Urupocidin C: a new marine guanidine alkaloid which selectively kills prostate cancer cells via mitochondria targeting

New bicyclic guanidine alkaloid, urupocidin C (Ur-C) along with the previously known urupocidin A (Ur-A) were isolated from the rare deep-sea marine sponge Monanchora pulchra, harvested in Northwestern Pacific waters. The unique structure of Ur-C was elucidated using 1D and 2D NMR spectroscopy as well as mass spectra. We discovered a promising selectivity of both alkaloids for human prostate cancer (PCa) cells, including highly drug-resistant lines, compared to non-malignant cells. In cancer cells, marine derived compounds were able to induce G1- and S-cell cycle arrest as well as caspase-mediated cell death. For the first time we have identified mitochondrial targeting as a central mechanism of anticancer action for these and similar molecules. Thus, treatment with the isolated alkaloids resulted in mitochondrial membrane permeabilization consequently leading to the release of cytotoxic mitochondrial proteins to cellular cytoplasm, ROS upregulation, consequent activation of caspase-9 and -3, followed by PARP cleavage, DNA fragmentation, and apoptosis. Moreover, synergistic effects were observed when Ur-A and Ur-C were combined with clinically approved PARP inhibitor olaparib. Finally, these alkaloids exhibited additive effects in combination with docetaxel and androgen receptor inhibitor enzalutamide, both applied in PCa therapy. In conclusion, urupocidin-like compounds are promising lead molecules for the development of new drugs for the treatment of advanced PCa.

demand for these substances for further clinical development and use synthetic and biotechnological methods as well as aquaculture rearing have been established 9,10 .
Marine guanidine alkaloids are a group of biologically active compounds found in some marine sponges. These compounds were suggested to be chemotaxonomic markers of the marine sponges belonging to the genera Ptilocaulis, Hemimycale, Crambe, Batzella, Clathria, and Monanchora 11,12 . Specifically, guanidine alkaloids extracted from the marine sponges of the Monanchora genus (family Crambeidae) are characterized by unique chemical structures and an impressive spectrum of biological activities (reviewed in 11,13 ). These molecules comprise pentacyclic, tricyclic, bicyclic or acyclic guanidine cores. In bicyclic alkaloids mono-, bi-, and trisubstituted skeleton systems can be found 14 . To date, only two members of the trisubstituted bicyclic guanidine alkaloids have been described, namely urupocidin A and B -both recently discovered by us 11,[13][14][15] . In addition to its unusual structural features, urupocidin A (Ur-A) exhibits promising biological properties, including anti-cancer activities [15][16][17] . However, our knowledge about the mechanism of action of these newly identified alkaloids is still incomplete.
To date, cytotoxic activity of Ur-A has been reported against different human cancer cell lines 17 . Ur-A was found to inhibit an EGF-induced neoplastic transformation of epithelial cells 17 and to induce a G 2 /M-phase cell cycle arrest as well as apoptosis of human cervical carcinoma HeLa cells. The latter was revealed by a significant increase of DNA fragmentation. However, only minor activation of caspase-3 and -7 was detected in the treated cells, suggesting a caspase-independent character of apoptosis induction 17 . In addition, Ur-A killed murine epithelial JB6 Cl41 cells without alteration of the transcriptional activity of p53, implying a p53-independent character of apoptosis 17 . Furthermore, Ur-A increased NO levels in murine macrophages treated at micromolar concentrations 15 . Finally, Ur-A and its semi-synthetic derivative inhibited TRPV receptors 16 . However, despite these findings, the current knowledge on the biological effects of marine natural compounds belonging to the urupocidin structural family is incomplete and cellular targets are still unknown.
