IFIH1 and DDX58 gene variants in pediatric rheumatic diseases

Abstract

BACKGROUND

The IFIH1 gene codes the MDA5 protein and the DDX58 gene codes the RIG-I receptor. Both proteins are parts of the interferon (IFN) I signaling pathway and are responsible for antiviral defense and innate immune response. IFIH1 and DDX58 polymorphisms are associated with a spectrum of autoimmune diseases. Rare gain-of-function IFIH1 mutations have been found in Singleton-Merten and Aicardi-Goutières syndrome, while DDX58 mutation can cause atypical Singleton-Merten syndrome.

AIM

To characterize children with pediatric rheumatic diseases (PRD) carrying DDX58 or IFIH1 variants.

METHODS

Clinical exome sequencing was performed on 92 children with different PRD. IFIH1 and DDX58 variants have been detected in 14 children. IFN-I score has been analyzed and the clinical characteristics of patients have been studied.

RESULTS

A total of seven patients with systemic lupus erythematosus (SLE) (n = 2), myelodysplastic syndrome with SLE features at the onset of the disease (n = 1), mixed connective tissue disease (MCTD) (n = 1), undifferentiated systemic autoinflammatory disease (uSAID) (n = 3) have 5 different variants of the DDX58 gene. A common non-pathogenic variant p.D580E has been found in five children. A rare variant of uncertain significance (VUS) p.N354S was found in one patient with uSAID, a rare likely non-pathogenic variant p.E37K in one patient with uSAID, and a rare likely pathogenic variant p.Cys864fs in a patient with SLE. Elevated IFN-I score was detected in 6 of 7 patients with DDX58 variants. Seven patients had six different IFIH1 variants. They were presented with uSAID (n = 2), juvenile dermatomyositis (JDM) (n = 1), SLE-like disease (n = 1), Periodic fever with aphthous stomatitis, pharyngitis, and adenitis syndrome (n = 1), and systemic onset juvenile idiopathic arthritis (n = 1). Three patients have VUS p.E627X, one patient has benign variant p.I923V. Rare VUS p.R595H was detected in the JDM patient. Another rare VUS p.L679Ifs*2 and previously not reported variant p.V599Ffs*5 were detected in the patient with uSAID. One patient with uSAID has rare VUS p.T520A. All patients had elevated IFN-I scores.

CONCLUSION

Rare compound-heterozygous IFIH1 variant (p.L679Ifs*2 and p.V599Ffs*5), heterozygous IFIH1 variant (p.T520A) and heterozygous DDX58 variant (p.Cys864fs) are probably disease causative for uSAID and SLE. The majority of patients with different DDX58 and IFI1 variants had hyperactivation of the IFN I signaling pathway.

Key Words: IFIH1, DDX58, Undifferentiated systemic autoinflammatory disease, Systemic lupus erythematosus, Interferon-I score

Core Tip: Interferon (IFN) I signaling pathway is the important part of innate immune system and antiviral defense. It’s known that defects in the components of IFN I signaling system can leads to its hyperactivation. This pathogenetic phenomenon is considered as one of the most important in the pathogenesis of immune-mediated diseases, such as systemic lupus erythematosus, dermatomyositis, systemic autoinflammatory diseases. From database containing 92 patients with different rheumatic diseases, whom clinical exome sequencing was performed we selected 14 children (10 girls and 4 boys): 7 patients had DDX58 and 7 had IFIH1 gene variants. The majority of patient with different DDX58 and IFI1 variants had hyperactivation of IFN I signaling pathway.

INTRODUCTION

The DDX58 gene encodes the RIG-I receptor (retinoic acid-induced protein I), which is responsible for the virus’s ribonucleic acid (RNA) recognition[1]. RIG-1 interacts with short-stranded RNA (ssRNA) of viruses that lead to mitochondrial antiviral-signaling protein (MAVS) activation, which in turn leads to the expression of interferon (IFN)-regulated genes and the secretion of type I IFNs. The IFIH1 gene encodes a cytoplasmic receptor that senses double-stranded RNA viral products to activate type I IFN signaling through the MAVS adaptor molecule. This can inhibit virus replication and modulate cellular immune responses. IFIH1 may also help recognize and limit the replication of ssRNA viruses[2].

