Changes in monocytic expression of aminopeptidase N/CD13 after major trauma

Abstract

HLA-DR expression on monocytes as marker for monocytic function is severely depressed after major trauma. The membrane enzyme aminopeptidase N/CD13 can trigger help in antigen processing by MHC class II molecules of antigen-presenting cells. We determined the simultaneous expression of HLA-DR and CD13 on peripheral blood monocytes of patients with major trauma (injury severity score of ≥16). 1 : 1 conjugates of phycoerythrin (PE)-to-monoclonal antibody were used in combination with QuantiBRITE^TM PE beads for a standardized quantification in terms of antibodies bound per cell (ABC). The very low expression of HLA-DR antigen on monocytes of patients at day 1 after major trauma confirmed previous results in the literature. Monocytic HLA-DR expression increased slowly to reach values in the lower range of healthy volunteers at day 14. Monocytic CD13 expression at day 1 showed values in the range of healthy volunteers, and a strong rise afterwards. Fourteen days after trauma, the monocytic expression of CD13 was still much higher than in the control group. Because lipopolysaccharide (LPS) and the anti-inflammatory cytokine interleukin (IL)-10 have been shown to be involved in the depressed HLA-DR expression on monocytes in trauma patients, we studied the in vitro effects of LPS and interleukin (IL)-10 on the expression of CD13 on monocytes prepared from the peripheral blood of healthy volunteers. Whereas a 3-day IL-10 treatment resulted in a down-regulation of both HLA-DR and CD13 expression on monocytes, LPS caused a down-regulation of HLA-DR but a rapid up-regulation of CD13 levels. Therefore we suggest that, with respect to monocytic CD13 expression, LPS rather than IL-10 could well be the explanation for monocytic surface molecules after severe injury, although other mediators with a CD13 regulating function have to be considered.

INTRODUCTION

Infectious complications remain a serious problem after critical injury, which may be related to disturbed cellular immune function after trauma [1]. Patients with temporary immunodeficiency are characterized by a depressed expression of the major histocompatibility class II antigens on monocytes, which plays a critical role in the induction of the cellular immune response to foreign antigen [2]. The magnitude of monocytic human leucocyte antigen-DR (HLA-DR) expression has been shown to correlate with clinical outcome of patients [3,4].

Whereas the behaviour of HLA-DR following injury has been studied extensively, the time–course of other monocytic markers following trauma has not been characterized similarly well. Aminopeptidase N (APN)/CD13 is a zinc-dependent ectoenzyme expressed on a wide variety of cells (for review, see [5]). The enzyme cleaves preferentially neutral amino acids from the unsubstituted N-terminus of oligopeptides. With respect to haematopoietic cells, CD13 has been considered specific for the myeloid lineage, because neutrophils, monocytes and myeloid dendritic cells but not lymphocytes of peripheral blood show a surface expression of CD13 antigen. CD13 expression seems to be regulated by a variety of different mechanisms. In addition to its developmentally regulated expression, CD13 is affected by cell growth and differentiation, e.g. it has been described that differentiation of the myeloid precursor cell line HL60 into monocytic cells leads to an increase in CD13 expression [6]. CD13 has been shown to be up-regulated by exposure of monocytes to lipopolysaccharide (LPS) [7]. T cell-derived cytokines, such as interleukin (IL)-4 and interferon (IFN)-γ, can up-regulate CD13 protein and mRNA expression in a variety of cells, among them monocytes [7,8].

The function of the transmembrane molecule varies depending on its location. Previous studies have shown that CD13 participates in antigen processing and presentation, trimming peptides protruding out of MHC class II molecules [9,10]. Furthermore, CD13 is involved in the scavenging of amino acids and catabolism of regulatory peptides, such as vasoactive and neuroactive peptides [11], and in tumour invasion and metastasis [12].

In this study, we investigated the monocytic expression of CD13 in comparison to HLA-DR during the first 14 days after major trauma. CD13 levels on neutrophils were studied in parallel. Trauma patients were compared with a group of age-matched healthy volunteers. We used the QuantiBRITE^TM flow cytometry system, which yields an absolute antigen expression value (antibodies bound per cell) and may be useful in standardizing surface antigen expression analysis. This system makes use of a highly purified phycoerythrin (PE)-labelled antibody with a 1 : 1 fluorochrome to protein (F : P) ratio, and multi-level calibrated beads with known absolute PE fluorescence.

