4-1BB Is Expressed on CD45RAhiROhiTransitional T Cell in Humans

CELLULAR IMMUNOLOGY ARTICLE NO.

169, 91–98 (1996)

0095

4-1BB Is Expressed on CD45RAhiROhi Transitional T Cell in Humans BETH ANN GARNI-WAGNER, ZANG H. LEE, YOUNG-JUNE KIM, CHARLES WILDE, CHANG-Y KANG,* AND BYOUNG S. KWON1 Department of Microbiology and Immunology and the Walther Oncology Center, Indiana University School of Medicine, Indianapolis, Indiana 46202-5120; and *Seoul National University College of Parmacy, Seoul, Korea Received August 17, 1995; accepted December 14, 1995

encounter peptide from the invading pathogen in context with self-MHC via the T cell’s own T cell receptor (TCR)2 complex and other costimulatory molecule(s), such as CD28 (1, 2). Without the engagement of the other costimulatory molecule(s) the T cell is rendered anergic (3). To date the best characterized costimulatory molecule has been CD28 (1, 2). More recently, however, other cell surface molecules have been suggested to play a costimulatory role, such as the molecule 4-1BB. 4-1BB, an Ç55-kDa homodimeric molecule expressed on activated T cells in the mouse is a member of the nerve growth factor receptor (NGFR)/ tumor necrosis factor receptor (TNFR) gene family (4). This family is characterized by the presence of cysteine-rich motifs in the extracellular domains. Other members of this family include NGFR, B cell activation molecule CD40, the T cell activation molecule OX-40 in rat and CD27, the two receptors for TNF called TNFR-1 and TNFR-II, the apoptotic inducing molecule called Fas, and CD30 which has been suggested to play a role in the regulation of cellular growth and transformation (4, 5). 4-1BB is expressed on a variety of murine T cell lines, thymocytes, and mature T cells upon activation with concanavalin A (Con A), phorbol myristate acetate (PMA) and ionomycin, or anti-CD3i (6, 7). In the mouse crosslinking of 4-1BB on the cell surface with an agonistic monoclonal antibody, 1AH2, resulted in an up to 6.9-fold enhancement of T cell proliferation when activated suboptimally with anti-CD3i. Syngeneic BALB/c splenic T cells were induced to produce IL-2 when cultured with paraformaldehyde-fixed K46J stimulator cells, which are B7-negative and 4-1BB ligand (4-1BBL)-positive, in combination with anti-CD3 (7, 8). IL-2 production was not blocked using anti-CD28 mAb, but was inhibited using 4-1BB/AP, a fusion protein between 4-1BB and alkaline phosphatase (7, 8).

Murine 4-1BB is an Ç30-kDa glycoprotein expressed on activated T cells and plays a role in T-cell-mediated proliferative response. To date the majority of work on 4-1BB has been conducted in the mouse. To assess the role of 4-1BB in humans, mAbs were made against the recombinant human (rh) 4-1BB protein. One such mAb 4B4-1 specifically binds SF-21 insect cells expressing rh4-1BB but not irrelevant control protein as measured by flow cytometry (FCM). 4B4-1 mAb stains PMAand ionomycin-stimulated CEM (human T lymphoma) cells and PHA-stimulated peripheral blood T cells, but not resting cells. 4B4-1 mAb immunoprecipitates both Ç32 and Ç80 kDa protein from rh4-1BB expressing SF21 cells and an Ç39- and Ç85-kDa protein from PMAstimulated CEM cells under reducing conditions by SDS–PAGE. As added proof of its specificity, binding of FITC-labeled 4B4-1 mAb to PHA-stimulated T cells was blocked by rh4-1BB protein. Together these data demonstrate that 4B4-1 is specific for 4-1BB in humans. Unlike in the mouse, 4-1BB is expressed much earlier (within 24 hr) peaking around 2–3 days following PHA stimulation. As in the mouse 4-1BB is induced on both CD4/ and CD8/ T cell subsets. 4-1BB expression is induced upon PHA stimulation in both the naive (CD45RAhi-CD45ROlo/0) and the memory (CD45RAlo/0ROhi) T cell populations. Virtually all CD45RAhiROlo/0 cells upon culture in PHA give rise to an intermediate CD45RAhiROhi 4-1BB/ transitional cell and subsequently CD45RAlo/0ROhi 4-1BBlo/0 and CD45RAlo/0ROhi 4-1BBhi cells. In contrast, approximately 27% of CD45RAlo/0ROhi 4-1BB0 cells when cultured in PHA for 24 hr acquire 4-1BB expression and all remain CD45RAlo/0ROhi. q 1996 Academic Press, Inc.

