Abstract
Reggiana and Modenese are dual-purpose local cattle breeds, but almost all milk produced by these two breeds is processed in mono-breed Protected Designation of Origin Parmigiano-Reggiano cheese. The main objective of the project is to merge phenotypic traits, some of which have never been previously analysed in Italian local breeds, with omics data to link phenotypes to genotype variability that could be useful to guide deci-sions in the management and breeding plans of Reggiana and Modenese genetic resources. In particular, about 500 animals of Reggiana and 100-150 animals of Modenese will be subjected to phenotyping of novel milk and cheese-making efficiency traits and other related proxies such as resilience indicators (health traits, longevity, fertility), and omic data (metabolome, and genome profiles). Statistical analysis of new data will be implemented (e.g. heritability and correlations with investigated phenotypes). Genome wide association study (GWAS) will be performed to link single nucleotide polymorphism (SNP) to productive and qualitative economically important traits, resilience relat-ed indicators and metabolome profiles. The surrounding region of significant SNPs will be investigated to identify candidate genes and putative causative mutations. The project is divided in 4 Work Packages (WPs) that will assure the correct work flow of all activities to reach the indicated objective: WP1: Coordination, management, measures to maximize impact, communication and dissemination. WP2: Milk and cheese-making traits and other related proxies. WP3: Production of omics (metabolomics, foodomics, and genomics) data. WP4: Integration of data and applications. Two research groups with extensive research experience in milk production, quality and cheese-making ability (UNIPR), and metabolomics and genomics in livestock (UNIBO), are proposing this project that is innovative and applied at the same time. The project is expected to have important impacts on different fields: efficiency of cheese-making process (yield and nutrient recovery in cheese), resilience (e.g., disease reduction) of the local cattle populations; new knowledge in basic and applied research fields.
Results achieved
Animals of Reggiana and Modenese were subjected to phenotyping (external phenotypes) of traditional and novel milk and cheese-making efficiency traits and other related proxies. Five main compositional traits (fat, protein, casein, lactose, and urea) were measured and casein index was then estimated. With respect to milk processing traits and cheese yielding ability, traditional coagulation parameters: rennet coagulation time, time from coagulation to a curd firmness of 20 mm (k20), and curd firmness (a30 and other related measures) were measured. Two cheese-yielding ability traits were derived from laboratory cheese-making procedures: cheese yield of processed milk, and recovery of milk components in the curd. Additional external phenotypes were analysed such as milk fat globule size (MFGS) and the content of macro- and micro-minerals. Traits related to udder health were determined, namely Somatic Cell Count (SCC), and Differential Somatic Cell Count (DSCC), representing the combined proportions of different somatic cell populations such as polymorphonuclear leukocytes (PMM), lymphocytes (LYM) and macrophages (MAC). Official milk recording data were obtained for the sampled animals. These included productive, functional, and health traits (e.g., milk yield, days open, calving interval, parity, and other available functional indicators). Pedigree data of the animals were also retrieved and integrated in a data set. Metabolomic profiles (internal phenotypes) were generated from plasma samples. An untargeted metabolomics approach was adopted, based on liquid chromatography-tandem mass spectrometry (LC-MS/MS) platforms. The resulting profiles included relative quantification of 905 biochemicals [796 compounds of known identity (named biochemicals) and 109 compounds of unknown structural identity (unnamed biochemicals)], belonging to 7 biological classes (e.g., lipids, amino acids, nucleotides, carbohydrates, etc.) and covering >30 major metabolic pathways. Metabolomic profiles of plasma samples were successfully obtained, providing a robust dataset for subsequent integration analyses. With regard to genomic data, SNP genotypes from 1,500 animals previously genotyped using the GeneSeek® GGP Bovine 150k Array were retrieved. Genomic data were subjected to quality control. Milk- and cheese-related traits were subsequently subjected to descriptive and statistical analyses, including comparative evaluations among traits and integration with genetic information. Statistical approaches included: Pearson’s correlation analyses, parametric (t-test) and non-parametric (Wilcoxon test) comparisons, and linear mixed models (LMM) to estimate genetic and genomic heritability of milk and cheese-related traits. Heritability estimates for selected novel phenotypes (moderate values were obtained) confirmed the potential inclusion of these traits in breeding strategies. In addition, significant correlations were identified between key milk composition parameters and cheese-making efficiency traits. Metabolomic data of plasma underwent rigorous quality control procedures. Missing values within metabolomic profiles were imputed using the Multiple Imputation by Chained Equations (MICE) approach to preserve statistical power and reduce bias. Prior to core integrative analyses, the impact of relevant fixed effects (including herd, month of sampling, parity, and days in milk) was evaluated using linear models implemented in R tool. This preliminary step allowed adjustment for systematic environmental and management factors, ensuring that subsequent analyses focused on biologically meaningful variation. Core analyses included the assessment of relationships between internal and external phenotypes (milk recording productive, functional, and health traits; milk processing parameters; and cheese-yielding ability traits). Associations were investigated through correlation and regression-based approaches. Principal Component Analysis (PCA) was performed to identify clusters of animals and to evaluate patterns potentially influenced by fixed effects or biological status. To further explore relationships and potential causal structures between external and internal phenotypes, Structural Equation Modelling (SEM) was applied. This approach enabled the modelling of direct and indirect effects among metabolomic, productive, and technological traits. All analyses were conducted using R tool. Genomic data (SNP genotypes) were subsequently integrated with external phenotypes using a Genome-Wide Association Study (GWAS) framework implemented in the GEMMA software. Linear mixed models were applied, incorporating both fixed and random effects to account for population structure and relatedness. Multiple testing correction was performed using the Bonferroni approach (P = 0.05 / 150,000 markers), and the most significant associations were retained. For each significant SNP, the surrounding genomic region was investigated to identify putative Quantitative Trait Loci (QTL) and candidate genes potentially exerting large genetic effects on the analysed traits. Several genomic regions were identified for different traits. Similarly, GWAS analyses were performed using metabolomic traits (mGWAS) as dependent variables. This analysis led to the identification of 20 genomic regions associated with specific metabolomic traits (e.g., lipids). Finally, genomic regions identified through GWAS and mGWAS were examined in whole-genome resequencing data to detect potential causative variants segregating within the studied population. The project demonstrate how multi-omics integration can enhance the understanding of genotype-phenotype relationships and support precision breeding strategies in local cattle populations.Dettagli del progetto
Responsabile scientifico: Stefania Dall'Olio
Strutture Unibo coinvolte:
Dipartimento di Scienze e Tecnologie Agro-Alimentari
Coordinatore:
ALMA MATER STUDIORUM - Università di Bologna(Italy)
Contributo totale di progetto: Euro (EUR) 206.614,00
Contributo totale Unibo: Euro (EUR) 126.500,00
Durata del progetto in mesi: 24
Data di inizio
12/10/2023
Data di fine:
28/02/2026
