In T cells, the PI3K pathway promotes proliferation and survival induced by Ag or growth factors, in part by inactivating the FOXO transcription factor 1. We now report that FOXO1 controls the expression of L-selectin, an essential homing molecule, in human T lymphocytes. This control is already operational in unprimed T cells and involves a transcriptional regulation process that requires the FOXO1 DNA-binding domain. Using transcriptional profiling, we demonstrate that FOXO1 also increases transcripts of EDG1 and EDG6, two sphingosine-1-phosphate receptors that regulate lymphocyte trafficking. Additionally, FOXO1 binds the promoter of the cell quiescence and homing regulator Krüppel-like factor 2 and regulates its expression. Together, these results reveal a new function of FOXO1 in the immune system and suggest that PI3K controls a coordinated network of transcription factors regulating both cell quiescence and homing of human T lymphocytes.

The PI3K pathway translates numerous extracellular stimuli into a wide range of essential cellular processes through 3′-phosphoinositide-dependent effectors like the serine/threonine kinase Akt. We and others previously reported that PI3K is strongly activated in naive T cells after Ag recognition (1, 2). In different cell systems Akt has numerous substrates, including transcription factors that shape essential biological outcomes triggered by PI3K activation. Among these are the three principal members of the forkhead box O (FOXO)4 family of transcription factors (FOXO1, FOXO3, and FOXO4) now considered as true tumor suppressor genes that can control cell quiescence in various cellular systems (3). FOXO transcription factors also promote stress resistance in a broad range of cell types and organisms and can induce apoptosis upon cytokine withdrawal in some cell types, including lymphocytes (4). Under conditions of low Akt activity, unphosphorylated FOXOs reside within the nucleus where they support the transcription of molecules that promote these different biological effects, depending on cell context. Activation of naive T cells by Ag triggers phosphorylation of FOXOs by Akt and localization to the cytosol. Using an active form of one FOXO member, FOXO1, we demonstrated that this compartmentalization process can affect T cell growth after Ag recognition (5). Thus, the need for sustained PI3K signaling to exclude FOXOs from the nucleus and neutralize their transcriptional activity, as already shown in various cell systems (6, 7, 8), is likely a key event used by primary T cells to initiate cell growth.

It is clear that in T cells exposed to Ag or cytokines, or in phosphatase and tensin homolog (PTEN)-deficient transformed T cell lines showing elevated D3-phosphoinositides, FOXO inactivation is a central mechanism by which the PI3K/Akt triggers proliferation and survival. This central role of FOXO proteins in opposing proliferation and survival was recently confirmed by the finding that the mouse genes Foxo1, Foxo3a, and Foxo4 are redundant tumor suppressor genes (9, 10). However, mice that conditionally lack these three principal FoxO genes develop tumors restricted to only certain tissues, like thymic lymphomas and hemangiomas. Moreover, individual disruption of FoxO genes in mice results in different phenotypes (11, 12). Thus, a broad consensus has now emerged that FOXO factors have tissue-specific and cell type-specific functions and that their transcriptional targets vary depending on context (9, 10, 12, 13, 14, 15, 16, 17, 18). Investigating this issue with FOXO1, we now report a new committed and unanticipated function of this transcription factor in human T cells to coordinately regulate expression of L-selectin (CD62L) and other homing-related molecules downstream of PI3K.

Human T lymphocytes were isolated from blood donors by Ficoll density gradient centrifugation followed by negative depletion on magnetic beads (T cell negative isolation kit; Dynal Biotech). Human primary T cells and leukemic T cell lines (Jurkat, CEM, and SupT1) were grown in RPMI 1640 medium (Invitrogen) supplemented with 10% FCS, 50 U/ml penicillin, 50 μg/ml streptomycin, 4 mM glutamine, and 1 mM sodium pyruvate.

FOXO1-GFP and the triple mutant T24A/S256A/S319A FOXO1-GFP (FOXO1A3-GFP) constructs were a gift from Dr. T. Unterman (University of Illinois College of Medicine, Chicago, IL) (8). FOXO1-cyan fluorescent protein (CFP) construct was obtained by subcloning FOXO1 as a NheI/BamHI restriction fragment into the pECFP-N2 vector (Clontech Laboratories). The FOXO1A3 H215R-GFP DNA binding mutant was generated using the QuickChange site-directed mutagenesis kit (Stratagene). The FOXO1A3-GFP plasmid was used as a template for the PCR and amplified with the following primers: 5′-CTGGAAGAACTCAATTCGTCGTAATCTGTCCCTACACAGC-3′ (upstream primer) and 5′-GCTGTGTAGGGACAGATTACGACGAATTGAGTTCTTCCAG-3′ (downstream primer). A Myr-PTEN-GFP construct was provided by Dr. H. Bourne (University of California, San Francisco, CA). Myr-PTEN-YFP was obtained after subcloning Myr-PTEN as EcoRI/EcoRI fragment into the EcoRI site of the pEYFP-N3 vector (Clontech Laboratories). The TRIPiziE-FOXO1-GFP and TRIPiziE-FOXO1A3-GFP lentiviral constructs were generated after subcloning FOXO1-GFP and FOXO1A3-GFP as a NheI/XbaI blunted restriction fragment into the blunted BamH1 site of TRIPΔU3-EF1α lentiviral vector (19). A PCR fragment coding for human KLF2 was obtained from Jurkat cDNA using a GC-rich PCR amplification kit (Roche Diagnostics) with the following primers: 5′-CATGGCGCTGAGTGAACCCATC-3′ (upstream primer) and 5′-CTACATGTGCCGTTTCATGTGCAGC-3′ (downstream primer). This PCR product was cloned into the TOPO PCRII vector (TA cloning kit; Invitrogen), then subcloned as an EcoRI/EcoRI fragment into the EcoRI site of pEGFP-N2 (Clontech Laboratories). A NheI/XbaI blunted restriction fragment of this plasmid was next cloned into the blunted BamH1 site of TRIPΔU3-EF1α lentiviral vector. All constructs were verified by sequencing.

Jurkat T cells (5 × 106 in RPMI 1640 medium) were electroporated in 4-mm polycarbonate cuvettes (Eurogentec) with 2 μg of pEGFP-C1 or 7 μg of the appropriate plasmids at 250 V, 960 μF. Freshly prepared human primary T cells were transfected with the Amaxa Nucleofector technology according to the manufacturer’s instructions with 5 μg of the appropriate construct.

Lentiviral production was obtained by transient calcium-phosphate co-transfection of the 293T cell line with the TRIPΔU3-EF1α vector, the p8.91 encapsidation vector (ΔVprΔVifΔVpuΔNef), and a vesicular stomatitis virus-G protein (VSV-G) envelop expression plasmid (pHCMV-G). Fresh medium was added after 24 h at 37°C. The culture supernatant containing viral particles was collected 48 h posttransfection and treated for 20 min at 37°C with 1 μg/ml DNase I (Boehringer Ingelheim). Virion particles were concentrated by ultracentrifugation and the resulting pellet was resuspended in PBS and frozen at −80°C under 20 μl of aliquots until use. The concentration of virion particles was normalized by measuring the p24 capsid protein content of supernatants by ELISA (HIV-1 p24 ELISA Ag assay; Beckman Coulter).

Before lentiviral transduction, human primary lymphocytes were activated for 24 h in 96-wells round-bottom plate (2 × 105/well) with anti-CD3/anti-CD28-coated Dynabeads (Invitrogen) in the presence of IL-2 (30 U/ml). We used a bead/cells ratio equal to 1:3 and left the stimulating beads in the culture medium during all the infection procedure. Stimulated primary T cells were then incubated with lentiviral vector particles. After a 24-h exposure to virus at 37°C, cells were centrifuged and resuspended in fresh RPMI medium with IL-2. Transduced cells were processed and analyzed after different periods of culture (see Results). In some experiments, transduced T cells were sorted based on GFP expression. T cell lines were infected in 24-well plate (0.3 × 106/well) in 0.5 ml of complete RPMI 1640 medium and after a 24-h exposure to virus at 37°C, washed and resuspended in fresh RPMI 1640 medium. Transduced cells were usually processed from 24 to 48 h after this washing step.

