Vitiligo is a progressive multifactorial skin disorder in which multifocal loss of pigmentation results from abnormal function of cutaneous melanocytes. From a clinical point of view, two major types of vitiligo are recognized: segmental and non-segmental (Yaghoobi et al., 2011). Segmental vitiligo is the rarer of these disorders, usually having an early onset and exhibiting a stable course. Non-segmental vitiligo is frequently associated with autoimmune diseases and occurs at sites of pressure or friction. Although the pathogenesis of the disorder remains unclear, recent genome-wide association studies (GWAS) revealed a highly significant association of familial cases of generalized vitiligo with polymorphic variants of the gene encoding NLRP1 (also known as NALP1) (Jin et al., 2007).
Some familial forms of the dermatological condition vitiligo have recently been linked to polymorphisms in the innate immunity gene, NLRP1. Here, we review what is currently known about the mechanisms that regulate activation of the NLRP1 protein and the downstream effects of NLRP1 on pathways impacting inflammation and apoptosis. How polymorphic variants of the NLRP1 gene contribute to the pathogenesis of vitiligo remains mysterious, requiring further investigation.
NLR1 is a member of the NLR family
NLRP1 is a member of the NLR family of proteins. These proteins derive their names from the observation that they uniformly possess a nucleotide-binding fold known as the NACHT domain (Koonin and Aravind, 2000), variable numbers of leucine-rich repeat (LRR) domains (NACHT + LRR = NLR) and also because of the founding member of the family, NOD1 (NOD-like receptors). NLRs are important components of the innate immune system, operating as intracellular sensors for pathogen recognition and ‘danger’ signals generated during tissue injury and cell stress. NLRs are highly reminiscent of the pathogen resistance genes (R-genes) of plants and are thought to represent the intracellular analogs of the Toll-like receptors (TLRs), important mediators of innate immunity in animals. NLRs, TLRs and plant R-gene products have in common that they contain LRRs, which are required for sensing pathogen-associated molecular patterns (PAMPs) (Martinon and Tschopp, 2004; Martinon and Tschopp, 2005; Stehlik and Reed, 2004). In TLRs, the binding of PAMPs to the LRRs of TLRs is thought to promote conformational changes that result in receptor activation, such that the intracellular TIR domains of these transmembrane receptors dimerize and recruit TIR-containing adapter proteins that link TLRs to downstream signaling proteins, typically protein kinases [reviewed in Medzhitov (2001)]. By comparison, the R-gene products of plants typically contain an N-terminal TIR or coiled-coil domain, which presumably links them to downstream effectors, a central nucleotide-binding domain (NB-ARC) and finally LRRs, which presumably sense pathogens [reviewed in (Cannon et al., 2002; Dangl and Jones, 2001)]. In NLRs, the association of PAMPs with LRRs triggers a conformational change that allows these proteins to bind ATP and oligomerize via their NACHT domains, creating a platform on which activation of caspase family proteases, kinases or other effector proteins occurs (Faustin et al., 2007; Martinon and Tschopp, 2005; Meylan et al., 2006).
In total, humans have 22 genes encoding proteins that possess the combination of NACHT and LRR (excluding three pseudogenes) (Martinon and Tschopp, 2004, 2005; Meylan et al., 2006; Pawlowski et al., 2001; Ting et al., 2006; Tschopp et al., 2003). Of the 22 human NLRs, 14 encode proteins that combine the NACHT and LRRs with a protein interaction domain called the PYRIN domain (PYD), while seven encoded proteins that contain a Caspase Recruitment Domain (CARD). The CARDs of some NLRs form heterotypic dimers with the CARD domains found in several pro-inflammatory caspase family proteases, such as caspase-1, caspase-4, and caspase-5 in humans (caspase-1 and caspase-11 in mice). These caspases activate by proteolytic processing cytokines interleukin-1beta (IL-1β) and IL-18, as well as causing a form of apoptosis (pyrotposis) in some situations (Fernandes-Alnemri et al., 2007; Miao et al., 2011). The PYRIN domains (PYDs) of certain NLRs indirectly interact with CARD-carrying caspases through a bipartite adapter protein, ASC, which contains both PYRIN and CARD domains. ASC thus serves as a critical linker of these two protein domain families.
