Trafficking and organization of cyst wall components into a robust biphasic structure in Entamoeba
Tam Kha Vo, Hiroki Yoshida, Fumika Mi-ichi
TL;DR
This study reveals how components of the Entamoeba cyst wall are organized into a two-layered structure during parasite development.
Contribution
The study identifies the temporal and spatial dynamics of cyst wall component synthesis and localization, revealing a biphasic structure.
Findings
Jacob1/3 lectins are post-translationally modified and localized to the cyst wall before Jessie3a and Chitinase1/4.
Jessie lectin localization is disrupted by a chitinase inhibitor, indicating a distinct mechanism compared to Jacob lectins.
Immunoelectron microscopy confirms a biphasic structure of the Entamoeba cyst wall with inner and outer layers.
Abstract
Entamoeba histolytica, a protozoan parasite, causes amebiasis. Amebiasis is mainly transmitted by oral ingestion of cysts. Cysts are produced in the large intestine of humans from proliferative trophozoites by a cell differentiation process called encystation. The Entamoeba cyst wall consists of chitins and proteins that include chitinase and Jacob and Jessie lectins. During encystation, these components are synthesized and layered around encysting Entamoeba cells. The structures of these components are well studied; however, the detailed timings of their synthesis (the transcription of the encoding genes and the translation of the resulting mRNAs) and of changes in their localization during encystation are poorly understood. Here, we performed quantitative RT-PCR and an approach combining western blotting and immunofluorescence, confocal, and immunoelectron microscopy to analyze…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Fig 1
Fig 2
Fig 3
Fig 4
Fig 5
Fig 6
Fig 7
Fig 8
Fig 9
Fig 10- —Ministry of Education, Culture, Sports, Science and Technology of Japan
- —Ministry of Education, Culture, Sports, Science and Technology of Japan
- —Ministry of Education, Culture, Sports, Science and Technology of Japan
- —AMED Japan Program for Infectious Diseases Research and Infrastructure
- —AMED Japan Program for Infectious Diseases Research and Infrastructure
- —Kaketsuken Foundation
- —http://dx.doi.org/10.13039/501100008971Nagase Science Technology Foundation
- —http://dx.doi.org/10.13039/100016925Takeda Foundation
- —http://dx.doi.org/10.13039/100007428Naito Foundation
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAmoebic Infections and Treatments · Parasites and Host Interactions · Parasitic infections in humans and animals
Introduction
Entamoeba histolytica is a protozoan parasite that causes amebiasis, which, globally, is the third deadliest parasitic disease [1]. This disease is mainly transmitted by oral ingestion of cysts via food or water that has been contaminated with fecal matter from infected individuals. Upon reaching the small intestine, cysts hatch into trophozoites and then move to the large intestine, where they proliferate [2,3]. The E. histolytica life cycle is maintained by alternating between these two forms, a trophozoite and a cyst. Differentiation from trophozoites into cysts is called encystation. In the large intestine, some trophozoites encyst to form mature cysts. Mature cysts are surrounded by a cyst wall, which is specifically synthesized by encysting E. histolytica cells to enable survival in harsh environments outside, as well as inside the human host [4,5].
Entamoeba invadens, a reptilian parasite, has been adopted as an encystation model for E. histolytica because laboratory strains of E. histolytica do not encyst after adaptation to in vitro culture conditions [5]. A laboratory strain of E. histolytica was recently shown to encyst in vitro [6]; however, this system has not been widely employed, and we were unable to reproduce this result. The Entamoeba cyst wall is composed of chitins and proteins, including chitinase and Jacob and Jessie lectins [7–9]. These three proteins all have a chitin binding domain (CBD) and are stage-specifically synthesized during encystation and layered around encysting Entamoeba cells [8,10]. Chitinase is a hydrolase that acts on chitin, a β-1,4-linked polymer of N-acetyl D-glucosamine. Chitinase genes are present in a broad range of organisms, such as humans, insects, plants, fungi, and bacteria [11]. We recently showed that chitinase activity is required to form a mature rounded Entamoeba cyst wall [12], which confers dormancy to the parasite, an essential characteristic for this parasite to be transmitted to a new host. One and four chitinase genes are encoded in E. histolytica and E. invadens genomes, respectively [12]. All five of these chitinases, EhCht and EiCht1–4, possess conserved catalytic domains, which are phylogenetically related; however, two E. invadens chitinases (EiCht2 and -3) lack a CBD [12]. The CBDs in EhCht and EiCht1 and -4 have eight cysteines and aromatic amino acids that are conserved and are therefore termed 8-Cys CBDs [13].
In contrast to chitinase, a ubiquitous protein in organisms, Jacob and Jessie lectins are unique to Entamoeba. Jacob lectins are composed of a signal peptide and tandemly arrayed ~50 amino acid CBD repeats. All E. histolytica and E. invadens Jacob lectins (EhJacob1 and 2 and EiJacob1–7, respectively) have two to six CBDs, each of which contains six conserved cysteines and conserved aromatic amino acids, and are therefore termed 6-Cys CBDs [9]. Jacob lectins are hypothesized to cross-link chitin fibrils via their CBDs [9]. Meanwhile, all Jessie lectins possess a signal peptide and 8-Cys CBDs similar to those in chitinases. In addition, Jessie 3 (EhJessie3 and EiJessie3a and -3b) also possess a C-terminal conserved domain, which is called the daub domain [8]. Jessie3 lectin is suggested to act as a mortar or a daub, which makes the Entamoeba cyst wall impenetrable to small molecules [8]. Previously, the “wattle and daub model” was proposed for the Entamoeba cyst wall [8]. In this model, chitin fibrils, likely cross-linked by Jacob lectins (the wattle or sticks), are constructed prior to the addition of Jessie3 lectins (the mortar or daub). However, further evidence is required to fully support this model, including detailed analyses of the gene transcription and mRNA translation of the chitins, chitinase, and Jacob and Jessie lectin cyst wall components, and of the changes in their localization and functional status during Entamoeba encystation.
In this study, we analyzed E. invadens cells that were sampled with short time intervals during encystation using quantitative RT-PCR and an approach combining western blotting and immunofluorescence and confocal microscopy. We also employed immunoelectron microscopy to visualize the subcellular localization of these proteins and chitins in encysting cells. Furthermore, we investigated the effect of D-B-09, an Entamoeba chitinase inhibitor [12], on the localizations of Jacob and Jessie lectins.
