Heart energetic efficiency in O,-exposed rats studied in isolated working heart R. ARIELI,
S. A. BEN-HAIM,
G. HAYAM,
AND
Y. EDOUTE
Israeli Naval Hyperbaric Institute, Haifa 31080; and Faculty of Medicine, Cardiovascular Research Group and The Rappaport Family Institute for Research in the Medical Sciences, Technion-Israel, Institute of Technology, Haifa 31096, Israel ARIELI, R., S. A. BEN-HAIM, G. HAYAM, AND Y. EDOUTE. Heart energetic efficiency in C&-exposed ruts studied in isolated working heart. J. Appl. Physiol. ?3(6): 2289-2296,1992.-Death in normobaric hyperoxia was related in the past to pulmonary insufficiency of the edematouslung. However, high arterial 0, tension on final collapseled to the suggestionthat the heart and not the lung is the first organ that fails. We measured aortic flow, coronary flow, left ventricular pressure, affluent and effluent PO,, Pco~, and pH in the working heart excised from control and normobaric O,-exposed rats (51-63 h). The oxygen consumption (TO,) of experimental hearts was not different from control, but mechanical power output (PVAP) (calculated from pressure-volumearea) was reducedasa function of 0, exposure time. Myocardial contractility indexes, maximal elastance and maximal time derivative of pressure, increasedas a function of 0, exposure time, being below control values after 50 h and above control values after 60 h. The individual slopesfor the regressionof ACIDvs. PVAP rose as a function of exposuretime from values below control after 50 h exposure to valyes above control after 60 h. Energetic efficiency (PVAPIVo,) decreasedas a function of 0, exposure time and points to possibleheart failure in the intact animal. After 50 h 0, exposure the heart was energetically more efficient than the control. Possiblechangesin the heart are discussed.
of peripheral perfusion. Similarily, Harabin and Farhi (10) showed in awake rabbits exposed to normobaric hyperoxia that Pao, was high at death and that metabolic acidosis preceded the precipitous drop in PO,. They suggested cardiac insufficiency as a possible main effect. A number of studies have investigated the physiology of the cardiovascular system during prolonged 0, exposure. Some have used the intact animal, in which the specific effect on the heart cannot be distinguished from other systemic effects and effects on the cardiovascular control system (1, 10, 11,18,23). Other postmortem studies have analyzed perfusion distribution, biochemical products (2, 12), and tissue ultrastructure (3,4, 15), but direct inference to the work of the heart is difficult. Effects of 0, toxicity on the heart can be assessed without the complication of accompanying systemic effects by using an isolated heart preparation. In the present study, we exposed rats to normobaric hyperoxia for 51-63 h, a period known to produce pulmonary symptoms: the 50% survival time is 76 h (5). At the end of the exposure, we studied heart function in the isolated working heart preparation. METHODS
coronary resistance; cardiac contractility; oxygen consumpAnimals tion; output power; oxygen toxicity
Twenty-five hearts from male White Sprague-Dawley rats weighing 250-350 g were studied. ARE TWO GENERAL, FORMS OF fatal 0, toxicity: “central nervous system 0, toxicity” at a high PO, and “pulmonary 0, toxicity” at a moderate PO,. The final collapse and ensuing death in pulmonary 0, toxicity has been related in the past to the “development of pulmonary 0, intoxication before other vital organs are fatally affected” (6). However, some recent studies on awake animals, in contrast to anesthetized animals, have cast doubt, on the lung as the organ leading to the final collapse. Matalon et al. (19) found that in awake sheep exposed to normobaric hyperoxia the arterial PO, (Pa,,) was 200 Torr at death. Thus, although the inspired-arterial 0, difference increased as the hyperoxic exposure progressed, the final Pa,, cannot be considered as hypoxic. Harabin et al. (11) showed in awake dogs exposed to nurmobaric hyperoxia that hemodynamic instability preceded the precipitous change in blood gas tension. They related the metabolic acidosis preceding the fall in Pa,, to the reduction in cardiac output and redistribution THERE
016f-7567/92
$2.00
Copyright
Working Heart Experimental system. Rats were anesthetized by an intraperitoneal injection of methoexitone sodium (30 mgl kg). After removal, the heart was perfused by the Langendorff technique in a nonrecirculating mode at a constant pressure of 80 cmH,O for 6 min. During this time, the left atrium was cannulated to permit atria1 perfusion (atria1 pressure = 14 cmH,O), according to a previously described modified working heart model (25). The right ventricle was then cannulated via the pulmonary artery to anaerobically sample coronary effluent (21) for effluent pH, PCO, (Pef,,,), and PO, (Pef,,). For the first 2 min, the left ventricle (LV) ejected against a constant pressure of 80 cmH,O and thereafter against a Starling resistor, in which the surrounding pressure (70-100 Torr) and the tangential stress of a collapsible latex tube had been adjusted to produce an aortic flow of -15 ml/ min (within the physiological range). Once a surrounding
0 1992 the American
Physiological
Society
2289
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
&ml
EXPOSURE:
Ved
FIG. 1. A reconstructed pressure-volume loop. OACDE is pressurevolume loop of an ejecting beat. Line OB is following occluded beat. Line VoB is end-systolic pressure-volume relationship (ESPVR). LVP, left ventricular pressure; V, volume. Important points are 0, end of diastole; A, onset of ejection; B, maximal pressure of occluded beat; C, end systole of ejecting beat; D, end of ejection; E, end of isovolumetric relaxation; Vo, volume-axis intercept of ESPVR line; and Ved, assumed constant end-diastolic volume.
pressure had been established, it remained fixed throughout the test regardless of the aortic flow. Values are means t SD. The average surrounding pressure applied was 79 t 9 Torr. Perfusate temperature was maintained at 37.5*C. The perfusion medium was KrebsHenseleit bicarbonate buffer containing (in mmol/l) 110.0 Na+, 5.0 K+, 2.6 Ca”, 1.2 Mg2+, 1.2 H,PO,, 129.0 Cl-, 25.0 HCO,, 1.2 SO:-, and 10.0 glucose, equilibrated with 95% O,-5% CO, at atmospheric pressure. LV and aortic pressures were measured by use of a narrow polyethylene catheter (0.4 mm) and pressure transducers (Gould P23). Aortic flow was measured by an electromagnetic flowmeter (model MVF-2100, Nihon Kohden) placed 1 cm above the aortic valve. Aortic flow and LV and aortic pressures were recorded on a Graphtec paper recorder, digitized with the use of a 12-bit 500-Hz analogto-digital converter (LabMaster, Tecmar), and stored for later processing. Coronary flow was measured using a stopwatch and a calibrated cylinder. Affluent and coronary effluent pH, Pco,, and PO, were determined using a Radiometer blood-gas analyzer. Signal processing. A detailed description of the signalprocessing algorithm has recently been published (25). Our method for deriving the mechanical indexes is similar to that reported by Igarashi and Suga (16) and will be briefly described here. The following time indexes were used for determining the principal pressure and flow values: time of end diastole, marks the onset of isovolumic contraction (Fig. 1, point 0); ejection onset time, defined as the first moment after which four consecutive flow values were positive (Fig. 1, point A); end ejection time, defined as the first moment after ejection onset time when the flow signal became negative (Fig. 1, point D); and time of end-isovolumetric relaxation, defined as the moment when left ventricular pressure reached a minimal value (Fig. 1, point E). The aortic flow signal was used to determine the peak
WORKING
HEART
aortic flow and to calculate blood volume changes during the ejection period. Left ventricular volume change was calculated by integration of aortic flow between ejection onset time and end of ejection time. Assuming the left ventricular end-diastolic volume does not change during steady-state operation, we assigned the left ventricular end-diastolic volume a unit value and calculated a relative volume change of the LV. Indexes calculated by use of the left ventricular pressure signal were mean pressure, end-systolic pressure, end-diastolic pressure, and maximal time derivative of left ventricular pressure (dP&na,). The systolic and the isovolumetric relaxation parts of the pressure-volume (PV) curve were reconstructed using the pressure and volume signals synchronized according to the above-mentioned time indexes. An approximated line connected the end-isovolumetric relaxation point to the end-diastolic point to yield a complete PV loop (Fig. 1). After four stable ejecting heart beats, we suddenly occluded the aorta, creating a single isovolumetric contraction. Because both the ejecting beat and the isovolumetric beat had similar end-diastolic volumes (stable action assumption: at steady state the measured parameters of the heart are similar for every beat), we overlaid the PV relationship of the occluded beat on the PV relationship of the ejecting beat (Fig. 1, line OB). The end-systolic PV relationship line (Fig. 1, line VoB) was obtained using a normal beat as well as the occluded beat. The end-systolic PV relationship’s slope is defined as the maximal elastance (E,,,). From the reconstructed PV loop we derived external work, as the area circumscribed by the loop, and total mechanical work output = pressure-volume area (PVA) (22, 28). Because heart frequency was not controlled, both estimates of work were multiplied by heart frequency to obtain external power (EP) and total mechanical power output (PVAP). Calculations. Oxygen consumption (VOJ was calculated from the affluent-effluent PO, difference, the solubility of 0, in the perfusion medium, and coronary flow. Coronary resistance was calculated by dividing the surrounding pressure (afterload) by coronary flow. The different energy variables (Vo2, EP, and PVAP) and coronary flow (and as a result coronary resistance) were normalized by dividing them by LV weight. The concentration of H+ ([H+]) in the perfusate was calculated from the pH. ProtocoL. Each heart was tested under three sets of conditions: two different work loads and hypoxia. This protocol was chosen to investigate whether hyperoxic exposure has different effects on the heart’s performance with regard to work load or hypoxia. Neely et al. (21) showed that the mechanical power of the isolated working heart increased by 57% when preload was increased from 10 to 20 cmH,O. Because of technical limitations, we selected 14 and 19 cmH,O preloads to obtain different work loads. Hypoxia was produced by equilibrating the perfusion medium with a mixture of 02, Nzt and CO, to yield an affluent PO, (Paf,,) of -300 Torr. Because of the low 0, solubility of the perfusate, the concentration of dissolved 0, at 300 Torr limits the VO, of the working heart compared with PO, >400 Torr (8, 9) and was further defined as hypoxia. Thus, the three sets of condi-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSURE:
tions were high O,-14 cmH,O, high O,-19 cmH,O, and The different PO, and PCO, levels in the perfusion medium were obtained by using two gas mixing pumps (H. Wosthoff). When the heart stabilized, we sampled LV pressure and aortic flow at 14 cmH,O preload. Immediately afterward we sampled the affluent and effluent medium for gases and pH and measured coronary flow. A second sample was taken under the same conditions. The first two samples were taken at 11 and 18 min of the working heart preparation. The preload was then raised to 19 cmH,O, and a further two samples were taken at 27 and 33 min of the working heart preparation. At the end of the fourth sample the preload was changed back to 14 cmH,O and the equilibrating gases were switched to the hypoxic mixture. The fifth and sixth samples were taken in hypoxia at 43 and 48 min of the working heart preparation. After the last sample, the heart was removed, the LV was separated, excess water absorbed, and the LV was weighed, dried at 105*C overnight, and reweighed to determine water content.
