Balancing N-linked glycosylation to avoid disease

Abstract

Complete loss of N-glycosylation is lethal in both yeast and mammals. Substantial deficiencies in some rate-limiting biosynthetic steps cause human congenital disorders of glycosylation (CDG). Patients have a range of clinical problems including variable degrees of mental retardation, liver dysfunction, and intestinal disorders. Over 60 mutations in phosphomannomutase (encoded by PMM2) diminish activity and cause CDG-Ia. The severe mutation R141H in PMM2 is lethal when homozygous, but heterozygous in about 1/70 Northern Europeans. Another disorder, CDG-Ic, is caused by mutations in ALG6, an α1,3glucosyl transferase used for lipid-linked precursor synthesis, yet some function-compromising mutations occur at a high frequency in this gene also. Maintenance of seemingly deleterious mutations implies a selective advantage or positive heterosis. One possible explanation for this is that production of infective viruses such as hepatitis virus B and C, or others that rely heavily on host N-glycosylation, is substantially inhibited when only a tiny fraction of their coat proteins is misglycosylated. In contrast, this reduced glycosylation does not affect the host. Prevalent functional mutations in rate-limiting glycosylation steps could provide some resistance to viral infections, but the cost of this insurance is CDG. A balanced glycosylation level attempts to accommodate these competing agendas. By assessing the occurrence of a series of N-glycosylation-compromising alleles in multi-genic diseases, it may be possible to determine whether impaired glycosylation is a risk factor or a major determinant underlying their pathology.