Exploring a flexible and cytotoxic drug carrier of cisplatin and 5-fluorouracil for a multitarget therapeutic approach in colorectal cancer

Abstract

The therapeutic potential of many anticancer drugs is frequently hindered by challenges such as non-specific distribution, suboptimal dosing, and premature degradation, which collectively compromise treatment efficacy. To overcome these limitations, advanced drug delivery systems capable of targeted, controlled, and synchronised release are essential. This study presents the development and complete characterisation of the flexible supramolecular metal–organic framework (SMOF) Cu₇Naph as a multifunctional carrier for the co-delivery of the chemotherapeutic agents cisplatin (cisPt) and 5-fluorouracil (5-FU), aiming to enhance therapeutic efficacy against cancer. The water-stable Cu₇Naph is assembled from heptanuclear copper–adenine units and naphthalene-2,6-dicarboxylate counterions, held together by π–π stacking and hydrogen bonding interactions, which confer high structural flexibility and porosity. Single-crystal X-ray diffraction analyses demonstrate that Cu₇Naph undergoes significant structural expansion or contraction depending on hydration stages and guest molecule inclusion, enabling simultaneous incorporation of cisPt and 5-FU within the same porous matrix. This co-loading results in synergistic effects, increasing 5-FU loading capacity to 14.1 wt% in the presence of cisPt and synchronising their release kinetics, thereby reducing kinetic disparity (K_5-FU/K_cisPt = 2.5 versus 4.2 when loaded separately). The first stage of the drug release follows pseudo-first-order kinetics under physiologically relevant conditions (35 °C). Cytotoxicity assays using HCT116 colorectal cancer cells cultured in the presence of Cu₇Naph reveal that Cu₇Naph exhibits intrinsic antiproliferative activity, which is enhanced upon 5-FU loading but attenuated with cisPt inclusion, suggesting a possible interaction between cisPt and the carrier's cytotoxic mechanism. Transcriptomic analysis via RNA sequencing identifies downregulation of AKR1A1 and PUF60 genes as contributors to the observed biological effects. Collectively, these findings highlight the potential of structurally adaptable SMOFs as versatile platforms for the synchronised co-delivery of multiple drugs with distinct release profiles and therapeutic mechanisms, offering a promising strategy for improved drug combination cancer therapies.

---

1. Introduction

Cancer is currently one of the leading causes of death worldwide, and as a result, an increasing amount of research is being devoted to its prevention and treatment. Among the latter, there is a push to develop more effective and specific therapeutic strategies¹ because, despite the progress made in this field, cancer treatment still faces significant limitations. These include the low specificity of conventional drugs, their rapid degradation in the body and the emergence of resistance by tumour cells due to their high capacity to adapt and evolve.^2–4 There are several works that have proved how drug carriers can help in dealing with low specificity and drug degradation. Drug carriers comprise a vast variety of materials: vesicles,⁵ nanoparticles,⁶ hydrogels,⁷ foams,⁸ and more recently metal–organic frameworks (MOFs).⁹ MOFs are porous structures made up of metal ions coupled with organic ligands. These hybrid materials are porous and tuneable, with an extensive surface area, a high degree of porosity, and remarkable chemical stability.¹⁰ MOFs have demonstrated effectiveness in the delivery of drugs by acting as transporters of nucleic acids,¹¹ proteins,¹² and drug molecules.¹³ A related family of porous materials is that of supramolecular metal–organic frameworks (SMOFs), which rely on discrete metal–organic entities that are assembled together by means of hydrogen-bonding and/or π–π stacking interactions to build up a porous supramolecular network.^14–16 Among them, those based on [Cu₇(μ-adeninato-κN3:κN9)₆(μ₃-OH)₆(μ-OH₂)₆]²⁺ entities show great stability in water (pH = 2–11), insolubility, tailorable crystal structure and porosity based on the counterion employed, and structural breathing behaviour.^17–20 As these features facilitate their use as drug delivery systems, some of these compounds were tested for the loading and release of several drug molecules (5-fluorouracil, 4-aminosalicylic acid, 5-aminosalicylic acid and allopurinol).^20,21 Cytotoxic studies carried out on human colon carcinoma cell line HCT116 proved the effectiveness of this approach. Interestingly, the cytotoxic behavior of these porous materials changes based on the counterion employed. The obtained results in an analogue compound with a Cu₆Cr metal center indicated that sulfate counterions impart a non-cytotoxic behavior to the SMOF, a proliferative effect is even detected;²¹ whereas the use of naphthalene-2,6-dicarboxylate counterions produces, in contrast, a cytotoxic response in intestinal HCT116 cell cultures.²⁰

Drug resistance is a great concern when designing therapeutic treatments. Once drug resistance by malignant tumour cells has emerged, it is usually addressed with the development of new drugs or modulators that hinder the resistance mechanism of the tumoral cell.²² However, to diminish the possibility of the emergence of this drug resistance it is common to use drug cocktails in which each drug has a different therapeutic mechanism making it complicated for the tumoral cell to generate resistance to all these drugs simultaneously.²³ In this sense, to ensure that all drug molecules in the cocktail therapy arrive at the tumoral cells simultaneously is very desirable to provide them with some form of delivery that ensures that concurrency. When the drug molecules are loaded into separate carriers, the release kinetics can be significantly different because the carriers are different or because the features of the drug molecules (size, shape, chemical nature) are different. Therefore, in this work we aim to load the drug molecules of the cocktail together in the same carrier, one of these SMOFs. We will demonstrate how this approach affects the release kinetics of drug molecules. Slower-releasing molecules hinder faster-releasing ones, bringing their release kinetics closer together. The structurally flexible nature of these SMOFs also helps in loading drugs with different sizes.

However, to go a step further, we decided that the carrier itself should also have some kind of therapeutic capacity. In addition to the selected drug molecules (5-fluorouracil, 5-FU, and cisplatin, cisPt), employed for the treatment of cancer because of their cytotoxicity, we have also selected a cytotoxic SMOF as a carrier, in an attempt to increase the effectiveness of this cocktail therapy. 5-FU and cisPt are widely used in many well established cancer therapies and, because of their distinct mechanism of action. 5-FU acts as an antimetabolite, inhibiting DNA, RNA and protein synthesis by blocking the conversion of deoxyuridylic acid to thymidylic acid. Cisplatin exerts its cytotoxic activity by forming covalent bonds with DNA, preventing cell replication and inducing apoptosis, particularly in rapidly proliferating cells. These two drugs are already being employed in combination to treat several different types of cancers including head and neck, esophageal and anal cancer.²⁴ The selected cytotoxic SMOF, [Cu₇(μ-adeninato)₆(μ₃-OH)₆(μ-OH₂)₆](naphthalene-2,6-dicarboxylate)·nH₂O, consists of heptameric cations and naphthalene-2,6-dicarboxylate anions that are held together not only by electrostatic interactions but also by π–π stacking interactions between the adeninato ligands of adjacent units and with the planar and aromatic naphthalene-2,6-dicarboxylate anion.²⁰ These π–π stacking interactions are less rigid with respect to the in plane rotation of the stacked adeninato ligands and the dicarboxylate anion, a feature that leads to a great flexibility in the resulting porous supramolecular structure. In addition to the above-mentioned cytotoxicity of this compound, it was also previously reported that it can be loaded, among other drug molecules, with 5-FU up to a 58 wt% and that the release follows a pseudo first-order kinetics with a t_1/2 of 2.6 h at 35 °C when it is in a saturated medium.

