Measurement of the Intracellular Concentration of Oxygen in a Cell Perfusion System Kai Chen, Cheng E. Ng, Jay L. Zweier, Periannan Kuppusamy, Jerry D. Glickson, Harold M. Swartz

[OJ was measured in the embedding material (alginate) in a typical apparatus for conducting studies of viable cells with NMR, using low frequency EPR. In suspension cultures res- piration was independent of [O,] in the perfusing media down to about 1 pM while in alginate beads, the comparable value was 70 pM, indicating that the alginate was a very substantial barrier to the free diffusion of oxygen. With knowledge of [O,] in the various compartments, [OJ in the perfusing medium can be increased and the full power of NMR can be used to provide information on metabolism under various conditions. These results also provide evidence supporting the feasibility and usefulness of EPR techniques using nitroxides to meas- ure [O,] in macroscopic samples such as NMR perfusion tubes. This technique is rapid, apparently nonperturbing, and enables one to differentiate between the concentrations of oxygen in different compartments. Key words: oxygen; EPR; perfusion systems; nitroxides.

INTRODUCTION

Oxygen plays a critical role in physiology and patho- physiology. The detection and characterization of altered oxygen concentrations, especially in regard to physi- ological stress, ischemia, and cancer, is of great potential value for understanding these processes, for clinical di- agnoses, and for guiding and understanding therapies. NMR spectroscopy provides a noninvasive method for studying metabolism, especially in relation to redox me- tabolism and the occurrence of hypoxia. It is a potential clinical tool for detecting the presence of ischemia and guiding its treatment, and for predicting and following the response of tumors to radiation and chemotherapy (1-3). However, NMR spectroscopy usually does not di- rectly measure the concentration of oxygen ([O,]); it de- tects a number of oxygen dependent metabolic processes which should serve as indices of cellular oxygenation when suitably calibrated with regard to [O,]. A critical step for the full and effective use of NMR spectroscopy in

MRM 31:668-672 (1994) From the Division of NMR Research, Department of Radiology and Radio- logical Sciences (K.C., C.E.N.. J.D.G., H.M.S.) and The EPR Laboratories, Department of Medicine, Cardiology Division (J.L.Z., P.K.), The Johns Hop- kins University School of Medicine, Baltimore, Maryland, and the Depart- ment of Radiology, Dartmouth Medical School, Hanover, New Hampshire (H.M.S.). Address correspondence to: Harold M. Swartz, Ph.D., 7252 Strasenburgh, Room 308, Dartmouth Medical School, Hanover, NH 03755-3863. Received December 13, 1993; revised February 9, 1994; accepted February 9, 1994. This work was supported by grants from the American Cancer Society PDT- 63063 (H.M.S.) and RD-277 (Fellowship, K.C.), and the National Institutes of Health GM 34250 (H.M.S.), CA 51935 and 51950 (J.D.G.), and HL 38324 (J.L.Z.).

Copyright 0 1994 by Williams 8 Wilkins All rights of reproduction in any form reserved.

0740-31 94/94 $3.00

clinical and experimental medicine, therefore, is the ca- pability of making such calibrations. Unfortunately, the existing methods of measuring [O,] in tissues have seri- ous limitations in regard to both sensitivity and accuracy.

Nitroxide spin labels and other paramagnetic materials recently have been developed as noninvasive probes for measuring [O,] which can be monitored by electron para- magnetic resonance (EPR or, equivalently, electron spin resonance, abbreviated as ESR) spectroscopy (4). The use of these probes, combined with recent instrumental de- velopments which extend the use of EPR to large, water- containing specimens, makes it feasible to study the con- centration of oxygen in the same types of samples used for NMR studies of cells, tissues, and animals (5, 6) . The aim of this study was to determine the validity of an approach that employed EPR methods to measure [O,] in perfused cells under conditions identical to those em- ployed in NMR experiments in one of our laboratories (7, 8), in order to be able to relate various NMR spectro- scopic parameters to [O,] in cells and to evaluate the level of oxygenation in the apparatus.

