Wiskott-Aldrich Syndrome at the nexus of autoimmune and primary immunodeficiency diseases

Abstract

Wiskott-Aldrich Syndrome (WAS) is a X-linked primary immunodeficiency disorder also marked by a very high (up to 70%) incidence of autoimmunity. Wiskott-Aldrich Syndrome arises from mutations in the Wiskott-Aldrich Syndrome protein (WASp), a cytoplasmic protein that links signaling by cell surface receptors such as the T-cell receptor and integrins to actin polymerization. WASp promotes the functions of multiple cell types that support immune responses, but also is important for the function of regulatory T cells and in TCR-induced apoptosis, two negative mechanisms of immune regulation that maintain peripheral immune tolerance. Here we review the nature of immune defects and autoimmunity in WAS and WASp deficient mice and discuss how this single gene defect can simultaneously impair immune responses to pathogens and promote autoimmunity. The myriad cellular immune defects found in WAS make this Mendelian syndrome an interesting model for the study of more complex immune diseases that arise from the interplay of environmental and multiple genetic risk factors.

The Wiskott-Aldrich Syndrome (WAS) is a X-linked primary immunodeficiency disorder occurring with a frequency of between 1–10 per million males. WAS results from mutations in the gene coding for the Wiskott-Aldrich Syndrome protein (WASp), a key regulator of branched actin chain polymerization. [1]. However, the clinical manifestations of WAS are not limited to immunodeficiency, thrombocytopenia and eczema, the classically defined trio of symptoms in this syndrome. A remarkably high percentage of WAS patients also develop autoimmunity. The protean clinical manifestations of this single-gene disease likely arise from the numerous cellular processes that rely on WASp to regulate dynamic changes in the cytoskeleton in response to signal transduction by a number of cell surface receptors. Here we will review the mechanisms underlying WAS with particular attention to the pathogenesis of autoimmunity in WAS and WASp-deficient mice. Understanding the aspects of immune system function which are disrupted in a single gene disorder such as WAS can shed light on signal transduction pathways which underlie generation of effective immune responses as well as those involved in maintaining immunological tolerance.

1. Wiskott-Aldrich Syndrome and the WASp protein

Mutations in the WASP gene coding for the Wiskott-Aldrich Syndrome protein (WASp) were identified as the cause of most cases of WAS in 1994 [2]. WASp is a 502 amino acid protein with five functional domains: N-terminal Drosophila-enabled/vasodilator-stimulated phosphoprotein homology 1 (EVH1) domain, a basic region, a GTPase-binding domain (GBD), a polyproline-rich region (PRR) and a C-terminal verprolin homology/central region/acidic region (VCA) domain [3, 4]. (Figure 1) The major function of WASp is to activate the Arp2/3 complex through its VCA domain which results in the polymerization of branched actin filaments. Regulated formation of branched actin filaments underlie formation of cellular structures such as lamellopodia. This regulated formation of branched actin filaments function in cell migration, endocytosis, and vesicle trafficking. Signals from cell surface receptors activate WASp through the activating the CDC42 GTPase. This GTPase binds WASP in its activated form, and induces a conformational change in WASp that activates its actin polymerizing activity. Because CDC42 can be rapidly activated and inactivated by various cellular signaling pathways, this mechanism allows the tight control over actin polymerization/depolymerization necessary for rapid cellular responses. Signals acting through other domains of WASp can synergize with CDC42 binding to optimize WASp activity. Other modifying signals that regulate WASp activity include Phosphatidylinositol 4,5-bisphosphate (PIP2) binding to the basic domain, WASp tyrosine phosphorylation by Src family kinases, and binding of SH3-domain containing proteins to the Proline Rich Domain (PRD)[3] (Figure 1).

Figure 1. Wiskott Aldrich Syndrome Protein domains and function.

The indicated domains of WASp are shown schematically, with proteins and other molecules binding WASp shown below the schematic, and functions shown above. EVH1/WH1 domain: enabled VASP (vasodilator-stimulated protein) homology 1)/WH1 (WASP Homology 1). B: Basic Domain. GBD: GTPase Binding Domain. Proline Rich Domain (PRD). The PRD of WASp has been reported to bind numerous Src Homology 3 (SH3) domain containing proteins, including Btk, Hck, Fyn, Lyn, Itk, Nck, c-Src. Grb2, p85a, PI-3K, PLC-g, PSTPIP1, PTP-PEST, profilin and VASP [38, 39]. PSTPIP1 binding has been implicated in trafficking of WASp to the immune synapse [15], and binding of Fyn and PTP-PEST has been implicated in regulating tyrosine 291 phosphorylation and the function of WASp in T cell activation [21]. VCA domain: Verprolin, cofilin, acidic domain.

