The de novo L-cysteine biosynthetic pathway is critical for the growth, antioxidative stress defenses, and pathogenesis of bacterial and protozoan pathogens, such as Salmonella typhimurium and Entamoeba histolytica. This pathway involves two key enzymes, serine acetyltransferase (SAT) and cysteine synthase (CS), which are absent in mammals and therefore represent rational drug targets. The human parasite E. histolytica possesses three SAT and CS isozymes; however, the specific roles of individual isoforms and significance of such apparent redundancy remains unclear. In the present study, we generated E. histolytica cell lines in which CS and SAT expression was knocked down by transcriptional gene silencing. The strain in which CS1, 2 and 3 were simultaneously silenced and the SAT3 gene-silenced strain showed impaired growth when cultured in a cysteine lacking BI-S-33 medium, whereas silencing of SAT1 and SAT2 had no effects on growth. Combined transcriptomic and metabolomic analyses revealed that, CS and SAT3 are involved in S-methylcysteine/cysteine synthesis. Furthermore, silencing of the CS1-3 or SAT3 caused upregulation of various iron-sulfur flavoprotein genes. Taken together, these results provide the first direct evidence of the biological importance of SAT3 and CS isoforms in E. histolytica and justify the exploitation of these enzymes as potential drug targets.
Critical metabolic pathways that are unique to pathogens and are significantly divergent from their hosts are rational targets for the development of new chemotherapeutic agents. In particular, sulfur-containing amino acid metabolism, particularly the de novo L-cysteine biosynthetic pathway, is a promising target for drug development against bacterial and parasitic infections, such as those caused by Mycobacterium tuberculosis, Salmonella typhimurium, and Entamoeba histolytica 1,2,3,4,5,6.
Amebiasis is an intestinal infection caused by the protozoan pathogen E. histolytica and is widespread worldwide [CDC, https://www.cdc.gov/parasites/amebiasis/index.html], particularly in countries with inadequate sewage treatment and poor water quality7. According to the WHO, an estimated 50 million people are infected with E. histolytica worldwide, resulting in 40,000–100,000 deaths annually8. Metronidazole is the drug of choice for treating amebiasis despite its low efficacy against asymptomatic cyst carriers1. Moreover, metronidazole is also teratogenic and causes adverse side effects, such as nausea, vomiting, headache, insomnia, dizziness, drowsiness and hypersensitivity reactions (urticaria, pruritus, erythematous rash)9. In addition, E. histolytica is capable of tolerating sub-therapeutic levels of metronidazole in vitro 10,11. Therefore, new drugs that target parasite-specific metabolic pathways and enzymes distinct from those targeted by metronidazole are urgently needed.
Sulfur-containing amino acid metabolism in E. histolytica differs markedly from that in humans with respect to three main features: (i) the absence of forward and reverse transsulfuration pathways and thus does not convert L-methionine to L-cysteine12,13 or vice versa; (ii) the presence of a sulfur-assimilatory de novo L-cysteine biosynthetic pathway14,15,16; and (iii) the presence of a unique enzyme, methionine γ-lyase (MGL), which is involved in the degradation of sulfur-containing amino acids17,18,19. As MGL and two enzymes involved in the cysteine biosynthetic pathway, serine O-acetyltransferase (SAT) and cysteine synthase (CS, O-acetylserine sulfhydrylase), are absent in mammals, these enzymes are potential suitable targets for chemotherapeutic agents against amebiasis.
The cysteine biosynthetic pathway plays an important role in the incorporation of inorganic sulfur into organic compounds1 and has been extensively studied in bacteria, plants, and protozoa20,21,22,23,24. In this pathway, SAT (EC 188.8.131.52) catalyzes the formation of O-acetyl-L-serine (OAS) from L-serine and acetyl-CoA15,16 (Fig. 1A). CS [O-acetyl-L-serine (thiol)-lyase] (EC 184.108.40.206) then catalyzes the production of L-cysteine/S-methylcysteine (SMC) through the modification of sulfide/methanthiol with the alanyl moiety of O-acetylserine13,14. However, using a metabolomics approach, we previously showed that CS enzymes in E. histolytica trophozoites cultured in the absence of exogenous L-cysteine are predominantly involved in SMC formation, but not L-cysteine13. E. histolytica SAT and CS have several unique features with respect to localization, complex formation and homology. For example, isozymes of SAT (EhSAT1-3) and CS (EhCS1-3) are localized to the cytosol14,15,16, whereas plant isoforms of SAT and CS are found in the mitochondria, plastids, and cytosol25. In addition, EhCS1 and EhSAT1 do not form a heteromeric complex26, whereas bacterial and plant SAT and CS form complexes that are involved in cross-talk between sulfur assimilation, carbon and nitrogen metabolism via the generation of OAS27. Further, EhSAT1-3 are biochemically divergent, showing 48–73% mutual sequence identity (Fig. 1B) and markedly different sensitivities to allosteric feedback by L-cysteine16. EhCS1-3 also exhibit sequence divergence with CS1 and CS2 being identical with the exception of two amino acid changes and CS3 having 83% amino acid identity with CS1 and CS2 (Fig. 1B).
