Adherent Ia+ murine cell lines with characteristics of dendritic cells

CELLUUR

IMMUNOLOGY

80,

Adherent

349-362 (1983)

la+ Murine Cell Lines with Characteristics of Dendritic Cells

II. Characteristics of I Region-Restricted Antigen Presentation’ DONALD Department

A. COHEN*

AND ALAN

of Medical Microbiology and Immunology, College of Medicine, Lexington. Kentucky Received

April

4. 1983; accepted

M. KAPLAN University 40536-0084

of Kentucky,

June 2. 1983

The ability of an adherent Ia+, interleukin I’ (IL]) tumor cell line (P388AD) to present turkey r-globulin (TGG) to primed T lymphocytes was demonstrated and compared with normal antigen-presenting cells (APC) found in mouse spleen. P388AD tumor cells presented TGG to long-term cultures of TGG-reactive T cells (LTTC) and to lymph nodederived T cells which were enriched on nylon wool columns and subsequently depleted of endogenous antigen-presenting cells with anti-Ia antisera and complement. MHC-restricted antigen presentation by P388AD was observed when long-term cultures of TGG-reactive T cells were used as the responding Tcell population. Furthermore, antisera directed against I-region determinants expressed on the P388AD tumor ce.Usinhibited TGG-specific Tcell proliferation in a dose-related fashion, suggesting a functional role for the tumor cell-associated Ia molecules. The kinetics of antigen presentation to LTTC by P388AD wem similar to the kinetics observed for splenic APC, aUbough the magnitude of the proliferative response to LTTC to TGG was generally lower when antigen (Ag) was presented by the tumor cells compared to splenic antigen-presenting cells (APC). However, the magnitude of T-cell proliferation of immune lymph node (LN) T ceUs was comparable when Ag was presented on tumor cells or splenic APC. Several experiments suggested that Ag uptake and/or prccessing may be less effective in P388AD tumor cells as compared to normal splenic APC. A nonadherent la+, ILl- tumor cell line (P388NA), which was isolated from the same parental tumor as P388AD, was also tested for the ability to present Ag to primed T lymphocytes and Ag-reactive LITC. In contrast, to P388AD, the nonadherent tumor cell failed to present TGG under identical culture conditions even though la molecules were expressed on the tumor cells and Ag uptake had occurred. However, the defect in Ag presentation by P388NA could be corrected if an exogenous source of purified interleukio I was supplied to the cultures. A unique opportunity thus exists with both the P388AD and P388NA tumor cell lines to decipher some of the molecular interactions leading to T-cell proliferation during antigen presentation.

INTRODUCTION The ability of antigen-specific helper T lymphocytes (Th)3 to recognize and proliferate in response to antigen (Ag) has been shown to be dependent upon the capacity of ’ This work was supported in part by NIH Grants CA28308 and CA32147, NFCR Giant 31362, and by the University of Kentucky, College of Medicine, PSP Fund. * To whom correspondence should be addressed. 3 Abbreviations used: Ab, antibody: AD.2, adherent clone 2 of the P388AD tumor, Ag, antigen; APC, antigen-presenting cells; 86, C57B1/6; IL-I, interleukin 1; IL-2, interleukin 2; LN, lymph node; LTTC, 349 0008-8749/83 $3.00

350

COHEN AND KAPLAN

an antigen-presenting cell (APC) to deliver at least three identifiable signals to the Th (l-4). One signal, the immunizing Ag, appears to be presented by the APC after it has been modified by a relatively obscure event called “antigen processing” (5, 6). In addition, the Th must also interact with membrane expressed gene products of the H-21 region on the APC, namely, the Ia molecules (7, 8). Finally, the Th must receive a soluble mitogenic signal (&I) probably secreted directly by the APC (9, 10). While our understanding of antigen presentation has been greatly extended by the identification of some of the signals for the induction of Ag-specific T-cell proliferation, many questions remain to be answered. For example, we do not yet understand the association which occurs on the membrane of the APC after Ag processing between the relative nominal antigenic fragments and the Ia molecules. In particular, it is not clear whether nominal Ag and Ia molecules are (or need to be) physically associated (covalently or noncovalently) on the membrane of the APC. It also is not known whether all signals (Ag, Ia, II- 1) need to simultaneously interact with the Th or whether a sequence of interactions occurs such that one signal may induce differentiative events in the Th so that it becomes receptive to a subsequent mitogenic signal. Recently, the entire concept of Ir gene product-antigen interaction has been questioned with respect to the necessity for Ag to be presented to the T cell in the context of Ia molecules (11). Rather, it has been suggested that Ir gene products serve not to present the Ag, but rather to participate in inducing the APC to release IL1 following interaction of the Ag-specific T cell. A major problem encountered in answering these questions has been that the APC population represents a small and variable population and is very likely comprised of several cell types. The ability to achieve homogenous populations of APC in sufficient quantities for molecular studies has been hampered by the fact that APC in the peripheral organs (macrophages or dendritic cells) appear to be terminally differentiated and, as such, are relatively resistant to most cloning techniques. An alternative approach which we have utilized to derive stable clones of APC was to develop tumor cell lines which express Ag-presenting capacity. We have recently described the isolation and characterization of a series of adherent tumor cell lines (P388AD) which constitutively express Ia molecules on their surface membranes, are inducible for IL- 1 and are morphologically and functionally similar to dendritic cells of the spleen (12, 13). A related series of nonadherent tumor cell lines (P388NA), which was isolated from the same parent tumor, also expresses surface Ia molecules but differs from P388AD in many morphological and functional characteristics and is not inducible for IL-l. P388AD but not P388NA has been shown to induce syngeneic (SMLR) and allogeneic mixed-lymphocyte reactions (MLR) ( 12). Furthermore, P388AD can act as accessory cells for per&late-induced mitogcnesis and, when conjugated with TNP, can induce hapten-specific cytolytic T cells in primary in vitro cultures.4 In this report we characterize the ability of the Ia+ IGl+ P388AD to present soluble Ag in a MHC-restricted fashion to primed Th derived from immune lymph long-term cultured T cells; MHC, major histocompatibility complex; MLR, mixed-lymphocyte reaction; Mb, macrophage; NA.10, nonadherent clone 10 of the P388NA tumor; PEC, peritoneal exudate cells; SMLR, syngeneic mixed-lymphocyte reaction; TCG, turkey y-globulin; Th, helper T cells. ’ L. A. Smith, D. A. Cohen, and A. M. Kaplan, Accessory cell functions of an Ia+ adherent tumor cell line. Fed. Pm. 42,961, 1983 (abstract).

