Originally published online as doi:10.1189/jlb.0303089 on October 2, 2003
Published online before print October 2, 2003
(Journal of Leukocyte Biology. 2003;74:971-981.)
© 2003
by Society for Leukocyte Biology
Notch signaling in the immune system
Gerard F. Hoyne1
ACRF Genetics Laboratory and Medical Genome Centre, John Curtin School of Medical Research, Australian National University, Canberra
1Correspondence: ACRF Genetics Laboratory and Medical Genome Centre, John Curtin School of Medical Research, Australian National University, Mills Road, P.O. Box 334, Canberra, ACT, Australia 2601. E-mail: Gerard.Hoyne{at}anu.edu.au
 |
ABSTRACT
|
|---|
Notch signaling plays a preeminent role during development in not only regulating cell fate decisions, but it can also influence growth and survival of progenitor cells. In the immune system, Notch is required for the maintenance of hematopoietic stem cells and in directing T- versus B-lineage commitment. In this review, I will summarize some of the recent findings relating to the function of Notch in the immune system during lymphocyte development and in the generation and function of mature cells.
Key Words: development • haematopoiesis • lymphopoiesis • lymphocytes
|
INTRODUCTION
|
|---|
The Notch signaling pathway (NSP) plays a key role in regulating tissue patterning during embryogenesis and can also influence cellular responses in the adult. Development of the immune system begins in embryonic life and continues throughout the lifetime of the individual. The immune system generates a wide diversity of cell types, each with distinct effector functions that are required to respond and eliminate pathogens. Notch signaling is critical during lymphocyte development, and dysregulation of the pathway can give rise to leukaemia. In this review, I will present a general description of the main signaling components involved in the Notch pathway and examine its role in lymphoid development and how it can affect the growth and function of mature lymphocytes.
|
NOTCH SIGNALING
|
|---|
The NSP regulates cell growth and differentiation in a wide range of organs, from insects to humans [1
]. Notch was first identified in Drosophila and encodes a single-pass transmembrane protein that contains multiple structural motifs. The receptors are expressed in a wide range of different cell types in temporally and spatially restricted patterns. The extracellular domain contains multiple tandem epidermal growth factor (EGF)-like repeats and three Lin-12/Notch motifs that prevent inappropriate cleavage of the receptor before it binds ligand (Fig. 1
) [1
]. The intracellular region contains a RAM domain that binds the CSL protein, which binds to the intracellular domain of Notch [1
]. It also contains iterated ankyrin repeats that participate in protein–protein interactions with CSL and other proteins and a C-terminal PEST domain that regulates protein stability [1
]. Notch can bind to two different ligands, Delta and Serrate/Jagged. Vertebrates express multiple Notch receptors (Notch1–4) and ligands including Delta-like (1–4) and Jagged 1 and Jagged 2 [1
] (Fig. 1
and Table 1
).
View larger version (50K):
[in this window]
[in a new window]
|
Figure 1. Structure of Notch receptors and ligands. Schematic representation of the receptors and ligands involved in Notch signaling. The extracellular domain of Notch receptors contains a variable number of EGF-like repeats and a Notch/Lin-12 (N/Lin12) motif. The intracellular domain contains the regulation of amino acid metabolism (RAM) domain and ankyrin repeats, which bind the chick brain factor-1, Su (H), Lag-1 (CSL) protein, a family of nuclear transcription factors. The noncoding region (NCR) domain helps regulate the functional activity of the Notch receptors. The Notch1 and Notch2 proteins also possess a transactivation domain (TAD), and all four receptors have a proline, aspartate, serine, and threonine (PEST) domain that is thought to regulate protein stability. NICD, Notch intracellular domain; DSL, Delta/Serrate/Lag-2.
|
|
The ligands tend to be expressed in a more highly restricted pattern than the receptors. They also contain EGF-like repeats and a transmembrane domain, and the ligands bind to the EGF repeats 11 and 12 of the Notch receptor via the DSL domain [2
]. The Jagged ligand also contains an additional cysteine-rich domain. Upon ligand binding, the NICD protein is cleaved and translocates to the nucleus, where it interacts with CSL protein (Fig. 2
). Normally, CSL proteins mediate transcriptional repression through binding to nuclear repressor protein complexes. The NICD can displace the repressor complex and convert the CSL protein into a transcriptional activator [3
] (Fig. 2)
. A major target of the CSL activation is the basic helix-loop-helix family of proteins encoded by the HES genes (Table 1)
. HES proteins contain a C-terminal WRPW sequence that is used to recruit transcriptional CoRs of the groucho family and thereby down-regulate transcription [4
].
