Structural-morphological and biological properties of selenium nanoparticles stabilized by bovine serum albumin

Article abstract of DOI:10.1134/S0036024407070291

Nanostructures formed during the reduction of ionic selenium in the selenite-ascorbate redox system in an aqueous solution of bovine serum albumin (BSA) were studied using static and dynamic light scattering and flow birefringence. It was established that this process results in the formation of stable aggregates of selenium nanoparticles that adsorb BSA molecules. It was found that highly-ordered superhigh-molecular-weight spherical nanostructures with high density and unique morphology are formed. Experiments with a cell culture of promyelocytic leukemia HL-60 showed that BSA adsorbed on selenium nanoparticles can inhibit the growth of tumor cells and deactivate free radicals with an efficiency comparable with that of sodium selenite.

Full text of DOI:10.1134/S0036024407070291

ISSN 0036-0244, Russian Journal of Physical Chemistry A, 2007, Vol. 81, No. 7, pp. 1170–1173. © Pleiades Publishing, Ltd., 2007.
Original Russian Text © S.V. Valueva, L.N. Borovikova, V.V. Koreneva, Ya.I. Nazarkina, A.I. Kipper, V.V. Kopeikin, 2007, published in Zhurnal Fizicheskoi Khimii, 2007,
Vol. 81, No. 7, pp. 1329–1333.
BIOCHEMICAL
CHEMISTRY
Structural–Morphological and Biological Properties
of Selenium Nanoparticles Stabilized by Bovine Serum Albumin
a
a
b
a
a
S. V. Valueva , L. N. Borovikova , V. V. Koreneva , Ya. I. Nazarkina , A. I. Kipper ,
†
a
and V. V. Kopeikin
a
Institute of High-Molecular Weight Compounds, Russian Academy of Sciences, St. Petersburg, Russia
b
Institute of Cytology, Russian Academy of Sciences, St. Petersburg, Russia
Received May 29, 2006
Abstract—Nanostructures formed during the reduction of ionic selenium in the selenite–ascorbate redox sys-
tem in an aqueous solution of bovine serum albumin (BSA) were studied using static and dynamic light scat-
tering and flow birefringence. It was established that this process results in the formation of stable aggregates
of selenium nanoparticles that adsorb BSA molecules. It was found that highly-ordered superhigh-molecular-
weight spherical nanostructures with high density and unique morphology are formed. Experiments with a cell
culture of promyelocytic leukemia HL-60 showed that BSA adsorbed on selenium nanoparticles can inhibit the
growth of tumor cells and deactivate free radicals with an efficiency comparable with that of sodium selenite.
DOI: 10.1134/S0036024407070291
The discovery that selenium is highly important for with ascorbic acid in the presence of globular BSA, to
human health [1] gave impetus to the development of discuss the nature of its interactions with nanoparticles, to
0
preventive and therapeutic drugs based on inorganic determine the morphology of BSA–Se nanostructures
and organic selenium compounds. The distinctions formed in aqueous solutions, and to assess in vitro the
between the vital, therapeutic, and toxic doses of these potentialities of these nanocomposites as antitumor and
forms of selenium are extremely small. The toxicity of antioxidant agent in comparison with sodium selenite.
selenium decreases in going from its ionic forms to
organic compounds, especially containing zero-oxida-
tion-state selenium. The assessment of the specific bio-
EXPERIMENTAL
logical activity of zero-valence selenium nanoparticles
obtained by chemical reduction of its ionic forms in the
absence of stabilizers [2] and by biotechnological
methods involving bacteria capable of accumulating
and reducing ionic selenium [3] showed that the activ-
ity of these nanoparticles does not exceed 2% of the
activity of sodium selenite, a medicinal agent widely
used in medicine. In [4], however, it was established
that addition of bovine serum albumin to the sodium
selenite—glutathione redox system results in the for-
We used selenious and ascorbic acids with main
substance content not less than 99.99%. Before experi-
ments, BSA (Calbiochem, fraction V, grade B) was
additionally purified by gel-filtration on a Sefadex G-
1
00 (the eluent was distilled water).
The reduction of ionic selenium was performed at a
selenium concentration in its aqueous solution of
0
.01% and a Se : BSA mass ratio of ν = 0.1, conditions
corresponding to a complete saturation of the adsorp-
tion capacity of nanoparticles in the poly(vinylpyrroli-
done)–selenium–water system [5].