WORKING
2291
HEART
hypoxia-14 cmH,O.
150
0
l l
high 02 -14cm
a
l
l
0
E L? 15op %
high
02019cm
poo
ls
t
5oLL
=hypoxia-14cm
a
Experimental Protocol
The experimental protocol was approved by the local committee. Groups of three to five experimental rats were placed in a 12-liter sealed Plexiglas chamber, through which pure 0, was pumped at a rate of 1.5 l/min. Replacement of the chamber gas in -8 min kept the chamber’s atmosphere dry. Chamber temperature was monitored and was controlled at 25-30°C. Soda lime was placed on the floor of the chamber to absorb CO,, and water and food were supplied ad libitum. After predetermined periods (between 51 and 63 h) rats were taken one by one for the isolated working heart preparation. The chamber was flushed with 0, to restore the 0, atmosphere for the remaining rats. Exposure time was marked for each rat. Although the longest exposure was 13 h below the time for 50% mortality (76 h) (5), we lost 10 animals during or immediately after the exposure. Hearts for control measurements (n = 8) were taken from rats from the same population at random order between the experimental hearts. Statistics
Two-variable analysis of variance with repeated measurement was used to compare the different physiological variables for control and experimental hearts and either three test conditions (high O,-14 cm, high O,-19 cm, and hypoxia-14 cm) or six measurements. The Pearson correlation test was applied to the different measured parameters and the 0, exposure time for experimental hearts. Multiple linear regression was used to test differences in the slope or intercept of VU, as a function of PVA power in experimental and control hearts. The SAS Institute (Cary, NC) statistical package was used for the calculations. RESULTS
Perfusate equilibration produced at high O,-14 cmH,O preload Pafoz = 463 t 27 Torr, affluent PCO, (Paf,,,) = 29.7 t 1.8 Torr, and affluent pH (pH,,) = 7.33 t 0.04.
I ’ c
;I” 50
I
I
1
1
52
54
56
58
A 60
A 62
O2 EXPOSURE TIMEA 2. Oxygen consumption (W f heart plotted as a function of exposu .re time of rat. Results from 3”sets of test con ditions (level of 0; aid preload pressure in cmH,O) are shown in 3 panels. Each dot represents 1 heart, and 8 control hearts are shown above C as means IL SD. FIG.
The different gaseous mixtures were bubbled through the same perfusate container, and at high O,-19 cmH,O preload, affluent oxygen was somewhat higher than it had been previously because of slow equilibration of the gaseous oxygen: Pafoz = 522 t 26 Torr, Pafoq, = 31.4 t 1.1 Torr, and pH,, = 7.36 t 0.01. Equilibration of the hypoxic mixture gave Paf,, = 321 t 13 Torr, Pafcoz = 35.6 t 2.4 Torr, and pH,, = 7.30 t 0.04. The lungs of the experimental rats were edematous, with fluid in the intrapleural space. However, the rats were not observed to be cyanotic. We took care to anesthetize rats immediately on removal from the chamber for fast excision of the heart before hypoxia developed in the intact animal in normoxic environment. Despite the edematous lungs in the experimental rats, there was no difference in the water content (84 t 2%) of the LV in control and experimental hearts at the end of the test on the working heart. VO, did not differ in control and experimental hearts (Fig. 2), nor was it related to exposure time in experimental hearts; but VO, was affected (P < 0.0001) by test conditions. As expected, increasing the preload resulted in an increase of 18% in VCI? (Fig. 2, middle). Hypoxia caused a 50% reduction in VO,. EP in the control hearts was 0.41 t 0.23, 0.47 t 0.23, and 0.20 t 0.13 J. min-1 g-l for high O,-14 cm, high O,l
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSI
RE: WORKING
high 02 -f&m
l
high 02’19cm a l l
.
l l l a
l
l 0 l
hypoxia44cm
c’
-50
52
54
O2 EXPOSURE
56
58
60
62
HEART
and mean systolic pressures were reduced in hypoxia (P < 0.0001). There was no difference in heart frequency for control and experimental hearts. The control heart frequency was 218 t 27,230 t 58, and 183 t 64 beats/min for high O,-14 cm, high O,-19 cm, and hypoxia-14 cm, respectively. The heart frequency was reduced in hypoxia (P < 0.0001). Coronary resistance was not affected by 0, exposure time, and there was no difference in coronary resistance for control and experimental hearts (Fig. 6). Hypoxia caused elevation of coronary resistance (P < 0.004). The coronary resistance results remained unchanged when it was calculated using aortic pressure during the nonejection phase. Because only at the beginning the aortic flow was adjusted and afterload pressure was fixed the cardiac output throughout the test could vary. In high 02, cardiac output decreased as a function of 0, exposure time (Fig. 7, P < 0.01). Cardiac output increased when preload pressure was increased from 14 to 19 cmH,O and decreased in hypoxia (P < 0.0001). The control stroke volume was 0.20 t 0.05, 0.22 t 0.05, and 0.14 t 0.07 ml for high O,-14 cm, high O,-19 cm, and hypoxia-14 cm, respectively. The hypoxic stroke volume was significantly lower than stroke volume in high 0, (P < 0.0001). The trend for
TIMLh
3. Total mechanical power output [pressure-volume area X frequency (PVAP)] plotted as a function of 0, exposure time. See explanation in Fig. 2. FIG.
19 cm, and hypoxia-14 cm, respectively. The EP and the PVAP (Fig. 3) decreased as a function of 0, exposure time at the high 0, levels (P < 0.01 and P c 0.0001, respectively). Mechanical power output (PVAP) after 50 h exposure was higher than control PVAP; after 60 h exposure, it was lower than the control value. Both EP and PVAP were affected by test conditions (P < 0.0002 and P < 0.0001, respectively). There was no difference in 0, extraction [(Paf,, PefoJ/PafoJ for control and experimental hearts, and only 0, extraction at high O,-19 cm was significantly lower than at high O,-14 cm. This may be due to the higher Pafog at high O,-19 cm for the PO, range that does not limit VOW (8). Effluent-affluent differences of PCO, and [H+] are shown in Fig. 4 as a function of measurement number. There was no difference between contpol and experimental hearts for both variables, but they were affected by test conditions (P < 0.03 and P < 0.02, respectively). The release of H+ into the myocardial circulation was greater in the first sample than in the following two to five samples, but an increase in Pef,,,-Paf,,, was found only in the second hypoxic sample (measurement 6). Peak systolic pressure is shown in Fig. 5. Control mean systolic pressure was 204 t 78, 197 t 65, and 162 t 73 Torr for high O,-14 cm, high O,-19 cm, and hypoxia-14 cm, respectively. Both systolic pressure parameters increase as a function of 0, exposure time in high O,-14 and -19 cm (P < 0.03 and P < 0.05, respectively). Peak
control hyperoxia
0 l
t: 0 Y *t-l
0
0 p4
c
measurement
number
Effluent and affluent differences in H+ concentration ([H+]) and in PCO~ plotted as a function of measurement number. Means 5 SD are shown for control and experimental hearts (hyperoxia). Test conditions are marked below measurement numbers. FIG.
