Differential expression of regulator of G-protein signalling transcripts and in vivo migration of CD4⁺ naïve and regulatory T cells

Mice

Six- to eight-week-old female C57Bl/6 (Ly5²), C57Bl/6.Ly5¹ congenic (Ly5¹) and AND T-cell receptor (TCR) transgenic mice²² were obtained from Iffa-Credo (Lyon, France) or the centre de développement des technologies avancées (CDTA) (Orleans, France), or were bred in the Commissarlat à l'Energie Atomique (CEA) animal facilities (Grenoble, France). All animals were kept in specific pathogen free (SPF) conditions in the Pasteur Institute and CEA animal facilities (Paris and Grenoble, France).

Flow cytometric analysis

The following monoclonal antibodies were used: anti-Ly5¹ (A20), anti-Ly5² (104), anti-CD4 (RM4-5), anti-CD25 (7D4), anti-CD45RB (16A), anti-CD69 (H1.2F3) and anti-CD62L (Mel-14) (Pharmingen-BD, Sunnyvale, CA). Cell surface staining was carried out as previously described and analyses were performed on a FACScalibur interfaced to a Macintosh computer with Becton Dickinson CellQuest software (Becton Dickinson, Sunnyvale, CA).²³ Dead cells were excluded from the analysis by light scatter and/or absence of propidium iodide (P1) staining. Cell sorting was performed on a Moflo (Cytomation, Fort Collins, CO) and the purity of sorted cells (greater than 98%) was verified by analysis on a FACScalibur.

Surgery

For parabiosis experiments, 8-week-old female mice were joined laterally under anaesthesia as previously described.²¹ Parabionts were killed 6–12 days after surgery. When required, thymectomy was performed on 6-week-old mice and parabiosis 2 weeks later. Absence of the thymus was confirmed at the time of analysis of the parabionts. All animal experiments were carried out according to institutional guidelines. A standard procedure was used to prepare single cell suspensions and the total number of cells in each lymphocyte subset was calculated as previously described.²³

In vitro and in vivo T-cell activation

For in vitro T-cell activation, spleen cells were cultured in a volume of 15 ml (2 × 10⁶ cells/ml) in 10-cm-diameter Petri dishes in complete RPMI 1640 (Life Technologies, Inc., Grand Island, NY) supplemented with 2 mml-glutamine, 50 µg/ml streptomycin, 50 U/ml penicillin, 5% heat-inactivated foetal calf serum and 5 × 10⁻⁵ m 2-mercaptoethanol. Cells were placed in a humidified atmosphere of 5% CO₂ in air. Concavalin A (ConA) (Sigma Chemical Co., St Louis, MO) (2 µg/ml) or peptide 88–104 of pigeon cytochrome c (Cyt-c) (Neosystem, Strasbourg, France) (1 µg/ml) was added at the start of the culture. At selected times, spleen cells were depleted of B cells and MHC Ia⁺ cells using immuno-magnetic negative selection. Cells were incubated with anti-Ia^b monoclonal antibodies (mAbs) for 15 min, washed and then added to magnetic particles conjugated to goat antimouse immunoglobulin G (IgG) (BioMag^® goat antimouse IgG; PerSeptive Biosystems, Framingham, MA). Surface Ig⁺ and Ia⁺ cells were eliminated by removal of the particles using a magnetic unit (BioMag^® Separator; Advanced Magnetics Inc., Cambridge, MA). The T-cell purity, analysed by flourescence activated cell sorter (FACS), was always higher than 93%. For in vivo T-cell activation, AND mice were immunized with peptide by the intravenous route in the tail vein (150 µg). At selected times, spleen cells were depleted of B cells and Ia⁺ cells as described above.

