Evaluation of Spectral Counting for Relative Quantitation of Proteoforms in Top-Down Proteomics

Spectral counting
is a straightforward label-free quantitation
strategy used in bottom-up proteomics workflows. The application of
spectral counting in label-free top-down proteomics workflows can
be similarly straightforward but has not been applied as widely as
quantitation by chromatographic peak areas or peak intensities. In
this study, we evaluate spectral counting for quantitative comparisons
in label-free top-down proteomics workflows by comparison with chromatographic
peak areas and intensities. We tested these quantitation approaches
by spiking standard proteins into a complex protein background and
comparing relative quantitation by spectral counts with normalized
chromatographic peak areas and peak intensities from deconvoluted
extracted ion chromatograms of the spiked proteins. Ratio estimates
and statistical significance of differential abundance from each quantitation
technique are evaluated against the expected ratios and each other.
In this experiment, spectral counting was able to detect differential
abundance of spiked proteins for expected ratios ≥2, with comparable
or higher sensitivity than normalized areas and intensities. We also
found that while ratio estimates using peak areas and intensities
are usually more accurate, the spectral-counting-based estimates are
not substantially worse. Following the evaluation and comparison of
these label-free top-down quantitation strategies using spiked proteins,
spectral counting, along with normalized chromatographic peak areas
and intensities, were used to analyze the complex protein cargo of
exosomes shed by myeloid-derived suppressor cells collected under
high and low conditions of inflammation, revealing statistically significant
differences in abundance for several proteoforms, including the active
pro-inflammatory proteins S100A8 and S100A9.