Computational insight into the interaction of oxaliplatin with insulin

Giuseppe Sciortino ab, José-Emilio Sánchez-Aparicio a, Jaime Rodríguez-Guerra Pedregal a, Eugenio Garribba b and Jean-Didier Maréchal *a
aDepartament de Química, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallés, Barcelona, Spain. E-mail:
bDipartimento di Chimica e Farmacia, Università di Sassari, Via Vienna 2, I-07100 Sassari, Italy

Received 27th November 2018 , Accepted 21st January 2019

First published on 21st January 2019

In an organism, cisplatin and its derivatives are known to interact with proteins besides their principal DNA target. These off-target interactions have major therapeutic consequences including undesired side effects, loss of bioavailability and emergence of resistance. Insulin is one of the prototypical protein targets of platinum drugs as it has been seen to be involved in bioavailability reduction and might also determine resistance in certain cancer lines. However, despite the interest in understanding the nature of the oxaliplatin–insulin adducts, no 3D models have been achieved so far. In this study, we apply our recent computational multiscale protocol optimized for bioinorganic interactions to provide structural insights into these systems. To do so, the initial structures are predicted by blind protein–metalloligand docking calculations optimized to account for a metal-containing species, and then refined using a Molecular Dynamics (MD) and Quantum Mechanics/Molecular Mechanics (QM/MM) integrated protocol. The results are consistent with experimental information obtained from fragment analysis, and also provide novel structural information like conformational changes occurring upon binding and potential effects on the biological functions of the protein. This study opens an avenue towards applying similar strategies to a wide ensemble of metallodrug–protein/peptide systems for which no structural data are available.

Significance to metallomics

Cisplatin and its derivatives are a particularly interesting case of metallodrug promiscuity, since they are known to bind not only to nucleic acids (their intended target), but also to peptide chains, which is suspected to be the cause of their multiple undesired side effects and drug resistance phenomena. In this work, we construct and apply a multiscale protocol supported by our recent software advancements to generate models of metalloligand–protein adducts. Therefore, we apply it to gain an insight into oxaliplatin–insulin interactions, where a 3D structure is yet to be characterized experimentally.


Since the approval of cisplatin as an anticancer agent, efforts for the development of metalloligands with relevance in biomedical applications, therapy or diagnosis have constantly been increasing.1–7 To name only a few, BiIII–citrate (De-Nol) is now investigated as an antiulcer agent, the triethylphosphine AuI complex (auranofin) is an antiarthritic species, RuIII NAMI-A is an antitumoral species and GdIII complexes are used for contrast enhancement in magnetic resonance imaging (MRI).8–10 However, looking for cisplatin alternatives and understanding their mechanism of action is still a vivid field of research. Second and third generations of Pt drugs aim at lowering toxicity, limiting side effects (such as kidney toxicity, nausea, vomiting, tinnitus and bone marrow suppression), improving bioavailability and reducing resistance.1,3,7–9

One of these alternatives is oxaliplatin with formula [Pt(dach)(ox)] – where dach is 1R,2R-diaminocyclohexane and ox is oxalate – whose clinically used infusion is named Eloxatin.8,9 Oxaliplatin is an anticancer drug especially effective against metastatic colorectal cancer which was approved for clinical use in the European Union in 1999 and in the United States in 2002.11–13 Major advantages of oxaliplatin versus first generation drugs are its much lower nephro- and ototoxicity, and, for this reason, it is often used to treat patients with cisplatin-resistant tumours. However, oxaliplatin has also displayed resistance profiles in certain patients and some of them have been shown to be insulin related.14

Although the main therapeutic target of cisplatin and its derivatives is cellular DNA, the pro-drugs contain labile ligands, which in blood provide the metal moiety with two free coordination positions. In serum, these positions can be coordinated by donors present in bioligands, proteins and peptides. These interactions can lead to a reduction of the amount of the drug available in the organism to exert its antitumor activity.15 Therefore, decoding the interaction of metallodrugs with proteins16 plays a key role in understanding their transport in blood and their cellular uptake, and could also help to shed light on the origin of the side-effects and drug resistance, leading – in the end – towards the design of improved treatments. Direct interactions of cisplatin and oxaliplatin with a large list of proteins, including insulin, have been demonstrated.17–21 Further evidence of the interaction between insulin and oxaliplatin has been shown in the relation between drug resistance and the insulin dependent cellular cascade of PI3K/Akt for colon cancer lines (although in this case the relevance of direct binding between the drug and the hormone has not been assessed).14 Despite the interest of decoding the nature of the oxaliplatin–insulin adducts, no models of 3D structures have been achieved so far.

