Regulatory Single Nucleotide Polymorphism of the Bovine IFITM3 Gene Induces Differential Transcriptional Capacities of Hanwoo and Holstein Cattle

Interferon-induced transmembrane protein 3 (IFITM3), a crucial effector of the host’s innate immune system, prohibits an extensive range of viruses. Previous studies have reported that single nucleotide polymorphisms (SNPs) of the IFITM3 gene are associated with the expression level and length of the IFITM3 protein and can impact susceptibility to infectious viruses and the severity of infection with these viruses. However, there have been no studies on polymorphisms of the bovine IFITM3 gene. In the present study, we finely mapped the bovine IFITM3 gene and annotated the identified polymorphisms. We investigated polymorphisms of the bovine IFITM3 gene in 108 Hanwoo and 113 Holstein cattle using direct sequencing and analyzed genotype, allele, and haplotype frequencies and linkage disequilibrium (LD) between the IFITM3 genes of the two cattle breeds. In addition, we analyzed transcription factor-binding sites and transcriptional capacity using PROMO and luciferase assays, respectively. Furthermore, we analyzed the effect of a nonsynonymous SNP of the IFITM3 gene using PolyPhen-2, PANTHER, and PROVEAN. We identified 23 polymorphisms in the bovine IFITM3 gene and found significantly different genotype, allele, and haplotype frequency distributions and LD scores between polymorphisms of the bovine IFITM3 gene in Hanwoo and Holstein cattle. In addition, the ability to bind the transcription factor Nkx2-1 and transcriptional capacities were significantly different depending on the c.-193T > C allele. Furthermore, nonsynonymous SNP (F121L) was predicted to be benign. To the best of our knowledge, this is the first genetic study of bovine IFITM3 polymorphisms.

In a previous study, a splicing variant inducing the single nucleotide polymorphism (SNP) rs12252 was shown to be associated with the severity of pandemic influenza A 2009 virus infection. The C allele of the rs12252 SNP was suggested to generate a truncated isoform of the IFITM3 protein (∆21 IFITM3) and reduce antiviral capacity [3,12]. In addition, the rs34481144 SNP, which is located on exon 1 of the IFITM3 gene, was found to be related to a mechanism by which the transcription of the IFITM3 gene is downregulated [13,14]. An allele inducing transcriptional variation was shown to cause the severity of pandemic influenza A 2009 virus infection. Furthermore, rs6598045 SNP, which is located on the proximal promoter region of the IFITM3 gene, is correlated with a mechanism by which transcription of the IFITM3 gene is regulated [1]. Susceptibility to pandemic influenza A 2009 virus infection was significantly increased with the T allele of the SNP rs6598045, and a difference in transcriptional capacity was also detected. In chicken, the c.298C > A (L100M) SNP was found to induce a variation in the topology of the IFITM3 protein and increased the length of transmembrane domain 2 (TM2) [2]. In addition, the IFITM3 gene is expressed in primordial germ cells (PGCs), and IFITM3 protein plays a pivotal role in germline development via PGCs localization [15]. However, although several polymorphisms of the IFITM3 gene in various species are strongly associated with antiviral capacity and genetic features of the IFITM3 protein, polymorphisms of the bovine IFITM3 gene have not been investigated thus far.
In the present study, we investigated the bovine IFITM3 gene in Hanwoo and Holstein cattle by using direct sequencing and compared the genotype and allele frequencies of the IFITM3 gene in the two cattle breeds. In addition, we analyzed linkage disequilibrium (LD) and haplotype frequency of polymorphisms of the bovine IFITM3 gene. We also assessed differences in a transcription factor-binding site and transcriptional capacity based on an allele of the IFITM3 gene with a regulatory SNP using PROMO and luciferase assays, respectively [16]. Furthermore, we annotated the effect of a nonsynonymous SNP of the IFITM3 gene using PolyPhen-2, PANTHER, and PROVEAN [17][18][19].

Ethical Statement
Tissue samples from 221 cattle of 2 breeds (108 Hanwoo and 113 Holstein cattle) were provided from slaughterhouses in the Republic of Korea. All experimental procedures were approved by the Institute of Animal Care and Use Committee of Chonbuk National University (CBNU 2018-079).

