Trait gene mapping in a cross between two divergent pig breeds - Duroc and Mangalica pigs
1997 Report

List of partners:

Scientific progress report

The Mangalica breed was developed in Hungary in the first half of the last century for lard production. Numbers of the Mangalica breed have declined sharply since the Second World War in response to the demand for lean meat. In 1936 86% of the 5 million swine population in Hungary was Mangalica, declining to 5% in 1950. The move to lean meat production has been effected by breed substitution with the Mangalica being replaced by Landrace and Large White pigs from Western Europe. Thus, the Mangalica is now threatened with extinction and its desirable alleles for other aspects of pig production could be lost. In this project we are primarily concerned with using the Mangalica to understand the genetic control of fatness in pigs. This greater understanding will allow the pig breeding industry to develop informed and flexible responses to changes in consumer demand for pig meat. The Duroc breed was developed in North America and is renowned for its high meat quality. Some of the Duroc’s reputation for meat quality may be associated with its intramuscular fat (or "marbled" meat). Thus, the fat in the Mangalica is predominantly found as backfat and in the Duroc as intramuscular fat. The former is currently considered undesirable, whilst the latter is perceived as the key to meat quality and cookability. Other potentially interesting differences between the Mangalica and the Duroc included prolificacy, growth rate, coat type and colour. The Duroc grows rapidly, but the Mangalica is late maturing. Whilst the Duroc’s prolificacy may not match that of the Min Chinese breeds it is better than that of the Mangalica (3-7 piglets per litter). The Mangalica has a "wooly" or "curly" dark coat whilst the Duroc is conspicuously red.

One of the reasons for establishing the Duroc/Mangalica cross was to develop a genotype which is suitable for extensive conditions, with grazing, in a low input environment, with low veterinary costs. The cross utilises the growth rate, the conformation and mainly the meat quality of the Duroc. The crossbred pigs will be sold on the Spanish market as a basis for the Serrano ham. The role of the Mangalica breed is to protect the ham from drying out during the 1-3 years long maturation process with its excessive intramuscular fat and the 2-3 cm depth of fat on the ham. The special nutrition, the management and the biochemical processes during the long maturation provide a flavour to the ham that is popular on the market and provides an extra profit for the producer. Our initial programme that was designed to protect the Mangalica from the extinction provided the base population for the present research started in 1997.

Recent experiments in pigs suggest that, contrary to the predictions of some sceptics, there are genes with significant effects on economically important traits in livestock. The rapidly improving maps of the porcine genome have provided the tools to locate these trait genes. For example, a region of chromosome 4 containing QTLs which control a large part of the variation in total body fatness in a Wild Boar x Large White cross (Andersson et al., 1994. Science 263:1771). QTLs influencing fat levels have also been found on chromosomes 4 and 7 in a Large White x Meishan cross (Walling et al., 1998: Proceedings of the 6th World Congress on Genetics Applied to Livestock Production, Armidale, NSW, Australia, 12-16 January, 1998 Vol 23, 519-522.)

The purpose of this research project is to build on the established maps of the porcine genome developed in the earlier EC-funded PiGMaP collaborations to locate chromosomal regions influencing traits of economic importance and to develop new mapping resources for the identification of trait genes. Such regions are termed quantitative trait loci (QTLs). We are seeking to identify QTL that control fatness, and hence product quality, in the pig. We are using a cross between two divergent pig breeds - Duroc and Mangalica - to study the genetic control of fatness.

The specific project objectives are -

• to identify QTLs for fatness (both total fatness and the level of intramuscular fat) in Mangalica x Duroc crosses.
• to develop library resources for exploring regions containing QTL, specifically a porcine adipose tissue cDNA library, a library enriched for CpG islands and a large fragment library in the BAC vector.
• to explore candidate genes identified on the basis of map location, function or comparative arguments for the causal molecular variation.

1.1 Establishing Mangalica x Duroc crosses - Mangalica x Duroc crosses have been established by partner 4 as three-generation F₂ intercrosses with the Mangalica and Duroc as the purebred grandparents. Briefly, a single Mangalica boar has been crossed to 20 Duroc sows to generate F₁ pigs. The F₁ pigs have been mated inter se to generate F₂ pigs.

For each pig body weight at the age of weaning (30-35 days), at 90 days, at 210 days and at the age of slaughter are measured. For the slaughtered pigs body weight after slaughter, fat depths were measured, killing out percentage, fat percentage were calculated and meat content was predicted.

