EtcPRS^Mut as a molecular marker of halofuginone resistance in Eimeria tenella and Toxoplasma gondii

Summary

The control of coccidiosis, causing huge economic losses in the poultry industry, is facing the stagnation of the development of new drugs and the emergence of drug resistance. Thus, the priority for coccidiosis control is to decipher the effect mechanisms and resistance mechanisms of anticoccidial drugs. In this study, we mined and validated a molecular marker for halofuginone resistance in Eimeria tenella through forward and reverse genetic approaches. We screened whole-genome sequencing data and detected point mutations in the ETH2_1020900 gene (encoding prolyl-tRNA synthetase, PRS). Then, we introduced this mutated gene into E. tenella and Toxoplasma gondii and validated that overexpression of this mutated gene confers resistance to halofuginone in vivo and in vitro. These results together show that mutations A1852G and A1854G on the ETH2_1020900 gene are pivotal to halofuginone resistance in E. tenella, encouraging the exploration of mechanisms of drug resistance against other anticoccidial drugs in eimerian parasites.

Graphical abstract

Highlights

• •Halofuginone-resistant Eimeria tenella strains harbor mutations within EtcPRS
• •EtcPRS^Mut could be a powerful selectable marker for genetic manipulation
• •These findings could enable studies of genotype-phenotype relationships in Eimeria

Molecular biology; Parasitology; Medical microbiology

Introduction

Chicken coccidiosis is caused by protozoan parasites of the genus Eimeria and is one of the most damaging parasitic diseases in the poultry industry, causing economic losses of approximately US$ 3 billion a year worldwide.^1,2 Effectively controlling chicken coccidiosis is largely dependent on anticoccidial drugs, and to an extent, vaccines.^3,4 To date, two categories of anticoccidial drugs, polyether antibiotics and synthetic compounds, are being used in the poultry industry.⁵ Unfortunately, drug-resistant eimerian parasites emerged quickly following the extensive use of anticoccidial drugs.^1,6 However, there are no pieces of evidence that have yet to unequivocally identify loci responsible for resistance against anticoccidial drugs when eimerian parasites resistant to nearly all drugs are reported.

Febrifugine is the active component of Dichroa febrifuga; halofuginone is a halogenated derivative of febrifugine and is currently used as a coccidiostat against apicomplexan parasites, including inhibiting the invasion and early development of Eimeria.^7,8,9 However, halofuginone-resistant strains have been reported, and no evidence has been reported describing the molecular mechanism of halofuginone resistance in Eimeria. Previous studies have shown that drug-resistant strains can be selected under a dose-escalation, and transcriptome analysis has identified some candidate genes that may confer resistance to anticoccidial drugs.^10,11,12 Thus, the lack of molecular markers hinders the rapid detection of these anticoccidial-resistant eimerian parasites in the field.

Recently, the prolyl-tRNA synthetase (PRS) enzyme, which is a member of the aminoacyl-tRNA synthetase family, was verified as the molecular target of halofuginone in Plasmodium falciparum.^{13,14,15,16,17,18} Interestingly, the high sequence conservation of the PRS protein among apicomplexan parasites suggested that these parasites can be attacked by targeting invariant essential proteins.^15,19 These breakthroughs provide potential clues for mining and validating molecular marker(s) of halofuginone resistance in Eimeria.

Directed evolution coupled with whole-genome sequencing has been successfully used for the identification of resistance targets and pathways for many antiparasitic compounds, such as artemisinin, quinoline, and atovaquone in P. falciparum and Toxoplasma gondii.^{13,20,21,22,23,24,25,26} Recently, reverse genetic approaches based on CRISPR/Cas9 gene editing and overexpression were also applied for the identification of mutations in candidate genes.^27,28,29 These successful shreds of evidence have suggested that the molecular markers of drug resistance in Eimeria could be verified through experimental evolution and genomic characterization of drug-resistant strains.

Here, we report, for the first time in eimerian parasites, the exploration of the potential markers of resistance to halofuginone. We used an accelerated evolution process to select mutated Eimeria tenella strains conferring resistance to halofuginone. We identified a single amino acid mutation in a protein of the resistant strains, with this mutation reaching fixation over time. We found that wild-type parasites obtained resistance to halofuginone when overexpressing this mutated gene, showing that the mutation confers resistance to halofuginone in E. tenella. Hence, our current work demonstrates that forward and reverse genetic approaches are valuable toolboxes for deciphering the molecular mechanisms of resistance in coccidia against more anticoccidial drugs.

Results

Establishment of halofuginone-resistant mutants of E. tenella

We speculated that eimerian parasites convert to halofuginone resistance via mutations of candidate genes, and those parasites could be enriched along with drug stress screening. To accelerate the selection of halofuginone-resistant E. tenella strains, we designed an experimental evolution strategy (Figure 1A).³⁰ In the first experiment, the peak oocyst output in Group 2 (chickens fed with halofuginone, 3 mg/kg) occurred at approximately 35 dpi (Figure 1B). The oocysts were collected at 35 dpi and confirmed as halofuginone-resistant strains when the oocysts were subjected to a drug resistance test (Table S1). Thus, we obtained halofuginone-resistant strains in less than 2 months.

Figure 1.

Selection and characterization of halofuginone-resistant E. tenella strains

(A) Schematic illustration of the experimental evolution strategy.

