Mitochondrial Pyruvate Carrier Subunits Are Essential for Pyruvate-Driven Respiration, Infectivity, and Intracellular Replication of Trypanosoma cruzi

Trypanosoma cruzi is the causative agent of Chagas disease. Pyruvate is the end product of glycolysis, and its transport into the mitochondrion is mediated by the mitochondrial pyruvate carrier (MPC) subunits.

In the mitochondrial matrix, the pyruvate dehydrogenase (PDH) complex catalyzes the oxidative decarboxylation of pyruvate into acetyl coenzyme A (acetyl-CoA), which is a major substrate for the tricarboxylic acid (TCA) cycle. Therefore, the availability of pyruvate in mitochondria, determined by MPC, links the glycolytic and oxidative phosphorylation pathways and is important for cellular homeostasis.
MPC has been characterized in several organisms (6)(7)(8)(9)(10). MPC is generally composed of two main types of small homologous proteins, MPC1 and MPC2. In Saccharomyces cerevisiae, there is a third member of the MPC family, MPC3, which shows high identity with MPC2 in amino acid sequence. Therefore, yeast MPC is constituted by the protomers MPC1 and either MPC2 or MPC3 (6,7). The MPC1/MPC2 heterocomplex is the active MPC under fermentative conditions, and the MPC1/MPC3 heterocomplex functions under respiratory conditions (11,12). MPC1 and MPC2 proteins have been reported to migrate as a 150-kDa heterocomplex in blue native polyacrylamide gel electrophoresis (6), but the oligomeric state of the functional MPC has been disputed (13). Yeast MPC has been described as an active heterodimer (MPC1/MPC3 dimer), whereas individual subunits can form homodimers but they are not able to transport pyruvate (13). On the other hand, human MPC2 protomers were able to form high-order MPC2 homocomplexes, which promoted effective pyruvate transport in proteoliposomes independently of MPC1 (14). A posttranslational modification that changes MPC activity has been reported. The MPC1 subunit is deacetylated at lysine residues by Sirt3, and consequently, MPC activity increases (15).
In Trypanosoma brucei, the causative agent of human African trypanosomiasis or sleeping sickness, TbMPC1 and TbMPC2 subunits have been described as essential proteins for mitochondrial pyruvate transport. Both proteins localize to the mitochondrion of procyclic forms (PCF). Knockout of TbMPC1 and TbMPC2 genes generated a 2-fold decrease in pyruvate uptake into isolated mitochondria, while the inhibitor UK5099 [2cyano-3-(1-phenyl-1H-indol-3-yl)-2-propenoic acid] reduced mitochondrial pyruvate import in wild-type PCF (9). UK5099 acts on MPC and plasma membrane monocarboxylate transporters (MCTs) (4) but exerts a specific inhibitory effect on MPC in isolated mitochondria or digitonin-permeabilized cells. In bloodstream forms (BSF) of T. brucei, ATP is mainly generated through glycolysis, and most pyruvate is excreted (16) through the pyruvate-proton symporter to maintain the intracellular pH (17). However, PDH activity is present in BSF trypanosomes (18,19) that also possess an MPC (9). TbMPC1 knockout (KO) BSF trypanosomes displayed a reduction in growth rate (in Creek's minimal medium) and in parasitemia and lethality in mice. It was not possible to knockout TbMPC2 in BSF trypanosomes (9).
In the present study, we evaluated the importance of T. cruzi MPC (TcMPC) in mitochondrial bioenergetics, growth, differentiation, infectivity, and intracellular replication by individual ablation of TcMPC1 and TcMPC2 genes.
Tagging of both alleles of TcMPC1 or TcMPC2 after 6 weeks of double selection with G418 and puromycin was confirmed by PCR using two specific primer sets for each gene (primers 9 to 16) (see Table S1) designed to discriminate between the intact TcMPC locus and the endogenous tagged gene. For TcMPC1, PCR analysis using genomic DNA (gDNA) and primers 9 and 10 (Table S1, primer set 1) showed the amplification of a 0.45-kb band in control cell lines (wild-type and transfected with scrambled single guide RNA [sgRNA] [scrambled]) and a 1.7-kb band in the TcMPC1-TAG cell line, as expected (see Fig. S2A and B). PCR analysis using primers 11 and 12 (Table S1, primer set 2) also confirmed the endogenous tagging of TcMPC1 ( Fig. S2C and D). For TcMPC2, PCR analyses showed expected bands of approximately 0.5-kb in control cell lines and 1.9-kb in the TcMPC2-TAG cell line ( Fig. S2E and F) using primers 13 and 14 (Table S1, primer set 1), and an expected band of approximately 1.6-kb in TcMPC2-TAG cell line using primers 15 and 16 (Table S1, primer set 2) ( Fig. S2G and H).
Western blot analysis confirmed the expression of the proteins TcMPC1-3Âc-Myc and TcMPC2-3Âc-Myc in their respective cell lines with molecular masses close to their expected values (TcMPC1-3Âc-Myc, 17 kDa; TcMPC2-3Âc-Myc, 21 kDa) (Fig. 1C). Localization of both proteins to the mitochondrion was confirmed by fluorescence microscopy. Figure 1D and E show colocalization of the green fluorescent signal from the expressed proteins using a monoclonal anti-c-Myc antibody with the mitochondrial marker MitoTracker in red. Thus, our results confirmed that TcMPC1 and TcMPC2 proteins localize to the mitochondria of T. cruzi epimastigotes.