Insulin is a small protein composed of two peptides, namely A and B, linked by two interchain and one intrachain disulphide bridges, respectively (Fig. 1).22 It is synthesized in the β-cells of pancreatic islets and plays a crucial role in the regulation of blood glucose levels. Its relatively low molecular mass (5.777 kDa in porcine insulin) and the presence of nitrogen and sulphur donors in the amino acidic sequence makes it a suitable and interesting model for studying its interaction with oxaliplatin. In the literature, several possible binding sites of oxaliplatin have been identified by the combined application of MALDI-ToF-MS (matrix-assisted laser desorption/ionization time-of-flight mass spectrometry) and LC-nESI-Q-ToF-MS (liquid chromatography-nano-electrospray ionization quadrupole time-of-flight mass spectrometry), revealing that His5, Cys7, and His10 residues would be involved in the metal coordination (Fig. 1).23 Moreover, MALDI-ToF/ToF-MS results demonstrated that the platinum drugs induce reduction of the disulphide bonds in porcine pancreatic insulin indicating that the metal complex would alter the native conformation of the protein and lead to its inactivation.18

image file: c8mt00341f-f1.tif
Fig. 1 Dipeptide structure of porcine insulin. The A and B chains, the three disulphide bridges and the potential binding sites (in the two spheres) are shown.

When experimental data on the structure of a protein are missing (i.e. X-ray or NMR determination) or partial (provided, for example, by using EPR, ESEEM, ENDOR, ESI-MS, CD and UV-Vis techniques), molecular modelling can give complementary information. However, dealing with metalloligand–protein binding is usually out of the scope of most molecular modelling frameworks, particularly if the amino acid residues involved in the coordination are not well characterized. Recently, we demonstrated that convenient improvements in docking techniques for dealing with metalloligands could fill this gap, especially if dockings are merged under an integrative computational framework with more accurate methodologies like classical force field MD or QM/MM calculations.24–27

In this study, the approach described above is applied to fully understand the mechanism of interaction of [PtII(dach)(OH)2], generated in water upon the release of the weak oxalate ligand,9 with insulin. One or both the hydroxide ions of [PtII(dach)(OH)2] can be replaced by amino acid residues of insulin.

Computational methods

All the calculations were carried out using the structure 1zni29 of insulin bound to Zn2+ ions available in the Protein Data Bank (PDB28), since it presents sidechain pre-organization and can be considered a valid starting point for modelling the metal interaction. Moreover, this is the structure with the highest resolution of the PDB (1.498 Å). A comparison between the selected and higher resolution unbound X-ray structures of the human form (5ena30 and 3w7y) is provided in Fig. S1 of the ESI. From the comparison of the two structures, an RMSD value lower than 0.55 Å is obtained indicating that the differences are very small and that the structure selected in this paper is a valuable model to study the interaction with oxaliplatin.

The protein was prepared by removing water, ions and crystallographic small molecules from the PDB structure to have free interaction of the insulin residues with the Pt moiety during the simulation. Finally, protons were then added using the algorithm implemented in the UCSF Chimera software.31 The Cartesian coordinates of the Pt moiety ([Pt(dach)]2+) were extracted from the structure with code 4s18.32

Docking calculations were performed by using GOLD 5.2 software33 with the GoldScore function34 and the new set of optimized scoring parameters presented in our recently published studies.24,25 Genetic algorithm (GA) parameters were set at 50 runs with a minimum of 100[thin space (1/6-em)]000 operations. The other parameters – pressure, number of islands, niche size, crossover, mutation and migration – were the default values. The simulations were carried out without any geometrical constraints or energy restraints. The two equatorial positions of oxaliplatin left free from the oxalate ligand (extracted from the structure with PDB code 4s1832) were activated for coordination by adding two dummy hydrogen atoms on the metal-donor axes, making them able to interact with the protein donors. The evaluation sphere was set up to contain the whole monomer of the dinuclear structure of insulin. The GOLD5.2 rotamer libraries35 were applied to all potentially coordinating side chains: His5, His10 and Glu13. Metal complex [Pt(dach)]2+ was treated as a flexible structure with the standard GOLD algorithm. The solutions were analysed by means of GaudiView.36,37

Molecular Dynamics (MD) simulations were set up using xleap, which was instructed to solvate the protein with a cubic box of pre-equilibrated TIP3P water molecules and balance the total charge with Cl ions (ions94.lib library), see Table S1 (ESI). The AMBER14SB force field38 was used for the standard residues, while the GAFF force field was adopted for the remaining atoms. Pt bonding force constants and equilibrium parameters were obtained using the Seminario's method, using Gaussian0939 to compute the geometry and harmonic frequencies at the DFT level with the B3LYP functional combined with the scalar-relativistic Stuttgart–Dresden SDD pseudopotential and its associated double-ζ basis plus a set of f polarization functions40 for Pt. The 6-31G(d,p) basis set was used for H, C, N and S. Point charges of the [Pt(dach)]2+ moiety and coordinated residues were derived using the RESP41 (restrained electrostatic potential) model. The force field building operations were carried out using (interested readers can find a more detailed description of the protocol in the ESI).42 The MD simulation of the unbound protein was performed by leaving the X-ray waters in their crystallographic positions and by following the procedure reported above. For all the MDs, the solvent and the whole system were sequentially submitted to 3000 energy minimization steps to relax possible steric clashes. Then, thermalization of water molecules and side chains was achieved by increasing the temperature from 100 K up to 300 K. MD simulations under periodic boundary conditions were carried out for 500 ns (for ligand-free and binding mode α), 700 ns (binding mode β) and 1000 ns (binding mode γ) with the OpenMM engine43 using the OMM Protocol.44 Analysis of the trajectories was carried out by means of CPPTraj implemented in AmberTools16.45 The MD trajectory was considered converged when a full exploration of the conformational space was achieved. In particular, a stable conformation or a pool of relative stable conformations visited for a statistically consistent number of times was considered as convergence indicators. Considering the alpha carbons of the three helices and the central loop, RMSD from the minimized structure, all-to-all frames RMSD and counting cluster analyses were performed. Moreover, to ensure that dynamic transitions occur between different conformations, a principal component analysis (PCA) was carried out by plotting the two principal modes relative to each other (for further details, interested readers can refer to the ESI).