Results
Transcriptional profiles of EiJacob, EiJessie, and EiChtinase cyst wall protein genes during encystation
In E. histolytica and E. invadens genomes, two and seven Jacob genes are present, respectively [S1 Fig, AmoebaDB (https://amoebadb.org/amoeba/app)]. Meanwhile, there are four and six Jessie genes in E. histolytica and E. invadens genomes, respectively (EhJessie1–4 and EiJessie1a–d, -3a, and -3b) (S1 Fig, AmoebaDB; [9,13]). EiJessie1a was previously reported [9] but was not registered in AmoebaDB at that time. EiJessie1d was registered as a hypothetical protein in AmoebaDB (ID, EIN_526480) and was confirmed in this study to be a EiJessie family member. Different open reading frames for EiJessie1b were registered in NCBI (https://www.ncbi.nlm.nih.gov/; Accession No., ABC59326) and AmoebaDB (ID, EIN_096900), and the open reading frame registered in NCBI (Accession No., ABC59326) was confirmed by RT-PCR-mediated cloning and sequencing in this study to encode EiJessie1b (S1 Fig). In addition, the daub domain of Jessie 3 (EhJessie3 and EiJessie3a and -3b) was revealed to be conserved in proteins encoded by Pelomyxa schiedti, Blastocystis, and Amoeboaphelidium protococcarum (S2 Fig). However, as these proteins are annotated as hypothetical proteins, the precise role of these proteins and domains are still unclear.
Most, but not all, members of the EiJacob and EiJessie family genes have previously been analyzed by RT-PCR [9] and transcriptomic [14,15] analyses during E. invadens encystation, and were shown to be upregulated. To comprehensively analyze all members of the EiJacob and EiJessie family genes during E. invadens encystation, we performed qRT-PCR using cells collected at the time points indicated in Fig 1. Transcription of all EiJacob and EiJessie family genes sharply increased immediately after encystation induction (Fig 1). The transcriptional profile of EiJessie1d gene, a newly identified EiJessie family gene, was similar to those of EiJessie1a, -1b, and -1c genes (Fig 1). The overall transcription trends of these genes were consistent with each other, but the degree and timing of upregulation fluctuated markedly across experiments (S3 Fig). Similarly, the overall trend was consistent with trends in previous studies, but the extent and timing of upregulation were slightly different. For instance, the peaks of upregulation of EiJacob and EiJessie family genes were 8–16 hr after encystation induction (Fig 1) whereas those were around 24 hr after induction [14,15]. We assume these fluctuations arise because encysting cell-specific genes are transcribed at very low levels in trophozoites (at 0 hr), and that the expression levels reached and the timing of upregulation vary with cell conditions [12]. Importantly, at 8–24 hr after encystation induction, the increased gene expressions were notable (Fig 1), and coincidently the cyst wall structure started to be observed by Calcofluor staining and electron microscopy [16].
Transcriptional changes of the genes encoding EiJacob and EiJessie 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 hr. Representative results from two independent experiments with duplicates are shown. The results of a second trial are shown in S3 Fig. (A) EiJacob1 to -7. (B) EiJessie family: 1a to -d, -3a, and -3b.
Expression of EiJacob1, EiJacob3, EiJessie3a, and EiChN1 and their post-translational modification during encystation
We then investigated the translational profiles of the major cyst wall proteins to confirm their transcriptional profiles and to analyze changes in levels and in post-translational modifications of the encoded proteins during the course of encystation. In this analysis, western blot analyses were performed using antisera against EiJacob1, EiJacob3, EiJessie3a, and EiChitinase1 (EiCht1). These proteins were previously determined by mass spectrometry to constitute ~90% of the proteins in E. invadens cyst walls [9]. The four required antigens were all produced in E. coli and prepared as affinity-purified His-tagged recombinant proteins (see Table 1). Recombinant EiJacob1 and -3 were prepared as full-length without signal peptide whereas rEiJessie3a and rEiCht1 were prepared as a truncated form (rEiJessie3a-N, ^19^Phe–^230^Val) and [rEiCht1-catalytic domain (rEiCht1-CD), ^151^Lys–^450^Gln], respectively, due to being unable to clone amplicons and low yield and instability of recombinant protein (see Table 1).
Table 1: List of cyst wall proteins analyzed in this study.
In addition to these four antigens, other rEiJacob, rEiJessie, and rEiCht proteins were also similarly prepared. The deduced molecular masses of each native and recombinant protein are listed in Table 1. Among cyst wall protein gene families in E. invadens (7 Jacobs, 6 Jessies, and 4 chitinases), EiJacob1–5 and EiJessie1a, -1c, -1d, and -3b, and EiCht1–4 were prepared as affinity-purified His-tagged recombinant full-length proteins without signal peptide. In contrast, we were unable to prepare rEiJacob6 and 7 because their corresponding amplicons could not be cloned in the expression plasmid, pColdI (TAKARA, Kyoto, Japan). We were also unable to prepare rEiJessie1b because the level of the recombinant protein was too low to be purified. The reactivities of the four prepared antisera against rEiJacob1 and -3, rEiJessie3a-N, and rEiCht1-CD to the relevant recombinant proteins were analyzed by western blotting (S4 Fig). The reactivities of the corresponding pre-immune sera to the recombinant protein used as antigen were also analyzed (S5 Fig). Although some cross-reactivities among the family members were detected, the highest reactivity of each antiserum was observed for the recombinant protein used as antigen. Furthermore, the corresponding pre-immune sera did not react to each antigen. Therefore, these four antisera, which hereafter are designated anti-EiJacob1 and -3, -EiJessie3A-N, and -EiCht1-CD, were used to analyze the protein profiles of encysting E. invadens cells.
