A novel mutation of IL1RN in the deficiency of interleukin-1 receptor antagonist syndrome: Description of two unrelated cases from Brazil




Monogenic autoinflammatory diseases are disorders of Mendelian inheritance that are characterized by mutations in genes that regulate innate immunity and whose typical features are systemic inflammation without high-titer autoantibodies or antigen-specific T cells. Skin and bone inflammation in the newborn period have been described in 3 of these autoinflammatory disorders: neonatal-onset multisystem inflammatory disease, Majeed syndrome, and deficiency of interleukin-1 (IL-1) receptor antagonist (DIRA) syndrome. This study was undertaken to present the characteristics of the DIRA syndrome in 2 cases from Brazil, and describe a novel mutation in IL1RN.


Two unrelated Brazilian patients were evaluated for the clinical signs and symptoms of these 3 disorders, and peripheral blood samples were assessed for mutations in NLRP3, LPIN2, and IL1RN by DNA resequencing analysis. A mutation in IL1RN that encodes a mutant protein was identified, and the expression and function of this mutant protein were compared to those of the wild-type protein.


Both patients presented with pustular dermatitis resembling generalized pustular psoriasis, recurrent multifocal aseptic osteomyelitis, and elevation in the levels of acute-phase reactants, all of which are features most consistent with the DIRA syndrome. Chronic lung disease was observed in 1 of the patients, and jugular venous thrombosis was observed in the other patient. Both patients showed a partial response to corticosteroid therapy, and 1 patient experienced an initial improvement of dermatitis with the use of acitretin. Both patients were homozygous for a novel 15-bp (in-frame) deletion on the IL1RN gene. The mutated protein expressed in vitro had no affinity with the IL-1 receptor, and stimulation of the patients' cells with recombinant human IL-1α or IL-1β led to oversecretion of proinflammatory cytokines, similar to the findings obtained in previously reported patients.


The presence of the same homozygous novel mutation in IL1RN in 2 unrelated Brazilian patients suggests that this genetic variant may be a founder mutation that has been introduced in the Brazilian population.

Autoinflammatory diseases are a group of disorders characterized by systemic inflammation without high-titer autoantibodies or antigen-specific T cells, and their etiology is thought to be attributable to dysregulation of innate immunity (1, 2). A number of the autoinflammatory diseases are single-gene disorders that are clinically characterized by features of recurrent or persistent systemic inflammation, such as fever and elevation in the levels of acute-phase reactants, and organ-specific manifestations, such as rashes and osteoarticular, serosal, neurologic, and ocular manifestations (1–3). The identification of these monogenic disorders in innate immune pathways has led to a better understanding of the key inflammation pathways, particularly the role of interleukin-1 (IL-1) in a number of these disorders (1).

Thus far, studies have identified 3 autoinflammatory disorders that manifest predominantly with skin and bone inflammation in the newborn period, including neonatal-onset multisystem inflammatory disease (NOMID; also called chronic infantile neurologic, cutaneous, articular syndrome) (4–7), Majeed syndrome (8–11), and the recently described deficiency of interleukin-1 receptor antagonist (DIRA) syndrome (12, 13). In addition to inflammation of the skin and bone, NOMID also has features of neurologic impairment and papilledema, chronic arthropathy, and persistent fever, and its etiology has been linked to mutations in NLRP3 (previously called CIAS1) (5, 6). In Majeed syndrome, the clinical features include microcytic congenital dyserythropoietic anemia, and a causal mutation has been identified in LPIN2 (8). The DIRA syndrome is also characterized by perinatal-onset pustular dermatitis, multifocal aseptic osteomyelitis, periosteitis, leukocytosis, and marked elevation in the levels of acute-phase reactants, and its etiology has been linked to a loss-of-function protein that truncates mutations in IL1RN (12), a gene that encodes the IL-1 receptor antagonist (IL-1Ra), or a major deletion that involves the IL1RN locus (12, 13). One of the nonsense mutations of IL1RN (identified in patients from The Netherlands), a frameshift mutation attributable to a 2-bp deletion (in patients from Newfoundland), and a genomic 175-kb deletion (in patients from Puerto Rico) are believed to be founder mutations in their respective populations (12).

In the present study, we describe 2 unrelated Brazilian patients whose clinical phenotype was consistent with the DIRA syndrome. Both patients presented with a severe but distinct clinical picture, mainly involving the skin, bones, and lungs, and both were homozygous for the same 15-bp (in-frame) deletion on IL1RN (a mutation not previously described). We hypothesize that this variant is likely to be a possible founder mutation in the Brazilian population. This novel mutation of IL1RN produces a protein that does not bind the IL-1 receptor, and thus lacks functional activity.


