Abstract
Protein glycosylation is an essential protein modification existing in all domains of life. About 2% of the human genome is involved in the glycosylation machinery and 50% of the human proteins are modified with glycans. In the endoplasmic reticulum, N-linked protein glycosylation begins with the assembly of an oligosaccharide precursor by step-wise transfer of sugar building blocks to a membrane-embedded, reduced polyprenol anchor called dolichol-phosphate. The fully assembled N-linked glycan precursor is then transferred co- or post-translationally to defined asparagine residues of target proteins.
Considering the abundance of N-glycosylated proteins throughout all human cell types, inherited glycosylation defects called congenital disorders of glycosylation (CDG) show multi-systemic involvement due to developmental defects in children. Most CDG involve a neurological component resulting in psychomotor retardation. Among the most common symptoms are epilepsy, hypotonia, hyporeflexia, strabismus, retinitis pigmentosa, polyneuropathy, myopathy, and cerebellar hypotrophy/hypoplasia. A common marker for CDG is the underglycosylation of blood serum proteins such as transferrin. Depending on the type of CDG, the N-linked glycans differ in their structure and can be detected in carbohydrate-deficient transferrin (CDT). Interestingly, alcoholic liver disease (ALD) is accompanied by a N-glycosylation defect as well. Besides a similar pattern of CDT, certain forms of CDG and ALD share another common symptom: liver fibrosis. Chronic alcoholism is estimated to be responsible for 4% of global death and the liver being the primary site of ethanol metabolism is particularly affected. Despite the long history of CDT in ALD, glycosylation deficiency in ALD has not yet been characterized at the molecular level.
So far, treatment of CDG is very restricted. In MPI-CDG, the mannose-phosphate isomerase is defective, an enzyme responsible for the conversion of fructose-6-phosphate to mannose-6-phosphate, thereby providing mannose for the glycosylation machinery. A simple supplementation of nutrition with mannose was shown to attenuate the manifestations of MPI-CDG.
In the first part of this thesis, we tested the potential of a cholesterol-lowering drug, Zaragozic acid A, to improve N-glycosylation in DPM1-CDG. DPM1 is a subunit of the dolichol-phosphate-mannose synthase which produces dolichol-phosphate-mannose, an important mannose donor for N-linked glycosylation. Zaragozic acid A inhibits an enzyme at a bifurcation of the anabolic pathway common to dolichol and cholesterol synthesis. At this bifurcation, the metabolic flux either goes towards cholesterol or dolichol synthesis. By blocking the enzyme for cholesterol synthesis, N-linked protein glycosylation improved as observed by increased dolichol-phosphate-mannose levels, increased dolichol-phosphate levels, normalized dolichol- and N-linked oligosaccharide distribution. The restored dolichol-phosphate-mannose pool also resulted in better GPI-anchor availability. We could compensate the lower dolichol-phopshate-mannose synthase activity by increasing substrate availability. Thus Zaragozic acid represents a possibility for drug treatment of DPM1-CDG.
The second part of the thesis is focusing on ethanol-induced N-glycosylation deficiency. We studied the effect of ethanol on N-glycosylation in two hepatoma cell lines. The two cell lines VA-13 and HepaRG are distinct because they express alcohol dehydrogenase and cytochrome P450 2E1, respectively, which confers ethanol-metabolizing properties usually lost in cultured hepatocytes. We found lower dolichol levels in ethanol-treated cells. Moreover, the dolichol-linked oligosaccharide pattern was disturbed with a lower fraction of the final N-glycan precursor. The changes in glycosylation were accompanied by transcriptional changes. DPM1 was downregulated while RPN2 was upregulated. The transcriptional changes could be regulatory responses since we observed an increased DPM synthase activity despite the lower DPM1 transcription.
In conclusion, we explored CDG and acquired deficiency in N-linked protein glycosylation by using a therapeutic approach to treat DPM1-CDG and by the characterization of N-glycosylation in an ethanol-induced glycosylation deficiency model. With this work we contributed to the search for treatment possibilities for CDG and promoted the understanding of ethanol-induced N-glycosylation deficiency.