- Email: [email protected]
In vivo electron paramagnetic resonance spectroscopy-imaging in experimental oncology: The hope and the reality
Int. J. Radiation Oncology Biol. Phys.. Vol. 29. No. 3. pp. 421-425, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in the USA. All rights reserved 0360-3016/94 $6.00 + .OO
Pergamon
0360-3016(93)EOOlO-4
??Special Lecture
IN VZVO ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPY-IMAGING IN EXPERIMENTAL ONCOLOGY: THE HOPE AND THE REALITY MARCO FERRARI, M.D., VALENTINA QUARESIMA, CINZIA L. URSINI, MARCELLO ALECCI AND ANTONELLO SOTGIU, PH.D. Department
of Biomedical Sciences and Technology, University L’Aquila, 67 100 L’Aquila, Italy
Purpose: Low frequency (280 MHz) electron paramagnetic resonance imaging is a new magnetic resonance technique, -being developed, that can map the in riro spatial distribution of paramagnetic species such as nitroxide free radicals. The reduction rate of these molecules is affected by oxygen concentration. This paper gives some examples of the use of electron paramagnetic resonance imaging methodology in whole rats in the framework of its possible use in experimental oncology. Methods and Materials: The 280 MHz apparatus based on a cylindrical 16 pole magnet was developed and designed specifically for 50-200 g laboratory animals. It generates the main field and the three field gradients required for three-dimensional (3-D) projections. A pyrrolidine nitroxyl (2,2,5,5,-tetramethylpyrrolidine-l-oxyl-3-carboxylic acid) was injected intravenously in rats to provide an electron paramagnetic resonance signal for in rive measurements. Electron paramagnetic resonance X-band spectrometer was used to monitor pyrrolidine nitroxyl decay in an external blood circuit during normoxia and moderate hypoxia (15% Or). Results and Conclusion: One-dimensional (1-D) transversal and longitudinal mapping of this nitroxide free radical distribution in rat whole body was obtained 7-9 min after injection. In circulating blood, nitroxide half-life decreased significantly during hypoxia. The present sensitivity (1O-4-1O-5 M), spatial resolution (3-10 mm) and collection time (3-5 min) could be drastically improved by narrow linewidth paramagnetic probes and pulsed techniques. Electron paramagnetic resonance, Free radicals, Oxygenation,
INTRODUCTION
Free radicals, and oxygen-derived free radicals in particular, have very short half-lives (i.e., OH’, 10e9 s; ROO
RO’, 1O-6-1O-8 s). Even in physiological circumstances the free radical production is consistent; that is, the endothelial production of nitric oxide is estimated at 3-4 pmol/kg of wet tissue (20). Due to the limitations of the EPR instrumentation currently available, their direct detection is impossible. Spin trapping methods have been used in vitro and in vivo (23). This technique involves the addition to the biological system under study of a diamagnetic compound (nitroso or nitrone compoundsspin trap). It reacts with the free radicals to form relatively stable radical adducts (more stable than the primary free radical) that are easily detectable by EPR. It is expected that it will be possible to use spin trapping methods on living animals once the sensitivity of the EPRI instrumentation is improved and more appropriate nontoxic spin traps are available. The EPR spectroscopy research presented here was performed using exogenous free radicals as the source of an EPR signal. The most commonly used probes are ni-
Magnetic resonance imaging ( 17) magnetic resonance spectroscopy (24), and positron emission tomography ( 19) are different techniques holding great promise for the evaluation of tumor oxygenation-metabolism and cancer therapy. In the last few years our laboratory and a few others have been involved in the development of a new medical imaging technique: electron paramagnetic resonance (EPR) imaging (EPRI) (6,8,11,15, 18,25,27,28). These advances have been recently reviewed (9, 12). Electron paramagnetic resonance imaging can determine the in vivo spatial distribution of free radicals. A free radical is any species capable of independent existence that contains one or more unpaired electrons. Free radicals are produced by different physiopathological processes such as inflammation, aging, reperfusion, and carcinogenesis. Substantial evidence suggests that both oxygen and organic-free radical intermediates in biomolecular interactions contribute to the initiation, promotion, and/ or progression stages of chemical carcinogenesis (29).
Reprint requests to: Marco Ferrari, M.D. Acknowledgements-This research has been supported by MURST
40%, INFM,
Nitroxides.
Accepted for publication in part
INFN and PF CNR ACRO. 421
13 October
1993.