4.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSURE:
250
l *
200
hypoxia-14cm
a
t
100
. 1 54 O2 EXPOSURE
5. Peak exposure time. FIG.
, 0 , 56 58
I 6;
1 62
TIME&
systolic pressure (PSP) plotted See explanation in Fig. 2.
as a function
of 0,
WORKING
2293
HEART
was markedly different in the first sample was the effluent pH (Fig. 2), which indicates an acidotic heart at the outset. However, this acidity was not related to the hyperoxic exposure and was probably produced during preparation of the heart. Other studies have shown focal necrosis in the hyperoxic heart (3,4,15), which should be reflected in effects of long duration (>ll min). However, the possibility still exists that effects of very short duration, which recover or are “washed out” in the perfused heart, add to the intoxication of the O,-exposed heart in the intact animal. Variation in preload at high 0, affected EP, VQ, and cardiac output, and hypoxia affected all measured parameters except E,,,. Hyperoxic exposure of the intact animal affected PVAP, EP, systolic pressure, stroke volume, and cardiac output. The tendency for coronary resistance to decrease after 60 h exposure (increased coronary flow) proves that coronary flow was not the cause of the reduction in mechanical power output after prolonged 0, exposure. An important finding of this study is the reduction in the mechanical power output of the heart as q2 exposure progresses, with no effect of 0, exposure on VO, (Figs. 2 and 3). Bergo et al. (1) calculated a pump work index (PWI = systolic blood pressure X heart frequency) for rats exposed to 5 bar 0, and found no change after 60 min of exposure. We calculated the PWI for the isolated hearts and found high variability and no correlation to exposure time in contrast to the PVAP (Fig. 3). It is accepted that PWI correlates with VO,, but in our experi-
reduced stroke volume after prolonged 0, exposure was not significant. The two estimates of cardiac contractility, E,, and are shown in Figs. 8 and 9, respectively, as a dPldt,,,, function of 0, exposure duration Test conditions affected dP/d&, which was reduced by hypoxia (P < O.OOOl), but did not affect E,,. Both estimates increased as a function of exposure time (p < 0.0001 and P < 0.006, respectively). After 50 h exposure, both estimates for contractility were lower than the contractility of control hearts (an exception to this being hypoxic dPldt,,,), and after 60 h exposure, both estimates were higher than the control value. DISCUSSION
In the present experiment, the first sample was taken when the isolated heart had been working for 11 min. Therefore toxic effects of 0, with either a short recovery time (41 min) or the effects of toxic substances, which can be washed out by perfusion, would not be measured by the present experimental system. An example of this is the fast inhibition of rat heart glyceraldehyde-3-phosphatase dehydrogenase (lo-15 min in 1 ATA 0,), which is reversible (13). If such a short-lasting effect had a recovery time close to 11 min, this effect would be apparent in the first but not in subsequent samples. Huwever, there was no variable that was different in the first and not in subsequent samples and was also different for experimental and control hearts. The only parameter that
8I
hypoxia-14cm I
ii
u
c
'SO O2
52 EXPOSURE
54
56
60
62
TIMEA
6. Coronary resistance (CR) plotted sure time. See explanation in Fig. 2. FIG.
58
as a function
of 0, expo-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
2294
OXYGEN
EXPOSURE:
l
20
b4 .
.
a a
%
a a
30
50
c
52
54
56
58
60
WORKING
HEART
Coronary blood flow is known to be reduced in hyperoxia in the intact animal (19). We have shown that coronary resistance was not affected after the 0, exposure and there was a tendency for coronary resistance to decrease after 60 h exposure (Fig. 6). Thus reduced coronary flow in the intact animal can be related to reduced cardiac output (1, lo), increased blood viscosity, decreased LV tension (7), and contraction of precapillary sphincters (19). A puzzling but consistent finding was the biphasic effect of 0, exposure. After 50 h of 0, exposure, mechanical power output and efficiency were higher than in control hearts (Figs. 3 and lo), and only after 60 h of exposure did the heart become energetically inferior to the control. Another puzzling finding was increased contractility as 0, exposure progressed (Figs. 8 and 9). Cardiac contractility was also biphasic with respect to control hearts. There is only limited information on the toxic effect of 0, on cardiac contractility. Reduction of isometric systolic tension has been observed in the anesthetized dog transferred from 25% 0, to 100% 0, (7). Although Daniel1 and Bagwell (7) relate this reduction to the initial manifestation of 0, toxicity, the duration of the exposure in their study was only 30 min. Hyperbaric oxygen (3.6 ATA for 15 min) caused decreased myocardial contractility (dP/dt,,) (18). These two reports can be correlated with the initial reduction in both estimates of cardiac contractility found in the present study. However, the
62
O2 EXPOSURE TIME,fi FIG. 7. Cardiac output (CO) plotted as a function of 0, exposure time. See explanation in Fig. 2.
mental system it did not convey the mechanical power output clearly. We used multiple regression (using all measurements from the 25 hearts) of Vo,lbeat against total PVA/beat and dummy variables (assigning 0 to control and I to experimental hearts) to find out whether there were differences in either intercept or slope for control and experimental hearts. Both VO, and PVA were given in units of joules per beat per gram. The results showed no difference in intercept but a significant difference in slope (P < 0.001) for control and experimental hearts. We therefore calculated the slope for each heart, using linear regression of its six meisurements. These slopes, which represent the reciprocal work efficiency of the heart, are shown as a function of 0, exposure time in Fig. 10. The slope of the control hearts (2.1) is very similar to that found by Suga et al. (26) for puppies’ hearts (2.2). The calculated intercept for control hearts (0.007 J beat-l g-l) is also similar to that of the puppies’ hearts (0.005 J *beat-l* g-l). The work efficiency of the heart deteriorates as the hyperoxic exposure progresses and would obviously lead to death of the intact animal. These physiological findings may correlate with reported microscopic observations. Busing et al. (3) found in the rabbit exposed to I ATA 0, for 40 h that there is advanced disintegration of the heart myofibrils, whereas the mitochondria remain intact. Hughson et al. (15) reported initial mitochondrial swelling in 5 ATA O,exposed rats that preceded local dissolution of the myofibrillar structure. l
+
a
high
a
l
@-l&n l
~1000
;
l9
I,..,
15oot
2 y
l
+ 0
*
#
l l
ti
500
z t cr;t I 20oot
.
l
1
a
l
l l
fl
l
0
!’
I
I
1
1
1
I 0
hypoxia-14cm
l
0 1OOO
l 0
0
l
l
a
500
_e 7
,'50
*ta52
l
54)
l
&,
l
58,
l
60,
62,
,
O2 EXPOSURE TIME,h FIG. 8. Maximal elastance (E,,,) plotted as a function of 0, exposure time. See explanation in Fig. 2.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSURE:
8000
WORKING
2295
HEART
06.
6000
a
04.
$8000 ii E t
0
0 00 ’
high 02-19cm
* (I,
$6000 PI a 4000
0 l
l
0
0
m
O
l
0
0 @
2000 ti
0
high 02’19cm
l
a
I
control hyperoxia
high 02-14cm
5
0
l
OO
l
*4
O & 0
0 l
8000 6000
hypoxia-14cm
t
hypoxia-14cm
e l
*a
so-@
t l
l
a 1
I
1
c'
i
50’
O2
02.
a l
8
52
54I
56 ,
EXPOSURE
l
a
I 62
,
58
.
