Intestinal Mucin Is a Chaperone of Multivalent Copper

Mucus protects the body by many mechanisms, but a role in managing toxic transition metals was not previously known. Here we report that secreted mucins, the major mucus glycoproteins coating the respiratory and intestinal epithelia, are specific copper-binding proteins. Most remarkably, the intestinal mucin, MUC2, has two juxtaposed copper binding sites, one that accommodates Cu2+ and the other Cu1+, which can be formed in situ by reduction with vitamin C. Copper is an essential trace metal because it is a cofactor for a variety of enzymes catalyzing electron transfer reactions, but copper damages macromolecules when unregulated. We observed that MUC2 protects against copper toxicity while permitting nutritional uptake into cells. These findings introduce mucins, produced in massive quantities to guard extensive mucosal surfaces, as extracellular copper chaperones and potentially important players in physiological copper homeostasis.


Met326 (
). Furthermore, mutation of methionines lowered the affinity for Cu 1+ as measured by competitive titrations ( Figure S5). Muc5b, which naturally lacks the methionines, was not stabilized by Cu 1+ ( Figure S4D). Together, the experiments described thus far provide structural and biochemical evidence for a two-tiered copper binding environment in MUC2, in which histidines dominate the Cu 2+ coordination and nearby methionines capture Cu 1+ .

MUC2 Blocks ROS Formation and Protects Colon Cells from Copper Toxicity
As shown, Cu 2+ bound to MUC2 D1 was reduced by ascorbate ( Figure 3A). However, unlike copper in solution, which catalyzes the consumption of oxygen and depletion of the antioxidant (Buettner & Jurkiewicz, 1996), copper bound to MUC2 D1 did not engage in futile cycling of electrons ( Figure 4A). This result indicates that MUC2 D1 stabilizes Cu 1+ even in the presence of oxygen in a manner that prevents accumulation of reactive oxygen species.
Based on this finding, we hypothesized that MUC2 might protect intestinal cells from toxic copper concentrations. E-cadherin labeling of Caco-2 human colorectal adenocarcinoma cells was used to report on the integrity of cell-cell contact sites and the health of the monolayer. Addition of 250 µM Cu 2+ in serum-free medium resulted in dramatic loss of the organized networks of E-cadherin seen in control cultures ( Figure 4B).

MUC2 Permits Copper Uptake into Cells
The ability of MUC2 D1 to neutralize the toxicity of excess copper led us to ask whether D1 completely sequesters extracellular copper, or rather allows bioavailable copper uptake for beneficial physiological processes. To address this question, we investigated whether nutritional levels of copper supplied with D1 can be taken up by copper-starved cells. Caco-2 cells were deprived of copper until a severe drop was seen in the level of the mitochondrial electron transport chain component cytochrome c oxidase subunit 1 (COX1), which requires copper for function (Getz et al., 2011). Limited amounts of copper, 0.1 or 1 µM, were then re-supplied, either alone or in the presence of stoichiometric or excess MUC2 D1, and the recovery of COX1 levels was monitored. Recovery proceeded even with a five-fold excess of D1 over copper, in a concentration range substantially over the dissociation constant, suggesting that D1 can release needed copper to cells (Figures 5A and S6). Hinting at a mechanism for Cu 1+ transfer, the Cu 1+ ligands in MUC2 D1, though buried within the protein, are adjacent to a small interior tunnel with two portals at the protein surface ( Figure 5B). Together these findings show that MUC2 guards cells from the toxicity of excess copper without blocking supply of the low levels of copper required to maintain physiological processes.