This work will focus firstly on a detailed crystallographic analysis of the structural flexibility of this SMOF (Cu₇Naph) when the water content inside the pores of the material changes and when a drug molecule (5-FU) is loaded inside these pores. Secondly, we will take advantage of this structural flexibility to load cisplatin together with 5-FU (Scheme 1) and to measure the difference on the release kinetics when each drug is loaded alone or in combination with the other to check if their kinetics become closer when they are loaded together. Thirdly, RNA-seq transcriptomic studies will be carried out to determine the mechanism responsible for the cytotoxic behaviour of this SMOF.

Scheme 1 Front view of the cationic heptanuclear entity of Cu₇Naph and naphthalene-2,6-dicarboxylate anion (colour code: H, white; C, grey; N, blue; O, red and Cu, orange). Supramolecular porous structure highlighting the π–π interactions (double black line), the contact surface of the pores and the drugs to be loaded (inset).

Finally, cytotoxic studies will be carried out to verify if the combination of these two drugs plus the carrier cytotoxic combination does in fact produce a greater cytotoxic effect than when employed separately.

2. Experimental

2.1. Chemicals

The following reagents were used for the preparation of the Cu₇Naph compound: copper(II) nitrate trihydrate (Sigma-Aldrich, 98%), adenine (Sigma-Aldrich, >99%), naphthalene-2,6-dicarboxylic acid (Sigma-Aldrich, 95%), methanol (Labkem, AGR), nitric acid (Labkem, 65% EPR) and sodium hydroxide (Labkem, 98%). The used drugs were 5-FU (5-fluorouracil, TCI, >99%) and cisplatin (cis-diamminedichloridoplatinum(II), Sigma-Aldrich, >99%).

2.2. Synthesis of compounds [Cu₇(μ-adeninato)₆(μ₃-OH)₆(μ-OH₂)₆](naphthalene-2,6-dicarboxylate)·∼nH₂O (Cu₇Naph)

The synthesis of the compound Cu₇Naph was carried out using a procedure previously reported.²⁰ A mixture of methanol–water (20 mL, 1 : 1) containing 0.081 g (0.60 mmol) of adenine was heated at 60 °C, and then added to a solution of Cu(NO₃)·3H₂O (0.122 g, 0.50 mmol) in 10 mL of deionized water. The resulting mixture, pH ∼4.0, was basified with a NaOH solution until pH ∼9. Concurrently, naphthalene-2,6-dicarboxylic acid (0.141 g, 0.65 mmol) was dissolved in 15 mL of deionized water and basified with NaOH to pH ∼11 and added to the first solution. Subsequently, the pH was adjusted to ∼9.2 by the dropwise addition of a dilute HNO₃ solution (1 : 3). After 6–7 days, blue square single crystals of Cu₇Naph were obtained (yield: 85%). The water content of this compound is variable depending on the storage conditions. The freshly obtained crystals contain 32 crystallization water molecules per formula (Cu₇Naph·32H₂O),²⁰ but after being kept filtering under vacuum, this value reduces to 20 (Cu₇Naph·20H₂O), and when kept under open atmosphere conditions for 24 h, the water content slightly reduces even more to 15 (Cu₇Naph·15H₂O). The samples with a reduced water content reverted to the initial Cu₇Naph·32H₂O when exposed to a moisture-saturated atmosphere. Fortunately, the crystals of this compound were able to withstand these conditions, as to be suitable for single-crystal X-ray diffraction (SC-XRD) and provide an insight into the structural differences that accompany these changes in the crystallization water molecule content.

2.3. Synthesis of the compound [Cu₇(μ-adeninato)₆(μ₃-OH)₆(μ-OH₂)₆](naphthalene-2,6-dicarboxylate)·2(5-fluorouracil)·26H₂O (Cu₇Naph·2(5-FU)·26H₂O)

Crystals of Cu₇Naph (50 mg) were placed in a capped vial containing an aqueous saturated solution of 5-fluorouracil. The vial with the resulting mixture was kept in a temperature-controlled box for one week at 30 °C. After that time, a crystal specimen was selected for SC-XRD structural characterization.

2.4. Loading of 5-fluorouracil and cisplatin in Cu₇Naph

The procedure for the loading of the drug molecules consisted of placing 50 mg of Cu₇Naph in 10 mL of an aqueous solution containing either 50 mg of 5-FU (5-FU@Cu₇Naph) or 21 mg of cisplatin (cisPt@Cu₇Naph) or a mixture of 50 mg of 5-FU and 21 mg of cisplatin (5-FUcisPt@Cu₇Naph). The vials were kept stirring in a rotatory mixer for 72 h at 40 °C. After that time, the particles were recovered by filtration. Both the drug@Cu₇Naph particles and the remaining solution were kept for further characterization.

2.5. Physical methods

The purity of the bulk samples was ascertained by powder X-ray diffraction (PXRD), Fourier-transform infrared (FTIR) spectroscopy and thermogravimetric analysis (TGA). Routine PXRD measurements were performed using a Phillips X’PERT diffractometer (equipped with Cu Kα radiation, λ = 1.5418 Å) over the range of 5 < 2θ < 70° with a step size of 0.02°, a variable automatic divergence slit, and an acquisition time of 2.5 s per step at 293 K (Fig. S1, SI). FTIR spectra of the samples (KBr pellets) were recorded at a resolution of 4 cm⁻¹ in the 4000–320 cm⁻¹ region using an FT/IR-8X Jasco spectrometer. TGA was performed on a Mettler Toledo TGA/SDTA851 thermal analyzer in a synthetic air (80% N₂, 20% O₂) flux of 50 cm³ min⁻¹, from room temperature to 600 °C with a heating rate of 5 °C min⁻¹ and a sample amount of about 10–20 mg per run.

SC-XRD data for structure determination were collected with an Agilent Technologies Supernova diffractometer (λ_CuKα = 1.54184) for Cu₇Naph·32H₂O and Cu₇Naph·20H₂O; λ_Mokα = 0.71073 Å for Cu₇Naph·15H₂O and Cu₇Naph·2(5-FU)·26H₂O. Data reduction was performed with the CrysAlisPro program.²⁵ The crystal structures were solved by direct methods using SIR92²⁶ or SUPERFLIP²⁷ and refined by full-matrix least-squares on F², including all reflections with SHELXS,²⁸ within the WINGX crystallographic software package.²⁹ Several structures exhibited disorder in adeninato ligands and the aromatic ring of the dicarboxylic ligands. The disorder was modeled by distributing the disordered atoms over two positions and constraining the sum of their occupation factors to one. Some of the highly disordered solvent molecules precluded their modeling, and consequently, the electron density was subtracted from the reflection data by the SQUEEZE method²⁹ as implemented in PLATON.³⁰ Full crystallographic details are provided in Table S1 (SI).

