PROTEINASE 3 EXPRESSION ON NEUTROPHIL MEMBRANES FROM PATIENTS WITH INFECTIOUS DISEASE

Abstract

Proteinase-3 (PR3) is an abundant serine proteinase stored in the azurophilic granules of neutrophils and released to the cell surface upon activation where it contributes to local tissue destruction and inflammation. The sub-population of membrane PR3 (mPR3) high expression (PR3-high) varies among individuals. There are many reports about PR3 in Wegener's granulomatosis, but few about PR3 expression in patients with common inflammatory disorders, such as sepsis. The mPR3 expression on neutrophils from 56 patients with inflammatory disorders and from 64 healthy volunteers was examined by flow cytometry. High variability in the percentage of PR3-high (%PR3-high) neutrophils was observed in healthy volunteers and patients with inflammatory disease, and the %PR3-high was significantly greater in the patients (72 ± 19% vs 55 ± 20%, P < 0.0001). Overall neutrophil PR3 expression in patients with infectious diseases, especially systemic inflammatory response syndrome (SIRS) was significantly high (P < 0.01) and showed a positive correlation with C-reactive protein (CRP). Even under inflammatory conditions not involving autoimmune vasculitis, there are significant increases in both the absolute surface expression of PR3 and the numbers of neutrophils expressing high levels of PR3 and these correlate with CRP levels. The data are consistent with a model in which neutrophil membrane expression of PR3 is greatly influenced by an in vivo inflammatory environment.

INTRODUCTION

Proteinase-3 (PR3) is an abundant serine proteinase stored in the azurophilic granules of neutrophils along with neutrophil elastase and cathepsin G, and is released to the cell surface upon activation (^1-3). This 29-kD glycoprotein also is stored in secretory vesicles and specific granules and is generally thought of as a soluble enzyme, in the sense of being released into the local inflammatory milieu upon neutrophil activation and disgorgement of granular contents. However, several recent studies have observed that most of the neutrophil PR3 remains associated with the cell membrane upon activation, with very little released into the medium (⁴).

PR3 is possibly best known as the primary target antigen of the PR3 anti-neutrophil cytoplasmic antibodies (PR3-ANCA) in Wegener's granulomatosis (WG), a debilitating autoimmune disease characterized by necrotizing vasculitis (^5-7). PR3-ANCA are also found in patients with microscopic polyangiitis, a systemic vasculitic disease (⁸). PR3 has activities that include degradation of extracellular matrix proteins (⁹), regulation of myeloid differentiation (^{10, 11}), potentiation of platelet activation (¹²), and antibacterial action that is independent of its enzymatic activity (¹³). Structurally, PR3 is very similar to neutrophil elastase (¹⁴), but it does have unique substrates, including the membrane-bound precursors of the pro-inflammatory tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) cytokines (^15-17). Thus, PR3 expression near a vascular surface would very likely contribute to local tissue destruction and inflammation.

New diseases entities have recently emerged, such as systemic inflammatory response syndrome (SIRS) (¹⁸) sepsis in the emergency department and intensive care unit, which frequently. become fatal outcome. SIRS is frequently associated with elevated levels of plasma cytokine, organ failure and disseminated intravascular coagulation (^{19, 20}). Neutrophils play a major role in mediating innate inflammatory reactions and PR3 activity contributes to production of pro-inflammatory cytokines. One study shows membrane PR3 expression in normal volunteers that is stable with time, but variable in magnitude, and a relatively high expression of PR3 on neutrophils of patients with WG and rheumatoid arthritis (²¹). In SIRS, expression of PR3 may play an important role in the progress of inflammation.

In this study, we measured the expression of PR3 on neutrophil membrane from patients with infectious diseases to examine the relationship between inflammation and PR3 expression.

