Novel insertion/deletion polymorphisms and genetic features of the shadow of prion protein gene (SPRN) in dogs, a prion-resistant animal

Prion diseases are fatal infectious neurodegenerative disorders that are induced by misfolded prion protein (PrPSc). Previous studies have reported that the shadow of prion protein (Sho) encoded by the shadow of prion protein gene (SPRN) plays a critical role in stimulating the conversion process of normal PrP (PrPC) into PrPSc, and genetic polymorphisms of the SPRN gene are significantly related to susceptibility to prion diseases. Recent studies have reported that dogs show prion resistance, and there have been several attempts to identify resistance factors to prion diseases in dogs. However, there has been no study of the canine SPRN gene thus far. We investigated genetic polymorphisms of the canine SPRN gene in 201 dogs using amplicon sequencing and compared the number of SPRN polymorphisms among prion-related species. In addition, we performed multiple sequence alignments of the amino acid sequences of Sho among prion-related species by ClustalW and analyzed the 3D structure of Sho using AlphaFold. Furthermore, we assessed the protein–protein interaction of canine PrP with canine Sho carrying wild-type and mutant alleles using HawkDock. We found four novel insertion/deletion polymorphisms of the SPRN gene in 201 dogs and identified a significant difference in the number of SPRN polymorphisms between prion-susceptible and prion-resistant animals. In addition, Sho has two α-helixes linked with the coil. Furthermore, we found different binding complexes and binding free energies between canine Sho and PrP according to SPRN polymorphisms. To the best of our knowledge, this is the first report of canine SPRN polymorphisms.

Prion diseases are fatal infectious neurodegenerative disorders that are induced by misfolded prion protein (PrP Sc ). Previous studies have reported that the shadow of prion protein (Sho) encoded by the shadow of prion protein gene (SPRN) plays a critical role in stimulating the conversion process of normal PrP (PrP C ) into PrP Sc , and genetic polymorphisms of the SPRN gene are significantly related to susceptibility to prion diseases. Recent studies have reported that dogs show prion resistance, and there have been several attempts to identify resistance factors to prion diseases in dogs. However, there has been no study of the canine SPRN gene thus far. We investigated genetic polymorphisms of the canine SPRN gene in dogs using amplicon sequencing and compared the number of SPRN polymorphisms among prionrelated species. In addition, we performed multiple sequence alignments of the amino acid sequences of Sho among prion-related species by ClustalW and analyzed the D structure of Sho using AlphaFold. Furthermore, we assessed the protein-protein interaction of canine PrP with canine Sho carrying wild-type and mutant alleles using HawkDock. We found four novel insertion/deletion polymorphisms of the SPRN gene in dogs and identified a significant di erence in the number of SPRN polymorphisms between prionsusceptible and prion-resistant animals. In addition, Sho has two α-helixes linked with the coil. Furthermore, we found di erent binding complexes and binding free energies between canine Sho and PrP according to SPRN polymorphisms. To the best of our knowledge, this is the first report of canine SPRN polymorphisms.

Introduction
Prion diseases are fatal and infectious disorders that are reported in a wide range of hosts, including humans, cattle, sheep, goats, deer, and elk. Prion diseases occur by misfolded prion protein (PrP Sc ) originating from normal prion protein (PrP C ) (1-3). To date, although the conversion process of PrP C to PrP Sc has been elusive, several cofactors that play a critical role in the exacerbation of prion diseases have been identified (4)(5)(6)(7)(8)(9). Among these cofactors, the shadow of prion protein (Sho) encoded by the shadow of prion protein gene (SPRN) stimulates the conversion process of PrP C to PrP Sc via interaction with PrP (7). Thus, previous studies have reported that genetic polymorphisms of the SPRN gene that affect the expression level, function, and structure of Sho are involved in susceptibility to several types of prion diseases. In humans, a null allele of the SPRN gene was observed in variant Creutzfeldt-Jakob disease (vCJD) patients (10). In addition, the insertion polymorphism in the 3' untranslated region (UTR) of the caprine SPRN gene is related to susceptibility to scrapie in goats (11). Furthermore, rare sequence variation in the open reading frame (ORF) of the bovine SPRN gene was found in L-type atypical bovine spongiform encephalopathy (BSE)-affected cattle (12). These studies indicate a clear association of Sho with pathomechanisms in various prion diseases. In recent studies, prion diseases have not been reported in dogs during the 1986 BSE outbreak. Thus, there have been several attempts to identify resistance to prion diseases in dogs (13-17). To date, Sho, a critical factor in PrP Sc conversion, has not been analyzed. Furthermore, as Sho has an unstructured 3D structure, comparative analysis of the 3D structure of Sho in prion-related animals has not been investigated thus far.
In the present study, we investigated genetic polymorphisms of the canine SPRN gene in 201 dogs using amplicon sequencing and compared the number of SPRN polymorphisms among prion-related species. In addition, we compared amino acid sequences of several prion-related species to find unique genetic characteristics of canine Sho by ClustalW and predicted the 3D structure of Sho of several prionrelated species using AlphaFold (18,19). Furthermore, we analyzed the protein-protein interaction of canine PrP with canine Sho carrying wild-type and mutant alleles using HawkDock (20).

