Animal Genetics 1992,23, 175-178
SHORT COMMUNICATION
Rapid p-lactoglobulin genotyping of cattle using the polymerase chain reaction R. J. WILKINS & Y. M. KUYS Ruakura Agricultural Centre, Hamilton, New Zealand
Summary. A polymerase chain reaction (PCR) assay has been developed for genotyping P-lactoglobulin A and B variants in dairy cattle. Either blood or semen samples can be used as a source of DNA. The method is accurate, faster than Southern blot analyses and should prove a useful tool in breeding programmes. Keywords: p-lactoglobulin, genotyping, PCR
In recent years, there has been considerable interest in genotyping P-lactoglobulin, the major serum protein in bovine milk. At least eight genetic variants have been identified but only one in particular, the B variant, is reported to be associated with superior milk production and cheesemaking characteristics, e.g. Marziali & Ng-Kwai-Hang (1986), McLean et al. (1987) and Tee et al. (1990). The A and B variants are by far the most common in dairy cattle (Lien et al. 1990a; Tee et al. 1990) and for all practical purposes a genotyping test need only distinguish between these two variants. Two amino acid substitutions, a Gly for an Asp at position 64 and an Ala for a Val at position 118, distinguish the B from the A variant. At the DNA level, these changes arise from two single base substitutions, A+G, and a T+C, respectively (Medrano & Aguilar-Cordova 1990). The two variants are readily distinguishable by polyacrylamide gel electrophoresis under non-denaturing conditions because of the greater negative charge in the A variant (due to the Asp at position 64) which gives about a 10% increase in mobility. Fortuitously, at the DNA level both the A+G and the T+C substitutions give rise to restriction fragment length polymorphisms (RFLPs) with, respectively, Hph I and HaeIII restriction sites occuring in DNA coding for the B but not the A variant. Lien etal. (1990a) make use of the HaeIII site to distinguish between the A and B alleles in a Southern blot test that they have developed for P-lactoglobulin genotyping. Tee et al. (1990) have developed a similar test, but have used both the HphI and the HaeIII sites. This last group report that the two sites are tightly linked and segregate together in all the 25 B variant genomic DNAs tested. They concluded that only one of the polymorphic sites needed to be assayed; their preference was the HaeIII site because of the 10-fold lower cost of the restriction enzyme. Correspondence: Dr R.J. Wilkins, Ruakura Agricultural Centre, Private Bag, Hamilton, New Zealand. Accepted 24 June 1991
175
176 R. J . Wilkins & Y . M . Kuys
Polymerase chain reaction (PCR) assays obviously have several advantages over Southern blot methods and one such assay was recently reported for p-lactoglobulin (Medrano & Aguilar-Cordova 1990), based on the polymorphism at the Hue111 locus. The assay we report here looks directly at the HphI polymorphisms and, as such, may offer some advantages over the aforementioned assay. DNA samples were prepared from either peripheral blood (Ciulla et al. 1988) or from frozen semen using a proteinase K variation of the extraction method reported by Lien et al. (1990b). For PCR of genomic DNA, primers were chosen which amplified a 961-bp region of the p-lactoglobulin gene from the junction of exon 2-intron 2 through to the end of exon 3 (Fig. 1). We used the cattle P-lactoglobulin cDNA sequence information of Alexander et uf. (1989) to design the primers and the ovine genomic sequence (Harris et al. 1988) to determine the intron-exon boundaries, the approximate size of intron 2 and the position of a non-polymorphic HphI site within this intron. The polymorphic site within exon 3 is cut by Hph I in the B but not the A variant allele. In order to amplify this 961-bp fragment, a total reaction mix of 20p1 containing 16.7 r n (NH4)*S04, ~ 66.7rnM Tris-HCI (pH 8-8), 6.7mM MgC12, 1OmM f3-mercaptoethanol, 170 &ml of bovine serum albumin, 6-7p,M EDTA, 6 0 0 of~ each ~ dNTP, 170nM of each primer, 1 U of Taq polymerase (Cetus) and approximately lOOng of DNA was subjected to 30 cycles of 94°C for 1min, 61°C for 30s and 72°C for 2.5 min in a Perkin Elmer Thermal Cycler. The DNA product was restricted by adding 2 U of HphI (New England Biolabs, Beverly, MA, USA) to lop1 of the mix and continuing incubation at 37°C for 1h. This sample was then electrophoresed through 3% agarose gels at 10V/cm for 15 min. DNA bands were detected by ethidium bromide fluorescence. P-lactoglobulin protein variants were detected by standard techniques, namely electrophoresing the supernatant from defatted, acid-precipitated milk samples through 15% non-denaturing polyacrylamide gels and then staining with Coomassie