343
CONCISE COMMUNICATION
Endotoxin Down-Modulates Granulocyte Colony-Stimulating Factor Receptor (CD114) on Human Neutrophils Ursula Hollenstein,1,2 Monika Homoncik,1 Petra Jilma Stohlawetz,3 Claudia Marsik,1 Anna Sieder,1 Hans-Georg Eichler,1 and Bernd Jilma,1
1
Departments of Clinical Pharmacology, The Adhesion Research Group Elaborating Therapeutics, 2Internal Medicine I, Division of Infectious Diseases, and 3Blood Group Serology and Transfusion Medicine, Division of Transfusion Medicine, Vienna University School of Medicine, Vienna, Austria
During infection, the development of nonresponsiveness to granulocyte colony-stimulating factor (G-CSF) may be influenced by the down-modulation of G-CSF receptor (G-CSFR) by cytokines. This down-modulation was studied during experimental human endotoxemia. Healthy volunteers received either 2 ng/kg endotoxin (lipopolysaccharide [LPS], n = 20 ) or placebo (n = 10) in a randomized, controlled trial. Endotoxin infusion increased the mean fluorescence intensity of the neutrophil activation marker CD11b 1300% after 1 h (P ! .001 vs. placebo). LPS infusion down-modulated G-CSFR expression in as early as 60 min (217%; P = .001 vs. placebo). Down-modulation was almost maximal at 90 min and persisted for 6 h (250% from baseline; P ! .0001 vs. placebo). Plasma levels of G-CSF started to increase only after G-CSFR down-modulation had occurred and peaked 37-fold above baseline at 4 h (P ! .0001 vs. placebo). In conclusion, LPS down-modulates G-CSFR expression in humans, which may render neutrophils less responsive to the effects of G-CSF and, thereby, compromise host defense mechanisms.
Granulocyte colony-stimulating factor (G-CSF) is crucial for proliferation of neutrophils [1], delays apoptosis of neutrophils [2], and enhances neutrophil functions, such as oxidative burst [3] and phagocytic and bactericidal activity [4]. Endogenous plasma levels of G-CSF are normally in the low picomolar range [1, 5]. In patients undergoing a successful response to infection, G-CSF levels are high in the acute phase and decrease consecutively [5]. G-CSF is removed by binding to its receptor (G-CSFR; CD114) and in reaction to increased neutrophil counts [1], which contribute substantially to clearance of G-CSF. Endogenous G-CSF concentrations are initially very high in bacteremic patients [5, 6]. This could be explained in part by decreased clearance due to so-called trans down-modulation of G-CSFR surface expression on neutrophils by mediators such as lipopolysaccharide (LPS) or cytokines (i.e., down-modulation by a stimulus other than G-CSF). This heterologous GReceived 14 February 2000; revised 27 March 2000; electronically published 6 July 2000. Informed consent was obtained from all volunteers, according to the Declaration of Helsinki, and study procedures conformed to good clinical practices. ¨ sterreichischen Nationalbank. Financial support: Jubila¨umsfonds der O Reprints or correspondence: Dr. Bernd Jilma, Dept. of Clinical Pharmacology, The Adhesion Research Group Elaborating Therapeutics, Vienna University Hospital School of Medicine, Waehringer Guertel 18-20, A-1090 Wien, Austria (
[email protected]). The Journal of Infectious Diseases 2000; 182:343–6 q 2000 by the Infectious Diseases Society of America. All rights reserved. 0022-1899/2000/18201-0048$02.00
CSFR down-modulation has been observed in vitro [7], and, if it occurs in vivo, it could play an important role in the development of nonresponsiveness to G-CSF and the occurrence of relative or absolute neutropenia in sepsis. It could prevent further recruitment of neutrophils from the bone marrow into the circulation and, thereby, decrease host defense mechanisms. We studied trans down-modulation of G-CSFR during experimental human endotoxemia [8]. We characterized the time course of G-CSFR expression and G-CSF plasma levels, in healthy, LPS-challenged volunteers, and their relation to leukocyte activation, as measured by expression of the b2 integrin CD11b. Methods Study design. The study design was 2:1 randomized (LPS: placebo), double blind, placebo controlled, and in parallel groups. Study subjects. Healthy male volunteers were invited to participate in this trial. All subjects were 19–35 years of age with a body mass index between the 15th and the 85th percentile. Determination of health status included medical history, physical examination, laboratory parameters, and virologic and standard drug screening. In addition, study subjects were tested for hereditary thrombophilia (i.e., factor V Leiden, protein C, and protein S deficiency), to minimize potential risks imposed by endotoxin-induced coagulation activation. Exclusion criteria were regular or recent intake of medication, including nonprescription medication, and relevant abnormal findings in medical history or laboratory parameters. Study protocol. The experimental procedures of our endotoxin
344
Hollenstein et al.
