The copper chaperone CCS facilitates copper binding to MEK1/2 to promote kinase activation

Normal physiology relies on the precise coordination of intracellular signaling pathways that respond to nutrient availability to balance cell growth and cell death. The canonical mitogen-activated protein kinase pathway consists of the RAF-MEK-ERK signaling cascade and represents one of the most well-de ﬁ ned axes within eukaryotic cells to promote cell proliferation, which underscores its frequent mutational activation in human cancers. Our recent studies illuminated a function for the redox-active micronutrient copper (Cu) as an

Normal physiology relies on the precise coordination of intracellular signaling pathways that respond to nutrient availability to balance cell growth and cell death. The canonical mitogen-activated protein kinase pathway consists of the RAF-MEK-ERK signaling cascade and represents one of the most well-defined axes within eukaryotic cells to promote cell proliferation, which underscores its frequent mutational activation in human cancers. Our recent studies illuminated a function for the redox-active micronutrient copper (Cu) as an intracellular mediator of signaling by connecting Cu to the amplitude of mitogen-activated protein kinase signaling via a direct interaction between Cu and the kinases MEK1 and MEK2. Given the large quantities of molecules such as glutathione and metallothionein that limit cellular toxicity from free Cu ions, evolutionarily conserved Cu chaperones facilitate efficient delivery of Cu to cuproenzymes. Thus, a dedicated cellular delivery mechanism of Cu to MEK1/2 likely exists. Using surface plasmon resonance and proximity-dependent biotin ligase studies, we report here that the Cu chaperone for superoxide dismutase (CCS) selectively bound to and facilitated Cu transfer to MEK1. Mutants of CCS that disrupt Cu(I) acquisition and exchange or a CCS small-molecule inhibitor were used and resulted in reduced Cu-stimulated MEK1 kinase activity. Our findings indicate that the Cu chaperone CCS provides fidelity within a complex biological system to achieve appropriate installation of Cu within the MEK1 kinase active site that in turn modulates kinase activity and supports the development of novel MEK1/2 inhibitors that target the Cu structural interface or blunt dedicated Cu delivery mechanisms via CCS.
The development of a multicellular organism depends on the precise synchronization of cell division, death, specialization, interactions, and movement. The orchestration of complex cellular processes in response to environmental changes is essential for organismal tissue homeostasis. Intrinsic molecular mechanisms enable the conversion of fluctuating extracellular and intracellular inputs into various outputs that drive cellular processes necessary for adaptation (1, 2). This interpretation and integration of inputs is controlled by signal transduction pathways that facilitate the precise coordination of cellular processes with dynamic spatial and temporal control (1, 2). One of the simplest evolutionarily conserved mediators of signal transduction is the protein kinase, an enzyme that traditionally phosphorylates a substrate at threonine, serine, and/or tyrosine residues (3). Kinases within complex protein modules directly respond to and/or sense growth factors, nutrients, and metabolites to influence signal amplification, duration, and frequency (1, 2).
Although the kinase signaling networks that integrate fluctuations in the abundance of many nutrients and metabolites are well-established (1, 2), the discovery of signaling molecules capable of mediating similar functions downstream of transition metal availability is underdeveloped. Traditionally, the redox-active transition metal copper (Cu) functions as a high affinity structural and/or catalytic cofactor within the active site of Cu-dependent enzymes (4)(5)(6). Phenotypic studies of the rare genetic diseases, Menkes and Wilson, solidified the physiological impact of aberrant Cu absorption and excretion (7)(8)(9)(10). Mechanistically, these diseases display deficiencies in cellular functions attributed to the dozens of known Cudependent enzymes and helped elucidate the cellular machinery responsible for the proper acquisition and distribution of Cu (5,7,8). Despite the need for strict homeostatic mechanisms to control Cu abundance, we know relatively little about direct kinase signaling pathways that respond to and or/ sense Cu abundance, especially those that integrate to influence cellular proliferation.
