Carboxy-fluorescein diacetate-succinimidyl ester
Notch receptor intracellular domain
N-[N-(3,5-Difluorophenacetyl-L-alanyl)]-S-phenylglycine t-butyl ester
The Notch pathway is involved in cell differentiation processes in various organs and at several developmental stages. The importance of Notch for early T lymphocyte development is well established. Recently, Notch has been implicated in directing naive T helper cell differentiation towards the Th1, Th2 or regulatory T cell lineages. However, the molecular events underlying these processes are poorly understood.We show that the Notch ligands Delta-like1, Delta-like4 and Jagged1 differentially affect early T cell activation and proliferation following T cell receptor cross-linking. Delta-like1 and Jagged1 induce a dose-dependent inhibition of early activation markers CD69 and CD25, as well as inhibition of proliferation after anti-CD3 stimulation of purified CD4+ T cells. Similarly, the rapid activation of transcription factors NF-AT, AP-1 and NF-κB is suppressed. In contrast, triggering of Notch by Delta-like4 enhances T cell activation and proliferation. The observed effects are dependent on simultaneous cross-linking of TCR and Notch but independent of γ-secretase-mediated cleavage of Notch. These data suggest direct interference between Notch and early TCR signal transduction events, independent of the classical Notch pathway via release of the Notch intracellular domain. A Notch-mediated alteration of TCR signaling strength may contribute to the recently described modulation of naïve T cell differentiation by Notch ligands.
The Notch pathway is highly conserved in evolution and is generally involved in cell fate decisions during differentiation 1. In mammals four isoforms of the transmembrane receptor Notch have been identified, and they are activated by members of two conserved groups of ligands, Jagged1 and Jagged2 or Delta-like (Dll)1, Dll3 and Dll4 2. Ligand binding induces the γ-secretase-mediated cleavage of the Notch receptor intracellular domain (NICD) 3, 4 and its translocation into the nucleus, where it associates with CBF-1/RBP-Jκ and converts it from a repressor into an activator of transcription 5, 6.
Notch signaling is known to regulate a multitude of differentiation processes in early embryonic development but is also involved in hematopoiesis 2, 7, 8. Its critical role in the regulation of multiple steps during early lymphocyte ontogeny in the thymus is well studied (for review see 8–11). Mature CD4+ T cells enter the periphery and further differentiate upon activation by their cognate antigen and costimulatory signals into various T helper cell subsets (Th1, Th2, Th3, and TR1). Recently, Notch has been described to be involved in these late antigen-dependent differentiation steps. Initial studies had reported the development of regulatory T cells in the mouse and human system by Jagged1-overexpressing APC 12–14 and the induction of IFN-γ-secreting Th1 cells as the result of a Dll1/Notch3 interaction 15. Recently, Amsen et al. 16 showed that APC generally use the Notch pathway to instruct Th cell differentiation. Interestingly, distinct effects were observed for the two families of Notch ligands, Dll inducing Th1 and Jagged inducing Th2 differentiation.
Apart from influencing the differentiation into effector cells, Notch has also been reported to have an impact on T cell activation upon antigen contact. However, currently available data are inconsistent. While Eagar et al. 17 reported inhibition of T cell activation, proliferation and cytokine production by Jagged1 or a stimulatory anti-Notch1 antibody, two other groups observed an inhibition of proliferation and CD25 expression upon specifically blocking Notch signaling, whereas the expression of a constitutively active Notch1 in their hands resulted in enhanced T cell activation 18, 19.
These conflicting results may arise from distinct effects mediated by individual Notch ligands. Their analysis during APC – T cell interaction is complex since both APC and T cells express various Notch ligands and receptors 12, 16, 20, 21. Moreover, a number of reports describe an impact of Notch signaling on differentiation and maturation of APC subsets 22–25. Therefore, we used in vitro stimulation of purified CD4+ T cells with anti-CD3 plus the recombinant Notch ligands Dll1, Dll4 and Jagged1 to specifically determine their impact on activation of T helper cells.
