SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

Objective

Hypocomplementemic urticarial vasculitis syndrome (HUVS) is characterized by recurrent urticaria along with dermal vasculitis, arthritis, and glomerulonephritis. Systemic lupus erythematosus (SLE) develops in >50% of patients with HUVS, although the pathogenesis is unknown. The aim of this study was to identify the causative DNA mutations in 2 families with autosomal-recessive HUVS, in order to reveal the pathogenesis and facilitate the laboratory diagnosis.

Methods

Autozygosity mapping was combined with whole-exome sequencing.

Results

In a family with 3 affected children, we identified a homozygous frameshift mutation, c.289_290delAC, in DNASE1L3. We subsequently identified another homozygous DNASE1L3 mutation leading to exon skipping, c.320+4delAGTA, in an unrelated family. The detected mutations led to loss of function, via either nonsense-mediated messenger RNA decay or abolished endonuclease activity, as demonstrated by a plasmid nicking assay.

Conclusion

These results show that HUVS is caused by mutations in DNASE1L3, encoding an endonuclease that previously has been associated with SLE.

Urticarial vasculitis is a clinical entity in which the gross cutaneous lesions resemble urticaria and appear histologically as vasculitis. The incidence of vasculitis in patients with apparent urticaria is 2–20% ([1]). Hypocomplementemia is a well-described laboratory finding in patients with urticarial vasculitis, although complement levels are normal in the majority of patients. The hypocomplementemic form of urticarial vasculitis is more often associated with systemic symptoms and has been linked to connective tissue diseases such as systemic lupus erythematosus (SLE) ([2]). Hypocomplementemic urticarial vasculitis syndrome (HUVS) was first described by McDuffie et al in 1973 as an immune complex disease characterized by recurrent urticaria and a variety of systemic manifestations ([3]).

HUVS usually presents in the third or fourth decade of life. Two major criteria (recurrent urticaria for >6 months and hypocomplementemia) and at least 2 minor criteria (dermal vasculitis on biopsy, arthralgia or arthritis, uveitis or episcleritis, recurrent abdominal pain, glomerulonephritis, and/or a decrease in the C1q level with positive results of a C1q precipitation test) are required for the clinical diagnosis of HUVS ([4, 5]). A diagnostic laboratory test does not exist. Small numbers of pediatric patients with HUVS have been reported in the literature ([6-9]).

A debate continues regarding whether HUVS is a rare subset or unusual type of SLE or is a separate entity entirely, because SLE develops in >50% of patients with HUVS ([10]). Familial occurrence of HUVS has not been reported other than in the patients described in the current study. Although no genetic studies in HUVS have been previously performed, Wisnieski et al ([11]) reported identical twins with HUVS and suggested that genetic background has a role in HUVS. The described twins were living in separate locations, and HUVS developed at different ages. Thus, those investigators proposed that, as in SLE, development of HUVS requires both genetic and environmental factors. The aim of this study was to identify the causative DNA mutations in 2 families with autosomal-recessive HUVS, in order to reveal the pathogenesis and assist in the laboratory diagnosis.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

The study was approved by the ethics committees at Ankara University and by the institutional review board at the University of Miami. Informed consent was obtained from all participants or, in the case of minors, from the parents.

Patients

HUVS was diagnosed in 5 children from 2 families (Table 1 and Figure 1). Clinical findings in the affected children in one family have recently been reported ([12]). All of the patients fullfilled the clinical diagnostic criteria for HUVS. Two additional multiplex families and 1 simplex family with isolated SLE (without HUVS) were also included. Parental consanguinity was present in all families, suggesting autosomal-recessive inheritance. Four of the clinically evaluated affected children in these 3 families were ages 11–17 years at the time of disease onset, and all satisfied the American College of Rheumatology criteria for the classification of SLE ([13]), including hematologic involvement, renal involvement (class III–V lupus nephritis), and antinuclear antibody and anti–double-stranded DNA positivity. None of them had recurrent urticaria, abdominal pain, or opthalmologic findings compatible with HUVS. The followup period was 8–30 months; disease remission was complete in 2 patients, and disease was in partial remission in the other 2 patients.

