Top

Is the Mammalian Cell Plasma Membrane
a Barrier to Oxygen Transport?

Source

WITOLD K. SUBCZYNSKI, LARRY E. HOPWOOD, and JAMES S. HYDE

From the National Biomedical Electron Spin Resonance Center, Department of Radiology and
Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, Wisconsin
53226; and Department of Biophysics, Institute of Molecular Biology, Jagiellonian University,
31-120 Krakow, Poland

ABSTRACT Oxygen transport in the Chinese hamster ovary (CHO) plasma membrane has been studied by observing the collision of molecular oxygen with nitroxide radical spin labels placed in the lipid bilayer portion of the membrane at
various distances from the membrane surface using the long-pulse saturationrecovery
electron spin resonance (ESR) technique. The collision rate was estimated for 5-, 12-, and 16-doxylstearic acids from spin-lattice relaxation times (TI) measured in the presence and absence of molecular oxygen. Profiles of the local oxygen transport parameters across the membrane were obtained showing that the oxygen diffusion-concentration product is lower than in water for all locations at 37°C.

From oxygen transport parameter profiles, the membrane oxygen permeability
coefficients were estimated according to the procedure developed earlier by Subczynski et al. (Subczynski, W. K., J. S. Hyde, and A. Kusumi. 1989. Proceedings of the National Academy of Sciences, USA. 86:4474-4478). At 37°C, the oxygen permeability coefficient for the plasma membrane was found to be 42 cm/s, about two
times lower than for a water layer of the same thickness as the membrane. The oxygen concentration difference across the CHO plasma membrane at physiological conditions is in the nanomolar range. It is concluded that oxygen permeation across the cell plasma membrane cannot be a rate-limiting step for cellular respiration.

Correlations of the form PM = cK~ between membrane permeabilities PM of small nonelectrolyte solutes of mol wt <50, including oxygen, and their partition coefficients K into hexadecane and olive oil are reported. Hexadecane: c = 26 cm/s, s = 0.95; olive oil: c -- 23 cm/s, s = 1.56. These values ofc and s differ from those reported in the literature for solutes of 50 < mol wt <300 (Walter, A., and J.
Gutknecht. 1986. Journal of Membrane Biology. 90:207-217). It is concluded that oxygen permeability through membranes can be reliably predicted from measurement of partition coefficients.

INTRODUCTION

Knowledge of the concentration, diffusion, and transport of molecular oxygen in
tissues, ceils, and subcellular structures is required to solve important oxygen-related physiological and toxicological problems (McCord, 1985) and to understand the basis for radiation (Hill and Pallavicini, 1983) and photodynamic (Kalyanaraman, Feix,
Sieber, Thomas, and Girotti, 1987) therapy.
The rate of oxygen consumption of a suspension of cells can be measured using
various approaches (Lai, Hopwood, Hyde, and Lukiewicz, 1982; Froncisz, Lai, and
Hyde, 1985, and citations therein).

It was found in these two studies that the rate of
oxygen consumption of Chinese hamster ovary (CHO) cells was independent of dissolved oxygen concentration down to levels as low as ~ 1 I~M. It is therefore apparent that the respiratory rate in this cellular system is under enzymatic control
over a wide range of oxygen tensions, and the question of whether or not the membrane is a barrier to oxygen transport can be relevant only at low concentrations.

Froncisz et al. (1985) developed arguments indicating that diffusion limitation in the
CHO system cannot be occurring at any level of oxygen concentration including
concentrations near the apparent Michaelis-Menten value. However, this conclusion was cautiously qualified. The question of whether or not diffusion limitation can occur either from diffusion limitation in the surrounding medium or in the cell membrane
is a fundamental one of considerable importance. The present study approaches this
question in another way by asking how the barrier to oxygen transport presented by
the membrane compares with that of the surrounding medium.

Because oxygen is constantly consumed, and oxygen consumption reactions are localized inside the cell, it follows that there must be a gradient in oxygen concentration across the cell plasma membrane. The value of the oxygen concentration
difference is determined by the rate of oxygen consumption by the cell and the
oxygen permeability coefficient of the cell plasma membrane. Values of oxygen concentration differences reported in the literature vary widely: from small, lower than 1 ~M (Katz, Wittenberg, and Wittenberg, 1984; Wittenberg and Wittenberg,
1985), to large, a few micromolar to as great as 40 ~M (Tamura, Oshino, Chance,
and Silver, 1978; Jones and Kennedy, 1982; Jones, 1984; Morse and Swartz, 1985;
Glockner, Swartz, and Pals, 1989; see also review papers by Swartz and Pals, 1988;
and Wittenberg and Wittenberg, 1989).

