An IgM antibody is a potent immunosuppressive agent that inhibits B cell proliferation

The results of this study demonstrate that the culture supernatant of lipopolysaccharide (LPS)-stimulated murine spleen cells is able to inhibit the growth of freshly isolated B lymphocytes. The inhibition is specific for B cells because the suppression of the LPS, Fc-fragment of human IgG, dextran sulfate, and anti-mu induced proliferation of B, but not the Concanavalin A response of T lymphocytes could be shown. The cells producing the inhibitor do not adhere to plastic, and are Thy-1 negative but surface Ig positive, i.e. they are B lymphocytes. The regulatory substance is heat resistant, sensitive to trypsin treatment, and has a high molecular weight of approximately 1000 kDa. Moreover, it is specifically adsorbed on, and can be eluted from, anti-mu and anti-Ig immunoaffinity columns. Thus, it seems to be an IgM antibody. Non-specific effects were excluded by the ineffectiveness of poly- and monoclonal IgM proteins. IgM-IgG complexes were also excluded. Thus, these results suggest the existence of a novel IgM antibody mediated control mechanism regulating B cell growth during polyclonal activation.

Stabilization by a disulfide bond of the N-terminal domain of a mutant troponin C (TnC48/82)

The regulatory activity of troponin C is reversibly inhibited by a disulfide bridge between cysteine residues introduced by site-directed mutagenesis in positions 48 and 82 (TnC48/82) in the N-terminal domain of rabbit skeletal troponin C (sTnC; Grabarek, Z., Tan, R.-Y., Tao, T., and Gergely, J. (1990) Nature 345, 132-135). In the present work we have investigated the effects of the disulfide on structural properties of TnC48/82 monitored by CD spectroscopy and limited trypsinolysis. The CD spectra of the mutant protein in the oxidized form (oxTnC48/82) with and without Ca2+ are similar to the corresponding ones of the reduced and carboxamidomethylated form (CAMTnC48/82), indicating that the disulfide has essentially no effect on the overall secondary structure. The N-terminal domain of oxTnC48/82 is resistant to thermal unfolding, but that of CAMTnC48/82 is only slightly more stable than the corresponding domain of sTnC. In the presence of Ca2+ oxTnC48/82 is more resistant to trypsinolysis than sTnC whereas the rate of tryptic digestion of CAMTnC48/82 is the same as that of sTnC, indicating that peptide bonds adjacent to lysine residues at position 84 and 88, the sites of tryptic attack, are protected by the disulfide. The disulfide cross-linked N-terminal peptide of TnC48/82 does not bind TnI, unlike its reduced or carboxamidomethylated forms. Our data indicate that the disulfide between Cys48 and Cys82 stabilizes the structure of the N-terminal domain of TnC and blocks its ability to interact with TnI. The effects of the disulfide appear to be restricted to the N-terminal domain of TnC.

The role of C3 in the immune response

The complement system, particularly the third component, plays an important modulatory role in the inductive phase of the immune response. As discussed here by Anna Erdei and colleagues, the picture that is emerging is that immobilized C3 split products facilitate the cooperation between immunocompetent cells and are co-stimulatory molecules in T- and B-cell activation, probably as a result of their ability to promote cell-cell adhesion. In contrast, soluble C3 products inhibit lymphocyte proliferation.

Cross-linking of residue 57 in the regulatory domain of a mutant rabbit skeletal muscle troponin C to the inhibitory region of troponin I

Interactions between troponin C (TnC) and troponin I (TnI) play an important role in the Ca(2+)-dependent regulation of vertebrate striated muscle contraction. In the present study, we investigated the sites of interaction between the N-terminal regulatory domain of TnC and the inhibitory region (residues 96-116) of TnI, using a mutant rabbit skeletal TnC (designated as TnC57) that contains a single Cys at residue 57 in the C-helix. TnC57 was modified with the photoreactive cross-linker 4-maleimidobenzophenone (BP-Mal), and, after formation of a binary complex with TnI, cross-linking between the proteins was induced by photolysis. The resulting product was cleaved with CNBr and several proteases, and peptides containing cross-links were purified and subjected to amino acid sequencing. The results show that Cys-57 of TnC57 is cross-linked to the segment of TnI spanning residues 113-121. Previously, we showed that Cys-98 of TnC can be cross-linked via BP-Mal to TnI residues 103-110 (Leszyk, J., Collins, J.H., Leavis, P.C., and Tao, T. (1987) Biochemistry 26, 7042-7047). Taken together, these results demonstrate that both the C- and the N-terminal domains of TnC interact with the inhibitory region of TnI and are consistent with the hypothesis that, in a complex with TnI, TnC adopts a more compact conformation than in the crystal structure.

