L-selectin has become established as a key molecule in the recirculation of naïve T cells from the blood to peripheral lymph nodes, yet little is known about its role in the migration of effector or memory cells. While differentiating naïve CD4+ T cells into Th1 and Th2 subsets in vitro, it was noted that L-selectin levels were maintained on the Th1 subset of cells. The expression of L-selectin on the Th1 cells appeared to be dependent on the presence of IL-12. Th2 cells, differentiated in the absence of IL-12, failed to maintain L-selectin expression. Coculture with IL-12, IL-18, IL-4, TNF-α, or IFN-α, -β, or -γ demonstrated a dependence on IL-12 alone for L-selectin expression. In addition, the inclusion of heat-killed Listeria monocytogenes in the cultures also maintained L-selectin expression on the Th1 cells. In all cultures, the maintenance of L-selectin on the T cell surface could be blocked by the inclusion of anti-IL-12 Abs. Analysis of the mRNA levels for L-selectin in T cells, differentiated in the presence or absence of IL-12, showed that the cytokine appears to exert its effect on L-selectin at the transcriptional level. Given the key role played by IL-12 in the differentiation of naïve T cells into the Th1 subset, the observation that IL-12 can also regulate L-selectin expression has implications for the migration of Th1 effector cells both through the lymphatic system and to sites of inflammation.
The activation of naïve T cells, with either specific Ag or polyclonal stimuli, results in the differential regulation of a series of cell adhesion/activation molecules. Cells transiently up-regulate a range of activation markers, such as CD25 (IL-2R), CD40 ligand (CD40L)2, LFA-1, phagocytic glycoprotein-1 (CD44), and CD69 (1, 2). In contrast, CD45RB and L-selectin (CD62L) are thought to decrease on T cell activation (3, 4, 5, 6, 7).
L-selectin is classically defined as the lymphocyte homing receptor and plays a key role in the migration of naïve T cells through the lymph nodes (8, 9, 10, 11). Naïve T cells that have not yet been stimulated by Ag express high levels of L-selectin (4, 7). These cells exit the thymus and circulate through peripheral lymph nodes via the high endothelial venules (HEV) (12, 13, 14). Ligation of L-selectin by ligands borne on the lymph node HEVs is essential for the initial arrest of the lymphocyte (15, 16). Firm attachment, mediated by the β2 integrins, is followed by transmigration from the blood stream into the lymph node paracortex (17, 18, 19). Antigenic stimulation of the naïve cells occurs within the lymphoid tissue, where the appropriate APC presents processed Ag complexed with MHC molecules to the T cell. Once the T cell has been activated to become an effector or memory cell, its homing potential will allow for extravasation at extra lymphoid sites, such as the skin, the lamina propria of the intestine, and inflamed joints. This is reflected in the up-regulation of a new array of adhesion molecules and homing receptors (20, 21, 22). It is generally considered that, once they have become activated, T cells shed L-selectin, preventing the subsequent recirculation of effector cells through the lymph nodes (7). However, recent data has shown that memory T cells can indeed recirculate in an L-selectin-dependent manner (23). This would suggest that L-selectin can be borne on activated T cells and is not solely expressed by naïve cells.
Naïve and memory CD4+ T cells mediate quantitatively and qualitatively distinct types of immune response. These are manifested by differences in cytokine secretion, adhesive properties, and activation requirements (2, 24, 25, 26). The generation of effector Th cells, of which the Th1 and Th2 subsets represent polarized phenotypes, is determined by several factors. T cells can be committed to the Th1 lineage in vitro by activation in the presence of IL-12, IFN-γ, and anti-IL-4, while Th2 cells are generated in the presence of IL-4, anti-IL-12, and anti-IFN-γ (27, 28, 29, 30). These polarized effector populations can subsequently be characterized by the cytokines that they secrete (Th1: TNF-β, IFN-γ, and IL-2; and Th2: IL-4, IL-5, IL-6, IL-10, and IL-13) (31). Th1 cells will go on to mediate cellular immunity, classically delayed-type hypersensitivity, while Th2 cells are most active as helpers for Ab production. The generation of Th1 cells is dependent on the presence of IL-12. This cytokine is produced by monocytes, dendritic cells, and macrophages in response to bacterial products and bacteria such as Listeria monocytogenes (32).
