Antibody response against Betaferon® in immune tolerant mice: involvement of marginal zone B-cells and CD4+ T-cells and apparent lack of immunological memory.

PURPOSE: The immunological processes underlying immunogenicity of recombinant human therapeutics are poorly understood. Using an immune tolerant mouse model we previously demonstrated that aggregates are a major trigger of the antidrug antibody (ADA) response against recombinant human interferon beta (rhIFNß) products including Betaferon®, and that immunological memory seems to be lacking after a rechallenge with non-aggregated rhIFNß. The apparent absence of immunological memory indicates a CD4+ T-cell independent (Tind) immune response underlying ADA formation against Betaferon®. This hypothesis was tested. METHODS: Using the immune tolerant mouse model we first validated that rechallenge with highly aggregated rhIFNß (Betaferon®) does not lead to a subsequent fast increase in ADA titers, suggesting a lack of immunological memory. Next we assessed whether Betaferon® could act as Tind antigen by inactivation of marginal zone (MZ) B-cells during treatment. MZ B-cells are major effector cells involved in a Tind immune response. In a following experiment we depleted the mice from CD4+ T-cells to test their involvement in the ADA response against Betaferon®. RESULTS: Inactivation of MZ B-cells at the start of Betaferon® treatment drastically lowered ADA levels, suggesting a Tind immune response. However, persistent depletion of CD4+ T-cells before and during Betaferon® treatment abolished the ADA response in almost all mice. CONCLUSION: The immune response against rhIFNß in immune tolerant mice is neither a T-cell independent nor a classical T-cell dependent immune response. Further studies are needed to confirm absence of immunological memory (cells).

Recombinant IL-12 prevents formation of blocking IgA antibodies to recombinant adenovirus and allows repeated gene therapy to mouse lung.

Enthusiasm for the use of recombinant adenoviruses in gene therapy has been tempered by the problematic immune responses that develop to the virus and virus-infected cells. Humoral immune responses to the input viral proteins generate neutralizing antibodies that thwart attempts to effectively administer the therapy more than once. Previous studies in murine models of gene therapy for cystic fibrosis (CF) have shown that the formation of adenoviral antibodies of the IgA subtype, a process that is dependent on T helper cells of the TH2 subset, contributes to a block in gene transfer that occurs following a second administration of virus. We show in this report that coadministration of interferon-gamma (IFN-gamma) (or interleukin-12, which activates TH1 cells to secrete IFN-gamma) with the recombinant adenovirus into the airway of C57BL/6 mice diminishes the activation of TH2 cells and formation of neutralizing antibody, allowing for efficient readministration of recombinant virus. This suggests a strategy for gene therapy of CF in which administration of a short-acting immune modulator at the time of gene therapy may be sufficient to overcome the problems of humoral immunity.

Nonspecific activation of murine lymphocytes. VI. Mediation of synergistic interaction between T and B lymphocytes by a cell-associated, reciprocally acting lymphocyte proliferation helper.

The mechanism by which B and T lymphocytes interact synergistically in the proliferative response to 2-mercaptoethanol (2-ME) as a mitogen was investigated in cultures of C3H/St spleen cells. The interaction between these cells required physical contact between the collaborating cell types, and was not mediated by the release of a soluble factor into the culture supernate. Sonicates of spleen cells which had been activated with optimal concentrations of 2-ME for 24 h and then washed extensively, stimulated the uptake of tritiated thymidine and morphological blast transformation of fresh, unstimulated cells. This activity was found to reside within the soluble fraction of the activated cells, and to activate cells optimally at a ratio of 1 naive cell: 1 activated cell-equivalent. This reciprocally-acting lymphocyte proliferation helper (RALPH) activity was produced by B cells as well as by T cells, with a kinetic peak at 48 h of culture. RALPH activity was produced by viable but not by nonviable cells incubated with 2-ME, and was nondialyzable. It could not be induced by the B-cell mitogens lipopolysaccharide, polyinosinic-polycytidilic acid, or purified protein derivative of tuberculin, or by the T-cell mitogen concanavalin A. RALPH isolated from T cells activated B cells exclusively, while that from B cells acted predominantly upon T cells, possibly with a nonspecific effect on B cells. A model for the cellular interactions involved in the amplification of the proliferative response to 2-ME is described.

Virus-replicating T cells in the immune response of mice. II. Characterization of T cells capable of replicating vesicular stomatitis virus.

