Viscum album agglutinin-I (VAA-I) increases cell surface expression of cytoskeletal proteins in apoptotic human neutrophils

Abstract

Viscum album agglutinin-I (VAA-I) is a plant lectin, which possesses anti-inflammatory properties, including the ability to induce neutrophil apoptosis by a mechanism that is not completely understood. Among the three actin-binding membrane-anchoring proteins ezrin/radixin/moesin (ERM), neutrophils are known to express ezrin and moesin. The behavior of these proteins in apoptotic neutrophils is not well established. In the present study, the expression and localization of ezrin and moesin by Western blot and immunofluorescence revealed a clear degradation and relocalization of both the proteins during VAA-I-induced apoptosis. Also, flow cytometry analysis revealed that VAA-I markedly and significantly induced the cell surface expression of ezrin and moesin and this was reversed when cells were pretreated with the Syk inhibitor piceatannol. The expression of ezrin and moesin on the cell surface of apoptotic neutrophils may represent a mechanism responsible for the appearance of autoantibodies directed against ERM proteins, which have been found in the serum of patients suffering from autoimmune diseases. Therefore, the ability of VAA-I to increase cell surface expression of cytoskeletal proteins in apoptotic neutrophils provides important insight into a possible toxic mechanism of this plant lectin and this has to be considered for its potential utilization for in vivo treatment.

Introduction

Polymorphonuclear neutrophils (PMNs) are the most abundant human leucocytes in circulation. They are particularly known for their role in innate immune response and are highly involved in inflammation. One of the best mechanisms for the resolution of inflammation involves apoptosis of PMNs, which are then eliminated by professional phagocytes including macrophages. Although there is extensive literature regarding caspase activation and cell signaling events involved in PMN apoptosis, ^1–5 the characterization, reorganization and behavior of cytoskeletal proteins are far from being fully understood. In the past few years, we have demonstrated that microfilament (MF)-associated proteins (MFAPs), such as gelsolin, paxillin and nonmuscle myosin heavy chain IIA (NMHC-IIA), but not vinculin, are cleaved during spontaneous apoptosis (SA) and during agent-induced cell apoptosis, including the potent proapoptotic plant lectin Viscum album agglutinin-I (VAA-I).^6–10 VAA-I is a 63-kDa galactoside-specific plant lectin, which belongs to the family of type II ribosome-inactivating proteins composed of two distinct subunits: the A chain (29 kDa) and the B chain (34 kDa). The A chain confers the property of inhibition of protein synthesis to the VAA-I molecule by acting as a ribosome-inactivating agent. The B chain allows the VAA-I molecule to bind to terminal galactoside residues on the membranes of various cells. VAA-I is a potent proapoptotic molecule not only in mature PMNs but also in promyelocytic PLB-985 cells, eosinophilic AML14.3D10 cells and primary mature eosinophils. ^9,11,12 Interestingly, while the two major microtubule (MT) proteins α- and β-tubulin are not cleaved by VAA-I, another proapoptotic agent, arsenic trioxide (ATO), induces cleavage of these two proteins in human PMNs, ¹³ indicating the presence of selective mechanisms involved in cytoskeleton breakdown during PMN apoptosis. In addition, the two intermediate filament (INFIL) proteins, namely vimentin and lamin B₁, the only INFIL proteins known to be expressed in human PMNs, are also cleaved during SA and in response to VAA-I and are expressed on the cell surface of apoptotic PMNs. ¹⁴ Therefore, to date, among the three major types of cytoskeletal filaments (MF, INFIL and MT) tested, only proteins of the INFIL, vimentin and lamin B₁, are known to be expressed on the cell surface of PMNs during SA by an as yet unknown mechanism; their role is still unknown but may represent a source of autoantigens.

