Ester hydrolysis of polysorbate 80 in mAb drug product: evidence in support of the hypothesized risk after the observation of visible particulate in mAb formulations.

An observation of visible particulate matter was made during formulation development of a mAb and investigations initiated to understand the colloidal instability of the formulation. It was observed that there was a loss of polysorbate in the IgG formulation and concurrent hydrolysis of polysorbate 80 (PS80) into fatty acid, polyethylene glycol (PEG), and pegylated sorbitan in the presence of mAb A. This observation was confirmed with two other mAb development programs (mAb B and mAb C) that used PS80 and formed particulate, but was absent in any placebo sample tested. Comparative analysis to acid, base and esterase hydrolysis, and exposure to the oxidation reagents Iron(II) and tert-Butyl hydroperoxide demonstrates that the observed reaction is reproduced by a biologic (enzymatic) mechanism. Monooleates of PS80, including the sorbitan and PEG oleates, are hydrolyzed first, showing a slower reaction with higher-order oleates. This leads to a change in the composition of the formulation over time where PS85 becomes the predominant component of the original surfactant remaining in solution. Data suggest that there is a lipid-specific mechanism rather than a general biologic hydrolysis mechanism that hydrolyzes oleate esters of PS80 increasing the risk that colloidal IgG particles will form in mAb drug product.

The oxidation of methionine-54 of epoetinum alfa does not affect molecular structure or stability, but does decrease biological activity.

Erythropoietin therapy is used to treat severe anemia in renal failure and chemotherapy patients. One of these therapies based on recombinant human erythropoietin is marketed under the trade name of EPREX and utilizes epoetinum alfa as the active pharmaceutical ingredient. The effect of oxidation of methionine-54 on the structure and stability of the erythropoietin molecule has not been directly tested. We have observed partial and full chemical oxidation of methionine-54 to methionine-54 sulfoxide, accomplished using tert-Butylhydroperoxide and hydrogen peroxide, respectively. A blue shift in the fluorescence center of spectral mass wavelength was observed as a linear response to the level of methionine sulfoxide in the epoetinum alfa molecule, presumably arising from a local change in the environment near tryptophan-51, as supported by potassium iodide quenching studies. Circular dichroism studies demonstrated no change in the folded structure of the molecule with methionine oxidation. The thermal unfolding profiles of partial and completely oxidized epoetinum alfa overlap, with a T(m) of 49.5 degrees C across all levels of methionine sulfoxide content. When the protein was tested for activity, a decrease in biological activity was observed, correlating with methionine sulfoxide levels. An allosteric effect between Met54, Trp51, and residues involved in receptor binding is proposed. These results indicate that methionine oxidation has no effect on the folded structure and global thermodynamic stability of the recombinant human erythropoietin molecule. Oxidation can affect potency, but only at levels significantly in excess of those seen in EPREX.

Recognition and identification of UV-absorbing leachables in EPREX pre-filled syringes: an unexpected occurrence at a formulation-component interface.

During the period of 1998 to 2002, there was an increase in the incidence of antibody-positive pure red cell aplasia (PRCA) in patients receiving subcutaneous administration of EPREX (epoetinum alfa). As part of the investigation of this event, the aqueous formulation containing polysorbate 80, introduced in 1998, facilitated the leaching of small-molecule, aromatic compounds from the uncoated rubber syringe stoppers. The leachables were identified using Liquid Chromatography-Mass Spectroscopy, Electrospray Ionisation-MS/MS, Dithiothreitol reduction, and Hydrogen/Deuterium exchange. The major leachable was identified as a dialkylphenol disulfide, and the majority of the remaining peaks were identified as structural variants containing different numbers of sulfur atoms in the sulfide bridge. In this report, we describe the strategies and experimental designs that were used to overcome the analytical challenges and that led to successful structural identification of the leachables in EPREX pre-filled syringes with uncoated syringe stoppers.

A comparative study of electrochemically and fluorometrically addressed molecular reporter groups: effects of protein microenvironment.

To probe the effects of protein microenvironment on electrochemically and fluorometrically addressed molecular reporter groups, genetically engineered apo-cytochrome c peroxidase derivatives W51C, A174C, K243C, and S246C, each containing a single cysteine residue, were labeled at identical sites with two kinds of microenvironment sensitive reporters, either an electrochemically active sulfhydryl-reactive reagent, [Ru(II)(NH(3))(4)(1,10-phenanthroline-5-maleimide)](PF(6))(2) [RuPA4] or a fluorescent 6-acryloyl-2-dimethylaminonaphthalene [acrylodan] probe. Two types of sites were labeled with each probe based on their predicted solvent accessibilities from the known structure for holo-cytochrome c peroxidase. One set of sites (K243C and S246C) was selected to be completely solvent exposed, while the other two sites (W51C and A174C) were less accessible, residing in or near the heme binding site. Spectroscopic properties of the fluorescent probe were consistent with predictions for relative solvent accessibilities; however, even the less solvent accessible probes reported a quite polar environment, suggesting that this region of the apo-protein is either substantially solvent exposed or undergoes significant dynamic motion. A linear correlation was observed between the lambda(max) of the metal to ligand charge-transfer (MLCT) absorption band of the RuPA4 complex and the acrylodan emission maximum for the four labeled apo-protein variants. The same trend occurred for the formal potential of RuPA4 versus the acrylodan emission maximum, with the exception of electrochemical probe behavior at position 174, possibly due to specific probe-protein interactions.