TMCnet News

Synthesis and Characterization of a Stable Copper(I) Complex for Radiopharmaceutical Applications [ChemPlusChem]
[October 30, 2014]

Synthesis and Characterization of a Stable Copper(I) Complex for Radiopharmaceutical Applications [ChemPlusChem]


(ChemPlusChem Via Acquire Media NewsEdge) A highly stable copper(I) complex was obtained starting from a copper(II) salt. This compound was characterized by a combination of several analytical techniques (UV/Vis spectroscopy, energy-dispersive X-ray spectroscopy, electrochemistry, and X-ray photoelectron spectroscopy) and was shown to present an N4 Cu structure. These results were confirmed by a density functional calculations study of the binding energy and the electronic structure of model ligand and copper complexes. Preliminary tests of complexation showed a high ability of the corresponding ligand to chelate 64 Cu in very diluted medium, which is of interest for developing new positron emission tomography imaging agents. The stability and the kinetic inertness of the complex are promising. In particular, it displayed good redox stability, which is important because in vivo reduction or oxidation of the copper of Cu complexes can lead to demetalation. The rapid microwave-assisted strategy used to synthesize the ligand was applied to the synthesis of more than ten ligands. One of these was functionalized by an amino group to form a bifunctional chelate for a future bioconjugation for applications in nuclear medicine.



Keywords : chelates · copper · density functional calculations · ligand design · radiopharmaceuticals Introduction Over recent years, medical imaging has become one of the main interests in therapeutic chemistry. Progress in instrumen- tation and the need to diagnose diseases at their early stages are some of the reasons for such an interest. Moreover, identi- fying the target(s) of a drug and understanding the mecha- nism of action of a new compound are becoming essential re- quirements for their acceptance in clinical-phase trials. This has led to the development of the new field of theranostics. The term theranostics has been coined to describe this emerging area of research, which focuses on agents that could be used in both imaging and therapy.[1] In our case, this consists of taking advantage of the duality of an element such as copper, which possesses both a b + emitter (64Cu), used as a positron emission tomography (PET) imaging agent, and a b^ emitter (67Cu), used as a radiotherapeutic entity.

The long half-life of 64Cu (t1= = 12.7 h) allows a satisfactory accumulation of radionuclides in the tumor.[2] Furthermore, 64Cu presents a good spatial resolution for PET imaging, equiv- alent to 18F.[3] Some studies have shown that it gives the high- est effective dose and, above all, that it can be used in prelimi- nary dosimetry studies for radioimmunotherapy (use of radiola- beled antibodies to target tumor cells and destroy them).[4] In fact, the key problem in radioimmunotherapy is to be able to measure the amount of radioactivity bound to the tumor cells (as this depends on the kind of cancer, the vector, and even on the patient). Thus, with 64Cu, a PET image can be obtained and the radioactivity can be quantified. As the same vector can be used to transport 67Cu to tumor cells, the pharmacoki- netics will be the same and the quantity of 67Cu linked to the tumor will be known.


To target copper selectively to cancerous tissues, a peptide, peptidomimetic, or an antibody can be used. The key element of this strategy is the bifunctional chelate (BFC), the entity that binds the radionuclide to the vector. An ideal BFC binds the targeted radiometal rapidly with a high specificity and yield. Furthermore, the resulting complex should be both thermody- namically stable and, more importantly, kinetically inert toward in vivo transchelation or transmetalation processes. To form a complex with copper, a large variety of acyclic, macrocyclic, and macrobicyclic polyamine BFCs have been studied and re- ported.[5] However, the choice of the chelator is critical given that recent studies have shown that it can influence the radio- labeling of the bioconjugate, its targeting, and its pharmacoki- netics.[6] Clearly, it is still important to develop new copper che- lators. Cyclen and cyclam macrocyclic derivatives have commonly been used as copper chelators, owing to their commer- cial availability, their good thermodynamic stability, and their relatively rapid complexation kinetics. Nevertheless, almost all of these BFCs have been shown to be susceptible to the re- duction and release of copper, which is then transchelated to proteins present in both blood and liver and results in high liver uptake.[7] To overcome this poor kinetic inertness, cross- bridged or macrobicyclic derivatives (e.g. , CB-TE2A or SarAr) have been studied as copper chelators but have shown either very slow complexation kinetics[8] or poor electrochemical sta- bility.[9] Copper complexes of bis(thiosemicarbazone) ligands represent another important class of compounds and have been investigated extensively, mainly for hypoxia imaging, in the form of Cu-ATSM.[10] These kinds of ligands are of interest because of their high complexation kinetics, the ease and effi- ciency of their synthesis, and their ability to form neutral com- plexes. Nevertheless, even though they show an excellent ther- modynamic stability, copper complexes of bis(thiosemicarba- zone) ligands have sometimes been reported as having a poor kinetic inertness.[11] Our group reported the synthesis of new tetradentate li- gands, with a bis(thiosemicarbazone) skeleton, which display a high ability for chelating 64Cu in very diluted medium.[12] Al- though these results were promising, some elements were missing to consider these ligands good BFCs. First, the copper complex obtained needed to be fully characterized : 1) ligand/ metal stoichiometry, 2) oxidation state of the copper, 3) number of different complexes formed, and 4) atoms chelat- ing the metal. Then, the stability of the complex had to be checked : 1) in acidic medium, 2) in the presence of other li- gands or metals, and 3) upon redox processes (the reactions CuII$CuI are the main explanation for the poor kinetic inert- ness of most copper complexes). Finally, the ligand needed to possess a group that could be functionalized, such as "NH2", for coupling with the vector.

Results and Discussion A preliminary study was performed with the ligand L1 (Scheme 1), which was chosen because the chlorine atoms on this chelating agent could be used as probes for further energy-dispersive X-ray spectrosco-obtained in a three-step micro- wave-assisted strategy described previously.[12] The copper complex CuL1 was synthesized by the reac- tion of one equivalent of ligand L1 with one equivalent of copper(II) diacetate in dimethylformamide (DMF) for one hour at room tem- perature or for five minutes at 110 8C under microwave (MW) irra- diation (Figure 1).

Characterization of the complex CuL1 Although various crystallogenesis experiments (gel permeation, "H-tube" crystallization, and slow solvent evaporation) were performed, a monocrystal suitable for X-ray analysis could not be obtained. Unfortunately, NMR and EPR studies of CuL1 were unsuccessful (relaxation was very fast and thus the NMR signals were very broad and no significant EPR signal was ob- served). Several analytical techniques were therefore combined to determine the structure of the complex CuL1.

