Ruthenium-based complex nanocarriers for cancer therapy

Biomaterials 33 (2012) 3770e3782

Contents lists available at SciVerse ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Ruthenium-based complex nanocarriers for cancer therapy Gaetano Mangiapia a, b, Gerardino D’Errico a, b, Luca Simeone a, Carlo Irace c, Aurel Radulescu d, Antonio Di Pascale c, Alfredo Colonna c, Daniela Montesarchio a, **, Luigi Paduano a, b, * a

Dipartimento di Scienze Chimiche, Università degli Studi di Napoli “Federico II”, Complesso Universitario di Monte S. Angelo, via Cintia, 80126 Naples, Italy CSGI e Consorzio interuniversitario per lo sviluppo dei Sistemi a Grande Interfase, Italy c Dipartimento di Farmacologia Sperimentale, Università degli Studi di Napoli “Federico II”, via D. Montesano 49, 80131 Naples, Italy d Jülich Centre for Neutron Science, Lichtenbergstraße 1, 85747 Garching bei München, Germany b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 January 2012 Accepted 31 January 2012 Available online 21 February 2012

A new organometallic ruthenium complex, named AziRu, along with three amphiphilic nucleoside-based ruthenium complexes, ToThyRu, HoThyRu and DoHuRu, incorporating AziRu in their skeleton, have been synthesized, stabilized in POPC phospholipid formulations and studied for their antineoplastic activity. Self-aggregation behavior of these complexes was investigated, showing that the three synthesized AziRu derivatives able to form liposomes and, under specific conditions, elongated micelles. The formulations prepared in POPC proved to be stable for months and showed high in vitro antiproliferative activity. The here described results open new scenarios in the design of innovative transition metalbased supramolecular systems for anticancer drugs vectorization. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Ruthenium-based complexes Antineoplastic agents Nanoaggregates Anti-proliferative activity

1. Introduction In the last decades, research has proposed a growing number of antineoplastic agents. Among these, transition metal-based complexes represent, nowadays, a very important class of chemotherapeutics, intensively used for clinical treatments. The antineoplastic activity of the most important of these, Cisplatin, was discovered in 1969 [1] and in thirty years this complex has become one of the most used drugs in the treatment of some tumoral diseases, such as testicular, breast, uterine and ovarian cancers [2]. Since ever, some analogues of Cisplatin have been tested and then approved as drugs, such as Carboplatin and Oxaliplatin [3]. In addition to extremely high and selective anti-proliferative activity, novel anticancer drugs should exhibit specific efficacy toward the formation and growth of metastases. In fact, many tumors can develop metastases already very extended at the diagnosis time, making scarcely effective the surgical treatment [4]. If the primary tumor can be surgically removed, the pharmacological therapy seems to be the best choice for the metastases treatment, because of their non specific localization in the body [5]. On the other hand, unlike the primary tumor cells, metastases are * Corresponding author. Dipartimento di Scienze Chimiche, Università degli Studi di Napoli “Federico II”, Complesso Universitario di Monte S. Angelo, via Cintia, 80126 Naples, Italy. Tel.: þ39 081 674250; fax: þ39 081 674090. ** Corresponding author. Tel.: þ39 081 674126; fax: þ39 081 674393. E-mail addresses: [email protected] (D. Montesarchio), [email protected] (L. Paduano). 0142-9612/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2012.01.057

not responsive to several chemotherapeutics, probably because of their different proliferation kinetics [6,7]. These limitations have prompted researchers operating in this field to develop more effective and less toxic metal-based anticancer agents. In this context, ruthenium complexes have attracted much interest as a promising alternative to platinum, showing a remarkable antitumoral and antimetastatic activity, associated with lower toxicity. Among Ru-based complexes, the most promising compounds have been found and studied by Sava and co-workers [8e13]. Since the early 90’s, Sava has been a pioneer in studying transition-metal complexes, developing, among others, the complex named NAMI-A, endowed with relevant anticancer activity. This compound, along with KP1019, has gone over the phase I clinical trial with good outcome [14,15]. Despite these encouraging results, a possible drawback of Ru complexes has also been highlighted. In fact, it has been shown that, under physiological conditions, the chloride ligands of the Ru complexes are replaced by hydroxide ligands in relatively few hours. This leads to partial hydrolysis of the complex and to poly-oxo species formation [16,17]. Although it is claimed that the formation of poly-oxo species does not really alter the Ru complexes anticancer activity [18], it is to consider that, generally speaking, the premature aquation and hydrolysis of other anticancer drugs, as in the case of Cisplatin, can deactivate or activate too early most of the administered complex [19,20]. Thus, the design of longlife Ru-based antineoplastc agents is still a mandatory goal. A large number of alternative Ru-based complexes have been published in the recent years [21]. In this frame, some of us recently

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

proposed a new concept for Ru-based anticancer therapy, centered on the design of amphiphilic nanovectors carrying ruthenium complexes [22,23]. As widely known, self-assembled amphiphiles allow an efficient bottom-up strategy in order to obtain nanosized aggregates whose size and shape are quite easily tunable. Further benefits related to the use of nanostructures containing Ru complexes concern their capability: i) to transport a larger amount of the metal inside the blood stream; ii) to be “stealth” to the human immune system, specifically increasing the complex permanence time in the blood; iii) to make the aggregates selective toward cancer cells by inserting within the ruthenium complexes some "marker" molecules, recognized by protein receptors specifically over-expressed by cancer tissues, iv) to tune the shape and size of the aggregates by acting either on their molecular structure or on external physicochemical parameters like pH and ionic strength. These last effects may be very useful in a stimuli-responsive scenario [24,25]. With these aims we have synthesized and characterized a series of amphiphilic Ru complexes able to form supramolecular aggregates: the structures of these complexes, baptized ToThyRu, HoThyRu and DoHuRu, are shown in Fig. 1 and are based on a pyrimidine deoxyribo-(Thymidine, as in the case of ToThyRu and HoThyRu) or ribonucleoside (Uridine, for DoHuRu). In the general design of these complexes, the nucleoside, chosen as the starting poly-functional scaffold, was decorated with one or two oleic acid chains, attached at the ribose secondary hydroxy groups, as well as with an oligo(ethylene oxide) chain at the 50 -end, so to confer the desired amphiphilicity to the nucleolipids, and to make the resulting aggregates resistant toward the enzymatic degradation. Finally, a pyridine residue was attached to the nucleobase as a chelating moiety able to complex Ru(III) ions. The minimal structure incorporating the Ru(III)-complex, with one pyridine ligand, named AziRu, reported in Fig. 1, was also studied for comparison. Nucleolipids have been selected as scaffold for building up the amphiphilic Ru complexes because of their capability to mime the molecular organizations of the biological systems, as well as for the possibility to form a wide variety of supramolecular systems such as liposomes/vesicles, cubic phases, ribbons, etc. [26e28], that have found an increasing application in the biomedical field. The synthesized molecules have been studied as pure aggregates, as well as in mixture with palmitoyl-2-oleoyl-sn-glycero-3phosphocholine (POPC), at selected POPC/Ru molar ratios. Indeed, the combination of Ru complexes with phospholipids can allow a fine tuning of the metal amount to be administered, as well as protection from degradation, since the ruthenium complex is lodged in the liposome bilayer. Among phospholipids, POPC is of particular interest because it is one of the components of natural membranes [29]. The aggregation behavior of the prepared nanoaggregates has been investigated through an experimental strategy which has been proved to be extremely informative [30,31]. It combines dynamic light scattering (DLS) to estimate aggregate dimensions, small angle neutron scattering (SANS) to analyze the aggregate morphology and to determine their geometrical characteristics, and electron paramagnetic resonance (EPR) to get information on the dynamics of the lipid hydrophobic tail in the bilayer. Finally, testing of the antiproliferative activity of the aggregates has been carried out.