Direct correlation of parvalbumin levels with myosin isoforms and succinate dehydrogenase activity on frozen sections of rodent muscle

Parvalbumin (PV) is a soluble Ca++ binding protein which is particularly concentrated in fast muscles of rodents. We have developed a new protocol to fix frozen sections of muscle by formaldehyde vapor, which enabled us to immunochemically stain serial frozen sections for PV. Fiber types were defined on the basis of myosin ATPase stability, and of isomyosins identified by a variety of antibodies because ATPase stability alone yielded ambiguous results in the mouse. Slow Type I fibers in mouse and rat were devoid of PV and had intermediate to high SDH levels. Fast fiber subtypes IIA, IIB, and IIX-like were defined in the mouse on the basis of the similarity of their myosin heavy chain immunoreactivity to these types in the rat. The soleus muscle was usually PV negative, but a small population of strongly PV-positive IIX-like fibers was present in the mouse. In mouse fast muscle, small diameter IIA fibers were PV negative with high SDH activity. In both mouse and rat, PV reactivities of IIB and IIX fibers were higher than those of IIA and I, whereas SDH levels of IIA, IIX, and I fibers were higher than those of IIB. Thus, PV content correlated with the type of myosin ATPase but not with SDH levels. The method described for immunocytochemistry of PV may be applicable to other highly soluble proteins.

Modulation of type II Fc gamma receptor expression on activated human B lymphocytes

We have monitored Fc gamma RII expression during the activation of human B lymphocytes by simultaneous analysis of monoclonal antibody (mAb) binding and EA rosetting. The expression of Fc gamma RII showed a biphasic time course. Initially, a transient increase of Fc gamma RII with no ligand-binding capacity was observed with mAb staining as early as 10 min after stimulation by the F(ab')2 fragment of anti-human IgM or phorbol 12-myristate 13-acetate and then after 3 to 24 h a decrease in the number of Fc gamma RII+ cells was seen. Trypsin-like serine protease activity also appeared in the lysate of activated B cells at this time. On the 2nd day of activation a significant enhancement of Fc gamma RII expression was observed, mainly on enlarged blast cells as monitored by both mAb and by ligand binding (EA rosette). At the same time, soluble fragments of Fc gamma RII with the ability to bind human Fc were detected in the supernatant of activated B cells, probably as a result of proteolytic cleavage. These findings suggest that activated B cells are identical with the population of mononuclear cells which shed Fc gamma R when incubated at 37 degrees C. The ability of activated but not resting B cells to release Fc gamma RII correlates with the expression of early activation markers and with the appearance of a trypsin-like serine protease activity of the same cells; thus, the release of Fc gamma RII occurs in the early G1 phase of cell cycle as a result of proteolysis. Later the release of Fc gamma RII is accompanied by the enhancement of Fc gamma RII expression before the cells reach the S phase. The fragments of cleaved Fc gamma RII had an apparent molecular mass of 33 and 14-18 kDa under nonreducing conditions, and upon reduction fragments of smaller size were observed.

Isolation and characterization of IgG2a-reactive autoantibodies from influenza virus-infected BALB/c mice

Repeated influenza virus infection induces the production of dominantly IgG2a-type virus-specific antibodies as well as the appearance of IgG2a-reactive autoantibodies in BALB/c mice characterized by low spontaneous rheumatoid factor-type autoantibody production. IgG2a-reactive autoantibody-producing hybridomas could be isolated from the spleen of influenza virus-infected BALB/c mice. The mAb produced by these clones represent not only IgM but also IgG and IgA isotypes and show strong isotype or isoallotype specificity. The common functional property of these autoantibodies is their preferential- and high-affinity binding to complexed, solid-phase-bound or heat-aggregated IgG2a when compared to native soluble or cell-bound IgG2a. The mechanism of induction and the possible biological function of these autoantibodies are discussed in the light of their fine specificity and functional properties.

