| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Biochemistry and Molecular Medicine, University of California, Davis, California
Submitted 9 June 2005 ; accepted in final form 19 December 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
left ventricular hypertrophy; myocardial ischemia; spontaneously hypertensive rat; nuclear magnetic resonance
Inflammatory and oxidative signals, including TNF-
or oxygen free radicals, might contribute to the increased susceptibility of hypertrophic myocardium to ischemia (31, 55). Yet many studies have also demonstrated an altered energy metabolism in LVH and have suggested a cellular hypoxia (15, 41, 61, 64). Consistent with possible O2 limitation, a depressed O2 consumption in the hypertrophic myocardium, especially under low perfusion pressure, has been reported (9, 61). These findings suggest that the LVH heart may be partially ischemic even under normal unstressed conditions, increasing its vulnerability to major ischemic events. In a word, O2 supply of the LVH heart might be so close to being limiting (or already limiting) that an ischemic episode will readily "tip it over the edge."
LVH O2 limitation may result from a compromised capillary-to-cell O2 diffusion. Even though the absolute number of capillaries can be increased, capillary proliferation in the pressure-induced LVH heart fails to match the myocyte size increase, resulting in lower capillary density and elongated capillary-to-cell O2 diffusion distance (15, 23, 43). At the same time, the increased cell volume-to-surface area ratio of hypertrophied myocytes decreases the effective area of the diffusional plane.
Furthermore, abnormalities of the coronary vasculature may also contribute to hypoxia in LVH. Pressure-induced hypertrophy leads to the loss of coronary flow reserve, narrowed arteriolar lumens, and increased minimal coronary vascular resistance, particularly in the subendocardial region (4, 61). In effect, either the defect in the conductive (diffusion) and/or convective (flow) component of O2 delivery can cause O2 shortage in LVH.
LVH hypoxia, implicated by histological and vascular evidences, nonetheless has eluded a direct demonstration until recently. One of the recent efforts to directly measure O2 level in the intact tissue is the 1H NMR detection of myoglobin (Mb) signal, which provides an intracellular O2 marker in heart and skeletal muscle (20, 60). Indeed, a 1H NMR study (5) of in situ canine hearts examined the oxygenation state in LVH and reported a normal O2 level. This report then suggests that the LVH O2 supply in vivo may not be as compromised as has been previously surmised.
Yet to be determined, however, is the effect of the LVH O2 condition on the ischemia-reperfused heart. As O2 returns to the O2-depleted heart, the abnormalities of O2 transport in LVH, either diffusive or convective, are likely to be magnified. Therefore, the current study aimed to test the hypothesis that a reduced availability of cellular O2 is responsible for poor energetic and functional recoveries documented for reperfused LVH hearts. Cellular O2 level was measured by 1H NMR signals of Mb. A simultaneous physiological monitoring and interleaved 31P NMR acquisition traced energetic and functional states of the heart.
Unlike the in situ canine 1H NMR work (5, 65), which utilized deoxygenated Mb signal as a counterindex of O2 concentration ([O2]), the present perfused heart study used the signal that arises from oxygenated Mb. When Mb is bound to O2, the ring current-shifted, three-proton signal from Val E11, originally assigned from the isolated protein studies (40, 51), resonates at –2.8 parts per million (ppm) in the 1H NMR spectra of the heart (11). Although the isolated perfused heart does not approximate in vivo physiology as closely as do the in situ animal models, the isolated perfused heart preparation allows a ready control of perfusion flow, free from the backdrop of nervous and hormonal interactions. A full control of perfusion flow should facilitate understanding the mechanisms behind O2 limitation, i.e., impaired diffusion versus compromised flow. To that end, the present study alternately utilized constant-flow and constant-pressure perfusion mode. Ischemic and hypoxic perturbations characterized myocardial responses.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
For the hypertrophic model, spontaneously hypertensive rats (SHR) (Taconic Farms, Germantown, NY), 7–9 mo old, were used. In SHR, elevated blood pressure is detected by the second month of age, and, concurrently, LVH develops. The spontaneous and gradual development of hypertension in SHR is considered to mimic human essential hypertension. Age-matched normotensive Wistar-Kyoto (WKY) rats were used as controls. SHR and WKY rats were maintained in a facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. The care and experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee at University of California at Davis.