Ia+ DENDRITIC-LIKE

CELL LINES

351

node (LN) cells depleted of endogenous APC and to Ag-specific long-term T-cell cultures (LTTC). Moreover, we demonstrate that in the presence of exogenous IL-1 the Ia+, IL-l- P388NA also presents antigen. These results substantiate previous findings which demonstrate the necessity of Ir gene expression and IL1 secretion in APC ( I- 10) and provide a series of Ag-presenting cell lines which allows these signals to be clearly dissociated. MATERIALS

AND

METHODS

Mice. DBA/Z, C3H/He, and C57Bl/6 (B6) mice were purchased from either Harlan Sprague-Dawley, Inc. (Indianapolis, Ind.) or Jackson Laboratories (Bar Harbor, Maine). All mice were housed and maintained by the Division of Animal Care Services at the University of Kentucky according to guidelines determined by the American Association for Accreditation of Laboratory Animal Care. Antigens and other reagents. Turkey y-globulin (TGG) was purified from pooled turkey serum by DEAE-celIulose chromatography as previously described (14). Purified human interleukin 1 was the generous gift of Dr. Lawrence Lachman (Immunex Corporation, Seattle, Wash.) and was prepared as previously described (15). Interleukin 2 (IL-2)containing supematants were prepared from the EL4 thymoma cell line as described by Farrar ef al. ( 16). Briefly, the IL-Zproducing subline of EL-4 was cultured for 72 hr in RPM1 1640 medium (Gibco Laboratories, Grand Island, N.Y.) containing 10% fetal calf serum (Sterile Systems, Inc., Logan, Utah) and phorbol myristate acetate (Sigma Chemical Co., St. Louis, MO.) at 10 rig/ml. Cell-free culture supematants were then absorbed twice with DEAE-activated charcoal pellets (1 pellet/ml of supematant) (West Chem Products, San Diego, Calif.). The resultant supematant was found to optimally support the proliferation of the IL-2dependent cytotoxic T-cell line, CT6 ( 17), when used at a 10% concentration. Cells and cell lines. The isolation and characterization of the adherent Ia’ P388AD.2 (AD.2) and the nonadherent Ia+ P388NA. 10 (NA. 10) has been described elsewhere (12, 13). The established macrophage cell line, P388D1, was obtained as a gift from Dr. Hillel Koren (Duke University, Chapel Hill, N.C.), and the EL-4 thymoma and CT6 cell lines were a gift from Dr. John Farrar (NIH, Bethesda, Md.). All tumor cell lines were maintained in RPM1 1640 medium (G&co) containing L-glutamine (2 mM) and gentamicin sulfate (40 &ml). Medium was supplemented with 10% fetal calf serum, which contained less than 0.07 @ml of endotoxin (Sterile Systems, Inc.) for all tumor cell lines, except P388Dl which contained 20% serum. Normal antigenpresenting cells were obtained from two sources: normal syngeneic or allogeneic spleens or peritoneal exudate cells (PEC) 72 hr after stimulation with thioglycolate. Spleens were teased and pressed through sterile stainless-steel screens and were depleted of erythrocytes by lysis with Tris-buffered ammonium sulfate (0.83%). Immunization and preparation of responder T cells. Mice were injected in both hind footpads with 100 pg of TGG emulsified 1:l in complete Freund’s adjuvant (Difco Laboratories, Detroit, Mich.) in a volume of 50 ~1. One to three weeks after immunization, mice were sacrificed and the draining popliteal LN were ascepticahy collected and prepared as single-cell suspensions by pressing through stainless-steel screens. T lymphocytes were enriched by passing LN cells through nylon wool columns as described by Julius et al. (18). The T-cell-enriched preparation was subsequently depleted of endogenous APC (19) by incubating the cell pellet for 30 min on ice in

