Immunological Activity of B Cell Subsets Responding to T-

Transcription

1 INFECTION AND IMMUNITY, Aug. 1984, p /84/ $02.00/0 Copyright C 1984, American Society for Microbiology Vol. 45, No. 2 Mechanism of Lipopolysaccharide-Induced Immunosuppression: Immunological Activity of B Cell Subsets Responding to T- Dependent or T-Independent Antigens in Lipopolysaccharide- Preinjected Mice TAKEHIKO UCHIYAMA,* YUSUKE KAMAGATA, AND MORIMASA YOSHIOKA Department of Microbiology, Tokyo Women's Medical College, Shinjuku-ku, Tokyo 162, Japan Received 19 January 1984/Accepted 27 April 1984 Spleen cells from mice preinjected with high doses of bacterial lipopolysaccharide did not generate antitrinitrophenyl (TNP) plaque-forming cells in vitro to the T-dependent antigen, TNP-sheep erythrocytes, but did generate fully plaque-forming cells to the T-independent antigens, TNP-Ficoll and TNP-Brucella abortus. The immunological activity of B cells from such lipopolysaccharide-preinjected mice was analyzed in the present study. T cell-depleted spleen cells from mice injected with 30,ug of lipopolysaccharide 3 days previously did not respond to combined stimulation with TNP-sheep erythrocytes and concanavalin A-induced T cell-replacing factor and had no suppressive activity on normal T cell-depleted spleen cells. Splenic B cells, which were separated from T cells and macrophages from mice injected with 30,ug of lipopolysaccharide 3 days previously, responded only partially (about 25% of the control response) to combined stimulation with TNP-sheep erythrocytes and concanavalin A-induced cell-replacing factor in the presence of normal macrophages, but responded fully to TNP-B. abortus, regardless of the presence of normal macrophages. These results indicate that B cells responding to the T-dependent antigens are rendered unresponsive to antigenic stimulation in mice preinjected with lipopolysaccharide, whereas B cells responding to the T-independent antigens are kept intact. Bacterial endotoxin or lipopolysaccharide (LPS) can either enhance or suppress the immune response. When LPS is injected into mice at the same time as antigen, antibody production is enhanced (7, 18). In contrast, when the antigenic stimulation is delayed several days after LPS injection, the response is suppressed (12-14, 19). The cellular basis for the adjuvant activity of LPS has been investigated extensively, and the effect of LPS on T cells (4, 18, 20), B cells (18), and macrophages (5, 18) has been reported. However, less is known about the suppressive effect of LPS at the cellular level. Recently, the regulatory role of LPS, which was thought to be derived from gastrointestinal bacterial flora, on the immune system of normal mice was reported; T cells from conventional LPS-responder mice suppressed the antibody response to trinitrophenyl (TNP)-LPS of B cells from germfree mice, but T cells from germfree mice did not (9). In addition, feeding sheep erythrocytes (SRBC) by gastric intubation induced immunological tolerance in LPS-responder mice, but the same treatment induced immunity rather than immunological tolerance in LPS-nonresponder mice (10). These results indicate that LPS may exert its role in an immunosuppressive way in LPS-responder animals, including humans, that have continuous exposure to LPS via gut bacteria. Therefore, analysis of the mechanism of the suppressive effect of LPS seems important for clarification of the actual effect of LPS on the immune system of animals. The present study was undertaken to clarify the suppressive effect of LPS on the antibody response at the cellular level, focusing on the immunological activity of B cells from mice that had been injected with LPS several days previously. The experimental results indicate that the B cell subpopulation responding to the T-dependent (TD) antigens is rendered unresponsive to antigenic stimulation in LPS-prein- * Corresponding author. 