Developing efficient strategies for localizing the enhanced yellow fluorescent protein subcellularly in transgenic Eimeria parasites

Eimeria species serve as promising eukaryotic vaccine vectors. And that the location of heterologous antigens in the subcellular components of genetically modified Eimeria may determine the magnitude and type of immune responses. Therefore, our study aimed to target a heterologous fluorescent protein to the cell surface or microneme, two locations where are more effective in inducing protective immunity, of Eimeria tenella and E. acervulina sporozoites. We used an enhanced yellow fluorescent protein (EYFP) as a tagging biomarker, fusing variously with some localization or whole sequences of compartmental proteins for targeting. After acquiring stable transgenic Eimeria populations, we observed EYFP expressing in expected locations with certain strategies. That is, EYFP successfully localized to the surface when it was fused between signal peptides and mature products of surface antigen 1 (SAG1). Furthermore, EYFP was efficiently targeted to the apical end, an optimal location for secretory organelle known as the microneme, when fused to the C terminus of microneme protein 2. Unexpectedly, EYFP exhibited dominantly in the apical end with only weak expression on the surface of the transgenic sporozoites when the parasites were transfected with plasmid with EYFP fused between signal peptides and mature products of E. tenella SAG 13. These strategies worked in both E. tenella and E. acervulina, laying a solid foundation for studying E. tenella and E. acervulina-based live vaccines that can be further tailored to the inclusion of cargo immunogens from other pathogens.

Like other apicomplexan parasites of medical and veterinary importance, Eimeria species use the surface antigen (SAG) with a C-terminal glycosylphosphatidylinositol (GPI)-anchored sequence to initially attach to its host cells 15,16 .Subsequently, the parasite uses its adhesive microneme protein (MIC), which is secreted by the anterior organelle microneme, to tighten attachment to host cells 17,18 .The invading parasite then injects its secretory rhoptry neck protein into the host cytoplasm to build moving junctions, and with the help of rhoptry protein (ROP), the invasion stage of each infection is completed 19,20 .
Early studies have shown that P30 is a major surface SAG protein of Toxoplasma gondii, and deletion of its GPI-anchored sequence results in a changed localization of fused green fluorescent protein (GFP) to the parasitophorous vacuole but not the membrane in stable transformants 21 .The E. tenella MIC-2 (EtMIC2) localizes to the cytoplasmic vesicles via its signal peptide (SP2); fusion of the GPI-anchor sequence of E. tenella SAG-1 (EtSAG1) to the SP2 C-terminus results in an altered localization to cell surface 13 .These observations have established an important role that GPI-anchor sequence plays in the cell membrane domain.Studies have also shown that the T. gondii ROP-1 (TgROP1) and the T. gondii MIC-3 (TgMIC3) could target fused GFP to the rhoptry or microneme, respectively, in T. gondii 22 .These targeting strategies could be considered to apply to develop effective vaccines with surface and/or microneme-localized antigens when genetically engineered Eimeria parasites are used as vectors.
Eimeria tenella is a model species of Eimeria.And E. acervulina causes chicken coccidiosis with a moderate pathogenesis; but interestingly, the infection results in a high level of host immunity.Thus, E. acervulina is also considered an adequate candidate for developing transgenic Eimeria-based vaccines 23 .In this study, we took various approaches for expressing heterologous antigens in two transgenic Eimeria organisms with surfacepresented or microneme-localized antigens.

Results
The signal peptide and mature peptide from SAGs targeted the EYFP to E. tenella cell surface EYFP was inserted between the signal peptide of EtSAG13 and the GPI-anchored sequence of EtSAG13 or the whole sequence minus the signal peptide, resulting in two constructs termed ss13-EYFP-GPI13 and ss13-EYFP-EtSAG13, respectively (Fig. 1A,B).In stably transfected parasites, i.e., sporulated oocysts, and released sporozoites, confocal fluorescent microscopy revealed that EYFP was expressed in RBs of transgenic sporozoites which were transfected with ss13-EYFP-GPI13 (Fig. 1A), but on transgenic sporozoites transfected with ss13-EYFP-EtSAG13, EYFP was unexpectedly expressed dominantly in apical end with only week expression on the surface of the parasite (Fig. 1B).

