A micromethod for the induction and assay of specific in vitro antibody responses by human lymphocytes

Interleukin 6 is not required for antigen-specific antibody responses by human B cells

Interleukin-6 (IL-6) is a late-acting differentiation factor for human B cells activated by polyclonal mitogens such as pokeweed mitogen (PWM) and Staphylococcus aureus Cowan strain I, but its role in specific antibody responses has not been established. We show here that IL-6 has no consistent effect on specific antibody responses by tonsillar mononuclear cells (TMC) stimulated with influenza virus. A blocking IL-6 antibody also had no effect on antibody production, suggesting that endogenous IL-6 production was not required. In control experiments, this antibody inhibited PWM-stimulated immunoglobulin secretion and proliferation of the IL-6-dependent B cell line B9. A requirement for IL-6 in responses of unfractionated TMC may have been disguised by the presence of T cells. To overcome this problem, we investigated the effect of IL-6 on specific antibody production by T-depleted B cells stimulated with antigen in the presence of IL-2, which is a T cell replacing factor (TRF) for human B cells. Specific antibody production was restored by IL-2, but not IL-6. Neither IL-6 nor anti-IL-6 antibody had any consistent effect on specific antibody production by purified B cells stimulated with antigen and TRF. These experiments show that IL-6 does not have a significant role in antigen (influenza virus)-specific antibody responses by human B lymphocytes.

The generation and activation of memory class I MHC restricted cytotoxic T cell responses to influenza A virus in vivo do not require CD4+ T cells

The cellular requirements for the generation and activation of anti-influenza memory class I MHC restricted Tc cell responses were studied by selectively reconstituting lethally irradiated mice. The generation of memory Tc cells was investigated by using unprimed splenocytes to reconstitute infected, lethally irradiated mice; the activation of memory Tc cells was tested by using primed splenocytes to reconstitute uninfected, lethally irradiated mice. It is shown here that depletion of CD4+ T cells from donor cells did not reduce Tc cell responses in recipient mice. Depletion of CD8+ T cells from donor splenocytes prevented the memory Tc cell responses. Thus depletion of CD4+ T cells had no effect on the generation and activation of the memory Tc cells in vivo.

Human T cell-replacing factor(s): a comparison of recombinant and purified human B cell growth and differentiation factors

Conditioned medium from phytohemagglutinin-activated T cells contains T cell-replacing factor(s) (TRF) able to restore specific antibody responses by human blood or tonsillar B cells which have been thoroughly depleted of T cells. Of twelve recombinant cytokines tested as possible candidates for TRF in conditioned media, namely human recombinant interleukin (hrIL) 1 alpha and beta, hrIL2, hrIL3, hrIL4, hrIL5, hrIL6, hrIFN-alpha and -gamma, hr granulocyte macrophage colony-stimulating factor (hrGM-CSF) and tumor necrosis factor (hr TNF)-alpha and -beta only IL2 was found to have TRF activity. In addition, a semi-purified low molecular weight B cell growth factor (BCGFlow) also had TRF activity. As the commercially available BCGFlow is known to contain low concentrations of IL2, IFN-gamma, TNF and GM-CSF as impurities, it was important to exclude these as being responsible for the TRF activity. At the concentrations present in BCGFlow (less than 0.2 U/ml), IL2 was not active in the TRF assay. In contrast, a combination of IL2 (0.2 U/ml), IFN-gamma (50 U/ml), TNF-alpha (50 U/ml) and TNF-beta (100 U/ml) did have TRF activity suggesting that B cells could be made to respond to low doses of IL2 by the presence of other cytokines. Although this finding raises important questions about the nature of TRF in conditioned medium, the TRF activity of BCGFlow was unlikely to be due to such a synergistic combination of cytokines for the following reasons. First, in several experiments, responses were obtained with BCGFlow, but not with IL2 or combinations of IL2 with IFN and TNF. Second, antibody to IL2 was found to inhibit the TRF activity of IL2 but not of BCGFlow. Taken together these findings show that two distinct cytokines (IL2 and BCGFlow) are TRF for human B cells. However, some combinations of cytokines can also have TRF activity underlining the complexities which can arise from working with semi-purified rather than recombinant factors.

