Intracellular metal binding and redox behavior of human DJ-1

  • Letizia Barbieri
  • Enrico Luchinat
  • Lucia Banci
Original Paper
Part of the following topical collections:
  1. Special issue to celebrate the 80th birthday of Helmut Sigel


DJ-1 is a conserved, ubiquitous protein associated to a large number of intracellular processes. Human DJ-1 has been linked to several pathologies, including hereditary forms of Parkinson’s disease, cancer, and amyotrophic lateral sclerosis. Several cytoprotective functions of DJ-1 have been reported, however, its actual mechanisms of action remain elusive. In vitro, DJ-1 has been shown to bind zinc and copper(II) at its active site, which contains a conserved cysteine (C106), and copper(I) at a different binding site. C106 is essential to DJ-1 function, and is easily oxidized upon oxidative stress. Here, we investigated the metal-binding- and redox properties of DJ-1 in living human cells by in-cell NMR. Intracellular DJ-1 is surprisingly free from interactions with any other cellular components and as such is clearly detectable by NMR. Metal-bound forms of DJ-1 were not observed upon treating the cells with excess zinc or copper. No copper binding was observed when co-expressing DJ-1 with the copper chaperone for superoxide dismutase 1 (SOD1). Co-expression of DJ-1 with SOD1 itself did not promote copper binding to SOD1, excluding a previously suggested function of DJ-1 as a copper chaperone. Overall, our data do not support the role of DJ-1 as a metalloprotein. Conversely, oxidative treatment to the cells caused the complete and selective oxidation of C106 to sulfinic acid, consistent with the reported role of DJ-1 as a redox sensor.


In-cell NMR DJ-1 Deglycase Metalloprotein Cysteine sulfinic acid 


DJ-1 is a homodimeric protein encoded by the Park7 gene, ubiquitously expressed and highly conserved in all biological kingdoms. It is present in the cytoplasm, in the nucleus and in mitochondria. Human DJ-1 was first identified as an oncogene [1], as its expression is altered in numerous cancers, including breast, lung, esophageal and prostate cancers [2]. Mutations in the DJ-1 sequence are associated with recessive, early-onset Parkinson’s disease [3, 4]. DJ-1 is also implicated in ischemic injury and amyotrophic lateral sclerosis and androgen receptor regulation [5]. Furthermore, DJ-1 binds RNA transcripts potentially involved in many cellular processes, both in vivo and in vitro [6].

DJ-1 is implicated in many processes, including transcriptional regulation, reaction against oxidative stress, and mitochondrial regulation [7]. Several functions have been attributed to DJ-1, including proteasome regulation [8], chaperone activity towards α-synuclein [9, 10], protease activity towards transthyretin [11, 12, 13], and glyoxalase activity [14]. DJ-1 has been shown to act as an intracellular deglycase, reacting with methylglyoxal- and glyoxal-glycated proteins [15]. Very recently, DJ-1 has been also shown to act as DNA deglycase, a long sought-after enzyme catalyzing the repair of methylglyoxal- and glyoxal-glycated guanine [16]. While these recent findings indicate that DJ-1 has a major cytoprotective role, it is not yet clear how it exerts this function [17].

Each DJ-1 monomer is a 20 kDa of globular protein containing two exposed cysteine residues: C53 and C106. C106 resides in a so-called nucleophilic elbow, which constitutes the active site of the protein, and is easily oxidized by reactive oxygen species (ROS) due to its low thiol pKa value [18]. Multiple oxidation states of C106 have been reported (mainly –SO2H and –SO3H) in response to oxidative stress [19, 20, 21]. Mutation of C106 results in the loss of most of the putative DJ-1 functions [20, 22, 23]. Consequently, redox-sensing and ROS-quenching activities of DJ-1, by direct reaction with C106, have been proposed [7, 22, 24].

DJ-1 has also been shown to bind copper, zinc and mercury in vitro [25]. Binding of these toxic metals would confer DJ-1 a protective function against metal-induced cytotoxicity. Zinc binding occurs with low micromolar affinity at the protein active site, with E18, C106 and two water molecules as ligands. Zinc is bound with higher affinity with respect to other divalent metal ions [26]. For copper, two distinct binding modes have been reported: Cu(II) binds to the same active site as zinc [27], whereas one Cu(I) per monomer binds with higher affinity at the homodimer interface, with two C53 residues, one from each monomer, as ligands [28]. It has been suggested that DJ-1 might act as copper chaperone for superoxide dismutase 1 (SOD1), in place of the known metallochaperone for SOD1, CCS [27, 29].

While potentially relevant, most of the available data on DJ-1 metal binding were obtained in vitro, in the absence of the physiological competitors present in the cell. In parallel, while the structure of DJ-1 upon oxidation of C106 has been extensively characterized in vitro, the actual oxidation state of C106 in cells exposed to oxidizing agents has not been clearly defined. In this work, we sought to clarify some of these aspects by in-cell NMR [30, 31]. We applied an established protocol to observe protein structural changes at the residue level in the cytoplasm of live human cells [32, 33]. By comparison with in vitro NMR data, we could analyze the change in conformation of intracellular DJ-1 based on chemical shift perturbation, that is extremely sensitive to metal binding and changes in redox state, as a result of different cell culture conditions (i.e., addition of metal ions, treatment with H2O2). In human cells, we did not observe any metal-bound form of DJ-1, neither with zinc nor with copper in any oxidation state. Consistently, we did not observe copper metallation of SOD1 upon co-expression of DJ-1 in copper-treated cells. These data pointed us to reconsider the potential role of DJ-1 as a metallochaperone. Conversely, DJ-1 was efficiently oxidized upon treatment with H2O2 and a single species was produced, corresponding to the C106-sulfinic acid state, thereby suggesting a physiological function for DJ-1 oxidation.

