INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

RECENT SPECTROPHOTOMETRIC MEASUREMENTS of fractional oxygen saturation of myoglobin (S_Mb) in near maximally exercising red skeletal muscles have indicated significant PO₂ gradients from capillary blood to the sarcolemma (6, 7, 10, 26). These PO₂ gradients appear so steep that PO₂ drops at a rate of ~7 Torr/µm even at moderate oxygen consumption (O₂) as the oxygen molecule travels a distance <2 µm from the erythrocyte to the cell surface (10). In contrast, cryospectrophotometric determination of S_Mb of the blood-perfused working heart in vivo (5) and ¹H NMR study of the isolated, perfused rat heart (14) reported much larger extracellular PO₂ gradients ranging from 15 to 30 Torr/µm even at normal resting O₂.

Because of the presence of surprisingly large extracellular PO₂ gradients, PO₂ at the sarcolemma may decrease to that at approximately half-saturation of myoglobin (P₅₀; 5.3 Torr at 37°C) in the beating heart (5). Hence, the intracellular PO₂ gradient from the sarcolemma to the mitochondrial membrane is a factor that critically determines oxygen transport to mitochondria and, therefore, oxidative phosphorylation. The magnitude of intracellular PO₂ gradients of the cardiomyocyte has been estimated from the simultaneous measurements of mitochondrial PO₂ and cytosolic PO₂ by using two separate intracellular oxygen probes (i.e., mitochondrial enzymes and myoglobin) in a suspension of isolated cardiomyocytes. Previous studies demonstrated quite shallow gradients between these compartments [<2 Torr at maximal O₂ (13, 24)], probably due to facilitation of oxygen diffusion by an intracellular oxygen carrier, myoglobin (25-27). If so, intracellular PO₂ of the normal beating heart would be substantially lower than that of the capillary blood but still high enough to sustain mitochondrial oxidative phosphorylation. These results seem to assign a minor role to the diffusional oxygen transport in the regulation of oxidative phosphorylation in the heart, at least at a moderate work rate.

Other than the PO₂ gradient between the cytosol and the mitochondria, gradual PO₂ changes perpendicular to the capillary direction may be present in the intracellular space (radial gradients). The magnitude and physiological significance of the radial intracellular PO₂ gradients have not been fully determined in the cardiac cell. We postulated that the radial PO₂ gradients might be a factor that limits oxidative phosphorylation in the cardiac cell at increased oxygen demand or during hypoxia (20). We undertook the present study to directly quantitate the intracellular radial PO₂ gradients in a single quiescent cardiomyocyte isolated from the rat when cellular oxygen demand was significantly increased.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Prior approval for the experiment was obtained from the Animal Research Committee, Yamagata University School of Medicine.

Cell isolation. Single cardiomyocytes were isolated from the ventricles of the pentobarbital sodium (50 mg/kg ip)-anesthetized adult Sprague-Dawley rat using the collagenase (type II, Worthington, Freehold, NJ) digestion method, as described previously (21). Isolated cardiomyocytes were suspended in a HEPES buffer solution containing (in mM) 150.0 NaCl, 3.8 KCl, 1.0 KH₂PO₄, 1.2 MgSO₄, 10.0 glucose, and 10.0 HEPES (pH adjusted to 7.35 at room temperature) supplemented with 0.1% BSA. Extracellular Ca²⁺ concentration was 1.0 mM except for the experiment using carbonyl cyanide m-chlorophenylhydrazone (CCCP), in which the buffer solution was Ca²⁺ free.

Measuring system. High spatial resolution spectrophotometry was conducted in a single individual cardiomyocyte using cytosolic myoglobin as an intrinsic oxygen probe. The previously described measuring system and image processing technique (20, 21) have been modified. The present measuring system consists of a light source, a computer-controlled monochrometer, and a microscope equipped with a digital charge-coupled device (CCD) camera. Light emitted from a direct current-powered 60-W metal halide lamp (LA-60Me, Hayashi Clock Works, Tokyo, Japan) was introduced to a computer-controlled monochrometer (SPG-100ST, Shimadzu, Kyoto, Japan). Monochromatic light at 410 nm (bandwidth 3 nm) was directed to a microscope (BH-2, Olympus, Tokyo, Japan), and the transmitted image of a single individual cell, via a ×40 objective lens (numerical aperture = 0.55; LWDCDPlan ×40, Olympus), was captured by a 10-bit digital CCD camera (C4742, Hamamatsu Photonics, Hamamatsu, Japan). Captured cell images were stored in a computer and subsequently processed as described below. We carefully selected the wavelength for the spectrophotometry to fulfill the following requirements: 1) the wavelength corresponds to the Soret band of oxymyoglobin, and 2) the spectrum of the light source is relatively flat around the selected wavelength. The spatial resolution of the final output image on the computer monitor was 0.14 µm/pixel. Image analysis software (IPLab, Signal Analytics, Vienna, VA) controlled the monochrometer and CCD camera.