In addition to the above techniques, X-ray fluorescence (XRF), high performance liquid chromatography (HPLC), inductively coupled plasma-atomic emission spectroscopy (ICP-AES) and scanning electron microscopy (SEM) were employed to analyse the material post-drug loading and to quantify the amount of drug that had been loaded/released. XRF measurements to semi-quantitatively analyse the platinum content in the samples were obtained using a MIDEX SD X-ray microfluorescence spectrometer (Spectro), ED-XRF energy dispersion for elemental analysis. The instrument is based on an automatic XYZ stage and collimator changer, a Mo X-ray tube with a maximum power of 40 W/voltage of 48 kV and a silicon drift detector (SDD) with an area of 30 mm². Calibration and calculations were performed using FP Plus fundamental parameters.

Quantitative analysis of 5-FU in the adsorption and desorption processes was performed using a high-performance liquid chromatography (HPLC) system model Shimadzu Nexera equipped with an SPD-20/40V ultraviolet-visible (UV-vis) detector. The separation was performed using a Shim-pack VP-ODS column (4.6 mm × 250 mm, 5 μm particle size) maintained at 40 °C. The UV-vis detector was set to a wavelength of 265 nm for the quantification of 5-FU. The mobile phase consisted of an isocratic 90 : 10 v/v mixture of water (phase A) and methanol (phase B). The flow rate was kept at 1 mL min⁻¹, with an injection volume of 60 μL. The calibration was carried out using 5-FU solutions in the concentration range of 10–70 ppm prior to the measurement of the pre- and post-adsorption 5-FU.

The platinum content in the solutions during the cisplatin loading and release experiments was quantified using an ICP-AES with an Agilent 5800 spectrometer in the axial mode. Argon plasma flow was 12 L min⁻¹, nebulization was 0.7 mL min⁻¹ and auxiliary gas input was 1.0 mL min⁻¹ with a power of 1300 W.

The distribution of copper, platinum and fluorine atoms in the sample crystallites was observed using a JEOL JSM-7000F scanning field emission SEM equipped with a Schottky-type gun, secondary electron (SE) detector, backscattered electron (BSE) detector, and Oxford's INCA X-sight Series Si(Li) penta-FET energy-dispersive X-ray spectroscopy (EDS) detector, enabling spot, line, and mapping microanalysis. SEM analysis was performed at different magnifications (200×, 1 k×) using an accelerating voltage of 20 kV and an approximate working distance of 10 mm. The samples for SEM were adhered to a holder using a double-sided adhesive carbon tape and coated with carbon by sputtering using a Q150T sample preparation kit (Quorum Technologies Ltd.)

2.6. Cytotoxicity and RNA-seq transcriptomic analysis

To test the cytotoxicity, the human colorectal cancer line HCT116 (ATCC–CCL247) was employed. Cells were seeded at 3000 cells per well in 96-well plates and incubated at 37 °C and 5% CO₂ in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS and antibiotics. After overnight incubation, the compounds were added to the cells and the cells were collected for crystal violet staining at 24, 48 and 96 h incubation time points. Subsequently, the incubated cells were fixed with paraformaldehyde and stained with 0.1% crystal violet. The stain was washed 3 times; 10% acetic acid was added, and the plates were incubated by shaking for 20 min. Absorbance was measured at 590 nm and relative proliferation was calculated. The following concentrations of the compounds were used to ensure equal concentration of the host Cu₇Naph and, as close as possible, of the released drugs that could provide comparative results: 27.0 μg/100 μL of compound Cu₇Naph, 29.3 μg/100 μL of 5-FU@Cu₇Naph, 29.0 μg/100 μL of cisPt@Cu₇Naph, 32.7 μg/100 μL of 5-FUcisPt@Cu₇Naph, 1.0 μg/100 μL of 5-FU, and 0.2 μg/100 μL of cisplatin.

For RNA sequencing studies, 30 000 cells were seeded in 12 well plates, compounds were added to the cells after overnight incubation and cells were grown for 72 h. Three subsequent cell passages were used for this experiment. After incubation, RNA was extracted from the cultures using the NucleoSpin RNA kit (Macherey Nagel) and submitted for RNA sequencing analyses to the Genomic Platform of the Biobizkaia Health Research Institute (Barakaldo, Spain).

RNA-sequencing (RNA-seq) transcriptomics data analysis was conducted as previously described.³¹ We used HISAT2³² to align the RNA-seq reads to the human reference genome hg38 and to calculate the Fragments Per Kilobase of transcript per Million fragments mapped (FPKMs) and Cufflinks³³ to annotate them. We merged the transcriptomics data into a single text file and used it in the downstream analysis using in-house functions developed in MATLAB (MathWorks).³⁴ We equalized the data and stabilized them through the log₂ transform of the data plus one – to avoid the undefined values of log₂ of zero; calculated the average values for each of the two groups: NT and Cu₇Naph – of three biological replicates each; selected the differentially expressed genes (DEGs) whose absolute difference in mean values between the two groups was less than the selection threshold θ_DEG = 1 of fold change (FC) in the log₂ scale; and selected the statistically significant DEGs using unpaired Student's t test with equal variances and with a significance threshold of α_DEG = 0.05.

3. Results and discussion

3.1. Structural flexibility upon crystallization water removal

In the present study, we present three novel structures for the system Cu₇Naph, all of which possess the same crystallographic space group P2₁/c (see Table 1 and Table S1). The building units of these structures: heptanuclear metal–organic cationic discrete entities and aromatic organic anions are rigid. The common wheel-shaped [Cu₇(μ-adeninato)₆(μ₃-OH)₆(μ-OH₂)₆]²⁺ heptanuclear entity consists of a central [Cu(OH)₆]⁴⁻ core connected to six outer copper(II) metal centers by μ₃-OH bridges in a radial and planar arrangement (Scheme 1 and Fig. S3). The peripheral copper atoms are connected by a double bridge of semicoordinated water molecules and bidentate μ-κN3:κN9 peripheral adeninato ligands. The copper(II) centers exhibit an octahedral geometry with the conventional Jahn–Teller tetragonal elongation, a feature that is more pronounced for the external Cu atoms. Therefore, the source of the structural flexibility must be attributed to the supramolecular interactions that stabilise these discrete entities. It is well documented that π-stacking interactions are characterised by a high degree of rigidity with respect to the interplanar distance (ca. 3.3–3.5 Å); however, no such rigidity is observed in terms of lateral placement and rotation.^35,36 Indeed, this is the behaviour that is observed in the crystal structures presented in this work when the water content is modified.

Table 1 More relevant crystallographic data of Cu₇Naph·nH₂O compounds

	Cu7Naph·32H2O	Cu7Naph·20H2O	Cu7Naph·15H2O
a (Å)	15.9866(12)	26.1051(10)	13.5818(10)
b (Å)	15.8405(7)	15.9114(6)	15.7993(6)
c (Å)	18.4074(9)	18.6613(8)	18.2487(7)
β (°)	95.424(5)	105.351(4)	96.350(5)
V (Å^3)	4640.5(5)	7474.8(5)	3891.8(4)
Z	2	4	2
V/Z (Å^3)	2320	1946	1869
Voids (%)	43.0	28.5	24.4

---

The structure of Cu₇Naph·32H₂O has been previously reported by our research group.²⁰ The naphthalene-2,6-dicarboxylate counterion is sandwiched by two adeninato ligands from two adjacent heptanuclear entities (Fig. 1a and Fig. S4). This interaction involves two adeninato ligands of the heptanuclear entities; the remaining four adeninato ligands establish direct offset face-to-face π-stacking interactions with other adeninato ligands from four adjacent heptanuclear entities (see Fig. 2a). It is also worth noting that hydrogen bonding interactions are also present, reinforcing the above-described π-stacking interactions, Fig. S5. In this case, a hydrogen-bonded R₂²(8) ring is formed by two O–H⋯N hydrogen bonds between the pyrimidinic N1 and exocyclic N6 nitrogen atoms of the Watson–Crick face of an adeninato ligand (acceptor) and a HO–Cu–OH₂ fragment of a cationic unit (donor), as shown in Fig. 1b. These supramolecular interactions result in the formation of cationic layers, from which the two adeninato ligands interact with the naphthalene-2,6-dicarboxylate anion placed almost perpendicularly to these sheets. The resulting 3D supramolecular architecture is characterised by a high degree of porosity (Fig. 3; void percentage: 43%; pores: 3.6–8.1 Å, Table S2 and Fig. S7 and S8a).

