IVIg - The Journal of Immunology

Therapeutic Preparations of Normal Polyspecific IgG (IVIg) Induce Apoptosis in Human Lymphocytes and Monocytes: A Novel Mechanism of Action of IVIg Involving the Fas Apoptotic Pathway1 Nagendra K. A. Prasad,2,3* Giuliana Papoff,2† Ann Zeuner,4† Emmanuelle Bonnin,* Michel D. Kazatchkine,* Giovina Ruberti,5† and Srini V. Kaveri5* Therapeutic preparations of normal human IgG for i.v. use (IVIg) exhibit a broad spectrum of immunoregulatory activities in vitro and in vivo. IVIg has been shown to inhibit the proliferation of activated B and T lymphocytes and of several autonomously growing cell lines. In this study, we demonstrate that IVIg induces apoptosis in leukemic cells of lymphocyte and monocyte lineage and in CD40-activated normal tonsillar B cells, involving, at least in part, Fas (CD95/APO-1) and activation of caspases. IVIginduced apoptosis was higher in Fas-sensitive HuT78 cells than in Fas-resistant HuT78.B1 mutant cells, and soluble Fas inhibited IVIg-induced apoptosis. IVIg immunoprecipitated Fas from Fas-expressing transfectants and recognized purified Fas/glutathioneS-transferase fusion proteins upon immunoblotting. Affinity-purified anti-Fas Abs from IVIg induced apoptosis of CEM T cells at a 120-fold lower concentration than unfractionated IVIg. Inhibitors of cysteine proteases of the caspase family, caspase 1 (IL1b-converting enzyme) and caspase 3 (Yama/CPP32b), partially inhibited IVIg-induced apoptosis of CEM cells. Furthermore, cleavage of poly(A)DP-ribose polymerase into an 85-kDa signature death fragment was observed in CEM cells following IVIg treatment. Thus, normal IgG induces apoptosis in lymphocytes and monocytes. Our results provide evidence for a role of Fas, bring new insights into the mechanisms of action of IVIg in autoimmune diseases, and suggest a role of normal Ig in controlling cell death and proliferation. The Journal of Immunology, 1998, 161: 3781–3790.

I

n addition to substitutive treatment of patients with primary and secondary Ab deficiencies, therapeutic preparations of normal polyspecific Ig G (i.v. Ig (IVIg)6) are used in a large number of immune-mediated conditions, including acute and chronic/relapsing autoimmune diseases and systemic inflammatory disorders (1– 4). Several mutually nonexclusive mechanisms of action have been proposed to account for the immunoregulatory ef-

*Institut National de la Sante´ et de la Recherche Me´dicale U430, and Universite´ Pierre et Marie Curie, Hopital Broussais, Paris, France; and †Institute of Cell Biology, National Research Council, Rome, Italy Received for publication October 16, 1997. Accepted for publication June 2, 1998. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by a grant from Hyland Division of Baxter Healthcare Corporation (Duarte, CA) and by Institut National de la Sante´ et de la Recherche Me´dicale, Centre National de la Recherche Scientifique, France, Associazione Italiana per la Ricerca sul Cancro, Italian National Institute of Health, AIDS (N. 9403-98), and CNR “Progetto Strategico,” Italy. N.K.A.P. was supported by a fellowship from French Association pour la Recheche contre le Cancer. G.P. was supported by a fellowship from Fondazione Italiana Ricerca sul Cancro. 2

The first two authors have contributed equally to this work.

3

Current address: Department of Hematology-Oncology, Istituto Superiore di Sanita`, Rome, Italy. 4

Current address: UCSF Cancer Center, 2340 Sutter St., San Francisco, CA 94115.

5

Address correspondence and reprint requests to Dr. Giovina Ruberti, Institute of Cell Biology, National Research Council, Via E. Ramarini 32, 00016 Monterotondo Scalo (Rome), Italy. E-mail address: [email protected], or Dr. Srini V. Kaveri, INSERM U430, Hopital Broussais, 96 rue Didot, 75014 Paris, France. E-mail address: [email protected] 6 Abbreviations used in this paper: IVIg, i.v. IgG; ECL, enhanced chemiluminescence; GST, glutathione-S-transferase; ICE, IL-1b-converting enzyme; MTT, 3-(4,5dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide; PARP, poly(A)DP-ribose polymerase; tTA, tetracycline-controlled transactivator.

Copyright © 1998 by The American Association of Immunologists

fects of IVIg (2, 5– 8). Recent work has emphasized the capacity of IVIg to regulate the activation and functions of monocytes, B cells, and T cells in vitro and to contribute to the selection of immune repertoires in vivo (6, 7, 9 –12). The observation that IVIg inhibits the proliferation of in vitro activated B and T lymphocytes (13) and of autonomously growing human cell lines (14) suggests that it interferes with the expression of genes and/or function of the gene products involved in cell growth and cell death. Several experimental and human autoimmune diseases are associated with altered regulation of cell proliferation or apoptosis (15–20). In this study, we report that IVIg induces cell death in human monocytic and lymphoblastoid cell lines and CD40-activated normal human B lymphocytes in vitro, and that the cell death is associated with nucleosomal cleavage of cellular DNA and the expression of phosphatidylserines on the cell surface, an early event of apoptosis occurring before membrane disruption (21). We demonstrate that IVIg-induced apoptosis of lymphocytes is dependent on Fas and on activation of the caspase family of proteases. We further show the presence of anti-Fas Abs in IVIg that can efficiently induce apoptosis upon affinity purification. Taken together, these observations provide new insights into the mechanisms of action of IVIg in autoimmune diseases and suggest a role of normal circulating Ig in controlling cell proliferation.

Materials and Methods Human Igs Intravenous Ig preparations were Gammagard (Baxter Hyland, Glendale, CA) and Sandoglobulin (Novartis, Basel, Switzerland). When reconstituted for therapeutic use, Gammagard contains 50 mg/ml IgG, 40 mg/ml glucose, 4 mg/ml polyethylene glycol, 0.6 M glycine, 6 mg/ml human serum albumin, and 0.15 M NaCl (osmolality 655 mOsmol). Sandoglobulin contains 50 mg/ml IgG, 25 to 35 mg/ml sucrose, 6 to 10 mg/ml glucose, and 40 to 0022-1767/98/$02.00

3782

INDUCTION OF APOPTOSIS BY NORMAL HUMAN Ig 37°C. GST fusion proteins were purified from bacterial lysates with glutathione-Sepharose 4B, according to the manufacturer’s instructions (Pharmacia).

Immunoaffinity purification of anti-Fas Abs from IVIg Purified human extracytoplasmic Fas-GST fusion protein was coupled to CNBr-activated Sepharose 4B (Pharmacia). A total of 50 mg of IVIg was allowed to interact with 1 ml of affinity gel in PBS, pH 7.4, on a rocking platform overnight at 4°C. After extensive washing, bound IgG was eluted using 0.2 M glycine-HCl, pH 2.8. Eluted fractions were immediately brought to pH 7 with 2 M Tris and dialyzed against PBS and once against serum-free RPMI 1640.

Cell lines and transfectants

FIGURE 1. Loss of cell viability induced by normal polyspecific IgG. CEM, Raji, and MM6 cells were cultured (5 3 104 cells/well in 96-well plates) for 24 h before the addition of increasing amounts of IVIg (Ig) or IVIg that had been dialysed against RPMI 1640 (Dial. Ig). Cell viability was assessed using the MTT assay after 12 h of treatment. The absorbance values of control untreated cells were standardized as 100%, and the viability values of test samples are expressed as percentage of untreated control cells. The figure depicts the mean values 6 SE obtained in two independent experiments conducted in duplicates.

100 mM NaCl (osmolality 515 mOsmol). For the present study, a stock solution of 100 mg/ml (0.6 mM) of either IVIg was prepared in serum-free RPMI 1640 containing L-glutamine, 50 U/ml penicillin, and 50 mg/ml streptomycin. IVIg was dialyzed twice against large volumes of serum-free RPMI 1640 at 4°C before use in MTT assay (Fig. 1). F(ab9)2 of IVIg were prepared by pepsin digestion and chromatography on protein G-Sepharose (Pharmacia, Uppsala, Sweden), and stock solutions (0.6 mM) were made in serum-free RPMI 1640 medium. The human IgG1k myeloma Prez. was obtained and purified as described (22).