Immunocytological properties of the splenic T cell (Tv) which develop into virus plaque-forming cells in response to the antigenic challenge in vitro were investigated in relation to the properties of helper T cells and suppressor T cells in antibody response. Tv was observed in spleen around 1 wk after the intravenous injection of mice with 10(7) sheep erythrocytes. This contrasted with the finding that both helper T cells and suppressor T cells developed as early as 3 days after the immunization. Tv was proliferative in response to the antigenic stimulation, whereas helper T-cell activity could be expressed without cell division. Development of Tv to virus plaque-forming cells was much more dependent on macrophages than the generation of helper activity. Tv was found in nylon wool adherent fraction, whereas helper T cell was found in both nylon adherent and nonadherent fractions. Tv belongs to the short-lived and nonrecirculating T-cell population (T1), whereas the major part of helper T cells belongs to the long-lived and recirculating T-cell population (T2). These results strongly suggest that vesicular stomatitis virus infect and replicate in the different subset(s) of T cell(s) to which the major part of helper T cells belong.

Cellular and genetic restrictions in the immunoregulatory activity of alpha-fetoprotein. II. Alpha-fetoprotein-induced suppression of cytotoxic T lymphocyte development.

Alpha-fetoprotein (AFP), a major component of fetal and newborn sera, was shown to exert significant immunosuppressive activity on the in vitro generation of cytotoxic T lymphocytes (CTLs). This suppression proved independent of the suppressibility of the mixed leukocyte culture activation phase, since strain combinations whose proliferative responses were refractive to AFP-induced suppression also failed to develop demonstrable CTLs in the presence of AFP. Several strain combinations were also found in which normal generation of CTLs occurred in cultures containing AFP. This refractive nature correlated with the presence of nonsuppressible lymphocyte-stimulating alloantigenic systems on the stimulating cell population. These data provide the basis for proposing several possible mechanisms for AFP-induced suppression of T-cell-mediated cytotoxicity, as well as suggesting that the primary target of this suppression is the proliferating helper T cell precommited to respond towards the major histocompatibility complex-associated lymphocyte-activating determinants.

Glucocorticoids administered in vivo inhibit human suppressor T lymphocyte function and diminish B lymphocyte responsiveness in in vitro immunoglobulin synthesis.

The effects of corticosteroid given in vivo on human lymphocyte subpopulation function were investigated using an in vitro system of pokeweek mitogen-stimulated immunoglobulin production. Peripheral blood lymphocytes were obtained from normal volunteers before and 4 h after the intravenous administration of methylprednisolone. Unfractioned peripheral blood lymphocytes showed a consistent decrease (mean congruent with 50%) in immunoglobulin and total protein synthesis after steroid administration. Utilizing separated thymus-derived (T) and bone marrow-derived (B) lymphocyte fractions, the pathophysiology of this alteration in immunoglobulin production was elucidated. B lymphocytes obtained after steroid treatment showed a markedly diminished immunoglobulin response (20% of normal) to normal T lymphocytes and to normal T cells that had been irradiated to remove suppressor T lymphocyte function. All major classes of immunoglobulin (IgG, IgM, and IgA) were affected. T lymphocytes procured after steroid administration were capable of providing normal amounts of T cell help for B cells in immunoglobulin production. However, suppressor T lymphocyte activity, observed with normal T lymphocytes at high T to B cell ratios, was absent from the post-steroid T lymphocytes. This loss of suppressor T lymphocyte function was not due to the presence of excess help as irradiated pre- and poststeroid T cells provided equal amounts of helper activity. On recombining the poststeroid treatment B cells, which are hyporesponsive in immunoglobulin synthesis, with the posttreatment T lymphocytes, which lack suppressor activity, diminished amounts of immunoglobulin were produced which correlate well with the effects observed with unseparated cells. Thus, corticosteroids have differential effects on the lymphocyte populations involved in immunoglobulin biosynthesis. B cell responsiveness is diminished, suppressor T lymphocyte activity is removed, and helper T lymphocyte function is unaffected.

Dysfunctions of pokeweed mitogen-stimulated T and B lymphocyte responses induced by gammaglobulin therapy.

