- Email: [email protected]
Mitochondrial death in sepsis: A failed concept
Mitochondrial EVAN R. GELLER,
M.D.,S.
Death in Sepsis: A Failed Concept JANKAUSKAS,M.D.,ANDJ.KIRKPATRICK,M.D.
Wayne State University, Department of Surgery, Detroit, Michigan 48202 Presented at the Annual Meeting of the Association for Academic Surgery, Cincinnati, Ohio, November 10-13, 1985 The concept of early selective mitochondrial injury has been proposed to explain the global metabolic dysfunction observed in the septic state. A two phase study was undertaken to test the validity of this hypothesis. In the initial phase, an endotoxin shock model was employed in the rat to delineate the function of skeletal muscle mitochondria. Mitochondrial function was determined polarimetrically, comparing state three and state four rates,respiratory control index (RCI) and ADPO ratios. No significant alteration in these parameters was observed in the endotoxic state. Phase II of the study was designed to investigate mitochondrial function in a bacterial peritonitis rat model. Both liver and skeletal muscle mitochondrial function were determined to control for possible alterations in liver metabolism. Neither muscle nor liver mitochondria exhibited functional impairment during sepsis. We conclude from this study that neither endotoxemia nor peritonitis selectively “kills” mitochondria as previously suggested. 0 1986 Academic
Press, Inc.
INTRODUCTION
METHODS
Despite advances in antimicrobial therapy and intensive care, sepsis remains a paramount cause of surgical mortality. Studies have revealed a global metabolic dysfunction unique to the septic state-defective oxygen utilization at the cellular level despite apparently adequate oxygen delivery [3]. The concept of a selective mitochondrial injury has been proposed to account for this metabolic dysfunction. Previous studies utilizing endotoxic models in rats have suggested such an impairment in mitochondrial function [ lo- 131. However, many investigators have been unable to demonstrate mitochondrial dysfunction in various bacterial models of sepsis [2,5,6]. These latter studies have examined liver mitochondrial function. In an effort to determine whether endotoxemia or bacterial models of sepsis produced mitochondrial dysfunction, a two phase study was undertaken. In Phase I, skeletal muscle mitochondrial function in an endotoxic model was examined. Phase II was designed to determine whether mitochondrial dysfunction occurred in bacterial peritonitis, either in liver or in skeletal muscle tissues. 0022-4804/86
$1.50
Copyright 0 1986 by Academic Pres, Inc. AI1 right.3 of reproduction in any form reserved.
514
Phase I
Seventeen male Sprague-Dawley rats (Charles River) weighing 400-500 g were randomly assigned to control or septic groups following placement of a jugular venous catheter under methoxyfluorane anesthesia. Three milliliters of blood was withdrawn for study at the time of catheterization. Rats in the septic group received an intraperitoneal injection of Escherichia coli endotoxin (Difco Laboratories) 4 mg/ 100 g body wt. A constant infusion of Lactated Ringers solution (60 cc/24 hr) was maintained per catheter during the study period. Animals were reanesthetized 18 hr following endotoxin injections and skeletal muscle of the hind limb was immediately harvested as described below. Phase II
Twenty male Sprague-Dawley rats (Charles River) weighing 250-350 g were randomized to control or septic groups following placement of a jugular venous catheter under pentobarbital anesthesia. Two milliliters of blood was withdrawn for study at the time of catheter-
GELLER,
JANKAUSKAS,
AND
KIRKPATRICK:
ization. Rats in the septic group received an intraperitoneal injection of 0.5 cc of a pooled fecal inoculum mixed with barium sulfate adjuvant as described by Lang et al. [7]. Control rats received an intraperitoneal injection of a sterilized inoculum with barium sulfate adjuvant. All rats received a constant intravenous infusion of 8% dextrose with Lactated Ringers solution (60 cc/24 hr). Rats were sacrificed at 24 hr by decapitation; those dying prior to completion of the study period were replaced in series using the same randomization schedule. Liver and hind limb skeletal muscle were harvested immediately and processed as described below.