INTRODUCTION T cells are critically important in acquired immunity, providing protection against viral, bacterial, and parasitic infection. T cells are induced into action when they

2 Abbreviations used: Recombinant human, rh; human, h; flow cytometry, FCM; ligand, L; goat anti-mouse gm-specific mAb, GAMGM; TCR, T cell receptor; NGFR, nerve growth factor receptor; TNFR, tumor necrosis factor receptor.

1

To whom correspondence should be addressed. Fax: (317) 2744090.

91

/ 6c0d$$8085

03-11-96 14:45:02

cias

0008-8749/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

AP: Cell Immuno

92

GARNI-WAGNER ET AL.

The ligand for 4-1BB (4-1BB/L) is found on activated macrophages and mature B cells (9). 4-1BB/L shows homology to TNF, LT-A, LT-B, CD40L, and CD27L, which form a merging family of molecules that bind to members of the NGF/TNF receptor family and likewise may play a role in cellular activation (8, 10). 4-1BB/L in CV-1 cells has been shown to play a primary role in the costimulation of the Con A-activated thymocytes and of PHA-activated splenic T cells (11). Furthermore, binding of the ligand by 4-1BB/AP has been shown to block T cell activation by K46J lymphomas in both an antigen-specific system and with polyclonally (antiCD3) activated T cells and to inhibit antigen presentation by normal spleen cells (11). Collectively, these data suggest that 4-1BB and its ligand are important in murine T-cell-mediated response. Furthermore, 4-1BB was found to specifically coimmunoprecipitate p56lck and up-regulate its expression in Con A-stimulated thymocytes, suggesting its role as a signaling molecule in murine T cells (12). The human homolog of 4-1BB (h4-1BB) was isolated from a cDNA library from an activated human T cell line (10, 13–15). The gene encoding h4-1BB is 1.4 kb and shows 60% identity of amino acid sequence to murine 4-1BB with high conservation in the cytoplasmic domain. h4-1BB mRNA is undetectable in unstimulated peripheral blood T cells but is induced within 1.5 hr following stimulation. Furthermore, it is inducible in T cell lines such as Jurkat and CEM (13–15). To characterize the expression and biological function of 4-1BB in humans, a panel of mAbs was made against recombinant human 4-1BB (rh4-1BB). Here we show that one such mAb, 4B4-1, is specific for human 4-1BB and its expression is induced on virtually all CD45RAhiROlo/0 and only a subpopulation of CD45RAlo/0ROhi cells. MATERIALS AND METHODS Cells. Human peripheral blood mononuclear cells (PBMC) were isolated by centrifugation on Histopaque 1077 (Sigma) at 400g for 30 min. The resulting interface cells were washed twice prior to use. Human T cells were isolated using Lympho-kwik (One Lambda, Inc., Canoga Park, CA) according to manufacturer’s instructions. Briefly, PBMCs were treated with Lymphokwik T cell isolation solution for 20 min at 377C followed by a 5-min centrifugation at 1500g. The resulting cell pellet was washed twice prior to use. PBMC or freshly isolated T cells were cultured in basic culturing medium consisting of RPMI 1640 (Gibco Laboratories), 10% FCS (Hyclone, UT), 11 MEM nonessential amino acids (Sigma), 1 mM sodium pyruvate (Gibco), 2 mM glutamine (Sigma), and 50 U/ml penicillin and 50 mg/ ml streptomycin (Gibco) supplemented with 5 mg/ml PHA (Calbiochem) and 50 mM 2-ME (Sigma). CEM and