In this study, in continuation of our search for bioactive natural products isolated from the northwestern Pacific marine sponge Monanchora pulchra 18 , we report on the isolation and structure elucidation of the new bicyclic guanidine alkaloid Ur-C along with previously known Ur-A. Mechanism of action and anticancer activity alone or in combination with established drugs were examined in human prostate cancer (PCa) cells. Prostate cancer was chosen because of the persisting high unmet medical need in advanced stages. In fact, aggressive phenotypes and impaired clinical outcome can be frequently found with increasing treatment duration due to development of resistance to standard therapies 19 . Both compounds showed cytotoxic activity in cell lines representing these advanced stages mainly mediated by mitochondrial targeting. Thus, Ur-A and Ur-C represent novel and promising candidates for development of effective drugs against aggressive and highly resistant PCa.

Analysis of apoptosis (annexin-V-FITC/PI double staining).
The experiment was performed as described before 21 . In brief, 0.2 × 10 6 cells/well were seeded in 6-well plates, incubated overnight, and pre-treated with 100 µM z-VAD(OMe)-fmk or an equal volume of vehicle for 1 h in 2 mL of fresh media/well. Cells were then treated with the investigated drugs for 48 h, harvested with trypsin, immediately stained with propidium iodide and annexin-V-FITC for 15 min. Further analyses were performed using FACS Calibur (BD Bioscience, San Jose, CA, USA) followed by quantification using the BD Bioscience Cell Quest Pro v.5.2.1. software.
MTT assay. The experiment was performed as described before 22 . In brief, 6000 cells/well were seeded in a 96-well plate, incubated overnight and treated with the tested drugs in 100 μl/well of fresh media for 48 h, unless otherwise stated. The cells were incubated with MTT reagent (3-(-4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) and the viability was measured using a spectrophotometer Infinite F200PRO reader (TECAN, Männedorf, Switzerland).
Western blotting. The experiment was performed as described before 20 . In brief, cells (1 × 10 6 cells/well) were seeded in ø 6 cm Petri dishes (TC Dish, Sarstedt, Numbrecht, Germany) in 5 mL/dish of media. The cells were incubated overnight and treated with the tested compounds for 48 h. Cells were harvested with the scraper, washed with ice cold PBS, and resuspended in Western blotting lysis buffer. The samples were frozen overnight at −20 °C, centrifuged at 10,000 g and the protein concentrations in the supernatants were determined using the Bradford assay. The total protein extracts (20-30 μg/ sample) were subjected to electrophoresis in gradient Mini-PROTEAN TGX Stain-Free gels (Bio-Rad, Hercules, CA, USA) at 200 V. The proteins then were transferred to a ø 0.2 μm pore PVDF membrane. The membrane was blocked with 5% BSA/TBS-T solution and then incubated with the primary antibody overnight according to the manufacturers' protocols. The membrane was washed with TBS-T, incubated with the secondary antibody, and again washed with TBS-T. The signal was detected using the ECL chemiluminescence system (Thermo Scientific, Rockford, IL, USA) according to the manufacturers' protocol. β-Actin was used as a loading control. For the list of used antibody see Supplementary information. The images were proceeded with the CorelDRAW X7 software (V. 17.1.0.572, Corel Corporation, Ottawa, Canada).
cell fractionation. The separation of cellular nuclear, mitochondrial and cytosolic fractions was performed using the Cell Fractionation Kit ab109719 (abcam, Cambridge, MA, USA) as reported before 21 . In brief, 4 × 10 6 cells were seeded in T75 culture bottles containing 20 mL of media and incubated overnight. Cells were treated with investigated compounds for another 48 h and harvested using a cell scraper. Further procedures were performed according to the manufacturer's protocol with slight modification. Note, cytosolic and mitochondrial fractions were concentrated using the Amicon Ultra-2 Centrifugal Filter device (Cat. No. UFC203024, Merck, Darmstadt, Germany). Analysis of mitochondrial membrane potential (ΔΨ m ). The alteration of the mitochondrial membrane potential (ΔΨm) was evaluated using ΔΨm-sensitive JC-1 dye (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraet hyl-imidacarbocyanine iodide) as previously described 20 . In brief, 0.1 × 10 6 cells/well were seeded in 12-well plates and incubated overnight in 1 mL media/well. The media was exchanged with fresh PBS (1 mL/well) containing the investigated compounds. The cells were incubated for 2 h (37 ° C, 5% CO 2 ), trypsinized, pelleted and stained with 100 μL/well of 2 µM JC-1/PBS. The cells were further incubated for 1 h in the dark (37 °C, 5% CO 2 ) and analysed using FACS Calibur device (BD Bioscience, San Jose, CA, USA) and BD Bioscience Cell Quest Pro v.5.2.1. software.