Lymphocytes obtained from nonimmuno deficient children treated with commercially available preparations of gammaglobulin failed to proliferate and to mature into plasma cells in vitro after stimulation with pokeweed mitogen. The influence of the treatment on lymphocyte functions varied according to the cell population considered. A T helper cell activity was detected in these patients but only in the cell subset bearing receptors for IgG after irradiation. T lymphocytes exerted a suppressive effect that disappeared after irradiation or incubation at 37 degrees C. The suppressive cells were found among E rosette-forming cells depleted of leukocytes bearing receptors for IgG. Their suppressive effect was expressed only in the presence of normal radioresistant T lymphocytes that did not bear Fc receptors for IgG. Similar dysfunctions could be induced in vitro by incubation of normal T and B lymphocytes with gammaglobulin preparations. Because F(ab)'2 fragments or deaggregated preparations of gammaglobulin failed to activate T suppressor lymphocytes, this activation was likely triggered by attachment of Fc portion of denatured IgG to the corresponding membrane receptor. This activation step was prostaglandin E(2)-dependent, suggesting that activated monocytes were involved in the activation process. B lymphocyte responses appeared directly inhibited by attachment of denatured gammaglobulin on membrane Fc receptor. Our observations suggest that immunological effects of gammaglobulin therapy are not limited to antibody transfer, since it also induces subtle modifications of in vitro pokeweed mitogen-stimulated T and B cell responses. These modifications must be considered in interpreting results obtained in immunodeficient patients investigated under gamma-globulin therapy.

Characterization of a non-T, non-B human lymphocyte (L cell) with use of monoclonal antibodies. Its regulatory role in B lymphocyte function.

These studies investigate the role of L lymphocytes in regulating terminal B lymphocyte differentiation. L cells have abundant Fc IgG receptors and comprise 10--15% of human peripheral blood mononuclear cells (PBMC). L cells lack the conventional markers of B and T lymphocytes and in culture, do not develop into B cells, T cells, or macrophages. Additionally, use of monoclonal antibodies failed to detect on L cells, surface antigens specific for B cells, T cells, and macrophages. In these studies, purified L cell subpopulations depleted of macrophages were co-cultured with autologous PBMC in the presence of pokeweed mitogen and at the end of 8 d, development of intracytoplasmic immunoglobulin (Ig) was determined. L cells were depleted of B and T cells by rosetting techniques and, in addition, by cytotoxicity techniques using monoclonal-specific antisera to T cells. In 14 individuals, L cells when co-cultured with PBMC, enhanced Ig synthesis by 83% +/- 62 SD, and also enhanced cell proliferation. Radiated L cells lost enhancing properties. To study the role of their high density Fc IgG receptors, L cells pretreated with IgG antibody-sensitized erythrocytes were used (i.e., after lysis of rosettes). Such L cells significantly inhibited Ig synthesis (by greater than 50%) despite promoting cell proliferation. Antibody-sensitized erythrocyte-rosetted macrophages did not inhibited Ig synthesis. Thus, positive and negative influences can be mediated by the same cell, depending on the state of Fc-receptor stimulation. Such cells may play a more prominent role in "feed-back" regulation of Ig synthesis by virtue of having abundant Fc IgG receptors.

Studies of T- and B-cell interactions in adult patients with combined immunodeficiency.

Cellular interactions involved in the pathogenesis of hypogammaglobulinemia were studied in six patients with common variable immunodeficiency. Amounts of immunoglobulin (Ig)G and IgM in the supernate of pokeweed mitogen-stimulated cocultures of normal and immunodeficient mononuclear cells were measured by radioimmunoassays. Mononuclear cells from three of six patients inhibited Ig production of normal B cells (P < 0.005). When purified patient and normal T cells were added to B cells in various autologous or allogeneic combinations, it was observed that immunodeficient T cells (AT) from four patients suppressed normal IgM synthesis. Allogeneic normal T cells did not provide help for B cells from these same immunodeficient patients. In two patients, autologous T cells were able to help autologous B-cell IgM synthesis in vitro. In five patients, AT cells inhibited normal B-cell IgG synthesis. Removal of T cells bearing Ia determinants or T cells with Fc-IgG receptors did not diminish the suppressive effect of AT cells on normal B-cell Ig synthesis. Addition of indomethacin, a prostaglandin synthetase inhibitor, did not abrogate the suppressive effect of immunodeficient mononuclear cells. Addition of hydrocortisone succinate (10 muM) did reverse the suppressive effect of AT cells on IgM production in one patient; however, no in vitro reversal of suppressor cell effect was recorded in five. Suppression by immune-deficient T cells was eliminated by 2,000 rad of x-ray irradiation in three patients. After x-ray irradiation immunedeficient T cells could function as helpers of normal B cells.