422
1. J.
Radiation
Oncology 0
Biology 0 Physics
Volume 29, Number 3, 1994
troxide free radicals, which combine low toxicity with the possibility to determine oxygen concentration. This can be done either through the determination of their linewidth and/or through the study of their reduction rate. Nitroxides are reduced to their corresponding diamagnetic hydroxylamines by enzymatic and nonenzymatic mechanisms (9). Nitroxides are introduced intravenously or intraperitoneally into the biological system and mapped by magnetic field gradients and reconstruction techniques. EPRI can be performed in a wide range of operating frequencies: X-band (9- 10 GHz) for sample volumes of a few ~1, L-band (l-2 GHz) for samples of a few ml, and very low frequencies (200-300 MHz) for larger samples up to 200 ml. The low frequency instrumentation and lumped parameter resonators (8, 13, 15, 26) make it possible to collect in vivo EPR spectra of nitroxides from localized regions. Different EPRI spectrometers have recently been used to obtain two-dimensional (2-D) images of mouse lung at 1.2 GHz (28), and 3-D images of a rat head at 700 MHz (18). Electron paramagnetic resonance spectra of a pyrrolidine nitroxide spin probe, 2,2,5,5,-tetramethylpyn-olidine1-oxyl-3-carboxylic acid (PCA) were obtained in the rat whole body by our group using a low frequency EPR spectrometer operating at 280 MHz. The uptake, distribution and reduction of PCA were investigated and 2-D transversal projections were obtained on rats (25), as well as on phantoms (2). The purpose of this paper is to give some examples of the use of EPRI methodology in 50-70 g rats. Spectra with transversal and longitudinal field gradients and the corresponding 1-D deconvolutions necessary for the 2-D projection reconstruction, are presented to demonstrate the possibility of applying EPRI in experimental animals. METHODS
on a cylindrical 16-pole magnet, was developed and designed specifically for 50-200 g laboratory animals. It generates the main field up to 0.02 T and two of the three field gradients (100 mT/m) required for 3-D reconstruction. The third gradient (28 mT/m) is provided by air coils inside the magnet. Field and gradient direction can be rotated under computer control, making image reconstruction very similar to that used in computer assisted tomography. Wistar rats (50- 180 g) were anesthetized with urethane (lg/kg, intraperitoneally). The jugular veins and carotid arteries were isolated and cannulated, respectively, for PCA injection and for blood pressure recording. Temperature was maintained at 38°C by the use of a water jacket. Unrestrained rats were placed in the resonator in a supine position (Fig. 1). The nitroxide free radical PCA was chosen for its relatively low toxicity (LDsO = 15 mmol/ kg). In in vivo experiments, only the peak-to-peak amplitude of the low field component of the PCA triplet was used for quantitative measurement of the signal. The lowest concentration at which it was possible to obtain 2-D projections was about lop5 M. The signal decay over time was monoexponential and the measured half-life was dose dependent. For PCA monitoring in blood an external circuit (dead volume 160 ~1) was inserted between the carotid arteries of artificially ventilated rats. The PCA signal intensity as a function of time was monitored in the circuit by an Xband EPR spectrometer equipped with an ER 4 108 TMH cavity. Spectrometer conditions were: frequency, 9.72 GHz; field modulation frequency, 100 KHz; field modulation amplitude, 0.08 mT; power, 30 mW; scan range, 0.8 mT; scan time, 20 s; time constant, 0.1 s. Two identical doses of PCA were given and the first dose was administered during normoxia. After 20 min, the animal was submitted to moderate hypoxia ( 15% 02) and the second PCA injection was given. During hypoxia the mean blood pressure was maintained by continuous adrenaline infusion. Spectra were recorded consecutively every 30 s for 20 min. The intensity of the signal was estimated from the peak-to-peak height. The same tubing used to monitor circulating blood was used to evaluate, at about 37”C,
AND MATERIALS
The 280 MHz apparatus consists of three main sections: the radio frequency bridge, the magnet and gradient system, and the receiving-control section that drives the experiment. The technical details of the apparatus have been reported in recent papers (I, 3, 4). This apparatus, based
carotid for
artery
catheter
loop
BP recording
-
resonator
hap \
temperature Jugular veln cat-. for drugs
23
thermostatic
system
Fig. 1. Experimental setup for rat whole body measurements. .The loop gap resonator is positioned in the core of the multipolar magnet. BP = blood pressure.
423
In vivo electron paramagnetic resonance 0 M. FERRARI er al.
the PCA EPR signal time course of arterial blood collected 4 min after injection. Nitroxide kinetics showed a biexponential decay with a slow and a fast component.
O-O
8 min
A--.
6 min
RESULTS Figure 2 shows EPR spectra from rat whole body at different times after PCA injection and the corresponding 1-D spin density obtained by deconvolutions. Spectra without gradients allowed us to calculate the PCA halflife (30 min) and to obtain the deconvolutions relating to each spectrum recorded in the presence of the field gradient. The best picture of PCA distribution was obtained by axial gradient projection showing two regions. The use of a stable free radical positioned on the rat’s abdomen as spatial reference made it possible to identify one of the regions observed as the liver. The duration of the PCA half-life allowed us to collect 4-8 spectra in the presence of a field gradient necessary to reconstruct each 2-D transversal or longitudinal projection, which was recorded in 2.5-5 min. The nitroxyl PCA decay, which is due mainly to its reduction to diamagnetic hydroxylamine by the organs responsible for nitroxide metabolism (liver and kidneys), was monoexponential (25). The effect of oxygenation on the PCA reduction rate in vivo was investigated by Xband EPR monitoring of circulating rat blood. Figure 3 shows the effect of the fraction of inspired oxygen on the reduction of PCA. Since the PCA water soluble nitroxide equilibrates quickly with the intracellular compartment, an initial fast phase of PCA distribution was followed by a slow phase due to PCA metabolism and clearance. The decay half-life of the metabolism-elimination phase decreased significantly during hypoxia. No increase was found in an arterial blood sample collected 4 min after the administration of PCA to the rat and no substantial DECONVOLUTIONS
m 2. _--____
9 min __-*
DISTANCE
( a.“. )
Fig. 2. (Left panel) EPR spectra obtained from 50 g rat whole body after 4 mmol/kg injection of PCA in the following conditions: without gradient (continuous line), with radial gradient (12 mT/m) (dotted line) and with axial gradient (12 mT/m) (dashed line). Acquisition parameters were: center field, 8.35 mT; scan range, 1.7 mT; scan time, 25 s; time constant, 300 ms; power level, 110 mW; modulation frequency, 8 kHz; width amplitude, 0.04 mT. (Right panel) The 1-D spin distribution of PCA in rat whole body along the radial (dotted line) and axial (dashed line) directions. The spectra in the presence of a gradient (left panel) were deconvoluted with the integral zero gradient line (continuous line) to obtain the 1-D spin density.
min
Fig. 3. Typical semilogarithmic plot of blood PCA signal vs. time after rapid intravenous administration (0.044 mmol/kg) in normoxia (0) followed by hypoxia (A). The corresponding halflives of the slow phase are reported. Signal amplitude was the peak-to-peak height of the central component of the PCA triplet spectrum.
