In Vivo Detection of Free Radicals in Real Time

by Low-Frequency Electron Paramagnetic

Resonance Spectroscopy
Gerald M. Rosen, Sovitj Pou, and Howard J. Halpern

1. Introduction
During his studies on the properties of oxygen, Priestley (I) noted that this
gas, an essential ingredient for life processes, appears to “burn out the candle
of life too quickly.” More than two centuries would elapse, however, before
this observation would be associated with Grubbe’s (2) accounts of redness
and irritation on the hands of his workers testing X-ray tubes. By 1954,
Gerschman et al. (3) suggested that free radicals were the common element
linking the observed toxtcity of oxygen to the harmful effects of ionizing
radiation. The implication of this hypothesis seemed remote at that time. How-
ever, within a decade, the search for brologically generated free radicals would
lead to the discovery of superoxide and an enzyme that attenuated cellular lev-
els of this free radical (45). In the intervening years, free radicals have been
recognized as common intermediates in cellular metabolism (6,7), found to
play an essential role in host immune response (8) and demonstrated to regu-
late many essential physiologic functions (9).
As the biological significance of free radicals, such as superoxide and
hydroxyl radical, became apparent, methodologies were developed to identify
these reactive species. One such approach, known as spin trapping (JO),
involved the addition of free radicals to either nitrosoalkanes or nitrones
(11,12). In so doing, the initial unstable free radical is “trapped” as a long-lived
nitroxide, which can be observed by electron paramagnetic resonance (EPR)
spectroscopy at ambient temperature (13). In the intervening years, spm trap-
ping has been found to be able to simultaneously measure and distinguish
among a variety of important biologically generated free radicals (24,15). With
From Methods IR Molecular Ehology, vol 108 Free Radmal and Anfrox/dant Protocols
Edited by D Armstrong Q Humana Press Inc , Totowa. NJ

27
28 Rosen, Pou, and Halpern

the advent of lower frequency EPR spectrometers (16), attention has focused
on the m VIVOin situ spin trapping of free radicals in real time (17-19). In this
review, we will describe our recent studies measuring hydroxyl radical pro-
duction at the site where it evolved using a low-frequency EPR spectrometer in
combination with m viva spin-trapping techniques (18).

2. Materials
1. Chelex 100 (Bio-Rad, Hercules, CA).
2. Conalhumm.
3. Diethylenetriammepentaacetlc acid (DTPA).
4 Ethylenedlaminetetraacetic acid (EDTA)
5. 4-Pyridyl-1-oxide-N-tert-butylmtrone (CPOBN).
6. $5Dtmethyl-1-pyrrolme N-oxide (DMPO).
7. Phenyl N-tert-butyl mtrone (PEN).
8 EPR spectrometer.
9 Striplme plastic sample holder, ABS, resin GPX3700 (General Electric, Pitts-
field, MA).
10. Nitrones and mtrosoalkanes can be obtained from several commercial sources
To obtain accurate data, these spin traps should be of the highest purity possrble.
As iron and copper salts are common contaminants of most laboratory buffers,
removal of these ions can be accomplished by passing the buffer through an ion-
exchange resin, such as Chelex 100 (Bio-Rad) (20) In some experiments, conal-
bumin has proven to be a more rehable method at removing iron salts (21).
Verification that redox active metal ions are no longer in the buffer can be
accomplished by using a simple ascorbate assay (22). Inclusion of the metal ion
chelator, diethylenetriaminepentaacetic acid (DTPA), but not EDTA renders Fe+*,
Cu+* and Mn+* inefficient as hydroxyl radical catalysts (23). All other
reagents should be of the highest quality obtainable.

