Human Natural Immunoglobulin M Antibodies Induce Apoptosis of Human Neuroblastoma Cells by Binding to a M_r 260,000 Antigen¹ Free

NB,⁶ the most common extracranial solid neoplasm in infancy, is characterized by a high spontaneous regression rate of human malignant disorders (1, 2). Although several mechanisms including the action of natural killer cells and specific antibodies are discussed as responsible factors for this unusual phenomenon (3, 4), the mechanism of NB tumor regression remains unclear. In a recent study, we reported the presence of natural IgM antibodies cytotoxic for human NB cells in the sera of healthy humans (5). Approximately one-third of the sera exhibited considerable complement-mediated cytotoxicity (5). The incidence of high titers of the anti-NB IgM is even greater in healthy children than in healthy adults (6). The recognized antigen was identified as a M_r 260,000 surface protein (NB-p260) that appears to be expressed only on human NB cells because it could neither be detected on various normal cells nor on cancer cells from different origins such as melanoma, osteosarcoma, lymphoma, and various carcinomas (5). Interestingly, only negligible levels of natural anti-NB IgM antibodies were detected in the sera of NB patients (n = 11) with active disease (5). These observations suggest a role of anti-NB IgM in the biology of human NB. This is supported by our recent observation that natural human anti-NB IgM from healthy donors is capable of effecting growth arrest of even large solid human NB tumors in nude rats (7). Because only a small fraction of the injected total polyclonal human IgM represents specific anti-NB antibodies, the rather dramatic antitumor effect indicates the outstanding cytotoxic potential of these antibodies.

In this study, we show that natural human anti-NB IgM antibodies are very potent mediators of apoptotic cell death both in vitro and in vivo. The apoptotic response of human NB cells to these antibodies is mediated by binding to the NB-p260 surface protein as confirmed by inhibition experiments. On the basis of the observation that NB-p260 is expressed on all human NB cell lines analyzed thus far but not on other cancer cells from different origins, the induction of apoptosis by natural anti-NB-p260 IgM antibodies may represent one molecular mechanism involved in the spontaneous NB tumor regression.

Human NB cell lines were obtained from R. C. Seeger (LA-N-1; University of California, Los Angeles, CA; Ref. 8), N-K. V. Cheung (NMB-7; Memorial Sloan-Kettering Cancer Center, New York, NY; Ref. 9), and the American Type Culture Collection (IMR-32). The human embryonal rhabdomyosarcoma RD cell line and the human osteogenic sarcoma cell line TE-85 were also obtained from American Type Culture Collection. The human melanoma cell lines SK-MEL-93-2 and SK-MEL-170 have been described (10). All cells were cultivated in RPMI 1640 supplemented with 10% (v/v) heat-inactivated FCS, 100 IU/ml penicillin, and 100 μg/ml streptomycin (5). All cell culture reagents were obtained from Life Technologies, Inc. (Berlin, Germany).

Human sera exhibiting either >80% complement-mediated cytotoxicity against human NB cells (positive donors) or <20% cytotoxicity (negative donors) were used to purify the IgM fraction by a combination of size exclusion (Sephacryl S-300 HR; Amersham Pharmacia Biotech, Freiburg, Germany) and anion exchange chromatography (Macro-Prep High Q; Bio-Rad, Munich, Germany) as described (7). Using this procedure, IgM was purified to homogeneity as confirmed by SDS-PAGE. Purified IgM fractions were concentrated by ultrafiltration (Diaflo XM 300 membrane; Amicon, Witten/Ruhr, Germany) to ∼1.5 mg/ml and stored at 4°C after sterile filtration through a 0.2-μm pore size cellulose acetate filter. The term anti-NB IgM refers to IgM fractions from positive donors.

NB-p260 was isolated from extracts of LA-N-1 cells prepared by treatment of 1 × 10⁹ cells with 40 ml of 20 mm Tris-HCl, 150 mm NaCl, 5 mm EDTA, 1% (v/v) Triton X-100, 2 mm PMSF, 1.0 μg/ml leupeptin, and 1.0 μg/ml pepstatin (pH 8.3). After removal of insoluble components by centrifugation, the supernatant was applied to a 5-ml anion exchange chromatography column (Econo Q; Bio-Rad). The NB-p260 was eluted in the breakthrough with 20 mm Tris-HCl, 5 mm EDTA, 0.1% (v/v) Triton X-100, 2 mm PMSF, 1.0 μg/ml leupeptin, and 1.0 μg/ml pepstatin (pH 8.3). NB-p260-containing fractions were pooled and desalted on a PD-10 column (Amersham Pharmacia Biotech) equilibrated with 50 mm sodium phosphate, 0.1% (v/v) Triton X-100, 2 mm PMSF, 1.0 μg/ml leupeptin, and 1.0 μg/ml pepstatin (pH 6.5) and subsequently applied to a 5-ml cation exchange chromatography column (Econo S; Bio-Rad) equilibrated in the same buffer. The breakthrough containing the NB-p260 was concentrated by ultrafiltration (Diaflo Ultrafilter XM300; Amicon, Witten, Germany) and further purified to homogeneity by preparative SDS-PAGE. Preparative SDS-PAGE was performed with a dedicated instrument (Model 491 Prep Cell; Bio-Rad) using a 4% (w/v) polyacrylamide separating gel and following the manufacturer’s instructions (11). The NB-p260 was identified by Western blotting with anti-NB IgM during the purification procedure.