CRISPR/Cas9-induced knockout of TcMPCs. To investigate the role of the pyruvate carrier, MPC knockout cell lines were generated using CRISPR/Cas9 gene editing (28) (Fig. 1F and G). DNA donors contained blasticidin resistance marker (Bsd; gene encoding blasticidin S deaminase, 399 bp) and two 100-bp homology arms corresponding to the TcMPC1 or TcMPC2 flanking regions. After 6 weeks of selection with G418 and blasticidin, knockouts of TcMPC1 and TcMPC2 were validated by PCR and Southern blot analyses.
Three primer sets were designed to carry out PCR analysis using gDNA from doubly resistant parasites as the template. Using primers (primer set 1 [primers 23 and 24], Table S1) designed to discriminate between the intact TcMPC1 locus and the one replaced by Bsd, we amplified bands of approximately 0.75-kb with gDNA from control cell lines and 0.85-kb with gDNA from the TcMPC1-KO cell line, indicating the generation of a homogeneous knockout population in which both TcMPC1 alleles were ablated ( Fig. S3A and B). TcMPC1 knockout was also validated by PCR analysis using different primers (primer set 2 [primers 25 and 26] and primer set 3 [primers 27 and 28], Table S1) (Fig. S3C, D, E, and F). Southern blot analysis using a probe comprising the entire TcMPC1 (294 bp) also confirmed knockout of both TcMPC1 alleles, revealing a 2kb band only for HindIII-digested gDNA from scrambled control cells (Fig. 1H).
A similar strategy was followed to knockout TcMPC2. To validate the knockout of TcMPC2 by PCR, two primer sets were designed. With the first primer set (primer set 1  Table S1), there was an expected amplification of a 0.6-kb band from gDNA from TcMPC2-KO parasites and no amplification for wild-type and scrambled cell lines ( Fig. S3G and H). PCR analysis using another primer set (primer set 2 [primers 31 and 32], Table S1) resulted in amplification of a band with expected size of approximately 0.4 kb for control gDNAs ( Fig. S3I and J). Using a TcMPC2 probe (414 bp) for Southern blot analysis, TcMPC2 knockout was also successful in generating a band of 2.4 kb only for BamHI-digested gDNA from scrambled control cells (Fig. 1I). Further PCR assays were performed to confirm the specific ablation of each TcMPC gene (see Fig. S4).
Effect of TcMPC-KOs on epimastigote growth. To assess the role of MPC proteins in epimastigote growth, TcMPC1-KO and TcMPC2-KO epimastigotes were grown in liver infusion tryptose (LIT) medium in the absence or presence of additional glucose and compared to control parasites (transfected with scrambled sgRNA). There was no significant reduction in the growth rates of both knockout epimastigotes in either standard ( Fig. 2A) or low-glucose (Fig. 2B) LIT medium.
Effect of TcMPC-KOs on stimulation of mitochondrial oxygen consumption. To demonstrate that pyruvate could not be transported to the mitochondria in the knockout parasites, we measured oxygen consumption by mitochondria in situ of digitonin-permeabilized epimastigotes in the presence of pyruvate or succinate as the respiratory substrate. Oxygen consumption rates (OCRs) were evaluated under ADP-stimulated (state 3, respiration stimulated by oxidative phosphorylation [OXPHOS]), oligomycin-inhibited (state 4, minimal respiratory rate), and carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP)-stimulated (state 3u [uncoupled], maximal respiratory capacity) conditions and compared to those in parasites transfected with the scrambled sgRNA (control cells).
Control cells showed similar OCR (Fig. 2C) and oxygen uptake (Fig. 2F) tracings in the presence of either pyruvate (blue lines) or succinate (red lines) as the substrate. In contrast, in the presence of pyruvate, both TcMPC-KO mutants showed only basal OCRs upon additions of ADP, oligomycin, and FCCP ( Fig. 2D, E, and I) and diminished oxygen uptake ( Fig. 2G and H). However, TcMPC1-KO and TcMPC2-KO mutants showed comparable OCRs to those of the control cells upon additions of ADP, oligomycin, and FCCP when succinate was used (Fig. 2D, E, and J). These results suggest that both TcMPC1 and TcMPC2 are important for mitochondrial pyruvate import.
To rule out any influence of the potential presence of endogenous substrates in the preparations, we also measured oxygen consumption in the absence of exogenous substrates. Under these conditions, there was no OCR stimulation by ADP, and the basal OCRs were maintained in control and knockout cells (see Fig. S5). Addition of the mitochondrial complex III inhibitor antimycin A that results in an interruption of the electron flux from cytochrome b to cytochrome c 1 decreased OCR to almost zero (Fig. S5).
Pyruvate is commonly used as an electron donor for complex I of the respiratory chain through the formation of NADH catalyzed by PDH. However, complex I appears to be functionally limited in T. cruzi (29), although pyruvate was able to stimulate oxygen consumption in permeabilized cells as well as succinate (Fig. 2C). To investigate whether this stimulation was due to the previous conversion of pyruvate into succinate, we measured oxygen consumption driven by pyruvate in the absence or presence of malonate. Malonate is a competitive inhibitor of succinate dehydrogenase, which is a component of complex II localized at the matrix face of the inner mitochondrial membrane. Succinate dehydrogenase catalyzes the conversion of succinate into fumarate, and the resultant reduced flavin adenine dinucleotide (FADH 2 ) feeds electrons into complex II. OCR analysis of digitonin-permeabilized control cells (wild-type Representative OCR (C to E) and oxygen concentration (F to H) tracings of digitonin-permeabilized scrambled (C and F), TcMPC1-KO (D and G), and TcMPC2-KO (E and H) epimastigotes when pyruvate (blue lines) or succinate (red lines) was used as the mitochondrial substrate. Additions of 100 mM ADP (respiration stimulated by oxidative phosphorylation [OXPHOS]), 1 mg ml 21 oligomycin (Oligo, minimal respiratory rate), and 1 mM FCCP (maximal respiratory capacity) are indicated by arrows. Normalized OCRs upon addition of ADP, oligomycin, and FCCP when pyruvate (I) and succinate (J) were used as the substrate. Normalized OCR of 1 means that OCR upon an addition is equal to the basal OCR (prior to ADP addition). Values are means 6 SEMs (n = 6). *, P , 0.05; ****, P , 0.0001; n.s., no significant differences. Two-way ANOVA test with multiple comparisons. and cells transfected with scrambled sgRNA) showed a complete inhibition of pyruvate-driven oxygen consumption when malonate was added to the reaction medium (light blue lines and bars, Fig. 3), indicating that pyruvate needs to be converted into succinate in order to donate electrons to the respiratory chain.