Materials and methods

Gene cloning

The plasmid “DJ-1 TEV site pET15b”, used for bacterial expression, was a gift from Mark Wilson (Addgene plasmid # 60687). For mammalian expression, the WT DJ-1 gene was amplified by PCR and cloned into the pHLsec vector [34] between EcoRI and XhoI restriction enzyme sites. The clone was verified by DNA sequencing. pHLsec vectors encoding WT human SOD1 and CCS had been previously obtained [32].

Human cell culture and transfection

HEK293T (ATCC CRL-3216) cells were maintained in DMEM high glucose (Life Technologies) supplemented with l-glutamine, antibiotics (penicillin and streptomycin) and 10% FBS (Gibco) in uncoated 75 cm2 plastic flasks and incubated at 37 °C, 5% CO2 in a humidified atmosphere. The cells were transiently transfected with the pHLsec plasmid containing the gene of interest using polyethylenimine (PEI), as previously described [32, 33]. For co-expression of two proteins, cells were transfected with a mixture of pHLsec plasmids containing the genes of interest. [U–15N]-BioExpress6000 medium (Cambridge Isotope Laboratories) was used to produce [U–15N]-labeled samples, supplemented with 2% FBS and antibiotics. For [15N]cysteine labeling, a reconstituted medium was prepared following the reported composition of DMEM (Sigma), in which [15N]cysteine was added together with the other unlabelled components. Zinc-treated samples were obtained by adding 10 µM of ZnSO4 to the expression medium immediately after transfection. Cu(II)-treated cells were obtained by adding 100 µM of CuCl2 48 h after transfection, followed by 24 h treatment prior to cell collection. To monitor DJ-1 oxidation, cells expressing DJ-1 were treated with 10 mM of H2O2 for 1 h before cell collection.

In-cell NMR sample preparation

Samples for in-cell NMR were prepared following a reported protocol [33]. Transfected cells were detached with trypsin–EDTA 0.05%, suspended in DMEM + 10% FBS, washed once with PBS and re-suspended in one pellet volume of DMEM supplemented with 90 mM of glucose, 70 mM of HEPES and 20% D2O. The cell suspension was transferred in a 3 mm Shigemi NMR tube, which was gently spun to sediment the cells. Cell viability before and after NMR experiments was assessed by trypan blue staining. After the NMR experiments, the cells were collected and the supernatant was checked for protein leakage by NMR. The cell lysates were prepared by freeze–thaw cycles in PBS buffer followed by centrifugation to remove the insoluble fraction.

Expression and purification of DJ-1

E. coli BL21(DE3) Gold cells (Stratagene) were transformed with the plasmid “DJ-1 TEV site pET15b”. Cells were grown overnight at 37 °C in LB, harvested and re-suspended in 15N-labeled M9 medium. Protein expression was induced with 0.5 mM of IPTG and carried on at 37 °C; the cells were harvested after 4 h from induction. Protein purification was performed using a nickel-chelating HisTrap 5 ml column (GE Healthcare). After digestion with AcTEV protease (Invitrogen) overnight at 25 °C, the protein was separated from the affinity tag in the HisTrap column. Finally, the protein was exchanged against in 20 mM of KPi buffer pH 7, and 5 mM of DTT was added to prevent protein aggregation.

Metal binding and oxidation of DJ-1 in vitro

Zinc-DJ-1 and Cu(II)-DJ-1 were prepared by titrating aliquots of [U–15N]-DJ-1 with up to 1 equivalent-per-monomer of ZnSO4 and CuCl2, respectively, in KPi 20 mM buffer, pH 7. The spectral changes were monitored by NMR. Cu(I)-DJ-1 was prepared in an oxygen-free glove box, by titrating [U–15N]-DJ-1 (in oxygen-free buffer) with up to 0.5 equivalents-per-monomer of a 50 mM stock solution tetrakis(acetonitrile)copper(I) hexafluorophosphate, [Cu(I)(CH3CN)4]PF6, dissolved in acetonitrile. The spectral changes were monitored by NMR. Oxidized DJ-1 was obtained as follows: an aliquot containing 100 µM DJ-1 was washed with 20 mM of KPi buffer pH 7 in a 10 kDa-cutoff centricon to remove DTT and subsequently incubated with 1 mM of H2O2 overnight at 4 °C. Excess of unreacted H2O2 was removed by washes in centricon.