Experimental protocol. A 15-µl cell suspension containing ~1,500 cells was placed on the poly-L-lysine (Sigma, St. Louis, MO)-coated glass slide of an airtight measuring cuvette that provided gas inlet and outlet ports. The measuring cuvette was then transferred to the stage of the microscope and was connected to the outlet port of a computer-controlled gas blender consisting of mass-flow controllers (SEC-320, STEC, Kyoto, Japan). The gas blender supplied a mixed gas of desired oxygen concentration via a humidifier at 2 ml/min. We continuously monitored PO₂ at the gas outlet port of the measuring cuvette using a conventional oxygen electrode (model 17026, Instrumentation Laboratory, Lexington, MA).

Spectrophotometric determination of S_Mb. We conducted digital image processing on the captured cell images to determine the light absorption of the cell with a subcellular spatial resolution. The basic assumption was that changes in light absorption associated with changes in oxygenation level of the cell are exclusively attributable to changes in light absorption of cytosolic myoglobin and thus report cytosolic oxygenation. Before the cell image was digitized, the analog circuit in the video amplifier (C4742) subtracted a part of the light intensity signal (voltage corresponding to the transmitted light intensity that is not affected by changes in myoglobin light absorption) and amplified the remaining part of the signal. This procedure is mathematically represented as

(1)

where X_ij and Y_ij denote the values for a pixel [coordinate (i, j)] of the transmitted cell image and the final digitized cell image, respectively, and  and  are gain and offset of the video amplifier, respectively. According to the Beer-Lambert law, light absorption by intracellular pigments can be represented as

(2)

where I_ij is the intensity of incident light;  and C_ij denote the molar absorption coefficient and the myoglobin concentration at pixel (i, j), respectively; superscripts oxy and deoxy refer to oxygenated and deoxygenated myoglobin, respectively; M_ij represents light absorption by intracellular pigments other than myoglobin; and L_ij is the length of the light path. By taking advantage of the fact that light absorption in a single cardiomyocyte is extremely small, Eq. 2 can be linearized as

(3)

where k is a constant. Finally, the following calculation conducted for each pixel converts the transmitted light intensity to S_Mb at coordinate (i, j)

(4)

Data analysis. The position of each pixel relative to the cell must be identical for all three separately obtained images. To avoid error arising from slight movement of the cell during image acquisitions, we aligned the three cell images, if necessary, before calculation of S_Mb as follows. First, we chose five to seven marker points in a small region of interest (ROI) that were peculiar with respect to light absorption, and the absolute positions of these points were recorded. These marker points appeared to correspond to locations of mitochondria according to the results of rhodamine-123 staining (data not shown). We then checked whether these marker points were in fact found at the same absolute coordinates in the remaining two images. If not, the deflections of the markers from the expected coordinates were calculated. Finally, the affine transform (linear mapping of an image including shifts and rotation) was conducted over the remaining two cell images so that the square errors of the deflections could be minimized. Also, the images were low-pass filtered three times.

Calibration. We conducted a calibration that relates local S_Mb to local PO₂. We added 2 mM NaCN to the suspension medium and conducted S_Mb measurements for superfusion gas containing either 0.25, 0.51, 0.96, 2.09, or 3.14% oxygen. We assumed that NaCN abolishes the consumption of oxygen by the cell, thereby abolishing PO₂ gradients from the extracellular medium to the intracellular space. Hence, intracellular PO₂ is in equilibrium with gas PO₂. The relationship between PO₂ of the superfusion gas and measured S_Mb was fitted to the Hill equation.

O₂ measurement. Because the magnitude of intracellular PO₂ gradients would be proportional to flux of oxygen to the cell (10), we used 1 µM CCCP (an uncoupler of oxidative phosphorylation) to amplify the intracellular PO₂ gradients. Therefore, we needed to determine O₂ of the cell in the presence of 1 µM CCCP. Five milliliters of the cell suspension were placed in the airtight measuring cuvette, in which an oxygen electrode (model 17026, Instrumentation Laboratory) was inserted. The cell suspension was vigorously stirred using a magnetic stirrer. The time-dependent decrease in PO₂ of the suspension medium was recorded, and the rate of fall of PO₂ (PO₂/t in Torr/min) was converted to the rate of oxygen consumption (nmol O₂ · min¹ · 10⁶ cells¹) using the following equation where N is the number of cells per milliliter and _w is the solubility of oxygen in water (1.62 µmol · l¹ · Torr¹ at 27°C).

