The Lewis ␣(1,3/1,4)-fucosyltransferase, Fuc-TIII, encoded by the FUT3 gene is responsible for the final synthesis of Le a and Le b antigens. Various point mutations have been described explaining the Lewis negative phenotype, Le(a؊b؊), on erythrocytes and secretions. Two of these, T202C and C314T originally described in a Swedish population, have not been found as single isolated point mutations so far. To define the relative contribution of each of these two mutations to the Lewis negative phenotype, we cloned and made chimeric FUT3 constructs separating the T202C mutation responsible for the amino acid change Trp 68 3 Arg, from the C314T mutation leading to the Thr 105 3 Met shift. COS-7 cells were transfected and the expression of Fuc-TIII enzyme activity and the presence of Lewis antigens were determined. There was no decrease in enzyme activity nor of immunofluorescence staining on cells transfected with the construct containing the isolated C314T mutation compared with cells transfected with a wild type FUT3 allele control. No enzyme activity nor immunoreactivity for Lewis antigens was detected in FUT3 constructs containing both mutations in combination. The T202C mutation alone decreased the enzyme activity to less than 1% of the activity of the wild type FUT3 allele. These results demonstrate, that the Trp 68 3 Arg substitution in human Fuc-TIII is the capital amino acid change responsible for the appearance of the Le(a؊b؊) phenotype on human erythrocytes in individuals homozygous for both the T202C and C314T mutations.The Lewis histo-blood group system comprises different complex carbohydrate structures, which participate in different biological processes such as embryogenesis, tissue differentiation, tumor metastasis, inflammation, and bacterial adhesion (1). Le a1 and Le b (2, 3) are the major Lewis antigens found on human erythrocytes. These fucosylated glycosphingolipids are synthesized by exocrine epithelial cells (4) and are secondarily passively adsorbed onto erythrocytes in the peripheral circulation giving these blood cells their Lewis phenotype (5).It was shown as early as the 1950's, that the Lewis phenotype of erythrocytes is influenced by the ABH secretor status of the individual (6), and the erythrocyte phenotype is the result of the epistatic interaction of the Lewis (Le-le) and the salivary ABH secretor (Se-se) loci (7). Fucosyltransferases (Fuc-Ts) encoded by genes at these loci compete and interact with each other and with other glycosyltransferases to determine the individual's final Lewis and secretor phenotypes.Since 1990, seven human fucosyltransferase genes (FUT1-7) have been cloned, sequenced, and characterized according to acceptor specificities (8 -18). The human ␣(1,2)-Fuc-T gene family comprises FUT1 encoding for the H enzyme and FUT2 encoding for the human secretor ␣(1,2)-fucosyltransferase (9). FUT3-7 encode for ␣(1,3)-Fuc-Ts. Two of these human enzymes, respectively (10,14), also express ␣(1,4)-fucosyltransferase activities (15,19,20).Five main missense mutations have be...

The human Lewis gene encodes an alpha(1,3/1,4)-fucosyltransferase responsible for synthesis of the Le(a) and a Le(b) antigens. To define the molecular background for non-functional Lewis genes we have sequenced PCR-amplified DNA fragments from two Le(a-b-) individuals. One was homozygously mutated at nucleotides 202(T --> C) and 314 (C --> T), altering Trp68 to Arg and Thr105 to Met, and the other was homozygously mutated at nucleotides 59 (T --> G) and 1067 (T --> A), altering Leu20 to Arg and Ile356 to Lys. Using PCR we screened for these and additionally one other mutation at nucleotide 508 (G --> A) among 40 Caucasians. Of 15 Le(a-b-) individuals, 7 typed as le59/1067le202/314, 4 as le202/314le202/314 and 1 as le59/1067le59/1067. Of 21 Le(a-b+) and 4 Le(a+b-), 17 typed as LeLe and 7 as Lele202/314. A pedigree study of 8 Lewis-positive individuals showed that the mutations at nucleotides 202 and 314 were located on the same allele.

Screening the FUT6 gene of 40 Swedish individuals, originally selected for genotyping of FUT3, revealed an unexpected high frequency of mutations. Four were originally typed as homozygous for the enzyme lethal mutation G739A by Taqal restriction pattern, but only one lacked plasma a(l,3)fucosyltransferase activity. Cloning and sequencing of FUT6 from 2 of them revealed a new allele, without the G739A mutation, but with two new point mutations C738T and G977A. Segregation of this allele was confirmed in Swedish and Indonesian families. Since G739A and C738T mutations are only one nucleotide apart and induce the same modification of Taqal cleavage, a new screening strategy for FUT6 was adopted. The homozygous inactivating G739A mutation was for the first time identified in Caucasian and Polynesian individuals, both lacking plasma enzyme activity. The mutation C370T was present in 25 of the 40 Swedish individuals and the inactivating mutation C945A was not found at all. These findings stress the dangers of transferring restriction enzyme genotype strategies from one population to another and of inferring phenotypes from genotypes without phenotyping and/or performing confirmatory cloning and sequencing.

No abstract

Screening the FUT6 gene of 40 Swedish individuals, originally selected for genotyping of FUT3, revealed an unexpected high frequency of mutations. Four were originally typed as homozygous for the enzyme lethal mutation G739A by Taq alpha I restriction pattern, but only one lacked plasma alpha(1,3)fucosyltransferase activity. Cloning and sequencing of FUT6 from 2 of them revealed a new allele, without the G739A mutation, but with two new point mutations C738T and G977A. Segregation of this allele was confirmed in Swedish and Indonesian families. Since G739A and C738T mutations are only one nucleotide apart and induce the same modification of Taq alpha I cleavage, a new screening strategy for FUT6 was adopted. The homozygous inactivating G739A mutation was for the first time identified in Caucasian and Polynesian individuals, both lacking plasma enzyme activity. The mutation C370T was present in 25 of the 40 Swedish individuals and the inactivating mutation C945A was not found at all. These findings stress the dangers of transferring restriction enzyme genotype strategies from one population to another and of inferring phenotypes from genotypes without phenotyping and/or performing confirmatory cloning and sequencing.