The translation path should start with target-validation rather than medicinal chemistry. Use behavioral and molecular assays in stress paradigms with ADORA2A antagonists, agonists, and genetic loss-of-function to determine whether parthenolide's antidepressant signal collapses when A2A signaling is fixed experimentally.
If the signal truly routes through ADORA2A, the indirect extracellular-adenosine model is the most developable because it suggests measurable biomarkers: adenosine tone, phospho-CREB, DARPP-32 state, and astrocyte/microglial inflammatory markers. Direct receptor chemistry is less likely given parthenolide's scaffold, and off-target covalency plus PK liabilities would complicate a pure receptor program unless structure-activity work cleans up the electrophile.