DISCUSSION
Mucins are multifunctional molecules that promote mucosal health by restricting access of pathogens to the epithelial cell surface, storing factors that participate in innate immunity, and managing the microbiome (McGuckin et al., 2011;Johansson & Hansson, 2016;Chairatana & Nolan, 2017;Benam et al., 2018;Bankole et al., 2021). Moreover, secreted mucins were hypothesized to act not only as a protective barrier but also as regulators of small molecule access to membrane-embedded receptors and transporters (Strous & Dekker, 1992). However, investigating the molecular mechanisms of mucin function has long been limited by the lack of high-resolution mucin structures. Remarkably, a specific role for mucins in managing copper was revealed when structural information became available to inspire and inform the experiments reported in this work.
The studies presented here show that mucin D1 assemblies are specific for Cu 2+ over a variety of other metals tested. Weak binding to zinc was also detected, but no evidence could be obtained from crystallography that this ion binds at the Cu 2+ site. The specialization of the mucin D1 assembly for copper binding contrasts with a separate, conserved calcium binding site present in all three D assemblies in the amino-terminal region of MUC2 (Javitt et al., 2020), as well in the three homologous D assemblies of the related blood clotting glycoprotein von Willebrand factor (Dong et al., 2019;Javitt et al., under revision). Calcium was not observed in the crystals of Cu 2+ -bound MUC2 D1, perhaps because it was stripped by the citrate present in the crystallization solution, though MUC2 D3 was also crystallized with citrate and retained its calcium (Javitt et al., 2019).
All solution measurements of Cu 2+ binding were likely performed on Ca 2+ -loaded MUC2 D1, as addition of Ca 2+ had no effect on the protein in MST (Figure S1B) or DSF ( Figure   S4A) experiments.
Copper and calcium binding occur in different manners in MUC2. Ca 2+ is bound locally within the VWD domains of the D assemblies, whereas Cu 2+ is bound at the interface between three separate domains in D1: VWD1, C8-1, and TIL-1 (Figure 2A). One Cu 2+ ligand is provided by each of these domains, and the fourth ligand is supplied by the flexible region at the amino terminus of the protein, just upstream of the VWD1 domain ( Figure 2B). The Cu 1+ ligands, in turn, are provided by the VWD1 and TIL-1 domains, without the participation of C8-1. The binding of copper between multiple domains suggests that affinity for copper is tunable by factors or events, such as interactions with other proteins, that may affect the relative orientations of those domains. Indeed, the TIL-1 and E-1 domains are flexible relative to the VWD1 and C8-1 domains, affecting the availability of the key copper ligand H324 ( Figure 2C).
With detailed structural and biochemical information now available, a physiological role for copper binding by mucins can be considered. Just as mucins themselves have many functions, copper binding by mucins may serve various purposes. MUC2 is expressed at high levels in all parts of the intestine, including both aerobic and anaerobic environments and regions with different digestive, absorptive, and excretive roles (Audie et al., 1993;Paone & Cani, 2020). In addition to protecting the intestines and other mucosal surfaces against copper toxicity, copper binding by mucins may facilitate the conversion of dietary copper (largely Cu 2+ ) to the form transported across the enterocyte membrane (thought to be Cu 1+ ). The mechanisms by which dietary copper passes from the lumen of the digestive tract into enterocytes are debated (Zimnicka et al., 2007;Nose et al., 2010;Pierson et al., 2019). Hypothetically, MUC2 may interface structurally or functionally with the transmembrane copper transporter Ctr1, which uses methionines to coordinate Cu 1+ in its pore (Ren et al., 2019), or with other proteins putatively involved in intestinal copper uptake (Nose et al., 2006). Mucins may also participate in nutritional immunity (Lopez & Skaar, 2018) by depriving pathogens of copper, and MUC2 may cooperate with bile (Linder, 2020) for the safe excretion of excess copper. These activities need not be mutually 8 exclusive. Further exploration of these various possibilities will help integrate mucins into models of physiological copper management (Shanbhag et al., 2021).
It is well known that copper is carefully regulated in biology. Transfer of copper within eukaryotic cells and in the bacterial periplasm is done by direct protein-protein contacts and ligand exchange rather than by release of copper into solution (Robinson & Winge, 2010;Nevitt et al., 2012). The players and mechanisms of copper chaperoning inside cells and in blood are relatively well understood (Nevitt et al., 2012;Linder, 2016;Magistrato et al., 2019), but how the body manages copper before it reaches those sites has been obscure. The lung and intestinal mucosa are the largest, most important, and most vulnerable of exposed physiological surfaces, so it is reasonable that a mechanism would have evolved for restraining an essential but toxic element in these environments. The discovery that multiple gel-forming mucins are copper binding proteins, with an elaborate two-tiered site for distinct copper redox states in intestinal mucin ( Figure 6), introduces intriguing new players into the field of copper regulation and utilization in the human body.

DATA AND CODE AVAILABILITY
Coordinates and structure factors for MUC D1 bound to Cu 2+ have been deposited in the Protein Data Bank (PDB ID: 7PRL).

