Scatter Correction for UV Absorbance Assays: A Cautionary Comment
It seems to have become industry practice to apply a “scatter correction” to absorbance measurements (A280) for protein concentration. Typically, these involve either subtracting the absorbance from a single, nearby wavelength (e.g., 320 nm) or by linear extrapolation using two or more wavelengths (e.g., 320 nm and 360 nm). The premise is that any apparent absorbance at these wavelengths represents light scattering due to particulate load and measurements of protein concentration at 280 nm will be improved by subtracting out the contribution from this scatter.
Let’s examine the assumption that absorbance at wavelengths greater than 280 nm is due to scatter. This assumption ignores absorbance due to advanced glycation end products (AGEs or Maillard reaction end products), or leached plasticizers. AGEs are at least as likely a source of UV absorbance as is scattering, and I have seen instances where leached plasticizers (e.g., bisphenol A) have resulted in significant absorbance in the range of 300 nm to 400 nm. As a result, applying “scatter corrections” in situations where AGEs or leachables are the cause of the absorbance results in underestimation of the protein concentration.
Moreover, the AGEs may continue to develop over time, potentially introducing further errors. In one stability study, this manifested as both loss of protein and increasing specific activity over time! The protein concentration was not changing, nor was the activity; but blindly applying the scatter correction resulted in inaccurate stability trending.
Applying any type of “correction” to a measurement requires that the root cause for the correction be justified. If you suspect your sample has significant scatter due to particulates, the correct course of action should be to centrifuge or filter the sample and then compare the UV absorbance profile before and after the sample treatment. If the absorbance persists, you likely don’t have a particulate problem. If you do have a particulate problem, simply subtracting an estimate of the contribution to absorbance should not be viewed as a long-term fix. You have a sample stability or formulation problem, which should not be ignored by data “correction” tricks.
In summary, good methods are based on understanding the measurement itself and the sample constraints. Taking shortcuts by applying “corrections” without confirming why there is a need for such corrections will invariably lead you down the wrong path and likely cause more problems in the long run.
Blog article by: Joe Siemiatkoski