MITOCHONDRIAL
DEATH
515
sisting of 100 mhf KCl, 500 mA4 Tris-HCl
(pH 7.5), 0.2 m&f ATP, 1 mM MgClz, 0.2
mM EDTA, and centrifuged at 3500g for 10 min. This step was repeated and the final pellet suspended in 0.15 M KC1 yielding approximately 3.5 mg protein per milliliter. Respiratory measurements. Oxygen uptake was measured polarographically in a closed temperature-controlled (30°C) chamber using a Clark oxygen electrode (Yellow Springs Instrument Co.) and Gilson Oxygraph (Gilson Medical Electronics). The assay medium contained a solution of 0.25 M sucrose, 0.5 M Tris-HCl (pH 7.5), and 200 rnit4 K phosphate. Mitochondrial suspension (50 ~1) was added to the assay medium and either 5 ~1 of 1 M Preparation of Mitochondria succinate or 5 ~1 of 1 M pyruvate added as Liver. Immediately following sacrifice, the substrate. Pyruvate assayswere primed with 5 liver was totally excised and immersed in ice- ~1 of 0.5 M malate. The final volume of the cold 0.25 M sucrose. The tissue was weighed assay medium was 1.7 ml. Measurements were (approximately 10 g tissue), coarsely cut-up, performed both in the presence and absence and homogenized in a chilled medium con- of EDTA in the assay medium. taining 0.25 M sucrose with 0.5 mM diamiThe rate of State 3 respiration was deternoethanetetraacetic acid (EDTA). The ho- mined following the addition of 300 nmole of mogenate was centrifuged at 600g for 10 min. ADP. State 3 is defined as mitochondrial resThe supernatant was decanted and centrifuged piration in the presence of oxygen, substrate, at 15,OOOgfor 5 minutes. The pellet then pro- and ADP. It represents respiration coupled to duced by these steps was suspended in 20 ml ATP synthesis. State 4 is respiration in the 0.25 M sucrose solution and spun at 15,OOOg presence of substrate and oxygen but lacking for 5 min. The final mitochondrial pellet was ADP; it is not linked to ATP synthesis. The suspended in 0.25 M sucrose to yield approxState 4 respiration rate was measured after the imately 15 mg protein per milliliter as deter- expenditure of ADP as recommended by mined by the method of Lowry [8]. Chance [ 11. The respiratory control index Skeletal muscle. Skeletal muscle mitochon(RCI) was obtained by the ratio of State 3 to dria were isolated in the manner described by State 4 respiratory rates for each addition of Makinen and Lee [9]. Five to six grams of ADP and the average RCI calculated for each skeletal muscle was immediately harvested complete run. The ADP/O was calculated as from the hind limb following sacrifice and described by Estabrook [4]. Results were stasuspended in a cold isolation medium of 0.15 tistically analyzed by Student’s t test. M KCl, 50 mM Tris-HCl (pH 7.5), 1 mM RESULTS ATP, 5 mM MgCl*, 1 mM EDTA. The tissue was digested with a Bacillus subtilis crystalline During Phase I of this study, 7 of the 10 proteinase (Nagarse, Sigma Laboratories) so- rats survived the endotoxic challenge for 18 lution (5 mg/g tissue) for 7 min and mechanhr and were available for study. The remaining ically homogenized (Ultra-Turrax disintegrathree (30%) died before 18 hr. At sacrifice, the tor for 15 set). The homogenate was centrisurviving rats exhibited piloerection, tachyfuged at 600g for 10 min and the supernatant pnea, and diarrhea; two animals appeared was filtered and spun at 14,000g for 10 min. prostrate. Table 1 summarizes the mitochonThe pellet was suspended in a medium con- drial assay of the harvested skeletal muscle of
516
JOURNAL OF SURGICAL RESEARCH: VOL. 40, NO. 5, MAY 1986 TABLE 1
TABLE 3
EFFECTOFENDOTOXEMIAONOXIDATIVE PH~~~HOR~LATION~FRATSKELETAL MUSCLE MIT~CHONDRIA
