Converting bile acids into mitocans

A B S T R A C T Cholic acid ( 1 , CD), deoxycholic ( 3 , DCA), chenodeoxycholic acid ( 5 , CDCA), ursodeoxycholic acid ( 7 , UDCA), and lithocholic acid ( 9 , LCA) were acetylated and converted into their piperazinyl spacered rhodamine B conjugates 16 – 20 . While the parent bile acids showed almost no cytotoxic effects for several human tumor cell lines, the piperazinyl amides were cytostatic but an even superior effect was observed for the rhodamine B conjugates. Extra staining experiments showed these compounds as mitocans; they led to a cell arrest in the G1 phase.

Bile acids are natural detergents. They facilitate the absorption of fat in the intestine but they are also essential in the maintenance of the intestine epithelium homeostasis. Depending on type and concentration they show a dual behavior both as pro-survival or pro-death molecules [16][17][18]. Recently, cytotoxic activity [19] was established for bile acidpaclitaxel hybrids [20], a camptothecin conjugate [21] as well as for a dihydroartemisinine [22] analog. Furthermore, some bile carboxylamides have been shown to exert pro-apoptotic effects in human colon adenocarcinoma cells DLD-1, HCT-116 and HT29. Pro-apoptotic activity has also been observed on multiple myeloma as well as on glioblastoma multiforme [23]. We have previously demonstrated the cytotoxic activity of triterpenoid-rhodamine B conjugates [4]. It was therefore reasonable to extend our investigations to bile acids, especially since these compounds show a priori a better solubility in biological systems than triterpenoids by comparison due to their amphiphilic nature. In addition, bile acids have emerged as important starting materials for a variety of different bioactive conjugates [24,25].
Finally, the desired rhodamine B conjugates were easily synthesized employing the well-established EDC . HCl / HOBt method to afford the conjugates in 16-20 in 47 %-66 % isolated yields, respectively.

Biology
The cytotoxicity of the synthesized compounds was evaluated using photometric SRB assays employing several human tumor cell lines (cutoff at 30 µM; Table 1), and the results of which are summarized in Table 1. While the parent bile acids 2, 4, 6 and 8 as well as rhodamine B (Rhd B) showed no activity towards the cell lines used in the SRB-assay, their acetates held increased biological activity towards the cell lines A375, MCF7, and A2780. Compound 10 proved to be insoluble under the conditions of the assay. At first glance, this seems surprising since compounds 2, 4, 6 and 8 were readily soluble. However, the poor solubility of 10 also follows the solubility behavior reported for the unsubstituted GAs. Here, too, LCA is the most poorly soluble with 0.38 mg/L while CA, for example, has a solubility of 175 mg/L in water. [26] For the piperazinyl amides 12 and 13, however, even lower EC 50 values were observed. The observation that piperazinyl amides hold lower EC 50 values than their corresponding carboxylic acids parallels earlier results having been observed also in the field of terpenoid, in particular of triterpenoid carboxylic acids [1,2,27]. Finally, for the rhodamine B conjugates 16-20 and the malignant cell lines very low EC 50 -values in the range of 0.2-1.2 µM were measured. These conjugates, however, showed only low selectivity for the non-malignant cell line NIH 3 T3 with 16 being the best overall compound holding a tumor cell/nontumor cell selectivity ranging from S = 1.33 -3.27. The best selectivity for this series of compounds was measured for the breast adenocarcinoma cell line MCF-7; this again, parallels previous finding for similar steroid conjugates [28].
To further investigate the mode of action of the piperazinyl amides 12 and 13, as well as of the rhodamine B conjugate 16, annexin-V-FITC/ PI staining assays were used to assess the triggered mode of cell death onto the tumor cell line A2780 at double the EC 50 concentrations (Fig. 2). Thereby, the samples (sixfold sample repetitions) were incubated for 24 h and 48 h with and without added compounds. Two technical repetitions in reference to the control group were used for calculation. Interestingly, compound 16 showed no significant difference from the control group (Table 2). In contrast to this compound, compounds 12 and 13 showed a significantly lower number of necrotic and vital cells and, in addition, they also held significantly more apoptotic cells at 24 h incubation. These trends continued after 48 h of incubation with compound 12 being the best with an average of 34 % more apoptotic and − 26 % vital cells in comparison to the control cell line.
Furthermore, cell cycles were analyzed by FACS employing the ovarian cancer cell line A2780 applying an incubation time of 24 h and 48 h, respectively. Analysis of the results from these experiments showed compounds 12 and 16 to lock treated cells in the G1-phase and a decreased number of cells was found in the S-phase at 24 h incubation ( Fig. 3; Fig. 4).
This parallels most recent findings for 12; R. Yang et al. have previously shown that this compound arrests HepG2 hepatoma cells in G0/ G1 and induces apoptosis by the PI3K/AKT/mTor pathway [29]. Compound 12 also induced nearly 39 % of apoptosis as compared to the control cell line (Fig. 3). After an incubation time of 48 h, cells treated with 12 showed 51.38 % apoptotic cells, while the number of cells in the G1 phase had decreased (Fig. 4). The increased cytostatic effect of compound 16 seems to be the result of fewer cells being able to enter the S and G2/M phase.
To assess whether 16 acts as a mitocan (an acrynom describing compounds exerting their anti-cancer activity via their molecular targets within mitochondria), some extra staining experiments were performed, the results of which are depicted in Fig. 5. The dye rhodamine 123 is known to stain mitochondria specifically; this dye emits green light after excitation; compound 16 emits red light.
Consequently, a merged image of the two excitations would cause an observed orange color, and gives evidence whether 16 is accumulated in the mitochondria of A2780 cells (Fig. 5). A microscopic investigation indeed showed a good match of the rhodamine 123 dye with 16, and an orange color was observed. Moreover, staining with Hoechst 33,342, a blue-emitting nucleus targeting dye, showed 16 not to enter the nucleus of the cancer cells.

Conclusion
Several bile acids, i.e. cholic acid (1), deoxycholic (3), chenodeoxycholic acid (5), ursodeoxycholic acid (7), and lithocholic acid (9) were converted into their acetylated piperazinyl amides. The latter were coupled with rhodamine B to yield conjugates. These conjugates were cytostatic for a panel of human tumor cell lines; they led to a cell arrest in G1, and are accumulated in the mitochondria of the tumor cells. The conjugates do not enter the nucleus. Albeit their cytostatic effect is lower than that of pentacyclic triterpenoid analogs, they represent interesting starting materials for the development of analogs of even higher cytotoxicity and improved tumor cell/non-malignant cell selectivity while still retaining their good solubility properties.

Methods and equipment
Cholic (1), deoxycholic (2), chenodeoxycholic (3), ursodeoxycholic (4), and lithocholic acid (5) were obtained from Carl Roth and abcr GmbH and were used as received. Equipment and lab equipment was used as previously described. Details can be found in the Supplementary materials file. For the Annexin V-FITC/PI assay as well as for the cell cycle analysis the Attune® Cytometric Software (1.2.5) and MSExcel were used for calculations.

General procedure (GP1) for the acetylation of bile acids
The bile acid (1-5, 1 eq) and cat. amounts of DMAP were dissolved in a minimal amount of dry pyridine (20 mL), and acetic anhydride (3-7 eq, depending on the number of hydroxyl groups) was added. After stirring at 25 • C for 24 h, the solution was diluted with DCM (100 mL), washed with HCl (0.1 M, 50 mL), and water (2 × 100 mL). The organic phase was dried (MgSO 4 ), the solvent evaporated under reduced pressure, and the residue was purified by column chromatography (SiO 2 , nhexane/ethyl acetate).

General procedure (GP2) for the synthesis of acetylated piperazinyl amides
The acetylated bile acid (2, 4, 6, 8, 10, 1 eq) was dissolved in a minimal amount of dry DCM (10 mL) under argon and cooled to 0 • C. Cat. amount of dry DMF and oxalyl chloride (4 eq) were added, and the reaction was allowed to warm to 25 • C. After stirring for an additional 2 h, the volatiles were removed under reduced pressure. To a solution containing piperazine (3.8 eq) and triethylamine (1.1 eq) in a minimal amount of dry DCM at -21 • C under argon was added dropwise the acid chloride, dissolved in dry DCM (20 mL). The reaction mixture was stirred at 25 • C for 24 h, quenched with water (100 mL), extracted with DCM (3 × 100 mL), the combined organic phases were dried (MgSO 4 ), and the solvent was evaporated under reduced pressure to obtain a solid which was purified by column chromatography (SiO 2 , CHCl 3 /MeOH, 9:1).