The ezrin/radixin/moesin (ERM) proteins are three closely related cytoskeletal proteins of similar molecular weight (∼75 kDa) principally located in microvilli where the plasma membrane is linked to the cytoskeleton. ¹⁵ ERM proteins are known to be enriched in membrane ruffles, cell–cell junctions, the cleavage furrow of dividing cells and link integral membranes proteins to the cortical actin cytoskeleton that lies just beneath the plasma membrane. ^15,16 In their active conformation, the N-terminal domain of ezrin, radixin and moesin binds to the cytoplasmic tail of transmembrane proteins, while the C-terminal domain binds to the F-actin cytoskeleton. ERM proteins also exist in an autoinhibited conformation, in which the N-terminal domain is masked by the remainder of the molecule. Ezrin and moesin, but not radixin, are expressed in lymphocytes, monocytes and neutrophils, moesin being quantitatively predominant. In PMNs, members of the ERM family play an important role because reversible actin-membrane linkage is essential for the maintenance of cell shape, cell adhesion and migration of these leucocytes. ¹⁵ Curiously, presently there is no data regarding the role and behavior of ERM proteins during PMN apoptosis, an essential mechanism for modulating PMN homeostasis and inflammation. In the present study, we report that VAA-I induced the cleavage and relocalization of both moesin and ezrin. In addition, we report that the ability of VAA-I to increase cell surface expression of ezrin and moesin occurs by a Syk-dependent mechanism.

Materials and methods

Chemicals, agonists and antibodies

Roswell Park Memorial Institute (RPMI) 1640, 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES), penicillin/streptomycin (P/S), bovine serum albumin (BSA), VAA-I, ATO, mouse monoclonal antigelsolin antibody (clone GS-2C4) and piceatannol (Pic) were purchased from Sigma-Aldrich Chemical Co. (St Louis, Missouri, USA). Mouse monoclonal antivimentin (clone V9), mouse monoclonal anti-human ezrin (3C12) and mouse monoclonal anti-Syk antibodies were purchased from Santa Cruz Biotechnologies (Santa Cruz, California, USA). Mouse monoclonal anti-human moesin antibody was from Abcam (San Francisco, California, USA). Horseradish peroxidase (HRP)-conjugated goat anti-mouse and goat anti-rabbit antibodies, goat and rabbit immunoglobulin G (IgG; F(ab′)₂) fragment-specific phycoerythrin (PE)-conjugated AffiniPure F(ab′)₂ fragments were purchased from Jackson Immunoresearch (West Grove, Pennsylvania, USA). Control isotypic IgG1 antibody was obtained from PharMingen (Mississauga, Ontario, Canada). Recombinant human granulocyte–macrophage colony stimulating factor (GM-CSF) was purchased from PeproTech Inc. (Rocky Hill, New Jersey, USA). Goat anti-mouse Alexa Fluor 488 conjugate was obtained from Molecular Probes (Camarillo, California, USA). Vectashield containing 4′,6-diamidino-2-phenylindole (DAPI) was ordered from Vector Laboratories (Burlingame, California, USA).

Human neutrophil isolation

PMNs were isolated from the venous blood of healthy volunteers by dextran sedimentation followed by centrifugation over Ficoll-Paque (Amersham Pharmacia Biotech Inc., Baie d’Urfé, Québec, Canada), as described previously. ¹⁴ Blood donations were obtained from individuals who provided informed consent according to our institutionally approved procedures. Cell viability (>98%) was monitored by trypan blue exclusion, and the purity (>98%) was verified by cytology from cytocentrifuged preparations colored with the Hema 3 stain set (Biochemical Sciences Inc., Swedesboro, New Jersey, USA).

Assessment of neutrophil apoptosis

Freshly isolated human PMNs (10 × 10⁶ cells/ml in RPMI 1640–HEPES–P/S, supplemented with 10% autologous serum) were incubated for 22 h in the presence or absence of PMN agonists. VAA-I was used at 500 ng/ml throughout the study. ⁶ Cytocentrifuged samples of PMNs were prepared using a Cyto-tek® centrifuge (Miles Scientific, Naperville, Illinois, USA) and processed as documented previously. ¹³ Cells were examined by light microscopy at 400× final magnification, and apoptotic PMNs were defined as cells containing one or more characteristic, darkly stained pyknotic nuclei.

