ISSN 0006-3509, Biophysics, 2018, Vol. 63, No. 3, pp. 325–329. © Pleiades Publishing, Inc., 2018.

Original Russian Text © A.G. Akopdzhanov, A.I. Borisova, G.A. Frolov, T.F. Shaikhutdinov, N.D. Shaikhutdinova, T.A. Fedotcheva, N.L. Schimanowsky, 2018, published in
Biofizika, 2018, Vol. 63, No. 3, pp. 438–443.

MOLECULAR BIOPHYSICS

The Dependence of the Magnetic-Resonance Contrast Imaging

Properties of Ultrasmall Nanoparticles of Complex Iron Oxide
on Their Chemical Composition
A. G. Akopdzhanova, *, A. I. Borisovaa, **, G. A. Frolovb, T. F. Shaikhutdinovb, N. D. Shaikhutdinovab,
T. A. Fedotchevaa, and N. L. Schimanowskya
a
Russian National Pirogov Research Medical University, ul. Ostrovityanova 1, Moscow, 117997 Russia
b
MISiS National University of Science and Technology, Leninskii pr. 4, Moscow, 119049 Russia
*e-mail: artura777@mail.ru
**e-mail: aarisa2012@yandex.ru
Received April 4, 2018

Abstract⎯In this paper, we considered the optimization of the magnetic-resonance imaging properties of
nanosystems that are based on ultrasmall nanoparticles of complex iron oxide by modifying their chemical
composition with Zn, Mn, Mg, or Co. The relationship between the chemical composition, crystal structure,
magnetic phase, and contrast-imaging properties of the nanoparticles was investigated. The comparative
cytotoxicity of the nanoparticle solutions on HeLa cells was studied. The prospects for these types of nano-
systems for the development of contrast MRI agents with the predominant effect on the T1 or T2 parameters
of the body tissues were shown.

Keywords: nanoparticles, complex iron oxide, chemical synthesis of preparations, MRI contrast agents, pro-
ton relaxometry
DOI: 10.1134/S0006350918030028

Magnetic nanoparticles of complex iron oxide and should form a stable colloid system, be biocompatible,
systems on their basis are promising materials for the and have a fixed pH that is close to the physiological
creation of novel contrast agents for magnetic-reso- value. The cytotoxic properties of nanomaterials
nance imaging. Our studies [1] and analysis of the lit- depend not only on the chemical composition, but
erature [2] demonstrate that systems of different-sized also on the characteristic size of the objects and sur-
nanoparticles of complex iron oxide make it possible face state. A method for synthesis of solutions of
to effectively increase the quality and contrast of MRI ultrasmall nanoparticles suggests the active use of sta-
images and have an important ability to accumulate in bilizers that prevent aggregation and make it possible
the tumor nidus [3]. to reduce the activity of the developed surface.
The magnetic-resonance contrast properties of For the optimum clinical application of ultrasmall
paramagnetic nanoparticles are determined by their nanoparticles it is necessary to increase their relax-
chemical composition, characteristic size, crystal ation characteristics and to achieve the dominant
structure, magnetic phase, and surface state. It may be effect of one of the main biophysical parameters, that
possible to control all the mentioned factors by select- is, the T1 or T2 proton relaxation time of the biological
ing modes for chemical synthesis of the nanoparticles. tissue. This paper considers the effect of modification
The main factor is the characteristic size, with 7– of the chemical composition of magnetite nanoparti-
12 nm nanoparticles forming superparamagnetic sys- cles, that is, partial replacement of divalent iron atoms
tems and being T1 and T2 contrast agents at the same in the magnetite structure with a divalent ligand atom,
time and 20–40 nm ferrimagnetic nanoparticles pre- on their relaxation activity. The ligands were magnetic
dominantly affecting the T2 relaxation time. These metals with a characteristic radius of the nucleus close
nanoparticles that exceed 30 nm are actively absorbed to that of the iron atom. The formation of nanoparti-
by the reticuloendothelial system and are efficient for cles with different chemical compositions and influ-
diagnostics of liver and lymph-node diseases [4, 5]. ence of the local magnetic fields that are created by
The use of nanoparticles as pharmaceutical sub- their solutions on the longitudinal and transverse
stances has several requirements beside the presence relaxation times of water-containing media in an
of the necessary physical properties: the nanoparticles external magnetic field have been discussed. The

