Cryptosporidium parvum Elongation Factor 1α Participates in the Formation of Base Structure at the Infection Site During Invasion

Abstract

Background

Cryptosporidium is a genus of apicomplexan parasites, the causative agents of cryptosporidiosis in humans and/or animals. Although most apicomplexans parasitize within the host cell cytosols, Cryptosporidium resides on top of host cells, but it is embraced by a double-layer parasitophorous vacuole membrane derived from host cell. There is an electron-dense band to separate the parasite from host cell cytoplasm, making it as an intracellular but extracytoplasmic parasite. However, little is known on the molecular machinery at the host cell-parasite interface.

Methods

Cryptosporidium parvum at various developmental stages were obtained by infecting HCT-8 cells cultured in vitro. Immunofluorescence assay was used to detect CpEF1α with a polyclonal antibody and host cell F-actin with rhodamine-phalloidin. Recombinant CpEF1α protein was used to evaluate its effect on the invasion by the parasite.

Results

We discovered that a C parvum translation elongation factor 1α (CpEF1α) was discharged from the invading sporozoites into host cells, forming a crescent-shaped patch that fully resembles the electron-dense band. At the same time, host cell F-actin aggregated to form a globular-shaped plug beneath the CpEF1α patch. The CpEF1α patch remained for most of the time but became weakened and dissolved upon the completion of the invasion process. In addition, recombinant CpEF1α protein could effectively interfere the invasion of sporozoites into host cells.

Conclusions

CpEF1α plays a role in the parasite invasion by participating in the formation of electron-dense band at the base of the parasite infection site.

Cryptosporidium is an intracellular but extracytoplasmic parasite separate from host cell by a unique pedestal structure. In this study, a Cryptosporidium protein (EF1α) was discovered to participate in the formation of the pedestal during parasite invasion into host cells.

Cryptosporidium parvum is a globally distributed zoonotic protozoan parasite. The parasite is transmitted via fecal-oral route, such as by drinking contaminated waters [1]. When human or animal hosts ingest oocysts, sporozoites are released from oocysts to invade intestinal epithelial cells. During the invasion, a sporozoite attaches to host cell surface, where an electron-dense band is formed at the parasite-host cell interface, and the host cell membrane forms a fold to encircle the apical end of the zoites. The membrane fold/rim gradually rises up along the zoite and fuses together to fully cover the zoite [2–5]. The process differs from other groups of apicomplexans that enter into the host cytoplasm (eg, Plasmodium, Toxoplasma, and Eimeria). At the same time, the invading sporozoite undergoes morphological change to become a round trophozoite, starting its intracellular but extracytoplasmic development. The developing parasites are embraced by a host cell-derived membrane termed parasitophorous vacuole membrane (PVM) facing intestinal lumen and an electron-dense band facing host cytoplasm, while the dense band is padded with a layer of F-actin-rich structure [6].

The morphology at the infection sites has been more extensively studied [2–5]. However, the molecular mechanisms regulating the parasite attachment and invasion are poorly understood. A number of sporozoite surface glycoproteins have been implied for their involvement in attachment and invasion, including some lectin/mucin-like glycoproteins, thrombospondin-related adhesive proteins, immunodominant antigens, and a circumsporozoite-like protein with undefined sequence (see reviews [7–9]).

Eukaryotic elongation factor 1α (EF1α), the counterpart to the prokaryotic elongation factor thermo unstable (EF-Tu), is a family of conservative proteins that catalyzes the binding of aminoacyl-tRNA to the ribosome [10, 11]. Besides the canonical role, EF-Tu and EF1α also play a role in modulating cytoskeleton by interacting with the actin-like MreB protein in prokaryotes and F-actin in eukaryotes [12, 13]. We previously observed that the CpEF1α gene (cgd6_3990) was highly expressed in the oocysts of C parvum (ie, top 5.5% among the 1924 expressed protein-encoding genes by microarray analysis) [14], followed by the confirmation of its high level of expression in sporozoites by quantitative reverse-transcription polymerase chain reaction (qRT-PCR) and by data mining the proteomics datasets (Figure 1). The high abundance of C parvum translation EF1α (CpEF1α) in sporozoites, together with the ability of EF1α to modulate cytoskeleton, prompted us to investigate its potential noncanonical roles in C parvum (eg, as a structure protein and involvement in parasite invasion and infection).

