Current treatment for advanced PCa is mainly based on androgen ablation therapies like chemical or surgical castration or application of anti-androgens. Although hormonal therapy is initially very effective, the benefit of androgen withdrawal is only transitory. While still expressing high levels of functional AR, some PCa cells acquire the ability to grow and survive under castrate levels of circulating androgens. The mechanisms involved in an altered AR-signalling of CRPC cells include an over-expression/amplification of the AR (hypersensitive pathway), point mutations broadening ligand-specificity of the receptor (promiscuous pathway) as well as activation of the AR by peptide growth factors or cytokines (outlaw pathway) . A recently identified mechanism enabling CRPC cells to bypass the requirements for physiological levels of androgens is the expression of C-terminal truncated, constitutively active AR splicing variants termed AR-V. Due to the loss of the C-terminus, AR-Vs are lacking essential parts of the ligand binding domain . In the absence of androgens, AR-Vs were shown to induce AR-signalling either as AR-V/AR-V-homodimers or AR-V/AR-heterodimers [18, 48]. Whereas AR-V/AR-V homodimers are unaffected by hormone ablation in general, some AR-V/AR-heterodimers were insensitive to the commonly used anti-androgens . Given the fact that both, AR wild type as well as AR-V, play a crucial role in a high percentage of CRPC there is an urgent need for an AR-blockade that does not involve the LBD.
Nitric oxide, a free radical gas, has been repeatedly shown to impair the function of nuclear receptors via targeting their cys4-type zinc fingers, located in the DBD of these receptors . This is well exemplified by studies showing a NO-induced nitrosation of the cystein residues within the zinc finger structures of steroid receptors like the estrogen receptor (ER) . As the DBD is a crucial part of functional full length AR as well as constitutively active AR-Vs, it is tempting to speculate that NO-donors able to release NO in a spatial and timely controllable manner could be of special interest for the treatment of deregulated AR-signalling in CRPC cells.
Inhibitory effects of NO on AR-function in PCa cells was first described in vitro using the spontaneous NO-donor Deta/NO . In a recent clinical study Siemens et al. were able to show an increase of the PSA doubling time in patients treated with the NO-donor glyceryl trinitrate (GTN) . No cardiovascular toxicities or serious side effects were encountered during treatment with GTN . Although the number of patients enrolled in the study was relatively low (29 patients), our data and the clinical data with GTN [22, 51] support the assumption that NO might be suitable for the treatment of advanced PCa.
Commonly used NO-Donors like Deta/NO, spontaneously dissociate in a pH dependent process thereby releasing the free radical NO. Unfortunately NO exhibits only a relatively short half-life in the tissue culture medium or body fluids. Therefore, we tested a potent NO-prodrug, the non-ionic nitroaromatic diazeniumdiolate JS-K. The compound was shown to be activated by glutathione S-transferase (GST), a key phase II detoxification enzyme that is frequently over-expressed in cancer tissue. Among the various isoforms of GSTs which are located in the cytoplasm and the mitochondria, mainly the π isoform is able to form the so called < Meisenheimer complex >  which finally delivers NO directly within the cell. First experimental in vivo studies using tumor xenografts in a mouse model, showed that JS-K exhibits a strong anti-tumor activity and is generally well tolerated by the host organism [5, 23, 24].
In order to test the effects of JS-K on AR signalling, we treated the AR-positive CRPC cell line 22Rv1 grown in presence of DHT with increasing concentrations of JS-K. Generation of intracellular NO was monitored indirectly by detection of nitrotyrosine using fluorescence microscopy. JS-K was able to inhibit the genomic functions of the AR in 22Rv1 cells. The inhibition of AR-signalling was neither due to an inhibition of AR-dimerization nor to an inhibition of nuclear translocation. Interestingly, JS-K was able to diminish intracellular levels of all AR-isoforms in PCa cells. However, in contrast to the dramatic downregulation of AR-V and AR in 22Rv1 cells the intracellular AR levels of LNCaP were only slightly affected.
So far, studies on intracellular steroid receptor levels following NO-treatment are sparse. In a previous study using Deta/NO as an NO-donor the concentration of the endogenous AR in LNCaP remained unaffected . In contrast, DETA/NO was able to downregulate the intracellular levels of the Drosophila melanogaster ecdysteroid receptor (EcR) when overexpressed in chinese hamster ovary cells, CHO-K1 . Although the heterologeous expression of an insect steroid receptor in CHO-K1 cells differs largely from the situation found in LNCaP cells that express high levels of intracellular AR under physiological conditions the present data suggest that there are cell specific differences in the sensitivity of cells towards NO-treatment. Moreover, the generation of NO by JS-K largely depends on the intracellular GST-activity that differs between different cell lines or cell types.
The fact that NO is able to inhibit the canonical WNT-pathway in different tumor cells [42–44] including prostate cancer cells (as seen in this study) offers a more intriguing explanation for the massive down regulation of AR and AR-V in 22Rv1 cells. There is experimental evidence that the WNT/β-catenin-pathway is able to up-regulate AR-mRNA expression in prostate cancer cell lines through interaction with TCF/LEF binding sites situated in the promoter region of the AR . In contrast to the highly aggressive castration resistant 22Rv1 cells which are able to drive a WNT-typical T-cell factor (TCF)-dependent reporter gene activity, the androgen sensitive LNCaP are unable to do so . In consequence, the inhibition of the canonical WNT-pathway would lead to a decrease of AR and AR-V mRNA in 22Rv1 cells but not or to a lesser extent in LNCaP.
In proliferation assays castration resistant 22Rv1 and LNCaP-SSR cells were more susceptible to the growth inhibitory effects of JS-K than LNCaP. Whereas LNCaP cells largely depend on androgenic stimuli, the 22Rv1 and LNCaP-SSR cells are able to grow under androgen deprived conditions. LNCaP cells express a full length AR with a point mutation at position 877 (T877A), enabling the AR to be stimulated by different steroids (promiscuous AR) . In contrast to the hormone dependent LNCaP, the castration resistant LNCaP-SSR cells were shown to exhibit high levels of nuclear AR in the absence of hormonal stimuli . Increased nuclear AR-levels were paralleled by elevated PSA-levels suggesting that the AR is active in these cells . In 22Rv1 cells two AR-forms can be found: A larger AR-form expressing 3 zinc finger motifs due to the duplication of exon 3 (AREx3dup) and a C-terminally truncated, constitutively active AR-V . In the absence of androgenic stimuli, some AR-Vs have been shown to form constitutively active AR-V-homodimers or AR/AR-V-heterodimers, thereby uncoupling the need of CRPC cells for physiological levels of androgens [18, 48]. Although it is unknown whether AREx3dup is able to form androgen independent heterodimers with AR-V in 22Rv1 cells, its expression of 3 zinc finger structures makes it probably also more susceptible to the effects of NO. The fact that NO targets the zinc finger structures of wild type AR, mutated AR-forms as well as AR-Vs suggests that NO-donors are promising compounds for the treatment of deregulated AR-activity. Moreover, the observation that CRPC cells like 22Rv1 or LNCaP-SSR, expressing either constitutively active AR-Vs or functionally deregulated full length AR, are more susceptible to the effects of JS-K than LNCaP cells further supports this assumption.