E. histolytica trophozoites, strain HM1: IMSS, were axenically grown at 37°C in TYI‐S‐33 medium (Diamond et al., 1978) and harvested at the logarithmic growth phase. The culture flasks were then chilled at 4°C, and trophozoites were collected by centrifugation. All experiments reported here were performed at least three times in independent experiments, with two technical replicates per experiment.

The amino acid sequence of EhTom1 (access numbers: C4LXU1), described by Bañuelos et al. (2022) as a component of ESCRT, was used to design a specific peptide to generate mouse polyclonal antibodies against a specific EhTom1 peptide (N_244- EQIKTTLERHKKLTEK-C₂₅₉). Male BALB/c mice (from an already-existing collection in the Unidad de Producción y Experimentación de Animales de Laboratorio-Centro de Investigación y de Estudios Avanzados del Instituto Politécnico Nacional (UPEAL-CINVESTAV) were immunized with 80 µg of the peptide resuspended in TiterMax Gold adjuvant (1:1) (Sigma Aldrich). Three more immunizations were then performed at 15-day intervals, followed by bleeding to obtain the α-EhTom1 antibody. Pre-immune serum was obtained before the immunizations.

The other primary antibodies used in this study were mouse α-EhVps35, which specifically recognizes the EhVps35C isoform (onwards the EhVps35 protein) (Díaz-Valdez et al., 2024), mouse monoclonal α-Ubiquitin (α-Ub) (Santacruz), rabbit α-EhADH (Galindo et al., 2022), mouse monoclonal α-human actin (kindly given by Dr. Manuel Hernandez, CINESTAV IPN), rabbit α-EhCP112 (García-Rivera et al., 1999), rat α-EhVps23 (Galindo et al., 2021), mouse α-EhVps32 (Avalos-Padilla et al., 2015), and rabbit α-EhVps36 (Díaz-Hernández et al., 2023). As secondary antibodies, we used HRP-labeled α-mouse IgG, α-rabbit IgG, and α-rat IgG (Zymed) for the Western blot assays; and Pacific Blue-labeled or Alexa Fluor 647-labeled α-mouse IgG and Cy5-labeled α-rat IgG (Life Technologies) for the immunofluorescence assays. For the immunoelectron microscopy experiments, we used α-mouse IgG conjugated with 20 nm or 30 nm gold particles, α-rat IgG conjugated with 10 nm gold particles, and α-rabbit IgG conjugated with 30 nm gold particles (TED Pella Inc).

Immunoprecipitation was carried out with 200 µl of protein-G-agarose (rProtein-G; Invitrogen) previously incubated for 2 h at 4°C with the α-EhVps35 antibody or preimmune serum (Díaz-Valdez et al., 2024). Trophozoite lysates were prepared in the presence of 10 mM Tris-HCl, 50 mM NaCl, and a protease inhibitors cocktail during freeze-thawing in liquid nitrogen and vortexing cycles. The lysates were then pre-cleared with 200 µl of rProtein-G (previously blocked with 2% BSA) and incubated for 2 h at 4°C under gentle stirring (Avalos-Padilla et al., 2015; Galindo et al., 2021). The previously cleared cell lysates were incubated overnight (ON) at 4°C with rProtein-G bound to α-EhVps35 antibody, and then the complex: rProtein-G/α-EhVps35/EhVps35-associated proteins were recovered by centrifugation. After washing with PBS, 60 µl of 4 x sample buffer (40% glycerol, 240 mM Tris-HCl, pH 6.8, 8% SDS, 0.04% bromophenol blue, and 5% of β-mercaptoethanol). Samples were boiled for 3 min and centrifuged again at 11,600 g for 2 min at 4°C. The supernatant (30 μl) was loaded into 10% SDS–PAGE and subjected to Western blot assays using the α-EhVps35, α-EhTom1, α-EhVps23, α-EhVps36, α-EhVps32, and α-EhADH antibodies. In addition, replicas of the same experiment were analyzed in the Proteomics Units of the LaNSE (National Laboratory of Experimental Services) of CINVESTAV to massively identify the proteins that bind to EhVps35.

