Mannose-binding lectin engagement with late apoptotic and necrotic cells

Authors


Abstract

The serum opsonin mannose-binding lectin (MBL) has been shown to be involved in the handling of apoptotic cells. However, at what stage in the process this happens and whether this mediates activation of complement is unknown. Cells rendered apoptotic or necrotic were incubated with purified MBL/MBL-associated serine protease (MASP) complexes and assessed by flow cytometry and fluorescence microscopy. MBL bound specifically to late apoptotic cells, as well as to apoptotic blebs and to necrotic cells, but not to early apoptotic cells. Binding of MBL could be inhibited by EDTA as well as with an antibody against the CRD region. Addition of C1q, another serum opsonin involved in the handling of apoptotic cells, prior to MBL partly inhibited MBL binding to apoptotic cells and vice versa. MBL/MASP could initiate deposition of purified complement C4 on the target cells. However, addition of MBL/MASP to whole serum deficient for both C1q and MBL did not enhance deposition of C4, but MBL enhanced phagocytosis of apoptotic cells by macrophages. These results demonstrate that MBL interacts with structures exposed on cells rendered late apoptotic or necrotic and facilitates uptake by macrophages. Thus, MBL may promote non-inflammatory sequestration of dying host cells.

Abbreviations:
SLE:

Systemic lupus erythematosus

MBL:

Mannose-binding lectin

MASP:

MBL-associated serine protease

APC:

Allophycocyanin

SP-A:

Surfactant protein A

SP-D:

Surfactant protein D.

1 Introduction

Systemic lupus erythematosus (SLE) is a prototypical autoimmune disease, which affects many organ systems. During the course of the disease, an array of clinical symptoms and immunological manifestations is seen. Environmental, hormonal and genetic factors appear to play a role in the pathophysiology. Common HLA haplotypes and rare deficiencies in the components of the classical pathway of complement are observed with increased frequency in SLE patients. In particular, deficiency of the classical complement pathway component C1q is associated with an extraordinarily elevated risk of SLE, because more than 95% of the individuals with this condition are being affected. In addition, deficiencies of C4 and C2 confer increased risks to develop SLE.

Until recently, it has been the common notion that deficiencies in the classical complement pathway lead to SLE because of defective clearance of immune complexes. However, this view was questioned by Korb and Ahearn 1, suggesting that early classical complement components and especially C1q are crucial in handling of cells undergoing programmed cell death (apoptosis),which are thought to be the reservoir for autoantigens in SLE. This hypothesis has been substantiated by the development of C1q-deficient mice lacking C1q and exhibiting a phenotype that resembles several features of human SLE 2. In particular, these mice developed autoantibodies against nuclear antigens and glomerulonephritis. Moreover, it has been shown that the binding of C1q to apoptotic cells and blebs induces activation of the complement system 3.

MBL is a molecule that shares many features with C1q 4. Both molecules are composed of trimeric subunits with collagenous tail domains that are involved in receptor interactions. Whereas C1q is the central recognition molecule in the classical complement pathway, MBL is one of the recognition molecules in the lectin pathway of complement activation, binding to mannose and N-acetylglucosamine sugar groups on different microorganisms. MBL is present in the blood in complexes with different MASP, inducing activation of the lectin pathway of complement. Inter-individual differences in the MBL serum concentration are due to common variant alleles in the structural as well as in the regulatory part of the MBL gene (mbl2) 5. It has been shown that MBL is important for the protection against infections, particularly during infancy 6, but also in adult patients with a concomitant disease or immunodeficiency 7. Moreover, the presence of MBL variant alleles is associated with increased risk of SLE (for a comprehensive analysis see 8). Recent findings show that MBL binds to apoptotic cells and that both C1q and MBL use calreticulin and CD91 to initiate macropinocytosis and uptake of apoptotic cells by phagocytes 9.

In the present study we have investigated (i) whether MBL binds to a range of cells rendered apoptotic as well as to apoptotic blebs, and at which stage in the process the binding takes place; (ii) whether MBL binds to necrotic cells; (iii) whether C1q and MBL may bind to adjacent structures on apoptotic cells; (iv) whether MBL facilitates complement activation upon binding to apoptotic cells; and (v) whether MBL enhances phagocytosis of apoptotic cells.

