Senolytic compounds control a distinct fate of androgen receptor agonist- and antagonist-induced cellular senescent LNCaP prostate cancer cells

Background The benefit of inducing cellular senescence as a tumor suppressive strategy remains questionable due to the senescence-associated secretory phenotype. Hence, studies and development of senolytic compounds that induce cell death in senescent cells have recently emerged. Senescent cells are hypothesized to exhibit different upregulated pro-survival/anti-apoptotic networks depending on the senescent inducers. This might limit the effect of a particular senolytic compound that targets rather only a specific pathway. Interestingly, cellular senescence in prostate cancer (PCa) cells can be induced by either androgen receptor (AR) agonists at supraphysiological androgen level (SAL) used in bipolar androgen therapy or by AR antagonists. This challenges to define ligand-specific senolytic compounds. Results Here, we first induced cellular senescence by treating androgen-sensitive PCa LNCaP cells with either SAL or the AR antagonist Enzalutamide (ENZ). Subsequently, cells were incubated with the HSP90 inhibitor Ganetespib (GT), the Bcl-2 family inhibitor ABT263, or the Akt inhibitor MK2206 to analyze senolysis. GT and ABT263 are known senolytic compounds. We observed that GT exhibits senolytic activity specifically in SAL-pretreated PCa cells. Mechanistically, GT treatment results in reduction of AR, Akt, and phospho-S6 (p-S6) protein levels. Surprisingly, ABT263 lacks senolytic effect in both AR agonist- and antagonist-pretreated cells. ABT263 treatment does not affect AR, Akt, or S6 protein levels. Treatment with MK2206 does not reduce AR protein level and, as expected, potently inhibits Akt phosphorylation. However, ENZ-induced cellular senescent cells undergo apoptosis by MK2206, whereas SAL-treated cells are resistant. In line with this, we reveal that the pro-survival p-S6 level is higher in SAL-induced cellular senescent PCa cells compared to ENZ-treated cells. These data indicate a difference in the agonist- or antagonist-induced cellular senescence and suggest a novel role of MK2206 as a senolytic agent preferentially for AR antagonist-treated cells. Conclusion Taken together, our data suggest that both AR agonist and antagonist induce cellular senescence but differentially upregulate a pro-survival signaling which preferentially sensitize androgen-sensitive PCa LNCaP cells to a specific senolytic compound.