Figure 1.

Expression levels of CpEF1α gene (gene identification [ID] cgd6_3990) in various developmental stages of Cryptosporidium parvum. (A) Ranks of CpEF1α transcript level in oocysts in a transcriptomics dataset, and C parvum translation elongation factor 1α (CpEF1α) protein level in sporozoites in a proteomic dataset, in comparison with those of an elongation releasing factor 3 (CpERF3; cgd2_2070). The oocyst transcriptomics and sporozoite proteomics datasets were derived from previous publications and available at the CryptoDB. (B) Levels of CpEF1α transcripts in oocysts, sporozoites, and intracellular stages at various postinfection time points in comparison with those of CpEF1β1, CpEF1γ, CpEFR3, and CpLDH genes. Gene transcription levels were determined by quantitative reverse-transcription polymerase chain reaction, converted to the fluorescent units at cycle zero (F₀) by a 4-parametric sigmoidal curve fitting, normalized with F₀ values of Cp18S, and expressed relative to the medium value of all genes. hpi, hours postinfection; mRNA, messenger ribonucleic acid.

In the present study, we investigated the dynamic distributions of CpEF1α during parasite invasion and intracellular development, and we observed that CpEF1α at the sporozoite apex was discharged into host cells and participated in the formation of the base structure at the infection site during the parasite attachment and invasion. The involvement of CpEF1α in the parasite invasion was further supported by the inhibition of CpEF1α protein on the invasion of sporozoites into host cells.

MATERIALS AND METHODS

Materials and methods are concisely described here. A more detailed description of materials and methods is provided in Supplementary Material.

Parasite Material and In Vitro Cultivation

The oocysts of C parvum was purchased from Bunch Grass Farm (Deary, ID) with a subtype IIaA17G2R1 at gp60 locus [15–17]. Oocysts used in experiments were <3 months old with an in vitro excystation rate of ~70% or higher. Before use, oocysts were purified as described [14, 18]. Free sporozoites were prepared by incubation in phosphate-buffered saline (PBS) containing .25% trypsin and .5% taurodeoxycholic acid at 37°C for 1 hour followed by washes with PBS.

In vitro cultivation of C parvum was hosted in HCT-8 cells as previously described [18, 19]. In brief, HCT-8 cells were seeded in plates in Roswell Park Memorial Institute 1640 medium containing 20% fetal bovine serum (FBS) in an incubator with 5% CO₂ at 37°C until they reached ~80% confluence or as specified. Oocysts or freshly prepared sporozoites were used to infect host cells for various times as specified.

Cloning of CpEF1α Gene and Expression of Recombinant Protein

The C parvum genome encodes a single-copy CpEF1α gene (cgd6_3990). Its open-reading frame was amplified by PCR from C parvum genomic deoxyribonucleic acid (DNA) using a high-fidelity Pfu DNA polymerase (Agilent Technologies), and cloned into a pCR4Blunt-TOPO vector (Thermo Fisher Scientific), followed by sequencing of plasmids to confirm the identity of inserts. The CpEF1α insert was then subcloned into a pMAL-c2E-TEV-His vector for expressing as a maltose-binding protein (MBP) fusion protein [20]. The expression of recombinant MBP-CpEF1α protein was carried out in a Rosetta strain of Escherichia coli (Novagen). The induction of expression and purification of recombinant protein using amylose resin-based affinity chromatography followed manufacturers’ standard protocols as described [21, 22].

CpEF1α Gene Expression in Various Developmental Stages of Cryptosporidium parvum

The relative levels of transcripts of CpEF1α and several elongation factor-related genes (ie, CpEF1β, CpEF1γ and CpERF3) were evaluated by qRT-PCR using a one-step QuantiTect SYBR Green RT-PCR kit. The C parvum lactate dehydrogenase (CpLDH) gene was also included as a reference [22]. The parasite 18S ribosomal ribonucleic acid ([rRNA] Cp18S) levels were used for normalization. Each reaction (10 μL) contained .6 ng of total RNA from oocysts or sporozoites, or 10 ng of total RNA from intracellular stages, and 500 nM of each of the primers for specified genes (Supplementary Table S1).