IFN I signaling pathway is an important part of the innate immune system and antiviral defense[3]. It’s known that defects in the components of the IFN I signaling system can lead to its hyperactivation. This pathogenetic phenomenon is considered one of the most important in the pathogenesis of immune-mediated diseases, such as systemic lupus erythematosus (SLE), dermatomyositis (DM), and systemic autoinflammatory diseases (SAID)[4]. The roles of DDX58 and IFIH1 gene variants in the pathogenesis of immune-mediated diseases are not fully studied. Our study aimed to describe children with different rheumatic diseases who have variants in DDX58 or IFIH1 genes.

MATERIALS AND METHODS

From a database containing 92 patients with different rheumatic diseases, for whom clinical exome sequencing was performed we selected 14 children (10 girls and 4 boys) having the variants of DDX58 or IFIH1 genes. They have the following diagnosis: SAID (n = 6), SLE (n = 5), juvenile DM (JDM) (n = 1), systemic onset juvenile idiopathic arthritis (n = 1), mixed connective tissue disease (n = 1).

IFN-I score was measured in the majority of patients. Real-time polymerase chain reaction with previous reverse transcription of RNA was used for IFN-I score assessment. Amplification was carried out using hybridization probes (TaqMan probes); changes in gene expression were evaluated by the ratio of signals “studied gene/referee gene”); the value of the IFN-I score corresponds to the median relative expression of 5 genes studied (IFI44L, IFI44, IFIT3, LY6E, MX1). The IFN-I score value of ≥ 2.0 was taken as a cut-off, indicating increased transcription of IFN-I-regulated genes i.e., the presence of an IFN signature.

RESULTS

Five different variants of the DDX58 gene were found in 7/14 patients included in the study, among which the p.D580E variant was found in 4 patients. One patient had 2 different variants of DDX58, including p.D580E. The variant p.D580E is a frequent polymorphism [minor allele frequency (MAF): 0.11]. Two patients had likely non-pathogenic variants DDX58 according to American College of Medical Genetics (ACMG) (p.S144F and p.E37K). A rare variant of uncertain significance (VUS) of the DDX58 (p.N354S) gene was detected in one patient. Only one patient with SLE had a likely pathogenic variant of the DDX58 gene - p.Cys864fs. Patients had the following distribution according to the diagnosis: SLE (n = 2), myelodysplastic syndrome with SLE features at the onset of the disease (n = 1), mixed connective tissue disease (n = 1), undifferentiated SAID (uSAID) (n = 3). IFN-I score was evaluated in 6/7 patients with DDX58 variants. IFN-I score was higher by 4 or more times compared to reference values in 5 patients, whereas in one patient IFN-I score was slightly increased - 2.13 UE/mL (normal value is less than 2.0 UE/mL). Table 1 shows variants of the DDX58 genes and the values of the IFN-I scores. Brief clinical characteristics of patients with rare DDX58 variants will be presented below.

Table 1 DDX58 variants in children with immune-mediated diseases.

No.	DDX58 variant	MAF	ClinVar	ACMG	IFN-I score (n.v.0-1.9 UE/L)	Diagnosis
1	c1061A>G (p.N354S)	0.0001	VUS	VUS	2.13	uSAID
2	c109G>A (p.E37K)	0.000008	Not reported	Likely non-pathogenic	N/A	uSAID
3	c.2587_2590dup (p.Cys864fs)	0.0006	Not reported	Likely pathogenic	21.0	SLE
	c.1740T>A (p.D580E)	0.11	Non-pathogenic	Non-pathogenic
4	c.431C>T (p.S144F)	0.0277	Non-pathogenic	Likely non-pathogenic	13.31	MCTD
5	c.T1740A (p.D580E)	0.11	Non-pathogenic	Non-pathogenic	8.5	SLE and MDS
6	c.1740T>A (p.D580E)	0.11	Non-pathogenic	Non-pathogenic	24.35	SLE
7	c.1740T>A (p.D580E)	0.11	Non-pathogenic	Non-pathogenic	18.93	uSAID

ACMG: The American College of Medical Genetics; GnomAD: The Genome Aggregation Database; MAF: Minor allele frequency; MCTD: Mixed connective tissue disease; MDS: Myelodysplastic syndrome; n.v.: Normal value; SLE: Systemic lupus erythematosus; uSAID: Undifferentiated systemic autoinflammatory disease; VUS: A variant of uncertain significance; N/A: Not applicable.