2 Results

2.1 MBL binding to apoptotic cells

Analysis of apoptotic cells by flow cytometry showed changes in morphology, i.e. a decrease in forward scatter and an increase in side scatter compared to vital cells. Staining with FITC-conjugated annexin V, 7-amino actinomycin D staining or propidium iodide staining indicated altered membrane exposure of phospholipids and alterations in membrane integrity. Fig. 1 upper panel shows a forward versus side scatter dot plot of gamma-radiated Jurkat T cells treated as described in Sect. 4. Based on their forward and side scatter characteristics, the cells could be divided into two distinct populations: R1 and R2. Clear binding of MBL was observed to the apoptotic cell population (R2), and no binding was seen to vital cells (R1), based on the characteristics derived from the forward and side scatter (Fig. 1, lower panel). Similar results were obtained with gamma-radiated Ramos B cells, etoposide-treated Jurkat T cells, gamma–radiated PBMC and staurosporin-treated HUVEC (Fig. 2). In alternative experiments, apoptosis was also induced with an anti-FAS antibody and camptothecin. The binding of MBL to apoptotic cells was further characterized by fluorescence microscopy. Apoptotic Jurkat T cells were identified on the basis of nuclear condensation and fragmentation (Fig. 3). No staining was observed when MBL was omitted.

Figure 1.

Binding of purified human MBL (1.0 μg/ml) to gamma-radiated Jurkat T cells gated on basis of forward and side scatter. The upper panel shows the forward and side scatter, and the regions used to detect MBL binding is shown in the lower panel. Detection of MBL was performed by a biotinylated anti-MBL mAb (clone 131-10) followed by streptavidin conjugated with PE. The results shown are representative for three individual experiments.

Figure 2.

Binding of MBL to different types of apoptotic cells. Binding of MBL to (A) gamma-radiated apoptotic Ramos B cells, (B) etoposide-treated apoptotic Jurkat T cells, (C) gamma-radiated PBMC and (D) staurosporin-treated apoptotic HUVEC on the basis of forward and side scatter. Histogram shows apoptotic cells incubated with MBL (filled histogram) or in the absence of MBL (open histogram). Detection of MBL was performed either by a biotinylated anti-MBL mAb (clone 131-10) followed by streptavidin conjugated with PE in (A) and (C), or by an anti-MBL mAb (clone 3E7) followed by a PE-conjugated polyclonal goat anti-mouse antibody in (B) and (D), respectively. The results shown are representative for three individual experiments.

Figure 3.

MBL binds specifically and directly to apoptotic Jurkat T cells. Apoptotic Jurkat T cells were incubated with or without purified MBL, followed by detection with an anti-MBL mAb and nuclear staining with Hoechst 33258. In the upper left panel, MBL binding by fluorescence microscopy is shown, and in upper right panel from the same magnification field nuclear staining is shown. Lower panels show corresponding control fluorescence microscopy without MBL added.

2.2 MBL binding to late apoptotic cells

Early and late apoptotic cells can be discriminated by staining the cells with FITC-conjugated annexin V and 7-amino actinomycin D when tested in flow cytometry. In Fig. 4, it is shown that MBL does neither bind to vital cell regions (R1) nor to cells staining only for FITC-conjugated annexin V (R2). Only cells staining both for FITC-annexin-V and 7-amino actinomycin D (R3) bound MBL. These results indicate that the ligand for MBL binding is exposed on late apoptotic cells.

Figure 4.

MBL binding to late apoptotic Jurkat T cells. Gamma-radiated Jurkat T cells were prepared as described in Sect. 4. Detection of MBL was performed by a biotinylated anti-MBL mAb (clone 131-10) followed by a PE-conjugated streptavidin. Counter staining was performed with annexin V conjugated to FITC, and 7-amino actinomycin D (7-AAD). Based on annexin V and 7-AAD-staining, the cells were divided into three populations (upper panel). R1 indicates vital cells, R2 indicates early apoptotic cells and R3 indicates late apoptotic cells. Histograms show MBL binding to each population (lower panel). The results shown are representative for three individual experiments.

2.3 MBL binding to apoptotic blebs

Microparticles were derived from the supernatant of apoptotic Jurkat T cells, which were cultured in the presence or absence of MBL during induction of apoptosis. The blebs stained positive with antibodies against CD2 and CD3 as well annexin V and propidium iodide. Following cell culture in the presence of MBL, more than 40% of the blebs stained positive for MBL as detected by flow cytometry, whereas the controls, generated during culture without MBL, were negative (Fig. 5).

Figure 5.

