218
B.R. Bloom, G. Stoner, J. Gaffney, E. Shevach and I. Green
tion method the sensitivity increased only 10-30 times; this advantage, however, was achieved by very simple means requiring almost n o change of the laboratory protocol. Polymers like dextran and PEG have been known t o enhance the precipitation of antibodies with molecular antigens [ 12, 131. We have shown that the presence of PEG affects both the interaction of DNP-bacteriophage with anti-DNP antibodies (present in concentrations 1 - 10 ng/ml), and the interaction of complexes with anti-Ig antibodies. The effect exerted o n the latter interaction seems t o present the major contribution to the final result.
Eur. J. Immunol. 1975.5: 218-220
5. References 1 Makela, O.,Immunology 1966. 10: 81. 2 Haimovich, J., Hurwitz, E., Novik, N. and Sela, M., Biochim. Biophys. Acta 1970. 207: 125.
3 Andrieu, J.M., Mamas, S . and Dray, F., Eur. J. Immunol. 1974. 4 : 417. 4 Eisen, N.H., Belman, S . and Carsten, M.E., J. Amer. Chem. SOC. 1953. 75: 4583. 5 Franzk, F. and Simek, L., Eur. J. Immunol. 1971. 1 : 300. 6 Segal, S., Globerson, A., Feldman, M., Haimovich, J. and Sela, M., J. Exp. Med. 1970.131: 93.
7 Jormalainen, S. and Makela, O., Eur. J. Immunol. 1971. 1: 471.
It is likely that the effect of PEG is of general nature, because it has been observed both with the system porcine anti-DNP plus ovine anti-pig Ig antibodies and rabbit antiDNP plus porcine anti-rabbit Ig antibodies. The fact that lower antigen concentrations inhibit DNP-bacteriophage inactivation in the presence of PEG is conceivable, since a lower antibody concentration is employed when the antigen assay is performed in the presence of PEG.
12 Hellsing, K. and Laurent, T.C., Acta Chem. Scund. 1964. 18: 1303.
Received October 7, 1974.
1 3 Harrington, J.C., Fenton, J.W.11 and Pert, J.H., Immunochemistry 1971.8: 431.
B.R. Bloom', G. Stoner'. Judith Gaffney+, E. Shevach" and I. Greeno Albert Einstein College of Medicine', of Allergy Bronx and and Infectious Diseaseso,National Institutes of Health, Bethesda
8 Haimovich, J., Hurwitz, E., Novik, N. and Sela, M., Biochim. Biophys. Acta 1970. 207: 115. 9 Krummel, W.M. and Uhr, J.W., J. Immunol. 1969.102: 772. 10 Skvortsov, V.T., Immunochemistry 1972. 9 : 366. 1 1 Pestka, S., Rosenfeld, H. and Harris, R., Immunochemistry 1974. 11: 213.
Production of migration inhibitory factor and lymphotoxin by non-T cells" Treatment of tuberculin-sensitive guinea pig spleen or lymph node cells with a burro anti-T cell serum plus complement diminished markedly the number of functionally detectable T cells, but did not affect the amount of migration inhibitory factor (MIF) o r lymphotoxin produced by the residual T cell-depleted populations.
1. Introduction
A variety of in v i m tests have been utilized in recent years to assay cell-mediated immunity, and a general correlation between blast cell transformation, production of soluble mediators and lymphocyte cytotoxicity in v i m and the appropriate cell-mediated reactivity in vivo has been observed. Because cell-mediated immune reactivity has been shown to be a thymus-dependent immune response in vivo, it has widely been assumed that the in v i m reactions commonly studied reflect the activity of T lymphocytes. .While this is likely t o be true in many systems the question whether these functions can be carried out as well by B lymphocytes remains problematic. The present experiments were undertaken t o study t h e production of mediators from tuberculin-sensitive guinea pig lymph [I 9481
* Supported by A1 07118 and A1 10702. Correspondence: Barry R. Bloom, Department of Microbiology and Immunology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461, USA Abbreviations: MIF: Migration inhibitory factor LT: Lymphotoxin BAT: Burro anti-guinea pig thymus serum MEM: Eagle's minimum essential medium BSA: Bovine serum albumin NBS: Normal burro serum PPD: Protein derivative of tuberculin NGPS: Normal guinea pig serum PHA: Phytohemagglutinin C o d : Concanavalin A CRL: Complement receptor lymphocytes
node and spleen cell populations depleted of T lymphocytes by burro anti-guinea pig thymus (BAT) specific antisera prepared against strain 2 thymus and absorbed with the strain 2 B cell leukemia as described previously [ 11.