The type of mutation in the WASP gene strongly influences the clinical severity of Wiskott-Aldrich Syndrome [1, 3]. Premature termination and deletion mutations in WASp that ablate expression of Wiskott-Aldrich Syndrome protein usually result in severe, classical WAS. These patients present with neonatal thrombocytopenia and dysfunctional platelets that can be accompanied by excessive bleeding, eczema, and severe infections early in life. These complications can be fatal if not treated with bone marrow transplantation or gene therapy. Missense mutations in the WAS gene that allow WASp protein expression with reduced function usually result in phenotypically milder cases. Some of these cases are limited to thrombocytopenia, termed X-linked thrombocytopenia (XLT) that can be intermittent [1, 5, 6]. Interestingly, activating mutations in WASp can result in a distinct clinical syndrome of severe congenital x-linked neutropenia (XLN) [7].

The immune deficiencies in WAS result from effects on a remarkably broad spectrum of cell types leading to defects in both cellular and humoral immunity. IgG levels are generally normal, but depressed IgM levels while elevated levels of IgA and IgE occur. Patients with severe WAS have reduced responses to polysaccharide antigens and class switching in response to protein antigens [1]. Lymphocyte numbers tend to be normal at birth, but then decline with age with lymphopenia below 1000 cells/mm³ in about a quarter of patients [8]. Depressed T cell function as measured by Delayed-Type Hypersensitivity responses and mitogen-induced proliferation has been reported. In a number of retrospective case studies, 40–72% of WAS patients have autoimmunity. Autoimmunity associated with WAS manifested most frequently by autoimmune hemolytic anemia, but other complications such vasculitis resembling Henoch-Schönlein purpura with IgA nephropathy, uveitis and inflammatory bowel disease have also been reported [9–11]. While many thought that autoimmunity was more common in patients with complete WASp protein deficiency, recent reports show that patients with only the mild form of X-linked thrombocytopenia have autoimmunity [5].

2. Cellular Immune defects in WAS and WASp deficiency in mice

Investigation of WASp deficient mice has uncovered a large number of distinct defects in immune function in multiple cell types that are similar to those seen in WAS patients (Figure 2). WASp deficient mice have thrombocytopenia and lymphopenia similar to WAS patients, but do not develop eczema, hematopoietic malignancies or abnormal bleeding that can be seen in WAS [12, 13]. One colony of WASp deficient mice developed spontaneous colitis [12] but in other facilities, colitis in these mice is mainly seen after irradiation or other stresses, suggesting the involvement of environmental factors. WASp deficient T cells exhibit poor T cell proliferation after mitogen or antigen stimulation that could be rescued by addition of IL-2 [14, 15]. This may stem from reduced efficiency of immune synapse formation between T cells and APC, but WASp also appears to independently regulate the capacity of T cells to secrete effector cytokines [16].

Figure 2. WASp affects multiple cellular processes in immunity.

Green arrows represent processes with positive influence on immune responses and red arrows represent processes that reduce immune responses through cytotoxicity or other regulatory mechanisms. Immune processes influenced by WASp are numbered in the figure with green numbers denoting effects of WASp that enhance immune responses and red denoting effects of WASp which regulate immune responses. Specific functions of WASp are numbered as follows: 1) WASp promotes dendritic cell (DC) podosome formation and migration [40], presentation of particulate antigens [41] and T-DC contacts in the immune synapse in a DC intrinsic manner [42, 43] 2) WASp in T cells can enhance the efficiency of immune synapse formation which contributes to T cell activation [14, 15] 3) WASp also promotes cyokine secretion which also contributes to T cell activation and effector function [16] 4) WASp promotes restimulation-induced cell death and release of cytotoxic granules and FasL[24]. 5) WASp promotes B cell adhesion, migration, and germinal center formation [33] 6) WASp promotes T reg homeostasis, Treg control of effector T cell proliferation and Treg killing of B cells [25, 28, 35]. 7) WASp promotes cytotoxicity of NK cells [37] 8) WASp may also promote uptake of apoptotic blebs by macrophages [44], although we have not reproduced this defect in WASp deficient macrophages [24].

