Last update: September 8, 2011 09:46:16 AM E-mail Print

 

A REVIEW ON THE USE OF MOLECULAR TECHNOLOGY FOR THE GENETIC IMPROVEMENT OF SOUTH AFRICAN ANGORA GOATS

 

C. Visser# & E. van Marle-Köster

Department of Animal & Wildlife Sciences, Faculty of Natural & Agricultural Sciences, University of Pretoria, Pretoria, 0002

#E-mail: Carina Visser

 

 

INTRODUCTION

Selection based on quantitative theory has been the basis for genetic improvement of livestock for many decades. Livestock breeders world-wide have over time replaced subjective scores and judging of animals with the use of objective animal recording of traits of economic importance. Animal recording and development of quantitative methodology ultimately resulted in selection being based on best linear unbiased prediction of estimated breeding values (BLUP EBVs) in many breeds. Selection based on recorded data is a major advantage and lead to genetic progress in many economically important traits. The application of standard EBVs can, however, be a relatively slow and costly procedure (Allan & Smith, 2008) and the additive model of inheritance used in BLUP analysis has been shown to have inherent limitations (Dodds et al., 2007). DNA technology can overcome some of these limitations and could be applied in genetic improvement programs in various ways, from parentage verification and traceability to the identification of quantitative trait loci (QTL) and whole-genome selection.  The major aim of these practices is to improve EBV accuracies and selection efficiency, resulting in faster genetic progress and reduced cost of altering genetics (Allan & Smith, 2008). 

 

The South African Angora goat industry has relied on mass selection to genetically improve mohair-producing goats over the past few decades. Selection practices have evolved over time from an over-emphasis on fleece weight to a more balanced selection index approach. Currently three traits (fleece weight, fibre diameter and body weight) are included in the index with moderate success in the national herd. An evaluation of the goats subjected to selection based on this index has indicated that fibre diameter can be decreased without a decrease in body weight (Snyman, 2002). Quantitative selection is generally practiced in the stud breeding industry with the application of BLUP breeding values. The South African Angora goat stud breeders have, however, very limited participation in the National Small Stock Recording and Improvement Scheme and limited use is made of EBVs.

 

Although the global mohair industry has a relatively stable niche market, natural fibres are under severe pressure from competing synthetic textiles. The mohair industry needs to align its production profile to the market demands, which include a greater focus on quality and uniformity of fibres. In order to maintain their global stronghold, the South African mohair industry assigned priority to projects for investigation and application of new technology, including molecular technology. Biotechnology and genomics can benefit the goat industry in a number of ways, including pedigree verification, traceability of products, diagnostic tests and marker assisted or gene assisted selection (MAS or GAS). The aim of this review is to provide an overview of the recent research performed by the Department of Animal and Wildlife Sciences at the University of Pretoria on South African Angora goats using molecular technology.

 

MOLECULAR TECHNOLOGY FOR GENETIC IMPROVEMENT

Even though the advantages of molecular technology have been proven in other species, small stock and especially goats have received limited interest from researchers. The mohair industry of South Africa acknowledged both the limitations of quantitative selection and the opportunity that DNA technology poses for livestock improvement.  The Department of Animal and Wildlife Sciences at the University of Pretoria has established a focus area on goat research, including the genetic improvement of the Angora using a molecular approach. Discussions between Mohair South Africa, the Department of Agriculture, Fisheries and Forestry (mainly the Grootfontein Agricultural Development Institute, GADI) and the University of Pretoria resulted in the initiation of a DNA biobank for small stock in 2005, with the further intent of molecular research. In 2006, the GADI-biobank was established at GADI (Snyman, 2011). Blood is currently collected annually after kidding seasons and stored together with the related phenotypic and pedigree records of animals at this facility.