Binding of MBL to blebs derived from apoptotic Jurkat cells. Apoptotic blebs were prepared as described in Sect. 4. Blebs were incubated with (A) or without MBL (B). Detection of MBL was performed by anti-MBL mAb (clone 3E7) followed by a PE-conjugated polyclonal goat anti-mouse antibody. The results shown are representative for three individual experiments.

2.4 MBL binding to apoptotic cells involves the lectin domain

To determine whether the binding of MBL to apoptotic cells was mediated via its collagen-like or its lectin domain, purified MBL/MASP complexes were incubated with apoptotic cells in the presence of an EDTA-containing buffer. EDTA abrogated the binding of MBL to the apoptotic Jurkat T cells, to HUVEC (Fig. 6) and also to apoptotic Ramos B cells (data not shown). Moreover, when MBL was incubated in the presence of an anti-MBL mAb against the lectin domain (clone 3F8), no binding to apoptotic Jurkat T cells or HUVEC was observed (Fig. 6). By contrast, MBL binding to Jurkat T cells and HUVEC was not affected when a non-inhibitory mAb against the collagen-like domain of MBL (clone 1C10) was used under the same experimental conditions (Fig. 6).

Figure 6.

MBL binding to apoptotic cells via the lectin domain. Apoptotic Jurkat T cells and HUVEC were incubated with MBL in the presence or absence of EDTA, an mAb (clone 3F8) against the lectin domain of MBL or a control anti-MBL antibody (1C10), respectively. Detection of MBL was performed by anti-MBL mAb (clone 3E7) followed by a PE-conjugated polyclonal goat-anti-mouse antibody. Results are expressed as relative fluorescence intensity, calculated by dividing the mean fluorescence intensity in the presence of MBL and an inhibitor with that obtained in the presence of anti-MBL alone. The results are presented as the mean ± SD of duplicates of one representative out of two experiments.

2.5 MBL binding to necrotic cells

In order to study the binding of MBL to necrotic cells, necrosis was induced in Jurkat T cells by heat shock. A fraction of the cells maintained its membrane integrity as shown by the absence of staining by FITC-conjugated annexin V and 7-amino actinomycin D. Uniform and strong binding of MBL was observed to the necrotic cell population (Fig. 7). Also MBL binding to necrotic cells involves the lectin domain, since this binding could be inhibited with EDTA, indicating its calcium dependence.

Figure 7.

Binding of MBL to necrotic cells. Necrosis was induced by exposing Jurkat T cells to heat shock by incubating the cells at 56°C for 30 min. Detection of MBL was performed by a biotinylated anti-MBL mAb (clone 131-10) followed by streptavidin conjugated with PE. Histograms show fluorescence in the presence of MBL, with MBL and EDTA and in the absence of MBL. The results shown are representative for three individual experiments.

2.6 Lack of MBL binding to ionomycin-treated erythrocytes

To investigate in more detail whether MBL binding to apoptotic cells could be due to an interaction with externally exposed membrane phospholipids, ionomycin-treated erythrocytes were used to imitate apoptotic cell membranes. Erythrocytes treated with ionomycin, flip-flop their membrane due to increased intracellular calcium concentration 3. As they lack a nucleus, any eventual binding can only occur to a membrane component. Upon treatment with ionomycin, aminophospholipids are exposed on the erythrocytes. Although more than 90% of the erythrocytes bound FITC-conjugated annexin V, no binding of MBL could be detected (Fig. 8).

Figure 8.

Lack of MBL binding to flip-flopped erythrocytes. Erythrocytes were treated in the presence (A) or absence (B) of ionomycin, followed by incubation with MBL, and double stained for binding of MBL and annexin V conjugated to FITC. The quadrant setting was derived from a negative control staining obtained in the absence of MBL and annexin V. Detection of MBL was performed by anti-MBL mAb (clone 3E7) followed by a PE-conjugated polyclonal goat anti-mouse antibody. The results shown are representative for three individual experiments.

2.7 Competition between MBL and C1q in binding to apoptotic cells

Since C1q has a similar binding pattern on apoptotic cells 3 as we now show for MBL in the present paper, we speculated whether MBL and C1q could bind to the same or adjacent structures on the apoptotic cells. The results outlined in Fig. 9 indicate that both MBL and C1q bind to a subpopulation of same size of late apoptotic cells, indicating that they nearly have the same binding capacity towards apoptotic Jurkat T cells when added in equal concentrations (1 μg/ml). Incubation of the cells with C1q prior to MBL markedly diminished MBL binding when evaluated by flow cytometry. Furthermore, also the binding of C1q was strongly inhibited by pre-incubation of the cells with MBL, clearly indicating that both MBL and C1q bind to the same or adjacent structures on apoptotic Jurkat T cells.