EFFECTOFFECALFTRITONTISONOXIDATIVE F'HOSPHOR~ATIONOFRATLIVERMITOCHONDR~A n atoms O2 uptake/min/mg protein
n atoms Oz uptake/min/mg protein
Control mean SD Endotoxic mean SD
State 3
State 4
RCI
ADP:O
16.1 +4.0
4.0 kO.66
5.1 20.79
2.48 kO.12
19.8 +2.5
4.0 kO.82
4.4 +0.70
2.51 +0.12
septic and control animals. No statistically significant difference was observed between these groups for State 3 or State 4 respiratory rates, RCI, or ADP:O ratios. In Phase II of the study 2 of 12 rats who developed fecal peritonitis were dead at 24 hr (17%). (In a parallel study using this same model, carotid artery catheters were utilized to document that septic rats were normotensive at 24 hr (BP 124 +- 16 Torr.) At sacrifice, the rats demonstrated tachypnea, piloerection, and diarrhea. Results of liver and skeletal muscle mithochondria assay for septic and control rats are summarized in Tables 2 and 3. No statistically significant differences were seen between septic and control groups regardless of whether endotoxin (Phase I) or fecal peritonitis (Phase II) produced the septic insult. TABLE 2 EFFECTOFFECALPERITONTISONOXIDATIVE PH~~PHORYLATIONOFRATSKELETAL MU~CLEMITOCHONDRIA n atoms Oz uptake/min/mg protein
Control mean SD Septic mean SD
State 3
State 4
RCI
16.6 +0.91 15.7 kO.66
2.76 20.27 2.40 f0.25
6.42 kO.38 7.39 *0.72
ADl?O 2.52 +0.07 2.51 +o. 10
State 3
State 4
RCI
ADPO
Pyruvate Control mean SD Septic mean SD
3.14 f0.75 3.70 +1.6
1.33 20.54 1.43 kO.72
2.59 20.70 2.69 +0.45
2.09 +0.40 2.15 +0.30
Succinate Control mean SD Septic mean SD
10.21 +1.3 9.93 +2.5
2.54 +0.57 2.74 +1.2
4.22 Zk1.1 3.85 +0.94
1.65 +0.13 1.62 +0.19
DISCUSSION
Previous investigations have suggested an inhibition in energy production in both liver and skeletal muscle mitochondria in endotoxic rats [lo-l 31. If true, then the basic cause of the energy deficit of sepsis would be a “faulty furnace.” If false, then the cause of this energy deficit must lie elsewhere. The present study was undertaken to confirm or reject the concept that either endotoxin (Phase I) or bacterial peritonitis (Phase II) produce a selective disruption of mitochondrial function. The results of the present study are quite striking-at a time when the animal is clinically septic and many of its cohorts are dead, mitochondrial function is normal. While at variance with other earlier reports, the reproducibility of our findings in many animals using two different septic models suggests their validity. One factor, tissue perfusion, might partially reconcile the differences between our studies and the earlier reports. We looked at this parameter in a small subset of animals (Phase II) by monitoring carotid pressure until sacrifice. No hypotension was noted. Differences in methodology might also explain the differences in results. Previous investigators who used various bacterial sepsis models did not find the impairment of liver mitochondrial respiration described by others
GELLER,
JANKAUSKAS,
AND KIRKPATRICK:
MITOCHONDRIAL
DEATH
517
using the endotoxic model. The present study, tochondria was seen in either model. Liver as well, demonstrates nornml respiration of mitochondrial function was unimpaired in the peritonitis model as well. The results suggest liver mitochondria in bacterial sepsis. Our findings when viewed in the context of that the metabolic derangements of sepsis are previous work raise several points of discus- not due to any impairment of mitochondrial sion. Though endotoxemia may impair liver oxidative phosphorylation. mitochondrial respiration, this effect is not a ACKNOWLEDGMENT generalized one, since the mitochondria of endotoxic rat skeletal muscle appear to function The authors thank Miss Haydee Provido for her dedinormally. Further, no such impairment of ei- cation and expertise in assisting with these studies. ther liver or skeletal muscle mitochondria has been seen in the various septic models. This REFERENCES raises the question of the relation of endotox1. Chance, B., Williams, G. R., McLean, A. P., et al. emia to clinical sepsis. In view of the current The respiratory chain and oxidative phosphorylation. study and the body of literature cited, it would Adv. Enzymol. 