In pursuit of providing a molecular explanation for the observations that biological systems convert Cu abundance into cellular decisions like proliferation (11), we recently reported an unexpected link between the cellular acquisition of Cu and a mitogenic signaling cascade instrumental in cell proliferation (12). Namely, genetic ablation of the primary Cu transporter gene Ctr1, which in multiple organisms results in growth retardation and embryonic lethality (13)(14)(15), reduced mitogen-activated protein kinase (MAPK) signaling downstream of growth factor stimulated receptor tyrosine kinase activation (12). The canonical MAPK pathway consists of the RAF-MEK-ERK signaling cascade and represents one of the most well-defined axes within eukaryotic cells to promote cell proliferation (16). In response to receptor tyrosine kinase activation, GTP-bound rat sarcoma virus GTPase engages the serine/threonine rapidly accelerated fibrosarcoma protein kinases via their rat sarcoma virus binding domains to promote phosphorylation and activation of the MEK1 and MEK2 kinases, which in turn phosphorylate and activate the ERK1 and ERK2 kinases to drive cell proliferation (17,18). Layered on top of this signaling cascade are the multiple positive and negative regulatory mechanisms that fine-tune the frequency, duration, localization, and amplitude of canonical MAPK signaling (19), underscoring the importance of tightly regulating this pathway at the cellular and organismal level for a host of responses. Genetic, molecular, and biochemical interrogation of the requirement of Cu for MAPK pathway activation revealed a direct metal-kinase interaction between Cu and MEK1/2 that was sufficient to augment phosphorylation and activation of the ERK1 and ERK2 kinases (12). This is the first example of Cu directly regulating the activity of a mammalian kinase and has exposed a signaling paradigm that directly connects Cu to the signaling pathway components (6). Further, targeting aberrant MAPK signaling in melanomas and other cancers (20)(21)(22) contributes to the efficacy of Cu chelators (23)(24)(25)(26) that are traditionally used to treat Cu overload in Wilson disease patients (7) and highlights the importance of dissecting the molecular mechanism of MEK1/2 activation via Cu.
In the present study, we set out to interrogate the mechanism by which selective Cu delivery and subsequent MEK1/2 activation is achieved. In vitro and in mammalian cells, Custimulated MEK1/2 kinase activity required the Cu chaperone for superoxide dismutase (CCS) and could be blocked with the small molecule inhibitor DCAC50 that targets CCS. Taken together, the data provide a more precise understanding of how Cu cooperates with MEK1/2 that can be used to expand the landscape of Cu in kinase signal transduction.

CCS binds MEK1 to facilitate Cu binding and increased kinase activity
We first explored the existence of a dedicated cellular Cu delivery mechanism capable of supporting Cu-stimulated MEK1/2 activity. The evolutionarily conserved Cu chaperones facilitate the efficient delivery of Cu to cuproenzymes located in the cytosol, trans-Golgi network, or mitochondria (27). Based on the predominant cytosolic localization of MEK1/2, we tested whether the cytosolic Cu chaperones ATOX1 or CCS bind to MEK1 in single-cycle surface plasmon resonance (SPR) experiments. The recombinant MEK1 was coupled to a SPR sensor followed by injection of increasing concentrations of parvalbumin (Fig. 1A), as a negative control, or the Cu chaperones ATOX1 or CCS (Fig. 1, B and C). CCS interacted with MEK1 at an apparent dissociation constant (K D ) of 2.09 ± 0.48 μM ( Fig. 1C; Table 1) when compared with ATOX1 (Kd = 10.9 ± 5.65 μM), confirming a direct interaction between CCS and MEK1. However, both electrophoretic mobility shift assay and size-exclusion chromatography (SEC) of excess MEK1 or CCS suggest that the association between apo-CCS or Cu-CCS and MEK1 is transient based on the absence of a stable complex that could be coeluted (Fig. 1, D-G).