Chemotaxis assay

Splenocytes were stained with anti-CD4, anti-CD25, anti-CD62L monoclonal antibodies; the staining remain stable at 37° for > 4 hr (data not shown). Then 5 × 10⁵ stained cells were added to the upper chamber of transwell inserts (5-µm pore size; Costar, Cambridge, MA) which were then placed in wells containing medium alone or medium plus chemokine. The chemokines MIP3α/CCL20 and MIP3β/CCL19 and the stromal cell-derived factor SDF1α/CXCL12 were purchased from R&D Systems (Oxon, UK) and used at the mouse effective dose 50 (ED₅₀), as recommended by the manufacturer. After 2 hr of culture at 37° in a 5% CO₂ atmosphere, inserts were removed, and cells from inserts and wells were counted and phenotyped by flow cytometry. Triplicate cultures were analysed for each chemokine. In some experiments, 10⁶ cells were pretreated with pertussis toxin (PTX) at 100 ng/ml. As expected, following this treatment, chemotaxis was abrogated (data not shown).

Polymerase chain reactions (PCRs)

Cells were sorted based on their CD4, CD25 and CD45RB expression levels (Treg: CD4⁺ CD25⁺ CD45Rb^low; intermediate T cells: CD4⁺ CD25^– CD45Rb^low; naïve T cells: CD4⁺ CD25^– CD45RB^high) and total RNA was extracted with RNAzolB (AMS Biotechnology, Abingdon, UK). cDNA synthesis was performed with a reverse transcription kit (N808-0234; Applied Biosystems, Foster City, CA). Real-time quantitative PCRs were performed with the TaqMan probe technology and the ABI PRISM 7000 apparatus (PE Applied Biosystems).²⁴ Duplex PCRs were performed for a reference gene (VIC™-TAMRA™ (VIC-TAMRA) labelled TaqMan probe) and a target gene (FAM™-TAMRA™ (FAM-TAMRA) labelled TaqMan probe). PCRs were performed in duplicate and the reaction mixes were as follows: ± 200 ng cDNA mixed with 12·5 µl Master Mix 2X Buffer (Applied Biosystems), 400 nm of each primer, and 200 nm of TaqMan probe in a final volume of 25 µl. The PCR cycles were: 2 min at 50° followed by 10 min at 95° and 45 amplification cycles (95° for 45 s and 60° for 60 s). Relative quantification of the target gene was performed with the comparative threshold cycle (Ct) method and the amount of target relative to the reference gene is given by: 2^{−(Ct target − Ct reference)}. Primers and probes were designed using Primer Express Software (PE Applied Biosystems). The primers and TaqMan probes used to distinguish the 21 members of the RGS family and their various isoforms are presented in Table 1.

For non-quantitative PCR reactions, 1 µl of cDNA was used as a template in a 19-µl reaction mixture containing: 1·5 mm MgCl₂, 0·2 mm dNTP, 1 µm forward and reverse primers, and 1 U of Taq DNA polymerase (Roche Diagnostics, Basel, Switzerland). PCR was performed on a thermocycler (Perkin-Elmer, Boston, MA). For β-actin amplifications, PCR consisted of a 5-min denaturation step at 94° followed by 40 cycles of 30 s at 94°, 30 s at 55° and 30 s at 72°, and finally one cycle of 10 min at 72°. For Foxp3, reactions were conducted as described above except that the annealing temperature was 61°. The primer sequences were as follows: β-actin: 5′-ACG CAC GAT TTC CCT CTA G-3′ and 5′-ATG GTG GGA ATG GGT CAG AA-3′; Foxp3: 5′-CAG CTG CCT ACA GTG CCC CTA G-3′ and 5′-CAT TTG CCA GCA GTG GGT AG-3′. PCR products were separated on 1·5% agarose gels and visualized under UV illumination.

