FTY720 (S) phosphate (FTY720-P) was purchased from Cambridge Bioscience (Cambridge, UK). It was dissolved in ethanol at 1 mg/ml for storage at −20°C. On the day of use 1 mg/ml FTY720-P was diluted to 100 µg/ml in sterile water with 2% β-cyclodextrin (Sigma-Aldrich; Poole, UK) and then mixed thoroughly. SEW2871 was purchased from Enzo Life Sciences (Exeter, UK).

Female BALB/c (H2^d) and C57BL/6 (H2^b) mice (6–8 weeks old; Charles River, Margate, UK) were maintained under pathogen-free conditions. All procedures were performed in accordance with UK Home Office and EU Institutional Guidelines and within the parameters of current personal and project licences.

C57BL/6 mouse footpad injections (on day zero) were unilateral, subcutaneous, and comprised 5×10⁶ BALB/c splenocytes suspended in 25 µl RPMI 1640 medium (Sigma-Aldrich). Between days 2 and 5, some mice were injected daily, intraperitoneally, with 100 µl 100 µg/ml FTY720-P or an equal volume of drug vehicle. The mice were killed on day six and the popliteal lymph nodes draining the injected footpads removed. Cell suspensions were prepared from the nodes by pressing the tissue through 70 µm cell strainers (BD Biosciences; Oxford, UK) into RPMI 1640 medium. Popliteal lymph node-derived cells were washed twice with RPMI 1640 medium before further use.

BALB/c splenocytes were prepared as follows. Spleens were mechanically disrupted then the tissue forced through cell strainers into RPMI 1640 medium. The cells were purified by density centrifugation (Histopaque-1083; Sigma-Aldrich).

Ninetysix-well format Immobilon MultiScreen-P plates (Millipore; Watford, UK) were coated with IFN-γ capture antibody [clone AN18] (Mabtech; SE-131 28 Nacka Strand, Sweden) diluted into carbonate-bicarbonate buffer (Sigma-Aldrich) overnight at 4°C. These plates were washed twice with PBS and blocked with RPMI 1640 medium supplemented with 10% FBS, 100 U/ml penicillin and 0.1 mg/ml streptomycin (all Sigma-Aldrich) for 1 h at room temperature. Mixed leukocyte reaction assays were then performed in each well by mixing 1×10⁴ C57BL/6 popliteal lymph node-derived cells with 2×10⁵ BALB/c splenocytes in a total volume of 200 µl RPMI 1640 culture medium containing 50 µM 2-mercaptoethanol (Sigma-Aldrich). After incubation for 18 h at 37°C the cells were discarded and the plates washed six times with PBS. A biotinylated IFN-γ detection antibody [clone R4–6A2] (Mabtech), diluted in PBS, was applied overnight at 4°C. The plates were washed six times with PBS. The streptavidin-alkaline phosphatase conjugate (Mabtech) was diluted in PBS and applied for 1 h at room temperature. The plates were washed another six times with PBS and 50 µl of BCIP/NBT liquid substrate system for membranes (Sigma-Aldrich) was added to each well. After development for 5 min at room temperature, the reaction was stopped by removal of substrate and washing with tap-water. Developed spots were enumerated using an ELISPOT reader (AiD; Strassberg, Germany).

Total CD3⁺ T cells were derived from human peripheral blood or platelet-depleted leukocyte cones (NHS Blood and Transplant Service, UK). They were isolated using an erythrocyte-rosetting negative-selection kit (StemCell Technologies; Grenoble, France) and cell separation by centrifugation across Lympholyte H (Cedarlane Laboratories; Ontario, Canada). T cells were cultured in X-VIVO 15 medium (Lonza; Slough, UK) and activated with Dynabeads Human T Activator CD3/CD28 (Life Technologies; Paisley, UK), at a ratio of one bead per cell and a starting culture density of 1×10⁶ cells per ml.