Determination of the stoichiometry of the complex CuL1 The significant color change during the reaction (see Figure 1) prompted us to study its evolution by UV/Vis spectrometry. The UV/Vis spectrum of L1 displayed two main absorption bands at 395 and 505 nm. On recording the spectrum of the complex CuL1, only one main absorption band appeared at 440 nm (Figure 2). A deconvolution of this spectrum showed that no signal of L1 remained. Moreover, monitoring by UV/Vis spectroscopy over time clearly showed isosbestic points, and the addition of a further one or two equivalents of copper did not change the spectrum (Figures S1 and S2 in the Supporting Information). It could thus be concluded that with one equivalent of copper(II), all the ligand reacted with the copper (stoi- chiometry L1/Cu : 1:1).

To confirm this conclusion, an EDX analysis linked to scan- ning electron microscopy (SEM) was performed on L1 and CuL1. The results pointed to the presence of two equivalents of sulfur or of chlorine and one of copper. That meant that the complex had a ligand/copper stoichiometry of 1:1. Moreover, we took advantage of this analysis to take a picture of L1 and CuL1, which displayed striking differences in appearance (Figure 3). This stoichiometry was confirmed by mass spec- trometry ; the spectra (MALDI, 2,5-dihydroxybenzoic acid (DHB) matrix) for CuL1 possess an isotopic pattern corresponding to (CuL1 + H+)atm/z = 667.

Determination of the oxidation state of the metal Copper often undergoes oxidation and reduction reactions, so it is worth checking its oxidation state. In the complexation re- action, copper(II) was employed. To determine if it was Cu0, CuI,orCuII in CuL1, a voltamperometric study was performed. The electrochemical behavior of both free ligand L1 and copper complex CuL1 was examined by cyclic voltammetry in DMF containing 0.1 mol L^1 tetrabutylammonium hexafluoro- phosphate. The free ligand L1 displayed an irreversible process at a potential of about 0.85 V versus a saturated calomel elec- trode (SCE ; Figure 4) that could be attributed to the electro- chemical oxidation of S or N centers. The electrochemical be- havior of the complex CuL1 was thus examined within the potential window in the range of ^0.9 to 0.7 V, at which no elec- trochemical response was observed for L1. Figure 5 displays a quasi-reversible process for CuL1 with anodic and cathodic peaks at potentials around 0.55 and 0.30 V versus SCE, respec- tively. From the comparison between the voltammograms of L1 and CuL1, a quasi-reversible redox process can be assigned to the copper complex indicating a stable binding between L1 and cationic copper metal. On performing cyclic voltammetry at different scan rates (10, 20, 50, 100, and 150 mV s^1), the peak intensity increased with the square root of the scan rate, thus indicating that the electrochemical response is controlled by the diffusion process with anodic peak current intensity Ipa=3.02v=2 +0.01 in which 3.02 is the slope (mA) and v is the scan rate (mV s^1). Moreover, the ratio of the cathodic to anodic peak was close to 1 at a low scan rate (10 mV s^1) and decreased to around 0.7 at a high scan rate (150 mV s^1). The peak-to-peak potential DE also increased from 121 to 203 mV, respectively, when the scan rate increased from 10 to 150 mV s^1. Based on these results, one can conclude a quasi- reversible behavior of the complex CuL1 at the electrode/solu- tion interface.

On scanning the potential in the cathodic direction, note that no electrochemical response was observed until ^1.2 V versus SCE (Figure 5 A). Below this value, an anodic peak was observed at reverse scan at around ^0.240 V with a sharp shape. On the other hand, the electrochemical behavior of free Cu2 + under the same experimental conditions showed cathodic and anodic waves at ^0.05 and ^0.367 V, respectively (Fig- ure 5 B). Based on this, we attributed the electrochemical be- havior observed in Figure 5 A to the anodic redissolving of Cu0, previously deposited by the reduction of CuL1 at potential values below ^1.2 V versus SCE. These results indicate that the quasi-reverse electrochemical behavior displayed in Figure 4 may easily be attributed to the redox process of the CuII com- plex/CuI complex system.

The oxidation state of the copper was analyzed by X-ray photoelectron spectroscopy (XPS). The part of the CuL1 XPS spectrum corresponding to copper displayed two peaks at 953.5 and 933.6 eV (Figure S3). These values and the absence of two satellite peaks (shake-up) confirmed the oxidation state of the metal, and the formation of a CuI complex.[13] The reduction of CuII salt to CuI was unexpected. It was par- ticularly surprising to obtain a CuI complex that was totally air- stable both in the solid state and in solution (for several months). This phenomenon is not easy to explain. According to the L1 voltammogram, L1 cannot be the reducing agent. We also studied the redox behavior of L1 in basic medium to see if the acetate ligands of Cu(OAc)2 could affect the redox properties of L1. The work of Minaev and Bondarchuk perfectly describes the numerous mechanisms that can explain the re- duction of CuII to CuI.[14] One, in particular, attracted our atten- tion : CuCl2 and Cu(OAc)2 can both be reduced in situ by the solvent (e.g. , acetone).[15] We think that such a reaction could occur in our case and the high stability of the CuI complex formed could act as the driving force. This result is noteworthy because, although the chemistry of copper coordination is do- minated by the complexes of CuII, many CuI derivatives are of some interest.[5e, 16] Moreover, it is important to have a CuI com- plex for medical applications as the reduction of CuII com- plexes in vivo often leads to demetalation.

Using Cu(p-CF3C6H4SO3), Cu(OAc), and Cu(CH3CN)4BF4 as the source of copper(I), similar complexation experiments were performed with L1, in DMF under microwave irradiation, to see if the same type of UV/Vis spectra were obtained as with Cu(OAc)2. Unfortunately, we came up against the problem of the varying degree of solubility of copper(I) salts under the ex- perimental conditions employed. The best results were ob- tained with Cu(CH3CN)4BF4 for which the same absorption band at 440 nm was observed, but at a lower intensity than that of Cu(OAc)2 owing to the partial insolubility of CuI (Fig- ure S4). This again supports the formation of a copper(I) com- plex.

