1
Supplementary Methods
Linkage analysis and inbred strain genotyping
An initial genome screen for NTN QTLs was performed on 92 F2 animals at the extreme of the phenotype distribution for crescent formation. Genotypes were generated from spleen DNA for 128 microsatellites spread across all autosomes and the X chromosome, using fluorescently labelled primers separated on an ABI 3700 (Applied Biosystems) and scored using Genotyper 2.0 (Applied Biosystems). Animals at either extremes were treated as affected and unaffected and microsatellite markers with genotypes showing P0.01 on chi square were used to genotype all 177 F2 animals. Around 10% of markers showed heterozygosity in the Lewis parental strain due to apparent contamination of the Lewis breeding stock. These markers were discarded from the linkage analysis. Multipoint linkage analysis was carried out with QTL Cartographer1. Fcgr3 was placed on the genetic linkage map using exon 5 genotypes from the entire F2 cross analysed by Mapmaker QTL2. Genetic variance attributable to individual QTLs was extracted from multipoint linkage analysis using Mapmaker QTL. Narrow sense heritability was calculated as the ratio of the genetic to phenotypic variance.
Fcgr3 exon 5 genotype was determined by PCR and agarose gel electrophoresis on the following rat strains: SHR (Charles River, UK); ACI, AS (cryo), BDII, BDIX, BDV, BH, BN, BS, DA, E3, F344, LE, LEW-a2, LOU/C, MNS, MWF, NAR, OM, WC, WF, EXBH, LEW-ci2, SPRD-Cu3 and SPRD-zi (all donated by H Hedrich, Hannover); SHRSP (by A Dominiczak, Glasgow); SS rats (by J Rapp, Ohio); and SHR/Kyo, WTC, WKYO, DON, and IS/Kyo (by T Serikawa, Kyoto). The PCR primers used to amplify exon 5 genomic DNA were 5'-GTCCCTAAATTCTGAATTTC-3' and 5'-AAAGAAGTCACAGAAAGGAG-3'.
Reverse Transcription
RT-PCR of Fcgr3-rs exon5_226- was performed with primers 5'-CGGCATCCACCAGCTCTCTAGTTGG-3' and5'-CCCAGAGCATCATGTGTCCTGGT-3'as described3.
Sequencing
Column-purified PCR products were directly sequenced on both strands using BigDye sequencing kit on an ABI Prism 3700 DNA Sequencer. PCR products were cloned using the p-GEM-T Easy Vector System II kit (Promega) and extracted DNA from selected clones sequenced with T7 and SP6 p-GEM primers.Presence of the G129 variant was verified by direct sequencing and by sequence analysis of cloned cDNA, genomic DNA and gel-purified PCR products.
Southern Analysis
The probe for the Southern analysis of rat Fcgr3 was created from gel-extracted Fcgr3 exon5_226- amplified from Lewis genomic DNA using primers 5'-TGAGTCCAGGACAGCTCCAGAGC-3' and5'-TGGTGATCCTCCTGCCTCAGCC-3'. The probe for Southern analysis of human FCGR3A and FCGR3B was amplified from human genomic DNA using primers 5'-CTGGATTTGCAACCCCATCAT-3' and 5'-CCACTACGGCTAATGTTCTA-3' and column purified to yield a 520 bp product from the 3' untranslated region that was 96% identical betweenFCGR3A and FCGR3B.
Haplotype map
Genotypes and a subsequent haplotype map of the Ntn1 QTL were generated from genomic DNA of several rat strains (Fig. 3c) for 17 markers spread across the 28 Mbp region of linkage (D13Rat32, D13Rat129, D13Rat33, D13Rat133, D13Csc1, Fcgr2 (G388T), D13Csc2, D13Csc3, D13Csc4, Fcgr3 intron 4 (CCAA)n, D13Csc5, D13Mgh8, D13Rat86, D13Rat34, D13Arb15, D13Got87, D13Rat51) Five new informative microsatellite markers flanking the Fcgr3 gene were identified from the rat genome sequence (using the Tandem Repeats Finder tool – and designated with the prefix D13Csc. A further multiallelic microsatellite was identified by inspection of the rat genome sequence within Fcgr3 intron 4. The multiple peaks observed within the Fcgr3 intron 4 microsatellite always co-segregated according to parental strain in the F2 cross, supporting the presence of tandem duplication at this locus. PCR primers amplifying Fcgr3 exon 5 as above were used to identify strains that carried the exon 5 deletion ΔG129. The Fcgr2 T388G SNP was genotyped by direct sequence analysis of PCR products amplified from Fcgr2 exon 3. Primer sequences of the genotype assays developed in-house are given below. Fluorescently labelled microsatellite PCR products were separated on an ABI 3700 or 3730XL (Applied Biosystems) and scored using GeneMapper Software v3.7.
