Hypoxia-sensitive NMR contrast agents
Descrição do Produto
MAGNETIC RESONANCE IN MEDICINE 3,
Lihat lebih banyak...
169- 174 ( 1986)
Hypoxia-Sensitive NMR Contrast Agents HAROLDM. SWARTZ, KAI CHEN, MARKPALS, MARJETA SENTJURC,* AND PHILIP D. MORSE11 University of Illinois College of Medicine, Urbana. Illinois 61801, and *J. Stephan Institute, Ljuboana, Yugoslavia Received August 16, 1985; revised October 28, 1985 The rate of reduction of nitroxides is shown to be more rapid in hypoxic cells. The rate of reduction and the effect of hypoxia on the reduction rate vary for different nitroxides. These findings indicate that it may be feasible to develop in vivo NMR contrast agents that selectively will indicate areas of hypoxia and thereby aid in the detection of disease processes such as neoplasia, ischemia, and inflammation. 0 1986 Academic press, Inc.
In vivo NMR imaging techniques already have the potential of significantlyextending physicians’ capabilities to detect and follow various disease processes. As in other imaging techniques, the use of in vivo NMR may be extended by the use of “contrast agents” that enhance differences between regions by differentially affecting a parameter of the imaging process. In NMR imaging, contrast agents are usually paramagnetic substances which shorten the relaxation times of water protons; the two relaxation times (termed T I and T,) of water are often the most important parameters in NMR imaging. Most of the contrast agents currently under study for NMR imaging use paramagnetic metal ions which have relatively large magnetic moments (1-3). These agents provide contrast by their differential distribution in tissues. They have provided effective contrast in preliminary studies and are being very actively pursued, although there are some concerns about their potential toxicity. Nitroxide spin labels have been suggested as potential contrast agents for in vivo NMR. They have a number of attractive properties including chemical versatility and little evidence of toxicity (4-7). They have not been as vigorously pursued as contrast agents as have metal ions, in part because of concern about the metabolic conversion in vivo of the nitroxides to nonparamagnetic states with consequent loss of their ability to provide contrast. Metabolic conversion need not be a liability. We report here data that indicate that the reduction of nitroxides by cells can be utilized to extend rather than diminish their clinical utility. We find that the degree of hypoxia is a major variable for the rate of metabolism of many nitroxides. This result raises the very attractive possibility of using the differential metabolism of nitroxides in hypoxic areas to provide increased contrast for disease processes that involve hypoxia (such as occurs in tumors, inflammation, and ischemia). The cells (“TB cells”) used in these studies were a gift from Dr. Paul Wong, Department of Microbiology, University of Illinois. These cells were established in tissue 169
0740-3 194/86 $3.00 Copyright 0 1986 by Academic Press, Inc. AII rights of reproduction in any form reserved.