3. Methods
3.1. In Vivo In Situ Detection of Free Radicals
For the better part of a decade, we have refined techniques to allow the
detection of free radicals in living animals, culmmatmg in the recent m vtvo in
situ identification of hydroxyl radical in an irradiated tumor of a hvmg mouse
(18). This requires an EPR spectrometer, which can detect free radicals deep in
hvmg tissue Loss of signal with depth is obviated by operating at much lower
frequencies, 100-300 MHz (24). To use the low frequencies necessary for
in viva measurement of free radicals m hvmg tissue deep m animals, we chose
to develop ab initio a low frequency EPR spectrometer (16). The frequency
used, 250 MHz, is capable of measurements 7 cm deep in tissues (16,25,26)
EPR Spectroscopy 29

Because this is a magnetic resonance technique, the spectrometer is capable of

imaging (16). We have found it possible to detect spin trapped adducts at these
frequencies (27). Even though m vivo spm traps compete for free radicals with
other chemical species, the resultant effects may often result in disparate bio-
logic endpoints. Because of this, spin traps should be considered reporters of
specific free-radical events, as well as pharmaceutical agents, altermg physi-
ologic outcomes.
3.2. Enhanced Stability of Spin-Trapped Adducts
Poor cellular stability of many spin-trapped adducts have made it difficult to
verify the presence of specific free radicals (28), even with the enhanced sensl-
tivity afford by using deuterium-labeled spin traps (27,29). The search for more
stable spin trapped adducts, especially for superoxide and hydroxyl radicals,
directly led to the discovery of the spin trapping system 4-POBN plus ethanol
(EtOH) (30). In this reaction, hydroxyl radical abstracts a hydrogen atom from
ethanol, resulting in formation of a-hydroxyethyl radical. Spm trapping of this
free radical with 4-POBN gives 4-POBN-CH(CH,)OH, exhibiting remarkable
stability in homogenous solutions and cell preparations (30,31).
HO’ + CH3CH20H -> CH3CHOH + HZ0
CH3CHOH + 4-POBN -> 4-POBN-CH(CH3)0H
Formation of 4-POBN-CH(CH,)OH is specific for hydroxyl radical, because
superoxide does not catalyze formation of a-hydroxyethyl radical; rapid with a
second-order rate constant of 3.1 x 107/M/s; more sensitive towards hydroxyl
radical than either DMPO plus EtOH or PBN plus EtOH (30).
3.3. Design Characteristics of the Low-Frequency EPR
Spectrometer
We have developed ab initio a low frequency EPR spectrometer, which has
the capability to detect free radicals within animal tissues and organs (Id). The
schematic of the low-frequency EPR spectrometer is shown m Fig. 1. Five
design criteria were followed:

1. Very low frequency. 50-500 MHz, presently at 260 t- 0 MHz Skrn depth IS
7 cm (16)
2 Continuous-wave operation. This IS desirable owmg to the short relaxation time
of mtroxldes. It meansthat, unlike pulsed systems,the spectrometerIS a rela-
tively highly tuned system. As a result, we have implemented electromc feed-
back stablhzatlon of both the frequency and the coupling. There 1s point by pomt
momtormg of and correctlon for frequency change.
30 Rosen, Pou, and Ha/per-n

Fig. 1 Design of a low-frequency EPR spectrometer. (A) Striplme sample holder,

(B) modulatron coils, (C) capactive couplmg, (D) radiofrequency shield, (E) mam
magnet, and (F) horizontal plane defmmg coils.