Analysis of cytoplasmatic ca 2+ level change. Alterations of the cytoplasmatic calcium concentration
were investigated by Fluo-4 NW Calcium Assay Kit (Cat. No. F36205, Molecular probes, Invitrogen, Eugene, OR, USA). In brief, 20 × 10 3 cells/well were seeded in a black 96-well plate with transparent bottom in 100 µL media/ well. The cells were incubated overnight and treated with investigated compounds in 100 µL fresh media/well for 2 h in the dark (37 °C, 5% CO 2 ). The media was then removed, the plates were washed with PBS (100 µL/well), and 100 μL of the dye solution was added per well. The plates were incubated in the dark (37° C, 5% CO 2 ) for 3 h, the dye solution was then exchanged by PBS (100 µL/well) and the measurement was performed using spectrophotometer Infinite F200PRO reader (TECAN, Männedorf, Switzerland) according to manufacturer's protocol.
The results were normalized against the cell viability, measured by MTS assay performed as described before 23 .
Determination of cytotoxic effects in combination with clinically established drugs. Synergistic, antagonistic or additive effects of the isolated compounds in combination with cisplatin, carboplatin, docetaxel, olaparib and enzalutamide were evaluated by the Chou-Talalay method 24,25 . The data were generated using the MTT test as described above. The cells were treated with the individual drugs or their combinations in a non-constant molar ratio. A combinational index (CI) was calculated using the CompuSyn v.  (Fig. 1b). The subfraction eluted with [EtOH: H 2 O (1: 9)] was further purified using a reversed-phase HPLC and the elution system [EtOH: H 2 O (62: 38) + TFA (0.1%)] to obtain the two pure compounds 1 and 2 (Fig. 1c). elucidation of the chemical structure. Compound 2 was identified as the previously known Ur-A based on its NMR and HRESIMS data and a comparison with the authentic sample of the previously isolated compound 15 (Fig. 1c).
The  (Table 1) suggested a close relationship with Ur-A (2). Further analysis of the COSY, HSQC and HMBC data (Fig. 1d) of 1 indicated that the structures of the aliphatic chains in 1 were the same as in 2. The ester carbonyl carbon C-24 at δ C 165.3 was assigned by HMBC correlations from H-25, which was also correlated to C-26 and C-27. The chemical shift value of CH 2 -28 at δ H 3.64/δ c 52.7 indicated a hydroxylated N-29a nitrogen atom of guanidine group in 1 15 . However, the other carbon signals observed in the spectra of 1, were not in agreement with those observed for 2. Specifically, no N-substituted methine signal was observed, and the HMBC spectra showed the presence of nonprotonated carbons at δ C 114.5, 168.5, and 180.9. The chemical shift value of proton at C-10 at δ H 4.89 supported the protonated form of the tertiary N-23a in 1 26 . The downfield signal at δ C 180.9 along with the nine degrees of unsaturation suggested by the molecular formula, showed that compound 1 was a dehydro-analogue of Ur-A. HMBC correlations from the methylene protons at δ C 3.14 (H 2 -16) to carbons at δ C 114.5 and 180.9, as well as to C-17 and C-18, and from H 2 -11 to C-12 and C-13 confirmed this notion. The bicyclic guanidine portion of 1 is thus similar to the bicyclic guanidine structures found in dihydrocrambescin A2 418 27 and dehydrocrambine A 28 .
