Oxaliplatin Binding to Human Copper Chaperone Atox1 and Protein

Jun 15, 2016 - Synopsis. The interaction between the copper chaperone Atox1 and oxaliplatin, a platinum-based anti-neoplastic agent, shows little ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/IC

Oxaliplatin Binding to Human Copper Chaperone Atox1 and Protein Dimerization Benny D. Belviso,†,⊥ Angela Galliani,‡,∥ Alessia Lasorsa,‡ Valentina Mirabelli,†,§ Rocco Caliandro,*,† Fabio Arnesano,*,‡ and Giovanni Natile*,‡ †

Institute of Crystallography, Consiglio Nazionale delle Ricerche, via Amendola 122/o, 70126 Bari, Italy Department of Chemistry, University of Bari “A. Moro”, via E. Orabona 4, 70125 Bari, Italy § Department of Economics, University of Foggia, Via A. Gramsci 89/91, 71122 Foggia, Italy ‡

S Supporting Information *

ABSTRACT: Copper trafficking proteins have been implicated in the cellular response to platinum anticancer drugs. We investigated the reaction of the chaperone Atox1 with an activated form of oxaliplatin, the third platinum drug to reach worldwide approval. Unlike cisplatin, which contains monodentate ammines, oxaliplatin contains chelated 1,2-diaminocyclohexane (DACH), which is more resistant to displacement by nucleophiles. In solution, one or two {Pt(DACH)2+} moieties bind to the conserved CXXC metal-binding motif of Atox1; in the latter case the two sulfur atoms likely bridging the two platinum units. At longer reaction times, a dimeric species is formed whose composition, Atox12·Pt2+2, indicates complete loss of the diamine ligands. Such a dimerization process is accompanied by partial unfolding of the protein. Crystallization experiments aiming at the characterization of the monomeric species have afforded, instead, a dimeric species resembling that already obtained by Boal and Rosenzweig in a similar reaction performed with cisplatin. However, while in the latter case there was only one Pt-binding site (0.4 occupancy) made of four sulfur atoms of the CXXC motifs of the two Atox1 chains in a tetrahedral arrangement, we found, in addition, a secondary Pt-binding site involving Cys41 of the B chain (0.25 occupancy). Moreover, both platinum atoms have lost their diamines. Thus, there appears to be little relationship between what is observed in solution and what is formed in the solid state. Since full occupancy of the tetrahedral cavity is a common feature of all Atox1 dimeric structures obtained with other metal ions (Cu+, Cd2+, and Hg2+), we propose that in the case of platinum, where the occupancy is only 0.4, the remaining cavities are occupied by Cu+ ions. Experimental evidence is reported in support of the latter hypothesis. Our proposal represents a meeting point between the initial proposal of Boal and Rosenzweig (0.4 Pt occupancy) and the reinterpretation of the original crystallographic data put forward by Shabalin et al. (1 Cu occupancy), and could apply to other cases.



and α1 and is highly conserved among metallochaperones and soluble domains of Cu ATPases.13−15 A lysine-rich region, KKTGK, located in the loop between α2 and β4, represents the nuclear localization signal for the translocation of Atox1 to the nucleus.16 In a previous investigation we applied a combined approach, using solution and in-cell NMR spectroscopy methods, to probe the intracellular interaction of cisplatin with Atox1.17 The intracellular environment provides suitable conditions for the preservation of the protein in its active form, while cell-free systems usually require an exogenous reducing agent, which, however, has itself good coordinating ability toward Pt and thus can strongly compete with the protein for formation of the Pt adducts. Molecules of reducing agent were found in one of the

INTRODUCTION

Among anticancer therapeutics, Pt-based drugs have a prominent role. They perform their antitumor activity by forming stable adducts with DNA, thus interfering with replication and transcription processes.1,2 Cellular uptake of these drugs appears to be tightly connected to copper transport.3,4 The major Cu influx transporter Ctr1 has been proposed, although with some controversy,5 to mediate transport of cisplatin and its analogues.6 Evidence also suggests that Cu ATPases ATP7A and ATP7B mediate cisplatin sequestration and/or efflux from cells, thus influencing drug resistance.7−9 Finally, the cytosolic Cu chaperone Atox1, which delivers the metal to ATP7A/B,10 has also been reported to be implicated in cisplatin resistance.11,12 Antioxidant-1 (Atox1; 68 amino acids) has a β1α1β2β3α2β4 ferredoxin-like structure. The metal-binding site, CXXC, is located in the loop between secondary structure elements β1 © XXXX American Chemical Society

