WO2009126740A1

WO2009126740A1 - Methods for screening for genetic predisposition to type i diabetes

Info

Publication number: WO2009126740A1
Application number: PCT/US2009/039958
Authority: WO
Inventors: William Hagopian; Hui Peng
Original assignee: Pacific Northwest Research Foundation; PACIFIC NORTHWEST Res Inst
Current assignee: Pacific Northwest Research Foundation; PACIFIC NORTHWEST Res Inst
Priority date: 2008-04-09
Filing date: 2009-04-08
Publication date: 2009-10-15
Anticipated expiration: 2010-10-09
Also published as: US20090311697A1; EP2279411A4; US8268561B2; EP2279411A1; CA2718986A1

Abstract

A method of genetically screening large numbers of individuals to identify those individuals requiring follow-up testing for active Type I diabetes (T1D) is provided. The method includes obtaining a nucleic-acid containing biological sample from each individual and testing for the presence of specific combinations of HLA II alleles in the sample.

Description

METHODS FOR SCREENING FOR GENETIC PREDISPOSITION TO TYPE I DIABETES

FIELD OF THE INVENTION

This application relates to methods for identifying individuals at risk of developing type I diabetes mellitus (TlD). More particularly, this application relates to methods for rapid, cost-effective, genetic screening of large numbers of individuals in order to identify those who should receive subsequent testing for active TlD.

BACKGROUND

Diabetes is a disorder characterized by persistent variable hyperglycemia due to either inadequate production of insulin by the body and/or an inadequate response by the body to insulin. Type I diabetes mellitus (TlD; also known as juvenile onset diabetes or insulin dependent diabetes) is an autoimmune disorder that typically develops in susceptible individuals during childhood, and involves progressive destruction of insulin- producing cells in the Islets of Langerhans of the pancreas. Patients with clinical TlD require regular insulin replacement therapy. Currently millions of people suffer from Tl D with the overall incidence increasing at about 3-5% per year in most populations. While approximately 50% of the background risk of TlD is believed to be due to environmental factors, the remainder is due to genetic causes with up to 20 different genes influencing susceptibility to the disorder. Of the genetic influence, approximately 50% appears to involve genetic variations within the human leukocyte antigen (HLA) class II alleles HLA-DR and HLA-DQ.

Large-scale destruction of insulin-producing cells will already have taken place by the time clinical symptoms of TlD appear. The sub-clinical phase of the disease is characterized by the presence of auto-antibodies which target the individual's islet cells (ICAs), insulin (IAAs), glutamic acid decarboxylase (or GADAs) and/or tyrosine phosphatase (IA-2As).

Although there is currently no cure for TlD, early detection can reduce the likelihood of long-term complications, thereby both improving the quality of life and reducing costs resulting from repeated hospitalization. For example, it has been shown that children previously identified as being autoantibody-positive had a much lower hospitalization rate at the time of diagnosis (3.3% versus 44%), lower mean glycohemoglobin one month later, and lower mean insulin dose one year later. Predictive testing thus appears to lessen morbidity and medical costs at diagnosis and may lead to better metabolic function in the early period after diagnosis ( 1). Although no immunoprevention therapy has yet been identified that will delay or prevent disease, it is likely that such therapies will be more effective when implemented early, for example in the pre-clinical period before the autoimmune response is well advanced and the remaining islets are stressed.

Cases with a positive family history for Tl D represent only 10- 15% of newly diagnosed patients (2-5), therefore effective public health testing must test all children in order to identify pre diabetes. While the presence of islet autoantibodies is a sensitive and specific predictor of future Tl D, autoantibodies appear at varying ages in different individuals, so periodic testing throughout childhood is necessary for prompt detection. Testing for autoantibodies is invasive and expensive, and thus a separate initial screening step is essential for overall cost-effectiveness. Since the peak incidence range for TlD is from about 4 to 15 years, predictive strategies must be applied early in life if they are to be informative.

The HLA-DR-DQ genetic locus is by far the most informative for Tl D susceptibility and is estimated to account for approximately 50% of the genetic susceptibility to the disease(6). It has been suggested that HLA Class II genotyping might provide sufficient information for an initial screening step (6-8). Such genetic screening may be performed as early as the newborn period, well before disease onset. Initial HLA genetic screening can be used to identify susceptible children, who are then offered periodic autoantibody testing to detect activation of islet autoimmunity. This approach has been successfully demonstrated in large research studies, such as the Diabetes Auto-Immunity Study of the Young (DAISY) in Colorado (7), the Prospective Assessment of Newborns for Diabetes Autoimmunity (PANDA) in Florida (9), the Diabetes Prediction and Prevention (DIPP) study in Finland (10), and the Diabetes Evaluation in Washington (DEW-IT) study ( 1 1). The latter study further minimized cost and invasiveness by performing the HLA screening in coordination with a Washington State Dept. of Health Newborn Screening Program. These studies confirmed the ability of HLA screening to identify high-risk subjects for intervention or follow-up studies, but no consensus strategy for population-based TlD public health screening has emerged from them. Developing such a strategy is challenging because HLA haplotypes and genotypes form a continuum between highly susceptible and highly protective. Kiviniemi et al. (12) describe a system for screening large numbers of individuals for genetic risk for TlD that employs multiple screening steps requiring large numbers of probes to identify different HLA alleles.

Assays employing DNA hybridization probes directed to a specific region of the HLA-DQβ region to detect a person's susceptibility to autoimmune diseases, such as Tl D, are described in US Patent 5,665,548, with probes directed to the HLA-DR4 region and their use in diagnosing susceptibility to Tl D being disclosed in US Patent 5,039,606. US Published Patent Application no. US2004/0126794 discloses methods for detecting increased or decreased risk for T l D by detecting the presence of specific HLA-C and/or HLA-A alleles. Methods for predicting autoimmune diabetes by detecting specific HLA Class II alleles are also described in US Patent 6,534,272. US Patent 5,567,809 describes specific primers and probes for HLA-DRβ DNA typing. US Published Patent Application no. US 2008/0026378 describes a method for predicting the onset of Tl D comprising determining a subject's HLA genotype, assigning the subject's risk of developing Tl D on the basis of the determined HLA genotype, measuring the concentration of at least one amino acid in a biological sample taken from the individual and combining the resulting information to predict the likelihood of onset of TlD.

While several methods for determining susceptibility to Tl D using HLA DR-DQ have been described, such methods are high resolution and therefore not cost-effective for routine screening of large numbers of individuals. There thus remains a need in the art for materials and methods that may be effectively employed to screen populations for individuals at risk of developing Tl D.