As a next step, we studied the in vitro effects of LPS and IL-10, two mediators known to down-regulate monocytic HLA-DR expression, on the expression of CD13 on monocytes prepared from the peripheral blood of healthy volunteers.

MATERIALS AND METHODS

Patients and controls

Patients who fulfilled clinical criteria of severe trauma (n = 30) were prospectively enrolled in the study, which was performed with approval from the Ethics Committee of the University Witten/Herdecke, Germany. Informed consent was obtained initially from a guardian or relative and as soon as appropriate from the patient.

Thirty-three patients admitted consecutively to the intensive care unit (ICU) of the Hospital BG Kliniken Bergmannstrost, Halle (Germany) were recruited in 2001 and 2002. Inclusion criteria were major trauma after accident [injury severity score (ISS) ≥16], admission within 24 h after trauma, age between 18 and 80 years and a minimum stay of 7 days at the ICU. Three of 33 patients did not meet the inclusion criteria (due to death or leaving ICU early). The APACHE II score or the sepsis score according to Elebute and Stoner [13] were calculated each day.

On days 1, 3, 5, 7 and 14, both HLA-DR antigen and CD13 expression were measured in fresh EDTA-treated venous blood after lysis of erythrocytes. Fifteen healthy individuals with a mean age of 32·8 ± 2·8 years (range 23–67 years) served as controls.

Antibodies and sample preparation

The expression of HLA-DR and CD13 on CD14⁺ monocytes was evaluated by FACS analysis with antibodies labelled on a protein/fluorophore ratio of 1/1 (QuantiBRITE^TM reagents; BD Biosciences, Heidelberg, Germany). The measurement of multi-level calibrated QuantiBRITE^TM fluorescent beads enables the construction of a standard curve for antigen quantification. Using a Microsoft Excel^TM-spreadsheet referring to CellQuest^TM, provided by BD Biosciences, the measured sample fluorescence can be converted into the term ‘antibody molecules bound per cell’ (ABC).

The anti-HLA-DR (clone L243) PE/anti-CD14 PerCP-Cy5·5 antibody (catalogue no. 340827; BD Biosciences) was used according to the manufacturer's protocol. Briefly, 50 µl blood sample was incubated with 20 µl antibody at room temperature for 30 min, treated with FACS lysing solution (BD Biosciences) for lysis of erythrocytes and fixation of leucocytes and measured in 1 ml phosphate buffered saline (PBS) without any further washing steps. The CD13 antigen was detected using antibody Leu-M7 at 2·5 µg/ml final concentration with a protein/PE ratio of 1/1 (catalogue no. 347837, prepared as a customer service, BD Biosciences). For gating strategy, a CD14-specific FITC-labelled antibody (Beckmann Coulter GmbH, Krefeld, Germany) was additionally included. Incubation of blood cells with the CD13 antibody was done similarly as with HLA-DR, with the exception of two washing steps after lysis of erythrocytes.

All samples were analysed on a FACS Calibur^TM using the software CellQuest^TM (BD Biosciences). The monocytes and granulocytes were separated on the basis of their forward scatter and side scatter patterns, and the staining with anti-CD14 was used to check the identification of the monocytes. At least 2000 monocytes were analysed per sample. Additionally, a minimum of 10 000 events in the granulocyte gate was counted. The PE fluorescence within the gates was measured as specific geometric mean fluorescence intensity of the whole population of cells and converted into the term ABC with help of the Microsoft Excel^TM-spreadsheet.

In vitro culture

Peripheral blood mononuclear cells (PBMC) were isolated from the heparinized blood of healthy volunteers by standard density gradient centrifugation. Briefly, PBMC were enriched by diluting blood 1/1 with sterile PBS, pH 7·2, and then layered onto 15 ml of endotoxin tested Ficoll-Paque^TM PLUS (Amersham Biosciences, Sweden). Cells were centrifuged at 18°C for 20 min without braking at ∼ 2000 r.p.m. (800 g). Interface layer cells were pooled into larger tubes, washed twice with fresh RPMI-1640 and centrifuged at 4°C at 1400 r.p.m. (400 g) for 10 min. Cells were then resuspended in RPMI-1640 containing 10% fetal calf serum (Biochrom KG, Berlin, Germany) at a density of 5 × 10⁶ cells/ml in specially coated six-well plates to prevent the firm adhesion of monocytes (ultra-low cluster, Costar 3471, Corning, NY, USA).