Direct measurement of the oxygen concentration difference across a cell plasma membrane of oxygen-consuming cells is difficult because it must be done using oxygen-sensitive probes both outside and inside the cell (Swartz and Pals, 1988).

Another approach is to calculate the concentration difference on the basis of a measured oxygen consumption rate and the plasma membrane permeability coefficient
for oxygen.
Attempts have been made to determine the oxygen membrane permeability
coefficient using stop-flow rapid-mixing apparatus to create an oxygen gradient.
These studies have been criticized because the presence of a thick (~ 2 ~m) unmixed
water layer on the cell surface prevents immediate contact of the oxygenated solution
with the cell membrane (Coin and Olson, 1979; Huxley and Kutchai, 1981, 1983).
Other studies have been reported using the quenching of fluorescence or phosphorescence
probes by oxygen in red blood cell membranes, mitochondrial membranes,
or artificial membranes to obtain insight into oxygen permeability. However, these
methods give only the average oxygen diffusion-concentration product across the
membrane because of the large size of the probe and its uncertain localization in the
membrane (Fischkoff and Vanderkooi, 1975; Vanderkooi, Wright, and Erecinska,1990). Nuclear magnetic resonance techniques give a good spatial distribution of the oxygen diffusion-concentration product across the membrane, but the sensitivity of
the method is low (McDonald, Vanderkooi, and Oberholtzer, 1979).

In our previous papers we developed a methodology to estimate the membrane oxygen permeability coefficient from the profile of the oxygen diffusion-concentration product across the membrane (Subczynski, Hyde, and Kusumi, 1989, 1991a;
Subczynski and Markowska, 1992). Our method is based on a theory by Diamond and Katz (1974) and does not need creation of fast decaying oxygen gradients. Profiles of the oxygen diffusion-concentration product across the membrane can be obtained using different electron spin resonance (ESR) spin-label approaches: line-broadening
(Windrem and Plachy, 1980; Subczynski and Markowska, 1992), continuous wave
saturation (Subczynski and Hyde, 1981), or saturation recovery (Kusumi, Subczynski,
and Hyde, 1982; Subczynski et al., 1989, 1991a). The present study uses the long-pulse saturation-recovery ESR technique to investigate oxygen transport in the plasma membrane of CHO cells by observing the collision of molecular oxygen with
nitroxide radical spin labels placed at various distances from the membrane surface.
The value of the membrane oxygen permeability coefficient is determined from the profile of the oxygen diffusion-concentration product across the cell plasma membrane, and the oxygen concentration difference under physiological conditions is then calculated. A description of the mathematical procedure is also given.
It is commonly accepted that stearic acid spin labels reside in the lipid bilayer portion of the cell membrane (Kaplan, Canonico, and Caspary, 1973; Bales, Lesin, and Oppenheimer, 1977). However, the exact location of the spin labels within the
cell has long been a subject of discussion. Recently, Nettleton, Morse, Dobrucki, Swartz, and Dodd (1988), based on concentration-dependent broadening of the ESR spectra of the nitroxide 5-doxylstearic acid, concluded that the spin label must
distribute into most cellular membranes of intact cells. There is also a body of evidence that supports the hypothesis that either a substantial portion of the label resides in the plasma membrane or the ESR signal arises mainly from the cell
surface-located spin labels (Kaplan et al., 1973; Gaffney, 1975; Bales et al., 1977; Struve, Arneson, Chenevey, and Cartwright, 1977; Bales and Leon, 1978; Lai, Hopwood, and Swartz, 1980a; Dodd, Schor, and Rushton, 1982). Support for the
plasma membrane location of stearic acid spin labels comes from the facts that: (a)
spin labels are rapidly reduced at physiological conditions when entering the cell interior, (b) K~Fe(CN)6, which is impermeable to cell membranes, can quickly reoxidize nitroxides that are reduced by the cell, resulting in revived ESR signals that are identical to the original signals, and (c) order parameters measured within
minutes after label introduction are the same as those measured later. In this paper
we assume the plasma membrane location of stearic acid spin labels, but we will return to that subject in the conclusion.