Cells were lysed at 4°C for 30 min in a lysis buffer (20 mM Tris-HCl, pH 7.5, 140 mM NaCl, 1 mM EDTA, 50 U/ml aprotinin, 1 mM PMSF, 1 mM sodium orthovanadate, 50 mM sodium fluoride) containing 1% (v/v) Nonidet P-40 detergent. Nuclei and cellular debris were removed by centrifugation at 10,000 × g for 10 min, and the amount of proteins in each postnuclear supernatant was determined. All samples were diluted in Laemmli buffer (500 mM Tris-HCl, pH 6.8, 10% SDS, 10% glycerol, 5% 2-ME, 10% bromphenol blue) and boiled for 3 min. Proteins were loaded onto 10% SDS-polyacrylamide gels and electrophoretically transferred for 75 min at 65 V to a polyvinylidene difluoride membrane (Amersham Biosciences). Immunoblotting was performed using the indicated Ab at the appropriate dilution followed by peroxidase-labeled goat anti-mouse or goat anti-rabbit antiserum (Bio-Rad), and revealed using an ECL system (Amersham Health).

Cells were fixed for 10 min at room temperature in PBS-4% paraformaldehyde, washed twice in PBS, and stored at 4°C in PBS. Analyses were performed on an Eclipse TE3200 inverted microscope (Nikon) with a ×40 oil objective. Images were collected with a cooled charge-coupled device camera (CoolSNAPfx; Roper Scientific) and the Metavue Imaging system (Universal Imaging) and then analyzed with MetaMorph software (Universal Imaging).

Transfected cells were analyzed for expression of both GFP and surface markers using the following PE-conjugated mouse mAbs: anti-CD62L, anti-CD69, anti-CD25, or anti-CD71 (BD Biosciences). In brief, 2.5 × 105 cells were washed twice with PBS-2% FCS and labeled with 5 μl of Ab in 100 μl of PBS-2% FCS for 60 min at 4°C. Cells were then washed twice in PBS, resuspended in PBS, and directly analyzed on a BD FACScan flow cytometer.

Jurkat cells were infected with GFP or FOXO1A3-GFP as described above. Cells were resuspended at 106 cells in 200 μl of fresh complete medium 2 days postinfection and cultured for 24 h. Supernatants were then collected and soluble CD62L (sCD62L) was measured by ELISA (Diaclone) following manufacturer’s instructions.

Total RNA was isolated from infected cells by TRIzol extraction, according to the manufacturer’s instructions (Invitrogen) and purified on silica gel columns using the RNeasy MinElute Cleanup kit (Qiagen). Purified RNA was checked on a Nanodrop spectrophotometer ND-1000 (NanoDrop Technologies).

Total RNA were directly compared using Agilent oligonucleotide dual color technology in running dye swap and duplicate experiments. Probe synthesis and labeling were performed by Agilent’s low fluorescent low input linear amplification kit. Hybridization was performed on human whole genome 44,000 oligonucleotide microarrays using reagents and protocols provided by the manufacturer. Feature extraction software provided by Agilent (version 7.5) was used to quantify the intensity of fluorescent images and to normalize results using the linear and lowest subtraction method. Primary analysis was performed using Resolver software (Rosetta Biosoftware) to identify genes differentially expressed between GFP and GFP-FOXO1A3 with a fold change >2 and a value of p < 10−5.

Total RNA was prepared using RNeasy mini kit (Qiagen) and checked for purity as described above. cDNA was produced with the Advantage RT-for-PCR kit (Clontech Laboratories) using 1 μg of total RNA and random hexamer priming in a final volume of 100 μl. Real-time quantitative PCR was performed by using the LightCycler FastStart DNA Master plus SYBRGreen kit (Roche Diagnostics). Genes of interest were detected using primers that had been designed with the Oligo6 software (Molecular Biology Insights) and optimized to generate a single amplicon of 80–130 nucleotides. The following primers were used: CD62L forward, 5′-CTGGCACATCATGGAACCGAC-3′; CD62L reverse, 5′-GTGTAATTGTCTCGGCAGA-3′; EDG1 forward, 5′-TTCTGCGGGAAGGGAGTATGT-3′; EDG1 reverse, 5′-AAGAGGCGGAAGTTATTGCT-3′; EDG6 forward, 5′-GGAACTGCCTGTGCGCCTTT-3′; EDG6 reverse, 5′-AGGACGCCGGCGAAGATCA-3′; KLF2 forward, 5′-CACCGGGTCTACACTAGAGG-3′; KLF2 reverse, 5′-AAATGCCGCAGACAGTACAA-3′; CCR7 forward, 5′-GCCAACTTCAACATCACCA-3′; CCR7 reverse, 5′-AAGGCGTACAAGAAAGGGTT-3′; Bim forward, 5′-ATTGCAGCCTGCGGAGAG-3′; Bim reverse, 5′-ATGGGTGCTGGGCTCCT-3′; p27 forward, 5′-CTCCCTTCCACCGCCATA-3′; p27 reverse, 5′-CGGAGAGGGTGGCAAAGC-3′. Reactions were performed in glass capillaries in a final volume of 20 μl. In brief, 5 μl of cDNA or water were added to 15 μl of Light Cycler DNA Master SYBR Green I mix containing 0.5 μM of the appropriated primers. Each sample was processed in duplicate with initial incubation at 95°C for 10 min followed by 40 cycles of denaturation at 95°C for 10 s, primers annealing at 55°C for 8 s, and extension at 72°C for 10 s. SYBR Green I fluorescence emission was determined after each cycle. Melting curve analysis was performed to assess the specificity of the PCR products. Collected data were filtered to remove background with the light cycler data analysis software and crossing point determination was performed using the fit point method. An external standard curve was generated for each set of primers by amplification of serial dilutions of a cDNA from T lymphocytes. This curve provided an estimation of primer sets efficiency (according to the general PCR equations, Eff = 10 −1/slope where Eff represents the primer efficiency) and was stored as a coefficient file used for the analysis. Relative expression levels of the target genes were then generated by an efficiency-corrected calibrator-normalized relative quantification strategy using the so-called coefficient files, according to the manufacturer’s recommendations (RealQuant software; Roche Diagnostics, technical note LC 13/2001). To correct for variations in mRNA recovery and reverse transcription yield, the target gene expression was normalized to peptidylprolyl isomerase A expression, a housekeeping gene, in the same samples. The final normalized ratio was the result of the efficiency corrected target/reference ratio of each sample divided by the target/reference ratio of a calibrator. We used the GFP-infected cells sample as calibrator.

Gene lists were assessed for known biological functions using Ingenuity Pathways Analysis software (Ingenuity). Significance was expressed as a p value, calculated using the right-tailed Fisher’s exact test. In this method, the p value was calculated by comparing the number of induced or repressed genes that participate in a given function or pathway, relative to the total number of occurrences of these genes in all functional/pathway annotations stored in the Ingenuity Knowledge Database.

EMSA was done as described (20). The upper oligonucleotides used for labeling were as follows: hKLF2A: 5′-TGCAAAAGTTTTGTTTTGTTTGTTTTTTTGAGA-3′, hKLF2C: 5′-ACTGGAGCGCGTGTTGACAACGTCGCCGGGGAG-3′. The upper strand oligonucleotide for nonspecific competition was: 5′-GCATGCGCAATCCAGCCAGAGTTGCTAAGG-3′.

ChIP assay was performed using the EpiQuik ChIP kit (Epigentek) per the manufacturer’s instruction. One million PBL cells were used for each Ab reaction. Anti-rabbit IgG was purchased from Santa Cruz Biotechnology, anti-FOXO1 and anti-phospho FOXO1 Abs were purchased from Cell Signaling Technology. Eluted DNA was analyzed by PCR amplification for 25 cycles (94°C for 1 min, 58°C for 1 min, and 72°C for 1 min), separated on 1% agarose gel, transferred to nylon membranes, and hybridized to each probe (same as the ChIP PCR fragments) as described (20). The primer sets used for PCR were as follows: KLF2 forward, 5′-TGCACTTGACGGGCTTATTGAGGT-3′; KLF2 reverse, 5′-ACCGTGTGCACATCACCCTGTAAA-3′; actin forward, 5′-GACCCACCCAGCACATTTAG-3′; actin reverse, 5′-TCGAGCCATAAAAGGCAA-3′.

Overexpression of an active nuclear mutant of FOXO1 has been very informative to uncover the genes regulated by this transcription factor in PTEN null cell lines derived from various tissues (15). Jurkat cells are leukemic T lymphocytes, lacking the lipid phosphatase PTEN (21). In these cells nuclear exclusion of FOXO1 is constitutive (see Fig. 3) because the cells accumulate phosphatidylinositol-3,4,5-trisphosphate and thereby exhibit permanent activation of Akt, the serine/threonine kinase that inactivates FOXO1. Hence, restoring nuclear expression of FOXO1 in these cells with a constitutively active mutant having the three Akt phosphorylation sites mutated to alanine would be predicted to reinstate most of its transcriptional activity.

Jurkat cells were infected with a lentiviral construct coding for FOXO1A3-GFP to carry out expression profiling of isolated mRNA from these cells compared with GFP-expressing cells. RNA was isolated 48 h after infection, before the initiation of the strong antiproliferative and proapoptotic effects of FOXO1 usually observed from day 3 postinfection (data not shown). Several genes encoding receptors involved in homing or trafficking were among those showing the highest changes in expression (Table I). L-selectin was strongly up-regulated; in addition, the two S1P receptors EDG1 and EDG6 were highly induced. We also identified KLF2, a transcription factor that regulates both T cell growth and homing receptor expression (22, 23, 24, 25, 26, 27). By comparison, typical FOXO1 targets like Bim and p27 were less strongly induced.

Table I.

FOXO1A3-inducible genes in Jurkat T cellsa

GeneDescription [Accession No.]Fold Changep Value
LTB Lymphotoxin β (TNF superfamily, member 3) (LTB), transcript variant 1, mRNA [NM_002341] 15.8 2.75E-04 
SELL Selectin L (lymphocyte adhesion molecule 1), mRNA [NM_000655] 13.6 8.00E-04 
C6orf32 mRNA for KIAA0386 gene, partial cds. [AB002384] 13.1 7.80E-04 
EDG6 Endothelial differentiation, G-protein-coupled receptor 6 (EDG6), mRNA [NM_003775] 10.7 4.73E-04 
EDG1 Endothelial differentiation, sphingolipid G-protein-coupled receptor, 1 (EDG1), mRNA [NM_001400] 8.1 5.59E-04 
KLF2 Krüppel-like factor 2 (lung) (KLF2), mRNA [NM_016270] 7.8 4.92E-04 
CD52 CDW52 Ag (CAMPATH-1 Ag) (CDW52), mRNA [NM_001803] 7.7 1.92E-03 
IL16 Interleukin 16 (lymphocyte chemoattractant factor) (IL16), transcript variant 1, mRNA [NM_004513] 7.3 3.32E-04 
CDKN1B Cyclin-dependent kinase inhibitor 1B (p27, Kip1) (CDKN1B), mRNA [NM_004064] 1.8 8.87E-03 
THC2172359 BIM_HUMAN (O43521) Bcl-2-like protein 11 (Bcl2 interacting mediator of cell death), complete [THC2172353] 1.6 1.33E-03 
FASLG Fas ligand (TNF superfamily, member 6) (FASLG), mRNA [NM_000639] 1.1 1.31E-02 
GeneDescription [Accession No.]Fold Changep Value
LTB Lymphotoxin β (TNF superfamily, member 3) (LTB), transcript variant 1, mRNA [NM_002341] 15.8 2.75E-04 
SELL Selectin L (lymphocyte adhesion molecule 1), mRNA [NM_000655] 13.6 8.00E-04 
C6orf32 mRNA for KIAA0386 gene, partial cds. [AB002384] 13.1 7.80E-04 
EDG6 Endothelial differentiation, G-protein-coupled receptor 6 (EDG6), mRNA [NM_003775] 10.7 4.73E-04 
EDG1 Endothelial differentiation, sphingolipid G-protein-coupled receptor, 1 (EDG1), mRNA [NM_001400] 8.1 5.59E-04 
KLF2 Krüppel-like factor 2 (lung) (KLF2), mRNA [NM_016270] 7.8 4.92E-04 
CD52 CDW52 Ag (CAMPATH-1 Ag) (CDW52), mRNA [NM_001803] 7.7 1.92E-03 
IL16 Interleukin 16 (lymphocyte chemoattractant factor) (IL16), transcript variant 1, mRNA [NM_004513] 7.3 3.32E-04 
CDKN1B Cyclin-dependent kinase inhibitor 1B (p27, Kip1) (CDKN1B), mRNA [NM_004064] 1.8 8.87E-03 
THC2172359 BIM_HUMAN (O43521) Bcl-2-like protein 11 (Bcl2 interacting mediator of cell death), complete [THC2172353] 1.6 1.33E-03 
FASLG Fas ligand (TNF superfamily, member 6) (FASLG), mRNA [NM_000639] 1.1 1.31E-02 
a

Only genes with a fold increase ≥ 7 between FOXO1A3-GFP and GFP control cells are shown in the upper part of the table. Fold values for p27 (CDKN1B), Bim (THC2172359), and FasL (FasLG) are shown in the lower part.

To explore the functional relationships among the genes with altered expression and determine the significance of these observations, we used Ingenuity Pathway Analysis software. From the list of genes with a fold increase or decrease (see Table II) ≥3 between FOXO1A3 and control, we conducted a p value analysis of the cellular functions listed in the database. This analysis identified immune response and cellular movement categories as the top functions ordered by significance (Fig. 1,A). Both categories shared several of the previously mentioned homing genes, but also additional key molecules for T cell recirculation like the chemokine receptor CCR7 and integrin α1 (ITGA1) (Fig. 1,B). These results were consistently reproduced in two additional microarray experiments with different cell samples (data not shown). Real-time quantitative RT-PCR further confirmed these results (Fig. 1,C). In accordance with the microarray data Bim and p27 were much less efficiently induced by active FOXO1. We also measured KLF2 protein expression levels in Jurkat T cells expressing active FOXO1 by Western blot. Jurkat cells spontaneously express very low levels of KLF2. The results showed a clear increase 48 h postinfection (Fig. 1 D) showing that increased KLF2 mRNA expression by FOXO1 resulted in a parallel increase at the protein level.

Table II.

FOXO1A3-repressed genes in Jurkat T cellsa

GeneDescriptionFold Changep Value
EGR1 Early growth response 1 (EGR1), mRNA [NM_001964] 7.97 6.46E-04 
AK3 Adenylate kinase 3 (AK3), nuclear gene encoding mitochondrial protein, transcript variant 2, mRNA [NM_013410] 7.03 6.66E-04 
EGR2 Early growth response 2 (Krox-20 homolog, Drosophila) (EGR2), mRNA [NM_000399] 4.73 6.95E-04 
RODH-4 Microsomal NAD+-dependent retinol dehydrogenase 4 (RODH-4), mRNA [NM_003708] 4.07 1.31E-03 
FOS v-fos FBJ murine osteosarcoma viral oncogene homolog (FOS), mRNA [NM_005252] 3.86 3.54E-03 
THC2051867 Q6UZF6 (Q6UZF6) ceramide kinase-like protein, partial (4%) [THC2051867] 3.74 5.35E-04 
PPM1H mRNA for KIAA1157 protein, partial cds. [AB032983] 3.56 3.99E-04 
CD69 CD69 Ag (p60, early T-cell activation Ag) (CD69), mRNA [NM_001781] 3.47 4.88E-04 
EGR3 Early growth response 3 (EGR3), mRNA [NM_004430] 3.36 3.00E-04 
CTSG Cathepsin G (CTSG), mRNA [NM_001911] 3.24 1.26E-04 
A_23_P57836 Unknown 3.20 1.34E-04 
HSPA4L Heat shock 70-kDa protein 4-like (HSPA4L), mRNA [NM_014278] 3.17 7.11E-04 
LOC284422 HSPC323 mRNA, partial cds. [AF161441] 3.00 2.48E-04 
GeneDescriptionFold Changep Value
EGR1 Early growth response 1 (EGR1), mRNA [NM_001964] 7.97 6.46E-04 
AK3 Adenylate kinase 3 (AK3), nuclear gene encoding mitochondrial protein, transcript variant 2, mRNA [NM_013410] 7.03 6.66E-04 
EGR2 Early growth response 2 (Krox-20 homolog, Drosophila) (EGR2), mRNA [NM_000399] 4.73 6.95E-04 
RODH-4 Microsomal NAD+-dependent retinol dehydrogenase 4 (RODH-4), mRNA [NM_003708] 4.07 1.31E-03 
FOS v-fos FBJ murine osteosarcoma viral oncogene homolog (FOS), mRNA [NM_005252] 3.86 3.54E-03 
THC2051867 Q6UZF6 (Q6UZF6) ceramide kinase-like protein, partial (4%) [THC2051867] 3.74 5.35E-04 
PPM1H mRNA for KIAA1157 protein, partial cds. [AB032983] 3.56 3.99E-04 
CD69 CD69 Ag (p60, early T-cell activation Ag) (CD69), mRNA [NM_001781] 3.47 4.88E-04 
EGR3 Early growth response 3 (EGR3), mRNA [NM_004430] 3.36 3.00E-04 
CTSG Cathepsin G (CTSG), mRNA [NM_001911] 3.24 1.26E-04 
A_23_P57836 Unknown 3.20 1.34E-04 
HSPA4L Heat shock 70-kDa protein 4-like (HSPA4L), mRNA [NM_014278] 3.17 7.11E-04 
LOC284422 HSPC323 mRNA, partial cds. [AF161441] 3.00 2.48E-04 
a

Genes showing the highest decrease (≥3) between FOXO1A3-GFP and GFP control cells.

