- Email: [email protected]
Rapid reduction of intestinal cytochrome a,a3 during lethal endotoxemia
JOURNAL
OF SURGICAL
RESEARCH
51, 382-391 (19%)
Rapid Reduction of Intestinal Cytochrome during Lethal Endotoxemia CARL F. SCHAEFER,PH.D.,’
a,+
M.D.,PH.D., MEGANR. LERNER,FRANSF. J~BSIS-VANDERVLIET,PH.D., ANDLENNARTFAGRAEUS, M.D.,PH.D.
BJ~~RNBIBER,
Department of Anesthesiology, University of Oklahoma Health Sciences Center and Research Service, Veterans Administration Medical Center, Oklahoma City, Oklahoma 73190 Submitted
for publication
In vivo near-infrared
spectrophotometry was used to determine whether lethal endotoxemia impairs small intestinal oxidative phosphorylation as reflected by the redox state of mitochoadrial cytochrome ala, (AA3). Adult male Sprague-Dawley rats were anesthetized with 2.1% isoflurane in 30% 0,:70% N20, and the small intestine was partially exteriorized for spectrophotometric monitoring (OMNI-3). By 5 min after an iv bolus of Escherichia coli endotoxin (40 mg/kg, LDgo , n = 7) a significant shift toward reduction in intestinal AA3 had occurred in association with hypotension and a marked fall in both superior mesenteric artery blood flow (SMAF) and cardiac output. In a separate group (n = 7) SMAF was kept at the baseline level by periodic infusions of donor rat plasma begun 1 min after endotoxin injection, and the reduction in AA3 was again found despite the fluid loading intervention which successfully maintained not only organ blood flow, but also cardiac output and mean arterial pressure in their normal ranges. Further experiments (n = 34) measuring SM vascular bed oxygen consumption indicated that intestinal 90, remained unchanged during early endotoxemia. These findings suggest a rapid impairment of oxidative phosphorylation by endotoxin which seems to occur through direct (and/or indirect) toxic cellular effects rather than through impaired tissue perfusion. 0 1991
Academic
Press,
February
ularly suited to studying the development of shock. This technology employs multiwavelength near-infraredspectrophotometry to monitor the redox state of mitochondrial cytochrome ~,a, (AA3) in uivo and in real time [4]. We have developed a rat model of septic shock [5] in which a very severe reduction in superior mesenteric arterial blood flow (SMAF) occurs in association with rapidly developing hemorrhagic pathology of the small intestine 161. The OMNI was used in the present study to determine whether redox changes in AA3 occur in the small intestine of the anesthetized rat given a lethal intravenous (iv) bolus dose of endotoxin. When a rapidly developing and sustained shift toward reduction of small intestinal mitochondrial AA3 was found in association with the fall in SMAF, we further tested whether the use of fluid loading to prevent the decrease in SMAF would also prevent the reduction of AA3 after endotoxin challenge. Finally, to pursue the mechanism of this AA3 redox shift, we investigated whether changes in systemic and regional (superior mesenteric vascular bed) oxygen consumption occurred in this rat model during early endotoxemia when the AA3 redox shift first occurred. METHODS Optical
Inc.
Circulatory shock is often defined as a state of inadequate perfusion of the tissues [l, 21. A new technology (Oxidative Metabolism Near-Infrared (OMNI) monitoring) developed by Jobsis-VanderVliet [3] for monitoring the state of tissue oxygen sufficiency seems partic‘To whom correspondence and reprint requests should be addressed at Department of Anesthesiology, University of Oklahoma Health Sciences Center, P.O. Box 26901, Res. Bldg. 29R, Oklahoma City, OK 73190.
Copyright 0 All
rights
1991 by of reproduction
382 Academic Press, Inc. in any
form
reserved.
Experiments
Surgical preparation. Twenty-two male SpragueDawley rats (Sasco, Omaha, NE) weighing 293 f 4 g (mean f SEM) were deprived of food but not water for 24 hr prior to the study. The rats were anesthetized with 4% isoflurane in 0, and endotracheally intubated [7]. Anesthesia was maintained with 2.1% isoflurane (1.5 MAC) [8] using a Harvard Rodent Respirator (arterial PCO, initially between 30-40 mm Hg). A thermocouplecatheter was then implanted (tip distal to aortic valve) to measure blood pressure, blood temperature (thermodilution cardiac output), and to collect arterial blood samples (0.2 ml) for pH, PCO,, and PO, (IL 1301 BGA), blood glucose (Ames Glucometer), plasma lactate (Kontron Lactate Analyzer), and hematocrit. Another catheter was inserted into the right atrium to measure cen-
INTRODUCTION
0022-4804/91$1.50
14, 1990
SCHAEFER
ET AL.: ENDOTOXIN-INDUCED
tral venous pressure and to inject endotoxin, fluid loads, and room temperature saline boluses (0.1 ml) for cardiac output (Cardiomax IIR, Columbus Instruments, Inc.). Heart rate was recorded from the arterial pulse using a tachograph. Following a midline laparotomy, an electromagnetic flow probe was placed on the superior mesenteric artery (SMA) to measure blood flow (FM 501, Carolina Medical Electronics) to the small intestine. About half the length of the small intestine (jejunum and ileum) was then cradled just above the abdominal wall in a troughlike construction made out of saline-soaked gauze. In vivo spectrophotometry. Multiwavelength, nearinfrared, differential spectrophotometry for in vivo monitoring of the redox state of cytochrome (AA3) has been described in detail previously [3, 41. The present study employed the OMNI-3 Spectrophotometer (International Instr. Lab, Inc., Durham, NC), which is a 3 wavelength (775,815, and 905 nm) instrument. Briefly, this technology is based on the principle that AA3 changes its near-infrared (NIR) spectral absorbance depending on whether it is oxidized or reduced. AA3 is the terminal cytochrome in the mitochondrial electron transport chain which transfers electrons to oxygen. Hemoglobin is the other major component of tissues which absorbs NIR light in an oxygen-dependent manner [4]. Due to the overlap of the absorption spectra for AA3, oxyhemoglobin (HbO,) , and deoxyhemoglobin (Hb) , three wavelengths of NIR light are used in the OMNI together with various algorithms to gauge changes in these molecules according to the relative changes in amounts of the light transmitted. The amount of AA3 in the preparation is constant and can only change proportional to the fraction that is oxidized (absorptive) or reduced (nonabsorptive). The Hb and HbO, can change both in relative proportion and amount, so a fourth signal, blood volume (BV), is derived from the algebraic sum of the Hb and HbO, signals. (Parenthetically, tissue myoglobin is detected in the same fashion as hemoglobin in both its oxygenated and deoxygenated states, but it is ignored here due to its low concentration (~5% of the tissue hemoglobin concentration) and because it cannot be distinguished from the corresponding hemoglobin signals.) The four signals (AA3, Hb, HbO,, and BV) are displayed by the OMNI in units (k1.0 V) of “variations in density” (V./d.), where one unit represents a lo-fold change in the algorithm value for the derived signal. The OMNI is thus a trend-monitoring instrument which produces data in terms of changes relative to baseline conditions, i.e., AAA3, AHb, AHbO,, and ABV. The exteriorized length of small intestine was transilluminated by positioning the OMNI optrodes on each side of the rat. The fiber bundle (7 mm diam.) projecting the near-infrared light was aimed across the abdomen through several intertwined loops of intestine directly at the clad glass collecting rod (9 mm diam.) of the photo-
MITOCHONDRIAL
REDOX
RESPONSE
383
multiplier tube (Hamamatsu R936) positioned approximately 25 mm on the other side of the intestine. The intestine was continuously moistened with saline (1 ml/ hr) using a Raze1 infusion pump. Beginning at laparotomy, fluid losses were offset by an iv infusion of 0.9% saline (3.5 ml/kg/hr). The intestine was then covered with plastic wrap, and a drape of dark cloth and aluminum foil was used to cover the optrodes and the animal to exclude extraneous light and to conserve body heat. Body temperature was maintained near 37°C using a heating pad and a blue light positioned above the drapes. Experimental protocol. The intestinal preparation was allowed to stabilize for 30 min, and the experiment was then begun with a 30 min control period. Measurements included mean aortic pressure (MAP), central venous pressure (CVP), heart rate (HR), superior mesenteric artery blood flow (SMAF), cardiac output (reported as cardiac index (CI), ml/min/lOO g BW), AA3 redox level, and tissue HbO,, Hb, and BV changes. Superior mesenteric arterial resistance (SMAR = (MAP - CVP)/SMAF, mm Hg/ml/min) and systemic vascular resistance (SVR = (MAP - CVP)/CI, mm Hg/ml/min/ 100 g BW) were calculated as indicated. Readings were taken at 30 and 25 min before endotoxin (designated as -30 and -25 min), and then the carrier gas for isoflurane was changed from 0, to 30% 0,:70% N,O until the end of the study. After time zero an iv bolus of endotoxin (E, n = 7) (40 mg/kg, Difco, LPS B Escherichia coli 0127:B8 dissolved in 0.9% saline, 1 ml/kg vol) was given over 30 set, a dose that when tested in conscious rats killed 9 of 10 animals within 24 hr. Sham (control) rats received saline (S, n = 8) (1 ml/kg), and the volumeloaded group (E + PL, n = 7) received E at time zero followed by intermittent infusions of donor rat plasma starting at +l min, i.e., prior to the time (+2 min) when blood pressure and blood flow normally begin to fall in untreated endotoxin-challenged rats. The rat plasma was collected previously from donor rats. Readings were continued until 2 hr after endotoxin injection when the rat was killed by anoxia (2.1% isoflurane in 100% N,O) to establish the maximal state of reduction of AA3. Blood samples were taken at -30, -5, +15, +60, and +120 min. At autopsy catheter placements were verified, and the severity of macroscopic pathology of the small intestine was assessed using a five-point rating scale where 0 signified no observable petechiae or hemorrhages and 4 indicated extensive, severe hemorrhagic banding of the tissue. Oxygen Consumption
Studies
Surgical preparation. Thirty-four male SpragueDawley rats weighing 263 + 3 g were fasted for 24 hr with water available ad lib. Anesthetic and surgical procedures were conducted in the same fashion as in the optical experiments except as follows. As soon as the rat was intubated, the carrier gas was changed from 0, to 30% 0,:70% N,O. Since no optical data were collected, the
384
JOURNAL
OF SURGICAL
RESEARCH:
small intestine was only temporarily exteriorized for catheterization of the superior mesenteric vein (SMV). An &cm length of Silastic catheter (Dow Corning No. 602-1050.3 mm ID, 0.6 mm OD) was inserted into the ileocolic vein (draining part of the cecum) and fed approximately 2 cm into the SMV for collecting blood samples. The flow probe was then positioned on the SMA, and the intestine was carefully placed back in the abdomen. Experimental protocol. A 30-min equilibration period was followed by a 30-min baseline period. Immediately after the first hemodynamic measurements at -30 min, blood samples (0.35 ml each) were drawn simultaneously from the three catheterized sites, aorta (A), right atrium (RA), and SMV, to measure blood oxygen contents using an IL 482 CO-Oximeter. Systemic (VO,) and regional (SMVO,). 0, consumption were calculated by the Fick principle (VOz.= (A - RA) X CI, reported in ml/min/kg BW, and SMVO, = (A - SMV) X SMAF, reported in ml/min). Immediately after the time zero readings endotoxin (E, n = 12) or saline (S, n = 12) was given to the rat (as above), and in the case of the volumeloaded groups (E + PL and S + PL, n = 5 in each group) intermittent plasma infusions were started at +l min. E + PL and S + PL rats were tested in alternating order and each S + PL rat received its plasma infusions according to the manner (timing and volume) in which the previous E + PL rat required volume in order to maintain its SMAF at baseline. A final set of three blood samples was collected at +15 min. Statistics. All results except the pathology scores are reported as mean f SEM. Repeated measures analysis of variance 19, lo] was used to evaluate the hemodynamic, OMNI, and blood data. When significant (P < 0.05) main or interactive effects were found, simple contrasts between-groups or within-groups (relative to time zero) were made using the appropriate paired or unpaired t test. The Median Test [ll], a nonparametric test, was used to analyze the macroscopic intestinal pathology scores. RESULTS
Optical Experiments Hemodynamic changes. Volume-loading (E + PL group) prevented the precipitous fall in MAP (Fig. 1A) that was seen in the untreated group (E) and maintained pressure near the saline control (S) level throughout the study. As expected SMAF (Fig. 1B) of the E group fell to about one-third of the baseline level by 5 min after endotoxin and showed essentially no recovery over the next 2 hr. Intermittent infusions of plasma (total = 9.3 + 1.2 ml) successfully maintained SMAF at the baseline level in the E + PL group as planned. The CI data (Fig. 1C) indicate that volume loading based on maintaining a normal SMAF produced a hyperdynamic state in the E + PL group relative to S, while the usual hypodynamic state was observed in the E group. A modest but significant increase in HR (Fig. 1D)
VOL.
