Different Roles of Mitochondrial Calcium Uniporter Complex Subunits in Growth and Infectivity of Trypanosoma cruzi

ABSTRACT Trypanosoma cruzi is the agent of Chagas disease, and the finding that this parasite possesses a mitochondrial calcium uniporter (TcMCU) with characteristics similar to that of mammalian mitochondria was fundamental for the discovery of the molecular nature of MCU in eukaryotes. We report here that ablation of TcMCU, or its paralog TcMCUb, by clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 led to a marked decrease in mitochondrial Ca2+ uptake without affecting the membrane potential of these cells, whereas overexpression of each gene caused a significant increase in the ability of mitochondria to accumulate Ca2+. While TcMCU-knockout (KO) epimastigotes were viable and able to differentiate into trypomastigotes, infect host cells, and replicate normally, ablation of TcMCUb resulted in epimastigotes having an important growth defect, lower rates of respiration and metacyclogenesis, more pronounced autophagy changes under starvation, and significantly reduced infectivity. Overexpression of TcMCUb, in contrast to what was proposed for its mammalian ortholog, did not result in a dominant negative effect on TcMCU.

T rypanosoma cruzi is the etiologic agent of Chagas disease, an enormous burden on human health in the American continent, which has four major developmental stages that alternate between an insect vector and a mammalian host. Two are replicative forms, the epimastigote found in the insect vector intestine and the intracellular mammalian form or amastigote, and two are nonreplicative, the metacyclic trypomastigote found in the rectum and urine of the vector and the bloodstream trypomastigote found in the mammalian host. All these forms appear to have functional mitochondria with an active oxidative metabolism (1,2).
The finding that mitochondrial Ca 2ϩ transport in T. cruzi is electrogenic, has low affinity and high capacity, and is inhibited by ruthenium red, as occurs with vertebrate mitochondria, identified the presence of a mitochondrial calcium uniporter (MCU) in trypanosomatids (3,4). The presence of MCU in trypanosomes together with its absence in yeast (5) led to the identification, first, of the gene encoding an MCU modulator (mitochondrial calcium uptake 1 [MICU1]) (6) and then of the gene encoding the MCU of mammalian cells (7)(8)(9). In recent work, we demonstrated that the MCU of Trypanosoma brucei, which belongs to the group of parasites that produce African trypanosomiasis or sleeping sickness, is essential for the regulation of the bioenergetics of the parasite and its growth and infectivity (10).
After the discovery of the molecular identity of MCU, other components of the MCU complex were described (11,12). One of them, MCUb, was shown to exert a dominant negative effect in HeLa cell mitochondria, reducing the mitochondrial Ca 2ϩ increase evoked by agonist stimulation (13).
In this work, we report that knockout of the T. cruzi MCU (TcMCU) gene by clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 in T. cruzi epimastigotes (14) abolishes mitochondrial calcium uptake without affecting their mitochondrial membrane potential (Δ⌿ m ) and reduces growth in low-glucose medium. However, epimastigotes conserve their ability to differentiate into metacyclic trypomastigotes and infect mammalian cells. In contrast to a previous report on HeLa cells (13), overexpression of TcMCUb does not have a dominant negative effect but increases mitochondrial Ca 2ϩ uptake without affecting the Δ⌿ m . Knockout of TcMCUb by CRISPR/ Cas9 abolishes mitochondrial Ca 2ϩ transport, reduces respiration, has a significant effect on epimastigote growth, and increases autophagy. These cells have a reduced ability to differentiate into metacyclic trypomastigotes and are unable to efficiently infect cells, underscoring the relevance of TcMCUb for the parasite life cycle.

RESULTS
Ca 2؉ uptake by TcMCU and TcMCUb knockouts. After the recent characterization of MCU as the channel-forming subunit of the mitochondrial calcium uniporter complex (15), several pore regulators were reported, among them mitochondrial calcium uptake 1 (MICU1), MICU2, MCUb, essential MCU regulator (EMRE), and MCU regulator 1 (MCUR1) (11). It has been suggested that MCUb is a dominant negative regulator of the uniporter complex (13). However, evidence of its influence on MCU regulation is lacking. MCUb has been identified in the T. cruzi genome, and similarly to its paralog MCU, it has two predicted transmembrane domains (16). Therefore, we aimed at investigating the effect of downregulation and overexpression of TcMCUb on the physiological role of the MCU complex in T. cruzi.
We used the recently developed CRISPR/Cas9 system with a DNA donor cassette for DNA repair (14) to knock out TcMCU ( Fig. 1A and B) and TbMCUb ( Fig. 1E and F) in epimastigotes. After 5 to 6 weeks of selection under G418 and blasticidin, we obtained resistant populations of epimastigotes transfected with specific single guide RNAs (sgRNAs) and blasticidin cassettes. These were constructed with long (~500-bp) flanking untranslated regions (UTRs) cloned in pGEM-T Easy vector to obtain TcMCU-KO cells or using long oligonucleotides (ultramers) with shorter homology regions (100 bp) to obtain the TcMCUb-KO cell line. Primers (see Table S1 in the supplemental material) were used in PCR experiments to check TcMCU and TcMCUb disruption in G418/ blasticidin-resistant cells. As shown in Fig. 1C, TcMCU was ablated and replaced by the blasticidin resistance gene in TcMCU-KO cells. Disruption of the TcMCUb gene was also demonstrated using specific primers (Fig. 1G). In addition, Southern blot analyses demonstrated that TcMCU (Fig. 1D) and TcMCUb (Fig. 1H) were absent in genomic DNA of the KO cell lines.
To determine the capacity of TcMCU-KO and TcMCUb-KO cell lines to take up Ca 2ϩ , we monitored Ca 2ϩ uptake with calcium green-5N probe in digitonin-permeabilized epimastigotes. A decrease in fluorescence indicates decreasing medium Ca 2ϩ or increasing vesicular Ca 2ϩ . Figures 1I and K show that addition of 50 M digitonin in the presence of 5 mM succinate and 20 M Ca 2ϩ produced a fast decrease in Ca 2ϩ concentration starting after a period of about 60 s, which continued until addition of the uncoupler p-trifluoromethoxyphenylhydrazone (FCCP), which released the mitochondrial Ca 2ϩ taken up. This Ca 2ϩ -transporting activity was fully eliminated by the addition of 5 M ruthenium red, indicating that it is due to the uniporter. Knockout of localizes to the mitochondria of T. cruzi (17). TcMCU-OE protein also localized to the mitochondria, as demonstrated by immunofluorescence microscopy (see Fig. S1A). Antibody to T. brucei MCU (10) colocalized with antibody against the voltage-dependent anion channel (TbVDAC), which is a major protein of the outer mitochondrial membrane of eukaryotes. Western blot analysis detected a band of the expected size (32 kDa) in the overexpressing cell line ( Fig. 2A). The band was not visible in wild-type (WT) cells, probably as a consequence of its low expression levels. TcMCU-OE and TcMCUb-OE (Fig. S1B) cells exhibited the same growth rate as control cells transfected with the pTREX-n empty vector. Western blot analysis of TcMCUb-OE cells using antibodies against its hemagglutinin (HA) tag confirmed its expression (Fig. 2B), and immunofluorescence analysis demonstrated its mitochondrial localization (Fig. S1C).
To investigate whether TcMCU and TcMCUb could form oligomers as previously described with their orthologs in HeLa cells (13), we used a T. cruzi cell line that coexpresses the c-Myc-tagged TcMCU (TcMCU-3ϫc-Myc) and the HA-tagged TcMCUb (TcMCUb-3ϫHA), as described in Materials and Methods. Cells were lysed, and immunoprecipitation was done with either anti-c-Myc or anti-HA antibody. Western blot analyses showed that the anti-HA antibody immunoprecipitated TcMCU-3ϫc-Myc and the anti-c-Myc antibody immunoprecipitated TcMCUb-3ϫHA, revealing the MCU-MCUb interaction in situ (Fig. 2C).
TcMCU-OE ( Fig. 2D and E) and TcMCUb-OE ( Fig. 2H and I) epimastigotes showed increased Ca 2ϩ uptake. The mitochondria of permeabilized control epimastigotes were capable of buffering multiple pulses of exogenously added Ca 2ϩ , and overexpression of TcMCU ( Fig. 2F and G) or TcMCUb ( Fig. 2J and K) increased significantly the ability of their mitochondria to accumulate Ca 2ϩ in response to Ca 2ϩ pulses. Overexpression of TcMCUb in epimastigotes possessing endogenously tagged TcMCU did not induce its overexpression (Fig. S1D).
Analysis of the mitochondrial membrane potential (⌬⌿ m ) of mutant cell lines. Safranin O was used to measure Δ⌿ m in digitonin-permeabilized epimastigotes in the presence of succinate as the mitochondrial substrate. To be certain that the Δ⌿ m was not affected in the mutants, we calibrated the Δ⌿ m using valinomycin and potassium as we described before (18) and expressed the changes in millivolts instead of fluorescence arbitrary units. The magnitude of this membrane potential of epimastigote mitochondria respiring on succinate was 177 Ϯ 2.0, 179.6 Ϯ 2.8, 177.6 Ϯ 2.3, 177.7 Ϯ 1.6, and 177.9 Ϯ 1.7 mV for wild-type, TcMCU-KO, TcMCU-OE, TcMCUb-KO, and TcMCUb-OE cells, respectively (n ϭ 3). When using safranin O, an increase in fluorescence after addition of digitonin indicated stacking of the dye to the energized inner mitochondrial membrane ( Fig. S2A to C). Addition of ADP produced the expected small decrease in membrane potential, indicating ADP phosphorylation. Δ⌿ m returned to its initial level after addition of the adenine nucleotide translocator inhibitor carboxyatractyloside (CAT). Addition of FCCP collapsed the membrane potential. Neither knockout nor overexpression of either TcMCU (Fig. S2A to C) or TcMCUb (Fig. S2D to F) affected the Δ⌿ m at the steady state or ADP phosphorylation.
