Systemic lupus erythematosus (SLE) is an important systemic autoimmune disease characterised by the presence of autoantibodies binding double stranded DNA (henceforth dsDNA ab) and decreased levels of serum complement components C3 and C4 [1]. It has been postulated that defective apoptotic clearance of immune complexes incorporating histone associated DNA predisposes to SLE [1]. DNASE I, is an endonuclease [2] believed to facilitate the removal of cellular debris thereby preventing aberrant inflammation [3-5]. There have been reports that DNASE I deficiency is a predisposing factor for SLE. Lower levels of DNASE I enzyme activity may be associated with lupus nephritis [6,7] and correlate with disease activity state and clinical parameters [8]. Yasuda and colleagues [2,9-12] and Iida et al. [13] have identified exomic polymorphisms of the DNASE I gene, some of which appear to be associated with susceptibility to SLE [14]. Furthermore, Koji and colleagues [15] have described an A-G transversion in exon two that results in defective synthesis of DNASE I in an SLE patient manifesting lower DNASE I activity. The DNASE I gene also contains a polymorphic 56bp variable number tandem repeat (VNTR-HumDN1) locus within intron four manifesting six alleles corresponding to molecular product sizes of 469, 525, 581, 637 and 693 for two, three, four, five and six repeats respectively [16-18]. Yasuda and colleagues [16] reported that the three repeats allele of VNTR (HumDN1) was in linkage disequilibrium with DNASE I*1 allele whilst the VNTR alleles four and five were in linkage disequilibrium with DNASE I*2.

In this study, genetic polymorphisms at both the DNASE I locus and at the VNTR (HumDN1) locus were typed in a Western Australian population of SLE patients and controls with known serum concentrations of DNASE I protein. The genetic diversity of DNASE I polymorphisms in these cohorts was compared with other studies and associations sought with disease and serological parameters indicative of SLE pathology.

Patients, controls and ethics

Fifty-six (56) treated systemic lupus erythematosus patients aged 47 ± 15 yrs. were recruited between March 2014 and December 2015 from Sir Charles Gairdner (SCGH) and Royal Perth Hospitals (RPH-Perth, Western Australia). The diagnoses were based upon the American College of Rheumatology (ACR) classification criteria. In addition to the SLE diagnosis, patients had co-morbidities including renal disease, liver pathologies as well as Sjögren and Raynaud’s syndromes. SLE patients in this cohort were treated with various therapeutic regimens including steroids (low dose prednisolone (≤ 5 mg/day)), immunosuppressant therapies (azathioprine, cyclosporine, cyclophosphamide, methotrexate, mycophenolate, mofetil) and antimalarial therapies (hydroxychloroquine). Thirty-three age and sexmatched healthy control subjects (47 ± 15 yrs.) were recruited from the Perth metropolitan area. Human ethics approval: Curtin University, Sir Charles Gairdner Hospital and Royal Perth Hospital Human Ethics Committees approved the human studies (approval numbers HR 202/2013, HREC 2013-174). Written and informed consent was obtained from all study participants.

Sample collection

Peripheral blood from SLE patients was collected at one time point during the course of their disease monitoring and management into ethylenediaminetetraacetic acid (EDTA), heparin and serum tubes at PathWest collection centres. Serum samples were separated at PathWest laboratories within one hour of collection using centrifugation and frozen to between -70°C and -80°C. Samples from healthy human controls were also collected into EDTA, heparin and serum vacutainer tubes (Beckton Dickinson (BD), San Jose, USA) and processed in the same fashion at the Curtin University laboratory (Perth, WA).