Received: March 25, 2016

A

DOI: 10.1021/acs.inorgchem.6b00750 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

and centrifuged for 20 min at 4 °C. The cellular pellet was suspended in 20 mM 2-(N-morpholino)ethanesulfonic acid (MES; pH 5.5), 1 mM ethylenediaminetetraacetic acid (EDTA), 5 mM DTT, and 1 mM phenylmethylsulfonyl chloride (proteases inhibitor), and it was lysed by sonication. The cellular lysate was centrifuged at 20 000 rpm at 4 °C for 1 h. The supernatant, containing the recombinant protein, was loaded onto a HiTrap Sepharose Fast Flow (SP FF) cation exchange column (GE Healthcare) pre-equilibrated with 20 mM MES, 5 mM DTT, and 5 mM EDTA (pH 5.5) and eluted with a column volume gradient of 20 mM MES (pH 5.5), 5 mM DTT, 5 mM EDTA, and 500 mM NaCl. Atox1-containing fractions, eluted at ∼200 mM NaCl, were concentrated and further purified on a Superdex 75 column (GE Healthcare) equilibrated with 20 mM MES, 5 mM DTT, 5 mM EDTA, and 150 mM NaCl (pH 6.0). Fractions containing pure Atox1 were pooled and washed with 20 mM MES and 5 mM DTT (pH 6.0), by using Amicon Ultra centrifugal filters with 3 kDa cutoff (Millipore, U.S.), to remove NaCl and EDTA. The protein sample was concentrated to 200 μM (protein concentration was determined by UV−visible absorbance at 280 nm) and stored at −20 °C. Protein purity was determined by gel electrophoresis and ESI-MS, while inductively coupled plasma mass spectrometry (ICP-MS) analysis was performed to determine the amount of residual Cu. Atox1 Sample Preparation. All purification steps of Atox1 were performed in the presence of an excess of DTT as reducing agent to preserve the protein in its active form by keeping reduced its cysteine residues. However, DTT is able to bind Pt drugs with its sulfur and oxygen atoms, forming stable adducts and competing with the protein for metal binding. To overcome the latter problem, after reduction of the protein with DTT, the reducing agent was removed under strictly anaerobic conditions. Hence, Atox1 was washed several times with the chosen deoxygenated buffer inside a nitrogen-filled glovebox (ItecoEng, Italy) by using Amicon Ultra centrifugal filters (3 kDa cutoff) and, after the last wash, it was concentrated and immediately used for the incubations. Preparation of Pt Solutions. The activated form of oxaliplatin, [Pt(R,R-1,2-DACH)(H2O)(SO4)], and 15N-labeled cisplatin (cis[PtCl2(15NH3)2]) were prepared by reported procedures.22 The compounds were dissolved immediately prior to use in ultrapure deoxygenated water at 2 and 15 mM final concentrations, respectively. [Pt(R,R-1,2-DACH)(H2O)(SO4)], used for the crystallization experiment, was dissolved in pure deoxygenated MES solution (20 mM, pH 6.0) at 15 mM final concentration. The Pt complex solutions were extensively vortexed and sonicated, and the precise metal concentration was determined using a Varian 880Z atomic absorption spectrometer. The manipulation of all samples was performed in a glovebox under N2 atmosphere. Incubations of the activated form of oxaliplatin and cisplatin with the protein were conducted at 25 °C, under strictly analogous conditions. One experiment was performed starting from oxaliplatin, [Pt(R,R-1,2-DACH)(oxalate)]: after 24 h of incubation with the protein, 10-fold excess glutathione (GSH) was added, and the reaction was monitored by ESI-MS. Size-Exclusion Chromatography. Atox1 (230 μM) was incubated with equimolar Pt complex in phosphate buffer (25 mM, pH 7.0) at 25 °C. The reaction progress was monitored by SEC analysis of samples (150 μL) taken from time to time. Gel filtration was performed with a Superdex 75 10/300 column (GE Healthcare, column volume 24 mL) on an AKTA purifier (GE Healthcare). The run was performed by using 20 mM ammonium acetate (pH 6.5) as running buffer, with a flow rate of 0.4 mL min−1, and monitoring the protein elution by measuring absorbance at 280 nm. The protein fractions corresponding to peaks of monomeric and dimeric Atox1 were individually collected, concentrated with Amicon Ultra centrifugal filters (3 kDa cutoff), and immediately analyzed by mass spectrometry and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Electrospray Mass Spectrometry. The concentrated elution fractions (containing either monomeric or dimeric Atox1) were twofold diluted with pure water, treated with 1% (v/v) acetic acid, and analyzed with a dual electrospray interface and a 6530 Series AccurateMass Quadruple Time-of-Flight LC/MS (Agilent, U.S). Ionization was