Abs and reagents The DX2 IgG1 murine anti-Fas mAb was kindly provided by Dr. Testi (University of Rome “Tor Vergata,” Italy). Anti-Fas Ab CH-11 (IgM) was purchased from Upstate Biotechnology (Lake Placid, NY). FITC-conjugated mAbs to CD3, CD14, and CD19, and phycoerythrin-conjugated mAb to CD95 (7C11) were purchased from Immunotech (Marseilles, France). Anti-CD40 Ab was obtained from PharMingen (San Diego, CA). Polyclonal rabbit Abs to Fas, FasN18, and FasC20 were from Santa Cruz Biotechnology (Santa Cruz, CA). FasN18 is specific for an epitope of Fas corresponding to amino acids 21 to 38 mapped to the amino terminus of Fas, and FasC20 recognizes an epitope mapped to amino acids 316 to 335 at the carboxyl terminus of the molecule. GST-specific Ab was from Pharmacia. mAb C-2-10 directed against poly(A)DP-ribose polymerase (PARP) was obtained from Dr. Poirier (University of Laval, Ste-Foy, Canada). Human serum albumin was obtained from Laboratoire Franc¸ais des Biotechnologies (LFB, les Ulis, France). The peptide inhibitors of ICE, Ace-YVAD-cho, and of Yama, AceDEVD-cho, were obtained from Neosystem Laboratories (Strasbourg, France). The peptides were dissolved in serum-free RPMI 1640 to a final concentration of 20 mM. Leupeptin and the calpain inhibitor E64 were from Sigma (St. Louis, MO).

Soluble Fas molecules The DAP.3 and DAP.3 Fas-soluble variant-transfected cells were grown in DMEM containing 10% FCS (23). Supernatants containing Fas-soluble proteins were produced and assayed by an ELISA, as previously described (23).

GST fusion proteins The cDNA sequences encoding the extracytoplasmic and intracytoplasmic regions of human Fas were cloned into the EcoRI and XhoI sites of pGEX4T-1 in frame with the open reading frame of GST and expressed in BL21 by induction with 0.1 mM isopropyl-b-D-thiogalactopyranoside for 4 h at

The human T cell line CEM, B lymphoblastoid line Raji, and the promonocytic cell line MM6 were maintained in RPMI 1640 supplemented with 1% L-glutamine, 50 U/ml penicillin, 50 mg/ml streptomycin, and 10% FCS (HyClone). The Fas-sensitive human T cell lymphoma line HuT78 and the Fas-resistant subclone HuT78.B1 (24) were maintained in RPMI 1640 medium supplemented with 5% FCS. The Fas BW transfectant cell lines were established by using the tetracycline-controlled transactivator (tTA) system (25). The pUHD15-1 neo and the pUHD10-3 (26) plasmids were provided by Dr. Bujard (Heidelberg, Germany). The full-length coding region of the human Fas cDNA with a point mutation, Val/Asn238, in the intracellular domain (F3) with the addition of a FLAG (F) epitope tag at the N terminus was cloned into the XbaI site of pUHD10-3 vector (pUHFF3). Such mutation is analogous to an identified mutant allele of Fas in the cg strain of lpr autoimmune mice, which is deficient in Fas-mediated induction of apoptosis (27). The FF3 cDNA was produced by PCR with the following primers: HFas59 (GGC GATCTAGAATGCTGGGCATCTGGACCCT; the XbaI site is underlined) and GR91 (CTTGTCATCGTCGTCCTTGTAGTCTTTGGACG ATAATCTAGC; the sequence coding for the Flag peptide is underlined); GR108 (AAAGACTACAAGGACGACGATGACAAGAGTGTTAATG CCCAAGTG; the sequence coding for the Flag peptide is underlined) and HFasRev (GGCGATCTAGACCAAGCTTTGGATTTC; the XbaI site is underlined) using pcDNA3FP1 plasmid as template. pcDNA3FP1 was also constructed by PCR essentially as previously described (27). Plasmid pUHD15-1 neo linearized with ScaI (5 mg) was transfected by electroporation. Clones were selected in the presence of 800 mg/ml of active G418 (Geneticin; Life Technologies) and tested for their ability to induce, in transient transfection experiments with the pUHG16-3 plasmid (26) tetracycline-regulated expression of b-galactosidase. One such clone, BWtTA, was selected as recipient for the pUHFF3. A linearized pUHFF3 plasmid (5 mg) was transfected together with the pBABEPuro vector (0.3 mg) (28) into BwtTA. Clones were selected against puromycin and in the presence of tetracycline. However, we were unable to isolate clones with tetracycline-inducible expression of FF3. A clone that constitutively expressed high levels of Fas protein, as assessed by cytofluorometry using the DX2 anti-Fas Ab, was selected and used in all experiments. The cells were maintained in RPMI 1640 with the addition of nonessential amino acids, sodium pyruvate, and 0.1 mM b-mercaptoethanol.

B cell preparations Normal B cells were purified from tonsils, as described previously (29). T cells were eliminated by rosetting with SRBC. Purified B cell preparations contained .90% B cells and ,7% T cells and monocytes, as determined by cytofluorometry using fluorochrome-conjugated anti-CD19, anti-CD3, and anti-CD14 mAbs. Ligation of B cells with CD40 was conducted by incubating the cells with soluble anti-CD40 mAb (1 mg/ml) for 18 h under culture conditions, as described above.

Assays for apoptosis To perform the MTT assay (colorimetric cell viability assay based on conversion of tetrazolium salts to formazan crystals by mitochondrial enzymes), target cells that had been cultured in the presence of inducers of apoptosis were interacted with the MTT reagent (Boehringer Mannheim, Mannheim, Germany) before extraction using an SDS/HCl solution. Absorbance was recorded at 490 to 650 nm using an ELISA reader, and values normalized by subtracting the background corresponding to samples in the absence of cells. The viability data were expressed as percentage of control cells cultured in the absence of apoptosis-inducing agents. For measuring propidium iodide (PI) uptake, cells that had been cultured in the presence of apoptosis-inducing agents in RPMI 1640 containing 10% FCS were harvested by centrifugation at 750 3 g, washed in

The Journal of Immunology

3783

ice-cold PBS before resuspending the pellet in PBS containing 0.02% sodium azide (PBS-azide). PI (50 mg/ml) was then added, and the dye uptake was analyzed by fluorescence analysis for red fluorescence (30). Analysis of DNA fragmentation was performed following extraction of DNA and agarose gel electrophoresis, as described (31). In brief, cells were harvested by centrifugation, washed with PBS, and lysed in lysis buffer containing 20 mM Tris, pH 7.4, 0.4 mM EDTA, and 0.4% Triton X-100 (500 ml/2 3 106 cells). Solubilized cells were centrifuged at 10,000 3 g for 5 min. DNA fragments in the supernatants were precipitated overnight with 0.5 M NaCl and an equal volume of isopropanol at 270°C. Samples were thawed, centrifuged at 10,000 3 g, and washed with 70% ethanol. Dried pellets were resuspended in 10 to 20 ml of 13 TE (10 mM Tris.HCl, pH 7.4, and 1 mM EDTA) containing 0.1 mg/ml RNase and incubated at 37°C for 30 min. DNA samples from equivalent number of cells were then electrophoresed with an appropriate volume of 103 loading buffer (50% glycerol, 10% bromophenol blue, and 1% xyalene cyanol in 13 TE) in 0.8% agarose minigels for 1 h before revelation with ethidium bromide. Annexin V labeling was conducted by staining cells with Annexin VFITC (Bender Medsystems, BioWhittaker, Gagny, France) (2.5 mg/ml) for 30 min on ice. As negative control, duplicate samples were incubated in staining buffer without CaCl2 instead containing 2 mM EGTA. Cells were washed with 13 PBS-azide with 2 mM CaCl2 (for negative control samples, PBS-azide with 2 mM EGTA was used) and resuspended in PBS-azide before adding PI (50 mg/ml) to each sample. Samples were analyzed for green fluorescence (Annexin V labeling) and for red fluorescence (PI uptake) using a FACScan flow cytometer (Becton Dickinson, Mountain View, CA).

