Cellular aspects of non-specific stimulation of antibody production by capsular polysaccharide of Klebsiella pneumoniae

Comparative studies were made of the increase in the numbers of plaque-forming cells (PFC), rosette-forming cells (RFC) and haemolytic foci for erythrocyte antigens in the spleens of mice given a non-specific stimulus (the capsular polysaccharide of Klebsiella pneumoniae (CPS-K)) and an antigenic stimulus (sheep red blood cells (SRBC)). The number of direct PFC for SRBC was increased by injection of CPS-K to as high a level as that obtained by injection of SRBC. In contrast, by injection of CPS-K the numbers of indirect PFC, RFC (probably the antibody-forming cell precursors) and haemolytic foci were not increased significantly, whereas all of them were increased markedly by injection of SRBC. The maximal number of PFC in mice injected with CPS-K was approximated to the number of background RFC of the same mice. Injection of CPS-K generated 25–130 times more direct PFC for each of three kinds of erythrocyte antigens, SRBC, rabbit red blood cells and chick red blood cells, than background PFC, whereas the total number of spleen cells was not increased significantly or increased very slightly. Repeated injections of CPS-K were not significantly more effective for increase in the number of direct PFC than a single injection of CPS-K. Injection of CPS-K could generate many direct PFC in mice which had been thymectomized, irradiated and reconstituted with foetal liver cells. In mice injected with CPS-K, increase in (or maintenance of) the numbers of direct PFC and RFC were inhibited by injection of a mitogen inhibitor, vinblastine sulphate, but their sensitivities to the drug were less than those found in mice immunized with SRBC. It has been concluded from these results that in mice injected with CPS-K a large number of antibody-forming cell precursors are differentiated to direct PFC through one division or a few divisions of the individual cells, and that the inability of CPS-K to induce sufficient cell divisions of the individual precursor cells is the cause of the lack of increase in the number of indirect PFC and in immunological memory for secondary PFC responses in mice injected with CPS-K.     

Comparative studies on the actions of antigen and polyclonal B-cell activator in differentiation and proliferation of B-cells and B memory cells.

Using the capsular polysaccharide of Klebsiella pneumoniae (CPS-K) as a polyclonal B-cell activator (PBA) and sheep red blood cells (SRBC) as a T cell-dependent antigen, we compared the ability of PBA and antigen to differentiate (generate antibody-forming cells, AFC) and proliferate (generate immunological memory) virgin B cells and B memory cells. In vitro CPS-K induced the differentiation of IgM virgin B cells, IgM B memory cells and IgG B memory cells to AFC, as well as or better than SRBC. The differentiation of B memory cells to AFC by CPS-K did not require the participation of macrophages or T cells, whereas the action of SRBC depended strictly upon the helper actions of these cells. The responsiveness to CPS-K and SRBC of normal and antigen-primed spleen cells as judged by anti-SRBC PFC responses in vitro was markedly decreased after stimulation of virgin B cells and B memory cells in vivo by CPS-K injection into normal or primed mice but greatly increased after the injection of SRBC. The decrease in the responsiveness to CPS-K of spleen cells from mice treated with CPS-K appeared principally due to exhaustion of the functions of B cells and B memory cells. From the present data it has been concluded that the signals required for the differentiation and proliferation of B cells of B memory cells are different from each other, the signal for differentiation being provided by either antigen (SRBC) or PBA (CPS-K), while the signal for proliferation only by antigen.     

Effect of antigen doses and time intervals between antigen infections on secondary, tertiary and quaternary antibody responses: Establishment of hyperimmunization with bovine serum albumin in mice treated with capsular polysaccharide of Klebsiella pneumoniae

Injection of bovine serum albumin (BSA) with the capsular polysaccharide of Klebsiella pneumoniae (CPS-K) as adjuvant strongly primes mice for booster injections of BSA alone. The best doses of antigen for booster injections varied according to the frequency of antigen injections and the time interval between antigen injections. Increase in memory following the injection of BSA mixed with CPS-K adjuvant was first detected 10 days later. Memory continued to increase as the time interval after priming was increased to 90 days. Memory for the tertiary response increased after the optimum secondary antigenic stimulus. Increase in memory after the secondary antigenic stimulus was detected as early as 7 days after injection. Even secondary injection of the high dose of 50 mg of BSA did not exhaust immunological memory. Memory was also increased after the tertiary antigenic stimulus. Extremely high antibody responses (up to 49.2 mg protein per ml of antibody) could be obtained by eliciting tertiary and quaternary antibody responses optimally at the time memory was developed to the maximum level.     