17: 65, 1956. appear that the septic state does not cause the 2. Decker, G. A., Daniel, A. M., Blevings, S., et al. Effect of peritonitis on mitochondrial respiration, J. Surg. selective mitochondrial injury proposed to acRes. 11: 528, 1971. count for the global metabolic dysfunction 3. Duff, J. H., Groves, A. C., McLean, A. P., et al. Deseen in sepsis. fective oxygen consumption in septic shock. Surg. It must be noted, however, that this does Gynecol. Obstet. 128: 1051, 1969. not mean a fundamental defect does not exist 4. Estabrook, R. W. Mitochondrial respiratory control and the polarographic measurement of ADP:O ratios. in the metabolism of the septic organism. On In R. W. Estabrook and M. E. Pullman (Eds.), Meththe contrary, such a flaw or constellation of ads in Enzymology, 1967. Vol. 10, P. 41. flaws must indeed account for the altered bio5. Fry, D. E., Silva, B. B., Rink, R. D., et al. Hepatic chemical state known to accompany sepsis. mitochondrial function in intraperitoneal sepsis.Rev. The studies considered address only the ability Surg. 34 214, 1977. 6. Fry, D. E., Silva, B. B., Rink, R. D., et al. Hepatic of mitochondria to perform oxidative phoscellular hypoxia in murine peritonitis. Surgery 85: 652, phorylation in an efficient and tightly regulated 1979. manner when provided in vitro with various I. Lang, C. L., Bagby, G. J., Bomside, G. H., et al. Sushigh-energy substrates. This ability appears tained hypermetabolic sepsisin rats: Characterization unimpaired in sepsis. However, the ability of of the model. J. Surg. Res. 35: 201, 1983. 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and the septic organism to perform the intermeRandall, R. J. Protein measurement with the folin diate metabolism necessary to provide the miphenol reagent. J. Biol. Chem. 193: 265, 195 1. tochondria with high-energy substrate, and the 9. Makinen, M. W., and Lee, C. Biochemical studies of ability to process such substrates into a form skeletal muscle mitochondria. Arch. Biochem. Biosuited to entry into the oxidative phosphoryphys. 126: 75, 1968. lation chain remain to be investigated. Such 10. Mela, L., Bacalzo, L. V., Jr., and Miller, L. D. Defective oxidative metabolism of rat liver motochondria investigations are currently underway in our in hemorrhagic and endotoxin shock. Amer. .I Physiol. laboratory. 220: 571, 1971. SUMMARY
In a two phase study, mitochondrial respiration of rat hind limb skeletal muscle was studied in both endotoxic and fecal peritonitis rat models. Liver mitochondrial function was also studied in the peritonitis model. No impairment of function of skeletal muscle mi-
11. Mela, L., Nicholas, G. G., and Miller, L. D. Inhibition of mitochondrial energy metabolism in hypovolemic and endotoxin shock. In T. M. Glen (Ed.), Steroids and Shock. Baltimore: University Park Press, 1974. P. 301. 12. Schumer, W., Dasgupta, T. K., Moss, G. S., and Nyhus, L. M. Effect of endotoxemia on liver cell mitochondria in man. Ann. Surg. 171: 875, 1970. 13. Schumer, W., Erve, P. R., and Obemolte, R. P. Endotoxemic effect on cardiac and skeletal muscle mitochondria. Surg. Gynecol. Obstet. 133: 433, 197 1.