Purified NB-p260 was transferred by Western blot technique onto a nitrocellulose sheat. After staining with India ink, the bands corresponding to NB-p260 were cut out and subsequently sonicated in PBS. The suspension was mixed with an equal volume of either complete (first injection) or incomplete (subsequent injections) Freund’s adjuvant (Sigma Chemical Co., Deisenhofen, Germany), and ∼250 μl of each emulsion was injected i.p. into BALB/c mice (Harlan Winkelmann, Borchen, Germany). After two booster injections, the antisera were collected, and the IgG fractions were purified using a protein G column (Sigma). The binding specificity of antisera and IgG fractions was tested by Western immunoblot and cytofluorometry.

Chicken anti-NB-p260 antisera were developed as a custom service by Dr. J. Pineda (Antikörper-Service, Berlin, Germany) using purified NB-p260. The IgY fractions were purified as described (12) and characterized by Western immunoblot and cytofluorometry.

Binding assays of human IgM, murine IgG, or chicken IgY fractions to NB cells were performed as described (5). The secondary antibodies, including goat anti-human IgM (Fc5μ specific) conjugated to DTAF (1:50 dilution), goat anti-murine IgG conjugated to DTAF (1:50 dilution), and donkey anti-chicken IgY conjugated to DTAF (1:50 dilution), were obtained from Dianova (Hamburg, Germany). For competitive binding analysis, 5 × 10⁵ LA-N-1 cells were preincubated with 100 μl of anti-NB IgM for 45 min on ice. The cells were washed twice and subsequently resuspended in 100 μl of murine anti-NB-p260 IgG. After incubation for 45 min on ice, the cells were washed, resuspended in 100 μl of the DTAF-conjugated secondary goat anti-murine IgG antibody at a 1:50 dilution, and analyzed as described (5).

At subconfluency, LA-N-1 cells were detached and aliquots (2 × 10⁵ cells) in supplemented RPMI 1640 were incubated at 37°C in Permanox chamber slides (Nunc, Wiesbaden, Germany) in the presence of 125 μg of purified human IgM obtained from anti-NB IgM-positive or -negative donors (7). Unless otherwise indicated, the cells were incubated for 8 h and subsequently analyzed for apoptotic parameters as described below.

Microscopic analysis of apoptotic NB cells was performed by light microscopy or fluorescence microscopy after staining with Hoechst 33258 (13). Photographs were taken with Ektachrome 64T EPY and 320T EPJ artificial light films (Kodak, Stuttgart, Germany).

Quantification of apoptotic cells was performed by FACScan analysis (Becton Dickinson, Heidelberg, Germany) after staining with propidium iodide, propidium iodide and Annexin V, or monoclonal antibody APO2.7.

For propidium iodide staining, anti-NB IgM-treated cells were washed in PBS, fixed in 200 μl of ice-cold 70% (v/v) ethanol at 4°C for 30 min, and incubated with propidium iodide as described (14). Apoptotic cells were identified via their hypodiploid DNA content.

For Annexin V staining, anti-NB IgM-treated cells were washed in PBS, incubated with Annexin V-FITC (Annexin V-FITC kit; R&D Systems GmbH, Wiesbaden, Germany) for 15 min at room temperature, and analyzed for bound Annexin V as described (15). Annexin V-positive cells that did not take up propidium iodide were identified as early apoptotic, whereas Annexin V-positive cells that could also be stained with propidium iodide were classified as late apoptotic. If not indicated otherwise, the percentage of both cell populations was used to calculate the extent of apoptosis after subtraction of corresponding background values obtained with cells treated with IgM from negative donors.