NMR data collection and analysis

In-cell NMR data were collected at a 950 MHz Bruker Avance spectrometer equipped with a TCI CryoProbe. For each cell sample, 1D 1H and 1H–15N SOFAST-HMQC [35] spectra were recorded at 308 K (total acquisition time ~ 1 h). The same spectra were recorded on the corresponding cell lysates. In vitro NMR data were collected on samples of purified DJ-1 at either a 950 MHz or a 900 MHz Bruker Avance spectrometer, both equipped with TCI CryoProbes. 1H–15N SOFAST-HMQC spectra of apo DJ-1, Cu(II)-DJ-1 and oxidized DJ-1 were acquired at 308 K; 1H–15N SOFAST-HMQC spectra of zinc-DJ-1 and Cu(I)-DJ1 were acquired at 298 K. The spectra were processed with Topspin NMR data processing software and analyzed with CARA. The previously reported backbone amide resonance assignment of DJ-1 [36] (BMRB Entry 17507) was transferred to the 1H–15N NMR spectra of reduced and oxidized apo DJ-1 at 298 K with the aid of 3D NOESY-HSQC spectra, and subsequently transferred to the spectra recorded at 308 K, by collecting a series of spectra at increasing temperatures. Combined chemical shift difference (CCSD) values for each amide signal were calculated with the formula: \( {\text{CCSD}} = \sqrt {\frac{{\left( {\Delta \delta^{1} {\text{H}}} \right)^{2} + \left( {\Delta \delta^{15} {\text{N}}/5} \right)^{2} }}{2}} \).

Mass spectrometry

Electrospray ionization mass spectrometry (ESI–MS) analysis was performed with a Thermo LTQ Orbitrap (FT-MSn) spectrometer on unlabelled samples of 37 µM reduced and oxidized DJ-1. Both samples were dialyzed against MilliQ H2O prior to analysis.


Intracellular DJ-1 is soluble and free from interactions

To determine its intracellular conformation in the absence of external stimuli, DJ-1 was overexpressed in HEK293T cells in [U–15N]-growth medium, and analyzed by in-cell NMR. The resulting 1H–15N NMR spectra were of surprisingly high quality considering the relatively high molecular size of the DJ-1 dimer (Fig. 1a). Most signals arising from DJ-1 could be resolved in the spectra, indicating that the protein does not interact with slowly tumbling cellular components and behaves as a free solute in the cytoplasm. Comparison with in vitro NMR spectra (Fig. 1c) revealed that, in untreated cells, intracellular DJ-1 is dimeric and metal free, with C106 in the reduced state. Upon cell lysis, a spectrum virtually identical to that recorded on DJ-1 in vitro was obtained (Fig. 1b). Minor chemical shift differences were observed between in vitro and in-cell (Fig. 1d) and lysate resonances (Fig. 1e), which are likely due to small differences in pH, ionic strength and to small, weakly interacting co-solutes.
Fig. 1

a 1H–15N NMR spectrum of cells expressing [U–15N]-DJ-1; b 1H–15N NMR spectrum of the corresponding lysate; c 1H–15N NMR spectrum of purified apo, reduced [U–15N]-DJ-1; d CCSD calculated for each residue between DJ-1 in-cell and in vitro; e CCSD calculated for each residue between DJ-1 in the cell lysate and in vitro

DJ-1 does not bind divalent metal ions in the cell

DJ-1 has been shown to bind zinc and copper ions in vitro [26, 27, 28]. To determine the metal-binding behavior of the protein in the cellular environment, we analyzed the conformation of DJ-1 in 1H–15N NMR spectra recorded on cells treated with zinc or copper in excess. The in-cell NMR spectra of zinc-treated cells did not reveal any appreciable changes in chemical shift or intensity of DJ-1 resonances with respect to untreated cells, indicating that no metal binding occurred in the cell (Fig. 2a, c). Comparison with the NMR spectra of [U–15N]-DJ-1 titrated in vitro with zinc confirmed this hypothesis: zinc binding induced clear chemical shift changes in the proximity of the metal binding site previously identified (Fig. 2b, d). In addition, in the case of Cu(II) treatment, no changes in chemical shift or intensity were observed with respect to untreated cells (Fig. 2e, g), whereas in vitro titration of DJ-1 with Cu(II) caused the complete disappearance of the residues close to the binding site, and severe broadening of farther residues, due to paramagnetic relaxation enhancement (Fig. 2f, h). Comparison of the signal intensity ratios (Fig. 2g, h) clearly shows that no paramagnetic effect is observed in cells, indicating that intracellular Cu(II) binding does not occur.
Fig. 2

a 1H–15N NMR spectra of cells expressing [U–15N]-DJ-1 in the absence (black) or in the presence (red) of excess zinc; b 1H-15N NMR spectra of purified apo-DJ-1 (black) and Zn-DJ-1 (red); c CCSD plot between DJ-1 in zinc-treated cells and DJ-1 in untreated cells; d CCSD plot between Zn-DJ-1 and apo-DJ-1 in vitro; e 1H-15N NMR spectra of cells expressing [U-15N]-DJ-1 either untreated (black) or treated (blue) with excess copper; f 1H-15N NMR spectra of purified apo-DJ-1 (black) and Cu(II)-DJ-1 (blue); g signal intensity ratio calculated for each amide signal between DJ-1 in copper-treated and DJ-1 in untreated cells; h signal intensity ratio between Cu(II)-DJ-1 and apo-DJ-1 in vitro

