Functional investigation of a QTL region affecting resistance to Haemonchus contortus in sheep Guillaume Sallé 2, Carole Moreno 1, Julien Ruesche 1, Frédéric Bouvier 1, Mathias Aletru 1, Jean-Louis Weisbecker 1, Françoise Prévot 2, Jean-Paul Bergeaud 2, Cathy Trumel 3, Christelle Grisez 2, Dominique François 1, Andres Legarra 2, Emmanuel Liénard 2,3, Philippe Jacquiet 2,3 1 INRA Animal Genetics Department (UR631, UE632, UE) 2 INRA Animal Health Department (UMR 1225) 3 Toulouse University, INP - Ecole Nationale Vétérinaire de Toulouse EAAP 2013, Nantes, France
INTRODUCTION Gastro-intestinal nematodes : a curse for sheep breeding Impact sheep health and welfare Strong economical burden for breeders (~85 M in the UK Nieuwhof & Bishop 2005) Anthelminthic resistant populations have been selected for worldwide Breeding for more resistant animals could help controlling nematodes in a more sustainable fashion ~30 % of the observed variation in resistance due to the genetic background Strong between breeds differences Need to know the genes : good understanding of mechanisms + improve breeding schemes.02
Our working basis (Aumont et al. 2003, Terefe et al., 2007) Martinik Black-Belly Subtropical breed Resistant Romane French meat breed Susceptible Martinik sheep are far more resistant than the Romane sheep to H. contortus 3 to 10 fold less excreted H. contortus eggs Phenotypic differences also observed on the immunological background Higher and quicker eosinophilic infiltration of the abomasal mucosa IL5 and IL13 over-expressed in Martinik abomasal mucosa Key differences observed at 1 st infection Next question: what are the genetic mechanisms controlling these differences?.03
SUMMARY OAR12: an interesting candidate Investigating the functional properties of this region Conclusion & perspectives.04
Looking for regions associated to H. contortus resistance Martinik Romane F 1 1000 Back-cross B F C B F C Identify regions of the genome affecting resistance to H. contortus 1,000 lambs experimentally challenged twice Measured for Fecal Egg Counts and Haematocrit Genotyped for 50K SNP Choose one interesting region.05
Association of 4 SNPs with resistance to Haemonchus contortus One region affecting FEC each time the sheep faced H. contortus Second highest effect (0.19 σ p ) Found in other sheep populations: Sarda*Lacaune (Moreno et al., 2006) Merino (Beh et al., 2002) Soay sheep (Beraldi et al., 2007) G. Sallé et al., J. Anim. Science, 2012.06
The 4-SNP allelic effects Effect (phen. std) 0,2 0,1 0-0,1-0,2-0,3-0,4-0,5-0,6-0,7-0,8-0,9-1 -1,1-1,2-1,3 Allelic effect (phenotypic standard deviation) 0 2 4 6 8 10 12 14 Other alleles: 4SNP_Un Alleles with favorable effect 4SNP_R Two extreme alleles with unfavorable effects 13 alleles have been identified 3 favorable alleles versus 2 very unfavorable alleles One allele from the Romane breed within the favorable alleles (!) Martinik (resistant) origin Romane (susceptible) origin.07
SUMMARY OAR12: an interesting candidate Investigating the functional properties of this region Conclusion & perspectives.08
The experimental design Martinik Romane F 1 1000 Back-cross B F C B F C Identify regions of the genome affecting resistance to H. contortus Choose one interesting region BC x BC matings to generate offsprings Selection of allelic carriers Functional comparison.09
4SNP_R Functional validation BC x BC sheep 4SNP_Un Genotyping Two groups of BCxBC sheep selected on the 4SNP alleles they inherited : 4SNP_R animals carried the favorable alleles 4SNP_Un animals inherited neutral alleles - Determination of the inherited alleles - Estimation of the genomic background effect Effect (phen. std) 0,2 0,1 0-0,1-0,2-0,3-0,4-0,5-0,6-0,7-0,8-0,9-1 -1,1-1,2-1,3 Allelic effect (phenotypic standard deviation) Not found in the BCxBC flock 0 5 10 15 Alleles with high favorable effect : 4SNP_R Other alleles: 4SNP_Un.010
Functional validation 2 BC*BC groups based on their 4SNP alleles Equivalent genomic value between the two groups Comparison of egg excretions Comparison of worm burden, females fertility and gene expressions 4SNP_R 4SNP_Un 44 challenged (16 vs 28) D0 15 dpi 18 dpi 21 dpi 24 dpi 27 dpi 30 dpi FEC FEC FEC FEC FEC Hematology Hematology Necropsy.011