Table 1. Clinical and laboratory findings in the 2 families with HUVS*
 Family 1Family 2
IV-1IV-4IV-5IV-1IV-2
  1. HUVS = hypocomplementemic urticarial vasculitis syndrome; SLE = systemic lupus erythematosus; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; NA = not available; anti-dsDNA = anti–double-stranded DNA; ANA = antinuclear antibody; ANCA = antineutrophil cytoplasmic antibody; GN = glomerulonephritis.

Age at disease onset, years62263
Followup period10 months6 years4 years13 years4 years
OutcomeDied of HUVS; no SLE developmentDeveloped SLE at age 5.5 years; SLE is in remission; HUVS is activeOngoing active HUVS; no SLE developmentDiagnosis of HUVS plus SLE at age 7.5 years; now in remissionDiagnosis of HUVS plus SLE at age 3 years; died due to active SLE crisis
Clinical findings
Recurrent urticaria for >6 months+++++
Recurrent abdominal pain+++
Fatigue+++++
Fever++++
Uveitis++
Arthritis/arthralgia++++
Lymphadenopathy++++
Glomerulonephritis++++
Laboratory findings
Continuous [UPWARDS ARROW]ESR and [UPWARDS ARROW]CRP+++++
Anemia+++++
Thrombocytopenia++
Thrombocytosis++
[DOWNWARDS ARROW]C3+++++
[DOWNWARDS ARROW]C4+++++
[DOWNWARDS ARROW]C1qNA++NANA
Anti-C1q antibody positiveNA++NANA
Anti-dsDNA positive+++
ANA positive+++
ANCA positive++NANA
Biopsy findings
Leukocytoclastic vasculitis, skinNA++++
Lupus nephritis, kidneyClass IIIClass IIClass II
Crescentic GN, kidney+
image

Figure 1. A, Pedigrees of the 2 studied families and segregation of the DNASE1L3 mutations. Solid symbols represent affected subjects. +/+ = homozygous wild-type; +/− = heterozygous; −/− = homozygous mutant. B, Clinical and biopsy findings in affected children. Left, Uveitis and urticarial rash in an affected child. Middle, Hematoxylin and eosin (H&E)–stained punch biopsy specimen of an urticarial lesion, showing leukocytoclastic vasculitis. Original magnification × 200. Right, H&E-stained renal biopsy specimen showing focal segmental glomerulonephritis and severely inflamed interstitium compatible with class III lupus nephritis. Original magnification × 40.

Download figure to PowerPoint

Autozygosity mapping

DNA was extracted from blood and a kidney biopsy specimen obtained from a deceased child with HUV, using standard methods. Genome-wide SNP genotyping was performed with Affymetrix 5.0 arrays in 2 affected siblings and their unaffected parents in family 1 (IV-4, IV-5, III-1, and III-2). The DNA samples were processed according to Affymetrix protocols. Before analysis, overall sample call rate, sex consistency checks, relationship inference (using a Graphical Representation of Relationships program) ([14]), and Mendelian inconsistency rates were used for quality assessment. Copy number variants (CNVs) were assessed by determining the relative loss or gain of fluorescence signal intensity from SNP or CNV probes on the array using the PennCNV approach ([15]). Regions of autozygosity from the SNP screening were first sought using Plink ([16]). Haplotypes were constructed for the shared homozygous genomic segments in affected children, and their cosegregation with the phenotype was assessed visually.