367 jected mice, whereas other 1B cell subpopulations responding to the T-independent (TI) antigens are kept intact. MATERIALS AND METHODS Animals. C57BL/6 mice were purchased from Shizuoka Laboratory Center, Hamamatsu-shi, Shizuoka-ken, Japan. Female mice were used at 6 to 8 weeks of age. LPS. A phenol extract of Escherichia coli 055:B5 obtained from Difco Laboratories (Detroit, Mich.) was repurified with the hot phenol method (16) and suspended in phosphate-buffered saline (PBS), and the indicated doses were injected intraperitoneally into mice. Antigens. As the TD antigen, heavily substituted TNP- SRBC were used; as the TI antigens, TNP-Ficoll and TNP- Brucella abortus (TNP-Ba) were used. TNP-SRBC was prepared according to a slight modification of the method of Rittenberg et al. (15). One milliliter of packed SRBC was reacted with 80 mg of sodium trinitrobenzenesulfonic acid (Tokyo Kasei Kogyo, Tokyo) in 10 ml of PBS (ph 7.4) for 30 min at room temperature and washed three times with PBS. TNP-Ba was prepared by the method of Mond et al. (11) and was a gift of T. Tadakuma, Keio University, Tokyo. The preparation of TNP-Ficoll was described previously (17). Lightly substituted horse erythrocytes (TNP-HRBC) used for detection of anti-tnp plaque-forming cells (PFC) were prepared as follows: 1 ml of packed HRBC was reacted with 10 mg of sodium trinitrobenzenesulfonic acid in 10 ml of PBS for 10 min and washed three times with PBS. In vitro tissue culture and PFC assay. Nucleated spleen cells (7 x 106 to 15 x 106 per culture) were cultured in duplicate in petri dishes (35 by 10 mm; Falcon Plastics, Oxnard, Calif.) at 37 C in a 5% CO2 incubator. Click's medium, used as culture medium, contained 7.5% fetal calf serum and 5 x 10-5 M 2-mercaptoethanol (17). The cultures were stimulated with TNP-immunogens, and in some experi-

2 368 UCHIYAMA, KAMAGATA, AND YOSHIOKA ments concanavalin A (ConA)-induced T cell-replacing factor (TRF) was added at several volumes. At 96 h after the start of the culture, duplicate cultures were assayed individually for direct PFC by the hemolytic plaque technique (17) with TNP-HRBC and HRBC, and the mean PFC per culture was given. Specific anti-tnp PFC were calculated by subtracting the number of plaques formed with HRBC from those formed with TNP-HRBC. Removal of macrophages and T cells from spleen cells. Macrophages were removed by serial passages through petri dishes and Sephadex G10 columns (18). Macrophages present in the recovered cells after these procedures were present in less than 0.1% as estimated by nonspecific esterase staining. T cells were removed by treating spleen cells with the monoclonal anti-theta antibody produced from hybridoma H013 cells (8). The antibody was a gift of H. Ishikawa, Keio University, Tokyo. Spleen cell suspensions (2 x 107 cells per ml) in Hanks solution were reacted with the antibody at a final dilution of 1:200 at 4 C for 30 min, centrifuged, and suspended in guinea pig serum at a final dilution of 1:10. After incubation at 37 C for 30 min, the treated cells were washed three times with Hanks solution. The recovered cells had lost completely the ability to respond to T-cell mitogens, ConA, or phytohemagglutinin, but had kept the ability to respond to LPS to a degree comparable to untreated spleen cells. Preparation of macrophage monolayers. Macrophage monolayers were prepared by incubating spleen cells (8 x 106 cells per dish) in 35-mm petri dishes for 2 h at 37 C. Nonadherent cells were removed, and the dishes were washed with fresh medium and used immediately. The dishes contained about 5 x 105 adherent cells, 80% of which were nonspecific esterase staining positive. Preparation of TRF. ConA-induced TRF (21) was prepared by incubating spleen cells (8 x 106 cells per ml) with 3,ug of ConA per ml and harvesting the culture fluid 20 h later. The culture fluid had the ability to enhance the in vitro antibody response to TNP-SRBC of T cell-depleted spleen cells from normal mice. Graded volumes of the ConAinduced TRF were added to cultures 48 h after start of the culture, since at this time TRF induced the highest PFC response. RESULTS Effect of LPS preinjection on in vitro PFC response of whole spleen cells to TD and TI antigens. Preliminary experiments were done to define conditions for the LPS-induced immunosuppression and to learn whether suppression was exerted in response to the TI antigens as well as to the TD antigen, TNP-SRBC (19). Spleen cells from mice injected with graded doses of LPS ranging from 0.1 to 50,ug 3 days previously were cultured in vitro and stimulated with TNP-SRBC or the TI antigens, TNP-Ficoll and TNP-Ba. The anti-tnp PFC response was assayed on day 4 of culture. Ten experiments were done, and similar results were obtained in all of them. The response to TNP-SRBC was sharply reduced by preinjection of increasing doses of LPS (Fig. 1). In contrast, the response to TNP-Ficoll and TNP-Ba was not suppressed at all over this LPS dose range. The duration of the effect of LPS preinjection was determined by injecting 30,ug of LPS on various days before culture. Spleen cells from each group were cultured with TNP-SRBC or