Staining of NB cells with monoclonal antibody APO2.7 (Coulter-Immunotech, Hamburg, Germany) was performed as described (16). Briefly, NB cells were permeabilized with digitonin for 20 min on ice, washed, incubated with APO2.7 for 45 min, washed again, and then resuspended in PBS (pH 7.3) containing goat anti-mouse IgG-DTAF conjugate at a 1:50 dilution. After another 45 min on ice, the cells were examined in a flow cytometer by recording 2.5 × 10³ events.

Aliquots of 2 × 10⁶ LA-N-1 cells in a tissue culture flask were prelabeled with BrdUrd for 16 h at 37°C. Subsequently, aliquots of 5 × 10⁴ cells were transferred to a microtiter plate, incubated with purified human anti-NB IgM (100 μg/well) at 37°C for various periods of time, and analyzed for DNA fragmentation by ELISA according to a procedure of the manufacturer (Boehringer Mannheim, Mannheim, Germany). Briefly, BrdUrd-labeled cells were lysed in the presence of EDTA and Tween 20, which released cytoplasmic BrdUrd-labeled DNA fragments into the culture supernatant. After centrifugation, supernatants were analyzed using an anti-BrdUrd peroxidase-conjugate (Boehringer Mannheim).

Aliquots of 5 × 10⁴ LA-N-1 NB cells in a microtiter plate were incubated with purified human anti-NB IgM (100 μg/well) at 37°C. After various periods of time, the DNA was isolated and analyzed by agarose gel electrophoresis (17) or field inversion electrophoresis (18).

LA-N-1 cells (5 × 10⁵) were preincubated with 100 μl of murine anti-NB-p260 IgG for 45 min on ice. The cells were washed twice and subsequently incubated with 125 μg of human anti-NB IgM for 6 h at 37°C. Apoptosis was subsequently determined by staining with the monoclonal antibody APO2.7.

Nitrocellulose (1 cm², cut into small pieces) was incubated with purified NB-p260 (50 μg) in 0.2 ml of 50 mm Tris-HCl (pH 8.0) for 2 h at 4°C, washed with Tris-buffered saline, blocked with 10% (w/v) nonfat dry milk, washed again with Tris-buffered saline, and incubated with purified human anti-NB IgM (80 μg) for 2 h at 4°C. Nonbound IgM was used for the induction of apoptosis as described above. Control experiments were performed by incubating anti-NB IgM with nitrocellulose pretreated with nonfat dry milk only.

Nude rats (nu/nu; male), 4–6 weeks of age, were obtained from the breeding facility of the University Hospital Eppendorf. Animals were injected s.c. with 2 × 10⁷ LA-N-1 NB cells in the left flanks. Animals (n = 8) with tumors (0.5 × 10⁴ to 2.8 × 10⁴ mm³ in volume) received a single i.v. injection of 600 μg (corresponding to ∼3 mg/kg body mass) human IgM from a donor with a high titer of anti-NB IgM. Tumors were excised 24 h later, cryopreserved, and cut into serial sections for subsequent histological examination after staining with H&E (7). The control group of tumor-bearing animals (n = 8) did not receive any treatment.

DNA strand breaks in serial cryostat sections (4–6 mm) of human NB xenograft tumors fixed with acetone were identified in situ by labeling free 3′OH termini with fluorescein-dUTP in the presence of terminal deoxynucleotidyl transferase (TUNEL method) as described (19). Incorporated fluorescein was detected by an anti-fluorescein antibody conjugated to horseradish peroxidase (Boehringer Mannheim). Tissue sections were subsequently counterstained with hematoxylin.

SDS-PAGE, silver staining, and Western immunoblotting were performed as described (5).

The induction of apoptosis of NB cells by binding of natural anti-NB IgM was first investigated in vitro. After treatment with the purified IgM fraction from donors with high natural anti-NB activity (>80% cytotoxicity; termed anti-NB IgM), human LA-N-1 cells showed morphological signs of extensive apoptosis including nuclear fragmentation, chromatin condensation, membrane blebbing, and the formation of apoptotic bodies (Fig. 1). Control IgM from donors with negligible natural anti-NB activity (<20% cytotoxicity) did not induce morphological changes in these NB cells. Two other human NB cell lines (NMB-7 and IMR-32) included in this study showed identical morphological signs of apoptosis after treament with human anti-NB IgM, whereas no signs of apoptosis were observed with various other human tumor cell lines of other origin (not shown).

The microscopic observations could be confirmed by quantitative flow cytometry using a combined staining procedure with FITC-labeled Annexin V and propidium iodide. As shown in Fig. 2, treatment of LA-N-1 cells with human anti-NB IgM for 4 h yielded ∼40% early apoptotic cells, as indicated by positive staining with FITC-labeled Annexin V and negative staining with propidium iodide. Apoptosis could not be induced by anti-NB IgM in other human tumor cell lines of different origins including SK-MEL-93–2 melanoma cells and RD rhabdomyosarcoma cells (Fig. 2).