We found that the various ligands extensively differ in their binding capacity for Notch. More interestingly, they differentially regulate T cell activation. Whereas Dll4 enhances proliferation and the expression of early activation markers CD69 and CD25, Dll1 induces a partial and Jagged1 an almost complete inhibition of T cell activation. The observed effects are independent of γ-secretase activity. Ligand-mediated cross-linking of Notch has to coincide with a stimulatory TCR signal in order to be effective. Taken together, these data suggest a direct crosstalk between TCR and Notch signal transduction and imply Notch as a costimulatory molecule with the ability to differentially regulate the activation state of T cells.
Soluble Notch ligands
To investigate the effect of ligand-induced cross-linking of Notch on T cell activation, we used recombinant IgG fusion proteins of Dll1, Dll4 and Jagged1 22, 26–29 for in vitro stimulation assays. Immobilized ligands were functionally active as shown by induction of luciferase activity from a CBF1-Luc reporter construct in Jurkat T cells (Fig. 1). Dll4 showed the strongest induction of Notch signaling followed by Dll1 and Jagged1. The induction of the reporter gene could be blocked by using the γ-secretase inhibitor, N-[N-(3,5-difluorophenacetyl-L-alanyl)]-S-phenylglycine t-butyl ester (DAPT), which is known to inhibit the release of NICD, a crucial step in the Notch signaling pathway (Fig. 1).
Notch ligands have different binding potential for Notch on peripheral T helper cells
The expression of Notch receptors on T cells has been demonstrated on the mRNA level and by Western blotting. Recently, Notch1 was also detected by immunofluorescent staining with a specific antibody 30. However, for T helper cells the receptor binding capacity of individual Notch ligands has not been analyzed. Therefore, we used soluble Notch ligands for immunofluorescent staining of purified CD4+ T cells.
Dll4 binding to resting CD4+ T cells was readily detectable and the binding intensity was 4 to 6-fold up-regulated upon T cell activation using plate-bound anti-CD3 and anti-CD28 for 16 h (Fig. 2A). This is in accordance with published mRNA data 19. In comparison, binding of Dll1 was intermediate, whereas Jagged1 yielded only a weak receptor staining on activated cells as compared to an isotype control (Fig. 2B). As an additional specificity control, we blocked the Ca2+-dependent binding of the Notch ligands to their receptor by adding 4 mM EDTA 31, 32 (data not shown). Thus, the binding characteristics correspond well to Notch-dependent transcriptional activation. An increased receptor binding after T cell activation compared to non-activated cells was detected for all ligands (data not shown). Although this approach does not allow a discrimination between the different Notch isoforms on the T cell surface, it clearly demonstrates the homogenous expression of Notch receptors on all CD4+ T cells, its activation-induced increase and different binding capacities of the three ligands tested.
Notch ligands differentially regulate early T cell activation
Next, we analyzed the impact of individual Notch ligands on T cell activation. To rule out the unpredictable influence of naturally expressed Notch ligands on both APC and T cells, we used immobilized recombinant Notch ligands for in vitro costimulation of highly purified CD4+ T cells activated with plate-bound anti-CD3 and soluble anti-CD28. Interestingly, distinct Notch ligands differentially affected the expression of the early activation marker CD69 when compared to control IgG. As shown in Fig. 3A, after 24 h of stimulation Dll4 increased the frequency of CD69+ T cells, whereas Dll1 and Jagged1 led to a marked decrease in CD69 expressing T cells. The expression of CD25, the α-chain of the IL-2 receptor, was affected accordingly (data not shown). The effects of Notch ligands could not be abrogated by 10 µM of the γ-secretase inhibitor DAPT, indicating a mechanism independent of release of NICD. Soluble Notch ligands added to the culture at concentrations of up to 50 µg/mL had no effect on anti-CD3/anti-CD28-induced T cell activation, suggesting that Notch cross-linking is necessary for signal induction (33–35 and data not shown).
A kinetic analysis of CD69 and CD25 expression during the first 6 h of stimulation revealed that the Notch ligand-mediated effects can be observed already at the earliest time points of detectable T cell activation (1–4 h). This indicates that Notch cross-linking by its natural ligands directly interferes with the early signaling events of T cell activation (Fig. 3B and C).
Notch ligands differentially regulate T cell proliferation
To examine the impact of Notch ligands on T cell proliferation, CD4+ T cells were labeled with CFDA-SE and stimulated as described above. After 55 hours, cells were analyzed for proliferation by FACS and the average number of cell divisions per precursor was calculated (see 36 and Materials and methods section for details). In accordance with the CD69 and CD25 expression data, Notch ligands also differentially affected the proliferation of Th cells (see Fig. 4). Whereas Dll4 increased the average number of cell divisions per precursor, Dll1 and even more pronounced Jagged1 led to a dose-dependent reduction of T cell proliferation (Fig. 4A–C). Again, the effects were independent of γ-secretase-mediated cleavage of NICD as shown by the identical proliferation observed in the presence of DAPT (Fig. 4A).
Ligand-induced Notch activation interferes with T cell receptor signaling
TCR triggering results in activation and nuclear translocation of the transcription factors NF-κB, AP-1 and NF-AT, leading to transcription of IL-2 and other genes involved in T cell activation and proliferation 37–39. To analyze the Notch ligand-induced modulation of T cell activation on the molecular level, we performed electrophoretic mobility shift assays (EMSA) for all three transcription factors and compared their activity in the presence or absence of Jagged1. Strikingly, when T helper cells were stimulated with anti-CD3/anti-CD28 in the presence of Jagged1 the induction of NF-κB as well as AP-1 and NF-AT was completely abolished (Fig. 5A–C).
Physiological T cell activation can be mimicked by PMA and ionomycin, which directly activate protein kinase C (PKC) and thereby bypass the need for TCR triggering. In order to test whether the inhibitory effect of Jagged1 is mediated downstream of PKC, we analyzed NF-κB activity and CD69 expression in response to PMA/ionomycin. There was no inhibition of NF-κB (Fig. 5A) or CD69 expression (Fig. 5D) observed even when limiting concentrations of PMA/ionomycin were used for stimulation. These data indicate that Notch signaling interferes with early TCR signaling events upstream of PKC(theta) activation. To rule out that the inhibition of proliferation in the presence of Dll1 and Jagged1 is merely a secondary effect due to a lack of IL-2 production during initial activation we titrated exogenous IL-2 to the cultures. Neither the suppression of CD69 nor the inhibition of proliferation could be reversed by the addition of IL-2 (0.01 –10 ng/mL) (Fig. 5E and data not shown).
These data suggest that Notch ligands can interfere with early events of TCR signaling upstream of PKC(theta) activation via a mechanism which is independent of Notch cleavage and the induction of Notch-regulated target genes. In accordance with this, stimulation of CD4+ T cells with the various Notch ligands prior to activation by anti-CD3 had no effect on expression of CD69, CD25 or proliferation (data not shown), indicating that TCR cross-linking and Notch receptor/ligand interaction have to occur simultaneously.
The Notch signaling pathway is involved in cell fate decisions at many steps in the development of multicellular organisms. Lineage commitment within the hematopoietic system, including T cell development in the thymus, is critically regulated by Notch (reviewed in 8–11). Several recent reports suggest a role for the Notch pathway during differentiation of mature naïve T cells into functionally distinct effector T cell subsets in the periphery. Due to the complexity of potential Notch receptor/ligand interactions the molecular details of these processes are not yet clearly defined; a fact that is illustrated by a number of inconsistent reports published on the function of Notch in peripheral T cell activation and differentiation (reviewed in 11, 40).
Components of the Notch pathway are expressed by both APC and T cells. Naive T cells express Notch1 and Notch2 12, 16. Moreover, mRNA levels of all four receptor isoforms are increased upon T cell activation implying a potential involvement of more than one Notch isoform 19. On the other hand APC, in particular DC, express a number of Notch ligands such as Jagged1, Jagged2, Dll1 und Dll4 16, 21. Generally, all ligands are capable of interacting with all Notch receptors 41–43. The multitude of possible receptor/ligand interactions and the potential effects on both T cells as well as APC, which also express Notch receptors 22–25, make it rather difficult to define the contribution of an individual Notch receptor/ligand interaction to T cell activation and differentiation.
In an attempt to clarify the influence of distinct Notch ligands on T cell activation, we used several recombinant Notch ligands for costimulation of highly purified CD4+ T cells in an in vitro stimulation assay. We have demonstrated that individual ligands markedly differ in their capacity (i) to bind to CD4+ T cells, (ii) to activate the classical Notch signaling pathway in T cells, and (iii) to modulate anti-CD3-induced T cell activation. We show that Dll4 promotes activation, whereas Dll1 and Jagged1 largely inhibit activation of naïve CD4+ T cells.
The modulation of T cell activation is independent of the classical Notch pathway
All ligands tested were able to induce the classical Notch signaling pathway, including cleavage of NICD and transcriptional activation of CBF-1. While a specific γ-secretase inhibitor efficiently blocked this pathway, it did not affect the observed modulation of T cell activation by Notch ligands. A number of reports point to an alternative Notch signaling pathway which appears to be independent of CBF-1 and probably also of receptor cleavage by the γ-secretase complex (reviewed in 44). Amsen et al. 16 demonstrated that the Notch-induced Th2 differentiation is CBF-1 dependent, whereas the Th1 polarization instructed by Dll ligands is not influenced by a γ-secretase inhibitor and therefore probably not mediated by the classical Notch signaling pathway. However, these alternative signaling pathways are thus far not clarified.
Whereas classical Notch signaling involves de novo transcription of several target genes, the Notch-induced suppression of T cell activation occurred without any detectable time delay and could be observed after 45 min for NF-AT/AP-1 activation and after 1 h for CD69 expression. Stimulation in the presence of Jagged1 led to a simultaneous inhibition of all three major transcription factors, NF-AT, AP-1 and NF-κB, which are induced by TCR signaling. This is in accordance with other reports showing that Notch1 can inhibit NF-AT/AP-1 promotor activity in Jurkat cells 45 and AP-1 activity in a number of other cell lines 46. Notch triggering had to occur synchronously to T cell activation in order to be effective. Furthermore, PMA/ionomycin stimulation was not inhibited, indicating that the inhibitory effect observed after anti-CD3 stimulation occurs early in the TCR signaling cascade, i.e. upstream of PKC(theta) activation.
Altogether, these data indicate a direct interference of Notch with TCR signaling most likely at a membrane-proximal site. This probably involves recruitment of Notch into the immunological synapse, as has been demonstrated recently for Notch1 and the regulator of Notch activity, Numb 30, 47.
How can distinct Notch ligands differentially affect T cell activation?
Currently the role of Notch signaling in T cell activation is not clearly defined. Eagar et al. 17 demonstrated that a stimulatory anti-Notch antibody, as well as a Jagged1 or Dll1 expressing B cell line, suppressed T cell proliferation in vitro. In sharp contrast, two other reports showed that a γ-secretase inhibitor, a specific inhibitor of Notch signaling, drastically decreased T cell proliferation 18, 19. In this case, overexpression of a constitutively active Notch1 led to increased proliferation; suggesting that an activation of the Notch pathway enhances rather than inhibits T cell proliferation 19. Our data provide evidence that these opposing effects may at least partially be accounted for by the nature of the Notch ligand. In fact, Dll4 is capable of promoting T cell activation, whereas Dll1 and Jagged1 exhibit inhibitory effects. It has been described before that different Notch ligands can induce differential effects in the same target cell, e.g. for the differentiation of lymphoid precursors 48, 49 and for Th1 vs. Th2 polarization of naive peripheral T cells 16, but the molecular basis for these differences are not clear. Interestingly, we find considerable differences in the binding capacity of distinct Notch ligands to both resting and activated T cells. Dll4 exhibits the strongest binding followed by Dll1, whereas Jagged1 shows very weak receptor binding. Individual ligands may selectively bind distinct Notch isoforms which vary in their expression level and thus mediate qualitatively or quantitatively distinct effects. In line with such a model the overexpression of constitutively active Notch3 was found to mediate Th1 differentiation 15, whereas Notch1 induced Th2 16. However, binding and recruitment of different Notch isoforms under physiological conditions are currently difficult to analyze due to the lack of specific antibodies.