SUMMARY OF THE INVENTION The present invention provides efficient, cost-effective and non-invasive methods for screening large numbers of individuals in order to identify individuals who are at risk of developing Tl D and who should therefore receive follow-up testing for development of T l D, such as testing for the presence of autoantibodies indicative of pre-clinical T l D. Materials for use in such methods are also provided. The disclosed methods provide a practical means for screening large populations in order to implement public health strategies aimed at minimizing the occurrence and associated costs of clinical Tl D. As described in detail below, the inventors have employed a simple algorithm to convert the risk of developing Tl D due to the presence of specific HLA Class II haplotypes into genotype-based risk in order to maximize performance of a HLA DR-DQ screen. The methods disclosed herein comprise obtaining a nucleic acid-containing (for example, DNA-containing) biological sample from an individual to be tested, and amplifying DNA in the sample using primers specific for exons 2 of the HLA DQB l , DQAl and/or DRB l loci. The amplified DNA is then used for identification of specific alleles by one of several well-known methods. For example, the amplified DNA can be contacted with combinations of oligonucleotide probes directed against specific alleles, in the method generally known as SSOP, or "sequence specific oligonucleotide probe". The presence or absence of binding between the probe(s) and the amplified DNA sample is indicative of the presence or absence of the specific allele(s). Alternatively, the amplified DNA can be sequenced directly, for example on an Applied Biosystems 310 genetic analyzer (Perkin Elmer Applied Biosystems, Foster City, CA) or similar equipment. A method employing selective primer extension can also be used to identify specific alleles. Finally, the initial DNA amplification can employ primers which, instead of amplifying exons 2 of all DQA l and/or DQB l alleles, are designed with even greater specificity to amplify only specific alleles, in the method known as "sequence specific primers". In this case, successful DNA amplification itself implies allele identity.

The specific DQAl and/or DQB 1 alleles which are identified by one of the above methods are then used to detect and/or infer haplotypes previously identified as being indicative of either resistance (R) or susceptibility (S) to TlD development, or as being neutral (N) with regards to TlD predisposition. In general, an individual is not recommended for follow-up autoantibody testing if the genetic screen indicates the presence of a resistant (R) haplotype and/or the absence of a susceptible (S) allele. However, the presence of a certain allele may "forgive", or negate, the presence of another specific allele. Specific combinations of alleles to be tested for, or probed, are discussed below and identified in Table 1 below. The combinations of alleles (and therefore the probes) are selected in order to maximize the number of future Tl D cases included in autoantibody screening (i.e. maximize the sensitivity of the screening), while also minimizing the number of individuals recommended for follow-up autoantibody screening (i.e. maximizing the specificity). The disclosed methods are homogeneous for all samples (i.e. all samples may be tested with the same combination(s) of probes), and allow determining of genotype at the minimum resolution needed to define high and low risk DR-DQ haplotypes in order to determine Tl D genetic risk. Unlike prior art methods, individuals requiring follow-up testing for autoantibodies are identified using a simple method that can be readily employed for high-throughput screening of large numbers of individuals. In specific embodiments, methods for identifying an individual in need of follow- up testing for T 1 D are provided, the methods comprising testing for the presence of a first HLA Class II allele, a second HLA Class II allele and a third HLA Class II allele in a nucleic acid sample obtained from the individual, wherein the first allele is DQBl *0301 , the second allele is DQAl *020X, and the third allele is selected from the group consisting of: (a) DQB 1 *0602/0603; (b) DQB l *050X/060X; and (c) DQA 1 *01 , where X = any integer. The presence of any one of the first, second and third alleles indicates that the individual is not in need of follow-up testing for TlD.

In one embodiment, the third allele is DQB 1 *050X/060X or DQAl *010X (where X = any integer) and the method further comprises testing for the presence of a fourth HLA Class II allele selected from the group consisting of: (i) DQB 1 *0604; and (ii) DQB 1 *0501 , wherein the presence of the fourth allele negates the use of the third allele to indicate that the individual is not in need of follow-up testing for TlD.

In further embodiments, the methods additionally include testing for the presence of a fifth HLA Class II allele, wherein the fifth allele is DQB 1 *0503/0601, and wherein the presence of any one of the first, second, third and fifth alleles indicates that the individual is not in need of follow-up testing for TlD. The nucleic acid-containing sample may be further tested for the presence of a sixth HLA Class II allele, and/or for the presence of a seventh allele wherein the sixth allele is DQB 1 *0602/0603 and the seventh allele is DRBl *0403, and wherein the presence of any one of the first, second, third, fifth, sixth and seventh alleles indicates that the individual is not in need of follow- up testing for TlD.

In other embodiments, the methods further comprise testing for the presence of an eighth HLA Class II allele and a ninth HLA Class II allele, wherein the eighth allele is DQB 1 *0302 and the ninth allele is DQB l *020X (also referred to as DQB l *020X), and wherein the presence of any one of the first, second, third, fifth, sixth and seventh alleles or the absence of any one of the eighth and ninth alleles indicates that the individual is not in need of follow-up testing for Tl D. In a related embodiment, the nucleic-acid containing sample may further be tested for the presence of a tenth HLA Class II allele, wherein the tenth allele is DQB 1 *040X (where X = any integer) and wherein the presence of the tenth allele negates the use of the ninth allele to indicate that the individual is in need of follow-up testing for Tl D.

Materials for use in the disclosed methods, such as oligonucleotide probes that specifically hybridize, or bind, to the HLA Class II alleles of interest, are also provided. In one embodiment, kits are provided for identifying individuals at increased risk for developing TlD, such kits comprising combinations of oligonucleotide probes that are capable of hybridizing to the alleles of interest. The probes may already be labelled to facilitate detection of the presence or absence of binding between the probes and their targeted alleles, or the kits may include reagents for labelling the probes. The kits may also optionally include reagents to detect the label, and/or instructions for their use.

In yet a further embodiment, arrays, such as microarrays, are provided for use in the disclosed methods, such arrays comprising oligonucleotide probes that are capable of hybridizing to the specific combinations of alleles disclosed herein. The oligonucleotide probes may be immobilized on a substrate, such as a membrane or glass. Techniques and materials for preparing microarrays are well known in the art. Microarrays are available commercially and include those available from Affymetrix (Santa Clara, CA).

These and additional features of the present invention and the manner of obtaining them will become apparent, and the invention will be best understood, by reference to the following more detailed description and the accompanying drawings.

BRIEF DESCRIPTION OF THE FIGURES

Fig. 1 shows the percentage of population needing autoantibody screening (specificity) versus percentage of TlD cases detected (sensitivity) for possible hypothetical risk strategies generated for the haplotypes shown in Table 2 using a computer program.

Fig. 2 shows the cost-effectiveness of immunogenetic screening for Tl D.

Fig. 3 shows specificity versus sensitivity for a subset of the strategies of Fig. 1 , including for three strategies of most interest.

