
thin in C. zofingiensis were analysed by HPLC and shown in Fig. 1.
During the cultivation, the level of astaxanthin started to increase
from the 2nd day (0.159 mg g1) and achieved the peak at the 14th
day (0.877 mg g1). Therefore in the present work, cells were har-
vested at the 2nd and 14th day for extraction, respectively, repre-
senting above two distinct stages.
The two C. zofingiensis extracts were analysed for antioxidant
activities at first. DPPH is a stable radical and often used for the
evaluation of antioxidant ability of natural compounds. The IC50
values for DPPH inhibition by two extracts are shown in Table 1.
Meanwhile, TEAC method was adopted to assess the ABTS+ radical
scavenging ability of two extracts. Data from both assays clearly
demonstrated that the 14th day (red) extract possessed higher
antioxidant capacity than the 2nd day (green) extract, suggesting
that astaxanthin might be the major component responsible for
the antioxidant properties of C. zofingiensis extracts.
It is known that oxidative reactions and reactive oxygen species
(ROS) are involved in the nonenzymatic glycation process and
greatly promote the AGE formation. Under the influence of transi-
Fig. 1. The accumulation of total carotenoids and astaxanthin by heterotrophic C.
zofingiensis. (j) total carotenoids; (h) astaxanthin.
1632 Z. Sun et al. / Food Chemistry 126 (2011) 1629–1635

Table 1 Table 2
Antioxidant activity of the 2nd day (green) and the 14th day (red) C. zofingiensis Effects of the 2nd day (green) and the 14th day (red) C. zofingiensis extracts
extracts determined by DPPH and TEAC methods. (1.0 mg mL1) as well as fractions separated from the 14th day (red) C. zofingiensis
extract on protein carbonyl formation and thiol group content of glucose-modified
Sample TEACa DPPH (IC50)b BSA.
The 2nd day (green) extract 3.04 ± 0.21 2.96 ± 0.05 Sample Protein carbonyl content Thiol group
The 14th day (red) extract 3.85 ± 0.27 1.04 ± 0.02 (nmol mg1 protein) (pmol mg1 protein)
Each value represents the mean ± SD (n = 3). Controla 0.79 ± 0.17 14.27 ± 0.78
a
TEAC was expressed as mg Trolox equivalent g1 dried extract. Controlb 5.75 ± 0.28 7.64 ± 0.37
b
IC50 was expressed as mg mL1. The 2nd day 3.85 ± 0.18* 8.05 ± 0.35*
(green) extract
The 14th day (red) 3.37 ± 0.14* 8.96 ± 0.25*
extract
tion metal ions, especially copper and iron ions, the glucose autox-
Neoxanthin 5.63 ± 0.16 7.45 ± 0.51
idation occurs, producing numerous hydroxyl radicals which in Free astaxanthin 4.59 ± 0.16* 7.56 ± 0.43
turn contribute to protein impairments (Wolff & Dean, 1987). To Free adonixanthin 5.15 ± 0.13 7.82 ± 0.33
investigate whether C. zofingiensis extracts could inhibit glucose Astaxanthin 3.61 ± 0.13* 8.43 ± 0.31*
monoester
autoxidation, they were reacted with sodium benzoate with the
Astaxanthin diester 2.87 ± 0.12* 9.76 ± 0.33**
presence of glucose and Cu2+. The amount of benzoate hydroxyl- Adonixanthin ester 4.14 ± 0.08* 9.25 ± 0.30**
ation induced by glucose autoxidation was detected by measuring Lutein + zeaxanthin 4.70 ± 0.22* 8.78 ± 0.35*
its characteristic fluorescence. As shown in Fig. 2A, both green and AG (1 mM) 2.58 ± 0.14* 9.34 ± 0.48*
red extracts exhibited significant inhibitory effects on the forma- Each value represents the mean ± SD (n = 3).
tion of benzoate hydroxylation. Increasing the concentration a
Reaction mixture without glucose.
(500–1000 lg mL1) was associated with lower values of hydrox- b
Reaction mixture in the presence of glucose.
ylated benzoate. At the concentration of 1000 lg mL1, algal ex-
*
p < 0.05 significant differences with controlb
**
p < 0.01 significant differences with controlb
tracts suppressed the formation of benzoate hydroxylation by
62.33% and 52.95%, respectively. These results suggested that C.

