- Email: [email protected]
Impaired neutrophil microbicidal activity in rat cholestasis
GASTROENTEROLOGY 1997;112:1692–1698
Impaired Neutrophil Microbicidal Activity in Rat Cholestasis KARTIKA TJANDRA,* RICHARD C. WOODMAN,‡ and MARK G. SWAIN* *Gastrointestinal Research Group and ‡Immunology Research Group, Department of Medicine, University of Calgary, Calgary, Alberta, Canada
Background: Previously, we have documented impaired neutrophil recruitment to inflammatory sites in cholestatic rats. However, there may be additional neutrophil defects that could account for the increased incidence of septic complications in cholestatic patients. The aim of this study was to investigate neutrophil functional defects in cholestasis. Methods: Sprague–Dawley rats were either bile duct resected (BDR) or sham resected (sham). Five days after surgery, peripheral blood neutrophils were assayed for bacterial killing, phagocytic activity, superoxide anion (O20) production, and degranulation. Results: BDR neutrophils showed several functional defects. An in vitro killing of Staphylococcus aureus (5 1 106 CFU/mL) showed that BDR neutrophils were less efficient at killing bacteria than sham neutrophils. Furthermore, bacterial killing by sham and BDR neutrophils was significantly attenuated in the presence of BDR sera. Phagocytosis and neutrophil degranulation did not seem to contribute to impaired killing in BDR neutrophils. However, a rightward shift was observed in the dose-response curve of N-formyl-methionyl-leucyl-phenylalanine–stimulated BDR neutrophil O20 production. Conclusions: O20 generation and bacterial killing are depressed in BDR neutrophils, and BDR sera appear to accentuate the defect in BDR neutrophil bacterial killing. These defects may contribute to lowered resistance to microbial invasion in cholestasis.
P
atients with obstructive jaundice are predisposed to postsurgical septic complications. The rate of postoperative septic complications is higher in jaundiced patients than in the general surgical population.1 – 3 Despite antibiotic pretreatment and improvements in surgical procedures and prophylactic measures such as oral bile salts and preoperative biliary drainage, infective complications remain a significant problem contributing to increased mortality in jaundiced patients after surgery.1,4 The mechanisms underlying the high incidence of postoperative infective complications in cholestatic patients are not well understood. Polymorphonuclear leukocytes (neutrophils) play a major role in the first line of host defense. The ability of neutrophils to migrate from the vasculature to affected tissues is essential in targeting and eliminating pathogens. Furthermore, neu/ 5e1c$$0008
04-11-97 16:55:04
gasa
trophil bactericidal activity is mediated by their ability to generate and release oxidants or proteases. The high incidence of postoperative septic complications in patients with obstructive jaundice may be related to neutrophil dysfunction at the level of neutrophil recruitment or microbicidal activity. The microbicidal function of neutrophils is mediated by two mechanisms. The first involves an oxygen-dependent pathway with the generation of reactive oxygen intermediates (ROIs) and the second is an oxygen-independent pathway involving the release of granular proteins with microbicidal activity (reviewed by Babior,5 Klebanoff,6 and Elsbach and Weiss7). After stimulation by an agonist, neutrophils show a sharp increase in oxygen uptake and within seconds after stimulation begin to release large quantities of superoxide anion (O20) and hydrogen peroxide (H2O2 ) into their surroundings, a process termed the respiratory burst. Further reactions involving O20 and H2O2 result in the formation of more potent microbial oxidants, such as HOCl and •OH.6,8 Although the microbicidal function of neutrophils largely depends on the respiratory burst, several lines of evidence suggest that neutrophils are capable of killing bacteria independent of oxygen radical formation (reviewed by Selsted9). Oxygen-independent neutrophil microbial killing agents (microbicides) have been found in the primary (azurophil) or secondary (specific) granules of neutrophils.10 Previously, we have documented impaired neutrophil recruitment to inflammatory sites in cholestatic rats.11 However, there may be additional neutrophil defects that account for the increased incidence of septic complications in cholestatic patients. We therefore hypothesized that cholestasis may impair neutrophil function and consequently reduce neutrophil bacterial killing. The purpose of this study was to investigate whether cholestasis is Abbreviations used in this paper: BDR, bile duct resected; DHCB, dihydrocytochalasin B; FMLP, N-formyl-methionyl-leucyl-phenylalanine; NBT, nitroblue tetrazolium; O20, superoxide anion; OZ, opsonized zymosan; PMA, phorbol myristate acetate; ROI, reactive oxygen intermediate. q 1997 by the American Gastroenterological Association 0016-5085/97/$3.00
WBS-Gastro
May 1997
IMPAIRED MICROBICIDAL ACTIVITY IN CHOLESTASIS 1693
associated with an impairment of neutrophil microbicidal activity and to determine which neutrophil functions (phagocytosis, O20 production, and degranulation) are impaired in cholestasis.
Materials and Methods
Phagocytosis
Animal Model of Cholestasis Male Sprague–Dawley rats (200–225 g) were either bile duct resected (BDR) or sham resected (sham) as described previously.12 The animals were kept in a light-controlled room with a 12-hour day-night cycle and given free access to rat chow and water. Animals were handled humanely under the University of Calgary Animal Care Committee Guidelines. The animals were studied on the 5th postoperative day, when BDR animals were overtly cholestatic.13
Isolation of Neutrophils Rat neutrophils were isolated 5 days after surgery by dextran sedimentation of anticoagulated peripheral blood (anticoagulant, consisting of 0.14 mol/L citric acid, 0.2 mol/L sodium citrate, and 0.22 mol/L dextrose) followed by hypotonic lysis and centrifugation over histopaque (Sigma Chemical Co., St. Louis, MO) at 47C.14 This isolation procedure yielded ú95% neutrophils with ú90% viability as measured by trypan blue dye exclusion.14 Neutrophils were suspended in cold phosphate-buffered saline (PBS) containing 0.5 mmol/L MgCl2 , 0.9 mmol/L CaCl2 , and 7.5 mmol/L glucose at a final suspension of 2 1 107 cells/mL before being used for functional assays.
Bacterial Killing Assay Staphylococcus aureus (strain 502A) was prepared in Mueller–Hinton broth (Sigma) and grown overnight at 377C.15 The following day, the bacteria were centrifuged (2000g for 15 minutes), washed, and resuspended in PBS with 1% gelatin (pH 7.4) to a final volume of 1 mL. To determine bacterial density, the optical density of the bacterial suspension was measured at 620 nm (U-2000 double beam spectrophotometer; Hitachi Ltd., Tokyo, Japan). The bacterial suspension was adjusted to a density of approximately 5 1 106 CFU/mL. Freshly isolated neutrophils were diluted with PBS (containing 0.5 mmol/L MgCl2 , 0.9 mmol/L CaCl2 , and 7.5 mmol/ L glucose) to 5 1 106 cells/mL and coincubated with bacteria in 1 mL of PBS containing 1% gelatin and 10% autologous serum and shaken end-over-end at 377C for 1 hour (model 1105 Adams Nutator; Clay Adams, Parsippany, NJ). Aliquots of 50 mL were withdrawn from each tube at 0 and 1 hour of incubation and immediately added to tubes containing 5 mL of distilled water with 1% gelatin (pH 7.4). The tubes were held for 10 minutes to ensure neutrophil lysis and vortexed, and then a further 10-fold dilution was made in 10 mmol/L Tris buffer with 1% gelatin (pH 7.4). Each condition was run in duplicate. Viable bacteria were determined by spreading 50 mL from each dilution tube onto tryptic soy agar plates. After an overnight incubation at 377C, viable bacteria were quanti-
/ 5e1c$$0008
04-11-97 16:55:04
tated as individual colonies. To determine the percentage of in vitro bacterial killing by neutrophils, total counts of bacteria on control plates (assuming no killing at time 0) were compared with the number of bacterial colonies remaining on tryptic soy agar plates after 1 hour incubation.