Determination of the number of different complexes formed When designing our ligand, we wanted an ambidentate chelat- ing agent able to bind an N4,N3S, or N2S2 ligand. To be ambi- dentate, the ligand contains two thiazole rings. They are bound to the ligand core by a single bond so that they can rotate. Our hypothesis was that a hard metal would be chelat- ed by the four nitrogen atoms (N4 complex), whereas a softer metal would entail a rotation of the thiazole rings, thus ena- bling sulfur atoms to replace one or two of the nitrogen atoms (N3SorN2S2 complexes ; Figure 6).

This special feature increases the number of different com- plexes that can be obtained. UV/Vis spectroscopy analyses and an electrochemical study seemed to indicate the formation of just one complex. This was confirmed by analyzing the part of the CuL1 XPS spectrum corresponding to copper : there was just one system of two signals for copper (corresponding to 2p=2 and 2p =2 copper electrons), which confirmed that just one type of complex was formed.

Determination of the atoms chelating the metal The environment around the copper was determined by XPS. The signals displayed by the sulfur atom in the XPS spectra of L1 and CuL1 were similar (Figure S5). Their environment is thus very similar, which probably excludes a chelation of copper by the sulfur atoms. These results are in agreement with the recent work of Jin et al. , who showed the formation of copper(I) complexes by chelation of the nitrogen atom of a thiazole.[17] On the contrary, the signals displayed by nitrogen atoms in the spectra of L1 and CuL1 were very different (Figure 7). A deconvolution of the "N-L1" signal led to three peaks with the same area. Thus, each peak represents two nitrogen atoms of the ligand. This result was expected because of the symmetry of the molecule. However, the symmetry was lost after complexation and three different peaks were ob- served after the deconvolution of the "N-CuL1" signal. These results suggest that the complexation probably took place by nitrogen atoms.

According to the work of Deng et al. ,[18] the binding energy of the peak at 402.1 eV (peak A, Figure 7) may correspond to 1s electrons involved in a nitrogen-copper covalent bond. Thus, an explanation of the three different peaks for the CuL1 complex could be that one of the nitrogen atoms was cova- lently bound to copper whereas the others formed longer-dis- tance links. Following this hypothesis, the formation of the co- valent bond could result from the deprotonation of NH of one of the hydrazones by the acetate anion. Thus, by electronic de- localization, a Cu^N bond would form with the nitrogen of a thiazole.[19] Infrared spectra of L1 and CuL1 were obtained to give more clues. The decrease in the N^H band signals was significant for the CuL1 spectrum but the signals did not dis- appear. These observations are consistent with the deprotona- tion of one of the two N H bonds.

In view of these analyses, it could be assumed that the com- plex CuL1 forms N4 chelate rings in which the copper(I) is probably coordinated by N atoms with one covalent bond (Figure 6). Although two-coordinated linear and three-coordi- nated trigonal arrangements are known, CuI complexes are mostly four-coordinated species adopting a tetrahedral geom- etry.

To rationalize the experimental results, a model of ligand and copper complexes was optimized in vacuo by using density functional theory (DFT). The selected geometric parameters of computed complexes are presented in Table 1. In all these conformers, namely (CuLH2)+-N4, (CuLH2)+-N3S1, and (CuLH2)+ -N2S2, the metal cation interacts mainly with two centers of the ligand. The distances between ligating atoms and the Cu+ ion lie between 1.92 and 1.95 | for the binding modes with the nitrogen of the thiazole and between 2.27 and 2.30 | for the binding modes with the sulfur of the thiazole, respectively (Table 1). These values are partially in agreement with experimental structural data from the Cambridge Structural Database (CSD) for small molecules and the Protein Data Bank (PDB): the copper(I)-binding complexes in the CSD have coordinating distances of 1.90(6) | for CuN and 2.17(2) | for CuS in bidentate ligands.[20] The binding mode with the sulfur of the thiazole is disadvantaged in such complexes. In fact, the stabilization energy is lower when the sulfur is involved in the complexation of copper.

The characteristics of hydra- zone complexes are often relat- ed to the deprotonation/proto- nation behavior of the hydrazone ligands. We report here the results obtained for the mon- odeprotonated (CuLH) and dideprotonated (CuL)^ complexes (Table 1).

As for the diprotonated form, the structures of lowest energy are the N4 conformers (Figure 8). For the monodeproto- nated complex, the structure is not quite the same as for the dideprotonated structure. The distances to the nitrogen shell in the first coordination sphere of the copper are 1.98 and 1.94 ^ for the (CuLH)-N4 and 1.94 ^ for the (CuL)^-N4 com- plexes. In both cases, the ligands chelate in a bidentate manner, with only the thiazole groups coordinating through nitrogen atoms. The copper stabilization energy is 788.75 and 1057.21 kJ mol^1 for the (CuLH)-N4 and (CuL)^-N4 complexes, respectively. The complexation of cations with anionic hydra- zones may lead to more stable complexes than with neutral hydrazones (CuLH2) + -N4 because of the presence of electro- static interactions (calculations are detailed in the Supporting Information).

This theoretical study shows that, in vacuo, the copper ligand complex is 1:1 regardless of the state of protonation. In addition, the complex formed is of the type N4 (Figure 6) in which copper is complexed mainly by the two nitrogen atoms of the thiazoles, thus confirming the results obtained by XPS.

Stability of complexes Stability and kinetic inertness are the essential properties needed for complexes used in nuclear medicine. They prevent compounds from releasing radionuclides in the human body because of molecules able to chelate copper (e.g. , albumin, metallothionein) or free metallic ions. First, the stability of the complex in acidic conditions was checked. To do this, HCl (100 mL, concentration 0.1 m) was added directly to a solution of CuL1. Immediately, a broad absorption band appeared from 380 to 400 nm whereas the 440 nm band almost vanished. This new band must correspond to a protonated species of the complex. Moreover, the neutralization of the solution by the same amount of NaOH (100 mL, concentration 0.1 m) led to a spectrum very similar to the initial spectrum of CuL1 (the ac- quisition was done immediately after the addition ; Figure S6). Even after several hours in the presence of acid, the initial spectrum was always recovered. It is worth remembering that the complexation was complete after one hour at room tem- perature, so acid could not have destroyed the complex, and that after addition of base, the complex was formed again in less than three minutes (time needed to acquire a spectrum). We thus concluded that CuL1 is stable at pH 4.

