Myocyte Adaptation to Chronic Hypoxia and Development of Tolerance to Subsequent Acute Severe Hypoxia
Abstract Studies in animal models and humans suggest that myocardium may adapt to chronic or intermittent prolonged episodes of reduced coronary perfusion. Stable maintenance of partial flow reduction is difficult to achieve in experimental models; thus, in vitro cellular models may be useful for establishing the mechanisms of adaptation. Since moderate hypoxia is likely to be an important component of the low-flow state, isolated adult rat cardiac myocytes were exposed to 1% O2 for 48 hours to study chronic hypoxic adaptation. Hypoxic culture did not reduce cell viability relative to normoxic controls but did enhance glucose utilization and lactate production, which is consistent with an anaerobic pattern of metabolism. Lactate production remained transiently increased after restoration of normal O2 tension. Myocyte contractility was reduced (video-edge analysis), as was the amplitude of the intracellular Ca2+ transient (indo 1 fluorescence) in hypoxic cells. Relaxation was slowed and was accompanied by a slowed decay of the Ca2+ transient. These changes were not due to alterations in the action potential. Tolerance to subsequent acute severe hypoxia occurred in cells cultured in 1% O2 and was manifested as a delay in the time to full ATP-depletion rigor contracture during severe hypoxia and enhanced morphological recovery of myocytes at reoxygenation. The latter was still seen after normalization of the data for the prolonged time to rigor, suggesting a multifactorial basis for tolerance. An intervening period of normoxic exposure before subsequent acute severe hypoxia did not result in loss of tolerance but rather increased the delay to subsequent ATP depletion rigor. Cellular glycogen was preserved during chronic hypoxic exposure and increased after the restoration of normal O2 tension. As mitochondrial cytochromes should be fully oxygenated at levels well below 1% O2, hypoxic adaptation may be mediated by a low-affinity O2-sensing process. Thus, adaptations that occur during prolonged periods of moderate hypoxia are proposed to poise the myocyte in a better position to tolerate impending episodes of severe O2 deprivation.
Experiments in animal models suggest that myocardium is capable of adapting to chronic low-flow states. A similar process may occur in humans in the setting of atherosclerotic coronary artery disease. Early studies described a reversible depression of contractility occurring during mild-to-moderate reductions in coronary perfusion without associated myocardial necrosis or major changes in high-energy phosphates.1 2 This phenomenon was termed “myocardial hibernation.” Although alterations in metabolism and contractile function have been demonstrated and proposed to allow myocardial survival in the face of reduced O2 supply, the cellular basis of adaptation and the signaling pathways involved in the process have yet to be defined. A major impediment to studying the phenomenon has been the difficulty of creating a stable chronic low-flow state in an experimental animal model. Because chronic hypoxia may represent an important component of these states, we recently developed a cellular model to study hypoxic adaptive responses and their basis by using isolated adult rat ventricular myocytes maintained for 48 hours in serum-free medium equilibrated with 1% O2. We reasoned that myocytes would survive the moderately hypoxic state and perhaps show adaptations consistent with those reported in low-flow states. We further hypothesized that the adaptive process would make cardiac myocytes more tolerant of subsequent acute severe hypoxic insults. Prior studies performed in our laboratory3 showed that brief exposure of myocytes to hypoxia did not result in protection during a later severe hypoxic episode. This finding is to be contrasted with studies in the intact heart in which marked reductions in ischemic injury were seen with brief “preconditioning” pulses of ischemia or hypoxia.4 5 One recent study demonstrated that chronic hypoxia increased tolerance to myocardial ischemia in a manner dissimilar to preconditioning6 and showed that preconditioning and “adaptation” were additively protective, suggesting that their effects are likely to be mediated by different mechanisms. In the present study, we examine the effects of chronic hypoxia on cell viability, metabolism, contractility, and intracellular Ca2+ regulation. We also test the hypothesis that chronic hypoxia induces tolerance to later acute severe hypoxic insults. Since prior studies showed that O2 consumption became O2-limited only below 0.1% O2 in isolated rat cardiac myocytes7 and below 0.3% in isolated rat hearts,8 physiological or metabolic changes observed in myocytes at 1% O2 would be unlikely to result from limited oxygenation of mitochondrial cytochromes.