3771

then [trans-RuCl4(DMSO)2] Naþ (0.033 mmol) was added. The mixture was stirred at 40  C and the reaction was monitored by TLC on alumina. After 4 h, TLC showed in all cases the total disappearance of the starting material, with concomitant formation of the desired salt in almost quantitative yields, and the solvent was removed in vacuo. The obtained Ru(III) complexes were then fully characterized by 1H and 13C NMR spectroscopy and ESI-MS analysis. All the collected spectral data are reported in the Supplementary Material Section. 2.2. Sample preparation Ru-containing samples have been prepared by dissolving a suitable amount of the complex in pure chloroform, in order to get a concentration of w1 mg ml1. For pseudo-ternary systems containing POPC, an appropriate amount of this phospholipid was added to the Ru-complex solution, in order to have the pre-fixed molar ratio. The dissolution has been favored by a slight warming (w40  C) and a very short sonication treatment (w5 min). Subsequently, an appropriate amount of this solution has been transferred in round-bottom glass tubes. A thin film was obtained through evaporation of the solvent and vacuum desiccation. The samples have been then hydrated with different media, namely, pure H2O (or D2O), a 0.9% wt NaCl solution and, finally, a pH 7.4 buffer for miming physiological conditions. This buffer has been prepared by dissolving sodium dihydrogenphosphate (NaH2PO4) and disodium hydrogenphosphate (Na2HPO4) in D2O or H2O at concentrations equal to 0.0773 mol dm3 and 0.123 mol dm3, respectively. The pH has been checked to be within 0.1 pH units by means of a Radiometer pHM220 pH-meter, equipped with a AgCl/Ag electrode and a glass electrode previously calibrated with IUPAC standard buffer solutions [33]. All the solutions were vortexed, and the suspensions were then sonicated and repeatedly extruded through polycarbonate membranes of 100 nm pore size, for at least 11 times. The final amount of Ru complex was 0.1 mmol kg1 in both binary and ternary systems; POPC concentration, in systems containing this component, was chosen accordingly to the pre-fixed Ru/POPC molar ratio. Samples prepared for EPR experiments also included 1% (w/w) of spin-labeled phosphatidylcholine (1-palmitoyl-2-[n-(4,4-dimethyloxazolidine-N-oxyl)]stearoylsn-glycero-3-phosphocholine, nePCSL, n ¼ 7,14), purchased from Avanti Polar Lipids and stored at 20  C in ethanol solutions. Samples prepared for fluorescence microscopy were prepared as reported above by adding 2% mol of 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) ammonium salt, here abbreviated as Rhod, purchased from Avanti Polar Lipids and used as received. 2.3. Dynamic light scattering (DLS) DLS measurements were performed with a home-made instrument composed by a Photocor compact goniometer, an SMD 6000 Laser Quantum 50 mW light source operating at 5325 Å, a photomultiplier (PMT-120-OP/B) and a correlator (Flex02-01D) from Correlator.com. All the measurements were performed at (25.00  0.05)  C with the temperature controlled through the use of a thermostat bath. In DLS, the intensity autocorrelation function g(2)(t) is measured and related to the electric field autocorrelation function g(1)(t) by the Siegert relation [34]. This latter function can be written as the Laplace transform of the distribution of the relaxation rate G used to calculate the translational diffusion coefficient D: gð1Þ ðtÞ ¼

ZþN N



t



sAðsÞexp  d ln s s

(1)

where s ¼ 1/G. Laplace transforms were performed using a variation of CONTIN algorithm incorporated in Precision Deconvolve software. From the relaxation rates, the zeaverage of the diffusion coefficient D may be obtained as: D ¼ lim

q/0

G q2

(2)

q ¼ 4pn0/lsin(q/2) is the modulus of the scattering vector, n0 is the refractive index of the solution, l is the incident wavelength and q represents the scattering angle. Thus D is obtained from the limit slope of G as a function of q2, where G is measured at different scattering angles. For each sample, relaxation times were acquired at five to six angles at least, and three or more measurements per angle were performed in order to improve statistical accuracy. For spheres diffusing in a continuum medium at infinite dilution, the diffusion coefficient DN is dependent on the sphere radius RH, called hydrodynamic radius, through the StokeseEinstein equation [35]: kT 6ph0 DN

2. Materials and methods

RH ¼

2.1. General procedure for the synthesis of the nucleolipids Ru(III) complexes ToThyRu, HoThyRu, DoHuRu

where k is the Boltzmann constant, T is the absolute temperature and n0 is the solvent viscosity. For not spherical particles, RH represents the radius of equivalent spherical aggregates. Due to the high dilution, for non interacting species it is possible to make the approximation: D y DN and h y h0, where h represents the solution viscosity. In this hypothesis, equation (3) can be reasonably used to estimate the hydrodynamic radius of the aggregates [36].

The selected nucleolipid (ToThy, HoThy or DoHu, 0.033 mmol), synthesized in our laboratories as previously described [32], was dissolved in 1.0 ml of the appropriate anhydrous solvent (CH3CN for ToThy; CH2Cl2 for HoThy and DoHu) and

(3)

3772 G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

Fig. 1. Functional nucleolipids: ToThyRu (A), HoThyRu (B) and DoHuRu (C), along with the complex AziRu (D).

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782 2.4. Small angle neutron scattering (SANS) SANS measurements were performed at 25  C with the KWS2 instrument located at the Heinz Meier Leibtnitz Source, Garching Forschungszentrum (Germany). Neutrons with a wavelength spread Dl/l  0.2 were used. A twodimensional array detector at three different wavelength (W)/collimation (C)/ sample-to-detector(D) distance combinations (W7ÅC8mD2m, W7ÅC8mD8m and W19ÅC8mD8m), measured neutrons scattered from the samples. These configurations allowed collecting data in a range of the scattering vector modulus q ¼ 4p/lsin (q/2) between 0.0019 Å1 and 0.179 Å1, with q scattering angle. The investigated systems were contained in a closed quartz cell, in order to prevent the solvent evaporation and kept under measurements for a period sufficient to have w2 million counts. The raw data were then corrected for background and empty cell scattering. Detector efficiency correction, radial average and transformation to absolute scattering cross sections dS/dU were made with a secondary plexiglass standard [37,38]. For a system composed by a collection of monodisperse bodies, the scattering cross section dS/dU, that contains information on their interactions, sizes and shapes, can be expressed as [39]   dS dS ¼ np PðqÞSðqÞ þ dU dU incoh

(4)

where np represents the number density of the scattering objects in the system, P(q) and S(q) are the form and the structure factors of the scattering particles respectively, whereas (dS/dU)incoh takes into account the incoherent scattering contribution, mainly due to the presence of hydrogen atoms. The form factor contains information on the shape of the scattering objects, whereas the structure factor accounts for inter-particle correlation and is normally important for concentrated or charged systems. Provided that solutions are quite diluted (c < 1 mmol kg1), the structure function S(q) can be approximated to the unity, and the scattering cross section reduces to   dS dS ynp PðqÞ þ dU dU incoh

(5)

Structural parameters of the aggregates have been obtained by applying the appropriate models to the experimental SANS data as described in the Results and Discussion section. 2.5. Electron paramagnetic resonance (EPR) EPR spectra were recorded on a 9 GHz Bruker Elexys E-500 spectrometer (Bruker, Rheinstetten, Germany). Capillaries containing the samples were placed in a standard 4 mm quartz sample tube containing light silicone oil for thermal stability. The temperature of the sample was regulated at 25  C and maintained constant during the measurement by blowing thermostated nitrogen gas through a quartz Dewar. The instrumental settings were as follows: sweep width, 120 G; resolution, 1024 points; modulation frequency, 100 kHz; modulation amplitude, 1.0 G; time constant, 20.5 ms, incident power, 5.0 mW. Several scans, typically 16, were accumulated to improve the signal-to-noise ratio. Values of the outer hyperfine splitting, 2Amax, were determined by measuring, through a home-made MATLAB-based routine, the difference between the low-field maximum and the high-field minimum. This parameter is a useful empirical measure of the lipid chain dynamics and order in both gel and fluid phases of lipid bilayers [40,41]. The main source of error on the 2Amax value is the uncertainty in composition of samples prepared by mixing few microliters of mother solutions. For this reason, reproducibility of 2Amax determination was estimated by evaluating its value for selected independently prepared samples with the same nominal composition. The uncertainty affecting the 2Amax parameter was 0.2 G. 2.6. Cell cultures Human WiDr epithelial colorectal adenocarcinoma cells, MCF-7 breast adenocarcinoma cells, and rat C6 glioma cell line were purchased from ATCCÒ (American Type Culture Collection, Manassas, Virginia, USA). C6 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen, Paisley, UK) containing high glucose (4.5 g/l), while MCF-7 and WiDr were grown in RPMI 1640 medium (Invitrogen, Paisley, UK). Media were supplemented with 10% fetal bovine serum (FBS, Cambrex, Verviers, Belgium), L-glutamine (2 mM, Sigma, Milan, Italy), penicillin (100 units/ml, Sigma) and streptomycin (100 mg/ml, Sigma), according to ATCC recommendations. All cells were cultured in a humidified 5% carbon dioxide atmosphere at 37  C. 2.7. Anticancer activity The anticancer activity of ruthenium-containing nucleolipidic nanoparticles and of AziRu was investigated by the estimation of a "cell survival index", arising from the combination of cell viability evaluation with cell counting. WiDr, MCF-7 and C6 cells were washed with PBS buffer solution (Sigma), collected by trypsine (Sigma) and then inoculated in a 96-microwell culture plates at density of 104 cells/well. Cells were allowed to grow for 24 h, then the medium was replaced with fresh

3773

medium and cells were treated for further 48 h with a range of concentrations (10e100 mM) of AziRu and of the liposomes POPC, DoHuRu/POPC, HoThyRu/POPC and ToThyRu/POPC. Cell viability was evaluated with an MTT assay procedure, which measures the level of mitochondrial dehydrogenase activity using the yellow 3-(4,5dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT, Sigma) as substrate. The assay was based on the redox ability of living mitochondria to convert dissolved MTT into insoluble purple formazan. Briefly, after the treatments the medium was removed and the cells were incubated with 20 ml/well of an MTT solution (5 mg/ml) for 1 h in a humidified 5% CO2 incubator at 37  C. The incubation was stopped by removing the MTT solution and by adding 100 ml/well of DMSO to solubilize the purple formazan. Finally, the absorbance was monitored at 530 nm by using a multiwell plate reader in a PerkineElmer LS 55 Luminescence Spectrometer (PerkineElmer Ltd, Beaconsfield, UK) [42]. Cell number and proliferation was determined by TC10 automated cell counter (Bio-Rad, Milan, Italy), providing an accurate and reproducible total count of mammalian cells and a live/dead ratio in one step. The calculation of the concentration required to inhibit the net increase in the cell number and viability by 50% (IC50) is based on plots of data carried out in triplicates and repeated three times. IC50 values were obtained using a dose response curve by nonlinear regression using a curve fitting program, GraphPad Prism 5.0, and are expressed as mean  SEM. 2.8. Light microscopy In microphotography experiments, MCF-7, WiDr and C6 cell lines were grown on 100 mm standard culture dishes (Tissue Culture Dish, Falcon) by plating 3  106 cells in 10 ml of culture medium (see above). After reaching the confluence, cells were incubated for 48 h with 100 mM of AziRu and of the liposomes POPC, DoHuRu/POPC, HoThyRu/POPC and ToThyRu/POPC. Finally, cells were placed on a contrast-phase light microscope (Labovert microscope, Leizt). Microphotographs at an 100 magnification were taken with a standard VCR camera (Nikon). 2.9. Fluorescence microscopy and cellular uptake of liposomes Sterile coverslips were placed in six-well plates. MCF-7 cells were seeded at a concentration of 2  104 per ml in the same six-well plates. Following a growth period of 24 h at 37  C in RPMI 1640 medium containing 10% FBS, the medium was replaced with serum-free RPMI 1640 medium, which was followed by the addition of fluorescent ToThyRu/POPC liposomes or free PBS saline solution (fluorescent ToThyRu/POPC final concentration 100 mM) to each well. The cells were incubated for additional times (30 min, 1, 3 and 6 h) with liposomes and washed with PBS three times to remove unassociated liposomes. The cells were then fixed at room temperature in 4% paraformaldehyde for 20 min, which was followed by adding 0.2% Triton X-100 (Sigma) to disrupt the cell membranes. After washing with PBS three times, the cells were treated with diaminophenylindole (DAPI) (Sigma) to stain the cell nuclei. The coverslip from each well was mounted onto a glass microslide with 80% fluorescence-free glycerol mounting medium. Finally, the interaction of liposomes with MCF-7 cells and the cellular uptake was monitored using a fluorescent microscope (Leica Microsystems GmbH, Wetzlar, Germany) to visualize DAPI (345/ 661 nm) and fluorescent liposomes (557/571 nm). Images were taken using an AxioCam HRc video-camera (Zeiss) connected to an Axioplan fluorescence microscope (Zeiss) using the AxioVision 3.1 software. 2.10. Statistical analysis All data were presented as mean  SEM. The statistical analysis was performed using Graph-Pad. Prism (Graph-Pad software Inc., San Diego, CA) and ANOVA test for multiple comparisons was performed followed by Bonferroni’s test.

3. Results and discussion 3.1. Synthesis of AziRu AziRu was obtained by reaction of pyridine with an equimolar amount of [trans-RuCl4(DMSO)2] Naþ in anhydrous CH3CN, kept at 40  C under stirring. After 4 h the reaction mixture was taken to dryness, giving the desired salt in a pure form, as confirmed by ESIMS and 1H NMR analysis (data not shown). 3.2. Synthesis of the nucleolipid Ru(III)-complexes ToThyRu, HoThyRu and DoHuRu The general design of the starting nucleolipids ToThy, HoThy and DoHu, here used to build the amphiphilic, self-assembling

3774

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

Ru(III)-complexes ToThyRu, HoThyRu and DoHuRu (Fig. 1), was centered on thymidine or uridine, chosen as the central scaffolds anchoring three diverse structural motifs: 1) one pyridine-methyl arm, inserted at the N-3 position, as the functional reporter group, allowing the transition-metal complexation; 2) one (in the case of thymidine derivatives ToThy and HoThy) or two (in the case of uridine analog DoHu) lipophilic oleic acid residues, attached at the ribose secondary hydroxyl group(s) to induce self-assembly in aqueous solutions; 3) one hydrophilic oligo(ethylene glycol) chain, linked at the 50 OH ribose moiety, with the role of optimizing the “hydrophilic/ lipophilic balance” within the hybrid molecules and of preventing their extracellular enzymatic degradation by the reticuloendothelial system. The synthesis of ToThyRu, HoThyRu and DoHuRu was accomplished in one, quantitative step by coupling the functional amphiphilic nucleosides here named as ToThy, HoThy and DoHu, respectively (Scheme 1) previously prepared in our laboratories [32] with [trans-RuCl4(DMSO)2] Naþ added in equimolar amounts. The obtained Ru(III) salts ToThyRu, HoThyRu and DoHuRu were then characterized by ESI-MS analysis, which in all cases confirmed the presence of the target complexes as pure compounds. An indepth NMR characterization of these compounds could not be undertaken, since both 1H and 13C NMR spectra showed large, broadened signals; this effect, attributable to the presence of the paramagnetic Ru(III) atom, confirmed the effective complexation of the nucleolipids, which, before the reaction with [transRuCl4(DMSO)2] Naþ, showed easily interpretable spectra. Remarkably, 1H NMR spectra of the three complexes dissolved in CDCl3 exhibited diagnostic, dramatically upfield-shifted signals at ca. d ¼ 2 and 10 ppm, respectively attributed to the pyridine and DMSO-methyl protons coordinating the Ru(III) atom. 3.3. Physico-chermical characterization of pure ToThyRu, HoThyRu and DoHuRu nanoaggregates Amphiphilic Ru-complexes ToThyRu, HoThyRu and DoHuRu nanovectors were investigated in pure water and in NaCl solution at physiological ionic strength, by DLS, SANS and EPR techniques. These systems have not been susceptible of a detailed investigation because of the fast hydrolysis and poly-oxo species formation,

followed by precipitation of the species, as detailed in section 3.4. Because of these limitations, all the measurements described in this subsection have been performed on freshly prepared samples. The hydrodynamic radius distributions of ToThyRu, HoThyRu and DoHuRu, as pure aggregates in water and at physiological ionic strength, were determined by equation (3) and are reported in Fig. 3A and Fig. SI1 (See Supplementary Material Section for figures and tables labeled as SI), respectively. The diffusion coefficients D along with hydrodynamic radii RH are reported in Table 1 and SI1: a comparison of the distribution functions as well as of the results obtained from data analysis shows a substantial similarity between the systems in water and the corresponding ones in physiological conditions. The morphology of these aggregates, as well as their geometrical characteristics, has been obtained by means of SANS measurements. Scattering cross sections for all the binary systems in water are reported in Fig. 3B (scattering data measured at physiological ionic strength are not reported). Examination of the figure shows the presence, in the low q region, of a power law (dS/ dU) f qa where a is an exponent ranged between 2 and 4. These values are characteristic of systems containing multilamellar vesicles (MLV), being the exponent a strictly connected to the mean lamellarity of the supramolecular structure [43]: in the limit case of a ¼ 2, the presence of unilamellar vesicles occurs. Additionally, in the intermediate q region, i.e. 0.025 < q/Å1 < 0.070, the HoThyRu scattering cross sections follow the law: (dS/dU) f q1, typical of aggregates in which one of the dimensions is prevalent over the other two, i.e. cylindrical or highly elongated micelles [25]. This evidence is in agreement with DLS results highlighting the presence in solution of smaller and faster objects, namely the cylindrical micelles showed by SANS, coexisting with bigger and slower aggregates, the vesicles detected by SANS as well. Surprisingly, the behavior of the ToThyRu-based system, expected to be very close to that of HoThyRu, even showing a hydrodynamic radius bimodal distribution function, does not exhibit a detectable (dS/dU) f q1 power law in Fig. 3B. This is likely due to the very small volume fraction of micelles that are dominated by the large MLV, as also supported by a comparison of the ratio (small/large aggregates) between the area of the hydrodynamic radius distribution functions of ToThyRu and HoThyRu. This is larger for the HoThyRu system than for ToThyRu suggesting a large presence of smaller aggregates. In contrast, in the case of DoHuRu the power law (dS/dU) f q1 in the intermediate q region is absent, so that only the presence of

O R1

-

O R1

N

Cl

N N

N N

R4 O

O

R3

R1 = R2 = R3 = R4 =

O

+ R2 [trans-RuCl4(DMSO)2] Na

40 *C, stirring CH3CN or CH2Cl2

N

R4 O

O

R3

O

Ru

Cl

R2

Cl O S

Cl CH 3 CH 3

-CH3 for ToThy and HoThy; -H for DoHu, -H for ToThy and HoThy; -R3 for DoHu; oleic acid residue; CH3O(CH2CH2O)3CH2CO- for ToThy ; BnO(CH2CH2O)6CH2CO- for HoThy and DoHu. Scheme 1. e Synthetic procedure for the preparation of ToThyRu, HoThyRu and DoHuRu.

Na+

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

3775

Fig. 2. Example of time evolution obtained for aqueous dispersions of DoHuRu in water as a freshly prepared solution (A) and after 120 h (B), and in POPC formulation in pseudophysiological conditions 10 months after its preparation (C).

vesicles is admitted in the model. Also in this case, SANS evidence is confirmed by the monomodal distribution of DoHuRu observed by DLS. Quantitative analysis of scattering data, through equation (5), has been performed by modeling the systems as a collection of one-dimensional paracrystalline stacks [44] and, where present, cylindrical micelles [45]. From the analysis is possible to obtain, for MLV, the average number of lamellae per liposome N, the mean layer thickness s and, finally, for the average distance between the centers of two consecutive layers dl. Finally, for micelles, the cylinder radius Rcyl is evaluable and, by combining this latter information with the cylindrical hydrodynamic radius, the micellar linear extension Lcyl is obtainable as well, through the relation [46]

RH

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  3 3 ¼ R2cyl Lcyl 1:0304 þ 0:0193x þ 0:06229x2 þ 0:00476x3 4 

þ 0:00166x4 þ 2:66,106 x7

ð6Þ

where x ¼ ln[Lcyl/(2Rcyl)]. All these parameters are reported in Table 1 and Table SI1, also showing for comparison the values detected for POPC liposomes. Again, as for DLS results, quantitative analysis of the data does not show large differences among the systems in water and those at physiological ionic strength. Liposomial aggregates are mainly oligolamellar, with a number of lamellae ranging from 5 to 10 units. Thickness of pure POPC and Ru-complex bilayers are located around 4.0 O 5.0 nm. In this context it is worthy to note that the extension of the C18 alkyl chain in an all-trans conformation is w2.5 nm. Moreover, the chain extension is fairly close to the radius, Rcyl measured for the cylindrical aggregates (2.7 nm). A microstructural investigation on the aggregates formed by HoThyRu, ToThyRu and DoHuRu in water was performed by EPR. Two amphiphilic spin probes were alternatively inserted in the systems: 5ePCSL and 14ePCSL. The former bears the radical nitroxide group close to the hydrophilic moiety of the molecule, while in the latter one the reporter group is located at the end of the hydrophobic tail. Consequently, 5-PCSL provides information on the aggregate microdomain just below the external surface, while the

14-PCSL gives information on the inner core. The spectra of 5-PCSL in HoThyRu, ToThyRu and DoHuRu self-assemblies in water, recorded on the freshly prepared samples, are shown in Fig. 3C. In all cases, an anisotropic lineshape is observed, which is indicative of the formation of layered structures typical of vesicular meso-structures [23]. Interestingly, the spectra of 14-PCSL in the same systems show an analogous, although less resolved, anisotropic line shape as shown in Fig. 3D. This result is quite unexpected for aggregates whose inner hydrophobic domain is composed by oleyl chains and indicates an ordered and poorly dynamic packing of the molecules forming the aggregates. All the EPR spectra were also registered after 3 months from sample preparation. No signal was observed, because of the instability of the aggregates formed by HoThyRu, ToThyRu and DoHuRu in water, see below. 3.4. Stability of the complexes As for NAMI-A [16] and other ruthenium complexes, the stability of the synthesized ToThyRu, HoThyRu and DoHuRu is limited. In fact, depending on the environmental conditions, a clear change of the Ru-complex properties in few hours (pseudo-physiological conditions) or few days (pure water) has been observed. In particular, in all samples the formation of small dark particles after y78 h in water and after y5 h in pseudo-physiological conditions has been detected. The degradation process is imputable to the replacement of chloride ions, as well as of the DMSO ligand, with water molecules and/or hydroxide ions, followed by the formation of poly-oxo species [16]: these processes result in a visible change of the solution color, changing from yellow to dark green. As time goes, precipitation of brown particles is observed, as displayed in Fig. 2. These phenomena are more pronounced as the pH increases and don’t allow a full characterization of the formed nanoaggregates. With the aim to increase the lifetime of the Ru-based nanoaggregates, the synthesized molecules have been lodged in POPC liposomes. In this case, nanoaggregates have been observed to be stable for months, i.e. no degradation and precipitation phenomena occur as depicted in Fig. 2, if the Ru amount does not exceed 15% in

3776

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

Fig. 3. Physico-chemical data obtained for freshly prepared binary systems in water containing POPC (0.2 mmol/kg, green data), ToThyRu (0.1 mmol/kg, red data), HoThyRu (0.1 mmol/kg, orange data) and DoHuRu (0.1 mmol/kg, violet data). (A) Hydrodynamic radius distribution functions, (B) scattering cross sections, (C) EPR spectra of 5ePCSL, (D) EPR spectra of 14ePCSL. In (B) cross sections have been multiplied for a scale factor, as indicated, and reported along the fitting lines; in (C) solid lines have been obtained for freshly prepared samples, whereas dotted lines have been registered 24 h after the preparation (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

molar fraction with respect to total amphiphilic solutes. In the next section, the physico-chemical characterization of mixed aggregates is presented. 3.5. Physico-chemical characterization of POPC/amphiphilic Ru complexes nanoaggregates With the aim to slow down Ru complexes degradation, as well as to increase the biocompatibility of the formulations and

modulate the amount of ruthenium carried, the synthesized amphiphilic molecules were co-aggregated with amounts of 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) at 15/ 85 M ratio. These aggregates were then studied in three different systems: i) in pure water; ii) in a physiological ionic strength solution; iii) in pseudo-physiological conditions. DLS investigations, whose results are reported in Figs. 4A and SI2, suggest that, in contrast to the binary systems, the ternary dispersions exhibit the presence of a single mode corresponding to

Table 1 Mean diffusion coefficients, hydrodynamic radii and geometrical parameters for aggregates obtained from SANS data for ToThyRu, HoThyRu, DoHuRu and POPC binary systems in water. The meaning of the symbols is reported in the paper. DLS

SANS

108 D cm2 s1 ToThyRu HoThyRu DoHuRu POPC

Fast Slow Fast slow

27 2.7 35 4.1 6.1 3.5

     

3 0.3 5 0.3 0.3 0.3

s

RH nm 9 90 7 60 40 70

     

N

nm 1 10 1 4 2 7

dl nm

4.3  0.2

81

9.1  0.3

4.9  0.3 4.8  0.3 4.4  0.1

10  1 10  1 51

9.2  0.3 10.0  0.3 8.7  0.4

Rcyl nm

Lcyl nm

2.8  0.4

45  7

2.7  0.3

37  8

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

3777

Fig. 4. Physico-chemical data obtained in buffer at pH 7.4 for ternary systems POPC/Ru complex at 85/15 phospholipid/ruthenium complex molar ratio and 0.1 mmol/kg concentration of ruthenium complex. (A) Hydrodynamic radius distribution functions, (B) scattering cross sections, (C) EPR spectra of 5ePCSL, (D) EPR spectra of 14ePCSL. In (B) cross sections have been multiplied for a scale factor, as indicated, and reported along the fitting lines; in (C) solid lines have been obtained for freshly prepared samples, whereas dashed curves have been registered 90 days after the preparation.

a single translational diffusive species. Furthermore, all the distributions are approximately located in the same RH region, around w60 nm. This value is comparable to that obtained for pure POPC liposomes, according to the circumstance that the relative amount of POPC is large. Thus the presence of POPC does not give rise to the formation of micelles and the ruthenium complexes are fully lodged in the liposome bilayer. From the point of view of Ru

complexes, this means that the synthesized molecules do not destabilize the POPC liposomes. This is also confirmed by SANS data, reported in Fig. 4B, that show, for all the systems, the presence of MLV vesicles, whose structure shows some rigidity, because of the presence of oscillations observable at q y 0.007 Å1. The characteristics of detected liposomes are reported in Tables 2, 3 and SI2. Inspection of the

Table 2 Mean diffusion coefficients, hydrodynamic radii and geometrical parameters for aggregates obtained from SANS data for ToThyRu/POPC, HoThyRu/POPC and DoHuRu/POPC ternary systems in NaCl 0.9% wt at 85/15 ratio between POPC and ruthenium complex.

Table 3 Mean diffusion coefficients, hydrodynamic radii and geometrical parameters for aggregates obtained from SANS data for ToThyRu/POPC, HoThyRu/POPC and DoHuRu/POPC ternary systems in the pH ¼ 7.4 pseudo-physiological buffer at 85/15 ratio between POPC and ruthenium complex.

DLS 8

10 D/(cm s ToThyRu/POPC HoThyRu/POPC DoHuRu/POPC

4.2  0.2 4.4  0.2 4.6  0.3

DLS

SANS 2 1

)

s

RH nm

nm

58  3 56  3 53  3

4.1  0.4 4.4  0.4 4.8  0.4

N 10  1 13  1 91

108D/(cm2s1)

dl nm 8.7  0.3 7.9  0.4 7.6  0.3

ToThyRu/POPC HoThyRu/POPC DoHuRu/POPC

4.5  0.2 4.5  0.3 4.2  0.2

SANS

s

RH nm

nm

54  2 54  3 58  3

4.5  0.5 4.4  0.3 4.2  0.7

N

dl nm

12  1 15  1 10  1

8.5  0.5 8.4  0.5 8.7  0.4

3778

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

Table 4 Values of the outer hyperfine splitting 2Amax, in gauss (G), of 5ePCSL and 14ePCSL in bilayers of POPC and POPC/ruthenium complexes at 85/15 M ratio. 5ePCSL POPC ToThyRu/POPC HoThyRu/POPC DoHuRu/POPC

52.4 49.5 51.5 53.2

   

0.1 0.1 0.2 0.2

14ePCSL 32.6 33.5 33.3 32.9

   

0.1 0.1 0.1 0.1

Tables allows observing similar values for all the systems and a similarity among them and the binary systems containing only POPC as the amphiphilic component. This is ascribable, again, to the abundance of POPC in the investigated systems. Finally, for both DLS and SANS measurements no difference has been observed comparing the systems analyzed in 0.9% NaCl or in pseudophysiological solutions. A further microstructural investigation of the liposomes formed by HoThyRu, ToThyRu and DoHuRu in mixtures with POPC was performed by EPR. The spectra, depicted in Figs. 4CeD, show an anisotropic lineshape for 5ePCSL and an almost isotropic lineshape in the case of 14ePCSL. This is a characteristic hallmark of lipid bilayers in the liquid-crystalline fluid phase. A quantitative analysis of the spectra has been performed by determining the outer hyperfine splitting, 2Amax, whose values are reported in Table 4. Inspection of the Table shows that insertion of HoThyRu or ToThyRu in POPC causes a reduction of 2Amax for 5ePCSL, indicating a loosening of the local microstructure. In contrast, DoHuRu causes an increase of 2Amax, i.e., a stiffening of the bilayer. These evidences suggest that the inclusion of single-tailed molecules can

perturb the bilayer, while this is not the case for double-tailed molecules. The EPR spectra, performed on the same samples after three months, showed no variation of the signals, confirming the stability with time of the bilayers formed by POPC hosting HoThyRu, ToThyRu or DoHuRu. Definitively, the presence of POPC increases the aggregates stability, namely no precipitation due to poly-oxo species formation is observed. 3.6. Cellular uptake of liposomes by means of fluorescence microscopy Fluorescence microscopy was used to evaluate the liposomal uptake and localization into human MCF-7 cancer cells (see Fig. 5). In microphotographs, the blue areas indicate the cells nuclei stained by DAPI, whereas the red areas represent the liposomeassociated Rhod localization within cells. As shown by timecourse experiments, ToThyRu/POPC liposomes were taken up significantly by MCF-7 cells after 3 h of incubation. Merely small amounts of liposomes were detected into cells after shorter incubation times such 30 min (data not shown) and 1 h. After 6 h of incubation liposomes were taken up massively by cells as highlighted by the red fluorescence emission, and localized widespread in the nuclei, cytoplasm and close to membranes. In merged images, the purple fluorescence, which was enhanced when blue and red fluorescence overlapped at the same location, suggested also a nuclear localization of fluorescent aggregates. Rutheniumcontaining liposomes were taken up possibly via endocytosis and/or membrane fusion in a nonspecific pattern involving multiple molecular mechanisms [47].

Fig. 5. Fluorescent microphotographs of monolayers showing the cellular uptake of liposomes by human MCF-7 breast adenocarcinoma cells. MCF-7 were incubated with 100 mM of the fluorescent ToThyRu/POPC liposome for 1, 3 and 6 h. DAPI is used as a nuclear stain (shown in blue). Rhod-dependent fluorescence of ToThyRu/POPC liposomes is shown in red. In merged images (Merge), the purple areas represent the localization of liposomes in the nuclei. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

3779

Fig. 6. Representative microphotographs by phase-contrast light microscopy of MCF-7 (A), WiDr (B) and C6 cell lines (C) untreated (control cells, left column) or treated for 48 h with ToThyRu/POPC (right column) showing the morphological changes of cells and the cytotoxic effects on cellular monolayers. Inset: higher magnifications of injured cells following incubations with the anti-proliferative ToThyRu/POPC liposome.

3.7. Cell morphological changes and cytotoxic effect by microscopic analysis Throughout the in vitro experiments, confluent MCF-7, WiDr and C6 cancer cells treated with DoHuRu/POPC, HoThyRu/POPC and ToThyRu/POPC were examined under phase-contrast light microscopy. As reported in the representative micrographs of Fig. 6, following incubation with ToThyRu/POPC, in vitro modifications of the cell monolayers morphology clearly appeared in conjunction with a well detectable cytotoxic effect. Besides losing their normal morphology, apoptotic features - in particular membrane blebs and cell shrinkage - seemed to appear in most of the cells as well as the number of rounded-up cells clearly increased after 48 h of incubation with ruthenium-based nanocarriers. These findings are substantially in accordance with recent reports showing the in vitro apoptosis-inducing activities of ruthenium complexes to explain anti-proliferative effect and show that the amphiphilic nature of the synthesized Ru(III) complexes, and the consequent self-aggregation, do not perturb the metal-induced biological effects [48,49]. 3.8. Antitumor activity First, in our in vitro experimental model we analyzed the antiproliferative effects of the low molecular weight ruthenium

complex AziRu, analog of NAMI-A. From the analysis of the concentration/effect curves depicted in Fig. 7 emerged that AziRu shows a moderate cytotoxicity, with IC50 values of about 305 mM and 440 mM measured on human MCF-7 and WiDr cancer cells, respectively, and of about 318 mM in rat C6 glioma cells (see Table 5). These data are broadly in agreement with literature reports regarding the anti-proliferative effects of ruthenium-based drugs [50]. However, from our data it is interesting to note that AziRu seem to be more cytotoxic than the well known NAMI-A complex. One possible reason is that ligands could confer a higher lipophilic character to AziRu complex thereby leading to enhanced cellular uptake efficiency. Additionally, the ligands may also play an important role in biomolecular interactions and recognition processes. Indeed, hydrophobicity, cellular uptake efficiency and cytotoxic effects of anti-proliferative drugs on cancer cells are often strictly correlated [49]. Since the AziRu and NAMI-A mechanism of action should be the same, this finding emphasizes the importance of the transitionemetal complex physical and chemical properties which can play a critical role in the vehiculation of the active metal to the molecular targets, thus modulating its bioavailability. Under the same incubation conditions, predictably the liposomes composed exclusively of POPC showed no significant interference with the cell viability and proliferation. Starting from these preliminary experiments, the cells were then treated for 48 h with

3780

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

Fig. 7. Concentration/effect curves and cell survival index. Cell survival index, evaluated by MTT assay and total cell count, in MCF-7 (panel A), WiDr (panel C) and C6 (panel E) cell lines incubated for 48 h with POPC, with different Ru-containing formulations and with AziRu, as indicated in the legend. In panel B, D and F, for MCF-7, WiDr and C6 cells respectively, the concentration-effect curves are reported, by normalizing for the actual amount of ruthenium contained within DoHuRu/POPC, HoThyRu/POPC and ToThyRu/POPC liposomes. Data are expressed as percentage of untreated control cells and are reported as mean of three independent experiments  SEM.

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782 Table 5 Ruthenium IC50 values (mM) relative to NAMI-A, AziRu and DoHuRu/POPC, HoThyRu/ POPC and ToThyRu/POPC liposomes in the indicated cell lines following 48 h of incubation. IC50 values are reported as mean  SEM. Values for NAMI are taken from [49] in the main text. MCF-7 NAMI AziRu DoHuRu/POPC HoThyRu/POPC ToThyRu/POPC

620 305 71 7 9

    

WiDr 30 16 6 4 4

441 99 40 75

   

C6 20 5 5 4

318 24 81 36

   

12 5 7 8

different concentrations (10, 25, 50 and 100 mM) of DoHuRu/POPC, HoThyRu/POPC and of ToThyRu/POPC, which transport in liposomal form amphiphilic nucleolipidic ruthenium complexes derived from AziRu. The evaluation of the "cell survival index" emerging from concentration/effect curves (Fig. 7, panels A, C and E for MCF-7, WiDr and C6 cells, respectively) suggested that these nanocarriers produce a cytotoxic effect substantially similar to that of AziRu, albeit they enclose only 15% of ruthenium in moles. In short, these formulations lodged in POPC liposomes allowed reaching the same cytotoxicity shown by AziRu, but at a ruthenium concentration of about 6 times smaller. This is the most interesting result we expected. In fact, by normalizing the results to the actual ruthenium amounts contained within the nanoaggregates, we obtained new concentration/effect curves clearly showing that our nanocarriers are more effective in inhibiting the growth of cancer cells than AziRu complex (Fig. 7, panels B, D and F for MCF-7, WiDr and C6 cells, respectively). The higher anti-proliferative effect was obtained with the nanocarriers HoThyRu/POPC and ToThyRu/POPC on MCF-7 cells, the most sensitive cells to the ruthenium complex action in our model. Consequently, due to ruthenium vectorization, the IC50 values resulted markedly reduced compared to that of AziRu (Table 5). Less sensitive to the effect of ruthenium were found to be WiDr and C6 cells, albeit in a context in which all the ruthenium-based nanocarriers resulted more effective than AziRu, as evidenced by the calculation of IC50 values. Overall, bioactivity screenings, in addition to indicate a higher cytotoxicity for AziRu with respect to the well known NAMI-A complex, demonstrated that the nucleolipid Ru(III) complexes DoHuRu, HoThyRu and ToThyRu, lodged in POPC liposomes, show higher in vitro anti-proliferative activities toward cancer cells of different histogenesis than AziRu itself. This means that the three synthesized AziRu derivatives turn out to exhibit higher in vitro cytotoxicity, presumably as a result of enhanced cellular bioavailability when administered as POPC liposomes, thereby opening new perspectives in the design of innovative transition metalbased supramolecular systems for anticancer therapy. 4. Conclusions We have reported on new Ru-based antineoplastic agents, namely the organometallic complex AziRu and three amphiphilic nucleolipid derivatives, incorporating AziRu in their structure. In contrast to several other ruthenium complexes known in the literature that show limited stability in physiological environment, the amphiphilic molecules we have proposed when lodged in the POPC lipid bilayer exhibit long life stability, as long as months. The obtained nanoaggregates have a mono- or multilayer vesicular morphology. Remarkably, AziRu and the derivatives Ru(III) nucleolipidic complexes DoHuRu, HoThyRu and ToThyRu lodged in POPC liposomes have demonstrated high in vitro anti-proliferative activity. To the best of our knowledge, these molecules are among the most promising ruthenium-based targets as anticancer and

3781

antimetastatic agents currently described in the literature. In fact, on the investigated panel of human and non-human cell lines, AziRu shows IC50 values lower than most of its companion organometallic complexes; in the same conditions, a further improved behavior has been displayed by all the amphiphilic ruthenium complexes stabilized in POPC formulations, showing in turn IC50 values w6 times lower than those of AziRu. Acknowledgments This work was supported by MIUR (PRIN 2008-prot. 20087K9A2J). The authors thank the Forschungszentrum Jülich for provision of beam time. SANS experiments were supported by the European Commission, NMI3 contract RII3-CT-2003-505925. Appendix. Supplementary material Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.biomaterials.2012.01. 057. References [1] Rosenberg B, VanCamp L, Trosko JE, Mansour VH. Platinum compounds: a new class of potent antitumor agents. Nature 1969;222:385e6. [2] Boulikas T, Vougiouka M. Cisplatin and platinum drugs at the molecular level (review). Oncol Rep 2003;10:1663e82. [3] Graham J, Muhsin M, Kirkpatrick P. Oxaliplatin. Nat Rev Drug Discov 2004;3: 11e2. [4] Ramaswamy S, Ross KN, Lander ES, Golub TR. A molecular signature of metastasis in primary solid tumors. Nat Genet 2003;33:49e54. [5] Pantel K, Brakenhoff RH. Dissecting the metastatic cascade. Nat Rev Cancer 2004;4:448e56. [6] Adjei AA, Rowinsky EK. Novel anticancer agents in clinical development. Cancer Biol Ther 2003;2:S5e15. [7] Germanov E, Berman JN, Guernsey DL. Current and future approaches for the therapeutic targeting of metastasis (Review). Int J Mol Med 2006;18:1025e36. [8] Bergamo A, Cocchietto M, Capozzi I, Mestroni G, Alessio E, Sava G. Treatment of residual metastases with Na[trans-RuCl4(DMSO)Im] and ruthenium uptake by tumor cells. Anti-Cancer Drugs 1996;7:697e702. [9] Sava G. Ruthenium compounds in cancer therapy. Met Compd Cancer Ther; 1994:65e91. [10] Sava G, Pacor S, Bregant F, Ceschia V. Metal complexes of ruthenium: a potential class of selective anticancer drugs. Anticancer Res 1991;11: 1103e7. [11] Sava G, Pacor S, Bregant F, Ceschia V, Mestroni G. Metal complexes of ruthenium: antineoplastic properties and perspectives. Anti-Cancer Drugs 1990;1:99e108. [12] Sava G, Pacor S, Mestroni G, Alessio E. Na[trans-RuCl4(DMSO)Im], a metal complex of ruthenium with antimetastatic properties. Clin Exp Metastasis 1992;10:273e80. [13] Sava G, Pacor S, Zorzet S, Alessio E, Mestroni G. Antitumor properties of dimethyl sulfoxide ruthenium(II) complexes in the Lewis lung carcinoma system. Pharmacol Res 1989;21:617e28. [14] Lentz F, Drescher A, Lindauer A, Henke M, Hilger RA, Hartinger CG, et al. Pharmacokinetics of a novel anticancer ruthenium complex (KP1019, FFC14A) in a phase I dose-escalation study. Anti-Cancer Drugs 2009;20:97e103. [15] Rademaker-Lakhai Jeany M, van den Bongard D, Pluim D, Beijnen Jos H, Schellens Jan HM. A Phase I and pharmacological study with imidazoliumtrans-DMSO-imidazole-tetrachlororuthenate, a novel ruthenium anticancer agent. Clin Cancer Res 2004;10:3717e27. [16] Bouma M, Nuijen B, Jansen MT, Sava G, Flaibani A, Bult A, et al. A kinetic study of the chemical stability of the antimetastatic ruthenium complex NAMI-A. Int J Pharm 2002;248:239e46. [17] Mestroni G, Alessio E, Sava G, Pacor S, Coluccia M, Boccarelli A. Water-soluble ruthenium(III)-dimethylsulfoxide complexes: chemical behavior and pharmaceutical properties. Met-Based Drugs 1994;1:41e63. [18] Sava G, Bergamo A, Zorzet S, Gava B, Casarsa C, Cocchietto M, et al. Influence of chemical stability on the activity of the antimetastasis ruthenium compound NAMI-A. Eur J Cancer 2002;38:427e35. [19] Galanski M, Keppler BK. Searching for the magic bullet: anticancer platinum drugs which can be accumulated or activated in the tumor tissue. Anti-Cancer Agents Med Chem 2007;7:55e73. [20] Kelland L. The resurgence of platinum-based cancer chemotherapy. Nat Rev Cancer 2007;7:573e84. [21] Levina A, Mitra A, Lay PA. Recent developments in ruthenium anticancer drugs. Metallomics 2009;1:458e70.

3782

G. Mangiapia et al. / Biomaterials 33 (2012) 3770e3782

[22] Mangiapia G, Vaccaro M, D’Errico G, Frielinghaus H, Radulescu A, Pipich V, et al. Cubosomes for ruthenium complex delivery: formulation and characterization. Soft Matter 2011;7:10577e80. [23] Vaccaro M, Del Litto R, Mangiapia G, Carnerup AM, D’Errico G, Ruffo F, et al. Lipid based nanovectors containing ruthenium complexes: a potential route in cancer therapy. Chem Commun; 2009:1404e6. [24] Mangiapia G, Accardo A, Lo Celso F, Tesauro D, Morelli G, Radulescu A, et al. Mixed micelles composed of peptides and gadolinium complexes as tumorspecific contrast agents in MRI: a SANS study. J Phys Chem B 2004;108: 17611e7. [25] Vaccaro M, Accardo A, Tesauro D, Mangiapia G, Lof D, Schillen K, et al. Supramolecular aggregates of amphiphilic gadolinium complexes as blood pool MRI/MRA contrast agents: physicochemical characterization. Langmuir 2006;22:6635e43. [26] Barthelemy P. Nucleoside-based lipids at work: from supramolecular assemblies to biological applications. CR Chim 2009;12:171e9. [27] Berti D, Bombelli FB, Fortini M, Baglioni P. Amphiphilic self-assemblies decorated by nucleobases. J Phys Chem B 2007;111:11734e44. [28] Fortini M, Berti D, Baglioni P, Ninham BW. Specific anion effects on the aggregation properties of anionic nucleolipids. Curr Opin Colloid Interface Sci 2004;9:168e72. [29] Arriaga LR, Lopez-Montero I, Monroy F, Orts-Gil G, Farago B, Hellweg T. Stiffening effect of cholesterol on disordered lipid phases: a combined neutron spin echo þ dynamic light scattering analysis of the bending elasticity of large unilamellar vesicles. Biophys J 2009;96:3629e37. [30] D’Errico G, Silipo A, Mangiapia G, Molinaro A, Paduano L, Lanzetta R. Mesoscopic and microstructural characterization of liposomes formed by the lipooligosaccharide from Salmonella minnesota strain 595 (Re mutant). Phys Chem Chem Phys 2009;11:2314e22. [31] D’Errico G, Silipo A, Mangiapia G, Vitiello G, Radulescu A, Molinaro A, et al. Characterization of liposomes formed by lipopolysaccharides from Burkholderia cenocepacia, Burkholderia multivorans and Agrobacterium tumefaciens: from the molecular structure to the aggregate architecture. Phys Chem Chem Phys 2010;12:13574e85. [32] Simeone L, Mangiapia G, Irace C, Di Pascale A, Colonna A, Ortona O, et al. Nucleolipid nanovectors as molecular carriers for potential applications in drug delivery. Mol Biosyst 2011;7:3075e86. [33] Covington AK, Bates RG, Durst RA. Definition of pH scales, standard reference values, measurement of pH and related terminology. Pure Appl Chem 1983; 55:1467e76. [34] Berne BJ, Pecora R. Dynamic light scattering: with applications to chemistry, biology, and physics; 1975. [35] Vergara A, Paduano L, Sartorio R. Multicomponent diffusion in systems containing molecules of different size. 4. Mutual diffusion in the ternary system tetra(ethylene glycol)-di(ethylene glycol)-water. J Phys Chem B 2001;105: 328e34.

[36] Paduano L, Sartorio R, Vitagliano V. Diffusion coefficients of the ternary system a-cyclodextrin−sodium benzenesulfonate−water at 25  C: the effect of chemical equilibrium and complex formation on the diffusion coefficients of a ternary system. J Phys Chem B 1998;102:5023e8. [37] Russell TP, Lin JS, Spooner S, Wignall GD. Intercalibration of small-angle x-ray and neutron scattering data. J Appl Crystallogr 1988;21:629e38. [38] Wignall GD, Bates FS. Absolute calibration of small-angle neutron scattering data. J Appl Crystallogr 1987;20:28e40. [39] Kotlarchyk M, Chen SH. Analysis of small angle neutron scattering spectra from polydisperse interacting colloids. J Chem Phys 1983;79:2461e9. [40] D’Errico G, D’Ursi AM, Marsh D. Interaction of a peptide derived from glycoprotein gp36 of feline immunodeficiency virus and its lipoylated analogue with phospholipid membranes. Biochemistry 2008;47:5317e27. [41] D’Errico G, Vitiello G, Ortona O, Tedeschi A, Ramunno A, D’Ursi AM. Interaction between Alzheimer’s A beta(25-35) peptide and phospholipid bilayers: the role of cholesterol. BBA-Biomembranes 2008;1778:2710e6. [42] Irace C, Scorziello A, Maffettone C, Pignataro G, Matrone C, Adornetto A, et al. Divergent modulation of iron regulatory proteins and ferritin biosynthesis by hypoxia/reoxygenation in neurons and glial cells. J Neurochem 2005;95: 1321e31. [43] Frielinghaus H. Small-angle scattering model for multilamellar vesicles. Phys Rev E Stat Nonlinear Soft Matter Phys 2007;76. 051603/1e/8. [44] Kotlarchyk M, Ritzau SM. Paracrystal model of the high-temperature lamellar phase of a ternary microemulsion system. J Appl Crystallogr 1991;24:753e8. [45] Triolo R, Magid LJ, Johnson Jr JS, Child HR. Small-angle neutron scattering from aqueous micellar solutions of a nonionic surfactant as a function of temperature. J Phys Chem 1982;86:3689e95. [46] Hansen S. Translational friction coefficients for cylinders of arbitrary axial ratios estimated by Monte Carlo simulation. J Chem Phys 2004;121:9111e5. [47] Campbell RB, Fukumura D, Brown EB, Mazzola LM, Izumi Y, Jain RK, et al. Cationic charge determines the distribution of liposomes between the vascular and extravascular compartments of tumors. Cancer Res 2002;62: 6831e6. [48] Groessl M, Reisner E, Hartinger CG, Eichinger R, Semenova O, Timerbaev AR, et al. Structure-activity relationships for NAMI-A-type complexes (HL)[transRuCl4L(S-dmso)ruthenate(III)] (L ¼ imidazole, indazole, 1,2,4-triazole, 4amino-1,2,4-triazole, and 1-methyl-1,2,4-triazole): aquation, redox properties, protein binding, and antiproliferative activity. J Med Chem 2007;50: 2185e93. [49] Tan C, Wu S, Lai S, Wang M, Chen Y, Zhou L, et al. Synthesis, structures, cellular uptake and apoptosis-inducing properties of highly cytotoxic rutheniumNorharman complexes. Dalton Trans 2011;40:8611e21. [50] Pluim D, van Waardenburg RCAM, Beijnen JH, Schellens JHM. Cytotoxicity of the organic ruthenium anticancer drug NAMI-A is correlated with DNA binding in four different human tumor cell lines. Cancer Chemother Pharmacol 2004;54:71e8.