(B) Oocyst output curves for Group 1 and Group 2. Oocysts in the litter were measured every single day from day 7 to 52 post inoculation.

(C) Total oocyst output of chickens inoculated with 10⁴ oocysts of the wild-type strain or one halofuginone-resistant strain. Birds were supplemented in feed with or without 30 mg/kg halofuginone. ∗∗∗p < 0.001.

(D) H&E stained and counted the number of vacuoles in the cecum sections. During the assay, birds were fed with or without halofuginone (30 mg/kg). The black arrow shows parasites at different developmental stages. Scale bar, 50 μm. The number of parasites was counted at different times.

See also Table S1 and Figure S1.

To acquire a batch of halofuginone-resistant strains, we carried out a second assay with 150 new one-day-old Arbor Acres broilers. These birds were used to propagate halofuginone-resistant oocysts. In this experiment, 80 halofuginone-resistant strains were collected from individual birds under 30 mg/kg halofuginone selection. We then used halofuginone at 30 mg/kg in feed for the propagation of some of these strains. After an additional 6 generations of propagation, 12 strains were ultimately attained.

We assessed the reproductivity of halofuginone-resistant strains (halofuginone at 30 mg/kg) and found no significant difference in oocyst output compared to wild-type E. tenella without drug exposure (Figure 1C). We also compared the endogenous development of the wild type and halofuginone-resistant strain by H&E staining of tissue sections. We observed first-generation schizonts in halofuginone-resistant parasites and found that parasites could complete whole endogenous development under drug pressure (Figures 1D and S1). In contrast, we did not find first-generation schizonts in wild-type parasites treated with halofuginone (Figure 1D). Overall, we acquired halofuginone-resistant strains by using the experimental evolution strategy.

Parasites resistant to halofuginone harbor mutations within EtcPRS

To determine the candidate gene associated with halofuginone resistance that had developed during selection, genomic DNA from all 12 selected strains, as well as that from the wild-type strains, was extracted and subjected to whole-genome sequencing. The whole-genome sequencing performed for these strains achieved 100 × average nucleotide coverage. Mapping of the reads to the reference genome was used to identify changes from parental lines that showed altered alleles due to nonsynonymous mutations in the selected populations. Selective sweeps occur when beneficial genetic variants increase in frequency due to positive selection together with linked neutral sequence variants.³¹ This results in genomic islands of reduced heterozygosity and increased differentiation between populations around the selected site. Using SNPs from resequencing accessions of halofuginone-resistant strains, we estimated the population differentiation statistics (F_ST) and heterozygosity (Hp) between the wild type and halofuginone-resistant strains, in a 50-kb sliding window across the genome. There were 7 chromosomes containing loci with F_ST > 0.95; thus, we excluded intergenic regions and regions of high heterozygosity, and there were just 2 outliers with F_ST > 0.95 and Hp < 0.05 (Figures 2A–2C and S2). The majority of these changes were found at the highest level of selection pressure (30 mg/kg), where they were present with an allele frequency of >90% (Table 1). Then, we excluded synonymous mutation sites. Surprisingly, mutations appeared in ETH2_1020900 in 100% of the reads in resistant strains but in 0% of the mutant allele reads in the wild-type strains (Figure S3A, Table 1).

Figure 2.

Whole-genome sequencing of selective sweep among halofuginone-resistant and wild-type strains

(A) Population differentiation (F_ST) between halofuginone-resistant and wild-type strains for 50-kb windows with 20-kb step is plotted along the chromosomes. Each dot represents a 50-kb window. The red dashed line indicates the significance threshold (p = 1) and the blue dashed line indicates the significance threshold (p = 0.95). The red dots show the loci with F_ST = 1, while the green dots show the loci with F_ST > 0.95.

(B) The F_ST values for the corresponding traits are plotted against positions on chromosome 10.

(C) A sliding window of average heterozygosity (Hp) on chromosome 10 of the wild type (red dot) and drug-resistant strain (blue dot). A 50-kb sliding window with a 20-kb step is depicted. The average heterozygosity (Hp) of halofuginone-resistant strains is also shown as blue dots for comparison.

See also Figures S2 and S3.

Table 1.

Mutations found in 4 candidate genes identified by whole-genome sequencing

Chromosome	Gene ID	Annotation	Frequency, %	Coding region change	Amino acid change
10	ETH2_1020900	prolyl-tRNA synthetase	100	A1852G	T618A
10	ETH2_1020900	prolyl-tRNA synthetase	100	A1854G	T618A
10	ETH2_1021000	Hypothetical protein	55	C928T	R310C
12	ETH2_1240900	Apicomplexan specific protein	100	C375T	H127H
12	ETH2_1240900	Apicomplexan specific protein	100	C1830T	A610A
12	ETH2_1241000	Alpha/beta hydrolase family/alpha/beta hydrolase fold	15	C923T	T308I

Based on the whole-genome sequencing data, we focused on the ETH2_1020900, which is annotated as prolyl-tRNA synthetase(Database: ToxoDB) (https://toxodb.org/toxo/) (Table 1, Data S1). A previous study in P. falciparum reported that the mutated PfcPRS gene (PF3D7_1213800, cytoplasmic prolyl-tRNA synthetase) could confer resistance to halofuginone.¹⁴ As it has a high similarity to PfcPRS (Figure 3A), it is referred to here as an EtcPRS ortholog. We amplified and sequenced this gene from drug-resistant and wild-type strains and then identified independently selected mutations, A1852G and A1854G, corresponding to a single amino acid substitution T618A (Figures S3B and S3C).