2. Materials and methods
Guinea pigs were sensitized t o tuberculin and lymph node and spleen cells were obtained as described previously [2]. After removing erythrocytes by centrifuging in a one-step discontinuous centrifugation in 29 % Pathocyte 4 bovine serum albumin (BSA) (Pentex-Miles Labs, Elkhart, Ind.) under a layer of Eagle's minimum essential medium (MEM), the white cells were treated either with normal burro serum (NBS) or burro anti-thymus serum (BAT) at a final cell concentration of 10 x 106/ml with a final concentration of antiserum at 1 /60 and guinea pig complement 1 /4 for 30 min at 37 O C . Cell killing was ascertained by trypan blue dye exclusion, and viable cells were recovered following a second discontinuous gradient centrifugation at the 2 9 7% BSA - MEM interface. After three washings, the viable cells were cultured in MEM at a concentration of 15 x 1 O6 cells/ml in Leighton tubes in the presence o r absence of PPD ( 2 5 pg/ml) as described previously. The supernatants were harvested after 20 h, PPD was restored to the control unstimulated supernatant ( t o 25 pg/ml), 15 % normal guinea pig serum (NGPS) wa i added and the supernatants were tested directly or frozen at -70 "C. Migration inhibitory factor (MIF) activity was tested
Eur. J. Immunol. 1975.5: 218-220
Production of MIF and LT by non-T cells
on normal guinea pig peritoneal cells 7 2 h after oil induction. For assaying lymphotoxin (LT) activity, a slight modification of the Granger assay [3] was used. A monolayer of 1.25 x 1 O5 L cells in the test supernatants or dilutions thereof, in 16 mm diameter wells of Linbro plates (cat. no. IS-FB-96-TC/55S) was cultured for 48 h, and the uptake of I4C amino acids (0.4 pCi of algal hydrolysate, Schwarz-Mann, Orangeburg, N.Y., cat. no. 31 22-10) over a 4 h period into TCA precipitable protein was measured.
in round bottom microtiter plates in RPMI 1640 supplemented with 10 % NGPS. The optimal concentration of phytohemagglutinin (PHA, Welcome, Kent) or concanavalin A (Nutritional Biochemicals, Cleveland, Ohio) was added in 10 1.11. Cells were cultured for 72 h at 3 7 "C and 18 h prior t o harvesting, 1.0 pCi of [3H]thymidine (New England Nuclear, Boston, Mass.) was added t o each culture. The amount of [ 3H]thymidine incorporation into DNA was measured with the MASH I1 semiautomated harvester (Microbiological Associates, Bethesda, Md.). As can be seen from Table 2, cells treated with NBS + C responded vigorously both t o PHA and t o Con A. In contrast, cells treated with BAT + C failed to respond t o PHA and had only a feeble response t o Con A. These results indicated that under the conditions used, most of the T cells in guinea pig spleen were killed by the antiBAT serum and in this regard it may be noted that 9 7 %99 % of the cells remaining alive after treatment of guinea pig lymph node cells after treatment with BAT + C bore surface immunoglobulins [I]. The approximately 30 % of spleen cells surviving after this treatment which d o not bear surface immunoglobulin are unlikely t o be mature lymphoid cells.
3. Results T cell-depleted spleens or lymph nodes were as efficient, o n an equal cell basis, at producing MIF and LT as the intact spleen cells (Table 1). In the initial studies with lymph node cells, a high degree of cytotoxicity by the BAT was found (70-85 %). The possibility that small numbers of residual T cells would be heavily weighted upon reconstituting equal numbers of viable cells motivated us to carry out similar experiments on spleen cells, where approximately 34-44 % of the cells were killed by the BAT plus complement (C). Following treatment with anti-T serum, the number of spleen cells staining for immunoglobulin with a fluorescent goat anti-Ig rose from approximately 3 4 % t o 6 7 %. To control for the possibility that the production of the mediators by the non-T cell population could result from a mitogenic activity of PPD o n B cells, parallel experiments were carried out with tuberculin negative guinea pigs and no MIF production was detected. Lastly, supernatants were subjected t o gel filtration o n Sephadex (3-75 and the broad second peak about Kd 0.1 5 (encompassing mol.wt. markers between 67 000 and 35 000) was tested to demonstrate that MIF activity actually produced by T-depleted spleens was not simply a complex of antibody and tuberculin. The MIF produced by non-T cells eluted in this peak, ruling out the possibility of antigen-antibody complexes. These results strongly indicate that MIF and LT can be produced by non-T lymphocytes.