Human CD57+ germinal center-T cells are the major helpers for GC-B cells and induce class switch recombination.

BACKGROUND: The function of CD57+ CD4+ T cells, constituting a major subset of germinal center T (GC-Th) cells in human lymphoid tissues, has been unclear. There have been contradictory reports regarding the B cell helping function of CD57+ GC-Th cells in production of immunoglobulin (Ig). Furthermore, the cytokine and co-stimulation requirement for their helper activity remains largely unknown. To clarify and gain more insight into their function in helping B cells, we systematically investigated the capacity of human tonsil CD57+ GC-Th cells in inducing B cell Ig synthesis. RESULTS: We demonstrated that CD57+ GC-Th cells are highly efficient in helping B cell production of all four subsets of Ig (IgM, IgG, IgA and IgE) compared to other T-helper cells located in germinal centers or interfollicular areas. CD57+ GC-Th cells were particularly more efficient than other T cells in helping GC-B cells but not naive B cells. CD57+ GC-Th cells induced the expression of activation-induced cytosine deaminase (AID) and class switch recombination in developing B cells. IgG1-3 and IgA1 were the major Ig isotypes induced by CD57+ GC-Th cells. CD40L, but not IL-4, IL-10 and IFN-gamma, was critical in CD57+ GC-Th cell-driven B cell production of Ig. However, IL-10, when added exogenously, significantly enhanced the helper activity of CD57+ GC-Th cells, while TGF-beta1 completely and IFN-gamma partially suppressed the CD57+ GC-Th cell-driven Ig production. CONCLUSIONS: CD57+CD4+ T cells in the germinal centers of human lymphoid tissues are the major T helper cell subset for GC-B cells in Ig synthesis. Their helper activity is consistent with their capacity to induce AID and class switch recombination, and can be regulated by CD40L, IL-4, IL-10 and TGF-beta.

Ganglioside GD1a enhances immunoglobulin production by human peripheral blood mononuclear cells.

We previously reported that ganglioside GD1a greatly enhanced spontaneous immunoglobulin (Ig) production by human peripheral blood mononuclear cells (PBMC) in vitro. We herein examined the mechanism for the stimulatory effect of GD1a.PBMC from healthy volunteers were cultured with GD1a. The amounts of IgG, IgM, and IgA and cytokine activity in the culture supernatants were measured by enzyme-linked immunosorbent assays. Proliferation was determined by [3H] thymidine uptake.GD1a at 10(-6) M increased IgG, IgM, and IgA production by PBMC 2.10-fold, 2.10-fold, and 2.23-fold above the control values, respectively. GD1a did not affect the proliferation and viability of PBMC. GD1a did not alter Ig production of B cells alone. Anti-interleukin-6 (IL-6) or anti-IL-10 antibody each partially blocked the GD1a-induced enhancement of Ig production by PBMC, and the addition of both antibodies completely blocked the enhancement. GD1a increased IL-6 and IL-10 production of monocytes without altering those of T cells or B cells. The supernatant from GD1a-treated monocytes enhanced B cell Ig production to a greater extent than that from medium-treated monocytes. The supernatant-mediated effect of GD1a was partially blocked by anti-IL-6 or anti-IL-10 antibody, and the addition of both antibodies completely blocked the GD1a effect. GD1a-induced increases of IL-6 and IL-10 production in monocytes were both blocked by Ca(2)+/calmodulin (CaM)-dependent phosphodiesterase (PDE) inhibitors, 8-methoxymethyl-3-isobutyl-1-methylxanthine and vinpocetin, but not by other signal-transducing enzyme inhibitors. The culture with GD1a enhanced Ca(2)+/CaM-dependent PDE activity in monocytes. These results suggest that GD1a may indirectly enhance B cell Ig production in whole PBMC by increasing IL-6 and IL-10 production of monocytes via promoting Ca(2)+/CaM-dependent PDE activity.

Pyridoxine supplementation: effect on lymphocyte responses in elderly persons.