/ 6c0d$$8085

03-11-96 14:45:02

cias

4B4-1, 1G5, and 4B11 hybridoma cells were maintained in basic culturing medium. CEM cells were stimulated overnight to 48 hr with 10 ng/ml PMA and 1 mM ionomycin in culturing medium. SF-21 cells, an insect cell line, were grown in Grace Media (Gibco) supplemented with 10% FCS, 3.3 g/liter yeastolate (Gibco), 3.3 g/liter lactoalbumin hydrolysate (Gibco), 2 mM glutamine (Sigma), and antibiotics at 277C. SF21 cells were infected with baculovirus containing the cDNA encoding for the expression of rh4-1BB or irrelevant protein for 1 hr in Grace Media and following a wash, cells were cultured for 48 hr prior to use in an assay. Antibodies and reagents. FITC-labeled goat antimouse mAb with specificity for the heavy chain of IgG and IgM (GAMGM) (Jackson Laboratories, MA) were used at 1:150. Tricolor-labeled anti-CD4, CD8, CD45RO, PE-labeled anti-CD3 mAbs, and isotype controls were purchased from Caltag (S. San Francisco, CA) and used according to manufacturer’s recommendations. 4B4-1 (IgG1) was purified over a protein G– Sepharose column and directly labeled by incubating the purified mAb in 0.4 mg/ml FITC (Pierce) on ice for 3 hr followed by filtration through Centricon 30 Microconcentrators (Amicon) to remove free-unbound FITC. FITC-labeled 4B4-1 was used at 3–4 mg per 5.0 1 105 cells for staining. Recombinant human 4-1BB. The cDNA sequence encoding the ectomonain of h4-1BB was fused in frame with the GST-binding domain of glutathionine S-transferase (GST) in PGEX-3 expression vector (Pharmacia). The 4-1BB–GST fusion protein was expressed in Escherichia coli strain Top1 (Stratagene). The 4-1BB–GST fusion protein allowed for efficient purification of rh41BB when isolated by GST–Sepharose and a Sepharose 4B column chromatographies. The rh4-1BB portion was purified by GST–Sepharose column and Sepharose 4B column chromatographies after cleavage of the fusion protein with factor Xa. Immunization and antibody production. Eightweek-old BALB/c mice were immunized with 50 mg of h4-1BB-GST fusion protein emulsified in Titermax (Cytech) or complete Freund’s adjuvant. Three intraperitoneal (ip) injections were administered 2 weeks apart (16). Three days following the last injection, the mice were sacrificed and their spleens were removed. Spleen cells were fused with Sp2/o myeloma cells as previously described (16). Briefly, spleen cells and Sp2/o cells were mixed at 5:1 ratio and fused using 50% PEG. Cells were washed, resuspended in OptiMEM (Gibco), 10% FCS, 5 mM hypoxanthine, 1% aminopterin, and 0.8 mM thymidine (HAT) and cultured in 96-well U-bottom plates (Corning). Resulting cell supernatants were screened by ELISA for rh4-1BB reactivity, using purified rh4-

AP: Cell Immuno

93

4-1BB EXPRESSION ON CD45RAhiROhi T CELL

1BB ectodomain. rh4-1BB-reactive hybridoma clones were isolated and subcloned (16). Flow cytometric (FCM) studies. Fresh cells were washed once and cultured cells were washed three times in staining medium consisting of PBS, 1% BSA and 0.1% glucose prior to staining. Approximately 2.5– 5 1 105 cells were resuspended in 50–100 ml of diluted mAb in staining medium and incubated at 47C for 30– 60 min. If mAb was not directly labeled, the cells were subsequently washed, resuspended in a 1:150 dilution of FITC-labeled GAMGM secondary mAb (Jackson Laboratories, MA) incubated 30 min, and washed three to four times. If primary mAb was directly labeled, cells were simply washed three to four times. Samples were fixed with 1% paraformaldehyde overnight prior to analysis on the FACScan by Becton Dickinson using Consort 30 or FACScan software. Appropriate isotype controls were included for all primary antibodies used. Gates were set on live cells only, based on forward versus side scatter profiles. Five to 15,000 events were collected for each sample. Competition assay. PHA-stimulated T cells were incubated 30 min on ice with staining media or varying concentrations of either rh4-1BB protein or irrelevant recombinant protein that was produced in using PGEX3 vector. Without washing 2.5 mg of 4B4/FITC-labeled mAb or isotype-matched FITC-labeled mAb was added at suboptimal concentrations and the cells were further incubated 30 min on ice. Cells were washed three times with staining media, fixed with 1% paraformaldehyde, and analyzed by FCM as described above. Immunoprecipitation studies. Cell surface proteins were labeled with biotin as previously described (17). Briefly, 1 1 107 cells were washed in ice-cold PBS, resuspended in freshly prepared 0.5 mg/ml animohexanoyl–biotin–N-hydroxysuccinimide ester (AH-BNHS, Zymed Lab., Inc) in PBS for 30 min at 47C with constant agitation. Cells were washed twice with 0.2 M glycine in PBS. Cells were lysed on ice for 15 min in 200 ml lysis buffer containing 1.0% NP-40, 20 mM Tris–HCl, 0.15 NaCl, 1 mM sodium orthovanadate, 5 mg/ml aprotinin, and 1 mg/ml leupeptin, pH 7.5. Suspension was centrifuged at 14,000 rpm for 10 min, 47C to remove debris. Twenty microliter of hybridoma culture supernatants was added for 1 hr at 47C followed by the addition of 30 ml of a 50% suspension of protein G–Sepharose 4B (Zymed Lab, Inc.) and incubated for 1 hr at 47C. Suspension was centrifuged and the resulting protein G–Sepharose pellet was washed. The immunoprecipitates were boiled 5 min in 20 ml SDS sample buffer and run on 10% SDS–PAGE. Proteins were transferred to Immobilon-P membrane (Millipore Corp., Bedford, MA) and blocked with 5% BSA in buffer containing 25 mM Tris–HCl, 0.15 NaCl, 0.05% Tween 20, pH 7.5

