Background

Chicken coccidiosis, caused by apicomplexan parasites of the genus Eimeria, is a widespread and highly pathogenic avian disease posing a significant threat to the poultry industry worldwide [1, 2]. There are seven species of Eimeria responsible for morbidity and socioeconomic burdens including Eimeria necatrix, E. tenella, E. maxima, E. praecox, E. mitis, E. brunetti and E. acervulina [3]. Besides, three cryptic Eimeria operational taxonomic units (OTUs) were recently identified as endemic to Australian chicken populations and later given the names Eimeria lata, E. nagambie and E. zaria based on their genotypic and phenotypic properties [4,5,6]. Infections with Eimeria spp. can damage the digestive tract of chickens and lead to malabsorption of nutrients and lethal diarrhea, which is reflected in clinical symptoms: poor weight gain, low egg production and even death [1, 3]. There are many reports of coccidiosis prevalence in countries worldwide, with 70–92% prevalence in Romania (92%), Greece (86%), Brazil (91%), India (81%), Algeria (72%), Turkey (70%) and China (88%), and multi-Eimeria species infections predominate [7,8,9,10,11,12,13,14]. The annual economic loss for chicken farmers has been estimated at over £10.4 billion worldwide [15].

Because of similar oocyst morphotypes and overlapping biological features of chicken Eimeria spp. [16, 17], species identification based on morphological characteristics is difficult. Therefore, there is an urgent need to develop an efficient approach to identify Eimeria infections for clinical diagnosis; moreover, phylogenetic relationships and biogeographic range evolution of diverse strains can also provide valuable information to support diagnosis and control of chicken coccidiosis. Therefore, it might be possible to achieve this goal by utilizing next-generation sequencing (NGS) approach-based species identification, phylogenetic constructions and population genetic analyses. Encouragingly, Vermeulen and colleagues have established a ribosomal protein 18 (18S)-based NGS methodology to assess Eimeria communities of the Australian marsupial [18]. Hinsu et al. have also reported the application of Illumina MiSeq deep sequencing to 18S amplicons of chicken Eimeria and achieved detection of all seven validated species including their cryptic genotypes [19]. In addition, increased genetic evidence showed that the taxonomy and diversity of certain taxa or specific groups of chicken Eimeria can be equally revealed by using the mitochondrial (mt) gene-based NGS owing to matrilineal inheritance, absent recombination and rapid evolution rate of these molecules [20,21,22]. For example, Hauck et al. showed that using NGS to sequence mt cytochrome c oxidase subunit 1 (cox1) was efficient for identification of Eimeria species that circulated in chicken flocks [13]. Similarly, Snyder et al. also established the mt cytochrome c oxidase subunit 3 (cox3) amplicon library-based NGS pipeline analysis to differentiate E. acervulina, E. tenella, E. maxima, E. mitis, E. necatrix and E. praecox from mixed infected samples [23]. Nevertheless, it appears noteworthy that the single gene/locus-based NGS approach fails to provide sufficient genetic information to understand variations at the species and genus levels. Consequently, using NGS-based mitogenomic data becomes an ideal candidate strategy because it can not only provide comprehensive molecular insights into intra-/interspecific variability but also yields a broad taxonomy-range phylogenetic and evolutionary justification for Eimeria parasites [24, 25].

Herein, we sequenced complete mitogenomes of 33 Eimeria isolates infecting domestic chickens in China and genetically characterized these isolates by comparisons with other chicken eimerian parasites for which mitogenomic datasets are available in public databases. Furthermore, genome- and gene-level phylogenetic analyses were achieved to better understand the evolutionary relationships of chicken Eimeria at the species level. This comprehensive molecular information will provide valuable resources for systematic, population genetic and evolutionary biological studies of Eimeria parasites in poultry.

