Introduction

Przewalski’s horse genome falls outside a domestic horse’s group, approximately 38,000–72,000 year BP [1]. Przewalski horses are considered the last living population of wild horses [2], however, the previous research based on whole genome sequencing indicated that they are feral descendants of the horses domesticated about 5000 years ago by the Botai culture in present-day Kazakhstan [3]. The species’ history remains unknown for nearly five millennia until Przewalski horses were rediscovered as a free-range population deemed to be wild about 150 years ago [4]. In the twentieth century, Przewalski horses population went through several bottlenecks and almost became extinct but survived from a captive herd as a result of enormous conservation and reintroduction efforts. After a drastic decrease in individual numbers, and based on pedigree data, 11 pure species founders are distinguished including three Old Askania-Nova line founders, one New Askania-Nova line founder, three Old Prague line founders, and four Munich (US) line founders [5]. However, the population cannot be genetically pure, as in the early 1900s, several zoos interbred their Przewalski horses with domestic horses to save the species and started to reproduce another two founders: a domestic horse and domestic/Przewalski hybrid [4, 6]. Now, their population consists of no more than 2500 individuals worldwide including 900 horses in Europe and 1360 in Asia [7].

One of the largest breeding centers for Przewalski horses is Askania-Nova Biosphere Reserve (Ukraine). The first 10 Przewalski horses were imported to Askania-Nova from Mongolia in 1899–1903 and by 1945, 37 live foals were born but a decline in interest in horses, inadequate breeding conditions, extremely difficult exchange of producers and, as a result, breeding in strict inbreeding caused the degeneration of the entire population and almost extinction in the 1940s. Purposeful breeding work on this species, which led to its preservation, began after the publication in 1959 of the first Przewalski horse studbook [8]. Nowadays, the population in Askania-Nova is maintained at about 60–70 individuals which are kept under semi-free conditions in large enclosures with natural steppe vegetation as well as in small zoo exhibition paddocks without grass [9].

Previous studies on mitochondrial DNA (mtDNA) identified three unique haplotypes in Przewalski horses, none of which were found in modern horses—two similar ones and one substantially divergent from them [10]. However, recent studies indicate evidence of four Finnhorse and one Latvian horse individuals carrying mitochondrial haplogroup previously confined only to Przewalski horse [11].

The basic coat colors genetic control in horses resides at two genetic loci, namely Extension (E) and Agouti (A) which are responsible for the three basic colors occurrence—black, bay, and chestnut [12]. The most common color seen on Przewalski horses is bay with Dun dilution. It has been shown that Przewalski horses, as well as the Siberian horses of that time, do not have the black pigmentation recessive alleles ‘a’ in ASIP gene, which is possibly related to their living in the Asian steppe’s adaptation [13, 14]. The archaeological samples studies indicated that during domestication the selection pressure had a great impact on coat color variations [13]. It was hypothesized that ancestral horses were characterized by black-based patterns [12], and all Siberian and European Pleistocene horses were bay or bay-dun [13]. The ‘fox’ color carriers caused by MC1R recessive alleles ‘ee’ should not be present in Przewalski horses, however, the previous research showed the ‘fox’ color alleles presence in 15% of the tested Przewalski horses [14]. The results may be the consequence of the population crossbreeding in the early days of the conservation efforts. In TBX3 gene for Dun-related trait; there is only one characteristic genotype observed in all Przewalski horses. Horses with this genotype show a diluted body color and have dark points called primitive markings including dorsal stripe and leg barring [15].

The research purpose was to establish the maternal variation of the Przewalski horses population maintained in Askania-Nova Reserve based on whole mtDNA hypervariable regions profiling, as well as, analysis of Y chromosome SNP unique for Przewalski horses, and coat color markers: MC1R (‘fox’) and TBX3 (Dun).