The stability of the CuL1 complex compared to other chelat- ing agents of copper(II) (1,4,7,10-tetraazacyclododecane- 1,4,7,10-tetraacetic acid (DOTA) and ethylenediaminetetraacetic acid (EDTA)) and copper(I) (bicinchonic acid and bathocuproine disulfonate) was investigated. Complex CuL1 was reacted with an excess of ligands in a DMF/acetate buffer (1:1 v/v) or in DMF. In every case, no bands characteristic of ligand L1 ap- peared, even after 24 hours, which indicated that copper was not released. These experiments suggest a good stability for CuL1 under the conditions employed. To determine the risk of release of copper in the body because of another metal, a transmetalation reaction was tried using Zn(OAc)2. The choice of zinc was justified by the similarity of the hard-soft acid-base (HSAB) character of copper and zinc and the similari- ty of their biological targets (it is known that a trace-element cure containing a large amount of zinc causes copper deple- tion). The complex ZnL1 was synthesized by reacting one equivalent of L1 with one equivalent of Zn(OAc)2. The lmax dis- played by the spectrum of ZnL1 was clearly different from that of CuL1.IfCuL1 was reacted with 15 equivalents of Zn(OAc)2, no ZnL1 signal appeared even after one day (Figure 9). This implies an excellent stability of CuL1 in the presence of zinc. The same experiment was performed with Ca(OAc)2·H2O, Pb(OAc)4, Fe(OAc)2, Mg(NO3)2·6H2O, Mn(OAc)3·2H2O, and KNO3, and in every case, no transmetalation was observed (Figur- es S7-S12). Thus, we can conclude that the risk of release of metal in the organism is very low. These encouraging results prompted us to perform a preliminary complexation test using radioactive copper.

Radiolabeling Chelating a radionuclide is very demanding for a ligand. It in- volves a complexation in acidic buffer medium (to prevent the formation of metallic hydroxides), at room temperature (to be compatible with biological vectors), in a short time (short radi- onuclide half-life), and in a very diluted medium (high dilution of radionuclide source: 10^6 to 10^8 m).[21] We have described the excellent abilities of L1 to chelate radioactive copper (64Cu) in acid medium at room temperature, in 30 minutes at 10^9 m.[12] Thin-layer chromatography (TLC) associated with an appropriate eluent enabled the free form of the radionuclide to be separated from the complexed one, as the former does not migrate in the established conditions. Radioactivity on the TLC plates was quantified by using a phosphorimager. As the staining intensity is a function of the locally emitted radioactiv- ity, the rate of complexation can be measured. The experimen- tal results revealed a high chelation rate of 70 % when one equivalent of L1 was used, and more than 80 % when ten equivalents were used. As radioactive sources and buffered sol- utions contain unknown and varying amounts of nonradioac- tive isotopes that compete with the radionuclide, an excess of ligand was used to ensure the total complexation of the cat- ionic species. In comparison with currently used chelating agents, this appears to be a very competitive ligand.[22] These experiments also suggest that the complexation kinetics is very fast even in hostile conditions. These results encouraged us to synthesize the bifunctional analogues of ligand L1.

Synthesis of the bifunctional chelate We were able to synthesize a fine-tunable ligand through a two- or three-step microwave-assisted strategy (Scheme 2, and Schemes S1 and S2). The R1 group provides a binding site and the variation of R2,R3,andR4 enables the chelation prop- erties of the ligand (changing electronic and steric effects) to be modified.

We broadened the scope of our strategy[12] by using 2,6-diacetyl- pyridine instead of isophthalalde- hyde (L7) and by functionalizing the central aromatic ring (L8-L10 ; Table 2). However, our objective was the synthesis of an efficient BFC. This ligand needs a functional group to be bound to a vector. The one most often used is "NH2" because it can be involved in the formation of a peptide bond, or be transformed into isothiocyanate to form a thiourea bond.[23] We chose to introduce the functional group on the arene bearing the have the smallest influence on the chelation. Starting from 1a- c, obtained in a three- or four- step process (syntheses are pre- sented in Scheme S3), the li- gands L8-L10 were isolated in good yields by the microwave- assisted sequence described pre- viously (Scheme 3).

The different functionaliza- tions on position 5 of the aro- matic ring, especially when 1a was used, proved that steric hin- drance did not disturb the con- struction of the arms of the ligand. The second strategy, which provided 1band 1c, enabled the easy introduction of different groups or linkers on the chelating agent. The Boc group of L10 could easily be removed in TFA at room temperature in less than one hour to afford the free amino ligand L11 in 91 % yield. It is important to highlight the fact that the hydrazones are not at all acid-sensitive. This means that they will be stable in a biological medium, especially in the slightly acidic sur- roundings of tumors.

Preliminary complexation tests were performed on L11. This BFC displayed a similar behavior to L1: quantitative complexa- tion in diluted media, similar UV/Vis spectra for the ligand and the complex CuL11 (Figure S13), and a similar mass spectrum.

Thus, we have developed a versatile and efficient method of designing new chelating agents, which enables a linker to be easily constructed on the aromatic ring. The overall yields were good to excellent, even for compounds that required a multi- step synthesis (Table 2). Moreover, we succeeded in synthesiz- ing our target L11, a BFC, which could be bound to a vector through its amino group.

Conclusion By combining different analytical techniques, we have demon- strated the formation of N4 copper(I) complexes from a copper(- II) salt. This work was completed by a theoretical molecular modeling study, which confirmed the N4 complexes in which the copper seems to be complexed mainly by the two nitro- gen atoms of the thiazoles. The electrochemical study of the complex CuL1 displayed a quasi-reversible process with anodic and cathodic peaks at potentials of about 0.55 and 0.30 V versus SCE, respectively, thus indicating a stable CuII/CuI com- plex system. This latter point is important because in vivo re- duction of stable CuII complexes to form CuI products can lead to demetalation. Furthermore, this investigation has demon- strated a good kinetic inertness for CuL1 against competing metals and ligands (Zn, Fe, Ca, Mg, Pb, K, bathocuproine disul- fonate, DOTA, and EDTA). One of these ligands was functional- ized by an amino group to form a bifunctional chelate that could be bound to a biological vector to be used in nuclear medicine. Moreover, a preliminary study with radioactive copper has shown the great potential of these ligands, given their very rapid complexation kinetics. These new derivatives thus offer a viable and original route for copper binding for targeted imaging.