To compare the transcript levels between different genes, we calculated the fluorescent units at cycle zero (F₀) for each gene transcript by a 4-parametric sigmoidal curve fitting of the plots between the fluorescence readouts and thermal cycles [23–25]. Because F₀ values were direct reflection of the initial target quantities expressed in fluorescence units, they could be directly used to compare the levels of different gene transcripts without prior knowledge of PCR amplification efficiencies [25]. In this study, the F₀ values of gene transcripts were first normalized with those of Cp18S in individual samples. The relative levels of gene transcripts were then compared with the median value derived from all genes in all samples.

Anti-Cryptosporidium parvum Elongation Factor 1α Antibody Preparation and Western Blot Analysis

Polyclonal anti-CpEF1α antibodies were prepared against a synthetic peptide specific to CpEF1α (¹⁵⁶CEYKQSRFDEIFNEVDGYLKK¹⁷⁶) in 2 specific pathogen-free rabbits at a commercial service (Alpha Diagnostics International). Antibodies were affinity-purified using agarose-resin conjugated with the synthetic peptide. The specificity of the polyclonal antibody was evaluated by Western blot analysis of C parvum sporozoite total proteins. Preimmune rabbit serum (1:500 dilution) and the anti-CpEF1α antibody neutralized by presoaking it with recombinant CpEF1α protein (~0.5 μg/mL) were also used as controls.

Distributions of Cryptosporidium parvum Elongation Factor 1α in Sporozoites and in the Parasite During Invasion and Intracellular Development and Colocalization With Host Cell F-actin

Various life cycle stages of C parvum were prepared. Free sporozoites were prepared by excystation, fixed with 4% paraformaldehyde, washed with PBS, and loaded onto glass coverslips coated with poly-l-lysine. For studying sporozoite invasion, fresh sporozoites (1 × 10⁶/well) were incubated with HCT-8 cells grown on coverslips for 25 to 45 minutes at 37°C. For preparing intracellular parasites after invasion, oocysts were incubated with HCT-8 cells for 3 hours at 37°C, followed by an exchange of culture medium to remove free parasites and continuous cultivation at 37°C for a total of 6, 10, and 18 hours. Cell monolayers were fixed with 4% paraformaldehyde in PBS and washed with PBS.

After permeabilization with .1% Triton X-100 and blocking with 5% bovine serum albumin (BSA), specimens were stained for host cell F-actin using rhodamine-phalloidin (1:500 dilution; Thermo Fisher Scientific) and stained for CpEF1α using affinity-purified antibody, followed by incubation with a goat antirabbit immunoglobulin G antibody conjugated with fluorescein-5-isothiocyanate (FITC). Samples were mounted onto glass slides with an antifade reagent containing 4′,6-diamidino-2-phenylindole (DAPI) to counterstain nuclei. Slides were examined with an Olympus BX51 research microscope equipped with a 100×/1.3 oil objective lens and filter sets for FITC and rhodamine. All images were digitally captured and uniformly manipulated with Adobe Photoshop CC for signal contrast and intensity.

Evaluation of the Effect of Cryptosporidium parvum Elongation Factor 1α and Actin on the Parasite Invasion

In this assay, recombinant CpEF1α ([rCpEF1α] 50 μg/mL or 12.5 to 100 μg/mL; MBP-tag uncleaved), rabbit skeletal muscle G-actin (50 μg/mL), F-actin (50 μg/mL), or MBP-tag were first mixed with oocysts in culture medium containing BSA (50 μg/mL) for 10 minutes. Oocysts suspensions were then added to host cells cultured in 48-well plates at ~90% confluence (oocysts/host cells ratio at 1:2; 2.5 × 10⁴ oocysts/well). After incubation at 37°C for 3 hours to allow parasite excystation and invasion, uninfected parasites were removed by a medium exchange.

After cultivation for additional 15 hours in the absence proteins (total 18 hours infection time), cells were lysed in ice-cold iScript qRT-PCR sample preparation reagent (200 µL/well) (Bio-Rad Laboratories), and supernatants were used as RNA templates after dilution for qRT-PCR or stored at −80°C until use [26]. The parasite loads were evaluated by detecting the levels of Cp18S rRNA transcripts [26]. Melting curves were analyzed between 65°C and 95°C at the end of PCR. Amplification curves and melting peaks were examined to assess the quality and specificity of the reactions.