Table 2. Response of BAT serum-treated cells to mitogens Incorporation of [3H]thyrnidine (cpm)a) NBS + C treated BAT + C treated
Mitogen
1892 35 785 24 276
None PHA
Con A
4. Discussion The first observation suggesting that mediators could be produced b y cells other than T lymphocytes is that of Papageorgiou, Henley and Glade [5] which demonstrated a migration inhibitory activity in the culture supernatants of long-term human lymphoblastoid lines, all of which tested being, in fact, derived from B lymphocytes. Similarly, Oates et al. [6] reported that production of a mitogenic factor in chickens was diminished by bursectomy, and suggested that it might be produced both by B and T cells. More recently Yoshida et al. [7] reported the production of MIF t o PPD in the guinea pig b y complement receptor lymphocytes (CRL) thought t o contain primarily B lymphocytes, as well as by T cells. Curiously, MIF was produced equally well by CRL
Table 1 . Mediator production by non-T-cells
Cclls
PPD-scnsitivc Normal
treatment
NBS BAT NBS BAT
Inhibition of migration (%J Lvmuh . - node supers') Spleen Supers
52.39 43,36
41,48,63.56 36,42.61,56
N.T.
0.12
N.T.
0.0
a) Supers = supernatants. b) Peak 2 of Sephadex G-75 chromatography.
pk2
927 1053 3 864
a) Results represent the mean of three determinants.
It could b e argued that all of the MIF activity observed above might have been produced by residual T cells which remained after treatment with BAT serum. Therefore, further attempts were made to characterize the cells remaining after BAT treatment. Preliminary experiments demonstrated that the use of the rabbit erythrocyte rosette marker for guinea pig T cells [4] was unsuccessful because treatment of guinea pig spleen cells with BAT even in ihe absence o f C abolished their ability to form rabbit rosettes. Therefore, the functional ability of cells remaining after BAT and C treatment to respond t o the T cell mitogens, PHA o r Con A, was assessed. Spleen cells either treated with BAT o r NBS as described above were cultured at a density of 1 x 106/ml
Scrum
219
43 51
Lymphotoxin (5% killJ Spleen
Supers
(undiluted)
(1/3)
76 66
70 83
0 8
220
Eur. J. Immunol. 1975.5: 220-223
D. Naor, R. Saltoun and F. Falkenberg
from nonsensitized guinea pigs as from specifically immunized animals, in contrast to results obtained here and in our previous work. Epstein et al. [8], using the fluorescenceactivated cell sorter, have found interferon t o be produced by enriched B lymphocyte populations, although later than in the case of T cells. MacDermott et al. [9] have separated human lymphocytes into T and B cell fractions by filtration through Sephadex-conjugated anti-F columns, and found MIF production t o be produced by both B and T-rich fractions. Accepting the present fact that no single method of separation of lymphoid cells into B and T cells is perfect, t h e multiplicity of approaches used t o test the point renders the conclusion compelling that some mediators can be produced b y non-T as well as T lymphocytes. However, the results presented here do not exclude the possibility that a small percentage of T cells is necessary for activation of MIF production by B cells. Indeed, it appears that as few as 5 % T cells can induce human B lymphocytes t o respond t o PHA [ 101. Clearly, considerable caution must be observed in interpreting results using these in vitro assays.