Genetic analysis.

DNA was extracted from the whole blood using standard procedures. In blood samples obtained from the 2 patients and their parents, all exons, splice sites, and flanking sequences of LPIN2 (GenBank accession no. NM_014646) and IL1RN (isoform 1, GenBank accession no. NM_173842) were amplified using standard polymerase chain reaction (PCR) methodology, and were then resequenced in both directions using dye-terminator chemistry and an Applied Biosystems 3730 automated sequencer. The variants identified were genotyped in 100 ethnically matched control subjects, as well as in 100 ethnically unmatched controls, by PCR, agarose gel electrophoresis, and fragment analysis. The genotypes were confirmed by resequencing in the forward direction in a randomly chosen sample constituting 20% of controls.

Functional analyses.

Quantitative real-time reverse transcription–PCR (RT-PCR).

RNA was prepared from the patients' peripheral blood, collected in PAXgene tubes (Qiagen), and converted to complementary DNA using Superscript II (Invitrogen). Quantitative real-time RT-PCR was performed to assess the expression of messenger RNA (mRNA) for the IL-1Ra gene, using TaqMan probes for IL1RN (Hs00277299_m1; Applied Biosystems) in comparison with 18S as an endogenous control.

Western blotting.

Western blot analysis of protein expression was performed using 150 mg of total protein obtained from the patients' peripheral blood mononuclear cells, in Ficoll solution, and analyzed using a polyclonal goat anti-human antibody directed against the IL-1Ra N-terminus (N-19; Santa Cruz Biotechnology). In addition, 0.1 mg of recombinant IL-1Ra was loaded as a positive control.

Expression/purification of untagged and C-terminal FLAG-(His)6–tagged human IL-1RaΔ72–76.

Both wild-type human IL-1Ra (huIL-1RaWT) and human IL-1Ra lacking residues 72–76 (huIL-1RaΔ72–76) were subcloned into an Escherichia coli expression vector as N-terminal small ubiquitin-like modifier (SUMO) fusions. The SUMO fusion tag contains a polyhistidine tag at the N-terminus to facilitate purification via nickel–nitrilotriacetic acid (Ni-NTA) immobilized metal-ion affinity chromatography (IMAC). Following purification, the SUMO tag may be cleanly excised from IL-1Ra by digestion with SUMO protease I.

Since the huIL-1RaΔ72–76 is expressed at lower levels than the wild-type form and also is less soluble and does not bind Ni-NTA efficiently, we decided to attempt to express and purify C-terminal FLAG-(His)6–tagged versions of huIL-1RaWT and huIL-1RaΔ72–76. Both huIL-1RaWT and huIL-1RaΔ72–76 were subcloned into an E coli expression vector as C-terminal FLAG-(His)6 fusions, to produce huIL-1RaWT-FpH and huIL-1RaΔ72–76-FpH. The 2 constructs were introduced into E coli DH10, and 5 hours after induction, cell pellets were collected from 400-ml cultures, frozen overnight at −80°C, and lysed in 30 ml EMD BugBuster Master Mix lysis buffer. Cleared lysates were run over 1 ml Qiagen IMAC columns and washed 4 times with 5 ml of 50 mM sodium phosphate (pH 8.0)/300 mM NaCl/10–40 mM imidazole. Both protein forms were eluted from the washed columns 3 times with 50 mM sodium phosphate (pH 8.0)/300 mM NaCl/250 mM imidazole, and the 3 × 1 ml–elution fractions were collected and pooled. The purification efficiency of each elution was assessed by running samples of starting lysate, cleared lysate, Ni-NTA flow-through, wash fractions, and pooled eluate on 4–12% Bis-Tris-MES NuPAGE gels, followed by staining with Coomassie blue.

Quantitation of eluted human IL-1RaΔ72–76-FpH.

Varying amounts of huIL-1RaWT-FpH and huIL-1RaΔ72–76-FpH were run on a 4–12% Bis-Tris NuPAGE gel along with varying amounts of a previously quantitated preparation of huIL-1RaWT-FpH. The gel was blotted to nitrocellulose and probed with a goat anti-human IL-1Ra polyclonal antibody, with detection via LiCor analysis. The band intensities of each construct were measured, and the concentrations of huIL-1RaWT-FpH and huIL-1RaΔ72–76-FpH were interpolated by comparison to the standards. To assess whether the contaminating protein in the huIL-1RaΔ72–76-FpH preparation interferes nonspecifically with downstream assays, we included huIL-1RaWT-FpH diluted to 52 μg/ml, using the huIL-1RaΔ72–76-FpH preparation as diluent.

Assay for human IL-1RaΔ72–76-FpH antagonistic activity.