ACKNOWLEDGMENTS
We thank Prof. Byung-Eun Kim and Dr. David Gokhman for helpful suggestions.

AUTHOR CONTRIBUTIONS
DF, KF, and NR conceptualized and designed research. NR, ADG, KR, YF-S, GJ, NAN, KC, KWR, and DF carried out experiments. YF-S and TI provided methodology and expertise, NR, ADG, KWR, KC, and DF analyzed data. DF wrote the original manuscript draft, and DF, KF, and NR reviewed and edited the manuscript with input from all authors.

DECLARATION OF INTERESTS
The authors declare no competing interests. (D) Sequences of human gel-forming mucins and the related blood clotting protein von Willebrand factor (VWF). MUC5B and MUC5AC are gel-forming mucins in airway mucus (Benam et al., 2018).

Figure 2. MUC2 Binds Cu 2+
(A) MUC2 D1 Cu 2+ -bound crystal structure. A backbone cartoon is shown within a semitransparent surface, and the dark red sphere is the copper ion. Disulfide bonds are yellow.
The His32 side chain was disordered in the crystal.
(C) Comparison of the MUC2 D1 crystal structure with D1 in the cryo-EM structure of the filamentous assembly intermediate ( Figure 1B). A cartoon representation of one D1 assembly is colored magenta within a semi-transparent surface representation of a unit in the filament ( Figure 1B). The crystal structure of MUC2 D1 bound to Cu 2+ is superposed in dark teal. The D1 assembly is rotated by about 180° compared to panel A. Calcium ions bound at specific acidic motifs found in mucins and the related von Willebrand factor are indicated (Dong et al., 2019;Javitt et al., 2019). The zoomed view below shows how movement of the TIL1 domain brings His324 closer for Cu 2+ coordination.
(D) Illustration of the Cu 2+ -coordination site with ligand-metal distances obtained from EXAFS. 12 (E) Cu 2+ binding increased the midpoint of thermal denaturation of MUC2 D1 by about 6 °C (black bar), measured using DSF at pH 5.7. The differences between the denaturation temperature with and without Cu 2+ (ΔTm) are shown for wild type (WT) and the indicated mutants, colored according to amino acid type mutated. His34 was mutated also to arginine because the UniProt entry Q02817 contains arginine at this position. The label "-2His" refers to a double H277A/H324A mutant. The label "-3Met" refers to a triple M146L/M154L/M326V mutant. Errors are standard deviation, n=3.

Figure 3. Reduced Copper is Transferred to a Nearby Methionine Cluster in MUC2
(A) The indicated additions were made to MUC2 D1. Buffer was rapidly exchanged on a size exclusion column (SEC) to remove excess reductant or unbound copper, and Cu 1+ retained by 20 µM protein was detected using BCA.
(B) Fo-Fc difference map density, calculated using the diffraction data from ascorbatetreated MUC2 D1-Cu 1+ crystals phased using MUC2 D1 coordinates from which the Cu 2+ ion had been removed. Map is displayed at 4.5σ (dark blue). For comparison, Fo-Fc map density of the Cu 2+ crystal calculated using phases from the same model is superposed and displayed at 8σ. The ribbon diagram was made using the coordinates of the Cu 2+ -bound form without displaying the Cu 2+ .  (B) Caco-2 colon cells were subjected to the indicated treatments in serum-free medium and then stained with DAPI (nuclei, blue) and labeled for E-cadherin (green). Cultures treated with a high copper concentration (250 µM) in the presence of equimolar wild-type MUC2 D1, but not of a H324A/M326V mutant, retained a normal appearance. Scale bar is 100 µm.

Figure 5. MUC2 D1 can release copper to cells
(A) Supplying copper bound to MUC2 D1 restored COX1 levels in copper-starved Caco-2 cells. "SF rec" indicates that only serum-free medium was supplied during the recovery period, whereas the "no rec" sample was harvested without recovery. Dashed gray lines separate sets of samples with higher and lower copper and D1 concentrations during recovery. See also Figure S6. (B) Met326 is exposed to the exterior of the protein through two small portals. A surface representation of the Cu 2+ -bound MUC2 D1 crystal structure is shown with sulfur atoms colored yellow.