Part of the reason WASp deficiency produces such varied and often mild defects is that there may be partial redundancy with other related proteins which promote actin polymerization termed type I nucleation promoting factors (NPF), including N-WASp, WAVE, WASH, WHAMM, and JMY proteins. Unlike WASp, whose expression is restricted to hematopoietic cells, N-WASp is ubiquitously expressed in all cells and, N-WASp deficiency results in defective endocytosis, exocytosis, and hair-follicle cycling [17]. WASH mediates endosome trafficking, but functions at different times and in different vesicles than N-WASp. WAVE proteins mediate cell motility and membrane ruffling, WHAMM mediates ER-Golgi transport and Golgi organization, and JMY mediates cell motility [18]. Although N-WASp deficiency is lethal in mice, the redundancy of N-WASp and WASp in lymphocytes was investigated by creating mice which WASp, N-WASp, or both proteins were ablated specifically in T cells or in the lymphocytes of chimeras with RAG deficient blastocysts [19]. Remarkably, no defects were seen in cells deficient only in N-WASp, indicating that N-WASp may be redundant for many of functions of lymphocytes that are affected by WASp. However when B and T cells were doubly deficient in WASp and N-WASp, lymphopenia and a block in thymic development prior to the CD4⁺CD8⁺ double positive T cell stage was seen. Thymocytes from these mice displayed reduced proliferation and chemotactic responses to CCL12, a critical chemokine positioning cells within the thymus [19]. Of note, colitis in WASp/N-WASp doubly-deficient mice was more severe than in WASp deficient mice, possibly because of the lymphopenia and reduced numbers of (Treg) T cells present in the doubly deficient mice.

Expression of targeted mutations of WASp in T cells and in transgenic and knock-in mice have helped to define the functions of each domain of WASp and dissect their contribution to the clinical manifestations of WAS (Figure 1). Deletion of the EVH1 domain reduces the ability of WASp to bind to WIP, which binds to and stabilizes WASp [20]. Since the majority of WAS point mutations in patients are at the N-terminal region of WASp, including many within the EVH1 domain, WIP binding appears necessary in vivo to stabilize WASp expression in vivo. In mice expressing WASp missing the GBD, binding of Cdc42 is ablated as would be predicted, but T cells proliferate normally after activation, suggesting that WASp acts in T cell activation independently of the GBD [21]. Instead, a key role of phosphorylation of Tyrosine at position 291 in human (293 in mouse) in the GBD has been uncovered [21, 22].Mice expressing a mutant WASp protein which cannot be phosphorylated at this position are similar to WASp deficient animals, suggesting that phosphorylation of this tyrosine in the GBD is required for WASp activation independently of CD42 binding. Deletion of the Proline Rich Region (PRR) in WASp reduces formation of T cell:APC conjugates and actin polymerization suggesting that the PRR is important for TCR signaling [15]. Interestingly, deletion of the C terminal VCA domain in WASp results in defective thymopoiesis, which is not seen in WASp deficient mice [23], suggesting that WASp lacking this C-terminal domain important for actin polymerization acts as a dominant negative for other N-WASp and other NPF family proteins.

3. Autoimmunity in WASp deficiency: Cellular and molecular mechanisms

The study of autoimmunity in WASp deficient mice may help inform our understanding of the pathogenesis of the autoimmune manifestation in WAS patients. Other than colitis, autoimmunity in WASp deficient mice had not been studied until recently. We and others have found that older WASp deficient mice develop anti-nuclear and anti-dsDNA antibodies at much higher rates than isogenic controls with titers approaching those of other autoimmune-prone mouse strains such as MRL/lpr and NZB/NZWF1 [24, 25]. In WASp deficient mice on a 129SvEv background over six months of age, we found circulating immune complexes, immune complex deposition in the kidney, and mild nephritis resembling the IgA nephropathy seen in some WAS patients [24]. Recently, anti-platelet antibodies have been found in up to 40% of WASp deficient mice accompanied by evidence of increased platelet turnover [26].

Defective homeostasis and function of regulatory T cells (Treg) may account for some of the observed autoimmunity seen in WASp deficient mice and WAS patients. While numbers of FoxP3(+) Tregs are normal in the periphery and thymus of WASp deficient mice, in bone marrow chimera experiments in mice, WASp deficient Treg were out-competed by normal Treg, suggesting a cell-intrinsic role for WASp in Treg survival. WASp deficient Treg proliferated poorly and failed to produce TGF-β when stimulated in vitro. These defects could not be rescued by exogenous IL-2 [25, 27]. WASp deficient Treg exhibited weaker suppression of normal T cell proliferation and had reduced ability to prevent colitis induced by naïve T cells transferred into immunodeficient hosts. Treg from WAS patients display similar abnormalities in activation and function [27, 28]. Why Treg are particularly affected by WASp deficiency is not clear. Defective Treg homeostasis may stem from the dependency of Treg on Interleukin-2 for survival, which may be limiting in WASp deficiency. Functional defects in Treg may be related to known T cell activation defects, or problems with secretion of cytotoxic granules by WASp deficient Treg.

