- Email: [email protected]
Activated Human T-Cells Bestow T-Cell Antigens to Non-T-Cells by Intercellular Antigen Transfer
Activated Human T-Cells Bestow T-Cell Antigens to Non-T-Cells by Intercellular Antigen Transfer Ruth Rabinowitz, Russell Pokroy, Yongmao Yu, and Michael Schlesinger ABSTRACT: The mechanism of the appearance of T-cell antigens on B-cells, following in vitro activation of peripheral blood lymphocytes, was analyzed using the following model: Purified T-cell suspensions were activated by exposure to phytohemagglutinin (PHA) for 3 days, and then incubated for one hour in the presence of cells of either Raji or K562 cells. The expression of T-cell antigens on the cell lines was determined using immunofluorescent F(ab)2 fragments of monoclonal antibodies (mAbs). Following exposure of the CD191 Raji cells to activated T lymphocytes, 87.6% of the CD191 cells coexpressed CD2. A large proportion of the CD191 cells also expressed CD4, CD5, and CD8 antigens. Similar results were obtained with Raji cells that were prelabeled
with calcein AM. In Raji cells, which were rendered CD51 following incubation with activated T cells, only a negligible level of CD5 mRNA was detected with a sensitive RT-PCR technique, probably attributable to contamination with T cells. K562 cells incubated with activated T cells acquired CD2 but not the CD4 and CD8 antigens. Exposure of either Raji or K562 cells to mAb against CD58 inhibited the transfer of CD2. The present study indicates that following their activation, T-cells gain the capacity to transfer T-cell antigens to non-T cells and that CD2 and CD58 molecules are involved in this process. Human Immunology 59, 331–342 (1998). © American Society for Histocompatibility and Immunogenetics, 1998. Published by Elsevier Science Inc.
ABBREVIATIONS FITC fluorescein isothiocyanate mAb monoclonal antibody PBL peripheral blood lymphocytes PBS phosphate-buffered saline
PE PHA RT-PCR
INTRODUCTION Activation of lymphocytes from a distinct subset can elicit the appearance of antigens that are characteristic for another subset. Thus, CD8 antigenicity can be detected on the surface of lymphocytes of the CD4 cell lineage following mitogenic stimulation of human peripheral blood lymphocytes (PBL) [1, 2]. Neoplastic B-cells cultivated for 3–5 days in the presence of T cells and From the The Hubert H. Humphrey Center for Experimental Medicine and Cancer Research, The Hebrew University-Hadassah Medical School, Jerusalem, 91120 Israel. Present address of Russell Pokroy: Department of Ophthalmology, Kaplan Medical Center, Rehovot, Israel. Address reprint requests to: Prof. Michael Schlesinger, The Hubert H. Humphrey Center for Experimental Medicine and Cancer Research, The Hebrew University—Hadassah Medical School, Jerusalem 91120, Israel; Tel: 972-2-6758334; Fax: 972-2-6414583. Received February 11, 1998; revised March 23, 1998; accepted March 23, 1998. Human Immunology 59, 331–342 (1998) © American Society for Histocompatibility and Immunogenetics, 1998 Published by Elsevier Science Inc.
phycoerythrin phytohemagglutinin reverse transcription polymerase chain reaction
mitogens acquire concomitantly the CD2 marker characteristic for T cells and the capacity to form rosettes with sheep red blood cells (SRBC) [3]. The expression of receptors for SRBC on B-cells depends on the presence of T cells in the culture [3]. Similarly, normal human B-cells become CD2-positive following culture with either autologous or allogeneic T cells and mitogen [4]. A number of studies suggested that acquisition of the CD8 antigen by CD41 cells [1, 2, 5] and of CD2 by B-cells [3, 4] may result from induction of the expression of these markers. In contrast, other studies indicated that various antigens can be transferred from one lymphocyte population to another. Thy-1 molecules become adsorbed to the surface of murine B cells when exposed in vitro to cells of Th1 clones [6]. Glycosyl-phosphatidylinositol-(GPI)-linked-CD4 showed a high level of intercellular transfer [7]. Cloned lines of alloreactive murine 0198-8859/98/$19.00 PII S0198-8859(98)00029-9
332
T lymphocytes acquired class I and class II MHC antigens from the stimulating cell population [8]. Human platelets adsorb class I HLA molecules from plasma [9] and a considerable portion of platelet HLA could be eluted [10, 11]. We have recently provided evidence that the appearance of CD41CD81 cells among activated human PBL, may result from the transfer of CD8 molecules to CD41 cells [12]. In a previous study [13], we have observed that in vitro stimulation of PBL for three days by phytohemagglutinin (PHA) led to expression of T-cell markers on human B cells. The majority of human B cells in PHA-stimulated cultures displayed CD2 on their cell surface, and a high proportion also reacted with the CD8 and CD4 monoclonal antibodies (mAbs). The use of fluorescent F(ab9)2 fragments of mAbs ensured that the staining of B cells by mAbs to T-cell markers did not result from their binding to Fc receptors. CD4 and CD8 were detected on B-cells only when T cells expressing these antigens were present in the culture. It was suggested, therefore, that following activation T-cell antigens were transferred to B cells [13]. The aim of the present study was to analyze the mechanism of the expression of T-cell antigens on B cells. Instead of exposing both T and B cells to a mitogenic stimulus, only purified human T lymphocytes were stimulated by exposure to PHA for three days. The activated T cells were incubated for one hour in the presence of a human non-T-cell lines and the capacity of activated T cells to bestow the expression of T-cell antigens on non-T-cells was determined.
MATERIALS AND METHODS Cell Lines The Raji [14] B-cell line and the K562 chronic myelogenous leukemia cell line [15] were used. The cells were grown in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine, penicillin G (100 mg/ml) and streptomycin (100 mg/ml). Isolation of Peripheral Blood T Cells Buffy coat from blood donated by volunteers was obtained from the Blood Bank of the Hadassah Hospital. Lymphocytes were isolated by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density gradient centrifugation. T cells were enriched by applying 1 3 108 mononuclear cells in a volume of 4 ml to a column containing 1.2 g combed, scrubbed nylon wool, packed to a volume of 12 ml (Robbins Scientific Corp., Sunnyvale, CA). The column was incubated for 60 min at 37°C, and the nonadherent cell fraction was eluted. The T-cell-enriched fraction contained less than 1% CD191 lymphocytes.
R. Rabinowitz et al.
Selection of Peripheral Blood B Lymphocytes Lymphocyte suspensions were incubated with MACS magnetic microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) coupled covalently to CD19 mAbs. The cell suspension was passed through a magnetized matrix. The CD19 antigen-bearing cells, which bound to magnetic microbeads, were concentrated by passage through the magnetizable matrix. Over 80% of these cells expressed CD20. Stimulation of Peripheral Blood T Cells Suspensions enriched for T lymphocytes, containing 1 3 106 cells/ml were prepared in RPMI-1640 medium supplemented with 15% FCS, 2 mM L-glutamine and antibiotics. Five ml aliquots of cell suspensions were introduced into 75 ml tissue culture flasks, and 10 mg phytohemagglutinin (PHA) (Sigma Chemical Co., St. Louis, MO) was added to each flask. The cultures were incubated at 37°C in a humidified atmosphere containing 5% CO2 and 95% air for 3 days. In some experiments T cells were exposed for 3 days to insolubilized agarose-bound PHA at a concentration of 25 mg/ml (Sigma Chemical Co.). Co-Incubation of Activated T Cells and Non-T-Cell Lines To determine if activated T cells can transfer T-cell antigens to Raji B cells, 5 3 106 activated T cells were co-incubated for one hour at 37°C with 2.5 3 106 Raji cells with intermittent agitation. In control experiments, non-activated T cells were mixed with Raji cells. Similar experiments were carried out also with mixtures of activated T cells and K562 or normal, unstimulated B cells. After incubation, the mixtures were washed twice with PBS and stained with immunofluorescent mAbs. To prevent direct cell-cell contact between Raji cells and activated T cells, in some experiments suspensions of purified T lymphocytes and Raji B cells were kept separated by a permeable membrane with a 0.4 mm pore size (Corning Hydrophilized PTEE). Membrane inserts with a 30 mm diameter were used in conjunction with 6-well tissue culture plates. The total volume of the culture medium was 5 ml. Suspensions of Raji cells (3 3 106 cells) were introduced above the filter, while T-cell suspensions (10 3 106 cells) were kept in the wells below the filter, and the cultures were exposed to PHA for 3 days. Monoclonal Antibodies The presence of T-cell antigens on various non-T-cell lines was determined with fluorescent F(ab9)2 fragments of mAbs. These were: PE anti-CD2, PE anti-CD4, PE anti-CD8, and FITC anti-CD19, purchased from Zymed Laboratories, South San Francisco, CA. Anti-CD5 mAbs
T-Cell Antigen Intercellular Transfer
FIGURE 1 Flow cytometric analysis of mixtures of calcein AM labeled Raji cells and activated T-cells. Raji cells, which were prelabeled with calcein AM, were incubated for one hour in the presence of PHA-activated T cells. The fluorescence elicited by calcein AM was recorded as fluorescence 1. The cell mixtures were exposed in addition to the following PE-labeled reagents: F(ab9)2 fragments of IgG1 (A), of anti-CD2 mAb (B), of anti-CD4 (C), of anti-CD8 (D) and PE-labeled whole antiCD5 mAb. Log10 fluorescence 2 (ordinate): staining with PE-labeled reagents. The two right hand squares constitute calcein AM-positive cells. The percentage of calcein AMpositive cells that reacted with PE-labeled reagents was: A: 1.9%, B: 94.9%, C: 62.6%, D: 34.3%, E: 62.6%.
were either PE-Leu-1, obtained from Becton-Dickinson (Mountain View, CA) or PE-CD5, obtained from Pharmingen (San Diego, CA). FITC anti-CD15 mAb was purchased from Becton-Dickinson, and FITC anti-CD20 mAb (B1) from Coulter Immunology (Hialeah, FL). Various mAbs were tested for their capacity to block the transfer of T-cell antigens. These mAbs included the following: anti-CD72 obtained from Serotec (Oxford, England), mAbs to CD11a (LFA-1 a chain, clone TS 1/22), HLA-DR (clone L243) and HLA-ABC (clone W6/ 32) which were produced by hybridomas obtained from the American Type Culture Collection (ATCC). mAbs to CD58 (LFA-3), were produced by hybridoma clone 1C3 obtained from Prof. S. Meuer.