Materials and methods

The analyses were carried out on hair follicle samples derived from 23 Przewalski horses population. The analysed individuals belonged to three Przewalski horses lines (Prague, Munich, and New Askania-Nova) living in the Askania-Nova reserve (Ukraine). The tested horses (13 stallions and 10 mares) represented all the lines of the Askania-Nova reserve population. The 21 animals were sampled after death and the other two when disturbed according to the classic procedure for zoos including branding animals and assigning them an international breeding number. During the veterinarian procedure, the hair follicle samples were taken from the mane or tail of the horses.

The DNA was isolated using Sherlock AX kit (A&A Biotechnology) according to the manufacturer’s protocol and stored at −20 °C. For the mitochondrial DNA hypervariable regions amplification, three PCR products were designed covering a total region of 1062 bp (X79547.1; Table 1). Amplification was performed on all 23 individuals using the Phanta Ready Mix (Vazyme Biotech). PCR products were cleaned with the enzymatic method using the EPPiC Fast reagent (A&A Biotechnology) and used as a template for sequencing by the Sanger method. The PCR for the sequencing reaction was performed for each amplicon (69 samples in total) with the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific) and the products were repurified with the BigDye XTerminator Purification Kit (Thermo Fisher Scientific). The capillary electrophoresis was performed on 3500xL Genetic Analyzer (Thermo Fisher Scientific) using POP-7 ™ Polymer for 3500/3500xL (Thermo Fisher Scientific) and the results were analysed using FinchTV (Geospiza, Inc.), BLAST and Variant Analysis (Thermo Fisher Cloud) as well as compared with the GenBank reference sequences belonging to the species: Equus caballus (X79547.1), Equus asinus (MK896302.1), Equus asinus somalicus (MG885769.1), Haringtonhippus francisci (KT168329.2), Equus grevyi (NC020432.2), Equus kiang (NC020433.1), Equus hemionus (NC018782.1), and Equus przewalskii (KT221845.1).

To analyse Y chromosome SNP, first, the amplification of X chromosome part was performed for all samples using fluorescently labeled probe unique complementary to the X chromosome (PLP1 gene, ENSECAG00000015446) and the TaqMan™ Gene Expression Master Mix kit (Thermo Fisher Scientific). The X chromosome were used as an internal control. Next, all of the male samples were genotyped using the allelic discrimination method with assay complementary to the Y chromosome (designed using Primer Express™) and TaqPath™ ProAmp™ Master Mix (Thermo Fisher Scientific). The SNP localized on the Y chromosome and distinguishing Equus caballus from Equus przewalskii was investigated (MH341179.1, g731821T>C), [16]. Samples were analysed qualitatively, the probes with X and Y signals were recognized as male (Table 1). Real-time PCR reaction was performed on QuantStudio™ 7 Flex (Thermo Fisher Scientific).

Table 1 The primer sequence and detailed information about analysed polymorphisms in Przewalski horses
Full size table

Moreover, two coat color genes were investigated: TBX3 and MC1R. All PCRs were carried out with Phanta Ready Mix (Vazyme Biotech). Two polymorphisms at the TBX3 gene (chr8:18,227,267 + 1066G>T, chr8:18,226,905A>G) were detected by Sanger sequencing and the heterozygosity state of deletion (1.6 kb in/del, chr8:18,227,267) was investigated by the amplification length polymorphism method using method described by Stefaniuk-Szmukier et al. [17]. Melanocortin-1-receptor (MC1R) gene polymorphism responsible for ‘fox’ color (ECA3g.36,259,552C>T) was detected by Sanger sequencing.

Results and discussion

The sequencing of mitochondrial DNA hypervariable region 1 and hypervariable region 2 assigned Przewalski horses to the three distinctly different haplotypes (Fig. 1a). All samples were clustered in three homogeneous groups presenting unique SNPs patterns (Table 2). The obtained sequences have been submitted to GenBank and received accession numbers: ON393914, ON393915, ON393916. The mtDNA haplotypes comparison with other Equidae mtDNA sequences available in GenBank was determined qualitatively using the evolutionary distance values (in Figs. 1b and c) and showed an interesting association. The haplotype 3 is uniquely clustered with Equus caballus reference, while the haplotype 2 is separated from the genus Equus, showing almost the same value of evolutionary distance with Equus przewalskii reference (KT221845.1; 0.1142 and 0.11005 respectively) and is probably low admixtured with other Equidae. The haplotype 1 presented the evolutionary distance value similarity to the Haringtonhippus (KT168329.2; 0.22801 and 0.19402 respectively)—extinct species, that lived in North America during the Pleistocene.

Fig. 1
figure 1

a The neighborhood tree illustrates mtDNA sequence similarities between 23 analysed Przewalski horses and the haplotypes which they are presenting (Mega 11.0.9 software). b The neighborhood tree illustrates mtDNA distances between three Przewalski horse haplotypes and the species selected as potentially having the highest mtDNA sequence similarity to them based on NCBI reference with the substitution parameter (Mega 11.0.9 software). c The neighborhood tree illustrates mtDNA distances between three Przewalski horse haplotypes and the species selected as potentially having the highest mtDNA sequence similarity to them (Mega 11.0.9 software)

Full size image
Table 2 The mitochondrial DNA haplotypes occurrence in different species
Full size table

Haringtonhippus was described by Heintzman et al. [18] as the result of the full mitochondrial and partial nuclear genomes analysis from late Pleistocene ‘New World stilt-legged’ equids endemic to North America. They also demonstrated that Haringtonhippus falls outside of a crown group Equus however, based on our results, it can be suspected that the Haringtonhippus may have interbred with selected equines, including Przewalski horses. The geographical barrier concerning both species occurrence could be overcome through the land existing in the Pleistocene and connecting today’s Siberia in Asia with North America—called Beringia. Therefore, the gene flow between Haringtonhippus and Equus przewalskii populations may have occurred regularly at least periodically during the Pleistocene. Moreover, the latest research confirms a small proportion of North American ancestry in Przewalski horses and observed that in present Przewalski horses some of their genomic origin is derived from relatively recent gene flow from extinct North American horses [19].

According to the Y chromosome analysis, 13 of 23 samples were male and all of them presented C allele in MH341179.1 g731821T>C locus, characteristic for Przewalski horses. The research previously conducted on Przewalski horses [16] also presented C allele in this location and confirmed the g731821T>C substitution characteristics only for Przewalski horses.

The coat color genes analyses indicated only native, wild genotypes. The TBX3 gene polymorphisms were analysed. SNP1 (G in Dun, T in non-dun1) is located within the region deleted in non-dun2, 1067 bp downstream of the deletion breakpoint at chr8:18,227,267. SNP2 (G in Dun, A in non-dun1 and non-dun2), is located 362 bp upstream of the deleted region in non-dun2 at chr8:18,226,905. The results assigned all horses to the genotype with G/G alleles in both SNP loci, and deletions were not observed. MC1R variants were also identified; none of Przewalski horses presented C to T missense mutation associated with an unfavorable ‘fox’ color variant for this species. The ability to produce phaeomelanin is attributed to the recessive alleles ‘ee’ and is not in concordance with the recommendations for the Przewalski horses [12].

Conclusion

The presented genetic analysis of Przewalski horses population living in the Askania-Nova reserve adds useful information for the conservation of the breed. The mtDNA results, according to the literature, confirm three pure and fixed remaining maternal lineages in Przewalski horses, assigning the individuals to three distinctly different haplotypes. The Y chromosome polymorphism analysis confirmed the presence of genotype C characteristics for Przewalski horses in all 13 male individuals, moreover, the polymorphisms within the coat color genes MC1R and TBX3 indicated only native, wild genotypes. Unexpected polymorphisms’ appearance in genes determining the coat color and in Y chromosome could indicate population crossbreeding, therefore the homogeneity in these genes and the Y chromosome in the Askania-Nova population needed to be confirmed.