two crystalline forms of Atox1−cisplatin derivatives reported by Boal and Rosenzweig.18 In such a form (PDB code 3IWL)18 the protein was monomeric, and the metal-binding motif, which consists of two cysteine residues (Cys12 and Cys15), was tricoordinated to the platinum ion through the N and S atoms of Cys12 and the S atom of Cys15, with the two sulfur atoms in trans positions. Conceivably, all donor groups had lost a proton. The fourth coordination position of Pt was occupied by a molecule of tri(2-carboxyethyl)phosphine (TCEP) used as exogenous reducing agent in the preparation of the sample. In the second crystalline form (PDB code 3IWX),18 the protein was dimeric (chains A and B), with the metal-binding motifs of the two chains close to one another and forming an almost regular tetrahedron with average S···S distances of 3.8 ± 0.4 Å. The electron density inside the tetrahedral cavity was assigned to a {Pt(NH3)22+} moiety, which completes the square-planar coordination shell of Pt by coordinating Cys15 S atoms of the two protein molecules. The Pt···S(Cys12) nonbonding distances (2.47 ± 0.01 Å) were only slightly longer than the bonding Pt−S (Cys15) distances (2.2 ± 0.1 Å). Moreover, there were short nonbonding distances between the ammine groups and the Cys12 sulfur atoms (N···S distances of 2.3−2.6 Å). The dimeric arrangement of protein molecules was very similar to that described for several other Atox1 adducts with a variety of metal ions (Cu+, Cd2+, Hg2+),19 with a tetrahedral arrangement of the four sulfur atoms (average S···S distance of 4.0 ± 0.8 Å) and only a small dependence of the size of the tetrahedron upon the nature of the metal ion. In a recent report by Shabalin et al.20 it was pointed out that the crystallographic data of the Atox1 dimer by Boal and Rosenzweig (PDB code 3IWX) could also be interpreted on the basis of full occupancy of the tetrahedral cavity by Cu+ instead of a {Pt(NH3)22+} occupancy of only 0.4. In the present investigation, we monitored by size-exclusion chromatography (SEC) coupled with electrospray ionization mass spectrometry (ESI-MS) the 1:1 reaction between Atox1 and an activated form of oxaliplatin, the third-generation Pt drug in clinical use, under strictly anaerobic conditions and in the absence of any exogenous reducing agent (such as dithiothreitol (DTT) or TCEP), all of which have good coordinating ability toward Pt and can thus interfere with the formation of Pt−protein adducts. Contrary to cisplatin, oxaliplatin proved to be resistant to amine-ligand displacement by a methionine-rich peptide.21 Solution experiments showed that initially monomeric Atox1 adducts containing one or two {Pt(DACH)2+} (DACH = 1,2-diaminocyclohexane) units are formed, while at longer reaction time (one week) formation of Atox12·Pt2+2 species, in which the Pt cores have lost the DACH ligands and the proteins are partially unfolded, takes place. Crystallization experiments aiming to the characterization of the Atox1 monomer(s) have instead afforded a dimeric species (with perfectly folded protein molecules) resembling that already reported by Boal and Rosenzweig but with a secondary Pt-binding site involving Cys41 of one chain. In both sites there was no trace of DACH ligand.