Materials and Methods
Myocyte Isolation and Culture
Adult (2- to 4-month-old) Sprague-Dawley rats were euthanized with pentobarbital sodium in accordance with institutional guidelines, and the heart was rapidly excised and attached to a gravity-based sterile perfusion apparatus. The heart was perfused at 37°C with Ca2+-free Krebs-Ringer bicarbonate buffer containing 0.1% collagenase B (Boehringer-Mannheim) and 0.004% protease type XIV (Sigma Chemical Co). The perfusate also contained (mmol/L) taurine 60, glutamic acid 8, and carnitine 2, with added 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL amphotericin, according to a modification of Kirshenbaum et al.9 After roughly 20 minutes, the heart was removed, and the left ventricle was dissected free and minced. Myocytes were then mechanically dissociated and resuspended in buffers of gradually increasing [Ca2+]. To remove dead myocytes and residual contaminating cell types, the myocyte suspension was centrifuged through a discontinuous Percoll gradient, usually resulting in roughly 80% to 90% rod-shaped cells (some residual dead myocytes). Myocytes were cultured relatively sparsely at a density of 200 to 300 μg protein/mL (50 000 to 80 000 cells/mL) in medium 199 (containing 26 mmol/L sodium bicarbonate and 25 mmol/L HEPES as buffers and 5.6 mmol/L glucose, 0.6 mmol/L acetate, and amino acids as metabolic substrates) supplemented with 5 mmol/L carnitine, 5 mmol/L taurine, 5 mmol/L creatine, 100 U/mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL amphotericin. No lactate was present in the medium. Plastic culture dishes (either 2.5- or 10-cm diameter) were first coated with laminin (20 μg/mL culture medium), and cells were added after 1 hour. Laminin coating of dishes facilitated the attachment of cells and permitted the “wetting” of dishes, allowing us to use small volumes of culture medium (we were able to maintain a minimal height of 0.64 mm) to minimize O2 diffusion gradients. For experiments in which cells were later studied by microscopy, four to six 0.5-mm-diameter glass coverslips were added to each plastic dish before laminin coating. Cultures were then equilibrated with either 5% CO2/95% air (normoxia) or 1% O2/5% CO2/94% N2 (moderate hypoxia) for 48 hours at 37°C and maintained in the absence of electrical stimulation. Because cells were cultured sparsely and the height of the medium was kept to a minimum, virtually no medium surface–to–cell surface O2 gradient was created (see “Discussion”).
Determination of Culture Medium Glucose and Lactate
Glucose was measured in culture medium by coupled enzymatic reactions catalyzed by hexokinase and glucose-6-phosphate dehydrogenase.10 During these reactions, added NAD was reduced to NADH in an amount equimolar to glucose. The increase in absorbance at 340 nm was monitored and is proportional to glucose concentration. Lactate was measured in culture medium by an assay using lactate dehydrogenase to enzymatically convert lactate to pyruvate with equimolar conversion of added NAD to NADH.11 Formed pyruvate was trapped with hydrazine. The increase in absorbance at 340 nm was monitored and is proportional to lactate concentration. Cellular lactate release was quantified by normalizing data to total protein using a Bio-Rad DC protein assay (based on the method of Lowry et al12 ).
Assay of Glycogen Content
Myocytes were washed twice with ice-cold glucose-free HEPES buffer. Glycogen was acid-extracted from myocyte cultures and digested with amyloglucosidase as outlined by Haworth et al.13 The resultant glucose was then measured as detailed above. Glycogen content is expressed as equivalents of glucose per milligram of total protein.
Assessment of Myocyte Contractility and [Ca2+]i
Glass coverslips with attached myocytes were removed from culture and placed in a small chamber on the stage of a customized inverted microscope (Nikon Diaphot). The cells were field-stimulated at 0.2 Hz, 75 V, and 5-millisecond pulse width (Grass Instruments) at 37°C. Contractility was assessed independently (non–dye-loaded cells) or simultaneously with [Ca2+]i (in myocytes loaded with indo 1). When indo 1 was used, myocytes were loaded by exposure to the membrane-permeant form, indo 1-AM (25 μmol/L) for 10 minutes. The cell was visualized by bright-field illumination (long-pass filter at 695 nm). The image was directed to a TV camera and viewed on a monitor. Cell length was assessed continuously by a Crescent Electronics video-edge detector. Fluorescence excitation was from a xenon lamp at 350±5 nm. Emitted light was directed to two photomultiplier tubes selecting wavelengths of 405±17 nm (Ca2+-bound indo) and 495±10 nm (Ca2+-free indo). The signal was passed to two amplifiers/discriminators and then to a custom-built analog circuit with antialiasing filters. Data were digitized at 200 Hz. Custom software was used for analysis of length and indo 1 data. After subtraction of cellular autofluorescence, the ratio of fluorescence at the two wavelengths was calculated and served as an index of [Ca2+]i. Because the AM ester of indo 1 is taken up almost equally into the cytosol and mitochondria, we did not attempt to calibrate the signal. Therefore, data are reported as the raw ratio of fluorescence in the two channels.
Measurement of Action Potentials
Membrane potential was assessed by use of an Axopatch-1C amplifier and a 1/100 CV-3 head stage. Experimental control, data acquisition, and data analysis were accomplished by use of the software package pClamp 6.0 with the Digidata 1200 acquisition system. Action potentials were assessed using 5- to 10-MΩ pipettes pulled from 1.5-mm borosilicate glass. They were filled with (mmol/L) KCl 120, NaCl 10, MgCl2 1, and HEPES 20 (pH 7.2 with KOH). Recordings were obtained in current-clamp mode.
