Single-cell profiling quantifies the dose gap between population-average and complete pY-STAT1/3 suppression by JAK inhibitors

Inhibitor potency is commonly assessed from population-average responses, but how much single-cell signaling remains under strong average suppression is less clear. We measured nuclear STAT1 and STAT3 tyrosine phosphorylation in two human monocytic cell lines stimulated with IFN-γ; and IL-10 and treated with six JAK-pathway inhibitors. Across 24 profiles, we compared arithmetic-mean inhibition and inhibition of the mean of log10-transformed single-cell responses (log-mean) with the fractions of cells leaving the stimulated range (partial inhibition) or entering the non-stimulated range (complete inhibition). At the lowest tested doses where arithmetic-mean or log-mean inhibition reached at least 80%, median fractions remaining above the non-stimulated range were 76% and 37%, respectively. Thus, strong average suppression frequently left substantial fractions of cells with tangibly elevated pY-STAT responses. Log-mean inhibition generally approximated partial-inhibition occupancy, whereas the apparent IC50 for complete-inhibition occupancy was a median 2.4-fold higher than the log-mean IC50 and 7.4-fold higher than the arithmetic-mean IC50. These relationships provide an empirical guide for estimating cell fractions from conventional inhibition curves.