culture from a mixed culture of thymus and bone marrow cells (8)from CFW/D mice (9).For our studies we used passages 13-20 of a single clone that we isolated. All ESR spectra were taken on a Varian E 109-E spectrometer equipped with a Varian gas flow temperature controller. The ESR cavity (E-238 TM) was turned on its side to prevent cells from settling out of the active region of the cavity during the time course of the experiment. All samples were run at 37°C. TB cells (1 X 10’) in 1 ml of McCoy’s medium supplemented with 10% fetal calf serum were centrifuged at 200g for 1 min. The excess medium was removed to provide a final sample volume of 0.10 ml after addition of spin label. For spin labels with appreciablewater solubility (all except the doxyl stearates), the spin labels were dissolved in 150 mMNaC1-5.0 mM phosphate buffer, pH 7.3, and added to the cell sample to give a final concentration of 0.1 mM. The resultant pH was verified to be 7.3. For the experimentswith the doxyl stearates 1-3 nmol of the spin labels in ethanol were placed in a 6 X 50-mm culture tube, the ethanol was removed by vacuum, cells were added, and the sample was intermittently vortexed for 2 min. Following addition of the spin label, the sample was quickly mixed and then taken up into a gas-permeable Teflon tube (Zeus Industries, Raritan, New Jersey) with an inside diameter of 0.034 in. and a wall thickness of 0.001 in. This tube was folded in half and placed in a standard quartz 4 mm i.d. ESR tube which was open on both ends. A copper-constantan thermocouple was placed next to the sample tube to measure sample temperature and to prevent the tube from being pushed out of the active region of the cavity by the flowing gas. Air, nitrogen, or a combination of both was mixed with COz (final C02 concentration was 5%) and used as the temperature control gases. Gas flow rates were maintained at 12 scfh. To measure oxygen concentration of the gas a portion of the gas stream was bubbled into a stirred beaker held at the same temperature as the sample and the oxygen concentration was measured with a YSI model 54 APB oxygen meter and standard YSI probe. Figure 1 illustrates the changes that occur in the amount of a paramagnetic nitroxide as air or nitrogen is flowed around the sample. Initially, when the molecule is mainly in the paramagnetic (nitroxide) form and air is flowed, the concentration remains high. When nitrogen replaces the air, the concentration of nitroxide decreases rapidly (Fig. 1A). Figure 2 indicates the relationship between the rate of reduction and the concentration of oxygen for 5-doxy1 stearate. We find that the rate of reduction is a very sharp function of the concentration of oxygen (Fig. 2) and the type of nitroxide (Table 1). After the cells have decreased the concentration of nitroxide to very low levels, the concentration rises if air is reintroduced (Fig. 1B). These and other data (10-13) indicate that cells can reversably reduce nitroxides to hydroxylamines and reoxidize the hydroxylamines to nitroxides. The rate of reoxidation depends upon the amount of reduced spin label present and the observed “rate of reduction” is actually the sum of reduction and reoxygenation processes. Neither the reduction nor the reoxidation process occurs if the cells are heated to 60°C for 5 min and then studied at 37°C. Table 1 summarizes results obtained for several spin labels when flowing either air or nitrogen. The nitroxides are grouped into three classes: nitroxides which are uncharged, nitroxides which are ionizable, and nitroxides which are completely charged. Within the limits of the number of nitroxides studied, it appears that those which are
Reduction and reoxidation of a nitroxide spin label by mammalian cells. The sample consisted of lo7 TB cells with approximately 2 nmol of 5-doxy1 stearate, in a gas-permeable tube held at 36.9"C. Timedependent changes in the concentration of the spin label were obtained by setting the recorder at the peak of the middle ( m = 0) line of the ESR spectrum. Each part of the figure shows approximately 25-30 min of the experiment. (A) Line height when the sample was perfused by air (to left of arrow) and by nitrogen (to right of arrow). The initial ESR spectrum is also shown. (B) Line height in the same sample upon reintroduction of air after the signal height decreased to the level of the noise. Amplitude 4X that in (A). The ESR spectra at the start of this part of the experiment (essentially a straight line) and at the end of the experiment are also shown. FIG. 1.
uncharged or ionizable at physiological pHs are reduced more rapidly than those which are permanently charged; this suggests that membrane permeability has a role in this process. Comparing the two amine derivatives, the spin label constructed on a piperidine ring reduced more rapidly than that constructed on a pyrrolidine ring, consistent with published data on the relative susceptibility of nitroxides to reduction (14-18). In previous studies of the interactionsof nitroxides with cells and tissues, the apparent reduction of the spin labels was noted but the role of hypoxia was not discussed (15, 18,19). Our data indicate that severe hypoxia can increase the rate of reduction of some spin labels to the nonparamagnetic state by a factor as high as thirty (Fig. 2). We found that different nitroxides have maximum relative changes in reduction rates at different oxygen concentrations. Inasmuch as our study covered only a limited number of nitroxides it seems likely that suitable nitroxides can be synthesized that may be even more sensitive to the concentration of oxygen.