Methods

Litter sampling, parasite analysis and DNA extraction

From 2019 to 2021, 153 commercial farms that were in the finishing phase of the productive cycle were selected and sampled as part of ongoing surveys of coccidia on commercial broiler farms in China: 32 in Sichuan, 30 in Henan, 25 in Anhui, 23 in Chongqing, 18 in Guangdong, 15 in Fujian and 10 in Zhejiang. Litter samples were randomly collected from each shed while walking in a “zigzag” pattern [26]. About 250-g litter from each sampling was bagged and transported to the laboratory where it was kept at 4 °C. Oocysts per gram (OPG) of each litter was quantitated using a McMaster chamber. For species identification, positive litter samples were mashed and suspended with running water, followed by filtration through a series of sieves (212, 180, 75 and 45 mm; Endecotts, London, UK) and culture in a 5% (w/v) potassium dichromate solution for 72 h at 28 °C under forced aeration. Oocysts were sporulated and microscopically speciated using morphometric keys [27, 28]. Furthermore, pure sporulated oocysts from 33 isolates that represent all known Eimeria species infecting chickens were enriched with saturated sodium nitrate flotation technique. One hundred sporulated oocysts from each isolate were used for genomic DNA extraction. After vortex breakdown, the total genomic DNA was extracted from oocysts using the Genomic DNA Kit (TIANGEN, Beijing, China). Following the PCR amplification using Eimeria-specific primers ERIB1 (forward: 5′-ACCTGGTTGATCCTGCCAG-3′) and ERIB10 (reverse: 5′-CTTCCGCAGGTTCACCTACGG-3′) [29] and comparing with the targeted 18S regions, those isolates were determined to share > 99.5% identities with query sequences in GenBank nos. KT184333 (E. acervulina), KT184337 (E. brunetti), KT184346 (E. maxima), KT184349 (E. necatrix), KT184351 (E. praecox), KT184354 (E. tenella) and FR775307 (E. mitis).

Genome sequencing, assembly and annotation

After species identification, approximately 2 million sporulated oocysts from each of the 33 isolates were used for mitogenome sequencing. Following quality and quantity assessment, ~ 3 µg high-quality DNA from each sample was fragmented to construct 350-bp paired-end (PE) libraries and sequenced on an Illumina HiSeq X-TEN platform (BerryGenomics, Beijing, China). The clean reads (~ 1.8 Gb) were assembled with MITObim [30] using the available mitogenomes of other Eimeria species from GenBank (GenBank nos. JN864949, KX094949, KX094954, HQ702479, HQ702480, HQ702481, HQ702483 and KC409031) as the references. These assembled mitogenomes were also validated by PCR amplifications using four overlapping fragments. These four overlapping fragments were located between cytochrome b (cytb) and cox1 (~ 2.0 kb), between cox1 and large subunit A (LSUA) (~ 3.0 kb), between LSUA and large subunit E (LSUE) (~ 2.0 kb) and between LSUE and cytb (~ 1.0 kb), respectively (Fig. 1), and their corresponding PCR primers were designed based on the aforementioned mitogenome sequences of Eimeria species and are shown in Additional file 1: Table S1. PCR reactions were achieved using a 50-μl reaction volume containing 3 μl genomic DNA (≥ 10 ng), 25 μl 2 × HiFi TransTaq PCR SuperMix (TaKaRa, Biotech, Dalian, China), 1 µl sense primer (10 pmol), 1 µl anti-sense primer (10 pmol) and 20 µl nuclease-free water. The PCR conditions were: 95 °C for 5 min denaturation, 95 °C for 45 s for 30 cycles and 45 ~ 55 °C for 45 s, and 68 °C for 1 ~ 3 min according to the Tm values and the product lengths, and a final extension at 68 °C for 10 min. After agarose gel detections, all target amplicons were column-purified and sequenced either directly or following sub-cloning into the pMD19-T vector (TaKaRa, Biotech, Dalian, China). Each amplicon was sequenced three times to ensure maximum accuracy. Mitogenome annotation was performed with MITOS [31] and manual inspection based on a whole genome-guided alignment using poultry Eimeria spp. for which complete mitogenome sequences were available so far.