gasa
Neutrophil phagocytosis was determined using a modified flow cytometric assay.16 S. aureus (Woodstrain without protein A) bioparticles conjugated with fluorescein isothiocyanate (BOPIDY FL from Molecular Probes Inc., Eugene, OR) were diluted with sonication in 1 mL of assay buffer (PBS containing 0.1% gelatin, 5 mmol/L glucose, 0.5 mmol/L MgCl2 ); 500 mL was then added into polypropylene tubes containing 1 mL of untreated and undiluted rat sera (sham or BDR). The samples were centrifuged (20 minutes at 1000g at 47C) after an incubation of 1 hour at 377C. The S. aureus bioparticles were washed twice with ice-cold PBS buffer and then aliquoted in assay buffer at a final concentration of 4 1 108 bioparticles/mL for storage at 0607C. S. aureus bioparticles were thawed once on the day of use. The bioparticles (1 1 107 bioparticles) opsonized with either sham or BDR sera were then added to 100 mL of rat neutrophils (1 1 106) and incubated to permit phagocytosis to occur. After incubations of 0 and 60 minutes, the cells were mixed briefly and 50-mL aliquots of cells were removed for analysis by FACScan (Becton Dickinson, San Jose, CA). The channel number (log scale) representing the mean fluorescence intensity of 10,000 cells was used. For each aliquot of cells, the percentage of cells expressing an increased mean fluorescence intensity, not seen at 0 minutes, was determined to be those cells phagocytosing S. aureus bioparticles. Cells alone were used as a negative control in each experiment.
Superoxide Anion Assay Superoxide anion production was quantitated using a continuous kinetic assay of superoxide dismutase inhibition of ferricytochrome c reduction.17 Superoxide-dependent cytochrome c reduction was determined spectrophotometrically at 550 nm (U-2000 double-beam spectrophotometer; Hitachi, Ltd.). Briefly, 100 mL of cell suspension (1 1 106 cells) was added to sample and reference cuvettes containing 890 mL of PBS and 0.08 mmol/L cytochrome c/ddH2O. Cells were activated by either phorbol myristate acetate (PMA; 200 ng/mL; Sigma), opsonized zymosan (OZ; 0.1 mg/mL), or N-formylmethionyl-leucyl-phenylalanine (FMLP; 10010 to 1005 mol/L; Sigma). In some assays, neutrophils were primed with dihydrocytochalasin B (DHCB; 2.5 mg/mL; Sigma) for 2 minutes at 377C before FMLP addition. OZ was prepared by incubating 10 mg zymosan A (Sigma) per milliliter of fresh plasma for 30 minutes at 377C. The mixture was diluted with Hank’s balanced salt solution (Sigma), centrifuged at 450g for 6 minutes, washed twice with Hank’s balanced salt solution, and suspended at a final concentration of 1 mg/mL. To examine O20 release by neutrophils in a more physiological milieu, 10% of either sham or BDR sera was added to each cuvette
WBS-Gastro
1694 TJANDRA ET AL.
GASTROENTEROLOGY Vol. 112, No. 5
for some assays. The data from the latter experiments were used to determine whether neutrophil dysfunction in generating ROIs results from a defect in the respiratory burst or possibly a suppressive effect of sera. To examine differential release of O20 by sham and BDR neutrophils, a nitroblue tetrazolium (NBT) slide test was performed.15 Freshly isolated neutrophils were incubated at 377C on glass slides with NBT dye (Sigma) and PMA (200 ng/mL) or buffer alone (resting). Neutrophils were fixed with absolute methanol before they were stained with safranin-o (BDH Inc., Toronto, Ontario, Canada). In the presence of O20, the NBT dye is reduced chemically to yield a dark purple insoluble compound (formazan), which can be clearly discerned microscopically and the percentage of cells containing the reduced NBT dye (formazan) can be determined easily. A comparison of staining intensity also can be made to control cells.
Degranulation Assay Microbicidal proteins and proteases released from the granules of neutrophils may also mediate neutrophil microbicidal activity. We examined exocytic neutrophil degranulation by measuring b-glucuronidase activity in the supernatant of stimulated and unstimulated neutrophils.14 Exocytic degranulation was determined by subtracting the amount of enzyme released in the presence of stimuli from that released in their absence. Briefly, neutrophils (1 1 107 cells) were incubated at 377C with DHCB (2.5 mg/mL) and stimulated with either FMLP (1007 mol/L) or PMA (200 ng/mL) for 30 minutes in a shaker bath (model 2564; Forma Scientific, Marrietta, OH). DHCB was added to promote exocytic degranulation. The supernatant from each tube was collected after centrifugation at 420g for 10 minutes and placed on ice. The b-glucuronidase levels were quantitated spectrofluorometrically (SPF-500 Spectrofluorometer; SLM Instruments Inc., Urbana, IL) against a known b-glucuronidase standard using excitation and emission wavelengths of 365 and 460 nm, respectively. Cell lysis was analyzed by measuring glucose-6-P-dehydrogenase activity spectrophotometrically at 340 nm (U-2000 double-beam spectrophotometer; Hitachi Ltd.).
Statistical Analysis
was further reduced in the presence of BDR (cholestatic) sera with an additional 50% reduction relative to the percentage of bacterial killing by BDR neutrophils in the presence of sham sera. In addition, the antibactericidal effect of BDR sera also was observed in bacterial killing by sham neutrophils (Figure 1). These data suggest that bacterial killing is impaired in cholestatic rats and that there is an inhibitory effect of BDR sera on neutrophil microbicidal activity. To elucidate the mechanisms of impaired bactericidal activity in BDR neutrophils, neutrophil phagocytosis, O20 production, and degranulation were investigated. Neutrophil Phagocytic Activity
Data are expressed as mean { SEM. An unpaired student’s t test was used for comparisons between two means, and analysis of variance followed by Student–Newman–Keuls test for comparisons between more than two means. A P value of °0.05 was considered significant.
Results Neutrophil Bactericidal Activity In the presence of autologous sera, sham neutrophils were more efficient at bacterial killing in vitro than BDR neutrophils (Figure 1). The ability of BDR neutrophils to kill bacteria was decreased by approximately 20% relative to sham neutrophils (Figure 1). Furthermore, impaired bacterial killing by BDR neutrophils / 5e1c$$0008
Figure 1. Percentage of S. aureus killed by sham (j) and BDR ( ) neutrophils after a 1-hour incubation. Neutrophils, sera, and bacteria were incubated in PBS with 1% gelatin. Aliquots were placed in water containing 1% gelatin for at least 10 minutes to ensure neutrophil lysis. Further dilutions were made into Tris buffer with 1% gelatin. Bacteria remaining in suspension were grown on tryptic soy agar plates overnight and individual colonies counted the next day. Data are mean { SEM of five experiments. *P õ 0.05 vs. sham neutrophils / BDR sera; /P õ 0.05 and //P õ 0.001 vs. sham neutrophils / sham sera; **P õ 0.001 vs. BDR neutrophils / sham sera.
04-11-97 16:55:04
gasa
Because phagocytosis is essential in neutrophil microbicidal function, a kinetic phagocytic assay of sham and BDR neutrophils was performed. Phagocytic activity was similar in sham and BDR neutrophils after 1-hour incubation with either S. aureus bioparticles opsonized with sham sera (sham: 8.72% { 3.3% vs. BDR: 5.51% { 1.14%; n Å 4) or S. aureus bioparticles opsonized with BDR sera (sham: 8.82% { 1.2% vs. BDR: 6.00% { 1.42%; n Å 4). O20 Release Neutrophils from BDR rats showed an approximate 30%–50% decrease in the rate of O20 release when stimulated with either 1007 or 1008 mol/L FMLP comWBS-Gastro
May 1997
IMPAIRED MICROBICIDAL ACTIVITY IN CHOLESTASIS 1695
Figure 2. O20 production by sham ( ) and BDR (j) rat neutrophils at increasing FMLP concentrations (10010 to 1005 mol/L). Neutrophils were primed with DHCB (2.5 mg/mL) for approximately 2 minutes at 377C before FMLP addition. Data are mean { SEM of three experiments (each with duplicate runs). **P õ 0.005 vs. sham value, *P õ 0.05 vs. sham value.
pared with sham neutrophils (Figure 2). However, at higher concentrations of FMLP (i.e., 1006 and 1005 mol/ L), there was no significant difference in the ability of neutrophils to generate O20. There was no appreciable release of O20 by either sham or BDR neutrophils when stimulated with lower doses of FMLP (1009 and 10010 mol/L). When stimulated with either PMA or OZ, sham and BDR neutrophils showed similar rates of O20 release (data not shown). The addition of either 10% sham or BDR sera to FMLP-stimulated sham neutrophils decreased the rate of O20 release by approximately 35% compared with the rate when neither sera were present, but without a further reduction in the rate of O20 release by BDR neutrophils (Figure 3). In addition, the rates of O20 release were not significantly different between sham and BDR neutrophils in the presence of either sera (Figure 3). The NBT staining indicated that the decreased rate of O20 release by BDR neutrophils was caused by a homogenous reduction of O20 release by BDR neutrophils. The results of NBT tests were comparable between sham and BDR neutrophils (93.8% { 0.2% and 92.1% { 1.1%, respectively), suggesting that the decrease in O20 release observed in BDR neutrophils is not caused by an inability to produce O20.