Western blot

PMNs (10 × 10⁶ cells/ml) were incubated in the presence or absence of agonists for the indicated periods of time, and the expression of cytoskeletal proteins was performed by Western blot as published previously. ⁹ Briefly, cells were harvested for the preparation of cell lysates in 2× Laemmli sample buffer. Aliquots corresponding to 1 × 10⁶ cells were loaded onto 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and transferred from gel to polyvinylidene difluoride membranes. Nonspecific sites were blocked with the blocking solution (5% BSA in Tris-buffered saline–Tween (25 mM Tris-HCl, pH 7.8, 190 mM sodium chloride (NaCl), 0.15% Tween-20)) for 1 h at room temperature. Membranes were washed and incubated with anti-human cytoskeletal antibodies (mouse monoclonal anti-ezrin (1:1000), mouse monoclonal anti-moesin (1:1000) and mouse monoclonal anti-gelsolin (1:1000)), overnight at 4°C. After several washes, membranes were incubated with HRP-labeled goat anti-mouse IgG antibody (1:25,000) or HRP-labeled goat anti-rabbit IgG antibody (1:25,000) for 1 h at room temperature in blocking solution. Bands were revealed with the enhanced chemiluminescence–Western blotting detection system (Amersham Pharmacia Biotech Inc.).

Immunofluorescence microscopy

After incubation with indicated agents, cells were cytocentrifuged on glass coverslips (Fisher Scientific, Ottawa, Canada) and then fixed and permeabilized in 3.7% paraformaldehyde (Sigma) + 0.1% digitonin (Sigma) at room temperature for 20 min. After four washes with PBS, cells were incubated with primary antibodies (mouse anti-human specific for ezrin or moesin (5 µg/ml)) for 40 min at 37°C. After four washes with PBS, cells were incubated with 3 µg/ml of goat anti-mouse Alexa Fluor-488 conjugate for 40 min at 37°C. After four washes with PBS, coverslips were mounted with Vectashield + DAPI (Vector Laboratories). Fluorescent-labeled cells were captured from high-power fields (400×) observed with a photomicroscope Leica equipped with an ebq 100 dc epifluorescent condenser. Images were taken with a Cooke Sensicam high-performance camera (Applied Scientific Instrumentation, Eugene, OR, USA) coupled to the Image Pro-plus (Version 4.0; Media Cybernetics, Rockville, MD, USA) program.

Cell surface expression of cytoskeletal proteins by flow cytometry

PMNs (10 × 10⁶ cells/ml RPMI–HEPES–P/S) were incubated at 37°C, 5% CO₂, in the presence or absence of proapoptotic agent VAA-I for 22 h. Cells were harvested in cold PBS and blocked with PBS containing 20% autologous serum for 30 min on ice. Cells were washed in PBS and incubated for 30 min on ice with 2 µg/ml mouse monoclonal anticytoskeletal antibodies (antivimentin (V9), anti-moesin or antiezrin). Appropriate isotypic control antibody (mouse IgG1) was used to compare with the proteins of interest. Cells were washed in PBS and incubated for 30 min on ice with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse. Cell surface expression was analyzed using a FACScan. Results are expressed as a mean fluorescence intensity (MFI; G-mean) obtained by subtracting the G-mean value of the corresponding isotypic control from the G-mean value obtained with the anticytoskeletal antibody directed against the protein of interest. In some experiments, cells were preincubated for 30 min at 37°C with the Syk inhibitor, Pic (30 µM), before investigating cell surface expression of cytoskeletal proteins.

Immunoprecipitation

PMN cell lysates (10 × 10⁶ cells/condition) were lysed in nondenaturing cold lysis buffer (50 mM Tris–HCl (pH 8), 50 mM NaCl, 1% Triton X-100, 125 mM phenylmethylsulfonyl fluoride (PMSF), 10 g/ml trypsin inhibitor, aprotinin, leupeptin and pepstatin, 1 mM orthovanadate) for 1 h on ice and sonicated three times for 20s. The lysates were preincubated with protein G-Sepharose® (GE healthcare, Upsala, Sweden). After 1 h, samples were centrifuged to remove sepharose beads and then incubated with 2 μg/ml mouse anti-human Syk at 4°C with gentle agitation overnight. Protein G-Sepharose was then added for 4 h with gentle agitation at 4°C. The solid matrix was collected and washed three times with lysis buffer before adding an equivalent volume of sample buffer. The samples were then boiled at 100°C for 10 min. Immunoprecipitates were electrophoresed on an SDS–polyacrylamide gel, followed by immunoblotting with anti-moesin or anti-Syk antibodies.

Statistical analysis

Statistical analyses were performed, using Student’s t test with SigmaStat for Windows (version 3.0). Statistical significance was established at p < 0.05.