We next interrogated the relationship between the Cu chaperones and MEK1 kinase activity and Cu binding in vitro. A selective 3-fold enhancement in MEK1 kinase activity was observed when increasing concentrations of Cu loaded CCS, but not ATOX1, were added to MEK1 kinase assays (Fig. 1H), indicating that CCS is directly involved in Cu-stimulated MEK1/2 phosphorylation of ERK1/2 (24). Based on the observation that CCS, which binds Cu(I), increased MEK1 kinase activity, we evaluated whether this increase in activity was driven by Cu transfer. Stoichiometric addition of Cu-CCS to MEK1 and subsequent separation by SEC to capture MEK1 positive fractions for analysis by inductively-coupled plasma mass spectrometry (ICP-MS) resulted in an average of 1.068 Cu atoms bound per MEK1 protomer, which is significantly above the as isolated MEK1 protein from bacteria ( Fig. 1, I and J). As an important control, we next tested whether CCS transfer of Cu to MEK1 can occur in the presence of the Cu(I) chelator bathocuproinedisulfonic acid (BCS) and found an average increase of 0.95 Cu atoms bound per MEK1 protomer ( Fig. 1, I and J), which suggests that CCS is acting as a bonafide Cu chaperone for MEK1 instead of MEK1 binding Cu in solution. Further, when compared with a control sample in which CCS was incubated with Cu to determine stoichiometry, we detected a reduction in CCS Cu binding from typical 1:1 to either 0.5 Cu atoms or 0.2 Cu atoms per protomer when incubated with MEK1 in either the presence or absence of BCS, respectively (Fig. 1, I and J). Taken together, these data support a scenario in which the cytosolic Cu chaperone CCS functions as a direct Cu delivery mechanism to MEK1 via a transient association that in turn facilitates increased kinase activity.

Cu integration into MAPK signaling requires CCS Cu binding
To directly address an association between CCS and MAPK signaling in mammalian cells, BirA proximity-dependent biotin identification (BioID) reagents were generated to capture the weak or transient interaction between ATOX1 or CCS with MEK1 in living cells. Specifically, myc-epitope tagged mutant (R118G) BirA biotin ligase, which conjugates biotin to CCS mediates copper activation of MEK1/2 proteins within the immediate vicinity (10 nm) (28), was fused in frame to the N-terminus of ATOX1 or CCS and stably expressed in HEK-HT cells ( Fig. 2A). As expected, exogenous biotin treatment of HEK-HT cells expressing myc-BirA-ATOX1 resulted in the streptavidin recovery of biotinylated ATP7A (5, 27) ( Fig. 2A), which facilitates the transport of Cu into the lumen of the trans-Golgi network where Cu loading of Cu-dependent enzymes occurs, whereas myc-BirA-CCS biotinylated SOD1 (5, 27, 29) ( Fig. 2A) uses Cu as a cofactor for catalyzing the disproportionation of superoxide to hydrogen peroxide and dioxygen. Endogenous MEK1 was only biotinylated in presence of biotin and myc-BirA-CCS, indicating that CCS is proximal to MEK1/2 and may be responsible for Cu-stimulated activation of MAPK signaling. Interestingly, streptavidin recovery of biotinylated MEK1 from exogenous biotin treatment of HEK-HT cells expressing myc-BirA-CCS was significantly reduced in response to treatment with the Cu(I) chelator BCS and was unaffected by exogenous CuCl 2 (Fig. 2B).
To investigate whether CCS Cu binding is necessary for endogenous MAPK signaling, CCS was knocked down with two independent doxycycline-inducible shRNAs in HEK-HT cells. The knockdown of CCS decreased basal and either CuCl 2 or epidermal growth factor stimulated activation of P-ERK1/2 ( Fig. 2, C and D). Given the well characterized contribution of CCS to SOD1 Cu binding, we tested whether the loss of CCS was primarily acting through a direct CCS-MEK1 interaction and not a parallel pathway in which SOD1 activity is reduced. In contrast to the knockdown of CCS (Fig. 2, C and D), we found that decreased expression of SOD1 reduced both MEK1/2 and ERK1/2 phosphorylation (Fig. 2E). These data are in agreement with a previously proposed mechanism in which loss of SOD1 activity should reduce H 2 O 2 and as a consequence, protect protein phosphatases from H 2 O 2 -mediated inactivation and promote dampened MAPK pathway activation (30). Thus, our data suggest that abrogated CCS function works via a differential mechanism that is specific to the level of MEK1/2 activity.
