Interaction of the novel DNA damaging anticancer agent 11β with the androgen receptor

A22

We have previously described the rationally designed anticancer compound 11β, a bifunctional compound composed of a DNA damaging N,N-bis(2-chloroethyl)-aniline mustard tethered to an estradienone androgen receptor (AR) ligand via a stable linker. This compound exhibits striking effects upon exposure to LNCaP prostate cancer cells, eliciting apoptosis within 15 hours. It is also very effective in preventing growth of tumors in a LNCaP xenograft mouse model (Marquis et al. 2005 Chem. Biol. 12: 779). We propose that this compound works by forming sites of DNA damage that can recruit AR and prevent access of repair enzymes, thus creating selective toxicity toward cells expressing high levels of AR. In addition, the removal of AR from its natural sites of transcription should lead to the reduced transcription of genes necessary for survival.
 To this end, we have explored the interaction of 11β with the AR. It has been determined that 11β is selectively toxic in favor of the presence of AR in a pair of otherwise isogenic cell lines (PC3-AR (AR+), and PC3-neo (AR-)). Additionally, apoptosis is induced to a greater extent and at an earlier time in PC3-AR cells than PC3-neo cells as evidenced by caspase 3 cleavage.
 A luciferase transcriptional reporter assay driven by prostate specific antigen (PSA) promoter was established and employed in LNCaP cells to determine the agonistic and antagonistic abilities of 11β toward the AR. It was determined that 11β could act as a weak agonist in the absence of endogenous androgens; 1μM 11β was able to drive luciferase expression with approximately 16% efficiency of a 1nM dose of DHT, or about 3 fold above background levels. This same 1μM concentration of 11β also displayed antagonistic ability, competing away approximately 56% of the AR transcription driven by 1nM DHT in a competitive assay format. Evidence for the specificity of this interaction includes the observations that in PC3-neo cells lacking AR, 11β caused no changes in luciferase expression, while PC3-AR cells responded in similar fashion as LNCaP. Additionally, the agonistic interaction of 11β with the AR in LNCaP cells could be competed away with the AR antagonist flutamide.
 These findings were confirmed by measuring expression of the androgen regulated genes KLK2, KLK3, and TMPRSS2 by RT-PCR, and also by measuring secreted PSA protein after cell treatments mirroring those in the luciferase assay. Furthermore, a non-DNA damaging version of our compound, 11β-dimethoxy, has been tested to determine effects of 11β irrespective of DNA damage. This compound performs similarly to 11β in these transcriptional assays.
 Levels of AR protein and mRNA have also been studied. The protein level is reduced by approximately 50% within 6 hours upon treatment with a 5μM dose of 11β which does not cause apoptosis. Levels of AR mRNA follow a similar trend, falling by approximately 50% within 6 hours, suggesting altered transcription from the AR promoter.
 Affymetrix HGU133plus2 microarrays have been employed in LNCaP cells to observe global changes in gene expression following treatment with 11β. A robust transcriptional response is elicited by 11β, including increased expression of genes associated with cell cycle control (ATF3, p21, GADD45A), cholesterol biosynthesis (HMGCS1, HMGCR, SQLE), and unfolded protein response (EIF2AK3, HERPUD1, DDIT3). Efforts are underway to understand these transcriptional changes.