Compounds 1 and 2 may be biosynthesized via the same pathway as both alkaloids were isolated from the same marine sponge; therefore, it is very likely that their absolute configurations are also identical. The specific rotation of 1 was similar to the reported value of compound 2 15 . Furthermore, as the absolute configuration of www.nature.com/scientificreports www.nature.com/scientificreports/ 2 has been established by chemical transformations and CD 15 , the absolute configuration of 1 was suggested to be 8R, 10R. Thus, 1 was identified as a new analogue of bicyclic guanidine alkaloids belonging to the urupocidin structural family 15 , and named urupocidin C (Ur-C). To the best of our knowledge, this is the third member of guanidine alkaloid class possessing trisubstituted bicyclic skeleton.
Thus, the isolated compound 1 was identified as a new natural product Ur-C, compound 2 -as a previously known Ur-A (Fig. 1c).

Ur-A and Ur-C exhibit cytotoxicity in human prostate cancer cells. To investigate the cytotox-
icity of Ur-A and Ur-C in vitro, we used the human castration resistant prostate cancer 22Rv1 cells (Fig. 2a), human hormone-sensitive prostate cancer LNCaP cells (Fig. 2b), as well as non-malignant human lung fibroblast MRC-9 cells (Fig. 2c). Cisplatin was used as a reference substance (positive control). The IC 50 s for both investigated compounds towards cancer cells were in the low micromolar range (Fig. 2a-c). For 22Rv1 cells the cytotoxic activity of Ur-C was comparable to cisplatin, whereas Ur-A was slightly more active (Fig. 2a-c). Notably, hormone-dependent LNCaP cells were ~6 times more sensitive to Ur-A and Ur-C in comparison to cisplatin (Fig. 2b). More important, the IC 50 s determined by MTT assay in cancer cells were significantly lower in comparison with non-malignant MRC-9 cells and the selectivity indexes (SI, MRC-9 versus 22Rv1 cells) for Ur-A and Ur-C were determined as 3.7 and 2.4, correspondently. Thus, we conclude both compounds to exhibit a selectivity towards human prostate cancer cells.

Effect on cell cycle progression and apoptosis induction.
Drug-resistant 22Rv1 cells were chosen as a main model for the further investigation of mode of action of the isolated alkaloids. In order to investigate the mechanisms which stipulate the anticancer activity of the compounds, we further examined their effects on cell cycle progression and apoptosis induction in cancer cells. First, alterations of the cell cycle were accessed by flow cytometry (Fig. 3a). Cells treated with 5 µM of either Ur-A or Ur-C revealed significant increase of the cellular populations in G1-and S-phases, indicating a drug-induced cell cycle arrest (Fig. 3b).
In addition, a dose dependent DNA fragmentation of prostate cancer cells treated with Ur-A and Ur-C for 48 h was detected in flow cytometry indicating apoptosis induction (Fig. 3c,d). In line with this observation, Western blot analyses of the protein extracts of the treated 22Rv1 cells revealed dose-dependent PARPand caspase-3 cleavage -two well-established apoptotic hallmarks (Fig. 3e). Furthermore, the expression of anti-apoptotic proteins, such as IAP protein survivin and Bcl-2 was suppressed under the treatment (Fig. 3e). Simultaneously, a phosphatidylserine externalization detected in the cells following 48 h treatment indicated the induction of a "classical" early apoptosis (Fig. 3f,g). Remarkably, induction of apoptosis was effectively inhibited by pre-treatment with pan-caspase inhibitor z-VAD(OMe)-fmk (Fig. 3f,g). In conclusion, these results indicate a caspase-dependent character of the Ur-A-and Ur-C-induced apoptosis in 22Rv1 cells.

Apoptosis is exerted via intrinsic (mitochondrial) pathway.
