Selective breeding for stress response in common carp (Cyprinus carpio L.) using androgenesis

Summary

The aim of the thesis was to explore the genetic background of stress response in common carp ( Cyprinus carpio L.) and produce homozygous and heterozygous isogenic strains with divergent stress responses. As stressor a rapid temperature decrease (= cold shock) was used. As a preparatory step, a number of experiments were carried out to 1) investigate the validity of the cold shock as a stressor, 2) investigate the possible influences of environmental factors on the stress response, and 3) define a selection criterion for the selection experiment. The stress response of common carp was studied by evaluating plasma cortisol, glucose and lactate after single or multiple rapid temperature drops (ΔT: 7, 9 or 11 °C). All three amplitudes used induced a significant rise in plasma cortisol levels. Peaks occurred within 20 min after onset of the cold shock. No stress-related related secondary metabolic changes were observed in any of the experiments described: plasma glucose levels remained unaffected and plasma lactate levels dropped. Carp of 60 days old showed a significant stress response, although plasma cortisol levels were lower than those observed in carp of 120 days. Furthermore, fish that had experienced multiple cold shocks showed an overall lower cortisol response than fish experiencing a single cold shock, indicating that habituation to this stressor occurred. Based on these results, the plasma cortisol concentration at 20 min after onset of the cold shock was set as selection criterion in our selection experiment and fish were tested at a minimal age of 100 days post hatching.

The first step in the actual selection experiment was the formation of the base population. This base population was an F ₁ cross between six sires from a wild strain originating from the Anna Paulowna (AP) polder and a highly domesticated homozygous E4 dam already present in our laboratory. The six sires, caught in the water system surrounding the Anna Paulowna (AP) Polder in The Netherlands, were characterised using allozyme and microsatellite markers. At the sMDH-A1,2* loci an allele was found, which was previously only found in wild River Rhine and wild Vietnamese common carp. Microsatellite allele frequencies showed that these AP carp were significantly different from a group of carps originating from several different domesticated strains. Based on both allozyme and microsatellite data, the AP carp most likely originated from a wild or feral self-sustaining population.

Thirty-three randomly picked sires from these six E4 × AP full-sib families (F ₁ ) were androgenetically reproduced to create the F ₂ generation, which thus consisted of 33 doubled haploids (DH) progeny groups. These 33 DH progeny groups (566 individuals) were subjected to the 9 °C cold shock, enabling us to estimate heritabilities for weight, length, condition factor (K), and plasma cortisol, glucose and lactate concentrations using Gibbs sampling and an animal model. Estimated heritabilities for the morphological traits weight and length were 0.09 (90% Highest Posterior Density range: 0.03 - 0.17) and 0.11 (0.04 - 0.21), respectively. The condition factor (K), showed a medium heritability of 0.37 (0.20 - 0.62). Heritabilities for basal plasma glucose and lactate were 0.19 (0.10 - 0.33) and 0.56 (0.33 - 0.85), respectively. For stress-related cortisol increase a high heritability estimate of 0.60 (0.37 - 0.90) was found. Although the height of this cortisol heritability has to be regarded with some reservation, due to confoundedness of some environmental effects with sire effects, the estimated heritability clearly shows that the stress response due to a cold shock is hereditary in the carp population used.

Because the model used to estimate the h² assumed a complete homozygous state of the DH individuals and to ensure that only homozygous individuals would be used for subsequent reproduction, all individuals within the 33 DH androgenetic progeny groups were analysed using 11 microsatellite markers to: 1) verify the homozygous status of the 566 androgenetic DH individuals, 2) analyse the microsatellite allele segregation, and 3) study possible association of microsatellite alleles with the phenotypic traits recorded. In total, 92% of the androgenetic DH individuals proved to be homozygous at all 11 loci. Forty-three out of the 47 heterozygous individuals were heterozygous at a single locus only. This heterozygosity was probably due to DNA fragments caused by UV-irradiation of the eggs, although the maternal origin of the fragments could not be proved beyond doubt. Screening with 11 microsatellites also revealed two linkage groups, a segregation distortion at two microsatellite loci and possible association of some microsatellites with weight, length, stress-related plasma cortisol levels and basal plasma glucose levels. The success of the linkage and association study could be explained by a low recombination frequency due to high chiasma interference. This would imply a relatively short genetic map length for common carp.

Selection of individual fish from the 33 DH progeny groups based on the response at 4 months was not possible. Therefore, three DH progeny groups with a high (H1-3) and three with a low (L1-3) mean plasma cortisol concentration were selected. The 154 DH fish in these six groups were individually tagged, mixed and subjected to a second cold shock at an age of 15 months. For each individual fish, a breeding value was estimated (EBV ₁₅ ) for stress-related cortisol using an animal model with a fixed h ² of 0.60. Two homozygous sires (two high and two low) and dams (high and low) were selected based on their EBV ₁₅ and used to produce four homozygous (HomIso) and eight heterozygous isogenic (HetIso) strains. These were used in two separate experiments to examine the genetic background of the stress-related cortisol response. In both experiments, the strains were subjected to the 9 °C cold shock at an age of 5 months. The ranking in plasma cortisol levels of the HomIso strains was identical to the ranking in EBV ₁₅ of the sires and the maximal difference of 350 nmol.l ^-1 was similar to the expected difference based on these EBV ₁₅ 's. Differences between the HetIso strains were smaller than expected, and influence of non-additive genetic effects could not be detected ( P_D×S = 0.14). Estimated breeding values based on the performance of the androgenetic progeny (EBV ₅ ) in experiment 1 and general combining abilities (GCA) of the sires and dams calculated in experiment 2 were positively correlated with the EBV ₁₅ ( r not significantly different from 1), providing no evidence that the stress response at 5 and 15 months are different traits.

Apart from the isogenic strain used in the first experiments, no complete profiles of the cortisol, glucose and lactate dynamics had been examined in other isogenic strains. Therefore, an additional experiment, parallel to the selection experiment, was carried out to investigate the 'complete' cortisol, glucose and lactate dynamics during the cold shock in four, readily available, isogenic at two different ages. The experiments showed that stress-related cortisol response patterns can differ consistently between genotypes of common carp. Significant dam and sire effects on the total amount of cortisol measured during the cold shock were found, but no significant dam × sire interaction effect. Although no significant difference was found between the cortisol response dynamics at 5 or 7 months, the results justify further research into that field. The observed differences in plasma glucose and lactate dynamics between control and shocked fish were most likely temperature related. Age did not have any apparent influence on either plasma glucose or lactate dynamics in both control and shocked fish.

Based on the results of the experiments described in this thesis, it can be argued that the best method to change the stress response of common carp would be through selective breeding (exploiting additive genetic effects) rather than through crossbreeding (exploiting non-additive genetic effects). The selection and the 'parallel' experiments resulted in several isogenic strains of common carp with at least two types of cortisol stress responses. Type I showed a relative short cortisol response with either a high or low peak at 20 min after onset of the shock. Type II showed a similar cortisol level at 20 min but no significant decrease in this level during the cold shock. These different isogenic strains will be valuable tools in future research into the stress response itself and its effects on other traits like growth, reproduction and health. This way, some of the problems related to the use of stress response as selection criterion in commercial breeding programmes in fish could be solved in the near future.

Residual heterozygosity was demonstrated to occur in androgenetic progenies, most likely due to maternal DNA fragments induced by the UV irradiation of the eggs. Improved control measures were implemented in the androgenesis procedure, but androgenetic progenies destined for further reproduction purposes should be screened for residual heterozygosity. Androgenetic reproduction proved to be a useful tool for dissection of phenotypic variance and heritability estimations for traits, especially in combination with selection experiments aimed at development of isogenic strains for this trait. Androgenesis might result in reduced fertility in female progeny, but the advantages are such that inclusion of androgenetic reproduction within larger commercial breeding programmes for faster dissemination of genetic progress and product protection should be considered as a promising option.