In total, 30 μg of the immunoprecipitated proteins with the α-EhVps35 antibody were enzymatically digested according to the protocol reported by Ramírez-Flores et al. (2019). Afterward, the peptides were loaded into a Symmetry C18 Trap V/M precolumn (Waters); 180 μm × 20 mm, 100 Å pore size, 5 μm particle size, and desalted using mobile phase A (0.1% formic acid in H2O) and mobile phase B (0.1% formic acid in acetonitrile) under the following isocratic gradient: 99.9% mobile phase A and 0.1% of mobile phase B at a flow of 5 μl/min for 3 min. Then, the peptides were loaded and separated on an HSS T3 C18 column (Waters); 75 μm × 150 mm, 100 Å pore size, 1.8 μm particle size, using an UPLC ACQUITY M-Class (Waters) with the same mobile phases under the following gradient: 0 min 7% B, 121.49 min 40% B, 123.15 to 126.46 min 85% B, and 129 to 130 min 7% B, at a flow of 400 nL/min and at 45°C. The spectra data were acquired in a mass spectrometer, Synapt G2-Si (Waters), with electrospray ionization and ion mobility separation using a data-independent acquisition approach through the HDMSE mode (Waters). The generated raw files containing MS and MS/MS spectra were deconvoluted and compared using ProteinLynx Global Server (PLGS) v3.0.3 software (Li et al., 2009) against a reversed E. histolytica database (downloaded from Uniprot). Only the proteins with ≥ 95% reliability (Protein AutoCurate green) were reported here, according to the methodology described by Ramírez-Flores et al. (2019). Additionally, the identified proteins were classified according to their function, as described in the literature, and analyzed using the PANTHER server and the GeneOntology (GO) database (http://geneontology.org/). GO analysis was conducted using the top 10 enriched GO terms in the biological processes, molecular functions, and cellular component branches. All the adjusted and statistically significant P values of the terms were log-normalized negative 10 bases.

For the phagocytosis assays, trophozoites were incubated with RBCs (1:25) from an already-existing collection for 2 min at 37°C. The cell mixture was then washed with TYI-water (2:1) at 37°C to remove the adhered and non-ingested RBCs. Subsequently, the cells were incubated at 37°C for 28 min and the samples were processed for the immunofluorescence assays (García-Rivera et al., 1982).

Trophozoites were grown on coverslips, fixed with 4% paraformaldehyde at 37°C for 1 h, permeabilized with 0.2% Triton X-100, and blocked with 10% fetal bovine serum in PBS. Preparations were incubated at 4°C ON with primary antibodies (1:50); and were incubated with the corresponding secondary antibodies for 30 min at 37°C: Pacific Blue-labeled α-mouse IgG for α-Ub, Cy5-labeled α-rat IgG for α-EhVps23 or Alexa Fluor 647-labeled α-mouse IgG for α-EhVps32 (1:100). We directly labeled the α-EhVps35 antibody with the FITC labeling fluorochrome kit (Molecular Probes-Thermo Fisher). All preparations were preserved using the Vectashield antifade reagent (Vector) and 0.5 µm laser sections were obtained and examined through the Carl Zeiss LMS 700 confocal microscope and processed with ZEN 2009 Light Edition Software (Zeiss). To evaluate the co-localization between molecules, Pearson’s coefficients were obtained from at least 30 confocal images using the ImageJ 1.45v software and the JACoP plugin.

Trophozoites (3x10⁶) in basal conditions or after being incubated with RBCs for 2 min as previously described (Galindo et al., 2021), were washed three times in PBS and incubated with 200 μl of PBS supplemented with 1 mg/ml of E64 (Sigma) and a protease inhibitor cocktail (Roche) for 2 h at 37°C. The samples were centrifuged at 13,000 x g for 10 min to obtain the secretion products (SP) in the supernatant fraction. The trophozoites in the pellet were lysed in the presence of protease inhibitors as previously reported (Bolaños et al., 2016; Galindo et al., 2022) to obtain the trophozoite extracts (TE). Samples were submitted to Western blot assays using α-EhVps35, α-EhVps23, α-EhCP112 or α-actin antibodies as described above. For further experiments, secretion products were processed to purify the extracellular vesicles (EVs), as described below.

