mitosomes play an important role in encystation by association with cholesteryl sulfate synthesis

Fumika Mi-ichia,1, Tomofumi Miyamotob, Shouko Takaoa, Ghulam Jeelanic, Tetsuo Hashimotod,e, Hiromitsu Haraa, Tomoyoshi Nozakic,d,1, and Hiroki Yoshidaa

aDivision of Molecular and Cellular Immunoscience, Department of Biomolecular Sciences, Faculty of Medicine, Saga University, Saga, Saga 849-8501, Japan; bDepartment of Natural Products Chemistry, Graduate School of Pharmaceutical Sciences, Kyushu University, Higashi-ku, Fukuoka 812-8582, Japan; cDepartment of Parasitology, National Institute of Infectious Diseases, Shinjuku-ku, Tokyo 162-8640, Japan; dGraduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan; and eCentre for Computational Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan

Edited by W. Ford Doolittle, Dalhousie University, Halifax, Canada, and approved April 28, 2015 (received for review December 11, 2014) Hydrogenosomes and mitosomes are mitochondrion-related or- as the tricarboxylic acid (TCA) cycle, electron transport, oxida- ganelles (MROs) that have highly reduced and divergent functions tive phosphorylation, and β-oxidation of fatty acids (1, 2). Fur- in anaerobic/microaerophilic . ,a thermore, unique features of mitosomes, unlike other MROs, have microaerophilic, parasitic amoebozoan species, which causes intes- not been linked to distinct roles in organisms. tinal and extraintestinal in humans, possesses mito- Mitosomes have been described exclusively in protistan para- somes, the existence and biological functions of which have sites including Entamoeba histolytica, the causative agent for in- been a longstanding enigma in the evolution of mitochondria. testinal and extraintestinal amoebiasis in humans. The infectious We previously demonstrated that sulfate activation, which is not diseases caused by this parasite are a serious public health generally compartmentalized to mitochondria, is a major function problem (6), and thus developing novel therapeutics to prevent of E. histolytica mitosomes. However, because the final metabo- these diseases is of great importance. We have shown that sulfate lites of sulfate activation remain unknown, the overall scheme of activation is a major metabolic pathway in E. histolytica mito- this metabolism and the role of mitosomes in Entamoeba have not somes (1, 5). Because in eukaryotes sulfate activation generally been elucidated. In this study we purified and identified choles- occurs in the cytoplasm or plastids (1, 7), its compartmentali- teryl sulfate (CS) as a final metabolite of sulfate activation. We zation to mitosomes is unprecedented. MROs are present in a then identified the encoding the cholesteryl sulfotransferase variety of unicellular eukaryotic organisms (2, 4); however, the responsible for synthesizing CS. Addition of CS to culture media relationship of sulfate activation to MROs is not clear. In increased the number of cysts, the dormant form that differenti- Mastigamoeba balamuthi, a microaerophilic, free-living , ates from proliferative trophozoites. Conversely, chlorate, a selec- which is a close relative of E. histolytica, the enzymes involved in tive inhibitor of the first enzyme in the sulfate-activation pathway, the sulfate-activation pathway are encoded in the genome, and inhibited cyst formation in a dose-dependent manner. These re- these proteins possess mitochondrial targeting signals; two of sults indicate that CS plays an important role in differentiation, an these enzymes are localized in MROs (8). Although other an- essential process for the of Entamoeba between aerobic/microaerophilic, parasitic organisms, such as Trichomo- hosts. Furthermore, we show that Mastigamoeba balamuthi, an nas vaginalis, Giardia intestinalis, and Cryptosporidium parvum, anaerobic, free-living amoebozoan species, which is a close rela- tive of E. histolytica, also has the sulfate-activation pathway in Significance MROs but does not possess the capacity for CS production. Hence, we propose that a unique function of MROs in Entamoeba con- Evolution and diversification of organelles is a central topic in tributes to its adaptation to its parasitic life cycle. biology. Mitochondrion-related organelles (MROs) are highly modified forms of mitochondria found in anaerobic eukary- mitochondrion-related organelles | protist | sulfolipids | differentiation otes. MROs show a spectrum of functions that are either re- duced or modified from those of canonical mitochondria by itochondrion-related organelles (MROs) are derived from environmental constraints and evolutionary selection. Hence, Mcanonical mitochondria and are found in a wide range of elucidation of MRO functions will improve our understanding anaerobic/microaerophilic eukaryotes (1, 2). During the course of organelle evolution and the speciation of eukaryotes. Here, of evolution MROs have undergone secondary loss of mito- we substantiate a role of the Entamoeba mitosome, a type of chondrial functions; this loss has occurred independently multi- MRO, by showing that cholesteryl sulfate synthesized through ple times, resulting in the broad phylogenetic distribution of a mitosomal pathway regulates differentiation that is essential organisms possessing MROs (2). Furthermore, MROs occa- for the parasite’s life cycle. These findings support the contri- sionally acquire novel functions from other organisms by lateral bution of an endosymbiont-derived organelle to , gene transfer (LGT) (1, 3). Hence, MROs are not simply rem- a previously unrecognized concept that casts new light on nants of mitochondria but rather are organelles that display a organelle evolution. variety of unique features (1–5). Unique features have been demonstrated in different types of Author contributions: F.M. designed research; F.M., T.M., S.T., G.J., and T.H. performed MRO (2–4). Some anaerobic lineages of eukaryotes possess research; F.M., H.H., T.N., and H.Y. analyzed data; and F.M., T.N., and H.Y. wrote MROs (hydrogenosomes or hydrogen-producing mitochondria) the paper. that have remodeled their mitochondria drastically to couple The authors declare no conflict of interest. ATP generation with hydrogen production (2–4). Mitosomes, This article is a PNAS Direct Submission. 1To whom correspondence may be addressed. Email: [email protected] or nozaki@ another type of MRO maintained in some organisms that inhabit nih.go.jp. anaerobic/microaerophilic environments, do not produce ATP This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. or hydrogen and have lost typical mitochondrial functions, such 1073/pnas.1423718112/-/DCSupplemental.

