Background The heritability of autism spectrum disorder is currently estimated at 55%. Identification of the molecular basis of patients with syndromic autism extends our understanding of the pathogenesis of autism in general. The objective of this study was to find the gene mutated in eight patients from a large kindred, who suffered from autism spectrum disorder, arthrogryposis and epilepsy. Methods and results By linkage analysis and exome sequencing, we identified deleterious mutations in SLC35A3 in these patients. SLC35A3 encodes the major Golgi uridine diphosphate N-acetylglucosamine (UDP-GlcNAc) transporter. In Golgi vesicles isolated from patient fibroblasts the transport of the respective nucleotide sugar was significantly reduced causing a massive decrease in the content of cell surface expressed highly branched N-glycans and a concomitant sharp increase of lower branched glycoforms. Conclusions Spontaneous mutation in SLC35A3 has been discovered in cattle worldwide, recapitulating the human phenotype with arthrogryposis and additional skeletal defects known as Complex Vertebral Malformation syndrome. The skeletal anomalies in the mutant cattle and in our patients, and perhaps even the neurological symptoms are likely the consequence of the lack of high-branched N-glycans and the concomitant abundance of lower-branched glycoforms at the cell surface. This pattern has previously been associated with growth arrest and induction of differentiation. With this study, we add SLC35A3 to the gene list of autism spectrum disorders, and underscore the crucial importance of UDP-GlcNAc in the regulation of the N-glycan branching pathway in the Golgi apparatus.

Synthetic sugar analogs are widely applied in metabolic oligosaccharide engineering (MOE) and as novel drugs to interfere with glycoconjugate biosynthesis. However, mechanistic insights on their exact cellular metabolism over time are mostly lacking. We combined ion-pair ultrahigh performance liquid chromatography–triple quadrupole mass spectrometry mass spectrometry using tributyl- and triethylamine buffers for sensitive analysis of sugar metabolites in cells and organisms and identified low abundant nucleotide sugars, such as UDP-arabinose in human cell lines and CMP-sialic acid (CMP-NeuNAc) in Drosophila. Furthermore, MOE revealed that propargyloxycarbonyl (Poc)-labeled ManNPoc was metabolized to both CMP-NeuNPoc and UDP-GlcNPoc. Finally, time-course analysis of the effect of antitumor compound 3F_ax-NeuNAc by incubation of B16-F10 melanoma cells with N-acetyl-D-[UL-¹³C6]glucosamine revealed full depletion of endogenous ManNAc 6-phosphate and CMP-NeuNAc within 24 h. Thus, dynamic tracing of sugar metabolic pathways provides a general approach to reveal time-dependent insights into the metabolism of synthetic sugars, which is important for the rational design of analogs with optimized effects.

Congenital disorders of glycosylation type 1 (CDG‐I) comprise a group of 27 genetic defects with heterogeneous multisystem phenotype, mostly presenting with nonspecific neurological symptoms. The biochemical hallmark of CDG‐I is a partial absence of complete N‐glycans on transferrin. However, recent findings of a diagnostic N‐tetrasaccharide for ALG1‐CDG and increased high‐mannose N‐glycans for a few other CDG suggested the potential of glycan structural analysis for CDG‐I gene discovery. We analyzed the relative abundance of total plasma N‐glycans by high resolution quadrupole time‐of‐flight mass spectrometry in a large cohort of 111 CDG‐I patients with known (n = 75) or unsolved (n = 36) genetic cause. We designed single‐molecule molecular inversion probes (smMIPs) for sequencing of CDG‐I candidate genes on the basis of specific N‐glycan signatures. Glycomics profiling in patients with known defects revealed novel features such as the N‐tetrasaccharide in ALG2‐CDG patients and a novel fucosylated N‐pentasaccharide as specific glycomarker for ALG1‐CDG. Moreover, group‐specific high‐mannose N‐glycan signatures were found in ALG3‐, ALG9‐, ALG11‐, ALG12‐, RFT1‐, SRD5A3‐, DOLK‐, DPM1‐, DPM3‐, MPDU1‐, ALG13‐CDG, and hereditary fructose intolerance. Further differential analysis revealed high‐mannose profiles, characteristic for ALG12‐ and ALG9‐CDG. Prediction of candidate genes by glycomics profiling in 36 patients with thus far unsolved CDG‐I and subsequent smMIPs sequencing led to a yield of solved cases of 78% (28/36). Combined plasma glycomics profiling and targeted smMIPs sequencing of candidate genes is a powerful approach to identify causative mutations in CDG‐I patient cohorts.

The SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) protein syntaxin-5 (Stx5) is essential for Golgi transport. In humans, the STX5 mRNA encodes two protein isoforms, Stx5 Long (Stx5L) from the first starting methionine and Stx5 Short (Stx5S) from an alternative starting methionine at position 55. In this study, we identify a human disorder caused by a single missense substitution in the second starting methionine (p.M55V), resulting in complete loss of the short isoform. Patients suffer from an early fatal multisystem disease, including severe liver disease, skeletal abnormalities and abnormal glycosylation. Primary human dermal fibroblasts isolated from these patients show defective glycosylation, altered Golgi morphology as measured by electron microscopy, mislocalization of glycosyltransferases, and compromised ER-Golgi trafficking. Measurements of cognate binding SNAREs, based on biotin-synchronizable forms of Stx5 (the RUSH system) and Förster resonance energy transfer (FRET), revealed that the short isoform of Stx5 is essential for intra-Golgi transport. Alternative starting codons of Stx5 are thus linked to human disease, demonstrating that the site of translation initiation is an important new layer of regulating protein trafficking.

Phosphoglucomutase 1 (PGM1) is a key enzyme for the regulation of energy metabolism from glycogen and glycolysis, as it catalyzes the interconversion of glucose 1-phosphate and glucose 6-phosphate. PGM1 deficiency is an autosomal recessive disorder characterized by a highly heterogenous clinical spectrum, including hypoglycemia, cleft palate, liver dysfunction, growth delay, exercise intolerance, and dilated cardiomyopathy. Abnormal protein glycosylation has been observed in this disease. Oral supplementation with D-galactose efficiently restores protein glycosylation by replenishing the lacking pool of UDP-galactose, and rescues some symptoms, such as hypoglycemia, hepatopathy, and growth delay. However, D-galactose effects on skeletal muscle and heart symptoms remain unclear. In this study, we established an in vitro muscle model for PGM1 deficiency to investigate the role of PGM1 and the effect of D-galactose on nucleotide sugars and energy metabolism. Genome-editing of C2C12 myoblasts via CRISPR/Cas9 resulted in Pgm1 (mouse homologue of human PGM1, according to updated nomenclature) knockout clones, which showed impaired maturation to myotubes. No difference was found for steady-state levels of nucleotide sugars, while dynamic flux analysis based on ¹³C6-galactose suggested a block in the use of galactose for energy production in knockout myoblasts. Subsequent analyses revealed a lower basal respiration and mitochondrial ATP production capacity in the knockout myoblasts and myotubes, which were not restored by D-galactose. In conclusion, an in vitro mouse muscle cell model has been established to study the muscle-specific metabolic mechanisms in PGM1 deficiency, which suggested that galactose was unable to restore the reduced energy production capacity.