Rat Heart Perfusion
Rat heart perfusion at 36°C followed the reported methods (7, 25, 54) and was previously described (11). The isolated heart was placed in a 20-mm diameter NMR tube and anchored on a Teflon plug with holes to permit perfusate overflow. The perfusion medium was Krebs-Henseleit buffer containing (in mM) 118 NaCl, 4.7 KCl, 1.2 KH2PO4, 1.8 CaCl2, 1.2 MgSO4, 20 NaHCO3, and 15 glucose. The perfusate was saturated with 95% O2-5% CO2, which maintained the pH at 7.4 ± 0.05 (SE), and passed through 5-µm and 0.45-µm Millipore filters. The gas equilibration was achieved by initially bubbling the gas mixture in the perfusate reservoir and later again passing the perfusate through a home-built membrane lung chamber, comprising gas-permeable Dow Corning Silastic tubing (ID, 0.058 in.; OD, 0.077 in.) wrapped around a heat exchanger.
The perfusate was delivered by a peristaltic pump (Rainin RP-1) to the top of the magnet, where the head pressure was set at 100 mmHg, a pressure high enough to allow hypertensive coronary vessels to autoregulate the flow (61). Coronary flow was measured by collecting the effluent perfusate returning from the heart. For the constant-flow perfusion mode, the buffer chamber on top of the magnet was bypassed, forming a closed delivery system. In both constant-pressure and constant-flow experiments, the perfusate was not recirculated.
When global ischemia was induced, perfusion flow was stopped at two clamping points. During ischemia, the heart and the surrounding perfusate remaining in the NMR tube were maintained at 36°C, with warm air provided by the variable temperature unit in the magnet. For hypoxia experiments, a Cameron gas-mixing flowmeter (GF-4/MP) controlled the fraction of 95% O2-5% CO2 and 95% N2-5% CO2. The precision of gas mixing was ± 0.1%. With a switch to the different gas-mixing ratio, the equilibration between the gas mixture and the perfusion buffer in the membrane lung unit was completed within 
2 min.
Approximately 50% of the perfusate flowing from the heart was withdrawn through a polyethylene catheter inserted close to the pulmonary artery, and the O2 concentration of the perfusate was measured by a Yellow Springs Instrument (YSI) 5300 O2 meter. Parallel non-NMR experiments determined empirically the O2 loss in the perfusion tubing that extends from the heart placed in the magnet to the O2 meter set up outside the magnet. The measured partial pressure of O2 (PO2) was then corrected accordingly to reflect venous and arterial value. Myocardial O2 consumption rate (M
O2) was calculated from the corrected inflow and outflow O2 measurements and the coronary flow rate.
A saline-filled latex balloon inserted in the LV monitored LV pressure and heart rate (HR) via a strain-gauge transducer (Statham P23 XL), which were traced by an oscillographic recorder (Gould Windograf). The balloon volume was adjusted to give a resting end-diastolic pressure (EDP) of 8 ± 1 mmHg. The rate-pressure product (RPP) was calculated by multiplying HR by LV developed pressure (LVDP). After the perfusion-NMR experiments, hearts were dissected, dried at 80°C for 72 h, and weighed.
Perfusate Metabolite and Enzyme Assays
Perfusate lactate concentration was determined in triplicates by a lactate oxidase method (YSI 2700 Bioanalyzer). Mb leakage from the heart was assessed by optically scanning the perfusate sample (pH 8.2, equilibrated with room air) at the Soret band (410–420 nm) (3). Creatine kinase (CK) activity in the perfusate was measured by a spectrophotometric enzyme assay (16). For CK and Mb, total leakage amount was calculated by integrating the measured concentration over the respective time periods of interest.
NMR
An AMX 400-MHz Bruker spectrometer was used to record 1H/31P signals with a 20-mm 1H-{X} probe, where X represents nuclei from 15N to 31P. A modified 1331 binomial pulse sequence suppressed the H2O peak and selectively excited oxy-Mb (MbO2) Val E11 (
-CH3) resonance at –2.8 ppm (18). 1H 90° pulse (65 µs) was calibrated on the perfusate H2O signal. Observation of MbO2 signal required a 40-ms acquisition time and a 45° pulse angle. The spectral width was 8,065 Hz; the data size was 512. Six thousand transients were averaged for a typical 1H spectrum, requiring 5 min of signal accumulation (50-ms repetition time). The free induction decays (FIDs) were zero-filled to 2 K and multiplied by an exponential-Gaussian window function. A spline fit smoothed the baseline. The MbO2 signal was referenced to H2O peak as 4.67 ppm at 36°C, which was in turn calibrated against sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4.