CCS-mediated metalation of SOD1 involves three distinct domains of CCS that are each required for SOD1 activity. Domain 1 (D1) facilitates Cu acquisition, domain 2 (D2) mimics SOD1 to promote protein-protein interactions, and domain 3 (D3) may be involved in D1 to D3 Cu transfer while also generating the disulfide bond necessary for SOD1 activation (27,31,32). To definitively test the contribution of CCS D1 and D3 to MAPK signaling, HEK-HT stably expressing nontargeting control or doxycycline-inducible CCS shRNA in the presence of exogenous WT or mutant CCS at the Cu binding interface of D1, D3, or both were generated (Fig. 2F). Based on its inability to acquire Cu regardless of the source, mutation of CCS D1 resulted in reduced P-ERK1/2 (Fig. 2F). Surprisingly, D3 of CCS was also required for basal P-ERK1/2 levels (Fig. 2F), suggesting that a similar mechanism by which CCS activates SOD1 may be used to facilitate Cu-dependent CCS-mediated activation of MEK1/2 (31,32) and is supported by the combined D1/D3 mutant not exhibiting an exacerbated phenotype. Despite our expectations, the knockdown of CCS, which specifically resulted in reduced ERK1/2 phosphorylation (Fig. 2, C and D), had little to no effect on SOD1 function, as measured via in gel activity assays (Fig. S1A). Taken together, these results indicate that CCS is an integral component of MEK1/2 responsiveness to both Cu and growth factors.
Targeting CCS with the small molecule inhibitor DCAC50 reduces MEK1/2 activity A recent study described the development and characterization of a small molecule inhibitor, DCAC50, targeting the Cu chaperones ATOX1 and CCS (33). Mechanistically, DCAC50 binds ATOX1 and CCS with K D of 6.8 ± 1.7 μM and 8.2 ± 2.7 μM, respectively, and prevents Cu transfer between the chaperones and their downstream interacting target proteins (33). Thus, to investigate whether acute inhibition of CCS would decrease MAPK pathway activation in an analogous fashion to genetic deletion of CCS (Fig. 2, B and C), the HEK-HT cells were treated with increasing concentrations of DCAC50. Similar to loss of CCS, P-ERK1/2 was reduced in response to DCAC50 treatment in a dose-dependent fashion (Fig. 3A). In agreement with our findings that genetic ablation of CCS has little impact on SOD1 function, we similarly observed no change in SOD1 activity in the HEK-HT cells treated with increasing concentrations of DCAC50 (Fig. S1B). Further, the kinetics of MAPK pathway activation by CuCl 2 or epidermal growth factor were dampened and delayed in the presence of DCAC50 (Fig. 3, B and C). These results support our findings that CCS is directly required for MAPK signaling by contributing to MEK1/2 Cu binding and suggest another avenue to pharmacologically target aberrant MAPK signaling required for malignant transformation (Fig. 3D).

Discussion
Our previous discovery that Cu selectively regulates the canonical MAPK pathway at the level of the MEK1/2 kinases, in both Drosophila and mammalian cell settings, suggests that there is an evolutionarily conserved pressure for this integration (12). Further, it established a critical mechanistic function for Cu as an intracellular mediator of signaling, which was previously reserved for redox-inactive metals like Zn 2+ , K + , Na + , and Ca 2+ that have well appreciated roles in cell signaling (6). However, the cellular delivery mechanism of Cu to the dual specificity protein kinases MEK1/2 remained to be elucidated.
In support of the hypothesis that kinases within complex protein modules may directly respond to and/or sense metals like Cu, we found that short exposure to CuCl 2 was sufficient to robustly increase ERK1/2 phosphorylation (Figs. 2 and 3). This observation indicates that MEK1/2 are not fully Cu bound in cells and are thus primed to rapidly respond to fluctuations in intracellular Cu levels. The cellular contexts in which labile Cu pools are liberated are under active investigation. Two recent reports provide evidence of this phenomena in which depletion of the intracellular glutathione pools (34) or amino acid deprivation (35) were sufficient to elevate Cu(I) levels. Further, both basal and growth factor stimulated MAPK signaling were diminished when point mutants of the Cu coordinating ligands were introduced, which suggests that Cu is a hardwired component of the pathway that is necessary for the other inputs.
