Metabolome Analysis Revealed Increase in S-Methylcysteine and Phosphatidylisopropanolamine Synthesis upon l-Cysteine Deprivation in the Anaerobic Protozoan Parasite Entamoeba histolytica*

l-Cysteine is ubiquitous in all living organisms and is involved in a variety of functions, including the synthesis of iron-sulfur clusters and glutathione and the regulation of the structure, stability, and catalysis of proteins. In the protozoan parasite Entamoeba histolytica, the causative agent of amebiasis, l-cysteine plays an essential role in proliferation, adherence, and defense against oxidative stress; however, the essentiality of this amino acid in the pathways it regulates is not well understood. In the present study, we applied capillary electrophoresis time-of-flight mass spectrometry to quantitate charged metabolites modulated in response to l-cysteine deprivation in E. histolytica, which was selected as a model for examining the biological roles of l-cysteine. l-Cysteine deprivation had profound effects on glycolysis, amino acid, and phospholipid metabolism, with sharp decreases in the levels of l-cysteine, l-cystine, and S-adenosylmethionine and a dramatic accumulation of O-acetylserine and S-methylcysteine. We further demonstrated that S-methylcysteine is synthesized from methanethiol and O-acetylserine by cysteine synthase, which was previously considered to be involved in sulfur-assimilatory l-cysteine biosynthesis. In addition, l-cysteine depletion repressed glycolysis and energy generation, as it reduced acetyl-CoA, ethanol, and the major nucleotide di- and triphosphates, and led to the accumulation of glycolytic intermediates. Interestingly, l-cysteine depletion increased the synthesis of isopropanolamine and phosphatidylisopropanolamine, and it was confirmed that their increment was not a result of oxidative stress but was a specific response to l-cysteine depletion. We also identified a pathway in which isopropanolamine is synthesized from methylglyoxal via aminoacetone. To date, this study represents the first case where l-cysteine deprivation leads to drastic changes in core metabolic pathways, including energy, amino acid, and phospholipid metabolism.

and led to the accumulation of glycolytic intermediates. Interestingly, L-cysteine depletion increased the synthesis of isopropanolamine and phosphatidylisopropanolamine, and it was confirmed that their increment was not a result of oxidative stress but was a specific response to L-cysteine depletion. We also identified a pathway in which isopropanolamine is synthesized from methylglyoxal via aminoacetone. To date, this study represents the first case where L-cysteine deprivation leads to drastic changes in core metabolic pathways, including energy, amino acid, and phospholipid metabolism.
Sulfur-containing amino acids are essential for all living organisms from bacteria to higher eukaryotes and play indispensable roles in various cellular processes, such as methylation and the generation of polyamines, iron-sulfur clusters, and antioxidants. L-Cysteine in particular is essential for the structure, stability, and various protein functions, including catalysis, electron transfer, redox regulation, nitrogen fixation, and sensing for regulatory processes (1).
Entamoeba histolytica is an enteric protozoan parasite that causes hemorrhagic dysentery and extraintestinal abscesses in millions of inhabitants of endemic areas (2). This parasite is generally considered as anaerobic/microaerophilic and has been shown to consume oxygen and tolerate low levels of oxygen pressure but lacks most of the components of antioxidant defense mechanisms, such as catalase, peroxidase, glutathione, and the glutathione-recycling enzymes glutathione peroxidase and glutathione reductase (3,4). L-Cysteine, which replaces glutathione as a major thiol in E. histolytica, is synthesized via a sulfur assimilatory de novo cysteine biosynthetic pathway (5)(6)(7)(8)(9) that is typically present in bacteria and plants. This pathway consists of two steps that are catalyzed by serine acetyltransferase (SAT, EC 2.3.1.30) 6 (7,8) and cysteine synthase (CS; OAS (thiol) lyase; EC 4.2.99.8) (5). In addition to the presence of prokaryotic/plant-like L-cysteine biosynthesis, E. histolytica is also unique because the forward and reverse trans-sulfuration pathways are absent and interrupted, respectively. Furthermore, through lateral gene transfer from archaea, E. histolytica has acquired methionine ␥-lyase (EC 4.4.1.11), an enzyme that degrades L-methionine, L-homocysteine, and L-cysteine (10 -12). Thus, although typical parasitic protists show degenerated amino acid metabolic pathways, particularly those associated with catabolism, because of the parasitic lifestyle, sulfur-containing amino acid metabolism appears to have uniquely evolved in E. histolytica. However, the specific role of this pathway in this organism remains unclear.
L-Cysteine is the principal low molecular weight thiol in E. histolytica and is involved in the survival, growth, attachment, elongation, motility, gene regulation, and antioxidative stress defense of this organism (13)(14)(15)(16)(17). Because sulfur-containing amino acid metabolism differs significantly between E. histolytica and its mammalian host, the molecular dissection and characterization of this pathway may lead to the development of new chemotherapeutics against this parasite (18).
Here, to gain further insight into the roles and regulatory mechanisms of sulfur-containing amino acid metabolism and individual metabolites in E. histolytica, we utilized capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS) (19 -21) for the metabolomic profiling of this parasite. We observed drastic changes in the metabolome as a result of L-cysteine depletion, which led to the discovery of novel L-cysteine-mediated regulation of several metabolic pathways in E. histolytica.