/ 6c0d$$8085

03-11-96 14:45:02

cias

(TBST). The blot was probed with avidin–horseradish peroxidase conjugate and developed using chemiluminescence detection kit (ECL, Amersham Corp., Arlington Heights, IL). RESULTS 4B4-1 mAb specifically binds human 4-1BB. To assess the role of 4-1BB on human T cells it was necessary to develop mAb specific for human 4-1BB. BALB/c animals were immunized with purified rh4-1BB protein and their splenocytes were fused with the Sp2/o fusion partner. The resulting hybridoma supernatants were screened for binding to rh4-1BB protein by ELISA (data not shown). ELISA-positive clones were subsequently assessed for staining of CEM (human T lymphoma) cells by FCM (Fig. 1). 4B4-1 mAb specifically bound CEM cells above isotype control, but only after stimulation with 10 ng/ml PMA and 1 mM ionomycin for 24 hr (Fig. 1). All other mAb positive by ELISA for rh4-1BB did not appear to bind CEM cells regardless of the cells activation state. To further address whether 4B4-1 mAb specifically recognized 41BB and to assess the specificity of the other mAb picked up by ELISA, SF-21 insect cells were infected with baculovirus containing cDNA encoding h4-1BB or an irrelevant protein. Infected SF-21 insect cells were then analyzed by FCM (Fig. 2). Only cells infected with the baculovirus expressing the h4-1BB and not irrelevant protein were found to bind 4B4-1, 4B11, and 1G5 mAbs. Interestingly, the 4B11 and 1G5 mAb also bound 4-1BB expressing SF-21 cells, but not the stimulated CEM cells. By immunoprecipitation, 4B4-1 mAb precipitated an Ç32-kDa protein from 4-1BB expressing SF-21 cells, but not from cells expressing an irrelevant protein. Precipitations from CEM cells stimulated with PMA and ionomycin, but not unstimulated cells produced a band of Ç39 kDa under reducing conditions (Fig. 3). Under nonreducing conditions, 4B4-1 mAb precipitated larger bands of Ç80–89 kDa from both the 4-1BB expressing SF-21 cells and stimulated CEM cells (Fig. 3B). These larger bands are most likely dimeric forms of 4-1BB. 4-1BB is induced on human peripheral blood T cells following PHA stimulation. To evaluate 4-1BB expression in human peripheral blood, mononuclear cells were isolated by passage over Ficoll, washed, and stained with 4B4-1 mAb. Of the lymphocyte-sized fraction, õ2% expressed 4-1BB, but upon PHA stimulation Ç16–19% expressed 4-1BB at 48 hr (data not shown). By multicolor analysis, õ2% of CD3/ (T) cells from peripheral blood were found to expressed 4-1BB (Fig. 4A), but by 48 hr PHA stimulation 18.5–21.5% of purified T cells (CD3/) from peripheral blood were 4-1BB positive (Fig. 4B). Without PHA stimulation õ2% of

AP: Cell Immuno

94

GARNI-WAGNER ET AL.

