Alteration of Chemoattractant Receptor Expression Regulates Human Neutrophil Chemotaxis In Vivo

Abstract

Objective

To elucidate the mechanisms that regulate human neutrophil delivery in vivo, as well as the mechanisms that lead to observed reduction in polymorphonuclear (PMN) delivery to remote sites in septic patients.

Methods

Alterations in human PMN chemoattractant receptor expression and chemotactic function in vivo were evaluated in two distinct experiments: exudate PMNs (PMNs that have undergone transmigration to skin window blisters in controls) and septic PMNs (circulating PMNs from septic patients in the intensive care unit) were both separately compared with control circulating PMNs.

Results

Exudate PMNs displayed increased C5a receptors and C5a chemotaxis, and reduced interleukin-8 receptors (both IL-8 RA and IL-8 RB) and IL-8 chemotaxis. Septic PMNs displayed reduced C5a and IL-8 receptors and decreased C5a chemotaxis but no change in IL-8 chemotaxis. IL-8 but not C5a receptor gene expression decreased in parallel to receptor alteration.

Conclusions

These results suggest that change in PMN chemoattractant receptor expression serves to regulate PMN chemotaxis in vivo; that exudate PMN chemotaxis depends more on C5a than IL-8; and that diminished chemoattractant receptors and chemotaxis in septic PMNs may explain decreased PMN delivery in these patients.

An appropriate and effective host response to infection is of fundamental importance to host survival. Thus, a thorough understanding of the alterations in host response in critically ill patients has long been of particular interest. ^1,2 Appropriate recruitment of polymorphonuclear neutrophils (PMNs) to a site of inflammation is a principal component of effective host defense against bacterial and fungal infection. ³ We have previously shown that septic patients have reduced delivery of neutrophils to skin blisters, ^4,5 and we believe that diminished PMN delivery to remote sites may contribute to sepsis-related immunosuppression, leading to “second-front” infections, subsequent organ dysfunction, and death. Given the essential role of the PMN in both health and disease, our investigations have focused on the regulation of human PMN delivery in vivo.

Directional migration or chemotaxis along the concentration gradient of a leukocyte chemoattractant is essential for effective leukocyte delivery to a site of infection. PMNs undergo directional migration or chemotaxis toward an increasing concentration of a variety of chemoattractant substances. C5a, the most potent proinflammatory and chemotactic anaphylatoxin, is formed during complement activation and serves as a nonspecific chemoattractant for monocytes, neutrophils, eosinophils, and basophils. ⁶ The effect of C5a on PMN is mediated by the C5a receptor (C5aR), identified on neutrophils and other peripheral blood leukocytes. ⁷ It is a member of the superfamily of rhodopsin-type receptors, all containing seven transmembrane loops. ⁸ In addition to chemotaxis, C5aR also mediates a proinflammatory response in PMN, including the production of superoxide anions, ⁹ and the release of proteolytic enzymes. ¹⁰ The capacity of mediators to alter C5aR expression in vitro is well known. ^11,12 When C5a binds to C5aR, both ligand and receptor are internalized, with the potential for cell surface reexpression of the receptor approximately 100 minutes later. ^13,14 Although C5a receptors have been shown to be decreased in anergic patients, ⁵ the functional significance of increased or decreased C5aR expression in vivo is unknown.

Interleukin (IL)-8 is the prototype of a supergene family of chemokines (chemotactic cytokines) that possess specific chemotactic activity for neutrophils and is implicated in the pathogenesis of inflammatory disease. ¹⁵ IL-8 is produced by virtually all nucleated cells in response to inflammatory stimuli such as endotoxin, tumor necrosis factor-alpha, and IL-1. ¹⁶ There are two receptors for IL-8: type A receptor or IL-8 RA (also named CXC R1) and type B receptor or IL-8 RB (CXC R2). Both IL-8 receptors have seven transmembrane regions, are coupled to G proteins, and share 29% amino acid sequence homology with C5aR. ¹⁷ IL-8 receptors are expressed on monocytes, a subset of NK cells and T cells, but are most strongly expressed on neutrophils. ¹⁸ IL-8 RA and IL-8 RB differ in their selectivity for other ligands, and their biologic differences remain to be clarified. ¹⁹ Although numerous mediators modulate IL-8 R expression, including granulocyte colony stimulating factor (G-CSF), lipopolysaccharide, and IL-8, ^20,21in vivo regulation of IL-8 RA and IL-8 RB and their functional significance remain to be determined.