EXPERIMENTAL PROCEDURES
Chemicals and Reagents-All of the chemicals of analytical grade were purchased from either Wako or Sigma-Aldrich unless otherwise mentioned. 2Ј,7Ј-Dichlorodihydrofluorescein di-acetate (2Ј,7Ј-DCF-DA) was purchased from Invitrogen. 1-Aminoacetone hydrochloride was obtained from United States Biologicals. High performance thin layer chromatography silica gel 60 plates were purchased from Merck.
[U-13 C 5 , 15 N]L-Methionine and [U-13 C 3 , 15 N]L-serine were purchased from Cambridge Isotope Laboratories. Stock solutions of metabolite standards (1-100 mmol/liter) for CE-MS analysis were prepared in either Milli-Q water, 0.1 mol/liter HCl, or 0.1 mol/liter NaOH. A mixed solution of the standards was prepared by diluting stock solutions with Milli-Q water immediately before CE-TOFMS analysis.
Microorganisms and Cultivation-Trophozoites of the E. histolytica clonal strain HM-1: IMSS cl 6 were maintained axenically in Diamond's BI-S-33 medium at 35.5°C, as described previously (22,23). Trophozoites were harvested in the late logarithmic growth phase 2-3 days after the inoculation of medium with one-thirtieth to one-twelfth of the total culture volume.
Metabolic Labeling and Metabolite Extraction-E. histolytica trophozoites were cultivated in either standard BI-S-33 medium containing 8 mM L-cysteine or L-cysteine-deprived medium for 48 h. For the metabolic labeling, trophozoites were cultured in the presence of either 3 mM stable isotope-labeled [U-13 C 5 , 15 N]L-methionine or 6 mM [U-13 C 3 , 15 N]Lserine in L-cysteine-deprived medium for 48 h as described above. To extract metabolites, ϳ1.5 ϫ 10 6 cells from each condition were harvested and washed twice with 5% mannitol. The cells were then suspended in 1.6 ml of methanol containing 16 M of each internal standard, 2-(N-morpholino)ethanesulfonic acid, methionine sulfone, and D-camphor-10-sulfonic acid and mixed with 1.6 ml of chloroform and 640 l of deionized water. After vortexing, the mixture was centrifuged at 4,600 ϫ g at 4°C for 5 min. The aqueous layer (1.6 ml) was filtrated using an Amicon Ultrafree-MC ultrafilter (Millipore Co.) and centrifuged at 9,100 ϫ g at 4°C for ϳ2 h. The filtrate was dried and preserved at Ϫ80°C until mass spectrometric analysis (24). Prior to the analysis, the sample was dissolved in 20 l of deionized water containing reference compounds (200 mol/liter each of 3-aminopyrrolidine and trimesic acid).
Instrumentation and CE-TOFMS Conditions-CE-TOFMS was performed using an Agilent CE capillary electrophoresis system equipped with an Agilent 6210 time-of-flight mass spectrometer, Agilent 1100 isocratic HPLC pump, Agilent G1603A CE-MS adapter kit, and Agilent G1607A CE-ESI-MS sprayer kit (Agilent Technologies, Waldbronn, Germany). The system was controlled by Agilent G2201AA ChemStation software for CE. Data acquisition was performed by Analyst QS software for Agilent TOF (Applied Biosystems and MDS Sciex).
CE-TOFMS Conditions for Cationic Metabolite Analysis-Cationic metabolites were separated in a fused silica capillary (50-m inner diameter ϫ 100-cm) filled with 1 mol/liter formic acid as the reference electrolyte (25). Sample solution (ϳ3 nl) was injected at 50 mbar for 3 s, and a positive voltage of 30 kV was applied. The capillary and sample trays were maintained at 20°C and below 5°C, respectively. Sheath liquid composed of methanol/water (50% v/v) that contained 0.1 mol/liter hexakis (2,2-difluorothoxy)phosphazene was delivered at 10 l/min. ESI-TOFMS was operated in the positive ion mode. The capillary voltage was set at 4 kV, and a flow rate of nitrogen gas (heater temperature, 300°C) was set at 10 p.s.i. For TOFMS, the fragmenter voltage, skimmer voltage, and octapole radio frequency voltage (Oct RFV) were set at 75, 50, and 125 V, respectively. An automatic recalibration function was performed using two reference masses of reference standards; protonated [ 13 C]methanol dimer (m/z 66.063061) and protonated hexakis (2,2-difluorothoxy)phosphazene (m/z 622.028963), which provided the lock mass for exact mass measurements. Exact mass data were acquired at the rate of 1.5 cycles/s over a 50 -1,000 m/z range.
CE-TOFMS Data Processing-Raw data were processed using the in-house software Masterhands (28). The overall data processing flow consisted of the following steps: noise filtering, baseline removal, migration time correction, peak detection, and integration of peak area from a 0.02 m/z-wide slice of the electropherograms. This process resembled the strategies employed in widely used data processing software for LC-MS and GC-MS data analysis, such as MassHunter (Agilent Technologies) and XCMS (29). Subsequently, accurate m/z values for each peak were calculated by Gaussian curve fitting in the m/z domain, and migration times were normalized using alignment algorithms based on dynamic programming (19,30). All of the target metabolites were identified by matching their m/z values and normalized migration times with those of standard compounds in the in-house library.