333
Antibody Treatment of T Cells or of B-Cell Line Saturating quantities of various mAbs were added to 2.5 3 106 T cells or Raji cells suspended in 0.5 ml culture medium and the mixture was kept at 4°C for 45 min. Antibody-coated cells were washed 3 times in PBS before co-incubation with other cells. Control cells were subjected to similar treatment, the only difference being the omission of mAbs. Immunofluorescent Assays As described previously [12–13], two fluorescent mAbs (PE and FITC labeled) were added to 50 ml cell suspension containing 107 cells/ml. The mixture of cell suspension and mAbs was kept for 30 min at 4°C, and washed twice with cold PBS containing 0.1% NaN3 and 0.1% bovine serum albumin. The washed cell pellet were resuspended in 1% paraformaldehyde and kept at 4°C until they were analyzed by flow cytometry. In some experiments, either Raji cells or K562 cells were stained prior to their incubation in the presence of activated T cells. Raji cells were stained either with FITC anti-CD19 F(ab9)2 fragment or with calcein AM. Raji cells were incubated with calcein AM (Molecular Probes, Netherlands) at a final concentration of 0.01 mg/ml for 30 min at 37°C. In some experiments, activated T cells
334
R. Rabinowitz et al.
incubated for 1 hour with activated T lymphocytes, and the expression of T-cell antigens on the cells was determined as described above. Flow Cytometry Acquisition and analysis were carried out with a FACScan flow cytometer (Becton-Dickinson, Mountain View, CA) using the Facscan program. For each test, 104 cells were acquired and tested by two-color flow cytometry. PE labeled mAbs were used to detect T-cell markers expressed on non-T cells while FITC labeled markers were used to detect the target cell population. Either the whole cell population was analyzed or only the subpopulation of target cells. This was accomplished by first gating according to the typical forward light scatter and right angle light scatter of these cells. Then the FITC marker was set so as to collect over 98% of the FITC positive cells. In some experiments, cells showing either high or low staining with PE labeled mAbs were analyzed separately. The cut-off point between bright and dim cell subsets was set at 516 fluorescence intensity units. Cell sorting was performed with a FACStar flow cytometer (Becton-Dickinson). Raji cells were gated according to their typical forward and right angle light scatter. The FITC marker was set so as to collect over 98% of the CD191 cells, and the PE marker was set to collect PE negative and PE-dimly positive cells.
FIGURE 2 Flow cytometric analysis of CD191 cells among mixtures of Raji cells and activated T-cells co-incubated for 1 h. Scattergrams in left column: all the cells in the mixture were analyzed. Scattergrams in right column: only cells gated for CD19 positivity were analyzed. All of the cells were exposed to FITC labeled F(ab9)2 fragments of anti-CD19 mAb. In addition, the cells were exposed to PE F(ab9)2 fragments of IgG1 (A,a), anti-CD2 mAb (B,b), anti-CD4 (C,c), or anti-CD8 (D,d) and to PE-labeled whole anti-CD5 mAb (E,e). Log10 fluorescence 1 (abscissa): staining with FITC-labeled F(ab9)2 fragments of anti-CD19 mAb, Log10 fluorescence 2 (ordinate): staining with PE-labeled reagents.
were incubated with dihydroethidine (Molecular Probes, Netherlands) at a final concentration of 2 mg/ml for 30 min at 37°C. K562 cells were stained with FITC antiCD15 mAb. Thereafter, Raji or K562 cells were washed,
Reverse Transcription Polymerase Chain Reaction (RT-PCR) The CD5 specific primers described by Kasaian et al. [16] were used for PCR. Upstream: 59 AGG ACG GAT GGC ACA TGG TTT 39, downstream: 59 TTG TCC TGG GCC TCA TAG CT 39. The expected RT-PCR product is 420 bp, which encompasses the sequence corresponding to residues 230 – 647 of pT1-1 CD5 cDNA [17]. CD5 mRNA was reverse transcribed and the cDNA was amplified in a 50 ml volume, using the Access RT-PCR system (Promega). The amount of total RNA template ranged from 0.5 pg to 1 mg, while the final concentration of each of the primers was 2 mM. The other components included in each reaction tube were avian myeloblastosis virus (AMV) reverse transcriptase (0.1 u/ml), Thermus flavus (Tfl) DNA polymerase (0.1 u/ml), AMV/Tfl reaction buffer, MgSO4 (2 mM), and dNTP mixture (0.2 mM). For reverse transcription, the reaction mixture was incubated at 48°C for 45 min. After inactivation of the AMV reverse transcriptase at 94°C for 5 min, the CD5 cDNA was amplified by 30 cycles of denaturing (94°C, 1 min), annealing (55°C, 1 min) and extension (72°C, 2 min). Aliquots of RT-PCR products, at a volume of 20 ml were loaded to 1.5% agarose gels, containing 0.5 mg/ml ethidium bromide,
T-Cell Antigen Intercellular Transfer
FIGURE 3 Flow cytometric analysis of Raji cells and activated T cells incubated either alone or as a mixture. Raji cells and PHA-activated T cells were either kept alone or mixed and incubated for one hour. All the cells were exposed to FITC F(ab9)2 fragments of anti-CD19 mAb. In the top row, the cells were co-incubated with PE-labeled F(ab9)2 fragments of IgG1, while in the bottom row, the cells were co-incubated with PE-labeled F(ab9)2 fragments of anti-CD2 mAb. Log10 fluorescence 1 (abscissa) and Log10 fluorescence 2 (ordinate): see legend Fig. 2.
and electrophoresed. Bands of RT-PCR products were visualized in a UV transilluminator. Southern Blotting Separated DNA was transferred onto positively charged nylon membranes (Boehringer, Mannheim) and hybridized with digoxigenin (DIG) labeled oligo-probe (59TTG GAG GTG TTG TCT TCT GG-39, encompassing residues 447– 496 of the anti-sense strand of CD5 cDNA sequence [17]. The DIG-labeled hybrids were detected with anti-DIG-alkaline phosphatase conjugate and the substrates NBT (nitroblue tetrazolium salt) and BCIP (5-bromo-4-chloro-3-indoyl phosphate toluidimium), added according to the recommendations of the producer (Boehringer, Mannheim).
335
RESULTS The Appearance of T-Cell Antigens on Raji Cells The Raji B-cell line expresses CD19 and other B-cell surface markers but no T-cell antigens [18]. Within 1 h of incubation in the presence of activated human T cells, Raji cells acquired T-cell antigens on their surface. Figure 1 depicts results obtained with calcein-AM-labeled Raji cells that were incubated for 1 hour with activated T cells and were then stained with PE labeled F(ab9)2 fragments of anti-CD2, CD4, and CD8 or PE anti-CD5 mAbs. The majority of the Raji cells acquired the CD2 antigen following exposure to activated T lymphocytes. A high proportion of the Raji cells also reacted with anti-CD4, CD5 and CD8 mAbs (Fig. 1). Similar results were obtained when Raji cells incubated with activated T-cells were stained with FITC-labeled F(ab9)2 fragments of anti-CD19 mAb and PE-labeled F(ab9)2 fragments of anti-CD2, CD4, and CD8 or PE anti-CD5 mAb (Fig. 2). Raji cells by themselves were not stained by any of the mAbs against T-cell antigens, while purified Tcells were not stained by anti-CD19 mAb. Figure 3 illustrates the fact that F(ab9)2 fragments of anti-CD2 mAb stained activated T-cells but not Raji cells, while F(ab9)2 fragments of CD19 mAb stained Raji cells but
336
R. Rabinowitz et al.
TABLE 1 The expression of T-cell markers on CD191 Raji cells
T-cell antigen CD2 CD4 CD8 CD5
Proportion of CD191 expressing T-cell antigen*
Proportion of CD191 expressing high levels of T-cell antigen**
Proportion of CD191 expressing low levels of T-cell antigen**
87.63 6 2.30 59.93 6 5.50 52.03 6 6.32 72.65 6 6.53
27.83 6 1.70 19.30 6 2.62 10.53 6 1.30 30.14 6 2.30
59.80 6 8.38 40.63 6 4.89 41.50 6 4.17 42.51 6 6.37
Raji cells were incubated for one hour in the presence of T cells which had been stimulated for 3 days by exposure to PHA. The cells were stained with FITC-labelled F(ab)2 fragments of CD19 mAb and PE-labelled F(ab)2 fragments of mAb against T-cell antigens. * The proportion of cells expressing various T-cell antigens among CD191 cells (i.e., among cells in quadrants 2 and 4). ** The proportion of cells showing either high or low levels of T-cell antigens were calculated separately. Results shown are means (6 S.E.) obtained in 10 separate experiments.
failed to stain activated, purified T cells. In contrast, following incubation of T cells and Raji cells for one hour the majority of the CD191 Raji cells co-expressed CD2. Table 1 summarizes findings obtained in 10 repeated experiments in which Raji cells were exposed for one hour to activated T cells. Among CD191 cells, 87.6% co-expressed the CD2 marker. The majority of the CD191 cells also expressed the CD4, CD5, and CD8 markers, although the proportion of CD191 cells expressing these antigens was lower than the proportion of those expressing CD2. Among the Raji cells which expressed T-cell antigens following exposure to activated T lymphocytes, the majority showed dim staining with mAbs to T-cell antigens. In addition, following exposure to activated T lymphocytes, a low proportion of Raji cells showed bright staining with mAbs to T-cell antigens. This was seen regardless of whether the Raji cells were stained with CD19 mAb or calcein AM (Figs. 1–3, Table 1). The staining intensity of the Raji cells which reacted strongly with mAbs to CD2, CD4, CD5, or CD8 resembled the staining intensity of T cells with these mAbs. It seems that the cells which showed bright staining with antiT-cell mAbs represent small aggregates of activated T cells and Raji cells. This conclusion is supported by the finding that the sum of the percentage of CD4bright CD191 and CD8bright CD191 cells equaled that of the percentage of either CD2bright CD191 or CD5bright CD191 cells. Furthermore, when activated T lymphocytes were rendered red fluorescent by labeling with dihydroethidine prior to their incubation with calceinAm-labeled Raji cells, 20 –23% of the Raji cells were found to acquire red fluorescence (not shown). It should be noted that it proved impossible to determine the presence of small T-B aggregates by FSC-SSC plots. While the FSC and SSC of most Raji cells was larger than that of activated T cells, the overlap between
large T cell blasts and Raji cells prevented the detection of small aggregates. In contrast to Raji cells which stained brightly with mAbs against T cell antigens, the staining intensity of Raji cells which displayed dim staining with mAbs to T cell antigens was considerably lower than among T cells. Moreover, the sum of the proportion of CD191 cells which were CD4dim and of those that were CD8dim clearly exceeded the proportion of CD191 cells that were either CD2dim or CD5dim. These observations indicate that CD191 cells which showed dim staining with mAbs against T cells constitute B cells which acquired T-cell antigens. When normal, non-activated B cells, rather than Raji cells, were incubated with activated T cells for 1 hour, less than 6% of the B cells expressed T-cell antigens (not shown). In the experiments described above, activation of T cells was accomplished by their exposure to soluble PHA. Although activated T lymphocytes were washed thoroughly prior to their incubation with Raji cells, the possibility could be raised that the mitogen caused sticking of T-cell surface fragments to the surface of B cells. In order to rule out this possibility, T cells were activated for 3 days by exposure to insoluble PHA prior to incubation with Raji cells. The same acquisition of T-cell antigens by Raji cells was noted as with T cells activated by soluble PHA (not shown). The question arose whether Raji cells had to be in direct contact with activated T cells for T-cell antigen acquisition. In order to elucidate this issue, suspensions of T lymphocytes and Raji B cells were exposed for 3 days to PHA, while being kept separated by a permeable membrane with a 0.4 mm pore size. In repeated experiments, no significant acquisition of CD2, CD4, CD5, or CD8 by Raji cells could be detected (not shown). In order to exclude the possibility that activated T
T-Cell Antigen Intercellular Transfer
FIGURE 4 The effect of cycloheximide on the staining of Raji cells by mAbs to CD2 and CD5. Raji cells were incubated for one hour in the presence of PHA-activated T cells, either in the absence (top row A,B,C) or in the presence of cycloheximide (20 mg/ml) (bottom row a,b,c). All the cell mixtures were exposed to FITC F(ab9)2 fragments of anti-CD19 mAb. In addition, the cells were incubated with PE-labeled F(ab9)2 fragments of IgG1 (A,a), PE-labeled F(ab9)2 fragments of anti-CD2 mAb (B,b) or PE-labeled anti-CD5 mAb (C,c). The percentage of FITC-labeled, CD191 cells that reacted with PE-labeled reagents was: A: 3.7%, a: 4.1%, B: 96.9%, b: 96.5%, C: 76.0%, c: 86.5%. Log10 fluorescence 1 (abscissa) and Log10 fluorescence 2 (ordinate): see legend Fig. 2.
