Absence of Strong Genetic Linkage Disequilibrium between Single Nucleotide Polymorphisms (SNPs) in the Prion Protein Gene (PRNP) and the Prion-Like Protein Gene (PRND) in the Horse, a Prion-Resistant Species

Prion disease is a fatal neurodegenerative disorder caused by a deleterious prion protein (PrPSc). However, prion disease has not been reported in horses during outbreaks of transmissible spongiform encephalopathies (TSEs) in various animals in the UK. In previous studies, single nucleotide polymorphisms (SNPs) in the prion protein gene (PRNP) have been significantly associated with susceptibility to prion disease, and strong linkage disequilibrium (LD) between PRNP and prion-like protein gene (PRND) SNPs has been identified in prion disease-susceptible species. On the other hand, weak LD values have been reported in dogs, a prion disease-resistant species. In this study, we investigated SNPs in the PRND gene and measured the LD values between the PRNP and PRND SNPs and the impact of a nonsynonymous SNP found in the horse PRND gene. To identify SNPs in the PRND gene, we performed direct sequencing of the PRND gene. In addition, to assess whether the weak LD value between the PRNP and PRND SNPs is a characteristic of prion disease-resistant animals, we measured the LD value between the PRNP and PRND SNPs using D’ and r2 values. Furthermore, we evaluated the impact of a nonsynonymous SNP in the Doppel protein with PolyPhen-2, PROVEAN, and PANTHER. We observed two novel SNPs, c.331G > A (A111T) and c.411G > C. The genotype and allele frequencies of the c.331G > A (A111T) and c.411G > C SNPs were significantly different between Jeju, Halla, and Thoroughbred horses. In addition, we found a total of three haplotypes: GG, AG, and GC. The GG haplotype was the most frequently observed in Jeju and Halla horses. Furthermore, the impact of A111T on the Doppel protein was predicted to be benign by PolyPhen-2, PROVEAN, and PANTHER. Interestingly, a weak LD value between the PRNP and PRND SNPs was found in the horse, a prion disease-resistant animal. To the best of our knowledge, these results suggest that a weak LD value could be one feature of prion disease-resistant animals.


Ethical Statement
The hair samples of 148 Jeju and 100 Halla horses were collected from Jeju Island. All experimental procedures were approved by the Institute of Animal Care and Use Committee of Jeonbuk National University (JBNU 2016-65). All experiments using horses were carried out following the Korea Experimental Animal Protection Act.

Genetic Analysis
The genomic DNA of Halla and Jeju horses was isolated from 10 hair bulbs using a HiYield™ Genomic DNA Mini Kit (Real Biotech Corporation, Banqiao city, Taiwan) according to the manufacturer's instructions. Polymerase chain reaction (PCR) on the horse PRND coding region located at exon 3 was performed with sense and antisense primers as follows: horse PRND forward 5 -GCCCGTTGCAGCTTCTTATCT-3 and horse PRND reverse 5 -GCTGGAGGAGAGAAGTGGGAT-3 . PCR was performed in a final volume of 25 µL, including 1 µL of genomic DNA, 10 pmol of each primer, 2.5 µL of 10X Taq DNA polymerase buffer, 0.5 µL of a 0.2 µM dNTP mixture, 2.5 µL of 5X band helper, and 0.2 µL Taq DNA polymerase. The PCR conditions were as follows: denaturing at 95 • C for 2 min, 33 cycles of 95 • C for 20 s, 61 • C for 40 s, 72 • C for 1 min 30 s, and 1 cycle 72 • C for 5 min. The purification of PCR products was performed using a FavorPrep GEL/PCR Purification Mini Kit (FAVORGEN, Pingtung city, Taiwan). The PCR products were directly sequenced using an ABI 3730 sequencer (ABI, Foster City, CA, USA), and sequencing results were analyzed using Finch TV software (Geospiza, Inc., Seattle, WA, USA).

Statistical Analysis
The genotype, allele, and haplotype frequencies of the horse PRND gene were compared between 3 horse breeds by the chi-square test using SAS 9.4 software. In addition, we examined LD calculated by Lewontin's D' (|D'|), and pairwise linkage disequilibrium (r 2 ) was examined using Haploview version 4.2 (Broad Institute, Cambridge, MA, USA). The Hardy-Weinberg equilibrium (HWE) test and haplotype analysis were carried out using Haploview version 4.2 (Broad Institute, Cambridge, MA, USA).

Analysis of Genetic Linkage between SNPs in the PRNP and PRND Genes
We performed LD analysis between PRNP and PRND SNPs in 135 Jeju and 72 Halla horses. In addition, we examined LD calculated by Lewontin's D' (|D'|), and pairwise linkage disequilibrium (r 2 ) was examined using Haploview version 4.2 (Broad Institute, Cambridge, MA, USA).

Measurement of Protein Functional Alterations Induced by Nonsynonymous SNPs
We predicted the damaging effect of amino acid substitution induced by nonsynonymous SNPs using the PolyPhen-2 v.2.2.2, PROVEAN v.1.1.3, and PANTHER 15.0 programs. PolyPhen-2 evaluates the effect of amino acid changes from an independent count (PSIC), represented as a score ranging from 0 (tolerated) to 1 (deleterious). The prediction range of 0.0 to 0.15 indicates 'benign', 0.15 to 1.0 indicates "possibly damaging", and 0.85 to 1.0 indicates "damaging". PROVEAN estimates nonsynonymous or indel variants; a score below −2.5 is considered "deleterious", and a score above −2.5 is considered "neutral". PANTHER classifies amino acids into groups with similar molecular biology. In PANTHER, a score below −3 indicates "deleterious" and a score above −3 indicates "neutral". . We obtained sequence information from GenBank at the National Center for Biotechnology Information (NCBI).

Investigation of Polymorphisms in the Horse PRND Gene in 140 Jeju and 80 Halla Horses
The horse PRND gene consists of two exons, and exon 2 contains the ORF. We investigated polymorphisms in exon 2 of the PRND gene in two horse breeds, Jeju and Halla. The PRND gene sequences of Halla and Jeju horses are identical to that of Equus caballus registered in GenBank (Gene ID: 100048937). We discovered a total of two novel SNPs, c.331G > A and c.411G > C. c.331G > A (A111T) is a nonsynonymous SNP, and c.411G > C is a synonymous SNP (Figure 1a,b). The genotype and allele frequencies of the PRND SNPs are described in Tables 1 and 2. Both SNPs were in HWE.
Next, we compared the genotype and allele frequencies of the c.331G > A (A111T) and c.411G > C SNPs in the horse PRND gene in Jeju, Halla, and Thoroughbred horses. Regarding the c.331G > A (A111T) SNP, the genotype and allele frequencies in the Jeju horse were significantly different from those in the Halla (p < 0.0001) and Thoroughbred (p < 0.0001) horses. However, regarding the c.411G > C SNP, the genotype distribution in the Jeju horse showed no significant difference with that in the Halla horse (p = 0.06) but showed a significant difference with that in the Thoroughbred horse (p < 0.0001). In addition, the allele frequency in the Jeju horse showed no significant difference with that in the Halla horse (p = 0.0631) but showed a significant difference with that in the Thoroughbred horse (p = 0.0022) (   * NA, Not available. We performed a haplotype analysis of the two PRND SNPs in Jeju and Halla horses. A total of three haplotypes, GG, AG, and GC, were observed in these two horses. The GG haplotype was the most frequently observed in the Jeju (74.3%) and Halla (98.1%) horses. The haplotype frequencies between Jeju and Halla horses were significantly different (p < 0.0001) ( Table 3). We also estimated the biological impact of the nonsynonymous SNP c.331G > A (A111T) using PolyPhen-2, PROVEAN, and PANTHER. A111T was predicted to be "benign", with a score of 0.006 by PolyPhen-2, "neutral", with a score of −0.969 by PROVEAN, and "probably benign", with a score of 176 by PANTHER (Table 4). Next, we measured LD of the two horse PRND SNPs by (|D'|) and r 2 values. Both SNPs were strongly linked, with a D' value of 1.0 according to the (|D'|) value in Jeju and Halla horses. However, the r 2 value showed weak LD, with values below 0.02 (Table 5).

Analysis of Genetic Linkage between SNPs in the PRNP and PRND Genes
To examine whether horse PRND SNPs have genetic linkage with the PRNP gene, we measured the LD values between SNPs in the PRNP and PRND genes. All PRND SNPs showed weak LD with the PRNP SNPs (r 2 value: below 0.1) (Figure 3). The detailed r 2 values are described in Table 6.

Analysis of Genetic Linkage Between SNPs in the PRNP and PRND Genes
To examine whether horse PRND SNPs have genetic linkage with the PRNP gene, we measured the LD values between SNPs in the PRNP and PRND genes. All PRND SNPs showed weak LD with the PRNP SNPs (r 2 value: below 0.1) (Figure 3). The detailed r 2 values are described in Table 6.  Table 6. Linkage disequilibrium (LD) between single nucleotide polymorphisms (SNPs) in the PRNP and PRND genes with r 2 values in Jeju and Halla horses.

Measurement of the Physical Map Distance between the PRNP and PRND Genes among Several Species
We measured the physical map distance between the PRNP and PRND genes in prion disease-susceptible species, including humans, sheep, and goats, and prion disease-resistant species, including dogs and horses. Notably, prion disease-resistant species, including dogs (17 kb) and horses (16 kb), showed relatively short physical map distances between the PRNP and PRND genes compared to those of prion disease-susceptible animals, including humans (20 kb), sheep (25 kb), and goats (25 kb) (Figure 4).
Genes 2020, 11, x FOR PEER REVIEW 10 of 12 value between horse PRNP and PRND SNPs. A weak LD value may be regarded as a common feature of prion disease-resistant species (e.g., dogs and horses) [29]. However, the study is limited to a few breeds of only one alleged resistant species and should be extended to more breeds and/or other resistant species. Thus, careful validation is needed in future study.
In addition, prion disease-resistant species showed weak LD values despite the short physical map distance between the PRNP gene and the PRND gene (humans: 20 kb; sheep: 25 kb; goats: 25 kb; dogs: 17 kb; horses: 16 kb) ( Figure 4). Interestingly, although there was a short physical map distance between the PRNP and PRND genes, weak LD between these two genes was identified in prion disease-resistant animals ( Table 7). Strong genetic linkage between PRNP and PRND genes in prion disease-susceptible animals indicated that prion disease-susceptible animals were selected with prion disease-contributable alleles of PRNP and PRND genes. Thus, evaluation of susceptibility to prion disease according to alleles of polymorphisms in the PRNP and PRND genes will be highly needed in the future.

Conclusions
In conclusion, we found a total of two novel SNPs, one nonsynonymous SNP (c.331G > A, A111T) and one synonymous SNP (c.411G > C), in the PRND gene in Jeju and Halla horses. The genotype and allele

Genetic Linkage between PRNP and PRND Genes among Several Species
Lastly, we summarized the genetic linkage between PRNP and PRND genes in prion disease-susceptible species, including humans, sheep, and goats, and prion disease-resistant species, including dogs and horses. Notably, prion disease-susceptible animals showed strong genetic linkage between the PRNP and PRND genes compared to prion disease-resistant species (Table 7). Table 7. Genetic linkage between PRNP and PRND genes in several animals.

Discussion
A previous study surveyed PRND gene polymorphisms in the Thoroughbred horse. However, we did not find polymorphisms in the PRND gene [30]. Because mating in the Thoroughbred horse is strictly regulated, it may cause nonpolymorphic genetic features of the PRND gene. Therefore, we investigated polymorphisms in the horse PRND gene in outbred breeds such as Jeju and Halla.
The Jeju horse is a Korean native horse and is registered and protected as Natural Monument No. 347. According to molecular genetic studies of Korean native Jeju horses, some of the ancestors of Jeju are believed to be of Mongolian origin [31], and some Jeju horses are estimated to have an independent maternal lineage. Halla horses are a crossbreed of the Jeju and Thoroughbred horses, which has been deemed a better breed of horses by the National Institute of Animal Science for horse racing. The lineage of Jeju horses is managed by the Jeju Horse Database System [32], operated by the Livestock Promotion Agency of Jeju Special Self-Governing Province. According to the Livestock Promotion Agency, 5223 Jeju horses were registered and are being raised in 613 breeding farms so far. In addition, 5201 Halla horses were registered and are being raised in 662 breeding farms. The habitat of Jeju and Halla horses is Jeju Island, which is a small island with an area of 1845 km 2 . Interestingly, we found a total of two novel SNPs, including one nonsynonymous SNP (c.331G > A, A111T) and one synonymous SNP (c.411G > C), in the ORF of the horse PRND gene. Unlike Thoroughbred horses, which are distributed in a wide range of habitats, SNPs were first discovered in Jeju and Halla horses despite their limited range of habitats. It can be assumed that the Jeju horse may have evolved longer than the Thoroughbred horse or maintained by free crossing.
According to a previous study, prion disease-susceptible species, such as sheep and goats, show strong LD between SNPs in the PRNP and PRND genes [27,28]. In contrast, prion disease-resistant species, such as dogs, show low LD values between these two genes. In this study, we first discovered a weak LD value between horse PRNP and PRND SNPs. A weak LD value may be regarded as a common feature of prion disease-resistant species (e.g., dogs and horses) [29]. However, the study is limited to a few breeds of only one alleged resistant species and should be extended to more breeds and/or other resistant species. Thus, careful validation is needed in future study.
In addition, prion disease-resistant species showed weak LD values despite the short physical map distance between the PRNP gene and the PRND gene (humans: 20 kb; sheep: 25 kb; goats: 25 kb; dogs: 17 kb; horses: 16 kb) ( Figure 4). Interestingly, although there was a short physical map distance between the PRNP and PRND genes, weak LD between these two genes was identified in prion disease-resistant animals (Table 7). Strong genetic linkage between PRNP and PRND genes in prion disease-susceptible animals indicated that prion disease-susceptible animals were selected with prion disease-contributable alleles of PRNP and PRND genes. Thus, evaluation of susceptibility to prion disease according to alleles of polymorphisms in the PRNP and PRND genes will be highly needed in the future.

Conclusions
In conclusion, we found a total of two novel SNPs, one nonsynonymous SNP (c.331G > A, A111T) and one synonymous SNP (c.411G > C), in the PRND gene in Jeju and Halla horses. The genotype and allele distributions of horse PRND SNPs were significantly different between Jeju, Halla, and Thoroughbred horses. In addition, we measured LD between SNPs in the PRNP and PRND genes and found a weak LD value. Furthermore, we investigated the physical map distance between the PRNP and PRND genes in prion disease-resistant and -susceptible species. Notably, although there was a short physical map distance between the PRNP and PRND genes, weak LD between these two genes was identified in prion disease-resistant animals. To the best of our knowledge, we provide evidence that a weak LD value could be a feature of prion disease-resistant animals.