Quantitation of Reactive Oxygen Species-Fluorescence spectrophotometry was used to measure the production of intracellular reactive oxygen species using 2Ј,7Ј-DCF-DA as a probe as previously described (31). Briefly, E. histolytica cells were washed in PBS, and 5.0 ϫ 10 5 cells were then incubated in 1 ml of PBS containing 20 M 2Ј,7Ј-DCF-DA for 30 min at 35.5°C in the dark. The intensity of fluorescence was immediately read at excitation and emission wavelengths of 492 and 519 nm, respectively.
L-Cysteine/SMC Synthase Assay-L-Cysteine/SMC synthase was assayed by measuring acetate production through the coupling reaction of this enzyme with acetate kinase, pyruvate kinase, and lactate dehydrogenase. Acetate kinase generates ADP and acetyl-phosphate from acetate and ATP. The ADP production was coupled with the oxidation of NADH (⑀ 340 ϭ 6.22 mM Ϫ1 cm Ϫ1 ) through pyruvate kinase and lactate dehydrogenase (32). The standard reaction mixture contained 50 mM of Tris-Cl, pH 8.0, 3 mM OAS, 3 mM sodium sulfide or sodium methanethiolate, 4 units each of acetate kinase, pyruvate kinase, and lactate dehydrogenase, 0.5 mM ATP, 0.3 mM NADH, and 1-2 g of recombinant cysteine synthase. The reactions were initiated by the addition of recombinant cysteine synthase, and optical absorbance was read at 340 nm on a Shimadzu spectrophotometer. Kinetic parameters were determined using various concentrations (0.1-6 mM) of sodium sulfide, sodium methanethiolate, and OAS. The kinetic parameters were estimated using the nonlinear regression function obtained from the GraphPad Prism software (GraphPad Software Inc., San Diego, CA).
Choline and Ethanol Quantitation-The amount of choline (Cho) in the metabolite extracts was quantitated enzymatically using components of the Amplex Red sphingomyelinase assay kit (Invitrogen). Briefly, Cho was first oxidized by Cho oxidase to betaine and hydrogen peroxide. The produced hydrogen peroxide was then reacted with Amplex Red reagent in a 1:1 stoichiometry in the presence of horseradish peroxidase to generate the highly fluorescent product resorufin, which was read in a fluorescence spectrophotometer (model F-2500; Hitachi) at excitation and emission wavelengths of 545 and 590 nm, respectively. Ethanol production by trophozoites cultured in either normal or L-cysteine-deprived medium was determined as described previously (33).
Extraction of Lipids, Thin Layer Chromatography, and Phospholipid Quantitation-Cells cultured in either normal or cysteine-deprived medium for 48 h were collected by centrifugation, and lipids were then extracted by the Bligh and Dyer's method (34). The extracted lipids were analyzed by two-dimensional high performance thin layer chromatography using a solvent system of chloroform:methanol:28% ammonium hydroxide (65:25:5 v/v/v) in the first direction and chloroform:acetone:methanol:Acetic acid:water (50:20:10:10: The phosphorus content of phospholipids was determined after scraping representative spots from the plate, as described previously (35). The lipids were visualized by exposing TLC plates to iodine vapor.

L-Cysteine Deprivation Caused Accumulation of O-Acetylserine and S-Methylcysteine-
We first verified that L-cysteine deprivation affected intracellular L-cysteine/L-cystine concentrations in E. histolytica. Under normal culture conditions (8 mM L-cysteine), the intracellular concentrations of L-cysteine and L-cystine were 431 Ϯ 52 and 202 Ϯ 40 pmol, respectively, per 2 ϫ 10 5 cells. Approximately two-thirds (68 Ϯ 7%) of Lcysteine/L-cystine was present in a reduced form, whereas the remaining third was present in an oxidized form. Upon Lcysteine deprivation for 48 h, both L-cysteine and L-cystine decreased to nearly undetectable levels (88 Ϯ 11 and 79 Ϯ 10% decrement, respectively) (Fig. 1A). These results suggest that the intracellular L-cysteine/L-cystine concentrations in E. histolytica are greatly affected by the composition of the extracellular milieu.
We also examined whether oxidative stress induced by paraquat affected intracellular L-cysteine/L-cystine concentrations. Treatment of the amebae with 2 mM paraquat for 10 h led to 60.6 Ϯ 8.8 and 41.4 Ϯ 7.3% decreases in the levels of L-cysteine and L-cystine, respectively. Under conditions of L-cysteine limitation, the intracellular levels of reactive oxygen species increased by Ͼ4-fold, which was comparable with the 3.3-fold increase observed in paraquat/air-treated cells (Fig.  1B). These results suggest that L-cysteine may be an important scavenger of reactive oxygen species in E. histolytica.
Among the ϳ90 intermediary metabolites that were measured by CE-TOFMS-based metabolomic analysis, which include amino acids, organic acids, and nucleotides (19 -21), L-cysteine depletion caused drastic changes in the metabolites of E. histolytica involved in sulfur-containing amino acid metabolism (Fig. 1C). L-Cysteine depletion resulted in a sharp increase in O-acetylserine (OAS) (nearly undetectable under normal conditions), an activated form of L-serine that is synthesized from L-serine and acetyl-CoA by SAT. We also ob-served a marked increase (nearly undetectable under normal conditions) in S-methylcysteine (SMC), which is suggested to be a storage compound for sulfide and methyl groups in plants (36). L-Cysteine deprivation also caused a 44 Ϯ 6% decrement in the level of S-adenosylmethionine (SAM), whereas the level of L-methionine remained unchanged. SMC can be formed by the methylation of L-cysteine using either SAM or S-methylmethionine as a methyl group donor or by the transfer of the alanyl moiety of OAS to methanethiol (CH 3 SH) by CS (37). To differentiate between these possibilities, we performed metabolic labeling of E. histolytica trophozoites with stable isotope U-13 C, 15 N-labeled L-serine and Lmethionine in normal and L-cysteine-depleted media for 48 h. Upon the addition of [U-13 C 3 , 15 N]Ser to the L-cysteine-depleted culture medium, comparable levels of [ 13 C 3 , 15 N]SMC and unlabeled SMC, derived from [U-13 C 3 , 15 N]OAS and un-labeled OAS, respectively, were detected ( Fig. 2A). Similarly, when trophozoites were cultured in the presence of [U-13 C 5 , 15 N]Met under the L-cysteine-deprived conditions, we also detected comparable levels of [ 13 C 1 ]SMC and unlabeled SMC (Fig. 2B). In contrast, under normal conditions, neither SMC nor OAS was detected after [U-13 C 3 , 15 N]Ser or [U-13 C 5 , 15 N]Met labeling (data not shown). Taken together, these data clearly indicate that SMC is not synthesized by SAM-or S-methylmethionine-dependent methylation of Lcysteine; rather, SMC is synthesized in E. histolytica from the backbone of Ser and thiomethyl group of methanethiol.
Surprisingly, [ 13 C 3 , 15 N]OAS was not incorporated into either L-cysteine or L-cystine (data not shown), or their levels were too low to be detected by CE-TOFMS. To determine whether the lack of OAS incorporation into L-cysteine was due to the low sulfide concentrations under the in vitro axenic growth conditions, we deprived trophozoites of L-cysteine for 45 h and then continued their culture in medium supplemented with 2 mM sulfide for a further 3 h. However, sulfide supplementation did not affect the level of L-cysteine, whereas the levels of SMC and OAS markedly decreased (90.6 Ϯ 3.4 and 84.8 Ϯ 7.7% decrement, respectively) compared with the unsupplemented medium (Fig. 2C). These data suggest that sulfide negatively regulates OAS and SMC synthesis and also imply that the pathway formally called the "L-cysteine biosynthetic pathway" is primarily involved in the synthesis of SMC, but not L-cysteine, at least under in vitro culture conditions.
In Vitro Examination of S-Methylcysteine Synthesis-To elucidate the enzyme(s) involved in the formation of SMC from OAS and methanethiol, we examined whether different CS isotypes could catalyze the synthesis of SMC. Among the three examined CS isotypes (EhCS1-3), two CS proteins (EhCS1 and EhCS2) are very similar (99% amino acid identity, with two conserved amino acid changes) (5, 6), whereas EhCS3 shares only 83% amino acid identity with the other two isotypes. Both recombinant EhCS1 and EhCS3 efficiently catalyzed the synthesis of SMC using OAS and methanethiol as substrates. As revealed from the kinetic parameters ( Fig.  2D), EhCS1 and EhCS3 did not show any preference for either methanethiol or sulfide, because the K m , V max , k cat , and k cat /K m values for both of these substrates were comparable.
L-Cysteine Depletion Repressed Glycolysis and Energy Generation-Similar to other anaerobic and microaerophilic parasitic protozoa, such as Giardia lamblia and Trichomonas vaginalis, E. histolytica lacks features of aerobic eukaryotic metabolism, including the TCA cycle and oxidative phosphorylation, and primarily generates energy by substrate level phosphorylation (10). The CE-TOFMS-based metabolomic analysis demonstrated that L-cysteine depletion affected the levels of the majority of metabolites involved in glycolysis and its associated pathways (Fig. 3). L-Cysteine-depleted amebae generally contained higher amounts of glycolytic intermediates, with the exception of acetyl CoA and ethanol, than cells cultured under normal conditions. The largest changes caused by L-cysteine depletion were the increment in the levels of glycerol-3-phosphate (2.  DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50 and fumarate (1.60 Ϯ 0.20-fold). Several other metabolites involved in glycolysis, including glucose 6-phosphate, glucose 1-phosphate, fructose 6-phosphate, and phosphoenolpyruvate also showed slightly elevated levels (1.2-1.5-fold), whereas the levels of fructose 1,6-bisphosphate and dihydroxyacetonephosphate remained unchanged. In contrast to the significant increases in the glycolytic intermediates upstream of pyruvate in amebae cultured under L-cysteine-limited conditions, we observed reduced levels of acetyl CoA (29.4 Ϯ 7.1%) and ethanol (40.7 Ϯ 6.7%), suggesting a decrease in glycolytic flux and ATP generation by L-cysteine depletion. A number of other metabolites downstream of acetyl CoA, such as N-acetyl-glutamate, N-acetyl ␤-alanine, N-acetyl-leucine, and N-acetylphenylalanine, were also decreased (supplemental Fig. S1), supporting the premise that the glycolytic flux downstream of pyruvate was repressed.

Response of E. histolytica to L-Cysteine Depletion
Because glycolysis is the major source of energy generation in E. histolytica, a reduced glycolytic flux was thought to result in a decrement in the energy storage molecules of the trophozoites. As expected, the levels of the nucleotide triphosphates ATP, GTP, UTP, and CTP were significantly lower (p Յ 0.05) in the L-cysteine-depleted cells than in the trophozoites maintained under normal conditions (Fig. 3). We also observed slight deceases in the levels of ADP and GDP, whereas the levels of AMP and GMP were unchanged (Fig. 3).
L-Cysteine Depletion Altered Amino Acid Pools-Because amino acids are also used for energy production in E. histolytica (38), we examined the effects of L-cysteine deprivation on amino acid levels (supplemental Fig. S1). Next to L-cysteine and L-cystine, L-threonine and L-serine were the most highly modulated by L-cysteine depletion (1.63 Ϯ 0.25-and 2.07 Ϯ 0.29-fold increases, respectively) among the 20 amino acids. In E. histolytica, L-threonine and L-serine are catabolized by threonine dehydratase (39) to yield 2-oxobutyrate and pyruvate, respectively, which are in turn used by pyruvate:ferredoxin oxidoreductase for energy generation (40). L-Cysteine depletion also resulted in a slight increase in the intracellular concentration of L-alanine, which is synthesized from pyruvate by L-alanine:2-oxoglutarate aminotransferase (EHI_096750 (EAL50292.1) and EHI_159710 (EAL44861.1)). The levels of the remaining amino acids were not significantly affected by L-cysteine depletion.
L-Cysteine Depletion Caused Increases in Isopropanolamine, Aminoalcohol Phosphates, and Phosphatidylisopropanolamine-The metabolomic analysis of E. histolytica also revealed that L-cysteine depletion caused marked changes in amino alcohol metabolism (Fig. 4A). L-Cysteine depletion led to a dramatic increase in the levels of isopropanolamine (1aminopropan-2-ol, Ispn) (5.44 Ϯ 0.76-fold) and isopropanolamine phosphate (Ispn-P, undetected under normal conditions) (Fig. 4A). In addition, trophozoites cultured in L-cysteine-limited conditions showed 7.01 Ϯ 1.38-and 2.8 Ϯ 0.21-fold increases in ethanolamine phosphate (Etn-P) and choline phosphate (Cho-P) levels, respectively, whereas the levels of ethanolamine (Etn) and Cho were unchanged. Both Etn-P and Cho-P are intermediates in the Kennedy pathway, where phospholipids, including phosphatidylethanolamine and phosphatidylcholine, are produced.
Because L-cysteine limitation affected Ispn-P, Etn-P, and Cho-P concentrations, we next investigated whether L-cysteine deprivation influenced phospholipid synthesis by performing lipid profiling of amebic trophozoites cultured under L-cysteine-deprived or normal conditions using two-dimensional TLC (Fig. 4B). We found that in the absence of L-cysteine, E. histolytica synthesized an unconventional phospholipid that was verified to be phosphatidylisopropanolamine (PtdIspn) and was undetectable under normal conditions. Quantitation of individual lipids indicated that phosphatidylethanolamine (PtdEtn) decreased by 39.9 Ϯ 6.9%, whereas other phospholipids, such as phosphatidylcholine (PtdCho), phosphatidylserine, phosphatidylinositol, and phosphatidic acid, were unchanged (Fig. 4C). These data are consistent with the premise that PtdIspn was formed in a competition for the formation of PtdEtn, the level of which decreased by approximately the identical amount that PtdIspn increased (Fig. 4C). To further demonstrate that PtdIspn was formed from Ispn, E. histolytica trophozoites were cultured in normal medium containing 5 mM Ispn for 24 h. Under this condition, trophozoites produced an appreciable amount of PtdIspn (Fig.  4B, panel c).
As described above, L-cysteine depletion increased the level of reactive oxygen species. We therefore examined whether oxidative stress caused the observed changes in amino alcohols and phospholipids. It was observed that the lipid profiling of E. histolytica trophozoites cultured with 2 mM paraquat in ambient air for 10 h did not increase PtdIspn (Fig. 4B, panel  e). Furthermore, the addition of D-cysteine to the L-cysteinelacking medium did not reverse the effects of L-cysteine deprivation on the phospholipid profiles (Fig. 4B, panel d). These results confirmed that the generation of PtdIspn caused by L-cysteine depletion was not a result of oxidative stress but represents a specific response to L-cysteine deprivation.
Examination of Isopropanolamine Biosynthesis in E. histolytica-Next, we investigated the synthesis route of Ispn in E. histolytica. From studies of Escherichia coli, it is known that Ispn is synthesized from 1-aminoacetone by the action of Ispn:NAD ϩ oxidoreductase (41). 1-Aminoacetone is formed by the breakdown of L-threonine by L-threonine dehydrogenase (42) or is alternatively synthesized from methylglyoxal by monoamine oxidase, which catalyzes the interconversion of methylglyoxal and aminoacetone (43). Methylglyoxal is a byproduct of several metabolic pathways, with glycolysis being the most important source (44). Methylglyoxal is synthesized either enzymatically or nonenzymatically from dihydroxyacetone phosphate or glyceraldehyde 3-phosphate (44).
To examine the Ispn synthesis pathway in E. histolytica, we cultured amebae in medium supplemented with either methylglyoxal, aminoacetone, or L-threonine and examined the resulting lipid profiles. We found that supplementation with 2 mM methylglyoxal or 4 mM aminoacetone, but not 50 mM Lthreonine, led to the synthesis of PtdIspn (Fig. 5A). These results are consistent with the premise that E. histolytica is capable of Ispn synthesis from methylglyoxal and possesses the  DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50 JOURNAL OF BIOLOGICAL CHEMISTRY 39165 enzymatic activities of monoamine oxidase and Ispn:NAD ϩ oxidoreductase (Fig. 5B).

Identification of SMC and OAS as the Major Metabolites
Increased upon L-Cysteine Deprivation-In the present study, using a CE-TOFMS-based approach (19 -21), we identified novel metabolic changes caused by L-cysteine deprivation in the anaerobic/microaerophilic protozoan parasite E. histolytica. The major advantages of CE-MS analysis include its extremely high resolution and ability to simultaneously quantify charged low molecular weight compounds (19 -21). We demonstrated that L-cysteine deprivation causes a dramatic accumulation of SMC and OAS (Fig. 1C). SMC is a sulfurcontaining amino acid that has never been detected in protozoa but is widely present in relatively large amounts in several legumes, where it is considered to serve as a sulfur storage compound (36,37). Using stable isotope-labeled L-serine and L-methionine, we showed that SMC is synthesized from these amino acids in E. histolytica via OAS and methanethiol, respectively, which is similar to the pathway reported in A. thaliana (36). Interestingly, the increase in both SMC and OAS was mitigated by supplementation of the culture medium with 2 mM sulfide. These results have solved one enigma concerning the biological roles of the sulfur assimilatory de novo L-cysteine biosynthetic pathway in E. histolytica.
Role of L-Cysteine Biosynthetic Pathway-Although E. histolytica is a unique organism that constitutively expresses high Asterisks (*, **, and ***) denote statistically significant differences with p Յ 0.05, p Յ 0.01, and p Յ 0.001, respectively, as determined by Student's t test. B, profiles of phospholipids derived from trophozoites cultured under various conditions (panels a-e), analyzed by two-dimensional TLC. Trophozoites were cultured in normal culture medium (panels a, c, and e) or L-cysteine-deprived medium supplemented (panels b and d), without (panels a and b) or with 5 mM Ispn (panels c), or 8 mM D-cysteine (panels d) for 48 h or 2 mM paraquat for 10 h (panels e). Circles in the bottom left corners indicate the spots where the samples were applied. C, quantitation of the phospholipid species (percentage) derived from trophozoites cultured using normal (black bars) or L-cysteine-depleted (gray bars) medium. Asterisks (* and ***) denote statistically significant differences with p Յ 0.05 and p Յ 0.001, respectively, as determined by Student's t test. PtdSer, phosphatidylserine; PtdIns, phosphatidylinositol; PA, phosphatidic acid. levels of multiple cytosolic isotypes of CS and SAT, the physiological significance of the L-cysteine pathway and its redundancy are not well understood (8,9,18). In vitro cultivation of amebic trophozoites requires high concentrations of L-cysteine, which cannot be replaced by other thiols (16), indicating that the synthesis pathway may not be sufficient for the production of L-cysteine and might play an unknown role. Our metabolomic study using labeled L-serine did not support the hypothesis that L-cysteine is formed from L-serine and sulfide via OAS by the sequential action of SAT and CS, because labeled L-serine was not incorporated into L-cysteine (data not shown). We also demonstrated that amebic CS isotypes can catalyze the formation of SMC from OAS and methanethiol, unlike the CS from T. vaginalis (45), and also possess robust L-cysteine forming activity (Fig. 2D).
We also revealed that OAS is exclusively directed for the synthesis of SMC, but not L-cysteine, even in the presence of high concentrations of substrates. The apparent inability of E. histolytica to incorporate OAS into L-cysteine under L-cysteine-deprived conditions cannot be explained by the limiting concentration of sulfide under axenic culture conditions because the addition of sulfide did not increase L-cysteine levels, whereas the accumulation of OAS and SMC was immediately ceased by sulfide supplementation (Fig. 2C). In fact, the amebic trophozoites cultured under normal conditions contained appreciable concentrations of sulfide (134 M) (46). The lack of OAS incorporation into L-cysteine is also not attributable to the low substrate specificity of CS toward sulfide, because the K m values of CS isotypes for sulfide were comparable with those for methanethiolate (Fig. 2D). Thus, the preferred utilization of OAS by E. histolytica for SMC synthesis, but not for L-cysteine production, suggests that the L-cysteine biosynthetic pathway plays a major role in SMC production, whereas the apparent defect of L-cysteine production by this pathway in vivo remains puzzling.
