Solving glycosylation disorders: fundamental approaches reveal complicated pathways

Abstract

Over 100 human genetic disorders result from mutations in glycosylation-related genes. In 2013, a new glycosylation disorder was reported every 17 days. This trend will probably continue given that at least 2% of the human genome encodes glycan-biosynthesis and -recognition proteins. Established biosynthetic pathways provide many candidate genes, but finding unanticipated mutated genes will offer new insights into glycosylation. Simple glycobiomarkers can be used in narrowing the candidates identified by exome and genome sequencing, and those can be validated by glycosylation analysis of serum or cells from affected individuals. Model organisms will expand the understanding of these mutations' impact on glycosylation and pathology. Here, we highlight some recently discovered glycosylation disorders and the barriers, breakthroughs, and surprises they presented. We predict that some glycosylation disorders might occur with greater frequency than current estimates of their prevalence. Moreover, the prevalence of some disorders differs substantially between European and African Americans.

Figure 1

Identification of Glycosylation-Related Disorders The graph shows the cumulative number of human glycosylation disorders in various biosynthetic pathways and the year of their identification. In most cases, the year indicates the definitive proof of specific linked genes or causative mutations. Prior to ∼2000, discovery of glycosylation disorders was based on compelling biochemical evidence and confirmed by the identification of mutations in the genes. In 2013 alone, a genetically proven glycosylation disorder was reported, on average, every 17 days.

Figure 2

N-Linked Glycosylation Pathway Showing Activation of Precursor Molecules and Synthesis of Dolichol-P-P-GlcNAc₂Man₉Glc₃ and Transfer to Protein Early steps in the N-linked pathway require conversion of the lipid molecule polyprenol to the oligosaccharide carrier molecule dolichol, which serves as a scaffold for LLO. Monosaccharides activated in the cytoplasm are used as substrates for several glycosyltransferases that extend the growing chain. Once dolichol-P-P-GlcNAc₂Man₅ is formed, it flips into the ER lumen and is further extended to the final product with Dol-P-Man and Dol-P-Glc as donors for different mannosyl- and glucosyl-transferases, respectively. The oligosaccharyltransferase complex then transfers the completed structure onto nascent polypeptides. The processing of N-glycans is not shown. Genetic defects within this pathway are shown by an X. This figure was adapted from Figure 42.1 in Essentials of Glycobiology, Second Edition from Cold Spring Harbor Laboratory Press and is used with permission.

Figure 3

GPI-Anchor Synthetic Pathway The initial two steps of GPI-anchor biosynthesis in mammalian cells require the formation of GlcNAc-PI and then de-N-acetylation of GlcNAc-PI to form GlcN-PI, which occurs on the cytoplasmic face of the ER. GlcN-PI then flips into the ER lumen, where further extension of the glycan occurs by the addition of mannose and ethanolamine phosphate. The completed structure is transferred to multiple acceptor proteins. One of the acyl chains is removed. The GPI-anchored proteins move to the Golgi, where one of the acyl chains is removed and another is added. Genetic defects with in this pathway are noted in red.

Figure 4

Burden of Putatively Deleterious Alleles in Glycosylation-Pathway Genes We queried 1,828 EAs (A) and 1,828 AAs (B) in the ESP6500 data set. Strict filtering criteria for identifying putatively deleterious alleles are shown in black, and those using moderate filtering criteria are highlighted in color. Under moderate filtering, 20.7% of EAs carry one predicted CDG allele, 3.6% carry two alleles, and <1% carry three or more alleles. In comparison, 26.3% of AA carry one allele, 6.9% carry two alleles, 1.4% carry three alleles, and <1% carry four or more alleles.

Figure 5

Expected Frequency of Individuals with Two Putatively Deleterious Alleles in 53 Glycosylation-Pathway Genes EAs (A) and AAs (B). Strict filtering criteria for identifying putatively deleterious alleles are shown in black, and moderate filtering criteria are highlighted in color. Under moderate filtering, 9/10,000 EAs and 21/10,000 AAs are expected to carry two predicted CDG alleles in one gene.