325
326 AKOPDZHANOV et al.

results of the investigation of the cytotoxic properties all the cases remained constant for adequate compari-
of nanoparticles that were alloyed with different mag- son of the characteristics of the solutions.
netic atoms are given. To control the growth of the nucleus, citric acid
(2-hydroxy-1,2,3-propantricarbonic acid), which is
MATERIALS AND METHODS biologically acceptable and makes it possible to con-
trol the dynamics of the intermediate pH values, was
The nanoparticle solutions that we studied were used. Sodium citrate was used as a stabilizer of the
obtained by modified chemical synthesis that was structure of the colloidal solution. When the stabilizer
based on hydrolysis of divalent and trivalent iron salts concentration was selected we followed the principle
and described earlier in details in [6]. of its minimum concentration (taking its negative
In selecting the mode of the chemical synthesis of influence on the contrast characteristics into account)
the nanoparticles, conditions that make it possible to that was sufficient to reach the necessary parameters
preserve the characteristic size of the main crystal and of the preservation of the size of the nanoparticles and
magnetic phase structure of magnetite were chosen. the solution stability. According to the results of toxic-
The selection of the ligand was determined by its mag- ity studies we selected the optimum concentrations
netic properties, the presence of divalent atoms, an and modes of introduction of the stabilizers, whose
atomic mass as close to that of iron as possible, and by concentrations for all the solutions were 12 g/L for cit-
the conditions of the biocompatibility of the end ric acid and 44 g/L for sodium citrate [7]. The large
nanoparticles. The chemical ligands for modification fraction of the nanoparticles was separated by centrif-
of the iron oxide nanoparticles were divalent atoms of ugation taking the minimization of the losses in the
paramagnetic manganese and magnesium, ferromag- concentration of the main active agent into account.
netic cobalt, and diamagnetic zinc. The initial The optimum modes, three 5-minute series of centrif-
reagents for the chemical synthesis were the corre- ugation at 800 rpm and filtration, were chosen. The
sponding sulfates of the divalent metals. The magnetic final pH value of the solution for all the samples
phase of the selected ligands was determined accord- was 7.2.
ing to the tabulated value of the magnetic permeabil-
The study of the stability of the solutions by peri-
ity. Manganese compounds are preferable to increase
odic separation of the large fraction by centrifugation
relaxation, since they strongly affect the proton relax-
revealed the stability of the final structure and the
ation rate. Data are also available on the clinical appli-
main physical properties of the solutions for 6 months.
cation of manganese compounds as a contrast agent.
No exfoliation and intensive precipitate formation
We studied the mechanisms of the formation of were observed during this period.
nanoparticles of complex iron oxide in detail earlier
[6]. The size of the nanoparticles is determined by the The particle size and the phase composition were
dynamics of the change in the intermediate pH of the determined using a JEOL JEM-1011 electron micro-
solution during the chemical synthesis and by the scope. The images were statistically processed using
selection of the stabilizers. It is important that the the Scan Soft 2000 program. The phase composition
concentration ratio of the initial reagents shift toward of the nanoparticles was determined by comparing the
the region of divalent iron. The dispersion of the sedi- electron diffraction patterns of the samples with a bulk
ment that occurs during the chemical reaction is reference sample for complex oxide Fe3O4. The con-
determined by two processes, that is, nucleation and centration of the active agent of the solutions was
particle growth. Each of the processes is characterized determined on an X-art M X-ray fluorescence ana-
by the optimum dynamics of a change in the solution lyzer (ZAO Komita, Russia). The parameters of the
pH that is controlled by selection of the synthesis experiment were a voltage of 8 kV and a current of
mode. The initial ratio of the salts of the divalent iron 100 μA. The analytic parameter was the area of the
and ligands was determined by calculations according peak of the analyzed element with the background
to the corresponding ion ratio so that the level of area under the peak subtracted. In the case of the pres-
ligand atoms in the nanoparticles did not exceed 50%. ence of two or more atoms in the composition of the
A further increase in the initial concentration of the nanoparticles, the experimental spectrum was the
ligand resulted in significant distortions of the crystal superposition of the peaks of the initial metals. The
structure of the ultrasmall nanoparticles of complex metal concentration was determined according to the
iron oxide and thus led to the destruction of its mag- ratio of the areas under the spectrum for the sample
netic structure. As in the case of obtaining solutions of and a reference sample with a known concentration.
nanoparticles of complex iron oxide, ammonium Magnetic-resonance contrast properties were stud-
hydroxide, NH4OH, was added at an excess concen- ied by proton relaxometry using a Minispec mq NMR
tration with intermediate administration of citric acid relaxometer (Bruker Optic GmBH, Germany). The
into the solution to stabilize the surface and control main frequency of the magnet was 20 MHz (470 mT).
the pH value of the intermediate solution. The The internal air circulation made it possible to main-
amounts of the initial substances that did not change tain the temperature of 40°C. Inversion-recovery
in the chemical reactions, including the stabilizers, in pulse sequences were used to record T1 times, while