Acute Severe Hypoxic Exposure
Individual cells (stimulated at 0.2 Hz) were studied in a specially designed chamber described in detail by Stern et al.14 Ultrahigh-purity argon entered the chamber at a precisely controlled flow rate, creating a laminar barrier to atmospheric O2. Myocytes were studied in glucose-free buffer containing (mmol/L) NaCl 144, KCl 5, MgSO4 1.2, and HEPES 10, with 1 mmol/L Ca2+ and 0.5 mmol/L octanoate as respiratory substrate. The chamber allowed rapid changes in O2 from atmospheric to 0.003%, while permitting continuous field stimulation. The profoundly low O2 tensions achieved in this chamber were necessary to elicit acute cellular effects, since mitochondria only become O2-limited below 0.1% O2.7 An early study14 has shown that during exposure to glucose-free severe hypoxia, myocytes remain excitable with preserved contractile amplitude for a variable period of time (10 to 50 minutes) and then develop failure of contraction over 1 minute as action potential (AP) narrowing occurs because of the activation of ATP-sensitive K+ channels. Roughly 5 minutes later, cells develop ATP-depletion rigor contracture. The variable delay to contractile failure and rigor contracture is likely due to differences in intracellular glycogen stores, which serve as the sole fuel source for anaerobic metabolism in these protocols. Studies conducted by Haworth et al13 in myocyte suspensions confirmed that ATP depletion occurs when glycogen is exhausted and that the hallmark of this event at the cellular level is the development of rigor contracture. A more recent study15 has demonstrated, at the single-cell level, that ATP depletion occurs abruptly as contraction fails and rigor develops. Myocytes shorten to two thirds of their original length but retain a square shape as they undergo rigor contracture.16 Remarkably, some cells partially relengthen and resume contraction on electrical stimulation (recovery) if reoxygenated promptly after full ATP depletion. On the other hand, cells that are reoxygenated >20 to 30 minutes after rigor contracture invariably hypercontract to an irreversibly damaged state. The probability of cell survival at reoxygenation declines not with total hypoxic exposure but with the time of continued hypoxia after full ATP depletion. After rigor onset, myocytes become progressively Ca2+-loaded, and their fate at reoxygenation has been correlated with the level of [Ca2+]i achieved just before reoxygenation.17
Data are presented as mean±SEM. Comparisons were made by Student’s t test or the Mann-Whitney rank sum test for nonparametric data. Differences in the four groups exposed to acute severe hypoxia were assessed by Kruskal-Wallis ANOVA on ranks, with multiple comparisons made by Student-Newman-Keuls testing. Differences in electrophysiological data between the three groups were assessed by factorial ANOVA with Fisher’s least significant difference post hoc analysis. Values of P<.05 are considered significant.
Fig 1⇓ demonstrates the morphological appearance of adult rat myocytes maintained in serum-free culture for 48 hours in 5% CO2/95% air (normoxia) and in 1% O2/5% CO2/94% N2 (moderate hypoxia). The gross appearance of these cells is identical, with myocytes maintaining their rod shape and clear sarcomere pattern. When the cells are initially placed in culture, roughly 80% to 90% of them are rod-shaped, and at 48 hours viability is unaffected by hypoxia (65±3% in normoxia, n=15 cultures; 64±3% in hypoxia, n=11 cultures; P=NS).
Metabolism in Hypoxic Cultures
In 15 paired cultures (from 15 rats), culture medium glucose concentration was reduced from an initial 5.6 to 4.68±0.12 mmol/L in the normoxic cells and to 3.52±0.30 mmol/L in the hypoxic cells (P=.002). Lactate accumulated in the cultures, and at 48 hours lactate concentration was roughly four times greater in the hypoxic group (n=20 cultures per group), suggesting a shift to an anaerobic pattern of metabolism. When lactate released into the medium was normalized for myocyte protein content (micromoles lactate released per milligram protein), the values for the hypoxic group remained four times greater than for the normoxic group. Despite lactate accumulation, pH of the medium was not affected, probably because of significant buffering due to the presence of HEPES and bicarbonate. These data are summarized in Fig 2⇓. The relationship between medium glucose and lactate for both the chronically normoxic and hypoxic myocyte cultures is shown in Fig 3⇓. Each data point represents a single culture. In the normoxic group, glucose concentration varied over a narrow range of 3.94 to 5.43 mmol/L and lactate concentration remained below 1 mmol/L in all but one case. The hypoxic group was strikingly different, with glucose concentration ranging widely from 1.12 to 4.77 mmol/L and lactate concentration ranging from 0.58 to 7.51 mmol/L. A linear relationship existed between medium lactate and glucose concentrations, with a rise in lactate concentration of 1.83 mmol/L per 1 mmol/L fall in glucose concentration (r=.987). This increase is close to the stoichiometric conversion of 1 glucose to 2 lactate by anaerobic glycolysis. Six additional cultures exposed to 1% O2 for 48 hours were washed, and fresh culture medium was added. Cultures were then placed in the normoxic incubator for 3 hours, and then medium glucose and lactate concentrations were assessed to establish whether marked lactate production and release continued after restoration of normoxia. The resultant medium lactate concentration was 0.41±0.17 mmol/L, and glucose concentration was 4.86±0.27 mmol/L. Lactate levels were 0.10±0.06 mmol/L (P<.05 versus above) in five cultures, which were exposed to normoxia for 48 hours, washed, and returned to normoxic culture for another 3 hours before assay.