In vivo manipulation of IgG2a production by isotype-specific autoantibodies

Repeated intranasal infection of Balb/c mice with A/PR/8 influenza virus induced an intense antiviral IgG response dominated by the IgG2a subclass, and accompanied by the appearance of IgG2a reactive autoantibodies. Cells producing IgG2a reactive autoantibodies could then be cloned as hybridomas from the virus infected animals. Monoclonal antibodies produced by selected hybridomas U28, Z26 and Z41 produced IgM-type antibodies with strong specificity for the IgG2a isotype bearing "a" allotypic determinants on the Fc region. These IgG2a specific autoantibodies showed highly preferred binding to solid phase bound or aggregated IgG2a, compared to soluble native IgG2a. Based on these characteristics they were classified as mono-reactive rheumatoid factor (RF)-like autoantibodies. Passive administration of IgM type IgG2a-specific autoantibodies to influenza virus infected animals resulted in a long-term reduction in the secondary antiviral response. This could be demonstrated by decreased virus neutralizing activity of the serum and diminished level of IgG2a-type anti-viral antibodies. A similar effect was observed in Balb/c mice contact sensitized with oxazolone: passive administration of RF-like antibodies resulted in reduced IgG2a response to oxazolone while the level of antibodies belonging to other isotypes was not influenced. These results suggest an isotype-specific regulatory function of these RF-like autoantibodies presumably acting via antigen-antibody complexes.

The two binding-site models of human IgG binding Fc gamma receptors

Fc receptors (FcR) are immunoglobulin-binding molecules that enable antibodies to perform several biological functions by forming a link between specific antigen recognition and effector cells. FcRs are involved in regulating antibody production as well. Most FcRs belong to the immunoglobulin superfamily, and show structural homology with each other and with their ligands. Recent data on the structure of IgG binding FcRs obtained from monoclonal antibodies and gene cloning studies, as well as on ligand binding capacity and fine specificity of the receptor binding site (or sites), are reviewed. The binding capacity and fine specificity of receptor binding sites, as well as the structure and conformation of the immunoglobulin ligands, play important roles in triggering FcR-mediated signals. In induction of signals, the interaction of the FcR with the CH2 domain of the IgGFc is decisive. The high-affinity Fc gamma RI possess one active binding site specific for contact residues that is located at the N-proximal end of the CH2 domain and is able to mediate both binding and signal transfer. The low-affinity Fc gamma RIII has two active binding sites: the CH3 domain-specific site, which mediates only binding; and the CH2 domain-specific site, which is responsible for binding and signaling. Similarly, the low-affinity Fc gamma RII on resting B cells has one site for CH2 and another for CH3 binding. The expression, release, and fine specificity of Fc gamma RII on B cells correlates with the cell cycle.

Fine specificity of a rabbit antibody interacting with human IgG Fc receptor-like molecules

A polyclonal rabbit antibody raised against an Fc receptor (FcR)-like membrane glycoprotein fraction of chronic leukaemic lymphocytes has previously been prepared and partially characterized. This antibody, called AbA, was found to precipitate a 70-kDa and a 45-kDa fraction of the detergent lysate of U937 cells and to inhibit ligand binding to Fc gamma R on the P388D1 murine macrophage cell line. In the present work we have characterised this antibody further. All Fc gamma RII-positive B lymphoblastoid cell lines, as well as resting human B lymphocytes, were positively stained with the AbA antibody. U937 cells were found to be negative, but after stimulation with phorbol ester (PMA), 50% of the cells became positive. AbA antibody did not react with human T cell lines or with the T + 0 cell subset of peripheral blood. Monocytes were also negative. On the other hand, AbA antibody exhibited a dose-dependent inhibition of antibody-mediated cytotoxic reaction (ADCC) of monocytes, while not affecting K cell-mediated ADCC. It had an inhibitory effect of EA rosette formation of B cells and stimulated U937 cells. Furthermore, it interacted with the soluble form of Fc gamma RII released by activated B lymphocytes, and--similarly to IgG--precipitated a 33 kDa fraction from the supernatant of B cells.

Inhibition of mutant troponin C activity by an intra-domain disulphide bond

Triggering of contraction in striated muscles involves a conformational transition in the N-terminal domain of troponin C, the calcium-binding component of thin filaments. We have designed a mutant troponin C in which the key conformational transition and the calcium-regulatory activity are reversibly blocked by the formation of a disulphide bridge. Our results may be applicable to other proteins of the same family of calcium-binding proteins.