Results

VAA-I induces degradation of moesin and ezrin in apoptotic neutrophils

Since cytoskeletal breakdown occurs in 22 h-aged PMNs (SA), as evidenced by cleavage of several cytoskeletal proteins, ^9,13 we first decided to compare potential cleavage of moesin and ezrin during SA. Unexpectedly, we found that, unlike ezrin, the expression of moesin remained stable during SA (Figure 1(a)). Since ezrin alone is cleaved during SA, in which about 30–50% of PMNs are typically apoptotic, ^13,9 we next decided to examine whether or not moesin cleavage could be observed under experimental conditions in which more cells are in apoptosis. To do so, cell apoptosis was induced with the neutrophil proapoptotic VAA-I molecule. ^6,9 To be sure that the PMNs were apoptotic, apoptosis was assessed in parallel by cytology where pyknotic nuclei, a hallmark of apoptosis, are easily observable in these cells. As illustrated in Figure 1(b), only VAA-I led to a decrease in moesin expression under conditions in which almost all cells (99%) were apoptotic (inset). Therefore, moesin may be degraded during PMN apoptosis but probably not as rapidly and/or as strongly as ezrin. As expected, ^6,17 the MFAP gelsolin was cleaved during SA: this cleavage was prevented in the presence of GM-CSF. Both ATO and VAA-I also induced cleavage of gelsolin.

Figure 1.

Moesin, unlike ezrin, is not degraded during neutrophil spontaneous apoptosis. (a) PMNs were freshly isolated (F) and incubated for 22 h in the absence of any agonists (SA). Cell lysates were prepared for both fractions (F) and SA, and the expression of ezrin (upper panel) and moesin (lower panel) was monitored by Western blotting as indicated in Materials and methods section. The corresponding Coomassie blue-stained membrane shows the equal loading of proteins. (b) Cells were incubated in the presence of buffer (SA), the antiapoptotic cytokine GM-CSF (GM, 65 ng/ml), the proapoptotic anticancer agent ATO at 5 or 10 μM or with the potent proapoptotic molecule VAA-I (500 ng/ml) for 22 h and degradation of moesin (upper panel) and gelsolin (positive control) was assessed by immunoblotting. The apoptotic rate, as assessed by cytology, is indicated at the bottom of the figure. Note that the cleavage of moesin was evident when PMNs were treated with VAA-I, where almost all cells were in apoptosis (inset). Results are from one representative experiment out of three. VAA-I: Viscum album agglutinin-I; PMN: polymorphonuclear neutrophil cell; SA: spontaneous apoptosis; ATO; arsenic trioxide; GM-CSF: granulocyte-macrophage colony stimulating factor.

Behavior of moesin and ezrin in VAA-I-induced neutrophil apoptosis

Knowing that moesin and ezrin are cleaved in VAA-I-induced apoptotic human PMNs, we investigated the behavior of these proteins by immunofluorescence microscopy. As illustrated in Figure 2, moesin and ezrin are preferentially concentrated at the plasma membrane in nonapoptotic cells and are also distributed diffusely in the cytoplasm (panels (b) and (d), respectively). However, in apoptotic PMNs, the signal of fluorescence observed at the plasma membrane faded and a clear cytoplasmic translocation of moesin (Figure 2(f)) and ezrin (Figure 2(h)) was observed.

Figure 2.

Localization of moesin and ezrin during spontaneous and VAA-I induced apoptosis. Localization of moesin and ezrin was investigated by immunofluorescence in fresh PMNs (a–d) or in VAA-I-induced apoptosis (e–h). PMNs were then fixed and permeabilized in 3.7% paraformaldehyde + 0.1% digitonin, incubated with antibody directed against moesin (b and f) or ezrin (d and h), and DNA was stained with DAPI (a, c, e and g) as indicated in Materials and methods section. Results are from one representative experiment out of three. VAA-I: Viscum album agglutinin-I; PMN: polymorphonuclear neutrophil cell; DAPI: 4′,6-diamidino-2-phenylindole.

Cell surface expression of moesin and ezrin during neutrophil apoptosis

We have previously documented that among several cytoskeletal proteins, the two INFIL proteins, vimentin and lamin B₁, are expressed on the cell surface of human PMNs during SA. ¹⁴ Therefore, we next investigated whether or not moesin and ezrin were expressed on the cell surface of apoptotic PMNs, using vimentin as a positive control. ¹⁴ As illustrated in Figure 3, cell surface expression of moesin and ezrin was observed in VAA-I-induced PMN apoptosis but not during SA. The percentage of cells in apoptosis was determined in parallel by cytology. As expected, cell surface expression of vimentin was observed in both the conditions. Of note, no expression of ezrin, moesin and vimentin was observed in fresh nonapoptotic cells (Figure 3(c)).

