Does the IL-2 receptor α chain induced on dendritic cells have a biological function?

The IL-2 receptor (IL-2R) α chain (CD25), but not the IL-2R β chain, is induced on dendritic cells (DC) by brief periods of culture. To test if this IL-2Rα is important for DC function, DC were isolated from the spleens of mutant mice with the IL-2Rα gene disrupted and compared with normal DC for ability to stimulate proliferation of allogeneic CD4 and CD8 T cells in culture. The IL-2Rα null DC and the normal DC produced nearly identical proliferative responses from CD4 and from CD8 T cells. When the CD8αF and CD8α– subsets of the IL-2Rα null DC were separated, they also produced proliferative responses similar to that of their normal DC counterparts. Overall there was no evidence that the inducible IL-2Rα on DC was required for DC development, for stimulation of T cells or for regulation of T cell responses. Dendritic cells (DC) are sparsely but widely distributed bone marrow-derived cells whose function is to present peptide antigens to T lymphocytes (1). After a developmental phase when they collect and process antigens, DC move to lymphoid tissue where they mature to a form specialized for T cell activation. The DC of lymphoid tissues have cytoplasmic extensions to maximize T cell contact, they express adhesion molecules to bind to T cells, express high levels of MHC molecules to optimize antigen presentation and express costimulator molecules (such as CD80 and CD86) to ensure T cell activation (2). The signal exchange is two-way, with T cells expressing the ligand to activate DC via CD40 and with T cells producing cytokines such as granulocyte macrophage colony stimulating factor (GM-CSF). We have recently shown that certain subgroups of DC also regulate the response of the T cells they activate. A population of CD8α1 mouse spleen DC expresses the Fas ligand and kills the activated CD4 T cells by Fas-mediated apoptosis (3). The same ‘regulatory’ CD8α1 DC subset induces only a limited proliferative response from CD8 T cells, compared to that induced by CD8α– DC, as a result of restricted endogenous IL-2 production (4). It is likely that a number of as yet unidentified signaling systems are involved in DC–T cell interactions. One molecule of interest that may be induced at high levels on DC is CD25, the α-chain of the IL-2 receptor (IL-2R). After incubation in culture, CD25 can be detected on mouse splenic (5), thymic (6) and lung (7) DC as well as on mouse epidermal Langerhans’ cells (8), on rat thymic (9,10), lung (11) and Correspondence to: V. Kronin Transmitting editor: A. Kelso Received 18 July 1997, accepted 4 November 1997 afferent lymph (12) DC, and on human blood DC (13). Despite these repeated observations there is no data on the functional significance of CD25 on DC. In the present paper we therefore ask whether CD25 on DC affects their interaction with T cells. First, we determined whether the IL-2R induced on DC was the low-affinity αα homodimer or the high-affinity IL-2R which is normally formed with the participation of β and γ chains (14). While the evidence showing the expression of the α chain of IL-2R on DC (5–12) is convincing, there is no published data on expression of the IL-2R β chain. Accordingly, DC were purified from mouse spleens, thymuses and lymph nodes (LN), and stained using conjugated mAb against IL-2Rα or IL-2Rβ, either immediately after purification or after 16 h incubation of DC in culture with GM-CSF. The expression of IL-2Rα and IL-2Rβ on the DC is shown in Fig. 1. The IL2R α chain was generally very low on most DC before incubation, but the expression increased strikingly after overnight incubation with GM-CSF. These results confirm those obtained by others (5–12), with extension of the data to LN DC. The expression of IL-2Rα was induced on both CD81 and CD8– populations of splenic DC, as we have reported elsewhere (2). In marked contrast, mouse DC did not show any expression of IL-2Rβ, neither directly after enrichment from mouse lymphoid organs nor after overnight incubation in culture in the presence of GM-CSF (Fig. 1). E. Kämpgen (pers. commun.) has also noted the absence of IL-2Rβ on murine splenic DC and cultured Langerhans cells. In the absence of the β chain, the IL-2R should be only a 238 Role of CD25 in dendritic cell function Fig. 1. Expression of IL-2Rα (CD25) and IL-2Rβ (CD122) on the surface of DC. DC were isolated based on the method described in detail previously (17). Briefly, the lymphoid organs were pooled from eight C57BL/6 mice and cut into fragments, after which collagenase digestion followed by EDTA treatment was applied to release DC from lymphoid tissues. Low-density cells were then enriched by density centrifugation, followed by removal of non-DC lineages using the immunomagnetic bead depletion. Samples were taken at that stage to determine the initial surface expression of IL-2Rα and IL2Rβ. The remaining cells were incubated for 16 h in a modified RPMI 1640/10% FCS medium supplemented with 200 U/ml GM-CSF (Immunex, Seattle, WA). The cells (either before or after incubation) were stained with: allophycocyanin-conjugated anti-MHC class II, M5/114 (19); fluorescein-conjugated anti-CD8α, 53-6.7 (20); and either phycoerythrin-conjugated anti-IL-2Rβ, TM-β1 (21), or biotinylated anti-IL-2Rα, PC/61 (22), followed by phycoerythrinconjugated avidin as second stage. Propidium iodide was included to stain dead cells. The samples were then analyzed on modified dual-laser FACS II (Becton Dickinson, San Jose, CA). The cytometer was calibrated using CaliBRITE beads (Becton Dickinson) and setting of the calibration points was done by eye. The figure shows the distribution of CD25 and CD122 on DC either before or after incubation. DC were gated as class II MHC highly positive cells with high forward and side scatter; propidium iodide-positive cells were excluded. The results are typical of three experiments conducted. low-affinity receptor (Kd 5 3.2 nM, t1/2(association) 5 5 s, t1/2(disassociation) 5 35 s), unable to transmit signals into the DC itself (14). However, it seemed possible that this lowaffinity IL-2R on DC after activation could serve to concentrate and focus IL-2, passing it to the high-affinity (Kd 5 2.5 pM, t1/2(association) 5 20–40 s, t1/2(disassociation) 5 255 min) (14), complete IL-2R induced on the activated T cells of the DC–T cell interaction complex. We used mutant mice deficient in the α chain of the IL-2R (15) to assess such potential functional roles. Before using IL-2Rα null DC in tests of T cell activation we asked whether the absence of IL-2Rα affected the development of DC. DC were purified from the spleen, then stained and counted to determine if normal numbers of both CD81 Table 1. Numbers of DC in the spleens of normal and IL2Rα–/– mice DC per spleen (310–3) Control C57BL/6 IL-2Rα–/– CD81 DC 227 6 1 370 6 57 CD8– DC 117 6 7 800 6 258 DC were isolated from the spleens of normal C57BL/6 and IL2Rα–/– ‘knockout’ mice as described previously (17) and in Fig. 1. The enriched DC were stained with fluorescein-conjugated antiCD11c, N418 (18), phycoerythrin-conjugated anti-CD8α, 53-6.7, and propidium iodide to stain dead cells. The numbers of CD81 and CD8– DC were then calculated as the numbers per mouse of gated CD11c1CD81 cells or CD11c1CD8– cells with high forward and side scatter with propidium iodide-positive cells excluded. Both populations expressed similar high levels of CD11c. Results are the mean of three experiments (6SEM). and CD8– DC were present in the spleens of the IL-2Rα knockout mice. Both CD8α1 and CD8α– DC subsets were present (Table 1) and their light-scatter characteristics (higher for CD8α1 DC than CD8α– DC) were identical with those from the normal control mice (data not shown). However, the total numbers of DC, especially of CD8– DC, were increased in spleens of IL-2Rd null mice, although there was substantial variability (Table 1). This increase correlated with an increase in spleen size up to 3-fold, but variable, in IL-2Rα–/– mice at this age (15). Such changes in total and relative DC numbers may be a secondary consequence of the marked effects of the mutation on T cells, perhaps including the production of GM-CSF. However, despite the disturbances to the spleens in the IL-2R–/– mice, both types of DC were present, so we could test for changes in their interaction with T cells. We first investigated the ability of the IL-2Rα null DC to stimulate allogeneic CD4 T cells. Because the ability of certain DC subsets to kill CD4 T cells by Fas-mediated apoptosis (3) could confuse the analysis of T cell activation and expansion, we used Fas-deficient T cells to eliminate these complications caused by T cell death (4). CD4 T cells were purified from the LN of young C3H.lpr mice and stimulated in culture with DC purified from the spleens of either normal C57BL/6 or IL2Rα–/– C57BL/6 mice; the resultant T cell proliferation was measured by [3H]thymidine uptake. As shown in Fig. 2, the DC from the IL-2Rα–/– mice induced a CD4 T cell response that was identical to that produced by DC from normal mice. This result was obtained over the DC range from 500 to 2000 per culture, indicating that the presence of the IL-2Rα chain did not promote T cell proliferation even when DC were limiting. This equivalence of stimulatory ability between IL2Rα null and normal DC was confirmed in a subsequent experiment where the CD8α– and CD8α1 DC subsets were separated and tested individually (data not shown). Overall these cultures provided no evidence that the IL-2Rα on DC played a role in CD4 T cell activation. We then tested the ability of the DC from the IL-2Rα–/– mice to stimulate allogeneic CD8 T cells. In view of the IL-2related differences in T cell proliferation late in culture when CD8 T cells were stimulated with CD8α– versus CD8α1 DC Role of CD25 in dendritic cell function 239 Fig. 2. Comparison of the proliferation of CD4 T cells stimulated by allogeneic splenic DC isolated from normal or IL-2Rα–/– ‘knockout’ mice. DC were enriched from spleens of normal C57BL/6 or IL2Rα–/– mice as described in Fig. 1 and Table 1, then stained with: fluorescein-conjugated anti-CD11c, N418; propidium iodide. The CD11c1 DC were then sterile sorted, using also the forward and side scatter and propidium iodide exclusion gates as in Table 1. The sorted DC were .95% pure on reanalysis. CD4 T cells were isolated from LN of C3H.lpr mice by immunomagnetic bead depletion as described previously (3), depleting for CD81 T cells, B cells, MHC class II1 cells, erythroid cells and any cells bearing CD44. The resulting population contained at least 98% CD41 cells. Replicate DC T cell cultures were set up and the proliferative response determined as described in detail previously (3). Briefly, 23104 CD4 T cells were cultured with 1000 DC in the wells of 96-well V-bottomed plates in 200 ml of a modified RPMI 1640/10% FCS medium, at 37°C, in a 10% CO2-in-air incubator. Proliferation of CD4 T cells was assessed by [3H]thymidine incorporation. Then, 1 μCi [3H]thymidine was added per well and after a 6 h pulse the content of each well was transferred to glass fiber filters. The levels of [3H]thymidine incorporation were then measured in a gas scintillation β-counter. The figure shows the mean of the pooled data from two experiments (6SEM) each with three cultures per point. Cultures with T cells alone, or DC alone, gave ,100 c.p.m. Stimulation indices at the peaks of proliferation were .100. from normal mice, these subsets were both isolated from the spleens of normal C57BL/6 or IL-2Rα–/– C57BL/6 mice and tested separately for stimulatory ability. Since CD8 T cells were not susceptible to killing by FasL on DC (4), CD8 T cells purified from LN of normal CBA mice were used as responder cells. The proliferative responses, measured by [3H]thymidine uptake, are shown in Fig. 3. The DC purified from the IL2R–/– mice were as good at stimulating CD8 T cells as those from normal mice. This result was obtained with both the CD8α1 and CD8α– subpopulations. From these culture results there is no evidence that IL-2Rα on DC plays a role in the activation of CD8 T cells nor any evidence that it was involved in the differences between CD8α– and CD8α1 DC in their ability to maintain CD8 T cell proliferation late in culture. We have found that the induction of IL-2Rα (CD25) without IL-2Rβ is a feature common to most mouse DC when cultured with GM-CSF. We have also found that the expression of IL2Rα is induced on ~40% of splenic DC when cultured with allogeneic T cells in mixed leukocyte cultures similar to those in which function was tested in Figs 2 and 3 (data not shown). This induction of IL-2Rα is also likely to occur naturally in vivo, as DC are activated along with T cells in the DC–T cell interaction complex. However, our culture studies have not revealed any deficit in DC function if DC are genetically Fig. 3. Comparison of the proliferation of CD8 T cells stimulated by allogeneic splenic DC isolated from normal or IL-2Rα–/– ‘knockout’ mice. The procedure was similar to that described in Fig. 2, except that CD8 T cells were derived from CBA mice and were purified using anti-CD4 instead of anti-CD8 mAb in the immunomagnetic bead depletion procedure. DC were stained with fluoresceinconjugated anti-CD11c, N418; phycoerythrin-conjugated anti-CD8α, 53-6.7; propidium iodide. The DC were then sorted into CD11c1CD81 and CD11c1CD8– fractions and used at dose of 500 DC per well. The figure shows the mean of the pooled data from two experiments (6SEM) each with three cultures per point. T cells cultured alone, or DC cultured alone, gave ,100 c.p.m. Stimulation indices for CD8– DC were .100 at day 3.5. unable to produce IL-2Rα. Accordingly it is possible that the expression of IL-2Rα is a developmental ‘accident’, perhaps being activated in a co-ordinate manner with some other gene which is essential, and that IL-2Rα itself serves no useful role at all. It is equally possible that the subtleties of the function of IL-2Rα on DC have escaped us. In particular, since the levels of IL-2 available in a closed culture system may be higher and be maintained longer than in the local environment in lymphoid tissue, the ability of DC to ‘focus’ IL-2 via CD25 could have a small selective advantage in vivo. Such a marginal advantage would be difficult to test. Another potential role for the induced IL-2Rα is focusing IL-2 for the IgM production by B cells, or for growth of B cells, since a subgroup of human DC has been found to enhance these aspects of CD40-activated B cell function (16). Finally, it is conceivable that an IL-2 IL-2Rα interaction is in some way involved in the uptake or processing of exogenous antigens, an aspect our mixed leukocyte cultures do not address. At present, however, the only experimental evidence available and presented here suggests that the IL-2R α chain is not an essential component of the DC surface, not being required for either DC development or DC–T cell interaction. 240 Role of CD25 in dendritic cell function