51, NO. 5, NOVEMBER
1991
was observed during the first 90 min after endotoxin infusion in both E groups. SVR (Fig. 1E) and SMAR (Fig. 1F) were significantly elevated above control during most of the study in the E group, but remained at or below (+5 min) control in the E + PL group after endotoxin. Optical Data. Figure 2A shows that in the E group, intestinal mitochondrial AA3 shifted rapidly toward a maximally reduced level in concert with the fall in SMAF (Fig. 1B) induced by endotoxin and showed little recovery over the following 2 hr. Importantly, maintaining the blood supply to the small intestine at a normal level did not prevent the reduction of AA3 after endotoxin in the E + PL group. Between +30 and +120 min a gradual oxidation of AA3 occurred in the E + PL group so that by +105 min the redox level was not statistically different from control. When the animals were sacrificed at the end of the study, the maximal reduction of AA3 produced by anoxia was reached within 3 min at which point (+135 min, Fig. 2A) the level of reduction of AA3 was comparable in all three groups. The carrier gas shift from 100% 0, to 30% 0, at -25 min produced offsetting decreases and increases in small intestinal HbO, (Fig. 2B) and Hb (Fig. 20, respectively, as can be further appreciated by inspecting the BV data (Fig. 2D) which remained constant during the baseline period. HbO, fell significantly in the E group after endotoxin, recovered slightly, and then gradually decreased further during the remainder of the study. The tissue HbO, signal of the E + PL group showed the same tendency to decrease but was statistically different from control only at +45 min after endotoxin and was clearly comparable to S by +120 min. Both E groups showed increases in tissue Hb at +5 min which were essentially offsetting the HbO, decreases as can be seen in the BV data. Thereafter, Hb fell significantly below control until +90 min in the E + PL group but not in the E group. The tissue BV data provide an interesting contrast to the blood flow data (SMAF, Fig. 1B). While SMAF was held constant in the E + PL group by periodic volumeloading, tissue BV fell quite markedly after a latent period of more than 5 min and then slowly recovered to the control level. Furthermore, although BV and SMAF changes were more congruent in the E group, the marked delay before a decrease occurred in BV differed sharply from SMAF where the maximum decrease was seen by +5 min. Blood sample data. Data obtained from the arterial blood samples are presented in Table 1. Acidosis developed progressively in both E groups although more severely in the untreated E group. Despite constant ventilator settings, arterial PCO, dropped significantly after endotoxin in the E group while PO, increased slightly but significantly compared to controls. The E + PL group was comparable to S in both PCO, and POZ throughout the study. Volume-loading in the E + PL group prevented the increase in hematocrit which was noted in the E group after endotoxin and, relative to S,
SCHAEFER
ET AL.:
ENDOTOXIN-INDUCED
MITOCHONDRIAL
REDOX
385
RESPONSE
A.
9. 150 125 20
100 Q E '5
g 15
50
1 E '0
25
5 0’
0 -30
-15
15
0
30
45
60
90
75
105
-30
120
-15
0
15
I’ 60
75
90
105
120
D.
C. 600 I
:
1
:
500
E
L I’
100
-30
I 45
TIME (minutes)
TIME (minutes)
60
30
-15
15
0
30
45
60
75
90
105
-30
120
-15
0
E
E
45
60
i
”
15
30
75
90
105 120
TIME (minutes)
TIME (minutes)
E. F.
310
u
m 006 I
i
25
0 -30
-15
0
15
30
45
60
TIME (minutes)
75
90
105
120
-30
-15
0
15
30
45
60
75
90
105
120
TIME (minutes)
FIG. 1. Effects of endotoxin (40 mg/kg iv) or 0.9% saline (1 ml/kg iv) on hemodynamic variables in rats receiving endotoxin alone (E, n zz 7), endotoxin plus intermittent plasma infusions (E + PL, n = 7), or saline (S, n = 8). The vertical dashed line indicates the bolus infusion of endotoxin or saline. Plasma infusions were begun 1 min after the endotoxin bolus in the E + PL group. Values are mean + SEM. Significant differences (P 6 0.05) from the S group are indicated by E or P for the E and E + PL groups, respectively. (A) Mean aortic pressure; (B) superior mesenteric arterial flow; (C) cardiac index; (D) heart rate; (E) systemic vascular resistance; and (F) superior mesenteric arterial resistance.
386
JOURNAL
OF SURGICAL
RESEARCH:
VOL.
51, NO. 5, NOVEMBER
B.
A.
g 0.53 P 2 o.o-
;i z 5 Ok-
i I:
P 2 o.o-
h z
I : :
r, -5
g-0.5n .G Lo>
1 ” r ” :” -30 -15 0
I ” I ” I 15 30 45
’ t ” t ” t ” ! ” t , 60 75 90 105 120 135
s -05 0 .r i-l.0 >”
1 ” I ” F” I ” I ” I -30 -15 0 15 30 45
TIME (minutes)
C
:
$“I “F,’ -30 -15 0
/“““.“““l”“““’ 15 30 45
60
75
xI 60
I ” 1 I’ I 1, 75 90 105 120 135
TIME (minutes)
A 3 gj o-on ‘Lo.5-
1991
1 90 105 120 135
TIME (minutes)
>I % E-0.513 O-OS .E 5 -l.o- 1 1 -30 -15 >
: : 0
1 ( 15 30
TIM
, 45
1 1 60 75
2 I 1 I 90 w)5 w) u5
(minutes)
FIG. 2. Effects of endotoxin (40 mg/kg iv) or 0.9% saline (1 ml/kg iv) on optically derived variables in rats receiving endotoxin alone (E, n = 7), endotoxin plus intermittent plasma infusions (E + PL, n = 7), or saline (S, n = 8). The vertical dashed line indicates the bolus infusion of endotoxin or saline, Plasma infusions were begun 1 min after the endotoxin bolus in the E + PL group. The units (V./d.) stand for “variations in density,” where one unit represents a lo-fold change in the algorithm value for the derived signal. Values are mean + SEM. Significant differences (P < 0.05) from the S group are indicated by E or P for the E and E + PL groups, respectively. (A) Cytochrome AA3; (B) oxyhemoglobin (HbOx); (C) deoxyhemoglobin (Hb); and (D) blood volume (Hb + HbO,).
