Endotoxemia and enhanced generation of oxygen radicals by neutrophils from patients undergoing cardiopulmonary bypass

J THORAC CARDIOVASC SURG 1989;98:381-5

Endotoxemia and enhanced generation of oxygen radicals by neutrophils from patients undergoing cardiopulmonary bypass Plasma endotoxin concentrations and oxidative burst response of peripheral blood polymorphonuclear leukocytes were examined in 12 patients undergoing coronary artery bypass. The measurements were made just before the operation, 5 minutes after removal of the aortic crosscIamp, and 24 hours after the operation. Endotoxin was quantitated by a combination of a sensitive Limulus amebocyte lysate assay and rocket immunoelectrophoresis measuring picogram amounts of endotoxin. Peripheral blood neutrophils were purified by a two-step dextran sedimentation and metrizoate sodium FicoU (Lymphoprep., Nyegaard, Oslo, Norway) centrifugation. The oxidative burst response of tbese ceUs was measured for their ability to generate superoxide anion and was determined by a cytochrome c reduction assay. Preoperatively, aU the plasma samples except one were free of endotoxin. The endotoxin levels reacbed 100 pg/ml 5 minutes after removal of the aortic crosscIamp, and except in one sample they had decreased 24 bours after the operation. Studies on the generation of superoxide by neutropbils showed a decline in the response 5 minutes after removal of the aortic crosscIamp and an enhancement of the response to f-Met-Leu-Phe by ceUs obtained from 11 of 12 patients 24 hours postoperatively. In vitro addition of bacterial lipopolysaccharide to blood from bealthy individuals also enhanced the superoxide response of the neutrophils. We conclude that during cardiopulmonary bypass the circulating blood is contaminated by endotoxin and the neutropbils are primed for enhanced generation of oxygen radicals. The released oxygen radicals may be involved in the tissue damage observed in these patients.

Arsalan Kharazmi, PhD," Lars W. Andersen, MD,b Leif Baek, MD,' Niels H. Valerius, MD, PhD," Michael Laub, MD,b and Jens P. Rasmussen, MD, PhD,b

Copenhagen. Denmark

During recent years there has been considerable evidence in the literature suggesting that oxygen radicals may be responsible for cell and tissue injury in the lung,1,2 intestine,' liver,' and heart.i-" In vivo the oxygen radicals are generated either by simple chemical reactions within living cells? or by activation of the polymorphonuclear leukocytes and monocyte macrophages.v? Endotoxins have been known

From the State Serum Institute. Department of Clinical Microbiology, Rigshospitalct.: Departments of Anesthesiology and Thoracic Surgery. Gentofte Hospital," and Department of Infectious Diseases. Rigshospitalct,' Copenhagen, Denmark. Received for publication Jan. II. 1988. Accepted for publication Dec. 15. 1988. Address for reprints: A. Kharazmi, Department of Clinical Microbiology. Rigshospitalet 7806. Tagensvej 20. DK-2200. Copenhagen 1\. Denmark.

12/1/10283

to be potent activators of the oxidative burst response in phagocytic cellS.ID- 12 Endogenous vascular endothelial oxygen radical generation by endotoxin, as well as neutrophil-mediated inflammatory tissue damage, is believed to occur in endotoxemic states such as in adult respiratory distress syndrome associated with infection," in animal models of adult respiratory distress syndrome," and in endotoxic shock. 15. 16 A temperature rise after cardiac operations is common, and several mechanisms have been proposed.P:" It is known that the host defense system in patients undergoing major operations is temporarily impaired.'? We 2D have previously demonstrated the presence of endotoxin in the blood of patients undergoing coronary artery bypass. The present study was designed to study the oxidative burst response of blood neutrophils from these patients, to correlate the response with serum endotoxin levels, and to make comparisons with the in vitro studies. 38 I

The Journal of

3 8 2 Kharazmi et al.

Thoracic and Cardiovascular Surgery

Table I. Endotoxin concentrations* in the plasma (Aj and total blood leukocyte count given as I X 109 cellsfl. (Bj of patients undergoing cardiopulmonary bypass

13.0

.,

a; 12.0 o ~O

~

5 min

11.0

after

0.