D. Naor’,
Rita Saltoun’ and F. Falkenbergoo
Lautenberg Center for General and Tumor Immunology, The Hebrew UniversityHadassah Medical School, Jerusalem’, and Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot’
5. References 1 Shevach, E., Green, I., Ellman, L. and Maillard, J.,Nuture-New Biol 1972.235: 19.
2 Bloom, B.R. and Bennett, B., in Bloom, B.R. and Glade, P. (Eds.) In Vitro Methods in Cell Mediated Immunity, Academic Press, New York 1971, pp. 235-248. 3 Granger, G. and Williams, T.W., Nature 1968. 218: 1253. 4 Stadecker, M.J., Bishop, G. and Wortis, H.H., J. Immunol. 1973. 111: 1834. 5 Papageorgiou, P.S.. Henley, W.L. and Glade, P.R., J. Immunol. 1972.108: 494. 6 Oates, C.M., Bissenden, J.F., Maini, R.N., Payne, L.N. and Dumonde, D.C., Nature New-Biol. 1972. 239: 137. 7 Yoshida, T., Sonozaki, H. and Cohen, S . , J. Exp. Med. 1973.138: 784. 8 Epstein, L.B., Kresth, H.W. and Herzenberg, L.A., Cell. Immunol. 1974. 12: 407. 9 MacDermott, R.P., Rocklin, R.E., Chess, L., David, J.R. and Schlossman, S.T., J. Clin. Invest. 1974.53: 49a. 10 Lohrmann, H.P., Novikovs, L. and Graw, R.G., J. Exp. Med. 1974. 139: 1 5 5 3 .
Lack of requirement for thymocytes for efficient antibody formation to ttinitrophenylated mouse red cells in mice: role for thymocytes in suppression of the immune response* The requirement of thymus-derived cells for the jmmune response to trinitrophenylated syngeneic mouse red cells (TNP-MRC) was investigated. In three sets of experiments the following results were obtained: a) irradiated mice which were reconstituted with bone marrow cells alone showed a better anti-TNP response after injection with TNP-MRC than those reconstituted with both bone marrow cells and thymocytes. b) Anti-thymocyte serum augmented the mouse anti-TNP response to TNP-MRC. c) Nude thymusless mice showed a better anti-TNP response t o TNP-MRC than their normal littermate controls. These results indicate that the anti-TNP response of mice t o TNP-MRC does not require thymus-derived helper cells. Moreover, thymus-derived cells have a suppressive effect o n the anti-TNP response.
1. Introduction Most of the antigens which are capable of stimulating an antibody response b y triggering bone marrow cells directly are distinguished, at least partly, by some unique features: they induce, merely, IgM responses [ 1-41, they possess repeating antigenic determinants [ 2 , 3 , 5-81, are slowly metabolized [2, 3 , 7 , 91 and exhibit mitogenic properties [lo]. The im[I 9031
Correspondence: David Naor, Lautenberg Center for General and Tumor Immunology, The Hebrew University- Hadassah Medical School, Jerusalem, Israel Abbreviations: TNP-MRC: Trinitrophenylated mouse red cells TNPSRC: Trinitrophenylated sheep red cells TNP-DRC: Trinitrophenylated donkey red cells PFC: Plaque-forming cells TNBS: Trinitrobenzene sulfonic acid ATS: Anti-thymocyte serum NRS: Normal rabbit serum
* Supported by a grant from Mr. Frank Lautenberg. Recipient of a postdoctoral fellowship of the Stiftung Volkswagenwerk, on leave from the lnstitut fb Biochemie, Ruhruniversitat Bochum, Germany.
mune response t o some of these immunogens is depressed by thymus-derived cells [ 1 1 - 131. Thus, such immunogens are not only thymus independent, but can be negatively thymus dependent as well. It is widely accepted that the immun: response t o “conventional” thymus-dependent immunogens is also regulated t o some degree by a suppressive subpopulation of thymus-derived cells [ 151. It was shown in a previous communication [ 161 that TNPMRC induce low direct anti-TNP PFC responses 4 days after their injection into the syngeneic host. No indirect response was found 4 o r 7 days after immunization. Since TNP-MRC cannot provide sufficient carrier determinants for thymusderived carrier-seeking cells, a thymus-independent response t o this immunogen could be expected. The limited IgM response to TNP-MRC supported this assumption, This munication will present evidence that t h e anti-TNP response t o TNP-MRC is not only thymus independent but also suppressed by thymocytes of t h e syngeneic host. We use the term “thymus independence” for an immune response which does not require thymocytes when tested by transfer experiments in irradiated mice.