M.D.,S.
Death in Sepsis: A Failed Concept JANKAUSKAS,M.D.,ANDJ.KIRKPATRICK,M.D.
Wayne State University, Department of Surgery, Detroit, Michigan 48202 Presented at the Annual Meeting of the Association for Academic Surgery, Cincinnati, Ohio, November 10-13, 1985 The concept of early selective mitochondrial injury has been proposed to explain the global metabolic dysfunction observed in the septic state. A two phase study was undertaken to test the validity of this hypothesis. In the initial phase, an endotoxin shock model was employed in the rat to delineate the function of skeletal muscle mitochondria. Mitochondrial function was determined polarimetrically, comparing state three and state four rates,respiratory control index (RCI) and ADPO ratios. No significant alteration in these parameters was observed in the endotoxic state. Phase II of the study was designed to investigate mitochondrial function in a bacterial peritonitis rat model. Both liver and skeletal muscle mitochondrial function were determined to control for possible alterations in liver metabolism. Neither muscle nor liver mitochondria exhibited functional impairment during sepsis. We conclude from this study that neither endotoxemia nor peritonitis selectively “kills” mitochondria as previously suggested. 0 1986 Academic
Press, Inc.
INTRODUCTION
METHODS
Despite advances in antimicrobial therapy and intensive care, sepsis remains a paramount cause of surgical mortality. Studies have revealed a global metabolic dysfunction unique to the septic state-defective oxygen utilization at the cellular level despite apparently adequate oxygen delivery [3]. The concept of a selective mitochondrial injury has been proposed to account for this metabolic dysfunction. Previous studies utilizing endotoxic models in rats have suggested such an impairment in mitochondrial function [ lo- 131. However, many investigators have been unable to demonstrate mitochondrial dysfunction in various bacterial models of sepsis [2,5,6]. These latter studies have examined liver mitochondrial function. In an effort to determine whether endotoxemia or bacterial models of sepsis produced mitochondrial dysfunction, a two phase study was undertaken. In Phase I, skeletal muscle mitochondrial function in an endotoxic model was examined. Phase II was designed to determine whether mitochondrial dysfunction occurred in bacterial peritonitis, either in liver or in skeletal muscle tissues. 0022-4804/86
$1.50
Copyright 0 1986 by Academic Pres, Inc. AI1 right.3 of reproduction in any form reserved.
514
Phase I
Seventeen male Sprague-Dawley rats (Charles River) weighing 400-500 g were randomly assigned to control or septic groups following placement of a jugular venous catheter under methoxyfluorane anesthesia. Three milliliters of blood was withdrawn for study at the time of catheterization. Rats in the septic group received an intraperitoneal injection of Escherichia coli endotoxin (Difco Laboratories) 4 mg/ 100 g body wt. A constant infusion of Lactated Ringers solution (60 cc/24 hr) was maintained per catheter during the study period. Animals were reanesthetized 18 hr following endotoxin injections and skeletal muscle of the hind limb was immediately harvested as described below. Phase II
Twenty male Sprague-Dawley rats (Charles River) weighing 250-350 g were randomized to control or septic groups following placement of a jugular venous catheter under pentobarbital anesthesia. Two milliliters of blood was withdrawn for study at the time of catheter-
GELLER,
JANKAUSKAS,
AND
KIRKPATRICK:
ization. Rats in the septic group received an intraperitoneal injection of 0.5 cc of a pooled fecal inoculum mixed with barium sulfate adjuvant as described by Lang et al. [7]. Control rats received an intraperitoneal injection of a sterilized inoculum with barium sulfate adjuvant. All rats received a constant intravenous infusion of 8% dextrose with Lactated Ringers solution (60 cc/24 hr). Rats were sacrificed at 24 hr by decapitation; those dying prior to completion of the study period were replaced in series using the same randomization schedule. Liver and hind limb skeletal muscle were harvested immediately and processed as described below.
MITOCHONDRIAL
DEATH
515
sisting of 100 mhf KCl, 500 mA4 Tris-HCl
(pH 7.5), 0.2 m&f ATP, 1 mM MgClz, 0.2
mM EDTA, and centrifuged at 3500g for 10 min. This step was repeated and the final pellet suspended in 0.15 M KC1 yielding approximately 3.5 mg protein per milliliter. Respiratory measurements. Oxygen uptake was measured polarographically in a closed temperature-controlled (30°C) chamber using a Clark oxygen electrode (Yellow Springs Instrument Co.) and Gilson Oxygraph (Gilson Medical Electronics). The assay medium contained a solution of 0.25 M sucrose, 0.5 M Tris-HCl (pH 7.5), and 200 rnit4 K phosphate. Mitochondrial suspension (50 ~1) was added to the assay medium and either 5 ~1 of 1 M Preparation of Mitochondria succinate or 5 ~1 of 1 M pyruvate added as Liver. Immediately following sacrifice, the substrate. Pyruvate assayswere primed with 5 liver was totally excised and immersed in ice- ~1 of 0.5 M malate. The final volume of the cold 0.25 M sucrose. The tissue was weighed assay medium was 1.7 ml. Measurements were (approximately 10 g tissue), coarsely cut-up, performed both in the presence and absence and homogenized in a chilled medium con- of EDTA in the assay medium. taining 0.25 M sucrose with 0.5 mM diamiThe rate of State 3 respiration was deternoethanetetraacetic acid (EDTA). The ho- mined following the addition of 300 nmole of mogenate was centrifuged at 600g for 10 min. ADP. State 3 is defined as mitochondrial resThe supernatant was decanted and centrifuged piration in the presence of oxygen, substrate, at 15,OOOgfor 5 minutes. The pellet then pro- and ADP. It represents respiration coupled to duced by these steps was suspended in 20 ml ATP synthesis. State 4 is respiration in the 0.25 M sucrose solution and spun at 15,OOOg presence of substrate and oxygen but lacking for 5 min. The final mitochondrial pellet was ADP; it is not linked to ATP synthesis. The suspended in 0.25 M sucrose to yield approxState 4 respiration rate was measured after the imately 15 mg protein per milliliter as deter- expenditure of ADP as recommended by mined by the method of Lowry [8]. Chance [ 11. The respiratory control index Skeletal muscle. Skeletal muscle mitochon(RCI) was obtained by the ratio of State 3 to dria were isolated in the manner described by State 4 respiratory rates for each addition of Makinen and Lee [9]. Five to six grams of ADP and the average RCI calculated for each skeletal muscle was immediately harvested complete run. The ADP/O was calculated as from the hind limb following sacrifice and described by Estabrook [4]. Results were stasuspended in a cold isolation medium of 0.15 tistically analyzed by Student’s t test. M KCl, 50 mM Tris-HCl (pH 7.5), 1 mM RESULTS ATP, 5 mM MgCl*, 1 mM EDTA. The tissue was digested with a Bacillus subtilis crystalline During Phase I of this study, 7 of the 10 proteinase (Nagarse, Sigma Laboratories) so- rats survived the endotoxic challenge for 18 lution (5 mg/g tissue) for 7 min and mechanhr and were available for study. The remaining ically homogenized (Ultra-Turrax disintegrathree (30%) died before 18 hr. At sacrifice, the tor for 15 set). The homogenate was centrisurviving rats exhibited piloerection, tachyfuged at 600g for 10 min and the supernatant pnea, and diarrhea; two animals appeared was filtered and spun at 14,000g for 10 min. prostrate. Table 1 summarizes the mitochonThe pellet was suspended in a medium con- drial assay of the harvested skeletal muscle of
516
JOURNAL OF SURGICAL RESEARCH: VOL. 40, NO. 5, MAY 1986 TABLE 1
TABLE 3
EFFECTOFENDOTOXEMIAONOXIDATIVE PH~~~HOR~LATION~FRATSKELETAL MUSCLE MIT~CHONDRIA
EFFECTOFFECALFTRITONTISONOXIDATIVE F'HOSPHOR~ATIONOFRATLIVERMITOCHONDR~A n atoms O2 uptake/min/mg protein
n atoms Oz uptake/min/mg protein
Control mean SD Endotoxic mean SD
State 3
State 4
RCI
ADP:O
16.1 +4.0
4.0 kO.66
5.1 20.79
2.48 kO.12
19.8 +2.5
4.0 kO.82
4.4 +0.70
2.51 +0.12
septic and control animals. No statistically significant difference was observed between these groups for State 3 or State 4 respiratory rates, RCI, or ADP:O ratios. In Phase II of the study 2 of 12 rats who developed fecal peritonitis were dead at 24 hr (17%). (In a parallel study using this same model, carotid artery catheters were utilized to document that septic rats were normotensive at 24 hr (BP 124 +- 16 Torr.) At sacrifice, the rats demonstrated tachypnea, piloerection, and diarrhea. Results of liver and skeletal muscle mithochondria assay for septic and control rats are summarized in Tables 2 and 3. No statistically significant differences were seen between septic and control groups regardless of whether endotoxin (Phase I) or fecal peritonitis (Phase II) produced the septic insult. TABLE 2 EFFECTOFFECALPERITONTISONOXIDATIVE PH~~PHORYLATIONOFRATSKELETAL MU~CLEMITOCHONDRIA n atoms Oz uptake/min/mg protein
Control mean SD Septic mean SD
State 3
State 4
RCI
16.6 +0.91 15.7 kO.66
2.76 20.27 2.40 f0.25
6.42 kO.38 7.39 *0.72
ADl?O 2.52 +0.07 2.51 +o. 10
State 3
State 4
RCI
ADPO
Pyruvate Control mean SD Septic mean SD
3.14 f0.75 3.70 +1.6
1.33 20.54 1.43 kO.72
2.59 20.70 2.69 +0.45
2.09 +0.40 2.15 +0.30
Succinate Control mean SD Septic mean SD
10.21 +1.3 9.93 +2.5
2.54 +0.57 2.74 +1.2
4.22 Zk1.1 3.85 +0.94
1.65 +0.13 1.62 +0.19
DISCUSSION
Previous investigations have suggested an inhibition in energy production in both liver and skeletal muscle mitochondria in endotoxic rats [lo-l 31. If true, then the basic cause of the energy deficit of sepsis would be a “faulty furnace.” If false, then the cause of this energy deficit must lie elsewhere. The present study was undertaken to confirm or reject the concept that either endotoxin (Phase I) or bacterial peritonitis (Phase II) produce a selective disruption of mitochondrial function. The results of the present study are quite striking-at a time when the animal is clinically septic and many of its cohorts are dead, mitochondrial function is normal. While at variance with other earlier reports, the reproducibility of our findings in many animals using two different septic models suggests their validity. One factor, tissue perfusion, might partially reconcile the differences between our studies and the earlier reports. We looked at this parameter in a small subset of animals (Phase II) by monitoring carotid pressure until sacrifice. No hypotension was noted. Differences in methodology might also explain the differences in results. Previous investigators who used various bacterial sepsis models did not find the impairment of liver mitochondrial respiration described by others
GELLER,
JANKAUSKAS,
AND KIRKPATRICK:
MITOCHONDRIAL
DEATH
517
using the endotoxic model. The present study, tochondria was seen in either model. Liver as well, demonstrates nornml respiration of mitochondrial function was unimpaired in the peritonitis model as well. The results suggest liver mitochondria in bacterial sepsis. Our findings when viewed in the context of that the metabolic derangements of sepsis are previous work raise several points of discus- not due to any impairment of mitochondrial sion. Though endotoxemia may impair liver oxidative phosphorylation. mitochondrial respiration, this effect is not a ACKNOWLEDGMENT generalized one, since the mitochondria of endotoxic rat skeletal muscle appear to function The authors thank Miss Haydee Provido for her dedinormally. Further, no such impairment of ei- cation and expertise in assisting with these studies. ther liver or skeletal muscle mitochondria has been seen in the various septic models. This REFERENCES raises the question of the relation of endotox1. Chance, B., Williams, G. R., McLean, A. P., et al. emia to clinical sepsis. In view of the current The respiratory chain and oxidative phosphorylation. study and the body of literature cited, it would Adv. Enzymol. 17: 65, 1956. appear that the septic state does not cause the 2. Decker, G. A., Daniel, A. M., Blevings, S., et al. Effect of peritonitis on mitochondrial respiration, J. Surg. selective mitochondrial injury proposed to acRes. 11: 528, 1971. count for the global metabolic dysfunction 3. Duff, J. H., Groves, A. C., McLean, A. P., et al. Deseen in sepsis. fective oxygen consumption in septic shock. Surg. It must be noted, however, that this does Gynecol. Obstet. 128: 1051, 1969. not mean a fundamental defect does not exist 4. Estabrook, R. W. Mitochondrial respiratory control and the polarographic measurement of ADP:O ratios. in the metabolism of the septic organism. On In R. W. Estabrook and M. E. Pullman (Eds.), Meththe contrary, such a flaw or constellation of ads in Enzymology, 1967. Vol. 10, P. 41. flaws must indeed account for the altered bio5. Fry, D. E., Silva, B. B., Rink, R. D., et al. Hepatic chemical state known to accompany sepsis. mitochondrial function in intraperitoneal sepsis.Rev. The studies considered address only the ability Surg. 34 214, 1977. 6. Fry, D. E., Silva, B. B., Rink, R. D., et al. Hepatic of mitochondria to perform oxidative phoscellular hypoxia in murine peritonitis. Surgery 85: 652, phorylation in an efficient and tightly regulated 1979. manner when provided in vitro with various I. Lang, C. L., Bagby, G. J., Bomside, G. H., et al. Sushigh-energy substrates. This ability appears tained hypermetabolic sepsisin rats: Characterization unimpaired in sepsis. However, the ability of of the model. J. Surg. Res. 35: 201, 1983. 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and the septic organism to perform the intermeRandall, R. J. Protein measurement with the folin diate metabolism necessary to provide the miphenol reagent. J. Biol. Chem. 193: 265, 195 1. tochondria with high-energy substrate, and the 9. Makinen, M. W., and Lee, C. Biochemical studies of ability to process such substrates into a form skeletal muscle mitochondria. Arch. Biochem. Biosuited to entry into the oxidative phosphoryphys. 126: 75, 1968. lation chain remain to be investigated. Such 10. Mela, L., Bacalzo, L. V., Jr., and Miller, L. D. Defective oxidative metabolism of rat liver motochondria investigations are currently underway in our in hemorrhagic and endotoxin shock. Amer. .I Physiol. laboratory. 220: 571, 1971. SUMMARY
In a two phase study, mitochondrial respiration of rat hind limb skeletal muscle was studied in both endotoxic and fecal peritonitis rat models. Liver mitochondrial function was also studied in the peritonitis model. No impairment of function of skeletal muscle mi-
11. Mela, L., Nicholas, G. G., and Miller, L. D. Inhibition of mitochondrial energy metabolism in hypovolemic and endotoxin shock. In T. M. Glen (Ed.), Steroids and Shock. Baltimore: University Park Press, 1974. P. 301. 12. Schumer, W., Dasgupta, T. K., Moss, G. S., and Nyhus, L. M. Effect of endotoxemia on liver cell mitochondria in man. Ann. Surg. 171: 875, 1970. 13. Schumer, W., Erve, P. R., and Obemolte, R. P. Endotoxemic effect on cardiac and skeletal muscle mitochondria. Surg. Gynecol. Obstet. 133: 433, 197 1.