The experiments were performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals, Ministry of Science and Technology, China. All experimental proceduresw ere approved by the Institutional Animal Care and Use Committee of China Agricultural University (under the certificate of Beijing Laboratory Animal employee ID: 18049). The mice were humanely euthanized by cervical dislocation after anesthetization by subcutaneous injection of atropine (0.02 mg/kg) when they were unable to reach food or water for more than 24 h and lost 20% body weight.

Vero cells and human foreskin fibroblast (HFF) cells were purchased from the ATCC (Manassas, VA, USA) and serially passaged in our laboratories as previously described (Gaskell et al., 2009). The cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM, MACGENE, Beijing, China) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco, USA) in a humidified incubator containing 5% CO₂ at 37°C. The cells were passaged every 3–6 days by trypsinization.

The N. caninum Nc-1 wild-type strain and the newly constructed NcROP5-deficient strain (described below) were cultured as tachyzoites by serial passages in HFF or vero cells as previously described (Gaskell et al., 2009). Briefly, the parasites were cultured in DMEM (pH 7.4) supplemented with L-glutamine, 2% heat-inactivated FBS, penicillin (100 U/mL) and streptomycin (100 μg/mL) (Sigma–Aldrich, St. Louis, MO, USA) at 37°C in an atmosphere with 5% CO₂. The parasites were harvested by filtering through a 5.0 μm pore filter (Millipore, MA, USA), centrifuged, washed twice with cold phosphate-buffered saline (PBS), and centrifuged (2,000 rpm for 10 min each time) (Zhang et al., 2007).

The pTCR-CD plasmid was used at the Key Laboratory of Animal Parasitology (Beijing City, China) as previously described (Li et al., 2016). The plasmid contains the chloramphenicol resistance gene (CmR), red fluorescence protein gene (RFP), 5-fluorine cytosine gene (CodA), and ampicillin resistance gene (Amp). Nctublin promoter was used to drive the CmR-RFP fusion gene and CodA gene expression. The fragments of the NcROP5 5′ (2826 bp) and 3′ UTR regions (2297 bp) were amplified using the primer pairs F1/R1 and F2/R2 (Supplementary Table S1), which were designed based on the N. caninum Liverpool protein coding gene sequence¹ (FR823393, Chromosome: XII). Primers F1, R1, F2, and R2 introduced KpnI, XhoI, XmaI, and SpeI sites, respectively (Supplementary Table S1). After amplification, the fragments were cloned into a pEASY-T-Blunt vector (TransGen Biotech Co., Ltd, Beijing). Then the DNA insert was sequenced (Beijing Sunbiotech Co. Ltd., China) and analyzed by blast in ToxoDB¹ to confirm the authenticity of the cloned sequence. All alignments were performed using DNAMAN version v5.2.2 and the correct plasmids were double digested with KpnI and XhoI for the 5′ UTR and XmaI and SpeI for the 3′ UTR (NEB, USA). Then, the fragments were ligated to the pTCR-CD vector to produce the NcROP5 gene deleted vector, pTCR-NcROP5-CD KO.

All constructs were verified by sequencing (Beijing Sunbiotech Co. Ltd, China). After digestion with NotI (NEB, Beijing), the linearized pTCR-NcROP5-CD KO plasmids were purified using ethanol precipitation and then resuspended using cytomix (Hui et al., 2014). Nc-1 tachyzoites (1 × 10⁷), mixed with 50 μg of the above plasmids, were transferred to a 0.4 mm gap cuvette and electroporated with 2 kV at 25 μFd and 50 Ω with the Gene Pulser Xcell electroporation system (Bio-Rad, USA). Then the parasites cultured in an atmosphere containing 5% CO₂ at 37°C for 24 h prior to the addition of chloramphenicol (20 μM) and 5-fluorine cytosine (40 μM). The parasites were cultured consecutively to the 8th generation and then screened to confirm the purity of the selected strains.

For complementation analysis, pDMG plasmid (Hui et al., 2014; Tao et al., 2014; Li et al., 2016) was used to express NcROP5 in the ΔNcROP5 strain. The complete coding sequence of NcROP5 was amplified using primers F3/R3 (Supplementary Table S1) that appended the flanking EcoRV and AvrII restriction endonuclease sites. The amplification products were introduced into the same sites of a modified pDMG plasmid with GFP replaced by HA (Li et al., 2016). The pDMG-NcROP5-HA vector was electroporated into the ΔNcROP5 strain. The transgenic parasites were grown under pyrimethamine selection pressure. The selected NcROP5 complementary strain, named as iΔNcROP5, was identified by western blotting and immunofluorescence assays (IFAs).

