The lymphoid chemokine CCL21 costimulates naïve T cell expansion and Th1 polarization of non-regulatory CD4+ T cells

ellular Immunology Cellular Immunology 231 (2004) 75–84 www.elsevier.com/locate/ycimm

The lymphoid chemokine CCL21 costimulates naïve T cell expansion and Th1 polarization of non-regulatory CD4+ T cells Kenneth Flanagana, Dorota Moroziewiczb, Heesun Kwakb, Heidi Hörigb, Howard L. Kaufmana,b,¤ a

Department of Pathology, Columbia University, New York, NY 10032, USA Department of Surgery, Columbia University, New York, NY 10032, USA

b

Received 28 September 2004; accepted 8 December 2004 Available online 21 January 2005

Abstract CCL21 (SLC/6Ckine) is constitutively expressed by secondary lymphoid tissue and attracts CCR7-expressing mature dendritic cells and naïve T cells. Recent studies demonstrated that intra-tumoral delivery of CCL21 induces tumor regression in a T cell dependent manner. CCL21 is known to mediate T cell traYcking but little is known about its function as a costimulatory molecule. Herein, we demonstrate that CCL21 costimulates expansion of CD4+ and CD8+ T cells and induces Th1 polarization. These eVects were speciWc for naïve T cells, and we show that CD4+CD25+ regulatory T cells were hyporesponsive to CCL21 induced migration, and unresponsive to CCL21 costimulation. These unique functions of CCL21 to both attract naïve T cells as well as costimulate their proliferation and diVerentiation, suggests that CCL21 is a pivotal molecule for priming T cell responses and has therapeutic implications for local delivery of CCL21. The coordinated eVects of CCL21 on T cell migration and activation may also represent a more comprehensive paradigm for the activity of other chemokines as well.  2004 Elsevier Inc. All rights reserved. Keywords: Chemokines; Costimulation; Chemotaxis; Regulatory T cells; T lymphocytes; T cell diVerentiation

1. Introduction The CC chemokines are a family of small, structurally related molecules that mediate migration of immune cells. Ectopic expression of CCL21 within established tumors results in increased numbers of T cells at the site of expression, and has been associated with signiWcant anti-tumor responses in vivo [1,2]. In these models, inWltrating T cells in regressing tumors demonstrate an activated phenotype, including increased secretion of IFN-
[3]. Furthermore, in at least some models the anti-tumor eVects were completely abrogated when CD4+ T cells were depleted prior to CCL21 treatment [2]. The mecha-

¤

Corresponding author. Fax: +1 212 342 0234. E-mail address: [email protected] (H.L. Kaufman).

0008-8749/$ - see front matter  2004 Elsevier Inc. All rights reserved. doi:10.1016/j.cellimm.2004.12.006

nism of tumor regression was initially explained by enhanced migration of leukocytes to the site of CCL21 expression, yet it is also possible that CCL21 could directly induce proliferation of a smaller number of inWltrating leukocytes. In support of this, in vivo data demonstrated that T cells inWltrating CCL21 treated tumors had the phenotype of activated, proliferating cells (our own unpublished observations). Thus, we sought to evaluate the direct eVects of CCL21 on T cells in vitro and determine if CCL21 functions as a true costimulatory molecule. The full activation and diVerentiation of eVector T cells depends on the delivery of two signals, the Wrst provided through the TCR and the second imparted by costimulatory molecules via a variety of ligands [4]. The importance of costimulation can be demonstrated in T cells receiving only TCR signaling; such T cells may be

76

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

rendered unresponsive or anergic [5]. While most of the deWned costimulatory molecules are members of the Ig and TNFR superfamilies, such as B7 and CD40L, respectively, several inXammatory chemokines, including CCL3 and CCL5, have demonstrated the ability to provide activation signals to T cells [6,7]. The expression of these chemokines is limited to sites of antigen insult, where local expression may potentiate, rather than prime T cell responses. In contrast to the inXammatory chemokines, CCL21 is constitutively expressed in secondary lymphoid tissue and has not previously been shown to eVect T cell activation. Under normal physiological conditions, CCL21 is expressed by high endothelial venules within lymph nodes and ensures the temporal and spatial colocalization of mature DC and naïve T cells within the T cell zones [8]. An estimated 100–200 naïve T cells exist for a given antigen and the probability that one of these circulating T cells will contact its cognate antigen on a migrating DC is minute [9]; therefore, CCL21 functions to coordinate the migration of CCR7-expressing mature DC as well as naïve T cells, thus ensuring the required cell contact for T cell priming. While CCL21 is known to mediate traYcking through CCR7, direct signaling also results in activation of the integrin leukocyte function associated antigen 1 (LFA1-1) within nanoseconds of binding CCL21 [10,11]. Recent evidence suggests that CCR7 may function through a variety of other signaling pathways in lymphocytes as well, including phosphatidylinositol 3-OH kinase (PI3K), protein kinase C (PKC), phospholipase C 2/3, focal adhesion kinase, src kinases and Janus kinase (Jak) family members [12–14]. This widespread signaling provides a framework for CCL21 mediated lymphocyte arrest under high Xow conditions in vivo and suggests that CCL21 may be important for other T cell functions. Stimulation of T cells with CCL21 through CCR7 precedes stimulation via conventional costimulatory pathways, and thus CCL21 could provide a very early signal for priming of T cells. In this report, we provide evidence that CCL21 is a costimulatory molecular for T cells in vitro. We focused on CCL21 as a paradigm for lymphoid chemokines due to its high concentration in lymphoid tissues and previous evidence suggesting a strong anti-tumor eVect of this chemokine after local intra-tumoral injection [1,2]. CCL21 costimulation aVects both CD4+ and CD8+ T cells, resulting in T cell expansion, Th1 cytokine release, and upregulation of early T cell activation markers. Since it has recently been determined that the suppression mediated by CD4+CD25+ regulatory T cells is inversely related to CCR7 expression in mice [15], we also examined the eVects of CCL21 on these cells. Unlike naive T cells, CD4+CD25+ regulatory T cells did not migrate nor proliferate in response to CCL21 exposure. Thus, CCL21 may preferentially induce migration and

