M. Soller

A genetic linkage map of the bovine genome

A cattle genetic linkage map was constructed which marks about 90% of the expected length of the cattle genome. Over 200 DNA polymorphisms were genotyped in cattle families which comprise 295 individuals in full sibling pedigrees. One hundred and seventy–one loci were found linked to one other locus. Twenty nine of the 30 chromosome pairs are represented by at least one of the 36 linkage groups. Less than a 50 cM difference was found in the male and female genetic maps. The conserved loci on this map show as many differences in gene order compared to humans as is found between humans and mice. The conservation is consistent with the patterns of karyotypic evolution found in the rodents, primates and artiodactyls. This map will be important for localizing quantitative trait loci and provides a basis for further mapping.

Simple sequence repeats as a source of quantitative genetic variation

Most traits in biological populations appear to be under stabilizing selection, which acts to eliminate quantitative genetic variation. Yet, virtually all measured traits in biological populations continue to show significant quantitative genetic variation. The paradox can be resolved by postulating the existence of an abundant, though unspecified, source of mutations that has quantitative effects on phenotype, but does not reduce fitness. Does such a source actually exist? We propose that it does, in the form of repeat-number variation in SSRs (simple sequence repeats, of which the triplet repeats of human neurodegenerative diseases are a special case). Viewing SSRs as a major source of quantitative mutation has broad implications for understanding molecular processes of evolutionary adaptation, including the evolutionary control of the mutation process itself.

Selective genotyping for determination of linkage between a marker locus and a quantitative trait locus

Summary“Selective genotyping” is the term used when the determination of linkage between marker loci and quantitative trait loci (QTL) affecting some particular trait is carried out by genotyping only individuals from the high and low phenotypic tails of the entire sample population. Selective genotyping can markedly decrease the number of individuals genotyped for a given power at the expense of an increase in the number of individuals phenotyped. The optimum proportion of individuals genotyped from the point of view of minimizing costs for a given experimental power depends strongly on the cost of completely genotyping an individual for all of the markers included in the experiment (including the costs of obtaining a DNA sample) relative to the cost of rearing and trait evaluation of an individual. However, in single trait studies, it will almost never be useful to genotype more than the upper and lower 25% of a population. It is shown that the observed difference in quantitative trait values associated with alternative marker genotypes in the selected population can be much greater than the actual gene effect at the quantitative trait locus when the entire population is considered. An expression and a figure is provided for converting observed differences under selective genotyping to actual gene effects.

Genetic polymorphism in varietal identification and genetic improvement

SummaryNew sources of genetic polymorphisms promise significant additions to the number of useful genetic markers in agricultural plants and animals, and prompt this review of potential applications of polymorphic genetic markers in plant and animal breeding. Two major areas of application can be distinguished. The first is based on the utilization of genetic markers to determine genetic relationships. These applications include varietal identification, protection of breeders rights, and parentage determination. The second area of application is based on the use of genetic markers to identify and map loci affecting quantitative traits, and to monitor these loci during introgression or selection programs. A variety of breeding applications based on these possibilities can be envisaged for Selfers, particularly for those species having a relatively small genome size. These applications include: (i) screening genetic resources for useful quantitative trait alleles, and introgression of chromosome segments containing these alleles from resource strain to commercial variety; (ii) development of improved pure lines out of a cross between two existing commercial varieties; and (iii) development of crosses showing increased hybrid vigor. Breeding applications in segregating populations are more limited, particularly in species with a relatively large genome size. Potential applications, however, include: (i) preliminary selection of young males in dairy cattle on the basis of evaluated chromosomes of their proven sire; (ii) genetic analysis of resource strains characterized by high values for a particular quantitative trait, and introgression of chromosome segments carrying alleles contributing to the high values from resource strain to recipient strain.

On the power of experimental designs for the detection of linkage between marker loci and quantitative loci in crosses between inbred lines.

SummaryThe power of experiments aimed at detecting linkage between a quantitative locus and a marker locus, both segregating in the backross or F2 generation of a cross between two inbred lines, is examined. Given that the two lines are close to fixation for alternative alleles of both marker locus and quantitative locus, it is concluded that experiments involving a few thousand offspring should be able to detect close linkages involving quantitative loci (or groups of loci) having rather modest effects (i.e., that contribute, say, 1% of the total phenotypic variance in the F2).

Restriction fragment length polymorphisms in genetic improvement: methodologies, mapping and costs

SummaryRecently a new class of genetic polymorphism, restriction fragment length polymorphisms (RFLPs), has been uncovered by the use of restriction endonucleases which cleave DNA molecules at specific sites and cloned DNA probes which detect specific homologous DNA fragments. RFLPs promise to be exceedingly numerous and are expected to have genetic characteristics — lack of dominance, multiple allelic forms and absence of pleiotropic effects on economic traits — of particular usefulness in breeding programs. The nature of RFLPs and the methodologies involved in their detection are described and estimated costs per polymorphism determination are derived. The anticipated costs of applying RFLPs to genome mapping are considered in terms of the number of RFLPs required for a given degree of genome coverage, the number of probe × enzyme combinations tested per polymorphism uncovered, and the total number of individuals and polymorphisms scored for mapping purposes. The anticipated costs of applying RFLPs to genetic improvement are considered in terms of the number of individuals and the number of polymorphisms per individual that are scored for the various applications. Applications considered include: varietal identification, identification and mapping of quantitative trait loci, screening genetic resource strains for useful quantitative trait alleles and their marker-assisted introgression from resource strain to commercial variety, and marker-assited early selection of recombinant inbred lines in plant pedigree breeding programs and of young sires in dairy cattle improvement programs. In most cases anticipated costs appear to be commensurate with the scientific or economic value of the application.

Trait-based analyses for the detection of linkage between marker loci and quantitative trait loci in crosses between inbred lines.

SummaryMethods are presented for determining linkage between a marker locus and a nearby locus affecting a quantitative trait (quantitative trait locus=QTL), based on changes in the marker allele frequencies in selection lines derived from the F-2 of a cross between inbred lines, or in the “high” and “low” phenotypic classes of an F-2 or BC population. The power of such trait-based (TB) analyses was evaluated and compared with that of methods for determining linkage based on the mean quantitative trait value of marker genotypes in F-2 or BC populations [marker-based (MB) analyses]. TB analyses can be utilized for marker-QTL linkage determination in situations where the MB analysis is not applicable, including analysis of polygenic resistance traits where only a part of the population survives exposure to the Stressor and analysis of marker-allele frequency changes in selection lines. TB analyses may be a useful alternative to MB analyses when interest is centered on a single quantitative trait only and costs of scoring for markers are high compared with costs of raising and obtaining quantitative trait information on F-2 or BC individuals. In this case, a TB analysis will enable equivalent power to be obtained with fewer individuals scored for the marker, but more individuals scored for the quantitative trait. MB analyses remain the method of choice when more than one quantitative trait is to be analyzed in a given population.

Restriction fragment length polymorphisms and genetic improvement of agricultural species

Evidence is accumulating demonstrating the ubiquity and abundance of a new class of genetic markers, restriction fragment length polymorphisms (RFLPs). These markers should allow the genetic map of agricultural species to be saturated in the near future. This holds great promise for useful applications, including: protection of breeders rights, and more effective means for the characterization and manipulation of individual genetic loci affecting traits of economic importance.

Controlling the proportion of false positives in multiple dependent tests.

Genome scan mapping experiments involve multiple tests of significance. Thus, controlling the error rate in such experiments is important. Simple extension of classical concepts results in attempts to control the genomewise error rate (GWER), i.e., the probability of even a single false positive among all tests. This results in very stringent comparisonwise error rates (CWER) and, consequently, low experimental power. We here present an approach based on controlling the proportion of false positives (PFP) among all positive test results. The CWER needed to attain a desired PFP level does not depend on the correlation among the tests or on the number of tests as in other approaches. To estimate the PFP it is necessary to estimate the proportion of true null hypotheses. Here we show how this can be estimated directly from experimental results. The PFP approach is similar to the false discovery rate (FDR) and positive false discovery rate (pFDR) approaches. For a fixed CWER, we have estimated PFP, FDR, pFDR, and GWER through simulation under a variety of models to illustrate practical and philosophical similarities and differences among the methods.

A deletion in the myostatin gene causes the compact (Cmpt) hypermuscular mutation in mice.

Mouse myostatin (GDF-8) is a member of the TGF-b superfamily of secreted growth and differentiation factors (McPherron et al. 1997). Disruption of the myostatin gene by gene targeting in mice caused a large and widespread increase in skeletal muscle mass resulting from a combination of muscle cell hyperplasia and hypertrophy. These results suggest that myostatin functions specifically as a negative regulator of skeletal muscle growth (McPherron et al. 1997). Mutations in bovine myostatin were subsequently shown to be present in the Belgian Blue and Piedmontese breeds of double-muscled (mh) cattle (Grobet et al. 1997; Kambadur et al. 1997; McPherron and Lee 1997). We have recently reported the mode of inheritance of a new hypermuscular mouse mutation termed ‘‘compact’’ (Cmpt), which was uncovered in the course of a long-term selection program for high carcass protein content (Varga et al. 1997). Homozygous Cmpt lines (denoted HCI) were produced by selection and inbreeding, and crossed to normal BALB/c mice to produce a very large F2 population for mapping. Within the F2 population, the compact trait, as scored by visual inspection, showed a continuous normal distribution. Mapping, therefore, was through genotyping and haplotype analysis of 370 male and 274 female F2 mice showing the highest degree of compact phenotype development (denoted C5). On this basis, Cmpt was mapped to an 8.2-cM region between D1Mit375 and D1Mit21 on mouse Chromosome (Chr) 1 (Varga et al. 1997). Subsequent genotyping with additional microsatellites in this region showed that Cmpt is located very close to the D1Mit237 microsatellite marker. For the most part this region corresponds to the 2q32-35 segment of human Chr 2 (Mouse Genomic Database, MGD), which shows synteny homology to the centromeric region of bovine Chr 2 to which the double-muscle mh gene was mapped (Charlier et al. 1995). On the basis of the phenotypic and comparative mapping information, therefore, myostatin became a strong candidate gene for Cmpt. Two primer pairs were designed to amplify the coding sequence of myostatin cDNA (McPherron et al. 1997). Reverse transcription-PCR (RT-PCR) was performed with total RNA isolated from skeletal muscle of compact and normal laboratory mice (NMRI), and PCR products corresponding to myostatin cDNA were cloned. Sequencing of two of the clones derived from the second primer pair revealed a 12-bp deletion from positions 775– 786 (numbering according to the GenBank sequence for GDF-8, accession number U84005). This deletion, denoted Mstn, eliminates the following amino acids: Leu (224), Gly (225), Ile(226), Glu(227), Ile(228) and at the same time creates a new Phe residue (Fig. 1). The deletion was verified by sequencing from genomic DNA of an individual exhibiting strong expression of the compact phenotype. Sequencing from genomic DNA of a normal (BALB/c) individual did not reveal the deletion. This finding, in conjunction with the Genebank sequence, confirms that the 12-bp deletion is not solely a polymorphic difference between the HCI and NMRI strains. For genotyping the deletion, a reverse primer was designed and used with the forward primer (MSTN2F) of the second primer pair. Amplification products were separated by PAGE electrophoresis and visualized by silver staining. Identification of the 206-bp allele carrying the deletion and the normal 218-bp allele was unambiguous. In order to map the mouse myostatin locus and determine whether it cosegregates with Cmpt, we genotyped 334 C5 male and 261 female mice of the above-mentioned large F2 population with respect to the 12-bp deletion. Combined results of the male and female animals are shown in Fig. 2. These results place Mstn and thus the mouse myostatin gene between D1Mit375 and D1Mit237, very close to D1Mit237. In addition to the animals included in Fig. 2, there were 11 individuals heterozygous for the HCI and BALB/c alleles at all the three marker loci and at Mstn. It is very likely that these heterozygous animals were classified to the C5 phenotypic category owing to variable expressivity of the compact trait. This is supported by the observation that 9 of the 11 C5 heterozygotes were males, in which expressivity of the compact phenotype in the heterozygote is stronger, and only 2 were females. There was also a single individual homozygous for the normal BALB/c (non-deletion) allele at Mstn and also homozygous for the BALB/c allele at all three marker loci. Considering that over 8200 F2 animals were scored and recorded for the compact trait in the course of these experiments, we presume that this is an individual with normal muscularity that was included in the C5 phenotypic class as a result of recording error. Except for these animals, the Mstn mutation was completely associated with the Cmpt mutation. In particular, Mstn and Cmpt were associated in all 50 recombinant haplotypes involving D1Mit375, Mstn and D1Mit237. We conclude, therefore, that the Mstn mutation in the myostatin gene is the causative mutation responsible for the hypermuscular compact phenotype in mice. The distance between Mstn and D1Mit237 is 0.3 cM on the basis of the C5 haplotype results, which places myostatin 27.7 cM from the centromere on Chr 1 (Mouse Genome Database 1998). The Cmpt deletion is in the propeptide region (residues 224– 228) that precedes the proteolytic processing site of mouse myostatin. The structure of the biologically active growth factor domain is thus unaffected by this mutation. Consequently, the apparent loss of myostatin activity is not owing to disruption of the growth factor domain, as is the case for the Belgian Blue and Piedmontese double-muscled cattle (Grobet et al. 1997; Kambadur Correspondence to: L. Varga Mammalian Genome 9, 671–672 (1998).