Genetic Analysis of the IFITM3 Gene
Genomic DNA was extracted from 20 mg of brain tissue sample using a HiYield genomic DNA mini kit (Real Biotech Corporation, Banqiao Taiwan). Polymerase chain reaction (PCR) was carried out to amplify the bovine IFITM3 gene using BioFACT™ Taq DNA Polymerase (BioFACT, Daejeon, Korea). Information on bovine IFITM3 genespecific primers and experimental conditions is provided in Table 1. The PCR mixture contained 2.5 µL of 10× Taq DNA polymerase reaction buffer, 1 µL of genomic DNA, 10 pmol of each primer, 0.5 µL of a 0.2 µM dNTP mixture, 0.2 µL of 0.04 units of Taq DNA polymerase, and sterile deionized water in a total volume of 25 µL. PCR amplicons were directly sequenced by using an ABI 3730 sequencer (ABI, Foster City, California, USA), and sequencing electropherograms were visualized by using Finch TV software (Geospiza, Inc., Seattle, WA, USA).

Cell Culture
Embryonic bovine tracheal (EBtr) cells were provided by the Korea Cell Line Bank and maintained in Eagle's minimum essential medium (ATCC, Manassas, VA, USA). In order to prepare the complete growth medium, 10% (v/v) fetal bovine serum (Gibco, Gaithersburg, MD, USA) was added. EBtr cells were cultured at 37 • C in a humidified atmosphere of 5% CO 2 (v/v) in air.

Plasmids and Luciferase Assay
The promoter sequences based on alleles of the IFITM3 gene were synthesized and inserted into the pGL4.10 [luc2] vector (Promega, Fitchburg, WI, USA). Plasmid construction and preparation followed standard protocols. The plasmids were transfected using Lipofectamine (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The transfected cells were incubated for 30 h, and the promoter activity of the IFITM3 gene was measured with a Glomax 20/20 luminometer (Promega, Fitchburg, WI, USA) using a luciferase assay system (Promega, Fitchburg, WI, USA).

Statistical Analysis
Statistical analyses were performed using SAS version 9.4 (July 2013, SAS Institute, Inc., Cary, NC, USA). The differences in genotype and allele frequencies of the IFITM3 gene between cattle breeds were compared using the χ 2 test. The Hardy-Weinberg equilibrium (HWE) test and haplotype and LD analyses of 23 polymorphisms of the bovine IFITM3 gene were performed using Haploview version 4.2 (September 2009, Broad Institute, Cambridge, MA, USA). Luciferase assays were carried out in three independent experiments, and statistical significance was determined by p-value calculated by two-tailed Student's t-test for single comparisons. The symbol "***" indicates p < 0.001.

Identification of Polymorphisms of the Bovine IFITM3 Gene
In order to investigate polymorphisms of the bovine IFITM3 gene, we performed direct sequencing with IFITM3 gene-specific primers in 108 Hanwoo and 113 Holstein cattle (Table 1). We identified a total of 23 polymorphisms of the bovine IFITM3 gene, including 1 nonsynonymous SNP (c.361T > C, p.Phe121Leu) and 2 insertion/deletion polymorphisms (c.249+350_351delCA and c.249+395delG) (Figure 1).

Comparison of Genotype and Allele Frequencies of Polymorphisms of the Bovine IFITM3 Gene among Cattle Breeds
We investigated the differences in allele and genotype frequencies of the bovine

Comparison of Genotype and Allele Frequencies of Polymorphisms of the Bovine IFITM3 Gene among Cattle Breeds
We investigated the differences in allele and genotype frequencies of the bovine IFITM3 gene between Hanwoo and Holstein cattle. In brief, a total of 19 polymorphisms showed significantly different genotype and allele distributions between Hanwoo and Holstein cattle: 361T > C and c.396C > T. Among the 23 polymorphisms, the c.136C > T, c.249+350_351delCA, c.249+372C > T, c.249+395delG, c.249+455G > T, and c.249+472G > C polymorphisms were specific to Hanwoo cattle. In addition, the c.249+320G > C, c.249+64C > A, c.249+359G > A, and c.396C > T polymorphisms were specific to Holstein cattle ( Table 2).

LD and Haplotype Analyses of the Bovine IFITM3 Gene
Since breed-specific polymorphisms were identified, we carried out LD analysis of the 23 polymorphisms of the bovine IFITM3 gene in Hanwoo and Holstein cattle. The LD scores for the Hanwoo and Holstein cattle are shown in Tables 3 and 4, respectively. In brief, 27 high LD scores were identified in Hanwoo cattle, including two Hanwoo-specific LD scores (between c.249+320G > C and c.249+367G > A and between c.249+395delG and c.249+472G > C). In addition, 26 high LD scores were identified in Holstein cattle, including one Holstein-specific LD score (between c.105C > G and c.249+32G > C).
We also examined the haplotype distribution of the 23 polymorphisms of the bovine IFITM3 gene in Hanwoo and Holstein cattle. Detailed information on the haplotype distributions of bovine IFITM3 gene polymorphisms in Hanwoo and Holstein cattle is provided in Tables 5 and 6, respectively. In summary, a total of 12 major haplotypes were identified in Hanwoo cattle (Table 5). Among the 12 haplotypes, the TCCGCGCGWtG-GCWtGCAGTGGGCC haplotype had the highest frequency (14.8%), followed by the CCCGGTCGWtGGCWtGCAGTGGGCC (11.6%) and CCCGGTCGDelGACWtCAGTGAG-GTC (7.9%) haplotypes. In Holstein cattle, a total of six major haplotypes were identified (Table 6). Among these six haplotypes, the TCCGCGCGWtGGCWtGCAGTGGGCC haplotype had the highest frequency (43.4%), followed by the CCCGGTCGWtGGCWt-GCAGTGGGCC (16.4%) and TCCGCGCGWtAGCWtGCAGTGGGCC (9.7%) haplotypes. Table 3. Linkage disequilibrium (LD) scores among 23 polymorphisms of the bovine IFITM3 gene in Hanwoo. 0

The Transcription Factor-Binding Capacity of the Bovine IFITM3 Gene
We found two regulatory SNPs in the proximal promoter region of the bovine IFITM3 gene, and c.-193T > C showed significantly different genotypes and allele distributions in Hanwoo and Holstein cattle ( Table 2). We analyzed the transcription factor-binding capacity of the bovine IFITM3 gene according to the c.-193T > C alleles using PROMO. Interestingly, the haplotype with the T allele and the haplotype with the C allele differed in their ability to bind transcription factor Nkx2-1 (Figure 2).

The Transcription Factor-Binding Capacity of the Bovine IFITM3 Gene
We found two regulatory SNPs in the proximal promoter region of the bovine IFITM3 gene, and c.-193T > C showed significantly different genotypes and allele distributions in Hanwoo and Holstein cattle ( Table 2). We analyzed the transcription factorbinding capacity of the bovine IFITM3 gene according to the c.-193T > C alleles using PROMO. Interestingly, the haplotype with the T allele and the haplotype with the C allele differed in their ability to bind transcription factor Nkx2-1 (Figure 2).

Promoter Activities Based on Alleles of Regulatory SNPs in the Proximal Promoter Region of the Bovine IFITM3 Gene
We investigated the differences in promoter activity according to the alleles of promoter SNPs, which showed different genotype and allele frequencies in Hanwoo and Holstein cattle ( Table 2). The T-type promoter, which contained the T allele of c.-193T > C, was more prevalent in Hanwoo cattle. The C-type promoter, which contained the C allele of c.-193T > C, was more prevalent in Holstein cattle. Notably, the C-type promoter significantly increased the expression level of mRNA compared to that of the T-type promoter (Figure 3).

In Silico Annotation of a Nonsynonymous SNP of the Bovine IFITM3 Gene
The impact of the nonsynonymous SNP c.361T > C (F121L) of the bovine IFITM3 gene identified in this study was analyzed by PolyPhen-2, PANTHER, and PROVEAN. Notably, the F121L mutation was predicted to be benign by all three programs (Table 7). stein cattle ( Table 2). The T-type promoter, which contained the T allele of c.-193T > C, was more prevalent in Hanwoo cattle. The C-type promoter, which contained the C allele of c.-193T > C, was more prevalent in Holstein cattle. Notably, the C-type promoter significantly increased the expression level of mRNA compared to that of the T-type promoter (Figure 3).

In Silico Annotation of a Nonsynonymous SNP of the Bovine IFITM3 Gene
The impact of the nonsynonymous SNP c.361T > C (F121L) of the bovine IFITM3 gene identified in this study was analyzed by PolyPhen-2, PANTHER, and PROVEAN. Notably, the F121L mutation was predicted to be benign by all three programs (Table 7).

Discussion
Previous studies have reported that overexpression of the IFITM3 protein inhibits an extensive range of viruses under experimental conditions [20][21][22]. This propensity has been consistently reported under natural conditions. The duck species that are known to be resistant to avian influenza virus showed elevated expressions of the IFITM3 protein compared to that in the chicken species that are known to be susceptible to avian influenza virus. In addition, the genetic polymorphisms of Ross chickens, a kind of broiler, were significantly different from those of Dekalb White chickens, a kind of layer. Broilers show more resistance to viral infection than layers, a notable feature of the IFITM3 gene [2]. In humans, the SNPs rs34481144 and rs6598045 of the human IFITM3 gene, which are related to the modulated expression of this gene, showed prominent associations with the severity and susceptibility of pandemic influenza A 2009 virus infection, respectively [1,13,14]. In addition, the SNP rs12252, which influences the length of the IFITM3 protein, was found to be related to susceptibility to pandemic influenza A 2009 virus infection. Furthermore, recent studies have reported that the IFITM3 protein is also involved in not only immunerelated functions but also embryogenesis and feed efficiency [15,23].
Thus, in the present study, we investigated polymorphisms of the bovine IFITM3 gene, which can affect the expression level or function of the IFITM3 protein. We found a total of 23 polymorphisms in the bovine IFITM3 gene. In addition, Hanwoo and Holstein cattle showed significantly different genotype and allele frequencies and breed-specific polymorphisms ( Table 2). The LD scores and haplotype frequencies of the polymorphisms also showed different distributions between the two cattle breeds (Tables 3-6). These results indicate that the IFITM3 genes in Hanwoo and Holstein cattle have significantly different genetic properties, including genotype, allele, and haplotype frequencies and LD scores. In addition, we annotated a regulatory SNP (c.-193T > C) and a nonsynonymous SNP (c.361T > C, F121L) of the bovine IFITM3 gene. Strikingly, the C-type haplotype with the C allele of c.-193T > C and the T-type haplotype with the T allele of c.-193T > C differed in their ability to bind transcription factor Nkx2-1, and the C-type haplotype exhibited elevated expression of the IFITM3 gene compared to the T-type haplotype (Figure 3). Since the T allele is frequently observed in Holstein cattle while the C allele is frequently observed in Hanwoo cattle, this result suggests a significant difference in the expression of the bovine IFITM3 gene between the cattle breeds. Indeed, Nkx2-1, a member of the Nkx-homeodomain factor family, is related to the regulation of organ development. In addition, Nkx2-1 is associated with several diseases, including benign hereditary chorea, choreoathetosis, congenital hypothyroidism, and neonatal respiratory distress. Furthermore, Nkx2-1 has a function in organ development, and it is involved in morphogenesis [24,25]. Since the IFITM3 protein is also related to cellular developmental processes, further investigation of the relationship between IFITM3 and Nkx2-1 is highly desirable in the future. We also annotated a nonsynonymous SNP (c.361T > C, F121L) by using in silico annotation tools. The F121L mutation was predicted to be benign (Table 7). Since several viruses have been reported to be associated with the IFITM3 gene, future study of the characterization of the bovine IFITM3 gene in other local cattle breeds is highly desirable.

Conclusions
In conclusion, we finely mapped the bovine IFITM3 gene and annotated regulatory and nonsynonymous SNPs. We identified 23 polymorphisms of the bovine IFITM3 gene and significantly different genotype, allele, and haplotype distributions and LD scores for these polymorphisms of the bovine IFITM3 gene between Hanwoo and Holstein cattle. In addition, transcription factor-binding ability and transcriptional capacity were significantly different depending on regulatory SNP alleles. A nonsynonymous SNP (F121L) was predicted to be benign. To the best of our knowledge, this is the first genetic report of bovine IFITM3 polymorphisms.