infusion studies have been described in detail elsewhere [9]. In brief, volunteers were admitted to the study ward at 8:00 AM after an overnight fast, because the response to endotoxin varies with the time of day [10]. Throughout the study period, subjects were confined to bed rest and kept fasting for 8.5 h after LPS infusion. A 5% glucose infusion (Leopold Pharma, Vienna) was started at 8:30 AM and continued over 8.5 h at 3 mL/kg/h, to maintain adequate blood glucose levels and urinary output. Thirty minutes after the start of glucose, 20 subjects received an intravenous bolus of 2 ng/kg LPS (National Reference Endotoxin, Escherichia coli; USP Convention, Rockville, MD), 10 other subjects served as a control group, receiving normal saline solution instead of LPS. Sampling and analysis. Analysts were blinded with regard to group allocation. Sampling times were selected on the basis of the kinetics of G-CSFR down-regulation in pilot subjects challenged with LPS and on the kinetics of homologous G-CSFR down-regulation (i.e., down-modulation of G-CSFR by infusion of G-CSF in humans; unpublished data). Blood samples were obtained by venipuncture and collected into EDTA-anticoagulated Vacutainer tubes (Becton Dickinson, Vienna) before LPS infusion; thereafter, blood was obtained at times indicated in the figures (except samples for leukocyte counts, which were obtained from an indwelling venous line on the contralateral arm, into which LPS had been administered). Plasma samples were processed immediately by centrifugation at 2000 g at 47C for 15 min and stored at 2807C before analysis. G-CSF plasma levels were analyzed with a high-sensitivity EIA (R&D Systems, Oxon, UK) [11], and samples from individual subjects were run in the same assay. G-CSFR expression and CD11b expression were quantified by flow cytometry (FACSCalibur; Becton Dickinson). The anti G-CSFR antibody (Serotech, Kidlington, UK) [11] and the anti-CD11b antibody were phycoerythrin labeled (Becton Dickinson). Results are presented as mean fluorescence intensity (MFI). Neutrophil counts were obtained with a cell counter (Sysmex, Kyoto, Japan). Data analysis. Data are expressed as the mean and the 95% confidence intervals for description in the text. Nonparametric statistics were applied. All statistical comparisons within groups were done with the Friedman analysis of variance and the Wilcoxon signed ranks test for post hoc comparisons. The Mann-Whitney U test was used to test changes in end points between groups for statistical significance. P ! .05 (2-tailed) was considered significant.
Results Endotoxin infusion rapidly induced neutrophil activation within 1 h, as measured by surface expression of CD11b, an adhesion molecule and marker of leukocyte degranulation (figure 1). Basal MFI of CD11b ranged from 49 to 1114, with no differences between groups. In line with the observation that LPS increases CD11b expression at 3 h [12], we have now characterized in more detail the time course of CD11b (figure 1): LPS infusion had already enhanced CD11b expression 1300% after 60 min (P ! .0005 vs. baseline; P = .0002 vs. placebo), and the increase in CD11b expression persisted for at least 6 h. Baseline MFI of G-CSFR expression averaged 56 (range,
JID 2000;182 (July)
Figure 1. A, Effect of endotoxin infusion on surface expression of b2-integrin (CD11b). Effect of granulocyte colony-stimulating factor receptor (G-CSFR) on neutrophils (B) and on plasma levels of endogenous granulocyte colony-stimulating factor (G-CSF) (C). Volunteers received either 2 ng/kg Escherichia coli endotoxin (n = 20) or placebo (n = 10). As indicated by the vertical dashed line, G-CSFR down-modulation occurred before an increase in G-CSF plasma levels was detectable. LPS, lipopolysaccharide; MFI, mean fluorescence intensity. Data are mean 5 SE. P ! .001 vs. baseline and between groups for all parameters.
