(M0930-06-31) Identifying Protein Aggregation through UV-Visible Absorption Spectroscopy

Purpose: Protein aggregation commonly occurs in samples containing unstable proteins.^1,2 These aggregates form irreversibly and often lead to a turbid solution. For pharmaceutical applications, including protein therapeutics, the presence of aggregated species has been linked to a lowered immune response, and may cause allergic reactions in patients.³ Consequently, it is important the presence of aggregates be identified early in the formulation process. Herein, we demonstrate the ability to identify protein aggregation through UV-Visible absorption spectroscopy. Through this analysis, we were able to analyze samples with both minimal and extensive aggregation, as well as determine the amount of protein content which is not aggregated in solution.

Methods: Bovine gamma globulin (BGG) was chosen as a model protein for this analysis as it is robust, allowing for a control sample to be made which contains no protein aggregates. A 0.55 mg/mL control sample was made by diluting a 2 mg/mL solution of BGG with phosphate buffer (1X pbs). Protein aggregation was induced through two different methods: increasing the ionic strength of the solution and heating. For the first method, a 0.55 mg/mL BGG sample containing NaCl (2.6 M) was made. A separate 0.55 mg/mL BGG sample was heated to 75 ˚C for the second aggregation method. Following sample preparation, the absorption spectrum of all samples was measured using a UV-Visible absorption spectrophotometer in a 1.0 cm quartz cuvette. For the BGG sample heated to 75 ̊C, as well as the control BGG sample, a spectrum was collected using an integrating sphere accessory in transmission geometry as well. Following analysis, the BGG sample aggregated through heating was then filtered using a PTFE syringe filter and measured using the UV-Visible spectrophotometer without the integrating sphere.

Results: The resulting absorption spectrum collected for the control BGG sample agrees with literature,⁴ however both samples in which protein aggregation was induced exhibit an increased baseline at wavelengths greater than 310 nm, then onset for absorption from BGG. As this protein does not absorb in this region, this feature is attributed to light scatter from the aggregates in solution. Notably, the solution of BGG which included NaCl does not appear to be turbid by eye, however the use of UV-Visible absorption spectroscopy allows the presence of these aggregates to become apparent. The scattering artifact present in the spectrum is not expected to be the same across the entire UV-Visible spectrum, requiring careful correction. By modifying Beer’s law to account for the wavelength dependence of scatter (according to the Rayleigh Equation),⁵ an estimate for the scattering contribution in the absorption spectrum can be calculated and used to determine spectrum of the non-aggregated proteins in solution. Using the corrected spectrum, the concentration of non-aggregated BGG was found to be 0.54 mg/mL. The samples which were heated to 75 ˚C are visibly cloudy. This degree of scatter cannot be corrected for using the estimate described previously, however the scatter can be corrected by using an integrating sphere to collect the scattered and transmitted light. The integrating sphere allows for collection of the scattered light by the detector, however the design of the integrating sphere used does not allow for the collection of all transmitted light, preventing collection of a true absorption spectrum. By measuring the intensity of the collected light and using the Kubelka-Munk (KM) transformation, a correlation between the KM signal and the true concentration can be made. As this collected signal, once transformed, is proportional to the absorption coefficient, it is also proportional to the concentration of the material according to Beer’s law. Through measuring the both the control BGG as well as the heated BGG sample, a system of equations can be set up which correlates the concentration of BGG and measured KM signal of the control sample with the concentration and measured KM signal of the non-aggregated species in the heated BGG sample. Through this method, the concentration of non-aggregated BGG was found to be 0.20 mg/mL for the heated sample. To confirm that the concentration determined through the integrating sphere method was accurate, the heated BGG sample was filtered and analyzed using the UV-Visible spectrum without the integrating sphere. Using Beer’s law and the collected spectrum, the concentration of non-aggregated BGG was confirmed to be 0.20 mg/mL

Conclusion: The ability to use UV-Visible absorption spectroscopy to identify the presence of protein aggregates in a sample with both a low and high degree of aggregation has been demonstrated through the experiments described herein. Through this method, the concentration of non-aggregated BGG was determined to be 0.54 mg/mL for samples aggregated through increasing ionic strength, and 0.20 mg/mL for samples aggregated through heat. For highly scattering samples, the integrating sphere was effective in removing scattering contributions in the measured spectrum.

References: (1) Chen, Z.; Huang, C.; Chennamsetty, N; Xu, X.; Li, Z.J., J. Chromatogr. A, 2016, 1460, 110 – 122.
(2) Jubete, Y.; Maurizi, M.R., Gottesman, S., J. Biol. Chem., 1996, 271, 48, 30798-30803.
(3) Lundahl, M.L.E.; Fogli, S.; Colavita, P.E.; Scanlan, E.M., RSC Chem. Biol., 2021, 2, 1004-1020.
(4) Smith, E.L.; Coy, N., J. Biol. Chem., 1946, 164, 1, 367 – 370.
(5) Yao, G.; Li, K.A.; Tong, S.Y., Anal. Chim. Acta, 1999, 398, 319 – 327.


Figure 1 - (a) UV-Visible absorption spectra of 0.55 mg/mL BGG collected using a 1 cm quartz cuvette (blue curve = control, no heat or NaCl, red curve = BGG with 2.65 M NaCl, yellow curve = BGG heated to 75 C). (b) BGG control solution. (c) BGG with NaCl solution. (d) BGG after heating to 75 C.


Figure 2 - (a) Kubelka-Munk (green) and absorption (brown) spectra of 0.55 mg/mL BGG following heating to 75 C. Samples were measured using a 1 cm quartz cuvette, with and without the use of an integrating sphere, respectively. (B) Absorption spectra of the BGG control (orange) and heated BGG after filtration (black).