For analysis of mouse cells, the antibodies were CD3 [clone KT3] (AbD Serotec) and CXCR3 [clone 220803] (R&D Systems). The antibodies used for human cells were S1PR1 [clone 218713] (R&D Systems) and CD69 [clone FN50] (BD Biosciences). In some cases T cells were labelled by incubation with 1 µM CFSE-DA (carboxyfluorescein diacetate N-succinimidyl ester; Sigma-Aldrich) for 5 min at 37°C in PBS +0.1% (v/v) FBS. They were then chilled rapidly on ice and washed with cold RPMI 1640 before use. Data was acquired on a FACSCanto II instrument (BD Biosciences) and analysis performed using FACSDiva 6.1.3 (BD Biosciences) and FlowJo 7.6 (Treestar; Ashland, Oregon, USA) software.

RNA was prepared from cell pellets using TRI Reagent (Sigma-Aldrich). cDNA was synthesised from approximately 1 µg RNA per sample using SuperScript III Reverse Transcriptase and random hexamers as primers (Life Technologies). Expression of S1PR1 was determined semi-quantitatively using a specific TaqMan Gene Expression assay (Hs00173499_m1), with 18S ribosomal RNA as the reference (Hs99999901_s1; both Applied Biosystems, Life Technologies; Paisley, UK). The amplifications were run on an ABI Prism 7000 instrument (Applied Biosystems).

T cells were lysed in a buffer containing CelLytic M (Sigma-Aldrich) + Mini Complete Protease Inhibitor cocktail (Roche; Welwyn, UK) + Halt Phosphatase Inhibitor cocktail (Thermo Fisher Scientific; Loughborough, UK). The concentration of protein in cell lysates was determined by BCA assay (Thermo Fisher Scientific). Equal masses of protein (routinely 10–50 µg) were loaded into each lane of a single gel and subjected to 12% resolving SDS-PAGE. Transfer to polyvinylidene difluoride membrane (GE Healthcare; Little Chalfont, UK) was by wet electroblotting. Membranes were blocked with 5% immunoglobulin-depleted BSA (Sigma-Aldrich) for 1 h before immunoblotting. The primary antibodies used were: phospho(S473)-Akt [clone D9E] and pan Akt [clone C68E7] (both New England Biolabs; Hitchin, UK). The secondary antibody was horseradish peroxidase-conjugated anti-rabbit IgG (Sigma-Aldrich).

Phospho(S473)-Akt and total Akt were also detected using a cell-based ELISA (R&D Systems). Approximately 10 000 T cells were added to each well of a 96-well format plate (previously coated with 10 µg/ml poly-lysine (Sigma-Aldrich) for 30 min) and left to attach for 1 h. The cells were treated for the desired length of time then fixed with 8% formaldehyde. Phospho(S473)-Akt and total Akt were detected by double immunoenzymatic labelling. The relative quantities of the proteins were deduced from the intensities of spectrally distinct fluorescences associated with each target. The detector was a Dynex MFX instrument (Worthing, UK) operating with excitation and emission wavelengths of 540 nm and 600 nm for phospho(S473)-Akt, and 360 nm and 450 nm for Akt.

Resting T cells were transfected with 100 pmol siRNA per million cells by electroporation (Nucleofector I instrument; Lonza). For CD69 knockdown, an equimolar mixture of three 21 nucleotide duplexes (MISSION siRNA; Sigma-Aldrich) was used. The single strand sequences were:

5′ GAGUUAGAUGUUGGUACUA 3′

5′ CUACUCUUGCUGUCAUUGA 3′

5′ CUCUCAUUGCCUUAUCAGU 3′

Separately, as a control, cells were transfected with an equal mass of MISSION siRNA Universal Negative Control 1 (Sigma-Aldrich).

After transfection, the T cells were rested for about 5 h in X-VIVO 15 before addition of Dynabeads Human T Activator CD3/CD28 at a ratio of one bead per cell. 24 hours later the cells were analysed directly for CD69 expression or viable cells were sorted using a FACS Aria II (BD Biosciences) instrument for phospho-protein analysis.

Prism 4.0c (GraphPad Software; La Jolla, California, USA) was used for statistical analyses. Comparisons between groups were made using the Student’s t-test for unpaired data. P values ≤0.05 were considered significant.