DJ-1 does not promote Cu(I) transfer to SOD1

Treatment of cultured human cells with Cu(II) is expected to increase the intracellular availability of Cu(I), in accordance with the known pathway of copper uptake in eukaryotes [37]. Consistently, we investigated also the Cu(I) binding properties of DJ-1 in Cu(II)-treated cells. Cu(I) binding in vitro induced subtle but clearly visible changes in the 1H–15N NMR spectra of DJ-1: few signals arising from C53 and from residues structurally adjacent to it decreased in intensity, possibly broadened by chemical exchange, while the other protein signals were unperturbed (Fig. 3a). In Cu(II)-treated cells, the corresponding signals of DJ-1 did not decrease in intensity (Fig. 2e), suggesting that DJ-1 did not bind Cu(I) at the C53 binding site.
Fig. 3

a 1H–15N NMR spectra of purified apo-DJ-1 (black) and Cu(I)-DJ-1 (orange). The signals disappearing upon Cu(I) binding are labeled; b the imino region of 1H NMR spectra acquired, from top to bottom, on copper-treated cells co-expressing SOD1 and CCS, on copper-treated cells expressing SOD1 alone, on copper-treated cells co-expressing SOD1 and DJ-1, on purified DJ-1 in vitro. Histidine signals from E,Zn-SOD1 and Cu(I),Zn-SOD1 are indicated; c 1H-15N NMR spectra of cells expressing DJ-1 in the absence of copper (black) or co-expressing DJ-1 and CCS in the presence of copper (violet). The signals which disappear upon Cu(I) binding in vitro are labeled

The strict cellular control over copper trafficking is mediated by specific metallochaperones that allow Cu(I) to reach its final acceptors [38]. We have previously shown that Cu(I) delivery to intracellular SOD1 can be promoted by co-expressing the copper chaperone CCS [32, 39]. Therefore, if DJ-1 acted as an alternative copper chaperone for SOD1 as suggested, one would expect that co-expression of DJ-1 and SOD1 in the presence of sufficient levels of both zinc and copper (as SOD1 needs to bind zinc to fold correctly) would lead to the observation of fully mature SOD1 (Cu,Zn-SOD1), mediated by a transient Cu(I)-DJ-1 species. However, this was not the case, as copper-free SOD1 was still mostly detected in the imino region of the 1D 1H in-cell NMR spectra, essentially as much as a control sample in which DJ-1 was not co-expressed (Fig. 3b). A sample of cells in which DJ-1 and CCS were co-expressed in the presence of copper was also analyzed, to assess whether CCS could deliver Cu(I) to DJ-1. In addition in this case, the in-cell NMR spectra did not reveal any significant changes compared to untreated cells, set aside for few additional signals belonging to CCS (Fig. 3c).

Cysteine 106 is quantitatively oxidized to sulfinic acid

The thiol group of C106 in the active site of DJ-1 is known to be sensitive to oxidation in cells exposed to oxidative stress, however, multiple oxidation states have been reported. We, therefore, analyzed the conformation of intracellular DJ-1 in cells treated with H2O2, and compared it with in vitro samples at defined C106 oxidation states. The in-cell NMR data revealed that DJ-1 in cells treated with 10 mM of H2O2 for 1 h is quantitatively oxidized, resulting in a single species being produced (Fig. 4a). The same species could be reproduced in vitro by incubating DJ-1 with 1 mM of H2O2 overnight at 4 °C (Fig. 4b), as shown by chemical shift analysis (Fig. 4d, e). ESI–MS analysis of unlabelled protein samples identically prepared indicated that upon treatment in vitro with H2O2, reduced DJ-1 (Fig. 4f) was oxidized to sulfinic acid (C106–SO2H, Fig. 4g), and that the reaction did not proceed to sulfonic acid (C106–SO3H). To determine whether other oxidation states were produced in H2O2-treated cells, lysates from [15N]-cysteine-labeled cells were analyzed. The resulting 1H–15N NMR spectra confirmed that a single species (C106–SO2H) was produced following treatment with H2O2, as no additional signals arising from C106 in other redox sates of were present (Fig. 4c).
Fig. 4

a 1H–15N NMR spectra of cells expressing [U–15N]-DJ-1 without (black) or with (magenta) treatment with H2O2; b 1H–15N NMR spectra of purified reduced DJ-1 (DJ-1–SH, black) and oxidized DJ1 (DJ-1–SO2H, magenta); c 1H–15N NMR spectra lysates from cells expressing [15N-cysteine]-DJ-1 without (black) or with (magenta) treatment with H2O2. The signals arising from DJ-1 cysteines are labeled, and the oxidation states of C106 as determined in vitro are indicated; d CCSD plot of DJ-1 between H2O2-treated and untreated cells; e CCSD plot between DJ-1 and SO2H and DJ-1 and SH in vitro; f ESI–MS spectrum of purified DJ-1–SH; g ESI–MS spectrum of purified DJ-1–SO2H