Figure 3. The rates of superoxide anion generated by sham (j) and BDR ( ) neutrophils in the presence of either 10% sham or BDR sera. Neutrophils were primed with DHCB (2.5 mg/mL) for 2 minutes at 377C before stimulation with 1007 mol/L FMLP. Data are mean { SEM of three experiments (each with duplicate runs). *P õ 0.05 vs. sham value.
FMLP (23.6% { 2.1% vs. 27.4% { 1.0%, respectively; NS) or 200 ng/mL PMA (21.2% { 3.7% vs. 25.5% { 3.1%, respectively; NS). The contribution of lysis to neutrophil degranulation, as determined by glucose-6-Pdehydrogenase activity in the supernatant, was negligible (data not shown).
Discussion To examine the ability of sham and BDR neutrophils to eliminate microorganisms, in vitro killing of S. aureus was analyzed. We observed a striking difference
Neutrophil Degranulation Neutrophil degranulation, quantitated as the percentage of b-glucuronidase released from primary granules, is shown in Figure 4. The percentage of b-glucuronidase released from sham and BDR neutrophils was similar whether they were stimulated with 1007 mol/L / 5e1c$$0008
04-11-97 16:55:04
gasa
Figure 4. b-Glucuronidase release from sham (j) and BDR ( ) neutrophils after incubation with DHCB (2.5 mg/mL) and stimulation with either 1007 mol/L FMLP or 200 ng/mL PMA for 30 minutes at 377C. After centrifugation at 420g for 10 minutes, the supernatant of each sample was collected and assayed for b-glucuronidase content using a spectrofluorometer (excitation l Å 365 nm; emission l Å 460 nm). Data are mean { SEM of five experiments.
WBS-Gastro
1696 TJANDRA ET AL.
GASTROENTEROLOGY Vol. 112, No. 5
in bacterial killing between sham and BDR neutrophils in that BDR neutrophils were less able to kill bacteria than sham neutrophils (P õ 0.05). Impaired ability to kill bacteria by BDR neutrophils may be caused by defective neutrophil function or the effect of cholestatic sera or both. Because BDR sera also exhibited an antimicrobicidal effect as shown by a significant reduction in bacterial killing by both sham and BDR neutrophils, we showed that impaired killing in BDR neutrophils arises from a combination of neutrophil functional defects and the antimicrobicidal effects of BDR sera. Phagocytosis is essential for bactericidal functions of neutrophils. A kinetic phagocytic assay using opsonized S. aureus, the same type of bacteria used in our killing studies, showed comparable phagocytic activity in sham and BDR neutrophils. This suggests that BDR neutrophils have normal phagocytic activity and are not responsible for the impaired neutrophil bactericidal activity we observed in BDR rats. To eliminate infectious agents, neutrophils generate ROIs and release their granule constituents. Therefore, impaired in vitro bacterial killing by BDR neutrophils may be caused by neutrophil defects in generating ROIs or releasing microbicidal proteins or a combination of both pathways. BDR neutrophils generated only half of the O20 released by sham neutrophils when stimulated with 1008 or 1007 mol/L FMLP (Figure 2). The NBT slide test results suggest that the impairment in O20 release by BDR neutrophils probably was caused by an effect on the entire population of BDR neutrophils, not caused by a specific subset of cells. In contrast, supraphysiological concentrations of FMLP (i.e., 1006 and 1005 mol/L), PMA, or OZ did not result in any significant difference in the rate of O20 release between sham and BDR neutrophils, suggesting that receptor and soluble agonists can equally activate sham and BDR neutrophils to produce O20. Higher levels of saturated fatty acids and cholesterol in plasma membranes of cholestatic patients and BDR rats18,19 may result in more rigid neutrophil membranes. This may influence FMLP receptor mobilization to the cell surface. In human neutrophils, low- and high-affinity binding sites for FMLP have been determined, whereas in rats, FMLP receptors appear to be of the high-affinity type, with a dissociation constant ranging from 1 to 30 nmol/L.20 Therefore, lower concentrations of FMLP would provide more sensitivity with respect to O20 production than higher concentrations. Similar O20 production in sham and BDR neutrophils at higher concentrations of FMLP may suggest a similar maximal capacity to produce O20. Because PMA bypasses membrane-associated neutrophil activation, similar O20 production in / 5e1c$$0008
04-11-97 16:55:04
gasa
BDR and sham neutrophils would be expected if there was a membrane- or receptor-determined defect. With respect to OZ, interpretation of our results is somewhat difficult. Although there is evidence to suggest that O20 production by human neutrophils is OZ-concentration dependent,21 detection of a concentration-dependent increase in O20 production by sham and BDR neutrophils stimulated with increasing OZ concentrations was difficult to assess because of increasing solution turbidity with higher OZ concentration. Our results contrast with those of Levy et al.22 who showed that neutrophils isolated from common bile duct–ligated rats generated more O20 than sham animals.22 Despite these differences, it is important to note that the levels of O20 produced in their study were much lower than what we observed. For example, using the same stimulus (1007 mol/L FMLP), the amount of O20 generated by their BDR neutrophils (which was significantly higher than sham neutrophils) was approximately 85% lower than what we observed with our BDR neutrophils. Furthermore, the range of O20 release in our study falls within a similar range observed with human neutrophils.15 Therefore, we are not certain about the physiological significance of such low levels of O20 production as reported by Levy et al.22 on neutrophil microbicidal function. The discrepancies between results may be explained by the use of different methods of blood collection and neutrophil isolation as well as assays to measure O20 generation. We chose the continuous assay because it is more sensitive in quantitating O20 release than the end point assay used by Levy et al.22 and permits one to observe directly the kinetics of neutrophil O20 production. To examine the effects of BDR sera on O20 release, 10% of either sham or BDR sera was included in the assay system. In contrast to the results of Ohshio et al.,23 but in agreement with Levy et al.,22 we were unable to see any increase in O20 release by neutrophils when incubated with either sham or BDR sera. In fact, we observed a 35% reduction in O20 release by sham neutrophils when exposed to either sera compared with its absence; there was comparable O20 production by sham and BDR neutrophils in the presence of either sera (Figure 3). A significant reduction in O20 release by sham neutrophils in the presence of either sera may be caused by antioxidant activity of the sera as physiological antioxidants (e.g., glutathione) and bystander antioxidants (e.g., plasma proteins) scavenge ROIs.24 – 26 Therefore, BDR sera did not appear to affect O20 release by BDR neutrophils. Some cytokines have been shown to be capable of priming neutrophils to enhance O20 release.17,27 – 29 WBS-Gastro
May 1997
IMPAIRED MICROBICIDAL ACTIVITY IN CHOLESTASIS 1697
Woodman et al.17 documented that preincubation of human neutrophils with granulocyte-macrophage colony– stimulating factor caused an augmentation in the rate of O20 production but without a significant increase in the maximal amount generated. Because glucocorticoids can inhibit cytokine release,30 the elevated levels of endogenous glucocorticoids previously documented in BDR animals11 may depress cytokine production. This in turn may affect neutrophil priming and activation. The role of cytokines in neutrophil O20 release in BDR rats warrants further investigation. Strong evidence for the contribution of granule constituents in neutrophil microbicidal activity has been shown in studies using neutrophils obtained from patients with chronic granulomatous disease who are deficient in functional reduced nicotinamide adenine dinucleotide phosphate oxidase.21,31 Therefore, we examined the degranulation process in sham and BDR neutrophils. The release of granule contents by stimulated sham and BDR neutrophils, measured as b-glucuronidase release, was similar between the two types of neutrophils (Figure 4). These data suggest that lowered microbial resistance in cholestasis is not caused by an impaired neutrophil degranulation. In addition, impaired O20 production in BDR neutrophils did not seem to be caused by a defect in neutrophil receptor function. In conclusion, we show that in vitro bactericidal function of neutrophils isolated from cholestatic rats is impaired. Specifically, we identified the following two functional defects that may contribute to increased infections in cholestasis: (1) decreased O20 generation and bacterial killing and (2) an unidentified antimicrobicidal effect of BDR sera that further impairs the bactericidal ability of neutrophils. The combined effects of neutrophil dysfunction and impaired recruitment, which we have identified previously in cholestatic rats,11 may explain the increased incidence of infective complications in cholestatic patients after surgery. These observations need to be confirmed in humans, but merit further investigation.