 

All animals from four Angora goat producers (three stud breeders and one commercial producer), as well as the goats at the Jansenville Experimental Station of the Eastern Cape Department of Agriculture, were sampled as a base population for participation in the GADI-biobank. These animals formed the Angora goat reference population and results based on this population will therefore be applicable to the whole Angora goat population in South Africa. Several research questions were answered using this population to evaluate how molecular information could contribute to the genetic improvement of the South African Angora goat. 

 

GENETIC DIVERSITY

The estimation and maintenance of genetic variability in conservation of breeds is one of the advantages of using DNA-based technology.  Genetic diversity, phylogenetic relationships between breeds and co-ancestry can be investigated using genotypic information.  To date, genetic diversity studies on South African goats were limited to a study on commercial meat goat breeds (Visser et al., 2004) and one broader phylogenetic study that included Angora goats, three meat goat breeds as well as indigenous goat populations (Pieters, 2007). Using DNA marker technology, the genetic variability in the SA reference population was found to be relatively high. Ninety-four microsatellite markers spread over the goat genome were amplified in 1067 Angora goats. Genetic variability of the families selected for the reference population was analysed using MS Toolkit (Park, 2001). Genetic parameters estimated included allelic frequencies, mean number of alleles and heterozygosity values per locus and for each population.  The polymorhpic information content (PIC) for each locus and across loci were estimated using Cervus 3.0 software (Marshall et al., 1998).  The FSTAT2.9.3 program (Goudet, 1995) was used to compute Wright’s F-statistics for each locus, including F, θ and f (analogous to Wright’s (1978) FIT, FST and FIS, respectively). 

 

A total of 800 alleles from 94 loci were detected in the 1067 individuals genotyped.  All markers were found to be polymorphic in each of the four herds evaluated.   The number of alleles identified per locus averaged 7.99, with variation from two to 23. The mean PIC value across loci was 0.57, indicating a medium level of information. Genetic variability in the SA reference population was found to be relatively high with the average number of alleles varying between 5.41 and 7.21 in the four herds.  The estimated unbiased heterozygosity or gene diversity was well above 60%, except for one herd with a value of 56.5% (Table 1). With regards to population subdivision the FST value (0.182) for herd 2 indicated a reduction of heterozygosity supporting the unbiased heterozygosity estimation (Hartl, 1988). These levels of heterozygosity exceeded expectations as the Angora goat population in South Africa is relatively small and high selection pressure has been applied to the animals over several generations. The observed and expected heterozygosity values over all loci for all herds averaged 0.63 and 0.62 respectively. 

 

Table 1. Measures of genetic variation in the population studied

Herd

Sample size

Loci typed

Unbiased heterozygosity ± SD

Observed heterozygosity ± SD

N Alleles

FST

1

400

94

0.627 ± 0.015

0.637 ± 0.003

6.98

0.0658

2

218

93

0.565 ± 0.018

0.592 ± 0.004

5.41

0.1818

3

338

94

0.633 ± 0.014

0.652 ± 0.003

7.21

0.0659

4

111

93

0.634 ± 0.016

0.671 ± 0.005

6.87

0.0486

 

The different herds included showed higher than expected heterozygosity values and very limited inbreeding. The levels of genetic diversity compared favourably with similar studies on global goat populations, indicating the possibility to exploit natural variation on a molecular level for improved production.

 

PARENTAGE VERIFICATION      

DNA based parentage verification has found wide application in the livestock industry, especially for seed stock producers where accurate pedigrees are essential for estimation of breeding values (Van Eenennaam et al., 2007; Van de Goor et al., 2009). The predicted loss in genetic progress using incorrect pedigree information has been shown to vary between 2 and 15% with a parentage error rate of between 10 to 15% (Geldermann et al., 1986; Pollak, 2005). In commercial sheep populations it was estimated that up to 15% of parentage recordings might be incorrect (Dodds et al., 2007). Accurate recording of parentage in South African Angora goats may be quite burdensome for breeders farming under extensive systems with large herd sizes. Most stud breeders have both stud and commercial herds. It had been estimated that of the Angora kids born between 2000 and 2005, 23% had incomplete or inaccurate pedigree data, with unknown sires posing the main limitation. Inaccurate parentage recording over time result in lower selection efficiency due to mating based on incorrect pedigree data.