changes in half-life occurred after two repeated doses of PCA in normoxic conditions. The different PCA reduction rate in hypoxia suggests that PCA could be used in vivo to generate images of differently oxygenated organs. DISCUSSION The role of free radicals and reactive oxygen species in many pathways and multiple steps leading to human cancer is as yet unresolved ( 14). Free radical damage has long been believed to be a risk factor for many processes that accompany cancer in a variety of animal species. Much research has been directed toward establishing correlations between oxidative damage, antioxidant defense systems, and cancer. Although a few modest correlations have been observed, attempts to use antioxidant manipulation (superoxide dismutase, catalase, glutathione peroxidase, vitamin E, vitamin C) in animal and cell culture models to modify the therapeutic response of malignant tumors have yielded few definitive results. Data continue to accumulate about the ubiquity of free radicals and their significant destructive capacity in living tissues. In addition, several anti-cancer drugs such as adriamycin, bleomycin, mitomycin-c, and vincristine are known to perform their tumoricidal action by a free radical-dependent mechanism and to increase free radical generation and lipid peroxidation in vitro and in vivo. It is probable that a better understanding of free radical processes will lead to the discovery of measures (dietary, pharmacology) to improve cancer care and increase life expectancy. The role of EPRI in oncology is currently under investigation in many centers throughout the world. Table 1 reports some of the most important results obtained by EPR spectroscopy and EPRI on experimental tumor models. Non-invasive in vitro and in viva technique at Lband made it possible to estimate quantitatively the effects on B-l 6 melanoma cells and growing tumors of radiosensitizing agents used separately or combined with others (22). When the biological effects of hyperthermia on B16 melanomas growing on the tail of mice were investi-
424
1. J. Radiation Oncology 0 Biology 0 Physics
Table 1. Examples of experimental tumor models studied by EPR at different frequencies Frequency 9-10 GHz (X-band)
Samples
Spin probes
Authors
Spheroids
15N-PDT
Dobrucki
Spheroids
15N 16-DS
Swartz
2-4 GHz (S-band)
Murine melanoma
TEMPOL
Lukiewicz
l-2 GHz (L-band)
Murine melanoma
CTPO
Berliner
Melanoma Murine adenocarcinoma
CTPO Fusinite
Lukiewicz Bacic
Murine fibrosarcoma
mHCTP0
Halpern
0.25 GHz
“N-PDT: 2,2,6,6,-tetramethylpiperidine-d,~-’5N-oxyl-4-one; 15N 16-DS: 15N-16-doxyl stearic acid; TEMPOL: 2,2,6,6,-tetramethylpiperidine-N-oxyl-4-01; CTPO: 3-carbamoyl-2,2,5,5-tetramethyl-3-pyrroline-I-yloxyl; mHCTP0: 4-protio-3-carbamoyl-2,2,5,5-tetraperdeuteromethyl-3-pyrrolinylI-oxy.
gated, the half-lives of TEMPOL were found to increase with temperature (21). The first EPR image of an in situ
growing tumor in a living mouse was obtained in 1987 (6). The cross-sectional image was obtained perpendicular to the tail axis using an L-band EPR spectrometer. The reconstructed image showed different concentrations of a nitroxide probe in the tumor tissue. EPRI microscopy (9 GHz) was used to study the time course of the loss of viability of spheroids at different stages of growth. To obtain 2-D images perdeuterated Tempone was used as the source of the EPR signal. For each image 64 projections were collected with a magnetic field gradient of 90 G/cm (10). Measurements of oxygen concentrations in the extracellular body water of fibrosarcoma in living mice were studied using a perdeuterated CTPO as spin probe and
Volume 29, Number 3, 1994
an L-band apparatus ( 16). This preliminary study demonstrated that in vivo the EPR technique can permit discrimination between different radiobiologic hypoxic fractions. Electron paramagnetic resonance determination of partial pressure of oxygen was recently done using fusinite, a carbon-based, inert and sensitive intratumoral probe. Changes in partial pressure of oxygen vs. localization of the probe and postirradiation were analyzed on murine mammary adenocarcinoma (5). Each new biomedical instrumentation takes cancer as its first and most formidable challenge. Electron paramagnetic resonance imaging has certain advantages over other imaging techniques. Firstly, the technique might be able to detect free radical generation in situ; secondly, EPRI does not use radiation, which is important for a cancer patient. The use of EPRI in oncology is still in its infancy and its development has been much slower than other imaging techniques currently available, but its role could be unique. Electron paramagnetic resonance imaging is still far from being a satisfactory imaging tool on account of the sensitivity limitations and the too large linewidth of the paramagnetic probes currently used. Research is needed to establish the critical relationships between free radical sources, protective systems and cancer phenomena that may be tractable to intervention. The value of EPRI in experimental oncology is beginning to be recognized and there is little doubt that further advances in technology and the development of narrow linewidth specific spin probes will allow EPRI to detect trapped free radicals. The ongoing development of pulsed EPRI techniques in our and other laboratories (7) should reduce the acquisition time of the EPR spectra reported in Figure 2 to the ms range. The fact that a different PCA reduction rate was found in hypoxia (Fig. 3) suggests that fast EPRI and appropriate spin probes might yield 3-D images with sufficient resolution to discriminate spatial heterogeneity of tumor oxygenation.
REFERENCES 1. Alecci, M.; Colacicchi, S.; Sotgiu, A.; Testa, L.; Varoli, V.
2.
3.
4.
5.
6.
