In addition to large-scale genetic association studies aimed at typing hundreds of thousands of single nucleotide polymorphisms (SNPs), smaller-scale projects focus on typing selected candidate SNPs. As large numbers of DNA samples have to be typed to assign genetic differences to distinct phenotypes, efficient high-throughput methods have to be employed. The Roche LightTyper and the LightCycler 480 Real-Time PCR instruments (Roche Diagnostics GmbH, Mannheim, Germany) have been developed for detection of SNPs, and protocols have been provided for the typing of deletions. LightTyper/LightCycler-based analyses can be performed rapidly with either 96- or 384-well microtiter plates (1), and thousands of SNP calls can be achieved per day at a reasonable cost (2).

Both systems allow post-amplification genotyping by fluorescence resonance energy transfer (FRET) and high-resolution DNA melting analysis with perfectly matched and discretely mismatched probes. The danger of contamination is minimized, as all reagents—including fluorescently labeled probes—are added to the microtiter plate at the reaction start. In addition, the FRET technology has been established and evaluated for the determination of DNA tetraplexes (3,4). The LightTyper and the LightCycler devices consist of a temperature-regulated heat block and a high-resolution charge-coupled device (CCD) camera. While the reaction mixture undergoes slow and continuous heating at steadily increasing temperature ramps, diodes emit blue light to excite the labeled probes in the LightTyper system (475-nm excitation filter), and fluorescence emission is recorded by the CCD camera (600-nm emission interference longpass filter). In the LightCycler system, excitation is achieved by a Xenon reflector lamp (483-nm excitation filter; 670-nm emission filter). Samples are grouped according to the melting temperature (T_m) and curve shapes, and scored in relation to an internal standard. Melting analyses are performed by the Roche LightTyper Software, Version 1.5.

Tandem repeat sequences are patchily disseminated throughout proand eukaryotic genomes and frequently exhibit a high degree of polymorphism. Among the techniques that have successfully been applied for the typing of such repeats are gel electrophoresis and capillary arrays (5,6,7), microchip capillary electrophoresis (8), mass spectrometry (9,10), and hybridization arrays (11). High-throughput melting analyses of PCR-amplified repeats would, for several reasons, be an interesting alternative and the need to develop appropriate protocols has recently been addressed (12). While significantly expediting analyses, it would reduce the danger of contamination, but also the expense of assays.

We have developed a LightTyper/LightCycler-based assay for rapid assessment of a tetranucleotide GTTT-repeat that occurs in the promoter of the human gene encoding the immunity-related GTPase family, M (IRGM) molecule (IRGM; OMIM *608212, gene map locus 5q33.1; rs60800371, position -308 to -311 relative to the ATG start codon), which plays a role in autophagy. We genotyped 2263 individuals originating from Ghana, West Africa, in a case-control study on susceptibility to pulmonary tuberculosis (TB). DNA was isolated from blood samples of study participants (AGOWA mag Maxi DNA Isolation Kit; AGOWA, Berlin, Germany) according to the manufacturer's instructions. In order to detect all IRGM variants occurring with a statistically high probability in our African study population, bidirectional resequencing of 1000 bp upstream of the ATG start codon (ABI Prism 6100; Applied Biosystems, Foster City, USA), 775 coding bp, and 1053 bp of the first intron of the IRGM gene was performed in 69 individuals (23 TB cases, 23 healthy controls with a positive tuberculin response, and 23 healthy controls with a negative tuberculin response). Sequence alignment showed that the GTTT repeat occurred as GTTT₁, GTTT₃, and GTTT₄ alleles. The GTTT₃ variant had not yet been recognized.

Amplification primers and sensor/anchor (S/A) oligonucleotides were designed with Roche LightCycler ProbeDesign, Version 2.0.R.33. The sensor oligonucleotide perfectly matched the sequence of the GTTT₄ 16-bp repeat with an overlap of 15 bp in the 5′ orientation and 9 bp in the 3′ orientation and required for reaching the highest T_m. The sensor then would not be completely complementary to a matrix sequence containing the alternative GTTT₁ or GTTT₃ alleles and form nonbinding loops of GTTTGTTTGTTT nucleotides (GTTT₁; Figure 1A) or GTTT nucleotides (GTTT₃; not shown in Figure 1A), respectively. Forward (F) and reverse (R) PCR primers (GTTT₄: 507 bp) and S/A oligonudeotides were 5′-CATCCTAACTCACCTGCTCTTCTA-3′, 5′-TGTTAGAGATCAC CTCTGGCAA-3′, 5′-TAAGCATTGG G T T T T G T T T G T T T G T T T G TTTGTTTTGAGA-6-FAM-3′ (6-FAM, 6-carboxyfluorescein) and 5′-Cy5-GGAGTCTTGCTCTGTTG CCAGGCTGGAG-phosphate-3′ (Cy5) (6-FAM-and Cy5-labeled nucleotides were synthesized by biomers.net GmbH, Ulm, Germany). Microtiter plates were prepared in the 384-well format with the Tecan GENESIS Workstation 200 pipetting instrument controlled by Gemini steering software v3.50 (Tecan, Crailsheim, Germany).

Figure 1. Design, DNA resquencing results and melting peaks of the assay of the IRGM tetranucleotide promoter oligopeat typing by fluorescence resonance energy tranfer (FRET). (Click to enlarge)

PCR conditions were optimized for the annealing temperature and Mg^2 + concentration. The final PCR conditions (Primus 96 Thermocycler with 384-well block; MWG Biotech AG, Ebersberg, Germany) were 94°C for 3 min, 45 cycles of 94°C for 20 s, 60°C for 30 s, 72°C for 30 s, with an extension at 72°C for 5 min, with final concentrations of 3 mM MgCl₂, 0.1 µM forward primer, 0.5 µM reverse primer, 0.2 µM sensor, 0.2 µM anchor, 200 µM dNTPs, 1 U Taq DNA polymerase, and 10 ng matrix DNA per reaction. The 10-µL reaction mixtures were covered with 10 µL of light mineral oil (Sigma Aldrich, Munich, Germany). The PCR success rate was 96.6%. Prior to subsequent melting, amplicons were denatured for 3 min at 95°C.

In the LightTyper instrument, samples were heated from 40°C to 80°C at a ramp rate of 0.1°C/s with simultaneous continuous fluorescence monitoring. Melting data was plotted as a first negative derivation of fluorescence properties depending on T_m (−dF/dT). Different T_m of perfect or reduced binding of the sensor oligonucleotide to the matrix DNA (Figure 1A) of different alleles allows for unambiguously distinguishing homozygous from heterozygous individuals and for assessing the number of repeats. The melting curves of the IRGM promoter genotypes are shown in Figure 1C. Six genotypes were identified (GTTT_1/1, n = 364; GTTT_3/3, n = 85; GTTT_4/4, n = 303; GTTT_1/3, n = 425; GTTT_1/4, n = 717; GTTT_3/4, n = 292). Alleles with two repeats were not observed at all in our study population, neither in the resequencing experiments nor in the analysis of melting curves.

Compared with the alternative techniques that are currently applied for oligotandem repeat typing, the LightTyper/LightCycler FRET-based system provides the advantages of high throughput, a reduced potential of DNA contamination, and less timeconsuming implementation and performance at a reasonable cost.