Regulation of IgM memory expression by spleen cells cultured in diffusion chambers

In intact mice IgM memory to sheep erythrocytes (SRBC) has an early peak followed by a declining phase. The aim of this study was to determine whether the decay of IgM memory is regulated by the early appearance of cells producing antibodies of IgG classes. The expression of IgM memory was studied in SRBC-primed mouse spleen cells restimulated with SRBC in Millipore diffusion chambers implanted in irradiated hosts. Restimulation of spleen cells obtained from mice primed with SRBC 2 days before the beginning of the culture (D₂ cells) induced a ten-fold increase in IgM PFC above the number expected in cultures of non-primed cells. The number of IgM PFC significantly decreased in cultures of cells obtained 4 days after priming (D₄ cells). Inhibition of IgM PFC occurred in cultures of D₂ cells co-cultured with D₄ cells in single or double compartment diffusion chambers. In primed spleen cells cultured in double compartment chambers there was an inverse relationship between the number of early IgG PFC in one side and appearance of IgM PFC in the other. A 500-fold increase in antigen concentration in chambers containing D₄ cells induced a slow, but consistent, rise in IgM PFC. The data indicate the expression of IgM memory in this culture system is inhibited by the appearance of cells producing antibody of IgG classes as early as 4–10 days after primary antigenic stimulation. It is postulated that a similar process may also regulate the kinetics of IgM memory decay in the intact animal.     

Depressed antibody responses to a thymus-dependent antigen in toxoplasmosis.

The immunodepressive effect of Toxoplasma gondii infection in mice was studied, using sheep erythrocytes (SRBC) as the testing antigen and serum hemagglutinins, hemolysins, and both direct and indirect splenic plaque-forming cells (PFC) to SRBC as assays. In the primary antibody response, immunoglobulin M (IgM), hemagglutinins, and hemolysins and both IgM- and IgG-secreting PFC were depressed in animals immunized after infection. Maximum immunodepression occurred during the first 3 weeks of Toxoplasma infection. When the secondary antibody response was studied, results varied. Mice immunized with SRBC after being infected with T. gondii had a depression in both IgM and IgG PFC. Mice immunized with SRBC before being infected with T. gondii and then given a challenge dose of SRBC had a delay, but no an actual depression, in IgG hemagglutinins and hemolysins and IgG-secreting PFC. These studies show that the immunodepression associated with Toxoplasma infection is complicated, and they provide no definitive explanation for the mechanism.     

The effect of maternal antigenic stimulation upon the active immune responsiveness of their offspring.

Pregnant mice were stimulated by sheep erythrocytes (SRBC) and the active immune responses of their offspring were investigated. The offspring whose mothers were stimulated with SRBC did not develop either IgM or IgG plaque-forming cell (PFC) target cells. From the dose response of pregnant mice for inducing suppression, enough doses (10(8)-10(10) cells of SRBC) for inducing primary anti-SRBC PFC could establish suppression in the young. Both intravenous and intraperitoneal administration of SRBC induced almost complete suppression of the specific PFC response (98.6-95.2%), but only partial suppression (57.8%) was induced by subcutaneous injection. For suppression to take place, female mice had to be injected with SRBC from 2 days before fertilization to day 16 of gestation. Suppression of the PFC response was not obtained when SRBC were given 3 days before fertilization or just 24 hr before delivery. This suppressive effect on the PFC response persisted until the 15th week after birth. In these newborn mice, detectable amounts of specific anti-SRBC antibodies were found. After exchanging mothers and newborns, the stimulated newborns fostered by normal mothers were still unresponsive following antigenic stimulation, even though a specific antibody from the mother was not detected. However, normal newborns nursed by stimulated mothers could respond to SRBC injection, no matter how the specific antibodies were transferred. Possible mechanisms of immunosuppression of the PFC response by antibody transmitted by the mother to her offspring are discussed.     

