Human Molecular Genetics, 2015, Vol. 24, No. 2 doi:10.1093/hmg/ddu468 Advance Access published on September 11, 2014
480–491
Tissue-specific responses to the LRPPRC founder mutation in French Canadian Leigh Syndrome Florin Sasarman{, Tamiko Nishimura{, Hana Antonicka, Woranontee Weraarpachai and Eric A. Shoubridge∗ , LSFC Consortium Montreal Neurological Institute and Department of Human Genetics, McGill University, Montreal, QC H3A 2B4, Canada Received June 30, 2014; Revised August 1, 2014; Accepted September 9, 2014
French Canadian Leigh Syndrome (LSFC) is an early-onset, progressive neurodegenerative disorder with a distinct pattern of tissue involvement. Most cases are caused by a founder missense mutation in LRPPRC. LRPPRC forms a ribonucleoprotein complex with SLIRP, another RNA-binding protein, and this stabilizes polyadenylated mitochondrial mRNAs. LSFC fibroblasts have reduced levels of LRPPRC and a specific complex IV assembly defect; however, further depletion of mutant LRPPRC results in a complete failure to assemble a functional oxidative phosphorylation system, suggesting that LRPPRC levels determine the nature of the biochemical phenotype. We tested this hypothesis in cultured muscle cells and tissues from LSFC patients. LRPPRC levels were reduced in LSFC muscle cells, resulting in combined complex I and IV deficiencies. A similar combined deficiency was observed in skeletal muscle. Complex IV was only moderately reduced in LSFC heart, but was almost undetectable in liver. Both of these tissues showed elevated levels of complexes I and III. Despite the marked biochemical differences, the steady-state levels of LRPPRC and mitochondrial mRNAs were extremely low, LRPPRC was largely detergent-insoluble, and SLIRP was undetectable in all LSFC tissues. The level of the LRPPRC/SLIRP complex appeared much reduced in control tissues by the first dimension bluenative polyacrylamide gel electrophoresis (BN-PAGE) analysis compared with fibroblasts, and even by second dimension analysis it was virtually undetectable in control heart. These results point to tissue-specific pathways for the post-transcriptional handling of mitochondrial mRNAs and suggest that the biochemical defects in LSFC reflect the differential ability of tissues to adapt to the mutation.
INTRODUCTION Deficiencies in oxidative phosphorylation (OXPHOS) are associated with a diverse array of multisystem disorders that are often referred to as mitochondrial encephalomyopathies because of the prominent involvement of the nervous system and striated muscle (1 – 3). The molecular basis for the extraordinarily broad clinical spectrum observed in OXPHOS disorders remains an enduring mystery. Mutations in mtDNA are the most frequent cause of mitochondrial disease in adults and more than 100 such pathogenic mutations have been identified. In the pediatric population, the majority of OXPHOS disorders (80%) are transmitted as autosomal recessive traits, usually with a severe phenotype and a fatal outcome. Leigh syndrome is one of the most common presentations in this population. First described by Denis Leigh in a single case
(4), the so-called classical form of the disease is characterized by early-onset neurodegeneration, particularly in the brain stem and basal ganglia. The genetic underpinnings of this disorder are found in structural subunits of the OXPHOS complexes, associated assembly factors or in pyruvate dehydrogenase (5,6). A variant known as Leigh Syndrome French Canadian (LSFC) has a characteristic pattern of tissue involvement and a different clinical evolution, with acute acidotic crises usually leading to very early fatality (7 –9). This disorder is caused by a founder missense mutation in LRPPRC, a gene coding for a pentatricopeptide repeat (PPR) family protein (10). PPR proteins, which belong to the alpha-solenoid protein family, have undergone a huge radiation in land plants where they number several hundred (11). Virtually all of them localize to mitochondria or chloroplasts where they mediate a large variety of RNA transactions including splicing, editing, stability and translation
∗
To whom correspondence should be addressed. Tel: +1 5143981997; Fax: +1 5143981509; Email: [email protected] Co-first authors.
†
# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Human Molecular Genetics, 2015, Vol. 24, No. 2
initiation. A combinatorial code that specifies the recognition of specific RNA substrates has been determined for the chloroplast PPR protein PPR10 (12), a result that was substantiated by the solution of the crystal structure of the protein bound to its mRNA substrate (13). Humans have only seven identified PPR proteins, all of which are mitochondrial (14). LRPPRC forms a ribonucleoprotein (RNP) complex of about 250 kDa with the stem-loop RNA-binding protein SLIRP, and this complex appears necessary to stabilize the majority of polyadenylated mitochondrial mRNAs (15– 17), protecting them against 3′ exoribonuclease digestion (16). In vitro studies with purified LRPPRC also show a stimulation of polyA addition to RNA substrates, suggesting that the complex might play a dual role in the synthesis and maintenance of the 3′ ends of mitochondrial mRNAs (16). The protein carrying the A354V substitution appears unstable in fibroblasts, as the steady-state level is approximately 30% of control (15). Depletion of wild-type LRPPRC to the same level phenocopies the defect seen in LSFC fibroblasts, suggesting that the mutant protein is functional; however, further siRNA-mediated depletion results in an assembly defect in all OXPHOS complexes with mtDNA-encoded subunits (15). These data suggest that tissue-specific differences in the level of mutant LRPPRC might account for the tissuespecific biochemical deficiencies characteristic of LSFC. We tested this hypothesis in cells and tissues obtained from LSFC patients and show that the situation is much more complex. Tissue-specific involvement appears to reflect, at least in part, the ability of different tissues to adapt to the genetic defect.
RESULTS LRPPRC and SLIRP are interdependent and interact with a 1:1 stoichiometry Previous work in human fibroblasts (15) and mouse liver (17) demonstrated that LRPPRC functions in a RNP complex with the stem-loop RNA-binding protein SLIRP. A decrease in steady-state levels of either protein results in a similar decrease in the levels of the other, and a reduction of the levels of either LRPPRC or SLIRP causes a similar destabilization of the mitochondrial mRNAs (18). To further confirm this functional interdependence, we assessed the pattern of mitochondrial translation in control fibroblasts in which SLIRP was knocked down using two different siRNA constructs, both of which reduced SLIRP to barely immunodetectable levels (Fig. 1B). The mitochondrial translation defect phenocopied that observed in LSFC fibroblasts (15), where the most severe reductions were observed in the synthesis of COX I, II and III, and an anomalous translation product appeared (Fig. 1A). As expected, the decrease in the level of SLIRP was paralleled by a decrease in the level of LRPPRC. We determined the stoichiometry of interaction between LRPPRC and SLIRP by immunoprecipitating SLIRP in mitochondrial extracts from control fibroblasts, followed by mass spectrometry analysis. Using the abundance of five tryptic peptides in each protein, the relative ratio of LRPPRC (130 kDa) and SLIRP (12 kDa) was found to be 1:1, suggesting that 250 kDa RNP complex in fibroblasts may be a heterotetramer. PPR10, the only PPR protein whose crystal structure has been determined, was found to dimerize when bound to its RNA substrate (13).
481
LSFC cultured muscle cells have a combined OXPHOS deficiency To investigate the nature of the biochemical phenotype in another cell type, we cultured myoblasts obtained from two LSFC cases, differentiated them into myotubes, and immunoblotted for LRPPRC and SLIRP. Similar to LSFC fibroblasts, the level of mutant LRPPRC was reduced to 20– 30%, and the levels of SLIRP, to 10– 20% of control in LSFC myoblasts and myotubes (Figs 2A and 3A). The synthesis of the COX subunits was particularly affected in LSFC muscle cells, and an anomalous translation product appeared in post-natal LSFC myoblasts and myotubes (Figs 2B and 3B). However, unlike LSFC fibroblasts, all of the cultured muscle cells from LSFC subjects (fetal and post-natal myoblasts and myotubes) showed a combined OXPHOS defect, with reduced levels of assembled complex I and complex IV as shown by blue-native polyacrylamide gel electrophoresis (BN-PAGE) analysis (Figs 2C and 3C). To determine whether a further reduction in the steady-state level of LRPPRC would lead to a more severe OXPHOS deficiency in cultured muscle cells (as it does in fibroblasts), we used siRNA to deplete LRPPRC below 10% of control levels in control and LSFC myoblasts and myotubes. This resulted in a generalized OXPHOS defect in which all complexes with mtDNA-encoded subunits were affected, as demonstrated by immunoblot (Fig. 3A), mitochondrial translation (Fig. 3B) and BN-PAGE analyses (Fig. 3C). The smaller molecular weight complex V species in Figure 3C represent F1-containing subcomplexes. The reduction in steady-state levels of mitochondrial mRNAs does not correlate with the biochemical phenotype in LSFC tissues Given the role of LRPPRC in binding mature mitochondrial mRNAs, we started our analysis of LSFC tissues by northern blot analysis of the steady-state levels of mitochondrial mRNAs in the skeletal muscle, heart and liver tissues from an LSFC patient (Fig. 4A). In all three tissues, there was a generalized decrease in the levels of all analyzed mitochondrial mRNAs, the most severe reduction in mRNA levels being observed in skeletal muscle. Additionally, an increase in the levels of the 12S mitochondrial ribosomal RNA was observed in the three tested LSFC tissues, and an increase in the 16S mitochondrial ribosomal RNA, in the LSFC heart and liver, presumably indicating a compensatory response to the decrease in mitochondrial mRNA levels. The levels of mitochondrial tRNAs were indistinguishable from control in all three LSFC tissues (Fig. 4B). Loss of LRPPRC and SLIRP results in tissue-specific OXPHOS deficiencies BN-PAGE analysis of the LSFC tissues revealed an almost complete lack of assembled complex IV in LSFC liver (not detectable even in overexposed blots), and 50% of control levels of assembled complex IV in heart (Fig. 5A). Additionally, both of these tissues displayed an increase in the levels of assembled complexes I and III. The differences in intensity of the bands for complexes I, III, IV and V in the control samples reflect real
482
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 1. Knockdown of SLIRP in control fibroblasts phenocopies the mitochondrial translation defect in LSFC fibroblasts. (A) Control fibroblasts were transiently transfected with two different siRNA constructs specific to SLIRP, or with a fluorescent control siRNA (Alexa). On the sixth day post-transfection, the mitochondrial translation products were pulse-labeled with a mix of [35S]-methionine and cysteine in the presence of emetine, an inhibitor of cytoplasmic translation, and analyzed by PAGE. The 13 mitochondrial translation products are indicated to the left of the figure: seven subunits of complex I (ND), one subunit of complex III, three subunits of complex IV (COX) and two subunits of complex V (ATP). The asterisk indicates an anomalous, unidentified translation product typically detected in LSFC fibroblasts. (B) The knockdown of SLIRP and the concomitant reduction in steady-state levels of LRPPRC were confirmed by immunoblotting with specific antibodies. The 70 kDa subunit of complex II was used as a loading control.
difference in the relative abundance of these complexes between heart and liver, as the same amount of mitochondrial protein was loaded in all lanes and the level of complex II is similar in all lanes. On the other hand, in skeletal muscle we observed a combined complex I and IV deficiency, similar to that seen in cultured muscle cells (Fig. 5A). This pattern of OHPHOS deficiency was confirmed by immunoblotting cleared mitochondrial extracts from LSFC and control tissues, which showed decreases in the steady-state levels of structural subunits for the same complexes that showed assembly defects by BN-PAGE analysis (Fig. 5B). Despite the tissue-specific OHPHOS defects, mutant LRPPRC and SLIRP were virtually undetectable in all three analyzed LSFC tissues. An immunoblot analysis of the endogenous levels of LRPPRC and SLIRP in cleared mitochondrial extracts from control tissues showed comparable levels, relative to proteins in the OXPHOS complexes, in all three analyzed LSFC tissues (Fig. 5C). Detergent solubility of LRPPRC and a subset of proteins involved in mitochondrial translation is altered in LSFC liver Since mutant LRPPRC appeared to be present at levels below immunodetection in LSFC tissues, yet with such tissue-specific biochemical consequences, we asked whether residual LRPPRC might be present in a detergent-insoluble fraction, that is a fraction that could not be extracted with the concentration of the non-ionic detergent dodecyl maltoside (DDM), that we routinely use for immunoblot analysis of proteins, in some or all of these tissues. To test this, mitochondria from LSFC and control
tissues extracted in DDM were mixed with the loading buffer [containing sodium dodecyl sulfate (SDS)], electrophoresed by PAGE as usual, and subsequently immunoblotted. As shown in Figure 6A, while LRPPRC was still not detected in total extracts of LSFC muscle, residual mutant protein was present in LSFC heart and liver at 10% of control levels. It is important to point out that mutant LRPPRC could easily be extracted from fibroblasts or myoblasts with DDM, so this is a tissue-specific and not a technical problem. While the above findings might explain a combined complex I and IV deficiency in LSFC muscle versus an isolated complex IV defect in LSFC heart and liver, it still does not clarify why complex IV deficiency is severe in liver and mild in heart. Hence, we further investigated the distribution of LRPPRC and other proteins involved in mitochondrial translation between the soluble and insoluble fractions following detergent extraction, in LSFC and control liver. Interestingly, while LRPPRC was found predominantly (80 –90%) in the supernatant (soluble fraction) in control liver, the residual amount of mutant LRPPRC was detected exclusively in the pellet (insoluble fraction) in the LSFC liver (Fig. 6B). A similar reversal in distribution between supernatant and pellet and, in some cases, a decrease in total amounts was also observed for a subset of proteins involved in mitochondrial translation, including two translation factors (TACO1, EFG1), a mitochondrial ribosomal protein (MRPS27) and the mitochondrial formyltransferase, MTFMT (Fig. 6B). By contrast, another mitochondrial translation factor (CCDC56), the ribosomal protein MRPL44, the mitochondrial peptide deformylase (PDF) and two subunits of OXPHOS complexes (SDHA, ATP5A1) displayed a normal distribution between the soluble
Human Molecular Genetics, 2015, Vol. 24, No. 2
483
Figure 2. Mitochondrial translation defect and combined OXPHOS deficiency in LSFC cultured muscle cells. (A) Immunoblot analysis showing reduced levels of LRPPRC and SLIRP in extracts from fetal and post-natal cultured muscle cells from LSFC subjects. Prohibitin was used as loading control. The sample for LSFC fetal myotubes was slightly under loaded. (B) Mitochondrial translation products in fetal and post-natal myoblasts and myotubes from LSFC subjects were pulse-labeled with a mix of [35S]-methionine and cysteine in the presence of emetine, an inhibitor of cytoplasmic translation and analyzed by PAGE. The 13 mitochondrial translation products are indicated to the left of the figure: seven subunits of complex I (ND), one subunit of complex III, three subunits of complex IV (COX) and two subunits of complex V (ATP). The asterisk indicates a translation product with anomalous migration also detected in LSFC fibroblasts. (C) Blue-native PAGE analysis of the five OXPHOS complexes revealed reduced levels of assembled complex IV (COX) and complex I in cultured muscle cells from LSFC subjects. Individual complexes were detected with subunit-specific antibodies.
and insoluble fractions in the LSFC liver, demonstrating that this was not a general problem of protein aggregation. The mitochondrial apoptosis-inducing factor AIFM1 also showed a control pattern of distribution between fractions following detergent extraction of LSFC liver mitochondria, suggesting that the change in solubility of the subset of proteins in patient tissue is not linked to pro-apoptotic events. The different distribution of LRPPRC between fractions in LSFC liver does not appear to be the consequence of compromised mitochondrial membranes, as transmembrane proteins of the outer (VDAC1) and inner
(ATP5A1) membranes were extracted the same way in LSFC and control. Strikingly, SLIRP, which appeared only in the soluble fraction in control tissues, was not detectable by immunoblot analysis in any LSFC tissue. We confirmed the presence of LRPPRC in the insoluble fraction following DDM extraction of LSFC liver mitochondria by repeating the analysis with two additional controls (Fig. 6C). While a certain percentage of LRPPRC (between 20 and 50%) is associated with the pellet in control liver, in the LSFC liver, the residual, mutant LRPPRC is found exclusively in the
484
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 3. Generalized OXPHOS defect following knockdown of LRPPRC in LSFC cultured muscle cells. LSFC and control myoblasts or myotubes were transiently transfected either with an siRNA construct specific to LRPPRC, or with a fluorescent control siRNA (Alexa), and then (A) analyzed by immunoblotting with antibodies against LRPPRC, SLIRP or subunits of the OXPHOS complexes, (B) pulse-labeled with a mix of [35S]-methionine and cysteine in the presence of emetine and (C) analyzed by BN-PAGE (as described in Fig. 2).
pellet. A similar analysis of heart mitochondria revealed that all tested proteins are found predominantly (80%) in the insoluble fraction following DDM extraction, in both control and LSFC tissues (data not shown). LRPPRC and SLIRP are partially associated with mitochondrial membranes The presence of LRPPRC in the insoluble fraction following DDM extraction, even in control mitochondria, prompted us to investigate whether LRPPRC and/or SLIRP are associated with the mitochondrial membranes. While neither of these two proteins is predicted to contain a transmembrane domain, alkaline carbonate extraction of isolated mitochondria revealed that up to 50% of LRPPRC and SLIRP are found in the membrane fraction in HEK293 cells (Fig. 7A) and in fibroblasts, whether control or LSFC (Fig. 7B).
The levels of mitochondrial small ribosomal subunits and the monosome are decreased in LSFC liver The alteration of the detergent solubility of a subset of proteins involved in mitochondrial translation in the LSFC liver, including a protein of the small ribosomal subunit (mt-SSU), MRPS27 (Fig. 6B), might reflect a qualitative or quantitative change of the mitochondrial ribosomes in this tissue. To test this, we performed sucrose gradient centrifugation of cleared mitochondrial extracts from control and LSFC liver. Consistent with the decrease in levels and detergent solubility of MRPS27, the amounts of fully assembled small subunit (mt-SSU) and the monosome were decreased in the LSFC liver, while the levels of assembled large ribosomal subunits (mt-LSU) were normal (Fig. 8A). In contrast, in the LSFC heart mitochondria there was an increase in the steady-state levels of subunits of the small and large ribosomal subunits and of assembled small and
Human Molecular Genetics, 2015, Vol. 24, No. 2
485
Figure 4. Generalized reduction of steady-state levels of mitochondrial mRNAs in LSFC tissues. Total RNA was extracted from muscle, heart and liver of controls and an LSFC subject, followed by either standard northern blotting analysis with probes specific for the mitochondrial mRNAs or rRNAs indicated at the left of the panel (A), or by urea-PAGE and hybridization with oligonucleotide probes complementary to the mitochondrial and cytoplasmic tRNAs indicated at the right of the panel (B).
large ribosomal subunits (Fig. 8B and C), suggesting an adaptive response to the decrease in the levels of mRNAs in this tissue.
The LRPPRC/SLIRP complex is not detectable in heart tissue We next asked whether the LRPPRC/SLIRP complex is preferentially stabilized in the LSFC heart. First dimension blue-native PAGE analysis showed that 250 kDa complex of LRPPRC and SLIRP was undetectable with antibodies against LRPPRC in the LSFC liver as compared with control (Fig. 9A), whereas in the heart, it was undetectable in both LSFC and control mitochondria (data not shown). This led us to hypothesize that the reason why heart is spared in LSFC might be a quantitative or qualitative difference of the LRPPRC/SLIRP complex in this tissue, as opposed to liver or muscle. To test this, we analyzed mitochondria from heart, liver and muscle of two controls, and from two control fibroblast lines, by the first and second dimension BN-PAGE. Indeed, the LRPPRC – SLIRP complex was undetectable in the heart of both controls in the first dimension and was greatly reduced in muscle or liver as compared with fibroblasts (Fig. 9B). This contrasts with the levels of assembled complex IV, which are highest in heart and muscle, and lowest in liver and fibroblasts (Fig. 9B and (19)). Additionally, the LRPPRC–SLIRP complex in muscle migrates at a slightly higher molecular mass than in the other analyzed tissues and cells, suggesting it might have different properties or composition. Heart was the only tissue in which LRPPRC and SLIRP were equally undetectable in the second dimension (Fig. 9C). These observations suggest that heart has alternative pathways to the LRPPRC/SLIRP complex for post-transcriptional handling of mitochondria-encoded proteins.
DISCUSSION The results presented in this study demonstrate striking differences in the post-transcriptional handling of mitochondrial mRNAs both among tissues and between cultured cells and fully differentiated tissues. In fibroblasts, myoblasts and differentiated myotubes, we show that the LRPPRC–SLIRP complex is essential for mitochondrial translation and, as a consequence, assembly of all OXPHOS complexes containing mtDNA-encoded structural subunits. Reducing the level of LRPPRC to levels undetectable by immunoblot analysis abolishes translation and completely abrogates assembly of a functional respiratory chain. Depletion of SLIRP phenocopies the translation defect, emphasizing the interdependence of the two proteins. What was unexpected, and remains to be explained, is that despite very similar translation defects, LSFC fibroblasts and muscle cells have quite distinct biochemical phenotypes, the latter showing a marked reduction in complex I assembly, similar to that observed in patients with isolated complex I deficiencies, in addition to a complex IV assembly defect. This suggests that the basis for the different biochemical defects lies downstream of mitochondrial translation. What is most extraordinary, however, is the fact that in the three LSFC tissues that we were able to study, the steady-state levels of the mutant form of LRPPRC were extremely low, the protein was largely detergent-insoluble and SLIRP was undetectable (Note that we have no evidence that tissue-specific isoforms or paralogues exist for either protein). Despite this rather severe and relatively homogeneous phenotype, it seems evident that many of the mitochondrial mRNAs are translated to a degree that underwrites normal (or increased as for complexes I and III in heart and liver) assembly of some of the OXPHOS complexes, but that marked tissue-specific deficiencies persist. This occurs despite the fact that the steady-state
486
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 5. Tissue-specific OXPHOS deficiencies in LSFC tissues. Mitochondrial extracts from muscle, heart and liver of one LSFC subject, several controls and four disease controls were analyzed by BN-PAGE (A) and by western blotting (B) with antibodies against subunits of the OXPHOS complexes or against LRPPRC and SLIRP, as indicated. In (A), ‘Co I deficiency’ refers to two subjects with assembly defects of complex I; in (B), ‘COX15’ and ‘EFG1’ refer to two subjects with mutations in an assembly factor of COX and a mitochondrial translation factor, respectively. (C) Cleared mitochondrial extracts from control liver, heart and muscle were analyzed by immunoblotting with antibodies against LRPPRC, SLIRP or subunits of the OXPHOS complexes.
levels of the mRNAs were severely reduced in all three tissues, as might have been predicted from the results we obtained in fibroblasts showing the interdependence of LRPPRC, SLIRP and mRNA (15). The only overt phenotype reported in a recent study of a mouse germ line deletion of SLIRP was a small decrease in male fertility, although the study did not include any biochemical analyses of OXPHOS complex assembly or function (20). It thus appears that SLIRP may not be an essential component of the molecular machinery involved in the processing and handling of mitochondrial mRNAs, or that some
Figure 6. Altered detergent solubility of LRPPRC and of other mitochondrial translation proteins in LSFC liver. (A) Mitochondria from muscle, heart and liver samples of the LSFC subject and controls were extracted in dodecyl maltoside, then mixed with loading buffer containing SDS without extract clearing by centrifugation, followed by immunoblot analysis with antibodies against LRPPRC or the 70 kDa subunit of complex II. (B, C) Mitochondria isolated from liver samples of one LSFC subject and controls were extracted in dodecyl maltoside or taurodeoxycholate (for TACO1 detection), then loading buffer containing SDS was added to the extracts without extract clearing by centrifugation (‘Input’). Duplicate samples were centrifuged following extraction in dodecyl maltoside or taurodeoxycholate, and each soluble (‘Supernatant’) and insoluble (‘Pellet’) fraction was then mixed with loading buffer containing SDS. All samples were analyzed by immunoblotting with antibodies against LRPPRC, SLIRP, mitochondrial ribosomal proteins (MRPS27, MRPL44), factors or enzymes involved in mitochondrial translation (EFG1, TACO1, MTFMT, PDF, CCDC56) and other mitochondrial proteins, as detailed in the text.
component(s) of the translation can compensate for its absence. In any case, we were not able to convincingly demonstrate the presence of the LRPPRC/SLIRP complex in control heart tissue by either first or second dimension BN-PAGE analysis.
Human Molecular Genetics, 2015, Vol. 24, No. 2
487
Figure 7. LRPPRC and SLIRP are partly associated with the mitochondrial membrane fraction. Alkaline carbonate extracts of mitochondria from HEK293 cells (A) or from control and LSFC fibroblasts (B) were analyzed by immunoblotting with antibodies against LRPPRC, SLIRP and the following protein controls: COX I (multiple pass, integral inner membrane), SCO1 (single pass, inner membrane), VDAC (outer membrane), Co II-70 kDa (soluble, inner membrane-associated), HSP 60 and 75 (soluble).
These data lead us to two main conclusions. First, our data show that the way in which mitochondrial mRNAs are handled is tissue-specific. Although LRPPRC does not contain a predicted transmembrane domain, a significant fraction of it stays in the pellet after alkaline carbonate extraction in HEK cells and fibroblasts, and most of it remains DDM-insoluble in heart, but less so in liver. This suggests that LRPPRC may shuttle between a soluble matrix fraction and an inner membrane fraction that might depend on its association with SLIRP, and that this equilibrium shifts in some tissues in the presence of the mutant protein. The fact that the LRPPRC/SLIRP complex was more difficult to detect in the first versus second dimension BN-PAGE gels in muscle and liver, suggests that the epitope recognized by the LRPPRC antibody is less accessible in the native complex in tissues, perhaps because of a posttranslational modification, or the presence of another protein in the complex, but this remains to be investigated. Second, it seems clear that some tissues have the ability to adapt to the LRPPRC mutation better than others. The increase in the total number of monosomes in the heart is presumably a response to increase the efficiency of translation at lower steady-state levels of mRNAs. This signaling does not appear to function in the liver, as there were fewer monosomes in LSFC liver, due to a decrease in the assembled mt-SSU. We previously reported a re-organization of the translation elongation machinery in the heart of a patient with mutations in the translation elongation factor EFG1 (19). In this case, the ratio of EFTu:EFTs (the proteins involved in bringing the charged tRNAs to the ribosome) was completely reversed in the patient heart compared with control, driven by transcriptional reprogramming. These data suggest that there is a feedback mechanism that can adjust the functioning of the translation apparatus in some tissues, in this case in the face of a crippled polypeptide elongation factor EFG1. Why such adaptation does not occur in all tissues, to minimize the impact of the genetic mutation, remains unknown. Recently, it has become clear that the polycistronic transcripts transcribed from mtDNA are processed in discrete granules that have been defined by the presence of the RNA-binding protein
GRSF1 (21,22). Newly synthesized RNAs, visualized by pulse labeling with the nucleotide analog BrU, co-localize with GRSF1 (21). These foci also contain Suv3 and PNPase, the enzymes that are involved in mRNA degradation (23). Mitochondrial transcripts are matured by the activities of RNAseP and Z, which process the 5′ and 3′ ends of tRNAs, respectively (24,25). As tRNAs punctuate most of the open reading frames in mtDNA (26), this results in the release of mature mRNAs. It is not known how LRPPRC and SLIRP then associate with these processed RNAs, but it does seem clear that this association is necessary for mRNA stability. Immunofluorescence experiments with antibodies directed against either protein do not localize them to foci, but rather label the entire mitochondrial matrix. It is possible that either or both proteins shuttle between the foci and a free pool in order to deliver the mRNAs to ribosomes, which are thought to be tightly associated to the inner mitochondrial membrane where the mitochondrial polypeptides are co-translationally inserted into the inner membrane. This could explain why LRPPRC appears to exist both in a soluble and membrane-associated form. The nature of the biochemical defects in LSFC patients that we report here shows that it is not a pure complex IV defect as had been previously thought. The study raises many questions that will require transcriptomic and proteomic analyses to resolve. We have recently developed a knock-in mouse model, in which we have replaced the wild-type allele with the murine homolog of the LSFC founder mutation, and this should allow us to address the molecular basis of tissue specificity in a definitive fashion.
MATERIALS AND METHODS Cell culture and RNA interference experiments Myoblasts were cultured from muscle biopsies of two LSFC patients according to an established procedure (27). Primary myoblasts were FACS-sorted and then immortalized by transduction with two retroviral vectors, one expressing the HPV-16 E6 and E7 genes, and the other expressing the catalytic
488
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 8. Analysis of ribosome assembly in LSFC liver and heart. Cleared mitochondrial lysates from control and LSFC liver (A) and heart (B) were loaded on a discontinuous sucrose gradient and centrifuged, after which fractions were collected for immunoblot analysis with antibodies against subunits of the large (mitochondrial ribosomal protein, large) and small (mitochondrial ribosomal protein, small) mitochondrial ribosomal subunits. The approximate migration of the small (28S) and of the large (39S) mitochondrial ribosomal subunits and the monosome (55S) are indicated for reference. (C) Western blotting analysis of cleared mitochondrial extracts from control and LSFC heart with antibodies against small and large mitochondrial ribosomal subunits. One subunit of complex II and one of complex V were used as loading controls.
component of human telomerase (28). Myoblasts were grown in Ham’s F10-based, supplemented growth medium (29). Myotubes were generated by allowing confluent myoblasts to fuse for 2 – 4 days in dulbecco’s modified eagle medium (DMEM) medium supplemented with 2% horse serum and 0.4 mg/ml dexamethasone. Fibroblasts and HEK293 cells were grown in DMEM supplemented with 10% fetal bovine serum. For knockdown of SLIRP, two Stealth RNA interference duplexes (Invitrogen) were used: CGG CGU CGA GUC AGC UGA AAG AAC A and ACA GAG GUU UGG GUU GGG UUC AGU U. LRPPRC was knocked down with the Stealth RNA interference duplex Stealth1173 (15). Stealth RNAi duplexes were transiently transfected at a concentration of 25 nM according to manufacturer’s instructions for reverse
transfection. Transfection was repeated on day 3, and on day 6 cells were either harvested and analyzed (fibroblasts and myoblasts), or transferred to fusion medium and allowed to fuse for 2–4 days (myotubes) prior to analysis.
Radioactive labeling of mitochondrial translation products Mitochondrial translation products in cultured muscle cells or fibroblasts were pulse-labeled for 1 h with 200 mCi/ml of a mixture of [35S]-methionine and cysteine (Perkin Elmer) in DMEM lacking methionine and cysteine and containing 100 ml/ml emetine, an inhibitor of cytoplasmic translation, as detailed elsewhere (30).
Human Molecular Genetics, 2015, Vol. 24, No. 2
489
Figure 9. The LRPPRC– SLIRP complex is undetectable in normal heart tissue by the first and second dimension BN/SDS-PAGE analysis. (A) First dimension BN-PAGE analysis of the LRPPRC–SLIRP complex with an antibody against LRPPRC, and of complexes II and IV (COX) with subunit-specific antibodies, in control and LSFC liver. (B) Comparison of the levels of LRPPRC– SLIRP complex in control tissues and cells by the first dimension BN-PAGE analysis. The LRPPRC–SLIRP complex was detected with an antibody against LRPPRC and complex IV was detected with an antibody against its nuclear subunit COX IV. (C) Extracts from control tissues and cells were run in the first dimension on a non-denaturing, blue native gel (BN-PAGE), after which individual lanes were cut out and run on the second dimension on a denaturing gel (SDS-PAGE). Antibodies were then used to detect LRPPRC and SLIRP, as indicated.
Denaturing, native and second dimension PAGE Denaturing SDS-PAGE was used for the electrophoretic separation of whole-cell extracts, isolated mitochondrial extracts from tissues or fractions from either alkaline carbonate extraction or sucrose gradient sedimentation. Whole cells or isolated mitochondria were extracted on ice with either 1.5– 1.75% dodecyl maltoside/phosphate buffered saline (PBS) or 1% taurodeoxycholate/PBS, centrifuged, after which the cleared extract (‘supernatant’ fraction) and pellet (‘pellet’ fraction) were mixed with Laemmli sample buffer. Alternatively, the detergent extract was mixed with sample buffer without prior centrifugation (‘input’ fraction). Twenty to forty micrograms of protein were then electrophoresed, transferred to a nitrocellulose membrane and analyzed by immunoblotting with the antibodies described in (21), as well as with anti-TACO1 (in-house), anti-MTFMT (Abcam), anti-PDF (Abcam), anti-HSP60 and 75 (Stressgen), anti-SCO1 (in-house) and anti CCDC56 (a kind gift from Peter Rehling, Go¨ttingen).
Mitoplasts prepared from fibroblasts, myoblasts or myotubes by treatment with 0.8, 1.2 or 0.4 mg digitonin/mg protein, respectively, or mitochondria isolated from tissues were run in the first dimension on non-denaturing, blue-native gels (BN-PAGE) as detailed elsewhere (31). The native complexes we visualized with the following antibodies against the following subunits in all native gels: complex I, ND1 (a gift from Anne Lombes, Paris); complex II, SDHA (Abcam); complex III, core1 (Abcam); complex IV, COX IV (Abcam) and complex V, ATPalpha (Abcam). Second-dimension, denaturing PAGE was carried out as described previously (32).
Isolation of mitochondria and submitochondrial localization experiments Mitochondria were isolated from cultured cells or from tissue specimens by differential centrifugation after homogenization in ice-cold SET buffer [250 mM sucrose/10 mM Tris– HCl/
490
Human Molecular Genetics, 2015, Vol. 24, No. 2
1 mM ethylene diamine tetraacetic acid (EDTA)] containing protease inhibitor cocktail (Roche). EDTA was omitted from the buffer when mitochondria were isolated for sucrose gradient fractionation. For cells, 10 confluent 150 mm culture dishes were resuspended in 5–6 ml SET buffer and homogenized with 7–10 strokes in a pre-chilled, zero clearance homogenizer. Tissue samples were cut in small pieces and resuspended in SET buffer at a concentration of 10–15% (w/v) and homogenized with seven to ten strokes in a pre-chilled glass/teflon homogenizer. Homogenates were centrifuged twice at 48C, 600 g for 10 min to generate post-nuclear supernatants, from which mitochondria were pelleted by centrifugation at 48C, 8–10 000 g for 10– 15 min. Mitochondrial pellets were washed once in cold SET buffer, then used immediately or stored at 2808C for later use. For submitochondrial localization experiments, 100 mg mitochondria were extracted in 100 ml sodium carbonate, pH 11.5, for 30 min on ice. Half of the extract volume was saved for immunoblot analysis (‘input’ fraction), while the other half was centrifuged at 25 000g for 1 h at 48C. The resulting supernatant fraction (‘supernatant’ or ‘soluble’) and pellet fraction (‘pellet’ or ‘membrane’) were analyzed along with the ‘input’ fraction by immunoblotting as described.
Sucrose gradient sedimentation Mitochondria isolated from liver or heart specimens were extracted in lysis buffer containing 260 mM sucrose, 100 mM KCl, 20 mM MgCl2, 10 mM Tris– HCl, pH 7.5, 1% Triton X-100, 5 mM b-mercaptoethanol and protease inhibitor cocktail without EDTA, for 20 min on ice, then the extract was cleared by centrifugation at 9400g for 45 min at 48C prior to loading on a 10– 30% discontinuous sucrose gradient (21). After centrifugation, fractions were collected from the top and analyzed by immunoblotting.
RNA blotting analysis Total RNA was isolated with Trizol reagent (Invitrogen) from muscle, heart or liver samples, and 1.5 –10 mg were run on either MOPS/formaldehyde gels for detection of mitochondrial mRNAs (33), or on polyacrylamide/urea gels for detection of mitochondrial tRNAs with 24-mer oligonucleotide probes complementary to the 3′ end of mitochondrial and cytoplasmic tRNAs (15).
ACKNOWLEDGEMENTS The LSFC Consortium is a group of scientists and clinicians dedicated to investigating the causes of and developing treatments for LSFC. The members of the consortium are (in alphabetical order) Bruce Allen, Yan Burelle, Guy Charron, Lise Coderre, Christine DesRosiers, Catherine Laprise, Charles Morin, John Rioux and Eric A. Shoubridge. Conflict of Interest statement. None declared.
FUNDING This research was supported by a CIHR emerging team grant (CPG 102168) to the LSFC consortium and by a CIHR operating grant (MT 15460) to E.A.S.
REFERENCES 1. Koopman, W.J., Willems, P.H. and Smeitink, J.A. (2012) Monogenic mitochondrial disorders. N. Engl. J. Med., 366, 1132–1141. 2. Nunnari, J. and Suomalainen, A. (2012) Mitochondria: in sickness and in health. Cell, 148, 1145–1159. 3. Schon, E.A. and Przedborski, S. (2011) Mitochondria: the next (neurode)generation. Neuron, 70, 1033–1053. 4. Leigh, D. (1951) Subacute necrotizing encephalomyelopathy in an infant. J. Neurochem., 14, 216 –221. 5. Zhu, Z., Yao, J., Johns, T., Fu, K., De Bie, I., Macmillan, C., Cuthbert, A.P., Newbold, R.F., Wang, J., Chevrette, M. et al. (1998) SURF1, encoding a factor involved in the biogenesis of cytochrome c oxidase, is mutated in Leigh syndrome. Nat. Genet., 20, 337– 343. 6. Rahman, S., Blok, R.B., Dahl, H.H., Danks, D.M., Kirby, D.M., Chow, C.W., Christodoulou, J. and Thorburn, D.R. (1996) Leigh syndrome: clinical features and biochemical and DNA abnormalities. Ann. Neurol., 39, 343–351. 7. Merante, F., Petrova-Benedict, R., MacKay, N., Mitchell, G., Lambert, M., Morin, C., De Braekeleer, M., Laframboise, R., Gagne, R. and Robinson, B.H. (1993) A biochemically distinct form of cytochrome oxidase (COX) deficiency in the Saguenay-Lac-Saint-Jean region of Quebec. Am. J. Hum. Genet., 53, 481–487. 8. Debray, F.G., Morin, C., Janvier, A., Villeneuve, J., Maranda, B., Laframboise, R., Lacroix, J., Decarie, J.C., Robitaille, Y., Lambert, M. et al. (2011) LRPPRC mutations cause a phenotypically distinct form of Leigh syndrome with cytochrome c oxidase deficiency. J. Med. Genet., 48, 183–189. 9. Morin, C., Mitchell, G., Larochelle, J., Lambert, M., Ogier, H., Robinson, B.H. and De Braekeleer, M. (1993) Clinical, metabolic, and genetic aspects of cytochrome C oxidase deficiency in Saguenay-Lac-Saint-Jean. Am. J. Hum. Genet., 53, 488–496. 10. Mootha, V.K., Lepage, P., Miller, K., Bunkenborg, J., Reich, M., Hjerrild, M., Delmonte, T., Villeneuve, A., Sladek, R., Xu, F. et al. (2003) Identification of a gene causing human cytochrome c oxidase deficiency by integrative genomics. Proc. Natl. Acad. Sci. USA, 100, 605 –610. 11. Barkan, A. and Small, I. (2014) Pentatricopeptide repeat proteins in plants. Ann. Rev. Plant Biol., 65, 415– 442. 12. Barkan, A., Rojas, M., Fujii, S., Yap, A., Chong, Y.S., Bond, C.S. and Small, I. (2012) A combinatorial amino acid code for RNA recognition by pentatricopeptide repeat proteins. PLoS Genet., 8, e1002910. 13. Yin, P., Li, Q., Yan, C., Liu, Y., Liu, J., Yu, F., Wang, Z., Long, J., He, J., Wang, H.W. et al. (2013) Structural basis for the modular recognition of single-stranded RNA by PPR proteins. Nature, 504, 168– 171. 14. Davies, S.M., Lopez Sanchez, M.I., Narsai, R., Shearwood, A.M., Razif, M.F., Small, I.D., Whelan, J., Rackham, O. and Filipovska, A. (2012) MRPS27 is a pentatricopeptide repeat domain protein required for the translation of mitochondrially encoded proteins. FEBS Lett., 586, 3555–3561. 15. Sasarman, F., Brunel-Guitton, C., Antonicka, H., Wai, T. and Shoubridge, E.A. (2010) LRPPRC and SLIRP interact in a ribonucleoprotein complex that regulates posttranscriptional gene expression in mitochondria. Mol. Biol. Cell., 21, 1315–1323. 16. Chujo, T., Ohira, T., Sakaguchi, Y., Goshima, N., Nomura, N., Nagao, A. and Suzuki, T. (2012) LRPPRC/SLIRP suppresses PNPase-mediated mRNA decay and promotes polyadenylation in human mitochondria. Nucleic Acids Res., 40, 8033–8047. 17. Ruzzenente, B., Metodiev, M.D., Wredenberg, A., Bratic, A., Park, C.B., Camara, Y., Milenkovic, D., Zickermann, V., Wibom, R., Hultenby, K. et al. (2011) LRPPRC is necessary for polyadenylation and coordination of translation of mitochondrial mRNAs. EMBO J., 31, 443–456. 18. Baughman, J.M., Nilsson, R., Gohil, V.M., Arlow, D.H., Gauhar, Z. and Mootha, V.K. (2009) A computational screen for regulators of oxidative phosphorylation implicates SLIRP in mitochondrial RNA homeostasis. PLoS Genet., 5, e1000590. 19. Antonicka, H., Sasarman, F., Kennaway, N.G. and Shoubridge, E.A. (2006) The molecular basis for tissue specificity of the oxidative phosphorylation
Human Molecular Genetics, 2015, Vol. 24, No. 2
20.
21.
22. 23.
24.
25. 26.
deficiencies in patients with mutations in the mitochondrial translation factor EFG1. Hum. Mol. Genet., 15, 1835– 1846. Colley, S.M., Wintle, L., Searles, R., Russell, V., Firman, R.C., Smith, S., Deboer, K., Merriner, D.J., Genevieve, B., Bentel, J.M. et al. (2013) Loss of the nuclear receptor corepressor SLIRP compromises male fertility. PLoS One, 8, e70700. Antonicka, H., Sasarman, F., Nishimura, T., Paupe, V. and Shoubridge, E.A. (2013) The mitochondrial RNA-binding protein GRSF1 localizes to RNA granules and is required for posttranscriptional mitochondrial gene expression. Cell Metab., 17, 386 –398. Jourdain, A.A., Koppen, M., Wydro, M., Rodley, C.D., Lightowlers, R.N., Chrzanowska-Lightowlers, Z.M. and Martinou, J.C. (2013) GRSF1 regulates RNA processing in mitochondrial RNA granules. Cell Metab., 17, 399–410. Borowski, L.S., Dziembowski, A., Hejnowicz, M.S., Stepien, P.P. and Szczesny, R.J. (2013) Human mitochondrial RNA decay mediated by PNPase-hSuv3 complex takes place in distinct foci. Nucleic Acids Res., 41, 1223– 12240. Holzmann, J., Frank, P., Loffler, E., Bennett, K.L., Gerner, C. and Rossmanith, W. (2008) RNase P without RNA: identification and functional reconstitution of the human mitochondrial tRNA processing enzyme. Cell, 135, 462– 474. Brzezniak, L.K., Bijata, M., Szczesny, R.J. and Stepien, P.P. (2011) Involvement of human ELAC2 gene product in 3′ end processing of mitochondrial tRNAs. RNA Biol., 8, 616– 626. Ojala, D., Montoya, J. and Attardi, G. (1981) tRNA punctuation model of RNA processing in human mitochondria. Nature, 290, 470–474.
491
27. Shoubridge, E.A., Johns, T. and Boulet, L. (1996) Use of myoblast cultures to study mitochondrial myopathies. Methods Enzymol., 264, 465–475. 28. Lochmuller, H., Johns, T. and Shoubridge, E.A. (1999) Expression of the E6 and E7 genes of human papillomavirus (HPV16) extends the life span of human myoblasts. Exp. Cell Res., 248, 186–193. 29. Boulet, L., Karpati, G. and Shoubridge, E.A. (1992) Distribution and threshold expression of the tRNA(Lys) mutation in skeletal muscle of patients with myoclonic epilepsy and ragged-red fibers (MERRF). Am. J. Hum. Genet., 51, 1187–1200. 30. Sasarman, F. and Shoubridge, E.A. (2012) Radioactive labeling of mitochondrial translation products in cultured cells. Methods Mol. Biol., 837, 207– 217. 31. Leary, S.C. and Sasarman, F. (2009) Oxidative phosphorylation: synthesis of mitochondrially encoded proteins and assembly of individual structural subunits into functional holoenzyme complexes. Methods Mol. Biol., 554, 143– 162. 32. Antonicka, H., Ogilvie, I., Taivassalo, T., Anitori, R.P., Haller, R.G., Vissing, J., Kennaway, N.G. and Shoubridge, E.A. (2003) Identification and characterization of a common set of complex I assembly intermediates in mitochondria from patients with complex I deficiency. J. Biol. Chem., 278, 43081– 43088. 33. Weraarpachai, W., Antonicka, H., Sasarman, F., Seeger, J., Schrank, B., Kolesar, J.E., Lochmuller, H., Chevrette, M., Kaufman, B.A., Horvath, R. et al. (2009) Mutation in TACO1, encoding a translational activator of COX I, results in cytochrome c oxidase deficiency and late-onset Leigh syndrome. Nat. Genet., 41, 833–837.
480–491
Tissue-specific responses to the LRPPRC founder mutation in French Canadian Leigh Syndrome Florin Sasarman{, Tamiko Nishimura{, Hana Antonicka, Woranontee Weraarpachai and Eric A. Shoubridge∗ , LSFC Consortium Montreal Neurological Institute and Department of Human Genetics, McGill University, Montreal, QC H3A 2B4, Canada Received June 30, 2014; Revised August 1, 2014; Accepted September 9, 2014
French Canadian Leigh Syndrome (LSFC) is an early-onset, progressive neurodegenerative disorder with a distinct pattern of tissue involvement. Most cases are caused by a founder missense mutation in LRPPRC. LRPPRC forms a ribonucleoprotein complex with SLIRP, another RNA-binding protein, and this stabilizes polyadenylated mitochondrial mRNAs. LSFC fibroblasts have reduced levels of LRPPRC and a specific complex IV assembly defect; however, further depletion of mutant LRPPRC results in a complete failure to assemble a functional oxidative phosphorylation system, suggesting that LRPPRC levels determine the nature of the biochemical phenotype. We tested this hypothesis in cultured muscle cells and tissues from LSFC patients. LRPPRC levels were reduced in LSFC muscle cells, resulting in combined complex I and IV deficiencies. A similar combined deficiency was observed in skeletal muscle. Complex IV was only moderately reduced in LSFC heart, but was almost undetectable in liver. Both of these tissues showed elevated levels of complexes I and III. Despite the marked biochemical differences, the steady-state levels of LRPPRC and mitochondrial mRNAs were extremely low, LRPPRC was largely detergent-insoluble, and SLIRP was undetectable in all LSFC tissues. The level of the LRPPRC/SLIRP complex appeared much reduced in control tissues by the first dimension bluenative polyacrylamide gel electrophoresis (BN-PAGE) analysis compared with fibroblasts, and even by second dimension analysis it was virtually undetectable in control heart. These results point to tissue-specific pathways for the post-transcriptional handling of mitochondrial mRNAs and suggest that the biochemical defects in LSFC reflect the differential ability of tissues to adapt to the mutation.
INTRODUCTION Deficiencies in oxidative phosphorylation (OXPHOS) are associated with a diverse array of multisystem disorders that are often referred to as mitochondrial encephalomyopathies because of the prominent involvement of the nervous system and striated muscle (1 – 3). The molecular basis for the extraordinarily broad clinical spectrum observed in OXPHOS disorders remains an enduring mystery. Mutations in mtDNA are the most frequent cause of mitochondrial disease in adults and more than 100 such pathogenic mutations have been identified. In the pediatric population, the majority of OXPHOS disorders (80%) are transmitted as autosomal recessive traits, usually with a severe phenotype and a fatal outcome. Leigh syndrome is one of the most common presentations in this population. First described by Denis Leigh in a single case
(4), the so-called classical form of the disease is characterized by early-onset neurodegeneration, particularly in the brain stem and basal ganglia. The genetic underpinnings of this disorder are found in structural subunits of the OXPHOS complexes, associated assembly factors or in pyruvate dehydrogenase (5,6). A variant known as Leigh Syndrome French Canadian (LSFC) has a characteristic pattern of tissue involvement and a different clinical evolution, with acute acidotic crises usually leading to very early fatality (7 –9). This disorder is caused by a founder missense mutation in LRPPRC, a gene coding for a pentatricopeptide repeat (PPR) family protein (10). PPR proteins, which belong to the alpha-solenoid protein family, have undergone a huge radiation in land plants where they number several hundred (11). Virtually all of them localize to mitochondria or chloroplasts where they mediate a large variety of RNA transactions including splicing, editing, stability and translation
∗
To whom correspondence should be addressed. Tel: +1 5143981997; Fax: +1 5143981509; Email: [email protected] Co-first authors.
†
# The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Human Molecular Genetics, 2015, Vol. 24, No. 2
initiation. A combinatorial code that specifies the recognition of specific RNA substrates has been determined for the chloroplast PPR protein PPR10 (12), a result that was substantiated by the solution of the crystal structure of the protein bound to its mRNA substrate (13). Humans have only seven identified PPR proteins, all of which are mitochondrial (14). LRPPRC forms a ribonucleoprotein (RNP) complex of about 250 kDa with the stem-loop RNA-binding protein SLIRP, and this complex appears necessary to stabilize the majority of polyadenylated mitochondrial mRNAs (15– 17), protecting them against 3′ exoribonuclease digestion (16). In vitro studies with purified LRPPRC also show a stimulation of polyA addition to RNA substrates, suggesting that the complex might play a dual role in the synthesis and maintenance of the 3′ ends of mitochondrial mRNAs (16). The protein carrying the A354V substitution appears unstable in fibroblasts, as the steady-state level is approximately 30% of control (15). Depletion of wild-type LRPPRC to the same level phenocopies the defect seen in LSFC fibroblasts, suggesting that the mutant protein is functional; however, further siRNA-mediated depletion results in an assembly defect in all OXPHOS complexes with mtDNA-encoded subunits (15). These data suggest that tissue-specific differences in the level of mutant LRPPRC might account for the tissuespecific biochemical deficiencies characteristic of LSFC. We tested this hypothesis in cells and tissues obtained from LSFC patients and show that the situation is much more complex. Tissue-specific involvement appears to reflect, at least in part, the ability of different tissues to adapt to the genetic defect.
RESULTS LRPPRC and SLIRP are interdependent and interact with a 1:1 stoichiometry Previous work in human fibroblasts (15) and mouse liver (17) demonstrated that LRPPRC functions in a RNP complex with the stem-loop RNA-binding protein SLIRP. A decrease in steady-state levels of either protein results in a similar decrease in the levels of the other, and a reduction of the levels of either LRPPRC or SLIRP causes a similar destabilization of the mitochondrial mRNAs (18). To further confirm this functional interdependence, we assessed the pattern of mitochondrial translation in control fibroblasts in which SLIRP was knocked down using two different siRNA constructs, both of which reduced SLIRP to barely immunodetectable levels (Fig. 1B). The mitochondrial translation defect phenocopied that observed in LSFC fibroblasts (15), where the most severe reductions were observed in the synthesis of COX I, II and III, and an anomalous translation product appeared (Fig. 1A). As expected, the decrease in the level of SLIRP was paralleled by a decrease in the level of LRPPRC. We determined the stoichiometry of interaction between LRPPRC and SLIRP by immunoprecipitating SLIRP in mitochondrial extracts from control fibroblasts, followed by mass spectrometry analysis. Using the abundance of five tryptic peptides in each protein, the relative ratio of LRPPRC (130 kDa) and SLIRP (12 kDa) was found to be 1:1, suggesting that 250 kDa RNP complex in fibroblasts may be a heterotetramer. PPR10, the only PPR protein whose crystal structure has been determined, was found to dimerize when bound to its RNA substrate (13).
481
LSFC cultured muscle cells have a combined OXPHOS deficiency To investigate the nature of the biochemical phenotype in another cell type, we cultured myoblasts obtained from two LSFC cases, differentiated them into myotubes, and immunoblotted for LRPPRC and SLIRP. Similar to LSFC fibroblasts, the level of mutant LRPPRC was reduced to 20– 30%, and the levels of SLIRP, to 10– 20% of control in LSFC myoblasts and myotubes (Figs 2A and 3A). The synthesis of the COX subunits was particularly affected in LSFC muscle cells, and an anomalous translation product appeared in post-natal LSFC myoblasts and myotubes (Figs 2B and 3B). However, unlike LSFC fibroblasts, all of the cultured muscle cells from LSFC subjects (fetal and post-natal myoblasts and myotubes) showed a combined OXPHOS defect, with reduced levels of assembled complex I and complex IV as shown by blue-native polyacrylamide gel electrophoresis (BN-PAGE) analysis (Figs 2C and 3C). To determine whether a further reduction in the steady-state level of LRPPRC would lead to a more severe OXPHOS deficiency in cultured muscle cells (as it does in fibroblasts), we used siRNA to deplete LRPPRC below 10% of control levels in control and LSFC myoblasts and myotubes. This resulted in a generalized OXPHOS defect in which all complexes with mtDNA-encoded subunits were affected, as demonstrated by immunoblot (Fig. 3A), mitochondrial translation (Fig. 3B) and BN-PAGE analyses (Fig. 3C). The smaller molecular weight complex V species in Figure 3C represent F1-containing subcomplexes. The reduction in steady-state levels of mitochondrial mRNAs does not correlate with the biochemical phenotype in LSFC tissues Given the role of LRPPRC in binding mature mitochondrial mRNAs, we started our analysis of LSFC tissues by northern blot analysis of the steady-state levels of mitochondrial mRNAs in the skeletal muscle, heart and liver tissues from an LSFC patient (Fig. 4A). In all three tissues, there was a generalized decrease in the levels of all analyzed mitochondrial mRNAs, the most severe reduction in mRNA levels being observed in skeletal muscle. Additionally, an increase in the levels of the 12S mitochondrial ribosomal RNA was observed in the three tested LSFC tissues, and an increase in the 16S mitochondrial ribosomal RNA, in the LSFC heart and liver, presumably indicating a compensatory response to the decrease in mitochondrial mRNA levels. The levels of mitochondrial tRNAs were indistinguishable from control in all three LSFC tissues (Fig. 4B). Loss of LRPPRC and SLIRP results in tissue-specific OXPHOS deficiencies BN-PAGE analysis of the LSFC tissues revealed an almost complete lack of assembled complex IV in LSFC liver (not detectable even in overexposed blots), and 50% of control levels of assembled complex IV in heart (Fig. 5A). Additionally, both of these tissues displayed an increase in the levels of assembled complexes I and III. The differences in intensity of the bands for complexes I, III, IV and V in the control samples reflect real
482
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 1. Knockdown of SLIRP in control fibroblasts phenocopies the mitochondrial translation defect in LSFC fibroblasts. (A) Control fibroblasts were transiently transfected with two different siRNA constructs specific to SLIRP, or with a fluorescent control siRNA (Alexa). On the sixth day post-transfection, the mitochondrial translation products were pulse-labeled with a mix of [35S]-methionine and cysteine in the presence of emetine, an inhibitor of cytoplasmic translation, and analyzed by PAGE. The 13 mitochondrial translation products are indicated to the left of the figure: seven subunits of complex I (ND), one subunit of complex III, three subunits of complex IV (COX) and two subunits of complex V (ATP). The asterisk indicates an anomalous, unidentified translation product typically detected in LSFC fibroblasts. (B) The knockdown of SLIRP and the concomitant reduction in steady-state levels of LRPPRC were confirmed by immunoblotting with specific antibodies. The 70 kDa subunit of complex II was used as a loading control.
difference in the relative abundance of these complexes between heart and liver, as the same amount of mitochondrial protein was loaded in all lanes and the level of complex II is similar in all lanes. On the other hand, in skeletal muscle we observed a combined complex I and IV deficiency, similar to that seen in cultured muscle cells (Fig. 5A). This pattern of OHPHOS deficiency was confirmed by immunoblotting cleared mitochondrial extracts from LSFC and control tissues, which showed decreases in the steady-state levels of structural subunits for the same complexes that showed assembly defects by BN-PAGE analysis (Fig. 5B). Despite the tissue-specific OHPHOS defects, mutant LRPPRC and SLIRP were virtually undetectable in all three analyzed LSFC tissues. An immunoblot analysis of the endogenous levels of LRPPRC and SLIRP in cleared mitochondrial extracts from control tissues showed comparable levels, relative to proteins in the OXPHOS complexes, in all three analyzed LSFC tissues (Fig. 5C). Detergent solubility of LRPPRC and a subset of proteins involved in mitochondrial translation is altered in LSFC liver Since mutant LRPPRC appeared to be present at levels below immunodetection in LSFC tissues, yet with such tissue-specific biochemical consequences, we asked whether residual LRPPRC might be present in a detergent-insoluble fraction, that is a fraction that could not be extracted with the concentration of the non-ionic detergent dodecyl maltoside (DDM), that we routinely use for immunoblot analysis of proteins, in some or all of these tissues. To test this, mitochondria from LSFC and control
tissues extracted in DDM were mixed with the loading buffer [containing sodium dodecyl sulfate (SDS)], electrophoresed by PAGE as usual, and subsequently immunoblotted. As shown in Figure 6A, while LRPPRC was still not detected in total extracts of LSFC muscle, residual mutant protein was present in LSFC heart and liver at 10% of control levels. It is important to point out that mutant LRPPRC could easily be extracted from fibroblasts or myoblasts with DDM, so this is a tissue-specific and not a technical problem. While the above findings might explain a combined complex I and IV deficiency in LSFC muscle versus an isolated complex IV defect in LSFC heart and liver, it still does not clarify why complex IV deficiency is severe in liver and mild in heart. Hence, we further investigated the distribution of LRPPRC and other proteins involved in mitochondrial translation between the soluble and insoluble fractions following detergent extraction, in LSFC and control liver. Interestingly, while LRPPRC was found predominantly (80 –90%) in the supernatant (soluble fraction) in control liver, the residual amount of mutant LRPPRC was detected exclusively in the pellet (insoluble fraction) in the LSFC liver (Fig. 6B). A similar reversal in distribution between supernatant and pellet and, in some cases, a decrease in total amounts was also observed for a subset of proteins involved in mitochondrial translation, including two translation factors (TACO1, EFG1), a mitochondrial ribosomal protein (MRPS27) and the mitochondrial formyltransferase, MTFMT (Fig. 6B). By contrast, another mitochondrial translation factor (CCDC56), the ribosomal protein MRPL44, the mitochondrial peptide deformylase (PDF) and two subunits of OXPHOS complexes (SDHA, ATP5A1) displayed a normal distribution between the soluble
Human Molecular Genetics, 2015, Vol. 24, No. 2
483
Figure 2. Mitochondrial translation defect and combined OXPHOS deficiency in LSFC cultured muscle cells. (A) Immunoblot analysis showing reduced levels of LRPPRC and SLIRP in extracts from fetal and post-natal cultured muscle cells from LSFC subjects. Prohibitin was used as loading control. The sample for LSFC fetal myotubes was slightly under loaded. (B) Mitochondrial translation products in fetal and post-natal myoblasts and myotubes from LSFC subjects were pulse-labeled with a mix of [35S]-methionine and cysteine in the presence of emetine, an inhibitor of cytoplasmic translation and analyzed by PAGE. The 13 mitochondrial translation products are indicated to the left of the figure: seven subunits of complex I (ND), one subunit of complex III, three subunits of complex IV (COX) and two subunits of complex V (ATP). The asterisk indicates a translation product with anomalous migration also detected in LSFC fibroblasts. (C) Blue-native PAGE analysis of the five OXPHOS complexes revealed reduced levels of assembled complex IV (COX) and complex I in cultured muscle cells from LSFC subjects. Individual complexes were detected with subunit-specific antibodies.
and insoluble fractions in the LSFC liver, demonstrating that this was not a general problem of protein aggregation. The mitochondrial apoptosis-inducing factor AIFM1 also showed a control pattern of distribution between fractions following detergent extraction of LSFC liver mitochondria, suggesting that the change in solubility of the subset of proteins in patient tissue is not linked to pro-apoptotic events. The different distribution of LRPPRC between fractions in LSFC liver does not appear to be the consequence of compromised mitochondrial membranes, as transmembrane proteins of the outer (VDAC1) and inner
(ATP5A1) membranes were extracted the same way in LSFC and control. Strikingly, SLIRP, which appeared only in the soluble fraction in control tissues, was not detectable by immunoblot analysis in any LSFC tissue. We confirmed the presence of LRPPRC in the insoluble fraction following DDM extraction of LSFC liver mitochondria by repeating the analysis with two additional controls (Fig. 6C). While a certain percentage of LRPPRC (between 20 and 50%) is associated with the pellet in control liver, in the LSFC liver, the residual, mutant LRPPRC is found exclusively in the
484
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 3. Generalized OXPHOS defect following knockdown of LRPPRC in LSFC cultured muscle cells. LSFC and control myoblasts or myotubes were transiently transfected either with an siRNA construct specific to LRPPRC, or with a fluorescent control siRNA (Alexa), and then (A) analyzed by immunoblotting with antibodies against LRPPRC, SLIRP or subunits of the OXPHOS complexes, (B) pulse-labeled with a mix of [35S]-methionine and cysteine in the presence of emetine and (C) analyzed by BN-PAGE (as described in Fig. 2).
pellet. A similar analysis of heart mitochondria revealed that all tested proteins are found predominantly (80%) in the insoluble fraction following DDM extraction, in both control and LSFC tissues (data not shown). LRPPRC and SLIRP are partially associated with mitochondrial membranes The presence of LRPPRC in the insoluble fraction following DDM extraction, even in control mitochondria, prompted us to investigate whether LRPPRC and/or SLIRP are associated with the mitochondrial membranes. While neither of these two proteins is predicted to contain a transmembrane domain, alkaline carbonate extraction of isolated mitochondria revealed that up to 50% of LRPPRC and SLIRP are found in the membrane fraction in HEK293 cells (Fig. 7A) and in fibroblasts, whether control or LSFC (Fig. 7B).
The levels of mitochondrial small ribosomal subunits and the monosome are decreased in LSFC liver The alteration of the detergent solubility of a subset of proteins involved in mitochondrial translation in the LSFC liver, including a protein of the small ribosomal subunit (mt-SSU), MRPS27 (Fig. 6B), might reflect a qualitative or quantitative change of the mitochondrial ribosomes in this tissue. To test this, we performed sucrose gradient centrifugation of cleared mitochondrial extracts from control and LSFC liver. Consistent with the decrease in levels and detergent solubility of MRPS27, the amounts of fully assembled small subunit (mt-SSU) and the monosome were decreased in the LSFC liver, while the levels of assembled large ribosomal subunits (mt-LSU) were normal (Fig. 8A). In contrast, in the LSFC heart mitochondria there was an increase in the steady-state levels of subunits of the small and large ribosomal subunits and of assembled small and
Human Molecular Genetics, 2015, Vol. 24, No. 2
485
Figure 4. Generalized reduction of steady-state levels of mitochondrial mRNAs in LSFC tissues. Total RNA was extracted from muscle, heart and liver of controls and an LSFC subject, followed by either standard northern blotting analysis with probes specific for the mitochondrial mRNAs or rRNAs indicated at the left of the panel (A), or by urea-PAGE and hybridization with oligonucleotide probes complementary to the mitochondrial and cytoplasmic tRNAs indicated at the right of the panel (B).
large ribosomal subunits (Fig. 8B and C), suggesting an adaptive response to the decrease in the levels of mRNAs in this tissue.
The LRPPRC/SLIRP complex is not detectable in heart tissue We next asked whether the LRPPRC/SLIRP complex is preferentially stabilized in the LSFC heart. First dimension blue-native PAGE analysis showed that 250 kDa complex of LRPPRC and SLIRP was undetectable with antibodies against LRPPRC in the LSFC liver as compared with control (Fig. 9A), whereas in the heart, it was undetectable in both LSFC and control mitochondria (data not shown). This led us to hypothesize that the reason why heart is spared in LSFC might be a quantitative or qualitative difference of the LRPPRC/SLIRP complex in this tissue, as opposed to liver or muscle. To test this, we analyzed mitochondria from heart, liver and muscle of two controls, and from two control fibroblast lines, by the first and second dimension BN-PAGE. Indeed, the LRPPRC – SLIRP complex was undetectable in the heart of both controls in the first dimension and was greatly reduced in muscle or liver as compared with fibroblasts (Fig. 9B). This contrasts with the levels of assembled complex IV, which are highest in heart and muscle, and lowest in liver and fibroblasts (Fig. 9B and (19)). Additionally, the LRPPRC–SLIRP complex in muscle migrates at a slightly higher molecular mass than in the other analyzed tissues and cells, suggesting it might have different properties or composition. Heart was the only tissue in which LRPPRC and SLIRP were equally undetectable in the second dimension (Fig. 9C). These observations suggest that heart has alternative pathways to the LRPPRC/SLIRP complex for post-transcriptional handling of mitochondria-encoded proteins.
DISCUSSION The results presented in this study demonstrate striking differences in the post-transcriptional handling of mitochondrial mRNAs both among tissues and between cultured cells and fully differentiated tissues. In fibroblasts, myoblasts and differentiated myotubes, we show that the LRPPRC–SLIRP complex is essential for mitochondrial translation and, as a consequence, assembly of all OXPHOS complexes containing mtDNA-encoded structural subunits. Reducing the level of LRPPRC to levels undetectable by immunoblot analysis abolishes translation and completely abrogates assembly of a functional respiratory chain. Depletion of SLIRP phenocopies the translation defect, emphasizing the interdependence of the two proteins. What was unexpected, and remains to be explained, is that despite very similar translation defects, LSFC fibroblasts and muscle cells have quite distinct biochemical phenotypes, the latter showing a marked reduction in complex I assembly, similar to that observed in patients with isolated complex I deficiencies, in addition to a complex IV assembly defect. This suggests that the basis for the different biochemical defects lies downstream of mitochondrial translation. What is most extraordinary, however, is the fact that in the three LSFC tissues that we were able to study, the steady-state levels of the mutant form of LRPPRC were extremely low, the protein was largely detergent-insoluble and SLIRP was undetectable (Note that we have no evidence that tissue-specific isoforms or paralogues exist for either protein). Despite this rather severe and relatively homogeneous phenotype, it seems evident that many of the mitochondrial mRNAs are translated to a degree that underwrites normal (or increased as for complexes I and III in heart and liver) assembly of some of the OXPHOS complexes, but that marked tissue-specific deficiencies persist. This occurs despite the fact that the steady-state
486
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 5. Tissue-specific OXPHOS deficiencies in LSFC tissues. Mitochondrial extracts from muscle, heart and liver of one LSFC subject, several controls and four disease controls were analyzed by BN-PAGE (A) and by western blotting (B) with antibodies against subunits of the OXPHOS complexes or against LRPPRC and SLIRP, as indicated. In (A), ‘Co I deficiency’ refers to two subjects with assembly defects of complex I; in (B), ‘COX15’ and ‘EFG1’ refer to two subjects with mutations in an assembly factor of COX and a mitochondrial translation factor, respectively. (C) Cleared mitochondrial extracts from control liver, heart and muscle were analyzed by immunoblotting with antibodies against LRPPRC, SLIRP or subunits of the OXPHOS complexes.
levels of the mRNAs were severely reduced in all three tissues, as might have been predicted from the results we obtained in fibroblasts showing the interdependence of LRPPRC, SLIRP and mRNA (15). The only overt phenotype reported in a recent study of a mouse germ line deletion of SLIRP was a small decrease in male fertility, although the study did not include any biochemical analyses of OXPHOS complex assembly or function (20). It thus appears that SLIRP may not be an essential component of the molecular machinery involved in the processing and handling of mitochondrial mRNAs, or that some
Figure 6. Altered detergent solubility of LRPPRC and of other mitochondrial translation proteins in LSFC liver. (A) Mitochondria from muscle, heart and liver samples of the LSFC subject and controls were extracted in dodecyl maltoside, then mixed with loading buffer containing SDS without extract clearing by centrifugation, followed by immunoblot analysis with antibodies against LRPPRC or the 70 kDa subunit of complex II. (B, C) Mitochondria isolated from liver samples of one LSFC subject and controls were extracted in dodecyl maltoside or taurodeoxycholate (for TACO1 detection), then loading buffer containing SDS was added to the extracts without extract clearing by centrifugation (‘Input’). Duplicate samples were centrifuged following extraction in dodecyl maltoside or taurodeoxycholate, and each soluble (‘Supernatant’) and insoluble (‘Pellet’) fraction was then mixed with loading buffer containing SDS. All samples were analyzed by immunoblotting with antibodies against LRPPRC, SLIRP, mitochondrial ribosomal proteins (MRPS27, MRPL44), factors or enzymes involved in mitochondrial translation (EFG1, TACO1, MTFMT, PDF, CCDC56) and other mitochondrial proteins, as detailed in the text.
component(s) of the translation can compensate for its absence. In any case, we were not able to convincingly demonstrate the presence of the LRPPRC/SLIRP complex in control heart tissue by either first or second dimension BN-PAGE analysis.
Human Molecular Genetics, 2015, Vol. 24, No. 2
487
Figure 7. LRPPRC and SLIRP are partly associated with the mitochondrial membrane fraction. Alkaline carbonate extracts of mitochondria from HEK293 cells (A) or from control and LSFC fibroblasts (B) were analyzed by immunoblotting with antibodies against LRPPRC, SLIRP and the following protein controls: COX I (multiple pass, integral inner membrane), SCO1 (single pass, inner membrane), VDAC (outer membrane), Co II-70 kDa (soluble, inner membrane-associated), HSP 60 and 75 (soluble).
These data lead us to two main conclusions. First, our data show that the way in which mitochondrial mRNAs are handled is tissue-specific. Although LRPPRC does not contain a predicted transmembrane domain, a significant fraction of it stays in the pellet after alkaline carbonate extraction in HEK cells and fibroblasts, and most of it remains DDM-insoluble in heart, but less so in liver. This suggests that LRPPRC may shuttle between a soluble matrix fraction and an inner membrane fraction that might depend on its association with SLIRP, and that this equilibrium shifts in some tissues in the presence of the mutant protein. The fact that the LRPPRC/SLIRP complex was more difficult to detect in the first versus second dimension BN-PAGE gels in muscle and liver, suggests that the epitope recognized by the LRPPRC antibody is less accessible in the native complex in tissues, perhaps because of a posttranslational modification, or the presence of another protein in the complex, but this remains to be investigated. Second, it seems clear that some tissues have the ability to adapt to the LRPPRC mutation better than others. The increase in the total number of monosomes in the heart is presumably a response to increase the efficiency of translation at lower steady-state levels of mRNAs. This signaling does not appear to function in the liver, as there were fewer monosomes in LSFC liver, due to a decrease in the assembled mt-SSU. We previously reported a re-organization of the translation elongation machinery in the heart of a patient with mutations in the translation elongation factor EFG1 (19). In this case, the ratio of EFTu:EFTs (the proteins involved in bringing the charged tRNAs to the ribosome) was completely reversed in the patient heart compared with control, driven by transcriptional reprogramming. These data suggest that there is a feedback mechanism that can adjust the functioning of the translation apparatus in some tissues, in this case in the face of a crippled polypeptide elongation factor EFG1. Why such adaptation does not occur in all tissues, to minimize the impact of the genetic mutation, remains unknown. Recently, it has become clear that the polycistronic transcripts transcribed from mtDNA are processed in discrete granules that have been defined by the presence of the RNA-binding protein
GRSF1 (21,22). Newly synthesized RNAs, visualized by pulse labeling with the nucleotide analog BrU, co-localize with GRSF1 (21). These foci also contain Suv3 and PNPase, the enzymes that are involved in mRNA degradation (23). Mitochondrial transcripts are matured by the activities of RNAseP and Z, which process the 5′ and 3′ ends of tRNAs, respectively (24,25). As tRNAs punctuate most of the open reading frames in mtDNA (26), this results in the release of mature mRNAs. It is not known how LRPPRC and SLIRP then associate with these processed RNAs, but it does seem clear that this association is necessary for mRNA stability. Immunofluorescence experiments with antibodies directed against either protein do not localize them to foci, but rather label the entire mitochondrial matrix. It is possible that either or both proteins shuttle between the foci and a free pool in order to deliver the mRNAs to ribosomes, which are thought to be tightly associated to the inner mitochondrial membrane where the mitochondrial polypeptides are co-translationally inserted into the inner membrane. This could explain why LRPPRC appears to exist both in a soluble and membrane-associated form. The nature of the biochemical defects in LSFC patients that we report here shows that it is not a pure complex IV defect as had been previously thought. The study raises many questions that will require transcriptomic and proteomic analyses to resolve. We have recently developed a knock-in mouse model, in which we have replaced the wild-type allele with the murine homolog of the LSFC founder mutation, and this should allow us to address the molecular basis of tissue specificity in a definitive fashion.
MATERIALS AND METHODS Cell culture and RNA interference experiments Myoblasts were cultured from muscle biopsies of two LSFC patients according to an established procedure (27). Primary myoblasts were FACS-sorted and then immortalized by transduction with two retroviral vectors, one expressing the HPV-16 E6 and E7 genes, and the other expressing the catalytic
488
Human Molecular Genetics, 2015, Vol. 24, No. 2
Figure 8. Analysis of ribosome assembly in LSFC liver and heart. Cleared mitochondrial lysates from control and LSFC liver (A) and heart (B) were loaded on a discontinuous sucrose gradient and centrifuged, after which fractions were collected for immunoblot analysis with antibodies against subunits of the large (mitochondrial ribosomal protein, large) and small (mitochondrial ribosomal protein, small) mitochondrial ribosomal subunits. The approximate migration of the small (28S) and of the large (39S) mitochondrial ribosomal subunits and the monosome (55S) are indicated for reference. (C) Western blotting analysis of cleared mitochondrial extracts from control and LSFC heart with antibodies against small and large mitochondrial ribosomal subunits. One subunit of complex II and one of complex V were used as loading controls.
component of human telomerase (28). Myoblasts were grown in Ham’s F10-based, supplemented growth medium (29). Myotubes were generated by allowing confluent myoblasts to fuse for 2 – 4 days in dulbecco’s modified eagle medium (DMEM) medium supplemented with 2% horse serum and 0.4 mg/ml dexamethasone. Fibroblasts and HEK293 cells were grown in DMEM supplemented with 10% fetal bovine serum. For knockdown of SLIRP, two Stealth RNA interference duplexes (Invitrogen) were used: CGG CGU CGA GUC AGC UGA AAG AAC A and ACA GAG GUU UGG GUU GGG UUC AGU U. LRPPRC was knocked down with the Stealth RNA interference duplex Stealth1173 (15). Stealth RNAi duplexes were transiently transfected at a concentration of 25 nM according to manufacturer’s instructions for reverse
transfection. Transfection was repeated on day 3, and on day 6 cells were either harvested and analyzed (fibroblasts and myoblasts), or transferred to fusion medium and allowed to fuse for 2–4 days (myotubes) prior to analysis.
Radioactive labeling of mitochondrial translation products Mitochondrial translation products in cultured muscle cells or fibroblasts were pulse-labeled for 1 h with 200 mCi/ml of a mixture of [35S]-methionine and cysteine (Perkin Elmer) in DMEM lacking methionine and cysteine and containing 100 ml/ml emetine, an inhibitor of cytoplasmic translation, as detailed elsewhere (30).
Human Molecular Genetics, 2015, Vol. 24, No. 2
489
Figure 9. The LRPPRC– SLIRP complex is undetectable in normal heart tissue by the first and second dimension BN/SDS-PAGE analysis. (A) First dimension BN-PAGE analysis of the LRPPRC–SLIRP complex with an antibody against LRPPRC, and of complexes II and IV (COX) with subunit-specific antibodies, in control and LSFC liver. (B) Comparison of the levels of LRPPRC– SLIRP complex in control tissues and cells by the first dimension BN-PAGE analysis. The LRPPRC–SLIRP complex was detected with an antibody against LRPPRC and complex IV was detected with an antibody against its nuclear subunit COX IV. (C) Extracts from control tissues and cells were run in the first dimension on a non-denaturing, blue native gel (BN-PAGE), after which individual lanes were cut out and run on the second dimension on a denaturing gel (SDS-PAGE). Antibodies were then used to detect LRPPRC and SLIRP, as indicated.
Denaturing, native and second dimension PAGE Denaturing SDS-PAGE was used for the electrophoretic separation of whole-cell extracts, isolated mitochondrial extracts from tissues or fractions from either alkaline carbonate extraction or sucrose gradient sedimentation. Whole cells or isolated mitochondria were extracted on ice with either 1.5– 1.75% dodecyl maltoside/phosphate buffered saline (PBS) or 1% taurodeoxycholate/PBS, centrifuged, after which the cleared extract (‘supernatant’ fraction) and pellet (‘pellet’ fraction) were mixed with Laemmli sample buffer. Alternatively, the detergent extract was mixed with sample buffer without prior centrifugation (‘input’ fraction). Twenty to forty micrograms of protein were then electrophoresed, transferred to a nitrocellulose membrane and analyzed by immunoblotting with the antibodies described in (21), as well as with anti-TACO1 (in-house), anti-MTFMT (Abcam), anti-PDF (Abcam), anti-HSP60 and 75 (Stressgen), anti-SCO1 (in-house) and anti CCDC56 (a kind gift from Peter Rehling, Go¨ttingen).
Mitoplasts prepared from fibroblasts, myoblasts or myotubes by treatment with 0.8, 1.2 or 0.4 mg digitonin/mg protein, respectively, or mitochondria isolated from tissues were run in the first dimension on non-denaturing, blue-native gels (BN-PAGE) as detailed elsewhere (31). The native complexes we visualized with the following antibodies against the following subunits in all native gels: complex I, ND1 (a gift from Anne Lombes, Paris); complex II, SDHA (Abcam); complex III, core1 (Abcam); complex IV, COX IV (Abcam) and complex V, ATPalpha (Abcam). Second-dimension, denaturing PAGE was carried out as described previously (32).
Isolation of mitochondria and submitochondrial localization experiments Mitochondria were isolated from cultured cells or from tissue specimens by differential centrifugation after homogenization in ice-cold SET buffer [250 mM sucrose/10 mM Tris– HCl/
490
Human Molecular Genetics, 2015, Vol. 24, No. 2
1 mM ethylene diamine tetraacetic acid (EDTA)] containing protease inhibitor cocktail (Roche). EDTA was omitted from the buffer when mitochondria were isolated for sucrose gradient fractionation. For cells, 10 confluent 150 mm culture dishes were resuspended in 5–6 ml SET buffer and homogenized with 7–10 strokes in a pre-chilled, zero clearance homogenizer. Tissue samples were cut in small pieces and resuspended in SET buffer at a concentration of 10–15% (w/v) and homogenized with seven to ten strokes in a pre-chilled glass/teflon homogenizer. Homogenates were centrifuged twice at 48C, 600 g for 10 min to generate post-nuclear supernatants, from which mitochondria were pelleted by centrifugation at 48C, 8–10 000 g for 10– 15 min. Mitochondrial pellets were washed once in cold SET buffer, then used immediately or stored at 2808C for later use. For submitochondrial localization experiments, 100 mg mitochondria were extracted in 100 ml sodium carbonate, pH 11.5, for 30 min on ice. Half of the extract volume was saved for immunoblot analysis (‘input’ fraction), while the other half was centrifuged at 25 000g for 1 h at 48C. The resulting supernatant fraction (‘supernatant’ or ‘soluble’) and pellet fraction (‘pellet’ or ‘membrane’) were analyzed along with the ‘input’ fraction by immunoblotting as described.
Sucrose gradient sedimentation Mitochondria isolated from liver or heart specimens were extracted in lysis buffer containing 260 mM sucrose, 100 mM KCl, 20 mM MgCl2, 10 mM Tris– HCl, pH 7.5, 1% Triton X-100, 5 mM b-mercaptoethanol and protease inhibitor cocktail without EDTA, for 20 min on ice, then the extract was cleared by centrifugation at 9400g for 45 min at 48C prior to loading on a 10– 30% discontinuous sucrose gradient (21). After centrifugation, fractions were collected from the top and analyzed by immunoblotting.
RNA blotting analysis Total RNA was isolated with Trizol reagent (Invitrogen) from muscle, heart or liver samples, and 1.5 –10 mg were run on either MOPS/formaldehyde gels for detection of mitochondrial mRNAs (33), or on polyacrylamide/urea gels for detection of mitochondrial tRNAs with 24-mer oligonucleotide probes complementary to the 3′ end of mitochondrial and cytoplasmic tRNAs (15).
ACKNOWLEDGEMENTS The LSFC Consortium is a group of scientists and clinicians dedicated to investigating the causes of and developing treatments for LSFC. The members of the consortium are (in alphabetical order) Bruce Allen, Yan Burelle, Guy Charron, Lise Coderre, Christine DesRosiers, Catherine Laprise, Charles Morin, John Rioux and Eric A. Shoubridge. Conflict of Interest statement. None declared.
FUNDING This research was supported by a CIHR emerging team grant (CPG 102168) to the LSFC consortium and by a CIHR operating grant (MT 15460) to E.A.S.
REFERENCES 1. Koopman, W.J., Willems, P.H. and Smeitink, J.A. (2012) Monogenic mitochondrial disorders. N. Engl. J. Med., 366, 1132–1141. 2. Nunnari, J. and Suomalainen, A. (2012) Mitochondria: in sickness and in health. Cell, 148, 1145–1159. 3. Schon, E.A. and Przedborski, S. (2011) Mitochondria: the next (neurode)generation. Neuron, 70, 1033–1053. 4. Leigh, D. (1951) Subacute necrotizing encephalomyelopathy in an infant. J. Neurochem., 14, 216 –221. 5. Zhu, Z., Yao, J., Johns, T., Fu, K., De Bie, I., Macmillan, C., Cuthbert, A.P., Newbold, R.F., Wang, J., Chevrette, M. et al. (1998) SURF1, encoding a factor involved in the biogenesis of cytochrome c oxidase, is mutated in Leigh syndrome. Nat. Genet., 20, 337– 343. 6. Rahman, S., Blok, R.B., Dahl, H.H., Danks, D.M., Kirby, D.M., Chow, C.W., Christodoulou, J. and Thorburn, D.R. (1996) Leigh syndrome: clinical features and biochemical and DNA abnormalities. Ann. Neurol., 39, 343–351. 7. Merante, F., Petrova-Benedict, R., MacKay, N., Mitchell, G., Lambert, M., Morin, C., De Braekeleer, M., Laframboise, R., Gagne, R. and Robinson, B.H. (1993) A biochemically distinct form of cytochrome oxidase (COX) deficiency in the Saguenay-Lac-Saint-Jean region of Quebec. Am. J. Hum. Genet., 53, 481–487. 8. Debray, F.G., Morin, C., Janvier, A., Villeneuve, J., Maranda, B., Laframboise, R., Lacroix, J., Decarie, J.C., Robitaille, Y., Lambert, M. et al. (2011) LRPPRC mutations cause a phenotypically distinct form of Leigh syndrome with cytochrome c oxidase deficiency. J. Med. Genet., 48, 183–189. 9. Morin, C., Mitchell, G., Larochelle, J., Lambert, M., Ogier, H., Robinson, B.H. and De Braekeleer, M. (1993) Clinical, metabolic, and genetic aspects of cytochrome C oxidase deficiency in Saguenay-Lac-Saint-Jean. Am. J. Hum. Genet., 53, 488–496. 10. Mootha, V.K., Lepage, P., Miller, K., Bunkenborg, J., Reich, M., Hjerrild, M., Delmonte, T., Villeneuve, A., Sladek, R., Xu, F. et al. (2003) Identification of a gene causing human cytochrome c oxidase deficiency by integrative genomics. Proc. Natl. Acad. Sci. USA, 100, 605 –610. 11. Barkan, A. and Small, I. (2014) Pentatricopeptide repeat proteins in plants. Ann. Rev. Plant Biol., 65, 415– 442. 12. Barkan, A., Rojas, M., Fujii, S., Yap, A., Chong, Y.S., Bond, C.S. and Small, I. (2012) A combinatorial amino acid code for RNA recognition by pentatricopeptide repeat proteins. PLoS Genet., 8, e1002910. 13. Yin, P., Li, Q., Yan, C., Liu, Y., Liu, J., Yu, F., Wang, Z., Long, J., He, J., Wang, H.W. et al. (2013) Structural basis for the modular recognition of single-stranded RNA by PPR proteins. Nature, 504, 168– 171. 14. Davies, S.M., Lopez Sanchez, M.I., Narsai, R., Shearwood, A.M., Razif, M.F., Small, I.D., Whelan, J., Rackham, O. and Filipovska, A. (2012) MRPS27 is a pentatricopeptide repeat domain protein required for the translation of mitochondrially encoded proteins. FEBS Lett., 586, 3555–3561. 15. Sasarman, F., Brunel-Guitton, C., Antonicka, H., Wai, T. and Shoubridge, E.A. (2010) LRPPRC and SLIRP interact in a ribonucleoprotein complex that regulates posttranscriptional gene expression in mitochondria. Mol. Biol. Cell., 21, 1315–1323. 16. Chujo, T., Ohira, T., Sakaguchi, Y., Goshima, N., Nomura, N., Nagao, A. and Suzuki, T. (2012) LRPPRC/SLIRP suppresses PNPase-mediated mRNA decay and promotes polyadenylation in human mitochondria. Nucleic Acids Res., 40, 8033–8047. 17. Ruzzenente, B., Metodiev, M.D., Wredenberg, A., Bratic, A., Park, C.B., Camara, Y., Milenkovic, D., Zickermann, V., Wibom, R., Hultenby, K. et al. (2011) LRPPRC is necessary for polyadenylation and coordination of translation of mitochondrial mRNAs. EMBO J., 31, 443–456. 18. Baughman, J.M., Nilsson, R., Gohil, V.M., Arlow, D.H., Gauhar, Z. and Mootha, V.K. (2009) A computational screen for regulators of oxidative phosphorylation implicates SLIRP in mitochondrial RNA homeostasis. PLoS Genet., 5, e1000590. 19. Antonicka, H., Sasarman, F., Kennaway, N.G. and Shoubridge, E.A. (2006) The molecular basis for tissue specificity of the oxidative phosphorylation
Human Molecular Genetics, 2015, Vol. 24, No. 2
20.
21.
22. 23.
24.
25. 26.
deficiencies in patients with mutations in the mitochondrial translation factor EFG1. Hum. Mol. Genet., 15, 1835– 1846. Colley, S.M., Wintle, L., Searles, R., Russell, V., Firman, R.C., Smith, S., Deboer, K., Merriner, D.J., Genevieve, B., Bentel, J.M. et al. (2013) Loss of the nuclear receptor corepressor SLIRP compromises male fertility. PLoS One, 8, e70700. Antonicka, H., Sasarman, F., Nishimura, T., Paupe, V. and Shoubridge, E.A. (2013) The mitochondrial RNA-binding protein GRSF1 localizes to RNA granules and is required for posttranscriptional mitochondrial gene expression. Cell Metab., 17, 386 –398. Jourdain, A.A., Koppen, M., Wydro, M., Rodley, C.D., Lightowlers, R.N., Chrzanowska-Lightowlers, Z.M. and Martinou, J.C. (2013) GRSF1 regulates RNA processing in mitochondrial RNA granules. Cell Metab., 17, 399–410. Borowski, L.S., Dziembowski, A., Hejnowicz, M.S., Stepien, P.P. and Szczesny, R.J. (2013) Human mitochondrial RNA decay mediated by PNPase-hSuv3 complex takes place in distinct foci. Nucleic Acids Res., 41, 1223– 12240. Holzmann, J., Frank, P., Loffler, E., Bennett, K.L., Gerner, C. and Rossmanith, W. (2008) RNase P without RNA: identification and functional reconstitution of the human mitochondrial tRNA processing enzyme. Cell, 135, 462– 474. Brzezniak, L.K., Bijata, M., Szczesny, R.J. and Stepien, P.P. (2011) Involvement of human ELAC2 gene product in 3′ end processing of mitochondrial tRNAs. RNA Biol., 8, 616– 626. Ojala, D., Montoya, J. and Attardi, G. (1981) tRNA punctuation model of RNA processing in human mitochondria. Nature, 290, 470–474.
491
27. Shoubridge, E.A., Johns, T. and Boulet, L. (1996) Use of myoblast cultures to study mitochondrial myopathies. Methods Enzymol., 264, 465–475. 28. Lochmuller, H., Johns, T. and Shoubridge, E.A. (1999) Expression of the E6 and E7 genes of human papillomavirus (HPV16) extends the life span of human myoblasts. Exp. Cell Res., 248, 186–193. 29. Boulet, L., Karpati, G. and Shoubridge, E.A. (1992) Distribution and threshold expression of the tRNA(Lys) mutation in skeletal muscle of patients with myoclonic epilepsy and ragged-red fibers (MERRF). Am. J. Hum. Genet., 51, 1187–1200. 30. Sasarman, F. and Shoubridge, E.A. (2012) Radioactive labeling of mitochondrial translation products in cultured cells. Methods Mol. Biol., 837, 207– 217. 31. Leary, S.C. and Sasarman, F. (2009) Oxidative phosphorylation: synthesis of mitochondrially encoded proteins and assembly of individual structural subunits into functional holoenzyme complexes. Methods Mol. Biol., 554, 143– 162. 32. Antonicka, H., Ogilvie, I., Taivassalo, T., Anitori, R.P., Haller, R.G., Vissing, J., Kennaway, N.G. and Shoubridge, E.A. (2003) Identification and characterization of a common set of complex I assembly intermediates in mitochondria from patients with complex I deficiency. J. Biol. Chem., 278, 43081– 43088. 33. Weraarpachai, W., Antonicka, H., Sasarman, F., Seeger, J., Schrank, B., Kolesar, J.E., Lochmuller, H., Chevrette, M., Kaufman, B.A., Horvath, R. et al. (2009) Mutation in TACO1, encoding a translational activator of COX I, results in cytochrome c oxidase deficiency and late-onset Leigh syndrome. Nat. Genet., 41, 833–837.