Interleukin 2 and low molecular weight B cell growth factor are T cell-replacing factors for different subpopulations of human B cells

Both recombinant human interleukin 2 (rhIL 2) and low molecular weight B cell growth factor (BCGFlow) were shown to be T cell-replacing factors (TRF) in specific antibody responses to influenza virus by human blood and tonsillar B cells. When B cells were separated into high and low-density populations on Percoll gradients at 1.074 kg/l, IL 2 was found to act as a TRF only on the low-density B cells, whereas BCGFlow was a TRF for high-density B cells with a lesser effect on low-density B cells. Both populations of B cells responded well in the presence of T cells. The high-density B cells could not be activated to respond to IL 2 by either IL 1, rhIL 4 or by a CD22 monoclonal antibody known to enhance B cell activation. In contrast, a 24-h preincubation with T cells and antigen appeared to prime high-density B cells to respond to IL 2. These results show that high-density B cells can in fact respond to TRF, and that IL 2 and BCGFlow act on different populations of B cells which may be defined by prior exposure to T cells.

Recombinant human interleukin 5 is an eosinophil differentiation factor but has no activity in standard human B cell growth factor assays

Following the observation that mouse interleukin 5 (IL5) is active as a B cell growth factor (BCGF) as well as an eosinophil differentiation factor, this work was carried out to test recombinant human IL5 for BCGF activity. A highly active, partially purified batch of recombinant human IL5 was prepared and tested for BCGF activity in four laboratories. This batch gave a 50% endpoint of 1:77,450 in the human eosinophil differentiation assay, 1:983 in the mouse eosinophil differentiation assay and 1:42 in the mouse BCL1 assay, thus demonstrating that, like mouse IL5, human IL5 has cross-species activity. By comparison with the assays in the mouse this batch would be expected to have 50% maximal human BCGF activity of about 1:4000. In each assay a known positive factor was used as a positive control, and there was no inhibitory activity in the preparation. However, despite the activity towards the mouse B cell lymphoma, the results showed no detectable activity in a panel of assays used to identify human BCGF and B cell differentiation factors. These assays included (a) proliferation assays with tonsillar or splenic B cells in the presence of the co-stimulators anti-mu or phorbol myristate acetate; (b) a restimulation assay in which tonsillar B cells are first activated with either Staphylococcus aureus Cowan 1 or a mixture of phorbol dibutyrate and ionomycin, or splenic B cells are first activated with anti-mu; (c) production of immunoglobulin by B cells in a restimulation assay with Staphylococcus aureus Cowan 1; (d) production of immunoglobulin by the Epstein-Barr virus-transformed B lymphoblastoid CESS cell line; (e) the ability to stimulate proliferation of chronic lymphocytic leukemia (B-CLL) cells freshly explanted from three different patients; (f) the ability to stimulate the B lymphoma (L4) cell line and the mature B cell (HBF1) line, and (g) the ability to replace T cells in specific antibody responses. It therefore seems unlikely that recombinant human IL5 is either a growth or a differentiation factor for human B cells, and raises the interesting question of the biological significance of the BCGF activity of this factor in the mouse.

Influenza-specific antibody-secreting cells and B cell memory in the murine lung after immunization with wild-type, cold-adapted variant and inactivated influenza viruses

The development of regional B cell responses was studied in mice immunized intranasally with different influenza virus vaccines. The ca-variant virus was 100-fold less efficient than the parental virus in the induction of influenza virus-specific antibody secreting cells (ASCs) in the lung and failed to induce ASCs in the spleen. The ca-variant virus was also less efficient in priming for secondary IgG and IgA responses generated in vitro in both lung and spleen cell cultures. Protection against homotypic challenge in mice immunized by different vaccine strategies correlated with the development of pulmonary B cell responses rather than splenic responses. In particular, protection correlated with the presence of ASCs and IgG and IgA memory in the lung at the time of challenge.

T cell help in human antigen-specific antibody responses can be replaced by interleukin 2