Automatic trimming technique for multipolar magnets. J. Appl. Phys. 71:3053-3055; 1992. Alecci, M.; Della Penna, S.; Sotgiu, A; Testa, L. R. F. (280 MHz) EPR imaging of extended samples: Apparatus and preliminary results. Appl. Magn. Reson. 3:909-9 15; 1992. Alecci, M.; Della Penna, S.; Sotgiu, A.; Testa, L.; Vannucci, I. Electron paramagnetic resonance spectrometer for threedimensional in vivo imaging at very low frequency. Rev. Sci. Instrum. 63:4263-4270; 1992. Alecci, M.; Gualtieri, G.; Sotgiu, A.; Testa, L.; Varoli, V. Multipolar magnet for low-frequency ESR imaging. Meas. Sci. Technol. 2:32-37; 1991. Bacic, G.; Liu, K. G.; O’Hara, J. A.; Goda, F.; Harris, R. D.; Szybinski, K.; Swartz, H. M. An EPR and MRI study of oxygen tension in a murine tumor (Abstr.). Society Magnetic Resonance Medicine, Twelfth Annual Scientific Meeting 1: 105; 1993. Berliner, L. J.; Fujii, H.; Wan, X.; Lukiewicz, S. J. Feasibility
study of imaging a living murine tumor by electron paramagnetic resonance. Magn. Reson. Med. 4:380-384; 1987. Bourg, J.; Krishna, M. C.; Mitchell, J. B.; Tschudin, R. G.; Pohida, T. J.; Friauf, W. S.; Smith, P. D.; Metcalfe, J.; Harrington, F.; Subramanian, S. Radiofrequency Ff EPR spectroscopy and imaging. J. Magn. Reson. B 102: 112-l 15; 1993. Brivati, J. A.; Stevens, A. D.; Symons, M. C. R. A radiofrequency ESR spectrometer for in vivo imaging. J. Magn. Reson. 92:480-489; 199 1. 9. Colacicchi, S.; Ferrari, M.; Sotgiu, A. In vivo electron paramagnetic resonance spectroscopy/imaging: First experiences, problems, and perspectives. Int. J. Biochem. 24:205-214; 1992. 10. Dobrucki, J. W.; Demsar, F.; Walczak, T.; Woods, R. K.; Bacic, G.; Swartz, H. M. Electron spin resonance microscopy of an in vitrotumor model. Br. J. Cancer 6 1:22 l-224; 1990. 11. Dobrucki, J. W.; Sutherland, R. M.; Swartz, H. M. Nonperturbing test for cytotoxicity in isolated cells and spheroids,
In vivo electron paramagnetic resonance 0 M. FERRARIetal.
12. 13.
14. 15.
16.
17. 18.
19.
20.
using electron paramagnetic resonance. Magn. Reson. Med. 19:42-55; 1991. Eaton, G. R.; Eaton, S. S.; Ohno, K. EPR imaging and in vivo EPR. Boca Raton: CRC Press; 199 1. Froncisz, W.; Hyde, J. S. The loop-gap resonator: A new microwave lumped circuit ESR sample structure. J. Magn. Reson. 47:515-521; 1982. Goldstein, B. D.; Witz, G. Free radicals and carcinogenesis. Free Rad. Res. Comms. 11:3-10; 1990. Halpern, H. J.; Spencer, D. P.; van Polen, J.; Bowman, M. K.; Nelson, A. C.; Dowey, E. M.; Teicher, B. A. Imaging radio frequency electron-spin-resonance spectrometer with high resolution and sensitivity for in vivo measurements. Rev. Sci. Instrum. 60:1040-1050; 1989. Halpern, H. J.; Yu, C.; Peric, M.; Barth, E. Measurement of molecular oxygen concentrations in the extracellular body water of tumors of live mice with in vivo spin label EPR oxymetry (Abstr. 102). 34th Rocky Mountain Conference on Analytical Chemistry, 1992. Husband, J.; Guy, R. Magnetic resonance imaging in oncology. Gut 33:1587-1589; 1992. Ishida, S.; Matsumoto, S.; Yokoyama, H.; Mori, N.; Kumashiro, H.; Tsuchihashi, N.; Ogata, T.; Yamada, M.; Ono, M.; Kitajima, T.; Kamada H.; Yoshida, E. An ESR-CT imaging of the head of a living rat receiving an administration of a nitroxide radical. Magn. Reson. Imag. 10: 109- 114; 1992. Jones, T.; Tilsley, D. W. 0.; Wilson, C. B. J. H.; Lammertsma, A. A.; Brown, G.; Brady, F.; Price, P. M. Positron emission tomography for tumor assessment. NMR Biomed. 5:265-269; 1992. Kubrina, L. N.; Caldwell, W. S.; Mordvintcev, P. I.; Malenkova, I. V.; Vanin, A. F. EPR evidence for nitric oxide
21.
22.
23. 24, 25.
26. 27.
28.
29.
425
production from guanidino nitrogens of L-arginine in animal tissues in vivo. Biochem. Biophys. Acta 1099:233-237; 1992. Lukiewicz, S. J.; Cieszka, K.; Wojcik, K.; Lackowska, B.; Markowska, E.; Pajak, S.; Elas, M.; Dubis, K. In vivo ESR studies on the influence of hyperthermia on biological halflives (BHL) of nitroxides in B- I6 murine melanoma. Phys. Med. 5:3 15-320; 1989. Lukiewicz, S. J.; Elas, M.; Cieszka, K. In vitro and in vivo ESR studies on multimodality radiosensitization of B-16 melanoma cells and tumors (Abstr.). Society Magnetic Resonance Medicine, Tenth Annual Scientific Meeting 1~474; 1991. McCay, P. B. Application of ESR spectroscopy in toxicology. Arch. Toxicol. 60:133-137; 1987. Negendank, W. Studies of human tumors by MRS: A review. NMR Biomed. 5:303-324; 1992. Quaresima, V.; Alecci, M.; Ferrari, M.; Sotgiu, A. Whole rat paramagnetic resonance imaging of a nitroxide free radical by a radiofrequency (280 MHz) spectrometer. Biochem. Biophys. Res. Comm. 183:829-835; 1992. Sotgiu, A. Resonator design for in vivo ESR spectroscopy. J. Magn. Reson. 65:206-214; 1985. Swartz, H. M.; Glockner, J. F. Measurement of oxygen by EPRI and EPRS. In: Eaton, G. R., Eaton, S. S., Ohno, K., eds. EPR imaging and in vivo EPR. Boca Raton: CRC Press; 199 1:260-289. Takeshita, K.; Utsumi, H.; Hamada, A. ESR measurement of radical clearance in lung of whole mouse. Biochem. Biophys. Res. Comm. 177:874-880; 1991. Trush, M. A.; Kensler, T. W. An overview of the relationship between oxidative stress and chemical carcinogenesis. Free Rad. Biol. Med. 10:201-209; 1991.