:. 0.0 0.000
0.030 0.060 0.090 0.120 0.150 Intracellular [ O I i n rnrnoles/l 2
FIG. 2. Effect of concentration of oxygen on the rate of reduction. The indicated oxygen concentration is the intracellular oxygen content. The reduction rates are the initial rates, assuming zero-order kinetics. Aliquots of the same suspension of cells were used. 2 nmol of 5-doxy1 stearate were deposited on the wall of a test tube to which lo7TB cells were added. Spectra were obtained at 37.0%
The data in this paper indicate that the rate of reduction can be vaned considerably by varying the structure of nitroxides so that development of nitroxides with the desired stability for detectable in vivo effects seem likely. We have also found that the rate of reduction of the nitroxides is affected by a number of other factors including cell permeability of nitroxides, concentration of the nitroxides, pH, energy metabolism, and the level of reducing substances in cells. It seems essential to carry out systematic studies to investigate these various factors. An important variable for this use of NMR contrast agents is the actual intracellular oxygen concentration. We have recently developed a method to measure intracellular oxygen (20). In the experiments described in this paper, when the concentration of oxygen of the perfusing gas (air) was approximately 6.1 mg/liter (0.190 a), the extracellular oxygen concentration of oxygen was approximately 3.8 mg/liter (0.120 mM) and the intracellular oxygen concentration was approximately 3.0 mg/liter (0.095 mM). Differences of this magnitude between intracellular and extracellular oxygen concentrations probably are typical for fairly dense cell cultures and in tissues in vivo. We conclude that it is feasible to use NMR contrast agents to reflect metabolic activity in vivo, especially hypoxia. In view of the data presented here on the range of responsiveness to hypoxia of various nitroxides and the results of preliminary studies on the range of relaxivity of various nitroxides (21), it seems reasonable to expect that suitable nitroxides can be developed that will be both metabolically responsive and good relaxers. Experiments are in progress to develop more optimum nitroxide contrast agents and to determine if it is possible to develop paramagnetic metal ion agents that also are metabolically responsive.
COMMUNICATIONS TABLE I Reduction of Spin Labels by Mammalian Cells in the Presence and Absence of Oxygen" Rates Spin labels
Uncharged Tempone Tempol 2N4 Mal-3' 5-Doxy1 stearated 12-Doxy1stearated
2.20 2.85 0.97 3.11 2.12 1.92
6.00 2.84 3.25 4.30 29.48 3.79
Partially ionized 5-Ternpamine' PCA 3-CP' Ternpamineb
0.36 0.16 0.3 1 4.09
1.43 0.92 0.56 4.55
Charged Temposulfate Zwitb Catl'
0.1 1 0.27 0.19
0.25 0.40 0.39
Abbreviations used: Tempone, 2,2,6,6-tetramethylpiperidine-N-oxyl-4-one; Tempol, 2,2,6,6-tetramethylpiperidine-N-oxyl-4-01; 2N4, 2,2',4-trimethyl-4'-ethyloxazolidine-N-oxyl; Mal-3, 2,2,5,5-tetramethylpyrrolidine-Noxyl-3-maleimide; n-Doxy1 stearate, 4,4-dimethyl-oxazolidine-N-oxyl derivative of n-ketostearic acid 5-Tempamine, 2,2,5,5-tetramethylpyrrolidine-N-oxyl-3-amine; PCA, 2,2,5,5-tetramethylpyrrolidine-N-oxyl-3-carboxylic acid; 3-CP, 2,2,5,5-tetramethylpyrrolidine-N-oxyl-3c bomyl; Tempamine, 2,2,6,6-tetramethylpiperidine-N-oxyl4-amine; Temposulfate, 2,2,6,6-tetramethylpiperidine-Noxyl4-sulfate; Zwit, 2,2,6,6-tetramethylpiperidine-N-oxyl4-ammonium-(N,Ndimethyl-N-(3-sulfopropyl)); Cat I ,
2,2,6,6-tetramethylpiperidine-N-oxyl-4-trimethylammonium. a Concentrations of spin labels are 0.1 mM except as noted. Spin labels are divided into classes which reflect their charge properties at physiological pHs. Rates are initial rate in units of molecules/min/cell x lo-''. Piperidine ring. Pyrrolidine ring. Located only in hydrophobic regions of cells, i.e., membranes; concentration not directly comparable to water soluble spin labels. ACKNOWLEDGMENTS This research was supported by NIH Grants GM-35534 and RR-01811 and the National Foundation for Cancer Research.
