The effect of mitochondrial DNA on milk production and health of dairy cattle

Livestock Production Science, 37 (1994) 283-295

283

Elsevier Science B.V., Amsterdam

The effect of mitochondrial DNA on milk production and health of dairy cattle M.M. Schutza, A.E. Freemana, G.L. Lindberga, C.M. Koehler b and D.C. Beitz a'b aDepartment of Animal Science, Iowa State University, Ames, IA, USA bDepartment of Biochemistry and Biophysics, Iowa State University, Ames, IA, USA (Accepted 20 April 1993 )

ABSTRACT Maternal lineage effects, probably indicative of mitochondrial DNA (mtDNA) differences, may be important for milk production and reproductive success in dairy cattle (Bos taurus). Sequence variation of mtDNA in 36 maternal lineages of dairy cattle was studied with animal models to assess effects on milk production and reproductive traits. Cattle within maternal lineages defined by registered pedigrees were assumed to be uniform for the nucleotide sequences examined. Sequence polymorphisms of bovine mtDNA were shown to be associated with milk production, reproduction, and health costs incurred. One particular base-pair substitution was associated with additional production of 842 kg milk and 37 kg milk fat per cow per lactation. Another single base-pair substitution was associated with a decrease of 36 days and one unsuccessful breeding between successive calvings. Effects of this size are economically important and have broad implications in genetic selection of dairy cattle. Key words: Dairy cow; Cytoplasmic inheritance; Maternal lineage; Mitochondrial DNA

INTRODUCTION

Dairy cattle breeders have made remarkable and consistent progress toward improving milk production. The current estimate of annual genetic gain for the entire US dairy herd is 139 kg milk per cow per lactation (Powell, 1992). Such gains have been accomplished by using Mendelian principles and statistical methodology to accurately predict breeding values of superior individuals for selection as parents of the next generation. The genetic increase in production has resulted almost entirely from use of additive genetic variation of nuclear origin. Several studies have demonstrated the existence of cytoplasmic inheritance of measures of production and reproduction in dairy cattle (Bell et al., 1985; Correspondence to and present address: M.M. Schutz, Animal Improvement Programs Laboratory, United States Department of Agriculture, Beltsville, MD 20705, USA.

0301-6226/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved.

SSDI 0 3 0 1 - 6 2 2 6 ( 9 3 ) E 0 0 5 5 - W

284

M.M. SCHUTZ ET AL.

Huizinga et al. 1986; and Schutz et al., 1992). Analyses of 4461 cows representing 102 maternal lineages revealed that 2.0, 1.8, and 3.5% of phenotypic variation in milk yield, milk fat yield, and percentage of fat in milk, respectively, was explained by cytoplasmic inheritance (Bell et al., 1985). Other studies have shown even higher percentages of phenotypic variation in milk and milk-component production to be attributable to maternal lineages (Huizinga et al. 1986; and Schutz et al., 1992 ). Effects of cytoplasmic inheritance on reproductive measures have been shown previously for number of days open (days from calving to next conception), days from calving to first detected estrus, first service conception rate, and number of services until conception (Bell et al., 1985; Huizinga et al., 1986; Faust et al., 1989). It is well known that dairy cattle deficient in energy have lower conception rates (Harrison et al., 1990); and, of course, mitochondria are intimately associated with cellular energy metabolism. These results imply that mitochondrial DNA (mtDNA) genotype is important because mitochondria are transmitted only from female parents to ensuing offspring (Gyllensten et al., 1985 ). Limited biparental inheritance has been reported in Mytilus [mussels (Hoeh et al., 1991 ) ] and Mus spretus [mice (Gyllensten et al., 1991 ) ], but not in other mammals. Mitochondrial lineage influences on health differences of cattle have not been examined previously. Much work associating human diseases with mtDNA sequence differences has been reported (Merrill and Harrington, 1985 ). Kearns-Sayres Syndrome (KSS) and Leber's hereditary optic neuropathy (LHON) are examples of such diseases (Grivell, 1989; Wallace, 1989 ). Indeed, LHON is correlated to a single guanine-adenine transition that converts an arginine to a histidine in NADH dehydrogenase subunit 4 gene of mtDNA (Wallace, 1989). MtDNA sequence substitutions may also affect health in cattle. Molecular variation in bovine mtDNA has been demonstrated through RFLP analysis (Watanabe et al., 1985; Brown et al., 1989; Koehler et al., 1991 ) and comparison of nucleotide sequences (Olivo et al., 1983). Displacement loop (D-loop) sequences of mtDNA [ approximately 910 base-pair (bp) ] from 36 distinct registered maternal lineages available for our study were compared previously (Lindberg, 1989). Fifty-one sequence differences were located, including 48 single bp substitutions, one 9 bp deletion, and two variable length poly G-C runs. Where possible, D-loops from two or more animals of the same maternal lineage were sequenced to verify accuracy of mtDNA isolation and nucleotide sequencing and to confirm constancy of mtDNA within maternal lineages (Lindberg, 1989 ). Previous reports citing evidence for cytoplasmic or maternal lineage effects in cattle have used statistical associations between lineages to infer an implied relationship between mtDNA and phenotypic expression of economically important traits. Assignment of cattle to lineages was according to maternal pe-