View larger version (24K):
[in this window]
[in a new window]
|
Figure 2. Signaling from the activated Notch receptor. Upon ligand binding, the Notch receptor undergoes proteolytic cleavage. The first cleavage occurs within the extracellular domain and then within the cytoplasmic domain, which involves a furin-like protease, and a  -secretase activity, which requires presenilin. The NICD is released from the membrane and can bind to the CSL protein. In the absence of Notch signaling, the CSL protein associates with nuclear corepressor (CoR) proteins in a complex and mediates transcriptional repression. Upon Notch activation, the NICD can displace the nuclear repressors and allow CSL to bind coactivators (CoAs) and thus mediate transcriptional activation leading to hairy/enhancer of split (Hes) gene expression. Notch signaling can also function in a CSL-independent manner that involves the deltex (Dx) protein. Dx can bind to the ankyrin repeats on NICD, and this complex can inhibit Ras activation. The details of the CSL-independent pathway of Notch activation are still poorly understood. Modulators of Notch signaling include Dx, Fringe (Frng), and numb. In lymphocytes, these three proteins act to inhibit Notch signaling. ADAM, Tumor necrosis factor  (TNF-)-converting enzyme; Mam, Mastermind.
|
|
|
NOTCH CONTROLS CELL FATE DECISIONS
|
|---|
There are three different ways in which Notch signaling can control cell fate decisions in the embryo. Lateral inhibition is a process that occurs during neurogenesis in insects and vertebrates. It is a mechanism in which isolated cells within a field of cells that share the same developmental potential can undergo specification and simultaneously inhibit the neighboring cells from following the same developmental pathway. This process requires the expression of the Notch ligand Delta on the cell that has committed to become a neuron, and the specified cell can engage the Notch receptor on neighboring cells [5
]. Lateral inhibition works via a positive-feedback mechanism that amplifies small differences in the expression of the receptors and ligands on individual cells. Activation of Notch signaling between neighboring cells normally involves induction of the HES family of transcription repressors that inhibits the expression of tissue-specific transcription factors. The process of lateral inhibition effectively allows cells to choose between two distinct cell fates. In this context, Notch signaling is thought to inhibit a default program of cell differentiation and thus can promote the adoption of a secondary cell fate [6
]. A second process in which Notch signaling is involved is lateral specification [6
]. This usually involves progenitor cells that have the capacity to differentiate along multiple lineages, for example, during haematopoiesis [7
]. In this process, cells are directed to adopt a particular cell fate as a result of receiving a signal through the Notch receptor. Thus, the access of receptor-bearing cells to a source of Notch ligand will drive cellular differentiation. Bone marrow stromal cells would provide a source of Notch ligands, and these would interact with the Notch receptor on progenitor cells. Thus, the presence or absence of a Notch signal would help direct the progenitors along the correct lineage path. Border formation is also controlled in part by Notch signaling [8
]. During development, tissues need to be specified into different compartments such as anterior-posterior and dorsal-ventral (D-V). This is particularly relevant to limb development, as establishment of these different compartments ensures the proper proximal-distal outgrowth of the limb. It is not surprising that wing development in the fly has proven an excellent experimental model system to study the role of Notch signaling in this process.
Notch signaling can also influence cell growth and differentiation without having to influence cell fate decisions. There are several examples in Drosophila and Caenorhabditis elegans, whereby activation of the NSP can induce cell-cycle arrest, promote cellular proliferation, or even promote apoptosis [9
10
11
]. Alternatively, activation of Notch signaling can protect cells from apoptosis [12
, 13
]. Thus, the pleiotropic effects that Notch signaling can exert on progenitor cells will be determined in part by the strength and duration of the Notch signal delivered, as well as on the developmental context of the cell.
|
MODULATORS OF NOTCH SIGNALING
|
|---|
Known modulators of Notch signaling include Fringe (Frng) and Deltex (Dx). The Frng proteins form a family of acetyl glucosaminyl transferases that add O-linked fucose to the Notch extracellular domain [14
]. There are three Frng proteins in vertebrates—Lunatic, Manic, and Radical—and Drosophila has a single Frng protein [14
] (Table 1
, Fig. 2
). Recent studies have identified that Frng proteins function in the Golgi, and they modify the extracellular domain of Notch before its transit to the cell membrane [15
16
17
18
]. This modification can affect Notch ligand binding, such that Frng-modified Notch receptors bind preferentially to Delta rather than Serrate/Jagged ligands [15
16
17
18
]. Furthermore, the different Frng proteins in vertebrates appear to modify distinct EGF repeats on the Notch receptor [19
]. Frng plays a key role in D-V patterning of the wing disc during Drosophila embryonic development, where it acts in a cell-autonomous manner to establish a feedback loop involving Notch signaling and to direct the formation of the compartment border [20
]. In vertebrates, Frng proteins are involved in patterning a variety of tissues, and studies with gene knockout mice suggest that there is functional redundancy between the different proteins [21
, 22
].
Studies on somite formation in the embryo have highlighted that Frng proteins can also act as negative regulators of Notch signaling. A number of NSP genes are expressed in the presomitic mesoderm (PSM) and undergo regular oscillations to direct the temporal development of somites. These include the Notch ligands Delta-like1 and Delta-like3 as well as Lunatic Frng (L-Frng), and Hes-1 [23
24
25
26
27
]. Recent studies in the chick PSM have indicated that the L-Frng protein helps to establish a negative-feedback loop, whereby L-Frng acts to inhibit Notch signaling and thus down-regulate L-Frng expression [23
]. Analysis of Frng protein function in the vertebrate-immune system has been limited to thymocyte development. L-Frng is expressed in the thymic medulla and when overexpressed in T cells, can abrogate commitment to the T cell lineage but allows B cell differentiation to proceed. This phenotype is identical to that obtained with conditional Notch1 loss-of-function in thymocytes (see below) and indicates that like in the PSM, L-Frng can act as a negative regulator of the NSP in thymocytes [28
].