Cell invasion assays were performed using the following protocol (Tao et al., 2014). Freshly egressed parasites (Nc-1, ΔNcROP5 and iΔNcROP5 strains) were allowed to settle onto the HFF monolayers on 12-well plates for 30 min at 37°C prior to incubation. Extracellular parasites were removed by washing three times with PBS. The cells were incubated for 20 h prior to IFA using anti-NcSRS2 primary and FITC-conjugated secondary antibodies (1:100 dilution). The number of vacuoles representing successful invasion parasites was counted under a fluorescence microscope. A total of three replicates (each clone in duplicate) were performed.

Plaque assays were set up immediately following the initiation of the invasion assays using the same parasite suspensions. Nc-GFP strain (kindly provided by Professor Xuenan Xuan, Obihiro University of Agriculture and Veterinary Medicine, Japan), as a null vector constructing group, which is a transgenic parasite by transfecting Nc-1 wild-type strain with the pDMG plasmid, was propagated as tachyzoites by serial passages in HFF cells as previously described (Gaskell et al., 2009). The plaque assay was performed as previously described (Hui et al., 2014). Five hundred parasites were added to each well of a 6-well plate containing confluent host cell HFF monolayers. The plate was incubated for 7 days in a 37°C incubator with 5% CO₂, fixed with 4% paraformaldehyde for 15 min and stained with 2% crystal violet for 15 min at room temperature. The stained wells were washed with deionized water, air dried and visualized using a fluorescence microscope (IX71, Olympus, Japan). The plaque area measurement was performed as previously described (Roos et al., 1994).

The Nc-1, ΔNcROP5 and iΔNcROP5 parasites were harvested and counted. Five-hundred parasites were added to confiuent vero and HFF cells on 25 mm coverslips. The coverslips were allowed to sit at 37°C for 30 min in DMEM with 20% FBS and then washed three times with PBS. The coverslips were incubated at 37°C for 24 h in DMEM with 2% FBS and then fixed with 4% paraformaldehyde for 15 min. IFAs were performed using anti-NcSRS2 primary (Tao et al., 2014) and FITC-conjugated secondary antibodies (Proteintech, USA). The number of parasites per vacuole was counted under a fluorescence microscope. The experiments were performed in triplicates and repeated three times, and 200 vacuoles were counted per coverslip.

Neospora caninum tachyzoites, as well as T. gondii, must form parasitophorous vacuole (PV) after invading the host cell (Tilley et al., 1997). They proliferate in the PV and then egress out of PV. Then, the tachyzoites invades other cells again (Soldati and Meissner, 2004). In our experiments, parasites were inoculated in 12-well plates containing prepared HFF cells for 30 h. Egress was stimulated with 10 μM of the Ca²⁺ ionophore A23187 from Streptomyces chartreusensis (Sigma, USA) for 2 min at 37°C prior to fixation with paraformaldehyde (Williams et al., 2015). The IFA was performed as described above using the mouse anti-NcSRS2 primary and FITC-conjugated secondary molecules. The average number of ruptured vacuoles was determined by counting a minimum of 100 vacuoles per slide and three slides for each experiment. Three independent experiments were performed.

Total RNA was extracted from 1 × 10⁷ tachyzoites of the Nc-1 wild-type strain and three ΔNcROP5 clones with TRIzol reagent (Invitrogen, USA). cDNA was synthesized using the EasyScript First-Strand cDNA Synthesis SuperMix kit (TransGen, China). Specific primers were designed using Primer Premier 5.0 (Hui et al., 2014), including primers for rhoptry necks (RON2 and RON4), rhoptrys (ROP4, ROP5, ROP7, ROP16, and ROP17), dense granules (GRA2, GRA6, and GRA7) and the endogenous reference gene NcActin (Supplementary Table S1). The specificity of these primers was evaluated using conventional quantitative real-time PCR (qRT-PCR). The qRT-PCR was conducted using the ABI Prism 7500 System (Biosystems Inc., USA) with SYBR Green II (Takara Biotechnology, Dalian, Co., Ltd) following manufacturer’s instructions. The resulting RNA concentrations were normalized using Ncactin (Wang et al., 2017), and the relative expression levels of the target genes were analyzed using the ABI Prism 7500 software v2.0.5 (Biosystems Inc., USA). The RT-PCR conditions were as follows: 94°C for 5 s, 40 cycles of 94°C for 5 s and 60°C for 30 s. The relative expression of genes was calculated using the 2^-ΔΔCt method and standard deviation was calculated from three replicates (Cárdenas-Mondragón et al., 2016; Gagnaire et al., 2016).