0
mation of red amorphous nanoselenium (a-nano-Se ),
which is seven times less toxic than sodium selenite but
retains the profile and level of the biological activity of
sodium selenite. Hence, it was supposed that selenium
is bound to albumin by means of relatively weak nonco-
valent interactions, the nature of which was not exam-
ined by the authors of [4]. The assessment of the biolog-
ical activity of selenium resuspended after centrifuging
the precipitate [4] showed that, in the absence of albu-
min, selenium does not exhibit the biological activity
characteristic of the selenium–albumin composite.
The kinetic measurements were performed at 20°ë
and pH 3.5 on a Specord M 40 spectrophotometer. The
time evolution of the optical density of the solution dur-
ing the reduction of selenous acid was recorded at λ =
3
20 nm.
The method of static light scattering [6] was used to
determine the molecular masses (MMs) (M and M*
w
w
for BSA and nanostructures, respectively) and root-
mean-square inertia radii R* of the nanostructure and
g
The aim of this work is to synthesize nanoparticles
of zero-valence selenium by reduction of selenious acid
its affinity for water (determined from the second virial
coefficient A* ). The number of BSA macromolecules
2
†
Deceased.
(N*) adsorbed on the surface of Se nanoparticles was
1
170

STRUCTURAL–MORPHOLOGICAL AND BIOLOGICAL PROPERTIES
1171
determined from the ratio between the MMs of BSA
In the general case, at dn/dc ≠ 0 (dn/dc is the refrac-
and the nanostructure formed. To measure the reduced tive index increment of the solution, 0.146 for the
0
scattering intensity of the solutions R , we used a Fica BSA–Se –ç é system), the experimental value of [n]
θ
2
photogoniodiffusometer. The wavelength of vertically is the sum of three contributions: [n] = [n]
+ [n]fs + [n]
is the intrinsic anisotropy, [n]fs is the micro-
is the macroform effect [6]. The
depends on the particle shape asymme-
,
e
f
polarized incident light was 546.1 nm. Measurements where [n]
e
were performed for angles of scattering within θ = 30°– form effect, and [n]
f
1
50°. The refractive index increment dn/dcBSA (cBSA is quantity [n]
f
try p as
the BSA concentration) was determined from refrac-
tometric measurements performed on an IRF-23
instrument.
2
s
2
2
[
n] = ((n + 2)/3) (M (dn/dc) f(p))/30πRTn
*
f
w
s
The experimental data on light scattering for the
nanocomposite solutions were processed in accordance
with the Zimm procedure, by double extrapolation (to
(2)
2
=
const × M*(dn/dc) f (p),
w
2
where n is the refractive index of the solvent, T is the
cBSA = 0, θ = 0) of the KÒ /R versus sin (θ/2) + kÒ
s
BSA
θ
BSA
absolute temperature, R is the universal gas constant,
and f(p) is the tabulated function of the ratio between
the lengths of the axes of the solid ellipsoid that mod-
eled the particle [6].
dependence (K and k are the calibration and numerical
constants, respectively).
The method of dynamic light scattering [7] was used
to determine the average hydrodynamic size R* of the
h
The biological tests were performed with cells of
promyelocytic leukemia HL-60 from the bank of cell
cultures at the Institute of Cytology of the Russian
Academy of Sciences. The cells were cultured in an
incubator (5% ëé , 37°ë) on plastic Petri dishes in an
RPMI medium (Biolot, St. Petersburg) containing 15%
nanostructure at ÒBSA
0. The ratio between the
experimental values R* and R* was used to determine
g
h
the parameter ρ* = R* /R* , which characterizes the
g
h
2
conformation of the nanostructure [8].
The optical equipment of the setup used for measur- of cattle embryo blood serum (Gibco). To exclude
ing the dynamic light scattering consisted of an ALV-SP microbial contamination of the cultivation medium, we
goniometer (Germany) and a helium–neon (He–Ne) introduced antibiotic gentamycin in a concentration of
laser (Spectra-Physics), as a light source (λ = 632.8 nm, 80 µg/ml. The cells were preliminary planted in plastic
W ~ 20 mW). The correlation function of the scattered cups 35 mm in diameter or in 24-socket plates.
light intensity was measured on a Photo Cor-FS
Anteks, Russia) correlator with 288 channels. The cor-
relation function was analyzed using the Dynals soft-
ware package (Gelios, Russia).