Following the detection of caspase-3 activation as well as a caspase-dependent character of the induced apoptosis, we examined the effect of compounds on caspase-9. Remarkably, a cleavage of caspase-9 was detected following treatment with both, Ur-A and Ur-C. Of note, while an activation of caspase-9 was detected already after 12 h of treatment, no cleavage of caspase-3 was observed at this time (Fig. 4a). Therefore, we assume that treatment with Ur-A and -C first leads to an activation of initiator caspase-9, which later causes the cleavage of caspase-3 suggesting the intrinsic (mitochondrial) apoptotic pathway to be involved in the cytotoxic action of the compounds 29 .
To confirm this hypothesis we further investigated the effects on the mitochondria of the cancer cells. Mitochondrial stress is a primary event of the intrinsic apoptotic pathway, which is accompanied by mitochondrial membrane permeabilization and release of specific cytotoxic proteins 29 . Thus, we demonstrated a translocation of cytochrome C (an activator of procaspase-9) and the apoptosis inducing factor (AIF) from mitochondria to cytoplasm of treated 22Rv1 cells (Fig. 4b). This suggests the permeabilization of the mitochondrial membrane under treatment. In line with this, a down-regulation of the anti-apoptotic protein Bcl-2 was found (Fig. 3e). Bcl-2 down-regulation is known to cause release of apoptogenic proteins from mitochondria 30 .

Mitochondria are a primary target of Ur-A and Ur-C in prostate cancer cells.
Based on the observations of the cleavage of caspase-9 prior to caspase-3, as well as the release of mitochondrial proteins to cytoplasm, www.nature.com/scientificreports www.nature.com/scientificreports/ mitochondria were assumed to be a primary target of Ur-A and Ur-C in cancer cells. For further confirmation, we investigated the initial effects, taking place in 22Rv1 cells shortly after treatment initiation. Interestingly, we could show that most of the cells exhibited the drop down of Δψ m after 2 h treatment (Fig. 5a,b). Disturbance of mitochondrial integrity and permeabilization of mitochondrial membrane result in ROS induction and ROS production 31,32 . Indeed, an increase of the ROS level was detected shortly after drug exposure (Fig. 5c). Simultaneously, increase of the cytoplasmatic Ca 2+ levels was observed (Fig. 5d). It is known that an elevated ROS level can lead to ER-stress and consequently to calcium release into cellular cytoplasm 33 . Therefore, the observed effect may result from ER targeting caused by ROS upregulation secondary to mitochondrial targeting. www.nature.com/scientificreports www.nature.com/scientificreports/ Anticancer in vitro effects of Ur-A and Ur-C in combination with established anticancer drugs. The anticancer effects of Ur-A and Ur-C were examined in combination with standard anti-cancer therapies. Thus, we examined the effects of the isolated alkaloids together with DNA-binding (cross-linking) drugs cisplatin and carboplatin, microtubuline stabilizing agent docetaxel, PARP inhibitor olaparib (Fig. 6a), as well as androgen receptor targeting drug enzalutamide (Fig. 6b).
The combination of Ur-A and Ur-C with platinum based agents cisplatin and carboplatin showed additive effects in the range of high Fa (fraction affected) values, i.e. at cytotoxic doses of the combo drug (Fig. 6a). At the same time slight signs of antagonism were observed in the range of lower Fa values (Fig. 6a). This effect should be considered and carefully examined prior to further in vivo experiments or clinical trials. Combination of Ur-C with the taxol derivative docetaxel showed promising additive/synergistic effects, whereas the combination of Ur-A with docetaxel was less active (Fig. 6a). Most promising results were obtained for the combination with olaparib (Fig. 6a). For both Ur-A and Ur-C well pronounced synergistic effects were observed (CI <0.5, Fig. 6a).