E2884–E2890 | PNAS | Published online May 18, 2015 www.pnas.org/cgi/doi/10.1073/pnas.1423718112 Downloaded by guest on October 1, 2021 possess MROs, the genomes of these organisms apparently lack We attempted to identify the gene responsible for producing PNAS PLUS involved in the sulfate-activation pathway (1). Moreover, CS in E. histolytica through a gene knockdown approach. Each of phylogenetic analyses revealed that E. histolytica and M. bala- the 10 SULT genes was knocked down in separate E. histolytica muthi appear to have acquired the enzymes in the sulfate-acti- transformants by antisense small RNA-mediated transcriptional vation pathway from distinct prokaryotic and eukaryotic lineages gene silencing (5, 13), and these transformants were designated by LGT (1, 8). Therefore sulfate activation is not a conserved “SULT1–10gs.” Quantitative RT-PCR (qRT-PCR) verified a function of MROs but may be a unique feature of MROs main- 92.8–99.9% reduction of the steady-state transcript level of each tained by the Entamoeba and Mastigamoeba lineages (5). target gene (Fig. S4). As expected, in SULT2–4gs, SULT6gs, and Sulfate activation is achieved by sequential reactions mediated SUTL8–10gs, only the target genes were knocked down. How- by ATP sulfurylase (AS) and adenosine 5′-phosphosulfate kinase ever, in SULT1gs, SULT5gs, and SUTL7gs, not only the target (APSK) to produce 3′-phosphoadenosine 5′-phosphosulfate genes but also two other genes (SULT5 and -7 in SULT1gs; (PAPS) (7). PAPS then acts as an activated sulfur donor to SULT1 and -7 in SULT5gs; and SULT1 and -5 in SULT7gs) were synthesize various sulfated metabolites through sulfotransferase knocked down simultaneously, probably because of high sequence (SULT) reactions. The metabolites thus produced have impor- identity (70.0–77.9%) among the regions used for gene si- tant roles in a variety of cellular events (9). In E. histolytica,we lencing in SULT1,-5,and-7. previously demonstrated that a major fraction of sulfated me- We then analyzed the profile of sulfolipids produced in tabolites are sulfolipids (1, 5). However, these sulfolipids were SULT1–10gs and control mock transformants (wild-type strain not identified, and thus the overall scheme of this metabolism in transfected with an empty vector) by metabolic labeling. Among Entamoeba remained unknown. More importantly, the role of the 10 transformants analyzed, only SULT6gs showed a differ- Entamoeba mitosomes remained an enigma. In this study, to ence in the profile of sulfolipids produced, i.e., a significant re- address these issues, we identified a sulfolipid synthesized in duction only of CS (Fig. 1). This result indicated that SULT6 E. histolytica and the gene responsible for producing it. We then encodes the cholesteryl SULT that is solely responsible for elucidated their biological significance in Entamoeba and ex- synthesizing CS. Moreover, SULT6 is functionally linked with amined whether this function is conserved in . the E. histolytica mitosomal sulfate-activation pathway via PAPS, which is a product of the pathway and a substrate for SULT6. Results Sulfolipid-I Is Cholesteryl Sulfate. We previously demonstrated that CS and SULT6 Are Dispensable in the Proliferative Trophozoite Stage the sulfate moiety produced by the sulfate-activation pathway in of E. histolytica. We next investigated the phenotype of SULT6gs E. histolytica mitosomes is incorporated mainly into sulfolipids to determine the role of CS in the proliferative trophozoite stage. (1, 5). In this study we first established a purification procedure Because SULT6gs, which has lost almost all capacity for pro- for different sulfolipids using a TLC plate with improved sepa- ducing CS, is maintained in the trophozoite stage, we conducted ration conditions (SI Methods and Fig. S1). Using a combination analyses that are available for the trophozoite stage, i.e., cell of Oasis WAX columns, by which sulfolipids can be enriched growth and CHO cell cytopathic activity. In all the analyses efficiently (as described in SI Methods and Fig. S1), and recovery performed, phenotypes did not significantly differ from wild type from TLC plates, we obtained ∼3 mg sulfolipid-I (SL-I), ∼2mg (Fig. S5), indicating that CS and SULT6 are dispensable in the EVOLUTION sulfolipid-V (SL-V), and ∼2 mg sulfolipid VI (SL-VI) from 1 × proliferative trophozoite stage of E. histolytica. 109 Entamoeba trophozoite cells. Our inability to purify sulfoli- Up-Regulation of CS Production During Encystation. The life cycle of pids II–IV (SL-II–SL-IV) by this procedure suggests that they Entamoeba consists of the proliferative trophozoite stage and the are secreted. We determined SL-I to be cholesteryl sulfate (CS) dormant cyst stage, which is responsible for transmission be- by NMR and MS analyses (SI Methods and Fig. S2). tween hosts. Both stages are essential to maintain the Entamoeba SULT6 Is Solely Responsible for Synthesizing CS in E. histolytica. CS, a life cycle. We examined a potential role of CS in the dormant cyst stage, especially during differentiation from the trophozoite constituent of cellular membranes, originally was isolated from “ ” human plasma, and its synthesis is mediated by an SULT that to the cyst, a process called encystation (14). We monitored uses cholesterol and PAPS as substrates (10, 11). To date, several the production of CS during encystation in Entamoeba invadens, genes for cholesteryl SULTs have been identified in mammals which, unlike laboratory strains of E. histolytica (14), encysts (10, 12). In the E. histolytica genome, 10 genes encoding putative efficiently in in vitro culture. E. invadens is a sibling species of SULTs are present [AmoebaDB, amoebadb.org/amoeba/ (1)], and amino acid sequences of these genes can be aligned, dem- onstrating their homology (Table S1). In the currently available public databases there are putative SULTs from a wide range of prokaryotes and from several eukaryotic lineages in addition to mammals. Amino acid residues essential for PAPS binding, i.e., 5′-phosphosulfate– and 3′-phosphate–binding motifs (9, 10, 12), are conserved in all SULTs, including those in Entamoeba (Fig. S3A). However, except for these motifs and their flanking re- gions, the mammalian cholesteryl SULTs cannot be aligned unambiguously with other putative SULTs. Poor resolution of the phylogenetic tree combined with the significant sequence divergence of the Entamoeba homologs relative to other organ- isms makes it difficult to infer the evolutionary history of this protein. However, it is clear that the Entamoeba paralogs of this Fig. 1. Profiles of sulfolipids produced in SULT1–10gs and in controls. HPTLC 35 – protein have been expanded drastically along the Entamoeba of methanol extracts from [ S]-labeled SULT1 10gs and the control. Each column represents a profile of sulfolipids (SL-I–VI) produced in each trans- lineage as compared with other lineages, suggesting that an an- formant (SULT1–10gs) in which a single SULT (SULT1–10) was knocked down cestor of Entamoeba acquired an ancestral SULT from a phy- or in which only vector was introduced (control). The autoradiographic logenetically distantly related organism by LGT (SI Methods and image shown is representative of the results from three independent Fig. S3B). experiments.