352

COHEN AND KAPLAN

a 1:5 dilution (0.5 ml) of an alloantiserum directed against the entire I region of the d haplotype (BlO.LG X A.TFR-4) anti-BlO.DZ. Note that this antiserum is also reactive against Ia specificities 7, 8, and 15 and as such could also be utilized for depletion against C3H (H-2k) and C57B1/6 (H-2b). Cells were washed after antiserum treatment and then incubated for 30 min at 37’C in 1 ml of a 1: 10 dilution of rabbit complement which was preabsorbed against mouse spleen cells (Cedarlane low-toxM, Accurate Chemical and Scientific Corp., Westbury, N.Y.). Viabilities after the Ab and complement treatments were generally greater than 90%. Preparation of antigen-presenting cells. Cells to be used as APC were suspended in polypropylene tubes at a concentration of I- 10 X lO’/ml in 5 ml of medium containing 25 @ml of mitomycin C (Sigma Chemical Co.) and incubated at 37’C for 60 min. Cells were washed by centrifugation four times and then pulsed with TGG by incubation at 37°C for 60 min in 1 ml of TGG (5 mg/ml) or as designated in the figure legends or tables. Free antigen was removed by washing the cells five times by centrimgation. Control APC were not pulsed with TGG but otherwise treated in an identical manner. T-cell proliferation assay. Preparations of responder T cells were cultured in 96well flat-bottomed microtest plates (Falcon Plastics, Becton-Dickenson and Co., Cockeysville, Md.) in RPM1 1640 medium containing 5% fetal calf serum, 2-mercaptoethanol(50 mM), Hepes buffer (10 mM), L-glutamine (2 mM), and gentamicin sulfate (40 &ml) (T-cell medium). Ag-pulsed or unpulsed APC were added to the T cells at a concentration indicated for each experiment; the final volume in each well was 0.2 ml. Plates were incubated at 37°C for 3-5 days as indicated in each experiment. Each well was pulsed with 2 &i/well of [methyl-3H]thymidine, ICN Pharmaceuticals, Inc., Irvine, Calif.) during the last 18 hr of culture. Cells were collected onto glass fiber filters with a semiautomatic cell harvester (Titertek, Flow Laboratories, Rockville, Md.) and incorporated radioactivity was determined in a liquid scintillation spectrometer (Beckman Instruments, Irvine, Calif.). Preparation of long-term cultures of antigen-spec@ic T cells. LTTC were established from TGG-immune DBA/2 and C3H mice by culturing 1 X 10’ cells from draining LNs in 2-cm* tissue culture wells (Costar, Cambridge, Mass.) in T-cell medium containing soluble TGG at 100 &ml. After 5 days, viable cells were collected by centrifugation over a Ficoll-Hypaque cushion (Pharmacia Inc., Piscataway, N.J.). Viable cells (2 X 106) were cultured with 4 X 10’ mitomycin C-treated (25 &ml, 37°C 30 min) syngeneic spleen cells in upright 25-cm* tissue culture flasks (Falcon Plastics, Becton-Dickenson) in T-cell medium containing soluble TGG (100 &ml). Viable cells were recultured every 7- 10 days in an identical manner. LTTC were not used for experiments until they had been maintained in culture for at least four passages, at which time alloreactivity was generally eliminated. Furthermore, the LTTC which were utilized in experiments were incubated 7-14 days prior to separation on FicollHypaque. In one experiment (Table 2), LTTC were collected 48 hr after culture and subsequently recultured every 72 h in T-cell medium containing 10% IL-Zcontaining supematant, but in the absence of spleen cells. This preparation of LTTC were used after four passages in IL-2. Antisera and monoclonal antibodies. The following alloantisera directed against H2-encoded surface structures were obtained from the National Institute of Allergy and Infectious Disease, National Institutes of Health, Bethesda, Maryland: anti-HZ Id (BlO.LG X A.TFR-4) anti-BlO.D2; anti-I-A.1 1, 16, (BlO X A) anti-BlO.D2; anti-

Ia+ DENDRITIC-LIKE

CELL LINES

353

I-A.2, (A.CA X BlO.HTP) anti-A.TL. Several hybridoma cell lines reactive against H-2 gene products were obtained as a generous gift of Dr. Donna King (Stanford University, Stanford, Calif.): MK-D6 (anti-I-Ad); 1O-2.16 (anti-I-Ak); 14-4-4 (anti-IEdpk); 34-l-2 (anti-H-2Kd). The isolation of these hybridomas and characterization of the monoclonal Ab have been described by the investigators who developed the hybrids (20-23). All monoclonal Abs were used as ammonium sulfate-precipitated fractions of culture supematants. RESULTS P388AD.2 Tumor Cells Can Present TGG to Primed TGG-Specljic Long- Term T-Cell Cultures

Lymph Node T Cells and to

Previous reports from this laboratory have indicated that a cloned Ia+ adherent tumor cell line of DBA/2 origin (P388AD.2) induced syngeneic and allogeneic mixedlymphocyte reactions in unprimed spleen cells (12) and preliminary experiments suggested that this cell line presented soluble Ag to TGG-primed LN T cells (13). In contrast, the cloned Ia+ nonadherent tumor line P388NA.10 did not induce T-cell proliferation under identical conditions. The current experiments were undertaken to determine the Ag-presenting characteristics of the P388AD cell lines and to identify the lesion(s) in the P388NA cells which prevented these Ia+ cell lines from presenting Ag. We have previously shown that under conditions where endogenous APC were removed from TGG-immune LN T-cell preparations with anti-Id and complement (as evidenced by the failure of T cells to proliferate in response to soluble TGG), Agpulsed AD.2 tumor cells presented Ag better than APC derived from normal DBA/ 2 spleen (13). T-Cell proliferation induced by AD-2 was Ag specific in that AD.2 would present turkey but not human y-globulin to TGG-primed T cells (data not presented). In contrast, the related Ia+ NA.10 failed to present Ag to the primed T cells (13). For a tumor cell to be used as a model for Ag presentation it should be MHC restricted as are normal APC. In order to determine if AD.2 presented Ag in an MHC-restricted manner, Ag-pulsed or unpulsed AD.2 cells were cultured with TGGprimed T cells derived from either syngeneic DBA/2 or allogeneic B6 LN. As seen in Table 1, AD.2 presented Ag to both syngeneic and allogeneic LN T cells even after endogenous APC were depleted (Table 1). Under identical culture conditions, Ag presentation by PEC was H-2 restricted. Since pretreatment of T cells with antiId and complement eliminated proliferation to soluble Ag alone, it was unlikely that endogenous Ia+ APC remained in the T-cell preparation. However, proliferating T cells have been shown to release a factor which can induce Ia expression in Ia- M+s and consequently convert them into APC (24-26). Since AD.2 has been shown to induce MLR in allogenic T cells (12), it was felt that this might lead to conversion of Ia- M@s which remained in the T-cell preparation after depletion of Iaf M+s. As such, this would mean that the apparent lack of MHC restriction by AD.2 might be an artifact of the culture conditions. Therefore, to determine the Ag-presenting capacity of the AD.2 and NA. 10 clones in an environment free from any potential APC, we next determined if AD.2 would present Ag to Ag-specific LTTC. TGG-specific LTTC were established from DBA/2 and C3H mice. Evidence that AD.2 effectively presented Ag to LTTC is presented in Fig. 1. Proliferation of DBA/2 LTTC was directly related to the number of AD.2