	RESULTS

Top Abstract Introduction Methods Results Discussion References

We attempted a visualization of intracellular oxygen with a subcellular spatial resolution in the presence of an uncoupler of oxidative phosphorylation, 1 µM CCCP. When the cell suspension was superfused with 2.09 (15.2 Torr) or 3.14% (22.8 Torr) O₂ gas, intracellular S_Mb averaged over the cell was ~0.4-0.7 (corresponding to ~1.9-8.7 Torr; see the results of calibration below) (Fig. 1, solid curve). These results indicate the presence of large PO₂ gradients in the extracellular medium, presumably resulting from the absence of the specific oxygen carrier myoglobin and the unstirred layer surrounding the cell surface (18). It should be noted that intracellular oxygenation in this condition appears comparable to the volume-averaged S_Mb reported in the working cardiac cell in vivo (3, 5).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Profiles of radial changes in myoglobin oxygen saturation (SMb) and intracellular PO2 in a single individual cardiomyocyte. Changes in SMb along short axis of cell (solid curve) were estimated from curve fitting of measured points () to a hyperbolic function. Estimated SMb was subsequently converted to intracellular PO2 (broken curve) using the relationship indicated in Fig. 6. The cell was treated with 1 µM carbonyl cyanide m-chlorophenylhydrazone (CCCP). The width of this particular cell was 23.9 µm. PO2 of superfusing gases were 15.2 and 22.8 Torr in A and B, respectively.

Figure 2 shows the representative data demonstrating the visualization of intracellular oxygenation in a single individual cardiomyocyte. For PO₂ of the superfusion gas of 15.2 Torr, significant gradients of S_Mb from the sarcolemma toward the center of the cell were demonstrated (indicated in pseudo colors). To quantitatively analyze these radial heterogeneities of S_Mb, we calculated S_Mb in small rectangular ROIs within a cell (Fig. 3A). Histograms of the S_Mb value in these ROIs were generated (Fig. 4A) and fitted to normal distributions, and the mean S_Mb was determined. Local S_Mb near the center of the cell (Fig. 4A, b) was significantly lower than those calculated near the sarcolemma (Fig. 4A, a and c). We assumed that these variations in the intracellular oxygenation reflect the intracellular PO₂ gradients as oxygen molecules diffuse from the surface into the core of the cell. We then tested this hypothesis by abolishing the oxygen flux using 2 mM NaCN. Because the suppression of oxygen flux by cyanide also eliminated extracellular PO₂ gradients, PO₂ of the superfusion gas was reduced until the average S_Mb was ~0.5. As shown in Figs. 3B and 4B, heterogeneities of intracellular S_Mb were eliminated after application of NaCN. Figure 5 summarizes radial changes in S_Mb in the presence of 2 mM NaCN. Data were collected from three and six individual cardiomyocytes exposed to extracellular PO₂ of 3.7 and 15.0 Torr, respectively. No significant change in S_Mb was found along the short axis of the cell.

View larger version (84K): [in this window] [in a new window]   Fig. 2.   Representative data demonstrating reconstruction of SMb. Transmitted images of a single cardiomyocyte (left) were converted to an SMb image (right) using Eq. 4. SMb is represented in pseudo colors. The cell was incubated with 1 µM CCCP to augment intracellular PO2 gradients. PO2 of superfusion gas was 15.2 Torr. Cell boundaries were manually traced.

View larger version (96K): [in this window] [in a new window]   Fig. 3.   Representative data demonstrating effect of cellular oxygen consumption on intracellular oxygenation. SMb is indicated in pseudo colors. A indicates SMb of a cell treated with 1 µM CCCP (same cell as in Fig. 2), whereas B indicates SMb of a cell treated with 2 mM NaCN to abolish oxygen consumption. X-X ' and Y-Y ' indicate cell boundaries, which were manually traced; a, b, and c are regions of interest (ROIs) within each cell.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   SMb histograms of ROIs indicated in Fig. 3 (a, b, and c correspond to a, b, and c in Fig. 3). A and B show histograms for CCCP- and NaCN-treated cells, respectively. Histograms are fitted to a normal distribution, and the mean is indicated in each panel.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Mean SMb values calculated from subcellular ROIs in 2 mM NaCN-treated cells. PO2 of superfusing gas is 3.7 Torr in A and 15.0 Torr in B. No correlation was found between SMb and distance from sarcolemma. Data were collected from 3 and 6 individual cells, respectively.

