|Home | About | Journals | Submit | Contact Us | Français|
The maturation and activation of the anti-oxidant Cu,Zn superoxide dismutase (SOD1) are highly regulated processes that require several post-translational modifications. The maturation of SOD1 is initiated by incorporation of zinc and copper ions followed by disulfide oxidation leading to the formation of enzymatically active homodimers. Our present data indicate that homodimer formation is a regulated final step in SOD1 maturation and implicate the recently characterized copper homeostasis protein COMMD1 in this process. COMMD1 interacts with SOD1, and this interaction requires CCS-mediated copper incorporation into SOD1. COMMD1 does not regulate disulfide oxidation of SOD1 but reduces the level of SOD1 homodimers. RNAi-mediated knockdown of COMMD1 expression results in a significant induction of SOD1 activity and a consequent decrease in superoxide anion concentrations, whereas overexpression of COMMD1 exerts exactly the opposite effects. Here, we identify COMMD1 as a novel protein regulating SOD1 activation and associate COMMD1 function with the production of free radicals.
Cu,Zn superoxide dismutase (SOD1) is a highly conserved and ubiquitously expressed anti-oxidant enzyme involved in the conversion of toxic superoxide anions into molecular oxygen and hydrogen peroxide (1). Active, mature SOD1 is a 32-kDa homodimeric protein containing two zinc (Zn2+) and two copper (Cu2+) ions needed for its stability and activity, respectively (2). The majority of cellular SOD1 is present as a preexisting apo-SOD1 pool, which contains zinc but lacks copper (3, 4). Consequently, apo-SOD1 lacks enzymatic activity. Although the mechanism of zinc incorporation into nascent SOD1 is elusive, its activation by copper is known to be a complex and highly regulated process. SOD1 subunits acquire copper predominantly via direct transient binding to the copper chaperone for SOD1 (CCS) and subsequent transfer of copper from CCS to SOD1 (5,–7). CCS-independent activation of SOD1 has also been described but accounts for a minor fraction of SOD1 activation (8, 9). Incorporation of copper is completed by O2-dependent disulfide isomerization resulting in an intrasubunit disulfide bond between SOD1Cys-57 and SOD1Cys-146, commonly referred to as disulfide-oxidized SOD1 subunits (10, 11). The final step in SOD1 maturation is the formation of enzymatically active homodimers, composed of two oxidized SOD1 subunits (4, 12, 13). Although copper incorporation and disulfide bond formation are tightly regulated by CCS, it is unknown whether the process of SOD1 subunit dimerization is subjected to regulation.
The recently identified copper metabolism Murr1 domain containing 1 (COMMD1) protein is a member of the COMMD protein family. The COMMD family includes 10 proteins sharing a unique sequence of 50–65 amino acids referred to as the COMM domain (14). The 21-kDa COMMD1 protein is considered as the prototype of the COMMD family and is implicated in the regulation of copper homeostasis, termination of NF-κB activity, HIF-1α signaling, and regulation of sodium transport (15,–28). COMMD1 engages in many different protein-protein interactions and acts as a scaffold protein to regulate the turnover of its interaction partners, such as the NF-κB subunit p65, HIF-1α, and the Wilson disease-associated protein ATP7B (15, 18, 19, 27,–31). Expression of COMMD1 itself is regulated on the post-translational level by the X-linked inhibitor of apoptosis (XIAP)3 and the redox sensor protein HSCARG (20, 24, 25). Consistent with the notion that COMMD1 functions in multiple cellular pathways, COMMD1 is ubiquitously expressed and located at different subcellular sites. COMMD1 has been detected in the nucleus (23, 32) and in the cytoplasm (33), where it may be associated with intracellular membranes (29, 33). In this study, we identified COMMD1 as a novel interaction partner of SOD1, and we show that COMMD1 plays an important role in the maturation of SOD1.
The following antibodies were used for immunoprecipitation and immunoblotting: rabbit anti-COMMD1 antiserum (33); monoclonal mouse anti-COMMD1 (Abnova, Taipei City, Taiwan); polyclonal rabbit anti-SOD1 (Santa Cruz Biotechnology, Santa Cruz, CA); polyclonal rabbit anti-CCS (Santa Cruz Biotechnology); polyclonal mouse anti-FLAG M2 HRP-conjugated (Sigma); polyclonal rabbit anti-GST (Santa Cruz Biotechnology); polyclonal rabbit anti-HA (Sigma); and polyclonal rabbit anti-β-tubulin (Abcam plc, Cambridge, UK). Purified SOD1 from bovine liver was obtained from Sigma.