HL-60 cells were treated 24 h after transplantation
by adding the corresponding agents to the cultivation
medium at certain time intervals (24 h to 3 days).
The cell distribution over the DNA content was
determined using flow cytofluorometry. To enhance the
(
The mean density of the spherical nanostructure was
calculated from M* and the root-mean-square radius membrane permeability, we treated the cells with
w
X-100 triton at a final concentration of 0.01% for 0.5 h
at room temperature (~20°C), added propidium iodide
of inertia by the formula
3
sph
Φ* = 3M* /4πN R
,
(1) (10 µg/ml, Sigma, USA), incubated the cells for 15 min
at 37°ë, and analyzed them on a ATC 300 flow cytoflu-
orometer (Bruker) at a flow rate of 20 µl/min for 3 min.
w
a
where Rsph = 1.29R* [9].
g
To determine the content of active oxygen species
AOS) in the cells, the cultivation medium was treated
with dihydroethidium bromide (30 min before the anal-
ysis) so as to attain a final concentration of 5 µmol/l.
The cells were washed twice with a PBS solution
The molecular dispersity of the test solutions of
nanostructures was determined by the flow birefrin-
gence method [6] from the dependences of the birefrin-
gence ∆n on the velocity gradient g and the BSA con-
centration Ò. We used a titanium dynamooptimeter with
a h = 4.0 cm and a rotor–stator gap of ∆r = 0.03. The
flow birefringence measurements were performed with
thermostatted solution (at 21°C) to avoid viscosity vari-
ations and optical distortion caused by temperature gra-
(
(
20 mmol/l phosphate buffer, pH 7.4, 0.1 mol/l NaCl).
The fluorescence of ethidium bromide was analyzed on
an ATC 300 flow cytofluorometer (Bruker).
The AOS yield in the cells was stimulated by adria-
dients. The setup was calibrated with phenylethyl alco- mycin at a concentration of 10 ng/ml.
hol which has a significant birefringence increment
1
1
(
∆n/g = 1.7 × 10 ), and with polystyrene–bromoform
RESULTS AND DISCUSSION
solution. The error in determination of the characteris-
The reduction of ionic selenium was found to be a
first-order reaction in the selenious acid concentration,
tic birefringence, [n] = lim (∆n /gÒBSAη0] (where
g → 0, c → 0
–3
–1
η0 is the solvent viscosity) did not exceed 10%. The
with rate constants k* of 2.5 × 10 s (Table 1) and
–3
–1
measurements were performed at g < g , where g is the 1.7 × 10
s
in the presence and absence of BSA,
k
k
velocity gradient of onset of flow turbulence.
respectively. Thus, the protein accelerates the reaction.
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 81 No. 7 2007

1
172
VALUEVA et al.
Table 1. Rate constant of the first-order reaction in selenious acid in the presence of BSA and the structura–morphological
characteristics of the selenium-containing BSA-based nanostructures under study
3
× 104,
A
*
3
Φ*,
k* × 10 ,
Mw × 10–6
2
R* , nm
R* , nm
ρ*
p
–
1
N*
g
h
3
s
2
g/cm
cm mol/g
0.2
2
.5
70
1000
50
50
1.0
1.2
0.1
According to the static light scattering data, the reduced optical anisotropy ∆n/gÒ η , we determined
BSA
0
4
–7
4
2
2
molecular mass M of BSA is 7.0 × 10 while that of the the characteristic birefringence [n] = 5 × 10 cm s /g .
w
7
nanostructure formed, 7.0 × 10 , i.e., 1000 times higher
Table 1).
0
The value of the parameter p For the BSA–Se –ç é
2
(
system calculated in the approximation of [n] ~ [n]f
suggest that the nanostructure is almost spherical:  =
For free BSA macromolecules in an aqueous solu-
tion, the mean hydrodynamic radius R was found to be
1
.2 (Table 1), as is the case with the BSA–H é system
2
h
(
at pH 6.0,  = 1.0 [12]).