The implementation of the K-medoid clustering algorithm46 of the MSMBuilder software47 was used to generate ten clusters for each trajectory, taking the RMSD between frames as a distance metric (oxaliplatin and alpha carbons of the three helices and the central loop were considered in the calculations). The representative structure of the most populated cluster for each mode was used in the further QM/MM study. The data of the clustering are provided in the ESI (Tables S2–S4).

All the QM/MM geometry optimizations and frequency calculations were performed on the Garleek48 framework combining Gaussian0939 and Tinker49 for QM and MM calculations, respectively. The QM region was described at the DFT level of theory using the B3LYP functional combined with the scalar-relativistic Stuttgart–Dresden SDD pseudopotential and its associated double-ζ basis set plus a set of f polarization functions40 for Pt, while the 6-31G(d) basis set was used for the rest of the atoms. The MM region was described using the AMBER99SB force field implemented in Tinker.49

Results and discussion

The protein–ligand docking was carried out on the species formed upon the hydrolysis of oxaliplatin. In aqueous solution at physiological pH, it produces [PtII(dach)(OH)2] (upon the release of the weak oxalate ligand) which can yield adducts with insulin after the replacement of one or both OH ions.23,50 The initial docking analysis was performed considering a general case, in which the [PtII(dach)]2+ moiety was treated by activating both the equatorial coordination vacancies through dummy hydrogen atoms.

The results suggest the presence of two potential binding sites for oxaliplatin in insulin: a primary one (site α, 56.75 GoldScore Fitness units) involving two amino acid residues, Cys7B and His5B, and a secondary one (β site, with a score of 55.24 units), in which the [Pt(dach)]2+ moiety is coordinated by only His10B (Fig. S2 of ESI). Site α is characterized by a square planar arrangement and shows stabilization through hydrogen bonds between the amino groups of [Pt(dach)]2+ and amino acid acceptor Cys7A. The predicted bond lengths of Pt–S(Cys7B) and Pt–N(His5B) are 2.297 and 2.288 Å, respectively. The bond angles for N(dach)–Pt–S(Cys7B) and N(dach)–Pt–N(His5B) are 91.2 and 91.3°.

After this preliminary stage, further calculations on the identified regions were also performed considering the interaction with insulin of the monohydroxide hydrolysis product; in other words, only one coordination site is accessible for the protein. The results show that if the [Pt(dach)(OH)]+ fragment is taken into account, a single binding site involving His10B as a coordinating residue is predicted. For this site, two possible binding modes are identified: the first one (score 55.49, population 92%) is stabilized by a strong hydrogen bond between the hydroxide ligand and Glu13B, while on the secondary mode (score 51.64, population 8%) the stabilization takes place with Gln25B, and the complex is rotated 90° with respect to the plane of the ring. The most favoured docking solution for site β shows a square planar arrangement, two strong stabilizations through hydrogen bonds between the hydroxide ligand and the carboxylate group of Glu13B, and the amino group of the [Pt(dach)(OH)]+ moiety and the carbonyl group of His10B. The predicted bond distance of Pt–N(His10B) is 2.387 Å, while the bond angle of N(dach)–Pt–N(His10B) is 99.9°. The amino acids involved in the metal coordination and second sphere stabilization are shown in Fig. 2.

image file: c8mt00341f-f2.tif
Fig. 2 Two docking proposals for the interaction of [Pt(dach)]2+ and [Pt(dach)(OH)]+ moieties with insulin: the α primary binding site with the coordination of Cys7B and His5B and the β secondary binding site with PtII coordinated to His10B. The hydrogen bonds are also highlighted with solid lines.

Some years ago, Møller et al. digested the oxaliplatin–insulin adduct using endoproteinase Glu-C.23 In the peptides generated, several possible binding sites were identified by the combined application of MALDI-ToF-MS (matrix-assisted laser desorption/ionization time-of-flight mass spectrometry) and LC-nESI-Q-ToF-MS (liquid chromatography-nano-electrospray ionization quadrupole time-of-flight mass spectrometry). In particular, several adducts such as Pt(dach), Pt(dach) + OH, 2Pt(dach) + OH and 2Pt(dach) + 2OH formed with insulin were revealed. The fragment of Phe1–Glu13 was identified as the primary binding region and the side chains of Gln4B, His5B, Leu6B, Cys7B, Ser9B, His10B, Leu11B or Val12B and Cys19B were proposed as possible side chain donors,23 in good agreement with our results. The principal peptides identified by Møller et al. are summarized in Table 1.