Regulation of OAS Synthesis-The marked increase in OAS observed under L-cysteine deprivation is also worthy of attention. Similar to SMC, the level of OAS under normal culture conditions was nearly undetectable. Unlike other organisms, E. histolytica possesses three apparently functionally redundant, cytosolic SAT isozymes (SAT1-3) (9). Because these SAT isozymes have low to high sensitivity to feedback inhibition by L-cysteine, OAS and SMC were presumed to be formed even in the presence of high concentrations of L-cysteine, mainly by L-cysteine-insensitive SAT3. Thus, the fact that OAS and SMC were undetectable in the amebae cultured under normal conditions indicates that the activity of SAT, particularly SAT3, is repressed by unknown mechanisms. The fact that CS activity is a few orders of magnitude higher than SAT activity in the amebae may explain why OAS was not detected under the normal conditions, but it does not explain why SMC is not synthesized. The observed increase in OAS under L-cysteine depletion also indicates that L-cysteine-sensitive SAT1 and SAT2 are derepressed (i.e. L-cysteine-mediated feedback inhibition of SAT1/2 was reversed) under Lcysteine depletion.
In addition to the feedback inhibition, the cysteine biosynthetic pathway is also regulated by the bi-enzyme complex of SAT and CS (47). This complex is not involved in the metabolic channeling of OAS from SAT to CS, because OAS freely diffuses out of the complex. The formation of the SAT⅐CS complex (cysteine synthase complex) was shown to modulate the kinetic parameters of both enzymes. It was shown that the complex was dissociated by the elevated OAS levels (47). It is possible that OAS affects the formation and dissociation of the cysteine synthase complex in E. histolytica and the SMC but not L-cysteine forming activity of the complex. Although OAS also acts as an inducer of the L-cysteine regulon in bacteria (48), the gene expression of SAT and CS isotypes was not affected upon L-cysteine depletion in E. histolytica (data not shown).
Role and Fate of SMC-Because genes encoding other enzymes that utilize methanethiol as a substrate, such as O-acetylhomoserine sulfhydrylase (EC 2.5.1.49) and methanethiol oxidase (EC 1.8.3.4), are absent in the E. histolytica genome (10), SMC synthesis is likely the major salvaging pathway of methanethiol. SMC is present in relatively large amounts in several legumes, and there are a few lines of evidence demonstrating that the methyl moiety of SMC is incorporated into various metabolites (methionine, choline, and creatine) and proteins (pectin) (37, 49 -51). In plants (e.g. Brassica pekinensis), SMC can also be demethylated to generate L-cysteine (52). Metabolic labeling with SMC revealed that L-methionine and L-cysteine are formed from SMC in Neurospora crassa grown in low sulfur medium (53). In addition, methionine and cystathionine-auxotrophic mutants of N. crassa were able to grow when supplemented with SMC (53). Despite evidence from the studies, the fate of SMC in E. histolytica remains to be established because neither labeled Lcysteine nor L-methionine was detected using isotope-labeled serine in our metabolomic analysis. The role of methionine ␥-lyase in E. histolytica could be to generate methanethiol for the synthesis of SMC. It has been shown in other organisms that the methyl and thiomethyl moieties of SMC are transferred to unidentified metabolites or proteins (49 -53).
L-Cysteine Deprivation Affected SAM and Amino Acid Concentrations, Glycolysis, and Energy Generation-L-Cysteine depletion results in reduced levels of SAM, a precursor for polyamine biosynthesis and the essential methyl donor for numerous transmethylation reactions, including DNA methylation. This decrement in the SAM level is likely caused by either the reduction of SAM production by methionine adenosyltransferase or increased utilization of SAM. Because the amount of polyamines, such as putrescine, spermidine (supplemental Fig. S1), and N-acetylputrescine (data not shown), remained unchanged, SAM-dependent methylation may have increased upon L-cysteine deprivation. Because the methionine adenosyltransferase activity from various organisms is inhibited by nitric oxide-mediated nitrosylation of the cysteine residues in its active site (54), it is conceivable that methionine adenosyltransferase activity is inhibited by L-cysteine deprivation. The observed increase in L-threonine and L-serine can be attributed to their increased uptake, which is supported by the fact that L-cysteine is a strong inhibitor of Lthreonine and L-serine uptake in the BSC-1 epithelial cell line (55). DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50

JOURNAL OF BIOLOGICAL CHEMISTRY 39167
We also demonstrated that L-cysteine deprivation repressed glycolysis and energy generation. Upon L-cysteine depletion, pyruvate and other upstream glycolytic intermediates accumulated that appeared to be rerouted toward the associated pathways. For example, the metabolites linked to pyruvate and phosphoenolpyruvate (i.e. alanine, malate, and fumarate), 3-phosphoglycerate (i.e. O-phosphoserine), and dihydroxyacetone-phosphate (i.e. Gly 3-P) increased in response to L-cysteine depletion. In contrast, the level of acetyl-CoA, ethanol, and the major nucleotide triphosphates significantly decreased. In E. histolytica, pyruvate is utilized by pyruvate:ferredoxin oxidoreductase, a highly oxygen-sensitive iron-sulfur cluster-containing protein (56). Our data are consistent with the premise that L-cysteine depletion-mediated oxidative stress inactivates pyruvate:ferredoxin oxidoreductase and other redox-sensitive enzymes, which results in the overall reduction in the glycolytic flux and the accumulation of upstream glycolytic intermediates. It has been shown in E. histolytica that under oxidative or nitrosative stress, pyruvate, glucose 6-phosphate, and fructose 6-phosphate accumulate, whereas ethanol and ATP decrease (33,56). Unlike nitrosative stress, L-cysteine depletion for 48 h did not induce apoptosis (data not shown), and the decrease in ATP content appears to be primarily a result of the reduced glycolytic flux. Whole genome microarray analysis has revealed that L-cysteine depletion does not affect the expression of most of the genes involved in energy metabolism, with the exception of phosphoglycerate mutase and malate dehydrogenase, which were slightly down-regulated by 1.6-and 2.1-fold, respectively. 7 Down-regulation of these two genes may also contribute, at least in part, to the overall reduction in the glycolytic flux.
Discovery of Isopropanolamine and PtdIspn Synthesis upon L-Cysteine Deprivation-We have demonstrated for the first time that the Kennedy pathway, the major pathway for phospholipid biosynthesis, is regulated by the level of L-cysteine in E. histolytica. L-Cysteine deprivation resulted in the accumulation of an unusual phospholipid, PtdIspn, and also affected the composition and ratio of the major phospholipids. Under L-cysteine-depleted conditions, the synthesis of Ispn, Ispn-P, Etn-P, and Cho-P was elevated, PtdEtn synthesis was downregulated, and the levels of Etn, Cho, PtdCho, phosphatidylserine, phosphatidylinositol, and phosphatidic acid were unaffected (Fig. 4).
Based on the findings related to phospholipid biosynthesis, we propose the following scheme for the involvement of Lcysteine. When E. histolytica is cultured under L-cysteinedepleted conditions, Ispn synthesis is increased. Ispn-P, formed from Ispn and ATP in a reaction catalyzed by Etn/ Cho kinase (EHI_148580 (EAL52090.1); EHI_152340 (EAL51511.1)), then competes with Etn-P for Etn-P cytidyltransferase (EHI_095120 (EAL44415.1); EHI_140590 (EAL48799.1)), which appears to be the rate-limiting enzyme for the production of CDP-Ispn. This competition leads to an accumulation of Etn-P and a decrease in PtdEtn. The fact that PtdCho level was not affected by either L-cysteine depletion or Ispn supplementation, whereas Cho-P level increased upon L-cysteine depletion (Fig. 4A), suggests that Ispn-P does not compete with Cho-P for Etn-P/Cho-P cytidyltransferase. This observation also indicates that the increase in Cho-P is a consequence of increased production from accumulated Etn-P by SAM-dependent methylation, which may also contribute to the decrement in the SAM level. Alternatively, Ispn-P may compete with Cho-P for Etn-P/Cho-P cytidyltransferase, but the contribution of de novo synthesized PtdCho is negligible compared with the PtdCho incorporated from the culture milieu (Fig. 4).
Significance of PtdIspn Production upon L-Cysteine Deprivation-One of the consequences of the L-cysteine-dependent increase in Ispn synthesis is the concomitant increment in Etn-P, which is a known scavenger of free radicals (57). However, the increment of PtdIspn by L-cysteine depletion is not associated with either oxidative stress or changes in the redox status, because D-cysteine did not alleviate the PtdIspn synthesis, and paraquat/air treatment did not increase PtdIspn synthesis.
Similar to PtdEtn, phospahtidyl-propanolamine, an analog of PtdIspn, is a nonbilayer or hexagonal phase-forming phospholipid; however, it is not known whether PtdIspn is also similar to PtdEtn and phospahtidylpropanolamine in its nonbilayer or hexagonal phase-forming nature (58). Hexagonal phase-forming phospholipids have been proposed to be important for membrane fluidity, protein translocation, and membrane fusion events (59,60). In PtdEtn methylation-defective mutants of Saccharomyces cerevisiae, supplementation of the culture medium with propanolamine leads to phosphatidylpropanolamine production and thus presumably compensates for the role of PtdCho and its N-methylated phospholipid precursors (58). However, mitochondrial PtdEtn cannot be completely replaced by phosphatidylpropanolamine, suggesting that PtdEtn is essential for the structure and function of mitochondrial membranes (58). It has been shown that changes in the PtdCho/PtdEtn ratio affect membrane integrity of large unilamellar vesicles in mouse hepatocytes, and this ratio is inversely correlated with leakage across the membrane (61). Because L-cysteine depletion also increases the PtdCho/PtdEtn ratio, it is conceivable that this changes membrane integrity and fluidity and affects protein translocation across the plasma membrane. In addition, because PtdEtn also plays various metabolic roles in cells, a decrease in PtdEtn level may also affect other cellular processes, including the synthesis of GPI anchors and protein modification.
To date, this is the first report to show that PtdIspn synthesis is increased by changes in environmental conditions. Pt-dIspn and phosphatidylpropanolamine have been identified in various organisms, including yeast, protozoa, and animals, and are considered to be unnatural phospholipids synthesized only under conditions where Ispn or propanolamine are supplied in the culture medium (58,62) or administered intraperitoneally (63). Recently, PtdIspn has been shown to be naturally synthesized in BHK cells through the decarboxylation of the rare phospholipid phosphatidylthreonine (64).
In conclusion, we have demonstrated that L-cysteine regulates various metabolic pathways in E. histolytica and thus affects the concentrations of the amino acids, phospholipids, and intermediary metabolites involved in central energy metabolism. Further investigation on the physiological role and fate of SMC and PtdIspn will help to better understand sulfurcontaining amino acid metabolism, which is considered an attractive drug target for the development of new chemotherapeutics against this pathogen (18,65). Future research is also needed to understand the function of PtdIspn in the plasma membrane and membrane-bound organelles and in the regulation of phospholipid metabolism.