Figure 3.

Visualization of mutation in EtcPRS through molecular docking

(A) Molecular phylogenetic tree of PRS protein from different species. The neighbor-joining phylogenetic tree was constructed by MEGA 11. Alignment was performed using ClustalW. The scale bar is substitutions per site.

(B) Gene model of ETH2_1020900. Green modules represent different domains. Redline, mutation site. See also Data S1.

(C) Crystal structure of EtcPRS^Mut showing views of the point mutation. Zoom the view of mutated residue.

Molecular docking suggests a relationship between EtcPRS^Mut and the halofuginone resistance mechanism

To further confirm the mutation site, molecular docking was performed to visualize whether the docking form changed after introducing this mutation. Multiple sequence alignments showed high overall sequence conservation within the aminoacyl-tRNA editing domain, tRNA synthetase class II, anticodon-binding domain, and prolyl-tRNA synthetase domains of the EtcPRS protein within apicomplexan parasites (Figures 3A and 3B). The notable difference among different apicomplexan parasites is the presence of loop regions. Then, we generated the mutant EtcPRS protein and visualized the mutation site, and the mutation site altered the amino acid polarity and hydrophilicity (Figure 3C). When visualizing the mutation, the mutated residue was placed on the “active pocket” and localized in the prolyl-anticodon-binding domain, which is responsible for specificity in tRNA binding. We anticipated that this mutation may have a direct effect on its binding and activity. Based on these results, we speculated that the T618A mutant remains resistant to halofuginone in E. tenella.

Overexpression of mutated EtcPRS confers halofuginone resistance in E. tenella and T. gondii

To establish a functional link between the candidate gene and halofuginone resistance, we constructed an overexpression transgenic E. tenella strain harboring EtcPRS^Mut to introduce mutations into the wild-type parasites and we used a drug selection strategy and fluorescence to monitor growth more precisely (Figure 4A, Table 2). The fluorescence rates of the F1 and F2 generations showed that pyrimethamine (150 mg/kg) and halofuginone (30 mg/kg) had similar selective efficiencies (Table 2). The reporter-positive transgenic Eimeria tended to be stable after 5 generations under selection pressure (Figure 4B). The expression of EtcPRS^Mut transgenic parasites was demonstrated by western blotting using FLAG tag-specific antibody (Figure 4C). To further localize the EtcPRS in transgenic parasites, E. tenella sporozoites were stained with anti-Flag monoclonal antibody, and we found that the EtcPRS gene was mainly expressed in the cytoplasm of the transgenic Eimeria sporozoites (Figure 4D). Comparing the oocyst output curves between wild type and transgenic strains, the data showed that the expression of EtcPRS^Mut transgenic parasites could complete whole endogenous development under drug pressure (30 mg/kg) (Figures 4E and 4F). Hence, the introduction of the EtcPRS^Mut seen in the halofuginone-resistant strains could significantly increase resistance to halofuginone (30 mg/kg).

Figure 4.

The gene-editing approach identified EtcPRS^Mut as conferring halofuginone resistance in E. tenella

(A) Schematic illustration of the plasmid expressing EtcPRS^Mut. Redline, mutation site.

(B) Fluorescence observation of transgenic oocysts under the microscope. Scale bar, 5 μm.

(C) Western blot of total protein extract from the parental and EtcPRS^Mut-overexpressing strains. PVDF membranes were probed with anti-Flag to detect the presence of EtcPRS^Mut protein (upper panel), and anti-EtActin was used as a control for normalization (lower panel).

(D) Indirect immunofluorescence microscopy showing the localization of EtcPRS^Mut. Mouse anti-Flag mAb was used to detect the Flag-tagged EtcPRS^Mut protein followed by Cy3-conjugated goat anti-mouse IgG. Scale bar, 5 μm.

(E) Oocyst output curves of the wild type and EtcPRS^Mut-overexpressing strain. Each bird was inoculated with 500 oocysts.

(F) Oocysts outputs of the wild type, halofuginone-resistant and EtcPRS^Mut-overexpressing strain. During the assay, birds were fed with or without halofuginone (30 mg/kg).

Table 2.

Isolation of stable transgenic strain under drug selection

	Generation	Birds/survival	Age (d)	Drug concentration (mg/kg)	Oocyst yield	Fluorescence rate, %
EtcPRSMut-OE Strain	F1	4/4	10	Pyrimethamine (150 mg/kg)	1.88×107	0.57
	F1	4/4	21	Halofuginone (30 mg/kg)	8.15×106	0.3
	F1	4/4	14	Halofuginone (2 mg/kg)	1.57×107	0.07
	F2	5/5	32	Pyrimethamine (150 mg/kg)	1.88×107	23.1
	F2	5/5	34	Halofuginone (30 mg/kg)	2.42×107	37.5
	F3	4/4	21	Halofuginone (30 mg/kg)	4.793×106	57
	F4	5/5	7	Halofuginone (30 mg/kg)	3.82×107	81.3
	F5	4/4	19	Halofuginone (30 mg/kg)	6.39×107	97

The aforementioned data provide evidence that EtcPRS^Mut confers resistance to E. tenella. Phylogenetic analysis of the coding protein showed homology between E. tenella and T. gondii, and we sought to determine whether EtcPRS^Mut may also confer resistance to halofuginone in T. gondii (Figure 3A). We expressed the EtcPRS^Mut in the Δku80 RH strain, under the T. gondii tubulin promoter at the uracil phosphoribosyl transferase locus (Figure 5A). Combined with western blot and indirect immunofluorescene assay (IFA), we found that the expression of mutated EtcPRS in T. gondii exhibits the same protein size and localization compared to E. tenella (Figures 5B and 5C). The growth rate of the TgEtcPRS^Mut was distinguishable from that of the wild-type strain when assessed using the fibroblast plaque assay under different drug concentrations (0, 1, 2, 5, 8, and 10 nM) (Figure 5D). In contrast, the wild-type strain was completely blocked by the addition of halofuginone (5 nM) (Figure 5D). Further phenotypic analyses revealed that the parental line led to a severe defect in replication capacity under halofuginone (5 nM) in vitro, while the transgenic parasites showed normal growth (Figure 5E). Altogether, these data further indicate the role of EtcPRS mutations in drug resistance.

Figure 5.

EtcPRS^Mut confers halofuginone resistance in T. gondii

(A) Schematic illustration demonstrating the strategy used for generating the Tg-EtcPRS^Mut-3HA strain by targeting the UPRT locus and replacing it with exogenous EtcPRS^Mut-3HA under the control of the tubulin promoter.

(B) Western blot of total protein extract from the Δku80 RH and TgEtcPRS^Mut-overexpressing strains. Anti-HA was used to detect the presence of EtcPRS^Mut protein (upper panel), and TgActin was used as a control for normalization (lower panel).

(C) Confocal imaging demonstrates the expression of the EtcPRS protein using the anti-HA antibody under halofuginone (5 nM). The inner membrane complex was stained with anti-GAP45 (green). Scale bar, 5 μm.

(D) Plaque assay images represent the viability of TgEtcPRS^Mut-overexpressing and Δku80 RH strains under halofuginone. Parasites were treated with 0, 1, 2, 5, 8, and 10 nM halofuginone.

(E) Quantification of plaque assays for Δku80 RH, and TgEtcPRS^Mut-overexpressing parasites under different concentrations. Data are presented as box and whisker plots (median with min to max, n = 3 biologically independent experiments).

(F) Intracellular replication of Δku80 RH and TgEtcPRS^Mut-overexpressing parasites. Δku80 RH or TgEtcPRS^Mut-overexpressing parasites invaded fresh HFFs in the presence or absence of halofuginone (5 nM). The number of parasites in each parasitophorous vacuole was counted 24 h after tachyzoite invasion. Means ± standard deviations (n = 100) of results from four independent experiments (each with triplicates).

EtcPRS^Mut could be a powerful selectable marker for the genetic manipulation of T. gondii

To further assess whether EtcPRS^Mut can be used as a stable selective marker in T. gondii, we constructed three transgenic parasites that harbor this mutant gene (Figure 6). Western blot and IFA were used to identify these transgenic parasites (Figure 6). To determine the frequency of stable halofuginone resistance and to isolate transgenic clonal lines, the progeny of transfected parasites was tested after three passages in different concentrations of halofuginone. Multiple clones were obtained from independent cultures, and IFA and PCR were then performed to determine the flexibility of halofuginone selection and to confirm the efficiency of screening at different concentrations (Table 3). The results showed that the positive rates of all transfected parasites reached more than 92%, which suggested that EtcPRS^Mut could be used as a stable screening marker in T. gondii (Table 3, Figures S4 and S5).

Figure 6.

Insertion of the EtcPRS^Mut gene into T. gondii through CRISPR/Cas9-induced homologous recombination

(A) Schematic illustration of the genetic manipulation.

(B) Schematic of the CRISPR/cas9 strategy used to knockout the H2AX gene by inserting EtcPRS^Mut. The arrow in the 3′UTR represents the region targeted by the sgRNA.

(C) Intracellular replication of Δku80 RH and Tg-H2AX-KO parasites. Δku80 RH or Tg-H2AX-KO parasites invaded fresh HFFs in the presence or absence of halofuginone (0, 5, 8, and 10 nM). The number of parasites in each parasitophorous vacuole was counted 24 h after the invasion. Means ± standard deviations (n = 100) of results from four independent experiments (each with triplicates).

(D) Schematic of the CRISPR/cas9 strategy integrating the EtcPRS^Mut fusion protein into the 3′ end of the gene (H2AX/SDHB). The arrow in the 3′UTR represents the region targeted by the sgRNA.

(E) Western blot of total protein extracts from the Δku80 RH and transgenic strains. Western blots were probed with anti-Flag to detect the expression of the tagged gene (upper panel). TgActin was used as a control for normalization (lower panel). C1 and C2 represent the two clones, H2AX-Tag and SDHB-Tag, respectively.

(F) Confocal imaging demonstrates the expression of the tagged gene protein using an anti-Flag antibody under halofuginone (8 nM). The inner membrane complex was stained with anti-GAP45 (green). Scale bar, 5 μm.

(G) Plaque assay images representing the viability of Tg-H2AX-Tag, Tg-H2AX-Tag, Tg-H2AX-KO, and Δku80 RH strains. These transgenic strains were cultured under halofuginone for three passages.

See also Figures S4 and S5.

Table 3.

