WO2010102387A1 - Interleukin-12 polymorphisms for identifying risk for primary biliary cirrhosis - Google Patents

Abstract

Primary biliary cirrhosis and related autoimmune disorders are characterized by polymorphisms in various loci including IL-12-related genes, and by reduced levels of IL- 12A transcripts. Methods useful to screen for these defects are described for use in diagnosing these conditions. Methods useful to treat subjects afflicted with these conditions are also described.

Description

INTERLEUKIN-12 POLYMORPHISMS FOR IDENTIFYING RISK FOR PRIMARY BILIARY CIRRHOSIS

Field of the Invention

This invention relates to diagnosis and treatment of liver disease. More particularly, the invention relates to detection of genetic markers of liver disease, and to the treatment of subjects presenting with such markers. Embodiments of the invention are applied particularly for diagnosis and treatment of primary biliary cirrhosis.

Background to the Invention Primary biliary cirrhosis (PBC) is the most common autoimmune liver disease, affecting up to one in 1000 women over the age of 40 years . Treatment for this chronic cholestatic condition remains limited and empiric ². The granulomatous destruction of interlobular bile ducts that characterizes PBC is almost always associated with anti- mitochondrial antibodies (AMA, specific for the E-2 subunit of the pyruvate dehydrogenase complex, PDC-E2) ³. Hepatic accumulation of autoreactive T lymphocytes in PBC, as well as data from animal models of autoimmune cholangitis, suggest an important role for T lymphocytes and particularly, CD4+ T helper subpopulations in PBC pathogenesis ^4"6. A genetic predisposition for PBC has been revealed by both family aggregation and twin concordance data. The monozygotic twin concordance rate in PBC is 60% ⁷ and the sibling relative risk is estimated to be 10.5 . The co-existence of other autoimmune diseases in patients and raised prevalence of such diseases in their families is also in keeping with an important genetic contribution to PBC ⁹'¹⁰. Among the genes studied as PBC susceptibility candidates, only HLA has consistently shown an association with PBC, with the odds ratios associating PBC with HLA-DRBl *0801 in Caucasians estimated 2.4-3.3 and the derived population attributable fraction for HLA-DRB* 08 ranging from 2.8 to 8.8%. ^π'¹² PBC has also been associated in some, but not all studies, with alleles in the gene encoding the cytotoxic T immunoregulatory lymphocyte antigen 4 (CTLA-4) receptor. ^13"15

Summary of the Invention

To explore the genetic basis of PBC, there was conducted a genomewide association (GWA) screen involving the analysis of 2072 (536 cases/1536 controls) North American Caucasian subjects. This GWA data together with replication and fine-mapping data confirm the association of PBC with HLA class II alleles and suggest important roles for the interleukin (IL) 12/IL 12 receptor (IL 12R) signaling axis iα risk for PBC.

In one general aspect of the present invention, a subject presenting with, or at risk for, autoimmune disease including autoimmune liver disease and particularly including primary biliary cirrhosis (PBC) is identified by screening a biological sample obtained from the subject to identify at least one of the following characteristics:

(a) at least one polymorphism within the gene for one or both of IL- 12 A and IL 12-RB; and

(b) a level of IL- 12A protein, IL- 12 A protein complex or IL- 12 A transcript that is reduced relative to a healthy individual, the presence of either characteristic indicating the subject has, or is at risk for PBC.

In related aspects, the present invention provides kits comprising primers and instructions for the use thereof in genotyping the subject's DNA to reveal polymorphisms indicative of PBC.

In another general aspect of the present invention, there is provided a method useful to treat a subject presenting with, or at risk for, autoimmune disease including autoimmune liver disease such as PBC, the method comprising the step of administering to the subject a pharmaceutical composition comprising an agent effective to modulate, e.g., to stimulate or to inhibit, signaling via the IL-12/IL-12R pathway. In embodiments, the subject so treated is a subject that has tested positive for at least one polymorphism within a gene encoding one or both of IL-12A and IL- 12RB, or for a reduced level of endogenous IL- 12A protein or transcript.

These and other aspects of the present invention are now described with reference to the accompanying drawings in which:

Reference to the Figures Figure 1 provides a flow chart demonstrating quality control and study design. Panel A depicts the genotyping quality control for SNPs and subjects whilst Panel B outlines the stages undertaken during the study from stage I GWA screening, through stage II replication analysis, to fine-mapping. SNP - single nucleotide polymorphism; HWE - Hardy Weinberg Equilibrium; MAF- Minor allele frequency;

Figure 2 illustrates genomewide association results, in which the y axis represents the level of significance for each single-nucleotide polymorphism (logio-transformed P values; EIGENSTRAT) at the relative genomic position on each chromosome along the x axis. Chromosomes are shown in alternating colors. The dashed line shows the suggested GWA significance threshold.39 The P values are after adjusting in EIGENSTRAT for the 10 eigenvectors having the highest eigenvalues;

Figure 3 depicts gene structures and LD blocks for IL12A (A) and IL12RB2 (B). The top panels show the results of population-based genomewide, fine-mapping and combined association analyses. The results of chi-square analysis are presented and the -logio transformation of P values are plotted, so that more significant P values appear higher. Middle panels show the organization of the human IL12A (A) and IL12RB2 (B) genes and surrounding loci. Location of the genes is not drawn to exact scale. Bottom panels demonstrate the haplotype block structure, based on D2 and rc values for 25 genotyped IL12A locus SNPs and 30 IL12RB2 locus genotyped SNPs. Haplotype block structure was defined according to the criteria established by Gabriel et al. ²⁰ using the pairwise estimates of standardized Lewontin's coefficient (D'), while the LD among pairs of SNPs was characterized using the square of the correlation coefficient (n). The regions of high r2 are shown in dark green and regions with lower n are light grey, i.e. intensity decreases as n values decrease; Figure 4 is a reproduction of Table Sl herein, showing the complete association data for HLA region SNPs. Results are shown for SNPs in the HLA region yielding a stage I GWA significant association (P<lxlO^~4) sorted according to EIGENSTRAT P values. Statistical tests for association were carried out with the use of EIGENSTRAT and PLINK (http://pngu.mgh.harvard.edu/purcell/plink/). Genome positions are based on the NCBI database build 36 (hgl 8);

Figure 5 is a reproduction of Table S2 herein, showing the complete associate for non- HLA region SNPs (GWA study). Results are shown for SNPs in non-HLA regions yielding a stage I GWA significant association (P<lxlO^~ ) sorted according to

EIGENSTRAT P values. Statistical tests for association were carried out with the use of EIGENSTRAT and PLINK (http://pngu.mgh.harvard.edu/purcell/plink/). Genome positions are based on the NCBI database build 36 (hgl 8);

Figure 6 is a reproduction of Table 1 herein, showing results of the genomewide association study and stage II replication;

Figure 7 is a reproduction of Table 2 herein, showing fine mapping results from the IL- 12A and IL-12RB2 loci, and

Figure 8 is a histogram showing detection of IL- 12A transcripts in human PBMs, with our without prior LPS stimulation, in subjects genotyped either with risk (TAGTG) or non-risk (AGTCA) haplotypes.

Detailed Description of the Invention

There has now been discovered an association between autoimmune disease and certain genetic polymorphisms present in one or more of the genes that produce proteins involved in the IL-12/IL-12R signaling axis. These genetic polymorphisms associate particularly with autoimmune liver disease, and especially with primary biliary cirrhosis (PBC). The particular genes harbouring these one or more polymorphisms are one or both of (1) the human IL- 12A gene (UniProt reference P29459) that produces the interleukin-12 p35 protein (having UniProt reference P29459-] and which, together with the IL- 12 p40 protein also known as IL- 12B, combine to form functional heterodimeric IL- 12 protein), and (2) the human IL-12RB2 gene that produces the IL- 12 receptor beta-2 protein, (having UniProt reference Q99665-1, which together with the IL-12RBl protein combine to form functional IL- 12 receptor).

In the case of the IL- 12 genes, the one or more polymorphisms associated with PBC comprise the single nucleotide polymorphisms (SNPs) associated with the following SNP IDs: For the IL- 12A gene:

Minor/Major

SNP ID Location Alleles Physical Location rsl 7810546 5 'flanking G/A 161147744 rs583911 intron 2 G/A 161193084 rs668998 3 'flanking G/A 161 198245 rs6441284 3 'flanking G/A 161200962 rs485497 3 'flanking G/A 161201826 rs4680536 3 'flanking G/A 161202965 rs9852519 3 'flanking T/C 161203322 rs4679867 3 'flanking A/T 161206597 rs4679868 3 'flanking A/G 161206848 rs6441286 3 'flanking G/T 161211572 rs574808 3 'flanking C/T 161215677 rs589545 3 'flanking A/G 161216294

For the IL- 12RB2 gene:

Minor/Major

SNP ID Location Alleles Physical Location rsl 1209050 intron 3 A/C 67564324 rsl 908632 intron 8 G/T 67578394 rs3790565 intron 8 C/T 67583944 rs946685 intron 8 A/G 67588303 rs6679356 intron 9 C/T 67592782 rsl 0749775 intron 9 C/A 67594675 rs3790567 intron 9 A/G 67594965 rs6695348 intron 9 T/C 67599604

In the lists above, and elsewhere in this specification, the SNP ID refers to the identification number attached to a reference SNP cited and searchable in the single nucleotide polymorphism database provided by NCBI at http://www.ncbi.nlm.nih.gov/SNP. The "location" refers to a region within the subject gene, whereas the "physical location" refers to the location of the polymorphism within the whole of the human genomic sequence, as defined on the NCBI database, build 36 (hgl 8). The noted "minor/major alleles" identify the particular single nucleotide polymorphism, using underlining to identify the specific risk allele identified in subjects presenting with PBC, as revealed in greater detail in the examples herein.

Each of the alleles (polymorphisms) noted above is associated with PBC patients at a P value that is at least 10^"6. In embodiments of the present invention, the SNP targeted to identify subjects having, or at risk for, PBC is associated with PBC patients at a statistically more significant P value of at least 10^"7, preferably at least 10^"8 and most preferably at least 10^"9. In this context, and according to preferred embodiments of the present invention, a subject presenting with PBC, or at risk therefor, is identified when the polymorphism revealed by analysis of the DNA thereof comprises at least one SNP selected from the group of IL-12A SNPs consisting of rs6441286 and rs574808 and the group of IL-12RB2 SNPs consisting of rs3790567 and rs3790565. In a particularly preferred embodiment, the subject is so identified when the revealed polymorphism comprises the IL- 12A SNP rs4679868 located in the 3'UTR and/or the IL-12RB2 SNP 6679356 located in intron 9.

It will be appreciated that the present method, useful in identifying subjects presenting with PBC, and those at risk therefor, is adapted to reveal at least one of the noted SNPs in the subject's DNA, and desirably is adapted to reveal two or more such SNPs in any combination, thereby to enhance the confidence with which the diagnosis is made, and to understand the potential severity with which the disease may affect the subject.

In embodiments, the genotyping is applied to identify particular haplotypes or haplotype blocks, i.e. genetic sequences that incorporate at least two single nucleotide polymorphisms. In one embodiment, the method seeks to identify an IL- 12A or IL- 12RB2 haplotype, such as a haplotype including an IL- 12A haplotype selected from the five locus haplotypes AGTCA (SEQ ID No.l), TGTTG (SEQ ID No. 2) and, preferably, TAGTG (SEQ ID No.3) (where the five loci consist in order of the rs4679867, rs4679868, rs6441286, rs574808 and rs589545 SNPs), and/or an IL-12RB2 haplotype that is either a 10-locus haplotype selected from TGTAGTCAGC (SEQ ID No.4), TTTAGTTAGC (SEQ ID NO.5), GGCAACCCAT (SEQ ID NO.6), GGTAGCCAAT (SEQ ID No.7), GGCAATCAAT(SEQ ID NO.8), GCGATCAAT (SEQ ID NO.9), (where the 10 loci consist in order of the rsl908632, rs790564, rs3790565, rsl0489624, rs946685, rs6679356, rsl0889684, rsl0749775, rs3790567 and rs6695348 SNPs) or a 3 locus haplotype selected from ACA, GTC, GTA, GCA, ATC (where the 3 locus haplotypes consist of the ordered SNPs rs4297265, rs2270614 and rs2229546). In a preferred embodiment, the haplotype is a 3 -locus haplotype, preferably a 3 -locus IL- 12RB2 haplotype, particularly including GTC.

It should also be appreciated that detection of polymorphisms occurring within the IL- 12 genes noted herein can be used alone or in combination with detection of polymorphisms within other genes for which PBC-related polymorphisms are identified. In particular embodiments, PBC diagnosis is made by detecting PBC-associated polymorphisms that occur either per se or also within the human leukocyte antigen (HLA) gene, the gene for signal transducer and activator of transcription 4 (STAT4, UniProt Q 14765) and/or the gene for cytotoxic T-lymphocyte protein 4 (CTLA-4, UniProt P 16410), as well as in one or more of the genes for other proteins noted herein as harbouring SNPs correlating with PBC, as set out in Tables 1 and 2 herein (see Figures 4 and 5, respectively).

In embodiments, the HLA region-based polymorphisms that can usefully be detected to aid in PBC diagnosis are one or more of the following: Risk

SNP ID Location Gene Allele

Rs2856683 32763196 HLA-DQBl C Rs2395148 32429532 CόorflO A

Rs9277535 33162839 HLA-DPBl G Rs3806156 32481676 BTNL-2 A In a preferred embodiment, the polymorphism is the HLA-DQBl risk allele C noted above, and haplotypes comprising this SNP.

In other embodiments, the STAT4 gene-based polymorphisms that can usefully be detected to aid in PBC diagnosis are one or more of the following:

Minor/Major SNP ID Location Alleles Physical Location

RsI 0168266 intron 5 T/C 191644049

Rsl6833239 intron 4 A/G 191648505

Rs3024921 intron 3 _ A/T 191651517

In other particular embodiments, PBC diagnosis is made by detecting PBC-associated polymorphisms that occur either per se or also, together with IL- 12 gene-based polymorphism(s), within the 17ql2-21 locus (P - 3.50 x 10^{" 13}), within the IRF5-TNPO3 risk locus (P = 8.66xl O^{" 13}) and within the MMELl locus (P = 2.28 x 10^"9).

Polymorphisms that correlate with PBC and lie within the 17q 12-21 risk locus include the following. This locus includes four genes, i.e., zone pellucida-binding protein 2 (ZPBP2), the IKAROS family zinc finger 3 protein involved in leukocyte development and IgE production(IKZF3), Gasdermin-B (GSDMB) involved in epithelial barrier function, and ORMl -like protein 3, an uncharacterized protein (ORMDL3) belonging to a family of transmembrane proteins anchored in the endoplasmic reticulum. All eight SNPs across this region achieved significance in replication analysis:

SNP Position Gene Risk Allele rs907092 35175785 IKZJ- 3 Λ rs9303277 35229995 IKZF3 C rsl 1557467 35282160 ZPBP2 G rs8067378 35304874 GSDMB A rs2305480 35315722 GSDMB r rs2290400 35319766 GSDMB A rs7216389 35323475 GSDMB I rs4795405 35341943 ORMDL3 I

Of the 17q 12-21 risk loci noted above, the strongest association signal came from the ZPBP2 SNP, rsl 1557467, which is thus a preferred marker for PBC diagnosis in this series. Polymorphisms that correlate with PBC and lie within the MMELl locus, located on chromosome Ip36, include the following. This locus comprises the gene for membrane metallo-endopeptidase-like 1 (MMELl).

SNP Position Gene Risk Allele rs3748816 2516606 MMELl C rs3890745 2543484 MMELl G Other polymorphisms that correlate with PBC lie within the IRF5-TNPO3 locus. This locus comprises the genes for interferon regulatory factor 5-transportin 3).

SNP Position Risk Allele rs35000415 128372852 T

IRF5_rs2070197 128376236 C rslO488631 128381419 C rs 12539741 128384041 T

Among the associations identified, the strongest signals were provided by rsl2539741 and rs2070197 alleles. These variants map just 3 ' to the IRF 5 coding region, are in tight linkage disequilibrium (r²>0.95) with one another, and their associations with PBC reach P values of 1.65xlO^'10 (OR=1.63) and 3.74xlO^~10 (OR-1.62), respectively, in the combined cohort analysis. These polymorphisms are also among the SNPs at this locus highly associated with SLE and are both correlated with changes in IRF5 expression in transformed B cells [4]. By contrast, an insertion-deletion polymorphism 64bp upstream of IRF5 exon IA representing a putative disease-causal variant for SLE, showed only modest association with PBC (P=LOOxIO^"3) and no associations were detected between PBC and another SLE-associated SNP, rs2004640, or a 3^' UTR variant, rs 10954213, identified as a key predictor of IRF 5 expression level [4]. Analyses also identified a number of haplotypes across the IRF5-TNPO3 locus that were more strongly associated with PBC than any of the individual test SNPs. Among these, a two-allele haplotype involving the rs35000415 and rs 12539741 SNPs showed the strongest association (omnibus P=7.79xlO^'31, OR=I .97 for the TT haplotype) with risk for PBC, as shown below:

Two SNP Haplotype

HAPLOTYPE P SNPs

TT 5.58xl O^"19 rs35000415|rsl2539741 TC 4.62xlO^"16 rs35000415|rsl2539741 CC 2.82x1 (T⁰⁸ rs35000415|rs 12539741 OMNIBUS 7.79x10^"" rs35000415|rsl 2539741

Three SNP Haplotype

HAPLOTYPE P SNPs

TCT 4.24x10^"19 rs35000415|IRF5_rs2070197|rsl2539741

TTC 8.87xlO^"16 rs35000415|IRF5_rs2070197|rsl2539741

CTC 2.85x10^'08 rs35000415|lRF5 rs2070197|rsl2539741

OMNIBUS 1.13xlO^"30 rs35000415|IRF5 rs2070197|rsl2539741

Thus, in further embodiments, the genotyping is applied to identify particular haplotypes or haplotype blocks, i.e. genetic sequences that incorporate at least two single nucleotide polymorphisms particularly in the IRF5-TNPO3 locus, and particularly the two SNP haplotype shown supra. Such haplotypes further include the 3-locus risk allele haplotypes within the IRF5-TNOP3 also shown supra.