IL-1Ra activity was assayed by measurement of IL-1β–induced phosphorylated JNK (phospho-JNK) nuclear translocation in KB epithelial cells (ATCC no. CCL-17). IL-1Ra at various concentrations was added to the cells for 15 minutes prior to stimulation with 10 ng/ml of IL-1β for an additional 30 minutes. The cells were fixed, permeabilized, and stained with a rabbit anti-human phospho-JNK antibody (8404002; Thermo Fisher Scientific) followed by a goat anti-rabbit IgG conjugated to Alexa 488 (A-11008; Invitrogen). The nuclei were counterstained using Hoechst 33258 (H-3569; Invitrogen). Fluorescence intensity values from the nuclear and cytoplasmic areas were measured using a Cellomics ArrayScan VTI. The mean difference in fluorescence intensity, after sampling 100 cells per well in each of 4 wells, was determined as the nuclear fluorescence intensity − the cytoplasmic fluorescence intensity; data were normalized relative to the results of control experiments using unstimulated and IL-1β–stimulated KB cells in the absence of IL-1Ra.

BIAcore analysis of human IL-1RaΔ72–76-FpH binding to human IL-1 receptor type I (IL-1RI).

The binding affinity of the same samples to immobilized IL-1RI was assessed via BIAcore analysis. Briefly, Biosensor analysis was conducted at 25°C in an HBS-EP buffer system (HEPES buffered saline with 3 mM EDTA and 0.005% Surfactant P20) using a BIAcore 3000 optical biosensor equipped with a CM5 sensor chip (GE Healthcare). All reagents were kept at 4°C prior to injection. Goat anti-human IgG (Jackson ImmunoResearch) was immobilized (∼3,500 resonance units [RU]) over flow cells 1 and 2 via standard amine coupling and followed with ethanolamine blocking and IL-1RI. Fc was captured (∼125 RU) to flow cell 2. Flow cell 1 was used as the reference flow cell. The association rate (over 3 minutes) and dissociation rate (over 10 minutes) were monitored at a flow rate of 50 μl/minute. A separate dissociation experiment of longer duration (90 minutes) was performed to determine an accurate dissociation constant (kd) value. The data were fit to a 1:1 binding model (global Rmax) via Scrubber2* software (available from the University of Utah Center for Biomolecular Interaction Analysis Web site at http://www.cores.utah.edu/interaction/scrubber.html). The kd value determined from the long-term dissociation experiment was used as a constant to fit the short-term dissociation data.

Leukocyte stimulation assay.

Cells from heparinized whole blood were washed with phosphate buffered saline (PBS), and the leukocytes were cultured at 2 × 106 per ml in RPMI with 5% fetal calf serum for 3 hours at 37°C, with or without 50 ng/ml recombinant human IL-1β (rhIL-1β) or 50 ng/ml recombinant human IL-1α (rhIL-1α) (PeproTech) or 1 μg/ml ultrapure lipopolysaccharide (LPS) (Invivogen). Supernatants were collected and concentrations of cytokines (IL-1β, IL-1α, tumor necrosis factor α [TNFα], IL-6, IL-8, and macrophage inflammatory protein 1α [MIP-1α]) were measured using the Bio-Plex platform (Bio-Rad). Unpaired t-tests were used to compare group data.


Patient presentation.

Patient 1.

Patient 1, a 30-month-old girl born to healthy parents, had a birth weight of 3.34 kg and an Apgar score of 9 at both 1 minute and 5 minutes after birth. She developed disseminated pustular cutaneous lesions a few hours after birth. Laboratory investigations showed a hemoglobin (Hgb) level of 7.9 gm/dl, a white blood cell (WBC) count of 21,300 cells/mm3 with 88% neutrophils, a platelet count of 681,000 cells/mm3, an erythrocyte sedimentation rate (ESR) of 51 mm in the first hour, and a C-reactive protein (CRP) level of 54 mg/liter. Despite several courses of intravenous antibiotics, she developed 3 abscesses on her head and thorax, and skin lesions persisted, while cultures from drained material were consistently negative for bacteria, fungi, and mycobacteria.

At the age of 1 month, she presented with pain on manipulation. Her laboratory evaluation then showed mild anemia, high WBC and platelet counts, and elevated CRP levels (63.7 mg/liter). Blood, urine, and cerebrospinal fluid cultures were all negative for bacteria, fungi, and mycobacteria. Bone scintigraphy revealed increased radiotracer uptake in the left clavicle and right seventh and ninth ribs, a finding suggestive of osteomyelitis. Bone drainage was performed and cultures for bacteria, fungi, and mycobacteria remained negative. Extensive testing for infectious diseases and primary immunodeficiency (tests for cytomegalovirus, hepatitis B and C, rubella, syphilis, toxoplasmosis, and human immunodeficiency virus, immunophenotyping of lymphocytes, dihydrorhodamine tests for neutrophil function, determination of immunoglobulin, CD40 and CD40 ligand, and IL-12 levels, and functional assays for interferon-γ) all yielded normal results.