Trophozoites lysates were obtained in the presence of protease inhibitors (PHMB 10 mM, E-64 10 μg/ml, and a protease inhibitor cocktail). Samples were electrophoresed in 10% SDS-PAGE, transferred to nitrocellulose membranes, and probed with α-EhVps35 (1:500), α-actin (1:3000), α- CP112 (1:3000), α-EhVps23 (1:500), α-EhTom1 (1:500), α-EhVps36 (1:500), α-EhADH (1:500), or α-EhVps32 (1:500) antibodies. Membranes were washed, incubated with the respective HRP-labeled secondary antibodies (Sigma, 1:1 000) according to the species, and revealed with the ECL Prime detection reagent (GE-Healthcare, Chicago, IL, USA), according to the manufacturer´s instructions.

The secretion products of trophozoites were processed to purify the extracellular vesicles (Théry et al., 2006; Galindo et al., 2022). Briefly, the secretion products were centrifuged at 10,000 x g for 30 min to remove cell debris. The final supernatant was then ultracentrifuged at 100,000 × g for 70 min to pellet the small vesicles. The pellet was washed in one volume of PBS to remove contaminating proteins and centrifuged at the same high speed. The purified samples were analyzed by transmission electron microscopy (TEM).

EVs purified from trophozoites in basal conditions were prepared for TEM as previously described (Mancilla-Olea et al., 2018). Briefly, the preparations were fixed with 4% paraformaldehyde and 0.5% glutaraldehyde in PBS for 1 h at room temperature, washed with PBS, and dehydrated with increasing concentrations of ethanol. After infiltration, samples were embedded in LR White resin (London Resin Co) and polymerized at 56°C ON to obtain thin sections (60 nm) that were mounted on Formvar‐covered nickel grids followed by ON incubation at 4°C with the α‐EhVps35, α‐EhVps23, or α‐EhADH antibodies (1:50). The samples were then incubated ON with the corresponding gold-labeled secondary antibodies (1:50) according to the specific species: α-mouse IgG was conjugated with 20 nm or 30 nm gold particles, α-rat IgG was conjugated with 10 nm gold particles, and α-rabbit IgG was conjugated with 30 nm gold particles (TED Pella Inc). The samples were then contrasted with uranyl acetate and lead citrate and observed through a Joel JEM‐1011 transmission electron microscope.

The predicted and refined 3D structures of the EhVps32 and EhVps23 (Montaño et al., 2017; Galindo et al., 2021) proteins were used for docking analysis as was the 3D structure of the EhVps35 protein (Díaz-Valdez et al., 2024). EhTom1, a 3D model, was obtained from the I-TASSER server using the VHS domain of the Tom1 protein (PDB: ELK1) from Homo sapiens as the template. The 3D structure was refined through 200 ns of MDS by NAMD2.8 (Phillips et al., 2020) with the force field CHARMM36 to create the topologies of the protein (Huang and Mackerell, 2013). The TIP3 model was applied for the water molecules. The system was solvated using the psfgen software in the VMD program (Humphrey et al., 1996). Next, 17,229 water molecules and 6 chlorine ions were added to neutralize the system which was minimized for 10,000 steps, followed by equilibration under constant temperature and pressure (NPT) conditions for 1 ns with the protein and lipid atoms restrained. Afterward, MDS was run for 200 ns, considering EhTom1 a soluble protein, without position restraints under periodic boundary conditions (PBC), and using an NPT ensemble at 310 K and 200 ns of MD simulation. The structures were visualized using the UCSF Chimera software.

The snapshots were obtained using the clustering analysis of 200 ns MDS with the Carma software (Koukos and Glykos, 2013). The protein-protein docking was done employing different conformers with the Cluspro server (Comeau et al., 2004; Kozakov et al., 2013). The conformers with the highest cluster members and the lowest energy were analyzed on the PDBSum server (Laskowski et al., 1997) and the 3D structures visualization was performed by VMD (Humphrey et al., 1996).

