In 3 unrelated children with recurrent infections, opsonization defect, and low serum MBL protein concentrations (MBLD; 614372), Sumiya et al. (1991) identified a homozygous 230G-A transition in the MBL2 gene, resulting in a gly54-to-asp (G54D) substitution. The change is known as the 'B' allele. The change occurs in the fifth collagen repeat of the protein, which the authors predicted may disrupt the collagen helix sufficiently to hinder MBL assembly and secretion by hepatocytes. Study of 16 members of the 3 families showed autosomal dominant co-inheritance of the G54D mutation and low serum MBL concentrations.
Super et al. (1992) found that among more than 100 randomly selected patients 5% were homozygotes for asp54. They showed that recombinant MBL with the asp54 substitution can oligomerize to yield a functional protein of about 650 kD that binds ligand and functions as an opsonin. However, although the recombinant MBL could form hexamers and therefore, like serum MBL, would be expected to act as a surrogate for C1q in classical pathway activation, this function was defective. The results indicated that the G54D substitution does not account for a deficiency state, but instead suggested that MBL has 2 predominant allelic forms that have overlapping function and that differ only in their ability to activate the classical pathway of complement.
By immunoassay, Lipscombe et al. (1992) found reduced serum MBL levels in individuals carrying the G54D change. However, 3 wildtype individuals had no detectable MBL protein. The G54D change showed a reduced capacity to activate complement through the MPB-initiated classical pathway. The authors suggested that both homozygous and heterozygous individuals would have reduced serum levels of the protein.
In a study of 123 healthy Danish individuals, Garred et al. (1992) found that 93 (75.6%) were gly54 homozygotes, 28 (22.8%) were heterozygous, and 2 (1.6%) were asp54 homozygotes; the frequency of the asp54 allele was 0.13. The median MBL concentration in the group of subjects with the asp54 allele was 6.4 times lower than in the gly54 homozygous group (195 and 1234 microg/l, respectively); however, the range in plasma concentrations of MBL was wide and overlapped between the groups. MBL protein was detected in the 2 asp54 homozygotes (9 and 387 microg/L), and there was no difference in relative mass and biologic activity between the 2 forms of the protein. Garred et al. (1992) concluded that the asp54 allele is able to produce a functional MBL protein that can be detected in serum at low concentrations.
In 2 adult patients with severe and unusual infections, Summerfield et al. (1995) identified homozygosity for the G54D change. A third patient was compound heterozygous for the G54D and R52C (154545.0003) changes. The authors concluded that MBL deficiency may confer a life-long risk of infection.
Turner et al. (2000) used MBL polymorphisms in the indigenous Australian population to date the appearance of the MBL B allele. They found a paucity of MBL structural gene mutations in 2 population groups from geographically distinct regions. Of 293 individuals tested, 289 were wildtype, and 4 were heterozygous for either the B or D allele. In each individual with an MBL mutation, the HLA haplotype profile suggested some Caucasian admixture. The authors hypothesized that the B mutation probably arose between 50,000 and 20,000 years ago, and that its absence from the founder gene pool of indigenous Australians may also partly explain their vulnerability to intracellular infections such as tuberculosis.
To test the hypothesis that a genetic predisposition to a proinflammatory state could favor the appearance of gestational diabetes mellitus (GDM), Megia et al. (2004) studied the R52C (154545.0003) and G54D polymorphisms of the MBL2 gene and plasma MBL levels from 105 consecutive women with GDM and 173 healthy pregnant women. They found an association between G54D and GDM (odds ratio = 2.03 (1.18-3.49); P less than 0.01), and this association remained significant when the presence of both mutated alleles was considered (odds ratio = 1.76 (1.04-2.96); P less than 0.05). GDM patients who carried the G54D mutation required insulin therapy more frequently and had heavier infants than GDM women homozygous for the wildtype allele. An inverse correlation in GDM patients between neonatal weight and plasma MBL levels was found, remaining significant after adjustment for confounding variables.