Patient’s description

Patient 1 (Table 1) is 12-year-old the boy developed fever, abdominal pain, diarrhea, livedo reticularis, and edema of the 4^th proximal interphalangeal (PIP) joint of the left hand and 3^rd PIP joint of the right hand. Shigella spp. was detected and he was treated with antibiotics. Abdominal symptoms and fever resolved, but he had periodic fever episodes and livedo reticularis. Blood tests showed elevated inflammatory markers. We ruled out autoimmune diseases, inflammatory bowel disease. Genetic testing was performed and VUS in the DDX58 gene (p.N354S) was detected. The patient was treated with short-course corticosteroids. Fever, livedo reticularis, and laboratory inflammation resolved.

Patient 2 (Table 1) is a 7-year-old boy who has had 7 episodes of exudative pericarditis for one year preceding our hospital admission. He required pericardiocentesis twice. Pericardial effusion usually was accompanied by systemic inflammation (leukocytosis, increased C-reactive protein) and has been resolved on treatment with corticosteroids. Infectious, oncologic, and autoimmune diseases have been excluded. Multigene-targeted sequencing revealed a pathogenic mutation in the JAK1 gene and likely a non-pathogenic variant of the DDX58 gene (p.E37K). Colchicine and canakinumab were initiated, however, due to four following flares during the year of treatment, he was switched to tocilizumab.

Patient 3 (Table 2) is a 16-year-old girl with SLE. Clinical manifestations in the disease onset included fever, lymphadenopathy, pancytopenia, and minimal proteinuria (0.2 g/d) were noted. The malar rash, stomatitis, and polyserositis (by computed tomography) were developed. She had positive ANA, anti-dsDNA antibodies, and hypocomplementemia. She’s been treated with corticosteroids (IV, oral), hydroxychloroquine, and rituximab. Disease remission was achieved and corticosteroids were stopped. IFN-I score was increased 10 times than the normal value (21.0 UE, normal value less than 2.0 UE), and following multigene targeted sequencing revealed two variants in the DDX58 gene (p.Cys864fs, p.D580E). A rare variant of p.C864Ffs*9 was inherited from a healthy mother. There is no information about the pathogenicity of this variant in the ClinVar database. Variant DDX58 p.D580E refers to frequent benign variants of the gene.

Table 2 IFIH1 variants in children with immune-mediated diseases.

No.	IFIH1 variant	MAF	ClinVar	ACMG	IFN-score (n.v. 0-1.9 UE/L)	Diagnosis
1	c.2035_2036del (p.L679Ifs*2)	0.0001	VUS	VUS	5.4	uSAID
	c.1795delG (p.V599Ffs*5)		Not reported	Not reported
2	c.1558A>G (p.T520A)	0.0002	CIP	VUS	5.52	uSAID
3	c.1784G>A (p.R595H)	0.0001	Likely benign	VUS	29.0	JDM
4	c.1879G>T (p.E627X)	0.003	CIP	VUS	20.2	SLE-like disease
5	c.1879G>T (p.E627X)	0.003	CIP	VUS	11.3	SLE
6	c.1879G>T (p.E627X)	0.003	CIP	VUS	2.95	PFAPA
7	c.A2767G (p.I923V)	0.0126	Benign/likely benign	Benign	10.0	soJIA

ACMG: The American College of Medical Genetics; CIP: Conflicting interpretations of pathogenicity; GnomAD: The Genome Aggregation Database; JDM: Juvenile dermatomyositis; MAF: Minor allele frequency; n.v.: Normal value; PFAPA: Periodic fever with aphthous stomatitis, pharyngitis, and adenitis; SLE: Systemic lupus erythematosus; soJIA: Systemic onset juvenile idiopathic arthritis; uSAID: Undifferentiated systemic autoinflammatory disease; VUS: A variant of uncertain significance.

Different variants in the IFIH1 gene were found in 7/14 patients (Table 2). All patients had elevated IFN scores. Three patients had the p.E627X variant. This variant has conflicting interpretations of pathogenicity according to ClinVar and uncertain significance according to ACMG. One patient with systemic onset juvenile idiopathic arthritis had a benign p.I923V variant, and one patient with JDM had a rare (MAF 0.0001) likely benign coding sequence variant p.R595H.