Identification of phenotypes and genotypes at the DNASE I and intron 4 VNTR (HumDN1) loci

Stored DNA samples were used in a PCR reaction using MyFiTM DNA polymerase (Bioline, USA) to amplify exons (or part of) two, five, six, seven and eight using primers sets (from GeneWorks, Australiatable S4). The DNASE I gene was detected by PCR amplification using primers as indicated in Supplementary Table S4 in PCR microfuge tubes containing 1 μL of test DNA, together with 5 μL of 5x MyFi reaction buffer (Bioline, USA), 0.5 L of each forward and reverse primers 10 μM and 1 μL of MyFi DNA polymerase (Bioline, USA) as well as 17 μL of nuclease free water for a total volume of 25 μL per reaction. PCR assays were performed with cycle parameters set at 95°C for 1 min (1 cycle) and 35 cycles consisting of 95°C for 15 s followed by 61°C for 15 s (DNASE I) or 57°C for 15 s (VNTR-HumDN1).

PCR products were then sequenced (MACROGEN, Seoul, Korea) to identify exonic alleles of Human DNASE I as indicated in Table 1.

Table 1: PCR products sequenced to identify exonic alleles of Human DNASE I.

The number of VNTR (HumDN1) repeats within intron 4 of the Human DNASE I was determined using a PCR reaction with MyFi™ DNA polymerase (Bioline, USA) to amplify intron four spanned by exons four and five using the primers pair shown in Supplementary Table S4 (GeneWorks, Australia). The lengths of the amplicons were then estimated on a 4% agarose gel electrophoresis using a 1kb DNA ladder. As a control measure, five samples were sent for sequencing (details as above) and the results compared with the gels. Alleles manifested three, four, five and six repeats with corresponding molecular weights of 525, 581, 637 and 693bp respectively [19].

DNASE I enzyme concentrations

A sandwich ELISA kit (Creative Diagnostics, USA) was used to assay human DNASE I serum concentrations per manufacturer’s instructions. Briefly, stored patient sera were incubated in microtiter wells coated with a DNASE I specific antibody. DNASE I present in the samples was immobilized and unbound products removed. Biotin labelled DNASE I specific antibody bound to immobilized DNASE I was detected with avidin conjugated Horseradish Peroxidase (HRP). Test sample concentrations were interpolated using a four-parameter logistic calibration curve fitted to the standard sera provided by the manufacturer (100, 50, 25, 12.5, 6.25, 3.12, 1.56 and 0 ng/mL).

Other serological parameters associated with SLE

We have previously reported the methods for assaying C1q, C3, and C4 concentrations in the serum of the SLE and control cohorts [20].

Data analysis

Nucleotide and peptide sequences were analysed using Geneious version 10 [21]. Linkage disequilibrium and population parameters were determined using “Genepop on the web” [22,23], LinkDos [24] and SNPStats [25]. Statistical significance was calculated using Mann- Whitney rank sum U-test with the GraphPad PRISM 7 application (GraphPad Software Inc., California, USA). P values of

DNASE I protein concentrations in SLE patients and controls

DNASE I protein concentrations observed in serum samples of SLE patients and controls are shown in Figure 1. The distribution of concentrations in both cohorts was skewed; the mean concentration of DNASE I protein in SLE patients was lower than that of the controls (44.2 U/mL vs. 56.4 U/mL) although the difference was not significant. A preliminary investigation of associations between DNASE I serum concentrations and co-morbidities and drug therapies in the SLE patients was undertaken using heat maps is shown in Figures 2C and 2D respectively.

Figure 1: Serum DNASE I concentrations. A: purple bars-the frequency distribution (U/mL) in the SLE (mean=44.2 U/mL) and light grey bars-control (mean=56.4 U/mL) populations. B: Ranks of serum DNASE I concentrations in SLE and control cohorts showing means ± standard deviation-not significantly different.

Figure 2: Heat maps showing associations between mean concentrations of serum DNASE I in SLE patients stratified in three groups defined as low concentration range, mid-concentration range and high concentration range with mean concentrations of DNASE I=29 U/mL top row, DNASE I=59 U/mL middle row, and DNASE I=103 U/mL bottom row respectively. Figures 2A and 2B: DNASE I concentrations versus DNASE I genotypes and VNTR (HumDN1) repeats genotypes respectively. Figures 2C and 2D: DNASE I concentrations in SLE patients versus co-morbidities and drug therapies respectively.