The N. caninum Nc-1 and ΔNcROP5 strains were cultured as tachyzoites in HFF cells as previously described (Gaskell et al., 2009) for 96 h. For each parasite, a total of 5 × 10⁷ tachyzoites were harvested and isolated by washing in cold PBS, centrifugation, resuspension in cold PBS, syringing three times through a 27-gaure needle, filtering through a 5.0 μm pore filter (Millipore, MA, USA), washing three times with PBS, and finally centrifugation at 1800 rpm for 10 min (Tao et al., 2014). Then the parasites were ultrasonically lysed in protein lysis buffer (8.0 M urea, 100 mM pH 8.0 Tris-HCl, 1×cocktail) using an ultrasonic liquid processor (Sonic & Materials INC., USA). The lysates were sonicated five times for 2 s each at 30% amplitude. Then the sonicated lysates were centrifuged for 30 min at 4°C at 12000 g and obtained the supernatants (Deutsch et al., 2014). After quantitative assay by the Bradford method (Bradford Protein Assay kit) (Deutsch et al., 2014), 60 μg protein was reduced, alkylated and trypsin digested. Each sample was separated by a high performance liquid chromatography (HPLC) system (Ultimate 3000, Thermo Scientific), and analyzed by mass spectrometry (Q-Exactive HF, Thermo Scientific). The spectra were identified and analyzed by the Mascot software (Thermo Scientific, USA) and Proteome discoverer 2.1 (Thermo Scientific, USA) on the basis of database listed in Supplementary Table S2 (Sandberg et al., 2012).

To compare the virulence of the ΔNcROP5 strain with the parental Nc-1 strain and complemented iΔNcROP5 strain, tachyzoites (3 × 10⁶) were injected i.p. into 6-week-old female BALB/c mice (five mice per strain), which were purchased from the Laboratory Animal Center of Academy of Military Medical Sciences (Beijing, China). Survival was monitored for 30 days. To examine the cerebral load of parasites, tachyzoites were injected i.p. into 6-week-old female BALB/c mice at a dose of 1 × 10⁶ (five mice). The mouse brains were examined for parasites on the 30th day (Pan et al., 2014).

The statistical analysis of all of the data was performed using GraphPad Prism 5 v. 5.01 (San Diego, CA, USA). The results were expressed as mean ± SD and evaluated by non-parametric tests. Values of P < 0.05 and P < 0.01 were considered statistically significant and very significant, respectively.

Using the knockout plasmid, pTCR-CD, we successfully constructed the NcROP5 knockout strain. To use the plasmid for deleting a gene, we ligated the 5′ UTR and 3′ UTR regions of the target gene. Then, we analyzed the genome of N. caninum Liverpool strain in ToxoDB² and identified the sequences with high identity to TgROP5, which were NcLIV_060730, NcLIV_060740, and NcLIV_060741. Compared to the tandem cluster of polymorphic alleles of TgROP5 (TgROP5A, TgROP5B, and TgROP5C) (Behnke et al., 2011), we believed that NcROP5 would contain NcROP5A, NcROP5B, and NcROP5C, coded by the above three genes, respectively. Homology of nucleotide and protein sequence between NcLIV_060730 and TgROP5A were 52.72 and 76.76%, respectively. Since TgROP5 is an important virulence factor of T. gondii (Behnke et al., 2012), we suspected that NcROP5 might play an active role in N. caninum virulence. To examine the physiological role of ROP5 in the Nc-1 strain, NcROP5 KO parasites were generated by targeted gene disruption. The non-coding region (UTR) sequences of NcROP5, NcROP5 5′ UTR and 3′ UTR, were cloned (Supplementary Figure S1A) and ligated to the pTCR-CD vector, forming pTCR-NcROP5-CD KO (Supplementary Figure S1B). The plasmid was identified by restriction enzyme digestion (Supplementary Figure S1A) and sequenced and then transfected into Nc-1. Theoretically, the entire ROP5 coding region, containing NcROP5A, NcROP5B, and NcROP5C, could be replaced with the CmR and RFP genes (Supplementary Figure S1C). RFP was expressed at normal levels in the suspected ROP5 mutants (Figure 1A). Following CmR selection of the RFP and 5-fluorine cytosine-negative parasites, we isolated three mutants out of 218 clones.