FIG. 1. Expression of 4-1BB on resting and activated CEM cells. CEM cells were cultured in media (A) or media containing PMA and ionomycin (B) for 24 hr. Cells were stained with 4B4-1, 4B11, and 1G5 hybridoma supernatants plus FITC-labeled GAMGM and analyzed by FCM.

cultured T cells expressed 4-1BB. Of PHA-stimulated T cells, 4-1BB was expressed on both CD4/ and CD8/ cells with Ç50% of 4-1BB/ cells expressing CD4 and Ç50% expressing CD8 (data not shown). B cells do not appear to express 4-1BB. Isotype-matched irrelevant mAb were used for each fluorescence parameter to correct for nonspecific binding of the mAbs. Furthermore, when competitive binding assays were conducted, binding of FITC-labeled 4B4-1 mAb by PHA-stimulated T cells was almost completely blocked when cells were

preincubated with 10 mg/ml purified rh4-1BB protein, but not irrelevant control protein at similar concentration (Fig. 5). This competitive blocking of 4B4-1/FITC mAb by rh4-1BB was dependent upon rh4-1BB concentration. Together these data demonstrate that 4B4-1 mAb is specific for human 4-1BB and that 4-1BB is expressed on activated not resting T cells. When purified T cells were stimulated with PHA for 72 hr, Ç25% of cells expressed 4-1BB of which nearly all appeared to express CD45RA and CD45RO as deter-

FIG. 2. Specificity of mAb for 4-1BB. SF-21 insect cells were infected with a Baculovirus expression vector containing the construct encoding for an irrelevant recombinant protein (A) or rh4-1BB (B). SF-21 cells were subsequently stained with 4B4-1, 4B11, and 1G5 hybridoma supernatants plus FITC-labeled GAMGM and analyzed by FCM.

/ 6c0d$$8085

03-11-96 14:45:02

cias

AP: Cell Immuno

95

4-1BB EXPRESSION ON CD45RAhiROhi T CELL

mined by FCM (data not shown). To further assess the possible association of 4-1BB and CD45RAhiROhi cells, in three separate experiments, peripheral blood T cells were sorted by FCM to obtain CD45RAhiRO0/lo cells, cultured in 5 mg/ml PHA, and analyzed by multicolor FCM for expression of 4-1BB (FITC-labeled), CD45RA (PE-labeled), and CD45RO (Tri-color-labeled) over a 7to 10-day period. In agreement with data presented by Picker et al., the larger more blasting cells appeared to up-regulate surface expression of CD45RO and undergo a transition from CD45RAhiRO0/lo (Days 0–1) to CD45RAhiROhi (Days 2–4) and ultimately resulted in a blasting population that had down-regulated expression of CD45RA and was predominately CD45RO (Fig. 6). In contrast the resting, nonblasting cells appeared to have limited up-regulation of CD45RO and until later in culture (5–9 days) remained largely CD45ROlo/0. 4-1BB is expressed within 24 hr of stimulation, reach-

FIG. 4. Expression of 4-1BB on fresh and stimulated T cells. (A) Freshly isolated PBMCs or (B) cells that had been cultured 48 hr in PHA were stained with PE-labeled anti-CD3 and FITC-labeled 4B4-1 and analyzed by FCM. A gate was set on CD3/ (FL2/) cells and analyzed for 4-1BB (FL1) expression. Isotype-matched controls were run.