and clinical correlates. 2nd ed. New York: Raven, 1992:541– 588. Elsbach P, Weiss J. Oxygen-independent antimicrobial systems of phagocytes. In: Gallin JI, Goldstein M, Snyderman R, eds. Inflammation: basic principles and clinical correlates. 2nd ed. New York: Raven Press, 1992:603–636. Babior BM. The respiratory burst (pp. 132–134, Lehrer RI, moderator). Neutrophils and host defense. Ann Intern Med 1988; 109:127–142. Selsted ME. Nonoxidative killing by neutrophils (pp. 129–132, Lehrer RI, moderator). Neutrophils and host defense. Ann Intern Med 1988;109:127–142. Babior BM. Neutrophil function as related to neutrophil-endothelial cell interactions. Nouv Rev Fr Hematol 1992;34(Suppl):S29– S35. Swain MG, Tjandra K, Kanwar S, Kubes P. Neutrophil adhesion is impaired in a rat model of cholestasis. Gastroenterology 1995; 109:923–932. Cameron GR, Oakley CL. Ligation of the common bile duct. J Pathol Bacteriol 1932;35:769–798. Tjandra K, Kubes P, Rioux K, Swain MG. Endogenous glucocorticoids inhibit neutrophil recruitment to inflammatory sites in cholestatic rats. Am J Physiol 1996;270:G821–G825. Woodman RC, Reinhardt PH, Kanwar S, Johnston FL, and Kubes P. The effects of human neutrophil elastase (HNE) on neutrophil function in vitro and in inflamed microvessels. Blood 1993;82: 2188–2195. Woodman RC, Newburger PE, Anklesaria P, Erickson RW, Rae J, Cohen MS, Curnutte JT. A new X-linked variant of chronic granulomatous disease characterized by the existence of a normal clone of respiratory burst-component phagocytic cells. Blood 1995;85: 231–241. Schiff DE, Rae J, Davis BH, Curnutte JT. In vivo recombinant human interferon-g (rhIFN-g) treatment of normal human subjects is associated with increased FcgRI expression by phagocytes and improved phagocytosis of S. aureus (abstr). Blood 1995;86:28a. Woodman RC, Curnutte JT, Babior BM. Evidence that de novo protein synthesis participates in a time-dependent augmentation of the chemotactic peptide induced respiratory burst in neutrophils—effects of recombinant human colony stimulating factors and dihydrocytochalasin B. Free Radical Biol & Med 1988;5: 355–361. Scriven MW, Horrobin DF, Puntis MCA. Study of plasma and red cell phospholipid fatty acids in extrahepatic cholestatic jaundice. Gut 1994;35:987–990. Lang ME, Jourd’heuil D, Meddings JB, Swain MG. Increased opioid binding to peripheral white blood cells in a rat model of acute cholestasis. Gastroenterology 1995;108:1479–1486. Walker BAM, Seiler AJ, Owens CA, Hagenlocker BE, Ward PA. Absence of FMLP receptors on rat macrophages. J Leukocyte Biol 1991;50:600–606. Root RK, Cohen MS. The microbicidal mechanisms of human neutrophils and eosinophils. Rev Infect Dis 1981;3:565–598. Levy R, Schlaeffer F, Keynan A, Nagauker O, Yaari A, Sikuler E. Increased neutrophil function induced by bile duct ligation in a rat model. Hepatology 1993;17:908–914. Ohshio G, Miyachi Y, Kudo H, Niwa Y, Manabe T, Tobe T. Effects of sera from patients with obstructive jaundice on the generation of oxygen intermediates by normal polymorphonuclear leukocytes. Liver 1988;8:366–371. Dobrinich R, Spagnuolo PJ. Binding of C-reactive protein to human neutrophils. Arthritis Rheum 1991;34:1031–1038. Bucurenci N, Blake DR, Chidwick K, Winyard PG. Inhibition of neutrophil superoxide production by human plasma a1-antitrypsin. FEBS Lett 1992;300:21–24. Kim SJ, Han D, Moon KD, Rhee JS. Measurement of superoxide dismutase-like activity of natural antioxidants. Biosci Biotechnol Biochem 1995;59:822–826.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16.
17.
18.
19.
20.
21.
References 22. 1. Amstrong CP, Dixon JM, Taylor TV, Davies GC. Surgical experience of deeply jaundiced patients with bile duct obstruction. Br J Surg 1984;71:234–238. 2. Pain JA, Cahill CJ, Bailey ME. Perioperative complications in obstructive jaundice: therapeutic considerations. Br J Surg 1985; 72:942–945. 3. Cainzos M, Potel J, Puente J. Anergy in jaundiced patients. Br J Surg 1988;75:147–149. 4. Pitt HA, Cameron JL, Postier RG, Gadacz TR. Factors affecting mortality in biliary tract surgery. Am J Surg 1981;141:66–72. 5. Babior BM. Oxidants from phagocytes: agents of defense and destruction. Blood 1984;64:959–966. 6. Klebanoff SJ. Oxygen metabolites from phagocytes. In: Gallin JI, Goldstein M, Snyderman R, eds. Inflammation: basic principles
/ 5e1c$$0008
04-11-97 16:55:04
gasa
23.
24. 25.
26.
WBS-Gastro
1698 TJANDRA ET AL.
GASTROENTEROLOGY Vol. 112, No. 5
27. Weisbart RH, Kwan L, Golde DW, Gasson JC. Human GM-CSF primes neutrophils for enhanced oxidative metabolism in response to the major physiological chemoattractants. Blood 1987;69:18–21. 28. Bajaj MS, Kew RR, Webster RO, Hyers TM. Priming of human neutrophil functions by tumor necrosis factor: enhancement of superoxide anion generation, degranulation, and chemotaxis to chemoattractants C5a and F-Met-Leu-Phe. Inflammation 1992; 16:241–250. 29. Wozniak A, Betts WH, Murphy GA, Rokicinski M. Interleukin-8 primes human neutrophils for enhanced superoxide anion production. Immunology 1993;79:608–615. 30. Hawes AS, Rock CS, Keogh CV, Lowry SF, Calvano SE. In vivo effects of the antiglucocorticoid RU 486 on glucocorticoid and cytokine responses to Escherichia coli endotoxin. Infect Immunol 1992;60:2641–2647.
/ 5e1c$$0008
04-11-97 16:55:04
gasa
31. Kaplan EL, Laxdal T, Quie PG. Studies of polymorphonuclear leukocytes from patients with chronic granulomatous disease of childhood: bactericidal capacity for streptococci. Pediatrics 1968;41:591–599.
Received June 21, 1996. Accepted December 26, 1996. Address requests for reprints to: Mark G. Swain, M.D., Gastroenterology Research Group, Health Sciences Center, 3330 Hospital Drive Northwest, Calgary, Alberta, Canada T2N 4N1. Fax: (403) 270-0995. Supported by grants from Alberta Heritage Foundation for Medical Research (AHFMR) and Medical Research Council (MRC) of Canada. R.C.W. is an AHFMR Medical Scholar and M.G.S. is an MRC Scholar and AHFMR Clinical Investigator. The authors thank M. Maric, D. Teoh, and F. Johnston for technical assistance.
