Abolishment of N-glycan mannosylphosphorylation in glyco-engineered Saccharomyces cerevisiae by double disruption of MNN4 and MNN14 genes.

Mannosylphosphorylated glycans are found only in fungi, including yeast, and the elimination of mannosylphosphates from glycans is a prerequisite for yeast glyco-engineering to produce human-compatible glycoproteins. In Saccharomyces cerevisiae, MNN4 and MNN6 genes are known to play roles in mannosylphosphorylation, but disruption of these genes does not completely remove the mannosylphosphates in N-glycans. This study was performed to find unknown key gene(s) involved in N-glycan mannosylphosphorylation in S. cerevisiae. For this purpose, each of one MNN4 and five MNN6 homologous genes were deleted from the och1Δmnn1Δmnn4Δmnn6Δ strain, which lacks yeast-specific hyper-mannosylation and the immunogenic α(1,3)-mannose structure. N-glycan profile analysis of cell wall mannoproteins and a secretory recombinant protein produced in mutants showed that the MNN14 gene, an MNN4 paralog with unknown function, is essential for N-glycan mannosylphosphorylation. Double disruption of MNN4 and MNN14 genes was enough to eliminate N-glycan mannosylphosphorylation. Our results suggest that the S. cerevisiae och1Δmnn1Δmnn4Δmnn14Δ strain, in which all yeast-specific N-glycan structures including mannosylphosphorylation are abolished, may have promise as a useful platform for glyco-engineering to produce therapeutic glycoproteins with human-compatible N-glycans.

Data for analysis of mannose-6-phosphate glycans labeled with fluorescent tags.

Mannose-6-phosphate (M-6-P) glycan plays an important role in lysosomal targeting of most therapeutic enzymes for treatment of lysosomal storage diseases. This article provides data for the analysis of M-6-P glycans by high-performance liquid chromatography (HPLC) and matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry. The identities of M-6-P glycan peaks in HPLC profile were confirmed by measuring the masses of the collected peak eluates. The performances of three fluorescent tags (2-aminobenzoic acid [2-AA], 2-aminobenzamide [2-AB], and 3-(acetyl-amino)-6-aminoacridine [AA-Ac]) were compared focusing on the analysis of bi-phosphorylated glycan (containing two M-6-Ps). The bi-phosphorylated glycan analysis is highly affected by the attached fluorescent tag and the hydrophilicity of elution solvent used in HPLC. The data in this article is associated with the research article published in "Comparison of fluorescent tags for analysis of mannose-6-phosphate glycans" (Kang et al., 2016 [1]).

Comparison of fluorescent tags for analysis of mannose-6-phosphate glycans.

Mannose-6-phosphate (M-6-P) glycan analysis is important for quality control of therapeutic enzymes for lysosomal storage diseases. Here, we found that the analysis of glycans containing two M-6-Ps was highly affected by the hydrophilicity of the elution solvent used in high-performance liquid chromatography (HPLC). In addition, the performances of three fluorescent tags--2-aminobenzoic acid (2-AA), 2-aminobenzamide (2-AB), and 3-(acetyl-amino)-6-aminoacridine (AA-Ac)--were compared with each other for M-6-P glycan analysis using HPLC and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. The best performance for analyzing M-6-P glycans was shown by 2-AA labeling in both analyses.

Increased mannosylphosphorylation of N-glycans by heterologous expression of YlMPO1 in glyco-engineered Saccharomyces cerevisiae for mannose-6-phosphate modification.