Finally, additive effects of the isolated compounds in combination with anti-androgen drug enzalutamide were found (Fig. 6b). It is important to note that 22Rv1 cells used for the experiments express androgen receptor splice variant V7 (AR-V7) 34 , which can autoactivate the AR-pathway in the absence of androgens 35 and ultimately lead to resistance to enzalutamide. Interestingly, we found a down-regulation/degradation of AR-V7 as well as AR full length (AR-FL) in cells treated with the investigated alkaloids (Fig. 6c), which could partially explain the observed recall of sensitivity to enzalutamide. In accordance with this, a down-regulation of PSA in treated cells was detected indicating a suppression of AR-signaling by Ur-A and Ur-C (Fig. 6c).

Discussion
Mitochondria play an important role in energy metabolism, cell cycle regulation and survival of eukaryotic cells. The idea to develop molecules which are able to specifically target mitochondria raises from 1950 when the mitochondrial structure has been described and therefore the molecular basics of mitochondrial-targeting properties of small molecules have been first understood 36,37 . The concept of mitochondria targeting is based on the 3-to 5-fold difference between mitochondrial membrane potential (Δψ m ) and plasma membrane potential 38 . Therefore, positively charged molecules, in particular of those containing so called delocalized lipophilic cations (DLC), are able to target and accumulate in mitochondria 36 .
Mitochondria are an attractive target for anticancer therapy. These organelles play a critical role in several apoptosis-related processes by mediating cytotoxic ROS production and release of cytotoxic proteins (e.g. cytochrome C, AIF, and others) 39 , which can be induced by a number of different drugs and stimuli. Additionally, they control vital ATP production and metabolic pathway signaling, which are critical for cell proliferation 40 . Thus, alteration of mitochondria function or homeostasis may easily initiate programmed cell death. Mitochondria of cancer cells exhibit special features which discriminates them from non-malignant cells. Thus, in some cases the mitochondria of cancer cells exhibit significantly higher transmembrane potential (Δψ m ) in comparison to non-malignant cells 41,42 . This difference gives an opportunity to develop selective and effective drugs targeting cancer cell mitochondria and therefore the tumor cells 36 . Additionally, mitochondria targeting agents may cause less resistance in cancer cells due to the rather downstream position of this target in the apoptotic cascade 43 . Apoptotic cell death mediated by the mitochondria-targeting agents is often observed in fast proliferating cancer cells (due to the lower antioxidant capacity of mitochondria in these cells), while non-dividing or slowly proliferating cells are less affected or even intact 44 .
A well-known example for mitochondria-targeting is FDA-approved Bcl-2 inhibitor Venetoclax, which is used for treatment of recurrent leukemia and lymphoma 40 . Several other Bcl-2 family protein inhibitors are currently in different stages of clinical or preclinical development 40 . Additionally, standard chemotherapies such as paclitaxel, etoposide and vinorelbine may directly or indirectly affect mitochondria contributing to cancer cell death 39 .
In the current study, we isolated a new alkaloid Ur-C and previously known Ur-A from marine sponge M. pulchra. Along with urupocidins A and B 15 , Ur-C is the 3 rd known representative of trisubstituted bicyclic guanidine alkaloids. Ur-C exhibited a promising cytotoxic activity by targeting mitochondria of cancer cells. Thus, Ur-C and "mother" alkaloid Ur-A induced signs of mitochondria targeting such as Δψ m loss, ROS production and ER stress shortly after treatment. This further lead to the release of cytotoxic mitochondrial proteins to cytoplasm, www.nature.com/scientificreports www.nature.com/scientificreports/ consequent caspase-9 activation, which further provoked caspase-3 and PARP cleavage, DNA degradation and ultimately resulted in caspase-dependent apoptosis. Both compounds also induced G1-and S-phase cell cycle arrest of 22Rv1 cells. Interestingly, we previously observed the G2/M-phase arrest in human cervical carcinoma  www.nature.com/scientificreports www.nature.com/scientificreports/ HeLa cells treated with Ur-A 17 . Therefore, this effect seems to be cancer type specific (or even cell line specific). Accordingly, we did not detect any caspase-3 activation in the Ur-A-treated HeLa cells in our previous study 17 , whereas a cleavage of caspase-3 and moreover a caspase-dependent apoptosis was observed in prostate cancer cells within the current project.