Table 1 Principal peptides identified by Møller et al. and binding site assignment using the combined application of MALDI-ToF-MS and LC-nESI-Q-ToF-MSa
Adduct Peptide Side chain donors
a In boldface, the regions corresponding to the most intense peaks.
A1 + Pt(dach) Gly1A–Gln5A Not assignable
A1 + Pt(dach) + OH Gly1A–Gln5A Not assignable
A2 + Pt(dach) + OH Cys6A–Leu16A Not assignable
B1 + Pt(dach) Phe1 B –Glu13 B Gln4, His5, Leu6 or Cys7
B1 + Pt(dach) + OH Phe1 B –Glu13 B Cys7
B1 + 2Pt(dach) + OH Phe1 B –Glu13 B His5 or Leu6 and Ser9, His10, Leu11 or Val12
B2 + Pt(dach) + OH Ala14B–Glu21B Cys19

Further docking calculations were performed in order to evaluate the possibility of adducts with more than one PtII moiety bound to insulin. Using the previously obtained Pt-insulin structures (see Fig. 2), subsequent docking calculations were carried out, showing the formation of an adduct with two Pt (one Pt(dach)2+ and one Pt(dach)(OH)+) bound to insulin. The predicted structure (Fig. 3) unveils the compatibility of site α and site β, leading to the conclusion that it could be the B1 + 2Pt(dach) + OH adduct observed by Møller et al. (Table 1).

image file: c8mt00341f-f3.tif
Fig. 3 Docking proposal for the adducts 2Pt(dach) + OH with insulin. The B1 peptide is highlighted in blue.

Based on the docking proposals and considering the close scoring values obtained for the discussed solutions, three different scenarios are possible for the mono-adduct formation: (i) the [Pt(dach)]2+ moiety binds to His5B and Cys7B leaving the disulphide bridge between Cys7A and Cys7B intact (binding mode α); (ii) the [Pt(dach)(OH)]+ moiety binds to His10B (binding mode β) and (iii) the [Pt(dach)]2+ moiety binds to His5B and Cys7B favouring the reduction of the disulphide bridge (additional binding mode γ).23

To assess the stability of the Pt–insulin adducts and the effect of the three potential binding modes on the peptide folding, MD calculations were carried out as described in the Computational section and compared with the unbound state of insulin (supplementary details and Fig. S3 in the ESI).

The original folding of the monomer is highly conserved along the simulation in the first proposal α, probably due to the three disulphide bridges that connect the helices. Concerning the β model, the protein presents a more flexible structure, with two additional conformation sets (see Fig. S3 of the ESI, representatives 9 and 12). Finally, in mode γ the highest flexibility of the monomer is shown with two further ensembles of conformations (Fig. S3 of the ESI, representatives 15 and 16).

The MD trajectory convergence was assessed through PCA analysis (Fig. 4). Ligand-free insulin shows several transitions between the different sub-states explored and binding mode α has a very compact zone and also different transitions between the sub-states that can ensure a high probability of convergence at 500 ns. Binding mode β, however, seemed to start exploring a new sub-state at the end of 500 ns, so the simulation was extended to 700 ns, when the new space had already been explored and a transition to a previous zone had occurred. Finally, the MD trajectory of binding mode γ was elongated until 1.0 μs, expecting a transition to a previous sub-state (eventually observed at ∼950 ns). Further details and complementary studies of the convergence are provided in the ESI (Fig. S4–S8).

image file: c8mt00341f-f4.tif
Fig. 4 PCA analysis for ligand-free insulin, binding mode α, binding mode β, and binding mode γ. Plot of the two principal components (PC1 and PC2). Sub-states visited along the trajectories are depicted with a colour map as a function of the simulation time.

For both binding sites, hydrogen bond network appears to be an important stabilization factor along the MD trajectory. Complex Pt(dach)(OH)(His10B)–insulin exhibits two hydrogen bonds of intermediate strength at different times of the dynamics. Up to ∼160 ns, the interaction predicted by docking between the OH ligand and the carboxylate of Glu13 (see Fig. S9 of the ESI) shows a series of bound and unbound states until the hydrogen bond is definitely broken due to the reorientation of the Pt(dach)(OH)+ moiety. From 500 up to 700 ns, a hydrogen bond between the amino group of dach and the His5 backbone carbonyl stabilizes the adduct. Complex Pt(dach)(His5B){Cys7B(-SS-)}–insulin shows a unique strong hydrogen bond between the amino groups of dach and Cys7A with a mean distance of 2.8 Å (see Fig. S9 of the ESI).

Concerning the hypothesis of interaction with the Cys7B-SS-Cys7A disulphide bridge, we considered, at the DFT level of theory, the possibility of the disulphide reduction modifying the charge of the model system during the optimization. Under these conditions and without any restrain, the bridge rupture and the reorientation of the Cys7B side chain occur spontaneously. Subsequently, from the optimized geometry of this model, we extracted the force constants relative to the metal centre and performed MD simulations. The MD trajectory shows the unfolding of the B chain containing the [Pt(dach)]2+ moiety after 10 ns, and a reorientation of one of the helices in chain A (Gly1–Thr8) is also observed after ∼200 ns of simulation. Both processes are shown in Fig. 5.

image file: c8mt00341f-f5.tif
Fig. 5 (a) Schematic representation of the unfolding process after S–S disulphide reduction during the MD trajectory. (b) Schematic representation of the reorientation process of the helix in chain A (Gly1–Thr8).