Allele and genotype frequencies at the DNASE I and intron 4 VNTR (HumDN1) loci

Allele frequencies at the DNASE I exomic and VNTR (HumDN1) loci are summarised in Table 2.

Table 2: Allele frequencies for the DNASE I locus and the intron 4 VNTR (HumDN1) locus in Australian SLE patients and healthy controls (numbers in brackets indicate the number of alleles in each group).

Alleles at both intragenic loci exhibited Hardy-Weinberg equilibrium and linkage disequilibrium. The frequencies of haplotypic combinations at both the DNASE I and VNTR (HumDN1) loci are shown in Figure 3. Haplotypes with VNTR alleles of 4, 5 and 6 repeats

Figure 3: DNASE I (left number) & VNTR (HumDN1) (right number) Loci Haplotypes. The figure shows percentage haplotype frequencies in both loci for the SLE patients and control cohort.

were more prevalent in both the SLE and control cohorts.

Ethnic variation in DNASE I polymorphisms

The allele frequencies reported in this study are in general agreement with the metastudy of ethnic allele frequency variation shown (Table 3). Only the Ovambos’ of African origin have higher frequencies of the DNASE I*1 relative to the DNASE I*2 allele. No apparent significant differences between DNASE I allele frequencies were observed for either the SLE patients or the healthy controls. A comparison of ethnic variation in the frequencies of the internal VNTR (HumDN1) locus within the different ethnic groups is shown in Table 4. The frequencies of the six alleles reported in this study were similar to those observed in other population with the exception that the long six repeat allele was more frequent in both the control and SLE populations reported herein. The genetic differentiation of the Ovambo and Japanese populations was evident with the three-repeat allele being more prevalent than in the other populations reported.

Table 3: Frequency data for the DNASE alleles *1 and *2 are derived from published reports. The rare DNASE I*3; *4; *5 and *6 alleles are not included due to insufficient published data. “Mixed other” control group contained individuals of Middle Eastern and Indian ethnicity whereas the “Mixed other SLE” group included individuals from African, South American and Australian Indigenous ethnicities which were grouped together as there were only a small number of individuals in each.

Table 4: Frequency data for the VNTR (humDN1) number of repeats are derived from referenced published reports. The rare DNASE I*3; *4; *5 and *6 alleles are not included due to insufficient data. “Mixed other” control group contained individuals of Middle Eastern and Indian ethnicity whereas the “Mixed other SLE” group included individuals from African, South American and Australian Indigenous ethnicities which were grouped together as there were only a small number of individuals in each.

Are serum DNASE I concentrations dependent on genotype?

DNASE I serum concentration distributions are shown ranked by genotype in Figure 4.

Figure 4: Shows the mean concentrations (U/mL ± SD) of the DNASE I protein in SLE patients (A & C) and controls (B & D) associated with DNASE I phenotype DNASE I*1 or DNASE I*2 (A and B) and with the VNTR (HumDN1) number of repeats (C and D).

There were no significant associations between DNASE I serum concentrations and either the DNASE I exomic genotypes or the repeat number variation of the VNTR (HumDN1) locus. The lowest concentration levels of serum DNASE I were associated with DNASE I*2 genotype, and with the higher number of repeats at the VNTR (HumDN1) locus.

Associations between serological parameters and DNASE I and VNTR (HumDN1) alleles

A correlation matrix was generated for age, ethnicity, [C1q], [C3], [C4] and [DNASE I] versus VNTR (HumDN1) and DNASE I alleles (Supplementary Table S1). No significant correlations between the various parameters were observed except for serum concentrations of DNASE I and C1q complement protein (r=0.653; p=0.05). In the SLE cohort, the VNTR (HumDN1) repeat number was significantly correlated with [C3] and [C4] (0.304; p=0.03 and 0.279; p=0.04 respectively). However, specific alleles at both the VNTR (HumDN1) and DNASE I loci were significantly correlated in the SLE and control cohorts (r=0.760; p=0.0001; 0.569; p=0.0005 respectively) (Supplementary Tables S2 and S3).