cells induced the expression of T-cell antigens on Raji cells, Raji cells were exposed to cycloheximide (20 mg/ ml) for 2 h prior to and during their incubation with activated T cells. This treatment had no effect on the capacity of Raji cells to acquire T-cell antigens, such as CD2 and CD5 (Fig. 4). T-Cell Antigens Appear on Raji Cells Without Synthesis of mRNA CD191 Raji cells which expressed low levels of CD5 following incubation for 1.5 hours in the presence of activated T cells, were separated by FACS sorting (Fig. 5A–C). As shown in Fig. 5C, over 90% of the isolated
337
CD191 cells displayed low levels of CD5. CD51CD192 T cells constituted 3.3% of the sorted suspension. The expression of CD5 mRNA was determined by RT-PCR with CD5 primers (Fig. 6). The 420 bp band characteristic for CD5 cDNA was seen in RT-PCR tests with RNA from Jurkat cells (lane J) and from peripheral blood T cells (lanes a– d). With decreasing concentrations of RNA from T cells, the cDNA band became weaker. A faint line could still be seen with RT-PCR product obtained with 5 ng of RNA derived from T cells (lane d). In contrast, with RNA derived from sorted, dim CD51 Raji cells, a faint band could only be seen with RT-PCR product obtained with 1 mg RNA (lane 1). Thus, it can be estimated that the concentration of CD5 mRNA in the sorted Raji cells attained about 0.5% of that found in T cells. In Southern blots with a DIG-labeled oligo-probe for CD5, the lowest concentration of RT-PCR product from T cells that still gave a faint line was obtained with 0.5 ng RNA (lane e). In contrast, the lowest concentration of CD5 specific RT-PCR product obtained with 50 ng of RNA from sorted CD51 Raji cells (lane 3). Thus, both the direct visualization of RT-PCR products and the results of Southern blotting indicated that the concentration of CD5 mRNA detected in Raji cells that ac-
338
FIGURE 5 Flow cytometric sorting of CD5dim Raji cells. Flow cytometric analysis of either Raji cells alone (A) or Raji cells which were incubated in the presence of activated T cells (B). The population of CD52 and CD5dim cells collected by FACS sorting is indicated in B, while the flow cytometric analysis of the sorted cell population is shown in C. Log10 fluorescence 1 (abscissa): staining with FITC labeled F(ab9)2 fragments of anti-CD19 mAb, Log10 fluorescence 2 (ordinate): staining with PE-labeled CD5 mAb.
quired CD5 antigenicity constituted between 0.5 to 1% of the concentration detected in T cells. This very low level of CD5 mRNA can be attributed to the low contamination of T cells present among the sorted Raji cells. The Acquisition of T-Cell Antigens by K562 Cells A number of mechanisms could underlie the transfer of T-cell antigens to the surface of B cells. It could be envisioned that each of the T-cell markers transferred to B cells would be bound by a ligand specific for that marker. Raji cells express CD58, the ligand of CD2, HLA-DR, the ligand of CD4, HLA-ABC, the ligand of CD8, and CD72, the ligand for CD5 [cf. 19]. It was, therefore, of interest to determine whether T-cell antigen transfer would occur with cells of the K562 line which express only CD58 but none of the other ligands. Incubation of K562 cells for 1 h with activated T-lymphocytes resulted in the transfer of CD2 to the majority of the K562 cells. PE labeled F(ab9)2 fragments of antiCD2 mAb stained 50.6% of the CD151 K562 cells, following their incubation with activated T cells (Fig. 7). Only a negligible percentage of K562 cells incubated with activated T cells expressed CD4, CD8 (Fig. 7) and CD5 (not shown). The Effect of Various Antibodies on T-Cell Antigen Transfer Antibodies against various ligands were tested for their capacity to inhibit the transfer of T-cell antigens from
R. Rabinowitz et al.
activated T lymphocytes to non-T-cell lines. Exposure of K562 cells to mAb against CD58, prior to their incubation with activated T cells, decreased the proportion of CD21 K562 cells from 84.6 6 2.0% to 23.6 6 2.4% (mean 6 S.E. from 6 experiments, p , .0001). Treatment of Raji cells with CD58 mAb, prior to incubation with activated T cells, significantly inhibited the proportion of CD191 co-expressing CD2 from 87.6 6 2.3% to 63.1 6 6.2% (p 5 .0016). CD58 mAb strikingly reduced the proportion of CD191CD2dim cells from 59.8 6 8.4 to 35.2 6 3.6 (p , .0001) but had no significant effect on the appearance of CD191CD2bright cells. Treatment of Raji cells with mAbs against HLAABC, HLA-DR, CD72, or CD11a had no effect on the expression of any of the T-cell antigens tested. DISCUSSION Exposure of PBL for a number of days to T-cell mitogens lead to the appearance of T-cell antigens on B cells [3, 4, 13]. Some studies suggested that the CD2 molecules detected on the surface of B cells cultured with activated T cells were produced by B cells [3, 4]. We showed previously [13] that B cells expressed either the CD4 or the CD8 antigen on their surface depending on the T-cell subpopulation with which they were co-cultured. We suggested, therefore, that T-cell marker were transferred from activated T cells to B cells with which they were in direct contact [13]. The present study showed that nonT-cell lines acquired cell surface T-cell antigens following a brief incubation in the presence of T lymphocytes which had been activated separately for three days by exposure to PHA. CD2, CD4, CD5, and CD8 antigens were expressed on B cells after their exposure for as little as one hour to activated T cells. Most Raji cells which were exposed to activated T cells expressed low levels of T-cell antigens. Some Raji cells displayed, however, high levels of T-cell antigens. A
T-Cell Antigen Intercellular Transfer
FIGURE 6 Analysis of CD5 mRNA expression. CD5dim Raji cells were obtained by FACS sorting after incubation of Raji cell in the presence of activated T cells. RNA isolated from CD5dim Raji cells was analyzed by RT-PCR, upper part, and by Southern blotting, lower part of figure. M: 100 bp DNA ladder, J: RNA from Jurkat cells as positive control, R: RNA from untreated Raji cells, as negative control. Lanes 1– 8: RT-PCR products from CD51 sorted Raji cells. The amount of total RNA used for RT-PCR for each lane was 1: 1000 ng, 2: 500 ng, 3: 50 ng, 4: 5 ng, 5: 0.5 ng , 6: 0.05 ng, 7: 5 pg, 8: 0.5 pg. Lanes a– h: RT-PCR products from PBL-T cells. The amount of total RNA used for RT-PCR for each lane was the same as that used for lanes 1– 8. The position of 420 bp DNA, the expected size of CD5 cDNA, is indicated.
priori, these cells could represent either T cells that acquired B-cell antigens, or aggregates of B and T cells. It proved impossible to determine the presence of small aggregates by FSC-SSC plots. While most Raji cells had a higher FSC and SSC than activated T cells, there was some overlap between large T-cell blasts and Raji cells, making the detection of small T-B aggregates infeasible. Experiments in which Raji cells were labeled with calcein AM prior to their incubation with activated T cells excluded the possibility that Raji cells which displayed high levels of T-cell antigens were T cells that acquired B-cell antigens. When activated T lymphocytes were rendered red fluorescent by labeling with dihydroethidine prior to their incubation with calcein AM labeled Raji cells, 20 –23% of the Raji cells were found to acquire red fluorescence. It seems, therefore, that the
339
co-expression of B-cell markers and high levels of T-cell antigens, similar to those displayed by T cells, reflects the presence of small conjugates between activated T cells and Raji cells. This conclusion is supported by the fact that the sum of the percentage of CD4bright CD191 and CD8bright CD191 cells equaled that of the percentage of either CD2bright CD191 or CD5bright CD191 cells. In contrast, Raji cells expressing low levels of T-cell antigens, strikingly lower than their level on T cells, clearly constitute B cells which acquired T-cell antigens. The expression of T-cell antigens on B cells did not result from of an inductive process. The short time which was sufficient for the appearance of T-cell antigens on non-T-cell lines was too swift to set an inductive process into motion. Moreover, exposure of Raji cells to cycloheximide prior to and during their incubation with activated T lymphocytes did not decrease the expression of T-cell antigens by the Raji cells. The fact that antiCD58 mAb inhibited the appearance of CD2 antigen on non-T cells also support the conclusion that the expression of CD2 on these cells did not result from an inductive process. Raji cells which acquired dim staining with anti-CD5 mAb following incubation with activated T cells were purified by FACS sorting and the presence of CD5 mRNA was tested with a sensitive RT-PCR technique. Only a very low level of CD5 message was detected in preparations from these Raji cells, which could be attributed to the residual contamination with T cells. Thus although over 90% of these Raji cells expressed
340
FIGURE 7 Flow cytometric analysis of the CD151 population among mixtures of K562 cells and activated T cells. K562 cells were coated with anti-CD15 mAb prior to incubation for 1 h with activated T cells. The scattergrams were obtained with cells gated for the FSC and SSC characteristic of K562 cells and for CD15 positivity. Shown are results obtained either with F(ab9)2 fragments of mAb to CD2, CD4, and CD8 or with cells not exposed to mAb to any T-cell antigen (controls).