60 I
0
400
Emax,
TIME,h
l
0
4% 800
0 0
0
1200
torrpn1
plotted as a
FIG. 11. Energy efficiency (PVAP/vo, in J/min) plotted against cardiac contractility (E,,,) for control and experimental hearts (hyperoxia).
mechanism of the later increase in these two estimates is not clear. Decreased work efficiency has been measured and predicted in different situations when E,,, increases (22,27, 28). Our experimental apparatus and results are not the same as in these previous studies, because neither the EP nor the stroke volume was constant. However, we tested the relationship between cardiac contractility and work efficiency (PVAP/%,), as shown in Fig. 11. It seems that no clear relationship exists between E,,, and efficiency in the control hearts. However, there is a clear relationship between E,,, and work efficiency in the hearts of O,-exposed rats: work efficiency decreases as E,,, increases. What might be the cause of the tight relation-
ship of E,,, and efficiency of hearts from O,-exposed rats compared with that of the control hearts? One possibility is changes in muscle myosin isoenzymes during 0, exposure, as has been found in other types of stress: heat, exercise, thyroid state, and ventricular pressure overload (14, 17, 24). In that case the turnover time should be shortened because of 0, exposure. Another possible stress-related changes in the heart are density and affinity of P-adrenoreceptors. It is possible that 0, exposure made the heart more work efficient in the early stages of 0, exposure. This intriguing question is a topic for further study. Our findings on the effect 0, toxicity on the heart support earlier suggestions (10, 11,20) that the cardiovascular system is the system that led to the final collapse in pulmonary 0, toxicity.
FIG. 9. Maximal time derivative of pressure (dP/dt,,,) function of 0, exposure time. See explanation in Fig. 2.
i’oz=A
+BxPVAP
0
The authors thank R. Lincoln for skillful editing. This study was supported in part by the McRamer Foundation. The opinions and assertions made herein are the private ones of the authors and do not represent the official views of the Israeli Naval Hyperbaric Institute. Address for reprint requests: R. Arieli, Israeli Naval Hyperbaric Institute, PO Box 8040, Haifa 31080, Israel.
50
c! O2
l aI 52
EXPOSURE
Received 22 October 1990; accepted in final form 8 June 1992. I 54
l
I 56
1 58
I 60
1 62
TIME,h
FIG. 10. Slope (B) of linear regression of PVAP/beat against OO,/ beat (both in J beat?. g-l in equation) for individual hearts, plotted as a function of 0, exx)osure time. See exnlanation in Fig. 2. l
REFERENCES 1. BERGO, G. W., J. RISBERG, AND I. TY~SEBOTN. Effect of 5 bar oxygen on cardiac output and organ blood flow in conscious rats. Undersea Biomed. Res. 15: 457-470, 1988. 2. BLOUNT, D. H. Cardiac metabolic response to hyperbaric oxygen. Proc. Sot. EXD. BioL. 2Med. 133: 1129-1131, 1970.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
2296
OXYGEN
EXPOSURE:
3. Busr~c, C. M., U. KREENSEN, F. BUHLER, AND U. BLEYL. Light and electron microscopic examinations of experimentally produced heart muscle necroses following normobaric hyperoxia. Virchows Arch.
A Puthol.
Anat.
Histol.
366: 137-147,
1975.
4. CAULFIELD, J. B., R. W. SHELTON, AND J. F. BURKE. Cytotoxic effects of oxygen on striated muscle. Arch. Patho2. 94: 127-132, 1972. CLARK, J. M. Effects of acute and chronic hypercapnia on oxygen tolerance in rats. J. Appl. Physiol. 50: 1036-1044, 1981. CLARK, J. M., AND C. J. LAMBERT~EN. Pulmonary oxygen toxicity: a review. Pharmacok Rev. 23: 37-133, 1971. DANIELL, H. B., AND E. E. BAGWELL. Effects of high oxygen on coronary flow and heart force. Am. J. Physiol. 214: 1454-1459, 1968. 8. EDOUTE, Y., AND R, ARIELI. Effect of different degrees of hypoxia and reoxygenation on myocardial energetics. Isr. J. Med. Sci. 25: 3&2-3&8,19&9. 9. EDNJTE, Y., R. ARIELI, AND E. NEVO. Evidence for improved myocardial oxygen delivery and function during hypoxia in the mole rat. J. Camp. Physiol. B Biochem Syst. Environ. Physiul. 158: 5755&2,1988. IO. HARABIN, A. L., AND L, E. FARHI. Blood-gas transport in awake rabbits exposed to normobaric hyperoxia. Undersea Biomed. Res. L. D. HOMER, AND M. E. BRADLEY. Pulmonary oxygen toxicity in awake dogs: metabolic and physiological effects.
J. Appl. Physiol. 57: 1480-1488, 1984. 12. HAUGAARD, N., M. E. HESS, AND H. ITSK~VITZ.
The toxic action of oxygen on glucose and pyruvate oxidation in heart homogenates. J.
Biol. Chem. 227: 605-616,1957. 13. HORN, R. S., E. S. HAUGAARD, Biuchim. Biophys. 14. HOROWITZ, M.,
AND N. HAUGAARD. The mechanism of glycolysis by oxygen in rat heart homogenate. Acta 99: 549-552,
1965.
Y. M. PEYSER, AND A. MUHLRAD. Alterations in cardiac myosin isoenzymes distribution as an adaptation to chronic environmental heat stress in the rat. J. lMoZ. Cell. Cardiot. 18: 511-
515, 1986. 15. HUGHSON,
M., J. D. BALENTINE, AND H. B. DANIELL. The ultrastructural pathology of hyperbaric oxygen exposure: observations on the heart. Lab. Invest. 37: 516-525, 1977. 16. IGARASHI, Y., AND H. SUGA. Assessment of slope of end-systolic pressure-volume line of in situ dog heart. Am. J. Physiol. 250 (Heart
Circ. PhysioZ.
19): H6&5-H692,
1986.
HEART
S.-I., R. MATSUOKA, E. HIRATSUKA, M. KIMURA, T. AND A. TAKAO. Local response to cardiac overload on myosin heavy chain gene expression and isozyme transition. Circ.
17. IMAMURA. NISHIKAWA,
Res. 66: 1067-1073,199O. 18. KIUSCHOS, J. M., V. S. BEHAR, H. A. SALTZMAN, H. N. E. MYERS, W. W. SMITH, AND H. D. MCINTOSH.
perbaric oxygenation on left ventricular
K. THOMPSON, Effect of hy-
function. Am. J. Physiol. 216: 161-166, 1969. 19. KIRKMAN, A. D. Cardiovascular system and skeletal muscle. In: Pathology of Oxygen Toxicity, edited by J. D. Balentine. New York: Academic, 1982, p. 214-238. MATALON, S., M. S. NESARAJAH, AND L. E. FARHI. Pulmonary and circulatory changes in conscious sheep exposed to 100% O2 at 1 ATA. J. Appl. Physiok 53: 110-116, 1982. 21. NEELY, J. R., H. LIEBERMEISTER, E. J. BAWERSBY, AND H. E. MORGAN. Effect of pressure development on oxygen consumption by isolated rat heart. Am. J. Physiol. 212: 804-814, 1967. 22. NOZAWA, T., Y. YASUMURA, S, FUTAKI, N. TANAKA, Y. IGARASHI, Y. GOTO, AND H. SUGA. Relation between oxygen consumption and pressure-volume area of in situ dog heart. Am. J. Physiol. 253 (Heart Circ. Physiol. 22): H31-H40, 1987. 23. ONARHEIM, J., AND I. TYSSEBOTN. Effect of high ambient pressure and oxygen tension on organ blood flow in the anesthetized rat. Undersea
14: 133-147,19&7. 11. HARABIN, A. L.,
of the inhibition
WORKING
Biomed.
Res. 7: 47-60,
1980.
24. PAGANY, E. D., AND R. J. SOLARO. Swimming exercise, thyroid state, and the distribution of myosin isoenzymes in rat heart. Am. J. Physiol.
245 (Heart
Circ. Physiol.
14): H713-H720,
1983.
25. SHWARTZ,
L., S. A. BEN-HAIM, A. ELAMI, A. HASSAN, AND Y. EDOUTE, An algorithm for calculating left ventricular mechanical and energetic characteristics. Med. Biol. Eng. Comput. 29: 318-323, 1991.
26. SUGA, H., 0. YAMADA, Y. GOTO, Y. IGARASHI, Y. YASUMURA, T. NOZAWA, AND S. FUTAKI. Left ventricular 0, consumption and pressure-volume area in puppies. Am. J. Physiol. 253 (IIeart Circ. Physiol.
22): H770-H776,
1987.
27. SUGA, H., Y. YASUMURA, T. NOZAWA, S. FUTAKI, Y. IGARASHI, AND Y. GOTO. Prospective prediction of 0, consumption from pressurevolume area in dog hearts. Am. J. Physiol. 252 (Heart Circ. Physiol. 21): H1258-H1264,19&7. 28. TANAKA, N., Y. YASUMURA, T. NOZAWA, S. FUTAKI, M. UENISHI, K. HIRAMORI, AND H. SUGA. Optimal contractility and minimal oxygen consumption for constant external work of heart. Am. J. Physiol.
254
(Regulatory
Integrative
Comp.
Physiol.