Site-specific integration of 3Flag-EtcPRS^Mut into the 3′ end of the gene (H2AX and SDHB) under different concentrations of halofuginone

sgRNA target	Size (bp) of homology arms	Drug concentration (nM)	Population	%, Frequency	Clone	%, Frequency	Frequency of clones by PCR, %
			Positive/Total		Positive/Total
H2AX-3′UTR	42,42	5	130/140	92.9	13/14	92.9	92.9
H2AX-3′UTR	42,42	8	276/283	97.5	12/13	92.3	100
SDHB-3′UTR	42,42	8	256/260	98.5	8/8	100	100
SDHB-3′UTR	42,42	10	102/103	99.1	10/10	100	100

Discussion

The lack of drug-resistant molecular markers hinders the rapid detection of drug-resistant parasites in the field. In this study, we took advantage of the experimental evolution strategy and whole-genome sequencing to explore genomic mutations during halofuginone selection. We detected a strong selective sweep at the EtcPRS (ETH2_1020900) locus in halofuginone-resistant strains and identified an amino acid substitution of T618A in the coding sequence of EtcPRS. Then, we genetically engineered mutations into a wild-type strain and confirmed that these mutations confer resistance to halofuginone in E. tenella and T. gondii.

Obtaining drug-resistant strains is a prerequisite to studying the mechanisms of drug resistance in Eimeria parasites. At present, there are two classic approaches to acquire drug-resistant strains. Drug-resistant parasites can be directly isolated from fields or selected by using a stepwise increase in drug concentration to derive parasites with elevated drug tolerance.^{21,23,32,33,34} However, strains isolated from the field have complex genetic backgrounds, resulting in confusing SNP signals which will interfere with resequencing analysis.³⁵ In previous studies, drug-resistant strains of E. tenella were generated using a dose-escalation strategy, for example, maduramycin (20 generations), diclazuril (18 generations), and monensin (35 generations).^10,36 The limitation of this strategy is the low rate of spontaneous mutations per generation and thus it will take a long time to yield drug-resistant strains. In our work, we accelerated the induction of halofuginone-resistant strains and acquired strains with resistant phenotypes within just 55 days. Compared to the two approaches mentioned above, this novel procedure could yield many resistant strains in a relatively short time for the selection of resistant strains to other anticoccidial drugs.

To date, candidate genes corresponding to several antiprotozoal drugs have been reported in the apicomplexan parasites P. falciparum and T. gondii.^{13,20,21,22,27,28,29,37} In plants and animals, the sliding window-based method was used to detect selection signatures.^38,39 To better exclude the inference of irrelevant SNPs, we used a sliding window-based approach to analyze the candidate loci and we also excluded the other mutation sites by evaluating the frequency and heterozygosity of loci. Because nonsynonymous mutations always affect the phenotype, we also deleted the synonymous mutation sites. After that, we successfully characterized several selective signals related to halofuginone resistance and finally narrowed them to two mutations in ETH2_1020900 (EtcPRS), which are the most likely loci conferring resistance in Eimeria. Surprisingly, only A1852G was identified from resistant strains other than those induced in the laboratory when we amplified and sequenced this gene. Notably, there is prior evidence that a point mutation in PF3D7_1213800 (PfcPRS), resulting in L482F, correlates with halofuginone resistance in P. falciparum.^13,14 In accordance with these studies of halofuginone resistance, there is strong evidence of the correlation between EtcPRS^Mut and the halofuginone-resistance phenotype.

We also validated that EtcPRS^Mut confers halofuginone resistance in T. gondii when this mutated gene was introduced into wild-type parasites. Together with the above findings and core sequence homology analysis, it is tempting to speculate on the broad-spectrum antiprotozoal activity of halofuginone within the phylum Apicomplexa. A recent study showed that the development of malarial parasite tolerance to halofuginone is attributed to the upregulation of intracellular proline through the adaptive proline response.^40,41 Further work will be needed to elucidate the detailed mechanism in E. tenella.

Taken together, we identified that EtcPRS (ETH2_1020900) is directly related to halofuginone resistance in E. tenella. This work also proves that experimental evolution and genome sequencing can be useful to study the relationship between the phenotype and genotype of Eimeria parasites. As PRS^Mut is conserved in these apicomplexan parasites, it would be a druggable target for developing new chemical drugs.

Limitations of the study

In this study, we deciphered that the mutations in the E. tenella PRS gene (EtcPRS) are directly linked to halofuginone resistance through forward and reverse genetic approaches. Our approaches, though detailed, have some limitations. Halofuginone-resistant strains in this study were obtained by laboratory induction, with clear background and low spontaneous mutation rate of the resistant strains induced in the laboratory setting as previous studies demonstrated.⁴² However, resistant strains obtained from field may have more mutations which may also confer resistance to drugs.^43,44 Confirming the correlation of these mutations with halofuginone resistance is challenging because it requires many strains collected from different regions.²¹ In addition, the lack of conditional gene knockout approaches limits the further exploration of essential gene function in E. tenella. Last but not least, the lack of successive in vitro culture system for coccidia also limits the detection of the detailed changes after drug treatment. Though the future studies warrant elucidating the exact mechanisms that mediate drug resistance in coccidia, the current study still provide insight into the strategy to study the relationship between phenotype and genotype.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse Flag monoclonal antigen	Sigma	RRID: AB_1960908
Cy3-conjugated goat anti-mouse IgG	Proteintech	RRID: AB_10892835
FITC-conjugated goat anti rabbit IgG	Proteintech	RRID: AB_1512558
Hoechst 33258	Macgene	RRID: AB_2651133
HRP-conjugated goat anti-mouse IgG	Macgene	RRID: AB_895481
Mouse HA monoclonal antigen	Sigma	RRID: AB_2751029
Chemicals, peptides, and recombinant proteins
Trypsin	Biotopped	9002-07-7
Sodium taurodeoxycholate hydrate	Sigma	207737-97-1
Q5® High-Fidelity DNA Polymerases	NEW ENGLAND Biolabs	M0491
Deoxynucleotide (dNTP) Solution Mix	NEW ENGLAND Biolabs	N0447S
Q5 High GC buffer	NEW ENGLAND Biolabs	B9028A
Q5 Reaction buffer	NEW ENGLAND Biolabs	B9027S
Percoll	CYTIVA	17089101
SnaBI restriction enzyme	NEW ENGLAND Biolabs	R0130L
4% paraformaldehyde (PFA)	NOVON	SS0312
RIPA	Macgene	MP015
DMSO	Applichem	A3672
0.25% Triton X-100	Sigma	MC0711
Albumin Bovine V	Macgene	CA003.50
Halofuginone	MCE	CAS No.: 55837-20-2
DMEM	Macgene	CM15019
TRIZOL Reagent	Ambion	15596018
Critical commercial assays
pEASY®-Uni Seamless Cloning and Assembly Kit	TransGen Biotech	CU101-01
Deposited data
Raw and analyzed data	This paper	PRJCA010267
PRS structure	This paper	PDB:5XIP
Experimental models: Cell lines
HFF cell	ATCC	SCRC-1041
Oligonucleotides
Primers for this paper, see Table S2	This paper	N/A
Recombinant DNA
Plasmid: CRISPR/Cas9	Shen et al.45	N/A
Software and algorithms
GATK v4	Clevenger et al.46	https://github.com/broadinstitute/gatk
BWA	Clevenger et al.46	https://github.com/lh3/bwa
PyMOL	Azuara et al.47	https://pymol.org/2/
Other
Reference genome data (E. tenella Houghton)	This paper	https://www.ncbi.nlm.nih.gov/genome/28?genome_assembly_id=1585209
The sequence of EtcPRSMut	This paper	Data S1

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Xianyong Liu (liuxianyong@cau.edu.cn).

Materials availability

This study did not generate new unique reagents.

Experimental model and subject details

Animals and parasites

One- to six-week-old Arbor Acres broilers, used for proliferation and drug resistance line selection, were purchased from Beijing Arbor Acres Poultry Breeding (Beijing, China). All birds were given anticoccidial drug-free diet and water ad libitum. The wild type E. tenella Xinjiang strain (XJ strain) used was maintained in the laboratory.⁴⁸ XJ strain was sensitive to the anticoccidial drug halofuginone. The procedures for oocyst collection, sporulation and purification were carried out as in the previous report. The cervical dislocation was performed for chickens necessary for sacrifice, which aims to lose consciousness of chickens rapidly. All chickens in this experiment were performed in accordance with the China Agricultural University Institutional Animal Welfare and Animal Experimental Ethical Inspection.

Method details

Selection and characterization of halofuginone-resistant strains

To induce halofuginone-resistant strains rapidly, 300 one-day-old Arbor Acres broilers were equally divided into two groups, an infection group (Group 1) and a treatment group (Group 2). Every bird in Group 1 was inoculated with 500 fresh oocysts of the XJ strain on the first day. Birds in Group 2 were fed with only 3 mg/kg halofuginone in feed during the entire experiment. From Day 7, litter containing feces with shed oocysts was randomly sampled in five locations from Group 1 and dispersed to the litter of Group 2. The collection and dispersion were performed every 7 days from Day 7 to Day 49 after inoculation. The oocyst outputs per gram of the two groups were measured daily. Details for the successive selection are shown in Figure 1A. After the first experiment, the oocysts output in group 2 was randomly collected and performed drug resistance tests to evaluate whether the oocysts acquired resistant phenotype. To acquire adequate halofuginone-resistant strains, the litter of Group 2 in the first experiment was kept and applied to infect another 150 one-day-old Arbor Acres broilers in a second trial. During the second trial, the chickens were consistently fed 30 mg/kg (10-fold higher than that normally added in feed) halofuginone, and chickens were infected by the oocysts in the litter. To acquire more resistant strains, the caeca of 20 chickens were removed for the collection of oocysts every 7 days. Subsequently, propagation of the progeny parasites derived from the second trial was performed in chickens fed 30 mg/kg halofuginone. This propagation procedure was successively carried out for the next 6 generations. Then, halofuginone-resistant parasites from the last propagation further proliferated for 3 successive generations in birds without halofuginone addition. These resistant parasites were tested for their halofuginone-resistant phenotype.

Comparative studies of endogenous development and reproductivity between the wild type and halofuginone-resistant strain were performed with modifications to previously described methods.⁴⁹ Sections of caeca of chickens at 24, 36, 48, 60, 72, 84, 96, 108, 120, 132, 144 and 156 h post inoculation were checked after H&E staining. The reproductivity of wild type and drug-resistant strains was tested by measuring oocyst output. For this experiment, 4 groups of 16-day-old chickens were infected with 10⁴ oocysts/bird for each strain and were fed with 30 mg/kg halofuginone or without halofuginone as a control. To count the number of parasitophorous vacuoles in the cecal sections, two cecal intestines were sectioned separately, and vacuoles were counted in 5 different locations in each section under microscope.

Collection of second-generation merozoites

To collect a mass of second-generation merozoites, twenty groups of 4-week-old Arbor Acres broilers (two broilers per group, including 14 halofuginone-resistant groups and 10 wild type groups) were dosed with 2.5×10⁵ fresh oocysts/chicken. At 120 h after infection, all chickens were sacrificed and the ceca were harvested. The second-generation merozoites of different strains were extracted and purified. Briefly, the ceca were dissected longitudinally with scissors and the intestinal contents of the cecum were rinsed with cold PBS (pH 7.4). Then the ceca were chopped up and digested with 0.25% trypsin and 0.5% sodium taurodeoxycholate hydrate (digestive fluid was prewarmed to 37°C) for 45 minutes at 37 °C. Finally, merozoites were purified using a 300-mesh sieve for filtration. After centrifugation, the purified merozoites were precipitated and stored at −80°C.⁵⁰

Preparation of samples for whole genome sequencing

For whole genome sequencing, drug-resistant strains (two biological replicates), wild type strains (ten biological replicates) were isolated and genomic DNA (∼2 μg) was extracted by using the cetyltrimethylammonium bromide (CTAB) method as previously reported.⁵¹ Then genomic DNA was randomly sheared into ∼300 bp fragments and all libraries were sequenced to an average coverage of 100× using Hiseq-PE150 Sequencer.

Genotype calling

Raw sequence files were filtered and trimmed, and the trimmed reads from fastq files were mapped to the whole-genome sequencing data against the E. tenella Houghton reference genome (pEimTen1.1) using BWA mem (http://bio-bwa.sourceforge.net/) under the default parameters. All 12 samples were excluded SNP by Genome Analysis Toolkit GATK v4 (https://gatk.broadinstitute.org) with default parameters. The raw SNPs were filtered using SelectVariants and VariantFilteration module, with variants required to pass the following criteria ‘QD < 2.0 || MQ < 40.0 || FS > 60.0 || SOR >3.0 || MQRankSum < -12.5 || ReadPosRankSum < -8.0’.⁵²

The filtered variants in VCF format were annotated by SnpEff v5.0 (https://pcingola.github.io/SnpEff/) with E. tenella Houghton (pEimTen1.1) as the reference.

Allele frequency and differentiation analysis

Heterozygosity (Hp) and population differentiation statistics (F_ST) values between populations were estimated using in-house python scripts with a window size of 50-kb and 20-kb step size. The matrices of pairwise distances were then plotted using a custom R script.

Construction of vectors containing EtcPRS^Mut used for transfecting E. tenella and T. gondii

To verify the relevance of T618A to halofuginone resistance, we generated a transgenic E. tenella strain harboring EtcPRS^Mut. The transfected plasmid consisted of two cassettes, and the EYFP sequence was fused with DHFR-Ts2m3m (pyrimethamine-resistance cassette), under the EtMIC2 promoter. The plasmid also harbors the candidate gene cassette, and the EtcPRS^Mut gene with a C-terminal Flag epitope was guided for the EtActin promoter. The point mutation in EtcPRS was verified by PCR amplification and Sanger sequencing, the sequence was listed (see Data S1).

The procedures for the establishment of the transgenic T. gondii were carried out as previously reported.⁵³ The Δku80 RH line was overexpressed with an EtcPRS^Mut under the tubulin promoter at the uracil phosphoribosyl transferase (UPRT) locus. The pTub1- EtcPRS^Mut -3HA plasmid was cotransfected with the CRISPR/Cas9 plasmid into the Δku80 RH strain, and 5 μM 5-fluorodeoxyuridine (FUDR) was added to select positive parasites.

To study the efficiency of the selectable marker, we generated three plasmids. To disrupt H2AX (TGME49_261580) in the Δku80 RH strain, we cotransfected the sgRNA and CRISPR plasmid along with an amplicon containing H2AX homology regions and pTub1-EtcPRS^Mut-3Flag. To integrate the EtcPRS^Mut fusion protein to the 3′ end of target genes, we transfected the Δku80 RH parasites with the sgRNA CRISPR plasmid to target integration to the 3′UTRs of target genes (H2AX and SDHB, TGGT1_215280), along with the amplicon (5HR-linker-3Flag-3UTR-pTub1- EtcPRS^Mut-3UTR-3HR) separately. These three transgenic parasites are cultured under different concentrations of halofuginone (5, 8, and 10 nM) for three passages. In parallel, single-cell clones were obtained by limiting dilutiont.⁵⁴ Single clones were screened by IFA and further analyzed by PCR.

The PCRs were performed in a total volume of 50 μL, consisting of 10 μL High GC Buffer, 10 μL Q5 Reaction Buffer, 0.5 μL Q5 DNA Polymerase, 1.5 μL Deoxynucleotide (dNTP) Solution Mix, 2 μL (10 pmol/μL) reverse and forward primers, 1 μL template DNA and 23 μL double-distilled water (ddH₂O) under the reaction conditions of initial denaturation at 98 °C for 1 min, 35 cycles for the following three steps, denaturation at 98 °C for 10 sec, annealing at 60 °C for 20 sec, and extension at 72 °C for 2 min, with a final extension at 72 °C for 5 min. The primers were designed with Primer Premier 5, and all primers are provided (Table S2).

All fragments and the pEasy-Blunt-Simple cloning vector in this plasmid were linked by a seamless assembly strategy (pEASY®-Uni Seamless Cloning and Assembly Kit).

Transfection and establishment of EtcPRS^Mut-overexpression E. tenella

The sporozoites of E. tenella were extracted from oocysts by Percoll centrifugation and digested by trypsin-bile. The transfection procedures were mediated by SnaBI restriction enzyme as in the previous report.^55,56,57 The 20 μg linearized overexpression plasmid was used for transfection with 1×10⁷ sporozoites. Then the transfected sporozoites were inoculated into five 2-week-old chickens through the cloaca. Halofuginone (30 mg/kg) was fed to chickens during the whole experiment.

To enrich the positive transgenic oocysts, the positive sporocysts from the first generation were sorted by flow cytometry and then inoculated positive sporocysts with chickens under halofuginone (30 mg/kg) selection.

Indirect immunofluorescence assay and immunoblotting

Indirect immunofluorescence assay (IFA) and immunoblotting were performed to verify the expression of EtcPRS^Mut, the methods were carried out as the following protocol. The transgenic sporozoites of E. tenella were extracted and purified through a cellulose filter, then infected with MDBK cells for 4 h. Intracellular sporozoites were fixated with 4% paraformaldehyde (PFA) and permeabilized with 0.25% Triton X-100, then stained with Mouse Flag monoclonal antigen (1:200), Cy3-conjugated goat anti-mouse IgG (1:200) and Hoechst 33258(1:100). The monolayers were observed by the Leica confocal microscope (Leica, YCS SP52, Germany) with 63× magnification, and high-content imaging and analyses were performed with the LAS AF lite 2.2.0 software.

For western blot, proteins of EtcPRS^Mut-overexpressing or wild type strains were extracted from ∼3×10⁶ second-generation merozoites. Western blot analysis was performed as a standard procedure. Anti-Flag mAb (1:1000) and horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:2000) were used to detect Flag-tagged EtcPRS^Mut protein. The Actin of E. tenella was used as a control. Chemiluminescence detection of bands was carried out by using Tanon 1600 (Tanon, Shanghai, China).

The IFA and western blot were also carried out in T. gondii. Mouse HA monoclonal antigen (1:500), anti-Flag mAb (1:1000), rabbit GAP45 monoclonal antigen (1:5000) (this antibody was a gift from Professor Dominique Soldati, University of Geneva), Cy3-conjugated goat anti-mouse IgG (1:200), FITC-conjugated goat anti-rabbit IgG (1:200) and Hoechst 33258(1:100) were used for IFA. Mouse HA monoclonal antigen (1:1000), horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:2000) were used for western blot. The Actin of T. gondii was used as an internal control.

Growth assay of T. gondii

For the growth assay, 10⁵ Δku80 RH or EtcPRS^Mut -3HA parasites per well of a 12-well plate were inoculated on HFF cells grown on coverslips for 1h in normal media treated with vehicle (DMSO) or media treated with 5 nM of halofuginone, then washed the extracellular parasites. After 24 h post-infection, the coverslips were fixated with 4% paraformaldehyde and stained with Hoechst33258 and TgGAP45, and the number of parasites per vacuole was counted. A total of 100 vacuoles were counted for each replicate. Three independent experiments were performed.

For the plague assay, 75 parasites of Δku80 RH and EtcPRS^Mut -3HA parasites were inoculated on HFF cells grown in a 12-well plate for 9 days in normal media or media treated with 0 nM, 1 nM, 2 nM, 5 nM, 8 nM and 10 nM of halofuginone. Parasites were fixated and then stained by Crystal Violet. Lysis plaque areas were quantified using Adobe Photoshop software (version 2022). 50-lysis plaque was quantified for each strain in each replicate. Quantification represents the mean (±SD) from three independent experiments. Statistical significance was assessed by t-test on GraphPad Prism 8 software.

cPRS phylogeny

EtcPRS homologs were identified by BLAST searches against all sequenced species. Homologous putative protein sequences were aligned using the ClustalW program of MEGA11 software. The phylogenetic tree based on the cPRS coding protein was constructed with the Neighbor-Joining method and the resulting tree in Newick format was visualized and processed using MEGA11.

Molecular docking

All structural superposition and preparation of figures were done using UCSF Chimera and PyMOL (http://www.pymol.org). Homology modeling visualization of the E. tenella PRS mutated residues was performed using the wild type E. tenella EtcPRS crystal structure (PDB: 5XIP) as a structural model basis and then anticipated in AlphaFold2. The method was described as previous report.⁵⁸

Quantification and statistical analysis

GraphPad Prism 8.0 was performed to generate graphs and analyze statistical data. All data were analyzed with the two-tailed Student’s t-test. p < 0.05 was considered to represent statistical significance.

Published: March 4, 2023

Supplemental information

Table S2. Primers used for this study, related to STAR methods

Data and code availability

The WGS data for these isolates is available in the NGDC SRA database, under accession number PRJCA010267 (https://ngdc.cncb.ac.cn/bioproject/browse/insdc/PRJCA010267).

This study did not generate any custom code.

Any additional information required to reanalyze the data reported in this article is available from the lead contact on request.