Ethical statements
All experimental protocols were approved by the Institute of Animal Care and Use Committee of Jeonbuk National University (JBNU 2018-062).

Samples
The whole-blood samples were byproducts of completing blood count to evaluate overall health and testing several diseases, including anemia, infection, and leukemia in dogs and were collected from 201 dogs by a qualified veterinarian. The samples were provided by Anyang Cool-Pet Animal Hospital in South Korea. The dogs consisted of eight dog breeds. Detailed information on the dog breeds is provided in Supplementary Table 1. The samples were obtained depending upon hospital visitation based on the most common dog breed in the South Korea. The dog breed was determined by a pedigree document.

Genomic DNA extraction
Genomic DNA was extracted from whole-blood samples using the Labopass Blood Genomic DNA Isolation Kit Mini (Cosmogenetech, Seoul, Korea) following the method provided by the manufacturer. The quality and concentration of genomic DNA samples (15-20 ng/µl, A260/280 = 1.8-2.0) were assessed by Nanodrop (Thermo Scientific, Waltham, USA).

Multiple sequence alignments
The amino acid sequences of Sho were collected from GenBank at the National Center for Biotechnology Information (NCBI). Detailed information on Sho is described in Supplementary Table 2. The sequences were compared by ClustalW based on progressive alignment methods. PROVEAN PROVEAN (http://provean.jcvi.org/seq_submit.php) was used to evaluate the effect of amino acid substitution on the structure/function of proteins. PROVEAN assigns scores by calculating BLAST hits of homologs extracted from the NCBI. PROVEAN scores were classified as follows: above −2.5 indicates "neutral, " and below −2.5 indicates "deleterious." . /fvets. .

AlphaFold
The 3D structure of Sho was predicted by AlphaFold based on machine learning (https://colab.research.google.com/github/ deepmind/alphafold/blob/main/notebooks/AlphaFold.ipynb). The confidence of modeling was assessed by the predicted local distance difference test (pLDDT) score on a scale from 0 to 100.

HawkDock
Protein-protein interactions and complex structures were predicted by HawkDock based on the docking algorithm "ATTRACT". MM/GBSA calculated the structure evaluation of protein-protein interactions and binding free energy. The 3D structure of canine PrP (PDB ID: 1XYK) was deposited as the receptor. 3D structures of canine Sho with wild-type and mutant alleles were generated by AlphaFold and deposited as the ligands.

TM-Align
The structural similarity of two query proteins was evaluated by TM-align based on heuristic dynamic programming iteration. TM-align calculated TM-score and average root mean square deviation (RMSD). TMalign scores below 0.2 indicate "randomly chosen unrelated protein." TM-align scores above 0.5 indicate "in about the same fold."

Identification of four novel insertion/deletion polymorphisms of the canine PRNP gene in dogs
To investigate canine SPRN gene polymorphisms, we performed PCR to amplify the ORF region of the canine SPRN gene. We identified four novel insertion/deletion polymorphisms, including c.209_214DelCGGCGG (70_71DelAA) in the ORF region and c.444+30_444+31InsGGCCCCC, c.444+30_444+31InsGGCCCCCGGCCCC, and c.444+31_444+32InsGGCCCC in the intron (Figure 1). The genotype and allele distributions of the insertion/deletion polymorphisms of the canine SPRN gene are described in Table 1. We also investigated the LD values among the three canine SPRN polymorphisms with r 2 values (Table 2). Notably, strong LD (r 2 >0.333) was found between c.444+30_444+31InsGGCCCCC and c.444+31_444+32InsGGCCCC. In addition, we performed a haplotype analysis of four polymorphisms of the canine SPRN gene (Table 3). The WT-WT-WT-Ins haplotype was most frequently observed (31.2%), followed by Del-Ins-WT-WT (22.3%) and WT-WT-WT-WT (22.2%) in the canine SPRN gene (Table 3).

Comparison of Sho of dogs with that of several species
We compared amino acid sequences of Sho among seven species, including humans, cattle, sheep, goats, red deer, horses, and dogs, using multiple sequence alignments (Figure 3). Notably, we found a high conservation of the PrP interaction domain of Sho (red box), the NXT glycosylation motif (black box), and the omega site (serine) and signal sequence (green box) of the C-terminal domain among all species. The canine Sho showed a total of eight canine-specific amino acids, including D95, G99, A102, G115, F116, S128, R142, and P146. In addition, equine Sho showed a total of 11 equine-specific amino acids, including R84, S86, G87, V93, D99, S103, Q111, S117, E124, L129, and C130. However, these codons were not in the same location as polymorphisms. Among several species, human Sho has the highest identity with canine Sho (80.8%) followed by equine Sho (80.3%), bovine Sho (77.2%), and ovine Sho (75.3%) (Supplementary Table 2).

In silico evaluation of the e ect according to canine Sho substitutions using PROVEAN
To evaluate the effect of the substitutions on the function of canine Sho, PROVEAN was used (Table 4). Notably, 70_71DelAA was predicted to be "deleterious" with a score of "−5.125." However, all substitutions of canine-specific amino acids with interspecies conserved amino acids were predicted to be "neutral."

Prediction of the D structure of Sho among various species using AlphaFold
AlphaFold was used to analyze the 3D structure of Sho among seven species, including humans, cattle, sheep, goats, red deer, horses, dogs, and chickens ( Figure 4). Notably, all Sho was predicted to have two α-helixes linked with the coil Frontiers in Veterinary Science frontiersin.org . /fvets. .  Figures 4A-I). In addition, we predicted the 3D structure of canine Sho carrying the 70_71DelAA allele. We did not find a significant difference between the canine Sho carrying the wildtype and the 70_71DelAA alleles ( Figures 4G,H). We evaluated the structural similarity of wild-type Sho of dogs with Sho of several species using TM-align (Supplementary Table 3). All Sho proteins have a higher TM-align score above 0.2 and lower RMSD below 6Å. TM-align indicates that the wild-type Sho of dogs has a similar structure to Sho of humans, cattle, sheep, goats, red deer, and horses.

Protein-protein interaction of canine Sho with canine PrP using HawkDock
As 70_71DelAA is located in the interaction region of Sho with PrP, we assessed the impact of 70_71DelAA on the protein-protein interaction of canine Sho with canine PrP ( Figure 5). A different binding conformation was observed in the complex of the protein-protein interaction of canine PrP with canine Sho carrying wild-type and 70_71DelAA alleles. Although the canine Sho carrying wild-type and 70_71DelAA alleles have similar 3D structures of protein (Figure 4), the TM-align (0.64840) and RMSD (3.74 Å) scores indicate that there are some structural differences between canine Sho carrying wild-type and 70_71DelAA alleles (Supplementary Table 3). As expected, the binding free energy of canine PrP with canine Sho carrying the wild-type allele (−39.96 kcal/mol) is lower than that of canine PrP with canine Sho carrying the 70_71DelAA allele (-22.53 kcal/mol). We also observed the difference in TMalign and RMSD score between canine Sho carrying wildtype allele (TM-align score: 0.23865; RMSD: 3.92 Å) and canine Sho carrying 70_71DelAA allele (TM-align score: 0.24947; RMSD: 3.74 Å) (Supplementary Table 4), indicating that binding complexes have structural differences according to the Indel polymorphism. In addition, we investigated the conformational difference of the PrP-Sho complex in other species (Supplementary Figure 1) and observed that the PrP-Sho complex of each species has similar conformation but a different binding free energy.

Discussion
In the present study, we found only one insertion/deletion polymorphism in the ORF of the canine SPRN gene. Previous studies have reported that prion-susceptible animals, including cattle, sheep, cats and goats, have highly polymorphic SPRN genes (11,12,(21)(22)(23)(24)(25)(26). In contrast, a prion-resistant animal, the horse, has one polymorphism of the SPRN gene (27, 28). The number of prion-resistant animals is adequate to find the rare genetic polymorphisms with < 1% frequency, even though genetic polymorphisms have been explored only in a relatively small sample size in prion-resistant animals compared to prionsusceptible animals. As genetic polymorphisms may affect the function and structure of Sho, the difference in the number of these polymorphisms is noticeable. Future research is needed on the relationship between the number of SPRN polymorphisms and the susceptibility since pathogenic mutations are more likely to occur in highly polymorphic SPRN genes.
We also found three insertion/deletion polymorphisms located downstream of the SPRN gene. However, because the structure of the SPRN gene has not been confirmed, further annotation of these 3 SPRN polymorphisms is not available. Thus, it is highly necessary to further analyze the downstream region of the SPRN gene using 3 ′ rapid amplification of cDNA ends (RACE) to investigate the impact of the 3 SPRN polymorphisms. We observed strong LD between c.444+30_444+31InsGGCCCCC and c.444+31_444+32InsGGCCCC (Table 2). This indicates that LD block is present between these 2 SPRN polymorphisms. As Sho was predicted to have an unstructured 3D structure, the whole 3D structure of Sho has not been determined thus far. Thus, we analyzed the whole 3D structures of Sho in prion-related species. Notably, all species have a similar 3D structure with 2 α-helixes linked with the coil (Figure 4). This indicates that a significant difference in the 3D structure of Sho was not observed between prion-susceptible and prionresistant animals.
In the present study, the canine SPRN polymorphism found in the ORF is located in the interaction region of Sho with PrP. Thus, we investigated the difference in .
/fvets. .  protein-protein interactions according to alleles of the canine SPRN polymorphisms. Notably, we found different binding complexes of canine PrP with canine Sho according to SPRN polymorphisms ( Figure 5). In addition, the binding free energy of Sho and PrP is also different according to SPRN polymorphisms. The canine Sho with 70_71DelAA allele showed higher binding free energy compared to wild-type canine Sho ( Figure 5). This result indicates that the canine Sho with 70_71DelAA allele has an unstable complex compared to wildtype Sho. In addition, a complex generated by canine PrP and Sho with the 70_71DelAA allele differs compared to the interaction of that with wild-type canine Sho, and to those shown in other species ( Figure 5, Supplementary Figure 1). All interactions seem to proceed between the α-helical motifs from PrP and Sho, except the interaction of canine Sho with 70_71DelAA allele and canine PrP, where PrP appears as interacting with the unstructured region of Sho. However, the role of this polymorphism in pathology of prion diseases is elusive. Further bioassay using prion-infected dogs carrying the wild-type and 70_71DelAA alleles of canine SPRN gene is needed to confirm the role of the this polymorphism. In addition to prion diseases, Sho has various physiological roles, including the development and differentiation of embryos and mammary glands. The SPRN-knockout pups showed augmented embryonic lethality rates. Furthermore, SPRN knockout female mice showed a lactation defect (29-31).
As the difference in Sho according to SPRN polymorphisms was identified, further investigation of the impact of the polymorphisms on the physiology of the embryo and mammary gland is highly desirable in the future.
Although we predicted the 3D structure of Sho using AlphaFold, the coil between the two α-helices has a low confidence (pLDDT < 70) in all species. In addition, conserved putative interaction sites with PrP (codons 63-77 in dogs) also showed low confidence (pLDDT < 50). To confirm the 3D structure of Sho and obtain reliable results for comparison of putative interaction site with PrP, the identification of Sho structure using nuclear magnetic resonance (NMR) and X-ray crystallography is needed in the future. Furthermore, while we investigated the 3D structure of Sho in several species, we did not analyze the impact of prion-resistant polymorphisms found in prion-susceptible species. Future research on the impact of prion-resistant genetic polymorphisms on the 3D structures of the Sho and PrP-Sho binding complex is highly desired.

Conclusion
In summary, we identified four insertion/deletion polymorphisms of the SPRN gene in 201 dogs and found a significant difference in the number of SPRN polymorphisms located on the ORF between prion-susceptible Frontiers in Veterinary Science frontiersin.org . /fvets. . and prion-resistant animals. In addition, we found that the 3D structure of Sho is similar to that of prion-related animals. Furthermore, we found a significant difference in the protein-protein interaction of canine PrP with canine Sho according to SPRN polymorphisms. To the best of our knowledge, this is the first report of canine SPRN polymorphisms.

Data availability statement
All other data supporting the findings of this study are available from the corresponding author on reasonable request.

Ethics statement
The animal study was reviewed and approved by Institutional Animal Care and Use Committee (IACUC) of Jeonbuk National University.