Fig. 1
figure 1

Structure views of mitogenomes of 33 chicken Eimeria isolates identified in this study. a–g Mitochondrial genome organizations of the Eimeria isolates 1–33 corresponding to seven chicken-infecting Eimeria species. Isolates in the same species are arranged together and contrast with the reported mitogenome organization. Red color indicates protein-coding genes, blue color indicates fragments of rRNA genes, green and purple colors indicate positive and negative GC-skew, respectively, and black color indicates the GC content of every position in the genome. h The average AT-skew and GC-skew values of the complete mitogenomes (Entire), concatenated PCG datasets (PCGs), LSU and SSU rRNAs of each species

Full size image

Sequence characterization

Nucleotide composition and codon usage of 33 eimerian mitogenomes were measured using Geneious and PhyloSuite [32, 33]. The nucleotide skewness of mitogenomes was performed using the formulas [34]: AT skew = (A − T)/(A + T) and GC skew = (G − C)/(G + C). Nucleotide and amino acid divergences of protein-coding genes (PCGs) were determined using DNAstar [35]. The multi-alignment of PCGs obtained from 33 sequenced eimerian isolates and available chicken eimerian parasites including newly recognized E. lata, E. nagambie and E. zaria were separately aligned using MEGA software [36]. Based on those alignments, the sliding window analysis was used to compute the nucleotide diversity Pi (π) using a 200-bp window and 20-bp steps, followed by assessment of genetic structure using the Wright’s fixation index (FST) with 1000 replicates as a permutation test in DnaSP ver. 5.10 [37]. In addition, the evolutionary rate and the ratio of the nonsynonymous substitution (Ka) and synonymous substitution (Ks) of each PCG were calculated using KaKs_Calculator [38]. Genetic distances among these 10 Eimeria species mitogenomes were calculated based on Kimura-2-parameter (K2P) [39] with MEGA X.

Phylogenetic characterization

The phylogenies were reconstructed based on the mitogenomic datasets of 33 Eimeria isolates and other related species (Additional file 2: Table S2). Nucleotide sequence alignments were yielded from the complete mitogenomes, concatenated datasets of dual/triple PCGs or individual PCG of coccidian parasites using MAFFT ver. 7.271 [40], followed by filtration of the ambiguous regions with GBLOCKS ver. 0.91b [41]. Three algorithms including the maximum parsimony (MP), maximum-likelihood (ML) and Bayesian inference (BI) were used to reconstruct the phylogenetic relationships of chicken Eimeria species, and Isospora sp. (GenBank no. KP658103) was treated as the outgroup and included in each analysis. For the MP analysis, the complete nucleotide sequence dataset of mitogenomes or multi-alignments of the nucleotide sequence for the single, dual or triple PCGs were analyzed using equally weighted parsimony and heuristic searches with a tree-bisection-reconnection (TBR) branch-swapping in PAUP* [42]. One thousand replicates of Wagner trees (using random addition sequences) were chosen, and five trees of each replication were saved, followed by obtaining the optimal topology with the Kishino-Hasegawa method. Bootstrap resampling with 1000 replications was computed for each nodal support. The ML analysis was implemented with PHYML ver. 3.0.1 [43] using the optimal evolutionary models “TIM + F + I + G4” for the complete mitogenomic dataset and “GTR + F + I + G4” for the dual and triple concatenated PCGs and individual PCG datasets that were selected with the “Auto” option on W-IQ-TREE web server (http://iqtree.cibiv.univie.ac.at) according to Bayesian information criterion (BIC). ML trees were reconstructed with a 10,000-replicate and an ultra-fast bootstrap approximation. Within the BI trees, the optimal evolutionary model “CAT + GTR + G” was selected for all mtDNA datasets using ModelFinder [44]. The BI analyses were carried out using MrBayes ver. 3.2.7 [45], with four independent Markov chains, running for 1,000,000 (complete mitogenomic dataset) and 100,000 (dual and triple concatenated PCGs and individual PCG datasets) metropolises coupled with Monte Carlo generations, sampling a tree every 0.1% generations. The first 25% of the trees were eliminated as “burn-in” when the average standard deviation (SD) of the split frequencies decreased to < 0.01, and the remaining trees were used to calculate Bayesian posterior probabilities (PPs). The evolutionary distance was estimated using the MrBayes order (aamodelpr = mixed) with default parameters. A consensus tree was obtained and visualized using TreeviewX (https://www.linuxlinks.com/treeviewx/).