3 Open design to accommodate varied samples mcludmg animals

4. A wide variety of resonators of striplme design Each designed to accommodate
a specific sample. This allows fillmg of the resonator with the sample, grvmg a
high filling factor (filling fraction) to which the EPR signal is directly propor-
tional. We use machmeable, platable plastic, ABS.
5. Low mam magnetic field. 90 Gauss (O.O09T), substantial gradients and the open
design argue that mam magnetic-field production and gradient production be
combined m a smgle multi-purpose magnet of nonferrous desrgn. Operation wrth
these low magnetic fields allows standard laboratory equipment to give extremely
high absolute stability--= 0.5 mdhgauss-with relatrvely moderate fractional sta-
bility (5 parts m 106). The Helmholtz coils provide similar uniformity over the
tumor With these low magnetic fields, eventual designs could include currents
Installed m walls of clinical rooms wtthout danger to personnel (the maximum
magnetic field is less than l/100 that of an magnetic resonance imaging [MRI]
spectrometer) and to other equipment and with mmimal interference if any To
generate gradrents along the axes of the resonator, we have chosen to display the
magnetrc coils. This provides uniquely linear gradients. Gradients in the perpen-
dicular horizontal direction are generated by current imbalances.
EPR Spectroscopy 31

T ,,* =267 s

800 1200 1600

Time of irradiation (s)
Rg. 2. Spectral half-life of the EPR signal derived from repeated measurements of
4-POBN-CH(CH,)OH peak height while the tumor on the leg of a mouse resided m
the spectrometer (adduct signal vs time). Half-life was calculated to be 267 s

3.4. Results
3.4.1. In Vivo In Situ Spin Trapping of Radiation Generated Free
Radicals with Very Low-Frequency EPR Spectroscopy
Several years ago, we reported the use of spur trapping to image and identify
hydroxyl radicals produced by ionizing radiation m buffered solutions (27).
These solutions mimic the drfficult spectroscopic penetrability of living tissue.
The same low frequency-250 MHz-spectrometer described herein and deu-
terated spin traps DMPO-dr 1and DMPO-d, were used. Partial deuteration sim-
plified the spectral stgnature (DMPO-d3) and full deuteratton (DMPO-dll)
increased the signal intensity. These results established the feasibility of in
vivo spin trapping (27), even though in vivo studies would require the more
stable spin-trapped adduct, 4-POBN-CH(CHs)OH. For these initial expert-
ments, we chose an extremity tumor to deliver high, toxic doses of radiation to
a substantial bulk of the tissue with mmimal effect on the physiology of the
rest of the animal.
The identification of radrolytrc-generated hydroxyl radical, as 4-POBN-
CH(CH,)OH, is dependent on the pharmacodynamics properties of the tumor.
As 4-POBN IS a zwitterion, with an octanol/water partition coefficient of only
32 Rosen, Pou, and Halpern

In Vivo IOOOGy
B

In Vzvo 3000 Gy
C

Solution 3000 Gy
D

Fig. 3. EPR spectrum of 4-POBN-CH(CH3)0H obtained by irradiating a tumor m

the leg of a mouse, as described m ref 18. (A), 500 Gy (B), 1000 Gy. (C), 3000 Gy.
(D), EPR spectrum of 4-POBN-CH(CH,)OH, generated by irradiating a solution of 4-
POBN and ethanol with 3000 Gy.

0.09 (30), the detection of 4-POBN-CH(CH3)0H will depend upon the

concentration of hydroxyl radical and EtOH in vascular compartments and the
rate of spin-trapped adduct diffusion from the site of irradiation. The dynamics
of this process is illustrated by followmg the pharmacokinetics of the posl-
EPR Spectroscopy 33

tively charged nitroxide, 3-trimethylaminomethyl-2,2,5,5-tetramethyl- l-

pyrrolidinyloxyl iodide, with an octanol/water partition coefficient co.02 (18).
Its TtlZ of elimination from a leg tumor increased from 8 to 35 min after placing
a tourniquet around the tumor (18). In a similar experimental design, 4-POBN-
CH(CH,)OH, whose octanol/water partition coefficient is 0.22 (18), was elimi-
nated from the tumor more rapidly, with Tt12 approaching 4.5 min (Fig. 2).
Taking into account these limitations, has allowed us to directly detect radt-
olytic-generated hydroxyl radical in a leg tumor in a living mouse with the
very low-frequency EPR spectrometer and the 4-POBN plus EtOH spin trap-
ping system (18). Fig. 3A depicts the EPR spectrum of 4-POBN-CH(CHs)OH,
generated by irradiating a leg tumor of a mouse with 500 Gy. Fig. 3B,C
was derived under similar experimental conditions, except the radiation was
increased to 1000 and 3000 Gy, respectively. Verification of the reaction was
obtained by irradiating an aqueous solution of 4-POBN plus EtOH. These data
demonstrate the feasibility of spin trapping hydroxyl radical in viva and the
ability to detect this spin trapped adduct in a living animal in real time. This is
the first example of the detection of hydroxyl radical in a living animal in real
time. Nevertheless, the in vivo in situ identification of other free radicals has
begun (17,19) and the future is even more promising. With advances in the
design of lower frequency EPR spectrometers and new spin traps, in viva in
situ identification of free radicals in real time is within reach.

Acknowledgment
This research was supported in part from a grant from the National Institutes
of Health, CA-69538.

References
1. Gilbert, D. L. (1981) Perspective on the History of Oxygen and Life, in Oxygen
and Living Processes. An Znterdisciplinary Approach (Gilbert, D. L., ed.),
Springer-Verlag, NY, pp l-43.
2. Grubbe, E. H. (1933) Priority in the therapeutic use of X-rays. Radiology 21,
156-162.
3. Gerschman, R., Gilbert, D. L., Nye, S. W., Dwyer, P., and Fenn, W. 0 (1954) Oxy-
gen poisoning and x-irradiation: a mechanism m common Science 119,623-626.
4. McCord, J. M and Fridovich, I (1968) The reduction of cytochrome c by milk
xanthine oxidase. J Biol. Chem. 243, 5753-5760.
5. McCord, J. M. and Fridovich, I. (1969) Superoxide dismutase. an enzymic func-
tion for erythrocuprein (hemocuprem) J. Biol. Chem. 244,6049-6055.
6. Fridovich, I. (1978) The biology of oxygen radicals. Science 201, 875-880
7. Guengerich, F. P. and Macdonald, T. L. (1984) Chemtcal mechanisms of catalysis
by cytochrome P-450: a unified view. Act. Chem. Res. 17,9-16.
34 Rosen, Pou, and Halpern