CD5 on their surface, these molecules were not synthesized by these B cells. Exposure of T cells to PHA bound to L-Agarose led to expression of T-cell markers on B cells similar to that induced by soluble PHA, thus indicating that the appearance of T-cell markers on the surface of B cells was mediated by T-cell activation rather than by PHA molecules. Stimulation of PBL with pokeweed mitogen, a T cell dependent B cell mitogen, likewise elicited the acquisition of T-cell antigens by B cells (unpublished data). The transfer of T-cell antigens to B cells seems to require not only that T cells be activated, but also that B cells will undergo activation. Thus, nonstimulated B cell failed to acquire T-cell antigens following exposure to activated T cells. The capacity of Raji cells to acquire
R. Rabinowitz et al.
T-cell antigen suggests that immortalized, neoplastic B cells constitutively express receptors enabling the transfer of T-cell markers from activated T cells. Numerous receptor-ligand pairs have been shown to mediate the interaction between T and B cells [19]. It could be envisioned that each of the T-cell markers transferred to B cells would be bound by a ligand specific for that marker. The role of CD58 for the transfer of CD2 was obvious from the pattern of receptors transferred to K562 cells. High levels of CD2, but negligible levels of CD4, CD5, and CD8 became expressed on K562 cells which were exposed to activated T lymphocytes. This was in line with the fact that cells of the K562 line possess CD58 determinants but fail to express HLA-DR, HLA-ABC, and CD72, which are the ligands of CD4, CD8, and CD5, respectively [19]. The involvement of CD58 molecules in the transfer of CD2 molecules to non-T cells was also demonstrated by the inhibitory effect of anti-CD58 mAb on the acquisition of CD2 by Raji and K562 cells. In contrast, mAbs against HLADR, HLA-ABC, and CD72 failed to inhibit the transfer of CD4, CD8, and CD5, respectively. It is possible, however, that these mAbs were not directed against epitopes involved in the receptor transfer. It was previ-
T-Cell Antigen Intercellular Transfer
ously shown that CD2 molecules enable activated T cells to bind to other cells [20, 21] via the CD58 ligand [22], and that interactions of the CD2 and CD58 molecules regulate immune reactivity [23–26]. The present study shows that the interaction between CD2 and CD58 molecules plays a role in the transfer of T-cell surface antigens from activated T cells to non-T cells, which in turn could exert an immunoregulatory function. T-cell activation results both in the release of soluble T-cell markers [27–31] and in the appearance of T-cell markers on B cells. No evidence could be found, however, in the present study for the binding of soluble T-cell antigens to the surface of B cells. Raji cells separated by a filter from T cells undergoing activation did not acquire such antigens. The present evidence supports the conclusion that following interaction of activated T cells with activated B cells, T-cell molecules are transferred to the surface of B cells. Cell membrane preparations were shown to exert a regulatory functions in the interactions among lymphoid subsets [32]. The intercellular transfer of surface proteins was demonstrated in a number of systems [6, 8, 12]. It was recently suggested that the release of GPI-linked surface molecules and their incorporation into other cells constitutes a form of intercellular communication [7]. We have presented evidence that following T-cell activation, the CD8 antigenicity detected on CD41CD81 cells may result from the transfer of CD8 molecules to CD4 cells [9]. Moreover, B lymphocytes from HIV-1-infected individuals were found to acquire the CD8 antigen in vivo [33]. Thus, the intercellular transfer of surface molecules among activated cells may constitute a common regulatory mechanism in the immune system. It is possible that intercellular receptor transfer could provide the recipient cells with the transient capacity to react with stimuli with which they would otherwise be unable to react. Alternatively, if the transferred receptors do not become effectively attached to the signaling machinery of the recipient cells, receptor transfer could be envisioned to function as a down-regulatory mechanism. Possibly stimuli that would otherwise activate lymphoid cells would react with transferred ineffective receptors, thus diluting down the effectiveness of the stimulus. Further studies are under way to discriminate between these alternatives. ACKNOWLEDGMENTS
This study was supported in part by a grants from the F. Goldhirsch Foundation and from the Paul Ehrlich Center for the Study of Normal and Leukemic WBC, and by a contribution from A. Adler. Dr. Russell Pokroy received a scholarship from the Chaim and Bracha Scholarship Fund. We are grateful to Prof. S. Meuer, Heidelberg Germany, and to the American Type Culture Collection, Rockville MD, for the supply of
341
hybridoma cells and to Prof. H. Ben-Bassat for the generous provision of leukemic cell lines. The technical help of Mrs. R. Hadar is gratefully acknowledged.
REFERENCES 1. Farcet JP, Gourdin M-F, Calvo C, Oudrihiri N, Divine M, Bouguet J, Fradelizzi D, Senik A, Reyes F: A subset of OKT41 peripheral T cells can generate colonies containing mixed progeny with OKT41 helper and OKT81 suppressor cells. Eur J Immunol 15:1067–1073, 1985. 2. Blue M-L, Daley JF, Levine H, Craig KA, Schlossman SF: Biosynthesis and surface expression of T8 by peripheral blood T41 cells in vitro. J Immunol 137:1201–1207, 1986. 3. Worman CP, Mills KHG, Cawley JC: Non T lymphoid cells can express T cell surface characteristics: Normal, HCL, and a range of leukemic B cells express T cell antigens after appropriate in vitro stimulation. Sem Oncol 11:401– 404, 1984. 4. Mills KHG, Worman CP, Cawley JC: Normal human B cells express endogenous sheep erythrocyte (E) receptor after appropriate in vitro culture. Eur J Immunol 13:379 – 382, 1983. 5. Shenker BJ, Vitale L, King C: Induction of human T cells that coexpress CD4 and CD8 by an imunomodulatory protein produced by actinobacillus actinomycetemcomitans. Cell Immunol 164:36 – 46, 1995. 6. Snapper CM: Regulation of murine B cell Thy-1 expression by IL-4, IFN-t, and CD41 T cell subsets. Cell Immunol 129:80 –94, 1990. 7. Anderson SM, Yu G, Giattina M, Miller JL: Intercellular transfer of glycosylphosphatidylinositol (GPI)-linked protein: Release and uptake of CD4-GPI from recombinant adeno-associated virus-transduced HeLa cells. Proc Natl Acad Sci USA 93:5894 –5898, 1996. 8. Lorber MI, Loken MR, Stall AM, Fitch FW: I-A antigens on cloned alloreactive murine T lymphocytes are acquired passively. J Immunol 128:2798 –2803, 1982. 9. Lalezari P, Driscoll AM: Ability of thrombocytes to acquire HLA specificity from plasma. Blood 59:167–170, 1982. 10. Blumberg N, Masel D, Mayer T, Horan P, Heal J: Removal of HLA-A,B antigens from platelets. Blood 63: 448 – 450, 1984. 11. Novotny VMJ, Doxiadis INN, van Doorn R, Brand R: The kinetics of HLA class I elution and the relevance for the use of HLA-eluted platelet transfusions. Br J Haematol 95:416 – 422, 1996. 12. Rabinowitz R, Hadar R, Schlesinger M: The appearance of the CD41CD81 phenotype on activated T cells: Possible role of antigen transfer. Human Immunol 55:1–10, 1997. 13. Rabinowitz R, Massiah E, Hadar R, Schlesinger M: In vitro activation leads to the binding of T-cell markers to
342
14.
15.
16.
17.
18.
19. 20.
21.
22.
R. Rabinowitz et al.
the surface of B-lymphocytes. Clin Immunol Immunopathol 76:148 –154, 1995. Ben–Bassat H, Weiss L, Shlomai Z, Slavin S: Reactivity of human monoclonal antibody Campath-1 with human leukemia and lymphoma cell lines of varying maturity. Leukemia Lymphoma 1:249 –255, 1990. Lozzio CB, Lozzio BB: Human chronic myelogenous leukemia cell line with positive Philadelphia chromosome. Blood 45:321–334, 1975. Kasaian MT, Ikematsu H, Casali P: Identification and analysis of a novel human surface CD5- B lymphocyte subset producing natural antibodies. J Immunol 148: 2690 –2702, 1992. Jones NH, Clabby ML, Dialynas DP, Huang HJ, Herzenberg LA, Strominger JL: Isolation of complementary DNA clones encoding the human lymphocyte glycoprotein T1/Leu-1. Nature 323:346 –349, 1986. Shubinsky G, Ben–Bassat H, Schlesinger M: The effect of IL-4 on the phenotype of a human B-cell lymphoma line (Farage) lacking immunoglobulin expression. Immunol Lett 36:37– 42, 1993. Clark EA, Ledbetter JA: How B and T cells talk to each other. Nature 367:425– 428, 1994. Schlesinger M, Levy J, Laskov R, Hadar R, Weinstock J, Ben-Bassat H, Rabinowitz R: The role of E-receptors in the attachment of thymocytes and T-lymphocytes to human target cells. Clin Immunol Immunopathol 29:349 – 358, 1983. Schlesinger M, Haubenstock HS, Bekesi JG: E-receptors mediate the attachment of activated T-lymphocytes to human red blood cells: evidence from studies with monoclonal antibodies and simple sugars. J Clin Lab Immunol 15:27–33, 1984. Springer, TA: Adhesion receptors of the immune response. Nature 346:425– 434, 1990.
23. Bierer BE, Burakoff SJ: T-lymphocyte activation: the biology and function of CD2 and CD4. Immunol Rev 111:267–294, 1989. 24. Virella G, Rugeles MT, Hyman B, La Via M, Goust JM, Frankis M, Bierer BE: The interaction of CD2 with its LFA-3 ligand expressed by autologous erythrocytes results in enhancement of B cell responses. Cell Immunol 116: 308 –319, 1988. 25. Hynes, RO: Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69:11–25, 1992. 26. Dustin ML, Springer TA: Role of lymphocyte adhesion receptors in transient interactions and cell locomotion. Annu Rev Immunol 9:27– 66, 1991. 27. Schlesinger M: Soluble lymphocyte receptors. In Kurth R, Schwerdtfeger WK (eds): Current topics in biomedical research, pp. 35–51. Berlin, Springer Verlag, 1992. 28. Rubin LA, Kurman CC, Fritz ME, Biddison WE, Boutin B, Yarchoan R, Nelson DL: Soluble interleukin 2 receptors are released from activated human lymphoid cells in vitro. J Immunol 135:3172–3177, 1985. 29. Oh SK, Lapenson D, Morgan AC: Quantification of soluble E-receptor in the serum of patients with various diseases and its accompanying immunosuppression in neoplasia. Scand J Immunol 22:51– 60, 1985. 30. Symons JA, McCulloch JF, Wood NC, Duff GW: Soluble CD4 in patients with rheumatoid arthritis and osteoarthritis. Clin Immunol Immunopathol 60:72– 82, 1991. 31. Tomkinson BE, Brown MC, Ip SH, Carrabis S, Sullivan JL: Soluble CD8 during T cell activation. J Immunol 142:2230 –2236, 1989. 32. Hodgkin PD, Kehry MR: The mechanism of T and B cell collaboration. Immunol Cell Biol 70:153–158, 1992. 33. Schlesinger M, Rabinowitz R, Levy P, Maayan S: The expression of CD8 on B-lymphocytes in HIV-infected individuals. Immunol Lett 50:23–27, 1996.
with calcein AM. In Raji cells, which were rendered CD51 following incubation with activated T cells, only a negligible level of CD5 mRNA was detected with a sensitive RT-PCR technique, probably attributable to contamination with T cells. K562 cells incubated with activated T cells acquired CD2 but not the CD4 and CD8 antigens. Exposure of either Raji or K562 cells to mAb against CD58 inhibited the transfer of CD2. The present study indicates that following their activation, T-cells gain the capacity to transfer T-cell antigens to non-T cells and that CD2 and CD58 molecules are involved in this process. Human Immunology 59, 331–342 (1998). © American Society for Histocompatibility and Immunogenetics, 1998. Published by Elsevier Science Inc.