ing peak expression around 48–72 hr and thereafter declines (Fig. 6B). 4-1BB is preferentially expressed on the blasting cells and virtually all CD45RAhiROhi cells (Fig. 6B). Increased expression of 4-1BB on Day 9 is due to restimulation of 4-1BB on CD45RAlo/0ROhi cells with PHA. Interestingly, when CD45RA0/loROhi sorted cells were cultured in PHA for 24 hr, approximately 27% of CD45RA0/loROhi cells expressed 4-1BB, although expression varied significantly (Fig. 7). Furthermore, the kinetics of 4-1BB expression on CD45ROhi sorted cells were similar to that observed on CD45RAhiRO0/lo sorted cells. CD45RA expression remained õ5% throughout culture (0–9 days). DISCUSSION FIG. 3. Immunoprecipitation by 4B4-1 mAb from SF-21 and CEM cells. (A) 4B4-1 mAb was used to precipitate cell lysates from SF-21 cells expressing 4-1BB (/) or irrelevant protein (0) and lysates from CEM cells with (/) and without (0) PMA and ionomycin stimulation. Precipitates were run on SDS–PAGE under reducing conditions. (B) Cells lysates from PMA and ionomycin-stimulated CEM cell and 4-1BB expressing SF-21 cells were precipitated using 4B41 mAb and run on 10% SDS–PAGE under reducing (R) or nonreducing (NR) conditions. h4-1BB was detected by streptavidin–horseradish peroxidase conjugate and chemiluminescence detection kit (ECL Amersham).

/ 6c0d$$8085

03-11-96 14:45:02

cias

4-1BB is expressed on actived but not resting murine T cells (6, 7). Crosslinking of 1AH2 mAb directed against murine 4-1BB has been shown to enhance antiCD3-induced T cell proliferation (7, 18). Normal splenic cell antigen presentation and T cell activation can be blocked by inhibiting the binding of 4-1BB on T cells to its ligand on B cells and macrophages with 4-1BB/ AP, a fusion protein containing the extracellular domains of 4-1BB and alkaline phosphatase (11). Although we are beginning to understand the role of 4-

AP: Cell Immuno

96

GARNI-WAGNER ET AL.

FIG. 5. Specificity of 4B4-1 mAb for 4-1BB on human T cells. Purified T cells cultured in 5 mg/ml PHA for 48 hr were preincubated with varying concentrations of purified rh4-1BB (A) or irrelevant protein (B) and subsequently stained with FITC-labeled 4B4-1 mAb and analyzed by FCM. Isotype-matched controls were run.

1BB in the mouse, little is known about the expression or role of 4-1BB in humans. In order to further characterize 4-1BB in humans, we developed and here describe the characterization of 4-1BB-specific mAbs, 4B4-1, 4B11, and 1G5. All mAb specifically bind 4-1BB expressing SF-21 cells (Figs. 2A and 2B). The 4B4-1 mAb specifically stain the T cell line, CEM (Figs. 1A and 1B), and peripheral blood T cells following activation (Fig. 6), but does not bind resting cells. In this regard, 4-1BB expression in man is similar to that observed in the mouse (6, 7). 4B4-1 mAb immunoprecipitates a homodimeric protein of Ç32 kDa under reducing conditions and Ç85 kDa under nonreducing conditions from 4-1BB expressing SF21 cells (Figs. 3A and 3B) and a protein of Ç35 and Ç85 kDa from activated CEM cells (Figs. 4A and 4B). The observed molecular weight of 4-1BB in human is similar to that anticipated based on amino acid sequence and anticipated glycosylation and is similar to that observed in the mouse (4, 18). Furthermore, 10 mg of rh4-1BB protein, but not irrelevant protein, completely blocks the binding of FITC-labeled 4B4-1 mAb to PHA-stimulated T cells (Figs. 7A and 7B). Collectively, these data conclusively show that 4B4-1 mAb is specific for human 4-1BB. Although 4B11 and 1G5 mAbs specifically recognize 41BB expressed on SF-21 cells (Fig. 2), they do not recognize 4-1BB expressed on activated T cells (Fig. 1). This is most likely due to the mAb having specificity for a cryptic or unique binding site(s) that is not exposed or present on T cells but is accessible or present on SF-21 cells due to slight differences in glycosylation and processing between human T cells and insect cells (SF-21). In mouse, neither 4-1BB mRNA nor surface expression is detectable on resting splenocytes or unstimulated cloned T cells. However, upon activation of T cells by anti-CD3 or anti-TCRab, 4-1BB mRNA is