DNA sequencing

Genomic DNA obtained from a blood sample from one of the affected individuals in family 1 (IV-4) was used to sequence the whole exome. An Agilent SureSelect Human All Exon 50Mb Kit was used for in-solution enrichment of coding exons and flanking intronic sequences, according to the manufacturer's standard protocol. Adapter sequences for the Illumina HiSeq 2000 were ligated, and the enriched DNA samples were subjected to standard sample preparation for the HiSeq 2000 instrument. An Illumina CASAVA v1.8 pipeline was used to produce 99-bp sequence reads. A Burrows-Wheeler Aligner (BWA) algorithm ([17]) was used to align sequence reads to the human reference genome (hg19), and variants were called using GATK software ([18, 19]).

All variants were submitted to SeattleSeq Annotation 134 for further characterization. The Genomes Management Application (GEM.app) tool ([20]) was used for filtering of variants. The variants were filtered according to autosomal-recessive inheritance (both homozygous and compound heterozygous); the variant function class including missense, nonsense, splice sites, in-frame indels and frame-shift indels; and <0.01 minor allele frequency at the dbSNP137 and NHLBI (http://evs.gs.washington.edu/EVS/) databases. Variants were also filtered for absence of the homozygous state in our internal database that includes >100 ethnicity-matched exomes that have been obtained through various studies for the identification of Mendelian disease genes. These ethnicity-matched individuals do not have SLE or any other autoimmune disease.

The Sanger sequencing method was used to confirm a candidate variant obtained through whole-exome sequencing as well as to evaluate cosegregation of the identified variant in all families. All exons of the gene containing the variant were sequenced in other families. Polymerase chain reaction (PCR) products were visualized on agarose gels, cleaned over Sephadex columns, and applied to BigDye reaction mix according to the manufacturer's recommendations (Applied Biosystems). An ABI 3730 DNA Sequencer (Applied Biosystems) was used to detect the mutation. Sequences were analyzed using Mutation Surveyor software v3.20 (SoftGenetics).

RNA analyses

RNA was extracted from the blood of homozygotes and heterozygotes in families 1 and 2, using a PAXgene RNA kit (Qiagen) according to the manufacturer's protocol. Complementary DNA (cDNA) was synthesized with SuperScript III Reverse Transcriptase (Invitrogen). PCR primers were designed with Primer3, using reference sequence NM_004944.3. Amplification of cDNA was performed with an Applied Biosystems 7900HT Fast Real-Time PCR System. The quantitative PCR protocol was as follows: 2 minutes at 50°C and 10 minutes at 95°C, followed by 40 cycles of 15 seconds at 95°C and 1 minute at 60°C. PCR mixtures were prepared with Fast SYBER Green Master Mix (Applied Biosystems) in accordance with the manufacturer's protocol. The expression of DNASE1L3 was normalized to that of GAPDH, and each experimental condition was performed in quadruplicate.

DNase activity assay

The c.320+4delAGTA mutation identifed in this study was cloned from cDNA and transfected into COS-7 cells. DNase activity was assessed with a plasmid nicking assay using supercoiled pBR322, as previously described ([21]). Comparisons were made between wild-type and mutant proteins. A positive control was also included.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

Identification of DNASE1L3 mutations

Autozygosity mapping in family 1 revealed a 10.9-Mb region on chromosome 3, from 49,685,073 bp (rs2329020) to 60,625,101 bp (rs2856056) (Hg19), that was homozygous in both affected siblings and heterozygous in the parents. A whole-exome sequencing experiment produced 86,215,534 reads in individual IV-4. When measured at a minimum depth of 8×, 85% of the target region was covered, with an average depth of 64×. Likewise, when measured at 1× and 20× coverage, 97% and 70%, respectively, of the intended target was covered. In terms of variant calls, the BWA software predicted 24,837 variants, either homozygous or heterozygous. The ratio of transitions to transversions was 2.223. By filtering as described above, only one homozygous variant mapped to the autozygous region on chromosome 3 (chr3:58191226 according to Hg19) corresponding to c.289_290delAC (NM_004944.2), causing reading-frame shifts and leading to truncation (p.Thr97Ilefs*2) of DNASE1L3. Sanger sequencing showed cosegregation of the 2-bp deletion in the entire family (Figures 1A and 2A). The identified mutation was absent in the dbSNP137 and the NHLBI Exome Sequencing Project databases, and in >200 chromosomes from Turkish control chromosomes.

