Semaphorin 7A Is a Negative Regulator of T Cell Responses

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector Immunity 24, 591–600, May 2006 ª2006 Elsevier Inc. DOI 10.1016/j.immuni.2006.03.013 Semaphorin 7A Is a Negative Regulator of T Cell Responses

Agnieszka K. Czopik,1,2,5 Margaret S. Bynoe,1,6 responses, and Sema4D binds CD72 and turns off its in- Noah Palm,1 Cedric S. Raine,4 hibitory signaling during T cell-mediated B cell activa- and Ruslan Medzhitov1,3,* tion (Kumanogoh and Kikutani, 2001). Another immune 1 Section of Immunobiology semaphorin, Sema4A is expressed by DCs, B cells, 2 Department of Molecular, Cellular and Developmental and by activated T cells. Sema4A costimulates T cell ac- Biology tivation by interaction with receptor Tim-2 (Kumanogoh 3 Howard Hughes Medical Institute et al., 2002a). Studies of Sema4A-deficient mice show Yale University that DC-derived Sema4A is important for T cell priming, New Haven, Connecticut 06520 while T cell-derived Sema4A is involved in developing 4 Departments of Pathology, Neurology Th1 responses (Kumanogoh et al., 2005). and Neuroscience Semaphorin 7A, the only GPI-linked protein in sema- Albert Einstein College of Medicine phorin family, is prominently expressed in the embryo New York, New York 10461 and in the lymphoid organs and the nervous system of adult mice (Sato and Takahashi, 1998). Sema7A is a cellular homolog of viral semaphorins encoded by vaccinia Summary and herpesvirus (Comeau et al., 1998; Lange et al., 1998; Xu et al., 1998) and was demonstrated to bind to Semaphorins play an essential role in axonal guidance, the cellular receptor Plexin C1 in vitro (Tamagnone and emerging evidence points to diverse functions of et al., 1999). Sema7A can induce monocyte chemotaxis several Semaphorin family members in the immune and cytokine production and is expressed in activated system. Semaphorin 7A (Sema7A) promotes axonal lymphocytes and thymocytes (Holmes et al., 2002; Mine growth in the central nervous system. Here, we show et al., 2000), suggesting an immune function for this mol- that Sema7A also plays a critical role in negative regu- ecule. In addition, Sema7A is highly expressed on most lation of T cell activation and function. T cells deficient human T lymphocytes and natural killer cells (Angelisova in Sema7A exhibit enhanced homeostatic and antigen- et al., 1999). Sema7A knockout animals exhibited dimin- induced proliferative response. Moreover, autoreactive ished axonal tracts formation, while treatment with solu- Sema7A-deficient T cells mediate aggressive auto- ble Sema7A enhanced axonal outgrowth (Pasterkamp immune disease. The deficiency in Sema7A leads to et al., 2003). Sema7A binding to neurons induced activa- defective TCR downmodulation and T cell hyperres- tion of focal adhesion kinase (FAK) and extracellular reg- ponsiveness. These results demonstrate an important ulated kinases (ERKs) (Pasterkamp et al., 2003). These ef- role of Sema7A in limiting autoimmune responses fects of Sema7A were mediated by b1-integrin and were and add to growing evidence of shared signaling path- independent of Plexin C1 (Pasterkamp et al., 2003). Thus, ways used by the immune and nervous systems. Sema7A binds to at least two receptors with very different activities, and the significance of these interactions in Introduction various tissue types remains to be clarified. The expression of Sema7A on T lymphocytes and the Semaphorins are neuronal guidance factors that were acquisition of its homolog by two viral families suggest initially characterized by their ability to inhibit axonal mi- that this molecule may have an important specific func- gration via chemorepulsive mechanisms, which induce tion in lymphocytes. Here, we investigate the role of cytoskeletal rearrangement and growth cone collapse Sema7A in T cells and demonstrate that it plays a T (Kolodkin et al., 1992, 1993; Luo et al., 1993). Semaphor- cell-intrinsic inhibitory role and is essential for limiting ins transduce the repulsive signal through receptors of T cell-mediated autoimmunity. the plexin family (Tamagnone et al., 1999), and in addition, two of the class IV semaphorins interact with recep- Results tors structurally unrelated to plexins (Kumanogoh et al., 2000, 2002b). Sema7A-Deficient T Cells Are Hyperresponsive Several members of the semaphorin family have been Sema7A is expressed in naive CD4 and CD8 T cells, and characterized with respect to their function in immunity. its expression is increased in activated CD4 T cells and Sema4D is a 150 kDa transmembrane protein, which be- remains elevated in differentiated T helper 1 (Th1) lym- longs to the class IV semaphorin subfamily (Delaire phocytes (Figure S1A). Sema7A mRNA is detected in et al., 1998). In lymphoid organs, Sema4D is abundant dendritic cells (DCs) and macrophages but is not promi- on resting T cells. The receptor for Sema4D in lymphoid nently expressed in B cells (Figures S1A and S1B). To address the function of Sema7A in the immune system, tissues is CD72 (Elhabazi et al., 2003; Kumanogoh et al., 2/2 2000). CD72 functions as a negative regulator of B cell we first compared T cell responses in Sema7A animals and their wt littermates. Mice were immunized with a model protein antigen, chicken ovalbumin (OVA) *Correspondence: [email protected] emulsified in adjuvant. Six days after immunization, 5 Present address: Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139. CD4 T cells were purified from the draining lymph nodes 6 Present address: Department of Microbiology and Immunology, and restimulated with wt antigen-presenting cells (APC) Cornell University, Ithaca, New York 14853. pulsed with different doses of OVA. Sema7A-deficient Immunity 592