The means of cycle threshold (C_T) values from technical replicates of each biological replicate were first used to calculate ΔC_T values between Cp18S and Hs18S (ΔC_T = C_T[Cp18S]-C_T[Hs18S]) and then ΔΔC_T between samples and controls (ΔΔC_T = ΔC_T[sample]-ΔC_T[control]). The relative level of gene expression was then calculated using an empirical formula 2^-ΔΔCT and converted to percentage values [26]. Statistical significance on the relative levels of parasite 18S rRNA was evaluated by one-way analysis of variance followed by Holm-Sidak’s multiple comparisons test.

To further validate the effect of rCpEF1α on parasite invasion, we microscopically counted the number of invaded sporozoites or trophozoites after 3-hour infection. HCT-8 cells were cultured on round coverslips in 24-well plates to ~90% confluence. Oocysts were preincubated with rCpEF1α or MBP (each 50 μg/mL) in culture medium at 37°C for 10 minutes and then added into individual wells (1 × 10⁵ oocysts/well). After incubation at 37°C for 3 hours, cell monolayers were gently washed 3 times with PBS and fixed with 4% paraformaldehyde followed by washes with PBS. Slides were mounted with an antifade mounting medium with DAPI. The number of intracellular sporozoites or trophozoites were counted with Olympus BX51 research microscope. The experiments were repeated 3 times, each containing 3 biological replicates. For each biological replicate, 15 randomly selected microscopic fields were counted. The data were subjected to statistical analysis by Student’s t test for significance between experiment and control groups [18].

RESULTS

CpEF1α Was a Highly Expressed Gene in Oocysts and Sporozoites and Up-Regulated During the Parasite Infection

In an earlier microarray-based transcriptomics analysis [14], CpEF1α was found to be one of the highly expressed genes in oocysts, ie, top no. 106 (or top 5.5%) among 1924 protein-coding genes (Figure 1A). This observation was supported by the published proteomics data available at the CryptoDB (http://cryptodb.org/cryptodb/) [27, 28], in which CpEF1α was ranked no. 23 (top 2.1%) of the most abundant proteins in sporozoites (Figure 1A). For comparison, the level of a related cgd2_2070 gene that encoded an EF1-related eukaryotic peptide chain release factor eRF3 (CpERF3) was only ranked top no. 836 (top 43.6%) in oocyst transcriptomics and no. 391 (top 34.9%) in sporozoite proteomics (Figure 1A). It is noticed that cgd2_2070 was mistakenly annotated as EF1α at CryptoDB (release 45), rather than eRF3, likely due to the presence of a highly conserved EF1α domain (ie, cd01883 at the NCBI’s Conserved Domain Database) (Supplementary Figure S1). It is noticeable that the C hominis EF1α gene (Chro.60459) is also highly expressed in oocysts, ranked at 87.4 percentile among 3886 genes in an RNA-seq-based transcriptomics dataset available at CryptoDB.

The high-level expression of CpEF1α in oocysts and sporozoites were supported by qRT-PCR analysis, in which CpEF1α transcript levels were 5- to 13-fold or 15- to 83-fold higher than those of CpEF1β, CpEF1γ, and CpERF3 in oocysts or sporozoites (Figure 1B). The CpEF1α transcript levels were also much higher than those of CpEF1β, CpEF1γ, and CpeRF3 in intracellular developmental stages (ie, 51- to 546-fold higher) (Figure 1B). Although all EF1-related genes generally increased their expression levels from oocysts to sporozoites or to intracellular stages, the expression of CpEF1α was higher than that in other genes (eg, 60- or 133-fold increases from oocysts to sporozoites or to intracellular stage at 6 hours postinfection time vs 16- to 47-fold for CpEF1β, 19- to 7-fold for CpEF1γ, and 3.5- to 1.5-fold for CpERF3, respectively) (Figure 1B). As an additional comparison, the expression profile of the CpLDH gene differed from CpEF1α and other EF1-related genes by its highest expression in oocysts and sporozoites, which agreed with previously reported data [14, 22], and confirmed the reliability of the data in this study. Our observations were also supported by the transcriptomics data from Mauzy et al [28], in which the expressions of CpEF1α and other EF1-related genes followed similar patterns (data not shown but are available at CryptoDB).