Differential RGS expression in CD25⁺ and CD25^– CD4⁺ T cells

Among CD4⁺ T cells, two major subsets have been described: naïve CD4⁺ CD25^– T cells and regulatory CD4⁺ CD25⁺ T cells. In humans, these subsets are known to migrate differentially in vitro towards chemokines.¹³ To determine whether naïve and regulatory CD4⁺ mouse T cells also display differential migration properties, we performed an in vitro chemotaxis assay (Fig. 1). Basal migration and migration induced by CXCL12/SDF1α and CCL20/MIP3α were almost the same for the two subsets. However, we found that Treg migrated 3 times less than naïve T cells in response to CCL19/MIP3β. The peculiar migration properties of Treg can be explained by differential chemokine receptor expression: in humans, it has been shown that specific expression of CCR4 and CCR8 by regulatory T cells confers an increased sensitivity to CCL17/thymus and activation-regulated chemokine (TARC) and CCL1/I-309.^12,13 Chemokine receptor expression is difficult to evaluate in the mouse system because of the scarcity of specific antibodies. Moreover, modulation of chemokine receptor expression can only partially explain the migration properties of Treg as many molecules, including RGS, regulate the chemokine receptor transduction pathways.⁷

In order to examine the RGS gene expression profiles of naïve and regulatory T cells, splenic T cells were sorted according to their combined expression of several surface markers: Treg are CD4⁺ CD25⁺ CD45RB^low whereas naïve T cells are CD4⁺ CD25^– CD45RB^high. Cells with an intermediate phenotype (CD4⁺ CD25^– CD45RB^low) were also identified and sorted. For the first time, a global analysis of RGS expression was performed by quantitative PCRs on cDNA derived from these cells (RGS1 to RGS20, lbc (GEF protein) second cousin (lsc), and RGS domain and PHOX domain containing protein (RGS-PX1) (Fig. 2). The expression levels of the 18S rRNA, β-actin and elongation factor 1α (EF1α) genes were, as expected, identical in naïve and regulatory T cells (less than 20% difference). We confirmed by reverse transcriptase PCR (RT-PCR) that Treg express specifically the Foxp3 transcript (Fig. 3).² We observed that several RGS genes were differentially expressed in Treg compared to naïve T cells: RGS1, RGS9 and RGS16 (Fig. 2).

Our global RGS profiling showed that CD25⁺CD45RB^low regulatory T lymphocytes express 30 times more RGS1 than CD25⁻ CD45RB^high naïve T lymphocytes, as recently described.^1,25 In addition, we found that levels of transcripts for RGS9 were at least 70 times greater in regulatory than in naïve T cells. The expression of this gene in lymphocytes has not been described before. The absence of RGS9 expression in naïve T cells represents an exception among RGS gene family, members of which are usually expressed ubiquitously. CD25⁺ CD45RB^low regulatory T lymphocytes contained 30 times more RGS16 transcripts. RGS16 was cloned as an IL-2-induced gene from human T cells; thus, its overexpression in CD25⁺ cells was perhaps expected.²⁶

RGS profile expression is modified after T-cell activation

In some respects, Treg appear to have an activated phenotype.²⁷ Because expression of several RGS genes was significantly lower in naïve T cells, we wished to determine whether expression could be increased following activation.

Splenic T cells from C57Bl/6 mice were activated in vitro with ConA, and expression of CD69 and CD25 (two activation markers) was monitored by FACS (Fig. 4a). As expected, CD69 expression appeared early (3 hr) on CD4⁺ T cells following activation, whereas CD25 appeared slightly later (at 6 hr). Analysis of RGS gene expression (Fig. 4b) showed that levels of RGS1, RGS9 and RGS16 transcripts increased 3 hr following ConA activation. The expression levels of RGS1 and RGS16 progressively increased to reach a maximum at 12 hr after ConA treatment (Fig. 4b). Addition of cyclosporin A abrogated the up-regulation of RGS1, RGS9 and RGS16, suggesting that the calcineurin/nuclear factor of activated T cells (NF-AT) pathway may repress this process (data not shown and Fong et al.²⁸). Up-regulation of RGS1 following T-cell activation has been previously reported,¹ but our global analysis shows that RGS9 and RGS16 are also up-regulated. Other RGS genes were not affected following activation, showing less than a 2-fold change (data not shown). The only exception was the RGS11 gene, whose expression decreased 7 times 6 hr after ConA treatment. We noted that the expression of RGS11 was slightly decreased in Treg compared to naïve T cells (3 times) (Fig. 2). In summary, in vitro activated CD4⁺ T cells displayed an RGS expression profile similar to that of ex vivo Treg.

