Modification of uptake and subcellular distribution of doxorubicin by N-acylhydrazone residues as visualised by intrinsic fluorescence.

This is a preor post-print of an article published in Effenberger-Neidnicht, K., Breyer, S., Mahal, K., Sasse, F., Schobert, R. Modification of uptake and subcellular distribution of doxorubicin by N-acylhydrazone residues as visualised by intrinsic fluorescence (2011) Cancer Chemotherapy and Pharmacology, pp. 1-6. Article in Press. Modification of uptake and subcellular distribution of doxorubicin by N-acylhydrazone residues as visualised by intrinsic fluorescence Katharina Effenberger-Neidnicht • Sandra Breyer • Katharina Mahal • Florenz Sasse • Rainer Schobert Correspondence to R. Schobert, Organic Chemistry Laboratory, University Bayreuth, Universitaetsstrasse 30, NW 1, 95447 Bayreuth, Germany e-mail: [email protected] fax: +49(0)921 552671 F. Sasse, Helmholtz Centre for Infection Research (HZI), Department of Chemical Biology, Inhoffenstrasse 7, 38124 Braunschweig, Germany Abstract Purpose: Doxorubicin (1) is commonly used in the treatment of a wide range of cancers. Some N-acylhydrazones of 1 were previously found to have an improved tumour and organ selectivity. In order to clarify the molecular basis for this effect, the cellular uptake into various cancer cells and the localisation in PtK2 potoroo kidney cells of 1 and its N-acylhydrazones derived from heptadecanoic acid (2) and 11-(menthoxycarbonyl)undecanoic acid (3) were studied drawing on their intrinsic fluorescence. Methods: The uptake of compounds 1–3 into human cells of HL-60 leukaemia, 518A2 melanoma, HT-29 colon, and resistant KB-V1/Vbl and MCF-7/Topo breast carcinomas was determined fluorimetrically from their residual amounts in the supernatant. Their time-dependent accumulation in PtK2 potoroo kidney cells was visualised by fluorescence microscopy. 1 Results: The uptake, though not the cytotoxicity, of 2 in multi-drug resistant MCF-7/Topo breast cancer cells was conspicuously greater than that of 1 and 3, probably due to an attractive lipophilic interaction with the lipid-rich membranes of these cells. In non-malignant PtK2 cells both 1 and 3 accumulated initially in the nuclei. Upon prolonged incubation their fluorescent metabolites were visualised in lysosomes neighbouring the nuclei. In contrast, conjugate 2 was not observed in the nuclei at any time. After two hours it had accumulated in vesicles scattered all over the cells and upon prolonged incubation its fluorescent metabolites were concentrated in the cellular membrane. Conclusions: Long unbranched fatty acyl residues when attached to doxorubicin via a hydrazone can act as lipophilic membrane anchors. This allows an increased uptake of such derivatives into lipid-rich membranes especially of multi-drug resistant cancer cells, a retarded release from there into the cytosol and the eventual storage of their metabolites again in the cell membrane rather than in lysosomes. Keywords Doxorubicin • Fluorescence spectroscopy • PtK2 cells • Conjugates • Membrane lipids • Fatty acids Conflicts or disclosures: NONE 2 Introduction Doxorubicin (1) – a Streptomyces metabolite – is used in chemotherapy against a wide range of cancers such as haematological malignancies, soft tissue sarcomas, lymphomas and various types of carcinomas despite its clinical limitations such as cardiotoxicity and induction of multi-drug resistance [1–6]. In earlier studies, we had shown that some N-acylhydrazones of 1 derived from long-chain fatty acids or terpenes have a higher tumour and organ selectivity than 1 while sharing with it the same mechanism of apoptosis induction, characterised by elevated levels of reactive oxygen species and apoptosis-relevant caspases and a loss of the mitochondrial membrane potential. Some of these new hydrazone conjugates not only had a greater selectivity than the parent drug 1 [7, 8], but were also less good substrates for the ABC-transporters of multi-drug resistant cancer cells. In order to clarify the molecular basis for these effects we now studied the cellular uptake of 1 and its N-acylhydrazones derived from heptadecanoic acid (2) and 11- (menthoxycarbonyl)undecanoic acid (3) by various cancer cells and their localisation in flat, easy to visualise PtK2 potoroo (Potorous tridactylis) kidney epithelial cells by means of their intrinsic fluorescence (Fig. 1). ((Figure 1 here)) Materials and methods General The hydrazones 2 and 3 were prepared as described previously [7]. Fluorescence spectra were recorded on an FP-6500 fluorescence spectrophotometer (JASCO, Tokyo, Japan) between 500 and 700 nm. Uptake of the derivatives was measured using a CM Infinite F200 fluorescence plate reader with excitation and emission wavelengths of 485/20 and 590/20 nm (TECAN, Crailsheim, Germany). 3 Fluorescence analyses of treated cells were conducted with an Axioplan fluorescence microscope equipped with an Axiocam camera (ZEISS, Jena, Germany) and evaluated with the software AxioVision 3.1. Cell lines and culture conditions Leukaemia HL-60 cells were obtained from the German Collection of Biological Material (DSMZ), Braunschweig (Germany), melanoma 518A2 cells from the Department of Oncology and Hematology of the Martin Luther University, HalleWittenberg (Germany), KB-V1/Vbl cervix and MCF-7/Topo breast carcinoma cells from the Institute of Pharmacy of the University Regensburg (Germany), and HT-29 colon carcinoma cells as well as human foreskin fibroblasts (HF) from the University Hospital Erlangen (Germany). The HL-60 and HT-29 cells were grown in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS), 100 IU mL-1 penicillin G, 100 μg mL-1 streptomycin sulfate, 0.25 μg mL-1 amphotericin B and 250 μg mL-1 gentamycine (all GIBCO). The 518A2, the HF and the KBV1/Vbl cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, GIBCO), containing 10% FCS, 100 IU mL-1 penicillin G, 100 µg mL-1 streptomycin sulfate, 0,25 µg mL-1 amphotericin B and 250 µg mL-1 gentamycine. The MCF-7/Topo cells were grown in Eagle’s Minimal Essential Medium with Earle´s salts (MEM; SIGMA) supplemented with 2.2 g L-1 NaHCO3, 110 mg L-1 sodium pyruvate and 5% FCS. The potoroo kidney cells PtK2 were obtained from the American Type Culture Collection (ATCC) and cultivated in MEM (GIBCO) supplemented with 10% FCS and non-essential amino acids (GIBCO). They were maintained in a moisture-saturated atmosphere (5% CO2) at 37°C in 50-mL culture flasks (NUNC, Germany), and serially passaged after tripsinisation. 4 Fluorescence spectroscopy A JASCO FP-6500 fluorescence spectrophotometer was used for fluorescence measurements of compounds 1–3 dissolved in phosphate buffered saline (5 µM). The excitation wavelength was 490 nm, emission wavelengths ranged from 500 to 650 nm. Cellular drug uptake The cellular uptake of the test compounds was ascertained using a TECAN CM Infinite F200 fluorescence plate reader (Exc.: 485/25 nm, Em.: 590/20 nm). Typically, 500 µL cells of HL-60 leukaemia (5  105 mL-1), HT-29 colon carcinoma, 518A2 melanoma, KB-V1/Vbl cervix carcinoma and MCF-7/Topo breast carcinoma (each 5  104 mL-1) were incubated in 24-well plates with the compounds 1, 2 or 3 (5 µM) for 3 h. The medium was removed by centrifugation for 5 min at 150 g. The residual fluorescence intensity was measured in the supernatant and the corresponding concentration of test compounds taken up by the cells was calculated by comparison with identically treated references [9]. Statistical analysis of measurement results The results are expressed as means ± standard deviation (SD). The StudentNewman-Keuls test was used to determine statistical significance with P value < 0.05 considered significant (*). Cellular localisation The distribution of the test compounds in PtK2 cells was visualised via fluorescence microscopy. Typically, the cells were grown in DMEM (750 µL) in 4-well plates (Nunc) on glass coverslips and incubated with the test compounds 1, 2 or 3 (10 µM) for periods ranging from 2 h to 16 h. The medium was removed, 5 the cells were washed twice with phosphate buffered saline (PBS) and mounted vital in PBS. The Hoechst dye 33342 was optionally added (5 µg mL-1) to stain the nuclei and the cells were imaged using a ZEISS Axioplan fluorescence microscope [9, 10]. Results Fluorescence spectroscopy Fluorescence spectroscopy has been frequently employed to study the interaction of doxorubicin (1), e.g., with DNA in the course of intercalation [11, 12] or with lipid membranes [13]. We now measured the fluorescence spectra of the test compounds 1–3 in phosphate buffered saline (PBS; 5 µM) at pH 7.4. Since the fluorophore of 1 has a known absorption maximum between 480 nm and 500 nm [14] we excited the test compounds at 490 nm and observed three congruent fluorescence spectra (Fig. 2). They are each characterised by an overlay of three bands resulting in an envelope curve with a maximum around 550 nm and a shoulder at ca. 580 nm. However, the intensities of the spectra of the hydrazone conjugates 2 and 3 were only about half of that of 1 itself. This is qualitatively in keeping with a report by Chourpa et al. on the changes of the intrinsic fluorescence of 1 upon addition of sodium oleate [15]. The authors explained this pHand ratio-dependent effect by lipophilic interactions of the aromatic fluorophore of 1 with the long alkenyl chain of oleic acid. A similar and obviously more intense lipophilic interaction could result from a backfolding of the covalently attached heptadecanoyl chain of 2 or of the tethered menthyl residue of 3, respectively, on top of the anthraquinone fluorophore. ((Figure 2 here)) Cellular drug uptake 6 Next, we ascertained the uptake of 1 and its derivatives 2 and 3 by five human cancer cell lines of entities that are typically treated with 1, namely leukaemia, melanoma as well as breast, cervix and colon carcinomas. The cells were incubated with 5 µM of 1–3 for 3 h and the residual fluorescence intensity in the supernatant following incubation was measured after a centrifugation step. The percental cellular drug uptake relative to reference solutions was calculated and plotted as shown in Figure 3 [9]. All test compounds accumulated strongly in the 518A2 and the HL-60 cells which corresponds well with their previously reported high cytotoxicity against these cell lines [7]. In contrast, their accumulation in the remaining three multi-drug resistant cancer cell lines was comparatively low, which fact matches their low cytotoxicity in these cells. However, fatty acid conjugate 2 is an exception since it was selectively and significantly (> 20%) accumulated in multi-drug resistant MCF-7/Topo breast carcinoma cells. ((Figure 3 here)) Cellular localisation Finally, to analyse the cellular localisation of 1 and its derivatives 2 and 3, nonmalignant PtK2 potoroo kidney cells were incubated with 10 µM concentrations of them for up to 16 h, then mounted vital in PBS and imaged using fluorescence microscopy. The results of these autofluorescence analyses are shown in Figures 4 and 5. ((Figure 4 here)) While after 2 h of incubation 1 was localised only in PtK2 nuclei (Fig. 4 / A, B), upon prolonged exposure it or its fluorescent metabolites were accumulated in vesicles which were initially neighbouring the nuclei (Fig. 4 / C, D) but later were distributed throughout the cells (Fig. 4 / E, F). This is in line with the previously 7 reported uptake and subcellular distribution of 1 in other cells, e.g., rat embryo fibroblasts and human myeloid leukaemia cells [9, 10]. For the menthylterminated hydrazone 3 we found a subcellular localisation similar to that of 1, albeit proceeding more slowly. The accumulation in the nuclei required twice as long (4 h) for completion as in the case of the parent drug. However, again some differences emerged for conjugate 2. Here, no drug was found in PtK2 nuclei at any time but in vesicles throughout the cell yet after incubation periods as short as 2 hours (Fig. 5 / A, B). This was confirmed by co-localisation experiments with the nuclei-staining Hoechst dye 33342. Upon prolonged incubation 2, or its fluorescent metabolites, penetrated into the cellular membrane giving rise to intensive fluorescence there (Fig. 5 / C, D). We assume that the long linear fatty acyl residue acts as a lipophilic anchor thus immobilising the conjugate or a metabolite of it at the membrane. This would not be possible for 1 or 2 which lack such a straight lipophilic tether. ((Figure 5 here)) Discussion The cell membrane of resistant cancer cells is rich in lipids with saturated hydrophobic acyl chains, neutral and sphingomyelin lipids, which can attractively interact with doxorubicin [16]. Very recently, Labhasetwar et al [17] reported that doxorubicin penetrates into the highly organised lipid monolayers of resistant MCF-7 breast cancer cell membranes and is retained there. This strong lipid-drug interaction is also believed to reduce the ability of the drug to diffuse across the cell membrane and into the cytosol, and to enhance the likelihood of near-by ABC-efflux transporters expelling the drug [18]. The fatty acyl conjugate 2 can intertwine even more effectively than 1 with the membrane lipids of resistant 8