Mi-ichi et al. PNAS | Published online May 18, 2015 | E2885 Downloaded by guest on October 1, 2021 Fig. 2. CS level in Entamoeba is tightly controlled during encystation. (A–D) Representative autoradiographic images of results from three independent experiments. (A) Comparison of the HPTLC pattern of sulfolipids produced in E. histolytica (Eh) and E. invadens (Ei) trophozoites. (B, Right) Sulfolipid profiles during encystation are shown for the time points indicated. (Left) A trophozoite profile is shown for comparison. (C) The amount of CS accumulated in cells for every 6 h in encystation medium. (D) Pulse-chase experiment. (E) Modulation of the E. invadens SULT6 transcript level during encystation. Expression levels are shown as fold changes at the indicated time points after the induction of encystation relative to the level at time 0 h. Experiments were performed in triplicate; n = 2.

E. histolytica, possesses mitosomes in both trophozoites and cysts phozoites cultured in the presence of various concentrations (15), and in causes diseases similar to those caused by (0–500 μM) of CS. Addition of CS (50–200 μM) to the culture E. histolytica (16). Sulfolipids produced in E. invadens trophozoites, medium increased the number of cysts formed at 48 h (Fig. 3A). including CS, are comparable to those in E. histolytica (Fig. 2A). However, the addition of 500 μM CS or prolonged incubation In addition, the E. invadens genome encodes only one ortholog with 200 μM CS had a significant negative effect on encystation of E. histolytica SULT6 (Fig. S3C), reinforcing the premise that efficiency (Fig. 3A). These results indicate that CS, when present E. invadens can synthesize CS. Upon induction of encystation, at optimum concentration, enhances cyst formation. the amount of CS associated with cells increased significantly To confirm the role of CS in encystation, we used chlorate, a (Fig. 2B). A time-course analysis of metabolic labeling with selective inhibitor of AS, the first enzyme in the sulfate-activa- constant time intervals showed that the amount of CS associated tion pathway (5). Unfortunately we could not take advantage of with cells increased immediately upon the initiation of encysta- gene silencing in E. invadens, and no inhibitor for SULT6 is tion and that this increase was transient. CS levels reached a available. In view of chlorate IC50 values of 16.7 ± 6.3 mM and maximum 18–24 h after initiation and decreased from 24 h on- 69.3 ± 6.6 mM against the growth of E. histolytica and E. inva- ward, and new production of CS was almost undetectable after dens trophozoites, respectively (ref. 5 and this study), chlorate 48 h (Fig. 2 B and C). Furthermore, pulse-chase experiments was added to cultures in the concentration range of 0–450 mM. showed that the CS accumulated during the first 24 h was stable Chlorate indeed inhibited the synthesis of CS in a dose- for up to 120 h (Fig. 2D). These results indicate that CS synthesis dependent manner (Fig. 3B), and a significant reduction in the is up-regulated immediately upon the initiation of encystation amount of CS associated with cells was observed with chlorate and that the CS level is maintained at a constant level from 48 h >100 mM (Fig. 3B). Concomitantly, chlorate >100 mM reduced after initiation. Consistent with the observed up-regulation of CS the number of cysts formed in a dose-dependent manner (Fig. synthesis during encystation, the transcription of E. invadens 3C), reinforcing the premise that sulfate activation, CS pro- SULT6 also was up-regulated immediately upon induction of duction, and encystation are causally connected. Although the encystation. However, its expression level continued to increase role of CS in E. histolytica encystation needs to be confirmed, it up to 120 h (Fig. 2E) (17). The inconsistency in the changes of was demonstrated previously that E. histolytica strains that were SULT6 transcript and CS levels can be explained by the exoge- isolated from dysentery patients and retained high encystation nously added CS, which inhibited accumulation of CS, probably efficiency showed high expression levels of SULT6 (18). This because of product inhibition of the SULT6 enzyme (Fig. S6). observation, together with our results (Fig. 2E) (17), indicates Taken together, these data suggest that CS concentration in the that SULT6 expression is up-regulated during encystation in cell is strictly controlled. Entamoeba. Collectively, these findings indicate that CS plays an important role in the encystation of Entamoeba. CS Plays an Important Role in Entamoeba Encystation. To investigate the role of CS in encystation more precisely, its effect was M. balamuthi Does Not Possess the Capacity for CS Production. An evaluated directly by monitoring encystation of E. invadens tro- important question then arises: How ubiquitous is the involvement

E2886 | www.pnas.org/cgi/doi/10.1073/pnas.1423718112 Mi-ichi et al. Downloaded by guest on October 1, 2021 PNAS PLUS

Fig. 3. CS plays an important role in the encystation of Entamoeba.(A) The number of cysts formed in medium containing different concentrations of EVOLUTION exogenous CS. The experiment was performed in duplicate; n = 3. (B) Effect of chlorate, an inhibitor of the sulfate-activation pathway, on CS production during encystation. Shown are HPTLC of methanol extracts of [35S]-labeled cells cultivated in encystation medium in the presence of different concentrations of chlorate for 120 h. (C) Effect of chlorate on cyst formation. The number of cysts formed and the total number of cells remaining in the encystation medium in the presence of different concentrations of chlorate are shown The experiment was performed in duplicate; n = 3. Representative data among consistent results from three independent experiments are shown.

of CS and MROs in differentiation, i.e., is this association con- tained by Entamoeba organisms possess the sulfate-activation served in the supergroup Amoebozoa or is it restricted to the pathway, which was acquired by LGT from bacteria and eu- genus Entamoeba? To address this question, we extended the study karyotes (1, 5). However, in Entamoeba organisms, the biological to M. balamuthi, a free-living amoebozoan protist that is a close roles of mitosomes and the sulfate-activation pathway have relative of E. histolytica (19) and, like Entamoeba,isknownto remained an enigma. The present study provides several lines of possess MROs (20, 21). In addition, the M. balamuthi life evidence indicating that the mitosomal sulfate-activation path- cycle, including the cyst stage, is similar to that of Entamoeba (22). way, which coordinates with SULT6 to produce CS, plays an Immunofluorescence microscopy showed that M. balamuthi important role in Entamoeba encystation. Direct supplementa- APSK, the second enzyme of the sulfate-activation pathway (1), is tion of CS into medium, which presumably results in elevated CS localized in MROs (Fig. S7). These findings, together with MRO level in cells, enhanced the encystation efficiency of E. invadens localizations of two other enzymes in this pathway [AS and in- within an optimal concentration range. This finding suggests that organic pyrophosphatase (8)], indicate that the sulfate-activation the intracellular CS level must be controlled for encystation. pathway functions in M. balamuthi MROs. However, metabolic Furthermore, a reductive approach using an inhibitor of sulfate labeling shows that, in contrast to Entamoeba, M. balamuthi tro- activation showed that impairment of CS synthesis results in the phozoites do not produce sulfolipids (Fig. 4A). In agreement with abolition of encystation, indicating that a minimal CS level is this observation, TBLAST searches of the draft genome sequence required for triggering Entamoeba encystation. Entamoeba can of M. balamuthi reveal that homologs of E. histolytica SULTs are synthesize CS via SULT6, using PAPS and cholesterol as sub- lacking (mastigamoeba.img.cas.cz/). These results imply that strates. PAPS is synthesized by the Entamoeba mitosomal sul- the role of the sulfate-activation pathway differs in M. balamuthi fate-activation pathway, whereas there is no de novo synthetic and Entamoeba. pathway for cholesterol in Entamoeba (23, 24). Thus, Entamoeba rely on the external milieu as the source of cholesterol. In light of Discussion these findings, we suggest that, to trigger encystation, Entamoeba MROs, which include mitosomes and hydrogenosomes, are de- raises intracellular CS levels via the catalytic action of SULT6 rived from canonical mitochondria and exist with highly reduced and substrates that are provided intra- and extracellularly. and divergent functions in a broad range of anaerobic/micro- Unfortunately, an in vivo encystation model is not available; aerophilic eukaryotic organisms (1–4). The mitosomes main- therefore, investigation of the process is difficult. However, data

Mi-ichi et al. PNAS | Published online May 18, 2015 | E2887 Downloaded by guest on October 1, 2021 Fig. 4. Proposed models for the importance of the sulfate-activation pathway in MROs during the course of evolution of the anaerobic amoebae Entamoeba and Mastigamoeba.(A) HPTLC of methanol extracts of [35S]-labeled E. histolytica and M. balamuthi. The autoradiographic image shown is representative of the results from three independent experiments. (B) Illustration showing the importance of the sulfate-activation pathway in Entamoeba mitosomes. AAC, ATP/ADP carrier; APS, adenosine 5′-phosphosulfate; APSK, APS kinase; AS, ATP sulfurylase; Cpn, chaperonin; Hsp, heat shock protein; IPP, inorganic pyro- phosphatase; PAP, adenosine-3′,5′-bisphosphate; PiC, Pi carrier; SAM, sorting and assembly machinery; TOM, translocase of the outer membrane. (C) Proposed roles of the sulfate-activation pathway in MROs.