The effect of pyridoxine supplementation on lymphocyte responsiveness was investigated in 15 persons aged 65-81 y. Eleven subjects received 50 mg/d pyridoxine HCl (PN). Four subjects received a placebo. Lymphocyte proliferation to T and B cell mitogens, lymphocyte subpopulations with monoclonal antibodies, and plasma pyridoxal 5'-phosphate (PLP) were measured before and after 1 and 2 mo of supplementation. After 1 and 2 mo plasma PLP levels increased by 195 +/- 88 nM and 201 +/- 84 nM, respectively, in subjects receiving PN. With PN supplementation, lymphocyte proliferation increased significantly in response to phytohemagglutinin (p less than 0.01), pokeweed mitogen (p less than 0.01), and Staphylococcus aureus (Cowain I) (p less than 0.05). For PN-treated subjects with low presupplement plasma PLP levels, lymphocyte blastogenesis also increased significantly (p less than 0.01) in response to concanavalin A. Percentages of T3+ and T4+ but not T8+ cells increased significantly (p less than 0.05) in PN-treated subjects. These results suggest that improving vitamin B-6 status is important in stimulating immunocompetence in the elderly.

Suppressive effect of mizoribine on humoral antibody production in DBA/2 mice.

The mode of action of mizoribine (MZR) as a B cell inhibitor was studied using DBA/2 mice. Its in vitro administration significantly delayed the primary response in hemagglutinin production against sheep erythrocytes by suppressing the IgM antibody formation. In vitro plaque-forming cell (PFC) response against both T-dependent and T-independent antigens, such as TNP-SRBC and TNP-Brucella abortus, was dose-dependently suppressed by MZR. Since PFC formation by the T-depleted fraction of splenocytes was likewise suppressed, MZR may inhibit humoral antibody response by directly affecting the B cells (and/or macrophages) as well as by modulating the regulatory T cells. MZR may only act on a certain stage of the cell cycle of B lymphocytes following antigenic stimulation. It may not interfere with the initial antigen recognition or with mature B cells.

Lack of influence of cyclosporine on antigen presentation to lysozyme-specific T cell hybridomas.

Cyclosporine (CsA) was tested for its ability to inhibit antigen presentation by spleen cells or the B lymphoma line A20 to T cell hybridomas specific for hen egg-white lysozyme (HEL). Antigen-presenting cells (APC) were treated with CsA or nonimmunosuppressive derivatives thereof during or prior to encounter with antigen. The APC were then washed extensively and incubated in CsA-free medium for 6 hr before the T hybridoma cells were added. Under these conditions, CsA had no effect on antigen presentation up to the cytostatic regimen (1 microgram/ml). Omission of the 6-hr interval between CsA treatment of APC and the addition of T hybridoma resulted in inhibition of interleukin 2 production, although the CsA concentrations required were 10-75-fold higher than the ones inhibiting T cells directly (IC50: 100-150 ng/ml vs. 2-10 ng/ml). The responses to both HEL and a synthetic peptide of HEL sequence 105-120 were inhibited, indicating that the step influenced by the drug was not antigen-processing. The nonimmunosuppressive derivatives remained ineffective under these conditions. The results illustrate that the carryover of CsA from APC to T cells can mimic a drug effect on antigen presentation. Therefore, the demonstration of a CsA effect on antigen presentation can only be considered as conclusive when the readout of APC function is not a T cell response.

Clonal activation of cytotoxic lymphocyte precursors by monoclonal anti-CD3 antibody: analysis of feeder cell requirements.

Activation of cytotoxic lymphocyte precursors (CLP) by the mitogenic monoclonal anti-CD3 antibody OKT3 was studied under limiting dilution (LD) culture conditions. One out of 2-6 E-rosette-purified T cells gave rise to a cytotoxic T cell (CTL) clone when cultured in the presence of OKT3 (0.2-2 ng/ml), recombinant IL-2 (100 U/ml), and irradiated feeder cells. Clonal CLP activation was optimally supported by a combination of E-rosette-depleted non-T feeder cells with small numbers of T cells added back. Among the cell lines tested, Fc-receptor-bearing monocytic cell lines U937 and HL-60 were efficient feeder cells whereas T cell lines (Jurkat, Molt-4, Ke37) did not support clonal CLP activation. These data indicate that clonal activation of CLP and differentiation into cytotoxic effector cells under LD culture conditions are critically influenced by the type and number of feeder cells used.

HLA class-II-mediated homotypic aggregation: involvement of a protein tyrosine kinase and protein kinase C.