Specific delivery of drugs to mitochondria (and therefore their targeting) may be provided by conjugation of mitochondria-targeting ligands (e.g. delocalized lypophilic cations, DLC) either directly to an active molecule or to a nanocarrier 39 . A unique feature of Ur-A and Ur-C is a guanidine moiety in the structure of both compounds. Guanidine belongs to the DLC family; it has a delocalized positive charge, which allows this molecule to enter the highly negative charged mitochondrial matrix 36 . Therefore, these bioactive alkaloids combine mitochondrial affinity and cytotoxic properties, which stipulates their mode of action and selectivity to cancer cells. Indeed, Ur-A and Ur-C revealed higher selectivity for PCa cells compared to non-malignant cells.
We observed a synergistic effect of the investigated alkaloids in combination with PARP (poly(ADP-ribose) polymerase) inhibitor olaparib. PARP detects ssDNA breaks and induces the repair cascade. A blocked repair of a single strand break (via PARP inhibition) causes more severe double strand breaks during DNA replication and consequently leads to apoptosis, which is also known as synthetic lethality 45,46 . Olaparib is effective in tumors harboring DNA repair deficiencies, especially those with double strand break repair defects, like BRCA1/2 47,48 . 22Rv1 cells are known to bear a BRCA2 mutation and therefore have an impaired dsDNA break repair mechanism making them sensitive to olaparib 49 . In our experiments, we have shown that treatment with Ur-A and Ur-C dramatically increases the intracellular ROS level. ROS are well-known inducers of ssDNA breaks 50 . Co-treatment with olaparib prevents a restoration of DNA, finally leading to dsDNA breaks. Due to the BRCA2 defect of 22Rv1 cells a so called synthetic cell death is ultimately induced. Of note, DNA repair defects are acquired during the course of treatment of advanced prostate cancer. Thus, while in localized disease only up to 3% of the patients harbor a BRCA2 defect, up to 13.3% carry this most common alteration in advanced stages of metastatic, castration resistant PCa 51 .
Finally, an additive effect was observed for the combination of Ur-A/C with the androgen receptor (AR) inhibitor enzalutamide. This anti-androgen drug binds to the AR thus blocking binding of this receptor to testosterone and dihydrotestosterone 49,52 . Thus, enzalutamide inhibits the AR-signaling cascade, which is essential for the survival and progression of prostate cancer cells 52,53 . However, in advanced disease stages prostate cancer cells may exhibit resistance to this drug. A major mechanism of resistance is alternative splicing leading to the expression of the androgen receptor splice variant V7 (AR-V7). AR-V7 lacks the C-terminal androgen binding domain, which leads to the permanent activation of AR as a transcription factor and results in the promotion of cancer cell growth 35,54 . We have demonstrated the ability of both Ur-A and Ur-C to increase the effect of enzalutamide in 22Rv1 cells, which are known to be resistant to enzalutamide-like AR-targeting agents due to the expression of AR-V7. Thus, the isolated alkaloids were able to re-sensitize the cells to AR-targeted therapy. This can be explained by the drug-induced decrease of AR-V7 expression, as well as general decrease of AR-signaling.
conclusions Ur-C is a new cytotoxic bicyclic guanidine alkaloid, which was isolated by us from the deep-water marine sponge Monanchora pulchra along with the previously known Ur-A. The mechanism of action of both compounds includes mitochondria targeting, which further leads to the caspase-dependent apoptosis exerted via intrinsic pathway. Moreover, these alkaloids showed selectivity to human prostate cancer cells. Of high clinical impact is the observed synergism with olaparib as well as the ability to overcome AR-V7-mediated drug resistance to androgen receptor targeting drugs. Our current research significantly contributes to the understanding of the mechanism of action and therapeutic potential of urupocidins and similar compounds.