EFFECT OF MITOCHONDRIAL DNA ON MILK PRODUCTION

285

digrees based on a limited number (3 to 22 ) of generations. Only two previous reports have attempted to determine the effects of mtDNA polymorphism on economically important traits of any domestic livestock species (Brown et al., 1989; Schutz et a1.,1993). Brown et al. (1989) compared cows with or without a single nucleotide substitution, later determined to be highly heterogeneous even in single maternal lineages of dairy cattle (Koehler et al., 1991 ). Schutz et al. (1993) compared lineages identical or polymorphic at a single nucleotide site with previously noted evolutionary implications (Lindberg, 1989). They reported small significant effects on fat production and on energy levels in milk. The objective of this study was to determine whether actual mtDNA D-loop sequence polymorphisms are associated with differences in production, reproduction, and health traits of dairy cattle. MATERIALS

Dairy cows in this study were from a selection experiment founded at Iowa State University in 1968. Heifers for this herd were purchased from 38 Holstein breeders throughout Iowa to keep the herd as genetically broad-based as possible. Cows were bred artificially to sires from commercial artificial insemination organizations, allowing a continuous influx of nuclear genes. Females were assigned to groups and artificially mated to bulls with either high or average estimated additive genetic transmitting ability for milk yield. Females born in each group were mated to new bulls chosen for that group, thus forming divergent selection lines that differed by 1308 kg of milk per cow per lactation when these data were analyzed. Frequencies of bovine lymphocyte antigen phenotypes were similar to frequencies in the U.S. Holstein population, implying that these nuclear genes are likely to be representative of the entire U.S. Holstein population (Weigel et al., 1990). Ancestral pedigrees of registered foundation females in the herd were tracked backward through the Holstein-Friesian Herd Book (Wales, 1885 ). Eighty-one distinct maternal lineages were defined by convergence of maternal pedigrees after 1885. Probably, these lineages would have been found to converge to fewer lineages had registration records been kept before importation of these cows from Europe in about 1885. Thirty-six maternal lineages still had surviving members in the herd when samples for nucleotide sequencing were collected. Each of these lineages originated from as many as six purchased foundation females in the herd. Mitochondrial DNA was isolated from blood; and nucleotide sequence polymorphism data were obtained (Koehler et al., 1991 ). All cows within the same maternal lineage that were ever in the herd were assumed to have identical mtDNA. Spot checking of lineages has strengthened the validity of this assumption.

286

M.M. SCHUTZ ET AL.

Table 1 has location, type, and frequency of the 17 most common sequence variants of the mtDNA D-loop in this herd of Holstein dairy cattle. Nomenclature for specific bp sites is according to Anderson et al. (1982). Only those substitutions occurring in at least 4% of cows in the herd are listed because information on markers occurring in a very small number of cows would not be statistically informative. Nucleotide transitions at bp 169 and 216 occurred in 80 and 84 percent of cattle, respectively. The probable explanation is that the cow originally sequenced by Anderson et al. ( 1982 ) had the rarer genotype at those two sites. Milk production records of all cows in the 36 maternal lineages with known mtDNA D-loop sequences were considered. Milk and fat yield records adjusted to a uniform age and lactation length, or mature-equivalent (ME), basis were obtained. Percentages in milk of fat and solids-not-fat (SNF), which is total solids in milk less fat in milk, were known for each record of each cow. Up to seven production records were used for individual cows. Because mitochondria play an extensive role in energy metabolism, mtDNA polymorphism may alter energy content in milk. Fat, protein, and lactose are the carriers of energy in milk. However, information was complete since 1968 only for fat and SNF, which combines protein, lactose, and minerals. Net energy concentration in milk, which is based on fat and SNF (Tyrell and Ried, 1965 ), was calculated in terms of kilocalories per kilogram and multiplied by lactation milk yield to approximate lactation energy production in terms of megacalories. Number of days open and number of breedings (artificial inseminations), along with reproductive costs per lactation, were the measures of reproducTABLE 1 Location, type and frequency of seventeen most common sequence polymorphisms of mtDNA Dloops in a herd of dairy cattle Location in D-loop ~ 8 106 169 216 363 16022 16049 16057

Polymorphic event

Frequency

Location in D-loop ~

Polymorphic event

Frequency

G-A T-C A-G Var. length G-C run C-G* G-A C-T G-A

0.07 0.14 0.80 0.84

16058 16074 16085 16111 16113 16141 16230 16231 16247

C-T T-C T-C A-C* T-C T-C C-T C-T C-T

0.12 0.07 0.05 0.04 0.11 0.11 0.06 0.12 0.13

0.46 0.14 0.08 0.12

~Location is defined by the first published mtDNA sequence (Anderson et al., 1982 ). Polymorphic event and frequency is also with regard to that sequence. *Base pair substitution is a transversion.

EFFECT OF MITOCHONDRIAL DNA ON MILK P R O D U C T I O N

287

tion considered. Reproductive costs included costs of insemination and semen, as well as costs of reproductive examinations and treatments for postcalving disorders such as metritis and retained placenta. Health differences were measured by m a m m a r y costs and total health costs. Using costs of health disorders was necessary because incidences of specific events or diseases occur too infrequently to be of use in analyses of data from a single herd. Health costs were grouped by body systems, and total health cost was the sum of reproductive, digestive, mammary, respiratory, and skin and skeletal costs per lactation. M a m m a r y costs reflected costs of treatment and medications required for cows with mastitis or injured teats, but discarded milk value was not included.

METHODS To evaluate the effect of m t D N A D-loop sequence differences, each cow was assigned a value of 0 or 1 for each of the 17 locations considered. The assigned value was 0 if the determined base-pair sequence matched the first published bovine m t D N A sequence (Anderson et al., 1982 ) or 1 if it did not match (Table 1 ). Each production trait was analyzed individually with the animal model (Henderson, 1984; Westell et al., 1988 ): Yijk,m = fl+YSi +Pj +Xk +ill SI + ...+fll7S17 +PEI +a~ +em,

(1)

where Yijknp was the milk, fat, fat percentage, SNF, SNF percentage, energy concentration, or lactation energy record; # was an overall mean; YSi was the effect c o m m o n to all cows calving in year-season i; Pj was the effect c o m m o n to all cows in parity j; Xk was the effect c o m m o n to cows in either the high or average selection line; fl~ to fl17 were regressions of the production record on m t D N A D-loop sequence polymorphisms S, to S,7, (Sn= 0 or l ); PE~ was permanent environmental effect on all records of cow l; a~ was the effect of animal 1 and was composed of the additive genetic contribution of sire and dam breeding values and a Mendelian sampling effect; and em was a random residual. The effect of m t D N A D-loop bp substitutions on production traits was of primary interest. Effects of the overall mean, year-season of calving, parity, and selection line were treated as fixed effects in the mixed model to account for explainable environmental background. Additive genetic covariances among related individuals were incorporated in this model (Henderson, 1984; Westell et al., 1988 ). Permanent environmental and additive genetic effects were treated as random and had properties of Best Linear Unbiased Prediction (BLUP) (Henderson, 1984). R a n d o m effects were assumed to be nor-

288

M . M . S C H U T Z E T AL.

TABLE2 Literature estimates of variance ratios used for Best Linear Unbiased Prediction (BLUP) analyses Trait

(7 e2 /O. a2

2 ae2 / ape

Source

Milk (kg) Fat (kg) Solids not fat (kg) Lactation energy (mcal) Fat (%) Solids not fat (%) Energy concentration Days open (d) Number of breedings (n) Reproduction costs ($) Mammary costs ($) Total Health Costs ($)

1.7 1.1 1.9 0.4 0.4 0.6 1.5 33.8 43.5 40.5 12.8 6.6

1.3 1.3 1.3 1.4 1.5 10.8 1.3 9.4 ~ 7.9 7.2 2.1 5.0

Schutz et al., 1992 Schutz et al., 1992 Schutz et al., 1992 Schutz et al., 1992 Schutz et al., 1992 Schutz et al., 1992 Schutz et al., 1992 Hansen et al., 1983 Lyons et al., 1991 Lyons et al., 1991 Lyons et al., 1991 Shanks et al., 1982

~Variance ratio from Sylva et al., 1976.

mally and independently distributed with mean expectations of zero. Variance among permanent environments was assumed to be var(PE)=Irrp2E, where I is an identity matrix, and a~E is permanent environmental variance. That is permanent environmental effects between cows were assumed to be uncorrelated. Variance among animals was assumed to be var (a) = A rraz, where A is the numerator relationship matrix, representing additive genetic covariances among related individuals, and aa2 is additive genetic variance (Henderson, 1984). A included sires and dams of all cows back to foundation cows and included information for sires and paternal grandsires of bulls with daughters in the herd. Estimates of variance ratios for the BLUP analyses were from the literature and are in Table 2. Regression coefficients for sequence polymorphism, which were Best Linear Unbiased Estimates (BLUE), and tests of significance were obtained by iterative methods described elsewhere (Schutz et al., 1991 ). Altogether, 1800 records of 728 cows were used in this study. Along with effects of 17 sites of m t D N A variation, there were 33 year-season, 7 parity, 2 selection line, 728 permanent environment, and 950 animal effects. Animal, or additive genetic, effects were for 728 cows with records, 197 relatives without records, and 25 phantom parent groups (Westell et al., 1988 ) to account for genetic similarities among animals without relationship or production information. Animals were assigned to these two-year unknown parent groups according to year of birth. RESULTS

Table 3 has overall means of production traits along with their regressions on nucleotide sequence differences. The polymorphism at bp 363 previously

EFFECT OF MITOCHONDRIAL DNA ON MILK PRODUCTION

289

TABLE 3 Regression coefficients of production on sequence polymorphisms in the mtDNA displacement-loop (D-loop) and overall production means in a herd of dairy cattle I Location in D-loop 2 8 106 169 216 16022 16049 16057 16058 16074 16085 16111 16113 16141 16230 16231 16247 Overall mean 3

Milk (kg)

Fat (kg)

SNF (kg)

Lactation energy (mcal)

Fat (%)

SNF (%)

Energy concentration (kcal/kg)

235 464 482* -157 -113 989* -577 t 39 842* -197 107 32 -336 383 -650* 351

1 16 24** 1 -6 29 -21 20 37** -20 -5 3 -8 33 -28 21

19 47 51" -7 -26 102* -48 9 85* -36 15 -10 -35 39 -61* 32

85 343 456* -10 -162 756 -395 225 749** -343 52 3 -198 490 -522 t 346

-0.21 -0.03 0.05 0.07 0.04 -0.07 0.01 0.30* O. 14 -0.21' -0.05 0.09 0.10 0.39** -0.12 0.17"

-0.04 0.08 0.05 0.09 t 0.01 0.17 0.11 0.06 0.05 -0.19 t 0.05 -0.21' -0.05 0.06 -0.01 -0.01

-21 2 6 10 3 1 7 3 It 16 -29 t -3 -3 6 39* - 12 16 t

8085 (1771)

288 (60)

745 (165)

5888 (1225)

3.63 (0.44)

9.21 (0.40)

732 (50)

tiPs< 0.10; *P~<0.05; **P~<0.01. 2Location is defined by the first published mtDNA sequence (Anderson et al., 1982 ). 3Overall standard deviations are in parentheses.

has been associated with milk and fat yield and fat percentage (Brown et al., 1989 ). This site has since been shown to be highly heterogeneous within maternal lineages (Koehler et al., 1991 ) and is therefore unstable for use as a marker. Regressions of production on the binomial indicator of polymorphism at bp 363 are not reported because, unlike for other sites, the sequence difference cannot be assumed uniform within maternal lineages. Eleven nucleotide sequence substitutions were associated significantly with at least one production trait, and traits were influenced in both positive and negative directions (Table 3). A single adenine-to-guanine transition at bp 169 related to increased production of 482 kg of milk, 24 kg of fat, 51 kg of SNF, and 456 megacalories of energy per cow per lactation. Base pair substitution at site 16074 had a large positive effect on milk, fat, and SNF yield and lactation energy, whereas the substitution at site 16231 had a negative effect on the same traits (Table 3 ). Fat percentage in milk and energy concentration of milk were significantly affected in cows with poly-

290

M.M. SCHUTZ ET AL.

morphism at sites 16058, 16085, 16230, and 16247. Effects o f b p changes on SNF percentage were small. Table 4 has regressions of reproduction and health costs on binomial representations of sequence polymorphisms in the mtDNA D-loop. The transition at bp 169 was associated with increased days open, and reproductive costs and days open were increased in animals polymorphic at bp 16058 (Table 4 ). There was a very large favorable impact on the reproductive complex in cows with m t D N A D-loop sequence difference at bp 16085. This single T to C transition was related to a decrease of 36 days open, one insemination, and $12.82 in reproductive costs. Thirty-six days open is nearly two reproductive cycles. Mammary costs were altered in cows with m t D N A D-loop sequence variation at four locations (Table 4). None were locations significantly affecting milk or milk fat yield or fat percentage. Total health costs were larger when transitions at bp 106 and 16074 occurred, and smaller when transitions at bp TABLE4

Regression coefficients of reproduction and health costs per lactation on sequence polymorphisms in the mtDNA Displacement-loop (D-loop) and overall means of reproduction and health cost means in a herd of dairy cattle ~ Location in

Days open

Number of breedings

Reproduction costs

Mammary costs

D-loop 2

(d)

(n)

($)

($)

8 106 169 216 16022 16049 16057 16058 16074 16085 16111 16113 16141 16230 16231 16247

-31.7 0.5 14.6' -0.5 - 13.4 16.1 -9.7 28.3* 3.7 -36.3* - 14.1 -5.5 - 11.7 -4.4 - 16.0 - 8.2

Overall mean 3

135 (72)

-0.73 0.13 0.34 0.12 -0.12 0.28 0.14 0.68 0.34 -0.99* -0.64 -0.63 -0.36 -0.02 -0.53 0.02 2.61 (2.24)

--6.12 - 2.66 6.05* 0.07 -3.89 0.88 -2.90 2.44 6.28 - 12.82' -4.06 - 1.64 -3.01 -2.04 -3.53 2.47 38.73 (34.49)

24.79* 16.24* 0.84 -9.70* 9.77 4.58 -9.31 2.94 13.56 -2.31 -8.94 1.18 - 15.67* 13.10 -4.81 3.19 19.07 (43.8)

l'P-..< 0.10, *P~< 0.05.

2Location is defined by the first published mtDNA sequence (Anderson et al., 1982). 3Overall standard deviations are in parentheses.

Total health costs

($)

10.22 21.28* 11.47 - 14.54' 1.83 17.94 - 15.93 18.44 25.93* - 19.70 - 11.19 -2.24 - 19.58 12.12 - 11.30 7.71 77.67 (69.14)

EFFECT OF MITOCHONDRIAL DNA ON MILK PRODUCTION

2 91

216 occurred. Significant effects were not observed for digestive, respiratory, and skin and skeletal cost categories. Incidences of health disorders in these systems were sparse. DISCUSSION

The D-loop region of mtDNA does not code for any known gene products; hence, sequence variation in this region should not alter the specific amino acid sequence of metabolic chain subunits. Promoters for transcription of both heavy and light strands of mtDNA as well as the origin of heavy strand replication, however, lie within the D-loop region. Thus, sequence differences in the mtDNA D-loop may alter transcription or replication rates. Alternatively, such D-loop polymorphisms may serve as indirect markers for differences elsewhere on the mtDNA genome in coding regions of genes directly affecting phenotypic expression of some traits. On a purely evolutionary basis, bp169 previously has been found to demarcate two distinct mitochondrial families of cattle in the Holstein population (Lindberg, 1989). Schutz et al. ( 1993 ) reported effects of nucleotide substitution at bp 169 on milk and component yields that were similar but slightly smaller than those reported herein. They analyzed the bp 169 genotypes in the absence of information about other nucleotide sites. In that report, the effect of substitution at bp 169 had a significant effect on fat percentage. In this study, when several sites were analyzed at once, the effect was in the same direction but not significant at a 0.05 level. This points out that substitution effects in the D-loop should be considered while simultaneously accounting for other substitutions. These D-loop sequence polymorphisms probably do not directly affect production traits but mark differences in gene coding regions elsewhere on the mtDNA genome, which is transmitted in its entirety. Different D-loop polymorphisms would be associated with alterations in gene coding regions that arose in the same maternal lineages over time. Results from this study should be interpreted with some caution. Small frequencies of occurrence of some substitutions (Table 1 ) and the limited population of cows in this study reduced the power of tests to determine significant differences related to bp substitutions. For example, the effect of substitution at bp 16049 was 989 kg of milk (Table 3) or 0.55 standard deviations. Yet the effect was not significant (P> 0.05 ) because the substitution occurred in only 8 percent of the cows. Confirmation of these results with more information will be useful. The total number of statistical comparisons made in this study was 192 ( 16 bp substitutions X 12 traits). Thus a number of significant results are expected by chance alone. Indeed, the number of significant tests expected by chance would be 19, 10, and 2 for P<0.10, P<0.05, and P<0.01, respectively. The actual number of significant outcomes was 38, 14, and 4, respec-

292

M.M. SCHUTZ ET AL.

tively, which was distinctly more than expected. This is especially so for the production traits (Table 3), and therefore the results strongly indicate true mitochondrial effects on production. Evidence for a particular substitution having an important effect is less well supported but could be strengthened by further replication in other herds. It should also be noted that correlations among traits may have played a role in determining the number of significant results. But even after grouping closely related traits (such as milk, fat, and SNF kg or days open, number of breedings, and reproduction costs), the number of significant outcomes is greater than anticipated for any chosen level of P. Generally, reproductive success is negatively associated with yield levels which may be related to energy status of lactating cows (Harrison et al., 1990 ). Bell et al. ( 1985 ) found small effects of maternal lineages on days open and pointed out the role of mitochondria in the biosynthesis of steroids. In this study, the nucleotide sites with largest effects on milk, fat, and SNF yield (bp169, and bp16074), affected reproduction traits in the same (unfavorable) direction. That is, increased yield was associated with poorer reproduction, which is an expected result. The substitution with the largest impact on reproduction (bp16085) influenced reproduction measures and yield in a negative direction. Despite this association between yield and reproductive traits, cause and effect relationship obviously cannot be established. The milk components examined in this study were fat and SNF. The effect on SNF is a composite of effects on protein and lactose, which also are energy rich components of milk. Possibly, larger effects would have been observed for lactose and protein if data for those components were individually available. Effects as large as these found are certainly economically important especially for fat yield and percentage. IMPLICATIONS

Current dairy cattle breeding programs center around selection of bulls used for artificial insemination. The dam-to-cow pathway, on the other hand, traditionally has been selected least intensely (VanTassell and Van Vleck, 1991 ). New developments in reproductive technology and embryo manipulation seem poised to make this pathway of selection more viable. Differences in mtDNA could be incorporated into embryo transfer breeding programs to better choose donor and recipient females to produce replacement heifers. Current cloning techniques require nuclear transplantation into an enucleated ovum without regard to cytoplasmic content. Potential exists for using mtDNA sequence polymorphism to identify ova of females with inferior nuclear genetics in superior mtDNA background as candidates for enucleation and subsequent introduction of nuclei with greater genetic potential. Multiple ovulation and embryo transfer (MOET) breeding programs should continue to increase in number. Biases in estimation of the transmitting abil-

EFFECTOF MITOCHONDRIALDNAON MILKPRODUCTION

293

ity of young sires in selection assisted by MOET would result when female sib selection is employed. This occurs because mitochondrial effects are included in females' estimated transmitting abilities, but are not passed through male sibs to their daughters. Two-stage selection is practiced for bulls entering artificial insemination organizations in most national breeding schemes. The first stage is pedigree selection, that is based on sire and dam information. Bulls are finally chosen for extensive use on the basis of a progeny test. The largest selection differentials are for the sire-to-bull pathway (VanTassell and Van Vleck, 1991 ), where mtDNA polymorphism is not important if the mitochondrial genome is transmitted only from female parents. The dam-to-bull pathway is equally important, but the accuracy of selection is less than at the sire-to-bull pathway. A bull's estimated transmitting ability based on pedigree may be biased if the contribution from his dam is not adjusted for mitochondrial influence on her records. This is most important in the first stage of bull selection. Although the bull would acquire mtDNA from the dam, it would not be transmitted to his offspring. Adjustment of bull dam's records for mtDNA influences would allow more accurate prediction of expected genetic contribution of a bull to his daughters. Such adjustment could be based upon genetic markers of the bull's mitochondria, because the entire genome is inherited only from the dam. ACKNOWLEDGEMENT

Appreciation is given to the National Association of Animal Breeders, Columbia, MO, and 21 st Century Genetics Cooperative, Shawano, WI for partial financial support. Johann Detilleux and Martin Sieber are acknowledged for their help in manuscript preparation. This paper is number J-14755 and project number 1053 of the Iowa Agriculture and Home Economics Experiment Station, Ames.

REFERENCES Anderson, S., deBruijn, M.H.L., Coulson, A.R., Eperon, I.C., Sanger, F. and Young, I.G., 1982. Complete sequence of bovine mitochondrial DNA. J. Mol. Biol., 156:683-717. Bell, B.R., McDaniel, B.T. and Robison, O.W., 1985. Effects of cytoplasmic inheritance on production traits of dairy cattle. J. Dairy Sci., 68:2038-2051. Brown, D.R., Koehler, C.M., Lindberg, G.L., Freeman, A.E., Mayfield, J.E., Meyers, A.M., Schutz, M.M. and Beitz, D.C., 1989. Molecular analysis of cytoplasmic genetic variation in Holstein cows. J. Anim. Sci., 67:1926-1932. Faust, M.A., Robison, O.W. and McDaniel, B.T., 1990. Animal model estimates of cytoplasmic line constants for yield in Holsteins. J. Anim. Breed. Genet., 107:401-410. Grivell, L.A., 1989. Small, beautiful, and essential. Nature 341: 569 - 571.

294

M.M. SCHUTZ ET AL.

Gyllensten, U., Wharton, D. and Wilson, A.C., 1985. Maternal inheritance of mitochondrial DNA during backcrossing of two species of mice. J. Hered., 76:321-324. Gyllensten, U., Wharton, D., Josefson, A. and Wilson, A.C., 1991. Paternal inheritance of mitochondrial DNA in mice. Nature 352: 255-257. Hansen, L.B., Freeman, A.E. and Berger, P.J., 1983. Yield and fertility relationships in dairy cattle. J. Dairy Sci., 66: 293-305. Harrison, R.O., Ford, S.P., Young, J.W., Conley, A.J. and Freeman, A.E., 1990. Increased milk production versus reproductive and energy status of high producing dairy cows. J. Dairy Sci., 73: 2749-2758. Henderson, C.R., 1984. Application of Linear Models in Animal Breeding. University of Guelph, Guelph Canada. Hoeh, W.R., Blakely, K.H. and Brown, W.M., 1991. Heteroplasmy suggests limited biparental inheritance ofMytilus Mitochondrial DNA. Science 251: 1488-1490. Huizinga, H.A., Korver, S., McDaniel, B.T. and Politiek, R., 1986. Maternal effects due to cytoplasmic inheritance in dairy cattle. Influence on milk production and reproduction traits. Livest. Prod. Sci., 15:11-26. Koehler, C.M., Lindberg, G.L., Freeman, A.E., Myers, A.M., Mayfield, J.E. and Beitz, D.C., 1989. Complete replacement of a bovine mitochondrial genotype in a single generation. Genetics 129: 247-253. Lindberg, G.L., 1989. Sequence heterogeneity of bovine mitochondrial DNA. Ph.D. diss. Iowa State University, Ames. Univ. Microfilm Order No. DA-9014925. Ann Arbor, MI. Lyons, D.T., Freeman, A.E. and Kuck, A.L., 1991. Genetics of health traits in Holstien cattle. J. Dairy Sci., 74:1092-1100. Merrill, C.R. and Harrington, M.G., 1985. The search for mitochondrial inheritance of human diseases. Trends Genet., 1: 140-144. Olivo, P.D., Van de Walle, M.J., Laipis, P.J. and Hauswirth, W.W., 1983. Nucleotide sequence evidence for rapid genotypic shifts in the bovine mitochondrial DNA D-loop. Nature 306: 4OO. Powell. R.L., 1992. Breeding for profitable Holsteins in the USA. Hol. Sci. Rept., 1-4. Schutz, M.M., Freeman, A.E., Beitz, D.C.,and Mayfield, J.E., 1992. The importance of maternal lineage on milk production of dairy cattle. J. Dairy Sci., 75:1331-1341. Schutz, M.M., Freeman, A.E., Lindberg, G.L. and Beitz, D.C., 1993. Effects of maternal lineages grouped by mitochondrial genotypes on milk yield and composition. J. Dairy Sci., 76:621-629. Shanks, R.D., Berger, P.J., Freeman, A.E. and Dickinson. F.N., 1982. Genetic and Phenotypic aspects of health traits by lactation. J. Dairy Sci., 65:1601-1611. Sylva, H.M., Wilcox, C.J., Spurlock, A.H. and Martin, F.G., 1976. Factors affecting days open, generation length and calving interval of Florida dairy cows. J. Dairy Sci., 59 (Suppl., 1 ): (Abstr.) Tyrrell, H.F. and Reid, J.T., 1965., Prediction of the energy value of cow's milk. J. Dairy Sci., 48:1215. VanTassell, C.P. and Van Vleck, L.D., 1991. Estimates of genetic selection differentials and generation intervals for four paths of selection. J. Dairy Sci., 74:1078. Wales, T.P., Jr., 1885. Holstein Friesian Herd Book. Egbart, Fidlar and Chambers. Davenport, IA. Wallace, D.C., 1989. Mitochondrial DNA mutations and neuromuscular disease. Trends Genet., 5: 9-13. Watanabe, T., Hayashi, Y. Semba, R. and Ogasawora, N., 1985. Bovine mitochondrial DNA polymorphism in restriction endonuclease cleavage patterns and the location of the polymorphic sites. Bioch. Genet., 23: 947-957. Weigel, K.A., Freeman, A.E., Kehrli, Jr., M.E., Stear, M.J. and Kelley, D.H., 1990. Association

EFFECTOF MITOCHONDRIALDNAON MILKPRODUCTION

295

of class I bovine lymphocyte antigen complex alleles with health and production traits in dairy cattle. J. Dairy Sci., 73: 2538-2546. Westell, R.A., Quass, R.L. and Van Vleck, L.D., 1988. Genetic groups in an animal model. J. Dairy Sci., 71:1310. RESUME Schutz, M.M., Freeman, A.E., Lindberg, G.L., Koehler, C.M., et Beitz, D.C., 1994. Les effets de I'ADN mitochondrial sur la production laiti~re et la sant6 des bovins laitiers. Livest. Prod. Sci., 37:283-295 (en anglais). Les lign6es maternelles, sans doute indicatrices de diff6rences au niveau de I'ADN mitochondrial (mtADN), peuvent &re importantes pour la production laiti~re et pour l'avenir reproducteur des bovins laitiers (Bos taurus). La variation au niveau des sequences du mtADN ont 6t6 6tudi6es dans 36 lign6es maternelles bovines avec des modules animaux pour d6terminer leurs effects sur la production laiti~re et les caract~res reproducteurs. Les bovins avec des lign6es maternelles p6digr6-enregistr6es ont 6t6 concid6r6s comme uniformes pour les sequences de nucl6otides examin6es. Des polymorphismes du mtADN ont 6t6 associ6s avec la production laiti~re, la reproduction, et les coots de sant6 rencontr6s. En particulier, une substitution dans une paire de bases fut associ6e avec une production laiti~re additionelle de 842 kg et de 37 kg de graisse de lait par vache et par lactation. Une autre substitution darts une paire de bases rut associ6e avec une diminution de 36 jours et un accouplement sans succ~s entre velages successifs. Des effets de cet ordre sont 6conomiquement importants et ont de large implications dans la s61ection des bovins laitiers. KURZFASSUNG Schutz, M.M., Freeman, A.E., Lindberg, G.L., Koehler, C.M., und Beitz, D.C., 1994. Der Einfluss mitrochondrialer DNS auf Milchleistung und Gesundheit yon Milchvieh. Livest. Prod. Sci., 37:283-295 (aufenglisch). Der Einfluss miitterlicher Linien, wahrscheinlich ein Zeichen von mitrochondrialen DNS Unterschieden, kann wichtig sein ftir die Milchleistung und erfolgreiche Reproduktionsleistungen bei Milchktihen (Bos Taurus). Folgvariationen der mtDNS wurde bei 36 miJtterlichen Linien mit Hilfe eines Tiermodells untersucht um den Einfluss auf die Milchleistung und Reproduktionsmerkmale festzustellen. Es wurde angenommen, dass Kiihe innerhalb der miJtterlichen linien, die durch die Herdbuch Abstammung festgelgt wurden, gleich seien bez0glich der unterzuchten Nukleotid Folgen. Es konnte gezeigt werden, dass Folge polymorphismus der mtDNS mit Milchleistung, Reproduktion und Gesundheits-Kosten in Zusammenhang stehen. Der Austauch eines bestimten Basenpaarres wurde in Zusammenhang mit der zus~itzlichen Produktion von 842 kg Milchmenge und 37 kg Milchfett pro Kuh pro Laktation gesetzt. Ein anderer Basenpaar Austausch wurde in Verbindung gesetzt mit einer Abnahme yon 36 Tagen Zwischenkalbezeit und einer erfolglosen Besamung weniger, bei zwei aufeinanderfolgenden Kalbungen. Einfliisse dieser Gr6ssenordnung sind wirtschaftlich von Bedeutung und haben eine weite Anwendung bei der genetischen Auswahl von Milchkiihen.