Marker / Forward Primer / Reverse PrimerD13Csc1 / GACCCCTATGTGCCAACAGT / TCAAGATGTCGGTGCAAGTC
Fcgr2 (G388T) / AGGTTCTTGGGGCTTCCTTC / TGCAGTAAACAGATCCACCC
D13Csc2 / AGGGGAAGGAAGGAGAAACA / AGCCGGAGCTGCATAGAGTA
D13Csc3 / GTCTCCTGTGCTTGCCTTTC / AGGCAGAACTGCACCTCCTA
D13Csc4 / CCGTGCACAAAAGTTCTCAA / ATTATTGCCACCCACATGGT
Fcgr3 intron 4 (CCAA)n / GACAGCTCCAGAGCTCTGT / AGGAGCAGATGAATCAGGAA
D13Csc5 / GCCCTCTTCTGGAGTGTCTG / AGAAATGCCCATCCATTGAG
Western analysis
Lysis of cells was performed in homogenisation buffer (Tris base 0.1M, TritonX100 1%V/V, and EDTA 5mM, pH 7.5) with complete mini protease inhibitor cocktail (Roche Applied Science). Blots were blocked in phosphate buffered saline with 5% dried milk (Marvel)(PBSM) for one hour at room temperature followed by an overnight incubation at 4ºC with primary antibody in PBSM. Blots were washed twice in phosphate buffered saline in 0.05% Tween-20 (PBST) at room temperature for 5 minutes each, then incubated in secondary antibody in PBST at room temperature for one hour. Blots were washed four times at 10 minute intervals with PBST and proteins visualized by chemiluminescence with ECL Plus Western Blotting Detection Reagents (RPN2132, Amersham) and Hyperfilm ECL (RNP2103, Amersham) according to manufacturer's instructions. For detection of myc tag the primary antibody was a 1:1000 dilution of anti-myc 4A6 (05-724, Upstate) and the secondary was a 1:400 dilution of peroxidase labelled anti-mouse antibody (NA931, Amersham). For detection of Fcgr3-rs the primary was a 1:1000 dilution of rabbit antiserum raised against the ten C-terminal amino acids of the predicted cytoplasmic domain of Fcgr3-rs and the secondary was a 1:20,000 dilution of peroxidase-labelled anti-rabbit antibody (NA934vs, Amersham).
ADCC by visual assessment
Thioglycollate-elicited peritoneal exudate cells4 were added to 8-well chamber slides (Nunc) at 104 cells/well and purified by adherence. Macrophages were rested for 24 hours, labelled by incubation with green fluorescent dye di-octadecyloctacarbocyanine, washed and co-cultured with 103 rat mesangial cells (passage 5) coated with 10g ml-1 of a monoclonal IgG2a antibody to the mesangial cell antigen Thy1.15, donated by W Bagchus, Groningen. After 48 hours, death of mesangial cells was identified by nuclear morphology following staining with Hoechst 33258 and propidium iodide (10g ml-1 each). All cells were counted within every 40x objective high-powered field along the diagonal line between opposite corners of each square well.
Semi-automated calcein release assay of macrophage killing
Thioglycollate-elicited peritoneal exudate cells4(PEC) were harvested. Contaminating RBCs were removed by brief incubation in hypotonic lysis buffer (Serotec). Cells were then counted and resuspended in 10% foetal calf serum (FCS), phenol-red free DMEM. PEC and cultured rat mesangial cells were co-incubated in the presence or absence of anti-Thy1.1 antibody. 5x106 rat mesangial cells (passage 8-12) were trypsinised and re-suspended in 1ml PBS with 1g calcein (calcein–AM, Cat No C3100, Molecular Probes) and incubated at 37oC for 30mins. Cells were washed 3 times with PBS and resuspended in 10% FCS phenol-red free DMEM with antibiotics (Sigma D5921). In each well of a sterile polystyrene 96-well tissue culture plates was added: 200l 10% FCS DMEM, 10g ml-1 anti-Thy1.1 antibody; 2x104 calcein-labelled mesangial cells and 2x105 peritoneal exudate cells. The plates were incubated in 37oC, 5% CO2 and 50l supernatant from each well transferred to fresh 96 well plates for reading after 16 hours incubation. Fluorescence was measured by a multi-well fluorimeter (absorption 480nm, emission 535nm).
Creation of Fc Receptor Constructs
Fcer1g was PCR-amplified from WKY spleen cDNA, using the primer pair 5'-AAGCTTGCCACCATGATCCCAGCGGT-3'and 5'-TCTAGATTACTATTGGGGTGGTTTC-3', and then cloned into pUSEamp(+) (Upstate). Fcgr3 and Fcgr3-rs were amplified from Lewis spleen cDNA with forward primer 5'-AAGCTTGCCACCATGACTTTGGAG-3' and reverse primer 5'-tctagattaTTAGCCATACGATGGGAT-3'. Clones were re-amplified with forward primer 5'-CGGAAGCTTGCCACCATGGAGCAGAAACTCATCTCTGAAGAGGATCTGACTTTGGAGACCCAGATGTTTCAG-3' and reverse primer 5'-CCTTGAGCACCTGGATCCATGGGG-3' and recloned. Site-directed mutagenesis was performed on the Fcgr3 construct for G129 with primers 5'-GAGAAATCTTCAAACCTCGGGGAGGACTGGAGGAAATCCC-3' (forward) and 5'-GGGATTTCCTCCAGTCCTCCCCGAGGtTTGAAGATTTCTC-3' (reverse). All alpha subunit constructs were cloned into the pcDNA3.1/Hygro©(+) expression vector (Invitrogen; V870-20).
For inserting an N-terminal myc tag into the Fcgr3 constructs, 1.0µg of the Fcgr3, Fcgr3-rs and Fcgr3-G constructs were amplified with forward primer 5'-CGGAAGCTTGCCACCATGGAGCAGAAACTCATCTCTGAAGAGGATCTGACTTTGGAGACCCAGATGTTTCAG-3' and reverse primer 5'-CCTTGAGCACCTGGATCCATGGGG-3'. Constructs and PCR products were cut with HindIII and BamHI and ligated to each other.
Transfection
COS-1 cells, a gift from H Davidson, Cambridge, were maintained at 37ºC and 5% CO2 in DMEM (Invitrogen; 41966-029), supplemented with 10% FCS (Sera Laboratories International; EU-000-F), 100U ml-1 penicillin, 100µg ml-1 streptomycin, 0.25µg ml-1 amphotericin B (Invitrogen; 15240-062). 10µg of Fcer1g construct was transfected into COS-1 cells using the calcium phosphate method following the manufacturer’s instructions (BD Biosciences; K2051-1). The cells were maintained with 700µg ml-1 G418, (Geneticin®, Invitrogen; 11811-064). COS-1 cells stably expressing Fcer1g construct were then transfected via electroporation with 10µg of Fcgr3, Fcgr3-rs or the mutagensied Fcgr3 constructs using 260kV and 960µF. Cells were maintained with 700µg ml-1 G418 and 200µg ml-1 Hygromycin B (Invitrogen; 10687-010). Transfection was confirmed by western blot.
Phagocytosis of sheep red blood cells (SRBCs) by COS-1 transfectants.
COS-1 transfectants were trypsinised, harvested and washed with PBS. In each well of 8 well chamber slide (LabTek-II®, Nunc cat.no.177402), 2x104 COS-1 cells were diluted in 300l 10% FCS-DMEM with G418 or G418 and hygromycin as appropriate, and cultured at 37oC for 40 hours. Then, 0.2 ml SRBCs (Rockland, cat. No. R311-0050) were incubated with 9 mg anti-SRBC antibody (Rockland, cat. No. 113-4139) at 4 oC for 1 hour. Control SRBCs were incubated with 1% BSA. SRBC were washed three times with PBS after incubation and diluted in 200l 10% DMEM. 50l SRBCs were added to the COS-1 transfectants in each chamber well. The supernatant was removed 30 minutes later. SRBC which were not taken up by RBC were removed by gentle washing in lysis buffer. The slides were fixed with 4% paraformaldehyde and stained with Quick Diff solution (Glamidor Technical Services LTD) and internalised SRBC counted. For co-transfection of Fcgr3 constructs, cells expressing Fcer1g and Fcgr3 were newly transfected with either Fcgr3 or Fcgr3-rs in a bicistronic expression vector, pAdKS17, with lipofectamine. Fcgr3 or Fcgr3-rs alpha-chains were driven from a CMV promoter and green fluorescent protein was expressed in a second open reading frame to confirm transfection efficiency which was >70%.
Surface Expression of Fcgr3 and Fcgr3-rs
COS-1 cells transfected with respectively Fcer1g, Fcer1g andFcgr3, Fcer1g and Fcgr3-rs, and Fcer1gand Fcgr3-Gwere analysed for surface expression of the receptors. The transfected alpha-chains included a c-myc tag sequence to assess transfection efficiency. For flow cytometry staining 2x105 COS cells per stain were incubated with 1:50 anti-myc monoclonal IgG1 antibody (Upstate Cat 05-724) or isotype controls, in 2%FCS on ice, one hour. For permeabilised controls, 0.25% saponin was included. Then 1:50 FITC conjugated anti-mouse antibody in 2% FCS, was added at 4C, one hour. Following washing, the samples were fixed in 1% PFA and analyzed using a FACScalibur flow cytometer and CellQuest software (Becton Dickinson, Mountain View).
Quantification of FCGR3B copy number by quantitative real-time PCR (qPCR)
25 l PCRs were carried out using an Opticon 2 real-time PCR System (MJ Research). PCR conditions were 2 min at 94oC followed by 40 cycles of 15 sec at 94oC, 1 min at 60oC and 1 min at 72 oC. Melting curve analysis and sequence analysis were performed to check PCR product specificity. Standard curves were created by serial two-fold dilution of a top standard of 25ng to 0.78ng of a single genomic DNA sample and run in triplicate. Samples (2.5ng genomic DNA) were run in duplicate. Oligonucleotides were designed to specifically amplify the FCGR3B gene using primers selective for the promoter region: 5'-CACCTTGAATCTCATCCCCAGGGTCTTG-3'and 5'-CCATCTCTGTCACCTGCCAG-3'). The CD36 gene was used as a single copy control (primers 5'-TAAGTTCAGGTTCCTGGAATGC-3' and 5'-CAAATTATGGTATGGACTGTGC-3'). Normalisation to the control gene Forkhead Box P2 (FOXP2) (primers 5'-TGACATGCCAGCTTATCTGTTT-3' and 5'-GAGAAAAGCAATTTTCACAGTCC-3') was used to give an estimate of copy number6.The reproducibility of the qRT-PCR assay for each sample was calculated by estimating the within-sample variation measured through the coefficient of variation (C.V. % = 100*(standard deviation)/mean). Values which lay more than 1.5 times the inter-quartile range above the third quartile of the C.V. % were treated as outliers and removed from the dataset.
Logistic regression analysis
To calculate the odds ratio associated with copy number variation at FCGR3B and to test the hypothesis that other polymorphisms in the FCGR locus are associated with lupus nephritis, the dataset was fitted with logistic regression models.Age and gender were included as covariates in all presented logistic regression models. The goodness-of-fit 2 was used to estimate whether the models adequately fit the data and the Wald test was applied to estimate the significance (P-value) of each predictor variable.
The simple logistic regression model included only gene copy number at FCGR3B as the predictor variable, with FCGR3B gene copy number treated as an ordinal variable and lupus nephritis phenotype as the outcome (Supplementary Table 1A). The sequential logistic regression (SLR) was carried out to show whether adding data of FRGR2A and FCGR3A genes improves the model derived from analysis of FCGR3B copy number alone. In the SLR, genotype data for FCGR2A-G548A, FCGR3A-T559G and gene copy number at FCGR3B were used as predictor variables with lupus nephritis phenotype as the outcome. Genotypes of FCGR2A and FCGR3A were treated as categorical predictors by grouping the two homozygous and the heterozygous states into three distinct classes. The reference category was the most frequent (homozygous) genotype group. Comparison of log likelihood ratios between sequential models was used to evaluate improvement in the model by addition of new predictors that could be independently correlated with the outcome.
References
1.Basten, C. A., Weir, B. S. & Zeng, Z. B. QTL Cartographer User's Manual (Department of Statistics, North Carolina State University, Raleigh) (1999).
2.Lincoln, S. E., Daly, M. J. & Lander, E. S. Mapmaker/QTL Users' Manual (Whitehead Institute for Biomedical Research, Cambridge, Massachussetts) (1993).
3.Aitman, T. J. et al.Identification of Cd36 (Fat) as an insulin-resistance gene causing defective fatty acid and glucose metabolism in hypertensive rats. Nat Genet21, 76-83 (1999).
4.Cook, H. T., Smith, J., Salmon, J. A. & Cattell, V. Functional characteristics of macrophages in glomerulonephritis in the rat. O2- generation, MHC class II expression, and eicosanoid synthesis. Am J Pathol134, 431-437 (1989).
5.Bagchus, W. M., Donga, J., Rozing, J., Hoedemaeker, P. J. & Bakker, W. W. The specificity of nephritogenic antibodies. IV. Binding of monoclonal antithymocyte antibodies to rat kidney. Transplantation41, 739-745 (1986).
6.Iafrate, A. J. et al.Detection of large-scale variation in the human genome. Nat Genet36, 949-951 (2004).