Quantitative Trait Loci Mapping for Porcine Backfat Thickness

A partial genome scan using porcine microsatellites was carried out to detect quantitative trait loci (QTL) for backfat thickness (BFT) in a pig reference population. This population carried QTL on chromosomes 1, 13 and 18. The QTL on chromosome 1 was located between marker loci S0113 and SW1301. The QTL corresponded to very low density lipoprotein receptor gene (VLDLR) in location and in biological effects, suggesting that VLDLR might be a candidate gene. The QTL found on chromosome 13 was found between marker loci SWR1941 and SW864, but significance for the marker-trait association was inconsistent by using data with different generations. The QTL on chromosome 18 was discovered between markers S0062 and S0117, and it was in proximity of the regions where IGFBP3 and GHRHR were located. The porcine obese gene might be also a candidate gene for the QTL on chromosome 18. In order to understand genetic architecture of BFT better, fine mapping and positional comparative candidate gene analyses are necessary. (Asian-Aust. J. Anim. Sci. 2002. Vol 15, No. 7 : 932-937)


INTRODUCTION
Backfat thickness (BFT) is an important trait to the pig industry because of consumer demands for lean pork.The BFT is highly heritable (0.51-0.61) and genetically correlated with other fatness traits such as intramuscular fat content (Warris et al., 1990).
The objective of this study was to search for QTL that contributed to BFT by a partial genome scan in a pig reference population.Six chromosomes (1, 2, 11, 13, 17, and 18) were examined.

Reference population
The pig reference population used in this study was initiated with two founder pigs.They were a Landrace boar and a Yorkshire sow.The backcross and intercross progeny were produced by mating the animals from the founder and F1 generations.Subsequent generations were produced by advanced backcross and sib-mating.A total of 253 pigs were included in this study.BFT was measured for each individual at slaughter.

Microsatellite analysis and genotyping
Genomic DNAs were isolated following a conventional method (Sambrook et al., 1989).Sixty five microsatellite markers on chromosomes 1, 2, 11, 13, 17 and 18 were used in the PCR amplification, of which 47 were polymorphic in the founder generation (table 1).These microsatellite markers gave a reasonable coverage of these 6 chromosomes, which belonged to one of three types of swine chromosomes.Chromosomes 1 and 2 were submetacentric, 11 was metacentric, and 13, 17 and 18 were acrocentric.
Each pooled sample representing 0.4 µl of PCR products was heated to 95°C after adding 1.1 µl of internal standard-dye-formamide mixture (0.25:0.25:0.60).PCR products were loaded on 4.25% polyacrylamide denaturing sequencing gel using an ABI PRISM TM 377 DNA sequencer (Perkin-Elmer, USA).The size of the PCR products was analyzed using the GENSCAN and GEMMA software (Applied Biosystems, USA).

Statistical analysis
The linkage maps of frame markers were constructed using CRIMAP (Green et al., 1990).First, the TWOPOINT option was used to find linkage between markers with LOD scores larger than three.Then, the BUILD option was used to construct the framework map and the remaining markers were incorporated using the ALL option.Finally, the marker genotypes were checked using the CHROMPIC option.Parents with two or more double recombinants between adjacent marker loci with an interval less than 20 cM were re-examined, and unidentifiable recombinant genotypes were coded as unknown.Marker genotypes inconsistent with the pedigree were also re-examined, and unidentifiable marker genotypes were coded as unknown.The individuals with genotypes inconsistent with the pedigree information at two or more marker loci were all removed.
The QTL analysis was carried out using Sequential Oligogenic Linkage Analysis Routines (SOLAR) for Linux system (Almasy and Blangero, 1998).The analytical model included sex and age effects.The Two-Point Analysis was implemented using a user-defined script to produce LOD score for each marker locus.Marker-specific identity-bydescent (IBD) matrix was computed using the Monte Carlo method.A genome-wide significance threshold was empirically obtained by permutation test (Churchill and Doerge, 1994).5,000 replicates were used to get the threshold value at significance level of 0.05.

Linkage map
The linkage analysis assigned the 47 polymorphic microsatellite markers into six linkage groups with the total length of 622.2cM (table 1).These linkage maps were basically in agreement with USDA-MARC pig map database (http://www.genome.iastate.edu/pig).

QTL detection
We found QTL segregated on porcine chromosomes 1 and 18 in this pig population (figure 1A and F), and they were referred as QTL1 and QTL3.Additional evidence for QTL affecting BFT was also found on chromosome 13 (figure 1D), but the marker-trait association was not consistently significant by using data with different generations.This was referred as QTL2.No QTL was significant in the other chromosomes (figure 1B, C and E).The QTL mapped on chromosome 2 by Jeon et al. (1999) and Nezer et al. (1999) was not found in the current study.The QTL might not be segregated in this pig population.On the other hand, no QTL for BFT on chromosome 17 discovered in the current study confirmed the results from Wu et al. (2000).They found a QTL on the chromosome for growth trait, but not for fatness trait.

QTL on chromosome 1
The QTL1 was located between the marker loci S0113 and SW1301 on chromosome 1, and it was a region of 58.8cM in the linkage length (table 2).The largest LOD score was found (9.39) at marker locus S0113, and the second largest was 8.10 at marker locus SW974 (figure 1).This QTL1 concurred with QTL for BFT found in other   Furthermore, this QTL1 corresponded with a QTL for growth found by Paszek et al. (1999).He discovered the QTL for post-weaning average daily gain between markers SW373 and SW1301.This correspondence might suggest that they were the same QTL, yet further investigation was in order.The influence of this QTL1 on growth and fatness might contribute to the phenotypic correlation of those traits observed in the field data.The QTL for growth and body composition possibly affected metabolism or energy partitioning.It implied pleiotropic effects of the QTL.Understanding the pleiotropic effects of a QTL ought to be a clue to the biochemical pathway, and would provide valuable insights into selection of positional candidate genes.This region was homologous to a QTL on HSA9q of human chromosome 9 (Goureau et al., 1996).
The dominance effect of this QTL was suspected because the phenotypic mean (12.67 mm) for heterozygous genotype at the marker S0113 was smaller than those (14.10 mm and 12.90 mm) for homozygous marker genotypes.A large effect of the QTL1 for BFT suggested that it might be of great importance in genetic improvement of pigs.
Recently, very low density lipoprotein receptor gene (VLDLR) was physically mapped to 1q26 on porcine chromosome 1 (Pinton et al., 2000).The VLDLR might be a candidate gene for this QTL because the QTL1 corresponded to VLDLR in both location and biological effects.

QTL on chromosome 13
The QTL2 was found between marker loci SWR1941 and SW864 (table 2).The LOD score for the marker region ranged from 1.85 to 2.47 (p<0.05) using the first five generations of the pig population, but it decreased to below 1.60 (p>0.05) using the whole population.The lack of significance with later generations might be due to the bias caused by inbreeding depression.The accumulative inbreeding in later generations tended to change mean and variance for each genotype.As the result, it might shrink or enlarge the QTL effect.Yu et al. (1999) reported a QTL affecting backfat in proximity of PIT1, an essential transcriptional regulatory factor of growth hormone, prolactin, and thyrotropin b subunit (Ingraham et al., 1990;Radovick et al., 1992).The QTL2 was located 20-60 cM apart from the QTL found by Yu et al. (1999).

QTL on chromosome 18
The QTL3 was located between the marker loci S0062 and S0117 on chromosome 18 (table 2).LOD scores ranged from 4.80 to 7.34 (p<0.05) at the region (figure 1).This concurred with the study of Wu et al. (2001) where estimates were obtained by a single marker analysis using a mixed model.The QTL3 was in proximity of the regions where some growth hormone pathway genes were located.One was the insulin-like growth factor binding protein 3 (IGFBP3) mapped with the regional assignment of SSC18q24 (Lahbib-Mansais et al., 1996).Another was the growth hormone releasing hormone receptors (GHRHR) mapped to the same region with strong linkage to markers S0062 and S0120 (Sun et al., 1997).Effective and regulated expression of the growth hormone pathway is essential to maintain homeostasis of fat metabolism (Cogan and Phillips, 1998).Furthermore, the porcine obese (leptin) gene on chromosome 18 (Sasaki et al., 1996) might be a candidate gene for QTL3.In human chromosome 7 (HSA7), GHRHR and IGHBP3 were close to each other (Gaylinn et al., 1994), and they were localized in the homologous region of porcine chromosome 18.
A hypothesis for the effect of QTL3 on porcine chromosome 18 was established with a mutation model.The allele Q linked to marker alleles B and D was normal, and the allele q linked to marker alleles A and C was a mutant responsible for physiological lipase deficiency favoring obesity.The A, B, C, and D referred to the 4 alleles at S0120, and their numbers of amplified base pairs were 153 bp, 161 bp, 165 bp, and 171 bp, respectively.Individuals carrying qq were observed with increased fat deposition.On the other hand, the individuals carrying QQ or Qq were normal because the normal allele was able to code the normal mRNAs which in turn were responsible for synthesizing the normal lipase.This hypothesis agreed with the study of Taylor and Phillips (1997) where the expression of monogenic obesity was mostly recessive.From this hypothesis, although performance records were not available for the founder animals, the founder sow carrying the allele A and C was suspected to have a high content of fat.

Future directions
QTL mapping is a major step toward identifying and positional cloning of causative genes affecting quantitative traits.It is also important for effective introgression and marker-assisted breeding.The current results of QTL would add to understanding genetic architecture of porcine BFT.

Table 1 .
Heterozygosity of microsatellite markers used in QTL mapping 1 Koning et al. (1999)98)een the loci S0113 and SW974 included the region of a QTL suggested byRohrer and Keele (1998), and it was close to a QTL reported by deKoning et al. (1999).

Table 2 .
QTL for backfat thickness in pigs 1 Further research will be aimed at fine mapping of the regions found in this experiment and positional comparative candidate gene analysis.Eventually, it will lead to the characterization and isolation of the genes.