Surface biotinylation and immunoprecipitation of Fas Biotinylation of surface proteins was conducted as described (32). Briefly, 107 cells were washed with PBS and resuspended in 1 ml PBS, pH 8.5. D-biotinyl-e-amido caproic acid N-hydroxysuccinimide ester (Boehringer Mannheim) was dissolved in DMSO and incubated with the cell suspension at 50 mg/ml final concentration. The reaction was terminated by the addition of 10 mM NH4Cl. Fifty milliliters of 50 mM Tris/HCl, pH 7.4, 25 mM KCl, 5 mM MgCl2, and 1 mM EGTA were added; the cells were centrifuged and lysed in 1 ml lysis buffer (50 mM Tris/HCl, pH 8, 150 mM NaCl, 1% Nonidet P-40, 1 mM PMSF, aprotinin (2 mg/ml), and leupeptin (2 mg/ml)) for 20 min on ice. A total of 5 mg of DX2 Ab or 100 mg of preadsorbed IVIg (obtained by incubating 4 3 106 cells BWtTA cells with 800 mg of IVIg in RPMI 1640 for 12 h at 37°C) was added to lysates (300 –500 mg) and rotated for 90 min at 4°C. Beads were washed once in 13 lysis buffer, once in 0.13 lysis buffer, and in water. Immune complexes captured on beads were boiled in Laemmli’s buffer and loaded onto a reducing 10% SDS-PAGE gel. Immunoprecipitated biotinylated surface proteins were transferred to a Hybond nitrocellulose membrane, incubated with streptavidin horseradish peroxidase, and developed using ECL chemiluminescence system (Amersham, Arlington Heights, IL). GST-Fas fusion proteins were separated on a 10% SDS-PAGE, transferred to nitrocellulose membrane, and detected with N18 (0.5 mg/ml) and C20 (0.5 mg/ml) anti-Fas Abs (specific respectively for the extracytoplasmic and intracytoplasmic domains of Fas), preadsorbed IVIg (100 mg/ml), antiGST Ab (1 mg/ml) (Pharmacia), followed by protein A-horseradish peroxidase and ECL (Amersham).

Western blot analysis for PARP Cell extracts were analyzed on 8% SDS-PAGE, as described (33). Separated proteins were transferred onto nitrocellulose membranes, blocked for 1 h in PBS-MT (PBS with nonfat dried milk 5% and 0.1% Tween-20), and incubated overnight with anti-PARP mouse mAb C2-10. Membranes were washed with PBS-MT, incubated with secondary goat anti-mouse IgG (The Jackson Laboratory, Bar Harbor, ME) before detection using the ECL chemoluminescence system (Amersham).

Results Normal polyspecific IgG induces cell death in autonomously growing cell lines Cells of the promonocytic cell line MM6, the T cell line CEM, and the B cell line Raji were cultured in RPMI 1640/10% FCS for 24 h before the addition of IVIg at increasing concentrations and further culturing the cells for 12 h. Viability of the cells was assessed using the MTT assay. A dose- and time-dependent loss of viability was observed in the presence of IVIg in all three cell lines tested (Fig. 1 and data not shown). In 12 h, there was a 50 to 60%

FIGURE 2. PI uptake by IVIg-treated cells. Left panels, Cells were cultured in the presence of 50 mg/ml (0.3 mM) IVIg, black squares; 30 mg/ml (0.18 mM) IVIg, open circles; or in the absence of IVIg (open squares) for indicated durations. Cell death was quantitated by measuring the uptake of PI and expressed as percentage of dye-positive cells. Right panels, Cells cultured with equimolar amounts (0.3 mM) of IVIg (filled bars), F(ab9)2 fragments of IVIg (open bars), human serum albumin (vertical lines), medium alone (slanted hatches), or myeloma Ig (horizontal lines) for indicated durations.

reduction in viability in all three cell types, following incubation with 50 mg/ml of IVIg. No significant reduction in cell viability was observed with equimolar amounts of human serum albumin or 50 mg/ml of control myeloma protein (not shown). A similar loss of cell viability was observed with IVIg of either of the two sources used (Sandoglobulin and Gammagard) (not shown), and with IVIg that had been dialyzed against RPMI 1640 before incubation with the cells, ruling out a role for the stabilizers in the IVIg preparations in induction of cell death (Fig. 1). The induction of cell death by IVIg was further documented by measuring the uptake of PI by the cells, followed by FACS analysis (Fig. 2). Cell death was detected within 6 h of IVIg treatment. Of the three lines tested, CEM T cells were most sensitive to IVIg-mediated cell death. F(ab9)2 fragments of IVIg induced between 35 and 50% mortality in the cell lines under conditions in which equimolar amounts of intact IVIg induced more than 70% cell death (Fig. 2). Human serum albumin and myeloma protein did not induce significant loss of cell viability. Cell death induced by normal polyspecific IgG is apoptosis IVIg-induced cell death was associated with characteristic DNA ladder formation in the three cell lines that we tested (Fig. 3). DNA fragmentation was due to oligonucleosomal cleavage, as evidenced by the appearance of DNA fragments of approximately 200 bases

3784

INDUCTION OF APOPTOSIS BY NORMAL HUMAN Ig

FIGURE 3. DNA fragmentation of IVIgtreated cells. DNA isolated from cells treated with 50 mg/ml (0.3 mM) IVIg (indicated by 1) or untreated cells (indicated by 2) for 24 h were analyzed by agarose gel electrophoresis, as described in Materials and Methods. m.w. markers are indicated as MW. The arrowhead indicates the migration of the 565-bp marker. Lanes from left to right for CEM cells and from right to left for Raji cells represent increasing amounts of DNA loaded.

and its multiples. DNA fragmentation upon activation of endonucleases is known to be associated with apoptosis. Kinetics of induction of DNA fragmentation differed with the cell line and correlated with the kinetics observed by dye uptake analysis (data not shown). DNA fragmentation was detectable, within 4 h in CEM cells (not shown), 24 h in MM6 cells, and 36 h in Raji cells, of IVIg treatment. Exposure of phosphatidylserines on the cell surface is an early event in apoptosis and precedes dye uptake due to membrane damage (21, 34). Annexin V specifically binds to phosphatidylserine in a calcium-dependent manner, and may consequently be used for the detection of early apoptotis. Loss of membrane integrity, a late event in apoptosis, occurs in necrosis as well, leading to uptake of nonvital dyes. Therefore, dye-negative cells with an intact membrane that bind annexin V are considered to be in early apoptosis. On the other hand, late apoptotic cells and necrotic cells will be positive for both annexin V binding and dye uptake. Figure 4 depicts the double staining with annexin V-FITC and PI of cells that had been cultured in the presence of 50 mg/ml IVIg for 12 h. All three cell lines tested demonstrated two major population of cells: one that contained dye-negative/annexin V-positive (early apoptosis) cells, which represented 39% of MM6 cells, 8% of Raji cells, and 17.5% of CEM cells; another that contained double-positive (late apoptosis or necrosis) cells, which represented 11% of MM6 cells, 17% of Raji cells, and 61% of CEM cells. The kinetics of apoptosis differed with the cell type, and the relative proportion of dye-negative/annexin V-positive cells varied with the duration of IVIg treatment. The kinetics of annexin V labeling indicated that cells are progressing through dye-negative/annexin V-positive (early apoptosis) to double-positive (late apoptosis or necrosis) status (data not shown). We thus interpret that the double-positive cells in Figure 4 correspond to late apoptotic and not to necrotic cells, since these cells, after prolonged IVIg treatment, displayed characteristic DNA ladder formation, as exemplified in Figure 3. A role for Fas in IVIg-induced apoptosis Cytokines of the TNF ligand family and their cognate receptors such as TNFR-1 and Fas (also known as APO-1 or CD95) trigger apoptosis (35, 36). Fas is constitutively expressed on a variety of tumor cell lines of hemopoietic and nonhemopoietic origin (35). Fas cross-linking induced by its natural ligand or by agonistic Abs induces apoptosis of Fas-positive cells (37– 41). Thus, we inves-