The Ehvps35-KD gene silencing was performed using the bacteria-expressed double-stranded RNA (dsRNA) that was soaked with the trophozoites as previously described (Solis et al., 2009). Briefly, HT115 bacteria were transformed with pL4440/Ehvps35 (Díaz-Valdez et al., 2024) and grown at 37°C in LB broth in the presence of ampicillin (100 mg/ml) and tetracycline (10 mg/ml) (Takiff et al., 1989). Ehvps35-dsRNA expression was induced by 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) ON at 37°C. dsRNA was then isolated from the bacteria using Trizol Reagent (Invitrogen), according to the manufacturer’s recommendations. DNase I (Invitrogen) and RNase A (Ambion) were added to remove ssRNA and dsDNA molecules; Ehvps35-dsRNA was washed with isopropanol and ethanol, analyzed by agarose gel electrophoresis, and its concentration was determined by spectrophotometry. Finally, trophozoites (3 x 10⁴) in TYI-S-33 medium were incubated with purified Ehvps35-dsRNA molecules to a final concentration of 5 μg/ml, and the cultures were left for 48 h at 37°C, the time at which we previously determined the silencing effect. Cells growing under standard conditions (without dsRNA) were used as controls.

Serum-starved (3 h) trophozoites (7.5 x 10⁴) were placed in the upper chamber of Transwell inserts (5 μm pore size, 24 well, Costar) and 500 μl of bovine serum was added to the lower chamber. Trophozoites were incubated for 3 h at 37°C and trophozoite migration was determined by counting the number of cells in the lower chamber of the Transwell (Bolaños et al., 2016).

Four-week-old male hamsters (Mesocricetus auratus) weighing 40 ± 5g were fasted for 24 h prior to surgery (Galindo et al., 2022). Subsequently, they were anesthetized with 3% isoflurane and anesthetized again with 1.5% of the same anesthetic during the surgical procedure. The abdominal surfaces of the hamsters were shave, and a longitudinal incision of the abdominal wall was made, to expose the port vein and livers. Subsequently, 1.5×10⁶ trophozoites in 200 μl TYI-S-33 without bovine serum were intraportally inoculated into the animals. The hamsters were sacrificed with an overdose of anesthetic 7 days after the challenge; the whole liver was weighed, and the liver lesion was dissected and weighed to calculate the percentage of damaged tissue in relation to the total liver weight (Pais-Morales et al., 2016; Galindo et al., 2022).

Values for all experiments were expressed as the mean and standard error of at least three independent assays carried out by duplicate. Statistical analyses were done with the GraphPad Prism v5.01 software by a paired Student’s t test. *p< 0.05; **p< 0.01, and ***p< 0.001.

CINVESTAV fulfills the standard of the Mexican Official Norm (NOM-062-ZOO-1999) “Technical Specifications for the Care and Use of Laboratory Animals”, based on the Guide for the Care and Use of Laboratory Animals (“The Guide,” 2011, NRC, USA with the Federal Register Number BOO.02.03.02.01.908), awarded by the National Service for Agrifood Health, Safety and Quality (SENASICA). This organization verifies the state of compliance of such NOM in Mexico and belongs to the Ministry of Agriculture and Rural Development. The Institutional Committee for Animal Care and Use (IACUC/Ethics committee) from CINVESTAV, the regulatory office for research protocols approval involving the use of laboratory animals, reviewed and approved all animal experiments (Protocol Number 0505-12, CICUAL 001).

We continued with the characterization of the C isoform of the EhVps35 protein (hereinafter EhVps35) in virulence mechanisms. We investigated by immunoprecipitation experiments and mass spectrometry analyses the proteins that interact with EhVps35, using α-EhVps35 antibodies (Díaz-Valdez et al., 2024). After immunoprecipitation, the proteins were separated by SDS-PAGE, examined by mass spectrometry, and classified according to their function, as described in the literature. The results of the analysis revealed the presence of 300 proteins, from which we selected those related to vesicular trafficking (13 proteins) ( Supplementary Table S1 ), motility (40 proteins) ( Supplementary Table S2 ), phagocytosis (90 proteins) ( Supplementary Table S3 ), and secretion (99 proteins) ( Supplementary Table S4 ) ( Figures 1A, B and Supplementary Figure S1 ). Some of these proteins participate in more than one function ( Figure 1B ), the majority of them (33%) were related to the secretion process in E. histolytica trophozoites.

