Ovine Acyl Coa:Diacylglycerol Acyltransferase 1 – Molecular Characterization, Polymorphisms and Association with Milk Traits

doi:10.1111/j.1365-2052.2009.01909.x Ovine acyl CoA:diacylglycerol acyltransferase 1 – molecular characterization, polymorphisms and association with milk traits

M. C. Scata`*, F. Napolitano*, S. Casu†, A. Carta†, G. De Matteis*, F. Signorelli*, G. Annicchiarico‡, G. Catillo* and B. Moioli* *CRA-PCM, via Salaria 31, 00016 Monterotondo, Italy. †AGRIS – Sardegna, Loc. Bonassai, 07040 Olmedo, Italy. ‡CRA-ZOE, via Napoli, km 12-71020 Foggia, Italy

Summary The objective of this work was to characterize the complete coding region of the ovine acyl CoA:diacylglycerol acyltransferase 1 (DGAT1) gene of three Italian sheep breeds: Sarda, Altamurana and Gentile di Puglia. Characterization was accomplished by direct sequencing of 8676 bp of the relevant DNA, including introns and partial 5¢ and 3¢ untranslated regions (UTRs). We detected ﬁve novel SNPs; one SNP (g.5553C>T) is located in intron 2, has similar frequencies in the three breeds and showed a negative association with milk fat content. More interesting is an SNP in the 5¢ UTR (g.127C>A), the occurrence of which is rare in the higher milk-fat breeds (Altamurana and Gentile di Puglia); it is located in the core sequence of Sp1, a putative binding site of transcription factors. This SNP showed a signiﬁcant negative association with milk fat content in the Sarda sheep. Because DGAT1 plays a fundamental role in triacylglycerol synthesis, the novel detected SNP in the 5¢ UTR of the DGAT1 gene might explain, at least partially, the variation of fat content in the milk of Sarda sheep.

Keywords dairy sheep, DGAT1 gene, milk traits, single nucleotide polymorphisms.

genomic sequence of the bovine DGAT1 gene is available Introduction (Winter et al. 2002), as well as partial sequences of the Acyl CoA:diacylglycerol acyltransferase 1 plays a key role in caprine gene (Angiolillo et al. 2006); in contrast, the triacylglycerol synthesis; it catalyses the esterification of a sequence of the ovine gene has not been characterized. fatty acyl-CoA to the sn-3 position of a diacylglycerol. In The main objective of this work was to sequence the whole cattle, the DGAT1 gene maps to bovine chromosome 14 genomic region of DGAT1 in several sheep of different breeds (BTA 14) and contains a lysine to alanine polymorphism in with the aim of identifying polymorphisms that might be exon 8 (p.Lys232Ala) that explains 50% of the genetic used to evaluate the association of DGAT1 genotypes with variation in milk fat percentage (Grisart et al. 2002; milk traits. Schennink et al. 2007). Barillet et al. (2005), in a population of Lacaune and Manech sheep breeds, structured Materials and methods according to a granddaughter design, detected a significant QTL for dairy traits at the proximal end of ovine chromo- Animal material and recording of phenotypic traits some 9 (OAR 9), the sheep homolog of BTA14. However, they could not more precisely map the QTL, which had a SNP detection and association analysis of DGAT1 genotypes large confidence interval. with milk traits were performed on 108 sheep of three The essential role of DGAT1 in milk fat metabolism makes breeds: Altamurana (37), Gentile di Puglia (37) and Sarda the DGAT gene an interesting candidate for explaining the (34). genetic variation of milk traits in dairy sheep. The whole Altamurana is a dairy sheep belonging to the subgroup of South European milk-sheep (Pieragostini & Dario 1996). Address for correspondence The Altamurana is a local breed from Apulia (South-eastern B. Moioli, CRA-PCM, via Salaria 31, 00016 Monterotondo, Italy. Italy) and is subject to a rather harsh environment. Gentile E-mail: [email protected] di Puglia is an historical triple-purpose Merino-type breed, Accepted for publication 19 March 2009 whose origin may be traced back to the Roman times

2009 The Authors, Journal compilation 2009 Stichting International Foundation for Animal Genetics, Animal Genetics, 40, 737–742 737 738 Scata` et al.