Figure 9.

MBL and C1q compete for binding to apoptotic cells. Apoptotic cells were preincubated with or without 1.0 μg/ml purified C1q or MBL, followed by incubation with purified MBL or C1q, respectively. Detection of MBL and C1q was performed with an anti-MBL mAb (clone 3E7) or with an anti-C1q mAb (clone 2204), respectively, followed by a PE-conjugated polyclonal goat anti-mouse antibody. The results shown are representative for two individual experiments.

2.8 MBL binding and complement deposition

To demonstrate the capacity of MBL to activate the lectin complement pathway, we first used purified MBL-containing MASP complexes and purified C4. When MBL/MASP complexes were incubated with apoptotic Jurkat T cells in different concentrations, an MBL dose-dependent deposition of C4 was detected by flow cytometry (Fig. 10A). However, when we used full serum from individuals either homozygous (n=10) for the normal MBL allele (A/A), heterozygous (n=8) or homozygous (n=2) for the variant allele in codon 54 (A/B or B/B), no difference in C4 deposition was observed (Fig. 10B). In control experiments, no difference between the different genotypes was observed when we used IgM as a ligand for classical pathway activation, while as expected a much higher deposition was observed for sera from normal A/A individuals than from individuals carrying MBL variant alleles when we investigated deposition of C4 on mannan, an MBL pathway activator (Fig. 10B). Moreover, a close correlation between the capacity to deposit C4 on IgM and on apoptotic cells were observed, while no correlation was observed between the capacity to deposit C4 on mannan and on apoptotic cells (Fig. 10B). In additional experiments, we used serum from an individual homozygous deficient for MBL (B/B), which was depleted of C1q. No increase in C4 deposition was observed when purified MBL/MASP complexes were added to the serum (Fig. 10C, right panel). Control experiments showed that the binding of MBL to the apoptotic cells was unaffected by the presence of serum (Fig. 10C, left panel). In contrast, addition of purified C1q to the C1q-depleted serum resulted in binding of C1q and a marked increase in deposition of C4.

Figure 10.

Complement deposition on apoptotic cells. (A) Apoptotic Jurkat T cells were incubated in the presence of increasing amounts of purified MBL/MASP complexes, followed by incubation with purified C4 (1.0 μg/ml). Detection of C4 was performed by an anti-C4 mAb (clone C4-4A) followed by a PE-conjugated polyclonal goat-anti-mouse antibody. (B) Plates were coated with IgM or mannan. Sera obtained from donors with a wild-type (n=10) or mutant MBL genotype (n=10) were incubated in a calcium and magnesium-containing buffer [tested at 1:1,500 (IgM), 1:100 and 1:400 (mannan)]. Results were expressed in U/ml on basis of results obtained with serial dilutions of a standard serum arbitrarily set at 1,000 U/ml. Alternatively, serum was incubated with apoptotic Jurkat T cells (dilution 1/10), and C4 deposition was assessed (expressed as mean fluorescence intensity, MFI). A horizontal line indicates means of each group. Correlations are shown for the classical pathway activity (C4 deposition on IgM) and the MBL pathway activity (C4 deposition on mannan) in sera with C4 deposition on apoptotic cells (lower panel). (C) Apoptotic Jurkat T cells were incubated with MBL or C1q, C1q-deficient serum from a donor with a mutant MBL genotype (MBL/C1qD serum) supplemented with MBL or with C1q or MBL/C1qD serum alone, and assessed for deposition of MBL, C1q (left panel), and C4 (right panel) by flow cytometry, as described in Sect. 4. Data are presented as the mean ± SD of one representative out of two experiments.

2.9 MBL enhances phagocytosis of apoptotic cells

To determine whether opsonization of apoptotic cells with MBL would enhance the uptake of these cells by macrophages, we performed phagocytosis experiments with macrophages. As shown in Fig. 11, opsonization of apoptotic cells with MBL enhanced their uptake by macrophages.

Figure 11.

MBL enhances phagocytosis of apoptotic cells. CFSE-labeled Jurkat cells were incubated with 1 μg/ml MBL in the presence of etoposide for 18 h. The cells were washed in serum-free medium and then incubated with MΦ for 2 h. The MΦ were detached and then labeled with an anti-CD14 mAb and analyzed by flow cytometry. The results are expressed as the percentage of MΦ that bound or internalized CFSE-positive cells. One representative experiment of four identical experiments is shown.