Fig. 1 Arrangement and π-stacking interactions (double dashed black lines) established by the naphthalene-2,6-dicarboxylate anions (green coloured molecule) in the crystal structures of Cu7Naph·32H₂O (a) and (b), Cu7Naph·20H₂O (c) and (d) and Cu7Naph·15H₂O (e) and (f). Double dashed black and blue lines: π-stacking interactions and hydrogen bonds, respectively.

Fig. 2 Supramolecular cationic layer of Cu₇Naph·32H₂O (a), Cu₇Naph·20H₂O (b) and Cu₇Naph·15H₂O (c) formed by heptameric complexes assembled by π-stacking interactions. Double dashed black and blue lines: π-stacking interactions and hydrogen bonds, respectively.

Fig. 3 Crystal packing viewed along the crystallographic b, c and c axes, respectively, in Cu₇Naph·32H₂O (a)–(c), Cu₇Naph·20H₂O (d)–(f) and Cu₇Naph·15H₂O (g)–(i). The organic anion has been depicted in green for clarity. The packing representation in the third row illustrates the contact surface of the pore.

In Cu₇Naph·20H₂O, where the water content has been reduced by one third, this scheme of π-stacking supramolecular interactions is maintained, although the naphthalene-2,6-dicarboxylate anion has rotated almost 90°, disrupting the hydrogen-bonding interaction scheme observed previously and enabling a closer approach between the heptanuclear entities along this direction (crystallographic a axis). This results in a contraction of the unit cell volume (because of the a-axis contraction), which also entails a reduction in the void percentage (28.5%, Fig. S8b) and the pore diameter (2.6–6.8 Å) (see Table S2).

Further reduction of the water content in compound Cu₇Naph·15H₂O results in a crystal structure phase transition, accompanied by the formation of a new unit cell. This new unit cell can be described as being approximately doubled along the crystallographic a axis when compared to those of Cu₇Naph·32H₂O and Cu₇Naph·20H₂O (see Table 1). The heptanuclear entities are held together by means of π-stacking interactions between the adeninato ligands (Fig. 2c). However, it is notable that each heptanuclear entity now interacts with merely two adjacent ones by means of double π-stacking, involving four adeninato ligands (two from each interacting heptanuclear entity). This interaction results in the formation of 1D supramolecular chains of heptanuclear entities, within which the naphthalene-2,6-dicarboxylate anions are incorporated. The organic anion is surrounded by three adeninato ligands from three heptanuclear entities (see Fig. 1c). It is observed that at one side of the dicarboxylate anion, two adenine ligands are positioned in close proximity, with a separation of approximately 3.3 Å for the shortest distance between the two entities. On the opposite side, a single adenine ligand is present, exhibiting a pronounced tilt relative to the mean plane of the dicarboxylate anion (43°, shortest distance: 3.35 Å). It is interesting to note that both carboxylate groups of this dianion establish the same interaction with HO–Cu–OH₂ fragments from the cationic entities as that found in Cu₇Naph·32H₂O. However, one of the carboxylate groups loses its coplanarity with respect to the naphthalene core. As a consequence of the aforementioned alterations to the preceding configuration, the naphthalene-2,6-dicarboxylate anions are closer to the heptanuclear entities, resulting in a reduction of the interstitial space (void percentage: 24.4%; pores: 2.4–4.0 Å, Fig. S9).

As demonstrated above, the supramolecular interactions involving the naphthalene-2,6-dicarboxylate anion play a significant role in the shrinkage and enlargement process of the porous material. In addition, it has been demonstrated that these structural alterations are reversible, as these crystals, when exposed to a humid saturated atmosphere, return to the initial Cu₇Naph·32H₂O structure.

The supramolecular assembly process results in the formation of a three-dimensional network of interconnected pores for all the Cu₇Naph compounds (see Table S2 and Fig. S7). The total volume of these pores ranges from 973 to 1994 ų, which constitutes between 24.4 and 43% of the volume of the crystallographic unit cell of the compounds previously examined. This process of crystal structure expansion and contraction must also be understood as the capacity of these compounds to open their pores, which renders them ideal materials for loading molecules greater than solvent molecules inside them.

3.2. Structural flexibility upon the incorporation of 5-FU molecules

Up to this point, it has been proven that the crystal structure is flexible with respect to shrinkage caused by the loss of water molecules during the crystallisation process. Now we aim to demonstrate, through single-crystal X-ray diffraction structural characterisation, that this flexibility also enables the expansion of the unit cell, thus facilitating the incorporation of larger molecules, such as the drug 5-FU. It should be noted that an attempt was made to use cisplatin; however, no suitable single crystals could be isolated. As outlined in the experimental section, the incorporation procedure can be summarised as the passive diffusion of 5-FU molecules into single crystals of Cu₇Naph from an aqueous solution. The resulting Cu₇Naph·2(5-FU)·26H₂O crystal structure (a: 16.9872(6) Å, b: 15.9344(6) Å, c: 18.2038(7) Å, β: 97.989(4)°, Z: 2) demonstrates that 5-FU molecules interact with the π-stacking adeninato/naphthalene-2,6-dicarboxylate/adeninato synthon (Fig. 4a). The 5-FU establishes complementary hydrogen bonding interactions with the Hoogsteen side of the adeninato ligand N6_adeninato–H⋯O2_5-FU and N1_5-FU–H⋯N7_adeninato, and simultaneously both the adeninato ligand and 5-FU establish π-stacking interactions with the naphthalene-2,6-dicarboxylate anion (Fig. 4b). This interaction pattern happens both above and below the plane of the dicarboxylate anion. As seen in the Cu₇Naph·32H₂O, the carboxylate group of the organic anion is attached to the nearest cationic layer by two O_anion⋯H–O_wcoord/OH hydrogen bonds, one with a hydroxide group and the other with a water molecule coordinated to the same copper atom, forming a supramolecular R₂²(8) synthon. The interactions under consideration result in a lengthening of the cluster–cluster distance along the a axis and a 5.2% increase in unit cell volume with respect to the Cu₇Naph·32H₂O (4879.6 vs. 4640.5 ų). It is noteworthy that the porosity of the material is largely retained, as evidenced by a void percentage of 33.4% and pore dimensions ranging from 2.9 to 6.7 Å, Fig. S9. Indeed, in a previous work using a more concentrated 5-FU solution, a greater amount of 5-FU was loaded, probably occupying these remaining voids.²⁰ However, the passive diffusion experiment performed under these conditions did not yield a single crystal suitable for structural characterisation. Consequently, the position of 5-FU within the structure of Cu₇Naph·2(5-FU)·26H₂O should be considered as its preferential adsorption site. The structural flexibility described above, which allows the expansion and contraction of the crystal structure and the pores within it, will play a crucial role when incorporating even bigger cisplatin drug molecules.

