Scond. J. Immunol. 10, 267-274, 1979
HLA-D-restricted Antigen Activation of Sensitized T Lymphocytes: Studies on the Abihty of HLA-D/DR-expressing B Lymphocytes to Substitute for Macrophages in Antigen Activation B. O. BERGHOLTZ & E. THORSBY Tissue Typing Laboratory, Rikshospilalel, The National Hospital. Oslo, Norway
Bergholtz, B.O. & Thorsby, E. HLA-D-reslricted Antigen Aciivaiion of Sensitized T Lymphoeytes: Studies on ihc Ability of HLA-D/DR-expressing B Lymphocytes to Substitute for Macrophages in Antigen Aciivaiion. Scand. J. Immunol. 10, 267-274, 1979. The proliterative response of sensiii/.ed human T Kmphocytes to puritied protein derivative (PPD) and lo trinitrophenyl (TNP)-conjuBatcJ autologous cells in vilro is restricted by self HLA-D DR determinanls. Here we report ihal ihc PPD-specilic response Is siricil) related lo the content of phagoc\iosing cells Imacrophages. M9I in the cullures and that an opiimal PPD response occurred al a T/M-p ratio between 10:1 and 5;l. B-ccll-enric!ied suspensions, which also express the HLA-D/DR deierminanis. were nol able lo replace the macrophages in ihis HI.A-D;DR-restricted response. On tbe other hand. TNP-ireated similarly prepared B cells were in most instances effective in inducing a secondary TNP-specilic response of in vitro-scnsiiized T cells. B. O. Bergholt:. Tissue Typing Lahorator\, Rik.\hospitatef. The National Hospital. Oslo I. Sorwax.
Most T cell immune responses seem to be macrophage-dependent [ 11 ]. We have previotisly reported that this is also the case for the activation of T cells from bacillus C'almetteCiuerin (BCG)-imnumi/ed humans hy ptirificd protein derivative (PPD) in vitro. A proliferative response is only obtained when the antigen is presented by viable autologous macrophages (M9) or allogeneic M9 which share at least one HLA-D gene prodtict 1 HLA-D/DR determinani) with the T cell dotior [T'-S]. Specific alloaniibodies recognizing HLA-DR antigens were able to block this M^p/T cell cooperation provided they were reactive with an HLA-DR antigen shared by the M9 and the T cell donors [4]. Tbus, the prolirerative T cell response to PPD is HLA-D-restrictcd, i.e. it requires the combined recognition of the specific antigen and "self" HLA-D gene products expressed in the macrophage cell membrane. Similar obser0300-9475/79/0900-0267 $02.00
vations were made by restimuiating T cells primed to aulologous trinitrophenyl (TNP)conjugated non-T cells in vitro. The haptenspecific secondary proliferative response of these sensitized T eells is also, at least in part, restricted by self HLA-D 115]. The macrophage fractions used in our previous experiments consisted mainly of phagoeytosing cells of monocyte origin. However. B cells also express the HLA-D/DR antigens [14]. The question then arises whether B cells in conjunction with antigen might also be able to elicit an HLA-D/DR restricted response of immune T cells, or if this ability is an exclusive property of the macrophages. ln the present paper we present results indicating that B ctlls are not able to substitute for macrophages in the activation of sensitized T cells by PPD. On the olher hand, similarly prepared TNP-conjugated niacrophage-depleted autologous B cells were
r 1979 Blackwell Scientific Publications
267
268
B. O. Berffhottx & E. Thor.shy
able in most instances to induce a secondary response of T cells primed to TNP-conjugated autologous non-T cells.
MATERIALS AND METHODS Celt donnrs. Cell donors were healthy voltintecrs whose Cull HL'\ phenutypes (including HL.A-DR) were established by serologital typing |2| and by tionitwygous cell typing 18|. All donors used in the c\pcriments with PPD were tuherculin-positive as judged by skin testing. PPD. Purilied pniicin derivaiive of tuberculin was purchased from the Veterinary Institute, University of Oslo(PPD-l. 2 mg/ml). TNP treattnent. Cells were TNP-treaied following the protocol of Shearer, as previously described [15]. Briefly, cells were mixed with O.OI M TNP (picr>i sulphonic acid. Sigma cat. no. P 5878) in phosphatebutTered saline. pH 7.4. at .17 C for 10 min, Tollowed by two washes in medium. Macrophages. A macrophage-enrichcd population was prepared from peripheral blood mononuclear cells (PBM). separated Irom delibrjnated blood by means of Lymphoprep (Nyco. Oslo. Norway). The PUM were incubated (.17 C. 5",, CO,) for 2 h in tissue culture (lasks with medium RPMI 1640 and 2tr^ pooled normal human serum. The non-adherent cells were then removed by washing the flasks twice wiih medium and were subsequently used for the preparalioii of T and B cells (below), Theadhcrent ceils were Ihcn washed again. incubated for 90 min on ice with medium RPMI 1640 containing 30"^ normal human serum and EiDTA (3..^ ;^g ml), and could then be detached by vigorous pipeiting. 85-95"o of the cells thus prepared were able to ingest latex particles (Bacto-Latex 0.81, Difco. USA) and showed monocyte morphology [7]. T lymphocytes. A T cell-enriched population was prepared by allowing the non-adherenl cclK obtained after culturing PBM for maerophage preparation to form rosettes wiih 2-an>inoeihyl-isoihiuroniuni bromide {AET)-treated sheep erythrocyies (SRBC) [I2|. The rosettes were separated from the non-roseiling cells by cent rifu gat ion on a Lymphoprep gradient. The non-rosctting cells were used for ihe preparation ol' B cells (below). The roselies were then suspended in 1 ml of fresh human serum and incubated for 15 min at 37 C. resulting in lysis of the SRBC by the naturally occurring anti-sheep antibodies in human serum. After incubation of the cells in tissue culture flasks overnight to remove more of Ihe adherent cells. the rosetting/separaiion/hacniolysis setiiience was repealed once, resulting in a T cell fraction containing more than 85" „ H rosciie-forming cells, less than 5°,, Ig'positive cells, and less ihan 0.5''» latex-ingesting cells. B lymphocytes. A B cell-enriched population was purified from the non-rosettlng cells of ihe first rosetting procedure above. The cells were cultured in tissue culture llasks overnight (37 C. 5",, CO^) followed by another rosetling and Lymphoprep separalion to remove any remaining T cells. The cells thus obtained contained more than 65" „ Ig positive cells, 6 II" „ lalex-
ingesting cells.and less than 2"„ E roselte-forming cells. This fraciion will be referred to as B, cells. In some experiments ihis fraction was purilied further in order 10 remove the rest of the adherent cells. 2 5 - 10" B, cells were suspended in 5 ml of medium RPMI 1640 with 5" „ pooled normal human serum and incubaied for 30 min at 37 C in a glass bead column ic. 20 ml of glass beads, 0.5 mm diameter). The non-adherenl B cells were eiuted by flushing the columns with 50 ml medium. This fraction (B.jl always contained less than !"„ latexingesting cells and more ihan 60" „ of the cells carried cell-membrane bound Ig and reacted with ami DR antisera. In the TNP experiments, B cell-enriched suspensions were produced from PBM by first removing adherent cells {2 h culture in tissue culture (lasks) and then rosetting by AET-treated SRBC. The interface cells were then incubated in ihe glass bead column, as described. These B-cell-enriched ceil suspensions also always contained less than I",, latex-ingesting cells and 60-9^)"„ Ig positive and DR positive cells. Non-T cells. A non-T-cell fraction (B cells • macrophages) was prepared as described above by rosetting PBM cells with AFT-lrealed SRBC, wiihoui previous ineubalion in tissue culture flasks. The inierface cells after Lymphoprep notation contained 50 60"o Igposilive cells and 20-30",, latex-ingesting cells. Cell ctdture techniques, (a) The FPD experiments were performed in flal-boltomcd microtitre plates (TC dish 3596. Costar. USA). The following protocol was followed unless otherwise stated: triplicates of 5 < \0* T lymphocytes were co-cullured with 0.2-50 .•; 10' Xirnidiaied (20(K) rad) accessory cells (M9. B, or B, fraction) in a toial volume of 170 \i\ of tissue culture medium (RPMI 1640 supplemented with antibiotics and 20" „ heal-inaelivated pooled normal human serum), with or without PPD (2.5 [ig/ml). at 37 C and 5'„ COj. 'H-thymidine was added on day 4 and the cultures were harvested on day 5 by means of a semi-aulomaiic multiple culture harvester (Skatron. Lierbyen. Norway). *H incorporalion was assessed by means of a liquid scintillation counter. The results were expressed as mean cpm : SE. By 'incremental PPD response' is meant ihe response in cpm with PPD minus the response of the same culture wiihoui PPD. (bl The TNP experiments were performed as previously described [15). Briefly. 3 5 • 10" T cells were mixed with an equal number of TNP-treated aulologous 2000-rad-irradiated non-T cells in medium in 10 ml lubes (Falcon cat. no. 3033). Afler 9-10 days. during which period half of the medium bad been replaced once (days 3 5). 0.51 ^ 10* cells were transferred to each well of round-bottomed Linbro microlitre plates. They were ihen mixed with an equal amount of 20CX)-rad-irradialed non-lreated or TNPirealed cells prepared as described above from u fresh bleeding of ihe donor The cells were then cultured for 48 72 h. including 24 h after addition of 0.2 ml of physiological saline containing lOuCi of *H-lhymidine. and harvested as described above. In some experiments we followed the protocol described by Charmol & Mawas [6]. priming T cells with non-T cells in tissue ciillure llasks (Falcon no. 3013) for 6 days, followed by restimiilaiion in microiiire plates for 48 h. as described above
HLA-D Restriction of T Cell Activation The results were expressed as median cpm of iriplicatcs fto save space). As TNP-speeific response (A) was used the median cpm in cultures restimubted witb TNP-treated cells, afier subtraction of the median cpm values obtained after restimulation with non-lreaied cells.
269
results consistent with the above. The results of one experiment are shown in Fig. I. Both Table I and Fig. I demonstrate that an optimal T cell response occurred at a T/M9 ratio between 10:1 and 5:1 dropping markedly at higher Mtp concentrations. The response to PPD in conjunction with B, cells, on the other hand, was strongest at a T/B ratio of 1:1. Higher B cell concentrations were not tested.
RESULTS PPD presentation by different tioii- T cett fractions
Relation between PPD-presenting activity attd conleiil of phagocytositif! celts
In two experimenis 5x 10* T cells from a total of four different donors were stimulated with PPD in the presence of various numbers of aulologous non-T accessory eells (M 9. B, or B^,). Both experiments yielded consistent results. Table I shows the results of one of these experiments involving cells from two different donors and comparing the PPD-presenting capacities of the M-p. Bl and B^. fractions. Both the M^and the B, fractions were able to present PPD to the T cells, whiie the Bj fraction was not. Three more experiments were performed testing various numbers of B, cells and M9 from a total of seven donors. These experiments yielded
The above results indicate that the ability of a B cell fraction to present PPD to T cells is determined by its content of phagocytosing ceiis. This is more clearly demonstrated in Fig. 2, where the results of two of the experiments are presented, relating the incremental PPD response to the absolute number of phagocyiosing (latex ingesting) cells among the accessory cells (B| or M9). It appears that the PPD-presenting ability of non-T cells is a function of the content of phagocytosing cells, irrespective of whether or not additional B cells are present.
I. Purified protein derivative (PPD) presentation by different autologotis non-T fractions
No. of irradiated aulologous accessory cells
.•\UH)I.
T^eeH E.S.
E.T.
access. cells
Pi'D ((jg/m!)
0
200
1000
5000
10.000
50.000
(91" J t
0 2.5
66 : 12* 206 : 36 140:
77 :..25 1065 • .16 988
307 4-30 2990 : 174 2683
355 • 12 6350 ' 274 5995
409* 18 4780 143 4371
395:1^50 1911-59 1516
B, (6%)
0 2.5
24-i 5 480 : 40 456
155^7 615 t 102 460
153 ^ 6 1402 1 146 1249
470- 127 3111 f^29l 2641
795 i 83 .1419 ±176 2624
Bl ( 60"„ Ig and HLA-DR positive cells. t > 95" n latex ingesting cells. § Median cpm of triplicates. "^ TNP-specific response.
clear TNP-specific response was always seen using 10.000 or more B cells for restimulation. Also, the same number of TNP-treated B cells usually caused a stronger restimulation than ihe same number of TNP-treated macrophages.
In other experiments, 100,000 primed T cells were restimulated with an equal amount of non-treated or TNP-treated non-T cells. B cells or TNP-treated B cells together with 20% non-treated macrophages. In nine of twelve
272
B. O. Berghollz & E. Thorsby
experiments performed, a clear TNP-specific response was obtained after restimulation with TNP-treated autologous non-T cells, and in six of these restimulation with TNP-treated B cells gave rise to a TNP-specific response in the range of 50-IOO/o of the response seen using TNP-trcatcd non-T cells. This response was not or only slightly increased after addition of non-treated autotogous macrophages (data not shown).
DISCUSSION In previous studies we have shown that the in vitro PPD activation of T lymphocytes from sensitized individuals is HLA-D/DR-restricted; i.e. it requires that the antigen be recognized in the context of "self HLA-D/DR determinants of viable non-T accessory cells [3-5]. The PPD-responsive T cells do not express the HLA-D/DR determinants in detectable concentrations [1]. In conformity with the current concepts of the role of macrophages (M9) in the immunological network [11], we used the plasticadberent non-T population as our source of accessory cells. In this report we present evidence that the ability to present antigen (PPD) to T cells is an exclusive property of this M(? population. Although certainly expressing the HLA-D/DR determinants, B cells devoid of phagocytosing cells have no sucb ability (B^ fraction. Table I). The ability of a non-T cell population to present PPD to T cells is mainly determined by its content of phagocytosing cells (Fig. 2). It is mainly PPD bound to the accessory cells that stimulates T cells [5]. The simplest explanation of the inability of B cells to substitute for macrophages in antigen presentation would therefore be that B ceils lack the capacity to unspecifically bind antigen to their stirface. A selective purification of B cells bearing specific Ig receptors for PPD might yield a B cell fraction capable o\' binding and presenting PPD in the context of self HLA-D to T ceils. We have not been able to perform this kind of experiment. The T cell response to PPD was shown to be inhibited by an excess of M9—that is, at a T/Mtp ratio lower than 5:1. The presence of surplus B cells, on the other hand, had only a
weak inhibitory effect, if any at all (Tables i and II, Figs. 1 and 2). Inhibition of T cell responses in the presence of an excess of M9 has been described in various systems and may be explained in a variety of ways, such as production of inhibitory metabolites or enzymes, a direct inhibitory effect on T cells, or simply exhaustion of the medium [9, 10, 16]. The poor T cell response to PFD in conjunction with HLA-D/DR-incompatible allogeneie Mtp was not improved by the addition of autologous (HLA-D/DR-bearing) B cells (Table II). The possibility that suppressor cells were generated as the result of MLC activation by allogeneic M9 has been ruled out in previous experiments [4]. Assuming that the activation of T cells by Mtp-associated PPD involves two separate, independently distributed T cell receptor types, one with specificity for PPD and one for "self HLA-D/DR, one might have expected a response to take place under these experimental conditions, since the T cells had access both to M9-associated PPD and to autologous HLA-D/DR determinants. However, the above results indicate that the activation of T cells by PPD requires that bolh PPD and autologous HLA-D/DR antigens be displayed on the surface of the same cell. In previous experiments we showed that the specific secondary response of T lymphocytes primed to TNP-treated autologous cells in vitro is. at least in part, also restricted by seif-HLAD [15]. This was shown by using allogeneic cells expressing different HLA-D/DR determinants for restimulation and by the observation that TNP-treated autologous T cells, not expressing HLA-D/DR, were unable to induce a secondary response. Here we investigated whether TNP-treated B lymphocytes alone, depleted of macrophages, were able to restimulate TNP-primed ceils. Our data strongly suggest that TNP-treated B cells alone in most instances are able to induce a secondary response of primed T cells. When the same number of TNP-treated B cells and macrophages were compared, the B cells induced a stronger secondary TNP-specific response than macrophages. Furthermore, our B cell suspensions contained less than 1";, latex-ingesting cells, and this possible contamination of macrophages (500-1000 cells) was not able to induce a secondary TNP-specific response. In other experiments we found that the B cell
HLA-D Restriction of T Cell Activation suspensions were able to induce a response of allogeneic PBM and T cell-enriched lymphocytes of a degree similar to that induced by PBM stimulating cells prepared from the same donors [unpublished]. Thomas et at. [13] have studied these questions, using T cells from guinea-pigs sensitized in vivo with picryl-chloride and restimulating them with TNP-treated la positive EN-L2C B cell-like leukaemia cells. In contrast to similarly treated macrophages, these TNP-treated leukaemia cells were unable to induce a TNPspecific response. The authors state, however, that it might be that different results would have been obtained using normal and not malignant B cells. We are also currently studying whether TNP-treated B cell lines might induce a secondary TNP-specific response. The detailed events of T cell recognition of antigen are still poorly understood. However, the present results yield some information as to the nature of the participating accessory cells. The expression of HLA-D/DR determinants is necessary but, at least for soluble antigens, is not by itself sufficient since B cells are unable to present PPD to autologous T cells. It may be concluded that the proliferative T cell response to PPD requires non-T accessory cells which are characterized by monocyte morphology, phagocytosing capacity and ability to adhere to plastic and glass surfaces, in addition to their ability to bind and display PPD in conjtinction with the HLA-D/DR molecules. In contrast, haptens such asTNP are able to induce an HLA-D/DR-restricted specific secondary response of in vitro primed T cells also when conjugated to B cells. The explanation may be that TNP binds to most cell membranes and that modification of the TNPcell membrane complex by phagocytosing cells are not necessary, at least for inducing a response of primed T cells. However, as demonstrated in this and a previous article [15], the TNP-conjugated cells must express the self HLA-D/DR determinants to be able to induce a TNP-specific secondary response.
ACKNOWLEDGMFNTS This work has been supported by grants from the Norwegian Council for Science and the Humanities, Mr Anders Jahre's Fund for the
273
Promotion of Science and Laurine Maarschalk's Ftind. We wish to thank A. B. Thoresen and H. Nottsiainen for excellent technical assistance and E. Garmann and K. Goyer for typing and correcting the manuscript. REFERENCES 1 Albrechtsen. D.. Bratlie. A., Nousiainen. H., Solheim. B.G., Winther. N. & Thorsby, E. Serological typing of HLA-D; predictive value in mixed lymphocyte cultures (MLC). Immunogenetics. 6, 91, 1978. 2 Albrechtsen. D.. Solheim, B.G. & Thorsby, E. The presence of la-like determinants on a subpopulaCion of human T lymphocytes. Immunogeneiics. 5, t49, 1977. .1 Bergholl/, B.O. & Thorsby, E. Macrophagedependent response of immune human T lymphocytes to PPD in vitro. Scand. J. Immunol. 6, 779, 1977. 4 Bergholtz, BO. & Thorsby, E. HLA-D reslriclion of the miicrophagc-depenilent response of immune human T lymphocyfes to PPD in viiro: inhibition by anii-HLA-DR anlisera. Scand. J. humunol. 8, 63. 1978. 5 Bergholtz. B.O. & Thorsby. E. Macrophage/T lymphocyte interaction in Ihe immune response to PPD in humans. Scand. J. Immunol. 9, 511. t979. 6 Charmol. D. & Mawas, C. The in vitro cellular response of human lymphocytes to trinilrophenylated autologous cells: HLA-D restriction of proliferation but apparent absence of HLA restriction of cytolysis. Eur. J. Immunol, in press. 7 Furth. R. van. Cohn, Z.A.. Hirsch. J,B.. Spector. W.G. & Langcvoort. H.L. The mononuclear phagocyte system: a new classification of macrophages. monocytes, and their precursor cells. Bull. WId HItk Org. 46, 845. 1972, 8 Kaakingn, A.. Helgesen. A., Nousiainen, H., Solheim. B.G. & Thorsby, E. The HLA-D determinanls. A population study of Norwegians. Immurwgenetics. 4, 205, 1977. 9 Krtiisbcek. A.M. Helper and suppressor effects of macrophages in lymphocyte activation: a quantitative phenomenon? In Proceedings of the 12th International Leuf
HLA-D-restricted Antigen Activation of Sensitized T Lymphocytes: Studies on the Abihty of HLA-D/DR-expressing B Lymphocytes to Substitute for Macrophages in Antigen Activation B. O. BERGHOLTZ & E. THORSBY Tissue Typing Laboratory, Rikshospilalel, The National Hospital. Oslo, Norway
Bergholtz, B.O. & Thorsby, E. HLA-D-reslricted Antigen Aciivaiion of Sensitized T Lymphoeytes: Studies on ihc Ability of HLA-D/DR-expressing B Lymphocytes to Substitute for Macrophages in Antigen Aciivaiion. Scand. J. Immunol. 10, 267-274, 1979. The proliterative response of sensiii/.ed human T Kmphocytes to puritied protein derivative (PPD) and lo trinitrophenyl (TNP)-conjuBatcJ autologous cells in vilro is restricted by self HLA-D DR determinanls. Here we report ihal ihc PPD-specilic response Is siricil) related lo the content of phagoc\iosing cells Imacrophages. M9I in the cullures and that an opiimal PPD response occurred al a T/M-p ratio between 10:1 and 5;l. B-ccll-enric!ied suspensions, which also express the HLA-D/DR deierminanis. were nol able lo replace the macrophages in ihis HI.A-D;DR-restricted response. On tbe other hand. TNP-ireated similarly prepared B cells were in most instances effective in inducing a secondary TNP-specilic response of in vitro-scnsiiized T cells. B. O. Bergholt:. Tissue Typing Lahorator\, Rik.\hospitatef. The National Hospital. Oslo I. Sorwax.
Most T cell immune responses seem to be macrophage-dependent [ 11 ]. We have previotisly reported that this is also the case for the activation of T cells from bacillus C'almetteCiuerin (BCG)-imnumi/ed humans hy ptirificd protein derivative (PPD) in vitro. A proliferative response is only obtained when the antigen is presented by viable autologous macrophages (M9) or allogeneic M9 which share at least one HLA-D gene prodtict 1 HLA-D/DR determinani) with the T cell dotior [T'-S]. Specific alloaniibodies recognizing HLA-DR antigens were able to block this M^p/T cell cooperation provided they were reactive with an HLA-DR antigen shared by the M9 and the T cell donors [4]. Tbus, the prolirerative T cell response to PPD is HLA-D-restrictcd, i.e. it requires the combined recognition of the specific antigen and "self" HLA-D gene products expressed in the macrophage cell membrane. Similar obser0300-9475/79/0900-0267 $02.00
vations were made by restimuiating T cells primed to aulologous trinitrophenyl (TNP)conjugated non-T cells in vitro. The haptenspecific secondary proliferative response of these sensitized T eells is also, at least in part, restricted by self HLA-D 115]. The macrophage fractions used in our previous experiments consisted mainly of phagoeytosing cells of monocyte origin. However. B cells also express the HLA-D/DR antigens [14]. The question then arises whether B cells in conjunction with antigen might also be able to elicit an HLA-D/DR restricted response of immune T cells, or if this ability is an exclusive property of the macrophages. ln the present paper we present results indicating that B ctlls are not able to substitute for macrophages in the activation of sensitized T cells by PPD. On the olher hand, similarly prepared TNP-conjugated niacrophage-depleted autologous B cells were
r 1979 Blackwell Scientific Publications
267
268
B. O. Berffhottx & E. Thor.shy
able in most instances to induce a secondary response of T cells primed to TNP-conjugated autologous non-T cells.
MATERIALS AND METHODS Celt donnrs. Cell donors were healthy voltintecrs whose Cull HL'\ phenutypes (including HL.A-DR) were established by serologital typing |2| and by tionitwygous cell typing 18|. All donors used in the c\pcriments with PPD were tuherculin-positive as judged by skin testing. PPD. Purilied pniicin derivaiive of tuberculin was purchased from the Veterinary Institute, University of Oslo(PPD-l. 2 mg/ml). TNP treattnent. Cells were TNP-treaied following the protocol of Shearer, as previously described [15]. Briefly, cells were mixed with O.OI M TNP (picr>i sulphonic acid. Sigma cat. no. P 5878) in phosphatebutTered saline. pH 7.4. at .17 C for 10 min, Tollowed by two washes in medium. Macrophages. A macrophage-enrichcd population was prepared from peripheral blood mononuclear cells (PBM). separated Irom delibrjnated blood by means of Lymphoprep (Nyco. Oslo. Norway). The PUM were incubated (.17 C. 5",, CO,) for 2 h in tissue culture (lasks with medium RPMI 1640 and 2tr^ pooled normal human serum. The non-adherent cells were then removed by washing the flasks twice wiih medium and were subsequently used for the preparalioii of T and B cells (below), Theadhcrent ceils were Ihcn washed again. incubated for 90 min on ice with medium RPMI 1640 containing 30"^ normal human serum and EiDTA (3..^ ;^g ml), and could then be detached by vigorous pipeiting. 85-95"o of the cells thus prepared were able to ingest latex particles (Bacto-Latex 0.81, Difco. USA) and showed monocyte morphology [7]. T lymphocytes. A T cell-enriched population was prepared by allowing the non-adherenl cclK obtained after culturing PBM for maerophage preparation to form rosettes wiih 2-an>inoeihyl-isoihiuroniuni bromide {AET)-treated sheep erythrocyies (SRBC) [I2|. The rosettes were separated from the non-roseiling cells by cent rifu gat ion on a Lymphoprep gradient. The non-rosctting cells were used for ihe preparation ol' B cells (below). The roselies were then suspended in 1 ml of fresh human serum and incubated for 15 min at 37 C. resulting in lysis of the SRBC by the naturally occurring anti-sheep antibodies in human serum. After incubation of the cells in tissue culture flasks overnight to remove more of Ihe adherent cells. the rosetting/separaiion/hacniolysis setiiience was repealed once, resulting in a T cell fraction containing more than 85" „ H rosciie-forming cells, less than 5°,, Ig'positive cells, and less ihan 0.5''» latex-ingesting cells. B lymphocytes. A B cell-enriched population was purified from the non-rosettlng cells of ihe first rosetting procedure above. The cells were cultured in tissue culture llasks overnight (37 C. 5",, CO^) followed by another rosetling and Lymphoprep separalion to remove any remaining T cells. The cells thus obtained contained more than 65" „ Ig positive cells, 6 II" „ lalex-
ingesting cells.and less than 2"„ E roselte-forming cells. This fraciion will be referred to as B, cells. In some experiments ihis fraction was purilied further in order 10 remove the rest of the adherent cells. 2 5 - 10" B, cells were suspended in 5 ml of medium RPMI 1640 with 5" „ pooled normal human serum and incubaied for 30 min at 37 C in a glass bead column ic. 20 ml of glass beads, 0.5 mm diameter). The non-adherenl B cells were eiuted by flushing the columns with 50 ml medium. This fraction (B.jl always contained less than !"„ latexingesting cells and more ihan 60" „ of the cells carried cell-membrane bound Ig and reacted with ami DR antisera. In the TNP experiments, B cell-enriched suspensions were produced from PBM by first removing adherent cells {2 h culture in tissue culture (lasks) and then rosetting by AET-treated SRBC. The interface cells were then incubated in ihe glass bead column, as described. These B-cell-enriched ceil suspensions also always contained less than I",, latex-ingesting cells and 60-9^)"„ Ig positive and DR positive cells. Non-T cells. A non-T-cell fraction (B cells • macrophages) was prepared as described above by rosetting PBM cells with AFT-lrealed SRBC, wiihoui previous ineubalion in tissue culture flasks. The inierface cells after Lymphoprep notation contained 50 60"o Igposilive cells and 20-30",, latex-ingesting cells. Cell ctdture techniques, (a) The FPD experiments were performed in flal-boltomcd microtitre plates (TC dish 3596. Costar. USA). The following protocol was followed unless otherwise stated: triplicates of 5 < \0* T lymphocytes were co-cullured with 0.2-50 .•; 10' Xirnidiaied (20(K) rad) accessory cells (M9. B, or B, fraction) in a toial volume of 170 \i\ of tissue culture medium (RPMI 1640 supplemented with antibiotics and 20" „ heal-inaelivated pooled normal human serum), with or without PPD (2.5 [ig/ml). at 37 C and 5'„ COj. 'H-thymidine was added on day 4 and the cultures were harvested on day 5 by means of a semi-aulomaiic multiple culture harvester (Skatron. Lierbyen. Norway). *H incorporalion was assessed by means of a liquid scintillation counter. The results were expressed as mean cpm : SE. By 'incremental PPD response' is meant ihe response in cpm with PPD minus the response of the same culture wiihoui PPD. (bl The TNP experiments were performed as previously described [15). Briefly. 3 5 • 10" T cells were mixed with an equal number of TNP-treated aulologous 2000-rad-irradiated non-T cells in medium in 10 ml lubes (Falcon cat. no. 3033). Afler 9-10 days. during which period half of the medium bad been replaced once (days 3 5). 0.51 ^ 10* cells were transferred to each well of round-bottomed Linbro microlitre plates. They were ihen mixed with an equal amount of 20CX)-rad-irradialed non-lreated or TNPirealed cells prepared as described above from u fresh bleeding of ihe donor The cells were then cultured for 48 72 h. including 24 h after addition of 0.2 ml of physiological saline containing lOuCi of *H-lhymidine. and harvested as described above. In some experiments we followed the protocol described by Charmol & Mawas [6]. priming T cells with non-T cells in tissue ciillure llasks (Falcon no. 3013) for 6 days, followed by restimiilaiion in microiiire plates for 48 h. as described above
HLA-D Restriction of T Cell Activation The results were expressed as median cpm of iriplicatcs fto save space). As TNP-speeific response (A) was used the median cpm in cultures restimubted witb TNP-treated cells, afier subtraction of the median cpm values obtained after restimulation with non-lreaied cells.
269
results consistent with the above. The results of one experiment are shown in Fig. I. Both Table I and Fig. I demonstrate that an optimal T cell response occurred at a T/M9 ratio between 10:1 and 5:1 dropping markedly at higher Mtp concentrations. The response to PPD in conjunction with B, cells, on the other hand, was strongest at a T/B ratio of 1:1. Higher B cell concentrations were not tested.
RESULTS PPD presentation by different tioii- T cett fractions
Relation between PPD-presenting activity attd conleiil of phagocytositif! celts
In two experimenis 5x 10* T cells from a total of four different donors were stimulated with PPD in the presence of various numbers of aulologous non-T accessory eells (M 9. B, or B^,). Both experiments yielded consistent results. Table I shows the results of one of these experiments involving cells from two different donors and comparing the PPD-presenting capacities of the M-p. Bl and B^. fractions. Both the M^and the B, fractions were able to present PPD to the T cells, whiie the Bj fraction was not. Three more experiments were performed testing various numbers of B, cells and M9 from a total of seven donors. These experiments yielded
The above results indicate that the ability of a B cell fraction to present PPD to T cells is determined by its content of phagocytosing ceiis. This is more clearly demonstrated in Fig. 2, where the results of two of the experiments are presented, relating the incremental PPD response to the absolute number of phagocyiosing (latex ingesting) cells among the accessory cells (B| or M9). It appears that the PPD-presenting ability of non-T cells is a function of the content of phagocytosing cells, irrespective of whether or not additional B cells are present.
I. Purified protein derivative (PPD) presentation by different autologotis non-T fractions
No. of irradiated aulologous accessory cells
.•\UH)I.
T^eeH E.S.
E.T.
access. cells
Pi'D ((jg/m!)
0
200
1000
5000
10.000
50.000
(91" J t
0 2.5
66 : 12* 206 : 36 140:
77 :..25 1065 • .16 988
307 4-30 2990 : 174 2683
355 • 12 6350 ' 274 5995
409* 18 4780 143 4371
395:1^50 1911-59 1516
B, (6%)
0 2.5
24-i 5 480 : 40 456
155^7 615 t 102 460
153 ^ 6 1402 1 146 1249
470- 127 3111 f^29l 2641
795 i 83 .1419 ±176 2624
Bl ( 60"„ Ig and HLA-DR positive cells. t > 95" n latex ingesting cells. § Median cpm of triplicates. "^ TNP-specific response.
clear TNP-specific response was always seen using 10.000 or more B cells for restimulation. Also, the same number of TNP-treated B cells usually caused a stronger restimulation than ihe same number of TNP-treated macrophages.
In other experiments, 100,000 primed T cells were restimulated with an equal amount of non-treated or TNP-treated non-T cells. B cells or TNP-treated B cells together with 20% non-treated macrophages. In nine of twelve
272
B. O. Berghollz & E. Thorsby
experiments performed, a clear TNP-specific response was obtained after restimulation with TNP-treated autologous non-T cells, and in six of these restimulation with TNP-treated B cells gave rise to a TNP-specific response in the range of 50-IOO/o of the response seen using TNP-trcatcd non-T cells. This response was not or only slightly increased after addition of non-treated autotogous macrophages (data not shown).
DISCUSSION In previous studies we have shown that the in vitro PPD activation of T lymphocytes from sensitized individuals is HLA-D/DR-restricted; i.e. it requires that the antigen be recognized in the context of "self HLA-D/DR determinants of viable non-T accessory cells [3-5]. The PPD-responsive T cells do not express the HLA-D/DR determinants in detectable concentrations [1]. In conformity with the current concepts of the role of macrophages (M9) in the immunological network [11], we used the plasticadberent non-T population as our source of accessory cells. In this report we present evidence that the ability to present antigen (PPD) to T cells is an exclusive property of this M(? population. Although certainly expressing the HLA-D/DR determinants, B cells devoid of phagocytosing cells have no sucb ability (B^ fraction. Table I). The ability of a non-T cell population to present PPD to T cells is mainly determined by its content of phagocytosing cells (Fig. 2). It is mainly PPD bound to the accessory cells that stimulates T cells [5]. The simplest explanation of the inability of B cells to substitute for macrophages in antigen presentation would therefore be that B ceils lack the capacity to unspecifically bind antigen to their stirface. A selective purification of B cells bearing specific Ig receptors for PPD might yield a B cell fraction capable o\' binding and presenting PPD in the context of self HLA-D to T ceils. We have not been able to perform this kind of experiment. The T cell response to PPD was shown to be inhibited by an excess of M9—that is, at a T/Mtp ratio lower than 5:1. The presence of surplus B cells, on the other hand, had only a
weak inhibitory effect, if any at all (Tables i and II, Figs. 1 and 2). Inhibition of T cell responses in the presence of an excess of M9 has been described in various systems and may be explained in a variety of ways, such as production of inhibitory metabolites or enzymes, a direct inhibitory effect on T cells, or simply exhaustion of the medium [9, 10, 16]. The poor T cell response to PFD in conjunction with HLA-D/DR-incompatible allogeneie Mtp was not improved by the addition of autologous (HLA-D/DR-bearing) B cells (Table II). The possibility that suppressor cells were generated as the result of MLC activation by allogeneic M9 has been ruled out in previous experiments [4]. Assuming that the activation of T cells by Mtp-associated PPD involves two separate, independently distributed T cell receptor types, one with specificity for PPD and one for "self HLA-D/DR, one might have expected a response to take place under these experimental conditions, since the T cells had access both to M9-associated PPD and to autologous HLA-D/DR determinants. However, the above results indicate that the activation of T cells by PPD requires that bolh PPD and autologous HLA-D/DR antigens be displayed on the surface of the same cell. In previous experiments we showed that the specific secondary response of T lymphocytes primed to TNP-treated autologous cells in vitro is. at least in part, also restricted by seif-HLAD [15]. This was shown by using allogeneic cells expressing different HLA-D/DR determinants for restimulation and by the observation that TNP-treated autologous T cells, not expressing HLA-D/DR, were unable to induce a secondary response. Here we investigated whether TNP-treated B lymphocytes alone, depleted of macrophages, were able to restimulate TNP-primed ceils. Our data strongly suggest that TNP-treated B cells alone in most instances are able to induce a secondary response of primed T cells. When the same number of TNP-treated B cells and macrophages were compared, the B cells induced a stronger secondary TNP-specific response than macrophages. Furthermore, our B cell suspensions contained less than 1";, latex-ingesting cells, and this possible contamination of macrophages (500-1000 cells) was not able to induce a secondary TNP-specific response. In other experiments we found that the B cell
HLA-D Restriction of T Cell Activation suspensions were able to induce a response of allogeneic PBM and T cell-enriched lymphocytes of a degree similar to that induced by PBM stimulating cells prepared from the same donors [unpublished]. Thomas et at. [13] have studied these questions, using T cells from guinea-pigs sensitized in vivo with picryl-chloride and restimulating them with TNP-treated la positive EN-L2C B cell-like leukaemia cells. In contrast to similarly treated macrophages, these TNP-treated leukaemia cells were unable to induce a TNPspecific response. The authors state, however, that it might be that different results would have been obtained using normal and not malignant B cells. We are also currently studying whether TNP-treated B cell lines might induce a secondary TNP-specific response. The detailed events of T cell recognition of antigen are still poorly understood. However, the present results yield some information as to the nature of the participating accessory cells. The expression of HLA-D/DR determinants is necessary but, at least for soluble antigens, is not by itself sufficient since B cells are unable to present PPD to autologous T cells. It may be concluded that the proliferative T cell response to PPD requires non-T accessory cells which are characterized by monocyte morphology, phagocytosing capacity and ability to adhere to plastic and glass surfaces, in addition to their ability to bind and display PPD in conjtinction with the HLA-D/DR molecules. In contrast, haptens such asTNP are able to induce an HLA-D/DR-restricted specific secondary response of in vitro primed T cells also when conjugated to B cells. The explanation may be that TNP binds to most cell membranes and that modification of the TNPcell membrane complex by phagocytosing cells are not necessary, at least for inducing a response of primed T cells. However, as demonstrated in this and a previous article [15], the TNP-conjugated cells must express the self HLA-D/DR determinants to be able to induce a TNP-specific secondary response.
ACKNOWLEDGMFNTS This work has been supported by grants from the Norwegian Council for Science and the Humanities, Mr Anders Jahre's Fund for the
273
Promotion of Science and Laurine Maarschalk's Ftind. We wish to thank A. B. Thoresen and H. Nottsiainen for excellent technical assistance and E. Garmann and K. Goyer for typing and correcting the manuscript. REFERENCES 1 Albrechtsen. D.. Bratlie. A., Nousiainen. H., Solheim. B.G., Winther. N. & Thorsby, E. Serological typing of HLA-D; predictive value in mixed lymphocyte cultures (MLC). Immunogenetics. 6, 91, 1978. 2 Albrechtsen. D.. Solheim, B.G. & Thorsby, E. The presence of la-like determinants on a subpopulaCion of human T lymphocytes. Immunogeneiics. 5, t49, 1977. .1 Bergholl/, B.O. & Thorsby, E. Macrophagedependent response of immune human T lymphocytes to PPD in vitro. Scand. J. Immunol. 6, 779, 1977. 4 Bergholtz, BO. & Thorsby, E. HLA-D reslriclion of the miicrophagc-depenilent response of immune human T lymphocyfes to PPD in viiro: inhibition by anii-HLA-DR anlisera. Scand. J. humunol. 8, 63. 1978. 5 Bergholtz. B.O. & Thorsby. E. Macrophage/T lymphocyte interaction in Ihe immune response to PPD in humans. Scand. J. Immunol. 9, 511. t979. 6 Charmol. D. & Mawas, C. The in vitro cellular response of human lymphocytes to trinilrophenylated autologous cells: HLA-D restriction of proliferation but apparent absence of HLA restriction of cytolysis. Eur. J. Immunol, in press. 7 Furth. R. van. Cohn, Z.A.. Hirsch. J,B.. Spector. W.G. & Langcvoort. H.L. The mononuclear phagocyte system: a new classification of macrophages. monocytes, and their precursor cells. Bull. WId HItk Org. 46, 845. 1972, 8 Kaakingn, A.. Helgesen. A., Nousiainen, H., Solheim. B.G. & Thorsby, E. The HLA-D determinanls. A population study of Norwegians. Immurwgenetics. 4, 205, 1977. 9 Krtiisbcek. A.M. Helper and suppressor effects of macrophages in lymphocyte activation: a quantitative phenomenon? In Proceedings of the 12th International Leuf
