Variable fragment single-chain antibodies (scFv) are aggregation prone proteins with important medical applications in tumor imaging and targeted drug delivery. Although complex expression systems have been developed that can provide soluble and functional scFv, the yield and concentration obtained from these systems are undesirable. Bacterial expression funnels large amounts of scFv into inclusion bodies, preventing the scFv from folding into an active form. Methods to recover functional scFv from inclusion bodies suffer drawbacks such as aggregate formation and require the use of large quantities of denaturants such as guanidine hydrochloride.
The structural similarities of scFv with proteins implicated in aggregation-driven human diseases and the need for a highly efficient, fast, inexpensive, and aggregate-free recovery method for scFv from bacterial systems warrant research into the aggregation behavior of these proteins. The inability of current chemical techniques to sufficiently hinder aggregation has focused our attention on physical treatments that show promise at reversing aggregation. Previous work in our laboratory has shown that aggregates of the homotrimeric tailspike protein break apart and subsequently regain activity following exposure to high pressure. High pressure treatments, in conjunction with current chemical methods, may provide a tractable solution to the aggregation problem in scFv production.
The proposed aggregation study consists of four primary research aims: 1) Investigate the effect of high pressure on scFv aggregates; 2) Characterize the scFv folding pathway to identify aggregation prone conformations; 3) Develop a scalable folding protocol for scFv from Escherichia coli inclusion bodies that minimizes aggregate production; 4) Develop a method to convert inactive scFv aggregates into active scFv using a combination of high pressure and chemical techniques.
The well characterized anti-fluorescein scFv 4-4-20 and 4M5.3, a directed evolution mutant of 4-4-20 selected for its increased affinity for fluorescein, are model proteins for this study because they display a high propensity for aggregation and activity can be verified by a simple fluorescence assay. Our lab has successfully recovered active scFv from E. coli inclusion bodies using several different folding strategies that result in different active/aggregate distributions. Size exclusion HPLC can separate the active form of scFv 4-4-20 from aggregate. In the case of 4M5.3, this technique is able to resolve two active monomeric conformations and an inactive monomeric conformation from aggregate (Figure 1). Nondenaturing polyacrylamide gel electrophoresis (NPAGE) can also be used to separate monomeric from aggregated species and an intra-gel activity assay can identify active components.
A primary goal of this study is to characterize activity and conformational changes in the monomeric and aggregated scFv species upon the application of high pressure. Our laboratory is equipped with a high pressure cell capable of reaching 50000 psig and a fluorescence pressure cell rated to 35000 psig. Preliminary high pressure experiments with scFv samples containing both active and aggregate forms to 40000 psig for 1 hour showed a notable decrease in the amount of aggregate. High resolution size exclusion HPLC analysis verified the concomitant appearance of a conformation that appears to be between that of the active and inactive monomer state for 4M5.3.
To fully understand the effect of high pressure on the aggregated scFv system, we need to study the effects of pressure on each component independently. Therefore, studies are planned to vary the duration and magnitude of pressure treatment on not only the mixed aggregate/monomer samples, but also on purified samples of aggregate and monomer. The effect of pressure cycling from very high pressure to atmospheric or lower pressure on aggregate dissolution will also be examined in terms of the resultant species and fraction of pressure-resistant aggregate after each cycle. Size exclusion HPLC and the solution activity assay will be used as complementary techniques to quantify the distribution of scFv species following pressure treatment.
Intrinsic fluorescence measurements at atmospheric pressure and size exclusion chromatograms will be used to draw comparisons between monomeric states obtained by pressure treatment and those obtained as a direct consequence of the inclusion body folding strategy. Using the fluorescence pressure cell, the effect of moderate pressures on active and inactive monomer conformations will be determined by monitoring the center of mass shift in intrinsic fluorescence as pressure increases.
Characterization of the conformation and species distribution of the pressure-treated, aggregated scFv samples will provide new insight into scFv aggregate composition. These experiments will demonstrate the effectiveness of a physical, rather than chemical, solution to an industrially relevant aggregation problem.