Although chemoattractant receptors are known to be responsible for mediating PMN response to various chemotactic factors, including both C5a and IL-8, in vivo alteration of chemoattractant receptors, the mechanism for their alteration, and the functional significance of receptor alteration are unclear. We hypothesized that PMN chemoattractant receptor expression displays significant alteration in humans in vivo, that receptor alteration may be increased or decreased, and that PMN receptor alteration is associated with a parallel functional change in PMN chemotaxis. To investigate these hypotheses, we evaluated the simultaneous alterations in PMN chemoattractant receptors and PMN chemotaxis in two human PMN populations that are functionally distinct from control circulating PMNs. Thus, in two separate but related experiments, we measured the change from control circulating PMNs in chemoattractant receptor expression and PMN chemotaxis in exudate PMNs (PMNs that have necessarily undergone transmigration to skin blisters) and septic PMNs (circulating PMNs exposed to increased levels of circulating proinflammatory mediators); changes in C5a and IL-8 receptor expression were compared with changes in PMN function, or chemotaxis to C5a and IL-8. To evaluate the mechanism for receptor alteration, receptor gene expression was evaluated in all three PMN populations. The purpose of these investigations includes the further elucidation of the mechanisms that regulate human neutrophil delivery in vivo, as well as the mechanisms that lead to observed reduction in PMN delivery to remote sites in septic patients.

METHODS

Subjects

The subjects under evaluation in this study were healthy human controls and septic patients. Circulating PMNs were isolated from septic patients, whereas both circulating and exudate PMNs were collected simultaneously from healthy controls. Sepsis was defined by the presence of active infection requiring antibiotic treatment along with the systemic inflammatory response syndrome (identified by having two or more of the following criteria: body temperature >38°C or <36°C; heart rate more than 90; tachypnea [respiratory rate >20 or Paco₂ <32 mm Hg]; white cell count >12.0 × 10⁹/L or <4.0 × 10⁹/L). ²² All patients were being treated for active infection. Exclusion criteria for septic patients included the following: having received a transfusion of greater than 5 units blood or blood products within 48 hours, chemotherapy or radiotherapy in the previous 3 months, steroid administration, or hemodialysis; liver failure (Child B or C), known HIV positivity, hypovolemic shock, or hypotension requiring vasoactive drugs (e.g., dopamine at >8 μg/kg/min or norepinephrine >8 μg/min). Exclusion criteria for the control cohort included a history of infection within the previous 48 hours, severe chronic illness, immunosuppressive medication, and known malignancy. Informed consent was obtained from all patients and controls. The study was approved by the Committee on Human Experimentation, McGill University Health Center Research Institute.

Reagents

All preparations of neutrophils were kept in polypropylene tubes to prevent adherence. Neutrophil incubation was in Dulbecco’s phosphate-buffered saline (PBS) without calcium and magnesium (Flow Laboratories, Mississauga, Ontario). Recombinant human C5a and IL-8 (both Sigma Chemical Co., St. Louis, MO) were stored at −70°C in 0.1-mL aliquots at a concentration of 100 ng/mL and 10 ng/mL, respectively. All glassware was baked at 250°C for at least 6 hours. Disposable sterile plastic pipettes and polypropylene tubes were used whenever possible.

Isolation of Circulating Neutrophils

Seven milliliters of whole blood was obtained from the subjects using heparinized vacuum-sealed tubes (Becton Dickenson, Franklin Lakes, NJ). The blood was immediately added to 1.5 mL Macrodex/Dextran-70 (Pharmacia Laboratories, Piscataway, NJ), and gently mixed. The erythrocytes were gravity sedimented for 60 minutes at room temperature. The leukocyte-rich supernatant was removed and centrifuged at 400 g for 5 minutes at 4°C. The pellet was gently resuspended in Dulbecco’s PBS without calcium and magnesium (Flow Laboratories) and layered on 3.0 mL Ficoll-Hypaque (Pharmacia Laboratories). Centrifugation continued at 400 g for 25 minutes at 4°C. The plasma, lymphocyte interface, and Ficoll were removed, and the pellet was resuspended in Elyse (Cardinal Associates, Santa Fe, NM) to lyse erythrocytes. After repelleting the neutrophils (400 g for 25 minutes at 4°C), they were washed with and then resuspended in iced PBS. Cells were counted with a hemocytometer after staining with Turk’s solution and suspended in media at a concentration of 1 × 10⁶cells/mL. Cell viability (measured by propidium iodide [Sigma Chemical Co.] or trypan blue exclusion) and purity (assessed by flow cytometry or microscopic field examination) exceeded 90% in all experiments.