8. Babior, B. M , Krpnes, R. S., and Curnutte, J. T. (1973) Biological defense mecha-

nisms the productron by leukocytes of superoxide, a potential bactericidal agent.
.I. Clin. Invest. 52,741-744
9. Moncada, S. and Higgs, A (1993) The L-argmme-nitric oxide pathway. N Engl
.I. Med. 329,2002-2012.
10. Janzen, E. G. and Blackburn, B. J. (1968) Detection and Identification of short-
lived free radicals by an electron spm resonance trapping technique. J. Am Chem
sot 90,5909-5910
11. Mackor, A., Wajer, Th A J W., and de Boer, Th. J (1967) C-Nitroso compounds
Part III Alkoxy-alkyl-mtroxides as intermediates m the reaction of alkoxyl-radi-
cals with nitroso compounds. Tetrahedron Lett. 385-390.
12. Iwamura, M and Inamoto, N (1970) Reacttons of mtrones with
free radicals. II Formation of mtroxtdes. Bull. Chem. Sot. Japan 43, 860-863
13 Janzen, E. G. (1971) Spin trapping. Acct. Chem. Res 4, 31-40.
14. Fmkelstem, E., Rosen, G. M., and Rauckman, E J. (1980) Spur trapping of
superoxide and hydroxyl radical. practical aspects. Arch. Biochem. Bzophys.
200,1-16.
15. Janzen, E. G. (1980) A critical review of spur trapping m btologrcal systems, m
Free Radicals cn Biology, (Pryor, W. A., ed ), vol 4, Academic Press, NY, pp
116-154
16. Halpern, H. J., Spencer, D. P , vanPolen, J., Bowman, M. K., Nelson, A C.,
Dowey, E. M., and Teacher, B. A (1989) Imagmg radio frequency electron-spm-
resonance spectrometer with high resolution and sensitivity for m vlvo measure-
ments. Rev. Scz. Znstrum 60, 1040-1050
17. Lai, C-S. and Komarov, A. M. (1994) Spm trapping of nitric oxrde produced m
vivo in septic-shock m mice FEBS Letts. 345, 120-124
18. Halpern, H. J , Yu, C., Barth, E , Perrc, M., and Rosen, G. M. (1995) In sztu detec-
tion, by spin trapping of hydroxyl radical markers produced from ionizing radia-
tron m the tumor of a hvmg mouse. Proc. Nat1 Acad Set USA 92,796-800
19. Jiang, J. J., Liu, K. J., Shr, X., and Swartz, H. M. (1995) Detection of short-lived
free radicals by low-frequency electron paramagnetrc resonance spin trapping m
whole living animals. Arch. Biochem. Biophys. 319, 570-573.
20. Poyer, J. L. and McCay, P. B (1971) Reduced trrphosphopyrrdme nucleotrde oxr-
dase-catalyzed alterations of membrane phosphohprds. IV* dependence on Fef3.
J. Biol. Chem. 246,263-269.
21. Gutterrdge, J. M. C (1987) A method for removal of trace iron contammatron
from biological buffers. FEBS Lett 214, 362-364.
22. Buettner, G R. (1988) In the absence of catalytic metals ascorbate does not
autoxldrze at pH 7. ascorbate as a test for catalytic metals. J. Biochem. Biophys
Methods 16,27-40
23. Buettner, G. R , Oberley, L. W., and Leuthauser, S. W. H. C. (1978) The effect of
iron on the distrrbutron of superoxrde and hydroxyl radrcals as seen by spin trap-
pmg and on the superoxide drsmutase assay. Photochem. Photobtol. 28,693-695.
EPR Spectroscopy 35
24. Bottomley, P A., and Andrew, E. R. (1978) RF filed penetration, phase shaft, and
power drssrpatlon in brologrcal tissues: implications for NMR Imaging. Phys. Med.
Biol 23,630-643.
25. Johnson , C. C. and Guy, A. W. (1972) Nomonizmg electromagnetic wave effects
m brologrcal materials and systems. IEEE 60, 692-7 18.
26. Roschmann, P. (1987) Radlofrequency penetration and absorptron m the human
body: hmitatlons to high-field whole-body nuclear magnetic resonance imagmg.
Med. Phys 14,922-928
27. Halpern, H. J., Pou, S., Pent, M., Yu, C., Barth, E., and Rosen, G. M (1993)
Detection and imaging of oxygen-centered free radicals with low-frequency elec-
tron paramagnetrc resonance and signal-enhancing deuterium contammg spur
traps. J Am. Chem Sot. 115,218-223
28 Pou, S , Cohen, M. S , Britigan, B E., and Rosen, G. M. (1989) Spin trapping and
human neutrophrls: Lrmits of detection of hydroxyl radrcal J. Blol Chem. 264,
12299-12302.
29. Pou, S., Rosen, G. M., Wu, Y., and Keana, J. F W. (1990) Synthesis of deute-
rium- and t5N-contammg pyrroline I-oxides: a spin trapping study. J Org. Chem
55,4438-4443.
30. Pou, S., Ramos, C.L., Gladwell, T., Renks, E., Centra, M., Young, D , Cohen, M
S , and Rosen, G. M (1994) A kinetic approach to the selection of a sensrtrve spin
trapping system for the detection of hydroxyl radical Anal. Blochem. 217,76-83
31. Ramos, C. L , Pou, S , Brrtigan, B E., Cohen, M S, and Rosen, G M. (1992) Spm
trappmg evidence for myeloperoxidase-dependent hydroxyl radical formation by
human neutrophils and monocytes. J. Biol. Chem 267, 8307-8312