MATERIALS AND METHODS

Patients and controls

The patient group consisted of 56 patients with infectious diseases (34 men; 22 women; mean age, 59.0 ± 17.3 yr; range, 18 to 86 years) seen at Mie University Hospital between December 24, 2003 and February 17, 2004. Sixty-four healthy volunteers (40 men; 24 women; mean age, 26 ± 8 year; range 19 to 65 yr) were included in control group. Patients' diagnoses were as follows: 24 pneumonia, 19 peritonitis, 9 tonsillitis, 2 phlegmon and 2 pyelonephritis. None of the patients had vasculitis or ANCA titres in those patients whom we could examine (data not shown). Disseminated intravascular coagulation (DIC) was diagnosed according to the criteria established by the Japanese Ministry of Health and Welfare (²²). SIRS was diagnosed by the criteria delineated by the American College of Chest Physicians/Society of Critical Care Medicine (ACCP/SCCM) Consensus Conference Committee (¹⁹). Briefly, patients demonstrated more than 3 of 4 criteria: (1) temperature > 38° or < 36°; 2) heart rate > 90 beats/minute; 3) respiratory rate 20 breaths/minute or PaCO₂ < 32 mmHg; and 4) white blood cell count >12,000/mm³, or < 4,000/mm³, or > 10 % immature (band) forms). All patients and healthy volunteers gave informed consent before participation in the study and the study was approved by the Mie University's Review Board for human studies.

Materials

Dextran was purchased from Wako (Osaka, Japan). Hanks Balanced Salt Solution (HBSS) was from Gibco BRL (NY, USA). Trypan blue was from Acros Organics (New Jersey, USA). Recombinant TNF-α was from Peprotech EC LTD (London, UK). N-formyl-L-methionyl-L-phenylalanine (FMLP) was from Sigma-Aldrich Co. (St. Louis, USA).

Monoclonal antibody 1549 recognizing human PR3 was obtained from hybridoma supernatants using mice immunized with PR3 purified from human neutrophils (Athens Research, GA). Supernatants were screened for binding to neutrophil-derived PR3 immobilized in wells of 96 well plates and bound antibody was detected with horseradish peroxidase-conjugated goat anti-mouse IgG and substrate. The 1549mAb does not recognize neutrophil elastase (data not shown). For production, hybridoma cells were cultured in bags (I-MAB Monoclonal Antibody Production Kit, Diagnostic Chemicals Ltd) in RPMI-1640, 10% bovine calf serum, 50 U/ml penicillin, 50 μg/ml streptomycin, 0.35 mg/ml L-glutamine. Antibodies were purified from conditioned media using Mep Hypercel resin (Life Technologies, Inc) equilibrated in 50 mM Tris HCl, ph 7.5, 0.1 M NaCl, 0.02% sodium azide and eluted with 50 mM sodium acetate, pH 4.0. Antibody was labeled for flow cytometry studies with the Alexa488-succinimidyl ester conjugate (Molecular Probes, Eugene, OR) using standard methods and free probe was removed with extensive dialysis.

Neutrophil isolation

For experiments using purified neutrophils, 3 mL of EDTA-anticoagulated whole blood was mixed with a half volume of 2% dextran in 0.9% NaCl. RBCs were allowed to sediment for 45-60 min at room temperature. The supernatant was removed and centrifuged at 400 ×g (5 min, 4°C). The pellet was re-suspended in 5 ml of ice-cold 0.2% NaCl with mixing for 25 s, followed by 5 ml of ice-cold 1.6% NaCl. The cells were centrifuged and the pellet gently re-suspended in 5 ml of HBSS with 3mM calcium chloride containing 1% bovine serum albumin (HBSS+ buffer). This was transferred to a 15 ml conical centrifuge tube and underlaid with 5 ml of Lymphocyte Separation Media (density 1.077 ± 0.001 g/ml at 22°C; Cellgro, Herndon, VA). The sample was centrifuged at 400 ×g for 30 min at 4°C. The pellet of purified neutrophils was re-suspended in 5-10 ml HBSS+ buffer. Cell viability was determined for every cell preparation by trypan blue exclusion and was consistently greater than 99%.