In order to study a homogenous naïve T-cell population and to determine whether cells activated in vivo also modified their RGS gene expression, we activated TCR transgenic AND T cells with the activating pigeon Cyt-c peptide 88–104. After in vitro peptide stimulation, RGS1, RGS11 and RGS16 expression was, respectively, modified by factors of 2·0, 0·25 and 7·0 at 6 hr and 2·5, 0·12 and 6·8 at 12 hr (Table 2). In vivo experiments corroborated our in vitro results, although differences in RGS expression were not as important as for in vitro ConA or in vitro peptide stimulation. In AND transgenic mice, the vast majority of splenic and lymph node Vβ3⁺ cells displayed the typical phenotype of resting T cells: CD69^–, CD25^–, CD44^low and CD62L^high. Following peptide immunization, all CD4⁺ Vβ3⁺ cells transiently expressed CD25, CD44 and CD69 molecules (data not shown and Mascarell & Truffa-Bachi²⁹). Shortly after in vivo immunization (6 hr), RGS1 and RGS16 expression was increased by factors of 1·7 and 3·7, respectively, whereas RGS11 expression decreased by a factor of 3·0 (Table 2).

Differential in vivo migration of CD25⁺ and CD25^– CD4⁺ T cells

The differences in the RGS profiles of CD25⁺ and CD25⁻ CD4⁺ T cells suggest that RGS may play a role in their migration. Thus, RGS gene up-regulation in Treg may confer decreased sensitivity to chemokine receptor signalling at the molecular level and decreased migration capacity at the cellular level. In order to compare directly the in vivo rates of migration of regulatory and naïve CD4⁺ T cells, we performed parabiosis experiments. Cross-circulation between the two parabionts is established very early after surgery, and at day 2 serum components are completely mixed (data not shown). Between 6 and 12 days, cellular exchange between parabionts is the result of mature lymphocyte migration and not the migration of their progenitors, as the degree of chimerism in the thymic precursor compartment is less than 1·5%.²¹ Among peripheral T cells, we found that the level of chimerism was different in various lymphocyte compartments.

As shown in Fig. 5a (6 days post-parabiosis), CD4⁺CD25⁻ CD45RB^high naïve T cells migrated more than CD4⁺ CD25⁺ CD45RB^low Treg. Thus, 28 ± 6% (mean ± standard deviation) of the naïve T cells found in the spleen originated in the partner parabiont, whereas this value was only 6 ± 2% for Treg. The best way to compare migration of regulatory and naïve T cells is to determine the ratio of chimerism for these cells in each parabiont. We found that, in individual parabionts, naïve cells migrated 6 times more than Treg (percentage of chimerism for naïve cells/percentage of chimerism for regulatory cells = 5·5 ± 1·5). Intermediate cells migrated more than Treg, but less than naïve T cells, the spleen chimerism being 12 ± 4% for this cell population (Fig. 5a). Results obtained in the lymph nodes were comparable to those of the spleen; the chimerism in the naïve T-cell compartment was 26 ± 5%, whereas that of Treg was very low (3·0 ± 1·5%). The ratio of chimerism for these cell populations in lymph nodes of each parabiont was 12 ± 4.

Chimerism observed in peripheral lymphoid organs could be influenced by seeding of recently produced thymic emigrants.³⁰ To exclude a role for emigrants, we thymectomized mice prior to parabiosis. Results obtained in thymectomized parabionts, shown in Fig. 5b, were comparable to those obtained with non-thymectomized animals (Fig. 5a), showing that, within the CD4⁺ population, peripheral Treg migrated less than naïve T cells (Fig. 5b). As expected, the migrating CD4⁺ cells were mostly CD62L^high (data not shown and Fisson et al.²⁷). All these results show that there is a correlation between the RGS gene expression profile and the in vivo migration of naïve and regulatory CD4⁺ T cells. Regulatory T cells migrated less than naïve T cells and expressed higher levels of the chemokine-receptor-negative regulators RGS1, RGS9 and RGS16.

