The metabolic repair enzyme phosphoglycolate phosphatase regulates central carbon metabolism and fosmidomycin sensitivity in Plasmodium falciparum

The asexual blood stages of the malaria parasite, Plasmodium falciparum are highly dependent on glycolysis for ATP synthesis, redox balance and provision of essential anabolic precursors. Recent studies have suggested that members of the haloacid dehalogenase (HAD) family of metabolite phosphatases may play an important role in regulating multiple pathways in P. falciparum central carbon metabolism. Here, we show that the P. falciparum HAD protein, phosphoglycolate phosphatase (PfPGP), which is homologous to yeast Pho13 and mammalian PGP, regulates glycolysis in asexual blood stages by controlling intracellular levels of several intermediates and novel end-products of this pathway. Deletion of the P. falciparum pgp gene significantly attenuated asexual parasite growth in red blood cells, while comprehensive metabolomic analysis revealed the accumulation of two previously uncharacterized metabolites, as well as changes in a number of intermediates in glycolysis and the pentose phosphate pathway. The two unknown metabolites were assigned as 2-phospho-lactate and 4-phosphoerythronate by comparison of their mass spectra with synthetic standards. 2-Phospho-lactate was significantly elevated in wildtype and ΔPfPGP parasites cultivated in the presence of methylglyoxal and D-lactate, but not L-lactate, indicating that it is a novel end-product of the methylglyoxal pathway. 4-Phosphoerythronate is a putative side product of the glycolytic enzyme, glyceraldehyde dehydrogenase and the accumulation of both 4-phosphoerythronate and 2-phospho-D-lactate were associated with changes in glycolytic and the pentose phosphate pathway fluxes as shown by 13C-glucose labelling studies and increased sensitivity of the ΔPfPGP parasites to the drug fosmidomycin. Our results suggest that PfPGP contributes to a novel futile metabolic cycle involving the phosphorylation/dephosphorylation of D-lactate as well as detoxification of metabolites, such as 4-phosphoerythronate, and both may have important roles in regulating P. falciparum central carbon metabolism. Author summary The major pathogenic stages of the malaria parasite, Plasmodium falciparum, develop in red blood cells where they have access to an abundant supply of glucose. Unsurprisingly these parasite stages are addicted to using glucose, which is catabolized in the glycolytic and the pentose phosphate pathways. While these pathways also exist in host cells, there is increasing evidence that P. falciparum has evolved novel ways for regulating glucose metabolism that could be targeted by next-generation of anti-malarial drugs. In this study, we show the red blood cell stages of P. falciparum express an enzyme that is specifically involved in regulating the intracellular levels of two metabolites that are novel end-products or side products of glycolysis. Parasite mutants lacking this enzyme are viable but exhibit diminished growth rates in red blood cells. These mutant lines accumulate the two metabolites, and exhibit global changes in central carbon metabolism. Our findings suggest that metabolic end/side products of glycolysis directly regulate the metabolism of these parasites, and that the intracellular levels of these are tightly controlled by previously uncharacterized metabolite phosphatases.


Introduction
Plasmodium falciparum is the major cause of malaria, a disease that continues to kill  445,000 people each year and has significant impacts on the development of some of the poorest countries (1).The symptoms of malaria arise from the progressive ~ 48-hour cycles of parasite invasion into host red blood cells (RBC), rapid growth of asexual parasite stages within RBC and subsequent RBC lysis.RBC provide intracellular parasite stages with abundant supplies of glucose, amino acids and other carbon sources derived from the serum and/or breakdown of RBC proteins and lipids.P. falciparum asexual blood stages primarily use glucose as their major carbon source and are largely dependent on glycolysis for generation of ATP, although they retain a low flux TCA cycle for generation of the mitochondrial membrane potential (2)(3)(4).Rates of glucose utilization and L-lactate production from glycolysis are increased up to 100-fold in P. falciparum-infected RBC compared to uninfected RBC (5).The high glycolytic flux of P. falciparum asexual stages provides these cells with sufficient ATP and biosynthetic precursors to sustain anabolic processes and high rates of replication.However, high rates of glycolysis can generate toxic metabolic end-products, such as methylglyoxal, which can lead to increased oxidative stress, chemical modification and denaturation/inactivation of proteins, lipids and DNA (6)(7)(8).Thus, it is likely that P. falciparum has evolved ways to regulate pathways such as glycolysis under nutrient excess conditions.P. falciparum parasites express only a limited number of transcription factors and possess limited nutrient-stimulated transcriptional control (9)(10)(11)(12), indicating that regulation of central carbon metabolism primarily occurs at the post-transcriptional level.It is intriguing that biochemical investigations have uncovered a lack of conventional eukaryote allosteric regulatory/feedback mechanisms, suggesting dependence on other post-transcriptional regulatory mechanisms (13,14).One group of proteins that has a role in regulating glycolysis in P. falciparum asexual stages is the haloacid dehalogenase (HAD) family of metabolite phosphatases.The first member of this family to be functionally characterized was PfHAD1, identified in a screen for P. falciparum mutant lines that were resistant to the isoprenoid biosynthesis inhibitor fosmidomycin (15).HAD1 exhibited broad in vitro phosphatase activity against a range of sugar-phosphates and triose-phosphates that are connected to glycolysis, suggesting that it might have a role in vivo in the promiscuous dephosphorylation of glycolytic intermediates and negatively regulate glycolytic fluxes.Mutational inactivation of HAD1 leads to increased glycolytic flux and flow of intermediates into anabolic pathways such as isoprenoid biosynthesis, with associated increase in fosmidomycin resistance.Interestingly, a second HAD enzyme, HAD2 was identified in the same screen and was shown to exhibit in vitro phosphatase activity against glycolytic intermediates, suggesting that HAD enzymes may regulate multiple pathways in P. falciparum central carbon metabolism (16).
One of the HAD family members in the P. falciparum genome is homologous to the enzyme phosphoglycolate phosphatase (PGP), which is involved in regulating intracellular levels of several metabolites, including glycerol-3-phosphate (Gro3P), 2-phosphoglycolate, 2-phospho-Llactate and 4-phosphoerythronate (4-PE) (17,18).2-Phospho-L-lactate and 4-PE are thought to be minor side-products of the high flux reactions catalyzed by pyruvate kinase and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), respectively (17).PGP may thus act as a metabolite repair enzyme that detoxifies metabolites that would otherwise accumulate and allosterically affect key enzymes in central carbon metabolism (19,20).In this study we have investigated the role of P. falciparum PGP in asexual blood stages.In contrast to the situation in animal cells, we find that P. falciparum accumulates 2-phospho-D-lactate, rather than 2phospho-L-lactate, indicating synthesis through the methylglyoxal pathway rather than via enzymes in lower glycolysis.We show that PfPGP regulates the level of this novel end-product and contributes to a metabolic futile cycle that may regulate intracellular ATP levels.PfPGP is also involved in detoxifying 4-PE and we provide evidence that the accumulation of 4-PE leads to dysregulation of the pentose phosphate pathway and glycolysis, as well as increased sensitivity to fosmidomycin.Overall, these data highlight novel aspects of P. falciparum glycolysis and a key role for PfPGP in regulating central carbon metabolism in asexual blood stages.