At 8 hr after induction, ~ 50 kDa EiJacob1 and ~44 kDa EiJacob 3 were detected (Fig 2A and 2B), which are close to the deduced molecular masses of native mature forms (Table 1). From 16 to 72 hr, the bands of EiJacob1 and -3 were shifted to ~80 kDa and ~60 kDa molecular masses, respectively (Fig 2A and 2B), indicating that both are post-translationally modified, which is consistent with Jacobs being glycoproteins [17]. The amounts of modified Jacob1 and Jacob3 at 16, 24, and 72 hr were lower than those of unmodified Jacob1 and Jacob3 at 8 hr. The amounts of modified Jacob1 and Jacob3 did not change from 16 hr onwards (Fig 2A,and 2B). These results indicate that only limited amounts of Jacob1 or Jacob3 are modified and that the modified forms of Jacob1 and Jacob3 are stable during encystation. The main bands detected by anti-EiJacob1 and -3 in encysting cells corresponded to EiJacob1 and -3 (Fig 2A and 2B and Table 1). These bands were not detected by corresponding pre-immune sera (S5 Fig). Although anti-EiJacob3 weakly cross reacted with rEiJacob1 and rEiJacob2 (S4B Fig), the corresponding bands of native proteins could almost not be detected (Fig 2B). Furthermore, truncated forms of EiJacob1 and -3 were not detected (Fig 2A and 2B). These results are inconsistent with a previous study that showed Jacob lectins cleaved at conserved sites between CBDs by an endogenous Cys protease [9]. We suspect that the cleavage of Jacob lectins observed in the previous study [9] occurred during sample preparation and that such cleavage did not occur in this study, probably because of the addition of a protease inhibitor cocktail during cell lysis. Meanwhile, the major bands at molecular masses of ~75 kDa and ~65 kDa, which were respectively, seen by anti-EiJessie3A-N, and -EiCht1-CD, started to be detected at 16 hr after encystation induction and their levels increased until 24 hr (Fig 2C and 2D and Table 1). Although the predicted molecular masses of EiJessie3a and -3b are very similar (Table 1), in view of very weak cross-reactivity of anti-EiJessie3A-N to rEiJessie3b (S4C Fig), the major band detected by anti-EiJessie3a-N is considered to be EiJessie3a. Meanwhile, the major band detected by anti-EiCht1-CD in encysting E. invadens cells corresponded to the deduced molecular masses of EiCht1 and -4, which are also very similar (Table 1). All these bands were not detected by corresponding pre-immune sera (S5 Fig). Although anti-EiCht1-CD cross-reacted with EiCht2 and -3 (S4D Fig), the corresponding bands were almost undetectable (Fig 2D); therefore, the major band detected by anti-EiCht1-CD (Fig 2D) is considered a mixture of EiCht1/4. Furthermore, the molecular masses of the EiJessie3a and EiCht1/4 bands were slightly higher than the deduced molecular masses of native mature proteins (Fig 2C and 2D). These shifts may result from the modification of EiJessie3a and EiCht1/4 or may be attributed to their structures in the presence of SDS.
Western blots of encysting E. invadens cells at 0, 8, 16, 24, and 72 hr after induction.Rat anti-rEiJacob1, and guinea pig anti-rEiJacob3, anti-rEiJessie3a-N, and anti-rEiChN1-CD antisera (A–D). Arrows indicate native proteins recognized by each antiserum. Arrowheads indicate the sizes of predicted native proteins. The positions of EiJacob1, EiJacob3, and EiCht1-4 are predicted from the corresponding full-length recombinant protein without signal peptide. The position of EiJessie3a is predicted from the full-length rEiJessie3b protein without signal peptide as EiJessie3a could not be prepared, and the calculated molecular masses of rEiJessie3a and rEiJessie3b are highly similar. Western blots showing the antigen specificities of these four antisera are shown in S3 Fig. Molecular mass standards are indicated on the left. Representative results from two independent experiments are shown.
These results, together with those of a previous study [9] indicate that the protein synthesis and post-translational modification (probably glycosylation) of Jacob family members precede the protein synthesis of Jessie and chitinase family members.
EiJacob1 and EiJacob3 co-localize with chitin while EiJessie3a does not
To analyze the localization changes of each cyst wall protein during encystation, indirect immunofluorescence assays of encysting E. invadens cells at the time points indicated in Fig 3 were performed using anti-EiJacob1 and -3, -EiJessie3a-N, and -EiCht1-CD. Note that Fig 3 shows representative differently progressed cyst-forming cells at each time point because synchronization of the in vitro culture is not currently possible (S6 Fig). Strong fluorescence signals of EiJacob1 and -3 and EiJessie3a started to be detected on the cyst wall from 16 hr and 24 hr, respectively, after induction of encystation (Fig 3). By contrast, fluorescence signals corresponding to EiCht1/4 were not detected. Furthermore, no fluorescence signals were detected when all corresponding pre-immune sera of the above four antisera were used (S7 Fig). During encystation, the EiJacob1 and -3 signals overlapped substantially with that of chitin, which was visualized with WGA-Alexa647 (Fig 3). Co-localization of EiJacob1 and -3 with chitin was confirmed by confocal microscopy [Fig 4A; Pearson correlation coefficients, 0.91 and 0.95, respectively (S8 Fig)]. In addition, faint intracellular WGA-Alexa647 signals were also detected at early time points (8 hr in Fig 3). WGA also binds to the monomeric form of N‑acetyl‑D‑glucosamine (GlcNAc), the precursor of chitin and substrate of chitin synthase; therefore, these weak intracellular signals may reflect binding to GlcNAc. In contrast, strong signals on the cell surface reflect binding to the GlcNAc polymer.
Time course analysis of the cellular localization of EiJacob1, EiJacob3, and EiJessie3a during differentiation from trophozoites to cysts.Immunofluorescence images of encysting E. invadens cells using rat anti-rEiJacob1 (A), guinea pig anti-rEiJacob3 (B), and guinea pig anti-rEiJessie3a-N (C) antisera. Three independent cells are shown at 72 hr in (C). Each protein is visualized by either Alexa Fluor 488 conjugated secondary antibodies. Chitin is counter stained with Wheat Germ Agglutinin (WGA) Alexa flour 647. Images are obtained under x400 magnification (x40 objective lens) using Zeiss Axio Imager. M2 (Carl Zeiss). Scale bars indicate 10 µm.
Confocal microscopy analysis of E. invadens cells.(A) Immunofluorescence images of encysting E. invadens cells at 24 hr after encystation induction using rat anti-rEiJacob1, guinea pig anti-rEiJacob3, and guinea pig anti-rEiJessie3a-N antisera. Each protein is visualized by either Alexa Fluor 488 conjugated secondary antibodies. Chitin is counter stained with WGA Alexa flour 647. Images are obtained under x1000 magnification (x100 objective lens) using a confocal microscope (A1R HD25; Nikon, Japan). Scale bars indicate 10 μm. (B) Time course analysis of the cellular localization of EiJessie3a with short time interval sampling. Immunofluorescence images of encysting E. invadens cells using rat anti-rEiJessie3a-N antiserum. Cells were collected at 16, 20, and 24 hr after encystation induction, and stained cells were observed under a confocal microscope (AXR; Nikon). Two independent stained cells are shown. Protein is visualized by Alexa Fluor 488 conjugated secondary antibody. Chitin is counter stained with WGA Alexa flour 647. Images are obtained under x1000 magnification (x100 objective lens) using a confocal microscope (AXR). Scale bars indicate 10 µm.