Cluster analysis was performed to obtain the most representative structure of each MD trajectory. For each MD trajectory, the geometry and the reliability of the α, β and γ hypothetical binding modes were computed using QM/MM calculations describing PtII and its first coordination sphere at the QM level (as reported in the Computational section), which involves Cys7B, Cys7A and His10B for binding mode α, Cys7B, His5B and OH ligands for binding mode β, and Cys7B, and His10B for binding mode γ. The AMBER99SB force field was applied for the MM region. In Fig. 6 the optimized structures of α, β and γ binding modes are shown, while Table 2 contains selected bond lengths and angles. The refined geometries of the predicted adducts appear structurally coherent with well-established X-ray structures of oxaliplatin co-crystalized with lysozyme (PDB codes 4zee,514z46,51 and 4ppo52) and bovine ribonuclease (PDB code 4s1832) for which three types of coordination compounds have been characterized: Pt(dach)(OH/OH2)(COO-Asp), and Pt(dach)(OH/OH2)(N-His) or Pt(dach)(OH/OH2)(S-Met), respectively. A comparison of the most relevant parameters of the predicted adducts and the X-ray structures is reported in Table 2.

image file: c8mt00341f-f6.tif
Fig. 6 Three QM/MM optimized structures of the α, β, and γ binding modes for the interaction of [Pt(dach)]2+ and [Pt(dach)(OH)]+ moieties with insulin. The hydrogen bonds are highlighted with solid blue lines.
Table 2 Selected bond lengths (Å) and angles (degrees) for the adducts Pt(dach)(His5B){Cys7B(-SS-)}–insulin (binding mode α), Pt(dach)(OH)(His10B)–insulin (binding mode β) and Pt(dach)(His5B){Cys7B(S)}–insulin (binding mode γ) QM/MM optimized using Garleek and the X-ray parameters extracted from PDB 4zee, 4z46, 4ppo and 4s18
Parametere Mode α Mode β Mode γ 4zeea 4z46b 4ppoc 4s18 (a)d 4s18 (b)d
Calcd Calcd Calcd Exptl Exptl Exptl Exptl Exptl
a Ref. 51. b Ref. 51. c Ref. 52. d Ref. 32. e The atom indices are those reported in Fig. 6 and Fig. S10 of the ESI.
Pt–N(1) 2.094 2.115 2.084 2.175 2.035 1.980 2.031 2.034
Pt–N(2) 2.099 2.069 2.167 2.021 2.054 2.095 2.054 2.055
Pt–N/O(3) 2.071 2.034 2.050 1.661 2.058 1.978 2.304 2.235
Pt–O/S/N(4) 2.396 1.985 2.332 2.653 2.335 2.054 2.209 2.188
N(1)–Pt–N(2) 81.2 82.4 81.8 86.3 84.5 86.1 88.4 84.4
N(2)–Pt–N/O(3) 92.3 97.1 94.6 110.9 109.9 94.3 85.4 95.5
N/O(3)–Pt–O/S(4) 88.6 92.7 84.2 101.3 76.4 94.5 99.0 71.5
O/S/N(4)–Pt–N(1) 98.0 88.0 99.0 61.5 92.4 89.5 87.6 108.6


Metallodrug design attracts increasing attention especially because of the success of cisplatin and its derivatives in anticancer therapy. Nonetheless, the dynamical behaviour of the first coordination sphere of the metal and the lability of its ligands in vivo allow numerous off-target interactions that could have major significance in phenomena ranging from bioavailability to resistance. Here, we applied to the interaction of oxaliplatin with insulin our recent strategies for the prediction of metallodrug bindings to proteins and peptides. Our approach provides not only candidate binding sites of the drug to the target consistent with the experimental information obtained from fragment analysis, but also extensive 3D structures. The docking analysis reveals three apparently isoenergetic binding sites, the stabilities of which are assessed throughout MD simulations followed by QM/MM calculations. Only one of these sites could lead to substantial conformational changes of insulin if the reduction of the disulphide bridge occurs after the drug binding. Moreover, the simultaneous binding of two Pt moieties, [Pt(dach)]2+ and [Pt(dach)(OH)]+, to insulin was also assessed and confirmed. This study presents a multiscale approach for predicting the binding of metallodrugs to protein hosts, its implications in conformational changes of bioligands and the potential effects on their biological functions.


CDCircular dichroism
DFTDensity functional theory
ENDORElectron nuclear double resonance
EPRElectron paramagnetic resonance
ESEEMElectron spin echo envelope modulation
ESI-MSElectrospray ionization mass spectrometry
GAFFGeneralized Amber force field
LCLiquid chromatography
MALDI-ToF-MSMatrix-assisted laser desorption/ionization time-of-flight mass spectrometry
MDMolecular dynamics
MRIMagnetic resonance imaging
nESI-Q-ToF-MSNano-electrospray ionization quadrupole time-of-flight mass spectrometry
PCAPrincipal component analysis
QM/MMQuantum mechanics/molecular mechanics
RESPRestrained electrostatic potential
RMSDRoot mean square deviation

Conflicts of interest

There are no conflicts to declare.


