German WG patients were clinically diagnosed according to international standards by applying the 1990 classification criteria of the American College of Rheumatology 29 and the definitions of the 1994 Chapel Hill Consensus Conference 3. All patients have defined ANCA-status (n_positive = 151; n_negative = 30). WG was biopsy-proven in the German reference centre for vasculitis (Department of Pathology, University of Schleswig-Holstein Campus Luebeck, Germany) by 2 different observers. Control groups comprised healthy adults from northern Germany and the Ruhr area (Germany). The institutional Ethics Committee of Bochum Ruhr-University approved this study. Informed consent was obtained from all individuals.

DNA pooling was performed as described before 20. Briefly, DNA concentrations were measured spectrophotometrically in 4 steps, and samples were diluted accordingly. The final DNA concentration in pools was 50 ng/μl. In order to control for artefacts during pooling, 150 ANCA-positive patients were divided into 3 sub-pools and 100 controls from the northern German cohort into 2 sub-pools, respectively.

A panel of 94 markers was designed encompassing the WG associated 6p21.32 and flanking regions (annotated by Santa Cruz genome browser (UCSC): chro6:31255920-34938760, May 2004 freeze, see 30). Microsatellites were chosen utilizing the simple-repeat option of UCSC. The average distance between microsatellite markers was 41 kb (standard deviation ~30 kb) with five gaps of distances ≥100 kb (<152 kb). Most of the microsatellites harboured dinucleotide repeats (68%), the remaining comprised tri-, tetra- or penta-nucleotide repeat motifs. Further information about the microsatellites is available as additional file 1 (website: 31). No marker revealed significantly differing 'intra sub-pool' distributions as reflected by the p value (see below).

For fragment analysis we used fluorescence 5'FAM labelled, tailed oligonucleotide added to the 5'-part of the sequence specific primer as described before 20. Primers were designed with the Primer Express 2.0 Software (Applied Biosystems, Foster City, USA). Melting temperature was set to 55°C. Three primers were used for PCR: tailed forward primer (tailed F), reverse primer and labelled primer (labelled F) corresponding to the 5'-tail sequence of tailed F. Reaction mix consisted of 1 × PCR buffer (Qiagen GmbH, Hilden, Germany), 1.5 pmol labeled F, 0.2 mM each dNTP, 3 mM MgCl₂, 0.2 pmol tailed F, 1.5 pmol reverse primer, 0.25 U Qiagen Hot Start Taq (Qiagen) and 50 ng DNA. PCR was performed in a Biometra T-Gradient Thermoblock (Biometra, Göttingen, Germany) using the following characteristics: activation step at 95°C for 15 min; 35 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min and extension at 72°C for 1 min; and a final extension at 72°C for 10 min.

Electrophoresis and data analysis were performed on a 48-well ABI377 slab-gel system using ABI Genotyper software as described before 20. Data analysis generated marker-specific allele image profiles (AIPs) representing distributions of alleles within each sub-pool.

AIPs of WG pools were compared with those of the control cohort. Peak heights were cumulated to 100% (representing 100 alleles in each pool, respectively). Afterwards allele frequencies (corresponding to peaks height) were calculated according to the estimated total allele count. Association was tested for each marker by comparison of the AIPs from WG-DNA pools with those of controls by contingency tables using a significance level of p = 0.05. In order to focus the statistics on major effects, alleles with a frequency of less than 5% were summarized to one allele. Next, frequency distributions were compared by means of contingency tables as mentioned above. The resulting p values were corrected for multiple testing using QValue software 3233 with a 'cut off' of 5%. Nevertheless, non-corrected p values were simply ranked according to their evidence for association in order to select markers for further studies. In addition, intra-population differences were tested by comparing WG pools with each other as well as both control pools.