A nontargeting pool of siRNAs (siControl) and siRNA pools targeting hCCS, hCOMMD1, and mCommd1 (targeting sequences described in supplemental Table 1) were purchased as ON-TARGETplus SMARTpools from Dharmacon (Chicago).
Human embryonic kidney 293T cells (HEK293T), human hepatocellular liver carcinoma cells (HepG2), and mouse neuroblastoma Neuro2A (all obtained from ATCC, Manassas, VA) were cultured in high glucose Dulbecco's modified Eagle's medium GlutaMAXTM (4.5 g/liter d-glucose and pyruvate; Invitrogen) supplemented with 10% fetal bovine serum (FBS), l-glutamine, penicillin, and streptomycin at 37 °C in 5% CO2. Monoclonal HEK293T cell lines, stably transfected with pSUPER-RETRO vector (shControl) or a plasmid encoding short hairpin RNA (shRNA) targeting COMMD1 mRNA sequence (shCOMMD1), were described previously (18) and maintained in HEK293T medium supplemented with 1 μg/μl puromycin dihydrochloride (Sigma). In various experiments, cells were incubated overnight with 10–150 μm CuCl2 (Sigma), 150 μm ZnCl2 (Sigma), 10–200 μm bathocuproinedisulfonic acid (BCS; Sigma), 50 μm paraquat dihydrochloride (Sigma), or 50 μm menadione sodium bisulfite (Sigma) or for 1 and 2 h with 0.1 or 1 mm CuCl2 and 50 μm menadione, respectively, prior to lysis.
Cells were transiently transfected by the calcium phosphate precipitation method as described elsewhere (34). Transfection of siRNA ON-TARGETplus SMARTpools was performed by means of Lipofectamine RNAiMAX (Invitrogen), according to the manufacturer's protocol. Cells were harvested 48 h after transfection for biochemical analysis.
pEBB-HA-COMMD1, pEBB-COMMD1-FLAG, and pEBB-COMMD1-GST constructs were described previously (35). pEBB-COMMD1-GST amino acids 1–118 and 119–190 deletion constructs were kindly provided by Dr. E. Burstein (Dallas, TX) (25). Full-length coding sequences of CCS and SOD1 were amplified from human liver cDNA (Invitrogen) by using the designed primers indicated in supplemental Table 2 and subcloned into pEBB-FLAG and pEBB-GST vectors. Mutations in pEBB-CCS-FLAG were introduced by the QuickChange site-directed mutagenesis method (Stratagene, primer sequences indicated in supplemental Table 2). pEBB-CCS-GST deletion constructs were kindly obtained from Dr. C. Duckett (Ann Arbor, MI) and described previously (36). A pSUPER-RETRO vector encoding short hairpin RNAs (shRNAs) against COMMD1 (shCOMMD1) was described previously (15, 18). Sequences were verified by automated sequence analysis.
Precipitation of GST-tagged proteins by means of GSH-Sepharose beads was performed as described previously (20). In short, transiently transfected cells were rinsed once in PBS prior to lysis in lysis buffer A (25 mm HEPES, pH 7.9, 100 mm NaCl, 1 mm NaEDTA, 1% Triton X-100, 10% glycerol) supplemented with 1 mm Na3VO4, 1 mm PMSF, 10 mm DTT, and protease inhibitors (Complete; Roche Applied Science). For lysis of copper-treated, zinc-treated, or BCS-treated cells, lysis buffer A was supplemented with 1 mm CuCl2, ZnCl2, or BCS, respectively (15, 37), or with variable concentrations of CuCl2 or BCS, as indicated in Fig. 2B. Protein concentrations were determined by protein assay (Bradford, Bio-Rad).
Immunoprecipitation of endogenous proteins was performed essentially as described previously (20). Antibody conjugation to protein A beads (Sigma) was performed for 1 h at room temperature. Prior to immunoprecipitation, cell lysates were precleared by incubation with 40 μl of protein A beads.