354

COHEN AND KAPLAN TABLE

1

P388AD.2 Presents Antigen in an Unrestricted Manner to Primed Lymph Node T Cells T-Cell proliferation (cpm + SE) b

untreated

anti-Ia + C

Ag treatment

DBA/Z T cells

B6 T cells

DBA/2 T cells

B6 T cells

AD.2

Pulsed Unpulsed

63,877 k 5200 4,616 + 967 (59,261)

83,110 I? 4213 40,146 f 5314 (42,964)

31,716 + 747 3,686 + 358 (28,030)

71,343 + 9342 41,925 I? 8132 (29,416)

DBA/2

Pulsed Unpulsed

29,726 f 5300 1,606 + 171 (28,120)

14,921 -+ 603 13,122 + 706 (1,799)

11,240 + 534 570 + 78 (10,670)

2,146 + 275 12,965 + 4019 (0)

B6

Pulsed Unpulsed

24,254 +- 3201 3,981 + 275 (20,273)

19,293 k 1689 1,235 + 63 (18,058)

1,837 zk 137 2,267 + 21 (0)

12,667 f 926 434 rt_ 57 (12,233)

APC’

’ P388AD.2 or 48-hr thioglycolate-elicited peritoneal exudate cells were mitomycin C treated and then incubated in turkey -r-globulin (TGG) for 60 min at 37°C at 5 mg/ml. Cells were then washed five times to remove soluble TGG. b Nylon wool-enriched T cells from TGG-primed popliteal lymph nodes (DBA/2 or C57Bl/6) were treated with anti-Iad antiserum (BlO.LG X TPR-4) anti-BlO.D2 for 30 min at 4°C and subsequently with rabbit complement (1: 10) for 30 min at 37’C. Removal of endogenous antigen-presenting cells was indicated by the fact that only anti-Ia plus complement treatment completely eliminated T-cell response to soluble TGG. T cells (3 X 10’) were cultured for 5 days with antigen-presenting cells (3 X 10’). Cultures were pulsed with 2 &i/well of [‘H]thymidine during the last 18 hr of culture. Values in parentheses represent the Acpm, which is the difference in incorporation between cultures containing TGG-pulsed or unpulsed antigen-presenting cells.

tumor cells in the culture and only seen when AD.2 was pulsed with the immunizing Ag. Demonstrable Ag presentation was Seen by AD.2 with as few as 2.5 X lo4 tumor cells. Based on the assumption that spleen contains approximately 5- 10% macrophages, it appeared that splenic APC were more efficient on a per cell basis than AD.2 in that the peak response with AD.2 was always substantially lower than the peak response with splenic APC. When Ag-pulsed AD.2 tumor cells were cultured with syngeneic DBA/2 or allogeneic C3H LTTC, a clear pattern of MHC restriction was apparent (Table 2) in that AD.2 only presented Ag to syngeneic DBA/2 LTTC. Under identical conditions, NA. 10 tumor cells were ineffective APC when LTTC were used as responders and LTTC also failed to proliferate in response to soluble TGG alone. Characteristics of Antigen Presentation Long-Term T-Cell Cultures

by P388AD.2

Tumor Cells to TGG-Specific

In addition to the similarity between the P388AD clones and the splenic APC with regard to MHC restriction, we felt that other characteristics of their Ag-presenting capabilities should be examined to establish the similarities and/or differences between AD.2 and normal APC. Therefore, several studies were done to characterize Ag presentation by AD.2 and to compare AD.2 as APC with normal splenic APC. The kinetics of Ag presentation by AD.2 tumor cells to TGG-specific DBA/2 or C3H

Ia+ DENDRITIC-LIKE

Number

of APC

per

well

CELL LINES

x 104(AD)

or

355

x IO5 (spleen)

FIG. 1. Presentation of TGG to antigen-reactive long-term T-cell cultures by P388AD.2 tumor cells. TN&reactive long-term cultured DBA/2 T cells (5 X 10’) were incubated at 37“C for 72 hr with the indicated numbers of mitomycin C-treated P388AD.2 tumor cells (solid circles) or DBA/Z spleen cells (open circles). The antigen-presenting cells were previously pulsed with 5 mg of TGG (solid tine) or left unpulsed (dashed line). Cultures were pulsed during the last 18 hr of incubation with 2 @/well of [3H]thymidine. Incorporation by T cells alone in the presence of TCG (100 &ml) was 536 cpm.