ABBREVIATIONS FITC fluorescein isothiocyanate mAb monoclonal antibody PBL peripheral blood lymphocytes PBS phosphate-buffered saline
PE PHA RT-PCR
INTRODUCTION Activation of lymphocytes from a distinct subset can elicit the appearance of antigens that are characteristic for another subset. Thus, CD8 antigenicity can be detected on the surface of lymphocytes of the CD4 cell lineage following mitogenic stimulation of human peripheral blood lymphocytes (PBL) [1, 2]. Neoplastic B-cells cultivated for 3–5 days in the presence of T cells and From the The Hubert H. Humphrey Center for Experimental Medicine and Cancer Research, The Hebrew University-Hadassah Medical School, Jerusalem, 91120 Israel. Present address of Russell Pokroy: Department of Ophthalmology, Kaplan Medical Center, Rehovot, Israel. Address reprint requests to: Prof. Michael Schlesinger, The Hubert H. Humphrey Center for Experimental Medicine and Cancer Research, The Hebrew University—Hadassah Medical School, Jerusalem 91120, Israel; Tel: 972-2-6758334; Fax: 972-2-6414583. Received February 11, 1998; revised March 23, 1998; accepted March 23, 1998. Human Immunology 59, 331–342 (1998) © American Society for Histocompatibility and Immunogenetics, 1998 Published by Elsevier Science Inc.
phycoerythrin phytohemagglutinin reverse transcription polymerase chain reaction
mitogens acquire concomitantly the CD2 marker characteristic for T cells and the capacity to form rosettes with sheep red blood cells (SRBC) [3]. The expression of receptors for SRBC on B-cells depends on the presence of T cells in the culture [3]. Similarly, normal human B-cells become CD2-positive following culture with either autologous or allogeneic T cells and mitogen [4]. A number of studies suggested that acquisition of the CD8 antigen by CD41 cells [1, 2, 5] and of CD2 by B-cells [3, 4] may result from induction of the expression of these markers. In contrast, other studies indicated that various antigens can be transferred from one lymphocyte population to another. Thy-1 molecules become adsorbed to the surface of murine B cells when exposed in vitro to cells of Th1 clones [6]. Glycosyl-phosphatidylinositol-(GPI)-linked-CD4 showed a high level of intercellular transfer [7]. Cloned lines of alloreactive murine 0198-8859/98/$19.00 PII S0198-8859(98)00029-9
332
T lymphocytes acquired class I and class II MHC antigens from the stimulating cell population [8]. Human platelets adsorb class I HLA molecules from plasma [9] and a considerable portion of platelet HLA could be eluted [10, 11]. We have recently provided evidence that the appearance of CD41CD81 cells among activated human PBL, may result from the transfer of CD8 molecules to CD41 cells [12]. In a previous study [13], we have observed that in vitro stimulation of PBL for three days by phytohemagglutinin (PHA) led to expression of T-cell markers on human B cells. The majority of human B cells in PHA-stimulated cultures displayed CD2 on their cell surface, and a high proportion also reacted with the CD8 and CD4 monoclonal antibodies (mAbs). The use of fluorescent F(ab9)2 fragments of mAbs ensured that the staining of B cells by mAbs to T-cell markers did not result from their binding to Fc receptors. CD4 and CD8 were detected on B-cells only when T cells expressing these antigens were present in the culture. It was suggested, therefore, that following activation T-cell antigens were transferred to B cells [13]. The aim of the present study was to analyze the mechanism of the expression of T-cell antigens on B cells. Instead of exposing both T and B cells to a mitogenic stimulus, only purified human T lymphocytes were stimulated by exposure to PHA for three days. The activated T cells were incubated for one hour in the presence of a human non-T-cell lines and the capacity of activated T cells to bestow the expression of T-cell antigens on non-T-cells was determined.
MATERIALS AND METHODS Cell Lines The Raji [14] B-cell line and the K562 chronic myelogenous leukemia cell line [15] were used. The cells were grown in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine, penicillin G (100 mg/ml) and streptomycin (100 mg/ml). Isolation of Peripheral Blood T Cells Buffy coat from blood donated by volunteers was obtained from the Blood Bank of the Hadassah Hospital. Lymphocytes were isolated by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density gradient centrifugation. T cells were enriched by applying 1 3 108 mononuclear cells in a volume of 4 ml to a column containing 1.2 g combed, scrubbed nylon wool, packed to a volume of 12 ml (Robbins Scientific Corp., Sunnyvale, CA). The column was incubated for 60 min at 37°C, and the nonadherent cell fraction was eluted. The T-cell-enriched fraction contained less than 1% CD191 lymphocytes.
R. Rabinowitz et al.
Selection of Peripheral Blood B Lymphocytes Lymphocyte suspensions were incubated with MACS magnetic microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) coupled covalently to CD19 mAbs. The cell suspension was passed through a magnetized matrix. The CD19 antigen-bearing cells, which bound to magnetic microbeads, were concentrated by passage through the magnetizable matrix. Over 80% of these cells expressed CD20. Stimulation of Peripheral Blood T Cells Suspensions enriched for T lymphocytes, containing 1 3 106 cells/ml were prepared in RPMI-1640 medium supplemented with 15% FCS, 2 mM L-glutamine and antibiotics. Five ml aliquots of cell suspensions were introduced into 75 ml tissue culture flasks, and 10 mg phytohemagglutinin (PHA) (Sigma Chemical Co., St. Louis, MO) was added to each flask. The cultures were incubated at 37°C in a humidified atmosphere containing 5% CO2 and 95% air for 3 days. In some experiments T cells were exposed for 3 days to insolubilized agarose-bound PHA at a concentration of 25 mg/ml (Sigma Chemical Co.). Co-Incubation of Activated T Cells and Non-T-Cell Lines To determine if activated T cells can transfer T-cell antigens to Raji B cells, 5 3 106 activated T cells were co-incubated for one hour at 37°C with 2.5 3 106 Raji cells with intermittent agitation. In control experiments, non-activated T cells were mixed with Raji cells. Similar experiments were carried out also with mixtures of activated T cells and K562 or normal, unstimulated B cells. After incubation, the mixtures were washed twice with PBS and stained with immunofluorescent mAbs. To prevent direct cell-cell contact between Raji cells and activated T cells, in some experiments suspensions of purified T lymphocytes and Raji B cells were kept separated by a permeable membrane with a 0.4 mm pore size (Corning Hydrophilized PTEE). Membrane inserts with a 30 mm diameter were used in conjunction with 6-well tissue culture plates. The total volume of the culture medium was 5 ml. Suspensions of Raji cells (3 3 106 cells) were introduced above the filter, while T-cell suspensions (10 3 106 cells) were kept in the wells below the filter, and the cultures were exposed to PHA for 3 days. Monoclonal Antibodies The presence of T-cell antigens on various non-T-cell lines was determined with fluorescent F(ab9)2 fragments of mAbs. These were: PE anti-CD2, PE anti-CD4, PE anti-CD8, and FITC anti-CD19, purchased from Zymed Laboratories, South San Francisco, CA. Anti-CD5 mAbs
T-Cell Antigen Intercellular Transfer
FIGURE 1 Flow cytometric analysis of mixtures of calcein AM labeled Raji cells and activated T-cells. Raji cells, which were prelabeled with calcein AM, were incubated for one hour in the presence of PHA-activated T cells. The fluorescence elicited by calcein AM was recorded as fluorescence 1. The cell mixtures were exposed in addition to the following PE-labeled reagents: F(ab9)2 fragments of IgG1 (A), of anti-CD2 mAb (B), of anti-CD4 (C), of anti-CD8 (D) and PE-labeled whole antiCD5 mAb. Log10 fluorescence 2 (ordinate): staining with PE-labeled reagents. The two right hand squares constitute calcein AM-positive cells. The percentage of calcein AMpositive cells that reacted with PE-labeled reagents was: A: 1.9%, B: 94.9%, C: 62.6%, D: 34.3%, E: 62.6%.
were either PE-Leu-1, obtained from Becton-Dickinson (Mountain View, CA) or PE-CD5, obtained from Pharmingen (San Diego, CA). FITC anti-CD15 mAb was purchased from Becton-Dickinson, and FITC anti-CD20 mAb (B1) from Coulter Immunology (Hialeah, FL). Various mAbs were tested for their capacity to block the transfer of T-cell antigens. These mAbs included the following: anti-CD72 obtained from Serotec (Oxford, England), mAbs to CD11a (LFA-1 a chain, clone TS 1/22), HLA-DR (clone L243) and HLA-ABC (clone W6/ 32) which were produced by hybridomas obtained from the American Type Culture Collection (ATCC). mAbs to CD58 (LFA-3), were produced by hybridoma clone 1C3 obtained from Prof. S. Meuer.
