Key points

Introduction

Autoinflammatory diseases (AID) are a group of disorders characterized by seemingly unprovoked inflammation that may be recurrent or sometimes nearly continuous. The term, autoinflammatory , first appeared in the literature in 1999 to describe 2 monogenic disorders with recurrent fevers and episodes of systemic inflammation without high-titer autoantibodies or antigen-specific T cells: familial Mediterranean fever (FMF) and the then newly described TNF receptor–associated periodic syndrome. At present more than 20 monogenic AID have been reported. The clinical manifestations of AID are typically driven by genetically determined dysregulation of innate immunity, which results in overproduction of inflammatory cytokines, such as interleukin (IL)-1β, IL-6, IL-18, tumor necrosis factor (TNF), and type I interferon (IFN). Specific treatments targeting these cytokine signaling pathways have proved effective in many AID patients, highlighting the importance of accurate genetic diagnosis and detailed molecular pathophysiology.

This article reviews some of the recent advances in the field of AID over the past 3 years, including the discovery of several newly identified monogenic disorders ( Table 1 ). It also focuses on recent insights into the pathogenesis of FMF to demonstrate how genetics and basic biology have synergized to demystify one important mechanism of host-pathogen interaction.

Introduction

Autoinflammatory diseases (AID) are a group of disorders characterized by seemingly unprovoked inflammation that may be recurrent or sometimes nearly continuous. The term, autoinflammatory , first appeared in the literature in 1999 to describe 2 monogenic disorders with recurrent fevers and episodes of systemic inflammation without high-titer autoantibodies or antigen-specific T cells: familial Mediterranean fever (FMF) and the then newly described TNF receptor–associated periodic syndrome. At present more than 20 monogenic AID have been reported. The clinical manifestations of AID are typically driven by genetically determined dysregulation of innate immunity, which results in overproduction of inflammatory cytokines, such as interleukin (IL)-1β, IL-6, IL-18, tumor necrosis factor (TNF), and type I interferon (IFN). Specific treatments targeting these cytokine signaling pathways have proved effective in many AID patients, highlighting the importance of accurate genetic diagnosis and detailed molecular pathophysiology.

This article reviews some of the recent advances in the field of AID over the past 3 years, including the discovery of several newly identified monogenic disorders ( Table 1 ). It also focuses on recent insights into the pathogenesis of FMF to demonstrate how genetics and basic biology have synergized to demystify one important mechanism of host-pathogen interaction.

The deubiquitinase deficiencies

NF-κB denotes a group of transcription factors that regulate the expression of genes involved in the cell cycle, immune response, differentiation, and DNA repair. This signaling pathway is in part regulated by ubiquitination, a protein post-transcriptional modification process. The DUBs are a group of enzymes that specifically remove ubiquitin (Ub) moieties from target proteins, and their dysregulation has been reported to result in various human diseases. Several DUBs, including A20, CYLD, OTULIN, and OTUD7B (Cezanne), act as negative regulators of NF-κB signaling. Prior to 2016, CYLD was the only DUB for which germline mutations had been implicated in a Mendelian human disease.

Haploinsufficiency of A20

A20 is a DUB that plays a key inhibitory role in the NF-κB proinflammatory pathway. The inhibitory function of A20 is coordinately effected by its N-terminal ovarian tumor (OTU) domain-mediated DUB activity and by its C-terminal zinc finger-mediated E3 Ub ligase activity. Thus, A20 removes lysine 63 (K63)-linked Ub chains from proinflammatory signaling complexes, leading to their disassembly, and then conjugates the constituent proteins with lysine 48 (K48)-linked Ub chains, marking them for proteasomal degradation. Hence, the net effect of A20 is anti-inflammatory and a deficiency of A20 is predicted to cause unchecked inflammation.

In 2016, Zhou and colleagues reported 6 families with dominantly inherited truncating mutations in the TNFAIP3 gene, which encodes A20. Clinical manifestations included early-onset fevers, arthralgia, oral and genital ulcers, and ocular inflammation, in some cases resembling Behçet disease. Five of the mutations were in the OTU domain, whereas 1 was in a zinc finger domain. Mutant A20 demonstrated no inhibitory effect on the NF-κB pathway, whereas a mixture of wild-type and mutant A20 had substantial inhibitory activity, suggesting that the mutant proteins are likely to act through haploinsufficiency rather than a dominant-negative effect. In vitro reconstitution experiments showed accumulation of K63-Ub on RIPK1, one of the A20 substrates, an effect that was also confirmed in patients’ cells. Patients’ peripheral blood mononuclear cells and fibroblasts also demonstrated strong phosphorylation of IκBα, IKKα/β, and p38 with and without TNF stimulation, consistent with constitutive NF-κB activity. Spontaneous NLRP3 inflammasome activation leading to IL-1β release was observed in peripheral blood mononuclear cells, and 1 of the patients showed a good clinical response to IL-1β inhibition, consistent with previous reports suggesting the role of A20 as a negative regulator of the NLRP3 inflammasome in mice. Zhou and colleagues dubbed this novel Mendelian disease, HA20 ( Fig. 1 A ).