B.R. Bloom, G. Stoner, J. Gaffney, E. Shevach and I. Green
tion method the sensitivity increased only 10-30 times; this advantage, however, was achieved by very simple means requiring almost n o change of the laboratory protocol. Polymers like dextran and PEG have been known t o enhance the precipitation of antibodies with molecular antigens [ 12, 131. We have shown that the presence of PEG affects both the interaction of DNP-bacteriophage with anti-DNP antibodies (present in concentrations 1 - 10 ng/ml), and the interaction of complexes with anti-Ig antibodies. The effect exerted o n the latter interaction seems t o present the major contribution to the final result.
Eur. J. Immunol. 1975.5: 218-220
5. References 1 Makela, O.,Immunology 1966. 10: 81. 2 Haimovich, J., Hurwitz, E., Novik, N. and Sela, M., Biochim. Biophys. Acta 1970. 207: 125.
3 Andrieu, J.M., Mamas, S . and Dray, F., Eur. J. Immunol. 1974. 4 : 417. 4 Eisen, N.H., Belman, S . and Carsten, M.E., J. Amer. Chem. SOC. 1953. 75: 4583. 5 Franzk, F. and Simek, L., Eur. J. Immunol. 1971. 1 : 300. 6 Segal, S., Globerson, A., Feldman, M., Haimovich, J. and Sela, M., J. Exp. Med. 1970.131: 93.
7 Jormalainen, S. and Makela, O., Eur. J. Immunol. 1971. 1: 471.
It is likely that the effect of PEG is of general nature, because it has been observed both with the system porcine anti-DNP plus ovine anti-pig Ig antibodies and rabbit antiDNP plus porcine anti-rabbit Ig antibodies. The fact that lower antigen concentrations inhibit DNP-bacteriophage inactivation in the presence of PEG is conceivable, since a lower antibody concentration is employed when the antigen assay is performed in the presence of PEG.
12 Hellsing, K. and Laurent, T.C., Acta Chem. Scund. 1964. 18: 1303.
Received October 7, 1974.
1 3 Harrington, J.C., Fenton, J.W.11 and Pert, J.H., Immunochemistry 1971.8: 431.
B.R. Bloom', G. Stoner'. Judith Gaffney+, E. Shevach" and I. Greeno Albert Einstein College of Medicine', of Allergy Bronx and and Infectious Diseaseso,National Institutes of Health, Bethesda
8 Haimovich, J., Hurwitz, E., Novik, N. and Sela, M., Biochim. Biophys. Acta 1970. 207: 115. 9 Krummel, W.M. and Uhr, J.W., J. Immunol. 1969.102: 772. 10 Skvortsov, V.T., Immunochemistry 1972. 9 : 366. 1 1 Pestka, S., Rosenfeld, H. and Harris, R., Immunochemistry 1974. 11: 213.
Production of migration inhibitory factor and lymphotoxin by non-T cells" Treatment of tuberculin-sensitive guinea pig spleen or lymph node cells with a burro anti-T cell serum plus complement diminished markedly the number of functionally detectable T cells, but did not affect the amount of migration inhibitory factor (MIF) o r lymphotoxin produced by the residual T cell-depleted populations.
1. Introduction
A variety of in v i m tests have been utilized in recent years to assay cell-mediated immunity, and a general correlation between blast cell transformation, production of soluble mediators and lymphocyte cytotoxicity in v i m and the appropriate cell-mediated reactivity in vivo has been observed. Because cell-mediated immune reactivity has been shown to be a thymus-dependent immune response in vivo, it has widely been assumed that the in v i m reactions commonly studied reflect the activity of T lymphocytes. .While this is likely t o be true in many systems the question whether these functions can be carried out as well by B lymphocytes remains problematic. The present experiments were undertaken t o study t h e production of mediators from tuberculin-sensitive guinea pig lymph [I 9481
* Supported by A1 07118 and A1 10702. Correspondence: Barry R. Bloom, Department of Microbiology and Immunology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461, USA Abbreviations: MIF: Migration inhibitory factor LT: Lymphotoxin BAT: Burro anti-guinea pig thymus serum MEM: Eagle's minimum essential medium BSA: Bovine serum albumin NBS: Normal burro serum PPD: Protein derivative of tuberculin NGPS: Normal guinea pig serum PHA: Phytohemagglutinin C o d : Concanavalin A CRL: Complement receptor lymphocytes
node and spleen cell populations depleted of T lymphocytes by burro anti-guinea pig thymus (BAT) specific antisera prepared against strain 2 thymus and absorbed with the strain 2 B cell leukemia as described previously [ 11.