A 80
% Inhibition of benzoate hydroxylation

**
**
60

40 **
**
**
20
*

0
200 500 1000
Concentration (µg mL-1)

The 2nd day (Green) The 14th day (Red)

% Inhibition of benzoate hydroxylation

B 80
**
60

**
40
*
* *
20 *

0
in
r

r
n

ste

ste

te
hi

hi

hi

th
es
nt

nt

-e

an
an

o-

in
xa

xa

di

ax
x

on

th
eo

sta

ni

in

e
an
m

+z
do

h
N

ea

nt

ix
in

in
ea

xa
e

n
th

te
Fr

do
e

sta
n

Lu
Fr

xa

A
A
sta
A

Fig. 2. Inhibitory effects of the 2nd day (green) and the 14th day (red) C. zofingiensis extracts (Panel A) as well as fractions separated from the 14th day (red) C. zofingiensis
extract (Panel B) on the generation of benzoate hydroxylation induced by glucose autoxidation. Each value represents the mean ± SD (n = 3). Fluorescent intensities of
solutions with inhibitors were significantly different from that of the control solution, as marked with ⁄p < 0.05 or ⁄⁄p < 0.01.
 Z. Sun et al. / Food Chemistry 126 (2011) 1629–1635 1633

zofingiensis extracts especially the red one might possess capacities detected. Results showed that the original BSA maintained a very
to scavenge hydroxyl radical production, and/or chelate transition low level of carbonyl content (0.79 nmol mg1). However, after the
metals, resulting in the decrease of benzoate hydroxylation. incubation with glucose for 14 days, the carbonyl content signifi-
In addition to glucose autoxidation, ROS may also lead to the cantly increased to 5.75 nmol mg1 in the glycated BSA. The
modification of proteins, which is known as the oxidative modifica- 1.0 mg mL1 red and green extracts significantly decreased it to
tion. It takes place when protein side chains are attacked by free rad- 3.37 nmol mg1 and 3.85 nmol mg1, respectively. On the other
icals accompanying by the generation of carbonyl groups (aldehydes hand, thiol groups were measured to determine the protein modifi-
and ketones). Some protein carbonyl derivatives (ketoaldehydes and cation by excessive free radicals. Results showed that due to the gly-
ketoamines) could also be generated during the glycoxidation pro- cation-induced protein oxidation, a sharp decrease of thiol group
cess (Dalle-Donne, Rossi, Giustarini, Milzani, & Colombo, 2003). Car- content occurred in BSA, reducing from 14.27 to 7.64 pmol mg1.
bonyl groups are chemically stable and therefore regarded as useful However, both C. zofingiensis extracts displayed significant effects
biomarkers of oxidative damage (Stadtman & Levin, 2000). Another to stop the loss of thiol groups at the concentration of 1.0 mg mL1,
commonly used indicator for protein oxidation is the loss of thiol whereas the green extract displayed relatively weaker effects. In
groups, which is often employed to quantify the modification of cys- both assays, 1 mM AG solution, a commonly used antiglycative
teine residues (Bourdon, Loreau, & Blache, 1999). To determine the agent was used as a positive control, which exhibited good anti-pro-
effects of C. zofingiensis extracts on glycation-induced protein oxida- tein oxidation capacities.
tion, the extents of protein carbonyl formation and thiol groups were To assess the effects of C. zofingiensis extracts on the formation of
measured (Table 2). Carbonyl groups could be derivatised with 2,4- AGEs, a commonly used in vitro model, BSA-glucose system was
dinitrophenylhydrazine (DNPH), forming stable 2,4-dinitrophenyl adopted. The sugar-mediated fluorescence intensity, which is the
(DNP) hydrazone products, which were spectrophotometrically characteristic of AGEs was measured at excitation and emission of

100
% Inhibition of AGE

** **
80 **
formation

60
**
40 **
*
20 *

0
100 200 500 1mM AG
Concentration (µg mL-1)
The 2nd day (Green) The 14th day (Red)
% Inhibition of AGE formation