Mannosylphosphorylated N-glycans found in yeasts can be converted to those containing mannose-6-phosphate, which is a key factor for lysosomal targeting. In the traditional yeast Saccharomyces cerevisiae, both ScMNN4 and ScMNN6 genes are required for efficient mannosylphosphorylation. ScMnn4 protein has been known to be a positive regulator of ScMnn6p, a real enzyme for mannosylphosphorylation. On the other hand, YlMpo1p, a ScMnn4p homologue, mediates mannosylphosphorylation in Yarrowia lypolytica without the involvement of ScMnn6p homologues. In this study, we show that heterologous expression of YlMpo1p can perform and enhance mannosylphosphorylation in S. cerevisiae in the absence of ScMnn4p and ScMnn6p. Moreover, the level of mannosylphosphorylation of N-glycans enhanced by YlMpo1p overexpression is much higher than that with ScMnn4p overexpression, and this is highlighted further in Scmnn4- and Scmnn6-disrupted mutants. When YlMpo1p overexpression is applied to glyco-engineered S. cerevisiae in which the synthesis of immunogenic glycans is abolished, a great increase of bi-mannosylphosphorylated glycan is observed. Through an in vitro process involving the uncapping of the outer mannose residue, this bi-mannosylphosphorylated structure is changed to a bi-phosphorylated structure with high affinity for mannose-6-phosphate receptor. The superior ability of YlMpo1p to increase bi-mannosylphosphorylated glycan in yeast shows promise for the production of therapeutic enzymes with improved lysosomal targeting capability.

Molecular characterization of acidic peptide:N-glycanase from the dimorphic yeast Yarrowia lipolytica.

Peptide:N-glycanase (PNGase) A is used preferentially to cleave the glycans from plant and insect glycopeptides. Although many putative PNGase A homologous genes have been found in the plant and fungus kingdoms through sequence similarity analyses, only several PNGases from plants and one from a filamentous fungus have been characterized. In this study, we identified and characterized a PNGase A-like enzyme, PNGase Yl, in the dimorphic yeast Yarrowia lipolytica. The corresponding gene was cloned and recombinantly expressed in Pichia pastoris. The purified enzyme cleaved glycans from glycopeptides with the maximum activity at pH 5. No metal ions were required for full activity, and rather it was repressed by three metal ions (Fe(3+), Cu(2+) and Zn(2+)). Using glycopeptide substrates, PNGase Yl was shown to release various types of N-glycans including high-mannose and complex-type glycans as well as glycans containing core-linked α(1,3)-fucose that are frequently found in plants and insects. Moreover, in comparison with PNGase A, PNGase Yl was able to cleave with higher efficiency the glycans from some denatured glycoproteins. Taken together, our results suggest that PNGase Yl, the first biochemically characterized yeast PNGase A homologue, can be developed through protein engineering as a useful deglycosylation tool for N-glycosylation study.

N-glycan analysis of human α1-antitrypsin produced in Chinese hamster ovary cells.

Human alpha-1-antitrypsin (α1AT) is a glycoprotein with protease inhibitor activity protecting tissues from degradation. Patients with inherited α1AT deficiency are treated with native α1AT (nAT) purified from human plasma. In the present study, recombinant α1AT (rAT) was produced in Chinese hamster ovary (CHO) cells and their glycosylation patterns, inhibitory activity and in vivo half-life were compared with those of nAT. A peptide mapping analysis employing a deglycosylation reaction confirmed full occupancy of all three glycosylation sites and the equivalency of rAT and nAT in terms of the protein level. N-glycan profiles revealed that rAT contained 10 glycan structures ranging from bi-antennary to tetra-antennary complex-type glycans while nAT displayed six peaks comprising majorly bi-antennary glycans and a small portion of tri-antennary glycans. In addition, most of the rAT glycans were shown to have only core α(1 - 6)-fucose without terminal fucosylation, whereas only minor portions of the nAT glycans contained core or Lewis X-type fucose. As expected, all sialylated glycans of rAT were found to have α(2 - 3)-linked sialic acids, which was in sharp contrast to those of nAT, which had mostly α(2 - 6)-linked sialic acids. However, the degree of sialylation of rAT was comparable to that of nAT, which was also supported by an isoelectric focusing gel analysis. Despite the differences in the glycosylation patterns, both α1ATs showed nearly equivalent inhibitory activity in enzyme assays and serum half-lives in a pharmacokinetic experiment. These results suggest that rAT produced in CHO cells would be a good alternative to nAT derived from human plasma.