produced only a slight hemodilution by the +120 min sample time. Endotoxin induced severe hyperlactatemia in the E group. Although plasma lactate increased relative to baseline in the E + PL group also, the plasma infusions moderated this increase to the degree that it was only statistically different from the S group at +60 min. Blood glucose, on the other hand, declined comparably in both E groups so that by +120 min significant hypoglycemia had developed. Mild to moderate peteSmall intestinal pathology. chial hemorrhage of the intestinal wall was noted in both endotoxin groups (median scores: E = 3; E + PL = 2). While no significant difference in pathology scores was found between endotoxin groups, both groups were significantly different from the control group (S = 0). Oxygen Consumption
Studies
Hemodynamic data. The cardiovascular changes observed in the oxygen consumption studies (Fig. 3) closely replicate those of the optical studies presented above
(Fig. 1). Of new interest here are the results of the S + PL group which was added for completeness of design. While volume-loading had no effect on MAP (Fig. 3A) of the S + PL rats, SMAF (Fig. 3B) and CI (Fig. 3C) were increased significantly above baseline by the volume of plasma (total = 9.9 +- 0.5 ml) needed to maintain SMAF at baseline in the E + PL group. A sustained increase of 3 to 5 mm Hg occurred in CVP (Fig. 3D) with matched volume-loading in the S + PL group while in the E + PL rats CVP was elevated at +5 min only. Calculations of SVR (Fig. 3E) and SMAR (Fig. 3F) again showed that the increase in vascular resistance produced by endotoxin (E) was prevented by volume-loading (E + PL) and furthermore that equivalent plasma infusions in the controls (S + PL) significantly reduced vascular resistance. Oxygen consumption data. The systemic and regional oxygen consumption data are shown in Table 2. Both systemically and regionally the A-V 0, difference widened in the E group and narrowed in the S + PL group by +15 min to compensate for opposite changes
SCHAEFER
ET AL.:
ENDOTOXIN-INDUCED
MITOCHONDRIAL
TABLE Arterial
REDOX
387
RESPONSE
1
Blood Data for the Optical Experiments” Time, min
PH E S E + PL PCO, (mm Hg) E S E+PL POP (mm Hd E S E+PL HCRT (%) E S E+PL LAC (mg/dl) E S E + PL GLU (mg/dl) E S E+PL
-30
-5
f15
+60
f120
7.39 + .02 7.35 f .Ol 7.37 f .02
7.39 * .Ol 7.35 f .Ol 7.35 * .02
7.32 + .Ol 7.33 + .Ol 7.30 f .02
7.19 f .02t 7.30 + .Ol 7.23 f .02t
7.00 * .05t 7.29 f .Ol 7.16 f .03t
33 + 1 35 + 1 30 + 1
32 f 1 33 + 1 33 ? 2
28 + 1 36 f 1 36 f 2
31+ 2t 38 + 2 37 + 2
22 + 5t 38 f 2 40 f 2
409 + 14 347 + 23 340 + 26
128 k 4 121 Ik 5 123 k 7
125 k 5t 107 * 5 112 + 8
129 + 4t 97 f 5 110 + 8
155 + lot 94 f 6 105 * 7
44 + 1 44 + 1 44 + 2
45 Ii 1 44 + 1 44 + 1
50 + 1t 43 + 1 42 k 1
49 t 2t 42 2 1 39 k 2
51 f 3t 42 + 0.3 35 f 1t
16 + 2 17 f 3 16 f 2
18 + 4 18 + 2 16 + 2
27 t 3t 18 + 2 24 k 4
36 f 2t 19 + 3 29 * 4t
81 2 13t 25 + 5 41 * 8
191 t 36 112 * 8 143 + 13
148 -c 16 101 f 5 126 f 10
134 + 23 103 * 4 131+ 8
98 + 25 105 + 2 104 + 8
48 3~ 19t 112 k 7 64 rt~8t
“Values are means + SEM. E, endotoxin group, n = 7; S, saline group, n = 8; E + PL, endotoxin hematocrit; LAC, plasma lactate; GLU, blood glucose. t Significant difference (P 60.05) from the S group.
that had occurred in blood flow (Figs. 3B and 3C). As a result whole body and superior mesenteric regional oxygen consumption remained unchanged relative to baseline. The regional oxygen consumption of the endotoxin groups was significantly greater at +15 min than in the saline controls, but this difference seemed to stem at least in part from the tendency toward higher baseline values in the endotoxin groups. Other blood sample data. The arterial blood gas, hematocrit, and other metabolic data are listed in Table 3. Generally speaking, by +15 min only minor though statistically significant changes were produced by endotoxin in the acid-base and blood gas values, consistent with results in the optical study above (Table 1). Plasma infusion in the S + PL group resulted in significant hemodilution and a consequent large drop in arterial 0, content. As noted previously only a small decrease in hematocrit was seen in the E + PL group, and this had a correspondingly minor effect on arterial 0, content. Although plasma lactate in the E + PL group was significantly increased at +15 min relative to baseline, the value was no different from the S + PL group and much lower than in the E group. In contrast to the decrease in blood glucose in the E group, blood glucose increased slightly by +15 min in the E + PL group although less than in the S + PL group.
plus plasma infusion,
n = 7; HCRT,
Small intestinal pathology. Both endotoxin groups were observed at autopsy to have light to moderate petechial hemorrhage of the intestinal wall (median scores: E = 1 vs S = 0, P < 0.05; E + PL = 3 vs S + PL = 0, P G 0.05). DISCUSSION The rapid reduction of small intestinal AA3 in conjunction with the fall in SMAF after endotoxin infusion seemed on first consideration to be highly compatible with the traditional view [l] of endotoxin-induced hypoperfusion leading to tissue hypoxia and consequent redox shift of mitochondrial AA3. However, according to this view maintenance of normal organ blood flow should have prevented the reduction of tissue AA3. Since this maneuver did not block the endotoxin-induced reduction of intestinal AA3, we seem to be left with two major alternatives to explain the AA3 response. On the one hand, endotoxin may act rapidly to poison the cells, directly [12] and/or indirectly [13]. On the other hand, endotoxin may choke off nutritive blood flow, for example, by inducing disseminated intravascular coagulation (DIC) [ 14,151. In the latter case maintenance of total organ blood flow presumably would not equate to maintenance of nutritive blood flow, since the
388
JOURNAL
OF SURGICAL
RESEARCH:
VOL. 51, NO. 5, NOVEMBER
1991
CS-QS*PL
A-AE
5.
I 1
A-AE*PL
0
-15
-30
TIME (minutes)
TIME (minutes)
D. 99 70
6..
D-OS
3
13160 $0
(3-QS+PL
7-
A-AE
6
A-A
E+PL
O-OS D-cIS+PL Lt--AE
25.
6-A
E+PL
E4 32
dl
I
-30
-15
1 I 1
0
, 15
1 0 -30
-15
.O
15
TIME (minutes)
TIME (minutes)
F.
E.
o--u
S+PL
TIME (minutes1
TIME (minutes)
FIG. 3. Effects of endotoxin (40 mg/kg iv) or 0.9% saline (1 ml/kg iv) on hemodynamic variables in rats receiving endotoxin alone (E, n = 12), endotoxin plus intermittent plasma infusions (E + PL, n = 12), saline alone (S, n = 5), or saline plus intermittent plasma infusions (S f PL, n = 5). The vertical dashed line indicates the bolus infusion of endotoxin or saline. Plasma infusions were begun 1 min after the endotoxin or saline bolus, and the infusions given the S + PL group matched those given the E + PL group. Values are mean t SEM. Significant differences (P & 0.05) from the S or S + PL group are indicated by E or P for the E and E + PL groups, respectively, while significant changes from baseline (time 0) within any group are noted by an asterisk. (A) Mean aortic pressure; (B) superior mesenteric arterial flow; (C) cardiac index; (D) central venous pressure; (E) systemic vascular resistance; and (F) superior mesenteric vascular resistance.
blood would be shunted through the few remaining unoccluded capillaries which then act as thoroughfare channels [ 161.
At first glance, the DIC explanation seems supported by the perplexing BV data (Fig. 2D) where, despite the maintenance of total organ blood flow (SMAF, Fig. lB),
SCHAEFER
ET AL.:
TABLE Systemic and Regional
ENDOTOXIN-INDUCED
2
Oxygen Consumption
Data”
Time, min -30 GO, W/dU E S E + PL s + PL A-VO, (ml/dl) E S E+PL s + PL VO, (ml/min/kg) E S E+PL s+PL A-SMVO, (ml/dl) E S E + PL s+PL SMVO, (ml/min) E S E+PL s + PL
+15
17.8 18.1 18.6 17.5
-I f f *
0.5 0.5 0.8 0.5
19.1 17.2 15.9 11.3
f 2 + *
0.5**t 0.3*,t 0.5*q 0.4*
5.5 5.1 5.7 6.0
k * ? +
0.4 0.3 0.8 0.9
10.9 5.7 4.7 2.7
+ + * k
0.9*-f 0.4 0.5t 0.4*
13.9 13.4 14.1 16.5
f Ii t r
1.2 0.7 1.9 1.5
14.9 14.9 14.5 14.8
f f + k
1.1 1.1 1.1 3.3
5.5 4.8 5.4 5.4
+ t + +
0.4 0.5 0.9 0.5
11.3 5.1 5.5 2.8
* + + +
0.4*st 0.4 0.7t 0.3*
0.82 0.61 0.76 0.69
k k -t f
0.11 0.07 0.16 0.08
0.87 0.58 0.83 0.55
+ + f +
0.12t 0.04 0.09t 0.06*
MITOCHONDRIAL
REDOX
389
RESPONSE
than the related shift in AA3 in terms of variation in density units [4]. To provide further information in regard to the above alternate hypotheses, we collected oxygen consumption data with the expectation that an impairment in nutritive blood flow would result in a decrease in regional VO, while a toxic effect which uncoupled mitochondrial oxidative phosphorylation should increase regional VO, . As often seems to happen, the actual result fell in between these two scenarios, since both regional (superior mesenteric vascular bed) and systemic VO, remained unchanged relative to baseline at f15 min after endotoxin, the time by which the maximal reduction in intestinal AA3 had occurred in the E group. To determine whether the decrease in SMAF had blunted an increase in VO,
TABLE
3
Arterial Blood Data for the Oxygen Consumption Experiments” Time, min -30
’ Values are means f SEM. E, endotoxin group, n = 12; S, saline group, n = 12; E + PL, endotoxin plus plasma infusion, n = 5; S + PL, saline plus plasma infusion, n = 5; C,O,, aortic blood oxygen content; A-VO,, aortic minus right atria1 blood oxygen concentration difference; VO,, systemic oxygen consumption rate; A-SMVO*, aortic minus superior mesenteric vein oxygen concentration difference; SMVO,, superior mesenteric vascular bed oxygen consumption rate. * Significant difference (P < 0.05) within a group relative to baseline. t Significant difference (P < 0.05) for E vs S or for E + PL vs S + PL.