Isolation of Exudate Neutrophils

Skin window chambers were manufactured at the McGill University Workshop and used as previously described. ²³ The technique follows the one described by Zimmerli and Galin. ²⁴ Briefly, exudate neutrophils were collected from skin windows placed on the volar aspect of the subject’s forearm, previously sterilized with 10% povidone-iodine topical antiseptic and 70% isopropyl alcohol. Vacuum suction (360 mm Hg) was applied using a Plexiglas template (contains four separate 1.0-cm-diameter chambers attached to two side ports that connect to suction) for 45 to 90 minutes until four 1.0 × 1.0-cm blisters formed. These blisters were unroofed using sterile scissors, and a second template consisting of four open-bottomed chambers (1.0-cm-diameter inferior opening next to skin) was tightly applied using wide adhesive tape. The chambers were individually filled with 10% autologous serum through superior portholes (0.2 cm in diameter). The portholes were subsequently sealed with a sterile covering (OpSite, Smith & Nephew, Hull, England). After 18 to 22 hours, the sterile covering was removed and the exudate fluid was aspirated through the portholes. Neutrophil purity within the exudate fluid was more than 98% as assessed by microscopic field examination; no red blood cells were present. To maximize cell yield, chambers were rinsed three times with normal saline. Neutrophils were sedimented at 400 g at 4°C for 5 minutes. Neutrophil viability was confirmed to be more than 95% using propidium iodide exclusion. Neutrophils were counted using a hemocytometer after staining with Turk’s solution and resuspended in iced PBS at a concentration of 1 × 10⁶ cells/mL for immediate quantification of surface receptors. Using this technique, a population of approximately 2 to 10 × 10⁶ pure and viable exudate PMNs were collected from each healthy control.

Quantification of Surface Receptors

Circulating neutrophils were collected in 7-mL sterile Vacutainer tubes containing EDTA as anticoagulant (Becton Dickinson). The whole blood was kept on ice until analysis (begun immediately). Exudate neutrophils were analyzed as soon as they were suspended in iced PBS (1 × 10⁶ cells/mL). Chemoattractant receptor density was analyzed using specific immunofluorescent-labeled monoclonal antibodies to the C5aR (CD88), IL-8 RA (CXC R1), and IL-8 RB (CXC R2). Mouse antihuman (IgG₁) CD88-FITC (Serotec Ltd., Oxford, England), mouse antihuman (IgG_2b) IL-8 RA-PE (Pharmingen Canada, Ontario), and mouse antihuman (IgG₁) IL-8 RB-PE (Pharmingen Canada) were used to detect the chemoattractant receptors in the following manner. In 5-mL tubes, 10 μL (1–2 μg) immunofluorescent monoclonal antibody was incubated along with 100 μL whole blood for 30 minutes at room temperature in the dark. Subsequently, 2 mL monoclonal lysing solution (1 L dH₂O, 8.26 g NH₄Cl, 1 g KHCO₃, 0.037 g Na₂EDTA) was added to the mixture, and again it was stored in the dark for 10 minutes. After neutrophil sedimentation (400 g at 4°C for 5 minutes), cells were washed twice with PBS with 1% fetal bovine serum and 0.1% Azide (Sigma, Oakville, Ontario), and were finally suspended in 2% paraformaldehyde (Fisher, Ontario) and promptly analyzed by flow cytometry. Exudate neutrophils were also washed with PBS with 1% FBS and 0.1% Azide, and suspended in 2% paraformaldehyde in the same manner. The median intensity of fluorescence for each receptor, which is directly proportional to the density of surface receptors per cell, was recorded for each PMN population after at least 5,000 counts (Fig. 1). Median intensity of fluorescence was chosen over mean because the fluorescence distributions were often skewed. Cells were plotted on forward scatter versus side scatter; a gate was used to isolate the neutrophils from other cellular elements in whole blood. Monoclonal isotype controls included FITC-conjugated mouse IgG₁ (C5aR), phycoerythrin (PE)-conjugated IgG_2b (IL-8 RA), and PE-conjugated IgG₁ (IL-8 RB).