Results and discussion

General features of 33 Eimeria mitogenomes

The complete mitogenomes of 33 eimerian isolates representing all seven species known to infect chickens in China varied from 6148 to 6480 bp in size (GenBank accession nos. OP800493–OP800525; Table 1). Each genome encoded three mt PCGs including cox1, cox3 and cytb as well as seven fragmented small subunit (SSU) and 12 fragmented LSU rRNAs, except for E. mitis isolates, which contained another 56-bp LSU rRNA (LSU15) located between LSU13 and LSUD (Fig. 1a–g). The overall base composition of each mitogenome consistently exhibited high AT bias (64.52–67.37%) with T being the most favored base and G the least favored, similar to those observed in other Eimeria members [46,47,48,49,50].

Table 1 List of Chinese Eimeria isolates sequenced in the present study
Full size table

Within SSU rRNAs, seven gene fragments including SSUA, SSUF, SSUD, SSU9, SSU8, SSUB and SSUE ranged from 37 (SSUF) to 116 bp (SSUB) in size. For LSU rRNAs, 12 gene fragments including LSUF, LUSG, LUSC, LSU10, LSU13, LSUD, LSU2, LSUA, LSU1, LUSB, LSU3 and LSUE ranged from 16 (LUSC) to 188 bp (LSUE) in size, very similar to those of other Eimeria species reported to date [25, 46,47,48,49,50]. Furthermore, AT-skew and GC-skew values of rRNAs are shown in Fig. 1h, and it was clear that these mitogenomes shared a same skewness pattern: a slightly higher A than T (except for LSU rRNAs) in E. necatrix (− 0.003), E. maxima (− 0.021) and E. brunetti (− 0.008) and a slightly lower G than C (except for SSU rRNAs) in E. necatrix (− 0.021) and E. tenella (− 0.029).

For three PCGs, cox1, cox3 and cytb located between cytb and LSUF, LSUA and LSU1 and SSUA and cox1, respectively, with an obvious bias towards AT (Fig. 1). Such AT bias was also reflected in their RSCU and codon usage patterns of PCGs (Fig. 2); for instance, within the initiation codon choice, the cox1 gene was inferred to start with codons TTG, ATN (ATT and ATG) and GTN (GTG and GTT), the cox3 gene was inferred to start with codons ATT and TTA, and the cytb gene was inferred to start with codon ATG (Fig. 2h). Correspondingly, the standard stop codon TAG was used to terminate the cox1 and cox3 genes and the cytb gene used TAA as the stop codon, consistent with the previous study [49]. Moreover, RSCU comparisons showed that these 33 eimerian isolates consistently used TAA as the stop codon, except for E. acervulina isolates, which used TGA and TAA as stop codons (Fig. 2), inconsistent with mitogenomes of some other Apicomplexa parasites in which an abbreviated stop codon (TA or T) was used as the stop codon [50]. It was also evident that the most frequently used codon of these 33 eimerian isolates was AGA (RSCU ranging from 4.08–4.86), followed by UUA (2.67–3.65), GGU (2.44–2.62) and CCA (1.95–2.67). These codon patterns were similarly reflected in the amino acid usage frequency of PCGs, and the most frequently used amino acid was Leu (count ranging from 166–171), followed by Ser (102–112), Phe (98–104) and Ile (93–108) (Fig. 2a-g).