Synthesis and functional studies of tuftsin analogs containing isopeptide bond

In the present paper a new approach is reported, to increase the resistance of tuftsin toward enzymatic cleavage by the introduction of an isopeptide bond into the molecule. The tetrapeptides H-Lys(Thr)-Pro-Arg-OH and H-Lys(Ala)-Pro-Arg-OH, the pentapeptides H-Thr-Lys(Ala)-Pro-Arg-OH, H-Thr-Lys(Thr)-Pro-Arg-OH and H-Ala-Lys(Ala)-Pro-Arg-OH and their For- and Boc-protected derivatives were built up by stepwise elongation of the chain, using conventional solution-phase methods. Preliminary experiments confirmed that from the Lys residue in position 2 of tuftsin the alpha-peptide bond between the Thr and Lys is cleaved with a significantly higher rate by leucine aminopeptidase than the epsilon-peptide bond. Several of the isopeptide derivatives increased to a higher extent the interleukin (IL-1) secretion by monocytes than tuftsin or [Ala1]-tuftsin.

Ca2(+)-dependent interactions between the C-helix of troponin-C and troponin-I. Photocross-linking and fluorescence studies using a recombinant troponin-C

We have used in vitro mutagenesis to synthesize in Escherichia coli a recombinant rabbit skeletal troponin-C (designated as TnC57) in which Cys-98 was replaced with leucine, and Ala-57 in the C-helix of the N-terminal domain was replaced with cysteine. TnC57 labeled with the bifunctional photocross-linker benzophenone-4-maleimide could be photocross-linked with troponin-I in both the binary complex with troponin-I and in the ternary complex with troponin-I and troponin-T. The fluorescence lifetime of TnC57 labeled with the probe N-iodoacetyl-N'-(5-sulfo-1-naphthyl)ethylenediamine decreased from 13.2 +/- 0.1 to 11.8 +/- 0.1 ns when Ca2+ bound to the low affinity triggering sites. Complexation with either troponin-I or both troponin-I and troponin-T resulted in significant increases in this lifetime both in the absence and the presence of Ca2+. In either the binary or the ternary complex, this lifetime increased from 15.5 to 18.0 ns upon Ca2+ binding to the low affinity sites. Complementary acrylamide-quenching studies yielded results that are consistent with the fluorescence lifetime results. Our results show that the C-helix of troponin-C interacts with troponin-I, in confirmation of recent zero-length cross-linking results (Leszyk, J., Grabarek, Z., Gergely, J., and Collins, J.H. (1990) Biochemistry 29, 299-304). Moreover, they are in support of a model (Herzberg, O., Moult, J., and James, M.N.G. (1986) J. Biol. Chem. 261, 2638-2644) in which the binding of Ca2+ to the triggering sites in the N-terminal domain of troponin-C results in the movement of the B- and C-helices away from the central helix, thereby exposing a putative hydrophobic binding site for troponin-I.

Zero-length crosslinking procedure with the use of active esters

A two-step zero-length crosslinking procedure for studying protein-protein complexes has been developed. One component of a complex is briefly incubated with 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC) in the presence of N-hydroxysuccinimide resulting in the conversion of some of the protein carboxyls into succinimidyl esters. The reaction is stopped by addition of beta-mercaptoethanol and other interacting proteins are then added. Crosslinking arises from substitution of lysine epsilon-amino groups of these proteins for the succinimidyl moieties during a 1- to 2-h incubation period. The advantage of this method versus one-step zero-length crosslinking is that only one component of the complex is exposed to the crosslinker, which eliminates complications arising from the formation of crosslinks among several proteins of a multicomponent complex. Furthermore, crosslinks can be formed even in the presence of reagents, such as dithiothreitol and EDTA, that would interfere with direct crosslinking with EDC.