Figure 1. Flow cytometer analysis of chemoattractant receptor expression. Each graph is a histogram with intensity of fluorescence as a log scale on the x axis; the left peak in dark gray represents the isotype negative control; the right black peak represents control circulating polymorphonuclear neutrophils (PMNs); and the gray peak represents exudate PMNs (left) or septic PMNs (right side of the figure). Histograms for interleukin (IL)-8 RA, IL-8 RB, and C5a receptor (C5aR) expression are shown from top to bottom.

Flow Cytometer Calibration

The FACScan was calibrated biweekly to ensure no overlap of FL1 and FL2 spectra using Calibrite beads (Becton Dickinson, Mississauga, Canada). In addition, QC3 microbeads (Becton Dickinson) were used to standardize the intensity of fluorescence on a biweekly basis or sooner if instrument maintenance was performed, thus correcting for any variations in flow cytometer performance. By making small alterations in the cytometer settings, the cytometer was calibrated such that the FACScan read the same FITC and PE target channels for the QC3 beads from week to week.

In addition to FACScan calibration, standardized PE-conjugated beads were used to convert from intensity of fluorescence to number of receptors per cell. Quantibrite beads (Becton Dickinson) with known numbers of PE molecules per bead have been previously described and validated. ^25,26 A mixture of four precalibrated beads was used to create the calibration curve. Antibodies to the IL-8 receptor used in this experiment had one PE molecule per antibody. By verifying that the receptors on both circulating and exudate PMNs were saturated with antibody (doubling the concentration of antibody did not alter fluorescence), the number of antibodies bound to the cells were assumed to be identical to the number of receptors per cell. Thus, a standardized curve relating intensity of fluorescence to number of receptors per cell was used to calculate IL-8 receptors on neutrophils; the correlation coefficient for the linear regression performed on the calibration curve was 0.99993 (P = .00015). The C5a receptor was marked with an FITC-labeled antibody; thus, it was not possible to convert from intensity of fluorescence to number of C5a receptors per PMN.

Chemotaxis Assay

The chemotaxis assay was carried out using a modified Boyden’s chemotaxis assay, originally described by Boyden. ²⁷ Briefly, neutrophils were placed in the top wells of a Boyden chamber (Neuro Probe, Gaithersburg, MD). The chemoattractants, C5a, IL-8, Zymosan-activated human serum (ZAS; positive control), or PBS (negative control), were placed in the bottom wells. The two wells were separated by a 25 × 80-mm, 3-μm pore size polyvinylpyrrolidone-free polycarbonate membrane (Neuro Probe). After a 55-minute incubation period, the membrane was removed, stained using Hema 3 staining solutions for leukocytes (Biochemical Sciences, Swedesboro, NJ), and examined under a microscope at 400× magnification. Three microscopic fields were observed for each well, and average cell counts per field were calculated to yield chemotaxis values. Samples were always evaluated in triplicate. A dose-response curve was performed for IL-8 and C5a to establish the appropriate chemoattractant concentrations for use in the study.

Assay of Gene Expression

RNA isolation was performed on PMN suspensions; 1 to 3 million PMNs were collected and suspended in 1 mL PBS. After sedimentation (400 g at 4°C for 5 minutes), cells were lysed in 1 mL TRIzol (Life Technologies, Burlington, Ontario) and total RNA was collected as described by the manufacturer’s protocol. The RNA pellet was then suspended in 20 μL DEPC-treated double-distilled water.

Reverse transcription (RT) was performed on 2 μg total RNA using the ThermoScript RT-PCR System (Life Technologies) and the manufacturer’s oligo dT primer. The manufacturer’s suggested conditions were adhered to. A total reaction volume of 20 μL was used.