DETAILED DESCRIPTION

As outlined above, the present invention provides methods and materials for genetic screening of large populations to identify individuals who have increased genetic risk of developing TlD and should therefore receive follow-up screening for islet autoantibodies known to be indicative of the development of T l D. The methods include testing for the presence of HLA Class II haplotypes previously identified as being indicative of either resistance (R) or susceptibility (S) to TlD development, or as being neutral (N) with regards to Tl D predisposition by contacting DNA obtained from the individuals with combinations of oligonucleotide probes that hybridize with specific S, N or R alleles. As described in detail in Example 1 below, the inventors obtained extended HLA Class II DRB l -DQAl -DQB l haplotype information for over 1000 individuals previously diagnosed to have TlD and over 1000 healthy control individuals. Based on this information, the haplotypes were ranked from those being most Tl D resistant to those being most Tl D susceptible. While the continuum of Tl D risk in haplotypes ranges from highly susceptible to moderately susceptible to neutral to moderately resistant to highly resistant, the haplotype risk was assigned more concisely to three categories, namely susceptible (S), neutral (N) and resistant (R). Based on the known dominant protection of previously identified resistant haplotypes, a paradigm was developed to combine the haplotypes to provide two categories of genotypes, namely those associated with a high risk of developing Tl D (S/S, S/N; i.e. individuals who should receive follow- up autoantibody screening), and those associated with a low risk of developing TlD (N/N, R/S, R/N, R/R; i.e. individuals to be excluded from follow-up autoantibody screening). Including "N" as a third haplotype risk level allowed for greater stratification, while reducing the genotype risk levels to two makes the autoantibody follow up practical. Intermediate, or moderate, risk classifications are not considered when determining whether or not an individual should receive follow-up screening.

Groups of alleles in the S, N and R categories were subsequently established for several different sensitivity/specificity strategies. In order to create a cost-effective method for detecting TlD cases in a large population of individuals, strategies were developed that minimized the number of individuals receiving autoantibody screening (i.e. maximized the specificity) while maximizing the number of future TlD cases detected (i.e. maximizing the sensitivity).

In these strategies, detection of various combinations of the following R alleles is used to identify the presence of the R haplotypes and therefore identify individuals who can rapidly be excluded from the group requiring follow-up autoantibody testing: DQB l *0301 ; DQA 1 *O2OX (also referred to as DQB 1 *O2); DQB 1 *0602/0603; DQB 1 *O5OX/O6OX; DQB 1 *0602/0603; DQA 1 *01 OX; and DRB 1 *0403 (or DRB 1 *0403/0406/0407/041 1 ), where X = any integer. Examples of specific strategies employing R alleles are shown in Table 1. Table 1 includes the total numbers of individuals tested in different populations (row 1) and the number of TlD cases captured in each population for each specific strategy, wherein WA Cau = number of healthy Caucasians tested; WA all race = total number of healthy subjects tested (regardless of race); all race, all DM = total number of T1D subjects tested; all race. Tl D ons<22 = total number of Tl D subjects with an age at onset of less than 22 tested (regardless of race); Cau, all DM = number of Caucasian TlD subjects tested; and Cau, Tl D ons<22 - number of Caucasian TlD subjects with an age at onset of less than 22 tested.

TABLE IA

Detection of the following S alleles is used to identify the presence of S haplotypes, and therefore identify individuals who should be included in the group requiring follow-up autoantibody testing: DQB 1 *0302 and DQB 1 *020X (also referred to as DQB 1 *02). However, it should be noted that DQB l *020X is only useful when a probe for DQA1 *O2OX is included in the test as the presence of DQA 1 *O2OX disqualifies DQB l *020X as representing a susceptible haplotype. Similarly, in those cases where DRB 1 *0403 is probed, its presence disqualifies DQB 1 *0302 as representing a susceptible haplotype. R alleles can be grouped in several ways for use with S alleles as shown in Table 1.

Other N alleles, such as DQB 1 *0501 and DQB 1 *0604, can be added to relieve elimination by DQB 1 *050X/060X or by DQA 1 *01OX (the latter two are equivalent entities), as shown in Table 1. In addition, the allele DQB l *040X can be added to eliminate individuals who test positive for the S allele DQB 1 *O2OX. The sensitivity and specificity of the screening method vary depending on the combination of alleles tested for. For example, testing only for the R alleles DQB 1 *0503/0601 , DQB 1 *0301/0304, DQA l *020X results in a specificity of 28% and a sensitivity of 74.8%, while testing for the R alleles DQB 1 *0503/0601 , DQBl *0301/0304, DQA 1 *020X, B0602/0603, and the S alleles DQB 1 *0302/0304 and DQB l *020X results in a specificity of 16% and a sensitivity of 71.7%. Testing for the R alleles DQB1 *O5OX/O6OX, DQB l *0301/0304, DQA l *020X, and for the S alleles DQB 1 *0302/0304 and DQB l *020X, as well as for the S allele modifier DQB l *040X, yields a specificity of 5.7% and a sensitivity of 51.1 %, as does testing for R alleles DQB 1 *0301/0304, DQA l *020X and DQA 1 *01 and S alleles DQB 1 *0302/0304 and DQB 1 *020X and S allele modifier DQB 1 *040X.

Cost effectiveness of the overall prediction strategy (HLA screening and autoantibody follow-up) is a key factor in design considerations and is greatly affected by the stringency (sensitivity, specificity) of the HLA screening step. The cost per TlD case identified is higher when a higher number of follow-up autoantibody tests need to be performed. It is known that performing initial genetic screening to determine which individuals should receive autoantibody screening provides significant cost-savings compared to autoantibody screening alone. Defining genotype risk through converting haplotype information helped to identify the majority of future Tl D cases while minimizing the proportion of the population needing autoantibody follow up by maximizing performance of HLA DR-DQ in a Tl D genetic screening role, thereby increasing sensitivity while maintaining specificity compared to strategies that employ simply the alleles or haplotypes previously identified. The impact of the improvement in the sensitivity on cost savings over a longer term will be sizeable, as it is known that children who participate in prospective follow-up autoantibody testing are less often hospitalized and have milder metabolic abnormalities at diagnosis. In a large population, some deaths and permanent morbidity are likely to be prevented by early diagnosis.

