Introduction

The respiratory and reproductive syndrome (PRRS) is a worldwide disease that causes economic losses in the porcine industry estimated at approximately $3.08 American dollars per pig at market¹.

The etiology agent of PRRS is an RNA virus of single chain belonging to Artevirus gender, whose main characteristics are an elevated mutation rate that confers a high antigenic variability, and its capacity to induce persistent infections^2,3. The initial report of PRRS disease in Mexico described that infected sows showed a case of reproductive problem and mortality in the production line⁴. The first clinical, epidemiological and productive description of the disease, as well as the first isolation of the PRRS virus, were reported in the states of Mexico, Guanajuato, Veracruz and Puebla⁵.

The PRRS virus infection is characterized by poor feed conversion that leads to a low weight in pigs, as well as fertility alterations in breeding sows such as estrus return, fetal mortality, mummification, abortions induction and low viability of piglets at birth¹.

Vaccination is the most common method for PRRS control and currently it has achieved to prevent in some extent the PRRS infection. However, the efficiency of the vaccines is still far to be universal because of the virus has the ability to avoid the immune response of the host; moreover, there exist genetic differences among hosts in response to vaccine virus exposition^6,7. The vaccination is able to show some efficiency against homologous PRRS strains, but its efficiency against heterologous strains is drastically reduced. Therefore, the vaccination against the PRRS virus at present only guarantee a reduction in the length of the viremia and the elimination of the virus cycle, as well as a decrease in the intensity of signs and the appearance of clinical symptoms⁸.

The existence of genetic variants associated to the interaction between the PRRS virus and the host, as well as the evidence of a natural variability in the tolerance and/or susceptibility to the PRRS in the commercial porcine lines, they are opened the door for using molecular technologies as a valuable tool to battle the PRRS disease⁹. In this regard, the marker assisted selection (MAS) can be used to study candidate genes in order to identify those animals that possess a superior genetic ability for the expression of economically important traits, which include resistance or tolerance to diseases¹⁰. First examples of the application of these technologies in pigs were the selection against the halothane gene, and the identification of a significant association between the estrogen receptor gene and the number of live-born piglets¹¹. However, the current development of more robust computer systems has allowed to perform the whole genome selection, which involves an extensive use of molecular markers that cover the entire genome, in such a way that hundreds of thousands of molecular variants can be simultaneously studied in order to explain the generic variation of a phenotypic trait¹². This method allows to perform associative studies that include the simultaneous analysis of a great amount of markers through the use of low- (10k=10,000) or high-density devices (50k=50,000 to 60K=60,000 SNP).

Several studies have been developed in pigs with the objective to identify regions within the DNA related to economically important traits such as the resistance to the PRRS virus^13,14,15. Initial reports suggest the existence of a genetic basis associated to the PRRS disease. In this regard, pigs from the breed Hampshire infected with PRRS showed pulmonary damages more serious than pigs from the breeds Duroc and Meishan¹⁶. Furthermore, pigs from the synthetic line Large White-Landrace showed a lower rectal temperature and a reduction in the viremia after be infected with the PRRS virus, in comparison with pigs from the synthetic line Hampshire-Duroc¹⁷. Recently, it has been reported within the chromosome 4 a genomic region associated to the resistance to the PRRS virus, which evidenced the existence of a strong genetic component associated to such ability³.

Currently, there is scarce information that report genes and/or genetic variants related to the phenotypic differences observed in the reproductive efficiency of sows infected with the PRRS virus. Therefore, the genetic foundation analyses of the reproductive response of these sows could lead to the identification of genetic markers associated to an appropriate reproductive performance, which would be very useful for the implementation of more efficient selection programs that include sows with superior genetic ability to tolerate and/or resist the infection of the PRRS virus.

Based on the previous information, the objective of the present study was to identify single nucleotide polymorphisms associated to the number of live-born piglets in the first (LBP1) and second birth (LBP2) in breeding sows infected with the PRRS virus.

Material and methods

Results

Genome-wide association study

A total of 8,856 SNP fulfilled the criteria of quality to be included in the associative genomic analyses that detected genomic regions associated to LBP1 (chromosomes 6 and 7; Figure 1) and to LBP2 (chromosomes 2, 5, 7 and 8; Figure 2), which explain 3.6 and 4.1 % of the variation associated to the traits LBP1 and LBP2, respectively. The individual genomic analyses detected three SNP associated to the variable LBP1 (P<0.001) and five SNP associated to the variable LBP2 (P<0.001). Only two of the eight SNP mentioned before are located within functional gene regions (introns), the rs81276080 within the gen TTR (Transthyretin) and the rs80893794 within the gen CWH43 (Cell wall biogenesis 43 C-terminal homolog). The record of the eight SNP detected as associated to the traits LBP1 and LBP2, as well as the genes and biological processes related to such SNP, are showed in Table 2.