23):
R933-
R943,198&.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
S. A. BEN-HAIM,
G. HAYAM,
AND
Y. EDOUTE
Israeli Naval Hyperbaric Institute, Haifa 31080; and Faculty of Medicine, Cardiovascular Research Group and The Rappaport Family Institute for Research in the Medical Sciences, Technion-Israel, Institute of Technology, Haifa 31096, Israel ARIELI, R., S. A. BEN-HAIM, G. HAYAM, AND Y. EDOUTE. Heart energetic efficiency in C&-exposed ruts studied in isolated working heart. J. Appl. Physiol. ?3(6): 2289-2296,1992.-Death in normobaric hyperoxia was related in the past to pulmonary insufficiency of the edematouslung. However, high arterial 0, tension on final collapseled to the suggestionthat the heart and not the lung is the first organ that fails. We measured aortic flow, coronary flow, left ventricular pressure, affluent and effluent PO,, Pco~, and pH in the working heart excised from control and normobaric O,-exposed rats (51-63 h). The oxygen consumption (TO,) of experimental hearts was not different from control, but mechanical power output (PVAP) (calculated from pressure-volumearea) was reducedasa function of 0, exposure time. Myocardial contractility indexes, maximal elastance and maximal time derivative of pressure, increasedas a function of 0, exposure time, being below control values after 50 h and above control values after 60 h. The individual slopesfor the regressionof ACIDvs. PVAP rose as a function of exposuretime from values below control after 50 h exposure to valyes above control after 60 h. Energetic efficiency (PVAPIVo,) decreasedas a function of 0, exposure time and points to possibleheart failure in the intact animal. After 50 h 0, exposure the heart was energetically more efficient than the control. Possiblechangesin the heart are discussed.
of peripheral perfusion. Similarily, Harabin and Farhi (10) showed in awake rabbits exposed to normobaric hyperoxia that Pao, was high at death and that metabolic acidosis preceded the precipitous drop in PO,. They suggested cardiac insufficiency as a possible main effect. A number of studies have investigated the physiology of the cardiovascular system during prolonged 0, exposure. Some have used the intact animal, in which the specific effect on the heart cannot be distinguished from other systemic effects and effects on the cardiovascular control system (1, 10, 11,18,23). Other postmortem studies have analyzed perfusion distribution, biochemical products (2, 12), and tissue ultrastructure (3,4, 15), but direct inference to the work of the heart is difficult. Effects of 0, toxicity on the heart can be assessed without the complication of accompanying systemic effects by using an isolated heart preparation. In the present study, we exposed rats to normobaric hyperoxia for 51-63 h, a period known to produce pulmonary symptoms: the 50% survival time is 76 h (5). At the end of the exposure, we studied heart function in the isolated working heart preparation. METHODS
coronary resistance; cardiac contractility; oxygen consumpAnimals tion; output power; oxygen toxicity
Twenty-five hearts from male White Sprague-Dawley rats weighing 250-350 g were studied. ARE TWO GENERAL, FORMS OF fatal 0, toxicity: “central nervous system 0, toxicity” at a high PO, and “pulmonary 0, toxicity” at a moderate PO,. The final collapse and ensuing death in pulmonary 0, toxicity has been related in the past to the “development of pulmonary 0, intoxication before other vital organs are fatally affected” (6). However, some recent studies on awake animals, in contrast to anesthetized animals, have cast doubt, on the lung as the organ leading to the final collapse. Matalon et al. (19) found that in awake sheep exposed to normobaric hyperoxia the arterial PO, (Pa,,) was 200 Torr at death. Thus, although the inspired-arterial 0, difference increased as the hyperoxic exposure progressed, the final Pa,, cannot be considered as hypoxic. Harabin et al. (11) showed in awake dogs exposed to nurmobaric hyperoxia that hemodynamic instability preceded the precipitous change in blood gas tension. They related the metabolic acidosis preceding the fall in Pa,, to the reduction in cardiac output and redistribution THERE
016f-7567/92
$2.00
Copyright
Working Heart Experimental system. Rats were anesthetized by an intraperitoneal injection of methoexitone sodium (30 mgl kg). After removal, the heart was perfused by the Langendorff technique in a nonrecirculating mode at a constant pressure of 80 cmH,O for 6 min. During this time, the left atrium was cannulated to permit atria1 perfusion (atria1 pressure = 14 cmH,O), according to a previously described modified working heart model (25). The right ventricle was then cannulated via the pulmonary artery to anaerobically sample coronary effluent (21) for effluent pH, PCO, (Pef,,,), and PO, (Pef,,). For the first 2 min, the left ventricle (LV) ejected against a constant pressure of 80 cmH,O and thereafter against a Starling resistor, in which the surrounding pressure (70-100 Torr) and the tangential stress of a collapsible latex tube had been adjusted to produce an aortic flow of -15 ml/ min (within the physiological range). Once a surrounding
0 1992 the American
Physiological
Society
2289
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
&ml
EXPOSURE:
Ved
FIG. 1. A reconstructed pressure-volume loop. OACDE is pressurevolume loop of an ejecting beat. Line OB is following occluded beat. Line VoB is end-systolic pressure-volume relationship (ESPVR). LVP, left ventricular pressure; V, volume. Important points are 0, end of diastole; A, onset of ejection; B, maximal pressure of occluded beat; C, end systole of ejecting beat; D, end of ejection; E, end of isovolumetric relaxation; Vo, volume-axis intercept of ESPVR line; and Ved, assumed constant end-diastolic volume.
pressure had been established, it remained fixed throughout the test regardless of the aortic flow. Values are means t SD. The average surrounding pressure applied was 79 t 9 Torr. Perfusate temperature was maintained at 37.5*C. The perfusion medium was KrebsHenseleit bicarbonate buffer containing (in mmol/l) 110.0 Na+, 5.0 K+, 2.6 Ca”, 1.2 Mg2+, 1.2 H,PO,, 129.0 Cl-, 25.0 HCO,, 1.2 SO:-, and 10.0 glucose, equilibrated with 95% O,-5% CO, at atmospheric pressure. LV and aortic pressures were measured by use of a narrow polyethylene catheter (0.4 mm) and pressure transducers (Gould P23). Aortic flow was measured by an electromagnetic flowmeter (model MVF-2100, Nihon Kohden) placed 1 cm above the aortic valve. Aortic flow and LV and aortic pressures were recorded on a Graphtec paper recorder, digitized with the use of a 12-bit 500-Hz analogto-digital converter (LabMaster, Tecmar), and stored for later processing. Coronary flow was measured using a stopwatch and a calibrated cylinder. Affluent and coronary effluent pH, Pco,, and PO, were determined using a Radiometer blood-gas analyzer. Signal processing. A detailed description of the signalprocessing algorithm has recently been published (25). Our method for deriving the mechanical indexes is similar to that reported by Igarashi and Suga (16) and will be briefly described here. The following time indexes were used for determining the principal pressure and flow values: time of end diastole, marks the onset of isovolumic contraction (Fig. 1, point 0); ejection onset time, defined as the first moment after which four consecutive flow values were positive (Fig. 1, point A); end ejection time, defined as the first moment after ejection onset time when the flow signal became negative (Fig. 1, point D); and time of end-isovolumetric relaxation, defined as the moment when left ventricular pressure reached a minimal value (Fig. 1, point E). The aortic flow signal was used to determine the peak
WORKING
HEART
aortic flow and to calculate blood volume changes during the ejection period. Left ventricular volume change was calculated by integration of aortic flow between ejection onset time and end of ejection time. Assuming the left ventricular end-diastolic volume does not change during steady-state operation, we assigned the left ventricular end-diastolic volume a unit value and calculated a relative volume change of the LV. Indexes calculated by use of the left ventricular pressure signal were mean pressure, end-systolic pressure, end-diastolic pressure, and maximal time derivative of left ventricular pressure (dP&na,). The systolic and the isovolumetric relaxation parts of the pressure-volume (PV) curve were reconstructed using the pressure and volume signals synchronized according to the above-mentioned time indexes. An approximated line connected the end-isovolumetric relaxation point to the end-diastolic point to yield a complete PV loop (Fig. 1). After four stable ejecting heart beats, we suddenly occluded the aorta, creating a single isovolumetric contraction. Because both the ejecting beat and the isovolumetric beat had similar end-diastolic volumes (stable action assumption: at steady state the measured parameters of the heart are similar for every beat), we overlaid the PV relationship of the occluded beat on the PV relationship of the ejecting beat (Fig. 1, line OB). The end-systolic PV relationship line (Fig. 1, line VoB) was obtained using a normal beat as well as the occluded beat. The end-systolic PV relationship’s slope is defined as the maximal elastance (E,,,). From the reconstructed PV loop we derived external work, as the area circumscribed by the loop, and total mechanical work output = pressure-volume area (PVA) (22, 28). Because heart frequency was not controlled, both estimates of work were multiplied by heart frequency to obtain external power (EP) and total mechanical power output (PVAP). Calculations. Oxygen consumption (VOJ was calculated from the affluent-effluent PO, difference, the solubility of 0, in the perfusion medium, and coronary flow. Coronary resistance was calculated by dividing the surrounding pressure (afterload) by coronary flow. The different energy variables (Vo2, EP, and PVAP) and coronary flow (and as a result coronary resistance) were normalized by dividing them by LV weight. The concentration of H+ ([H+]) in the perfusate was calculated from the pH. ProtocoL. Each heart was tested under three sets of conditions: two different work loads and hypoxia. This protocol was chosen to investigate whether hyperoxic exposure has different effects on the heart’s performance with regard to work load or hypoxia. Neely et al. (21) showed that the mechanical power of the isolated working heart increased by 57% when preload was increased from 10 to 20 cmH,O. Because of technical limitations, we selected 14 and 19 cmH,O preloads to obtain different work loads. Hypoxia was produced by equilibrating the perfusion medium with a mixture of 02, Nzt and CO, to yield an affluent PO, (Paf,,) of -300 Torr. Because of the low 0, solubility of the perfusate, the concentration of dissolved 0, at 300 Torr limits the VO, of the working heart compared with PO, >400 Torr (8, 9) and was further defined as hypoxia. Thus, the three sets of condi-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSURE:
tions were high O,-14 cmH,O, high O,-19 cmH,O, and The different PO, and PCO, levels in the perfusion medium were obtained by using two gas mixing pumps (H. Wosthoff). When the heart stabilized, we sampled LV pressure and aortic flow at 14 cmH,O preload. Immediately afterward we sampled the affluent and effluent medium for gases and pH and measured coronary flow. A second sample was taken under the same conditions. The first two samples were taken at 11 and 18 min of the working heart preparation. The preload was then raised to 19 cmH,O, and a further two samples were taken at 27 and 33 min of the working heart preparation. At the end of the fourth sample the preload was changed back to 14 cmH,O and the equilibrating gases were switched to the hypoxic mixture. The fifth and sixth samples were taken in hypoxia at 43 and 48 min of the working heart preparation. After the last sample, the heart was removed, the LV was separated, excess water absorbed, and the LV was weighed, dried at 105*C overnight, and reweighed to determine water content.