In contrast to EiJacob1 and -3, EiJessie3a did not co-localize with chitin during encystation. EiJessie3a localization varied among analyzed cells. In some cells, the signal was detected across the whole cell surface, while in other cells, the surface was only partially stained with different signal intensities. Confocal microscopy showed that the EiJessie3a signal was detected as a dot beneath or on the surface of the chitin fibers [Fig 4A; Pearson correlation coefficients, 0.60 (S8 Fig)]. Immunofluorescence and confocal microscopy of encysting cells sampled with short time intervals revealed changes in the localization of Jessie3a on the cyst wall surface (Fig 4B). At 16 hr, only partially stained cells were dominant. At 20 hr, cells in which the EiJessie3a signal surrounded the chitin fibers became dominant. At 24 hr, the EiJessie3a signals of dominant cells overlapped those of chitin. In these cells, the EiJessie3a signal became faint, although the levels of EiJessie3a were stable (Fig 2C), indicating that EiJessie3a was localized within the chitin fibers and that the antibodies were unable to reach it. These results indicate that EiJessie3a initially appears on certain areas of the cell surface, then covers the entire cell surface, and then penetrates the chitin fibers.
In D-B-09 treated cells, EiJacob1 and EiJacob3 localize on the chitin wall, whereas EiJessie3a localizes throughout the cell, particularly on the inner surface of the chitin wall
The localization of EiJacob1 and EiJacob3 on, and EiJessie3a near, chitin fibers prompted us to consider whether chitin fiber compression affected their localizations. To address this question, the effect of D-B-09, an Entamoeba chitinase inhibitor, on the localizations of EiJacob1 and -3, and EiJessie3a was investigated. Inhibition of the chitinase impairs E. invadens cyst wall compression, resulting in the formation of abnormal cyst walls, most characteristically a pot-like structure [12]. In D-B-09-treated pot-like cells, the EiJacob1 and -3 signals still overlapped with that of chitin fibers [Fig 5A and 5B, 5D-5B-09; Pearson correlation coefficients, 0.87 and 0.88, respectively (S8 Fig)], indicating that EiJacob1 and -3 are co-localized regardless of the state of chitin fiber compression. In contrast, the EiJessie3a signal was affected by D-B- 09 treatment. In D-B-09-treated pot-like cells, the EiJessie3a signal accumulated on the periphery of the cell membrane [Fig 5C, 5D-5B-09; Pearson correlation coefficients, 0.40 (S8 Fig)], indicating that the localization relationship between EiJessie3a and chitin fibers is affected by the state of chitin fiber compression.
Effects of D-B-09 on the localization of EiJacob1, EiJacob3, and EiJessie3a at 24 h after induction of encystation.Immunofluorescence images of E. invadens cells treated with 100 μM D-B-09 (lower panels) or DMSO (upper panels) for 24 hr after encystation induction using rat anti-rEiJacob1 (A), guinea pig anti-rEiJacob3 (B), and anti-rEiJessie3a-N (C) antisera. Each protein is visualized by either Alexa Fluor 488 conjugated secondary antibodies. Chitin is counter stained with WGA Alexa flour 647. Images are obtained under x1000 magnification (x100 objective lens) using a confocal microscope (A1R HD25; Nikon, Japan). Scale bars indicate 10 μm.
Immunoelectron microscopy
To investigate the subcellular localization of cyst wall components in encysting E. invadens cells, we performed immunoelectron microscopy for Jacob, Jessie, Chitinase, and chitin using ultrathin sections. We visualized each component in cells at 16, 24, and 72 hr after induction of encystation (Figs 6, 7, 8, 9). At the early stage of cyst formation (16 and 24 hr), gold colloidal particles (GCPs) conjugated secondary antibody that bind to EiJacob1 via anti-EiJacob1 were distributed all around the cyst wall. GCPs were detected in areas with relatively high electron density (arrows in 16 and 24 hr in Fig 6). Additionally, GCPs were detected in small electron-lucent vesicles in the cytosol (arrowheads in Fig 6), indicating that Jacobs are transported to the cell membrane surface by vesicle trafficking. At the late stage (72 hr in Fig 6) when mature cysts predominate in in vitro encystation culture, almost all GCPs were distributed in electron dense areas in the outer layer. Importantly, electron dense areas were consistently separated and distinguishable from the cell membrane by electron light areas (72 hr in Fig 6), indicating that the Entamoeba cyst wall consists of two phases, and that Jacobs are enriched in outer layer (electron dense area) of the cyst wall. Of note, at 72 hr, almost all GCPs conjugated with WGA that bind to chitin were also highly enriched in the electron-dense area (72 hr in Fig 7), indicating that chitin is enriched in the outer layer of the cyst wall. This result, together with evidence from immunofluorescence and confocal microscopy, indicates that Jacobs and chitin fibers are localized together in the outer layer of the Entamoeba cyst wall. Furthermore, consistent with the fluorescence microscopy results (Fig 3), WGA-GCP signals, probably from binding to the chitin synthase substrate, GlcNAc, were also detected in cytoplasm at early time points (arrowheads in 16 and 24 hr in Fig 7)
Immunoelectron microscopy using anti-EiJacob1.Images from two independent cells are shown. In each set, right panel is higher (x16800) magnifications of left panel (x4210).
Immunoelectron microscopy using WGA-GCPs.Images from two independent cells are shown. In each set, right panel is higher (x16800) magnifications of left panel (x4210).
Immunoelectron microscopy using anti-EiJessie3a-N.Images from two independent cells are shown. In each set, right panel is higher (x16800) magnifications of left panel (x4210).
Immunoelectron microscopy using anti-EiCht1-CD.Images from two independent cells are shown. In each set, right panel is higher (x16800) magnifications of left panel (x4210).