Here, we describe the expression of L-selectin on Ag-stimulated mouse Th1 cells. Analysis of the cell culture environment revealed that the maintenance of L-selectin expression on Th1 cells was dependent on the inclusion of IL-12. L-selectin expression could not be maintained by the inclusion of IL-4, IL-18 (IFN-γ-inducing factor), or the IFNs (α, β, or γ) and could be abrogated by the inclusion of anti-IL-12 Abs. In addition, the production of endogenous IL-12 within the culture, through the inclusion of heat-killed L. monocytogenes, could also maintain L-selectin expression. With the key role played by L-selectin in T cell recirculation, it is possible that this increased expression of L-selectin on Th1 cells will affect the ability of these cells to migrate through lymphoid and extra lymphoid sites.
Materials and Methods
Mice transgenic for the DO11.10 αβ TCR (33) specific for the OVA peptide (OVA323–339) were maintained on a BALB/c genetic background. Heterozygous mice were bred with BALB/c mice and bled at 4 wk of age. The RBC were lysed, and the remaining cells were stained with Vβ8 PE and CD4+ FITC. Mice, at least 70% positive for Vβ8 expression, were maintained in isolators. BALB/c mice (6–8 wk old) were purchased from Charles River Breeding Laboratories (Margate, U.K.).
Preparation of CD4+ T cells
Single cell suspensions isolated from DO11.10 spleens were pooled and the RBC lysed. Splenocytes were purified on a CD4 T subset column (R&D Systems, Abington, U.K.) according to the manufacturer’s instructions. Briefly, 2 × 108 cells were resuspended in 1 ml wash buffer and treated with 1 ml Ab mixture for 15 min at room temperature. The cells were washed and placed onto a CD4 T subset column in 1 ml buffer. After 10 min at room temperature, a CD4+-enriched cell population (85% naïve cells) was eluted. Naive CD4+ CD45RB+ cells (99% positive) were sorted by flow cytometry using a Coulter (Luton, U.K.) EPICS ELITE flow cytometer.
Preparation of T cell-depleted APCs
Single cell suspensions of sex-matched BALB/c spleens were pooled and the RBC lysed. Splenocytes were treated with a mixture of anti-CD8-biotin, anti-CD4-biotin, and anti-Thy.1-biotin (PharMingen, San Diego, CA) at 0.5 μg Ab/106 cells for 20 min on ice. Cells coated with Ab were magnetically removed using streptavidin-coated Dynabeads (Dynal, Bromborough, U.K.) at a ratio of 4 beads/cell. The T cell-depleted splenocytes were irradiated with 30 Gy.
T cell differentiation
T cells were differentiated in vitro by Ag stimulation (34). Stimulations were performed using 1 μM OVA peptide (323ISQAVHAAHAEINEAGR339) with 0.5 × 106 naïve CD4+ DO11.10 cells and 2 × 106 APC. Cells were maintained in RPMI supplemented with 10% (v/v) FCS, 10 mM l-glutamine, 200 U/ml penicillin, 200 U/ml streptomycin, and 0.5 μM 2-ME (Sigma, Poole, U.K.). After 7 days, the cells were washed and restimulated with peptide in the same manner. For differentiation into Th1 cells, IL-12 (10 ng/ml), IFN-γ (1000 U/ml), and anti-IL-4 (10 μg/ml) were added to the cultures; for Th2 cells, IL-4 (10 ng/ml), anti-IL-12 (10 μg/ml), and anti-IFN-γ (10 μg/ml) were added (35). All cytokines and anti-cytokine Abs were purchased from R&D Systems, with the exception of IFN-γ (Serotec, Oxford, U.K.). IFN-α and -β and heat-killed L. monocytogenes (all kindly donated by G. Bancroft, London School of Tropical Hygiene and Medicine, London, U.K.) were added to cultures at the concentrations indicated in the text (36). Cytokine secretion into the culture supernatant was measured by ELISA using Ab pairs, according to the manufacturer’s instructions (PharMingen).
Flow cytometry and intracellular cytokine staining
Cells were removed from Ag-stimulation cultures at the times indicated in the text and washed in FACS buffer (PBS containing 0.1% (w/v) BSA, 0.01% (w/v) sodium azide). Staining of 2 × 105 cells was routinely performed in round-bottom 96-well plates (Dynex Technologies, Billingshurst, U.K.) with FITC- or PE-conjugated anti-CD4 (H129.19), anti-CD62L (Mel-14), anti-IL2Rα (7D4), and anti-Vβ8 (MR5) (PharMingen) for 20 min on ice. All FACS analyses were performed using a Coulter EPICS XL-MCL flow cytometer. Results are expressed either in terms of the percentage of positive cells based on positioning of a marker at 2% of a FITC- or PE-conjugated isotype control Ab or median fluorescence intensity.
Intracellular cytokine staining was performed as described (37). Differentiated cells were stimulated with 50 ng/ml PMA and 500 ng/ml ionomycin (Sigma) at 1 × 106 cells/ml for 4 h at 37°C. A total of 10 μg/ml Brefeldin A (Sigma) was added for the final 2 h. Cells were stained with anti-CD4-Quantum red (Sigma) before fixation using a “Fix and Perm” cell permeabilization kit (Caltag, South San Francisco, CA). After washing the cells were permeabilized and incubated with anti-IFN-γ-FITC (0.5 μg), anti-IL-4-PE (1.0 μg), or control Abs (PharMingen). Cells were gated on the CD4-Quantum red fluorescence before analysis of intracellular cytokine expression.
RNA extraction and Northern blotting
Total RNA was extracted from DO11.10 CD4+ cells that had been Ag-stimulated in the presence of either 10 ng/ml IL-12 or in the absence of IL-12. Cells (1 × 107) were harvested by centrifugation at given time points, washed in PBS, and the cell pellet lysed. Total RNA was extracted using an RNeasy kit (Qiagen, Crawley, U.K.). RNA (2 μg/lane) was electrophoresed in 1% (w/v) agarose gels containing 6% (v/v) formaldehyde at 50 V for 4 h and was blotted on nitrocellulose (Boehringer Mannheim, Lewes, U.K.) by capillary action. Nitrocellulose membranes were probed with an L-selectin digoxygenin (DIG)-labeled riboprobe (40 ng/ml) for 15 h at 68°C. The L-selectin riboprobe was prepared by in vitro transcription from L-selectin cDNA (kindly provided by D. Haskard, Royal Postgraduate Medical School, Hammersmith, U.K.) and DIG-labeled using a DIG RNA labeling kit (Boehringer Mannheim). After probing for L-selectin mRNA, the blot was stripped with 0.2 M NaOH and reprobed for β-actin. Although the levels of β-actin were seen to change during cell activation, this control was included to ensure equal loading of RNA in each sample at the given time points. Northern blots were developed using an anti-DIG-HRP conjugate (Boehringer Mannheim) and visualized using enhanced chemiluminescence (Amersham, Little Chalfont, U.K.).
The variation observed between experiments was calculated as the SEM at each data point. Differences in expression levels of L-selectin between two different cell populations were calculated using Students’s paired t test. Significance was assumed if p < 0.001.
Expression of L-selectin is maintained on Th1 cells but not Th2 cells
Polarized Th1 and Th2 cells were generated in vitro by the stimulation of naive CD4+ T cells from OVA-specific, I-Ad-restricted TCR transgenic mice (DO11.10) with OVA peptide (OVA) in the presence of cytokines and anti-cytokine Abs (Th1: 10 ng/ml IL-12, 1000 U/ml IFN-γ, and 10 μg/ml anti-IL-4; Th2: 10 ng/ml IL-4, 10 μg/ml anti-IL-12, and 10 μg/ml anti-IFN-γ). After culture for 7 days in a primary stimulation by Ag, the phenotype of the differentiated cells was investigated using intracellular cytokine staining. Fig. 1 shows the intracellular cytokine staining profiles for the Th1 and Th2 cells. Clearly, the majority of the cells, cultured under conditions designed to produce Th1 cells, secrete IFN-γ, a characteristic of Th1 cells. In contrast, the Th2 cells synthesized high levels of IL-4. These data confirm that Th1- and Th2-like cells have been generated in vitro. Following a secondary stimulation with peptide, the cells were stained for surface L-selectin expression. It was found that the Th1 subset continued to express L-selectin during Ag stimulation with OVA peptide, while the Th2 subset did not (Fig. 2). The same pattern of L-selectin expression is also observed when cells are stimulated with polyclonal stimuli, such as anti-CD3 and PMA/ionomycin (data not shown).
IL-12 maintains L-selectin on activated T cells
Since the T cells were cultured in a defined cytokine environment, the control of L-selectin expression could be readily investigated. IFN-γ shares many functional properties with IL-12 and is dependent on IL-12 for its induction in NK and T cells (38, 39). Since high levels of IFN-γ were synthesized by the Th1 cultures (Fig. 1) that were generated in the presence of IL-12, it was important to establish whether IFN-γ or IL-12 or a combination of both cytokines could be mediating the regulation of L-selectin on the cell surface. Consequently, cells were cultured with combinations of IL-12 and IFN-γ, and anti-IL-12 and anti-IFN-γ. Fig. 3 shows that L-selectin expression was maintained only when IL-12 was available to the cells. The addition of IFN-γ alone did not maintain L-selectin on the cell surface (Fig. 3,A). When IL-12 was absent, or neutralizing Abs to IL-12 were present, L-selectin expression was lost (Fig. 3,B). The slight decrease in the levels of L-selectin observed in the presence of anti-IFN-γ, in combination with IL-12 in the culture (Fig. 3,A), coupled with the increased levels of expression seen on inclusion of both IL-12 and IFN-γ in the primary stimulation (Fig. 3 B), suggest a slight synergistic effect of the two cytokines in the regulation of L-selectin expression.
Recent reports have suggested a synergy between IL-12 and IL-18 on the development of Th1 cells (36). In addition, the effects of IFN-α and -β on the regulation of human L-selectin have recently been described (40). Consequently, the effect of these cytokines on mouse L-selectin expression was investigated. In contrast to the regulation of human L-selectin, no maintenance of mouse L-selectin expression was observed when IFN-α, -β, or IL-18 were included in the culture medium during antigenic stimulation in the absence of IL-12 (Fig. 4,A). Although IL-18, IFN-α, and -β had no effect on L-selectin expression, they were all found to promote IL-2Rα expression (Fig. 4,B) and to induce the synthesis of IFN-γ (Fig. 4 C), confirming their functionality.
The effect of IL-12 on the maintenance of L-selectin expression was found to be dose-dependent (Fig. 5), with the levels of cell surface L-selectin strongly associated with the concentration of available IL-12 in the culture (Student’s t test, p < 0.001). A number of other cell surface markers were examined, including CD4, CD11a, CD11b, CD18, CD25 (IL-2R), CD40, CD43, CD45RB, CD48, CD54, CD69, and phagocytic glycoprotein-1 (data not shown). Only L-selectin and the IL-2R were maintained on the T cell surface while differentiating Th1 cells in the presence of IL-12.
The effect of IL-12 on L-selectin expression was also observed using Ag-stimulated CD8+ cells and on both CD4+ and CD8+ cells, isolated from both BALB/c and C57BL/6 mice, when activated with low levels of polyclonal stimuli (data not shown). In all cases, the inclusion of IL-12 prevented the loss of L-selectin from the cell surface.
Addition of heat-killed L. monocytogenes to cultures also maintains L-selectin expression
In addition to using cytokines to influence cellular responses directly, microorganisms can also be used in the cell culture to bias the phenotype toward Th1 or Th2 cell populations. For example, heat-killed Listeria is known to promote IL-12 release from BALB/c macrophages (32). When heat-killed Listeria (107 CFU/ml) was included in the culture, up to 7 ng/ml IL-12 was endogenously produced by the APCs within the culture in the response to the bacterium (data not shown). Fig. 6 demonstrates that the addition of heat-killed Listeria to the cell culture has the same effect as the addition of IL-12 in maintaining the L-selectin expression on the T cell surface. As with the exogenously added IL-12, the effect of the heat-killed Listeria can also be blocked by anti-IL-12.
Transcriptional regulation of L-selectin by IL-12
The observed modulation of L-selectin on the cell surface by IL-12 allowed the investigation of possible mechanisms for the regulation of L-selectin by IL-12. Cell cultures were established, as described previously, in the presence (10 ng/ml) or absence of IL-12 throughout a primary and secondary response. Seven days after the primary or secondary stimulation, RNA was extracted from the cells and probed for L-selectin mRNA using a DIG-labeled riboprobe. As seen in Fig. 7, the levels of L-selectin mRNA correlate with the presence of IL-12 in the culture. High levels of L-selectin mRNA are seen in the naïve cells. These levels are maintained through the primary stimulation, although more L-selectin mRNA is apparent in the cells cultured in the presence of IL-12. Following the secondary stimulation, the results are more striking with high levels of L-selectin mRNA observed in the cells treated with IL-12 and little detectable mRNA in the cells cultured in the absence of the cytokine. Some L-selectin-specific mRNA was still detectable, but this was significantly reduced compared with levels of mRNA from cells cultured with IL-12. Although the levels of β-actin mRNA are known to change during the stimulation of T cells, the measurement of β-actin mRNA was included as a comparative control.
While studying the differentiation of naïve T cells into Th1 and Th2 subsets, it was noted that the Th1 subset retained a high level of L-selectin expression. Analysis of the cell culture components indicated a role for IL-12 in the regulation of L-selectin expression on the T cell surface. L-selectin expression could be maintained either by direct addition of IL-12 to the cultures or through the addition of heat-killed L. monocytogenes. Heat-killed Listeria is known to promote IL-12 release from BALB/c macrophages (32), which is important for the in vivo generation of a protective Th1 response to the bacterium (41).
Since IFN-γ shares many properties with IL-12, particularly in its ability to drive Th1 responses in vitro (38, 42, 43, 44, 45), it was important to determine whether IFN-γ was responsible for the maintenance of L-selectin expression on CD4+ T cells. The use of anti-IFN-γ demonstrated that IFN-γ had little effect on the maintenance of L-selectin, which, under these culture conditions, appears to be solely regulated by IL-12. However, when cells were cultured in the presence of both IL-12 and IFN-γ, there was a slight increase in L-selectin expression in the primary stimulation, suggesting a synergy between these two cytokines in the initial events of naïve cell differentiation. Indeed, it has been proposed that IFN-γ contributes to the development of Th1 cells by promoting the responsiveness of the naïve cell to IL-12 (46). Although there is controversy surrounding the direct roles of IFN-γ and IL-12 in Th1 development (42, 43, 44, 47) it appears that both cytokines are required for optimal Th1 differentiation.
In contrast to the reported effects of cytokines on L-selectin expressed by human lymphocytes, or human lymphocytic cell lines, IFN-α and -β had no effect on the expression of L-selectin on mouse CD4+ T cells. Ligation of the IL-12R is known to result in STAT4 phosphorylation (48, 49). In the human, but significantly not the mouse, IFN-α and -β have been found to cause phosphorylation of STAT4 (50). This lack of mouse STAT4 phosphorylation by IFN-α or -β may explain why these cytokines do not appear to affect L-selectin on mouse T cells. Thus, it appears that there may be a species difference between mouse and man, with respect to the regulation of L-selectin by cytokines, possibly due to differences in cytokine receptor signaling pathways.
The observation that mouse L-selectin expression is regulated at the mRNA level is in accord with findings published recently that demonstrated that the loss of L-selectin on cell activation is regulated at the transcriptional level (51). In this report, we propose that IL-12 may represent a potential mediator of transcriptional regulation, as this cytokine appeared to have a direct effect on the levels of L-selectin mRNA. Additional control mechanisms, such as the stability of L-selectin mRNA and translational control, cannot be excluded. Similarly, the effect of IL-12 on the activity of the L-selectin sheddase (52, 53) has not been investigated.
On cell activation, naïve CD4+ mouse T cells are reported to shed L-selectin, and it has been suggested that the memory subset is exclusively associated with the L-selectin-negative subset (7). This data conflicts with the results presented herein, which clearly show the presence of L-selectin-positive Th1-type effector cells in vitro. In support of this finding, there are reports of increased numbers of L-selectin-positive cells both after in vivo stimulation with heat-killed L. monocytogenes and following in vitro secondary antigenic stimulation with the bacterium (54). Studies of recall responses of human CD4+ T cells also indicate that memory cells can be associated with the L-selectin-positive and -negative-expressing subsets (55), but, in contrast to the findings presented here, one study has shown that L-selectin is a marker for Th2 cells (56). Consequently, the effector subset on which L-selectin is expressed may also represent a species difference between mouse and man. Nevertheless, these data do suggest that L-selectin may not represent a stable marker for distinguishing naïve and effector cell populations.
Although L-selectin is typically characterized as the naïve T cell-homing receptor (8, 10), there is also some evidence, obtained using the Mel-14-blocking Ab, that L-selectin is involved in the homing of memory cells (23). Given that the homing of the memory cells is L-selectin-dependent, it might be expected that Th1 cells, expressing L-selectin, are also capable of recirculating through the lymph nodes. Although effector cells are likely to have reduced levels of L-selectin relative to naïve cells, cells with lower L-selectin levels have been observed to traffic through the lymph nodes (57).
Other molecules involved in lymphocyte migration are also differentially expressed between Th1 and Th2 cells. Th1 cells, but not Th2 cells, express functional P- and E-selectin ligands (58, 59, 60), allowing the Th1 cells to traffic to sites of inflammation. Conversely, the chemokine receptor CCR3 appears to be expressed on Th2 but not Th1 cells (61), consistent with the observed migration of Th2 cells to the lung. Since L-selectin is implicated in the recruitment of leukocytes to sites of inflammation (62), in addition to lymphocyte homing, our data might contribute to the differential migration observed when Th1 and Th2 effector cells are adoptively transferred into recipient mice (63). Although lymphocyte migration to inflammatory sites was shown to be dependent on both P- and E-selectin, the role of L-selectin was not investigated. Studies on skin grafts, using L-selectin-deficient mice, showed that L-selectin is certainly required for the effective migration of effector cells to inflamed skin (64). It might therefore be predicted that Th1 and Th2 cells display differential homing patterns based on their expression of L-selectin.
We thank Greg Bancroft and Debbie Smith (London School of Hygiene and Tropical Medicine, U.K.) for providing the IFN-α and -β and the heat-killed Listeria, Dorian Haskard (Royal Postgraduate Medical School, U.K.) for providing the L-selectin cDNA, and Gillian Amplett (Glaxo Wellcome, U.K.) for her help with the statistical analyses.
Abbreviations used in this paper: L, ligand; DIG, digoxygenin.