J.-D. M., G. S., J. R.-G. and J.-E. S.-A. are thankful for the support given by the Spanish grant CTQ2017-87889-P and the Generalitat de Catalunya grant 2017SGR1323. The support of COST Action CM1306 is kindly acknowledged. G. S. thanks the Universitat Autònoma de Barcelona for its support of his PhD program. J.R.-G. acknowledges the Generalitat de Catalunya and the European Social fund for his PhD grant (2017FI_B2_00168). E. G. acknowledges the Fondazione di Sardegna (project FdS15Garribba) and the FFABR 2017 “Fondo per il finanziamento delle attività base di ricerca” for the financial support.


  1. Z. Guo and P. J. Sadler, Metals in Medicine, Angew. Chem., Int. Ed., 1999, 38, 1512–1537 CrossRef.
  2. N. Farrell, in Comprehensive Coordination Chemistry II, ed. J. A. McCleverty and T. J. Meyer, Pergamon, Oxford, 2003, vol. 9, pp. 809–840 Search PubMed.
  3. Metallotherapeutic Drugs and Metal-Based Diagnostic Agents. The Use of Metals in Medicine, ed. M. Gielen and E. R. T. Tiekink, John Wiley & Sons Ltd, Chichester, 2005 Search PubMed.
  4. K. H. Thompson, in Encyclopedia of Inorganic and Bioinorganic Chemistry, ed. R. B. King, John Wiley & Sons, Ltd, Chichester, 2011 Search PubMed.
  5. Interrelations between Essential Metal Ions and Human Diseases, ed. A. Sigel, H. Sigel and R. K. O. Sigel, Springer Science + Business Media, Dordrecht, 2013 Search PubMed.
  6. K. D. Mjos and C. Orvig, Metallodrugs in Medicinal Inorganic Chemistry, Chem. Rev., 2014, 114, 4540–4563 CrossRef CAS PubMed.
  7. S. Medici, M. Peana, V. M. Nurchi, J. I. Lachowicz, G. Crisponi and M. A. Zoroddu, Noble Metals in Medicine: Latest Advances, Coord. Chem. Rev., 2015, 284, 329–350 CrossRef CAS.
  8. C. Jones and J. Thornback, Medicinal Applications of Coordination Chemistry, The Royal Society of Chemistry, Cambridge, 2007 Search PubMed.
  9. J. C. Dabrowiak, Metals in Medicine, John Wiley & Sons, Ltd, 2009 Search PubMed.
  10. Bioinorganic Medicinal Chemistry, ed. E. Alessio, Wiley-VCH Verlag GmbH & Co. KGaA, 2011 Search PubMed.
  11. J. Graham, M. Muhsin and P. Kirkpatrick, Oxaliplatin, Nat. Rev. Drug Discovery, 2004, 3, 11–12 CrossRef CAS PubMed.
  12. S. G. Chaney, S. L. Campbell, E. Bassett and Y. Wu, Recognition and Processing of Cisplatin– and Oxaliplatin–DNA Adducts, Crit. Rev. Oncol. Hematol., 2005, 53, 3–11 CrossRef PubMed.
  13. L. Kelland, The Resurgence of Platinum-Based Cancer Chemotherapy, Nat. Rev. Cancer, 2007, 7, 573–584 CrossRef CAS PubMed.
  14. J. Chen, X.-F. Huang, L. Qiao and A. Katsifis, Insulin Caused Drug Resistance to Oxaliplatin in Colon Cancer Cell Line Ht29, J. Gastrointest. Oncol., 2011, 2, 27–33 CAS.
  15. D. Hagrman, J. Goodisman, J. C. Dabrowiak and A.-K. Souid, Kinetic Study on the Reaction of Cisplatin with Metallothionein, Drug Metab. Dispos., 2003, 31, 916–923 CrossRef CAS PubMed.
  16. J. Costa Pessoa, E. Garribba, M. F. A. Santos and T. Santos-Silva, Vanadium and Proteins: Uptake, Transport, Structure, Activity and Function, Coord. Chem. Rev., 2015, 301–302, 49–86 CrossRef CAS.
  17. A. Casini, G. Mastrobuoni, C. Temperini, C. Gabbiani, S. Francese, G. Moneti, C. T. Supuran, A. Scozzafava and L. Messori, ESI Mass Spectrometry and X-Ray Diffraction Studies of Adducts between Anticancer Platinum Drugs and Hen Egg White Lysozyme, Chem. Commun., 2007, 156–158 RSC.
  18. J. Li, L. Yue, Y. Liu, X. Yin, Q. Yin, Y. Pan and L. Yang, Mass Spectrometric Studies on the Interaction of Cisplatin and Insulin, Amino Acids, 2016, 48, 1033–1043 CrossRef CAS PubMed.
  19. E. Moreno-Gordaliza, B. Canas, M. A. Palacios and M. M. Gomez-Gomez, Novel Insights into the Bottom-up Mass Spectrometry Proteomics Approach for the Characterization of Pt-Binding Proteins: The Insulin–Cisplatin Case Study, Analyst, 2010, 135, 1288–1298 RSC.