MATERIAL AND METHODS Cell Culture

RIF-1 mouse fibrosarcoma cells were routinely main- tained by alternate in vivo and in vitro passage according to the procedure of Twentyman et al. (9). Cultured cells were grown in Waymouth’s medium (GIBCO Laborato- ries, Grand Island, NY) supplemented with 10 mMHepes buffer, 10 units ml penicillin, 10 pg/ml streptomycin and 10% (v/v) fetal bovine serum (FBS; Flow Laboratories, Costa Mesa, CA) in a humidified atmosphere of 95% air/5% CO,. For EPR studies and measurement of cellular respiration experiments, cells in the late exponential phase of growth were harvested by trypsinization, and an aliquot was stained with crystal violet and cell counts made with a hemocytometer. 1.0 X lo8 cells in 0.9 ml of calcium and magnesium-free Hank’s balanced salt solu- tion were then uniformly mixed with 1.1 ml of 2.4% (w/v) sodium alginate essentially as described by Narayan et al. (10). Spherical calcium alginate beads of 2.6 2 0.1 mm diameter were formed by extruding the mixture of cells and alginate through a 27-gauge needle into 100 mM CaC1,. After several washes with perfusion medium, the beads containing cells were transferred to the perfusion system.

Perfusion for EPR Spectroscopy

Cells in alginate beads were perfused under aseptic con- ditions in a system suitable for either EPR or NMR exami- nation (Fig. l), for further details, see references 7 and 8) .

668

Measurement of Intracellular 0, in a Cell Perfusion System

95%/5% 02 IC02

Oxygen Mlcroelectroda

I ' l l I

Porous End

Standard

a 10 mm NYR Tube Fiber Optic Thermometer

f 95% 0215% Cop

Aspiration Line

Perfusate Inflow

Capillary containing Standard

b

Gas Line

Fiber Optic Probe

Perforated Teflon Plug

Sample Chamber

A commercial 10-mm screw cap NMR tube (Wilmad, Butma, NJ) was fitted with perforated inserts and tubing to create a 2.1-ml sample chamber. Perfusate (350 ml of a modified Waymouth's medium prepared with inorganic phosphate and supplemented with 10% FBS, 10 mM Hepes, and antibiotics, pH 7.4) was recirculated at a rate of 10 to 16 ml/min using a peristaltic pump; it was pumped to the bottom of the tube, flowed upward through the cell sample, and then returned to the reservoir. Gas equilibration (0, fraction as indicated, with 5% CO,) was maintained in the head space above the sample and in the perfusate reservoir. All tubing was of low gas permeabil- ity; polyethylene was used in the sample chamber and NorpreneB tubing (Norton Co.) was used for perfusion lines and pump tubing. The perfusate was warmed by placing the reservoir in a constant temperature bath and the sample chamber was maintained at 35 ? 1°C.

Measurement of Cellular Respiration Rate

The cellular respiration rate was measured in a sealed chamber that was maintained at 37°C. Cells or beads containing cells were stirred by a very small magnetic stirring bar. The oxygen concentration was measured by

669

Partusate

FIG. 1 . Schematic description of the perfusion system used for EPR and NMR experiments. (a) The Overall System. (b) Details of the perfusion chamber.

a Clark oxygen electrode and was recorded by a chart recorder.

EPR Measurements

The oxygen sensitive nitroxide, 4-0~0-2,2,6,6-tetrameth- ylpiperidine-d,,-l-l-oxyl ("N PDT) was purchased from MSD Isotopes (St. Louis, MO) and was used without fur- ther purification. EPR spectra were obtained at L-band (-1.2 GHz) using a specially designed microwave bridge containing a cavity stabilized transistor oscillator as the frequency source, a directional coupler with a test port connected to a frequency counter, a variable attenuator with a 60-dB range, a three port circulator, a Shotkey diode detector, and a preamplifier stage with output con- nected to the input of the spectrometer signal channel lock-in amplifier. The oscillator had a maximum power output of 100 mW and was locked to the resonator fre- quency with an AFC feedback loop. This instrumentation has previously been described in detail (6). The sample resonator used was a recessed gap loop gap resonator of radius 13 mm identical in design to that described pre- viously for use in perfused heart studies. The L-band

670 Chen et al.

bridge was interfaced to an IBM-Bruker ER 300 EPR spec- trometer with a 13-inch magnet.

RESULTS AND DISCUSSION

The EPR spectrum of 2 mM I5N PDT in a perfusion sys- tem under conditions identical to those in NMR experi- ments (7, 8 ) is shown in Fig. 2. The I5N PDT which is added to the perfusate distributes throughout the extra- cellular and intracellular space, inside and outside the beads. In order to observe selectively the EPR spectrum of intracellular PDT, 30 mM potassium ferricyanide, a charged, membrane-impermeable paramagnetic broaden- ing agent, was added to the perfusate to broaden out the extracellular EPR signal. The potassium ferricyanide was mixed with the diluted medium and the perfusate was adjusted to be isotonic. Under these conditions, two com- ponents of the signal are seen: the sharp one is from the intracellular compartment and the broad one is from the extracellular compartment. These results demonstrate that we can obtain adequate data from the intracellular environment.

