The authors raise an important question of what is the relevant concentration of SSRIs and where is the site of action. Casarotto et al. do not claim that the extracellular fluoxetine concentrations reach micromolar, but that brain concentrations do. Contrary to what is claimed in Table 1, Johnson, Lewis and Angier, 2007 did not use MRS but gas chromatography-MS in postmortem brain samples and found fluoxetine (norfluoxetine was excluded in this assay) concentrations between 1-30 microM. This paper confirms and validates the MRS measurements that the authors largely disregard. So while extracellular concentrations remain in nanomolar, SSRIs accumulate into brain, presumably in membranes, and specifically in lipid rafts, as shown by Rasenick and coworkers (but not cited by the authors) and can reach micromolar concentrations there. Membrane accumulation is also consistent with the data by Nichols et al, cited by the authors. Our unpublished data suggests that fluoxetine enters its binding site in TrkB not from the aqueous phase, but from the membrane phase after having equilibriated there at high-enough concentrations, therefore, membrane concentrations are more relevant to the SSRI action to TrkB than the extracellular levels. If extracellular concentrations are elevated to micromolar, membrane concentrations certainly increase to toxic levels, as demonstrated by the authors. So, the authors should modify their conclusion on page 3: while extracellular SSRI concentrations do not reach micromolar, brain concentrations do, and they may be more relevant for the mechanisms of action than the extracellular levels.