In contrast to EiJacob1, in the early stage of cyst formation (16 hr), only a few GCPs conjugated secondary antibodies that bind EiJessie3a and EiCht1/4 via anti-EiJessie3a-N and -EiCht1-CD, respectively, were detected outside the cell membrane (16 hr in Figs 8 and 9). As cyst formation proceeds (at 24 hr), GCPs were detected in the area with relatively high electron density in cyst wall (arrows in 24 hr in Figs 8 and 9). Furthermore, electron dense vesicles in which GCPs were enriched were clustered near the cell membrane (arrowheads in 24 and 72 hr in Figs 8 and 9). These structures were not observed at 16 hr (16 hr in Figs 8 and 9). Notably, only one or zero clusters were observed per section, indicating that Jessie and chitinase are secreted together. These results suggest that Jessies and chitinases are transported to a certain position of the cell membrane by vesicle trafficking and secreted together.
Discussion
The formation of round, cyst-walled cells is a crucial process for Entamoeba to maintain its parasitic life cycle because mature cysts are the sole form that can be transmitted to a new host. The cyst wall is an essential structure that protects Entamoeba against environmental stresses. It is composed of chitin fibrils, bundles of a β-1,4-linked polymer of N-acetyl D-glucosamine, and chitin-binding proteins, Jacob and Jessie lectins, and chitinases. Here, to elucidate the molecular dynamics of cyst wall formation in Entamoeba, including the timing of gene expression and mRNA translation, and localization changes of these three groups of proteins during the encystation, we analyzed time course experiment samples of encysting E. invadens cells using an approach combining western blotting and immunofluorescence and -electron and confocal microscopy.
This approach revealed that in the mature Entamoeba cyst wall, Jacob1 and Jacob3 exist as post-translationally modified forms, probably through glycosylation. These post-translational modifications occur at 8–16 hr after the first appearance of Jacob1 and Jacob3 proteins. As the amounts of modified Jacob1 and Jacob3 proteins increase, and as cyst wall synthesis proceeds, these proteins combine with the surface of encysting cells. Modified Jacob1 and Jacob3 stably exist in the cyst wall until mature cysts are formed. In these cells, Jacob1 and Jacob3 are strongly co-localized with chitin fibers. Furthermore, this strong co-localization is not disrupted by D-B-09 treatment, which deforms the compressed structure of the cyst wall by inhibiting chitinases [12]. We therefore suggest that modified Jacob1 and Jacob3 act as a chitin carrier and/or as principal components to maintain the structure of chitin fibrils in Entamoeba cyst formation, and that modification of Jacobs may be required for these important roles (Fig 10). Meanwhile, in the early stage of encystation, in which the cyst wall structure is not visible, Jacob1 and Jacob3 are observed as small dots in the cytoplasm. Furthermore, because we used immunoelectron microscopy on ultrathin sections, the possibility that chitin fibers might trap antibodies was eliminated, allowing us to analyze the intra‑ and extracellular localization of these proteins. Immunoelectron microscopy revealed that Jacob1 was localized within small electron-lucent vesicles in the cytosol, whereas Jessie3a and Cht1 were enriched in electron-dense vesicles that were clustered near the cell membrane. These results indicate that the secretion of Jacob and Jessie and Chitinase occur via different trafficking pathways. In Entamoeba, chitin synthase 1 and 2 are responsible for chitin synthesis during encystation [18] and chitin synthase 1 synthesizes chitin fibrils at the cell surface [7], but the mechanism underlying the release of synthesized chitin is not known. The primary structures of Entamoeba chitin synthase 1 and 2 show high similarity to that of Saccharomyces cerevisiae chitin synthase 1 [19]. This high degree of conservation led us to assume an analogous function of Entamoeba chitin synthases; that they are localized at the plasma membrane and synthesize chitin chains that are extruded across the plasma membrane (see Fig 10). Fluorescence and immunoelectron microscopy support our assumption because signals that probably reflect GlcNAc, the precursor of chitin, were detected in the cytoplasm directly beneath the plasma membrane. Once chitin chains are released outside the cell, Jacobs act as a chitin carrier and/or as principal components to maintain the structure of chitin fibrils. Although other members of the Entamoeba Jacob family were not analyzed in this study, structural conservation among the family members may lead to functional redundancy.
Schematic depiction of cyst wall formation in Entamoeba.ChS, chitin synthase; Cht, chitinase; JA1, Jacob1; JA3, Jacob3; and JE3a, Jessie3a.
In a previous study [7], a Jacob, corresponding to EiJacob2 (AmoebaDB ID number, EIN_230100), was detected on the surface of chitin wall-less or aberrant chitin walled cells, indicating that Jacob was deposited on the cell surface of cysts in a chitin independent manner. However, we detected neither EiJacob1 nor EiJacob3 on the surface of chitin wall-less cells. This inconsistency can be attributed mainly to low protein abundance and variety of chitin levels among encysting E. invadens cells. The amount of EiJacob2 is much lower than those of EiJacob1 and EiJacob3 [9]. Furthermore, we observed some encysting cells with no chitin wall on which EiJacob1 or EiJacob3 was localized on the surface. However, this observation was subsequently determined to be an experimental artifact because, as exposure time increased, signals for chitin walls became visible, probably because the level of chitins among the encysting cells varied over 100-fold [20]. In addition, previously described cysts with incomplete chitin walls [7] could not be observed. However, the reason for this inconsistency needs to be clarified.
Jessie3a and Cht1/4 proteins first appeared after encysting cells were surrounded by the complex containing chitin fibers and Jacobs. This finding is both consistent [8] and inconsistent [7] with previous studies. However, the timing of Jessie expression in these previous studies was determined by immunofluorescence assays, unlike the western blotting analysis used in this study. The differences among these three studies need to be explained. Jessies and chitinases are exist stably in the cyst wall until mature cysts are formed. In contrast to Jacob proteins, the level and localization pattern of Jessie3a varied in each encysting cell. This may reflect differences in the progression of cyst formation among encysting cells in our asynchronous in vitro cultures. In normal encysting cells, confocal microscopy revealed that Jessie appears on the surface of chitin walls and is localized as dots and then penetrates in and is colocalized with the chitin fibers. Of note, in abnormal encysting cells, Jessie3a was atypically accumulated in the space between the cell membrane and the cyst wall, resulting in unusual, strong fluorescence signals.