regarding CS in vivo can be combined with the mechanism whereas our previous study demonstrated that the sulfate-acti- suggested here. CS concentration in human plasma ranges from vation pathway, a major role of which is providing a substrate for 3.7–7 μM (11), which is far too low to enhance the efficiency of SULTs, including SULT6, is important for E. histolytica tro- Entamoeba encystation. Conversely, cholesterol, a substrate for phozoite proliferation (1, 5). This apparent inconsistency may be SULT6 to produce CS, is rich in vertebrate intestines (25, 26); explained by other sulfolipids, which probably are synthesized by such an environment therefore provides not only a niche for SULTs other than SULT6, being important for cell proliferation Entamoeba but also an external milieu for its growth. Moreover, and/or complementing the function of CS. Hence, we suspect in E. histolytica strains isolated from dysentery patients, both the that SULTs and sulfolipids have versatile roles in the life cycle of expression levels of SULT6 and encystation efficiency are high Entamoeba. To corroborate this inference, molecular identifi- (18), suggesting that in E. histolytica the CS level needs to be high cation of SL-II–VI and developing methods for multiple gene during encystation in vivo. These notions are consistent with the knockdown would be helpful. mechanism we have proposed in the present in vitro study. Encystation involves the differentiation of proliferative tro- Taken together, our results demonstrate that regulating encysta- phozoites into dormant cysts and is thus crucial for maintaining tion is an important role of the sulfate-activation pathway in the life cycle of Entamoeba parasites. Encystation proceeds with Entamoeba mitosomes (Fig. 4B). This regulation occurs in asso- a variety of specific, simultaneous changes; for instance, cells ciation with the synthesis of CS by cholesteryl SULT (SULT6) and become surrounded by chitin walls (14). Encystation traits make provides an answer to the longstanding enigma as to the function Entamoeba organisms resistant to environmental changes and of Entamoeba mitosomes. Furthermore, this study provides evi- thus help the high transmission rate of this parasite (14). dence for the previously unrecognized concept that endosymbiont- Therefore understanding the molecular mechanism underlying derived MROs are involved in cellular differentiation. encystation is important from a therapeutic perspective as well as The present study suggests that SULT6 and CS are dispens- from that of basic biology. However, this important mechanism able in the proliferative trophozoite stage of E. histolytica, remains largely unknown. In this study, we identified CS as a key

E2888 | www.pnas.org/cgi/doi/10.1073/pnas.1423718112 Mi-ichi et al. Downloaded by guest on October 1, 2021 molecule for the regulation of encystation. In mammals, the were determined as described previously (29). CS was dissolved in DMSO at PNAS PLUS roles of CS in cellular functions are relatively well characterized 50 mM, and chlorate was dissolved in BI-S-33 or encystation medium at (e.g., regulation of lipid metabolism, protection of erythrocytes 500 mM, as stock solutions. All the stock solutions were prepared freshly just from osmotic lysis, supporting platelet adhesion, and acting as a before use. CS or chlorate was added to encystation medium at time 0 h. For cultures to which DMSO was added, including control cultures treated with nuclear receptor ligand) (11, 27). However, there have been few DMSO alone, the final DMSO concentration was fixed at 1% (vol/vol). For reports concerning CS, or even its production, in unicellular metabolic labeling, E. invadens trophozoites were labeled with [35S]-labeled eukaryotes (28). To our knowledge, this is the first report dem- sulfate (12.5 μCi/mmol) in 2 mL of BI-S-33 or encystation medium for 24 or onstrating the precise role of CS in a unicellular . 120 h, respectively, at 26 °C. For pulse-chase experiments, cells were labeled Further investigation of the molecular mechanisms in which CS with [35S]-labeled sulfate in 2 mL of encystation medium for 24 h, followed participates would lead not only to new insight into CS functions by further incubation in 2 mL of isotope-free encystation medium for 24, 48, and encystation but also to a rationale for the development of 72, or 96 h after a single washing with PBS. Cell densities used in the BI-S-33 × 5 × 5 novel therapeutics to prevent the infectious diseases caused by or encystation medium were 1.5 10 /mL or 6 10 /mL, respectively. Analysis of E. invadens sulfolipids was performed essentially as described for Entamoeba parasites. E. histolytica. Surprisingly, the involvement of CS and MROs in differenti- ation is restricted to Entamoeba. Mastigamoeba does possess the Cultivation, Metabolic Labeling, and Indirect Immunofluorescence Analysis of sulfate-activation pathway in MROs but does not have the ca- M. balamuthi. M. balamuthi trophozoites were grown in BI-S-33 medium at pacity for producing sulfolipids, including CS (this study and ref. 26 °C. For metabolic labeling, M. balamuthi trophozoites were labeled with 8). These results imply that the role of the sulfate-activation [35S]-labeled sulfate (25 μCi/mmol) in 6 mL of BI-S-33 medium for 24 h at pathway has diverged between M. balamuthi and Entamoeba, and 26 °C. Analysis of sulfolipids of M. balamuthi was performed essentially as thus it is of great importance to unravel the role of the sulfate- described for E. histolytica. Indirect immunofluorescence analysis was per- activation pathway in Mastigamoeba MROs. Given that the sul- formed essentially as previously described (30). fate-activation pathway is well conserved in Entamoeba and Mastigamoeba, it is likely that the sulfate-activation pathway is Purification of Sulfolipids. For small-scale purification, 0.5 mL of methanol cell extract of [35S] sulfate-labeled E. histolytica trophozoites (1.2 × 105 cells) was exploited differently in these organisms, depending upon the applied onto a 1 cc Oasis WAX cartridge (30 mg; Waters), which had been enzyme(s) using PAPS. This notion was partially corroborated by equilibrated with 1 mL methanol, followed by 1 mL water. After the cell our finding that SULT6 is solely responsible for producing CS extract was passed through the cartridge, the cartridge was washed with in Entamoeba by using PAPS, whereas neither SULT6 nor 1 mL of 2 mM citrate, followed by 1 mL methanol. All the retained com- other SULT counterparts are encoded in the Mastigamoeba ge- pounds were eluted with 1 mL of 5% (vol/vol) ammonium hydroxide in meth- nome. It would be intriguing to identify an enzyme that uses anol. After solvent was evaporated in a spin-dryer centrifuge (VC-36R; TAITEC), PAPS in Mastigamoeba. More interestingly, ancestral genes for the pellet was dissolved in 50 μL methanol. The purified lipids were separated Entamoeba SULTs are suggested to have been transferred by by HPTLC [chloroform/methanol/28% (wt/wt) ammonium hydroxide (65:35:8; vol/vol/vol)] (5). The TLC plates were analyzed as described for E. histolytica. LGT to an ancestor of this organism. It is plausible that the For large-scale purification, 10 mL of methanol cell extract of E. histolytica acquisition of genes encoding SULT by a lineage of Entamoeba trophozoites (2 × 108 cells) and the 6 cc Oasis WAX cartridge (500 mg; Wa- resulted in the ability to synthesize CS, a regulatory molecule for ters) were used. The procedure for column chromatography was essentially