activation of non-regulatory T cells. These results provide evidence that CCL21 attracts naïve, non-regulatory T cells and acts as an early costimulatory molecule for priming and directing antigen-speciWc T cell responses prior to antigen recognition.

2. Materials and methods 2.1. Cell separation Single cell suspensions of pooled spleens and lymph nodes were obtained from six- to eight-week-old female C57BL/6 mice (Charles River Laboratories). Whole T cells, or puriWed populations of CD4+, CD8+ or CD4+CD25+ regulatory T cells, as indicated, were magnetically separated according to manufacturer’s protocols (Miltenyi Biotech). CD4+, CD8+, and CD4+CD25+ cells were separated by positive selection methods and were routinely 96–99% pure as analyzed by Xow cytometry. Whole T cells and CD4+CD25¡ T cells were separated by negative selection and were routinely 85–90% pure. 2.2. Proliferation assays Ninety-six-well plates were coated with anti-CD3 mAb (1 g/ml, BD Pharmingen, clone 145-2C11) with or without polyclonal anti-CD28 mAb (5 g/ml, BD Pharmingen, clone 37.51). 3 £ 105 of the indicated cells were cultured in complete media (CM, RPMI supplemented with 10% FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin) and added to the plates in the presence or absence of CCL21 (a generous gift from Chiron). The concentration of CCL21 used in proliferation assays was 2 g/ml unless otherwise indicated. Plates were incubated for 72 h at 37 °C with [3H]thymidine added for the Wnal 6 h and thymidine incorporation was measured using a Wallac Microbeta Trilux scintillation counter (Perkin–Elmer). In other studies, splenocytes were labeled with 2.5 M CFSE using manufacturer’s protocols (Molecular Probes) prior to cell culture. After 72 h of incubation, CFSE labeled cells were collected and stained with PE-labeled anti-CD4, CD8, or B220 mAbs (all from BD Pharmingen) and analyzed by Xow cytometry for dilution of CFSE intensity. For proliferation assays with regulatory T cells, 5 £ 104 of either CD4+CD25¡ or CD4+CD25+ T cells (magnetically separated to over 90% purity using the mouse CD4+CD25+ regulatory T cell isolation kit from Miltenyi Biotec) was incubated with 5 £ 104 irradiated T cell depleted splenocytes from syngeneic mice in CM supplemented with 10 M BME in U-bottom plates. Soluble anti-CD3 mAb was added at a Wnal concentration of 0.1 g/ml, with or without anti-CD28 mAb or CCL21

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

(5 and 2 g/ml, respectively), the cells were incubated for 72 h, and incorporation of [3H]thymidine was determined as above. Data from these experiments comparing diVerent cell types is normalized by expression as a simulation index, indicating the fold increase in a given treatment group compared to cells treated with anti-CD3 mAb alone. In restimulation experiments, the cells costimulated as above, for 72 h, were collected, washed and rested in IL-2 (10 ng/ml) for 72 h, and restimulated with anti-CD3 alone (1 g/ml) for an additional 48 h before supernatants were collected. In some experiments 5 £ 104 CD4+CD25¡ T cells were incubated with increasing numbers of CD4+CD25+ T cells in the presence of soluble anti-CD3 (1 g/ml) with or without soluble antiCD28 (5 g/ml). All proliferation data are representative of the mean § SEM of 3–4 wells per treatment. 2.3. Mixed leukocyte reaction Bone marrow derived dendritic cells (BMDCs) were generated as previously described with slight modiWcations [16]. Bone marrow, collected from tibias and femurs of 6–8-week-old female BALB/c mice, was resuspended in CM supplemented with 15 ng/ml of GM-CSF and IL-4 (Peprotech). Two-thirds of the medium was replaced on days 2 and 4 and BMDC were collected by scraping on day 6, irradiated (3000 Gy) and incubated in graded doses together with 3 £ 105 puriWed allogeneic T cells in round bottom 96-well culture plates in CM for 5 days. Where indicated, CCL21 was preincubated for 30 min at 25 °C, with 10 g/ml of neutralizing polyclonal goat anti-mouse CCL21 Ab or goat IgG control (both from R&D Systems). Where indicated, cells were preincubated for 45 min on ice with antiLFA-1 neutralizing mAb (clone M17/4), anti-ICAM-1 neutralizing mAb (clone 1A29) or rat IgG2a isotype control (R&D Systems, clone 54447) at 10 g/ml prior to culture. Proliferation of T cells was determined as above. 2.4. Flow cytometry Isolated T cells, costimulated with CCL21 as above, were collected at indicated time points and stained with APC conjugated anti-L-selectin, and anti-CD25, PE conjugated anti-CD69 and FITC conjugated anti-LFA1 mAbs, or the appropriate Xuorescently labeled isotype control mAbs (all from BD Pharmingen) and analyzed by Xow cytometry. For measurement of intracellular cytokines, cells were Wrst surface stained for CD4 as described, permeabilized using a CytoWx/Cytoperm kit (BD Pharmingen). Cells were subsequently stained with APC labeled IFN-
or IL-4. Data are gated on live CD4+ gated cells.

77

2.5. Cytokine measurement Supernatant samples were collected from in vitro cultures at indicated time points and the concentrations of IFN-
, TNF-, IL-2, IL-4, and IL-5 were simultaneously quantitated using a Mouse Th1/Th2 Cytokine Cytometric Bead Array (CBA) Kit (BD Biosciences) according to manufacturer’s protocols. 2.6. Microchemotaxis T cell migration was measured across a 5 m polycarbonate membrane (Costar) in a microchemotaxis assay as described [1]. BrieXy, CCL21 protein (1 g/ml) was added to the lower chamber of a transwell plate, and 3 £ 105 of the indicated puriWed population of T cells were added to the upper chamber and migration proceeded for three h at 37 °C. Cells in the lower chamber were collected and stained with APC-labeled anti-CD4 mAb (BD Pharmingen) for Xow cytometric analysis after the addition of 50,000 15 m unlabeled polystyrene beads (Bangs Laboratories). Flow cytometry proceeded by counting 5000 bead events, and the number of migrating cells was determined by the following formula: (# of counted cell events/5000) £ 50,000), and is expressed as the percentage of cells initially added to the transwell. 2.7. Statistical analysis Statistical analyses were performed using the Student’s t test (*p < 0.05 and **p < 0.005).

3. Results 3.1. CCL21 costimulates lymphocyte proliferation To determine whether CCL21 costimulates leukocyte expansion, splenocytes were stimulated with a sub-optimal dose of plate-bound anti-CD3 mAb antibody along with the lymphoid chemokines CCL19, CCL21 or CXCL13. Expansion of leukocytes cultured with CCL21, or the structurally and functionally related chemokine CCL19 [17,18], was greatly enhanced (Fig. 1A). Proliferation of murine splenocytes was observed when exposed to CCL21 at concentrations as low as 20 ng/ml (Fig. 1B). Importantly, the range of CCL21 concentrations analyzed are similar to concentrations inducing T cell migration, and is far below the estimated concentration of CCL21 in lymphoid tissue [19,20]. No proliferation of cells occurred in the absence of TCR stimulation, consistent with a costimulatory, rather than mitogenic, function. Though the CCL21 protein tested free of endotoxin (<1 EU/mg protein), we conWrmed that the above eVects were mediated by CCL21 as addition of CCL21 speciWc

78

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

Fig. 1. CCL21 costimulates leukocyte proliferation. (A) Splenocytes were stimulated with sub-optimal doses of plate-bound anti-CD3 mAb (1 g/ml) in the presence or absence of the indicated chemokines (2 g/ml). (B) Splenocytes were plated as described in Materials and methods without antiCD3 or in the presence of plate-bound anti-CD3 mAb (1 g/ml). CCL21 was added to the cultures, at the concentrations indicated. Proliferation was observed only in the presence of TCR engagement conWrming a co-stimulatory role for CCL21. (C) Splenocytes cultured with plate-bound anti-CD3 mAb (1 g/ml) in the presence or absence of CCL21 (2 g/ml) with or without addition of anti-CCL21 neutralizing antibody or control IgG (10 g/ ml). Representative results of three separate experiments. Statistical signiWcance is compared to cells receiving no CCL21.

neutralizing antibody inhibited CCL21 mediated proliferation (Fig. 1C). Furthermore, the results were conWrmed with a commercially available source of CCL21 (R&D Systems) with similar results (data not shown).

activation of T cells was more pronounced on CD4+ T cells. 3.3. CCL21 mediates polarization of CD4+ T cells toward production of Th1 cytokines

3.2. CCL21 speciWcally and directly costimulates T cells To detect the eVects of CCL21 on deWned subpopulations of lymphocytes, we labeled splenocytes with CFSE and measured proliferation of CD4+ and CD8+ T cells by Xow cytometry. After 72 h, only 30% of CD4+ T cells cultured in the absence of CCL21 had divided more than once. In contrast, nearly 90% of CCL21 costimulated CD4+ T cells divided a minimum of two times (Fig. 2A). Similarly, nearly all CD8+ T cells (86%) had undergone four or more divisions when costimulated with CCL21, compared to only 25% of CD8+ T cells cultured in the absence of CCL21 (Fig. 2B). Interestingly, CD4+ and CD8+ T cells cultured with anti-CD3 mAb and CCL21 resulted in a comparable number of cell divisions as seen in cultures with anti-CD3 and anti-CD28 mAbs. As expected, proliferation of B cells or MHC II expressing cells was unaVected by addition of CCL21 (data not shown), indicating that the costimulatory eVects of CCL21 were speciWc for T cells, expanding both CD4+ and CD8+ T cells. To further conWrm that CCL21 costimulation did not require non-T cells such as APC, puriWed CD4+ (Fig. 2C) or CD8+ (Fig. 2D) T cells were cultured in the presence or absence of CCL21. CCL21 directly costimulated both populations of T cells, as proliferation was enhanced in these puriWed T cell cultures. CCL21 resulted in decreased expression of L-selectin, a marker of naïve T cells, and a lymphoid homing molecule, with a more pronounced decrease on CD4+ T cells than CD8+ T cells (Fig. 3). In addition, expression of CD69, an early T cell activation marker, and CD25 an intermediate marker of T cell activation were both upregulated on CD4+ and CD8+ T cells incubated in the presence of CCL21, though the eVects of CCL21 on the

Once activated, CD4+ T cells undergo a process of diVerentiation into either Th1 or Th2 type eVector cells which can be characterized by their cytokine secretion proWles [21]. To analyze the role of CCL21 in directing Th1/Th2 polarization of CD4+ T cells, we sampled supernatants of CCL21 costimulated cells for representative cytokines. As expected, CCL21 costimulated CD4+ T cells demonstrated increased secretion of IL-2 within 24 h, coincident with the onset of proliferation (Fig. 4A). In addition, increased secretion of Th1 cytokines, TNF- and IFN-
was observed, while levels of the Th2 cytokines, IL-4 and IL-5 were low and virtually unchanged from control T cell supernatants. Since it remained possible that increased production of Th1 cytokines resulted from the increased expansion of CCL21 costimulated cells, we conWrmed Th1 diVerentiation of CD4+ T cells by evaluating intracellular levels of IFN-
(Fig. 4B) or IL-4 (Fig. 4C). While there was no detectable intracellular IL-4 in any of the samples, IFN
levels in CCL21 costimulated CD4+ T cells were higher than in cells cultured with TCR stimulation alone, indicating increased production of IFN-
. To further analyze polarization of CD4+ T cells in response to CCL21, CD4+ T cells were treated as above for three days, rested in IL-2 for three days, and restimulated with platebound anti-CD3 (in the absence of costimulation). As has been previously reported, costimulation of CD4+ T cells via CD28 polarized T cells toward secretion of Th2 cytokines (Fig. 4D) [22]. In contrast, secondary stimulation of CD4+ T cells exposed to CCL21 during primary stimulation resulted in secretion of IFN-
, with no increase in production of Th2 cytokines. This data suggests that CCL21 costimulation polarizes CD4+ T cells toward cell mediated Th1 immune responses.

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

79

Fig. 2. CCL21 speciWcally costimulates T cells. CFSE labeled splenocytes, cultured in the presence of plate-bound anti-CD3 mAb (1 g/ml) in the presence or absence of CCL21 (2 g/ml) or plate-bound anti-CD28 mAb (5 g/ml) for 72 h were subsequently labeled with Xuorescent anti-CD4 mAb (A), anti-CD8, (B) and proliferation of subsets of lymphocytes was determined by dilution of the CFSE label. Representative results of two separate experiments. PuriWed CD4+ (C) or CD8+ (D) T cells were cultured with plate-bound anti-CD3 mAb (1 g/ml) in the presence or absence of CCL21 (2 g/ml) and proliferation was determined. The number of cell divisions is shown in graphical form below the histograms for media, CCL21, and anti-CD28. Representative results of three separate experiments. Statistical signiWcance is compared to cells receiving no CCL21.

3.4. CCL21 does not costimulates CD4+CD25+ regulatory T cells Though little is known about the migratory and costimulatory properties of CD4+CD25+ regulatory T cells, recent reports suggest that these regulatory cells can be segregated based upon expression of L-selectin as well as CCR7 [15]. The population of CD4+CD25+ T cells with regulatory function was found to lack expression of CCR7 and to have reduced capacity to migrate to lymphoid tissue [15]. Accordingly, we discovered that while CCL21 attracted over 50% of

CD4+CD25¡ T cells in in vitro assays (Fig. 5A), less than 20% of CD4+CD25+ T cells migrated in response to CCL21, demonstrating that CD4+CD25+ T cells are hyporesponsive to CCL21. Although CD4+CD25+ T cells are known to be unresponsive to anti-CD3 mAb stimulation [23], a variety of stimuli are capable of inducing CD4+CD25+ T cells proliferation, including IL-2 and anti-CD28 mAb [24]. To determine the eVect of CCL21 on CD4+CD25+ T cells we evaluated CCL21 induced proliferation in separated CD4+ T cell populations. Non-regulatory T cells demonstrated a two-fold enhancement in proliferation in

80

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

Fig. 3. CCL21 costimulation induces early T cell activation. T cells were cultured with plate-bound anti-CD3 mAb (1 g/ml) in the presence (blue lines) or absence (green lines) of CCL21 (2 g/ml) for 24 h. Cells were collected, and stained for Xow cytometry as described in Materials and methods. Cells were gated on CD4+ or CD8+ T cells, and data indicates the expression levels of activation markers in each T cell subset. Data represents one of two individual experiments.

Fig. 4. CCL21 mediated costimulation of CD4+ T cells results in production of Th1 cytokines and Th1 polarization. (A) Supernatant from CD4+ T cells cultured with plate-bound anti-CD3 mAb (1 g/ml) in the presence or absence of CCL21 (2 g/ml) or anti-CD28 (5 g/ml) was assayed by cytometric bead array and Xow cytometry for the indicated cytokines. Data represent one of two individual experiments. Intracellular levels of the representative cytokines of IFN-
(B) or IL-4 (C) after stimulation with anti-CD3 alone (thin lines) or anti-CD3+ CCL21 (thick lines) compared to isotype control stained cells (Wlled histograms). (D) CD4+ T cells were cultured with plate-bound anti-CD3 mAb in the presence or absence of CCL21 or anti-CD28 for three days, rested in IL-2, and re-exposed to anti-CD3 for 48 h, and supernatants were assayed for the indicated cytokines as above.

response to CCL21 costimulation. CD4+CD25+ T cells, however, proliferated only in response to CD28 costimulation and demonstrated no ability to expand in the presence of CCL21 (Fig. 5B). These results suggest that CCL21 preferentially costimulates CD4+CD25¡ T cells.

The fundamental nature of CD4+CD25+ regulatory T cells is the polyclonal inhibition of naive T cell expansion [25]. As our data demonstrated that CCL21 preferentially costimulates non-regulatory T cells, we wished to determine whether CCL21 could function to

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

81

Fig. 5. CCL21 mediated costimulation is speciWc for non-regulatory T cells. (A) Migration toward CCL21 in puriWed CD4+CD25¡ T cells or CD4+CD25+ regulatory T cells was determined by chemotaxis assay. Data represents one of two individual experiments. (B) PuriWed populations of CD4+ T cells separated into CD4+CD25¡ T cells (left panel) or CD4+CD25+ regulatory T cell (right panel) populations were cultured with soluble anti-CD3 mAb (0.1 g/ml) with or without anti-CD28 mAb (5 g/ml) or CCL21 (2 g/ml). Data are expressed as the fold increase compared to cells cultured in anti-CD3 mAb and media. Data represent mean § SD of two separate experiments. (C) Suppression of the proliferation of CD4+CD25¡ T cells mediated by the addition of increasing numbers of CD4+CD25+ regulatory T cells is partially overcome by the presence of either CCL21 or anti-CD28 costimulation. Statistical signiWcance is compared to cells receiving no CCL21.

overcome CD4+CD25+ regulatory T cell mediated suppression. As shown previously, increased numbers of CD4+CD25+ regulatory T cells added to a Wxed number of CD4+CD25¡ T cells results in suppression of naive T cell expansion (Fig. 5C). However, in the presence of CCL21, the ability of CD4+CD25+ regulatory T cells to inhibit naive T cell proliferation was reduced, at least at higher ratios of naive to regulatory T cells. However, CCL21 did not overcome CD4+CD25+ T cell mediated suppression of T cell responses to the same extent that anti-CD28 costimulation did. 3.5. Physiological CCL21 costimulation requires LFA-1/ICAM-1 interactions To determine whether CCL21 would provide costimulation to T cells in response to antigen we exposed T cells to CCL21 in an allogeneic system. Immature DCs generated from BALB/c (H-2d) mice were incubated along with allogeneic T cells from C57BL/6 (H-2b) mice at a DC:T cell ratio of 1:10 in the presence or absence of CCL21. Addition of CCL21 resulted in signiWcantly enhanced proliferation of allogeneic T cells (Fig. 6). This data strongly supports the hypothesis that CCL21 functions to costimulate CD4+ and CD8+ T cells under physiologic conditions.

Fig. 6. CCL21 costimulates allogeneic T cell proliferation. T cells from C57BL/6 (H-2b) mice were incubated in the presence of the indicated neutralizing or control antibody and cultured in the presence of irradiated DC generated from the bone marrow of BALB/c (H-2d) mice at a DC:T cell ratio of 1:10 in the presence or absence of CCL21 for 5 days, and T cell proliferation was determined as described. Representative results of two separate experiments.

The primary downstream event of CCL21 signaling through CCR7 is activation of the integrin LFA-1, a critical adhesion molecule mediating cell arrest in the vasculature and immunological synapse formation following TCR engagement [26]. We conWrmed previous reports that CCL21, in conjunction with TCR stimulation results in increased surface expression of LFA-1 in

82

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

our system (data not shown). To determine if CCL21 could function as a costimulatory molecule in the absence of functional LFA-1/ICAM-1 interactions, we repeated the allogeneic reaction in the presence of neutralizing antibodies speciWc for these molecules. When LFA-1 or ICAM-1 was blocked, the costimulatory eVects of CCL21 were abolished, suggesting that CCL21 is incapable of costimulating T cells in the absence of LFA-1/ICAM-1 interactions, consistent with a recent report demonstrating that LFA-1 signaling lowers the threshold necessary for T cell activation [27].

4. Discussion We have previously found that local vaccination with a CCL21-expressing vaccinia virus induce a strong antitumor eVect dependent on CD4+ T cell responses [2]. While this could be explained, in part, on the known T cell chemotactic properties of CCL21, in this report we provide evidence that CCL21 also acts as a costimulatory molecule in promoting expansion of CD4+ and CD8+ T cells. In addition, CCL21 induces Th1 polarization of CD4+ T cells, at least under in vitro conditions. The profound eVects on CD4+ T cells prompted us to explore the eVects of CCL21 on CD4+CD25+ regulatory T cells. Although this subpopulation of CD4+ T cells is known to express lower levels of CCR7, we documented for the Wrst time that these cells were, in fact, hyporesponsive to CCL21-induced migration. Furthermore, in contrast to CD28 costimulation, CCL21 did not induce proliferation of CD4+CD25+ T cells. These Wndings suggest that CCL21 is not only involved in T cell and DC traYcking but may also play a pivotal role in the activation and diVerentiation of T cells. This possibility was strongly supported by the enhanced proliferation observed in allogeneic cultures (Fig. 6) and requires in vivo conWrmation. The observation that CD4+ T cells underwent Th1 polarization was particularly interesting, especially in light of the importance of CD4+ T cells in tumor models of local CCL21 administration [2,3]. The reasons for this are not clear, but it is interesting to speculate that this eVect may relate to subtle diVerences in CCR7 signaling by CCL19 and CCL21, which are both expressed in secondary lymphoid tissue. While both chemokines bind CCR7, slight conformational changes in CCR7 following binding by the diVerent chemokines results in diVerential T cell signaling [28]. In fact, T cell stimulation by CCL19, but not CCL21, results in receptor desensitization and greater activation of the Erk 1/2 pathway. This could provide a mechanism for regulating Th1/Th2 T cell diVerentiation by CCL19 and CCL21. Under physiologic conditions, it is likely that DCs are also important in orchestrating T cell diVerentiation and recently it has been suggested that CCL19–CCR7 interactions may be

critical for DC-T cell contact in lymph node tissue [29]. Thus, DC could inXuence the relative levels of CCL19 in contrast to the constitutive levels of CCL21 provided by the high endothelial venules, thus inXuencing T cell diVerentiation. The ability of CCL21 to speciWcally costimulate CD4+CD25¡ naïve T cells, but not CD4+CD25+ regulatory T cells, was another important observation. This is in contrast to IL-2 and anti-CD28 mAb, which stimulate expansion of both subsets of CD4+ T cells [30]. The physiologic consequences of this are not clear, but preferential costimulation of non-regulatory CD4+ T cells would provides a mechanism by which CCL21 could shift the ratio of naïve and regulatory T cells to favor T cell priming. CCL21 was able to partially overcome CD4+CD25+ T cell mediated suppression, at least when low numbers of regulatory T cells were present. This could explain how T cells may be primed in the lymphoid microenvironment despite the presence of regulatory T cells. The costimulatory eVects of CCL21 were diminished when lower ratios of naive to regulatory T cells were co-cultured (2:1 or lower), but these ratios likely represent non-physiological conditions since regulatory T cells typically account for only 5–10% of peripheral CD4+ T cells in normal hosts [31]. Thus, it is tempting to contemplate a mechanism whereby CD4+CD25+ regulatory T cells are not necessarily switched oV during T cell priming, but rather outcompteted by the preferential expansion of naïve T cells. The preferential migration of non-regulatory cells toward CCL21 may further ensure that the ratio of naive to regulatory T cells remains high in the T cell zones of the lymph node. The concentration of CCL21 in the lymph node has been estimated to be 11–12 g/ml, and is likely considerably higher within the T cell zones [20]. Our in vitro data were obtained with concentrations of CCL21 considerably lower than this, and it is possible that the eVects of CCL21 on inhibition of CD4+CD25+ regulatory T cells may be further exaggerated at physiological concentrations. Notably, CCL21 did not overcome CD4+CD25+ regulatory T cell mediated suppression to the same extent as anti-CD28 costimulation. Since both costimulatory pathways would be activated in the lymph node, it would be interesting to determine whether these pathways work in concert in overcoming the suppression of T cell responses mediated by CD4+CD25+ regulatory T cells. Although we deWned the direct eVects of CCL21 on T cells in vitro, the cumulative eVects of CCL21 costimulation in vivo are no doubt much more complex. The fact that CCL21 promoted allo-reactivity suggests that these eVects will be important in vivo and will likely depend on the integrated eVects of CCL21 on both T cell and DC. A likely mediator under these conditions is LFA-1, since CCL21 mediates LFA-1 aYnity and avidity on T cells and since LFA-1 is also required for immunological

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84

synapse formation between DC and T cells [11,26,28]. There is also a recent report demonstrating that LFA-1 signaling results in increased phosphorylation of the intracellular portion of the LFA-1 2 integrin chain in T cells, leading to IL-2 production and Th1 diVerentiation through the Erk 1/2 pathway [27]. In this context, it is possible that CCL21 mediates a costimulatory eVect through enhanced LFA-1 signaling, which promotes interactions between T cells and DC. However, our data implies that CCL21 can also mediate direct T cell signaling since puriWed populations of CD4+ or CD8+ T cells could proliferate and diVerentiate without APCs. These direct eVects on T cells may be related to CCR7 signaling of the JAK/STAT pathway as recently described, possibly explaining how CCL21 mediates T cell activation and cytokine release [12]. The costimulatory function of CCL21 has signiWcant implications for CCL21 as a therapeutic agent in priming T cells at extranodal sites. Local administration of CCL21 into established murine tumors has already been shown to induce signiWcant tumor regression associated with an increase in tumor inWltrating T cells [1,2,32]. The ability to induce migration of naïve CD4+ and CD8+ T cells, and expand such cells, at sites of tumor antigen expression provides a powerful tool for priming T cell responses in potentially immunosuppressed hosts [33]. The exclusion of CD4+CD25+ regulatory T cells from the local tumor microenvironment and their inability to respond to CCL21 may also provide a mechanism for tipping the balance back in favor of initiating an anti-tumor immune response. Further investigation is needed to fully deWne the eVects of CCL21 in vivo under normal physiologic conditions and in the face of active disease. Thus, in addition to the well deWned eVects on T cell migration, CCL21 also appears to be a potent T cell costimulatory molecule. The rapid kinetics of CCL21 mediated signaling within T cells suggests that CCL21 plays a major role in the early destiny of T cell immune responses.

[3]

[4] [5]

[6] [7]

[8] [9]

[10] [11]

[12]

[13]

[14]

[15]

Acknowledgments This work was supported by the National Institutes of Health Grant K08 79881 (to H.L.K.). The authors thank Robert T. Glover and Dr. Anne Davidson for critical review of the manuscript.

[16]

[17]

References [1] C.J. Kirk, D. Hartigan-O’Connor, B.J. NickoloV, J.S. Chamberlain, M. Giedlin, L. Aukerman, J.J. Mule, T cell-dependent antitumor immunity mediated by secondary lymphoid tissue chemokine: augmentation of dendritic cell-based immunotherapy, Cancer Res. 61 (2001) 2062–2070. [2] K. Flanagan, R.T. Glover, H. Hörig, W. Yang, H.L. Kaufman, Local delivery of recombinant vaccinia virus expressing secondary

[18]

[19]

83

lymphoid chemokine (SLC) results in a CD4 T cell dependent antitumor response, Vaccine 22 (2004) 2894–2903. C.J. Kirk, D. Hartigan-O’Connor, J.J. Mule, The dynamics of the T-cell antitumor response: chemokine-secreting dendritic cells can prime tumor-reactive T cells extranodally, Cancer Res. 61 (2001) 8794–8802. E.A. GreenWeld, K.A. Nguyen, V.K. Kuchroo, CD28/B7 costimulation: a review, Crit. Rev. Immunol. 18 (1998) 389–418. L.J. Appleman, D. Tzachanis, T. Grader-Beck, A.A. van Puijenbroek, V.A. Boussiotis, Helper T cell anergy: from biochemistry to cancer pathophysiology and therapeutics, J. Mol. Med. 78 (2001) 673–683. S.A. Luther, J.G. Cyster, Chemokines as regulators of T cell diVerentiation, Nat. Immunol. 2 (2001) 102–107. D.D. Taub, S.M. Turcovski-Corrales, M.L. Key, D.L. Longo, W.J. Murphy, Chemokines and T lymphocyte activation: I. Beta chemokines costimulate human T lymphocyte activation in vitro, J. Immunol. 156 (1996) 2095–2103. B. Moser, P. Loetscher, Lymphocyte traYc control by chemokines, Nat. Immunol. 2 (2001) 123–128. J.N. Blattman, R. Antia, D.J. Sourdive, X. Wang, S.M. Kaech, K. Murali-Krishna, J.D. Altman, R. Ahmed, Estimating the precursor frequency of naive antigen-speciWc CD8 T cells, J. Exp. Med. 195 (2002) 657–664. S. Sozzani, P. Allavena, A. Vecchi, A. Mantovani, Chemokines and dendritic cell traYc, J. Clin. Immunol. 20 (2000) 151–160. G. Constantin, M. Majeed, C. Giagulli, L. Piccio, J.Y. Kim, E.C. Butcher, C. Laudanna, Chemokines trigger immediate beta2 integrin aYnity and mobility changes: diVerential regulation and roles in lymphocyte arrest under Xow, Immunity 13 (2000) 759– 769. J.V. Stein, S.F. Soriano, C. M’Rini, C. Nombela-Arrieta, G.G. de Buitrago, J.M. Rodriguez-Frade, M. Mellado, J.P. Girard, A.C. Martinez, CCR7-mediated physiological lymphocyte homing involves activation of a tyrosine kinase pathway, Blood 101 (2003) 38–44. A. Poggi, R. Carosio, D. Fenoglio, S. Brenci, G. Murdaca, M. Setti, F. Indiveri, S. Scabini, E. Ferrero, M.R. Zocchi, Migration of V{delta}1 and V{delta}2 T cells in response to CXCR3 and CXCR4 ligands in healthy donors and HIV-1 infected patients: competition by HIV-1 TAT, Blood 103 (2003) 2205–2213. E. Scandella, Y. Men, D.F. Legler, S. Gillessen, L. Prikler, B. Ludewig, M. Groettrup, CCL19/CCL21-triggered signal transduction and migration of dendritic cells require prostaglandin E2, Blood 103 (2004) 1595–1601. J. Huehn, K. Siegmund, J.C. Lehmann, C. Siewert, U. Haubold, M. Feuerer, G.F. Debes, J. Lauber, O. Frey, G.K. Przybylski, U. Niesner, M. de la Rosa, C.A. Schmidt, R. Brauer, J. Buer, A. ScheVold, A. Hamann, Developmental stage, phenotype, and migration distinguish naive- and eVector/memory-like CD4+ regulatory T cells, J. Exp. Med. 199 (2004) 303–313. K. Inaba, M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, R.M. Steinman, Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor, J. Exp. Med. 176 (1992) 1693–1702. C.H. Kim, L.M. Pelus, J.R. White, E. Applebaum, K. Johanson, H.E. Broxmeyer, CK beta-11/macrophage inXammatory protein-3 beta/EBI1-ligand chemokine is an eYcacious chemoattractant for T and B cells, J. Immunol. 160 (1998) 2418–2424. S.A. Kellermann, S. Hudak, E.R. Oldham, Y.J. Liu, L.M. McEvoy, The CC chemokine receptor-7 ligands 6Ckine and macrophage inXammatory protein-3 beta are potent chemoattractants for in vitro- and in vivo-derived dendritic cells, J. Immunol. 162 (1999) 3859–3864. M.D. Gunn, K. Tangemann, C. Tam, J.G. Cyster, S.D. Rosen, L.T. Williams, A chemokine expressed in lymphoid high endothelial

84

[20]

[21] [22]

[23]

[24]

[25]

[26] [27]

K. Flanagan et al. / Cellular Immunology 231 (2004) 75–84 venules promotes the adhesion and chemotaxis of naive T lymphocytes, Proc. Natl. Acad. Sci. USA 95 (1998) 258–263. S.A. Luther, A. Bidgol, D.C. Hargreaves, A. Schmidt, Y. Xu, J. Paniyadi, M. Matloubian, J.G. Cyster, DiVering activities of homeostatic chemokines CCL19, CCL21, and CXCL12 in lymphocyte and dendritic cell recruitment and lymphoid neogenesis, J. Immunol. 169 (2002) 424–433. M. Moser, K.M. Murphy, Dendritic cell regulation of TH1–TH2 development, Nat. Immunol. 1 (2000) 199–205. S. Oki, N. Otsuki, T. Kohsaka, M. Azuma, Stat6 activation and Th2 cell diVerentiation [correction of proliferation] driven by CD28 [correction of CD28 signals], Eur. J. Immunol. 30 (2000) 1416–1424. T. Takahashi, T. Tagami, S. Yamazaki, T. Uede, J. Shimizu, N. Sakaguchi, T.W. Mak, S. Sakaguchi, Immunologic self-tolerance maintained by CD25(+)CD4(+) regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4, J. Exp. Med. 192 (2000) 303–310. A.M. Thornton, E.M. Shevach, CD4+CD25+ immunoregulatory T cells suppress polyclonal T cell activation in vitro by inhibiting interleukin 2 production, J. Exp. Med. 188 (1998) 287–296. T. Takahashi, S. Sakaguchi, Naturally arising CD25+CD4+ regulatory T cells in maintaining immunologic self-tolerance and preventing autoimmune disease, Curr. Mol. Med. 3 (2003) 693–706. M.L. Dustin, The immunological synapse, Arthritis Res. 4 (Suppl. 3) (2002) S119–S125. O.D. Perez, D. Mitchell, G.C. Jager, S. South, C. Murriel, J. McBride, L.A. Herzenberg, S. Kinoshita, G.P. Nolan, Leukocyte

[28]

[29]

[30]

[31]

[32]

[33]

functional antigen 1 lowers T cell activation thresholds and signaling through cytohesin-1 and Jun-activating binding protein 1, Nat. Immunol. 4 (2003) 1083–1092. T.A. Kohout, S.L. Nicholas, S.J. Perry, G. Reinhart, S. Junger, R.S. Struthers, DiVerential desensitization, receptor phosphorylation, beta arrestin recruitment, and ERK1/2 activation by the two endogenous ligands for the CC chemokine receptor 7, J. Biol. Chem. 279 (2004) 23214–23222. F. Katou, H. Ohtani, T. Nakayama, H. Nagura, O. Yoshie, K. Motegi, DiVerential expression of CCL19 by DC-Lamp+ mature dendritic cells in human lymph node versus chronically inXamed skin, J. Pathol. 199 (2003) 98–106. K. Nakamura, A. Kitani, W. Strober, Cell contact-dependent immunosuppression by CD4(+)CD25(+) regulatory T cells is mediated by cell surface-bound transforming growth factor beta, J. Exp. Med. 194 (2001) 629–644. S. Sakaguchi, Naturally arising CD4+ regulatory T cells for immunologic self-tolerance and negative control of immune responses, Annu. Rev. Immunol. 22 (2004) 531–562. A.P. Vicari, S. Ait-Yahia, K. Chemin, A. Mueller, A. Zlotnik, C. Caux, Antitumor eVects of the mouse chemokine 6Ckine/SLC through angiostatic and immunological mechanisms, J. Immunol. 165 (2000) 1992–2000. S. Sharma, M. Stolina, J. Luo, R.M. Strieter, M. Burdick, L.X. Zhu, R.K. Batra, S.M. Dubinett, Secondary lymphoid tissue chemokine mediates T cell-dependent antitumor responses in vivo, J. Immunol. 164 (2000) 4558–4563.