31P NMR spectra were acquired using 60° pulse angle and 2-s repetition time, a pulsing condition known to minimize the signal saturation (54). The fully relaxed spectra utilized 90° pulse angle with a 20-s repetition time. The 31P 90° pulse (72 µs) was calibrated on 0.1 M phosphate solution. The spectral width was 6,494 Hz; the data size was 4 K. The FIDs were zero-filled to 8 K and apodized with an exponential-Gaussian function to better resolve intra- and extracellular Pi signals. 31P signals were referenced to creatine phosphate (CrP) as 0 ppm.
Experimental Protocols
Ischemia experiment. During the baseline perfusion period (20 min), interleaved 1H (5 min)/31P (5 min) spectra traced the basal metabolism. Subsequently, hearts were subjected to a normothermic global ischemia for 20 min. To follow O2 time course changes during the early stage of ischemia, only 1H signals with shorter accumulation time (2.5 min) were collected in the first 15 min of an ischemic period, yielding six MbO2 data points. During the last 5 min of ischemia, a 31P spectrum was collected. Similarly, in the first 15 min of a reperfusion period, 1H signals were accumulated every 2.5 min, followed by two sets of interleaved 1H (5 min)/31P (5 min) spectra (total 35 min of reperfusion).
Hypoxia experiment. In a separate set of experiments, after normoxic baseline perfusion, hearts were subjected to varying degrees of hypoxia (5–90% O2) for 40 min. Hearts were perfused in the constant-flow mode to minimize the flow responses and to focus primarily on the diffusion limitation. One hypoxic step was tested for each (n = 1) experiment to ensure that prehypoxic conditions for all data points were equivalent.
Cellular PO2 Analysis
Myocardial O2 level was assessed by the 1H NMR signals of MbO2 (Fig. 1). Fractional Mb oxygenation,
 | (1) |
 | (2) |
|
78 ppm), which becomes detectable under ischemic or high-workload conditions (12, 20), was not visible, confirming the high basal oxygenation level of both SHR and WKY hearts. However, 100% preischemic MbO2 saturation was not assumed, because NMR detection threshold for deoxy-Mb peak can be as high as 20% (12). In addition, because the current study reports the first attempt to measure 1H NMR signals of MbO2 in the constant-pressure perfusion mode, additional experiments were carried out to establish the validity of high level baseline MbO2 saturation. First, the hearts were perfused initially with constant pressure, followed by a constant-flow perfusion at the flow rate 30% above the flow resulting from constant-pressure perfusion. No further increase in MbO2 intensity was observed by this perfusion. Second, at this increased flow level, perfusate O2 concentration was lowered. Until the level of hypoxia reached 60% of normoxic O2 level, the MbO2 signal intensity did not decrease. These results are thus consistent with Mb-O2 saturation approaching a plateau.
However, because of the nonlinear nature of Mb-O2 binding, only a drastic increase in flow rate (2 times) will reveal NMR detectable signal changes, as long as Mb is 
75% O2 saturated. For example, to raise %MbO2 from 80 to 90%, a 2.3 times increase in PO2 (therefore, 2.3 times flow rate) would be required. Because an excessive perfusion flow can damage the integrity of the heart, such manipulation was not carried out. Therefore, given the nonlinear nature of Mb-O2 binding and 10% error in NMR detection of MbO2 signal, one is still left with the possibility that the baseline MbO2 saturation can range between 75 and 100%. As a result, the baseline Y value was assessed numerically by retrofitting MbO2 data points obtained from the varying hypoxic measurements to the set of hyperbolic Mb-O2 binding curves predicted for the hypothetical baseline Y values. On the basis of this estimation, Mb was 75% O2 saturated during baseline perfusion in both SHR and WKY (Table 3) rats. Consistent with this numerical estimation, maximal dilation of the coronary arteries with adenosine under potassium arrest (to maximize supply and minimize demand for O2) increased MbO2 signal intensity by 20% (results not shown). Postischemic Y values were then evaluated by calibrating respective MbO2 peaks to the baseline spectra.