40–67; P 1 .05 between groups). LPS infusion down-modulated G-CSFR expression as early as 60 min (217%; P ! .001 vs. baseline; P = .001 between groups; figure 1). Down-modulation was almost maximal at 90 min and persisted for 6 h (250% from baseline), whereas no significant decrease was observed in the placebo group. Baseline G-CSF plasma levels averaged 19 pg/mL (range, 7–60 pg/mL; P 1 .05 between groups). Of note and in agreement with data elsewhere [12], G-CSF plasma levels did not change during the first hour after LPS infusion (figure 1), but they did increase after 2 h. LPS increased G-CSF plasma levels 37-fold at 4 h, whereas placebo had no effect (P 1 .05). As expected [9], neutrophil counts dropped sharply 60 min after LPS infusion and increased significantly above baseline from 2 h to at least 7 h (figure 2). Basal activation of neutrophils (i.e., CD11b expression) cor-
JID 2000;182 (July)
LPS Down-Modulates G-CSF Receptor
Figure 2. Time course of neutrophil counts after endotoxin infusion. Volunteers received either 2 ng/kg Escherichia coli endotoxin (n = 20) or placebo (n = 10). LPS, lipopolysaccharide. Data are mean 5 SE.
related negatively with expression of G-CSFR (r = 20.53, P = .0027). G-CSFR expression and G-CSF levels were inversely correlated (r = 20.38, P = .044) at baseline.
Discussion The major aim of this study was to examine whether LPS infusion induces heterologous trans down-modulation of GCSFR on neutrophils in vivo, as was observed in vitro [7]. Indeed, our results demonstrate that LPS down-modulates CD114 expression on the neutrophil surface within 60 min after intravenous LPS (figure 1). This decrease in G-CSFR occurred before any increase was seen in plasma levels of G-CSF (figure 1), which cause homologous G-CSFR down-modulation (unpublished data). Concomitant with the G-CSFR down-modulation, we observed massive activation of neutrophils, as evidenced by the increase in CD11b expression (figure 1). Our results thus confirm in vitro observations elsewhere that LPS trans down-modulates G-CSFR either directly [7] or by a mediator, such as tumor necrosis factor [13]. Down-modulation of G-CSFR by endotoxin is of clinical interest for several reasons. First, decreased expression of GCSFR is expected to enhance G-CSF levels by decreasing clearance. Together with increased G-CSF release during inflammation, this may partly contribute to the high G-CSF concentrations seen in bacteremic patients [5, 6]. Second, GCSF concentrations decrease with recovery in survivors of sepsis but remain increased in those who do not survive [14]. We postulate that this difference is primarily due to a difference in G-CSF plasma clearance because of a lack of availability of G-CSFR. Thus, heterologous G-CSFR down-modulation could play an important role in the development of nonresponsiveness to G-CSF and the occurrence of neutropenia in
345
sepsis: it could prevent further recruitment of neutrophils from the bone marrow into the circulation and thereby decrease host defense mechanisms. As far as neutrophil counts are concerned, it appears that the early LPS-induced neutrophilia observed at 2 h was not due to G-CSF, because infusion of exogenous GCSF increases neutrophil counts only after a time lag of 2 h (unpublished data). Last, the observed correlations, although not proof of a cause-effect relationship, may reflect the following biologic mechanisms: since G-CSF levels and G-CSFR expression were inversely correlated, it appears that the G-CSFR/ligand pair exert a negative feedback on each other. The highly significant negative correlation between the expression of CD11b and GCSFR at baseline indicates that neutrophil activation and GCSFR expression are tightly coregulated even in healthy subjects. The time course of CD11b expression was examined in more detail in this study because CD11b can function as a receptor for a large number of molecules, including ICAM-1, LPS, iC3b, high–molecular-weight kininogen, fibrinogen, and coagulation factor X [15]. As a consequence, it may be hypothesized that early up-regulation of this adhesion molecule plays an essential role in the very early phase of adhesion to and phagocytosis of bacteria. Further, CD11b expression may contribute to an alternative pathway of LPS-induced coagulation [15]. Thus, CD11b could represent an interesting therapeutic target in sepsis. In conclusion, LPS trans down-modulates G-CSFR expression in humans, which may render neutrophils less responsive to the effects of G-CSF and thereby compromise host defense mechanisms.
