Isoprenoid-phospholipid conjugates as potential therapeutic agents: Synthesis, characterization and antiproliferative studies

The aim of this research was to extend application field of isoprenoid compounds by their introduction into phospholipid structure as the transport vehicle. The series of novel isoprenoid phospholipids were synthesized in high yields (24–97%), their structures were fully characterized and its anticancer activity was investigated in vitro towards several cell lines of different origin. Most of synthesized compounds showed a significantly higher antiproliferative effect on tested cell lines than free terpene acids. The most active phosphatidylcholine analogue, containing 2,3-dihydro-3-vinylfarnesoic acids instead of fatty acids in both sn-1 and sn-2 position, inhibits the proliferation of colon cancer cells at 13.6 μM.


Introduction
In recent years, numerous epidemiological studies about the correlation between consumption of fruits, vegetables and other plant products and reduction of cancer incidence have been presented [1,2]. In the group of bioactive constituents of plants, isoprenoids are mentioned as one of the most important cancer-protective molecules [3]. They represent a broad class of mevalonate-derived phytochemicals, which exhibit many pharmacological and chemopreventive effects. They influence on the life cycle of the cells on the molecular level, activation of certain genes and apoptosis of tumor cells, what is the subject of extensive studies [4,5]. The results of the research from this area show that isoprenoid compounds inhibit proliferation of cancer cell lines, causing inhibition of cell division and transition of cells into the phase of programmed cell death. Limonen has documented antitumor activity against rodent mammary, liver, lung, stomach and skin cancers [6][7][8][9]. Chemopreventive activity of perillyl alcohol was also demonstrated against pancreatic [10] and prostate cancer cell lines [11]. For that compound even several clinical trials, including phase I and II trials have been recently conducted [12,13]. It was confirmed that geraniol has also high cytotoxic activity in vitro and in vivo against murine leukemia, hepatoma and melanoma cells [14,15], similar properties have been documented for farnesol [14,16,3]. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Replacement of fatty acids in phospholipids with therapeutic or desirable fatty acids improves the physico-chemical, nutritional and pharmaceutical functions. Phosphatidylcholine is particularly attractive lipid for this type of modification because of its health-promoting activity. It is a source of biologically active choline, which is a substrate in the synthesis of acetylcholine strictly responsible for many vital functions of organism such as breathing, heart rate or memory processes [17,18]. Phospholipids with defined molecular structure have been used in drug delivery and gene therapy [19][20][21].
Therefore it was of interest to synthesize the phospholipid molecules having isoprene acid as the acyl residue in the phospholipid backbone and to combine the valuable activities of these two groups of compounds. In our previous paper we have reported on the synthesis of phosphatidylcholine analogues containing geranic and citronellic acids in the sn-1 and sn-2 position [22] and present the results of their cytotoxic studies on leukemia, lung, breast, liver, colon and deoxorubicin-resistant colon cancer cell lines [22]. Encouraged by these results in the current paper we extended our earlier studies, preparing novel isoprenoid-phospholipid conjugates with several new isoprenoid acids. We report on their synthesis, characterization and anticancer activity trying to find relationship between some structural feature, i.e. the number of unsaturated bonds, the chain length, the cyclic or acyclic structure of isoprenoids and their antitumor activities. The synthesized molecules could be potential prodrugs or diet supplements.

Materials and methods General
HPLC chromatography was performed on an Ultimate 3000 Dionex chromatograph equipped with a DGP-3600A dual-pump fluid control module, a TCC-3200 thermostatted column compartment, and a WPS-3000 autosampler. A Corona charged aerosol detector (CAD) from ESA Biosciences was used, with the following parameters: acquisition range: 100 pA, digital filter set to none, N 2 pressure: 0.24 MPa. The system and data acquisition were carried out using the Chromeleon 6.80 software (Dionex Corporation). Analysis was carried out using a Betasil DIOL 5-mm column (Thermo, 150 2.1mm). The injection volume was 15 mL in all of the experiments and the cooling temperature for the samples was 20˚C. The column temperature was maintained at 30˚C. The total time of analysis was 19 min. The mobile phase had a constant flow of 1.5 mL min -1 . Solvent A (1% HCOOH, 0.1% triethylamine (TEA) in water), solvent B (hexane), and solvent C (propan-2-ol) were used in gradient mode starting with 3:40:57 (A:B:C (vol-%/ vol-%/vol-%)), at 4 min 10:40:50, at 9 min 10:40:50, at 9.1 min 3:40:57 and at 19 min 43:40:57.
Column chromatography was performed on silica gel (Kieselgel 60, 230-400 mesh, Merck). Spectroscopic measurements were carried out on a Bruker Avance II 600 MHz spectrometer. Chemical shifts are given in ppm downfield from tetramethylsilane (TMS) as the internal standard. In 31 P NMR spectra, chemical shifts were referenced to 85% H 3 PO 4 as a standard. Coupling constant (J) values are given in Hertz. High-resolution mass spectra (HRMS) were recorded using electron spray ionization (ESI) technique on Waters ESI-Q-TOF Premier XE spectrometer.
The series of new isoprenoid phospholipids (6a-6d; 7a-7d; 8a-8d; 9a-9d) were obtained according to the methods described earlier by Gliszczyńska et al. [22]. The yields of the reaction, their physical and spectroscopic data are presented below:

Biological studies
The biological studies were performed in vitro using human cancer cell lines: MV4-11 (human biphenotypic B myelomonocytic leukaemia), A-549 (non-small cell lung cancer), MCF-7 (breast cancer), LoVo (colon cancer), LoVo/DX (colon cancer drug resistant), HepG2 (liver cancer) and BALB/3T3 (normal mice fibroblasts). These lines were obtained from American Type Culture Collection (Rockville, Maryland, U.S.A.). The cell line is being maintained in the Institute of Immunology and Experimental Therapy, Wroclaw, Poland. All cell lines were grown at 37˚C with 5% CO 2 humidified atmosphere and were cultured in cultured medium according to the method described before [25].
Prior to usage, the compounds were dissolved in DMSO or in mixture of 99.8% ethanol and DMSO (1:1) to the concentration of 25 or 50 mM, and subsequently diluted in culture medium to reach the required concentrations (ranging from 5 to 625 μM).
Twenty-four hours prior to the addition of tested compounds, the cells were plated in 96-well plates (Sarstedt, Germany) at a density of 1 × 10 4 cells per well in 100 μL of culture medium. The assay was performed after 72 h of exposure to varying concentrations of the tested agents. The in vitro cytotoxic effect of tested agents was examined using the MTT (for MV4-11 cells) or SRB assay which were described previously [25]. Each compound in each concentration was tested in triplicate in a single experiment, which was repeated 3-5 times.
The results of antiproliferative activity were given as IC 50 values. Using this parameter we calculated also the resistance indexes (RI) dividing the IC 50 values of selected compounds estimated for drug resistant cell LOVO/DX line by respective values designated for drug sensitive LoVo line. According to Harker et al. [26] three categories of the cells could be distinguished: (a) the cells are drug-sensitive-if the ratio approaches 0-2; (b) the cells are moderately drugresistant-if the ratio ranges from 2 to 10; (c) the cells are markedly drug-resistant-if the ratio is higher than 10.
Statistical analysis. Statistical analysis was performed using STATISTICA version 10 (StatSoft Inc., USA). Mann-Whitney U Test was used in the analysis and the results in Tables 1  and 2 are given with statistical significance in comparison to free isoprenoid acids p < 0.05.

Results and discussion
Fundamental obstacle in pharmaceutical applications of therapeutic compounds are often their low bioavailability, incomplete absorption or too rapid absorption and too fast excretion as well as difficulties with preparing its appropriate formulations. Good strategy to solve these types of problems is designing the lipidic prodrugs. In this way the biologically active molecules can be covalently bound to the transport moiety and then release in vivo providing the effective treatment after oral administration. The transport molecules can be fatty acids, triglycerides or phospholipids. The simplest method of lipophilization is the esterification of  as the health promoting molecules. Recently we synthesized novel phosphatidylcholines containing isoprenoid compounds in the sn-1 and sn-2 positions of PC [22]. This strategy based on the introducing a drug molecule instead of fatty acids has been investigated so far by only a few research groups [33,39,40].
In our previous paper we reported on the higher activities of geranic and citronellic acids introduced in apolar part of PC towards selected cancer cell lines [22]. For better understanding the effect of conjugation of phospholipids with isoprenoid acids on the antiproliferative activity in living cells we designed and prepared more terpene phospholipids derivatives.
Acids were obtained according to known two-step synthesis from the corresponding alcohols: geraniol, farnesol, β-dihydrodamascol and phytol. Synthesis of acids 2 and 3 from farnesol and β-dihydrodamascol was described in the literature [23,24], the same procedure was applied for the synthesis of known [41] acid 1 from geraniol while the structure of ester (4) and acid (5) obtained from natural compound phytol is first time reported. The main step of synthesis of isoprenoid acids was the Johnson-Claisen rearrangement of allyl alcohols. Then formed esters were hydrolyzed in ethanolic solution of KOH to give corresponding acids (Fig 2).
Four groups of terpene-modified phosphatidylcholines were synthesized (Fig 3). We have started from the synthesis of symmetrical phosphatidylcholines with the same acyl residues in both sn-1 and sn-2 positions (6a-6d). For this purpose we used the cadmium salt of sn-glycero-3-phosphocholine (GPC×CdCl 2 ) as a starting material and series of isoprenoid acids previously prepared previously, following the procedure published before [22]. Novel 1,2-diisoprenoyl-sn-glycero-3-phosphocholines (6a-6d) were obtained in good 62%-97% yields after 72 hours of reaction. The asymmetrically substituted terpene-phospholipids (7a-7d) were obtained by the esterification of the hydroxy group of 1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine with isoprenoid acids (1-3, 5). A detailed description of this reaction conditions has been presented previously [22]. The 1-pamitoyl-2-isoprenoyl-sn-glycero-3-phosphocholines (7a-7d) has not been reported in the literature and was obtained in 41%-58% yields. Lysophosphatidylcholines with isoprenoid acids in the sn-1 position were synthesized from GPC, which was first transformed into cyclic stannylene ketal by treatment with DBTO and then selectively acylated with isoprenoyl chlorides. According to this methodology four new 1-isoprenoyl-2-hydroxy-sn-glycero-3-phosphocholines (8a-8d) in 24%-52% yields were obtained. The last synthetic path involved the key intermediates 8a-8d in acylation of free hydroxy groups in the sn-2 position with palmitic acid (PA) in the presence of DCC and DMAP. The final products 9a-9d were obtained in 47-68% yields. The structures of the final products were confirmed by their spectroscopic data and the detailed assignments of signals are given in the supplementary material.
All synthesized compounds 6a-6d, 7a-7d, 8a-8d and 9a-9d were evaluated for their antiproliferative activity. The preliminary biological studies were performed towards human leukemia cell line (MV4-11). The results are summarized in Table 1 and are expressed as the IC 50 -concentration of the compounds (in μM) that inhibits proliferation of the cells by 50% compared to the untreated control cells. The clinical chemotherapy drug cisplatin was used in this evaluation as a positive control. In order to observe the expected increase of activity of isoprenoids after their conjugation with phospholipid molecule free terpene acids were also tested. On the other hand trying to estimate the benefits resulting from the connection of two valuable groups of compounds the cytotoxic effect of palmitic acid and phosphatidylcholine with this fatty acid in both sn-1 and sn-2 positions as well as lysophosphatidylcholine with palmitic acid only in sn-1 position were also evaluated.
Analyzing the results presented in Table 1 significant increase of antiproliferative activity of monoterpene (1) and sesquiterpene (2, 3) acids derivatives attached to phospholipid moiety can be observed. The activity of isoprenoid phospholipids 6a-6d, 7a-7d, 8a-8d and 9a-9d is generally from 3 to 8-fold higer towards MV4-11 compared to free terpene acids. These results suggest that the phospholipid moiety increases inhibition of the proliferation of cancer cells by isoprenoid acids. That anticancer effect is also related to the presence and position of isoprene moiety in the phospholipid molecule. The best results for sesquiterpene acyl derivatives are observed for diisoprenoyl-PC (entries 10, 15) whereas for derivatives of monoterpene acid 1-isoprenoyl-2-palmitoyl-PC (entry 5) exhibits the highest activity. The highest activity towards leukemia cells was achieved for derivative 6b when 2,3-dihydro-3-vinylfarnesoic acids was directly connected to glycerol skeleton of PC in both sn-1 and sn-2 positions (IC 50 = 27.78 μM). Diterpene phytol derivative 5 does not exhibit higher cytotoxic effect on leukemia cells after introduction to sn-glycero-3-phosphocholine and from that reason it was not selected to further studies.
In the second step of evaluation studies we tested the selectivity of phospholipid derivatives of mono-and sesquiterpene acids 6a-6c, 7a-7c, 8a-8c and 9a-9c towards human cancer cell lines with different origin. The experiments were also carried out towards normal mice fibroblasts BALB/3T3 and synthesized compounds were evaluated at four concentration levels 5, 25, 125 and 625 μM. As it is shown in the Table 2 most of investigated terpene acids are more active after their introduction into GPC. Consequently, diisoprenoyl-PC of sesquiterpene acid derivatives (2, 3) and monoisoprenoyl-PC of monoterpene acid derivatives (1) turned out to be the most active compounds towards tested cancer cell lines. The compounds 6b and 6c that contained 2,3-dihydro-3-vinylfarnesoic and 2-(2-butylidene-1,3,3-trimethylcyclohexyl)acetic acids in both sn-1 and sn-2 positions of PC were more active than the corresponding derivatives substituted with other acyl donors. Compound 6b showed the highest cytotoxic effect against lung (A549), colon (LoVo) and liver (HepG2) cancer cell lines among all tested derivatives (IC 50 = 13.6-57.8 μM). It is worth noting that 6b derivative the highest activity against colon cancer cell line (IC 50 = 13.6 μM) is only 5-fold lower than those exhibited by cisplatin. Apart from that compound 6b had a lower cytotoxicity against the normal mice fibroblasts cell BALB/3T3 (57.8 μM) than cisplatin. The cells of breast cancer (MCF-7) and drug resistant colon cancer (LoVo/DX) were the most sensitive on the phosphatidylcholine 9a contained 2,3-dihydro-3-vinylgeranic acid in sn-1 position (IC 50 = 40.0 and 52.7 μM respectively). No significant differences in antiproliferative activity against normal and cancer cell lines was found for tested derivatives however for the most active compounds 9a and 6b the doses active towards leukemia, lung, breast and colon cancer cell lines are lower in comparison to cytotoxic doses active against BALB/3T3.
IC 50 values for newly synthesized terpene phospholipids were compared with the results previously reported for phospholipid derivatives of geranic and citronellic acids [22]. Based on these data we tried to determine the structure-activity relationship for prepared compounds. Particularly it can be noticed that the antitumor activity depends on the length of the isoprenoid moiety attached to PC. Whereas the mono-and sesquiterpene derivatives are active towards selected cancer cell lines, incorporation diterpene phytol derivative into PC does not increase its biological activity. In addition, it can be observed that linear terpene acids joined to PC are more active than the cyclic one. The number of unsaturated bounds in the structure of isoprenoid acyl donors also affects the activity of studied products. The derivatives of geranic and 2,3-dihydro-3-vinylgeranic acids are more active than derivatives of citronellic acid. However the differences in the antiproliferative activity between isoprenoids with three and four double bounds seem to have selective effect depending on the type of cancer cell line.
Additionally we calculated also the resisitance indexes (RI) for all tested compounds and we found that almost all of them are able to overcome the barrier of P-gp-dependent resistance with exception of three derivatives 7a, 6b and 7c (Table 3).
In summary, a series of sixteen new isoprenoid phospholipid conjugates containing isoprenoid acids were synthesized and their antiproliferative activity towards selected cancer cell lines was investigated. Our results clearly showed that incorporation isoprenoids acids with phosphatidylcholine significantly improves its selective cytotoxic activity. Short chain isoprenoid acids like geranic, citrolellic or 3,7-dimethyl-3-vinyloct-6-enoic are more active in the form of monoisoprenoilo-PC whereas 2,3-dihydro-3-vinylfarnesoic and 2-(2-butylidene-1,3,3-trimethylcyclohexyl)acetic acids are active after their incorporation in both sn-1 and sn-2 position of PC. However, studying the structure-activity relationship for isoprenoid-phospholipids it is nessecery to prepare more derivatives. At this stage promising results of in vitro activity qualify 6b, 6c and 9a as anticancer agents for further in vivo studies and enhance investigation of their mechanism of action.