The C(66)W Mutation in the Deafness-Dystonia Peptide 1 (DDP1) affects the formation of functional DDP1/TIM13 complexes in the mitochondrial intermembrane space

Mohr-Tranebjaerg syndrome is a progressive, neurodegenerative disorder caused by loss-of-function mutations in the DDP1/TIMM8A gene. DDP1 belongs to a family of evolutionary conserved proteins which are organized in hetero-oligomeric complexes in the mitochondrial intermembrane space. They mediate the import and insertion of hydrophobic membrane proteins into the mitochondrial inner membrane. All of them share a conserved Cys4 metal binding site proposed to be required for the formation of zinc fingers. So far, the only missense mutation (C66W) known to cause a full-blown clinical phenotype directly affects this Cys4 motif. Here, we show that the mutant human protein is efficiently imported into mitochondria and sorted into the intermembrane space. In contrast to wild-type DDP1 it does not complement the function of its yeast homologue Tim8. The C66W mutation impairs binding of Zn2+ ions via the Cys4 motif. As a consequence, the mutated DDP1 is incorrectly folded and loses its ability to assemble into a hetero-hexameric 70 kDa complex with its cognate partner protein human Tim13. Thus, an assembly defect of DDP1 is the molecular basis of Mohr-Tranebjaerg syndrome in patients carrying the C66W mutation.