LTTC was compared with normal spleen cells and the peak Ag-specific proliferative response was found to occur on Day 3 for both tumor cells and spleen cells (unpublished observations). In addition, MHC-restricted presentation by AD.2 to LTTC was an absolute phenomenon in that AD.2 did not present Ag to C3H LTTC at any time during the culture period. As previously mentioned, Ag presentation by AD.2 generally was not as efficient as splenic APC with LTTC as responders. One possibility for this difference was that Ag retention by AD.2 during the pulsing period was less than for splenic APC. If this were so, then Ag may be limiting in the cultures containing AD.2. When experiments were performed with soluble TGG in the cultures containing LTTC and APC rather than with TGG-pulsed APC, so that Ag concentration was not limiting, presentation by AD.2 was seen to be linearly related to the concentration of Ag in the culture (Fig. 2). As previously seen, presentation by AD.2 was less efficient than by splenic APC even though the concentration of TGG required to achieve optimal T-cell proliferation was comparable for both splenic and tumor APC. We next determined whether AD.2 and DBA/2 splenic APC required similar Ag concentrations during the pulsing period. The data in Fig. 3 demonstrate that AD.2 required a high concentration (5000 &ml) of TGG during the pulsing period for maximum Ag presentation to LTTC. In contrast, maximum Ag-presentation activity could be demonstrated with splenic APC pulsed with a log less Ag (500 &ml) and activity comparable to the maximum Ag presentation of the AD.2 cells occurred with splenic APC pulsed with 50 Ccg/ml.

356

COHEN AND KAPLAN TABLE 2 Antigen Presentation to Long-Term T-Cell Cultures by P388AD.2 Is MHC-Restricted T-Cell proliferation (cpm + SE)’ T cell”

APCb

TCG pulsed

Unpulsed

DBA/Z

AD/2 NA.10 DBA/Z spleen C3H spleen

19,513 4,051 12,155 7,616

* + + 2

1403 877 2136 691

1,448 625 12,169 7,150

r 1594 k 56 f 508 iz 790

12,065 3,426 59,986 0

C3H

AD.2 NA.10 DBA/2 spleen C3H spleen

6,159 1,863 25,915 63,086

k 654 + 525 + 2262 zk 2691

6,827 2,019 42,758 9,459

+ 1077 + 637 + 1295 + 152

0 0 0 53,627

‘Cells from draining lymph nodes of mice immunized with TGG were passed three times at weekly intervals in the presence of syngeneic, mitomycin C-treated spleen cells plus soluble TGG (100 &ml). Forty-eight hours after the third passage, T-cell blasts were cultured in medium containing 10% IL-2containing supernatant in the absence of any antigen-presenting cells. LTTC were washed, split 1:2, and fed fresh medium and IL2 every 3 days. After four passages in B-2, viable cells were isolated and tested for antigen-specific proliferation at a concentration of 5 X 104 LTTC per microwell. b Antigen-presenting cells were prepared by incubating 5 X lo6 tumor cells or 2 X 10’ spleen cells in mitomycin C for 60 min, at 37”C, followed by 60 min at 37°C in the presence or absence of TGG (5 mg/ ml). APC were washed five times prior to culture with LTTC at a concentration of 1 X 10’ tumor cells or 1 X 106 spleen cells per microwell. c Cultures were incubated 4 days at 37°C and were pulsed for the final 18 hr with [‘Hlthymidine (2 pCi/ well). Incorporation by LTTC alone was 2403 k 47 1 for DBA/2 and 823 + 325 for C3H. No response to soluble TGG was detected in LTTC in the absence of APC. Incorporation by APC alone was
In order to ascertain whether the expression of Ia molecules on AD.2 was functionally involved in presentation of TGG to LTTC, alloantisera directed against H-2 specificities I-A. 11, 16 (H-2d specific) and I-A.2 (H-2” specific) were added to LTTC containing TGG-pulsed AD.2. As seen in Fig. 4, only antisera directed against I-A specificities expressed on AD-2 (i.e., I-A. 11, 16) were effective in blocking Ag-specific proliferation in a dose-related manner. Blocking studies have also been performed with monoclonal Abs against I-A and I-E subregions (Table 3) and the data clearly indicate that the A subregion on AD.2 participated in presentation of TGG to L’MC. Results with monoclonal Ab directed against the E subregion remain equivocal and it is not yet clear whether the marginal inhibition of T-cell proliferation was because gene products of the E subregion are not involved in presentation of TGG or because I-E molecules appear to be expressed in low levels on AD.2 tumor cells (data not presented). Another feature of AD.2 tumor cells which may be involved in Ag presentation, is the capacity of this tumor cell to be induced to synthesize IL-1 (13). We have previously determined with the help of Dr. Stephanie Vogel (Uniformed Services University, Bethesda, Md.) that all clones of P388AD could be induced to secrete IL 1 following stimulation by lipopolysaccharide ( 13). In contrast, none of the clones of P388NA secreted IL-1 under identical stimulation. Since the role of IL-1 in Ag presentation has been established for normal APC, we sought to determine if IL1 was necessary for Ag presentation by these Ia+ tumor cell lines. In order to determine if IL-1 was necessary for these Ia+ tumor cells to act as APC, we utilized the fact that

Ia+ DENDRITIC-LIKE

TGG (Jg /ml

CELL LINES

357

1

FIG.2. Presentation of soluble TGG by P388AD.2. TGG-reactive, long-term cultured DBA/Z or C3H T cells (5 X 104) were incubated for 72 hr with 2 X 10’ mitomycin C-treated P388AD.2 tumor cells or spleen cells in the presence of soluble TGC at the indicated concentrations. Plates were pulsed during the last 18 hr of culture with 2 j&i/well of [3H]thymidine. Incorporation by T cells alone in the presence of soluble TGG (100 &ml) was less than 2000 cpm. Plotted values represent the difference in incorporation between cultures containing TCGpu1se.d and unpulsed antigen-presenting cells. 0 0, P388AD.2 with DBA/Z T cells; l - - - 0, P388AD.2 with C3H T cells; n n , DBA/Z spleen with DBA/Z T cells; A - - - A, C3H spleen with C3H T cells.