In order to identity the presence or absence of specific alleles in an individual, a nucleic acid (DNA and/or RNA) containing biological sample is first obtained from the individual. The biological sample may be, for example, blood, urine, saliva or sera. DNA and/or RNA may also be obtained from hair, skin, nails or other body tissue. The nucleic acid-containing sample is then subjected to polymerase chain reaction (PCR) to amplify exon 2 of the HLA DQA l and DQB l genes. Primers and techniques for use in PCR are well known to those in the art and include, but are not limited to, those described below in Example 1. The presence of alleles of interest can be detected using methods known in the art, including, but not limited to, contacting the amplified nucleic acid-containing sample with one or more oligonucleotide probes that hybridize under stringent hybridization conditions to one or more polymorphisms associated with the alleles and detecting the hybridized, or bound, oligonucleotide probes. Oligonucleotide probes that may be effectively employed to detect the HLA II alleles of interest are well known in the art and include, for example, those described in US Patent 5,567,809, US Patent Publication no. 2004/0126794, Kiviniemi et al. Diabetes Technology & Therapeutics, 9:460-472 (2007)), and ltoh et al. Immunogenetics 57:717-729 (2005), the disclosures of which are hereby incorporated by reference. Such oligonucleotide probes and primers are substantially complementary to one or more polymorphisms associated with the allele of interest. Two single stranded sequences are said to be substantially complementary when the nucleotides of one strand, optimally aligned and compared, with the appropriate nucleotide insertions and/or deletions, pair with at least 80%, preferably at least 90% to 95%, and more preferably at least 98% to 100%, of the nucleotides of the other strand. Alternatively, substantial complementarity exists when a first DNA strand will selectively hybridize to a second DNA strand under stringent hybridization conditions.

As employed herein the term "stringent hybridization conditions" includes salt conditions of less than about 1 M, more usually less than about 500 mM and preferably less than about 200 mM. Hybridization temperatures can be as low as 5°C, but are generally greater than about 22⁰C, more preferably greater than about 3O⁰C and most preferably greater than about 37°C. Longer DNA fragments may require higher hybridization temperatures for specific hybridization. Since the stringency of hybridization may be affected by other factors such as probe composition, presence of organic solvents and extent of base mismatching, the combination of parameters is more important than the absolute measure of any one alone. In one specific example, "stringent hybridization conditions" refers to prewashing in a solution of 6X SSC, 0.2% SDS; hybridizing at 65⁰C, 6X SSC, 0.2% SDS overnight; followed by two washes of 30 minutes each in IX SSC, 0.1 % SDS at 65⁰C and two washes of 30 minutes each in 0.2X SSC, 0.1 % SDS at 65⁰C.

In certain embodiments, sequence-specific oligonucleotide probes (SSOP; i.e. probes that hybridize specifically to the allele of interest) are immobilized on a solid substrate, such as a nylon membrane, using methods well known to those of skill in the art. The bound SSOP are contacted with the PCR-amplified nucleic acid sample for a period of time sufficient for the SSOP to hybridize to the target allele(s). The substrate is then washed to remove unhybridized sample and the presence of the bound SSOP is detected using methods known to those of skill in the art. For example, the oligonucleotide probes may be labelled with a moiety that allows detection of probe by spectroscopic methods. In one method, the nucleic acid-containing sample is amplified using biotinylated primers and the bound biotinylated PCR product is detected using streptavidin-horseradish peroxidase.

In an alternative embodiment, the presence or absence of specific alleles of interest is detected using a Delfia™ system (Perkin Elmer, Boston, MA), in which up to three different SSOPs, each labelled with a different detection reagent such as europium (Eu), terbium (Tb) or samarium (Sm), are simultaneously contacted with an amplified DNA sample. Binding of alleles to the labeled probes may then be detected using time- resolved fluorometry. For example, Sjδroos et al. ( 13) described a method for detecting two Tl D susceptibility (*0201 , *O3O2) and two TlD protective (*0301 , *0602/0603) alleles of the HLA-DQB l gene employing DNA amplification with PCR followed by simultaneous, allele-specific triple-label hybridization performed in microtitration wells using the Delfia™ system. Use of this type of system offers significant advantages, in that it enables simultaneous testing for the presence or absence of multiple alleles of interest.

Other techniques that may be effectively employed to detect hybridization between SSOP and alleles of interest include SSOP-Luminex™ methods as described by Itoh et al. (15). Such methods employ a flow cytometry dual-laser system to quantitatively detect fluorescently labelled oligonucleotides attached to color-coded microbeads and have previously been employed in high-throughput, high-resolution genotyping studies. SSOP may also be employed in a high-throughput ELISA technique to detect the presence of alleles of interest. Such techniques are well known in the art and include, for example, those described in (16).

A plurality of oligonucleotide probes may be provided in a kit form. Such kits generally comprise multiple oligonucleotide probes, with each probe being specific for an allele of interest. In one embodiment useful for high-throughput assays, the oligonucleotide probe kits disclosed herein comprise multiple probes in an array format, wherein each probe is immobilized in a predefined, spatially addressable location on the surface of a solid substrate. Array formats which may be usefully employed in the present invention are disclosed, for example, in U.S. Patents No. 5,412,087, 5,545,531 , 6,586, 168, 6,284,460, 6,268, 152, 6, 156,501, 6,045,996, the disclosures of which are hereby incorporated by reference.

Oligonucleotide probes for use in the disclosed methods may be constructed synthetically using techniques well known in the art (See, for example, Gait, ed., Oligonucleotide synthesis a practical approach, IRL Press: Oxford, England, 1984). Automated equipment for the synthesis of oligonucleotides is available commercially from such companies as Perkin Elmer/Applied Biosystems Division (Foster City, CA) and may be operated according to the manufacturer's instructions. Alternatively, the probes may be constructed directly on the surface of an array using techniques taught, for example, in PCT Publication No. WO 95/00530. Those of skill in the art will appreciate that alternative methods, such as restriction-fragment length polymorphism (RFLP), may also be employed to detect the presence or absence of specific alleles, or combinations of alleles, of interest in nucleic acid-containing biological samples. An additional method to determine the alleles present is to directly sequence the amplified exons 2 of DRBl , DQA l and/or DQBl using di- deoxy labeling followed by analysis on the Applied Biosystems 310 Genetic Analyzer or similar apparatus. Alternatively, a method employing selective primer extension can be used to identify specific alleles ( 14-16)

Another method that may be employed to detect specific alleles, is for the initial DNA amplification to employ primers which, instead of amplifying exons 2 of all DQA 1 and/or DQB l alleles, are designed with even greater specificity to amplify only specific alleles, in the method known as "sequence specific primers". In this case, successful DNA amplification itself implies allele identity.

The following examples are offered by way of illustration and not by way of limitation.

5

EXAMPLE 1

A large case-control cohort from Washington State was HLA-DQ genotyped to determine: a) what are the best strategies for population-wide HLA Class II screening in a typical U.S. population, and b) whether the best such strategy is sufficiently predictive to l ϋ be useful in screening for a subsequent autoantibody testing program for cost-effective, public health-based preclinical Tl D prediction as a prelude to risk counseling and ultimately to delay or prevent the onset of disease.