encystation. This event could be key for adaptation to a parasitic as described above, except that the solvent volumes for equilibrium, wash- EVOLUTION lifestyle in the host. This assumption is supported by the fact that ing, and elution were 3, 3, and 4 mL, respectively. After solvent was evap- Entamoeba, which is known to have no pathways for de novo orated, the purified lipids in the pellet were dissolved in 200 μL methanol cholesterol synthesis (23, 24), inhabits a niche in vertebrate in- and separated by 2D HPTLC [first dimension, chloroform/methanol/28% testines that is rich in cholesterol (25, 26), an essential precursor (wt/wt) ammonium hydroxide (65:35:8) (vol/vol/vol); second dimension, for synthesizing CS (11, 12). Hence, we propose a scenario for chloroform/methanol/acetone/acetic acid/water (50:20:10:10:5) (vol/vol/vol/ vol/vol)] (31). Subsequently the lipids were visualized by exposure to iodine the evolution of anaerobic amoeba in line with MRO functions vapor, and then the spots corresponding to the sulfolipids were individually (Fig. 4C). scraped from HPTLC plates. The samples were extracted with 3 mL methanol twice, followed by the removal of silica gel by centrifugation at 780 × g for Methods 5 min. After solvent was evaporated in a spin-dryer centrifuge, the purified Materials. [35S]-labeled sulfate (10 mCi/mmol) was purchased from American sulfolipids were stored at −30 °C under nitrogen gas until use. For detection Radiolabeled Chemicals, Inc. Sodium CS (synthetic product; >99% purity), of sulfolipids on a TLC plate, a dried silica gel plate was sprayed with 0.05% potassium chlorate (99.4% purity), and pinacryptol yellow (98% purity) were (wt/vol) pinacryptol yellow (32). Sulfur-containing anionics give bright or- from Sigma-Aldrich. ange fluorescence when reacted with pinacryptol yellow and viewed under long-wavelength UV light (365 nm). Cultivation and Metabolic Labeling of E. histolytica. In vitro cultures of E. histolytica HM-1:IMSS cl6 and G3 strains were maintained routinely as Mass Spectrometric Analysis of SL-I. SL-I, one of the sulfolipids purified, was

described previously in Diamond’s BI-S-33 medium at 37 °C (1, 5). Metabolic dissolved in 0.1% (vol/vol) HCOOH/CH3CN and subjected to electrospray labeling of E. histolytica was performed as described previously (1, 5), with ionization (ESI)-TOF MS analysis with a micrOTOF II mass spectrometer slight modifications. Briefly, 1.2 × 105 E. histolytica trophozoites were la- (Bruker). The operation was conducted in the negative ion mode. beled with [35S]-labeled sulfate (25 μCi/mmol) in 0.75 mL BI-S-33 medium at

37 °C for 24 h under anaerobic conditions using Anaerocult A (Merck). After NMR Analysis of SL-I. The purified SL-I was dissolved in CD3OD/CDCl3 [1/1 (vol/ the labeled cells were harvested, lipids were extracted with 0.5 mL metha- vol), 600 μL]. The NMR spectrum of SL-I was recorded on a Varian INOVA 600 nol. The extracted lipids were separated by high-performance TLC (HPTLC) spectrometer (Varian) at 303 K, and the 1H NMR chemical shifts were ref- [chloroform/methanol/28% (wt/wt) ammonium hydroxide (65:35:8; vol/vol/ erenced to tetramethylsilane. Sodium CS was used as an authentic sample. vol)] on a silica gel plate (no. 1.05641.0001; Merck). Each spot on the TLC plates was quantified using a Fuji imaging analyzer and Multi Gauge 2.2 Phylogenetic Analysis of SULTs. To analyze the phylogenetic relationship software (FLA-7000; Fujifilm). among the SULTs that exist in a wide variety of organisms, we performed BLAST searches of AmoebaDB (amoebadb.org/amoeba/) and the UniProt Cultivation, Encystation, and Metabolic Labeling of E. invadens. Trophozoites (www.uniprot.org/) databases with the E. histolytica SULT6 sequence as a of the E. invadens IP-1 strain were cultivated axenically as described pre- query to retrieve as many SULT sequences as possible. The SULTs used for viously in BI-S-33 medium at 26 °C (17, 29). To induce encystation, 1-wk-old analyses are listed in Tables S2 and S3. Amino acid sequences of these SULTs trophozoite cultures were harvested and transferred to encystation medium were automatically aligned by ClustalW (33), followed by manual modifi- at a final concentration of 6 × 105/mL as described (17, 29), with slight cation. After exclusion of ambiguously aligned positions, two datasets were modifications. Cells were sampled at the indicated times after exposure to produced. One contained all but mammalian cholesteryl SULTs that cannot the encystation medium. The number and the percentage of cysts formed be aligned with other SULTs, and the other contained only Entamoeba

Mi-ichi et al. PNAS | Published online May 18, 2015 | E2889 Downloaded by guest on October 1, 2021 SULTs. Both datasets were subjected to maximum likelihood (ML) phyloge- points were washed three times with PBS. Cells then were suspended in 3 mL netic analyses with the LG+ Γ model (Fig. S3B) or the LG+F+ Γ model (Fig. of RNAiso Plus and disrupted using a Dounce homogenizer (∼300 strokes) S3C) using RAxML version 8.0.0 (34). An appropriate model for each analysis until the majority of cysts were lysed. was selected on the basis of Akaike Information Criterion using the AMINOSAN program (35). The percent bootstrap support value from the ML Cell-Growth Assay. The cell-growth assay was performed as described pre- analyses with 100 bootstrap replicates was mapped onto each internal viously (36) except that the starting cell number was set at 3,000 tropho- branch of the optimal ML tree for an original dataset. zoites/mL, and growing cells were counted at 48, 72, and 96 h.