Cells were treated for varying times with the following mediators: LPS from Salmonella minnesota (catalogue no. L-6261, Sigma, Deisenhofen, Germany) at 40 ng/ml; or IL-10 at 10 ng/ml (specific activity >5 × 10⁶ U/mg; Strathmann Biotech GmbH Hamburg, Germany). Cells were detached from six-well plates by rinsing with PBS, pH 7·2. After a washing step with PBS, cells were stained with the CD13 (clone Leu-M7)-, or the HLA-DR-specific antibody (clone L243), or a PE-labelled isotype control for IgG1 and IgG2a (all antibodies from BD Biosciences) for 15 min at room temperature. After fixation with 1% paraformaldehyde, cells were washed with PBS; 3000 cells were measured in a scatter gate on monocytes. The PE fluorescence within the gate was measured as mean fluorescence intensity of the whole population of cells.

Statistical analysis

Results are given as means ± s.e.m. The statistical analysis was done using the commercial software SPSS 11·0 (SPSS Inc., Munich, Germany). Differences in monocytic expression between patients and healthy volunteers of HLA-DR and CD13, and differences at each point in time were analysed using the Student's t-test (paired/unpaired) with Bonferroni correction. Because we compared the values of the same patient at different points in time, we also used the general linear model (GLM) to create a two-factor analysis with repeated measurements. Pearson's correlation was used to evaluate the strength of linear relationship. P-values < 0·05 were considered significant.

RESULTS

Patients had a mean age of 39 ± 3 years and an ISS of 31·0 ± 1·8 (mean ± s.e.m.). Eighty per cent were men. The main cause for trauma was traffic accident. Twenty-two patients had polytrauma (eight of 22 polytraumatized patients additionally with severe head trauma) and eight patients suffered from an isolated head trauma. Patient characteristics at the day of admission are presented in Table 1. The patients spent an average of 22·3 ± 1·7 days in intensive care. All patients needed mechanical ventilation (mean 17·1 ± 1·6 days). Twenty-eight of 30 patients developed a manifest or a suspected ventilator-associated infection during their stay in ICU and received antibiotics for 12 ± 0·5 days. The mean fever days for patients were 7·9 ± 0·6 (body temperature >38°C). Twelve of 30 trauma patients showed a systolic blood pressure < 90 mmHg and received fluid boluses and/or vasopressors to maintain adequate arterial blood pressure. Sixteen of 30 trauma patients had a severe head trauma with a systolic blood pressure >90 mmHg and received vasopressors to maintain adequate cerebral perfusion pressure. Mean C reactive protein (CRP) values were estimated each day and showed values between 130 and 206 mg/l with a peak at day 9. Patients’ mean white blood cell counts ranged from 11 300 to 16 100 cells/mm³ with a peak at day 11. The mean sepsis score of all patients showed values between 8·5 and 12·2 within the 14 days. Because of spinal cord injury three of 30 patients received a high single dose of corticosteroids (5 g methyl prednisolone within 24 h).

Table 1.

Patients’ score characteristics at beginning of study

Characteristics (n = 30)	Mean ± s.e.m.
Injury severity score (ISS)	31·0 ± 1·8
Abbreviated injury scale (AIS)
Head/neck	2·8 ± 0·4
Face	0·4 ± 0·2
Chest	2·6 ± 0·3
Abdomen	1·7 ± 0·3
Extremities	1·5 ± 0·3
External	0·2 ± 0·1
Apache II	31·0 ± 1·8
Sepsis score; Elebute and Stoner [13]	8·5 ± 0·6

Monocytic HLA-DR on admission (day 1) revealed a marked attenuation in comparison with age-matched healthy volunteers as a control group (9980 ± 750 ABC versus 29 880 ± 2080 ABC, P < 0·01). Figure 1 illustrates the time–course of monocytic expression of HLA-DR molecules during ICU stay. Monocytic HLA-DR expression slowly increased to reach values in the lower range of healthy volunteers at day 14. The time–course of the monocytic CD13 expression showed a different pattern (Fig. 2). Starting with values in the range of healthy volunteers (12 880 ± 1600 ABC), CD13 expression on monocytes rose rapidly. Fourteen days after trauma, the expression of CD13 on monocytes was still much higher than in normal controls (33 230 ± 4150 ABC in patients versus 12 520 ± 1540 ABC in controls, P < 0·01). When using the general linear model (GLM) we found significant differences between both HLA-DR and CD13 expression on monocytes of patients and healthy controls (P < 0·01).