References 1. Weiss M, Moldawer LL, Schneider EM. Granulocyte colony-stimulating factor to prevent the progression of systemic nonresponsiveness in systemic inflammatory response syndrome and sepsis. Blood 1999; 93:425–39. 2. Dalhoff K, Hansen F, Dromann D, Schaaf B, Aries SP, Braun J. Inhibition of neutrophil apoptosis and modulation of the inflammatory response by granulocyte colony-stimulating factor in healthy and ethanol-treated human volunteers. J Infect Dis 1998; 178:891–5. 3. Allen RC, Stevens PR, Price TH, Chatta GS, Dale DC. In vivo effects of recombinant human granulocyte colony-stimulating factor on neutrophil oxidative functions in normal human volunteers. J Infect Dis 1997; 175: 1184–92. 4. Roilides E, Walsh TJ, Pizzo PA, Rubin M. Granulocyte colony-stimulating factor enhances the phagocytic and bactericidal activity of normal and defective human neutrophils. J Infect Dis 1991; 163:579–83. 5. Pauksen K, Elfman L, Ulfgren AK, Venge P. Serum levels of granulocytecolony stimulating factor (G-CSF) in bacterial and viral infections, and in atypical pneumonia. Br J Haematol 1994; 88:256–60. 6. Cebon J, Layton JE, Maher D, Morstyn G. Endogenous haemopoietic growth factors in neutropenia and infection. Br J Haematol 1994; 86: 265–74. 7. Nicola NA. Receptors for colony-stimulating factors. Br J Haematol 1991; 77: 133–8. 8. Suffredini AF, Hochstein HD, McMahon FG. Dose-related inflammatory
346
Hollenstein et al.
effects of intravenous endotoxin in humans: evaluation of a new clinical lot of Escherichia coli O:113 endotoxin. J Infect Dis 1999; 179:1278–82. 9. Jilma B, Blann A, Pernerstorfer T, et al. Regulation of adhesion molecules during human endotoxemia: no acute effects of aspirin. Am J Respir Crit Care Med 1999; 159:857–63. 10. Pollmacher T, Mullington J, Korth C, et al. Diurnal variations in the human host response to endotoxin. J Infect Dis 1996; 174:1040–5. 11. Jilma B, Hergovich N, Stohlawetz P, Eichler HG, Bauer P, Wagner OF. Circadian variation of granulocyte colony stimulating factor levels in man. Br J Haematol 1999; 106:368–71. 12. Kuhns DB, Alvord WG, Gallin JI. Increased circulating cytokines, cytokine
JID 2000;182 (July)
antagonists, and E-selectin after intravenous administration of endotoxin in humans. J Infect Dis 1995; 171:145–52. 13. Root RK, Dale DC. Granulocyte colony-stimulating factor and granulocytemacrophage colony-stimulating factor: comparisons and potential for use in the treatment of infections in nonneutropenic patients. J Infect Dis 1999; 179 (Suppl 2):S342–S52. 14. Kragsbjerg P, Holmberg H, Vikerfors T. Dynamics of blood cytokine concentrations in patients with bacteremic infections. Scand J Infect Dis 1996; 28:391–8. 15. Corbi AL. Adhesive functions of the leukocyte integrins and their ligands. In: Corbi AL, ed. Leukocyte integrins: structure, expression and function. 1st ed. New York: Springer, 1996:79–98.