/ 6c0d$$8085

03-11-96 14:45:02

cias

detected within 3 hr of stimulation and is first detectable on the cell surface 2 – 3 days following stimulation. Maximum surface expression is reached around Day 6 poststimulation (6, 7, 19). Similar to the mouse, 4-1BB is not detected on the surface of freshly isolated peripheral blood T cells in man, but is readily detected (within 1.5 hr) following PHA stimulation (Fig. 5). Unlike in the mouse, 4-1BB is expressed much more rapidly in humans reaching peak surface expression between 12 and 48 hr and begins to decrease by 72 hr poststimulation as do the number of cells expressing 4-1BB on their cell surface (Fig. 6B). In both mouse and humans, 4-1BB is approximately equally expressed on CD4/ and CD8/ T cell subsets (data not shown) (19). Furthermore, 4-1BB was not present on the cell surface nor was its expression induced upon activation of two other T cell lines (Jurkat and MT4), B cell line (Priess), a monocytic cell line (U937), the erythroleukemia line (K562), nor the myeloid cell line (MO7e) (data not shown). Given that 4-1BB is expressed on only 25% of PHA-stimulated T cells, it is not surprising that the Jurkat and MT4 T cell lines do not express 4-1BB even after PHA stimulation. Although freshly isolated PBMCs do not express 41BB, within 24–48 hr of PHA stimulation 16–19% of PBMCs and Ç25% of purified PHA-stimulated T cells expressed 4-1BB (data not shown). 4-1BB expression is induced on both CD4 and CD8 T cell subpopulations (data not shown). Furthermore, its expression is induced on both the CD45RAhiROlo/0 (naive phenotype) and CD45RAlo/0ROhi (memory phenotype) subpopulations of T cells. Expression of 4-1BB appears to define a major subpopulation of CD45RAhiROlo/0 PHA-activated T cells of blasting size that undergo the transition from CD45RAhiROlo/0 (naive phenotype) to CD45RAlo/0ROhi (memory phenotype) through an intermediate popula-

AP: Cell Immuno

4-1BB EXPRESSION ON CD45RAhiROhi T CELL

97

FIG. 6. Expression of 4-1BB on CD45RAhi and CD45ROhi-activated T cells. (A) CD45RAhi and CD45RO0/lo T cells were purified by cell sorting, cultured in 5 mg/ml PHA, stained at various times with FITC-labeled 4B4-1 mAb, PE-labeled CD45RA, and Tri-color-labeled CD45RO, and analyzed by FCM. Isotype-matched controls were run. Based on forward vs side scatter profile, black dots are the blasting cells while gray dots are the more resting cells. (B) Of the these cultured cells, gates were set so that only 4-1BBhi// cells were further analyzed for expression of CD45RA and CD45RO.

tion of cells which are CD45RAhiROhi. The expression pattern of 4-1BB was different from that of CD27 which was preferentially expressed on the CD45RAhiROlo/0 subset of CD4 T cells (21, 22). Picker et al., in an elegant set of FCM experiments, demonstrated that naive T cells undergo a ‘‘stepwise, unidirectional progression’’ from a naive (CD45RAhiROlo/0) to a memory (CD45RAlo/0ROhi) phenotype through a distinct CD45RAhiROhi intermediate cell type (23). In peripheral blood few cells expressing this intermediate phenotype were detected. However, in secondary lymphoid tissue, such as tonsil where the immune response is actively occurring, 2–

/ 6c0d$$8085

03-11-96 14:45:02

cias

10% of T cells were shown to be CD45RAhiROhi. Our data are consistent with their finding that cells undergo a sequential transition through a CD45RAhiROhi cell and that this transition is unidirectional given that purified CD45RA0/loROhi cells in culture do not reexpress CD45RA. These data demonstrate that 4-1BB is induced in both naive and memory populations as well as both major T cell subpopulations (CD4/ and CD8/). Given the apparent role of 4-1BB as an activation molecule in mouse and its apparent role in T cell proliferation and survival in humans as demonstrated by Schwarz

AP: Cell Immuno

98

GARNI-WAGNER ET AL. 3. 4. 5. 6. 7. 8.

9. 10.

11. hi

FIG. 7. Expression of 4-1BB on sorted CD45RO -activated T cells. CD45ROhi and CD45RA0/lo T cells were purified by cell sorting, cultured in 5 mg/ml PHA, and stained at 24–48 hr of culture with FITC-labeled 4B4-1 mAb, PE-labeled CD45RA, and Tri-color-labeled CD45RO. Cells were analyzed by FCM. Isotype-matched controls were run.

12. 13. 14. 15.

et al. (14), 4-1BB mAb will serve as an invaluable tool in further assessing the role of 4-1BB in man.