Inhibition of TcMPC by UK5099 decreases pyruvate-driven respiration. We tested whether pharmacological inhibition of TcMPC by an inhibitor of MCTs and MPC, UK5099 ( Fig. 4A) (4), resulted in similar changes in pyruvate-driven respiration to those observed in the knockout epimastigotes. We tested different concentrations of the inhibitor (10, 25, and 50 mM) on permeabilized cells and found the best inhibition with 50 mM UK5099 (see Fig. S6). Addition of this concentration of UK5099 significantly decreased the OCRs of wild-type cells and cells transfected with scrambled sgRNA in the presence of ADP, oligomycin, and FCCP (light blue lines and bars in Fig. 4B, C, F, and G), while it had little effect on the TcMPC-KO mutants (Fig. 4D, E, H, and I).
Effects of TcMPC-KOs on a-ketoglutarate consumption. It has been reported that when the mammalian MPC is deficient, increased glutaminolysis can compensate for the deficit in pyruvate-derived carbon to fuel the TCA cycle (30)(31)(32). Glutamine is converted into glutamate by a glutaminase, and then glutamate is converted into a-ketoglutarate, a TCA cycle metabolite. However, a glutaminase gene orthologue to the human enzyme is apparently missing from the T. cruzi genome (33), and other sources of glutamate might be available. We therefore tested whether a-ketoglutarate synthesis and utilization were altered in the MPC-KO mutant cells. We first measured changes in cellular respiration in digitonin-permeabilized control, TcMPC1-KO, and TcMPC2-KO cells upon addition of ADP, oligomycin, and FCCP in the presence of a-ketoglutarate as the substrate ( Fig. 5A and B). Both TcMPC-KO mutant epimastigotes showed increased respiration stimulated by OXPHOS and maximal respiratory capacity compared to those of control cells ( Fig. 5A and B). In agreement with these results, a-ketoglutarate dehydrogenase (a-KGDH) activity in lysates from TcMPC-KO mutants was higher than that of control cells (Fig. 5C).
Role of Mitochondrial Pyruvate Carrier in T. cruzi ® dephosphorylation by a Ca 21 -stimulated PDP (20). It has also been reported that decreased expression of MPC2 in Alzheimer's disease-related models led to decreased Ca 21 and pyruvate uptake into mitochondria and reduced Ca 21 -mediated stimulation of the TCA cycle, although the reason for the decrease in Ca 21 import was unknown (34). We therefore tested the phosphorylation state of the PDH E1a subunit in lysates from knockout mutants by Western blotting using a commercial antibody against the PDH E1a subunit (phosphopeptide Ser-293) of human cells that recognizes the phosphorylated site 1 (Ser-284) of the T. cruzi enzyme (34). Our results indicate that T. cruzi PDH phosphorylation is higher in TcMPC2-KO epimastigotes than in control and TcMPC1-KO parasites ( Fig. 6A and B).
Using pyruvate as mitochondrial substrate, 2-and 1.5-fold decreases in mitochondrial Ca 21 uptake were detected in digitonin-permeabilized TcMPC1-KO and TcMPC2-KO epimastigotes, respectively, compared to that in the control cells ( Fig. 6C and D). However, when succinate was used as the substrate, Ca 21 uptake was not affected in either TcMPC1-KO or TcMPC2-KO mutants ( Fig. 6E and F). Therefore, defective mitochondrial import of pyruvate resulted in decreased Ca 21 uptake. We then used safranine O to evaluate mitochondrial membrane potential (Dc m ) in digitonin-permeabilized epimastigotes in the presence of pyruvate or succinate as the mitochondrial substrate. A decrease in fluorescence intensity after permeabilization with digitonin indicates stacking of the dye to the energized inner mitochondrial membrane. Fluorescence intensity in the steady state was decreased in both knockout epimastigotes in the presence of pyruvate ( Fig. 6G and H), but not succinate ( Fig. 6J and K), evidencing an alteration in the mitochondrial membrane potential of knockout parasites when pyruvate was used as the substrate. However, phosphorylation of ADP was not affected in either case ( Fig. 6I and L).
Effect of TcMPC-KOs on the infective stages. The ability of TcMPC-KO epimastigotes to differentiate in vitro into metacyclic trypomastigotes was also analyzed. Interestingly, the percentage of metacyclogenesis was approximately 3-fold higher in TcMPC1-KO parasites and 2-fold lower in TcMPC2-KO parasites than in parasites transfected with a scrambled sgRNA (Fig. 7A), suggesting other potential functions for these carriers. The infectivity of both TcMPC1-KO and TcMPC2-KO trypomastigotes was approximately 3-fold lower than that of scrambled control trypomastigotes (Fig. 7B). Moreover, the replication of intracellular amastigotes was also significantly impaired for both TcMPC-KO cell lines (Fig. 7C). Therefore, TcMPC subunits are important for differentiation, host cell invasion, and intracellular replication.
Overexpression of TcMPC proteins. We also tested whether individual overexpression of the TcMPC subunits with a 3Â hemagglutinin (HA) tag produced significant phenotypic changes. Both expressed proteins had molecular masses close to their expected values (TcMPC1-3ÂHA, 17 kDa; TcMPC2-3ÂHA, 21 kDa), as indicated by Western blotting (see Fig. S7A). TcMPC-OE epimastigotes did not grow at a higher rate compared to the control parasites (transfected with pTREX-n/3ÂHA empty vector [EV]) (Fig. S7B). Immunofluorescence analysis showed a mitochondrial localization for both proteins ( Fig. S7C and S7D). Slight changes were detected in the OCRs in oxygen consumption stimulated by OXPHOS, minimal respiratory rate, and maximal respiratory capacity using pyruvate as the mitochondrial substrate ( Fig. S7E and F).

DISCUSSION
Our studies indicate that both MPC1 and MPC2 localize to the mitochondrion of T. cruzi where they have an essential role in pyruvate stimulation of respiration, with previous conversion of pyruvate into succinate. Individual knockout of each subunit results in ablation of this activity, suggesting that the function of one subunit cannot be compensated by the presence of the other and therefore both proteins must be present. Epimastigotes deficient in any of the two subunits have no growth deficiencies in either standard or low-glucose LIT medium, but trypomastigotes are severely impeded in their ability to invade host cells and replicate intracellularly as amastigotes.
In T. cruzi epimastigotes, pyruvate can enter the mitochondrion to be converted into acetyl-CoA and feed the tricarboxylic acid cycle or can be converted into alanine in the cytosol by the reaction catalyzed by alanine aminotransferase or in the glycosome by the action of an L-alanine dehydrogenase, with reoxidation of NADH needed to balance its reduction by glyceraldehyde 3-phosphate dehydrogenase (35). Although T. cruzi epimastigotes have a high rate of glucose consumption, they are characterized by the production and excretion of partially reduced products of glucose metabolism, such as succinate, alanine, and acetate, instead of oxidizing glucose completely to CO 2 and water (35)(36)(37). This metabolism is known as "aerobic fermentation of glucose" and could be responsible for the lack of growth defects in the MPC1/MPC2 knockout mutants: instead of being used for the TCA cycle, pyruvate could contribute to the maintenance of the redox balance in the glycosomes or could be excreted as alanine, while the increased a-ketoglutarate consumption, revealed by the increased activity of a-ketoglutarate dehydrogenase and the increase in respiration by a-ketoglutarate, could replace the pyruvate needed for the TCA cycle. It is possible that cytosolic accumulation of pyruvate leads to its reaction with glutamate catalyzed by alanine aminotransferase, resulting in the formation of a-ketoglutarate and alanine. Glutamate and  Role of Mitochondrial Pyruvate Carrier in T. cruzi ® alanine are the most abundant amino acids present in epimastigotes (38). a-Ketoglutarate could then enter the mitochondrion through the a-ketoglutarate carrier that exchanges a-ketoglutarate for malate and that has been characterized in T. brucei (39, 40) (T. cruzi gene orthologue, TriTrypDB gene ID, TcCLB.509805.190). Cytosolic malate can then be converted into pyruvate catalyzed by the malic enzyme. The malic enzyme is stimulated by aspartate (41) and, as a result, amino acid production would stimulate a reaction for the efficient utilization of amino acids for energy. This combination of reactions, shown in Fig. 8, would produce a-ketoglutarate that enters the TCA cycle and alanine that could be excreted.
A different situation would occur with the trypomastigote and amastigote stages. Attachment to host cells and their invasion by trypomastigotes are active processes that require energy, and it has been shown that glycolysis (2-deoxyglucose) or respiratory chain (antimycin A) inhibitors prevent these activities (42), which perhaps could not be compensated for by increased a-ketoglutarate consumption. Lack of MPC1 or MPC2 appears to affect less the intracellular replication of amastigotes (approximately 1.5-fold reduced) than the mammalian host cell invasion by trypomastigotes (approximately 3-fold reduced). This is in agreement with our results with TcPDP (20). The predominance of fatty acid oxidation and amino acid oxidation (43, 44) over glycolysis has been proposed for these intracellular stages, although it has recently been reported that they are also able to use glucose (45). Amastigotes can also use glutamate from the host cells (33).
An important difference in the functionality of the respiratory chain between the mitochondrion of T. cruzi and the mitochondria of mammalian cells is the limited function of complex I in the parasite (29). Our results indicate that pyruvate needs to be converted into succinate in order to provide electrons to the respiratory chain and are in agreement with the limited role of complex I. It has been suggested that reducing equivalents enter the respiratory chain as succinate through the action of an NADH-dependent fumarate reductase (46,47).
We found that TcMPC1-KO and TcMPC2-KO epimastigotes have decreased Ca 21 uptake into the mitochondrion when pyruvate, but not when succinate, was used as the substrate. A decrease in MPC2 expression in Alzheimer's disease models is associated with less Ca 21 uptake (34), and studies in hepatocytes and embryonic fibroblasts have shown the Ca 21 import into mitochondria is also decreased following inhibition of MPC activity (48). In the latter case, the effect is mediated by increased expression of mitochondrial calcium uptake 1 (MICU1), which together with MICU2, is the gatekeeper of the mitochondrial calcium uniporter (MCU). However, in T. cruzi epimastigotes, overexpression of either MICU1 or MICU2 does not affect Ca 21 import into the mitochondrion (49), and the reason for this lower Ca 21 uptake could be the lower mitochondrial membrane potential when pyruvate is used as the substrate. Decreased mitochondrial Ca 21 uptake in TcMPC2-KO cells could explain the higher phosphorylation of the TcPDH E1a subunit due to decreased phosphatase activity of the Ca 21 -stimulated PDP. The effect of the KOs on mitochondrial respiration is then expected to be due to a diminished flux from pyruvate to the TCA cycle that is affected by both the lower mitochondrial uptake and the presence of inactive PDH.