In summary, with respect to PBC risk alleles, the present invention identifies numerous loci associated with this condition, and permits diagnosis of the condition through genotyping based on any one, two, three, four, five, six, seven, eight, nine, ten or more of the following risk alleles, and any combinations thereof, including combinations that consist only of a selection from among the alleles noted below, and combinations that comprise at least an IL- 12A risk allele, or an IL- 12RB risk allele, and any combination of one or more other risk alleles noted below:

SNP ID Allele Physical Location Gene rs 17810546 A 161 147744 IL-12A rs58391 1 G 161 193084 IL- 12A rs668998 G 161 198245 IL- 12A rs6441284 A 161200962 IL-12A rs485497 G 161201826 IL- 12A rs4680536 A 161202965 IL-12 A rs9852519 C 161203322 IL- 12A rs4679867 T 161206597 IL- 12A rs4679868 A 161206848 IL- 12A rs6441286 G 16121 1572 IL- 12Λ rs574808 T 161215677 IL- 12A rs589545 G 161216294 IL- 12A rs 1 1209050 A 67564324 1L- 12RB2 rs 1908632 G 67578394 IL- 12RB2 rs3790565 C 67583944 IL-12RB2 rs946685 A 67588303 IL-12RB2 rs6679356 C 67592782 IL-12RB2 rs 10749775 C 67594675 IL-12RB2 rs3790567 A 67594965 IL-12RB2 rs6695348 T 67599604 IL-12RB2

rs2856683 C 32763196 HLA-DQB l rs2395148 A 32429532 Cόorf 10 rs9277535 G 33162839 HLA-DPB l rs3806156 A 32481676 B ΓNL-2 rsl 0168266 r 191644049 S I A I 4 r_ 16833239 G 191648505 S 1 A 1 4 rs3024921 A 191651517 STAT4 rs907092 A 35175785 IKZF3 rs9303277 C 35229995 IKZF3 rsl 1557467 G 35282160 ZPBP2 rs8067378 A 35304874 GSDMB rs2305480 T 35315722 GSDMB rs2290400 A 35319766 GSDMB rs7216389 T 35323475 GSDMB rs4795405 T 35341943 ORMDLi rs3748816 C 2516606 MMELl rs3890745 G 2543484 MMELl rs35000415 T 128372852 IRF 5 -TN P 03 rs2070197 C 128376236 IRF5 rsl 0488631 C 128381419 1RF5-TNPO3 rsl 2539741 T 128384041 1RF5-TNPO3 It will thus be appreciated that the present invention relates to a method useful to identify subjects presenting with, or at risk for, autoimmune disease including PBC which comprises the step of genotyping DNA from the subject to identify one or more of the polymorphisms herein identified, wherein the presence thereof indicates the disease or that the subject is at risk for the disease. The ultimate diagnosis of the disease will of course consider other disease hallmarks such as clinical symptoms characteristic of the disease when manifest, knowledge of hereditary factors, histological examinations, the presence of protein markers, and the like. In the case of PBC, detection of one or more polymorphisms herein identified can be used to confirm a clinical diagnosis which is based on the appearance in the subject of signs of altered liver function including such symptoms as the presence in blood of markers of deranged liver function, including anti- mitochondrial antibodies as well as antibodies to such nuclear proteins as glycoprotein 210 and p62. Patients suffering from PBC may also present with fatigue, itchy skin, jaundice resulting from bilirubin elevation, local collections of cholesterol in the skin and around the eyes (xanthoma) as well as symptoms generally associated with cirrhosis and portal hypertension.

Patients having one or more of these symptoms, and patients otherwise suspected for whatever medical reason of having or being at risk for PBC, are genotyped in accordance with a method of the present invention to identify one or more of the polymorphisms described herein, as an aid or adjunct to diagnosing this disease.

The genotyping process is generally well established in the art. It entails an examination of the sequence or other physical properties of genomic DNA extracted from the subject to identify either the specific polymorphisms herein identified or the hallmarks thereof. Genotyping can be established by extracting DNA or RNA from a patient sample such as blood, tissue biopsy, and the like, and then analyzing the DNA by any one or more methods including hybridization-based techniques, enzyme-based methods, and other post-amplification methods based on physical properties of DNA.

Useful hybridization-based methods make use of DNA probes that are complementary in sequence to, and hence hybridize with, a target DNA region comprising a SNP of interest. Such methods include dynamic allele-specific hybridization, which takes advantage of the differences in the melting temperature in DNA that results from the instability of mismatched pairs. In this method, a segment of target DNA is amplified and attached to a bead through a PCR reaction with a biotinylated primer. The amplified product is then attached to a streptavidin column and washed to remove the unbiotinylated strand. An allele/SNP-specific oligonucleotide probe is then added in the presence of a molecule that fluoresces when bound to double stranded DNA. The intensity is then measured as temperature is increased, until the melting point can be determined. A SNP will result in a lower than expected melting point. In another hybridization-based approach, oligonucleotide SNP arrays are used, in which probes are arrayed on a small chip. Several redundant probes are used to interrogate each SNP, to deal with mismatched hybridization, the probes being designed to have the SNP site in several different locations as well as containing mismatches to the SNP allele. By comparing the differential amount of hybridization of target DNA to each of these redundant probes, it is possible to determine specific homozygous and heterozygous alleles. The Affymetrix Human SNP 5.0 GeneChip® system is particularly useful for this purpose.

Still other hybridization-based techniques are based on specifically engineered single stranded oligonucleotide probes which have complementary regions at each of end of a central probe sequence. One end of the probe is labeled with a fluorophor, and the other end is labeled with a fluorophor quencher. The probe takes on a hairpin structure in its isolated state, and accordingly does not fluoresce, unless and until the target sequence is encountered by and hybridized to the probe. This elegant design allows detection of a single SNP in a DNA sample.

Enzyme-based genotyping techniques include those based on restriction fragment length polymorphism (RFLP), which makes use of the many different endonucleases and their high affinity to unique and specific restriction sites within target DNA. In this approach, target DNA is subjected to a particular one or more endonucleases, and the resulting fragment pattern is compared against a control sample lacking the SNP of interest. A difference in gel-resolved fragments between the target DNA and control indicates that the target DNA incorporates an alteration, such as a SNP, that renders it either vulnerable or resistant to digestion. In a variation of this technique, a particular region of the target DNA being queried for the presence of a given SNP can be amplified, and the amplified DNA can then be digested, to reduce the complexity of the result and point more directly to the SNP of interest as being responsible for the difference in endonuclease-mediated fragmentation.

Still other enzyme-based genotyping methods include polymerase chain reaction (PCR) based methods, such as tetra primer ARMS-PCR, in which two pairs of primers are used to amplify two alleles in one PCR reaction. The primers are designed such that the two primer pairs overlap at a SNP location, but each match perfectly to only one of the possible SNPs. As a result, if a given allele is present in the PCR reaction, the primer pair specific to that allele will produce product but not to the alternative allele with a different SNP. The two primer products are designed so that their PCR products have lengths sufficiently different to allow their discrimination by gel electrophoresis.

Other methods use flap endonuclease (FEN), which catalyzes structure-specific cleavage that is highly sensitive to mismatches and can be used to interrogate SNPs with a high degree of specificity. In the basic assay, a FEN called cleavase is combined with two specific oligonucleotide probes that together with the target DNA form a tripartite structure recognized by cleavase. The first probe is complementary to the 3 'end of the target DNA, and has a final base that is a non-matching base that overlaps the SNP nucleotide. The second probe is an allele-specific probe which is complementary to the 5 'end of the target DNA, but also extends past the 3' side of the SNP nucleotide. The allele-specific probe will contain a base complementary to the SNP nucleotide. If the target DNA contains the desired allele, the probes will bind to the target DNA forming the tripartite structure which can then be cleaved by cleavase. If the SNP nucleotide in the target DNA is not complementary with the allele-specific probe, the correct structure does not form and no cleavage occurs. The assay is usually coupled with a FRET system to detect the cleavage event.

The primer extension approach can also be used in genotyping. This is a two step reaction in which a probe that hybridizes upstream of a SNP nucleotide is used followed by a mini-sequencing reaction in which DNA polymerase extends the hybridized primer by adding a base complementary to the SNP nucleotide. This incorporated base is detected and determines the SNP allele. Various detection methods including mass spectrometry (e.g., MALDI-TOF) and ELISA-like formats are employed. Illumina Inc.'s Infinium® assay is an example of a whole genome genotyping pipeline that is based on primer extension, and allows for genotyping of over 100,000 SNPs. The assay uses hapten-labeled nucleotides in a primer extension reaction, wherein the hapten label is detected using antibodies which in turn are coupled to a detectable signal.

One particularly suitable approach entails PCR-based amplification of a DNA region in which a suspected SNP is located, using suitable forward and reverse primers to generate an amplicon having a size that is conveniently small to allow direct sequencing of the amplicon, and then comparison of the determined sequence against a wild type sequence, so that the presence of the risk allele can be revealed.

It will be appreciated that a wide variety of techniques thus are available for use in genotyping DNA from a subject to be diagnoses for PBC based on the SNPs and haplotypes herein described. In particular embodiments of the present invention, the DNA of a given subject can be interrogated to identify one or more of the present SNPs by implementing PCR-based sequencing methods. The IL- 12 SNP sequences exemplified herein were all extracted from the NCBI SNP database, and annotation of each SNP position in the human genome is based on the March 2006 human reference sequence (NCBI Build 36.1) produced by the International Human Genome Sequencing Consortium. SNP genotyping is performed for instance using the Sequenom MassARRAY platform using the iPLEX method. SNP assay design can be performed with the Sequenom Assay Design v3.1, and eXTEND available at MySequenom.com under Tools. For each target SNP analysis, the proximal variation (ProxSNP) can be masked to ensure that PCR and extension primers can bind, amplify and extend a unique region. For this purpose, the table below provides primers that are suitable for producing amplicons that incorporate IL- 12 SNPs of particular interest (the risk alleles), the specific presence of which can then be determined by amplicon sequencing according to well established methods.

In embodiments, genomic DNA is extracted from whole blood using standard phenol/chloroform extraction procedures and resuspended in IX TE buffer (pH 8.0).

PCR amplification is carried out using 5-10 ng of template DNA, in 5 ul reactions containing 1.25X PCR Buffer (Qiagen), 1.625 mM MgC12 (Qiagen), 500 uM dNTP mix (Fermentas), 100 nM primer mix (IDT) and 0.5 U Hotstar Taq polymerase (Qiagen). The reactions are incubated in a standard thermocycler using the following cycling conditions: initial denaturation at 94⁰C for 15 minutes, followed by 45 cycles of 94⁰C for 20 seconds, 56°C for 30 seconds, 72⁰C for 1 minute, followed by a final extension at 72⁰C for 3 minutes.

Unincorporated dNTPs are dephosphorylated by treatment with shrimp alkaline phosphatase following PCR. The 5 ul PCR reaction is incubated with 2 ul of SAP mix, containing 0.85X SAP buffer and 0.3 U SAP enzyme (Sequenom), in a standard thermocycler at 37⁰C for 20 minutes, followed by a 5 minute heat inactivation at 85⁰C.

For analysis on the Sequenom iPLEX system, for instance, 2 ul of iPLEX extension cocktail are added to the PCR reaction to a final concentration of 0.222X iPLEX buffer, IX iPLEX termination mix, 0.625 uM, 0.833 uM, 1.04 uM or 1.25 uM of each primer, and IX iPLEX enzyme (Sequenom). The primer concentrations in the multiplex reactions are adjusted based on the primer mass, with higher concentrations (1.04 uM or 1.25 uM) used for high mass primers. The reaction conditions for primer extension are as follows: initial denaturation at 94°C for 30 seconds, followed by a 40 cycle program consisting of a single denaturation at 94°C for 5 seconds and 5 cycles of 52°C for 5 seconds and 8O⁰C for 5 seconds. A final extension is performed at 72°C for 3 minutes.

To remove salts from the iPLEX reaction products, the samples are diluted with 16 ul of water and 6 mg of Clean Resin (Sequenom) was added to each reaction. The reactions are rotated for at least 10 minutes, followed by centrifugation at 5000 rpm for 5 minutes. The reaction products are dispensed onto a 384-element SpectroCHIP bioarray (Sequenom) using the Sequenom RS-1000 MassARRAY Nanodispenser and analyzed using the Sequenom MassARRAY Analyzer Compact.

In another of its aspects, the present invention provides a kit useful in the diagnosis of primary biliary cirrhosis (PBC), the kit comprising nucleic acid primers useful to amplify DNA obtained from a subject having or suspected of being at risk for PBC, wherein the primers have a nucleic acid sequence adapted to amplify a region of said DNA comprising at least one PBC risk allele as defined herein. The primers typically are in the form of DNA oligonucleotides, and prime the region comprising the SNP/risk allele by hybridizing therewith to produce an amplicon that is then sequenced to reveal the presence of the SNP. The primers can be packaged in the kit within sealed vials comprising the DNA primers in lyophilized or other suitable form. The kit further comprises instructions for the use of such primers in detecting risk alleles associated with PBC. Primers suitable for amplifying patient DNA to reveal IL- 12 SNPs associated with PBC are set out below:

SNP ID IL-12A rsl7810546 Genome Position 161147744

2^nd PCR Primer ACGTTGGATGCAATGAGGTTTTCTGTATGGC(SEQ ID No.10)

1 ^st PCR Primer ACGTTGGATGCAGATCAAGAACCCAACAAG(SEQ ID No.1 1 )

Amplicon length 99

Extension Primer GTATGGCTTTCTGTGGTA(SEQ ID NO.12)

Extl call/Ext2 call G/A SNP ID IL- 12A rs4679867

Genome Position 161206597 2^nd PCR Primer ACGTTGGATGGAAGTTCTGGCAATTCAGTC(SEQ ID No.13) 1^st PCR Primer ACGTTGGATGCACAGGTTAAGGGCTCAGTC(SEQ ID No.14) Amplicon length 97 Extension Primer TTCTGGCAATTCAGTCTTATAAC(SEQ ID No.15) Extl call/Ext2 call A/T

SNP ID 1L-12A rs4680536

Genome Position 161202965 2^nd PCR Primer ACGTTGGATGGGTCCTGTTTTAAAGCATCC(SEQ ID No.16) 1^st PCR Primer ACGTTGGATGCCAACTATTGCCAAAGAGCC(SEQ ID No.17) Amplicon length 83 Extension Primer AGCATCCTATTCCTTGG(SEQ ID No.18) Extl call/Ext2 call G/A

SNP ID IL- 12A rs9852519

Genome Position 161203322 2^nd PCR Primer ACGTTGGATGGCATTCTTCCCCATGAAAAG(SEQ ID No.19) 1^st PCR Primer ACGTTGGATGCCTGGAGAAGTACAACTGGC(SEQ ID NO.20) Amplicon length 101 Extension Primer CCCCATGAAAAGTAAATAATTCCAC(SEQ ID NO.21 ) Extl call/Ext2 call C/T

SNP ID IL-12A rs4679868

Genome Position 161206848 2^nd PCR Primer ACGTTGGATGTTTATGGGGCCAATGATTTG(SEQ ID NO22) 1^st PCR Primer ACGTTGGATGGATTTTTGCACTCTATCTCC(SEQ ID NO.23) Amplicon length 95 Extension Primer TGAATTGGTAGACATATATGCC(SEQ ID NO.24) Extl call/Ext2 call G/A

SNP ID IL-12A rs485497

Genome Position 161201826 2^nd PCR Primer ACGTTGGATGGATTTCGAGGGCTTACCGAG(SEQ ID NO.25) 1^st PCR Primer ACGTTGGATGCTGATCTGCTGGTTTCCAAG(SEQ ID NO.26) Amplicon length 99 Extension Primer TCTGACTTGCTCCTGG(SEQ ID NO.27) Extl call/Ext2 call G/A

SNP ID IL-12A rs574808

Genome Position 161215677 2^nd PCR Primer ACGTTGGATGTCAACTGCTGTCAATCCCAC(SEQ ID NO.28) I^s1 PCR Primer ACGTTGGATGGAAAGTAGCAAAGTAGACAG(SEQ ID NO.29) Amplicon length 100 Extension Primer gCAGAAAGTGGGATTTGTAG(SEQ ID No.30) Extl call/Ext2 call C/T SNP ID IL-12A rs58391 1

Genome Position 161 193084 2^nd PCR Primer ACGTTGGATGAGCTTGTCTTAAGGGTTTGC(SEQ ID NO.31 ) 1^st PCR Primer ACGTTGGATGCAAGTATAACTTCTAAAGGG(SEQ ID NO.32) Amplicon length 100 Extension Primer TTTGCATGTTTGTTATATCCATCA(SEQ ID NO.33) Extl call/Ext2 call A/G

SNP ID IL-12A rs589545 Genome Position 161216294 2^nd PCR Primer ACGTTGGATGTCAAATAAGATAGGCCTCCC(SEQ ID NO.34) 1^st PCR Primer ACGTTGGATGACTGCGCCTGACTCCAATTT(SEQ ID NO.35) Amplicon length 97 Extension Primer TAACTAATTTCAGAGCAATTTTT(SEQ ID NO.36) Extl call/Ext2 call G/A

SNP ID IL- 12A rs6441284

Genome Position 161200962 2^nd PCR Primer ACGTTGGATGCTGTCTAAGTCAAGGGAATG(SEQ ID NO.37) 1^st PCR Primer ACGTTGGATGAAGAGAGAGAAGGGAGATAC(SEQ ID NO.38) Amplicon length 100 Extension Primer GAATGCTTTCCCTATAACC(SEQ ID NO.39) Extl call/Ext2 call A/G

SNP ID IL-12A rs6441286

Genome Position 161211572 2^nd PCR Primer ACGTTGGATGGGAGGTGTATCTGGTGTTTC(SEQ ID NO.40) 1^st PCR Primer ACGTTGGATGGAGAAGCTGGAGAGAGCTA(SEQ ID NO.41 ) Amplicon length 99 Extension Primer CTCCCAGAGGAGGACCTCA(SEQ ID NO.42) Extl call/Ext2 call G/T

SNP ID IL-12A rs668998

Genome Position 161 198245 2^nd PCR Primer ACGTTGGATGGGGCAAGAATTTGAAAAACG(SEQ ID NO.43) 1^st PCR Primer ACGTTGGATGTCTTTCTCCCTGTAAATGTG(SEQ TD NO.44) Amplicon length 99 Extension Primer AATTTGAAAAACGAGAATGTTTG(SEQ ID NO.45) Extl call/Ext2 call G/A

SNP ID IL-12RB2 rs 10749775

Genome Position 67594675 2^nd PCR Primer ACGTTGGATGGATGGGGTAGGAAATTCTGG(SEQ ID NO.46) 1^st PCR Primer ACGTTGGATGCTTCCTCCCCAGTGAGGTTC(SEQ ID NO.47) Amplicon length 94 Extension Primer CAGAGTCACTGTCAAGG(SEQ ID NO.48) Extl call/Ext2 call C/A SNP ID IL-12RB2 rsl908632