3 Discussion

Evidence indicates that dysfunction in sequestration of apoptotic material is important for the pathophysiology of SLE (waste disposal hypothesis) 1012. This notion has led to extensive research, showing that opsonins of importance for innate immunity such as C1q, the short pentraxins. C-reactive protein, serum amyloid P component, the long pentraxin PTX3, the lung surfactant proteins SP-A and SP-D as well as MBL share the ability to bind to apoptotic cells (for a recent review see 13). In the present study, we have shown that MBL binds to a range of different apoptotic cell types, independent of the methods used for inducing apoptosis. This binding seems to appear via the lectin domain in the globular heads, indicated by the necessity of calcium to obtain binding and by the fact that an antibody known to inhibit specific MBL binding of the CRD region of MBL to ligands inhibited also the binding of MBL to apoptotic cells.

As has been shown for C1q and SP-D 3, 14, we were able to demonstrate that MBL binds to late apoptotic and necrotic cells. MBL binds only to cells positive both for annexin V and nuclear dyes when induced to apoptosis, as well to cells becoming directly necrotic after heat shock. Moreover, MBL bound strongly to apoptotic bodies derived from the apoptotic cells.

The failure of MBL to bind to ionomycin-treated erythrocytes which flip-flop their membranes, suggests that phosphatidylserine is not the ligand for MBL. This does not seem to be the case for C1q either 3. Therefore, it is of particular interest that we were able to demonstrate that the binding of MBL to apoptotic cells could be partly inhibited by C1q and vice versa. Thus, MBL and C1q may bind to the same or adjacent structures, which become accessible during the late phase of the apoptotic process. In this regard, it should be mentioned that also the ligand for SP-D is only accessible in late apoptotic cells and necrotic cells, indicating that the opsonins are specialized to handle apoptotic bodies that are not primarily sequestered by phagocytes recognizing early signals in the apoptotic process.

One mechanism by which MBL may mediate sequestration of the apoptotic cells is by activation of the complement system through MASP and subsequent recognition through complement receptors on phagocytes. Consistent with such a view, deposition of C4 was observed when purified MBL-MASP complexes and purified C4 was added to apoptotic cells. However, when we used full serum, no difference was observed between sera containing normal MBL and mutated MBL. Moreover, when we used an MBL-deficient serum depleted for C1q, it was only when we reconstituted with purified C1q and not with MBL that we could observe an enhanced deposition of C4 on the cell surface. By contrast, when mannan was used as ligand, MBL/MASP-mediated C4 deposition could easily be detected. This result indicates that the lectin pathway of complement may require a certain epitope density on the ligand before full complement activation may take place in full serum. At present, we do not know the epitope density on the apoptotic cells, but it is probably much lower than on mannan, which consists almost exclusively of high mannose residues. The physiological relevance of such a phenomenon is obvious in relation to the removal of apoptotic material, which is believed not to mediate amplification of inflammatory mechanisms 15. However, this leaves us with the obvious question, what kind of effector machinery does MBL utilize to remove apoptotic cells? Two recent studies have brought further insight to this question indicating that both C1q, SP-A, SP-D as well as MBL utilize the calreticulin/CD91 complex on phagocytes as acceptor site for sequestration of apoptotic material 9, 16. Consistent with these studies, we demonstrated that opsonization of apoptotic cells with MBL facilitated their uptake by macrophages. Therefore, MBL may promote the clearance of apoptotic cells in a noninflammatory way. It may become particularly relevant in situations where the availability of other opsonins is limited. An important message, which can be drawn from the present study, is that only C1q, but not MBL, could mediate deposition of C4, and this even correlated closely with deposition of C4 on IgM. Accordingly, the classical pathway appears to have a dominant role in complement activation by apoptotic cells.

In conclusion, MBL binds directly to late apoptotic cells, apoptotic blebs as well as to necrotic cells by its lectin domain. MBL and C1q may recognize a joint or adjacent structure exposed on the cells in the late phase of the apoptotic process. MBL binding to apoptotic cells does not lead to activation of complement by the lectin pathway. However, MBL facilitates the uptake of apoptotic cells by macrophages. Thus, MBL may promote non-inflammatory sequestration of dying host cells.