image

Figure 2. A, Left, The c.289_290delAC mutation identified in family 1. The 2 deleted nucleotides are shown in red. Right, Results of quantitative polymerase chain reaction showing the relative quantity of DNASE1L3 RNA for family 1. Bars show the mean ± SD. B, Left, The c.320+4delAGTA mutation identified in family 2. Middle, Agarose gel and Sanger sequencing of cDNA showing skipping of exon 3 in family 2. Right, Plasmid nicking assay of wild-type (WT) and mutant DNASE1L3 using supercoiled pBBR322 plasmid as substrate. Lane 1, pBR322 plasmid; lane 2, undiluted medium from cells transfected with WT DNASE1L3; lane 3, 1:10 dilution of medium from cells transfected with WT DNASE1L3; lane 4, undiluted medium from cells transfected with c.320+4delAGTA DNASE1L3; lane 5, 1:10 dilution of medium from cells transfected with c.320+4delAGTA DNASE1L3; lane 6, Eco RI–digested pBR322. DNA size markers (M) are shown at the far left and far right. The top 3 bands are 20,000, 10,000, and 7,000, followed by a brighter 5,000-bp band. Arrows indicate mutation points; broken vertical lines show exon–intron and exon–exon junctions. HET = heterozygous; HOM = homozygous. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/doi/10.1002/art.38010/abstract.

Download figure to PowerPoint

We subsequently sequenced all exons of DNASE1L3 via the Sanger method in the 1 affected child identified in 4 additional families. The experiment revealed a 4-bp deletion at the splice donor site of intron 3 (c.320+4delAGTA) that cosegregated in the family 2 members with HUVS (Figures 1A and 2B). This mutation was absent in the dbSNP137 and NHLBI Exome Sequencing Project databases and in >200 Turkish control chromosomes. No DNASE1L3 mutation was detected in the remaining 3 families, in which the affected children had a diagnosis of SLE only.

Loss of DNASE1L3 function caused by the identified mutations

Quantitative PCR analysis of the cDNA samples obtained from homozygote, heterozygote, and wild-type members of family 1 showed severely reduced levels of DNASE1L3 in homozygotes and a moderate reduction in heterozygotes, as expected due to nonsense-mediated messenger RNA decay that is triggered by the premature termination codon (Figure 1A). These analyses did not show a difference between homozygotes, heterozygotes, and wild-type individuals with the c.320+4delAGTA mutation identified in family 2 (data not shown). In silico analysis via BDGP Splice (www.fruitfly.org/seq_tools/splice.html) and NetGene2 Splice (www.cbs.dtu.dk/services/NetGene2/) tools suggested that the 4-bp deletion identified in family 2 compromised the splice donor site following exon 3 of DNASE1L3. Amplification of exons 2–4 from the cDNA of family 2 showed that the splice site mutation resulted in the skipping of exon 3 (Figure 2B). Sanger sequencing of the homozygous mutant cDNA confirmed that the 90-bp exon 3 was being entirely skipped (Figure 2B, middle). Exon 3 codes for a catalytic domain of the DNASE1L3 enzyme ([18]). An assay of DNase activity showed that the c.320+4delAGTA mutation leads to the lack of DNA nicking activity of DNASE1L3 (Figure 2B).

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

This study is the first to demonstrate that mutations in DNASE1L3 cause HUVS. The identified mutations lead to loss of function via nonsense-mediated decay or exon skipping, resulting in a missing functional domain in the protein. It is important to note that the 5 children with DNASE1L3 mutations in this study presented with HUVS at a very young age. SLE was present in 3 of these 5 children (Table 1). We did not observe DNASE1L3 mutations in children from 3 other families who had autosomal-recessive isolated SLE with multiple affected family members and/or parental consanguinity. Another loss-of-function mutation in DNASE1L3 has recently been reported to cause autosomal-recessive SLE ([21]). Surprisingly, clinical findings in the 6 families described in that study did not include HUVS. All of those patients had pediatric-onset SLE (in 13 of the 17 patients, disease onset occurred at age 6 years or younger) and low C3 and C4 levels. Four of the 6 families had ANCA positivity, and lupus nephritis was frequently detected in patients with that mutation ([21]).

Although the genetic basis of chronic urticaria and urticarial vasculitis has remained largely unknown, SLE is a complex autoimmune disease caused by a combination of genetic susceptibility and environmental factors. Genome-wide association studies have identified >30 associated loci in patients with SLE. The clinical heterogeneity of SLE is mirrored by the diversity of the pathways reported to contain the associated loci identified by the genetic studies, including apoptosis, innate immune response, ubiquitination, and phagocytosis ([22]). The hallmark of SLE is the formation of autoantibodies against structures that are in the interior of the cells, especially nucleosomes.

Nucleosomes are formed through cleavage of chromatin by nucleases during apoptosis. Abnormal apoptosis and deficiency in the clearance of apoptotic cell debris are of capital importance in the loss of immune tolerance ([23]). Although rare mutations that cause SLE in a Mendelian manner account for a very small fraction of SLE cases, they are extremely helpful in identifying pathogenetic pathways in this complex disease ([22]). The protein encoded by DNASEIL3 is one of the human homologs of DNase I and functions as an endonuclease capable of cleaving both single- and double-stranded DNA ([21, 24]). In 2000, Napirei et al ([25]) generated DNase I–deficient mice by gene targeting. These mice show the classic symptoms of SLE. DNase I is suggested to be responsible for the removal of DNA from nuclear antigens at sites of high cell turnover and thus for the prevention of SLE. In 2001, Yasutoma et al ([26]) described 2 SLE patients with a heterozygous nonsense mutation in DNASE1, and both patients had decreased DNASE1 activity.

The debate continues as to whether HUVS is a separate entity or a rare subset/unusual type of SLE. Some investigators state that the findings for HUVS are atypical for SLE, and that HUVS is to be regarded as a separate entity based on the absence of classic serologic test results in most cases. Furthermore, the skin lesions and pathologic states appear to be different ([27]). In contrast, other investigators state that HUVS resembles SLE both clinically and immunologically, and that it is accepted as an SLE-associated syndrome ([28]).

The autoantibody that binds only to the collagen-like region of C1q was observed in all tested patients with HUVS (100%) and was also detected in 30–35% of patients with SLE. The absence of anti-C1q antibody in a random selection of patients with other rheumatic diseases supports the relative specificity of anti-C1q antibody for HUVS and SLE ([29]). In both diseases, the autoantibody binds to the same collagen-like region epitopes of C1q ([30]). It is proposed that antiC1q antibodies form immune complexes and initiate the disease process.

These findings support the notion that HUVS is an autoimmune syndrome related to SLE. In addition, HUVS is present in 7–8% of patients with SLE, and SLE is observed in ∼50% of patients with HUVS during followup ([10, 28]). Our results strongly suggest that the pathogenesis of SLE and HUVS is related, and that DNASE1L3 mutations cause a familial form of HUVS and HUVS associated with SLE. Although none of the 5 affected individuals in the 2 families reported here presented with isolated SLE, and all had diagnostic findings of HUVS, further studies are needed to reach firm conclusions regarding the relationship between HUVS and SLE.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

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. Tekin 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. Özçakar, Kasapcopur, Fan, Yalçinkaya, Tekin.

Acquisition of data. Özçakar, Foster, Diaz-Horta, Kasapcopur, Fan, Yalçinkaya, Tekin.

Analysis and interpretation of data. Foster, Diaz-Horta, Fan, Tekin.

Acknowledgments

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES

We are grateful to the participating families and to Juan Young, PhD (University of Miami) for his assistance in laboratory analyses.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgments
  8. REFERENCES