The initial stage in recovering active scFv from inclusion bodies involves solubilization using harsh chemical denaturants such as 6M guanidine hydrochloride or 8M urea that destroy most of the protein secondary structure. This is a traditional approach to refolding inclusion bodies, but results in low yields of active scFv. I have developed an improved folding procedure for solubilized scFv inclusion bodies that is faster and less expensive than traditional folding methods10. Although the distribution between active and aggregate states is improved with this protocol, significant aggregation is still observed. If a folding strategy is to be optimized for maximum yield of active product, the aggregation propensity of the folding intermediates must be quantified.
We propose using high resolution size exclusion chromatography to determine the denaturant concentrations at which significant conformational changes occur in the folding scFv chain as well as the aggregation propensity of each conformation. A stepwise dialysis folding procedure was used to incrementally reduce denaturant concentration from 6M guanidine HCl to 0M guanidine HCl11. Using this approach, a sample was removed after each dialysis step and run on a size exclusion column equilibrated to the sample conditions. The sample was also run on a size exclusion column equilibrated to 0M guanidine HCl to examine the stability of the intermediates. Experimental data has been collected for 4M5.3 and 4-4-20 folding samples at guandine hydrochloride concentrations of 6M, 3M, 2M, 1M, 0.5M and 0M.
Comparison of the chromatograms at 0M guanidine HCl with those at the sample guanidine HCl concentrations reveal that the scFv is aggregating as it moves through the chromatography column. At low guanidine HCl concentrations active conformations are detected in the 0M guanidine HCl chromatograms but not in the traces obtained at the guanidine HCl concentration of the sample, indicating that the final folding steps to the active conformation occurred during the chromatography run at 0M (Figure 2). As the guanidine concentration is reduced, a gradual shift in the primary monomer peak to a smaller volume conformation is observed in the 0M guanidine HCl runs.
A very rich data set has been collected on the folding pathway of these scFv. This data needs to be mined with modeling techniques to identify key intermediate states responsible for aggregate formation. We propose developing a model that incorporates a cluster-cluster aggregation reaction and first order active species formation for scFv folding intermediates traveling through the chromatography column. The model must explicitly account for inactive monomer, active monomer, dimer, and large aggregates. Aggregates that are trimeric or larger cannot be individually resolved. This work will give us our first glimpse at the kinetics of aggregate formation during the folding process. Using the kinetic information from the model and the mobility shift of the monomeric peak, we can identify the denaturant concentrations at which key conformational shifts occur. We can also identify critical denaturant concentrations where the aggregation propensity of the intermediate is stronger than the destabilizing force of the guanidine solution, which allow the aggregates to persist in solution. This information will allow us to construct a folding pathway for scFv recovered from inclusion bodies.
Integration of the high pressure experimental results and the folding pathway developed from the aggregation kinetics will allow the development of an improved folding strategy for scFv that minimizes aggregate formation, and reduces cost. Application of the most effective pressure treatment techniques for disrupting aggregates to folding intermediates will provide a way to dramatically increase active yields and provide insight into the effects of changes in pressure on protein conformation. The monomer mobility shift observed in the folding intermediate chromatograms suggests that pressure may be able assist the conversion of inactive monomeric forms into active conformations at low guanidine concentrations in addition to aggregate disruption. The stage at which folding and redox additives are added will also need to be further explored. While the optimized folding protocol for the anti-fluorescein scFv may not be directly applicable to all scFv, a framework for the optimization process has been created.
We also propose investigating the use of high pressure treatments in combination with intermediate guanidine hydrochloride concentrations to solubilize the initial inclusion body rather than 6M guanidine hydrochloride. The character of the inclusion body before solubilization is largely unknown. Solubilization with strong denaturants may be unnecessarily destroying certain elements of the scFv secondary structure. It should be possible to obtain increased active yields by preserving some of the secondary structure of the original inclusion body.
A final goal of this work is to recover activity from samples with aggregate and active monomer conformations. As the aggregates are at least partially comprised of folding intermediates, addition of low levels of denaturant and/or folding additives prior to pressure treatment may return the sample to monomeric conformations rather than the completely unfolded states obtained through chemical denaturation. Removal of the remaining denaturant can be done following the optimized folding protocol beginning at the appropriate denaturant concentration. We propose a series of experiments working with purified aggregates as well as aggregate/active mixtures to optimize the procedure for activity recovery from aggregated scFv samples.
This material is based upon work supported by NIH P20 RR15588-01. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of NIH.