333
Antibody Treatment of T Cells or of B-Cell Line Saturating quantities of various mAbs were added to 2.5 3 106 T cells or Raji cells suspended in 0.5 ml culture medium and the mixture was kept at 4°C for 45 min. Antibody-coated cells were washed 3 times in PBS before co-incubation with other cells. Control cells were subjected to similar treatment, the only difference being the omission of mAbs. Immunofluorescent Assays As described previously [12–13], two fluorescent mAbs (PE and FITC labeled) were added to 50 ml cell suspension containing 107 cells/ml. The mixture of cell suspension and mAbs was kept for 30 min at 4°C, and washed twice with cold PBS containing 0.1% NaN3 and 0.1% bovine serum albumin. The washed cell pellet were resuspended in 1% paraformaldehyde and kept at 4°C until they were analyzed by flow cytometry. In some experiments, either Raji cells or K562 cells were stained prior to their incubation in the presence of activated T cells. Raji cells were stained either with FITC anti-CD19 F(ab9)2 fragment or with calcein AM. Raji cells were incubated with calcein AM (Molecular Probes, Netherlands) at a final concentration of 0.01 mg/ml for 30 min at 37°C. In some experiments, activated T cells
334
R. Rabinowitz et al.
incubated for 1 hour with activated T lymphocytes, and the expression of T-cell antigens on the cells was determined as described above. Flow Cytometry Acquisition and analysis were carried out with a FACScan flow cytometer (Becton-Dickinson, Mountain View, CA) using the Facscan program. For each test, 104 cells were acquired and tested by two-color flow cytometry. PE labeled mAbs were used to detect T-cell markers expressed on non-T cells while FITC labeled markers were used to detect the target cell population. Either the whole cell population was analyzed or only the subpopulation of target cells. This was accomplished by first gating according to the typical forward light scatter and right angle light scatter of these cells. Then the FITC marker was set so as to collect over 98% of the FITC positive cells. In some experiments, cells showing either high or low staining with PE labeled mAbs were analyzed separately. The cut-off point between bright and dim cell subsets was set at 516 fluorescence intensity units. Cell sorting was performed with a FACStar flow cytometer (Becton-Dickinson). Raji cells were gated according to their typical forward and right angle light scatter. The FITC marker was set so as to collect over 98% of the CD191 cells, and the PE marker was set to collect PE negative and PE-dimly positive cells.
FIGURE 2 Flow cytometric analysis of CD191 cells among mixtures of Raji cells and activated T-cells co-incubated for 1 h. Scattergrams in left column: all the cells in the mixture were analyzed. Scattergrams in right column: only cells gated for CD19 positivity were analyzed. All of the cells were exposed to FITC labeled F(ab9)2 fragments of anti-CD19 mAb. In addition, the cells were exposed to PE F(ab9)2 fragments of IgG1 (A,a), anti-CD2 mAb (B,b), anti-CD4 (C,c), or anti-CD8 (D,d) and to PE-labeled whole anti-CD5 mAb (E,e). Log10 fluorescence 1 (abscissa): staining with FITC-labeled F(ab9)2 fragments of anti-CD19 mAb, Log10 fluorescence 2 (ordinate): staining with PE-labeled reagents.
were incubated with dihydroethidine (Molecular Probes, Netherlands) at a final concentration of 2 mg/ml for 30 min at 37°C. K562 cells were stained with FITC antiCD15 mAb. Thereafter, Raji or K562 cells were washed,
Reverse Transcription Polymerase Chain Reaction (RT-PCR) The CD5 specific primers described by Kasaian et al. [16] were used for PCR. Upstream: 59 AGG ACG GAT GGC ACA TGG TTT 39, downstream: 59 TTG TCC TGG GCC TCA TAG CT 39. The expected RT-PCR product is 420 bp, which encompasses the sequence corresponding to residues 230 – 647 of pT1-1 CD5 cDNA [17]. CD5 mRNA was reverse transcribed and the cDNA was amplified in a 50 ml volume, using the Access RT-PCR system (Promega). The amount of total RNA template ranged from 0.5 pg to 1 mg, while the final concentration of each of the primers was 2 mM. The other components included in each reaction tube were avian myeloblastosis virus (AMV) reverse transcriptase (0.1 u/ml), Thermus flavus (Tfl) DNA polymerase (0.1 u/ml), AMV/Tfl reaction buffer, MgSO4 (2 mM), and dNTP mixture (0.2 mM). For reverse transcription, the reaction mixture was incubated at 48°C for 45 min. After inactivation of the AMV reverse transcriptase at 94°C for 5 min, the CD5 cDNA was amplified by 30 cycles of denaturing (94°C, 1 min), annealing (55°C, 1 min) and extension (72°C, 2 min). Aliquots of RT-PCR products, at a volume of 20 ml were loaded to 1.5% agarose gels, containing 0.5 mg/ml ethidium bromide,
T-Cell Antigen Intercellular Transfer
FIGURE 3 Flow cytometric analysis of Raji cells and activated T cells incubated either alone or as a mixture. Raji cells and PHA-activated T cells were either kept alone or mixed and incubated for one hour. All the cells were exposed to FITC F(ab9)2 fragments of anti-CD19 mAb. In the top row, the cells were co-incubated with PE-labeled F(ab9)2 fragments of IgG1, while in the bottom row, the cells were co-incubated with PE-labeled F(ab9)2 fragments of anti-CD2 mAb. Log10 fluorescence 1 (abscissa) and Log10 fluorescence 2 (ordinate): see legend Fig. 2.
and electrophoresed. Bands of RT-PCR products were visualized in a UV transilluminator. Southern Blotting Separated DNA was transferred onto positively charged nylon membranes (Boehringer, Mannheim) and hybridized with digoxigenin (DIG) labeled oligo-probe (59TTG GAG GTG TTG TCT TCT GG-39, encompassing residues 447– 496 of the anti-sense strand of CD5 cDNA sequence [17]. The DIG-labeled hybrids were detected with anti-DIG-alkaline phosphatase conjugate and the substrates NBT (nitroblue tetrazolium salt) and BCIP (5-bromo-4-chloro-3-indoyl phosphate toluidimium), added according to the recommendations of the producer (Boehringer, Mannheim).
335
RESULTS The Appearance of T-Cell Antigens on Raji Cells The Raji B-cell line expresses CD19 and other B-cell surface markers but no T-cell antigens [18]. Within 1 h of incubation in the presence of activated human T cells, Raji cells acquired T-cell antigens on their surface. Figure 1 depicts results obtained with calcein-AM-labeled Raji cells that were incubated for 1 hour with activated T cells and were then stained with PE labeled F(ab9)2 fragments of anti-CD2, CD4, and CD8 or PE anti-CD5 mAbs. The majority of the Raji cells acquired the CD2 antigen following exposure to activated T lymphocytes. A high proportion of the Raji cells also reacted with anti-CD4, CD5 and CD8 mAbs (Fig. 1). Similar results were obtained when Raji cells incubated with activated T-cells were stained with FITC-labeled F(ab9)2 fragments of anti-CD19 mAb and PE-labeled F(ab9)2 fragments of anti-CD2, CD4, and CD8 or PE anti-CD5 mAb (Fig. 2). Raji cells by themselves were not stained by any of the mAbs against T-cell antigens, while purified Tcells were not stained by anti-CD19 mAb. Figure 3 illustrates the fact that F(ab9)2 fragments of anti-CD2 mAb stained activated T-cells but not Raji cells, while F(ab9)2 fragments of CD19 mAb stained Raji cells but
336
R. Rabinowitz et al.
TABLE 1 The expression of T-cell markers on CD191 Raji cells
T-cell antigen CD2 CD4 CD8 CD5
Proportion of CD191 expressing T-cell antigen*
Proportion of CD191 expressing high levels of T-cell antigen**
Proportion of CD191 expressing low levels of T-cell antigen**
87.63 6 2.30 59.93 6 5.50 52.03 6 6.32 72.65 6 6.53
27.83 6 1.70 19.30 6 2.62 10.53 6 1.30 30.14 6 2.30
59.80 6 8.38 40.63 6 4.89 41.50 6 4.17 42.51 6 6.37
Raji cells were incubated for one hour in the presence of T cells which had been stimulated for 3 days by exposure to PHA. The cells were stained with FITC-labelled F(ab)2 fragments of CD19 mAb and PE-labelled F(ab)2 fragments of mAb against T-cell antigens. * The proportion of cells expressing various T-cell antigens among CD191 cells (i.e., among cells in quadrants 2 and 4). ** The proportion of cells showing either high or low levels of T-cell antigens were calculated separately. Results shown are means (6 S.E.) obtained in 10 separate experiments.
failed to stain activated, purified T cells. In contrast, following incubation of T cells and Raji cells for one hour the majority of the CD191 Raji cells co-expressed CD2. Table 1 summarizes findings obtained in 10 repeated experiments in which Raji cells were exposed for one hour to activated T cells. Among CD191 cells, 87.6% co-expressed the CD2 marker. The majority of the CD191 cells also expressed the CD4, CD5, and CD8 markers, although the proportion of CD191 cells expressing these antigens was lower than the proportion of those expressing CD2. Among the Raji cells which expressed T-cell antigens following exposure to activated T lymphocytes, the majority showed dim staining with mAbs to T-cell antigens. In addition, following exposure to activated T lymphocytes, a low proportion of Raji cells showed bright staining with mAbs to T-cell antigens. This was seen regardless of whether the Raji cells were stained with CD19 mAb or calcein AM (Figs. 1–3, Table 1). The staining intensity of the Raji cells which reacted strongly with mAbs to CD2, CD4, CD5, or CD8 resembled the staining intensity of T cells with these mAbs. It seems that the cells which showed bright staining with antiT-cell mAbs represent small aggregates of activated T cells and Raji cells. This conclusion is supported by the finding that the sum of the percentage of CD4bright CD191 and CD8bright CD191 cells equaled that of the percentage of either CD2bright CD191 or CD5bright CD191 cells. Furthermore, when activated T lymphocytes were rendered red fluorescent by labeling with dihydroethidine prior to their incubation with calceinAm-labeled Raji cells, 20 –23% of the Raji cells were found to acquire red fluorescence (not shown). It should be noted that it proved impossible to determine the presence of small T-B aggregates by FSC-SSC plots. While the FSC and SSC of most Raji cells was larger than that of activated T cells, the overlap between
large T cell blasts and Raji cells prevented the detection of small aggregates. In contrast to Raji cells which stained brightly with mAbs against T cell antigens, the staining intensity of Raji cells which displayed dim staining with mAbs to T cell antigens was considerably lower than among T cells. Moreover, the sum of the proportion of CD191 cells which were CD4dim and of those that were CD8dim clearly exceeded the proportion of CD191 cells that were either CD2dim or CD5dim. These observations indicate that CD191 cells which showed dim staining with mAbs against T cells constitute B cells which acquired T-cell antigens. When normal, non-activated B cells, rather than Raji cells, were incubated with activated T cells for 1 hour, less than 6% of the B cells expressed T-cell antigens (not shown). In the experiments described above, activation of T cells was accomplished by their exposure to soluble PHA. Although activated T lymphocytes were washed thoroughly prior to their incubation with Raji cells, the possibility could be raised that the mitogen caused sticking of T-cell surface fragments to the surface of B cells. In order to rule out this possibility, T cells were activated for 3 days by exposure to insoluble PHA prior to incubation with Raji cells. The same acquisition of T-cell antigens by Raji cells was noted as with T cells activated by soluble PHA (not shown). The question arose whether Raji cells had to be in direct contact with activated T cells for T-cell antigen acquisition. In order to elucidate this issue, suspensions of T lymphocytes and Raji B cells were exposed for 3 days to PHA, while being kept separated by a permeable membrane with a 0.4 mm pore size. In repeated experiments, no significant acquisition of CD2, CD4, CD5, or CD8 by Raji cells could be detected (not shown). In order to exclude the possibility that activated T
T-Cell Antigen Intercellular Transfer
FIGURE 4 The effect of cycloheximide on the staining of Raji cells by mAbs to CD2 and CD5. Raji cells were incubated for one hour in the presence of PHA-activated T cells, either in the absence (top row A,B,C) or in the presence of cycloheximide (20 mg/ml) (bottom row a,b,c). All the cell mixtures were exposed to FITC F(ab9)2 fragments of anti-CD19 mAb. In addition, the cells were incubated with PE-labeled F(ab9)2 fragments of IgG1 (A,a), PE-labeled F(ab9)2 fragments of anti-CD2 mAb (B,b) or PE-labeled anti-CD5 mAb (C,c). The percentage of FITC-labeled, CD191 cells that reacted with PE-labeled reagents was: A: 3.7%, a: 4.1%, B: 96.9%, b: 96.5%, C: 76.0%, c: 86.5%. Log10 fluorescence 1 (abscissa) and Log10 fluorescence 2 (ordinate): see legend Fig. 2.