-e

" "e o

."

o

"E

10.0

Before opera/ion

9.0

0

1: o 0

;:,

8.0

o

0

E

c:

7.0

5.0

4.0

3.0

A

B

A

B

A

B

I 2 3

0 0 0 0 10 0 0 0 0 0 0 0 0

9.0 7.9

100 100 50 50 100 100 100 100 0 0 50 50 0

4.9 3.9 1\;0 0.0 4.4 5.1 4.9 4.7 0.0

25 1000 0 12.5

10.0 11.0 :\0 13.7 9.2 7.8

Buffer

2.0

:!4 IIr after opera/ion

Patients

4 5 6 7 8 9 10 II 12

6.0

removal oj aortic crosse/amp

I\D 7.0 7.8 12.0 6.8 8.5 7.6

:\ID 7.5 6.0

:\:D 3.0 3.8

~D

50

1\iD 25 0 0 50 0 0

7J 8.0 7.3

:\D 8.9 9.2

:" D. :"ot determined.

"Given as picograrns Escherichia coli 055:B5 standard LPS equivalent per milliliter plasma.

1.0 A

B

C

Fig. 1. Neutrophil superoxide response of cells obtained from patients before operation (0), 5 minutes after removal of aortic crossclamp (e), and 24 hours after the operation (..i). Panel A shows the response of 106 cells/min to PMA; panel B. the response of 106 cells/min to f-Met-Leu-Phe; and panel C. the total response of 106 cells to f-Met-Leu-Phe in 15 minutes. The superoxide response is given as nanomoles cytochrome c reduced per 106 cells.

Patients The study comprised 12 patients undergoing coronary artery bypass. The mean age was 56 years (range 47 to 70). None of the patients had preoperative signs of infections clinically or biochemically. All patients were given cefuroxime (Zinaceff) as a prophylactic antibiotic during the operation and for the following 4 days. It was not possible to demonstrate any preoperative or postoperative hemodynamic complications. All the patients had an elevated temperature postoperatively, in the range of 37.5° to 38.2° C, except for patient 2, who had a temperature of 39° C. which declined to normal limits within 7 days. None of the blood cultures was positive for bacteria or other microorganisms. Operative procedure. The operation consisted of median sternotomy and placement of saphenous vein grafts. Anesthetic management was uniform in all patients. Cardiopulmonary bypass was performed with systemic hypothermia (25 0 C). A bubble oxygenator (BOS-I05, Bentley Laboratories, Inc., subsidiary of Baxter Healthcare Corporation, Irvine, Calif.)

and a Pemco heart-lung machine (Pemco, lnc., Cleveland. Ohio) with an Ohio roller pump (Ohmcda. Madison, Wis.) and nonpulsatile flow were used. A cardiotomy reservoir (BCR-35oo, Bentley), a blood infusion filter (PFF-I 00, Bentley), and a biologic gas-line filter (PF-302, Bentley) were placed in the extracorporeal circuit. All tubings in the circuit were made of polyvinylchloride except the tubing in the arterial roller pump box, which was made of silicone rubber. Cold cardioplegic fluid was used for myocardial protection combined with external cooling. The extracorporeal circuit was primed with Ringer's lactate 3000 ml (the dispensary, Gentofte Hospital), human albumin 200 ml (Nordic Gcntofte, Copenhagen, Denmark), sodium bicarbonate 60 ml, and antifoaming agent 2 ml. The mean perfusion time was 157 minutes (range 100 to 240). The radial artery was cannulated before anesthesia and a Swan-Ganz catheter (Baxter Edwards Divisions, Irvine. Calif.) was inserted aseptically via the internal jugular vein after induction of anesthesia and endotracheal intubation. Postoperatively the patients were admitted intubated to the intensive care unit and the tubing was connected to a ventilator. When the clinical situation and blood gas analyses were satisfactory, the patients were extubated within 16 hours postoperatively. Endotoxin measurement. Plasma endotoxin was quantitated by a combination of a sensitive Limulus amebocyte lysate (LAL) assay and rocket immunoelectrophoresis, as previously described." The sensitivity of the test is I pgj'ml measured by Escherichia coli 055:B5 standard LPS (M.A. Bioproducts. Walkersville, Md.). To 500 III plasma was added 500 III pyrogen-free water, and the solution was boiled for 10 minutes.