Genetic manipulation in E. acervulina
The success of stable transfection opens up genetic manipulation of E. acervulina 23 .For enhanced manipulation, several promoters and signal peptides were screened for using in E. acervulina.We tried to clone promoters and signal sequences of 9 candidate genes from E. acervulina, including microneme adhesive repeat domain containing protein (EaMCP1), EaMIC1, EaMIC2 and 6 rhoptry proteins (EaROP0, EaROP2, EaROP17, EaROP23, EaROP30, EaROP35) (Table 1).Then we constructed EYFP-expressing plasmids containing each of these elements and conducted transfection in E. acervulina sporozoites (Fig. 3A, M/M/R represents one of the above 9 genes).Finally, we obtained EYFP positive progeny oocysts from transfection with a plasmid containing EaMCP1 or EaMIC2 promoters and signal sequences, respectively.The results suggested that the EYFP expression profile was similar in these 2 transfectants, which seemed in the cavity within the sporocyst in the sporulated oocysts (Fig. 3B,C).The location was confirmed by fluorescence microscopy of sporocyst (Fig. 3D), and immunoelectron microscopy (Fig. 3E-H).

Targeting proteins to cell surface or microneme in E. acervulina
In follow-up studies that aimed to validate our strategies for a targeted delivery of cargo proteins, we fused EYFP between the signal peptide and mature peptide of EaSAG for cell surface expression in E. acervulina (Fig. 4A).After transfection, we obtained EYFP positive parasites, and the proportion of EYFP positive oocysts in the 2nd generation of progeny was up to 70-80% through double selection of pyrimethamine press and fluorescenceactivated cell sorting (FACS).We observed the surface expression of EYFP within sporulated oocysts (Fig. 4B) or released sporocysts (Fig. 4C), or even the sporozoites mechanically released from sporocysts (Fig. 4D).Further, we also observed the surface expression of EYFP during the endogenous developmental stages, including noninvaded sporozoites (Fig. 4E), invaded sporozoites (Fig. 4F), trophozoites (Fig. 4G) and schizonts (Fig. 4H) on  38 , a self-cleavage peptide of porcine teschovirus-1, was used to separate expression of flanking genes; DHFR, dihydrofolate reductase-thymidylate synthase mutants of Toxoplasma gondii, resistant to pyrimethamine, for drug selection of transgenic populations; 3′EtActin, the downstream untranslated sequence from E. tenella actin protein (EtActin); 5′EtSAG1, the upstream untranslated sequence from EtSAG1, as promoter; L, linker peptide, the sequence is GS(GGGS) 2 GS; 3′EtSAG1, the downstream untranslated sequence from EtSAG1; BF bright field.
Vol:.( 1234567890 Vol.:(0123456789) www.nature.com/scientificreports/slides with duodenum smears.Thus, EYFP mainly concentrated on the parasite surface both in vitro and in vivo, confirming that the in-frame fusion of EaSAG1 between signal peptide and mature peptide could make surface display for heterologous proteins.For microneme location in E. acervulina, we transfected the plasmid EtMIC2-EaMIC2-EYFP (Fig. 2D) in which EYFP was fused to the C-terminal of EaMIC2 under the control of 5'EtMIC2 to E. acervulina (Fig. 5A).We obtained EYFP positive parasites, and the proportion of EYFP positive oocysts in the 3rd generation of progeny was 60% with the selection of pyrimethamine and FACS for progeny oocysts (Fig. 5B).We found that the fluorescent signal mainly accumulated in the apical end of EYFP positive sporozoites but fewly co-localized with a monoclonal antibody against the EtMIC2 (Fig. 5C).

