SummaryFour patients with hereditary glucose phosphate isomerase (GPI) deficiency and their parents have been studied by means of various enzymatic, immunologic, electrophoretic, and stability methods. The four defective mutants enzymes (identified here as “GPI Barcelona,” “GPI Utrecht,” “GPI Nijmegen,” and “GPI Kortrijk”) were antigenically identical with normal enzyme as judged by double immunodiffusion and microcomplement fixation tests. Immunologic specific activity (i.e., the ratio of enzyme activity to antigen concentration) was slightly lowered for three variants (between 60 and 75% of normal). The various methods used allowed us to establish that all of the patients were heterozygous for two different mutated alleles inherited from each parent. One of them was a silent allele which did not seem to code for any enzymatic cross-reacting material; the other mutated gene coded for a structurally modified glucose phosphate isomerase subunit.In two parents heterozygous for a structurally modified gene, only two enzymatic forms were detected, even in the young cells synthesizing actively proteins such as granulocytes: one was the normal homodimer, and the other corresponded to the heterodimer “normal subunit-mutant subunit”; the mutant homodimer was not detected. The assumption that, in these heterozygotes, the mutant subunit dimerized more easily with a normal subunit than with another mutant subunit could be proposed. In contrast, the three expected enzyme forms were detected in the leukocytes from a woman heterozygous for GPI Nijmegen.From these results the defect in activity observed in the patients could be explained in the following way: 50% of the defect was due to the inheritance of a silent gene. The decrease below 50% of the residual activity in red cells was due to the molecular instability of the mutant enzyme associated in three cases (GPI Barcelona, GPI Utrecht, and GPI Kortrijk) with the decreased immunologic specific activity.SpeculationIn spite of the relatively high frequency of the mutations which are responsible for a silent GPI structural gene, GPI deficiency due to the inheritance of two silent genes has never been described. It can be speculated that the homozygous state of such a mutation, which would lead to null enzyme activity in all the tissues, could be lethal. As in most cases of genetic disorders due to silent genes, the nature of the mutation involved is unknown. A deletion, a polar mutation, a structural mutation leading to a highly labile product, or a cis-dominant regulator mutation are the different genetic hypotheses which could be put forward.