WORKING
2291
HEART
hypoxia-14 cmH,O.
150
0
l l
high 02 -14cm
a
l
l
0
E L? 15op %
high
02019cm
poo
ls
t
5oLL
=hypoxia-14cm
a
Experimental Protocol
The experimental protocol was approved by the local committee. Groups of three to five experimental rats were placed in a 12-liter sealed Plexiglas chamber, through which pure 0, was pumped at a rate of 1.5 l/min. Replacement of the chamber gas in -8 min kept the chamber’s atmosphere dry. Chamber temperature was monitored and was controlled at 25-30°C. Soda lime was placed on the floor of the chamber to absorb CO,, and water and food were supplied ad libitum. After predetermined periods (between 51 and 63 h) rats were taken one by one for the isolated working heart preparation. The chamber was flushed with 0, to restore the 0, atmosphere for the remaining rats. Exposure time was marked for each rat. Although the longest exposure was 13 h below the time for 50% mortality (76 h) (5), we lost 10 animals during or immediately after the exposure. Hearts for control measurements (n = 8) were taken from rats from the same population at random order between the experimental hearts. Statistics
Two-variable analysis of variance with repeated measurement was used to compare the different physiological variables for control and experimental hearts and either three test conditions (high O,-14 cm, high O,-19 cm, and hypoxia-14 cm) or six measurements. The Pearson correlation test was applied to the different measured parameters and the 0, exposure time for experimental hearts. Multiple linear regression was used to test differences in the slope or intercept of VU, as a function of PVA power in experimental and control hearts. The SAS Institute (Cary, NC) statistical package was used for the calculations. RESULTS
Perfusate equilibration produced at high O,-14 cmH,O preload Pafoz = 463 t 27 Torr, affluent PCO, (Paf,,,) = 29.7 t 1.8 Torr, and affluent pH (pH,,) = 7.33 t 0.04.
I ’ c
;I” 50
I
I
1
1
52
54
56
58
A 60
A 62
O2 EXPOSURE TIMEA 2. Oxygen consumption (W f heart plotted as a function of exposu .re time of rat. Results from 3”sets of test con ditions (level of 0; aid preload pressure in cmH,O) are shown in 3 panels. Each dot represents 1 heart, and 8 control hearts are shown above C as means IL SD. FIG.
The different gaseous mixtures were bubbled through the same perfusate container, and at high O,-19 cmH,O preload, affluent oxygen was somewhat higher than it had been previously because of slow equilibration of the gaseous oxygen: Pafoz = 522 t 26 Torr, Pafoq, = 31.4 t 1.1 Torr, and pH,, = 7.36 t 0.01. Equilibration of the hypoxic mixture gave Paf,, = 321 t 13 Torr, Pafcoz = 35.6 t 2.4 Torr, and pH,, = 7.30 t 0.04. The lungs of the experimental rats were edematous, with fluid in the intrapleural space. However, the rats were not observed to be cyanotic. We took care to anesthetize rats immediately on removal from the chamber for fast excision of the heart before hypoxia developed in the intact animal in normoxic environment. Despite the edematous lungs in the experimental rats, there was no difference in the water content (84 t 2%) of the LV in control and experimental hearts at the end of the test on the working heart. VO, did not differ in control and experimental hearts (Fig. 2), nor was it related to exposure time in experimental hearts; but VO, was affected (P < 0.0001) by test conditions. As expected, increasing the preload resulted in an increase of 18% in VCI? (Fig. 2, middle). Hypoxia caused a 50% reduction in VO,. EP in the control hearts was 0.41 t 0.23, 0.47 t 0.23, and 0.20 t 0.13 J. min-1 g-l for high O,-14 cm, high O,l
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSI
RE: WORKING
high 02 -f&m
l
high 02’19cm a l l
.
l l l a
l
l 0 l
hypoxia44cm
c’
-50
52
54
O2 EXPOSURE
56
58
60
62
HEART
and mean systolic pressures were reduced in hypoxia (P < 0.0001). There was no difference in heart frequency for control and experimental hearts. The control heart frequency was 218 t 27,230 t 58, and 183 t 64 beats/min for high O,-14 cm, high O,-19 cm, and hypoxia-14 cm, respectively. The heart frequency was reduced in hypoxia (P < 0.0001). Coronary resistance was not affected by 0, exposure time, and there was no difference in coronary resistance for control and experimental hearts (Fig. 6). Hypoxia caused elevation of coronary resistance (P < 0.004). The coronary resistance results remained unchanged when it was calculated using aortic pressure during the nonejection phase. Because only at the beginning the aortic flow was adjusted and afterload pressure was fixed the cardiac output throughout the test could vary. In high 02, cardiac output decreased as a function of 0, exposure time (Fig. 7, P < 0.01). Cardiac output increased when preload pressure was increased from 14 to 19 cmH,O and decreased in hypoxia (P < 0.0001). The control stroke volume was 0.20 t 0.05, 0.22 t 0.05, and 0.14 t 0.07 ml for high O,-14 cm, high O,-19 cm, and hypoxia-14 cm, respectively. The hypoxic stroke volume was significantly lower than stroke volume in high 0, (P < 0.0001). The trend for
TIMLh
3. Total mechanical power output [pressure-volume area X frequency (PVAP)] plotted as a function of 0, exposure time. See explanation in Fig. 2. FIG.
19 cm, and hypoxia-14 cm, respectively. The EP and the PVAP (Fig. 3) decreased as a function of 0, exposure time at the high 0, levels (P < 0.01 and P c 0.0001, respectively). Mechanical power output (PVAP) after 50 h exposure was higher than control PVAP; after 60 h exposure, it was lower than the control value. Both EP and PVAP were affected by test conditions (P < 0.0002 and P < 0.0001, respectively). There was no difference in 0, extraction [(Paf,, PefoJ/PafoJ for control and experimental hearts, and only 0, extraction at high O,-19 cm was significantly lower than at high O,-14 cm. This may be due to the higher Pafog at high O,-19 cm for the PO, range that does not limit VOW (8). Effluent-affluent differences of PCO, and [H+] are shown in Fig. 4 as a function of measurement number. There was no difference between contpol and experimental hearts for both variables, but they were affected by test conditions (P < 0.03 and P < 0.02, respectively). The release of H+ into the myocardial circulation was greater in the first sample than in the following two to five samples, but an increase in Pef,,,-Paf,,, was found only in the second hypoxic sample (measurement 6). Peak systolic pressure is shown in Fig. 5. Control mean systolic pressure was 204 t 78, 197 t 65, and 162 t 73 Torr for high O,-14 cm, high O,-19 cm, and hypoxia-14 cm, respectively. Both systolic pressure parameters increase as a function of 0, exposure time in high O,-14 and -19 cm (P < 0.03 and P < 0.05, respectively). Peak
control hyperoxia
0 l
t: 0 Y *t-l
0
0 p4
c
measurement
number
Effluent and affluent differences in H+ concentration ([H+]) and in PCO~ plotted as a function of measurement number. Means 5 SD are shown for control and experimental hearts (hyperoxia). Test conditions are marked below measurement numbers. FIG.