Contractile Performance of Cultured Myocytes
We confirmed the findings of Ellingsen et al,18 who reported that adult myocytes maintained in normoxic culture for 1 to 4 days show a mild-to-modest reduction in contractility. In our hands, contraction amplitude was reduced by roughly 25% at 48 hours of normoxic culture compared with fresh culture medium (data not shown). Immediately after removal of myocytes from culture (either normoxic or hypoxic [1% O2 for 48 hours] culture) where they remained at rest, cells contracted in response to electrical stimulation. In hypoxic cultured cells, the amplitude of contraction (extent of shortening as a percentage of diastolic cell length) was reduced relative to that of normoxic cultured cells (Fig 4⇓, n=30 cells per group, *P<.05 versus normoxic cells). Relaxation was slowed in the myocytes that were exposed to chronic hypoxia, whereas no significant differences in diastolic cell length or time to peak contraction were seen. These findings are remarkably similar to those defined in low-flow “hibernating myocardium” in experimental animal models.19 20 21
Intracellular Ca2+ Transient and Action Potential
Contractility can be reduced by a reduction in intracellular Ca2+ availability and/or a reduction in the myofilament response to Ca2+. To examine the intracellular Ca2+ transient, indo 1–loaded myocytes were studied, and fluorescence ratios (410/490 nm) were assessed after autofluorescence correction. Fig 5⇓ shows representative Ca2+ transients for single normoxic and hypoxic cells. No differences in averaged cellular autofluorescence were seen among normoxic (n=25) and hypoxic (n=26) myocytes (autofluorescence at 410 nm=0.19±0.01 for normoxic cells and 0.20±0.01 for hypoxic cells and at 490 nm=0.18±0.01 for normoxic cells and 0.19±0.01 for hypoxic cells, P=NS). Significant differences were seen between hypoxic and normoxic cells in diastolic [Ca2+] (indo 1 fluorescence ratio of 0.82±0.01 for hypoxic myocytes [n=37] versus 0.77±0.01 for normoxic myocytes [n=26], P<.05), the amplitude of the Ca2+ transient (amplitude of indo 1 fluorescence transient of 0.15±0.01 for hypoxic cells versus 0.22±0.01 for normoxic cells, P<.05), and the time from the peak of the Ca2+ transient to 50% recovery (43±3 milliseconds for hypoxic myocytes versus 35±2 milliseconds for normoxic myocytes, P<.05). In summary, hypoxic myocytes show an increased diastolic [Ca2+], a reduced Ca2+ transient amplitude, and slowed decay of the Ca2+ transient. Since these differences in [Ca2+]i could have their basis in differences in the cellular AP, APs were examined in both groups of myocytes. Fig 6⇓ shows representative raw data traces from cultured normoxic (panel a), cultured hypoxic (panel b), and freshly isolated (panel c) cells. No significant differences were seen in resting membrane potential, peak potential, or time from peak to 50%, 75%, or 90% recovery of the AP in 20 chronically hypoxic cells and 19 chronically normoxic cells (P=NS, Table⇓). Thus, the differences in contractility (reduced contractility with slowed relaxation) seen among hypoxic and normoxic cells appears to be due, at least in part, to differences in the intracellular Ca2+ transient, yet these latter alterations do not appear as a consequence of alterations in the AP. It should be noted that both chronically normoxic and hypoxic myocytes were mildly depolarized relative to their freshly isolated counterparts (−61.9±0.7 mV in chronic normoxic cells and −61.7±2.3 mV in chronic hypoxic cells versus −69±0.8 in freshly isolated cells, n=21, P<.01 for both comparisons). Although early repolarizations (AP durations at 50% and 75% repolarization) were unaffected, time from peak to 90% repolarization was delayed after culture (133±21 milliseconds for chronic normoxic cells, 138±34 milliseconds for chronic hypoxic cells, and 65.7±11.1 milliseconds for freshly isolated cells; P<.05 for both comparisons). There were no differences between the two chronic groups in any measured electrophysiological parameter.