Although the sulfur-assimilatory cysteine biosynthetic pathway in plants, bacteria, and protozoa has been extensively studied and exploited for drug development, the role of individual SAT and CS isozymes and significance of the apparent redundancy of this pathway in E. histolytica remain to be elucidated. In the present study, we investigated the role of the cysteine biosynthesis pathway in E. histolytica using parasites in which genes for the enzymes involved in cysteine biosynthesis were silenced by antisense RNA-mediated transcriptional attenuation. Using transcriptomic and metabolomic analyses, we demonstrated that EhCS and EhSAT3 are critical for SMC/cysteine production and cell growth. Furthermore, we examined the fate of SMC unique metabolite in E. histolytica and revealed that this unique metabolite is involved in the antioxidative stress mechanism.
Establishment of CS and SAT gene-silenced strains
To investigate the role of the L-cysteine biosynthesis pathway in E. histolytica, we utilized antisense small RNA-mediated epigenetic gene silencing to repress the CS1/2 (CS1 and CS2 are 99% identical at the amino acid level), CS3, SAT1, SAT2, and SAT3 genes in E. histolytica strain G3 (Fig. 1C)28,29. In the CS1/2 and CS3 gene-silenced strains, CS1/2 and CS3 gene expression were simultaneously repressed, likely due to the high sequence similarity (83% amino acid identity) between these genes (Fig. 1B). Similarly, the SAT1 and SAT2 genes, whose products share 73% amino acid identity, were simultaneously silenced in the SAT1 and SAT2 gene-silenced transformants (Fig. 1B), whereas the SAT3 gene was not silenced in either of these transformants because of low (48% at the amino acid level) identity between SAT1 and SAT3 and between SAT2 and SAT3. In the SAT3 gene-silenced strain, only the SAT3 gene was silenced (Fig. 1C), and neither SAT1 nor SAT2 was affected. In subsequent analyses, the SAT1/2 and CS1/3 gene-silenced transformants, designated SAT1/2gs and CSgs, respectively, were used for further analyses.
Effects of CS and SAT gene silencing on E. histolytica growth
To examine if the L-cysteine biosynthesis plays a role in the proliferation of E. histolytica, the growth kinetics of trophozoites of the gene-silenced and control transformants (cell line transfected with psAP2G plasmid) were compared in normal BI-S-33 medium containing 8 mM L-cysteine (Fig. 2A) or BI-S-33 medium without L-cysteine which we called as L-cysteine lacking medium (Fig. 2B). However this medium may still contain trace amounts of cysteine from yeast extract and/or tryptone. When cultured in L-cysteine lacking medium, CS gene-disrupted transformants displayed a severe growth defect, whereas SAT3 gene-disrupted transformants showed a mild growth defect (Fig. 2B). In contrast, SAT1/2gs transformants appeared to grow normally in L-cysteine lacking medium (Fig. 2B). However, in normal BI-S-33 medium, none of the gene-silenced strains showed defective growth (Fig. 2A). These results indicate that CS and SAT3 are essential for growth in the absence of exogenous L-cysteine and therefore contribute to cell proliferation.
Metabolomic analysis of gene-silenced transformants cultured in normal and L-cysteine lacking BI-S-33 medium
A total of 48 intermediary metabolites, including amino acids, nucleotides, and organic acids, were measured by CE-TOFMS-based metabolomic analysis in the SAT1/2gs, SAT3gs and CSgs transformants under different culture conditions (Supplementary Table S2). Silencing of the CS genes caused drastic changes in the metabolites involved in sulfur-containing amino acid metabolism (Fig. 3). Specifically, the L-cysteine concentration in CSgs trophozoites was approximately 60% lower than that in control trophozoites when cultured under normal BI-S-33 containing 8 mM L-cysteine and L-cysteine-lacking BI-S-33 medium, consistent with the speculation that CS is involved in L-cysteine production. CS gene silencing also resulted in a marked increase in OAS, an activated form of L-serine that is synthesized from L-serine and acetyl-CoA by SAT, in both normal and L-cysteine lacking conditions. In addition, SMC formation was completely abolished by CS gene silencing (Fig. 3), suggesting that CS enzymes are indispensable for SMC production.