Alternatively, different affinities of the various Notch ligands for the same Notch isoform(s) may induce quantitatively different signals. For example it is known that glycosylation of Notch by Fringe enhances binding of Dll ligands but reduces binding of Jagged family members 50, 51. Such quantitative differences may affect recruitment of Notch into the immunological synapse and provide an explanation how a weakly binding ligand such as Jagged1 can profoundly affect T cell activation. It is known that low-affinity receptor-ligand interactions, e.g. the TCR/MHC interaction, can induce efficient recruitment of receptors to the contact site when integrated into two-dimensional surfaces. In this way, a low-affinity ligand could in fact provoke a prolonged half-life of Notch within the synapse as compared to high-affinity ligands. As binding might not immediately result in receptor cleavage and its subsequent clearance from the cell surface, this might prolong the possible interaction between Notch and membrane-proximal TCR signaling events.
Modulation of T cell differentiation by Notch
T cell differentiation is a multilayered process integrating the quality and quantity of TCR/MHC interactions and co-stimulatory signals derived from membrane-bound molecules and/or cytokines. Eventually, this leads to the activation of master transcription factors T-bet in Th1 or GATA-3 in Th2 cells. Evidence now exists that this process is modulated by Notch at several levels, either by affecting the T cell activation or by directly instructing differentiation. Amsen et al. 16 demonstrate direct induction of GATA-3, and hence Th2 differentiation, by Jagged1 via the classical pathway. On the other hand, the instruction of Th1 by Dll was found to be clearly independent of CBF-1. However, another study describes the induction of Th1 by constitutively active Notch3 15, implying an involvement of the classical pathway. In conclusion, differentiation of naive T cells seems to be affected by Notch via both CBF-1-dependent and independent pathways.
Our finding that Notch can directly modulate TCR signaling strength provides an explanation for such CBF-1-independent effects. In fact, the impact of TCR signal strength on T cell differentiation has been demonstrated in many experimental systems. Several studies suggest that weak signals lead to Th2 differentiation 52–54, whereas Th1 development is favored by stronger signals 52, 55, 56. Accordingly, individual Notch ligands could instruct distinct T cell fates by differentially modulating TCR signaling favoring either Th1 or Th2 differentiation.
It is now clear that Notch regulates both activation and differentiation of peripheral T cells at multiple levels and via different mechanisms. In a complex network of Notch receptor/ligand interactions, in which the constituents are expressed by both APC and T cells, opposing effects can be exerted on the same cell. Therefore, it remains to be determined how and when one particular branch of the network becomes active or dominating and how Notch signaling is orchestrated in order to regulate an ongoing immune response.
Materials and methods
BALB/c mice were purchased from the BgVV (Berlin) and housed under specific pathogen-free (SPF) conditions. Mice were used at 6-to-12 weeks of age.
The following antibodies were either conjugated in house or purchased as indicated: purified anti-mCD3 (145–2C11, BD-PharMingen, San Diego, CA), anti-mCD28 (37.51, BD-PharMingen), biotinylated anti-mCD69 (H1.2F3, BD-PharMingen), biotinylated anti-mCD25 (7D4, BD-PharMingen), fluorescein isothiocyanate (FITC)-conjugated anti-mCD4 (GK1.5), phycoerythrin (PE)-conjugated anti-mMHCII (M5/114), PE-conjugated anti-mCD62L (MEL-14). For coating of cell-culture plates, polyclonal anti-hamster IgG and anti-human IgG were purchased from Jackson ImmunoResearch.
Isolation of murine naïve CD4+ T cells was performed as follows: spleen and lymph node cells were labeled with anti-FITC multisort microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) and sorted for CD4 expression using the AutoMACS™ (Miltenyi Biotec). After release of anti-FITC microbeads, the cells were labeled with anti-CD62L microbeads (Miltenyi Biotec) and sorted for naive T cells. The purity of the various sorted cell populations was higher than 96%.
Isolated CD4+ T cells were washed with PBS, resuspended in a 1-µM solution of carboxy-fluorescein diacetate-succinimidyl ester (CFDA-SE) (Sigma, St Louis, MO) at a density of 1 × 107/mL and incubated for 3 min at room temperature. The labeling reaction was stopped by washing with RPMI 1640 culture medium (PAA Laboratories, Cölbe, Germany) containing 10% FCS.