Cryptosporidium parvum Elongation Factor 1α Was Mainly A Cytosolic Protein in Sporozoites and Meronts

A rabbit polyclonal antibody against a CpEF1α-specific peptide was produced to investigate the distribution of CpEF1α protein in the parasite. Western blot analysis using this antibody detected single bands at ~47 kDa from sporozoites and ~90 kDa from rCpEF1α with MBP-tag, which agreed with the theoretical masses of native (~47 kDa) and rCpEF1α (~90 kDa) (Figure 2). The broad band from rCpEF1α contained signals derived from incompletely translated recombinant proteins, because presoaking of the antibody with rCpEF1α could completely eliminate the signals in both native and rCpEF1α (Figure 2). These observations confirmed the specificity of the antibody.

Figure 2.

Western blot analysis of native Cryptosporidium parvum translation elongation factor 1α (CpEF1α) and recombinant CpEF1α (rCpEF1α) proteins using a polyclonal antibody produced in rabbits against a synthetic peptide. The detection of native CpEF1α used C parvum sporozoites prepared by in vitro excystation (lanes labeled with Spz). Controls included the detection of CpEF1α in sporozoites with preimmune serum (Pre-Imm) and that of native CpEF1α and rCpEF1α using antibodies presoaked with rCpEF1α protein (presoaked anti-CpEF1α) to validate the antibody specificity. SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis.

Immunofluorescence microscopy revealed that CpEF1α was mostly distributed in the cytosol of free sporozoites, trophozoites, and meronts (Figure 3). There was evidence that some CpEF1α were associated with the sporozoite plasma membrane (Figure 3A‒C). No signals were observed in the parasites and host cells incubated with preimmune serum (Supplementary Figure S2).

Figure 3.

Indirect immunofluorescence microscopic detection of Cryptosporidium parvum translation elongation factor 1α (CpEF1α) in various developmental stages of C parvum. (A–C) The CpEF1α distribution in sporozoites under normal (A) or ruptured morphology (B and C). (D–F) The CpEF1α distribution in intracellular meronts with 1, 2, and 4 nuclei. Actin, host cell F-actin-labeled with rhodamine-phalloidin; CpEF1α, parasite CpEF1α protein labeled with a rabbit anti-CpEF1α antibody and a fluorescein-5-isothiocyanate-conjugated goat antirabbit immunoglobulin G antibody; CpEF1α/Actin/Nucl, superimposed images; DIC, differential interference contrast microscopy; Nucl, nuclei stained by 4’, 6-diamidino-2-phenylindole. Bar = 2 μm.

By costaining with rhodamine-conjugated phalloidin, we observed the accumulation of host cell F-actin filaments at the infection site (Figure 4D‒F), which agreed with the observations by other investigators [6]. Phalloidin is known to bind to F-actin from animals, plants, and many other organisms, but it is unable to visualize F-actin in Plasmodium and Toxoplasma [30–32]. The present study also confirmed that phalloidin-fluorophore was unable to visualize F-actin in Cryptosporidium (Supplementary Figure S3).

Figure 4.

Indirect immunofluorescence microscopic detection of Cryptosporidium parvum translation elongation factor 1α (CpEF1α) during the attachment and invasion of C parvum sporozoites. Specimens were labeled with fluorescein-5-isothiocyanate for CpEF1α, rhodamine-phalloidin for host cell F-actin (Actin), and 4′,6-diamidino-2-phenylindole for nuclei (Nucl). Bright-field images were captured by differential interference contrast (DIC) microscopy. Fluorescence images were superimposed from images labeled for CpEF1α, F-actin, and nuclei. A more comprehensive set of images, including individual images without superimposition, were presented in Supplementary Figure S4. Also see Figure 3 for abbreviations.

Cryptosporidium parvum Elongation Factor 1α Participated in the Formation of the Base Structure at the Infection Site During the Invasion by C parvum Sporozoites

The invasion of C parvum sporozoites takes place in minutes, along with the engulfment of sporozoites by host cell plasma membrane-derived PVM [3–5]. The invasion is also accompanied by the shortening of sporozoites from banana/rod-like shape to round-shaped trophozoites fully covered by PVM. We observed that a significant amount of CpEF1α protein at the sporozoite apex was discharged into host cells upon the attachment of sporozoites (Figure 4A, Supplementary Figure S4A). The aggregated CpEF1α then formed a crescent-shaped patch corresponding to the base structure (pedestal) formed at the invasion site (Figure 4A‒I, Supplementary Figure S4). The stain of the CpEF1α patch maintained the crescent shape for most of the time during the parasite invasion, but it weakened and dissolved after sporozoites were fully transformed into round trophozoites (Figure 4M‒R, Supplementary Figure S4). Upon discharge of CpEF1α, an anterior region in the sporozoite with no or weak fluorescence signals appeared (Figure 4C‒J), corresponding to the “anterior vacuole” observed by earlier ultrastructural studies [4, 5].