Fig. 1.

HLA-DR antigen expression on CD14⁺ monocytes of trauma patients compared with healthy volunteers (hatched area). All results are expressed as antibodies bound per cell (ABC, mean ± s.e.m.). Percentiles are shown for healthy volunteers with P25 (25th percentile), P50 (50th percentile) and P75 (75th percentile). N = number of patients.

Fig. 2.

Monocytic CD13 expression of trauma patients compared with healthy volunteers (hatched area). Values are expressed as ABC (mean ± s.e.m.). Percentiles are shown for healthy volunteers with P25 (25th percentile), P50 (50th percentile) and P75 (75th percentile). N = number of patients.

The Pearson correlation coefficients (r) showed a positive correlation between CD13 and HLA-DR expression on monocytes on each day (ranging from 0·32 to 0·47). We did not observe any significant correlation between CRP values or the leucocyte count (r = −0·29 to 0·2) with either HLA-DR or CD13 expression on monocytes (HLA-DR and CRP: r =−0·27 to 0·2; HLA-DR and leucocyte count: r =−0·27 to 0·25; CD13 and CRP: r =−0·09 to 0·45; CD13 and leucocyte count: r =−0·38 to 0·08).

Since granulocytes do also express CD13, we determined their CD13 expressions in our group of patients and compared them with monocytic expression and with granulocytic expression in healthy volunteers. CD13 expression of monocytes and neutrophils on an ABC level was similar in the control group. Neutrophils of patients started with a CD13 expression lower than in healthy controls (8500 ± 680 ABC in patients versus 12 550 ± 870 ABC in controls, P < 0·01). Granulocytic CD13 levels increased more slowly to values in the upper range of healthy volunteers (18 383 ± 1941 ABC), as shown in Fig. 3.

Fig. 3.

Granulocytic CD13 expression of trauma patients compared with healthy volunteers (hatched area). Values are expressed as ABC (mean ± s.e.m.). Percentiles are shown for healthy volunteers with P25 (25th percentile), P50 (50th percentile) and P75 (75th percentile). N = number of patients.

To study the effects of known HLA-DR down-regulating mediators on monocytic CD13 expression, PBMC were exposed to LPS or IL-10 in vitro. A 24-h in vitro culture of PBMC resulted already in a strong increase of both HLA-DR (87 000–210 000 ABC) and CD13 (64 000–200 000 ABC) molecules, in parallel with an increase in cell size (scatter properties in flow cytometry). As ABC values were outside the beads-defined linear range of PE fluorescence, we used mean channel values for presentation of results. Whereas an IL-10 treatment for 1–3 days resulted in a down-regulation both of HLA-DR and CD13 expression on monocytes (Fig. 4), LPS caused a delayed down-regulation of HLA-DR molecules. Day 1 values of monocytic HLA-DR expression indicated high interindividual differences with values ranging from −15 to + 17% of the control. As expected, LPS treatment of PBMC resulted in a rapid up-regulation of monocytic CD13 levels.

Fig. 4.

Influence of IL-10 and LPS treatment on monocytic HLA-DR () and CD13 (▪) expression. PBMC were cultured in the presence of LPS (40 ng/ml) or IL-10 (10 ng/ml). The mean fluorescence intensity (MFI) after staining with a PE-labelled anti-HLA-DR or CD13 antibody is given as a percentage of the control (culture without mediator = 100%). Values are means ± s.e.m. of a minimum of four experiments.

DISCUSSION

The principal aim of this study was to elucidate the time–course of the monocytic expression of the membrane peptidase CD13 in parallel with HLA-DR expression in patients with severe injury.