100
** **
80

60
*
40
* *
20 *

0
in
r

r
n

in

te

ste

te

AG
th
hi

hi
h

es

es
-e

an
nt

nt

nt

o-

M
n
di
xa

xa

xa

x
hi
on

1m
ea
in
eo

ta

ni

nt
m

+z
as

h
o

xa
N

nt
ad

in
ee

ni
hi

xa

te
ee

do
Fr

nt

sta

Lu
Fr

xa

A
A
sta
A

1 2 3 4 5 6 7

Fig. 3. Inhibitory effects of the 2nd day (green) and the 14th day (red) C. zofingiensis extracts (Panel A) as well as fractions separated from the 14th day (red) C. zofingiensis
extract (Panel B) on the formation of AGEs in BSA-glucose system. Each value represents the mean ± SD (n = 3). AG solution (1 mM) was used as a positive control. Fluorescent
intensities of solutions with inhibitors were significantly different from that of the control solution, as marked with ⁄p < 0.05 or ⁄⁄p < 0.01. Panel C: SDS–PAGE profile of
glycated protein. BSA (20 mg mL1) was incubated with glucose (800 mM) and NaN3 (0.2 mg mL1) in phosphate buffer (0.2 M, pH 7.4). All incubations were carried out at
37 °C for 2 weeks. 1. Mw marker; 2. glycated BSA; 3. AG (1 mM); 4. BSA; 5. the 14th day (red) C. zofingiensis extract (500 lg mL1); 6. the 2nd day (green) C. zofingiensis extract
(500 lg mL1); 7. astaxanthin diester.
1634 Z. Sun et al. / Food Chemistry 126 (2011) 1629–1635
Astaxanthin diester
Degraded Chlorophyll
Astaxanthin monoester
Adonixanthin ester
Chlorophyll a
Chlorophyll b
Free astaxanthin
Free adonixanthin
Lutein + zeaxanthin
Neoxanthin
Fig. 4. Fractions separated by thin layer chromatography (TLC) from the 14th day
(red) C. zofingiensis extract.(For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
330 and 410, respectively (Cervantes-Laurean, Jacobson, & Jacob-
son, 1996). In Fig. 3A, at different concentrations (100–
500 lg mL1), both extracts significantly suppressed the fluores-
cence intensity in a dose-dependent manner, whereas the red
one, i.e. the astaxanthin-rich extract, exhibited much stronger
inhibitory effects. Its inhibitory rate was 80.64% at the concentra-
tion of 500 lg mL1, higher than the effect of the green extract
(inhibitory rate: 67.26%) and comparable to 1 mM AG solution
(inhibitory rate: 81.23%). Meanwhile, the formation of AGE prod-
ucts was detected using SDS–PAGE (Fig. 3C). After incubation at
37 °C for 14 days, BSA was glycated and displayed a larger molecu-
lar weight (90 kDa) than the original one (66 kDa), suggesting the
occurrence of cross-linking reactions. The intensity of protein bands
was calculated. Results showed that at the concentration of
500 lg mL1, both green and red extracts markedly suppressed
the formation of glycation products.
Based on these evidences, the potential antiglycoxidative effects
of C. zofingiensis have been strongly supported. The algal extracts
with antioxidant activity can significantly reduce the glucose
autoxidation, diminish glycation-induced protein oxidation and in-
hibit the formation of AGEs. These data supported the view that
antioxidant agents could be used as powerful glycation inhibitors.
Through the comparison between two extracts of C. zofingiensis, it
was shown that the higher the astaxanthin content, the stronger
the antiglycoxidative capacity. These evidences suggested that
antiglycoxidative properties of C. zofingiensis might be due to the
presence of astaxanthin. To confirm whether astaxanthin is the
main compound responsible for observed activities, and to get a
clear understanding of its and/ or other components’ contributions
in glycoxidation reactions, the astaxanthin-rich extract was frac-
tionated by TLC. Seven major fractions were obtained. HPLC results
revealed that they were astaxanthin diester (containing tiny
amounts of beta-carotene), astaxanthin monoester, adonixanthin
ester, free astaxanthin, free adonixanthin, lutein + zeaxanthin and
neoxanthin, respectively (Fig. 4). These fractions were subse-
quently re-dissolved in 50 lL of acetone, of which 30 lL were used
in following assays.
In the BSA-glucose system, fractions showed very diverse abili-
ties to suppress the fluorescence intensity (Fig. 3B). The astaxanthin
diester was found to have the strongest antiglycative capacity
(inhibitory rate: 83.59%), higher than the effect of 1 mM AG solution
(inhibitory rate: 81.23%). Such results were also confirmed on the
SDS–PAGE profile (Fig. 3C). Inhibitory rates of remaining fractions
decreased in the order of astaxanthin monoester (29.48%) > free
