A new strategy of heterosis research in mice-approach and results on chromosome 19

The following steps were performed to analyse heterosis and QTL effects in litter size of mice: intercross of mouse inbred strains C57BL/6J and Balb/cJ in order to produce a F2 generation with 948 female animalsselection of trait groups with extreme high ((13 offspring) and extreme low litter size (5 offspring))typing of 56 microsatellites with an average distance of 32 cM; detection of different chromosome regions with associations to heterosis in litter size. Chromosome 19 was associated to heterosis in litter size. Additional animals with extreme high and low htter sizes were then typed for four DNA markers on chromosome 19 and used for QTL r o T t ^ Q T L W3S l d e n t i f l e d f o r l i t t e r s i 2 e in s e 8 m e i " D19Mit28 D19Mit99 with a maximum at 15 cM (p<0.05). The QTL explains about 11 % of the phenotypic variance in the F2 generation. With a degree of dominance of 4.09 the QTL shows that superdominance can explain heterosis in litter size. Key wgrds: heterosis, fertility, mouse, DNA marker, QTL


Introduction
Heterotic effects are highly significant for animal breeding.As well known, frait values, which improve fitness of animals, are highly influenced by non-additive effects.For example, fertility is a frait with low heritability and indicates large heter-otic effects (GÖTZ et al., 1991).Different types of crosses are suitable for the utilisation of heterosis whereas recurrent reciprocal selection (RRS) can improve nonadditive effects.For RRS, the combination effects are measured from trait values after test matings of appropriate parents in the cross offspring.Test matings in farm animals increase the generation interval, cause financial losses and need high Organization effort (LEE, 1997;BARBOSA-NETO et al., 1997).However, test crosses can be avoided, if homozygosity of DNA loci responsible for heterosis can be analysed in parents (LEUTHOLD, 1968).Superdominance, dominance, linkage and epistasis are assumed to be reasons for heterosis.Hardly another phenomenon of genetics is conflicted with numerous confradictory hypotheses as heterosis (POONI and TREHARNE, 1994;XIAO et al., 1996;STUBER, 1997).Molecular methods offer the possibility to verify the hypotheses of heterosis and to analyse the genetic basis of heterosis (TSAFTARIS, 1995).QTLs for traits with heterotic effects have been identified in plants (e.g.MITCHELL-OLDS, 1995;ARMSTEAD et al., 1997).Similar investigations are still missing in animals.In the work presented, DNA markers were used for QTL analysis of heterosis in litter size.

Approach of heterosis analysis
The long term heterosis project is divided in three parts (Fig. 1).Litter size (number of total born animals, two litters) was used as selection criterium for fertility.Project part I is described in detail by BRUNSCH et al. (1997) andPHILIPP (1997).Reciprocal recurrent selection in part II ofthe project is directed to complementary homozygosity in the two lines.
In project part I, associations between a number of loci and litter size were analysed.Additional loci were considered for positions of chromosomes which had significant effects on litter size and genotyped in a larger number of animals (project part III).In the following contribution, first results are presented from project parts I and III.

4.2
Cross ofthe inbred lines C57BL/6J x Balb/cJ The inbred lines C57BL/6J and Balb/cJ (Fig. 1) were mated reciprocally and subsequently the F, animals were intercrossed.Animals of inbred lines were purchased from Bomholtgard Breeding and Research Centre, Denmark.The F 2 generation consisted of 948 animals.Two extreme frait groups of litter size were selected from the 948 F 2 mice (18 animals with 13 offspring and 14 animals with 5 offspring).In part III additional animals with high or low litter sizes were included for QTL analysis of chromosome 19 (49 animals: 12 offspring; 49 animals: 7 offspring).

DNA analysis
In project part I two sets ofDNA loci have been considered (PHILIPP, 1997): (1) 19 microsatellites which are localized close to loci with known associations to fertility, (2) 32 microsatellites evenly distributed throughout the genome.Additional markers were typed according to the results of project part I on chromosome 19 for which associations to heterosis were found.The microsatellites had an average distance of 14 cM on chromosome 19 (total length 55.7 cM) and the positions 3 cM (D19Mit93), 12 cM (D19Mit28), 20 cM (D19MÜ98) and 41 cM (D19Mit99).Distances between loci are taken from Mouse Genome Database, August 1998.

Statistical analysis
The analysis of associations between genotypes of microsatellite loci and heterosis was performed using two models: superdominance (AA<Aa>aa) or dominance (AA=Aa>aa.).Associations between the degree of heterozygosity and litter size were investigated by Chi square tests according to PEARSON (SPSS, Version 6.1).In cases of observations per genotype smaller than five, the Exact Test has been used.The CRI-MAP Software of GREEN et al. (1990) was applied for linkage mapping.QTL mapping was performed using an interval mapping approach.Chromosome specific thresholds were estimated via permutation analysis.A Bonferroni correction was applied to obtain genome-wide thresholds.

Results
-effects.Litter sizes for inbred strains, F, and F 2 generations are shown in Figure 2. Heterotic effects were 45.3 % in the F, and 15.1 % in the F 2 generation.Associations between DNA markers on chromosome 19 and heterosis in litter size.Table 1 gives the significant associations between the degree of heterozygosity of microsatellites and litter size in the extreme trait groups.Linkage mapping of chromosome 19.The linkage map of chromosome 19 is shown in Figure 3.
QTL mapping for litter size.The plot of the test statistic (Figure 4) suggested a QTL for litter size on chromosome 19 in the interval D19Mit28 -D19Mit99 with the most likely position at 15 cM.Table 2 summarises the results of QTL analysis.

Discussion
Mice are advantageous for heterosis research because of a short generation interval, small expenditure of housing, numerous inbred strains and abundant well characterised DNA markers.The litter size is a major parameter of fertility in mice as well as in all farm animals like pig.In agreement with literature data, high heterotic effect for litter size have been observed after cross of the inbred lines C57BL/6J and Balb/cJ (Fig. 2).So far four of 20 analysed chromosomes were associated with heterosis in litter size.Chromosome 19 was chosen for this study since the microsatellites D19Mit28 and D19Mit61 differed largely in their degrees of heterozygosity between the high and low trait groups and are in agreement with superdominance.
For linkage and QTL mapping of chromosome 19 additional animals and markers have been included.The calculated linkage map is in agreement with the MGD (1998).The results of QTL mapping strongly suggest the presence ofa QTL on chromosome 19 at position 15 cM.The QTL explains about 11 % ofthe phenotypic variance in the F 2 generation.As a potential candidate gene of this chromosome region, the gene Relaxin with association to fertility is located at position 21 cM (MGD, August 1998).The high degree of dominance (4.09) shows that superdominance causes heterosis in litter size.Information on QTL obtained from mice can be fransferred to farm animals.It seems possible to define DNA markers which are homologous between mice and other mammalian species and test them for heterosis in fertility.After identification of the responsible genes they can be used for breeding on non-additive gene effects.