Figure 3.

Cell surface expression of moesin and ezrin in VAA-I-induced neutrophils. PMNs were treated with buffer (SA) or with 500 ng/ml VAA-I for 22 h and cell surface expression of vimentin (positive control), moesin or ezrin was determined by flow cytometry as detailed in Materials and methods section. (a) Representative data. (b) A bar graph illustrating the cell surface expression of the indicated protein expressed as MFI during SA or in VAA-I-induced PMNs (VAA). Results are means ± SEM (n = 6). *p < 0.05 versus SA. (c) One representative experiment illustrating that vimentin (upper panel), moesin (middle panel) or ezrin (bottom panel) are not detected on the cell surface of fresh, nonapoptotic PMNs. VAA-I: Viscum album agglutinin-I; PMN: polymorphonuclear neutrophil cell; MFI: mean fluorescence intensity; SA: spontaneous apoptosis.

Evidence that VAA-I induces cell surface expression of ezrin, moesin and vimentin by a Syk-dependent mechanism

Since Syk is involved in PMN apoptosis ¹⁸ and since this enzyme is known to directly associate with ezrin and moesin in leukocytes, ¹⁹ we next determined whether or not moesin (the quantitatively predominant ERM protein in PMNs) can associate with Syk. As illustrated in Figure 4, immunoprecipitation of Syk followed by immunoblotting with anti-moesin antibody revealed that moesin can physically be associated with Syk in human PMNs. Next, we decided to use a Syk inhibitor, Pic and evaluated cell surface expression of both moesin and ezrin in VAA-I-induced PMNs. As illustrated in Figure 5, the ability of VAA-I to increase the cell surface expression of moesin and ezrin was significantly decreased by pretreating cell with Pic. Interestingly, this was also observed for vimentin cell surface expression. Therefore, VAA-I increased cell surface expression of cytoskeletal proteins by a Syk-dependent mechanism.

Figure 4.

Moesin is physically associated with Syk in human neutrophils. (a) Cell lysates were prepared from freshly isolated human PMNs, and Syk was immunoprecipitated as detailed in Materials and methods section and then immunoblotting was performed with an anti-moesin antibody. (b) Syk was immunoprecipitated and then immunoblotting was performed with an anti-Syk antibody for visualizing protein loading. Lanes 1 and 2 are the preparations of two different blood donors. The same experiment was repeated once with two other blood donors. Note that the association of moesin to Syk can vary from donor to donor. PMN: polymorphonuclear neutrophil cell.

Figure 5.

VAA-I increases the cell surface expression of ezrin, moesin and vimentin in apoptotic human PMNs by a Syk-dependent mechanism. PMNs were pretreated with Pic for 30 min and then treated with buffer (SA) or with 500 ng/ml VAA-I for 22 h and cell surface expression of ezrin, moesin or vimentin was determined by flow cytometry as detailed in Materials and methods section. (a) A bar graph illustrating the cell surface expression of ezrin (upper panel), moesin (middle panel) or vimentin (lower panel) expressed as MFI in VAA-I-induced PMNs. Results are means ± SEM (n = 3). *, p < 0.05 versus VAA-I. (b and c) Representative data illustrating that VAA-I increased cell surface expression of ezrin, moesin and vimentin (b) versus SA (c) and that Pic reversed the effect of VAA-I (b). For clarity, only the results from (b) were plotted in the bar graph (a). Note that pretreatment of Pic did not alter SA (both curves overlapped). VAA-I: Viscum album agglutinin-I; PMN: polymorphonuclear neutrophil cell; MFI: mean fluorescence intensity; SA: spontaneous apoptosis; Pic; piceatannol.