In order to exclude false positive significances, patients and controls were individually genotyped for respective markers. Genotyping was performed on the Beckman Coulter CEQ8000 8-capillary system using 'Fragment Analysis Module' software (Beckman Coulter, Inc., Fullerton, USA). Reaction mix contained 1 × PCR buffer (Qiagen), 0.75 pmol fluorescent labelled F, 0.2 mM each dNTP, 3 mM MgCl₂, 0.2 pmol tailed F, 1.5 pmol reverse primer, 0.25 U Qiagen Hot Start Taq (Qiagen) and 50 ng DNA. Parameters for amplification were used as described above. Allele and genotype frequency of WG patients and controls were compared by χ²-testing. Additionally, Hardy-Weinberg equilibrium was tested using Genepop software 20.

The coding and promoter regions (593 bp) of the RXRB gene were initially screened by single strand conformation polymorphism (SSCP) with 48 patient (ANCA-positive) and 48 control DNAs. The RXRB sequence was downloaded from UCSC database (NM021976, July 2003 Freeze). Oligonucleotides were designed using Primer Express 2.0 Software (Applied Biosystems) with an optimal annealing temperature of 57°C (see table 1). PCR was performed using Qiagen HotStar Taq (Qiagen) with an initial denaturation at 95°C for 15 min, 35 cycles of 94°C for 30 sec -57°C for 30 sec -72°C for 30 sec and a final extension at 72°C for 10 min. PCR conditions were chosen as recommended by Qiagen. Fragments were labelled with α-dATP (exons 2–10; Hartmann Bioanalytics, Braunschweig, Germany) and α-dCTP (exon 1a and 1b; Hartmann Bioanalytics). SSCP were analysed on 6% and 5% polyacrylamide gels with either 5% or 10% glycerol. The gels were run at room temperature at 35 W. Amplificates containing obvious band shifts were sequenced using dye terminator cycle sequencing on an ABI377 automatic sequencer (Applied Biosystems). Statistical comparisons of allele frequencies and genotypes were based on the χ² test. In addition, Hardy-Weinberg equilibrium has been tested for all SNPs.

The sequence variation rs6531 was genotyped in 151 ANCA-positive patients and 201 controls by RFLP analysis. A restriction site for DdeI (NewEnglandBiolabs, Ipswich, USA) was generated by using the tailed mismatch oligonucleotide, mmantisense-CATCGCTGATTCGCACATCATCAATGGATCGGTCTGA. PCR was performed by 'the tailed primer' method using three oligonulceotides: the mmantisene-primer, an oligonucleotide corresponding to the tail and the forward oligonucleotide sense-CCGATCTTTAGTGACCCCAGT. Restriction with DdeI results in a 226 bp fragment (T allele) and in case of the C allele in 2 fragments of 192 and 34 bp, respectively. Electrophoresis was done on a 3% agarose gel containing ethidium bromide. The results were analysed with the χ² test and Hardy-Weinberg equilibrium was controlled.

The previously reported the Val95Ala polymorphism 34 was genotyped by RFLP analysis in 94 patients and 90 controls. Fragments were amplified using oligonucleotides Val95Ala sense-CGGTGGGGTATTAGAGAATT and Val95Ala antisense-CCCATGGAAGAACTGATGACTGG. PCR was performed with Qiagen HotStar Taq (Qiagen) as described above with an annealing temperature of 55–60°C and different MgCl₂-concentrations generating a 300bp fragment. The 2 alleles (Val, T allele; Ala, C allele) are detected as 232 bp and 193 bp bands, respectively.

Electrophoresis was done in ethidium bromide containing 2% agarose gels. In order to control the results of the RFLP analysis, DNAs of 15 patients and 20 controls were sequenced as described above. Association was verified with the χ² test and Hardy-Weinberg equilibrium.

Primer extension analysis was used to genotype the truncating splice site mutation in 180 patients and 261 controls. Two primers were designed for amplification, forward TCCAGATACTCAGTGCCAGA, reverse: TTGTCCAGGAACTAGCATATT. PCR conditions were: 1 × PCR buffer (Qiagen), 0.2 mM each dNTP, 0.2 pmol forward primer, 0.2 pmol reverse primer, 0.25 U Qiagen Hot Start Taq (Qiagen) and 50 ng DNA. The reaction was performed with initial denaturation at 95°C for 15 min, 35 cycles of 94°C for 30 sec -60°C for 30 sec -72°C for 30 sec and a final extension at 72°C for 10 min The product was purified with magnetic beads system (AMPure; ABgene, Epsom, UK). A primer extension reaction was performed with the interrogation primer: GCCCAGTTTGGATCTGAAGGTGGTA using Beckmann CEQ™ SNP-Primer Extension Kit (Beckman Coulter) under recommended conditions. Electrophoresis and data analysis was performed using Beckman Coulter CEQ8000. Statistical analysis was based on the χ² tests, and allele frequencies were tested for Hardy-Weinberg equilibrium.

We calculated pair-wise LD between HLA-DPB1 alleles (genotypes obtained from a previous study 20, see additional file 2, website: 35), rs6531 and the significantly associated microsatellite #1.0.3.7 in our control and patient cohorts. Microsatellite, SNP and HLA-DPB1 alleles with frequencies below 0.05 were excluded. For pairwise LD calculation we presumed each microsatellite allele to be biallelic, i.e. one distinct allele against all others pooled. The same procedure was applied for the most frequent HLA-DPB1 alleles *0401, *0402, *0201 and *0301. The strength of LD between all three markers was quantified by D' and r² values. LD analyses were calculated by Haploview 3.2 36.

Eighty four markers were analysed in an initial scan (after exclusion of non-polymorphic and recurrently artefact-producing microsatellites). Thereof, five microsatellites revealed significant differences in allele distribution between WG patients and controls (see table 2). Correction for multiple comparisons with QValue rejected significance for all results. Nonetheless, markers were simply ranked according to their p values. Thereafter, four markers were selected for individual genotyping (see figure 1). One microsatellite showed only marginal significance (#1.0.3.4) after summation of rare alleles (<5%) and thus was excluded from further investigations. Individual genotyping confirmed exclusively one marker (adjacent to RXRB), the positive control, as statistically significantly different: #1.0.3.7 (p = 3 × 10^-4, see table 2).

SSCP analyses in the RXRB gene did not reveal any variations in exons 1–5, 8–10 and the 593 bp promoter region. Similarly, the Val95Ala polymorphism was neither detectable by RFLP nor by sequence analyses in our patient and control cohorts. Yet, the analysis of the PCR product comprising exon 6 revealed a band shift. Subsequent sequencing showed an intronic CT-INDEL polymorphism which has been reported before (rs10548957). Statistical analysis of the allele frequencies revealed a p value of 0.02 with a minor allele frequency of 0.07 in the control cohort (n = 47) and total absence in the patients (n = 43). In exon 7 a single nucleotide polymorphism (SNP) (rs6531) was detected using SSCP and sequencing. Comparison of allele frequencies of patients (n = 151) vs. controls (n = 201) revealed significant p values (see table 3 for detailed analyses). Additionally 28 ANCA-negative patients were genotyped for this SNP and compared to either the ANCA-positive patients or the control cohort. There is a significant deviation between the ANCA-negative and -positive patients regarding this SNP. The ANCA-negative cohort did not differ from the control group.

Several alleles between HLA-DPB1, rs6531 and microsatellite marker #1.0.3.7 showed LD with r² values exceeding 0.10 (for detailed description see figure 2A and 2B). For example, strong LD exists between HLA-DPB1*0401 and microsatellite allele *3 as well as the rs6531 polymorphism in the patient and control cohorts.

Genotyping of rs2076530 polymorphism in 180 patients and 261 controls (Northern Germany and Ruhr area populations) did not reveal any statistically significant difference in allele, phenotype and genotype frequencies between all WG patients and healthy individuals (see table 4). In addition, when dividing the patient cohort into ANCA⁺ and ANCA^-patients, significant differences were not detected between the respective patient and control cohorts (data not shown).