In contrast to CS, silencing of SAT1/2 did not markedly alter the levels of sulfur-containing metabolites, particularly OAS, SMC, L-cysteine and L-methionine, in E. histolytica, suggesting that SAT3 can compensate for the loss of SAT1/2. However, upon silencing of the SAT3 gene, the levels of SMC and L-cysteine were decreased approximately 40–50% compared to the control strain despite the presence of high levels of the precursor metabolite OAS, which is formed by SAT1 and SAT2 in strain SAT3gs. The reduced level of SMC/cysteine with the concurrent higher OAS level (approximately 60% increase) in strain SAT3gs may be due to a decreased level of CS protein in strain SAT3gs. To investigate this possibility, we examined CS expression at the protein level in the SAT1/2gs, SAT3gs, and control transformant strains. Immunoblot analysis using anti-rEhCS1, anti-rEhCS3, and anti-rEhCPBF130 antibodies showed that the relative amounts of these proteins were comparable between these strains (Supplementary Fig. S1), suggesting that SAT3 may positively regulate CS activity, but not gene expression or protein stability, whereas SAT1/2 do not regulate CS activity.
Gene silencing of SAT1/2, SAT3, or CS1-3 caused global transcriptomic changes
To determine if the silencing of the CS and SAT genes affected the expression of other genes, global gene expression in the SAT1/2gs SAT3gs and CSgs transformants was analyzed using a whole-genome DNA microarray. However, the analysis revealed that after the removal of redundant or obsolete genes (those represented with probe sets with ‘_x_at’ and those for which corresponding NCBI entries were removed after genome reannotation)31, only a limited number of genes had three-fold or higher changes in expression (Supplementary Table S3).
In CSgs, 34 genes were up-regulated and 25 genes were down-regulated when compared to the control (Table 1). CS1-3 transcript levels were reduced by 104, 128 and 20.4 fold, respectively, in CSgs. Among the genes that were significantly down-regulated included those encoding for several hypothetical proteins (EHI_020830, EHI_196760, and EHI_066720), Rab family GTPase, RabH232 (EHI_128180), and a nonpathogenic pore-forming peptide precursor (EHI_169350), which may belong in the saposin-like protein31 (SAPLIP1) family. In contrast, Ras family GTPase (EHI_074750_at), methylene-fatty-acyl-phospholipid synthase (EHI_153710_at), and deoxyuridine 5′ triphosphate nucleotide hydrolase domain-containing protein (EHI_091670_at) were up-regulated in all three gene-silenced transformants (Supplementary Table S3), suggesting that the increased expression of these genes may compensate for the impairment of the cysteine biosynthetic pathway. The most highly upregulated gene related to sulfur metabolism in CSgs was a gene encoding a member of the NADPH-dependent FMN reductase domain-containing protein family (Table 1). Genes encoding NADPH-dependent oxidoreductase 2 (EHI_045340), which was previously shown to be involved in cystine reduction33, was also upregulated in CSgs strain (Table 1).
In strain SAT1/2gs, 39 genes were up-regulated and 13 genes were down-regulated compared to the control strain (Table 2). The SAT1 and SAT2 transcript levels were reduced by 529 and 4.2 fold, respectively, whereas the expression of the SAT3 gene remained unchanged. The genes encoding phosphoserine aminotransferase (EHI_026360), which catalyzes the formation of L-phosphoserine from 3-phosphohydroxypyruvate in the phosphorylated pathway of L-serine biosynthesis34, were down-regulated more than five fold (Table 2). Among the most highly upregulated genes was sulfotransferase (EHI_031640), which was up-regulated more than 8 fold, and Fe hydrogenase, which was induced more than 4 fold in strain SAT1/2gs (Table 2). In strain SAT3gs, 16 genes were up-regulated and 19 were down-regulated compared to the control (Table 3). The most highly repressed gene in SAT3gs was SAT3, which had 187-fold lower transcript levels compared the control, whereas SAT1 and SAT2 gene expression remained unchanged. Among the genes that were up-regulated by SAT3 gene silencing were several genes encoding NADPH-dependent FMN reductase domain-containing protein and iron-sulfur flavoprotein (ISF) genes, which were among the most highly up-regulated genes by CS3 gene silencing (Table 1).
Confirmation of differential gene expression by qRT–PCR
The microarray results were validated by qRT–PCR. Table 4 shows a comparison of the qRT-PCR and microarray data of six representative differentially expressed genes identified by the transcriptome analysis, with the RNA polymerase II gene used as reference33. The results of the qRT-PCR analysis agreed well with the microarray data for all examined gene transcripts (Table 4).
S-Methylcysteine production leads to increased oxidative stress tolerance
To investigate whether the SMC accumulation observed in trophozoites cultured in L-cysteine lacking BI-S-33 medium protects E. histolytica against oxidative stress, the CSgs, which does not produce SMC and control (harboring plasmid psAP2G) transformants were compared for oxidative stress sensitivity by culturing the two strains in L-cysteine lacking medium. After 48-h cultivation in L-cysteine lacking medium, SMC had accumulated in the control transformant, but remained absent in the CSgs transformant (Fig. 3). The CSgs and control transformants were next exposed to different concentrations of H2O2 (0–6.4 mM) for 1 h and viability was then determined. The CSgs transformant showed slightly, but significantly (Student’s t-test), lower sensitivity to 0.8–4.0 mM H2O2 compared to the control transformant, suggesting that CS1–3 or SMC synthesis may be involved in protecting the cells against oxidative stress (Fig. 4).
The identification and functional characterization of the molecular components involved in essential metabolic pathways contribute to the overall understanding of parasite biology, but also aid in the rational design of novel therapeutics. L-Cysteine is indispensable for the survival of virtually all living organisms and plays a major role in maintaining the redox balance of thiol compounds in microaerophiles18. The cysteine biosynthetic pathway exists in bacteria, plants, and several parasitic protozoa, including Leishmania major, Trypanosoma cruzi, and Trichomonas vaginalis 1, and enzymes involved in this pathway are suitable targets for the development of novel drugs to prevent disease caused by these parasites2,3,4,5,6.
In the present study, we investigated the specific role of individual isotypes of SAT and CS using a gene silencing approach. Although SAT26 and CS35 have been biochemically13,14,15,16 and structurally characterized, the specific role of individual SAT and CS isotypes in proliferation, pathogenesis, and parasitism remains to be elucidated. Although we attempted to silence the expression of single genes, all the three CS isotypes were simultaneously silenced in E. histolytica due to their high similarity at the nucleotide and amino acid levels (CS1-3, 83–99%). The global repression of CS expression resulted in impaired trophozoites proliferation in L-cysteine lacking BI-S-33 medium, but not in normal BI-S-33 medium containing 8 mM L-cysteine (Fig. 2B). Metabolomic analysis of the CS-silenced transformant under the two culture conditions further revealed that SMC was not detectable and that the level of OAS was markedly reduced, demonstrating that CS is involved in SMC synthesis from OAS (Fig. 3). We previously showed that E. histolytica trophozoites produce SMC, rather than L-cysteine, when cultured in L-cysteine lacking BI-S-33 medium13. The present metabolomic analysis further revealed that L-cysteine levels were also decreased when the CSgs (and SATgs) transformants were cultured in normal BI-S-33 medium containing 8 mM L-cysteine, suggesting that E. histolytica synthesizes both L-cysteine and SMC, and that the flux towards cysteine synthesis likely depends upon the intracellular availability of sulfides (i.e., methanethiol and sulfide).
In contrast to CS, silencing of SAT1/2 and SAT3 was specific to the targeted SAT isotypes. Unlike other organisms, E. histolytica possesses three apparently redundant SAT isozymes16. These three SAT isotypes differ from one other in their regulatory properties. The isoenzymes SAT1 and SAT2 are regulated through allosteric feedback by L-cysteine15,16, whereas SAT3 is relatively insensitive to L-cysteine inhibition16. Consistent with these findings, EhSAT1-3 showed different levels of sensitivity to allosteric feedback by L-cysteine16 [inhibition constant (Ki) values of EhSAT1-3 are 4.7, 28, and 460 μM, respectively]. We previously showed that under cysteine lacking conditions, OAS and SMC expression levels in E. histolytica increase, whereas the expression of SAT and CS isotypes are not affected by L-cysteine depletion13,36. As OAS and SMC were undetectable under normal conditions, OAS, which is produced by SAT1-3, appears to be readily converted to cysteine, but not SMC. Alternatively, the in-vivo activities of SAT3 and cysteine-sensitive SAT1/2 may be repressed by unknown mechanisms. Under cysteine lacking conditions, L-cysteine-sensitive SAT1 and SAT2, together with cysteine-insensitive SAT3, were derepressed, leading to increased production of OAS. The mechanism by which SMC, but not cysteine, accumulates in response to cysteine deprivation in strains SAT1/2gs and SAT3gs remains unknown. However, it is conceivable that in strain SAT1/2gs, SAT3 compensates for the loss of SAT1/2 by producing sufficient cellular OAS and thereby contributes to the maintenance of high SMC levels under cysteine lacking conditions.