An effect on fecal egg counts FEC in 4SNP-based groups 45000 40000 * FEC (eggs/g of feces) 35000 30000 25000 20000 15000 10000 5000 * * 0 FEC18 FEC21 FEC24 FEC27 FEC30 4SNP_R 4SNP_Un 4SNP_R sheep excreted less eggs than 4SNP_Un.012
The effect on FEC could be explained by a difference in female worms fecundity For each BC*BC infected sheep: The length of 35 H. contortus females was measured The number of in utero eggs of 20 H. contortus females counted Avg. female worms length( mm) 22 21 20 19 18 17 16 15 Female length * 4SNP_R 4SNP_Un No. eggs in utero 800 700 600 500 400 300 200 100 0 Avg. fertility of female worms * 4SNP_R 4SNP_Un.013
No differences for worm burden 4SNP_R lambs also exhibited significantly less blood loss at 30 dpi.014
Any candidate genes? Selection of the most frequent alleles in each group: GAAG MBB vs GACA RMN (N = 9) (N =8) Effect (phen. std) Allelic effect (phenotypic standard deviation) 0,2 0,1 0-0,1 0-0,2 5 10 15-0,3-0,4-0,5-0,6-0,7-0,8-0,9-1 -1,1-1,2-1,3 Gene expression compared between carriers - Set of differentially expressed genes from micro-array experiment between pure breeds MBB and RMN sheep (Liénard et al., 2011) - PAPP-A2 (pappalysin) gene underlying the 4SNP region (regulation of Insulin-like Growth Factor (IGF) activity).015
Any candidate genes? IL-13 IL-4 No. IL3 mrna copies /100 REF mrna copies * No. IL4 mrna copies /10 4 REF mrna copies * Stronger Th2 response Abomasal mucosa, 30 days post-infection PAPP-A2 PAPP-A2 was not differentially expressed between the two groups No. PAPP-A2 mrna copies /10 4 REF mrna copies.016
SUMMARY OAR12: an interesting candidate Investigating the functional properties of this region Conclusion & perspectives.017
CONCLUSION & PERSPECTIVES Take home message A 4SNP region in OAR12 was associated to sheep resistance This functional study seems to confirm its role This region seems to impact worms fertility (1 σp) This region was associated to a Th-2 biased over-expression Perspectives Refine : association analysis in Martinik and Romane sheep Investigating the host-nematode system as a whole.018
ACKNOWLEDGEMENTS Supervisors Pr. P. Jacquiet Dr. C. Moreno Quantitative genetics S.C. Bishop J.M. Elsen A. Legarra O. Filangi Parasitology J.-P. Bergeaud C. Grisez E. Liénard F. Prévot Lab Farm interface D. François J. Ruesche Experimental farms staff M. Aletru Y. Bourdillon F. Bouvier T. Fassier D. Marcon J-C. Thouly J. L. Weisbecker Financial support Thank you for your attention G. SALLĒ / Functional study of a QTL affecting resistance to H. contortus.019 EAAP 2013, Nantes, France
.020
Merging knowledge together Positional knowledge Regions involved No gene FEC1 FEC2 OAR12 Differential gene expression (Liénard et al., WAAVP 2011) Breed comparison Control and 8 dpi Tissue samples from abomasum and lymph nodes Comparison of the gene expression of ~15,000 genes (microarray) Uninfected control Infected 8 dpi
Merging knowledge together Uninfected control Infected 8 dpi Gene network analysis 157 genes with fold change > 1.5 FEC1 OAR12 Gene set FEC2 Significant QTL Annotated genes contained within LRTmax position +/- 4 Mbp Networks
Results of the bioinformatic approach Within the networks related to haemonchosis: a high number of genes mapped on OAR12 No. genes involved in immunity/hemostasis/plc-a2 activity No. genes 10 9 8 7 6 5 4 3 2 1 0 7 8 9 12 13 14 15 17 20 21 22 23 26 OAR
Association analysis & breeds Due to the two pure breeds histories, SNP alleles may not mirror the same QTL alleles * Relationship between r of the Martinik breed and the Romane breed (0-10 Kb) m Q* n M Q* N M q N M Q* N * The mutation can have appeared after breeds diverged
Population genetics input Different environments different selection pressure Stronger selection pressure due to GIN in MBB sheep Is there any gene that has been selected in the MBB sheep and not in the RMN breed?
Sweep detection In collaboration with S. Boitard Additional genotyping of MBB and RMN individuals One suggestive sweep at 42 M Fitting the sweep genotype in the QTL detection model erodes the QTL profile Classical model Sweep as a fixed effect