Time-dependent DNA fragmentation analyses revealed that the induction of apoptosis in human NB cells is a rapid process. Nuclear fragmentation became apparent as early as 60–120 min after the addition of natural anti-NB IgM to LA-N-1 cells and reached a plateau after ∼8 h (Fig. 3,A). At that time, ∼90% of the LA-N-1 cells showed typical apoptotic changes in the light microscope. Dose-response analysis demonstrated that the extent of apoptosis is dependent on the concentration of natural anti-NB IgM (Fig. 3 B).

Analyses by agarose gel electrophoresis (Fig. 4,A) and field-inversion electrophoresis (Fig. 4 B) revealed the generation of high molecular weight DNA fragments after treatment of human LA-N-1 cells with natural anti-NB IgM. Such high molecular weight DNA fragments were not present in untreated control cells and after treatment of LA-N-1 cells with the IgM fraction from a negative donor. Oligonucleosome-length DNA fragments could not be detected, even after extended periods of incubation (up to 16 h) of the cells with high concentrations of anti-NB IgM (up to 150 μg/well).

To demonstrate that the NB-p260 antigen is indeed responsible for the mediation of apoptosis, NB-p260 was purified to homogeneity from LA-N-1 cells by sequential ion exchange chromatography, followed by preparative SDS-PAGE (Fig. 5, Lane 1) and used to generate antisera in mice and chicken. After isolation of IgG fractions from the murine and IgY fractions from the chicken antisera, respectively, the specificity of the antibodies for NB-p260 was verified by immunoblot (Fig. 5, Lanes 4 and 6).

The chicken anti-NB-p260 IgY antibodies were used in cytofluorometric binding studies to assess the expression of NB-p260 on the NB cell lines included in this study (LA-N-1, IMR-32, and NMB-7) and on several human control tumor cell lines derived from different organs. Significant binding of the anti-NB-p260 IgY antibodies was observed with all NB cell lines, whereas the control cell lines exhibited only very little or background binding (Fig. 6,A). The binding pattern of chicken anti-NB-p260 IgY corresponded to that observed with natural human anti-NB IgM (Fig. 6 B).

The murine anti-NB-p260 IgG antibodies were used for a competitive cytofluorometric binding assay with human anti-NB IgM. As is evident from Fig. 7, pretreatment of LA-N-1 cells with human anti-NB IgM reduced the binding of murine anti-NB-p260 IgG to background values obtained with murine preimmune IgG.

The murine anti-NB-p260 IgG antibodies were also tested for their effects on NB cell apoptosis. Incubation of LA-N-1 cells with murine anti-NB-p260 IgG did not induce apoptosis. However, when LA-N-1 cells were preincubated with murine anti-NB-p260 IgG and subsequently treated with natural anti-NB IgM, the extent of anti-NB IgM-induced apoptosis was significantly decreased as compared with the extent of apoptosis after preincubation with murine preimmune IgG (Fig. 8). The inhibition of apoptosis is not only evident by a reduced absolute number of apoptotic cells (18% versus 44%) but also by a reduced extent of expression of the apoptotic antigen 7A6 recognized by the APO2.7 antibody (mean channel 38 versus mean channel 131).

To further corroborate the role of NB-p260 in serving as apoptosis-inducing receptor on human NB cells, natural human anti-NB IgM was preadsorbed with purified NB-p260 and subsequently tested for its ability to induce apoptosis of human NB cell lines. As shown in Fig. 9,A, in all three NB cell lines the induction of apoptosis could be completely abolished by preadsorption of human anti-NB-IgM with purified NB-p260. Identical results were obtained with natural anti-NB IgM from three different donors (Fig. 9 B). Collectively, these data demonstrate that NB-p260 serves as the apoptosis-mediating receptor on the surface of human NB cells for natural human anti-NB IgM.

On the basis of the potent induction of apoptosis by natural anti-NB IgM in vitro, the capability of these antibodies to mediate apoptosis in vivo was investigated. Nude rats bearing large human xenograft NB tumors (approximately 1.5–3 cm in diameter) received a single i.v. human anti-NB IgM injection through the tail vein. Tumors were excised 24 h later and cryopreserved for subsequent analyses. Histological analyses of all xenografted NB tumors after anti-NB IgM injection (eight animals) revealed large areas of pyknotic cells with condensed or fragmented nuclei, a typical picture of an apoptotic process (Fig. 10,A). Tumors of control animals showed unaltered cellular morphology (Fig. 10,B). The presence of apoptotic cells was verified by staining of serial tumor sections of natural anti-NB IgM-treated and untreated animals using the TUNEL method. Anti-NB IgM-treated tumors contained many areas of cells that stained positive with TUNEL (Fig. 10, C and E). The remaining normal cells within the treated tumors were often surrounded by an extended extracellular space. In contrast, tumors of untreated animals contained only few regions of weakly staining cells (Fig. 10 D). These data indicate that natural human anti-NB IgM given i.v. induces apoptosis of NB tumors in vivo within 24 h.

