An Embryo Transfer Study of Reciprocal Cross Differences in Growth and Carcass Traits of Duroc and Landrace Pigs' T. M. Wilken, L. L. Lo, D. G. McLaren2, R. L. Fernando, and P. J. Dziuk

ABSTRACT: Reciprocal cross differences have been reported for growth rate and carcass traits in F1 pigs with the Duroc (D1 as a parent breed. Such differences are synonymous with maternal effects if effects of sex linkage and genomic imprinting are negligible. In the present study, transfer of embryos (ET1 to paternal breed recipients partitioned effects occurring at or before fertilization from postfertilization effects for growth and carcass traits in F1 D-Landrace (Ll pigs. Fifteen boars sired 115 F1 litters, 49 produced by ET. Growth rate of 349 barrows and 361 gilts and carcass measurements on 256 barrows and 159 gilts were analyzed assuming mixed linear models with animal and litter as random effects. Contrasts among genotype (D x L, L x D) - treatment (ET, non-ET) means were tested. Reciprocal cross differences were not detected for growth rate or for carcass weight, length, average backfat thickness, estimated carcass lean, or lean per day of age. Reciprocal cross differences for loth rib backfat thickness (BFI and longissimus muscle area (LMA) were detected only in barrows. The sexual dichotomy for reciprocal

cross differences followed expectations for a Ylinked gene(s1, consistent with the fact that reciprocal D - L crossbred barrows exhibited a paternal effect, with responses more like the sire breed than the dam breed. Barrows that were non-ET from D sires and L dams had 3.9 cm2 larger LMA and 5.8 mm less BF than barrows from L sires and D dams P < .0011. Barrows from ET sired by D boars had 3.8 cm2 larger LMA than did barrows from ET sired by L boars CP < .001), although no difference was detected for BF. Barrows sired by D boars reared in a D postfertilization environment (ET) had 6.2 cm2 greater LMA and 4.1 mm less BF ( P < .05) than barrows sired by L boars gestated and reared by D dams (non-ET). Barrows sired by D boars reared by L dams (non-ET1 had 1.5 cm2 greater LMA and 2.3 mm less BF ( P > .lo) than barrows sired by L boars reared by L dams (ET).In conclusion, reciprocal cross differences detected for BF and LMA in barrows were established before or a t fertilization and seemed to be Ylinked.

Key Words: Carcasses, Embryo Transfer, Maternal Effects, Paternal Effects, Pigs, Reciprocal Effects

J. Anim. Sci. 1992. 70:2349-2358

Introduction The dam has been shown to influence postnatal traits of her offspring beyond her nuclear genetic contribution in many mammals (Hohenboken, 1985). This effect of the dam on the offspring is

'This study was supported in part by a grant from the National Pork Producers Council and by the Commodity Credit Corp., Washington, DC. 2To whom correspondence should be addressed. Present address: Pig Improvement Co., P. 0. Box 348, Franklin, KY 42135-0348.

Received August 28, 1991. Accepted March 21, 1992.

referred to as a maternal effect (Willham, 1972). Maternal effects may be exerted at fertilization by cytoplasmic inheritance or may result from the dam's uterine and(or1 postuterine environments. In pigs, significant reciprocal cross differences have been reported for both postweaning growth and carcass traits in crosses involving the Duroc breed (Ahlschwede and Robison, 1971; Bereskin et al., 1971; Young et al., 1976; Bereskin and Davey, 1978; Schneider et al., 1982; Bereskin, 1983; McLaren et al., 1987). Reciprocal cross differences are considered synonymous with maternal effects assuming that effects of sex linkage and genomic imprinting are negligible. There has, however, been little work to determine the cause(s1 of these effects. The

2349

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

Department of Animal Sciences, University of Illinois, Urbana 6 1801

2350

WILKEN ET AL.

objective of this study was to partition reciprocal cross differences for growth and carcass traits in F1 Duroc (DI- Landrace (L)barrows and gilts into effects occurring at or before fertilization vs postfertilization effects by use of embryo transfer (ET).