Growth of TcMCU and TcMCUb mutants. We initially evaluated the growth rate of epimastigotes in liver infusion tryptose (LIT) medium. TcMCU-KO cells grew at a slightly lower initial rate than wild-type cells or cells transfected with a scrambled sgRNA (Fig. 3A). We then tested their growth in low-glucose LIT medium. In this case, the initial growth rate of TcMCU-KO cells was significantly lower. Wild-type epimastigotes reached the stationary phase earlier (day 8) and then started dying, while TcMCU-KO cells reached the stationary phase later (day 11) and continued at steady state for a longer period (Fig. 3B). The longer survival of TcMCU-KO cells in low-glucose medium suggested that these cells could have access to an energy reserve. Amino acids or fatty acids derived from triglycerides have been proposed as a possible source of energy of epimastigotes under low-glucose conditions (1). We therefore investigated the presence of lipid droplets in TcMCU-KO cells under these conditions. Lipid droplets were more abundant in TcMCU-KO cells grown in LIT medium or in low-glucose LIT medium than in wild-type cells at the beginning of the stationary phase, suggesting triglycerides as a potential energy reserve of these cells ( Fig. S3A and B).
In contrast to TcMCU-KO cells, TcMCUb-KO cells grew in LIT medium at a significantly lower rate than wild-type cells or cells transfected with a scrambled sgRNA (Fig. 3C).
Oxidative stress in TcMCU-OE and TcMCUb-OE cells. Overexpression of TcMCU or TcMCUb resulted in higher basal mitochondrial Ca 2ϩ concentration, as measured by Rhod-2 fluorescence (Fig. 3D). Mitochondrial Ca 2ϩ overload is known to generate oxidative stress. Accordingly, reactive oxygen species (ROS) generation was increased in digitonin-permeabilized TcMCU-OE or TcMCUb-OE epimastigotes compared with control cells (wild type [WT]) when in the presence of 0.5 mM Ca 2ϩ ( Fig. 3E and F). They reached levels comparable to those produced by addition of antimycin A 1 , a known complex III inhibitor of the electron transport chain ( Fig. 3E and F).
Differentiation and infectivity of mutant cells. To induce the differentiation of mutant epimastigotes into infective metacyclic trypomastigotes (metacyclogenesis), we incubated these cells in triatome artificial urine (TAU) medium as described in Materials and Methods. TcMCU-KO cells were able to differentiate to metacyclic trypomastigotes in a higher proportion than wild-type or TcMCU-OE cells (Fig. 3G). The TcMCU-KO metacyclic trypomastigotes were able to infect host cells. After several cycles of infection to obtain a sufficient amount of culture-derived trypomastigotes, we conducted in vitro infection assays to determine the percentage of infected cells after incubation with wild-type or TcMCU-KO trypomastigotes. Figures 3H and I show that there was no significant difference between infections with wild-type and TcMCU-KO trypomastigotes, and Fig. 3J shows that replication of amastigotes was not significantly affected by TcMCU knockout, demonstrating that TcMCU is not essential for differentiation, infectivity, or intracellular replication. PCR analysis of the TcMCU-KO trypomastigotes obtained from these tissue cultures confirmed the absence of TcMCU in these Differences were not significant. Values in panels A to D and F to J are means Ϯ SD (n ϭ 3). *, P Ͻ 0.05; **, P Ͻ 0.01; ***, P Ͻ 0.001; n.s., not significant. cells (Fig. S3C). In contrast, induction of metacyclogenesis in TcMCUb-OE cells was similar to that of wild-type cells but greatly diminished in TcMCUb-KO cells (Fig. 3G). Several attempts (n ϭ 4) to infect tissue culture cells with TcMCUb-KO differentiated cells failed to recover significant amounts of trypomastigotes, suggesting that TcMCUb is important for infectivity.
Complementation of the KO mutants. To exclude the possibility that the TcMCU silencing phenotype is due to off-target effects, we investigated whether an exogenous TcMCU gene could complement the ablation of endogenous TcMCU. First, we demonstrated that the exogenous TcMCU (with changes in the PAM sequence to prevent disruption by CRISPR/Cas9), bearing a hemagglutinin (HA) epitope tag (TcMCU -PAM -HA), was targeted to the epimastigote mitochondrion. Immunofluorescence assays in Fig. S4A show that exogenous TcMCU-HA colocalizes with MitoTracker in the mitochondria. Then, we observed that mitochondrial Ca 2ϩ transport in permeabilized TcMCU-KO epimastigotes expressing exogenous TcMCU-HA was significantly higher than that of wild-type cells ( Fig. 4A and B). The expression of TcMCU-HA was confirmed by Western blot analysis (Fig. 4C). Then, we introduced mutations in highly conserved residues to evaluate their ability to restore Ca 2ϩ transport in TcMCU-KO cells (Fig. S4B). Complementation with a TcMCU R214W,D219V mutant, but not with the TcMCU D223N,E226Q mutant of the DIME region of TcMCU, was able to restore mitochondrial Ca 2ϩ uptake ( Fig. 4D and E), indicating the importance of Asp 223 and Glu 226 of the DIME domain for Ca 2ϩ transport. TcMCU-KO epimastigotes overexpressing TcMCUb failed to rescue Ca 2ϩ uptake, suggesting that TcMCUb cannot replace TcMCU. Western blot analysis confirmed the expression of these proteins (Fig. 4F). We also performed the complementation of TcMCU-KO cells with human MCU (Homo sapiens MCU [HsMCU]), and we found that it was not effective in restoring mitochondrial Ca 2ϩ transport ( Fig. 4G and H), although the protein colocalized with MitoTracker in the mitochondria of epimastigotes (Fig. S4C) and Western blot analysis confirmed its expression (Fig. 4I).
Reconstitution of the mitochondrial calcium uniporter in Saccharomyces cerevisiae. Saccharomyces cerevisiae does not possess an MCU, which makes it an ideal in vivo reconstitution system for studying the uniporter in a physiologically relevant organellar membrane (5,19). Because the minimal components sufficient for in vivo MCU activity in trypanosomatids are unknown, we transformed S. cerevisiae with TcMCU to assess whether this protein is able to reconstitute yeast mitochondrial calcium uniporter activity on its own. We transformed yeast with MCU from Dictyostelium discoideum (DdMCU) to use as a positive control, as this protein was reported to be sufficient alone to reconstitute mitochondrial calcium uniporter activity in yeast (19). To ensure expression of TcMCU in yeast mitochondria, we created a chimeric gene, consisting of the mitochondrial targeting sequence (MTS) of yeast cytochrome oxidase 4 (ScCox4p) fused to TcMCU lacking its predicted MTS ( ScMTS TcMCU) to obtain the expression of TcMCU in the yeast mitochondrial inner membrane. Extramitochondrial calcium levels were monitored with calcium green-5N, and we observed that mitochondria of yeast spheroplasts were able to take up Ca 2ϩ in the presence of mitochondrial substrates (5 mM succinate, 5 mM malate, 5 mM pyruvate, 5 mM ␣-ketoglutarate, and 1 mM glutamate) only when complemented with DdMCU ( Fig. 4J and K). However, neither TcMCU-nor ScMTS TcMCU-transformed cells were able to take up Ca 2ϩ under the experimental conditions tested, although the cells were able to take up Ca 2ϩ after addition of the membrane-potential-dependent calcium ionophore ETH-129 (19) (Fig. 4J). The enrichment of these proteins in the mitochondrial fraction was confirmed by Western blot analysis ( Fig. 4L and M). Our results suggest that TcMCU alone is unable to reconstitute mitochondrial Ca 2ϩ uptake in yeast.
Respiration of TcMCU-KO and TcMCUb-KO mutants. We measured oxygen uptake (   though OCRs in the presence of ADP, oligomycin, and FCCP were significantly higher in TcMCU-KO and significantly lower in TcMCUb-KO mitochondria (Fig. 5C to E). Respiratory control rates (state 3/state 4) were 1.93 Ϯ 0.05, 1.79 Ϯ 0.09, and 1.92 Ϯ 0.07 for WT, TcMCU-KO, and TcMCUb-KO cells, respectively (n ϭ 4). The lower OCR of TcMCUb-KO mitochondria correlated with lower staining of the mitochondria with MitoTracker Deep Red FM, a dye currently used to estimate mitochondrial mass (20) (Fig. 5F and G), as well as with lower activity of citrate synthase, another mitochondrial mass indicator (21) (Fig. 5H), suggesting a lower mitochondrial mass in these cells.
Autophagy in TcMCUb-KO cells. Autophagy has been reported as a starvation response in different trypanosomatids (22,23). In addition, block of mitochondrial Ca 2ϩ uptake was shown to increase autophagy as a survival mechanism (24,25). Huang et al. (10) recently showed that knockdown of MCU in T. brucei procyclic forms increased the expression of autophagy markers under starvation condition. We evaluated autophagy in these cells by immunofluorescence microscopy (Fig. 6A) using antibodies against the autophagy marker Atg8.1, which is the ortholog of LC3-II in mammalian cells (26). Then, we quantified the number of autophagosomes per cell (Fig. 6B) and the number of cells containing autophagosomes (Fig. 6C). TcMCUb-KO cells had significantly increased numbers of autophagosomes per cell when incubated under starvation conditions, while lower levels were found in TcMCU-KO cells than in wild-type cells (Fig. 6A to C). Moreover, TcMCUb-KO cells showed an increased number of autophagosomes/cell and a higher percentage of cells with autophagosomes even when cultured in rich medium. It has been proposed that in vertebrate cells a lower mitochondrial Ca 2ϩ transport capacity leads to an increase of the AMP/ATP ratio and stimulation of the AMPdependent protein kinase (AMPK) signaling axis that stimulates autophagy (24). However, knockout of both TcMCU and TcMCUb abolished mitochondrial Ca 2ϩ transport while only TcMCUb-KO cells showed increased autophagy. In agreement with these results, there was no correlation between the AMP/ATP ratio of these cells (Fig. 6D) and autophagy ( Fig. 6A to C), ruling out this potential mechanism of autophagy stimulation in T. cruzi.

DISCUSSION
Our studies have shown that the mitochondrial calcium uniporter (TcMCU) is not essential for T. cruzi epimastigote growth in rich medium, for differentiation (metacyclogenesis) under well-established starvation conditions, and for trypomastigote host cell infection and intracellular replication, although it is needed for optimal exponential growth of epimastigotes. No mitochondrial Ca 2ϩ transport was detected in TcMCU-KO epimastigotes, while the mitochondrial membrane potential (Δ⌿ m ) was unaffected in these cells. In contrast to these results, TcMCUb was required for optimal growth in rich medium, for metacyclogenesis under starvation conditions, and for trypomastigote host cell infection. However, as occurs with TcMCU-KO cells, mitochondria of TcMCUb-KO epimastigotes were unable to transport Ca 2ϩ while their Δ⌿ m was not altered. Additional differences between these KO cells were also found. TcMCUb-KO cells had lower respiratory activity, a higher AMP/ATP ratio in rich medium, reduced mitochondrial mass, and increased autophagy, suggesting that MCUb has additional bioenergetic roles independent of Ca 2ϩ transport. Overexpression of TcMCUb or TcMCU resulted in higher rates of Ca 2ϩ transport, no effects on Δ⌿ m , and no dominant negative effect of TcMCUb on mitochondrial Ca 2ϩ transport, in contrast to previous results described for HeLa cells (13). The two proteins interact, as shown by their coimmunoprecipitation.  been previously reported that epimastigotes use glucose from the medium during their exponential phase of growth and that once glucose is depleted and cells enter the stationary phase, they use tricarboxylic acid (TCA) cycle intermediates (29). All the main stages of T. cruzi have a high endogenous rate of respiration, suggesting that the organisms contain an energy reserve. Early reports ruled out the presence of a carbohydrate polymer (1,30). Oliveira et al. (31) first suggested the possibility that triglycerides could be the energy reserve, which would be similar to the situation in another stercorarian trypanosome, Trypanosoma lewisi, which also has a high endogenous rate of respiration and high triglyceride content (32). However, we found that lipid droplets increase, not decrease, in TcMCU-KO cells. As the presence of lipid droplets was evaluated at the beginning of the stationary phase, one possibility is that they could allow TcMCU-KO cells to survive longer because they were not consumed yet at that point. On the other hand, it is known that epimastigotes have a large pool of free amino acids that they use to maintain their osmolarity (33) and that protein degradation occurs under starvation conditions (1). Increased amino acid oxidation through the activity of the TCA cycle would bypass the need for the Ca 2ϩ -regulated formation of acetyl coenzyme A (CoA) from pyruvate (10). The higher ATP production via this pathway would explain the better survival of the cells when grown in lowglucose medium. The intracellular amastigotes appear to use fatty acids as an energy source, taking into account the increased levels of fatty acid oxidation enzymes in these stages (34). Amastigotes could then also bypass the need for the Ca 2ϩ -regulated pyruvate dehydrogenase to obtain energy through the TCA cycle. The predominance of fatty acid oxidation over glycolysis in the intracellular stages would explain the nonessential role of TcMCU for infection and replication.
Metacyclogenesis is usually stimulated under glucose-depleted conditions, and we detected a higher rate in TcMCU-KO cells. This higher rate of differentiation could be related to the lower ability of these cells to use endogenous substrates in low-glucose medium in contrast to TcMCU-OE cells. The increased metacyclogenesis observed in TcMCU-KO cells could also explain its longer survival during the stationary phase in low-glucose medium. On the other hand, the reduced metacyclogenesis of TcMCUb-KO cells correlated well with their inefficient ability to invade host cells.
To our knowledge, we report for the first time the complementation of a gene knockout (TcMCU-KO) obtained by CRISPR/Cas9 in T. cruzi. Complementation of TcMCU-KO cells with an exogenous copy of TcMCU but not with the gene mutated in the nucleotides encoding amino acids Asp 223 and Glu 226 of the DIME domain was able to restore mitochondrial Ca 2ϩ transport, confirming that as demonstrated in mammals, these two conserved acidic residues of the DIME motif are also key for TcMCU-mediated Ca 2ϩ uptake (8,13). However, a mutant including a substitution at the acidic residue Asp 219 (TcMCU R214W/D219V ; Glu 257 in HsMCU), which is highly conserved within MCU orthologs and has been proposed as a critical residue for Ca 2ϩ transport (8,13), rescued Ca 2ϩ transport in TcMCU-KO epimastigotes. Similarly, a recent report showed that the HsMCU E257A mutant was able to rescue Ca 2ϩ transport in HEK-293T MCU-KO cells, which is consistent with the predicted nuclear magnetic resonance (NMR)/electron microscopy (EM) structure of the MCU that exposes this residue outside the channel entrance (35).
Moreover, human MCU was unable to restore mitochondrial Ca 2ϩ transport in TcMCU-KO cells. TcMCU alone, in contrast to DdMCU (19), could not reconstitute mitochondrial Ca 2ϩ transport in yeast mitochondria. As T. cruzi lacks orthologs of EMRE (36) or MCUR1 (37,38), this result suggests that, as in the case of human MCU (19), other, still-unknown subunits of the complex might be necessary for full reconstitution of mitochondrial Ca 2ϩ transport in yeast by TcMCU.
An increase in autophagy was observed in TcMCUb-KO but not in TcMCU-KO cells when incubated under either rich or starvation conditions. In vertebrate cells, the lack of MCU-mediated Ca 2ϩ uptake promotes autophagy (37,38). An increase in the AMP/ATP ratio produced by downregulation of the TCA cycle by inhibition of the MCU could stimulate the phosphorylation of the AMP-dependent kinase (AMPK) and promote autophagy (24). However, we did not find such a correlation in T. cruzi, as TcMCU-KO cells have higher AMP/ATP ratios under starvation conditions and no difference in autophagy while TcMCUb-KO cells do not have differences in AMP/ATP ratios but exhibit higher autophagy than wild-type cells. These results suggest that this pathway might not be operative in T. cruzi and are in agreement with a reported AMPK-independent autophagy pathway reported recently in T. brucei (39).
It is possible that phenotypic changes in the TcMCUb-KO mutants may have been caused indirectly by adaptation to the loss of the gene that were not seen in the TcMCU-KO mutants because of the occurrence of more beneficial compensatory mutations in these cells. However, our attempts to complement TcMCUb-KO mutants with an exogenous TcMCUb gene devoid of the PAM motif did not result in selected parasites, probably due to the fragility of these cells and their inability to resist the electroporation stress (data not shown).
The mitochondria of TbMCU-OE epimastigotes have an increased ability for Ca 2ϩ uptake, which results in Ca 2ϩ overload and oxidative stress, and similar results were observed in TcMCUb-OE cells. These results agree with those obtained in HeLa cells (7) and T. brucei (10). Mitochondrial Ca 2ϩ overload has also been proposed as the link between complement deposition and triggering of cell death in T. cruzi epimastigotes (40). Accumulation of Ca 2ϩ in the mitochondrion leads to a decrease in cell respiration, dissipation of the inner membrane potential, and increased reactive oxygen species production (40).
In contrast with the results obtained in HeLa cells (13), where overexpression of MCUb has a dominant negative effect on mitochondrial Ca 2ϩ transport, overexpression of TcMCUb led to increased Ca 2ϩ uptake without affecting the mitochondrial membrane potential. Knockout of TcMCUb abolished mitochondrial Ca 2ϩ transport, and the cells had a lower growth rate than control cells. Moreover, as mentioned above, we introduced in TcMCU amino acid substitutions (R214W/D219V) that are conserved in its paralog TcMCUb and considered critical to determine the dominant negative effect of the mammalian protein (13). However, we found that TcMCU R214W/D219V , but not TcMCUb, was able to restore mitochondrial Ca 2ϩ transport in TcMCU-KO cells, suggesting that these substitutions are not enough to abolish Ca 2ϩ transport.
In summary, our results indicate that TcMCU is not essential for growth, differentiation, infectivity, and intracellular replication of T. cruzi, although it could be important under stress situations, such as for growth in low-glucose medium. This could be relevant for epimastigotes grown in the insect's gut, in which glucose is supposed to be scarce (41). In contrast to what has been described in some mammalian cells, where MCUb acts as a dominant negative subunit of the MCU complex (13), mitochondrial Ca 2ϩ transport is enhanced in TcMCUb-OE epimastigotes and abolished in TcMCUb-KO cells. TcMCUb-KO cells have additional phenotypic alterations that suggest other bioenergetic roles besides mitochondrial Ca 2ϩ transport.

MATERIALS AND METHODS
Culture methods. T. cruzi epimastigotes (Y strain) were grown at 28°C in liver infusion tryptose (LIT) medium (42)  . To obtain a low-glucose LIT medium, glucose was not added. We determined the growth rate of T. cruzi epimastigotes by counting cells in a Neubauer chamber with a starting culture of 1 ϫ 10 6 epimastigotes. Tissue culture cell-derived trypomastigotes were obtained from Vero cells infected with metacyclic trypomastigotes obtained as described below. T. cruzi trypomastigote forms were collected from the culture medium of infected host cells, using a modification of the method of Schmatz and Murray (43) as described previously (44). Vero cells were grown in RPMI supplemented with 10% fetal bovine serum and maintained at 37°C with 5% CO 2 .
Metacyclogenesis. We followed the protocol described by Bourguignon et al. (45) with minor modifications. Epimastigotes were obtained after 4 days in LIT medium and submitted to a stress (incubation for 2 h in a medium containing 190 mM NaCl, 17 mM KCl, 2 mM MgCl 2 , 2 mM CaCl 2 , 0.035% sodium bicarbonate, 8 mM phosphate, pH 6.9, at room temperature; triatome artificial urine [TAU] medium). After this stress, parasites were incubated for 96 h in TAU 3AAG medium (which consists of the above-described TAU medium supplemented with 10 mM L-proline, 50 mM sodium L-glutamate, 2 mM sodium L-aspartate, and 10 mM glucose). To increase the number of metacyclic forms to infect Vero cells, the contents of the flask were collected and resuspended in medium containing fresh fetal bovine serum and incubated at 37°C for 20 h. The complement in the FBS kills epimastigotes while metacyclic trypomastigotes survive. Samples were harvested from the TAU 3AAG plus FBS-containing medium at days 5 and 10 of cultivation.
In vitro infection assay. Gamma-irradiated (2,000-rad) Vero cells (4.5 ϫ 10 5 cells) were plated onto sterile coverslips in a 12-well plate and incubated overnight at 35°C, 7% CO 2 , in RPMI medium plus 10% fresh fetal bovine serum. Tissue culture-derived trypomastigote collections were incubated at 4°C overnight to allow amastigotes to settle from swimming trypomastigotes. Trypomastigotes from the supernatants of these collections were counted and used to infect the coverslips at a ratio of 50 parasites to 1 host cell. At 4 h postinfection, coverslips were washed extensively with Dulbecco's Hanks' solution, followed by phosphate-buffered saline (PBS), pH 7.4, to remove any extracellular parasites. Coverslips were fixed immediately in 4% paraformaldehyde in PBS, pH 7.4, at 4°C for 30 min. Coverslips were washed once with PBS and mounted onto glass slides in Fluoromount G containing 15 g/ml of 2-(4-aminophenyl)-1H-indole-6-carboxamidine (DAPI), which stains host and parasite DNA. Coverslips were viewed on an Olympus BX60 microscope to quantify the number of host cells that contained intracellular parasites and the number of intracellular parasites per cell in randomly selected fields. Three hundred host cells were counted per sample in three independent experiments. To quantify amastigote replication, the following modifications were used: host cells were infected at a ratio of 10 parasites to 1 host cell, and coverslips were allowed to incubate for 48 h postinfection at 35°C, 7% CO 2 , prior to fixation and DAPI staining.
TcMCU-KO. Chimera single guide RNA (sgRNA) sequences to target the TcMCU gene (TryTripDB identifier [ID] TcCLB.503893.120) were PCR amplified from plasmid pUC_sgRNA, containing the sgRNA backbone sequence (82 bp) (46). One specific protospacer was included in the forward primer (primer 1 [see Table S1 in the supplemental material]) while using a common reverse primer (primer 2 [ Table S1]) for sgRNA amplification. These primers also contained a BamHI restriction site for cloning into Cas9/ pTREX-n (14) upstream of the HX1 transsplicing signal (47) to generate the TcMCU-sgRNA/Cas9/pTREX-n construct. The sgRNA orientation was verified by PCR using the specific TcMCU-sgRNA forward primer and the HX1 reverse primer (14), which annealed on the HX1 region of the vector (primer 3 [ Table S1]). Positive clones that generate a 190-bp PCR fragment were also sequenced. A scrambled sgRNA (Sc-sgRNA/ Cas9/pTREX-n), obtained with primers 2 and 4 (Table S1), was used as a control. The Cas9/pTREX-n vector contains the Streptococcus pyogenes Cas9 sequence with a twice-repeated simian virus 40 (SV40) nuclear localization signal (2ϫNLS) and green fluorescent protein (GFP) (14), thus generating mutant cell lines with green fluorescent nuclei after transfection. The molecular construct (DNA donor) to promote homologous directed repair and ablation of the target gene after the double-strand break (DSB) induced by Cas9 was obtained using a recombinant PCR strategy. Briefly, the 5= and 3= UTRs of the TcMCU gene (490 and 529 bp, respectively) were amplified by PCR using T. cruzi genomic DNA as the template (primers 7 to 10 [ Table S1]). The open reading frame (ORF) of the blasticidin S-deaminase gene (Bsd) was also amplified by PCR from pTREX-b vector as the template using primers containing 20 nucleotides (nt) overlapping with the 5= and 3= flanking sequences of the TcMCU gene (primers 5 and 6 [ Table S1]). The three PCR fragments (5=-TcMCU flanking sequence, blasticidin-resistant gene, and 3=-TcMCU flanking sequence) were linked together by sequential PCR using primers 7 and 10 ( Table S1). The final PCR product was cloned into pGEM-T Easy vector (Promega). After sequencing verification of positive clones, circular constructs TcMCU-sgRNA/Cas9/pTREX-n and the Bsd cassette in pGEM-T Easy were used to cotransfect T. cruzi epimastigotes. After 5 weeks of selection with 250 g/ml G418 and 10 g/ml blasticidin, TcMCU gene ablation was verified by PCR using primers 11 and 12 (Table S1). Alternatively, to confirm TcMCU gene KO in trypomastigotes obtained from infected cells, we amplified the entire TcMCU ORF (909 bp) by PCR (primers 15 and 19 [ Table S1]).
TcMCUb-KO. An sgRNA to target the sequence coding for the hypothetical TcMCUb protein (TryTripDB ID TcCLB.416883.9) was amplified by PCR (primers 31 and 2 [ Table S1]). Following the same strategy mentioned above for TcMCU, we obtained the TcMCUb-sgRNA/Cas9/pTREX-n construct. The DNA donor cassette designed to promote homologous directed repair and disruption of TcMCUb was obtained by PCR using a set of long primers (ultramers) containing 120 nucleotides, from which 100 nucleotides correspond to the region from nt Ϫ20 to ϩ80 (forward ultramer) and nt ϩ510 to ϩ610 (reverse ultramer) of the TcMCUb ORF. Each one of the ultramers had 20 nt annealing on the Bsd gene (primers 32 and 33 [ Table S1]). Gene disruption of TcMCUb was verified by PCR using primers 34 and 35 (Table S1).

Generation of TcMCU/TcMCUb-overexpressing cell lines in T. cruzi.
For gain-of-function analysis of TcMCU/TcMCUb, T. cruzi epimastigotes were transfected with the trypanosome expression vector pTREX-n (47,48) containing the corresponding genes. The molecular construct for TcMCU overexpression was made as follows: the full sequence of TcMCU was PCR amplified using primers 15 and 16 (Table S1) and T. cruzi gDNA as the template. The purified fragment of 909 bp was digested with EcoRI and HindIII restriction enzymes and cloned into the pTREX-n vector linearized with the same enzymes. The molecular construct for TcMCUb overexpression was made as follows: the full sequence of TcMCUb was PCR amplified using primers 38 and 39 (Table S1) and T. cruzi gDNA as the template. For this purpose, we used the genomic sequence (KB205522 from nucleotide 139,788 to 140,579, TryTripDB) from T. cruzi strain Esmeraldo to design the reverse primer, because the MCUb gene 3= end and 3= UTRs are missing in the TcCLB.416883.9 sequence. The purified fragment of 792 bp was digested with XbaI and SalI restriction enzymes to be cloned into a version of the pTREX-n vector that includes a 3ϫHA epitope tag coding sequence. The pTREX-n vector was linearized with XbaI and XhoI enzymes, and the TcMCUb gene was cloned in frame with the 3ϫHA sequence for C-terminal tagging, allowing the detection of the overexpressed protein with anti-HA antibody. For this purpose, we previously cloned the 3ϫHA tag sequence, excised from pMOTag-4H vector (49) with XhoI and SalI enzymes, into pTREX-n by using the XhoI restriction site.
Molecular constructs to complement TcMCU-KO mutants. To revert the phenotype exhibited by TcMCU-knockout epimastigotes, we obtained the TcMCU -PAM -HA construct to reconstitute the ablated TcMCU gene. Following a two-step-PCR strategy, we eliminated the PAM sequence (CGG to CCT) specific for the TcMCU-sgRNA used to obtain the TcMCU-KO cells, therefore avoiding constitutively expressed Cas9 to target the inserted sequence (Table S1, primers 15 and 17 to 19). The fragment obtained by overlap extension PCR was digested with EcoRI and HindIII and then cloned into pTREX-h (which confers hygromycin resistance) digested with the same enzymes. We also included a C-terminal HA tag in order to detect the overexpressed protein using anti-HA antibody.
Furthermore, we attempted to reconstitute the ablated TcMCU gene in TcMCU-KO mutants with the following. For TcMCUb, primers 38 and 39 (see Table S1 in the supplemental material) were used to amplify from T. cruzi gDNA the TcMCUb gene that was digested with XbaI and SalI enzymes and cloned in frame with the 3ϫc-Myc tag sequence into pTREX-h digested with XbaI and XhoI. For TcMCU D223N,E226Q and TcMCU R14W,D219N , the generation of pTREX-h-TcMCU D223N,E226Q -3ϫc-Myc was done by overlap extension PCR using the TcMCU -PAM -HA plasmid as a template and primers 20 to 23 (Table S1). Following a similar strategy, the pTREX-h-TcMCU R14W,D219N -3ϫc-Myc was obtained using primers 20, 21, 24, and 25 (Table S1). For the human MCU gene, we purchased from GenScript (USA) a cDNA clone corresponding to the Homo sapiens MCU variant 1 (GenBank accession number NM_138357) of 1,100 bp. The HsMCU fragment amplified with primers 26 and 27 (Table S1) was digested and cloned into pTREX-p (which confers puromycin resistance) by EcoRI and XhoI restriction sites in frame with the 3ϫc-Myc tag sequence.
For this purpose, we first generated pTREX-h and pTREX-p vectors that were not available (Table S1, primers 40 to 48). As templates to amplify GAPDH UTRs and hygromycin and puromycin genes, we used pTREX-n, pMOTag-4H (49), and pMOTag-23M (49) plasmids, respectively. The 3ϫc-Myc tag was excised from the pMOTag-23M vector (49) with XhoI and SalI enzymes and subsequently cloned into pTREX-h and pTREX-p by the XhoI restriction site.

Statistical analysis.
Values indicated in the figure legends are means Ϯ standard deviations (SD) of n biological experiments (n is indicated in the figure legends), and the significance of their differences was evaluated by the Student t test.