The most useful haplotype algorithms for maximum performance of HLA DR-DQ in a Tl D genetic screening role were identified as described below. Medium-resolution

15 inferred HLA DQAl -DQBl haplotypes from 907 TlD cases and 1 163 healthy subjects from Washington State were combined into genotypes to test all risk assessment strategies based on relative risk groupings (susceptible, neutral, resistant) of the individual haplotypes. DQB 1 *0302 haplotypes were further stratified by DRBl *040X subtypes before analysis. Computerized simulations tested all risk strategies representing0 all possible haplotype risk assignment combinations. Results were interpreted in light of desirable general pediatric T l D risk screening goals, namely sufficient sensitivity to include most future cases among subjects screening positive, and sufficient specificity to minimize the overall number who must undergo subsequent follow-up. Strategies with the highest combined sensitivity/ specificity (% of future cases within % of pediatric 5 population) were 51.1 % within 6.8%, 65.1 % within 12.1 %, 72.5% within 15.1 %, and 76.0% within 18.0%. There was slightly less sensitivity if adult onset, as well as childhood-onset, Tl D cases were considered, and slightly greater sensitivity if only Caucasians were included. 0 Subjects- The unrelated healthy control cohort consisted of 1 163 randomly selected subjects from a 4505-subject Washington State general population study (17) and a similar Washington State general population diabetes screening study, and excluded subjects with diabetes, those who were first degree relatives of current diabetes patients, and those with persistent islet autoantibodies. Type 1 diabetes subjects were recruited,5 consented and sampled from hospital wards or clinics, or specialty medical practices in Washington State. Blood was drawn for serum autoantibody testing and genomic DNA testing. Of 1094 consented and sampled diabetes cases (median diabetes duration 1 1.2 years), 1062 had sufficient DNA sampled for genotyping. For 72 cases, a serum C- peptide measurement was available. Inclusion criteria for childhood TlD (onset age < 22 years), was positivity for any of the 3 islet autoantibodies OR random C-peptide <0.8 (18) OR first degree relative with autoantibody-confirmed TlD OR all of the following (BMI<25, no T2D first degree relatives, AND on continuous insulin therapy since diagnosis). Inclusion criteria for adult TlD (onset age 22 years or older) was positive Tl D autoantibodies OR random C-peptide <0.8 OR a first degree relative with autoantibody-confirmed TlD. After application of the above inclusion criteria, a total of 907 cases (650 childhood TlD cases and 257 adult TlD cases) were included in the analysis. For all Tl D cases, 93.3% were Non-Hispanic White, 0.8% Hispanic White, 1.5% Black, 1.1 % Asian, and 3.3% other/undetermined. For healthy controls, 78.3% were Non-Hispanic White, 5.2% Hispanic White, 2.7% Black, 7.3% Asian and 6.4% other/undetermined.

Autoantibodies - Serum autoantibodies to the human diabetes islet autoantigens GAD65, the full cytoplasmic domain of IA2, and insulin were measured using separate radiobinding assays as described by Woo et al. (19).

Sequence-Based DQ Genotvping - Genotyping of HLA DQAl and DQBl utilized direct sequencing of amplified exon 2 of each gene using a Perkin Elmer/Applied Biosystems Inc. 310 automated sequencer. PCR templates consisted of either 1/8" dried bloodspots fixed in MeOH as described (1 1 ) or genomic DNA purified from whole frozen blood (QiaAmp, Qiagen). PCR primers for DQBl exon 2 were GH29 and DB 130 (20) and for DQAl exon 2 were DQAAMP-A and DQAAMP-B (21 ). Allele frequencies and frequency of homozygosity were similar to those found in 1 102 subjects combined from two large Washington State bone marrow transplant registries (22) and in published 1 lth International HLA Workshop data on North American Caucasians, Blacks and Japanese (23). This indicates that all DQAl and DQB l alleles were well amplified by our methods. DRB 1 *04 subtvping - Subjects with DQ haplotypes expected to have DRB 1 *04 were further examined by low-resolution DR4 subtyping using Restriction Fragment Length Polymorphism (24). Published DRB 1 *04-specific PCR primers AB54 (sense) and AB60 (antisense) (20) were used to generate 257 bp amplicons for digestion with SacII and visualization by agarose gel electrophoresis. Cleavage to 199 bp and 58 bp fragments indicated common Tl D-susceptible DR4 alleles (DRB 1 *0401/0402/0404/0405) while no cleavage indicated DR4 alleles generally conferring Tl D resistance (DRB 1 *0403/406/407/41 1).

Algorithm- After identifying allele sequences for each gene, DQA l -DQB l haplotypes were inferred based on published frequencies in Caucasians (25) and from the HLA 1991 workshop for Asian- Americans and Black Americans (23). Haplotypes were assigned into one of three categories, namely resistant (R), neutral (N) and susceptible (S), and each pair of haplotypes was then combined into an individual genotype. Based on the known dominance of resistant over susceptible haplotypes (26; 27), as well as the necessity of having at least one susceptible haplotype to be at risk of disease (28; 29), the genotypes R/S, R/N, R/R, and N/N were assigned to the low risk cohort, while S/S and S/N were assigned to the high risk cohort for which follow-up autoantibody testing was recommended.

Modest simplification of the haplotype list was done to allow each listing to contain sufficient numbers of subjects. Where practical typing refinements existed (e.g. DR4 subtyping), these were included to better resolve the haplotypes. In the case of ambiguous assignments, the most frequent DQAl-DQB l combinations were chosen, which in all cases were at least 50-fold more prevalent overall than the non-chosen combinations. The list of haplotypes was then further simplified in three ways. First, in a limited number of cases, haplotypes identical at DQB l but with minor differences at the fourth digit of DQAl (e.g. 0102 and 0103) were grouped, provided that the grouped haplotypes did not differ substantially in relative disease risk based on published data (23). Second, DQA 1 *03 haplotypes with DQB 1 *0302 or 0304 were grouped together since these DQB l alleles are structurally similar and do not differ substantially in relative disease risk. This grouping was particularly important since these grouped DQA l *030X- DQBl *0302/4 haplotypes were then divided into three groups based on DR4 subtyping (DRB 1 *0403 group, not DRB 1 *0403, and not DRB P040X). Finally, four rare haplotypes (DQA l *030X-DQB 1 *0402, DQA 1 *0501 -DQB 1 *0302/4, DQA 1 *0102- DQB 1 *0504 and DQA 1 *0101 -DQB 1*0608) which were each observed at a frequency of less than 1 in 500, were combined as "rare haplotypes" which included a total of 18 haplotype counts out of 4140 total haplotype counts.