Another modulator of the NSP is a novel cytoplasmic protein, Dx (Table 1
, Fig. 2
). Dx was first identified in Drosophila as a positive regulator of Notch signaling that functions independently of CSL activation [29
, 30
]. The Dx protein has three structural motifs: The N terminus is required for its interaction with the ankyrin repeats on NICD; it has a proline-rich motif that can bind to Src homology 3 domain-containing proteins; and the C terminus has a Really Interesting New Gene-H2 (RING-H2) domain that is involved in protein–protein interactions and recruiting E3 ubiquitin ligases [30
, 31
]. However, Dx can also act as a negative regulator of Notch signaling. Overexpression of Dx in hematopoietic stem cell (HSC) precursors leads to inhibition of T cell development in the thymus but allows precursors to differentiate instead as B cells [32
]. Exactly how Dx functions in lymphocytes is currently unknown, but it is thought that it can dampen Notch signaling by preventing the NICD from recruiting transcriptional CoA proteins (Fig. 2)
. A recent report describes that the human family of Dx proteins (Dtx1–3), as well as a related member B-lymphoma- and B aggressive lymphoma-associated protein, could all act as E3 ubiquitin ligases [33
]. The Dx proteins can form homo- and heterodimers [34
], and it has been noted that heteromeric RING-finger protein complexes can display increased E3 ubiquitin ligase activity [35
, 36
]. Therefore, as a result of its ubiquitin ligase activity, Dx proteins may be able to target the Notch receptor or the activated NICD for protein degradation.
|
NOTCH SIGNALING AND REGULATION BY UBIQUITIN
|
|---|
Several studies have highlighted the important role for ubiquitin-mediated degradation of Notch receptors, ligands, and intracellular signaling components [37
] (Table 1)
. Recently, the neuralized gene product in Drosophila and the mind bomb gene in zebra fish were demonstrated to contain E3 ubiquitin ligase activity and are required for the internalization of the ligand Delta. This enzymatic activity appears to be a prerequisite for efficient Delta-Notch signaling [38
39
40
41
]. Additional negative regulators of Notch activity include Sel-10 and Su(Dx). These two proteins also possess E3 ubiquitin ligase activity that promotes ubiquitin-mediated degradation of NICD [42
, 43
].
A mutation was identified in the agouti locus (a18H) in mice that resulted in coat-color alterations and a constant itching of the skin. The itchy mice are characterized by profound immune defects [44
, 45
]. On a C57BL/6J background, itchy mice develop immune disorders that result in death at 6–8 months of age. The mutant mice die from hypoxia caused by chronic pulmonary interstitial inflammation and alveolar proteinosis; inflammation of the glandular stomach and skin, resulting in scarring as a result of constant itching; and hyperplasia of lymphoid cells, hematopoietic cells, and the forestomach epithelium [45
]. The spleen, lymph nodes, and thymus are all enlarged in the itchy mouse. On the JU/Ct background, the a18H mice develop an inflammatory disease of the large intestine [45
]. Further analysis of the itchy mice revealed that CD4+ T cells display augmented T helper cell type 2 (Th2) responses, indicating that the itch protein plays a role in regulating Th cell differentiation [46
]. The a18H mutation resulted from an inversion of the itch gene, and it was shown to encode a novel E3 ubiquitin protein ligase, a protein involved in ubiquitin-mediated protein degradation [44
]. In fact, the itch protein is the vertebrate homologue of the Su(Dx) protein. The itch protein can bind to NICD in vitro to mediate its ubiquitination and target it for degradation.
Cbl proteins function as E3 ubiquitin ligases that regulate cell-surface expression of activated receptors with receptor tyrosine kinase activity (e.g., the EGF receptor) through endocystosis and targeting them for ubiquitin-mediated degradation [47
]. Thymocytes predominantly express c-Cbl, and Cbl-b is more abundant in mature T cells [48
, 49
]. Cbl proteins are required for T cell receptor (TCR) sequestration and down-regulation following antigen receptor engagement and signaling in thymocytes and peripheral T cells [50
, 51
]. More recently c-Cbl has been shown to bind to a unique site at the C terminus of the NICD and can direct its ubiquitination [52
]. It remains to be seen how the relationship between Cbl and Notch proteins might affect the developmental processes of lymphopoiesis and haematopoiesis, as well as the function of mature lymphocytes.