Next, we explored the mechanism used by the trophozoites to secrete EhVps35, the effect of RBCs-stimulus in the protein secretion, and its relationship with ESCRT machinery proteins. Confocal images of trophozoites in basal and phagocytic conditions evidenced that EhVps35 appeared together with EhVps23 in cytoplasmic and extracellular vesicles. EhVps23 has been reported as an EVs marker in E. histolytica (Galindo et al., 2022) ( Figure 2A ). We then analyzed by SDS-PAGE the supernatants with the secreted proteins that were transferred to nitrocellulose membranes and probed with α-EhVps35, α-EhVps23, α-EhCP112, and α-actin, and with the corresponding secondary antibodies ( Figure 2B ). As we previously reported, the amount of EhVps35 protein decreased between 20% and 40% after the RBCs-stimulus to the trophozoites (Díaz-Valdez et al., 2024). Our results confirmed that this reduction remained after 2 h, the time required for the secretion assays. The decrease of EhVp35 in total extracts is possibly due to an increase in secretion ( Figure 2B ). However, we cannot rule out partial degradation of EhVps35 during the assays. In these experiments, the α-EhVps23 antibody served as an internal control (Galindo et al., 2022), the α-EhCP112 as a positive control for secretion (Bolaños et al., 2016), while the α-actin antibody, used as an internal negative control, indicates the integrity of the trophozoites during the experiments ( Figure 2B ) (Bolaños et al., 2016; Galindo et al., 2022). TEM confirmed EhVps35 and EhVp23 localization in MVBs, formed through the ESCRT machinery ( Figures 2C–G ). These structures contain intraluminal vesicles, which are released to generate exosomes.

We proceeded to purify the EVs from the supernatant as described in the materials and methods section to analyze them by TEM. As previously reported (Galindo et al., 2022), the EVs presented in sizes from 32 to 297 nm ( Figure 3 ). Thin sections of the EVs were then labeled with α-EhVps23, α-EhVps35, and α-EhADH antibodies and the corresponding secondary antibodies conjugated to gold particles, namely, EhVps35 (20 nm), EhVps23 (10 nm), and EhADH (30 nm), and the samples were then examined by TEM. The EhVps35 protein was located near vesicle membranes and inside EVs ( Figures 3A, C, E ). Recently, Galindo et al. (2022) reported that EhVps23 was also found in EVs, determining that this protein is a key factor in E. histolytica secretion, as has also been reported for human cells (Anand et al., 2019). In addition, PDCD61P, an ALIX family member, has been considered an exosome marker in humans (Anand et al., 2019). Thus, we also looked for EhADH in EVs. The images showed that the three proteins appeared in the vesicular membranes and inside them ( Figures 3A–J ). In some vesicles, EhVps35 and EhVps23 ( Figures 3A–D ) or EhVps35 and EhADH ( Figures 3E–G ) were observed together. EhADH was also localized alone in certain vesicles ( Figures 3H–J ), however, it being found alone in vesicles or their solitude may be due to the thin section observed. Interestingly, this is the first time that EhADH has been found in EVs, as has been reported for other ALIX family members. Therefore, our results highlight the importance of secretion for cellular communication and virulence processes (van Niel et al., 2022), suggesting that the EhVps35, EhVps23, and EhADH proteins in EVs might have a role in the E. histolytica attack on the host cell. Due to EhVps23 and EhVps35 co-localization in EVs and in the cytoplasm of trophozoites ( Figure 2A ), we proceeded to analyze the interaction between EhVps35 with different proteins of the ESCRT machinery.