Another unique aspect of E. histolytica SAT1 is the lack of protein-protein interaction with CS26. It is well known that in bacteria and plants, CS and SAT form a heteromeric complex with a molecular mass of several hundred kilodaltons37. However, EhCS1 and EhSAT1 form a homodimer and homotrimer, respectively26, but these proteins do not interact under physiological conditions26. The lack of interaction between EhCS1 and EhSAT1 was structurally elucidated26,35. The apparent reduction of cysteine/SMC production in SAT3gs strain despite high level of OAS may explain the possible formation of a SAT3-CS complex that regulates cellular CS activity. Here, silencing of SAT3 resulted in the loss of complex formation, reduction of CS activity, and decreased production of SMC and L-cysteine (Fig. 3). Metabolomic analysis confirmed that the similar response occurred in the CS gene-silenced transformants. SAT3 possesses a unique 25–30 amino acids extension at the carboxyl terminus and has a low isoelectric point compared to SAT1 and SAT216. These features may favor the interaction with E. histolytica CS, particularly EhCS3, which possesses the highest pI (8.17) among the three CS isotypes. However, this hypothesis needs to be experimentally proven.
The present metabolomic analyses combined with the results of the growth kinetic assay demonstrated that neither the concentrations of OAS, L-cysteine or SMC in the two culture conditions, nor trophozoites growth under L-cysteine lacking conditions were affected by SAT1/2 gene silencing (Fig. 3). These data suggest that SAT3 is a robust enzyme that likely compensated for the loss of SAT1/2 under in-vitro conditions. In contrast to SAT1/2, repression of SAT3 had more marked effects on growth than the repression of SAT1/2, suggesting that SAT3 is critical for survival under stressful conditions, whereas SAT1/2 are involved in more general house-keeping roles. This speculation is also supported by the fact that the levels of both L-cysteine and SMC were decreased in strain SAT3gs.
To determine whether other genes, particularly those involved in sulfur metabolism, compensate for the loss of CS and SAT gene expression, we compared the transcriptomes of the CSgs, SAT1/2, SAT3gs, and control strains grown in normal medium. Notably, several genes from a family of the NADPH-dependent FMN reductase domain-containing proteins, also known as iron-sulfur flavoproteins (ISFs), which are commonly found in anaerobic prokaryotes, were highly upregulated in CSgs and SAT3gs (Tables 1 and 3). To date, the only eukaryotic species that have been found to possess ISF homologs are E. histolytica and Trichomonas vaginalis 38,39. A search of the genome database of E. histolytica revealed the presence of seven independent ISF genes40, which were previously shown to be upregulated in E. histolytica cells cultured in L-cysteine lacking BI-S-33 media, suggesting that these genes are regulated in response to L-cysteine deprivation36.
In contrast to CSgs strain, we found that in SAT1/2 gene-silenced strain one of the sulfotransferase, SULT9 (XP_653539, EHI_031640) (Table 2), was up-regulated more than eight fold, suggesting its involvement in L-cysteine biosynthesis and/or redox-related metabolism. The E. histolytica genome contains 10 genes that encode putative sulfotransferases (SULTs), which are localized in the cytosol and are involved in the production of sulfated molecules41. For example, SULT6 (XP_649714, EHI_146990) is responsible for synthesizing cholesteryl sulfate, an important compound for the encystation process in the Entamoeba life cycle41. However, the function of other SULTs in E. histolytica remains largely unknown. In Arabidopsis roots, a plasma membrane sulfate ion transporter (SULTR) physically interacts with CS to coordinate internalization of sulfate ions based on the energetic/metabolic state of root cells42. Here, we also determined that Fe hydrogenase, which belongs to a distinct class of hydrogen-producing metalloenzymes and is found in a wide variety of prokaryotes and eukaryotes43, was up-regulated more than four fold in strain SAT1/2gs. Fe hydrogenase contributes to the utilization of hydrogen as a growth substrate and for the disposal of excess electrons through combination with protons to form hydrogen43. Although the role of Fe hydrogenase in Entamoeba is unclear, it is possible that this enzyme is regulated in response to oxygen levels, as was shown in Chlamydomonas reinhardtii, which contained increased transcript levels of Fe hydrogenase upon shifting from an aerobic to anaerobic atmosphere44.
The present metabolome data of strain CSgs suggest that in addition to L-cysteine, CS enzymes are involved in SMC production (Fig. 3). SMC is a sulfur-containing amino acid that is found in relatively large amounts in several legumes, where it is considered to be a sulfur storage compound45. However, the fate and physiological significance of SMC in protozoa, particularly E. histolytica, is not yet fully understood. Previously, we investigated metabolic responses to hydrogen peroxide - and paraquat-mediated oxidative stress in E. histolytica trophozoites and reported that SMC levels are increased more than two fold under both stress conditions46, suggesting the involvement of this metabolite in the oxidative stress response. To confirm this speculation, we compared the oxidative stress tolerance between the control and CSgs transformant because SMC was undetected in CSgs strain (Fig. 3), and demonstrated that the CSgs transformant was more sensitive to oxidative stress. In Brassica exposed to H2O2 or O2 stress, SMC is non-enzymatically converted to SMC sulfoxide47, which is further enzymatically catabolized into pyruvate, ammonia, and alkylthiosulfinates48. The enzyme that catalyzes the last reaction is cystine lyase (EC 220.127.116.11) and behaves similarly to allinase (EC 18.104.22.168) in garlic, with the exception that cystine lyase also has the ability to cleave L-cystine49. Based on this observation, we propose that under oxidative stress conditions, SMC is converted to SMC sulfoxide and is further degraded by a lyase enzyme, such as methionine γ-lyase17 (MGL), to pyruvate and sulfenic acid.
In summary, the present metabolomic analysis revealed that CS and SAT3 are key enzymes for cysteine/SMC production in E. histolytica and are also essential for parasite survival under oxidative stress conditions. Transcriptomic analysis of the constructed CSgs and SAT3gs strains revealed that compensatory mechanisms in which ISFs play key roles operate under conditions where the CS and SAT3 pathway(s) are inactivated. These findings corroborate the metabolic and physiological importance of the L-cysteine pathway in E. histolytica and suggest that CS and SAT3 represent good targets for drug development. Further work is needed to demonstrate the specific role of these ISFs in E. histolytica.
Microorganisms and cultivation
Strain G3 and plasmid psAP2 were kindly provided by Dr. David Mirelman (Weisman Institute, Israel)28,29. Gene silencing was performed as previously described28,29,36,52. Briefly, 420-bp fragments containing the entire open reading frames of the E. histolytica CS1, CS3, SAT1 SAT2 and SAT3 genes starting at the initiation codon were amplified by PCR from cDNA using the oligonucleotide primers listed in Supplementary Table S1. The obtained PCR products were digested with StuI and SacI and inserted into StuI/SacI-digested pSAP2G to produce psAP2G-CS1, psAP2G-CS3, psAP2G-SAT1, psAP2G-SAT2 and psAP2G-SAT3. The constructed plasmids were introduced into E. histolytica strain G3 by liposome-mediated transfection15, and the resulting transformants (designated psAP2G [control], CS1gs, CS3gs, SAT1gs, SAT2gs and SAT3gs) were selected and maintained in normal BI-S-33 medium supplemented with 7 μg/ml geneticin (Invitrogen). The expression of the respective genes was confirmed by semi-quantitative RT-PCR as described previously using RNA polymerase II mRNA (GenBankTM accession number XM_643999) as a reference33, as the expression of this gene was invariant in all of the transformants. The transformants were designated psAP2G (control), CS1gs, CS3gs, SAT1gs, SAT2gs and SAT3gs.
Extraction of metabolites from E. histolytica
E. histolytica trophozoites were cultured for approximately 24 h in standard BI-S-33 medium containing 8 mM L-cysteine. The medium was replaced with either normal BI-S-33 medium or medium lacking L-cysteine13, and trophozoites were cultured for a further 48 h. To extract metabolites, approximately 1.5 × 106 cells were then harvested and immediately suspended in 1.6 mL of −75 °C methanol to quench metabolic activity. To minimize the effects of experimental artifacts, such as ion suppression, on metabolite levels, 2-(N-morpholino) ethanesulfonic acid, methionine sulfone, and D-camphor-10-sulfonic acid were added to each sample as internal standards13,53,54. The samples were sonicated for 30 s and then mixed with 1.6 mL chloroform and 640 µl deionized water. After vortexing, the mixed samples were centrifuged at 4600 g for 5 min at 4 °C. The aqueous layer (1.6 mL) was filtrated using an Amicon Ultrafree-MC ultrafilter (Millipore Co., Massachusetts, USA) and the collected sample was centrifuged at 9100 g at 4 °C for approximately 2 h. The filtrate was vacuum dried and stored at −80 °C until needed for mass spectrometric analysis55. Prior to the analysis, the sample was dissolved in 20 μl de-ionized water containing 200 μmol/L of two reference compounds (3-aminopyrrolidine and trimesic acid).
Instrumentation and capillary electrophoresis-time-of-flight mass spectrometry (CE-TOFMS)
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 using Analyst QS software for Agilent TOF (Applied Biosystems, CA, USA; MDS Sciex, Ontario, Canada).
CE-TOFMS conditions for cationic metabolite analysis
Cationic metabolites were separated in a fused-silica capillary (50 μm i.d. × 100 cm total length) filled with 1 mol/L formic acid as the reference electrolyte56. Sample solution (~3 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) and 0.1 μmol/L hexakis (2,2- difluorothoxy) phosphazene was delivered at 10 μL/min. ESI-TOFMS was operated in positive ion mode. The capillary voltage was set at 4 kV and the flow rate of nitrogen gas (heater temperature 300 °C) was set at 10 psig. 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 the masses of two reference compounds, protonated 13C 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 50 to 1,000 m/z.
CE-TOFMS conditions for anionic metabolite analysis
Anionic metabolites were separated in a cationic-polymer–coated COSMO(+) capillary (50 μm i.d. × 110 cm) (Nacalai Tesque) filled with a 50 mmol/L ammonium acetate solution (pH 8.5) as the reference electrolyte57,58. Sample solution (~30 nL) was injected into the system at 50 mbar for 30 s and a negative voltage of −30 kV was applied. Ammonium acetate (5 mmol/L) in methanol/water (50% v/v) containing 0.1 μmol/L hexakis (2,2-difluorothoxy) phosphazene was delivered as sheath liquid at 10 μL/min. ESI-TOFMS was performed in negative ion mode at a capillary voltage of 3.5 kV. For TOFMS, the fragmenter voltage, skimmer voltage, and Oct RFV were set at 100, 50, and 200 V, respectively58. An automatic recalibration function was performed using the masses of two reference compounds: deprotonated 13C acetate dimer (m/z 120.038339) and an acetate adduct of hexakis (2,2-difluorothoxy) phosphazene (m/z 680.035541). The other conditions were identical to those used for the cationic metabolome analysis.
CE-TOFMS data processing
Raw data were processed using in-house Masterhands software59. The overall data processing flow consisted of the following steps: noise-filtering, baseline-removal, migration time correction, peak detection, and peak area integration from a 0.02 m/z-wide slice of the electropherograms. The data processing resembled the common strategies used for LC-MS and GC-MS data analysis software, such as MassHunter (Agilent Technologies) and XCMS60. 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 programming61,62. All target metabolites were identified by matching their m/z values and normalized migration times with those of standard compounds in the in-house library.
RNA isolation and Affymetrix microarray hybridization
Trophozoites were grown in BI-S-33 medium containing 8 mM L-cysteine for approximately 48 h. The collected cell pellets were resuspended in Trizol reagent (Invitrogen, Carlsbad, CA, USA) and RNA was isolated according to the manufacturer’s protocol. The RNA concentration for each sample was measured using a Nanodrop Spectrophotometer 1000 (Thermo Scientific, Wilmington, DE, USA). RNA integrity was checked using an Experion Automated Electrophoresis System (RNA StdSens analysis kit, Bio-Rad). All reagents and protocols followed those described in the Affymetrix user manuals. Using the One-Cycle cDNA synthesis kit, 5 μg total RNA was reverse transcribed using a T7-Oligo (dT) primer for first strand cDNA synthesis. After second strand synthesis, the double-stranded cDNA template was used for in-vitro transcription (IVT) in the presence of biotinylated nucleotides (GeneChip IVT labeling kit) to produce Biotin-labeled cRNA. Unincorporated NTPs were removed from the biotinylated cRNA (GeneChip sample cleanup module), which was then purified, quantified and fragmented. A hybridization cocktail consisting of eukaryotic hybridization controls and fragmented, labeled cRNA (GeneChip Hybridization, Wash and Stain Kit) were hybridized for 16 h at 45 °C in a Hybridization Oven 640 (Affymetrix) onto a custom-generated Affymetrix platform microarray (49–7875) with probe sets consisting of 11 probe pairs, each representing 12,384 E. invadens 63 (Eh_Eia520620F_Ei) and 9,327 E. histolytica 36 (Eh_Eia520620F_Eh) open reading frames. The array chips were washed and stained (GeneChip Hybridization, Wash and Stain Kit) with Streptavidin–phycoerythrin Biotinylated anti-streptavidin antibody using a GeneChip Fluidics Station 450 (Affymetrix) for 1.5 h. After washing and staining, the GeneChip arrays were scanned using a Hewlett-Packard Affymetrix Scanner 3000.
