Kinetic Folding Mechanism of Erythropoietin

Abstract

This report describes what to our knowledge is the first kinetic folding studies of erythropoietin, a glycosylated four-helical bundle cytokine responsible for the regulation of red blood cell production. Kinetic responses for folding and unfolding reactions initiated by manual mixing were monitored by far-ultraviolet circular dichroism and fluorescence spectroscopy, and folding reactions initiated by stopped-flow mixing were monitored by fluorescence. The urea concentration dependence of the observed kinetics were best described by a three-state model with a transiently populated intermediate species that is on-pathway and obligatory. This folding scheme was further supported by the excellent agreement between the free energy of unfolding and m-value calculated from the microscopic rate constants derived from this model and these parameters determined from separate equilibrium unfolding experiments. Compared to the kinetics of other members of the four-helical bundle cytokine family, erythropoietin folding and unfolding reactions were slower and less susceptible to aggregation. We tentatively attribute these slower rates and protection from association events to the large amount of carbohydrate attached to erythropoietin at four sites.

The Effect of Glycosylation on the Folding Kinetics of Erythropoietin

“…(3)] under strong refolding conditions (b 1 M urea) as shown before. 24 The fast phase, which was previously attributed to the unfolding of an off-pathway burst-phase (BP) species, carried less than 5% of the total folding amplitude and increased in SF-FL intensity, while the intermediate and slow kinetic phases carried the bulk of the folding signal and decreased in SF-FL intensity (Fig. 3a).…”

“…This made it difficult to determine the magnitude and stability of the off-pathway BP intermediate due to the steepness and scatter of the unfolded baseline region of this equilibrium curve. 24 The faster unfolding kinetics under the low-ionicstrength conditions of the present study made BP analysis more tractable, as SF folding closer to the transition region became possible. A considerable discrepancy was observed between the expected WT-EPO unfolded signal at low denaturant concentrations and the signal calculated at the time of mixing, consistent with the presence of a hyperfluorescent BP populated within the dead time of the SF instrument (Fig.…”

“…A similar approach was used in the previous WT-EPO folding study where a BP was detected and considered to be an offpathway species due to the positive m ‡ value of the subsequently observed fast folding phase, and the ability to derive the equilibrium unfolding parameters using only the observed intermediate and slow kinetic phases fit to an on-pathway model. 24 The stabilizing solvent conditions and slow folding kinetics of the previous study, however, necessitated that the final SF-FL folding signals be normalized to an equilibrium unfolding curve. This made it difficult to determine the magnitude and stability of the off-pathway BP intermediate due to the steepness and scatter of the unfolded baseline region of this equilibrium curve.…”

“…Biophysical and functional characterization of the nonglycosylated Escherichia coli-expressed EPO have already demonstrated that it is well folded, has the same in vitro biological activity as the Chinese hamster ovary cell-derived glycosylated EPO (WT-EPO), but is thermodynamically less stable and tends to precipitate after shortterm storage at 20°C. [21][22][23] Finally, kinetic folding studies of the glycosylated WT-EPO have been performed in near-physiological conditions, and the urea concentration dependence of the observed rate constants was shown to be consistent with a threestate on-pathway model, 24 making it plausible that the effects of glycosylation on the intermediate species, as well as the unfolded and native states, may be elucidated by comparison with the nonglycosylated EPO.…”

“…21 Despite these mutations, a direct comparison to the WT-EPO folding data collected previously could not be made due to visible precipitation following dialysis of MKLys-EPO into the same solution conditions used in the previous study: 140 mM sodium chloride and 20 mM sodium phosphate (pH 6.9). 24 Equilibrium and kinetic data for both WT-and MKLys-EPOs were therefore collected under the same low-ionic-strength solvent conditions that maintained the solubility of the nonglycosylated MKLys-EPO mutant: 10 mM sodium phosphate (pH 6.9), a strategy similar to the one employed for the NMR structure determination. 25 Under these same solution conditions, glycosylation did not alter the folding mechanism of EPO but did slow the folding process and increase the stability of the onpathway intermediate species to a greater extent than it increased the equilibrium stability, suggesting that EPO glycosylation has evolved to optimize biological function rather than folding efficiency.…”