WBS-Gastro
Impaired Neutrophil Microbicidal Activity in Rat Cholestasis KARTIKA TJANDRA,* RICHARD C. WOODMAN,‡ and MARK G. SWAIN* *Gastrointestinal Research Group and ‡Immunology Research Group, Department of Medicine, University of Calgary, Calgary, Alberta, Canada
Background: Previously, we have documented impaired neutrophil recruitment to inflammatory sites in cholestatic rats. However, there may be additional neutrophil defects that could account for the increased incidence of septic complications in cholestatic patients. The aim of this study was to investigate neutrophil functional defects in cholestasis. Methods: Sprague–Dawley rats were either bile duct resected (BDR) or sham resected (sham). Five days after surgery, peripheral blood neutrophils were assayed for bacterial killing, phagocytic activity, superoxide anion (O20) production, and degranulation. Results: BDR neutrophils showed several functional defects. An in vitro killing of Staphylococcus aureus (5 1 106 CFU/mL) showed that BDR neutrophils were less efficient at killing bacteria than sham neutrophils. Furthermore, bacterial killing by sham and BDR neutrophils was significantly attenuated in the presence of BDR sera. Phagocytosis and neutrophil degranulation did not seem to contribute to impaired killing in BDR neutrophils. However, a rightward shift was observed in the dose-response curve of N-formyl-methionyl-leucyl-phenylalanine–stimulated BDR neutrophil O20 production. Conclusions: O20 generation and bacterial killing are depressed in BDR neutrophils, and BDR sera appear to accentuate the defect in BDR neutrophil bacterial killing. These defects may contribute to lowered resistance to microbial invasion in cholestasis.
P
atients with obstructive jaundice are predisposed to postsurgical septic complications. The rate of postoperative septic complications is higher in jaundiced patients than in the general surgical population.1 – 3 Despite antibiotic pretreatment and improvements in surgical procedures and prophylactic measures such as oral bile salts and preoperative biliary drainage, infective complications remain a significant problem contributing to increased mortality in jaundiced patients after surgery.1,4 The mechanisms underlying the high incidence of postoperative infective complications in cholestatic patients are not well understood. Polymorphonuclear leukocytes (neutrophils) play a major role in the first line of host defense. The ability of neutrophils to migrate from the vasculature to affected tissues is essential in targeting and eliminating pathogens. Furthermore, neu/ 5e1c$$0008
04-11-97 16:55:04
gasa
trophil bactericidal activity is mediated by their ability to generate and release oxidants or proteases. The high incidence of postoperative septic complications in patients with obstructive jaundice may be related to neutrophil dysfunction at the level of neutrophil recruitment or microbicidal activity. The microbicidal function of neutrophils is mediated by two mechanisms. The first involves an oxygen-dependent pathway with the generation of reactive oxygen intermediates (ROIs) and the second is an oxygen-independent pathway involving the release of granular proteins with microbicidal activity (reviewed by Babior,5 Klebanoff,6 and Elsbach and Weiss7). After stimulation by an agonist, neutrophils show a sharp increase in oxygen uptake and within seconds after stimulation begin to release large quantities of superoxide anion (O20) and hydrogen peroxide (H2O2 ) into their surroundings, a process termed the respiratory burst. Further reactions involving O20 and H2O2 result in the formation of more potent microbial oxidants, such as HOCl and •OH.6,8 Although the microbicidal function of neutrophils largely depends on the respiratory burst, several lines of evidence suggest that neutrophils are capable of killing bacteria independent of oxygen radical formation (reviewed by Selsted9). Oxygen-independent neutrophil microbial killing agents (microbicides) have been found in the primary (azurophil) or secondary (specific) granules of neutrophils.10 Previously, we have documented impaired neutrophil recruitment to inflammatory sites in cholestatic rats.11 However, there may be additional neutrophil defects that account for the increased incidence of septic complications in cholestatic patients. We therefore hypothesized that cholestasis may impair neutrophil function and consequently reduce neutrophil bacterial killing. The purpose of this study was to investigate whether cholestasis is Abbreviations used in this paper: BDR, bile duct resected; DHCB, dihydrocytochalasin B; FMLP, N-formyl-methionyl-leucyl-phenylalanine; NBT, nitroblue tetrazolium; O20, superoxide anion; OZ, opsonized zymosan; PMA, phorbol myristate acetate; ROI, reactive oxygen intermediate. q 1997 by the American Gastroenterological Association 0016-5085/97/$3.00
WBS-Gastro
May 1997
IMPAIRED MICROBICIDAL ACTIVITY IN CHOLESTASIS 1693
associated with an impairment of neutrophil microbicidal activity and to determine which neutrophil functions (phagocytosis, O20 production, and degranulation) are impaired in cholestasis.
Materials and Methods
Phagocytosis
Animal Model of Cholestasis Male Sprague–Dawley rats (200–225 g) were either bile duct resected (BDR) or sham resected (sham) as described previously.12 The animals were kept in a light-controlled room with a 12-hour day-night cycle and given free access to rat chow and water. Animals were handled humanely under the University of Calgary Animal Care Committee Guidelines. The animals were studied on the 5th postoperative day, when BDR animals were overtly cholestatic.13
Isolation of Neutrophils Rat neutrophils were isolated 5 days after surgery by dextran sedimentation of anticoagulated peripheral blood (anticoagulant, consisting of 0.14 mol/L citric acid, 0.2 mol/L sodium citrate, and 0.22 mol/L dextrose) followed by hypotonic lysis and centrifugation over histopaque (Sigma Chemical Co., St. Louis, MO) at 47C.14 This isolation procedure yielded ú95% neutrophils with ú90% viability as measured by trypan blue dye exclusion.14 Neutrophils were suspended in cold phosphate-buffered saline (PBS) containing 0.5 mmol/L MgCl2 , 0.9 mmol/L CaCl2 , and 7.5 mmol/L glucose at a final suspension of 2 1 107 cells/mL before being used for functional assays.
Bacterial Killing Assay Staphylococcus aureus (strain 502A) was prepared in Mueller–Hinton broth (Sigma) and grown overnight at 377C.15 The following day, the bacteria were centrifuged (2000g for 15 minutes), washed, and resuspended in PBS with 1% gelatin (pH 7.4) to a final volume of 1 mL. To determine bacterial density, the optical density of the bacterial suspension was measured at 620 nm (U-2000 double beam spectrophotometer; Hitachi Ltd., Tokyo, Japan). The bacterial suspension was adjusted to a density of approximately 5 1 106 CFU/mL. Freshly isolated neutrophils were diluted with PBS (containing 0.5 mmol/L MgCl2 , 0.9 mmol/L CaCl2 , and 7.5 mmol/ L glucose) to 5 1 106 cells/mL and coincubated with bacteria in 1 mL of PBS containing 1% gelatin and 10% autologous serum and shaken end-over-end at 377C for 1 hour (model 1105 Adams Nutator; Clay Adams, Parsippany, NJ). Aliquots of 50 mL were withdrawn from each tube at 0 and 1 hour of incubation and immediately added to tubes containing 5 mL of distilled water with 1% gelatin (pH 7.4). The tubes were held for 10 minutes to ensure neutrophil lysis and vortexed, and then a further 10-fold dilution was made in 10 mmol/L Tris buffer with 1% gelatin (pH 7.4). Each condition was run in duplicate. Viable bacteria were determined by spreading 50 mL from each dilution tube onto tryptic soy agar plates. After an overnight incubation at 377C, viable bacteria were quanti-
/ 5e1c$$0008
04-11-97 16:55:04
tated as individual colonies. To determine the percentage of in vitro bacterial killing by neutrophils, total counts of bacteria on control plates (assuming no killing at time 0) were compared with the number of bacterial colonies remaining on tryptic soy agar plates after 1 hour incubation.
gasa
Neutrophil phagocytosis was determined using a modified flow cytometric assay.16 S. aureus (Woodstrain without protein A) bioparticles conjugated with fluorescein isothiocyanate (BOPIDY FL from Molecular Probes Inc., Eugene, OR) were diluted with sonication in 1 mL of assay buffer (PBS containing 0.1% gelatin, 5 mmol/L glucose, 0.5 mmol/L MgCl2 ); 500 mL was then added into polypropylene tubes containing 1 mL of untreated and undiluted rat sera (sham or BDR). The samples were centrifuged (20 minutes at 1000g at 47C) after an incubation of 1 hour at 377C. The S. aureus bioparticles were washed twice with ice-cold PBS buffer and then aliquoted in assay buffer at a final concentration of 4 1 108 bioparticles/mL for storage at 0607C. S. aureus bioparticles were thawed once on the day of use. The bioparticles (1 1 107 bioparticles) opsonized with either sham or BDR sera were then added to 100 mL of rat neutrophils (1 1 106) and incubated to permit phagocytosis to occur. After incubations of 0 and 60 minutes, the cells were mixed briefly and 50-mL aliquots of cells were removed for analysis by FACScan (Becton Dickinson, San Jose, CA). The channel number (log scale) representing the mean fluorescence intensity of 10,000 cells was used. For each aliquot of cells, the percentage of cells expressing an increased mean fluorescence intensity, not seen at 0 minutes, was determined to be those cells phagocytosing S. aureus bioparticles. Cells alone were used as a negative control in each experiment.