Characterization of zero-length cross-links between rabbit skeletal muscle troponin C and troponin I: evidence for direct interaction between the inhibitory region of troponin I and the NH2-terminal, regulatory domain of troponin C

Interactions between troponin C (TnC) and troponin I (TnI) play an important role in the Ca2(+)-dependent regulation of vertebrate striated muscle contraction. Previous attempts to elucidate the molecular details of TnC-TnI interactions, mainly involving chemically modified proteins or fragments thereof, have led to the widely accepted idea that the "inhibitory region" (residues 96-116) of TnI binds to an alpha-helical segment of TnC comprising residues 89-100 in the nonregulatory, COOH-terminal domain. In an attempt to identify other possible physiologically important interactions between these proteins, 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide (EDC) was used to produce zero-length cross-links in the complex of rabbit skeletal muscle TnC and TnI. TnC was activated with EDC and N-hydroxysuccinimide (NHS) and then mixed with an equimolar amount of TnI [Grabarek, Z., & Gergely, J. (1988) Biophys. J. 53, 392a]. The resulting cross-linked TnCXI was cleaved with cyanogen bromide, trypsin, and Staphylococcus aureus V8 protease (SAP). Cross-linked peptides were purified by reverse-phase HPLC and characterized by sequence analysis. The results indicated that residues from the regulatory Ca2(+)-binding site II in the NH2-terminal domain of TnC (residues 46-78) formed cross-links with TnI segments spanning residues 92-167. The most highly cross-linked residues in TnI were Lys-105 and Lys-107, located in the inhibitory region. These results yield the first evidence for an interaction between the N-terminal domain of TnC and the inhibitory region of TnI.

Phosphorylation of type II Fc gamma receptor on activated human B lymphocytes

Activation of resting human B lymphocytes either by cross-linking their membranal IgM or by phorbol esters has been previously demonstrated to modulate the type II receptor for Fc gamma domains (Fc gamma RII): shortly after stimulation a decrease in IgG binding capacity and an enhancement of Fc gamma RII expression were observed. These were followed by the release of Fc gamma RII fragments from the cell membrane. Since protein phosphorylation is well-established signal transduction element, we examined whether Fc gamma RII may be a target of such activation induced phosphorylation. Resting (high density) and activated (low density) human tonsil B lymphocytes were stimulated either by cross-linking their surface IgM (sIgM) or by the phorbol ester TPA. This treatment induced specific phosphorylation of a 36 kd membrane protein. This polypeptide was shown to specifically bind to IgG-coated Sepharose beads or to monoclonal Fc gamma RII-specific antibody-coated Affi-Gel 10 beads; thus, it most probably corresponds to the Fc gamma RII of these cells. In addition, phosphorylation of a 20 kd protein with similar binding characteristics was also observed in several experiments. Both serine and tyrosine were the amino acids that underwent phosphorylation in the 36 kd Fc gamma RII. The extent of Fc gamma RII phosphorylation correlated with the increase in receptor expression as monitored by specific mAb binding and, at the same time, with the decrease in the capacity to bind IgG-sensitized erythrocytes. These results suggest that stimulation-induced phosphorylation of Fc gamma RII on B cells is an early signal transduction element involved in controlling B cell response.

Structural characteristics influencing the carrier function of synthetic branched polypeptides based on poly[Lys-(DL-Ala)3)]backbone

Effective carrier function of selected representatives of new branched polypeptides covalently coupled with the synthetic monovalent hapten, oxazolone was studied. The effectiveness of oxazolone-synthetic polypeptide conjugates in inducing oxazolone-as well as carrier-specific antibody responses in inbred mice was compared to that of bovine serum albumin (BSA)- and KLH-oxazolone conjugates. The synthetic polypeptides, poly[Lys-(D-Leui-DL-Alam)] (D-LAK), LAK and FAK, as well as the common poly[Lys-(DL-Alam)](AK) core covalently coupled to oxazolone (Ox) induced a T cell-dependent antibody response when repeatedly administered with or without Freund's adjuvant in mice. This was evidenced by: the increasing titer of oxazolone-specific IgG during the course of the memory response; the appearance of all IgG subclasses; the effective oxazolone-specific priming by the conjugates; and the induction of an intense oxazolone- and carrier-specific DTH reaction. Although the oxazolone-specific antibody response was 10-100 times lower than that induced by KLH- or BSA-oxazolone conjugates, it was accompanied by a lower level or no detectable carrier-specific antibody response despite an effective carrier-specific T cell-mediated response. Significant differences were observed between the effectiveness of synthetic polypeptides used as carrier: highest oxazolone-specific antibody titers were observed using the AK, LAK and FAK conjugates. The intensity and specificity of the DTH reaction and antibody response induced by the carrier-oxazolone conjugates suggested that the distinct effectiveness of L- and D-amino acid-containing conjugates (LAK vs D-LAK and FAK vs D-FAK) was dependent on altered B cell recognition of the haptenic group. Circular dichroism (CD) spectra indicating different local orientation of oxazolone, when coupled to L or D side chain-terminating amino acids, support this suggestion.