Inhibition of chitinase activity by D-B-09 treatment reveals differences in the roles of Jacobs and Jessies during the formation of Entamoeba cysts. Chitinase is required for Jessie to be transported to the cyst wall via compression of the chitin fibrils while Jacobs are transported in a chitinase-independent manner to their final destinations and localized in a complex with chitin fibrils. The unusual, strong Jessie3a signal observed in D-B-09-treated encysting Entamoeba cells may indicate abnormal accumulation of Jessie3a under the cyst wall, resulting in an unusual space between the cell membrane and the swollen cyst wall. This finding is further supported by the observation that the appearance of chitinase and Jessie proteins started at the same time during encystation. We were unable to elucidate a subcellular change in chitinase localization during Entamoeba encystation by indirect immunofluorescence because the anti-chitinase antiserum did not work. Anti-Jacobs and -Jessie antisera, in contrast, worked in this analysis, ruling out that chitin fibrils nonspecifically trap antibody. This assumption is further supported by the results from immunoelectron microscopy that anti-EiCht1-CD indeed worked on the ultrathin section samples. Previously, significant chitinase activity was detected in intact encysting E. invadens cells, not permeabilized cells, until mature cysts were formed [12]. These results indicate that chitinase is secreted and localized in the cyst wall and forms a complex with other cyst components (probably chitins, Jessies and/or Jacobs). Notably, immunoelectron microscopy revealed that chitinase and Jessie proteins are localized in similar structures, electron dense outer layer in cyst wall as well as electron dense vesicles that form clusters. These clustered structures are retained in mature cysts. Considering the high conservation of their N-terminal structures, chitinase and Jessies may interact and function cooperatively during cyst wall formation in Entamoeba.
Immunoelectron microscopy revealed that the mature cyst wall is composed of two layers, an electron-light inner layer and an electron-dense outer layer. The outer layer is enriched with chitin, Jacobs, Jessies, and chitinases. The inner layer lacks these components and appears to be a space. We assume that this space is generated by the removal of organophilic materials, such as lipids, during the organic solvent treatments of the immunoelectron microscopy protocol. This assumption is supported by our observation that the cyst wall components, chitin, Jacobs, and Jessie, were barely detectable by immunofluorescence microscopy of mature cyst samples that were fixed using standard conditions (4% paraformaldehyde for 10 min), in contrast to immature cyst samples. This observation indicates that the mature cyst wall structure is more easily destroyed by paraformaldehyde treatment than the immature cyst wall; therefore, for this study, we optimized the fixation conditions to be 2% paraformaldehyde for 3–5 min. Recently, we showed that very long chain dihydroceramides are stage specifically synthesized in encysting Entamoeba cells and accumulate at high levels in mature cysts [21]. It is therefore likely that very long chain dihydroceramides are the organophilic materials localized in the space between the cyst wall component-enriched area and the cell membrane. However, the precise localization of very long chain dihydroceramides in the cyst wall requires confirmation.
In conclusion, we provide several lines of evidence detailing the molecular dynamics of cyst wall components during Entamoeba encystation. Our findings indicate the functional links among the three kinds of cyst wall protein, chitinase and Jacob and Jessie lectins, with chitin fibrils (bundles of a β-1,4-linked polymer of N-acetyl D-glucosamine) during encystation. Our findings were consistent with the “wattle and daub model” [8] in that: (1) Jacob and chitin fibers are simultaneously localized on the cell surface, and (2) Jessie3a protein first appears after encysting cells are surrounded by chitin fibers and Jacobs. In addition, we provided new evidence that: (3) The expression levels and intra- and extracellular trafficking of Jessie and chitinase are highly similar, and (4) the cyst wall consists of a biphasic structure, an electron-light inner layer, and an electron-dense outer layer. The outer layer is enriched with chitin, chitinases, and Jacob and Jessie lectins, while the inner layer is not enriched with these components. We summarized our findings as schematic depiction of cyst wall formation in Entamoeba (Fig 10).
Materials and methods
Ethics statement
Production of antisera by SCRUM Inc. (Tokyo, Japan) was approved by Institutional Animal Care and Use Committee (Chairman Isamu Fukamachi) in Protein Purify Co., Ltd. This institute certifies that Protein Purify Co., Ltd., with which SCRUM Inc. partners, has formulated ethical regulations for conducting animal experiments from the perspective of animal welfare, with reference to the required guidelines.
Materials
DDD85646 (2,6-dichloro-4-[2-(1-piperazinyl)-4-pyridinyl]-N-(1,3,5-trimethyl-1H-pyrazol-4-yl)-benzenesulfonamide), which is identical to D-B-09 [ID number in Pathogen Box from Medicines for Malaria Venture (https://www.mmv.org/)], was purchased from Cayman Chemical (Ann Arbor, MI, USA), dissolved in dimethyl sulfoxide (DMSO) at 10 mM as a stock solution, and stored at −30°C in 100 µl aliquots.
Production and purification of recombinant E. invadens proteins
Plasmid construction, recombinant Entamoeba protein production in Escherichia coli, and subsequent protein purification were performed according to the procedure previously described for producing histidine (His)-tagged recombinant enzymes [12]. E. invadens (IP-1) cDNAs and appropriate primer sets (S1 Table) were used to construct plasmids.
Production of antisera
Four antisera used in this study, anti-EiJacob1, -EiJacob3, -EiJessie3a-N and -EiCht1-CD, were all custom made by SCRUM Inc. (Tokyo, Japan). Anti-EiJacob1 antiserum was prepared in rat, whereas the other three antisera were prepared in guinea pig. Briefly, rats and guinea pigs were immunized five times with each antigen mixed with Freund’s complete adjuvant. Antigens were prepared as described above (in Production and purification of recombinant E. invadens proteins) and whole blood was collected from each animal at one week after the last immunization.
Parasite cultures and induction of encystation
Maintenance of E. invadens (IP-1) in routine cultures, induction of encystation, and time course sampling at indicated time points were performed essentially as described previously [4]. Briefly, trophozoites suspended in encystation medium (6 × 10^5^ cells/ml) were seeded in 96-well culture plates (240 µl per well) and sealed using Plate Seal (P96P01S; Stem Corporation, Tokyo, Japan). Plates were then incubated at 26°C for the period indicated in the text and figures. For time course experiments, log-phase trophozoites were prepared from routine cultures and induced for encystation. For D-B-09-treated experiments, stationary-phase trophozoites from routine cultures were induced for encystation and incubated for 24 hr in the presence of 100 µM D-B-09.
