Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press REVIEW Livestock Genomics Comes of Age Michel Georges and Leif Anclersson 2 1Department of Genetics, Faculty of Veterinary Medicine, University of Liege, 4000-Liege, Belgium; 2Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Uppsala Biomedical Centre, 24 Uppsala, Sweden It is estimated that man 1 first domesticated ani- as by an undefined number of polygenes or quan- mals as early as 10,000 BP. Since then, farmers titative trait loci (QTL). Their heritabilities typi- have been unwittingly manipulating livestock cally range from less than 5% to over 50%. Ani- genes by selective breeding. This early genetic en- mal geneticists therefore have had a long-stand- gineering generated a wealth of variation for a ing interest in the genetics of complex inherit- myriad of traits in the different livestock species. ance, the relevance of which is being recognized The dramatic size difference between the Shire increasingly in medical genetics as well. and Shetland pony or the plethora of dog breeds The implementation of breeding schemes are just two illustrations of the diversification ob- that proved so efficient would have been impos- tained by artificial selection. The resulting carica- sible without the organization of extensive phe- ture of naturally occurring variation has played a notypic record keeping. Particularly illustrative major role in Darwin's formalization of his in this respect is the collection of individual re- theory of the evolution of species by natural se- cords (milk yield and composition, type traits, lection. The first chapter of his acclaimed The Ori- health traits, etc.) that is performed on a monthly gin of Species (1859) is devoted to "Variation Un- basis for millions of cows as part of dairy herd der Domestication." improvement programs in the United States, In the beginning of this century, the coales- Western Europe, and several other parts of the cence of biometrics and Mendelian genetics pio- world. Likewise, breeding companies carefully neered by R.A. Fisher, J.B.S. Haldane, and S. monitor their pig and poultry breeding stock for Wright led to the foundation of quantitative ge- a whole range of phenotypic measurements. netics, which in turn allowed development of the Given the tradition of eagerly adopting mod- theory of animal breeding that is still imple- ern technology that might increase genetic re- mented to this date. This biometrical era of ani- sponse as well as the availability of unique mate- mal genetics has led to spectacular increases in rial, it is quite surprising that animal geneticists productivity in all major livestock species during have been reluctant to invest in genomics when the second half of this century. As an example, in compared with plant breeders or human medical the United States milk yield has increased from geneticists. This is likely due in part to the real- approximately 4500 kg per cow per year to more ization that the majority of economically impor- than 6800 kg in less than 20 years (Pearson et al. tant traits in livestock are typical multifactorial 1990). traits and are therefore the most difficult ones to It is noteworthy that the vast majority of pro- tackle using genomic strategies. Moreover, given duction traits undergoing selection are typical the spectacular genetic progress achieved by quantitative traits, that is, they exhibit a continu- means of conventional breeding programs, some ous rather than a discrete distribution and they skepticism has prevailed regarding the cost- are influenced by environmental factors as well effectiveness of biotechnology in livestock pro- duction. It is only during the last 5 years that we have witnessed a growing interest of animal ge- neticists in genomics. This has undoubtedly been 3Corresponding author. catalyzed by the successes of the Human Genome E-MAIL michel@stat'fmv'ulg'ac'be; FAX 32-41-66-41-22. Initiative. This review will summarize the state of 6:907-921 9 by Cold Spring Harbor Laboratory Press ISSN 1054-9803/96 $5.00 GENOME RESEARCH ~ 907 Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press GEORGES AND ANDERSSON the art in livestock genomics, while emphasizing tion provided by linked markers is considered some specificities of animal studies, and will jointly. Moreover, animal geneticists may have identify some of the major challenges for the the option to arrange matings between breeds future. and even subspecies in a manner reminiscent of the Mus musculus domesticus x Mus spretus or cas- Current Status of Genome Analysis taneus crosses, enhancing the information con- tent at the marker as much as trait loci (see be- Primary Microsateilite-basedMaps Are Being low). Nevertheless, continued marker develop- Generated for Host Livestock Species ment seems advisable in most domestic species to The description of microsatellites as an abundant ensure adequate information content across the source of highly polymorphic, well-dispersed, respective genomes. and conveniently typed markers has boosted the Sequence conservation has proven to be suf- generation of primary maps in livestock species ficiently high between cattle and sheep to allow (for review, see Beattie 1994). Individual as well for -50% of primer sequences developed in one as internationally coordinated efforts have lead species to work in the other (Moore et al. 1991; A. to the generation of a number of linkage maps in Crawford and N. Cockett, pers. comm.). This per- the different species: cattle (Barendse et al. 1994; centage is considerably higher than what has Bishop et al. 1994; Ma et al. 1996), pig (Ellegren been observed for the mouse and rat, which are et al. 1994; Rohrer et al. 1994; Archibald et al. thought to be as closely related evolutionarily. As 1995; Marklund et al. 1996; Rohrer et al. 1996), expected, this cross-species use of microsatellite sheep (Crawford et al. 1995), and poultry (Bum- markers has proven inefficient for more distantly stead and Palyga 1992; Levin et al. 1994; Cheng related domestic species despite the occasional et al. 1996; Crooijmans et al. 1996) in particular. demonstration of remarkable conservation of mi- When several parallel efforts have been under- crosatellite position (Moore et al. 1991; Ellegren taken in a given species, the number of common et al. 1993; Sun and Kirkpatrick 1996). markers fortunately has usually been sufficiently Most of the linkage groups have been an- high to allow for efficient cross-referencing (Eg- chored and oriented to specific chromosomes in gen and Fries 1995; Marklund et al. 1996). At this cattle and pig using fluorescence in situ hybrid- point, the available maps number -1100 markers ization (FISH) mapping. In poultry, establishing for cattle and pig (C. Beattie, pers. comm.; Rohrer connections between the linkage and cytogenetic et al. 1996), 700 for sheep (A.M. Crawford, pers. maps is complicated by nonidentifiable mini- comm.), and 450 for poultry (M.A.M. Groenen chromosomes, which might represent as much as and N. Bumstead, pets. comm.). These maps pro- 30% of the genome. vide very adequate genome coverage, especially These mapping data are being compiled in a for cattle and pig, where the average between number of data bases conveniently accessed via marker interval is now -2.5-5 cM. the internet (e.g., http://www.ri.bbsrc.ac.uk/ Artificial breeding schemes often result in a genome_mapping.html; http://dirk.invermay. reduction in effective population size. Assuming cri.nz; http://locus.jouy.inra.fr/cgi-bin/bovmap/ selective neutrality of microsatellite alleles, one Bovmap/intro.pl; http://sol.marc.usda.gov/marc/ could therefore expect reduced allele numbers html/genel.html; http://probe.nalusda.gov: and heterozygosity when compared with the hu- 8000/index.html). man. In addition, because fewer markers have Thus, with the available genetic maps, we been developed in animal species, less selection have reached the point where the marker re- has taken place based on information content in sources are no longer the limiting factor when putting together marker panels for genome scans. attempting to map trait loci by exploiting within- As an example, the battery of markers used to family linkage disequilibrium. It can be argued, scan the bovine genome in a search for QTL af- however, that in a number of instances the map- fecting milk production had an average hetero- ping method of choice would be based on linkage zygosity of 56% in the studied Holstein-Friesian disequilibrium existing across the population population (Georges et al. 1995), compared with (such as in identity-by-descent mapping meth- heterozygosities typically >70% in human micro- ods). Efficient implementation of such strategies satellite-based studies. Part of the segregation in- will require further development of a substantial formation can be recovered by performing mul- number of additional markers in the different tipoint linkage analyses in which the informa- livestock species. 908 ~ GENOME RESEARCH Downloaded from genome.cshlp.org on September 29, 2021 - Published by Cold Spring Harbor Laboratory Press LIVESTOCK GENOMICS Comparative Mapping Confirms the Remarkable cessfully applied to the bovine and porcine Conservation of Synteny Amongst karyotypes, allowing for the majority of chromo- Mammalian Species somes to be sorted as pure fractions (Schmitz et al. 1992, 1995; Langford et al. 1993; Yerle et al. Although microsatellites undoubtedly form the 1993). These advances should facilitate the devel- scaffold of all livestock linkage maps, a sufficient opment of chromosome-specific libraries and number of evolutionarily conserved Type I mark- markers in the near future. Likewise, a number of ers (O'Brien 1991; O'Brien et al. 1993) are inter- successful applications of chromosome or chro- spersed in these maps to confirm the extensive mosome-band microdissection for the develop- conservation of synteny among distantly related ment of region-specific markers are being re- mammals that was predicted from early mapping ported (Schmutz et al.