Development of Tolerance to Subsequent Acute Severe Hypoxia
We reasoned that myocytes that had been exposed to chronic moderate hypoxia (1% O2) might have developed adaptations that would permit them to better tolerate subsequent acute episodes of severe hypoxia. Therefore, we took myocytes that had been in culture for 48 hours and subjected them to severe hypoxia (0.003% O2) in the absence of glucose using a unique open chamber that allows single myocytes to be studied while being stimulated electrically.14 Four groups of cells were examined in the present study, including (1) cells kept in normoxic culture for 48 hours, (2) cells kept in hypoxic culture for 48 hours and then immediately subjected to acute severe hypoxia, (3) cells kept in hypoxic culture for 48 hours and then subjected to washing, the addition of new medium, and placement in a normoxic environment for 10 minutes before acute severe hypoxic exposure, and (4) cells kept in hypoxic culture for 48 hours and then subjected to washing, the addition of new medium, and placement in a normoxic environment for 3 hours before acute severe hypoxic exposure. The latter two groups were studied to examine the persistence of the response after removal from the chronic hypoxic environment. As in freshly isolated cells, contractility was retained during severe hypoxia for a variable period, and then shortening occurred (to roughly two thirds of the original length) soon after contractile failure. At reoxygenation, cells either partially relengthened, responding once again to electrical stimulation (recovery), or hypercontracted to irreversibly injured forms. Two important observations emerged from these studies: First, the time to ATP-depletion rigor contracture was increased for all groups of chronic hypoxic cells relative to normoxic cultured cells (group 1, 27.6±1.2 minutes; group 2, 42.2±1.2 minutes; group 3, 51.3±1.7 minutes; and group 4, 71.3±2.5 minutes; P<.05 versus normoxic group for all hypoxic groups by ANOVA). The time to rigor for each cell and the mean±SEM are shown in Fig 7⇓ (n=54 to 97 cells per group). Interestingly, the effect on the time to ATP-depletion rigor persisted and lengthened as the time of the intervening period of normoxia between hypoxic culture and acute severe hypoxic exposure increased. Furthermore, cells that had adapted to hypoxia had longer times to ATP-depletion rigor contracture than did the normoxic controls even when they were studied in a quiescent state, suggesting that the modest reduction in contractility in that group was not responsible for slowed ATP depletion due to reduced ATP demand (data not shown). Second, myocyte recovery at reoxygenation was markedly enhanced in all chronic hypoxic groups. This is true whether recovery was determined as a function of total hypoxic exposure or as a function of the time from ATP-depletion rigor contracture to reoxygenation (Fig 8⇓). The latter analysis essentially normalizes results for the time to rigor contracture. It demonstrates that the chronically hypoxic groups manifested improved recovery even after comparable periods of ATP depletion. The enhanced response to acute severe hypoxia cannot therefore be attributed simply to improved energetics. Rather, recovery at reoxygenation is enhanced in all groups of cells adapted to chronic hypoxia by virtue of a prolonged time to ATP depletion contracture and an additional as-yet-undefined factor.
Glycogen Content of Myocytes
To explore one possible mechanism for the prolonged time to ATP-depletion rigor contracture seen in chronically hypoxic cells upon subsequent exposure to acute severe hypoxia, glycogen content was assessed in myocytes cultured for 48 hours under normoxic conditions or hypoxic conditions. Another group of cells that had been exposed to 1% O2 for 48 hours and then subjected to washing, the addition of new medium, and reexposure to normal O2 tensions for 3 hours was also studied; this latter group had the longest times to ATP-depletion rigor onset. No difference in glycogen content was seen between normoxic and chronically hypoxic cells (68.8±6.7 versus 94.4±9.7 nmol glucose/mg protein, respectively; n=11 cultures per group; P=.12). Myocytes that had been reexposed to O2 for 3 hours after chronic hypoxic culture had increased glycogen content relative to their normoxic counterparts (190.3±45.2 nmol glucose/mg protein, n=5 cultures, P<.05 versus chronic normoxic). Thus, although hypoxic myocytes do not become glycogen-depleted, stores of glycogen actually increase when O2 levels are restored after chronic hypoxia. Three other normoxic cultures were washed, new medium was added, the cultures were replaced in the normoxic incubator for 3 hours, and then glycogen content was assayed. Glycogen content (99.5±11.0 nmol glucose/mg protein) did not increase in this group, showing that the change in medium was not responsible for the increase in cell glycogen content. Unlike the case for the relationship between medium lactate and glucose concentrations, no relationship existed between myocyte glycogen content and culture medium glucose concentration for cells cultured under either normoxic or hypoxic conditions.
The present study shows the feasibility of developing a model of chronic modest hypoxia in isolated adult rat cardiac myocytes. Compared with normoxically cultured cells, myocytes exposed to 1% O2 for 48 hours survived without a decline in viability. We are unaware of any studies of prolonged hypoxic exposure in adult myocyte cultures. The major finding of the present study is that adult cardiac myocytes adapt to chronic modest hypoxia and become more tolerant of subsequent severe acute hypoxic insults. These adaptations include a shift to an anaerobic pattern of metabolism that persists at least early after return to normal O2 tension, a preservation of glycogen stores, a reduction in contractility with a smaller amplitude intracellular Ca2+ transient, and a slowing of myocyte relaxation accompanied by a slowed decay of the Ca2+ transient. Tolerance to subsequent severe hypoxia appears to be multifactorial, with a delay in the onset of ATP-depletion rigor contracture and enhanced posthypoxic recovery that is evident even when normalized for the longer time to ATP-depletion rigor.
Adaptive Responses to Ischemia in Animal Models and in Humans
The clinical observation that coronary artery bypass grafting restores mechanical function in some individuals with areas of chronically dysfunctional myocardium led to studies that have attempted to define the basis of the reversible impairment in contraction.22 Studies in intact animals have usually focused on relatively brief periods of low coronary flow (up to 5 hours23 ) because of the difficulty of creating stable chronic low-flow states. After 3 to 5 days of partial coronary occlusion in rats, left ventricular end-diastolic pressure was elevated, and peak systolic pressure and dP/dt were reduced, consistent with contractile dysfunction.20 21 24 A more recent study showed that myocytes isolated from these hearts showed reduced contraction and increases in diastolic [Ca2+] and reductions in peak systolic [Ca2+].24 Myocytes in the present study showed a reduction in contraction amplitude and Ca2+ transient amplitude when assessed soon after removal from hypoxic culture, suggesting a strong parallel between this in vitro model of chronic hypoxia and models of acute or chronic partial coronary flow reduction. In addition to the depressed contractile response, a slowed decay of the intracellular Ca2+ transient was also noted. The cellular AP was unaltered relative to that of normoxic cultured cells, suggesting that the observed alterations in the Ca2+ transient and contraction were not due to major changes in sarcolemmal ion regulation but could perhaps have been due to altered sarcoplasmic reticulum function.
Studies also suggest that adaptations to chronic flow reduction likely occur in humans.25 26 In one study,26 samples of mechanically dysfunctional collateral-dependent myocardium obtained at surgery did not indicate infarction but showed myocytes with reduced contractile filaments, numerous small mitochondria, and glycogen accumulation. Tillisch et al25 found increased glucose uptake in dysfunctional areas of myocardium, which demonstrated improvement after coronary artery bypass grafting. These studies support the notion that chronic or chronic intermittent low-flow states may exist in animals and humans and that adaptations occur in response to the low-flow state. The specific adaptive responses that have thus far been defined include (1) decreased myocardial contractility, (2) a reduction in the amplitude of the intracellular Ca2+ transient, (3) en- hanced glycolytic metabolism, and (4) increased intracellular glycogen stores. In the present study, myocytes cultured in 1% O2 showed a 4-fold increase in lactate production relative to cells maintained under normoxic conditions. This increase in lactate is compatible with an “anaerobic” pattern of metabolism. Importantly, lactate production and efflux continued for at least a few hours after the return of hypoxic myocytes to a normoxic environment. This observation is consistent with an adaptation rather than a simple response to mitochondrial O2 limitation.
Effects of Chronic Hypoxia on Cellular Metabolism
The shift in metabolism from an oxidative pattern to a glycolytic pattern with chronic hypoxic exposure has been studied in a variety of cell types. Though glycolysis is activated transiently and acutely in severe hypoxia, a slower developing activation is seen after roughly 24 to 48 hours of modest hypoxia (1% to 3% O2) and has been associated with an increase in glycolytic enzyme content and activity in fibroblasts27 as well as in skeletal myoblasts and mouse lung macrophages.28
Recent studies offer some insight into the potential molecular basis of the coordinate increase in glycolytic enzyme expression with hypoxia. A protein known as hypoxia-inducible factor 1 (HIF-1) is synthesized rapidly during hypoxia and then activates transcription of several genes, including erythropoietin.29 A variety of cell types show HIF-1 binding activity that is induced by exposure of cells to 1% O2,30 suggesting that this pathway may serve as a common O2-sensing pathway in all mammalian cells. Semenza and colleagues31 showed that coordinate induction of glycolytic gene expression in Hep3B and HeLa cells exposed to 1% O2 was likely due to HIF-1 binding and that several of the genes encoding these enzymes had HIF-1–like DNA binding sites.
Cellular “Sensing” of Modest Hypoxia
That myocytes show these adaptive responses to modest hypoxia is remarkable, since experiments with suspensions of resting adult rat myocytes showed that mitochondrial cytochromes were fully oxidized above 1 mm Hg O2 (≈0.1% to 0.2% O2), that O2 consumption became O2-limited only at <0.3 mm Hg, and that lactate accumulation occurred at <0.1 mm Hg (0.01% O2) during acute hypoxic exposure.7 Thus, mitochondria in myocytes exposed to 1% O2 (7 to 8 mm Hg) should not be O2-limited. Cells studied in high-density culture consume O2 and may develop a deep unstirred layer of medium. A significant O2 gradient may therefore exist from the medium surface to the cell surface, which could reduce O2 tension to levels that become O2 limiting for mitochondria. To ensure that little or no O2 gradient exists in our system, cells were sparsely cultured, and the height of overlying culture medium was minimized. The lack of a significant O2 gradient is suggested by calculation of the O2 tension predicted under our conditions using Fick’s law32 : P2=P1−(ah/sk), where P2 is the O2 tension at the cell surface (% O2), P1 is the O2 tension of the incubator (1% O2), a is O2 consumption of myocytes (1.4×10−6 mL·min−1·cm−2), h is the height of the medium layer (0.064 cm), s is the O2 solubility constant (0.024 cm3 gas/cm3), and k is the O2 diffusion coefficient (0.0035 cm2/min). The above calculations are based on a culture in a 10-cm-diameter dish at the conditions used in the present study, assuming an O2 consumption rate for resting myocytes of 10 nanoatom O·min−1·mg−1 as reported by Haworth et al33 and show that the predicted O2 tension at the myocyte should be roughly 0.9% if the incubator is equilibrated with 1.0% O2. We have found that myocyte behavior is identical whether the cultures are unstirred (as above) or if they rest on a rotary shaker, thus preventing any unstirred layers, offering further evidence against a significant O2 diffusion gradient.
The fact that hypoxic adaptation occurs at these rather modest levels of hypoxia suggests that there may be a cellular “O2 sensor” that binds O2 with a far lower affinity than mitochondrial cytochromes.
Development of Tolerance to Subsequent Acute Severe Hypoxia
Numerous studies have focused on the mechanisms underlying ischemic preconditioning in which brief (5- to 10-minute) episodes of ischemia protect against longer severe episodes (reviewed by Lawson and Downey5 ). Although preconditioning is one of the most powerful forms of myocardial ischemic protection, its effects are relatively short-lived. Protection is lost after 1 hour of delay between the preconditioning stimulus and the ischemic episode in the rat.34 In contrast, in our present study of hypoxic tolerance, relatively long episodes of modest hypoxia (hours) induced tolerance to subsequent anoxia, which persisted for at least 3 hours after returning myocytes to a normoxic environment.
Tolerance was manifested in two specific ways: First, the time to ATP-depletion rigor contracture on exposure of chronically hypoxic myocytes to a subsequent acute severe hypoxic insult was markedly prolonged relative to that for myocytes cultured for a similar time under normoxic conditions. Second, when cells were reoxygenated at similar times after ATP-depletion rigor, enhanced morphological and functional recovery was observed relative to the cultured normoxic counterparts. Although the delay in ATP depletion during acute severe hypoxia may be due in part to increased intracellular glycogen stores, this is unlikely to fully account for the delayed time to ATP-depletion rigor observed in the present study. Myocytes that were immediately subjected to severe hypoxia after 48 hours in 1% O2 demonstrated no significant increase in glycogen content, yet these cells showed a statistically significant (roughly 50%) increase in the time to ATP-depletion rigor. Although it may be that baseline levels of ATP or creatine phosphate could be increased in myocytes that had been cultured under hypoxic conditions relative to those cultured under normoxic conditions, this seems not only unlikely but also not sufficient to delay ATP depletion by any significant amount during subsequent severe acute hypoxia. Instead, a downregulation in metabolism may account for the slowed ATP depletion seen during later severe hypoxia.
As mentioned previously, recovery was enhanced in those cells that had been cultured in 1% O2, with and without an intervening period of normoxia, even when myocytes were reoxygenated at the same time after the development of rigor contracture. This adaptation would be one independent of energetics, since myocytes remain fully depleted of ATP until they arereoxygenated. Studies from our laboratory17 and others35 show an association between acute hypoxic cellular Ca2+ loading and failure of recovery once hypoxia is terminated. Thus, it may be that an adaptation occurs that reduces the extent to which myocytes load with Ca2+ during this subsequent severe hypoxic period or reduces the deleterious response to Ca2+ loading.
The present study shows that in response to acute hypoxia, adaptations occur that may facilitate cell survival and impart a natural defense to subsequent severe episodes of hypoxia. These processes may also occur in humans, as has been suggested by a variety of studies. Since the isolated cell model allows precise environmental control and is stable for several days, it offers certain advantages over other models of low-flow states. We recognize that it does not precisely mimic the in vivo state. An issue that remains to be resolved is whether adaptive responses in some animal models and in humans occur in response to sustained hypoxia or to intermittent periods of hypoxia, which are separated by periods of relatively normal tissue oxygenation, as has been suggested by some groups.26 36 The present study suggests that some responses result purely from chronic hypoxia, whereas others depend on subsequent restoration of normal oxygenation.
This study was supported by National Heart, Lung, and Blood Institute grants R01-HL-42050 and K08-HL-02539.
The opinions expressed in this article are those of the authors and do not necessarily reflect the policies of the US Air Force, the Department of Defense, or the Uniformed Services University.
- Received September 25, 1996.
- Accepted January 3, 1997.
- © 1997 American Heart Association, Inc.
Cave AC, Adrian S, Apstein CS, Silverman HS. Anoxic preconditioning fails to reduce subsequent anoxic injury in isolated rat cardiac myocytes. Circulation. 1993;88(suppl I):I-544. Abstract.
Murry CE, Jennings RB, Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation. 1986;74:1124-1136.
Lawson CS, Downey JM. Preconditioning: state of the art myocardial protection. Cardiovasc Res. 1993;27:542-550.
Tajima M, Katayose D, Bessho M, Isoyama S. Acute ischaemic preconditioning and chronic hypoxia independently increase myocardial tolerance to ischaemia. Cardiovasc Res. 1994;28:312-319.
Wittenberg BA, Wittenberg JB. Oxygen pressure gradients in isolated cardiac myocytes. J Biol Chem. 1985;260:6548-6554.