Experimental Section Materials and methods NMR spectra were recorded at room temperature with a Bruker Avance 300 Ultra Shield or eBruker Avance III 400 spectrometer. Chemical shifts are reported in parts per million (ppm) ; coupling constants are reported in Hertz (Hz). Infrared (IR) spectra were re- corded with a Bruker Vector 22 FTIR spectrometer using KBr films or KBr pellets. Low-resolution mass spectra were recorded with a Thermo Electron DSQ spectrometer. High-resolution mass spec- trometry (HRMS) was performed with a Bruker Autoflex III Smart- Beam spectrometer (MALDI). Microanalyses were performed on a Thermo Scientific FLASH 2000 Series CHNS/O Analyzer. Melting points were determined in open capillary tubes and are uncorrect- ed. EDX elemental analyses were measured on a JEOL 5800 LV in- strument with an EDX probe ; analyses are semiquantitative be- cause of the use of element standards. XPS spectra were recorded on a Leybold LHS11 MCP apparatus. SEM experiments were per- formed on a JEOL 6400F microscope. All reagents were purchased from Acros Organics or Aldrich and were used without further pu- rification. Column chromatography was conducted on silica gel Kieselgel SI60 (40-63 mm) from Merck. Reactions requiring anhy- drous conditions were performed under argon. Dichloromethane was distilled from calcium hydride under nitrogen prior to use. Tet- rahydrofuran was distilled from sodium/benzophenone under argon prior to use. Microwave experiments were conducted in sealed vials in commercial microwave reactors especially designed for synthetic chemistry (MultiSYNTH, Milestone). The instrument featured a special shaking system that ensured high homogeneity of the reaction mixtures. It was equipped with an indirect pressure control through precalibrated springs at the bottom of the vessel shields and with a fiber-optic contact thermometer (FO) for accu- rate temperature measurement. Cyclic voltammetry was performed using a three-electrode system. Potential was applied between a glassy carbon electrode and a Pt counter electrode, and a saturat- ed calomel electrode (SCE) was used as the reference. A solution of tetrabutylammonium hexafluorophosphate in anhydrous and de- oxygenated DMF was used as the supporting electrolyte.

Synthesis For the synthesis of 1b, 1c, and L2-L10, see the Supporting Infor- mation.

2,2'-[1,3-Phenylenebis(methan-1-yl-1-ylidene)]bis(hydrazin-1-yl-2-yli- dene)bis(thiazole-5,2-diyl)bis[(4-chlorophenyl)methanone] (L1): 2,2'- [1,3-Phenylenebis(methan-1-yl-1-ylidene)]bis{N-[(dimethylamino)- methylene]hydrazinecarbothioamide} (see the Supporting Informa- tion, compound 3) was dissolved in DMF (4 mL) in a 10 mL micro- wave reactor and the appropriate a-bromoketone (1.42 mmol, 2 equiv) was added. The solution turned orange. After 2 min of stir- ring, distilled triethylamine (2.13 mmol, 3 equiv) was added. The solution turned dark red. The reactor was heated from room tem- perature to 50 8C (FO) for 1 min by monomode microwave irradia- tion (power: 70 W, stirring: 30 %, ventilation: 1/3), maintained at 508C for 9min (power: 30 W, stirring: 50%, ventilation: 1/3), and allowed to cool to room temperature (power: 0 W, stirring: 50 %, ventilation: 3/3). The solution was concentrated under reduced pressure. The residue was suspended in dichloromethane (20 mL) and washed five times with water (30 mL). The organic layer was evaporated under reduced pressure. The residue was suspended in dichloromethane, isolated by filtration, and washed with dichloro- methane, which gave compound L1 as a pale yellow powder (70%). M.p. 205-2068C; 1H NMR (300 MHz, [D6]DMSO): d=7.56 (t, J=7.7Hz,1H; Har), 7.60(d, J=8.5Hz, 4H; Har),7.81 (m, 2H; Har), 7.83(d, J=8.5Hz, 4H; Har), 7.94 (s, 2H; Hthiaz), 7.97 (s,1H;Har), 8.24 (s, 2H; CH=N), 12.90ppm (bs,2H;NH);13CNMR (100MHz, [D6]DMSO): d= 125.6, 127.2, 127.6, 127.9, 128.6, 130.3, 134.4, 136.4, 136.8, 145.2, 150.1, 173.2, 184.4 ppm; IR (KBr): ñ=3451, 3187, 3059, 2761, 1614, 1588, 1568, 1507, 1485, 1438, 1397, 1323, 1278, 1257, 1200, 1174, 1100, 1090, 1013, 933, 906, 879, 842, 792, 750, 687, 628, 596 cm^1; UV/Vis (DMF): lmax (10^3 e)=395 (46), 505 nm (15 mol^1 m3 cm^1); MS (MALDI, DHB/CH3CN): m/z (%): 627 [M+Na] +, 605.1 [M+H] + ; HRMS (MALDI DHB/CH3CN): m/z calcd for C28H18Cl2N6O2S2 + Na+: 627.0207 [M+Na+]; found: 627.0234.

2,2'-[5-(2-Aminoethoxy)-1,3-phenylene]bis(methan-1-yl-1-ylidene)- bis(hydrazin-1-yl-2-ylidene)bis(thiazole-5,2-diyl)bis[(4-chlorophenyl)- methanone] (L11): Ligand L10 (0.2 mmol, 1 equiv) was dissolved in TFA (3 mL). The reaction mixture was stirred for 1 h at room tem- perature. Solvent was removed under reduced pressure. The resi- due was dissolved in dichloromethane (1 mL) then water (2 mL) was added. The solution was treated with 1 m sodium bicarbonate solution until effervescence stopped and the mixture was slightly basic. Dichloromethane (50 mL) was added and then the mixture was extracted with three 15 mL portions of dichloromethane. The organic layers were evaporated at reduced pressure to give com- pound L11 as an orange powder (91%). M.p. 247-2498C;1H NMR (300 MHz, [D6]DMSO): d = 3.29 (m, 2H; CH2), 4.27 (m, 2H; CH2), 7.39 (s,2H;Har),7.43(s,1H;Har), 7.55(d,J=8.4Hz,4H; Har),7.73(d, J=8.4 Hz, 4H; Har), 7.80 (s, 2H; Hthiaz), 8.16 ppm (s, 2H; CHN); 13C NMR (100 MHz, [D6]DMSO): d = 38.5, 65.2, 111.6, 120.2, 124.5, 128.5, 130.0, 135.7, 137.0, 137.7, 144.6, 152.3, 158.5, 177.7, 184.3 ppm; IR (KBr): ñ=3182, 3069, 2931, 2759, 1679, 1615, 1568, 1504, 1432, 1326, 1279, 1256, 1201, 1173, 1109, 1090, 1012, 878, 840, 800, 749, 680, 471 cm^1; UV/Vis (DMF): lmax (10^3 e) = 394 (22), 509 nm (15 mol^1m3 cm^1); MS (MALDI, DHB): m/z (%): 686.1 [M+Na] +, 666.1. [M+H] +; HRMS (MALDI DHB/CH3CN): m/z calcd for C33H23Cl2N7O3S2 + Na+: 686.0579 [M+Na]+; found: 686.0563.