Fig. 4 (a) and (b) Details of the supramolecular interactions between the adeninato ligands, naphthalene-2,6-dicarboxylate anions and 5-FU drug molecules in Cu₇Naph·2(5-FU)·26H₂O. Crystal structure packing viewed along the b crystallographic axis (c) and the contact surface of the voids along the c crystallographic axis (d). Naphthalene-2,6-dicarboxylate anion is represented in green and 5-FU in blue for clarity.

3.4. Drug loading and release capacity

At this point, it was decided to exploit the structural flexibility of the system and the findings of this previous study^20,21 to develop a controlled drug delivery system. This system incorporates two well-known anti-tumour drug molecules: 5-FU and cisplatin, which are used in combined therapies for certain cancers.^37–39 The loading of both molecules was monitored by analysing the concentration of the solution containing the drug before and after the loading procedure. In a previous study, the maximum 5-FU loading that Cu₇Naph can achieve was established at 57.7 wt% with respect to the initial mass of Cu₇Naph (the drug solution that was utilised in this study was five times more concentrated).²⁰ However, for this particular study, it was hypothesised that a reduced dosage of 5-FU would be more appropriate, with the objective of preserving the capacity for subsequent loading of cisplatin. The outcomes of the study demonstrate that, under the more diluted conditions outlined in the experimental section, a reduced quantity of 5-FU and cisplatin is loaded (7–15%, Table 2). Interestingly, a significant increase in the 5-FU loading was observed, increasing from 8.1 wt% when loaded alone to 14.6 wt% when loaded in combination with cisplatin, while the amount of loaded cisplatin only decreased very slightly (7.0 wt% vs. 8.0 wt%). The observed changes are attributed to a synergistic effect between both drug molecules when located inside the voids of Cu₇Naph. It has been proposed that the inner surface of the channels is optimally suited to interact with 5-FU when cisplatin is concomitantly present in these pores. In summary, the functionalisation of the pore surface due to the sorption of cisplatin is conducive to a better interaction/capture of 5-FU.

Table 2 Loaded drug content (%) in the Cu₇Naph compound

	5-FU	cisPt
5-FU@Cu7Naph	8.1 ± 0.3	—
cisPt@Cu7Naph	—	8.0 ± 0.3
5-FUcisPt@Cu7Naph	14.6 ± 0.4	7.0 ± 0.2

---

The drug-loaded samples were characterised using SEM-EDS (Fig. 5a and Fig. S12–S15). The studies indicate that the size and morphology of the Cu₇Naph particles are retained, with a homogeneous distribution of 5-FU and cisplatin (monitored through the presence of F and Pt, respectively). The PXRD patterns of the drug-loaded samples reveal changes to the supramolecular structure. As expected, the major changes that appear are the variations in the peak's intensity due to the different contents within the pores of the material. There are no significant changes in the peak position for 5-FU@Cu₇Naph and cisPt@Cu₇Naph, probably due to the low loading amounts in these samples which do not imply a significant change in the unit cell parameters. However, the 5-FUcisPt@Cu₇Naph sample shows significantly greater changes affecting the first diffraction peak (2θ ≈ 5.5°), which corresponds to the overlapping diffraction signals of the (100) and (010) crystallographic planes. These two peaks almost perfectly coincide for the low loaded samples 5-FU@Cu₇Naph and cisPt@Cu₇Naph (a and b axes are almost identical in Cu₇Naph·32H₂O), but it is significantly broadened in 5-FUcisPt@Cu₇Naph as the greater amount of drug loaded implies a significant lengthening of a axis while b axis remains relatively unchanged as observed in the SC-XRD structure of Cu₇Naph·2(5-FU)·26H₂O.

Fig. 5 (a) SEM (BSE detector) and EDS elemental mapping images (Cu, F and Pt) for the surface of the crystals of compound 5-FUcisPt@Cu₇Naph at 1kX magnifications and (b) PXRD patterns of the Cu₇Naph after loading the different drugs (separately and mixed).

The release kinetics of these samples were measured at 35 °C to simulate conditions closer to the human body (Fig. 6 and Fig. S16).^40,41 Only a fraction of the loaded drug is released (47% in 5-FU@Cu₇Naph, 4% in cisPt@Cu₇Naph and in 5-FUcisPt@Cu₇Naph 8% and 4% of 5-FU and cisplatin, respectively). These values are indicative of the presence of irreversible interactions, particularly those involving cisplatin, or of pore collapse, which is not evident in the PXRD patterns. Another interesting feature is that a lower amount of 5-FU is released from 5-FUcisPt@Cu₇Naph, despite the loading of 5-FU increasing from 8.1% to 14.6%. This seems to indicate a strong and irreversible interaction between 5-FU and cisplatin. The data from the first 2 hours of drug release can be fitted to first-order kinetics (Fig. S17). Analysis of this data indicates that, for samples loaded with a single drug, the release is relatively fast in both cases, although faster for 5-FU in 5-FU@Cu₇Naph (K_5-FU = 1.6439 h⁻¹; t_1/2 = 0.42 h) than for cisplatin in cisPt@Cu₇Naph (K_cisPt = 0.3908; t_1/2 = 1.77 h), Table S4. It appears that the low loading values facilitate the diffusion of the drug molecules along the channels of this compound. In the case of 5-FUcisPt@Cu₇Naph, however, the release of the two drug molecules has slowed. The release of 5-fluorouracil is still faster than that of cisplatin (K_5-FU = 0.8733 h⁻¹; t_1/2 = 0.79 h vs. K_cisPt = 0.3474; t_1/2 = 2.0 h), but the difference between the two in relative terms has been reduced significantly (K_5-FU/K_cisPt = 2.5 for simultaneous loading and 4.2 for separated loading). This is an expected result of the simultaneous loading of both drug molecules, as they must both diffuse through the same channels before being released into the aqueous medium. The slower molecule (cisplatin) hinders the diffusion of the faster 5-FU, causing their release kinetics to converge. The slower release is probably due to a greater accumulation of drug molecules within the channels, hindering diffusion. Apart from the fact that less abrupt drug release is always desirable, closer release kinetics for both drugs is advantageous for delivering a drug cocktail therapy, as it is better that all the different drug molecules arrive at the tumour cells simultaneously or as close as possible to maximise the therapeutic effect. Conversely, the ratio of released 5-FU to cisplatin in 5-FUcisPt@Cu₇Naph is 4.4, which can be acceptable, keeping in mind the higher cytotoxicity of cisplatin (rat oral LD₅₀ = 25.8 mg kg⁻¹) compared to 5-FU (rat oral LD₅₀ = 230 mg kg⁻¹).⁴² In fact, several medical procedures combine 5-FU and cisplatin to treat different cancers, and all of them use an even higher dose of 5-FU than cisplatin.^43–45

Fig. 6 Kinetic desorption curves until release stabilization in water at 35 °C for (a) 5-FU in 5-FU@Cu₇Naph, (b) cisPt in cisPt@Cu₇Naph and (c) 5-FU and (d) cisPt in 5-FUcisPt@Cu₇Naph, respectively. Inset: fitting to first-order kinetics considering the first 2 h of the desorption process.