Pergamon
0360-3016(93)EOOlO-4
??Special Lecture
IN VZVO ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPY-IMAGING IN EXPERIMENTAL ONCOLOGY: THE HOPE AND THE REALITY MARCO FERRARI, M.D., VALENTINA QUARESIMA, CINZIA L. URSINI, MARCELLO ALECCI AND ANTONELLO SOTGIU, PH.D. Department
of Biomedical Sciences and Technology, University L’Aquila, 67 100 L’Aquila, Italy
Purpose: Low frequency (280 MHz) electron paramagnetic resonance imaging is a new magnetic resonance technique, -being developed, that can map the in riro spatial distribution of paramagnetic species such as nitroxide free radicals. The reduction rate of these molecules is affected by oxygen concentration. This paper gives some examples of the use of electron paramagnetic resonance imaging methodology in whole rats in the framework of its possible use in experimental oncology. Methods and Materials: The 280 MHz apparatus based on a cylindrical 16 pole magnet was developed and designed specifically for 50-200 g laboratory animals. It generates the main field and the three field gradients required for three-dimensional (3-D) projections. A pyrrolidine nitroxyl (2,2,5,5,-tetramethylpyrrolidine-l-oxyl-3-carboxylic acid) was injected intravenously in rats to provide an electron paramagnetic resonance signal for in rive measurements. Electron paramagnetic resonance X-band spectrometer was used to monitor pyrrolidine nitroxyl decay in an external blood circuit during normoxia and moderate hypoxia (15% Or). Results and Conclusion: One-dimensional (1-D) transversal and longitudinal mapping of this nitroxide free radical distribution in rat whole body was obtained 7-9 min after injection. In circulating blood, nitroxide half-life decreased significantly during hypoxia. The present sensitivity (1O-4-1O-5 M), spatial resolution (3-10 mm) and collection time (3-5 min) could be drastically improved by narrow linewidth paramagnetic probes and pulsed techniques. Electron paramagnetic resonance, Free radicals, Oxygenation,
INTRODUCTION
Free radicals, and oxygen-derived free radicals in particular, have very short half-lives (i.e., OH’, 10e9 s; ROO
RO’, 1O-6-1O-8 s). Even in physiological circumstances the free radical production is consistent; that is, the endothelial production of nitric oxide is estimated at 3-4 pmol/kg of wet tissue (20). Due to the limitations of the EPR instrumentation currently available, their direct detection is impossible. Spin trapping methods have been used in vitro and in vivo (23). This technique involves the addition to the biological system under study of a diamagnetic compound (nitroso or nitrone compoundsspin trap). It reacts with the free radicals to form relatively stable radical adducts (more stable than the primary free radical) that are easily detectable by EPR. It is expected that it will be possible to use spin trapping methods on living animals once the sensitivity of the EPRI instrumentation is improved and more appropriate nontoxic spin traps are available. The EPR spectroscopy research presented here was performed using exogenous free radicals as the source of an EPR signal. The most commonly used probes are ni-
Magnetic resonance imaging ( 17) magnetic resonance spectroscopy (24), and positron emission tomography ( 19) are different techniques holding great promise for the evaluation of tumor oxygenation-metabolism and cancer therapy. In the last few years our laboratory and a few others have been involved in the development of a new medical imaging technique: electron paramagnetic resonance (EPR) imaging (EPRI) (6,8,11,15, 18,25,27,28). These advances have been recently reviewed (9, 12). Electron paramagnetic resonance imaging can determine the in vivo spatial distribution of free radicals. A free radical is any species capable of independent existence that contains one or more unpaired electrons. Free radicals are produced by different physiopathological processes such as inflammation, aging, reperfusion, and carcinogenesis. Substantial evidence suggests that both oxygen and organic-free radical intermediates in biomolecular interactions contribute to the initiation, promotion, and/ or progression stages of chemical carcinogenesis (29).
Reprint requests to: Marco Ferrari, M.D. Acknowledgements-This research has been supported by MURST
40%, INFM,
Nitroxides.
Accepted for publication in part
INFN and PF CNR ACRO. 421
13 October
1993.