Interestingly, the higher phosphorylation of the E1a subunit was only detected in TcMPC2-KO but not in TcMPC1-KO cells. Since these experiments reveal the phosphorylation state of PDH in vivo, they suggest that epimastigotes are using more pyruvate than succinate under physiological conditions, and when TcMPC2 is knocked out, there is higher phosphorylation of the TcPDH E1a subunit because pyruvate is not transported to the mitochondrial matrix. The lack of increased E1a subunit phosphorylation in TcMPC1-KO cells could be due to the selection of clones that depend more on succinate than on pyruvate as the mitochondrial substrate, and there is no difference in Ca 21 uptake when succinate is the substrate.
Our results are compatible with the formation of functional heterooligomers of TcMPC1 and TcMPC2, in contrast with the postulated formation of active homocomplexes  (14). When yeast MPC1/MPC2 were originally identified, it was reported that blue native polyacrylamide gel electrophoresis showed that both proteins migrated as part of an ;150-kDa complex, and it was concluded that MPC1 and MPC2 form a multimeric complex, with MPC2 as the major structural subunit (6). However, recent work has found no evidence of a multimeric complex and characterized them as a heterodimer (13).
In summary, the mitochondrial pyruvate carrier of T. cruzi is essential for pyruvatedriven respiration, trypomastigote invasion of host cells, and intracellular amastigote replication.

MATERIALS AND METHODS
Chemicals and reagents. Heat-inactivated fetal bovine serum (FBS) and penicillin-streptomycin were from Vitrocell (Campinas, Brazil). G418 was from KSE Scientific (Durham, NC). Puromycin, blasticidin S HCl, Platinum Taq DNA polymerase high fidelity, subcloning efficiency DH5a competent cells, and MitoTracker Deep Red FM were from Invitrogen (Eugene, OR). Antarctic phosphatase, BamHI, XbaI, XhoI, and HindIII restriction enzymes were from New England BioLabs (Ipswich, MA). T4 DNA ligase, GoTaq G2 Flexi DNA polymerase, Wizard Plus SV miniprep purification system, Wizard SV gel, and PCR clean-up system were from Promega (Madison, WI). Four-millimeter electroporation cuvettes, Zeta-Probe GT membrane, Precision Plus protein dual color standards, and nitrocellulose membranes were from Bio-Rad (Hercules, CA). North2South chemiluminescent hybridization and detection kit, North2South biotin random prime labeling kit, bicinchoninic acid (BCA) protein assay kit, SuperSignal West Pico chemiluminescent substrate, mouse anti-HA monoclonal antibody, and calcium green-5N were from Thermo Fisher Scientific Cell cultures. T. cruzi epimastigotes (Y strain) were grown at 28°C in liver infusion tryptose (LIT) medium (51) supplemented with 10% heat-inactivated fetal bovine serum (FBS), penicillin (100 U ml 21 ), and streptomycin (100 mg ml 21 ). LIT medium contained 2 g liter 21 glucose. To prepare a low-glucose LIT medium, glucose was not added. The glucose in the supplemented low-glucose LIT medium (approximate glucose concentration of 0.07 g liter 21 ) was from the FBS. Cell density was verified by counting parasites using Muse cell analyzer (Millipore, Burlington, MA). Parasites from the control cell lines transfected with scrambled sgRNA (scrambled) or pTREX-n-3ÂHA empty vector (EV), and overexpressing (OE) cell lines TcMPC1-OE and TcMPC2-OE were grown in the presence of 250 mg ml 21 G418. TcMPC1-TAG and TcMPC2-TAG cell lines were maintained with 250 mg ml 21 G418 and 5 mg ml 21 puromycin. TcMPC1-KO and TcMPC2-KO cell lines were grown with 250 mg ml 21 G418 and 10 mg ml 21 blasticidin. Tissue culture-derived trypomastigotes were collected from the culture medium of infected Vero cells, using a modification of the method of Schmatz and Murray (1982) as previously described (52). Vero cells were grown in RPMI medium supplemented with 10% FBS and maintained at 37°C with 5% CO 2 .
Endogenous C-terminal tagging of TcMPCs. CRISPR/Cas9-mediated endogenous C-terminal tagging of TcMPC proteins was performed as previously described (26,27). To perform individual tagging FIG 8 Legend (Continued) complex II of the respiratory chain via FADH 2 , as demonstrated by our findings of a decrease in oxygen consumption by the succinate dehydrogenase inhibitor malonate. Individual knockout of TcMPC1 and TcMPC2 (bottom) impairs pyruvate transport across the inner mitochondrial membrane (IMM), causing accumulation of pyruvate in the cytosol. Pyruvate and glutamate are converted into alanine and a-ketoglutarate (a-KG) by alanine aminotransferase. a-KG is transported to the mitochondrion through the a-ketoglutarate carrier (yellow) that exchanges a-KG for malate. Pyruvate can then be regenerated from cytosolic malate by the malic enzyme. The increase in oxygen consumption with a-ketoglutarate is in agreement with increased a-KGDH activity in TcMPC-KO epimastigotes. Ca 21 uptake is diminished in the mitochondria of knockout cells, which decreases phosphatase activity of PDP, and consequently, this leads to a higher PDH phosphorylation by PDK (in at least TcMPC2-KO cells). Knockout of TcMPCs in infective stages resulted in decreased ability to infect host cells and replicate as intracellular amastigotes. Solid arrows indicate enzymatic reactions, and dashed black arrows represent transport of metabolites through mitochondrial carriers. OMM, outer mitochondrial membrane.
of TcMPC1 and TcMPC2, we chose protospacer targeting regions close to the 39 end of the putative TcMPC1 gene (294 bp, TriTrypDB gene ID, TcCLB.511577.144) and the putative TcMPC2 gene (414 bp, TriTrypDB gene ID, TcCLB.508265.44). sgRNAs were amplified by PCR using a forward oligonucleotide (primers 1 and 2, Table S1 in the supplemental material) and a common reverse primer (Rv_sgRNA, primer 3, Table S1) and further cloned into the Cas9/pTREX-n vector. To amplify DNA donor cassettes containing 3Âc-Myc tag sequence and marker for puromycin resistance (Pac gene), a set of long primers (ultramers) (primers 5 to 8, Table S1) for each gene was designed. DNA donors were amplified by PCR using pMOTag23M vector as the DNA template. Wild-type T. cruzi epimastigotes (Y strain) were cotransfected with the gene-specific sgRNA/Cas9/pTREX-n plasmid and linear DNA donor. After 6 weeks of selection with G418 and puromycin, tagging of TcMPC1 and TcMPC2 was confirmed by PCR analysis using specific primer sets (primers 9 to 16, Table S1) and by Western blotting. A scrambled sgRNA cell line was used as control (53).
Knockout of TcMPCs. Molecular construction of sgRNA/Cas9/pTREX-n vector was performed as previously described (28). For knocking out TcMPC1 and TcMPC2 individually, we chose protospacer sequences closest to the 59 end of the gene of interest (GOI), and sgRNAs were amplified by PCR using a forward oligonucleotide (primers 17 and 18, Table S1) and a common reverse primer (Rv_sgRNA, primer 3, Table S1). Ultramers (primers 19 to 22, Table S1) for each gene were designed to amplify DNA donors containing the blasticidin resistance gene (Bsd) using pGEM-Bsd vector as the DNA template. Wild-type T. cruzi epimastigotes (Y strain) were cotransfected with TcMPC-specific sgRNA/Cas9/pTREX-n vector and linear DNA donor. A scrambled sgRNA cell line was used as the control. After 6 weeks of selection with G418 and blasticidin, knockout of TcMPC1 and TcMPC2 was confirmed by PCR analysis using specific primer sets (primers 23 to 32, Table S1) and by Southern blotting.
Overexpression of TcMPC proteins. For overexpression of 3ÂHA-tagged TcMPC1 and TcMPC2 proteins, the primers 33 to 36 (Table S1) were designed (XbaI and XhoI restriction sites in the primers forward and reverse, respectively). The pTREX-n/3ÂHA vector, which provides resistance to G418, and PCR products were doubly digested with XbaI and XhoI. The inserts were independently cloned into the digested pTREX-n/3ÂHA vector using T4 DNA Ligase, generating pTREX-n/TcMPC1-3ÂHA and pTREX-n/ TcMPC2-3ÂHA plasmids. The recombinant constructs were confirmed by PCR analysis with Fw_HX1-pTREX primer (primer 37, Table S1) and the gene-specific reverse primer, digestion with the restriction enzymes, and sequencing. Wild-type T. cruzi epimastigotes (Y strain) were transfected with recombinant constructs. Epimastigotes transfected with pTREX-n/3ÂHA empty vector were used as the control. After 6 weeks of selection with G418, overexpression of TcMPC1 or TcMPC2 was confirmed by Western blot analysis.
Cell transfections. Y strain wild-type T. cruzi epimastigotes were used for cotransfection of the sgRNA/Cas9/pTREX-n plasmid and linear DNA donor. Cell transfections were performed as reported previously (53) in triplicates. Negative controls were flasks containing cells without plasmid DNA that were subjected to the electroporation, and flasks containing cells without plasmid DNA that were not subjected to electroporation.
Transfectants TcMPC1-TAG and TcMPC2-TAG were cultured in LIT medium supplemented with 20% heat-inactivated FBS, penicillin, and streptomycin plus 250 mg ml 21 G418 and 5 mg ml 21 puromycin until stable cell lines were obtained. Transfectants TcMPC1-KO and TcMPC2-KO were cultured in the same medium plus 250 mg ml 21 G418 and 10 mg ml 21 blasticidin until stable cell lines were obtained.
For transfection with the overexpression vectors pTREX-n/TcMPC1-3ÂHA and pTREX-n/TcMPC2-3ÂHA, we followed the protocol described above using 25 mg of the plasmid. Transfectants TcMPC1-OE and TcMPC2-OE were cultured in LIT medium supplemented with 20% heat-inactivated FBS, penicillin, and streptomycin plus 250 mg ml 21 G418.
Southern blot analyses. To corroborate TcMPC1 and TcMPC2 gene knockouts by Southern blotting, 25 mg of gDNA from scrambled sgRNA and TcMPC1-KO cell lines was digested with HindIII enzyme and 25 mg of gDNA from scrambled sgRNA and TcMPC2-KO cell lines was digested with BamHI enzyme. Products of restriction enzyme digestion were resolved on a 1% agarose gel that was further subjected to alkaline blotting using a nylon membrane (Zeta-Probe GT membrane) according to the manufacturer's recommendations. The next steps were performed with a North2South chemiluminescent hybridization and detection kit. Probes for hybridization were amplified by PCR using the same primer sets used for the overexpression strategy (primers 33 to 36, Table S1) and pTREX-n/TcMPC1-3ÂHA or pTREXn/TcMPC2-3ÂHA vector as the DNA template. DNA probes were biotinylated using North2South biotin random prime labeling kit.