In order to confirm that the sharp spectrum in Fig. 2 was due to intracellular PDT, the same sample was freeze- thawed twice and another spectrum was obtained (Fig. 3); as expected, the sharp spectrum was now replaced by a broad resonance. Also, no sharp signal was seen in the absence of cells, and previous studies have demonstrated that when cells have damaged membranes, they cannot exclude ferricyanide (11). Therefore, the observed signal is from the intracellular compartment of viable cells.

The calibration curve for the relation in the perfusate between the line width of the EPR spectrum of PDT and the [O,] measured with a calibrated Clark micro-elec- trode in the NMR tube is shown in Fig. 4.

The calibration of the relationship between the line width of the EPR spectrum of intracellular PDT and measured [O,] was obtained from a sample in which 0, is in equilibrium between the intracellular compartment and the perfusate (open circles, Fig. 5). Equilibration was

FIG. 2. EPR spectrum of 2 mM I5N PDT in a perfusion system containing lo* RIF-1 cells and 30 mM potassium ferricyanide ex- perimental conditions included. The RIF-1 cells were embedded in the alginate beads and perfused with the medium while the spec- trum was taken. The sample was in a 10-mm NMR tube placed in a loop-gap resonator tuned at 1.2 GHz, the modulation amplitude was 0.5 G, the microwave power was 100 mW, and the scan range was 50 G.

FIG. 3. EPR spectrum of I5N PDT after freeze-thawing to allow ferricyanide to enter the intracellular compartment. The same sample in Fig. 2 was freeze-thawed twice before the spectrum was taken. Other experimental conditions were the same as those in Fig. 2.

40%- 750

350 7 L o o

- A

is00 3 - 0 ~450 c - 0 : a -400-

~ 3 5 0

1 0

-50

Extracellular Oxygen Conc. (pM)

FIG. 4. Calibration of the changes in line width of 15N PDT with the concentration of oxygen. The changes in the line width of the EPR signal were calculated by subtracting the line width at a specific concentration of oxygen from the line width at zero concentration of oxygen. The concentrations of oxygen were measured by a micro- electrode in the 10-mm NMR tube.

700 600 I

i , I " ' I " ' I ' " I ' " I " ' I ' "

0 100 200 300 400 500 600 700

Extracellular [02] (pM)

FIG. 5. Measurement of intracellular oxygen concentrations. Line widths were calibrated in cells whose respiration was inhibited with 5 mM cyanide (open circle). The concentration of oxygen in the perfusate was measured with a microelectrode intracellular oxygen concentration then measured in respiring RIF-1 cells embedded in alginate beads and perfused with the medium (closed circles). Other experimental conditions were the same as in Fig. 2.

obtained by inhibiting respiration with 5 mM sodium cyanide to eliminate the source of a potential gradient in

Intracellular [O,] under conditions of active metabo- lism was measured using the calibration curve for intra- cellular [O,]; the results are indicated in Fig. 5 (closed circles). Note that there was a gradient between the [O,]

[0,1 (121.

Measurement of Intracellular 0, in a Cell Peqfusion System 671

in the intracellular compartment and in the perfusate (extracellular compartment) with actively metabolizing cell:;. This is an important observation because in most NMR experiments with cells, it usually is assumed that the intracellular [O,] is the same or very close to extra- cellular [O,]. This observed gradient in [O,] between the intracellular and extracellular compartment was not in- fluenced by back diffusion of oxygen into and out of the head space above liquid in this system when the perfu- sion rate was 10 to 16 ml/min. This was confirmed by the oxygen consumption experiments described below.

For both experimental and theoretical reasons, it is important to understand the basis of the observed gradi- ent in [O,] between the intracellular and extracellular compartment. The intracellular oxygen concentration de- creased as the oxygen concentration in the perfusate was reduced, reaching less than 10 pM when [O,] in the per- fusate was 100 w. If the rate limiting step for the diffu- sion of 0, resides in the cells themselves, then a meas- urement of consumption of 0, versus [O,] in cells in suspension should show a relationship similar to that observed in the perfusion system. As indicated in Fig. 6, this is not the case; the rate of 0, consumption shows zero order kinetics down to very low [O,], indicating that there is little barrier to the passage of 0, into the cells under these conditions. In contrast, when the cells are embedded in alginate beads, there is a change in the rate of ccinsumption of 0, that follows approximate first order kinetics in respect to [O,] (Fig. 7); this is consistent with the alginate beads being the site of the limitation of dif- fusion of 0,.