3.5. Cytotoxicity and RNA-seq transcriptomic analysis

Finally, cytotoxic studies were carried out on samples containing one (5-FU@Cu₇Naph and cisPt@Cu₇Naph), two drugs (5-FUcisPt@Cu₇Naph) and aqueous solutions of each drug. The amounts selected (see Table 3) were chosen to ensure clarity of comparison and to always use the same amount of the drug carrier, Cu₇Naph.

Table 3 Conditions employed for the cytotoxic studies, indicating the drug@carrier amounts employed and the released drug concentration according to the desorption kinetic studies

	Concentration (μg/100 μL)	Released 5-FU (μg/100 μL)	Released cisplatin (μg/100 μL)
Cu7Naph	27.0	—	—
5-FU@Cu7Naph	29.3	1.1	—
cisPt@Cu7Naph	29.0	—	0.2
5-FUcisPt@Cu7Naph	32.7	0.4	0.1
5-FU	1.0	1.0	—
Cisplatin	0.2	—	0.2

---

Growth measurements performed using aqueous solutions of 5-FU and cisplatin, as well as an aqueous suspension of Cu₇Naph, indicate that the cytotoxic effect of 5-FU and Cu₇Naph in cell growth is visible after 96 hours of incubation, whereas the effect of cisplatin only emerges after 48 hours (Fig. 7). When Cu₇Naph is loaded with these drugs, 5-FU and cisplatin exhibit different behaviours. The cytotoxicity of 5-FU@Cu₇Naph during the first 48 hours is greater than that of 5-FU or Cu₇Naph separately. Conversely, the cytotoxicity of cisPt@Cu₇Naph is lower than that of neat Cu₇Naph, likely due to the strong interaction of cisplatin with the [Cu₇(μ-adeninato)₆(μ₃-OH)₆(μ-OH₂)₆]²⁺ heptanuclear entities, which probably hinders the mechanism of Cu₇Naph cytotoxicity. Finally, 5-FUcisPt@Cu₇Naph exhibits lower cytotoxicity than the carrier Cu₇Naph itself, but higher than cisPt@Cu₇Naph, likely due to the release of some 5-FU molecules. These results emphasise the importance of ensuring that the constituents of drug cocktail therapy do not interfere with their mechanism of action, as is the case between cisplatin and the Cu₇Naph drug carrier. This interference does not occur when Cu₇Naph is loaded with 5-FU.

Fig. 7 Relative growth of the HCT116 cells measured by crystal violet at different time points. Cells were grown at basal conditions or after the different compounds were added. Data represent the mean and standard error of 4 independent experiments. *p < 0.05 (red); **p < 0.01 (blue) calculated by two-tailed, paired Student's t-test against the basal condition.

However, the cytotoxic behaviour of Cu₇Naph still needs to be understood, particularly given that a related compound based on these heptanuclear cationic entities, but with a Cr³⁺ cation in place of the central Cu²⁺ cation and SO₄²⁻ as counterions, promotes the proliferation of tumour cells.²¹ Therefore, transcriptomic studies were conducted. Comparing NT (non-treated) cells with cells cultured in the presence of Cu₇Naph. These studies reveal no significant transcription differences (Fig. 8 and Fig. S18 and S19). However, careful analysis of the most under expressed genes revealed that: AKR1A1 (“aldo-keto reductase family 1, member A1”; fold change (FC) = 2.6; p = 0.00001) and PUF60 (“poly(U) binding splicing factor 60”; FC = 3.0; and p = 0.006), shed some light on the reason for the cytotoxic behaviour of the Cu₇Naph-cultured cells. AKR1A1 encodes an aldehyde reductase that reduces or protects against cytotoxicity by detoxifying aldehydes, which are reactive molecules that can cause cellular damage, protein carbonylation, and oxidative stress,⁴⁶ while PUF60 encodes a nucleic acid-binding protein that promotes mitotic cell cycle and cancer progression.⁴⁷ It has been proven that PUF60 depletion inhibits LUAD cell-cycle G2/M transition, cell proliferation, and tumor development.⁴⁸ Therefore, the observed cytotoxicity in colon cancer cells may be influenced by both underexpressed genes.

Fig. 8 Pairwise volcano plot for Cu₇Naph cultured cells vs. the non-treated (NT) cultured cells. The colour bar indicates the scattering density. Darker blue colour corresponds to higher scattering density.

4. Conclusions

This work demonstrates the potential of supramolecular porous materials as flexible drug delivery systems capable of simultaneously loading two drug molecules: 5-fluorouracil and cisplatin. The flexibility of the Cu₇Naph supramolecular structure originates from π-stacking interactions between the adeninato ligand and naphthalene-2,6-dicarboxylate counterions. These π-stacking interactions are rigid with respect to the parallel alignment of the interacting aromatic rings, but allow for the rotation of the entities while maintaining this parallel alignment. Therefore, the Cu₇Naph structure shrinks upon loss of crystallisation water molecules, but also expands upon the inclusion of larger molecules, such as 5-FU. This molecule is incorporated into the adeninato/naphthalene-2,6-dicarboxylate/adeninato synthon via π-stacking interactions with the dicarboxylate counterion, as well as complementary hydrogen bonding interactions with the Hoogsteen face of the adeninato ligands. This forces the distance between the heptanuclear entities along this synthon (the a-axis) to increase from 15.987 Å in Cu₇Naph·32H₂O to 16.987 Å in Cu₇Naph·2(5-FU)·26H₂O. This flexibility enables the incorporation of molecules such as 5-FU and cisplatin, despite their dimensions (5-FU: 8.3 × 7.1 × 3.5 Å; cisplatin: 7.7 × 7.7 × 3.8 Å; see Fig. S20) exceeding the pore windows of Cu₇Naph·32H₂O (3.7 Å).

Conversely, simultaneous loading of drug molecules does not equate to the sum of loadings for each drug alone, nor does it result in a lower loading value, due to competition for the same adsorption sites. In fact, the loading of 5-fluorouracil almost doubles (14.6 wt% vs. 8.1 wt%), while the loading of cisplatin remains almost the same (7.0 wt% vs. 8.0 wt%) when both drugs are loaded simultaneously. These results are interpreted as indicating a stronger interaction of cisplatin with the Cu₇Naph host than with 5-FU. However, the interaction of 5-FU with cisplatin appears to be at least as strong as that with the Cu₇Naph host, significantly altering the loading values of 5-FU. These interactions also influence drug release, with only a small amount of loaded cisplatin being released from both cisPt@Cu₇Naph and 5-FUcisPt@Cu₇Naph (4%), whereas the percentage of loaded 5-FU being released drops from 47% for 5-FU@Cu₇Naph to 8% for 5-FUcisPt@Cu₇Naph. The latter is probably related to some kind of irreversible binding of cisplatin to the adeninato ligands of the host and also to the loaded 5-FU molecules. Fitting the drug release data to a first-order kinetic equation also shows that, although removable 5-FU is released from 5-FU@Cu₇Naph 4.2 times faster than removable cisplatin from cisPt@Cu₇Naph, the release kinetics of both drugs become closer when they are loaded simultaneously. Currently, 5-FU is released only 2.5 times faster than cisplatin, and it is likely that, with greater loading values for both drugs, their release kinetics could become even closer.