422
1. J.
Radiation
Oncology 0
Biology 0 Physics
Volume 29, Number 3, 1994
troxide free radicals, which combine low toxicity with the possibility to determine oxygen concentration. This can be done either through the determination of their linewidth and/or through the study of their reduction rate. Nitroxides are reduced to their corresponding diamagnetic hydroxylamines by enzymatic and nonenzymatic mechanisms (9). Nitroxides are introduced intravenously or intraperitoneally into the biological system and mapped by magnetic field gradients and reconstruction techniques. EPRI can be performed in a wide range of operating frequencies: X-band (9- 10 GHz) for sample volumes of a few ~1, L-band (l-2 GHz) for samples of a few ml, and very low frequencies (200-300 MHz) for larger samples up to 200 ml. The low frequency instrumentation and lumped parameter resonators (8, 13, 15, 26) make it possible to collect in vivo EPR spectra of nitroxides from localized regions. Different EPRI spectrometers have recently been used to obtain two-dimensional (2-D) images of mouse lung at 1.2 GHz (28), and 3-D images of a rat head at 700 MHz (18). Electron paramagnetic resonance spectra of a pyrrolidine nitroxide spin probe, 2,2,5,5,-tetramethylpyn-olidine1-oxyl-3-carboxylic acid (PCA) were obtained in the rat whole body by our group using a low frequency EPR spectrometer operating at 280 MHz. The uptake, distribution and reduction of PCA were investigated and 2-D transversal projections were obtained on rats (25), as well as on phantoms (2). The purpose of this paper is to give some examples of the use of EPRI methodology in 50-70 g rats. Spectra with transversal and longitudinal field gradients and the corresponding 1-D deconvolutions necessary for the 2-D projection reconstruction, are presented to demonstrate the possibility of applying EPRI in experimental animals. METHODS
on a cylindrical 16-pole magnet, was developed and designed specifically for 50-200 g laboratory animals. It generates the main field up to 0.02 T and two of the three field gradients (100 mT/m) required for 3-D reconstruction. The third gradient (28 mT/m) is provided by air coils inside the magnet. Field and gradient direction can be rotated under computer control, making image reconstruction very similar to that used in computer assisted tomography. Wistar rats (50- 180 g) were anesthetized with urethane (lg/kg, intraperitoneally). The jugular veins and carotid arteries were isolated and cannulated, respectively, for PCA injection and for blood pressure recording. Temperature was maintained at 38°C by the use of a water jacket. Unrestrained rats were placed in the resonator in a supine position (Fig. 1). The nitroxide free radical PCA was chosen for its relatively low toxicity (LDsO = 15 mmol/ kg). In in vivo experiments, only the peak-to-peak amplitude of the low field component of the PCA triplet was used for quantitative measurement of the signal. The lowest concentration at which it was possible to obtain 2-D projections was about lop5 M. The signal decay over time was monoexponential and the measured half-life was dose dependent. For PCA monitoring in blood an external circuit (dead volume 160 ~1) was inserted between the carotid arteries of artificially ventilated rats. The PCA signal intensity as a function of time was monitored in the circuit by an Xband EPR spectrometer equipped with an ER 4 108 TMH cavity. Spectrometer conditions were: frequency, 9.72 GHz; field modulation frequency, 100 KHz; field modulation amplitude, 0.08 mT; power, 30 mW; scan range, 0.8 mT; scan time, 20 s; time constant, 0.1 s. Two identical doses of PCA were given and the first dose was administered during normoxia. After 20 min, the animal was submitted to moderate hypoxia ( 15% 02) and the second PCA injection was given. During hypoxia the mean blood pressure was maintained by continuous adrenaline infusion. Spectra were recorded consecutively every 30 s for 20 min. The intensity of the signal was estimated from the peak-to-peak height. The same tubing used to monitor circulating blood was used to evaluate, at about 37”C,
AND MATERIALS
The 280 MHz apparatus consists of three main sections: the radio frequency bridge, the magnet and gradient system, and the receiving-control section that drives the experiment. The technical details of the apparatus have been reported in recent papers (I, 3, 4). This apparatus, based
carotid for
artery
catheter
loop
BP recording
-
resonator
hap \
temperature Jugular veln cat-. for drugs
23
thermostatic
system
Fig. 1. Experimental setup for rat whole body measurements. .The loop gap resonator is positioned in the core of the multipolar magnet. BP = blood pressure.
423
In vivo electron paramagnetic resonance 0 M. FERRARI er al.
the PCA EPR signal time course of arterial blood collected 4 min after injection. Nitroxide kinetics showed a biexponential decay with a slow and a fast component.
O-O
8 min
A--.
6 min
RESULTS Figure 2 shows EPR spectra from rat whole body at different times after PCA injection and the corresponding 1-D spin density obtained by deconvolutions. Spectra without gradients allowed us to calculate the PCA halflife (30 min) and to obtain the deconvolutions relating to each spectrum recorded in the presence of the field gradient. The best picture of PCA distribution was obtained by axial gradient projection showing two regions. The use of a stable free radical positioned on the rat’s abdomen as spatial reference made it possible to identify one of the regions observed as the liver. The duration of the PCA half-life allowed us to collect 4-8 spectra in the presence of a field gradient necessary to reconstruct each 2-D transversal or longitudinal projection, which was recorded in 2.5-5 min. The nitroxyl PCA decay, which is due mainly to its reduction to diamagnetic hydroxylamine by the organs responsible for nitroxide metabolism (liver and kidneys), was monoexponential (25). The effect of oxygenation on the PCA reduction rate in vivo was investigated by Xband EPR monitoring of circulating rat blood. Figure 3 shows the effect of the fraction of inspired oxygen on the reduction of PCA. Since the PCA water soluble nitroxide equilibrates quickly with the intracellular compartment, an initial fast phase of PCA distribution was followed by a slow phase due to PCA metabolism and clearance. The decay half-life of the metabolism-elimination phase decreased significantly during hypoxia. No increase was found in an arterial blood sample collected 4 min after the administration of PCA to the rat and no substantial DECONVOLUTIONS
m 2. _--____
9 min __-*
DISTANCE
( a.“. )
Fig. 2. (Left panel) EPR spectra obtained from 50 g rat whole body after 4 mmol/kg injection of PCA in the following conditions: without gradient (continuous line), with radial gradient (12 mT/m) (dotted line) and with axial gradient (12 mT/m) (dashed line). Acquisition parameters were: center field, 8.35 mT; scan range, 1.7 mT; scan time, 25 s; time constant, 300 ms; power level, 110 mW; modulation frequency, 8 kHz; width amplitude, 0.04 mT. (Right panel) The 1-D spin distribution of PCA in rat whole body along the radial (dotted line) and axial (dashed line) directions. The spectra in the presence of a gradient (left panel) were deconvoluted with the integral zero gradient line (continuous line) to obtain the 1-D spin density.
min
Fig. 3. Typical semilogarithmic plot of blood PCA signal vs. time after rapid intravenous administration (0.044 mmol/kg) in normoxia (0) followed by hypoxia (A). The corresponding halflives of the slow phase are reported. Signal amplitude was the peak-to-peak height of the central component of the PCA triplet spectrum.
changes in half-life occurred after two repeated doses of PCA in normoxic conditions. The different PCA reduction rate in hypoxia suggests that PCA could be used in vivo to generate images of differently oxygenated organs. DISCUSSION The role of free radicals and reactive oxygen species in many pathways and multiple steps leading to human cancer is as yet unresolved ( 14). Free radical damage has long been believed to be a risk factor for many processes that accompany cancer in a variety of animal species. Much research has been directed toward establishing correlations between oxidative damage, antioxidant defense systems, and cancer. Although a few modest correlations have been observed, attempts to use antioxidant manipulation (superoxide dismutase, catalase, glutathione peroxidase, vitamin E, vitamin C) in animal and cell culture models to modify the therapeutic response of malignant tumors have yielded few definitive results. Data continue to accumulate about the ubiquity of free radicals and their significant destructive capacity in living tissues. In addition, several anti-cancer drugs such as adriamycin, bleomycin, mitomycin-c, and vincristine are known to perform their tumoricidal action by a free radical-dependent mechanism and to increase free radical generation and lipid peroxidation in vitro and in vivo. It is probable that a better understanding of free radical processes will lead to the discovery of measures (dietary, pharmacology) to improve cancer care and increase life expectancy. The role of EPRI in oncology is currently under investigation in many centers throughout the world. Table 1 reports some of the most important results obtained by EPR spectroscopy and EPRI on experimental tumor models. Non-invasive in vitro and in viva technique at Lband made it possible to estimate quantitatively the effects on B-l 6 melanoma cells and growing tumors of radiosensitizing agents used separately or combined with others (22). When the biological effects of hyperthermia on B16 melanomas growing on the tail of mice were investi-
424
1. J. Radiation Oncology 0 Biology 0 Physics
Table 1. Examples of experimental tumor models studied by EPR at different frequencies Frequency 9-10 GHz (X-band)
Samples
Spin probes
Authors
Spheroids
15N-PDT
Dobrucki
Spheroids
15N 16-DS
Swartz
2-4 GHz (S-band)
Murine melanoma
TEMPOL
Lukiewicz
l-2 GHz (L-band)
Murine melanoma
CTPO
Berliner
Melanoma Murine adenocarcinoma
CTPO Fusinite
Lukiewicz Bacic
Murine fibrosarcoma
mHCTP0
Halpern
0.25 GHz
“N-PDT: 2,2,6,6,-tetramethylpiperidine-d,~-’5N-oxyl-4-one; 15N 16-DS: 15N-16-doxyl stearic acid; TEMPOL: 2,2,6,6,-tetramethylpiperidine-N-oxyl-4-01; CTPO: 3-carbamoyl-2,2,5,5-tetramethyl-3-pyrroline-I-yloxyl; mHCTP0: 4-protio-3-carbamoyl-2,2,5,5-tetraperdeuteromethyl-3-pyrrolinylI-oxy.
gated, the half-lives of TEMPOL were found to increase with temperature (21). The first EPR image of an in situ
growing tumor in a living mouse was obtained in 1987 (6). The cross-sectional image was obtained perpendicular to the tail axis using an L-band EPR spectrometer. The reconstructed image showed different concentrations of a nitroxide probe in the tumor tissue. EPRI microscopy (9 GHz) was used to study the time course of the loss of viability of spheroids at different stages of growth. To obtain 2-D images perdeuterated Tempone was used as the source of the EPR signal. For each image 64 projections were collected with a magnetic field gradient of 90 G/cm (10). Measurements of oxygen concentrations in the extracellular body water of fibrosarcoma in living mice were studied using a perdeuterated CTPO as spin probe and
Volume 29, Number 3, 1994
an L-band apparatus ( 16). This preliminary study demonstrated that in vivo the EPR technique can permit discrimination between different radiobiologic hypoxic fractions. Electron paramagnetic resonance determination of partial pressure of oxygen was recently done using fusinite, a carbon-based, inert and sensitive intratumoral probe. Changes in partial pressure of oxygen vs. localization of the probe and postirradiation were analyzed on murine mammary adenocarcinoma (5). Each new biomedical instrumentation takes cancer as its first and most formidable challenge. Electron paramagnetic resonance imaging has certain advantages over other imaging techniques. Firstly, the technique might be able to detect free radical generation in situ; secondly, EPRI does not use radiation, which is important for a cancer patient. The use of EPRI in oncology is still in its infancy and its development has been much slower than other imaging techniques currently available, but its role could be unique. Electron paramagnetic resonance imaging is still far from being a satisfactory imaging tool on account of the sensitivity limitations and the too large linewidth of the paramagnetic probes currently used. Research is needed to establish the critical relationships between free radical sources, protective systems and cancer phenomena that may be tractable to intervention. The value of EPRI in experimental oncology is beginning to be recognized and there is little doubt that further advances in technology and the development of narrow linewidth specific spin probes will allow EPRI to detect trapped free radicals. The ongoing development of pulsed EPRI techniques in our and other laboratories (7) should reduce the acquisition time of the EPR spectra reported in Figure 2 to the ms range. The fact that a different PCA reduction rate was found in hypoxia (Fig. 3) suggests that fast EPRI and appropriate spin probes might yield 3-D images with sufficient resolution to discriminate spatial heterogeneity of tumor oxygenation.
REFERENCES 1. Alecci, M.; Colacicchi, S.; Sotgiu, A.; Testa, L.; Varoli, V.
2.
3.
4.
5.
6.