Proposed mechanisms of pathogenesis in DUB deficiencies. ( A ) After binding TNF, a signaling complex is recruited to the TNF receptor. The addition of K63-linked Ub chains to RIPK1 stabilizes its signaling complex and leads to phosphorylation of the IKK complex, degradation of IκBα, and translocation of the NF-κB heterodimer into the nucleus. Under normal conditions ( left ), A20 removes K63-linked Ub chains from the RIPK1 complex and instead conjugates its constituent proteins with K48-linked Ub chains to mark them for proteasomal degradation. In HA20, heterozygous loss of function of A20 results in the impaired suppressive function of A20, leading to excessive activation of NF-κB signaling ( right ). ( B ) LUBAC conjugates linear Ub to targets, including NEMO and RIPK1, which potentiate NF-κB signaling. OTULIN restricts this signaling by hydrolyzing linear Ub on LUBAC and its target proteins ( left ). In otulipenia/OTULIN-related autoinflammatory syndrome (ORAS), biallelic loss of function of OTULIN results in loss of this suppressive function, leading to accentuated activation of NF-κB signaling ( right ). For the sake of simplicity, ubiquitination on TRAF2 and TNFR1 are not shown in these figures.

Recent studies suggest an essential role of A20 in the development and appropriate regulation of immune cells, and its dysregulation has been linked to various human diseases. Common nucleotide variants in TNFAIP3 have been associated with multiple autoimmune diseases, including systemic lupus erythematosus (SLE), type I diabetes, inflammatory bowel disease, ankylosing arthritis, Sjögren syndrome, and rheumatoid arthritis. Furthermore, somatic loss-of-function mutations in A20 have been described in B-cell lymphoma, which suggests its role as a tumor-suppressor gene. Complete loss of A20 in mice ( Tnfaip3 ^−/− ) resulted in early lethality due to persistent NF-κB activation and severe multiorgan inflammation, whereas immune cell–specific ablation resulted in autoimmunity, such as SLE. One of the reported HA20 cases initially carried the diagnosis of SLE, and recently a new case of HA20 from Japan was reported to have autoimmune lymphoproliferative syndrome. Together, these reports underscore the importance of A20 in immune regulation, and suggest that further study is needed to define the full clinical spectrum of HA20.

Deficiency of adenosine deaminase 2

In 2014 Zhou and colleagues and Navon Elkan and colleagues identified biallelic loss-of-function CECR1 mutations in patients presenting with fevers and early-onset strokes and/or with vasculitis resembling PAN, a systemic necrotizing vasculitis typically affecting medium-sized muscular arteries. Zhou and her colleagues reported 9 patients with fevers, early-onset (<5 year old) lacunar strokes, livedoid rash, hepatosplenomegaly, cytopenia, and systemic vasculopathy, including 2 patients with PAN and 1 with small-vessel vasculitis; 8 of the 9 patients had histories of lacunar strokes mainly affecting the deep-brain nuclei and the brain stem. Several strokes were hemorrhagic or underwent hemorrhagic transformation, leading to long-term disability. Most of the strokes occurred during episodes of systemic inflammation. These patients also presented with various sequelae of systemic vascular disease, including livedo racemosa, myositis, portal hypertension, and ophthalmologic complications. Biopsy samples from skin, liver, and brain exhibited vasculopathic changes, including impaired endothelial integrity, endothelial cellular activation, and inflammation. Four patients had hypogammaglobulinemia, and 2 of them had multiple episodes of bacterial and viral infections.

Simultaneously and independently, Navon Elkan and colleagues reported 24 patients with PAN with biallelic CECR1 mutations, 19 of whom were of Georgian Jewish ancestry. Among them, 18 presented with childhood-onset PAN (<10 years old), including 6 who received the diagnosis during infancy (<1 year old). Most of the patients had cutaneous involvement, most commonly livedo racemosa. Five patients had episodes of either strokes or intracranial hemorrhage, whereas 10 had signs of peripheral neuropathy. Aneurysm formation in visceral arteries was observed in 6 patients, with associated renal hypertension and gastrointestinal manifestations.

CECR1 encodes ADA2, which can convert adenosine to inosine and 2′-deoxyadenosine to 2′-deoxyinosine. Although ADA2 has partial structural homology with ADA (ADA1), the deficiency of which causes human severe combined immunodeficiency (SCID) through the intracellular accumulation of toxic nucleotides, these 2 enzymes differ in many aspects. Whereas ADA1 acts as a monomer and is primarily localized intracellularly, ADA2 acts as a dimer and is secreted into the extracellular space. The patients described in both reports had a marked reduction of ADA2 protein concentrations and ADA2-specific enzymatic activity in the blood, suggesting that the detected mutations were loss of function. In the structural analysis, the missense mutations were predicted to affect the catalytic and dimerization domains or protein stability. Knock-down of cecr1b in zebrafish embryos resulted in intracranial hemorrhages and neutropenia, which were rescued by coinjection with human wild-type CECR1 (but not mutant CECR1 ), establishing the pathogenicity of the mutations. Zhou and colleagues, therefore, proposed the term, DADA2 , to denote the human disease.

ADA2 is expressed in the myeloid lineage and, once secreted, it induces differentiation of monocytes into macrophages, possibly by binding proteoglycan-like structures on the cellular surface. DADA2 patient monocytes showed impaired differentiation toward the anti-inflammatory (M2) macrophage population under standard culture conditions, thus leading to polarized differentiation toward the proinflammatory (M1) macrophage subset. A recent transcriptome-wide analysis using DADA2 patients’ blood samples displayed a strong up-regulation of neutrophil-related genes as well as a moderate IFN signature, and the investigators also reported the accumulation of myeloperoxidase in patients’ polymorphonuclear cells. Zhou and colleagues further reported that the DADA2 patients’ brain and skin samples showed substantial endothelial activation and damage, and up-regulation of inflammatory cytokines. Coculture of patients’ monocytes with a human primary endothelial cell layer led to considerable disruption of its integrity. These studies suggest that the deficiency of ADA2 results in vascular damage at least in part mediated by skewed monocytic differentiation and neutrophil activation.

All the Georgian Jewish patients in the report from Navon Elkan and colleagues were homozygous for a mutation encoding a p.Gly47Arg substitution. The carrier frequency of this mutation in the endogamous Georgian Jewish population was 0.102, which is consistent with the apparently high prevalence of this disease in this population. Conserved haplotypes were detected around several missense mutations, including p.Gly47Arg, suggesting the existence of a possible founder effect. A heterozygous p.Tyr453Cys CECR1 mutation was identified in 2 siblings with late-onset lacunar strokes in the Siblings with Ischemic Stroke Study. Although no genome-wide association study loci for stroke have been identified in the CECR1 gene region, the effect of this gene on non-Mendelian cases of stroke and other vascular diseases should be further pursued.

Recent reports have broadened the clinical spectrum of DADA2 beyond the typical clinical picture of systemic inflammation presented in the first articles. CECR1 mutations have been demonstrated in patients with autoimmunity, lymphoproliferation, and a combined immunodeficiency as well as patients presenting primarily with common variable immunodeficiency. The lymphoproliferative picture is shared by a mutation-positive patient who was diagnosed with Castleman disease and responded to anti–IL-6 treatment. Biallelic CECR1 mutations have also been found in patients presenting with anemia, thrombocytopenia, and splenomegaly, leading to the initial clinical diagnosis of Diamond-Blackfan anemia or storage disease. Recently, a biallelic 770-kb deletion of chromosome 22q11.1 encompassing both CECR1 and IL17RA (encoding the IL-17 receptor A) was reported in 2 siblings with a history of both mucocutaneous infection and early-onset systemic vasculitis. This finding, taken together with the 28-kb deletion in the CECR1 locus reported by Zhou and colleagues, indicates the importance of structural genomic analysis in the genetic diagnosis of DADA2.

Given the specter of stroke, vasculitis, and the other possible manifestations of DADA2, treatment strategies have been the subject of intense investigation. In many cases, corticosteroids, methotrexate, cyclophosphamide, azathioprine, and IL-1 inhibitors have been ineffective. Navon Elkan and colleagues reported the use of anti-TNF agents in 10 DADA2 vasculitis patients, among which 8 demonstrated a complete response. This strategy has been further supported by a recent report from Ombrello and colleagues, demonstrating that anti-TNF treatment has completely prevented the recurrence of strokes in 15 DADA2 patients with a previous history of stroke. Hematopoietic stem cell transplantation (HSCT) is another possible therapeutic strategy, because the pathogenesis of DADA2 seems to derive mainly from the myeloid cell lineage. The effectiveness of HSCT was initially reported by Van Eyck and colleagues, followed by several articles in which HSCT normalized the plasma level of ADA2 and suppressed disease manifestations. It will be important to establish the risk-benefit ratios for anti-TNF, HSCT, and other potential therapies in various clinical settings across the widening spectrum of DADA2.

Lastly, although the zebrafish data of Zhou and colleagues clearly establish the importance of CECR1 in vascular and myeloid development, the field has been hampered by the lack of an obvious murine CECR1 orthologue that would permit the development of a mouse model. Whether through the development of alternative animal models or through the more detailed study of leukocyte and endothelial biology in DADA2 patients, there is still much to be learned about the basic biology of this disease.