After birth when the individual piglets are tagged the colour of the animal is recorded together with the sex. (e.g. dirty white, roan, light brown, light brown with stripes, full brown with stripes, black with stripes and grey with stripes). Records on colour are also taken at seven months of age.

Full details of this experimental pig population are included in the laboratory report from partner 4.

1.2 DNA sampling, storage and distribution - Blood samples have been taken from 244 pigs and DNA prepared by partner 3 and distributed to the other participating laboratories (partners 1, 2 and 5) (see Table 1).

Table 1: Blood and DNA samples

* most of the F2 pigs were sampled in November 1997, DNA preparation and distribution is in progress

1.3 Optimising genotyping methods and primer sets - In order to map quantitative trait loci it is necessary to genotype the pigs for genetic markers that provide good coverage of the porcine genome. Microsatellite markers are ideal for QTL mapping studies as they are highly polymorphic and can be relatively rapidly using the polymerase chain reaction (PCR) and automated fluorescent DNA fragment analysers (sequencers). Markers appropriate for genotyping the Mangalica x Duroc crosses are selected on the basis of their known map locations and the alleles present in the purebred Mangalica and Duroc grandparents.

Approximately 100-150 informative markers are required for the genome scan. Methods have been optimised for multiplex PCR, pooling PCR products, and gel electrophoresis on ABI 377 and ABI 310 fluorescent DNA fragment analysers by partners 2 (ABI 377), 3 (ABI 310) and 5 (ABI 310). The purchase of this specialised equipment (ABI 310) for partners 3 and 5 was funded in part through this EC INCO Copernicus grant. To date protocols have been established for ninety markers (see Table 2). The results are encouraging as about seventy-fine percent of the markers tested are informative in the Mangalica x Duroc F₁ males.

Table 2: Markers optimised for genotyping

QTL mapping studies will identify chromosomal regions that contain genes controlling traits of economic and biological significance. There are two strategies for the subsequent identification and isolation of such trait genes - ‘positional cloning’ or the ‘positional candidate gene approach’. We are developing the key resources necessary for implementing these strategies - large fragment genomic libraries and catalogues of (mapped) expressed sequences (or genes). DNA library resources for exploring regions containing QTL have been established including, a porcine adipose tissue cDNA library, a library enriched for CpG islands and a large fragment library in the BAC vector.

2.1 A porcine adipose cDNA library - An adipose cDNA library has been produced by partner 1 from adipose mRNA isolated from a Large White x Meishan F₁ female pig. The cDNAs were directionally cloned into Lambda Zap II and excised as plasmid clones in pBluescript. This library is an obvious source of candidate genes controlling fatness.

Clones representing abundant adipose tissue transcripts were indentified by screening the library with labeled total adipose cDNA. Two hundred such abundant clones and two hundred clones picked at random were subjected to single pass sequencing from their 5’ ends. The resulting DNA sequences were used to search the EMBL, Genbank and ESTdb databases using BLAST. The initial results of these searches are summarised in Table 3.

Table 3: Classification of sequenced cDNA clones

2.2 A porcine CpG island library - A porcine genomic library enriched for CpG island sequences has been established (McQueen et al., 1997). Such a library provides a valuable complementary resource to a cDNA library. cDNA libraries only represent genes being expressed in a particular tissue at the time of production, and even in normalised libraries some rare transcripts may be missed. Around 60% of genes in man are associated with CpG islands, including all housekeeping genes and 40% of tissue-specific genes. They thus provide a useful adjunct to cDNA libraries in giving access to a large number of genes independent of developmental stage. The library is available through the UK HGMP Resource Centre, Hinxton, Cambridge CB10 1SB, UK (for details see –

URLs: http://www.hgmp.mrc.ac.uk/Public/Docs/Bio/CpG_island_libraries.html

and http://www.hgmp.mrc.ac.uk/).

This library was produced in a parallel collaborative project by partner 1.

2.2 A porcine BAC library - A large fragment genomic library has been established by partner 1 in the bacterial artificial chromosome (BAC) vector (pBeloBAC11). One hundred and two thousand five hundred clones have been individually picked in duplicate into 384-well plates and stored at -70ºC. As the average size of the cloned fragments is 150 kilobasepairs this library represents about 5-fold coverage of the pig genome. The library will be transferred to the UK HGMP Resource Centre who will prepare DNA pools for PCR screening and gridded filters for screening by hybridization. These resources will be available through the Resource Centre by summer 1998.