Figure 1 Manhattan plot showing the position of the SNP associated with the trait of LBP1. (significance threshold fixed to P<0.001) 

Figure 2 Manhattan plot showing the position of the SNP associated with the trait of LBP2. (significance threshold fixed to P<0.001) 

Table 2 Genes and biological processes related to the SNP associated with the traits of LBP1 and LBP2 in breeding sows positive and negative to PRRS 

Trait	SNP	Variant	Associated gene	Associated biological process
LBP1	rs81276080	Intronic	TTR KCNQ4 CTPS1 FLRT2 ISOC1 ADAMTS19 IRAK4 ADAMTS20 CD83 CWH43 FAT4	Hormonal activation Neuronal transmission Immune response Neuronal development Oogenesis Enzymatic activation Immune response Enzymatic activation Immune response Cellular activation Neuronal transmission
	rs81334603	Intergenic
	rs80947173	Intergenic
LBP2	rs81364943	Intergenic
	rs80859829	Intergenic
	rs80895640	Intergenic
	rs80893794 rs81245908	Intronic Intergenic

LBP1= Live-born piglets at 1^st birth; LBP2= Live-born piglets at 2^nd birth.

Discussion

The negative effect of the PRRS virus infection on the number of live-born piglets observed in the present study has been previously reported in sows from different parity^22,23. In a similar study conducted with PRRS infected sows which were compared to healthy sows, it was observed an increase in the average values of mummified and non-born piglets of 0.04 to 1.12 and 0.62 to 1.02, respectively, as well as a reduction in the number of live-born piglets of 10.3 to 9.8²⁴.

Furthermore, the existence of genetic variability associated to the reproductive performance in sows infected with the PRRS virus has been also described in previous research reports. In this regard, Rashidi et al¹⁵ reported a variation of 3.83 ± 0.31 in the number of live-born piglets infected with the PRRS virus, compared to a variation of 1.96 ± 0.06 observed in healthy sows. Such variability, mainly in sows infected with PRRS, suggests the existence of a genetic basis associated to the reproductive response in the face of the disease. Therefore, it has been pointed out as an important strategy to reduce the negative impact of the PRRS in breeding sows, the identification of molecular markers that allow a better understanding about the genetic control of the response to the virus, which would be eventually incorporated in marker assisted selection (MAS) programs²⁵.

In the present study, the genome-wide analyses identified genomic regions and a total of eight SNP associated to the variables LBP1 and LBP2 (P<0.001) in breeding sows infected and non-infected with the PRRS virus. Such SNP showed a minor allele frequency greater than 10 %, which in general terms is considered as a requirement to avoid false results in genotype to phenotype association studies²⁶.

In the individual statistical analysis, the eight SNP previously identified were validated as predictors for the variables LBP1 and LBP2, from which only two of them are located within functional gene regions. On the one hand, the SNP rs81276080 (associated with LBP1) is located within the 5’region from the TTR gene (Transthyretin), which codifies the synthesis of a transport protein of thyroid hormones in plasm and cerebrospinal fluid. The TTR gene has been proposed as a potential candidate gene associated with the physiological response in pigs exposed to heat stress, which seriously restrict their reproductive performance²⁷, because it reduces oocyte quality and embryo viability, and enhances the PRRS negative effects on fertility of infected sows. On the other hand, the SNP rs80893794 (associated with LBP2) is located within an intron region from the gen CWH43 (Cell wall biogenesis 43 C-terminal homolog). This gene is involved in lipids remodeling of the cell wall from yeasts²⁸, and it has been reported that this gene is homologous to the gene PGAP2 (Post-GPI attachment to proteins), which is involved in both di-acetylation and re-acetylation cycles of the proteins that synthesizes Phosphatidyl Inositol (PI) in mammal cells²⁹. The PI is a family of lipids that participates in the second messenger mechanism in the cell membrane. This mechanism is used by several hormones such as PGF2α which plays an important role in ovary function and uterine activity; then, it influences directly the variables LBP1 and LBP2.

The remaining six SNP are located within intergenic regions (positional). In this regard, is important to point out that, when exploring the entire genome, it is complicated to detect a causal variant or a variant directly responsible of phenotypic changes within populations; however, because of the property of linkage disequilibrium in the genome, it is possible to identify indirect associations among the SNP and specific phenotypes. This information supports the importance for considering genes whose chromosomal location is close to resulting significant SNP that possess an intergenic position (at least in a range of 100 thousand of base pairs; 100 kbp)³⁰. One of these SNP is the rs81334603 associated to LBP1, which is located to a distance of 40.26 kbp from the gene KCNQ4 (Potassium voltage-gated channel subfamily Q member 4). This gene is under-expressed in tracheobronchial lymph nodes from pigs infected with the North American strain VR-2332 of the PRRS virus³¹. Moreover, approximately to 72.77 kbp from the SNP rs81334603 is located the gene CTPS1 (CTP Synthase 1), which codifies the production of the enzyme CTP synthase; the function of this enzyme is the biosynthesis of pyrimidine nucleotides (UTP and CTP), as well as the synthesis of ciclopentenilcitosine, a wide-spectrum antiviral agent³².

In relation to the SNP rs80947173, also associated to LBP1, the closest gene to this SNP is the FLRT2 (Fibronectin leucine rich transmembrane protein 2) gene which is located at 889.05 kbp of distance. Even though it is true that this gene is located considerably far away from the SNP rs80947173, useful levels of linkage disequilibrium (>0.3) appear to extend in pigs to a higher distance than Holstein cows, which implies that low-density SNP panels could provide reliably results in genome-wide association studies³³. In addition, it has been reported that the gene FLRT2 is associated to the number of live-born piglets in populations of Large White and Landrace pigs³⁴.

In the case of the SNP associated to LBP2, one of them is the rs81364943, which is located to a distance of 88.74 kbp from the gene ISOC1 (isochorismatase domain containing 1) and 178.67 kbp from the gene ADAMTS19 (ADAM metallopeptidase with thrombospondin type 1 motif 19). The gene ISOC1 has been linked to a processes of catalytic activity in porcine oocytes according to a genetic co-expression study³⁵, whereas polymorphisms from the ADAMTS19 gene in women have been associated to the presence of the polycystic ovarian syndrome³⁶.

Another SNP associated with LBP2 was the rs80859829 which is located to 31.75 and 177.73 kbp from the genes ADAMTS20 (ADAM metallopeptidase with thrombospondin type 1 motif 20) and IRAK4 (Interleukin 1 receptor associated kinase 4), respectively. The possible explanation for the association of this polymorphism with the reproductive performance at birth in PRRS infected sows is because the gene ADAMTS20 is over-expressed in organs such as brain and gonads³⁷, which are affected after the infection of the PRRS virus³⁸. Likewise, the kinase associated with the interleucine-1 receptor (protein product from the gene IRAK4), it has been involved in the mechanisms of PRRS virus replication because its production is reduced by the action of a well-known micro RNA (miRNA-373) with pro-viral effects³⁹.

The SNP rs80895640 also associated with LBP2 is located to 34.3 kbp from the gene CD83 (CD83 molecule); this gene has been previously linked with the number of total born piglets and live-born piglets in hybrids pigs Iberic X Meishan^40,41. Interestingly, in the immunological context, the CD83 molecule has been recently indicated as a key piece of the scape mechanisms of the PRRS virus against the immune system, because this virus is able to regulate positively the expression in soluble form of the molecule CD83 (sCD83), which was associated to the immunosuppression of T-cell proliferation in the host⁴². Similarly, from 54.05 kbp of the SNP rs80895640 is located the gene RNF182, which is involved in neuronal apoptosis processes⁴³ that commonly occur after the infection of highly-infectious strains of the PRRS virus⁴⁴.

Finally, the SNP rs81245908 (associated with LBP2) is located to an approximate distance of 53.52 kbp from the gene FAT4 (FAT atypical cadherin 4). The association of this polymorphism could be explained from a study which provides evidence that FAT4 gene expression in humans has been detected in fetal and infant brain tissues⁴⁵, because it has been also proved that PRRS maternal infection is able to affect neuronal development in piglets, reducing the number of neurons from the hippocampus and increasing the number of glia⁴⁶.

Conclusions and implications

The detection of genomic regions that explain 3.6 and 4.2 % of the variance associated to the traits live-born piglets at first and second births, as well as the validation of eight SNP located within such regions, suggest the existence of a genetic basis that underlies the reproductive response in breeding sows infected with the PRRS virus. Therefore, this study proposes to consider these eight SNP associated with the variables LBP1 and LBP2, two functional and six positional, as genetic markers for selection programs focused to improve the reproductive efficiency in sows infected with the PRRS virus. It is suggested to conduct additional studies to evaluate the functionality of the detected SNP; in addition, it is important to consider the validation of the genomic regions and genes reported in this study in other breed populations of sows also infected with the PRRS virus.