  20. E. Moreno-Gordaliza, B. Cañas, M. A. Palacios and M. M. Gómez-Gómez, Top-Down Mass Spectrometric Approach for the Full Characterization of Insulin–Cisplatin Adducts, Anal. Chem., 2009, 81, 3507–3516 CrossRef CAS PubMed.
  21. M. Rupasri, M. B. Sawyer and L. Xing-Fang, Mass Spectrometry Study of Hemoglobin-Oxaliplatin Complexes in Colorectal Cancer Patients and Potential Association with Chemotherapeutic Responses, Rapid Commun. Mass Spectrom., 2006, 20, 2533–2538 CrossRef PubMed.
  22. A. P. Ryle, F. Sanger, L. F. Smith and R. Kitai, The Disulphide Bonds of Insulin, Biochem. J., 1955, 60, 541–556 CrossRef CAS PubMed.
  23. C. Møller, R. R. Sprenger, S. Stürup and P. Højrup, Determination of the Binding Sites for Oxaliplatin on Insulin Using Mass Spectrometry-Based Approaches, Anal. Bioanal. Chem., 2011, 401, 1619–1629 CrossRef PubMed.
  24. G. Sciortino, E. Garribba and J.-D. Maréchal, Validation and Applications of Protein–Ligand Docking Approaches Improved for Metalloligands with Multiple Vacant Sites, Inorg. Chem., 2019, 58, 294–306 CrossRef CAS PubMed.
  25. G. Sciortino, J. Rodríguez-Guerra Pedregal, A. Lledós, E. Garribba and J.-D. Maréchal, Prediction of the Interaction of Metallic Moieties with Proteins: An Update for Protein–Ligand Docking Techniques, J. Comput. Chem., 2018, 39, 42–51 CrossRef CAS PubMed.
  26. G. Sciortino, D. Sanna, V. Ugone, G. Micera, A. Lledós, J.-D. Maréchal and E. Garribba, Elucidation of Binding Site and Chiral Specificity of Oxidovanadium Drugs with Lysozyme through Theoretical Calculations, Inorg. Chem., 2017, 56, 12938–12951 CrossRef CAS PubMed.
  27. D. Sanna, V. Ugone, G. Sciortino, P. Buglyo, Z. Bihari, P. L. Parajdi-Losonczi and E. Garribba, VIVO Complexes with Antibacterial Quinolone Ligands and Their Interaction with Serum Proteins, Dalton Trans., 2018, 47, 2164–2182 RSC.
  28. P. W. Rose, A. Prlić, A. Altunkaya, C. Bi, A. R. Bradley, C. H. Christie, L. D. Costanzo, J. M. Duarte, S. Dutta, Z. Feng, R. K. Green, D. S. Goodsell, B. Hudson, T. Kalro, R. Lowe, E. Peisach, C. Randle, A. S. Rose, C. Shao, Y.-P. Tao, Y. Valasatava, M. Voigt, J. D. Westbrook, J. Woo, H. Yang, J. Y. Young, C. Zardecki, H. M. Berman and S. K. Burley, The Rcsb Protein Data Bank: Integrative View of Protein, Gene and 3d Structural Information, Nucleic Acids Res., 2017, 45, D271–D281 CrossRef CAS PubMed.
  29. G. Bentley, E. Dodson, G. Dodson, D. Hodgkin and D. Mercola, Structure of Insulin in 4-Zinc Insulin, Nature, 1976, 261, 166–168 CrossRef CAS PubMed.
  30. K. Mandal, B. Dhayalan, M. Avital-Shmilovici, A. Tokmakoff and S. B. H. Kent, Crystallization of Enantiomerically Pure Proteins from Quasi-Racemic Mixtures: Structure Determination by X-Ray Diffraction of Isotope-Labeled Ester Insulin and Human Insulin, ChemBioChem, 2016, 17, 421–425 CrossRef CAS PubMed.
  31. E. F. Pettersen, T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng and T. E. Ferrin, UCSF Chimera—a Visualization System for Exploratory Research and Analysis, J. Comput. Chem., 2004, 25, 1605–1612 CrossRef CAS PubMed.
  32. L. Messori, T. Marzo and A. Merlino, Interactions of Carboplatin and Oxaliplatin with Proteins: Insights from X-Ray Structures and Mass Spectrometry Studies of Their Ribonuclease a Adducts, J. Inorg. Biochem., 2015, 153, 136–142 CrossRef CAS PubMed.
  33. G. Jones, P. Willett, R. C. Glen, A. R. Leach and R. Taylor, Development and Validation of a Genetic Algorithm for Flexible Docking, J. Mol. Biol., 1997, 267, 727–748 CrossRef CAS PubMed.
  34. G. Jones, P. Willett and R. C. Glen, Molecular Recognition of Receptor Sites Using a Genetic Algorithm with a Description of Desolvation, J. Mol. Biol., 1995, 245, 43–53 CrossRef CAS PubMed.
  35. S. C. Lovell, J. M. Word, J. S. Richardson and D. C. Richardson, The Penultimate Rotamer Library, Proteins: Struct., Funct., Bioinf., 2000, 40, 389–408 CrossRef CAS.
  36. J. Rodríguez-Guerra, Insilichem/Gaudiview: Pre-Alpha Public Release, Zenodo, 2017,  DOI:10.5281/zenodo.556352.