2. Materials and methods
Guinea pigs were sensitized t o tuberculin and lymph node and spleen cells were obtained as described previously [2]. After removing erythrocytes by centrifuging in a one-step discontinuous centrifugation in 29 % Pathocyte 4 bovine serum albumin (BSA) (Pentex-Miles Labs, Elkhart, Ind.) under a layer of Eagle's minimum essential medium (MEM), the white cells were treated either with normal burro serum (NBS) or burro anti-thymus serum (BAT) at a final cell concentration of 10 x 106/ml with a final concentration of antiserum at 1 /60 and guinea pig complement 1 /4 for 30 min at 37 O C . Cell killing was ascertained by trypan blue dye exclusion, and viable cells were recovered following a second discontinuous gradient centrifugation at the 2 9 7% BSA - MEM interface. After three washings, the viable cells were cultured in MEM at a concentration of 15 x 1 O6 cells/ml in Leighton tubes in the presence o r absence of PPD ( 2 5 pg/ml) as described previously. The supernatants were harvested after 20 h, PPD was restored to the control unstimulated supernatant ( t o 25 pg/ml), 15 % normal guinea pig serum (NGPS) wa i added and the supernatants were tested directly or frozen at -70 "C. Migration inhibitory factor (MIF) activity was tested
Eur. J. Immunol. 1975.5: 218-220
Production of MIF and LT by non-T cells
on normal guinea pig peritoneal cells 7 2 h after oil induction. For assaying lymphotoxin (LT) activity, a slight modification of the Granger assay [3] was used. A monolayer of 1.25 x 1 O5 L cells in the test supernatants or dilutions thereof, in 16 mm diameter wells of Linbro plates (cat. no. IS-FB-96-TC/55S) was cultured for 48 h, and the uptake of I4C amino acids (0.4 pCi of algal hydrolysate, Schwarz-Mann, Orangeburg, N.Y., cat. no. 31 22-10) over a 4 h period into TCA precipitable protein was measured.
in round bottom microtiter plates in RPMI 1640 supplemented with 10 % NGPS. The optimal concentration of phytohemagglutinin (PHA, Welcome, Kent) or concanavalin A (Nutritional Biochemicals, Cleveland, Ohio) was added in 10 1.11. Cells were cultured for 72 h at 3 7 "C and 18 h prior t o harvesting, 1.0 pCi of [3H]thymidine (New England Nuclear, Boston, Mass.) was added t o each culture. The amount of [ 3H]thymidine incorporation into DNA was measured with the MASH I1 semiautomated harvester (Microbiological Associates, Bethesda, Md.). As can be seen from Table 2, cells treated with NBS + C responded vigorously both t o PHA and t o Con A. In contrast, cells treated with BAT + C failed to respond t o PHA and had only a feeble response t o Con A. These results indicated that under the conditions used, most of the T cells in guinea pig spleen were killed by the antiBAT serum and in this regard it may be noted that 9 7 %99 % of the cells remaining alive after treatment of guinea pig lymph node cells after treatment with BAT + C bore surface immunoglobulins [I]. The approximately 30 % of spleen cells surviving after this treatment which d o not bear surface immunoglobulin are unlikely t o be mature lymphoid cells.
3. Results T cell-depleted spleens or lymph nodes were as efficient, o n an equal cell basis, at producing MIF and LT as the intact spleen cells (Table 1). In the initial studies with lymph node cells, a high degree of cytotoxicity by the BAT was found (70-85 %). The possibility that small numbers of residual T cells would be heavily weighted upon reconstituting equal numbers of viable cells motivated us to carry out similar experiments on spleen cells, where approximately 34-44 % of the cells were killed by the BAT plus complement (C). Following treatment with anti-T serum, the number of spleen cells staining for immunoglobulin with a fluorescent goat anti-Ig rose from approximately 3 4 % t o 6 7 %. To control for the possibility that the production of the mediators by the non-T cell population could result from a mitogenic activity of PPD o n B cells, parallel experiments were carried out with tuberculin negative guinea pigs and no MIF production was detected. Lastly, supernatants were subjected t o gel filtration o n Sephadex (3-75 and the broad second peak about Kd 0.1 5 (encompassing mol.wt. markers between 67 000 and 35 000) was tested to demonstrate that MIF activity actually produced by T-depleted spleens was not simply a complex of antibody and tuberculin. The MIF produced by non-T cells eluted in this peak, ruling out the possibility of antigen-antibody complexes. These results strongly indicate that MIF and LT can be produced by non-T lymphocytes.