Intravenous antibiotics were ineffective. The spontaneous occurrence of pustular rashes and abscesses and the development of a pathergy reaction (formation of purulent collections after percutaneous intravenous catheter insertion) led to the suspicion of a pyogenic autoinflammatory disorder. A repeat bone scintigraphy showed persistent inflammatory lesions on the left clavicle and the seventh and ninth ribs. The CRP level continued to be elevated (77.5 mg/liter). At that time, prednisolone was started at 2 mg/kg/day, leading to impressive improvement of the pain on manipulation, amelioration of the skin rash, and a significant decrease in the CRP level (21 mg/liter). At the age of 4 months, she was discharged from the hospital.

At the age of 5 months and while still being treated with prednisolone, the patient became febrile, with a disseminated pustular and erythematous rash (Figures 1A–C). Her laboratory evaluation showed extremely high acute-phase reactant levels and a high WBC count. Broad-spectrum intravenous antibiotics were started, resulting in a progressive improvement in the general clinical and laboratory findings. At that time, her skin biopsy showed epidermal parakeratosis and subcorneal pustules, reminiscent of pustular psoriasis, which led to the initiation of treatment with 0.2 mg/kg/day of acitretin (14), the dose of which was progressively increased to 1.0 mg/kg/day, which resulted in a remarkable improvement of the rash. The patient was hospitalized for 4 months, during which she developed a deep venous thrombosis after jugular catheterization and was treated with low molecular weight heparin. Osteomyelitis was not observed during this period. The dose of prednisolone was tapered but never discontinued, because the patient experienced 2 disease flares, and was therefore kept on a low dose of prednisolone.

Figure 1.

Skin and bone manifestations in 2 Brazilian children with the deficiency of interleukin-1 receptor antagonist syndrome. A–C, The skin manifestations in patient 1 are confluent pustular lesions evident on the head, chest, and trunk (A) and extremities (B and C). D–F, The skin manifestations in patient 2, as observed on the chest and trunk, are less severe than those seen in patient 1 (D). In addition, in patient 2, a computed tomography scan shows the collapse of vertebral bodies secondary to osteolytic vertebral lesions at the level of T1 and T2, with gibbus formation (E), and odontoid nonfusion leading to atlantoaxial subluxation (arrow) (F).

At the age of 1 year, her weight (6.5 kg) and length (63 cm) were below the third percentile, but she only had sparse pustules on her head, chest, and extremities, without secondary infection or abscesses. Osteomyelitis was absent and the acute-phase reactant levels were normal. Her motor development was delayed and she could only sit up with support.

Between 12 and 18 months of age, she presented with 2 flares of severe skin inflammation that necessitated hospitalization and an increase in the prednisolone dose. She also presented with a new osteomyelitis focus on the left femur at the age of 15 months.

The family history was unremarkable for rheumatic diseases or periodic fever syndromes. Her father has mild and episodic pustules only on his head. The parents are second cousins, originally from the northeast of Brazil, and are not aware of European ancestry. The patient's half-brother is healthy.

Patient 2.

Patient 2, a 27-month-old girl who was born after an uneventful pregnancy, had a birth weight of 3.26 kg, length of 49 cm, and Apgar scores of 9 and 10 at 1 minute and 5 minutes after birth, respectively. She developed respiratory distress within the first 24 hours of life and remained in the neonatal intensive care unit for 25 days, due to oxygen requirement. At 30 days of life, she developed diffuse papular and pustular cutaneous lesions suggestive of a pyogenic infection and was treated with antibiotics. The laboratory evaluation showed an Hgb level of 7 gm/dl, a WBC count of 24,100 cells/mm3 with 53% neutrophils, a platelet count of 853,000 cells/mm3, an ESR of 47 mm in the first hour, and a CRP level of 21 mg/liter. A skin biopsy showed mixed superficial perivascular dermatitis with neutrophils, eosinophils, and subcorneal pustules.

At 2 months of age, the patient presented with fever and pain on manipulation and had osteomyelitis in the left distal radius and several ribs, as identified by computed tomography. During her 2-month hospitalization, she received several intravenous antibiotic courses, but without clinical improvement. Her CRP level remained mildly elevated at 11 mg/liter and her WBC count remained elevated at 24,000 cells/mm3. Osteolytic lesions in the T1 and T2 vertebral bodies resulted in vertebral collapse, with development of a gibbus deformity of the spine, and this was associated with a chronic inflammatory process, characterized by perivertebral fibrosis (Figure 1E).

