Contribution of rare and predicted pathogenic gene variants to childhood-onset lupus: a large, genetic panel analysis of British and French cohorts

Summary

Background

Systemic lupus erythematosus (SLE) is a rare immunological disorder and genetic factors are considered important in its causation. Monogenic lupus has been associated with around 30 genotypes in humans and 60 in mice, while genome-wide association studies have identified more than 90 risk loci. We aimed to analyse the contribution of rare and predicted pathogenic gene variants in a population of unselected cases of childhood-onset SLE.

Methods

For this genetic panel analysis we designed a next-generation sequencing panel comprising 147 genes, including all known lupus-causing genes in humans, and potentially lupus-causing genes identified through GWAS and animal models. We screened 117 probands fulfilling American College of Rheumatology (ACR) criteria for SLE, ascertained through British and French cohorts of childhood-onset SLE, and compared these data with those of 791 ethnically matched controls from the 1000 Genomes Project and 574 controls from the FREX Consortium.

Findings

After filtering, mendelian genotypes were confirmed in eight probands, involving variants in C1QA, C1QC, C2, DNASE1L3, and IKZF1. Seven additional patients carried heterozygous variants in complement or type I interferon-associated autosomal recessive genes, with decreased concentrations of the encoded proteins C3 and C9 recorded in two patients. Rare variants that were predicted to be damaging were significantly enriched in the childhood-onset SLE cohort compared with controls; 25% of SLE probands versus 5% of controls were identified to harbour at least one rare, predicted damaging variant (p=2·98 × 10⁻¹¹). Inborn errors of immunity were estimated to account for 7% of cases of childhood-onset SLE, with defects in innate immunity representing the main monogenic contribution.

Interpretation

An accumulation of rare variants that are predicted to be damaging in SLE-associated genes might contribute to disease expression and clinical heterogeneity.

Funding

European Research Council.

Introduction

The term systemic lupus erythematosus (SLE) describes a rare and heterogeneous set of phenotypes characterised by the presence of autoantibodies targeting nuclear autoantigens, and by type I interferon upregulation. Familial aggregation and higher concordance rates between monozygotic compared with dizygotic twins suggest a major hereditary component to disease pathogenesis.¹ Genome-wide association studies (GWAS) have identified more than 90 SLE-associated loci.2, 3 GWAS-defined variants are common, confer small effects on disease susceptibility, and typically fall outside of coding regions. By contrast, monogenic forms of SLE, of which almost 30 have been described in humans, and more than 60 in mice,⁴ involve highly penetrant rare variants in protein encoding DNA.⁵

Mendelian forms of disease can help define the involvement of discrete pathways in pathogenesis. For example, both classical complement gene and DNASE1L3 deficiency highlight the importance of efferocytosis in lupus pathology,6, 7 whereas the relevance of type I interferon signalling to SLE is underlined by the association with the mendelian type I interferonopathies.8, 9 B cells are also a key player in lupus causation,¹⁰ with PKCδ deficiency the first described B-cell-related form of monogenic lupus,¹¹ and heterozygous germline alterations in IKZF1, encoding the B-cell transcription factor IKAROS, identified as a cause of autoimmunity, including SLE.¹²

Research in context

Evidence before this study

We searched PubMed for studies published until Dec 31, 2013, with no start date or language restrictions and with the terms “monogenic lupus”, “GWAS AND lupus”, “mice models AND lupus”, and “immunodeficiencies AND lupus”. Published data highlight familial aggregation, and higher concordance rates between monozygotic (20–40%) versus dizygotic twins and other full siblings (2–5%), suggesting a major hereditary component to the pathogenesis of systemic lupus erythematosus (SLE). Although genome-wide association studies (GWAS) have identified more than 90 loci as being robustly associated with the lupus phenotype, it is an overlooked fact that mendelian forms of lupus have been described in the context of almost 30 discrete genotypes in humans, referred to here as the mendeliome, and more than 60 single gene defects in mice. At present, the importance of these observations for the understanding of the biology of lupus remains unclear.

Added value of this study

To address the issue of the genetic contribution to lupus in children, we analysed 147 genes at depth, selected on the basis of their involvement in lupus pathogenesis, in 117 unrelated probands with SLE demonstrating onset before the age of 16 years. Within the mendeliome, we observed mutant genotypes in 7% of probands. Furthermore, heterozygous, predicted pathogenic variants in genes previously implicated in lupus causation were significantly enriched in the SLE cohort compared with controls. Our study supports the concept of oligogenic inheritance in genetic susceptibility to SLE, with an accumulation of rare, predicted pathogenic variants in different genes in patients.

Implications of all the available evidence

Numerous clinical trials done in SLE cohorts have yet to show a positive effect with various drugs, and existing clinical and laboratory criteria are not sufficient to detect specific factors that determine treatment outcome. Our results are consistent with an underlying heterogeneity of the lupus phenotype, with dysfunctional innate immunity having an important role. Our data suggest that an exploration of genetic burden in single individuals might be informative in tailoring future personalised therapeutic interventions.

Recent sequencing studies suggest a false dichotomy in categorisation of diseases according to strict complex or mendelian models; an alternative possibility is to highlight the importance of combinations of small numbers of rare variants promoting disease in a single individual and at a population level.¹³ Immune responses are variable in humans, with up to 40% of this diversity estimated to be explained by genetic variation.¹⁴ In association with environmental factors, such genetic polymorphisms might promote breakdown of tolerance,¹⁵ as exemplified by the lupus phenotype.

We aimed to explore potential genetic factors contributing to childhood-onset SLE, by analysing the mutational spectrum directly causing, or strongly influencing, disease pathogenesis.