Changes in the population and functional properties of antigen-specific rosette-forming B cells in antigen-primed mice by administration of polyclonal B-cell activator

We have demonstrated that the number of rosette-forming cells (RFC) in the spleens of mice primed with sheep red blood cells (SRBC) was markedly decreased by administration of a polyclonal B-cell activator (PBA) such as the capsular polysaccharide of Klebsiella pneumoniae (CPS-K). Most of the RFC estimated were shown to be of B-cell type with Ig-receptors specific for SRBC. The precursor activity for the generation of anti-SRBC antibody-forming cells (AFC) (plaque-forming cells (PFC)) was closely associated with these RFC. Moreover, the precursor activity for the generation of AFC of RFC in the spleens of mice primed with SRBC and then treated with CPS-K (SRBC-primed and CPS-K-treated mice), as estimated by anti-SRBC PFC responsiveness in vitro to CPS-K, was much less than that of the same number of RFC in the spleens of SRBC-primed mice not treated with CPS-K especially at an early stage after injection of CPS-K. This low anti-SRBC PFC responsiveness of individual RFC in the spleens of SRBC-primed and CPS-K-treated mice resulted neither from an increase in some suppressing activity in the spleens of these mice nor from a relative increase in the number of RFC of non-B-cell type or non-SRBC-specific RFC. The percentage of the number of rosette-forming PFC in the total number of RFC seemed to be slightly increased in SRBC-primed and CPS-K-treated mice. However, this cannot totally explain the mechanism of the low responsiveness of RFC-fraction of spleen cells from SRBC-primed and CPS-K-treated mice. It has been concluded from these results that the signal mediated by PBA such as CPS-K acts on B cells bearing Ig-receptors specific for antigen (RFC) and changes a large number of them to the cells lacking Ig-receptors (non-RFC) and at least in part to the cells bearing Ig-receptors but showing low responsiveness to generate AFC following further stimulus (modified RFC).     

The enhancement of antibody response by IgM antibodies is dependent on antigen-specific T helper cells

The enhancement of antibody responses by IgM antibodies administered with low doses of antigen has been studied in a T-dependent (SRBC) and an T-independent (alpha 1,6 dextran) system. It has been found that IgM anti-SRBC antibodies do not enhance a SRBC response in nude mice. The T-cell dependency was also directly demonstrated by showing the effect of IgM on T-cell priming in transfer experiments. The simultaneous injection of antigen and IgM antibody also induced a polyclonal increase of IgM, PFC, which was not due to a non-specific "adjuvant" effect of IgM, as we could not detect a similar effect on an ongoing response to HRBC in mice simultaneously given SRBC and IgM anti-SRBC antibodies. The specificity of the helper cell for either the antibody or the antigen was investigated in a response to alpha 1, 6 dextran, in which we could demonstrate antibody-specific helper T cells, but no antigen-specific help. We have found that IgM anti-dextran antibodies do not enhance and rather suppress the response of normal, high-responder mice, to dextran, suggesting that the T cells mediating the "19S enhancement" are antigen-specific. The magnitude of the enhancement response, as compared to the responses induced by either antigen or antibody alone, implies a synergistic mechanism, possibly involving antigen-specific and antibody(idiotype)-specific T helper cells.     

Competition of the actions of antigen and polyclonal B-cell activator in the induction and amplification of B-memory cell function.

Using the capsular polysaccharide of Klebsiella pneumoniae (CPS-K) as a polyclonal B-cell activator (PBA) and sheep red blood cells (SRBC) as a T-dependent antigen, the correlation of the actions of PBA and T-dependent antigen on B cells in induction and amplification of immunological memory was studied. B-memory cell function, as judged by anti-SRBC responsiveness in vitro of spleen cells of CPS-K, was amplified by the secondary injection of SRBC into SRBC-primed mice, whereas it was decreased markedly by injection of CPS-K. When CPS-K was injected simultaneously with, or 1 or 2 days before the secondary injection of SRBC, B-memory cell function was also decreased markedly. On the other hand, CPS-K did not inhibit induction of B-memory cell function when injected simultaneouly with the primary injection of SRBC. However, CPS-K inhibited induction of B-memory cell function when injected 3 days before the primary injection of SRBC. The inhibition by CPS-K of amplification of B-memory cell function in response to SRBC when CPS-K was injected simultaneously with the secondary injection of SRBC occurred markedly in mice primed with SRBC 8 days or longer before the secondary injection, whereas it was not detectable in mice primed 3 days before. It is concluded that the CPS-K-mediated signal and the SRBC-mediated signal act competitively on the same subpopulations of B cells in induction and amplification of memory, and that the susceptibility of B cells to the CPS-K-mediated negative signal changes correspondingly with their maturation stage.     