The final result totaled 22 haplotypes or haplotype categories. Their frequencies among cases and controls, odds ratios (OR), and significance of association to Tl D, are shown in Table 2. Table 2 lists these haplotypes in order from those which confer the greatest resistance to TlD (at the top of the table), to those which confer the greatest susceptibility to T I D (at the bottom of the table). <ooooι

<o.oooι

Haplotypes were assigned into risk categories using a simple manual method and a computer program. For both methods, the top two haplotypes in Table 2 were seldom found in cases of Tl D and were fixed as R, while the bottom two haplotypes, which clearly conferred disease risk, were fixed as S, and the rare haplotype group was fixed as N based on insufficient data. The seventeen remaining intermediate risk haplotypes were allowed to wobble between different risk categories. For the manual assignment, these intermediate haplotypes were kept in order of relative risk. A variable number of them (from three to seventeen) was then assigned as N in a contiguous block. The contiguous N block was shifted up and down within the seventeen haplotypes, with all haplotypes above the block assigned R and all below the block assigned S. This resulted in 121 different "manual" strategies.

A computer program was developed that allowed every possible assignment of haplotypes into risk categories to be exhaustively evaluated. For the computer-generated strategies, the seventeen remaining intermediate risk haplotype groups were allowed to be assigned into one of the three risk categories. Each arrangement of haplotypes constituted a hypothetical risk-evaluation strategy. Every possible unique hypothetical risk strategy was iteratively tested for sensitivity (proportion of Tl D cases detected) and specificity (proportion of healthy controls excluded from autoantibody follow-up testing) on the WA state data. A non-parametric ROC (receiver operating characteristic) curve was plotted using sensitivity (percentage of Tl D cases that would be detected by autoantibody screening) versus 1 -specificity (percentage of population screened genetically who would receive follow-up autoantibody screening). Fig. 1 shows the results of sensitivity versus specificity for the computer strategy.

Due to the discrete nature of HLA haplotypes and genotypes, the curve is not continuous. Many strategies resulted in the same specificity but differed in their sensitivity - only the set of best strategies is presented (i.e. the highest sensitivity strategy for each given specificity, the highest specificity for each given sensitivity. As can be seen from Fig. 1 , the best strategies had a specificity between 4-30%. As expected, higher sensitivities were associated with lower specificities.

Fig. 2 is a graph of the cost per identified TlD case versus the prediction strategy sensitivity. From this graph, it can be seen that identification of 60-70% of Tl D offers the best cost-effectiveness. Based on these cost considerations, haplotype screening strategies which lead to detection of 60-80% of future TlD cases by performing follow- up autoantibody testing on 10-20% of the screened population were determined to be of most interest.

Fig. 3 presents a subset of the strategies shown in Fig. 1 , namely strategies that achieved 60-80% of TlD cases detected within 10-20% of the population. Three of the 5 most informative strategies are marked. These represent detection of 65.1 % of future Tl D cases by follow-up testing of 11.8% of the pediatric population for autoantibodies, 72.5% of cases by follow-up testing of 15.6% of children, and 76.0% of cases by follow- up testing of 17.9% of children. The attractiveness of higher sensitivity must be weighed against lower specificity, in this case the greater expense and invasiveness of following a 10 larger proportion of the general pediatric population for periodic autoantibody testing during childhood.

EXAMPLE 2

The presence of one or more specific alleles in a biological sample is identified as

I₅ follows. The method comprises PCR amplification followed by oligonucleotide probe hybridization using a commercially available time-resolved fluorescence (TRF) assay.

DNA amplification by polymerase chain reaction (PCR) was performed using either dried blood spot 1/8 inch punches or genomic DNA purified from whole frozen blood (QiaAmp, Qiagen) as template. DBS were amplified in 96-well microtiter plates0 by the PCR procedure. The reaction mixture was: DBS, IXPCR buffer (16 mM (NH₄)₂SO₄, 67 mM Tris-HCL (pH 8.8 at 25⁰C), 0.01% Tween 20), 5.5% glycerol, 2.0 mM MgCl₂, 0.2 mM each of dATP, dCTP, dTTP and dGTP, 0.35 pM DQA l primers and 0.25 pM DQBl primer, 3.5 unit DNA polymerase (Bioline, MA, USA) and DNAase free molecular grade water for a total volume of 100 ul. PCR primers for DQAl were Biotin-5 5'- TAT GGT GTA AAC TTG TAC CAG T-3'(sense; SEQ ID NO: 1), 5'-GGT AGC AGC GGT AGA GTT G-3'(antisense; SEQ ID NO: 2). PCR primers for DQB l were 5'- GCA TGT GCT ACT TCA CCA ACG-3'(sense; SEQ ID NO: 3), Biotin-5'-CCT TCT GGC TGT TCC AGT ACT-3'(antisense; SEQ ID NO: 4). PCR amplifications were performed on automated PCR thermal cycler (PTC-200, Peltier thermal cycler, MJ0 Research, New Jersey) with 34 cycles as follows: 10 minute at 95⁰C, followed by 34 cycles of 50 second at 95⁰C, 1 minute at 55⁰C, 1 minute at 72⁰C, then 5 minute at 72⁰C. A small portion of the amplified mixture was evaluated by 2% agarose gel electrophoresis to verify successful amplification.

For the Time-Resolved Fluorescence (TRF) assay, 10 ul of biotinylated PCR5 product was directly transferred to streptavidin-coated microtitration plates (Pierce), incubated with 50 ul of hybridization solution for 30 minutes at room temperature, and denatured with 20 mmol/L NaOH for 5 minutes at room temperature. The PCR products were then hybridized for 2 hours with a mixture of three allele sequence-specific probes (Delfia, Perkin-Elmer) that each carry a different lanthanide chelate (europium (Eu), samarium (Sm) or Terbium (Tb)). The probes were each used at a final concentration of 1.0 ng-1.5 ng/well). After incubation, stringent washes with wash solution (Delfia, Perkin Elmer) were performed at 45⁰C, and 200 ul of enhancement solution (Delfia, Perkin Elmer) was added to enhance the Eu and Sm fluorescence. Microtiter plates were counted on a Victor² fluorescence counter (Perkin-Elmer Wallac Oy, Turku, Finland) to measure the Eu and Sm TRF signals. Then 50 ul of enhancer solution (Delfia, Perkin Elmer) was added prior to measuring the Tb-fluorescence, also on a Victor² microtiterplate counter (Perkin-Elmer Wallac Oy). The details of the assay have been described previously (13). The lanthanide chelates probes, hybridization buffer, washing buffer, enhancement and enhancer were commercial reagents from PE-Wallac DELFIA system (Wallac OY, Turku, Finland). We modified the three allele specific probes combination of the multiplex assay to function accurately and cost-effectively for the specific research goal for the TlD general population screen, for example using a mixture of Eu-DQB 1 *05/06, Sm-DQB 1 *0301 and Tb-DQA 1 *0201. The DQB 1 *05/06 probe was designed by, Drs. Hagopian and Peng as 5'-Eu-CAG GGG CGG CT-3' (SEQ ID NO: 5), and then manufactured to their specification including Eu chelate labeling, by Perkin- Elmer Wallac. An alternative probe also manufactured to specification including Eu chelate labeling by Perkin-Elmer Wallac, was the DQB 1 *0503/0601 probe: 5'-Eu-GGC GGC CTG ACG-3" (SEQ ID NO: 6). The Sm-DQB 1 *0301 and Tb-DQAl *0201 probes are available from Perkin-Elmer Wallac as catalog items. In some strategies from Table 1 , two sets of probes were used in parallel assays on separate microtiter plates, with each set comprising up to three separate Lanthanide-labeled oligonucleotide probes.

