In the biopharmaceutical industry, thorough and reliable protein characterization methods are of critical importance. This is especially necessary because even the most minor structural change can affect the safety, quality, efficacy or manufacture of a biotherapeutic. Thus, regulatory authorities like the FDA and EMA have started placing an ever-increasing importance on objective, statistically-validated data. Though they are aware of such requirements, developers of biotherapeutics are faced with challenges, such as methodological limitations or sample availability, that prevent them from easily collecting such data.
In light of this, the FDA has retracted its draft guidance on statistical approaches to evaluate analytical similarity, a mere nine months after its release in September 2017. This action was taken as a result of industry concerns about the high number of reference product lots to be sourced and their lot-to-lot variability.
Nevertheless, though reference availability will remain a matter of concern, we can overcome the technical limitations. This is particularly true in the case of higher order structure (HOS) comparisons, which can now be carried out via CD spectroscopy. This type of analysis gathers information about changes in HOS by conducting pairwise comparisons of spectra obtained under varying conditions.
HOS changes must be detected and followed throughout the development of biotherapeutics in order to confirm that a drug is unaffected by a minor change of the formulation buffer, or to demonstrate the similarity of multiple production lots. In biosimilar development, early-stage identification of structural differences between biosimilars and originators minimizes residual uncertainties that would need to be resolved in downstream processes. Thus, the quick identification of differences is as crucial as confirming the similarity between biosimilars and originators at later stages, where such data and results can strengthen the argument for regulatory submission.
Although CD spectroscopy has the capacity to provide information about both secondary and tertiary structure, its usefulness for HOS comparisons has, thus far, been limited because comparisons were based on the subjective, visual evaluation of CD spectra. Moreover, HOS comparisons were neglected in the past – not because such data was considered useless, but because conventional CD spectroscopy could not ensure a quantitative analysis.
However, modern CD spectrometers such as the Chirascan Q100 can now provide objective HOS comparisons that are statistically-validatable. Thus, CD spectroscopy today is recognized as a crucial tool in the biophysical characterization process, where HOS comparisons are necessary to define the critical quality attributes of biotherapeutics.
CD instruments that make use of solid-state detectors are highly sensitive and can thus detect even minor changes – not just in the far-UV’s secondary structure, but also in the tertiary structure of the near‑UV, where a CD signal is typically lower by orders of magnitude.
Moreover, using flow cells in combination with automated liquid handling today eliminates the errors associated with manually handling samples and cuvettes. In addition to increased reproducibility, CD analyses can be conducted more accurately and precisely, reducing the number of replicates required for robust statistical analysis.
An added benefit includes the removal of the user-bias risks associated with visual comparisons, and the statistical significance of spectral differences can be confirmed objectively. CD spectroscopy is being elevated by these and other innovative developments from Tier 3 (raw data/graphical comparison) to Tier 2 (quality range method) for HOS comparison, as referred to in the tiered approach developed by the Office of Biostatistics and Office of Biotechnology Products, CDER/FDA.
Case Study: Forced Degradation of a Monoclonal Antibody
To demonstrate the detection and statistical analysis of minor changes in HOS, a monoclonal antibody – IgG1 – was subjected to various stress conditions as part of a forced degradation study. A number of sample pre-treatments were anticipated to induce oxidation, asparagine (Asn) deamidation/aspartate (Asp) isomerization or glycation of the antibody. Stress conditions were then stopped by dialysis against a common buffer (phosphate buffered saline pH 7.4). Each stress condition required the measurement of CD spectra for five independent replicates. For each replicate, the acquisition, averaging, and buffer-correction of three repeat scans was also implemented.
The simultaneous acquisition of absorbance along with the CD enabled normalization (by absorbance at 280 nm). This is a vital step in data analysis, as it excludes any apparent differences between replicate spectra as a direct result of minor differences in concentration rather than structural variations.
All samples’ near-UV spectra have a similar fingerprint profile, contributed by aromatic amino acids and disulfide bonds (Figure 1). A difference between the oxidation and reference spectrums - while fairly obvious upon visual inspection - is less easy to identify in alternative stress conditions. Particularly, the glycation spectrum appears to be as similar to the reference as the control spectrum is. Nevertheless, small alterations in tertiary structure can be caused by changes in the dihedral angles of disulfide bonds that are so subtle that they elude visual inspection.