Introduction
With the exception of a few components of the oxidative phosphorylation machinery, all mitochondrial proteins are encoded by nuclear genes and synthesized on cytosolic ribosomes.
The import of such preproteins into mitochondria and the correct sorting into mitochondrial subcompartments is mediated by a set of import systems in the outer and inner mitochondrial membrane. Three distinct preprotein import systems have been described (1-5). All preproteins most likely use the general translocase in the outer membrane, the TOM 1 complex. It mediates the recognition and binding of preproteins as well as their transfer across the outer membrane. Further movement of the translocation intermediates into and across the inner membrane is mediated by two distinct translocases in the inner membrane, the TIM23 and the TIM22 complexes. The TIM23 complex mediates import of preproteins with positively charged targeting signals at their N-termini into the mitochondrial matrix space and into the inner membrane (6-8). The transfer of preproteins across the inner membrane strictly requires both an electrochemical potential (∆ψ) across the inner membrane and ATP in the matrix as energy sources (9). The TIM22 complex is used by a class of hydrophobic inner membrane proteins with internal and so far less characterized targeting signals (10-15).
Typical substrates are the members of the mitochondrial carrier family that are synthesized without a matrix targeting signal. In addition, the TIM22 complex appears to mediate the import of precursors of other hydrophobic membrane proteins such as Tim23 and Tim22, which do not belong to the class of mitochondrial carriers (16)(17)(18). Insertion of these precursors into the inner membrane is strictly dependent on ∆ψ but does not require ATP in the matrix.
Proteins destined for the inner membrane require the help of small, structurally related Tim proteins in the intermembrane space. In the yeast S.cerevisiae, five members of this protein by guest on July 10, 2020 http://www.jbc.org/ Downloaded from family are expressed (19). Of these, Tim9, Tim10 and Tim12 specifically support the import of mitochondrial carrier preproteins. (11)(12)(13)(14)16,20). They form two distinct heterooligomeric complexes of 70 kDa which interact with the hydrophobic precursors thereby keeping them in an import competent conformation. The TIM9-10 complex interacts with the carrier proteins early in the import pathway when they are partially translocated across the TOM complex. The carrier proteins are then handed over to the TIM9-10-12 complex which is tightly associated with the TIM22 complex and mediates their insertion into the inner membrane.
Tim8 and Tim13 assist the import of a distinct subgroup of inner membrane proteins such as Tim23, the major component of the TIM23 translocase (16,(21)(22)(23). Like the other members of the family they exist as hetero-oligomeric 70 kDa complexes in the intermembrane space but, in contrast to Tim9, Tim10 and Tim12, are not essential for the cell viability in yeast. The TIM8-13 complex binds to the incoming Tim23 precursor still associated with the TOM complex. In yeast, it is only required when the membrane potential is low and the membrane insertion of Tim23 is inefficient (21). Under these conditions the TIM8-13 complex is necessary to accumulate the Tim23 precursor and present it to the TIM22 complex thereby facilitating its insertion into the inner membrane.
The human homologue of Tim8 is encoded by the DDP1 (deafness dystonia peptide 1) gene. Mutations in the DDP1/TIMM8A gene cause the Mohr-Tranebjaerg syndrome, a progressive, neurodegenerative disorder characterized by sensorineural hearing loss, dystonia, mental retardation and blindness (24,25). Most of the patients harbor loss-of-function mutations leading to a complete absence of the DDP1 protein. In human mitochondria, DDP1 forms a hetero-oligomeric complex of 70 kDa together with hTim13 (26). The DDP1-hTim13 complex specifically assists the import of the human Tim23 precursors into the inner by guest on July 10, 2020 http://www.jbc.org/ Downloaded from membrane (26). The human complex is able to complement the function of the TIM8-13 complex in yeast. Whereas import of yeast Tim23 does not require the TIM8-13 complex under normal conditions, import of human Tim23 appear to be dependent on the assistance of the DDP1-hTim13 complex under all conditions studied (26). It was therefore suggested that the pathomechanism underlying Mohr-Tranebjaerg syndrome may involve an impaired biogenesis of the human TIM23 complex.
The small Tim proteins belong to an evolutionary conserved protein family characterized by a common Cys 4 metal binding motif. Binding of Zn 2+ ions was proposed to be required for the formation of typical zinc finger structures (13,19,26). These zinc fingers may be crucial for the recognition and binding of translocation intermediates during their transfer through the aqueous environment of the intermembrane space.
In this report, we analyzed the structural and functional consequences of a mutation (C66W), directly affecting the Cys 4 metal binding motif. This cysteine to tryptophan exchange at amino acid position 66 is currently the only missense mutation known to cause . We show that the DDP1 C66W is efficiently imported into mitochondria and correctly sorted into the intermembrane space. However, the mutant protein is no more able to complement the function of Tim8 in yeast mitochondria lacking the endogenous TIM8-13 complex; in particular, the mutant DDP1 cannot restore import of Tim23. Analysis of purified recombinant DDP1 revealed that the C66W exchange impairs the ability of DDP1 to bind zinc. As a consequence, DDP1 C66W does not fold properly and is no longer able to interact with its partner protein hTim13. Thus, the C66W substitution leads to a defect in the formation of a functional DDP1-hTim13 complex. This may interfere with the biogenesis of the TIM23 complex and an impaired biogenesis of the mitochondria might by by guest on July 10, 2020 http://www.jbc.org/ Downloaded from the pathomechanistic basis of the Mohr-Tranebjaerg syndrome.

In vitro synthesis of precursor proteins and import into mitochondria -
In order to analyze the interaction between DDP1 and hTim13 in vitro, recombinant hisDDP1 or hisDDP1 C66W was bound to Ni-NTA resin and washed with 10 volumes of buffer (1 x PBS, pH 7.4, 20 mM Imidazol). Subsequently, E. coli suspensions expressing hTim13-GST were lysed and centrifuged at 10,000 g (20 min at 4°C). The supernatant was loaded onto the prebound columns. After an over-night incubation, the columns were extensively washed with washing buffer to remove unspecifically bound material. His-tagged proteins were then eluted using 5 volumes of equilibration buffer containing 300 mM imidazole. Aliquots (30 µl) of the elution fractions were removed for direct analysis. In a second purification step, elution fractions were incubated with glutathione sepharose beads (GSH beads). Material bound to the beads was released in Laemmli buffer, applied to SDS-PAGE and analyzed by Western blotting using affinity purified antibodies against DDP1 and hTim13.
Miscellaneous -Mitochondria from yeast were isolated as described (28). Mitochondria from mouse liver were prepared in a medium containing 0.25 M sucrose, 5 mM EDTA, 20 mM Tris/HCl pH 7.4 by differential centrifugation and subsequent sucrose step gradient ultracentrifugation as described previously (29). Antisera against the C-termini of DDP1, hTim13 and hTim23 were raised in rabbits by injecting the chemically synthesized peptides as described before (26,29). RNA was prepared from 10 ml liquid yeast cultures grown to an OD1.0. Cells were resuspended in AE buffer (50 mM Na-acetate, 10 mM EDTA, pH 5.3).
SDS was added to a final concentration of 1 % and RNA was isolated using a phenol/freeze protocol. Ethanol precipitated total RNA was used for first-strand cDNA synthesis using Superscript II reverse transcriptase (Gibco) and oligo d(T) primers.