~
[
15 nm, approximately equal to the value reported in
10, 11]. The value of R* for selenium nanoparticles
formed in the absence of the protein stabilizer was esti-
mated as 170 nm [5]. The size of the BSA–Se nano-
structure determined by dynamic and static light scat-
tering is smaller (the root-mean-square radius of inertia
and the mean hydrodynamic radius are equal: 50 nm,
Table 1). The ratio ρ* = R* /R* = 1 suggests that the
nanostructure formed is spherical [8].
The parameter Φ* was evaluated from the calcu-
h
0
3
lated density of the BSA–Se nanostructure: 0.1 g/cm
(
Table 1, Eq. (1)). This value substantially exceeds the
0
density of free BSA macromolecules in water [10].
A comparison of the hydrodynamic radius of the
free BSA macromolecule (R = 15 nm) with the
h
radius of the nanostructure (R* = 50 nm) showed
h
g
h
0
that R ꢀ R* . Thus, the BSA–Se superhigh-molecu-
h
h
lar-weight nanosystem can be thought of as a spherical
dense polynuclear nanostructure or a mononuclear
dense sphere composed of a selenium nucleus and mul-
tilayer protein shell comprised of about 1000 BSA
macromolecules.
0
The second virial coefficient A* for the BSA–Se –
2
–
5
3
2
ç O system was found to be 2 × 10 cm mol/g
Table 1), a value typical of ideal solutions. This value
is considerably lower than that for BSA in water or in
other solvents [10].
2
(
0
The formation of a BSA–Se nanostructure and sta-
bilization of selenium nanoparticles are determined by
the adsorption processes caused by hydrophobic inter-
actions of the nonpolar areas of macromolecules with
the high-energy surface of nanoparticles [5] and by
intermolecular electrostatic interactions of polyam-
pholytes of protein nature.
Note that the molecular-dispersed state of solutions
0
of the BSA–Se –ç é system remained unchanged over
2
the entire concentration range covered (ÒBSA), as can be
seen from the character of the ∆n = f(g) dependence: at
g < g , they are closely approximated by straight lines
k
passing through the origin of the coordinates at all Ò
BSA
It is well-known that sodium selenite exhibits an
antitumor effect, in particular, with respect to leukemia
values. From the concentration dependence of the
[
13, 14]. In this work, by the example of promyelocytic
leukemia HL-60 cells, we demonstrated that sodium
Table 2. Dynamics of changes in the number of live cells of
0
human promyelocytic leukemia in the absence and in the selenite and the BSA–Se nanocomposite reduce the
presence of selenium compounds during the first (I), second number of cancer cells in the G -phase of the cellular
1
(
II), and third (III) days
cycle (a diploid 2n DNA set) and increase the number
of them in the G -phase (a 4n DNA set).
2
Additives to the culture
medium
I
II
III
During the first 24 h, selenium nanocomposite and
sodium selenite inhibit the growth of HL-60 cells pro-
portionally to the increase in their concentration. The
Reference run (no additives)
BSA
499
350
442
48
780
758
955
24
1342
1170
1313
18
0
addition of sodium selenite and BSA–Se at concentra-
tions of 10 and 1–50 µmol/l, respectively, resulted in
enhancement of the cell proliferation on the second day
of incubation (Table 2). The addition of sodium selenite
and selenium nanocomposite so as to obtain a final con-
centration of 200 µmol/l resulted in cancer cells death
on the third day. Thus, nanoparticles of zero-valence
Sodium selenium, 10 µmol/l
Sodium selenium, 100 µmol/l
Sodium selenium, 200 µmol/l
36
22
0
0
BSA–Se , 1 µmol/l
250
250
250
40
702
276
168
18
1011
383
221
0
0
BSA–Se , 50 µmol/l
0
selenium entering into the composition of BSA–Se
0
BSA–Se , 100 µmol/l
nanocomposite retain the profile of biological activity
of ionic selenium, an observation indicative of the inter-
0
BSA–Se , 200 µmol/l
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 81 No. 7 2007

STRUCTURAL–MORPHOLOGICAL AND BIOLOGICAL PROPERTIES
1173
Table 3. AOS level in HL-60 cells as compared to the refer- as compared with adriamycin; for 100 µmol/l, this level
ence one: (A) in the absence and (B) presence of adriamycin
decreased to the control value (in the absence of adria-
mycin) but the number of live cells decreases (Table 4).
