Know more

As part of an interdisciplinary project, PICSAR-PARADIGM financed by INRAE (GISA metaprogramme), APIS-GENE and GDS-France, the healthy or diseased phenotypes (clinical or subclinical) of 2293 cows (1644 Holstein and 649 Normandy) were finely characterized with at least four consistent diagnoses with a case-control type experiment. After imputation of the complete genome sequence of each cow, an association analysis showed three QTL on chromosomes 12, 13 and 23, targeting in particular the genes ABCC4, CBFA2T2 and IER3 which are also good functional candidates.

Paratuberculosis is an infectious, contagious and incurable disease of ruminants caused by a mycobacterium, Mycobacterium avium paratuberculosis (MAP). Transmission mainly occurs by absorption of MAP of fecal origin by calves in the first weeks of life, whereas the disease develops much later after several years of latency. Contamination is therefore essentially intra-herd; transmission between herds is the result of exchanges of infected animals. Currently, control methods are relatively inefficient since most infected animals are not detected due to this long latency period and because diagnostic tests (Elisa and PCR) lack sensitivity and specificity. A large interdisciplinary research program on the control of this disease was launched in 2014 with financial support from INRAE (the PICSAR  project of the GISA metaprogramme), APIS-GENE and GDS France (PARADIGM project). In this program, a genetic component is aimed at analyzing the genetic determinism for resistance to paratuberculosis and selection of this trait.

An effort was made in particular to build a set of reliable phenotypes. Using control tests performed by the GDS in the affected farms, the animals considered as healthy and sick underwent two standard tests to confirm infection. Using these results, three phenotypes were defined: clinical cases, subclinical cases (infected but without clinical signs, confirmed positive by at least three tests), healthy animals (exposed, sufficiently old enough and confirmed negative by at least four tests). In the end, 1644 Holstein and 649 Normandy cows integrated the study with a reliable phenotype, their complete genome sequence was imputed by SNP array typings using reference sequences from the “1000 bovine genome” project. An association analysis on the sequence scale revealed three regions on chromosomes (BTA) 12, 13 and 23 presenting very significant effects on paratuberculosis resistance in the Holstein breed whereas only one was identified in the Normandy breed (BTA23) due to a reduced power. The most significant effects were found on BTA13 in a region reduced by at least 10 kb. The ABCC4 (BTA12), CBFA2T2 (BTA13), and IER3 (BTA23) genes appear as very good candidate genes, for both their position and function.

This study shows the value of basing association studies on highly reliable phenotypes. However, the number of animals is still low and should be increased to improve power of detection. In fact, the method must be applied on other breeds. A proposed solution may be the implementation of genomic selection, which would allow both the selection of animals for a better resistance and the use of the information produced in genetic control programs

INRAE Méeaprogramme: GISA