To further investigate the interaction between the retromer and the ESCRT machinery components, we performed immunoprecipitation assays using α-EhVps35, and we searched for components of the different ESCRT machinery subcomplexes using α-EhTom1 (ESCRT-0), α-EhVps23 (ESCRT-I), α-EhVps36 (ESCRT-II), α-EhVps32 (ESCRT-III), and α-EhADH (an ESCRT accessory protein). Western blot assays of immunoprecipitates evidenced that α-EhTom1, α-EhVps23, α-EhVps32, and α-EhADH specific antibodies recognized 60, 54, 32, and 75 kDa bands, respectively, in concordance with previous reports on the migration of each protein (Arroyo and Orozco, 1987; Avalos-Padilla et al., 2015; Galindo et al., 2021; Galindo-Olea, 2022). However, the 28 kDa band that corresponds to EhVps36 did not appear in the immunoprecipitates ( Figure 4A ). This is probably because EhVps36 is an atypical protein, lacking the ubiquitin-binding domain necessary in many systems for the recruitment of other ESCRT machinery proteins (Díaz-Hernández et al., 2023). However, we cannot rule out a very fast dynamic interaction between EhVps35 and EhVps36, since the ESCRT machinery proteins participate in highly dynamic cellular processes with very fast direct or indirect interactions (Avalos-Padilla et al., 2018). Furthermore, as an interaction between EhVps36 and EhVps23 has been previously reported (Díaz-Hernández et al., 2023), an indirect interaction between EhVps35-EhVps36 could be mediated by EhVps23, a highly labile protein that is frequently degraded during experimental procedures (Galindo et al., 2022), making it difficult to study its interactions.

To strengthen the evidence of the association among the ESCRT machinery components and the retromer, we performed docking analysis using 3D models of the proteins already reported (Galindo et al., 2021; Díaz-Valdez et al., 2024) and the EhTom1 3D model was obtained using the I-TASSER server. The global free energy of the EhVps35-EhTom1 interaction was calculated as ΔG = -986.8 ( Figure 4B ). The EhVps35 predicted residues that interacted with EhTom1 were K65, Y68, Q75, D157, D209, R212, Q216, R220, Q223, N227, S266, K302, E311, E312, D316, P317, and R321. Meanwhile, in EhTom1, there were M1, L2, M3, N4, L5, L8, D175, E176, C177, E178, Q179, I181, C184, Q185, F186, L187, Y193, L197, R227, and D230. R227 and D230 were located in the EhTom1GAT domain ( Figures 4B, C ), a ubiquitin-binding domain (Bañuelos et al., 2022). The majority of the EhTom1 residues that interacted with EhVps35 were located between the VHS and GAT domains ( Figure 4C ), suggesting that the binding of EhVps35-EhTom1 does not interfere with EhTom1 binding to other proteins.

The docking analysis also predicted that EhVps35 interacts with EhVps23 with an overall free energy of ΔG = -1082.46 ( Figure 4B ). The EhVps35 predicted residues that interact with EhVps23 were R3, E13, E14, Q16, Y105, R109, E137, K147, K192, L197, R196, R202, D203, and C251; while in EhVps23 they were E49, R66, E93, Q142, Q144, N145, S146, T149, Q152, S188, S197, Q201, Y202, T336, S345, and Y490. Only three residues appeared in the EhVps23 UEV domain (E49, R66, and E94), and one residue (Y490) in the Vps23 core domain. The majority of the EhVps23 predicted residues were located between the UEV and Vps23 core domains, suggesting that EhVps23-EhVps35 binding does not interfere with EhVps23 binding to other ESCRT proteins and other molecules such as EhUbiquitin (EhUb) and LBPA (Galindo et al., 2021) ( Figure 4C ). The EhVps23 UEV domain participates in the binding to EhUb, EhVps36, EhVps32, and EhADH (Galindo et al., 2021; Díaz-Hernández et al., 2023); meanwhile, in other organisms, the Vps23 core domain is responsible for congregating other ESCRT proteins (Kostelansky et al., 2007).