 

Eighteen microsatellite markers for parentage testing in goats were recommended by the International Society of Animal Genetics (ISAG) in 2002 and 2005 (http://www.isag.org.uk/Docs/2005_PanelsMarkers SheepGoats.pdf). These markers were tested in several laboratories and compared well with regard to individual performance, but test results of these marker panels vary in different goat breeds with regard to the polymorphicity and heterozygosity levels. These parameters have a direct impact on the exclusion probabilities and should therefore be tested in the specific population as no commercialised panel for goats is available. These microsatellite markers recommended by the international organisation together with markers used in other SA Angora goat projects were tested in a number of families with known parentage to verify the parameters for parentage verification (Friedrich, 2009). The final panel consisted of 14 markers and was tested in a test family where no information was provided on the family pedigree. Ninety-five percent of the animals could be allocated to the correct sire using this panel (Friedrich, 2009). The project was performed with the financial support of both Mohair SA and NRF (THRIP) and the technology and markers are now available in South Africa for application by Angora farmers. This is most useful for stud breeders where premiums may be involved in marketing of superior bloodlines. It is envisaged that this will benefit the industry and will result in greater selection accuracy and rate of genetic progress in future.

 

A GENETIC LINKAGE MAP FOR SA ANGORA GOATS

A genetic map forms the foundation for DNA research for any species, detailing the specific DNA marker positions on the genome.  The goat linkage map is relatively underdeveloped (Maddox & Cockett, 2007) and in effect still quite primitive when compared to most other livestock species, thus posing a limitation for any molecular study on goats. This prompted a project that aimed to improve the status of the map and to verify previously reported discrepancies with the sheep map. Microsatellite markers (n=134) were selected from the existing goat map database (http://locus.jouy.inra.fr/cgi-bin/lgbc/mapping/ common/intro2.pl?BASE=goat) and previously published literature to obtain sufficient genome coverage. Markers were first screened for polymorphicity and amplification success in 19 potential Angora sires, obtained from the GADI-biobank.  Finally, the 12 sires (with offspring) that expressed the highest levels of heterozygosity across all loci tested were selected for inclusion in the study.  Due to misidentification and recording errors that occur during mating under extensive production systems, the data were analysed using Cervus 3.0 (Marshall et al., 1998) and all aberrant individuals were removed from the study. Linkage analysis was performed with CRI-MAP 2.4 (Green et al., 1989) compiled for Windows XP. The use of a relatively large number of animals for this study resulted in a significant improvement in the number of informative meiosis and shorter mapping distances. New markers were mapped to the goat genome and previously reported inter-chromosomal inversions were corrected. An example of this is shown in Figure 1.

 

This lead to the generation of an improved linkage map, which provides the basis for an advanced linkage map. This more complete and accurate map will assist with comparative mapping and should lead to the opportunity of mapping genetic variation that is responsible for the phenotypic differences in economically important traits.

 

SEARCH FOR QUANTITATIVE TRAIT LOCI

The identification of specific regions of interest and the selection thereof (either through the use of closely linked markers (MAS) or of the causative mutation itself (GAS)) has been the aim of many molecular studies. The selection for chromosomal areas that directly contribute to the genetic variation of traits of economic importance will lead to increased genetic progress and offers the opportunity to better understand and exploit phenotypic variation (Dekkers, 2004).  This has however mainly remained a theoretical concept in many species.  Implementation of marker assisted selection in goats has been limited to the use of selection against diseases (scrapie, CAEV and Johne’s disease) and for casein genes in dairy goats (Van der Werf, 2007), but no MAS / GAS is currently practiced for fibre traits (Dodds et al., 2007). 