Automatic trimming technique for multipolar magnets. J. Appl. Phys. 71:3053-3055; 1992. Alecci, M.; Della Penna, S.; Sotgiu, A; Testa, L. R. F. (280 MHz) EPR imaging of extended samples: Apparatus and preliminary results. Appl. Magn. Reson. 3:909-9 15; 1992. Alecci, M.; Della Penna, S.; Sotgiu, A.; Testa, L.; Vannucci, I. Electron paramagnetic resonance spectrometer for threedimensional in vivo imaging at very low frequency. Rev. Sci. Instrum. 63:4263-4270; 1992. Alecci, M.; Gualtieri, G.; Sotgiu, A.; Testa, L.; Varoli, V. Multipolar magnet for low-frequency ESR imaging. Meas. Sci. Technol. 2:32-37; 1991. Bacic, G.; Liu, K. G.; O’Hara, J. A.; Goda, F.; Harris, R. D.; Szybinski, K.; Swartz, H. M. An EPR and MRI study of oxygen tension in a murine tumor (Abstr.). Society Magnetic Resonance Medicine, Twelfth Annual Scientific Meeting 1: 105; 1993. Berliner, L. J.; Fujii, H.; Wan, X.; Lukiewicz, S. J. Feasibility
study of imaging a living murine tumor by electron paramagnetic resonance. Magn. Reson. Med. 4:380-384; 1987. Bourg, J.; Krishna, M. C.; Mitchell, J. B.; Tschudin, R. G.; Pohida, T. J.; Friauf, W. S.; Smith, P. D.; Metcalfe, J.; Harrington, F.; Subramanian, S. Radiofrequency Ff EPR spectroscopy and imaging. J. Magn. Reson. B 102: 112-l 15; 1993. Brivati, J. A.; Stevens, A. D.; Symons, M. C. R. A radiofrequency ESR spectrometer for in vivo imaging. J. Magn. Reson. 92:480-489; 199 1. 9. Colacicchi, S.; Ferrari, M.; Sotgiu, A. In vivo electron paramagnetic resonance spectroscopy/imaging: First experiences, problems, and perspectives. Int. J. Biochem. 24:205-214; 1992. 10. Dobrucki, J. W.; Demsar, F.; Walczak, T.; Woods, R. K.; Bacic, G.; Swartz, H. M. Electron spin resonance microscopy of an in vitrotumor model. Br. J. Cancer 6 1:22 l-224; 1990. 11. Dobrucki, J. W.; Sutherland, R. M.; Swartz, H. M. Nonperturbing test for cytotoxicity in isolated cells and spheroids,
In vivo electron paramagnetic resonance 0 M. FERRARIetal.
12. 13.
14. 15.
16.
17. 18.
19.
20.
using electron paramagnetic resonance. Magn. Reson. Med. 19:42-55; 1991. Eaton, G. R.; Eaton, S. S.; Ohno, K. EPR imaging and in vivo EPR. Boca Raton: CRC Press; 199 1. Froncisz, W.; Hyde, J. S. The loop-gap resonator: A new microwave lumped circuit ESR sample structure. J. Magn. Reson. 47:515-521; 1982. Goldstein, B. D.; Witz, G. Free radicals and carcinogenesis. Free Rad. Res. Comms. 11:3-10; 1990. Halpern, H. J.; Spencer, D. P.; van Polen, J.; Bowman, M. K.; Nelson, A. C.; Dowey, E. M.; Teicher, B. A. Imaging radio frequency electron-spin-resonance spectrometer with high resolution and sensitivity for in vivo measurements. Rev. Sci. Instrum. 60:1040-1050; 1989. Halpern, H. J.; Yu, C.; Peric, M.; Barth, E. Measurement of molecular oxygen concentrations in the extracellular body water of tumors of live mice with in vivo spin label EPR oxymetry (Abstr. 102). 34th Rocky Mountain Conference on Analytical Chemistry, 1992. Husband, J.; Guy, R. Magnetic resonance imaging in oncology. Gut 33:1587-1589; 1992. Ishida, S.; Matsumoto, S.; Yokoyama, H.; Mori, N.; Kumashiro, H.; Tsuchihashi, N.; Ogata, T.; Yamada, M.; Ono, M.; Kitajima, T.; Kamada H.; Yoshida, E. An ESR-CT imaging of the head of a living rat receiving an administration of a nitroxide radical. Magn. Reson. Imag. 10: 109- 114; 1992. Jones, T.; Tilsley, D. W. 0.; Wilson, C. B. J. H.; Lammertsma, A. A.; Brown, G.; Brady, F.; Price, P. M. Positron emission tomography for tumor assessment. NMR Biomed. 5:265-269; 1992. Kubrina, L. N.; Caldwell, W. S.; Mordvintcev, P. I.; Malenkova, I. V.; Vanin, A. F. EPR evidence for nitric oxide
21.
22.
23. 24, 25.
26. 27.
28.
29.
425
production from guanidino nitrogens of L-arginine in animal tissues in vivo. Biochem. Biophys. Acta 1099:233-237; 1992. Lukiewicz, S. J.; Cieszka, K.; Wojcik, K.; Lackowska, B.; Markowska, E.; Pajak, S.; Elas, M.; Dubis, K. In vivo ESR studies on the influence of hyperthermia on biological halflives (BHL) of nitroxides in B- I6 murine melanoma. Phys. Med. 5:3 15-320; 1989. Lukiewicz, S. J.; Elas, M.; Cieszka, K. In vitro and in vivo ESR studies on multimodality radiosensitization of B-16 melanoma cells and tumors (Abstr.). Society Magnetic Resonance Medicine, Tenth Annual Scientific Meeting 1~474; 1991. McCay, P. B. Application of ESR spectroscopy in toxicology. Arch. Toxicol. 60:133-137; 1987. Negendank, W. Studies of human tumors by MRS: A review. NMR Biomed. 5:303-324; 1992. Quaresima, V.; Alecci, M.; Ferrari, M.; Sotgiu, A. Whole rat paramagnetic resonance imaging of a nitroxide free radical by a radiofrequency (280 MHz) spectrometer. Biochem. Biophys. Res. Comm. 183:829-835; 1992. Sotgiu, A. Resonator design for in vivo ESR spectroscopy. J. Magn. Reson. 65:206-214; 1985. Swartz, H. M.; Glockner, J. F. Measurement of oxygen by EPRI and EPRS. In: Eaton, G. R., Eaton, S. S., Ohno, K., eds. EPR imaging and in vivo EPR. Boca Raton: CRC Press; 199 1:260-289. Takeshita, K.; Utsumi, H.; Hamada, A. ESR measurement of radical clearance in lung of whole mouse. Biochem. Biophys. Res. Comm. 177:874-880; 1991. Trush, M. A.; Kensler, T. W. An overview of the relationship between oxidative stress and chemical carcinogenesis. Free Rad. Biol. Med. 10:201-209; 1991.