|
NOTCH SIGNALING AND LYMPHOCYTE DEVELOPMENT
|
|---|
Notch signaling plays a critical role in haematopoiesis and lymphocyte development in the embryo and adult, and this has been the subject of two recent reviews [7
, 53
]. HSCs give rise to common lymphoid progenitors (CLPs) in the bone marrow, which can specify various cell lineages including T and B cells as well as dendritic cells (DC) and natural killer (NK) cells. The CLPs can migrate from the bone marrow and enter the thymus to allow T cell differentiation (Fig. 3
). Upon entering the thymus, progenitor cells must negotiate a series of cell fate choices. The first decision is whether to become a T or a B cell. If thymocytes commit to the T cell lineage, they must then decide whether to become a TCR-ß or TCR-
cell. If thymocytes choose the ß T cell lineage, they can undergo a well-defined program of cellular differentiation as defined by the expression of the CD4+ and CD8+ coreceptors at the cell surface [54
]. Immature double-negative (DN; CD4–CD8–) cells progress to the CD4+CD8+ double-positive (DP) stage of differentiation after successfully rearranging the TCR-ß and TCR- chain genes, respectively, and express a functional antigen receptor on the surface [55
]. At the DP stage, immature thymocytes must finally choose between the CD4+ or CD8+ cell fate. At this time, the antigen receptors on DP cells are tested for specificity to self-antigen, present on thymic epithelial cells and DC. If a thymocyte expresses a suitable TCR with low-to-medium affinity for self-antigen/major histocompatibility complex (MHC), it will complete its differentiation program and become a single-positive (SP) cell [54
]. If a T cell expresses a TCR with high affinity, it will be deleted. Following positive selection, mature T cells are exported from the thymus and enter the peripheral circulation.
It is now apparent that Notch1 signaling is crucial in CLPs in enabling these cells to negotiate correctly the T- versus B-lineage cell fate choice [56
] (Fig. 3)
. Loss of Notch1 signaling in CLPs completely blocks T cell development in the thymus and leads to the accumulation of B cells. The loss of Notch1 can also block the development of thymus-independent, intestinal intraepithelial T cells [57
, 58
]. A similar abrogation of T cell development in the thymus has been observed in mice with a conditional mutation in the RBP-J (CSL) gene [57
, 59
]. Like the Notch1-/- mice, the RBP-J-/- mice show an accumulation of thymic B cells. These findings suggest that Notch1 signals occur through CSL in a nonredundant manner. The loss of Hes-1 in thymocytes causes an arrest in early DN cell development before the pro-T cell stage, but there is no indication that the Hes-1 deficiency causes an accumulation of B cells in the thymus [60
]. Thus, Notch1 signaling is critical in early hematopoietic precursors to induce commitment to the T cell lineage. Further support for the role of Notch signaling in the T versus B cell lineage cell fate choice in early lymphoid progenitors has been obtained with gain-of-function studies with overexpression of the Notch signal modulators Dx1 and L-Frng [28
, 32
]. Expression of both of these proteins in lymphoid progenitors was sufficient to promote B cell development in the thymus at the expense of T cells. Bone marrow chimera studies revealed that Dx1 could act in a cell-autonomous manner, whereas L-Frng acted in a cell-nonautonomous manner in vivo [28
, 32
]. Thus, both of these proteins appear to function as negative regulators for Notch signaling in the thymocytes.
|
NOTCH LIGANDS IN LYMPHOCYTE DEVELOPMENT
|
|---|
Jagged1 and Jagged2 are expressed in bone marrow and fetal liver stromal cells in thymic epithelium and subsets of hematopoietic cells of different lineages [61
]. The Delta-like1 ligand is reported to be expressed in the cortico-medullary region of the thymus and in fetal and adult human hematopoietic tissues [62
]. Studies by Anderson et al. [63
] have shown that Notch ligand genes, Jagged1, Jagged2, and Delta-like1, are expressed exclusively in the thymus by MHC class II+ epithelial cells, whereas DC and thymocytes lack expression of these genes. Culturing thymocytes with epithelial cells was sufficient to induce Notch signaling in thymocytes, as measured by expression of the downstream target genes Hes-1 and Dx1; in contrast, culturing thymocytes alone in the absence of epithelial cell contact was unable to induce Hes-1 or Dx1 gene transcription [63
]. These results are compatible with a lateral specification model in which epithelial cells provide an inductive Notch signal to thymocytes rather than thymocytes themselves engaging in lateral signaling through T–T interactions.
Until recently, little was known about the functional role of individual Notch ligands in T cell development. The different Notch ligands appear to have redundant functions in early hematopoietic cells to allow cellular expansion but prevent terminal differentiation, but their effects can be context-dependent [64
65
66
67
68
]. Using a coculture system with a stromal cell line S-17 transfected with the Delta-like1 or Jagged1 ligand, Jaleco et al. [69
] showed that culturing human CD34+ bone marrow progenitors with the Delta-like1 ligand could completely block B cell differentiation but allowed the emergence of cells with characteristics of a T/NK cell precursor. In contrast, the Jagged1 ligand had no effect on B or T/NK cell differentiation. In a complimentary study, Schmitt and Zuniga-Pflucker [70
], using a different stromal cell line transfected with the Delta-like1 ligand, showed that these cells rapidly lose the ability to induce B cell differentiation in HSCs. Instead the Delta-like1-expressing stromal cells could support commitment of mouse DN cells to the TCR-ß and TCR- lineages as well as production of mature SP cells in the absence of a thymus. Finally, studies with Jagged2-deficient embryos support a role for this ligand in directing TCR- cell differentiation in the thymus, as Jagged2-deficient mice display a 50% reduction in TCR-+ cells in the fetal thymus [71
]. It has now been established that Notch1 can function as a nonredundant receptor to regulate the T- versus B-lineage decision in lymphoid progenitor cells, but at present, it is not known if only a single-ligand help regulates this cell fate decision. To confirm that Delta-like1 can function as a nonredundant ligand to induce T cell differentiation in CLPs will require the conditional deletion of the Delta-like1 gene at an early stage of T cell development. This type of approach will be required, as Delta-like1 knockout mice die very early in embryonic life—well before the thymus is seeded with progenitor cells [72
].
