http://www.stockton-press.co.uk/ejhg deletions was detected, which removed between 1 and 30 exons from the gene. Most microdeletions and insertions occurred at homopolymeric tracts or direct repeats within the coding sequence. These features have not been observed in the other FA gene which has been cloned to date (FAC) and may be indicative of a higher mutation rate in FAA. This would explain why FA group A is much more common than the other complementation groups. The heterogeneity of the mutation spectrum and the frequency of intragenic deletions present a considerable challenge for the molecular diagnosis of FA. A scan of the entire coding sequence of the FAA gene may be required to detect the causative mutations, and scanning protocols will have to include methods which will detect the deletions in compound heterozygotes.

gastric intrinsic factor ͉ Imerslund-Grä sbeck syndrome ͉ intrinsic factor deficiency ͉ vitamin B 12 malabsorption ͉ megaloblastic anemia

The role of methylenetetrahydrofolate reductase (MTHFR C677T), glutathione Stransferases (GSTM1 and GSTT1 null, GSTP1 Ile105Val), and cytochromes p450 (CYP1A1*2A) genotypes in the etiology of childhood leukemia was simultaneously investigated. 144 Turkish children with acute lymphoblastic leukemia (ALL) and 33 with acute nonlymphoblastic leukemia (ANLL) were studied and compared with 185 healthy pediatric controls. The frequency of MTHFR genotype was insignificantly higher in ALL (7.7%) and ANLL (6.3%) than in controls (4.4%). Equal distribution of the GSTM1 null genotype was detected between ALL patients and controls (55%), while its incidence was slightly higher in ANLL patients (61.3%). Although GSTT1 null genotype was insignificantly lower in ALL patients (20.9%) than controls (22.7%), it was significantly underrepresented in ANLL patients (6.5%) (P = 0.05, OR 0.24, 95% CI 0.05-1.03). The homozygous frequency of GSTP1 genotype did not differ significantly between groups of ALL (3.7%), ANLL patients (9.1%) and controls (4.9%). Homozygous CYP1A1*2A genotype was underrepresented in ALL patients (1%) as compared to control (4.8%) but the differences did not reach to statistical significance (OR 0.21; 95% CI 0.03-1.72). Homozygosity for this genotype was not detected in ANLL patients. No particular association was noted between different combinations of combined genotypes and risk of development of childhood ALL and ANLL. These results suggested that there are no significant associations between the studied genotypes and the risk of developing either form of acute leukemia except GSTT1 null and homozygosity for CYP1A1 genotypes that may play protective roles in the development of ANLL in Turkish children. Am. J. Hematol. 73:154-160, 2003.

No abstract