In vitro stimulation and measurement of membrane PR3 expression by flow cytometry

Isolated neutrophils were re-suspended in HBSS+ buffer at a final concentration of 2 × 10⁶/ml. The cell suspension was incubated with TNF-α (10 ng/ml) and FMLP (1.0 μM) for 20 min at 37° and washed with 1 ml ice-cold HBSS+ buffer. Cells were pelleted and re-suspended in HBSS+ buffer before incubation with labeled antibodies. Data acquisition by flow cytometry and subsequent analysis was done with a FACScan using CellQuest software (Becton Dickinson, Heidelberg, FRG). Neutrophils were gated according to relative size (forward scatter) and relative granularity (side scatter) properties. Data is reported either as the percentage of mPR3 high expression (%PR3-high) neutrophil population, or the level of mPR3 expression is reported as mean fluorescence intensity (MFI) after correction for non-specific binding (NSB) as assessed using an irrelevant isotype-matched antibody and appropriate conjugate. Data are expressed as: (MFI-NSB) × %PR3-high to take into account differences in surface PR3 expression between individuals (²³).

Measurement of plasma concentrations of granulocyte elastase derived fibrin degradation products (GE-XDP), D-dimer, soluble fibrin (SF) and thrombomodulin (TM).

GE-XDP was measured by latex-agglutination assay using monoclonal antibody IF-123. IF-123 specifically recognizes elastase-digests of human fibrinogen and fibrin, but not their plasmin-digests. The epitope for this antibody is located at Aα Leu-196 to IIe-204. D-dimer, SF and sTM were determined with D-dimer test Kokusai-F (Kokusai-Shiyaku), Enzymum test (Boehringer Mannheim, Mannheim, Germany) and sTM-test Kokusai (Kokusai-Shiyaku), respectively.

Statistical analysis

Differences in continuous variables between two groups were analyzed by means of Mann Whitney U test. Differences in continuous variables between more than three groups were analyzed by means of Steel-Dwass test. Correlations were analyzed by the nonparametric Spearman correlation coefficients. A two-sided P < 0.05 was considered to be statistically significant.

RESULTS

mPR3 expression on activated neutrophils from healthy volunteers and septic patients

Similar to other studies, we observed PR3 antigen on the surface of purified neutrophils and a sub-population of neutrophils expressing higher membrane PR3 antigen appears after activation (Fig. 1). However, the distribution of PR3 antigen after neutrophil activation with TNF-α and FMLP varied considerably between individuals. For example, in cells from four healthy volunteers, the percentage of cells expressing high levels of membrane PR3 (%PR3-high) varied from 0.19% to 84.36% (Fig. 1A-D), and was not related to the phenotype before activation. This contrasts with results of a previous study in which the distribution of PR3 antigen between sub-populations remained constant regardless of stimulation (²¹). This latter study activated neutrophils with cytochalasin B and FMLP, conditions known to mobilize azurophilic granules, but also to disrupt the cytoskeleton (²⁴), which may result in aberrant re-distribution of surface antigens.

Fig. 1: mPR3 expression on neutrophils from four healthy volunteers as detected by flow cytometry. The isotype control (dashed line) showed no significant membrane staining. Purified neutrophils were stained for mPR3 with (filled) or without (bold line) stimulation by TNF-α and FMLP. The percentage of cells expressing high levels of mPR3 are shown as %PR3-high.

Similar to the data from healthy volunteers (N = 64), this variability in PR3 distribution after in vitro activation with TNF-α and FMLP also was observed in patients with acute infectious disease (N = 56) (Fig. 2). However, the percentage of cells in the PR3-high population was significantly higher in the patients than in healthy volunteers (72 ± 19% vs 55 ± 20%, P < 0.0001). In the healthy subjects population, the percentage of PR3-high expressing cells was distributed normally (bell-shaped curve), whereas the distribution of %PR3-high cells in patients with acute infectious disease was skewed to a higher phenotype.

Fig. 2: Histograms of PR3 high expression neutrophils population. (A) In a cohort of 64 healthy subjects, the percentage of PR3 high expression neutrophils (%PR3-high) was evenly distributed. (B) In 56 patients with infectious disease, %PR3-high was skewed to a higher phenotype.

In some cases, there was a significant increase in the percentage of cells in the PR3-high population over time both with and without in vitro activation with TNF/FMLP. Two representative cases are shown in Figure 3. At the onset of sepsis (Day 0), the %PR3-high was higher than at day -14, and this paralleled increases in C-reactive protein (CRP), an acute phase protein and biomarker of inflammation. Though both of the patients seemed to recover from SIRS, they eventually relapsed into SIRS again and died within four weeks.

Fig. 3: Changes in mPR3 expression with time in patients with sepsis. PR3 expression on neutrophils from sepsis patients was evaluated by flow cytometry with (filled) and without (bold line) stimulation by TNF-α and FMLP. The isotype control (dashed line) showed no significant membrane staining. (A) mPR3 expression in a patient with small cell lung cancer complicated by bacterial pneumonia. (B) mPR3 expression in a patient with cancer of the hypopharynx complicated by bacterial pneumonia. In both cases, mPR3 expression was bimodal before the onset of sepsis (left). At the time of infection, mPR3 expression increased and the %PR3-high also was greater (middle). After treatment for sepsis, mPR3 expression was lower than at the onset of infection whereas the %PR3-high was slower to resolve (right).

As shown in Figure 4, overall membrane expression of PR3 on resting neutrophils was higher in the patients than in the healthy volunteers (913 ± 160 vs 393 ± 75; P = 0.0003). After in vitro activation by TNF-α and FMLP, expression of PR3 remained significantly higher on neutrophils from patients with inflammatory disease (5242 ± 426 vs 3667 ± 547; P = 0.00003).

Fig. 4: Neutrophil PR3 expression in healthy volunteers (N = 64) and patients with infectious diseases (N = 56). Neutrophil PR3 expression is shown as the product of surface expression levels (MFI-NSB) and relative abundance (%PR3-high) because both parameters change as a consequence of inflammation. Expression levels were significantly higher in patients than in healthy volunteers with (A) and without (B) in vitro stimulation, (P = 0.0003 and P = 0.00003, respectively).

mPR3 expression and plasma levels of CRP

The mPR3 expression on resting neutrophils and after activation had a slightly correlation with CRP levels (r = 0.465, P < 0.001 and r = 0.442, P < 0.001, respectively).

It may be important to mention that we did see a few cases that did not follow the general trend described above. This is most likely due to the fact that PR3 expression is governed by multiple complex factors, including genetics, and the inflammatory environment is only one of many contributing factors.

mPR3 expression and SIRS

Each patient was reviewed daily and cardio-respiratory observations and daily blood tests were noted to determine the incidence of SIRS. Consequently 14 patients were diagnosed with SIRS on the same day when the blood samples were taken for measurement of membrane PR3 expression by flow cytometry. PR3 expression [(MFI-NSB) × %PR3-high] was significantly greater in the patients with SIRS (N = 14) than non-SIRS (N = 42) on either resting (478 ± 135 vs 1813 ± 361, P < 0.001) or activated neutrophils (4170 ± 493 vs 8862 ± 1060, P < 0.001), respectively (Fig. 5). We observed a trend that PR3 expression is higher in patients with DIC compared to those without DIC. CRP and D-dimer were significantly higher in larger PR3 than in less PR3, but GE-XDP and TM were not (Table 1).

Fig. 5: Neutrophil PR3 expression in SIRS (N = 14) and non-SIRS (N = 42) patients. Neutrophil PR3 expression is shown as the product of surface expression levels (MFI-NSB) and relative abundance (%PR3-high) without in vitro stimulation (A) or with stimulation (B). Membrane PR3 expression without in vitro stimulation was significantly greater in the patients with SIRS (478 ± 135 vs 1813 ± 361, P = 0.0007). After stimulation, PR3 was also significantly greater in the patients with SIRS (4170 ± 493 vs 8862 ± 1060, P = 0.0002).

Table 1: Clinical and Laboratory Data in Infectious Patients

DISCUSSION

Previous reports show that both the %PR3-high and level of mPR3 expression are related to disease and relapse in Wegener's granulomatosis patients with PR3-ANCA-associated vasculitis (^{25, 26}). The present study newly shows that, even in patients without vasculitis, but having systemic inflammatory conditions, both the %PR3-high and level of mPR3 expression are increased. Some patients showed a significant increase in %PR3-high with time and clinical severity. The patients with increasing %-PR3-high died from complications due to sepsis, suggesting a correlation with the severity of inflammation (data not shown). This change in neutrophil phenotype also is probably not related to mobilization of new neutrophils from bone marrow, because there was no significant changing in %PR3-high after glucagon injection in healthy volunteers (²⁷). In addition, we find that the %PR3-high does not change even after G-CSF injection (unpublished observations). Thus, the changes in %PR3-high more likely occur in the peripheral circulation as a consequence of an inflammatory environment. We have made many efforts to recreate this phenomenon in vitro, but with limited success, so there are as yet unidentified contributors to changes in %PR3-high observed in patients.

The current observation that PR3 antigen is expressed on the surface of resting, purified neutrophils from patients and healthy volunteers is consistent with previous studies (²⁸). The sub-populations of cells expressing considerable PR3 (PR3-high) or those with less PR3 (PR3-low) varied among individuals. In an earlier study, this expression pattern was found to be stable for each individual over prolonged time periods and may be genetically controlled because the %PR3-high was highly correlated in monozygonic twin pairs but there was no correlation in dizygotic twin pairs (^{27, 28}). After activation in vitro, mPR3 expression increases but the %PR3-high remains stable, again suggesting regulation of expression at the genetic level (²³). The current data in healthy volunteers is consistent with this model because the %PR3-high was constant in an individual regardless of the blood drawing time. However, in a pathophysiologic environment, the expression of mPR3 on neutrophils does differ and will change depending on the severity of the inflammatory challenge.

The present study shows that both the %PR3-high and the absolute level of mPR3 expression are increased in septic patients and related to their CRP concentration. CRP is an acute-phase protein and likely to reflect the presence as well as the severity of sepsis (²⁹). CRP is synthesized predominantly by the liver, mainly in response to interleukin-6 (IL-6) (^{30, 31}). TNF-α and IL-1β also are regulatory mediators of CRP synthesis (³⁰). Actually not only TNF-α, but also IL-1β and IL-6 augment mPR3 expression on neutrophils surface in vitro (data not shown). In addition, activated neutrophils and purified PR3 augment TNF-α and IL-1β release from a stimulated human monocytic cell line (³²). It was considered that mPR3 expression was up-regulated by some cytokines in inflammatory environment and contributing to the progress of the inflammation. Thus, neutrophil PR3 expression and activity plays a role in the cytokine networks that modulate an inflammatory response. Some patients did not fall into the general relationship between mPR3 expression and CRP. It is possible that some patients who have genetically dominant PR3-low neutrophils may have difficulty in increasing membrane PR3 expression on neutrophils even with similar inflammatory environment. It is equally possible that others with low CRP and high PR3 expression may be reflecting the fact that an increase in CRP is followed by neutrophil activation.

In the current study we observed that mPR3 expression was increased in patients with SIRS to a level greater than in patients without SIRS, suggesting that the neutrophils in patients with SIRS were either partially activated or shifted to an altered state. In vitro activation studies demonstrated that PR3 is also high in patients with SIRS suggesting that the neutrophils in those patients may be more primed by the inflammatory environment. These findings indicate that patients with SIRS have acquired the ability to express higher levels of PR3. Consequently the amount of PR3 expression was larger in the SIRS patients. It is not clear whether this difference in mPR3 expression is due to greater cellular activation levels in the face of an extreme inflammatory environment or to some other mechanism, but it is clear that elevated expression of mPR3 reflects activation of neutrophils in inflammatory disease.

CONCLUSION

PR3 expression on neutrophils is higher in patients with systemic inflammatory diseases, such as sepsis, in the absence of PR3-ANCA-positive vasculitis, suggesting that membrane expression of PR3 is greatly influenced by an in vivo inflammatory environment.