P. falciparum PGP is a cytoplasmic protein required for normal growth of asexual stages in RBC
The P. falciparum gene, PF3D7_0715000 (hereinafter referred to as PfPGP) shares homology to the phosphoglycolate phosphatase (PGP) family of HAD enzymes that are involved in regulating the intracellular levels of key phospho-intermediates or end-products generated in glycolytically active cells, and features all four characteristic HAD motifs in its sequence (Fig. S1) (21).Previous studies have suggested that PfPGP may be involved in vitamin B1 biosynthesis, although this role has yet to be confirmed (22).Given the strong dependence of P. falciparum asexual blood stages on glycolysis for generation of ATP, redox balance and generation of anabolic precursors, we reinvestigated the functional role of PfPGP.Consistent with previous studies (22), we show that a PfPGP-GFP fusion protein is exclusively located in the cytoplasm of asexual blood stages (Fig. 1a) and is readily extracted in either phosphate buffer saline/sodium dodecyl sulfate (SDS) or radioimmunoprecipitation assay (RIPA) buffer/SDS (Fig. S2a).
To characterise the role of PfPGP in vivo, a knock-out parasite line was generated using CRISPR/Cas9 (Fig. S2b) (23).PGP-deleted P. falciparum mutants (∆PfPGP) were recovered and loss of the gene was confirmed by PCR (Fig. S2c).Analysis of the growth of two clones of the ∆PfPGP line indicated that loss of this gene was associated with a significant decrease in growth rate (Fig. 1b; P = 0.0005 and 0.039 for clones 1 and 2, respectively).While the NF54 parental line had a doubling time of 1.04  0.12 days, ∆PfPGP clones 1 and 2 had doubling times of 1.48  0.24 days and 1.21  0.03 days, respectively.As the culture medium contains vitamin B1 and other vitamins, these data support the premise that PfPGP has a metabolic function distinct from vitamin B1 biosynthesis.
To define the function of PfPGP, we undertook a comprehensive metabolomic analysis of parental (NF54) and ∆PfPGP mutant parasite lines.Trophozoite-stage infected RBC were magnetically enriched (> 95% parasitemia), and metabolites were extracted and analyzed by liquid chromatography -mass spectrometry (LC-MS).The mutant parasite line showed selective increases in a number of metabolites associated with the pentose phosphate pathway (6phosphogluconate, ribose/ribulose-5-phosphate), flavin mononucleotide and two unknown peaks with accurate masses (m/z; mass over charge ratio) of 168.9909 and 214.9968, respectively (Fig. 1c).Candidate metabolites for 168.9909 m/z included the glycolytic intermediates, glyceraldehyde phosphate (GAP) and dihydroxyacetone phosphate (DHAP), or the two enantiomers of phospho-D/L-lactate (METLIN metabolite database).GAP or DHAP were discounted based on the lack of co-elution with authentic standards for these metabolites on LC-MS or gas chromatography -mass spectrometry (GC-MS), indicating that the 168.9909 m/z peak may be phospho-D/L-lactate.To distinguish between the two possible enantiomers, parasite-infected RBC were incubated in the presence of either L-or D-lactate and intracellular levels of the 168.9909 m/z peak determined by targeted GC-MS analysis (Fig. 1d).Consistent with the untargeted analysis, accumulation of 168.9909 m/z was increased 12-fold ( 2.4) in the ∆PfPGP line compared to the parental wildtype (WT) line (Fig. 1d, left panel).Addition of 2 mM L-lactate (the major enantiomer generated by glycolysis) to WT or ∆PfPGP parasite cultures had no effect on the intracellular levels of this metabolite.In contrast, incubation of either WT or ∆PfPGP parasites with D-lactate resulted in a marked increase of the 168.9909 m/z peak (Fig. 1d, right panel).The 168.9909 m/z peak was confirmed as phospholactate by synthesizing racemic 2-phospho-D/L-lactate.This standard had the same fragmentation profile and retention time as the peak of interest via GC-MS (Fig. S3).Taken together, these data indicate that WT parasites are capable of phosphorylating endogenous and exogenous D-lactate to form 2-phospho-D-lactate (Fig. 1e), and that PfPGP is involved in dephosphorylating this species back to D-lactate.