Discussion
Our study here clearly demonstrates that two Eimeria species are amenable to genetic modifications that target various cargo proteins (immunogens) to specific location (Table 2).Using confocal fluorescence microscopy, EYFP was observed on cell surface when EYFP is fused between signal peptide and mature peptide of EtSAGs, and in apical end, a potential microneme location when it was fused with MIC2 for both E. tenella or E. acervulina.
Various designs that ensure the expression of heterologous proteins intracellularly or extracellularly in transgenic T. gondii and E. tenella have been described before 6,12,21,24 .These targeting designs always rely on the organelle or surface proteins related to the trafficking mechanism.For surface localization, the C-terminal GPIanchored protein from TgSAG1 could relocate bacterial alkaline phosphatase 24 and TgMIC6 to the T. gondii tachyzoite surface 25 , which indicate that this sequence contains all the targeting signals required for surface.However, in our research, the GPI from EtSAG1 and EtSAG13 does not lead to EYFP expression on cell surface (Fig. 1A,D), but the whole or coding sequence minus signal peptide does (Fig. 1B,E), suggesting that sequences besides GPI of SAG proteins may also play a role in localization.In addition, when the EtSAG1 was fused to the N terminal of EYFP, EYFP was not shown on surface of sporozoites (Fig. 1C), suggesting that the fusing site relative to the target sequence is an important factor for the translocation of heterologous proteins.There were also puzzling results that EYFP were observed in RBs of extracellular sporozoites (Fig. 1A,C,D).Because there are no studies showing that location sequences of surface protein are related to target proteins to RBs, that are considered to be devoid of a membrane and usually involve in nutrient stores or metabolic energy reserves 26 .This phenomenon may be random because inappropriate fusions with the location sequences of SAGs, but this preference for RBs is worth further study.For microneme localization, the whole EtMIC2 targeting the foreign protein to the apical end have been reported 13 , which agrees perfectly with our observation: we also targeted EYFP to apical end by fusing EYFP to the C-terminal of EtMIC2 in extracellular sporozoites (Fig. 2A,B), and this position is the ideal location of the EtMIC2 27 .And we showed EaMIC2, which was located mainly at the apical tip of the sporozoite and in the merozoite 28 , also could target EYFP to apical end (Fig. 2C,D).In addition, we showed that the location of MIC2 was not affected by different promoters.The signal peptide of MIC proteins translocated the heterologous proteins to all over the cavity of sporocysts in E. tenella 12,13 .In this study, we showed that signal peptide of EaMIC2 and EaMCP1, could target EYFP to the cavity of sporocyst in E. acervulina (Fig. 3).And some of the above strategies for surface or microneme could target EYFP to the corresponding position in E. acervulina (Figs. 4 and  5).Though EYFP fewly co-localized with a monoclonal antibody against the EtMIC2, it mainly accumulated in the apical end of sporozoites (Fig. 5C), the position where EaMIC2 located 28 .But the colocalization need to be identified in further study.Even so, another species of Eimeria, E. acervulina was developed further for potential vaccine carrier.Although we achieved specific targeting of heterologous proteins in two Eimeria species, the transport mechanisms are still not fully understood.Our study showed that GPI-anchored sequence is not the only sequence that determines surface location.But the control mechanism of MICs trafficking seems clearer that the signal peptide is responsible for extracellular secretion, and the C-terminal sequence is responsible for targeting to the micronemes 29 .We did not make a further exploration for the specific selected proteins/region for targeting, which should be resolved in the future, especially for understanding the interactions between the Eimeria parasites and the host.On the other side, the using of the whole organelle proteins may play a positive role on inducing immune responses for the host, because that they were known as protective antigens 28,[30][31][32] .As research continues with the use of genetically engineered E. acervulina parasites as live vaccine vectors, additional data should resolve some of the issues related to mechanisms, as well as assessment of efficacy for vaccine-induced immune responses.Table 2. List of constructs used in the study and their location features.www.nature.com/scientificreports/Our studies provide methods for heterologous protein localization in transgenic Eimeria, which is the first implementation in E. acervulina.These strategies are expected to benefit future studies that use live Eimeria parasite as vaccine vectors that induce both cellular and humoral immune responses.

Ethics statement
The animal studies described in this work followed institutional guidelines for animal welfare and biosafety and were approved by China Agricultural University Institutional Animal Care and Use Committee (CAU20160628-2).The study is reported in accordance with the ARRIVE guidelines.