In the absence of well-defined associations between specific DNASE I genotypes and SLE, heat maps were used to visualize trends in disease severity that may be related to DNASE I genotypes or serum concentrations in both the SLE patient and control cohorts. The heat map analyses (Figure 2) show qualitatively that SLE patients with higher numbers of VNTR (HumDN1) repeats (≥ 5) had lower serum concentrations of DNASE I protein and required more intensive drug regimens for management of their disease than persons with low numbers of repeat motifs.

In this investigation only the two most frequent (i.e., DNASE I*1 and DNASE I*2) of the six exonic alleles defined at the DNASE I locus were observed, although the typing system (shotgun sequencing) would have detected them if present. No variations in the nucleic acid sequence within exons two, five and seven were present in any of the individuals tested. The remaining four alleles (DNASE I*3, *4, *5 and *6) are rare and there are few reports of their frequencies in the other populations [26]. The frequencies of the DNASE I*1 and *2 in persons of European and Asian ethnicity were similar to previously published reports [26-28]. Four alleles were observed at the VNTR locus with the number of repeats varying from three to six. A greater frequency of the longest six repeat allele was observed in both cohorts in this study which may reflect improved typing techniques. Only in the Ovambo and Ghanian populations did allele frequencies at both loci differ markedly from the values reported in other populations. This is consistent with an ancestral version of the DNASE I gene being present in modern day African populations [18].

Allele frequencies at both the DNASE I and VNTR (HumDN1) loci manifested Hardy-Weinberg equilibrium despite the relatively small sample size of the populations typed. As would be expected, both intragenic loci exhibited linkage disequilibrium (LD) as has been previously reported [16,37].

Of the 36 possible genotypes comprising observed alleles at these two loci, 13 were observed in the SLE patients and 12 in the control group. These data demonstrate the presence of a restricted number of haplotypes in our populations. Al Fadhli and co-workers have shown that higher number of the VNTR (HumDN1) repeats are a risk factor for SLE and that the allele with five repeats was associated with a reduction of DNASE I enzyme activity [17]. It was not possible to measure the DNASE I enzyme activity in this study, however serum protein DNASE I concentrations were assayed in both populations. All persons tested exhibited DNASE I protein in their serum. Significant differences in serum DNASE I protein concentrations between SLE patients and the control cohort were not observed, although the mean concentration in the SLE cohort was slightly lower than in the control group. If DNASE I activity is proportional to the concentration of DNASE I protein it is unlikely that non-expressed mutant genes were present in the cohorts investigated.

Associations between the VNTR (HumDN1) locus and complement C3 and C4 serum concentrations were observed in the SLE cohort but not in the control group. Hence, the number of VNTR (HumDN1) repeats in the SLE patients may influence clinical parameters. AlFadhli and co-workers have shown that higher number of the VNTR (HumDN1) repeats are a risk factor for SLE, and moreover, that the five-repeat allele was associated with a reduction of DNASE I enzyme activity [17].

The heat map analysis (Figure 2) showed qualitatively that SLE patients with higher numbers of VNTR (HumDN1) repeats (≥ 5) had lower serum concentrations of DNASE I protein than the controls and required more intensive drug regimens for management of their disease. However, quantitative analysis did not reveal any significant correlations between the serum concentrations of DNASE I protein and either phenotypes at the DNASE I and VNTR (HumDN1) loci or with disease severity (Figure 4). Other reports [17,36] in particular that of Al Fahdi [17] show that a high number of VNTR repeats are a risk factor for SLE and may cause decreased DNASE I activity. Positive correlations between DNASE I and C1q serum concentrations and the length of the VNTR allele with serum concentrations of complement components C3 and C4 were observed and noted. If low DNASE I activity and/or concentration impairs removal of apoptotic material thereby predisposing to SLE, it might be expected that VNTR alleles negatively correlate with C3, C4 and C1q serum concentrations in SLE patients. Hence this report does not support the hypothesis that specific two locus DNASE I genotypes predispose to SLE in the Western Australian cohort.