Real time PCR
Total RNA extraction from encysting E. invadens cells (see below), cDNA synthesis, and real-time qRT-PCR with appropriate primer sets (S1 Table) were performed as previously described [21].
Western blotting
Encysting E. invadens cells, whose numbers were adjusted after induction to 1.44 × 10^5^ per well of a 96-well culture plate, were harvested from 26 wells into a single 1.5-ml tube using 1 ml phosphate-buffered saline (PBS) and pelleted by centrifugation at 770 × g for 5 min at 4°C. The cell pellet was washed once with 1 ml PBS and resuspended in 1 ml PBS. The cell suspension was then dispensed into five 1.5-ml tubes (200 µl each). Cells were repelleted by centrifugation and stored at −80°C until use. Frozen cell pellets were mixed with loading buffer containing Protease inhibitor cocktail (Roche, Basel, Switzerland) just before loading onto SDS-polyacrylamide gel electrophoresis (SDS-PAGE) gels (loaded cell number was assumed to be 4 × 10^4^ cells per lane). After SDS-PAGE, proteins were transferred onto PVDF membranes (Immobilon-P PVDF Membrane; Millipore, MA, USA) using a WSE-4025M HorizeBLOT 4M (ATTO, Tokyo, Japan) with EzFastBlot buffer (ATTO, Tokyo, Japan) in accordance with the manufacturer’s instructions.
Membranes were incubated in freshly prepared TBST [Tris-buffered saline (TBS) containing 0.05% Tween-20 (vol/vol)] containing 5% non-fat dried milk (wt/vol) for 1 hr at ambient temperature. After washing with TBST four times, membranes were incubated in a primary antibody solution for 1 hr at ambient temperature. Primary antibody solutions for EiJacob1, EiJacob3, and EiJessie3a-N antisera and EiCht1-CD antiserum were used after 10,000-fold and 3,000-fold dilution in TBST containing 0.1% BSA [wt/vol; (BSA/TBST)], respectively. Corresponding pre-immune sera were used in same dilution ratio in BSA/TBST. After washing with TBST five times, membranes were incubated in an appropriate horseradish peroxidase-conjugated goat anti-rat or anti-guinea pig IgG solution at ambient temperature for 1 hr. Secondary antibodies were purchased from Abcam (ab97057 and ab6908; Cambridge, UK) and diluted 5,000-fold with BSA/TBST. After washing with TBST five times, ImmunoStar Zeta (in Figs 2A, 2C, 2D, and S3A–S3D) or Immunostar LD (in Fig 2B) (Fujifilm Wako, Osaka, Japan) were used for detection.
Indirect immunofluorescence assay
Encysting E. invadens cells from four wells of a 96-well plate were prepared as described above and harvested into a single 1.5 ml tube using 1 ml PBS and then centrifuged at 770 × g for 5 min at 4°C. The cell pellet was washed once with 0.5 ml PBS, resuspended in 0.5 ml 2% paraformaldehyde, which was diluted with PBS from 4% paraformaldehyde (Fujifilm, Tokyo, Japan), and incubated for 3–5 min at ambient temperature. After washing with PBS three times, cells were permeabilized with 0.1 ml of 0.1% Triton X-100 in PBS (vol/vol) for 10 min at ambient temperature. After washing with PBS twice, cells were incubated with one of the primary antibody or pre-immune sera solutions [anti-rEiJacob1, -rEiJacob3, or -rEiJessie3a-N antisera or corresponding pre-immune sera diluted 300-fold with 1% BSA in PBS (wt/vol)] for 1 hr at ambient temperature. After washing with 0.1% BSA in PBS (wt/vol) three times, cells were incubated with one of the appropriate Alexa Fluor 488 conjugated goat anti-rat or anti-guinea pig IgG (ab150157 and ab150185; Abcam) [500-fold dilution with 1% BSA in PBS (wt/vol)] for 1 hr at ambient temperature. Wheat Germ Agglutinin (WGA) Alexa flour 647 (Thermo Fisher Scientific, MA, USA) a final concentration at 10 µg/ml was added to the incubation for the final 15 min. After washing with 0.1% BSA in PBS (wt/vol) three times, cells were resuspended in 30 µl PBS and observed under a fluorescence microscope at x400 magnification using x40 objective lens (Zeiss Axio Imager. M2; Carl Zeiss, Germany in Fig 3) or a confocal microscope at x1000 magnification using x100 objective lens (A1R HD25; Nikon, Japan in Figs 4A and 5 and AXR; Nikon in Fig 4B) and images were processed using Zen software (Carl Zeiss) or NIS-elements software (Nikon), respectively. Pearson correlation coefficient was determined by NIS-elements software (Nikon) (S8 Fig).
Immunoelectron microscopy on ultrathin sections
Immunoelectron microscopy analysis, based on a rapid freezing and freeze-fixation method, was outsourced to Tokai Electron Microscopy, Inc. (Nagoya, Japan). The method was performed essentially as described previously [21] with slight modification. Encysting E. invadens cells from six wells of a 96-well plate were harvested in a single 1.5 ml tube and pelleted as described above. The cell pellet was then washed twice with 0.5 ml PBS. Samples were then freeze substituted with ethanol containing 2% tannic acid in distilled water (vol/vol) at −80°C for 48 h. Subsequently, they were kept at −20°C and 4°C for 3 hr and 1 hr, respectively. After the samples were dehydrated in absolute ethanol at 4°C overnight, they were soaked once for 30 min in a mixture of ethanol and a resin (LR whilte; London ResinCo. LTD., Berkshire, UK) (50:50; vol/vol) and three times for 30 min in 100% resin at 4°C. Then, the samples were freshly soaked in 100% resin, and incubated at 50°C for 24 h to polymerize the resin. The samples embedded in the polymerized resin were sectioned at an ultra-thin thickness of 80 nm using an ultramicrotome (Ultracut UCT; Leica, Vienna, Austria). Each section on a nickel grid was incubated at 4°C overnight with one of three primary antibody solutions, rat anti-EiJacob1, and guinea pig anti-EiJessie3a-N and -EiCht1-CD, which were diluted 25, 1000, and 1000-fold, respectively, in PBS containing 1% BSA (wt/vol) and 1.5% normal goat serum (vol/vol). The sections were then incubated at ambient temperature for 2 hr with an appropriate 10 nm colloidal gold-conjugated secondary antibody (BBI Solutions, Crumlin, UK) diluted 20-fold in PBS containing 1% BSA (wt/vol) and 1.5% normal goat serum (vol/vol). For chitin staining, sections were incubated at 4°C overnight with 15 nm colloidal gold conjugated WGA (WGA-GCP 15 nm; Laboratories, Inc., San Mateo, CA, USA) diluted 50-fold in PBS. All samples were stained with 2% uranyl acetate at ambient temperature for 10 min. Then, the samples were secondarily stained with lead stain solution (Sigma-Aldrich, St. Louis, MO, USA) at ambient temperature for 3 min. The resulting samples were then examined at 100 kV acceleration voltage under a transmission electron microscope (JEM-1400 Plus; JEOL, Tokyo, Japan). Images were acquired using a CCD camera (EM-14830RUBY2; JEOL).