To balance cellular toxicity from free Cu ions with metalation of cuproenzymes, the evolutionarily conserved Cu chaperones promote Cu distribution. Here, we show that the Cu chaperone for superoxide dismutase CCS transiently binds to and facilitates Cu loading into MEK1 and in turn increases MEK1/2 kinase activity (Fig. 1). In agreement with a whole genome RNAi screen in Drosophila S2 cells aimed at identifying modulators of MAPK signaling (36), the knockdown of CCS reduced ERK1/2 phosphorylation (Fig. 2). An intricate molecular mechanism requiring three distinct domains of CCS is required for SOD1 maturation and activation (27,31,32). Interestingly, D3 of CCS, which helps with Cu transfer into the SOD1 active site and generates a disulfide bond in SOD1 (27,31,32), was also necessary for robust MEK1/2 activation (Fig. 2). This finding suggests that the aforementioned Cumediated MEK1 disulfide formation may be installed by CCS in mammalian cells and mechanistically underlie MEK1 activation by Cu. Interestingly, in vitro experiments adding CuCl 2 to purified MEK1 protein resulted in inhibition, not activation, of the enzyme (data not shown), suggesting that Cu loading by CCS is required for activation, potentially to avoid Cu binding to other sites that might inhibit MEK1 activity or to facilitate complete maturation of the Cu bound state. However, the structural and molecular features of MEK1/2 that facilitate a specific interaction with CCS remain elusive. Notable alignment of either structure of sequence of CCS or SOD1 with MEK1 fails to provide an obvious explanation for the recognition of MEK1 by CCS. These preliminary analyses indicate that the similarity between CCS, SOD1, and MEK1 is weak but given the potential of targeting the structural CCS-MEK1 interface with small molecules, this line of investigation warrants future exploration. This result further supports the importance of CCS in controlling MEK1 activation via Cu.
Finally, pharmacologic targeting of CCS with the small molecule inhibitor DCAC50 (33) blunted both basal and stimulated activation of MEK1/2 (Fig. 3). Our previous work demonstrated that knockout of ATOX1 in human BRAF mutation positive melanoma cells also decreased MAPK signaling (37). Although DCAC50 targets both CCS and ATOX1, exogenous CuCl 2 was sufficient to rescue P-ERK1/2, indicating that ATOX1 impacts MAPK pathway activation indirectly of facilitating Cu binding of MEK1/2. The vital importance of intricate mechanisms to dynamically regulate MAPK signaling output is underscored by its dysregulation in approximately 85% of the human cancers (21,22). Although targeting the protein kinase catalytic activity of MEK1/2 is approved in the setting of BRAF V600E metastatic melanoma (22,38) and is under investigation in several other cancers, resistance mechanisms limited clinical durability. Therefore, our structural and molecular studies provide a framework for the development of MEK1/2 inhibitors that exploit its Cu dependence by targeting the binding interface or intracellular delivery components.

Experimental procedures
Cell lines HEK-HT cells were previously described (39) and provided by C.M. Counter (Duke University). The HEK-HT cells were maintained in the Dulbecco's Modified Eagle Media (DMEM, Gibco) supplemented with 10% v/v fetal bovine serum (FBS, GE Lifesciences) and 1% penicillin-streptomycin (P/S, Gibco). The HEK-HT cells were stably infected with lentiviruses derived from the pSMARTvector inducible lentiviral shRNA plasmids (Dharmacon, see Plasmids below) or retroviruses derived from the pWZL retroviral plasmid (see Plasmids below) described below using established methods.

Protein expression and purification
Proteins used for in vitro kinase assays GST-ERK2 K54R -GST-ERK2 K54R was expressed in BL21 (DE3) RIL cells in TB at 37 C until reaching an A 600 of 0.700 and then induced with 1 mM IPTG overnight at 17 C. The cells were then spun down the next day and lysed in lysis buffer 1 (25 mM Tris pH 7.5, 500 mM NaCl, and 5 mM βmercaptoethanol [BME]) treated with 1 mM PMSF and DNAseI. The lysate was then spun down at 19,000 rpm for 40 min, and the supernatant was added to 10 ml of glutathione agarose resin (Pierce, 16102BID) previously washed with lysis buffer 1 and left to incubate at 4 C for 2 h. The supernatant was eluted via gravity column, and the resin was then washed with 1 L of lysis buffer 1. The protein was then eluted with lysis buffer 1 supplemented with 20 mM glutathione (Sigma Aldrich # G4521), and the elution fractions were pooled together and dialyzed into dialysis buffer 1 (25 mM Tris pH 7.5, 150 mM NaCl, and 5 mM BME). The protein was then concentrated and applied to a Superdex S200 10/300 gel filtration column in a final buffer of 25 mM Tris pH 7.5, 150 mM NaCl, and 5 mM BME.The protein was concentrated to 2.4 mg/ml (35 μM), flash frozen in liquid nitrogen, and stored in −80 C freezer for future use.
6XHIS-ATOX1-6XHIS-ATOX1 was expressed in BL21 (DE3) RIL cells in LB (Millipore) at 37 C until reaching an A 600 of 0.700 and then induced with 1 mM IPTG overnight at 17 C. The cells were then spun down the next day and lysed in lysis buffer 1 (25 mM Tris pH 7.5, 500 mM NaCl, and 5 mM BME) treated with 1 mM PMSF and DNAseI. The lysate was then spun down at 19,000 rpm for 40 min, and the supernatant was added to 10 ml of HisPur nickel-nitrilotriacetic acid (Ni-NTA) resin previously washed with lysis buffer 1 and left to incubate at 4 C for 1 h. The supernatant was eluted via gravity column, and the resin was then washed with 1 L of lysis buffer 1 treated with 20 mM imidazole. The protein was then eluted with lysis buffer 1 supplemented with 300 mM imidazole, and the elution fractions were pooled together and treated with TEV protease overnight while dialyzing into dialysis buffer 1 (25 mM Tris pH 7.5, 20 mM NaCl, 5 mM BME). The eluent is then applied to a HiTrap Q HP anion exchange 5 ml column and eluted over a gradient of 20 column volumes ranging from 0% to 100% of buffer B (25 mM Tris pH 7.5, 1 M NaCl, and 5 mM BME). The peak fractions were run on an SDS-PAGE gel, pooled, concentrated, and applied to 10 ml of fresh Ni-NTA resin washed with reverse nickel buffer 1 (25 mM Tris pH 7.5, 250 mM NaCl, 20 mM imidazole, and 5 mM BME). The protein was allowed to pass over the nickel resin three times and then washed extensively with reverse nickel buffer 1 until no more protein was washed off the resin as monitored by Bradford Reagent. The protein was then concentrated and added to a Superdex S200 10/300 gel filtration column in a final buffer of 20 mM Hepes pH 7.0, 150 mM NaCl, and 5 mM BME. The protein was concentrated to 6 mg/ml (650 μM) and dialyzed into dialysis buffer 2 (20 mM Hepes pH 7.0, 150 mM NaCl, and 3.25 mM DTT) overnight. The next morning, 1M CuCl 2 was added to 650 μM of ATOX1 to a final concentration of 585 μM of CuCl 2 and mixed until dissolved. The protein loaded with Cu was then flash frozen in liquid nitrogen and stored in −80 C freezer for future use.
6XHIS-CCS-6XHIS-CCS was expressed in BL21 (DE3) RIL cells in LB at 37 C until reaching an A 600 of 0.700 and then induced with 1 mM IPTG overnight at 17 C. The cells were then spun down the next day and lysed in lysis buffer 1 (25 mM Tris pH 7.5, 500 mM NaCl, and 5 mM BME) treated with 1 mM PMSF and DNAseI. The lysate was then spun down at 19,000 rpm for 40 min, and the supernatant was added to 10 ml of HisPur Ni-NTA resin previously washed with lysis buffer 1 and left to incubate at 4 C for 1 h. The supernatant was eluted via gravity column, and the resin was then washed with 1 L of lysis buffer 1 treated with 20 mM imidazole. The protein was then eluted with lysis buffer 1 supplemented with 300 mM imidazole, and the elution fractions were pooled together and treated with TEV protease overnight while dialyzing into dialysis buffer 1 (25 mM Tris pH 7.5, 20 mM NaCl, and 5 mM BME). The eluent is then applied to a HiTrap Q HP anion exchange 5 ml column and eluted over a gradient of 20 column volumes ranging from 0% to 100% of buffer B (25 mM Tris pH 7.5, 1 M NaCl, and 5 mM BME). The peak fractions were run on an SDS-PAGE gel, pooled, concentrated, and applied to 10 ml of fresh Ni-NTA resin washed with reverse nickel buffer 1 (25 mM Tris pH 7.5, 250 mM NaCl, 20 mM imidazole, and 5 mM BME). The protein was allowed to pass over the nickel resin three times and then washed extensively with reverse nickel buffer 1 until no more protein was washed off the resin as monitored by Bradford Reagent. The protein was then concentrated and added to a Superdex S200 10/300 gel filtration column in a final buffer of 20 mM Hepes pH 7.0, 150 mM NaCl, and 5 mM BME. The protein was concentrated to 20 mg/ml (650 μM) and dialyzed into dialysis buffer 2 (20 mM Hepes pH 7.0, 150 mM NaCl, and 3.25 mM DTT) overnight. The next morning, 1M CuCl 2 was added to 650 μM of CCS to a final concentration of 585 μM of CuCl 2 and mixed until dissolved. The protein loaded with Cu was then flash frozen in liquid nitrogen and stored in −80 C freezer for future use.
Proteins used for surface plasmon resonance T7-HIS-ATOX1-T7-HIS-ATOX1 was expressed in BL21 (DE3) cells in LB with carbenicillin (100 μg/ml, Sigma Aldrich) and when reached the exponential growth, IPTG (Sigma Aldrich) was added. Protease inhibition cocktail (Roche) was added to prevent protein degradation and the cells were lysed by sonication. Nucleic acids digestion was performed with universal nuclease (Pierce). The proteins were then purified using an ÄKTA Purifier (GE Healthcare). First, the protein was purified with a HisTrap FF Ni-NTA column (GE Healthcare), using a gradient of imidazole (5 mM-1 M). Second, further purification was done with a Q Sepharose Fast Flow column (GE Healthcare) and a gradient of NaCl (50 mM-500 mM). Finally, the last purification step was size exclusion using a HiLoad 16/600 Superdex 75 pg column (GE Healthcare). The protein was concentrated with a 3 kDa Amicon Ultra Centrifugal filter (Sigma-Aldrich).
HIS-CCS-HIS-CCS was purchased from Protein Expression Platform at Umeå University.
Proteins used for electrophoretic mobility shift assays (EMSAs) and size exclusion chromatography (SEC) 8XHIS-CCS and 6XHIS-MEK1-8XHIS-CCS and 6XHIS-MEK1 proteins were expressed in BL21 (DE3) pLysS cells in LB media at 37 C until reaching an A 600 of 0.6 to 0.8 and then induced with 1 mM IPTG for an additional 4 h before being harvested. 8XHIS-CCS protein was purified using a HisTrap HP Ni affinity column (Amersham Biosciences) using buffer A (20 mM Tris, pH 8, 300 mM NaCl, and 2 mM DTT) and buffer B (20 mM Tris, pH 8, 300 mM NaCl, 2 mM TCEP, and 1M imidazole). The column was washed with 2% buffer B for 10 bed volume and CCS eluted with a gradient from 2% to 100% in 80 ml. The 8XHIS-tag was removed from the proteins using TEV protease produced in-house and engineered to contain its own noncleavable 8XHIS-tag. After digestion overnight at room temperature, the cleaved HIS-tag and TEV protease were removed from the sample by another pass through the nickel column. The 6XHIS-MEK1 proteins were purified using buffer A (50 mM Hepes, pH 7.5, 150 mM NaCl, and 1 mM TCEP) and buffer B (50 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM TCEP, and 1 M imidazole), as described above.

Surface plasmon resonance
The interaction between MEK1 and HIS-ATOX1 or HIS-CCS was examined using a Biacore X100 surface plasmon resonance instrument (GE Healthcare). MEK1 was covalently coupled to a CM5 chip (GE Healthcare) using the amine coupling kit according to manufacturer's instructions (GE healthcare). Briefly, the surface was activated using 0.05 M N-hydroxysuccinimide and 0.4 M 1-ethyl-3-(-3dimethylaminopropyl) carbodiimide hydrochloride before the injection of MEK1 (20 μg/ml in sodium acetate buffer, pH 4.5) until 2000 RU (2 ng/mm2) was reached and, finally, 1M ethanolamine was injected for deactivation of the surface. The interaction was measured at 25 C in HBS-P+ running buffer containing 10 mM Hepes, 150 mM NaCl, and 0.001% P20 detergent at pH 7.4 (GE Healthcare). Increasing concentrations (2x increase for every injection) of HIS-ATOX1 or HIS-CCS were flown over the surface in a single-cycle sequence without regeneration steps between injections. At the end of the cycle, regeneration was done by injection of 50 mM NaOH (GE Healthcare). The samples of HIS-ATOX1 and HIS-CCS were prepared with DTT (Sigma Aldrich) with a molar ratio of 5:1 (DTT:protein) and for experiments conducted with Cu (CuCl 2 from Sigma Aldrich), with a molar ratio of 0.9:1 for CuCl 2 :protein. The dissociation constant, K d , was determined with the BIAevaluation software using the 1:1 binding model. To confirm that the HIS-tag was not significant for the binding, an equivalent experiment using HIS-PARVALBUMIN was performed.