To generate the ROP5 gene-deficient Nc-1 strain, the NcROP5 gene was targeted and replaced by CmR-RFP. PCR was used to confirm the deletion of the NcROP5 gene. NcROP5 gene amplification was negative in the parasite genomes (Figure 1B). NcROP5 deficiency was confirmed using western blotting (Figure 1C) and RT-PCR (Figure 1D). We also successfully generated a strain with complete knockout of the NcROP5 gene. These results suggested that NcROP5 was not an essential, surviving gene in the parasites.

To verify the biological role of NcROP5, we obtained the complete coding sequence of NcROP5A without the termination codon TAA (NcLIV_060730, 1647 bp, Supplementary Figure S2A) and generated the complementary plasmid pDMG-NcROP5-HA. The expression of NcROP5 with an HA-tag and stop codon TAA at the C-terminus was driven under the control of the NcGRA1 promoter (Supplementary Figure S2B). We correctly identified the plasmid by restriction enzyme digestion and DNA sequencing (Supplementary Figure S2A), and then electroporated it into one of the ΔNcROP5 strains. After repeated passage, we selected strains that stably expressed NcROP5 (named as iΔNcROP5). NcROP5 expression in the complementary strain based on WB and RT-PCR was comparable to the expression of the wild strain (Figures 2A,B). IFA was used to analyze the subcellular localization of NcROP5 in the iΔNcROP5 strains using a mouse anti-HA monoclonal antibody. NcROP5 was co-localized with HA at the anterior end of the extracellular parasites (Figure 2C).

Plaque formation measures the viability of parasites in cell culture and reflects parasite motility on the surface of the host cell layer, invasion, intracellular growth and egress (Plattner et al., 2008). We assessed the ΔNcROP5 strain using the plaque assay. The ΔNcROP5 strain formed noticeably smaller and considerably fewer plaques than the parental strain and the complementary strain (Figures 3A,B). The ΔNcROP5 strain showed significant weakening of host cell invasion (Figure 3C). Intracellular replication assays were performed to investigate if deleting ROP5 in the Nc-1 strain weakened the ability of the parasite to proliferate in host cells. Parasites that invaded host cells 24 h after inoculation were analyzed for intracellular growth by counting the number of parasites per vacuole. Due to the lack of synchronization of host cell invasion, the intracellular vacuoles contained 1, 2, 4, or 8 parasites. However, intracellular replication of ΔNcROP5 in Vero and HFF cells was lower than that of the wild type parasite (Nc-1). As shown in Figure 3D, the distribution of the eight parasites per vacuole was significantly lower for the ROP5 knockout strain than that for Nc-1, Nc-GFP, and iΔNcROP5, indicating that the intracellular replication of the ΔNcROP5 strain was slower than that of the other three strains. However, there was no significant difference in egress between the four strains (Figure 3E). These findings indicated that NcROP5 expression in N. caninum is a critical determinant of parasite phenotype, including host cell invasion and proliferation.

To analyze the cause of the ΔNcROP5 phenotype changes, we compared protein expression levels in Nc-1 and ΔNcROP5. Nc-1 and ΔNcROP5 tachyzoites were analyzed by label-free LC-MS/MS for quantification of peptides. A total of 18,809 peptides with FDR ≤ 0.01 were identified, and a total of 1149 proteins with expression difference ≥ 1.2-fold between the two strains were functionally annotated according to the N. caninum genome annotation in the ToxoDB database (Supplementary Figure S3A and Table S3). The three main annotation types were biological process, cellular component and molecular function (Supplementary Figures S3B–D). The differentially expressed proteins were classified into 31 functional groups (Supplementary Figure S3E), with biological process possessing 12 GO (gene ontology annotation) terms, cellular component 10 GO terms, and molecular function 9 GO terms. Cellular process and metabolic process were major biological processes (32 and 31%, respectively), cell part and cell are major cellular components (25% each), and binding and catalytic activities were major molecular functions (53 and 37%, respectively).