Although the litter sizes in the Mangalica x Duroc population are within acceptable limits, the losses of pigs between birth and weaning has been particularly high in these cross bred litters. Thus, it will be necessary to reconsider the experimental design of the QTL mapping components of the project. For example, it may be necessary to switch from selective genotyping of extremes in the population to genotyping the entire three-generation cross.

The objectives for the second year of this project include -

• Generation of further F₂ pigs in the Mangalica x Duroc crosses
• Record performance of F₂ pigs in the Mangalica x Duroc crosses
• Genotyping microsatellite markers in order to effect a genome scan of the Mangalica x Duroc crosses
• Development of polymorphic genetic markers for candidate fat genes isolated from the adipose cDNA library

• A porcine adipose tissue cDNA library has been established
• Four hundred clones from the adipose cDNA library have been subjected to single pass DNA sequencing.
• A five-fold coverage large fragment genomic library has been established in a bacterial artificial chromosome (BAC) vector.
• A library enriched for porcine CpG islands has been established.
• A database has been established to hold the genotypes and pedigree data

Pig mRNA was isolated from adipose tissue from a Large White - Meishan F₁ pig and a cDNA library created using Stratagene's ZAP-cDNA Synthesis Kit. This procedure results in a library of cDNA inserts directionally cloned in pBluescript. This has three benefits: 1) it allows easy handling of the DNA as a plasmid, 2) Blue/white colour selection of recombinant clones, 3) The known orientation allows selection of appropriate primers for sequencing 5' and 3' ends.

Two hundred colonies were randomly selected plus another two hundred positives colonies, selected after a colony hybridisation using labelled total adipose cDNA as a probe. This latter group should represent abundant transcripts (mRNA) and therefore genes highly expressed in adipose tissue. The clones were sequenced from the 5’ end using T3 primer. The sequences were loaded into a programme which selected readable sequence, trimmed off vector sequence and then performed Blast searches (GenEmbl, Tags/EST, Fugu databases) with the data in order to identify homologous sequences and the identity of particular cDNAs.

After sequencing the first hundred "highly expressed" cDNAs it became apparent that there was a large percentage of mitochondrial (34%) and repetitive sequences (20%) represented. A second round of colony hybridisation was performed, this time probing replica filters with labeled genomic DNA. Colonies which lit up with both cDNA and DNA probes were assumed to be repetitive and those that hybridised only to the cDNA probe, the highly expressed genes. A further 100 colonies were picked from this group and sequenced.

This extra round of screening resulted in a drop in the number of repetitive sequences (3%) and the level of mitochondrially derived cDNAs (21%). When the level of these two types of sequences was compared to their representation in the 200 randomly picked clones (mit 14%, rep 6%) it appears that the selection is still preferentially picking out these types of sequences rather than highly expressed genes. This is confirmed further by the fact that there is not much difference in the percentage of adipose related genes identified (4%) compared with the random "low abundance" group (6%). "Adipose related genes" were identified as those that mention adipose/adipocyte/lipid/lipocyte in their names. Some of the other genes identified may be equally "adipose related". The unidentified portion of the selected clones could also contain highly expressed adipose-related genes.

The 103 cDNAs that have been assigned putative identities were sequenced from the 3' end using M13 Forward primer. These 3’ sequences that should correspond to the 3’-untranslated regions of these genes (cDNA) will be used for the development of polymorphic markers.

In all the cDNAs studied so far there were very few di/tri/tetranucleotide repeats, with only two larger than 6 repeating units ( [GTT]₁₀ and [TA]₉) although there were a lot of single polynucleotide runs. Primers are now being designed to amplify the 3' UTR of the adipose related cDNAs for SSCP analysis.

A large fragment genomic library for the pig in a BAC vector (pBeloBAC11) has been established. A library of 102,912 independent clones with an average insert size of 150 kilobasepairs (Kbps) has been established. As the size of the pig genome is estimated at 2.8 x 10⁹ basepairs (bps) such a library should provide about five-fold genome coverage. The cloned DNA was isolated from an F₁ Meishan/Large White boar from the Roslin QTL mapping populations. The clones were individually picked into 384-well microplates, replica plates made and both copies stored independently at -70ºC.

One complete copy of the library will be transferred to the MRC Human Genome Mapping Project Resource Centre on the Hinxton genome campus. Filters and PCR pools from library which has now been named the PigE BAC library (Edinburgh pig BAC library), will be made available through the MRC HGMP Resource Centre.

Six BAC clones have been subjected to prolonged culture equivalent to 100 cell generations in order to check the stability of the cloned material. These clones appear to be stable as verified by sizing restriction digest products.

A library enriched for CpG islands was prepared according to the procedures described by Cross et al., (Cross, S. H., Charlton, J. A., Nan, X. & Bird, A. P., 1994. Nature Genetics 6, 236-244.). Briefly, genomic DNA is cleaved with MseI. The MseI digested DNA is passed down a MBD (methyl binding domain) column. Heavily methylated DNA is retained on the column. The DNA in the column eluate is methylated in vitro and then passed down a second MBD column. The retained fraction represents DNA fragments with clusters of CpG (unmethylated in vivo, but now methylated experimentally). After repeated passing over the column the retained fraction is eluted in a salt gradient. The eluted fraction that is enriched for CpG islands is then cloned to constitute a CpG island library. The distribution of CpG islands throughout the pig genome was examined with fluorescent in situ hybridization in collaboration with INRA Toulouse. When viewed in the context of the known conservation of genome organisation between pigs and humans it is evident that the distribution of CpG islands is similar in both species. The pig CpG island library represents a valuable library of genomic fragments associated with genes. The library has been lodged at the MRC HGMP Resource Centre and is thus available to other research groups.

A generic database (RseSpecies) has been developed to hold data from linkage and QTL-mapping experiments. The database model supports the ongoing PiGMaP Linkage Consortium. This database model has also been implemented for this INCO-Copernicus project. Details of the animals, pedigrees and marker genotypes are stored. Data will be accessible to the project partners via the World Wide Web.

Dr. Alan L. Archibald (coordinator)

Dr. Judy F. Brown (post-doctoral scientist - molecular biology; funded through EC contribution to this project)

Dr. Andy Law (post-doctoral scientist - database)

Dr. Chris S. Haley (senior research scientist - quantitative genetics)

• Procedures have been established for semi-automated high throughput genotyping of microsatellite markers.
• Sixty different microsatellite markers have been screened for informativeness in the F1 sires.

The major task for the Uppsala group will be to analyse a large number of microsatellite markers. We have therefore optimised our methods for microsatellite analysis. We have established a method where we use an ABI877 robotic workstation for setting up multiplex PCR, for pooling PCR products, and for adding loading buffer and size markers before gel electrophoresis. Good specificity with a minimum of stuttering is obtained by using AmpliTaq Gold, Touch-down PCR and tailed primers. The fluorescently labeled PCR products are separated on sequencing gels using an ABI377 instrument. The procedure works very well in our hands and we are now set up for the genome scan to be carried out in 1998.

A set of 60 different microsatellite markers has been screened for informativeness using the three F₁ sires of the intercross pedigree. The individual genotypes are provided are listed in Table 2.1. The results are very good since as many as 54 markers were informative in at least one F₁ sire and a majority of markers were informative in all three sires.

Leif Andersson participated in the first project meeting in Hungary, March 1997

Leif Andersson, professor

Karin Sitte, post doctoral fellow, Feb-Apr (funded with EC contribution to this project)

Elisabetta Giuffra, July-Dec (funded with EC contribution to this project)

Table 2.1: Marker parameters and microsatellite genotypes for F₁ sires

Table 2.1 (contd): Marker parameters and microsatellite genotypes for F₁ sires

• DNA has been prepared from blood samples from the Mangalica x Duroc population.
• DNA has been distributed to partners 1, 2 and 5.
• Genotyping procedures have been optimised for the ABI 310 automated DNA sequencer.

We take part in four activities belonging to the work package 1 - Mapping QTLs:

1.3) DNA sampling, storage and distribution.

1.4) Optimising genotyping methods and primer sets.

1.5) Genome scanning to map QTL-s in Mangalica crosses.

1.7) Fine scale QTL mapping.

In 1997 we have made progress with activities in 1.3 and 1.4.

One of the major tasks of our group is DNA preparation from blood taken by partner 4. The quantity of anticoagulated blood samples was usually 15 ml. We tried simple salting out protocols for extracting DNA, instead of hazardous organic solvent extraction. First we tested the "WIZARD Genomic DNA Purification Kit" (Promega). The quality and quantity of DNA was similar to those obtained from phenol-chloroform extraction. In order to reduce the cost of the procedure, the simple procedure described by Miller et al 1988 was used with modifications. DNA extracted by this procedure yields on the average 20 µg of DNA from 300 µl blood, with 1.7 purity. We isolated DNA from small blood samples in order to be able to repeat preparation several times even from the irreplaceable samples. This method was economical, safe and rapid.

Table 3.1: Current data on blood and DNA samples:

* most of the F₂ samples arrived in November 1997, so we had not enough time to prepare DNA from these. We are going to send the second set of DNA samples with a part of these F₂-s in the near future.

There were several coagulated bloods among the samples mainly at the beginning of the experiment. Unfortunately there was no possibility to replace some of these by new blood samples because the animals were slaughtered meanwhile. For this reason we were looking for special DNA preparation methods from clotted blood. We applied the procedure using Chelex 100 (Walsh et al. 1991), but without success. This experiment will be repeated.

We have purchased in February 1997 an ABI Prism 310 Genetic Analyser (Perkin Elmer Applied Biosystems) which is an automated fluorescent DNA fragment analyser, suitable for microsatellite detection and genotyping.

Last year a project meeting was held in Hungary. Partners 1, 2, 3 and 5 divided the microsatellite sets available, to genotype in the four laboratories. Microsatellites suitable for coamplification were selected on the basis of similar reaction conditions and compatible allele size ranges. Our task is to genotype the following 9 microsatellites in three triplex reactions

Table 3.2: Chromosome location, size range and annealing temperatures for selected markers

Fluorescencent labeled primers were obtained from partner 1. Partner 2 suggested PCR component concentration and cycling parameters on the basis of their results using ABI 377 Sequencer. We used the following touchdown PCR cycling conditions for each locus, although SW951 and SW122 required higher annealing temperature compared to the other loci.

Table 3.3: PCR conditions

We found that the intensity of PCR products varied considerably within the triplex reactions. In order to obtain compatible signal intensities, we modified the suggested individual primer concentrations and also the MgCl₂ concentrations.

Table 3.4 : PCR reaction components were as follows:

Table 3.5: Final PCR conditions for selected markers

PCR products from the three triplex reaction were combined in a pool in the ratio of 1:1:1 (1 µl each), mixed with TAMRA-500 internal size standard and diluted before loading for electrophoresis. Examination of the electrophoretograms revealed that, the top of several peaks were missing, in spite of that we had used reduced primer concentrations. The reason for this is sample overloading, which causes spectral interference between the dye labels during analysis. Reducing the ratio of the TET-triplex (0.25 µl) in the pool and reducing the injection time from 5 sec - recommended by manufacturer - to 3 sec, prevented overloading, thereby eliminating sizing errors.

Using this modified protocol, we genotyped a small family form the Duroc x Mangalica cross - F₁ parents and 5 F₂ offspring - for the 9 microsatellites (Table 3.)

Table 3.6: Inheritance of marker types for Sw951, S0225, S0227

The amplified fragments correspond to the expected size ranges for the 9 loci. Loci labeled with the same dye were sufficiently different in size and resolved well enough so as not to overlap and cause difficulties in data interpretation. Genotyping was carried out by manual interpretation of each allele, having no Genotyper software yet.

Only one (S0228) microsatellite was homozygous out of the 9 loci examined in this family. It is very likely that even this locus will be informative in the genomes scan analysis. We would like to compare our results to those of the other participants involved in marker optimisation. The next project meeting will provide an opportunity of making these comparisons.

The capacity of the ABI Prism 310 is 48 samples daily. Using the optimised genotyping method the 244 samples could be genotyped within one week.

• Duroc - Mangalica crosses have been established
• Basic performance data has been recorded
• Blood samples have been taken and passed to partner 3 for the preparation of DNA

Background

The research started with the Duroc and Mangalica pig breeds. The Mangalica breed which is known for its intensive fat production, can be found mainly in Hungary.

The aim of the two-breed crossing was to develop a genotype that is suitable for extensive conditions, with grazing, in a low input environment, without veterinary cost. The cross utilises the growth rate, the conformation and mainly the meat quality of the Duroc. The crossbred pigs were planned to be sold on the Spanish market as a basis for the Serrano ham. The role of the Mangalica breed is to protect the ham from drying out during the 1-3 years long maturation process with its excessive intramuscular fat and the 2-3 cm depth of fat on the ham. The special nutrition, the management and the biochemical processes during the long maturation provide a flavour to the ham that is popular on the market and provides an extra profit for the producer.

Most of the F₁ pigs required for the three-generation crosses within the INCO-Copernicus project were generated prior to the start of the project. 144 inseminations were carried out on F₁ sows with F₁boars between the period of 01.09.1996 and 30.12.1997. These inseminations are yielding the necessary F₂ pigs.

Number of farrowings: (producing F₁ pigs)

Number of farrowings: (producing of F₂ pigs)

04.05.1996 – 22.11.1996 21

out of this from the "37" blond Mangalica boar 3

Number of litters in the research: 18

01.01. 1997– 15.12.1997 69

out of this that is not available for research 13

Number of litters in the research: 56

Total number of litters:

Genotype	year	litters
	1994	11
F1	1995	97
	1996	28
	1997	13
Total:		149

Out of these 44 F₁ sows are available for the research, 13 are not swallow bellied and are not used in the research.

Genotype	year	litters
F2	1996	182 + 16*
	1997	482 + 105*

* Not used in the research because the F₁ born from blond Mangalica boars.

Blood samples have been collected as follows and the samples passed to partner 3 for the preparation of DNA.

"14" swallow-bellied Mangalica

"47" Duroc x swallow-bellied Mangalica F₁

"76" Duroc x swallow-bellied Mangalica F₁

"77" Duroc x swallow-bellied Mangalica F₁

Body weight at the age of weaning (30-35 days), at 90 days, at 210 days and at the age of slaughter are measured. Survival rates according to litters are also recorded.

Pigs slaughtered

Body weight after slaughter and fat depths were measured, killing out percentage, fat percentage were calculated, meat content was predicted.

After the birth at the first treatment of the piglets when individual tagging is made the colour of the animal is recorded together with the sex. (e.g. dirty white, roan, light brown, light brown with stripes, full brown with stripes, black with stripes and grey with stripes). Records on colour are also taken at the age of seven months.

We can examine the number, and the weight of piglets born and weaned in the 72 farrowings. In the analysis of the prolificacy results concerning the negative effects of inbreeding, the F1 sow population was divided into three groups:

• like the littermates of the 47 F₁ boar (sows with the following ear notch 2(942503), 3(942509), 4(942508) and 5(942501)).
• The fullsibs of the 47 F₁ boar which were born of Duroc sow Id. no. 448 and Swallow-bellied Mangalica Id. no. 14, the 11 sow altogether. According to ear notch: 6(952515), 7(952516), 8(952517), 9(952521), 11(952513), 25(952522), 39(962513), 45(962514), 49(962512), 53(962511) and the 54(962507) sows.
• And those halfsib sows that were born of the Swallow-bellied Mangalica Id. no. 14 and 9 Duroc sows. Twenty-eight F₁ sows were bred from the 9 Duroc sows those are presented in table 4.1.

Table 4.1.

The prolificacy results according to the groupings above are presented in table 4.2., 4.3. and 4.4. We have the data of litters from 12 littermates, 18 fullsibs and 42 halfsibs. A large difference can be found in the number of piglets born. 129 F₂ piglets were born from the F1 littermate matings, with the littersize of 10.75, which can be considered as an excellent result even in the Large White breed (Table 4.2.). 165 F₂ piglets were born from the 18 F₁ fullsib mating with an average littersize of 9.17 (Table 4.3.). The least related - the halfsib (F₁ sows from different mothers and the F₁boar) - mating resulted 357 piglets in 42 litters, with an average litter size of 8.5 (Table 4.). 1.58 more piglets were born from the littermate matings than from the fullsib matings but from different, and 2.25 more piglets were born than from the halfsib matings. The number of piglets weaned were 9.0; 7.61; and 7.33 in the same order. The differences moderated compared to the differences found at birth to 1.39 and 1.67.

The heaviest weaning weight (4.82 kg) of piglets at the age of 30 days was found in the littermate matings compared to the fullsib (4.60 kg) and halfsib (4.55 kg) mating. The litterweight of the littermate matings were heavier with 8.4 kg and 10.0 kg compared to the other two mating respectively. So increased prolificacy meant increased mothering abilty too in the case of littermate matings.

Besides the acceptable prolificacy, the survival rate was also acceptable 83.7, 82.98 and 86.33% in littemate, fullsib and halfsib matings respectively. In the second and third farrowings the prolificacy significantly increased in the case of littermate and halfsib matings and significantly decreased in fullsib matings with 2 piglets.

The data of 72 farrowings are presented in table 4.5. In the experiment 651 piglets were born so far, with an average litter size of 9.04. Comparing this result to the prolificacy of Duroc and Mangalica this is acceptable especially knowing that these are from sib matings. (Figure 4.1.). The average litter size at weaning is 7.67 with a survival rate of 84.8 %, compared to the usual 88-90% till the age of 30 days.

The main problems of inbred piglets are the low viability, the low suckling activity. During the suckling and later phases the death rate is larger (Table 4.6, Figure 4.2.). Consequently the average weaning weight (4.6 kg) of inbred piglets is 2-3 kg lower (30-40%) than those of non-inbred piglets. The 90-day weight of F₂ porklings are 30% lower than of the purebred Duroc porklings (Figure 4.4.).

While at the age of 210 days the purebred Duroc pigs attained the slaughter weight (116.4 kg), the average weight of F2-s were only 84.36 kg – which is 28.5 % less in growth rate (Table 4.7.). Since the different genotypes differed in optimal slaughter weight and age, the ADG was chosen for the basis of comparison. The ADG of pureberd Duroc was 542 g, the ADG of Duroc x Mangalica F₁ was 453 g (-16.5 %), the ADG of F₂ genotype was 349 g (35.5 %) (Table 4.8., Figure 4.5.). The 100 g/day loss in the ADG of F1 is a consequence of the poor growth rate of the Mangalica. The mating of the closely related F₁ –s resulted in a further 100 g/day decrease, which is a result of inbreeding in the F₂ generation. The three genotypes were kept and fed in the same environment.

Out of the 134 pigs slaughtered 107 were F₂, 16 were F₁ and 11 were Blond Mangalica. The slaughter parameters were compared to 56 Duroc pigs. The results are presented in Table 4.9. The Duroc pigs are slaughtered when they reach 110 kg weight. At this weight gives the largest amount of meat, or lean %. Above this weight the lean % decreases. The optimal slaughter weights of Mangalica or Duroc x Mangalica F₁ and F₂ genotypes are higher.

Fifty years ago the Mangalica pigs as fat pigs were fattened till the weight of 180-250 kg, to get a larger amount of fat. The fat % attained 65-75 %. The Mangalica and Duroc x Mangalica F₁ –s slaughtered in the experiments were 136 and 145 kg. The average slaughter weight of the 107 Duroc x Mangalica F₂ were 130.4 kg.

From economic point of view it is important that the pigs attain the optimal slaughter weigth at an early age. Unfortunately neither the Mangalica, nor the F₁ or F₂ pigs did not attain the optimal weight but because of the low growth rate they have been slaughterd earlier. It can be seen in table 4.9, with increasing slaughter weight the loss decreases, that the killing out % increases. This is especially favourable in the case of purebred Mangalica and in the case of F₂ -s.

Reduced fat in the carcass is desirable. The fat increases the value only of special products. The least amount of fat (30.65%) were produced by the Duroc, the largest amount was produced by the Mangalica, 45.43 % (Figure 4.6.). This amount is favourable for producing special ham and lard, but market price does not allow to sell it in large amount. The fat % of the F₁ generation was 37.14 %, having the average of the purebreds, which shows intermediate inheritance. The fat % increases by 10 % in the F₂ generation compared to the F₁.

The predicted lean meat % shows a close relationship with the fat % in the case of the different genotypes. 54.6 % lean meat was found in the Duroc pigs and 40.2 % in the Mangalica pigs. 47.8 % lean meat was found in the F₁ generation that is close to the pureberd average. The meat deposition in the F₂ generation decreased to the Mangalica, independently from the same nutrition – one intensive phase from 30 kg to 100-130 kg.

No visible character of the Duroc can be seen in the F₁ pigs; the F₂ generation especially resemble the Mangalica. Short rounded middle, short and rounded neck. An excessive deposition of fat can be seen below the jaw. The legs are short and thin. While the F₁ –s have larger frames and straight, wild type hair, the F₂ –s have smaller frame and have thin, curly hair.

The colour of Duroc x Mangalica F₁ pigs from birth to weaning is uniform. The colour is dark brown with yellow stripes at birth. The stripes are very similar to those of the wild boars, but darker probably because of the Swallow bellied father. In the later phase of growth the colour of the F₁ –s is dark grey, and often similar to the colour of the wild boar especially at the top of shoulder. The colour of the F₂ –s is very variable. The following colours were differentiated in which similar colours, and colours with and without stripes were grouped.

The colours litters of the three F₁ boars [47(942504), 76(952554), 77(952569)] were recorded, and photographed. The colour was recorded at the first treatment of the piglets after birth. Twenty-four litters (173 piglets) of the three F₁ boars were recorded. The results are presented in table 4.10.

The most frequent colour was the brown 43.3% that is similar to the Duroc colour unicolour or striped. The frequency of the black was 25.4% which is a characteristic of the Swallow-bellied Mangalica. The frequency of the roan and the pale was 26% together (Figure 4.7.) The frequency of the grey that is the colour of the Blond Mangalica at the suckling age was 5.2 %. This shows that there were blond ancestors of the Id.14 boar and the Swallow bellied Mangalica. The frequency folows the Mendelian 1:2:1 distribution from which the blond colour was segregated. The ratio of unicolour and striped seems steadely 1:2 in all colours. All grey individuals are striped. There were twice as much roan coloured individuals in the males than in the females, while the dark colour was more frequent in the females. The ratio of unicolour:striped was 31.5:68.5 % in the males, close to 1:2 , while in the females was 26.8:73.2 %, also close to 1:2. The difference might arise from the small number of individuals, similarly in the sex ratio of 44 % males and 56 % females.

Number of F₂ pigs (31.12.1997.):

01-02.1998.	60
03-04.1998.	100
05-06.1998.	100
07-08.1998.	100
09-10.1998.	100
11-12.1998.	100
Total:	560

Statistical analysis will be carried out when more data available.

• Genotyping procedures have been optimised for the ABI 310 automated DNA sequencer.
• Procedures have been established for semi-automated high throughput genotyping of microsatellite markers.

We take part in three activities from work package 1 - Mapping QTLs:

1.4) Optimising genotyping methods and primer sets.

1.5) Genome scanning to map QTL-s in Mangalica crosses.

1.7) Fine scale QTL mapping.

In 1997 we have made progress with activity 1.4

During this year we purchased the ABI Prism 310 automatic sequencer (partially financed from this project) and we started with optimisation of genotyping procedure for the marker sets 9 through 12 (Table 5.1). These marker sets that were developed by Dr. Groenen, Wageningen Agricultural University within the context of the EC-funded PiGMaP II programme, were discussed at the project meeting in Hungary in March 1997. Primers for markers in sets 9 to 12 were provided by partner 1.

During the optimisation procedure we tried to group markers according to fragment size, dye label and annealing temperature in duplex reactions. If we can establish robust conditions for duplex or multiplex PCR, then we can make significant savings in the consumable (and labour) cost of the genotyping required for the QTL-mapping or genome scanning phase of the project. We have established duplex PCR conditions for the following pairs of markers.

We have also established PCR conditions that allow us to co-amplify one triplex: SW21, S0355 and S0091.

At some loci the resolution / identification of alleles is still uncertain and needs improvement. In total our laboratory has responsibility for 27 markers in the genome scan. At present we can successfully amplify 22 of these microsatellite loci.

Even using a "Touchdown" protocol we are not able to amplify the remaining five loci. If these difficulties cannot be resolved we will, in discussion with the other partners, select alternative closely linked markers from the published maps.

Our group would be interested in isolation of CYP450 cDNA from the cDNA library from adipose tissue. Due to its possible influence on reproduction traits it might be an interesting side product of this research project.

Table 5.1: Microsatellite markers being optimised for high throughput genotyping

During the year 1998 we will finish the genotyping of F₁ animals and proceed with genotyping of F₂ animals. If necessary we will include some new markers in the target area of the genome.

Peter Dovc participated at the 1^st project meeting in Gödöllö Hungary, March 1997.

Peter Dovc, professor

Tamara Milosevic Berlic

Roslin Institute Homepage
Back to the INCOproject homepage