Phospholactate is a product of the methylglyoxal pathway
In animal cells, 2-phospho-L-lactate is thought to be a by-product of the terminal glycolytic enzyme, pyruvate kinase (17).The finding that P. falciparum phospholactate is derived from the D-stereoisomer of lactate indicates, for this case, that an additional pathway for synthesizing this intermediate must operate.The only pathway known to generate D-lactate in P. falciparum is the methylglyoxal pathway, which is required for detoxification of methylglyoxal formed by non-enzymatic phosphate elimination of the glycolytic triose phosphates, GAP and DHAP (7,24).Methylglyoxal is converted to S-D-lactoyl-glutathione by glyoxalase I (GloI) and then further metabolized to D-lactate by glyoxalase II (GloII) (6) (Fig. 2a).To investigate whether the D-lactate generated in this pathway is converted to 2-phospho-D-lactate we generated P. falciparum knock-out lines lacking the cytoplasmic gloI gene (PF3D7_1113700) using the CRISPR/Cas9 system (Fig. S4a) (23).Knock-out of PfgloI was confirmed by PCR (Fig. S4b).
To investigate whether 2-phospho-D-lactate is synthesized by the methylglyoxal pathway, parental WT and ∆GloI parasite cultures were suspended in medium containing methylglyoxal (1 mM, 1 hour, 37C), then levels of 2-phospho-D-lactate were measured by GC-MS.Addition of methylglyoxal led to a 15-fold increase in 2-phospho-D-lactate levels, indicating that this metabolite is the end-product of the methylglyoxal pathway (Fig. 2d).Strikingly, similar levels of 2-phospho-D-lactate were present in both WT and ∆GloI parasites, before and after addition of methylglyoxal, indicating that synthesis of D-lactate is not strictly dependent on PfGloI.It is possible that the apicoplast isoform of GloI (GloI-like protein, GILP) could substitute for cytoplasmic GloI and/or that methylglyoxal can be converted to S-D-lactoyl-glutathione nonenzymatically.Alternatively, methylglyoxal may be converted to D-lactate by the RBC methylglyoxal pathway and D-lactate subsequently imported by the parasite and converted to 2phospho-D-lactate.Collectively, these studies suggest that 2-phospho-D-lactate is generated by the direct phosphorylation of D-lactate produced by either the host cell, or the parasite methylglyoxal detoxification pathways.

Accumulation of 2-phospho-D-lactate is toxic at non-physiological concentrations
Next, we determined whether the reduced growth rate of the ∆PfPGP mutant lines in RBC could be attributed to the toxic accumulation of 2-phospho-D-lactate.WT-and ∆PfPGP-infected RBC were suspended in medium containing 0, 1 and 5 mM D-lactate to elevate intracellular levels of 2-phospho-D-lactate and asexual parasite growth was monitored over a period of 13 days.
Growth of both WT and ∆PfPGP, were significantly reduced in the presence of a high (5 mM) concentration of D-lactate (Fig. 3a,b), with ∆PfPGP appearing to be more susceptible to the metabolic treatment (growth decreased by 49.49%  2.31 compared to 20.25%  5.24 for WT).This result suggests that accumulation of 2-phospho-D-lactate is toxic to the parasite and that PfPGP plays a key role in maintaining non-toxic levels.However, extracellular concentrations of D-lactate in P. falciparum cultures are generally below 1 mM indicating that 2-phospho-Dlactate toxicity may not be the major cause of the growth defect of the ∆PfPGP mutant under physiological growth conditions.An alternative possibility is that the inter-conversion of D-lactate and 2-phospho-D-lactate, mediated by an unknown kinase and PfPGP, has a metabolic function that is required for parasite asexual growth.We hypothesized that D-lactate may be converted to 2-phospho-Dlactate to prevent its secretion with L-lactate.Consistent with this proposal, analysis of the culture medium of uninfected and infected RBC showed that secretion of D-lactate was reduced by 75% in the latter, indicating that D-lactate may be sequestered within the parasite as 2phospho-D-lactate (Fig. 3c).This finding differed from the previously reported ~30-fold increase in D-lactate secretion following erythrocyte infection (24).Here we deproteinised samples before detection and suspect that without this step, misleadingly high levels of D-lactate are observed due to residual enzyme activity in the media.

