I. INTRODUCTION
In order to obtain information about the effects of microwave (MW) on living cells, different types of exposure systems have been developed during the last two decades. These exposure systems are used for MW medical applications or studies of MW biological effects and mechanisms [Reference De Vequi-Suplicy, Benatti and Lamy1–Reference Kuster and Schönborn6]. The difficulties in bioelectromagnetic studies arise from the control of the physical and biological parameters during exposure. Moreover, as bioelectromagnetic investigations involve different targets and protocols, exposure setups can hardly be standardized. Therefore, the design of each exposure device must be supported by numerical electromagnetic dosimetry and validated with experimental measurements.
In most exposure systems MW irradiation is performed prior to the biophysical analysis. For example, as presented in [Reference Gaber, El Halim and Khalil3], multilamellar liposomes permeability and lipids' conformational changes were studied after 5 h exposure to 900 MHz MW. The liposome samples were placed in a coaxial-type glass tube applicator and exposed to a specific absorption rate (SAR) of 12 W/kg. Among existing applicators, antenna-based systems [Reference Gerbaux, Glowacki, Ammann and Thorne4–Reference Kuster and Schönborn6] have been developed for studying the radio frequency/MW fields' interaction with biological samples or for MW medical applications. All these systems present the same major disadvantage: the biophysical parameters are measured after the irradiation was stopped, which prevents the observation of any transient effect.
The system presented in this paper permits the acquisition of temperature and fluorescence measurements while the biological sample is simultaneously exposed to the MW signal through an open coaxial applicator. In the present experiments, a suspension of giant unilaminar vesicles (GUV) was used as a model of living cells and exposed to MW in a spectrofluorometric cuvette while the GUV generalized polarization (GP) was measured.
For a rigorous analysis of biological experiments, the entire systems must be electrically and thermally characterized. Complementary characterization can also be obtained with numerical analysis. In this study, accurate dosimetry of the proposed MW exposure system was performed. The dosimetry is based on electromagnetic simulations that permit the estimation of SAR levels in well-controlled numerical models.
This paper is organized as follows. Section II provides a description of the MW setup with the open coaxial applicator, the fluorescence measurement technique, and the exposure protocol. Section III details elements about the performed experimental and numerical electromagnetic characterization. The dosimetric results in terms of SAR distribution and temperature measurements are presented in Section IV. An example of GUV membrane's GP measurements is also given in Section IV.
II. EXPOSURE SETUP AND PROTOCOLS
A) Open coaxial applicator system
The device is composed of a spectrofluorometer combined with a 2.45 GHz MW exposure system. The latter component comprises a generator (Sairem, France), a bidirectional coupler (Hewlett Packard 777D, CA, USA), an impedance matcher (Microlab, FXR SF-31N, NJ, USA), and an applicator placed in a holder which contains the biological sample. The generator delivers a 2.45 GHz MW signal with an adjustable output power (up to 120 W). As shown in Fig. 1, the generator is connected to the bidirectional coupler for incident and reflected powers measurements and to the impedance matcher. The latter permits the achievement of optimal impedance matching between the generator and the applicator. Once the impedance matching was performed after measuring the reflected power with a powermeter (Hewlett Packard 436 A, CA, USA) the experiment can be carried out.
Exposure setup: bidirectional coupler (a), impedance matcher (b), and open coaxial applicator (c).
The radiating element is an open coaxial cable ended by a central metallic pin conductor which is in direct contact with the biological suspension (Fig. 2(a)). The inner and outer diameters of the coaxial cable (0.8 and 2.75 mm, respectively) are designed for a 50 Ω impedance. The length of the pin (2 mm) is small compared to the wavelength λ (14.2 mm in the solution) and provides the best impedance matching at 2.45 GHz (Fig. 2). The location of the pin permits to deliver the MW power density in the center of the cuvette near the spectrofluorometer analysis area (light beam). The GUV suspension is placed in a 12 × 12 × 40 mm plastic cuvette (Fig. 2(a)). The thickness of the cuvette walls is 1 mm. The exposed solution volume is 10 × 10 × 23.5 mm which corresponds to 2.35 ml. As illustrated in Fig. 2(b), the spectrofluometric system (SPEX, Germany) is composed of a thermostated holder (TLC50, Quantum Northwest Inc., USA) which sustains the cuvette. Rectangular windows of 4 × 8 mm are found on three sides of the holder in order to allow fluorescence measurements (the fourth window at 180° with respect to the excitation light path is closed with an aluminum obturator). The Peltier elements of the holder are under computer control. A fiber optic temperature probe (Luxtron, Model One, CA, USA) immersed into the medium allows the temperature acquisition during the MW exposure. The tip of the fiber optic is located near to the applicator's pin (Fig. 2(a)). A magnetic stirrer placed at the bottom of the cuvette was also used during the experiments.