Fig. 2
figure 2

Codon usage in PCGs of mitogenomes of 33 chicken Eimeria isolates studied here at the species level. a–g RSCU and the number of codons of PCG genes used in each species based on their corresponding isolates. Codon families are plotted under the X axis and represented by different color bars, and the codon numbers are indicated by the black line graph (right axis scale). h Start and stop codon usages of three PCG genes in all isolate mitogenomes

Full size image

Mitogenome diversity

Most gene sizes of mitogenomes of these 33 Eimeria isolates were consistent with those of other congeneric species (Additional file 3: Table S3). Nevertheless, to further understand the evolutionary divergence among the genus Eimeria, the intra- and interspecific variability was determined across the 33 isolates using the nucleotide and amino acid sequences. The interspecific divergences were determined to range from 1.5 to 12.8% while intraspecific differences were < 0.3% (Additional file 4: Table S4). Both ranges were lower than those estimated in Australian chicken Eimeria by Morgan and Godwin [25], to some extent supporting a more general systematic congruence among Chinese chicken eimerian parasites sequenced in this study. Meanwhile, it seemed that the lowest interspecific divergence always occurred between E. tenella and E. necatrix, suggesting their closer genetic similarity among Eimeria species. Furthermore, divergences were also focused on various portions of the mitogenomes. For rRNAs, the nucleotide sequence divergences of SSU and LSU rRNAs ranged from 0.2 to 4.6% and 0.4 to 8.2%, respectively; within PCG genes, the nucleotide sequence divergences ranged from 2.1 to 17.6%, and amino acid sequence divergences ranged from 0.8 to 10.1%, with cox3 being the most variable gene, in line with previous findings in species of Plasmodium, Theileria and Babesia [51, 52].

To confirm the aforementioned sequence diversities within and between mt genes, sliding window analysis was also carried out by a combination of the 33 Eimeria isolates and other congeneric species, with a comparison between Chinese and Australian Eimeria populations. As shown in Fig. 3a–g, the intraspecific Pi values of seven Eimeria species, regardless of sources (China or Australia), all were relatively low (0 to 0.069) with a significantly increased trend in the following order: E. acervulina < E. tenella < E. necatrix < E. praecox < E. maxima < E. brunetti < E. mitis. It was significant that various numbers of peaks presented within comparisons of interspecific diversity depending on species pairs, suggesting that there is still a considerable number of alternative genetic loci to be determined as species-specific markers for identification and differentiation of chicken coccidian parasites. Current mt genetic loci used for PCR detection/diagnostics in the genus Eimeria include 805-bp cox1 and 954-bp cox3 [23, 53], and both genetic loci have been recently targeted for development of an amplicon library-based NGS pipeline analysis for diagnosis [13, 23]. From the analysis in this study, however, compared to the cox1 and cox3 genes, it appeared that the cytb gene could also be suitable as a genetic marker for diagnosis and identification between E. maxima/E. tenella and other congeneric species because of its higher variability (Fig. 3c and g). Furthermore, when the sliding window analysis was performed across all chicken Eimeria species, the PCG regions with pronounced peaks and troughs seemed to be more significant than other portions of the mitogenomes (Fig. 3h). Among these PCG regions, cox1 was determined to be the most conserved while cox3 was deemed to be the least conserved, supporting the aforementioned by sequence divergence analysis.

Fig. 3
figure 3

Sliding window analysis of mitogenomes for the nucleotide diversity (Pi) among chicken Eimeria. X-axis: position of the midpoint of a window, Y-axis: nucleotide diversity (Pi) of each window (window size: 200 bp; step size: 20 bp). PCG gene boundaries are indicated above the graph. Colors from gray (high diversity) to white (low diversity) indicate the different nucleotide diversities. a–g Sliding window analysis of each chicken Eimeria species including isolates sequenced in this study. h Sliding window analyses of all chicken Eimeria species including the 33 isolates sequenced in this study. The Pi value of each PCG is shown near the gene name