At the age of 5 months, she developed hypoxemia and dyspnea, requiring respiratory therapy and home oxygen therapy. She also developed periodic low-grade fever (body temperature of 38–38.5°C). Her skin involvement was mild and characterized by recurrent small pustular lesions on the face, chest, abdomen, and extremities (Figure 1D). In addition to that observed in the left radius, ribs, and vertebral bodies, osteomyelitis was also documented in both femurs, the right wrist, and left shoulder. The patient also had nonunion of the odontoid, with significant C1–C2 subluxation greater than 0.6 cm (Figure 1F). Moreover, a severe and persistent anemia was present, requiring 7 blood transfusions.

At the age of 11 months, her weight (5.5 kg) and length (65 cm) were below the third percentile. She had small pustules on the head and the chest, generalized muscle atrophy, and central cyanosis upon crying. Motor functions were delayed, as she could not stand up. She was started on treatment with prednisolone at a dose of 1 mg/kg/day, which resulted in substantial improvement. The prednisolone dose was tapered and finally stopped when the patient was age 26 months.

The family history was remarkable for juvenile idiopathic arthritis (JIA); in particular, her mother had been diagnosed with polyarticular-onset JIA at the age of 8 years. Her parents are nonconsanguineous and are of Portuguese and French descent. The patient has no siblings.

Genetic findings.

Resequencing analysis of LPIN2, performed to rule out Majeed syndrome in the 2 patients, did not show any mutations. Resequencing analysis of IL1RN (isoform 1, accession no. NM_173842) showed a homozygous in-frame 15-bp deletion (c.213_227delAGATGTGGTACCCAT; p.Asp72_Ile76del) in both girls (Figure 2A). This deletion was not described in previously reported patients with DIRA (12, 13). The deletion was confirmed in forward and reverse sequencing, was present in both sets of parents in a heterozygous state, and was not present in 200 ethnically matched control chromosomes and 200 ethnically unmatched control chromosomes.

Figure 2.

Genetic and functional analyses of the novel IL1RN mutation in patients with the deficiency of interleukin-1 receptor antagonist (DIRA) syndrome. A, The chromatograph indicates the length of the deletion in a wild-type sequence (panel a), a heterozygous carrier sequence (panel b), and an affected (DIRA) sequence (panel c) (denoted by the arrows). Agarose gel electrophoresis in homozygous affected and heterozygous carriers (panel d) indicates a difference in molecular weight between the wild-type and mutant DNA. In panel d, the C denotes a carrier (heterozygous), A denotes an affected individual (patients with DIRA), and N denotes a healthy control. B, The ribbon structure of interleukin-1 receptor antagonist (IL-1Ra) (in green) bound to IL-1 receptor type I (IL-1RI) (in yellow) is modeled using a method previously described by Schreuder et al (15). Values in the green ribbons indicate the 12 β-strands of the IL-1Ra structure, with the missing amino acids located in strand β4 (highlighted in pink). C, Levels of IL1RN mRNA were determined in the peripheral blood of healthy controls, heterozygous carriers, and patients with DIRA. Bars show the mean ± SD fold change in mRNA expression in both Brazilian patients with DIRA (DIRA1 and DIRA2), 1 patient with DIRA from Newfoundland (DIRA3), 1 patient with DIRA from The Netherlands (DIRA4), 1 patient with DIRA from Puerto Rico (DIRA5), and 7 heterozygous carriers (HET) who were the parents of patients in the previously reported populations, relative to that in 18 healthy controls (set at 1). D, Western blotting shows expression of mutant IL-1Ra protein in a patient with DIRA (DIRA5) who was a homozygous carrier (Homo) of the mutation, although in a much lower concentration than that in a parent who was a heterozygous carrier; 0.1 mg of recombinant IL-1Ra (anakinra) was loaded as a positive control.

Functional findings.

Binding affinity of IL-1Ra revealed by in silico modeling.

The ribbon diagram shown in Figure 2B illustrates the structure of IL-1Ra bound to IL-1RI, as previously determined (15). IL-1Ra has 12 β-strands, including a 6-stranded β-barrel with strands β1, β4, β5, β8, β9, and β12. The missing amino acids are in strand β4 (highlighted in pink in Figure 2B). Loss of strand β4 would likely affect the stability of strands β1 and β5, as well as the β4–β5 loop. This may disrupt formation of the entire β-barrel and, as a result, disrupt binding of IL-1Ra to IL-RI (Figure 2B).

Expression of IL1RN mRNA.

RT-PCR showed that, in contrast to the previously reported patients with DIRA, all of whom had deleterious truncating mutations, both Brazilian patients expressed about one-half the amount of mRNA for IL1RN compared to that observed in healthy controls, but the levels were similar to those found in the heterozygous parents of previously reported patients with DIRA (Figure 2C).

Expression of IL-1Ra protein.

Western blot analysis showed that IL-1Ra protein was secreted at lower levels in the 2 Brazilian patients, as compared to that in healthy controls and a heterozygous parent of a previously reported patient (Figure 2D).

Expression of mutant versus wild-type human IL-1Ra.

The wild-type and mutant (lacking amino acids 72–76) versions of IL-1Ra were expressed in E coli after the addition of a C-terminal tag to aid in protein purification. Previous experiments had indicated that addition of a C-terminal tag to IL-1Ra results in only a slight loss in potency. Expression of huIL-1RaWT-FpH was robust, and several milligrams of protein were recovered. In comparison with the wild-type construct, expression of huIL-1RaΔ72–76-FpH was comparable. However, ∼50% of the expressed protein is insoluble, and the remaining, soluble protein does not bind Ni-NTA efficiently. Nevertheless, a small amount of huIL-1RaΔ72–76-FpH was recovered.

Antagonistic activity of human IL-1RaΔ72–76-FpH.

The ability of the mutant form of IL-1Ra to antagonize IL-1 responses was analyzed by monitoring the movement of phospho-JNK protein from the cytoplasm into the nucleus. The huIL-1RaΔ72–76-FpH form completely lacked the ability to inhibit IL-1β–mediated phospho-JNK nuclear translocation. In contrast, the huIL-1RaWT-FpH form demonstrated typical antagonistic activity, regardless of whether it was diluted in PBS or with the huIL-1RaΔ72–76-FpH preparation, demonstrating that the mutant form does not inhibit the functioning of wild-type IL-1Ra. The activity of the wild-type protein was only slightly lower than that of the untagged recombinant human IL-1Ra (rhMet-IL-1Ra), as was expected for the C-terminally tagged protein (Figure 3A).

Figure 3.

Protein binding and functional assays of wild-type and mutant interleukin-1 receptor antagonist (IL-1Ra) protein. A, Wild-type human IL-1Ra (huIL-1RaWT-FpH), but not the mutant form (human IL-1Ra lacking residues 72–76 [huIL-1RaΔ72–76-FpH]), blocked IL-1–mediated phosphorylated JNK nuclear translocation, and neither the simultaneous presence of the mutant protein nor the C-terminal tag used for purification substantially affected its ability to inhibit the response. The mutant protein or phosphate buffered saline (PBS) was used as diluent. The 50% inhibitory concentration (IC50) values for each corresponding sample are given in the table. Bars show the mean ± SD percentage of control activity (POC) after sampling of 100 cells per well in each of 4 wells. N/A = not applicable (referent). B, Surface plasmon resonance reveals that wild-type IL-1Ra (with or without a C-terminal tag), but not mutant IL-1Ra, is able to bind to IL-1 receptor type I. Comparison of samples B and C demonstrates that mutant IL-1Ra does not interfere with receptor binding by the wild-type molecule. ka = association rate constant; kd = dissociation rate constant; KD = equilibrium binding constant (the ratio of kd to ka); RU = resonance units. C, Whole peripheral blood from healthy controls, heterozygotes, and patients with the deficiency of interleukin-1 receptor antagonist (DIRA) syndrome were stimulated with 1 ng/ml of lipopolysaccharide (LPS), 50 ng/ml recombinant human IL-1β (rhIL-1β), or 50 ng/ml rhIL-1α for 3 hours, followed by measurement of cytokine levels in the culture supernatants. Assays with LPS and rhIL-1β included 2 controls, 3 heterozygotes, and 2 patients. Assays with rhIL-1α included 1 control, 2 heterozygotes, and 1 patient. Bars show the mean ± SD. MIP-1α = macrophage inflammatory protein 1α; TNF = tumor necrosis factor α.

Binding of huIL-1RaΔ72–76-FpH to human IL-1RI.

The ability of wild-type and mutant IL-1Ra to bind to IL-1RI was analyzed using a surface plasmon resonance assay. Robust binding was observed for huIL-1RaWT-FpH and rhMet-IL-1Ra. IL-1RaΔ72–76-FpH, however, failed to bind IL-1RI, even at high concentrations (150 nM) (Figure 3B).

Cytokine expression in leukocytes.

Washed whole blood cells from both patients were stimulated with LPS, rhIL-1α, or rhIL-1β, and the results showed increased expression of TNFα, MIP-1α, IL-6, and IL-8 when compared to that in normal healthy controls in response to all 3 stimuli. The heterozygous carriers had the same response as that in controls (Figure 3C). In addition, stimulation with LPS or IL-1α led to increased amounts of IL-1β production, and stimulation with LPS or IL-1β led to increased amounts of IL-1α production. The cytokine pattern in the stimulated peripheral blood cells of these patients was very similar to that observed in previously reported patients with DIRA (12).

Response to treatment with anakinra.

Both patients had an immediate and persistent response to treatment with anakinra at an initial dose of 1 mg/kg/day, and the dose was increased over several months to keep patients in a state of remission of inflammation (CRP level <5 mg/dl) along with a decrease in clinical symptoms and normalized levels of acute-phase reactants. In patient 1, the skin rash cleared within a few days of treatment with anakinra, and the acute-phase reactant levels, WBC count, and platelet count dropped rapidly (Figure 4). Steroid doses were decreased and acitretin was discontinued. Both patients continue to do well while continuing to receive treatment with anakinra at a dose of 2.5 mg/kg/day (patient 1) and 3 mg/kg/day (patient 2), administered by subcutaneous injection.

Figure 4.

Treatment response to anakinra. Treatment with anakinra rapidly improved the clinical signs and symptoms in 2 Brazilian patients with the deficiency of interleukin-1 receptor antagonist syndrome. Top, In patient 1, a rapid improvement in the skin rash was seen after treatment. Bottom, In patients 1 and 2, normalization of the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level as well as a rapid improvement in the white blood cell (WBC) count, hemoglobin (Hb) level, and thrombocytosis were seen at 1 month after treatment.


The DIRA syndrome is an autosomal-recessive autoinflammatory disorder presenting with systemic inflammation and severe skin and bone inflammation in the neonatal period (12, 13). Of the 10 previously reported patients with DIRA, 3 had a fatal course, whereby 2 of the patients died in the perinatal period and 1 died at the age of 9 years. DIRA is caused by homozygous mutations in IL1RN, which encodes the IL-1Ra. All previously reported mutations were deleterious, being either a whole-gene deletion or protein-truncating mutations (frameshift 2-bp deletion or nonsense mutations). The reported mutations led to the absence of the production of a functional IL-1Ra, which is an endogenous antagonist that binds to the IL-1RI. Therefore, it can be concluded that in patients with DIRA, IL-1 signaling cannot be blocked. Treatment with anakinra, a recombinant IL-1Ra, is, in essence, a replacement of the very protein missing in these patients (12, 13). Treatment with anakinra uniformly leads to a rapid clinical improvement.

Herein we report the first cases of DIRA in South America, in 2 unrelated patients from Brazil, both of whom had the same novel homozygous, in-frame, 15-bp deletion. Since the mutation was absent in 400 control chromosomes, we could not glean insight into the frequency of this mutant allele. Unlike previously described deleterious mutations, the 15-bp deletion is expected to lead to the production of a mutant protein. In silico modeling suggested that secondary structures were altered, and in vitro functional assays confirmed the absence of binding of the mutant protein to the IL-1 receptor. The mutant protein lacks antagonistic activity and is not able to function as an inhibitor of the IL-1 receptor. The stimulation of whole blood cells by IL-1 and LPS led to the overproduction of a number of proinflammatory cytokines in patients compared to healthy controls and heterozygous carriers. Although the mutant protein was detected in both patients, its production was reduced. Similar to previously described patients with deleterious mutations, the 2 patients showed a rapid response to treatment with anakinra.

Both of our patients partially responded to corticosteroid therapy, as indicated by an improvement in the skin, bone, and lung abnormalities. One of the patients also had a marked improvement of her skin lesions after treatment with acitretin, which was previously described to lead to skin improvement in an infant with generalized pustular psoriasis (14). The response to steroids in our patients was more pronounced than what has been previously reported, and this raised the question as to whether the mutant protein in our patients might have a residual IL-1 blocking activity. In vitro experiments did not provide any evidence of residual IL-1 blocking activity of the mutant protein. The somewhat better response to steroids might be based on in vivo residual function or on differences in genetic background. It appears that this nondeleterious mutation produces a milder syndrome, although the mechanism of the milder presentation remains elusive.

The DIRA syndrome is characterized by a lack of symptoms in obligate and confirmed carriers (12). In this report, 1 parent from each family had mild inflammatory symptoms. It is quite difficult to interpret these symptoms as part of the phenotype in carriers, but it is prudent to investigate further the medical history of all reported carriers to resolve this issue.

Other monogenic autoinflammatory disorders that can present in the neonatal period include Majeed syndrome, NOMID, and mevalonate kinase deficiency/hyper-IgD syndrome. However, except for the severe presentations observed in patients with NOMID (5), the other genetically defined autoinflammatory diseases are not life-threatening in the neonatal period (2, 16). The presenting features of DIRA include low-to-high–grade fevers, rashes, blood neutrophilia, and neutrophilic infiltrates in the skin and bones, and these are often confused with congenital or perinatal infections in the newborn period (16). Typically, extensive evaluations for infectious causes yield negative results and patients do not respond to treatment with antibiotics. Due to the significant bone involvement, patients with DIRA are often diagnosed as having infectious osteomyelitis, but the negative bone and blood cultures and the unresponsiveness to antibiotics can help to raise the suspicion of an autoinflammatory syndrome.

Although NOMID and the DIRA syndrome share clinical similarities, there are some distinct differences. Patients with NOMID typically present with aseptic meningitis and cochlear inflammation, and mental retardation is quite frequent (17). Patients with DIRA typically do not have neurologic inflammation. Central nervous system vasculitis has been seen in 1 patient with DIRA, and 1 patient with longstanding, untreated disease had a mild cognitive delay. There was no evidence of inner ear inflammation, hearing loss, or eye inflammation in the previously reported patients (12) or in the patients in the current study. Another difference in the clinical presentations is the cutaneous eruptions, which, in patients with DIRA, are pustular, involving the epidermis; in contrast, in patients with NOMID, the epidermis is not involved and the rash is urticaria-like with a neutrophilic infiltrate in the dermis (4, 16). Patients with mevalonate kinase deficiency/hyper-IgD syndrome can also develop manifestations in the newborn period, but patients usually have significant lymphadenopathy, abdominal pain, vomiting, and diarrhea. In addition, a purpuric nonpustular exanthema is present in 80% of the patients (16). In the severe form of mevalonate kinase deficiency, known as mevalonic aciduria, developmental delay, facial dysmorphism, ataxia, hypotonia, myopathy, and cataracts can be clinical presentations (18–20).

The Majeed syndrome also shares clinical similarities with the DIRA sundrome. Patients present with a neutrophilic dermatosis and chronic recurrent multifocal osteomyelitis. However, congenital dyserythropoietic anemia is found only in Majeed syndrome, but not in DIRA (8).

The genetic mutations in NOMID and DIRA are both linked to IL-1. Both diseases respond very well to treatment with IL-1α and IL-1β blocking agents, such as anakinra, and patients with cryopyrin-associated periodic syndromes also respond to therapies directed to IL-1β only. Since IL-1Ra blocks both IL-1α and IL-1β signaling, it remains uncertain whether signaling through IL-1α contributes to the inflammation in patients with DIRA. If there is a specific role for IL-1α, then one might hypothesize that therapies specifically blocking IL-1β may not be sufficient to treat DIRA, but clinical trials to confirm this notion are needed. Patients with Majeed syndrome also present with multifocal osteolytic lesions, but the pathogenic pathways of inflammation linked to LPIN2 mutations, which are the purported genetic cause of Majeed syndrome, have not been resolved. It will be interesting to determine whether the pathogenesis of Majeed syndrome involves the IL-1 pathway, and thus whether patients with this disease respond to IL-1 blockade.

In conclusion, in this report, we have described 2 patients from Brazil with the new autoinflammatory syndrome DIRA. Both patients displayed the same mutation on IL1RN, causing severe cutaneous and bone manifestations. Our findings suggest that the identified novel mutation is a founder mutation that has been introduced in Brazil, which needs to be investigated, since proving a founder effect will facilitate screening for the frequency of this mutation in the Brazilian population and enable genetic screening and prenatal diagnosis and immediate treatment of patients.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. El-Shanti had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Jesus, Silva, Ferguson, Goldbach-Mansky, El-Shanti.

Acquisition of data. Jesus, Osman, Silva, Kim, Pham, Yang, Bertola, Carneiro-Sampaio, Renshaw, Schooley, Brown, Al-Dosari, Sims, Goldbach-Mansky, El-Shanti.

Analysis and interpretation of data. Jesus, Osman, Silva, Kim, Gadina, Yang, Bertola, Carneiro-Sampaio, Renshaw, Schooley, Brown, Al-Dosari, Alami, Sims, Goldbach-Mansky, El-Shanti.


Authors Renshaw, Schooley, Brown, and Sims are employees of Amgen. Amgen had no role in the study design or in the collection, analysis, or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by Amgen.


The authors would like to thank the families for their participation in the study. The authors also thank Nicole Plass, RN for her help with the logistics and clinical data collection, and Deborah Stone, MD and Dawn Chapelle, RN for their help with evaluating the patients and collecting clinical data.