In this instalment of Robert MacNeill’s column, he explores the fascinating manifestations of the phenomenon that we see in bioanalytical LC–MS calibration curves.
Robert MacNeill received his Bachelor’s degree with Honors in Chemistry from Heriot Watt University then his MSc in Analytical Chemistry from the University of Huddersfield, both in the United Kingdom. Robert is also a Chartered Chemist and a Fellow of the Royal Society of Chemistry. With 22 years of experience in all aspects of quantitative bioanalytical LC–MS/MS method development, 13 of these years heading method development activities within the Princeton site that has housed HLS/Envigo (now Covance), and a regular author and peer reviewer for the journal Bioanalysis, Robert is a recognized expert and innovator in the field.
During a time in which we’re all discussing how best to ‘flatten the curve’, in this article I look at the same kind of push but within a quite distinct context. The fascinating manifestations of the phenomenon that we see in bioanalytical LC–MS calibration curves, naturally.
In regulated bioanalysis, we love straight lines. There’s no getting away from it. Then simplest and most reliable concentration–response model. A given method constitutes a multi-parametric operation and prior to validation we will adjust a curvilinear-producing method as appropriate in order to produce a straight line. On the very odd occasion, we may be able to justify using a curved fit as a feature of an established method, but the resultant quadratic regression parameters would need to be demonstrably reproducible. This is a very infrequent situation, and to all intents and purposes we only ever use this in non-GLP bioanalysis, but even then we tend to be uncomfortable. Curves often have a horrid tendency to change characteristics, even become linear, when no deliberate change to the system has been input. Also, it is desirable to never be approaching a plateau where a given response does not correspond to one singular concentration but rather an asymptotic range, or similarly where a given concentration does not correspond to one singular response, again the asymptote being visualized. Moreover, we know we can use our knowledge and chemistry toolbox to make adjustments in order to attain a straight line.
In short, a non-linear response may be seen as an indication that we can take measures to bolster the reliability of the method.
The seasoned bioanalyst will be well-used to the generation of non-linear calibration curves, and by far the most common is what is typically referred to as a ‘saturation-type’ curve. This is where the slope begins to decrease with increasing concentration and most often it represents the onset of detector saturation, hence where the term originates. At any given point in a chromatogram, the intensity in an ion channel cannot be more than a certain threshold and signal saturation occurs when pushed towards this threshold. Solutions lie in reducing injection volumes, increasing extract dilutions, or even broadening chromatographic peaks if it doesn’t compromise resolution or performance. Reducing the linear dynamic range of the assay is also an alternative, although clearly less desirable as we don’t relish the prospect of increasing our chances of having to perform repeat analyses with dilution. Moreover, I would confidently say that this kind of curvature manifests more often down to simple chemical reasons. Dimerization of analytes is known to occur in the heat of the source during the ion evaporation process, leading to signal loss at higher concentrations. The aforementioned adjustments can be made to good effect, but a frequent winner here is adjusting the mobile phase pH incrementally closer to neutrality, or indeed sliding to the other side of neutrality. This may necessitate chromatographic adjustment, and may make a decent, selective sample extraction even more important, if any interferences end up hard to chromatographically resolve. Then there is the possibility of moving to the often-overlooked APCI gas phase ion production conditions, great for sensitivity and straight lines with hydrophobic small molecules and often deals soundly with matrix effects that are encountered with electrospray-based conditions. Last but not least, another important parameter that could be altered in order to rectify this curvature is the declustering potential, the voltage applied to the orifice leading to the vacuum interface at the head of the mass filter rail and which governs how forcefully ions are drawn in, as is necessary to remove weakly-held adducts and solvent ion clusters from the analyte. That brings us quite nicely to the opposite type of curvature as well.
Disclaimer: the opinions expressed are solely my own and do not express the views or opinions of my employer.