Superoxide Anion Assay Superoxide anion production was quantitated using a continuous kinetic assay of superoxide dismutase inhibition of ferricytochrome c reduction.17 Superoxide-dependent cytochrome c reduction was determined spectrophotometrically at 550 nm (U-2000 double-beam spectrophotometer; Hitachi, Ltd.). Briefly, 100 mL of cell suspension (1 1 106 cells) was added to sample and reference cuvettes containing 890 mL of PBS and 0.08 mmol/L cytochrome c/ddH2O. Cells were activated by either phorbol myristate acetate (PMA; 200 ng/mL; Sigma), opsonized zymosan (OZ; 0.1 mg/mL), or N-formylmethionyl-leucyl-phenylalanine (FMLP; 10010 to 1005 mol/L; Sigma). In some assays, neutrophils were primed with dihydrocytochalasin B (DHCB; 2.5 mg/mL; Sigma) for 2 minutes at 377C before FMLP addition. OZ was prepared by incubating 10 mg zymosan A (Sigma) per milliliter of fresh plasma for 30 minutes at 377C. The mixture was diluted with Hank’s balanced salt solution (Sigma), centrifuged at 450g for 6 minutes, washed twice with Hank’s balanced salt solution, and suspended at a final concentration of 1 mg/mL. To examine O20 release by neutrophils in a more physiological milieu, 10% of either sham or BDR sera was added to each cuvette
WBS-Gastro
1694 TJANDRA ET AL.
GASTROENTEROLOGY Vol. 112, No. 5
for some assays. The data from the latter experiments were used to determine whether neutrophil dysfunction in generating ROIs results from a defect in the respiratory burst or possibly a suppressive effect of sera. To examine differential release of O20 by sham and BDR neutrophils, a nitroblue tetrazolium (NBT) slide test was performed.15 Freshly isolated neutrophils were incubated at 377C on glass slides with NBT dye (Sigma) and PMA (200 ng/mL) or buffer alone (resting). Neutrophils were fixed with absolute methanol before they were stained with safranin-o (BDH Inc., Toronto, Ontario, Canada). In the presence of O20, the NBT dye is reduced chemically to yield a dark purple insoluble compound (formazan), which can be clearly discerned microscopically and the percentage of cells containing the reduced NBT dye (formazan) can be determined easily. A comparison of staining intensity also can be made to control cells.
Degranulation Assay Microbicidal proteins and proteases released from the granules of neutrophils may also mediate neutrophil microbicidal activity. We examined exocytic neutrophil degranulation by measuring b-glucuronidase activity in the supernatant of stimulated and unstimulated neutrophils.14 Exocytic degranulation was determined by subtracting the amount of enzyme released in the presence of stimuli from that released in their absence. Briefly, neutrophils (1 1 107 cells) were incubated at 377C with DHCB (2.5 mg/mL) and stimulated with either FMLP (1007 mol/L) or PMA (200 ng/mL) for 30 minutes in a shaker bath (model 2564; Forma Scientific, Marrietta, OH). DHCB was added to promote exocytic degranulation. The supernatant from each tube was collected after centrifugation at 420g for 10 minutes and placed on ice. The b-glucuronidase levels were quantitated spectrofluorometrically (SPF-500 Spectrofluorometer; SLM Instruments Inc., Urbana, IL) against a known b-glucuronidase standard using excitation and emission wavelengths of 365 and 460 nm, respectively. Cell lysis was analyzed by measuring glucose-6-P-dehydrogenase activity spectrophotometrically at 340 nm (U-2000 double-beam spectrophotometer; Hitachi Ltd.).
Statistical Analysis
was further reduced in the presence of BDR (cholestatic) sera with an additional 50% reduction relative to the percentage of bacterial killing by BDR neutrophils in the presence of sham sera. In addition, the antibactericidal effect of BDR sera also was observed in bacterial killing by sham neutrophils (Figure 1). These data suggest that bacterial killing is impaired in cholestatic rats and that there is an inhibitory effect of BDR sera on neutrophil microbicidal activity. To elucidate the mechanisms of impaired bactericidal activity in BDR neutrophils, neutrophil phagocytosis, O20 production, and degranulation were investigated. Neutrophil Phagocytic Activity
Data are expressed as mean { SEM. An unpaired student’s t test was used for comparisons between two means, and analysis of variance followed by Student–Newman–Keuls test for comparisons between more than two means. A P value of °0.05 was considered significant.
Results Neutrophil Bactericidal Activity In the presence of autologous sera, sham neutrophils were more efficient at bacterial killing in vitro than BDR neutrophils (Figure 1). The ability of BDR neutrophils to kill bacteria was decreased by approximately 20% relative to sham neutrophils (Figure 1). Furthermore, impaired bacterial killing by BDR neutrophils / 5e1c$$0008
Figure 1. Percentage of S. aureus killed by sham (j) and BDR ( ) neutrophils after a 1-hour incubation. Neutrophils, sera, and bacteria were incubated in PBS with 1% gelatin. Aliquots were placed in water containing 1% gelatin for at least 10 minutes to ensure neutrophil lysis. Further dilutions were made into Tris buffer with 1% gelatin. Bacteria remaining in suspension were grown on tryptic soy agar plates overnight and individual colonies counted the next day. Data are mean { SEM of five experiments. *P õ 0.05 vs. sham neutrophils / BDR sera; /P õ 0.05 and //P õ 0.001 vs. sham neutrophils / sham sera; **P õ 0.001 vs. BDR neutrophils / sham sera.
04-11-97 16:55:04
gasa
Because phagocytosis is essential in neutrophil microbicidal function, a kinetic phagocytic assay of sham and BDR neutrophils was performed. Phagocytic activity was similar in sham and BDR neutrophils after 1-hour incubation with either S. aureus bioparticles opsonized with sham sera (sham: 8.72% { 3.3% vs. BDR: 5.51% { 1.14%; n Å 4) or S. aureus bioparticles opsonized with BDR sera (sham: 8.82% { 1.2% vs. BDR: 6.00% { 1.42%; n Å 4). O20 Release Neutrophils from BDR rats showed an approximate 30%–50% decrease in the rate of O20 release when stimulated with either 1007 or 1008 mol/L FMLP comWBS-Gastro
May 1997
IMPAIRED MICROBICIDAL ACTIVITY IN CHOLESTASIS 1695
Figure 2. O20 production by sham ( ) and BDR (j) rat neutrophils at increasing FMLP concentrations (10010 to 1005 mol/L). Neutrophils were primed with DHCB (2.5 mg/mL) for approximately 2 minutes at 377C before FMLP addition. Data are mean { SEM of three experiments (each with duplicate runs). **P õ 0.005 vs. sham value, *P õ 0.05 vs. sham value.
pared with sham neutrophils (Figure 2). However, at higher concentrations of FMLP (i.e., 1006 and 1005 mol/ L), there was no significant difference in the ability of neutrophils to generate O20. There was no appreciable release of O20 by either sham or BDR neutrophils when stimulated with lower doses of FMLP (1009 and 10010 mol/L). When stimulated with either PMA or OZ, sham and BDR neutrophils showed similar rates of O20 release (data not shown). The addition of either 10% sham or BDR sera to FMLP-stimulated sham neutrophils decreased the rate of O20 release by approximately 35% compared with the rate when neither sera were present, but without a further reduction in the rate of O20 release by BDR neutrophils (Figure 3). In addition, the rates of O20 release were not significantly different between sham and BDR neutrophils in the presence of either sera (Figure 3). The NBT staining indicated that the decreased rate of O20 release by BDR neutrophils was caused by a homogenous reduction of O20 release by BDR neutrophils. The results of NBT tests were comparable between sham and BDR neutrophils (93.8% { 0.2% and 92.1% { 1.1%, respectively), suggesting that the decrease in O20 release observed in BDR neutrophils is not caused by an inability to produce O20.