In all interaction studies, equal amounts of proteins in lysis buffer A were used for precipitation. Input samples represent ~1% of protein amounts used for immunoprecipitation or GST-precipitation.
To study protein stability, cells were treated overnight with 15 μg/ml cycloheximide (CHX, Sigma). Cells were lysed in sample buffer (SB: 2.5% SDS, 20% glycerol, 0.2 m Tris-HCl, pH 6.8, bromphenol blue) supplemented with 2% β-mercaptoethanol (β-ME). Prior to gel loading, protein lysates were boiled at 95 °C and sheared using a 29-gauge needle. In all experiments, SDS-PAGE was followed by protein transfer onto nitrocellulose membranes (Schleicher & Schuell) for immunoblot analysis.
Experiments were performed essentially as described previously (38). Briefly, cells were lysed in lysis buffer B (50 mm HEPES, pH 7.2, 2.5% SDS, 0.1 mm NaEDTA) supplemented with 100 mm 2-iodoacetamide (Sigma) and incubated for 30 min at 37 °C. SDS-PAGE was performed under nonreducing conditions preceding boiling of the samples at 95 °C. After SDS-PAGE and prior to blotting, proteins were subjected to in-gel reduction by shaking in protein transfer buffer containing 2% β-ME. Input samples were prepared in sample buffer supplemented with 2% β-ME. Percentages of reduced SOD1 were quantified using densitometry (GS-700; Bio-Rad).
Visualization of SOD1 monomers and dimers was performed as described previously (39) with slight modifications. In short, cells were rinsed once in PBS prior to lysis in lysis buffer A supplemented with 1 mm Na3VO4, 1 mm PMSF, and protease inhibitors. 30 μg of protein was incubated for 30 min at 37 °C in sample buffer without β-ME and analyzed by SDS-PAGE and immunoblotting under nonreducing conditions. In the case of purified SOD1, samples were preincubated in PBS without sample buffer for 1 h at 37 °C. Samples were not boiled prior to gel loading.
For the in-gel SOD1 activity assay, cells were washed once in PBS prior to lysis in lysis buffer A supplemented with 1 mm Na3VO4, 1 mm PMSF, and protease inhibitors. 50 μg of protein lysates were analyzed by PAGE under native conditions, without boiling of the samples. Gels were incubated for 1 h in the dark on a shaking platform in 50 mm potassium phosphate buffer, pH 7.8, supplemented with 0.34 mm nitro blue tetrazolium chloride (Sigma), 14 mm riboflavin (Sigma), and TEMED (Bio-Rad). Subsequently, yellow-stained gels were transferred into deionized H2O and exposed to bright light to induce a color change by the generation of free radicals, which is inversely correlated to the SOD1 activity (1).
A quantitative SOD1 activity assay, based on the formation of formazan by the tetrazolium salt WST-1 (2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt) upon reduction with a superoxide anion, was performed according to manufacturer's product sheet (Sigma). 1 mm KCN (Sigma) was used as a specific inhibitor of SOD1 activity, and measured activities were corrected for protein concentration.
Cells were incubated with 3 μm dihydroethidium (hydroethidine, DHE; Invitrogen) in PBS for 30 min at 37 °C. Cells were incubated with menadione sodium bisulfite prior to DHE incubation. Cells were washed once with PBS, trypsinized, and immediately analyzed on a FACSCalibur flow cytometer (BD Biosciences; software CellQuest Pro).
The quantitative data in this paper are represented as means ± S.D. Statistical evaluation was made using one-way analysis of variance, followed by the post hoc Bonferroni test. Differences were considered to be significant at p < 0.05.
To investigate a putative role for COMMD1 in regulating SOD1 maturation, the potential binding of COMMD1 to SOD1 was examined. Endogenous COMMD1 was immunoprecipitated from HepG2 cells, followed by immunoblot analysis using SOD1 antibodies. This experiment revealed a specific interaction of endogenous COMMD1 with SOD1 (Fig. 1A). Transient expression of COMMD1-FLAG and SOD1-GST, followed by GST pulldown of the SOD1-GST fusion protein and immunoblot analysis of the presence of COMMD1, confirmed the interaction between these proteins in HEK293T cells (Fig. 1B) and in the neuroblastoma cell line Neuro2A (data not shown). These experiments indicate that COMMD1 interacts with SOD1 and validate the use of GST pulldown analysis for subsequent interaction studies in our cellular model systems.
