First, can you cite a study finding what you say bolded above? (Always best to proactively back up assertions with citations, BTW). I dug around some and could find no supporting evidence: the closest thing was a summary of earlier research in PMID: 27163728, citing PMIDs 21840301 and 19090755, neither of which actually finds this.
Michael, I think it is fruitful to discuss the action of the metabolites in connection with the in vivo mice results, especially under this topic.
Coming back to the study “Fisetin is a senotherapeutic that extends health and lifespan” I found the following results surprising:
The diet studies with a daily feed of 60mg/kg led to a reduction of senescence markers in multiple organs. Assuming the mice were eating three meals per day one can calculate with 20 mg/kg of oral administration to estimate the peak Fisetin plasma concentration. According to the results of (https://doi.org/10.1080/10717544.2016.1245366) it should be reasonable to estimate a peak plasma concentration of about 140 ng/ml three times daily with a first rapid alpha half-life of 0.09 h (5.4 min), followed by a longer terminal disposition half-life beta of 3.12 h (doi:10.1016/j.bcp.2011.07.097).
Following these data the Fisetin plasma concentration in the daily feed would fall to a level of about 70 ng/ml after 5.4 min.
Looking into the in vitro analysis of Fisetin I found the following:
The lowest concentration thresholds for any senolytic activity for HUVECs in vitro of Fisetin is 500 ng/ml with three days exposure, but the main activities are in the range of several µg/ml (http://doi.org/10.18632/aging.101202). Obviously there are much lower plasma concentrations in vivo than needed to activate any senolytic activity.
The second mode of Fisetin administration in the study reveals again surprising results. Here the mice were treated with 100 mg/kg Fisetin for 5 consecutive days by oral gavage, or vehicle. So there was only one daily peak with a first rapid alpha half-life of 0.09 h (5.4 min), followed by a longer terminal disposition half-life beta of 3.12 h. Looking at available study data for the peak plasma concentrations we can assume between 700 ng/ml and below 1 µg/ml (based on https://doi.org/10.1...247.2013.860131 and https://doi.org/10.1...247.2013.860131). Again these concentrations seem to be too low to meet the in vitro threshold data mentioned before.
Summarizing these considerations suggest that it is difficult to infer the in vivo effects from their in vitro activities.
Based on these findings I believe that the metabolites should play an important role in the senolytic activities.
A second mechanism could be the following:
However,ß-glucuronidase was found to be present in liver and increasingly released by neutrophils at inflammation sites (28, 29). Therefore, deglucuronidation might occur in various tissues, resulting in the formation of flavonoid aglycones, which warrant further studies
→ source (external link)
Coming back to the start of the considerations you stated “it would be very suprising if the detoxification products — which are formed, exactly, to chemically castrate them and ensure their rapid removal from the circualtion — had similar activity.” Can you cite a study finding what you mentioned above?
If this is true the only left explanation for the in vivo senolytic results are deglucuronidation activities in the tissues where Fisetin showed senolytic activities.
Following the rationale above I think it is reasonable to assume that even in humans with the stronger metabolism oral Fisetin administration could lead to senolytic activities.