Figure 3. The rates of superoxide anion generated by sham (j) and BDR ( ) neutrophils in the presence of either 10% sham or BDR sera. Neutrophils were primed with DHCB (2.5 mg/mL) for 2 minutes at 377C before stimulation with 1007 mol/L FMLP. Data are mean { SEM of three experiments (each with duplicate runs). *P õ 0.05 vs. sham value.
FMLP (23.6% { 2.1% vs. 27.4% { 1.0%, respectively; NS) or 200 ng/mL PMA (21.2% { 3.7% vs. 25.5% { 3.1%, respectively; NS). The contribution of lysis to neutrophil degranulation, as determined by glucose-6-Pdehydrogenase activity in the supernatant, was negligible (data not shown).
Discussion To examine the ability of sham and BDR neutrophils to eliminate microorganisms, in vitro killing of S. aureus was analyzed. We observed a striking difference
Neutrophil Degranulation Neutrophil degranulation, quantitated as the percentage of b-glucuronidase released from primary granules, is shown in Figure 4. The percentage of b-glucuronidase released from sham and BDR neutrophils was similar whether they were stimulated with 1007 mol/L / 5e1c$$0008
04-11-97 16:55:04
gasa
Figure 4. b-Glucuronidase release from sham (j) and BDR ( ) neutrophils after incubation with DHCB (2.5 mg/mL) and stimulation with either 1007 mol/L FMLP or 200 ng/mL PMA for 30 minutes at 377C. After centrifugation at 420g for 10 minutes, the supernatant of each sample was collected and assayed for b-glucuronidase content using a spectrofluorometer (excitation l Å 365 nm; emission l Å 460 nm). Data are mean { SEM of five experiments.
WBS-Gastro
1696 TJANDRA ET AL.
GASTROENTEROLOGY Vol. 112, No. 5
in bacterial killing between sham and BDR neutrophils in that BDR neutrophils were less able to kill bacteria than sham neutrophils (P õ 0.05). Impaired ability to kill bacteria by BDR neutrophils may be caused by defective neutrophil function or the effect of cholestatic sera or both. Because BDR sera also exhibited an antimicrobicidal effect as shown by a significant reduction in bacterial killing by both sham and BDR neutrophils, we showed that impaired killing in BDR neutrophils arises from a combination of neutrophil functional defects and the antimicrobicidal effects of BDR sera. Phagocytosis is essential for bactericidal functions of neutrophils. A kinetic phagocytic assay using opsonized S. aureus, the same type of bacteria used in our killing studies, showed comparable phagocytic activity in sham and BDR neutrophils. This suggests that BDR neutrophils have normal phagocytic activity and are not responsible for the impaired neutrophil bactericidal activity we observed in BDR rats. To eliminate infectious agents, neutrophils generate ROIs and release their granule constituents. Therefore, impaired in vitro bacterial killing by BDR neutrophils may be caused by neutrophil defects in generating ROIs or releasing microbicidal proteins or a combination of both pathways. BDR neutrophils generated only half of the O20 released by sham neutrophils when stimulated with 1008 or 1007 mol/L FMLP (Figure 2). The NBT slide test results suggest that the impairment in O20 release by BDR neutrophils probably was caused by an effect on the entire population of BDR neutrophils, not caused by a specific subset of cells. In contrast, supraphysiological concentrations of FMLP (i.e., 1006 and 1005 mol/L), PMA, or OZ did not result in any significant difference in the rate of O20 release between sham and BDR neutrophils, suggesting that receptor and soluble agonists can equally activate sham and BDR neutrophils to produce O20. Higher levels of saturated fatty acids and cholesterol in plasma membranes of cholestatic patients and BDR rats18,19 may result in more rigid neutrophil membranes. This may influence FMLP receptor mobilization to the cell surface. In human neutrophils, low- and high-affinity binding sites for FMLP have been determined, whereas in rats, FMLP receptors appear to be of the high-affinity type, with a dissociation constant ranging from 1 to 30 nmol/L.20 Therefore, lower concentrations of FMLP would provide more sensitivity with respect to O20 production than higher concentrations. Similar O20 production in sham and BDR neutrophils at higher concentrations of FMLP may suggest a similar maximal capacity to produce O20. Because PMA bypasses membrane-associated neutrophil activation, similar O20 production in / 5e1c$$0008
04-11-97 16:55:04
gasa
BDR and sham neutrophils would be expected if there was a membrane- or receptor-determined defect. With respect to OZ, interpretation of our results is somewhat difficult. Although there is evidence to suggest that O20 production by human neutrophils is OZ-concentration dependent,21 detection of a concentration-dependent increase in O20 production by sham and BDR neutrophils stimulated with increasing OZ concentrations was difficult to assess because of increasing solution turbidity with higher OZ concentration. Our results contrast with those of Levy et al.22 who showed that neutrophils isolated from common bile duct–ligated rats generated more O20 than sham animals.22 Despite these differences, it is important to note that the levels of O20 produced in their study were much lower than what we observed. For example, using the same stimulus (1007 mol/L FMLP), the amount of O20 generated by their BDR neutrophils (which was significantly higher than sham neutrophils) was approximately 85% lower than what we observed with our BDR neutrophils. Furthermore, the range of O20 release in our study falls within a similar range observed with human neutrophils.15 Therefore, we are not certain about the physiological significance of such low levels of O20 production as reported by Levy et al.22 on neutrophil microbicidal function. The discrepancies between results may be explained by the use of different methods of blood collection and neutrophil isolation as well as assays to measure O20 generation. We chose the continuous assay because it is more sensitive in quantitating O20 release than the end point assay used by Levy et al.22 and permits one to observe directly the kinetics of neutrophil O20 production. To examine the effects of BDR sera on O20 release, 10% of either sham or BDR sera was included in the assay system. In contrast to the results of Ohshio et al.,23 but in agreement with Levy et al.,22 we were unable to see any increase in O20 release by neutrophils when incubated with either sham or BDR sera. In fact, we observed a 35% reduction in O20 release by sham neutrophils when exposed to either sera compared with its absence; there was comparable O20 production by sham and BDR neutrophils in the presence of either sera (Figure 3). A significant reduction in O20 release by sham neutrophils in the presence of either sera may be caused by antioxidant activity of the sera as physiological antioxidants (e.g., glutathione) and bystander antioxidants (e.g., plasma proteins) scavenge ROIs.24 – 26 Therefore, BDR sera did not appear to affect O20 release by BDR neutrophils. Some cytokines have been shown to be capable of priming neutrophils to enhance O20 release.17,27 – 29 WBS-Gastro
May 1997
IMPAIRED MICROBICIDAL ACTIVITY IN CHOLESTASIS 1697
Woodman et al.17 documented that preincubation of human neutrophils with granulocyte-macrophage colony– stimulating factor caused an augmentation in the rate of O20 production but without a significant increase in the maximal amount generated. Because glucocorticoids can inhibit cytokine release,30 the elevated levels of endogenous glucocorticoids previously documented in BDR animals11 may depress cytokine production. This in turn may affect neutrophil priming and activation. The role of cytokines in neutrophil O20 release in BDR rats warrants further investigation. Strong evidence for the contribution of granule constituents in neutrophil microbicidal activity has been shown in studies using neutrophils obtained from patients with chronic granulomatous disease who are deficient in functional reduced nicotinamide adenine dinucleotide phosphate oxidase.21,31 Therefore, we examined the degranulation process in sham and BDR neutrophils. The release of granule contents by stimulated sham and BDR neutrophils, measured as b-glucuronidase release, was similar between the two types of neutrophils (Figure 4). These data suggest that lowered microbial resistance in cholestasis is not caused by an impaired neutrophil degranulation. In addition, impaired O20 production in BDR neutrophils did not seem to be caused by a defect in neutrophil receptor function. In conclusion, we show that in vitro bactericidal function of neutrophils isolated from cholestatic rats is impaired. Specifically, we identified the following two functional defects that may contribute to increased infections in cholestasis: (1) decreased O20 generation and bacterial killing and (2) an unidentified antimicrobicidal effect of BDR sera that further impairs the bactericidal ability of neutrophils. The combined effects of neutrophil dysfunction and impaired recruitment, which we have identified previously in cholestatic rats,11 may explain the increased incidence of infective complications in cholestatic patients after surgery. These observations need to be confirmed in humans, but merit further investigation.