4 Materials and methods

4.1 Purification of MBL and C1q

Purified human MBL containing MASP for therapeutic use was obtained from Statens Serum Institut (Copenhagen, Denmark) 17. In addition, MBL containing MASP was purified from healthy donors as described earlier 18, 19. C1q was isolated essentially as described 3, 20.

4.2 Antibodies and reagents

Biotinylated mouse anti-MBL mAb (clone 131-10, IgG1) was obtained from Statens Serum Institut. Mouse 3E7 anti-MBL mAb (IgG1) was kindly provided by Dr. T. Fujita (Medical University School of Medicine, Fukushima, Japan). Mouse anti-MBL IgG mAb (clone 3F8, inhibitory antibody directed against the CRD domain of MBL and clone 1C10, non-inhibitory anti-MBL antibody) were kindly provided by Dr. G. L. Stahl (Department of Anesthesiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA). A mouse anti-human C4 mAb (clone C4-4A) was kindly provided by Dr. C. E. Hack (Sanquin Blood Supply Foundation, Amsterdam, The Netherlands).

PE-labeled streptavidin and Viaprobe (7-amino actinomycin D) was purchased from Becton Dickinson (Erembodegem, Belgium). Fluorescein isothiocyanate-labeled annexin V and annexin-binding buffer was purchased from PharMingen (San Diego, CA). Tris-HCl buffer (10 mM Tris, 150 mM NaCl, pH 7.4) was purchased from Bie & Berntsen (Rødovre, Denmark). FBS was obtained from Gibco BRL, Life Technologies (Tåstrup, Denmark). RPMI 1640, L-glutamine (200 mM) and penicillin-streptomycin was obtained from Sigma Chemical Co. (St. Louis, MO). β-Mercaptoethanol was purchased from Bie & Berntsen. EDTA and EGTA, were obtained from Sigma. Medium* indicates RPMI 1640 supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine and 1% penicillin/streptomycin.

4.3 Cell lines and cell cultures

Jurkat, a human T lymphocytic leukemia cell line, was grown in suspension culture in 5% CO2 at 37°C in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, penicillin (90 U/ml) and streptomycin (90 μg/ml). Ramos, a Burkitt lymphoma cell line, was cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, penicillin (90 U/ml), streptomycin (90 μg/ml) and 50 μM β-mercaptoethanol in 5% CO2 at 37°C. Cells were in log-phase growth for 24 h when used in the experiments. HUVEC were isolated and cultured as described 21. PBMC were isolated by centrifugation on Ficoll-Hypaque (Pharmacia, Sweden) density gradient, washed and resuspended in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine and 1% penicillin (90 U/ml) and streptomycin (90 μg/ml). The following procedure was used: The obtained blood was diluted 1:1 with medium*. Five milliliter of the diluted blood were carefully placed on 4 ml of cold Ficoll-Hypaque and centrifuged at 360×g for 30 min. The inter-phase cells were then transferred to new vials containing 2 ml of medium* and centrifuged at 581×g for 10 min. The supernatant was discarded, and the pellet resuspended in medium* and centrifuged at 581×g for 10 min. The supernatant was again discarded, and the cells were resuspended in medium*.

4.4 Apoptosis induction

Cells from cell culture bottles were centrifuged at 360×g for 10 min and resuspended in growth medium prior to usage. Cell concentration was measured by the eosin method and then adjustedto 0.5×106 cells/ml in growth medium. To induce apoptosis, cells were exposed to gamma radiation at a dose of 24 Gray and at a rate of 3 Gray/min. In our set-up, this equaled 8 min using a cesium source. Irradiated cells and control cells were incubated at 37°C, 5% CO2 for 18 h. Alternatively, apoptosis was induced in Jurkat T cells in 24-wells plates (5×105 cells/ml; 1 ml/well) by incubation with 40 μM etoposide (Sigma) for 20 h, using RPMI 1640 medium containing 0.5% of the above solution with FBS or in serum-free AIMV medium (Gibco). Moreover,Jurkat T cells were incubated with 1.0 μg/ml of mAb against Fas (CD95), clone DX2 (BD Biosciences), and 1.0 μg/ml recombinant protein G (BD Biosciences) for 5 h. When PBMC were used, the incubation was preceded by 18 h incubation with 1 mg/ml PHA. In addition, Jurkat T cells or PBMC were incubated with 0.15-1.0 mM camptothecin for 3-5 h. Apoptosis was induced in HUVEC when the cells were cultured in the presence of staurosporin (1.0 μM, Sigma) for 20 h.

