Investigation of variation in genes influencing fertility in New Zealand sheep : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

An important trait in commercial sheep breeding is the number of lambs born per ewe, because the amount of meat produced per ewe is to a great extent determined by litter-size. Accordingly, the Identification of functional variation in genes that are responsible for improving fertility, would potentially allow for flocks to be bred for increased fertility, and thus increase profitability in the NZ sheep industry. Fertility if realised as increased fecundity, would not only be a determinant of profitability but may also affect the carbon footprint of New Zealand livestock production systems. In this research, three genes involved in regulating fertility in sheep were investigated. Polymerase Chain Reaction – Single Strand Conformational Polymorphism (PCR-SSCP) analyses were used to search for genetic variation in three genes, the Growth Differentiation Factor 9 gene (GDF9), the Bone Morphogenetic Protein 15 gene (BMP15), and the Bone Morphogenetic Protein Receptor Type 1B gene (BMPR1B). Once identified by PCR-SSCP the genetic variation was further characterised with DNA sequencing. Confirmation of the sequence variation, then enabled subsequent testing of whether the variation was associated with variation in fertility in three sheep breeds (Finnish Landrace, Finnish Landrace X Texel and composites) using best linear unbiased prediction (BLUP) and ASREML with both animal and sire models. In this study, the number of sheep studied was 1064 for the GDF9 gene and 852 for the BMP15 gene. A total of 241, 251 and 335 ewes were analysed for GDF9, BMP15 and BMPR1B respectively. These included NZ Finnish Landrace sheep, Finnish Landrace × Texel-cross sheep, and composite sheep (farm 1) (of varying breed background). These three breeds derived from a single large ewe flock farmed on pasture and all fed in the same way in North Canterbury. All ewes had records for the 2016 lambing season, hence the number of lambs born in 2016 were used for association study. In the sheep studied, variations in ovine GDF9 and BMP15 were associated with litter-size. Finnish Landrace × Texel-cross sheep with the c.1111A variant of GDF9 were found to be more fertile (P = 0.036) than those without c.1111A. In animal models, the effect of GDF9 appeared to be additive, with one copy of c.1111A increasing litter-size by 0.43 ± 0.202 in the Finnish Landrace x Texel-cross ewes, and two copies increasing litter-size by 0.86. No such effect was seen Finnish Landrace and composite sheep. However, the impact of a single copy of c.1111A led to an increase in litter-size of 0.34 ± 0.154 (p = 0.027) compared to those ewes with c.1111G, when all the sheep groups were analysed together. In contrast to the c.1111A>G results, litter-size did not differ between sheep with and without GDF9 c.994A in all three groups of sheep investigated. The c.31_33del in BMP15 was found to be associated with litter-size (P < 0.001) in composite sheep. The effect of the presence of one copy of c.31_33del was an increase of 0.26 ± 0.092 (P = 0.008) lambs compared to those ewes without c.31_33del using the animal model. The estimate for the effect of variant A (absence of the c.31-33del) in the composite sheep was -0.26 ± 0.092 (p = 0.008) and -0.22 ± 0.095 (p = 0.026) in both the animal and sire models, respectively. This association between the detected c.31-33del and litter-size was not observed for Finnish Landrace or the Finnish Landrace x Texel-cross (P > 0.05). It is possible that the effect of this deletion in the signal sequence seems to vary from study to study and breed to breed. Sequence analysis of a 394 bp fragment spanning the partial exon 9 and intron 8 and a 338 bp of exon 8 and intron 7 regions of BMPR1B in 335 sheep belonging to three groups of New Zealand sheep of differing background, revealed 5 variant sequences with a total of six single-nucleotide substitutions. The sequencing results revealed nucleotide substitutions c.1032T>C in the amplified region of exon 9/intron 8 and c.754-144G>A, c.754-88G>A, c.762G>A, c.754-31C>T and c.765G>A in the amplified region of exon 8/intron 7. Despite the presence of six nucleotide substitutions (found across two regions) in BMPR1B, no association was found between the sequence variation and litter-size (p > 0.05). This gene may not play a significant role in the fertility of the New Zealand sheep breeds investigated. The only modest (but not statistically significant, p = 0.162) association of intron 7/exon 8 was the effect of variant C on increased litter-size in composite sheep (0.23 ± 0.167). The impact of variant B in Finnish Landrace sheep (-0.04 ± 0.239 lambs, P = 0.861) is very similar to the effect of the variant B when all the groups were analysed together -0.04 ± 0.146 (P = 0.747). The identification of functional sequence variation in the breeds studied here, may at first be of limited value to breeds that do not have the observed variation, but it lays a strong foundation to further this type of analysis with more common New Zealand breeds.