General procedure for complexes Ligand Lx (0.05 mmol, 1 equiv) and monohydrate Cu(OAc)2 (0.05 mmol, 1 equiv) were dissolved in DMF (4 mL) in a 10 mL mi- crowave reactor, which was placed in the microwave oven. The re- actor was heated from room temperature to 120 8C (FO) for 2 min by monomode microwave irradiation (power: 150 W, stirring: 50 %, ventilation: 1/3), maintained at 1208C for 15min (power: 50W, stir- ring: 50 %, ventilation: 1/3), and allowed to cool to room tempera- ture (power: 0 W, stirring: 50 %, ventilation: 3/3). The solution was concentrated under reduced pressure. The residue was suspended in dichloromethane (10 mL) and washed five times with 15 mL of water. The organic layer was evaporated under reduced pressure. The residue was suspended in dichloromethane, isolated by filtra- tion, and washed with dichloromethane.

[5-(4-Chlorobenzoyl)-2-{[3-({2-[5-(4-chlorobenzoyl)thiazol-2-yl]hydra- zono}methyl)benzylidene]hydrazono}thiazol-3(2H)-yl]copper (CuL1): Brown powder; yield: 95%; IR (KBr): ñ=3446, 3062, 1588, 1526, 1496, 1477, 1398, 1348, 1288, 1255, 1199, 1174, 1088, 1013, 958, 918, 888, 840, 748, 686, 628, 600 cm^1; UV/Vis (DMF): lmax (10^3 e) = 440 nm (33 mol^1m3 cm^1); EDX: Cu/Cl and Cu/S ratios were 1:1; MS (MALDI, DHB/CH3CN) m/z (%): 667 [M+H] +, 605 [M^Cu+H] for C28H17Cl2N6O2S2Cu.

2,2'-[5-(2-Aminoethoxy)-1,3-phenylene]bis(methan-1-yl-1-ylidene)- bis(hydrazin-1-yl-2-ylidene)bis(thiazole-5,2-diyl)bis[(4-chlorophenyl)- methanone]copper (CuL11): Brown powder; yield: 99 %; IR (KBr): ñ= 3407, 3123, 1659, 11606, 1495, 1472, 1398, 1348, 1283, 1255, 1173, 1133, 1088, 1013, 958, 918, 874, 840, 745, 685 cm^1; UV/Vis (DMF):lmax(10^3e)=440nm (36mol^1m3cm^1);MS(MALDI,DHB/ CH3CN) m/z (%): 727 [M+H] +.

Radiolabeling Copper-64 dichloride in 0.1 m hydrochloric acid was obtained from ARRONAX cyclotron (Saint-Herblain, France). Radionuclide purity was determined by gamma spectroscopy using a DSPEC-JR-2.0 type 98-24B HPGE detector (AMETEK) and chemical purity was con- trolled by inductively coupled plasma-optical emission spectrosco- py with an iCAP 6500 DUO instrument (Thermo Fisher Scientific). TLC analyses were performed using silica gel on TLC-PET foils (Fluka Analytical). TLC plates were revealed on a storage phosphor screen by using a Cyclone Plus phosphor imager (PerkinElmer). The ligand was radiolabeled with 64Cu by addition of 64CuCl2 solution (40 to 60 MBq ; metal composition : 10 ppm of copper for 60 ppm total metals) in 2.5 m ammonium acetate (pH 5.5) to chelating agent L1 (1 or 10 equiv) in DMF. The solution was stirred at room temperature for 30 min and the radiolabeled 64CuL1 was obtained in 70 % (for 1 equiv) or 80 % (for 10 equiv) radiochemical yield. The radiochemical purity was >95 % as measured by TLC (the eluent used was a mixture of acetic acid and ethyl acetate (1:1)). The spe- cific activity was 49 kBq nmol^1.

Computational details All geometries reported herein are available in the Supporting In- formation. DFT was used to obtain the structure and energy of all the compounds under study under symmetry constraints, if appli- cable, to reduce computational time. We selected the widely used B3LYP density functional for our calculations.[24] These were per- formed using the simplified model compounds (LH2), in which the 4-chlorobenzaldehyde groups on the thiazole rings were replaced by hydrogen atoms to render the calculations computationally tractable. We used all-electron standard 6-311G basis sets for all the atoms, except copper. We used the pseudopotential and SDD basis sets, which are adapted for Cu and greatly reduce the com- putation time. Compounds with or without the Cu + cation were represented by a closed-shell singlet electronic ground state. The vibrational frequencies were calculated at the same level of theory to evaluate the zero-point energy (ZPE) correction and to deter- mine whether the structures found corresponded to true minima of the potential energy surface (PES) or to transition states (TSs). Unscaled values were used for zero-point vibrational energy (ZPVE) corrections for energy values. Thermal corrections were calculated for the evaluation of reaction enthalpies (DH) and Gibbs energies (DG) at 1.0 atm and 298.15 K, using standard statistical mechanics formulas in the independent mode, and harmonic oscillator. The copper stabilization energies were determined according to Equa- tion (1): (ProQuest: ... denotes formula omitted.) in which Ecomplex represents the total ZPVE corrected energy of the whole complex, Ei represents the ZPVE corrected energy of an indi- vidual subsynstem, BSSE represents the basis set superposition error, and " Cuþ = 1Cu+" represents the number of CuI ions in the complex.i Acknowledgements We are grateful to the French Ministry of Education and to the Centre National de la Recherche Scientifique (CNRS) for financial support. This research used the CPU resources of the CCIPL (Centre de Calcul Intensif des Pays de Loire). This study was also supported by a grant from the French National Agency for Re- search called "Investissements d'Avenir", Equipex ArronaxPlus no. ANR-11-EQPX-0004 and from the "Region Pays de la Loire" (NUCSAN project).