Recombinant IL 2, and immunosorbent/high performance liquid chromatography-purified interleukin 2 (IL 2) obtained from the human T cell leukemic line Jurkat, but not interferon-alpha or -gamma, were able to substitute for T cells in specific antibody responses to influenza virus by T cell-depleted (E-) human peripheral blood mononuclear cells, and resulted in antibody formation equivalent to that obtained in the presence of T cells. The antibody response was shown to be antigen specific by using two non-cross-reacting strains of influenza virus (A/X31 and B/HK). IL 2 in this assay therefore functions as a T cell-replacing factor. Less than 1% of T (UCHT1+) cells were present in the E- preparations, and this number did not increase during the 7-day culture with antigen and IL 2. Because the frequency of T helper cells for X31 is known to be less than 5 X 10(-5), this low number of contaminating cells excluded indirect action of IL 2 through antigen-specific T helper cells. Three to four times less IL 2 was required for antibody production by E- cells than was needed for optimal proliferation by an IL 2-dependent T cell line. Moreover, the concentration of anti-Tac required for 50% inhibition of the IL 2-induced antibody response was 50 times less than required for 50% inhibition of IL 2-dependent proliferation by the T cell line. But when T cells were added back to the E- cells, the anti-Tac inhibition curve shifted back to that obtained with the T cell line. In cell labeling experiments, Leu 11+ cells but not HNK1+ cells were increased in E- cells cultured with antigen and IL 2. This increase in Leu 11+ cells was abolished by prior passage of the E- cells through Sephadex G-10 columns without affecting the IL 2-induced antibody response. From these experiments we conclude that IL 2 can replace T cells in specific antibody responses, and that the IL 2 effect is not mediated indirectly through T cells or large granular lymphocytes.

Antigen-specific suppression of human antibody responses by allogeneic T cells. III. Role of the major histocompatibility complex

Specific antibody responses obtained in vitro from human blood mononuclear cells (PBM) were profoundly suppressed by allogeneic T cells. Experiments carried out with combinations of cells from HLA identical siblings, and HLA identical but unrelated donors, showed that suppression depended upon HLA incompatibility between responding PBM and allogeneic Ts. In order to map the specific HLA loci concerned, a series of experiments were undertaken using combinations of cells from a large number of HLA typed donors. Significant suppression was found to occur in every combination of HLA incompatible cells tested, including those with nonidentity at HLA-A, B, DR, A and DR, or B and DR, suggesting that suppression can be generated by nonidentity at class I or class II loci. With some HLA-A homozygous donors, however, a dominant role for class I (HLA-A) antigens was indicated by the finding of one-directional suppression in combinations where the HLA-A locus was seen as foreign by one partner only (A3,----A2,3; and A2----A2,26). Similar one-directional suppression was also seen with cells from a pair of siblings who were HLA identical except for a single A locus antigen arising from an HLA-A/B recombination (A3,----A3,1). These results indicate an important, but not exclusive role for class I MHC antigens in the activation of allogeneic Ts. The way in which this occurs is unknown, but one possibility is that it results from the activation of normal antigen-specific Ts by the interaction of their receptors for self-MHC with cross-reacting alloantigens.

Antigen-specific suppression of human antibody responses by allogeneic T cells. II. Cell interactions involved in the generation of suppression

Specific antibody responses to influenza virus were obtained in vitro from human blood mononuclear cells (PBMC). Antibody production in these cultures was profoundly suppressed by the addition of allogeneic T cells with the surface phenotype Leu2a+ (CD8+), Leu8-. Suppression by allogeneic T suppressor (Ts) cells required interactions only between T-depleted B (E-) cells and allogeneic Leu2a+. No evidence was obtained for T-T cell interactions, or for Ts inducer cells similar to those described for nonspecific antibody responses to pokeweed mitogen. Moreover, allogeneic E+, or allogeneic Leu2a+ cells were able to suppress specific antibody responses by E- cells when help was provided by T cell-replacing factor showing that the target of suppression was the responding E- cells, and not T helper cells. In contrast to allogeneic T cells, allogeneic E- cells did not suppress antibody production when added to cultures of unfractionated PBMC (E- + E+). That is, Ts cells activated to allogeneic E- were unable to suppress antibody production by the syngeneic E- cells present in the same culture tube. This result shows that alloactivated Ts cells were specific for the allogeneic E- target cells, and that suppression was not mediated by nonspecific allogeneic effects. Allogeneic Ts cells therefore differ from Ts cells in pokeweed mitogen responses by their specificity, and by their activation in the absence of Ts inducer cells.