Figure 6. Summary of Copper Chaperoning by MUC2
A schematic illustration of the MUC2 two-tiered copper binding site, in which a histidinerich site (His site) binds captures Cu 2+ and a methionine-rich site (Met site) captures Cu 1+ .
MUC2-bound Cu 2+ can be reduced to Cu 1+ by vitamin C (ascorbate) or other dietary antioxidants. Bound Cu 1+ is protected by MUC2 from oxidation in aerobic environments, but copper, putatively in the form of Cu 1+ , can be released by MUC2 for nutritional delivery to cells (dashed arrow).

LEAD CONTACT AND MATERIALS AVAILABILITY
Requests for further information and or reagents may be addressed to the Lead Contact, Deborah Fass (deborah.fass@weizmann.ac.il). The primary plasmids used in this study were deposited to Addgene (identifiers: xxx).

Caco-2 cell line
The human colon epithelial cells line Caco-2 was a kind gift from the laboratory of Prof.

Protein production and purification
The MUC2 D1 coding sequence (NCBI reference sequence NP_002448.4, amino acids 21-389) was cloned into the pCDNA3.1 vector following the signal sequence of the enzyme QSOX1. A His6 tag was appended to the carboxy terminus. The plasmid was transiently transfected into HEK293F cells (ThermoFisher) using the PEI Max reagent (Polysciences Inc.) with a 1:3 ratio (w/w) of DNA to PEI at a concentration of 1 million cells per ml.
Cells were maintained in FreeStyle 293 medium. To facilitate subsequent crystallization, 5 µM kifunensine (Cayman Chemical) was added at the time of transfection to obtain protein containing high-mannose, EndoH-cleavable glycans. Six days after transfection, cells were removed from the cultures by centrifugation for 10 min at 500g. The culture medium was then further clarified by centrifugation for 30 min at 9500g and filtration through a 0.45 µm pore-size membrane. The D1 assembly was purified from the medium by nickel-nitrilotriacetic acid (Ni-NTA) chromatography. Purified protein was exchanged into 25 mM HEPES, pH 7.5, 250 mM NaCl and concentrated to 10 mg/ml (240 µM).
For experiments aside from crystallization, the MUC2 D1 expression construct was modified to contain a His6 tag and tobacco etch virus (TEV) cleavage site following the signal sequence, and the His6 tag at the carboxy terminus was eliminated. Protein was produced without kifunensine. After Ni-NTA purification, protein was subjected to cleavage with TEV protease containing a His6 tag. The cleaved amino terminal tag and the TEV protease were subsequently removed using Ni-NTA beads. The protein was concentrated in 10 mM MOPS buffer, pH 7.0, 100 mM NaCl. Glycerol was added to 10%, and aliquots were stored at -80 °C until use. An expression plasmid for Muc5b D1 (UniProt E9Q5I3 amino acids 50-426) was prepared with a TEV cleavage site followed by a His6 tag at the carboxy terminus. Muc5b D1 was produced and purified as for MUC2 D1, subjected to TEV cleavage, and quantified (ε280nm for Muc5b D1 including the remaining segment of the TEV site = 50,360 M -1 cm -1 ).

Crystallization and structure solution
The MUC2 D1 assembly containing a carboxy-terminal His6 tag was crystallized using the hanging drop method over a well solution containing 30 mM MgCl2, 8% polyethylene glycol 3350, 100 mM citrate buffer, pH 5.4, and 10% glycerol. The protein stock solution for crystallization was supplemented with 480 µM copper sulfate. Drops were prepared by mixing 2 µl protein with 1 µl well solution. Data-quality crystals grew slowly over the course of a few months. About 10 min prior to flash freezing, crystals were transferred to a solution containing all components of the well except that the glycerol concentration was increased to 20%. For reduction of Cu 2+ , 5 mM ascorbic acid (Sigma) was included in this soak solution. Data were collected at the European Synchrotron Radiation Facility (ESRF) beamline ID23-1 at 100 K. The wavelength was 14.2 keV (0.8731 Å). Initial phases were provided by molecular replacement using the D1 assembly from the MUC2 aminoterminus cryo-EM structure (Javitt et al., 2020), and the structure model was improved by cycles of rebuilding using Coot (Elmsley et al., 2010) and refinement using Phenix (Adams et al., 2010). No Ramachandran outliers were present in the refined structure model, and 95.5% of the amino acids were in favored regions of Ramachandran space. Though the D1 assembly was previously seen to bind calcium in the cryo-EM structure (Javitt et al., 2020), no calcium was detected in the D1 crystal structure, most likely due to the high concentration of citrate in the crystallization buffer.