Parasites and animals
The wild type strain of E. tenella (EteWT) used in this study is the E. tenella Houghton strain, and the wild type strain of E. acervulina (EacWT) is the E. acervulina Beijing strain, as maintained in coccidia-free, Arbor Acre (AA) broilers by the National Key Laboratory of Veterinary Public Health and Safety at CAU. Procedures for the collection, purification, sporulation 33 , and extraction of sporocysts and sporozoites were carried out as previously described 34 , except that when detecting EYFP expressing in transgenic E. acervulina with surface location, sporulated oocysts in PBS were vortexed in the presence of glass beads until rupturing of most walls of oocysts and sporocysts, sporozoites were observed by fluorescence microscopy, in the mixing state with unbroken oocysts, sporocysts, oocyst shells, and sporocyst shell.
One-day-old AA broiler chickens were purchased from Beijing Arbor Acres Poultry Breeding Co., Ltd.They were fed with a pathogen-free diet and water ad libitum in coccidian-free housing units.After the experiment, the cervical dislocation was performed for chickens necessary for euthanasia, which aims to lose consciousness of chickens rapidly.
The genes DHFR and EYFP were amplified by PCR using plasmid pSDEP2ARS 38 as a template.And the backbone was acquired by linearizing the pSDEP2ARS 38 with SnaBI.
The amplification primers and the information of above products are shown in Supplementary Table S1.These products were cloned into pEASY ® -Blunt Cloning Vector (TransGen Biotech Co. Ltd, Beijing) for sequencing and extracting plasmids.Then these plasmids were used as templates for fragments needed by final plasmid construction using Seamless Cloning and Assembly kit (TransGen Biotech Co. Ltd, Beijing).

Transfection and selection of transgenic parasites
The constructed plasmids, linearized with SnaBI, were transfected into 1 × 10 7 EacWT or EteWT sporozoites by restriction enzyme-mediated integration as previously described 39 .After nucleofection (Program U-033, AMAXA, Switzerland), the transfected E. tenella sporozoites were inoculated into the ileocecal opening via the cloaca, and the transfected E. acervulina sporozoites were inoculated intravenously into the wing vein of five 1-week-old chickens equally for stable transfection 23 .The transgenic oocysts were selected in chickens by the MoFlo ® Cell Sorter (Dako-Cytomation, Fort Collins, CO) on the single-cell mode in vitro, and/or 150 mg/L pyrimethamine (Sigma-Aldrich Co., St. Louis, Mo., USA) press by drinking water in vivo, then inoculated into coccidian-free chickens for the propagation of next generation oocysts.

Passage in vivo
For E. tenella, 1 × 10 4 oocysts from the 1st generation transgenic population were orally inoculated to an average of five 10-day-old, coccidia-free AA broilers, and the next generation was collected during 6-9 days post inoculation under the drug press.For E. acervulina, 6 × 10 4 oocysts from the 1st generation transgenic population were orally inoculated to an average of three 7-day-old, coccidia-free AA broilers, and the next generation was collected during 5-8 days post inoculation under the drug press.

Observation of fluorescent reporter protein
Micrographs of the transgenic populations, including sporulated oocysts, sporocysts, or sporozoites expressing EYFP was observed using the fluorescent microscope (OLYMPUS, Japan) and/or confocal microscope (SP5, Leica, Germany).www.nature.com/scientificreports/Eight 7-day-old AA broilers were orally inoculated with 1 × 10 6 sporulated oocysts of transgenic E. acervulina.Every 12 h from 24 to 108 h post inoculation, one chicken was euthanized and necropsied to collect the middle part of duodenum with 3 cm, which was washed with cold PBS, then the mucosa was scraped for smears and observed for endogenous developmental stages using confocal microscope.
For in vitro observation, 2 × 10 5 freshly purified transgenic sporozoites were inoculated onto human foreskin fibroblast (HFF) monolayers grown on glass coverslips in 12-well plates.Then indirect immunofluorescence assay was conducted 12 h post invasion as previously described 23 , with a monoclonal antibody against EtMIC2 and Cy3-conjugated goat anti-mouse IgG (H + L) (Proteintech, America).5 × 10 6 freshly purified transgenic sporozoites were fixed in 2% glutaraldehyde for 24 h, following a routine concluding washing, dehydrating, embedding and sectioning, immunoelectron microscopy assay was conducted as previously described 12 .

Figure 4 .
Figure 4. Targeting EYFP to the cell surface of E. acervulina.(A) The construct ss-EYFP-EaSAG1, in which EYFP is fused between the signal peptide and C terminal mature peptide of EaSAG1.(B-D) Confocal microscopy for EYFP expression in sporulated oocysts (B), sporocysts (C), or sporozoites (D) mechanically released during oocyst ruptured by shaking with glass beads.Bar = 5 μm.(E-H) Confocal microscopy for EYFP expression observed during the endogenous developmental stages by duodenum smears.Including non-invaded sporozoites (E), invaded sporozoites (F), trophozoites (G) and schizonts (H).Bar = 5 μm.EYFP expresses on the surface of parasite in vivo and in vitro.

Table 1 .
Promoter and signal sequences tested for controlling EYFP expression in E. acervulina.