Genome Position 67578394 2^nd PCR Primer ACGTTGGATGGATCATTGCACACTTCTTCAG(SEQ ID NO.49) 1^st PCR Primer ACGTTGGATGTGTGTGTGTGAGAGAGACAG(SEQ ID NO.50) Amplicon length 98 Extension Primer ACTTCTTCAGTTTTAAGAAATGAGT(SEQ ID NO.5 1 ) Extl call/Ext2 call G/T

SNP ID IL-12RB2 rs6679356 Genome Position 67592782 2^nd PCR Primer ACGTTGGATGAGCTCGGCTCTGAAGAATTG(SEQ ID NO.52) 1^st PCR Primer ACGTTGGATGGCTAAGAATGAGGTTGATGC(SEQ ID NO.53) Amplicon length 95 Extension Primer CTTCTTTCCCTCCTGAC(SEQ ID NO.54) Extl call/Ext2 call T/C

SNP ID IL-12RB2 rs6695348

Genome Position 67599604 2^nd PCR Primer ACGTTGGATGAGCAAGCATCAGAACCATAC(SEQ ID NO.55) 1 ^st PCR Primer ACGTTGGATGATTCCTGGTCTGATAATTCC(SEQ ID NO.56) Amplicon length 80 Extension Primer CtTCAGAACCATACTTGGATA(SEQ ID No.57) Extl call/Ext2 call C/T

SNP ID 1L-12RB2 rs946685

Genome Position 67588303 2^nd PCR Primer ACGTTGGATGCATTTGCTGACTACAGIGTG(SEQ ID NO.58) I ⁸¹ PCR Primer ACGTTGGATGTTTCTTGGCAAGGCTGTCCC(SEQ ID NO.59) Amplicon length 100 Extension Primer ccccTGTGTCTTTCAGCAAAC(SEQ ID NO.60) Extl call/Ext2 call A/G

SNP ID IL-12RB2 rsl 1209050

Genome Position 67564324 2^nd PCR Primer ACGTTGGATGTCGTGGGAGGTACTCΛATGC(SEQ ID NO.61 ) 1^st PCR Primer ACGTTGGATGTGCATTGTCTCTTGGAGCAC(SEQ ID NO.62) Amplicon length 99 Extension Primer cACTCAATGCCTATGTACAC(SEQ ID NO.63) Extl call/Ext2 call C/A

SNP ID IL-12RB2 rs3790567

Genome Position 67594965 2^nd PCR Primer ACGTTGGATGACAGCTATAGCTCATGGTCC(SEQ ID NO.64) 1^st PCR Primer ACGTTGGATGTCTGGCCTGGAGCTTTCCAA(SEQ ID NO.65) Amplicon length 98 Extension Primer CTCATGGTCCTCCATGTATG(SEQ ID NO.66) Extl call/Ext2 call G/A SNP ID IL-12RB2 rs3790565

Genome Position 67583944 2^nd PCR Primer ACGTTGGATGCAGCAGATGGCACAATTTAG(SEQ ID No.67) 1^st PCR Primer ACGTTGGATGCTGTGATGGGCCAGGATTG(SEQ ID NO.68) Amplicon length 98 Extension Primer ATGGCACAATTTAGATTGTTT(SEQ ID NO.69) Extl call/Ext2 call C/T

SNP ID IRF5 rslO488631 PCR Primer 1 : ACGTTGGATGATTCACTGCCTTGTAGCTCG (SEQ ID NO. 70) PCR primer 2: ACGTTGGATGGTCTATCAGGTACCAAAGGC (SEQ ID No. 71 ) Extension Primer: TGTAGCTCGGAAATGGTTC(SEQ ID NO. 72) Alleles: T/C

SNP ID IRF5 rs35000415 PCR Primer 1 : ACGTTGGATGGGCACTTCCAGATTCAAACC(SEQ ID NO. 73) PCR primer 2: ACGTTGGATGGTGTGATCTCTAAGGAGGAC(SEQ ID NO. 74) Extension Primer: GGGTGCTCACCCTTATCC(SEQ ID NO. 75) Alleles: T/C

SNP ID IRF5 rsl2539741 PCR Primer 1 : ACGTTGGATGTCAATTCATACCTCCTGCCC(SEQ ID NO. 76) PCR primer 2: ACGTTGGATGCACTTTTATGACCTGAATGC(SEQ ID NO. 77) Extension Primer: CCTGCCCTGGTCAAAA(SEQ ID NO. 78) Alleles: T/C

SNP ID MMELI rs3748816

PCR Primer 1 : ACGTTGGATGTTGGAGTACTCCTCGTCCAG(SEQ ID NO. 79) PCR primer 2: ACGTTGGATGATGAGCATCCGGGAGCAGA(SEQ ID NO. 80) Extension Primer: CCTCGTCCAGGCGCCTGTTC(SEQ ID No. 81) Alleles: T/C

SNP ID MMELl rs3890745

PCR Primer 1 : ACGTTGGATGGCATTTTGTGACTGTAGCCC(SEQ ID NO. 82) PCR primer 2: ACGTTGGATGGGAATCACCTGGGGAAATTG(SEQ ID NO. 83) Extension Primer: CTGAGGGAGGGCCCAA(SEQ ID NO. 84) Alleles: A/G

SNP ID Chrl 7 region rsl 1557467 PCR Primer 1 : ACGTTGGATGGATGTATTCTTTAGAGTGCTG(SEQ ID NO. 85) PCR primer 2: ACGTTGGATGTGATGGCTCTATGACATGGC(SEQ ID NO. 86) Extension Primer: GCTGAAGAAAATCTTGGATA(SEQ ID NO. 87) Alleles: G/T

A subject that presents with one or more of the polymorphisms noted herein is a subject that is at risk for PBC or, if other PBC signs and symptoms are manifest, is a subject presenting with PBC.

In other aspects, the present invention provides for the detection of IL-12A protein in a patient sample as an additional or alternative way to identify subjects presenting with or at risk for PBC and related autoimmune disorders. Such subjects present with reduced levels of endogenous IL- 12A protein and/or transcript, relative to normal subjects. Detection of endogenous IL- 12A can be achieved using any of the methods established for protein species detection. This typically can utilize an antibody or other agent that binds selectively and specifically to the target protein. Antibodies that bind IL- 12A selectively are available commercially. These antibodies can be used in any useful assay format, including radioimmunoassay (RIA), or an enzyme immunoassay (EIA) or enzyme-linked immunosorbent assay (ELISA) format or with the aid of flow cytometry, to capture and detect IL-12A and IL-12A complexes (such as IL-12 and IL-35) present in a patient sample. Measurement of IL- 12A levels can then be achieved by quantifying the bound IL- 12 A, using for instance an agent that is labeled for detection.

Presently, there is no known cure for PBC but medication may slow the progression of the disease so that a normal lifespan may be attainable. Specific treatment for fatigue is unavailable, although ursodoxycholic acid is frequently used in treatment, to reduce cholestasis and to improve liver function and blood test results. Cholestyramine may also be prescribed to absorb bile acids. Multivitamins, including vitamin D and calcium, are also recommended as patients with PBC have poor lipid-dependent vitamin absorption. In advances cases of PBC, liver transplant is indicated and can result in favourable prognosis.

The genotyping results presented herein, showing a clear and significant association between PBC and polymorphism in IL-12-related genes, suggest that PBC subjects can benefit from IL-12A-based therapy. It is expected in particular that these subjects will benefit from administration of agents that stimulate the IL-12/IL-12R signaling axis. As will be appreciated from the genotyping data, PBC subjects exhibit polymorphisms not only within the IL- 12A gene, but also in the IL-12RB2 gene. The polymorphisms exist in gene regions that are unlikely to affect the sequences of the transcripts and proteins expressed from these genes. Rather, the polymorphisms reside in 5', 3' and intronic sequences suggesting that these patients will have difficulty with the levels at which they express the required proteins, relating to reduced or dysfunctional transcript production and other phenomena unrelated to proper protein function.

More particularly, further evidence, shown herein, indicates that the level of IL-12 A transcripts in human peripheral blood mononuclear (PBM) cells is reduced in healthy individuals homozygous for the 5 locus PBC risk haplotype ((TAGTG) compared with healthy individuals homozygous for the non-risk haplotype (AGTCA). The applications of other techniques useful to assess the relative levels of IL-12 A transcripts and/or translated protein in biological samples obtained from individuals thus form valuable additional aspects of the invention.

In one aspect, there is provided a method for identifying a subject having or at risk for PBC, or for an autoimmune condition related thereto such as an autoimmune liver disease, comprising measuring the level of an IL- 12A RNA transcript, an expression product thereof or IL- 12A complex, such as IL-12 and/or IL-35, in a biological sample obtained from that subject, and comparing the measured level with a corresponding reference level for a healthy control subject, wherein a reduction in the measured level relative to the reference level indicates autoimmune disease, including PBC.

Assay methods useful to determine the presence or amount of the IL-12A-based analyte can be applied to biological samples that include tissues, cells and extracts thereof including RNA and protein. The biological sample can thus be blood, including whole blood, plasma and serum, as well as saliva, urine, cerebrospinal fluid, cells particularly including white cells such as leukocytes and lymphocytes, peripheral blood mononuclear cells generally including antigen presenting cells and monocytes, tissues including particularly spleen, marrow, and the like. The biological sample can also be extracted from such fluids, cells and tissues, particularly including protein and RNA components thereof.

When applied to detect IL- 12A in tissue samples, the assay can be based on immunohistochemistry techniques established in the art, in which an IL- 12A antibody or any other agent that binds selectively to IL-12A is used essentially to target a conjugated stain or label to the site at which IL- 12A is localized on the tissue. The intensity with which the tissue is stained or labeled by the binding agent is then compared, either visually or preferably with instrumentation, against control tissue, and a reduction in the label intensity compared to the control tissue reveals tissue that produces a reduced level of IL-12A, and indicates the subject providing the tissue has or is at risk for PBC.

Similarly, agents that bind selectively to IL- 12A can be used to assay the presence or amount of IL-12 A protein that is present either in cell-bound or soluble form. The practice of flow cytometry can be applied to detect and measure the amount of IL-12A protein in solution or on the surface of cells. Similarly, the presence of soluble IL- 12A protein can be detected using the binding agent and well established assay platforms and formats, such as the ELISA, RIA, and EIA formats including sandwich formats, or any other assay format, suitably one that is automated and adapted to quantify the amount of IL- 12A detected by the assay.

A most useful reagent for use in these assays is an antibody that binds specifically to IL- 12A, including antibodies that bind specifically to IL- 12A when complexed to form IL- 12 and IL-35. Such antibodies reveal the presence of free and complexed IL- 12 A, and are desirable for that reason. A variety of such antibodies are available commercially as reagents for this purpose.

The IL- 12A specific antibodies can be provided as immunoconjugates in which the antibody is complexed with a detectable label, such as an enzyme label including peroxidase, a FITC label, gold sol, and the like to provide a means for revealing the presence and quantity of antibody that has accumulated at the desired target. In the alternative, the antibody without a label can be revealed using a labeled secondary antibody that binds to the IL- 12A primary antibody.

The presence and amount of IL-12A present in a biological sample can also be determined using an assay for IL-12A transcripts. By this method, RNA extracted from target cells, particularly liver cells or circulating immune cells including lymphocytes and leukocytes, is assayed to identify or quantify the presence of IL-12A encoding RNA or a signature fragment thereof. To achieve this, any standard method of amplifying RNA can be applied, in combination with primers specific for amplification of IL- 12 A- encoding RNA or any unique fragment thereof. One useful primer pair is described in the examples herein.

It will be appreciated that the assay is designed to reveal a relative reduction of IL-12 A in samples obtained from test subjects relative to normal subjects, thereby to identify test subjects having or at risk for PBC. The relative reduction sought is one having statistical significance. The reduction is relative to normal subjects who, using the genotyping method herein described for instance, are confirmed to be free from PBC or the risk thereof. A statistically significant number of normal subjects will of course be tested to arrive at a known reference level for IL- 12A in normal individuals for a given assay, and results from test subjects will be compared against these reference values when making a determination. In one study, comparing healthy subjects and subjects with alcoholic liver disease, it was revealed that the serum IL- 12 levels in healthy subjects was on the order of 39.3 +/- 8.3 pg/ml (p<0.02, n - 35) (see J. Chin. Med. Assoc, 2010 Feb; 73(2):67-71, abstract). Similarly, but in a study of asthma subjects, serum IL-12 levels in healthy controls, determined by ELISA, are reported as 67.42 +/- 20.58 ng/L (PO.01 , n=30) (see Zhongguo Dang Dai Er Ke Za Zhi, 2008 Apr; 10(2):146-8, abstract). These reports provide reference values for circulating IL-12, and are expected to correlate substantially with circulating levels of IL-12A. Thus, relative reductions in IL- 12A can be determined by measuring levels of IL-12 in test and control subjects. A more complete assessment of IL-12A levels will also take into account the levels of IL-35, a cytokine that also incorporates IL- 12A in a heterodimer. Accordingly, the determination of total IL- 12A protein level can be accomplished either by measuring the amount of IL- 12A detectable protein in a given sample, or by measuring the amount of IL- 12 A- complexed protein in the sample, such as by combining detectable levels of IL-12 and IL- 35. Antibodies that detect IL-12 and IL-35 are also well known and available commercially, including the IL-12 antibody known as ustekinumab, a human monoclonal antibody that binds to the p40 component of IL-12. Whatever measurement is taken, it will be appreciated that levels of the same analyte will also be determined by the same method but in normal subjects, and a relative reduction by comparison will reveal a patient at risk for PBC.

Antibodies that are commercially available and useful in the present, protein-based assays are listed below: Antibodies for p35 (IL- 12A)

Vendor Cat# Epitop/specificity source Note

Sigma GW22498 23-219aa chicken poly

Sigma HPAOOl 886 39-171 Rabbit poly

Monoclone

R&D MAB 1570 23-219aa Mouse 27537

R&D IC2191C 23-219 mouse mono, for FC

Abeam AB66462 c-terminal Rabbit poly

Abeam ab37635 23-219 chicken poly

Abnova PAB l 159 203-221 Rabbit poly

Abnova PAB8409 57-253 Sheep poly

Novus Biologicals NB 100-81023 c-terminal rabbit poly

Novus Biologicals NB 100-75535 23-219 Chicken poly

R & D Systems IC2191 A IL-12 / IL-35 p35 Mouse Mono, conjugated

R & D Systems IC2191 C IL-12 / IL-35 p35 Mouse Mono, conjugated

R & D Systems IC2191P IL-12 / IL-35 p35 Mouse Mono, conjugated

Santa Cruz sc-74149 IL-12A p35 Mouse unconjugated

Santa Cruz sc-65323 IL-12A p35 Mouse unconjugated

Antibodies for p40 (IL- 12B)

Vendor Cat# Epitop/specificity source Note

Sigma GW21394 161 -233 chicken poly

Sigma SAB2500536 278-291 Goat poly

R&D MAB 1510 unknown mouse mono

Everest Biotech EB07862 IL-12 /P40 Goat poly,unconjugated

IL12 / IL-23

Cell Sciences CSI 10018 (P4O) mouse mono, clone C8.6

Abeam ab20196 unknown mouse mono

Abeam ab62822 278-292 Goat poly

Abnova MAB2067 23-328 mouse mono

Abnova PAB7305 278-291 Goat poly

Novus Biologicals NB 100-61661 278-291 Goat poly

Novus Biologicals NB 100-64693 unknown mouse mono

Novus Biologicals NB 100-75276 IL 12B chain Chicken poly

GeneTex GTX82129 IL 12 beta Chicken poly

GeneTex GTX20050 IL12 p40 mouse mono

Conjugates -for

Biolegend 501809 IL-12/IL-23 p40 mouse FC

Biolegend 508801 IL-12/IL-23 p40 mouse Conjugated -biotin

Conjugates -for eBiosciences 46-7235-41 IL-12/IL-23 p40 mouse FC eBiosciences 14-7127 IL-12/IL-23 p40 mouse Mono, purified

BD BIOSCIENCES 559329 IL-12 (p40/p70> mouse mono,conjugates

BD Bioscience 554477 IL-12/IL-23 p40 mouse Mono, purified

BD Bioscience 554477 IL-12 P40/P70 mouse mono, conjugates

Santa Cruz Biotech sc-53120 IL-12 P40/P70 mouse Mono, purified

Santa Cruz Biotech sc-53348 IL 12 p40 mouse mono

Santa Cruz Biotech sc- 1282 IL 12B C-term Goat poly Antibodies for IL- 12

Vendor Cat# Epitop/specificity source Note

Santa Cruz sc-74147 FL 1L12 mouse Mono, purified

Santa Cruz sc-74150 FL IL12 mouse Mono, purified

Biolegend 51 1001 IL12 P70

LifeSpan BioSc LS-C42972-50 IL12 Goat poly, conjugated

LS-C24859-

LifeSpan BioSc 100 IL12 Goat poly

LS-C24876-

LifeSpan Bio 100 IL12 mouse Mono, purified

GeneTex GTX82388 FL 1L12 Chicken poly

GeneTex GTX20053 FL IL12 mouse mono

Fisher/ ABR PA 1-86600 1L12 poly eBiosciences 12-7235-71 IL12 mouse mono,conjugates eBioscience 14-7129 (1L-I2) p4θ/7O mouse mono eBioscience 14-7128 IL12 mouse mono

BD Bioscience 554658 IL12 P70 mouse mono,purified

BD Bioscience 559325 IL12 P70 mouse mono,conjugates

Interleukin 12 (IL-

Millipore 04-962 12) mouse clone 17.8

Invitrogen A08E6E5 IL12 mono, purified

The reduced level of IL-12A protein and/or transcripts in PBC and related diseases suggests that IL- 12A replacement will provide a therapy useful in their treatment.

Thus, in another aspect of the present invention, the present invention provides a method for treating a subject presenting with or at risk for PBC, optionally a subject identified by the genotyping or assay methods herein described, comprising the step of administering to the subject an agent effective to promote signaling via the IL-12/IL-12R axis. In related aspects, the invention provides the use of IL-12 A or a complex thereof such as IL- 12 and/or IL-35 in the preparation of a medicament for the treatment of PBC and related autoimmune liver conditions.

Agents that are effective to promote signaling via the IL-12/IL-12R axis particularly include the IL-12A protein and the IL-12RB2 protein, i.e., the proteins encoded by the genes in which PBC-associated polymorphisms have been identified as described herein. The amounts of such agents effective in the treatment of PBC are those amounts, delivered by any suitable treatment regimen, that result in the reduction or elimination of any one or more PBC symptoms, as identified hereinabove.

