Quantitative comparison of global carbohydrate structures of glycoproteins using LC-MS and in-source fragmentation.

A comparative method for the quantitative analysis of the ratio of oxonium fragment (reporter) ions derived from sialic acid and N-acetylhexosamine residues on a large intact glycoprotein, the B domain of recombinant human factor VIII (rhFVIII), was developed. The method utilized liquid chromatography-electrospray ionization mass spectrometry (LC-ESI MS) on a single-quadrupole instrument. During development, systematic approaches such as full-matrix and simplex strategies were used for the optimization of the signal-to-noise ratio by controlling source temperature and cone voltage. The method was found to be precise (RSD = 0.84%), sensitive (capable of differentiating 1 sialic acid residue change among at least 29 sialic acids on a 103-kDa glycoprotein that is 38% carbohydrate), applicable to a wide range of loading (11.6-372 micrograms of FVIII), and accurate according to a comparison to matrix-assisted laser desorption-ionization time-of-flight mass spectrometry. Combining the method with enzymatic removal of N-glycans, selective O-glycan analysis was also performed leading to differential fragment ion analysis ascribed to N- and O-glycans. Quantitative ESI in-source dissociation MS combined with LC can generally be used for glycoproteins, as one of the indicators, to compare the distribution of carbohydrate residues over N- and O-glycans, to investigate their isoforms, and compare batch-to-batch characteristics of biopharmaceuticals.

Analysis of recombinant human interleukin-11 fusion protein derived from Escherichia coli lysate by combined size-exclusion and reversed-phase liquid chromatography.

A two-dimensional size-exclusion-reversed-phase high-performance liquid chromatographic assay has been developed for the quantitation of recombinant human interleukin-11 fusion protein (rhIL-11 FP) expressed in E. coli cells. The sample preparation procedure included the optimization of lysis buffer components to achieve maximum rhIL-11 FP recovery through the disruption of associations between rhIL-11 FP and E. coli components. The E. coli cells were dialyzed into lysis buffer and lysed by a French Press prior to two-dimensional chromatographic analysis. A size-exclusion column was used first to remove high- and low-molecular-mass E. coli components. Then reversed-phase chromatography was used to separate and quantify the rhIL-11 FP. The assay was linear over the range of 0.0294 to 0.235 mg/ml. The limit of quantitation, 0.0294 mg/ml, was based on % normalized residuals and precision criteria not exceeding 10%. The reproducibility of the assay for lysate samples was good on a daily (% R.S.D. = 1.0; n = 5) and a day-to-day reproducibility was good (% R.S.D. = 2.2; n = 9). Selectivity and chromatographic peak identification were based upon gel electrophoresis and N-terminal sequencing of the rhIL-11 FP peak collected from the reversed-phase column.

Flow-through particles for the high-performance liquid chromatographic separation of biomolecules: perfusion chromatography.

This paper reports a new technique for reducing resistance to stagnant mobile phase mass transfer without sacrificing high adsorbent capacity or necessitating extremely high pressure operation. The technique involves the flow of liquid through a porous chromatographic particle, and has thus been termed "perfusion chromatography". This is accomplished with 6000-8000 A pores which transect the particle. Data from electron microscopy, column efficiency, frontal analysis and theoretical modelling all suggest that mobile phase will flow through these large pores. In this manner, solutes enter the interior of the particles through a combination of convective and diffusional transport, with convection dominating for Peclet numbers greater than one. The implications of flow through particles on bandspreading, resolution and dynamic loading capacity are examined. It is shown that the rate of solute transport is strongly coupled to mobile phase velocity such that bandspreading, resolution of proteins and dynamic loading capacity are unaffected by increases in mobile phase velocity up to several thousand centimeters per hour. The surface area of this very large-pore diameter material is enhanced by using a network of smaller, 500-1500 A interconnecting pores between the throughpores. Scanning electron micrographs show that the pore network is continuous and that no point in the matrix is more than 5000-10,000 A from a through-pore. As a consequence, diffusional path lengths are minimized and the large porous particles take on the transport characteristics of much smaller particles but with a fraction of the pressure drop. Capacity and resolution studies show that these materials bind and separate an amount of protein equivalent to that of conventional high-performance liquid chromatography as well as low performance agarose-based media at greater than 10-100 times higher mobile phase velocity with no loss in resolution.

Thermodynamic model for electrostatic-interaction chromatography of proteins.

A thermodynamic model derived by Record et al. [M. T. Record, Jr., Biopolymers, 14 (1975) 2137 and M. T. Record, Jr., C. F. Anderson and T. M. Lohman, Q. Rev. Biophys., 11 (1978) 103] from Wyman's linkage theory [J. Wyman, Adv. Protein Chem., 19 (1964) 223] using Manning's condensation model [J. Manning, J. Chem. Phys., 51 (1969) 924] was extended to electrostatic interaction chromatography. Mixed, electrostatic and hydrophobic interactions of a model protein, ovalbumin were characterized by ion and water release.

Phase ratio determination in an ion-exchange column having pores partially accessible to proteins.

A method is suggested for determination of the hold-up volume and the phase ratio of a protein on a strong anion-exchange chromatographic column, which is based on mercury porosimetry and size-exclusion calibration with polymer samples.

Molecular orientation of immunoglobulin G at high concentration on an ion-exchange sorbent.

The change of molecular orientation of IgG, bound on a strong-anion-exchange surface, was studied using a generalization of the stoichiometric displacement model, over the entire range of protein adsorption isotherms. The Z number was found to decrease with increasing stationary phase protein concentration, approaching a limiting value. The analogy of the multiple equilibria model within highly cooperative identical binding sites was suggested as a possible way to evaluate the observed change in Z number with the protein concentration.

Effects of large sample loads on column lifetime in preparative-scale liquid chromatography.

In preparative-scale liquid chromatography of proteins, the use of high sample concentration and large sample mass may result in irreversible adsorption to the support surface. This can change the stationary phase characteristics, reduce the capacity, shorten the column lifetime and diminish the economic viability of a particular separation method. Column recycling and regeneration can influence the throughput (mass purified per time unit) and selectivity, and affect the reproducibility. The effects of large sample loads on column lifetime and performance were evaluated for three strong anion-exchange columns: (1) a silica support with a quaternized polyethyleneimine (PEI) coating, (2) a polymeric support with an adsorbed PEI coating which also was quaternized, and (3) a polymeric support with a proprietary quaternary amine stationary phase. The column capacity for proteins was measured by frontal chromatography and monitored as a function of cycle number. The column lifetime was determined by examining chromatographic properties subsequent to the frontal chromatography. The change in protein binding capacity was then compared to the change in nitrate binding capacity. The column performance was evaluated under analytical conditions by measuring the change in resolution of standard protein mixtures.