Upon the attachment of sporozoites, there was also an aggregation of host cell F-actin into a globular structure (ie, Ø = ~0.5 μm) that partially overlapped with the CpEF1α patch (Figure 4, Supplementary Figure S4). During the course of invasion and morphogenesis of sporozoites, host cell F-actin transformed from the globular structure to form a layer beneath the CpEF1α patch, which was in agreement with previously reported observations [33, 34]. The phalloidin-stained actin filaments was not extended to the PVM, although strong signals surrounding meronts in later developmental stages were observed (eg, Figure 4K and P). However, we were unable to conclude the absence or presence of host cell F-actin in PVM because of the limited sensitivity of immunofluorescent assay in detecting actin in the thin layer of PVM as suggested by earlier studies [33–35]. The strong signals surrounding meronts were mainly derived from the actin-rich pedestal and host cell membrane folds rather than from the PVM [35].

Recombinant Cryptosporidium parvum Elongation Factor 1α Protein Interfered With the Invasion by C parvum Sporozoites

We further confirmed that rCpEF1α protein could interfere with the invasion by sporozoites. First, we used an 18-hour infection/qRT-PCR assay, in which the parasites were incubated with rCpEF1α, G-actin, or F-actin for 3 hours, followed by the removal of proteins and uninvaded parasites and continuous culture of the invaded parasites for 15 hours to increase the detection power by qRT-PCR. Therefore, the parasite loads detected by this assay were correlated to the numbers of invaded sporozoites. In this assay, we detected 49.4% reduction of invasion by rCpEF1α alone, or 41.0% to 52.8% reduction by rCpEF1α together with F- or G-protein, but not by G-actin or F-actin (each at 50 μg/mL) (Figure 5A). The inability for G- or F-actin to “neutralize” the inhibitory effect of rCpEF1α on parasite invasion suggested the lack of direct interaction between rCpEF1α and F-actin under this experimental condition that lacked sufficient Ca²⁺ and adenosine triphosphate (or guanosine-5'-triphosphate) needed for the interaction and for F-actin bundling. The inhibition by rCpEF1α on the parasite invasion was dose dependent, ranging from 21.2% at 12.5 μg/mL to 68.4% at 100 μg/mL concentrations (Figure 5B). Second, we directly counted the invaded parasites after 3 hours of invasion of sporozoites by an immunofluorescence assay, and we observed 59.5% reduction of invasion, which further confirmed the inhibitory effect of rCpEF1α on the parasite invasion (Figure 5C).

Figure 5.

Effects of treatment of recombinant Cryptosporidium parvum translation elongation factor 1α (rCpEF1α) and rabbit muscle actin on the invasion of C parvum sporozoites into HCT-8 cells. (A) Effect of rCpEF1α (as maltose-binding protein [MBP]-fusion protein), G-actin, and F-actin (50 μg/mL each) on sporozoite invasion as determined by quantitative reverse-transcription polymerase chain reaction (qRT-PCR) in 3-hour invasion + 15-hour infection assay. (B) Dose-dependent inhibition of CpEF1α on sporozoite invasion as determined by qRT-PCR in 3-hour invasion + 15-hour infection assay. (C) Effect of rCpEF1α (50 μg/mL) on sporozoite invasion as determined by microscopic counting of intracellular sporozoites/trophozoites after 3-hour invasion. In all experiments, proteins were individually preincubated with oocysts and host cells for 10 minutes and then mixed together to allow 3-hour invasion. All proteins were removed by a medium exchange, followed by additional 15 hours growth for qRT-PCR assay (A and B) or fixation for microscopic counting (C). Asterisks indicate level of statistical significance between experimental groups and control by Holm-Sidak’s multiple comparisons test (A and B) or by Student’s t test (C). *, P < .05; **, P < .01; ***, P < .001; ****, P < .0001.