Figure 6 shows the relationship between extracellular PO₂ and S_Mb measured in the presence of 2 mM NaCN. The P₅₀ determined by simple linear regression was 3.1 Torr (Fig. 6, line A), although the Hill parameter was not equal to 1. Because of the absence of heme-heme interaction in myoglobin, the Hill parameter is expected to be unity. Hence, we recalculated the regression line while fixing the Hill parameter to 1 in order to see the effect of errors in the measurement. We found a small decrement of the regression coefficient from 0.899 to 0.766 and a slight increase in the P₅₀ from 3.1 to 3.7 Torr (Fig. 6, line B).

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Relationship between PO2 of superfusion gas and SMb in a single cardiomyocyte treated with 2 mM NaCN. Data were represented using the Hill equation. Line A was obtained from simple linear regression of data (y = 0.813x  0.395, r = 0.899), whereas the Hill coefficient was fixed to a value of 1 in line B (y = x  0.571, r = 0.766).

In all cells in which the image processing was successful, a significant drop of S_Mb was demonstrated near the center of the cell (Figs. 1 and 7). The change in S_Mb was then approximated by a hyperbolic curve (Figs. 1 and 7, solid curves) and subsequently converted to PO₂ using the calibration data described above. We found quite steep gradients of the intracellular PO₂ in the vicinity of the sarcolemma, whereas PO₂ near the center of the cell was flat (Fig. 7). The lowest S_Mb estimated by this technique was 0.36 ± 0.07 (n = 8), which was unrelated to the size of the cell.

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Individual results of measurements of intracellular radial gradients of SMb () and estimation of intracellular PO2 (dotted curves) in which image processing was successfully carried out. Cells were incubated with 1 µM CCCP. PO2 of the superfusion gas was 15.2 Torr. Vertical dotted lines indicate width of each cell. Nos. above graphs represent individual experiments.

O₂ of the quiescent cardiomyocytes at 27°C was 26 ± 9 nmol O₂ · min¹ · 10⁶ cells¹ (n = 5) in the presence of 1 mM extracellular Ca²⁺. Incubation with 1 µM CCCP increased O₂ to 225 ± 38 nmol O₂ · min¹ · 10⁶ cells¹ (n = 7) in the absence of extracellular Ca²⁺. These values (after temperature is compensated for) are in good agreement with the O₂ of isolated rat cardiomyocytes reported by Wittenberg and Robinson (23). Thus the CCCP-treated cells in the present study represent metabolic oxygen demand of the beating rat heart at increasing (not maximal) work output (23).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In the previous study (21), we showed that average cytosolic PO₂ of a single quiescent cardiomyocyte can be quantitated from the measurement of fractional oxygen binding to myoglobin using three-wavelength spectrophotometry. Because light absorption of a single cell is extremely small, we needed to carry out the spectrophotometry at the Soret band of myoglobin, where the myoglobin light absorption is maximum. This requirement, however, elicited a problem in that a significant part of the measured light absorption contains that of cytochromes because light absorption peaks of myoglobin and cytochromes significantly overlap in the Soret band. Although we have demonstrated that the estimation of average light absorption may not be seriously affected by cytochromes (see Fig. 5 of Ref. 21), the calculation of myoglobin light absorption at subcellular resolution was subject to potential errors. Moreover, the previous method did not allow us direct estimation of S_Mb without knowledge of molar extinction coefficients of oxy- and deoxymyoglobin in vivo. In contrast, the present technique with the use of just one wavelength enables much more accurate estimation of S_Mb. It is theoretically correct that light absorptions of cytochromes, even if they are large in magnitude, can be completely subtracted in the calculation of absolute S_Mb values (Eq. 4). Even values for molar extinction coefficients are not required. Another benefit of using a single wavelength is that the effect of light scattering, which strongly depends on the wavelength, can be minimized. Thus precise high spatial resolution measurement of S_Mb was possible in the present study.