  37. J. Rodríguez-Guerra Pedregal, G. Sciortino, J. Guasp, M. Municoy and J.-D. Maréchal, GaudiMM: A Modular Multi-Objective Platform for Molecular Modeling, J. Comput. Chem., 2017, 38, 2118–2126 CrossRef PubMed.
  38. H. Viktor, A. Robert, O. Asim, S. Bentley, R. Adrian and S. Carlos, Comparison of Multiple Amber Force Fields and Development of Improved Protein Backbone Parameters, Proteins: Struct., Funct., Bioinf., 2006, 65, 712–725 CrossRef PubMed.
  39. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford, CT, 2010 Search PubMed.
  40. A. Ehlers, M. Böhme, S. Dapprich, A. Gobbi, A. Höllwarth, V. Jonas, K. Köhler, R. Stegmann, A. Veldkamp and G. Frenking, A Set of F-Polarization Functions for Pseudo-Potential Basis Sets of the Transition Metals Sc–Cu, Y–Ag and La–Au, Chem. Phys. Lett., 1993, 208, 111–114 CrossRef CAS.
  41. C. I. Bayly, P. Cieplak, W. Cornell and P. A. Kollman, A Well-Behaved Electrostatic Potential Based Method Using Charge Restraints for Deriving Atomic Charges: The RESP Model, J. Phys. Chem., 1993, 97, 10269–10280 CrossRef CAS.
  42. P. Li and K and M. Merz, A Python Based Metal Center Parameter Builder, J. Chem. Inf. Model., 2016, 56, 599–604 CrossRef PubMed.
  43. P. Eastman, J. Swails, J. D. Chodera, R. T. McGibbon, Y. Zhao, K. A. Beauchamp, L.-P. Wang, A. C. Simmonett, M. P. Harrigan, C. D. Stern, R. P. Wiewiora, B. R. Brooks and V. S. Pande, OpenMM 7: Rapid Development of High Performance Algorithms for Molecular Dynamics, PLoS Comput. Biol., 2017, 13, e1005659 CrossRef PubMed.
  44. J. R.-G. Pedregal, L. Alonso-Cotchico, L. Velasco and J.-D. Maréchal, Ommprotocol: A Command Line Application to Launch Molecular Dynamics Simulations with OpenMM, ChemRxiv, 2018 DOI:10.26434/chemrxiv.7059263.v1.
  45. R. M. B. D. A. Case, W. Botello-Smith, D. S. Cerutti, T. E. Cheatham, III, T. A. Darden, R. E. Duke, T. J. Giese, H. Gohlke, A. W. Goetz, N. Homeyer, S. Izadi, P. Janowski, J. Kaus, A. Kovalenko, T. S. Lee, S. LeGrand, P. Li, C. Lin, T. Luchko, R. Luo, B. Madej, D. Mermelstein, K. M. Merz, G. Monard, H. Nguyen, H. T. Nguyen, I. Omelyan, A. Onufriev, D. R. Roe, A. Roitberg, C. Sagui, C. L. Simmerling, J. Swails, R. C. Walker, J. Wang, R. M. Wolf, X. Wu, L. Xiao, D. M. York and P. A. Kollman, Amber 16 Search PubMed.
  46. Y. Dodge, Statistical Data Analysis Based on the L1-Norm and Related Methods, Birkhäuser, 2012 Search PubMed.
  47. M. P. Harrigan, M. M. Sultan, C. X. Hernández, B. E. Husic, P. Eastman, C. R. Schwantes, K. A. Beauchamp, R. T. McGibbon and V. S. Pande, MSMBuilder: Statistical Models for Biomolecular Dynamics, Biophys. J., 2017, 112, 10–15 CrossRef CAS PubMed.
  48. J. R.-G. Pedregal, I. Funes-Ardoiz, G. Sciortino, J.-E. Sánchez-Aparicio, G. Ujaque, A. Lledós, J.-D. Maréchal and F. Maseras, GARLEEK: Adding an Extra Flavor to ONIOM, J. Comput. Chem., 2019, 40, 381–386 CrossRef CAS PubMed.
  49. R. V. Pappu, R. K. Hart and J. W. Ponder, Analysis and Application of Potential Energy Smoothing and Search Methods for Global Optimization, J. Phys. Chem. B, 1998, 102, 9725–9742 CrossRef CAS.
  50. Y.-S. Kim, S.-M. Shin, M.-S. Cheong and S.-S. Hah, Mechanistic Insights into in vitro DNA Adduction of Oxaliplatin, Bull. Korean Chem. Soc., 2010, 31, 2043–2046 CrossRef CAS.
  51. D. Marasco, L. Messori, T. Marzo and A. Merlino, Oxaliplatin Vs. Cisplatin: Competition Experiments on Their Binding to Lysozyme, Dalton Trans., 2015, 44, 10392–10398 RSC.
  52. L. Messori, T. Marzo and A. Merlino, The X-Ray Structure of the Complex Formed in the Reaction between Oxaliplatin and Lysozyme, Chem. Commun., 2014, 50, 8360–8362 RSC.


Electronic supplementary information (ESI) available: This includes Supplementary Details, Supplementary Fig. S1–S10, and Supplementary Tables S1–S4. See DOI: 10.1039/c8mt00341f
These authors have contributed equally to this work.

This journal is © The Royal Society of Chemistry 2019