Taking the bull by the horns

While cattle have been domesticated for 10,000 years, it was not until the twentieth century that it became possible to develop a system for selecting sires based on traits in their offspring. While the method was reliable, assessing a bull’s genetic potential took five years — the time for its female offspring to reach breeding age. A breakthrough came in 2009 with genomic selection, allowing calves to be assessed from birth using their genomic data.

Health and wellbeing are focus areas

“What an amazing adventure it’s been!” says Didier Boichard, leader of the team of 30 that received the 2016 INRA Science with an Impact Award. “There’s a terrific team spirit and a healthy sense of competition. While there may be a lot of meetings everyday, our tight-knit group is always there to help each other.” Researchers from INRA and engineers from ALLICE and IDELE* work together at the Joint Technological Unit for Cattle Management, Genetics and Genomics (UMT 3G), whose work is supported by APIS-GENE. The group has built a shared work culture that’s at science’s leading edge — with more than 60 scientific publications to its name — and also attuned to industry needs.

Everything’s possible

In 2016, 80% of French breeding bulls are genomically indexed

Genomic selection is quicker and less costly, and makes it possible to use a wider variety of breeding animals, thereby favouring genetic diversity. It makes it possible to select sires based on a wide range of traits and complex characteristics, including those that are difficult to measure or that have low heritability, such as fertility or health, which are difficult to target with conventional breeding methods. Genomic selection can be applied to over 40 traits, making it possible to meet a highly diverse range of objectives, including those from alternative agriculture. New traits will be added to the process, such as for milk quality and meat tenderness, as well as for environmental criteria such as methane emission. Animal health is another focus area to improve animal wellbeing and longevity, and to reduce the use of antibiotics through traits for mastitis resistance, foot health, metabolic diseases or genetic defects. Selecting the gene for hornlessness will also make it possible to eliminate the need for dehorning.

A full-speed revolution

In a single year, 2009, genomic selection processes were developed and put in place for three breeds of dairy cattle: Holstein, Normande and Montbéliarde. This has since grown to cover 11 breeds, including the three most popular beef cattle breeds in France: Charolaise, Limousine and Blonde d’Aquitaine.

Genomic selection is a groundbreaking concept that looks at the statistical relationship, within a reference population, between an animal genotype, i.e., its DNA sequence, and its phenotype, which are its observed characteristics and traits. Using only its genotype, the genomic model makes it possible to “predict” an animal’s “value” as regards specific criteria from a very young age. An animal’s genotype is analysed using biological samples such as blood or ear biopsies and then measured against thousands of genetic markers that are distributed throughout the genome and that vary among individuals. The larger the reference population, both in terms of genotype and phenotype, the more accurate the predictions will be. Boichard calls this a “snowball effect”; the more breeders use genomic selection to test their animals, the more data will be available to inform predictions.
 

The genomic revolution is also informed by history. The genealogies and pedigrees that have been compiled over time were used to build the mathematical models for genomic selection. The process began in 2001 with a large-scale marker assisted selection (MAS) programme, of which INRA was a pioneer. This programme facilitated the rapid adoption of genomic selection, with the help of two key facilities: GIE LABOGENA for genotyping and the Genetic Data Processing Centre (CTIG) to compile data and to provide computing power.

Strong vision of the common good

The French Animal Breeding Act of 1966 assigned INRA the task of creating “breeding indices” for all cattle breeds. A major aim was to spread the use of this method to all cattle breeds and assess their sustainability. Through partnerships with other European countries, the reference population for Holsteins was quadrupled as a part of the EuroGenomics programme,** leading to a huge advance in the quality of predictions.
 

Another challenge was to make the method accessible for stakeholders in developing countries. “There is a growing divide between countries that use genomic selection and those that are not able to,” says Vincent Ducrocq. “The aim is to create win–win partnerships that would allow countries to build reliable, local assessment infrastructure and for France to better assess its bull populations for export overseas.” In 2014, a partnership was established with South Africa. Another project with Indian NGOs has been running for 12 years.
 

For Boichard, the foundation of the group’s work is to exploit the full potential of genome information. The next step will be to use the full sequencing of animal genomes. Instead of using markers associated with genes, the aim is to use the genes directly responsible for variation in a trait, thereby making breeding even more precise. The work will continue to be a team sport, one that UMT 3G is well trained for.

^{* UMT 3G: Joint Technological Unit for Cattle Management, Genetics and Genomics, was founded in 2009 and brings together INRA, ALLICE (French Union of Livestock Cooperatives) and IDELE (French Livestock Institute).
** Eurogenomics Consortium, established in 2009 by France, Germany, Netherlands, Denmark, Sweden and Finland, which Spain and Poland later joined.}