Selective gene amplification from the cDNA was achieved by performing polymerase chain reaction (PCR) using Platinum Taq DNA Polymerase (Life Technologies). The PCR conditions were as follows: GAPDH required 0.8 μL RT product, annealing temperature 60°C and 30 cycles; CXC R1 required 2.4 μL RT product, annealing temperature 55°C and 35 cycles; CXC R2 required 1.6 μL RT product, annealing temperature 53°C and 30 cycles; C5aR required 0.5 μL RT product, annealing temperature 55°C and 33 cycles. The specific oligonucleotide primers used and amplified fragment lengths were GAPDH, 5′ ACC ACC ATG GAG AAG GCT GG 3′ and 5′CTC AGT GTA GCC CAG GAT GC 3′ (527 bp) ²⁸; CXC R1, 5′ CAG ATC CAC AGA TGT GGG AT 3′ and 5′ TCC AGC CAT TCA CCT TGG AG 3′ (296 bp) ²⁹; CXC R2, 5′ CTT TTC TAC TAG ATG CCG C 3′ and 5′GAA GAA GAG CCA ACA AAG G 3′ (966 bp) ²⁹; and C5aR, 5′ATG AAC TCC TTC AAT TAT ACC 3′ and 5′TGG TGG AAA GTA CTC CTC CCG 3′ (551 bp). ³⁰

After amplification, 10 μL PCR product was electrophoresed on an ethidium bromide-stained agarose gel and imaged with the Alpha Imager 2000 (Alpha Innotech Corporation, San Leandro, CA). Band density was determined with the ChemiImager Software (Alpha Innotech). Analysis was performed by comparing the proband genes as ratios to the housekeeping gene GAPDH.

Statistical Analysis

All results are expressed as mean ± standard deviation. SYSTAT 8.0 for Windows was used for all statistical analysis. Statistical significance (P < .05) was evaluated with the Student t test or analysis of variance between groups, with Bonferroni correction when multiple tests were performed simultaneously. Statistical significance is indicated on the figures to follow. If the number of subjects used for a figure differs from the number listed below, it is stated in the legend for the figure.

RESULTS

All subjects were prospectively enrolled from January 2000 to May 2000. Twelve intensive care unit patients were identified and consent was obtained (APACHE II range 11–37, mean 23.6 ± 7.8). Patient infections were varied, including pneumonia, intraabdominal abscess with peritonitis, infected acute pancreatitis, pyelonephritis, endocarditis, and infected aortic prosthesis. Nineteen human healthy volunteers were studied during the same period.

Exudate Polymorphonuclear Neutrophils

Analysis of chemoattractant receptor expression in healthy controls showed that exudate neutrophils (see Figs. 1 and 2) display significantly greater expression of C5a receptors compared with control circulating PMNs of the same healthy controls (results expressed as median intensity fluorescence [MIF]: 31.5 ± 15.9 in exudate PMNs vs. 16.8 ± 7.0 in circulating PMNs;P < .001). In contrast to increased C5aR, exudate PMNs have decreased expression in both IL-8 RA and IL-8 RB compared with circulating PMNs (IL-8 RA: 12,200 ± 7,400 receptors per cell in exudate PMNs vs. 52,800 ± 6,300 in circulating PMNs; IL-8 RB: 7,200 ± 3,900 in exudate PMNs vs. 44,800 ± 3,100 in circulating PMNs; both P < .001). Evaluation of chemotaxis in exudate PMNs (see Fig. 2) from healthy controls showed increased migration to C5a and ZAS, both significantly higher than that of circulating PMNs. Exudate PMN migration to C5a was 40.86 ± 12.82 PMNs per microscopic field versus 32.17 ± 8.86 for circulating neutrophils (P = .028). Similarly, ZAS, a source of C5a, induced an exudate PMN chemotaxis of 50.84 ± 19.43 PMN per microscopic field, which was significantly higher than with circulating PMNs (34.90 ± 8.53;P = .004). Exudate PMNs displayed reduced chemotactic migration to IL-8 compared with circulating PMNs in the same individuals (29.5 ± 4.9 PMN per microscopic field in circulating PMNs vs. 16.48 ± 8.42 in exudate PMNs;P < .001). Chemotaxis to the negative control (PBS) was similar in exudate and circulating neutrophils (circulating PMNs, 5.28 ± 2.41 PMN per microscopic field vs. exudate PMNs, 6.3 ± 3.2;P = .3). Semiquantitative gene expression analysis revealed significant differences between IL-8 mRNA expression but similar C5aR mRNA expression in circulating versus exudate PMNs (Table 1).

