Defective Mitochondrial Protein Translocation Precludes Normal Caenorhabditis elegans Development*

We demonstrate biochemically that the genes identified by sequence similarity as orthologs of the mitochondrial import machinery are functionally conserved in Caenorhabditis elegans . Specifically, tin-9.1 and tin-10 RNA interference (RNAi) treatment of nematodes impairs import of the ADP/ATP carrier into isolated mitochondria. Developmental phenotypes are associated with gene knock-down of the mitochondrial import components. RNAi of tomm-7 and ddp-1 resulted in mitochondria with an interconnected morphology in vivo , presumably due to defects in the assembly of outer membrane fission/fusion components. RNAi of the small Tim proteins TIN-9.1, TIN-9.2, and TIN-10 resulted in a small body size, reduced number of progeny produced, and partial embryonic lethality. An additional phenotype of the tin-9.2 (RNAi) animals is defective formation of the somatic gonad. The biochemical demonstration that the protein import activity is reduced, under the same con-ditions that yield the defects in specific tissues and lethality in a later generation, suggests that the developmental abnormalities observed are a consequence of defects in mitochondrial inner membrane biogenesis.

The mitochondrion consists of an outer and inner membrane, which serves to separate two aqueous compartments, the matrix and the intermembrane space. There is an elaborate set of proteins within each sub-compartment of the mitochondrion for protein import (5,6). These translocons are specialized and recognize precursors that possess specific targeting and sorting information. In the budding yeast Saccharomyces cerevisiae, the outer mitochondrial membrane contains a single translocase for the passage of polypeptides. The translocase of the outer membrane (TOM) 1 complex facilitates translocation across the outer membrane (7)(8)(9). The sorting and assembly machinery, a second complex in the outer membrane, facilitates assembly of outer membrane proteins with complex topologies such as ␤-barrel structures (10). Once a precursor has passed through the TOM complex to the intermembrane space, the precursor can take one of several routes. Precursors destined for the matrix typically contain an amino-terminal targeting signal that is utilized by the translocase of the inner membrane (TIM)23 complex (11,12). The TIM23 translocon facilitates translocation across the mitochondrial inner membrane in a membrane potential dependent manner. Insertion into the inner membrane is accomplished by a related but functionally distinct translocon, the TIM22 complex (5,6,13,14). Precursors that utilize the TIM22 translocon such as the carrier family of proteins, Tim22p and Tim23p, have targeting information within the mature polypeptide and lack an aminoterminal targeting sequence. In addition, the TIM22 translocon utilizes two soluble complexes in the intermembrane space, the Tim9p-Tim10p complex and the Tim8p-Tim13p complex, that function to bind to the hydrophobic stretches found in TIM22 substrates as they are escorted across the aqueous intermembrane space (15)(16)(17)(18)(19).
Finally, many aspects of the translocation machinery appear to be conserved in mammalian systems (20,21). Mitochondrial dysfunction is a common factor in a broad range of diseases, and numerous mitochondrial proteins have been identified as regulators of the aging process (22,23). The first disease Mohr-Tranebjaerg syndrome associated with a defect in protein import is caused by mutations in the intermembrane space import component DDP-1 (deafness dystonia polypeptide) (24,25).
To demonstrate the conservation of import mechanisms in higher eukaryotes and to study the developmental consequences of mitochondrial defects, we investigated mitochondrial import in C. elegans. The reverse genetic manipulation used impairs mitochondrial protein translocation in biochemical assays. Gene knock-down of the homologous translocation components led to developmental defects in specific tissues as well. Specifically, knock-down of tin-9.1, tin-9.2, and tin-10 resulted in a small body (Sma) size, reduced brood size, and partial embryonic lethality. tin-9.2(RNAi) animals also had cell migration defects. tomm-7(RNAi) and ddp-1(RNAi) animals displayed defects in mitochondrial morphology. Thus, aspects of human mitochondrial diseases are mimicked in C. elegans and provide an excellent model for understanding the global impact of mitochondrial biogenesis on a multicellular organism.