The proposed metallochaperone function of DJ-1 has joined an already long list of biological roles that have been attributed to this ubiquitous protein [17]. The metal-binding properties of DJ-1 have been characterized in vitro, however, direct proof of protein metallation in situ has been lacking. By in-cell NMR, changes in protein conformation occurring in cells can be directly monitored at high resolution, making it an ideal approach to validate functional hypotheses in the physiological environment. Overall, our data indicate that in our experimental conditions DJ-1 does not bind intracellular zinc or copper ions, despite both having been supplemented in excess and being incorporated in the cells, as previously verified in the same conditions [32, 40]. This outcome can be explained by considering that the intracellular concentration of free metal, both copper and zinc, is kept negligible by the cellular homeostasis [41]. For the most part, these metals are bound to other molecules either involved in metal trafficking or being the final acceptors. In the case of zinc, the amount of metal available for binding to any given protein depends roughly on the binding affinity of the latter compared to that of all the other competitors. For zinc, most intracellular protein binding sites (zinc fingers and metallothioneins being the most abundant) have affinities estimated in the picomolar range (K d = 10−11 to 10−12 M [42]), while the intracellular glutathione (K d = 6.3 × 10−5 M) does not act directly as a zinc pool [42]. In addition, active transport systems shuttle the excess zinc to zinc storing organelles or remove it from the cell through a process called ‘muffling’ [43]. Given that the reported affinity of DJ-1 for zinc (K d = 6 × 10−7 M [26]) is relatively low, one can expect the amount of zinc available to DJ-1 to be negligible, even in cells treated with excess zinc. This outcome is opposite to that of SOD1, which has a much higher affinity (K d = 4.2 × 10−14 M [44]) and spontaneously binds zinc in the same cellular environment [32].

Copper trafficking is even more strictly regulated, and copper transfer only occurs via associative ligand substitution between copper-binding proteins [37]. Specific copper-delivery pathways exist for each of the few cellular copper proteins, in which metallochaperones relay copper ions from the plasma membrane transporter CTR1, which imports copper as Cu(I), to each final acceptor in a process driven by increasing affinities for copper [38, 45]. Such specific pathways also act as kinetic barriers, preventing unwanted molecules from receiving copper. All known intracellular copper chaperones bind Cu(I) with high affinity, ranging between 10−14 and 10−16 M [45]. In this context, a physiological role for Cu(II) binding of DJ-1 seems unlikely. On the other hand, the reported affinity of the Cu(I)-binding site at the homodimer interface of DJ-1 (K d = 6 × 10−16 M [28]) falls within the range expected for an intracellular copper protein. Despite this, in our experimental conditions, Cu(I)-DJ-1 was not formed, and co-expression of DJ-1 together with SOD1 did not promote the formation of Cu,Zn-SOD1, indicating that copper transfer mediated by DJ-1 did not occur. A possible explanation for this outcome is that DJ-1, like other copper proteins, needs a specific metallochaperone for its own maturation. Increased levels of the final acceptor (in this case, DJ-1) might overload the endogenous copper-delivery pathway, which has to be boosted by increasing the levels of metallochaperone, in analogy with SOD1 needing increased levels of CCS. Currently, however, no hypothesis has been made on the existence of such metallochaperone for DJ-1. The involvement of CCS in DJ-1 copper delivery can be excluded, as increasing CCS levels did not have any effect on DJ-1 metallation. While the possibility that DJ-1 binds Cu(I) under certain conditions cannot be ruled out, these data show that DJ-1 cannot act as a copper chaperone itself, as unlike CCS it cannot receive Cu(I) directly from CTR1 at the plasma membrane.

DJ-1 is known to be oxidized in response to cellular oxidative stress. The oxidation is mediated by reactive oxygen species (ROS) and occurs selectively on the reactive C106 residue in the active site of the protein. Intracellular C106 oxidation has been reported to produce both cysteine-sulfinic (–SO2H) and -sulfonic (–SO3H) acid [20, 21]. While there is consensus that C106–SO2H is the functionally relevant form, multiple oxidation states of C106 have been reported upon cell treatment with oxidizing agents [46, 47]. By in-cell NMR, the intracellular speciation of C106 could be directly accessed, as the chemical shift is highly sensitive to small changes in the chemical environment. In-cell NMR data revealed that H2O2-treated cells contain only one species of DJ-1, in which C106 is oxidized to sulfinic acid. [15N]-cysteine labeling further reduces spectral complexity and simplifies data analysis, allowing multiple redox species, if present, to be identified [48, 49]. As no other species were detected in NMR spectra of [15N]-cysteine-labeled samples, it can be estimated that > 90% of the total DJ-1 is in the C106-sulfinic acid form. Notably, these data were obtained with the highest concentration of H2O2 that preserved cell viability, and treating the cells with lower concentrations of H2O2 for the same period of time is unlikely to oxidize DJ-1 to a higher oxidation state. Therefore, in this cellular system, C106–SO3H does not form upon acute treatment with H2O2. These results are consistent with the reactivity of C106 in vitro, where C106–SO2H forms readily, sometimes spontaneously during sample handling [20], while oxidation to C106–SO3H requires harsher conditions [10]. Nevertheless, the fact that C106–SO3H DJ-1 has been observed under certain circumstances [21, 46] suggests that its formation might occur in other types of cells, or under stress conditions closer to the protracted, low-level stress occurring in dopaminergic neurons.


Using a direct, high-resolution approach, we have assessed the biological relevance of several post-translational modifications occurring to DJ-1 that have been linked to severe pathologies such as Parkinson’s disease, amyotrophic lateral sclerosis and cancer. The results obtained suggest that the metal-binding properties of DJ-1 active site (i.e., E18 and C106) towards copper and zinc should be reconsidered in light of the low-binding affinity. While thermodynamically allowed, Cu(I) binding to the C53 residues at the dimer interface does not occur, indicating that DJ-1 cannot autonomously act as a copper chaperone.

Furthermore, we have shown that the oxidation state of C106 of intracellular DJ-1 can be monitored directly by in-cell NMR. Oxidation of C106 to sulfinic acid occurs efficiently in H2O2-treated cells, and the reaction does not proceed further to sulfonic acid. With such approach, it will be possible to determine how DJ-1 responds to varying concentration of oxidant and to other types of cellular stress, such as treatment with rotenone. Moreover, the time dependence of the intracellular oxidation process could in principle be studied by real-time in-cell NMR spectroscopy.

In addition to specific conformational changes, in-cell NMR also reveals that the protein is free from any interaction with intracellular partners. In our experience, this behavior is somewhat unexpected when considering the many DJ-1 cellular functions. Indeed, a protein with chaperone activity and regulation roles in many pathways (such as transcriptional regulation, proteasome inhibition) would more likely be invisible by in-cell NMR due to the plethora of interacting partners found in the cell. An example of the such behavior is given by profilin, which is simultaneously interacting with actin, phosphoinositides and several other proteins [50]. In comparison, an enzymatic activity of DJ-1 such as protease, glyoxalase or deglycase would be more consistent with the observed behavior, as the intracellular efficiency of an enzyme depends on its ability to diffuse freely within a cellular compartment, reaching its substrates where their concentration is higher. In this respect, being able to observe structural features of intracellular DJ-1 makes it possible to further investigate the enzymatic activity of DJ-1 through in-cell NMR, especially in light of its recently reported activity as protein and DNA deglycase [15, 16].



This work was funded by Instruct, part of the European Strategy Forum on Research Infrastructures (ESFRI) and supported by national member subscriptions. Specifically, we thank the ESFRI Instruct Core Centre CERM-Italy. This work has also been supported by iNEXT, Grant number 653706, funded by the Horizon 2020 programme of the European Union.


  1. 1.
    Nagakubo D, Taira T, Kitaura H et al (1997) DJ-1, a novel oncogene which transforms mouse NIH3T3 cells in cooperation with ras. Biochem Biophys Res Commun 231:509–513. CrossRefPubMedGoogle Scholar
  2. 2.
    Cao J, Lou S, Ying M, Yang B (2015) DJ-1 as a human oncogene and potential therapeutic target. Biochem Pharmacol 93:241–250. CrossRefPubMedGoogle Scholar
  3. 3.
    van Duijn CM, Dekker MC, Bonifati V et al (2001) Park7, a novel locus for autosomal recessive early-onset parkinsonism, on chromosome 1p36. Am J Hum Genet 69:629–634. CrossRefPubMedPubMedCentralGoogle Scholar
  4. 4.
    Bonifati V, Rizzu P, van Baren MJ et al (2003) Mutations in the DJ-1 gene associated with autosomal recessive early-onset parkinsonism. Science 299:256–259. CrossRefPubMedGoogle Scholar
  5. 5.
    Wilson MA (2011) The role of cysteine oxidation in DJ-1 function and dysfunction. Antioxid Redox Signal 15:111–122. CrossRefPubMedPubMedCentralGoogle Scholar
  6. 6.
    van der Brug MP, Blackinton J, Chandran J et al (2008) RNA binding activity of the recessive parkinsonism protein DJ-1 supports involvement in multiple cellular pathways. Proc Natl Acad Sci USA 105:10244–10249. CrossRefPubMedPubMedCentralGoogle Scholar
  7. 7.
    Ariga H, Takahashi-Niki K, Kato I et al (2013) Neuroprotective function of DJ-1 in Parkinson’s disease. Oxid Med Cell Longev 2013:683920. CrossRefPubMedPubMedCentralGoogle Scholar
  8. 8.
    Moscovitz O, Ben-Nissan G, Fainer I et al (2015) The Parkinson’s-associated protein DJ-1 regulates the 20S proteasome. Nat Commun 6:6609. CrossRefPubMedGoogle Scholar
  9. 9.
    Shendelman S, Jonason A, Martinat C et al (2004) DJ-1 is a redox-dependent molecular chaperone that inhibits alpha-synuclein aggregate formation. PLoS Biol 2:e362. CrossRefPubMedPubMedCentralGoogle Scholar
  10. 10.
    Zhou W, Zhu M, Wilson MA et al (2006) The oxidation state of DJ-1 regulates its chaperone activity toward alpha-synuclein. J Mol Biol 356:1036–1048. CrossRefPubMedGoogle Scholar
  11. 11.
    Olzmann JA, Brown K, Wilkinson KD et al (2004) Familial Parkinson’s disease-associated L166P mutation disrupts DJ-1 protein folding and function. J Biol Chem 279:8506–8515. CrossRefPubMedGoogle Scholar
  12. 12.
    Koide-Yoshida S, Niki T, Ueda M et al (2007) DJ-1 degrades transthyretin and an inactive form of DJ-1 is secreted in familial amyloidotic polyneuropathy. Int J Mol Med 19:885–893PubMedGoogle Scholar
  13. 13.
    Chen J, Li L, Chin L-S (2010) Parkinson disease protein DJ-1 converts from a zymogen to a protease by carboxyl-terminal cleavage. Hum Mol Genet 19:2395–2408. CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Lee J, Song J, Kwon K et al (2012) Human DJ-1 and its homologs are novel glyoxalases. Hum Mol Genet 21:3215–3225. CrossRefPubMedGoogle Scholar
  15. 15.
    Richarme G, Mihoub M, Dairou J et al (2015) Parkinsonism-associated protein DJ-1/Park7 is a major protein deglycase that repairs methylglyoxal- and glyoxal-glycated cysteine, arginine, and lysine residues. J Biol Chem 290:1885–1897. CrossRefPubMedGoogle Scholar
  16. 16.
    Richarme G, Liu C, Mihoub M et al (2017) Guanine glycation repair by DJ-1/Park7 and its bacterial homologs. Science 357:208–211. CrossRefPubMedGoogle Scholar
  17. 17.
    Biosa A, Sandrelli F, Beltramini M et al (2017) Recent findings on the physiological function of DJ-1: beyond Parkinson’s disease. Neurobiol Dis. PubMedGoogle Scholar
  18. 18.
    Witt AC, Lakshminarasimhan M, Remington BC et al (2008) Cysteine pKa depression by a protonated glutamic acid in human DJ-1. Biochemistry 47:7430–7440. CrossRefPubMedPubMedCentralGoogle Scholar
  19. 19.
    Mitsumoto A, Nakagawa Y, Takeuchi A et al (2001) Oxidized forms of peroxiredoxins and DJ-1 on two-dimensional gels increased in response to sublethal levels of paraquat. Free Radic Res 35:301–310CrossRefPubMedGoogle Scholar
  20. 20.
    Canet-Avilés RM, Wilson MA, Miller DW et al (2004) The Parkinson’s disease protein DJ-1 is neuroprotective due to cysteine-sulfinic acid-driven mitochondrial localization. Proc Natl Acad Sci USA 101:9103–9108. CrossRefPubMedPubMedCentralGoogle Scholar
  21. 21.
    Kinumi T, Kimata J, Taira T et al (2004) Cysteine-106 of DJ-1 is the most sensitive cysteine residue to hydrogen peroxide-mediated oxidation in vivo in human umbilical vein endothelial cells. Biochem Biophys Res Commun 317:722–728. CrossRefPubMedGoogle Scholar
  22. 22.
    Taira T, Saito Y, Niki T et al (2004) DJ-1 has a role in antioxidative stress to prevent cell death. EMBO Rep 5:213–218. CrossRefPubMedPubMedCentralGoogle Scholar
  23. 23.
    Martinat C, Shendelman S, Jonason A et al (2004) Sensitivity to oxidative stress in DJ-1-deficient dopamine neurons: an ES- derived cell model of primary Parkinsonism. PLoS Biol 2:e327. CrossRefPubMedPubMedCentralGoogle Scholar
  24. 24.
    Mitsumoto A, Nakagawa Y (2001) DJ-1 is an indicator for endogenous reactive oxygen species elicited by endotoxin. Free Radic Res 35:885–893CrossRefPubMedGoogle Scholar
  25. 25.
    Björkblom B, Adilbayeva A, Maple-Grødem J et al (2013) Parkinson disease protein DJ-1 binds metals and protects against metal-induced cytotoxicity. J Biol Chem 288:22809–22820. CrossRefPubMedPubMedCentralGoogle Scholar
  26. 26.
    Tashiro S, Caaveiro JMM, Wu C-X et al (2014) Thermodynamic and structural characterization of the specific binding of Zn(II) to human protein DJ-1. Biochemistry 53:2218–2220. CrossRefPubMedGoogle Scholar