Figure 2. Exudate polymorphonuclear neutrophils (PMNs): C5a and interleukin (IL)-8 receptors (left graphs) and chemotaxis (right graphs) in exudate PMNs compared with control circulating PMNs. Results expressed as mean ± standard deviation. MIF, median intensity fluorescence. †P < .001; ‡P < .05 vs. circulating PMNs.

Table 1. C5a AND INTERLEUKIN (IL)-8 RECEPTOR mRNA IN CONTROL CIRCULATING, EXUDATE, AND SEPTIC POLYMORPHONUCLEAR NEUTROPHILS (PMN)

* NS, all vs. control circulating PMN;

†P < .001;

‡P < .07;

§P = .06.

Septic Polymorphonuclear Neutrophils

Septic PMNs from 12 septic patients showed a significantly decreased expression of all three chemoattractant receptors evaluated (see Figs. 1 and 3), including C5aR (MIF: 7.2 ± 3.6 in septic PMNs vs. 16.8 ± 7.0 in circulating PMNs;P < .001), IL-8 RA (number of receptors: 36,200 ± 13,800 in septic PMNs vs. 52,800 ± 6,300 receptors in circulating PMNs;P < .001) and IL-8 RB (21,500 ± 11,100 in septic PMNs vs. 44,800 ± 3,100 in circulating PMNs;P < .001). Evaluation of chemotactic function in circulating PMNs in 12 septic patients revealed that chemotaxis to C5a was significantly reduced in septic PMNs (see Fig. 3) compared with the control cohort (19.7 ± 4.0 PMNs per microscopic field in septic PMNs vs. 32.2 ± 8.9 in circulating PMNs;P < .001). The observed response to the positive control, ZAS, paralleled the alterations in C5a-mediated chemotaxis (23.87 ± 6.58 PMNs per microscopic field in septic PMNs vs. 34.89 ± 8.53 in circulating PMNs;P = .001). There was no alteration in chemotaxis to IL-8 observed in septic PMNs (27.8 ± 6.1 PMNs per microscopic field in septic PMNs vs. 29.5 ± 4.9 in circulating PMNs;P = .42). Chemotaxis to the negative control (PBS) in septic PMNs was similar to circulating neutrophils (septic PMNs, 8.23 ± 5.99 PMNs per microscopic field vs. control circulating PMNs, 5.28 ± 2.41;P = .08). Semiquantitative gene expression analysis revealed significant differences between IL-8 RA mRNA expression, slightly reduced IL-8 RB gene expression, and similar C5aR mRNA expression in circulating versus septic PMNs (see Table 1).

Figure 3. Septic polymorphonuclear neutrophils (PMNs): C5a and interleukin (IL)-8 receptors (left graphs) and chemotaxis (right graphs) in septic PMNs compared with control circulating PMNs. Results expressed as mean ± standard deviation. MIF, median intensity fluorescence. †P < .001; *NS vs. circulating PMNs.

DISCUSSION

Delivery of PMN to the inflammatory exudate microenvironment is essential for effective host response to infection. To investigate the importance of in vivo alteration of chemoattractant receptors in determining PMN chemotactic function, as well as to evaluate potential mechanisms underlying the observed reduction in PMN delivery to skin blisters in septic patients, we studied the alteration from control circulating PMNs in healthy subjects in chemoattractant receptor expression and chemotactic function in two distinct PMN populations: exudate neutrophils in healthy controls and circulating PMNs in septic patients. Both exudate PMNs (which have undergone transmigration from the intravascular to the interstitial environment and have been exposed to inflammatory mediators in the exudate milieu) and septic PMNs (which have been exposed to increased levels of various circulating mediator mediators, cells, and cytokines within the serum of septic patients) represent functionally separate populations of PMNs. They have undergone physiologic alteration from “resting” circulating PMNs in the bloodstream of healthy individuals. Comparison of both healthy exudate PMNs and septic circulating PMNs with control healthy circulating PMNs reveals that receptor alteration may be a physiologic mechanism occurring in vivo that leads to change in chemotactic function. In addition, the results have specific implications regarding PMN delivery in patients with sepsis and PMN chemotaxis in the exudate environment.