the Ia+ NA. 10 tumor cell is not inducible for IL- 1 secretion and that the related Ma tumor, P388D1, although inducible for IL-l, does not express Ia molecules constitutively (27). When these three tumor lines as well as DBA/2 and C3H spleen cells were utilized as APC with DBA/2 LTTC only AD.2 and DBA/2 spleen could present Ag (Table 4). However, when purified human IL-1 was added to the cultures, NA.10 also presented Ag, whereas P388Dl M@s did not. These data suggested that both Ia expression and IL-l were necessary for these tumor cells to present Ag to LTTC and that in contrast to P388AD tumor cells, P388NA failed to present Ag because of a lesion in the cell which prevented the synthesis and/or secretion of IL- 1 as a mitogenic signal. It should be noted that the addition of IL-l to cultures containing T cells and unpulsed AD.2 tumor cells led to nearly a 12-fold increase in proliferation. We have routinely observed this and are not certain whether this is related to a deficiency in IL-l induction relative to splenic APC during interaction with T cells. However, previous studies have determined that AD.2 produces less IL-l than the P388Dl M+ line following lipopolysaccharide stimulation (13). DISCUSSION The results presented in this report clearly indicate that the cloned adherent Ia+ tumor cell line, P388AD.2 can present soluble antigen to primed T lymphocytes in

358

COHEN AND KAPLAN

-

-7 5o I II 40A z - 30-

150

-120 s5 :: 2 - 90 e

N-l 0 ;

-

% -

20-

60

: z

6 -

IO-

0

ANTIGEN

5 50 CONCENTRATION

500

30

5000 (ughnl)

FTC+.3. Effect of TGG concentrations during antigen-pulse period on antigen presentation by P388AD.2 tumor cells. Mitomycin C-treated P388AD.2 (0 - - - l ) (2 X 106) or mitomycin C-treated DBA/2 spleen (2 X 10’) cells (M) were pulsed for 60 min at 37’C with the indicated concentrations of TGG. P388AD.2 (5 X 10’) or spleen (5 X 10s) cells were subsequently cultured for 72 hr with 5 X 104 TGGreactive, long-term cultured DBA/Z T cells. Plates were pulsed during the last 18 hr of culture with 2 &i/ well of [3H]thymidine. Incorporation by T cells in the presence of soluble TGG (100 &ml) was 804 cpm.

the absence of endogenous APC. Proliferation of the primed T cells was shown to be Ag-specifk and directly related to the Ag concentration and number of Ag-presenting tumor cells in the culture. It is widely accepted that APC must bind and process soluble Ag in an as yet undefined manner in order to present the Ag to an Ag-reactive T cell (5, 6). Our studies indicate that AD.2 tumor cells could present Ag effectively when either Ag pulsed or when Ag was continually present in the culture, suggesting Oc l-A.2

NONE

I:320 DILUTION

I:160

I:60

130

OF ANTISERUM

I%. 4. Blocking of antigen presentation by P388AD.2 with anti-Ia antisera. TGG-reactive LTTC (5 X 104) were cultured for 4 days with TGG-pulsed P388AD.2 tumor cells (1 X 104 (mitomycin C treated) in the presence of the indicated dilution of anti-I-A. 11, 16 (0 0) or anti-I-A.2 (0 - - - 0) alloantisera. Cultures were pulsed with [3H]thymidine during the last 18 hr of culture. Incorporation by LTTC alone was 175 + 275 cpm.

Ia+ DENDRITIC-LIKE

359

CELL LINES

TABLE 3 Inhibition of Antigen Presentation by P388AD.2 with Monoclonal Antibodies (mAb) against I Region Determinants T-Cell proliferation (cpm)” Expt No. 1

2

mAb Added

Ag-pulsed AD.2

Unpulsed AD.2

Acpm

17,560

15,569 5,196 8,286 12,104 13,343

67 47 22 14

8,955 0 7,107 5,880

100 21 34

None I-Ad I-E*’ I-Ak K-Dd

9,036 14,194 16,316

1,992 804 750 2,090 2,913

None I-Ad I-Edk 1-A’

12,611 4,931 11,172 10,005

3,656 5,195 4,065 4,125

ho00

% Inhibition

0 Long-term TGG-specific DBA/Z T cells (5 X 104) were cultured for 72 hr with 1 X lo5 antigen-pulsed or unpulscd AD.2 tumor cells (mitomycin C treated). Hybridoma culture supematants were added at a final dilution of 1:4. Cultures were pulsed with 2 pCi/well of [‘Hlthymidine during the last 18 hr of culture, T cells failed to respond to soluble TGG in the absence of antigen-presenting cells. Incorporation by T cells alone was 400 cpm. Percentage inhibition was determined by the formula { [(cpm without mAb) (cpm with mAb)]/(cpm without mAb)} X 100.

that both the initial binding and processing events occurred in AD.2 tumor cells. The ability of AD.2 to present Ag in an MHC-restricted fashion was consistently demonstrated when using Ag-reactive LTTC as the responding T-cell population. However, when T cells were obtained directly from primed LN, unrestricted presentation of Ag was frequently, though not always, seen even after endogenous APC were first removed with anti-Ia and complement. Schwartz et al. ( 19) have demonstrated that Ag-pulsed tumor cells can shed surface-bound Ag which can then be presented by endogenous APC which contaminate the T-cell preparation. Our data support the necessity of removing Ia+ M@s from the T-cell preparation when assaying Ag-pulsed tumor cells for Ag presentation. However, since the T-cell preparations which we used in our experiments were treated with anti-Ia and complement and failed to respond to soluble Ag at several concentrations, we are uncertain why allogenic T cells taken directly from LNs proliferate in response to Ag-pulsed AD.2. Although endogenous APC were depleted by this procedure, not all of the contaminating M+s (Ia-) would be removed. Since several reports (24-26) have demonstrated that proliferating T cells release a factor, which is capable of inducing Ia expression in IaM9s and because AD.2 stimulates a very strong MLR ( 12), it is possible that conversion of Ia- Ma into functional APC may have occurred in the allogeneic T-cell cultures. If true, this could lead to an erroneous conclusion that Ag presentation by AD.2 is unrestricted. Clearly, AD.2 tumor cells presented Ag in an MHC-restricted fashion when LTTC were used rather than primary LN T cells. The Ag-reactive LTTC which we used throughout these studies had little or no alloreactivity to AD.2 tumor cells and as such lymphokine production in these cultures would be minimal. In addition, the LTTC were routinely used for experiments after 7- 14 days of culture, and although LTTC were passaged in the presence of mitomycin C-treated spleen cells, our experience was that the number of viable M+s in these cultures after centrifugation on Ficoll-