Antibody synthesis by chicken spleen cells in vitro. I. Requirements of B cells at various stages after immunization for T cells,macrophages and antigen

The requirements for T cells, macrophages and antigen during the induction of in vitro antibody responses were ascertained with chicken spleen cells obtained at various times after immunization with sheep red blood cells (SRBC). The IgM plaque-forming cell (PFC) response was T cell independent exclusively in cultures initiated 3 days after priming, but macrophage dependent at all time intervals tested. In cultures started 4 to 10 days after priming the IgG response was both T cell and macrophage independent and PFC numbers remained at a high plateau level throughout the culture period. In contrast, IgG responses initiated more than 15 days after priming showed a reversal to complete T cell and macrophage dependence and were characterized by a sharp increase in PFC numbers between days 2 and 4 of culture. Formaldehyde-fixed SRBC were immunogenic for IgG PFC 4 to 10 days after priming but failed to stimulate later IgG memory and all IgM responses. Contrasting antigen dose requirements for IgM and IgG responses were found in cultures initiated at various periods after priming. The results suggest that direct contact with fixed antigen was sufficient to maintain IgG antibody synthesizing PFC in vitro, while native antigen and cell co-operation were required for late secondary IgG and all IgM responses. These results are interpreted in terms of separate pathways of differentiation for IgM and IgG antibody-producing cells. A distinct 3rd day stage of T cell-independent but macrophage-dependent responsiveness for both classes of antibody was also defined.     

Immune deficiency of cataract Shionogi (CTS) mouse. II. Impaired in vivo T cell-mediated immune response

Our previous study demonstrated that cataract Shionogi (CTS) mice, an inbred strain related to non-obese diabetic (NOD) mice, are T lymphocytopenic and that their T cell-mediated in vitro reactions, such as proliferative responses of spleen cells to T cell mitogens and alloantigens or production of IL 2 and IL 2 receptors after stimulation of spleen cells with Con A, are greatly reduced. To confirm these in vitro characteristics, in vivo immune responses of CTS mice to T-dependent and T-independent antigens were compared with those of some reference strains including NOD mice. Antibody responses of CTS mice after one injection of a high dose (10(8)) or one or two injections of a low dose (10(5)) of sheep red blood cells (SRBC) were markedly lower than those of the reference strains. The decrease was particularly striking in the IgM antibody production at primary response to both high and low doses, and the IgG antibody production at the secondary response to low dose. Similar lower antibody production was observed in CTS mice against bovine serum albumin (BSA). Little production of IgE antibody was observed from 1 through 3 weeks after an injection of BSA plus Bordetella pertussis. IgG1 response was observed at high incidence but lower in titer than those in the reference strains. Unexpectedly, in spite of the poor antibody production to BSA, potent systemic sensitization for anaphylactic shock was easily established; incidence of lethal shock being comparable with those in the reference strains. This suggests that CTS mice are highly susceptible to the effector phase of active anaphylactic shock. Cell-mediated immunity was also impaired. Delayed type of hypersensitivity to SRBC was low, and the rejection of the skin graft from NOD mouse did not occur. In contrast to the reduced T cell-mediated responses, no difference was found between CTS and reference strains with regard to the antibody production to LPS, a T-independent antigen. These in vivo findings are consistent with the previous in vitro study.     

The allogeneic effect in (Balb/c X AKR)F1 mice

Graft-versus-host reaction induced in 6-weeks-old (Balb/c X AKR)F1 hybrid mice by the injection of parental (Balb/c) spleen cells caused the allogeneic effect, i.e. the stimulation of host IgM and IgG antibody forming cells (PFC) to SRBC and LPS. Parental lymphocytes in vivo allosensitized to AKR antigens were used to find out if the alloreactive T lymphocytes mediate the allogeneic effect. It was found that these cells influenced differently the humoral response to the antigens. The stimulation of the response to LPS was not changed in comparison with that produced by normal parental cells. On the 5th day of immunization with SRBC the stimulation of IgM PFC was stronger than IgG PFC. On the 10th day, the stimulation of IgM and IgG PFC was decreased. Moreover, it was observed that when parental splenocytes normal or allosensitized were incubated and centrifuged on allogeneic fibroblast monolayer the ability of nonadherent cells to mediate allogeneic effect was decreased or completely disappeared.     