the volume of blood in the optical field which resides mainly in the capillaries 14) decreased strikingly in the E + PL group just as in the E group. However, the BV decrease did not occur until after the +5 min reading when most of the AA3 reduction had already occurred, and, thus, could not be causing the initial and major part of the AA3 shift. This decrease in BV in the E + PL group is unlikely to be a dilutional phenomenon, since (1) the timing is out of phase with the plasma loading which began at +l min and (2) the decrease in hematocrit was very gradual and minor (Table 1). More importantly, the HbO, data are inconsistent with the capillary plugging hypothesis in that the decrease in tissue HbO, following endotoxin injection was disproportionately small relative to the reduction in AA3. That is, if the reduction in AA3 was the result of an impairment in the 0, supply to the tissue, that impairment should be refleeted in a decrease in HbO, which is equal to or greater
PH E S E + PL s+PL PCO, (mmHg) E S E + PL s + PL
i-15
7.36 7.37 7.36 7.34
+ f 2 *
.Ol .Ol .Ol .Ol
7.30 7.34 7.31 7.30
35 37 33 34
+ + -c f
1 2 2 1
96 104 99 89
+ f f f
5 6 5 4
46 46 45 45
f f -c f
1 1 1 1
51 -t 45 t 41+ 31+
1*t 1* 1t 1*
18 + 21+ 19 -c 22 f
1 2 2 2
42 21 29 28
3*t 2 4* 5
z!z5 * 10 i 7 + 8
64 111 128 173
30 36 37 40
+- .01*t t .01* * .02 + .03 2 -t f t
1*t 1 2 2*
PO,bmHd E S E+PL s + PL HCRT (%) E S E + PL s + PL LAC (mg/dl) E S E + PL s + PL
97 32 5 100 k 5 104 * 12 go* 13
a f f t
GLUbdW E S E + PL s + PL
89 107 104 120
k 5*t k 10 + 8*t zk 14*
n Values are means + SEM. E, endotoxin group, n = 12; S, saline group, n = 12; E + PL, endotoxin plus plasma infusion, n = 5; S + PL, saline plus plasma infusion; n = 5; HCRT, hematocrit; LAC, plasma lactate; GLU, blood glucose. * Significant difference (P < 0.05) within a group relative to baseline. t Significant difference (P d 0.05) for E vs S or for E + PL vs S + PL.
390
JOURNAL
OF SURGICAL
RESEARCH:
that in vitro data [17] indicate should occur with a toxic uncoupling of the mitochondria, the plasma-loaded groups were included to unmask this possible flow-limitation of VO,. However, the results for E + PL confirmed the findings with E alone, since, with SMAF maintained constant, the A-V 0, difference remained at baseline rather.than increasing as in the E group, with the result that VO, remained constant. Our oxygen consumption findings agree with the results of others in septic animals [2] and in septic shock patients [la]. Previously, in vitro studies of oxidative phosphorylation have been performed with isolated mitochondria exposed directly to endotoxin or with mitochondria that have been isolated from animals challenged in vivo with endotoxin or bacteria. Unfortunately, results of these studies conflict, with some providing support for an impairment of oxidative metabolism, either direct [12] or indirect [13, 19,201, while others find no impairment of mitochondrial oxidative metabolism [21, 221. Nonetheless, our results are quite consistent with recent results of Astiz et al. [23] using a rat cecal ligation and puncture model to study the effect of sepsis on muscle oxidative metabolism. Fluid infusion which prevented the systemic hypoperfusion and muscle tissue hypoxia seen in untreated peritonitis failed to prevent a significant reduction in muscle ATP and concomitant increase in muscle lactate/pyruvate ratio. Thus, Astiz et al. [23] concluded that sepsis causes an early metabolic injury which is not dependent on tissue hypoxia and is nonetheless associated with increased lactate production. In addition to our results with intestinal AA3 several other findings are noteworthy. First, volume-loading with the single purpose of preventing a fall in SMAF had several unintended but interesting side-effects, such as (1) preventing hypotension, (2) converting the hypodynamic response into a hyperdynamic one (increased CI post-endotoxin), (3) preventing hemoconcentration, and (4) reducing hyperlactatemia. Second, as shown convincingly by others [24], a rapid loss of plasma into the extravascular space may explain part of the fall in venous return, since only a minor decrease in hematocrit occurred in either E + PL group while significant hemodilution and a prolonged increase in CVP occurred in the S + PL group given the same amount of plasma. The present results suggest that vascular resistance increases in response to decreased venous return, since no change in systemic or regional resistance was seen in the E + PL groups in which volume-loading prevented the fall in venous return as judged by CI (normal or increased). Clearly, the reduction of AA3 induced by endotoxin may be causally related to the decrease in venous return, but much work remains to be done to clarify this relationship. In summary, we have found that lethal endotoxin challenge rapidly (within 3 min) reduces intestinal AA3 through a mechanism which is not dependent on a re-
VOL.
51, NO. 5, NOVEMBER
1991
duction in organ blood flow. This finding together with a lack of change in regional (superior mesenteric vascular bed) VO, seem incompatible with the traditional view of impaired tissue perfusion. As suggested by the present data and the results of others [2, 231, the assumption that hyperlactatemia reflects tissue ischemia and hypoxia seems unwarranted. A direct and/or indirect toxic action of endotoxin to rapidly uncouple cellular oxidative phosphorylation may explain the initial pathological events in sepsis. ACKNOWLEDGMENTS Supported in part by the Veterans Administration Medical Research Service, the College of Medicine Alumni Association Research Fund, the Department of Anesthesiology Research Fund, and the Swedish Medical Research Council (Projects 6575 and 6938). Presented in part at the 1986 and 1988 Shock Society Meetings. We gratefully acknowledge the helpful assistance of Daniel Brackett, Mary Kaye Foster, Carolyn Getsinger, Theresa Lander, and the Research and Education Computer Center.
REFERENCES 1. 2.
3.
4.
5.
6.
7.
8. 9. 10. 11. 12.
Chaudry, I. H. Cellular mechanisms in shock and ischemia and their correction. Am. J. Physiol. 246: R117, 1983. Van Lambalgen, A. A., Bronsveld, W., Van Den Bos, G. C., and Thijs, L. G. Distribution of cardiac output, oxygen consumption and lactate production in canine endotoxin shock. Cardiouasc. Res. 18: 195, 1984. Jobsis, F. F. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 198: 1264, 1977. Jobsis-VanderVliet, F. F. Noninvasive, near infrared monitoring of cellular oxygen sufficiency in uivo. In F. Kreuzer, S. M. Cain, Z. Turek, and T. K. Goldstick (Eds.), Oxygen Transport to Tissue VII. New York: Plenum, 1985. Pp. 833841. Brackett, D. J., Schaefer, C. F., Tompkins, P., Fagraeus, L., Peters, L. J., and Wilson, M. F. Evaluation of cardiac output, total peripheral vascular resistance, and plasma concentrations of vasopressin in the conscious, unrestrained rat during endotoxemia. Circ. Shock 17: 273, 1985. Biber, B., Schaefer, C. F., Smolik, M. C., Lawrence, M. R., Lerner, M. R., Brackett, D. J., Wilson, M. F.. and Fagraeus, L. Dose-related effects of isoflurane on superior mesenteric vasoconstriction induced by endotoxemia in the rat. Acta Anaesthesiol. Stand. 31: 430, 1987. Schaefer, C. F., Bracket& D. J., Downs, P., Tompkins, P., and Wilson, M. F. Laryngoscopic endotracheal intubation of rats for inhalation anesthesia. J. Appl. Physiol. 56: 533, 1984. White, P. F., Johnston, R. R., and Eger, E. I., II. Determination of anesthetic requirement in rats. Anesthesiology 40: 52, 1974. SAS Institute, Inc. SAS User’s Guide: Statistics, Version 5 Edition. Cary, NC: SAS Institute, Inc., 1985. Pp. 478-506. Winer, B. J. Statistical Principles in Experimental Design. New York: McGraw-Hill, 1962. Pp. 298-378. Hays, W. L. Statistics for Psychologists. New York: Holt, Rinehart, and Winston, 1963. Pp. 620-623. Schumer, W., Dae Gupta, T. K., Moss, G. S., Nyhus, L. M. Effect of endotoxemia on liver cell mitochondria in man. Ann. Surg.
171:875,1970. 13.
Mela, L. Direct
and indirect
effects of endotoxin
on mitochon-
SCHAEFER
14.
15.
16.
17. 18.
ET AL.:
ENDOTOXIN-INDUCED
drial function. In J. A. Majde and R. J. Person (Eds.), Puthophysiological Effects of Endotonins at the Cellular Level. New York: A. R. Liss, 1981. Pp. 15-21. Hardaway, R. M., Husni, E. A., Geever, E. F., Noyes, H. E., and Burns, J. W. Endotoxin shock: A manifestation of intravascular coagulation. Ann. Surg. 154: 791, 1961. Movat, H. Z. Microcirculation in disseminated intravascular coagulation induced by endotoxins. In E. M. Renkin and C. C. Michel (Eds.), Handbook of Physiology, Section 2: The Cardiovascular System, Vol. IV, Microcirculation, Part 2. Bethesda: Am. Physiological Society. 1984. Pp. 1047-1076. Haljamae, H., Amundson, B., Bagge, U., Jennische, E., and Branemark, P. I. Pathophysiology of Shock. Puthol. Res. Pruct. 165: 200, 1979. Lehninger, A. L. Biochemistry. New York: Worth, 1975. Pp. 509540. Dahn, M. S., Lange, P., Lobdell, K., Hans, B., Jacobs, L. A., and Mitchell, R. A. Splanchnic and total body oxygen consumption differences in septic and injured patients. Surgery 101: 69,1987.
MITOCHONDRIAL 19.
20.
21.
22. 23.
24.
REDOX
RESPONSE
391
Corbucci, G. G., Gasparetto, A., Candiani, A., Crimi, G., Antonelli, M., Bufi, M., DeBlasi, R. A., Cooper, M. B., and Gohil, K. Shock-induced damage to mitochondrial function and some cellular antioxidant mechanisms in humans. Circ. Shock 15: 15, 1985. Mela, L., Bacalzo, L. V., Jr., and Miller, L. D. Defective oxidative metabolism of rat liver mitochondria in hemorrhagic and endotoxin shock. Am. J. Physiol. 220: 571, 1971. Dawson, K. L., Geller, E. R., and Kirkpatrick, J. R. Enhancement of mitochondrial function in sepsis. Arch. Surg. 123: 241, 1988. Geller, E. R., Jankauskas, S., and Kirkpatrick, J. Mitochondrial death in sepsis: A failed concept. J. Surg. Res. 40: 514, 1986. Astiz, M., Rackow, E. C., Weil, M. H., and Schumer, W. Early impairment of oxidative metabolism and energy production in severe sepsis. Circ. Shock 26: 311, 1988. Van Lambalgen, A. A., Rasker, M. T. E., Van Den Bos, G. C., and Thijs, L. G. Effects of endotoxemia on systemic plasma loss and hematocrit in rats. Microvasc. Res. 36: 291, 1988.
