Genetic study of Babesia bovis infection level and the association with tick resistance in Hereford and Braford cattle

Author(s)

Ligia Cavani

• Document type: Doctoral Thesis
• Press: Jaboticabal. 2019-04-03.
• Institution: Universidade Estadual Paulista (Unesp). Faculdade de Ciências Agrárias e Veterinárias. Jaboticabal
• Defense date: 2019-02-19
• Advisor: Henrique Nunes de Oliveira; Fernando Flores Cardoso; Rodrigo Giglioti

Abstract

Bovine babesiosis is a tick-borne disease, and the Babesia bovis is considered the most pathogenic species. Both parasites constitute major drawbacks for improvement of beef cattle productivity in the tropics, especially when purebred and crossbred taurine animals are used. This study analyzed a population of Hereford and Braford cattle and it was composed of four chapters with the following objectives: Chapter 1) Literature review; Chapter 2) Estimate genetic parameters for tick count (TC) and B. bovis using linear and generalized linear models; Chapter 3) Evaluate predictive ability and application of genomic selection and performed genome wide association studies (GWAS) for B. bovis infection level (IB) using single step GBLUP model; Chapter 4) Search for causal structures to investigate potential functional relationships among TC, IB, weight gain from birth to weaning (WG), and weight gain from weaning to yearling (YG) using structural equation modeling (SEM). Tick counts were performed by manually counting adult female ticks on one side of each animal. The B. bovis quantification was performed using a qPCR assay. In the Chapter 2, the tick count and B. bovis records were in log scale for analysis using linear model. A Poisson model was applied for tick count without log transformation and for probit model the phenotype was assessed by absence or presence of B. bovis based on three different thresholds (BBt1: IB using the threshold observed; BBt2: BB using threshold as a Cq (quantification cycle) mean from a qPCR assay equal to 38; BBt3: BB using threshold as a Cq mean from a qPCR assay equal to 37). In the Chapter 4, statistical analyses were conducted in three steps: 1) Partition of genetic and residual (co)variances using Bayesian multiple-trait modeling (MTM); 2) Search for plausible causal structures using the inductive causation (IC) algorithm applied to the residual (co)variances obtained from the MTM analysis; and 3) Final analysis using SEM. Estimates of heritability for TC were high using Poisson model (0.363) and moderate using linear model (0.189). While for B. bovis heritability estimates were low corresponding to 0.021 for BBt1, 0.027 for BBt2, 0.041 for BBt3 for probit model, and 0.121 for linear model. The estimated correlation between IB and TC was very low. The cross validation genomic prediction accuracy for IB ranged from 0.18 to 0.35, and from 0.29 to 0.32 for k-means and random clustering. The GWAS results showed that the top 10 SNP explained a total of 5.04% of the genetic variance for IB, and identified 40 candidate genes. The most plausible SEM comprised three direct links between traits: WG→YG, TC→WG, and WG→IB with structural coefficients posterior means equal to -0.3026, 6.3620, and 0.0004, respectively. The selection for TC and for animals that possibly maintain lower levels of infection by B. bovis will render short-term and long-term responses, respectively. Genomic predictions could be used as a tool to improve genetics for IB, especially for larger and randomly training populations. Some genomic regions associated with candidate genes related to immunity system were identified. The final inferred directed acyclic graph suggests that WG imposes a negative causal effect on YG, and TC has a positive causal effect on WG. (AU)

• FAPESP's process: 15/13024-0 - Genomic study on infection rate of Babesia bovis and association with resistance of tick in Hereford and Braford cattle
• Grantee: Ligia Cavani
• Support Opportunities: Scholarships in Brazil - Doctorate