MALDI-TOF mass spectrometry for high throughput SNP detection: Application to type 2 diabetes Erdos, M., Braun, A., Mohlke, K., Silander, K., Ghosh, S., Collins, F.S. for the FUSION (Finland-US Investigation of NIDDM) study group. NHGRI, NIH, Bethesda, MD, Sequenom Inc., San Diego, CA Matrix-Assisted Laser Desorption Ionization (MALDI) mass spectrometry has recently been introduced as a powerful analytical tool for the analysis of short DNA sequences by mass. Ions are generated by laser bombardment of DNA fragments embedded in a matrix of 3-hydroxypicolinic acid. These large DNA ions are accelerated through a path towards the detector measuring the time of flight (TOF) which is correlated with the mass of the DNA fragment. SNP analysis by MALDI-TOF is initiated by multiplex PCR amplification of genomic DNA, using primers flanking the SNPs of interest. A locus-specific primer which abuts the variable nucleotide of each SNP is then used in a primer oligo base extension reaction; primer extension is performed with DNA polymerase in the presence of dideoxynucleotides that terminate the reactions at a specific length determined by the allele present. The extended products are then purified and subjected to mass spectrometric analysis. In a pilot study, we have designed and successfully implemented the multiplex analysis by MALDI-TOF mass spectrometry of four genes with known polymorphisms that have been implicated in diabetes. The variants tested are SNPs in peroxisome proliferator-activated receptor gamma 2, insulin receptor substrate-1, uncoupling protein-2 and glucokinase genes. These genes are known to play a role in insulin sensitivity, insulin receptor signaling, obesity, and/or glucose metabolism, all traits which are associated with type 2 diabetes. Genotypes in the pilot study have been highly reproducible. We are currently extending this analysis to 582 probands of diabetic sib-pairs and 232 controls consisting of elderly non-diabetic Finnish subjects. Comparison of SNP allele frequencies between diabetic probands and controls should allow the identification or elimination of these candidate gene variants for association with type 2 diabetes in the Finnish population.