T cell stimulation
Flat-bottom 96-well cell culture plates were coated with 10 µg/mL anti-hamster IgG and 10 µg/mL anti-human IgG in PBS overnight at 4°C. Plates were washed three times with PBS and then coated with 1 µg/mL anti-CD3 antibody and the various Notch ligand fusion proteins (concentrations as indicated) in PBS for 4 h at 37°C. Plates were washed three times with RPMI 1640 Medium containing 10% FCS and directly used for T cell stimulation. The costimulatory anti-CD28 antibody was supplied at 1 µg/mL in the culture medium.
In some cases the γ-secretase inhibitor DAPT (Merck Biosciences, Schwalbach, Germany) was added at the indicated concentrations.
Proliferation was quantified as previously described 36. Briefly, the number of cell divisions was calculated by determining the number of non-divided precursor cells for cells in each generation, and calculation of the cell divisions undergone by those precursor cells. Numbers are given as cell divisions per precursor cell.
Staining with ligand fusion proteins
Notch staining using the ligand-human IgG fusion proteins was done in RPMI 1640 medium at room temperature for 15 min. Fusion proteins were used at a final concentration of 33 nM. After washing, the cells were incubated with a PE-conjugated anti-human IgG polyclonal antibody.
Electrophoretic mobility shift assay
A small scale method for the preparation of nuclear extracts was used. Cells were washed in PBS, pelleted and incubated with approximately 200 volumes low-salt buffer (10 mM HEPES pH7.9, 10 mM NaCl, 0.1 mM EDTA, 0.1 mM EGTA) for 15 min on ice. Subsequently, 0.6% NP40 was added and the lysate immediately centrifuged for 2 min at 13 000 rpm. The resulting nuclear pellet was washed with low-salt buffer and nuclear proteins were eluted with 10 volumes of high-salt buffer (350 mM NaCl, 1 mM MgCl, 0.5 mM EDTA, 0.1 mM EGTA, 1% NP40, 20 mM HEPES, 20% glycerol, pH 7.9). All protein lysis buffers contained protease inhibitor cocktail (Complete, Roche Diagnostics, Mannheim, Germany), 1 mM DTT, 10 mM NaF, 8 mM β-glycerophosphate and 0.2 mM Na-vanadat. Protein concentration was determined in each extract by Bradford assay and equal amounts of protein were assayed for DNA-binding by EMSA as described 57. Briefly, 3.5 µg of protein were incubated for 30 min at 30°C with 32-P-labeled probe (10 000 cpm) in 20 µL binding buffer (20 mM HEPES pH7.9, 60 mM KCl, 2 mM DTT, 4% Ficoll, 2 μg BSA and 2 µg poly dI/dC). Subsequently, the complexes were resolved by electrophoresis on a native 5% polyacrylamide gel (in 0.5xTBE). Probes used were H2 K: 5′-gatcCAGGGCTGGGGATTCCCCATCTCCACAGG-3′, AP1-TRE: 5′-agctAGCATGAGTCAGACAC-3′ and ARRE2: 5′- gatcAAAGAGGAAAATTTGTTTCATAC-3′ (only the sense strand is shown; lower case letters are overhangs to label the probe with Klenow-fragment polymerase).
Transfection of Jurkat cells
Jurkat cells (ATCC) were transfected using the nucleofection™ technology by amaxa biosystems (Cologne, Germany), according to the manufacturer's protocol. Briefly, 2 × 106 cells were suspended in 100 µL buffer V (amaxa biosystems) and transfected with 5 µg of 12xCBF1-Luc reporter construct and 0.3 µg of renilla luciferase control vector pRL-TK (Promega, Mannheim, Germany) using the nucleofector™ program C-16. Subsequently, cells were cultured in 1.5 mL pre-warmed RPMI 1640 medium containing 10% FCS.
Four hours after transfection, Jurkat cells were stimulated with plate-bound Notch ligands (as described above) for 16 h. Afterwards cells were harvested, washed once with PBS and lysed in Passive Lysis Buffer (Promega) by pipetting up and down and incubating for 15 min at room temperature on a shaker. Afterwards 20 µL of lysate was analyzed for luciferase activity using the Dual Luciferase Reporter Assay (Promega) at a Monolight™ 3096 Microplate Luminometer (BD Biosciences). Firefly luciferase activity was normalized in each sample to renilla luciferase activity.
Sascha Rutz was supported by a grant from the Boehringer Ingelheim Fonds. This work was supported by SFB 650.