HLA class II restriction governing cell cooperation between antigen-specific helper T lymphocytes, B lymphocytes and monocytes for in vitro antibody production to influenza virus

To study HLA class II compatibility requirement for in vitro antibody production to influenza virus, semipurified T lymphocytes, B lymphocytes and monocytes from HLA-typed responder donors were used. The presence of the three subpopulations was required for antibody production while a mixture of only two of those was ineffective. When using fresh T lymphocytes which exert an allogeneic suppressive effect and may also exhibit allogeneic helper activity, it was not possible to conclude an HLA class II-linked restriction of T-B cell cooperation although there was a suggestion of it. However, a grown H3 hemagglutinin-specific T cell line (L2), previously shown to be restricted by HLA-DR molecule (DR1) for interaction with antigen-presenting cells and devoid of allogeneic reactivity, exerts an HLA class II-restricted helper activity. This was demonstrated by various combinations of HLA-DR semi-compatible or incompatible B lymphocytes and/or monocytes with L2 T cells. The restriction element was identified as an HLA-DR determined since HLA-DC-compatible, HLA-DR-incompatible B lymphocytes were not helped by L2 T cells. In addition, monoclonal anti-HLA-DR but not anti-HLA-DC antibodies directed to the relevant specificity did inhibit the antigen-specific helper activity. We present evidence that not only T monocyte but also T-B and/or T-B-monocyte interactions are HLA class II restricted.

In vivo effects of cyclosporine on influenza A virus-infected mice

Cyclosporine (cyclosporin A, CsA) administered to mice substantially affects their immune response to an influenza A virus infection. If treated with CsA for 21 days, the mouse lungs contain high titers of virus which are cleared more slowly than in controls. Indicators of pathological damage--lung weight, extent of consolidation, fine morphology, and the extent of infiltration of dividing cells into the lung--showed that administration of CsA greatly decreased the level of inflammation. The production of hemagglutination-inhibiting (HI) antibody was delayed but reached almost control levels and NK cell activity in the lung was also comparable to control levels. In contrast, a delayed-type hypersensitivity (DTH) response to the virus could not be elicited in the CsA-treated, infected mice at 6 or 12 days after infection. Cytotoxic-T-cell (Tc-cell) activity was present in the lungs of such mice though its appearance was delayed and the activity recovered was less than that of the control infected mice. If administered with a dose of virus lethal for normal mice. CSA-treated mice survived, probably due to the greatly reduced level of immunopathological damage in the infected lung.

T cell-replacing factor in specific antibody responses to influenza virus by human blood B cells

In man, B cell maturation factors obtained from T cells or T cell lines have been shown to induce antibody formation in mitogen or anti-immunoglobulin activated B cells, and in some continuous B cell lines, but the relationships between these factors and B cell differentiation factors in antigen-specific antibody responses is unclear. We have now shown that supernatants from phytohemagglutinin-activated tonsil cells, or from the Gibbon Ape T cell line MLA-144, can substitute for T cells in the specific antibody response by human blood B cells to influenza virus. Thus, T cell-depleted non-rosette-forming (E-) cells prepared from peripheral blood mononuclear cells made antibody when cultured with antigen and factor together, whereas control cultures of E- cells with either antigen or factor alone did not. Moreover, E- cells cultured with factor and influenza virus strain A/X31 made antibody to A/X31, but not the non-cross-reacting strain, B/HK (and vice versa) showing that the response was antigen specific. The activity in these supernatants, therefore, fulfilled the functional definition of T cell-replacing factor (TRF). The possibility that interleukin 2 (IL 2) present in the TRF-containing supernatants was expanding residual T cells in the E- preparations to provide normal T cell help was excluded in three different ways. First, E- cells depleted of T (Leu4+) cells to undetectable levels made normal amounts of antibody when cultured with antigen and TRF. Secondly, a limiting dilution technique was employed to show that help in cultures of E- cells and TRF was not mediated by antigen-specific T helper cells. Thirdly, TRF-containing supernatants depleted of IL2 retained activity, whereas purified IL2 was inactive. Preliminary purification of TRF by gel filtration on Ultrogel AcA54 columns showed that all the activity eluted in a single peak between 35 000 and 43 000 molecular weight. This distinguishes human TRF from IL 2 and from other B cell maturation factors with a molecular weight range of 15 000-20 000 which act on continuous B cell lines. In addition to TRF, supernatants from phytohemagglutinin-activated tonsils also contained a factor which could induce polyspecific IgM production, but only in cultures containing significant numbers of T cells. This additional activity may have been due to IL 2, and provides an explanation for the apparent T cell-dependent effects sometimes observed in experiments designed to test B cell differentiation factors on T cell-depleted normal B cells.