TNP-Ba. Two experiments were done, and similar results were obtained in both. Unresponsiveness to TNP-SRBC was observed 1 day after LPS injection, and recovery from the unresponsiveness began by day 8 (Fig. 2). In contrast, the response to TNP-Ba 4i a) OP, zqll-. INFECT. IMMUN Dose of LPS (jg) FIG. 1. Effect of LPS dose preinjected on in vitro PFC response. Spleen cells (10 x 106 cells per culture) from mice injected with several doses of LPS ranging from 0.1 to 50,ug 3 days previously (two mice per group) were cultured and stimulated with (0) 3 x 105 TNP-SRBC, (A) 10 ng of TNP-Ficoll, or (0) a 3 x 104 dilution of TNP-Ba. was enhanced for the first 3 days after LPS injection and decreased to the control level by day 14. The different susceptibility to LPS preinjection of the in vitro PFC response to the TD and TI antigens did not depend on the amount of antigen added. Graded doses of TNP-SRBC, TNP-Ficoll, or TNP-Ba were added to cultures of control mice and mice injected with 50,ug of LPS 3 days previously. Four experiments were done, and similar results were obtained in all of them (Fig. 3 shows representative results). Spleen cells from LPS-preinjected mice responded only slightly to TNP-SRBC at high doses: 22 and 14% of the control at 10 x 105 and 20 x 105 TNP-SRBC, respectively. The response to TNP-Ficoll and TNP-Ba was not suppressed at all at any dose in this range; rather, an enhanced response was observed in the response to TNP-Ba. The findings of these three experiments indicate that the suppression induced by LPS preinjection is restricted to the response to the TD antigen. The following experiments focused the suppression of the immunological activity of B cells. Immunological activity of T cell-depleted spleen cells from LPS-preinjected mice to combined stimulation with TNP- SRBC and TRF. We investigated whether T cell-depleted spleen cells from LPS-preinjected mice responded to combined stimulation with TNP-SRBC and TRF. Spleen cells from control mice and mice injected with 1 or 30 ixg of LPS 3 days previously were depleted of T cells by anti-theta antibody and complement. Whole and T cell-depleted spleen cells from these mouse groups were stimulated with TNP- SRBC. The cultures of T cell-depleted spleen cells were stimulated further with ConA-induced TRF 48 h after the start of the culture. The anti-tnp PFC response was as-

3 VOL. 45, 1984 LPS EFFECT ON B CELLS RESPONDING TO TD OR TI ANTIGENS ) 4-) E-) 04 1 C-) z H 4-) C. None Days after LPS Injection FIG. 2. Duration of LPS effect on in vitro PFC response. Spleen cells (10 x 106 cells per culture) from control mice and mice injected with 30 p.g of LPS at various days before the start of the culture (two mice per group) were stimulated in vitro with (0) 6 x 105 TNP- SRBC or (0) a 3 x 104 dilution of TNP-Ba. sayed on day 4 of culture. Five experiments were done, and similar results were obtained in all of them. In a group preinjected with 30 p,g of LPS, whole and T cell-depleted spleen cells did not respond to TNP-SRBC and combined stimulation with TNP-SRBC and TRF, respectively (Table 1). In contrast, in a group preinjected with 1,ug of LPS, T cell-depleted spleen cells responded fully to combined stimulation with TNP-SRBC and TRF, whereas whole spleen cells did not respond to TNP-SRBC. In the following experiments, we investigated the unresponsiveness induced with high doses of LPS. The following three different mechanisms could account for the unresponsiveness of T cell-depleted spleen cells from mice preinjected with a high dose of LPS: generation of suppressor cells, inactivation of macrophages as accessory cells in antibody response, and inactivation of the B cell subpopulation responding to the TD antigen. Absence of suppressor activity in T cell-depleted spleen cells from LPS-preinjected mice. To test the possibility that suppressor cells might be generated in LPS-preinjected mice and cause the unresponsiveness of T cell-depleted spleen cells, the following experiments were done. T cell-depleted spleen cells from control mice and mice injected with 30 p.g of LPS 3 days previously were cultured separately or together and stimulated with TNP-SRBC and ConA-induced TRF. If suppressor cells were present in the LPS-preinjected group, the magnitude of response of the combined culture 100 O TNP-SRBC (xlo 5/culture TNP-Ba jxlo dilutipn/culture) TNP-Ficoll (ng/culture) FIG. 3. Independence on antigen concentration of LPS effect on in vitro PFC response. Spleen cells (7 x 106 cells per culture) from ( ) control mice and (--- ) mice injected with 50,ug of LPS 3 days previously (two mice per group) were cultured and stimulated with several doses of (0) TNP-SRBC, (A) TNP-Ficoll, or (0) TNP- Ba. from two groups would be lower than the mean value of the responses of the two groups. Two experiments were done, and similar results were obtained in both. The magnitude of the response of the combined culture was much higher than the mean value of the two groups (Table 2), indicating that suppressor cells were not responsible for the LPS-induced unresponsiveness of T cell-depleted spleen cells to combined stimulation with TNP-SRBC and TRF. Immunological activity of B cells from LPS-preinjected mice. As T cell-depleted spleen cells contain B cells and macrophages, the unresponsiveness could be caused by inactivation of macrophages or B cells. To test directly the immunological activity of B cells from LPS-preinjected mice, the anti-tnp PFC response of B cell preparations that were separated from T cells and macrophages was compared TABLE 1. Immunological activity of whole and T cell-depleted spleen cells from LPS-preinjected mice' Source of spleen Whole T cell-depleted spleen cells cells spleen TRF added (RI) cells Normal mice ,660 2,600 2,950 LPS-preinjected ,200 1,400 3,400 mice (1 jig) LPS-preinjected mice (30,ug) a Whole cell (1.5 x lo' cells per culture) and T cell-depleted spleen cells (8 x 106 cells per culture) from normal mice (six mice) and mice injected with 1 or 30 Rg of LPS 3 days previously (five mice per group) were stimulated in vitro with 6 x 10- TNP-SRBC. The cultures of T cell-depleted spleen cells were stimulated further with several doses of ConA-induced TRF 48 h after the start of the culture.

4 370 UCHIYAMA, KAMAGATA, AND YOSHIOKA TABLE 2. Inability of T cell-depleted spleen cells from LPSpreinjected mice to suppress the immune response of normal T cell-depleted spleen cells to the stimulation with TNP-SRBC and TRF" No. of T cell-depleted spleen cells from: Normal LPS- TRF added (,ul) mice preinjected mice x ,580 5, X ,350 5 x x ,580 (1,210)b 4,700 (3,195)b a T cell-depleted spleen cells from normal mice (five mice) and mice injected with 30 jig of LPS 3 days previously (four mice) were cultured either separately at 10 x 106 cells per culture or together at 5 x 106 cells of each group per culture and stimulated with combination of 6 x 105 TNP-SRBC and ConA-induced TRF. TRF was added 48 h after the start of the culture. b Mean value between the response of two groups cultured separately. between control and LPS-preinjected groups. Adherent cell monolayers (macrophages) were prepared from normal spleen cells. Splenic B cells from normal mice and mice injected with 30,ug of LPS 3 days previously were prepared by serial passages on glass petri dishes and Sephadex G10 columns, treated with anti-theta antibody and complement, and cultured with or without macrophages. The cultures were stimulated with TNP-SRBC or TNP-Ba. ConA-induced TRF was added 48 h after the start of culture in TNP-SRBCstimulated cultures. Four experiments were done, and similar results were obtained in all of them. The response to combined stimulation with TNP-SRBC and TRF was dependent on the presence of normal macrophages, and B cells from LPS-preinjected mice responded only partially to combined stimulation with TNP-SRBC and TRF in the presence of normal macrophages (Table 3). The PFC response of the LPS-preinjected group in the presence of normal macrophages was about 25% of the control group at TRF doses of 20 and 60 RI. In contrast, the same B cell preparation from the LPS-preinjected group responded to TNP-Ba to a degree comparable to B cells from the control group, regardless of the presence of normal macrophages. The results indicate that the B cell subpopulation responding to the TD antigen in mice preinjected with high doses of LPS is in a state unresponsive to antigenic stimulation, whereas other B cell subpopulations responding to the TI antigens are kept intact. DISCUSSION The present study was undertaken to clarify the suppressive effect of LPS on the antibody response at the cellular level. We focused the study on the immunological activity of INFECT. IMMUN. B cells from mice that had been injected with high doses of LPS (30 to 50,ug) several days previously. The results indicate that the B cell subpopulation responding to the TD antigen is rendered unresponsive to antigenic stimulation by LPS preinjection, whereas other B cell subpopulations responding to the TI antigens are kept intact. This conclusion is