Specific IgM Enhances and IgG Inhibits the Induction of Immunological Memory in Mice

The effect of priming mice with IgM anti-SRBC (sheep erythrocytes) together with SRBC or IgG anti-SRBC together with SRBC on the development and expression of memory cells was studied. Mice primed with specific IgM and SRBC showed a much more efficient secondary plaque-forming cell and serum antibody response after challenge with SRBC in an adoptive transfer system than did controls primed with SRBC only. The expression of this enhanced memory of IgM-primed spleen cells was counteracted by the high levels of internal IgG anti-SRBC (also the result of priming with IgM) when the mice, instead of being tested in adoptive transfers, were challenged directly. The antigen-specific feedback suppression of the primary antibody response by specific IgG antibodies was also seen to inhibit partially the development of memory cells. The suppressive effect on priming could be demonstrated both in adoptive transfer systems and after direct boost of the same mice that received the primary immunization. Both the IgM enhancement and the IgG suppression of memory cell development were antigen-specific, since no effect on the antibody response to a non-cross-reacting antigen, horse erythrocytes, was seen. The effect of these up- or down-regulations of immunological memory could be demonstrated after secondary injections as long as 90-280 days after priming.     

Immune Deficiency of Cataract Shionogi (Cts) Mouse. H. Impaired in Vivo T Cell-Mediated Immune Response

Our previous study demonstrated that cataract Shionogi (CTS) mice, an inbred strain related to non-obese diabetic (NOD) mice, are T lymphocytopenic and that their T cell-mediated in vitro reactions, such as proliferative responses of spleen cells to T cell mitogens and alloantigens or production of IL 2 and IL 2 receptors after stimulation of spleen cells with Con A, are greatly reduced. To confirm these in vitro characteristics, in vivo immune responses of CTS mice to T-dependent and T-independent antigens were compared with those of some reference strains including NOD mice. Antibody responses of CTS mice after one injection of a high dose (10⁸) or one or two injections of a low dose (10⁵) of sheep red blood cells (SRBC) were markedly lower than those of the reference strains. The decrease was particularly striking in the IgM antibody production at primary response to both high and low doses, and the IgG antibody production at the secondary response to low dose. Similar lower antibody production was observed in CTS mice against bovine serum albumin (BSA). Little production of IgE antibody was observed from 1 through 3 weeks after an injection of BSA plus Bordetella pertussis. IgG₁ response was observed at high incidence but lower in titer than those in the reference strains. Unexpectedly, in spite of the poor antibody production to BSA, potent systemic sensitization for anaphylactic shock was easily established; incidence of lethal shock being comparable with those in the reference strains. This suggests that CTS mice are highly susceptible to the effector phase of active anaphylactic shock. Cell-mediated immunity was also impaired. Delayed type of hypersensitivity to SRBC was low, and the rejection of the skin graft from NOD mouse did not occur. In contrast to the reduced T cell-mediated responses, no difference was found between CTS and reference strains with regard to the antibody production to LPS, a T-independent antigen. These in vivo findings are consistent with the previous in vitro study.     

Influence of Plasmodium chabaudi adami on the isotypic distribution of the antibody response of mice to sheep erythrocytes

The influence of an infection with P. chabaudi adami on the isotypic distribution of the in vivo antibody response to SRBC was investigated. Previous experiments suggested that the IgG1 isotype was poorly represented in the antibody response to plasmodial antigens and in the non-specific B cell response which accompanies an infection with P. chabaudi. The experiments described here indicated that although the magnitude of the total primary or secondary in vivo PFC response to SRBC was relatively unaffected by infection, the SRBC-specific IgG1 PFC response was depressed. Maximum depression of the IgG1 component of the response was observed when the priming dose of SRBC was administered at the same time as or after infection with P. chabaudi organisms. Coincident with the depression in the IgG1 response in infected mice was a corresponding increase in the SRBC-specific IgM response. The IgG1 depression was not a consequence of different kinetics of the generation of an IgG1 response, since at all times measured, the IgG1-PFC response was lower. In addition, the depressed IgG1 responses occurred only during a viable infection and could not be induced by inoculation of large amounts of irradiated erythrocytic stages of the parasite. These data suggest therefore, that there is a selective depression of IgG1 antibodies (but not those of other isotypes) regardless of antigenic specificity as a result of infection of C57BL/6 mice with P. chabaudi adami.     