Figure 2: HOS comparison: forced degradation of IgG1 samples assessed by near-UV CD. The monoclonal antibody was subjected to the stress conditions as indicated; near-UV spectra shown are normalized by absorbance at 280 nm. Data was recorded with a Chirascan Q100.
An objective, unbiased evaluation requires rigorous statistical analysis. The pharmaceutical industry currently prefers using the Weighted Spectral Difference (WSD) method  for a comparative analysis to assess the similarity between two spectra. This form of analysis is extremely transparent, and is increasingly being adopted by biopharmaceutical companies as well as regulatory authorities.
In short, a difference spectrum is calculated from the two spectra under comparison, following which a wavelength-dependent weighting is applied based on the reference spectrum (Box 4). Such a weighting lowers the relative contribution of low-signal regions, where spectral variation arise mainly as a result of noise. Finally, the spectral results are assigned a numerical value – WSD – to depict similarity. A smaller WSD means a greater similarity between the compared spectra.
To conduct statistical analysis, firstly the WSD values are computed in order to compare reference replicates among one another and assess the variability of the reference spectra. For example, in Figure 1, the resulting five WSD values (one for each reference replicate compared against the other reference replicates) are shown in black along with the mean of these WSD values and a range, which is defined as twice the standard deviation. The range functions as an acceptance criterion, as recommended by the FDA for Tier 2 critical quality attributes, thus allowing the data to be subjected to a quality range test.
For the purpose of this study, WSD values were calculated to compare individual sample replicates and the reference. Figure 1 shows that all WSD values for the control are within the acceptance range, while values for other stress conditions are not. This reveals that the glycation conditions, which resulted in spectra highly similar to the reference (and virtually indistinguishable from the control), led to a statistically significant change in tertiary structure. In addition, by plotting the WSD values, the detected, significant differences that resulted from the various stress conditions differ in their magnitude.
To summarize, it was found that minor differences between CD spectra were statistically significant, thereby reflecting changes in tertiary structure of a monoclonal antibody after being subjected to different stress conditions. To conclude, this forced degradation study illustrates how minor differences can be revealed by statistical analysis; differences that would have otherwise be overlooked during simple visual inspection.
Box 4: Weighted Spectral Difference
Comparison between two spectra by the Weighted Spectral Difference (WSD) technique yields a single value (a WSD), which measures similarity according to the following equation:
Obtaining the WSD to compare a reference spectrum (yAi) with a sample spectrum (yBi), first the sample spectrum is subtracted from the reference spectrum (1). The resulting difference spectrum (blue) is then squared, to ensure that negative values cannot cancel out positive values (2).
Then, a weighting based on the reference (gray) is applied to this squared difference. Firstly, absolute values of the reference are considered, also to account for negative signals (3). The absolute reference is then normalized (4) by dividing through its average (red) and multiplied by the squared difference spectrum (5) to obtain a weighted difference spectrum (orange, solid line).
Finally, the WSD is computed by calculating the average intensity of the weighted difference spectrum (orange, dashed line), by summing over all datapoints (i) and dividing by their total number (n), and taking its square root.
 N. N. Dinh, B. C. Winn, K. K. Arthur, and J. P. Gabrielson, “Quantitative spectral comparison by weighted spectral difference for protein higher order structure confirmation,” Anal. Biochem., vol. 464, no. July, pp. 60–62, Nov. 2014.
About Applied Photophysics
Applied Photophysics is a leading provider of systems and accessories for the biophysical characterization of biomolecules. Headquartered in Leatherhead, Surrey, UK, the Company has been established for more than 40 years.
The SX-range of stopped-flow spectrometers, used to monitor changes in absorbance and fluorescence during fast biological reactions, is acknowledged globally as the gold standard for kinetic studies. In 2005, the Company introduced the first Chirascan™ system, using the phenomenon of circular dichroism (CD) to characterize changes in the higher order structure of proteins.
Since then, the company has continued to incorporate its in-depth knowledge and understanding of CD into a range of Chirascan products that are used in cutting-edge research and to support the development of innovator drugs and biosimilars in the biopharmaceutical industry. Compared to conventional CD instruments, the new generation of Chirascan systems ensures that every scientist gets the most from every CD analysis.
Sponsored Content Policy: News-Medical.net publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of News-Medical.Net which is to educate and inform site visitors interested in medical research, science, medical devices and treatments.