EXPERIMENTAL PROCEDURES
Strains and Cloning-Strains were maintained at 20°C unless otherwise specified by using standard techniques (26). Wild-type genes were cloned from N2 Bristol. The mitochondrial outer membrane YFP marker under the expression of the myo-3 promoter was generously donated by the van der Bliek laboratory. Transgenic worms were obtained as described previously (27) by injecting expression constructs and the transformation marker, rol-6(su1006), into the gonad of adult wild-type C. elegans.
RNAi constructs were created by PCR amplification of genes from either genomic DNA or cDNA pools generated from purified poly(A) ϩ RNA (Ambion). The PCR products were then cloned into the L4440 vector, and the Escherichia coli host HT115 was used for feeding RNAi studies (28). A genomic clone of tin-9.1 and cDNA clones of tomm-7, ddp-1, tin-9.2, and tin-10 were used.
Microscopy and Image Analysis-Live worms were mounted on a film of dried agarose in a small volume of M9 medium. The worms were paralyzed with 10 mM aldicarb. Green fluorescent protein variants were viewed with fluorescein isothiocyanate, YFP, or CFP filter sets (Chroma Technologies Corp.). Worms were visualized on a Zeiss Axiovert 200 microscope.
RNAi Methods-Eggs were isolated from OP50 fed wild-type worms and allowed to hatch in liquid S-medium overnight in the absence of food. The synchronous L1 larvae were then seeded onto NGM plates containing E. coli expressing double-stranded RNA specific for mitochondrial import components. The plates were incubated at the temperatures stated. At adulthood, worms were moved to fresh RNAi plates and allowed to lay eggs for 4 h. The adult animals were then removed, and the second-generation animals were scored for morphological defects. Third generation animals were obtained in a similar manner.
Mitochondria Purification-For large scale RNAi, worms grown in liquid culture were prepared as above except the synchronous and starved L1 larvae were diluted to 2000 worms/ml of S-medium containing freshly induced E. coli expressing dsRNA. The worms were fed until adulthood, and then the animals were cleaned by sucrose flotation and washed thoroughly before utilization for mitochondria purification.
Mitochondria were purified from lactate-grown yeast cells (29) or from S-medium-cultured adult worms by a modified protocol derived from Jonassen et al. (30). The nematodes were resuspended in buffer STEG (250 mM sucrose, 5 mM Tris-HCl, 1 mM EGTA, pH 7.4) with 1 mM PMSF and protease inhibitor mixture (PIC, Roche Diagnostics). The samples were kept cold on ice throughout the fractionation procedure. The animals were homogenized with a Kontes ground glass tissue grinder using 15 strokes. The volume was increased to 25 ml with STEG ϩ 1 mM PMSF/PIC and centrifuged at 750 ϫ g for 10 min. The supernatants were saved, and another 10 ml of STEG ϩ1 mM PMSF, 1ϫ PIC was added to the pellets. The pellets were resuspended and homogenized again with another 15 strokes, and the volumes were increased to 25 ml and centrifuged at 750 ϫ g for 10 min. The supernatants were combined and centrifuged at 12,000 ϫ g for 10 min. The mitochondrial pellets were gently resuspended in 10 ml of STEG without protease inhibitors using a Potter-Elvehjem tissue homogenizer with PTFE pestle (Bellco). The mixture was centrifuged at 750 ϫ g for 10 min. The supernatants were collected, avoiding the pellets, and were centrifuged at 12,000 ϫ g for 10 min. The final mitochondrial pellets were resuspended in STEG. Purified mitochondria were used immediately for in vitro import assays because coupled import-competent organelles could not be recovered after freezing.
In Vitro Protein Import Assays-In vitro protein import assays were performed as described (31). Proteins were synthesized in a rabbit reticulocyte lysate in the presence of [ 35 S]methionine and [ 35 S]cysteine after in vitro transcription of the corresponding gene by SP6 polymerase. The reticulocyte lysate containing the radiolabeled precursor was incubated with isolated mitochondria at the indicated temperatures in import buffer (1 mg/ml bovine serum albumin, 0.6 M sorbitol, 150 mM KCl, 10 mM MgCl 2 , 2.5 mM EDTA, 2 mM ATP, 2 mM NADH, 20 mM K ϩ HEPES, pH 7.4). Where indicated, the potential across the mitochondrial inner membrane was dissipated with 1 M valinomycin. Nonimported radiolabeled proteins were removed by treatment with 100 g/ml trypsin for 30 min on ice; trypsin was inhibited with 400 g/ml soybean trypsin inhibitor. Import reactions were separated by SDS-PAGE followed by fluorography. Import was quantified by scanning laser densitometry (Personal Densitometer SI; Amersham Biosciences) and ImageQuant (version 4.2a; Amersham Biosciences).