BIOPHYSICS Vol. 63 No. 3 2018

 THE DEPENDENCE OF THE MAGNETIC-RESONANCE CONTRAST IMAGING 327

Table 1. The experimental values of the T1 and T2 relaxivities of the nanoparticle solutions and the nanoparticle size
Characteristic size, nm R1, L/(mmol s) R2, L/(mmol s)

MnFe2O4 8±1 3.23 ± 0.32 4.62 ± 0.46

MgFe2O4 9±1 1.01 ± 0.10 1.47 ± 0.15
ZnFe2O4 8±2 1.66 ± 0.17 3.08 ± 0.31
CoFe2O4 14 ± 3 1.47 ± 0.15 8.79 ± 0.88
Fe3O4 6±1 8.48 ± 0.85 17.63 ± 1.76
T = 40°C, pH 7.2.

CPMG (Carr–Purcell–Melboom–Gill) sequences compared to the reference sample, which was probably
were employed to record T2 times. The decay curves caused by the ultrasmall size (to 10 nm) of the
of the corresponding T1 and T2 relaxation times on nanoparticles and distortions of the standard crystal
the concentration of the nanoparticles that were intro- structure. The analysis of the experimental results for
duced into a water-containing medium made it possi- all the nanoparticles with the ligand revealed preserva-
ble to determine the experimental values of the relax- tion of the diffraction maximums that are typical for
ivities. the magnetite structure, which was evidence of
The toxic properties of the nanoparticles were incomplete replacement of atoms of divalent iron. In
determined by MTT assay using a HeLa cell culture at the case of MnFe2O4, the maximum spreading of the
incubation times of 24 and 48 h. The results were pro- diffraction rings was observed, which was probably
cessed using the nonparametric Mann–Whitney test due to additional distortions of the crystal lattice of
with a significance difference between the results for magnetite with the Mn ligand. The MgFe2O4 and
different concentrations at p ≤ 0.05. The viability of ZnFe2O4 nanoparticles were characterized by inter-
the cell culture in the presence of the studied prepara- mediate spreading of the diffraction rings. The sample
tions was determined in contrast to the control. The that contained Co was characterized by the clearest
main parameter for evaluation of the viability was the diffraction pattern that was the closest to the reference
IC50 value, that is, the semimaximum inhibitory con- sample. This result can be explained by the larger size
centration. of the nanoparticles resulting in a higher-order struc-
ture, as well as by the influence of the Co ligand on the
formation of the crystal structure of the nanoparticle
RESULTS AND DISCUSSION (Fig. 1).
The nanoparticles that we obtained were certified The concentration of the main active agent in the
using a complex of methods for determining their solution of the Fe3O4 nanoparticles was 27 g/L. The
phase composition, average particle size, particle-size ratios of the areas for all the nanoparticles with the
distribution, and the concentration of the main active ligand except for cobalt was 1.08–1.19, which was evi-
agent in the colloidal solution. The analysis of micro- dence of the similar contents of the nanoparticles in
photographs that we obtained by electron microscopy the samples. The ratio of the areas of the elements for
made it possible to determine the dispersion composi- the nanoparticles that contained cobalt was 1.56,
tion of the solution. All the solutions contained a main which indicated a smaller particle content compared
fraction of ultrasmall nanoparticles (7–15 nm) and an to the reference sample. It was found that the cytotox-
additional fraction with larger particles (40–50 nm). icity of the solution depended on the particle size, sur-
The percentage of the larger fraction did not exceed face state, and stabilizer concentrations; as the particle
20%; this fraction consisted of aggregates of the ultras- size increased, the nanoparticle toxicity significantly
mall nanoparticles. The characteristic sizes of the decreased [8, 9]. It is believed that the toxicity of
nanoparticles of the small fraction were experimen- nanoobjects is connected with the developed surface
tally determined (Table 1). The samples that con- of the nanoparticles and, as a rule, significantly
tained Co were significantly different, with the char- depends on the characteristic size of the nanoparticles.
acteristic size reaching 14–18 nm. The experimental studies that were carried out for the
The analysis of the phase composition demon- samples demonstrated that the cytotoxicity of the
strated that the ultrasmall Fe3O4 nanoparticles had a ultrasmall nanoparticles did not significantly depend
standard spinel structure that corresponds to the mag- on the incubation time; although there was a tendency
netite structure with the identical position of the dif- toward its decrease at an increase in the incubation
fraction rings. The rings were significantly diffuse duration from 24 to 48 h (Fig. 2). The IC50 values are

BIOPHYSICS Vol. 63 No. 3 2018

328 AKOPDZHANOV et al.

macologically significant dose; therefore, these solu-

tions can be considered to be biocompatible. It is
necessary to note the nanoparticles that contain zinc,
whose toxicity is an order of magnitude higher than for
the other samples, which appears to be connected with
the cytotoxicity of zinc itself and residual products of
the chemical reactions that contain it. The high cyto-
toxicity of the nanoparticles with zinc could be con-
nected with the peculiarities of their formation and
with the state of their surface. One of the main factors
that affect the possible toxicity of the nanoparticles is
the state of the modifying surface layer and the origin
of the modifier. The selection of the modifier should
also consider the biological safety and procedures for
further pharmacological application. The magnetic-
20 µm resonance contrast characteristics of the solutions
(а)
were determined by proton-resonance relaxometry,
with the experimental values of relaxivity given in
Table 1. According to [3], it was shown that the relative
closeness of the relaxivities R1 ≈ R2 for the magnetite
nanoparticles indirectly indicates their superparamag-
netic state. Therefore, it can be concluded that the
ultrasmall nanoparticles with Mn, Mg, and Zn
according to the data of proton relaxometry are in the
superparamagnetic state. The nanoparticles that con-
tain Co should be particularly noted, as their experi-
mental relaxivity R2 exceeds R1 by a factor of five. As
a rule, this indicates the ferrimagnetic structure of the
nanoparticles; however, with insignificant growth of
the characteristic size of the Co nanoparticles to 12–
16 nm with respect to the magnetite nanoparticles (6
20 µm nm), this small size is insufficient for the formation of
(b)
a multidomain structure. The determining role of the
structural factors is indicated by the fact that the
experimental values of the relaxivities R1 and R2 of the
nanoparticles with diamagnetic zinc are almost twice
as high as those for the nanoparticles with magnetic
Mg. The change in the relaxivities of the alloyed
nanoparticles compared to the nonalloyed magnetite
nanoparticles with a similar characteristic size appears
to be because of distortions of the crystal and domain
structures that are induced by the ligand introduction.
For the case of the considered ultrasmall nanoparticles
that form superparamagnetic systems, the effect of the
ligand on the magnetic phase composition is quite
limited. This hypothesis is mostly in agreement with
the data of electron microscopy: at the similar charac-
teristic sizes of the nanoparticles, the diffraction pat-
ters of the samples with the ligands have a pronounced
20 µm spreading of the characteristics rings compared to the
(c)
magnetite sample.
Fig. 1. The diffraction patterns of the samples: (a) Fe3O4,
(b) CoFe2O4, and (c) MnFe2O4. Thus, according to the requirements of modern
clinical magnetic-resonance diagnostics, solutions of
ultrasmall magnetite nanoparticles that are alloyed
given in Table 2. The analysis of the results demon- with manganese and cobalt make it possible to provide
strates that the critical concentrations for all the sam- predominant contrast with respect to one of the
ples are almost 1000 higher than the calculated phar- parameters (T1 or T2).

BIOPHYSICS Vol. 63 No. 3 2018

 THE DEPENDENCE OF THE MAGNETIC-RESONANCE CONTRAST IMAGING 329

(а) (b)

98.43
99.11
98.03

97.04
97.30
95.01

92.19
100 100

89.22

88.54

87.46
86.27

84.48

82.42
80.39
90 90

79.20
75.94
73.09
80 80

64.09
61.44
60.57
60.83
70 70
Cell viability, %

Cell viability, %
54.73

54.00
51.45
60 60

49.75

45.73
50 50
40 40

28.33
22.24
30 30
20 13.73 20

8.68
10 10
0 0
CoFe2O4 MgFe2O4 Fe3O4 CoFe2O4 MgFe2O4 Fe3O4
MnFe2O4 ZnFe2O4 MnFe2O4 ZnFe2O4

Fig. 2. The viability of HeLa cells at 24 h (a) and 48 h (b) incubation with nanoparticles in percent, % of the control values. The
nanoparticle concentration: light columns 10–2 g/L, gray columns 10–3 g/L, and dark columns 10–4 g/L.

Table 2. The experimental values of IC50 for the solutions at different incubation times
IC50, mg/mL; T = 24 h IC50, mg/mL; T = 48 h
MnFe2O4 4.35 ⋅ 10–5 1.57 ⋅ 10–5
MgFe2O4 1.51 ⋅ 10–3 3.09 ⋅ 10–5
ZnFe2O4 1.32 ⋅ 10–6 4.23 ⋅ 10–6
CoFe2O4 1.23 ⋅ 10–3 3.03 ⋅ 10–5
Fe3O4 3.69 ⋅ 10–5 2.52 ⋅ 10–5

ACKNOWLEDGMENTS 5. W. Jiang, Y. S. KimBetty, J. T. Rutka, et al., Nat. Nan-

otechnol. 3, 145 (2008).
This study was supported by the Russian Founda-
tion for Basic Research (project no. 17-03-00539). 6. A. G. Akopdzhanov, N. L. Shimanovskii, V. Yu. Nau-
menko, et al., Khim. Fiz. 33 (7), 1 (2014).
7. N. L. Shimanovskii, V. Yu. Naumenko, A. G. Akop-
REFERENCES dzhanov, et al., Vestn. Ross. Gos. Med. Univ., No. 2,
1. N. L. Shimanovskii, A. G. Akopdzhanov, A. I. Sergeev, 62 (2011).
et al., Eksp. Klin. Farmakol. 73 (6), 23 (2010).
8. T. A. Fedotcheva, N. I. Fedotcheva, V. V. Teplova,
2. J. Mosayebi, M. Kiyasatfar, and S. Laurent, Adv. et al., Vopr. Biol. Med. Farm. Khim. 11, 158 (2011)
Healthc. Mater. 6 (23) (2017). doi 10.1002/adhm.
201700306 9. A. G. Akopdzhanov, N. L. Shimanovskii, T. A. Fed-
3. A. G. Akopdzhanov, A. V. Babich, I. V. Bykov, et al., otcheva, et al., Biophysics (Moscow) 61 (6), 838
Biomed. Radioelektr., No. 9, 28 (2011). (2016).
4. P. Aggarwal, J. B. Hall, C. B. McLeland, et al., Adv.
Drug Delivery Rev. 61, 428 (2009). Translated by E. Berezhnaya

BIOPHYSICS Vol. 63 No. 3 2018