16. 17.

ACKNOWLEDGMENTS 18. We thank Dr. Janice S. Blum for her excellent advice and critical review of the manuscript. We are deeply in depted to Dr. Edward Srour, Susan Grigsby, and Jon McMahel and the Indiana Medical Center Cancer Center Flow Cytometry Laboritory in addition to Carol Kulesavage for their expert technical help. This work was supported by the NIH Grants AI28175 and DE10525.

19. 20. 21.

REFERENCES

22.

1. June, C. H., Ledbetter, J. A., Lindsey, P. S., and Thompson, C. B., Immunol. Today 11, 211, 1990. 2. Linsley, P. S., Brady, W., Grosmaire, L., Aruffo, A., Damle, N. K., and Ledbetter, J. A., J. Exp. Med. 173, 721, 1991.

23.

/ 6c0d$$8085

03-11-96 14:45:02

cias

Jenkins, M. K., Immunol. Today 13, 69, 1992. Smith, C. A., Farrah, T., and Goodwin, R. G., Cell 76, 959, 1994. Beutler, B., and vanHuffel, C., Science 264, 667, 1994. Kwon, B. S., Kestler, D. P., Eshkar, Z., Oh, K., and Wakulchik, M., Cell. Immunol. 121, 414, 1989. Pollok, K. E., Kim, Y.-J., Zhou, Z., Hurtado, J., Kim, K. K., and Kwon, B. S., J. Immunol. 150, 771, 1993. Goodwin, R. G., Din, W. S., Davis-Smith, T., Anderson, D. M., Gimpel, S. D., Sato, T. A., Maliszewski, C. R., Brannan, C. I., Copeland, N. G., Jenkins, N. A., et al. Eur. J. Immunol. 23, 2631, 1993. Pollok, K. E., Hurtado, J., Kim, Y.-J., Zhou, Z., Kim, K. K., Pickard, R. T., and Kwon, B. S., Eur. J. Immunol. 24, 367, 1994. Alderson, M. R., Smith, C. A., Tough, T. W., Davis-Smith, T., Armitage, R. J., Falk, B., Roux, E., Baker, E., Sutherland, G. R., Din, W. S., and Goodwin, R. G., Eur. J. Immunol. 24, 2219, 1994. DeBenedette, M. A., Chu, N. R., Pollok, K. E., Hurtado, J., Wade, W. F., Kwon, B. S., and Watts, T. H., J. Exp. Med. 181, 985, 1995. Kim, Y.-J., Pollok, K. E., Zhou, Z., Shaw, A., Bohlen, J. B., Fraser, M., and Kwon, B. S., J. Immunol. 151, 1255, 1993. Schwarz, H., Tuckwell, J., and Lotz, M., Gene 134, 295, 1993. Schwarz, H., Valbracht, J., Tuckwell, J., vonKempis, J., and Lotz, M., Blood, in press, 1995. Zhou, Z., Kim, S., Hurtado, J., Lee, Z. H., Kim, K. K., Pollok, K. E., and Kwon, B. S., Immunol. Lett. 45, 67, 1995. Harlow, E., and Lane, D. (Eds.), ‘‘Antibody Lab Manual,’’ Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1988. Pelchen-Matthews, A., Armes, J. E., Griffiths, G., and Marsh, M., J. Exp. Med. 173, 575, 1991. Kim, Y. J., Pollok, K. E., Kim, S. H., and Kwon, B. S., Submitted, 1995. Pollok, K. E., Kim, S. H., and Kwon, B. S., Eur. J. Immunol. 25, 488, 1995. Biffen, M., McMichael-Phillips, D., Larson, T., Venkitaraman, A., and Alexander, D., EMBO J. 13, 1920, 1994. Sugita, K., Hirose, T., Rothstein, D. M., Donahue, C., Schlossman, S. F., and Morimoto, C., J. Immunol. 149, 3208, 1992. Kurosawa, K., Kobata, T., Tachibana, K., Agematsu, K., Hirose, T., Schlossman, S. F., and Morimoto, C., Cell. Immunol. 158, 365, 1994. Picker, L. J., Treer, J. R., Ferguson-Darnell, B., Collins, P. A., Buck, D., and Terstappen, L. W. M. M., J. Immunol. 150, 1105, 1993.

AP: Cell Immuno