Ca2+ dependence of the distance between Cys-98 of troponin C and Cys-133 of troponin I in the ternary troponin complex. Resonance energy transfer measurements

We have used resonance energy transfer to study the spatial relationship between Cys-98 of rabbit skeletal troponin C and Cys-133 of rabbit skeletal troponin I in the reconstituted ternary troponin complex. The donor was introduced by labeling either troponin C or troponin I with N-(iodoacetyl)-N'-(5-sulfo-1-naphthyl)ethylenediamine, while the acceptor was introduced by labeling either protein with N-[4-(dimethylamino)phenyl-4'-azophenyl]maleimide. The extent of energy transfer was determined by measuring the quenching of the donor fluorescence decay. The results indicate first that the distance between these two sites is not fixed, suggesting that the protein regions involved possess considerable segmental flexibility. Second, the mean distance between the two sites is dependent on the metal-binding state of troponin C, being 39.1 A when none of the metal-binding sites are occupied, 41.0 A when Mg2+ ions bind at the high-affinity sites, and 35.5 A when Ca2+ ions bind to the low-affinity sites. Neither the magnitude of the distances nor the trend of change with metal ions differs greatly when the locations of the probes are switched or when steady-state fluorometry was used to determine the transfer efficiency. Since the low-affinity sites have been implicated as the physiological triggering sites, our findings suggest that one of the key events in Ca2+ activation of skeletal muscle contraction is a approximately 5-A decrease in the distance between the Cys-98 region of troponin C and the Cys-133 region of troponin I.

Interaction between C3 and IL-2; inhibition of C3b binding to CR1 by IL-2

We previously reported that C3 has a role in the enhancement of the IL-2 dependent proliferation of helper T cells. Because the IL-2R has a structural homology with the complement proteins, such as CR1 and CR2, we studied the possible ligand crossreactions on CR1 and IL-2-receptor, and the direct interaction between C3 and IL-2. While C3 has an enhancing effect on the IL-2 dependent proliferation of HT-2, a CR1-positive mouse T-cell line, the growth of the CTLL-16 line (CR1-negative) is not affected by C3. It has been proven that neither the insolubilized C3 nor the soluble C3b-like C3 react with the IL-2 binding epitope of the IL-2 receptor. However, using human RBC we have demonstrated that the binding of aggregated C3 to CR1 is inhibited by rIL-2, in a dose-dependent manner. When RBC were incubated with rIL-2 and FITC-labelled Fab-anti-CR1 simultaneously, there was no inhibition in the fluorescence intensity. As detected by ELISA, rIL-2 was bound to the same extent by insolubilized C3, C3b, and C3c, while C3d coat had lower binding capacity. The receptor-binding epitope of IL-2 is intact in the complex of complement proteins and rIL-2, as demonstrated by the binding of DMS1, a monoclonal antibody reacting with the receptor site of IL-2. It is strongly suggested that C3b may play a role in the growth of CR1 positive T cells.

Distinctive role of IgG1 and IgG3 isotypes in Fc gamma R-mediated functions

Polyclonal and monoclonal anti-Rh (D) antibodies of IgG1 and IgG3 subclass were evaluated for their capacity to sensitize erythrocytes and (i) to trigger monocyte and K-cell mediated antibody-dependent cellular cytotoxicity (ADCC); (ii) to mediate binding to monocyte and lymphocyte Fc gamma R; (iii) to stimulate phagocytosis by monocytes. All antibodies were equally effective in mediating monocyte or activated U937 cell ADCC but IgG1 was more active than IgG3 in K-cell mediated ADCC. IgG3-sensitized erythrocytes inhibited IgG1-induced lysis, suggesting that each subclass engages the same Fc gamma R receptor but that lysis requires a further 'signal' that the IgG3 molecule can not deliver. Two monoclonal IgG3 anti-D antibodies were shown to have higher binding (two times) and phagocytic (three times) indices than IgG1 antibody for monocytes; similar differences were observed for polyclonal IgG1 and IgG3 antibodies. The same pattern was observed in an EA rosette assay when a total lymphocyte population was used; however, this difference was not seen with a B-cell depleted (T+ null cell) lymphocyte population.