T cells showed increased proliferative response to the antigen, compared to the wt T cells (Figure 1A), suggesting that Sema7A may have a negative regulatory role in T cell activation. Sema7A deficiency leading to increased proliferation could reside either in antigen-presenting cells or in T cells. To distinguish between these possibil- ities, we performed in vitro T cell activation assays with naive OTII TCR transgenic CD4 T cells stimulated with cognate peptide presented by wt APCs. OTII CD4 T cells on Sema7A2/2 background showed dramatically increased proliferative response to cognate antigen presented by wt APC, compared to OTII Sema7A+/+ T cells (Figure 1B). This result demonstrated that the increased T cell responsiveness was due to the lack of Sema7A in T cells and suggested that this molecule may have an inhibitory function that regulates proliferation of T cells. In order to further study the inhibitory role of Sema7A in T cell activation, we prepared recombinant soluble Sema7A-Fc. Addition of 100 mg/ml of soluble Sema7A- Fc to wt OTII transgenic T cells stimulated with OVA- pulsed DCs resulted in a significant decrease in T cell proliferation, compared to addition of Fc alone (Figure 1C). This result provides further indication for an inhibitory function of Sema7A in T cell activation. Figure 1. Function of Sema7A in T Cell Proliferation We next addressed the role of Sema7A expression in (A and B) Enhanced CD4 T cell responses in Sema7A-deficient mice DCs by comparing the maturation and T cell activation compared with wild-type mice. (A) CD4 T cell responses from OVA 2/2 by wt and Sema7A2/2 DCs. No significant differences immunized mice. (B) Proliferation of naive OTII CD4 Sema7A or Sema7A+/+ T cells activated with wild-type APCs. were observed in DC maturation induced by various (C) Addition of soluble, recombinant Sema7A-FC (100 mg/ml) inhibits 2/2 TLR ligands in vitro (Figure S2A). Wt and Sema7A proliferation of naive OTII CD4 T cells stimulated with Ovalbumin- DCs had equivalent ability to activate naive T lympho- pulsed DCs. As a control, recombinant IgG-FC fragment was added cytes in vitro (Figure S2B) and in vivo (Figure S2C). These at 100 mg/ml. T cell-proliferative responses are expressed as the findings indicate that the increased proliferative re- average 6 SE of three animals per group in (A) or as the mean 6 sponse seen in Sema7A2/2 T cells is likely to be T cell SE of triplicate cultures in (B) and (C). autonomous and independent of Sema7A deficiency in APCs. measured on day 10 postimmunization were elevated (Figure 2C), and, consistently, a larger number of infil- Severe EAE in Sema7A-Deficient Mice trating lymphocytes was found in the spinal cord sec- To examine the in vivo function of Sema7A in CD4 tions from moribund Sema7A2/2 animals compared T cells, we employed the animal model of multiple scle- to wt animals on the same day postimmunization rosis, experimental autoimmune encephalomyelitis (Figure 2D). Additionally, the level of IL-12 in the serum (EAE), which is an inflammatory disease of the central of Sema7A2/2 mice was also found to be elevated nervous system (CNS) mediated primarily by Th1 cells (Figure S3A). As the aggravated autoimmunity could re- (Kuchroo et al., 2002; Wong et al., 1999). Upon induction sult from a deficiency in cytokines that play a regulatory of EAE, self-reactive activated T cells transmigrate into role in inflammation, we measured the levels of IL-10 the CNS, where they activate resident microglia and at- and TGF-b in the serum of animals prior to and post tract further influx of monocytes and other cells through EAE immunization but found no significant differences the secretion of proinflammatory cytokines (Raine, 1997; in their levels between the Sema7A+/+ and Sema7A2/2 Ruddle et al., 1990). To analyze the role of Sema7A in littermates (Figures S3B and S3C). Therefore, the high T cell responses in vivo, we immunized Sema7A2/2 mortality of Sema7A2/2 animals at the onset of EAE is and wt littermate mice with myelin oligodendrocyte likely a result of increased T cell activation, accelerated glycoprotein (MOG35–55) peptide in complete Freud’s infiltration of T cells into the CNS, and the elevated levels adjuvant (CFA). This immunization regimen leads to the of IFNg produced by activated T cells. This phenotype is induction of EAE within 12–15 days (see Experimental indicative of an exacerbated autoimmune response in Procedures). Comparison of mean disease scores (aver- Sema7A2/2 animals. aged from three independent experiments) showed The CD25 CD4 regulatory T cells play an important role a dramatically increased disease index in Sema7A2/2 in autoimmune disease protection (Paust and Cantor, mice (Figures 2A and 2E), which was due to the high 2005; Sakaguchi and Sakaguchi, 2005), and their defi- mortality (w60%) (Figure 2E) of these animals near the ciency could contribute to the severity of EAE observed in time of disease onset. Analysis of disease progression Sema7A2/2 animals. We therefore addressed the in vivo in surviving Sema7A2/2 mice showed a single, lingering functionality of regulatory T cells deficient in Sema7A by paralysis episode in the Sema7A-deficient animals, in using a T cell transfer model of colitis. In this model, intra- contrast to a later onset, earlier disease remittance, peritoneal injection of a small number of purified CD4 and typical consecutive episodes of relapse in the wt RBhigh CD25negative T cells into a lymphopenic host results littermates (Figures 2B and 2E). Serum levels of IFNg in colitis and is associated with a rapid weight loss in the Negative Control of T Cells by Semaphorin 7A 593

Figure 2. Increased Morbidity and Alteration in the Clinical Course of EAE in Sema7A2/2 Mice EAE was induced and scored in Sema7A+/+ (n = 14) and Sema7A2/2 (n = 14) mice with

MOG35–55. Pooled data from three independent experiments are shown. (A) Sema7A2/2 mice are prone to die at the onset of EAE. (B) Sema7A2/2 animals surviving past the disease onset exhibit a monophasic EAE with a delayed period of recovery, compared to relapsing-remitting EAE phenotype in wild- type animals. (C) Elevated levels of IFNg in the serum of Sema7A2/2 mice 11 days postimmunization (p < 0.0014, result of unpaired t test with one-tailed p value). (D) Increased inﬁltration of CD3 lymphocytes into CNS in moribund Sema7A2/2 animals 11 days postimmunization with MOG. (E) Summary of EAE course in Sema7A2/2 and Sema7A+/+ mice. Sema7A2/2 mice are prone to increased mortality (61%) at the onset of disease compared to Sema7A+/+ (0%) littermates. The onset of disease in Sema7A2/2 mice is accelerated (mean 12.1 days compared with 14.9 days in Sema7A+/+).

animal (Powrie et al., 1994). Rag22/2 host mice trans- mononuclear cells to the spinal cord at the same time ferred with Sema7A2/2 or wt CD4 RBhigh CD25negative points (Figure 3A, right panel). At day 26 postimmuniza- T cells developed comparable weight loss and clinical tion, wt and KO animals of equal clinical scores (EAE = signs of colitis (Figure S4A and data not shown). Cotrans- 0.5) were compared. Inflammation and demyelination fer of Sema7A2/2 or Sema7A+/+ CD25high CD4 T regula- were more pronounced in the lower spinal cord of tory cells equally suppressed the colitis-associated Sema7A2/2 mice compared to wt animals, with Waller- weight loss and clinical signs of the disease (Figure S4B ian degeneration, and remyelination (Figure 3B, lower and data not shown). This result suggests that right panel). However, while inflammation and demyelin- Sema7A2/2 regulatory T cells are functional and further ation were more widespread and appeared earlier in points to a specific defect in the naive CD4 T cells and Sema7A2/2 mice (Figure 3C), no additional neuronal to their abnormal activation as a cause of the aggravated changes unrelated to inflammation were observed. EAE in Sema7A2/2 animals. Therefore, Sema7A deficiency in the EAE model appears to affect only cells of the immune system and does not Exacerbated EAE Pathology in Sema7A-Deficient seem to play any detectable role in neuronal protection. Mice Sema7A is a neuronal factor highly expressed in the Enhanced Delayed Type Hypersensitivity Response brain, where it could potentially contribute to neuronal in Sema7A2/2 Mice protection during an autoimmune disease. In order to To further determine if the lack of Sema7A is responsible evaluate if Sema7A deficiency makes neurons more vul- for a generalized defect in activation of CD4 T cells nerable to damage during EAE, we examined histopath- in vivo, we analyzed the induction of a delayed type hy- ological changes in the CNS of wt and KO animals with persensitivity (DTH) response by sensitizing Sema7A2/2 similar clinical scores. Despite the similarity in clinical and wt animals with a chemical skin irritant, oxazolone. scores between the two groups of surviving animals Activated CD4 T cells play a major role in DTH by emi- (Figure 2B), there were differences at the level of histo- grating to the irritated skin areas and producing inflam- pathology of the lower lumbar region of the spinal matory cytokines, resulting in a local tissue swelling cord. In Sema7A+/+ mice in the early phase (12 days (Kobayashi et al., 2001). The extent of ear swelling, mea- postimmunization, prior to EAE onset), very little inflam- sured 48 hr postchallenge, was significantly greater in mation was detectable (Figure 3A, left panel). In contrast, the Sema7A2/2 than in wt animals upon secondary in Sema7A2/2 mice, there was a noticeable influx of challenge (Figure S5). This exaggerated DTH response Immunity 594

Sema7A2/2 T cell recipients had higher serum con- centrations of IFNg, IL-6, and KC-1, compared to Sema7A+/+ T cell-recipient mice (Figure S3D and data not shown). The immunofluorescent staining revealed higher numbers of Sema7A2/2 CD4 T cells in the cerebel- lar sections from transfer recipients, which were not seen in wt T cell recipients (Figure 4B, left panels). The bright CD11b staining, indicative of activated microglia, is a likely effect of IFNg production by infiltrating Sema7A2/2 CD4 T cells and was not observed in wt T cell recipients (Figure 4B, right panels). The activated Sema7A2/2 Tcells displayed a robust expansion, making up to 50% of total cells in the spleen, compared to a more modest proliferation (2%–3% of total spleen cells) in wt T cell recipients (Figures 4C and 4D). Sema7A2/2 CD4 T cells showed increased expression of activation markers, such as CD25, compared to Sema7A+/+ T cells (Figure 4E). Thus, the hyperactivity of Sema7A2/2 T cells is likely due to the lack of a T cell-autonomous inhibitory effect of Sema7A, as all other cells in the recipient mice were Sema7A sufficient.

Increased Homeostatic Proliferation of Sema7A-Deficient T Cells We next examined the role of Sema7A in T cells in homeostatic proliferation. Homeostatic proliferation in lymphopenic hosts is controlled primarily by cytokines and low-affinity interactions of T cells with self-antigens and has been linked with T cell-mediated autoimmunity (Gallegos and Bevan, 2004; Surh and Sprent, 2000). We transferred one million of naive Sema7A2/2 or littermate wt CD4 T cells into Rag22/2;Sema7A+/+ lymphopenic Figure 3. Histopathology of EAE in Sema7A+/+ and Sema7A2/2 Mice hosts to compare their homeostatic proliferation. Within 2/2 (A and B) Toluidine-blue-stained epoxy sections (1 mm) from the 6 days posttransfer, the Sema7A T cells showed lower lumbar spinal cord region. (A) At 12 days postinduction of greater expansion compared to wt T cells (Figures 5A EAE, Sema7A2/2 mice show inflammation and increased number and 5B). Time-course analysis revealed that Sema7A2/2 of lymphocytes in L5 region of spinal cord, which is not seen in wt CD4 T cells showed a greater degree and faster kinetics 2/2 littermates. (B) 26 days postimmunization, sections from Sema7A of proliferation compared to Sema7A+/+ T cells (Fig- mice (B, right) (S1 spinal cord) showed infiltration of lymphocytes, ure 5C). These data demonstrate that Sema7A ex- and demyelination (lower right). Demyelination was less pronounced in Sema7A+/+ mice (B, left) (S1 spinal cord), but an influx of inflamma- pressed in T cells negatively regulates lymphopenia- tory cells was observed. induced T cell proliferation. (C) Distribution and severity of histopathological signs. Sections of representative mice were evaluated blindly with scores between Sema7A Negatively Regulates TCR Downmodulation 0.5–4 for inflammatory cell infiltrates and demyelinated axons. Means and Signaling are based on three slides with two animals per group. TCR-mediated T cell activation is controlled by a number of parameters, including the duration of the TCR signal- is consistent with the hyperactive T cell phenotype seen ing. Upon TCR stimulation, cell surface expression of in EAE. TCR declines rapidly as components of the TCR signaling complex are targeted for degradation through the T Cell-Autonomous Role of Sema7A in Limiting endocytic pathway (Duan et al., 2004; Geisler, 2004). Autoimmunity Sustained surface expression of TCR can interfere with While the pathology of EAE is mediated in large part by termination of downstream signals and result in pro- T cells, the contribution of other cells types is essential longed triggering of the T cells, causing their hyperacti- (Kuchroo et al., 2002) and could potentially play a role vation (Naramura et al., 2002). To determine the effect of in the high mortality of Sema7A2/2 mice at the onset of Sema7A on TCR signaling, we activated T cells with EAE. To determine the contribution of Sema7A defi- plate-bound anti-CD3 antibody and analyzed changes ciency specifically in T cells, we purified naive CD4 in the TCR surface expression by flow cytometry. We T cells from Sema7A2/2 or wt littermates, labeled them found that anti-CD3 stimulation led to a downregulation with CFSE, and transferred one million cells per mouse of surface TCR expression in Sema7A+/+ CD4 T cells, into Sema7A+/+, T cell-deficient recipients (Rag22/2 or whereas this TCR downregulation was markedly de- TCRa2/2). Twenty-four hours later, mice were immu- layed in Sema7A2/2 T cells (Figure 6A). After 8 hr of nized with MOG35–55. Within 6–9 days, all the mice anti-CD3 stimulation, the wt cells lost about 90% of their receiving Sema7A2/2 T cell, but none of the Sema7A+/+ TCR surface expression. In contrast, the Sema7A2/2 T cell recipients, became moribund or died (Figure 4A). CD4 T cells had retained more than 40% of the TCR level Negative Control of T Cells by Semaphorin 7A 595

Figure 4. Increased Expansion of Sema7A2/2 CD4 T Cells in a Lymphopenic Host Rag22/2 mice received 1 3 106 FACS-sorted CFSE labeled CD4 cells from Sema7A2/2 or Sema7A+/+ littermates intravenously. 1 day posttransfer, mice were immunized with

MOG35–55. (A) Rag22/2 mice (n = 8) receiving Sema7A2/2 T cells became moribund and died rapidly within 6–9 days following immunization, while mice receiving Sema7A+/+ T cells (n = 8) re- mained healthy. (B) Immunofluorescent detection of Sema7A2/2 or Sema7A+/+ CD4 (red, left) and detection of CD11b (red, right) in brain sections of TCRa2/2 mice. Results are representative of three experiments. (C–E) Typical profiles of splenocytes from Rag22/2 recipients 6 days postimmunization. (C) Dilution of CFSE label in transferred CD4 T cells. (D) Expansion of CD T cells is greatly enhanced in Sema7A2/2 cell recipients (n = 4/ group 6 SD). (E) Activated CD4 T cells with high CD25 surface staining are more numer- ous in spleens of Sema7A2/2 CD4 than Sema7A+/+ CD4 recipients. of unstimulated cells (Figure 6B). Similar results were phorylation patterns between wt and Sema7A2/2 T cells observed in CD8 T cells (data not shown). revealed no global differences in major phosphorylated As sustained surface expression of TCR is known to bands, which include Zap70 and components of the correlate with increased calcium signaling, we com- TCR complex (data not shown). One of the downstream pared calcium mobilization upon anti-CD3 crosslinking targets of TCR activation is glycogen synthase kinase 3 in Sema7A+/+ and Sema7A2/2 CD4 T cells. We found (GSK-3), which negatively regulates antigen-specific T that the magnitude of Ca2+ mobilization in Sema7A-de- cell proliferation by acting on nuclear factor of activated ficient T cells was higher and decayed at a slower rate T cells (NF-AT) (Ohteki et al., 2000). In response to TCR than in the control wt cells (Figure 6C). signal, GSK-3 is inactivated by phosphorylation, which TCR activation leads to rapid phosphorylation of ZAP- releases the inhibitory effect on T cell activation, while 70 and activation of the ERK-MAP kinase pathway (Qian increased inhibition of GSK-3 leads to hyperproliferation and Weiss, 1997). The comparison of tyrosine phos- of T cells (Ohteki et al., 2000). We observed a modestly increased inhibitory phosphorylation of GSK-3a/b on Ser21/9 in response to TCR stimulation in Sema7A2/2 T cells, compared to wt T cells (Figures 6D and 6E). The S6 ribosomal kinase (S6K) is required for cell growth and G1 cell-cycle progression and is a target of TCR-induced PDK1 kinase activity (Lee et al., 2005; Pullen et al., 1998). We found that in Sema7A2/2, T cells the phosphorylation of S6K is modestly increased in strength (Figure 6E) and duration (Figure 6D), in comparison to the wt T cells. TCR-induced ERK activation was similar in Sema7A2/2 and wt T cells (Figures 6D and 6E). Activation of T cells through the T cell receptor (TCR) triggers b1-integrin signaling and induces phosphorylation of focal adhesion kinase (FAK) (Iwata et al., 2000). Considering the evidence of Sema7A in b1-integrin function in neurons (Pasterkamp et al., 2003), we analyzed the FAK activation in Sema7A-deficient cells. We activated T cells with a combination of anti-CD3/CD28 antibodies. Within 10 min, the wt T cells exhibited a robust phosphorylation of FAK at Y925, which was diminished in the Sema7A2/2 T cells (Figure 6F). We observed similar FAK Figure 5. Homeostatic Proliferation of CD4 T Cells phosphorylation defect in Sema7A2/2 macrophages ac- (A) Increased homeostatic proliferation among Sema7A2/2 CD4 tivated with collagen IV or with IGF-I (Figure S6A). TCR- +/+ cells compared to Sema7A cells in lymphopenic host. mediated phosphorylation of FAK at Y925 creates a bind- (B) Mean of CD4 T cells determined by surface antibody staining and ing site for the adaptor proteins Gads/Grb2 which could FACS analysis (n = 4/group 6 SD). (C) Time course of homeostatic proliferation of transferred naive negatively regulate T cell signaling by bringing inhibitory CD4 T cells labeled with CFSE and followed 2, 4, and 6 days after regulatory molecules into the proximity of the TCR com- transfer. Shown are CFSE-dilution profiles of gated CD4 CD3 cells. plex (Schlaepfer and Hunter, 1996). Immunity 596

Figure 6. Analysis of TCR Signaling in Wt and KO CD4 T Cells (A) Anti-CD3 stimulation induced TCR surface downregulation. T cells were stimulated with plate bound CD3 for indicated time, and surface TCR retention was determined by flow cytometry. Shown is TCRb surface staining of gated CD4-positive population (filled histograms, wt; open histograms, Sema7A2/2). (B) Quantification of anti-CD3 induced TCR downregulation. (C) Enhanced calcium mobilization in Sema7A2/2 CD4 T cells. Histograms repre- sent calcium flux measured by flow cytometry. Control: addition of secondary antibody alone. (D and E) Lysates of naive CD4+ Sema7A+/+ or Sema7A2/2 T cells (D) coated with anti-CD3 antibody on ice and incubated with crosslinking antibody for indicated times, or CD4+ Sema7A+/+ or Sema7A2/2 effector T cells (E) activated for 10 min with anti-CD3/anti-CD28 antibodies were analyzed for kinases phosphorylation by Western blotting. (F) Western blot of OTII CD4+ Sema7A+/+ or Sema7A2/2 effector T cells activated for 10 min with anti-CD3/CD28 antibodies shows diminished FAK phosphorylation. Results representative of three independent experiments.

Thus, with the exception of TCR downmodulation, proliferation of Sema7A2/2 T cells compared to wt cells Ca2+ mobilization, and FAK phosphorylation, Sema7A2/2 (Figure 7A), similarly to the antigen-specific stimuli T cell did not exhibit any dramatic enhancements in (Figure 1). This result implies an interaction of Sema7A signaling events when stimulated with plate bound with a receptor on DCs. Fixed DCs pulsed with SEA sim- anti-CD3/CD28 antibodies in the absence of APCs. Fur- ilarly resulted in an increased proliferation of Sema7A2/2 thermore, stimulation with anti-CD3, in contrast to the T cells, indicating that Sema7A2/2 T cells are not simply APC-induced hyperproliferation of Sema7A2/2 T cells more responsive to other activating signals, such as (Figure 1), led to a diminished Sema7A2/2 T cells prolifer- cytokines, produced by DCs. ation compared to wt T cells (Figure S6B). Activation of Sema7A may affect the TCR complex through a pro- T cells with APC-independent stimuli such as anti-CD3 tein-protein interaction on the cell surface, and that in- antibody crosslinking delivers a strong and nonspecific teraction may be stabilized by the presence of a Sema7A signal to the T cell and bypasses a number of signals receptor on the APC. Consistent with this prediction, we that are otherwise delivered by an APC. Therefore, it is noticed that while staining of the TCRb on the cell sur- likely that while anti-CD3 activation allows visualization face is identical in Sema7A2/2 and Sema7A+/+ cells T cell signaling in the absence of contaminating APCs, (Figure 6A, top panel), staining of CD3 is decreased on the actual differences in signaling between wt and Sema7A2/2 CD4 T cells (Figure 7B). As the levels of Sema7A2/2 T cells could only be revealed when T cells TCR and CD3 are equal on the plasma membrane (Call are activated by professional APCs. and Wucherpfennig, 2005), this result suggests that CD3 in Sema7A2/2 T cells is not fully accessible to the Sema7A2/2 Functions in T Cell Activation by APCs antibody, either because of a conformational change The hyperproliferation of Sema7A2/2 T cells stimulated or obstruction of the surface part of CD3 by another with DCs, but not with anti-CD3/CD28 antibodies in molecule. Thus, Sema7A appears to directly affect the the absence of APCs, suggested that interaction of components of the TCR complex. Sema7A with a receptor present on APC might be nec- These results lead us to propose a tentative mecha- essary to execute the inhibitory function of Sema7A in nism of Sema7A function in CD4 T cells. As shown in the T cell. To test this hypothesis, we activated Figure 7C, left panel, Sema7A normally present on the Sema7A+/+ and Sema7A2/2 T cells with DCs pulsed surface of a T cell interacts with the components of with superantigen, SEA (staphylococcal enterotoxin A). the TCR/CD3 complex during T cell activation by an This antigen-nonspecific activation resulted in hyper- APC. A putative receptor present on APC (presumably Negative Control of T Cells by Semaphorin 7A 597

animals. We did not find any obvious neuron-specific role for Sema7A in neuronal survival and protection in the EAE model. We also provide evidence of an enhanced T cell-mediated delayed-type-hypersensitivity reaction in Sema7A-deficient mice, and this observation is in agreement with the general hyperactive phenotype of Sema7A2/2 T cells. The hyperactivity of Sema7A2/2 T cells is not dependent on antigen stimulation because the enhanced T cell proliferation is evident even in homeostatically proliferating T cells from Sema7A2/2 mice. Collectively, these results demonstrate that Sema7A negatively regulates T cell activation and functions in a T cell-specific manner. Sema7A deficiency results in an enhanced T cell proliferation triggered by either antigen or lymphopenia. However, Sema7A-deficient mice do not develop any obvious autoimmune disorders, and T cell hyperactivity is most strongly pronounced under conditions of experimentally induced autoimmunity or lymphopenia, pointing to a specific, rather than general, inhibitory role of Sema7A. Based on the T cell transfer experiments, we conclude that the mechanism of Sema7A activity is largely T cell intrinsic. Engagement of T cell receptors by an antigen leads to activation, cell proliferation, differentiation, and effector functions. It has been proposed that the ini- Figure 7. Sema7A Regulates T Cell Activation by APCs tiation and propagation of the signaling events taking 2/2 place in immune cells occur in specialized membrane (A) Sema7A T cells activated with live of fixed DCs pulsed with staphylococcal enterotoxin A (SEA) show increased proliferation regions, lipid rafts, where the lymphocyte receptors lo- compared to wt T cells. T cell proliferative responses are expressed calize (Magee et al., 2002). The GPI (glycosylphosphati- as the mean 6 SE of triplicate cultures. dylinositol)-linked proteins localize to lipid rafts, and (B) Surface staining of CD33 is decreased in Sema7A2/2 CD4 T cells many of them were shown to play important roles in +/+ compared to Sema7A cells. T cell activation (Hwa, 2001; Loertscher and Lavery, (C) Model of Sema7A function in T cell signaling. Sema7A interacts 2002; Magee et al., 2002). We propose that similarly with the components of TCR complex and with a putative receptor on APCs. This interaction stabilizes TCR/CD3 complex and pro- to several other T cell-specific GPI-linked proteins, motes inhibitory signals that limit T cell proliferation. In the absence Sema7A may interact in cis and negatively regulate the of Sema7A, the TCR/CD3 complex is destabilized, and upon activa- TCR complex within lipid rafts. Furthermore, our data tion by an APC, negative regulatory signals are impaired, leading to indicates that Sema7A may somehow control the func- overt activation and T cell hyperproliferation. tional conformation of the TCR complex. While many different biochemical pathways regulate b1-integrin) engages Sema7A, and that interaction neg- T cell activation, TCR downmodulation and internaliza- atively regulates the TCR. In the absence of Sema7A, tion are important means of negative regulation. Previous this negative regulatory mechanism is lost, resulting work utilizing E3-ligase Cbl knockout animals showed in enhanced activation and hyperproliferation of CD4 a critical role of TCR downregulation in T cell responsive- T cells (Figure 7C, right panel). ness, as in the T cells from these mice the abnormal retention of T cell receptor on the cell surface causes sustained Discussion TCR signaling (Naramura et al., 2002). In addition, Cbl- b knockout T cells show a hyperproliferative phenotype In this report, we describe a crucial function of a neuronal and develop autoimmunity, indicating that alterations in guidance factor, Semaphorin 7A, in the negative regula- TCR downmodulation and negative regulation can lead tion of T lymphocyte function. We find that Sema7A- to detrimental effects in the host (Bachmaier et al., deficient T cells are hyperactive in response to activa- 2000; Chiang et al., 2000; Krawczyk et al., 2000). We tion in vivo and in vitro. This phenotype is particularly find that TCR downregulation upon stimulation is defec- accentuated in wild-type lymphopenic hosts receiving tive in Sema7A2/2 T cells. The surface retention of TCR small numbers of Sema7A-deficient T cells followed by correlates with an enhanced and sustained calcium immunization with self-antigen, which leads to a mas- mobilization. sive, unrestrained T cell expansion and a subsequent TCR-mediated cell signaling events are tightly regu- death of the host animal. This hyperactive phenotype lated, as even small alterations in the TCR sensitivity correlates well with our finding that Sema7A knockout can lead to changes in T cell activation and proliferation mice undergo a markedly different and more severe (Qian and Weiss, 1997). We found that CD4 T cells lack- course of EAE compared to their Sema7A-sufficient lit- ing Sema7A show slightly enhanced activation of S6K termates. The Sema7A-deficient animals are prone to and calcium influx, as well as increased inhibitory phos- die near the onset of clinical EAE, and that correlates phorylation of GSK-3. Increased inhibition of GSK-3 has with the increased infiltration of CNS by lymphocytes, been linked with T cell hyperproliferation (Ohteki et al., which we observed in the moribund Sema7A knockout 2000). We also demonstrate a defect in activation of Immunity 598

FAK, which is reportedly involved in recruitment of neg- function in the immune system, in addition to its function ative regulators to the proximity of TCR (Schlaepfer in the nervous system, and thus adds to a growing num- and Hunter, 1996), although the functional significance ber of signaling pathways shared by the two systems of defective FAK activation is currently unclear. Interest- (Trautmann and Vivier, 2001). ingly, hyperresponsiveness of Sema7A-deficient T cells could only be observed upon T cell stimulation by pro- Experimental Procedures fessional APCs, suggesting a contribution of DC-derived signal to Sema7A-mediated negative regulation of T cell Mice 2/2 activation. Generation of Sema7A mice was described previously (Paster- kamp et al., 2003). Male and female mice 4–6 weeks old were used With the use of superantigen SEA, we find that an for analyses. C57BL/6 TCRa2/2 and C57BL/6 RAG22/2 mice were antigen-independent but APC-dependent activation of purchased from the Jackson laboratory (Bar Harbor, ME) and Sema7A-deficient T cells leads to their hyperprolifera- Taconic Farms (Germantown, NY). Animals were housed in a patho- tion and that increased response is preserved if fixed gen-free facility at Yale University. All procedures involving mice DCs are used as APCs. This suggests that part of the were approved by Yale University Institutional and Animal Care normal activation of T cells by APCs is the inhibitory sig- and Use Committee. naling that limits T cell proliferation. It is likely that Reagents and Antibodies Sema7A interacts with a putative receptor expressed LPS, OVA, oxazolone, and staphlococcal enterotoxin A were from on DCs and that this interaction may be required to pro- Sigma (St. Louis, MO). MOG35–55 and OTII OVA peptide were synthe- mote inhibitory function of Sema7A in TCR signaling. sized and purified by Keck Facility (Yale University, CT). APC anti- Our data provide evidence for Sema7A function in TCR CD4, phycoerythrin (PE) anti-CD25, and rat anti-CD4 (L3T4) were all signaling that could involve either a direct association of from BD Biosciences (San Diego, CA). Anti-rat IgG-PE was from Sema7A/TCR or possibly another molecule serving as Jackson ImmunoResearch Labs (West Grove, PA). Goat anti-rat Alexa Fluor 594 was from Molecular Probes (Eugene, OR). Anti- the linker. The diminished antibody accessibility of CD4, anti-CD11c, and anti-CD19 microbeads were from Miltenyi Bio- 2/2 surface CD3 in Sema7A T cells indicates that the ab- tec (Auburn, CA). Mouse cells were cultured in complete RPMI-1640 sence of Sema7A has a direct effect on the TCR complex supplemented with 10% FCS, 100 U/ml penicillin, 100 mg/ml strepto- composition or conformation, and that may contribute to mycin, 2 mM L-glutamine, 10 mM HEPES, and 1 mM sodium pyruvate the defects of T cell signaling seen in Sema7A-deficient (all from Sigma). Recombinant Sema7A-FC was prepared as follows: T cells. The nature of the Sema7A counter structure re- Sema7A with a deletion of C-terminal 25 amino acids was cloned into a pcDNA3 vector upstreem and in frame with a human Fc fragment, mains puzzling. Recent report of the Plexin C1-deficient expressed in 293T cells, and purified from supernatant with protein animals demonstrated only a small defect in T cell func- A column (Amersham) according to manufacturer’s instructions. tion in these mice and completely normal function of The Fc fragment was expressed in pSecTag vector and purified APCs (Walzer et al., 2005). While it remains possible by the same method. Recombinant proteins were dialyzed exten- that Plexin C1 plays some role in Sema7A signaling in sively against PBS and purity was assayed by UV absorption mea- T cells, it is likely that another receptor, possibly an integ- surements. rin, plays a more relevant role. Our observations indicate EAE Induction and Scoring that such receptor may function in trans with Sema7A For EAE induction, MOG35-55 (3 mg/ml in PBS) was mixed with equal and may be expressed on the DCs. volume of complete Freund’s adjuvant (CFA, from Sigma), and 50 ml Based on the available evidence, we propose a model was injected subcutaneously in each flank of Sema7A2/2 or that predicts that Sema7A exists in a complex with the Sema7A+/+ littermates. Pertussis toxin (Ptx, 0.2 mg) (Biological Lab- TCR and/or CD3 on the T cell surface. During T cell acti- oratories, Inc., Campbell, CA) was given intravenously (i.v.) at the vation, Sema7A engages a receptor on the APC. This en- time of immunization and again 2 days later. Mice were scored daily for clinical signs of disease, and a numerical score was assigned gagement may serve to change the conformation of based on the severity of the disease symptoms: 0, no disease; 1, Sema7A to make it active, or alternatively it may serve limp tail; 2, weak tail and partial hind limb paralysis; 3, total hind to position Sema7A within the T cell: APC synapse. Ac- limb paralysis; 4, both hind limb and fore limb paralysis; 5, death. tivity of Sema7A manifests in limiting T cell proliferation Mice with a score of 4 were euthanized. Remission was defined by by promoting induction of the inhibitory signaling mech- a decrease in the score of at least one point for two consecutive anisms. In the absence of Sema7A, TCR/CD3 signaling days. EAE was considered remitting when at least one remission occurred within the first 26 days. and complex assembly is impaired. This impairment may in part stem from the altered CD3 activity, defective Adoptive Transfer TCR internalization, and altered intracellular signaling. In Pooled spleen and lymph node cells from Sema7A2/2 or wild-type the absence of Sema7A, this negative regulatory mech- littermates were depleted with TIB 146, TIB 164, Y3JP, and TIB anism is lost, resulting in an abnormally strong activa- 128 hybridoma supernatants and rabbit complement (Cedarlane tion and hyperproliferation of the T cells. Conversely, Lab). After depletion, cells were washed in PBS/1% fetal calf serum providing additional soluble Sema7A inhibits T cell sig- twice, stained with anti-CD4-APC, and sorted on FACSVantage. 1 3 106 CD4 cells were labeled with 5 mM Carboxy-fluorescein diacetate, naling and decreases proliferation. succinimidyl ester (CFSE) for 7 min, washed in sterile PBS three Previous studies have demonstrated immunological times, resuspended in a final volume of 200 ml PBS, and adminis- function of several members of Semaphorin family, in- tered intravenously to naive C57BL/6 TCRa2/2 or C57BL/6 RAG22/2 cluding Sema4A, Sema4D, and PlexinA1 (Bismuth and recipient mice, and EAE was induced 24 hr after transfer. For homeo- Boumsell, 2002; Kumanogoh and Kikutani, 2003; Wong static proliferation studies, identical protocol was used, except the et al., 2003). In each of these cases, the mechanism of immunization step was omitted. Semaphorin or Semaphorin receptor function in the Staining and Flow Cytometry immune system is different. The present study, together Cells were stained with relevant antibodies for 30 min on ice, with previous reports (Holmes et al., 2002; Mine et al., washed, and analyzed with a FACSCalibur (BD Biosciences). Data 2000), demonstrates that Sema7A also has an important were analyzed with FlowJo software. Negative Control of T Cells by Semaphorin 7A 599

Immunization secondary antibody for indicated times at 37ºC. Purified CD4 OTII For antigen-specific T cell responses, mice were immunized in hind transgenic T cells (wt or Sema7A2/2) were cultured with irradiated foot pads with 100 mg/mouse of OVA (Sigma) and 50 mg/mouse of wild-type spleen cells and OVA peptide (1 mg/ml) for 4 days. Cells LPS emulsified in Incomplete Freund’s adjuvant (IFA, from Sigma). were then purified, washed three times in PBS, and resuspended Draining lymph nodes were collected 6 days postimmunization. in RPMI medium without serum for 6 hr and activated with anti- CD3/anti-CD28 (at 10 mg/ml) for 10 min at 37ºC. 2 3 106 cells/sample Cell Purification were lysed immediately following activation; total protein (100 mg) Spleen and lymph nodes were harvested from 6–12 week old mice. was resolved on 10% SDS-PAGE gels. Blotting was performed Cell suspensions were incubated with anti-CD4 microbeads followed with phospho-S6 kinase T389, phospho-ERK T202/Y204, or phos- by purification with the AutoMACS sorter (Miltenyi Biotec). In all pho-GSK3a/b S21/9 (all from Cell Signaling Technology), or mono- experiments, more than 95% of purified cells were positive for CD4. clonal anti-b-actin (Sigma). Bands were visualized with the ECL System (all from Amersham). T Cell Proliferation Assay Purified CD4 T cells (1 3 105) from draining lymph nodes were Statistical Analysis cultured in flat bottom 96-well plates with spleen cells (3 3 105, Analysis of statistical significance for indicated data sets was 1500 rads irradiated) or bone marrow DCs (2 3 104) live or fixed, performed with Prism4 Graphpad software. as indicated, and titrating doses of antigen for 72–84 hr. Proliferation 3 of T cells was determined by incorporation of [ H]thymidine for the Supplemental Data last 12–16 hr of culture. The Supplemental Data include six figures and are available at http://www.immunity.com/cgi/content/full/24/5/591/DC1/. ELISA Serum samples were collected by eye bleeding from animals. Paired Acknowledgments antibodies for cytokine specific enzyme linked immunosorbent assays were from Pharmingen. Authors wish to thank Dr. A. Kolodkin for the generous gift of Sema7A-deficient mice; Miriam Parkingan for technical assistance Immunohistochemistry and Immunofluorescence with histopathology; members of the Medzhitov lab for comments 2/2 Wild-type and Sema7A animals were anesthetized, perfused with on this manuscript; C. Annicelli and S. Holley for assistance with ice-cold PBS, and fixed with 4% paraformaldehyde in PBS. Brains animals; Drs. O. Henegariu, J. Czyzyk, and C. Viret for helpful discus- and spinal cords were removed and frozen or paraffin-embedded sions; and Drs. A. Jamieson and M. Szczepanik for help with calcium 4–5 mm sections were prepared. Tissues were stained with anti- measurements. This work was supported by the Howard Hughes CD3 antibody, followed by histochemical development with DAB. Medical Institute (R.M.), National Institutes of Health grants AI46688 For immunofluorescence, tissues were stained with anti-CD4 or anti- and AI055502 (R.M.), and United States Public Health Service grants CD11b (BD Biosciences) and secondary anti-rat Alexa Fluor 594. NS 08952, NS 11920, and NMSS grant RG 1001-J-10 (C.S.R). Results representative of at least three independent experiments.

Received: May 5, 2005 Histopathology Revised: February 21, 2006 Slices (1 mm) of glutaraldehyde-perfused spinal cord (L5, L6, and Accepted: March 6, 2006 S1) were postfixed in 1% osmium tetroxide/Millonig’s buffer on ice Published: May 23, 2006 for 90 min, dehydrated in ethyl alcohol (70%, 90%, 95%, and 100%), cleared in propylene oxide, and embedded in Epon. 1 mm epoxy sections were stained with toluidine blue for light microscopy. References Sections were scored by an investigator blinded to the code, on a scale of 0 to 4 for cell infiltration, de- and remyelination, and Wal- Angelisova, P., Drbal, K., Cerny, J., Hilgert, I., and Horejsi, V. (1999). lerian degeneration. Representative lesions were examined from Characterization of the human leukocyte GPI-anchored glycopro- matching levels of spinal cord in animals from both groups tein CDw108 and its relation to other similar molecules. Immunobiol- (Sema7A+/+ and Sema7A2/2) with comparable disease courses ogy 200, 234–245. and clinical scores. Bachmaier, K., Krawczyk, C., Kozieradzki, I., Kong, Y.Y., Sasaki, T., Oliveira-dos-Santos, A., Mariathasan, S., Bouchard, D., Wakeham, Calcium Mobilization A., Itie, A., et al. (2000). Negative regulation of lymphocyte activation Purified CD4 T cells were resuspended in RPMI/10% FCS and and autoimmunity by the molecular adaptor Cbl-b. Nature 403, loaded with 3 mM Indo-1 (AnaSpec, San Jose, CA) for 30 min at 211–216. 37ºC. Cells were then washed with ice-cold PBS and surface stained Bismuth, G., and Boumsell, L. (2002). Controlling the immune system with anti-CD3 for 30 min at 4ºC, followed by ice-cold PBS wash. through semaphorins. Sci. STKE 2002, RE4. Immediately prior to cross linking with secondary antibody, each Call, M.E., and Wucherpfennig, K.W. (2005). The T cell receptor: crit- sample was warmed up to 37ºC for 10 min, and calcium flux was ical role of the membrane environment in receptor assembly and measured with a LSRII flow cytometer. Data were analyzed with function. Annu. Rev. Immunol. 23, 101–125. FlowJo software. Chiang, Y.J., Kole, H.K., Brown, K., Naramura, M., Fukuhara, S., Hu, TCR Downregulation Measurement and Analysis R.J., Jang, I.K., Gutkind, J.S., Shevach, E., and Gu, H. (2000). Cbl-b Single-cell suspensions of splenic cells were incubated for indicated regulates the CD28 dependence of T-cell activation. Nature 403, times with 2 mg/ml plate bound anti-CD33 at 37ºC. Collected cells 216–220. were surface stained with FITC anti-TCRb (H57-597), PE anti-CD4 Comeau, M.R., Johnson, R., DuBose, R.F., Petersen, M., Gearing, P., or anti-CD8, fixed, and analyzed on FACS Calibur flow cytometer. VandenBos, T., Park, L., Farrah, T., Buller, R.M., Cohen, J.I., et al. The percentage of TCR downregulation for each gated CD4 or (1998). A poxvirus-encoded semaphorin induces cytokine pro- CD8 population was determined based on MFI of TCRb staining duction from monocytes and binds to a novel cellular semaphorin on stimulated versus unstimulated cells. To calculate the percent receptor, VESPR. Immunity 8, 473–482. of TCR downregulation from the cell surface, the following formula Delaire, S., Elhabazi, A., Bensussan, A., and Boumsell, L. (1998). was used: (U, unstimulated; S, stimulated) CD100 is a leukocyte semaphorin. Cell. Mol. Life Sci. 54, 1265–1276. Duan, L., Reddi, A.L., Ghosh, A., Dimri, M., and Band, H. (2004). The %TCR downregulation = ð½MFI Utime 0 2 MFI Stime t=MFI Utime 0 3 100 Cbl family and other ubiquitin ligases: destructive forces in control of antigen receptor signaling. Immunity 21, 7–17. Western Blotting Elhabazi, A., Marie-Cardine, A., Chabbert-de Ponnat, I., Bensussan, Naive purified CD4 T cells were incubated on ice with anti-CD3 A., and Boumsell, L. (2003). Structure and function of the immune (10 mg/ml) for 30 min, washed with ice-cold PBS, and treated with semaphorin CD100/SEMA4D. Crit. Rev. Immunol. 23, 65–81. Immunity 600

Gallegos, A.M., and Bevan, M.J. (2004). Driven to autoimmunity: the Magee, T., Pirinen, N., Adler, J., Pagakis, S.N., and Parmryd, I. nod mouse. Cell 117, 149–151. (2002). Lipid rafts: cell surface platforms for T cell signaling. Biol. Geisler, C. (2004). TCR trafficking in resting and stimulated T cells. Res. 35, 127–131. Crit. Rev. Immunol. 24, 67–86. Mine, T., Harada, K., Matsumoto, T., Yamana, H., Shirouzu, K., Itoh, Holmes, S., Downs, A.M., Fosberry, A., Hayes, P.D., Michalovich, D., K., and Yamada, A. (2000). CDw108 expression during T-cell devel- Murdoch, P., Moores, K., Fox, J., Deen, K., Pettman, G., et al. (2002). opment. Tissue Antigens 55, 429–436. Sema7A is a potent monocyte stimulator. Scand. J. Immunol. 56, Naramura, M., Jang, I.K., Kole, H., Huang, F., Haines, D., and Gu, H. 270–275. (2002). c-Cbl and Cbl-b regulate T cell responsiveness by promoting ligand-induced TCR down-modulation. Nat. Immunol. 3, 1192–1199. Hwa, K.Y. (2001). Glycosyl phosphatidylinositol-linked glycoconju- gates: structure, biosynthesis and function. Adv. Exp. Med. Biol. Ohteki, T., Parsons, M., Zakarian, A., Jones, R.G., Nguyen, L.T., 491, 207–214. Woodgett, J.R., and Ohashi, P.S. (2000). Negative regulation of T cell proliferation and interleukin 2 production by the serine threo- Iwata, S., Ohashi, Y., Kamiguchi, K., and Morimoto, C. (2000). Beta nine kinase GSK-3. J. Exp. Med. 192, 99–104. 1-integrin-mediated cell signaling in T lymphocytes. J. Dermatol. Sci. 23, 75–86. Pasterkamp, R.J., Peschon, J.J., Spriggs, M.K., and Kolodkin, A.L. (2003). Semaphorin 7A promotes axon outgrowth through integrins Kobayashi, K., Kaneda, K., and Kasama, T. (2001). Immunopatho- and MAPKs. Nature 424, 398–405. genesis of delayed-type hypersensitivity. Microsc. Res. Tech. 53, 241–245. Paust, S., and Cantor, H. (2005). Regulatory T cells and autoimmune disease. Immunol. Rev. 204, 195–207. Kolodkin, A.L., Matthes, D.J., O’Connor, T.P., Patel, N.H., Admon, A., Powrie, F., Correa-Oliveira, R., Mauze, S., and Coffman, R.L. (1994). Bentley, D., and Goodman, C.S. (1992). Fasciclin IV: sequence, Regulatory interactions between CD45RBhigh and CD45RBlow expression, and function during growth cone guidance in the grass- CD4+ T cells are important for the balance between protective and hopper embryo. Neuron 9, 831–845. pathogenic cell-mediated immunity. J. Exp. Med. 179, 589–600. Kolodkin, A.L., Matthes, D.J., and Goodman, C.S. (1993). The sema- Pullen, N., Dennis, P.B., Andjelkovic, M., Dufner, A., Kozma, S.C., phorin genes encode a family of transmembrane and secreted Hemmings, B.A., and Thomas, G. (1998). Phosphorylation and acti- growth cone guidance molecules. Cell 75, 1389–1399. vation of p70s6k by PDK1. Science 279, 707–710. Krawczyk, C., Bachmaier, K., Sasaki, T., Jones, G.R., Snapper, B.S., Qian, D., and Weiss, A. (1997). T cell antigen receptor signal trans- Bouchard, D., Kozieradzki, I., Ohashi, S.P., Alt, W.F., and Penninger, duction. Curr. Opin. Cell Biol. 9, 205–212. M.J. (2000). Cbl-b is a negative regulator of receptor clustering and raft aggregation in T cells. Immunity 13, 463–473. Raine, C.S. (1997). Demyelinating diseases. In Textoook of Neuro- pathology, R.L. Davis and D.M. Robertson, eds. (Baltimore: Williams Kuchroo, V.K., Anderson, A.C., Waldner, H., Munder, M., Bettelli, E., & Wilkins), pp. 243–287. and Nicholson, L.B. (2002). T cell response in experimental auto- Ruddle, N.H., Bergman, C.M., McGrath, K.M., Lingenheld, E.G., immune encephalomyelitis (EAE): role of self and cross-reactive Grunnet, M.L., Padula, S.J., and Clark, R.B. (1990). An antibody to antigens in shaping, tuning, and regulating the autopathogenic lymphotoxin and tumor necrosis factor prevents transfer of experi- T cell repertoire. Annu. Rev. Immunol. 20, 101–123. mental allergic encephalomyelitis. J. Exp. Med. 172, 1193–1200. Kumanogoh, A., and Kikutani, H. (2001). The CD100-CD72 interac- Sakaguchi, S., and Sakaguchi, N. (2005). Regulatory T cells in immu- tion: a novel mechanism of immune regulation. Trends Immunol. nologic self-tolerance and autoimmune disease. Int. Rev. Immunol. 22, 670–676. 24, 211–226. Kumanogoh, A., and Kikutani, H. (2003). Roles of the semaphorin Sato, Y., and Takahashi, H. (1998). Molecular cloning and expression family in immune regulation. Adv. Immunol. 81, 173–198. of murine homologue of semaphorin K1 gene. Biochim. Biophys. Kumanogoh, A., Watanabe, C., Lee, I., Wang, X., Shi, W., Araki, H., Acta 1443, 419–422. Hirata, H., Iwahori, K., Uchida, J., Yasui, T., et al. (2000). Identification Schlaepfer, D.D., and Hunter, T. (1996). Evidence for in vivo phos- of CD72 as a lymphocyte receptor for the class IV semaphorin phorylation of the Grb2 SH2-domain binding site on focal adhesion CD100: a novel mechanism for regulating B cell signaling. Immunity kinase by Src-family protein-tyrosine kinases. Mol. Cell. Biol. 16, 13, 621–631. 5623–5633. Kumanogoh, A., Marukawa, S., Suzuki, K., Takegahara, N., Wata- Surh, C.D., and Sprent, J. (2000). Homeostatic T cell proliferation: nabe, C., Ch’ng, E., Ishida, I., Fujimura, H., Sakoda, S., Yoshida, how far can T cells be activated to self-ligands? J. Exp. Med. 192, K., and Kikutani, H. (2002a). Class IV semaphorin Sema4A enhances F9–F14. T-cell activation and interacts with Tim-2. Nature 419, 629–633. Tamagnone, L., Artigiani, S., Chen, H., He, Z., Ming, G.I., Song, H., Kumanogoh, A., Suzuki, K., Ch’ng, E., Watanabe, C., Marukawa, S., Chedotal, A., Winberg, M.L., Goodman, C.S., Poo, M., et al. (1999). Takegahara, N., Ishida, I., Sato, T., Habu, S., Yoshida, K., et al. Plexins are a large family of receptors for transmembrane, secreted, (2002b). Requirement for the lymphocyte semaphorin, CD100, in and GPI-anchored semaphorins in vertebrates. Cell 99, 71–80. the induction of antigen-specific T cells and the maturation of den- Trautmann, A., and Vivier, E. (2001). Immunology. Agrin—a bridge dritic cells. J. Immunol. 169, 1175–1181. between the nervous and immune systems. Science 292, 1667– Kumanogoh, A., Shikina, T., Suzuki, K., Uematsu, S., Yukawa, K., Ka- 1668. shiwamura, S., Tsutsui, H., Yamamoto, M., Takamatsu, H., Ko-Mita- Walzer, T., Galibert, L., and De Smedt, T. (2005). Dendritic cell func- mura, E.P., et al. (2005). Nonredundant roles of Sema4A in the immune tion in mice lacking Plexin C1. Int. Immunol. 17, 943–950. system: defective T cell priming and Th1/Th2 regulation in Sema4A- Wong, A.W., Brickey, W.J., Taxman, D.J., van Deventer, H.W., Reed, deficient mice. Immunity 22, 305–316. W., Gao, J.X., Zheng, P., Liu, Y., Li, P., Blum, J.S., et al. (2003). CIITA- Lange, C., Liehr, T., Goen, M., Gebhart, E., Fleckenstein, B., and regulated plexin-A1 affects T-cell-dendritic cell interactions. Nat. Ensser, A. (1998). New eukaryotic semaphorins with close homology Immunol. 4, 891–898. to semaphorins of DNA viruses. Genomics 51, 340–350. Wong, F.S., Dittel, B.N., and Janeway, C.A., Jr. (1999). Transgenes Lee, K.Y., D’Acquisto, F., Hayden, M.S., Shim, J.H., and Ghosh, S. and knockout mutations in animal models of type 1 diabetes and (2005). PDK1 nucleates T cell receptor-induced signaling complex multiple sclerosis. Immunol. Rev. 169, 93–104. for NF-kappaB activation. Science 308, 114–118. Xu, X., Ng, S., Wu, Z.L., Nguyen, D., Homburger, S., Seidel-Dugan, Loertscher, R., and Lavery, P. (2002). The role of glycosyl phospha- C., Ebens, A., and Luo, Y. (1998). Human semaphorin K1 is glycosyl- tidyl inositol (GPI)-anchored cell surface proteins in T-cell activation. phosphatidylinositol-linked and defines a new subfamily of viral-re- Transpl. Immunol. 9, 93–96. lated semaphorins. J. Biol. Chem. 273, 22428–22434. Luo, Y., Raible, D., and Raper, J.A. (1993). Collapsin: a protein in brain that induces the collapse and paralysis of neuronal growth cones. Cell 75, 217–227.