|
0.6-mm penetration depth), whereas NMR measures the whole tissue. Furthermore, the cited optical studies (32, 49) used substantially lower perfusion pressure than the present study, and the resulting coronary flow rate was approximately one-half the flow rate recorded in the current study. Indeed, low O2-carrying capacity of red blood cell-free buffer may not permit full O2 saturation of Mb in the actively metabolizing tissue, which is in agreement with pervasive concerns about saline perfusion. Nevertheless, my experiments (unpublished observations) and the graded hypoxia experiments of others (49) clearly demonstrate that submaximal MbO2 saturation does not have to be equated with submaximal heart function/metabolism. The critical PO2 level below which the function and energy metabolism of the heart start to deteriorate (if defined as 10% drop from the baseline values) corresponds to 55–65% MbO2 saturation level. Therefore, the buffer-perfused heart, despite submaximal MbO2 saturation, can provide a metabolically stable system within the limited time frame.
Another factor that can potentially undermine the accuracy of %MbO2 measurement is the possibility of Mb leakage from the heart as a result of ischemic/hypoxic cellular damage. Some Mb release did occur during postischemic reperfusion in both WKY and SHR. However, Mb loss over the entire experimental protocol was <5% of total rat heart Mb. Because the error of MbO2 measurement by 1H NMR is 10%, the inaccuracy introduced by Mb leakage stays within the overall experimental error. In the hearts subjected to hypoxic manipulations, no Mb leakage was detected.
Phosphate Metabolite Analysis
To quantify the absolute concentrations of high-energy phosphate (HEP) metabolites (CrP, Pi, and 
-ribonucleoside triphosphate), the ventricular balloon was filled with 100 mM phenylphosphonate solution, following the methods described by Kingsley-Hickman et al. (25). After the baseline 31P spectra were collected, the balloon was inflated by 0.04 ml, which corresponds to 4 µmol of phenylphosphonate signal. The change in phenylphosphonate signal intensity was considered to represent 4 µmol of 31P nuclei and was used to quantify other signals in the spectra. Because balloon inflation increases the workload of the heart and can potentially affect the outcome of ischemic experiments, 31P signal quantification was carried out in separate experiments (n = 10) that were not included in the experimental data pool. Once the average baseline ATP concentration was determined in this manner, other 31P signals were quantitated from the fully relaxed spectra obtained at the beginning of each experiment. With the assumption that no significant changes in the longitudinal relaxation time (T1) occur as a result of ischemia (35), the same T1 correction factor was applied for the partially saturated spectra from preischemic, ischemic, and postischemic periods.
Intracellular pH, reflected by the chemical shift position of the Pi peak, was evaluated from the equation
 | (3) |
A = ppm of H2PO4– at 3.290 ppm, B = ppm of HPO42– at 5.805, and o = ppm of observed Pi peak (19). The ADP level was calculated from the CK equilibrium
 | (4) |
 | (5) |
|
The data are reported as means ± SE. Statistical significance of t-test was accepted at P < 0.05. The least squares method (Sigma Plot, Jandel Scientific) determined slopes, intercepts, and correlation coefficients of linear regression. The slopes of the regression lines were compared by testing the distribution of t, t = (b1 – b2)/(Sb1 – b2), where (b1 – b2) is the difference of regression slopes and (Sb1 – b2) is the standard error of the difference of regression slopes.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The LV-to-body weight (BW) ratio (LV/BW) was increased by 48% in SHR, showing a moderate degree (30–50% increase) of LVH (Table 1). For the comparable age groups, WKY rats are substantially larger than SHR. To assess whether the larger body size of WKY rats contributed to the underestimation of LV/BW ratio, data from only the comparable weight ranges of SHR and WKY rats (n = 8 total, 400–500 g range) were also analyzed. The LV/BW ratio of this selected group again showed a significant difference between SHR and WKY rats (30% increase in SHR). Right ventricle (RV)/BW ratio of SHR heart was not changed from normal (Table 1).
Baseline Physiology and Metabolism
The RPP (LVDP x HR), an index of myocardial mechanical function, was higher in SHR because of higher LVDP. HR was the same as in WKY rats. Despite the increased RPP, the MO2 of SHR hearts was unchanged from normal. Consequently, cardiac aerobic efficiency (RPP/MO2) was increased in SHR. Coronary flow, normalized to the heart weight, was slightly lower in SHR, but the difference was not statistically significant. These results are summarized in Table 2.
|
75% saturated with O2 in both SHR and WKY hearts during baseline normoxic perfusion (see MATERIALS AND METHODS). 31P NMR spectra showed that CrP level, CrP/Pi ratio, and phosphorylation potential were all increased in SHR (Table 3). Cellular pH and lactate production rates were unchanged. Hypoxia
Perfusing with hypoxic buffer did not reveal any relative O2 limitation in SHR heart. Figure 2 depicts intracellular PO2 at varying venous PO2, of which slopes (<<1.0) indicate steep extra-intracellular O2 gradients. These slopes are the same for SHR and WKY hearts (SHR, y = 0.042x – 0.47, r = 0.95; and WKY, y = 0.045x – 0.67, r = 0.91), showing no further decrease in cellular O2 in SHR hearts throughout the entire extracellular O2 range tested. The x-intercepts reflect the threshold venous PO2 at which cellular PO2 approaches zero.
|
O2, CrP, and MbO2 (with a small overshoot in CrP and RPP) all returned to the prehypoxic levels (results not shown). Ischemia
As soon as perfusion flow stops (20-min normothermic global ischemia), RPP starts to decline in both SHR and WKY rats. After 4 min, no detectable (<0.5%) mechanical activity remains (Table 4). The early phase of the RPP drop is caused primarily by the decrease in the systolic pressure. Subsequently, HR declines. At 9–11 min after the ischemic onset, the contracture (defined in this study as a rise of resting LV pressure by 3 mmHg above preischemic EDP) develops. Although the contracture builds more slowly in SHR, the peak pressure reached during contracture is significantly higher (Table 4).
|
|
A period of ventricular fibrillation and arrhythmia (10 min) follows the onset of reflow until the normal sinus rhythm contractions resume. However, even after the normal contraction pattern returns, RPP continues to improve for another 10–15 min (Fig. 3B). In SHR, the final postischemic RPP level is significantly lower than preischemic RPP (Table 2; note that in SHR, even after being significantly reduced from the preischemic level, the absolute value of postischemic RPP is comparable to WKY rats, as a result of higher baseline value). In WKY rats, RPP almost fully recovers. The lower %RPP recovery in the SHR heart is the result of lower LVDP recovery (32% decrease), because HR changes (15% decrease) are similar in SHR and WKY. Coronary flow, on the other hand, recovers to the depressed level in both SHR and WKY (Table 2 and Fig. 3A). As a result, %RPP recovery is higher than %coronary flow recovery in WKY rats, whereas the reverse is true for SHR. The decrease in postischemic RPP is paralleled by the depressed MO2, leaving the aerobic efficiency (RPP/MO2) unchanged from the preischemic values in both SHR and WKY rats. In addition to the difference in RPP recovery, a notable difference in diastolic dysfunction is evident as more than two times higher postischemic EDP in SHR (Table 2).
In response to reflow, the return of O2 is swift (Fig. 1, spectrum c, and Fig. 3A), yet it recovers to the level significantly lower than baseline oxygenation, particularly in SHR. Concomitant with severely depressed O2, HEP metabolite recovery is depressed in SHR (Table 3). CrP returns to 50% of preischemic level in SHR, whereas it recovers fully in WKY rats. ATP, depressed in both SHR and WKY, is lower in SHR. As a result, postischemic phosphorylation potential in SHR heart is well below the preischemic level, whereas it is unchanged in WKY heart. Intracellular pH remains slightly acidic during reperfusion in both SHR and WKY hearts. Although the average pH values are similar in SHR and WKY hearts, SHR pH shows a pattern of heterogeneous distribution, indicated by the broadened line width of the Pi resonance. Lactate production rate, an index of anaerobic glycolysis, nearly doubles from preischemic level in WKY rats. In SHR, it remains the same (Table 2).
Irreversible Cell Damage
Although some CK release persists during postischemic reperfusion, total leakage amounts to only 2% of CK present in the rat heart tissue (44). Mb leakage is <5% of total cellular Mb. Taken together, these results suggest that no appreciable degree of irreversible cellular damage was induced by the current ischemic protocol (Table 2).
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The ventricular hypertrophy examined in the present study is a compensated hypertrophy. In SHR, up to 12 mo of age, LVH remains in a compensated state without overt hemodynamic failures (42). In the current study, 7- to 9-mo-old rats were used. The degree of LV hypertrophy (48%) was moderate, and the signs of heart failure, such as increased RV/BW ratio, pleural effusion, and hepatomegaly, were absent.
In line with the "compensated" nature of hypertrophic hearts in our study, physiological indexes were either normal (cellular PO2, coronary flow and MO2) or supranormal (RPP). Because RPP was increased and MO2 was unchanged, cardiac aerobic efficiency (defined as RPP/MO2 ratio) was increased in SHR.
Although the contractility of the SHR heart is often overcompensated (14, 24, 39, 53, 61) and the energy utilization can be more efficient in LVH (45, 61), why SHR hearts examined in the current study show higher baseline energy level is not clear. In fact, literature reports are mixed on the basal HEP level in moderate LVH: decreased, normal, or slightly increased (6, 41, 53, 57, 61). In the case of human LVH patients, either no change (34, 48, 62) or varying degrees of reduction in the CrP/ATP ratio (21, 28) have been observed. The latter case has been documented mostly in the symptomatic patients and correlates with the severity of hypertrophy. Some workers (34, 52) report that the primary hypertrophic cardiomyopathy is more likely to result in an energy decline than the secondary hypertrophy, such as hypertension-induced LVH. In addition, a coexisting heart failure seems to increase the incidence of HEP abnormalities (21, 47).
Similar to humans, the extent of hypertrophy and the absence/presence of heart failure affect the energy profile in SHR; the energy level of SHR hearts remains normal at 
18 mo of age and starts to decline as the signs of heart failure become more evident (39, 41, 56). Experimentally induced LVH models, on the other hand, seem to show the tendency for early metabolic and functional decline even when hypertrophy is moderate (9, 14, 57).
However, the increased, rather than normal, SHR HEP level observed in the current study is puzzling and may originate from the substrate condition of our heart perfusion. A shift in substrate preference toward glucose is commonly observed in LVH, including SHR (10, 26, 30, 45). In SHR, glucose transporter-1 (GLUT-1) expression is increased, whereas the GLUT-4 level is normal (26). Because GLUT-1-mediated transport is driven primarily by the glucose concentration gradient and glucose was the only substrate used in our study (in relatively high concentration and in the absence of insulin), there is a possibility that SHR hearts utilized the substrate more efficiently than WKY hearts.
Although the point was not emphasized, a report of a glucose-perfused heart study (41) showed an apparent increase in the CrP/ATP ratio in 10-mo-old SHR (1.83 ± 0.09) over that of WKY (1.67 ± 0.07) rats. A second perfusion study (7) that used glucose as the only substrate described the increased CrP/Pi and CrP/ATP ratios in 18-mo-old SHR, compared with 12-mo-old SHR, suggesting a possible progression with age of glucose preference (comparison with normotensive rat was not presented in this study). Given the vulnerability of LVH hearts to lower perfusion pressure (53, 61), different perfusion pressures used in various studies could also contribute to varying HEP results.
Baseline O2 Level in Hypertrophic Heart
MbO2 signal obtained from SHR hearts during baseline perfusion was of the same intensity as from WKY hearts (Table 3), indicating a normal basal O2 level in SHR. However, it is inherently difficult to determine PO2 differences under O2 abundant conditions; due to nonlinear Mb-O2 binding, wide-ranging values (from 11 to infinite mmHg) result when PO2 is calculated from 75% O2-saturated Mb (see Eq. 2 in MATERIALS AND METHODS). For a more precise comparison, O2 measurements were made in this study under varying degrees of hypoxic conditions that would bring Mb-O2 saturation below 70%, so that PO2 values can be derived from a relatively linear portion of the Mb-O2 binding curve. In these experiments, hearts were perfused in the constant-flow mode to minimize the flow responses to hypoxia and to focus primarily on the capillary-to-cell O2 diffusion.
Even hypoxic manipulations, however, failed to reveal any further cellular PO2 reduction in SHR (Fig. 2). The small, statistically insignificant difference in the x-intercepts is, in fact, in the opposite direction from what would be expected; WKY cellular O2 is depleted at 15.1 mmHg venous PO2 versus SHR at 11.3 mmHg, inconsistent with the presumed O2 handicap of the LVH heart.
In summary, SHR hearts show no reduced O2 availability in the baseline as well as hypoxic conditions.
Postischemic Functional Recovery and O2
Increased vulnerability to ischemia is well recognized among advanced stage LVH (1, 9, 15, 31, 55). Ischemic responses of fully compensated LVH hearts, however, are more controversial (24, 53). In the current study, despite its normal O2 level and well-compensated heart function/metabolism in the basal state, the response of SHR hearts to ischemia was drastically different from normal hearts, displaying a pronounced contractile and metabolic dysfunction. Certainly, preischemic myocardial O2 limitation is not a prerequisite for postischemic dysfunction in LVH.
The irrelevance of preischemic O2 condition, however, does not negate the importance of postischemic [O2]. The eventual postischemic RPP level correlates well with the reperfused O2 level. First, as both SHR and WKY RPPs are reduced from the preischemic level, so are the reperfused O2 levels. Second, [O2] is further decreased in SHR in proportion to the degree of RPP depression (Fig. 4A). Quantitatively matching with lower O2/RPP is the lower CrP level in SHR (Fig. 4B), which implies a reduced aerobic energy turnover. This point is reinforced by the lower MO2 (Table 2 and Fig. 4A). The excellent correlation between RPP, [O2], [CrP], and the oxidative energy turnover rate (MO2) then indicates a probable sequence of events occurring in the postischemic SHR hearts; poor O2 reperfusion leads to aerobic energy failure, which then leads to depressed heart function.
|
O2 coupling efficiency normalized back to the preischemic level in both SHR and WKY rats. This means that RPP and MO2 were reduced to the same degree from their respective preischemic levels. As a result, the RPP/MO2 ratio, which was higher in SHR hearts during baseline perfusion, remained higher in SHR even during reperfusion.
It is also evident that the kinetic control of oxidative phosphorylation by [ADP] or [Pi] is not a dominant mechanism during reperfusion. If this were the overriding mechanism, given higher levels of both, SHR hearts should be at an advantage. Nor is the thermodynamic control by phosphorylation potential at work. If it were, again, the oxidative phosphorylation of SHR hearts should be more robust than WKY hearts. A kinetic control by O2 supply, on the other hand, is conspicuous (Fig. 4A). Plotting [O2] in terms of %MbO2 linearizes the hyperbolic relation. Therefore, the linear relationship between MO2 and %MbO2 signifies a Michaelis-Menten-type kinetic control by the substrate concentration. Our results seem clear: Mitochondrial respiration in postischemic SHR heart is lower because [O2] is lower.
Nevertheless, the rate of reoxygenation, i.e., how fast O2 returns to the ischemic myocardium, is not likely a limiting factor. Although some reports (9) suggested that a delay in O2 return caused by elongated diffusion distance may contribute to postischemic dysfunction in LVH, our data show that MbO2 recovery is completed during the first postreflow 1H NMR acquisition. Given 2.5-min 1H NMR signal averaging and 10% error, if a linear MbO2 increase is assumed, cellular O2 would approach a new steady state within 30 s after reperfusion is commenced. In similar reasoning, O2 depletion during ischemia is also expected to be complete within 30 s. The latter estimation agrees well with the optically measured PO2 decline rate recently reported on ischemic mouse skeletal muscle (33).
In contrast to rapid O2 return, however, the steady-state RPP is not attained until 20–25 min after reperfusion (Fig. 3B). In the current study, ventricular fibrillation and tachycardia, which were resolved faster in SHR than WKY rats, characterized the period immediately following the reflow. Even after the sinus rhythm returned, RPP did not instantaneously return to the final level but gradually increased for another 10–15 min. The prevalence of postreflow arrhythmia and delayed RPP recovery has been a typical observation in many ischemia studies (2, 8, 53, 54).
The disparate recovery time courses between RPP and O2 implicate the role of O2 as necessary but not sufficient for the postischemic functional recovery. The role of O2 may actually change over the course of the recovery process. In the early reperfusion period, the presence of O2 is not enough to eliminate electrophysiological disturbances of the postischemic hearts. During 10–25 min of reperfusion, a degree of coupling inefficiency between oxidative metabolism and contractile function persists (Fig. 3, A and B). Finally, the steady-state RPP correlates well with the steady-state MO2 and %MbO2 (Figs. 3B and 4A), suggesting that the ultimate level of functional recovery is limited by O2 availability.