Patient 1 (Table 2) had compound-heterozygous variants in the IFIH1 gene (p.L679Ifs*2 and p.V599Ffs*5) that can contribute to disease development. She is 17-year-old girl with undifferentiated interferonopathy. Her clinical manifestations included fever, systemic inflammation, panniculitis, sialadenitis, nodular rash, hepatitis, migraine headaches, and growth failure. Treatment with corticosteroids, azathioprine, and mycophenolate mofetil was ineffective. The application of tofacitinib partially controlled her disease (no fever flares and rash) and allowed her to taper corticosteroids to 0.15 mg/kg.

Patient 2 (Table 2) is 16-year-old girl with fever, pancytopenia, petechial rash, hepatomegaly, lower extremities edema, inflammation, and hyperferritinemia. She initially was treated with IV corticosteroids and сyclosporine A. Then she developed panniculitis nodules in the back, the abdomen, and the lower extremities, and muscle weakness. Pancytopenia and inflammation had recurred. Biopsy of nodules revealed adipocytes and xanthoma cells infiltrating by lymphocytes. Electromyography detected axonal-demyelinating polyneuropathy of motor and sensory nerve fibers of the upper and lower extremities. Further treatment with different medications, including cyclosporine A, etanercept, tocilizumab, tofacitinib, and anakinra showed partial efficacy. It’s unknown does her heterozygous variant p.T520A lead to disease manifestations or not. Functional tests are necessary to prove its pathogenicity.

DISCUSSION

Studies of the DDX58 gene in patients with rheumatic diseases are mostly limited to the evaluation of DDX58 polymorphisms as one of the factors of susceptibility to viral diseases. It is known that some DDX58 gene variants can lead to atypical Singleton-Merton syndrome, which was first described in 2015 in family members who had glaucoma, aortic calcification, and skeletal abnormalities[5]. Subsequently, the DDX58 gene variants were studied in 100 patients with congenital glaucoma, and in one family with congenital glaucoma, dental abnormalities, and skeletal dysplasia, variant c.803G>T (p.Cys268Phe) was found. The functional tests confirmed the pathogenicity of this gene variant. Hyperactivation of the type I IFN signaling system was observed in all the above-mentioned patients[5]. Two families with the Signleton-Merten syndrome having glaucoma, psoriasis-like rashes, calcifications of the joints, and aorta had DDX58 gene variants[6].

In the study of 15 patients with IIM 5/5 patients with DM had an increased expression of DDX58, which was associated with excessive secretion of IFN-beta[7]. All patients with DDX58 variants had high levels of IFN-I scores in our study. The association between DDX58 p.D580E polymorphism and SLE has not been confirmed in the study of 344 patients with SLE and 641 healthy persons[8]. Three of our patients with SLE had the same DDX58 polymorphic variant which is unlikely to have any causal role.

A novel DDX58 pathogenic variant R109C has been recently identified in five unrelated families with lupus nephritis. Transcriptome analysis revealed an increased IFN signature in patient monocytes. One patient was effectively treated with baricitinib[9]. Gain-of-function IFIH1 gene variants lead to an inadequate perception of both own and viral nucleic acids and contribute to the hyperactivation of the IFN-I type[10]. Currently, less than 100 IFIH1 mutations are known.

It was recently shown that some IFIH1 polymorphisms are associated with a spectrum of autoimmune diseases[11]. Rare IFIH1 variants have been found in Aicardi- Goutières syndrome-7, Singleton-Merten syndrome, and MDA-5 immunodeficiency. All these diseases are characterized by IFN signaling upregulation[12]. In our group of patients with the IFIH1 variant, hyperactivation of the IFNI signaling pathway was also observed, but it is unclear whether it was caused by IFIH1 defects.

In our opinion, two patients with uSAID have probably causative IFIH1 variants (compound-heterozygous p.L679Ifs*2, p.V599Ffs*5, and heterozygous p.T520A). Functional tests are pending. Almlöf et al[13] detected two heterozygous missense IFIH1 variants (p.Arg77Trp and p.Arg374Cys) with high potential to contribute to SLE. IFIH1 gene variant p.E627X found in our patients showed a significant association with decreased risk of type 1 diabetes[14].

Combined exome sequencing, transcriptomic analysis, and in vitro functional tests have been analyzed to identify genetic variants causing predisposition to severe courses of common respiratory viral infections. Pathogen-restricted immunodeficiency due to loss-of-function variants in IFIH1 (p.Glu627Ter, p.Leu509_Glu547del, and p.Ile872Ter) have been detected. They result in defective innate recognition of RNA viruses, preventing the activation of an efficient antiviral IFN response[15].

Homozygous nonsense mutation, c.2665A>T (p.Lys889*) in the IFIH1 gene was identified in a patient with microcephaly, severe psychomotor retardation, seizures, cataracts, and recurrent episodes of prolonged and severe chest infections. The results of the western blot analysis of protein from cultured fibroblasts of the patient indicated the absence of wild-type MDA5/IFIH1, compatible with nonsense-mediated decay[16].

It is not fully understood whether influence detected variants IFIH1 and DDX58 on the severity and activity of rheumatic diseases. The patients with rare pathogenic or likely pathogenic variants have had clinical features of interferonopathy (fever, livedo, nodular rash, panniculitis). It seems that patients with IFIH1 and DDX58 variants may be considered candidates for agents blocking the IFN-I signaling pathway (IFN-I receptor antibodies, JAK inhibitors) especially when standard treatment is ineffective.

We believe that the results of our study allow us to consider some patients with rheumatic diseases with unusual manifestations or resistance to standard treatment as patients with monogenic disorders, in which other pathogenic signaling pathways are involved and another type of treatment can be assumed. This information added to the discussion.

Genetic testing is becoming an increasingly accessible and widely used method in clinical practice. Careful assessment of clinical signs and family history, selection of patients with unusual clinical manifestations, and paying more attention to patients who do not respond to standard treatment protocols should alert the doctor about possible genetic disorders. Evaluation of the activity of the IFN type I signaling pathway may be an additional criterion when selecting patients for genetic testing. However, the results of the IFN-I score analysis remain contradictory in several studies, a high concentration can be observed both in classical autoimmune (SLE, DM) and in rare forms of auto-inflammatory diseases[17]. Sönmez et al[18] proposed preliminary classification criteria for type I interferonopathy, which were developed based on the analysis of a small group of patients. Assessing the sensitivity of these criteria is difficult due to the rare occurrence of these conditions. The development of international registries of both patients with interferonopathies and with genetic variants is promising for the development of diagnostic tools.

Future perspectives are associated with the wider use of genetic testing in patients with seemingly classic rheumatic diseases. The creation of new drugs or clinical trials of existing drugs, but to new indications (e.g., JAK inhibitors in DM and SLE, IFN-I receptor antibodies for DM), seem promising. The limitations of our study are related to a very small number of patients, and the heterogeneity of the selected group of patients. The results are preliminary. Without functional studies, it is very difficult to judge the actual pathogenicity of these variants.

CONCLUSION

The majority of patients with DDX58 and IFI1 variants had hyperactivation of the IFN I signaling pathway. Thus, the use of RNA-based IFN-I score seems to help select candidates for further genetic analysis. This study extends the existing data on the spectrum of DDX58 and IFIH1-associated phenotypes. Further studies and collection of patients with IFIH1 and DDX58 variants are needed.

ARTICLE HIGHLIGHTS

Research background

The IFIH1 gene and the DDX58 gene both are involved in the interferon (IFN) type I signaling pathway. Monogenic diseases with the features of systemic rheumatic diseases were described in patients with these genetic variants. Rheumatic disease patients with IFN type I hyperactivation may have these variants.

Research motivation

Patients with rheumatic diseases with unusual manifestations (clinical features of interferonopathy) or resistant to standard treatment may have molecular variants in genes, regulating IFN I type pathway.

Research objectives

To describe children with different rheumatic diseases who have variants in DDX58 or IFIH1 genes.

Research methods

Clinical exome sequencing was performed 92 patients with different rheumatic diseases, and 14 children (10 girls and 4 boys) with DDX58 or IFIH1 genes were selected. They have the following diagnosis: Systemic autoinflammatory disease (n = 6), systemic lupus erythematosus (n = 5), juvenile dermatomyositis (n = 1), systemic onset juvenile idiopathic arthritis (n = 1), mixed connective tissue disease (n = 1). Real-time polymerase chain reaction with previous reverse transcription of RNA was used for IFN-I score assessment.

Research results

All patients had elevated IFN-I scores. Variants in both genes were described in the studied population.

Research conclusions

Patients with DDX58 and IFI1 variants had hyperactivation of the IFN I signaling pathway. The RNA-based IFN-I score is a good tool to select candidates for further genetic analysis.

Research perspectives

Research perspectives are more on molecular tests in rheumatic diseases, more functional studies confirming the role of genetic variants, and an assessment of the efficacy and safety of drugs on new indications.