Introduction

Close association to host surface is an essential requirement for successful induction of disease by a broad variety of pathogens. This association is often mediated by lectin recognition of host cell glycoconjugates (Karlsson 1998). Breeding for genetic resistance is an established routine in plant sciences and has become an attractive goal in animal sciences. Thus, selection for absence of receptor molecules or for genetic variants of receptor molecules to which the pathogen cannot bind would represent an attractive route to genetic resistance.

The two most economically dominant porcine enteric E. coli infections comprise diarrhea induced by diverse enterotoxins and enterotoxemia mediated by a shigatoxin. Both diseases have in common that the bacteria have to colonize the small intestine in high numbers where they produce the deleterious toxins. To build up huge populations, the bacteria are forced by the short transit time of ingesta to adhere to the small intestine’s inner surface. This adhesion is the result of a highly specific interaction between fimbriae produced by the bacteria and receptors present exclusively on the brush border of small-intestine enterocytes (Bertschinger and Fairbrother 1999). Of the five pig-specific types of fimbriae, designated F4, F5, F6, F18, and F41 (Osek 1999), so far it is known that only two, F18 and F4, are highly prevalent. Antigen variants have been detected for both types of fimbriae. Whereas the two variants F18ab and F18ac recognize the same receptor (Rippinger et al. 1995), the three variants of F4 vary in their receptor specificities (Bijlsma et al. 1982). However, F4ac is by far the most prevalent variant all over the world with the exception of central China, where F4ad is reported to be predominant (Wang et al. 2006). Fortunately, the causative mutation for E. coli F18 susceptibility is known (Meijerink et al. 1997, 2000) and elimination of the E. coli F18 susceptibility allele from the porcine population is currently done in Switzerland and other countries to increase the number of resistant piglets in commercial populations (Luther et al. 2009).

In pigs the resistant and susceptible phenotypes for adherence or infection by E. coli with F4ac fimbriae are inherited as a monogenetic trait, the susceptibility allele being dominant over the resistance one (Gibbons et al. 1977). The receptor for F4ac or a closely linked locus binds F4ab as well (Jørgensen et al. 2003; Python et al. 2002). As yet, attempts to develop practical methods for identification of breeding stock that inherit resistance have not been completely successful. Several investigators have mapped the ETEC F4ab/F4ac receptor gene (F4bcR) to SSC13 (Edfors-Lilja et al. 1995; Python et al. 2002). An overview of the marker positions and mapped genes encompassing the F4bcR is given in Fig. 1. Subsequent studies mapped the ETEC F4ab/F4ac susceptibility locus to the interval SW207-S0075 refined in a 5.7-cM region around the F4bcR locus (Joller et al. 2009). Jørgensen et al. (2004) reported an association between resistance and a mutation in intron 7 of the mucin 4 (MUC4 g.8227 G > C) gene. However, data presented later by Rasschaert et al. (2007) as well as observations in Switzerland (Joller 2009) raised doubt as to whether this mutation in the MUC4 gene is in some rare cases not in complete linkage disequilibrium (LD) with the F4ac receptor locus.

Fig. 1
figure 1

Location of 59 markers in the F4bcR candidate region at SSC13. The gene order and the scale are deduced from the Sscrofa9 assembly. The positions of the microsatellites are shown to the right of the chromosome 13 idiogram. The approximate positions of candidate genes are indicated at the far right side. The number of markers in each gene is given in parenthesis. The position of the six SNPs ALGA0072075, ALGA0106330, MUC13-226, MUC13-813, DIA0000584, and MARC000691, which were in complete LD, are shown in bold

Full size image

We recently discovered a Large White boar that carried a recombinant chromosome in the interval between F4bcR and MUC4. The aim of the study was to characterize the recombinant haplotype and also to discover more accurate markers for F4ab/F4ac resistance. Therefore, we tested a total of 59 markers (53 SNPs and 6 microsatellites) in a three-generation pedigree. The data indicate that the F4bcR receptor locus is distal to the MUC4 locus.

Materials and methods

Pigs

The animals used in this study originated from a Swiss experimental herd (SEH) at the University of Zurich comprising Large White, Landrace purebred pigs, and Large White/Landrace crossbreds. The boar 698B used for breeding purposes was found to be homozygous for the susceptibility allele at the F4bcR locus but heterozygous in ZDHHC19 g.4043, TNK2 g.7075, TNK2 g.7717, TNK2 g.11142, MUC4 g.8227, and KIAA0226 g.62250, which were so far in complete LD with the F4bcR alleles (Jacobsen et al. 2010). To be sure that 698B was homozygous susceptible to ETEC F4ab/F4ac, it was mated to three resistant sows generating a total of 45 offspring. The homozygosity for the resistant allele of the sows was proved by previous matings to resistant boars and confirmed by the microscopic adhesion test after slaughtering.

To precisely define the location of the recombination event, a three-generation family was generated, including boar 698B, its parents (boar 381B and sow 405B), two siblings of boar 698B (9499 and 9486), and its offspring (1977, 2122, 2731, and 2734) (Fig. 2).

Fig. 2
figure 2

Diagram of the three-generation family tree and haplotypes of eight pigs in the F4bcR region. SNP names and positions on SSC13 are given at the right side. The animal identity is shown on top of the haplotypes. The SNP genotypes are depicted on the right side of the colored columns. The digit 1 corresponds to nucleotide A, 2 to C, 3 to G, and 4 to T. Microsatellite markers are represented by 5, 6, etc., depending on the sizes of the allelic bands. S on top of the colored bars indicates the adhesive and s the nonadhesive haplotype. Each haplotype has its own color. The lines indicate the relationships between the pigs. The geometric figures above the animal identity indicate the sex, squares for males and circles for females. Filled figures indicate F4ab/F4ac susceptible and blank figures resistant pigs

Full size image

Sampling of intestinal tissue and microscopic adhesion test

Pigs (n = 53) were slaughtered at about 2 months of age, at 150 days of age, or when they were eliminated from breeding. Samples of blood and intestine were taken at that time. The interval between killing and sampling of intestine was between 10 and 30 min to avoid degeneration of the epithelial intestinal cells.

All animals were phenotyped with the antigenic variant F4ac fimbriae by the microscopic adhesion test according to the method of Vögeli et al. (1996). A more detailed description of the preparation of bacterial strains, sampling of intestinal tissues, purification of enterocytes, and the procedure of the microscopic adhesion test can be found in Joller (2009). In brief, starting at the upper mesentery root, the intestine was separated from the mesenterium and an empty segment in the middle part was taken anywhere between 3.5 and 7.5 m in young pigs and between 5 and 10 m in older pigs. Twenty intact enterocytes were scored under an optical microscope by the same person. A pig was classified as adhesive to F4ac if more than 15% of the enterocytes bound more than five bacteria on the brush borders.

Screening for genetic polymorphisms

Genomic DNA was extracted from EDTA-anticoagulated whole blood as described by Vögeli et al. (1994). A total of 59 markers on chromosome 13q41 were used to characterize the three-generation pedigree (53 SNPs listed in Tables 1 and 2) along with 6 microsatellite markers: SW207, MUC4GT, S0283, S0075, SW1876, and SW698. Twenty-nine markers from Table 1 were selected from Jacobsen et al. (2010), nine SNPs in MUC13 were taken from Zhang et al. (2008), and LMLN g.15920 was detected in this study by BLASTing the human sequence (NM_001136049) to the pig genome and by sequencing the genomic porcine homolog. The 14 SNPs listed in Table 2 were analyzed using the Illumina PorcineSNP60 BeadChip (Illumina, Inc., San Diego, CA, USA) (Ramos et al. 2009). The microsatellite markers were selected from Python et al. (2002).

Table 1 Description of the SNPs
Full size table
Table 2 SNP list from the PorcineSNP60 BeadChip
Full size table

PCR was carried out in a reaction volume of 25 μl containing 100 ng of genomic DNA, 10 mM Tris-HCl (pH 9), 50 mM KCl, 1.5 mM MgCl2, 200 μM of each deoxynucleotide, 0.4 μM of forward and reverse primers, and 1.25 U of Taq DNA polymerase. The annealing temperature was chosen according to the melting temperature of the primers. For microsatellites, the PCR consisted of 25 cycles and for the SNPs 35 cycles followed by a final extension at 72°C for 2 min. Thermal cycling conditions were incubation at 94°C for 2 min followed by incubation at the specific annealing temperature for 30 sec and extension at 72°C for 30 sec. The subsequent denaturation was at 94°C for 30 sec. DNA samples were then purified with Montage PCR Centrifugal Filter Devices (p36461) (Millipore, Billerica, MA, USA) and sequenced. Alleles of the microsatellites were analyzed on an ABI377 DNA analyzer (Applied Biosystems, Foster City, CA, USA). Haplotypes were determined using MERLIN (Abecasis et al. 2002) and HaploPainter software (Thiele and Nürnberg 2005).

Results

The distribution of the 59 markers in the candidate region of the F4bcR locus is shown in Fig. 1, and the approximate positions of genes and markers are according to the Sscrofa9 assembly and the physical map (http://pre.ensembl.org/Sus_scrofa_map/). These markers cover the 35-Mb interval from microsatellite SW207 to SW698. Fifty-five markers map to the core interval 100–104 Mb on chromosome 13.

Boar B698 was mated with three sows resistant to ETEC F4ab/F4ac and hence putative homozygotes for the absence of an adherent ETEC F4ab/F4ac receptor, generating a total of 45 offspring. All of the offspring were adhesive to ETEC F4ab/F4ac according to the microscopic adhesion test. Thus, all the offspring are presumed to have received (from boar B698) a susceptibility allele encoding the presence of a receptor to which E coli F4ac can adhere. Therefore, it can be deduced that boar B698 is homozygous for the susceptibility allele. However, the genotyping data showed that 23 piglets (51.1%) possessed the MUC4 g.8227 C/C genotype, which was previously reported to be associated with the absence of F4ab/F4ac adhesion (Jacobsen et al. 2010; Jørgensen et al. 2004), and 22 piglets were C/G (48.9%). Haplotype analysis of the genotype data reveals a putative recombination event between LMLM g.15920 and ALGA0072075 SNP marker loci (Fig. 2). The recombinant chromosome was transmitted to 23 of B698’s offspring. The recombinant chromosome was identical to all other susceptible haplotypes for the interval ALGA0072075–SW698. The LMLN–S0075 region was also covered with 11 SNPs of the PorcineSNP60 BeadChip (Table 2), 2 microsatellites, and 14 SNPs in MUC13. In contrast to the MUC4 markers, ALGA0072075, ALGA0106330, MUC13-226, MUC13-813, DIA0000584, and MARC0006918 were all in complete LD with F4bcR in our family (Fig. 2).

Discussion

Previous studies identified MUC4 as the most probable candidate gene for F4bcR based on its chromosomal location and evidence that the molecule to which the E. coli F4ac adheres is a mucin-like sialoglycoprotein (Erickson et al. 1994; Jacobsen et al. 2010; Joller et al. 2009; Jørgensen et al. 2004). As reported earlier, the resistance allele is presumed to represent the ancestral allele and is associated with greater haplotype heterogeneity (see 798B and 739B, Fig. 2) (Jacobsen et al. 2010). The large shared haplotype block associated with the susceptibility allele handicaps efforts to locate and identify the causative mutation. In addition, determination of the F4bcR genotypes is time-consuming as it has to be done by progeny phenotyping which leads to the loss of potential breeding animals. Thus, when we detected a boar with a recombination between the MUC4 markers and F4bcR, it represented an opportunity to refine the search for the causative mutation. Therefore, we mated this boar to homozygous resistant sows and analyzed more markers distal to MUC4.

Boar 698B inherited a recombinant haplotype from its father, boar 381B. The recombinant allele was transmitted from 698B to half of its progeny, as seen in pigs 1977 and 2734 (Fig. 2), and was not found in other pigs from the same herd. The recombination in 698B occurred in an interval of about 220 kb, between SNPs LMLN g.15920 and ALGA0072075. Boars 381B and 698B were found to be homozygous for the SNP MARC0067282, making it impossible to determine the position of the crossing over in 698B more precisely.

These results support the hypothesis that the locus for F4bcR is located distal to MUC4. Besides the experimental pigs, we phenotyped 78 unrelated pigs of 38 litters of the Swiss performing station and determined their MUC4 g.8227 genotype (Joller 2009). Five pigs were phenotyped as resistant, whereas the genotype was MUC4 g.8227 C/G and one susceptible pig was MUC4 g.8227 C/C. These pigs were typed with the markers used in this study and the resistant pigs were homozygous A/A in ALGA0106330, G/G in MUC13-226, and C/C in MUC13-813, consistent with a resistant genotype. The susceptible pig was heterozygous A/G, A/G, and C/T, respectively, for the same markers. These results suggest stronger LD between these markers and F4bcR than between MUC4 g.8227 and F4bcR and give further evidence that F4bcR is located distal to MUC4.

BLAST analyses of the markers that were in complete LD with F4bcR showed that only MUC13-813 was located in an exon. However, this C > T transition is silent. ALGA0072075 was mapped between SLC12A8 and HEG1. ALGA0106330 was located in an intron of MUC13. DIAS0000584 was mapped to an intron of gene KALRN and MARC0006918 to an intron of MYLK. These findings indicate that these markers may not be causative for ETEC F4ab/F4ac adhesion.

Joller et al. (2006) had initially mapped the most probable position of the F4bcR locus in the SW207–S0283 interval. Unfortunately, the family in this study did not allow the identification of a distal border closer to the unknown F4bcR than microsatellite S0283. Therefore, we propose the interval between LMLN g.15920 and S0283, consisting of around 620 kb, to be the most probable region for F4bcR.

Mucin-type proteins and lipids are present on almost every epithelial tissue and are important for defense against pathogens. However, mucins and other glycosylated structures are also targets for microbial attachment as they contain a variety of oligosaccharide structures providing binding sites for bacteria. In the case of bacteria expressing the F18-type fimbriae, the adhesion is dependent on the activity of fucosyltransferase 1 and the presence of a yet unknown fucosylated structure (Meijerink et al. 2000). Although the causative mutation in F4ac is still unknown, the genetic analyses performed in 698B’s family have determined that the causal mutation for ETEC F4ab/F4ac susceptibility is downstream of the gene LMLN and probably may be located around the region containing the MUC13 gene.