In this experiment, exudate PMNs of septic patients were not investigated. First, the skin blister method is associated with some discomfort, and critically ill patients’ families are often reluctant to grant consent. Second, our previous data had shown similar receptor alterations in exudate PMNs, regardless of whether the subject was septic or healthy. For example, in septic patients, we observed a similar loss of expression of both IL-8 RA (69% reduction;P < .00001) and IL-8 RB (72% reduction;P = .00007) in the exudate PMNs; however, as opposed to the current study, chemotactic function was not evaluated in this previous experiment. ³¹ Given that alterations found in exudate PMNs appear to be consistent in septic patients and controls, and wishing to minimize patient discomfort, we elected not to collect exudate PMNs in septic patients.

The first component of our study involved the evaluation of alterations in chemoattractant receptor expression and chemotaxis after the process of transmigration. Using a human in vivo model of PMN transmigration, we showed that exudate PMNs display an upregulation of both C5a receptors and chemotaxis to C5a compared with circulating PMNs; conversely, marked downregulation of both IL-8 RA and IL-8 RB and a parallel decrease in chemotaxis to IL-8 were observed in exudate PMNs (Table 2). Our observed augmentation of both C5aR and chemotaxis to C5a in exudate PMNs has not been previously reported. The increase in C5a receptors and chemotaxis and decrease in IL-8 receptors and chemotaxis in exudate PMNs suggests that C5a, a potent PMN activator, ^9,10 may be more important for neutrophil delivery and activation in the exudate environment once transmigration has occurred.

Table 2. PERCENTAGE CHANGE IN C5A AND INTERLEUKIN (IL)-8 RECEPTORS AND IN CHEMOTAXIS IN EXUDATE PMN COMPARED WITH CIRCULATING PMN IN HEALTHY CONTROLS

PMN, polymorphonuclear neutrophils.

All percentage changes are significant at P < .001, except

*P < .05.

Our observation of decreased IL-8 receptors and IL-8 chemotaxis in exudate PMNs confirms previous reports showing diminished IL-8 receptors in exudate PMNs harvested from bronchioalveolar lavage fluid in patients with chronic respiratory tract infections (chemotactic function was not evaluated), ³² and reduced chemotaxis to IL-8, FMLP, leukotriene B4, and C5a in pustule (exudate) PMNs in a single patient with relapsing bullous staphyloderma. ³³ The importance of IL-8 to the proximal component to neutrophil delivery, including upregulation of PMN integrins, leading to neutrophil transendothelial cell migration, is well established. ^34,35 The importance of IL-8 for circulating PMN chemotaxis in contrast to the downregulation of IL-8 receptors and chemotaxis in the exudate environment suggests that IL-8 may be representative of a class of PMN-specific regional chemoattractants, recruiting PMNs from the circulation to a particular region or specific area of infection or inflammation based on production of inflammatory mediators and endothelial cell activation as the first step in PMN delivery. C5a may be representative of a class of local chemoattractants, serving to attract and activate the PMNs within the inflammatory exudate microenvironment. Analogous to PMN–endothelial cell interactions guided by a stepwise progression of adhesive interactions, neutrophil delivery may also involve a stepwise progression of chemoattractant exposure.

The second component to the study included the evaluation of septic PMNs, namely circulating PMNs in patients with the presence of infection and altered host response. Because we have previously shown that neutrophil delivery to skin blister sites is reduced by 72% in septic patients with active infection and the systemic inflammatory response syndrome, ⁴ we hypothesized that neutrophil chemoattractant receptors and chemotaxis in the circulating PMNs of septic patients would be diminished. In this experiment, we found a significant reduction in C5a receptors on circulating PMNs from septic patients and a corresponding decrease in chemotaxis to C5a. There were significant reductions in both IL-8 RB (51% reduction) and IL-8 RA (31% reduction) in septic PMNs, but no alteration in IL-8-stimulated chemotaxis (Table 3). In addition, we have previously shown that both IL-8 and C5a are present in the skin window fluid of healthy subjects at concentrations much greater than serum, thus creating a chemotactic gradient for neutrophils undergoing transmigration. However, the C5a gradient is absent in septic patients secondary to elevation in serum C5a. ⁴ Thus, our data to date suggest a defect in C5a-mediated chemotaxis in sepsis resulting from both a lack of a C5a gradient and a significant loss of C5a receptors and C5a chemotactic response in circulating PMNs. These results provide a mechanism to explain the observed decrease in PMN delivery to skin window blisters in septic patients.