In one embodiment, the method comprises the step of administering an effective amount of IL-12A protein, or a polynucleotide encoding an IL- 12A protein, particularly to subjects presenting with a genotype that includes any one or more of the IL- 12A polymorphisms herein associated with PBC, or subjects presenting with reduced IL-12 A protein or transcript levels. In embodiments, the IL- 12A protein is the human IL- 12A protein, which is the mature form of the protein having UniProt reference P29459. In the alternative, the administered protein can be a variant of the IL- 12A protein that incorporates 1, 2, 3, 4, 5 or more amino acid additions, substitutions and/or deletions, particularly including natural variations, and further including functional groups including additional glycosylation, polyethyleneglycolation or other depot means, yet retains the activity of IL-12A, i.e., associates with wild type IL- 12B (UniProt reference 29460) to form a heterodimer that stimulates IL- 12R, or associates with wild type Ebi3 to form IL-35 (where Ebi3 is a protein expressed from Epstein-Barr virus induced gene 3, and is also the IL-27B chain, having UniProt reference Q 14213).

In another embodiment, the method comprises the step of administering an effective amount of IL- 12RB 2 protein, particularly to subjects presenting with a genotype that includes any one or more of the IL-12RB2 polymorphisms herein associated with PBC. In embodiments, the IL-12RB2 protein is the human IL-12RB2 protein, which is the mature form of the protein having UniProt reference Q99665). In the alternative, the administered protein can be a variant of the IL-12RB2 protein that incorporates 1,2,3,4,5 or more amino acid additions, substitutions and/or deletions yet retains the activity of IL- 12RB2, i.e., associates with wild type IL-12RB1 (UniProt reference P42701) to form a heterodimer that is stimulated by IL- 12.

In another embodiment, the method comprises the step of administering an effective amount of a polynucleotide encoding IL- 12A, or encoding IL-12RB2, in the form of gene therapy using established administration and dosing regimens.

Also provided by the present invention is the use of IL- 12 A, a protein complex comprising IL- 12A or a polynucleotide encoding IL- 12A in the preparation of a medicament for the treatment of PBC. Similarly, the present invention also provides the use of IL-12RB2, a protein complex comprising IL-12RB2 or a gene encoding IL-12RB2 in the preparation of a medicament for the treatment of PBC.

Patients presenting with PBC also frequently present with other autoimmune and/or infectious diseases and disorders, the presence of which can be associated with an altered IL- 12 signaling axis which may result from the polymorphisms herein identified. Thus, in a related aspect of the present invention, the present invention relates to a method for treating an autoimmune disease or disorder, which may or may not be associated with PBC in the treatment recipient, by administering an amount of IL-12 A or complex and/or IL-12RB2 effective to reduce or eliminate one or more symptoms thereof, wherein the autoimmune disease or disorder is selected from ulcerative colitis, Crohn's disease, rheumatoid arthritis, autoimmune cholangitis, diabetes mellitus, multiple sclerosis, psoriasis, myasthenia gravis, autoimmune uveitis, systemic lupus erythematosus, Sjogren's syndrome and ankylosing spondylitis, and the like. In a further related aspect of the present invention, there is provided a method for treating an infectious disease or disorder, by administering an amount of IL-12 A and/or IL-12RB2 effective to reduce or eliminate one or more symptoms thereof, wherein the infectious disease or disorder is selected from example from HSV, HIV, hepatitis B, hepatitis C, papilloma, as well as bacterial infections including tuberculosis, salmonellosis, listeriosis and parasite infection such as malaria, leishmaniasis and schistosmiasis. In still another of its aspects, the present invention provides a pharmaceutical composition comprising a pharmaceutically acceptable carrier and an agent effective to stimulate the IL-12/IL-12R axis, wherein the agent is selected preferably from human IL- 12A or a complex thereof, and human IL-12RB2.

As proteins, both IL- 12A and IL-12RB2 can be produced routinely for medical use by application of recombinant DNA methods, in which DNA coding for the desired protein is introduced, together with operably linked expression controlling elements, into a selected production host that can then be cultured to produce the desired protein. Production of IL-12RB2 is described, for instance, in US 5,919,903, incorporated herein by reference, where gene expression is accomplished in Ba/F3 cells. Production of IL- 12A is described, for instance, in WO92/005256 incorporated herein by reference, wherein expression of the IL- 12A gene is described in a variety of hosts and expression systems.

It will also be appreciated that the genes encoding these proteins can be administered per se to drive endogenous expression of the protein products. The genes can be in the form of genomic DNA, in which case the wild type gene is used instead of the polymorphic forms associated with PBC, or the RNA or cDNA forms thereof, wherein the genes are delivered using protocols established in the art of gene therapy.

Pharmaceutical compositions of IL- 12 protein or complex, such as IL- 12 or IL-35 and other p35 -containing complexes, are prepared for storage and for administration by mixing the protein having the desired degree of purity with optional pharmaceutically acceptable carriers, excipients or stabilizers (Remington's Pharmaceutical Sciences, 16^l edition, Osol, A. Ed. [1980]), in the form of lyophilized formulations or aqueous solutions. Acceptable carriers, excipients, or stabilizers are nontoxic to recipients at the dosages and concentrations employed, and include buffers such as phosphate, citrate, and other organic acids; antioxidants including ascorbic acid and methionine; preservatives (such as octadecyldimethylbenzyl ammonium chloride; hexamethonium chloride; benzalkonium chloride, benzethonium chloride; phenol, butyl, or benzyl alcohol; alkyl parabens such as methyl or propyl paraben; catechol; resorcinol; cyclohexanol; 3- pentanol; and m-cresol); low molecular weight (less than about 10 residues) polypeptides; proteins such as serum, albumin, gelatin, or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone; amino acids such as glycine, glutamine, asparagines, histidine, arginine or lysine; monosaccharides, disaccharides, and other carbohydrates including glucose, mannose, or dextrins; chelating agents such as EDTA; sugars such as sucrose, mannitol, trehalose or sorbitol; salt-forming counter-ions such as sodium; metal complexes (e.g., Zn-protein complexes); and/or non-ionic surfactants such as TWEEN, PLURONICS or polyethylene glycol (PEG).

The active ingredients to be used for in vivo administration must be sterile. This is readily accomplished by filtration through sterile filtration membranes.

Sustained-release preparations may be prepared. Suitable examples of sustained-release include semipermeable matrices of solid hydrophobic polymers containing the antibody, which matrices are in the form of shapes articles, e.g., films or microcapsules. Examples of sustained-release matrices include polyesters, hydrogels (for example, poly (2- hydroxyethyl-methacrylate), polyactides (U.S. Pat. No. 3,773,919), copolymers of L- glutamic acid and ethyl-L-glutamate, non-degradable ethylene-vinyl acetate, degradable lactic acid-glycolic acid copolymers such as injectable microspheres composed of lactic acid-glycolic acid copolymer and leuprolide acetate, and poly-D-(-)-3-hydroxybutyric acid. While polymers such as ethylene-vinyl acetate and lactic acid-glycolic acid enable release of molecules for over 100 days, certain hydrogels release proteins for shorter time periods. Rational strategies can be devised for stabilization depending on the mechanism involved. For example, if the aggregation mechanism is discovered to be intermolecular S-S bond formation through thio-disulfide interchange, stabilization may be achieved by modifying sulfhydryl residues, lyophilizing from acidic solutions, controlling moisture content, using appropriate additives, and developing specific polymer matrix compositions.

A protein therapeutic according to the invention may be administered with a pharmaceutically-acceptable diluent, carrier, or excipient, in unit dosage form. Any appropriate route of administration can be employed, for example, parenteral, intravenous, subcutaneous, intramuscular, intracranial, intraorbital, ophthalmic, intraventricular, intracapsular, intraspinal, intracisternal, intraperitoneal, intranasal, aerosol, or oral administration.

For the treatment of subjects presenting with PBC, the appropriate dosage will depend on the severity and course of the disease, whether the agent is administered for preventative or therapeutic purposes, previous therapy, the patients clinical history and response to the agent, and the discretion of the attending physician. The agent is suitably administered to the patient at one time or over a series of treatments. For example, depending on the type and severity of the disease, about 1 μg/kg to 15 mg/kg (e.g., 0.1-20 mg/kg) of protein is a candidate dosage for administration to the patient, whether, for example, by one or more separate administrations, or by continuous infusion. A typical daily dosage might range from about 1 μg/kg to 100 mg/kg or more, depending on the factors mentioned above. For repeated administrations over several days or longer, depending on the condition, the treatment is sustained until a desired suppression of disease symptoms occurs. However, other dosage regimens may be useful. The effect of the agent can be seen as an increase in the endogenous levels, e.g., blood levels, of interferon gamma or STAT4. This effect is expected to be seen for example when a unit dose of about l-50ug of the agent is administered subcutaneously. The progress of this therapy is easily monitored by conventional techniques and assays used currently in the care of PBC subjects. Other agents useful to stimulate the IL-12/IL-12 axis can be identified, for example, by screening combinatorial chemical libraries or small molecule or peptide/protein libraries including particularly IL- 12A analog libraries and IL-12RB2 analog libraries to identify agents having the ability to stimulate IL- 12 receptor signaling. Such libraries may be derived from natural products, synthetic (or semi-synthetic) extracts, or chemical libraries according to methods known in the art. Those skilled in the field of drug discovery and development will understand that the precise source of compounds is not critical to the screening procedure(s) of the invention. Examples of natural compound sources include, but are not limited to plant, fungal, prokaryotic, or animal sources, as well as modification of existing compounds. Numerous methods are also available for generating random or directed synthesis (e.g., semi-synthesis or total synthesis) of any number of chemical compounds, including, but not limited to, saccharide-, lipid-, peptide-, and nucleic acid-based compounds. Synthetic compound libraries may be obtained commercially or may be produced according to methods known in the art. Furthermore, if desired, any library or compound is readily modified using standard chemical, physical, or biochemical methods. These agents can useful be screened using standard bioassays that report IL- 12 activity, such as the NK cell-based assay described in US20080138833 published June 12, 2008 incorporated herein by reference, wherein the effect of a given agent is determined by detecting production of interferon gamma by NK cells having a functional IL-12/IL-12R axis, an increase indicating that the agent stimulates that axis.

It will thus be appreciated that the present invention provides a novel means for identifying subjects that present with or are at risk for autoimmune disease particularly including PBC. The genotyping of such subjects reveals polymorphisms within proteins responsible for the signaling that occurs via the IL-12/IL-12R axis. The present invention accordingly also provides a method for the treatment of such subjects, particularly including subjects that are first genotyped by the present method and found to have one or more risk alleles, by administering thereto an agent effective to modulate signaling via this axis, including agents such as IL- 12A and IL-12RB2 proteins and genes encoding them.

Examples

SUBJECTSAND CONTROLS

A two stage design was implemented (Figure 1 ) beginning with a GWA survey (stage I) incorporating Canada-based PBC cases and controls and additional US (M. D. Anderson) historic controls. This analysis was followed by replication analyses (stage II) incorporating US-based (Ha) and additional Canada-based (lib) PBC cases and controls, as well as fine-mapping studies on the Canada-based GWA-screened and replication cohorts. Institutional ethics committee approval was obtained for all study participants. All patients fulfilled AASLD criteria for PBC 16 and all subjeots were Caucasian of European origin. Canada-based cases had a mean age of 60.7 years (23-92) at recruitment and a diagnosis of PBC for an average of 8.9 years (1-30). 93% of cases were female, 95.5% AMA positive and 5.2% had received a liver transplant. Cases were recruited between 2005-2008 with the majority (372) ascertained from Toronto and the remaining from London (Ontario), Halifax, Edmonton and Calgary. US-based cases had a mean age at recruitment of 60.2 years (29-85) and a diagnosis of PBC for an average of 9.3 years (0-33). 92.4% of US cases were female and 87.6% AMA positive; 10.5% had received a liver transplant.

Canadian control subjects (n=1279, 87% female) were geographically matched healthy volunteers with no prior history of autoimmune disease (mean age 31.4 years, range 18- 88) also ascertained between 2005 and 2008. A total of 1137 healthy subjects from the M. D. Anderson Cancer Center Lung Cancer Study, a case-control study ongoing since 1999, were also used as controls. These "historic" controls (mean age 61.1 ; SD 8.9) were ascertained from Kelsey-Seybold Clinics in the Houston Metropolitan area. Caucasian ancestry was ascertained first by self-report and then by excluding any subjects whose genotypes clustered with other or mixed racial groups (i.e. African- Americans, Hispanics, Asians, and Caribbean Islanders were excluded).

Another 310 healthy "US-based" controls were from the Mayo Clinic PBC Genetic Epidemiology Registry and Biospecimen Repository. Controls were recruited from the Mayo Clinic Division of General Internal Medicine during annual visits for preventative medical examination and were matched by age (±2.5 years), gender, and state of residence to individual PBC patients. Control exclusion criteria included evidence of prior or current liver disease. 90% were female and the mean age 62.4 years (33-88).

A total of 536 PBC cases and 1536 healthy controls (399 Caucasian healthy volunteers with no prior history of autoimmune diseases collected in Toronto and an additional 1 137 historic control samples collected from the M. D. Anderson Medical Center Lung Cancer Study ¹⁷) were included in the stage I GWA analysis. Stage Ha replication included 410 PBC cases and 310 controls from the Mayo Clinic PBC Genetic Epidemiology Registry ¹⁴ and Stage lib included additional Canada-based subjects (1 16 PBC cases and 896 controls). All Canada-based cases and controls meeting quality control in stage I as well as the added 116 cases and 896 controls ascertained in Canada were included in the combined and fine -mapping analyses.

GENOTYPING AND QUALITY-CONTROL FILTERING

DNA purification and genotyping methods are described as follows. DNA was prepared using standard methods. A total of 536 samples from PBC subjects and 399 control subjects collected in Canada were genotyped on Illumina HumanHap370 Genotyping BeadChips. This platform was selected because of the genome coverage provided, reflecting the almost exclusive use of tag SNPs with high minor allele frequencies (median=0.25), density (mean inter SNP distance=7.8kb) and LD (mean n>0.8). A total of 1 137 samples from M. D. Anderson Cancer Center Lung Cancer Study (historic controls) were genotyped at the Centre for Inherited Disease Research (Johns Hopkins University) on the Illumina HumanHap300 chip. Cluster definitions for each SNP were determined using Illumina BeadStudio Genotyping Module v 2.3.41 and the combined intensity data from -90% of study samples. The resulting cluster definition file was used on all study samples to determine genotype calls and quality scores. Genotype calls were made when a genotype yielded a quality score (Gencall value) of 0.25 or higher in Texas and 0.15 in the Canadian samples. Among these markers, 0.46% of cells were missing. For the lung control study, fifty-four blind duplicate pairs were included, the concordance of SNP genotype calls was 99.99%. Based upon 106 duplicate HapMap samples (56 unique individuals) from all 4 populations (Caucasian, Yoruban, Chinese and Japanese) the concordance was 99.8%, and the Mendelian consistency rate calculated from 29 Hapmap trios (16 unique trios) was 99.96%. A Sequenom MassARRAY iPLEX platform was used for both replication and finemapping genotyping (http://www.sequenom.com/seq-genotyping.html). In brief, allelespecific extension products were plated onto a SpectroCHIP (Sequenom), subjected to mass spectrometric analysis (MassARRAY system, Sequenom, San Diego, CA) and the genotypes identified using SpectroCALLER software, all in accordance with the manufacturer's instructions.

In stage II, we genotyped 16 SNPs in 14 genes/loci based on GWA signals. For finemapping studies, initiated coincident with the stage II analysis, a total of 93 SNPs from three different loci (ILl 2A, IL12RB2-IL23R, and STAT4) were selected based on at least one of the following criteria: 1) HapMap Phase II data (http://www.hapmap.org/) identifying the SNP as a TagSNP with minor allele frequency >0.05 or a higher minor allele frequency (>0.01) and n threshold of 0.8 and/or 2) localization in potential functional regulatory gene regions (i.e. exon, exon-intron boundary, 5' or 3' flank). Out of 93 SNPs tested, 89 passed quality control filters including 25 across the STAT4 gene and 200kb flanking regions at chromosome 2q32.2-q32.3, 25 in the IL12A region of chromosome 3q25.33-q26, and 39 in the IL12RB2/IL23R loci at chromosome Ip31.1- 31.2.

In stage I the PBC cases and Canadian controls samples were genotyped for 373,400 SNPs using the Illumina HumanHap370 BeadChip. Control samples from the M. D. Anderson Cancer Center (historic controls) were genotyped on the Illumina HumanHap300 BeadChip. Genotyping data were filtered prior to data analysis. After setting Maximum per-person missing (MIND) to be no less than 0.05 and GENO (Maximum per-SNP missing) to be no less than 0.05, minor allele frequency to be no less than 0.01 and Hardy Weinberg equilibrium test to be no less than 0.0001 , 929 samples and 334,444 SNPs from the HumanHap370 data set (including 536 PBC cases and 399 controls) and 1 137 subjects and 312,547 SNPs from the HumanHap300 data set were retained. After merging, the same MIND and GENO settings were applied to all of the samples and SNPs. The final combined data retain 2012 subjects (505 PBC subjects and 1507 control subjects) and 305,724 common SNPs.

In brief, genotypes were called if they exceeded minimum quality control metrics. Individual samples with genotype call rates less than 95% and SNPs with call rates less than 95%, minor allele frequencies less than 1%, and/or deviating from Hardy- Weinberg equilibrium with P<0.0001, were removed. Replication and fine-mapping were performed using the Sequenom MassArray iPLEX genotyping platform.

STA TISTICAL ANAL YSIS

For GWA association analysis, pairwise identity-by- state analysis was performed with PLINK¹⁸ to identify individuals with excess identity-by-descent-sharing (PI_HAT >0.25). One subject from each of 15 pairs exceeding this threshold was removed from the association analysis (the individual in each pair with the higher SNP call rate was retained). Hierarchical cluster analysis was performed with PLINK to identify individuals showing similar genotypes, over the entire genome. From this analysis, 39 outlier samples showed more than 4 standard deviations from a nearest neighbor and were removed from the analysis. Subsequent association analyses were conducted using PLINK with a conditional analysis that adjusted for stratifications among groups of individuals identified by hierarchical clustering. To alternatively control for the potential confounding influence of population stratification, association analysis was also performed based on a principal-component analysis implemented in EIGENSTRAT¹⁹ with default parameters, in which we adjusted for the 10 eigenvectors having the highest eigenvalues (-k 10, -m 10, -t 10, and -s 6.0). The lambda values showed minimal inflation, being 1.085 and 1.056 before and after eigenvector adjustment; using PLINK the lambda values were 1.14 without, and 1.09 with adjustment for clusters.