Isothermal titration calorimetry (ITC)
MUC2 D1 was dialyzed against 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer, pH 7.4, containing 150 mM NaCl and 2 mM, 10 mM, or 30 mM glycine. These conditions were based on another study using glycine as a competitor for copper to enable accurate ITC measurements of high-affinity protein-copper interaction (Trapaidze et al., 2012). After dialysis, protein was diluted to a final concentration of 50 µM. CuSO4 was prepared at 500 µM in the same dialysis buffer as the corresponding protein. Experiments were conducted at 25 °C on a Malvern MicroCal PEAQ-ITC system starting with a single 0.4 µl injection of Cu 2+ solution into protein solution, followed by 18 2-µl injections, with 150 seconds spacing between each injection and continuous stirring at 750 rpm. For each glycine concentration, a blank run titrating the Cu 2+ into buffer was done and subtracted accordingly. The titration data, including the apparent KD, were analyzed using software provided by the manufacturer. The conditional KD values were derived according to previous methods (Trapaidze et al., 2012). The experiments were done in triplicate, and representative measurements are displayed in the figures.
The consistent deviation from a 1:1 stoichiometry in fits to the ITC data (Extended Data Table 2) is not yet explained. X-ray fluorescence counts at the copper edge showed that purified MUC2 D1 was not pre-loaded with Cu 2+ , and measurements after addition of Cu 2+ suggested stoichiometric binding. Protein concentrations were determined for ITC and EXAFS experiments using the same method and were done with care, such that it is unlikely that large errors in protein concentration determination explain the observations of n = ~0.5.

Differential scanning fluorimetry (DSF)
To compare thermal stabilization upon Cu binding by WT MUC2 D1 and mutants, as well as by Muc5b, DSF was performed using a Prometheus NT.48 (NanoTemper) instrument.
Unless otherwise indicated, proteins were used at a concentration of 7.5 µM in 50 mM 2-(N-morpholino)ethanesulfonic acid (MES) buffer, pH 5.7, 100 mM NaCl, 0.05% Tween-20. When present, Cu 2+ was added in the form of CuSO4 to a concentration of 7.5 µM (1:1 molar ratio) and ascorbic acid to a concentration of 150 µM. Samples were heated from 25 °C to 95 °C at 1°C/min. Tryptophan and tyrosine fluorescence was monitored by recording the 350/330 nm emission ratio after excitation at 280 nm. DSF experiments done at neutral pH showed the same trends; data are reported for mildly acidic pH because the effect of thermal stabilization by Cu 2+ was greater and because the D1 Cu 2+ crystal structure was determined at acidic pH.

Microscale thermophoresis (MST)
MUC2 D1 was used at a concentration of 500 nM in 50 mM 3-(Nmorpholino)propanesulfonic acid (MOPS) buffer, pH 7.0, 100 mM NaCl, 0.05% Tween-20. Protein was incubated with 16 serial dilutions of metal in the range of 50 µM to 1.5 nM for 10 min and centrifuged at 20,000g for 10 min at 4 °C. Tryptophan and tyrosine fluorescence was monitored and recorded using a Monolith LabelFree instrument (NanoTemper). Thermophoresis was analyzed at 70% LED and 20% laser intensity for all samples. MUC2 measurements were done in triplicate. Initial fluorescence was equal in all capillaries before thermophoresis measurements. Muc5b was measured and analyzed in the same manner.