EXPERIMENTAL SECTION

Atox1 Expression and Purification. BL21(DE3)Gold Escherichia coli (Stratagene, U.S.) containing pET21a with the gene encoding Atox1 were grown at 37 °C to an optical density of 0.5 at 600 nm, and then protein expression was induced with 1 mM isopropyl β-D-1thiogalactopyranoside. After 4 h of induction, the cells were harvested B

DOI: 10.1021/acs.inorgchem.6b00750 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry achieved in the positive ion mode by application of 3.5 kV at the entrance of the capillary; the pressure of the nebulizer gas was 35 psi. The drying gas was heated to 300 °C at a flow rate of 8 L/min. Fullscan mass spectra were recorded in the mass/charge (m/z) range of 100−3000. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis. For SDS-PAGE, the concentrated elution fractions were treated with DTT (1 mM) to remove possible Atox1 dimers sustained by intermolecular disulfide bridges. Then the samples were mixed with Laemmli loading buffer 1X (∼5 μg per lane) and loaded on 4−20% polyacrylamide gel (TGX precast gel, Biorad) using TGS-1X as running buffer. Electrophoresis was performed for 45 min at 120 V. Samples were not boiled, so as to preserve the structure of metalbridged dimeric structures of Atox1. After electrophoresis, the gel was silver stained. Circular Dichroism Spectroscopy. Circular dichroism (CD) experiments were performed on a Jasco J-810 spectropolarimeter (Jasco Inc., Easton, MD, U.S.) at 25 °C. Quartz cuvettes with path length of 0.1 cm were used in the far-UV region from 190 to 250 nm. Each spectrum, obtained by averaging of five scans, was baselinecorrected and then smoothed by applying adjacent averaging or a fast Fourier transform filter. The ellipticity is reported as mean residue molar ellipticity (deg cm2 dmol−1) according to [θ] = 100 [θ]obs/(C L N), where [θ]obs is the observed ellipticity in degrees, C is the molar concentration of the protein, L is the optical path length (in cm), and N is the number of amino acids (N = 68 for Atox1). Spectra were collected using monomeric and dimeric Atox1 samples obtained upon incubation of the protein with oxaliplatin at 1:1 molar ratio in phosphate buffer (25 mM, pH 7.0) at 25 °C, fractioned by gel filtration on Superdex 75 and eluted in ammonium acetate 20 mM (pH 6.7). The concentration of both samples was 10 μM. Nuclear Magnetic Resonance Spectroscopy. 15N-labeled Atox1 (230 μM) was incubated with equimolar [Pt(R,R-1,2-DACH)(H2O)(SO4)] or 15N-labeled cisplatin (cis-[Pt(15NH3)2Cl2]) in phosphate buffer (25 mM, pH 7.0) at 25 °C, and after 48 h of incubation, the reaction was analyzed by SEC following the procedure previously described. The protein fractions corresponding to peaks of Atox1 monomer and dimer were collected, concentrated with Amicon Ultra centrifugal filters (3 kDa cutoff), and analyzed by 1H and 15N HSQC NMR experiments. The final protein concentration for both monomeric and dimeric Atox1 samples was 100 μM. NMR samples also contained 10% v/v D2O for NMR spectrometer lock, and all NMR experiments were performed at 298 K. Resonance assignment of apoAtox1 was performed by using the available 1H and 15N chemical shifts.23 All spectra were collected on a 700 MHz Bruker AVANCE III HD NMR spectrometer equipped with a proton-optimized quadruple resonance “inverse” (QCI) CryoProbe and with pulsed-field gradients along the z-axis, processed using the standard Bruker software (TopSpin), and analyzed with the program CARA, developed at ETHZürich (www.cara.nmr.ch). Crystallization and Structure Determination of the Dimeric Atox1−Platinum Adduct. DTT-free samples of Atox1 (prepared as described above) were concentrated to 5.5 mg mL−1 and treated with an equimolar amount of [Pt(R,R-1,2-DACH)(H2O)(SO4)], previously dissolved in MES 20 mM, pH 6.0. The reaction solution was incubated at 25 °C in nitrogen-filled glovebox. After 24 h, an aliquot of the reaction solution was desalted by Zip-TipC18 pipet tips (Millipore, U.S.) and analyzed by ESI-MS. The spectrum showed the binding of both one or two {Pt(R,R-DACH)2+} residues to Atox1 (the Pt ion retains its carrier ligand). Once verified by ESI-MS that the reaction was complete, the incubation solution was further concentrated to 18 mg mL−1 by using Amicon Ultra centrifugal filters and used for the crystallization experiments. Crystallization experiments were performed at 20 °C by using the hanging drop technique: 1 μL of protein solution (concentration ranging from 12 to 20 mg/mL) mixed with 1 μL of the reservoir solution made of Li2SO4 (concentration ranging from 60 to 96% that of saturation), NaCl 60 mM, and MES 100 mM at pH 6.5. Small lanceolate crystals (dimensions