4.5 Necrosis induction

To induce cell necrosis, Jurkat cells were incubated at 56°C for 30 min and were subsequently used for experiments by the same MBL-binding procedure as for the apoptotic cells.

4.6 Binding of MBL to apoptotic and necrotic cells

Apoptotic or necrotic cells were incubated with 1.0-5.0 μg/ml MBL in Tris-HCl + 5 mM CaCl2 for 1 h, 37°C, 5% CO2. The cells were subsequently washed twice with Tris HCl +5 mM CaCl2 + 10% FBS, and centrifuged at 581×g for 5 min. In experiments using a calcium chelator, 10 mM EDTA was included in the buffer solution, when cells were incubated with MBL. Subsequently, 1×105 cells were incubated with biotinylated antibody 131-10 at a concentration of 0.3-1.0 μg/ml for 45 min at 4°C. The cells were washed, centrifuged at 478×g for 5 min and incubated for 30 min with streptavidin-PE (1:120 final dilution) at room temperature. In some experiments, cells were stained for 15 min with FITC-conjugated annexin V (1:120 final dilution) and 7-amino actinomycin D (1:50 final dilution). The cells were washed and resuspended in a calcium-enriched buffer (annexin V-binding buffer, PharMingen). Fluorescence intensity was measured by flow cytometry using a FACScan from Becton Dickinson, and 20,000 events were collected per sample. The FACScan was calibrated using beads before each experiment. In alternative experiments, cells were incubated with 1.0 μg/ml MBL in PBS containing 1% BSA (w/v), 0.01% NaN3 (w/v) and 10 mM CaCl2 for 30 min at 0°C. The cells were washed with the same buffer, and incubated with anmAb directed against MBL [clone 3E7, (mouse IgG1), 6.5 μg/ml]. After 30 min at 0°C, cells were washed and incubated with phycoerythrin (PE)-conjugated goat F(ab′)2 anti-mouse Ig (DAKO, Glostrup, Denmark).

In some experiments, MBL was pre-incubated with mAb 3F8 (20 μg/ml) or mAb 1C10 (20 μg/ml), followed by addition of the mixture to the cells. Alternatively, cells were preincubated with1 μg/ml C1q or 1 μg/ml MBL, followed by incubation of the cells with 1 μg/ml MBL or C1q, respectively. Binding of MBL was detected with anti-MBL mAb (clone 3E7), as described above. Binding of C1q was detected with anti-C1q mAb (clone 2204), as described earlier 3.

4.7 Isolation of apoptotic blebs and binding of MBL

Blebs were obtained from Jurkat T cell cultures as described 3. In brief, Jurkat T cells were induced to apoptosis by etoposide as described above in the presence or absenceof 1.0 μg/ml purified MBL. After centrifugation for 10 min at 1,550×g, culture supernatant was centrifuged for 45 min at 15,700×g at room temperature. The supernatant was carefully removed by aspiration, and the pellet was resuspended in 1.0 ml PBS containing 0.32% citrate (pH 7.4). The bleb solution was again centrifuged at 15,700×g for 30 min at room temperature. The resulting pellet was incubated with an anti-MBL mAb (clone 3E7) and allophycocyanin (APC)-labeled annexin V (5 μM; Nexins Research) in PBS containing 2.5 mM Ca2+ for 15 min at room temperature. Blebs were washed with PBS-Ca2+, followed by incubation for 15 min with PE-conjugated goat anti-mouse Ig Ab. Before analysis by flow cytometry using a FACSCalibur (Becton Dickinson), the blebs were washed again with PBS-Ca2+. Alternatively, blebs were stained with FITC-labeled anti-CD2 or anti-CD3 (DAKO, Copenhagen Denmark), APC-conjugated annexin V and propidium iodide.

4.8 Induction of membrane flip-flop in erythrocytes

Human erythrocytes were isolated from fresh heparinized whole blood by centrifugation at 1,300×g for 10 min, cleared from the buffy coat and washed with PBS to remove plasma and contaminating white blood cells. Erythrocytes were treated with ionomycin (Sigma) to induce exposure of phosphatidylserine in the outer leaflet of the membrane as described in 3.