To confirm that the observed dependence of the respi- ration rate on [O,] was due to limited diffusion in alginate rather than an artifact associated with stacking or packing of cells, the experiments were repeated with lower cell density in the beads. The dependency of [O,] remained similar to that shown in Fig. 7, indicating that the ob- served limitation of diffusion of 0, was due to the algi- nate beads.

The cellular respiration rate as a function of oxygen

200 : A - = 175;

P -

150:

125;

w -

0 0 11,o:

6 '75:

0 , - !so; I

\ I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0 100 200 300 400 Time (sec)

FIG. 6. Oxygen consumption by cells in suspension. The RIF-1 cells were in simple suspension, instead of being embedded in alginate beads. The oxygen concentrations were measured by a Clark electrode in a sealed NMR tube at 37"; the decreases in [O,], therefore, reflected the consumption of oxygen.

FIG. 7. Oxygen consumption by cells in alginate beads. Condi- tions were similar to Fig. 6 except the RIF-1 cells were embedded in alginate beads.

concentration was calculated from the digitized data in Fig. 7 and is shown in Fig. 8. The Statistical Analysis System (SAS Institute Inc., Raleigh, NC) was used to fit the data to saturation kinetics:

v = V,;PO,/(P,, + PO,) [ l I where Vis cellular respiration rate; V,, is maximal value for V; PO, is concentration of oxygen in the medium; and P,, is the concentration of oxygen at half-maximal respi- ration. The best fit value for half-maximal respiration (P5,) was 70 pM and for maximal respiration rate ( Vma) was 1.1 pM/s, for the cells trapped in the alginate beads. The V,, value was proportional to the concentration of cells in the experiment. The high value of 70 pA4 for P,, for cells in alginate (P5, for cells in suspension is approxi- mately 1 pM) is consistent with the presence of a signifi- cant gradient of oxygen concentration between the me- dium and the area near mitochondria in cells trapped in the alginate beads, and it is consistent with the results from the above EPR experiments.

. . ; 1.0 v, 0

P W

C 0.6 0

Q E 0.4 3 rn c

.- 4

0.2 0

6 0.0

0 0

I?<,, , , , I , , , , I , , , , , , , , , I , , , , , , , , , I , , , ,

0 25 50 75 100 125 150 175 2 Oxygen Conc. (pM)

0

FIG. 8. Calculated rate of respiration plotted against concentra- tion of oxygen. The solid line is a best fit of the data to Eq. [l].

672

CONCLUSIONS

These results clearly demonstrate the need for and value of making accurate and appropriate direct measurements of [O,] in complex experimental situations such as NMR perfusion tubes. They demonstrated that large gradients of [O,] occurred which were not expected. It seems likely that the need for such measurements is even greater in the much more complex situations that occur in vivo.

The results have specific implications for the use of cellular embedding techniques that use viscous materials such as alginate or Matrigel (13). As shown in these stud- ies on alginate, these media can be very substantial bar- riers to the free diffusion of oxygen and, therefore, can leiid to substantial errors in interpretation of results if these phenomena are not recognized. In suspension cul- tures the rate of respiration was independent of [O,] in the perfusing media down to about 1 pM (‘‘P50”) while in alginate beads, for the same cells, the comparable value was 70 w. That indicates that for cells in alginate the intracellular [O,] was similar to that of cells in suspen- sion when the [O,] in the perfusion medium was 70-fold higher than for cells in suspension. With knowledge of [O,] in the various compartments, [O,] in the perfusing medium can be increased and the full power of NMR can be used to provide information on metabolism under various conditions.

These results also provide evidence supporting the fea- sibility and usefulness of EPR techniques to measure [O,] in macroscopic samples such as NMR perfusion tubes, providing high accuracy and sensitivity in the range of 10 pM to more than 700 pM. This technique is rapid and apparently nonperturbing. Equally important, the tech- nique enables one to differentiate between the concen- trations of oxygen in different compartments; other meth- ods have not yet developed this capability. Under most circumstances, the relevant [O,] is that of the intracellu- lar compartment (12).

The development of the instrumentation needed to carry out EPR studies in macroscopic aqueous samples such as NMR perfusion tubes has occurred recently and appears to be increasing rapidly in its capabilities (5). Similar capabilities are becoming available for in vivo as well as in vitro studies (5, 14).