Hydrocortisone and the antibody response in mice. II. Correlations between serum and antibody and PFC in thymus, spleen, marrow and lymph nodes.

The discovery that corticosteroids can alter leucocyte distribution implies that suppression of PFC in the spleen is not necessarily indicative of suppression of the antibody response as a whole. Therefore we have examined IgM and IgG serum antibodies and PFC in the thymus, spleen, femoral marrow, popliteal, thoracic and mesenteric lymph nodes of mice given a single injection of hydrocortisone acetate at various times relative to primary immunization with SRBC. Contrary to certain reports, there was close correlation between suppression of serum antibody and of splenic PFC, from which it was predicted and verified that few PFC could be detected elsewhere. In our hands, therefore, splenic PFC measurements did give an accurate indication of suppression overall. The characteristics of suppression suggest some of the underlying causes, and these are discussed.     

In Vitro Studies on the Effect of Anti-Lymphocyte Serum on the Humoral Immune Response

The effect of ALS (I), a heterologous anti-lymphocyte serum prepared against lymph node cells from rats pre-immunised with sheep erythrocytes (SRBC), on plaque forming cells (PFC) to SRBC was studied in vitro. ALS (I) reduced the number of both IgM and IgG PFC when complement was included in the reaction. This ability of ALS (I) to inhibit FFCs in vitro was absorbed out by the IgG fraction of anti-SRBC serum. Thus ALS (I) was thought to possess an anti-idiotypic antibody directed against B-cells at a later stage of differentiation.     

Immune Deficiency of Cataract Shionogi (Cts) Mouse. H. Impaired in Vivo T Cell-Mediated Immune Response

Our previous study demonstrated that cataract Shionogi (CTS) mice, an inbred strain related to non-obese diabetic (NOD) mice, are T lymphocytopenic and that their T cell-mediated in vitro reactions, such as proliferative responses of spleen cells to T cell mitogens and alloantigens or production of IL 2 and IL 2 receptors after stimulation of spleen cells with Con A, are greatly reduced. To confirm these in vitro characteristics, in vivo immune responses of CTS mice to T-dependent and T-independent antigens were compared with those of some reference strains including NOD mice. Antibody responses of CTS mice after one injection of a high dose (10⁸) or one or two injections of a low dose (10⁵) of sheep red blood cells (SRBC) were markedly lower than those of the reference strains. The decrease was particularly striking in the IgM antibody production at primary response to both high and low doses, and the IgG antibody production at the secondary response to low dose. Similar lower antibody production was observed in CTS mice against bovine serum albumin (BSA). Little production of IgE antibody was observed from 1 through 3 weeks after an injection of BSA plus Bordetella pertussis. IgG₁ response was observed at high incidence but lower in titer than those in the reference strains. Unexpectedly, in spite of the poor antibody production to BSA, potent systemic sensitization for anaphylactic shock was easily established; incidence of lethal shock being comparable with those in the reference strains. This suggests that CTS mice are highly susceptible to the effector phase of active anaphylactic shock. Cell-mediated immunity was also impaired. Delayed type of hypersensitivity to SRBC was low, and the rejection of the skin graft from NOD mouse did not occur. In contrast to the reduced T cell-mediated responses, no difference was found between CTS and reference strains with regard to the antibody production to LPS, a T-independent antigen. These in vivo findings are consistent with the previous in vitro study.     

A comparative study on the humoral immune responses in germ-free and conventional mice.

The plaque-forming cell (PFC) responses to sheep red blood cell (SRBC) dinitrophenyl-lysine-Ficoll (DNP-lys-Ficoll), and dinitrophenylated bovine serum albumin (DNP-BSA) have been studied in both germ-free and conventionally reared ICR mice. In germ-free mice, the IgG response to SRBC and the IgM and IgG responses to DNP-BSA were lower than in conventional mice, but no difference was observed in the IgM response to SRBC or the IgM and IgG responses to DNP-lys-Ficoll. Further, the number of 0-bearing cells in the spleen was smaller, and the mitogenic response of spleen cells to PHA was lower in germ-free mice than in conventional mice. These observations suggest that T cells of germ-free mice remain functionally immature.