Studied exposure applicator. (a) Spectrofluorometric cuvette containing the GUV suspension. The dimensions are: a = 40 mm, b = 28 mm, c = 2 mm, d = 23.5 mm, e = 4 mm, f = 30 mm. (b) Cuvette inside the thermostated holder and the fluorescence system.
B) Membrane GP measurements
Membrane GP is a parameter which accounts for the presence of small, mobile, polarizable molecules (e.g. water) within a biological membrane. It can be measured by fluorescence techniques, using a specific dye, Laurdan (6-dodecanoyl-2-dimethylamino-naphthalene) [Reference Laval, Leveque and Jecko7–Reference Leveque, Reineix and Jecko9], which, in its spectroscopic ground state, is non-polar. The excited state of Laurdan (which has a high dipole moment) may interact with other dipolar molecules in the environment, shifting its emission spectrum toward lower energies (higher wavelengths) because part of the excitation energy is dissipated in the reorientation process of the dipoles. Basically, one has to measure the fluorescence intensity of Laurdan at two specific wavelengths (in our case 437 and 485 nm), called I B and I R, respectively. The numerical value of GP is given by the relation GP = (I B − I R) / (I B + I R). If the membrane is destabilized by an external factor, i.e. MW exposure, water molecules penetrate into the lipid bilayer, and GP values become smaller. In our experiments, GP values are measured in real time during the MW exposure.
C) Chemicals and exposure protocol
GUVs labeled with 10 µM Laurdan (Molecular Probes, Oregon, USA) (excitation wavelength 364 nm) are prepared from DMPC (1,2-dimyristoyl-sn-glycero-3-phosphocholine) (Sigma–Aldrich GmbH, Germany) by the reversed phase evaporation method [Reference Lucio, Vequi-Suplicy, Fernandez and Lamy10].
During MW exposure, the emission spectra of Laurdan are recorded and GP is calculated [Reference Laval, Leveque and Jecko7]. The GUV suspension is thermostated at an initial temperature of 15°C. Then, a MW signal is applied until a temperature of 40°C is reached inside the GUV suspension. The experiments are carried out at 1.2 W incident power. GP values are represented as a function of temperature.
III. ELECTROMAGNETIC CHARACTERIZATION
A) Numerical modeling analysis
In order to characterize the response of the biological sample to the 2.45 GHz MW exposure, numerical analysis is performed using in-house software based on the finite difference-time domain (FDTD) method [Reference O'Connor, Madison, Leveque, Roderick and Bootman11–Reference Schuderer, Spat, Samaras, Oesch and Kuster14]. The calculation requires a grid meshing of the exposure system geometry and takes into account the physical properties of the biological sample such as permittivity and conductivity. The numerically modeled geometry is composed of the electromagnetically relevant parts (Fig. 2(b)) which are, in this case, the open coaxial cable, the cuvette containing the GUV biological suspension, and the internal part of the holder (mostly metallic). The geometry is discretized with a uniform 0.2 × 0.2 × 0.2 mm grid. A 50 Ω localized EM excitation is placed in the open coaxial cable.
The FDTD method gives the EM fields inside the cuvette exposed to MW. In addition, it gives the SAR which is a commonly used unit in bioelectromagnetic studies [Reference Szoka and Papahadjopoulos15–Reference Zhao18]. The SAR quantifies the amount of absorbed power per unit mass (W/kg) in the exposed sample:
where σ (S/m) and ρ (kg/m3) are, respectively, the electric conductivity and the mass density of the exposed biological medium. E (V/m) corresponds to the electric field magnitude. In our study, the considered electromagnetic properties of the biological medium at 2.45 GHz are ɛr = 74, σ = 2.85 S/m, and ρ = 1000 kg/m3, in final suspension, at room temperature. The electromagnetic properties (permittivity and conductivity) of the biological medium are measured with a dielectric probe (85070E Dielectric probe kit, Agilent, USA) connected to a vector network analyzer (VNA 8753E, HP, USA). SAR values are then computed, as illustrated in Section IV, for each elementary cell of the discretized sample volume.