Antigen-specific suppression of human antibody responses by allogeneic T cells. I. Frequency and antigen specificity of allogeneic suppressor T cells and their role in major histocompatibility complex-controlled genetic restriction

Specific antibody responses to influenza virus were obtained in vitro from human blood mononuclear cells (PBM). The addition of allogeneic T cells to responding PBM profoundly suppressed antigen-induced antibody responses, but had no effect on pokeweed mitogen (PWM)-induced polyclonal Ig formation. This raised the possibility that suppression by allogeneic T cells may result from the activation of antigen-specific T suppressor (Ts) cells rather than nonspecific allogeneic effects. The frequency of allogeneic Ts in PBM from a number of different donors, estimated in a series of limiting dilution analyses, was found to range from 0.8 X 10(-5) to 4.5 X 10(-5), which is more typical of antigen-specific than alloreactive T cells. By adding limiting numbers of allogeneic T cells to antibody-forming cultures stimulated simultaneously with two non-cross-reacting antigens (influenza A and B strain viruses A/X31 and B/HK), it was possible to demonstrate suppression of the response to one antigen, but not the other, in the same culture well. Moreover, the frequency of wells in which the response to both antigens was suppressed was not significantly different from that predicted from the calculated frequency of specific allogeneic Ts. These results show that allogeneic suppression was antigen specific, and was not due to non-specific allogeneic effects. By separating T cell preparations into Leu-3a+ (helper) and Leu-2a+ (suppressor/cytotoxic) T cell subsets, suppression was shown to be mediated by a radiosensitive Leu-2a+ T cell. The removal of Leu-2a+ cells from T cell preparations abrogated the suppressor effect and permitted T cell collaboration with B cells, across an HLA-A, -B, and -DR barrier. This result shows that in at least some combinations, suppression rather than major histocompatibility complex restriction is the reason for the failure of allogeneic T and B cells to collaborate in T cell-dependent antibody responses.

Induction of tolerance in influenza virus-immune T lymphocyte clones with synthetic peptides of influenza hemagglutinin

Antigen-specific human T cell clones specific for defined peptides of influenza A hemagglutinin were found to be rendered unresponsive by incubation with moderately high concentrations of antigen. This was the case whether the synthetic peptide antigen was present for the duration of the culture or the cloned T cells were preincubated with antigen for 3-18 h at 37 degrees C, before stimulation with T-depleted irradiated sheep erythrocyte non-rosette-forming lymphocytes (E-) pulsed with the optimal dose of peptide. Tolerance could not be overcome by culture with various numbers of E- cells and antigen. The induction of unresponsiveness was antigen specific, since it depended upon incubation with the appropriate peptide recognized by that clone. In addition, the tolerant T cells remained unresponsive to stimulation with the specific peptide for at least 7 d after induction even though maintained in culture in the presence of T cell growth factor. This state of antigen-specific unresponsiveness is akin to immunological tolerance. Furthermore, the experiments reported here demonstrate that the helper T cell clone can be inhibited by the relevant peptide in the absence of any suppressor cells or their precursors. This suggests that antigen-induced unresponsiveness need not always depend on the presence of suppressor T cells. The induction of tolerance in T cell clones does not result in early T cell death, since cells that no longer proliferate in response to the specific antigen and accessory cells still proliferate in response to T cell growth factor.

An influenza virus matrix protein-specific human T cell line with helper activity for in vitro anti-hemagglutinin antibody production

A human helper T cell line (F14m) activated by the matrix protein purified from A/X31 influenza virus has been developed. After activation by antigen for 7 days, and reculture with matrix protein and irradiated autologous feeder cells, the cells obtained from an in vivo influenza virus-immunized donor have been growing in the presence of interleukin 2 for more than 7 months. The cells all belong to the helper-inducer T cell subset and most of them express surface membrane HLA-DR antigens. A small number (approximately 10(3)) of F14m T cells provided optimal help for 1 X 10(5) autologous T-depleted lymphocytes for production of anti-A/X31 but not anti-B/HK antibodies. The F14m T cells produce soluble factors (S14m) able to help B cells to secrete anti-A/X31 antibodies. F14m and S14m were shown to help antibody production to hemagglutinin when cultured with B cells and the whole virus. The specificity of the T cell line for type-A matrix protein was confirmed by the ability of S14m to provide help for anti-A/JAP (A/H2/N2) but not for anti-B/HK antibody production. These data provide evidence for matrix protein-specific T helper cells and factors able to provide help for antibody production against hemagglutinin, a distinct protein of the same virus.