Table 2. Response of BAT serum-treated cells to mitogens Incorporation of [3H]thyrnidine (cpm)a) NBS + C treated BAT + C treated
Mitogen
1892 35 785 24 276
None PHA
Con A
4. Discussion The first observation suggesting that mediators could be produced b y cells other than T lymphocytes is that of Papageorgiou, Henley and Glade [5] which demonstrated a migration inhibitory activity in the culture supernatants of long-term human lymphoblastoid lines, all of which tested being, in fact, derived from B lymphocytes. Similarly, Oates et al. [6] reported that production of a mitogenic factor in chickens was diminished by bursectomy, and suggested that it might be produced both by B and T cells. More recently Yoshida et al. [7] reported the production of MIF t o PPD in the guinea pig b y complement receptor lymphocytes (CRL) thought t o contain primarily B lymphocytes, as well as by T cells. Curiously, MIF was produced equally well by CRL
Table 1 . Mediator production by non-T-cells
Cclls
PPD-scnsitivc Normal
treatment
NBS BAT NBS BAT
Inhibition of migration (%J Lvmuh . - node supers') Spleen Supers
52.39 43,36
41,48,63.56 36,42.61,56
N.T.
0.12
N.T.
0.0
a) Supers = supernatants. b) Peak 2 of Sephadex G-75 chromatography.
pk2
927 1053 3 864
a) Results represent the mean of three determinants.
It could b e argued that all of the MIF activity observed above might have been produced by residual T cells which remained after treatment with BAT serum. Therefore, further attempts were made to characterize the cells remaining after BAT treatment. Preliminary experiments demonstrated that the use of the rabbit erythrocyte rosette marker for guinea pig T cells [4] was unsuccessful because treatment of guinea pig spleen cells with BAT even in ihe absence o f C abolished their ability to form rabbit rosettes. Therefore, the functional ability of cells remaining after BAT and C treatment to respond t o the T cell mitogens, PHA o r Con A, was assessed. Spleen cells either treated with BAT o r NBS as described above were cultured at a density of 1 x 106/ml
Scrum
219
43 51
Lymphotoxin (5% killJ Spleen
Supers
(undiluted)
(1/3)
76 66
70 83
0 8
220
Eur. J. Immunol. 1975.5: 220-223
D. Naor, R. Saltoun and F. Falkenberg
from nonsensitized guinea pigs as from specifically immunized animals, in contrast to results obtained here and in our previous work. Epstein et al. [8], using the fluorescenceactivated cell sorter, have found interferon t o be produced by enriched B lymphocyte populations, although later than in the case of T cells. MacDermott et al. [9] have separated human lymphocytes into T and B cell fractions by filtration through Sephadex-conjugated anti-F columns, and found MIF production t o be produced by both B and T-rich fractions. Accepting the present fact that no single method of separation of lymphoid cells into B and T cells is perfect, t h e multiplicity of approaches used t o test the point renders the conclusion compelling that some mediators can be produced b y non-T as well as T lymphocytes. However, the results presented here do not exclude the possibility that a small percentage of T cells is necessary for activation of MIF production by B cells. Indeed, it appears that as few as 5 % T cells can induce human B lymphocytes t o respond t o PHA [ 101. Clearly, considerable caution must be observed in interpreting results using these in vitro assays.