FIGURE 1.

FOXO1 induces a set of genes related to T cell homing in Jurkat T cells. A, Functional analysis genes with Ingenuity Pathway Analysis of expression profiling experiments comparing FOXO1A3-GFP- to GFP-transduced cells. The top cellular categories of the proteins in the first 10 networks with a ≥3-fold change are shown. Top functions were ordered by significance expressed as a p value, calculated using the right-tailed Fisher’s exact test. B, Illustrated in this diagram are the different genes in the immune response and cellular movement categories with a ≥3-fold change. Genes shared by the two categories are also indicated. C, CD62-L, EDG6, EDG1, KLF2, CCR7, Bim, and p27 transcripts were measured 36 h after lentiviral infection of Jurkat cells with GFP, FOXO1A3-GFP, or FOXO1A3-H215R GFP by quantitative real-time PCR analyses. The experiment was repeated four times with different cell preparations and duplicates for each condition. A representative experiment is shown. D, Jurkat cells were infected for 48 h with lentiviruses coding for either GFP or FOXO1A3-GFP. Top panel, GFP-fluorescence analysis of infected cells. Non-infected cells were used as control. Bottom panel, Western blot analysis of whole cell extracts with anti-human KLF2-specific Abs. Anti-human Sam68 blotting of the same membrane was used as a control.

FIGURE 1.

FOXO1 induces a set of genes related to T cell homing in Jurkat T cells. A, Functional analysis genes with Ingenuity Pathway Analysis of expression profiling experiments comparing FOXO1A3-GFP- to GFP-transduced cells. The top cellular categories of the proteins in the first 10 networks with a ≥3-fold change are shown. Top functions were ordered by significance expressed as a p value, calculated using the right-tailed Fisher’s exact test. B, Illustrated in this diagram are the different genes in the immune response and cellular movement categories with a ≥3-fold change. Genes shared by the two categories are also indicated. C, CD62-L, EDG6, EDG1, KLF2, CCR7, Bim, and p27 transcripts were measured 36 h after lentiviral infection of Jurkat cells with GFP, FOXO1A3-GFP, or FOXO1A3-H215R GFP by quantitative real-time PCR analyses. The experiment was repeated four times with different cell preparations and duplicates for each condition. A representative experiment is shown. D, Jurkat cells were infected for 48 h with lentiviruses coding for either GFP or FOXO1A3-GFP. Top panel, GFP-fluorescence analysis of infected cells. Non-infected cells were used as control. Bottom panel, Western blot analysis of whole cell extracts with anti-human KLF2-specific Abs. Anti-human Sam68 blotting of the same membrane was used as a control.

Close modal

L-selectin membrane expression is low in Jurkat cells. Twenty-four hours after transfection with active FOXO1 fused to GFP, we observed a strong up-regulation of L-selectin in the GFP-gated cell population only, confirming the array data. GFP alone had no effect (Fig. 2,A). Similar inductions by active FOXO1 were observed in CEM and SupT1, two other leukemic T cell lines actively sequestering the wild-type molecule within the cytoplasm (Fig. 2,B). Analysis at different time points in Jurkat cells showed a time-dependent and continuous increase of L-selectin induced by active FOXO1 (Fig. 2,C). The wild-type FOXO1 molecule had no effect, consistent with its complete nuclear exclusion. We investigated whether a DNA binding-defective mutant of the active form of FOXO1 (FOXO1A3 H215R-GFP) has a similar effect because several mechanisms of transcriptional regulation by FOXOs have been described in non-lymphoid tumor cells that do not necessarily require this conserved histidine to alter gene regulation (15). As illustrated in Fig. 2,C, the H215R mutation suppressed the effect of the active form of FOXO1 on L-selectin expression. L-selectin membrane expression is known to be partly regulated by the cleavage of its ectodomain by a not yet clearly identified metalloproteinase (maybe ADAM 17, also referred to as TACE, (TNF-α converting enzyme)) or a combination of proteases (28, 29). Because this shedding process could be altered by FOXO1 to explain L-selectin increase, we measured using an ELISA soluble L-selectin in supernatants of Jurkat cells expressing GFP or active FOXO1 (Fig. 2 D). GFP-infected cell supernatants contained very low amounts of sCD62-L, indicating that its spontaneous release was very low in this cellular system. In contrast, cells expressing active FOXO1 released high amounts of the molecule. Thus, increased L-selectin membrane expression induced by active FOXO1 cannot result from some shedding inhibition; on the contrary, our results show that FOXO1 promotes this process, likely by restoring high levels of L-selectin at the plasma membrane. Collectively, these results demonstrate that L-selectin up-regulation by active FOXO1 is at the transcriptional level only and requires its DNA-binding domain.

FIGURE 2.

L-selectin induction by FOXO1 requires its DNA-binding domain. A, Jurkat cells were transfected with GFP or FOXO1A3-GFP and analyzed by FACS 24 h posttransfection. L-selectin expression level histograms in the GFP- and GFP+-gated populations are shown. B, CEM and SupT1 leukemic T cells were transduced using lentiviruses coding for GFP, FOXO1-GFP, or FOXO1A3-GFP. L-selectin levels were measured in the GFP+-gated cell populations 2 days postinfection. Mean fluorescence intensities (MFI) are shown in brackets. FOXO1-GFP localization is also shown in the upper panel. C, Jurkat cells were transfected as in A with GFP, FOXO1-GFP, FOXO1A3-GFP, or FOXO1A3 H215R-GFP, a DNA-binding domain mutant of FOXO1, and the GFP- and GFP+-gated populations analyzed for L-selectin expression after different incubation periods. Data represent the mean ± SD of three independent experiments. D, Jurkat cells were transduced using lentiviruses coding for GFP or FOXO1A3-GFP. sCD62L levels in supernatants were measured by ELISA 2 days postinfection. The mean ± SD of three independent experiments is shown.

FIGURE 2.

L-selectin induction by FOXO1 requires its DNA-binding domain. A, Jurkat cells were transfected with GFP or FOXO1A3-GFP and analyzed by FACS 24 h posttransfection. L-selectin expression level histograms in the GFP- and GFP+-gated populations are shown. B, CEM and SupT1 leukemic T cells were transduced using lentiviruses coding for GFP, FOXO1-GFP, or FOXO1A3-GFP. L-selectin levels were measured in the GFP+-gated cell populations 2 days postinfection. Mean fluorescence intensities (MFI) are shown in brackets. FOXO1-GFP localization is also shown in the upper panel. C, Jurkat cells were transfected as in A with GFP, FOXO1-GFP, FOXO1A3-GFP, or FOXO1A3 H215R-GFP, a DNA-binding domain mutant of FOXO1, and the GFP- and GFP+-gated populations analyzed for L-selectin expression after different incubation periods. Data represent the mean ± SD of three independent experiments. D, Jurkat cells were transduced using lentiviruses coding for GFP or FOXO1A3-GFP. sCD62L levels in supernatants were measured by ELISA 2 days postinfection. The mean ± SD of three independent experiments is shown.

Close modal

FOXO1 is a major downstream target of PI3K via Akt. Hence, to uncover the role of this pathway in L-selectin regulation by FOXO1 and see whether its expression could be directly influenced by 3′-phosphoinositide metabolism, a construct encoding a constitutively active plasma membrane form of PTEN (Myr-PTEN) was expressed in Jurkat cells to hydrolyze accumulated phosphatidylinositol-3,4,5-trisphosphate. We confirmed first by imaging analysis that Myr-PTEN restored most of the nuclear localization of FOXO1 using Myr-PTEN YFP and FOXO1-CFP constructs (Fig. 3, left panel). As shown in Fig. 3, right panel, L-selectin expression was also clearly induced, suggesting that restoration of the nuclear localization of endogenous FOXO1 was sufficient to trigger L-selectin expression. Mainly, this expression was further enhanced when the cells were co-transfected with wild-type FOXO1. In these experimental conditions we obtained L-selectin expression levels similar to those usually achieved with the active mutant overexpressed alone, indicating that wild-type FOXO1 has an activity similar to its active mutant FOXO1A3 to elicit this effect provided that the activity of the PI3K/Akt pathway is repressed.

FIGURE 3.

FOXO1 up-regulates L-selectin membrane levels dependently of PI3K in Jurkat T cells. Left panel, Jurkat cells were transfected with FOXO1-CFP (upper row) or FOXO1-CFP plus Myr-PTEN-YFP (lower row). Cells were analyzed by fluorescence microscopy 24 h posttransfection. Right panel, Jurkat cells were transfected with the indicated constructs and at different time points after transfection L-selectin expression was analyzed in the GFP+-gated cell populations. A Myr-PTEN-GFP construct was used in these experiments to gate L-selectin expression measurements by FACS on the transfected cells only. Data are representative of four independent experiments.

FIGURE 3.

FOXO1 up-regulates L-selectin membrane levels dependently of PI3K in Jurkat T cells. Left panel, Jurkat cells were transfected with FOXO1-CFP (upper row) or FOXO1-CFP plus Myr-PTEN-YFP (lower row). Cells were analyzed by fluorescence microscopy 24 h posttransfection. Right panel, Jurkat cells were transfected with the indicated constructs and at different time points after transfection L-selectin expression was analyzed in the GFP+-gated cell populations. A Myr-PTEN-GFP construct was used in these experiments to gate L-selectin expression measurements by FACS on the transfected cells only. Data are representative of four independent experiments.

Close modal

We next investigated whether these findings could be extended to normal cells by expressing different constructs of FOXO1 in human resting T lymphocytes. As expected, cells transfected with either FOXO1A3 or the wild-type molecule did not show any significant change in L-selectin expression, which remained very high. However, the FOXO1A3 H215R-GFP mutant had a striking effect as it induced a major decrease of L-selectin (Fig. 4, A and C). Because L-selectin is down-regulated after activation, we checked in parallel the expression of several T cell activation molecules to monitor the activation status of the T cells used in these experiments (Fig. 4 B). Activation markers like CD25 and CD71 were not altered by FOXO1A3 H215R-GFP. However, we found that CD69 was strongly up-regulated, a result highly reminiscent of the phenotype described for some mouse KLF2-deficient T cells (see Discussion). Together, these findings suggest that FOXO1 is truly active in resting T cells to control L-selectin expression, with the DNA-binding mutant acting as a dominant negative molecule to inhibit transcriptional control by endogenous FOXO1.

FIGURE 4.

The H215R DNA binding mutant of FOXO1 impairs L-selectin expression on resting human T cells. A, Resting human T cells were transfected with GFP, FOXO1-GFP, FOXO1A3-GFP, or FOXO1A3 H215R-GFP. Viable cells in gate R1 were analyzed 40 h posttransfection for GFP and CD62L expression levels by flow cytometry. B, In the same experiment, cells were also labeled with CD69-, CD25-, or CD71-specific Abs. Overlaid histograms, gated on the GFP positive cells, are shown. C, CD62L and CD69 fluorescence levels within the GFP-positive and GFP-negative cell populations over three independent experiments are shown (mean ± SD, statistical significance assessed by Student’s t test).

FIGURE 4.

The H215R DNA binding mutant of FOXO1 impairs L-selectin expression on resting human T cells. A, Resting human T cells were transfected with GFP, FOXO1-GFP, FOXO1A3-GFP, or FOXO1A3 H215R-GFP. Viable cells in gate R1 were analyzed 40 h posttransfection for GFP and CD62L expression levels by flow cytometry. B, In the same experiment, cells were also labeled with CD69-, CD25-, or CD71-specific Abs. Overlaid histograms, gated on the GFP positive cells, are shown. C, CD62L and CD69 fluorescence levels within the GFP-positive and GFP-negative cell populations over three independent experiments are shown (mean ± SD, statistical significance assessed by Student’s t test).

Close modal

In separate experiments, resting T cells transfected with active FOXO1 were activated to see whether exogenous FOXO1 could block the L-selectin decrease. A significant but limited restoration was noticed 24 and 48 h after activation (data not shown). Thus, we used T cells activated with a combination of anti-CD3 and anti-CD28 Abs to down-regulate L-selectin first. These cells were infected with lentiviruses coding for FOXO1-GFP, FOXO1A3-GFP, or GFP alone and L-selectin expression followed on the GFP-expressing cells at different time points. As shown in Fig. 5,A, CD62L down-regulation induced by activation was strongly neutralized by the active mutant. In contrast, wild-type FOXO1 showed much less effect. As an internal control, no such recovery of CD62L membrane levels could be observed when the analysis was performed on the GFP-negative-gated cell populations corresponding to the uninfected cell fractions. CD62L transcripts were also measured at different time points in GFP-sorted activated T cells after infection with GFP or FOXO1A3-GFP. A time-dependent and spontaneous increase of CD62L transcripts was observed in the GFP-infected cells (Fig. 5 B). However, this recovery was strongly enhanced by active FOXO1, with higher levels of transcripts than in unstimulated T cells at the latest time point.

FIGURE 5.

Active FOXO1 restores L-selectin expression in activated human T cells. A, Human peripheral T cells were stimulated for 24 h with anti-CD3/anti-CD28-coated beads before infection with lentiviral vectors encoding GFP, FOXO1-GFP, or FOXO1A3-GFP. Cells were then analyzed for CD62L expression 24, 48, and 72 h postinfection. Analysis was performed on activated cell populations showing a cell-size increase in the forward and side-scatter analysis because no transduced cells were detected in cells refractory to stimulation (not shown). Transduced and non-transduced cells were discriminated according to GFP expression. The two histograms show CD62L mean fluorescence intensities at the different time points in the GFP-negative and GFP-positive cells. As a comparison, CD62L expression before stimulation, at time 0 of the experiment, is also shown in each histogram (hatched symbols). Data are representative of two independent experiments. B, Human peripheral T cells were stimulated for 24 h with anti-CD3/anti-CD28-coated beads before infection with lentiviral vectors encoding GFP or FOXO1A3-GFP. Four days postinfection GFP-positive cells were sorted and re-cultured, and CD62L transcripts measured at different time points. Results were expressed as the percentage of CD62L transcripts measured before stimulation. Data are representative of two independent experiments.

FIGURE 5.

Active FOXO1 restores L-selectin expression in activated human T cells. A, Human peripheral T cells were stimulated for 24 h with anti-CD3/anti-CD28-coated beads before infection with lentiviral vectors encoding GFP, FOXO1-GFP, or FOXO1A3-GFP. Cells were then analyzed for CD62L expression 24, 48, and 72 h postinfection. Analysis was performed on activated cell populations showing a cell-size increase in the forward and side-scatter analysis because no transduced cells were detected in cells refractory to stimulation (not shown). Transduced and non-transduced cells were discriminated according to GFP expression. The two histograms show CD62L mean fluorescence intensities at the different time points in the GFP-negative and GFP-positive cells. As a comparison, CD62L expression before stimulation, at time 0 of the experiment, is also shown in each histogram (hatched symbols). Data are representative of two independent experiments. B, Human peripheral T cells were stimulated for 24 h with anti-CD3/anti-CD28-coated beads before infection with lentiviral vectors encoding GFP or FOXO1A3-GFP. Four days postinfection GFP-positive cells were sorted and re-cultured, and CD62L transcripts measured at different time points. Results were expressed as the percentage of CD62L transcripts measured before stimulation. Data are representative of two independent experiments.

Close modal

Independent studies have shown that KLF2-deficient peripheral mouse T cells produced by gene targeting had altered migration capabilities (22, 25). In addition, KLF2 seems to transactivate the promoter of CD62L in murine T cells (25, 27). Thus, we investigated whether KLF2 expression also induced CD62L in human resting T cells. As shown in Fig. 6, KLF2 causes significant CD62L up-regulation. Strikingly, this restoration closely paralleled the effects of FOXO1A3, suggesting a coordinated regulation. The human KLF2 promoter contains two motifs with sequence similarity to the canonical FOXO binding site TTGTTTAC, at −879 and −275 relative to the transcription start site, both of which are conserved in mice. EMSAs showed that FOXO1 can bind to sites in the human KLF2 promoter in vitro (Fig. 7, left panel). Importantly, chromatin immunoprecipitation showed that endogenous FOXO1 is bound to the KLF2 promoter in resting human peripheral blood T cells (Fig. 7, right panel), but not to the human actin promoter region. We did not detect any specific binding of FOXO1 to the human SELL proximal promoter region, which contains no consensus FOXO binding motifs (data not shown).

FIGURE 6.

KLF2 and active FOXO1 induce parallel restoration of L-selectin levels in activated human T cells. Purified human peripheral T cells were stimulated for 24 h with anti-CD3/anti-CD28-coated beads before infection with lentiviral vectors encoding GFP, FOXO1A3-GFP, or KLF2-GFP. CD62L membrane expression was analyzed on GFP-gated populations at different time points postinfection.

FIGURE 6.

KLF2 and active FOXO1 induce parallel restoration of L-selectin levels in activated human T cells. Purified human peripheral T cells were stimulated for 24 h with anti-CD3/anti-CD28-coated beads before infection with lentiviral vectors encoding GFP, FOXO1A3-GFP, or KLF2-GFP. CD62L membrane expression was analyzed on GFP-gated populations at different time points postinfection.

Close modal
FIGURE 7.

FOXO1 binds the KLF2 promoter in vivo. Left panel, Extracts of Cos7 cells transfected with FOXO1 or empty expression vector were incubated with 32P-end-labeled oligonucleotide probes. Specificity for two sites (hKLF2-A, hKLF2-C) is shown using anti-FOXO1 Abs that supershift only the specific complex, whereas a 100-fold molar excess of cold competitor oligomers reduces the detection of all complexes. An excess of nonspecific (NS) cold oligomers had no effect. A representative result of three independent experiments is shown. Right panel, Chromatin proteins in human PBT cells were cross-linked to DNA by formaldehyde and immunoprecipitated using anti-FOXO1 or anti-phospho-FOXO1 Abs, or nonimmune rabbit IgG control. Eluted DNAs were diluted to 10- to 50-fold for PCR and were detected by Southern blotting for KLF2 and actin promoter amplicons. Comparable results were obtained using PBT from a second donor.

FIGURE 7.

FOXO1 binds the KLF2 promoter in vivo. Left panel, Extracts of Cos7 cells transfected with FOXO1 or empty expression vector were incubated with 32P-end-labeled oligonucleotide probes. Specificity for two sites (hKLF2-A, hKLF2-C) is shown using anti-FOXO1 Abs that supershift only the specific complex, whereas a 100-fold molar excess of cold competitor oligomers reduces the detection of all complexes. An excess of nonspecific (NS) cold oligomers had no effect. A representative result of three independent experiments is shown. Right panel, Chromatin proteins in human PBT cells were cross-linked to DNA by formaldehyde and immunoprecipitated using anti-FOXO1 or anti-phospho-FOXO1 Abs, or nonimmune rabbit IgG control. Eluted DNAs were diluted to 10- to 50-fold for PCR and were detected by Southern blotting for KLF2 and actin promoter amplicons. Comparable results were obtained using PBT from a second donor.

Close modal

In vivo, naive T lymphocytes travel between the bloodstream and secondary lymphoid organs tissues to be brought into contact with Ags. L-selectin is required for this constitutive trafficking. Its high expression on the surface of naive T cells is essential for tethering and subsequent rolling of circulating lymphocytes along endothelial cells lining the walls of the high endothelial venules of peripheral lymph nodes, a crucial step for transmigration (30). L-selectin is actively shed after activation, a process that has essential functions, especially to prevent activated T cells from reentering lymph nodes (31, 32). However, L-selectin is also regulated at the transcriptional level because T cells expressing a shedding-resistant molecule also display a strong reduction of L-selectin membrane levels after a few days of stimulation (31). Likewise, the transcription rate of L-selectin in murine T cells turns down late after activation to maintain the low expression phenotype, suggesting that a constitutive transcriptional control is switched off by activation (33). We show in this report that this control can be exerted by the opposing actions of PI3K and FOXO1 in human T lymphocytes.

Central to the regulation of FOXOs by PI3K is a shuttling process triggered after their phosphorylation by Akt. As a consequence, FOXO molecules are repelled from the nucleus and their transcriptional activities switched off in the activated cell. We showed previously that the nuclear exclusion of FOXO1 in human resting T cells activated by Ag was almost complete a few minutes after the first contact with the APC (5). Having shown herein that endogenous FOXO1 exerts a steady-state control on L-selectin expression in resting T cells, this strongly indicates that this transcriptional control is likely turned off very early at the onset of the response. In vitro PI3K activation and FOXO1 nuclear exclusion are two sustained processes lasting for hours, as long as the contact persists between the T cell and the APC (1, 2, 5). T cells also make prolonged contacts with APCs in secondary lymphoid organs in vivo (34). However, it is not known whether PI3K activation is also maintained in these conjugates and more generally how the status of the PI3K/Akt pathway evolves and controls FOXO1 activity during the different phases of the response from Ag recognition to clonal cell expansion and memory cell generation. A restoration of L-selectin levels at a given stage of the response when PI3K activation is ended and FOXO1 is back in the nucleus would be logical. However, the relationship between PI3K, FOXO1, and L-selectin might be more complex, especially in memory T cell subsets because central memory human T cells express high levels of L-selectin and exhibit significant phosphorylation of FOXO3, another FOXO protein (35).

Several mechanisms of transcriptional regulation by FOXOs have been described, especially in PTEN null tumor cells that collectively demonstrate the complexity of FOXO DNA binding and transcriptional capabilities depending on the promoter and the cellular context. Some gene regulation events require direct binding to DNA, whereas others do not require this kind of recognition to activate transcription. The latter mechanism was mainly inferred from profiling experiments comparing genes induced by FOXO1A3 and FOXO1A3 H215R, a mutant which lacks a critical residue in helix 3 of the so-called winged-helix domain of FOXO1 to bind DNA (15). Our data show that FOXO1 up-regulates T cell homing molecules in a manner dependent on FOXO1 binding to DNA. Active FOXO1 has been described to regulate T cell growth induced by Ag in normal lymphocytes (5, 36). Therefore we tried to determine whether the H215R mutant could also discriminate these two biological effects. Our results demonstrate that FOXO1A3 and FOXO1A3 H215R share the same potential to inhibit T cell growth (data not shown). Thus the H215R mutation uncouples the control of cell growth and homing, suggesting that distinct mechanisms support these biological functions of FOXO1 in human T lymphocytes. Finally, our results showing that FOXO1A3 H215R strongly reduced L-selectin expression in resting human T lymphocytes also suggest that FOXO1 requires some molecular aid to shape its DNA-binding-dependent activities. To explain this striking effect, it may be speculated that the mutant is competing away one specific partner that cooperate with endogenous FOXO1 to activate the L-selectin gene. Whatever the identity of the cooperating partner, this result supports the idea that FOXO1 is constitutively functioning and essential in resting T cells to maintain L-selectin levels.

Prototypic targets of FOXOs were not strongly induced by active FOXO1 in Jurkat T cells (see Table I), especially p27, a known inhibitor of cell cycle progression in various cell types. However, the induction of p27 is not mandatory for the regulation of proliferation by FOXO1 in non-lymphoid PTEN null tumor cells (15, 37). In these cell systems, FOXO1 appears instead to regulate cell cycle progression mainly through transcriptional inhibition of D-type cyclins. These genes were not found repressed in Jurkat cells (not shown); instead, we found a marked down-regulation of several members of the early growth response gene family of transcription factors (Table II). Notably, the most efficiently repressed molecule, EGR1, was first identified as a putative G0/G1 switch regulatory gene in human blood lymphocytes (38) and is described as a direct regulator of multiple tumor suppressors including TGFβ, PTEN, and p53 (39). We also found that FOXO1A3 moderately induced the expression of cyclin G2, an inhibitor of cell growth in tumor B cell lines and in activated B cells (20, 40) and thus might also account for the inhibitory effects of FOXO1A3 on cell cycling. Finally we cannot exclude the possibility that KLF2 induction by FOXO1 also participates in the regulation of T cell growth, especially in the Jurkat T cell leukemic model where KLF2 can block cell proliferation (23).

Deletion or overexpression of KLF2 has been found to cause profound changes in homing and recirculation in mouse T cells (25, 26, 27). Naive KLF2-deficient T cells are absent from lymphoid organs, which only contain T cell populations showing a spontaneously “activated” phenotype with high levels of CD69 along with reduced levels of CD62L (25, 26, 27). It was therefore striking to observe that resting human T cells overexpressing the H215R mutant share the same phenotypic changes associated with opposite regulation of CD62L and CD69 membrane expression levels. We also found a marked transcriptional inhibition of CD69 by active FOXO1 (see Table II), an observation that fits very well with recent reports showing a mutual antagonistic regulation of EDG1 and CD69 in T lymphocytes (41). Of note, we detected in additional array experiments an opposite ability of the active FOXO1 molecule and the H215R mutant to regulate EDG1 and CD69 in Jurkat cells (not shown). Finally, everything appears as if FOXO1A3 H215R overexpression recapitulates in human resting T cells some phenotypic aspects observed after KLF2 inactivation in the mouse system. One of the most highly induced genes by active FOXO1 was KLF2 itself, and FOXO1 binds to the KLF2 promoter in resting T cells. Hence, several genes identified in our array experiments could be induced by FOXO1 indirectly, through KLF2. This might be especially true for EDG1 which is a direct target of KLF2 (25, 27), but maybe also for additional genes like SELL (L-selectin), because its expression is strongly increased in human resting T cells overexpressing KLF2 (Fig. 6). However, preliminary experiments in Jurkat showed that active FOXO1 was still able to significantly induce L-selectin in cells where KLF2 has been partially knock-down (data not shown). Thus, a direct regulation of CD62L by KLF2 itself is still questionable, in agreement with the recent observation that most KLF2-deficient murine T cells express normal levels of L-selectin (26). Importantly, in line with our conclusions, a recent report has shown that activation of the PI3K pathway negatively regulates CD62L, perhaps via inhibition of KLF-2 transcription (42). An inhibition by rapamycin of this PI3K-dependent transcriptional effect was also found, suggesting a dual participation of mTOR and FOXO1 in the regulation of CD62L in T cells. However, because prolonged treatments with rapamycin were used in this study, not only mTOR-C1, but also mTOR-C2 could have been blocked, as previously reported (43). Thus the possibility remains that inhibition of mTOR-C2 could have led to an inhibition of Akt upstream of FOXO1.

In conclusion, together with previous observations showing that another forkhead transcription factor, FoxJ1, can regulate mouse T cell egress (44), our findings open a new unanticipated chapter in the biology of this large family of molecules to regulate T cell trafficking. Indeed, we have identified a new function of FOXO1 that regulates various immune response genes and is opposed by PI3K/Akt signaling in human T cells. This mechanism is likely a very important means by which T cells can shape the expression of several essential leukocyte membrane receptors involved in T cell migration and tissue localization. How the parallel regulation of L-selectin, EDG1/6, and other homing molecules or chemokine receptors by FOXO1 and KLF2 can adjust T cell entry and egress from lymphoid organs, but also presumably their itinerary at the periphery, deserves additional studies. However, it seems now that the harmonious functioning of these different processes can be endowed by the PI3K pathway in human T cells, a conclusion that also adds a new perspective to therapeutic strategies that interfere with the adaptive immune response. Considering the intensive pharmacological studies and the increasing number of specific drugs under development to target the PI3K pathway, possible effects on lymphocyte trafficking should be addressed.

We thank Huges Ripoche for deposition of the microarray data and the cytometry and sequencing core facilities at the Cochin Institute for technical assistance.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by grants from Ligue Nationale contre le Cancer, Institut National de la Santé et de la Recherche Médicale, and Centre National de la Recherche Scientifique. S.F. was supported by a Ligue Nationale contre le Cancer Fellowship and F.C. by a Ministère de l’Education Nationale et de La Recherche doctoral Fellowship.

4

Abbreviations used in this paper: FOXO, forkhead box O; PTEN, phosphatase and tensin homolog; CFP, cyan fluorescent protein; ChIP, chromatin immunoprecipitation; sCD62L, soluble CD62L.

1
Costello, P. S., M. Gallagher, D. A. Cantrell.
2002
. Sustained and dynamic inositol lipid metabolism inside and outside the immunological synapse.
Nat. Immunol.
3
:
1082
-1089.
2
Harriague, J., G. Bismuth.
2002
. Imaging antigen-induced PI3K activation in T cells.
Nat. Immunol.
3
:
1090
-1096.
3
Birkenkamp, K. U., P. J. Coffer.
2003
. Regulation of cell survival and proliferation by the FOXO (Forkhead box, class O) subfamily of Forkhead transcription factors.
Biochem. Soc. Trans.
31
:
292
-297.
4
Coffer, P. J., B. M. Burgering.
2004
. Forkhead-box transcription factors and their role in the immune system.
Nat. Rev. Immunol.
4
:
889
-899.
5
Fabre, S., V. Lang, J. Harriague, A. Jobart, T. G. Unterman, A. Trautmann, G. Bismuth.
2005
. Stable activation of phosphatidylinositol 3-kinase in the T cell immunological synapse stimulates Akt signaling to FoxO1 nuclear exclusion and cell growth control.
J. Immunol.
174
:
4161
-4171.
6
Biggs, W. H., 3rd, J. Meisenhelder, T. Hunter, W. K. Cavenee, K. C. Arden.
1999
. Protein kinase B/Akt-mediated phosphorylation promotes nuclear exclusion of the winged helix transcription factor FKHR1.
Proc. Natl. Acad. Sci. USA
96
:
7421
-7426.
7
Rena, G., S. Guo, S. C. Cichy, T. G. Unterman, P. Cohen.
1999
. Phosphorylation of the transcription factor forkhead family member FKHR by protein kinase B.
J. Biol. Chem.
274
:
17179
-17183.
8
Zhang, X., L. Gan, H. Pan, S. Guo, X. He, S. T. Olson, A. Mesecar, S. Adam, T. G. Unterman.
2002
. Phosphorylation of serine 256 suppresses transactivation by FKHR (FOXO1) by multiple mechanisms: direct and indirect effects on nuclear/cytoplasmic shuttling and DNA binding.
J. Biol. Chem.
277
:
45276
-45284.
9
Tothova, Z., R. Kollipara, B. J. Huntly, B. H. Lee, D. H. Castrillon, D. E. Cullen, E. P. McDowell, S. Lazo-Kallanian, I. R. Williams, C. Sears, et al
2007
. FoxOs are critical mediators of hematopoietic stem cell resistance to physiologic oxidative stress.
Cell
128
:
325
-339.
10
Paik, J. H., R. Kollipara, G. Chu, H. Ji, Y. Xiao, Z. Ding, L. Miao, Z. Tothova, J. W. Horner, D. R. Carrasco, et al
2007
. FoxOs are lineage-restricted redundant tumor suppressors and regulate endothelial cell homeostasis.
Cell
128
:
309
-323.
11
Castrillon, D. H., L. Miao, R. Kollipara, J. W. Horner, R. A. DePinho.
2003
. Suppression of ovarian follicle activation in mice by the transcription factor Foxo3a.
Science
301
:
215
-218.
12
Hosaka, T., W. H. Biggs, 3rd, D. Tieu, A. D. Boyer, N. M. Varki, W. K. Cavenee, K. C. Arden.
2004
. Disruption of forkhead transcription factor (FOXO) family members in mice reveals their functional diversification.
Proc. Natl. Acad. Sci. USA
101
:
2975
-2980.
13
Bakker, W. J., T. B. van Dijk, M. Parren-van Amelsvoort, A. Kolbus, K. Yamamoto, P. Steinlein, R. G. Verhaak, T. W. Mak, H. Beug, B. Lowenberg, M. von Lindern.
2007
. Differential regulation of Foxo3a target genes in erythropoiesis.
Mol. Cell. Biol.
27
:
3839
-3854.
14
Daly, C., V. Wong, E. Burova, Y. Wei, S. Zabski, J. Griffiths, K. M. Lai, H. C. Lin, E. Ioffe, G. D. Yancopoulos, J. S. Rudge.
2004
. Angiopoietin-1 modulates endothelial cell function and gene expression via the transcription factor FKHR (FOXO1).
Genes Dev.
18
:
1060
-1071.
15
Ramaswamy, S., N. Nakamura, I. Sansal, L. Bergeron, W. R. Sellers.
2002
. A novel mechanism of gene regulation and tumor suppression by the transcription factor FKHR.
Cancer Cell
2
:
81
-91.
16
Furuyama, T., K. Kitayama, Y. Shimoda, M. Ogawa, K. Sone, K. Yoshida-Araki, H. Hisatsune, S. Nishikawa, K. Nakayama, K. Ikeda, N. Motoyama, N. Mori.
2004
. Abnormal angiogenesis in Foxo1 (Fkhr)-deficient mice.
J. Biol. Chem.
279
:
34741
-34749.
17
Nakae, J., W. H. Biggs, 3rd, T. Kitamura, W. K. Cavenee, C. V. Wright, K. C. Arden, D. Accili.
2002
. Regulation of insulin action and pancreatic β-cell function by mutated alleles of the gene encoding forkhead transcription factor Foxo1.
Nat. Genet.
32
:
245
-253.
18
Lin, L., J. D. Hron, S. L. Peng.
2004
. Regulation of NF-κB, Th activation, and autoinflammation by the forkhead transcription factor Foxo3a.
Immunity
21
:
203
-213.
19
Dardalhon, V., B. Herpers, N. Noraz, F. Pflumio, D. Guetard, C. Leveau, A. Dubart-Kupperschmitt, P. Charneau, N. Taylor.
2001
. Lentivirus-mediated gene transfer in primary T cells is enhanced by a central DNA flap.
Gene. Ther.
8
:
190
-198.
20
Chen, J., I. Yusuf, H. M. Andersen, D. A. Fruman.
2006
. FOXO transcription factors cooperate with δ EF1 to activate growth suppressive genes in B lymphocytes.
J. Immunol.
176
:
2711
-2721.
21
Shan, X., M. J. Czar, S. C. Bunnell, P. Liu, Y. Liu, P. L. Schwartzberg, R. L. Wange.
2000
. Deficiency of PTEN in Jurkat T cells causes constitutive localization of Itk to the plasma membrane and hyperresponsiveness to CD3 stimulation.
Mol. Cell. Biol.
20
:
6945
-6957.
22
Kuo, C. T., M. L. Veselits, J. M. Leiden.
1997
. LKLF: A transcriptional regulator of single-positive T cell quiescence and survival.
Science
277
:
1986
-1990.
23
Buckley, A. F., C. T. Kuo, J. M. Leiden.
2001
. Transcription factor LKLF is sufficient to program T cell quiescence via a c-Myc-dependent pathway.
Nat. Immunol.
2
:
698
-704.
24
Wu, J., J. B. Lingrel.
2004
. KLF2 inhibits Jurkat T leukemia cell growth via upregulation of cyclin-dependent kinase inhibitor p21WAF1/CIP1.
Oncogene
23
:
8088
-8096.
25
Carlson, C. M., B. T. Endrizzi, J. Wu, X. Ding, M. A. Weinreich, E. R. Walsh, M. A. Wani, J. B. Lingrel, K. A. Hogquist, S. C. Jameson.
2006
. Kruppel-like factor 2 regulates thymocyte and T-cell migration.
Nature
442
:
299
-302.
26
Sebzda, E., Z. Zou, J. S. Lee, T. Wang, M. L. Kahn.
2008
. Transcription factor KLF2 regulates the migration of naive T cells by restricting chemokine receptor expression patterns.
Nat. Immunol.
9
:
292
-300.
27
Bai, A., H. Hu, M. Yeung, J. Chen.
2007
. Kruppel-like factor 2 controls T cell trafficking by activating L-selectin (CD62L) and sphingosine-1-phosphate receptor 1 transcription.
J. Immunol.
178
:
7632
-7639.
28
Smalley, D. M., K. Ley.
2005
. L-selectin: mechanisms and physiological significance of ectodomain cleavage.
J. Cell. Mol. Med.
9
:
255
-266.
29
Spertini, O., A. S. Freedman, M. P. Belvin, A. C. Penta, J. D. Griffin, T. F. Tedder.
1991
. Regulation of leukocyte adhesion molecule-1 (TQ1, Leu-8) expression and shedding by normal and malignant cells.
Leukemia
5
:
300
-308.
30
Warnock, R. A., S. Askari, E. C. Butcher, U. H. von Andrian.
1998
. Molecular mechanisms of lymphocyte homing to peripheral lymph nodes.
J. Exp. Med.
187
:
205
-216.
31
Venturi, G. M., L. Tu, T. Kadono, A. I. Khan, Y. Fujimoto, P. Oshel, C. B. Bock, A. S. Miller, R. M. Albrecht, P. Kubes, et al
2003
. Leukocyte migration is regulated by L-selectin endoproteolytic release.
Immunity
19
:
713
-724.
32
Galkina, E., K. Tanousis, G. Preece, M. Tolaini, D. Kioussis, O. Florey, D. O. Haskard, T. F. Tedder, A. Ager.
2003
. L-selectin shedding does not regulate constitutive T cell trafficking but controls the migration pathways of antigen-activated T lymphocytes.
J. Exp. Med.
198
:
1323
-1335.
33
Chao, C. C., R. Jensen, M. O. Dailey.
1997
. Mechanisms of L-selectin regulation by activated T cells.
J. Immunol.
159
:
1686
-1694.
34
Germain, R. N., M. K. Jenkins.
2004
. In vivo antigen presentation.
Curr. Opin. Immunol.
16
:
120
-125.
35
Riou, C., B. Yassine-Diab, J. Van Grevenynghe, R. Somogyi, L. D. Greller, D. Gagnon, S. Gimmig, P. Wilkinson, Y. Shi, M. J. Cameron, et al
2007
. Convergence of TCR and cytokine signaling leads to FOXO3a phosphorylation and drives the survival of CD4+ central memory T cells.
J. Exp. Med.
204
:
79
-91.
36
Stahl, M., P. F. Dijkers, G. J. Kops, S. M. Lens, P. J. Coffer, B. M. Burgering, R. H. Medema.
2002
. The forkhead transcription factor FoxO regulates transcription of p27Kip1 and Bim in response to IL-2.
J. Immunol.
168
:
5024
-5031.
37
Schmidt, M., S. Fernandez de Mattos, A. van der Horst, R. Klompmaker, G. J. Kops, E. W. Lam, B. M. Burgering, R. H. Medema.
2002
. Cell cycle inhibition by FoxO forkhead transcription factors involves downregulation of cyclin D.
Mol. Cell. Biol.
22
:
7842
-7852.
38
Forsdyke, D. R..
1985
. cDNA cloning of mRNAS which increase rapidly in human lymphocytes cultured with concanavalin-A and cycloheximide.
Biochem. Biophys. Res. Commun.
129
:
619
-625.
39
Baron, V., E. D. Adamson, A. Calogero, G. Ragona, D. Mercola.
2006
. The transcription factor Egr1 is a direct regulator of multiple tumor suppressors including TGFβ1, PTEN, p53, and fibronectin.
Cancer Gene Ther.
13
:
115
-124.
40
Martinez-Gac, L., M. Marques, Z. Garcia, M. R. Campanero, A. C. Carrera.
2004
. Control of cyclin G2 mRNA expression by forkhead transcription factors: novel mechanism for cell cycle control by phosphoinositide 3-kinase and forkhead.
Mol. Cell. Biol.
24
:
2181
-2189.
41
Shiow, L. R., D. B. Rosen, N. Brdickova, Y. Xu, J. An, L. L. Lanier, J. G. Cyster, M. Matloubian.
2006
. CD69 acts downstream of interferon-α/β to inhibit S1P1 and lymphocyte egress from lymphoid organs.
Nature
440
:
540
-544.
42
Sinclair, L. V., D. Finlay, C. Feijoo, G. H. Cornish, A. Gray, A. Ager, K. Okkenhaug, T. J. Hagenbeek, H. Spits, D. A. Cantrell.
2008
. Phosphatidylinositol-3-OH kinase and nutrient-sensing mTOR pathways control T lymphocyte trafficking.
Nat. Immunol.
9
:
513
-521.
43
Sarbassov, D. D., S. M. Ali, S. Sengupta, J. H. Sheen, P. P. Hsu, A. F. Bagley, A. L. Markhard, D. M. Sabatini.
2006
. Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB.
Mol. Cell
22
:
159
-168.
44
Srivatsan, S., S. L. Peng.
2005
. Cutting edge: Foxj1 protects against autoimmunity and inhibits thymocyte egress.
J. Immunol.
175
:
7805
-7809.