|
ß VERSUS T CELL DEVELOPMENT
|
|---|
After negotiating the T versus B cell lineage commitment step, immature pro-T cells begin to rearrange their TCR genes and must negotiate a second cell fate choice to decide whether to become a TCR-ß or a TCR- cell. Thymic DN1 cells begin to undergo rearrangement of their TCR-ß, , and genes. If a cell successfully rearranges its TCR-ß chain, this will pair with the pre-T chain and CD3 to form a pre-TCR complex, begin to proliferate, and then differentiate along the TCR-ß lineage. If a cell successfully rearranges its TCR- and TCR- genes, this promotes development as a TCR- cell. It still remains a contentious issue as to exactly where the TCR-ß/TCR- commitment point occurs during thymocyte development and whether signaling from the pre-TCR or TCR- alone is sufficient to induce differentiation along these separate lineages [73
]. Experimental studies have suggested that Notch1 signaling may play a role in this process. Studies by Washburn et al. [74
], using mixed bone marrow chimeras with mice reconstituted with Notch1+/+ and Notch1+/- marrow cells, showed that the Notch1+/- progenitors differentiated as TCR- cells in the presence of wild-type (Notch1+/+) cells. Second, enforced expression of an activated Notch1IC in thymocytes on a TCR-ß-/- genetic background, which prevents pre-TCR signaling, leads to an increased production of DP cells, indicating that a strong Notch1 signal could still allow commitment to the TCR-ß lineage [74
].
These results imply that the strength of a Notch signal received by a progenitor cell may help direct the TCR-ß versus TCR- cell fate choice (see Fig. 3
). A strong Notch signal should direct cells along the ß lineage, whereas cells that do not receive or perhaps receive only a weak Notch signal would differentiate along the TCR- lineage. Conditional deletion of Notch1 in thymocytes at the DN2 stage of differentiation using cre-lox technology leads to a severe block in TCR-ß development, but TCR- cell development was normal [75
]. There is still a possibility that the Notch1 gene was deleted after cells had gone beyond the ß/ cell commitment point. This means that further studies are required that would involve deleting Notch1 at an even earlier stage of development, preferably just after they had traversed the T/B-lineage commitment point.
|
INVOLVEMENT OF NOTCH IN THE DP-SP TRANSITION OF THYMOCYTES
|
|---|
Two models have been proposed to explain how DP cells differentiate to become SP cells. The instructive model predicts that if a DP cell receives a TCR signal in the context of MHC class I, then the cell will extinguish CD4 and maintain CD8 [76
, 77
]. Alternatively, if a cell were to receive a TCR signal through MHC class II, this would lead to down-regulation of CD8 and the maintenance of CD4. The stochastic model of T cell development predicts that DP cells randomly down-regulate their CD4 or CD8 coreceptors, and the cells that survive will only be those that coexpress the correct MHC-restricted TCR with an appropriate coreceptor [76
]. Several groups have provided evidence implicating a role for Notch signaling in the DP-to-SP transition, but this is still controversial, as results obtained from Notch1 gain-of-function studies appear to be at odds with data derived from Notch1 loss-of-function experiments. Two groups made separate transgenic mouse strains with T cell-specific expression of an activated Notch1 receptor. Robey et al. [78
] first reported that active Notch1 signaling in thymocytes skewed the T cell repertoire toward the CD8 lineage, and this could occur in the absence of MHC class I. The Bevan group, conversely, showed that activated Notch1 signaling in thymocytes could promote development of CD4+ and CD8+ cells, although CD8+ cell production was greatly enhanced, and that this could occur in the absence of interactions with TCR and MHC molecules [12
]. These discrepancies may be explained by the fact that different, activated Notch1 constructs were used in the separate mouse strains, and this could lead to alterations in the level of protein expression in thymocytes at different stages of development. In contrast, enforced expression of a full-length, intracellular Notch1IC construct in DP thymocytes by retroviral expression could prevent their differentiation into SP cells. The authors conclude that Notch1 signaling may help regulate TCR signal strength in DP thymocytes, and a failure to down-regulate Notch signaling at this stage would be incompatible with differentiation to the CD4 or CD8 lineages. The differences in findings between the retroviral and the transgenic mouse studies using constitutively activated Notch1 constructs were explained by the fact that higher levels of protein expression could be achieved in T cells using the full-length Notch1 retroviral construct compared with the constructs used with the transgenic mice. It is interesting to note that under physiological conditions, DP thymocytes normally down-regulate the levels of Notch1 receptor at the cell surface and re-express the receptor once they become mature SP cells [79
].
However, studies with conditional Notch1 knockout mice are difficult to reconcile with a role for Notch1 in later stages of thymocyte development. Removal of Notch1 at the DN3 stage of T cell development had no effect on the latter stages of T cell development, indicating that Notch1 is not required for the DP-SP transition [80
]. These latter results reinforce the idea that Notch1 signaling is critical to allow early thymic progenitors to negotiate the T versus B cell lineage cell fate choice rather than playing a major role in later stages of T cell development.