(Altobella & Muscio 1997). The breed was widely distrib- Sequencing and genotyping of the sheep DGAT1 gene uted throughout Southern Italy, when transhumance was the common sheep-production system. Until about the mid- The whole sequence of the DGAT1 gene was amplified using 1960s, the Altamurana and Gentile di Puglia were the most 18 primer pairs. Primers to amplify the DNA region important sheep breeds in Southern Italy, numbering about encoding exons 12–17 were designed based on the goat one million head for each breed. During recent decades, sequence (accession DQ380250). To amplify the rest of the they have undergone a consistent decline in numbers, lar- gene, we designed the primers based upon the exons of the gely substituted by the Sarda breed, so that now no more bovine sequence (accession AJ318490). Because the size of than 5000 head of each breed are left. The Sarda sheep, bovine intron 1 is about 4 kb, we started by designing the a specialized dairy breed, autochthonous to Sardinia, is first primer pairs on bovine exons 1 and 2, obtaining a rapidly spreading outside of its island of origin, so that its sheep genomic amplicon of 3750 bp. Direct sequencing, in numbers have increased from about 2.5 million head in both directions, of this first amplicon, enabled us to design 1963 to over 5 million in 2000. Sound milk recording and a second set of primer pairs according to the novel sheep selection systems have allowed this breed to increase aver- sequence, obtaining a smaller second amplicon that was age lactation milk yield by 30% in 20 years (AIA 1981– again submitted to direct sequencing in both directions. 2005). The experimental farm where we performed the Design of additional primer pairs and direct sequencing present trial maintains the local breeds with the aim of the corresponding amplicons continued until the full of conservation and sustainable use of animal genetic sequence of intron 1 was obtained. Table 1 shows the resources. primers used to amplify each amplicon, covering the whole The sheep used in this trial were raised in a traditional DGAT1 gene. We performed 35 PCR cycles at the annealing management system, consisting of lambing in November, temperatures indicated in Table 1, for either 45 s (amplicon suckling for 35–60 days, then regular machine milking size <600 bp) or 1 min (larger amplicons). For four ampli- twice a day. Mean body weight of the ewes of the three cons, we performed a three-step PCR (seven cycles for 45 s at breeds was similar, ranging between 40 and 45 kg. Ewes the higher temperatures and 21 cycles for 45 s at the lowest grazed on natural pasture with supplementation of 250 g one) at the indicated annealing temperatures (Table 1). We pellet concentrate, 150 g oat grains and 1.5 kg oat and used Taq polymerase Gold enzyme (Applied Biosystems, vetch hay. Ewes were all at their second or third lambing Foster City, CA, USA) with no addition of PCR enhancing and were submitted to milk recording following the regu- additives or agents. Direct sequencing of all amplicons, in lations of the International Committee for Animal Record- both directions, for six individuals (two for each breed), was ing (ICAR), allowing us to obtain test-day records for milk performed on a Perkin Elmer ABI Prism 310 DNA sequencer yield, fat and protein content. and using the BigDye Terminator v1.1 Cycle Sequencing kit To confirm the association of DGAT1 genotypes with milk (Applied Biosystems) according to the manufacturerÕs traits, data collected from a second experimental resource instructions. The same primers used for the initial amplifi- population, consisting of 675 ewes, were used. This popu- cation of amplicons were used for this direct sequencing. lation was created in Sardinia between 2001 and 2003 by Mutation detection was performed with the Denaturing mating 475 Sarda · Lacaune backcross ewes with 15 Sarda High-Performance Liquid Chromatography (DHPLC) rams, chosen from the proven rams of the artificial insem- Transgenomic WAVE system (Transgenomic, San Jose, ination centre of the Sarda breed. The rams originated from CA, USA) for nucleic acid fragment analysis, which allows 12 flocks from different areas of the island and were rep- the resolution of DNA fragments on the basis of differential resentative of the Sarda breed. The experimental population retention of double-stranded vs. single-stranded DNA (Un- was organized in half-sib families of 44 daughters per sire derhill et al. 1997). The amplicon size recommended for this on average. The ewes were fed with the same regimen based technology is <800 bp. Therefore, we applied this approach on 4–5 h of grazing on irrigated mixed swards of ryegrass only to the 13 amplicons with <800 bp. All 108 individuals and berseem clover, with supplementation of lucerne hay, were used. For each amplicon, all heterozygous samples maize silage and concentrates, particularly in winter and detected through DHPLC were submitted to direct late spring. The first lambing was in winter at the age of sequencing. Genotyping of the homozygous samples was around 15 months, whereas successive lambings occurred performed by mixing a homozygous sample of known se- in autumn. After 4 weeks of suckling, ewes were machine quence with each of the unknown homozygous samples, as milked twice a day. In July, they were progressively dried-off proposed by Moioli et al. (2005). and milked only in the morning. In this experimental pop- A novel primer pair (forward: GTGCTGGCTTCCTA- ulation, milk yield was recorded one or two times a month. CAGTCC; reverse: AGCTGCGTGTTCTGGAAATC), yielding a At each milk recording, individual milk samples were col- 315-bp amplicon, was designed and submitted to DHPLC lected to determine protein and fat content by the infrared analysis for the genotyping of all 108 individuals at a method with Milkoscan equipment (Foss Eletric, Hillerød, mutation that had been detected in intron 1 through direct Denmark). sequencing of the first six individuals. Fourteen more

2009 The Authors, Journal compilation 2009 Stichting International Foundation for Animal Genetics, Animal Genetics, 40, 737–742 Ovine acyl CoA:diacylglycerol acyltransferase 1 739

Table 1 Primers used for PCR ampliﬁcation of the DGAT1 gene.

Primer Annealing Amplicon pair Forward primer Reverse primer temperature (C) size (bp) Covered region

1 GGAACTACGCTTCCCAGGAC ACGTCTCCGTCCTTGTCTGT 57 360 5¢UTR, partial exon 1 2 CACGGCAGTGGCGTAATAG ATCTCCTTCCTGGTAGAGGACACT 60/58/56 652 5¢ UTR, exon 1 3 ACAGACAAGGACGGAGACGTAGA ACTGTCAGAACTGAACAGGGAGTC 65 3750 Partial exon 1, partial exon 2 4 AGATTTCTCATCCCTTGCAGAC GGACTTTCAGAGCCCACCTCT 61 2837 Intron 1 5 CAGGTCAGTCAGCAAGGATTT ATCCCCTCCAGGTCTGTCCTA 63/61/58 1935 Intron 1 6 GTCACTTGCCTGCTCTACACA GTTCCCAATCCTGACAACAAG 61 1262 Intron 1 7 CTGCTCACCTAGGTCTGGACA CAGGGCCATTCTAGTTCATCA 58 467 Intron 1 8 GTCTTGCATCACCAGCTCCT CAGGCATCTACTGGGATTCAG 60 445 Exon 2 9 GCATCCTGAATTGGTGTGTG GCATGGACTATTGGGTCTGG 58 915 Intron 2 10 CACCGCTCAGTGCTGATGT AGCAGCAGCAAAGGACAGAT 60 719 Intron 2 11 TGGTAAGCTGGCTGGTTAGG TCACGCCAAGCTCTGAGAC 60/58/56 676 Intron 2 12 GGCAGGCTTGGACTTCACT GTGAGGGCACTGCTTACCAC 57 686 Exon 3, 4, 5 13 CATCCAGGTGGTGTCTCTGTT CTCTCGGCACCAGAGGTTG 58 690 Exon 4, 5, 6 14 AACCTGGCCACCATTCTCT GGGAAGTTGAGCTCGTAGCA 55 549 Exon 6, 7, 8 15 TCGTAGCTTTGGCAGGTAAGA CTTGAAGGGCTTCATGGAGTT 58 505 Exon 8, 9, 10 16 AACTCACTGTCCGCTTGCTT TGCCAGAAGTAGGTGATGGAC 58 616 Exon 10, 11, 12, 13 17 CCACCTCATCTGGCTCATCT CCTCATGGAAGAAGGCAGAG 58 419 Exon 13, 14 18 AGGAACTCGGAGTCCATCAC CTCCTCAGGGGCAGAAAAG 61/59/57 604 Exon 14, 15, 16, 17, partial 3¢ UTR individuals for each amplicon were subjected to direct Sarda · Lacaune ewes) and their grand-dams (Sarda pure- sequencing, in both directions, to confirm the genotyping bred ewes). results obtained after DHPLC analysis. In total, 16 437 records per trait, corresponding to the For the second experimental population, only daughters original lactations adjusted for the population-specific of sires carrying mutations were genotyped by direct environmental effects, were analysed. The pedigree file was sequencing in both directions at the amplicon containing created using all available relationships between animals the mutation. This was performed using the primers indi- and included 11 927 individuals born between 1960 and cated in Table1. 2004. The five-trait analysis was performed by the MTDFREML computer program (Boldman et al. 1995). Following Aulchenko et al. (2007), EBV of the informative ewes were Statistical analysis adjusted for family effects by subtracting the average of their Milk recording data of the three breeds (Sarda, Altamurana parentsÕ EBV. Finally, the association between the DGAT1 and Gentile di Puglia) of the experimental farm were used to gene and milk traits was tested by analysing the adjusted estimate, for all detected SNPs, the genotype effect on milk EBV with a simple ANOVA, including the genotype as fixed production traits, by regressing the phenotype on the effect. number of copies of one allele of each SNP (Sherman et al. 2008). The mixed procedure of SAS software (SAS Institute Results Inc. 2007) was used for this analysis. The model used for the analysis of the 108 ewes was the following: In this work, we characterized the complete genomic region of the DGAT1 gene in Ovis aries by sequencing 8676 bp of Y ¼ l þ B þ D þ G ðB ÞþA þ e ijklm i j k i l ijklm the relevant DNA in six sheep (GenBank accession number where Ym = test day milk yield, fat percentage, protein EU178818). percentage; Bi = fixed effect of the breed; Dj = covariate of The sequenced region encompassed 200 bp in the 5¢ the days in milking at the test day; Gk = allele substitution UTR, 17 exons and the corresponding introns and 128 bp effect; Al = random animal effect; eijklm = residual. in the 3¢ UTR. The coding region of 489-amino acid residues In the second experimental population, breeding values shares a high nucleotide identity (97%) with the bovine (EBV) for milk yield (kg), fat and protein yield (kg) and orthologous sequence, with only five non-synonymous content (%) were estimated by using a five-trait repeatability mutations (Ovis vs. Bos): p.Arg26Gly in exon 2; p.Leu387Ile animal model. To maximize the accuracy of breeding value in exon 14; p.Gly407Ala in exon 15; p.Arg425Gly in exon estimation, the analysed dataset included lactation records 16; and p.Thr486Ala in exon 17. The p.Lys232Ala variant of the target population ewes, of their dams (backcross of Bos taurus, located in exon 8 and directly responsible for

2009 The Authors, Journal compilation 2009 Stichting International Foundation for Animal Genetics, Animal Genetics, 40, 737–742 740 Scata` et al.

Table 2 Detected SNPs in the DGAT1 gene (GenBank accession Frequency of the SNP in the 5¢UTR (g.127C>A) was too low number EU178818) and frequency of the mutated allele in each breed. in the Altamurana (0.05) and in the Gentile di Puglia (0.02)

Frequency of the mutated allele to justify any consideration regarding gene diversity. Its Position in frequency was 0.19 in the Sarda breed and its allelic SNP the gene Altamurana Gentile di Puglia Sarda diversity was 0.32 and heterozygosity was 0.41. The overall g.127C>A 5¢ UTR 0.05 0.02 0.19 greater genetic diversity of the Sarda might reflect the g.1655C>T Intron 1 0.02 0 0 strong dichotomy in overall population size of the breeds. g.5553C>T Intron 2 0.25 0.32 0.41 Association analyses with milk traits were then per- g.7492C>T Intron 10 0 0.07 0 formed for the SNP in intron 2 (g.5553C>T) in all three g.8539C>T Exon 17 0.06 0.07 0 breeds and for the SNP in the 5¢ UTR (g.127C>A) in the Sarda breed only. Results of the association analysis of milk production milk fat variation in some cattle breeds, was not found traits with the the g.5553C>T SNP are shown in Table 3. A in sheep, where all individuals were monomorphic for the positive effect of the mutated allele on fat content is evident lysine residue. in the Altamura and Gentile di Puglia breed (+0.45%, The non-coding region of the ovine DGAT1 gene has P = 0.04 and +0.43, P = 0.07), but no effect was found on moderate global similarity (90%) with the bovine sequence, either milk yield or protein content. but the sequences of the small introns 6, 10, 14 and 16 Results of the association analysis of milk production (altogether 310 bp) differ from the corresponding bovine traits with the g.127C>A SNP, performed only in the Sarda sequence by over 20%, and beyond the nucleotide differ- breed, indicated a negative ()0.29%), although non-signif- ences, they are the location of 20% of the insertions and icant, effect of the mutated allele on fat content. This SNP is deletions observed in the characterized sequence. The two located in the 5¢ UTR of the DGAT1 gene, 73 bp before the UTR regions (328 bp) share about 85% nucleotide identity ATG start codon. Results of in silico analysis (http://www- with the bovine sequence. bitmas.cit.nih.gov) performed on this fragment indicate that The DHPLC analysis revealed five SNPs, which were the g.C127A SNP falls in the core sequence of a putative located in the 5¢ UTR, intron 1, intron 2, intron 10 and binding site of transcription factors: Sp1 (GGCGGG). a non-synonymous mutation in exon 17. In Table 2, we Because the in silico analysis showed that the g.127C>A report the position of the detected SNPs on the GenBank SNP might affect the expression of the DGAT1 gene, the accession EU178818 and the frequencies of the mutated association analysis was performed on the second experi- alleles in the three studied breeds. Three of the detected mental population, in which two out of the fifteen rams SNPs, g.1655C>T, g.7492C>T and g.8539C>T, were were heterozygous for this mutation. Genotyping of their extremely rare (frequency lower than 0.07) and absent in daughters (n = 70) at the g.127C>A mutation revealed some breeds and were thus not studied further. Average that about half were heterozygous, and only two ewes were allelic diversity and heterozygosity were then calculated for homozygous for the mutant allele (one for each sire), SNPs in intron 2 (g.5553C>T) and in the 5¢ UTR in accordance with Hardy–Weinberg equilibrium. We (g.127C>A) with Proc Allele in SAS (SAS Institute Inc. assumed, therefore, that the heterozygous ewes likely 2007), which uses the methodology described by Weir received the mutated allele from their sire. (1996). Frequency of the SNP in intron 2 (g.5553C>T) was Of the 70 daughters, only 57 (32 and 25 respectively for the greatest in the Sarda breed (0.41) and the smallest in each of the two rams) had a regular productive career; one the Altamurana (0.25). Allelic diversity was 0.39, 0.43 and of them was g.127AA homozygous and was excluded from 0.48 and heterozygosity was 0.42, 0.55 and 0.53 respec- the analysis. Thus, 27 g.127CA ewes were tested against 29 tively in the Altamurana, Gentile di Puglia and Sarda. g.127CC ewes.

Table 3 Breed effect and substitution effect of allele T at g.5553C>T (intron 2) on daily milk yield, fat percentage and protein percentage.

Milk yield (g) Fat % Protein %

Effect/trait Estimate SE P-value Estimate SE P-value Estimate SE P-value

Altamurana 382.81 25.8 8.301 0.24 6.641 0.10 Gentile di Puglia 198.42 32.3 8.872 0.16 6.952 0.13 Sarda 531.93 23.1 7.083 0.13 5.533 0.10 g.5553C>T · Altamurana )14.69 42.4 NS +0.45 0.23 0.04 +0.16 0.19 NS g.5553C>T · Gentile )15.29 56.4 NS +0.43 0.24 0.07 )0.29 0.23 NS g.5553C>T · Sarda )49.48 34.4 NS +0.07 0.19 NS 0.14 0.14 NS

Different superscript numbers indicate signiﬁcant differences between mean at P < 0.0001.

2009 The Authors, Journal compilation 2009 Stichting International Foundation for Animal Genetics, Animal Genetics, 40, 737–742 Ovine acyl CoA:diacylglycerol acyltransferase 1 741

Table 4 Allele substitution effect of DGAT1 g.127C>A, from ANOVA More interesting is a SNP in the 5¢ UTR (g.127C>A) that 1 simple test, on the EBV of 56 daughters (n = 27 g.127CA; is located in the core sequence of Sp1, a putative binding site n = 29 g.127CC) of two heterozygous rams. of transcription factors. This SNP, which showed very low Allele substitution effect (g.127CA vs. g.127CC) frequency in the Altamurana and Gentile di Puglia breeds, had a significant negative association with milk fat and Raw SD Raw mean Raw SD protein content and a positive association with milk yield EBV value units units units P-value in the Sarda sheep. This result is in accordance with the Milk yield (litres) 8.2 0.65 0.03 0.10 0.014 previous studies in cattle that estimated the effect of the Fat % )0.11 0.67 0.02 0.22 0.005 p.Lys232Ala variant of of DGAT1 (Grisart et al. 2002; Protein % )0.09 0.71 0.02 0.19 0.006 Schennink et al. 2007). The Sp1 protein contains a zinc Fat yield (kg) 0.28 0.45 0.02 0.05 NS finger protein motif, by which it binds directly to DNA and Protein yield (kg) 0.20 0.38 0.02 0.05 NS enhances gene transcription. It has been shown in human 1Sarda · Lacaune backcross ewes · Sarda rams. and rodent cell lines to bind to a number of promoters, thus modifying their basal expression (Greene et al. 1987). In humans, Muredda et al. (2003) showed that, at the post- The allele substitution effect of SNP g.127C>A (Table 4), transcriptional level, the presence of GC-rich sequences calculated by contrasting LS means values of CA and CC capable of binding Sp1 within the 5¢UTR may contribute to genotypes, was significant for fat (P = 0.005), protein basal expression of the gene. In the Fatty Acid Synthase percentage (P = 0.006) and milk yield (P = 0.014). No (FASN) gene, a polymorphism altering the putative Sp1 significant effects on fat and protein yields were detected. transcription factor-binding site in the untranslated exon 1 Significant substitution effects were large, ranging from was significantly associated with milk fat content in dairy 0.65 to 0.71 units of phenotype SD. Selection potential for cattle (Roy et al. 2006). In addition, the low frequency of this allele can be appreciated when expressing the effect in SNP g.127C>A in the higher milk fat breeds (Altamurana raw SD units, about 20% for fat and protein content and and Gentile di Puglia) is consistent with the hypothesis that 10% for milk yield. the detected SNP has a direct negative effect on fat content. This hypothesis was corroborated in the second population analysed, although this analysis included only 56 ewes, Discussion because of the low frequency of the mutated allele. There- DGAT1 plays a fundamental role in triacylglycerol synthe- fore, our findings can only be regarded as an indication of sis; in cattle, a non-synonymous substitution in exon 8 of the possible association of this SNP with milk fat content. the DGAT1 gene influences enzyme activity, having a major This needs to be further confirmed in independent dairy effect on milk fat content (Grisart et al. 2002; Schennink sheep populations. et al. 2007). Different cattle breeds show highly variable Fat content of sheep milk may vary from 6 to 9% across frequencies of the mutation (Kaupe et al. 2004) and some and within breeds (Haenlein 2001). Therefore, the results dairy breeds are monomorphic (Fontanesi et al. 2007). presented here indicate that the SNP g.127C>A of the According to Weller (2007), there is a significant scope for DGAT1 gene explains only a portion of that variation. Cer- selecting for this allele in cattle. These features make DGAT1 tainly, other genes are expected to influence milk fat con- an interesting candidate gene for fat percentage in other tent. In particular, the genes encoding enzymes directly ruminants. In the current work, we sequenced the complete involved in fatty acid metabolism – Acetyl-CoA carboxylase; coding region (8676 bp), including introns and partial 5¢ Stearoyl-CoA desaturase; lipoprotein lipase – are being now and 3¢ UTRs, of the ovine DGAT1 gene in 20 sheep of three extensively studied in sheep (Moioli et al. 2007). breeds: Sarda, Altamurana and Gentile di Puglia. This is the first study in sheep reporting SNPs in the No SNPs were detected in exon 8, where all sheep were DGAT1 gene, and it was performed after having fully homozygous for the alleles encoding Lysine. We detected characterized the complete coding region of the gene. To five novel SNPs and one SNP (g.5553C>T) was located in our knowledge, there is no previous report showing asso- the middle of the long intron 2 (1.9 kb). To our knowledge, ciations of this gene with milk or meat production traits in potential splicing sites might extend no more than 20–30 sheep. We have shown in this work that the novel detected bases into the intron (Garrigan & Edwards 1999). Moreover, SNP in the 5¢ UTR of the DGAT1 gene might explain a this SNP had a significant association with milk fat content portion of the variation of milk fat content in the Sarda only in the higher milk fat breeds (Altamurana and Gentile sheep. di Puglia), although it had similar frequencies in the three breeds. Therefore, although it is not possible to ascribe to Acknowledgement this variant a direct contribution to the observed increase of fat, it could serve as a surrogate for causative SNPs that are This study is part of the ‘‘GENZOOT’’ research programme, yet to be detected. funded by the Italian Ministry of Agriculture.

2009 The Authors, Journal compilation 2009 Stichting International Foundation for Animal Genetics, Animal Genetics, 40, 737–742 742 Scata` et al.

References Moioli B., Napolitano F., Orru` L. & Catillo G. (2005) Single nucleotide polymorphism detection in Promoter III of the Acetyl-CoA AIA. (1981–2005) Milk Recording Activity: Official Statistics. Italian carboxylase-a gene in sheep. Journal of Animal Breeding and Breeders Association, Rome, Italy. Genetics 122, 418–20. Altobella C. & Muscio A. (1997) Origine della razza ovina Gentile Moioli B., DÕAndrea M. & Pilla F. (2007) Candidate genes affecting del Tavoliere di Puglia. In: Proceedings of the National Conference sheep and goat milk quality. Small Ruminant Research 68(1–2), ÔLÕallevamento Ovino in Capitanata tra Memoria e FuturoÕ, Fog- 179–92. gia, Italy, 2nd December 1995, pp. 29–58. Centrografica Fran- Muredda M., Nunoya K., Burtch-Wright R.A., Kurz E.U., Cole S.P. & cescano, Foggia, Italy. Deeley R.G. (2003) Cloning and characterization of the murine and Angiolillo A., Amills M., Urrutia B., Domenech A., Sastre Y., rat Mrp1 promoter regions. Molecular Pharmacology 64, 1259–69. Badaoui B. & Jordana J. (2006) Identification of a single nucle- Pieragostini E. & Dario C. (1996) Altamurana sheep in Apulian otide polymorphism at intro 16 of the caprine Acytl-Coenzyme hills: the best choice for a sustainable animal production. Large A:diacylglycerol acyltransferase 1 (DGAT1) gene. Journal of Dairy Animal Review 1, 65–71. Research 74, 47–51. Roy R., Ordovas L., Zaragoza P., Romero A., Altarriba J. & Rodellar Aulchenko Y.S., de Koning D.J. & Haley C. (2007) Genome-wide C. (2006) Association of polymorphisms in the bovine FASN gene rapid association using mixed model and megression: a fast and with milk-fat content. Animal Genetics 37, 215–8. simple method for genome-wide pedigree-based Quantitative SAS Institute Inc. (2007) SAS/STAT UserÕs Guide, Version 9.1. SAS Trait Loci association analysis. Genetics 177, 577–85. Institute Inc., Cary, NC. Barillet F., Arranz J.J. & Carta A. (2005) Mapping quantitative trat Schennink A., Stoop W.M., Visker M.H., Heck J.M., Bovenhuis H., loci for milk production and genetic polymorphisms of milk van der Poel J.J., van Valenberg H.J. & van Arendonk J.A. (2007) proteins in dairy sheep. Genetics Selection Evolution 37(Suppl. 1), DGAT1 underlies large genetic variation in milk-fat composition S109–23. of dairy cows. Animal Genetics 38, 467–73. Boldman K.G., Kriese L.A., Van Vleck L.D., Van Tassell C.P. & Sherman E.L., Nkrumah J.D., Murdoch B.M., Li C., Wang Z., Fu A. & Kachman S.D. (1995) Manual for Use of MTDFREML. A Set of Moore S.S. (2008) Polymorphisms and haplotypes in the bovine Programs to Obtain Estimates of Variances and Covariances. U.S. NPY, GHR, GHRL, IGF2, UCP2, and UCP3 genes and their asso- Department of Agriculture, Agricultural Research Service, Clay ciations with measures of growth, performance, feed efficiency and Center, NE. carcass merit in beef cattle. Journal of Animal Science 86, 1–16. Fontanesi L., Scotti E., Schiavini F., La Mattina V., Davoli R., Russo Underhill P.A., Jin L., Lin A.L., Mehdi S.Q., Jenkins T., Vollrath D., V. & Bagnato A. (2007) Investigation of the DGAT1 K232A and Davis R., Cavalli-Sforza L.L. & Oefner P.J. (1997) Detection of VNTR mutations in dairy and dual purpose cattle breeds. Italian numerous Y chromosome biallelic polymorphisms by denaturing Journal of Animal Science 6 (Suppl. 1), 136 (abstract). high performance liquid chromatography. Genome Research 7, Garrigan D. & Edwards S.V. (1999) Polymorphism across an exon- 996–1005. intron boundary in an avian MHC class II B gene. Molecular Weir B.S. (1996) Genetic Data Analysis II. Sinauer Associates, Inc., Biology and Evolution 16, 1599–606. Sunderland, MA. Greene J.M., Larin Z., Taylor I.C., Prentice H., Gwinn K.A. & Weller J. (2007) Current and future development in patents for Kingston R.E. (1987) Multiple basal elements of a human hsp70 quantitative trait loci in dairy cattle. Recent Patent on DNA and promoter function differently in human and rodent cell lines. Gene Sequences 1, 69–76. Molecular and Cellular Biology 7, 3646–55. Winter A., Kramer W., Werner F., Kollers S., Kata S., Durstewits G., Grisart B., Coppieters W., Farnir F. et al. (2002) Positional candi- Buitkamp J., Womack J.E., Thaller G. & Fries R. (2002) Associ- date cloning of a QTL in dairy cattle: identification of a missense ation of a lysine-232/alanine polymorphism in a bovine gene mutation in the bovine DGAT1 gene with major effect on milk encoding diacylglycerol acyltransferase (DGAT) with variation at yield and composition. Genome Research 12, 222–31. a quantitative trait locus for on milk fat content in cattle. Pro- Haenlein G.F.W. (2001) The nutritional value of sheep milk. ceedings of the National Academy of Sciences of the United States of International Journal of Animal Science (India) 16, 253–68. America 99, 9300–5. Kaupe B., Winter A., Fries R. & Erhardt G. (2004) DGAT1 polymorphism in Bos indicus and Bos Taurus cattle breeds. Journal of Dairy Research 71, 182–7.

2009 The Authors, Journal compilation 2009 Stichting International Foundation for Animal Genetics, Animal Genetics, 40, 737–742