To examine whether the interaction between COMMD1 and SOD1 was dependent on copper, GST pulldown assays were performed after incubation of HEK293T cells with CuCl2 or the copper chelator BCS, respectively. These studies showed that the interaction was markedly enhanced upon addition of copper, whereas copper depletion decreased the binding to almost undetectable amounts (Fig. 1C). Incubation with various concentrations of CuCl2 or BCS confirmed the copper dependence of the interaction, as binding of COMMD1 to SOD1 was enhanced by copper in a concentration-dependent manner (Fig. 1D). In contrast to copper, addition of ZnCl2 had no significant effect on the interaction (Fig. 1E).
Because we were unable to identify a direct interaction between recombinant purified SOD1 and COMMD1 (data not shown) and because SOD1 acquires its copper predominantly via its metallochaperone CCS (5, 6), we determined whether the interaction between COMMD1 and SOD1 was affected by siRNA-mediated depletion of CCS protein expression in HEK293T cells. GST pulldown analysis revealed a reduced interaction between COMMD1 and SOD1 after CCS depletion, which could be partially restored by addition of excess copper (Fig. 2A). Similar to the effects shown in Fig. 1, C and D, copper depletion resulted in a diminished interaction of COMMD1 with SOD1, and this was not further inhibited by depletion of CCS. Conversely, overexpression of CCS enhanced the interaction between COMMD1 and SOD1 in a CCS concentration-dependent manner (Fig. 2B). Incubation of cells with high copper concentrations in addition to CCS overexpression did not result in further enhancement of binding between COMMD1 and SOD1. This suggests that the copper-dependent binding of COMMD1 to SOD1 was predominantly mediated by CCS.
To assess if this effect of CCS on the COMMD1-SOD1 interaction could be attributed to its function in copper incorporation into SOD1, a CCSC244A/C246A mutant was used. In this mutant, the CXC copper-binding motif located in CCS domain III, which is essential for the metal transfer activity of CCS toward SOD1 (supplemental Fig. S1A) (40, 41), was mutated to AXA, resulting in a failure to load SOD1 with copper (see Refs. 40, 42 and confirmed by our own data not shown). Although this CCSC244A/C246A mutant interacted with SOD1 (supplemental Fig. S1B), CCSC244A/C246A could not increase the binding between COMMD1 and SOD1 as seen with the wild-type CCS protein (Fig. 2C). Taken together, the observation that the interaction SOD1-COMMD1 is dependent on CCS or excess copper implies that binding of COMMD1 to SOD1 requires CCS-mediated copper incorporation into SOD1.
Next, we set out to investigate whether COMMD1 is present in a complex with both SOD1 and CCS. A sequential immunoprecipitation was performed in HEK293T cells in which CCS-FLAG proteins were precipitated prior to the precipitation of SOD1-GST. Immunoblot analysis of HA-COMMD1 revealed that COMMD1 is in the complex with CCS and SOD1. In line with previous observations, the formation of this heterocomplex (COMMD1-CCS-SOD1) depended on copper (Fig. 2D). In addition, as shown in Fig. 2E, endogenous COMMD1 was able to physically interact with endogenous CCS. In more detail, domain 2 of the CCS protein, a region that shows high homology to SOD1, is essential for this interaction (supplemental Fig. S2A). The only identified domain in the COMMD1 sequence thus far, the COMM domain, was shown to bind to CCS (supplemental Fig. S2B). In contrast to SOD1-COMMD1 interaction, binding of COMMD1 to CCS was copper-independent (data not shown). Although CCS enhanced the interaction between SOD1 and COMMD1, COMMD1 did not affect the binding of SOD1 to CCS (supplemental Fig. S3).
Because COMMD1 reduces the protein stability of its interaction partners by promoting their proteasomal degradation (15, 19, 27, 28, 31), we determined its effect on SOD1 stability. SOD1 turnover in HEK293T cells was monitored after inhibition of protein biosynthesis by cycloheximide. Incubation with cycloheximide resulted in a marked decline of immunodetectable SOD1 compared with vehicle-treated cells within 16 h. The expression of exogenous COMMD1 did not enhance SOD1 turnover rates, either under basal or high copper conditions (Fig. 3).