B) Thermal dependencies of electromagnetic properties
SAR values are obtained for electromagnetic properties of the biological medium at ambient temperature. But, as described in Section IIC, the GUV suspension's temperature increases during the experiments. Therefore, in order to estimate the SAR deviation, ɛr and σ values at different temperatures are measured with an open-ended dielectric probe with the same composition as the GUV suspension except for the lipids. During the measurements, the probe solution is heated using a thermostated water bath. Results show that at 2.45 GHz, a temperature increase from 21 to 43°C induces a variation of 7.8% for the permittivity and 6.1% for the conductivity (Fig. 3). Consequently, a comparable deviation, which is within an acceptable range, can also be assigned to the SAR distribution. With parameters at 21 and 43°C, the calculated mean SAR varies less than 5% from the SAR with the selected permittivity and conductivity.
Variation of the biological sample' relative permittivity and conductivity at 2.45 GHz versus temperature.
IV. RESULTS
The mean SAR value calculated in the whole volume is 448 W/kg for 1 W incident power. The SAR distribution computed with the FDTD tool is shown in Fig. 4. The electromagnetic field values in the GUV suspension are higher near the open coaxial applicator (Fig. 4). As the SAR distribution is not homogenous, a 1 × 3 mm magnetic stirrer covered with PTFE (Cole-Parmer, Amex, Austria) is used to homogenize the temperature distribution in the biological medium. During the exposure, the MW incident power delivered to the biological sample is 1.2 W. For this power, the mean SAR value is 538 W/kg for the whole volume. This mean value is evaluated by averaging the SAR values computed in each voxel by FDTD simulation. In Fig. 4, only one vertical plane is represented to illustrate the numerical spatial distribution of the SAR.
SAR distribution at 2.45 GHz for 1 W incident power microwaves.
Figure 5 presents the SAR distribution in the region where the light beams are crossing the GUV suspension. The SAR value presents an expected variation with a maximum placed in the center of the cuvette near the open ended coaxial. However, the mean value along the light beams (534 W/kg for 1.2 W incident power) is similar to the one obtained for the whole volume.
SAR distribution along the light beam through the cuvette at 2.45 GHz for 1.2 W incident power microwaves.
During the exposure protocol, temperature measurements are recorded. As shown in Fig. 6, the temperature reaches 40°C in 600 s. Using the formula CΔT/Δt, where C represents the medium-specific heat (4187 J/kg/K) and ΔT/Δt corresponds to the initial slope of the temperature versus time, an estimation of the mean SAR is equal to 537 W/kg (for 1.2 W incident power). The measured SAR matches well to the numerical result. The temperature-based SAR measurement neglects the heat loss at the boundary of the cuvette in the initial period, which can account for the small differences between the experimental and computed SAR values.
Measured temperature in the GUV suspension heated by 1.2 W incident power microwaves.
As an illustration of the fluorescence measurements performed with this exposure system, Fig. 7 shows the GP evolution measured for an exposure to 1.2 W MW power. The GP curve presents regular values and variation as the temperature rises from 15 to 40° C.
General polarization of 1, 2-dimyristoyl-sn-glycero-3-phosphocholine GUV suspension heated by 1.2 W incident power microwaves.
This system allows the observation of the GP evolution versus temperature. In the current experiments, the temperature increase is induced by MW exposure. For a similar temperature evolution obtained with classical heating, the measured GP variation was slightly different from the one produced by MW exposure. However, due to the electromagnetic field in-homogeneity, it is difficult to state whether the GP variation is induced only by increase in temperature or also by the electromagnetic field. To determine the possible MW effects on the GP, experiments could be made at a constant temperature with a MW modulated signal that does not induce complementary heating.
V. CONCLUSION
In this work, we presented an exposure system for bioelectromagnetic studies involving real-time fluorescence and temperature measurements on biological media during MW exposure. The combination of a spectrofluorimeter with a coaxial exposure device is designed to study effects of 2.45 GHz MW on biological samples. Simultaneous measurements help to understand molecular mechanisms of the effects and dissociate thermal and non-thermal effects, but add up to the complexity of the exposure device as well as the dosimetry. We presented SAR and temperature evaluations of the 1.2 W MW exposure. The SAR is numerically characterized by the Gaussian-like inhomogeneous distribution strictly related to the position of the coaxial applicator. The SAR averaged in the entire sample volume is close to the 534 W/kg mean value along the fluorescence light beam. The experimentally measured temperature is in the range of 15–40°C and quantitatively compatible with the calculated SAR. The GUV GP is in a unilateral relation to the temperature. Both the SAR homogeneity and the temperature control of the exposure system need consideration in improved designs.
ACKNOWLEDGEMENT
This work was performed using HPC resources from GENCI-IDRIS (Grant 2010050646).