 

Phenotypic variation in economically important traits is caused by a large number of genes spread over the genome, commonly referred to as Quantitative Trait Loci (QTL). Identifying causative mutations for economically important traits is dependent on QTL identification and mapping and such a study was performed on the reference population of the GADI-biobank. The results of this study indicated several QTL of medium effect influencing mohair production and quality and explain some of the genetic variance in mohair traits (Table 2).

 

Figure 1. Alignment of marker orders on CHI 2, 4, 5, 11, 13 and 19 with orders compared between the current study, the goat map of Schibler et al. (1998) and SheepMap4.7 (http://www.ncbi.nlm.nih.gov/mapview/static/sheepsearch.html)

 

Table 2. Putative QTL identified for seven traits per chromosome

CHI

Trait

Position (cM)

F Statistic

F Threshold

Segregating

family

Effect / SD a

Variance b (%)

1

Standard deviation along length of fibre

121

5.15

5.13*

3

-1.3

33.6

2

Fleece weight

1

2.26

2.16*

10

0.4

8.8

3

Standard deviation of fibre diameter

62

2.68

2.67*

4

1.0

14.4

4

Fibre diameter

71

2.25

2.25*

12

1.7

8.3

5

Fleece weight

1

2.52

2.35*

5

0.5

10.3

8

Comfort factor

21

2.15

1.88*

5

1.6

9.9

Spinning fineness

4

2.22

2.18*

5

-1.3

10.5

12

Comfort factor

1

2.22

1.95*

6

0.9

10.5

13

Comfort factor

31

2.27

2.02*

6

0. 2

14.6

Spinning fineness

31

2.22

2.16*

6

0.03

14.0

16

Comfort factor

68

2.15

1.92*

6

0.8

9.9

18

Comfort factor

26

2.31

1.84*

6

0.9

11.3

Spinning fineness

26

2.07

2.07*

6

-1.1

9.3

20

Comfort factor

32

2.15

1.89*

4 & 6

1.3 & 1.1

9.9

Spinning fineness

32

2.33

2.24*

4 & 6

-1.3 & -1.4

11.5

24

Fleece weight

41

2.93

2.71**

4

0.3

13.4

Fibre diameter

31

2.02

1.99*

4

0.7

6.9

25

Coefficient of variation of fibre diameter

22

2.73

2.47*

4

1.0

14.8

a Effect on the value of the trait of each QTL allele, measured as standard deviation in the specific sire family

b % of phenotypic variance of the trait explained by the QTL

* P<0.05 experiment-wide significance

** P<0.01 experiment-wide significance

 

Two QTL influencing fibre diameter in mohair were reported in this study and one of these QTL was not linked to a QTL affecting fleece weight. This poses an opportunity to improve one trait without a negative correlated response in the other and so overcoming the unfavourable positive genetic correlation between the two traits. The unfavourable correlation between some quality traits and fleece yield can be addressed in the same way. QTL were identified for OFDA-measured quality traits, e.g. spinning fineness, comfort factor and coefficient of variation of fibre diameter. In order for the QTL identified in this study to have an impact on Angora goat genetics in South Africa, fine-mapping of the identified regions of interest should follow. Further studies should be conducted to fine-map these regions and detect favourable alleles, which can then be incorporated into selection strategies through marker assisted selection.

 

CONCLUSION

The results of these studies make a valuable contribution to the scientific knowledge of Angora goats and can be applied in the South African goat industry. Research projects underway include a quantification of the effect of incorrect parentage on EBV estimation, fine-mapping of chromosomal regions of interest and a QTL identification study for growth traits. The continued recording of phenotypic data and storage of DNA samples will assist future molecular studies.

 

It is believed that DNA marker information will assist conventional selection by increasing selection accuracy and improving the rate of genetic improvement, as well as leading to a better understanding of the physiological background of quantitative traits. The SA Angora industry is the first livestock sector in South Africa in which molecular technology was applied for genetic improvement.

 

REFERENCES

Allan, M.F. & Smith, T.P.L., 2008. Present and future applications of DNA technologies to improve beef production.  Meat Science 80, 79-85.

Dekkers, J.C.M., 2004.  Commercial application of marker- and gene-assisted selection in livestock: Strategies and lessons.  J. Anim. Sci. 82 (E Suppl), E313-328.

Dodds, K.G., McEwan, J.C. & Davis, G.H., 2007.  Integration of molecular and quantitative information in sheep and goat industry breeding programmes.  Small Rumin. Res. 70, 32-41.

Friedrich, H., 2009. Evaluation of microsatellite markers for parentage verification in South African Angora goats.  MSc thesis, University of Pretoria, Pretoria, South Africa.

Geldermann, H., Pieper, U. & Weber, W.E., 1986. Effect of misidentification of the estimation of breeding value and heritability in cattle. J. Anim. Sci. 63, 1759-1768.

Goudet, J., 1995.  FSTAT (version2.9.3): a computer programme to calculate F-statistics.  J. Hered. 8, 485-486.

Green, P., Falls, K. & Crooks, S., 1989.  Documentation for CRI-MAP, version 2.4. Washington University, St Louis.

Hartl, D.L., 1988. A primer of population genetics, 2nd Ed. Sinauer Associates, Inc. Publishers, Sunderland, Massachusetts, USA. p. 78.

Maddox, J.F. & Cockett, N.E., 2007.  An update on sheep and goat linkage maps and other genomic resources.  Small Rumin. Res. 70(1), 4-20.

Marshall, T.C., Slate, J., Kruuk, L.E.B. & Pemberton, J.M., 1998. Statistical confidence for likelihood-based paternity inference in natural populations. Molecular Ecology 7, 639-655.

Park, S.D.E., 2001.  Trypanotolerance in West African cattle and the population genetic effects of selection.  PhD thesis. University of Dublin.

Pieters, A., 2007. Genetic characterization of commercial goat populations in South Africa.  MSc thesis, University of Pretoria, South Africa.

Pollak, E.J., 2005. Application and impact of new genetic technologies on beef cattle breeding: a “real world” perspective. Aust. J. Exp. Agric. 45, 739-748.

Schibler, L., Vaiman, D. Oustry, A., Giraud-Delville, C. & Cribou, E.P., 1998.  Comparative gene mapping: A fine-scale survey of chromosome rearrangements between ruminants and humans.  Genome Res. 8, 901-915.

Snyman, M.A., 2002.  Evaluation of a genetically fine mohair producing herd.  Small Rumin. Res. 43, 105-113.

Snyman, M.A., 2011. Establishment of the South African biological reserve for small stock research and conservation at Grootfontein. Grootfontein Agric 11(1), 1-9.

Van de Goor, L.H.P., Panneman, H. & Van Haaeringen, W.A., 2009. A proposal for standardization in forensic bovine DNA typing: allele nomenclature of 16 cattle-specific short tandem repeat loci. Anim. Genet. 40, 630-636.

Van der Werf, J.H.J., 2007.  Marker assisted selection in sheep and goats. In: Marker-Assisted Selection: Current status and future perspectives in crops, livestock, forestry and fish. Eds. Guimarães, E., Ruane, J., Scherf, B., Sonnino, A. & Dargie, J. FAO, Rome, Italy. pp. 230-247.

Van Eenennaam, A.L., Weaber, R.L., Drake, D.J., Penedo, M.C.T., Quaas, R.L., Garrick, D.J. & Pollak, E.J., 2007. DNA-based paternity analysis and genetic evaluation in a large, commercial cattle ranch setting. J. Anim. Sci. 85, 3159-3169.

Visser, C., Hefer, C.A., Van Marle-Köster, E. & Kotze, A., 2004. Genetic variation of three commercial and three indigenous goat populations in South Africa. S. Afr. J. Anim. Sci. 34 (Suppl 1), 24-27.   

Wright, S., 1978.  Evolution and the genetics of populations.  Variability within and among natural populations. Vol 4.  University of Chicago Press, Chicago.

 

Published

Grootfontein Agric 11 (2): 38-46