and clinical correlates. 2nd ed. New York: Raven, 1992:541– 588. Elsbach P, Weiss J. Oxygen-independent antimicrobial systems of phagocytes. In: Gallin JI, Goldstein M, Snyderman R, eds. Inflammation: basic principles and clinical correlates. 2nd ed. New York: Raven Press, 1992:603–636. Babior BM. The respiratory burst (pp. 132–134, Lehrer RI, moderator). Neutrophils and host defense. Ann Intern Med 1988; 109:127–142. Selsted ME. Nonoxidative killing by neutrophils (pp. 129–132, Lehrer RI, moderator). Neutrophils and host defense. Ann Intern Med 1988;109:127–142. Babior BM. Neutrophil function as related to neutrophil-endothelial cell interactions. Nouv Rev Fr Hematol 1992;34(Suppl):S29– S35. Swain MG, Tjandra K, Kanwar S, Kubes P. Neutrophil adhesion is impaired in a rat model of cholestasis. Gastroenterology 1995; 109:923–932. Cameron GR, Oakley CL. Ligation of the common bile duct. J Pathol Bacteriol 1932;35:769–798. Tjandra K, Kubes P, Rioux K, Swain MG. Endogenous glucocorticoids inhibit neutrophil recruitment to inflammatory sites in cholestatic rats. Am J Physiol 1996;270:G821–G825. Woodman RC, Reinhardt PH, Kanwar S, Johnston FL, and Kubes P. The effects of human neutrophil elastase (HNE) on neutrophil function in vitro and in inflamed microvessels. Blood 1993;82: 2188–2195. Woodman RC, Newburger PE, Anklesaria P, Erickson RW, Rae J, Cohen MS, Curnutte JT. A new X-linked variant of chronic granulomatous disease characterized by the existence of a normal clone of respiratory burst-component phagocytic cells. Blood 1995;85: 231–241. Schiff DE, Rae J, Davis BH, Curnutte JT. In vivo recombinant human interferon-g (rhIFN-g) treatment of normal human subjects is associated with increased FcgRI expression by phagocytes and improved phagocytosis of S. aureus (abstr). Blood 1995;86:28a. Woodman RC, Curnutte JT, Babior BM. Evidence that de novo protein synthesis participates in a time-dependent augmentation of the chemotactic peptide induced respiratory burst in neutrophils—effects of recombinant human colony stimulating factors and dihydrocytochalasin B. Free Radical Biol & Med 1988;5: 355–361. Scriven MW, Horrobin DF, Puntis MCA. Study of plasma and red cell phospholipid fatty acids in extrahepatic cholestatic jaundice. Gut 1994;35:987–990. Lang ME, Jourd’heuil D, Meddings JB, Swain MG. Increased opioid binding to peripheral white blood cells in a rat model of acute cholestasis. Gastroenterology 1995;108:1479–1486. Walker BAM, Seiler AJ, Owens CA, Hagenlocker BE, Ward PA. Absence of FMLP receptors on rat macrophages. J Leukocyte Biol 1991;50:600–606. Root RK, Cohen MS. The microbicidal mechanisms of human neutrophils and eosinophils. Rev Infect Dis 1981;3:565–598. Levy R, Schlaeffer F, Keynan A, Nagauker O, Yaari A, Sikuler E. Increased neutrophil function induced by bile duct ligation in a rat model. Hepatology 1993;17:908–914. Ohshio G, Miyachi Y, Kudo H, Niwa Y, Manabe T, Tobe T. Effects of sera from patients with obstructive jaundice on the generation of oxygen intermediates by normal polymorphonuclear leukocytes. Liver 1988;8:366–371. Dobrinich R, Spagnuolo PJ. Binding of C-reactive protein to human neutrophils. Arthritis Rheum 1991;34:1031–1038. Bucurenci N, Blake DR, Chidwick K, Winyard PG. Inhibition of neutrophil superoxide production by human plasma a1-antitrypsin. FEBS Lett 1992;300:21–24. Kim SJ, Han D, Moon KD, Rhee JS. Measurement of superoxide dismutase-like activity of natural antioxidants. Biosci Biotechnol Biochem 1995;59:822–826.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16.
17.
18.
19.
20.
21.
References 22. 1. Amstrong CP, Dixon JM, Taylor TV, Davies GC. Surgical experience of deeply jaundiced patients with bile duct obstruction. Br J Surg 1984;71:234–238. 2. Pain JA, Cahill CJ, Bailey ME. Perioperative complications in obstructive jaundice: therapeutic considerations. Br J Surg 1985; 72:942–945. 3. Cainzos M, Potel J, Puente J. Anergy in jaundiced patients. Br J Surg 1988;75:147–149. 4. Pitt HA, Cameron JL, Postier RG, Gadacz TR. Factors affecting mortality in biliary tract surgery. Am J Surg 1981;141:66–72. 5. Babior BM. Oxidants from phagocytes: agents of defense and destruction. Blood 1984;64:959–966. 6. Klebanoff SJ. Oxygen metabolites from phagocytes. In: Gallin JI, Goldstein M, Snyderman R, eds. Inflammation: basic principles
/ 5e1c$$0008
04-11-97 16:55:04
gasa
23.
24. 25.
26.
WBS-Gastro
1698 TJANDRA ET AL.
GASTROENTEROLOGY Vol. 112, No. 5
27. Weisbart RH, Kwan L, Golde DW, Gasson JC. Human GM-CSF primes neutrophils for enhanced oxidative metabolism in response to the major physiological chemoattractants. Blood 1987;69:18–21. 28. Bajaj MS, Kew RR, Webster RO, Hyers TM. Priming of human neutrophil functions by tumor necrosis factor: enhancement of superoxide anion generation, degranulation, and chemotaxis to chemoattractants C5a and F-Met-Leu-Phe. Inflammation 1992; 16:241–250. 29. Wozniak A, Betts WH, Murphy GA, Rokicinski M. Interleukin-8 primes human neutrophils for enhanced superoxide anion production. Immunology 1993;79:608–615. 30. Hawes AS, Rock CS, Keogh CV, Lowry SF, Calvano SE. In vivo effects of the antiglucocorticoid RU 486 on glucocorticoid and cytokine responses to Escherichia coli endotoxin. Infect Immunol 1992;60:2641–2647.
/ 5e1c$$0008
04-11-97 16:55:04
gasa
31. Kaplan EL, Laxdal T, Quie PG. Studies of polymorphonuclear leukocytes from patients with chronic granulomatous disease of childhood: bactericidal capacity for streptococci. Pediatrics 1968;41:591–599.
Received June 21, 1996. Accepted December 26, 1996. Address requests for reprints to: Mark G. Swain, M.D., Gastroenterology Research Group, Health Sciences Center, 3330 Hospital Drive Northwest, Calgary, Alberta, Canada T2N 4N1. Fax: (403) 270-0995. Supported by grants from Alberta Heritage Foundation for Medical Research (AHFMR) and Medical Research Council (MRC) of Canada. R.C.W. is an AHFMR Medical Scholar and M.G.S. is an MRC Scholar and AHFMR Clinical Investigator. The authors thank M. Maric, D. Teoh, and F. Johnston for technical assistance.
WBS-Gastro