OF SURGICAL
RESEARCH
51, 382-391 (19%)
Rapid Reduction of Intestinal Cytochrome during Lethal Endotoxemia CARL F. SCHAEFER,PH.D.,’
a,+
M.D.,PH.D., MEGANR. LERNER,FRANSF. J~BSIS-VANDERVLIET,PH.D., ANDLENNARTFAGRAEUS, M.D.,PH.D.
BJ~~RNBIBER,
Department of Anesthesiology, University of Oklahoma Health Sciences Center and Research Service, Veterans Administration Medical Center, Oklahoma City, Oklahoma 73190 Submitted
for publication
In vivo near-infrared
spectrophotometry was used to determine whether lethal endotoxemia impairs small intestinal oxidative phosphorylation as reflected by the redox state of mitochoadrial cytochrome ala, (AA3). Adult male Sprague-Dawley rats were anesthetized with 2.1% isoflurane in 30% 0,:70% N20, and the small intestine was partially exteriorized for spectrophotometric monitoring (OMNI-3). By 5 min after an iv bolus of Escherichia coli endotoxin (40 mg/kg, LDgo , n = 7) a significant shift toward reduction in intestinal AA3 had occurred in association with hypotension and a marked fall in both superior mesenteric artery blood flow (SMAF) and cardiac output. In a separate group (n = 7) SMAF was kept at the baseline level by periodic infusions of donor rat plasma begun 1 min after endotoxin injection, and the reduction in AA3 was again found despite the fluid loading intervention which successfully maintained not only organ blood flow, but also cardiac output and mean arterial pressure in their normal ranges. Further experiments (n = 34) measuring SM vascular bed oxygen consumption indicated that intestinal 90, remained unchanged during early endotoxemia. These findings suggest a rapid impairment of oxidative phosphorylation by endotoxin which seems to occur through direct (and/or indirect) toxic cellular effects rather than through impaired tissue perfusion. 0 1991
Academic
Press,
February
ularly suited to studying the development of shock. This technology employs multiwavelength near-infraredspectrophotometry to monitor the redox state of mitochondrial cytochrome ~,a, (AA3) in uivo and in real time [4]. We have developed a rat model of septic shock [5] in which a very severe reduction in superior mesenteric arterial blood flow (SMAF) occurs in association with rapidly developing hemorrhagic pathology of the small intestine 161. The OMNI was used in the present study to determine whether redox changes in AA3 occur in the small intestine of the anesthetized rat given a lethal intravenous (iv) bolus dose of endotoxin. When a rapidly developing and sustained shift toward reduction of small intestinal mitochondrial AA3 was found in association with the fall in SMAF, we further tested whether the use of fluid loading to prevent the decrease in SMAF would also prevent the reduction of AA3 after endotoxin challenge. Finally, to pursue the mechanism of this AA3 redox shift, we investigated whether changes in systemic and regional (superior mesenteric vascular bed) oxygen consumption occurred in this rat model during early endotoxemia when the AA3 redox shift first occurred. METHODS Optical
Inc.
Circulatory shock is often defined as a state of inadequate perfusion of the tissues [l, 21. A new technology (Oxidative Metabolism Near-Infrared (OMNI) monitoring) developed by Jobsis-VanderVliet [3] for monitoring the state of tissue oxygen sufficiency seems partic‘To whom correspondence and reprint requests should be addressed at Department of Anesthesiology, University of Oklahoma Health Sciences Center, P.O. Box 26901, Res. Bldg. 29R, Oklahoma City, OK 73190.
Copyright 0 All
rights
1991 by of reproduction
382 Academic Press, Inc. in any
form
reserved.
Experiments
Surgical preparation. Twenty-two male SpragueDawley rats (Sasco, Omaha, NE) weighing 293 f 4 g (mean f SEM) were deprived of food but not water for 24 hr prior to the study. The rats were anesthetized with 4% isoflurane in 0, and endotracheally intubated [7]. Anesthesia was maintained with 2.1% isoflurane (1.5 MAC) [8] using a Harvard Rodent Respirator (arterial PCO, initially between 30-40 mm Hg). A thermocouplecatheter was then implanted (tip distal to aortic valve) to measure blood pressure, blood temperature (thermodilution cardiac output), and to collect arterial blood samples (0.2 ml) for pH, PCO,, and PO, (IL 1301 BGA), blood glucose (Ames Glucometer), plasma lactate (Kontron Lactate Analyzer), and hematocrit. Another catheter was inserted into the right atrium to measure cen-
INTRODUCTION
0022-4804/91$1.50
14, 1990
SCHAEFER
ET AL.: ENDOTOXIN-INDUCED
tral venous pressure and to inject endotoxin, fluid loads, and room temperature saline boluses (0.1 ml) for cardiac output (Cardiomax IIR, Columbus Instruments, Inc.). Heart rate was recorded from the arterial pulse using a tachograph. Following a midline laparotomy, an electromagnetic flow probe was placed on the superior mesenteric artery (SMA) to measure blood flow (FM 501, Carolina Medical Electronics) to the small intestine. About half the length of the small intestine (jejunum and ileum) was then cradled just above the abdominal wall in a troughlike construction made out of saline-soaked gauze. In vivo spectrophotometry. Multiwavelength, nearinfrared, differential spectrophotometry for in vivo monitoring of the redox state of cytochrome (AA3) has been described in detail previously [3, 41. The present study employed the OMNI-3 Spectrophotometer (International Instr. Lab, Inc., Durham, NC), which is a 3 wavelength (775,815, and 905 nm) instrument. Briefly, this technology is based on the principle that AA3 changes its near-infrared (NIR) spectral absorbance depending on whether it is oxidized or reduced. AA3 is the terminal cytochrome in the mitochondrial electron transport chain which transfers electrons to oxygen. Hemoglobin is the other major component of tissues which absorbs NIR light in an oxygen-dependent manner [4]. Due to the overlap of the absorption spectra for AA3, oxyhemoglobin (HbO,) , and deoxyhemoglobin (Hb) , three wavelengths of NIR light are used in the OMNI together with various algorithms to gauge changes in these molecules according to the relative changes in amounts of the light transmitted. The amount of AA3 in the preparation is constant and can only change proportional to the fraction that is oxidized (absorptive) or reduced (nonabsorptive). The Hb and HbO, can change both in relative proportion and amount, so a fourth signal, blood volume (BV), is derived from the algebraic sum of the Hb and HbO, signals. (Parenthetically, tissue myoglobin is detected in the same fashion as hemoglobin in both its oxygenated and deoxygenated states, but it is ignored here due to its low concentration (~5% of the tissue hemoglobin concentration) and because it cannot be distinguished from the corresponding hemoglobin signals.) The four signals (AA3, Hb, HbO,, and BV) are displayed by the OMNI in units (k1.0 V) of “variations in density” (V./d.), where one unit represents a lo-fold change in the algorithm value for the derived signal. The OMNI is thus a trend-monitoring instrument which produces data in terms of changes relative to baseline conditions, i.e., AAA3, AHb, AHbO,, and ABV. The exteriorized length of small intestine was transilluminated by positioning the OMNI optrodes on each side of the rat. The fiber bundle (7 mm diam.) projecting the near-infrared light was aimed across the abdomen through several intertwined loops of intestine directly at the clad glass collecting rod (9 mm diam.) of the photo-
MITOCHONDRIAL
REDOX
RESPONSE
383
multiplier tube (Hamamatsu R936) positioned approximately 25 mm on the other side of the intestine. The intestine was continuously moistened with saline (1 ml/ hr) using a Raze1 infusion pump. Beginning at laparotomy, fluid losses were offset by an iv infusion of 0.9% saline (3.5 ml/kg/hr). The intestine was then covered with plastic wrap, and a drape of dark cloth and aluminum foil was used to cover the optrodes and the animal to exclude extraneous light and to conserve body heat. Body temperature was maintained near 37°C using a heating pad and a blue light positioned above the drapes. Experimental protocol. The intestinal preparation was allowed to stabilize for 30 min, and the experiment was then begun with a 30 min control period. Measurements included mean aortic pressure (MAP), central venous pressure (CVP), heart rate (HR), superior mesenteric artery blood flow (SMAF), cardiac output (reported as cardiac index (CI), ml/min/lOO g BW), AA3 redox level, and tissue HbO,, Hb, and BV changes. Superior mesenteric arterial resistance (SMAR = (MAP - CVP)/SMAF, mm Hg/ml/min) and systemic vascular resistance (SVR = (MAP - CVP)/CI, mm Hg/ml/min/ 100 g BW) were calculated as indicated. Readings were taken at 30 and 25 min before endotoxin (designated as -30 and -25 min), and then the carrier gas for isoflurane was changed from 0, to 30% 0,:70% N,O until the end of the study. After time zero an iv bolus of endotoxin (E, n = 7) (40 mg/kg, Difco, LPS B Escherichia coli 0127:B8 dissolved in 0.9% saline, 1 ml/kg vol) was given over 30 set, a dose that when tested in conscious rats killed 9 of 10 animals within 24 hr. Sham (control) rats received saline (S, n = 8) (1 ml/kg), and the volumeloaded group (E + PL, n = 7) received E at time zero followed by intermittent infusions of donor rat plasma starting at +l min, i.e., prior to the time (+2 min) when blood pressure and blood flow normally begin to fall in untreated endotoxin-challenged rats. The rat plasma was collected previously from donor rats. Readings were continued until 2 hr after endotoxin injection when the rat was killed by anoxia (2.1% isoflurane in 100% N,O) to establish the maximal state of reduction of AA3. Blood samples were taken at -30, -5, +15, +60, and +120 min. At autopsy catheter placements were verified, and the severity of macroscopic pathology of the small intestine was assessed using a five-point rating scale where 0 signified no observable petechiae or hemorrhages and 4 indicated extensive, severe hemorrhagic banding of the tissue. Oxygen Consumption
Studies
Surgical preparation. Thirty-four male SpragueDawley rats weighing 263 + 3 g were fasted for 24 hr with water available ad lib. Anesthetic and surgical procedures were conducted in the same fashion as in the optical experiments except as follows. As soon as the rat was intubated, the carrier gas was changed from 0, to 30% 0,:70% N,O. Since no optical data were collected, the
384
JOURNAL
OF SURGICAL
RESEARCH:
small intestine was only temporarily exteriorized for catheterization of the superior mesenteric vein (SMV). An &cm length of Silastic catheter (Dow Corning No. 602-1050.3 mm ID, 0.6 mm OD) was inserted into the ileocolic vein (draining part of the cecum) and fed approximately 2 cm into the SMV for collecting blood samples. The flow probe was then positioned on the SMA, and the intestine was carefully placed back in the abdomen. Experimental protocol. A 30-min equilibration period was followed by a 30-min baseline period. Immediately after the first hemodynamic measurements at -30 min, blood samples (0.35 ml each) were drawn simultaneously from the three catheterized sites, aorta (A), right atrium (RA), and SMV, to measure blood oxygen contents using an IL 482 CO-Oximeter. Systemic (VO,) and regional (SMVO,). 0, consumption were calculated by the Fick principle (VOz.= (A - RA) X CI, reported in ml/min/kg BW, and SMVO, = (A - SMV) X SMAF, reported in ml/min). Immediately after the time zero readings endotoxin (E, n = 12) or saline (S, n = 12) was given to the rat (as above), and in the case of the volumeloaded groups (E + PL and S + PL, n = 5 in each group) intermittent plasma infusions were started at +l min. E + PL and S + PL rats were tested in alternating order and each S + PL rat received its plasma infusions according to the manner (timing and volume) in which the previous E + PL rat required volume in order to maintain its SMAF at baseline. A final set of three blood samples was collected at +15 min. Statistics. All results except the pathology scores are reported as mean f SEM. Repeated measures analysis of variance 19, lo] was used to evaluate the hemodynamic, OMNI, and blood data. When significant (P < 0.05) main or interactive effects were found, simple contrasts between-groups or within-groups (relative to time zero) were made using the appropriate paired or unpaired t test. The Median Test [ll], a nonparametric test, was used to analyze the macroscopic intestinal pathology scores. RESULTS
Optical Experiments Hemodynamic changes. Volume-loading (E + PL group) prevented the precipitous fall in MAP (Fig. 1A) that was seen in the untreated group (E) and maintained pressure near the saline control (S) level throughout the study. As expected SMAF (Fig. 1B) of the E group fell to about one-third of the baseline level by 5 min after endotoxin and showed essentially no recovery over the next 2 hr. Intermittent infusions of plasma (total = 9.3 + 1.2 ml) successfully maintained SMAF at the baseline level in the E + PL group as planned. The CI data (Fig. 1C) indicate that volume loading based on maintaining a normal SMAF produced a hyperdynamic state in the E + PL group relative to S, while the usual hypodynamic state was observed in the E group. A modest but significant increase in HR (Fig. 1D)
VOL.
51, NO. 5, NOVEMBER
1991
was observed during the first 90 min after endotoxin infusion in both E groups. SVR (Fig. 1E) and SMAR (Fig. 1F) were significantly elevated above control during most of the study in the E group, but remained at or below (+5 min) control in the E + PL group after endotoxin. Optical Data. Figure 2A shows that in the E group, intestinal mitochondrial AA3 shifted rapidly toward a maximally reduced level in concert with the fall in SMAF (Fig. 1B) induced by endotoxin and showed little recovery over the following 2 hr. Importantly, maintaining the blood supply to the small intestine at a normal level did not prevent the reduction of AA3 after endotoxin in the E + PL group. Between +30 and +120 min a gradual oxidation of AA3 occurred in the E + PL group so that by +105 min the redox level was not statistically different from control. When the animals were sacrificed at the end of the study, the maximal reduction of AA3 produced by anoxia was reached within 3 min at which point (+135 min, Fig. 2A) the level of reduction of AA3 was comparable in all three groups. The carrier gas shift from 100% 0, to 30% 0, at -25 min produced offsetting decreases and increases in small intestinal HbO, (Fig. 2B) and Hb (Fig. 20, respectively, as can be further appreciated by inspecting the BV data (Fig. 2D) which remained constant during the baseline period. HbO, fell significantly in the E group after endotoxin, recovered slightly, and then gradually decreased further during the remainder of the study. The tissue HbO, signal of the E + PL group showed the same tendency to decrease but was statistically different from control only at +45 min after endotoxin and was clearly comparable to S by +120 min. Both E groups showed increases in tissue Hb at +5 min which were essentially offsetting the HbO, decreases as can be seen in the BV data. Thereafter, Hb fell significantly below control until +90 min in the E + PL group but not in the E group. The tissue BV data provide an interesting contrast to the blood flow data (SMAF, Fig. 1B). While SMAF was held constant in the E + PL group by periodic volumeloading, tissue BV fell quite markedly after a latent period of more than 5 min and then slowly recovered to the control level. Furthermore, although BV and SMAF changes were more congruent in the E group, the marked delay before a decrease occurred in BV differed sharply from SMAF where the maximum decrease was seen by +5 min. Blood sample data. Data obtained from the arterial blood samples are presented in Table 1. Acidosis developed progressively in both E groups although more severely in the untreated E group. Despite constant ventilator settings, arterial PCO, dropped significantly after endotoxin in the E group while PO, increased slightly but significantly compared to controls. The E + PL group was comparable to S in both PCO, and POZ throughout the study. Volume-loading in the E + PL group prevented the increase in hematocrit which was noted in the E group after endotoxin and, relative to S,
SCHAEFER
ET AL.:
ENDOTOXIN-INDUCED
MITOCHONDRIAL
REDOX
385
RESPONSE
A.
9. 150 125 20
100 Q E '5
g 15
50
1 E '0
25
5 0’
0 -30
-15
15
0
30
45
60
90
75
105
-30
120
-15
0
15
I’ 60
75
90
105
120
D.
C. 600 I
:
1
:
500
E
L I’
100
-30
I 45
TIME (minutes)
TIME (minutes)
60
30
-15
15
0
30
45
60
75
90
105
-30
120
-15
0
E
E
45
60
i
”
15
30
75
90
105 120
TIME (minutes)
TIME (minutes)
E. F.
310
u
m 006 I
i
25
0 -30
-15
0
15
30
45
60
TIME (minutes)
75
90
105
120
-30
-15
0
15
30
45
60
75
90
105
120
TIME (minutes)
FIG. 1. Effects of endotoxin (40 mg/kg iv) or 0.9% saline (1 ml/kg iv) on hemodynamic variables in rats receiving endotoxin alone (E, n zz 7), endotoxin plus intermittent plasma infusions (E + PL, n = 7), or saline (S, n = 8). The vertical dashed line indicates the bolus infusion of endotoxin or saline. Plasma infusions were begun 1 min after the endotoxin bolus in the E + PL group. Values are mean + SEM. Significant differences (P 6 0.05) from the S group are indicated by E or P for the E and E + PL groups, respectively. (A) Mean aortic pressure; (B) superior mesenteric arterial flow; (C) cardiac index; (D) heart rate; (E) systemic vascular resistance; and (F) superior mesenteric arterial resistance.
386
JOURNAL
OF SURGICAL
RESEARCH:
VOL.
51, NO. 5, NOVEMBER
B.
A.
g 0.53 P 2 o.o-
;i z 5 Ok-
i I:
P 2 o.o-
h z
I : :
r, -5
g-0.5n .G Lo>
1 ” r ” :” -30 -15 0
I ” I ” I 15 30 45
’ t ” t ” t ” ! ” t , 60 75 90 105 120 135
s -05 0 .r i-l.0 >”
1 ” I ” F” I ” I ” I -30 -15 0 15 30 45
TIME (minutes)
C
:
$“I “F,’ -30 -15 0
/“““.“““l”“““’ 15 30 45
60
75
xI 60
I ” 1 I’ I 1, 75 90 105 120 135
TIME (minutes)
A 3 gj o-on ‘Lo.5-
1991
1 90 105 120 135
TIME (minutes)
>I % E-0.513 O-OS .E 5 -l.o- 1 1 -30 -15 >
: : 0
1 ( 15 30
TIM
, 45
1 1 60 75
2 I 1 I 90 w)5 w) u5
(minutes)
FIG. 2. Effects of endotoxin (40 mg/kg iv) or 0.9% saline (1 ml/kg iv) on optically derived variables in rats receiving endotoxin alone (E, n = 7), endotoxin plus intermittent plasma infusions (E + PL, n = 7), or saline (S, n = 8). The vertical dashed line indicates the bolus infusion of endotoxin or saline, Plasma infusions were begun 1 min after the endotoxin bolus in the E + PL group. The units (V./d.) stand for “variations in density,” where one unit represents a lo-fold change in the algorithm value for the derived signal. Values are mean + SEM. Significant differences (P < 0.05) from the S group are indicated by E or P for the E and E + PL groups, respectively. (A) Cytochrome AA3; (B) oxyhemoglobin (HbOx); (C) deoxyhemoglobin (Hb); and (D) blood volume (Hb + HbO,).
produced only a slight hemodilution by the +120 min sample time. Endotoxin induced severe hyperlactatemia in the E group. Although plasma lactate increased relative to baseline in the E + PL group also, the plasma infusions moderated this increase to the degree that it was only statistically different from the S group at +60 min. Blood glucose, on the other hand, declined comparably in both E groups so that by +120 min significant hypoglycemia had developed. Mild to moderate peteSmall intestinal pathology. chial hemorrhage of the intestinal wall was noted in both endotoxin groups (median scores: E = 3; E + PL = 2). While no significant difference in pathology scores was found between endotoxin groups, both groups were significantly different from the control group (S = 0). Oxygen Consumption
Studies
Hemodynamic data. The cardiovascular changes observed in the oxygen consumption studies (Fig. 3) closely replicate those of the optical studies presented above
(Fig. 1). Of new interest here are the results of the S + PL group which was added for completeness of design. While volume-loading had no effect on MAP (Fig. 3A) of the S + PL rats, SMAF (Fig. 3B) and CI (Fig. 3C) were increased significantly above baseline by the volume of plasma (total = 9.9 +- 0.5 ml) needed to maintain SMAF at baseline in the E + PL group. A sustained increase of 3 to 5 mm Hg occurred in CVP (Fig. 3D) with matched volume-loading in the S + PL group while in the E + PL rats CVP was elevated at +5 min only. Calculations of SVR (Fig. 3E) and SMAR (Fig. 3F) again showed that the increase in vascular resistance produced by endotoxin (E) was prevented by volume-loading (E + PL) and furthermore that equivalent plasma infusions in the controls (S + PL) significantly reduced vascular resistance. Oxygen consumption data. The systemic and regional oxygen consumption data are shown in Table 2. Both systemically and regionally the A-V 0, difference widened in the E group and narrowed in the S + PL group by +15 min to compensate for opposite changes
SCHAEFER
ET AL.:
ENDOTOXIN-INDUCED
MITOCHONDRIAL
TABLE Arterial
REDOX
387
RESPONSE
1
Blood Data for the Optical Experiments” Time, min
PH E S E + PL PCO, (mm Hg) E S E+PL POP (mm Hd E S E+PL HCRT (%) E S E+PL LAC (mg/dl) E S E + PL GLU (mg/dl) E S E+PL
-30
-5
f15
+60
f120
7.39 + .02 7.35 f .Ol 7.37 f .02
7.39 * .Ol 7.35 f .Ol 7.35 * .02
7.32 + .Ol 7.33 + .Ol 7.30 f .02
7.19 f .02t 7.30 + .Ol 7.23 f .02t
7.00 * .05t 7.29 f .Ol 7.16 f .03t
33 + 1 35 + 1 30 + 1
32 f 1 33 + 1 33 ? 2
28 + 1 36 f 1 36 f 2
31+ 2t 38 + 2 37 + 2
22 + 5t 38 f 2 40 f 2
409 + 14 347 + 23 340 + 26
128 k 4 121 Ik 5 123 k 7
125 k 5t 107 * 5 112 + 8
129 + 4t 97 f 5 110 + 8
155 + lot 94 f 6 105 * 7
44 + 1 44 + 1 44 + 2
45 Ii 1 44 + 1 44 + 1
50 + 1t 43 + 1 42 k 1
49 t 2t 42 2 1 39 k 2
51 f 3t 42 + 0.3 35 f 1t
16 + 2 17 f 3 16 f 2
18 + 4 18 + 2 16 + 2
27 t 3t 18 + 2 24 k 4
36 f 2t 19 + 3 29 * 4t
81 2 13t 25 + 5 41 * 8
191 t 36 112 * 8 143 + 13
148 -c 16 101 f 5 126 f 10
134 + 23 103 * 4 131+ 8
98 + 25 105 + 2 104 + 8
48 3~ 19t 112 k 7 64 rt~8t
“Values are means + SEM. E, endotoxin group, n = 7; S, saline group, n = 8; E + PL, endotoxin hematocrit; LAC, plasma lactate; GLU, blood glucose. t Significant difference (P 60.05) from the S group.
that had occurred in blood flow (Figs. 3B and 3C). As a result whole body and superior mesenteric regional oxygen consumption remained unchanged relative to baseline. The regional oxygen consumption of the endotoxin groups was significantly greater at +15 min than in the saline controls, but this difference seemed to stem at least in part from the tendency toward higher baseline values in the endotoxin groups. Other blood sample data. The arterial blood gas, hematocrit, and other metabolic data are listed in Table 3. Generally speaking, by +15 min only minor though statistically significant changes were produced by endotoxin in the acid-base and blood gas values, consistent with results in the optical study above (Table 1). Plasma infusion in the S + PL group resulted in significant hemodilution and a consequent large drop in arterial 0, content. As noted previously only a small decrease in hematocrit was seen in the E + PL group, and this had a correspondingly minor effect on arterial 0, content. Although plasma lactate in the E + PL group was significantly increased at +15 min relative to baseline, the value was no different from the S + PL group and much lower than in the E group. In contrast to the decrease in blood glucose in the E group, blood glucose increased slightly by +15 min in the E + PL group although less than in the S + PL group.
plus plasma infusion,
n = 7; HCRT,
Small intestinal pathology. Both endotoxin groups were observed at autopsy to have light to moderate petechial hemorrhage of the intestinal wall (median scores: E = 1 vs S = 0, P < 0.05; E + PL = 3 vs S + PL = 0, P G 0.05). DISCUSSION The rapid reduction of small intestinal AA3 in conjunction with the fall in SMAF after endotoxin infusion seemed on first consideration to be highly compatible with the traditional view [l] of endotoxin-induced hypoperfusion leading to tissue hypoxia and consequent redox shift of mitochondrial AA3. However, according to this view maintenance of normal organ blood flow should have prevented the reduction of tissue AA3. Since this maneuver did not block the endotoxin-induced reduction of intestinal AA3, we seem to be left with two major alternatives to explain the AA3 response. On the one hand, endotoxin may act rapidly to poison the cells, directly [12] and/or indirectly [13]. On the other hand, endotoxin may choke off nutritive blood flow, for example, by inducing disseminated intravascular coagulation (DIC) [ 14,151. In the latter case maintenance of total organ blood flow presumably would not equate to maintenance of nutritive blood flow, since the
388
JOURNAL
OF SURGICAL
RESEARCH:
VOL. 51, NO. 5, NOVEMBER
1991
CS-QS*PL
A-AE
5.
I 1
A-AE*PL
0
-15
-30
TIME (minutes)
TIME (minutes)
D. 99 70
6..
D-OS
3
13160 $0
(3-QS+PL
7-
A-AE
6
A-A
E+PL
O-OS D-cIS+PL Lt--AE
25.
6-A
E+PL
E4 32
dl
I
-30
-15
1 I 1
0
, 15
1 0 -30
-15
.O
15
TIME (minutes)
TIME (minutes)
F.
E.
o--u
S+PL
TIME (minutes1
TIME (minutes)
FIG. 3. Effects of endotoxin (40 mg/kg iv) or 0.9% saline (1 ml/kg iv) on hemodynamic variables in rats receiving endotoxin alone (E, n = 12), endotoxin plus intermittent plasma infusions (E + PL, n = 12), saline alone (S, n = 5), or saline plus intermittent plasma infusions (S f PL, n = 5). The vertical dashed line indicates the bolus infusion of endotoxin or saline. Plasma infusions were begun 1 min after the endotoxin or saline bolus, and the infusions given the S + PL group matched those given the E + PL group. Values are mean t SEM. Significant differences (P & 0.05) from the S or S + PL group are indicated by E or P for the E and E + PL groups, respectively, while significant changes from baseline (time 0) within any group are noted by an asterisk. (A) Mean aortic pressure; (B) superior mesenteric arterial flow; (C) cardiac index; (D) central venous pressure; (E) systemic vascular resistance; and (F) superior mesenteric vascular resistance.
blood would be shunted through the few remaining unoccluded capillaries which then act as thoroughfare channels [ 161.
At first glance, the DIC explanation seems supported by the perplexing BV data (Fig. 2D) where, despite the maintenance of total organ blood flow (SMAF, Fig. lB),
SCHAEFER
ET AL.:
TABLE Systemic and Regional
ENDOTOXIN-INDUCED
2
Oxygen Consumption
Data”
Time, min -30 GO, W/dU E S E + PL s + PL A-VO, (ml/dl) E S E+PL s + PL VO, (ml/min/kg) E S E+PL s+PL A-SMVO, (ml/dl) E S E + PL s+PL SMVO, (ml/min) E S E+PL s + PL
+15
17.8 18.1 18.6 17.5
-I f f *
0.5 0.5 0.8 0.5
19.1 17.2 15.9 11.3
f 2 + *
0.5**t 0.3*,t 0.5*q 0.4*
5.5 5.1 5.7 6.0
k * ? +
0.4 0.3 0.8 0.9
10.9 5.7 4.7 2.7
+ + * k
0.9*-f 0.4 0.5t 0.4*
13.9 13.4 14.1 16.5
f Ii t r
1.2 0.7 1.9 1.5
14.9 14.9 14.5 14.8
f f + k
1.1 1.1 1.1 3.3
5.5 4.8 5.4 5.4
+ t + +
0.4 0.5 0.9 0.5
11.3 5.1 5.5 2.8
* + + +
0.4*st 0.4 0.7t 0.3*
0.82 0.61 0.76 0.69
k k -t f
0.11 0.07 0.16 0.08
0.87 0.58 0.83 0.55
+ + f +
0.12t 0.04 0.09t 0.06*
MITOCHONDRIAL
REDOX
389
RESPONSE
than the related shift in AA3 in terms of variation in density units [4]. To provide further information in regard to the above alternate hypotheses, we collected oxygen consumption data with the expectation that an impairment in nutritive blood flow would result in a decrease in regional VO, while a toxic effect which uncoupled mitochondrial oxidative phosphorylation should increase regional VO, . As often seems to happen, the actual result fell in between these two scenarios, since both regional (superior mesenteric vascular bed) and systemic VO, remained unchanged relative to baseline at f15 min after endotoxin, the time by which the maximal reduction in intestinal AA3 had occurred in the E group. To determine whether the decrease in SMAF had blunted an increase in VO,
TABLE
3
Arterial Blood Data for the Oxygen Consumption Experiments” Time, min -30
’ Values are means f SEM. E, endotoxin group, n = 12; S, saline group, n = 12; E + PL, endotoxin plus plasma infusion, n = 5; S + PL, saline plus plasma infusion, n = 5; C,O,, aortic blood oxygen content; A-VO,, aortic minus right atria1 blood oxygen concentration difference; VO,, systemic oxygen consumption rate; A-SMVO*, aortic minus superior mesenteric vein oxygen concentration difference; SMVO,, superior mesenteric vascular bed oxygen consumption rate. * Significant difference (P < 0.05) within a group relative to baseline. t Significant difference (P < 0.05) for E vs S or for E + PL vs S + PL.
the volume of blood in the optical field which resides mainly in the capillaries 14) decreased strikingly in the E + PL group just as in the E group. However, the BV decrease did not occur until after the +5 min reading when most of the AA3 reduction had already occurred, and, thus, could not be causing the initial and major part of the AA3 shift. This decrease in BV in the E + PL group is unlikely to be a dilutional phenomenon, since (1) the timing is out of phase with the plasma loading which began at +l min and (2) the decrease in hematocrit was very gradual and minor (Table 1). More importantly, the HbO, data are inconsistent with the capillary plugging hypothesis in that the decrease in tissue HbO, following endotoxin injection was disproportionately small relative to the reduction in AA3. That is, if the reduction in AA3 was the result of an impairment in the 0, supply to the tissue, that impairment should be refleeted in a decrease in HbO, which is equal to or greater
PH E S E + PL s+PL PCO, (mmHg) E S E + PL s + PL
i-15
7.36 7.37 7.36 7.34
+ f 2 *
.Ol .Ol .Ol .Ol
7.30 7.34 7.31 7.30
35 37 33 34
+ + -c f
1 2 2 1
96 104 99 89
+ f f f
5 6 5 4
46 46 45 45
f f -c f
1 1 1 1
51 -t 45 t 41+ 31+
1*t 1* 1t 1*
18 + 21+ 19 -c 22 f
1 2 2 2
42 21 29 28
3*t 2 4* 5
z!z5 * 10 i 7 + 8
64 111 128 173
30 36 37 40
+- .01*t t .01* * .02 + .03 2 -t f t
1*t 1 2 2*
PO,bmHd E S E+PL s + PL HCRT (%) E S E + PL s + PL LAC (mg/dl) E S E + PL s + PL
97 32 5 100 k 5 104 * 12 go* 13
a f f t
GLUbdW E S E + PL s + PL
89 107 104 120
k 5*t k 10 + 8*t zk 14*
n Values are means + SEM. E, endotoxin group, n = 12; S, saline group, n = 12; E + PL, endotoxin plus plasma infusion, n = 5; S + PL, saline plus plasma infusion; n = 5; HCRT, hematocrit; LAC, plasma lactate; GLU, blood glucose. * Significant difference (P < 0.05) within a group relative to baseline. t Significant difference (P d 0.05) for E vs S or for E + PL vs S + PL.
390
JOURNAL
OF SURGICAL
RESEARCH:
that in vitro data [17] indicate should occur with a toxic uncoupling of the mitochondria, the plasma-loaded groups were included to unmask this possible flow-limitation of VO,. However, the results for E + PL confirmed the findings with E alone, since, with SMAF maintained constant, the A-V 0, difference remained at baseline rather.than increasing as in the E group, with the result that VO, remained constant. Our oxygen consumption findings agree with the results of others in septic animals [2] and in septic shock patients [la]. Previously, in vitro studies of oxidative phosphorylation have been performed with isolated mitochondria exposed directly to endotoxin or with mitochondria that have been isolated from animals challenged in vivo with endotoxin or bacteria. Unfortunately, results of these studies conflict, with some providing support for an impairment of oxidative metabolism, either direct [12] or indirect [13, 19,201, while others find no impairment of mitochondrial oxidative metabolism [21, 221. Nonetheless, our results are quite consistent with recent results of Astiz et al. [23] using a rat cecal ligation and puncture model to study the effect of sepsis on muscle oxidative metabolism. Fluid infusion which prevented the systemic hypoperfusion and muscle tissue hypoxia seen in untreated peritonitis failed to prevent a significant reduction in muscle ATP and concomitant increase in muscle lactate/pyruvate ratio. Thus, Astiz et al. [23] concluded that sepsis causes an early metabolic injury which is not dependent on tissue hypoxia and is nonetheless associated with increased lactate production. In addition to our results with intestinal AA3 several other findings are noteworthy. First, volume-loading with the single purpose of preventing a fall in SMAF had several unintended but interesting side-effects, such as (1) preventing hypotension, (2) converting the hypodynamic response into a hyperdynamic one (increased CI post-endotoxin), (3) preventing hemoconcentration, and (4) reducing hyperlactatemia. Second, as shown convincingly by others [24], a rapid loss of plasma into the extravascular space may explain part of the fall in venous return, since only a minor decrease in hematocrit occurred in either E + PL group while significant hemodilution and a prolonged increase in CVP occurred in the S + PL group given the same amount of plasma. The present results suggest that vascular resistance increases in response to decreased venous return, since no change in systemic or regional resistance was seen in the E + PL groups in which volume-loading prevented the fall in venous return as judged by CI (normal or increased). Clearly, the reduction of AA3 induced by endotoxin may be causally related to the decrease in venous return, but much work remains to be done to clarify this relationship. In summary, we have found that lethal endotoxin challenge rapidly (within 3 min) reduces intestinal AA3 through a mechanism which is not dependent on a re-
VOL.
51, NO. 5, NOVEMBER
1991
duction in organ blood flow. This finding together with a lack of change in regional (superior mesenteric vascular bed) VO, seem incompatible with the traditional view of impaired tissue perfusion. As suggested by the present data and the results of others [2, 231, the assumption that hyperlactatemia reflects tissue ischemia and hypoxia seems unwarranted. A direct and/or indirect toxic action of endotoxin to rapidly uncouple cellular oxidative phosphorylation may explain the initial pathological events in sepsis. ACKNOWLEDGMENTS Supported in part by the Veterans Administration Medical Research Service, the College of Medicine Alumni Association Research Fund, the Department of Anesthesiology Research Fund, and the Swedish Medical Research Council (Projects 6575 and 6938). Presented in part at the 1986 and 1988 Shock Society Meetings. We gratefully acknowledge the helpful assistance of Daniel Brackett, Mary Kaye Foster, Carolyn Getsinger, Theresa Lander, and the Research and Education Computer Center.
REFERENCES 1. 2.
3.
4.
5.
6.
7.
8. 9. 10. 11. 12.
Chaudry, I. H. Cellular mechanisms in shock and ischemia and their correction. Am. J. Physiol. 246: R117, 1983. Van Lambalgen, A. A., Bronsveld, W., Van Den Bos, G. C., and Thijs, L. G. Distribution of cardiac output, oxygen consumption and lactate production in canine endotoxin shock. Cardiouasc. Res. 18: 195, 1984. Jobsis, F. F. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 198: 1264, 1977. Jobsis-VanderVliet, F. F. Noninvasive, near infrared monitoring of cellular oxygen sufficiency in uivo. In F. Kreuzer, S. M. Cain, Z. Turek, and T. K. Goldstick (Eds.), Oxygen Transport to Tissue VII. New York: Plenum, 1985. Pp. 833841. Brackett, D. J., Schaefer, C. F., Tompkins, P., Fagraeus, L., Peters, L. J., and Wilson, M. F. Evaluation of cardiac output, total peripheral vascular resistance, and plasma concentrations of vasopressin in the conscious, unrestrained rat during endotoxemia. Circ. Shock 17: 273, 1985. Biber, B., Schaefer, C. F., Smolik, M. C., Lawrence, M. R., Lerner, M. R., Brackett, D. J., Wilson, M. F.. and Fagraeus, L. Dose-related effects of isoflurane on superior mesenteric vasoconstriction induced by endotoxemia in the rat. Acta Anaesthesiol. Stand. 31: 430, 1987. Schaefer, C. F., Bracket& D. J., Downs, P., Tompkins, P., and Wilson, M. F. Laryngoscopic endotracheal intubation of rats for inhalation anesthesia. J. Appl. Physiol. 56: 533, 1984. White, P. F., Johnston, R. R., and Eger, E. I., II. Determination of anesthetic requirement in rats. Anesthesiology 40: 52, 1974. SAS Institute, Inc. SAS User’s Guide: Statistics, Version 5 Edition. Cary, NC: SAS Institute, Inc., 1985. Pp. 478-506. Winer, B. J. Statistical Principles in Experimental Design. New York: McGraw-Hill, 1962. Pp. 298-378. Hays, W. L. Statistics for Psychologists. New York: Holt, Rinehart, and Winston, 1963. Pp. 620-623. Schumer, W., Dae Gupta, T. K., Moss, G. S., Nyhus, L. M. Effect of endotoxemia on liver cell mitochondria in man. Ann. Surg.
171:875,1970. 13.
Mela, L. Direct
and indirect
effects of endotoxin
on mitochon-
SCHAEFER
14.
15.
16.
17. 18.
ET AL.:
ENDOTOXIN-INDUCED
drial function. In J. A. Majde and R. J. Person (Eds.), Puthophysiological Effects of Endotonins at the Cellular Level. New York: A. R. Liss, 1981. Pp. 15-21. Hardaway, R. M., Husni, E. A., Geever, E. F., Noyes, H. E., and Burns, J. W. Endotoxin shock: A manifestation of intravascular coagulation. Ann. Surg. 154: 791, 1961. Movat, H. Z. Microcirculation in disseminated intravascular coagulation induced by endotoxins. In E. M. Renkin and C. C. Michel (Eds.), Handbook of Physiology, Section 2: The Cardiovascular System, Vol. IV, Microcirculation, Part 2. Bethesda: Am. Physiological Society. 1984. Pp. 1047-1076. Haljamae, H., Amundson, B., Bagge, U., Jennische, E., and Branemark, P. I. Pathophysiology of Shock. Puthol. Res. Pruct. 165: 200, 1979. Lehninger, A. L. Biochemistry. New York: Worth, 1975. Pp. 509540. Dahn, M. S., Lange, P., Lobdell, K., Hans, B., Jacobs, L. A., and Mitchell, R. A. Splanchnic and total body oxygen consumption differences in septic and injured patients. Surgery 101: 69,1987.
MITOCHONDRIAL 19.
20.
21.
22. 23.
24.
REDOX
RESPONSE
391
Corbucci, G. G., Gasparetto, A., Candiani, A., Crimi, G., Antonelli, M., Bufi, M., DeBlasi, R. A., Cooper, M. B., and Gohil, K. Shock-induced damage to mitochondrial function and some cellular antioxidant mechanisms in humans. Circ. Shock 15: 15, 1985. Mela, L., Bacalzo, L. V., Jr., and Miller, L. D. Defective oxidative metabolism of rat liver mitochondria in hemorrhagic and endotoxin shock. Am. J. Physiol. 220: 571, 1971. Dawson, K. L., Geller, E. R., and Kirkpatrick, J. R. Enhancement of mitochondrial function in sepsis. Arch. Surg. 123: 241, 1988. Geller, E. R., Jankauskas, S., and Kirkpatrick, J. Mitochondrial death in sepsis: A failed concept. J. Surg. Res. 40: 514, 1986. Astiz, M., Rackow, E. C., Weil, M. H., and Schumer, W. Early impairment of oxidative metabolism and energy production in severe sepsis. Circ. Shock 26: 311, 1988. Van Lambalgen, A. A., Rasker, M. T. E., Van Den Bos, G. C., and Thijs, L. G. Effects of endotoxemia on systemic plasma loss and hematocrit in rats. Microvasc. Res. 36: 291, 1988.