Using a different approach to examine the role of Notch signaling in thymocyte development, pharmacological inhibitors were used to block -secretase activity in fetal thymic organ culture. Under physiological conditions, -secretases are crucial for processing the Notch receptor following Notch ligand binding that releases the intracellular domain to translocate to the nucleus to induce signaling [81
] (Fig. 2)
. The -secretase inhibitors would be expected to reduce the amount of Notch signaling and potentially mimic a Notch1 loss-of-function phenotype. The use of high doses of inhibitors in vitro could block the transition of cells from the DN to DP stage of development, and lower doses blocked T cell development at later stages, in particular, by reducing the numbers of CD8+ cells [82
, 83
]. Using a reaggregated fetal thymic organ culture system, Yasutomo et al. [84
] showed that the addition of a blocking antibody to Notch1 or antisense RNA to reduce the levels of the Notch1 receptor in thymocytes leads to a decrease in CD8+ cells and an accumulation of CD4+. The data obtained from experiments using -secretase inhibitors suggest that some form of Notch signaling may be important in the DP-SP transition, but a role for Notch2 can now be excluded, as conditional Notch2 knockout mice show completely normal T cell development [85
]. Experiments with Notch3 and Notch4 knockout mice are keenly awaited to determine if one of these receptors is required for the later stages of T cell development.
|
NOTCH SIGNALING AND B CELL DEVELOPMENT
|
|---|
It is now evident that for B cell development to proceed in the bone marrow, Notch signaling must be extinguished in B cell precursors. This is highlighted by the finding that forced expression of activated Notch1 alleles in bone marrow precursors leads to precocious T cell development and inhibition of B cell development [86
]. Understanding how Notch signaling regulates the T- versus B-lineage decision will require a better understanding of the target genes that are regulated by Notch in B and T cells. The nuclear proteins Pu.1, E2A, early B cell factor, and Pax-5 regulate commitment to and maturation of B-lineage cells from the CLP stage [87
], and the latter transcription factor is required to inhibit Notch1 expression [88
]. However, Notch may also inhibit B cell development by inducing apoptosis or cell-cycle arrest, as was observed in a chicken B cell line [89
].
Mature B cells leave the bone marrow and circulate in peripheral blood and can enter the spleen and lymph nodes. B cells express Notch receptors at various stages of development, and Notch ligands are expressed in the chicken bursa, a site of B lymphopoiesis, as well as in the bone marrow and the spleen of rodents. B cells cultured in the presence of the Jagged1 ligand induce Hes-1 expression and up-regulate CD23. Two types of transitional, mature B cell precursors have been identified in the spleen, and these are referred to as Type 1 and Type 2 [90
]. In the spleen, the immature, transitional B cells can differentiate further to become follicular (FC) B cells or marginal zone (MZ) B cells, and these are distinguished on the basis of expression of the cell-surface markers CD21 and CD23 [91
] (Fig. 3)
. FC B cells respond to T-dependent antigens and undergo extensive proliferation to form germinal centers and undergo somatic hypermutation to allow isotype switching. MZ B cells are an enigmatic population of cells that are thought to be required for responses to microbial antigens. Loss of Notch signaling in B cells as a result of conditional mutation in the RBP-J protein causes a significant loss of MZ B cells but does not affect FC B cell formation [92
]. Mice with a B cell-specific loss of RBP-J display a heightened sensitivity to blood-borne bacterial infection, consistent with the loss of MZ B cells [92
]. The loss of Notch signaling in mature B cells does not affect responses to T-dependent or -independent antigens, as isotype switching was comparable with that observed in wild-type mice. As the loss of the RBP-J did not affect the rate of apoptosis of MZ B cells, their migration to other tissues, or the rate of plasma cell differentiation, the authors concluded that Notch signaling must regulate the lineage choice between the MZ and FC B cell fate. However, further work is required to decipher the physiological roles for the different Notch ligands in this process.
A specific negative regulator of Notch signaling (Mint) was isolated using a yeast two-hybrid screen for RBP-interacting proteins [93
]. Mint is expressed in FC B cells and is low in MZ B cells. Analysis of Rag2 chimeric mice reconstituted with Mint-deficient fetal liver cells showed that the loss of Mint activity caused an increase in MZ B cells and a decrease in FC B cells in the spleen [93
]. Notch2 is highly expressed in mature B cells, and conditional deletion of Notch2 resulted in a decrease in splenic MZ B cells, whereas FC B cells in the spleen and B1 cells in the peritoneal cavity accumulated normally [85
]. Therefore, Notch2 can act in a nonredundant manner to specify the MZ versus FC B cell fate in the spleen, and a specific inhibitor of the Notch signaling pathway can ensure that B cells make the correct lineage choice.