RESULTS
We performed a phylogenetic analysis of five predicted genes in the C. elegans genome with sequence similarity to the small Tim family of proteins and Tom7p with S. cerevisiae, Mus musculus, and Homo sapiens. Specifically, we assembled dendrograms for the annotated genes Y93A3CR.4, C06G3.11, a non-annotated gene found on BAC B0564 (32), Y66D12A.22, and ZK652.2 here referred to as ddp-1, tin-9.1, tin-9.2, tin-10, and tomm-7, respectively (Fig. 1A); trees were assembled using ClustalW (33). Similar to other eukaryotic systems, the C. elegans genome codes for at least one member of each small Tim family: one from the Tim8 family, two Tim9 isoforms, one Tim10/12 family member, and one from Tim13. Therefore, C. elegans shares a similar complement of components in the conserved TIM22 pathway for the import of polytopic inner membrane proteins.
The tin-9.2 gene had not been annotated previously, although an SL1 spliced tin-9.2 sequence is present on cDNA EST yk1059h04.5. To confirm that the gene product of this cDNA could be translated and was correctly targeted to the mitochondria, we placed the cDNA clone upstream of and inframe with CFP under the control of the myo-3 gene promoter for expression in muscle cells. This construct was co-injected with a plasmid bearing the rol-6(su1006) mutation which causes the animals carrying the transgenes to move twisting in a circle rather than crawling. TIN-9.2::CFP accumulated in tubular structures in the muscle cells of transgenic animals ( Fig. 1B) and co-localized with the mitochondria-specific dye rhodamine 6G (data not shown). Thus, TIN-9.2 localizes to mitochondria as expected.
Isolation of Import-competent Mitochondria from C. elegans-Gene knock-down of the translocation components was obtained by RNAi treatment. The transcript levels for the specific genes were down-regulated in RNAi-treated but not control animals when assessed by quantification of RT-PCR products (data not shown). To establish the effectiveness of RNAi treatment and to demonstrate that the C. elegans homologs function as they do in yeast, we developed a protocol that yields import-competent mitochondria. The mitochondria were purified from adult C. elegans essentially as described previously (34). However, C. elegans are grown in a medium containing high levels of Ca 2ϩ that can disrupt mitochondrial function, so we substituted the Ca 2ϩ -chelator EGTA for EDTA in the isolation buffer. In organello import assays are well established for fungi and have been instrumental for the fundamental understanding of mitochondrial assembly in a single cell (31). [ 35 S]Methionine and [ 35 S]cysteine precursors for porin, ADP/ATP carrier (AAC), and the synthetic precursor pSu9-DHFR were synthesized in vitro for in organello import assays (see "Experimental Procedures"). The established yeast assay uses porin to test for import into the mitochondrial outer membrane. A membrane potential is not required for this translocation. For all import assays nonimported precursor is removed by protease treatment. Porin was imported into the mitochondrial outer membrane in both nematode and control yeast mitochondria (Fig. 2). Insertion into the outer membrane was confirmed by alkali extraction and flotation through a sucrose density gradient (35).
The established yeast assay uses the AAC to test for import into the mitochondrial inner membrane. Succinate was added to the mitochondria to maintain a membrane potential and the uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone and K ϩ ionophore valinomycin were added to dissipate the membrane potential. The inner membrane substrate AAC was imported into the inner membrane of the purified C. elegans mitochondria in the presence of a membrane potential (⌬⌿) (Fig. 2).
The insertion was confirmed by alkali extraction (36). The synthetic precursor pSu9-DHFR was imported into the matrix of the purified C. elegans mitochondria in a ⌬⌿-dependent manner, as it is in yeast, and the presequence was cleaved by the matrixprocessing protease (37). In sum, these in organello import assays in samples from C. elegans render this system amenable to the study of mitochondrial protein import.
Mitochondria from Worms with RNAi-mediated Gene Knockdown of tin-9.1 and tin-10 Are Defective in Inner Membrane Protein Import-In yeast, Tim9p and Tim10p are essential proteins involved in the TIM22 import pathway that facilitates the insertion of polytopic inner membrane proteins, such as AAC (14,18). To test whether the nematode homologs TIN-9.1 and TIN-10 also mediate the import of AAC, we used feeding RNAi to down-regulate TIN-9.1 and TIN-10 in C. elegans, and subsequently mitochondria were purified from treated adults. As expected, pSu9-DHFR import and maturation were similar in the RNAi-treated and wild-type mitochondria (Fig. 3, top  panel). However, the import of AAC was decreased in mitochondria isolated from tin-9.1(RNAi) and tin-10(RNAi) worms by 76 and 81%, respectively, in comparison to mitochondria from control adults (Fig. 3, bottom panel). This severe decrease suggests that TIN-9.1 and TIN-10 are functionally conserved components of the TIM22 import pathway. In addition, these biochemical studies demonstrate that the RNAi treatment is successful for decreasing TIN-9.1 and TIN-10 expression levels.
Phenotypes Associated with Gene Knock-down of Mitochondrial Import Components-To better understand the role that mitochondrial protein translocation has on the development of a multicellular organism, we utilized RNAi to disrupt the function of TIN-9.1 and TIN-10 as in Fig. 3, as well as TOMM-7, DDP-1, and TIN-9.2. Decreases in nuclear encoded mitochondria-associated gene function lead to an increase in life span (23,38,39) and decreased fertility (30,40). The mean life span of tomm-7(RNAi), ddp-1(RNAi), and tin-9.1(RNAi) animals showed no significant difference compared with control animals (Fig. 4). The mean life span of tin-9.2(RNAi) and tin-10(RNAi) animals was decreased 10 and 18%, respectively. We interpret this decrease in life span to be a result of sickness, rather than accelerated aging because these animals exhibit ]cysteine-radiolabeled precursors, porin, AAC, and pSu9-DHFR, were synthesized in a reticulocyte lysate system and incubated with coupled mitochondria isolated from C. elegans or S. cerevisiae. Equal aliquots of the import reaction were removed at the specified time points and treated with trypsin to remove nonimported precursor. The membrane potential (⌬⌿) was dissipated with valinomycin treatment. Porin insertion into the outer membrane was confirmed by carbonate extraction and sucrose flotation (35). As a control, an aliquot of the porin import reaction was treated with 0.1% Triton X-100. Inner membrane insertion of AAC was confirmed by extraction with 0.1 M Na 2 CO 3 , pH 11.0 (36). The Su9-DHFR precursor (p) was processed to the mature (m) form when imported into the mitochondrial matrix in the presence of ⌬⌿. Reactions were separated by SDS-PAGE and visualized by fluorography. As a standard (S), 10% of the translation reaction added to mitochondria was included. slower movement and increased incidence of protruding vulva (Pvl) phenotypes.
To assess fertility, we determined the brood size and the fraction of embryonic lethality over three sequential generations (Table I). tomm-7(RNAi), ddp-1(RNAi), and control animals showed similar brood sizes and no embryonic lethality in all generations tested. In contrast, tin-9.1(RNAi) worms showed a decrease in average brood size of 51% in the first generation compared with control animals, 38% in the second generation, and 39% in the third generation. tin-10(RNAi) worms also had smaller brood sizes, and this phenotype was more penetrant than in tin-9.1(RNAi) worms in subsequent generations. Unlike the other translocation components tested, tin-10(RNAi) worms had decreased brood sizes and increased embryonic lethality in the second generation, and the third generation animals that grew to adulthood were sterile. First generation tin-9.2(RNAi) worms showed no defects in brood size. However, the subsequent F2 and F3 progeny had average brood sizes decreased by 59 and 47%, respectively, compared with control animals. The RNAi animals that presented a sterile phenotype were not included in the calculation of the average brood size. The seemingly high standard deviations recorded for the observed phenotypes probably result in part from the combined occurrence of decreased fertility and embryonic lethality. Most interestingly, there were shared reproductive defects in the second and third generation associated with tin-9.1(RNAi), tin-9.2(RNAi), and tin-10(RNAi) worms, which are presumed to function together in the import pathway for polytopic inner membrane proteins. The decrease in viable eggs and larvae is expected, because the homologous genes are essential for viability in yeast.
A Sma Phenotype Associated with Decreased tin-9.1, tin-9.2, and tin-10 -Gene knock-down of the respiratory complexes in C. elegans results in decreased ATP production and a Sma phenotype (23,38). Considering that the TIM22 pathway indirectly contributes to the assembly of the respiratory complexes, we were curious whether RNAi of the mitochondrial protein import complexes may also lead to smaller body sizes. We observed that tin-9.1(RNAi), tin-9.2(RNAi), and tin-10(RNAi) worms had a smaller body size (Sma phenotype) compared with the control, tomm-7(RNAi), and ddp-1(RNAi) animals (Fig. 5). In addition, to the Sma phenotype tin-9.1(RNAi), tin-9.2(RNAi), and tin-10(RNAi) animals reached adulthood and started laying eggs 24 -48 h later than the control, tomm-7(RNAi), and ddp-1(RNAi) animals. Thus, defective inner mitochondrial membrane import gives rise to similar phenotypes to those of RNAi of the respiratory complexes.
Gonad Cell Migration Defects in Worms fed dsRNA Against tin-9.2-In addition to decreased brood sizes and a Sma phenotype, feeding of dsRNA tin-9.2 resulted in defects in the symmetrical migration of the distal tip cells (Mig phenotype) that form the rotationally symmetrical gonad arms. The symmetry of each gonad arm is formed by migration of the distal tip cell (DTCs) at the anterior and posterior ends of the gonad primordium to form a U-shaped structure (41). Control fed worms had normal gonad structures where during larval development the DTCs of the anterior arm (Fig. 6, gray line) and the posterior arm (black line) migrate away from the mid-body on the ventral wall where the vulva is formed (black square). The DTCs then turn and migrate back toward the mid-body against the dorsal wall, and finally the two DTCs stop migrating before reaching the midbody. tin-9.2(RNAi) worms were defective in forming this wild-type U-shaped structure. Either one or both gonad arms failed to migrate properly, and four morphological variations (type A-D) to this patterning were observed both on anterior and posterior arms (Fig. 6).
The type A morphology occurred when the DTCs initially failed to migrate away from each other but instead followed the same path toward one end of the animal. The type B and C gonad morphologies were most wild type-like in that the DTCs initially migrated properly along the ventral wall and completed the turn but prematurely stopped migrating along the ventral wall in type B or in type C variants; one DTC failed to stop migrating at the mid-body resulting in one gonad arm overgrowing the other. The type D morphology was the most severe of the four and was present when one DTC formed a wild-type-like structure, but the second DTC after migrating along the dorsal wall completed a second 180°turn and migrated back toward the end of the worm. Within the somatic gonad, the C. elegans germ line is initially a noncellularized syncytium of nuclei surrounded by mitochondria. Possibly the somatic gonad fails to migrate properly if the syncytial mitochondria fail to replicate appropriately due to tin-9.2 RNAi.
Mitochondrial Morphology Phenotypes Associated with Gene Knock-down of Mitochondrial Import Components-Defects in mitochondrial protein import result in secondary defects in mitochondrial morphology (42). In S. cerevisiae Tom7p is not required for viability and is involved in the assembly of the TOM complex and integration of proteins into the outer membrane (43). No reproductive, life span, or anatomical defects were observed in tomm-7(RNAi) worms ( Fig. 5 and Table I). To   FIG. 3. RNAi of TIN-9.1 and TIN-10 results in an import defect specific for the inner membrane substrate AAC. Import reactions were performed as in Fig. 2, except that mitochondria were isolated from N2 worms fed E. coli expressing dsRNA for tin-9.1 or TIN-10 or no dsRNA (control). Import was quantitated by scanning laser densitometry, and the % import was normalized to the longest time point for control mitochondria in the presence of ⌬⌿. STD, a standard of 10% of the translation reaction added to mitochondria; p, precursor; m, mature.

FIG. 4. RNAi of mitochondrial import components does not influence life span.
Eggs from N2 OP50 fed worms were isolated and allowed to hatch on plates seeded with E. coli expressing dsRNA specific for mitochondria translocation components. The plates were incubated at 25°C and monitored every 24 h.

Defects Due to Mitochondrial Dysfunction in C. elegans
determine whether mitochondrial assembly is grossly normal or not with putatively impaired outer membrane import complexes, mitochondrial morphology was investigated. We fed dsRNA against tomm-7 to a C. elegans transgenic strain expressing a yeast Tom20::YFP fusion protein in muscle cells under the myo-3 promoter (27) (YFP::mito) (Fig. 7). Normal muscle cells have mitochondria that run parallel to the body wall across the length of the cell (Fig. 7, control). Mitochondria in the tomm-7(RNAi) animals are more interconnected (Fig. 7), similar to that observed in mitochondria from animals expressing mutant DRP-1 (27). Thus, there are defects in mitochondrial morphology with gene knock-down of tomm-7.
Taken together, reduced expression of components involved in mitochondrial protein import in worms results in varied developmental abnormalities. Down-regulation of tomm-7 of the TOM complex and dpp-1 of the TIM22 pathway resulted in mitochondrial morphology defects. The tin-9.2(RNAi) animals were defective in the formation of the somatic gonad. Finally, we observed strong defects in the in vitro import of AAC into mitochondria isolated from tin-9.1(RNAi) and tin-10(RNAi) animals. The in vivo result of this import defect includes reduced brood sizes, embryonic lethality, small body size, and slow developmental rates. DISCUSSION The protein translocation field has utilized S. cerevisiae as a model organism because it is amenable to genetic and biochemical manipulations. Yeast, however, is inadequate for developmental studies. C. elegans is a relatively simple multicellular organism that is well established for genetic manipulation but has been under utilized for biochemical approaches. We demonstrate that coupled mitochondria can be isolated from C. elegans adults for in organello import assays and that genes  , tomm-7, ddp-1, tin-9.1, tin-9.2, or tin-10, were fed to strain N2.
b Brood sizes were scored for 20 individual F2 and F3 worms given a tin-9.1, tin-9.2, or tin-10 RNAi E. coli food source. Of these 20, (n) developed into fertile adult hermaphrodites. Sets less than 20 are due to arrest of the animal at a sexually immature stage or complete sterility.
c None of the F3 worms fed a tin-10 RNAi E. coli food source developed past the L3 larval stage.
FIG. 5. Gene knock-down of TIN-9.1, TIN-9.2, and TIN-10 results in a Sma phenotype. RNAi fed worms were generated as in Fig.  4 except that at adulthood the worms were moved to a fresh dsRNA expressing E. coli plate, and the F2 progenies were scored for developmental defects. F2 progenies fed dsRNA complementary to tin-9.1, tin-9.2, and tin-10 were Sma and were developmentally delayed by ϳ1-2 days.
FIG. 6. Gonads of hermaphrodites fed dsRNA of TIN-9.2 show cell migration defects. N2 worms were fed dsRNA as in Fig. 5, and the F2 progenies were visualized by Nomarski microscopy. The anterior gonad arm is represented by a gray line and the posterior gonad arm by a black line. Larvae fed E. coli expressing tin-9.2 dsRNA develop abnormal adult gonadal morphologies where both gonad arms migrate toward the same end of the worm (type A), the anterior and posterior arms fail to fully migrate back to the center of the worm (type B), one gonad arm fails to stop migrating at the worm mid-body (type C), or one gonad arm after migrating back to the worm mid-body turns 180°and continues to migrate away from the mid-body (type D). The entire gonad was inspected and is represented by the line drawing to the right of each microscopic image. The Nomarski images display the area designated by the black boxes. The white arrows designate gonad arm bends, and the black arrows designate the distal tip cells.
homologous to those identified in S. cerevisiae are involved in this process. By using RNAi, we observed a variety of developmental abnormalities associated with the gene knock-down of these specific import components. This study thus increases the repertoire of model systems for studies in mitochondrial protein translocation and provides insight into the developmental consequences associated with defective mitochondrial biogenesis. Ultimately, this will aid in understanding some of the human mitochondrial diseases that are linked to mutations in nuclear coded proteins that are imported to their correct location within the mitochondria (40).
Conserved Function for the Small TIM Complexes in C. elegans-The import pathway for polytopic proteins into the mitochondrial inner membrane has been well characterized in yeast (18,19,45). The small Tim proteins, Tim8p, Tim9p, Tim10p, and Tim13p, function as intermembrane space chaperones to guide inner membrane substrates to the TIM22 insertion complex at the inner membrane. Tim9p and Tim10p are essential for viability and assemble into a 70-kDa complex and are components of the TIM22 insertion complex at the inner membrane. Tim8p and Tim13p are not required for viability and assemble into a distinct 70-kDa complex. Loss-of-function mutations in the human Tim8p homolog DDP1 cause the neurodegenerative Mohr-Tranebjaerg syndrome (24,25,46). Studies in mammals suggest that the import pathway is conserved (32). By utilizing the in organello import assay, we show that RNAi treatment of C. elegans tin-9.1 and tin-10 leads to a specific defect in the import of AAC, providing evidence that the TIM22 import pathway is conserved in nematodes. That similar phenotypes were observed from the tin-9.1 and tin-10 RNAi treatments is consistent with these gene products functioning together, which implies mechanistic conservation between species.
Developmental Defects Associated with Insufficient Import to the Inner Mitochondrial Membrane-We found that RNAi knock-down of import components leads to developmental defects. Furthermore, the decreased brood size and embryonic lethality observed in tin-9.1(RNAi), tin-9.2(RNAi), and tin-10(RNAi) worms were more potent in later generations. This is most likely due to a combination of factors. First, RNAi treatment generally reduces the mRNA such that mitochondrial function is not abolished. In addition, there are different efficacies of RNAi gene knock-down by cell type such that the extent of mitochondrial function would depend upon cell type (47). Furthermore, the effects of RNAi of most mitochondrial proteins may not become evident until the maternally contributed factors are depleted. Thus, eventually the maternal contribution of mitochondria to the subsequent generations provide mitochondria with insufficient function to sustain the rapid cell divisions of embryogenesis and lethality results.
In the first generation of RNAi for specific translocons in C. elegans, certain cell types were more affected than others. There is precedence for this in that many tissues are unaffected in human mitochondrial diseases. The tissue-specific phenotypes might be due to differing dependence upon mitochondrial functions (energy or other metabolic products) for the different cell types in a multicellular organism. Thus, the effects of reduced mitochondrial protein import would be predicted to present differing phenotypes for various tissues. Alternatively, the expression of the translocons components could be different in the different tissues, as is the case in humans for the small Tim proteins (20).
In particular, the most obvious tissue-specific phenotype was that of the misguided gonads (Fig. 6). The C. elegans gonad contains a noncellularized syncytium where the germ line nuclei divide, and they are surrounded with mitochondria. Perhaps because this organ is densely packed with mitochondria, reduction in mitochondrial biogenesis disrupts proper formation. We speculate that a key mitochondrial influence in relation to the Mig phenotype is altered steroid biogenesis. Mutations in the daf-12 and daf-9 genes present Mig phenotypes (48,49). The daf-12 gene encodes a polypeptide with sequence similarity to steroid hormone receptors (50,51). daf-9 encodes a polypeptide with sequence similarity to cytochrome P450s (52). Cytochrome P450s predominantly localize to the endoplasmic reticulum and are involved in steroid biosynthesis. However, steroidogenic acute regulatory protein and cytochrome P450s are mitochondrial proteins involved in the conversion of cholesterol to progenolone (53,54). Steroidogenic acute regulatory protein can transfer cholesterol from liposomes to the endoplasmic reticulum or to mitochondria (53,54). Therefore, the Mig defects observed in tin-9.2(RNAi) adults may be a result of defective cholesterol trafficking or steroid biogenesis.
Alternatively, the Mig phenotypes presented in the tin-9.2(RNAi) animals could be the result of decreased import of TIN-9.2 substrates into the inner membrane. The prohibitin (PHB) family of proteins form high molecular weight complexes in the mitochondrial inner membrane (55). RNAi of the prohibitin subunits in C. elegans results in defects in embryonic development, body size, and germ line differentiation in the gonad (56). The similar phenotypes observed for RNAi of TIN-9.2 in this study and RNAi of the PHB proteins suggest TIN-9.2 might be involved in the import of the PHB proteins into the inner membrane. Although RNAi of the PHB proteins did not result in a Mig phenotype, it is reasonable to assume in tin-9.2(RNAi) worms a decreased import of the PHB proteins in combination with a decrease in other TIN-9.2 substrates could exacerbate defects in gonad formation.
The expression level of the gene product of C06G3.11 (tin- FIG. 7. dsRNA against tomm-7 and ddp-1 disrupts mitochondrial morphology in C. elegans muscle cells. Worms expressing TOM-20 fused to yellow fluorescent protein driven by the myo-3 promoter were fed E. coli expressing dsRNA specific for tomm-7 and ddp-1. Worms were mounted on agarose slides, and mitochondrial morphology was visualized with a Zeiss Axiovision 200 microscope with a fluorescein isothiocyanate filter set. Muscle cells typically have tubular mitochondria that run parallel with the body axis (control). F1, F2, and F3 progeny fed dsRNA against tomm-7 or ddp-1 had similar mitochondrial fission defects resulting in more interconnected organelles. The morphology defects were similar for a matrix::CFP marker (data not shown). Scale bar ϭ 10 m. 9.1) and the upstream region of B0564.1 that contains the 3Ј end of the tin-9.2 transcript were analyzed in a genome-wide transcriptional profiling study by Hill et al. (57). Analysis of these data reveal that tin-9.1 is present in higher abundance (1.5-3.2-fold) throughout the development and life cycle of the animal. Consistent with the role of TIN-9.2 in the development of the gonad, the highest expression of the tin-9.2 message was detected in L2 and L3 larval stages. Gonad morphogenesis occurs in the L2-L4 stages (58).
Most interestingly, we did not observe an increase in the mean life span of tomm-7(RNAi), ddp-1(RNAi), tin-9.1(RNAi), tin-9.2(RNAi), and tin-10(RNAi) animals. It has been shown that defects in many genes associated with mitochondrial function lead to life span extension. RNAi of the protein translocases, which import all nuclear coded mitochondrial proteins, should ultimately decrease the abundance of multiple mitochondrial components. The apparent inconsistency in the length of adult life span in animals with impaired mitochondria is probably not a result of defects in the OXPHOS but rather a global reduction of mitochondrial processes. In our RNAi experiments it is likely a level of protein import occurs in the first generation, because some mitochondrial protein translocons are present as indicated in the translocation assay (Fig. 3). The adult life span was measured in this first generation that had sufficient mitochondrial function to develop to adulthood; therefore, they were Sma, slow growing, and less fertile than control animals.
Control of Mitochondrial Morphology by Protein Translocases-Mitochondrial morphology is a regulated process that is coordinated by fission and fusion events (56). But how would the protein translocases of the mitochondria contribute to the regulation of these events? In yeast, Fzo1p, Ugo1p, and Mgm1p regulate mitochondrial morphology at the outer membrane. These proteins likely interact with TOM translocase and possibly Tom7p for proper localization and integration into the outer membrane. Recent evidence has shown some outer membrane proteins may also interact with the small TIM complexes in the intermembrane space (59,60). In addition, a new group of proteins including Pcp1p, Mdm31p, Mdm32p, Mdm39p, and Ymr1p regulate mitochondrial morphology from the inner membrane in yeast (42,61,62). The localization of these inner membrane proteins should require an interaction with the small TIM complexes during import. Therefore, defects in the assembly of the protein translocation components could result in an aberrant morphology. Indeed, the depletion of the protein products of tomm-7 and ddp-1 disrupted the normal mitochondrial morphology (Fig. 7). Although the morphological defects associated with RNAi of ddp-1 and tomm-7 are different from the defects associated with depletion of the prohibitin complex and OXPHOS, these studies suggest that defects in mitochondrial morphology can manifest in the absence of direct manipulation of the mitochondrial fission and fusion machinery in C. elegans (23,55,56).
The work presented here combines reverse genetic and biochemical approaches to characterize the effects of defective mitochondrial protein import in a multicellular organism. The in vitro mitochondrial protein import assays demonstrate reduction of activity and are coincident with the developmental abnormalities observed in the RNAi-treated animals. These observations also prove that the import pathway for mitochondrial inner membrane proteins, which is shared among eukaryotes, is functionally conserved in metazoans.