Although the paramagnetic probes used for these stud- ies provided the accuracy and sensitivity needed for the purpose of these experiments, they may not provide suf- fic:ient sensitivity to measure [O,] at the lower concentra- tions associated with changes in radiation sensitivity (about 5 pM) or where [O,] becomes limiting for the rate of respiration (about 0.5 pM). Fortunately, paramagnetic probes with a more than lo-fold increase in sensitivity have become available recently (14). It now seems fea- sible to apply the new capabilities of EPR instrumenta- tion to obtain well-resolved spectra in samples which contain large amounts of water such as NMR tubes and tissues in vivo, enabling one to make measurements of [O,] over the entire range needed to relate NMR spectros-

Chen et al.

copy to actual [O,] and to differentiate between [O,] in extracellular and intracellular compartments.

ACKNOWLEDGMENTS

The authors thank Penny C. Jenkins for technical assistance with the cell cultures.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10,

11.

12

G. K. Radda, B. Rajagopalan, D. J. Taylor, Biochemistry in vivo: an appraisal of clinical magnetic resonance spectros- copy. Magn. Reson. Q. 5, 121-151 (1989). J. D. Glickson, J. P. Wehrle, S. S. Rajan, S. J. Li, R. G. Steen, NMK spectroscopy of tumors, in “NMR Principles and Ap- plications to Biomedical Research” (J. W. Pettegrew, Ed,), pp. 255-309, Springer-Verlag, New York, 1990. W. Negendank, Studies of human tumors by MRS: a review. NMR Biomed. 5, 303-324 (1992). J. Hyde, W. K. Subczynski, Spin label oximetry, in “Biologi- cal Magnetic Resonance, Spin Labeling: Theory and Appli- cation’’ (L.-J. Berliner, J. Reubens, eds.), vol. 8, pp. 399425, Plenum Press, New York, 1989. H. M. Swartz, J. F. Glockner, Measurement of oxygen by ESRI and ESRS, in “EPR Imaging and In Vivo EPR” (G. R. Eaton, S. S. Eaton, K. Ohno, eds.), pp. 261-290, CRC Press, Inc., Boca Raton, FL, 1991. J. L Zweier, P. Kuppusamy, Electron paramagnetic reso- nance measurements of free radicals in the intact beating heaIt: a technique for detection and characterization of free radicals in whole biological tissues. Proc. Natl. Acad. Sci.

C. E. Ng, K. A. McGovern, J, P. Wehrle, J. D. Glickson, 31P NMR spectroscopic study of the effects of y-irradiation on RIF-1 tumor cells perfused in vitro. Magn. Reson. Med. 27,

K. A. McGovern, J. S. Schoeniger, J. P. Wehrle, C. E. Ng, J. D. Glickson, An I g F NMR spectroscopic method for monitoring oxygen concentration in cell perfusion systems. Magn. Re- son. Med. 29, 196-204 (1993). P. R. Twentyman, J. M. Brown, J. W. Gray, A. J. Franko, M. A. Scoles, R. F. Kallman, A new mouse tumor model system (RIF-1) for comparison of end-point studies. J. Natl. Cancer Inst 64, 595-604 (1980). K. S. Narayan, E. A. Moress, J. C. Chatham, P. B. Barker, 31P NMR of mammalian cells encapsulated in alginate gels uti- lizing a new phosphate-free perfusion medium. NMR Biorned. 3, 23-30 (1990). J. D. Dobrucki, R. M. Sutherland, H. M. Swartz, Nonperturb- ing lest for cytotoxicity in isolated cells and spheroids using electron paramagnetic resonance. Magn. Reson. Med. 19,

J. Glockner, H. M. Swartz, M. Pals, Oxygen gradients in CHO cells: measurement and characterization by electron spin resonance. 1. Cell. Phvsiol. 140, 505-511 11989).

( U S A ) 85, 5703-5707 (1988).

296-309 (1992).

42-55 (1991).

13. P. F. Daly, R. C. Lyon, E. J. Straka, J. S. Cohen, 31P-NMR spectroscopy of human cancer cells proliferating in a base- ment membrane gel. FASEB J. 2, 2596-2604 (1988).

14. H. M. Swartz, S. Boyer, P. Gast, J. F. Glockner, H. Hu, K. J. Liu, M. Moussavi, S. W. Norby, T. Walczak, N. Vahidi, M. Wu, R. B. Clarkson, Measurements of pertinent concentra- tions of oxygen in vivo. Magn. Reson. Med. 20, 333-339 (1991).