Table 3. PERCENTAGE CHANGE IN C5A AND INTERLEUKIN (IL)-8 RECEPTORS AND IN CHEMOTAXIS IN SEPTIC PMN COMPARED WITH CIRCULATING PMN IN HEALTHY CONTROLS

PMN, polymorphonuclear neutrophils.

All percentage changes are significant (P < .001).

In this experiment, IL-8 and C5a receptor expression and chemotaxis to IL-8 and C5a showed parallel alterations in exudate PMNs (see Table 2) and in septic PMNs (see Table 3). These data support receptor alteration as a mechanism for change in cell function in vivo, a conclusion suggested but not confirmed by previous in vitro investigations. Although receptor downregulation occurs in conjunction with a concurrent decrease in chemotactic function in vitro, ²⁰ the effects of C5a receptor antagonists, which block C5aR both in vitro and in vivo, ³⁶ and inflammatory mediators, including tumor necrosis factor-alpha and GM-CSF, may attenuate C5a-mediated chemotaxis in PMNs in vivo despite the presence of the receptor. ¹¹ In this study, both upregulation and downregulation of C5a receptors were associated with parallel alterations in C5a chemotaxis. The reduction in IL-8 receptors with no alteration in IL-8 chemotaxis observed in circulating PMNs in septic patients shows that receptor alteration alone may not be necessarily sufficient to alter PMN chemotaxis; however, chemotaxis to other IL-8 receptor (IL-8 RB) ligands, which may be decreased, was not evaluated. Taken together, these observations support the hypothesis that chemoattractant receptor alteration is a physiologic mechanism by which chemotactic function is altered in vivo.

The biologic mechanisms leading to receptor alteration in human neutrophils have undergone extensive investigation. In this experiment, semiquantitative analysis of gene expression showed moderate downregulation of IL-8 receptor, but not C5a receptor gene expression was noted in parallel with receptor alteration. Despite the inherent limitations of semiquantitative mRNA isolation and densitometry using ratios of proband to housekeeping genes, these data support previous observations suggesting that decreased receptor gene expression mediates alterations in IL-8 receptors. ²⁰ In addition, other mechanisms result in IL-8 receptor loss in vitro, including incubation with IL-8 itself, ^21,37 tumor necrosis factor alpha, and lipopolysaccharide, ³⁸ whereas G-CSF will stimulate IL-8 receptor expression of human PMNs. ²⁰ Other factors, including hypoxia and/or hypoxia with reoxygenation, will affect matrix protein regulation of PMN IL-8 receptors. ³⁹ C5a is known to mediate a decrease in C5aR. When compared with controls, C5a concentration is elevated in the serum of septic patients; the C5a concentration is also elevated in the exudate fluid compared with serum levels in healthy controls. ⁴ Thus, multiple factors are likely involved in the in vivo regulation of cell surface chemoattractant receptor expression.

By evaluating the change in receptor expression and chemotactic function in exudate PMNs (skin window skin blister method in healthy controls) and septic PMNs (circulating PMNs in septic patients), and comparing both with control circulating PMNs (from healthy controls), we have shown the following: (1) chemoattractant receptor alteration is specific, dependent on the receptor in question, and variable capable of both significant increase and decrease; (2) receptor alteration is associated with a concurrent parallel change in chemotaxis, while the receptor may not be necessarily sufficient to alter PMN chemotaxis; (3) alteration in gene expression may explain the alteration in IL-8 but not C5a receptors; (4) exudate PMN chemotaxis relies more on C5a than IL-8; and (5) diminished chemoattractant receptors and chemotaxis in septic PMNs may explain decreased PMN delivery in these patients. These data support the conclusion that receptor alteration is a principal means by which PMN chemotaxis is regulated in humans in vivo.

Acknowledgments

The authors thank Francine Nole, RN, Mary Bouldadakis, and Kathy Patterson.