Strong Association of Combined Genetic Deficiencies in the Classical Complement Pathway With Risk of Systemic Lupus Erythematosus and Primary Sjögren's Syndrome

Abstract

Objective: Complete genetic deficiency of the complement component C2 is a strong risk factor for monogenic systemic lupus erythematosus (SLE), but whether heterozygous C2 deficiency adds to the risk of SLE or primary Sjögren's syndrome (SS) has not been studied systematically. This study was undertaken to investigate potential associations of heterozygous C2 deficiency and C4 copy number variation with clinical manifestations in patients with SLE and patients with primary SS.

Methods: The presence of the common 28-bp C2 deletion rs9332736 and C4 copy number variation was examined in Scandinavian patients who had received a diagnosis of SLE (n = 958) or primary SS (n = 911) and in 2,262 healthy controls through the use of DNA sequencing. The concentration of complement proteins in plasma and classical complement function were analyzed in a subgroup of SLE patients.

Results: Heterozygous C2 deficiency-when present in combination with a low C4A copy number-substantially increased the risk of SLE (odds ratio [OR] 10.2 [95% confidence interval (95% CI) 3.5-37.0]) and the risk of primary SS (OR 13.0 [95% CI 4.5-48.4]) when compared to individuals with 2 C4A copies and normal C2. For patients heterozygous for rs9332736 with 1 C4A copy, the median age at diagnosis was 7 years earlier in patients with SLE and 12 years earlier in patients with primary SS when compared to patients with normal C2. Reduced C2 levels in plasma (P = 2 × 10^-9 ) and impaired function of the classical complement pathway (P = 0.03) were detected in SLE patients with heterozygous C2 deficiency. Finally, in a primary SS patient homozygous for C2 deficiency, we observed low levels of anti-Scl-70, which suggests a risk of developing systemic sclerosis or potential overlap between primary SS and other systemic autoimmune diseases.

Conclusion: We demonstrate that a genetic pattern involving partial deficiencies of C2 and C4A in the classical complement pathway is a strong risk factor for SLE and for primary SS. Our results emphasize the central role of the complement system in the pathogenesis of both SLE and primary SS.

Figure 1

Heterozygosity of the 28‐bp C2 deletion rs9332736 in systemic lupus erythematosus (SLE) patients and primary Sjögren's syndrome (SS) patients. A, Prevalence of heterozygous carriers of the C2 loss‐of‐function variant rs9332736 in SLE patients (n = 955), primary SS patients (n = 910), and healthy controls (n = 2,262). Individuals homozygous for the rs9332736 variant (2 SLE patients and 1 primary SS patient) were excluded. B, Risk of association with SLE or primary SS according to number of C4A copies in rs9332736, with odds calculated relative to healthy controls. C, Risk of association with SLE or primary SS according to the combined effect of C4A copy number and rs9332736 heterozygosity (ref/del), with odds calculated relative to a C4A copy number of 2 and normal C2 (ref/ref). Due to rs9332736 segregating with C4A, no individuals heterozygous for rs9332736 have 0 C4A copies. Data were analyzed using logistic regression and adjusted for sex (A, B, and C) and C4B copy number (B and C). In B and C, bars show the 95% confidence intervals. D, Linkage disequilibrium (LD; r²) between the 28‐bp C2 deletion rs9332736 and HLA alleles/biallelic single‐nucleotide polymorphisms in the HLA region. LD with HLA alleles for 6 HLA genes are indicated by triangles, including HLA alleles A, C, B, DRB1, DQB1, and DPB1. The vertical gray‐shaded line indicates the genomic position of C2. LD was estimated using SweGen whole‐genome sequencing samples (n = 1,000). Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/doi/10.1002/art.42270/abstract.

Figure 2

Concentration of plasma complement components in systemic lupus erythematosus (SLE) patients with the 28‐bp C2 deletion rs9332736 (ref/del) and in SLE patients with normal C2 (ref/ref), including plasma C2 concentration relative to a reference serum (n = 261) (A), classical complement function (n = 140) (B), plasma C3 (C) and plasma C4 (D) concentration stratified by presence of the 28‐bp C2 deletion rs9332736 and plasma C2 concentration (n = 258 for C and D). Data were analyzed by analysis of variance and adjusted for sex and age at sampling (A, C, and D), and for copy number of C4A and C4B (B). Plasma concentration of C4 was square root transformed. Data are shown as box plots. Each box represents the 25th to 75th percentiles. Lines inside the boxes represent the median, and whiskers extend to 1.5 times the interquartile range. Solid circles represent individual SLE patients heterozygous for rs9332736; open circles represent individual SLE patients with normal C2. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/doi/10.1002/art.42270/abstract.

Figure 3

Age at diagnosis of systemic lupus erythematosus (SLE) and primary Sjögren's syndrome (SS) among patients with the 28‐bp C2 deletion rs9332736 (ref/del) relative to SLE patients and primary SS patients with normal C2 (ref/ref). Kaplan‐Meier plots depict the age at diagnosis of SLE in those with the 28‐bp C2 deletion rs9332736 and 1 C4A copy (n = 15) or 2 C4A copies (n = 18), relative to SLE patients with normal C2 and 1 C4A copy (n = 352) or 2 C4A copies (n = 446) (A), and age at diagnosis of primary SS in those with the 28‐bp C2 deletion rs9332736 and 1 C4A copy (n = 17) or 2 C4A copies (n = 11), relative to primary SS patients with normal C2 and 1 C4A copy (n = 412) or 2 C4A copies (n = 363) (B). Data were analyzed using a Cox proportional hazards regression model adjusted for sex and C4B copy number. HR = hazard ratio; 95% CI = 95% confidence interval. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/doi/10.1002/art.42270/abstract.

Figure 4

Age at first occurrence of nephritis in SLE patients. The Kaplan‐Meier plot depicts age at first nephritis event in SLE patients with the 28‐bp C2 deletion rs9332736 and 1 C4A copy (n = 14) or 2 C4A copies (n = 18), and in SLE patients with normal C2 and 1 C4A copy (n = 322) or 2 C4A copies (n = 412). Data were analyzed using a Cox proportional hazards regression model adjusted for sex and C4B copy number. See Figure 3 for definitions. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/doi/10.1002/art.42270/abstract.

Figure 5

Summary of the conduct of the study. Homozygosity of the 28‐bp deletion rs9332736 in C2 is associated with monogenic lupus. Heterozygous individuals are not at risk of SLE or primary SS if they have 2 or more copies of C4A, whereas individuals heterozygous for rs9332736 who have 1 copy of C4A are at substantial risk of both SLE and primary SS. OR = odds ratio (see Figure 3 for other definitions).