1. G. L. WOLF,K. R. BURNETT, E. J. GOLDSTEIN, AND P. M. JOSEPH,in “Magnetic Resonance Annual,” p. 231, Raven Press, New York, 1985. 2. Y. S. KANG,J. C. GORE,AND I. M. ARMITAGE, Magn. Reson. Med. 1,396 (1984). 3. S. H. KOENIGAND R. D. BROWN111, Magn. Reson. Med. 1,478 (1984). 4 . R. C. BRASCH,D. A. LONDON,G. E. WESBEY,T. N. TOZER,D. E. NITECKI,R. D. WILLIAMS, J. DOEMENY, L. D. TUCK,AND D. P. LALLEMAND, Radiology 147,773 (1983). 5. R. BRASCH, Rudiology 147, 781 (1983). 6. R. C. BRASCH,D. E. NITECKI,M. BRANT-ZAWADZKI, D. R. ENZMANN, G. E. WESBEY, T. N. TOZER, L. D. TUCK,C. E. CANN,J. R. FIKE,AND P. SHELDON, AJNR 4, 1035 (1983). 7. M. H. MENWNCA-DIAS, E. GAGELLI,AND P. C. LAUTERBUR, Sem. Nucl. Med. 13,364 (1983). 8. B. S. WRIGHT,Cancer Res. 27, 1672 (1967). 9. J. BALL,T. HUH,AND J. MCCARTER, Br. J. Cancer 18, 120 (1964). 10. G. SOSNOVSKY AND M. KONIECZNY, Synthesis, 619 (1977). 11. C. M. PALEOS AND P. DAIS,J. Chem. SOC.Chem Commun., 345 (1977). 12. J. KAPLAN,P. G. CANONICO, AND W. J. CASPARY, PNAS USA 70,66 (1973). 13. S. J. LUKIEWICZ AND S. G. LUKIEWICZ, Magn. Reson. Med. 1,297 (1984). 14. W. R. COUET,R. C. BRASCH,G. SOSNOVSKY, J. LUKSZO,I. PRAKASH,C. T. GNEWUCH, AND T. N. TOZER,Tetrahedron 41, 1165 (1985). 15. W. R. COUET,U. G. ERIKSSON,T. N. TOZER,L. D. TUCK,G. E. WESBEY,D. NITECKI,AND R. C. BRASCH,Pharm. Res. 1,203 (1984). 16. A. T. QUINTANILHA AND L. PACKER, PNAS USA 74, 570 (1977). 17. J. F. KEANA,in “Spin Labeling 11” (L. Berliner, Ed.), p. 115, Academic Press, New York 1979. 18. K. L. GRIFFETH,G. M. ROSEN,E. J. RAUKMAN, AND B. P. DRAYER, Inv. Radiol. 19, 553 (1984). 19. K. Y. CHENAND M. G. MCLAUGHLIN, Biochim. Biophys. Acta 845, 189 (1985). 20. P. D. MORSEI1 AND H. M. SWARTZ, Mugn. Reson. Med. 2, 114 (1985). 21. H. M.SWARTZ, H. BENNETT,R. D. BROWN111, P. D. MORSE11, M. PALS,AND S. H. KOENIG,Period. Biol. 87, 175 (1985).