0
0
Cells,
%
Na SeO Na SeO BSA–Se BSA–Se
2 3 2 3
Control
Thus, a conclusion may be drawn that selenium in
10 µmol/l 100 µmol/l 10 µmol/l 100 µmol/l
0
the composition of the Se –BSA nanosystem decreases
A
B
100
150
92
92
87
87
the AOS level in HL-60 cells.
100
150
150
100
ACKNOWLEDGMENT
This work was supported by the Russian Foundation
for Basic Research, project no. 05-03-32842.
Table 4. Content of live cells (LC) (A) in the absence and
B) presence of adriamycin, %
(
0
0
Na SeO Na SeO BSA–Se BSA–Se
2
3
2
3
LC Control
1
0 µmol/l 100 µmol/l 10 µmol/l 100 µmol/l
REFERENCES
1
2
. M. P. Rayman, Lancet 356 (9225), 233 (2000).
. C. E. Schlekat, P. R. Dowdle, B. G. Lee, et al., Environ.
Sci. Technol. 34, 4504 (2000).
A
B
90
85
85
88
85
65
85
88
85
72
Note: For notations, see Table 3.
3. G. F. Combs, C. Garbisu, B. G. Lee, et al., Biol. Trace
Elem. Res. 52 (3), 209 (1996).
4
. J. S. Zhang, X. Y. Gao, L. D. Zhang, and Y. P. Bao, Bio-
crossing of the metabolic ways of sodium selenite and
factors 15 (1), 27 (2001).
0
the BSA–Se nanosystem and of the effect of their com-
5. V. V. Kopeikin, S. V. Valueva, A. I. Kipper, et al.,
Vysokomol. Soedin., Ser. A 45 (4), 615 (2003) [Polymer
Science, Ser. A 45 (4), 374 (2003)].
mon metabolite.
One disadvantages of adriamycin, an anthracycline-
series antibiotic that is widely used in chemotherapy of
tumors, is the generation of AOS [15]. It damages DNA
or causes lipid peroxidation. The AOS level is normal-
ized by selenoproteins, the activity of which depends
on the selenium concentration. Therefore, we supposed
that adriamycin-generated AOS may be neutralized by
selenium.
6
. V. N. Tsvetkov, V. E. Eskin, and S. Ya. Frenkel’, Struc-
ture of Macromolecules in Solutions (Nauka, Moscow,
1
964) [in Russian].
7
8
. W. Brown, Dynamic Light Scattering: The Method and
Same Application (Clarendon, Oxford, 1993).
. B. W. Burchard, Laser Light Scattering in Biochemistry,
Ed. by S. E. Harding, D. B. Satelle, and V. A. Bloomfild
(Royal Soc. Chem., Cambridge, 1992).
To verify this assumption, we stimulated the cells by
adding adriamycin in a concentration of 10 ng/ml, i.e.,
within the range of its concentrations in blood 1 h after
an injection and later until the next injection [15]; this
is a typical effective therapeutic concentration. Note
that this concentration causes almost no damages in
DNA; the effect concerns only AOS [16].
9
. N. V. Pogodina and N. V. Tsvetkov, Macromolecules 30
17), 4897 (1997).
(
1
1
0. C. Tanford, Physical Chemistry of Macromolecules
Wiley, New York, 1961; Mir, Moscow, 1965).
1. A. L. Lehninger, Biochemistry: The Molecular Basis of
Cell Structure and Function, 2nd ed. (Worth, New York,
1975; Mir, Moscow, 1979).
(
At a sodium selenite concentration of 10 µmol/l, the 12. D. C. Carter and J. X. Ho, Adv. Protein Chem. 45, 153
1994).
cals by adriamycin corresponds to the control one in 13. D.-Y. Cho, U. Chung, and A. S. Chung, Biochem. Mol.
Biol. Int. 47 (5), 781 (1999).
00 µmol/l, the AOS level did not vary as compared 14. H. Zeng, J. Nutr. 132 (4), 674 (2002).
with that for pure adriamycin (Table 3). For the nano- 15. K. Sugimoto, M. Sasaki, K. Tamayose, and K. Oshimi,
(
AOS level associated with the induction of free radi-
the absence of adriamycin. At a concentration of
1
0
Se –BSA system, the addition of the complex at the
Br. J. Haematol. 105, 720 (1999).
concentration 10 µmol/l did not change the AOS level 16. D. A. Gewirtz, Biochem. Pharmacol. 57, 727 (1999).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 81 No. 7 2007