Loss of PfPGP leads to increased flux through the pentose phosphate pathway
Our untargeted LC-MS metabolomic studies identified a second unknown metabolite accumulating in the ∆PfPGP mutant of 214.9968 m/z (Fig. 1c).METLIN database searching suggested that this metabolite might be 4-phosphoerythronate (4-PE; Fig. 4a).4-PE is not an intermediate in canonical metabolic pathways but is thought to be generated by GAPDH acting on the pentose phosphate pathway intermediate erythrose-4-phosphate, instead of its preferred substrate, GAP (17).The identity of the 214.9968 m/z peak was confirmed as 4-PE through comparison of its MS spectrum and GC-MS retention time with those of a syntheticallyacquired standard (Fig. 4b and Table S1).The intracellular levels of 4-PE were not affected by addition of exogenous D-or L-lactate to WT-and ∆PfPGP-infected RBC cultures, indicating that this metabolite is generated independently of the D-lactate/ 2-phospho-D-lactate pathway (Fig. S5).
These results suggest that the accumulation of 4-PE in ∆PfPGP parasites (µM range) may partially inhibit activity of 6-PGD in vivo.
Paradoxically, our metabolomic studies indicated that levels of downstream intermediates in the oxidative and non-oxidative PPP (ribose-5-P/ribulose-5-P) were elevated, rather than decreased, in ∆PfPGP parasites (Fig. 1c).An increase in the pool size of these pentose-phosphates could reflect increased flux through the non-oxidative PPP and/or a compensating increase in flux through the oxidative PPP (overcoming the partial inhibition of 6-PGD).To address this question directly, enriched WT-and ∆PfPGP-infected RBC cultures were metabolically labelled with 13 C-1,2-glucose to measure fluxes through both arms of the PPP.Catabolism of 13 C-1,2glucose through the oxidative arm of the PPP leads to loss of 13 C on carbon-1 of 6phosphogluconate (as carbon dioxide) as this intermediate is converted to 13 C 1 -ribose-5-P.In contrast, conversion of 13 C-1,2-glucose to ribose-5-P via the non-oxidative pathway does not involve a decarboxylation step resulting in 13 C 2 -ribose-5-P (Fig. 4d).Purified WT-and ∆PfPGPinfected RBC were incubated with 13 C-1,2-glucose for 30 minutes at 37°C and 13 C-enrichment in ribose-5-P was determined by GC-MS.Ribose-5-P levels were elevated in ∆PfPGP parasite cultures (Fig. 4e, Table S1), consistent with the results of the initial metabolomic analyses (Fig. 1c).The overall rate of turnover of pentose-phosphates were significantly increased in the ∆PfPGP mutant (Fig. 4f, M0 fraction).This was entirely due to increased production of 13 C 1ribose-5-P (and 13 C 3 -ribose-5-P) and indicates an increased flux through the oxidative PPP.It is notable that the flux through the non-oxidative PPP (as indicated by levels of 13 C 2 -ribose-5-P) was relatively low in both WT and ∆PfPGP parasites (Fig. 4f).These results suggest that inhibition of 6-PGD in the ∆PfPGP parasite lines, due to the accumulation of 4-PE, is more than compensated for by increased flux through this pathway.Therefore, partial inhibition of the oxidative PPP in the ∆PfPGP line is an unlikely cause of the decreased growth rate of asexual stages in RBC.

Loss of PfPGP is associated with changes in glycolytic flux and increased sensitivity to fosmidomycin
The untargeted LC-MS profiling of ∆PfPGP parasites revealed that DHAP was one of the few metabolites to be significantly down-regulated in the mutant (Fig. 1c).The interconversion of DHAP and GAP is mediated by the glycolytic enzyme, triosephosphate isomerase (TPI) which, like 6-PGD, catalyzes a reaction utilizing an ene-diolate intermediate that might be sensitive to allosteric modulation by 4-PE (26,27).DHAP and other triose-phosphates are catabolyzed in the glycolytic pathway and imported into the apicoplast where they are used for synthesis of 1deoxy-D-xylulose-5-P, the first committed intermediate in isoprenoid biosynthesis.Conversion of 1-deoxy-D-xylulose-5-P into 2-C-methyl-D-erythritol-4-P is inhibited by fosmidomycin, a potent antimalarial (Fig. 5a) (28).We hypothesised that perturbation of glycolytic flux and/or balance of triose-phosphates in the ∆PfPGP mutant could lead to reduced isoprenoid biosynthesis and increased sensitivity to fosmidomycin.We evaluated the sensitivity of ∆PfPGP-infected RBC to fosmidomycin in a 72-hour drug treatment assay and measured effects upon parasite growth via flow cytometry (Fig. 5b).In comparison to WT, the ∆PfPGP mutant exhibited a significant 4-fold increase in fosmidomycin sensitivity (EC 50 WT = 358 nM  14; ∆PfPGP = 89 nM  9.8; P = 0.001).These findings suggest that elevated 4-PE levels in ∆PfPGP parasites lead to perturbations in glycolytic flux, reduced isoprenoid synthesis and concomitant increase in fosmidomycin sensitivity.
To confirm that the metabolic dysregulation of glycolysis and the oxidative pentose phosphate pathway observed in the ∆PfPGP parasite was due to the elevation of 4-PE and not 2-phospho-D-lactate, we incubated WT parasites with D-and L-lactate and performed untargeted LC-MS profiling (Fig. S7).D-lactate incubation led to a selective increase in the phospholactate pool whereas the 4-PE pool remained unchanged.The only other metabolite altered was the D/Llactate peak itself, indicating that the observed reduction of DHAP and increase in ribose-5-P (Fig. 1c) are most likely the result of 4-PE effects on parasite metabolism.