E histolytica Preparation for . Gene-Silenced Transformants. Plasmids for gene Cytopathic Activity Assay. The level of destruction of CHO cell monolayers was silencing were constructed by PCR amplification of 380- to 450-bp fragments assessed as described previously (37), with minor modifications. Briefly, corresponding to the 5′ end of the ORF of E. histolytica SULT genes with confluent CHO cells prepared by overnight culture in 24-well plates at 37 °C appropriate primer sets (Table S4) (5, 13). Amplicons then were digested and in 5% CO were washed with Opti-MEM (Life Technologies) containing with StuI and SacI and were inserted into the corresponding sites of the 2 1 mg/mL ascorbic acid and 5 mg/mL cysteine. SULT6gs and control E. histo- psAP-2-gunma plasmid (5). Lipofectin transfection of E. histolytica tropho- lytica cells (5 × 104 or 1 × 105 cells), which had been suspended in 0.5 mL of zoites with the constructed plasmids, drug selection, and maintenance of selected transformants was performed as described previously (5). the medium described above, were added to each well of the 24-well plates. The plates were incubated under anaerobic conditions at 37 °C for 60 min. qRT-PCR. qRT-PCR was performed with SYBR Premix Ex Taq II (Takara Bio, Inc.) After the plates were placed on ice for 10 min to detach trophozoites from and RNA polymerase II (Rnapol) as a house-keeping reference gene. The the CHO cell monolayer, the number of CHO cells remaining in each well was cDNA template for PCR was prepared as follows: Total RNA of Entamoeba measured using WST-1 reagent (Roche Diagnostics) as described previ- trophozoites was extracted with RNAiso Plus (Takara Bio, Inc.), and cDNA ously (37). was synthesized using the ReverTra Ace qPCR RT Kit (Toyobo). Real-time PCR was performed with the AB PRISM 7000 Sequence Detection System (Ap- ACKNOWLEDGMENTS. We thank Dr. Ei’ichi Iizasa for valuable discussions plied Biosystems, Life Technology) using appropriate primer sets (Table S4) and technical assistance; Dr. Shinjiro Hamano (Department of Parasitology, and the following cycling conditions: 95 °C for 20 s, 40 cycles of 95 °C for 5 s Nagasaki University) for valuable discussions; and Ms. Ritsuko Yoshida and Ms. Shizuko Furukawa for technical assistance. This work was supported by and 60 °C for 30 s. All reactions were run in triplicate, including reverse Ministry of Education, Culture, Sports, Science and Technology of Japan transcriptase-minus and cDNA-minus controls. Quantification for each target Grants-in-Aid for Scientific Research 22890136, 24117517, and 26117719 (to Δ gene was determined by the Ct method with Rnapol as reference gene. For F.M.), 25460594 (to H.Y.), 24390256 (to H.H.), and 23117001, 23117005, and analysis of E. invadens cysts, total RNA was prepared as described previously 14506236 (to T.N.) and by Institute of Tropical Medicine Cooperative Research (14), with minor modifications. Briefly, the cells harvested at various time Grants 2011 and 2013 (to F.M.).