Therefore, the postischemic functional recovery process consists of two distinct phases, an O2-independent transient phase and an O2-dependent steady phase. The dissociation of contractile force from [O2], observed during the transient phase, leads to the question of whether the return of cellular PO2 coincides with the return of O2 consumption. The PO2 and MO2 time courses (Fig. 3, A and B) show that it does. Whether mitochondrial ATP synthesis is uncoupled from MO2 is difficult to judge from the current data. The probability, however, seems low; other researchers reported a rapid CrP recovery (2, 9, 54, 56) or an unchanged P/O ratio (46) under similar experimental conditions. Thus the dissociation seems to occur along the steps between ATP synthesis and muscle contraction, which apparently is mediated by factor(s) other than O2.
Origin of O2 Limitation in Reperfused SHR Hearts
As discussed in introduction, an elongation of O2 diffusion distance in LVH can limit O2 supply. Such O2 limitation, however, was not observed under nonischemic (baseline and hypoxic) conditions. Therefore, the anatomical change, i.e., myocyte hypertrophy, is not likely the underlying mechanism for postischemic O2 deficit; cell size would not change selectively in SHR after ischemia. Although postischemic edema can potentially increase the physical dimensions of myocyte and interstitial space, there is no evidence that edema was more severe in SHR hearts, because the wet-to-dry heart weight ratio was virtually identical between SHR and WKY hearts (Table 1).
If it is not the diffusion limitation imposed by myocyte hypertrophy, is it then possible that SHR hypoxygenation is vascular in origin? Reversible or irreversible circulatory disturbances are frequently observed in the postischemic myocardium (9, 27, 29). Indeed, coronary flow decreased after ischemia in both SHR and WKY hearts. Yet, curiously enough, relative flow recovery in SHR is almost comparable to WKY (only 5% difference). Consequently, postischemic O2 level in SHR is disproportionately lower than what can be accounted for by flow reduction (Fig. 3A). Even constant-flow reperfusion (Fig. 4A), which significantly improved O2/RPP, failed to abolish the difference between SHR and WKY recoveries. Therefore, the postischemic underoxygenation of SHR hearts cannot be fully explained by the flow per se.
The quantitative mismatch between reperfusion flow and O2 suggests the possibility that the bulk flow is shunted away from the nutritive microvessels. Exacerbated diastolic dysfunction in SHR hearts may play a role here. In the isovolumically contracting hearts, as in the present study, the increased EDP is a direct manifestation of decreased diastolic chamber distensibility and increased wall stress, which should raise extravascular compressive forces (58). It is then conceivable that the elevated tissue pressure in SHR promotes channeling the flow through the arteriovenous shunts within epicardial conduit vessels, bypassing the compressed intramyocardial vessels. The apparent "dumping" of O2 into the venous drainage is also reflected in the significantly increased venous PO2, shown in SHR during reperfusion (Table 3).
Because the subendocardial layer of myocardium experiences the highest tissue pressure during ventricular contraction and possibly during the diastolic phase as well (13), if increased extravascular compressive force is the mechanism, hypoxic myocardial regions are likely to concentrate in the subendocardium. Indeed, it has been repeatedly documented that the subendocardium is functionally and energetically the most vulnerable myocardial region in LVH. The clue for regionally heterogeneous O2 reperfusion can be found in the line width of Pi resonance. After ischemia, SHR Pi peak broadens 2.7 times more than WKY, indicating more uneven pH distribution. Because it is expected that underoxygenated cells acidify, pH heterogeneity implies a presence of O2 heterogeneity.
In conclusion, the compensated LVH heart is well oxygenated during baseline conditions. Yet it is markedly underoxygenated during postischemic reperfusion, the level of which is critically related to the depressed functional recovery. The role of reperfused O2, however, is time dependent. During early reperfusion, factor(s) other than O2 appear to limit functional recovery. It is when the mechanical function of the heart approaches a new steady state that O2 becomes a dominant factor. Meanwhile, the finding of a normal baseline O2 level in SHR hearts indicates that preischemic hypoxia is not required to provoke postischemic dysfunction in LVH.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
improves postischemic recovery of hypertrophied hearts. Circulation 104, Suppl I: I350–I355, 2001.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
and dotted line; n = 16) and Wistar-Kyoto (WKY) (
and solid line; n = 18) hearts were subjected to varying degrees of hypoxia.
, SHR; 
, WKY) nor O2 fully recovers. Note that SHR O2 recovery is out of proportion to flow recovery because SHR postischemic flow is only 