PEG's interaction with the protein surface has been observed in several instances. Speculation about the behavior of an unstructured
random coil PEG when encountering a surface can be based on PEG's flexibility14 and previous work with macromolecules such as DNA,19 where for steric reasons, random coils orient preferentially parallel to a solid surface. Supporting evidence for a resulting
shell-like conformation also comes from a study on chemically modified cytochrome c with PEG, where the overall hydrophobicity and thermostability of the protein is increased upon coupling.35 These data agrees with the previous observation that PEG modification enhances protein stability by decreasing electrostatic
repulsion between surface charges.5 In addition, PEG's affinity for tryptophan was established and used to explain protein partitioning between phases.36 Work with PEG and Bovine serum albumin suggests the existence of an attractive interaction between protein and polymer.37 Another study found that covalently-bound PEG to protein reduces the interaction with free PEG.38 These observations strongly support non-specific interactions between PEG and protein, interactions that ultimately affect
the overall properties of the molecule.
We propose that upon changing the solution conditions the monocovalently bound polymer chain can be driven from the helical
or random worm-like conformation when in contact with the solvent, to contact the protein surface in a compact, spherical
conformation (Figure 2). The resulting shell-like structure of the monoPEGylated species could inhibit a further PEGylation
reaction. The model of such a transition would include typical random fluctuations of unstructured polymer loop until it finds
either a surface or its other end. In the spherical conformation, some degree of loop formation in surrounding water is acceptable.
When optimizing a process for the manufacturing of a protein drug, there are three main considerations: high product quality
(purity, stability, and activity), process robustness, and low cost.33 Purifying and characterizing a particular positional isomer are recognized challenges associated with manufacturing of a
PEGylated biomolecule. To lower analytical and downstream processing hurdles, it is preferred that the selectivity of the
PEGylation reaction be optimized for a certain PEGylation degree and site. For a large-scale production of a PEGylated protein
drug, a higher yield of the target product coupled with a lower yield of secondary PEGylation products is extremely advantageous.
However, because the progression of the PEGylation reaction depends on several variables, the stoichiometry and the attachment
site are hard to control. Ideally, the reaction parameters can be fine-tuned to achieve the desired stoichiometry of PEG conjugates,
to produce predominantly mono-, di-, or other target PEG conjugates.39
Various methods have been previously used to force the reaction toward a unique, site-specific, active product that has the
suitable degree of PEGylation. Such approaches include the use of protective agents, various types (branched or linear) and
sizes of the PEG, mutations of reactive residues, or taking advantage of a specific amino acid reactivity under various solution
pH conditions. 13,31,35,36,40-43 In most cases, a mixture of PEGylated isomers is obtained, and the most active isomer is then selectively separated by chromatographic
methods.44,45 A combined control of the PEGylation reaction with simultaneous size-exclusion separation of the products has been described.46 The scaling-up of size exclusion chromatography, however, is a significant challenge in the biopharmaceutical industry because
of its low throughput and high cost. In most of the cited cases, not only has the PEGylation yield been low, but additional
downstream purification of the desired product also has been needed. Besides adding to the production cost, further separation
of the PEGylated mixture is extremely difficult because the mixture is complex and the physicochemical characteristics of
the species are very similar. Although ion exchange chromatography is routinely used, the dominant PEG properties affect the
capacity and resolution of the method. These challenges in developing PEGylated protein drugs still have not been resolved.