Results
The C66W 1A and B).
Rescue of the cold-sensitive phenotype represents a model system to functionally analyze pathogenic DDP1 mutations.
Therefore, the C66W exchange was introduced by PCR-based mutagenesis and DDP1 C66W was expressed in ∆8/∆13 cells together with hTim13 (∆8/∆13+C66W/hTim13). When the transformed cells were grown at 15°C in the presence of glucose the mutant DDP1 C66W was not able to rescue the cold-sensitive phenotype of ∆8/∆13 ( Fig. 1A and B). This indicates that a mutation affecting the putative zinc finger of DDP1 leads to a non-functional protein.
As shown previously the import efficiency of yeast and human Tim23 precursors in vitro is reduced significantly in mitochondria from ∆8/∆13 cells (21,26). While import of yeast Tim23 is affected in mitochondria from ∆8/∆13 cells only under conditions of low membrane potential, import of human Tim23 is reduced in ∆8/∆13 mitochondria under all conditions.
When DDP1 and hTim13 are expressed together in ∆8/∆13 cells, the import of yeast and human Tim23 precursors is restored to wild type levels (26). We therefore investigated whether the mutant DDP1 is still able to restore the import of Tim23. Figure 2 (Fig. 3B).
To investigate whether the C66W substitution affects import of the mutant protein into the intermembrane space, DDP1 and DDP1 C66W were synthesized in reticulocyte lysates in the presence of [ 35 S]methionine. The radiolabeled proteins were then incubated with isolated yeast mitochondria. To assess translocation of radiolabeled precursors across the outer mitochondrial membrane the mitochondria were treated with proteinase K (PK). DDP1 and DDP1 C66W were protected from proteinase K digestion. Both proteins were, however, degraded when mitoplasts were generated and treated with PK (Fig. 4A, left panels). This by guest on July 10, 2020 http://www.jbc.org/ Downloaded from indicates that DDP1 and the mutant protein DDP1 C66W were efficiently imported into the intermembrane space of mitochondria. Both DDP1 and DDP1 C66W precursors were also equally well imported into mitochondria isolated from mouse liver (Fig. 4A, right panels).
Import of wild type and mutant DDP1 into the intermembrane space was not dependent on a membrane potential across the inner membrane (Fig. 4B). Thus, the C66W exchange does not impair import of DDP1 across the mitochondrial outer membrane and its correct sorting into the mitochondrial intermembrane space.
To assess the stability of the imported precursors in the intermembrane space, the kinetics of degradation of DDP1 and DDP1 C66W were compared. Therefore, radiolabeled DDP1 or DDP1 C66W precursor was imported into mitochondria isolated from mouse liver and import was stopped after 10 min by treatment with trypsin to remove non-imported precursor.
Mitochondria were re-isolated and incubated at 37°C in the presence of ATP and an ATPregenerating system (CK/Cr-P system) to stimulate degradation of proteins in the intermembrane space by the i-AAA protease Yme1p (30). The DDP1 level was reduced by 20 % within 90 min, while the amount of DDP1 C66W decreased by about 55 % within the same time period (Fig. 5) suggesting that the turn-over rate of the mutant DDP1 C66W was significantly higher than that of wild-type DDP1.  Fig. 6A and B). In contrast, the preparation of MBP-DDP1 C66W contained only background levels of zinc, comparable to zinc levels obtained with the control preparation of MBP-α ( Fig. 6A and B). Thus, the C66W exchange impairs zinc binding of DDP1.