based on the following observations: (i) whole spleen cells from LPS-preinjected mice responded poorly to the TD antigen, TNP-SRBC, but responded fully to the TI antigens, TNP-Ficoll and TNP-Ba (Fig. 1, 2, and 3); (ii) T cell-depleted spleen cells from LPS-preinjected mice did not respond to combined stimulation with TNP-SRBC and TRF (Table 1); and (iii) purified B cells from LPS-preinjected mice responded only partially to combined stimulation with TNP-SRBC and TRF in the presence of normal macrophages, whereas the same B cell preparation responded fully to TNP-Ba, regardless of the presence of normal macrophages (Table 3). The presence of suppressor cells, reported by us and others (13, 19), does not appear to be responsible for the findings presented in Tables 1 and 3; T cell-depleted spleen cells from LPS-preinjected mice did not suppress the response of normal T cell-depleted spleen cells to combined stimulation with TNP-SRBC and TRF (Table 2). B cells can be separated into two major subpopulations on the basis of responsiveness to either TD or TI antigens (1-3). The latter can be separated further into two subpopulations on the basis of responsiveness to TI-1 antigens such as TNP-Ba and TI-2 antigens such as TNP-Ficoll (6). Considering the presence of B cell subpopulations, the reasonable explanation for the results of the present study seems to be the selective inactivation of the B cell subpopulation responding to TD antigens. Previously we reported that T cell-depleted spleen cells from LPS-preinjected mice responded poorly in vitro to TNP-SRBC in the presence of splenic T cells from normal mice, but reponded fully in the presence of splenic T cells from SRBC-primed mice (19). Also, Portnoi et al. reported that whole spleen cells from LPS-preinjected mice did not respond in vitro to TNP-SRBC, but responded fully to TI antigens, TNP-LPS, and TNP-dextran, and that T celldepleted spleen cells from these mice responded fully to combined stimulation with TNP-SRBC and ConA-induced TRF (14). Findings in these two studies indicated that B cells responding to TD and TI antigens are equally intact in LPSpreinjected mice. We have no available data to explain the discrepancies between the present study and the other two studies, but we consider that the difference of relative strength of LPS activity in experimental animals in those studies may be one of the causes. The purity of LPS we obtained from a trading company seems not so high, because TABLE 3. Immunological activity of purified B cells from LPS-preinjected mice' Source of B cells Normal mac- Stimulated with TNP-SRBC, TRF added rophagesstmltdwh (al) Stimulated with TNP-Ba Normal mice Absent ,840 Present 300 3,700 8,700 1,450 LPS-preinjected Absent ,340 mice Present 230 (76.6%) 950 (25.6%) 2,250 (25.8%) 1,710 (117.9%) a Both macrophages and T cell-depleted spleen cells (10 x 106 cells per culture) from normal mice (seven mice) and mice injected with 30,ug of LPS 3 days previously (five mice) were cultured in the presence or absence of the macrophage monolayer obtained from 8 x 106 normal spleen cells. The cultures were stimulated with combination of 5 x 105 TNP-SRBC and ConA-induced TRF or with a 3 x 104 dilution of TNP-Ba. TRF was added 48 h after start of the culture. The numbers within parentheses represent the percentage of PFC response calculated from the formula (PFC in the LPS-preinjected group in the presence of macrophages/pfc in the control group in the presence of macrophages) x 100.

5 VOL. 45, 1984 LPS EFFECT ON B CELLS RESPONDING TO TD OR TI ANTIGENS 371 we found that the LPS was mitogenic to B cells from an LPSnonresponder C3H/HeJ mouse strain. We repurified the LPS with the hot phenol method and got LPS without mitogenicity to B cells from this strain. In the present study we used this repurified LPS, and in the previous study we used LPS without repurification. In the present study, in a group preinjected with a small dose of LPS (1,ug), T cell-depleted spleen cells responded fully to combined stimulation with TNP-SRBC and TRF, although whole spleen cells did not respond to TNP-SRBC (Table 1). The results are similar to the two studies in which LPS was injected at doses of 10,ug (14) and 50,ug (19). The strength of LPS activity seems to be higher in the present study than in the other two studies. The present study showing that a B cell subpopulation responding to a TD antigen is selectively inactivated in LPSpreinjected mice does not mean that T cells and macrophages are not affected by LPS preinjection. We