Blue. p-lactoglobulin A and B variant standards from Sigma (St Louis, MO, USA) were used as controls. All blood DNA samples gave the expected 961-bp fragment on amplification with no other spurious DNA being visible under these PCR conditions. On digestion with HphI, all samples gave the 741-bp constant fragment plus either a 220-bp fragment ( A allele) or a 166-bp ( B allele) or a combination of the two (AIB heterozygote) (Fig. 2). The additional 54-bp fragment produced by cleavage of the B allele could not be seen on standard gels because of its small size, weak fluorescence and proximity to the strongly fluorescing primers which also run near the dye front. Some HphI digests were not complete, as evidenced by a remnant of the original 961-bp PCR product in two of the lanes in Fig. 2, but this band never constituted more than 5-10% of the total DNA and did not affect the assay. Gels were purposely run fast to maximize the sharpness of the 166-bp band. We found that slower runs (of an hour or so) led to significant diffusion of this band and a much weaker signal. Identical results were obtained with sperm DNA (results not shown). Blood from a total of 77 cows was analysed. Readable results were obtained in all except four cases, which proved difficult to PCR and for which the DNA had to be repurified. Analysis of milk proteins gave genotypes identical to those deduced from
PCR analysis of P-lactoglobulin genotypes 177
A-Allele
741bp
B - Allele
741bp
220bp
I
Figure 1. The amplified 961-bp fragment from bovine DNA. Primer 1 has the sequence 5'-ACCTGGAGATCCTGCTGCAGAAATG-3' and primer 2 5'-CATCGATCITGAACACCGCAGGGAT-3'. The polymorphicHph I site is shown in parentheses. The estimated size of fragments may be in error by a few base pairs because the ovine sequence of Harris et at. (1988) was used to calculate the size of intron 2.
DNA analysis except in one case for which the discrepancy was found to arise from an error in labelling the milk sample. The frequencies of the A and B alleles were 0.45 and 0.55 respectively, which is similar to that found by others (e.g. Tee et al. 1990). Our cows were of Friesian-Jersey origin. We were aware of the possibility that non-identical twins could give erroneous results (Pinder et al. 1990) due to blood stem cells being shared, but the only such twins in our sample had A/B genotypes at the protein level so were were unable to test this possibility. The method described here is faster, simpler and more amenable to large-scale use than those using Southern blotting. Forty samples can easily be handled per day and the only real bottleneck is likely to be the preparation of the DNA itself. As yet we have not investigated rapid methods (Higuchi 1989) of isolating small amounts of crude DNA from blood, semen or, possibly, mouth scrapings. The last mentioned source, as well as being well suited to PCR, would circumvent possible errors arising from non-identical twins. The particular genomic fragment we have chosen for analysis has two attractions. First, by analysing the HphI rather than the HaeIII site, we are looking directly at the substitution which gives rise to the electrophoretic mobility shift of the B variant P-lactoglobulin protein (though there is no evidence that this substitution per se has any phenotypic effects on milk production or composition). Second, the presence of the constant 741-bp fragment provides a convenient internal control of the efficacy of the HphI digestion. One objection to using HphI enzyme is its relative expense but in practice the cost is minimal at less than US$1 per assay. Whether or not it is better to detect B alleles on the basis of the Hph I or the Hue111 polymorphism is probably a somewhat academic point, given the reportedly (Tee et al. 1990) tight linkage between the two sites. A more practical consideration, however, is the potential difficulty of reliably detecting the small 74-bp DNA fragment released from B allele DNA by Hue111digestion, although Medrano & Aguilar-Cordova (1990) double the
178 R. J . Wilkitis & Y . M . Kuys
Figure 2. Examples of genotypes analysed by electrophoresis of a restriction digest of PCR products (left) and of milk proteins (right). Bands were identified by comparison with A HindIIIIEcoRI-pBR322 Hinfl DNA markers and 9-lactoglobulin standards (Sigma) respectively.
sensitivity of their assay by designing their primers so that two 74-bp fragments are released when the B allele is cleaved. In conclusion, we have presented a simple method which measures an important genotype in cattle and should, along with other PCR tests for milk protein variants, allow the early testing of sires and dams, thusavoiding the time-consuming analyses based on measuring milk proteins in lactating cows.