Binding of Zn 2+ ions is required for stable folding of DDP1 -We tested whether zinc
binding is required to stabilize the folded conformation of DDP1. His-tagged versions of DDP1 or DDP1 C66W were expressed in E. coli and purified by Ni-NTA chromatography.
Aliquots (corresponding to 2 µg protein) were subjected to digestion with increasing amounts of trypsin (Fig. 7). DDP1 was clipped to a fragment of lower molecular weight which was resistant to protease treatment up to a concentration of 100 µg/ml of trypsin. In contrast, when DDP1 was pre-incubated with the metal-chelating agents EDTA and o-phenantroline (EDTA/o-phe) it was completely degraded by 10 µg/ml trypsin (Fig. 7). This indicates that zinc binding stabilizes the folded conformation of DDP1.
HisDDP1 C66W was sensitive to treatment with lower concentrations of trypsin in the absence of metal chelating reagents (Fig. 7) indicating that the mutant protein was loosely or unproper folded. Since hisDDP1 C66W does not bind zinc, the data demonstrate that the C66W substitution impairs stable folding of DDP1 by affecting the Cys 4 metal binding motif. The recombinant proteins were subjected to gelfiltration analysis in order to assess their oligomeric state. HisDDP1 eluted in a fraction corresponding to a molecular mass of 65 kDa, indicating that the recombinant protein assembles into homo-oligomeric complexes ( Fig. 8A and B). HisDDP1 C66W eluted predominantly in the void volume of the Superdex-75 column suggesting that the mutant form of DDP1 was aggregated (not shown). A small portion of hisDDP1 C66W was found in the 65 kDa fraction. This suggests that the C66W exchange impairs the ability of DDP1 to form homo-oligomers.
For further analysis, the recombinant proteins were subjected to chemical crosslinking with the aminospecific crosslinker disuccinimidyl suberate (DSS). The crosslinked adducts were analyzed by SDS-PAGE and Western blotting. In the absence of DSS, the electrophoretic mobility of hisDDP1 corresponded to a monomer (Fig. 8C). A small fraction (<3 %) of hisDDP1 migrated corresponding to a dimer which could not be cleaved by addition of DTT or β-mercaptoethanol (data not shown). This dimeric form may, therefore, represent a spontaneous crosslinking product. In the presence of DSS, adducts corresponding to dimeric, trimeric, tetrameric, pentameric and hexameric hisDDP1 were detected (Fig. 8C, arrows). No crosslinking products of higher oligomeric state were generated even when higher concentrations of DSS were used or when crosslinking was performed with glutaraldehyde, a less specific crosslinking reagent (data not shown). This supports that purified hisDDP1 can form homo-hexameric complexes in vitro. In vivo, DDP1 assembles into a heterooliogomeric complex with its structurally related partner protein hTim13 (26).
HisDDP1 C66W migrated predominantly in a monomeric form in the absence of DSS (Fig.   8C). Like hisDDP1, a small fraction of hisDDP1 C66W was unspecifically crosslinked to a dimer. In the presence of DSS the amount of the dimeric hisDDP1 C66W increased slightly, however, no adducts of higher oligomeric state were formed, supporting the conclusion that the C66W substitution affects the complex formation of the mutant protein.