Of mice and men
Mice contain 33 NLR family genes (Reed et al., 2003), where the variation relative to humans is attributable mostly to amplification of certain gene loci. Although several NLRs in the mouse have clear orthologs in humans, extensive genetic variation suggests that several NLRs are paralogs rather than homologs. In mice, two examples have already been found where strain-specific variations in NLR gene clusters are associated with differential sensitivity to pathogens – namely NLR family member Naip (BIRC1; CLR5.1) and Nlrp1 (NALP1;DEFCAP; NAC; CARD7; CLR17.1). For example, the Naip gene locus in mice consists of a tandem array of seven homologous genes or pseudogenes and controls sensitivity to Legionella (Diez et al., 2000). Macrophages in which a particular version of the Naip protein is expressed are killed by Legionella, through a caspase-dependent cell death mechanism (Ren et al., 2006). In contrast, Legionella can replicate in macrophages from non-susceptible strains without inducing apoptosis of the host cells. Similarly, in mice, strain-specific variations in a cluster of three genes homologous to human NALP1 determine sensitivity to anthrax toxin. Anthrax toxin consists of three proteins, including a zinc metalloproteinase called lethal factor (LF), which is released into the cytosol of target cells (Pannifer et al., 2001). Induction of macrophage apoptosis by LF is dependent on expression of Nlrp1b (Nalp1b) and caspase-1 (Boyden and Dietrich, 2006). Thus, mouse strains such as C57Bl/6 and SJL/J that lack Nlrp1b expression are resistant to anthrax toxin, while mouse strains such as 129S1 that express Nlrp1b are sensitive.
These examples of genetic variability in NLRs dictating host susceptibility to bacteria or their toxins illustrate a mechanism by which microorganisms exploit host cell machinery to subvert host defense mechanisms. In the case of both Naip and Nlrp1, the bacteria appear to trigger activation of the NLR family proteins, causing pathological levels of caspase-1 activity that result in macrophage apoptosis. By killing host macrophages, the bacteria are able to gain the upper hand and propagate in vivo to lethal levels.
NLRP1 is unique among NLRs
Human NLRP1 (NALP1) has a unique combination of domains compared to other members of the NLR family (Figure 1). The NLRP1 protein contains both PRYIN and CARD domains located on opposite ends of the protein, as well as an internal pair ZU5 and UPA domains (previously termed the FIIND domain) that confer intra-proteolytic activity, causing cleavage of the NLRP1 protein (D’osualdo et al., 2011).
The recently identified polymorphisms associated with vitiligo reside either between the N-terminal PYRIN and NACHT domain (L155H) or in the promoter region of the gene. Interestingly, the mouse paralog/ortholog of NLRP1 lacks the N-terminal PYD domain, illustrating the critical differences between mouse and human. To date, no functional analysis of the polymorphic variants of NLRP1 has been reported. Given its location in the NLRP1 protein, the missense mutation at L155H potentially could impact oligomerization of the protein, although other explanations are also possible. Polymorphisms in the promoter of the human NLRP1 gene presumably affect expression, but whether the result is increased versus decreased levels of NLRP1 protein is unknown.
The only PAMP thus far identified for NLRP1 is the bacterial peptidoglycan component muramyl dipeptide (MDP). In vitro, using purified NLRP1 protein, MDP induces a conformational change in NLRP1 that renders the protein competent to bind ATP. Nucleotide triphosphate then stimulates NLRP1 oligomerization, forming a donut-like oligomer (Faustin et al., 2007). The oligomerized NLRP1 scaffold provides a platform on which recruitment of caspases results in their activation by promoting dimerization to achieve active conformations of these proteases (Boatright and Salvesen, 2003).
The primary function of human NLRP1 is caspase activation. Upon activation by appropriate PAMPs, resulting in its oligomerization, human NLRP1 binds adapter protein ASC via its N-terminal PYD. ASC, in turn, binds pro-caspase-1, forming a multiprotein complex termed an ‘inflammasome’ (Martinon et al., 2002). The C-terminal CARD of NLRP1 can also bind both pro-caspase-1 and caspase-5 (Tschopp et al., 2003) (Figure 2). These members of the caspase family activate IL-1β and IL-18, as well as causing apoptosis, as highlighted previously.
NLRP1 may have other functions besides activation of caspase-1 in response to bacterial PAMP, MDP. For example, NLRP1 was reported to bind the patched receptor complex via adapter protein DRAL, stimulating caspase-9 activation (Mille et al., 2009). In this regard, the CARD of NLRP1 binds the CARD of Apaf-1, a caspase-9 activating protein that becomes activated by cytochrome c when released from mitochondria (Bruey et al., 2007; Chu et al., 2001) (Figure 3). Thus, it is possible that NLRP1 contributes to apoptosis via caspase-9 activation in some contexts. Also, it is possible that Apaf1 and components of patched receptor complex are intrinsic activators of NLRP1. Cytochrome c-mediated oligomerization of Apaf-1 conceivably could create a platform for recruitment of NLRP1, resulting in caspase-1 activation via a mechanism whereby the N-terminal PYD domain of NLRP 1 binds ASC, which in turn binds pro-caspase-1, while the C-terminal CARD of NRLP1 binds the CARD of Apaf1. Activated patched receptors also could theoretically create an oligomerized protein complex at the plasma membrane for NRLP1 recruitment, thus promoting caspase activation.
Regulators of NLRP1
Both cellular and viral regulators of NLRP1 have been identified. The anti-apoptotic Bcl-2 family proteins Bcl-2 and Bcl-XL are reported to bind NLRP1 (Figure 3), interacting with the NACHT-LRR region, to suppress caspase-1 activation (Bruey et al., 2007). The region within Bcl-2 responsible for binding and suppressing NLRP1 activation has been mapped to a 10′mer sequence in an unstructured loop of Bcl-2, suggesting opportunities for creation of drugs the mimic this inhibitory peptide motif that presumably serves as a ligand for a regulatory site on NLRP1. The suppression of NLRP1 by Bcl-2 and Bcl-XL is reminiscent of the antagonism of caspase-activating protein CED4 by the Bcl-2 homolog CED9 in C. elegans (Metzstein et al., 1998).
The Kaposi sarcoma herpesvirus virus (KSHV), also known as human herpesvirus 8 (HHV8) encodes a protein, Orf63, that binds and suppresses NLRP1 (Figure 3) and other selected members of the NLR family (Gregory et al., 2011). This finding suggests that viruses also are capable of inhibiting NLRP1. The implication of this finding is that certain viruses presumably carry PAMPs that activate NLRP1, or they induce damage-associated molecular patterns (DAMPs) that activate NLRP1. Thus, viral protein that target NLRP1 may help to ensure success of these pathogens. Moreover, understanding the molecular details of how viral proteins suppress NLRP1 may provide insights that guide strategies for achieving pharmacological suppression of NLRP1 for autoimmune and inflammatory diseases.
The mechanisms regulating intramolecular cleavage of NLRP1 and its cellular consequences are presently unknown. Interestingly, after cleavage, the N and C-terminal fragments remain associated (D’osualdo et al., 2011), suggesting that cleavage induces a conformational change in NLRP1 rather than generating fragments with disparate functions. It remains to be determined whether autocleavage activates or inactivates NLRP1.
The discovery of hereditary polymorphisms in the gene encoding NLRP1 in patients with non-segmental vitiligo provides fresh insights into the pathogenic mechanisms of this autoimmune disorder. The functional consequences of these polymorphic variants remain to be determined. The complexity of NLRP1 regulation and its various putative effector mechanisms (inflammation, apoptosis, etc.) raise several interesting possibilities for how sequence variations in NLRP1 might impact immune system function. Further investigations are needed to elaborate mechanistic insights that might yield new therapeutic strategies for combating vitiligo and possibly other autoimmune disorders involving NLRP1 and various members of the NLR family.
We thank the NIH for generous support (AI-56324; AI-78048).