Acknowledgement We thank Gillian Rajan for preparation of DNA samples. References Alexander L.J.. Hayes G . , Pearse M.J.. Beattie C . W . , Stewart A.F.. Willis I.M. & MacKinlay A.G. (1989) Complete sequence of the bovine P-lactoglobulin cDNA. Nucleic Acids Research 17, 6739. Ciulla T . A . , Sklar R . M . & Hauser S.L. (1988) A simple method for D N A purification from peripheral blood. Anulyiical Biochemistry 174, 385488. Harris S . . Ali S . , Anderson S . , Archibald A.L. &Clark A.J. (1988)Complete nucleotide sequence of the genomic ovine P-lactoglobulin gene. Nucleic Acids Research 16, 10379-80. Higuchi R. (1989) Simple and rapid preparation of samples for PCR. In: PCR Technology (ed. by H . A . Erlich). pp. 31-8. Stockton Press. New York. Lien S.. Alestrom P.. Steine T.. Langsrud T.. Vegarud G. 6t Rogne S. (1990a) A method f o r P-lactoglobuin genotyping of cattle. Livestock Production Science 25, 173-6. Lien S.. Rogne S . , Brovold M.J. & Alestrom P. (1990b) A method for isolation of DNA from frozen (AI) hulls semen. Journal of Animul Breeding mid Generics 107, 74. McLean D.M., Graham E . R . B . , Ponzoni R . W . & McKenzie H . A . (1987) Effectsof milk protein genetic variants and composition on heat stability of milk. Journal of Dairy Research 54,219-35. Marziali A.S. & Ng-Kwai-Hang K.F. (1986) Relationships between milk protein polymorphisrns and cheese yielding capacity. Journal of Dairy Science 69, 1193201. Mcdrano J .F. & Aguilar-Cordova E. (1990) Polymerase chain reaction amplification of bovine p-lactoglobulin genomic sequences and identification of genetic variants by RFLP analysis. Animul Biotechnology 1, 73-7. Pinder S . J . , Perry B.N.. Savva D . & Skidmore C.J. (1990) The polymerase chain reaction applied to identification of specific allcles of the bovine milk protein genes. Biochemical Society Trunsacrions 18, 675-6. Tee M.K.. Moran C. & Nicholas F.W. (1990) Genotyping of the A and B variants of cattle beta-lactoglobulin using restriction length polymorphisms. In: Proceedings of the Australasian Gene Mapping Workshop. p. 62. Macquarie University. Sydney.
SHORT COMMUNICATION
Rapid p-lactoglobulin genotyping of cattle using the polymerase chain reaction R. J. WILKINS & Y. M. KUYS Ruakura Agricultural Centre, Hamilton, New Zealand
Summary. A polymerase chain reaction (PCR) assay has been developed for genotyping P-lactoglobulin A and B variants in dairy cattle. Either blood or semen samples can be used as a source of DNA. The method is accurate, faster than Southern blot analyses and should prove a useful tool in breeding programmes. Keywords: p-lactoglobulin, genotyping, PCR
In recent years, there has been considerable interest in genotyping P-lactoglobulin, the major serum protein in bovine milk. At least eight genetic variants have been identified but only one in particular, the B variant, is reported to be associated with superior milk production and cheesemaking characteristics, e.g. Marziali & Ng-Kwai-Hang (1986), McLean et al. (1987) and Tee et al. (1990). The A and B variants are by far the most common in dairy cattle (Lien et al. 1990a; Tee et al. 1990) and for all practical purposes a genotyping test need only distinguish between these two variants. Two amino acid substitutions, a Gly for an Asp at position 64 and an Ala for a Val at position 118, distinguish the B from the A variant. At the DNA level, these changes arise from two single base substitutions, A+G, and a T+C, respectively (Medrano & Aguilar-Cordova 1990). The two variants are readily distinguishable by polyacrylamide gel electrophoresis under non-denaturing conditions because of the greater negative charge in the A variant (due to the Asp at position 64) which gives about a 10% increase in mobility. Fortuitously, at the DNA level both the A+G and the T+C substitutions give rise to restriction fragment length polymorphisms (RFLPs) with, respectively, Hph I and HaeIII restriction sites occuring in DNA coding for the B but not the A variant. Lien etal. (1990a) make use of the HaeIII site to distinguish between the A and B alleles in a Southern blot test that they have developed for P-lactoglobulin genotyping. Tee et al. (1990) have developed a similar test, but have used both the HphI and the HaeIII sites. This last group report that the two sites are tightly linked and segregate together in all the 25 B variant genomic DNAs tested. They concluded that only one of the polymorphic sites needed to be assayed; their preference was the HaeIII site because of the 10-fold lower cost of the restriction enzyme. Correspondence: Dr R.J. Wilkins, Ruakura Agricultural Centre, Private Bag, Hamilton, New Zealand. Accepted 24 June 1991