astaxanthin (16.89%) > lutein + zeaxanthin (15.76%) > adonixanthin
ester (13.63%), whereas neoxanthin and free adonixanthin showed
no effects. In the glucose autoxidation assay, except for the fraction
of neoxanthin showing no activity, other fractions significantly
inhibited the generation of benzoate hydroxylation (Fig. 2B). Asta-
xanthin diester was found to exhibit the highest inhibitory rate
(62.27%). Inhibitory rates of remaining fractions decreased in the or-
der of astaxanthin monoester (35.07%) > lutein + zeaxanthin
(20.90%) > adonixanthin ester (14.81%) > free astaxanthin (14.34%)
> free adonixanthin (9.24%). In the protein carbonyl assay, the frac-
tion of astaxanthin diester exhibited the highest inhibitory activity
(Table 2). It decreased the carbonyl content from 5.75 to
2.87 nmol mg1, followed by astaxanthin monoester (3.61 nmol
mg1) and adonixanthin ester (4.14 nmol mg1). Other fractions
showed relatively weaker effects in which neoxanthin failed to
diminish the carbonylation. In the thiol group assay, expect for frac-
tions of neoxanthin, free adonixanthin and free astaxanthin, all other
fractions displayed significant effects to stop the loss of thiol groups.
astaxanthin diester was found to exhibit the highest activity (thiol
group content: 9.76 pmol mg1), followed by adonixanthin ester
(9.25 pmol mg1), lutein + zeaxanthin (8.78 pmol mg1) and asta-
xanthin monoester (8.43 pmol mg1).
These results, taken together, clearly demonstrated the hypoth-
esised roles of astaxanthin in antiglycoxidative activities of C. zof-
ingiensis extracts. Although several compounds e.g. adonixanthin
and lutein + zeaxanthin were also involved in the blockage of gly-
cation cascade, astaxanthin diester was the major contributor, pro-
viding the most significant protective effects. The esterified form of
astaxanthin in C. zofingiensis is believed to be a more stabilised
structure to maintain the high antioxidant ability (Kobayashi &
Sakamoto, 1999). LC–MS was adopted to identify the fatty acid
composition of astaxanthin diester. According to the LC–MS re-
sults, the detected fragment of 862.4 was presumably identified
as astaxanthin-octadecanoic acid (C18:0), 282.4 as oleic acid
(C18:1), and 257.0 as palmitic acid (C16:0). These LC–MS results
suggested that the fatty acids in these astaxanthin diesters are
highly possible to be octadecanoic acid, oleic acid and palmitic
acid, which were also in accordant to the GC–MS detection and li-
brary searching results (Grynbaum et al., 2005).
4. Conclusions
The green microalga C. zofingiensis was systematically evaluated
for its antiglycoxidative properties for the first time in this study.
Results showed that the C. zofingiensis extracts possessed antioxi-
dant capacities and significant inhibitory effects on the glycation
cascade in various model systems. Further studies revealed that
astaxanthin was the major component contributing to observed
activities. Taken together, these findings supported the possibility
of using natural antioxidants as glycation inhibitors and suggested
the beneficial roles of astaxanthin, which is an essential nutritional
component for human health in diabetes. The astaxanthin-rich
microalga C. zofingiensis therefore might be regarded as a potential
therapeutic and preventive agent for diabetes and its complica-
 Z. Sun et al. / Food Chemistry 126 (2011) 1629–1635
tions. In future, detailed cell model studies and in vivo investiga-
tions are needed to provide more evidence.
Acknowledgement
This work was supported by the GRF Grand of the Research
Grants Council of Hong Kong.
References
Ahmad, M. S., & Ahmad, N. (2006). Antiglycation properties of aged garlic extract:
Possible role in prevention of diabetic complications. Journal of Nutrition, 136,
796–799.
Ardstani, A., & Yazdanparast, R. (2007). Inhibitory effects of ethyl acetate extract of
Teucrium polium on in vitro protein glycoxidation. Food and Chemical Toxicology,
45, 2402–2411.
Bar, E., Rise, M., Vishkautsan, M., & Arad, S. (1995). Pigments and structural changes
in Chlorella zofingiensis upon light and nitrogen stress. Journal of Plant Pysiology,
146, 527–534.
Baroli, I., Do, A. D., Yamane, T., & Niyogi, K. K. (2003). Zeaxanthin accumulation in
the absence of a functional xanthophyll cycle protects Chlamydomonas