360

COHEN AND KAPLAN TABLE 4 Effect of Exogenous IL1 on Antigen Presentation to Long-Term T-Cell Cultures by P388Derived Tumor Cells T-Cell proliferation (cpm + SE)’ +1L-1

No addition APC type

TGG-pulsed

Unpulsed

AD.2 (Ia+, ILl+)

22,686 (+2,084)

665 (-1275)

NA.10 (Ia+, IL-l-)

723 (+85)

P388Dl (Ia-, IL-l+)

Acpm

TGG-pulsed

Unpulsed

Acpm

22,02 1

54,414 (?1750)

1,884 (+766)

46,530

(Z)

123

12,673 (fl,l27)

1,879 (k533)

10,794

2,388 (+248)

1473 (k458)

915

1,897 (2106)

2,723 (+1,228)

0

DBA spleen

114,260 (?7,404)

5872 (+968)

108,388

161,976 (+6,854)

11,514 (+787)

150,462

C3H spleen

1,691 (?170)

2516 (&222)

0

3,082 (+388)

2,623 (+392)

459

’ Long-term DBA/Z T cells (5 X 104) (TCC reactive) were cultured with mitomycin C-treated tumor cells (1 X 10’) or spleen cells (1 X 106) for 4 days with or without purified human interleukin 1 (diluted 1:100). As indicated, the antigen-presenting cells were pulsed with turkey y-globulin (5 mg/ml) for 60 min, at 37”C, and washed extensively or left unpulsed. Cultures were pulsed with [3H]thymidine (2 &i/well) during the last 18 hr of culture. T cells alone fIL1 failed to proliferate in the presence of an optimal amount of soluble TGG (100 &well).

Hypaque was insufficient to permit T-cell responsiveness to soluble Ag. It is unlikely that contaminating McPs participated in the MHC-restricted presentation to DBA-2 LTTC, since Ag presentation by AD.2 was also demonstrated under conditions where DBA/2 LTTC were first cultured through four passages in IG2-containing supematants in the absence of any splenic filler cells (Table 2). A consistent finding throughout these studies was that AD-2 tumor cells generally induced less Ag-specific T-cell proliferation than normal spleen APC when LTTC were used as responding cells. This was evident even in experiments where the number of spleen cells was adjusted so that roughly equal numbers of APC (M@s or tumor cells) were in the cultures. The difference in Ag-presenting capacity was not due to a kinetic shift since similar kinetics were observed for both AD.2 and spleen cells. Likewise, under conditions where Ag was not limiting in the culture, i.e., when soluble Ag was present, AD.2 tumor cells were still less effective in Ag presentation than spleen. Studies by McKean et. al. (28) Walker et al. (29), and Birmingham et al. (30) have also noted less efficient Ag-presenting capacity by B-lymphocyte tumor cells and the monocytic tumor cell WEHIwhen presenting to either Ag-reactive LTTC or T-cell hybridomas. However, Glimcher et al. found that Ag presentation by the A20.3 B-cell tumor was clearly more efficient than splenic APC (3 1). Several possibilities exist at the level of the APC tumor or the responding T cell which could result in the decrease in Ag-presenting efficiency by AD.2 tumor cells. Although preliminary studies with iodinated TGG indicated that Ag uptake did occur in AD.2 (unpublished observations), how cell-bound Ag is processed and/or associated with surface Ia in