Cytotoxic studies confirmed the previously reported cytotoxic nature of the Cu₇Naph host and were complemented by transcriptomic analyses. These studies revealed the down-expression of two significant genes (AKR1A1 and PUF60) in tumour cell cultures, which could cause the cytotoxic effect. The cytotoxic effect of drug-loaded Cu₇Naph samples is also impacted by the strong interaction between cisplatin and the Cu₇Naph host. 5-FU@Cu₇Naph exhibits the anticipated increase in cytotoxicity due to the combined action of 5-FU and Cu₇Naph. However, the cytotoxic effect of cisPt@Cu₇Naph and 5-FUcisPt@Cu₇Naph is diminished below that of Cu₇Naph. Therefore, while the two-component drug cocktail approach represented by 5-FU@Cu₇Naph is successful, the three-drug cocktail approach represented by 5-FUcisPt@Cu₇Naph is hindered by the strong interaction between cisplatin and the Cu₇Naph carrier, leading to poor cytotoxic behaviour. Overall, this work has shed light on the phenomenon of the simultaneous loading and release of drug molecules, how their interaction with the host and with each other alters the expected loading and release capacities, and how this approach of simultaneous drug release can bring the release kinetics of each drug closer together without the use of a different host for each drug.

Author contributions

The manuscript was written through contributions from all authors who gave their approval to the final version. S. M-G. performed the adsorption and drug release experiments and together with J. P-C analysed the crystallographic section and prepared the manuscript and SI. O. C., A. L., G. B. and S. P-Y. designed the experiments and contributed to the formal analysis. A. C-R. L. B-M., D. G. and M. J. A.-B. performed and analysed the biological assays. O. C. and A. L. contributed to fundraising, project management, supervision, and proofreading and editing of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

Supporting data associated with this work area are included in the SI. Supplementary information: X-ray powder diffraction data, crystallographic and structural data, porosity analysis, drug loading data, FTIR spectroscopic data, thermogravimetric analysis, SEM images, EDS analysis, desorbed samples characterization, RNA sequencing analysis, and adsorbate molecule size and shape analysis. See DOI: https://doi.org/10.1039/d5tb01280e.

The transcriptomic data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus (GEO)⁴⁹ and are accessible through GEO Series accession number GSE298079 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE298079).

CCDC 2454556–2454558 contain the supplementary crystallographic data for this paper.⁵⁰ These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via https://www.ccdc.cam.ac.uk/structures.

Acknowledgements

The authors gratefully acknowledge the final support from the University of the Basque Country (EHU-N23/51), Basque Government (IT1722-22 and ELKARTEK KK-2022/00032) and the Spanish Ministry of Science and Innovation (PID2022-138968NB-C22 and PID2023-152752OB-I00/AEI/10.13039/501100011033 projects funded by MCIN/AEI/10.13039/501100011033). D. G. and M. J. A-B. are thankful to the IPerGlio project from the European Union's Horizon 2020 research and innovation programme under Grant Agreement ERAPERMED2022-245 co-financed by the Fundación Científica de la Asociación Española Contra el Cáncer (FCAECC) (Cod. PERME224571ARAU) and the Instituto de Salud Carlos III (ISCIII) of Spain (Exp. AC22/00043). S. M-G. thanks the Basque Government and European Union NextGeneration EU (Investigo Program PIFINVE22/47). J. P-C is thankful for the final support from the Basque Government with the postdoctoral grant (POS_2023_1_0059). Technical and human support provided by SGIker (UPV/EHU, MICINN, GV/EJ, ESF) is also acknowledged.

Notes and references

1. F. Bray, M. Laversanne, E. Weiderpass and I. Soerjomataram, Cancer, 2021, 127, 3029–3030.
2. S. U. Khan, K. Fatima, S. Aisha and F. Malik, Cell Commun. Signaling, 2024, 22, 109–135.
3. T. Hilal, M. Gonzalez-Velez and V. Prasad, JAMA Intern. Med., 2020, 180, 1108–1115.
4. G. P. Nagaraju and M. A. Kamal, Curr. Drug Metab., 2019, 20, 931–932.
5. X. Lu, S. Fan, M. Cao, D. Liu, K. Xuan and A. Liu, J. Pharm. Invest., 2024, 54, 785–802.
6. S. Islam, M. M. S. Ahmed, M. A. Islam, N. Hossain and M. A. Chowdhury, Results Surf. Interfaces, 2025, 19, 100529.
7. H. Hameed, S. Faheem, A. C. Paiva-Santos, H. S. Sarwar and M. Jamshaid, AAPS PharmSciTech, 2024, 25, 1–27.
8. M. Kumar, A. Thakur, U. K. Mandal, A. Thakur and A. Bhatia, AAPS PharmSciTech, 2022, 23, 244.
9. D. S. R. Khafaga, M. T. El-Morsy, H. Faried, A. H. Diab, S. Shehab, A. M. Saleh and G. A. M. Ali, RSC Adv., 2024, 14, 30201–30229.
10. S. Gautam, I. Lakhanpal, L. Sonowal and N. Goyal, Next Nanotechnol., 2023, 3–4, 100027.
11. L. He, M. Shang, Z. Chen and Z. Yang, Chem. Rec., 2023, 23, e202300018.
12. Y. Du, X. Jia, L. Zhong, Y. Jiao, Z. Zhang, Z. Wang, Y. Feng, M. Bilal, J. Cui and S. Jia, Coord. Chem. Rev., 2022, 454, 214327.
13. A. Vikal, R. Maurya, P. Patel, S. R. Paliwal, R. K. Narang, G. Das Gupta and B. Das Kurmi, Appl. Mater. Today, 2024, 41, 102443.
14. J. Tian, Z. Y. Xu, D. W. Zhang, H. Wang, S. H. Xie, D. W. Xu, Y. H. Ren, H. Wang, Y. Liu and Z. T. Li, Nat. Commun., 2016, 7, 11580–11589.
15. Q. Hu, B. Zhang, H. Ren, X. Zhou, C. He, Y. Shen, Z. Zhou and H. Hu, Acta Biomater., 2023, 168, 617–627.
16. J. Thomas-Gipson, R. Pérez-Aguirre, G. Beobide, O. Castillo, A. Luque, S. Pérez-Yáñez and P. Román, Cryst. Growth Des., 2015, 15, 975–983.
17. R. Pérez-Aguirre, G. Beobide, O. Castillo, I. De Pedro, A. Luque, S. Pérez-Yáñez, J. Rodríguez Fernández and P. Román, Inorg. Chem., 2016, 55, 7755–7763.
18. J. Pascual-Colino, G. Beobide, O. Castillo, I. Da Silva, A. Luque and S. Pérez-Yáñez, Cryst. Growth Des., 2018, 18, 3465–3476.
19. J. Pascual-Colino, R. Pérez-Aguirre, G. Beobide, O. Castillo, I. de Pedro, A. Luque, S. Mena-Gutiérrez and S. Pérez-Yáñez, Inorg. Chem. Front., 2023, 10, 2250–2261.
20. S. Mena-Gutiérrez, J. Pascual-Colino, G. Beobide, O. Castillo, A. Castellanos-Rubio, A. Luque, E. Maiza-Razkin, J. Mentxaka and S. Pérez-Yáñez, Inorg. Chem., 2023, 62, 18496–18509.
21. S. Mena-Gutiérrez, E. Maiza-Razkin, J. Pascual-Colino, M. J. Araúzo-Bravo, G. Beobide, O. Castillo, A. Castellanos-Rubio, D. Gerovska, A. Luque and S. Pérez-Yáñez, J. Mater. Chem. B, 2024, 2, 11156–11164.
22. P. Garg, J. Malhotra, P. Kulkarni, D. Horne, R. Salgia and S. S. Singhal, Cancers, 2024, 16, 2478.
23. M. Nikolaou, A. Pavlopoulou, A. G. Georgakilas and E. Kyrodimos, Clin. Exp. Metastasis, 2018, 35, 309–318.
24. S. J. Park, K. Shin, I. H. Kim, M. A. Lee, T. H. Hong and H. Kim, Sci. Rep., 2025, 15, 3386.
25. Agilent Technologies UK Ltd CrysAlisPRO, Oxford Diffraction; Agilent Technologies UK Ltd: Yarnton, England.
26. A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. C. Burla, G. Polidori and M. Camalli, J. Appl. Crystallogr., 1994, 27, 435–436.
27. L. Palatinus and G. Chapuis, J. Appl. Crystallogr., 2007, 40, 786–790.
28. G. M. Sheldrick, Acta Crystallogr., Sect. A: Found Adv., 2008, 64, 112–122.
29. A. L. Spek, Acta Crystallogr., Sect. C:Cryst. Struct. Commun., 2015, 71, 9–18.
30. A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7–13.
31. D. Gerovska and M. J. Araúzo-Bravo, Int. J. Mol. Sci., 2023, 24, 2736.
32. M. Pertea, D. Kim, G. M. Pertea, J. T. Leek and S. L. Salzberg, Nat. Protoc., 2016, 11, 1650–1667.
33. C. Trapnell, A. Roberts, L. Goff, G. Pertea, D. Kim, D. R. Kelley, H. Pimentel, S. L. Salzberg, J. L. Rinn and L. Pachter, Nat. Protoc., 2012, 7, 562–578.
34. D. Gerovska, G. Larrinaga, J. D. Solano-Iturria, J. Márquez, P. G. Gallastegi, A. M. Khatib, G. Poschmann, K. Stühler, M. Armesto, C. H. Lawrie, I. Badiola and M. J. Araúzo-Bravo, Cancers, 2020, 12, 2380.
35. C. Janiak, J. Chem. Soc. Dalton Trans., 2000, 3885–3896.
36. R. G. Huber, M. A. Margreiter, J. E. Fuchs, S. Von Grafenstein, C. S. Tautermann, K. R. Liedl and T. Fox, J. Chem. Inf. Model., 2014, 54, 1371–1379.
37. N. Tsavaris, K. Tentas, C. Bacoyiannis, M. Katsikas, N. Sakelaropoulos, C. Kosmas, D. Daliani and P. Kosmidis, Anticancer Drugs, 1995, 6, 599–603.
38. T. Hara, K. Nishikawa, M. Sakatoku, K. Oba, J. Sakamoto and K. Omura, Gastric Cancer, 2011, 14, 332–338.
39. M. Moehler, A. Maderer, P. C. Thuss-Patience, B. Brenner, J. Meiler, T. J. Ettrich, R. D. Hofheinz, S. E. Al-Batran, A. Vogel, L. Mueller, M. P. Lutz, F. Lordick, M. Alsina, K. Borchert, R. Greil, W. Eisterer, A. Schad, J. Slotta-Huspenina, E. Van Cutsem and S. Lorenzen, Ann. Oncol., 2020, 31, 228–235.
40. T. Yahara, T. Koga, S. Yoshida, S. Nakagawa, H. Deguchi and K. Shirouzo, Surg. Today, 2003, 33, 243–248.
41. H. Zhu, H. Xu, Y. Zhang, J. Brodský, I. Gablech, M. Korabečná and P. Neuzil, Adv. Sci., 2025, 12, 2402135.
42. J. Taïeb, E. Mitry, V. Boige, P. Artru, J. Ezenfis, T. Lecomte, M. C. Clavero-Fabri, J. N. Vaillant and P. Rougier, Ann. Oncol., 2002, 13, 1192–1196.
43. National Institute for Health and Care Excellence ( 2020) Chemotherapy protocol. Head and Neck Cancer Cisplatin-Fluorouracil. Available at: https://www.uhs.nhs.uk/Media/UHS-website-2019/Docs/Chemotherapy-SOPs1/Headandneckcancer/InP-Cisplatin-Fluorouracil.pdf (Accessed: 2 September 2025).
44. National Cancer Control Programme ( 2020) Cisplatin, 5-Fluorouracil and Radiation Therapy. Available at: https://www.bccancer.bc.ca/chemotherapy-protocols-site/Documents/Head%20and%20Neck/HNSAVFUP_Protocol.pdf (Accessed: 2 September 2025).
45. British Columbia Cancer ( 2024) BC Cancer Protocol Summary for Combined Modality Therapy for Squamous Cell Cancer of the Genitourinary System using Fluorouracil and CISplatin with Radiation. Available at: https://www.bccancer.bc.ca/chemotherapy-protocols-site/Documents/Genitourinary/GUFUPRT_Protocol.pdf (Accessed: 2 September 2025).
46. T. O’Connor, L. S. Ireland, D. J. Harrison and J. D. Hayes, Society, 1999, 343, 487–504.
47. D. O. Toader, R. Ursu, N. Bacalbasa, D. Cretoiu, L. G. Ppp, I. Balescu, F. Gherghiceanu, F. Furtunescu, D. Radavoi and V. Radoi, Cancer Diagnosis Progn., 2021, 1, 213–219.
48. N. Xu, Y. Ren, Y. Bao, X. Shen, J. Kang, N. Wang, Z. Wang, X. Han, Z. Li, J. Zuo, G.-H. Wei, Z. Wang, W.-X. Zong, W. Liu, G. Xie and Y. Wang, Cell Rep., 2023, 42, 113041.
49. R. Edgar, M. Domrachev and A. E. Lash, Nucleic Acids Res., 2002, 30, 207–210.
50. (a) S. Mena-Gutiérrez, M. J. Araúzo-Bravo, G. Beobide, L. Bergara-Muguruza, O. Castillo, A. Castellanos-Rubio, D. Gerovska, A. Luque, J. Pascual-Colino and S. Pérez-Yáñez, CCDC 2454556: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2nd57h; (b) S. Mena-Gutiérrez, M. J. Araúzo-Bravo, G. Beobide, L. Bergara-Muguruza, O. Castillo, A. Castellanos-Rubio, D. Gerovska, A. Luque, J. Pascual-Colino and S. Pérez-Yáñez, CCDC 2454557: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2nd58j; (c) S. Mena-Gutiérrez, M. J. Araúzo-Bravo, G. Beobide, L. Bergara-Muguruza, O. Castillo, A. Castellanos-Rubio, D. Gerovska, A. Luque, J. Pascual-Colino and S. Pérez-Yáñez, CCDC 2454558: Experimental Crystal Structure Determination, 2025 DOI:10.5517/ccdc.csd.cc2nd59k.

---