D. Naor’,
Rita Saltoun’ and F. Falkenbergoo
Lautenberg Center for General and Tumor Immunology, The Hebrew UniversityHadassah Medical School, Jerusalem’, and Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot’
5. References 1 Shevach, E., Green, I., Ellman, L. and Maillard, J.,Nuture-New Biol 1972.235: 19.
2 Bloom, B.R. and Bennett, B., in Bloom, B.R. and Glade, P. (Eds.) In Vitro Methods in Cell Mediated Immunity, Academic Press, New York 1971, pp. 235-248. 3 Granger, G. and Williams, T.W., Nature 1968. 218: 1253. 4 Stadecker, M.J., Bishop, G. and Wortis, H.H., J. Immunol. 1973. 111: 1834. 5 Papageorgiou, P.S.. Henley, W.L. and Glade, P.R., J. Immunol. 1972.108: 494. 6 Oates, C.M., Bissenden, J.F., Maini, R.N., Payne, L.N. and Dumonde, D.C., Nature New-Biol. 1972. 239: 137. 7 Yoshida, T., Sonozaki, H. and Cohen, S . , J. Exp. Med. 1973.138: 784. 8 Epstein, L.B., Kresth, H.W. and Herzenberg, L.A., Cell. Immunol. 1974. 12: 407. 9 MacDermott, R.P., Rocklin, R.E., Chess, L., David, J.R. and Schlossman, S.T., J. Clin. Invest. 1974.53: 49a. 10 Lohrmann, H.P., Novikovs, L. and Graw, R.G., J. Exp. Med. 1974. 139: 1 5 5 3 .
Lack of requirement for thymocytes for efficient antibody formation to ttinitrophenylated mouse red cells in mice: role for thymocytes in suppression of the immune response* The requirement of thymus-derived cells for the jmmune response to trinitrophenylated syngeneic mouse red cells (TNP-MRC) was investigated. In three sets of experiments the following results were obtained: a) irradiated mice which were reconstituted with bone marrow cells alone showed a better anti-TNP response after injection with TNP-MRC than those reconstituted with both bone marrow cells and thymocytes. b) Anti-thymocyte serum augmented the mouse anti-TNP response to TNP-MRC. c) Nude thymusless mice showed a better anti-TNP response t o TNP-MRC than their normal littermate controls. These results indicate that the anti-TNP response of mice t o TNP-MRC does not require thymus-derived helper cells. Moreover, thymus-derived cells have a suppressive effect o n the anti-TNP response.
1. Introduction Most of the antigens which are capable of stimulating an antibody response b y triggering bone marrow cells directly are distinguished, at least partly, by some unique features: they induce, merely, IgM responses [ 1-41, they possess repeating antigenic determinants [ 2 , 3 , 5-81, are slowly metabolized [2, 3 , 7 , 91 and exhibit mitogenic properties [lo]. The im[I 9031
Correspondence: David Naor, Lautenberg Center for General and Tumor Immunology, The Hebrew University- Hadassah Medical School, Jerusalem, Israel Abbreviations: TNP-MRC: Trinitrophenylated mouse red cells TNPSRC: Trinitrophenylated sheep red cells TNP-DRC: Trinitrophenylated donkey red cells PFC: Plaque-forming cells TNBS: Trinitrobenzene sulfonic acid ATS: Anti-thymocyte serum NRS: Normal rabbit serum
* Supported by a grant from Mr. Frank Lautenberg. Recipient of a postdoctoral fellowship of the Stiftung Volkswagenwerk, on leave from the lnstitut fb Biochemie, Ruhruniversitat Bochum, Germany.
mune response t o some of these immunogens is depressed by thymus-derived cells [ 1 1 - 131. Thus, such immunogens are not only thymus independent, but can be negatively thymus dependent as well. It is widely accepted that the immun: response t o “conventional” thymus-dependent immunogens is also regulated t o some degree by a suppressive subpopulation of thymus-derived cells [ 151. It was shown in a previous communication [ 161 that TNPMRC induce low direct anti-TNP PFC responses 4 days after their injection into the syngeneic host. No indirect response was found 4 o r 7 days after immunization. Since TNP-MRC cannot provide sufficient carrier determinants for thymusderived carrier-seeking cells, a thymus-independent response t o this immunogen could be expected. The limited IgM response to TNP-MRC supported this assumption, This munication will present evidence that t h e anti-TNP response t o TNP-MRC is not only thymus independent but also suppressed by thymocytes of t h e syngeneic host. We use the term “thymus independence” for an immune response which does not require thymocytes when tested by transfer experiments in irradiated mice.