cells induced the expression of T-cell antigens on Raji cells, Raji cells were exposed to cycloheximide (20 mg/ ml) for 2 h prior to and during their incubation with activated T cells. This treatment had no effect on the capacity of Raji cells to acquire T-cell antigens, such as CD2 and CD5 (Fig. 4). T-Cell Antigens Appear on Raji Cells Without Synthesis of mRNA CD191 Raji cells which expressed low levels of CD5 following incubation for 1.5 hours in the presence of activated T cells, were separated by FACS sorting (Fig. 5A–C). As shown in Fig. 5C, over 90% of the isolated
337
CD191 cells displayed low levels of CD5. CD51CD192 T cells constituted 3.3% of the sorted suspension. The expression of CD5 mRNA was determined by RT-PCR with CD5 primers (Fig. 6). The 420 bp band characteristic for CD5 cDNA was seen in RT-PCR tests with RNA from Jurkat cells (lane J) and from peripheral blood T cells (lanes a– d). With decreasing concentrations of RNA from T cells, the cDNA band became weaker. A faint line could still be seen with RT-PCR product obtained with 5 ng of RNA derived from T cells (lane d). In contrast, with RNA derived from sorted, dim CD51 Raji cells, a faint band could only be seen with RT-PCR product obtained with 1 mg RNA (lane 1). Thus, it can be estimated that the concentration of CD5 mRNA in the sorted Raji cells attained about 0.5% of that found in T cells. In Southern blots with a DIG-labeled oligo-probe for CD5, the lowest concentration of RT-PCR product from T cells that still gave a faint line was obtained with 0.5 ng RNA (lane e). In contrast, the lowest concentration of CD5 specific RT-PCR product obtained with 50 ng of RNA from sorted CD51 Raji cells (lane 3). Thus, both the direct visualization of RT-PCR products and the results of Southern blotting indicated that the concentration of CD5 mRNA detected in Raji cells that ac-
338
FIGURE 5 Flow cytometric sorting of CD5dim Raji cells. Flow cytometric analysis of either Raji cells alone (A) or Raji cells which were incubated in the presence of activated T cells (B). The population of CD52 and CD5dim cells collected by FACS sorting is indicated in B, while the flow cytometric analysis of the sorted cell population is shown in C. Log10 fluorescence 1 (abscissa): staining with FITC labeled F(ab9)2 fragments of anti-CD19 mAb, Log10 fluorescence 2 (ordinate): staining with PE-labeled CD5 mAb.
quired CD5 antigenicity constituted between 0.5 to 1% of the concentration detected in T cells. This very low level of CD5 mRNA can be attributed to the low contamination of T cells present among the sorted Raji cells. The Acquisition of T-Cell Antigens by K562 Cells A number of mechanisms could underlie the transfer of T-cell antigens to the surface of B cells. It could be envisioned that each of the T-cell markers transferred to B cells would be bound by a ligand specific for that marker. Raji cells express CD58, the ligand of CD2, HLA-DR, the ligand of CD4, HLA-ABC, the ligand of CD8, and CD72, the ligand for CD5 [cf. 19]. It was, therefore, of interest to determine whether T-cell antigen transfer would occur with cells of the K562 line which express only CD58 but none of the other ligands. Incubation of K562 cells for 1 h with activated T-lymphocytes resulted in the transfer of CD2 to the majority of the K562 cells. PE labeled F(ab9)2 fragments of antiCD2 mAb stained 50.6% of the CD151 K562 cells, following their incubation with activated T cells (Fig. 7). Only a negligible percentage of K562 cells incubated with activated T cells expressed CD4, CD8 (Fig. 7) and CD5 (not shown). The Effect of Various Antibodies on T-Cell Antigen Transfer Antibodies against various ligands were tested for their capacity to inhibit the transfer of T-cell antigens from
R. Rabinowitz et al.
activated T lymphocytes to non-T-cell lines. Exposure of K562 cells to mAb against CD58, prior to their incubation with activated T cells, decreased the proportion of CD21 K562 cells from 84.6 6 2.0% to 23.6 6 2.4% (mean 6 S.E. from 6 experiments, p , .0001). Treatment of Raji cells with CD58 mAb, prior to incubation with activated T cells, significantly inhibited the proportion of CD191 co-expressing CD2 from 87.6 6 2.3% to 63.1 6 6.2% (p 5 .0016). CD58 mAb strikingly reduced the proportion of CD191CD2dim cells from 59.8 6 8.4 to 35.2 6 3.6 (p , .0001) but had no significant effect on the appearance of CD191CD2bright cells. Treatment of Raji cells with mAbs against HLAABC, HLA-DR, CD72, or CD11a had no effect on the expression of any of the T-cell antigens tested. DISCUSSION Exposure of PBL for a number of days to T-cell mitogens lead to the appearance of T-cell antigens on B cells [3, 4, 13]. Some studies suggested that the CD2 molecules detected on the surface of B cells cultured with activated T cells were produced by B cells [3, 4]. We showed previously [13] that B cells expressed either the CD4 or the CD8 antigen on their surface depending on the T-cell subpopulation with which they were co-cultured. We suggested, therefore, that T-cell marker were transferred from activated T cells to B cells with which they were in direct contact [13]. The present study showed that nonT-cell lines acquired cell surface T-cell antigens following a brief incubation in the presence of T lymphocytes which had been activated separately for three days by exposure to PHA. CD2, CD4, CD5, and CD8 antigens were expressed on B cells after their exposure for as little as one hour to activated T cells. Most Raji cells which were exposed to activated T cells expressed low levels of T-cell antigens. Some Raji cells displayed, however, high levels of T-cell antigens. A
T-Cell Antigen Intercellular Transfer
FIGURE 6 Analysis of CD5 mRNA expression. CD5dim Raji cells were obtained by FACS sorting after incubation of Raji cell in the presence of activated T cells. RNA isolated from CD5dim Raji cells was analyzed by RT-PCR, upper part, and by Southern blotting, lower part of figure. M: 100 bp DNA ladder, J: RNA from Jurkat cells as positive control, R: RNA from untreated Raji cells, as negative control. Lanes 1– 8: RT-PCR products from CD51 sorted Raji cells. The amount of total RNA used for RT-PCR for each lane was 1: 1000 ng, 2: 500 ng, 3: 50 ng, 4: 5 ng, 5: 0.5 ng , 6: 0.05 ng, 7: 5 pg, 8: 0.5 pg. Lanes a– h: RT-PCR products from PBL-T cells. The amount of total RNA used for RT-PCR for each lane was the same as that used for lanes 1– 8. The position of 420 bp DNA, the expected size of CD5 cDNA, is indicated.
priori, these cells could represent either T cells that acquired B-cell antigens, or aggregates of B and T cells. It proved impossible to determine the presence of small aggregates by FSC-SSC plots. While most Raji cells had a higher FSC and SSC than activated T cells, there was some overlap between large T-cell blasts and Raji cells, making the detection of small T-B aggregates infeasible. Experiments in which Raji cells were labeled with calcein AM prior to their incubation with activated T cells excluded the possibility that Raji cells which displayed high levels of T-cell antigens were T cells that acquired B-cell antigens. When activated T lymphocytes were rendered red fluorescent by labeling with dihydroethidine prior to their incubation with calcein AM labeled Raji cells, 20 –23% of the Raji cells were found to acquire red fluorescence. It seems, therefore, that the
339
co-expression of B-cell markers and high levels of T-cell antigens, similar to those displayed by T cells, reflects the presence of small conjugates between activated T cells and Raji cells. This conclusion is supported by the fact that the sum of the percentage of CD4bright CD191 and CD8bright CD191 cells equaled that of the percentage of either CD2bright CD191 or CD5bright CD191 cells. In contrast, Raji cells expressing low levels of T-cell antigens, strikingly lower than their level on T cells, clearly constitute B cells which acquired T-cell antigens. The expression of T-cell antigens on B cells did not result from of an inductive process. The short time which was sufficient for the appearance of T-cell antigens on non-T-cell lines was too swift to set an inductive process into motion. Moreover, exposure of Raji cells to cycloheximide prior to and during their incubation with activated T lymphocytes did not decrease the expression of T-cell antigens by the Raji cells. The fact that antiCD58 mAb inhibited the appearance of CD2 antigen on non-T cells also support the conclusion that the expression of CD2 on these cells did not result from an inductive process. Raji cells which acquired dim staining with anti-CD5 mAb following incubation with activated T cells were purified by FACS sorting and the presence of CD5 mRNA was tested with a sensitive RT-PCR technique. Only a very low level of CD5 message was detected in preparations from these Raji cells, which could be attributed to the residual contamination with T cells. Thus although over 90% of these Raji cells expressed
340
FIGURE 7 Flow cytometric analysis of the CD151 population among mixtures of K562 cells and activated T cells. K562 cells were coated with anti-CD15 mAb prior to incubation for 1 h with activated T cells. The scattergrams were obtained with cells gated for the FSC and SSC characteristic of K562 cells and for CD15 positivity. Shown are results obtained either with F(ab9)2 fragments of mAb to CD2, CD4, and CD8 or with cells not exposed to mAb to any T-cell antigen (controls).