The ability of mature T cells to undergo programmed cell death when restimulated through the TCR is also disrupted by WASp deficiency and this may contribute to autoimmunity in WAS independently of defects in Treg. Restimulation of activated T cells through the TCR results in programmed cell death, mediated in large part by autocrine stimulation of the TNF-family receptor Fas by Fas Ligand (FasL) [29]. Humans and mice with genetic defects in cell death induced by restimulation with antigen (Restimulation-Induced Cell Death, or RICD), including those with defects in Fas or Fas Ligand, produce autoantibodies and suffer from autoantibody mediated disease similar to the complications of WAS [30]. Activated WASp deficient T cells underwent significantly reduced RICD compared to isogenic control cells. This occurred in T cells from young mice even before the onset of autoantibody production, suggesting that the TCR-indcued apoptosis defect is an intrinsic property of WASp deficient cells [24].

Rather than stemming from a cell intrinsic defect in Fas signaling, the mechanism of resistance to RICD in WASp deficiency appears to be related to production of Fas Ligand. WASp deficient T cells underwent normal programmed cell death when exposed to exogenous FasL, but produced significantly less FasL compared to wild-type cells. FasL, like all other TNF family cytokines, is produced initially as a type II transmembrane protein. FasL can be cleaved off the cell membrane by ADAM family metalloproteases to produce soluble FasL [31], or secreted in a high-molecular weight form believed to be derived from secretion of FasL anchored in exosome-like secretory vesicles [31]. Fas-dependent apoptosis is highly dependent on the oligomerization state of FasL, only membrane-bound and vesicular FasL can induce apoptosis, whereas soluble FasL does not, and can indeed antagonize apoptosis induced by membrane-bound FasL. Reinforcing this dichotomy, mice in which only soluble FasL are produced develop systemic autoimmunity as severe as that in FasL deficient mice [32]. Further analysis of FasL secretion in WASp deficient cells revealed normal surface FasL, but reduced amounts of secreted FasL, both in soluble and high-molecular weight form. Supernatants from restimulated WASp deficient T cells which had reduced cytotoxicity when applied to bystander Fas-sensitive tumor cells. Restimulated CD4+ T cells also released less beta-hexosaminadase, suggesting a more generalized defect in granule secretion in WASp deficient T cells restimulated CD4+ T cells [24]. Why WASp deficient T cells produce less functional FasL is less clear, WASp-dependent cytoskeletal remodeling may be necessary for efficient secretion of FasL and potentially other molecules, or a step in TCR signaling important in generating secreted vesicular FasL may depend on WASp. This protection from cell death in WASp deficient cells may occur in vivo: in irradiated mice receive bone marrow containing a mixture of WT and WASp deficient cells, WASp⁺ T and B cells accumulate in the recipients, suggesting that WASp provides a selective survival advantage to T and B cells [25]. Taken together the data on defective RICD and FasL secretion in WASp deficiency provide a striking example of how a similar cell biological defect in secretion of granule associated proteins can promote immunodeficiency, through reduced secretion of effector cytokines, and autoimmunity, through reduced secretion of negative immune modulators such as FasL. This paradigm may apply to other mutations affecting affect both positive and regulatory aspects of immunity.

Immune deficiency and autoantibody production in WASp deficiency likely has a B cell intrinsic component as well. WASp deficient mice have reduced germinal center formation and B cell migration and proliferation defects [33]. In bone-marrow chimeras, WASp deficiency impairs mature B cell survival and homeostasis in a B cell intrinsic fashion [34]. Wasp deficient mouse Treg are also defective in killing of activated B cells, suggesting that autoantibodies in WAS may arise because of failure to eliminate potentially autoreactive B cells [35]. Cell-intrinsic defects in neutrophils, dendritic cells, NK cells and other cell types in the innate immune system have also been reported in WAS patients or WASp deficient mice. [36, 37] (Figure 2). Collectively, these defects likely account for the complex immune dysfunction seen in patients with WAS and WASp deficient mice.

Perhaps because of the redundancy of WASp with other members of the NBD family of proteins regulating branched actin chain formation, the defects in WASp deficiency tend to be partial, rather than absolute, but the number of cell types affected may lead to the high penetrance of immunodeficiency and autoimmunity in all but those with the mildest molecular lesions in the WASP gene. Defects in both positive and negative regulation of immune responses in WAS lead to a situation of immune dysregulation akin to taking ones feet off both the accelerator and the brake while driving, with consequences for loss of control of immune responses. Because of the pleiotropic effects of WASp on immune function, WAS resembles complex autoimmune and immunodeficiency diseases that stem from the interplay of multiple genetic and environmental risk factors. As such, the study of WAS and WASp deficient mice is likely to continue to yield insights into the pathogenesis of immunodeficiency and autoimmunity with imlplications beyond the relatively small number of patients afflicted with this syndrome.