In this study, we demonstrate that sera of healthy individuals contain IgM antibodies capable of inducing apoptotic cell death in human NB tumor cells both in vitro and in vivo. The apoptosis was demonstrated in vitro by morphological changes such as nuclear fragmentation, chromatin condensation, membrane blebbing, and the formation of apoptotic bodies. The biochemical evidence of the apoptotic process included the cytofluorometric demonstration of phosphatidylserine expression on the cell surface as determined by Annexin V binding, propidium iodide uptake, and expression of the apoptotic antigen 7A6 as measured by the binding of the APO2.7 antibody. Furthermore, DNA fragmentation was shown by ELISA, agarose gel electrophoresis, and field-inversion electrophoresis. The anti-NB IgM-induced apoptotic pathway generates high molecular weight DNA fragments and not oligonucleosome-length DNA fragments as observed in most studies (20, 21). However, the generation of high molecular weight DNA fragments during apoptosis has also been reported for other cellular systems (22, 23, 24).

Anti-NB IgM-induced apoptosis was also demonstrated in vivo using nude rats with human NB tumor xenografts. After injection of anti-NB IgM, large areas of apoptotic cells with condensed or fragmented nuclei were seen. The presence of apoptotic cells was further demonstrated by positive TUNEL staining.

The apoptotic process is relatively fast. In vitro kinetic analysis revealed up to 90% of apoptotic cells within 8 h, which corresponds to the kinetics of apoptosis as reported for other systems (25, 26). Similarly, the apoptotic process was also rapid in vivo, where apoptotic changes could be demonstrated within 24 h of anti-NB IgM injection.

The surface protein NB-p260 serves as the apoptosis-mediating receptor on the surface of human NB cells for human anti-NB IgM antibodies. This could be demonstrated by inhibition of anti-NB IgM-induced apoptosis by both murine anti-NB-p260 IgG and the purified NB-p260 protein. Preadsorption of anti-NB IgM with the purified NB-p260 antigen virtually abolished the induction of apoptosis in all human NB cell lines included in this study, indicating that NB-p260 is the predominant, if not the only antigen on human NB cells recognized by human anti-NB IgM. The generality and significance of these data are demonstrated by the fact that natural anti-NB IgM from different donors yielded identical results. We certainly cannot exclude the possibility that on other cells, different antigens may serve as apoptosis-inducing receptors for natural human IgM. However, the lack of binding of human IgM to normal and tumor control cell lines used in this study as well as in our previous work (5) indicates that apoptosis by natural human IgM is certainly not a frequent phenomenon and, therefore, appears to be of particular importance for NB. In addition to the induction of apoptosis, we have shown previously that the anti-NB IgM antibodies exhibit potent complement-mediated cytotoxicity, both in vitro and in vivo (5, 7). We have also shown that treatment of anti-NB IgM of nude rats with human NB xenografts leads to infiltration of neutrophils into the NB tumors (7). The neutrophil accumulation is independent of complement (7) and is likely to be a third cytotoxic mechanism induced by anti-NB IgM (27, 28, 29). The relative importance of these three cytotoxic mechanisms for the in vivo cytotoxic effect of anti-NB IgM is presently unknown.

The biological function of naturally occurring anti-NB IgM antibodies remains to be elucidated. On the basis of the presence of significantly lower levels of these antibodies in NB patients (5, 6), these antibodies could represent a defense against NB tumor cells. This would be in accordance with the observation that human NB shows a very high spontaneous regression rate compared with other human malignancies (1). Apoptotic cells have indeed been demonstrated in NB tumor specimens (30). On the other hand, such a function raises several questions. Why does a high percentage of human sera from healthy individuals contain anti-NB IgM antibodies, and what are the mechanisms that cause the formation of these antibodies with specificity for NB, an otherwise rather rare tumor? Additional work will need to answer these questions. At present, we are investigating the structure of the NB-p260 surface protein and the apoptotic signaling pathway. In addition, we are evaluating the potential of the natural anti-NB IgM antibodies for immunotherapy of NB in a Phase I clinical trial (31).