Incorporation of copper is completed by disulfide oxidation of the SOD1 subunits (10, 11). To investigate if COMMD1 influenced holo-SOD1 biogenesis downstream of copper incorporation into SOD1, the effect of COMMD1 on SOD1 disulfide oxidation was monitored in HEK293T cells transiently transfected with COMMD1 and incubated with varying copper concentrations. As described previously (38), SOD1 disulfide oxidation was promoted by overexpression of CCS or excess copper, whereas copper depletion by BCS promoted the reduced state of the SOD1 monomer (Fig. 4A). Overexpression of COMMD1 did not interfere with the oxidation of SOD1 subunits (Fig. 4B). In addition, exogenous COMMD1 did not alter CCS-induced oxidation of SOD1, as the ratio between reduced and oxidized SOD1 was comparable in the presence and absence of COMMD1 (Fig. 4B and data not shown).
Because homodimerization is an essential step in SOD1 activation subsequent to SOD1 oxidation, the effect of COMMD1 on this process in SOD1 maturation was examined. First, we determined whether COMMD1-SOD1 binding was dependent on SOD1 homodimers. Protein lysates of HEK293T cells transfected with SOD1-FLAG and COMMD1-GST were incubated with increasing concentrations of dithiothreitol (DTT), which reduces the intramolecular disulfide bonds of SOD1 dimers and favors the monomeric form of SOD1 (13). As predicted, the amount of SOD1 dimers decreased upon DTT treatment. However, the interaction between COMMD1 and SOD1 was unaffected as determined by GST pulldown analysis (Fig. 5A). This demonstrates that COMMD1-SOD1 interaction does not depend on the dimerization state of SOD1.
Second, the effect of COMMD1 on SOD1 dimerization was studied. COMMD1 overexpression resulted in a decrease of detectable SOD1 dimers in HEK293T cells, whereas knockdown of COMMD1 resulted in an increased level of SOD1 dimers (Fig. 5B). Normally, excess copper promotes the formation of SOD1 dimers (supplemental Fig. S4). However, the SOD1 dimerization in COMMD1-overexpressing cells upon copper incubation could not be enhanced to a similar extent compared with control cells (Fig. 5C). Together, these data indicate that COMMD1 impairs the expression of SOD1 dimers, although the COMMD1-SOD1 interaction seems to be independent of SOD1 homodimers.
Next, we set out to determine the relevance of this interaction for the enzymatic activity of SOD1, and we examined the effect of COMMD1 on SOD1 activity by an in-gel SOD1 activity assay. Expression of exogenous COMMD1 in HEK293T cells resulted in reduced SOD1 activity (Fig. 6A), consistent with the observation that COMMD1 expression resulted in impaired SOD1 homodimerization (Fig. 5, B and C). The repression of SOD1 activity by COMMD1 under basal conditions was confirmed in Neuro2A cells (Fig. 6B). As shown previously, a rise in intracellular copper concentrations readily induced SOD1 activity (Fig. 6B) (43), but the COMMD1-mediated inhibition of SOD1 activity was not abrogated by copper incubation.
We next considered it important to investigate whether endogenous COMMD1 modulates SOD1 activity. We hypothesized that a reduction of COMMD1 expression would lead to an increase in SOD1 activity. To address this hypothesis, endogenous COMMD1 expression was knocked down by transient transfection of shRNA directed against COMMD1 mRNA (shCOMMD1). Consistent with our expectations, knockdown of COMMD1 resulted in enhanced SOD1 activity (Fig. 6, C and D) and further enlarged the copper-dependent increase in SOD1 activity (Fig. 6D).