CD5 on their surface, these molecules were not synthesized by these B cells. Exposure of T cells to PHA bound to L-Agarose led to expression of T-cell markers on B cells similar to that induced by soluble PHA, thus indicating that the appearance of T-cell markers on the surface of B cells was mediated by T-cell activation rather than by PHA molecules. Stimulation of PBL with pokeweed mitogen, a T cell dependent B cell mitogen, likewise elicited the acquisition of T-cell antigens by B cells (unpublished data). The transfer of T-cell antigens to B cells seems to require not only that T cells be activated, but also that B cells will undergo activation. Thus, nonstimulated B cell failed to acquire T-cell antigens following exposure to activated T cells. The capacity of Raji cells to acquire
R. Rabinowitz et al.
T-cell antigen suggests that immortalized, neoplastic B cells constitutively express receptors enabling the transfer of T-cell markers from activated T cells. Numerous receptor-ligand pairs have been shown to mediate the interaction between T and B cells [19]. It could be envisioned that each of the T-cell markers transferred to B cells would be bound by a ligand specific for that marker. The role of CD58 for the transfer of CD2 was obvious from the pattern of receptors transferred to K562 cells. High levels of CD2, but negligible levels of CD4, CD5, and CD8 became expressed on K562 cells which were exposed to activated T lymphocytes. This was in line with the fact that cells of the K562 line possess CD58 determinants but fail to express HLA-DR, HLA-ABC, and CD72, which are the ligands of CD4, CD8, and CD5, respectively [19]. The involvement of CD58 molecules in the transfer of CD2 molecules to non-T cells was also demonstrated by the inhibitory effect of anti-CD58 mAb on the acquisition of CD2 by Raji and K562 cells. In contrast, mAbs against HLADR, HLA-ABC, and CD72 failed to inhibit the transfer of CD4, CD8, and CD5, respectively. It is possible, however, that these mAbs were not directed against epitopes involved in the receptor transfer. It was previ-
T-Cell Antigen Intercellular Transfer
ously shown that CD2 molecules enable activated T cells to bind to other cells [20, 21] via the CD58 ligand [22], and that interactions of the CD2 and CD58 molecules regulate immune reactivity [23–26]. The present study shows that the interaction between CD2 and CD58 molecules plays a role in the transfer of T-cell surface antigens from activated T cells to non-T cells, which in turn could exert an immunoregulatory function. T-cell activation results both in the release of soluble T-cell markers [27–31] and in the appearance of T-cell markers on B cells. No evidence could be found, however, in the present study for the binding of soluble T-cell antigens to the surface of B cells. Raji cells separated by a filter from T cells undergoing activation did not acquire such antigens. The present evidence supports the conclusion that following interaction of activated T cells with activated B cells, T-cell molecules are transferred to the surface of B cells. Cell membrane preparations were shown to exert a regulatory functions in the interactions among lymphoid subsets [32]. The intercellular transfer of surface proteins was demonstrated in a number of systems [6, 8, 12]. It was recently suggested that the release of GPI-linked surface molecules and their incorporation into other cells constitutes a form of intercellular communication [7]. We have presented evidence that following T-cell activation, the CD8 antigenicity detected on CD41CD81 cells may result from the transfer of CD8 molecules to CD4 cells [9]. Moreover, B lymphocytes from HIV-1-infected individuals were found to acquire the CD8 antigen in vivo [33]. Thus, the intercellular transfer of surface molecules among activated cells may constitute a common regulatory mechanism in the immune system. It is possible that intercellular receptor transfer could provide the recipient cells with the transient capacity to react with stimuli with which they would otherwise be unable to react. Alternatively, if the transferred receptors do not become effectively attached to the signaling machinery of the recipient cells, receptor transfer could be envisioned to function as a down-regulatory mechanism. Possibly stimuli that would otherwise activate lymphoid cells would react with transferred ineffective receptors, thus diluting down the effectiveness of the stimulus. Further studies are under way to discriminate between these alternatives. ACKNOWLEDGMENTS
This study was supported in part by a grants from the F. Goldhirsch Foundation and from the Paul Ehrlich Center for the Study of Normal and Leukemic WBC, and by a contribution from A. Adler. Dr. Russell Pokroy received a scholarship from the Chaim and Bracha Scholarship Fund. We are grateful to Prof. S. Meuer, Heidelberg Germany, and to the American Type Culture Collection, Rockville MD, for the supply of
341
hybridoma cells and to Prof. H. Ben-Bassat for the generous provision of leukemic cell lines. The technical help of Mrs. R. Hadar is gratefully acknowledged.
REFERENCES 1. Farcet JP, Gourdin M-F, Calvo C, Oudrihiri N, Divine M, Bouguet J, Fradelizzi D, Senik A, Reyes F: A subset of OKT41 peripheral T cells can generate colonies containing mixed progeny with OKT41 helper and OKT81 suppressor cells. Eur J Immunol 15:1067–1073, 1985. 2. Blue M-L, Daley JF, Levine H, Craig KA, Schlossman SF: Biosynthesis and surface expression of T8 by peripheral blood T41 cells in vitro. J Immunol 137:1201–1207, 1986. 3. Worman CP, Mills KHG, Cawley JC: Non T lymphoid cells can express T cell surface characteristics: Normal, HCL, and a range of leukemic B cells express T cell antigens after appropriate in vitro stimulation. Sem Oncol 11:401– 404, 1984. 4. Mills KHG, Worman CP, Cawley JC: Normal human B cells express endogenous sheep erythrocyte (E) receptor after appropriate in vitro culture. Eur J Immunol 13:379 – 382, 1983. 5. Shenker BJ, Vitale L, King C: Induction of human T cells that coexpress CD4 and CD8 by an imunomodulatory protein produced by actinobacillus actinomycetemcomitans. Cell Immunol 164:36 – 46, 1995. 6. Snapper CM: Regulation of murine B cell Thy-1 expression by IL-4, IFN-t, and CD41 T cell subsets. Cell Immunol 129:80 –94, 1990. 7. Anderson SM, Yu G, Giattina M, Miller JL: Intercellular transfer of glycosylphosphatidylinositol (GPI)-linked protein: Release and uptake of CD4-GPI from recombinant adeno-associated virus-transduced HeLa cells. Proc Natl Acad Sci USA 93:5894 –5898, 1996. 8. Lorber MI, Loken MR, Stall AM, Fitch FW: I-A antigens on cloned alloreactive murine T lymphocytes are acquired passively. J Immunol 128:2798 –2803, 1982. 9. Lalezari P, Driscoll AM: Ability of thrombocytes to acquire HLA specificity from plasma. Blood 59:167–170, 1982. 10. Blumberg N, Masel D, Mayer T, Horan P, Heal J: Removal of HLA-A,B antigens from platelets. Blood 63: 448 – 450, 1984. 11. Novotny VMJ, Doxiadis INN, van Doorn R, Brand R: The kinetics of HLA class I elution and the relevance for the use of HLA-eluted platelet transfusions. Br J Haematol 95:416 – 422, 1996. 12. Rabinowitz R, Hadar R, Schlesinger M: The appearance of the CD41CD81 phenotype on activated T cells: Possible role of antigen transfer. Human Immunol 55:1–10, 1997. 13. Rabinowitz R, Massiah E, Hadar R, Schlesinger M: In vitro activation leads to the binding of T-cell markers to
342
14.
15.
16.
17.
18.
19. 20.
21.
22.
R. Rabinowitz et al.
the surface of B-lymphocytes. Clin Immunol Immunopathol 76:148 –154, 1995. Ben–Bassat H, Weiss L, Shlomai Z, Slavin S: Reactivity of human monoclonal antibody Campath-1 with human leukemia and lymphoma cell lines of varying maturity. Leukemia Lymphoma 1:249 –255, 1990. Lozzio CB, Lozzio BB: Human chronic myelogenous leukemia cell line with positive Philadelphia chromosome. Blood 45:321–334, 1975. Kasaian MT, Ikematsu H, Casali P: Identification and analysis of a novel human surface CD5- B lymphocyte subset producing natural antibodies. J Immunol 148: 2690 –2702, 1992. Jones NH, Clabby ML, Dialynas DP, Huang HJ, Herzenberg LA, Strominger JL: Isolation of complementary DNA clones encoding the human lymphocyte glycoprotein T1/Leu-1. Nature 323:346 –349, 1986. Shubinsky G, Ben–Bassat H, Schlesinger M: The effect of IL-4 on the phenotype of a human B-cell lymphoma line (Farage) lacking immunoglobulin expression. Immunol Lett 36:37– 42, 1993. Clark EA, Ledbetter JA: How B and T cells talk to each other. Nature 367:425– 428, 1994. Schlesinger M, Levy J, Laskov R, Hadar R, Weinstock J, Ben-Bassat H, Rabinowitz R: The role of E-receptors in the attachment of thymocytes and T-lymphocytes to human target cells. Clin Immunol Immunopathol 29:349 – 358, 1983. Schlesinger M, Haubenstock HS, Bekesi JG: E-receptors mediate the attachment of activated T-lymphocytes to human red blood cells: evidence from studies with monoclonal antibodies and simple sugars. J Clin Lab Immunol 15:27–33, 1984. Springer, TA: Adhesion receptors of the immune response. Nature 346:425– 434, 1990.
23. Bierer BE, Burakoff SJ: T-lymphocyte activation: the biology and function of CD2 and CD4. Immunol Rev 111:267–294, 1989. 24. Virella G, Rugeles MT, Hyman B, La Via M, Goust JM, Frankis M, Bierer BE: The interaction of CD2 with its LFA-3 ligand expressed by autologous erythrocytes results in enhancement of B cell responses. Cell Immunol 116: 308 –319, 1988. 25. Hynes, RO: Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69:11–25, 1992. 26. Dustin ML, Springer TA: Role of lymphocyte adhesion receptors in transient interactions and cell locomotion. Annu Rev Immunol 9:27– 66, 1991. 27. Schlesinger M: Soluble lymphocyte receptors. In Kurth R, Schwerdtfeger WK (eds): Current topics in biomedical research, pp. 35–51. Berlin, Springer Verlag, 1992. 28. Rubin LA, Kurman CC, Fritz ME, Biddison WE, Boutin B, Yarchoan R, Nelson DL: Soluble interleukin 2 receptors are released from activated human lymphoid cells in vitro. J Immunol 135:3172–3177, 1985. 29. Oh SK, Lapenson D, Morgan AC: Quantification of soluble E-receptor in the serum of patients with various diseases and its accompanying immunosuppression in neoplasia. Scand J Immunol 22:51– 60, 1985. 30. Symons JA, McCulloch JF, Wood NC, Duff GW: Soluble CD4 in patients with rheumatoid arthritis and osteoarthritis. Clin Immunol Immunopathol 60:72– 82, 1991. 31. Tomkinson BE, Brown MC, Ip SH, Carrabis S, Sullivan JL: Soluble CD8 during T cell activation. J Immunol 142:2230 –2236, 1989. 32. Hodgkin PD, Kehry MR: The mechanism of T and B cell collaboration. Immunol Cell Biol 70:153–158, 1992. 33. Schlesinger M, Rabinowitz R, Levy P, Maayan S: The expression of CD8 on B-lymphocytes in HIV-infected individuals. Immunol Lett 50:23–27, 1996.