reinhardtii from photooxidative stress. Plant Cell, 15, 992–1008.
Blois, M. S. (1958). Antioxidant determination by the use of a stable free radical.
Nature, 181, 1190–1200.
Bourdon, E., Loreau, N., & Blache, D. (1999). Glucose and free radicals impair the
antioxidant properties of serum albumin. The FASEB Journal, 13, 233–244.
Brownlee, M. (2001). Biochemistry and molecular cell biology of diabetic
complications. Nature, 414, 813–820.
Cervantes-Laurean, D., Jacobson, E. L., & Jacobson, M. K. (1996). Glycation and
glycoxidation of histones by ADP-ribose. Journal of Biological Chemistry, 217,
10461–10469.
Chen, F. (1996). High cell density culture of microalgae in heterotrophic growth.
Trends in Biotechnology, 14, 421–426.
Dalle-Donne, I., Rossi, R., Giustarini, D., Milzani, A., & Colombo, R. (2003). Protein
carbonyl groups as biomarkers of oxidative stress. Clinica Chemica Acta, 329,
23–38.
Duraisamy, Y., Gaffney, J., Slevin, M., Smith, C. A., Williamson, K., & Ahmed, N.
(2003). Aminosalicylic acid reduces the antiproliferative effect of
hyperglycaemia, advanced glycation endproducts and glycated basic
fibroblast growth factor in cultured bovine aortic endothelial cells:
Comparison with aminoguanidine. Molecular and Cellular Biochemistry, 246,
143–153.
Ellman, G. L. (1959). Tissue sulfhydryl groups. Archives of Biochemistry and
Biophysics, 82, 70–77.
1635
Grynbaum, M. D., Hentschel, P., Putzbach, K., Rehbein, J., Krucker, M., Nicholson, G.,
et al. (2005). Unambiguous detection of astaxanthin and astaxanthin fatty acid
esters in krill (Euphausia superba Dana). Journal of Separation Science, 28,
1685–1693.
Hunt, J. V., Bottoms, M. A., & Mitchinson, M. J. (1993). Oxidative alterations in the
experimental glycation model of diabetes mellitus are due to protein-glucose
adduct oxidation. The Biochemical Journal, 291, 529–535.
Hunt, J. V., Dean, R. T., & Wolff, S. P. (1988). Hydroxyl radical production and
autoxidative glycosylation. Glucose autoxidation as the cause of protein
damage in the experimental glycation model of diabetes mellitus and aging.
Biochemical Journal, 256, 205–212.
Ip, P. F., & Chen, F. (2005). Production of astaxanthin by the green microalga
Chlorella zofingiensis in the dark. Process Biochemistry, 40, 733–738.
Kobayashi, M., & Sakamoto, Y. (1999). Singlet oxygen quenching ability of
astaxanthin esters from the green alga Haematococcus pluvialis. Biotechnology
Letters, 21, 265–269.
Kuhl, A. (1962). Zur physiologie der speicherung kondensetem organischer
phosphate in Chlorella. In: Beiträge zur Physiologie und Morphologie der
Algen. Gustav Fischer Verlage, Stuttgart, West Germany.
Miki, W. (1991). Biological functions and activities of animal carotenoids. Pure and
Applied Chemistry, 63, 141–146.
Nathan, D. M. (1993). Long-term complications of diabetes mellitus. The New
England Journal of Medicine, 328, 1676–1685.
Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999).
Antioxidant activity applying an improved ABTS radical cation decolorization
assay. Free Radical Biology and Medicine, 26, 1231–1237.
Stadtman, E. R., & Levin, R. L. (2000). Protein oxidation. Annals of the New York
Academy of Sciences, 899, 191–208.
Sun, Z., Peng, X. F., Liu, J., Fan, K. W., Wang, M. F., & Chen, F. (2010). Inhibitory effects
of microalgal extracts on the formation of advanced glycation endproducts
(AGEs). Food Chemistry, 120, 261–267.
Sun, N., Wang, Y., Li, Y. T., Huang, J. C., & Chen, F. (2008). Sugar-based growth,
astaxanthin accumulation and carotenogenic transcription of heterotrophic
Chlorella zofingiensis (Chlorophyta). Process Biochemistry, 43, 1288–1292.
Thornalley, P. J. (2003). Use of aminoguanidine (pimagedine) to prevent the
formation of advanced glycation endproducts. Archives of Biochemistry and
Biophysics, 419, 31–40.
Vlassara, H., & Palace, M. R. (2002). Diabetes and advanced glycation endproducts.
Journal of Internal Medicine, 251, 87–101.
Wolff, S. P., & Dean, R. T. (1987). Glucose autoxidation and protein modification: The
potential role of ‘autoxidative glycosylation’ in diabetes mellitus. The
Biochemical Journal, 245, 243–250.
Xi, M., Hai, C., Tang, H., Chen, M., Fang, K., & Liang, X. (2008). Antioxidant and
antiglycation properties of total saponins extracted from traditional Chinese
medicine used to treat diabetes mellitus. Phytotherapy Research, 22, 228–237.