175
176 R. J . Wilkins & Y . M . Kuys
Polymerase chain reaction (PCR) assays obviously have several advantages over Southern blot methods and one such assay was recently reported for p-lactoglobulin (Medrano & Aguilar-Cordova 1990), based on the polymorphism at the Hue111 locus. The assay we report here looks directly at the HphI polymorphisms and, as such, may offer some advantages over the aforementioned assay. DNA samples were prepared from either peripheral blood (Ciulla et al. 1988) or from frozen semen using a proteinase K variation of the extraction method reported by Lien et al. (1990b). For PCR of genomic DNA, primers were chosen which amplified a 961-bp region of the p-lactoglobulin gene from the junction of exon 2-intron 2 through to the end of exon 3 (Fig. 1). We used the cattle P-lactoglobulin cDNA sequence information of Alexander et uf. (1989) to design the primers and the ovine genomic sequence (Harris et al. 1988) to determine the intron-exon boundaries, the approximate size of intron 2 and the position of a non-polymorphic HphI site within this intron. The polymorphic site within exon 3 is cut by Hph I in the B but not the A variant allele. In order to amplify this 961-bp fragment, a total reaction mix of 20p1 containing 16.7 r n (NH4)*S04, ~ 66.7rnM Tris-HCI (pH 8-8), 6.7mM MgC12, 1OmM f3-mercaptoethanol, 170 &ml of bovine serum albumin, 6-7p,M EDTA, 6 0 0 of~ each ~ dNTP, 170nM of each primer, 1 U of Taq polymerase (Cetus) and approximately lOOng of DNA was subjected to 30 cycles of 94°C for 1min, 61°C for 30s and 72°C for 2.5 min in a Perkin Elmer Thermal Cycler. The DNA product was restricted by adding 2 U of HphI (New England Biolabs, Beverly, MA, USA) to lop1 of the mix and continuing incubation at 37°C for 1h. This sample was then electrophoresed through 3% agarose gels at 10V/cm for 15 min. DNA bands were detected by ethidium bromide fluorescence. P-lactoglobulin protein variants were detected by standard techniques, namely electrophoresing the supernatant from defatted, acid-precipitated milk samples through 15% non-denaturing polyacrylamide gels and then staining with Coomassie Blue. p-lactoglobulin A and B variant standards from Sigma (St Louis, MO, USA) were used as controls. All blood DNA samples gave the expected 961-bp fragment on amplification with no other spurious DNA being visible under these PCR conditions. On digestion with HphI, all samples gave the 741-bp constant fragment plus either a 220-bp fragment ( A allele) or a 166-bp ( B allele) or a combination of the two (AIB heterozygote) (Fig. 2). The additional 54-bp fragment produced by cleavage of the B allele could not be seen on standard gels because of its small size, weak fluorescence and proximity to the strongly fluorescing primers which also run near the dye front. Some HphI digests were not complete, as evidenced by a remnant of the original 961-bp PCR product in two of the lanes in Fig. 2, but this band never constituted more than 5-10% of the total DNA and did not affect the assay. Gels were purposely run fast to maximize the sharpness of the 166-bp band. We found that slower runs (of an hour or so) led to significant diffusion of this band and a much weaker signal. Identical results were obtained with sperm DNA (results not shown). Blood from a total of 77 cows was analysed. Readable results were obtained in all except four cases, which proved difficult to PCR and for which the DNA had to be repurified. Analysis of milk proteins gave genotypes identical to those deduced from
PCR analysis of P-lactoglobulin genotypes 177
A-Allele
741bp
B - Allele
741bp
220bp
I
Figure 1. The amplified 961-bp fragment from bovine DNA. Primer 1 has the sequence 5'-ACCTGGAGATCCTGCTGCAGAAATG-3' and primer 2 5'-CATCGATCITGAACACCGCAGGGAT-3'. The polymorphicHph I site is shown in parentheses. The estimated size of fragments may be in error by a few base pairs because the ovine sequence of Harris et at. (1988) was used to calculate the size of intron 2.