1. Mi-ichi F, Abu Yousuf M, Nakada-Tsukui K, Nozaki T (2009) Mitosomes in Entamoeba 19. Bapteste E, et al. (2002) The analysis of 100 genes supports the grouping of three histolytica contain a sulfate activation pathway. Proc Natl Acad Sci USA 106(51): highly divergent amoebae: Dictyostelium, Entamoeba, and Mastigamoeba. Proc Natl 21731–21736. Acad Sci USA 99(3):1414–1419. 2. van der Giezen M (2009) Hydrogenosomes and mitosomes: Conservation and evolu- 20. Gill EE, et al. (2007) Novel mitochondrion-related organelles in the anaerobic amoeba – tion of functions. J Eukaryot Microbiol 56(3):221–231. Mastigamoeba balamuthi. Mol Microbiol 66(6):1306 1320. 3. Müller M, et al. (2012) Biochemistry and evolution of anaerobic energy metabolism in 21. Nývltová E, et al. (2013) NIF-type iron-sulfur cluster assembly system is duplicated and eukaryotes. Microbiol Mol Biol Rev 76(2):444–495. distributed in the mitochondria and cytosol of Mastigamoeba balamuthi. Proc Natl – 4. Makiuchi T, Nozaki T (2014) Highly divergent mitochondrion-related organelles in Acad Sci USA 110(18):7371 7376. anaerobic parasitic protozoa. Biochimie 100:3–17. 22. Chávez LA, Balamuth W, Gong T (1986) A light and electron microscopical study of a 5. Mi-ichi F, Makiuchi T, Furukawa A, Sato D, Nozaki T (2011) Sulfate activation in mi- new, polymorphic free-living amoeba, Phreatamoeba balamuthi n. g., n. sp. J Protozool 33(3):397–404. tosomes plays an important role in the proliferation of Entamoeba histolytica. PLoS 23. Das S, et al. (2002) Lipid metabolism in mucous-dwelling amitochondriate protozoa. Negl Trop Dis 5(8):e1263. Int J Parasitol 32(6):655–675. 6. Ralston KS, Petri WA, Jr (2011) Tissue destruction and invasion by Entamoeba histo- 24. Clark CG, et al. (2007) Structure and content of the Entamoeba histolytica genome. lytica. Trends Parasitol 27(6):254–263. Adv Parasitol 65:51–190. 7. Patron NJ, Durnford DG, Kopriva S (2008) Sulfate assimilation in eukaryotes: Fusions, 25. van der Wulp MY, Verkade HJ, Groen AK (2013) Regulation of cholesterol homeo- relocations and lateral transfers. BMC Evol Biol 8:39. stasis. Mol Cell Endocrinol 368(1-2):1–16. 8. Nývltová E, et al. (2015) Lateral gene transfer and gene duplication played a key role 26. Abumrad NA, Davidson NO (2012) Role of the gut in lipid homeostasis. Physiol Rev in the evolution of Mastigamoeba balamuthi hydrogenosomes. Mol Biol Evol 32(4): 92(3):1061–1085. 1039–1055. 27. Hanyu O, et al. (2012) Cholesterol sulfate induces expression of the skin barrier 9. Negishi M, et al. (2001) Structure and function of sulfotransferases. Arch Biochem protein filaggrin in normal human epidermal keratinocytes through induction of Biophys 390(2):149–157. RORα. Biochem Biophys Res Commun 428(1):99–104. 10. Fuda H, Lee YC, Shimizu C, Javitt NB, Strott CA (2002) Mutational analysis of human 28. Bakker-Grunwald T, Geilhorn B (1992) Sulfate metabolism in Entamoeba histolytica. hydroxysteroid sulfotransferase SULT2B1 isoforms reveals that exon 1B of the Mol Biochem Parasitol 53(1-2):71–78. SULT2B1 gene produces cholesterol sulfotransferase, whereas exon 1A yields preg- 29. Jeelani G, et al. (2012) Metabolic profiling of the protozoan parasite Entamoeba nenolone sulfotransferase. J Biol Chem 277(39):36161–36166. invadens revealed activation of unpredicted pathway during encystation. PLoS ONE 11. Strott CA, Higashi Y (2003) Cholesterol sulfate in human physiology: What’sitall 7(5):e37740. about? J Lipid Res 44(7):1268–1278. 30. Makiuchi T, Mi-ichi F, Nakada-Tsukui K, Nozaki T (2013) Novel TPR-containing subunit 12. Kohjitani A, Fuda H, Hanyu O, Strott CA (2006) Cloning, characterization and tissue of TOM complex functions as cytosolic receptor for Entamoeba mitosomal transport. expression of rat SULT2B1a and SULT2B1b steroid/sterol sulfotransferase isoforms: Sci Rep 3:1129. Divergence of the rat SULT2B1 gene structure from orthologous human and mouse 31. Husain A, et al. (2010) Metabolome analysis revealed increase in S-methylcysteine and phosphatidylisopropanolamine synthesis upon L-cysteine deprivation in the anaero- genes. Gene 367:66–73. bic protozoan parasite Entamoeba histolytica. J Biol Chem 285(50):39160–39170. 13. Bracha R, Nuchamowitz Y, Anbar M, Mirelman D (2006) Transcriptional silencing of 32. Nagasawa K, Ogamo A, Anryu K (1972) Fluorescent colouration or organic anions multiple genes in trophozoites of Entamoeba histolytica. PLoS Pathog 2(5):e48. with pinacryptol yellow on cellulose layers. J Chromatogr A 67(1):113–119. 14. Aguilar-Díaz H, Carrero JC, Argüello-García R, Laclette JP, Morales-Montor J (2011) 33. Larkin MA, et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23(21): Cyst and encystment in protozoan parasites: Optimal targets for new life-cycle in- 2947–2948. – terrupting strategies? Trends Parasitol 27(10):450 458. 34. Stamatakis A (2014) RAxML version 8: A tool for phylogenetic analysis and post- 15. Siegesmund MA, Hehl AB, van der Giezen M (2011) Mitosomes in trophozoites and analysis of large phylogenies. Bioinformatics 30(9):1312–1313. – cysts of the reptilian parasite Entamoeba invadens. Eukaryot Cell 10(11):1582 1585. 35. Tanabe AS (2011) Kakusan4 and Aminosan: Two programs for comparing non- 16. Bradford CM, Denver MC, Cranfield MR (2008) Development of a polymerase chain partitioned, proportional and separate models for combined molecular phylogenetic – reaction test for Entamoeba invadens. J Zoo Wildl Med 39(2):201 207. analyses of multilocus sequence data. Mol Ecol Resour 11(5):914–921. 17. De Cádiz AE, Jeelani G, Nakada-Tsukui K, Caler E, Nozaki T (2013) Transcriptome 36. Penuliar GM, Furukawa A, Sato D, Nozaki T (2011) Mechanism of trifluoromethionine analysis of encystation in Entamoeba invadens. PLoS ONE 8(9):e74840. resistance in Entamoeba histolytica. J Antimicrob Chemother 66(9):2045–2052. 18. Ehrenkaufer GM, Haque R, Hackney JA, Eichinger DJ, Singh U (2007) Identification of 37. Furukawa A, Nakada-Tsukui K, Nozaki T (2012) Novel transmembrane receptor in- developmentally regulated genes in Entamoeba histolytica: Insights into mechanisms volved in phagosome transport of lysozymes and β-hexosaminidase in the enteric of stage conversion in a protozoan parasite. Cell Microbiol 9(6):1426–1444. protozoan Entamoeba histolytica. PLoS Pathog 8(2):e1002539.

E2890 | www.pnas.org/cgi/doi/10.1073/pnas.1423718112 Mi-ichi et al. Downloaded by guest on October 1, 2021