MATERIALS AND METHODS
Chemicals
5-, 12-, and 16-doxylstearic acid spin labels (5-, 12-, and 16-SASL) were obtained from
Molecular Probes, Inc. (Eugene, OR) and 1-15N-l-oxylo-4-oxo-2,2,6,6,-tetramethylpiperidine - d16 (d-TEMPONE) was purchased from Merck Sharp and Dohme/Isotopes (Dorval, Quebec,
Canada).

Spin Labeling of CHO Cells

Labeling of CHO cells with fatty acid spin labels was carried out as described previously (Lai,
Hopwood, and Swartz, 1980b). Briefly, 3 x 106 cells from an asynchronous population grown in
spinner culture were centrifuged at 4°C at 300 g for 5 min and washed once with cold
phosphate-buffered saline. A spin-label solution was prepared by drying an ethanolic solution
of stearic acid spin label (SASL) under Nz in a 10-ml beaker, adding 2 ml of phosphatebuffered
saline, and stirring vigorously for 5 min at 23°C. 2 ml of cell suspension was gently
transferred to the spin-label solution and incubated at 23°C for 10 min with slow stirring (1
rev/s). The concentration of spin labels used in this study was in the range of 12-30 p.M. The
spin-labeling procedure did not affect cell viability as indicated by plating efficiency (70-80%)
or by the mitotic index determined using the hypotonic squash method. The spin-labeled CHO
cells were washed twice. The last rinse contained 5 mM KCN to inhibit cell respiration. The
thick cell suspension was transferred to a thin wall, 0.6-mm-i.d., gas-permeable capillary made
from the methylpentene polymer, TPX (Hyde and Subczynski, 1989). This plastic is permeable
to oxygen, nitrogen, and other gases and is substantially impermeable to water. The active
sample volume was ~ 2 p.l.

Saturation-Recovery ESR Measurements

The sample in a TPX capillary was positioned inside the loop-gap resonator (Froncisz and
Hyde, 1982) of the X-band saturation-recovery spectrometer. The spectrometer is based on
the design of Huisjen and Hyde (1974). Control of the concentration of oxygen in the sample
was achieved by equilibrating it with the same gas that was used for temperature control, i.e., a
mixture of nitrogen and air adjusted with flowmeters (model 7631H-604, Matheson Gas
Products, Secaucus, NJ). Spin-lattice relaxation times (Tl's) of spin labels were measured using
the long-pulse saturation-recovery technique (Kusumi et al., 1982; Subczynski et al., 1989,
1991a). With long and intense pulses, the spin system approaches a steady state in which the
populations of all energy levels tend to be equalized. After the saturating pulse is turned off,
the recovery of the spin system to Boltzmann equilibrium is observed with a weak observing
power. Typically, 1.25-2.5 × 104 decays per second were accumulated with 256 data points on
each decay. Total accumulation time was typically 10 or 20 rain. All measurements of TI were
made on the central line of the ESR spectrum for 5-, 12-, and 16-SASL, and d-TEMPONE at
10, 20, and 37°C. The values of Tj were obtained by fitting the tails of the saturation-recovery
signals to a single exponential.

Processing of Data

The rationale of the spin-label Tl method is that the molecular probe can be placed at a known
location in the membrane to observe local events and that the time scale of Tl (1-20 p.s) is in
the correct range to study many molecular processes (Yin and Hyde, 1987). One of these
processes is oxygen transport within the membrane. Bimolecular collisions of molecular oxygen (a fast relaxing paramagnetic species) and the nitroxide (a slow relaxing paramagnetic species)
induce spin exchange, which leads to faster effective spin-lattice relaxation of the nitroxide.
The bimolecular collision rate was evaluated by measuring the T~'s of the nitroxide as a
function of the partial pressure of oxygen. An oxygen transport parameter W(x) was introduced
as a convenient quantitative measure of the collision rate between the spin label and molecular
oxygen (Kusumi et al., 1982).

Downloaded from www.jgp.org on July 11, 2007

Address reprint requests to Dr. Witold K. Subczynski, National Biomedical ESR Center, 8701
Watertown Plank Road, Milwaukee, WI 53226.
J. GEN. PHYSIOL. © The Rockefeller University Press - 0022-1295/92/07/0069/19 $2.00
Volume 100 July 1992 69-87
69
Downloaded from www.jgp.org on July 11, 2007

Return To Top