Trauma patients are known to have an early-onset depression of the overall cellular immune response [1], which is associated with a high rate of infections and considerable mortality. Furthermore, early changes of the immune system caused by surgical stress contribute to postoperative complications such as sepsis and multiple organ failure. Several investigators have shown that the expression of monocyte HLA-DR is severely reduced and correlates with the outcome of patients [3,4,14]. We found a very low expression of HLA-DR antigen on monocytes within 24 h of major trauma, which conforms to previous results in trauma patients. A diminished monocytic HLA-DR expression for a period longer than 7 days has been described to be associated with increased mortality [4].

We suggest that in addition to HLA-DR, other molecules can be useful to characterize monocytic function, as has also been shown recently for the measurement of the monocytic expression of l-selectin/CD62L. Monocytic CD62L expression increased with injury severity and was highest in those patients who developed multiple organ dysfunction syndrome [15]. Our group has been interested in the membrane ectoenzyme aminopeptidase N/CD13 for several years. We described CD13 as an activation-associated marker on T lymphocytes [16]. Now we show that CD13 could be used as an activation marker of monocytes in major trauma patients. The slow rise of monocytic HLA-DR molecules at the days following ICU admission was paralleled by an obvious increase in the CD13 expression, starting from values in the range of our control group at day 1. Fourteen days after major trauma, the monocytic expression of HLA-DR was in the lower part of the values of healthy individuals, whereas CD13 expression was still much higher than in normal controls. The possible clinical and therapeutic consequences of these findings warrant further investigation. CD13 could take part in monocytic antigen processing, trimming peptides protruding out of MHC class II molecules [9,10]. Whether an increased CD13 expression during a suboptimal HLA-DR expression allows a better antigen presentation capacity of monocytes remains to be elucidated. Otherwise, CD13 can degrade biologically active peptide substrates, such as neuropeptides and kinins [17]. The chemotactic cytokine monocytic chemotactic protein (MCP)-1 is a possible CD13 substrate. Deletion of the aminoterminus converts the mediator from an activator of basophils to an eosinophil chemoattractant [18]. Last, but not least, we could show the direct involvement of monocytic CD13 in signal transduction processes recently: ligation or cross-linking of the molecule by CD13-specific antibodies results in an increase in intracellular calcium, an activation of various kinases and an augmented IL-8 mRNA synthesis [19].

In major tissue trauma, neutrophils are also important effector cells [20]. Activated neutrophils are redistributed from the peripheral blood into the tissues, where release of proteolytic enzymes and radicals participate in the development of systemic inflammation and organ dysfunction. We observed that, at day 1, granulocytes of patients started with a significantly lower CD13 expression than healthy controls. This observation might be explained as an effect of neuropeptides: enkephalins as peptide substrates of CD13 can down-regulate CD13 expression on granulocytes, possibly via the down-regulation both of the opioid receptor and the co-localized peptidase [21]. Granulocytic CD13 expression of our patients increased more slowly than in monocytes to values in the upper range of our control group. The moderate up-regulation of CD13 in comparison to monocytes could be a result of the shorter lifespan of these cells. Otherwise, CD13 could exert different functions in monocytes in comparison to neutrophils. Because neutrophils lack HLA-DR molecules, granulocytic CD13 molecules do not play a role in antigen presentation. CD13 may down-regulate inflammatory responses by co-operating with the membrane endopeptidase neprilysin/CD10 to hydrolyse N-formyl-Met-Leu-Phe, a microbial-derived chemotactic peptide stimulating neutrophil migration into sites of infection, phagocytosis, production of superoxide radicals and the release of granulocytic proteolytic enzymes [22]. The CD13 enzyme activity of granulocytes of healthy donors has been shown to vary remarkably [21]. Balog et al. reported that only granulocytes with high CD13 activity (preactivated?) release increased amounts of superoxide anion after preincubation with enkephalins [21]. During multiple sclerosis (MS) exacerbation and in the course of chronic progressive MS, granulocytes reveal several forms of preactivation, including significantly higher expression of CD11b/CD18, neprilysin/CD10 and CD13, in comparison with MS remission or neurological patients with noninflammatory diseases [23]. This study suggests CD13 to be an activation-associated molecule in granulocytes as well.