Electrophoretic mobility shift assays
Metal content of proteins was determined by ICP-MS before experimental set-up. Binding assays were performed in triplicate at pH 7.5 in 50 mM Hepes, 150 mM NaCl, and 1 mM TCEP. The assays were conducted using apo-CCS and Cu(I)-CCS with CCS in excess and MEK1 constant. Experimental concentrations were chosen based on the affinity determined previously by SPR at 8 μM. Protein reactions were allowed to incubate 20 min at room temperature before native gel analysis. 10% native gels were loaded with 10ul of each reaction with 2ul of 50% glycerol added. The gels were run at 150 V for 60 min and then stained with Coomassie blue for 20 min. The gels were de-stained overnight with agitation and visualized using the ChemiDoc MP system (BioRad).

In vitro kinase assay
The activity and activation of MEK1 WT by Cu chaperones ATOX1 or CCS were assessed using an ELISA assay adapted from a previously established assay (42). Briefly, GST-ERK2 K54R fusion protein was diluted to a final concentration of 5.33 μM in Tris-buffered saline treated with 0.05% Tween-20 (TBST) and added to a glutathione-coated 96 well plate (Pierce, #15240) and incubated at room temperature for 1 h with shaking. Purified and untagged full-length MEK1 WT and mutants were diluted to 400 nM concentrations in kinase buffer 1 (50 mM Hepes pH 7.0 and 50 mM NaCl), and 3 μl of various concentrations of CCS were added to 100 μl of diluted MEK in a 96-well "V" bottom plate (Corning, #2897) to final concentrations ranging from 1.25 μM to 10 μM. The MEK1/ ATOX1 or MEK1/CCS mixture was then incubated for 1 h at room temperature. The glutathione-coated plates were then washed extensively with TBST, and 50 μl of the MEK1 ATOX or MEK1/CCS mixture was added to the ERK2 K54R -bound plates along with 50 μl of 200 μM ATP in ATP dilution buffer (50 mM Hepes pH 7.0, 200 mM NaCl, and 20 mM MgCl 2 ) bring the final reaction concentrations to 200 nM MEK and 100 μM ATP. The plate was then mixed at room temperature for 5 min and then left to incubate in a 37 C incubator for 30 min. The reaction was then washed from the plate, and the plate was extensively washed with TBST to quench the reaction. A 1:5000 dilution of rabbit anti-phospho(Thr202/ Tyr204)-ERK1/2 (Cell Signaling Technology (CST), 9101) in TBST treated with 0.5% BSA was added to the plate and incubated for 1 h at room temperature with shaking. The plate was then treated to multiple TBST washes and then incubated with a 1:5000 dilution of secondary antibody (goat anti-rabbit IgG (H + L)-HRP (BioRad)) in TBST treated with 0.5% BSA for 1 h with shaking at room temperature. The plate was again washed with multiple TBST washes, and Supersignal ELISA Pico Chemiluminescent Substrate (Pierce, #37069) was added. The plate was read on a Promega GloMAX 96 Microplate Luminometer. Statistical analysis of P-ERK2 luminescence units was analyzed using a one-way ANOVA followed by a Dunnett's multi-comparisons test or a two-way ANOVA followed by a Tukey's multi-comparisons test in Prism 7 (GraphPad).
Size exclusion chromatography SEC of MEK1 and CCS were conducted to determine elution volume using Superdex 75 Increase 10/300 Gl (GE Healthcare). Cu(I)-CCS was mixed in equal ratio with MEK1 in the presence or absence of 5:1 M ratio BCS( Sigma Aldrich # 146625) to protein and the mixture subjected to SEC. The flow rate of each purification was 0.5 ml/min using 50 mM Hepes, pH 7.5, 150 mM NaCl, and 1 mM TCEP buffer. ICP-MS determined the metal content of proteins before experimental set-up. SOD1 activity assays SOD1 activity was visualized by an in situ gel assay through staining with nitro blue tetrazolium, as previously described (29,43).

Data availability
All data are contained within the article. The raw data for these studies is available from the senior author at bradyd@ pennmedicine.upenn.edu.
Supporting information-This article contains supporting information.