Western blot analyses. To confirm the expression of the proteins TcMPC1-3Âc-Myc, TcMPC2-3Âc-Myc, TcMPC1-3ÂHA, and TcMPC2-3ÂHA, mid-log-phase epimastigotes from the respective cell lines TcMPC1-TAG, TcMPC2-TAG, TcMPC1-OE, and TcMPC2-OE were harvested by centrifugation at 1,000 Â g for 7 min and washed twice with phosphate-buffered saline (PBS), pH 7.4, at room temperature (RT). The pellets were resuspended in 100 ml of RIPA buffer (150 mM NaCl, 20 mM Tris-HCl, pH 7.5, 1 mM EDTA, 1% sodium dodecyl sulfate [SDS], and 0.1% Triton X-100) plus 1 mM PMSF, 2.5 mM TPCK, 100 mM E64, and protease inhibitor cocktail (P8340, 1:250 dilution). Samples were incubated on ice for 1 h and then homogenized with a 1-ml syringe. Lysates were kept at 280°C until further use. Protein concentration in the lysates was determined using the BCA protein assay kit. Approximately 50 mg of each lysate was mixed with 6Â Laemmli sample buffer (300 mM Tris-HCl, pH 6.8, 6% [wt/vol] b-mercaptoethanol, 50% [vol/vol] glycerol, 12% [wt/vol] SDS, 4% [wt/vol] bromophenol blue) and then subjected to electrophoresis using a 12% SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Precision Plus protein dual color standard was applied on the gel. Electrophoresed proteins were transferred to a nitrocellulose membrane using a Bio-Rad transblot apparatus for 1 h at 100 V at 4°C. Following transfer, the membrane blots were blocked with 5% nonfat dry milk in PBS containing 0.1% (vol/vol) Tween 20 (PBS-T) at 4°C overnight. Blots were probed with mouse anti-c-Myc monoclonal antibody (1:100 dilution in PBS-T), mouse anti-HA monoclonal antibody (1:5,000 dilution in PBS-T), or mouse anti-tubulin monoclonal antibody (1:40,000 dilution in PBS-T) for 1 h at RT. After washing three times with PBS-T, the blots were incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG antibody (1:10,000 dilution in PBS-T) for 1 h at RT. The membranes were washed three times with PBS-T, and Western blot images were processed and analyzed using SuperSignal West Pico chemiluminescent substrate and a UVItec Cambridge Alliance 2.7 imaging system (UVItec, Cambridge, United Kingdom) according to the manufacturer's instructions.
Immunofluorescence analyses. For immunofluorescence assays of TcMPC1-TAG, TcMPC2-TAG, TcMPC1-OE, and TcMPC2-OE epimastigotes, cells were incubated with 100 nM MitoTracker Deep Red FM in culture medium for 30 min at 28°C. After staining with MitoTracker dye, cells were washed twice with PBS, pH 7.4 (PBS1), and fixed with 4% paraformaldehyde in PBS1 for 1 h at RT. Then, cells were allowed to adhere to poly-L-lysine-coated coverslips for 30 min. The coverslips were washed three times with PBS1 and incubated with 0.3% Triton X-100 in PBS1 for 3 min. Permeabilized cells were washed twice with PBS1 and blocked with PBS1 containing 5% goat serum, 50 mM NH 4 Cl, 3% bovine serum albumin (BSA), and 1% fish gelatin at 4°C overnight. After blocking, cells were incubated with a primary antibody, mouse anti-c-Myc monoclonal antibody (1:10 dilution, 1% BSA in PBS, pH 8.0 [PBS2]) for TcMPC-TAG epimastigotes or mouse anti-HA monoclonal antibody (1:500 dilution in PBS2) for TcMPC-OE epimastigotes, for 1 h at RT. The coverslips were washed three times with 1% BSA in PBS2 and then incubated with Alexa Fluor 488-conjugated goat anti-mouse antibody (1:1,000 dilution) diluted in 1% BSA in PBS2 for 1 h at RT in the dark. Cells were washed three times with 1% BSA in PBS2, washed once with PBS2, and counterstained with 49,6-diamidino-2-phenylindole (DAPI; 5 mg ml 21 ) in Fluoromount-G mounting medium on the slides. Negative controls were performed as described above but in the absence of the primary antibody. Differential interference contrast and fluorescence optical images were captured with a 100Â oil immersion lens objective (1.44 aperture) under nonsaturating conditions with a Leica TCS SP5 II confocal microscope, which uses photomultiplier tubes for the detection of emission. Digital images were acquired and processed with Leica LAS AF Lite software (Leica, Wetzlar, Germany).
Metacyclogenesis assay. The protocol described by Bourguignon et al. (55) was followed with minor modifications (20). Cells were harvested from TAU 3AAG medium at day 5 of cultivation. Three independent experiments were conducted.
Invasion and replication assays. Gamma-irradiated (2,000 radiation-absorbed doses) Vero cells (4.5 Â 10 5 cells) were plated onto sterile coverslips in a 12-well plate and incubated overnight at 35°C with 7% CO 2 in RPMI medium supplemented with 10% fresh FBS. 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 mammalian cells at a 10:1 ratio of parasites to host cells. At 4 h postinfection, coverslips were washed extensively with Hank's balanced salt solution, followed by PBS, pH 7.4, to remove any extracellular parasites. This protocol has been optimized with a multiplicity of infection (MOI) of 10 to ensure that, at most, one trypomastigote (Y strain) invades each irradiated host cell. Coverslips were immediately fixed in 4% paraformaldehyde in PBS, pH 7.4, at 4°C for 30 min and then washed once with PBS and counterstained with DAPI (15 mg ml 21 ) in Fluoromount-G mounting medium on glass slides, which stains host and parasite DNA. Coverslips were analyzed 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 40 randomly selected fields. Four hundred host cells were counted per sample in three independent experiments. To quantify amastigote replication, the following modifications were used: Vero cells were infected at a 10:1 ratio of parasites to host cells, and after washing them at 4 h postinfection as described above, coverslips were allowed to incubate for 48 h postinfection at 35°C with 7% CO 2 prior to fixation and DAPI staining.
Oxygen consumption assays. The oxygen consumption rate (OCR) of digitonin-permeabilized epimastigotes was measured using a high-resolution Oroboros Oxygraph-2k oxygraph (Oroboros Instruments GmbH, Innsbruck, Austria) with DatLab 4 software for data acquisition and analysis. The equipment was calibrated as reported by its manufacturer. Parasites from the cultures for wild-type, scrambled sgRNA, TcMPC1-KO, and TcMPC2-KO (1 Â 10 8 cells), pTREX-n/3ÂHA, TcMPC1-OE, and TcMPC2-OE (8.5 Â 10 7 cells) were washed with buffer A with glucose (BAG; 116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO 4 , 50 mM HEPES, pH 7.2, and 5.5 mM glucose). Then, cells were incubated at 28°C and 750 rpm in a 2-ml chamber containing 125 mM sucrose, 65 mM KCl, 10 mM HEPES-KOH, pH 7.2, 1 mM MgCl 2, 2.5 mM K 2 HPO 4 , 0.2% BSA, 50 mM EGTA, and 5 mM mitochondrial substrate (succinate, pyruvate, or a-ketoglutarate, as indicated in the figure legends). Sequential additions of digitonin (25 mM), ADP (100 mM), oligomycin (1 mg ml 21 ), and FCCP (1 mM) were performed during reading time. Inhibitors UK5099 (50 mM) and malonate (2 mM) were independently added into the chamber prior to reading in each experiment of inhibition using pyruvate as the mitochondrial substrate. OCR was calculated as the negative time derivative of O 2 concentration measured in the closed oxygraph chambers and expressed per milligram of protein. Data were recorded at 2-s intervals, and 10 data points were used to calculate the OCR slope plot through a polynomial fit with DatLab 4 software, as previously described (56). For comparison between OCRs of mutant cells and OCRs of control cells, normalized OCR was calculated from the normalization of OCR after an addition of ADP, oligomycin, or FCCP in relation to the basal OCR obtained right before ADP addition. A normalized OCR of 1 means that OCR after an addition is equal to the basal OCR; n independent biological experiments are indicated in the respective figure legends.
Mitochondrial calcium uptake assay. Mitochondrial Ca 21 uptake in permeabilized scrambled, TcMPC1-KO, and TcMPC2-KO epimastigotes was assayed by fluorescence of calcium green-5N probe at 28°C, as described previously (53), with the addition of 5 mM succinate or pyruvate to the reaction buffer (as indicated in the figures). Three independent experiments were performed.
Mitochondrial membrane potential assay. Mitochondrial membrane potential was assessed spectrofluorometrically using the indicator dye safranine O, as described previously (53,57) with the following modification: the reaction buffer contained 5 mM pyruvate or succinate (as indicated in the figure legends). Additions of ADP (250 mM), CAT (1.5 mM), and FCCP (4 mM) were performed at different time points. The adenine nucleotide translocator inhibitor CAT was used in the membrane potential assay instead of oligomycin, because the basal fluorescence is only reached with CAT following the addition of ADP. Three independent experiments were performed.
a-Ketoglutarate dehydrogenase activity assay. The measurements of a-ketoglutarate dehydrogenase (a-KGDH) activity in control, TcMPC1-KO, and TcMPC2-KO epimastigotes were performed as described previously (58) with some modifications. Briefly, protein extracts from scrambled, TcMPC1-KO, and TcMPC2-KO epimastigotes were obtained by centrifugation of 10 ml of culture in exponential phase (1 Â 10 7 to 2 Â 10 7 cells ml 21 ), washed with PBS, pH 7.4, at RT, and resuspended in 100 ml lysis buffer (0.1% Triton X-100, 50 mM K 2 HPO 4 [pH 7.35] and 5 U ml 21 Benzonase). Protein extracts were incubated for 10 min on ice before protein quantification with a BCA protein assay kit. The conversion of a-ketoglutarate, NAD 1 , and coenzyme A to succinyl-CoA, NADH, and CO 2 catalyzed by a-KGDH was measured in 96-well plates by following NADH fluorescence for 10 min at wavelengths of 340 and 460 nm for excitation and emission, respectively, using a Synergy H1 hybrid reader (BioTek, Winooski, VT). The a-KGDH activity assay was carried out in reaction buffer containing 50 mM K 2 HPO 4 (pH 7.35), 1 mM MgCl 2 , 0.2 mM thiamine pyrophosphate, 0.3 mM dithiothreitol, 100 mM EGTA, 50mM coenzyme A-SH, 2 mM NAD 1 , 1 mM a-ketoglutarate, and 5 mg protein extract in a final volume of 260ml, at RT. Blank wells containing lysis buffer instead of protein extracts were included in each assay, and their mean fluorescence was subtracted from that in the reaction wells. V max values were used to normalize the a-KGDH activity from each cell line using control parasites as a reference sample. Three independent experiments were performed.
Statistical analyses. Values are expressed as means 6 standard errors of the means (SEMs). Significant differences between treatments were compared using one-way and two-way analysis of variance (ANOVA) tests and the Kruskal-Wallis test, where indicated. Differences were considered statistically significant at P value of ,0.05, and n refers to the number of experiments performed. All statistical analyses were conducted using GraphPad Prism 6 (GraphPad Software, San Diego, CA).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.