Formation of a zinc finger structure is required for the interaction between DDP1 and hTim13
-In human mitochondria DDP1 and hTim13 assemble into hetero-oligomeric complexes in the intermembrane space (26). In order to analyze whether the C66W mutation also affects the association of DDP1 and hTim13 we performed a two-step interaction assay in vitro. In the first step hisDDP1 or hisDDP1 C66W were bound to Ni-NTA sepharose. Subsequently, the hisDDP1 or hisDDP1 C66W affinity resins were incubated with E. coli lysates containing hTim13-GST, a chimeric fusion protein of human Tim13 and glutathione S-transferase.
After extensive washing steps the bound material was eluted with imidazol and aliquots were analyzed. hTim13-GST co-eluted together with hisDDP1, whereas only traces of hTim13-GST were detected in the peak elution fraction together with hisDDP1 C66W (Fig. 9A and B, left panels). To exclude non-specific binding of GST fusion proteins to the Ni-NTA resin, the eluates were applied to a second purification step using glutathione sepharose beads (GSH beads). The GSH beads were extensively washed. Bound material was released using Laemmli buffer and analyzed by SDS-PAGE and Western blotting. The bound material originating from the hisDDP1 column contained hisDDP1 and hTim13-GST indicating that both proteins directly interact in vitro (Fig. 9A, middle panel). As estimated from Coomassie blue stained gels hisDDP1 and hTim13-GST were present in roughly stoichiometric amounts in these complexes (Fig. 9A, right panel). In contrast, no hisDDP1 C66W was detected after the second purification step using GSH beads (Fig. 9B, right panel). Thus, the C66W exchange impairs the direct interaction between DDP1 and its cognate partner protein hTim13 thereby compromising the ability to form functional hetero-oligomeric complexes. In the present study, we generated DDP1 constructs carrying the C66W mutation in order to analyze which function of DDP1 is affected. We combined in vitro analysis of purified recombinant protein with in vivo functional analysis of the mutant DDP1 in the yeast system.

Discussion
The ∆8/∆13 yeast strain, lacking both yTim8 and yTim13, enabled us to test the function of the mutated DDP1. As demonstrated previously, expression of wild-type DDP1 together with human Tim13 in ∆8/∆13 cells rescues the cold sensitive phenotype of the ∆8/∆13 yeast strain (26). In contrast, when DDP1 C66W instead of wild-type DDP1 was introduced into ∆8/∆13 cells, transformants failed to grow on glucose at 15°C indicating that the mutant DDP1 is nonfunctional in yeast. Furthermore, mutant DDP1 is apparently no more able to take over the function of its yeast homologue Tim8 during import of Tim23. We showed that the DDP1 C66W precursor is efficiently imported into mitochondria and correctly sorted into the intermembrane space. Although the mutant DDP1 appears to be about 2-fold more susceptible to degradation by endogeneous proteases, it accumulates in mitochondria to a level near that of wild-type DDP1. Thus, the inability of the mutant DDP1 to complement strain ∆8/∆13 does not result from a reduced protein level in mitochondria.
The C66W mutation affects a highly conserved element within DDP1 consisting of four cysteines which likely form a metal binding site. We recently demonstrated that DDP1 and also its partner protein hTim13 indeed bind zinc in a 1:1 stoichiometry and thus appear to form zinc fingers (26). We have shown that the C66W mutation impairs the ability of DDP1 to bind zinc ions. As a consequence, the C66W mutation leads to an incorrectly folded protein, as indicated by its increased susceptibility to externally added protease. Likewise, pretreatment of wild-type DDP1 with the metal-chelating agents EDTA/o-phe results in an increased sensitivity of the protein to trypsin. It is therefore likely, that the C66W mutation alters the three-dimensional structure of the zinc binding domain thereby influencing the folding properties of DDP1. In summary, we showed that exchange of one of the conserved cysteine residues leads to impaired zinc binding and subsequent incorrect folding of the zinc finger domain. The functional consequence of the C66W mutation is primarily a deficiency of DDP1 to form a 70 kDa complex together with hTim13 in the mitochondrial intermembrane space. Thus, although the mutated DDP1 protein is imported into mitochondria and correctly sorted, it is non-functional and leads to a full-blown clinical phenotype in Mohr-Tranebjaerg patients.

Figure 3
Expression of the DDP1 C66W protein in yeast mitochondria. . When indicated, samples were treated with 100 µg/ml proteinase K (PK) mitochondria.

Figure 5
Stability of imported DDP1 and DDP1 C66W precursors. Radiolabeled DDP1 or DDP1 C66W precursors were imported into isolated mouse liver mitochondria for 10 min at 25°C.
Mitochondria were treated with 100 µg/ml trypsin to remove not imported precursor, reisolated by centrifugation and incubated at 37°C in the presence of a ATP-regenerating system containing creatine kinase and creatine phosphate. Aliquots (each corresponding to 50 µg mitochondrial protein) were removed after the indicated time periods and analyzed after SDS-PAGE and blotting onto nitrocellulose membrane. A Fuji PhosphorImaging system was used for quantification. The signal intensity of radiolabeled imported DDP1 at t = 0 min was set to 100 %. Protein concentrations were determined by Bio-Rad protein assay, and molar ratios (mol of Zn 2+ per mol of fusion protein) were calculated.

Figure 7
Sensitivity of recombinant DDP1 and DDP1 C66W to digestion with trypsin. His-tagged versions of DDP1 and DDP1 C66W were expressed in E. coli and affinity-purified on Ni-NTA columns. Purified, recombinant protein (2 µg) was subjected to digestion in 200 µl PBS buffer using increasing amounts of trypsin and analyzed by SDS-PAGE and immunoblotting using a monoclonal antibody against penta-his. When indicated, recombinant protein was incubated with the metal-chelating agents EDTA/o-phe prior to protease treatment.

Figure 8
Oligomeric state of DDP1 and DDP1 C66W . A, Gel filtration analysis. Affinity-purified hisDDP1 or hisDDP1 C66W (5 µg) were subjected to gel filtration analysis on a Superdex-75 column. Elution fractions were analyzed by SDS-PAGE and immunoblotting using a pentahis antibody and quantified by densitometry. HisDDP1 eluted corresponding to an apparent molecular mass of 65 kDa. HisDDP1 C66W was predominantly found in the void volume (not shown). B, Calibration standard curve using cytochrome c, carbonic anhydrase and bovine serum albumin as calibration standards. C, Chemical crosslinking of wild-type and mutant DDP1. Two µg-aliquots of affinity-purfied and desalted hisDDP1 or hisDDP1 C66W were incubated in the absence or presence of the bifunctional amino-reactive crosslinker DSS at a concentration of 100 µM for 30 min at room temperature. TCA-precipitated crosslinking by guest on July 10, 2020 http://www.jbc.org/ Downloaded from products were analyzed by SDS-PAGE and immunoblotting using a penta-his antibody.

Figure 9
C66W exchange impairs the interaction between DDP1 and human Tim13. A, hisDDP1 was expressed in E. coli and pre-bound to a Ni-NTA resin in order to generate a DDP1-affinity column. The pre-bound resin was then incubated overnight with E. coli lysate containing hTim13-GST. After extensive washing, bound material was eluted using 300 mM imidazole and aliquots of elution fractions were analyzed by immunoblotting using affinity purified antibodies against DDP1 and hTim13. hTim13-GST co-eluted together with hisDDP1 (left panel).
In a second purification step, the peak elution fraction was incubated with GSH beads. After extensive washing, the entire material bound to GSH beads was analyzed by SDS-PAGE and immunoblotting (middle panel) or Coomassie blue staining (right panel). The bound material contained hTim13-GST and hisDDP1 in about equimolar amounts (right panel). B, E. coli lysate containing hTim13-GST was loaded onto a Ni-NTA column to which recombinant hisDDP1 C66W was pre-bound. Bound material was eluted and aliquots of elution fractions were analyzed by immunoblotting using anti-DDP1 and anti-hTim13 antibodies. Elutions fraction contained hisDDP1 C66W but only trace amounts of hTim13-GST (left panel). To exclude that minor amounts of hTim13-GST were bound to hisDDP1 C66W , pooled peak elution fractions were applied to GSH beads. Material released from GSH beads was analyzed by immunoblotting (right panel).
by guest on July 10, 2020