X-ray absorption spectroscopy
A solution of 800 µM MUC2 D1 in 50 mM MES buffer, pH 5.7, 100 mM NaCl was incubated with 0.8:1 equivalents of CuSO4 via syringe pump to ensure no excess metal was present, then was split into two further samples. One sample was treated with excess ascorbate to ensure the cuprous form of the metalloprotein, and both samples were measured as an aqueous glass in 20% ethylene glycol at 10 K. Cu K-edge (8.9 keV) extended X-ray absorption fine structure (EXAFS) and X-ray absorption edge data were collected at the Stanford Synchrotron Radiation Lightsource on beamline 7-3 with an Si 220 monochromator, in fluorescence mode using a high-count rate Canberra 30-element Ge array detector. A Ni filter and a Soller slit were placed in line with the detector to attenuate the elastic scatter peak. For energy calibration, a Cu foil was placed between the second and third ionization chambers. In the case of the Cu 2+ -loaded sample, scans were taken from fresh surfaces of the sample window and their edges compared to ensure no photoreduction occurred, with a shutter in place in between scan points. Six scans of a buffer blank were averaged and subtracted from the raw data to produce a flat pre-edge and remove any residual Ni fluorescence. Data reduction and background subtraction were done using the EXAFSPAK suite, and the data were inspected for dropouts and glitches before averaging. The EXCURVE program was used for spectral simulations.
Where indicated, CuSO4 was supplied at a concentration of 60 µM and, subsequently, ascorbic acid or quercetin was added to a concentration of 1.2 mM. Following a 5-minute incubation, samples were exchanged into 50 mM MES buffer, pH 5.7, 100 mM NaCl using Zeba™ Spin Desalting Columns 7K MWCO (size exclusion) to remove ascorbic acid or quercetin, when present, and any copper not bound to protein. After desalting, samples were diluted by a factor of three (i.e., to 20 µM) into the same buffer containing bicinchoninic acid (BCA) (final BCA concentration 1 mM). Absorbance was measured at 562 nm, and data were converted to Cu 1+ concentration using an extinction coefficient of 7,900 M -1 cm -1 for the Cu 1+ -(BCA)2 complex (Young & Xiao, 2021).

Competition Titrations and Data Fitting
UV/vis absorption spectra were recorded in a 1-cm quartz cuvette on an SI Photonics model 420 fiber optic CCD array spectrophotometer located inside a Siemens MBraun glove box under inert nitrogen atmosphere. Cu 1+ stock solutions were prepared by dissolving PCu + 2 L  P + Cu(L)2 (1) anti-E-cadherin antibody ab231303, 1:100 dilution) for 1 hr in room temperature, followed by an Alexa-488 conjugated secondary antibody. Images were taken using a Olympus IX51 microscope equipped with Olympus XM10 camera.

COX1 Recovery Assay
Caco-2 cells were passaged in DMEM containing 20% fetal bovine serum (FBS), penicillin/streptomycin, and L-glutamine. For copper starvation, medium was supplemented with 500 µM BCS for 10 days. During passaging under copper starvation conditions, cells were plated in parallel in 6 cm dishes. To initiate recovery, cells were washed once with PBS, and additives for testing were supplied in serum-free medium.
After a further 3 days, cells were harvested by trypsinization and pelleted. Cell lysates separate by polyacrylamide gel electrophoresis and analyzed for COX1 levels by western blot (Anti-MTCO1 #ab14705). Bands were quantified using ImageJ software and normalized to the untreated sample for each biological experimental replicate. Figure S1. MUC2 specifically binds copper, related to Figure 2.

SUPPLEMENTAL FIGURES
(A) The crystal structure of the Cu 2+ -bound MUC2 D1 assembly is shown rotated 90° around the x-axis compared to Fig. 2A   have similar sets of angles, and CucA exhibits some quercetin dioxygenase activity (Tottey et al., 2008). As indicated by all angles being close to 110°, the Cu 2+ coordination geometry of S100A12 is approximately tetrahedral. The three histidines of MUC2 D1 are arranged roughly as three corners of a square plane (angles ~90°, ~90°, ~180°), while the glutamate is ~120° off the plane. This geometry is different from both 1JUH/2XLG and 1ODB.   Competitive binding assays of MUC2 D1 with BCA or ferrozine for Cu 1+ binding under anaerobic conditions. Error bars represent standard deviation from the average of triplicate measurements; the lines represent the best fit to the equilibrium expressed in Eq 3 of the methods section. Dissociation constants (KD) obtained from these fits are indicated, except for mutants that did not compete well enough with ferrozine to provide good fits to the data. Caco-2 cells were starved of copper by addition of BCS to the medium where indicated (+) and passaging the cells for 10 days. "No rec" indicates the sample that was then harvested with no recovery. To all other samples, serum-free (SF) medium was supplied with the indicated additives, and cells were harvested 3 days later. When present, proteins were at 5 µM and Cu 2+ was a 1 µM. Cell lysates were subjected to SDS-PAGE and western blotted for COX1. No appreciable differences were seen between addition of copper together with the copper-binding D1 assembly compared to the non-copper-binding assemblies D2 and D3.