No abstract

Mid-infrared (MIR) spectrometry was used to estimate the fatty acid (FA) composition in cow, ewe, and goat milk. The objectives were to compare different statistical approaches with wavelength selection to predict the milk FA composition from MIR spectra, and to develop equations for FA in cow, goat, and ewe milk. In total, a set of 349 cow milk samples, 200 ewe milk samples, and 332 goat milk samples were both analyzed by MIR and by gas chromatography, the reference method. A broad FA variability was ensured by using milk from different breeds and feeding systems. The methods studied were partial least squares regression (PLS), first-derivative pretreatment + PLS, genetic algorithm + PLS, wavelets + PLS, least absolute shrinkage and selection operator method (LASSO), and elastic net. The best results were obtained with PLS, genetic algorithm + PLS and first derivative + PLS. The residual standard deviation and the coefficient of determination in external validation were used to characterize the equations and to retain the best for each FA in each species. In all cases, the predictions were of better quality for FA found at medium to high concentrations (i.e., for saturated FA and some monounsaturated FA with a coefficient of determination in external validation >0.90). The conversion of the FA expressed in grams per 100mL of milk to grams per 100g of FA was possible with a small loss of accuracy for some FA.

Genetic parameters for the major milk proteins were estimated in the 3 main French dairy cattle breeds (i.e. Montbéliarde, Normande, and Holstein) as part of the PhénoFinlait program. The 6 major milk protein contents as well as the total protein content (PC) were estimated from mid-infrared spectrometry on 133,592 test-day milk samples from 20,434 cows in first lactation. Lactation means, expressed as a percentage of milk (protein contents) or of protein (protein fractions), were analyzed with an animal mixed model including fixed environmental effects (herd, year × month of calving, and spectrometer) and a random genetic effect. Genetic parameter estimates were very consistent across breeds. Heritability estimates (h) were generally higher for protein fractions than for protein contents. They were moderate to high for α-casein, α-casein, β-casein, κ-casein, and α-lactalbumin (0.25 < h < 0.72). In each breed, β-lactoglobulin was the most heritable trait (0.61 < h < 0.86). Genetic correlations (r) varied depending on how the percentage was expressed. The PC was strongly positively correlated with protein contents but almost genetically independent from protein fractions. Protein fractions were generally in opposition, except between κ-casein and α-lactalbumin (0.39 < r < 0.46) and κ-casein and α-casein (0.36 < r < 0.49). Between protein contents, r estimates were positive, with highest values found between caseins (0.83 < r < 0.98). In the 3 breeds, β-lactoglobulin was negatively correlated with caseins (-0.75 < r < -0.08), in particular with κ-casein (-0.75 < r < -0.55). These results, obtained from a large panel of cows of the 3 main French dairy cattle breeds, show that routinely collected mid-infrared spectra could be used to modify milk protein composition by selection.

In the context of the PhénoFinLait project, a genome-wide analysis was performed to detect quantitative trait loci (QTL) that affect milk protein composition estimated using mid-infrared spectrometry in the Montbéliarde (MO), Normande (NO), and Holstein (HO) French dairy cattle breeds. The 6 main milk proteins (α-lactalbumin, β-lactoglobulin, and αS1-, αS2-, β-, and κ-caseins) expressed as grams per 100g of milk (% of milk) or as grams per 100g of protein (% of protein) were estimated in 848,068 test-day milk samples from 156,660 cows. Genotyping was performed for 2,773 MO, 2,673 NO, and 2,208 HO cows using the Illumina BovineSNP50 BeadChip (Illumina Inc., San Diego, CA). Individual test-day records were adjusted for environmental effects and then averaged per cow to define the phenotypes analyzed. Quantitative trait loci detection was performed within each breed using a linkage disequilibrium and linkage analysis approach. A total of 39 genomic regions distributed on 20 of the 29 Bos taurus autosomes (BTA) were significantly associated with milk protein composition at a genome-wide level of significance in at least 1 of the 3 breeds. The 9 most significant QTL were located on BTA2 (133 Mbp), BTA6 (38, 47, and 87 Mbp), BTA11 (103 Mbp), BTA14 (1.8 Mbp), BTA20 (32 and 58 Mbp), and BTA29 (8 Mbp). The BTA6 (87 Mbp), BTA11, and BTA20 (58 Mbp) QTL were found in all 3 breeds, and they had highly significant effects on κ-casein, β-lactoglobulin, and α-lactalbumin, expressed as a percentage of protein, respectively. Each of these QTL explained between 13% (BTA14) and 51% (BTA11) of the genetic variance of the trait. Many other QTL regions were also identified in at least one breed. They were located on 14 additional chromosomes (1, 3, 4, 5, 7, 15, 17, 19, 21, 22, 24, 25, 26, and 27), and they explained 2 to 8% of the genetic variance of 1 or more protein composition traits. Concordance analyses, performed between QTL status and sequence-derived polymorphisms from 13 bulls, revealed previously known causal polymorphisms in LGB (BTA11) and GHR (BTA20 at 32 Mbp) and excluded some other previously described mutations. These results constitute a first step in identifying causal mutations and using routinely collected mid-infrared predictions in future genomic selection programs to improve bovine milk protein composition.

Les acteurs des filières laitières bovine, caprine et ovine françaises se sont regroupés dans le programme PhénoFinlait autour d’un but commun : caractériser la composition du lait en Acides Gras (AG) et protéines afin de la maîtriser. La quantification des AG et des protéines devait être possible à grande échelle et à moindre coût avant d’identifier des leviers permettant d’adapter cette composition à la demande. PhénoFinlait s’est organisé autour de trois objectifs : i) caractériser précisément la composition du lait, ii) phénotyper et génotyper une large population de femelles sur l’ensemble du territoire français et iii)identifier les leviers génétiques et alimentairespermettant de maîtriser cette composition. La spectrométrie dans le Moyen InfraRouge (MIR) a été choisie comme méthode de quantification à haut débit des composants du lait. Elle permet la quantification précise en routine de 15 à 27 AG, des quatre caséines et des deux protéines majeures du lactosérum. Une collecte de données de grande ampleur a été mise en œuvre dans plus de 1 500 élevages bovins, caprins et ovins. Les données de production laitière, les spectres MIR du lait, les informations sur le stade physiologique des femelles et sur la composition de l’alimentation des troupeaux ont été recueillies. Plus de 12 000 vaches, chèvres et brebis ont été génotypées. Finalement, plus de 800 000 données représentatives des situations de l’élevage français ont été stockées dans une base de données destinée à l’étude du déterminisme génétique de la composition en AG et en protéines du lait, et des facteurs d’élevage l’influençant.