Prerequisites for the present single-wavelength spectrophotometry were three cell images taken at three different myoglobin oxygenation states with constant cytochrome light absorptions. To fulfill these requirements, the cell was treated with NaCN or CCCP to fix the cytochromes to the respective reduced states without regard to the oxygen level. Practical issues that we encountered were the slight movements of the cell during measurement at three different PO₂ values. Correction for the movement with the use of a mathematical method (affine transform) is very intricate and may not always be effective, particularly when the deflection of the cell is >5 pixels. Another methodological problem is that the current technique does not measure S_Mb of various intracellular points on the same cross-sectional plane of a myocyte. Instead, the calculated S_Mb is the volume average along the light path including S_Mb of the well-oxygenated sarcolemmal portion, the less oxygenated cell core, and another well-oxygenated sarcolemmal portion. Thus radial changes in S_Mb can in fact be detected, but the calculated desaturation at the center of the cell should be underestimated.

Radial changes in intracellular PO₂ have been extensively studied theoretically and experimentally. Recent model analyses basically utilized the classic Krogh model of oxygen diffusion that was supplemented with the model of myoglobin-facilitated diffusion of oxygen (4, 8, 9, 15, 17). These analyses predicted, at least qualitatively, quite similar radial intracellular PO₂ profiles. The study by Groebe (8) provides us with the most up-to-date knowledge about the model analysis. He predicted quite steep PO₂ gradient at ~10 Torr/µm in the extracellular carrier-free region in red skeletal muscle at nearly maximum performance. In contrast, the rate of fall of PO₂ quickly decreases as oxygen diffuses from the sarcolemma into the cell. Facilitation of oxygen diffusion by myoglobin increases as PO₂ decreases where maximum facilitation occurs at a PO₂ of ~0. Therefore, as oxygen migrates toward the cell core, a progressive increase in the fraction of deoxymyoglobin resulting from local oxygen consumption maintains oxygen flux with a smaller PO₂ gradient. In addition, oxygen flux density (oxygen flux per unit area) also decreases as oxygen molecules diffuse into the cell. As a net result, PO₂ gradients become remarkably shallow near the center of the cell. Finally, the intracellular PO₂ a few micrometers away from the sarcolemma becomes quite low but relatively uniform in the rest of intracellular space.

Gayeski and Honig (7) conducted measurements of S_Mb by cryospectrophotometry in dog gracilis muscles in situ. Using a sampling volume of ~30 µm³, they mapped S_Mb profiles of a cross section of single muscle cell during twitch contraction at maximal O₂. They found very low but relatively uniform S_Mb within a cell that could be accounted for by the mathematical model of oxygen diffusion. Gayeski and Honig (5) subsequently extended their in vivo cryospectrophotometry in cardiac muscle cells of various animals. They again demonstrated quite low S_Mb (~P₅₀ of myoglobin) in the beating heart at normal work rate. Their high spatial resolution measurement of S_Mb showed negligible radial intracellular gradients from within 2 µm of the sarcolemma to the center of a 16-µm-diameter cell. However, these investigators did not address the steep PO₂ gradients near sarcolemma that are usually predicted by model studies (4, 8, 9).

In the present study, we have directly visualized the profile of intracellular PO₂ with a radial spatial resolution of 4 µm in a single cardiomyocyte. The use of a single, isolated cardiomyocyte gave us an opportunity to specifically examine diffusional oxygen transport within a cell in a situation free from the confounding effects of capillary oxygen transport. We adjusted the extracellular PO₂ so that cytosolic myoglobin was partially deoxygenated as in in vivo cardiomyocytes (5). When O₂ of the cell was increased approximately eightfold, radial gradients of S_Mb, with the nadir located at the center of the cell, were clearly visualized (Fig. 2). When the measured S_Mb was converted to intracellular PO₂, we found a quite steep PO₂ drop near the sarcolemma and relatively constant PO₂ around the center of the cell (Figs. 1 and 7). These results are in good agreement with the previous theoretical studies.

We carried out these measurements at 27°C. At 37°C, on the other hand, the effective oxygen conductivity would be only slightly (~10%)¹ higher (8), whereas O₂ would increase by ~1.8 times (19). Because the PO₂ gradient is determined by dividing the oxygen flux density by the effective oxygen conductivity, the magnitude of intracellular PO₂ gradients in vivo (37°C) would be much larger than those demonstrated at 27°C. Therefore, the dependency of the observed gradients on oxygen flux (Fig. 4) suggests the importance of interplay between intracellular oxygen diffusion resistance and oxygen flux as one of the determinants of oxygen transport to mitochondria in vivo.