reported previously the generation of suppressor T cells as well as the failure of helper T cell induction in LPS-preinjected mice (19). With regard to the effect on macrophages, we found that macrophages from LPS-preinjected mice were ineffective in cooperating with T cells in ConA-induced TRF production and with B cells in an antibody production (manuscript in preparation). It seems likely that LPS affects the immune system by inducing multifocal immunosuppression. ACKNOWLEDGMENTS We thank Takushi Tadakuma and Hiromichi Ishikawa for providing us with TNP-Ba antigen and anti-theta monoclonal antibody. LITERATURE CITED 1. Andersson, B., and H. Blomgren T-cell response to polyvinylpyrrolidone is linked to maturity of B cells. Nature (London) 253: Galanaud, P., M.-C. Crevon, and J. Dormont Effect of azathioprine on in vitro antibody response. Differential effect on B cells involved in thymus-dependent and independent response. Clin. Exp. Immunol. 22: Gorczynski, R. M., and M. Feldman B cell heterogeneity-difference in the size of B lymphocytes responding to T- dependent and T-independent antigens. Cell. Immunol. 18: Hamaoka, T., and D. H. Katz Cellular site of action of various adjuvants in antibody response to hapten-carrier conjugates. J. Immunol. 111: Hoffman, M. K., S. Koenig, R. S. Mittler, H. F. Oettgen, P. Ralph, C. Galanos, and U. Hammerling Macrophage factor controlling differentiation of B cells. J. Immunol. 122: Huber, B. T B cell differentiation antigens as probes for functional B cell subsets. Immunol. Rev. 64: Johnson, A. G., S. Gaines, and M. Landy Studies on the 0 antigen of Salmonella typhosa. V. Enhancement of antibody response to protein antigens by the purified lipopolysaccharide. J. Exp. Med. 103: Kubota, E., H. Ishikawa, and K. Saito Evidence for a cloned hybridoma (HO ) producing 2 kinds of antibody molecules different in specificity, one directed to thyl.2 gene products and the other to antigen on a subpopulation thyl.1 thymocytes. J. Immunol. 127: McGhee, J. R., H. Kiyono, S. M. Michalek, J. L. Babb, D. L. Rosenstreich, and S. E. Mergenhagen Lipopolysaccharide (LPS) regulation of the immune response: T lymphocytes from normal mice suppress mitogenic and immunogenic response to LPS. J. Immunol. 124: Michalek, S. M., H. Kiyono, M. J. Wannemuehler, L. M. Mosteller, and J. R. McGhee Lipopolysaccharide (LPS) regulation of the immune response: LPS influence on oral tolerance induction. J. Immunol. 128: Mond, J. J., I. Scher, D. E. Mosier, M. Baese, and W. E. Paul T-independent response in B cell-defective CBA/N mice to Brucella abortus and to trinitrophenyl (TNP) conjugates of Brucella abortus. Eur. J. Immunol. 8: Nakano, M., M. J. Tanabe, M. Saito, and T. Shimizu Immunosuppressive effect of bacterial lipopolysaccharide on antibody response. Jpn. J. Microbiol. 20: Persson, U Lipopolysaccharide-induced suppression of the primary immune response to a thymus-dependent antigen. J. Immunol. 118: Portnoi, D., I. Motta, and P. Truffa-Bachi Immune unresponsiveness of spleen cells from lipopolysaccharide-treated mice to particulate thymus-dependent antigen. I. Evidence for differentiation signal defect. Eur. J. Immunol. 11: Rittenberg, M. P., and K. L. Pratt Anti-trinitrophenyl (TNP) plaque assay. Primary response of Balb/c mice to soluble and particulate immunogens. Proc. Soc. Exp. Biol. Med. 132: Tanabe, M., T. Saito, and M. Nakano Methods of LPS preparations, p In Methods in immunological experiments. Japanese Society for Immunology, Tokyo. 17. Uchiyama, T B cell tolerance: B cells rendered tolerant are present in the immune system in a potentially responsive form. Microbiol. Immunol. 24: Uchiyama, T Modulation of immune response by bacterial lipopolysaccharide (LPS): roles of macrophages and T cells in in vitro adjuvant effect of LPS on antibody response to T celldependent and T cell-independent antigens. Microbiol. Immunol. 26: Uchiyama, T., and D. M. Jacobs Modulation of immune response by bacterial lipopolysaccharide (LPS): multifocal effects of LPS-induced suppression of the primary antibody response to a T-dependent antigen. J. Immunol. 121: Uchiyama, T., and D. M. Jacobs Modulation of immune response by bacterial lipopolysaccharide (LPS): cellular basis of stimulatory and inhibitory effect of LPS on the in vitro IgM antibody response to a T-dependent antigen. J. Immunol. 121: Watson, J., L. A. Aarden, and I. L. Lefkovits The purification and quantitation of helper T cell-replacing factors secreted by murine spleen cells activated by concanavalin A. J. Immunol. 122: