Comparing organic and metallo-organic hydrazone molecular cages as potential carriers for doxorubicin delivery

Molecular cages are three-dimensional supramolecular structures that completely wrap guest molecules by encapsulation. We describe a rare comparative study between a metallo-organic cage and a fully organic analogous system, obtained by hydrazone bond formation self-assembly. Both cages are able to encapsulate the anticancer drug doxorubicin, with the organic cage forming a 1 : 1 inclusion complex with μM affinity, whereas the metallo-organic host experiences disassembly by interaction with the drug. Stability experiments reveal that the ligands of the metallo-organic cage are displaced in buffer at neutral, acidic, and basic pH, while the organic cage only disassembles under acidic conditions. Notably, the organic cage also shows minimal cell toxicity, even at high doses, whilst the doxorubicin-cage complex shows in vitro anti-cancer activity. Collectively, these results show that the attributes of the pure organic molecular cage are suitable for the future challenges of in vivo drug delivery using molecular cages.


General equipment
NMR Experiments. 1 H, 13 C and 19 F NMR spectra were recorded on a Bruker FT-NMR Avance 400 (Ettlingen, Germany) spectrometer at 300K, using TMS as an internal standard or a Bruker Neo500 spectrometer, with a 500 MHz Bruker (11.7 T) standard mouth (54mm) shielded magnet and two radio frequency channels, a magnetic field gradients unit equipped with BBOF Plus ATM.1H/BB/19F direct multinuclear probe (5mm) with field gradients on the z-axis and automatic frequency control, 1H channel, wideband channel (5N to 31P and direct observation of 19F), BBI.1H/BB/ inverse probe (5mm) with field gradients on the z-axis.Wideband channel frequency interval of 5N to 31P.BCUII, accessory for air temperature control up to -40ºC.Liquid nitrogen evaporator for experiments at lower temperatures.Chemical shifts (δ) are reported in parts per million (ppm) and referenced to residual solvent.Coupling constants (J) are reported in hertz (Hz).Standard abbreviations indicating multiplicity are used as follows: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet.
High resolution Mass Spectrometry.HR-ESI-MS experiments were performed using a SYNAPT XS high-definition mass spectrometer (Waters Corporation, Manchester, UK) equipped with an electrospray ionization (ESI) source.Capillary voltage was set to 1.5 kV operated in the positive ionization mode and in the resolution mode.Source settings were adjusted to keep intact the molecular cages.Typical values were cone voltage 20−40 V and source offset 4 V; source and desolvation temperatures were set to 110 and 350 °C, respectively.Cone and desolvation gas flows were 150 and 500 (L/h), respectively.Characterisation of molecular cages were performed by MALDI-TOF/TOF mass spectrometry experiments using a TIMS-TOFF Flex (Bruker) in MALDI operation, in reflector positive mode at 1000-5800 m/z rang and a laser intensity of 60 %.The analyses were performed in the mass spectrometry and proteomics facilities of SCSIE University of Valencia.
Absorption and emission Spectrometric measurements.Optical extinction spectra were recorded using a JASCO V-650 UV/vis spectrophotometer with a Single monochromator, 1200 lines/mm plane grating, Czerny-Turner mount, Double-beam equipped with a deuterium lamp (190 to 350 nm) and a Halogen lamp (330 to 900 nm), and wavelength range of 190 to 900 nm.Fluorescence spectroscopy was carried out on a JASCO FP-8300 spectrofluorometer (Hitachi High Technologies) with high resolution of 1.0 nm and a wavelength range of 200 to 750 nm.Titration experiments have been carried out with a Perkin Elmer EnSpire 2300 Multimode Plate Reader equipped with fluorescence (monochromator wavelength range of 230 to 850 nm with an excitation and emission bandwidth of 5 nm), absorbance (monochromator wavelength range of 230 to 1000 nm with monochromator bandwidth of 5 nm) and luminescence detection.Compounds 1, [S1] 2•NO3, [S2] 2•BF4, [S3] 3, [S4] S3 [S5] were prepared according to literature procedures.

Synthesis of ligands and cages
S6,S7,S8] 4,4'-oxydibenzoic acid (4.0 g, 15,50 mmol) was dissolved in anhydrous methanol (70 mL) under N2 atmosphere, and 1.0 mL of concentrated H2SO4 was added.The resulting solution was heated at 70 ºC over 24 h to obtain abundant white precipitate.The mixture was cooled down, vacuum filtered and washed with cold methanol to obtain S1 as a pure white solid (4.3 g, 96%).

Synthesis of the Pd (II) Molecular Cage with NO3 -counterion (C1•NO3
). 2•NO3 (100 mg, 151.8 μmol) and 1 (86.9 mg, 303.7 μmol) were completely dissolved in 30 mL of anhydrous DMSO.The resulting solution was stirred at room temperature for 18 h, and then slowly poured into DCM (70 mL) until the formation of abundant white precipitate.The suspension was centrifuged at 10,000 rpm for 10 minutes, the supernatant solution was decanted, and the precipitate was dried out to obtain C1•NO3 pure as a whitegrey solid (162 mg, 52 %).

Synthesis of the Pd (II) Molecular Cage with BF4 -counterion (C1•BF4
). 2•BF4 (100 mg, 141.2 μmol) and 1 (80.9 mg, 282.5 μmol) were completely dissolved in of anhydrous DMSO (8.0 mL).The resulting solution was stirred at room temperature for 18 h, and then slowly poured into DCM (100 mL) until the formation of abundant white precipitate and left to sediment.The supernatant solution was removed, and the precipitate was dried out to obtain C1•BF4 pure as a white-grey solid (52 mg, 52 %).

X-Ray crystallographic data for C1•NO3
Single crystals of C1•NO3 were grown by vapor diffusion of CH3OH to a solution of the cage in DMSO over a week.A suitable crystal was selected and placed on a Bruker D8 Venture Diffractometer.The crystal was kept at 120.0 K during data collection.Using Olex2 [S9] , the structure was solved with the SHELXS [S10] structure solution program using Direct Methods and refined with the SHELXL [S11] refinement package using Least Squares minimisation.The structure contains disordered solvent molecules that could not be modeled, the corresponding contribution of disordered solvent molecules was handled by the solvent mask command in Olex2.The structure has 4 nitrate counter anions, only two of them could be located inside the cavity and the other two nitrate anions are presumably placed outside the cavity with a high disorder making it not possible to find them.The two nitrate anions in the cavity could be located, one of them without disorder, and the second one was modeled with a disorder over 3 positions, being only possible to fully find it in one of these 3 disordered positions.A twin law was used to model the enantiomeric disorder, i.e. the cage has a small helicity at the 50-50 ratio.CCDC deposition number 2295536.Crystal Data for C1•NO3: monoclinic, space group P21 (no.4), a = 19.6620(13

NMR spectrometric host-guest experiments and cage disassembly
NMR titration experiments were carried out on an AVA400 NMR spectrometer equipped with a BBFO + room temperature probe featuring two channels: 1 H and X/ 19 F (optimised).
The titration experiments were performed using a 10 mM solution of DOXO in CD3OD and a 500 μM solution of C1•BArF in a DCM-d2/CD3OD (9:1) mixture in order to achieve the complete dissolution of the molecular cage and DOXO.The spectra stacked in Figure S24 show chemical shifts after the addition of increasing equivalents of DOXO to a C1•BArF solution.In previous experiments, it was observed that the presence of water promotes precipitation of the cage, which makes C1 not suitable for biological applications.Increasing equivalents of DOXO eventually leads to the disassembly of the molecular cage.The low solubility of C2 in DMSO-d6/D2O mixtures for NMR experiments led to carrying out the titration experiments with DOXO in a DMSO-d6/CD3OD (8:2) mixture (Figure S25).The chemical shifts of C2 suggest the encapsulation of DOXO.However, the weak interaction observed is likely associated with the high amount of DMSO-d6 present in the sample.Despite that the changes observed in the chemical shifts in the titrations, they show binding between the host at the guest (Figure S26).
The upfield shift of the inward-facing Hf protons evidences the encapsulation of DOXO inside the cavity.While downfield shift of the outer protons Hh and Hj upon DOXO binding are also observed, these could be due to subtle conformational changes (e.g., a twisting of the hydrazone group leading to a change in a N•••Hh-C H-bond) or electronic effect-communication through the ring system.S12,S13 The lack of chemical shift change in Hk suggests that the binding mode is different to what is normally observed in Pd2L4 cages, wherein the inward facing ortho-pyridyl H atoms create a H-bond donor pocket that that can interact with H-bond acceptor groups.S12 It should also be noted that increasing amounts of guest (>0.6 eq) produce a decrease of integration of cage signals suggestive of disassembly, possibly by complexation of Pd 2+ by the NH2 moiety of DOXO.This again highlights the challenges in using metallo-organic cages in bio-medical applications.
For cage C2, it was found that DMSO-d6/CD3OD (8:2) was an optimal mixture in terms of cage solubility to probe binding by NMR (Figure S26b).The addition of DOXO revealed upfield changes of the protons inside the cavity Hh (∆δ = -0.020ppm), Hk/k' (∆δ = -0.02ppm), Hl/l' (∆δ = -0.02ppm) and Hg (∆δ = 0.04 ppm).The upfield shift in three of these signals is consistent with the formation of CH••• interaction between the anthracycline system of DOXO and the inward-facing protons of C2.Similar shielding effects have been observed for many different cage systems.

Experimental details
All UV-visible and spectrofluorimetric titration experiments were carried out on a Perkin Elmer EnSpire 2300 Multimode Plate Reader.The titration experiments and the Job Plot method were performed at room temperature using a 96-well plate.The titration experiments were carried out by adding increasing equivalents (0-2.3 or 2.8) of 4 mM (C2) or 10 mM (C1) molecular cage stock solutions in DMSO to a 50 μM solution of DOXO in phosphate buffer (without NaCl) at pH 7.1.The measurements were made with maximal volumes of 250 μL and a 50 μM concentration.All experiments were performed with phosphate buffer (without NaCl) at pH 7.1 or milli-Q water with a percentage of DMSO (0-10%).All emission spectra were recorded in a 700-250 nm wavelength range, while excitation spectra were recorded in a 510-700 nm wavelength range for spectrofluorimetric measurements with an excitation wavelength of 470 nm (excitation wavelength for DOXO).The binding stoichiometry between the molecular cage (C1 or C2) with DOXO was determined by Job Plot's Method (or Method of Continuous Variation) by fluorimetry titration experiments.In a 96-well plate, serial molecular cage's (C1 or C2) solutions were placed in phosphate buffer (without NaCl) at pH 7.1 with decreasing concentration, to which an increasing number of equivalents of DOXO were added to have a 0-1.0 molar concentration series of both species with a 0.05 molar fraction range.The titrations were performed in quadruplicate, and the standard error was obtained for each titration point.A host-guest release experiment was carried out on a JASCO FP-8300 spectrofluorometer with an excitation wavelength of 470 nm and the emission spectra were recorded in the 510-700 nm range.Two 50 μM stock solutions of DOXO were prepared with phosphate buffer (without NaCl) at pH 7.2 and 6% of DMSO (one was used as reference).In one of the solutions, the 1.0 equivalent of C2 was added to completely encapsulate the DOXO inside the cage's cavity.To perform the release experiment, an increasing amount of DMSO was added (6-55%) reducing the hydrophobic effect, as well as producing a dilution of the samples from 50 μM to 25 μM.The fluorescence spectra were recorded after each addition of DMSO (Figure S31).

Rehm-Weller calculations
S17] ΔG° = NA e ([E°(D+ In Eq 1 e is the elementary charge, NA is the Avogadro constant, E°(D+•/D) is the standard electrode potential of the donor cation radical resulting from the electron transfer, E°(A/A-•) is the standard electrode potential of the acceptor and SE is the energy difference between the fundamental electronic state and the first singlet excited state in eV.
Calculations were done using values obtained from the literature to the constituent parts of the cages and hence there is some degree of uncertainty in the calculations.Nevertheless, the high exergonicity of the calculated ΔG° is qualitatively indicative of the feasibility of a PET process.
The obtained ΔG° values for the photoinduced electron transfer were -120 kJ/mol for the PET from resorcinol to doxorubicin, and -84 kJ/mol for the PET from doxorubicin to tetrakis(pyridine)palladium(II).

Molecular Modelling
The structure of cage C2 and the supramolecular complex of encapsulated doxorubicin in cage C2 were modelling with the Spartan' 20 software using the MMFF force field and a standard optimization.The cavity volume of the cages was determined from the corresponding cage structureds using the CageCavityCalc Python script, [S24] obtaining 1350 Å³ for cage C1, and 1500 Å³ for cage C2.The the volume of DOXO (498 Å³) was determined with the Wavefunciton Spartan software.

Cell culture
Murine 4T1 triple-negative breast cancer cells and human melanoma SK-Mel-103 cells were obtained from ATCC and cultured in DMEM medium (Sigma), supplemented with 10% FBS (Sigma).Cells were incubated in an atmosphere of 5% CO2 at 37 ⁰C.

Cell viability/Cytotoxicity assays
Both cell lines (4T1 and SK-Mel-103) were plated in 96-well plates at a density of 2500 cells per well.
For the assays with the components of the organic cage, SK-Mel-103 cells were plated in 96-well plates at a density of 2500 cells per well.After 24 hours, cells were treated with 50 μM of calixarene, with 100 μM of linker and with both components together.After 48 h of treatment, viability was measured using the WST-1 method.

Drug-encapsulation in vitro assay
Cells were plated in 96-well plates at a density of 5000 cells/well.24 h later they were treated either with free doxorubicin at concentrations 5 μM, 2.5 μM and 1.25 μM or with the complex organic cage-doxorubicin, prepared at a fixed concentration of 25 μM of organic cage and different concentrations of doxorubicin (5, 2.5 and 1.25 µM), resulting in a 95 %, 96 % and 96 % of encapsulation of the drug, respectively.After 24 h of treatment, WST-1 was added to the cells according to manufacturer's instructions and absorbance at 595 nm was measured in Wallac 1420 Victor2 spectrophotometer.
For the assays with the components of the organic cage, SK-Mel-103 cells were plated in 96-well plates at a density of 5000 cells/well.24 h later they were treated with 50 µM of calixarene, 100 µM of ligand and different concentrations of doxorubicin (5, 2.5 and 1.25 µM).After 24 h of treatment, viability was measured using the WST-1 method.
Internalization of the organic cage-doxorubicin complex SK-Mel-103 cells were cultured in 6-well plates at 350,000 cells/well for 24 h.Then, cells were incubated with the nuclei marker Hoechst 33342 at a concentration of 1 μg/mL for 30 min.For the time-lapse experiment, the cell membrane marker Wheat Germ Agglutinin, Alexa Fluor™ 647 Conjugate (Invitrogen, W32466) was added at a concentration of 1 μg/mL 10 minutes before starting imaging.After the incubation with markers, cells were washed with PBS (Merck, D8537) and treated either with the organic-cage complex, formed previously with a mix of 25 μM of organic cage and 5 μM of doxorubicin (95 % of encapsulation).Similarly, the cells were treated with free doxorubicin as a control.A time-lapse up to 10 S41 minutes since treatment was performed for every condition in a confocal Leica TCS SP8 HyVolution II microscope, equipped with CO2 and temperature control and a resonant scanner for live-cells studies.

Figure S30 .
Figure S30.Job plot obtained from the fluorescence emission spectra (H2O with 10% DMSO and phosphate buffer 1mM, pH 7.1, rt) of the titration of doxorubicin with cage C-2 (total concentration 50 μM).The blue line represents the fitting to the 1:1 Host-Guest binding model.

Figure S31 .
Figure S31.Doxorubicin release of cage encapsulated doxorubicin.Experiment conditions: equimolar solution containing cage (50 μM) and doxorubicin (50 μM) in phosphate buffer (1 mM, pH 7.2) with increasing amounts of DMSO from 6% to 55%.The green dot indicates the release at 2% DMSO as determined in the binding experiments.The blue line indicates the expected release changes by dilution from 50 μM to 25 μM in a solution containing 2% DMSO.

Figure S36 .
Figure S36.Cavity volume of the cages C1 (left) and C2 (right) determined from the corresponding cage structureds using a the CageCavityCalc Python script based on the rolling prove algorithm.

Table S1 .
Crystal data and structure refinement for C1•NO3.
[S23 .The geometries and the XYZ coordinates are provided below.