The EhVps35-EhVps32 interaction showed a global free energy of ΔG = -845.5 ( Figure 4B ). The predicted residues of EhVps35 interacting with EhVps32 were S2, R3, Q5, R6, D7, S8, V9, S12, Q16, K21, R25, and H69; while in EhVps32, they were D121, N126, N127, H131, L137, E139, D140, L141, Q142, D143, D145, E148, and L149, located in the EhVps32 Snf7 domain ( Figure 4C ). The majority of the EhVps32 residues (D121, N127, H131, L137, E139, D140, L141, Q142, D145, and E148) are in the same site as those that interact with EhVps23 (Galindo et al., 2021), suggesting that the EhVps35-EhVps32 interaction prevents the EhVps23-EhVps32 union, or that a fast connection of these pair proteins occurs at a certain time. The amino acid location of the selected ESCRT proteins to predict their binding to EhVps35 are depicted in the scheme in Figures 4C, D . Some data were obtained from our docking analysis and other data from reports in the literature (Nakada-Tsukui et al., 2005; Castellanos-Castro et al., 2016; Montaño et al., 2017; Srivastava et al., 2017; Galindo et al., 2021; Díaz-Hernández et al., 2023; Díaz-Valdez et al., 2024).

These results strongly suggest an association of the retromer proteins with the ESCRT machinery as shown in the scheme depicted in Figures 4C, D . It has already been reported that several of these proteins are involved in phagocytosis (Arroyo and Orozco, 1987; García-Rivera et al., 1999; Avalos-Padilla et al., 2015, 2018; Galindo et al., 2021) and in other virulence events of E. histolytica (Galindo et al., 2022).

We have recently reported that the Ehvps35-KD gene alters the trophozoite localization of EhADH (Díaz-Valdez et al., 2024). To study the effect of Ehvps35-KD trophozoites on EhVps23 and EhVps32, we obtained Ehvps35-KD trophozoites as previously described (Díaz-Valdez et al., 2024). In concordance with previous experiments (Díaz-Valdez et al., 2024), the Western blot and confocal immunofluorescence assays evidenced a 60% reduction in the EhVps35 protein expression in the newly obtained Ehvps35-KD trophozoites ( Figures 5A–D ). To explore the impact of Ehvps35-KD on the localization and function of the EhVps23 and EhVps32 proteins, we performed pulse-chase erythrophagocytosis assays as described in the materials and methods section. After 30 min of phagocytosis, we searched, by confocal microscopy, for the EhVps35 and EhVps23 proteins, using the α-EhVps35 and α-EhVps23 antibodies. We also analyzed the EhUb localization because EhVps23 frequently binds to ubiquitinated proteins for the recruitment of other ESCRT machinery components (Galindo et al., 2021). The results obtained using non-silenced (control) trophozoites in basal conditions showed a poor signal for α-Ub, whereas α-EhVps35 and α-EhVps23 recognized small spots in the plasma membrane and cytoplasmic vesicular structures. The antibodies co-localized the EhVps35, EhVps23, and EhUb proteins at discrete points in the plasma membrane, but poor co-localization was detected in the cytoplasm. However, EhVps35 and EhVps23 were co-localized in dots and in circular structures ( Figure 5E ) while EhVps23 and EhUb appeared as small points near the plasma membrane, as described (Galindo et al., 2021). In contrast, in Ehvps35-KD trophozoites, EhVps35 exhibited a low signal, while the localization of EhVps23 increased in the plasma membrane where it mostly co-localized with EhUb ( Figure 5E ). After 30 min of phagocytosis, the cellular location of these proteins changed. EhVps35, EhVps23, and EhUb appeared in the cytoplasm in non-silenced trophozoites, co-localizing around the phagosomes. In addition, the α-Ub signal was augmented, possibly due to the ubiquitination of several proteins involved in different steps of phagocytosis. We also observed the co-localization of the three proteins in structures that could correspond to MVBs ( Figures 5E, F ). In contrast, in Ehvps35-KD trophozoites, the EhVps35, EhVps23, and EhUb proteins were poorly co-localized and only appeared in small vesicular structures ( Figures 5E–H ). It is worth mentioning that EhUb recognition by EhVps23 initiates the ESCRT machinery component recruitment for MVBs formation (Galindo et al., 2021). Our data evidenced that Ehvps35 gene silencing affects the EhUb location and fluorescence intensity ( Figures 5E, I ) and it alters both the localization and recruitment of other members of the ESCRT machinery components and the subsequent MVBs formation.