[1] a) X. Chen, S. S. Gambhir, J. Cheon, Acc. Chem. Res. 2011, 44, 841 - 841; b) J. A. MacKay, Z. Li, Adv. Drug Delivery Rev. 2010, 62, 1003 - 1004 ; c) C. S. Cutler, H. M. Hennkens, N. Sisay, S. Huclier-Markai, S. S. Jurisson, Chem. Rev. 2013, 113, 858 - 883.

[2] a) M. Li, C. F. Meares, Bioconjugate Chem. 1993, 4, 275 - 283 ; b) C. Glauss, R. Rossin, M. J. Welch, G. Bao, Bioconjugate Chem. 2010, 21, 715 - 722 ; c) J. N. Bryan, F. Jia, H. Mohsin, G. Sivaguru, W. H. Miller, C. J. Anderson, C. J. Henry, M. R. Lewis, Nucl. Med. Biol. 2005, 32, 851 - 858.

[3] a) M. Shokeen, C. J. Anderson, Acc. Chem. Res. 2009, 42, 832 - 841; b) C. J. Anderson, R. Ferdani, Cancer Biother. Radiopharm. 2009, 24, 379 - 393 ; c) H. Williams, S. Robinson, P. Julyan, J. Zweit, D. Hastings, Eur. J. Nuclear. Med. Mol. Imaging. 2005, 32 , 1473 - 1480.

[4] a) S. Liu, J. Pietryka, C. E. Ellars, D. Edwards, Bioconjugate Chem. 2002, 13, 902 - 913 ; b) H. Mohsin, J. Fitzsimmons, T. Shelton, T. J. Hoffman, C. S. Cutler, M. R. Lewis, P. S. Athey, G. Gulyas, G. E. Kiefer, R. K. Frank, J. Simon, S. Z. Lever, S. S. Jurisson, Nucl. Med. Biol. 2007, 34, 493 - 502 ; c) A. Carrasquillo, J. D. White, C. H. Paik, A. Raubitschek, N. Le, M. Rotman, M. W. Brechbiel, O. A. Gansow, L. E. Top, P. Perentesis, J. C. Rey- nolds, D. L. Nelson, T. A. Waldmann, J. Nucl. Med. 1999, 40, 268 - 276.

[5] a) E. Boros, J. F. Cawthray, C. L. Ferreira, B. O. Patrick, M. J. Adam, C. Orvig, Inorg. Chem. 2012, 51, 6279 - 6284 ; b) T. J. Wadas, E. H. Wong, G. R. Weisman, C.J. Anderson, Chem. Rev. 2010, 110, 2858-2902; c)R. E. Mewis, S. J. Archibald, Coord. Chem. Rev. 2010, 254, 1686 - 1712 ; d) D. Sarko, M. Eisenhut, U. Haberkorn, W. Mier, Curr. Med. Chem. 2012, 19, 2667 - 2688 ; e) C. Santini, M. Pellei, V. Gandin, M. Porchia, F. Tisato, C. Marzano, Chem. Rev. 2014, 114, 815 - 862.

[6] a) M. S. Cooper, M. T. Ma, K. Sunassee, K. P. Shaw, J. D. Williams, R. L. Paul, P. S. Donnelly, P. J. Blower, Bioconjugate Chem. 2012, 23, 1029 - 1039 ; b) J. L. J. Dearling, S. D. Voss, P. Dunning, E. Snay, F. Fahey, S. V. Smith, J. S. Huston, C. F. Meares, S. T. Treves, A. B. Packard, Nucl. Med. Biol. 2011, 38, 29 - 38 ; c) Z. Cai, C. J. Anderson, J. Labelled Compd. Radio- pharm. 2014, 57, 224 - 230.

[7] a) C. A. Boswell, X. Sun, W. Niu, G. R. Weisman, E. H. Wong, A. L. Rhein- gold, C. J. Anderson, J. Med. Chem. 2004, 47, 1465 - 1474 ; b) D. N. Pandya, J. Y. Kim, J. C. Park, H. Lee, P. B. Phapale, W. Kwak, T. H. Choi, G. J. Cheon, Y.-R. Yoon, J. Yoo, Chem. Commun. 2010, 46, 3517 - 3519.

[8] a) D. Zeng, Q. Ouyang, Z. Cai, X.-Q. Xie, C. J. Anderson, Chem. Commun. 2014, 50, 43 - 45 ; b) S. R. Banerjee, M. Pullambhatla, C. A. Foss, S. Nim- magadda, R. Ferdani, C. J. Anderson, R. C. Mease, M. G. Pomper, J. Med. Chem. 2014, 57, 2657 - 2669.

[9] a) C. J. Anderson, T. J. Wadas, E. H. Wong, G. R. Weisman, Q. J. Nucl. Med. Mol. Imaging 2008, 52, 185 - 192 ; b) L. Kong, E. Mume, G. Triani, S. V. Smith, Langmuir 2013, 29, 5609 - 5616 ; c) B. M. Paterson, P. Roselt, D. Denoyer, C. Cullinane, D. Binns, W. Noonan, C. M. Jeffery, R. I. Price, J. M. White, R. J. Hicks, P. S. Donnelly, Dalton Trans. 2014, 43, 1386 - 1396 ; d) H. Cai, Z. Li, C.-W. Huang, A. H. Shahinian, H. Wang, R. Park, P. S. Conti, Bioconjugate Chem. 2010 , 21, 1417 - 1424.

[10] a) B. M. Paterson, J. A. Karas, D. B. Scanlon, J. M. White, P. S. Donnelly, Inorg. Chem. 2010,49, 1884-1893; b)R. Hueting,M.Christlieb,J.R. Dil- worth, E. G. Garayoa, V. Gouverneur, M. W. Jones, V. Maes, R. Schibli, X. Sun, D. A. Tourwe, Dalton Trans. 2010, 39, 3620 - 3632 ; c) K. A. Wood, W.L.Wong,M.I. Saunders, Nucl.Med.Biol. 2008, 35,393-400;d)L. Car- roll, R. Bejot, R. Hueting, R. King, P. Bonnitcha, S. Bayly, M. Christlieb, J. R. Dilworth, A. D. Gee, J. Declerck, V. Gouverneur, Chem. Commun. 2010, 46, 4052 - 4054 ; e) G. Buncic, J. L. Hickey, C. Schieber, J. M. White, P. J. Crouch, A. R. White, Z. Xiao, A. G. Wedd, P. S. Donnelly, Aust. J. Chem. 2011, 64, 244 - 252 ; f) J. R. Dilworth, R. Hueting, Inorg. Chim. Acta 2012, 389, 3 - 15.