Discussion
All eukaryotic and prokaryotic cells express members of the HAD family of metabolite phosphatases, although the function of these proteins in vivo are poorly defined.There is increasing evidence that these enzymes have important roles in regulating intracellular levels of a range of phosphorylated intermediates and metabolic fluxes in cells.In this study, we provide evidence that the P. falciparum HAD family member, PfPGP, has at least two functions.First, it participates in a novel metabolic cycle involving the phosphorylation and dephosphorylation of D-lactate.Second, it is required to detoxify 4-PE, an allosterically-active metabolic sideproduct.Targeted deletion of PfPGP resulted in the accumulation of both 2-phospho-D-lactate and 4-PE, attenuated growth of P. falciparum asexual blood stages, and increased sensitivity to the antimalarial drug, fosmidomycin.Our results identify important differences in the function of PGP in P. falciparum compared to other eukaryotes, consistent with a greater dependence of these parasites on non-transcriptional metabolic regulation.
The PGPs were initially identified in plants as enzymes involved in converting phosphoglycolate, a side product of photorespiration, to glycolate (29).The subsequent conversion of glycolate to 3-phosphoglycerate and its catabolism in the Calvin-Benson cycle increases the efficiency of plant photorespiration by ~ 25%.More recently, studies on the role of PGP in yeast and animal cells have indicated that these enzymes have broader substrate preferences, acting on metabolites such as glycerol-3-phosphate, 2-phospho-L-lactate and 4-PE (17,18).Our metabolomic analyses indicate that PfPGP has a similar substrate specificity to the yeast/animal PGPs, although with important differences.In particular, PfPGP appears to act predominantly on the D-enantiomer of 2-phospholactate rather than the L-enantiomer as proposed to occur in animals.L-lactate is the major end-product of glycolysis and 2-phospho-Llactate is thought to be a minor side product of the glycolytic enzyme, pyruvate kinase, which normally converts pyruvate to L-lactate (17).In contrast, in P. falciparum and other eukaryotes that lack a D-lactate-dehydrogenase, D-lactate is produced exclusively via the methylglyoxal pathway.We provide evidence that the 2-phospho-D-lactate detected in wildtype, ∆PfGloI and ∆PfPGP mutants is likely derived from this pathway.Specifically, incubation of wildtype or ∆PfGloI mutant parasites with methylglyoxal led to a concomitant increase in 2-phospho-Dlactate levels.Similarly, incubation of wildtype infected RBC with D-lactate, but not L-lactate, elevates intracellular levels of 2-phospho-D-lactate.The latter result indicates that 2-phospho-Dlactate is synthesized directly from D-lactate, via the action of an as yet unidentified kinase, rather than being a side product of pyruvate kinase.
Interestingly, targeted deletion of the cytoplasmic isoform of GloI, the first enzyme in the methylglyoxal pathway, had little effect on the production of 2-phospho-D-lactate.P. falciparum express a second GloI enzyme localized to the apicoplast (30,31), which could sustain production of D-lactate and 2-phospho-D-lactate in the ∆GloI mutant.Alternatively, or in addition, the RBC host retains a methylglyoxal pathway, which could convert parasite-and host-derived methylglyoxal to D-lactate.The presence of the host pathway may also explain why the parasite glyoxalases are non-essential, although loss of gloI results in significant attenuation of intracellular parasite growth over a long period of time (6 replication cycles).
Overall, these findings suggest that P. falciparum converts D-lactate generated by the parasite (and potentially host) methylglyoxal pathway(s), to 2-phospho-D-lactate via an as yet unidentified kinase.The 2-phospho-D-lactate is subsequently dephosphorylated by PfPGP to regenerate D-lactate.As D-lactate is not secreted by the parasite, PfPGP may contribute to an ATP-consuming futile metabolic cycle that is directly connected to glycolytic flux (Fig. 6).
What is the potential function of the D-lactate/ 2-phospho-D-lactate cycle in P. falciparum?It has recently been shown that 2-phospho-L-lactate (the major phospholactate species in animal cells) inhibits the bi-functional glycolytic enzymes, PFKFB1-4.The enzymes have both phospho-fructo-kinase (PFK) and fructose-2,6-biphosphatase activities and regulate the intracellular levels of the potent PFK allosteric regulator fructose-2,6-bisphosphate (17).Strikingly, P. falciparum lacks a PFKFB homologue, and PfPFK is insensitive to conserved metabolic inhibitors/activators (13), highlighting differences in the way these parasites regulate central carbon metabolism.The metabolic futile cycle catalyzed by PGP may have a direct role in maintaining cellular levels of ATP and in preventing excessive glycolytic flux under conditions of nutrient (glucose) excess.While it is difficult to directly estimate the capacity of this cycle to modulate intracellular ATP levels, it is noteworthy that nearly all of the D-lactate produced by infected RBC is retained intracellularly, despite a 100-fold increase in glycolytic flux (as indicated by glucose uptake and L-lactate secretion) (5).This finding suggests that most/all of the D-lactate produced via the parasite methylglyoxal pathway contributes to this cycle.Furthermore, enhanced flux through this pathway, induced by supplementation of infected RBC with D-lactate, led to reduced growth of asexual stages, possibly by increasing ATP cycling.Collectively, these data indicate that the D-lactate/ 2-phospho-D-lactate cycle is very active in these parasite stages and has a significant capacity to modulate ATP levels.
In addition to contributing to the D-lactate/ 2-phospho-D-lactate cycle, PfPGP has a role in regulating intracellular levels of 4-PE.In yeast and animal cells, 4-PE is thought to be generated by the glycolytic enzyme, GAPDH, acting upon the non-standard substrate erythrose-4-P (an intermediate of the pentose phosphate pathway) (17).4-PE is a potent negative regulator of the enzyme 6-phosphogluconate dehydrogenase in in vitro assays (17,32).In animal PGP-and yeast Pho13-null mutants, the accumulation of 4-PE was proposed to lead to inhibition of the oxidative arm of the PPP (17).In P. falciparum, loss of PfPGP resulted in a 9-fold ( 1.6) accumulation of 4-PE indicating that it has a similar role in asexual blood stages.However, in contrast to the situation in yeast and animal cells, significant inhibition of Pf6-PGD activity by 4-PE was only observed at 100 µM in vitro.In contrast, the ex vivo 13 C-1,2-glucose labelling experiments indicated increased flux through the oxidative PPP in the ∆PfPGP mutant line.
These analyses suggest that accumulation of 4-PE is insufficient to inhibit 6-PGD under normal in vivo growth conditions.It remains to be determined whether 4-PE inhibits other glycolytic/PPP enzymes in P. falciparum that have an ene-diolate intermediate, including triosephosphate isomerase (26,27) or glucose-6-phosphate isomerase (33).Inhibition of either enzyme could contribute to the reduced glycolytic flux (inferred by the increased sensitivity of ∆PfPGP mutant to fosmidomycin) or cause the increased flux into the oxidative PPP in the ∆PfPGP mutant and explain the growth defect observed for the ∆PfPGP mutant.This work highlights the important role of HAD enzymes in regulating P. falciparum central carbon metabolism.The role of two other P. falciparum HAD enzymes have also recently been examined (15,16).HAD1, the first member of this class to be functionally characterized in P. falciparum, promiscuously dephosphorylates a range of glycolytic intermediates in vitro, including triose-, pentose-and hexose-phosphates.Loss-of-function mutations in this protein lead to increased parasite resistance to fosmidomycin, and it was proposed that HAD1 has a role in negatively regulating glycolytic flux by removing glycolytic intermediates (15).Similarly, HAD2 dephosphorylates a range of glycolytic intermediates in vitro, and is linked to negative regulation of glycolysis (16).A common theme for all three P. falciparum HAD proteins characterized to date is their participation in ATP-depleting futile cycles through either depletion of high energy intermediates (HAD1/HAD2) or ATP-dependent cycles of phosphorylation/dephosphorylation (PfPGP).Although these futile cycles appear energetically wasteful, they are comparable to other ATP-dependent cycles, such as protein phosphorylation/dephosphorylation and/or constitutive turnover of mRNA and protein.
Metabolic regulation via futile cycling has the potential to be more responsive to subtle changes in carbon sources such as glucose and allows parasites to adapt to changing conditions faster than can be achieved with other forms of metabolic regulation (e.g.transcriptional regulation).
This study highlights the utility of metabolomic approaches in identifying new metabolic pathways and regulatory mechanisms in evolutionarily divergent eukaryotic parasites.A significant proportion of the genes in Plasmodium remain uncharacterized and a significant proportion of metabolites detected in comprehensive LC-MS and GC-MS profiling studies have yet to be structurally defined.Our study indicates that unanticipated complexity in cellular metabolism can arise as a result of enzymatic side reactions and/or chemical modification of canonical metabolites, leading to the evolution of new enzyme activities and pathways.We expect that further dissection of the side activities of the major enzymes in intermediary metabolism in P. falciparum, and associated repair enzymes will provide new insights into the evolution of novel metabolic regulatory processes in these parasites.Such advances will serve an important role in the identification of new, druggable targets for improved malaria therapeutics.
For the localization of PfPGP, the coding sequence of PfPGP was amplified by Phusion PCR (NEB) from freshly prepared NF54 genomic DNA and inserted into the pTEOE-GFP plasmid (34) using XhoI and AvrII restriction sites and InFusion cloning (Clontech).To stably integrate the overexpression construct randomly into the genome, the pHTH helper plasmid expressing the piggyBac transposase system was used (35).Plasmid stocks were prepared from midipreps (Macherey-Nagel).All sequences-of-interest were confirmed by Sanger sequencing (AGRF).
Primers used for cloning are presented in Table S2.
Validation of successful integration was confirmed by PCR on knock-out-NF54 genomic DNA (Bioline), using primer pairs that were specific to the integrated cassette and the genomic DNA.
Primers used for this purpose are presented in Table S2.

P. falciparum parasite growth assay
The parasitemia of synchronised ring cultures (0.8% haematocrit, 0.8% parasitemia) was assessed by flow cytometry each day for 13 days and identical dilutions between wildtype and knock-out infected RBC were applied regularly.Nucleic acids were stained with SYTO 61 (Invitrogen) and parasitemia was measured on a FACS Canto TM II (BD Biosciences).Data were analysed using the FlowJo software (BD Biosciences).Raw counts were normalized to the day 13 WT values and the dilutions made throughout the experiment were corrected for.

D-lactate in vitro assay
The concentration of D-lactate excreted from uninfected-and NF54 infected-RBC was measured using a D-lactate assay kit (Cayman chemical).All conditions were incubated for 26-29 hours (rings to trophozoite stage) in complete RPMI media.Media was collected, deproteinised and measured following the manufacturer's instructions.After incubation for 30 minutes at 37C, the plate was read at 590 nm (Ex.544 nm) on a FluoStar Omega plate reader (BMG Labtech) and analysed with the Omega data software.

in vitro assays
Infected-RBC were magnetically enriched at trophozoite stage (MAGNEX Cell Separator (Colebrook Bioscience); >95% parasitemia) and lysed in 0.1% saponin buffer.After three PBS washes and counting of isolated parasites, cell pellets were snap frozen in liquid nitrogen and stored at -80C.Pellets were resuspended in 200 µL per 1.10 8 cells of pH7.4 lysis buffer (5 mM HEPES, 2 mM DTT, protease inhibitor).Enzymatic reactions were setup in a 50:50 ratio cell lysate:reaction buffer.For measuring glyoxalase activity, the reaction buffer was composed of 100 mM Tris HCl, 5 mM NH 4 Cl, 2 mM MgCl 2 , 2 mM ATP, 2 mM methylglyoxal (MG) and 2 mM reduced glutathione.The control reaction consisted of the reaction buffer without MG (-MG condition) and used as the baseline for data normalization.Enzymatic reactions were setup at 37C and samples were collected in technical duplicates at 0, 5, 10, 30 and 60 minutes.6-PGD activity was measured as described above using a reaction buffer composed of 100 mM Tris HCl, 5 mM NH 4 Cl, 2 mM MgCl 2 , 1 mM 6-phospho gluconate, 1 mM NADP and 1mM 4phosphoerythronate (as annotated).Enzymatic reactions were setup at 37C, incubated for 60 minutes and collected in technical duplicates at 0 and 60 minutes.Samples were extracted for

GC-MS sample extraction and derivatisation
All in vitro GC-MS samples were extracted on ice with 100 µL chloroform and 400 µL 3:1 methanol:H 2 O containing 1 nM scyllo-inositol (internal standard).At the end of the sample collection, 200 µL H 2 O was added to each tube, samples were mixed thoroughly and centrifuged at 14 000 x g for 5 minutes at room temperature.The aqueous phase was transferred into a new tube, ready for GC-MS drying.
Thirty µL were transferred into mass spectrometry tubes (1/20 dilution) and dried (SpeedVac).50µL of pure methanol were used for an additional wash to remove any trace water and samples were dried in a SpeedVac.Samples were derivatised with 20 µL of 20 mg/mL methoxyamine (Sigma Aldrich) made up in pyridine (Sigma Aldrich) and left at room temperature overnight.
The next morning, 20 µL of a ready-to-use N,O-bis(trimethylsilyl) trifluoroacetamide (BSTFA) + 1 % trimethylchlorosilane (TMCS) solution (Supelco) was added to each sample.Metabolites were separated as described previously (40) using a BD5 capillary column (J&W Scientific, 30 m x 250 µM x 0.25 µM) on a Hewlett Packard 6890 system (5973 EI-quadrupole MS detector).Briefly, the oven temperature gradient was 70 °C (1 minute); 70 °C to 295 °C at 12.5 °C/minute, 295 °C to 320 °C at 25 °C/minute; 320 °C for 2 minutes.MS data was acquired using scan mode with a m/z range of 50-550, threshold 150 and scan rate of 2.91 scans/second.GC retention time and mass spectra were compared with authentic standards analysed in the same batch for metabolite identification.

C -glucose labelling and LC-MS analysis
Synchronised trophozoite cultures were magnetically enriched using a MAGNEX Cell Separator (Colebrook Bioscience).Purified cells (5x10 7 cells/sample) were incubated for 30 minutes at 37C in either RPMI containing 11 mM 13 C 1,2 -glucose (Cambridge Isotopes) (= fully labelled) or 11 mM 13 C-U-glucose mixed 1:1 with complete RPMI containing 11 mM unlabelled glucose (Sigma).Samples were centrifuged at 14 000 x g for 30 seconds, washed in 1 mL ice-cold PBS and pellets were extracted for GC-MS analysis for 13 C 1,2 -glucose labelled samples (as described above) or extracted for LC-MS analysis ( 13 C-U-glucose samples).LC-MS samples were resuspended in 100 µL of 80% acetonitrile (Burdick & Jackson) containing the internal standard 1 µM 4-13 C, 15 N-aspartate.After centrifugation at 14 000 x g for 5 minutes at 4C, supernatants were transferred into mass spectrometry vials.LC-MS analysis was performed as described previously (40), with the following modifications.Metabolite samples were separated on a SeQuant ZIC-pHILIC column (5 µM, 150 x 4.6 mm, Millipore) using a binary gradient with a 1200 series HPLC system (Agilent), with solvent A being water with 20 mM ammonium carbonate and solvent B 100% acetonitrile.The gradient ran linearly (at 0.3 mL/minute) from 80-20% solvent B from 0.5 to 15 minutes, then 20-5% between 15 and 20 minutes, before returning to 80% at 20.5 minutes and kept at 80% solvent B until 29.5 minutes.
MS detection was performed on an Agilent Q-TOF mass spectrometer 6545 operating in negative ESI mode.The scan range was 80-1200 m/z between 2 and 25 minutes at 0.9 spectra/second.An internal reference ion solution continually run (isocratic pump at 0.2 mL/minute) throughout the chromatographic separation to maintain mass accuracy.Other LC parameters were: autosampler temperature 4 °C, injection volume 10 µL and data were collected in centroid mode with Mass Hunter Workstation software (Agilent).
The untargeted profiling of ∆PfPGP-infected RBC compared to WT-infected RBC was performed as described above and the effect of L-and D-lactate on parasite metabolism was assessed by incubating purified infected RBC in complete RPMI containing +/-10 mM L-and D-lactate (Sigma) for one hour at 37C and analysed via LC-MS as described above.

Mass spectrometry data analysis
GC-MS data was processed using ChemStation (Agilent) or the in-house software package DExSI (41) and metabolites of interest were compared to authentic standards.LC-MS.dfiles were converted to .mzXMLfiles using MS convert and analysed using MAVEN (42).Following alignment, peaks were extracted with a mass tolerance of <10 ppm.Untargeted comparative profiling was performed to generate a list of m/z features of interest.The m/z value of each peak of interest was then queried against the METLIN metabolite database (METLIN reference) for only M-H adducts with a 10 ppm mass tolerance.Peaks of interest were positively identified using their exact mass and retention time (compared to a standards library of 150 compounds ran the same day).

Synthesis of sodium phospholactate
Pearlman's catalyst (20 % Pd(OH) 2 /C, 135 mg) was added to a solution of phospho(enol)pyruvic acid monosodium salt monohydrate (Sigma, 40 mg, 192 µmol) in THF/MeOH/AcOH 1:1:1 (3 mL) and the mixture was stirred at room temperature under an atmosphere of H 2 (40 bar) for 48 hours.The reaction mixture was diluted with MeOH (20 mL) then filtered through Celite, and evaporated at 10 mbar on a rotary evaporator at 40 °C.The residue was dissolved in deionized water, and the solution was passed through a C18 Sep-Pak.The eluent was evaporated at 10 mbar on a rotary evaporator at 40 °C, and the residue was dried at 10 -3 mbar at room temperature for 24 hours, affording D/L-phospholactate monosodium salt hydrate (40 mg) as a viscous oil.HRMS (ESI-TOF) calculated (calcd) for C 3 H 4 O 6 P -[M -Na] -, 168.9907 m/z, found 168.9917.
Cyclohexylamine (39 µL, 0.324 mmol) was added to the eluted product, and the volatiles were removed by rotary evaporation and further drying under an N 2 stream overnight.The crude material was purified on a Shimadzu 2020 LC-MS instrument, using an ACE Excel 5 Super C18 column (150 mm x 2.1 mm) with 0.1 % formic acid (solvent A) and methanol (solvent B).A linear gradient was performed from 80 % to 10 % across 15 minutes and elution of 4-PE was determined using the theoretical exact mass of 4-PE.Separation of 4-PE from contaminants was monitored using the MS in full scan mode and UV detection.Verification of the purified 4-PE was performed by re-running the collected aliquot on the instrument and confirming no other masses or UV peaks were observed (above background).

Protein analysis by western blotting
Trophozoite parasites were isolated by addition of 0.05% saponin in the culture media and spun down at 3 750 x g for 5 minutes.Pellets were washed twice in ice-cold PBS containing complete protease inhibitors (Roche) and resuspended in PBS or RIPA buffers containing protease inhibitors.RIPA-buffer samples were incubated on ice for 10 min and centrifuged at 16 000 x g for 10 minutes.The supernatant was collected and placed into a fresh tube.The saponin pellet (PBS) and the RIPA (supernatant) samples were mixed with Bolt 4X LDS and 10X reducing agent (Invitrogen) and heated at 85°C for 10 minutes prior SDS-PAGE on 4-12% BisTris gels separated in 1X MOPS running buffer (Invitrogen).Proteins were transferred onto nitrocellulose membranes using the iBlot 2 transfer system (Invitrogen) and membranes were blocked in 3.5% skim milk for at least one hour at room temperature.Membranes were probed with mouse anti-GFP (1:1000 -Roche) and rabbit anti-GAPDH (1:1000 -( 43)) primary antibodies.Secondary antibodies were horseradish-peroxidase conjugated: goat anti-mouse and anti-rabbit (1:20 000 -Promega).Membranes were incubated with Clarity ECL substrate (Bio-Rad) and imaged on a ChemiDoc MP system (BioRad).as the mean ± SEM of the cumulative parasitemia normalised to the 0 mM/ day 13 data point (100%).Dilutions of the cultures were taken into account and statistical significance was determined using paired t-testing at day 13 (* and *** denote P < 0.05 and 0.001 respectively).c) Extracellular levels of D-lactate were measured in uninfected RBC (unRBC) and WTinfected RBC (iRBC) after sample deproteinisation using a D-lactate plate assay (Cayman chemicals).Cultures were set at 2.5% haematocrit (and 1.8 to 4% parasitemia for infected RBC).Media was collected after 26-29 hours of culture (ring to trophozoite stages for infected RBC).Data are presented as the mean ± SEM from four independent repeats collected on different days and statistical significance was determined using unpaired t-testing (* denotes P < 0.05).Scatter plots of m/z ion counts are presented.Following both D-and L-lactate exposure, the intracellular lactate pool was significantly elevated (P < 0.05 Benjamini corrected).
Phospholactate was significantly elevated only following D-lactate incubation (P < 0.05 Benjamini corrected).Each scatter plot represents averaged data from three (L-lactate) and four (D-lactate) independent biological replicates.

Table S1. Repertoire of the metabolites-of-interest and analytical GC-MS features
Highlight of the quantified ion and retention time used for identifying specific metabolites in all GC-MS-based assays presented in this manuscript.

Table S2. Primer list
Primers are presented by section.For the CRISPR and pTEOE cloning sections, InFusion primers were designed.Upper cases are nucleotides that are part of the plasmid backbone, lower cases are nucleotides that are part of the gene-of-interest, underlined nucleotides correspond to restriction sites, and for GloI_HA1_rev_AflII the nucleotides in blue are mutated nucleotides.

Figure 1 .
Figure 1.PfPGP is required for normal growth of P. falciparum asexual stages and ∆PfPGP mutant parasites selectively accumulate metabolites in the PPP, as well as two novel metabolites including 2-phospholactate.

Figure 3 .
Figure 3. Accumulation of 2-phospho-D-lactate is only toxic at high concentrations

Figure 4 .
Figure 4. Loss of PfPGP leads to partial inhibition of 6-PGD, and enhanced flux through the oxidative PPP

Figure 5 .
Figure 5. Loss of PfPGP leads to increased sensitivity to fosmidomycin

Figure 6 .
Figure 6.PfPGP is a key metabolic repair enzyme

Figure S5 .
Figure S5.Levels of 4-phosphoerythronate are not affected by addition of exogenous D/L lactate

Figure S7 .
Figure S7.The increase in 2-phospho-D-lactate is not responsible for metabolic dysregulations observed in ∆PfPGP parasite cultures