DNA analysis except in one case for which the discrepancy was found to arise from an error in labelling the milk sample. The frequencies of the A and B alleles were 0.45 and 0.55 respectively, which is similar to that found by others (e.g. Tee et al. 1990). Our cows were of Friesian-Jersey origin. We were aware of the possibility that non-identical twins could give erroneous results (Pinder et al. 1990) due to blood stem cells being shared, but the only such twins in our sample had A/B genotypes at the protein level so were were unable to test this possibility. The method described here is faster, simpler and more amenable to large-scale use than those using Southern blotting. Forty samples can easily be handled per day and the only real bottleneck is likely to be the preparation of the DNA itself. As yet we have not investigated rapid methods (Higuchi 1989) of isolating small amounts of crude DNA from blood, semen or, possibly, mouth scrapings. The last mentioned source, as well as being well suited to PCR, would circumvent possible errors arising from non-identical twins. The particular genomic fragment we have chosen for analysis has two attractions. First, by analysing the HphI rather than the HaeIII site, we are looking directly at the substitution which gives rise to the electrophoretic mobility shift of the B variant P-lactoglobulin protein (though there is no evidence that this substitution per se has any phenotypic effects on milk production or composition). Second, the presence of the constant 741-bp fragment provides a convenient internal control of the efficacy of the HphI digestion. One objection to using HphI enzyme is its relative expense but in practice the cost is minimal at less than US$1 per assay. Whether or not it is better to detect B alleles on the basis of the Hph I or the Hue111 polymorphism is probably a somewhat academic point, given the reportedly (Tee et al. 1990) tight linkage between the two sites. A more practical consideration, however, is the potential difficulty of reliably detecting the small 74-bp DNA fragment released from B allele DNA by Hue111digestion, although Medrano & Aguilar-Cordova (1990) double the
178 R. J . Wilkitis & Y . M . Kuys
Figure 2. Examples of genotypes analysed by electrophoresis of a restriction digest of PCR products (left) and of milk proteins (right). Bands were identified by comparison with A HindIIIIEcoRI-pBR322 Hinfl DNA markers and 9-lactoglobulin standards (Sigma) respectively.
sensitivity of their assay by designing their primers so that two 74-bp fragments are released when the B allele is cleaved. In conclusion, we have presented a simple method which measures an important genotype in cattle and should, along with other PCR tests for milk protein variants, allow the early testing of sires and dams, thusavoiding the time-consuming analyses based on measuring milk proteins in lactating cows.
Acknowledgement We thank Gillian Rajan for preparation of DNA samples. References Alexander L.J.. Hayes G . , Pearse M.J.. Beattie C . W . , Stewart A.F.. Willis I.M. & MacKinlay A.G. (1989) Complete sequence of the bovine P-lactoglobulin cDNA. Nucleic Acids Research 17, 6739. Ciulla T . A . , Sklar R . M . & Hauser S.L. (1988) A simple method for D N A purification from peripheral blood. Anulyiical Biochemistry 174, 385488. Harris S . . Ali S . , Anderson S . , Archibald A.L. &Clark A.J. (1988)Complete nucleotide sequence of the genomic ovine P-lactoglobulin gene. Nucleic Acids Research 16, 10379-80. Higuchi R. (1989) Simple and rapid preparation of samples for PCR. In: PCR Technology (ed. by H . A . Erlich). pp. 31-8. Stockton Press. New York. Lien S.. Alestrom P.. Steine T.. Langsrud T.. Vegarud G. 6t Rogne S. (1990a) A method f o r P-lactoglobuin genotyping of cattle. Livestock Production Science 25, 173-6. Lien S.. Rogne S . , Brovold M.J. & Alestrom P. (1990b) A method for isolation of DNA from frozen (AI) hulls semen. Journal of Animul Breeding mid Generics 107, 74. McLean D.M., Graham E . R . B . , Ponzoni R . W . & McKenzie H . A . (1987) Effectsof milk protein genetic variants and composition on heat stability of milk. Journal of Dairy Research 54,219-35. Marziali A.S. & Ng-Kwai-Hang K.F. (1986) Relationships between milk protein polymorphisrns and cheese yielding capacity. Journal of Dairy Science 69, 1193201. Mcdrano J .F. & Aguilar-Cordova E. (1990) Polymerase chain reaction amplification of bovine p-lactoglobulin genomic sequences and identification of genetic variants by RFLP analysis. Animul Biotechnology 1, 73-7. Pinder S . J . , Perry B.N.. Savva D . & Skidmore C.J. (1990) The polymerase chain reaction applied to identification of specific allcles of the bovine milk protein genes. Biochemical Society Trunsacrions 18, 675-6. Tee M.K.. Moran C. & Nicholas F.W. (1990) Genotyping of the A and B variants of cattle beta-lactoglobulin using restriction length polymorphisms. In: Proceedings of the Australasian Gene Mapping Workshop. p. 62. Macquarie University. Sydney.