Analysis of microarray data
A minimum of two arrays was used for each test condition. Raw probe intensities were generated using Gene Chip Operating Software (GCOS) and the Gene Titan Instrument from Affymetrix. Normalized expression values for each probe set were obtained from R 2.7.0 downloaded from the Bioconductor project (http://www.bioconductor.org) using robust multiarray averaging with correction for oligosequence (gcRMA). Standard correlation coefficients were calculated using GeneSpring GX 10.0.2. One-way ANOVA analysis with Tukey’s Post Hoc test was performed to extract differentially expressed genes. P values were calculated using Welch’s t-test after multiple test correction by the Benjamini–Hochberg method. A post-hoc test using Tukey’s Honestly Significant Difference test was conducted to determine significant differences between samples.
Quantitative real-time PCR (qRT–PCR)
Total RNA extracted above were used for qRT–PCR. cDNA synthesis was performed using the SuperScript III First-Strand Synthesis System (Invitrogen) following the manufacturer’s instructions. cDNA was synthesized from 5 μg total RNA and oligo (dT) 20 primers using the Superscript III First-Strand Synthesis System (Invitrogen). PCR was performed with cDNA as the template and gene-specific primers using the ABI PRISM 7300 Sequence Detection System (Applied Biosystems, Weiterstadt, Germany). The genes whose expression was verified by qRT–PCR are listed in Supplementary Table S4. The RNA polymerase II gene was used as a control. The parameters for PCR were: an initial denaturation step at 95 °C for 9 min followed by 40 cycles of denaturation at 94 °C for 30 s, annealing at 50 °C for 30 s and extension at 65 °C for 1 min. A final step at 95 °C for 9 s, 60 °C for 9 s and 95 °C for 9 s was used to remove primer dimers. All test samples were run in triplicate. An RT-negative control was also used for each sample set along with a blank control consisting of nuclease-free water in place of cDNA.
Growth assay of E. histolytica trophozoites
A cell-growth assay was performed as described previously36. Briefly, approximately 6 × 104 exponentially growing SAT1/2, SAT3, or CS gene-silenced trophozoites and control transformants were inoculated into 6 mL normal BI-S-33 medium with and without L-cysteine supplemented with 7 μg/mL geneticin, and the number of parasites was counted every 24 h using a haemocytometer.
Hydrogen peroxide (H2O2) sensitivity assay
To examine sensitivity to H2O2, E. histolytica CS gene-silenced and control (harboring plasmid psAP2G) transformants were cultured in L-cysteine lacking BI-S-33 media containing 7 μg/mL geneticin for 48 h at 35.5 °C. After 48 h, approximately 104 trophozoites per well were seeded into the wells of a 96-well plate containing BI-S-33 medium supplemented with 7 μg/mL geneticin and further incubated for 1 h at 35.5 °C. The trophozoites were then exposed to H2O2 (0, 0.8, 1.6, 2.4, 3.2, 4, 4.8 and 6.4 mM) for 1 h. After incubation, the medium was removed and 90 μl of pre-warmed Opti-MEM I (Life Technologies) and 10 μl WST-1 solution64 (Roche Diagnostics, Mannheim, Germany) were added to each well. Viability of trophozoites was detected by measuring absorbance at 450 nm using a microplate reader (Model 550, Bio-Rad, Tokyo, Japan). The sensitivity assays were performed in triplicate and repeated at least three times.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work was supported in part by a grant for Research on Emerging and Re-emerging Infectious Diseases from the Japan Agency for Medical Research and Development (AMED), Grants-in-Aid for Scientific Research on Innovative Areas (23117001, 23117005), a Grant-in-Aid for Scientific Research (23390099) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), and a grant for Science and Technology Research Partnership for Sustainable Development from AMED and the Japan International Cooperation Agency to TN.