4.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSURE:
250
l *
200
hypoxia-14cm
a
t
100
. 1 54 O2 EXPOSURE
5. Peak exposure time. FIG.
, 0 , 56 58
I 6;
1 62
TIME&
systolic pressure (PSP) plotted See explanation in Fig. 2.
as a function
of 0,
WORKING
2293
HEART
was markedly different in the first sample was the effluent pH (Fig. 2), which indicates an acidotic heart at the outset. However, this acidity was not related to the hyperoxic exposure and was probably produced during preparation of the heart. Other studies have shown focal necrosis in the hyperoxic heart (3,4,15), which should be reflected in effects of long duration (>ll min). However, the possibility still exists that effects of very short duration, which recover or are “washed out” in the perfused heart, add to the intoxication of the O,-exposed heart in the intact animal. Variation in preload at high 0, affected EP, VQ, and cardiac output, and hypoxia affected all measured parameters except E,,,. Hyperoxic exposure of the intact animal affected PVAP, EP, systolic pressure, stroke volume, and cardiac output. The tendency for coronary resistance to decrease after 60 h exposure (increased coronary flow) proves that coronary flow was not the cause of the reduction in mechanical power output after prolonged 0, exposure. An important finding of this study is the reduction in the mechanical power output of the heart as q2 exposure progresses, with no effect of 0, exposure on VO, (Figs. 2 and 3). Bergo et al. (1) calculated a pump work index (PWI = systolic blood pressure X heart frequency) for rats exposed to 5 bar 0, and found no change after 60 min of exposure. We calculated the PWI for the isolated hearts and found high variability and no correlation to exposure time in contrast to the PVAP (Fig. 3). It is accepted that PWI correlates with VO,, but in our experi-
reduced stroke volume after prolonged 0, exposure was not significant. The two estimates of cardiac contractility, E,, and are shown in Figs. 8 and 9, respectively, as a dPldt,,,, function of 0, exposure duration Test conditions affected dP/d&, which was reduced by hypoxia (P < O.OOOl), but did not affect E,,. Both estimates increased as a function of exposure time (p < 0.0001 and P < 0.006, respectively). After 50 h exposure, both estimates for contractility were lower than the contractility of control hearts (an exception to this being hypoxic dPldt,,,), and after 60 h exposure, both estimates were higher than the control value. DISCUSSION
In the present experiment, the first sample was taken when the isolated heart had been working for 11 min. Therefore toxic effects of 0, with either a short recovery time (41 min) or the effects of toxic substances, which can be washed out by perfusion, would not be measured by the present experimental system. An example of this is the fast inhibition of rat heart glyceraldehyde-3-phosphatase dehydrogenase (lo-15 min in 1 ATA 0,), which is reversible (13). If such a short-lasting effect had a recovery time close to 11 min, this effect would be apparent in the first but not in subsequent samples. Huwever, there was no variable that was different in the first and not in subsequent samples and was also different for experimental and control hearts. The only parameter that
8I
hypoxia-14cm I
ii
u
c
'SO O2
52 EXPOSURE
54
56
60
62
TIMEA
6. Coronary resistance (CR) plotted sure time. See explanation in Fig. 2. FIG.
58
as a function
of 0, expo-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
2294
OXYGEN
EXPOSURE:
l
20
b4 .
.
a a
%
a a
30
50
c
52
54
56
58
60
WORKING
HEART
Coronary blood flow is known to be reduced in hyperoxia in the intact animal (19). We have shown that coronary resistance was not affected after the 0, exposure and there was a tendency for coronary resistance to decrease after 60 h exposure (Fig. 6). Thus reduced coronary flow in the intact animal can be related to reduced cardiac output (1, lo), increased blood viscosity, decreased LV tension (7), and contraction of precapillary sphincters (19). A puzzling but consistent finding was the biphasic effect of 0, exposure. After 50 h of 0, exposure, mechanical power output and efficiency were higher than in control hearts (Figs. 3 and lo), and only after 60 h of exposure did the heart become energetically inferior to the control. Another puzzling finding was increased contractility as 0, exposure progressed (Figs. 8 and 9). Cardiac contractility was also biphasic with respect to control hearts. There is only limited information on the toxic effect of 0, on cardiac contractility. Reduction of isometric systolic tension has been observed in the anesthetized dog transferred from 25% 0, to 100% 0, (7). Although Daniel1 and Bagwell (7) relate this reduction to the initial manifestation of 0, toxicity, the duration of the exposure in their study was only 30 min. Hyperbaric oxygen (3.6 ATA for 15 min) caused decreased myocardial contractility (dP/dt,,) (18). These two reports can be correlated with the initial reduction in both estimates of cardiac contractility found in the present study. However, the
62
O2 EXPOSURE TIME,fi FIG. 7. Cardiac output (CO) plotted as a function of 0, exposure time. See explanation in Fig. 2.
mental system it did not convey the mechanical power output clearly. We used multiple regression (using all measurements from the 25 hearts) of Vo,lbeat against total PVA/beat and dummy variables (assigning 0 to control and I to experimental hearts) to find out whether there were differences in either intercept or slope for control and experimental hearts. Both VO, and PVA were given in units of joules per beat per gram. The results showed no difference in intercept but a significant difference in slope (P < 0.001) for control and experimental hearts. We therefore calculated the slope for each heart, using linear regression of its six meisurements. These slopes, which represent the reciprocal work efficiency of the heart, are shown as a function of 0, exposure time in Fig. 10. The slope of the control hearts (2.1) is very similar to that found by Suga et al. (26) for puppies’ hearts (2.2). The calculated intercept for control hearts (0.007 J beat-l g-l) is also similar to that of the puppies’ hearts (0.005 J *beat-l* g-l). The work efficiency of the heart deteriorates as the hyperoxic exposure progresses and would obviously lead to death of the intact animal. These physiological findings may correlate with reported microscopic observations. Busing et al. (3) found in the rabbit exposed to I ATA 0, for 40 h that there is advanced disintegration of the heart myofibrils, whereas the mitochondria remain intact. Hughson et al. (15) reported initial mitochondrial swelling in 5 ATA O,exposed rats that preceded local dissolution of the myofibrillar structure. l
+
a
high
a
l
@-l&n l
~1000
;
l9
I,..,
15oot
2 y
l
+ 0
*
#
l l
ti
500
z t cr;t I 20oot
.
l
1
a
l
l l
fl
l
0
!’
I
I
1
1
1
I 0
hypoxia-14cm
l
0 1OOO
l 0
0
l
l
a
500
_e 7
,'50
*ta52
l
54)
l
&,
l
58,
l
60,
62,
,
O2 EXPOSURE TIME,h FIG. 8. Maximal elastance (E,,,) plotted as a function of 0, exposure time. See explanation in Fig. 2.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
OXYGEN
EXPOSURE:
8000
WORKING
2295
HEART
06.
6000
a
04.
$8000 ii E t
0
0 00 ’
high 02-19cm
* (I,
$6000 PI a 4000
0 l
l
0
0
m
O
l
0
0 @
2000 ti
0
high 02’19cm
l
a
I
control hyperoxia
high 02-14cm
5
0
l
OO
l
*4
O & 0
0 l
8000 6000
hypoxia-14cm
t
hypoxia-14cm
e l
*a
so-@
t l
l
a 1
I
1
c'
i
50’
O2
02.
a l
8
52
54I
56 ,
EXPOSURE
l
a
I 62
,
58
.
