Re-activation of AR expression in AR-negative prostate cancer
Because the AR pathway represents such a key therapeutic contact point in prostate cancer, researchers have set out to search for means to restore AR signaling in AR-negative prostate cancer (Fig. 2). As mentioned above, induced lineage plasticity mediates the onset of AR-negative prostate cancer. For this reason, silencing or inhibiting transcription factors of cellular plasticity, such as SOX2, could have an impact on AR re-expression. In line with this, SOX2 silencing abrogates the capability of cells with TP53 and RB1 loss to develop lineage plasticity and to express neuroendocrine markers [45].
Alternatively, drugs that affect epigenetic silencing in different ways have been successfully used to re-express AR in AR-negative cell lines. Among these are drugs that block DNA methylation (azacytidine) and increase active (HDAC inhibitors Trichostatin A) or diminish repressive histone marks (EZH2 inhibitors) at the AR promoter [73, 74, 79] (Fig. 2). Most notably, various EZH2 inhibitors (DZNep, GSK126, EPZ6438) have been reported to upregulate AR protein expression in human NEPC cell lines NCI-H660/MDA PCa 144-13 and in mouse prostate epithelial cells with ablation of Rb1 and Trp53 [44, 73]. That said, the pharmacological inhibition of EZH2 does not seem sufficient to re-gain AR positivity in recently generated patient-derived NEPC organoid lines [80].
Moreover, indirect perturbations have shown the ability to restore AR expression as well. Among these are the administering of the nerve growth factor (NGF) to DU145 cells [81] (Fig. 2). NGF has two receptors, p75NTR and TrkA. p75NTR is known in prostate cancer field since its tumor suppressor features [82]. Moreover, its expression decreases during disease progression. Conversely, Trka activation has shown oncogenic features in prostate cancer [83]. Although the authors did not explain the mechanism of NGF-mediated AR re-expression in prostate cancer cells, NGF administration can downregulate DNA methyltransferases in other cells [84]. It is tempting to speculate that NGF treatment could restore AR through downmodulation of DNA methylation.
More recently, the Forkhead Box C2 protein (FOXC2) has been shown to modulate AR in both AR-positive and -negative cell lines. Generally, its expression is negatively correlated with AR presence. Indeed, its silencing in AR-negative DU145 cells can strongly re-establish AR protein expression. The effects on FOXC2 targets are mediated by Zeb1, a known transcriptional repressor. Since FOXC2 is directly activated by P-p38 phosphorylation, inhibition of this pathway by the p38 inhibitor SB203580 can exert the same effect in restoring AR expression [85] (Fig. 2). Interestingly, SB203580 administration can also decrease SOX2 levels in AR-negative cells, possibly linking p38 activity, SOX2 expression, and cellular plasticity [85].
At this point, the studies mentioned above provide only incomplete insights into the feasibility and relevance of AR re-expression in a few cell line models. Moreover, the clinical relevance is yet to be fully discovered. Nevertheless, these findings confirm that it is possible to revert at least in part the process that leads to de-differentiation and AR negativity.
