An increasing number of genes have been linked to Alzheimer disease (AD) over the past decade. Recently, two novel polymorphisms of the methylene-tetrahydrofolate reductase (MTHFR) gene have been suggested to be associated with AD. The C677T polymorphism occurs in exon 4 of the MTHFR gene, whereas the A1298C polymorphism occurs in exon 7. The product of this gene, the enzyme 5,10-MTHFR catalyzes the formation of 5-tetrahydrofolate, the major circulating form of folate. Folate is involved in methylation of homocysteine and MTHFR gene mutations have been suggested to cause hyperhomocysteinemia. Elevated plasma Hcy levels have been also linked to the occurrence of AD. Initially, we developed semi-automated PCR-single-strand conformation polymorphism (PCR-SSCP) methods based on the use of the PhastSystem (Pharmacia Biotechnology), then for even more convenient genotyping, we implemented real-time PCR methods based on the use of the LightCycler (Roche) for detecting both the C677T and A1298C polymorphisms of the MTHFR gene. The PCR-SSCP and LighCycler methods were both validated against conventional PCR-RFLP methods and DNA sequencing for both polymorphisms. There was a complete match of genotypes with the PCR-RFLP, PCR-SSCP and LightCycler methods for both polymorphisms in 100 subjects studied. The allele frequencies were 0.70 C and 0.30 T for C677T, and 0.64 A and 0.36 C for A1298C, respectively. When homocysteine and cysteine were measured with an HPLC method in EDTA plasma of these subjects, there were some variations in the mean plasma homocysteine and cysteine levels, but the differences did not reach statistical significance (ANOVA and Mann-Whitney U-test). In summary, the PCR-SSCP and LightCycler methods studied here are reliable and convenient alternatives to conventional PCR-RFLP for large-scale screening of subjects for both the MTHFR C677T and A1298C polymorphisms. The observed genotype and allele frequencies are at variance with those reported previously. The lack of association between these polymorphisms and plasma homocysteine and cysteine levels is likely due to recent food supplementation with folate in the USA. Further studies with larger number of subjects are underway. In a collaborative study, healthy older control subjects with at least one apo E4 allele were found to have lower, more AD-like, levels of beta-amyloid(1-42) but not tau in their cerebrospinal fluid. This biologic finding was present despite these individuals being normal cognitively, and it may represent an early biomarker of AD pathophysiology. Longitudinal follow-up is needed to see if this finding is associa