60 I
0
400
Emax,
TIME,h
l
0
4% 800
0 0
0
1200
torrpn1
plotted as a
FIG. 11. Energy efficiency (PVAP/vo, in J/min) plotted against cardiac contractility (E,,,) for control and experimental hearts (hyperoxia).
mechanism of the later increase in these two estimates is not clear. Decreased work efficiency has been measured and predicted in different situations when E,,, increases (22,27, 28). Our experimental apparatus and results are not the same as in these previous studies, because neither the EP nor the stroke volume was constant. However, we tested the relationship between cardiac contractility and work efficiency (PVAP/%,), as shown in Fig. 11. It seems that no clear relationship exists between E,,, and efficiency in the control hearts. However, there is a clear relationship between E,,, and work efficiency in the hearts of O,-exposed rats: work efficiency decreases as E,,, increases. What might be the cause of the tight relation-
ship of E,,, and efficiency of hearts from O,-exposed rats compared with that of the control hearts? One possibility is changes in muscle myosin isoenzymes during 0, exposure, as has been found in other types of stress: heat, exercise, thyroid state, and ventricular pressure overload (14, 17, 24). In that case the turnover time should be shortened because of 0, exposure. Another possible stress-related changes in the heart are density and affinity of P-adrenoreceptors. It is possible that 0, exposure made the heart more work efficient in the early stages of 0, exposure. This intriguing question is a topic for further study. Our findings on the effect 0, toxicity on the heart support earlier suggestions (10, 11,20) that the cardiovascular system is the system that led to the final collapse in pulmonary 0, toxicity.
FIG. 9. Maximal time derivative of pressure (dP/dt,,,) function of 0, exposure time. See explanation in Fig. 2.
i’oz=A
+BxPVAP
0
The authors thank R. Lincoln for skillful editing. This study was supported in part by the McRamer Foundation. The opinions and assertions made herein are the private ones of the authors and do not represent the official views of the Israeli Naval Hyperbaric Institute. Address for reprint requests: R. Arieli, Israeli Naval Hyperbaric Institute, PO Box 8040, Haifa 31080, Israel.
50
c! O2
l aI 52
EXPOSURE
Received 22 October 1990; accepted in final form 8 June 1992. I 54
l
I 56
1 58
I 60
1 62
TIME,h
FIG. 10. Slope (B) of linear regression of PVAP/beat against OO,/ beat (both in J beat?. g-l in equation) for individual hearts, plotted as a function of 0, exx)osure time. See exnlanation in Fig. 2. l
REFERENCES 1. BERGO, G. W., J. RISBERG, AND I. TY~SEBOTN. Effect of 5 bar oxygen on cardiac output and organ blood flow in conscious rats. Undersea Biomed. Res. 15: 457-470, 1988. 2. BLOUNT, D. H. Cardiac metabolic response to hyperbaric oxygen. Proc. Sot. EXD. BioL. 2Med. 133: 1129-1131, 1970.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
2296
OXYGEN
EXPOSURE:
3. Busr~c, C. M., U. KREENSEN, F. BUHLER, AND U. BLEYL. Light and electron microscopic examinations of experimentally produced heart muscle necroses following normobaric hyperoxia. Virchows Arch.
A Puthol.
Anat.
Histol.
366: 137-147,
1975.
4. CAULFIELD, J. B., R. W. SHELTON, AND J. F. BURKE. Cytotoxic effects of oxygen on striated muscle. Arch. Patho2. 94: 127-132, 1972. CLARK, J. M. Effects of acute and chronic hypercapnia on oxygen tolerance in rats. J. Appl. Physiol. 50: 1036-1044, 1981. CLARK, J. M., AND C. J. LAMBERT~EN. Pulmonary oxygen toxicity: a review. Pharmacok Rev. 23: 37-133, 1971. DANIELL, H. B., AND E. E. BAGWELL. Effects of high oxygen on coronary flow and heart force. Am. J. Physiol. 214: 1454-1459, 1968. 8. EDOUTE, Y., AND R, ARIELI. Effect of different degrees of hypoxia and reoxygenation on myocardial energetics. Isr. J. Med. Sci. 25: 3&2-3&8,19&9. 9. EDNJTE, Y., R. ARIELI, AND E. NEVO. Evidence for improved myocardial oxygen delivery and function during hypoxia in the mole rat. J. Camp. Physiol. B Biochem Syst. Environ. Physiul. 158: 5755&2,1988. IO. HARABIN, A. L., AND L, E. FARHI. Blood-gas transport in awake rabbits exposed to normobaric hyperoxia. Undersea Biomed. Res. L. D. HOMER, AND M. E. BRADLEY. Pulmonary oxygen toxicity in awake dogs: metabolic and physiological effects.
J. Appl. Physiol. 57: 1480-1488, 1984. 12. HAUGAARD, N., M. E. HESS, AND H. ITSK~VITZ.
The toxic action of oxygen on glucose and pyruvate oxidation in heart homogenates. J.
Biol. Chem. 227: 605-616,1957. 13. HORN, R. S., E. S. HAUGAARD, Biuchim. Biophys. 14. HOROWITZ, M.,
AND N. HAUGAARD. The mechanism of glycolysis by oxygen in rat heart homogenate. Acta 99: 549-552,
1965.
Y. M. PEYSER, AND A. MUHLRAD. Alterations in cardiac myosin isoenzymes distribution as an adaptation to chronic environmental heat stress in the rat. J. lMoZ. Cell. Cardiot. 18: 511-
515, 1986. 15. HUGHSON,
M., J. D. BALENTINE, AND H. B. DANIELL. The ultrastructural pathology of hyperbaric oxygen exposure: observations on the heart. Lab. Invest. 37: 516-525, 1977. 16. IGARASHI, Y., AND H. SUGA. Assessment of slope of end-systolic pressure-volume line of in situ dog heart. Am. J. Physiol. 250 (Heart
Circ. PhysioZ.
19): H6&5-H692,
1986.
HEART
S.-I., R. MATSUOKA, E. HIRATSUKA, M. KIMURA, T. AND A. TAKAO. Local response to cardiac overload on myosin heavy chain gene expression and isozyme transition. Circ.
17. IMAMURA. NISHIKAWA,
Res. 66: 1067-1073,199O. 18. KIUSCHOS, J. M., V. S. BEHAR, H. A. SALTZMAN, H. N. E. MYERS, W. W. SMITH, AND H. D. MCINTOSH.
perbaric oxygenation on left ventricular
K. THOMPSON, Effect of hy-
function. Am. J. Physiol. 216: 161-166, 1969. 19. KIRKMAN, A. D. Cardiovascular system and skeletal muscle. In: Pathology of Oxygen Toxicity, edited by J. D. Balentine. New York: Academic, 1982, p. 214-238. MATALON, S., M. S. NESARAJAH, AND L. E. FARHI. Pulmonary and circulatory changes in conscious sheep exposed to 100% O2 at 1 ATA. J. Appl. Physiok 53: 110-116, 1982. 21. NEELY, J. R., H. LIEBERMEISTER, E. J. BAWERSBY, AND H. E. MORGAN. Effect of pressure development on oxygen consumption by isolated rat heart. Am. J. Physiol. 212: 804-814, 1967. 22. NOZAWA, T., Y. YASUMURA, S, FUTAKI, N. TANAKA, Y. IGARASHI, Y. GOTO, AND H. SUGA. Relation between oxygen consumption and pressure-volume area of in situ dog heart. Am. J. Physiol. 253 (Heart Circ. Physiol. 22): H31-H40, 1987. 23. ONARHEIM, J., AND I. TYSSEBOTN. Effect of high ambient pressure and oxygen tension on organ blood flow in the anesthetized rat. Undersea
14: 133-147,19&7. 11. HARABIN, A. L.,
of the inhibition
WORKING
Biomed.
Res. 7: 47-60,
1980.
24. PAGANY, E. D., AND R. J. SOLARO. Swimming exercise, thyroid state, and the distribution of myosin isoenzymes in rat heart. Am. J. Physiol.
245 (Heart
Circ. Physiol.
14): H713-H720,
1983.
25. SHWARTZ,
L., S. A. BEN-HAIM, A. ELAMI, A. HASSAN, AND Y. EDOUTE, An algorithm for calculating left ventricular mechanical and energetic characteristics. Med. Biol. Eng. Comput. 29: 318-323, 1991.
26. SUGA, H., 0. YAMADA, Y. GOTO, Y. IGARASHI, Y. YASUMURA, T. NOZAWA, AND S. FUTAKI. Left ventricular 0, consumption and pressure-volume area in puppies. Am. J. Physiol. 253 (IIeart Circ. Physiol.
22): H770-H776,
1987.
27. SUGA, H., Y. YASUMURA, T. NOZAWA, S. FUTAKI, Y. IGARASHI, AND Y. GOTO. Prospective prediction of 0, consumption from pressurevolume area in dog hearts. Am. J. Physiol. 252 (Heart Circ. Physiol. 21): H1258-H1264,19&7. 28. TANAKA, N., Y. YASUMURA, T. NOZAWA, S. FUTAKI, M. UENISHI, K. HIRAMORI, AND H. SUGA. Optimal contractility and minimal oxygen consumption for constant external work of heart. Am. J. Physiol.
254
(Regulatory
Integrative
Comp.
Physiol.
23):
R933-
R943,198&.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 16, 2019.