Full size image

In parallel with the sliding window analysis, the FST values obtained from PCGs of these Eimeria spp. according to their geographical sources (China vs. Australia) are shown in Additional file 5: Table S5. FST of 0–0.18 suggested species-specific variation in levels of interbreeding, with a higher level of genetic isolation between Chinese and Australian E. mitis populations. In addition, the ratio of non-synonymous (Ka) and synonymous (Ks) substitutions for each PCG in 33 Eimeria isolates and other congeneric species is shown in Fig. 4a–g. Notably, the Ka, Ks and Ka/Ks tied well with results of sliding window analysis, and the PCGs of the Eimeria species with high Pi values trended to have positive selection sites (Ka/Ks > 1). For example, Ka/Ks ratios of the cox1 gene among E. brunetti and E. necatrix isolates and the cox1 and cox3 genes among E. maxima isolates all were > 1. In contrast, the Ka/Ks ratios of the cytb gene in all isolates were observed nearby zero, suggesting its strong purifying selection. Moreover, when the Ka/Ks was calculated across all Eimeria species, all PCG genes exhibited low ratios (< 1) [54], suggesting that these genes were evolving under negative or purifying selection and, to certain extent, implied the conservation and validity of seven Eimeria species infecting chickens.

Fig. 4
figure 4

Evolutionary rates of chicken Eimeria including the 33 isolates identified in this study. Rate of non-synonymous substitutions (Ka), rate of synonymous substitutions (Ks) and ratio of rate of non-synonymous substitutions to rate of synonymous substitutions (Ka/Ks) are calculated for each PCG. a–g Average Ka, Ks and Ka/Ks ratios of three PCGs of isolates reported and sequenced in this study in each species. h Average Ka, Ks and Ka/Ks ratios of 66 Eimeria isolates including the 33 isolates sequenced in this study

Full size image

Genetic distances of chicken Eimeria

We calculated the intraspecific and interspecific genetic distances between these 33 isolates and each of ten chicken eimerian species using the concatenated PCGs. As shown in Fig. 5, the K2P model-based genetic distances between Eimeria isolates 1–7 and E. tenella, Eimeria isolates 8–13 and E. necatrix, Eimeria isolates 14–19 and E. maxima, Eimeria isolates 20–23 and E. praecox, Eimeria isolates 24–26 and E. mitis, Eimeria isolates 27–30 and E. brunetti and Eimeria isolates 31–33 and E. acervulina approached zero and suggested their species identity, in agreement with the morphological and molecular identifications. For all species clusters, the intraspecific variations were lower than those calculated on the basis of the 18S gene (0.000–0.002 vs. 0.002–0.013) [5]. Moreover, the interspecific genetic distance between E. tenella and E. necatrix was lower than each compared to the other five Eimeria species, once again suggesting that these two species had the closest relationship among chicken eimerian parasites [5, 6, 25, 46,47,48,49,50, 53]. By contrast, the genetic distance between E. maxima and other congeneric species appeared the farthest (0.138 to 0.176), consistent with the previous finding [53], supporting a distinct relationship between E. maxima and other chicken Eimeria species.

Fig. 5
figure 5

Patterns of K2P distance between 33 isolates and 10 known chicken Eimeria species. The edge of these decagons indicates the K2P distance between the isolate and its corresponding species (0.000). Gray lines indicate the same K2P distance from the center of these decagons. Different color dots show the relative K2P distances between isolates with ten Eimeria species. The abbreviations Eac, Ebr, Ela, Ema, Emi, Ena, Ene, Epr, Ete and Eza represent Eimeria acervulina, E. brunetti, E. lata, E. maxima, E. mitis, E. nagambie, E. necatrix, E. praecox, E. tenella and E. zaria, respectively

Full size image

Phylogenies of chicken Eimeria

The available mitogenomes of 33 isolates representing all seven known chicken eimerian species in China together with an additional 25 mitogenomes of Australian Eimeria spp. provided us an opportunity to study the evolutionary relationships of each isolate in each species and of each species in the genus Eimeria. As shown in Fig. 6a and b, it was clear that three identical trees (MP/ML/BI) inferred from either the complete mitogenomes or concatenated PCG datasets consistently revealed that the Eimeria isolates 1–7 were grouped with available species of E. tenella, the Eimeria isolates 8–13 were grouped with E. necatrix, the Eimeria isolates 14–19 were grouped with E. maxima, the Eimeria isolates 20–23 were grouped with E. praecox, Eimeria isolates 24–26 were grouped with E. mitis, Eimeria isolates 27–30 were grouped with E. brunetti isolates, and Eimeria isolates 31–33 were grouped with E. acervulina, with high statistical supports (all values ≥ 98 or = 1.00). Furthermore, both phylogenetic trees split chicken eimerian parasites into two clades: one included E. tenella and E. necatrix and another was composed of E. lata, E. maxima, E. nagambie, E. brunetti, E. zaria, E. mitis, E. praecox and E. acervulina. Compared to the stable topology of E. tenella and E. necatrix, the phylogenetic relationships among eight other Eimeria species varied by different datasets used here (Fig. 6a and b). Notably, E. maxima and E. mitis were more closely related to each other than to E. praecox and E. acervulina, in contrast with findings based on the nuclear 18S and internal transcribed spacer (ITS) and mitochondrial cox1 datasets [53, 55,56,57]. Nevertheless, these analyses provided a consistent, robust phylogenetic resolution for the 33 isolates and their congeneric species in the genus Eimeria: a paraphyletic relationship was shared among the 10 chicken Eimeria species, in agreement with our genetic distance study herein and the results of morphological and molecular biology studies [6, 56, 58,59,60].

Fig. 6
figure 6

Phylogenetic relationships of chicken Eimeria including the 33 isolates identified in this study. a Phylogenies were inferred on the basis of the complete mitogenome datasets. b Phylogenies were inferred on the basis of the concatenated nucleotide sequences of three PCGs. c Phylogenies were inferred on the basis of the concatenated nucleotide sequences of two PCGs. d Phylogenies were inferred on the basis of the nucleotide sequences of single PCG. 66 chicken Eimeria and one Isospora speecies (outgroup) were included in the aforementioned phylogenetic analyses using the BI, ML and MP methods. Ten chicken Eimeria species, E. acervulina, E. brunetti, E. lata, E. maxima, E. mitis, E. nagambie, E. necatrix, E. praecox, E. tenella and E. zaria, and their branches are shown in dark blue, blue, pink, red, green, cyan, yellow, orange, purple and flavogreen, respectively; 33 Eimeria isolates sequenced in this study are indicated in bold font (a and b) or branches (c and d). The numbers along the branches indicate bootstrap values/posterior probabilities resulting from different analyses in the order MP/ML/BI

Full size image

Furthermore, the single and concatenated PCG genes were separately used for phylogenetic analysis to screen out the optimal genetic marker candidates for phylogeny and species identification and differentiation of chicken eimerian parasites from other related species. As shown in Fig. 6c and d, although any single or dual combination of PCGs exhibited various phylogenetic topologies, the positions of the species E. tenella and E. necatrix were steady in these six (cox1, cox3, cytb, cytb + cox1, cytb + cox3 and cox1 + cox3) phylogenetic analyses. Notably, the concatenated cytb and cox1 gene- and single cox1-based phylogenetic analyses shared the same topology as that of the genome-based phylogeny, suggesting that the cox1 gene might be the most appropriate genetic marker and therefore could be used instead of mitogenomes for evolutionary and phylogenetic studies of chicken Eimeria species. Of course, the marker validity of the cox1 gene remains further validated when more additional apicomplexan parasite mitogenomes become available, especially those from avain coccidia, although the cox1 has been widely used as a DNA barcode for species identification and differentiation in Eimeria [53, 61].

Conclusions

In this study, we presented a comprehensive molecular characterization of evolutionary blueprint of chicken coccidia by Illumina sequencing complete mitogenomes of 33 isolates representing all seven known Eimeria species infecting chickens in China. Comparative genomics revealed the low genetic diversity of these Eimeria species and showed three mitochondrial protein genes under purifying selection with cox1 and cox3 genes being the lowest and highest evolutionary rates, respectively. Phylogenies divided these 33 Eimeria isolates into seven subgroups, and each represented one chicken Eimeria species. Furthermore, single and concatenated mitochondrial protein gene-based phylogenies supported the cox1 gene as the genetic marker for evolutionary and phylogenetic studies for avain eimerians. These Eimeria population-level mitogenomic datasets provide an updated understanding of systematic, population genetic and evolutionary biological studies of apicomplexan parasites in poultry and other animals.