Mohamad Kenaan was born in Quenia, Lebanon in 1984. He received the Master degree in circuits, micro-electronic systems and nanotechnology for high frequency and optical communications from the University of Limoges in 2007. He is currently a PhD student at the Xlim Research Institute, Centre National de la Recherche Scientifique (CNRS)-University of Limoges, France. His research interests are focused on the development of exposure systems and characterization of applicators for the study of nanosecond pulsed electric fields (nsPEF) effects on biological cells.
Mihaela G. Moisescu was born in 1975 in Curtea de Arges, Romania. In 2000 she graduated Carol Davila Medical University of Bucharest and in 2002 she received from the same university the M.S. degree in Biophysics and Cell Biotechnology. In 2007 she received the Ph.D. degree from University of Paris-Sud 11 and Medical University of Bucharest with Magna cum Laudae distinction. She is currently Lecturer of Carol Davila Medical University. Her research interests are currently focused on electric cell manipulation (dielectrophoresis) and electromagnetic field interactions with cell membrane and cellular processes like endocytosis and mitosis.
Tudor Savopol was born in Bucharest (Romania) in 1955. He graduated the Faculty of Chemistry at the Polytechnic University of Bucharest in 1980 and the PhD degree at the same university, with the thesis “Photoinduced proton transfer reactions of retinal proteins within biological membranes”. Between 1999 and 2002 he was awarded a Max-Planck post doctoral fellowship at the Max-Planck Institute for Biophysical Chemistry in Dortmund where he was involved in the study of the X-ray crystal structure of the membrane protein Sensory Rhodopsin II. His present status is Associate Professor of Biophysics at the Carol Davila Medical University of Bucharest. His recent main interests are related to the study of molecular mechanisms of the interaction of the living matter with different electromagnetic fields (microwaves, lasers).
D. I. Martin received the primary and Ph.D. degrees in electronics from the Bucharest Polytechnic University of Bucharest, Bucharest, Romania, in 1961 and 1972, respectively. She also received the specialization in particle accelerators from the “Laboratory National de Frascati”, Frascati, Italy. She is currently a Senior Researcher First Degree with the Accelerator Laboratory, National Institute for Lasers Plasma and Radiation Physics, Bucharest, Romania. Her main research interests are in physics and applications of electron accelerators, microwave theory and techniques, and electron beam and microwave technologies. Her research is developed in the framework of four projects within the National Program for Research, Development and Innovation. She is also currently a member of the Ph. D. Committees at the Polytechnic University of Bucharest and Bucharest University. She is authored or coauthored ~80 scientific journal papers and ~100 scientific paper published in international proceedings and is the holder of eight Romanian Patents. Dr. D. Martin was co-recipient of the Bronze Metal at the 48th World Exhibition and Innovation, Research and New Technology held in Brussels, Belgium, in 1999, for the invention “Polymerization and Copolymerization Procedure in the Radiation Fields”. She also received the “Horia Hulubei” Romanian Academy Prize in Physics for the work “Macroscopic Effects of Electron Beam and Microwave Interaction with Matter” in 2002.
Delia Arnaud-Cormos was born on May 19, 1978, in Cugir, Romania. In 2002, she obtained the Diplôme d'Ingénieur degree from the Institute of Computer Science and Communication (IFSIC), Rennes, France. She received her Master and Ph.D. degrees from the National Institute of Applied Sciences (INSA), Rennes, France, in 2003 and 2006, respectively. From 2006 to 2007, she was an Assistant Professor at the University Paris-Sud 11, Orsay, France. Since 2007, she joined the Bioelectrophotonics research group of the Xlim Institute (University of Limoges/CNRS) in Limoges, France as an Associate Professor. She is a member of the EBEA, BEMS, IEEE. Her current research interests concern nanosecond pulses/microwave exposure systems set-up and dosimetric characterization for bioelectromagnetic studies.
Philippe Leveque was born in Poitiers, France, in 1964. He received the Ph.D. degree from the University of Limoges, France, in 1994. In 1995, he joined the C.N.R.S., as associate scientist. His main area of interest concerns scattering problems of electromagnetic waves, particularly in the time domain. He is involved in the development of dosimetry and exposure setups for health-risk assessment in cooperation with biological and medical research groups. He is currently the group leader of Bioelectrophotonics team working in nanopulse application, in the XLIM Research Institute held by the University of Limoges and CNRS. He is a member of the EBEA, BEMS, IEEE, and he was the French official member of the URSI scientific commission K (Electromagnetics in Biology & Medecine).