Kreutzer U, Jue T. Critical intracellular O2 in myocardium as determined by 1H nuclear magnetic resonance signal of myoglobin. Am J Physiol. 1995;268:H1675-H1681.
Kirshenbaum LA, MacLellan WR, Mazur W, French BA, Schneider MD. Highly efficient gene transfer into adult ventricular myocytes by recombinant adenovirus. J Clin Invest. 1993;92:381-387.
Bondar RJL, Mead DC. Evaluation of glucose-6-phosphate dehydrogenase from Leuconostoc mesenteroides in the hexokinase method for determining glucose in serum. Clin Chem. 1974;20:596-601.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin Phenol reagent. J Biol Chem. 1951;193:265-275.
Haworth RA, Hunter DR, Berkoff HA. Contracture in isolated adult rat heart cells: role of Ca2+, ATP, and compartmentation. Circ Res. 1981;49:1119-1128.
Stern MD, Silverman HS, Houser SG, Josephson RA, Capogrossi MC, Nichols CG, Lederer WJ, Lakatta EG. Anoxic contractile failure in rat heart myocytes is caused by failure of intracellular calcium release due to alteration of the action potential. Proc Natl Acad Sci U S A. 1988;85:6954-6958.
Silverman HS, DiLisa F, Hui RC, Miyata H, Sollott SJ, Hansford RG, Lakatta EG, Stern MD. Regulation of intracellular free Mg2+ and contraction in single adult mammalian cardiac myocytes. Am J Physiol. 1994;266:C222-C233.
Stern MD, Chien AM, Capogrossi MC, Pelto DJ, Lakatta EG. Direct observation of the ‘oxygen paradox’ in single rat ventricular myocytes. Circ Res. 1985;56:899-903.
Miyata H, Lakatta EG, Stern MD, Silverman HS. Relation of mitochondrial and cytosolic free calcium to cardiac myocyte recovery after exposure to anoxia. Circ Res. 1992;71:605-613.
Ellingsen O, Davidoff AJ, Prasad SK, Berger H-J, Springhorn JP, Marsh JD, Kelly RA, Smith TW. Adult rat ventricular myocytes cultured in defined medium: phenotype and electromechanical function. Am J Physiol. 1993:265:H747-H754.
Capasso JM, Jeanty MW, Palackal T, Olivetti G, Anversa P. Ventricular remodeling induced by acute nonocclusive constriction of coronary artery in rats. Am J Physiol. 1989;257:H1983-H1993.
Capasso JM, Li P, Anversa P. Non-ischemic origin of myocardial damage induced by short-term nonocclusive constriction of the coronary artery in rats. Am J Physiol. 1991;260:H651-H661.
Bolukoglu H, Liedtke J, Nellis S, Eggleston A, Subramanian R, Renstrom B. An animal model of chronic coronary stenosis resulting in hibernating myocardium. Am J Physiol. 1992;263:H20-H29.
Matsuzaki M, Gallagher KP, Kemper WS, White F, Ross J. Sustained regional dysfunction produced by prolonged coronary stenosis: gradual recovery after reperfusion. Circulation. 1983;68:170-182.
Capasso JM, Li P, Anversa P. Cytosolic calcium transients in myocytes isolated from rats with ischemic heart failure. Am J Physiol. 1993;265:H1953-H1964.
Vanoverschelde J-L J, Wijns W, Depre C, Essamri B, Heyndrickx GR, Borgers M, Bol A, Melin JA. Mechanisms of chronic regional postischemic dysfunction in humans: new insights from the study of noninfarcted collateral-dependent myocardium. Circulation. 1993;87:1513-1523.
Hance AJ, Robin ED, Simon LM, Alexander S, Herzenberger LA, Theodore J. Regulation of glycolytic enzyme activity during chronic hypoxia by changes in rate-limiting enzyme content. J Clin Invest. 1980;66:1258-1264.
Wang GL, Semenza GL. Characterization of hypoxia-inducible factor-1 and regulation of DNA binding activity by hypoxia. J Biol Chem. 1993;29:21513-21518.
Wang GL, Semenza GL. General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. Proc Natl Acad Sci U S A. 1993;90:4304-4308.
Semenza GL, Roth PH, Fang H-M, Wang GL. Transcriptional regulation of genes encoding glycolytic enzymes by hypoxia-inducible factor 1. J Biol Chem. 1994;269:23757-23763.
Haworth RA, Hunter DG, Berkoff HA, Moss RL. Metabolic cost of the stimulated beating of isolated adult rat heart cells in suspension. Circ Res. 1983;52:342-351.
Allshire A, Piper HM, Cuthbertson KSR, Cobbold PH. Cytosolic free calcium in single rat heart cells during anoxia and reoxygenation. Biochem J. 1987;244:381-385.
Shen Y-T, Vatner SF. Mechanism of impaired myocardial function during progressive coronary stenosis in conscious pigs: hibernation versus stunning? Circ Res. 1995;76:479-488.