[11] T. J. Hoffman, C. J. Smith, Nucl. Med. Biol. 2009, 36, 579 - 585.

[12] E. Bodio, K. Julienne, S. G. Gouin, A. Faivre-Chauvet, D. Deniaud, Bioorg. Med. Chem. Lett. 2011, 21, 924 -927.

[13] a) S. Dian-Hong, S. Tie-Han, T. Nagasawa, K. Murakami, Chin. Phys. B 2002, 11, 393 - 398; b) J. L. Colaux, P. Louette, G. Terwagne, Nucl. Ins- trum. Methods Phys. Res. Sect. B 2009, 267, 1299 - 1302 ; c) T. Storr, P. Verma, R. C. Pratt, E. C. Wasinger, Y. Shimazaki, T. D. P. Stack, J. Am. Chem. Soc. 2008, 130, 15448 -15459.

[14] B. F. Minaev, S. V. Bondarchuk, Russ. J. Appl. Chem. 2009, 82, 840 - 845.

[15] a)J. K. Kochi, J. Am. Chem. Soc. 1955, 77, 5090-5092; b)J. K. Kochi, J. Am. Chem. Soc. 1962, 84, 2121 - 2127.

[16] a) Y. Hattori, M. Nishikawa, T. Kusamoto, S. Kume, H. Nishihara, Inorg. Chem. 2014, 53, 2831 - 2840 ; b) M. R. Malachowski, M. Adams, N. Elia, A. L. Rheingold, R. S. Kelly, J. Chem. Soc. Dalton Trans. 1999, 2177 - 2182 ; c) P. R. Chetana, B. S. Srinatha, M. N. Somashekar, R. S. Policegoudra, R. Rao, B. Maity, S. M. Aradhya, Int. J. Pharm. Sci. Rev. Res. 2013, 21, 355 - 363; d)M. Osawa, Chem. Commun. 2014, 50, 1801-1803; e) D. F. Zigler, E. Tordin, G. Wu, A. Iretskii, E. Cariati, P. C. Ford, Inorg. Chim. Acta 2011, 374, 261 -268.

[17] Q.-M. Qiu, M. Liu, Z.-F. Li, Q.-H. Jin, X. Huang, Z.-W. Zhang, C.-L. Zhang, Q.-X. Meng, J. Mol. Struct. 2014, 1062, 125 -132.

[18] a) S. Deng, R. Bai, J. P. Chen, Langmuir 2003, 19, 5058 - 5064 ; b) S. Deng, R. Bai, J. P. Chen, J. Colloid Interface Sci. 2003, 260, 265 - 272.

[19] T. S. Lobana, S. Indoria, M. Sharma, J. Nandi, A. K. Jassal, M. S. Hundal, A. Castineiras, Polyhedron 2014, DOI : 10.1016/j.poly.2014.01.002.

[20] A. Kaufman Katz, L. Shimoni-Livny, O. Navon, N. Navon, C. W. Bock, J. P. Glusker, Helv. Chim. Acta 2003, 86, 1320 -1338.

[21] a) S. G. Gouin, E. Benoist, J.-F. Gestin, J. C. Meslin, D. Deniaud, Eur. J. Org. Chem. 2004, 878 - 885 ; b) S. G. Gouin, J.-F. Gestin, L. Monrandeau, F. Segat-Dioury, J. C. Meslin, D. Deniaud, Org. Biomol. Chem. 2005, 3, 454 - 461.

[22] S. V. Smith, J. Inorg. Biochem. 2004, 98, 1874 - 1901.

[23] a) L. Morandeau, P. R.-L. Saec, A. Ouadi, K. Bultel-Rivi^re, M. Mougin-De- graef, A. de France-Robert, A. Faivre-Chauvet, J.-F. Gestin, J. Labelled Compd. Radiopharm. 2006, 49, 109 - 123 ; b) C. F. Meares, M. J. McCall, D. T. Reardan, D. A. Goodwin, C. I. Diamanti, M. McTigue, Anal. Biochem. 1984, 142, 68 - 78.

[24] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648 - 5652 ; b) P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys. Chem. 1994, 98, 11623 - 11627.

Received : February 24, 2014 Revised : May 22, 2014 Published online on June 24, 2014 Ewen Bodio,[a, b] Mohammed Boujtita,[a] Karine Julienne,[a] Patricia Le Saec,[c] S^bastien G. Gouin,[a] Jonathan Hamon,[d] Eric Renault,[a] and David Deniaud*[a] [a] Dr. E. Bodio, Dr. M. Boujtita, Dr. K. Julienne, Dr. S. G. Gouin, Dr. E. Renault, Dr. D. Deniaud LUNAM Universit^, CEISAM Chimie et Interdisciplinarit^, Synth^se, Analyse, Mod^lisation UMR CNRS 6230, UFR des Sciences et des Techniques 2 rue de la Houssini^re, BP 92208, 44322 Nantes Cedex 3 (France) E-mail : [email protected] [b] Dr. E. Bodio Institut de Chimie Mol^culaire de l'Universit^ de Bourgogne UMR CNRS 6302, Universit^ de Bourgogne 9 avenue A. Savary, BP 47870, 21078 Dijon (France) [c] P. Le Saec Centre de Recherche en Canc^rologie de Nantes-Angers, UMR CNRS 6299 INSERM U892, Institut de Recherche Th^rapeutique, Universit^ de Nantes 9 quai Moncousu, 44093 Nantes Cedex 1 (France) [d] J. Hamon IMN, Institut des Mat^riaux Jean Rouxel, UMR CNRS 6502 UFR des Sciences et des Techniques 2 rue de la Houssini^re, BP 92208, 44322 Nantes Cedex 3 (France) Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cplu.201402031.

(c) 2014 Blackwell Publishing Ltd.

[ Back To TMCnet.com's Homepage ]