The effect of Ehvps35-KD on the location of EhVps23 and EhUb could also impact EhVps32 cellular localization, an ESCRT-III component that regulates the intraluminal vesicles (ILVs) (Avalos-Padilla et al., 2018). We proceeded to analyze, by confocal microscopy, whether the cellular location of EhVps32 was also affected in Ehvps35-KD trophozoites. Our results showed no co-localization of EhVps35 and EhVps32 in basal conditions. Both proteins appeared in vesicular structures in the cytoplasm, and, in some cells, EhVps32 presented polarization to a membrane pole. In Ehvps35-KD trophozoites, fluorescence staining was poorly observed with the α-EhVps35 antibody, and the EhVps32 protein appeared dispersed in the cytoplasm ( Figure 6A ). Interestingly, after 30 min of erythrophagocytosis in the control trophozoites, EhVps35 and EhVps32 appeared together in MVBs ( Figures 6A–D ). In contrast, in Ehvps35-KD trophozoites, EhVps35 and EhVps32 were co-localized only in cytoplasmic vacuoles, and no MVBs were visualized ( Figures 6A, D ). Previous studies in human cells have reported that Hsvps35 gene silencing causes an alteration in the ESCRT machinery functions (Filippone et al., 2021a; Walsh et al., 2021; Tan et al., 2022). Our results suggest that in E. histolytica, EhVps35 also indirectly participates in the recruitment of the ESCRT machinery, because in Ehvps35-KD trophozoites, EhUb (a key initiator molecule for this process) had an aberrant cellular location. In addition, from our results, it was logical to assume that these alterations affect virulence processes in the parasite.

EhVps23 is key for trophozoite migration (Galindo et al., 2022) and the experiments performed in this study have shown that the Ehvps35-KD gene affects EhVps23 cellular location and MVBs formation. In addition, we have reported that Ehvps35-KD causes an alteration in the cytoskeleton, a key element in movement (Díaz-Valdez et al., 2024). Furthermore, there are reports that indicate that, in cancer human cells, Hsvps35 gene silencing reduces the ability of the cells to migrate (Liu et al., 2012; Tan et al., 2022). We proceeded to investigate whether the Ehvps35-KD trophozoites presented with decreased motility, using a chemoattractant in Transwell filters (Bolaños et al., 2016). Trophozoites were placed in the upper chamber of the filters, and the lower chamber was charged with 500 μl of bovine serum. Migration was measured by the ability of the trophozoites to move to the chemoattractant. After 3 h incubation at 37°C, the Ehvps35-KD trophozoites showed an 80% decrease in their capacity to migrate toward the chemoattractant compared to the migration of the control trophozoites ( Figures 7A, B ). These results strongly suggest that EhVps35 is strongly involved in the motility signaling pathways involved in migration. This could explain the high number of motility- and cytoskeleton-involved proteins detected in the mass spectrometry analysis presented in Figure 1 .

E. histolytica movement is a key element for trophozoites to invade and reach different organs such as the liver where they cause hepatic abscesses. We investigated the ability of Ehvps35-KD trophozoites to produce hepatic abscesses in hamsters. The animals were anesthetized and intraportally inoculated with either control or Ehvps35-KD trophozoites. After 7 days of inoculation, the control animals presented with bristly hair and abdominal inflammation, and appeared depressed, while the animals inoculated with Ehvps35-KD trophozoites appeared healthy. Thus, under deep anesthesia, we examined their livers. The results showed that the control trophozoites produced a huge number of abscesses, whereas the Ehvps35-KD trophozoites caused only a few small ones ( Figure 8A ). Quantification of the healthy and damaged tissue showed that the control trophozoites damaged approximately 58.8% of the liver tissue, while the hamsters inoculated with the same number of Ehvps35-KD trophozoites presented with only approximately 23.3% of the tissue damaged ( Figure 8B ). The decrease in liver damage presented by the Ehvps35-KD trophozoites could be related to their motility impairment and to the alteration of the cellular location of EhVps23. It has been reported that when the Ehvps23 gene is overexpressed, the hepatic abscesses are exacerbated (Galindo et al., 2022). Our results suggest that EhVps35 is a key protein for the correct localization and function of the ESCRT machinery components, and they strengthened the assumption that the relationship between both complexes is crucial for virulence mechanism expression in E. histolytica. In addition, they influence other vital functions of trophozoites.