Discovery of EMRE in fungi resolves the true evolutionary history of the mitochondrial calcium uniporter

Calcium (Ca2+) influx into mitochondria occurs through a Ca2+-selective uniporter channel, which regulates essential cellular processes in eukaryotic organisms. Previous evolutionary analyses of its pore-forming subunits MCU and EMRE, and gatekeeper MICU1, pinpointed an evolutionary paradox: the presence of MCU homologs in fungal species devoid of any other uniporter components and of mt-Ca2+ uptake. Here, we trace the mt-Ca2+ uniporter evolution across 1,156 fully-sequenced eukaryotes and show that animal and fungal MCUs represent two distinct paralogous subfamilies originating from an ancestral duplication. Accordingly, we find EMRE orthologs outside Holoza and uncover the existence of an animal-like uniporter within chytrid fungi, which enables mt-Ca2+ uptake when reconstituted in vivo in the yeast Saccharomyces cerevisiae. Our study represents the most comprehensive phylogenomic analysis of the mt-Ca2+ uptake system and demonstrates that MCU, EMRE, and MICU formed the core of the ancestral opisthokont uniporter, with major implications for comparative structural and functional studies.


Main text
Comparative genomics analyses, based on a few eukaryotic species, in combination with RNAi assays were instrumental in the identification of MCU and MICU1 as the founding members of the mt-Ca 2+ uniporter [5][6][7] . This discovery paved the way to the identification of other paralogous components of this channel in mammalian cells, including negative and positive regulatory and tissue-specific subunits (MCUb 17 , MICU2 18,19 , MICU3 20 , and MICU1.1 21 ). Instead, EMRE was found as a specific interactor of MCU in human cells, required for both conductivity and binding of the channel to MICU1 8 . Although MCU and MICU1 showed correlated evolutionary histories across 138 sequenced eukaryotic organisms, EMRE apparently lacked any homolog outside the metazoan lineage and it was therefore suggested to be an animal-specific innovation [9][10][11] . While those observations pointed to an ancient eukaryotic origin of mt-Ca 2+ uptake, they also implied a very different composition and regulation of the uniporter in different clades. A notable case is the identification of MCU as the only uniporter component in Basidiomycota and filamentous Ascomycota (e.g., Neurospora crassa and Aspergillus fumigatus), suggesting that either fungal MCUs are sufficient for mt-Ca 2+ uptake or they are regulated independently of MICU1 and EMRE 9,10 . Based on the assumption of an orthologous relationship between human and fungal MCUs 9,10,22 , several independent structural studies of Ascomycota MCUs have been performed to understand the basic principles of uniporter channel assembly and function [12][13][14][15] . However, those organisms had been shown to lack uniporter activity 23,24 and their MCU homologs were unable to mediate mt-Ca 2+ uptake when heterologously expressed in HeLa or yeast cells 14,25 . Not surprisingly, significant structural and sequence differences were found between fungal MCUs and their animal counterparts [12][13][14][15] , raising the question of whether fungal MCUs function as classical Ca 2+ uniporters at all.
To resolve this paradox, we assessed the evolution of each uniporter component across 1,156 fullysequenced eukaryotic genomes (see Supplementary Table 1), using a combination of profile-based sequence searches, protein domain composition assessment, and phylogenetics. As shown in Fig. 1 (Extended Data Fig. 1), the overall taxonomic distributions of MCU and MICU1 homologs were largely congruent with that of previous genomics surveys 9,10 . We confirmed the presence of MCU in at least some species of the major eukaryotic groups (Unikonts, the SAR clade, Plants and Euglenozoa), and its absence in all sequenced Apicomplexans, all yeasts in Saccharomycotina and most in Schizosaccharomyces clades, Microsporidia, Trichomonas and Giardia. Hence, mt-Ca 2+ uptake appeared to have been lost many times independently during the evolution of eukaryotes. A significant number of these losses correlated with extreme streamlining of mitochondrial metabolism, as most MCU/MICU-lacking lineages encompassed relict forms of anaerobic mitochondria, such as mitosomes (Microsporidians, Entamoeba, Giardia, Cryptosporidium) or hydrogenosomes (Trichomonas) 26 . Our homology-based results confirmed the abovementioned anomaly that most fungal genomes, for which our dataset is particularly rich -776 species as compared to 50 in previous studies 9,10 -encode for MCU but not MICU or EMRE. Unexpectedly, however, our analysis uncovered for the first time the presence of EMRE outside Holozoa, identifying reliable orthologs in three chytrid fungi -an early-diverging zoosporic fungal lineage: Allomyces macrogynus, Catenaria anguillulae, and Spizellomyces punctatus. Additional searches in public databases confirmed that EMRE was not present in other sequenced fungi.
To clarify the underlying evolutionary history of the uniporter, we reconstructed and inspected the molecular phylogenies of MCU (Fig. 2a, Extended Data Fig. 2a) and MICU1 (Fig. 2b, Extended Data Fig. 2b) homologs across eukaryotes. We found that the evolution of both MCU and MICU gene families was driven by numerous gene duplications and losses, some of them having occurred in parallel in different lineages, implying an ancient and tight functional relationship. Furthermore, we showed that evolutionary independent duplications at the base of several main eukaryotic lineages -Vertebrates, Streptophytes, Oomycetes, Kinetoplastids, and Ciliates -resulted in the existence of multiple MCU paralogous copies in each of these clades. As a result, orthology relationships within the MCU gene family are complex and of the type manyto-many 27 . This means, for instance, that human MCU and MCUb are equally distant evolutionarily (coorthologous) to each non-vertebrate animal MCUs, and also to all MCUs from protists and plants, explaining why Trypanosoma's MCUb does not functionally complement human MCUb 28,29 . It also implies that shared physical interactions between paralogous MCU proteins in hetero-oligomers in different organisms, as shown for human and Trypanosoma brucei MCUb proteins 17,28,29 , are the result of parallel evolution. Importantly, one duplication event in the MCU gene family occurred in the common ancestor of opisthokonts and was followed by differential losses that distinguished Holomycota (fungi and their relatives, including Fonticula alba) from the other opisthokonts, i.e. the Holozoa (animals and their unicellular close relatives). Most fungal species kept only one of the two MCU paralogs that is referred here as the "fungal-specific" MCU (fsMCU). Holozoa, instead, retained the other MCU paralog, the bona-fide "animal" MCU. Only three chytrid fungi in our dataset, A. macrogynus, C. anguillulae, and S. punctatus retained both the fsMCU and the "animal" MCU, and these are also the only fungi encoding MICU1 and EMRE homologs. This striking, previously undetected, co-evolution pattern between MCU, MICU and EMRE in fungi suggests a strong interdependence, and even stronger considering that Blastocladiales (Allomyces and Catenaria) and Spizellomyces do not form a monophyletic clade 30 . Based on these findings, we hypothesized that these "animal-like" MCUs present in chytrids should require EMRE to drive mt-Ca 2+ uptake, similarly to their human ortholog.
The above mentioned finding of bona-fide EMRE orthologs in all these three cythrids ( Fig. 1) reinforced this idea and placed back the origin of an animal-like mt-Ca 2+ uptake in the opisthokont ancestor, preceding the diversification of animals and fungi. Consistently, the heterologous expression of MCU from Dictyostelium discoideum, representing an amoebozoan lineage that diverged earlier than the origin of opisthokonts, is alone sufficient to reconstitute mt-Ca 2+ uptake in yeast mitochondria, while human (Hs-) MCU only does so in the presence of EMRE 31 . Similarly, we hypothesized that co-expression of "animal" MCUs and EMRE proteins from chytrids would be necessary and sufficient to reconstitute uniporter activity. The phylogenetic distribution profile (presence/absence) across the MCU complex components reveals a strong co-evolution pattern, when only the true orthologous sequences are considered (Fig. 3a). Strikingly, we detected mt-Ca 2+ uptake in yeast strains expressing "animal" MCUs from either A. macrogynus (Am-MCUa) or S. punctatus (Sp-MCU) with their respective EMREs (Am-EMRE, Sp-EMRE) (Fig. 3b,c, Extended Data Fig. 3). In contrast, we did not detect any mt-Ca 2+ uptake in yeast strains expressing fsMCU proteins from A. macrogynus (Am-fsMCU1) and S.punctatus (Sp-fsMCU), despite proper expression and localization (Extended Data Fig. 3 and Extended Data Fig. 4). Similar results would be expected in other Holozoa despite the inability to detect EMRE by similarity searches. Indeed, the co-expression of MCU from the sea anemone Nematostella vectensis (Nv-MCU) with Hs-EMRE in yeast was able to reconstitute mt-Ca 2+ uptake to a similar extent of a strain expressing Hs-MCU and Hs-EMRE (Extended Data Fig. 5). These results, together with the absence of MICU proteins in most fungal lineages, indicate that mt-Ca 2+ uptake in fungal mitochondria, if it exists, is not mediated by fsMCUs, or that a different -yet unknown-regulator is necessary. Instead, animal-like MCUs from chytrid fungi and Holozoa function similarly to the mammalian uniporter, in an EMRE-dependent fashion. Altogether, our evolutionary analyses and experimental results confirm that MCU-EMRE interaction is conserved, and was already present in the last common ancestor of fungi and animals.
Consistent with our hypothesis that fsMCUs do not represent true functional orthologs of Hs-MCU, when comparing MCU sequences across eukaryotes we found that fsMCUs lack key residues conserved in the animal-like MCUs, despite retaining a DXXE motif ( Fig. 4a and Extended Data Fig. 6a). Those residues have been previously shown to be important for MCU function and its interaction with EMRE 32 , Notably, animal and fungal EMREs appear highly divergent ( Fig. 4b and Extended Fig. 6b), although the MCU interacting domain GXXXA/S/G and the polyaspartate tail necessary for the binding to MICU1 32 are fully conserved. Interestingly, the fungal EMRE sequences contained an extra C-terminal domain that is not found in Holozoa, suggesting some degree of specialization. Thus, we hypothesized that Am-MCUa and Sp-MCU would have evolved to interact with EMRE proteins from the same or related species. Indeed, those animallike MCUs were neither able to increase mt-Ca 2+ uptake when expressed in a wild-type HeLa background nor to rescue mt-Ca 2+ uptake in MCU knock-down (shMCU) HeLa cells (Extended Data Fig. 7), despite showing proper expression and insertion into the inner mitochondrial membrane (Extended Data Fig. 8). Furthermore, the expression of Am-MCUa and Sp-MCU in yeast mitochondria was unable to reconstitute mt-Ca 2+ uptake in the presence of Hs-EMRE ( Fig. 4c and Extended Data Fig. 9a). Instead, Hs-MCU was functional when co-expressed with either Am-EMRE or Sp-EMRE ( Fig. 4d and Extended Data Fig. 9b). Altogether, these results suggest that the C-terminal domain of fungal EMREs is dispensable for a functional interaction with Hs-MCU but necessary to activate animal-like fungal MCUs. Accordingly, we observed that the co-expression of Am-MCUa and Sp-MCU with Am-EMRE and Sp-EMRE lacking the extra C-terminal domain (EMRE-t) was unable to efficiently reconstitute mt-Ca 2+ uptake ( Fig. 4e and Extended Data Fig.  10a). However, while the presence of the extra C-terminal domain in Hs-EMRE did not affect the function of Hs-MCU, it was not sufficient to reconstitute mt-Ca 2+ uptake when co-expressed with Am-MCUa and Sp-MCU ( Fig. 4f and Extended Data Fig. 10b). On the one hand, these findings hint a possible activating role  The phylogenetic distribution of MCU (red), MICU1 (green), and EMRE (gray) homologs across 1,156 eukaryotic genomes is shown on the NCBI taxonomy tree. Viridiplantae and Rhodophyta (red algae) have been grouped together as Archaeplastida, and Alveolates, Stramenopiles (Str/es) and Rhizaria as the SAR clade. In all cases where data from various strains of a species are present with the same pattern, these have all been collapsed to the species level, resulting in 969 terminal nodes shown. The mt-Ca 2+ uniporter complex has been completely lost in Apicomplexa within Alveolates, Rhizaria (5 genomes), red algae (3 genomes), Cryptophytes (3 genomes), Haptophytes (1 genome), and the Entamoeba clade within Amoebozoa. Within fungi (in purple), all major clades that have completely lost MCU homologs are indicated with a darker purple color, namely Onygenales, Saccharomycetales, Pucciniomycotina, Mucoromycotina (Mucor/a), and Microsporidia (Micro/a). The only three early diverging fungal species (A.macrogynus, C.anguillae, S.punctatus) that encode also MICU and EMRE are highlighted with a red arrow. The NCBI taxonomy and the presence/absence profile were visualized using the ETE toolkit 33 . For a version of the profile, which includes the species names, see Extended Data Figure 1. structural data are shown in black circles, while the numbers refer to the publication reference. fsMCU, "fungal-specific" MCU. In (b) all Aralar-related MICU homologs were excluded after a first pre-processing, but are shown here as reference (See also Methods).

Figure 3: Functional reconstitution of mt-Ca 2+ uptake by fungal MCU and EMRE orthologs. a,
Phylogenetic distribution profile (presence/absence) across MCU complex components. The distribution pattern of MICU and EMRE largely overlaps with that of the animal like MCU, but not the fsMCU. b,c, Representative traces and quantification of mt-Ca 2+ transients in yeast cells expressing animal-like or fungalspecific MCU orthologs from S. punctatus (b) and A. macrogynus (c) with either their respective EMRE proteins or empty vector (p425) upon glucose-induced calcium (GIC) stimulation in presence of 1 mM CaCl 2 . All data represent mean ± SEM (n=3); ***p < 0.0001, **p < 0.001, one-way ANOVA with Dunnett's Multiple Comparisons Test.

Figure 4: Evolution of MCU-EMRE interaction. a,b,
Phylogenetic trees of members of MCU (a) and EMRE (b) protein families, and sequence diversity of major domains. The sequence alignment of TM1 and TM2 of MCU sequences from 20 species is shown in (a). The program Multi-Harmony 34 was used to detect residues that are overall conserved but differ in the fsMCU members (highest scoring positions are indicated with red arrows). The fsMCU clade is shown in purple. The evolutionary point where the MCU proteins become EMRE-dependent is shown in gray. The degree of conservation across the animal related MCU members is very high in these loci, while few positions are Holomycota or Holozoa specific. Similarly in (b) EMRE's sequence diversity across opisthokonts is shown for the β-hairpin, the TM, and CAD domain. Residues found important for the interaction between MCU and EMRE in 11 and fully conserved positions are indicated with gray and red arrows, respectively. c,d,e,f Representative traces and quantification of mt-Ca 2+ transients in yeast cells expressing either human and animal-like S. punctatus and A. macrogynus MCU with human EMRE (c), species-specific EMRE with human MCU (d), animal-like S. punctatus and A. macrogynus MCU with their respective wild-type or truncated (-t) EMREs (e) or with human EMRE fused to the fungal extra C-terminal domain (f) upon glucose-induced calcium (GIC) stimulation in presence of 1 mM CaCl 2 . All data represent mean ± SEM; n=4; ***p < 0.0001, *p < 0.01, one-way ANOVA with Dunnett's Multiple Comparisons Test.  Fig. 3a). Losses of MCU in the various eukaryotic lineages are indicated in red. The color code and symbols of the different taxonomic groups are the same as in Fig. 1 and Fig. 3a, respectively.

Sequence data and Homology searches
The protein sequences encoded in 1,156 completely sequenced eukaryotic genomes were retrieved from Ensembl DB v91, and v37 of Ensembl Metazoa, Plants, Fungi, and Protists (see Supplementary Table 1). Only the genome of Catenaria anguillulae PL171 was added from Ensembl Fungi v41. For each of the protein families studied, homologs were selected on the basis of sequence similarity and phylogenetic analysis. HMMER searches were performed using HMMER 3.1b2 35 and using the Gathering Cut-Off threshold (--cut_ga) when the raw HMM profile from Pfam was used, an e-value threshold of 10 -2 otherwise. For all BLAST searches low complexity regions in the query sequence (default parameter) were filtered out to minimise the number of false positives and an E-value threshold of 10 -5 was used. Conserved domains in all retrieved sequences, were annotated using the HMM profiles of Pfam release 30.0. MCU Proteins in our database containing at least one MCU (Pfam:PF04678) domain were detected using an HMMsearch. 1,076 protein sequences were selected for subsequent analysis. MICU 2,105 protein sequences were retrieved from a BLAST search using Hs-MICU1 (Uniprot: Q9BPX6) as a query. HMMscan was used to search for additional domains in the retrieved sequences using all the Pfam domain profiles, and all the sequences with detected at least one Mito_carr domain (Pfam:PF00153) were classified as members of the mitochondrial carrier family (slc25a12-Aralar homologs). The Aralar-related sequences (Mito_carr domain containing) were clustered together clearly as a monophyletic clade in a phylogenetic tree, in the exclusion of known MICU sequences, and were excluded. The remaining 651 MICU sequences were re-aligned and a new phylogenetic tree with the same methods was reconstructed. EMRE EMRE sequences are characterized by a DDDD (Pfam:PF10161) domain. Their short length and low sequence conservation makes detection strikingly difficult, which explains why in some cases EMRE appears to be missing even from some animal genomes. HMMER searches with the "gathering" (--cut_ga) threshold were performed, and all detected homologs were retrieved and re-aligned, for a new HMM profile to be built and used to search back the genome database in a second iteration. One EMRE sequence was detected in A.macrogynus in the first search, while one more sequence in the same species, as well as in S.punctatus and C.anguillulae, were detected in the second iteration. The detected EMRE sequences in this second iteration are those that were further considered for all analyses. NCLX (NCKX6) Na_Ca_ex Pfam domain characterizes an ubiquitous superfamily of sodium/calcium exchangers that regulate intracellular Ca 2+ concentrations in many cell types. Therefore selecting on that domain of the NCKX6 (NCLX) homologs using HMMER returned 4024 hits. To narrow down the number of hits for more accurate alignment and phylogenetic reconstruction, we used the human NCLX sequence as a query (uniprot:Q6J4K2) for a blast search, retrieving 2,105 sequences for phylogenetic analysis. Using the human members as reference, 1,391 sequences across eukaryotes were selected as related to the NCLX clade (NCLX orthologs).

Isolation of Crude Mitochondria from HeLa Cells
Crude mitochondria were isolated from HeLa cells as previously described 25 . Briefly, HeLa cells were grown to confluency, rinsed with PBS and resuspended in ice-cold isolation buffer (IB: 220 mM mannitol, 70 mM sucrose, 5 mM HEPES-KOH pH 7.4, 1 mM EGTA-KOH pH 7.4, protease inhibitors). Cells were permeabilized by nitrogen cavitation at 600 psi for 10 minutes at 4°C and then centrifuged at 600 x g for 10 minutes. The supernatant was transferred into new tubes and centrifuged at 8000 x g for 10 minutes at 4 °C. The resulting pellet containing crude mitochondria was resuspended in IB for protein topology analysis.

Analysis of Mitochondrial Protein Topology
Proteinase K (PK) protection assay was performed on mitochondria isolated from HeLa cells as previously described 25 . Roughly, 30 µg of freshly isolated mitochondria were gently resuspended in 30 µl of IB buffer with either increasing concentrations of digitonin or 1% Triton X-100 in the presence of 100 µg/ml PK and incubated at room temperature for 15 minutes. The reaction was stopped by the addition of 5 mM PMSF, followed by incubation on ice for 10 minutes. Samples were mixed with 10 µl of 4 X Laemmli buffer containing 10 % 2-mercaptoethanol and boiled for 5 minutes at 98 °C for immunoblot analysis.

Subcellular Fractionation of Yeast Cells
Expression and subcellular localization of heterologous expressed proteins in yeast was tested by immunoblot analysis of cytosolic and mitochondrial fractions isolated from recombinant yeast strains as previously described 25 . Briefly, yeast cells were grown at 30°C in a selective lactate medium supplemented with the respective selection markers till an OD ~0.8. The cell pellet was re-suspended in a buffer containing buffer 0.6 M sorbitol, 20 mM HEPES/KOH pH 7.2, 80 mM KCl, and 1 mM PMSF, and vortexed five times for 30 seconds with glass beads (425-600 µm diameter), with a 30 seconds cooling interval in between to break cell wall and plasma membrane. After a first centrifugation step at 1000 g for 5 minutes at 4°C, the supernatant was further centrifuged at 20,000 g for 10 minutes at 4°C to obtain the mitochondrial fraction (pellet). The supernatant (cytosolic fraction) was precipitated with trichloroacetic acid at -20°C for 1 hour, washed once with cold acetone and centrifuged at 20,000 g for 10 minutes at 4°C.

Measurements of Mitochondrial Calcium Uptake in Yeast and HeLa Cells
In vivo analyses of mitochondrial Ca 2+ uptake in intact yeast cells were performed as previously described 25 . Briefly, yeast cells were collected at an OD ~0.8, washed three times with milliQ water and starved for 1.5 hours at room temperature in a nutrient-free buffer (NFB, 100 mM Tris, pH 6.5 (1x10 8 cells/mL). Afterwards, cells were collected at 3,500 rpm for 5 minutes and resuspended in NFB to a higher density (25x10 8 cells/mL) in the presence of 50 µM native coelenterazine (Abcam, ab145165) to reconstitute the photoprotein aequorin. After 30 minutes in the dark at room temperature, 0.5x10 8 cells/well were plated into a white 96-well plate and Ca 2+ -dependent light kinetics was recorded upon stimulation with 1 mM CaCl 2 and 100 mM glucose, at 0.5 seconds interval in a MicroBeta2 LumiJET Microplate Counter. At the end of each experiment, cells were lysed with 1 mM digitonin for 5 minutes at 37°C and any residual aequorin counts were collected upon the addition of CaCl 2 to a final concentration of 140 mM. Mitochondrial Ca 2+ uptake was measured in mt-AEQ HeLa cells as previously described 40 .

Quantification of Calcium Transients
Quantification of mt-Ca 2+ concentration was performed using a MATLAB software as previously described in 40 . The dynamics of mt-Ca 2+ -dependent luminescence signal was smoothed by the cubic spline function: Where, p is a smoothing parameter, controlling the tradeoff between fidelity to the data and roughness of the function estimate, f is the estimated cubic spline function to minimize the above function, and x i and y i are the dynamical data points. Here, p is set at 0.5. Parametrization of the Ca 2+ -dependent luminescence kinetics was performed in order to determine the maximal amplitude of the luminescence signal (peak) and the left slope of the bell-shaped kinetic trace. Aequorin-based luminescence signal calibration into mt-Ca 2+ concentration was performed using the algorithm reported in 43 for wild-type aequorin and native coelenterazine, with the following formula: Where l= 1, K R = 7.23x10 6 , K TR = 120 and n = 2.99 are the calibration values used for WT aequorin and native coelenterazine.

Data Analysis
Data are represented as mean ± SEM and the statistical analysis of each experiment is described in the figure legends including the statistical tests used and the exact value of biological replicates. For each biological replicate experiment at least 3 technical replicates were used for quantification and data analysis. Normal distribution was tested by Shapiro-Wilk normality test. Statistical tests between multiple datasets and conditions were carried out using one-way analysis of variance (ANOVA) followed by Dunnett's Multiple Comparison tests. Statistical analyses were performed using GraphPad Prism (GraphPad Software, version 7).            L MA VQ L A V I S R L T F VD L DWD I ME P V S Y F L G SG T S I L F L I Y L L RN  N W I R L G L A Y I V L Q AG L V A R L TWWE L SWD I M E P V T Y M L T F T TG I G AM A Y F TH T  R I IWAG L G Y L V L Q AG V IG R L TWWD L SWD IME P V TY F V S F G T V L I G Y T Y F T L T  R V I WL G L G Y L V AQ A A I I G R L TWWE L SWD I M E P V T Y F V S F G T V L I G Y I Y F T F T  T I I Y TG L G Y C F VQ A A I L G R L TWWD L SWD I I E P V S Y F L T F G S V L F G Y C Y F S F T  A I IWTG L G YC F AQ A A I L A R L TWWD L SWD I I E P V S Y F L TF G S V L IG Y TY F TM T  R VMWGGWV F S AGQ L A F I A R L TWWE F SWD VM E P I S Y VMG V AN V A L F S F L F Q F R  RMM F L G F AG V V AQ L G I I T R L TWWE F SWD I M E Figure 1. Phylogenetic distribution of the mitochondrial calcium transporter complex protein families. The tree is collapsed to the species level, resulting in 969 species. Extended version of Figure 1, including also the distribution of NCLX, which appears to be largely uncoupled to that of the MCU complex members.     Fig. 4. Heterologous expression of S. punctatus and A. macrogynus MCU and EMRE orthologs in yeast. a,b, Immunoblot analysis of cytosolic (C) and mitochondrial (M) fractions isolated from yeast clones (Cl.) expressing mt-AEQ together with either (a) S. punctatus MCU (Sp-fsMCU, Sp-MCU) and EMRE (Sp-EMRE) orthologs or (b) A. macrogynus MCU (Am-fsMCU1, Am-MCUa, Am-MCUb) and EMRE (Am-EMRE1, Am-EMRE2) orthologs fused to a C-terminal V5-tag, using the following antibodies: α-V5 (Life Technologies, R96025), α-AEQ (Merck/Millipore, MAB4405), α-YME1, PGK1 (Life Technologies, 459250). YME1 was used as control for yeast mitochondrial targeted protein and PGK1 was used as control for cytosolic protein.    -+ + + + + + + + + + + + + Extended Data Fig. 8. Expression and localization of fungal MCU orthologs in HeLa cells. Immunoblot analysis of whole cell lysates from wild-type (a) and MCU knock-down (shMCU) (b) HeLa mt-AEQ cells stably expressing human or fungal MCU proteins fused to a C-terminal V5 tag using the following antibodies: α-MCU (Sigma Aldrich, HPA01648), α-V5 (Life Technologies, R96025), α-EMRE (Santa Cruz Biotechnology, sc-86337), α-ACTIN (Sigma-Aldrich, A2228). NI, not infected. c, Analysis of protein topology by proteinase K (PK) treatment of mitochondria isolated from sh-MCU HeLa mt-AEQ cells expressing human and fungal MCU, using the following antibodies: α-V5 (Life Technologies, R96025), α-TIM23 (BD Bioscience, 611222), α-TOM20 (Abcam, ab56783), and α-Cyclophilin D (Abcam, ab110324). TOM20, TIM23, and CypD were used as controls for integral mitochondrial outer membrane, inner membrane and soluble matrix targeted proteins, respectively. T, triton (1%); Dig., digitonin.