Volume 98 Number 3

Endotoxemia

September 1989

383

Table n. Superoxide response to PMA and f-Met-Leu-Phe of neutrophils from 12 patients at three different periods

Stimuli

Before operation

5 min after removal of aortic crosse/amp

PMA f-Met-Leu-Phe (maximal rate) f-Met-Leu-Phe (total)

2.87 ± 0.25 1.88 ± 0.16 5.95 ± 0.61

2.79 ± 0.27 1.33±0.16* 3.86 ± 0.65t

24 hr after operation 2.62 ± 0.24 2.34 ± 0.27* 7.51 ± i.nr

The results arc given as mean ± standard error of the mean nanomoles cytochrome C reduced per 10' cells per minute. The total f-Met-Leu-Phe response represents nanornoles cytochrome C reduced per 10' cells in 15 minutes. 'f' <0.02.

tf' < 0.01.

After being boiled, the vials were centrifuged at 5000 g for 10 minutes. A 50 ILl portion of the supernatant was mixed with 50 ILl of LAL in test tubes and incubated for I hour in a 37° C water bath. The LAL reaction was stopped by addition of 0.7 ml of 0.2N HCI and incubated for 20 minutes at 22° C. Finally 0.7 ml of 0.2N NaOH was added to the test tubes to obtain neutral pH. From this solution 5 ILl was used for rocket immunoelectrophoresis. Electrophoresis was run with antibody against LAL incorporated in the gel. The gels were stained with Coomassie brilliant blue dye. Height of rockets was measured and the amount of endotoxin in the plasma was determined as standard LPS equivalent per milliliter. Blood neutrophils. Peripheral blood polymorphonuclear leukocytes (PMNs) were separated by a two-step dextran sedimentation and metrizoate sodium Ficoll (Lymphoprep., Nyegaard, Oslo, Norway) centrifugation. Remaining erythrocytes were removed by hypotonic lysis. More than 98% of the cells were neutrophils. The cells were suspended in Krebs Ringer solution containing 1% human serum albumin. Cell viability determiend by trypan blue dye exclusion was greater than 98%. Superoxide assay. Cytochrome c reduction assay was used as previously described." Briefly. 100 ILl of a 2 X 10' cells/rnl PMN suspension was activated with either formylated chemotactic peptide f-Met-Leu-Phe at a final concentration of 10- 6 mol/L or phorbol myristate acetate (PMA) at a final concentration of 20 ILg/ml in the presence of horse heart cytochrome c (4 mg/rnl), f-Met-Leu-Phe and PMA concentrations were chosen on the basis of the optimal responses obtained at our laboratory. The total volume was brought up to I ml with Krebs Ringer solution. The control plastic cuvette contained a similar number of cells and cytochrome c with no stimulant. The absorbance was determined at 550 nm in a spectrophotometer (Shimadzu, Double-Beam UV-190. Kyoto, Japan). Results were converted to nanomoles cytochrome c reduced by the extinction coefficient" E"o = 2.1 X 10· molr'cm". Statistics. The Wilcoxon rank sum test was used for statistical analysis of the data, and P values of less than 0.05 were considered significant.

Results The data on blood endotoxin concentrations and the total leukocyte counts are given in Table I. Preoperatively, blood from 11 patients was endotoxin free. In one patient, an endotoxin concentreation of 10

Table m. Superoxide response of granulocytes from blood after addition of standard LPS in vitro LPS

Stimulants

concentrations tpgfml;

100 50 10

o

PMA

2.55 2.25 2.22 2.18

± ± ± ±

0.47 0.47 0.56 0.30

f-Met-Leu-Phe

1.97*±0.15 1.68t ± 0.11 1.28 ± 0.09 1.13 ± 0.13

Peripheral blood obtained from healthy individuals was incubated with various concentrations of LPS for 60 minutes at 37 C. The granulocytes were then separated and tested for superoxidc anion production. The data are mean ± standard error of the mean from six experiments. The results are given as nanomoles cytochrome c reduced per 10' cells per minute.

*1' <0.01. tp

< 002.

pg/rnl was determined in the blood. Shortly after the beginning of total perfusion, a considerable amount of endotoxin was found in the blood of 10 of 12 patients. The level of endotoxin ranged from 50 to 100 pg/rnl 5 minutes after removal of the aortic crossclamp. Twentyfour hours postoperatively the endotoxin level had increased to 1000 pg/rnl in one patient, who had a temperature of 39° C, whereas it was lower than 50 pg/rnl in the others. The total leukocyte count dropped in all of the patients 5 minutes after removal of the aortic clamp. Twenty-four hours later the counts were similar to or slightly higher than the preoperative levels (Table I). Generation of oxygen radicals by purified neutrophils from the patients' blood was measured by production of superoxide anion by these cells after activation with either PMA or f-Met-Leu-Phe. The data are presented in Table II and Fig. I. The response to PMA of cells obtained before operation was similar to those obtained 5 minutes after removal of aortic crossclamp or 24 hours after the operation (2.87, 2.79, and 2.62 nmol/106 cells/min, respectively). On the other hand, the response

The Journal of

384

Thoracic and Cardiovascular

Kharazmi et al.

to f-Met-Leu-Phe was less in the cells obtained 5 minutes after removal of the aortic crossclamp (3.86 nmol/ 1O~ cells) and greater 24 hours after the operation (7.51 nmol/Iu" cells) as compared with those before the operation (5.95 nmol/LO" cells). The response to fMet-Leu-Phe was less in 11 of 12 patients 5 minutes after removal of the aortic crossclamp and was greater in 11 of 12 patients in the 24-hour postoperative samples as compared with those obtained preoperatively (Fig. I). In vitro experiments were performed to determine whether endotoxin in the blood of the patients was responsible for the observed neutrophil response. Comparable amounts of standard LPS were added to fresh blood drawn from healthy individuals. The blood was incubated for 60 minutes at 37° C and then the neutrophils were separated in the same way as patient cells. The superoxide production of these cells was measured after activation with PMA or f-Met-Leu-Phe. The results are shown in Table III. The PMA-induced response was unchanged, whereas at concentrations of 50 and 100 pg/rnl standard LPS the response to f-Met-Leu-Phe was significantly higher than the control response in the absence of LPS. Discussion

The present study demonstrates that blood from patients undergoing cardiopulmonary bypass contains considerable amounts of endotoxin and that the PMNs isolated from the peripheral blood of these patients are primed for enhanced release of oxygen radicals. The specificity of the LAL test used in this study has been thoroughly investigated and discussed previously." Hollow-fiber hemodialyzers containing cellulose-based membranes produce positive results with the LAL test." Our study indicated that LAL-reactive material was derived from cellulose and not pyrogens. However, we have recirculated pyrogen-free sterile water, obtained fresh from the factory, in the extracorporeal circuit and found only a small amount of LAL-reactive material (3 to 5 pgj'ml). We believe that the endotoxin found in the patients' blood is derived either from the priming fluids or from the gut (or both), as shown by our previous

study." Bacterial endotoxins at low concentrations are able to prime neutrophils and monocytes for enhanced release of oxygen radicals. 10. 11.1~ Therefore it is most likely that the enhanced release of superoxide anion by f-MetLeu-Phe-activated neutrophils from the patients in the present study is due to endotoxin in the blood after the operation. This hypothesis is further supported by the in vitro studies in which LPS is added to blood obtained