An initial series of experiments demonstrated that unilateral foot pad sensitisation of C57BL/6 mice with BALB/c splenocytes produced a difference (P<0.01) in popliteal node cellularity after 6 days, with a mean of 2.01×10⁶ cells (s.e.m. 0.17×10⁶) recovered from the sensitised side and 4.88×10⁵ (s.e.m. 2.92×10⁵) from the uninvolved contralateral node (data not shown). Efficacy of the drug FTY720-P in our animal model was tested by measuring the effect of its administration on the frequency of peripheral blood CD3⁺ cells. Daily dosing over two days resulted in a greater than 90% depletion of CD3⁺ cells from peripheral blood (Fig 1A and B). A second series of experiments showed that the overall yield of cells from the reactive popliteal node of sensitised animals was not changed from control values by daily treatment with FTY720-P (P>0.05; Fig 1C).

Measurement by ELISPOT of the frequency of IFN-γ producing, alloreactive T cells in the population recovered from reactive popliteal lymph nodes showed a difference (Fig 1D; P<0.01) between animals treated with the drug vehicle and FTY720-P, with a mean 1.9-fold enrichment of these cells in nodes from the drug treated animals. Analysis of the expression of the T cell activation-associated chemokine receptor CXCR3 in the popliteal node-derived cell population (Fig 1E) also showed a mean 1.9-fold enrichment of these cells in drug-treated animals (Fig 1F; P<0.05).

Longitudinal analysis of the expression of S1PR1 gene following T cell activation by stimulation of CD3 and CD28 (Fig 2) showed a marked reduction in transcription after 24 h; a similarly reduced level of mRNA encoding this receptor was observed 3 days after T cell activation.

To test for the presence of functional S1PR1 at the cell surface, T cells were stimulated with the specific agonist SEW2871 and activation of the downstream signalling component Akt measured. When resting cells were stimulated with the specific agonist SEW2871 there was rapid formation of the phosphoserine (473) derivative of Akt (pAkt) (Fig 3A). Stimulation of T cells with this agonist 24 h after the activation of T cells did not increase the level of pAkt. However, an increased level of pAkt was observed when T cells were stimulated with SEW2871 on either day 3 or 6 after mitogenic activation. Additionally, Cell-based ELISA was used to quantify the relative amounts of pAkt and Akt in each cell population before and after cell stimulation with SEW2871. Data from three independent donors (Fig 3B) show that the transient reduction in S1PR1-mediated intracellular signalling was only apparent 1 day after T cell activation.

Flow cytometric analysis of human T cell surface expression of CD69 showed that almost no resting cells expressed this antigen (Fig 4A). However, almost all these cells expressed CD69 when analysed 1 and 3 days after mitogenic T cell activation (Fig 4A). Although the T cells analysed on days 1 and 3 after activation were uniformly positive for CD69 expression, quantification of the median level of antigen expression per cell demonstrated a marked reduction between days 1 and 3 after T cell activation (Fig 4B).

The rate of loss of CD69 from the surface of activated T cells was analysed by fractionating the dividing cells on the basis of CFSE dilution (Fig 5A). This experiment showed that none of the cells had divided on day 1, maintaining maximum CD69 expression. However, the overall level of CD69 expression observed on day 3, for example, was the sum of separate levels expressed by T cell subpopulations which had divided 0, 1, 2 or 3 times (Fig 5B). On both days 3 and 6 after activation, analysis of the rate of decrease of CD69 expression between cell cycles 0 and 3 showed a similar half-life of 1.15 and 0.97 mitotic division cycles respectively (Fig 5C).

A series of experiments was performed to demonstrate the potential to reduce CD69 expression by transfection of resting human T cells with sequence-specific siRNA sequences prior to mitogenic activation. After gating to exclude the T cells damaged during transfection, the percentage of cells expressing CD69 1 day after mitogenic activation (Fig 6) was reduced from 35.6% (control siRNA) to 13.6% (CD69-specific siRNA); the corresponding median fluorescence values were 24.1 and 11.4 respectively.

A series of 3 separate experiments was then performed in which viable and non-viable T cells were separated after transfection by fluorescence-activated cell sorting. These cells were then activated by stimulation of CD3 and CD28 for 1 day and the pAkt/total Akt ratio was quantified after stimulation with the specific S1PR1 agonist SEW2871. This study demonstrated that treatment with the CD69-specific siRNA sequence markedly increased the pAkt signal generated by stimulation of S1PR1 (P<0.001; Fig 7).