Concentration categories do not represent equivalent exposure intensity in comparative PFAS transcriptomics

Sprenger et al. (2026) provide an extensive transcriptomic dataset covering 33 PFAS congeners in differentiated HepaRG cells and report broadly comparable molecular responses across structurally diverse substances. The dataset is valuable, but the cross-PFAS comparison requires a clearer distinction between relative concentration categories, numerical exposure concentrations and toxicological potency. For each PFAS, three non-cytotoxic concentrations were selected individually. The resulting numerical concentrations ranged from 5 to 250 µM. All 33 congeners were tested at 25 µM, 25 of 33 at 100 µM and 22 of 33 at 250 µM. Nevertheless, the subsequent comparative IPA analyses were based on the labels "low", "medium" and "high", rather than numerical concentrations. These labels therefore do not represent equivalent exposure intensities across PFAS. The distinction becomes consequential because the authors report concentration-dependent increases in differentially expressed genes for most compounds and because several additional pathways emerged predominantly at the high concentration, which was close to the cytotoxicity threshold for some PFAS but substantially below it for others. Thus, the high category can simultaneously encode chemical identity, nominal dose and position relative to cytotoxicity. This does not mean that category-based comparisons are invalid. Rather, they answer a narrower question: whether compounds perturb common pathways at their individually selected exposure levels. Concentration-resolved or potency-based analyses would complement this approach and help distinguish shared molecular response from comparable potency and from effects emerging only at different positions on the concentration-response continuum.