RGS expression is modified in trafficking cells

In order to determine whether a correlation exists between RGS gene expression and cell migration, we decided to purify both trafficking and static subpopulations of naïve and regulatory T cells from Ly5 congenic parabionts. Cells were sorted according to their phenotype and their origin, defined by the Ly5 allele: ‘exogenous regulatory T cells’, ‘endogenous regulatory T cells’, ‘exogenous naïve T cells’ and ‘endogenous naïve T cells’. The level of chimerism was found to be similar to that in Fig. 5. We then determine the expression levels of RGS genes. We observed that, among naïve T cells, those that had migrated from one animal to the other had significantly down-modulated RGS1 and RGS9 expression (Fig. 6, upper panel). In contrast, Treg migrated much less and there was no major change in RGS expression, even in the small fraction of cells that had migrated (Fig. 6, lower panel). Thus, migration of naïve cells was correlated with a down-modulation of RGS1 and RGS9, whereas Treg were characterized by elevated and unmodified expression of RGS1 and RGS9 (Figs 5 and 6). Similar results were obtained with pools of cells from C57Bl/6 mice (Ly5², left parabionts) or C57Bl/6.Ly5¹ mice (right parabionts) for a limited number of genes (β-actin, RGS1 and RGS16; data not shown).

As shown above and described in the literature,¹⁹ RGS1 and RGS9 appear to be major regulators of lymphoid cell migration in general and CD4⁺ naïve T cells in particular. RGS16 expression is increased in TCR activated cells and Treg. However, RGS16 expression is not modified in trafficking CD4⁺ naïve T cells. Thus, it seems that this gene does not participate in the regulation of T-cell migration in physiological conditions. It is likely that RGS16 regulates signalling of other GPCRs (non-chemokine receptors) or may play a role only in cell migration in pathological conditions (e.g. inflammation), as recently shown in the RGS16 TG mice.³¹

In conclusion, in parabiosis experiments, we demonstrated that naïve T cells are trafficking whereas regulatory T cells are mostly sessile. Regulatory T cells may be sequestrated in niches or may be less sensitive to chemokine-induced migration. Chemokines are known to regulate both cell retention and rate of immigration within tissues. This has been described recently in detail for neutrophils stored in the bone marrow. During inflammation, these cells migrate to the periphery in response to various factors, including the CXCR2 ligand keratinocyte-derived chemokine (KC) chemokine produced at inflammatory sites. However, neutrophils are also sensitive to CXCL12 (SDF1α), which favours their sequestration in the bone marrow.³² Regulatory T cells may be more static because they are more sensitive to chemokines produced locally and less sensitive to chemokines produced distantly. Our results support this hypothesis, as Treg were found to be less sensitive to CCL19, whereas they migrated slightly more towards CCL20.^1,12 Migration may be regulated by chemokine receptor expression or by modulation of chemokine receptor signals by regulators such as RGS molecules.¹³ Global RGS profiling by quantitative PCR revealed that these molecules are expressed ubiquitously but their level of expression is different in various cell types (F. Agenès, unpublished data). Some RGS genes are expressed at identical levels in naïve and regulatory CD4⁺ T cells, and as such characterize the whole CD4⁺ T cell population. However, some RGS are differentially expressed in naïve cells and Treg. Here, we showed that Treg display a specific RGS gene profile characterized by elevated RGS1, RGS9 and RGS16 expression when compared to naïve T cells. We showed that RGS1, RGS9 and RGS16 are up-regulated after TCR engagement. Surprisingly, these three RGS transcripts are those that are highly expressed in Treg. It is possible that Treg acquire, during their differentiation and maturation, a specific RGS gene expression profile. Higher expression of the inhibitory molecules RGS1, RGS9 and RGS16 may decrease sensitivity to chemokines and change migratory abilities. Whether modulation of RGS expression precedes induction of migration is unclear, but RGS molecules may serve as potential targets for modulating lymphocyte migration.