tigated the possibility that Fas was at least partially responsible for the apoptosis observed upon IVIg treatment by using two experimental approaches. First, Fas-sensitive and Fas-resistant cell lines were treated with IVIg, and cell death was evaluated. HuT78 is a human T cell lymphoma line, highly sensitive to Fas-mediated apoptosis, while the HuT78.B1 cells are resistant to Fas-mediated apoptosis due to the expression of a wild-type and a truncated Fas receptor (24). We observed a dose-dependent induction of apoptosis of HuT78 cells upon treatment with IVIg for 16 h, as measured by means of PI uptake and by using the MTT assay (Fig. 5 and data not shown). The extent of apoptosis in HuT78 cells was lower than that observed in the established lines described above, when PI uptake, rather than the MTT assay, was used to assess the mortality. Using both of the approaches, we observed that IVIg-mediated apoptosis in Fas-resistant HuT78.B1 cells was significantly lower than that in the parental Fas-sensitive HuT78 cells (Fig. 5A and data not shown). As a second approach, soluble Fas molecules that block Fas-mediated apoptosis induced by Fas agonistic Ab and Fas ligand recombinant protein (23) were used along with IVIg, and the outcome on apoptosis was studied. Supernatants containing Fas molecules were obtained from cultures of DAP.3 fibroblast clones transfected with expression vectors encoding soluble Fas molecules (FasDExo3, 4 and FasDExo4). Negative control supernatants were harvested from DAP.3 fibroblast clones transfected with the vector alone. CEM cells were cultured in the presence of IVIg (50 mg/ml) and different dilutions of the supernatants in RPMI 1640 for 6 h. Cell death was measured by PI uptake. Soluble Fas molecules inhibited up to 50%, of IVIg-mediated apoptosis, as compared with negative controls, providing further evidence for involvement of Fas in IVIg-induced cell death (Fig. 5). Anti-Fas Abs in IVIg To investigate the presence of Abs with Fas specificity in IVIg, we used a Fas transfectant mouse BW cell line that overexpresses the human Fas molecule. Stable transfectants were generated in the BW murine T cell line, using a cDNA of Fas with a point lpr mutation in the cytoplasmic tail, as detailed in Materials and Methods. Clone BWtTAFF3 that expressed high levels of surface Fas Ag was selected for further analysis (Fig. 6A). BWtTAFF3 and parental BWtTA cells were surface biotinylated and subjected to immunoprecipitation with Fas mAb, DX2, or IVIg. Both sources of Ig immunoprecipitated a unique 48-kDa protein band (Fig. 6B,


3785

FIGURE 4. Staining of apoptotic cells with Annexin V. Cells were treated with IVIg (50 mg/ml; 0.3 mM) for 12 h and stained with Annexin V-FITC (FL-1) and PI (FL-2). The percentage population of cells positive for PI uptake alone, double positive for PI and Annexin V, and positive for Annexin V alone are depicted for each cell line. Annexin V staining was performed in the presence (left and middle panels) or in the absence of CaCl2 (right panel), as described in Materials and Methods.

lanes 2 and 4) that was not detected in parental BWtTA cells (Fig. 6B, lanes 1 and 3). We then immunoblotted recombinant Fas-GST fusion proteins using IVIg that had been preadsorbed with BWtTA cells. As shown in Figure 7, IVIg recognized both Fas extracytoplasmic and Fas intracytoplasmic GST fusion proteins (Fig. 7A; lanes 1 and 2), but not GST-Jun or GST control proteins (Fig. 7A; lanes 3 and 4). Anti-Fas mAbs N18 and C20 blotted the extra- and intracytoplasmic Fas-GST proteins, respectively (Fig. 7B, lane 1, and Fig. 7C, lane 2). No reactivity of the latter Abs was observed toward GST-Jun or GST (Fig. 7, B and C, lanes 3 and 4), while the proteins were detected using a GST-specific Ab (Fig. 7D). Taken together, these results demonstrate that Fas-reactive Abs are present in IVIg. Anti-Fas IgG Abs were purified from IVIg by affinity chromatography using GST-Fas extracytoplasmic fusion protein bound to Sepharose. The Abs were added to cultures of CEM cells for 12 h. Affinity-purified anti-Fas IgG induced cell death in the cultures at a concentration of 380 mg/ml, which is a 120-fold lower concentration than that of unfractionated IVIg that resulted in similar cell mortality, as assessed by using the PI uptake and MTT assay (Fig. 8). MTT assays consistently indicated more than 90% mortality after 12-h treatment with affinity-purified anti-Fas preparations (380 mg/ml). Anti-Fas mAb CH-11 induced apoptosis to a similar extent when used at a concentration of 50 ng/ml. The difference in the ability of IgM CH-11 and IgG affinity purified from IVIg to induce apoptosis may reflect differences in the avidity of binding

FIGURE 5. Fas dependency of IVIg-mediated apoptosis. A, The HuT78 lymphoblastoid T cell line sensitive to Fas-mediated apoptosis and the HuT78.B1 line resistant to Fas-mediated apoptosis was cultured in the presence of IVIg (50 mg/ml; 0.3 mM and 25 mg/ml; 0.15 mM) for 16 h. Mortality of cells was measured using the PI uptake assay. The decrease in the level of apoptosis in HuT78.B1 is significant (ppp 5 0.003). B, CEM cells were cultured in the presence of IVIg (50 mg/ml, 0.3 mM) and of soluble Fas molecules (i.e., the supernatants of fibroblasts transfected with Fas cDNA, encoding FasD Exo3, 4 and FasD Exo4 soluble Fas molecules, or in the presence of control supernatants of cells transfected with the vector alone) for 6 h. Percentage of IVIg-induced cell death (i.e., [death in the presence of IVIg and the supernatant] 2 [spontaneous death in the presence of supernatant alone]) was measured by PI uptake, and the graph depicts the mean values 6 SE obtained in three separate experiments conducted in duplicates.

3786


FIGURE 6. A, Expression of Fas in the BWtTAFF3 transfectant. Fas Ag expression on the parental cell line BWtTA (a) and the BWtTAFF3 clone (b) was assessed by staining with the DX2 anti-Fas mouse mAb and an irrelevant IgG control mAb of the same isotype, followed by a FITCconjugated sheep anti-mouse Ig. Relative fluorescent intensities of individual cells were analyzed using a FACScan flow cytometer (Becton Dickinson). B, Characterization of anti-Fas Abs in IVIg. Lysates from surfacebiotinylated BWtTA parental cells (lanes 1 and 3) and the BWtTAFF3 Fas transfectant cells (lanes 2 and 4) immunoprecipitated with anti-Fas mAb DX2 (lanes 1 and 2), or IVIg (lanes 3 and 4), as indicated in Materials and Methods. Precipitated proteins were electrophoresed and revealed by protein blotting with avidin-peroxidase and the ECL detection system.

to membrane Fas of the two types of Abs (42). Dye uptake assays, on the other hand, displayed about 50% mortality of CEM T cells with both affinity-purified anti-Fas from IVIg and CH-11 mAb (Fig. 8). In a separate set of experiments, we observed that affinitypurified anti-Fas Abs induced a threefold higher extent of apoptosis in Fas-sensitive HuT78 cells, as compared with the HuT78.B1 Fas-resistant cells (not shown). The difference in apoptosis observed in Fas-sensitive and resistant cells was higher when affinity-purified anti-Fas Abs were tested than we previously observed with unfractionated IVIg (Fig. 5). Induction of apoptosis of CD40-activated tonsillar B cells by IVIg The capacity of IVIg and affinity-purified anti-Fas Abs from IVIg to induce apoptosis of normal, nontransformed cells was then examined. Ligation of CD40 is known to induce the expression of Fas on B lymphocytes and facilitate apoptosis through the Fas pathway (29). Purified human tonsillar B cells were incubated with anti-CD40 mAb for 18 h, which resulted in a dose-dependent upregulation of Fas expression, as demonstrated by staining of B cells with phycoerythrin-labeled anti-Fas mAb 7C11 (not shown). Further incubation of the CD40-activated cells with IVIg resulted in induction of apoptosis in a dose-dependent manner, as assessed by the uptake of PI (Fig. 9). At 45 mg/ml, IVIg induced apoptosis in both resting and CD40-activated cells. The extent of apoptosis was twofold higher in CD40-activated cells when a concentration of 22 mg/ml of IVIg was used. Affinity-purified anti-Fas Abs from IVIg consistently induced increased apoptosis in activated cells

FIGURE 7. Immunoblotting of GST fusion proteins. Fas extracytoplasmic-GST fusion protein (lane 1), Fas intracytoplasmic-GST fusion protein (lane 2), GST-Jun control fusion protein (lane 3), and GST (lane 4) were electrophoresed and immunoblotted with IVIg (A), the N18 Ab that recognizes the N terminus of Fas (B), the C20 Ab specific for a C-terminal epitope of Fas (C), and an anti-GST polyclonal Ab (D), followed by protein A-peroxidase and the ECL detection system.

expressing Fas, as compared with resting B cells. Incubation of activated cells with control anti-Fas mAb CH-11 also resulted in apoptosis. The overall extent of apoptosis of resting and activated cells was significantly lower than that observed with transformed cells. IVIg-mediated apoptosis is associated with activation of caspases Activation of caspases, a family of cysteine proteases, is an important event in the execution phase of apoptosis. Apoptosis induced by a variety of agents could be blocked by inhibiting the caspase activity (43– 45). To assess the possible role of cysteine proteases in IVIg-mediated apoptosis, CEM T cells (1.25 3 105 cells/well) were pretreated with inhibitors of the proteases for 3 h before culture with IVIg (50 mg/ml; 0.3 mM) for 6 h. Cell death was assessed by measuring PI uptake. The peptide inhibitors for caspase 1 (ICE) and for caspase 3 (CPP32b/Yama), YVAD, and DEVD, respectively, inhibited IVIg-induced apoptosis (Fig. 10). At optimal concentrations, the inhibition observed with the tetrapeptide DEVD was 35%, and that observed with YVAD was 25%. A combination of YVAD and DEVD resulted in inhibition


3787

Discussion

FIGURE 8. Induction of apoptosis in CEM cells by affinity-purified antiFas Abs from IVIg. Anti-Fas IgG was affinity purified from IVIg using Fas-GST-Sepharose column. CEM cells were cultured for 12 h in the presence of affinity-purified anti-Fas IgG (380 mg/ml), the anti-Fas monoclonal IgM CH11 (50 ng/ml), or in medium alone before assessing cell viability by means of the PI uptake (A) or by MTT assay (B).

similar to that observed with DEVD alone (data not shown). The control peptides, leupeptin and calpain inhibitor E64(d), failed to inhibit IVIg-dependent apoptosis of CEM cells. These data indicate that activation of the caspase family of proteases is, at least in part, associated with IVIg-induced apoptosis. Since caspase 3 has been involved in Fas-mediated as well in other apoptotic process (43, 45– 47), we further tested the fate of PARP, a key substrate of caspase 3, in IVIg-treated CEM T cells. CEM cells were treated with IVIg for up to 24 h. Whole cell extracts were then obtained, subjected to electrophoresis, and immunoblotted with the antiPARP Ab C2-10. The cleavage of PARP into an 85-kDa signature death fragment was evident within 6 h in IVIg-treated cells, and its amount progressively increased during the following 24 h. No such cleavage was observed in control cells that had been incubated in the presence of medium alone (Fig. 10B).

FIGURE 9. Induction of apoptosis in nontransformed cells by IVIg (A) and affinity-purified antiFas Abs from IVIg (B). Purified tonsillar B cell preparations were ligated with CD40 (1 mg/ml) by culturing the cells with soluble anti-CD40 mAb for 18 h. After washing, resting (light hatches) and CD40-activated cells (dark hatches) were incubated with increasing concentrations of IVIg, affinity-purified anti-Fas Abs from IVIg, and CH-11. Human serum albumin (25 mg/ml) was used as negative control. Mortality of cells was assessed by PI uptake.

We have characterized the growth-inhibitory effects of normal human IgG, IVIg, on human T and B lymphocyte and monocyte lines as well as on CD40-activated human tonsillar B cells. We have observed that IVIg induces apoptotic cell death in a dose- and time-dependent manner associated with characteristic DNA ladder formation and surface exposure of phosphatidylserines. We present evidence in support of a role for the Fas molecule in apoptosis, for the presence of anti-Fas Abs in IVIg, and for a role of caspases 1 and 3 or other closely related caspases in IVIg-mediated apoptosis. Involvement of Fas in IVIg-induced cell death was obtained by two approaches. The Fas-sensitive HuT78, a human T lymphoblastic cell line, was compared for its sensitivity to IVIg-mediated apoptosis with that of a Fas apoptosis-resistant variant, HuT78.B1. HuT78.B1 contains a wild-type and a truncated Fas receptor (24). In addition, these cells are defective in eliciting some early signaltransduction events upon Fas cross-linking, such as activation of the acidic sphingomyelinase (48). Our results indicate that the variant cells were resistant to IVIg-mediated apoptosis. This, together with the observation that IVIg-mediated apoptosis is inhibited in the presence of soluble Fas molecules, provided strong evidence for the interaction between IVIg and Fas. Taken together, these experiments suggest that the presence of anti-Fas molecules in IVIg is important for apoptosis. In addition, IVIg specifically immunoprecipitated a 48-kDa protein species from Fas-transfected cells and not from nontransfected parental cells. On Western blots, IVIg preparations specifically recognized Fas-GST fusion proteins and not GST alone. It is interesting that reactivity toward both the extracytoplasmic and intracytoplasmic regions of the Fas receptor was detected. A Fas-splicing variant coding for a secreted protein lacking the transmembrane domain, but identical to the membranebound receptor in both the extracytoplasmic and intracytoplasmic regions, has been described (49, 50). Thus, Abs in IVIg may recognize both wild-type and soluble forms of molecule. These data together provided a molecular basis implicating Fas in IVIg-induced cytotoxic effect. A conclusive evidence for the role of Fas in IVIg-induced apoptosis came from the fact that affinity-purified anti-Fas Abs from IVIg pool efficiently induced apoptosis of CEM cells and HuT78 cells. IVIg also induced apoptosis in normal,

3788


FIGURE 10. Activation of caspases in IVIg-mediated apoptosis. A, CEM cells were preincubated with specific peptide inhibitors of ICE (YVAD, 250 mM) and of YAMA/CPP32b (DEVD, 250 mM) for 3 h, before culture in the presence or absence of 0.3 mM of IVIg for 6 h. Cell death was quantitated by PI uptake. The effect of inhibitors was calculated as follows: (apoptosis in the presence of IVIg 2 apoptosis in medium alone) 2 (apoptosis in the presence of both IVIg and the inhibitor 2 apoptosis in the presence of inhibitor alone)/(apoptosis in the presence of IVIg 2 apoptosis in medium alone) 3 100. The values indicated correspond to means of three experiments conducted in duplicates. Positive values indicate inhibition of IVIg-induced cell death, while negative values correspond to increase in IVIg-induced death upon treatment with protease inhibitor. B, Western blot analysis of PARP proteolysis in CEM cells treated with IVIg. Cell lysates equivalent of 1 3 105 CEM cells were loaded per lane. The approximate position of migration of prestained m.w. markers is indicated on the left. Lanes 3, 5, 7, 9, 11, 13, and 15 are samples of cells treated with IVIg for 2, 4, 6, 12, 14, 18, and 24 h, respectively. Lanes 1, 2, 4, 6, 8, 10, 12, and 14 are those of untreated cells taken at 0, 2, 4, 6, 12, 14, 18, and 24 h, respectively. The 85-kDa breakdown fragment of PARP as result of Yama actvation is indicated by an arrowhead.

nontransformed tonsillar B cells. While unfractionated IVIg induced apoptosis in both resting and CD40-activated cells to a similar extent, affinity-purified anti-Fas Abs from IVIg consistently induced increased apoptotic cell death in activated cells with induced surface expression of Fas as compared with resting B cells. The apparent discrepancy between the almost similar capacity of unfractionated IVIg to induce apoptosis in CD40-stimulated and unstimulated B cells and the selective ability of anti-Fas Abs to induce apoptosis of CD40-activated cells, supports the involvement of several pathways in addition to the Fas pathway, in IVIg-mediated cell death. The overall extent of apoptosis of resting and activated B cells was significantly lower than that observed with transformed cells. These findings are consistent with the observations that the transformed cells are more sensitive to apoptosis (51, 52). Although our experiments clearly pointed out a role for Fas in IVIg-induced cell death, the fact that variant HuT78.B1 cells with mutant Fas underwent apoptosis, albeit to a lesser extent, indicates the involvement of additional apoptotic pathways. This is consistent with the observation that soluble Fas molecules partially inhibited, approximately 50%, IVIg-induced cell death. Moreover, apoptosis occurred in the presence of F(ab9)2 fragments of IVIg to a lesser extent than that of intact IVIg, indicating either the involvement of Fcg receptors on target cells, or a partial loss of Ab activity upon enzymatic digestion of IgG and purification of F(ab9)2 fragments. The indication that multiple apoptotic pathways may be activated upon IVIg treatment prompted us to test downstream signal-transduction events. Experiments to assess the role of caspase family members indicated a modest inhibition of apoptosis in the presence of peptide inhibitors for caspase 1 (Ac-YVAD-CHO) and caspase 3 (Ac-DEVD-CHO). Inhibition observed was more pronounced with DEVD, indicating a more

critical requirement for caspase 3 or a closely related caspase. However, the inhibition observed was only partial, and apoptosis still occurred in the presence of caspase inhibitors. This would indicate the involvement of other members of caspases and a possible redundancy in their action, and is consistent with the suggestion that multiple members of caspases might be activated, perhaps as a cascade of events (53–55). It is clear that IVIg actively induces apoptosis in lymphoid cells and through activation of classical apoptotic signaling pathways. On the other hand, our data rule out any role for calpain protease in IVIg apoptosis. The calpain inhibitor, E64(d), in fact, sensitized the cells for IVIg apoptosis. Several mechanisms of action have been proposed to explain the immunomodulatory properties of IVIg in autoimmune and inflammatory diseases (56). These include reversible functional blockade of FcR on phagocytic cells (57), modulation of B and T cell functions through interactions of IVIg with FcR on lymphocytes (58), inhibition of the binding of complement components to cellular targets of complement activation (59), modulation of the synthesis and release of cytokines by lymphocytes and monocytes (7), and interaction with several surface molecules of immunocompetent cells (2). Induction of apoptosis by IVIg is of interest in the context of the beneficial effects of IVIg in the treatment of several autoimmune diseases and lymphoproliferative disorders. Apoptosis plays an important role in selection of lymphocyte repertoires, and defective apoptosis is associated with the pathogenesis of several autoimmune diseases (60 – 64). It has been proposed that too little apoptosis would lead to persistence of autoreactive cells, resulting in the production of autoantibodies and subsequently autoimmune conditions (65). The importance of Fas/APO-1-mediated apoptosis in various immune disorders and lymphoproliferative disorders is well documented.

The Journal of Immunology MRL mice homozygous for lpr (lymphoproliferation) or gld (generalized lymphoproliferative disease) develop a systemic autoimmune disease resembling systemic lupus erythematosus, Sjo¨gren’s disease, and rheumatoid arthritis (66). Defective apoptosis has been implicated in the pathogenesis of several conditions in which IVIg treatment has proven beneficial (2, 3, 65, 67, 68). The presence of anti-Fas molecules in IVIg may be significant in interfering with the pathogenic processes involved in certain autoimmune diseases and lymphoproliferative disorders. These functionally active anti-Fas molecules may induce apoptosis in activated self-reactive T and B cell clones that express Fas on their cell surface. Furthermore, IVIg has also been used in the treatment of clinical stage III and IV chronic lymphocytic leukemia associated with autoimmune hemolytic anemia or immune thrombocytopenic purpura, in which a decrease in total lymphocyte counts was observed (69, 70). The persistent lowering of the lymphocyte count was associated with diminution in the size of the lymph nodes and the spleen. The plasma concentration of IgG reached in a recipient of IVIg is in the range of 20 to 35 mg/ml, that is within the range of concentrations that were used in in vitro experiments reported in this work. Recent evidence indicates that in autoimmune thyroiditis, thyrocytes that constitutively express Fas ligand were apoptosed upon induction of Fas expression by disease-associated local production of IL-1 (71). These findings suggest that some of the autoimmune conditions may be associated with increased apoptosis of target tissue cells in an indirect fashion mediated by altered functioning of immune system. In such circumstances, reported beneficial effects of IVIg could be explained by its observed effects on the modification of production of cytokines and/or the existence of functionally active molecules against several cell surface Ags. Thus, normal serum and pooled normal human IgG have been shown to contain autoantibody activity against a number of molecules involved in the regulation of the immune response, such as idiotypic determinants of Ig (72, 73), determinants of TCR (74), CD4 (75), CD5 (76), HLA class I (12), b2-microglobulin (77), and the RGD motif (78). Such natural autoantibodies are found in newborns in cord blood IgM in mice and humans, probably being positively selected by self ligands rather than resulting from crossreactive responses to nonself determinants (79). Information from this study provides valuable and novel insights into mechanisms of action of IVIg in autoimmune, inflammatory, and lymphoproliferative disorders. The observation that IVIg induces apoptosis in human lymphocytes and monocytes with varied degree of effectiveness might be important in view of the beneficial effects observed with IVIg therapy in a broad spectrum of autoimmune disorders. IVIg may in fact render autoreactive T cells of the patients more sensitive to apoptosis. This may be followed by deletion of these cells apparently through classical apoptotic pathways. The signal-transduction events in response to IVIg at the molecular level, although unclear at present, provide a plausible working hypothesis to further elucidate the mechanisms underlying the immunoregulatory effects of IVIg. In addition, our results suggest that normal Igs participate in the selection of immune repertoires.

Acknowledgments We thank N. Bervas, N. Jouy, E. Lancia, and I. Pauselli for technical help. We also thank Dr. Ve´ronique Fre´meaux-Bacchi for helpful suggestions; Dr. H. Bujard for providing the pUHD 15-1 neo, pUHD 10-3, and pUHD 16-3 plasmids; Dr. O. Segatto for the pGEX GST-Jun construct; and Dr. R. Testi for the DX2 Ab.

3789

References 1. Dwyer, J. M. 1992. Manipulating the immune system with immune globulin. N. Engl. J. Med. 326:107. 2. Kazatchkine, M. D., G. Dietrich, V. Hurez, N. Ronda, B. Bellon, F. Rossi, and S. V. Kaveri. 1994. V region-mediated selection of autoreactive repertoires by intravenous immunoglobulin (IVIg). Immunol. Rev. 139:79. 3. Dalakas, M. C. 1997. Intravenous immune globulin therapy for neurologic diseases. Ann. Intern. Med. 126:721. 4. Leung, D. Y. 1989. The immunologic effects of IVIG in Kawasaki disease. Int. Rev. Immunol. 5:197. 5. Meissner, H. C., P. M. Schlievert, and D. Y. M. Leung. 1994. Mechanism of immunoglobulin action: observations of Kawasaki syndrome and RSV prophylaxis. Immunol. Rev. 139:109. 6. Arend, W. P., and D. Y. Leung. 1994. IgG induction of IL-1 receptor antagonist production by human monocytes. Immunol. Rev. 139:71. 7. Andersson, U., L. Bjork, S. U. Skansen, and J. Andersson. 1994. Pooled human IgG modulates cytokine production in lymphocytes and monocytes. Immunol. Rev. 139:21. 8. Abe, Y., A. Horiuchi, M. Miyake, and S. Kimura. 1994. Anti-cytokine nature of human immunoglobulin: one possible mechanism of the clinical effect of intravenous therapy. Immunol. Rev. 139:5. 9. Poutsiaka, D. D., B. D. Clark, E. Vannier, and C. A. Dinarello. 1991. Production of IL-receptor antagonist and IL-1b by peripheral blood mononuclear cells is differentially regulated. Blood 78:1275. 10. Pereira, B. J. G., D. D. Poutsiaka, A. King, J. A. Strom, G. Narayan, A. S. Levey, and C. A. Dinarello. 1992. In vitro production of interleukin-1 receptor antagonist in chronic renal failure, CAPD and HD. Kidney Int. 42:1419. 11. Shimozato, T., M. Iwata, H. Kawada, and N. Tamura. 1991. Human immunoglobulin preparation for intravenous use induces elevation of cellular cyclic adenosine 39:59-monophosphate levels, resulting in suppression of tumor necrosis factor a and interleukin-1 production. Immunology 72:497. 12. Kaveri, S., T. Vassilev, V. Hurez, R. Lengagne, S. Cot, P. Pouletty, D. Glotz, and M. Kazatchkine. 1996. Antibodies to a conserved region of HLA class I molecules, capable of modulating CD8 T cell-mediated function, are present in pooled normal immunoglobulin for therapeutic use (IVIg). J. Clin. Invest. 97:865. 13. Amran, A., H. Renz, G. Lack, K. Bradley, and E. W. Gelfand. 1994. Suppression of cytokine-dependent human T-cell proliferation by intravenous immunoglobulin. Clin. Immunol. Immunopathol. 73:180. 14. Van Schaik, I. N., I. Lundkvist, M. Vermeulen, and A. Brand. 1992. Polyvalent immunoglobulin for intravenous use interferes with cell proliferation in vitro. J. Clin. Immunol. 12:325. 15. Baixeras, E., L. Bosca, C. Stauber, A. Gonzalez, A. C. Carrera, J. A. Gonzalo, and C. Martinez-A. 1994. From apoptosis to autoimmunity: insights from the signaling pathways leading to proliferation or to programmed cell death. Immunol. Rev. 142:53. 16. Howie, S., D. J. Harrison, and A. E. Wyllie. 1994. Apoptosis. Immunol. Rev. 142:141. 17. Watanabe-Fukunaga, R. C., C. I. Brannan, N. G. Copeland, N. A. Jenkins, and S. Nagata. 1992. Lymphoproliferative disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356:314. 18. Cohen, J. J. 1994. Apoptosis: physiologic cell death. J. Lab. Clin. Med. 124:761. 19. Ridgway, W. M., H. L. Weiner, and C. G. Fathman. 1994. Regulation of autoimmune response. Curr. Opin. Immunol. 6:946. 20. Debatin, K. 1996. Disturbances of the CD95 (APO-1/Fas) system in disorders of lymphohaemopoietic cells. Cell Death Differ. 3:185. 21. Koopman, G., C. P. Reutelingsperger, G. A. Kuijten, R. M. Keehnen, S. T. Pals, and M. H. van Oers. 1994. Annexin V for flow cytometric detection of phosphatidylserine expression on B cells undergoing apoptosis. Blood 84:1415. 22. Baudet, V., V. Hurez, C. Lapeyre, S. V. Kaveri, and M. D. Kazatchkine. 1996. Intravenous immunoglobulin (IVIg) enhances the selective expansion of Vb31 and Vb171 ab T cells induced by superantigen. Scand. J. Immunol. 43:277. 23. Papoff, G., I. Cascino, A. Eramo, G. Starace, D. H. Lynch, and G. Ruberti. 1996. An N-terminal domain shared by Fas/Apo-1 (CD95) soluble variants prevents cell death in vitro. J. Immunol. 156:4622. 24. Cascino, I., G. Papoff, R. De-Maria, R. Testi, and G. Ruberti. 1996. Fas/Apo-1 (CD95) receptor lacking the intracytoplasmic domain protects tumor cells from Fas-mediated apoptosis. J. Immunol. 156:13. 25. Gossen, M., and H. Bujard. 1992. Tight control of gene expression in mammalian cells by tetracycline responsive promoters. Proc. Natl. Acad. Sci. USA 89:5547. 26. Resnitzky, D., M. Gossen, H. Bujard, and S. I. Reed. 1993. Acceleration of the G1/S phase transition by expression systems in higher eukaryotic cells. Nucleic Acids Res. 21:4411. 27. Itoh, N., and S. Nagata. 1993. A novel protein domain required for apoptosis. J. Biol. Chem. 268:10932. 28. Morgestern, J. P., and H. Land. 1990. Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res. 12:3587. 29. Schattner, E. J., K. B. Elkon, D. Yoo, J. Tumang, P. H. Krammer, M. C. Krow, and S. M. Friedman. 1995. CD40 ligation induces Apo-1/Fas expression on human B lymphocytes and facilitates apoptosis through the Apo-1/Fas pathway. J. Exp. Med. 182:1557. 30. McGahon, A. J., S. J. Martin, R. P. Bissonnette, A. Mahboubi, Y. Shi, R. J. Mogil, W. K. Nishioka, and D. R. Green. 1995. The end of the (cell) line: methods for the study of apoptosis in vitro. In Cell Death, Methods in Cell Biology Series, Vol. 46. L. M. Schwartz and B. A. Osborne, eds. Academic Press, San Diego, p. 153.

3790 31. Tian, Q., J. L. Taupin, S. Elledge, M. Robertson, and P. Anderson. 1995. Fasactivated serine/threonine kinase (FAST) phosphorylates TIA-1 during Fas-mediated apoptosis. J. Exp. Med. 182:865. 32. Meier, T., S. Arni, S. Malarkannan, M. Poincelet, and D. Hoessli. 1992. Immunodetection of biotinylated lymphocyte-surface proteins by enhanced chemoluminescence: a nonradioactive method for cell-surface protein analysis. Anal. Biochem. 204:220. 33. Desnoyers, S., G. M. Shah, G. Brochu, and G. G. Poirier. 1994. Erasable blot of poly(ADP-ribose) polymerase. Anal. Biochem. 218:470. 34. Fadok, V. A., D. R. Voelker, P. A. Campbell, J. J. Cohen, D. L. Bratton, and P. M. Henson. 1992. Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Immunol. 148:2207. 35. Nagata, S., and P. Golstein. 1995. The Fas death factor. Science 267:1449. 36. Krammer, P. H., I. Behrman, P. Daniel, J. Dhein, and K. M. Debatin. 1994. Regulation of apoptosis in the immune system. Curr. Opin. Immunol. 6:279. 37. Itoh, N., S. Yonehara, A. Ishii, M. Yonehara, S. I. Mizushima, M. Sameshima, A. Hase, Y. Seto, and S. Nagata. 1991. The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis. Cell 66:233. 38. Trauth, B. C., C. Klas, A. M. J. Peters, S. Matzku, P. Mo¨ller, W. Falk, K. M. Debatin, and P. Krammer. 1989. Monoclonal antibody-mediated tumor regression by induction of apoptosis. Science 245:301. 39. Suda, T., and S. Nagata. 1994. Purification and characterization of the Fas ligand that induces apoptosis. J. Exp. Med. 179:873. 40. Lynch, D. H., M. L. Watson, M. R. Alderson, P. R. Baum, R. E. Miller, T. Tough, M. Gibson, T. Davis-Smith, C. A. Smith, K. Hunter, D. Bhat, W. Din, R. G. Goodwin, and M. F. Seldin. 1994. The mouse Fas ligand gene is mutated in gld mice and is part of a TNF family gene. Immunity 1:131. 41. Takahashi, T., M. Tanaka, C. I. Brannan, N. A. Jenkins, N. G. Copeland, T. Suda, and S. Nagata. 1994. Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76:969. 42. Fadeel, B., C. J. Thorpe, S. Yonehara, and F. Chiodi. 1997. Anti-Fas IgG antibodies recognizing the same epitope of Fas/Apo-1 mediate different biological effects in vitro. Int. Immunol. 9:201. 43. Miller, D. K. 1997. The role of the caspase family of cysteine proteases in apoptosis. Semin. Immunol. 9:35. 44. Henkart, P. A. 1996. ICE family proteases: mediators of all apoptotic cell death. Immunity 4:195. 45. Martin, S. J., and D. R. Green. 1995. Protease activation during apoptosis: death by a thousand cuts? Cell 82:349. 46. Enari, M., H. Hug, and S. Nagata. 1995. Involvement of an ICE-like protease in Fas-mediated apoptosis. Nature 375:78. 47. Los, M., M. V. de-Craen, L. C. Penning, H. Schenk, M. Westendorp, P. A. Baeuerle, W. Dro¨ge, P. H. Krammer, W. Fiers, and K. Schulze-Osthoff. 1995. Requirement of an ICE/CED-3 protease for Fas/APO-1 mediated apoptosis. Nature 375:81. 48. Cifone, M. G., P. Roncaioli, R. D. Maria, C. Grazia, A. Santoni, G. Ruberti, and R. Testi. 1995. Multiple pathways originate at the Fas/Apo-1(CD95) receptor: sequential involvement of phosphatidylcholine-specific phospholipase C and acidic sphingomyelinase in the propagation of the apoptotic signal. EMBO J. 14:5859. 49. Cheng, J., T. Zhou, C. Liu, J. P. Shapiro, M. J. Brauer, M. C. Kiefer, P. J. Barr, and J. D. Mountz. 1994. Protection from Fas-mediated apoptosis by a soluble form of the Fas molecule. Science 263:1759. 50. Cascino, I., G. Fiucci, G. Papoff, and G. Ruberti. 1995. Three functional soluble forms of the human apoptosis-inducing Fas molecule are produced by alternative splicing. J. Immunol. 154:2706. 51. Ucker, D. S., L. D. Hebshi, J. F. Blomquist, and B. E. Torbett. 1994. Physiological T-cell death: susceptibility is modulated by activation, aging, and transformation, but the mechanism is constant. Immunol. Rev. 142:273. 52. Massaia, M., P. Borrione, C. Attisano, P. Barral, E. Beggiato, L. Montacchini, A. Bianchi, M. Boccadoro, and A. Pileri. 1995. Dysregulated Fas and Bcl-2 expression leading to enhanced apoptosis in T cells of multiple myeloma patients. Blood 85:3679. 53. Takahashi, A., E. S. Alnemri, Y. A. Lazebnik, T. Fernendes-Alnemri, G. Litwack, R. Moir, R. D. Goldman, G. G. Poirier, S. H. Kaufmann, and W. C. Earnshaw. 1996. Cleavage of lamin A by Mch2a but not CPP32: multiple interleukin 1b-converting enzyme-related proteases with distinct substrate recognition properties are active in apoptosis. Proc. Natl. Acad. Sci. USA 93:8395.

INDUCTION OF APOPTOSIS BY NORMAL HUMAN Ig 54. Enari, M., R. V. Talanian, W. W. Wong, and S. Nagata. 1996. Sequential activation of ICE-like and CPP32-like proteases during Fas-mediated apoptosis. Nature 380:723. 55. Schulze-Osthoff, K., M. K. A. Bauer, M. Vogt, and M. Los. 1996. Role of ICErelated and other proteases in Fas-mediated apoptosis. Cell Death Differ. 3:177. 56. Moller, G. 1994. Immunoglobulin treatment: mechanisms of action. Immunol. Rev. 139. 57. Fehr, J., V. Hofmann, and U. Kappeler. 1982. Transient reversal of thrombocytopenia in idiopathic thrombocytopenic purpura by high-dose intravenous gammaglobulin. N. Engl. J. Med. 306:1254. 58. Hashimoto, F., Y. Sakiyama, and S. Matsumoto. 1986. The suppressive effect of gammaglobulin preparations on in vitro pokeweed mitogen-induced immunoglobulin production. Clin. Exp. Immunol. 65:409. 59. Basta, M., L. F. Fries, and M. M. Frank. 1991. High-dose intravenous immunoglobulin inhibits in vitro uptake of C4 fragments onto sensitized erythrocytes. Blood 77:376. 60. Cohen, J. J. 1993. Apoptosis. Immunol. Today 14:126. 61. Green, D. R. 1994. Activation-induced apoptosis in lymphocytes. Curr. Opin. Immunol. 6:476. 62. Sprent, J., and S. R. Webb. 1995. Intrathymic and extrathymic clonal deletion of T cells. Curr. Opin. Immunol. 7:196. 63. Moller, G. 1994. Apoptosis in immunity. Immunol. Rev. 142. 64. Krammer, P. H., J. Dhein, H. Walczak, I. Behrmann, S. Mariani, B. Matiba, M. Fath, P. T. Daniel, E. Knipping, M. O. Westendorp, and et al. 1994. The role of APO-1-mediated apoptosis in the immune system. Immunol. Rev 142:175. 65. Rudin, C. M., and C. B. Thompson. 1997. Apoptosis and disease: regulation and clinical relevance of programmed cell death. Annu. Rev. Med. 48:267. 66. Cohen, P. L., and R. A. Isenberg. 1991. Lpr and gld: single gene models of systemic autoimmunity and lymphoproliferative disease. Annu. Rev. Immunol. 9:243. 67. Bauer, J., H. Wekerle, and H. Lassman. 1995. Apoptosis in brain-specific autoimmune disease. Curr. Opin. Immunol. 7:839. 68. Mountz, J. D., T. Zhou, X. Su, J. Wu, and J. Cheng. 1996. The role of programmed cell death as an emerging new concept for the pathogenesis of autoimmune diseases. Clin. Immunol. Immunopathol. 80:S2. 69. Besa, E. C. 1984. Use of intravenous immunoglobulin in chronic lymphocytic leukemia. Am. J. Med. 76:209. 70. Besa, E. C., and D. Klumpe. 1992. Prophylactic immunoglobulin in chronic lymphocytic leukemia. N. Engl. J. Med. 326:139. 71. Giordano, C., G. Stassi, R. De-Maria, M. Todaro, P. Richiusa, G. Papoff, G. Ruberti, M. Bagnasco, R. Testi, and A. Galluzzi. 1997. Potential involvement of Fas and its ligand in the pathogenesis of Hashimoto’s thyroiditis. Science 275:960. 72. Rossi, F., and M. D. Kazatchkine. 1989. Antiidiotypes against autoantibodies in pooled normal human polyspecific Ig. J. Immunol. 143:4104. 73. Kaveri, S. V., C. Y. Kang, and H. Kohler. 1990. Natural mouse and human antibodies bind to a peptide derived from a germline variable heavy chain: evidence for evolutionary conserved self-binding locus. J. Immunol. 145:4207. 74. Marchalonis, J. J., H. Kaymaz, F. Dedeoglu, S. F. Schlutter, D. E. Yocum, and A. B. Edmundson. 1992. Human autoantibodies reactive with synthetic autoantigens from T-cell receptor b chain. Proc. Natl. Acad. Sci. USA 89:3325. 75. Hurez, V., S. V. Kaveri, A. Mouhoub, G. Dietrich, J. C. Mani, D. Klatzmann, and M. D. Kazatchkine. 1994. Anti-CD4 activity of normal human immunoglobulins G for therapeutic use (intravenous immunoglobulin, IVIg). Ther. Immunol. 1:269. 76. Vassilev, T., C. Gelin, S. V. Kaveri, M. T. Zilber, L. Boumsell, and M. D. Kazatchkine. 1993. Antibodies to the CD5 molecule in normal human immunoglobulins for therapeutic use (intravenous immunoglobulins, IVIg). Clin. Exp. Immunol. 92:369. 77. Vincent, C., and J.-P. Revillard. 1983. Auto-antibodies specific for b2-microglobulin in normal human serum. Mol. Immunol. 20:877. 78. Vassilev, T., M. D. Kazatchkine, J. C. Mani, F. Schriever, and S. V. Kaveri. 1997. Therapeutic intravenous immunoglobulin (IVIg) contains anti-RGD antibodies that inhibit cellular adhesion interactions. In AAAAI/AAI/CIS Joint Meeting San Francisco, CA. 79. Coutinho, A., M. D. Kazatchkine, and S. Avrameas. 1995. Natural autoantibodies. Curr. Opin. Immunol. 7:812.