DISCUSSIONS

A Working Model on the Participation of Cryptosporidium parvum Elongation Factor 1α in the Formation of the Electron-Dense Band at the Parasite Infection Site

The epi-cytoplasmic development of C parvum is supported by a unique base structure (pedestal) at the parasite-host cell interface, which is characterized by an electron-dense band padded with an F-actin-rich layer in the host cytoplasm [6]. An earlier ultrastructural study on the invasion of C parvum sporozoites has clearly showed that the electron-dense band is initially formed as a “stand-alone” structure in the host cell cytosol below the cell surface, but its edge was eventually fused with host cell plasma membrane before the completion of the invasion process [4]. The molecular events associated with the formation of the unique pedestal structure are poorly understood.

The present study provides direct evidence showing that CpEF1α protein at the apical region of C parvum sporozoites was discharged into host cell to participate in the formation of the pedestal structure. The discharged CpEF1α forms a crescent-shaped morphology that fully resembles the electron-dense band [3–5]. These observations lead us to build a working model and a hypothesis that CpEF1α protein participates in the formation of the electron-dense band at the parasite-host cell interface (Figure 6).

Figure 6.

A working model on the role of Cryptosporidium parvum translation elongation factor 1α (rCpEF1α) in the formation of the electron-dense band during the invasion of C parvum sporozoites into host cells. (A) A diagram illustrating the working model. (B) Comparison between the model and a fluorescence image of an invading sporozoite. DB, electron-dense band; N, sporozoite nucleus; PVM, parasitophorous vacuole membrane.

Speculation on the Potential Role of Cryptosporidium parvum Translation Elongation Factor 1α in the Reorganization of Actin Filaments

The electron-dense band at the C parvum infection site is padded with an F-actin-rich layer on the side of host cell, but the molecular mechanism triggering the formation of host cell actin filament is barely understood [8]. There were observations suggesting the involvement of host cell integrin α2/Src/cortactin pathway and PI3K/Cdc42/WASP pathway in the rearrangement of host cell actin filaments [8, 33, 36–38]. The present study has showed that host cell F-actin aggregates to form a “plug” underneath the CpEF1α patch immediately after the attachment of C parvum sporozoites (Figure 4, Supplementary Figure S4). Considering that (1) EF1α could also function as an actin-bundling protein and regulate the reorganization of F-actin during cell protrusion [13, 39], (2) CpEF1α is discharged into host cell upon the attachment of C parvum sporozoites (Figure 4, Supplementary Figure S4), and (3) the discharged CpEF1α has an opportunity to interact with host cell F-actin, it is plausible to speculate that the discharged CpEF1α might also play a role in initializing the bundling and aggregation of host cell F-actin.

The protrusion and extension of cytoplasmic membrane is typically driven by the rearrangements of actin filaments [40–42]. Therefore, host cell F-actin has been thought to be involved in the formation and growth of PVM during the parasite invasion and the maintenance of PVM during the parasite intracellular development [8, 33–35]. However, the presence or absence of host actin filaments within the double-layered PVM were not confirmed in earlier studies, likely due to the limitation of sensitivity in detecting F-actin in the thin PVM by either phalloidin or antiactin antibodies [30, 31]. Likewise, we were unable to conclude the presence or absence of CpEF1α and/or C parvum F-actin in the thin PVM due to the limitation of detection sensitivity and the inability of phalloidin-fluorophore to visualize parasite F-actin.

CONCLUSIONS

In summary, we found that CpEF1α was involved in the initial steps of attachment and invasion of Cryptosporidium sporozoites into host cells, a new noncanonical function for EF1α, via participating in the formation of base structure at the host cell-parasite interface. This discovery opens the door to further delineate and validate the role of CpEF1α in Cryptosporidium invasion including using recently developed genetic tools. In addition, it is also worthwhile to develop an assay for high-throughput screening of selective anti-CpEF1α inhibitors for potential anti-cryptosporidial drug development.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. We thank Drs. Robert Burghardt and Gonzalo Rivera at Texas A&M University for technical and scientific advice.

Financial support. This work was funded in part by a grant from the National Institute of Allergy and Infectious Diseases at the National Institutes of Health (award number R21 AI136663). This study was also funded in part by a Graduate Student Core Facility Grant at the College of Veterinary Medicine & Biomedical Sciences, Texas A&M University.

Potential conflict of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.