Monocytic CD13 and HLA-DR expression were determined by us quantitatively, on an antibodies-per-cell level. This system can quantify accurately the absolute antigen expression level, taking into account variations in reagent purity and instrument performance. Using the software provided by the manufacturer and calibrator beads that need to be run only once per day, this system represents a rapid and practical way to standardize expression analysis [24]. The advantage of the use of calibration systems in cell cytometry is longitudinal comparability between different laboratories [25]. Recently, Volk and co-workers published a comparable study on monocytic HLA-DR expression of patients undergoing cardiopulmonary bypass surgery [26]. Using a cut-off of 5792 HLA-DR molecules per cell, both sensitivity and negative predictive value for patients who developed microbiologically confirmed infection was 1·0. Authors suggest that a standardized immune monitoring at day 1 might be useful for early discrimination of patients at elevated risk for infections [26]. In our patients, we only observed three people with fewer than 5800 HLA-DR molecules per cell with varying infection susceptibility. In future, more patients have to be investigated to clarify the importance of this cut-off value in major trauma patients.

As the involvement of LPS and the anti-inflammatory cytokine IL-10 has been discussed in the depressed HLA-DR expression on monocytes in major trauma patients [27,28], we studied the in vitro effects of LPS and IL-10 on the expression of both HLA-DR and CD13 on monocytes prepared from the peripheral blood of healthy volunteers. Wolk et al. describe an initial up-regulation of monocytic HLA-DR expression within the first 12 h of LPS exposure, and a decrease after 24 h [28]. The decrease in monocytic HLA-DR expression that we observed seems to be delayed compared with results of Wolk et al. The discrepancy between our results may be caused by the LPS source and concentration. Otherwise, Wolk et al. agree that significant interindividual differences do exist in the time courses of LPS induced attenuation of monocytic HLA-DR molecules [28]. Our data confirm results from Koch, who as early as 1991 described CD13 as a monocytic activation-related antigen, which is up-regulated by LPS [7]. Neprilysin/CD10 expression on neutrophils has also been described as increasing within 1 h after LPS exposure [29]. IL-10 treatment of PBMC resulted in a down-regulation of both HLA-DR and CD13 expression on monocytes. Therefore we suggest that, with respect to monocytic CD13 expression, LPS rather than IL-10 could well be the explanation for monocytic surface molecules after severe injury, although other mediators with a CD13 regulating function have to be considered here. The pleiotropic cytokine IL-10 has been shown to attenuate NF-κB activation by interfering with breakdown of IκBα, resulting in an inhibition of cytokine synthesis [30]. Furthermore, IL-10 has been discussed to inhibit maturation of antigen-presenting cells, thus preserving their ability to take up antigen while, at the same time, hampering their migration to draining lymph nodes [31]. Because CD13 expression of antigen-presenting cells has been regarded as a marker of maturation [6,7], the IL-10-provoked CD13 down-regulation would be consistent with this hypothesis.

Glucocorticoid treatment for 24 h is known to down-regulate both monocytic CD14 (part of the LPS receptor) and CD13 expression [32]. Three of our 30 patients received a single bolus of glucocorticoids. Interestingly, in these patients corticosteroids did not appear to have any effect on CD13/HLA-DR expression in contrast to previous studies.

Further experiments have to deal with the question whether or not CD13 is regulated on a transcriptional level. The membrane peptidase could be translocated from intracellular storage membranes to the cell surface after contact with a hitherto unknown mediator. For example, within minutes the anaphylotoxin C5a increases the membrane expression of CD13, neprilysin/CD10, tyrosine phosphatase/CD45RO and the Fc receptor Fcγ-RIII/CD16 on granulocytes [33]. Also, monocytes responded to C5a with increases in CD13 and CD45/CD45RO expression [33]. Otherwise, transcriptional regulation is involved in several processes, such as the down-regulation of CD13 expression of myeloid leukaemia cells after direct cell-to-cell contact with bone marrow stromal cells [34].

In summary, for the first time we provide data of monocytic CD13 expression from patients with severe injury during intensive care. Future prospective studies have to confirm a possible association of LPS and the expression of CD13. Furthermore, other monocytic surface molecules, such as CD62L, have to be studied in parallel. Finally, the quantification of the monocytic CD13 expression could be a useful approach also for the evaluation of patients with other forms of a temporary compromised immune defence, such as sepsis or after cardiac surgery.