As demonstrated in the present study, an increase in cellular respiration may produce large PO₂ gradients, particularly in the vicinity of the sarcolemma. It is presumable that these large gradients of PO₂ tend to produce regions away from the sarcolemma where oxidative phosphorylation is compromised due to the relative deficiency of diffusional oxygen supply (hypoxic core) (1). The presence of a hypoxic core would be prominent in relatively large and metabolically active cells such as cardiomyocytes. In addition, the significance of a hypoxic core may be more important in some pathophysiological conditions such as cardiac hypertrophy. At average S_Mb comparable to that in vivo, we demonstrated an almost flat PO₂ profile near the core of a CCCP-treated single cardiomyocyte (Figs. 1 and 7). Because the rate of fall of PO₂ along the diffusion path is a function of local oxygen consumption (not strictly in proportion to local O₂ due to myoglobin-facilitated oxygen diffusion), the observed PO₂ profile near the center of the cell seems to suggest the presence of hypoxic depression of O₂ (i.e., a hypoxic core) in a single individual cardiomyocyte.

Previous animal and model studies have addressed the major importance of myoglobin-facilitated diffusion in maintaining intracellular space at low but stable oxygenation (10, 26). Among these studies, the model study conducted by Groebe (8) isolated the effect of a hypoxic core from the effect of myoglobin-facilitated diffusion on the regulation of intracellular PO₂. He compared intracellular PO₂ values calculated from the two different models. In one model, O₂ was constant irrespective of PO₂, whereas in the other model, local O₂ was changed as a function of the local PO₂, assuming Michaelis-Menten kinetics. Surprisingly, two types of local O₂ control schema, PO₂-independent O₂ versus PO₂-dependent O₂, showed virtually identical radial PO₂ profiles. This was because PO₂ and PO₂ gradients were already extremely low (<0.5 Torr, critical mitochondrial PO₂) at the region where O₂ started to decrease (>15 µm into the fiber) and O₂-dependent changes in PO₂, if any, were too small to be detected. These results, however, may not exclude the physiological importance of the hypoxic core. The mathematical model by Groebe does not include the PO₂ gradients between the cytosol and the mitochondrial membranes. Although these gradients are usually believed to be very small, they may increase up to 2 Torr in cardiomyocytes when mitochondrial respiration is maximally stimulated (13, 24). Presumably, the effect of hypoxic depression of O₂ on reducing PO₂ gradient (20) could be demonstrated at higher cytosolic PO₂ if cytosol-mitochondrial PO₂ gradients were considered. Furthermore, mitochondrial O₂ becomes dependent on oxygen supply at an extremely low PO₂ range (Michaelis-Menten constant = ~0.05-0.1 Torr, see Ref. 11). Hence, a reduction of PO₂ gradient, albeit very small in absolute magnitude, caused by the hypoxic reduction of local O₂ would significantly affect local oxidative phosphorylation.

Cytosolic myoglobin facilitates intracellular oxygen diffusion approximately sixfold in red skeletal muscles (~4-fold in cardiomyocytes, assuming a myoglobin content of cardiac tissue of 0.2 mM) at a PO₂ of ~0 (8). In the cardiac cell, oxygen molecules must diffuse over a distance ~5-15 times longer in the intracellular space compared with that in the extracellular carrier-free region, until it reaches the center of the cell. Whether the fourfold increase in oxygen diffusion capability and a gradual reduction of oxygen flux in the intracellular space can fully account for the remarkably small intracellular PO₂ gradients or whether depression of local O₂ within a cell needs to be considered remains to be explored.

In contrast to the previous studies in red skeletal muscles at nearly maximum O₂, we demonstrated that even a moderate increase in the cellular O₂ results in a significant drop of intracellular PO₂ near the sarcolemma. Similarly, Gayeski and Honig (5) suggested the presence of large PO₂ gradients (presumably located in the extracellular carrier-free region) in the heart even at the resting metabolic oxygen demand. Current mathematical models of oxygen transport in skeletal muscles do not appear to be fully compatible with the data obtained from the cardiac cell (8). In addition, several studies demonstrated that intracellular PO₂ is remarkably stable against changes in arterial oxygenation or cardiac work (5, 12). These results seem to suggest the presence of an active regulatory mechanism of intracellular oxygen transport in the heart other than simple diffusion. Such regulation may possibly be coupled with local cellular oxygen demand/metabolism (2, 22).

In conclusion, we have demonstrated in a single individual cardiomyocyte that large radial PO₂ gradients may be generated within an actively respiring cell. As a result, oxidative metabolism may be partially suppressed at the core of the cardiac cell even if cellular oxygenation averaged over the cell is adequate.