Allele and genotype associations were assessed using PLINK software (http://pngu.mgh.harvard.edu/purcell/plink/) and linkage disequilibrium (LD) between pairs of SNPs and haplotypes were evaluated by Haploview v.4.1 (http://www.broad.mit.edu/mpg/haploview). Haplotype block structure was defined according to the criteria established by Gabriel et al. ° using the pairwise estimates of standardized Lewontin's coefficient (D'), while the LD among pairs of SNPs was characterized using the square of the correlation coefficient (π).

Combined analyses were carried out using SAS to perform Cochrane Mantel-Haenszel tests allowing for potential confounding of the case/control frequency with genotype frequencies among groups of individuals at different stages/centers (analyses were stratified into GWA samples, US replication samples and Canada replication samples).

To carry out power calculations for a GWA screen, a genotypic relative risk analysis (http://statgen.iop.kcl.ac.uk/gpc/cc2.html) was conducted with the following parameters: disease prevalence, 0.001 for PBC, D' between disease and SNP alleles, 1.0; genotypic heterozygous relative risk, 1.5 under a log-additive model; alpha, 5 x 10^"5 (corresponding to 16/317,000 the probability to retain a marker from the first to second stage of analysis) marker allele frequency (A) = 0.20, as this is the mean minor allele frequency of the tested SNPs. The power to detect association was estimated as 81.7% based on the sample of 2072 subjects (536 cases and 1536 controls) in a GWA study (stage I), and 90.7% based on the total sample of 3744 subjects (1031 PBC cases and 2713 healthy controls) in a combined analysis using a genome-wide significance criterion of p<lxlθ^"7. For a heterozygous genotypic relative risk of 1.4, the corresponding powers are 50% in the first stage and 60% in the combined analysis.

HLA HAPLOTYPE ANALYSIS

For the HLA region for which linkage disequilbrium extends over a broad area, we used PLINK to jointly model the impact that inferred SNP haplotypes (derived from the 13 PBC-associated HLA SNPs identified in Table 1 , see Figure 6) have on disease risk. We first modeled the association of all haplotypes across the region with PBC risk, and then sequentially removed SNPs until the SNPs removed yielded significant haplotypic risks for disease. We then analyzed the remaining SNPs as haplotypes to evaluate the haplotypic effects on PBC risk.

POPULATION ATTRIBUTABLE FRACTION

Population attributable fraction (AFp) for single DRBl genotypes described in the introduction were calculated as :

PFψR -X)

1 + PF(OR -X) where OR is the odds ratio associated with DRBl * 0801 genotype and PF is the genotype frequency of the risk genotype.

Population attributable fraction (AFp) for a single SNP was calculated over genotypes i (indexed as 1 , 2 for the heterozygous and homozygous risk genotypes respectively) and for the Canadian and American replication populations (indexed as j=l,2).

In this analysis, PF is the risk allele frequency and OR is the odds ratio associated with having a risk genotype.

The population attributable risk fraction was computed over all possible combinations of genotypes among the three risk SNPs according to the formula:

where the factors are defined for each joint genotype (Hanley JA. A heuristic approach to the formulas for population attributable fraction. J Epidemiol Community Health 2001 ;55:508— 14). The odds ratios and frequencies used for these computations are shown in Table S 13

Table S13: Population attributable risk for the HLA, IL12A and 1L12RB2 loci

SNP Genotype Average PAR

HLA rs2856683 AA CA CC Total

Canada-based controls 540 300 48 888

Frequency 0.451 0.251 0.040

OR 1 1.6 1.67

US-based controls 186 104 19 309

Frequency 0.155 0.087 0.016

OR 1 1.7985 2.41 0.212

IL12A rs6441286 GG GT TT

Canada-based controls 148 399 334 881

Frequency 0.124 0.335 0.281

OR 1.69 1.23 1

US-based controls 34 164 1 11 309

Frequency 0.029 0.138 0.093

OR 2.98 1.39 1 0.215

1L12RB2 rs3790567 AA AG GG

Canada-based controls 53 329 509 891

Frequency 0.044 0.275 0.425

OR 2.67 1.35 1

US-based controls 22 107 178 307

Frequency 0.018 0.089 0.149

OR 1.53 1.51 1 0.184

Table Sl 3 Legend: Population attributable risk fractions were calculated as described in the supplementary appendix, using control subject (US and Canadian) genotyping data for the three most significant variants identified.

CONDITIONAL AND LINEAR TREND ANALYSES

10 To further investigate whether a single or multiple SNPs influence disease risk within the ILl 2A, IL12RB2 and STAT4 loci (extended Canada based cohort dataset), we conducted conditional analysis using multiple logistic regression with a stepwise selection procedure implemented in SAS version 9.1 (SAS Institute Inc, Cary, North Carolina, USA) for which the entry criterion allowed any SNP with P<0.15 to enter the model for consideration but at each step, each SNP had to meet a P<0.05 criterion to be retained. We tested for a linear trend in PBC risk using PLINK, with an additive genetic model, to identify gene dose-response effects for rs6441286/rs574808 and rs3790567 SNPs at the IL12A and IL12RB2 loci. A gene dose effect was also evaluated using SAS version 9.1 and a general model for genotypic effects.

RESULTS

STAGE I- GWA SCREEN

The most significant results of the stage I GWA survey are shown in Table 1 (full dataset provided in Supplementary Tables Sl and S2) and Figures 2 and 3, based on analysis of0 the 305,724 SNPs that passed quality control filters and the 505 PBC and 1507 control subjects retained after pairwise IBS estimation and population stratification correction. Thirteen SNPs in the HLA/HLA boundary regions on chromosome 6p21.3 and three SNPs from two distinct non-HLA regions at the ILl 2A (3q25.33-q26) and IL12RB2 (Ip31.2) loci, respectively, met a GWA significance threshold of P<5.0xl0^"7 (based on EIGENSTRAT P values). The most highly associated markers, rs2856683, rs9275312, rs9275390 and rs7775228, map to the region between the HLA-DQBl and HLA-DQA2 genes, the PLINK P values for these associations ranging between 1.7OxlO-io and 8.58xlO^"17 with odds ratios varying between 1.81-2.01. Nine other SNPs mapping in or near the CόorflO (P=5.62xlO^"10), HLA-DPBl (P=8.28xlO^"9), BTNL2 (P=1.27xlO^"9) and0 HLA-DRA (P=6.83xlO^'7) genes within the HLA region also gave highly significant association signals, the CόorflO locus having the greatest association with risk (OR=3.24). Haplotype analysis, however, revealed that 4 of the 13 SNPs (rs2395148, rs3135363, rs2856683, and rs9357152 in the CόorflO, BTNL2, HLA-DQBl and HLADQBl genes, respectively) were sufficient to explain all of the HLA -associated risk for PBC (see Table S3 below).

Table S3 Associations between HLA haplotypes and primary biliary cirrhosis

Frequency Frequency Degrees of

SNP Haplotype Cases Controls T Freedom P

CGAG 0.0535 0.09816 18.54 1 1.66x10 ^s

CAAG 0.1239 0.1643 9.244 1 2.36 xlO^"3

AACA 0.04533 0.0119 41.65 1 1.09 xlO^"10

CACA 0.2954 0.2009 37.94 1 7.29 xlO^"10

CGAA 0.159 0.1871 3.945 1 0.047

CAAA 0.3229 0.3377 0.7315 1 0.3924

Overall NA NA 98.7 5 9.93 xlO^"20

Table S3 Legend: The impact that SNPs and haplotypes across the MHC have on risk for0 PBC was analyzed using PLINK. For this analysis, we included the 13 most significant SNPs in the HLA region that are identified in Table 1, Figure 6 and jointly modeled the impact of the inferred haplotypes derived from these SNPs on disease risk.

Similar levels of association were also found in a separate analysis of the genome-wide data incorporating only Canadian controls (Table S4). Table S4. Comparison of original stage 1 GWA analysis with an analysis restricted to Canadian controls alone.

Risk a ele frequency Original GWA a nalysis incorporating Canadian and Alternate GWA analysis incorporating only Canadian historic controls controls n=S0S cases and 1507 controls (383 Cu iidnn and 1124 H=SOS cases and 383 Canadian controls

M D Anderson historic

US Contro s Canadian LIGhNSTRATP PI INK P OR (95°o CI) PLINK P OR (95% Cl)

SNPID Chr Gene Controls

LA regions rs28S6683 6|)213 HIA-DQBl 0221 0219 134 x 10 " 8S8xlθ" 199 (169-234) 1OS 191 (153-239) rs927S312 6p213 HIA-DQBl 0128 0121 199x10" 384 x 10 " 201(166-242) 1 IS 193 (148-253) rs9275390 6p2I3 HIA DQBl 0247 0239 573x10" 113x 10" 181(1 SS 211) 169 169(136-210) rs777S228 6p213 HIA DQISl 0121 0120 190 x 10 "' 170 x 10 "' 187 (154227) S 04 188 (143-247) rs239S148 6p213 CύorflO 0022 0009 150x10' 562 x 10 '" 324 (220477) 643 575(261 1266) rs9277S3S 6p213 HIA-DPBl 0232 0249 726 x 10' 828 x 10' 160 (136-187) 488 146(118181) rs38061S6 6p213 BTNU 03S0 036S 131 xlO" 127 x 10' 1 S8(l 37184) 200 146(120178) rs93571S2 61 > 21 J HIA-DQBl 0741 0745 388x 10* 72SxIO" 16S(I 37-197) 0010 135(108-172) rs3135363 6p2l 3 BTNI 2 0738 0712 446x10" 372 x 10' 1 S6(130 I 8S) 0006 137(110-172) rs9277565 6p213 HlA-DPBl 0193 0213 594 x 10" 832 x 10" 1 S8(l 34 I 87) 0001 144(1 IS 181) rs2281389 6p213 HIA-DPBl 0163 0184 1 S8x 10 ' 238 x 10' 1 S9(l 33 I 90) 0004 141 (112 179) rs660895 6p213 HIA-DKIIl 0200 0184 341 x 10' 468 x 10" 16I)(I 3S 190) 86Ox 10' I 60(127202) rs9S01626 6p213 HIA-DRA 0117 0092 134 x 10¹⁴ 8S8 x 10 " 199(169234) 6S7x 10* I 99(147270)

Legend: GWA analysis was repeated using Canadian control data only ("Alternate GWA analysis"). The results demonstrate that the associations detected in the initial GWA survey are not secondary to stratification between control sets.

Among the non-HLA loci, the strongest PBC-association signals observed were from SNPs in the IL12A (rs6441286 and rs574808, P=3.25xlO^"8; OR=I.51 and P=5.34xlO^"7;

10 OR=I.47 respectively) and IL12RB2 (rs3790567, P=8.60xl0^"8; OR=I.54) genes. A significant association with disease was also seen with a second IL12RB2 region SNP, rs3790565 (P=I.41x10 v^"6 ). Twelve other loci with association signals at P<5xlO were identified, including ST AT 4 (rsl6833239, P=2.60xl0^"5), a locus associated with risk for several other autoimmune diseases, and a downstream biological mediator of IL 12 signalling, 21 CTLA-4 (rs6748358, P=I.41x10-5), a locus previously associated with risk for PBC ., 1¹4⁴1¹5^Dand IRF5/TNPO3 (rslO488631, P=2.14xlO^"3), a locus associated with risk for systemic lupus erythematosus (SLE) 22 and inflammatory bowel disease (IBD). 23 These latter loci were more modestly associated with PBC, but their associations with risk were comparable to the IL12A and IL12RB2 loci, odds ratios ranging from 1.39-2.13.

20

STAGE II- REPLICATION AND COMBINED ANALYSIS

To validate the GWA results (Table 1), sixteen of the most strongly-associated SNPs were tested (stage II replication) in two independently-collected case/control sample sets from the US (Ha) and Canada (lib) Results of this analysis replicated (PO 05 in each of the two replication cohorts; Bonferroni corrected P-values <005 in a joint analysis of the two replication datasets) the GWA results for the rs2856683, rs2395148 and rs9277535 SNPs at three distinct sites within the HLA region as well as the rs6441286/rs574808 and rs3790567 SNPs at the IL12A and IL12RB2 loci, respectively. A combined analysis of the GWA and replication datasets was also undertaken and the results again provided strong

30 evidence of PBC association with the rs2856683/HLA-DQBl (P=1.78xlO^"19; OR=I.75), CόorflO (P=3.62xlO^"14; OR=2.87), HLA-DPBl (P=3.92xlO-n), BTNL2

(P=Ll lxlθ -v-^"9^v; OR=I.42), IL12A (P=2.42xlO -^"1¹4⁴; OR=I.54) and IL12RB2 (P=2.76xlO -v-l l OR=I 51) loci A repeat of this analysis using a PBC case subgroup stratified by AMA status revealed no significant effect of AMA status on these associations (Table S5)

Table S5. Combined (stage I and II) analysis restricted to AMA positive patients

Combined Combined

AMA positive cases n=92' 'cases and 2713 n=1031 cases and 2713 controls controls

P Odds Ratio P Odds Ratio

SNPID Chr Location Gene (95% CI) (95% CI)

A HI A regions rs28S6683 6p213 32763196 HLA-DQB 178 x 10 " 17S(I SS-I 98) 28x lO² " 179(1 S8-2O3) rs9275312 6p213 32773706 HLA-DQBl rs9275390 6p213 32777134 HLA-DQBl rs777S228 6p213 32766057 HLA-DQBl rs239SI48 6p213 32429532 CύorflO 362x 10 ¹⁴ 287(216-382) 3SSx IO " 300 (22S-400) rs9277535 6p213 33162839 HLA-DPBl 392 x 10 ¹¹ I SO (133-170) 212 x 10 ¹¹ 152 (13S-172) rs38061S6 6p213 32481676 BTNL2 111 x 10 ' 142 (127-1 S8) 12Sx IO ' 143 (127-160) rs93S71S2 6p213 32772938 HLA-DQBl rs3135363 6p213 32497626 BTN L2 rs9277565 6p213 33164875 HLA-DPBl rs2281389 6p213 33167774 HLA-DPBl rs660895 6p213 3268S3S8 HLA-DRBl rs9S01626 6p213 32508322 HLA-DRA

B Nan HI A region¹, rs6441286 3q2S 33-q26 161211572 1L12A 242 x K) ¹⁴ 1 S4 (138-172) 723 x 10 ¹³ 1 Sl (13S-170) rs3790567 Ip312 67594965 1L12RB2 276 x K) " 1 Sl (133-170) 1 11 x 10 " 153(135-173) rsS74808 3q2533-q26 161215677 IL12Λ 188 x 10 ¹³ 1 S4(l 37 173) I 34 x 10 ¹² 1 S3 (136173) rs6838639 4q27 123118615 TRPC3 23Sx K) ³ 123(108-140) 210x 10 ³ 123(108-141) rs379056S Ip312 67S83944 IL12RB2 124 x 10 ⁸ 146(128167) 4S6x 10 ' 149(130170) rs9964104 18q21 507SI69S CCDC68 213 x 10 ⁴ 12S(I 11-140) 1 S9x 10 * 126(1 12142) rs3124607 9q343 138534660 NOTCHl rsl6833239 2q32 191648505 STAT4 467 x 10 ' 16S(I 30-210) 2 S3 x 10 ^s 171 (133-219) rslO222962 4plS 32036553 PCDH7 rs2211312 13q333 107203928 tAMISSA 0608 1 OS (088-124) 038 108(091-129) rs907092 17q21 3517S78S IKZFi 761 x 10 ' 129(1 IS 144) 621 x 10 ⁶ 130(116-145) rs9303277 17q21 35229995 IKZFi rs4679904 3q261 161823S90 ARF7 1 13 x 10 ' I 38(121-1 S7) 476 x 10 ^■" 136(119-155) rs230S480 17ql2 3S31S722 GSDML rs6I40113 20pl3 691770 C20orf54 0338 108 (093 12S) 0238 1 10(094-128) rs 10488631 7q321 128381419 IRF5/T\PO1 1 S2 x K) ⁷ I S2 (130-178) 683 x 10 ¹ 150(128-176) rs67483S8 2q33 204465150 CTI 44

Legend A repeat of the combined analysis in Table 1 , restricted to AMA positive cases only, reveals no si gnificant effect of AMA status on the associations identified

Copy number effect was also assessed for the two major non-HLA associated loci, IL12A

10 and IL12RB2, and a gene dose effect identified for both loci (Tables S6 and S7). Table S6. Linear trend for risk variants across genotypes at the IL12A and IL12RB2 loci

Locus SNP Risk Allele Effect Odds Ratio (95% CI) P

IL12A rs6441286 G GG vs GT vs TT 1.527 (1.323-1.762) 7.17 X lo-⁹ rs574808 T TT vs CT vs CC 1.497 (1.288-1.74) 1.43 X 10 '

IL12RB2 rs3790567 A AA vs AG vs GG 1.511 (1.296-1.762) 1.41 X 10 ⁷

Legend: A gene dose effect can be demonstrated for the IL12A and IL12RB2 loci: using PLINK and an additive model, a clear linear trend for risk variants is seen across all genotypes.

Table S7. Genotypic effects for risk variants at the IL12A and IL12RB2 loci

Locus / SNP Risk allele Genotype Odds ratio (95% CI)

ILl 2 A I rs6441286 G GG vs TT 2.215 (1.677-2.926)

GT vs TT 1.756 (1.402-2.2)

ILl 2 A I rs574808 T TT vs CC 2.31 (1.669-3.197)

CT vs CC 1.68 (1.219-2.316)

ILl 2RB2 / rs3790567 A AA vs GG 2.373 (1.687-3.338)

AG vs GG 1.497 (1.221-1.835)

Legend: Using SAS and a general model for genotypic effects for IL12A, the 10 homozygous genotype effect is less than log-additive, while for IL12RB2 the effect is approximately log-additive.

STAGE III - FINE-MAPPING AND HAPLOTYPE ANALYSIS: IL12A, IL12RB2 & ST AT 4

To refine and further validate the two major non-HLA PBC-associated loci emerging from the GWA screen (JLl 2 A and 1L12RB2), the entire set of Canadian PBC cases and controls were geno typed with tag SNPs across these loci (Table 2 in Figure 5, and Table S8).