To further identify biological pathways of the differentially expressed proteins, the number of proteins expressed in Nc-1 and ΔNcROP5 and differences between the two strains are shown in Supplementary Figure S4A. Demographic analysis demonstrated that 76 proteins were increased and 89 proteins were decreased with a 2.0-fold change; 198 proteins were increased and 252 proteins were decreased with a 1.5-fold change; and 477 proteins were increased and 671 proteins were decreased with a 1.2-fold change. We predicted that 477 increased proteins participated in 92 biological pathways and 671 decreased proteins participated in 88 biological pathways by KEGG pathway analysis³. The differentially expressed proteins were matched with the proteins annotated in the KEGG pathway database, and they were compared to determine their involvement in the KEGG pathway. We found that five KEGG pathways were clearly affected (P ≤ 0.05) (Supplementary Figure S4B and Table S4), followed by mismatch repair, SNARE interactions in vesicular transport, glycosylphosphatidylinositol (GPI)-anchor biosynthesis, ABC transporters and RNA degradation.

To better comprehend the causes of phenotypic differences, we analyzed representative differentially expressed proteins and uncommented proteins by gene and sequence alignment. Blast analysis (ToxoDB-28_Neospora caninum_Annotated Proteins. fasta) showed that increased proteins included apetela 2 (AP2) family proteins (NCLIV_063920, NCLIV_032930, and NCLIV_059950) and dense granule protein (NCLIV_004260), and decreased proteins included cell division and proliferation proteins (NCLIV_004280, NCLIV_005020, NCLIV_069400, NCLIV_065020, NCLIV_067140, and NCLIV_009310), and secreted proteins (NCLIV_038100) (Table 1).

Rhoptry neck proteins (RONs), ROPs and dense granule proteins (GRAs) were related to invasion, proliferation, and formation of PV (Sinai and Joiner, 2001; Dubremetz and Lebrun, 2012; Nolan et al., 2015). To detect whether these genes were affected by NcROP5, we selected RON2, RON4, ROP4, ROP7, ROP16, ROP17, GRA2, GRA6, and GRA7 genes to evaluate their mRNA expression. Quantitative RT-PCR (qRT-PCR) analysis of 1 × 10⁷ Nc-1, ΔNcROP5 and iΔNcROP5 tachyzoites was conducted to determine the potential involvement of NcROP5 in transcriptional regulation of other related genes. The qRT-PCR analysis indicated that RON2 (R² = 0.9838, slope = -3.392), RON4 (R² = 0.9849, slope = -3.425), ROP4 (R² = 0.9961, slope = -3.381), ROP7 (R² = 0.9918, slope = -3.466), ROP16 (R² = 0.9907, slope = -3.372), ROP17 (R² = 0.9899, slope = -3.373), GRA2 (R² = 0.9911, slope = -3.407), and GRA6 (R² = 0.9951, slope = -3.475) were not altered at the transcription level in ΔNcROP5. Notably, NcGRA7 (R² = 0.9879, slope = -3.331) expression was significantly down-regulated by approximately twofold in ΔNcROP5 (Figure 4). All genes have similar amplification efficiency to the reference gene Ncactin (R² = 0.9912, slope = -3.392). The results suggested that NcROP5 expression might affect the expression of other genes, including NcGRA7.

To investigate if NcROP5 is necessary for Nc-1 virulence, mice were infected intraperitoneally with 1 × 10⁶ Nc-1, Nc-GFP, ΔNcROP5, or iΔNcROP5 tachyzoites, followed by observation of parasite load in the brain and survival rate. Significant differences were observed between the ΔNcROP5 group and the other groups (Nc-1, Nc-GFP and iΔNcROP5). Signs of illness such as rough coats, inactivity, or nervous system signs (hind limb weakness, head tilting, or walking in circles) were observed 2 or 5 days before death. In addition, rough hair coats and reduced activity were observed in 2 Nc-1-infected mice from day 28 post-infection, 1 Nc-GFP-infected mouse from day 29 post-infection and 1 iΔNcROP5-infected animal from day 26 post-infection. There were no clinical signs in ΔNcROP5-infected mice during the 30 days observation period. The cerebral parasite load of the ΔNcROP5 group was approximately one-fourth of the load in the Nc-1 and Nc-GFP groups. The iΔNcROP5 group had the most parasites in the brain (Figure 5A). Consistent with the low parasite load in the brain, the survival rate of the ΔNcROP5-infected animals was higher than that of the other groups. Actually, all ΔNcROP5-infected mice survived, compared with the survival rate of 40–60% for the other groups (Figure 5B). These results demonstrate a requirement of NcROP5 for Nc-1 virulence.