  27. 27.
    Girotto S, Cendron L, Bisaglia M et al (2014) DJ-1 is a copper chaperone acting on SOD1 activation. J Biol Chem 289:10887–10899. CrossRefPubMedPubMedCentralGoogle Scholar
  28. 28.
    Puno MR, Patel NA, Møller SG et al (2013) Structure of Cu(I)-bound DJ-1 reveals a biscysteinate metal binding site at the homodimer interface: insights into mutational inactivation of DJ-1 in Parkinsonism. J Am Chem Soc 135:15974–15977. CrossRefPubMedGoogle Scholar
  29. 29.
    Xu XM, Lin H, Maple J et al (2010) The Arabidopsis DJ-1a protein confers stress protection through cytosolic SOD activation. J Cell Sci 123:1644–1651. CrossRefPubMedGoogle Scholar
  30. 30.
    Luchinat E, Banci L (2017) In-cell NMR: a topical review. IUCrJ 4:108–118. PubMedGoogle Scholar
  31. 31.
    Plitzko JM, Schuler B, Selenko P (2017) Structural Biology outside the box-inside the cell. Curr Opin Struct Biol 46:110–121. CrossRefPubMedGoogle Scholar
  32. 32.
    Banci L, Barbieri L, Bertini I et al (2013) Atomic-resolution monitoring of protein maturation in live human cells by NMR. Nat Chem Biol 9:297–299. CrossRefPubMedPubMedCentralGoogle Scholar
  33. 33.
    Barbieri L, Luchinat E, Banci L (2016) Characterization of proteins by in-cell NMR spectroscopy in cultured mammalian cells. Nat Protoc 11:1101–1111. CrossRefPubMedGoogle Scholar
  34. 34.
    Aricescu AR, Lu W, Jones EY (2006) A time- and cost-efficient system for high-level protein production in mammalian cells. Acta Crystallogr D Biol Crystallogr 62:1243–1250. CrossRefPubMedGoogle Scholar
  35. 35.
    Schanda P, Brutscher B (2005) Very fast two-dimensional NMR spectroscopy for real-time investigation of dynamic events in proteins on the time scale of seconds. J Am Chem Soc 127:8014–8015. CrossRefPubMedGoogle Scholar
  36. 36.
    Malgieri G, Eliezer D (2008) Structural effects of Parkinson’s disease linked DJ-1 mutations. Protein Sci Publ Protein Soc 17:855–868. CrossRefGoogle Scholar
  37. 37.
    Nevitt T, Ohrvik H, Thiele DJ (2012) Charting the travels of copper in eukaryotes from yeast to mammals. Biochim Biophys Acta 1823:1580–1593. CrossRefPubMedPubMedCentralGoogle Scholar
  38. 38.
    Banci L, Bertini I, Ciofi-Baffoni S (2009) Copper trafficking in biology: an NMR approach. HFSP J 3:165–175. CrossRefPubMedPubMedCentralGoogle Scholar
  39. 39.
    Luchinat E, Barbieri L, Rubino JT et al (2014) In-cell NMR reveals potential precursor of toxic species from SOD1 fALS mutants. Nat Commun 5:5502. CrossRefPubMedGoogle Scholar
  40. 40.
    Luchinat E, Gianoncelli A, Mello T et al (2015) Combining in-cell NMR and X-ray fluorescence microscopy to reveal the intracellular maturation states of human superoxide dismutase 1. Chem Commun Camb Engl 51:584–587. CrossRefGoogle Scholar
  41. 41.
    Finney LA, O’Halloran TV (2003) Transition metal speciation in the cell: insights from the chemistry of metal ion receptors. Science 300:931–936. CrossRefPubMedGoogle Scholar
  42. 42.
    Petering DH, Mahim A (2017) Proteomic high affinity Zn(2+) trafficking: where does metallothionein fit in? Int J Mol Sci. PubMedPubMedCentralGoogle Scholar
  43. 43.
    Colvin RA, Holmes WR, Fontaine CP, Maret W (2010) Cytosolic zinc buffering and muffling: their role in intracellular zinc homeostasis. Met Integr Biometal Sci 2:306–317. CrossRefGoogle Scholar
  44. 44.
    Crow JP, Sampson JB, Zhuang Y et al (1997) Decreased zinc affinity of amyotrophic lateral sclerosis-associated superoxide dismutase mutants leads to enhanced catalysis of tyrosine nitration by peroxynitrite. J Neurochem 69:1936–1944CrossRefPubMedGoogle Scholar
  45. 45.
    Banci L, Bertini I, Ciofi-Baffoni S et al (2010) Affinity gradients drive copper to cellular destinations. Nature 465:645–648. CrossRefPubMedGoogle Scholar
  46. 46.
    Miyazaki S, Yanagida T, Nunome K et al (2008) DJ-1-binding compounds prevent oxidative stress-induced cell death and movement defect in Parkinson’s disease model rats. J Neurochem 105:2418–2434. CrossRefPubMedGoogle Scholar
  47. 47.
    Fernandez-Caggiano M, Schröder E, Cho H-J et al (2016) Oxidant-induced interprotein disulfide formation in cardiac protein DJ-1 occurs via an interaction with peroxiredoxin 2. J Biol Chem 291:10399–10410. CrossRefPubMedPubMedCentralGoogle Scholar
  48. 48.
    Banci L, Barbieri L, Luchinat E, Secci E (2013) Visualization of redox-controlled protein fold in living cells. Chem Biol 20:747–752. CrossRefPubMedGoogle Scholar
  49. 49.
    Mercatelli E, Barbieri L, Luchinat E, Banci L (2016) Direct structural evidence of protein redox regulation obtained by in-cell NMR. Biochim Biophys Acta 1863:198–204. CrossRefPubMedGoogle Scholar
  50. 50.
    Barbieri L, Luchinat E, Banci L (2015) Protein interaction patterns in different cellular environments are revealed by in-cell NMR. Sci Rep 5:14456. CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© SBIC 2017

Authors and Affiliations

  • Letizia Barbieri
    • 1
    • 2
  • Enrico Luchinat
    • 1
    • 3
  • Lucia Banci
    • 1
    • 4
  1. 1.Magnetic Resonance Centre (CERM)University of FlorenceFlorenceItaly
  2. 2.Interuniversity Consortium for Magnetic Resonance of Metallo Proteins (CIRMMP)FlorenceItaly
  3. 3.Department of Experimental and Clinical Biomedical Sciences “Mario Serio”University of FlorenceFlorenceItaly
  4. 4.Department of ChemistryUniversity of FlorenceFlorenceItaly

Personalised recommendations