Homotypic aggregation of B-lymphocytes, B-cell lines and class-II-positive T cells via HLA class II molecules was examined. Signaling via DR antigens induced rapid aggregation in a dose- and time-dependent manner, maximum and stable aggregation was induced within 20 minutes. On the contrary, rapid signaling via DP or DQ required prestimulation with either PMA or anti-sIg. Aggregation was temperature and energy dependent. [Ca2+] and [Mg2+] concentrations and an intact cytoskeleton were required while neither mRNA or protein synthesis were required. Furthermore, FACS analysis revealed that aggregation was not directly correlated with cell surface expression of HLA class II molecules. Our results demonstrate that aggregation was mediated through a protein tyrosine kinase (PTK)-dependent pathway that preceded activation of protein kinase C (PKC) and failure to generate either the PTK signal or the PKC signal prevented aggregation. The contribution of a tyrosine kinase was further demonstrated by the total inhibition of aggregation following treatment with an anti-CD45 mAb.

Adjuvant effects of orally administered saponins on humoral and cellular immune responses in mice.

We have described adjuvant effects of orally administered Quillaja saponins on the immune responses of mice fed inactivated rabies antigen (AG). The in vivo lymphocyte proliferation in mice fed antigen + saponin (AG + SAP) was significantly greater than that in mice fed antigen (AG) alone. Further, the mitogen-induced cell proliferative responses in animals primed with AG + SAP was markedly increased compared with those in the AG group. These changes in clonal expansion were associated with an enhanced helper T cell (Th) and B cell co-operation. The in vivo cell proliferation and in vitro mitogen-induced responses of mice fed AG + SAP correlated with enhanced antibody synthesis. In mice fed saponin alone, there were significant increases in clonal expansion and lymphocyte function. Our present data indicate that the immunocompetence in animals fed AG + SAP was indeed evoked by saponins. Cytotoxic T lymphocyte activity in mice fed SAP or AG + SAP was detected 7 days after booster, in contrast to 21 days in mice fed AG alone. The natural killer cell activity in mice fed SAP alone was greatly enhanced and persisted for an extended period of time.

The effects of indomethacin administration during pregnancy on women's and newborns' T-suppressor lymphocyte activity and on HLA class II expression by newborns' leukocytes.

It has been previously shown that T lymphocytes from human newborns and pregnant women exert a suppressive activity when assayed on the PWM-induced B cell maturation. The mechanisms of the suppression have remained entirely unknown. Prostaglandin E2, known to trigger T-cell mediated suppressive activity, may be involved. We took advantage of the treatment of pregnant women with indomethacin, because of premature labor or hydramnios, to investigate the role of prostaglandins in the activation of T suppressor (TS) activity. Administration of indomethacin (250 mg/day for 1-7 weeks, then 150 mg/day for 3-12 weeks) during the third trimester of pregnancy, abrogated the TS activity in the nine women and the three newborns tested. Abrogation of TS activity by indomethacin therapy led to normal PWM-induced B cell maturation in pregnant women but not in newborns. Moreover, the low expression of HLA class II antigens observed on normal newborn B lymphocytes and monocytes was corrected in newborns from indomethacin-treated mothers. Our results strongly suggest that prostaglandins may play a role in induction of TS activity observed in normal pregnant women and newborns and in the decreased expression of HLA class II antigens on newborns' leucocytes. Both phenomena could play a role in immunological interactions between mother and fetus.

The effects of T cell subpopulations and recombinant interleukin (IL)-2 on peripheral B cell function in patients with myasthenia gravis.

Myasthenia gravis is an autoimmune disease mediated by antibodies to acetylcholine receptors (AChR) of skeletal muscle. The production of anti-AChR antibodies has been shown to be T cell dependent. To elucidate the mechanism(s) of anti-AChR antibody production in myasthenic patients, we studied the effects of regulatory T cells and/or IL-2 on the differentiation of AChR-primed B cells, with use of AChR stimulation for the induction of anti-AChR antibodies in vitro. Our data suggest that CD8+ T cells possess some complicated functions. CD8+ T cells could not only provide help for B cells to secrete anti-AChR antibody, but also possibly inhibit response of CD4+ T cells or kill B cells, then repress anti-AChR antibody production in MG patients. There might be some defect either in the number or function of CD8+ T cell in MG patients. Exogenous IL-2 could completely restore the suppression activity of CD8+ T cells in anti-AChR antibody production in vitro.