|
NOTCH SIGNALING IN MYELOID AND DC DIFFERENTIATION
|
|---|
DC precursors differentiate in the bone marrow and are derived from a HSC that can give rise to two distinct lineages [94
]. They both express the DC marker CD11c and can be further distinguished by the expression of CD8, DEC205, and CD11b. The CD11b+ DEC205- CD8 - cells reside primarily in lymphoid tissues, and CD11b- DEC205+ CD8 + cells localize preferentially in the thymus [95
]. Original studies suggested that the CD8 + DC originated from an immature T cell precursor and hence were referred to as lymphoid DC, and CD8 - DC were of myeloid origin [96
]. However, it is now evident that the CD8 + DC are produced at normal levels in the thymus of mice that lack immature T cell progenitors [97
, 98
]. Therefore, myeloid precursors in the bone marrow must be capable of generating CD8 + and CD8 - DC [99
]. A third DC population has recently been described that can be distinguished by the expression of CD45 and Gr-1 markers and are thought to be the equivalent of human plasmacytoid DC [100
101
102
]. The developmental origin of these CD11c+ B220+ Gr-1+ cells is still not known, but they are present in significant numbers in the thymus and produce type1 interferons (IFNs) in response to viruses and CpG oligonucleotides [100
101
102
]. Studies with conditional Notch1-/- mice and bone marrow chimera studies have shown that Notch1 is not required for the generation of CD8 + or CD8 - DC or the plasmacytoid DC, as each of these cell populations accumulates normally in the absence of Notch1 [97
, 103
].
Thymic and splenic CD11c+ DC display differential expression of Notch ligand genes. Murine splenic CD11C+ DC express transcripts for Jagged1 and Jagged2 but are low in Delta-like1. Conversely, thymic CD11c+ DC express high levels of Jagged2 but lower levels of Jagged1 and Delta-like1 [104
, 105
]. CD11b macrophages express transcripts for Delta-like1 but are low for Jagged ligands. Peritoneal macrophages display only low-level expression of Jagged1 [104
]. Although the data suggest that Notch ligand expression is widespread amongst DC and macrophage populations, virtually nothing is known about the physiological role for the individual ligands in the function of these mature cell lineages.
In vitro experiments using primary cell culture with human peripheral blood monocytes suggest that Notch signaling may help regulate the macrophage/DC cell fate choice. Culturing blood monocytes in the presence of granulocyte macrophage-colony stimulating factor (GM-CSF) and an immobilized extracellular domain of Delta-like1 could inhibit their differentiation into macrophages [106
]. However, Delta-like1-induced Notch signaling could greatly increase the proportion of monocytes that differentiate into DC when cultured in the presence of GM-CSF and TNF- [106
]. The type of cytokine present during the time in which a precursor cell receives a Notch signal could also have important implications on the differentiation process. This is illustrated by the finding that Delta-like1-induced Notch signaling together with M-CSF can induce apoptosis of monocytes, whereas the delivery of a Delta-Notch signal together with GM-CSF protects cells from death [107
]. Therefore, the combinatorial effects of Notch and cytokine-induced signaling on immature cells can have distinct influences on the outcome of hematopoietic cell differentiation. Cell fate decisions, such as those that regulate macrophage versus DC lineage choice, will normally be made within defined microenvironmental niches and will be influenced by the presence of cytokines and Notch ligands present on stromal cells. Future studies should hopefully define more accurately which Notch receptor(s) and ligands contribute to the macrophage/DC cell fate decision under normal physiological conditions.
|
NOTCH AND PERIPHERAL IMMUNITY
|
|---|
Hoyne and colleagues [105
] demonstrated that forced expression of Jagged1 on splenic DC by retroviral gene transduction loaded with a peptide antigen could induce CD4+ T cells to develop as regulatory T cells (Tr), cells that could inhibit primary and secondary immune responses in an antigen-specific manner. In support of these findings, Yvon et al. [108
] have demonstrated a similar phenomenon with human naïve CD4+ T cells recognizing alloantigen on Epstein-Barr virus-transformed B cells transfected with the Jagged1 ligand. Recognition of antigen in the presence of the Jagged1 ligand induced T cells with a regulatory phenotype that could mediate suppression in vitro in an antigen-specific manner but did not affect the response of T cells specific for a third party antigen [108
]. These studies indicate that peripheral T cells are responsive to Notch signaling, but more work is required to understand how Notch signaling affects the outcome of TCR/costimulatory signaling in peripheral T cells and how Notch signaling can direct naïve T cells toward the regulatory T cell fate.
Studies by the Dallman group have also shown that L cells cotransfected with the Delta-like1 ligand and an allogeneic MHC antigen could induce long-lasting, antigen-specific tolerance in a vascularized heart allograft model in mice, and tolerance in this model required the presence of CD8+ T cells (Ken Wong et al., submitted). The recognition of the alloantigen on the Delta-likel-expressing L cells could decrease IFN- secretion by alloreactive CD4+ and CD8+ T cells in a mixed lymphocyte reaction in vitro [109
]. These latter studies hold promise that manipulation of Notch signaling in peripheral T cells might be advantageous under certain clinical settings where immune responses to self- or transplantation antigens or indeed foreign antigens such as allergens might need to be regulated.