4.9 Complement deposition on apoptotic cells

To determine whether MBL/MASP complexes in different concentrations added to apoptotic Jurkat T cells were able to mediate activation of complement, purified plasma-derived C4 19 was added at a concentration of 1 μg/ml. Cell deposition of C4 was detected with a mouse anti-human C4 mAb (clone C4-4A), followed by a PE-conjugated polyclonal goat-anti-mouse antibody. Deposition of C4 was expressed as mean fluorescence intensity.

To assess deposition of complement in full serum, samples were obtained from ten healthy donors with the normal A/A MBL genotype as well as from ten healthy donors being either heterozygous (n=8) or homozygous (n=2) for the mbl2 codon 54 mutation (genotype B). Serum was aliquoted and stored at –80°C.

Apoptotic Jurkat T cells were incubated with 10% serum, diluted in RPMI 1640 culture medium containing 10% FBS (w/v) for 30 min at 37°C. Cell deposition of C4 was detected with a mouse anti-human C4 mAb (clone C4-4A). Cells were analyzed by flow cytometry using a FACScan. In some experiments, apoptotic cells were incubated with or without 1 μg/ml C1q or 1 μg/ml MBL and 10% C1q-depleted serum from a donor homozygous for the B/B MBL genotype. Normal human serum was depleted for C1q by immune adsorption using Biogel-coupled rabbit-anti-human C1q antibodies. After C1q depletion, the serum showed a normal alternative pathway activity but no classical pathway activity in hemolytic assays, performed as described previously 22. Binding of C1q was detected with an anti-C1q mAb (clone 2204) 3, and the binding of MBL was detected as described above.

To analyze the activity of the classical complement pathway in whole human serum, we used an ELISA method as described 23. Activation of complement by mannan in whole serum was analyzed essentially as described previously 19. The obtained results were standardized using a calibration curve generated with pooled normal human serum, and expressed as arbitrary units per ml (U/ml). The standard serum was arbitrarily set at 1,000 U/ml.

4.10 Fluorescence microscopy

Apoptotic Jurkat T cells were incubated with 1 μg/ml of purified MBL in PBS containing 1% BSA (w/v) and 0.01% NaN3 (w/v) for 30 min at 0°C. Cells were washed and incubated with anti-MBL 3E7 mAb (concentration 25 μg/ml), followed by Oregon–Green-conjugated goat anti-mouse IgG (Molecular Probes). To co-stain for MBL and nucleus, cells were stained for MBL as mentioned above, washed and incubated with Hoechst 33528 (Molecular Probes, 1 μg/ml) in PBS for 3 min. Cells were fixed with 1% formaldehyde for 10 min at 4°C, and cytospin preparations were made. Photographswere taken on Kodak TX-400 films on a Leitz microscope.

4.11 Phagocytosis assay

Human monocytes were isolated on a Percoll gradient as described previously 24. Cells were allowed to mature into macrophages (MΦ) over a 7-day period, in RPMI 1640 with15% FBS, penicillin (90 U/ml), streptomycin (90 μg/ml), and 5 ng/ml GM-CSF. The macrophages were used after 7 days of culture. Apoptotic cells were generated as described above. In brief, Jurkat cells were washed with PBS and stained with 5 μM carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Leiden, The Netherlands). After staining, the cells were washed with RPMI 1640 containing 10% FBS and resuspended in serum-free AIMV culture medium (Gibco). Jurkat cells were opsonized by incubation for 18 h with 1 μg/ml purified human MBL in the presence of 40 μM etoposide to induce apoptosis. Labeled apoptotic cells were added to MΦ on day 7 of culture for 2 h at 37°C or 4°C in 250 μl of AIMV culture medium. MΦ were labeled with anti-CD14 and PE-conjugated goat anti-mouse Ig antibodies (DAKO). Uptake was analyzed by flow cytometry.

Acknowledgements

Grants support was obtained from EU (QLGI-CT-2001-01039), the Novo Nordisk Research Foundation, The Danish Medical Research Council, The Danish Rheumatism Association, Copenhagen Hospital Corporation Research Foundation, the Netherlands Organization for Scientific Research (091-12-095), and by the Dutch Kidney Foundation (PC95). We want to thank Drs. G. L. Stahl, T. E. Fujita and C. E. Hack for providing valuable antibodies. Ms. Francien Fallaux-van den Houten, Ms. M. C. Faber-Krol and Ms. Vibeke Weirup are acknowledged for providing excellent technical assistance.

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

  8. 8

    WILEY-VCH

  9. 9

    WILEY-VCH

  10. 10

    WILEY-VCH

  11. 11

    WILEY-VCH

Ancillary