20 Table S8. Association data from fine-mapping of the IL12RB2, STAT4, and ILl 2 A loci.

SNP Location Minor Allele Major MAF- MAF- P Value Odds Ratio (95%CI) Allele Cascs Control

IL12RB2 RS3762317 67545170 G A 0.114 0.123 4.18X10 ' 1.097 (0.876 - 1.374)

IU2RB2 RS1890741 67555340 C G 0.187 0.176 4.27X10 ' 1.076 (0.898 - 1.291)

IL12RB2 RS1546159 67561014 T C 0.015 0.021 2.70X10^"1 1.360 (0.786 - 2.352)

IL12RB2 RS11209050 67564324 A C 0.283 0.205 2.76X10^"7 1.532 (1.301 - 1.804)

IL12RB2 RS6693065 67572606 G A 0.179 0.218 7.13X10³ 1.278 (1.069 - 1.528) IL12RB2 RS4655703 67573960 A T 0.463 0.407 1.66X10 ' 1.256(1.089-1.447)

IL12RB2 RS746389 67575391 G A 0.463 0.403 7.60X10⁴ 1.275(1.107-1.469)

IL12RB2 RS1995147 67575647 G C 0.462 0.403 8.41X10⁴ 1.273(1.105-1.467)

IL12RB2 RS10889682 67576471 C G 0.195 0.224 5.26X10² 1.188(0.998-1.414)

IL12RB2 RS17129823 67577357 C T 0.0⁷6 0.055 1.30X10² 1.425(1.076-1.886)

IL12RB2_RS1908632 67578394 G I 0.333 0.248 1.22X10⁷ 1.516(1.299-1.769)

IL12RB2 RS3790564 67579042 T G 0.133 0.164 1.57X10² 1.281 (1.048- 1.567)

IL12RB2 RS3790565 67583944 C 1 0.252 0.183 2.44X10' 1.505(1.269-1.785)

IL12RB2 RS10489625 67587008 G A 0.046 0.038 2.64X10 ' 1.219(0.861 - 1.726)

IL12RB2 RS946685 67588303 A G 0.252 0.183 1.93X10' 1.506(1.271 -1.783)

IL12RB2 RS6679356 67592782 C I 0.270 0.190 7.02X10" 1.574(1.334-1.858)

IL12RB2 RS10889684 67594609 r C 0.143 0.172 2.80X10² 1.246 (1.024 - 1.516)

IL12RB2 RS10749775 67594675 C A 0.192 0.125 1.88X10⁷ 1.655(1.368-2.003)

IL12RB2 RS3790567 67594965 A G 0.337 0.250 8.32X10⁸ 1.519(1.303-1.771)

IL12RB2 RS6695348 67599604 T C 0.333 0.247 9.87X10⁸ 1.519 (1.302 - 1.772)

IL12RB2 RS12564159 67603882 G C 0.336 0.298 2.34X10² 1.190(1.024-1.383)

IL12RB2 RS2252596 67606089 A G 0.015 0.022 1.23X10' 1.541 (0.886-2.681)

IL12RB2 RS2307145 67606115 C G 0.046 0.035 1.32X10' 1.311 (0.921 -1.865)

IL12RB2_RS2307153 67606231 A G 0.019 0.024 2.99X10 ' 1.302(0.791 -2.146)

IL12RB2 RS3790568 67608648 A G 0.052 0.052 9.69X10 ' 1.006(0.734- 1.379)

IL12RB2 RS3828069 67612161 C T 0.141 0.173 1.60X10² 1.273(1.046-1.549)

IL12RB2 RS6676606 67615554 G A 0.094 0.072 2.73X10² 1.327(1.032-1.707)

IL12RB2 RS4297265 67624923 G A 0.476 0.418 1.20X10³ 1.264 (1.097 - 1.457)

IL12RB2 RS2270614 67628609 I C 0.475 0.421 2.42X10³ 1.244(1.080-1.433)

IL12RB2 RS2229546 67634108 C A 0.354 0.325 8.41X10² 1.139(0.983-1.321)

STAT4 RSl 1676659 191627599 G A 0.036 0.050 7.71X10² 1.380(0.964-1.975)

S TA 14 RSl 1893432 191630119 G C 0.227 0.205 1.52X10' 1.141 (0.953-1.366)

STAT4 RS3024866 191631086 C I 0.275 0.264 5.14X10' 1.054(0.899- 1.236)

STAT4 RS3024851 191637067 A T 0.036 0048 1.08X10' 1.349(0.935-1.947)

S I Al 4 RS3024847 191638535 A I 0.432 0.410 2.35X10 ' 1.092(0.944- 1.264)

STAT4 RS1517352 191639709 A C 0.425 0.396 9.34X10² 1.13(0.980- 1.303)

Sl AT4 RS 13017460 191640801 A G 0.437 0.410 1.34X10' 1.115(0.967-1.285)

STAT4 RS 1400656 191643278 G A 0.035 0.050 4.30X10² 1.454(1.01-2.092)

S1AT4_RS1O168266 191644049 T C 0.226 0.196 4.15* 10 ^s 1.196 (1.008 - 1.420)

STAT4 RS7594501 191646845 A G 0.041 0.077 3.49X10 ^s 1.973(1.422-2.737)

STAT4 RS2459611 191647432 C r 0.116 0.108 4.60X10' 1.088 (0.870 - 1.361)

ST4T4 RS16833238 191647744 T A 0.109 0.099 3.33X10 ' 1.121 (0.890-1.411)

STAT4 RS 16833239 191648505 A G 0.041 0.077 6.32 X 10 ^s 1.932(1.397-2.674)

STAT4 RS3024921 191651517 A I 0.097 0.052 2.08X 10' 1.945(1.488-2.543)

STAT4 RSl 1889341 191651987 T C 0.277 0.229 1.98X10³ 1.288(1.097-1.512)

SIAI4 RS13010752 191654848 A T 0.097 0.092 6.09X10' 1.064(0.838- 1.352)

STAT4 RS6434435 191662109 A G 0.147 0.176 2.98X10² 1.240(1.021 -1.505)

S1AT4 RS10931480 191662292 G A 0.153 0.187 1.23X10² 1.275(1.054-1.543)

STAT4 RS10931481 191663097 G A 0.336 0.323 4.36X10 ' 1.061 (0.914-1.232)

STAT4 RS 13011805 191664494 T C 0.095 0.087 4.64X10 ' 1.096(0.858- 1.398) STAT4 RS4274624 191666901 C I 0.261 0.230 5.06X10² 1.183(1.000-1.399)

STAl 4 RS7574865 191672878 T G 0.272 0.222 1.21X10³ 1.309(1.112-1.541)

STAT4 RS12463658 191673589 C A 0.434 0.423 5.12X10' 1.049(0.910-1.210)

STAT4 RS10181656 191678124 G C 0.278 0.230 1.90X10³ 1.289(1.098-1.514)

STAT4 RS10497711 191722166 G A 0.049 0.069 2.1 IXlO² 1.437(1.054-1.959)

IL12AJIS17810546 161147744 G A 0.069 0.117 9.77X10^"'' 1.784(1.376-2.312)

IL12A RS9811792 161179692 C T 0.416 0.424 6.30X10 ' 1.036(0.898-1.195)

IL12A RS16830960 161184588 G A 0.051 0.029 1.51X10³ 1.766(1.238-2.521)

IL12A RS2243123 161192345 C T 0.232 0.276 5.01X10³ 1.263(1.073-1.488)

IL12A RS583911 161193084 G A 0.503 0.421 4.34X10⁶ 1.392(1.208-1.603)

IL12A RS2243133 161194967 T C 0.045 0.028 9.49X10³ 1.662 (1.128 - 2.450)

IL12A RS2243135 161195687 C G 0.342 0.406 2.51X10⁴ 1.314(1.135-1.521)

IL12A RS568408 161196161 A G 0.104 0.145 7.65X10⁴ 1.457(1.169-1.816)

IL12A RS668998 161198245 G \ 0.501 0.422 1.12X10 ^s 1.372(1.191 - 1.581)

IL12A_RS2243154 161198936 Λ G 0.108 0.081 8.25X10³ 1.375(1.085-1.743)

IL12A RS4608735 161199702 C A 0.107 0.124 1.58X10' 1.177(0.938- 1.476)

IL12A RS11927521 161199779 G A 0.344 0.407 2.65X10⁴ 1.312 (1.134 - 1.519)

IL12A RS17826053 161200323 G T 0.104 0.145 7.91X10⁴ 1.459(1.169-1.82)

IL12A RS6441284 161200962 G A 0.339 0.411 4.47X10 ^s 1.359(1.173-1.575)

IL12A RS485497 161201826 G A 0.439 0.518 1.09X10' 1.374(1.192-1.583)

IL12A RS4680536 161202965 G A 0.340 0.421 4.08X10 '^• 1.410 (1.218 - 1.632)

IL12A RS9852519 161203322 T C 0.470 0.372 2.1 IXlO⁸ 1.502(1.302-1.732)

IL12A_RS598638 161203511 T C 0.170 0.173 8.32X10 ' 1.021 (0.847 - 1.23)

IL12A RS4679867 161206597 A T 0.317 0.405 4.52X10⁷ 1.464(1.262-1.699)

IL12A RS4679868 161206848 A G 0.494 0.387 1.58X10* 1.546(1.342-1.782)

IL12A RS6441286 161211572 G T 0.494 0.390 4.83X10' 1.525(1.324-1.757)

1L12A RS574808 161215677 C r 0.320 0.412 1.15X10⁷ 1.493(1.287- 1.733)

IL12A RS589545 161216294 A G 0.322 0.413 1.34X10⁷ 1.487 (1.283 - 1.724)

IL12A RS2936298 161226577 A G 0.170 0.193 9.97X10² 1.167 (0.971 - 1.404)

IL12A RS17217081 161229792 G A 0.015 0.018 5.22X10 ' 1.205(0.68-2.134)

Legend: Association data are shown for 80 of 93 SNPs tested in the fine-mapping of Canada-based cases and controls Tests for association were carried out using PLfNK MAF corresponds to minor allele frequency Data for 9 IL23R region SNPs are not shown (all P>005)

To study the IL12A locus, 25 SNPs spanning a 80 kb region encompassing the entire gene, 38 kb 5'flank and 32 kb 3'flank segments, were genotyped. Significant associations were observed for multiple SNPs across the region, the strongest signal (P=1.58xlO^"9; OR=I.55) coming from rs4679868 at the 3'flank, a variant not assessed in the stage I survey. Of note, rsl7810546 (P=9.77xlO-6), located in the IL12A 5'flank, has also been 10 reported as a risk variant for celiac disease.24 Haplotype analyses (Table S9) across the region also identified a five-allele 3 'flank-located haplotype containing the rs4679867, rs4679868, rs6441286, rs574808, and rs589545 TAGTG alleles as a major risk haplotype for PBC (P=4.82xlO^~29). A combined analysis of the Canada and US-based samples revealed this haplotype to be present in 49.1% of all PBC cases (as compared to 32.0% in controls) and confirmed a highly significant (P=LlSxIO^"34; OR=2.01) association with risk for PBC. Table S9 IL12A Haplotype associations with PBC

Extended Canada-Ba sed Collection US-Based Collection Combined Analysis

Haplotype Frequency Frequency Freqjency

Odds Ratio Cases Controls P Cases Controls P Cases Controls P (95% Cl)

(N=621) (N=1279) (N=410) (N=310) (N=1031) (N=1589)

TAGTG 0 491 0 305 4 82 X 10-" 0 489 0 375 I 82 x 10 ' 0 491 0 320 I I 5 x 10 '" 2 01 (1 83-2 30)

AGTCA 0.309 0.318 0.569 0.326 0.424 2.0O x 10 * 0.317 0.340 0.0836 0.90 (0.80-1.01)

TGTTC 0.179 0.125 9.37 x 10 ' 0.179 0.195 0 4378 0.179 0.141 3.0O x 10 ' 1.33 (1.14-1.54)

Other 0.021 0.252 0.006 0.006 0.013 0.199 haplotypes

Legend: ILl 2A haplotype analyses were performed in the extended Canada and US-based cases/control collections and a combined analysis then performed (Haploview v.4.1). The 5-locus haplotype analysis consisted in order of the rs4679867, rs4679868, 10 rs6441286, rs574808 and rs589545 SNPs. Three haplotypes accounted for 80% of the overall haplotypes in controls. Haplotypes with an estimated frequency <20% in the control group were designated as "other haplotypes". The most significantly-associated haplotype is shown in bold.

For fine-mapping of the 1L12RB2 locus, the Canadian sample set was genotyped with 39 tag SNPs spanning 176.1kb across the IL23R-IL12RB2 region, and including previously identified IL23R risk variants associated with Crohn's disease and psoriasis.²⁵'²⁶ Significant associations derived primarily from SNPs in the IL12RB2 3'flank and intronic regions, the rs6679356 SNP showing the most significant association (P=7.02xl0^"8) with 20 PBC. In addition, a three locus risk haplotype identified in the 3'flank (GTC; present in 35.6% of cases and 25.3% of controls) was significantly associated with PBC,

P=3.07xl0 ,-π OR=I .53 (Table SlO). No significant associations were seen for the IL23R SNPs.

Table SlO: Haplotype association results for IL12RB2

Legend: Haplotype analysis for the 1L12RB2 gene was performed on the extended Canada-based collection using Haploview v.4.1. Odds ratios and 95% confidence intervals (CI) are shown. The 10-locus haplotype analysis includes (in order) SNPs

Haplotype Frequency P Odds Ratio (95% CI)

Cases Controls

(N=621) (N= 1279)

10-Locus

TGTAGTCAGC 0.519 0.565 7.40 x 10 ³ 0.83 (0.72-0.95)

TTTAGTTAGC 0.129 0.158 0.020 0.79 (0.65-0.92)

GGCAACCCAT 0.192 0.119 2.72 x 10 ⁹ 1.75 (1.45-2.10)

GGTAGCCAAT 0.080 0.063 0.054 1.29 (1.00 -1.68)

GGCAATCAAT 0.018 0.040 3.00 x 10 ⁴ 0.43 (0.27-0.68)

GGCGATCAAT 0.042 0.021 4.00 x 10^"4 1.98 (1.35-2.91)

3-Locus

ACA 0.511 0.489 0.153 1.10 (0.96 - 1.26)

GTC 0.356 0.253 3.07 x 10^"" 1 .53 (1 .32 - 1 .77)

GTA 0.117 0.076 2.53 x 10^"5 1.63 (1.30 -20.40)

GCA 0.009 0.092 6.02 x 10^"23 0.09 (0.05-0.161)

ATC 0.006 0.073 7.27 x 10^"19 0.08 (0.04-0.162) rsl908632-rs790564-rs3790565-rsl0489625-rs946685-rs6679356-rsl0889684- rsl0749775-rs3790567-rs6695348, and the 3-locus haplotype analysis includes (in order) SNPs rs4297265-rs2270614-rs2229546. The most significant haplotype association is shown in bold. 10

Although association with the STAT4 rsl6833239 SNP was not replicated in the stage II analysis, this association remained significant (P=4.67xlO^~5; OR=I .65) in the combined analysis and is of particular potential relevance to the IL12A/IL12RB2 associations. Additional genotyping was thus also undertaken across the 94.6kb STAT4 locus, the results revealing several intronic SNPs as being modestly associated with disease, the strongest association emanating from an intron 3 SNP, rs3024921, P=5.76xlO^"8 in the combined sample set (Tables S8 and SI l).

Table SI l . Replication & combined analyses showing STAT4 associations with PBC

Chr. SNP ID (Location) Physical Minor/Major US-Based Extended Combined Analysis

Location Alleles Collection Canada-Based

Collection

P P P Odds Ratio (95%

CI)

2 rslO168266 (Intron 5) 191644049 T/C 0.074 4.15 x 10 ^"5 5.95 X 10 ² 1.15 (0.99-1.32)

2 rsl6833239 (Intron 4) 191648505 A/G 0.482 6.32 x 10 ⁵ 3.69 X 10^"4 1.53 (1.21-1.93)

2 rs3024921 (Intron 3) 191651517 Λ/T 0.013 2.08 x 10 ^"6 5.76 X 10^"8 1.81 (1.45-2.26)

Legend: Association testing between SNPs in STAT4 were carried out in the US-based collection and the extended Canada-based collection using PLINK. A combined analysis was performed and allelic odds ratios and the corresponding 95% confidence intervals (CI) were calculated. Genome positions are based on the NCBI database, build 36 (hgl 8). The risk allele for each SNP is underlined and the most significant data is bolded.

10 Joint analysis of the fine-mapping SNPs across each of the IL 12A, IL12RB2 and

STA T4\oci (Table S 12) was also undertaken and revealed that for each region multiple

SNPs are required for the observed associations.

Table S 12. Stepwise selection/conditional analysis based on highest allelic associations identified as risk variants for PBC

Locus SNP Step χ² (DF=2) P

ILI 2 A rs6441286 1 24.0482 <0.001 rs485497 2 10.247 0.006 rsl7810546 3 6.5781 0.0373

IL12RB2 rsl0749775 1 28.2004 <0.001 rs 1908632 2 6.3627 0.0415

STAT4 rs3024921 1 24.8585 <0.001 rs7574865 2 17.469 0.0002 rs7594501 3 10.8064 0.0045

Legend: Using stepwise forward selection, we demonstrate that multiple alleles across the loci IL12A, IL12RB2 and ST AT 4 are associated with PBC.

20 From these results, it will be appreciated that the GWA data identify the HLA, IL12A and IL12RB2 loci as susceptibility loci for PBC, with multiple SNPs across the HLA region and several 1L12A and IL12RB2 locus SNPs achieving genomewide significance. The successful replication in independent cohorts of the most strongly-associated SNP at each of these loci, the high significance levels achieved in the combined analysis, and the identification by ancillary genotyping of multiple other PBC-associated SNPs across the ILl 2A and IL12RB2 loci, provide compelling evidence for involvement of all three loci in PBC. Associations between PBC and class II major histocompatibility complex (MHC) alleles have been previously reported,¹² but inconsistently replicated across independent studies. The data provide conclusive evidence for involvement of this region in PBC, revealing variants in the four class II HLA genes (DQBl, DPBl, DRBl, DRA) and the CόorflO and BTNL2 genes at this locus to be highly associated with disease. Among these genes, DQBl, DPBl and DRBl have been previously implicated in PBC in some association studies ²⁷ and association of a BTNL2 truncating mutation identified initially in patients with sarcoidosis ²⁸ (another granulomatous disease) has also been observed in several autoimmune diseases. This latter association may reflect strong linkage disequilibrium of the BTNL2 mutation with disease-predisposing HLA DQBl-DRBl haplotypes.²⁹'³⁰ While strong LD across this region also confounds identity of the PBC-causal allele(s) underpinning the association signals, haplotype analyses (Table S3) suggest that most of the PBC risk associated with this region derives from two common haplotypes across the CόorflO, BTNL2 and HLA-DQBl genes, AACA (p=1.09xl(T¹⁰) and CACA (P=7.29xlO^"10). By validating the class II MHC region as a major contributor to risk for PBC, our data indicate that further dissection of this region is warranted and will enable definition of allele(s) etiologically relevant to PBC.