Supporting information
S1 TablePrimers used in this study.(PDF)
S1 FigSchematic depictions of Jacob and Jessie lectin and chitinase structures.(A) E. invadens. (B) E. histolytica.(PDF)
S2 FigDaub domain in Entamoeba Jessie3 (EiJessie3a, 161Met-613Lys; EiJessie3b, 153Met-589Lys; EhJessie3, 174Met-621Lys) is conserved in hypothetical proteins in Pelomyxa schiedti (KAH3731474.1, 61Met-525Thr), Blastocystis sp.(KAK8816815.1, 120Leu-586Val), and Amoeboaphelidium protococcarum (KAI3638213.1, 441Met-916Asp). The highly conserved domains are highlighted with colored boxes.(TIF)
S3 FigRelated to Fig 1.Transcriptional changes of the genes encoding (A) EiJacob and (B) EiJessie during encystation in a second trial.(TIF)
S4 FigAntisera reactivities.The reactivities of the four antisera against rEiJacob1 (A) and -3 (B), rEiJessie3a-N (C), and rEiCht1-CD (D) were analyzed by western blotting using the relevant recombinant proteins. Proteins in two independent gels that were arranged one above the other were transferred onto a single membrane and then the membrane was treated according to the specified steps. Arrowheads indicate the positions of all recombinant proteins loaded. Molecular mass standards are indicated on the left. Representative results from two independent experiments are shown.(TIF)
S5 FigNegative pre-immune sera controls in western blots.The reactivities of pre-immune sera of anti-EiJacob1 (A) and -3 (B), -EiJessie3a-N (C), and -EiCht1-CD (D) to each antigen and encysting E. invadens cells were analyzed by western blotting. Molecular mass standards are indicated on the left. Representative results from two independent experiments are shown.(TIF)
S6 FigLow‑magnification images showing multiple stained cells.Immunofluorescence images of encysting E. invadens cells at 24 hr after induction using rat anti-rEiJacob1, and guinea pig anti-rEiJacob3, anti-rEiJessie3a-N, and anti-rEiCht1 antisera. Scale bars indicate 50 μm.(TIF)
S7 FigNegative pre-immune sera controls in immunofluorescence assay.Immunofluorescence images of encysting E. invadens cells at 24 hr using rat anti-rEiJacob1, and guinea pig anti-rEiJacob3, anti-rEiJessie3a-N, and anti-rEiCht1 antisera and the corresponding pre-immune sera. Same concentration of secondary antibodies conjugated with Alexa Fluor 488 and same exposure time was used. Chitin is counter stained with Wheat Germ Agglutinin (WGA) Alexa flour 647. Images are obtained under x400 magnification (x40 objective lens) using Zeiss Axio Imager. M2 (Carl Zeiss). Scale bars indicate 10 µm.(TIF)
S8 FigThe aera of colocalization analysis by Pearson correlation coefficients with NIS-elements software (Nikon).All figures are same as in Fig 4.(TIF)
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Nasrallah J, Akhoundi M, Haouchine D, Marteau A, Mantelet S, Wind P, et al. Updates on the worldwide burden of amoebiasis: A case series and literature review. J Infect Public Health. 2022;15(10):1134–41. doi: 10.1016/j.jiph.2022.08.013 36155852 · doi ↗ · pubmed ↗
- 2Watanabe K, Petri WA Jr. Molecular biology research to benefit patients with Entamoeba histolytica infection. Mol Microbiol. 2015;98(2):208–17. doi: 10.1111/mmi.13131 26173474 · doi ↗ · pubmed ↗
- 3Haque R, Huston CD, Hughes M, Houpt E, Petri WA Jr. Amebiasis. N Engl J Med. 2003;348(16):1565–73. doi: 10.1056/NEJ Mra 022710 12700377 · doi ↗ · pubmed ↗
- 4Mi-Ichi F, Hamano S, Yoshida H. Links between cholesteryl sulfate-dependent and -independent processes in the morphological and physiological changes of Entamoeba encystation. Parasitol Int. 2024;99:102844. doi: 10.1016/j.parint.2023.102844 38103862 · doi ↗ · pubmed ↗
- 5Mi-Ichi F, Yoshida H, Hamano S. Entamoeba Encystation: New Targets to Prevent the Transmission of Amebiasis. P Lo S Pathog. 2016;12(10):e 1005845. doi: 10.1371/journal.ppat.1005845 27764256 PMC 5072687 · doi ↗ · pubmed ↗
- 6Wesel J, Shuman J, Bastuzel I, Dickerson J, Ingram-Smith C. Encystation of Entamoeba histolytica in Axenic Culture. Microorganisms. 2021;9(4):873. doi: 10.3390/microorganisms 9040873 33919506 PMC 8074197 · doi ↗ · pubmed ↗
- 7Krishnan D, Pandey M, Nayak S, Ghosh SK. Novel Insights into the Wattle and Daub Model of Entamoeba Cyst Wall Formation and the Importance of Actin Cytoskeleton. Pathogens. 2023;13(1):20. doi: 10.3390/pathogens 13010020 38251328 PMC 10818507 · doi ↗ · pubmed ↗
- 8Chatterjee A, Ghosh SK, Jang K, Bullitt E, Moore L, Robbins PW, et al. Evidence for a “wattle and daub” model of the cyst wall of entamoeba. P Lo S Pathog. 2009;5(7):e 1000498. doi: 10.1371/journal.ppat.1000498 19578434 PMC 2698119 · doi ↗ · pubmed ↗