While the present invention has been described with reference to the specific embodiments thereof, it should be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the true spirit and scope of the invention. In addition, many modifications may be made to adapt a particular situation, material, composition of matter, method, method step or steps, for use in practicing the present invention. All such modifications are intended to be within the scope of the claims appended hereto. All of the publications, patent applications and patents cited in this application are herein incorporated by reference in their entirety to the same extent as if each individual publication, patent application or patent was specifically and individually indicated to be incorporated by reference in its entirety.

SEQ ID NO: 1-6 are set out in the attached Sequence Listing. The codes for nucleotide sequences used in the attached Sequence Listing, including the symbol "n," conform to WIPO Standard ST.25 ( 1998), Appendix 2, Table 1.

References

I . Barker JM, Goehrig SH, Barriga K, Hoffman M, Slover R, Eisenbarth GS, Norris JM, Klingensmith GJ, Rewers M: Clinical characteristics of children diagnosed with type 1 diabetes through intensive screening and follow-up. Diabetes Care 27: 1399-1404, 2004 2. Dahlquist G, Blom L, Holmgren G, Hagglδf B, Larsson Y, Sterky G, Wall S: The epidemiology of diabetes in Swedish children 0- 14 years old: a six-year prospective study. Diabetologia 28:802-808, 1985

3. Tuomilehto J, Lounamaa R, Tuomilehto-Wolf E, Reunanen A, Virtala E, Kaprio EA, Akerblom HK: Epidemiology of childhood diabetes mellitus in Finland-background of a nationwide study of type 1 (insulin-dependent) diabetes mellitus. The Childhood Diabetes in Finland (DiMe) Study Group. Diabetologia 35:70-76, 1992

4. Pociot F, Norgaard K, Hobolth N, Andersen O, Nerup J: A nationwide population- based study of the familial aggregation of type 1 (insulin-dependent) diabetes mellitus in Denmark. Danish Study Group of Diabetes in Childhood. Diabetologia 36:870-875, 1993 5. Wagener DK, Kuller LH, Orchard TJ, LaPorte RE, Rabin B, Drash AL: Pittsburgh diabetes mellitus study. II. Secondary attack rates in families with insulin-dependent diabetes mellitus. ΛM J. Epidemiol. 1 15:868-878, 1982

6. Hagopian W, Sanjeevi C, Kockum I, Landin-Olsson M, Karlsen A, Sundkvist G, Dahlquist G, Palmer J, Lernmark A: Glutamate decarboxylase-, insulin- and islet cell- antibodies and HLA typing to detect diabetes in a general population-based study of Swedish children. J Clin Invest 95: 1505-151 1 , 1995

7. Rewers M, Bugawan TL, Norris JM, Blair A, Beaty B, Hoffman M, McDuffie RS, Jr., Hamman RF, Klingensmith G, Eisenbarth GS, Erlich HA: Newborn screening for HLA markers associated with IDDM: diabetes autoimmunity study in the young (DAISY). Diabetologia 39:807-812, 1996

8. Morales A, She J, Schatz D: Prediction and prevention of type 1 diabetes. Curr Diab Rep 1 :28-32, 2001

9. Bennett-Johnson S, Baughcum A, Carmichael S, She J, Schatz D: Maternal anxiety associated with newborn genetic screening for type 1 diabetes. Diabetes Care 27:392- 397, 2004

10. Hahl J, Simell T, Ilonen J, Knip M, Simell O: Costs of predicting IDDM. Diabetologia 41 :79-85, 1998

I I. Wion E, Brantley M, Stevens J, Gallinger S, Peng H, Glass M, Hagopian W: Population-wide infant screening for HLA-based type 1 diabetes risk via dried blood spots from the public health infrastructure. Ann N Y Acad Sci 1005:400-403, 2003 12. Kiviniemi M, Hermann R, Nurmi J, Ziegler AG, Knip M, Simell O, Veijola R, Lovgren T, Ilonen J: A high-throughput population screening system for the estimation of genetic risk for type 1 diabetes: an application for the TEDDY (the Environmental Determinants of Diabetes in the Young) study. Diabetes Technol Ther 9:460-472, 2007 13. Sjoroos M, Iitia A, Ilonen J, Reijonen H, Lovgren T: Triple-label hybridization assay for type-1 diabetes-related HLA alleles. Biotechniques 18:870-877, 1995 14. Nikiforov TT, Rendle RB, Goelet P, Rogers YH, Kotewicz ML, Anderson S, Trainor GL, Knapp MR: Genetic Bit Analysis: a solid phase method for typing single nucleotide polymoφhisms. Nucleic Acids Res 22:4167-4175, 1994 15. Chen J, Iannone MA, Li MS, Taylor JD, Rivers P, Nelsen AJ, Slentz-Kesler KA, Roses A, Weiner MP: A microsphere-based assay for multiplexed single nucleotide polymorphism analysis using single base chain extension. Genome Res 10:549-557, 2000

16. Han M, Tan YQ, Zhang Y, Tsai J, Vorhaben R, Moraes JR, Moraes ME, Stastny P: Multiplex single nucleotide extension: a robust and high throughput method for HLA-A locus typing. Hum Immunol 64: 1111- 1 122, 2003

17. LaGasse J, Brantley M, Leech N, Rowe R, Monks S, Palmer J, Nepom G, McCulloch D, Hagopian W: Successful prospective prediction of type 1 diabetes in schoolchildren through multiple defined autoantibodies: an 8-year follow-up of the Washington State Diabetes Prediction Study. Diabetes Care 25:505-51 1 , 2002 18. Berger B, Stenstrom G, Sundkvist G: Random C-peptide in the classification of diabetes. ScandJ Clin Lab Invest 60:687-693, 2000

19. Woo W, LaGasse J, Zhou Z, Patel R, Palmer JP, Campus H, Hagopian WA: A novel high-throughput method for accurate, rapid, and economical measurement of multiple type 1 diabetes autoantibodies. J. Imm. Methods 244:91-103, 2000 20. Erlich H, Bugawan T, Begovich AB, Scharf S, Griffith R, Saiki R, Higuchi R, Walsh

PS: HLA-DR, DQ and DP typing using PCR amplification and immobilized probes. Eur

J Immunogenet 18:33-55, 1991

21. Tsuji K, Aizawa M, Sasazuki T: HLA 1991. New York, Oxford University Press,

1992 22. Mori M, Beatty P, Graves M, Boucher K, Milford E: HLA gene and haplotype frequencies in the North American population: the National Marrow Donor Program

Donor Registry. Transplantation 64: 1017- 1027, 1997

23. Kimura A, Sasazuki T: 11th International Histocompatiblity Workshop protocols for

DNA-typing. In HLA 1991: 11th International Histocompatiblity Workshop Tsuji K, Aizawa M, Sasazuki T, Eds. Oxford, UK, Oxford University Press, 1992, p. 397-419 24. Ju L, Gu X, Bardie R, Krishnamoorthy R, Charron D: A simple nonradioactive method of DNA typing for subsets of HLA-DR4: prevalence data on HLA-DR4 subsets in three diabetic population groups. Hum Immunol 31 :251-258, 1991

25. Klitz W, Maiers M, Spellman S, Baxter-Lowe L, Schmeckpeper B, Williams T, 5 Fernandez- Vina M: New HLA haplotype frequency reference standards: high-resolution and large sample typing of HLA DR-DQ haplotypes in a sample of European Americans. Tissue Antigens 62:296-307, 2003

26. Pugliese A, Gianani R, Moromisato R, Awdeh ZL, Alper CA, Erlich HA, Jackson RA, Eisenbarth GS: HLA-DQBl *0602 is associated with dominant protection from

10 diabetes even among islet cell antibody-positive first-degree relatives of patients with IDDM. Diabetes 44:608-613, 1995

27. Roep BO, R S, W V, GJ B, GM S, RR d: HLA-DRB 1 *0403 is associated with dominant protection against IDDM in the general Dutch population and subjects with high-risk DQA 1 *0301 -DQB 1 *0302/DQA 1 *0501 -DQB 1 *0201 genotype. Tissue Antigens

I₅ 54:88-90, 1999

28. Thorsby E, Ronningen K: Particular HLA-DQ molecules play a dominant role in determining susceptibility or resistance to type 1 (insulin-dependent) diabetes mellitus. Diabetologia 36:371 -377, 1993

29. Pugliese A: Unraveling the genetics of insulin-dependent diabetes: the search must go 20 on. Diabetes Reviews 7:39-54, 1999

Claims

CLAIMSWe claim:

1. A method for identifying an individual in need of follow-up testing for TlD, the method comprising testing for the presence of a first HLA II allele, a second HLA II allele and a third HLA II allele in a nucleic acid-containing sample obtained from the individual, wherein the first allele is DQB 1 *0301, the second allele is DQAl*020X, and the third allele is selected from the group consisting of:

(a) DQB 1 *0602/0603;

(b) DQB 1 *050X /060X; and

(c) DQA 1 *010X, wherein X equals any integer and the presence of any one of the first, second and third alleles indicates that the individual is not in need of testing for TlD.

2. The method of claim 1 , wherein the third allele is DQB l *050X /060X or DQA 1 *01OX and the method further comprises testing for the presence of a fourth HLA II allele selected from the group consisting of:

(i) DQB 1 *0604; and

(ii) DQB 1 *0501, wherein the presence of the fourth allele negates the use of the third allele to indicate that the individual is not in need of follow-up testing for TlD.

3. The method of claim 1, further comprising testing for the presence of a fifth HLA II allele, wherein the fifth allele is DQB 1 *0503/0601 , and wherein the presence of any one of the first, second, third and fifth alleles indicates that the individual is not in need of testing for TlD.

4. The method of any one of claims 1 and 3, further comprising testing for the presence of a sixth and/or a seventh HLA II allele, wherein the sixth allele is DQB 1 *0602/0603 and the seventh allele is DRB 1 *0403, and wherein the presence of any one of the first, second, third, fifth, sixth and seventh alleles indicates that the individual is not in need of testing for TlD.

5. The method of any one of claims 1, 3 and 4 further comprising testing for the presence of an eighth HLA II allele and a ninth HLA II allele, wherein the eighth allele is DQB 1 *0302 and the ninth allele is DQB 1 *02, and wherein the presence of any one of the first, second, third, fifth, sixth and seventh alleles or the absence of any one of the eighth and ninth alleles indicates that the individual is not in need of testing for TlD.

6. The method of claim 5, further comprising testing for the presence of a tenth HLA II allele, wherein the tenth allele is DQB l *040X and wherein the presence of the tenth allele negates the use of the ninth allele to indicate that the individual is in need of follow-up testing for Tl D.

7. The method of any one of claims 1-6, wherein the presence or absence of the first HLA II allele, the second HLA II allele and the third HLA II allele in the nucleic acid- containing sample is determined by:

(a) amplifying DNA in the sample using oligonucleotide primers specific for exons two of HLA DQB l and DQAl loci to provide amplified DNA; and

(b) contacting the amplified DNA with oligonucleotide probes specific for the first HLA II allele, the second HLA II allele and the third HLA II allele for a period of time sufficient for the oligonucleotide probes to hybridize to the first HLA II allele, the second HLA II allele and the third HLA II allele.

8. The method of claim 7, wherein the oligonucleotide probes are labelled with a detection reagent.

9. The method of claim 8, wherein each of the oligonucleotide probes is labelled with a different detection reagent, and the amplified DNA is contacted with each of the oligonucleotide probes simultaneously.

10. The method of claim 7, wherein the oligonucleotide probes are immobilized on a solid substrate.

1 1. The method of any one of claims 1-6, wherein the presence or absence of the first HLA II allele, the second HLA II allele and the third HLA II allele in the nucleic acid- containing sample is determined using restriction-fragment length polymorphism.

12. The method of any one of claims 1- 1 1 , wherein the nucleic acid-containing sample is selected from the group consisting of: blood, urine, saliva and sera.

13. A kit for use in the method of any one of claims 1- 12, the kit comprising:

(a) oligonucleotide probes specific for the first HLA II allele, the second HLA II allele and the third HLA II allele; and

(b) instructions for their use.

14. A microarray for use in the method of any one of claims 1- 1 1, comprising oligonucleotide probes specific for the first HLA II allele, the second HLA II allele and the third HLA II allele, each oligonucleotide probe being immobilized in a predefined location on the surface of a solid substrate.