While clearly important in PBC pathogenesis, HLA contributions to PBC risk are proportionately less significant than for other autoimmune diseases, ' the association of the HLA-DQBl locus on risk (OR=I .75) being comparable to the associations of 1L12A (OR=I.54) and IL12RB2 (OR=I.51), and the population attributable fraction (PAF) fraction (Table S 13) for the ILl 2A rs6441286 GT and GG risk genotypes (21.5%) being slightly higher than that for the HLA-DQBl rs2856683 CC and CA risk genotypes (21.2%). The IL12RB2 AA risk genotype also provides a relatively high PAR of 18.4%. These estimates are in keeping with substantive contributions of all three of these loci to risk for PBC.

The data also suggest important contributions of the IL12A and 1L12RB2 loci to PBC susceptibility. This possibility is consistent with the major immunoregulatory roles of their protein products, IL12p35 and IL12RB2, which respectively associate with the IL12p40 and ILl 2RBl chains to generate the IL 12 cytokine and its receptor, IL12R. Understanding of the pathways whereby IL 12 is induced and functions is continuing to evolve, but IL 12 binding to IL12R is thought to modulate autoimmune responses by evoking interferon gamma production and thereby inhibiting interleukin 23/IL23R- mediated induction of proinflammatory IL17-producing T helper 17 cells.³²'³³ Relevance of this pathway to PBC is also suggested by the finding of PBC association with several SNPs across the STAT4 gene, which encodes an effector integral to IL12/IL12R signaling. Interestingly, a STAT4 intron 3 SNP, rs7574865, associated with risk for rheumatoid arthritis, SLE ^{2 I} and type 1 diabetes,³⁴ was also associated with PBC in the current study (P=I .2IxIO^"3; OR=I .31), although the strongest association signals at this locus came from intronic SNPs outside the LD block tagged by rs7574865.

Similarly, another PBC-associated SNP, rsl78105416 in the IL12A gene, has been previously implicated in risk for celiac disease,²⁴ while SNPs in two other key components of the IL12/IL12R and related IL23 immunomodulatory axis, IL12B and IL23R, are associated with risk for psoriasis, psoriatic arthritis and/or IBD, although not for PBC.²⁵'²⁶' ³⁵

The GWA study was well-powered (82%) to detect associations with an OR of 1.5 (Supplementary Appendix) and, as such, provided strong evidence for PBC association with a number of HLA, ILl 2 A and IL 12 RB 2 variants. However, lower power to detect associations with more modest effects, even in the combined analysis (power estimated, for example, to be 60% for an OR of 1.4) may have impeded discovery of some PBC risk alleles. Larger and prospectively-followed sample sets will be required to identify non- HLA loci influencing risk and to elucidate the relevance of the PBC-associated loci to other clinically-relevant disease subphenotypes such as disease severity/progression. Population stratification appeared to have a minimal impact in this analysis given little inflation of genome-wide chi-square tests, and the similar levels of association provided by an analysis restricted to Canadian cases and controls.

At present, the disease-causal alleles at the PBC-associated HLA and non-HLA loci identified here remain unknown. Interestingly, the strongest association signals at the IL12A and IL12RB2 loci emanated from SNPs in the 3' flank and intronic regions, suggesting these variants may influence IL12/IL12RB2 expression. This possibility is consistent with the development of autoimmune/lymphoproliferative disease in IL12RB2 knock-out mice 36 and a recent report describing the development of biliary cirrhosis in a child with IL 12 deficiency.³⁷

These observations as well as the cumulative data linking inherited IL 12, IL12R and IFNydeficiency to increased susceptibility and severity of mycobacterial and other infectious disease, suggest that PBC-associated IL12A/IL12RB2 variants engender both an impaired response to infection, as well as a potentiated risk for autoimmunity, the former possibly driving the latter. Precise characterization of the nature and functional sequelae of the HLA and

IL12A/IL12RB2 variants conferring risk for PBC is required, but the association of PBC with variants at these loci confirms the critical role for immunogenetic factors in genesis of this disease. These data also indicate that IL12/IL12R modulation can be of benefit in the treatment of PBC.

Evaluation of IL-12A Gene Expression Levels in Healthy Subjects

Patient samples were assayed for IL- 12A transcript levels. More particularly, human peripheral blood mononuclear cells were purified from healthy donors and genotyped so as to identify individuals homozygous for the primary biliary cirrhosis risk (TAGTG) and non-risk (AGTCA) IL12A haplotypes. Representative data from two subjects with non- risk genotype (1 and 2) and from 2 with the risk genotype (3 and 4) are shown. Cells were stimulated with lipopolysaccharide (LPS) and interferon-γ for 8 hours and harvested for total RNA purification using with Qiagen kit. A two-step qPCR were performed to detect the expression of IL12A (p35) and the housekeeping gene, glyceraldehyde 3- phosphate dehydrogenase (GAPDH). Primers for IL 12 A:

5TATTGATGAGCTGATGCAGGCS' (SEQ ID NO.88) 5'GTCACTGCCCGAATTCTGAAAS' (SEQ ID NO.89)

Primers for GAPDH:

5ACCTCAACTACATGGTTTAC3' (SEQ ID No.90) 5'GAAGATGGTGATGGGATTTCS' (SEQ ID NO.91) The results are presented in Figure 8 as IL12A relative to GAPDH expression levels. It will be appreciated from these results that subjects at risk for PBC, as determined by the genotyping method herein, also present with reduced IL- 12A transcript levels, and correspondingly reduced IL- 12A protein levels, relative to subjects having a non-risk genotype.

Further genotyping analysis

Continued replication and fine mapping studies have also revealed association of primary biliary cirrhosis with the 17ql2-21 and MMELl loci and a high risk two-allele haplotype at the IRF5-TNPO3 locus. More particularly, genome-wide association data have confirmed HLA, IL12A and IL12RB2 as risk loci for primary biliary cirrhosis (PBC). By genotyping additional subjects for SNPs showing suggestive GWA significance (P<lxlO^"4), we identified 17ql2-21 (P=3.50xl0^'13), IRF5-TNPO3 (P=8.66xlO^"13), and MMELl (P=2.28xlO^"9) as PBC risk loci and defined a high risk two- allele haplotype at the IRF5-TNPO3 locus (P=7.79xlCT³¹). These results expand identified PBC risk loci and the evidence for genetic overlap among autoimmune diseases.

In particular, we have now tested another panel of 857 PBC cases of Caucasian European descent, together with an additional 3198 controls of Caucasian European descent (including 1860 historic control subjects from the New York Cancer Project [43]) for PBC associations with 36 SNPs, across 24 loci, in which association signals at P <lxlθ^~ were detected in the initial GWA survey. The combination of these replication data and our prior genome- wide association data provided a genetic dataset derived from 1351 PBC cases and 4700 controls (See Supplementary Methods).

The following methods were applied in this subsequent study:

Subjects

The study was approved by local institutional ethics committees at each participant clinical center. All cases were of Caucasian European descent, and were diagnosed with PBC according to American Association for the Study of Liver Disease guidelines [1 , 2]. Overall 92% of patients were female, with an average age of 60 years, and 6.5% of patients were AMA negative. All cases were Caucasian of European descent who were ascertained via: a) the original Canadian GWAS case collection, n=494; b) an additional recruitment of North American cases, n=720 or; c) from Europe, n=137. Controls were all healthy adults of Caucasian European descent and included the original GWAS control collection (n=1502), a second set of Canadian controls (n=972), and additional controls ascertained from Europe (124) and the USA (2102), the latter of which include 1860 "historic" controls from the New York Cancer Project previously genotyped at the Feinstein Institute using Illumina 500K BeadChips.

SNP genotyping and marker selection Individual samples with genotype call rates of less than 95% and SNPs with call rates of less than 95%, minor allele frequencies of less than 1%, or deviation from the Hardy- Weinberg equilibrium at P<0.001, were removed from the analysis.

Genotyping for the replication studies was performed using a Sequenom MassARRAY iPLEX platform and SNP annotation was based on NCBI dbSNP Build 129. 36 SNPs were genotyped across 24 loci in which association signals at P <lxlθ^~ in the initial GWA survey had been identified. GWA data were obtained using the Illumina Human Hap370 BeadChip as previously reported [2]. Genotype data quality was assessed by genotype clustering using Sequenom Typer Analyzer, discordance in the duplicate genotyping samples and genotyping call rate (>95%). To delineate the genetic variation across the IRF 5 locus, resequencing of this gene (including intronic and exonic regions and 5' and 3' flanking regions) was performed in genomic DNA pools prepared from 100 subjects. Four long-range PCRs were designed with Primer3 (http://frodo.wi.mit.edu/primer3/input.htm) so as to cover the 24.8-kb IRF5 genomic region (chr7: 128,361 ,145-128,385,948), the DNA then amplified using the SequalPrep™ Long PCR Kit (Invitrogen) and the purified PCR products pooled for sequencing. DNA sequencing library preparation, cluster generation were performed using DNA library preparation and cluster generation kits from Illumina and sequencing then undertaken on the Illumina/Solexa GAII platform using a 36 bp sequencing protocol. Sequence alignment and polymorphism identification, performed using Velvet and SoftGenetics SeqAlign program (SoftGenetics), confirmed prior reports of 44 polymorphisms across this region, but failed to identify any novel variants. Of these 5 SNPs failed repeatedly in genotyping assays and 4 did not show polymorphisms (MAF <0.001 %). Thus the fine- mapping studies incorporated 35 test polymorphisms, including two the IRF5 promoter CGGGG (4/5) and 30-bp (rs60343456) indel polymorphisms that were assayed using an ABI3700 as previously reported [45]. Auto-antibody assays

Titres of anti-splOO and gp210 antibodies were evaluated by QUANTA Lite™ (ELISA) assays using frozen serum samples, and the recommended protocol (INOVA Diagnostics, Inc. CA, USA).

Statistical analyses

Allelic associations were assessed by means of PLINK software (vl .07), as were the haplotype associations. To identify a parsimonious model for association of haplotypes with risk for PBC, we used a forward-selection approach adding one SNP at a time to the top SNP or combination of SNPs. Tests of haplotypes included comparing the frequency of the presence of each specific haplotype versus its absence, which yielded a 1 -degree of freedom test, and an omnibus test which jointly compares the frequencies of each haplotype in the cases to the frequency in the controls. This yields a chi-square test having degrees of freedom equal to the number of haplotypes minus 1. An R-script (www.r-project.org/) was used to generate the upper panel of Figure 1. Linkage disequilibrium between pairs of SNPs was determined with the use of Haploview software, version 4.1 (www.broad.mit.edu/mpg/haploview). Haplotype block structure was defined according to the criteria established by Gabriel and the pairwise estimates of standardized Lewontin's disequilibrium coefficient (D'), whereas the linkage disequilibrium among pairs of SNPs was characterized according to the square of the correlation coefficient (r²). Combined analyses were carried out with Cochran-Mantel- Haenszel tests, with adjustment for potential confounding of the case or control frequency with genotype frequency among groups of subjects according to the stage of analysis or center. For replication analysis the North American and European samples were stratified, and for the combined analyses we stratified by North American, European and GWA samples. Conditional logistic regression analyses of multiple markers were performed using SAS v9.13 (SAS Institute Inc.).

We used an in silico approach to check if the candidate genes directly interact with each other. Pathway Studio software (www.ariadnegenomics.com) was used to construct a network of interaction between genes (direct regulation of gene expression, protein/protein binding or promoter binding).

Results

Fifteen SNPs replicated (P<1.4xlO^~3) in this cohort after Bonferroni adjustment. In addition to HLA, IL12A and IL12RB, genes at the 1RF5-TNPO3, MMELl and chromosome 17q 12-21 (IKZF3, ZPBP2, GSDMB and ORMDLS) loci showed significant association with PBC in both the replication analysis and in an analysis combining the replication and initial GWA datasets, as shown below: SNP Chr. Position Gene Risk Allele rs3748816 I p36 2516606 MMELl C rs3890745 I p36 2543484 MMELI G rs 1 1588497 I p35.1 32803048 ZBTB8 G rs3790565 I p31.2 67583944 IL 12 RB 2 C rs3790567 I p31.2 67594965 ILl 2 RB 2 A rsl6833239 2q32.2-q32.3 191648505 STAT4 A rs231735 2q33 204402121 CTLA4 G rs6748358 2q33 204465150 CTLA4 A rsl 869241 3p26.1 5729732 EDEMl T rs619481 3p26-p24 8182706 LMCDl C rslO25818 3p24.3 16940025 PLCL2 G rs6441286 3q25.33-q26 16121 1572 ILl 2 A G rs4679904 3q25.33 161823590 ARL l 4 T rs9277535 6p21.3 33162839 HLA-DPBl G rsl 3216266 6q23.3 140059450 CITED2 T rsl 0250643 7pl 5 24502175 MPP6 G rsl 0488631 7q32.1 128381419 IRE 5 -TN P 03 C rsl 760 8p23.3 432079 C8orf42 G r_s7834444 8q22 99455892 KCNS2 τ rs7838130 8q22 99456091 KCNS2 T rs4259391 8q22 99460461 KCNS2 T rs l 0504989 8q22.2 100501854 VPS13B T rs7477689 10pl l .2 30055806 SVIL G rs3858359 I l q21-q22 91 148975 MTNRlB C rs221 1312 13q33-q34 107203928 LIG4 C rs907092 17ql 2-21 35175785 IKZE3 A rs9303277 17ql 2-21 35229995 IKZF3 C rsl 1557467 17q l 2-21 35282160 ZPBP2 G rs8067378 17q l 2-21 35304874 GSDMB A rs2305480 17q l 2-21 35315722 GSDMB T rs2290400 17q l 2-21 35319766 GSDMB A rs7216389 17q l 2-2 l 35323475 GSDMB T rs4795405 17ql 2-21 35341943 ORMDU T rs9964104 18q21 50751695 CCDC68 C rs3745516 19q l3.3-ql3.4 55618554 SPIB A rs61401 13 2Op 13 691770 C20oφ4 T

Among the PBC loci confirmed by this analysis, the IRF5-TNPO3 (interferon regulatory factor 5-transportin 3) locus is of particular interest because of the integral immunoregulatory roles for IRF5, prior association of this locus with systemic lupus erythematosus [44, 45], systemic sclerosis [46] and Sjδgrens syndrome [47], and the strong effect on risk of the disease-associated SNP at this locus in the replication (PREP= 1.13xlO^"8) and combined datasets

10^{" 13}; OR 1.57). Fine-mapping studies were thus initiated, wherein 1330 cases and 1833 controls were genotyped for 35 SNPs across this locus. Among the associations identified, as shown below, the strongest signals were provided by rsl2539741 and rs2070197 alleles. SNP Position Risk allele rs729302 128356196 C

IRF5 rs 11768806 128360364 T rs4728142 128361203 A rsl 874330 128362460 G rs3778754 128362788 C rs6968563 128364967 C

STR CGGGG 128365169 G

IRF5 rs2004640 128365537 G rs3807307 128366438 G IRF5_rs752637 128366656 A rs3823536 128366902 T rs3778752 128367283 A rs3778751 128367284 A rs55643109 128367411 T rs3807306 128367916 C rsl 1767834 128368512 T rsl 1761 199 128369071 G rs6975315 128369721 A rs7808907 128371320 T rsl 874328 128372340 C rsl 874327 128372612 T rs35000415 128372852 T

IndeBO 128374500 A lRF5_rs2070197 128376236 C rsl 0954213 128376663 G rsl 1770589 128376724 A rsl 0954214 128376869 C IRF5_rs 13242262 128378600 A rslO48863O 128381184 G rsl 0488631 128381419 C rs2280714 128381961 G rs3847098 128382385 G rsl 1761242 128382842 C rsl 2539741 128384041 T rs l 0236569 128384823 T

These variants map just 3 ' to the IRF5 coding region, are in tight linkage disequilibrium (r >0.95) with one another, and their associations with PBC reach P values of 1.65x10 -10

(OR=I.63) and 3.74x10 v^"lO (OR=I .62), respectively, in the combined cohort analysis. These polymorphisms are also among the SNPs at this locus highly associated with SLE and are both correlated with changes in IRF5 expression in transformed B cells [44], By contrast, an insertion-deletion polymorphism 64bp upstream of IRF5 exon IA representing a putative disease-causal variant for SLE, showed only modest association with PBC (P=LOOxIO^"3) and no associations were detected between PBC and another SLE-associated SNP, rs2004640, or a 3 ^' UTR variant, rsl 0954213, identified as a key predictor of IRF 5 expression level [44]. The analyses also identified a number of haplotypes across the IRF5-TNPO3 locus that were more strongly associated with PBC than any of the individual test SNPs. Among these, a two-allele haplotype involving the rs35000415 and rsl2539741 SNPs showed the strongest association (omnibus P=7.79xlO^'31, OR=I.97 for the TT haplotype) with risk for PBC.

Haplotype analysis across the IRF5-TNPO3 locus reveals a 2 SNP and 3 SNP haplotype strongly associated with PBC 2-SNP Haplotype

HAPLOTYPE DF P SNPs

TT 1 5.58xlO^"19 rs35000415|rsl2539741

TC 1 4.62xlO^"16 rs35000415|rsl2539741

CC 1 2.82xlO^"08 rs35000415jrsl2539741

OMNIBUS 2 7.79XlO^'11 rs35000415|rsl2539741

3-SNP Haplotype

HAPLOTYPE DF P SNPs

TCT 1 4.24x l O-'⁹ rs35000415|IRF5_rs2070197|rsl2539741

TTC 1 8.87xlO^{' i6} rs35000415|IRF5j-s2070197irs 12539741

CTC 1 2.85xlO -,-^"0^U8^δ rs35000415|IRF5_rs2070197|rsl2539741 OMNIBUS 2 1.13x1 (T" rs35000415|IRF5_rs2070197|rs 12539741

PLINK was used to evaluate haploype association by a forward-selection approach adding one SNP at a time to the top SNP or combination of SNPs. The omnibus test evaluated deviation from an equal frequency of haplotypes in cases and controls, using a goodness of fit chi-square test. The two SNP haplotype rs35000415-rsl2539741 showed greatest association with PBC, when compared with controls, with an odds ratio of 1.97. CHISQ= χ² and DF=degrees of freedom.

A second region of interest emerging from these studies is the Hq 12-21 locus, with all 8 of the tested SNPs across this region achieving significance in the replication analysis (PREP values between 1.78xlO^"9 and 1.88xlO^"5). This chromosomal region has also been associated with asthma [48], Crohn's disease [49] and type 1 diabetes [50] and contains four genes, ZPBP2, IKZF3, GSDMB, and ORMDL3, which respectively encode the zona pellucida-binding protein 2, the IKAROS family zinc finger 3 protein involved in leukocyte development and IgE production, Gasdermin-B involved in epithelial barrier function and ORMl -like protein 3, an uncharacterized protein belonging to a family of transmembrane proteins anchored in the endoplasmic reticulum. All 8 of the SNPs tested at this locus were in linkage disequilibrium (pair-wise r² range of 0.66-0.96), but the strongest association signal came from a ZPBP2 SNP, rsl 1557467 (P_COM=3.50X10^{~ 13}).

