Freezing-induced perturbation of tertiary structure of a monoclonal antibody

Abstract

We studied the effects of pH and solution additives on freezing-induced perturbations in the tertiary structure of a monoclonal antibody (mAb) by intrinsic tryptophan fluorescence spectroscopy. In general, freezing caused perturbations in the tertiary structure of the mAb, which were reversible or irreversible depending on the pH or excipients present in the formulation. Protein aggregation occurred in freeze-thawed samples in which perturbations of the tertiary structure were observed, but the levels of protein aggregates formed were not proportional to the degree of structural perturbation. Protein aggregation also occurred in freeze-thawed samples without obvious structural perturbations, most likely because of freeze concentration of protein and salts, and thus reduced protein colloidal stability. Therefore, freezing-induced protein aggregation may or may not first involve the perturbation of its native structure, followed by the assembly processes to form aggregates. Depending on the solution conditions, either step can be rate limiting. Finally, this study demonstrates the potential of fluorescence spectroscopy as a valuable tool for screening therapeutic protein formulations subjected to freeze-thaw stress.

Keywords: excipients; formulation; liquid chromatography; monoclonal antibody; protein aggregation; proteins; stability; surfactants.

Figure 1

Representative intrinsic (Trp) fluorescence spectra of 10 mM potassium phosphate buffer, 0.5 mg/mL RNase A, and 0.5 mg/mL mAb at pH 3 before and after freezing. The excitation wavelength is 295 nm. All spectra are original signals without correction.

Figure 2

The wavelength of Trp fluorescence emission maxima (λ_max) for all samples at pH 3. Data represent mean ± standard deviation of triplicate samples. Prior to the determination of λ_max, each spectrum was corrected by subtracting the signal collected from its blank solution at the same temperature.

Figure 3

The wavelength of Trp fluorescence emission maxima (λ_max) for all samples at pH 4. Data represent mean ± standard deviation of triplicate samples. Prior to the determination of λ_max, each spectrum was corrected by subtracting the signal collected from its blank solution at the same temperature.

Figure 4

The wavelength of Trp fluorescence emission maxima (λ_max) for all samples at pH 8. Data represent mean ± standard deviation of triplicate samples. Prior to the determination of λ_max, each spectrum was corrected by subtracting the signal collected from its blank solution at the same temperature.

Figure 5

Representative size-exclusion chromatographs of mAb with or without additives at pH 4 after freeze–thawing, except control sample was the sample without additive and not subjected to freeze–thawing stress.

Figure 6

The effects of additives on freeze–thawing-induced aggregation of mAb by SE-HPLC. Data represent mean ± standard deviation of triplicate samples. HMW%: percentage of dimer and high molecular weight species. The average total peak area for protein monomer of three replicate samples—without additives and that were not freeze–thawed—served as a control value. The AUC for the monomer peak divided by the average total AUC of the control samples (×100) was taken as the percentage of monomer, and AUC for the peak representing aggregates divided by the average total AUC of the control samples (×100) was taken as percentage of aggregates. mAb +4 M Gdn HCl samples were not tested as per the reason explained in the text.

Figure 7

Representative intrinsic (Trp) fluorescence spectra of 0.5 mg/mL mAb (pH 3) with no additive, 150 mM KCl, 1 M sucrose, 45 M Gdn HCl, 4 M Gdn HCl, and 0.05% PS80 at −30°C. The excitation wavelength is 295 nm. Each spectrum was corrected by subtracting the signal collected from its relative blank solution at the same temperature.