Surgery

from healthy individuals who have not had an operation. We were able to prime the f-Met-Leu-Phe-induced neutrophil response with 50 to 100 pg/rnl of LPS in vitro, concentrations similar to those measured in the blood from the patients. The enhanced superoxide response was observed in the cells obtained 24 hours after the operation, whereas it was decreased in the samples obtained 5 minutes after removal of the aortic crossclamp. The low response observed in this period might be due to the extensive tissue damage caused by surgical trauma or to anesthesia, or both." Monocytes and macrophages take up Ll'S but are not able to degrade and liberate most of the phagocytized LPS until after 48 hours." The liberated LPS from these cells may be responsible for the enhanced neutrophil superoxide response in the 24-hour postoperative period. The enhanced release of superoxide anion was observed only with f-Met-Leu-Phe and not with PMA. Different mechanisms are involved in activation of oxidative burst response in neutrophils by f-Met-Leu-Phe and PMA. The chemotactic peptide f-Met-Leu-Phe binds to its receptor on the cell surface followed by activation of the NADPH* oxidase, which results in consumption of oxygen and production of moderate amounts of superoxide anion." By contrast, PMA stimulates protein kinase C to activate a pathway leading to generation of superoxide anion by activation of membrane NADPH oxidase." The PMA response is maximal, and this might be the reason why in the present study the cells were primed for f-Met-Leu-Phe response only. The ability of oxygen radicals to damage endothelial cells, erythrocytes, and even tumor cells is widely known.":" In more complex systems such as in vivo conditions, the role of oxygen radicals in tissue damage is documented." Therefore the activation of neutrophils and enhanced release of oxygen radicals in patients undergoing coronary artery bypass may result in tissue damage and in some of the pathologic phenomena observed in these patients after the operations. Expert technical assistance of Hanne Tamstorf and Aasc Stricker-Nielsen is greatly appreciated. REFERENCES

I. Johnson KJ. Fantone JC, Kaplan J, Ward PA. In vivo damage of rat lungs by oxygen metabolites. J Clin Invest 1981 ;67:983-93. 2. Fox R, Hoidal J. Brown D. Repine J. Pulmonary inflammation due to oxygen toxicity: involvement of

"Nicotinarnidc-adcninc dinucleotide phosphate. reduced form.

Volume 98 Number 3 September 1989

chemotactic factors and polymorphonuclear leukocytes. Am Rev Respir Dis 1981;123:521-3. 3. Parks DA, Bulkley GB, Granger ON. Role of oxygenderived free radicals in digestive tract diseases. Surgery

1983;94:415-n 4. Bulkley GB. The role of free radicals in human disease processes. Surgery 1983;94:407-11.

5. Gardner T J, Stewart JR, Casale AS, Downey JM, Chambers DE. Reduction of myocardial ischemic injury with oxygen-derived free radical scavengers. Surgery 1983;94:423-7. 6. Burton KP, McCord JM, Ghai G. Myocardial alterations due to free-radical generation. Am J Physiol 1984; 246:M776-83. 7. Del Maestro RF. An approach to free radicals in medicine and biology. Acta Physiol Scand (suppl) 1980; 492: 153-68. 8. Klebanoff SJ, Clark RA. The neutrophil, functional and clinical disorders. Amsterdam: North Holland, 1978. 9. Johnston RB. Enhancement of phagocytosis-associated oxidative metabolism as a manifestation of macrophage activation. Lymphokines 1981 ;3:33-56. 10. Guthrie LA, McPhail LC Henson PM, Johnston RB. The priming of neutrophils for enhanced release of oxygen metabolites by bacterial lipopolysaccharide: evidence for increased activity of the superoxide-producing enzyme. J Exp Med 1984;16:1656-71. II. Pabst MJ, Johnston RB Jr. Increased production of superoxide anion by macrophages exposed in vivo to muramyl dipeptide or lipopolysaccharide. J Exp Med 1980; 151:I0 1-14. 12. Kharazmi A. Rcchtnitzer C Shi¢tz PO, Jensen T, Back L, Heiby N. Priming of neutrophils for enhanced oxidative burst by sputum from cystic fibrosis patients with Pseudomonas aeruginosa infection. Eur J Clin Invest 1987;17:256-61. 13. Rinaldo JE, Rogers RM. Adult respiratory distress syndrome: changing concepts of lung injury and repair. N Engl J Med 1982;306:900-9. 14. Brigham KL, Bowers R, Haynes J. Increased sheep lung vascular permeability caused by Escherichia coli endotoxin. Circ Res 1979;45:292-7. 15. Heflin AC Jr. Brigham KL. Prevention by granulocyte depletion of increased vascular permeability of sheep lung following endotoxemia. J Clin Invest 1981 ;68: 1253-60. 16. Morrison DC Ulevitch RJ. The effects of bacterial endotoxins on host mediation systems. Am J Pathol 1978;93:527-617. 17. Mikolic J, Hiertz H, Hudec M, Laczhovics DE. Fever, leucocytosis and infection after open heart surgery: a log linear regression analysis of I 15 cases. Thorac Cardiovasc Surg 198432:45-8. 18. Goodmann JS, Schatfner W, Collins HA, Battersby EJ, Koenig MG. Infection after cardio-vascular surgery: clinical study including examination of antimicrobial prophylaxis. N Engl J Moo 1968;278: 117-23.

Endotoxemia

385

19. Bowers TK, O'Flaherty J, Simmons RL. Post-surgical granulocyte dysfunction: studies on healthy kidney donors. J Lab Clin Med 1977;90:720-7. 20. Andersen L W, Baek L, Degn H, Lehd J, Krasnik M, Rasmussen JP. Presence of circulating endotoxins during cardiac operations. J THoRAc CARDIOVASC SURG 1987;93:115-9. 21. Baek L. New sensitive rocket immunoelectrophoretic assay for measurement of the reaction between endotoxin and Limulis amoebocyte lysate. J Clin Microbiol 1983; 17:1013-20. 22. Kharazmi A, Doring G, Heiby N, Valerius NH. Interaction of Pseudomonas aeruginosa alkaline protease and elastase with human polymorphonuclear leukocytes in vitro. Infect Immunol 1984;43:161-5. 23. Massey V. Microestimation of succinate and the extinction coefficient of cytochrome c. Biochem Biophys Acta 195934:255-56. 24. Baek L, H¢iby N, Hertz JB, Espersen, F. Interaction between Limulus amoebocyte lysate and soluble antigens from Pseudomonas aeruginosa and Staphylococcus aureus studied by quantitative immunoelectrophoresis. J Clin Microbiol 1985;22: 1790-7. 25. Pearson FC, Bohon J, Lee W, et aI. Comparison of chemical analysis of hollow-fiber dialyzer extracts. Artif Organs 1984;8:291-8. 26. Haslett C Guthrie LA, Kopaniak MM, Johnson RB, Henson PM. Modulation of multiple neutrophil functions by preoperative methods or trace concentrations of bacterial lipopolysaccharide. Am J Pathol 1985;119:101-10. 27. Silva J Jr, Hoeksema H, Feketry FR Jr. Transient defects in phagocytic functions during cardiopulmonary bypass. J THORAC CARDIOVASC SURG 1974;67:175-83. 28. Duncan RL Jr, Morrison DC. The fate of E. coli by murine macrophages in vitro. J Immunol 1984; 132: 141624. 29. McPhail LC. Snyderman R. Activation of the respiratory burst enzyme in human polymorphonuclear leukocytes by chemoattractants and other soluble stimuli: evidence that the same oxidase is activated by different transductional mechanisms. J Clin Invest 1983;72:192-200. 30. Becton DL, Adams DO, Hamilton TA. Characterization of protein kinase C activity in interferon gamma treated murine peritoneal macrophages. J Cell Physiol 1985; 125:485-91. 31. Sachs T, Moldow CF, Craddock PR, Bowers JK, Jacob HS. Oxygen radical mediated endothelial cell damage by complement-stimulated granulocytes: an in vitro model of immune vascular damage. J Clin Invest 1978;61:1161-7. 32. Lynch ER, Fridovich I. Effects of superoxide on the erythrocyte membrane. J BioI Chem 1978;253:1828-45. 33. Hafeman DO, Lucas ZJ. Polymorphonuclear leukocyte mediated antibody dependent cellular cytotoxicity against tumor cells: dependence on oxygen and respiratory burst. J Immunol 1979; 123:55-62.