CD4+ and CD8+ T cells constitutively express transcripts for Notch receptors and some ligands. Culturing naïve CD4+ T cells in the presence of anti-CD3 and anti-CD28 antibodies leads to T cell activation and down-regulation of Notch ligand gene expression (Jagged1 and Delta-like1) [110
]. If CD4+ T cells are activated in the presence of the immunoregulatory cytokines interleukin-10 or transforming growth factor-ß1 at the same time as TCR triggering, this can promote Notch ligand gene expression [110
]. The addition of IFN- does not induce Notch ligand gene expression in activated T cells. These results suggest therefore that cytokines can exert differential effects on Notch ligand gene expression in CD4+ T cells. Although these studies identify the conditions that can modulate Notch ligand gene expression, they do not provide any insight into the functional role that the individual ligands have regulating the differentiation or effector function of mature T cells.
|
NOTCH AND REGULATORY T CELLS
|
|---|
Naturally occurring CD4+ CD25+ Tr cells play a critical role in the maintenance of immune homeostasis in rodents and humans through their ability to suppress the growth of self-reactive lymphocytes [111
]. Lechler and colleagues [112
] have shown that under resting conditions, human Tr cells constitutively express low levels of Notch ligands (e.g., Jagged1 or Delta-like1) and Hes-1 but have elevated levels of Dx. Following TCR triggering, there was a dramatic rise in Notch4, Delta-like1, and Hes-1 expression but a decrease in Dx [112
]. We have found that murine CD4+ CD25+ Tr cells show increased levels of Delta-like1 and Notch1 transcripts compared with resting CD4+ CD25- cells. Following activation, Tr cells down-regulate Delta-like1 and up-regulate Notch1 (Hoyne et al. unpublished). Although Tr cells display constitutive expression of the Notch receptors and the Delta-like1 ligand genes, there is no direct evidence at present to suggest that Tr cells mediate their suppression through activation of Notch signaling in a target cell.
However, it is intriguing that the naturally occurring CD4+ CD25+ Tr cell populations in humans and rodents display constitutive expression of Notch ligand genes. Tr cells can be produced in the thymus as part of the normal T cell repertoire, and recently, it was shown that the fox p3 transcription factor was critical for their generation in vivo [113
114
115
]. There is no evidence at present to indicate that Notch signaling can regulate fox p3 expression in naïve T cells or alternatively, that fox p3 might regulate Notch receptor and/or ligand expression in Tr cells.
|
CONCLUSION
|
|---|
The NSP plays a diverse role in regulating the development of a vast number of different cell types in the immune system. Although most work has focused on its role in the generation of different cell lineages, it is apparent that Notch signaling is also required in mature cells. The combined approach, using gain-of-function as well as loss-of-function methods to study Notch signaling in the immune system, has provided some important insight into the way this pathway helps to regulate growth and cell fate decisions. For differentiation to proceed, cells must acquire independence of extrinsic growth factor signals and become more reliant on intrinsic programs of differentiation. The challenge ahead is to understand how Notch can integrate with other signaling pathways to regulate differentiation and to identify the possible feedback loops that are required to maintain or extinguish Notch signaling. Also, future studies may provide some fundamental insights into how modulation of Notch signaling might be used in a therapeutic manner to turn off or dampen unwanted immune responses in different clinical settings.
Received March 2, 2003;
revised July 21, 2003;
accepted July 22, 2003.
|
REFERENCES
|
|---|
- Artavanis-Tsakonas, S., Rand, M. D., Lake, R. J. (1999) Notch signaling: cell fate control and signal integration in development Science 284,770-776[Abstract/Free Full Text]
- Rebay, I., Fleming, R. J., Fehon, R. G., Cherbas, L., Cherbas, P., Artavanis-Tsakonas, S. (1991) Specific EGF repeats of Notch mediate interactions with Delta and Serrate: implications for Notch as a multifunctional receptor Cell 67,687-699[CrossRef][Medline]
- Kao, H.Y., Ordentlich, P., Koyano-Nakagawa, N., Tang, Z., Downes, M., Kintner, C. R., Evans, R. M., Kadesch, T. (1998) A histone deacetylase corepressor complex regulates the Notch signal transduction pathway Genes Dev. 12,2269-2277[Abstract/Free Full Text]
- Fisher, A. L., Caudy, M. (1998) Groucho proteins: transcriptional corepressors for specific subsets of DNA-binding transcription factors in vertebrates and invertebrates Genes Dev. 12,1931-1940[Free Full Text]
- Chitnis, A. B. (1995) The role of Notch in lateral inhibition and cell fate specification Mol. Cell. Neurosci. 6,311-321[CrossRef][Medline]
- Greenwald, I. (1998) LIN-12/Notch signaling: lessons from worms and flies Genes Dev. 12,1751-1762[Free Full Text]
- Allman, D., Punt, J. A., Izon, D. J., Aster, J. C., Pear, W. S. (2002) An invitation to T and more: notch signaling in lymphopoiesis Cell 109(Suppl.),S1-S11
- Irvine, K. D., Rauskolb, C. (2001) Boundaries in development: formation and function Annu. Rev. Cell Dev. Biol. 17,189-214[CrossRef][Medline]
- Berry, L., Westlund, B., Schedl, T. (1997) Germ-line tumor formation caused by activation of glp-1, a Caenorhabditis elegans member of the Notch family of receptors Development 124,925-936[Abstract]
- Johnston, L., Edgar, B. (1998) Wingless and Notch regulate cell cycle arrest in the developing Drosophila wing Nature 394,82-84[CrossRef][Medline]
- Miller, D. T., Cagan, R. L. (1998) Local induction of patterning and programmed cell death in the developing Drosophila retina De
TOP
ABSTRACT
INTRODUCTION