No abstract

Presence or absence of N-acetylneuraminic acid (Neu5Ac) can change a sialylated glycoprotein's serum half-life and possibly its function. We evaluated the linearity, sensitivity, reproducibility, and accuracy of a HPAEC/PAD method to determine its suitability for routine simultaneous analysis of Neu5Ac and N-glycolylneuraminic acid (Neu5Gc). An effective internal standard for this analysis is 3-deoxy-d-glycero-d-galacto-2-nonulosonic acid (KDN). We investigated the effect of the Au working electrode recession and determined that linear range and sensitivity were dependent on electrode recession. Using an electrode that was 350 microm recessed from the electrode block, the minimum detection limits of Neu5Ac, KDN, and Neu5Gc were 2, 5, and 2 pmol, respectively, and were reduced to 1, 2, and 0.5 pmol using a new electrode. The response of standards was linear from 10 to 500 pmol (r2>0.99) regardless of electrode recession. When Neu5Ac, KDN, and Neu5Gc (200 pmol each) were analyzed repetitively for 48 h, area RSDs were <3%. Reproducibility was unaffected when injections of glycoprotein neuraminidase and acid digestions were interspersed with standard injections. Area RSDs of Neu5Ac and Neu5Gc improved when the internal standard was used. We determined the precision and accuracy of this method for both a recessed and a new working electrode by analyzing Neu5Ac and Neu5Gc contents of bovine fetuin and bovine and human transferrins. Results were consistent with published values and independent of the working electrode. The sensitivity, reproducibility, and accuracy of this method make it suitable for direct routine analysis of glycoprotein Neu5Ac and Neu5Gc contents.

Comprehensive characterization of the N-glycome of a therapeutic is challenging because glycans may harbor numerous modifications (e.g., phosphorylation, sulfation, sialic acids with possible O-acetylation). The current report presents a comparison of two chromatographic platforms for the comprehensive characterization of a recombinant human erythropoietin (rhEPO) N-glycome. The two platforms include a common workflow based on 2-AB-derivatization and hydrophilic interaction chromatography (HILIC) and a native N-linked glycan workflow employing high performance anion exchange (HPAE) chromatography. Both platforms were coupled to an Orbitrap mass spectrometer, and data dependent HCD fragmentation allowed confident structural elucidation of the glycans. Each platform identified glycans not revealed by the other, and both exhibited strengths and weaknesses. The reductive amination based HILIC workflow provided better throughput and sensitivity, had good isomer resolution, and revealed the presence of O-acetylated sialic acids. However, it exhibited poor performance toward phosphorylated glycans and did not reveal the presence of sulfated glycans. Furthermore, reductive amination introduced dehydration artifacts and modified the glycosylation profile in the rhEPO glycome. Conversely, HPAE provided unbiased charge classification (sialylation levels), improved isomer resolution, and revealed multiple phosphorylated and sulfated structures, but delivered lower throughput, had artifact peaks due to epimer formation, and loss of sialic acid O-acetylation. The MS based identification of phosphorylated and sulfated glycans was not possible in HILIC mode due to their poor solubility caused by the high acetonitrile concentrations employed at the beginning of the gradient. After analyzing the glycome by both approaches and determining the glycans present, a glycan library was created for site specific glycopeptide analyses. Glycopeptide analyses confirmed all the compositions annotated by the combined use of 2-AB- and native glycan workflows and provided site specific location of the glycans. These two platforms were complementary and in combination delivered a more thorough and comprehensive characterization of the rhEPO N-glycome, supporting regulatory conformance for the pharmaceutical industry.

No abstract

No abstract