Discussion

VAA-I is a very potent proapoptotic molecule in human PMNs, but its mode of action is still not fully understood. Although it was previously reported that VAA-I induced cleavage of cytoskeletal proteins during PMN apoptosis, ⁹ the behavior of ERM proteins has never been investigated prior to the present study. In contrast to the degradation of ezrin, which was easily observable during SA, the cleavage of moesin was only observed in VAA-I-induced PMNs, an experimental conditions where the number of apoptotic cells is very high (close to 100%). This suggests that degradation of moesin correlates with the number of apoptotic PMNs and, to date, this is unique to moesin, since this was not observed for NMHC-IIA, paxillin, gelsolin, vinculin, vimentin, lamin B₁, α- and β-tubulin ^6,7,9,12 and ezrin (this report) that were all cleaved during SA. The fact that moesin and ezrin are expressed at the cell surface of human PMNs during apoptosis may be an example of a mechanism that explains how intracellular cytoskeletal proteins (or epitopes) can become accessible and present to the immune system. Therefore, knowing that VAA-I can increase the cell surface expression of cytoskeletal proteins in apoptotic PMNs is an important novel effect of this lectin that has to be considered in different in vivo models investigating the potential use of VAA-I for therapeutic purpose. Presently, except necrosis or potential molecular mimicry, the presence of autoantibodies directed against intracellular cytoskeletal proteins in the sera of patients suffering from autoimmune diseases is difficult to explain. In addition, knowing that this novel effect of VAA-I occurs, at least partly, by a Syk-dependent mechanism, this information could be advantageously used for better designing future studies with this plant lectin. This will be of importance knowing that Syk is now increasingly known as a crucial player in diverse biological functions and since the role of Syk inhibitors in the management of several diseases, including rheumatoid arthritis (RA), ^20,21 is gaining increasing interest. However, several experiments related to the inflammatory process needed to be performed to better refined potential therapeutic strategies. Using an antisense strategy as well as a pharmacological approach, we have been able to clearly demonstrate the role of Syk in interleukin-4-induced suppression of PMN apoptosis. ¹⁸ It will be interesting to use this strategy in VAA-I-induced human PMNs. This remained to be performed.

Interestingly, using recombinant ERM proteins produced in Escherichia coli, one team reported, almost 15 years ago, that ∼35% of sera obtained from patients with RA reacted with at least one of the three ERM proteins (17%, 15% and 14% for ezrin, radixin and moesin, respectively), ²² suggesting that ERM proteins are potential autoimmune target antigens for RA. Since then, no other studies have reported the presence of autoantibodies directed against ERM proteins in RA. This could be explained by the fact that no significant correlation between the presence of the autoantibodies and clinical manifestation (such as disease duration or stage) was reported in the original study.

In another study, antibodies specific to moesin were detectable in the serum of ∼40% of patients with acquired aplastic anemia (AA), a syndrome characterized by pancytopenia and bone marrow hypoplasia. ²³ A case–control study conducted by International Agranulocytosis and Plastic Anemia Study revealed that history of RA is significantly associated with later development of AA, suggesting that AA and RA share pathogenetic mechanisms leading to a breakdown of immunologic tolerance to moesin. Interestingly, the authors discussed that the membrane-linking protein moesin is expressed by various blood cells, including granulocytes, but that its expression was localized to the inner side of cell membrane and not on to the cell surface. ²³ The data that we bring in the present study indicating that apoptotic PMNs do express moesin (and ezrin) on their cell surface may be helpful for future comprehension in AA and other diseases. Of note, the authors also reported that moesin was detected in the culture supernatant from four leukemia cell lines and synovial cells. More interestingly, in another study conducted by the same team, moesin was detected on the surface of THP-1 cells and various human blood cells, including T cells, natural killer cells and monocytes from healthy individuals, ²⁴ but not on B cells, PMNs or bone marrow CD34⁺ cells; however, they did not look at apoptotic PMNs. Moreover, as we described here, the presence of cell surface moesin is only detected with a population of PMNs that is close to 100% apoptotic. Antimoesin antibodies in the serum of patients with AA were found to stimulate peripheral blood mononuclear cell to secrete tumor necrosis factor-α and interferon-γ; the titer of antibodies was found to be dependent on the amount of cell surface moesin expressed by THP-1 cells. ²⁵

We conclude that the novel effect of VAA-I reported in the present study could represent an important potential toxic effect of this lectin. Whether or not moesin and other ERM proteins can be detected on the cell surface of apoptotic immune cells other than PMNs in response to VAA-I remained to be determined. Such information will certainly help for better understanding the mode of action of VAA-I.