The replication and combined association data also reveal a strong association of PBC with two SNPs (rs3890745 and rs3748816) at the MMELl (membrane metallo- endopeptidase-like 1) locus on chromosome Ip36 (combined P=2.28x10^" and P=3.15xlO^'8, respectively). These SNPs are in linkage disequilibrium with one another (r²>0.88) and one (rs3748816) is a non-synonymous SNP in exon 16 that encodes a potentially functional Met to Thr substitution, while the other (rs3890745) maps within intron 2 of the MMELl gene and has been associated with risk for rheumatoid arthritis and for celiac disease [51 ,52].

A suggestive association signal

OR=I .27) was also observed in the combined case/control cohort for rs3745516, an intronic SNP in SPIB, a gene encoding the Spi-B transcription factor. This SNP did not achieve significance in the replication cohort after Bonferroni correction (P_REP=6.21X10^~3), but the strength of the association in the combined analysis and Spi-B roles in dendritic cell development [53] and B-cell receptor signalling, are in keeping with a potential relevance of this locus to PBC pathogenesis.

Anti-mitochondrial antibodies (AMA) are found in most patients with PBC, but no correlation of specific PBC genotypes with AMA status was observed in our prior [2] or current association studies. PBC is also associated with specific anti-nuclear antibodies, some 20% of patients manifesting anti-glycoprotein-210 antibodies (anti-gp210) directed against the human nuclear pore complex, or anti-splOO antibodies that recognize a 53- kDa nuclear antigen [14]. Evaluation of genotype status in the subset of cases (462) typed for anti-nuclear antibodies revealed a strong association of the HLA locus (rs9277535 at HLA-DPl) with disease (P=4.25xlO^~8; OR=2.25) in anti-splOO antibody-positive patients. By contrast, no genetic distinctions were apparent for the anti-gp210 positive subgroup.

In conclusion, the current data confirm associations of PBC with the HLA, IL12A and IL12RB2 loci and identify IRF5-TNPO3, 17ql2-21, and MMELl as three additional risk loci for PBC. Our data also identify IRF5-TNPO3 haplotypes that are highly associated with PBC and suggest that some genetic substructure may exist in PBC in relation to anti- splOO antibody status. A search for functional relationships between the various PBC- associated loci was also undertaken using Pathway Studio software to construct a network of interactions between candidate genes identified in our study (data not shown). This revealed potential interactions between only 3 of the disease-associated genes (ILl 2A, IL12RB2 and STAT4). Thus, it appears that many of the candidate genes identified here does not directly interact, but instead, act independently in conferring risk for PBC.

All references cited herein, including all database references and the sequence information referenced therein, are hereby incorporated herein in their entirety. REFERENCES

1. Kaplan MM, Gershwin ME. Primary biliary cirrhosis. N Engl J Med 2005; 353: 1261-73.

2. Poupon RE, Poupon R, Balkau B. Ursodiol for the long-term treatment of primary biliary cirrhosis. The UDCA-PBC Study Group. N Engl J Med 1994; 330: 1342-47.

3. Gershwin ME, Mackay IR. The causes of primary biliary cirrhosis: Convenient and inconvenient truths. Hepatology 2008; 47: 737-45.

4. Irie J, Wu Y, Wicker LS et al. NOD.c3c4 congenic mice develop autoimmune biliary disease that serologically and pathogenetically models human primary biliary cirrhosis. J Exp Med 2006; 203: 1209-19.

5. Oertelt S, Lian ZX, Cheng CM et al. Anti-mitochondrial antibodies and primary biliary cirrhosis in TGF-beta receptor II dominant-negative mice. J Immunol 2006; 177: 1655-60.

6. Wakabayashi K, Lian ZX, Moritoki Y et al. IL-2 receptor alpha(-/-) mice and the development of primary biliary cirrhosis. Hepatology 2006; 44: 1240-49.

7. Selmi C, Mayo MJ, Bach N et al. Primary biliary cirrhosis in monozygotic and dizygotic twins: genetics, epigenetics, and environment. Gastroenterology 2004; 127: 485-92.

8. Jones DE, Watt FE, Metcalf JV, Bassendine MF, James OF. Familial primary biliary cirrhosis reassessed: a geographically-based population study. J Hepatol

1999; 30: 402-7.

9. Watt FE, James OF, Jones DE. Patterns of autoimmunity in primary biliary cirrhosis patients and their families: a population-based cohort study. QJM 2004; 97: 397-406.

10. Lazaridis KN, Juran BD, Boe GM et al. Increased prevalence of antimitochondrial antibodies in first-degree relatives of patients with primary biliary cirrhosis. Hepatology 2007; 46: 785-92.

1 1. Donaldson PT, Baragiotta A, Heneghan MA et al. HLA class II alleles, genotypes, haplotypes, and amino acids in primary biliary cirrhosis: a large-scale study. Hepatology 2006; 44: 667-74. T

12. Invernizzi P, Selmi C, PoIi F et al. Human leukocyte antigen polymorphisms in Italian primary biliary cirrhosis: A multicenter study of 664 patients and 1992 healthy controls. Hepatology 2008; 48: 1906-12.

13. Donaldson P, Veeramani S, Baragiotta A et al. Cytotoxic f-lymphocyte-associated antigen-4 single nucleotide polymorphisms and haplotypes in primary biliary cirrhosis. Clin Gastroenterol Hepatol 2007; 5: 755-60.

14. Juran BD, Atkinson EJ, Schlicht EM, Fridley BL, Lazaridis KN. Primary biliary cirrhosis is associated with a genetic variant in the 3' flanking region of the CTLA4 gene. Gastroenterology 2008; 135: 1200-6. 15. Walker EJ, Hirschfield GM, Xu C et al. CTLA-4/ICOS gene variants and haplotypes are associated with rheumatoid arthritis and primary biliary cirrhosis in the Canadian population. Arthritis and Rheumatism 2009; in press.

16. Heathcote EJ. Management of primary biliary cirrhosis. The American Association for the Study of Liver Diseases practice guidelines. Hepatology 2000; 31 : 1005-13.

17. Amos CI, Wu X, Broderick P et al. Genome-wide association scan of tag SNPs identifies a susceptibility locus for lung cancer at 15q25.1. Nat Genet 2008; 40, 616-22.

18. Purcell S, Neale B, Todd-Brown K et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81, 559- 75.

19. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 2006; 38: 904-9.

20. Gabriel SB, Schaffner SF, Nguyen H et al. The structure of haplotype blocks in the human genome. Science 2002; 296: 2225-9.

21. Remmers EF, Plenge RM, Lee AT et al. STAT4 and the risk of rheumatoid arthritis and systemic lupus erythematosus. N Engl J Med 2007; 357: 977-86.

22. Sigurdsson S, Nordmark G, Goring HH et al. Polymorphisms in the tyrosine kinase 2 and interferon regulatory factor 5 genes are associated with systemic lupus erythematosus. Am J Hum Genet. 2005;76:528-37.

23. Dideberg V, Kristjansdottir G, Milani L et al. An insertion-deletion polymorphism in the interferon regulatory factor 5 (IRF5) gene confers risk of inflammatory bowel diseases. Hum MoI Genet. 2007;16:3008-16.

24. Hunt KA, Zhernakova A, Turner G et al. Newly identified genetic risk variants for celiac disease related to the immune response. Nat Genet 2008; 40: 395-402.

25. Duerr RH, Taylor KD, Brant SR et al. A genome-wide association study identifies IL23R as an inflammatory bowel disease gene. Science 2006; 314: 1461-3.

26. Cargill M, Schrodi SJ, Chang M et al. A large-scale genetic association study confirms IL12B and leads to the identification of IL23R as psoriasis-risk genes. Am J Hum Genet 2007; 80: 273-90.

27. Invernizzi P, Selmi C, Mackay IR, Podda M, Gershwin ME. From bases to basis: linking genetics to causation in primary biliary cirrhosis. Clin Gastroenterol Hepatol 2005; 3: 401-10.

28. Valentonyte R, Hampe J, Huse K et al. Sarcoidosis is associated with a truncating splice site mutation in BTNL2. Nat Genet 2005; 37: 357-64.

29. Traheme JA, Barcellos LF, Sawcer SJ et al. Association of the truncating splice site mutation in BTNL2 with multiple sclerosis is secondary to HLA-DRBl * 15. Hum MoI Genet 2006; 15: 155-61.

30. Mochida A, Kinouchi Y, Negoro K et al. Butyrophilin-like 2 gene is associated with ulcerative colitis in the Japanese under strong linkage disequilibrium with HLA-DRBl * 1502. Tissue Antigens 2007; 70: 128-35.

31. Todd JA, Walker NM, Cooper JD et al. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet 2007; 39: 857-64. 32. Goriely S, Neurath MF, Goldman M. How microorganisms tip the balance between interleukin-12 family members. Nat Rev Immunol 2008; 8: 81-6.

33. Hoeve MA, Savage ND, de Boer T et al. Divergent effects of IL- 12 and IL-23 on the production of IL-17 by human T cells. Eur J Immunol 2006; 36: 661-70.

34. Fung E, Smyth DJ, Howson JM et al. Analysis of 17 autoimmune disease associated variants in type 1 diabetes identifies 6q23/TNFAIP3 as a susceptibility locus. Genes and Immunity. 2008 Dec 25. PMID 191 10536. 35. Fisher SA, Tremelling M, Anderson CA et al. Genetic determinants of ulcerative colitis include the ECMl locus and five loci implicated in Crohn's disease. Nat Genet 2008; 40: 710-2.

36. Airoldi I, Di Carlo E, Cocco C et al. Lack of I112rb2 signaling predisposes to spontaneous autoimmunity and malignancy. Blood 2005; 106: 3846-53.

37. Pulickal AS, Hambleton S, Callaghan MJ et al. Biliary cirrhosis in a child with inherited interleukin-12 deficiency. J Trop Pediatr 2008; 54: 269-71.

38. Filipe-Santos O, Bustamante J, Chapgier A et al. Inborn errors of IL-12/23- and IFN-gamma-mediated immunity: molecular, cellular, and clinical features. Semin Immunol 2006; 18: 347-61.

39. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 2007; 447: 661-78.

40. Lindor KD et al. Hepatology. 2009;50:291-308.

41. Hirschfield GM et al. N Engl J Med. 2009;360:2544-2555.

42. Mitchell MK, Gregersen PK, Johnson S, Parsons R, Vlahov D. J Urban Health. 2004;81 :301-310.

43. Graham RR et al. Proc Natl Acad Sci U S A. 2007;104:6758-6763.

44. Sigurdsson S et al. Hum MoI Genet. 2008;17:872-881.

45. Dieude P et al. Arthritis Rheum. 2009;60:225-233. 46. Miceli-Richard C et al. Arthritis Rheum. 2009;60: 1991 -1997.

47. Bouzigon E et al. N Engl J Med. 2008;359: 1985-1994.

48. Barrett JC et al. Nat Genet. 2008;40:955-962.

49. Barrett JC et al. Nat Genet. 2009; 41 :703-707.

50. Raychaudhuri S et al. Nat Genet. 2008;40: 1216-1223.

51. Coenen MJ et al. Hum MoI Genet. 2009; 18:4195-4203.

52. Schotte R, Nagasawa M, Weijer K, Spits H, Blom B. J Exp Med. 2004;200: 1503- 1509.

53. Muratori L, Granito A, Muratori P, Pappas G, Bianchi FB. Clin Liver Dis. 2008;12:261-76; vii. 54. Snir O et al. Ann Rheum Dis. 2009;68:736-743.

Claims

CLAIMS:

A method for identifying a subject having or at risk for primary biliary cirrhosis (PBC), wherein a biological sample obtained from the subject is assessed to identify at least one of the following characteristics:

(a) at least one polymorphism within a gene for one or both of IL-12A and IL 12-RB; and

(b) a level of IL- 12A protein, IL- 12A protein complex or IL- 12 A transcript that is reduced relative to a healthy individual, the presence of said at least one characteristic indicating the subject has, or is at risk for PBC.

2. The method according to claim 1, wherein the biological sample is assessed by a method comprising genotyping DNA from said subject to identify at least one polymorphism within a gene encoding one or both of IL-12A and IL- 12RB, wherein the presence thereof is indicative of PBC.

3. The method according to claim 2, wherein the at least one polymorphism occurs within a gene encoding IL- 12 A.

4. The method according to claim 3, wherein the at least one polymorphism in the IL- 12A gene is selected from the group of risk alleles consisting of:

Risk

SNP ID Location Allele Physical Location rsl7810546 5 'flanking A 161147744 rs583911 intron 2 G 161193084 rs668998 3 'flanking G 161 198245 rs6441284 3 'flanking A 161200962 rs485497 3 'flanking G 161201826 rs4680536 3 'flanking A 161202965 rs9852519 3 'flanking C 161203322 rs4679867 3 'flanking T 161206597 rs4679868 3 'flanking A 161206848 rs6441286 3 'flanking G 16121 1572 rs574808 3 'flanking T 161215677 rs589545 3 'flanking G 161216294

5. The method according to claim 4, wherein the at least one polymorphism in the IL- 12A gene is selected from the group of IL-12 A SNPs consisting of risk allele G at location 16121 1572 and risk allele T at location 161215677.

6. The method according to claim 5, wherein the at least one polymorphism in the IL- 12A gene is any one or more of the polymorphisms in the 5-locus haplotype represented by SNPs rs4679867, rs4679868, rs6441286, rs574808 and rs589545 SNPs.

7. The method according to claim 6, wherein the at least one polymorphism is comprises the 5 locus haplotype TAGTG.

8. The method according to claim 1, wherein the at least one polymorphism occurs within a gene encoding IL-12RB2.

9. The method according to claim 8, wherein the at least one polymorphism in the IL-12RB2 gene is selected from the group consisting of:

Risk

SNP ID Location Allele Physical Location rsl 1209050 intron 3 A 67564324 rsl908632 intron 8 G 67578394 rs3790565 intron 8 C 67583944 rs946685 intron 8 A 67588303 rs6679356 intron 9 C 67592782 rsl 0749775 intron 9 C 67594675 rs3790567 intron 9 A 67594965 rs6695348 intron 9 T 67599604

10. The method according to claim 8, wherein the at least one polymorphism in the IL-12RB2 gene is selected from the group of IL-12RB2 SNPs consisting of risk allele A at location 67594675 and risk allele C at location 67583944.

1 1. The method according to any preceding claim, comprising the further step of genotyping DNA from said subject to identify at least one polymorphism within a locus for one or more of an HLA region, STAT4 and CTLA4, IRF5-TNOP3, 17q 12-21, and MMELl, wherein the presence thereof is indicative of PBC.

12. A method for identifying a subject having or at risk for primary biliary cirrhosis (PBC) according to claim 1 , wherein a biological sample obtained from the subject is assayed to determine the level of IL-12 A protein or IL- 12A transcript, the presence of a reduced level thereof indicating the subject has, or is at risk for PBC.

13. The method according to claim 12, wherein the biological sample is assayed to determine the level of IL-12A protein.

14. The method according to claim 13, wherein the level of IL-12A protein is determined using an antibody that binds selectively therewith.

15. The method according to claim 13, wherein the biological sample is assayed to determine the level of IL-12A protein by determining the level of IL-12 A protein complex.

16. The method according to claim 15, wherein the level of IL-12 A complex is determined by assaying for IL- 12 and IL-35.

17. The method according to claim 12, wherein the biological sample is assayed to determine the level of IL-12 A transcript.

18. A kit useful to identify a subject having or at risk for primary biliary cirrhosis (PBC), the kit comprising primers useful to amplify an IL- 12A gene locus comprising at least one risk allele for PBC, and instructions for the use thereof in identifying at least one polymorphism indicative of PBC.

19. The use of an agent effective to stimulate the IL- 12/IL- 12R axis in the treatment of an autoimmune disease characterized by reduced levels of IL- 12 A protein or IL-12 A transcripts.

20. The use according to claim 19, wherein the agent is IL- 12A.

21. The use according to claim 20, wherein the agent is an IL- 12A protein complex selected from IL- 12 and IL-35.

22. The use according to any one of claims 19, 20 and 21, wherein the autoimmune condition is primary biliary cirrhosis.

23. A method for treating a subject having or at risk for primary biliary cirrhosis, comprising administering to a subject in need thereof a treatment effective amount of a composition comprising IL- 12A.

24. A method for identifying a subject having or at risk for primary biliary cirrhosis (PBC), wherein a biological sample obtained from the subject is assessed to identify at least one of the following alleles:

SNP ID Allele Physical Location Gene rs 17810546 Λ 161 147744 IL- 12A rs58391 1 G 161 193084 IL- 12A rs668998 G 161 198245 I L- 1 2Λ rs6441284 A 161200962 I L- 1 2Λ rs485497 G 161201826 IL- 12A rs4680536 Λ 161202965 IL- 12A rs9852519 C 161203322 IL- 12Λ rs4679867 T 161206597 IL- 12A rs4679868 A 161206848 IL- 12A rs6441286 G 16121 1572 IL- 12Λ rs574808 T 161215677 IL- 12A rs589545 G 161216294 IL- 12A rs 1 1209050 A 67564324 IL- 12RB2 rs 1908632 G 67578394 IL- 12RB2 rs3790565 C 67583944 IL- 12RB2 rs946685 A 67588303 IL- 12RB2 rs6679356 C 67592782 IL- 12RB2 rs 10749775 C 67594675 IL- 12RB2 rs3790567 A 67594965 IL-12RB2 rs6695348 T 67599604 IL- 12RB2 rs2856683 C 32763196 HLA-DQBl rs2395148 A 32429532 Cόorfl O rs9277535 G 33162839 HLA-DPBl rs3806156 A 32481676 BTNL-2 rslO168266 T 191644049 STA T4 rsl 6833239 G 191648505 S I A 14 rs3024921 A 191651517 STA T4 rs907092 A 35175785 IKZF3 rs9303277 C 35229995 IKZl- 3 rsl 1557467 G 35282160 ZPBP2 rs8067378 A 35304874 GSDMB rs2305480 T 35315722 GSDMB rs2290400 A 35319766 GSDMB rs7216389 T 35323475 GSDMB rs4795405 T 35341943 ORMDL3 rs3748816 C 2516606 MMELl rs3890745 G 2543484 MMELl rs35000415 T 128372852 IRF5-TNPO3 rs2070197 C 128376236 IRl 5 rsl0488631 C 128381419 IRΓ5- ΓNPO3, and rsl 2539741 T 128384041 IRΓS- I NPO3. the presence of said at least one allele indicating the subject has, or is at risk for PBC.