Ia+ DENDRITIC-LIKE

CELL LINES

361

an immunostimulatory form is unknown. Clearly, AD.2 must be pulsed with substantially greater amounts of TGG in order to achieve equivalent levels of T-cell proliferation. In this regard, if the role of Ag processing is to bring nominal antigenic fragments into close proximity with the Ia molecules on the APC, this may indicate that AD.2 does not process Ag and that high concentrations of Ag are merely required to “randomly” associate the Ag with surface Ia molecules. Alternatively, the events of Ag processing in AD.2 may be quite different than those occurring in splenic APC, such that the nominal antigenic determinants (5) expressed on AD.2 are only crossreactive with respect to those antigenic fragments on splenic APC. Finally, AD.2 tumor cells may release less IL-1 during Ag presentation than splenic APC. Any or all of these factors operating at the level of the APC could affect the level of T-cell proliferation. On the side of the responding T cell, less efficient Ag presentation by AD.2 tumor cells when using LTTC rather than primed LN T cells could be accounted for if the T-cell repertoire differs in those T-cell populations. Individual Ag-reactive clones which respond better to Ag associated with Ia on AD.2 than with Ia on splenic APC may exist in T-cell preparations derived directly from primed LNs. When LTTC were established by repetitive culturing with splenic hller cells, these clones would be lost and only clones which preferentially recognized Ag on splenic Ia would be selected. This difference could be due either to a difference in the fine structure of Ia on AD.2 or to an alteration in the manner in which Ia and Ag are associated on the surface of AD.2. Studies with LTTC established in the presence of AD.2 as APC rather than with normal splenic APC should resolve this question. Although Ag presentation to LTIC by AD.2 may be less effective than splenic APC, the mechanisms involved in triggering T-cell proliferation appeared to be similar. Ia molecules were shown to be functionally involved in presentation by AD.2 since alloantisera as well as monoclonal Abs against I region determinants inhibited induction of T-cell proliferation. However, it is clear that Ia alone was insufficient since the nonadherent Ia+ tumor, NA.lO, which bound similar amounts of Ag (unpublished observation), was unable to present the Ag to LTTC unless an exogenous source of purified IL-l was included. Since the Ia- M+ tumor, P388D1, was unable to present Ag even in the presence of IL- 1, it appeared that Ag presentation by these tumor cell lines required both Ia expression and IL-l synthesis in a manner similar to that described for normal APC (7-10). The requirements for Ag presentation by AD.2 were similar to those reported for presentation by the B-cell tumor A20.3 (3 1) in that Ag presentation required Ia expression and was independent of any exogeneous source of IL- 1. In contrast to the Iaf B-cell tumors, which did not present Ag, NA. 10 appeared only to lack the ability to synthesize IL-l, since presentation did occur when IL-l was added to the cultures. These Ia+ tumor cell lines, P388AD.2 and P388NA.10, should, therefore, be very useful tools to dissect the molecular and temporal events in Ag presentation, including processing, Ia association, and triggering of proliferation in Ag-specific T cells. ACKNOWLEDGMENTS The authors wish to extend their gratitude to Dr. Lawrence Lachman and the Immunex Corporation for graciously supplying purified human IL-1 for these studies. In addition, we wish to thank Dr. Donna King for supplying us with the B-cell hybridomas, Dr. John Farrar for sending the EL-4 thymoma and CT6 cytotoxic T-cell line, and Dr. Hillel Koren for sending the P388Dl macrophage cell line. Finally, we thank Ms. Kathy Rogers for invaluable secretarial assistance during the preparation of this manuscript.

362

COHEN AND KAPLAN

REFERENCES I. 2. 3. 4. 5. 6. 7. 8. 9.

10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Benacerraf, B., and Germain, R. N., Immunol. Rev. 38, 70, 1978. Schwartz, R. H., Yano, A., and Paul, W. E., Immunol. Rev. 40, 153, 1978. Unanue, E. R., Immunol. Rev. 40,227, 1978. Unanue, E. R., Advan. Immunol. 31, I, 1981. Thomas, D. W., and Shevach, E. M., J. Immunol. 121, 1145, 1978. Ziegler, K., and Unanue, E. R., J. Immunol. 127, 1869, 198 1. Schwartz, R. H., David. C. S., Sachs, D. H., and Paul, W. E., J. Immunol. 117, 531, 1976. Thomas, D. W., and Schevach, E. M., J. Exp. Med. 144, 1263, 1976. Germain, R. N., J. Immunol. 127, 1964, 1981. Larsson, E. L., Iscove, N. N., and Coutinho, A., Nature (London) 283, 664, 1980. Durum, S. K., and Gershon, R. K., Proc. Natl. Acad Sci. USA 79,4747, 1982. Cohen, D. A., and Kaplan, A. M., J. Exp. Med. 154, 1881, 1981. Cohen, D. A., Smith, L. A., and Kaplan, A. M., In “Macrophages and Natural Killer Cells” (S. J. Norman and E. Sorken, Eds.), pp. 549-556. Plenum, New York, 1982. Chiller, J. M., Habicht, G. S., and Weigle, W. O., Proc. Natl. Acad. Sci. USA 65, 551, 1970. Lachman, L. B., Page, S. O., and Metzger, R. S., J. Supramol. Struct. 13, 457, 1980. Farrar, J. J., Fuller-Farrar, J., Simon, P. J., Hilfiker, M. L. Stadler, B. M., and Farrar, W. L., J. Immunol. 125,2555, 1980. Hihiker, M., Moore, R., and Farrar, J., J. Immunol. 127, 1983, 1981. Julius, M. F., Simpson, E., and Het-zenberg, L. A., Eur. J. Immunol. 3, 645, 1973. Schwartz, R. H., Kim, K. J., Asofsky, R., and Paul, W. E., in “Macrophage Regulation of Immunity” (E. R. Unanue and A. S. Rosenthal, Ed.%), pp. 277-280. Academic Press, New York, 1980. Kappler, J. W., Skidmore, B., White, J., and Marrack, P., J. Exp. Med. 153, 1198, 1981. Ozato, K., Mayer, N. M., and Sachs, D. H., J. Immunol. 124, 533, 1980. Ozato, K., Mayer, N. M., and Sachs, D. H. Transplantation 34, 113, 1982. Oi, V. T., Jones, P. P., Goding, J. W., Herzenberg, L. A., and Herzenbetg. L. A., Curr. Top. Microbial. Immunol. 81, 142, 1978. Steeg, P. S., Moore, R. N., and Oppenheim, J. J., J. Exp. Med. 152, 1734, 1980. Scher, M. G., Beller, D. I., and Unanue, E. R., J. Exp. Med. 152, 1684, 1980. Steinman, R. M., Nogueira, N.. Witmer, M. D., Tydines, J. D., and Mellman, 1. S., J. Exp. Med. 152, 1248, 1980. Ralph, P., Lymphokines 4, 175, 198 I. McKean, D. J., Infante, A. J., Nilson, A., Kimoto, M., Fathman, C. G.. Walker, E., and Warner, N., J. Exp. Med. 154, 1419, 1981. Walker, E. B., Lanier, L. L., and Warner, N. L., J. Exp. Med. 155, 629, 1982. Birmingham, J. R., Chestnut, R. W., Kappler, J. W., Marrack, P., Kubo, R., and Grey, H. M., J. Immunol. 128, 1491, 1982. Glimcher, L. H., Kim, K. J., Green, I., and Paul, W. E., J. Exp. Med. 155, 445, 1982.