The effect of COMMD1 expression on SOD1 activity was further investigated using a quantitative SOD1 activity assay based on the superoxide anion-responsive production of formazan. As a control for these experiments, we demonstrated that the amount of detectable SOD1 was not affected by overexpression of COMMD1 and copper overload (Fig. 6, E and F, lower panels). Overexpression of COMMD1 in HEK293T cells resulted in a markedly decreased SOD1 activity under basal conditions (Fig. 6E), confirming that the COMMD1 mediated inhibition on SOD1 activity as seen in Fig. 6, A and B. Overexpression of CCS resulted in slightly enhanced SOD1 activity, but simultaneous overexpression of COMMD1 overruled this CCS-mediated increase of SOD1 activity. SOD1 activity was clearly and significantly decreased in these cells in a COMMD1 concentration-dependent manner. In line with the observation that binding of COMMD1 to SOD1 is mediated by CCS, the inhibitory effect of COMMD1 on SOD1 activity was more pronounced when CCS was co-expressed (Fig. 6E). Conversely, knockdown of COMMD1 expression in HEK293T cells resulted in a significant increase in SOD1 activity (~2.5-fold) on top of the induction of SOD1 function upon copper incubation (Fig. 6F). Together, these data demonstrate that endogenous COMMD1 represses cellular SOD1 activity. More specifically, these findings indicate an inverse correlation between COMMD1 expression and SOD1 activity in different cell lines and suggest that COMMD1 might help regulate cellular superoxide anion concentrations.
As readout for intracellular superoxide anion concentrations, the fluorescence of the superoxide-specific probe DHE was measured in cells. HEK293T cells stably transfected with shControl or shCOMMD1 were incubated with the pro-oxidant menadione prior to DHE labeling and analyzed by flow cytometry. Upon menadione incubation, the DHE fluorescence was markedly increased in shControl-expressing cells (Fig. 7, A and B). In COMMD1 knockdown cells, a significant decline in DHE fluorescence was detected both under basal and menadione-induced conditions (Fig. 7, A and B) corresponding to the observed increase in SOD1 activity (Fig. 6 and data not shown). The interaction between COMMD1 and SOD1 remained unaffected under these conditions (supplemental Fig. S5).
These data were confirmed in Neuro2A cells in which Commd1 expression was transiently decreased by siRNA targeting of Commd1 mRNA (siCommd1). Cells transfected with a nontargeting siRNA pool were used as control (siControl). Commd1 depletion did not result in differences in DHE fluorescence under basal conditions. Incubation with menadione showed a clear induction of DHE fluorescence in siControl transfected cells. However, siRNA-mediated decrease in Commd1 expression revealed a significant reduction of DHE fluorescence under these conditions (Fig. 7, C and D), which is in accordance with the enhanced SOD1 activity observed in COMMD1 knockdown cells (Fig. 6 and data not shown). In both cell lines, the SOD1 protein expression was not changed by a decrease in Commd1 expression (Fig. 7, B and D, lower panels). However, the differences in DHE fluorescence decline upon COMMD1 knockdown in untreated cell lines might be explained by variation in the efficiency of COMMD1 knockdown. Collectively, these data demonstrate that endogenous COMMD1 expression significantly induces intracellular superoxide anion concentrations in different cell lines, likely in response to its repression of SOD1.
The post-translational events required for activation of human SOD1 provide a unique and elegant mechanism to tailor the activity of this abundant anti-oxidant enzyme to the availability of oxygen and the transition metal copper (7, 12). Here, we identify COMMD1 as a novel interaction partner of SOD1. This interaction between COMMD1 and SOD1 is genuine because it was detected in multiple different cell types, using overexpressed as well as endogenous proteins. In addition, the interaction displayed specificity with respect to varying amounts of copper and the metallochaperone CCS. Moreover, COMMD1 impaired SOD1 activity by reducing the expression levels of enzymatically active SOD1 homodimers late in the post-translational maturation process of SOD1. Intriguingly, COMMD1 expression was correlated with an increase in intracellular superoxide anions. These findings extend the already impressive COMMD1 protein-protein interaction network and support previous suggestions that COMMD1 exerts its function through physical interactions with proteins that are structurally and functionally unrelated (15, 19,–21, 27). More importantly, our results provide the first indication that SOD1 dimer formation is a regulated final step in SOD1 activation and implicate COMMD1 in this process.
The current findings are integrated in a model that describes the effect of COMMD1 on CCS-mediated SOD1 maturation and activation (summarized in Fig. 8). The majority of cellular SOD1 is present as a preexisting apo-SOD1 pool, which contains zinc but lacks copper (E,Zn-SOD1 subunit (3, 4)). Consequently, the inactive apo-SOD1 requires copper insertion, disulfide oxidation, and homodimerization for its activation (Fig. 8, upper panel).
Several observations are consistent with the conclusion that COMMD1 acts downstream of copper incorporation and intramolecular disulfide oxidation. First, COMMD1 expression did not reduce the half-life time of SOD1, suggesting that it does not interfere with the stability of the E,Zn-SOD1 apo-pool. Second, COMMD1 expression had no effect on the ratio of disulfide oxidized to disulfide-reduced SOD1. As described previously, CCS overexpression or elevated copper concentrations favored transition of SOD1 to the disulfide-oxidized form, and COMMD1 did not affect this process.
Our data demonstrate that COMMD1 impairs the expression levels of SOD1 homodimers and activity in conditions of excess copper (Fig. 8, lower panel). Copper incorporation into SOD1 and disulfide isomerization are strictly coupled events (4, 13). Both excess copper and CCS expression lead to enhanced copper incorporation into apo-SOD1, thus increasing the amount of active holo-SOD1 (7, 43). We showed that excess copper and overexpression of CCS, but not a copper incorporation-deficient mutant of CCS (CCSC244A/C246A), induced COMMD1 binding to SOD1. In contrast, RNAi-mediated depletion of CCS diminished COMMD1-SOD1 interaction and demonstrated that CCS indeed predominantly mediated that copper-dependent binding of COMMD1 to SOD1. At the same time, COMMD1 binds to CCS copper independently, but in excess copper, the two proteins are found in a complex together with SOD1. Collectively, our data suggest that COMMD1 binding requires CCS-mediated copper incorporation in SOD1 to establish its inhibitory effect on SOD1 dimerization.
Perhaps the most striking finding presented here indicates that siRNA-mediated depletion of endogenous COMMD1 results in increased formation of enzymatically active SOD1 dimers. COMMD1 can therefore be considered as a novel endogenous regulator of the final step in SOD1 activation, which is the formation of active dimers. This COMMD1-mediated reduction in SOD1 dimers and activity is most abundant in conditions of excess copper and overrules the CCS-mediated copper-responsive maturation of SOD1. The biological relevance of these observations is reflected by the repression of intracellular superoxide anion concentrations upon depletion of endogenous COMMD1. It has previously been described that copper incorporation into SOD1 and disulfide oxidation results in a conformational change of the SOD1 subunit necessary for homodimerization (reviewed in Ref. 12). This conformational change may expose a binding site for COMMD1 that had previously been buried within the SOD1 structure. It might be possible that COMMD1 prevents SOD1 dimerization by directly competing for the same SOD1-SOD1 interaction interface. Structural studies should provide detailed mechanistic answers into the many unsolved questions raised by the findings presented in this paper.
Interestingly, XIAP has also recently been identified as a novel protein involved in the regulation of SOD1 activation. Brady et al. (36) showed that, depending on the intracellular copper status, CCS delivers copper either to SOD1 or to XIAP. Under basal conditions, XIAP enhances copper acquisition and delivery to SOD1 by promoting the nondegradative ubiquitination of CCS. However, in excess copper, CCS delivers copper to XIAP resulting in XIAP degradation, induction of caspases, and eventually apoptosis. Because XIAP also binds to COMMD1 and enhances the degradative ubiquitination of COMMD1 (20, 25), it might be possible that COMMD1 function is related to this CCS- and XIAP-mediated dynamic regulation of copper delivery and SOD1 activation. However, additional studies are needed to clarify the possible role of COMMD1 in this new association of copper metabolism and apoptosis.
Taken together, COMMD1 regulates SOD1 maturation and is a novel regulator of SOD1 activity. These findings will be important for further understanding of the molecular mechanism of post-translational modifications of SOD1 maturation and activation in its defense against toxic superoxide anions.
We are grateful to Dr. Inge van den Berg, Dr. Peter van den Berghe, Martijn Gloerich, and Jackie Senior for critically reading the manuscript. We thank the laboratories of Dr. E. Burstein and Dr. C. Duckett for kindly providing COMMD1- and CCS-deletion constructs, respectively.
3The abbreviations used are: