| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1
* Animal Breeding and Genomics Centre, Wageningen University, PO Box 338, 6700 AH Wageningen, the Netherlands
Dairy Science and Technology Group, Wageningen University, PO Box 8129, 6700 EV Wageningen, the Netherlands
2 Corresponding author: anke.schennink{at}wur.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: milk fatty acid composition • SCD1 • DGAT1 • dairy cow
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The cow’s diet plays a role in determining the degree of unsaturation of milk fat (Baumgard et al., 2000; Perfield et al., 2006, 2007). Dietary fatty acids are hydrogenated in the rumen by bacteria and transported via the blood. In the mammary gland, fatty acids originating from the blood or from de novo fatty acid synthesis can be desaturated. Eventually, the fatty acids that are secreted into the milk determine the degree of unsaturation of milk fat. This degree of unsaturation is often addressed by a so-called index: the concentration of the unsaturated product proportional to the sum of the unsaturated product and the saturated substrate.
Studies demonstrating a significant variation in unsaturation among breeds and cows on the same diet suggest that also genetics plays a role (Beaulieu and Palmquist, 1995; DePeters et al., 1995; Lawless et al., 1999; Sol Morales et al., 2000; Drackley et al., 2001; Kelsey et al., 2003; Lock and Garnsworthy, 2002, 2003). For example, Kelsey et al. (2003) found that the milk-fat content of C18:2 cis-9, trans-11 (conjugated linoleic acid, CLA) and the CLA index varied largely among individual cows on the same diet, namely over 3-fold. Lock and Garnsworthy (2002 and 2003) reported that the C14:1/C14:0 index differed significantly between cows. However, genetic parameters for milk fatty acid unsaturation indices are scarce in literature (Royal and Garnsworthy, 2005).
Animals are capable of desaturating saturated fatty acids to 
9 unsaturated fatty acids by the stearoyl-CoA desaturase (SCD) enzyme, which catalyzes the insertion of a double bond between carbon atoms 9 and 10 of a fatty acid (Pereira et al., 2003). Two SCD isoforms have been identified in cattle, SCD1 and SCD5; SCD1 is located on chromosome 26 and expressed in a variety of tissues among which are adipose and mammary tissue, and SCD5 is located on chromosome 6 and expressed primarily in the brain (Chung et al., 2000; Lengi and Corl, 2007). A nonsynonymous SNP in exon 5 of SCD1, causing the substitution of valine with alanine (A293V), has been associated with carcass fatty acid composition in Japanese Black cattle (Taniguchi et al., 2004) and with milk fatty acid composition in Italian Holstein, Piedmontese, and Valdostana cattle (Mele et al., 2007; Moioli et al., 2007). The SCD1 A allele was associated with a higher monounsaturated fatty acid content.
Another candidate gene that may affect unsaturation is acyl CoA:diacylglycerol acyltransferase 1 (DGAT1), which is located on chromosome 14 (Grisart et al., 2002). The DGAT1 enzyme plays a key role in triacylglycerol synthesis; it catalyzes the esterification of a fatty acyl-CoA to the sn-3 position of a diacylglycerol. A lysine to alanine polymorphism in DGAT1 (K232A) explains 50% of the genetic variation in milk-fat percentage and also has a strong effect on milk fatty acid composition. The DGAT1 K allele was associated with a larger fraction of C16:0; and smaller fractions of C14:0, unsaturated C18, and CLA (Schennink et al., 2007).
Our study aims to estimate genetic variation for milk fatty acid unsaturation indices for specific fatty acids and their phenotypic and genetic correlations. Furthermore, we study the effects of polymorphisms in 2 candidate genes, namely SCD1 A293V and DGAT1 K232A, on milk fatty acid unsaturation indices.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Phenotypes
Fat and protein percentages were measured by infrared spectroscopy, using a MilkoScan FT6000 (Foss Electric, Hillerød, Denmark) at the Milk Control Station (Zutphen, the Netherlands). Fat and protein yields were calculated by multiplying fat or protein percentage by the morning milk yield. Milk yield (kg) data were missing for 145 cows. Milk fatty acid composition was measured by gas chromatography at the COKZ laboratory (Netherlands Controlling Authority for Milk and Milk Products, Leusden, the Netherlands) as described by Schennink et al. (2007). With this method, the C18:1 cis-9 peak was probably slightly overestimated due to coelution (Jensen, 2002). The cis double bond of C10:1 and C12:1 could not be ascertained at the carbon-9 position. The fatty acids were expressed as weight-proportion of total fat weight. Fatty acid unsaturation indices were calculated by expressing each product as a proportion of the product plus substrate, multiplied by 100 (Kelsey et al., 2003):
 |
We calculated indices for the following product and substrate pairs: C10:1 and C10:0 (C10 index), C12:1 and C12:0 (C12 index), C14:1 cis-9 and C14:0 (C14 index), C16:1 cis-9 and C16:0 (C16 index), C18:1 cis-9 and C18:0 (C18 index), C18:2 cis-9, trans-11 (CLA), and C18:1 trans-11 (CLA index).
Genotypes
Blood samples for DNA isolation were collected between April and June 2005. Genotypes for the SCD1 A293V polymorphism were assayed by the SNaPshot single base primer extension method (Applied Biosystems, Foster City, CA). The primer designs were based on the GenBank sequence (AY241932): forward PCR primer, 5'-TCATTTAACCCCTCATTACCTCA-3'; reverse PCR primer, 5'-TGTAAAATACTAGGCTTTCT GG-3'; genotyping primer, 5'-TGGTTTCCCTGGGAG CTG-3'. To amplify the SCD1 fragment, 12-µL PCR reactions were set up containing 20 ng of genomic DNA, 0.2 µM of each primer, and 2x AccuPrime Supermix II (Invitrogen, Carlsbad, CA). The PCR cycling conditions were 94°C for 5 min, 36 cycles of 94°C for 30 s, 55°C for 45 s, 68°C for 90 s, followed by an extension cycle of 68°C for 10 min. The PCR products were purified by incubation with shrimp alkaline phosphatase (USB, Cleveland, OH) and Exo I (USB) at 37°C for 1 h and 72°C for 15 min. Extension reactions, using 3 µL of purified PCR product and 5 pmol of genotyping primer and SNaPshot multiplex Ready reaction mix (Applied Biosystems), were performed using 40 cycles of 96°C for 10 s, 50°C for 5 s, and 60°C for 30 s. The extension products were incubated with shrimp alkaline phosphatase at 37°C for 1 h and 72°C for 15 min. Two microliters of extension product were added to 8 µL of Hi-Di formamide and electrophoresed on an ABI 3730 DNA analyzer. Results were analyzed using the GeneMapper Software v4.0 (Applied Biosystems). In total, 1,725 animals were genotyped for the SCD1 A293V polymorphism. Genotypes were missing for 208 animals because either no DNA sample was available (n = 141) or the sample could not be genotyped unambiguously (n = 67).
Genotyping of the DGAT1 K232A dinucleotide polymorphism was performed as described by Schennink et al. (2007). In total, 1,779 animals were genotyped for the DGAT1 polymorphism. Genotypes were missing for 154 animals because either no DNA sample was available (n = 141) or the sample could not be genotyped unambiguously (n = 13).
Statistical Analysis
Analyses were performed first using SAS 9.1 (2002) procedures to determine significance of fixed effects. The model included DIM (days between calving and date of sample), age at first calving, season of calving, and an effect of the differences in genetic level between proven bull daughters and young bull daughters. Variance components and genetic parameters were estimated using an Animal model in ASReml (Gilmour et al., 2002):
 |
where y was the dependent variable, µ was the general mean, dim was the covariate describing the effect of DIM modeled with a Wilmink curve (Wilmink, 1987), afc was the covariate describing the effect of age at first calving, season was the fixed effect of the class of calving season (June–August 2004, September–November 2004, or December 2004–February 2005), scode was the fixed effect of the differences in genetic level between proven bull daughters and young bull daughters, herd was the random effect of groups of animals sampled in the same herd, an was the random additive genetic effect of animal n, and e was the random residual effect. Effects of the polymorphisms were estimated using the same model, but extended with effect g: the fixed effect of the SCD1 genotype (AA, AV, or VV) or the DGAT1 genotype (KK, KA, or AA). Ungenotyped individuals were included as a separate group and appeared to be randomly distributed across other effects in the model.
The variance-covariance structure of the additive genetic effects was Var(a) = A
, where A was a matrix of additive genetic relationships between individuals and was the additive genetic variance. Heritabilities were estimated using univariate analyses. Heritability was defined as
 |
where was the additive genetic variation, and 
was the residual variation. Heringstad et al. (2006) referred to this heritability as the intraherd heritability. The intraherd heritability is the parameter that is required to predict selection responses of alternative breeding strategies. The proportion of variance due to herd (hherd) was defined as
 |
where 
herd2 was the herd variance. The proportion of additive genetic variance explained by the polymorphism was defined as
 |
where polymorphism2 was calculated by estimating the variance of the polymorphism based on the estimated genotype effects and the genotype frequencies. Corrected means for the reference group in Tables 3
, 4, and 5 were estimated by the PREDICT function in ASReml. Genetic correlations were estimated using bivariate analyses. Estimates of skewness and kurtosis for the fatty acid indices are in the same range as the estimates for milk production traits.
|
|
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
). Unsaturation indices of the short and medium-chain fatty acids C10, C12, C14, and C16 were between 2.7 and 10.9. Indices of the long-chain fatty acids C18 and CLA were higher, 67.6 and 33.7, respectively. These values are in line with those reported for dairy cattle in other studies (Royal and Garnsworthy, 2005; Perfield et al., 2006, 2007; Mele et al., 2007). Almost all C10, C12, C14, and approximately 50% of C16 fatty acids are synthesized in the mammary gland, whereas the longer-chain fatty acids as well as a proportion of the unsaturated long-chain fatty acids are derived from the blood. The dual origin of the long-chain fatty acid unsaturation may play a role in the contrast between short/medium and long-chain fatty acid unsaturation indices. An alternative explanation for this contrast may be that long-chain fatty acids are unsaturated to a larger extent. The coefficient of variation was lowest for C18 index (6%) and highest for C12 and C16 indices (about 20%).
|
). The heritability for total index was 0.30. Repeatabilities for C14, C16, and C18 indices, which are considered to be the upper limit of heritabilities, were estimated between 40 and 45% by Soyeurt et al. (2006), suggesting moderate heritabilities as well. Only Royal and Garnsworthy (2005) reported heritabilities for fatty acid unsaturation indices, based on 1,520 Holstein-Friesian cows, and reported similar values for C14 (0.30), C18 (0.19), and CLA (0.29) indices, but much lower values for C16 (0.01) and total (0.02) indices. The proportion of total variance explained by herd was small, ranging from 0.06 to 0.09 of total variance in individual unsaturation indices. For total unsaturation index the herd effect was larger (0.26). The ratio of genetic variance to herd variance showed that for all indices the genetic variance was much larger than the herd variance, except for total index, for which the herd variance was slightly larger than the genetic variance. The total unsaturation index, in fact, mainly represents the ratio of C18:1 cis-9 to C16:0 because these fatty acids are the largest unsaturated and saturated fatty acid fractions in milk. The proportion of variance explained by herd was 0.28 for C18:1 cis-9 fraction and 0.29 for C16:0 fraction in milk, which explains the relative large herd variance of the total unsaturation index (Stoop et al., 2008). The moderate to high heritabilities for unsaturation indices in combination with the moderate to high coefficients of variation indicate that the unsaturation indexes can be changed by means of selection.
|
. In comparison to the A allele, the V allele was associated with a higher proportion of C10:0, C12:0, C14:0, C16:1, and CLA, and with a lower proportion of C10:1, C12:1, C14:1, C18:0, and C18:1 trans-11. Moioli et al. (2007), who studied 27 Piedmontese and 27 Valdostana cows, found similar results, but no significant effects on C10:0 and C12:0. Mele et al. (2007), who studied 297 Italian Holstein Friesian cows, only found significant effects on C14:1 and on total monounsaturated fatty acids, similar to ours, and on C18:1 cis-9, on which we found no significant effect.
The SCD1 A293V polymorphism had significant effects on the unsaturation indices for all individual fatty acids, but not on the total unsaturation index (Table 4). The V allele was associated with lower indices of C10, C12, and C14, and higher indices of C16, C18, and CLA. The negative effect of the V allele on C10 and C14 index is in agreement with Moioli et al. (2007); however, they reported no significant effects on the other indices, which could be due to their limited sample size. The negative effect on C14 index was also found by Mele et al. (2007), but they didn’t find a significant effect on other indices.
The finding that the SCD1 polymorphism has highly significant effects on individual indices, but not on the total index, suggests that the activity of the enzyme is not affected. An altered substrate specificity, however, is supported by the differential effects on the individual fatty acid indices. Substrate specificity of SCD enzymes has been demonstrated for SCD isoforms in knockout studies in mice (Miyazaki and Ntambi, 2003). In vitro studies in rat liver microsomal preparations have shown that acyl-CoA derivatives with 12 to 19 carbon atoms were required for activity of the SCD1 enzyme and that the enzyme has substrate specificity with preference for longer-chain fatty acids (Enoch et al., 1976). Based on Enoch et al. (1976), it would be unlikely that C10:1 is desaturated from C10:0, which is, however, suggested by our finding that the C10 index is affected by the SCD1 A293V polymorphism. The enzyme function of SCD1 may be affected by the polymorphism because it causes a valine to alanine substitution on position 293, which is located in the third histidine-rich region of the enzyme. These histidine-rich regions are important for catalytic activity (Shanklin et al., 1994).
The SCD1 A293V polymorphism contributed considerably to the genetic variance (Table 4): it explained 31 to 52% of the genetic variance for C10, C12, C14, and C16 indices, but only 18% for CLA and 6% for C18 index (Mosley and McGuire, 2007).
Effects of the DGAT1 K232A Polymorphism
A total of 1,779 cows was genotyped for the DGAT1 K232A polymorphism. The frequency of 232K was 0.40, and the genotypes were in Hardy-Weinberg equilibrium. The DGAT1 A allele was associated with lower indices of C10, C12, C14, and C16, and with higher indices of C18 and CLA (Table 5). The A allele was also associated with a higher total unsaturation index. The majority of fatty acids in milk is present in the form of triacylglycerols. DGAT1 catalyzes the last step in triacylglycerol synthesis: the esterification of a fatty acyl-CoA to the sn-3 position of a diacylglycerol. The DGAT1 K232A polymorphism was reported to have an effect on, among others, C14:0, C16:0, and unsaturated C18 fatty acids in milk. Given these effects of the polymorphism on milk fatty acid composition, an effect on milk fatty acid unsaturation indices was conceivable. The effects on composition and unsaturation may be explained by different causes: a higher activity of DGAT1 and an alteration of specificity of DGAT1 (Schennink et al., 2007).
Joint Effects of the SCD1 and DGAT1 Polymorphisms
The contribution of the DGAT1 K232A polymorphism to the genetic variance of unsaturation indices was lower than the contribution of the SCD1 A293V polymorphism for all indices, except for C18 and total index (Tables 4 and 5). When the statistical model was extended to include both the SCD1 genotype and the DGAT1 genotype as fixed effects, the effects of DGAT1 on the unsaturation indices did not change considerably, meaning that the 2 polymorphisms do not explain the same part of the genetic variation (results not shown). In other words, the genetic variation explained by DGAT1 and the genetic variation explained by SCD1 are additive. Cows with the SCD1 VV genotype and the DGAT1 AA genotype (n = 41) have 9, 6, and 14% higher indices of C16, C18, and CLA, respectively, compared with cows with the SCD1 AA and DGAT1 KK genotype (results not shown, frequencies of combined genotypes in Table 6). This shows that selective breeding can improve when both the polymorphisms in SCD1 and DGAT1 are used in genetic selection.
|
). Genetic correlations were low between the medium-chain and the long-chain fatty acid indices (0.05 to 0.26). The C16 index showed low to moderate genetic correlations with C10 (0.15), C12 (0.37), and C14 (0.24) indices. Genetic correlations of C16 index with C18 and CLA indices were also moderate, although slightly higher (0.45 and 0.60). Our observations are similar to the phenotypic correlations reported in previous studies (Peterson et al., 2002; Kelsey et al., 2003). Genetic correlations of fat percentage with C10 (0.25), C12 (0.26), C14 (0.31), and C16 (0.17) indices were low, but positive, whereas genetic correlations of fat percentage with C18 (–0.35) and CLA (–0.48) indices were negative. Accordingly, an increase in fat percentage will lead to slightly higher C10, C12, C14, and C16 indices, but to lower C18 and CLA indices. Fat percentage also showed a negative genetic correlation with total index (–0.52). This might partly reflect the effects of the DGAT1 K232A polymorphism, for which the K allele is known to have a very large effect on fat percentage. In this study, it has been shown that the K allele is also associated with higher C10, C12, C14, and C16 indices, and lower C18 and CLA indices. It is likely that breeding in the Dutch dairy in the past years has resulted not only in increased fat percentage and fat yield, but also in higher C10 to C16 indices and lower C18 and CLA indices.
|
). Three groups can be identified in this figure, which almost completely coincide with the 3 genotypes of SCD1 A293V (AA, AV, and VV). Correction for the SCD1 polymorphism, by adding the SCD1 genotype to the model as a fixed effect, results in a higher genetic correlation between C14 and C16 index (0.93 with vs. 0.24 without correction, Table 7), but also in higher correlations between C10 index and C16 index (0.69 vs. 0.15) and between C12 index and C16 index (0.91 vs. 0.37), and between the medium chain C10, C12, and C14 indices and the long chain C18 and CLA indices (0.2 to 0.4 higher after correction; data not shown). Thus, the correlation due to the SCD1 A293V polymorphism is negative for these indices, whereas the correlation due to other genetic effects is positive.
|
Our study gives more insight into the process of unsaturation. Not only does SCD1 play a significant role by desaturating saturated fatty acids into unsaturated fatty acids, but also DGAT1 is important by influencing the composition of the triacylglycerols. The entire pathway of lipogenesis, which next to SCD1 and DGAT1 involves other enzymes like fatty acid synthase and acetyl-coenzyme A carboxylase, is regulated by the transcription factor sterol regulatory element-binding protein (SREBP)-1c and activated by dietary fatty acids. We made an important start to unravel the genetics of unsaturation, and this could be the start of other studies looking into the interaction between genetics and feeding.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Received for publication November 3, 2007. Accepted for publication January 8, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
9-desaturase enzyme activity in dairy cows. Page 52 in Proc. Br. Soc. Anim. Sci., York, UK. Br. Soc. Anim. Sci., Penicuik, UK.This article has been cited by other articles:
|
|
 |
|
  R. M. Demeter, G. C. B. Schopen, A. G. J. M. O. Lansink, M. P. M. Meuwissen, and J. A. M. van Arendonk Effects of milk fat composition, DGAT1, and SCD1 on fertility traits in Dutch Holstein cattle J Dairy Sci, November 1, 2009; 92(11): 5720 - 5729. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. L. Heck, H. J. F. van Valenberg, J. Dijkstra, and A. C. M. van Hooijdonk Seasonal variation in the Dutch bovine raw milk composition J Dairy Sci, October 1, 2009; 92(10): 4745 - 4755. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Glantz, H. Lindmark Mansson, H. Stalhammar, L.-O. Barstrom, M. Frojelin, A. Knutsson, C. Teluk, and M. Paulsson Effects of animal selection on milk composition and processability J Dairy Sci, September 1, 2009; 92(9): 4589 - 4603. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. M. Stoop, A. Schennink, M. H. P. W. Visker, E. Mullaart, J. A. M. van Arendonk, and H. Bovenhuis Genome-wide scan for bovine milk-fat composition. I. Quantitative trait loci for short- and medium-chain fatty acids J Dairy Sci, September 1, 2009; 92(9): 4664 - 4675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Schennink, W. M. Stoop, M. H. P. W. Visker, J. J. van der Poel, H. Bovenhuis, and J. A. M. van Arendonk Short communication: Genome-wide scan for bovine milk-fat composition. II. Quantitative trait loci for long-chain fatty acids J Dairy Sci, September 1, 2009; 92(9): 4676 - 4682. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. C. B. Schopen, J. M. L. Heck, H. Bovenhuis, M. H. P. W. Visker, H. J. F. van Valenberg, and J. A. M. van Arendonk Genetic parameters for major milk proteins in Dutch Holstein-Friesians J Dairy Sci, March 1, 2009; 92(3): 1182 - 1191. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. L. Heck, A. Schennink, H. J. F. van Valenberg, H. Bovenhuis, M. H. P. W. Visker, J. A. M. van Arendonk, and A. C. M. van Hooijdonk Effects of milk protein variants on the protein composition of bovine milk J Dairy Sci, March 1, 2009; 92(3): 1192 - 1202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Kgwatalala, E. M. Ibeagha-Awemu, A. F. Mustafa, and X. Zhao Influence of stearoyl-coenzyme A desaturase 1 genotype and stage of lactation on fatty acid composition of Canadian Jersey cows J Dairy Sci, March 1, 2009; 92(3): 1220 - 1228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Steinberg, L. Pereira, G. A. Lacorte, M. G. C. D. Peixoto, R. S. Verneque, R. L. Teodoro, M. A. Machado, C. G. Fonseca, and M. R. S. Carvalho Technical note: A new and cost-effective method for detection of the bovine acyl-CoA:diacylglycerol acyltransferase 1 K232A polymorphism in cattle J Dairy Sci, February 1, 2009; 92(2): 773 - 776. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Mele, R. Dal Zotto, M. Cassandro, G. Conte, A. Serra, A. Buccioni, G. Bittante, and P. Secchiari Genetic parameters for conjugated linoleic acid, selected milk fatty acids, and milk fatty acid unsaturation of Italian Holstein-Friesian cows J Dairy Sci, January 1, 2009; 92(1): 392 - 400. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