Management and Data Collection. Five farrowing groups, each of approximately 14 D and 14 L sows, were maintained at the University of Illinois Moorman Swine Research Farm during the course of this experiment. In addition, a separate pool of approximately 20 D and 20 L gilts available for breeding each month was maintained for ET. Each month, of the 34 available sows and gilts of each breed, six were bred to a boar of their breed, eight were bred to a boar of the opposite breed, and 10 donors and 10 recipients were prepared for ET. Herd boars were chosen to represent as broad a genetic base within each breed as possible and gilt replacements were randomly selected on a withinlitter basis (Lo et al., 1992). Pigs used in this experiment came at random from boars and gilts in the herd. The 10 gilts of each breed destined to be embryo donors were mated to boars of the opposite breed. The remaining 10 gilts served as recipients of embryos collected from a gilt of the opposite breed. The same boars that were used to produce the non-ET F1 litters were mated to the ET donor dams as far as possible. Breeding took place during the 1st and 2nd wk of each month. Sows and gilts were double-mated using AI. Pregnant animals were maintained in confinement and fed a daily ration of 2.2 to 2.7 kg of a 16% CP cornsoybean meal diet. Animals for ET were chosen at approximately 6 mo of age, moved to their own gilt pool for further development in close proximity to a boar, and observed for signs of estrus. Each month, 20 gilts of each breed during the first 6 mo of the experiment and 10 of each breed each month thereafter were transferred to a nearby farm for a 21-d period. During this time they were penned individually in gestation crates and fed 2.2 to 2.3 kg of the diet top-dressed for 14 d with 15 mg of [email protected] gilts were then returned to the breeding farm and grouped randomly in pens of five. They were exposed to a mature boar for 2 30 min/d and observed twice daily for estrus. The date of estrus was recorded and gilts were then allowed to return to estrus approximately 21 d later when they were assigned into pairs of one L and one D when estrus occurred within 24 h of one another. One gilt in each pair was inseminated twice with semen from a boar of the opposite

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

Experimental Procedure

breed. In one-half of the L-D pairs the L gilt was mated to a D boar, and in the other half the D gilt was mated to a L boar. One unmated gilt of each pair served as a recipient for embryos from the mated gilt. Embryo transfer was performed following the methods described by Dziuk (1971). The uterus of the donor was surgically exposed and the embryos flushed from each uterine horn via a small glass cannula inserted into the lumen. Embryos recovered from the donor were assessed immediately under a 30x dissecting microscope for signs of cleavage. Embryos were considered to be viable if they were between the four-cell and morula stage with equal cell cleavage and well-defined cell membranes. Embryos were transferred to the uterus of the recipient. Based on a mean count of 14.5 corpora lutea per donor, the ET technique was 7 9 . 1YO successful in collecting these embryos, with a n average of 11.5 viable embryos transferred per complete ET surgery. Of 135 transfers, 49 (36%) produced viable litters with a n average of 6.1 pigs born alive (Table 11. During the first 6 mo of the experiment all donors that had recovered from surgery and were deemed suitable to rebreed were combined with any unused recipients and placed into the next available group of gilts to conserve numbers. In the final eight surgery groups, however, the number of transfers performed was reduced from 20 to 10 per month; thus, no animals were reused. This change in procedure was associated with a n increase in the number of litters farrowed each month (3.0 vs 3.9) and an increase in litter size from 5.8 to 6.2 pigs per litter. The first litters were farrowed in December 1988. A total of 115 F1litters of four different types, 49 by ET, sired by 15 boars were analyzed (Table 2). Embryo transfer litters were all produced by first-litter sows, whereas data from later-parity sows farrowing non-ET litters were included to improve precision. This represents a potential source of bias only if a genotype-treatment x parity interaction existed. Fourteen of the non-ET litters were repeat farrowings by the same dam. All litters were farrowed in confinement and all boar pigs were castrated within 24 h of birth. A 22% CP creep feed was made available beginning a t 3 wk of age and was available ad libitum through 7 wk of age. Litters were weaned at 32 d (SD = 3 dI and pigs penned by size in an environmentally controlled nursery on fully slatted, expanded-metal floors. While pigs were in the nursery, the protein level of the diet was gradually reduced from 22 to 20% CP at 7 to 9 wk to 18% at 9 to 12 wk. The pigs remained in the nursery for approximately 7 wk and then were moved to one of six fully enclosed growing-finishing barns with either fully or partially slatted floors and penned 8

2351

RECIPROCAL CROSS DIFFERENCES IN PIGS

Table 1. Embryo transfer results No. attempted* Date

DCtoLd

LtoD

No. of transfersb

Litters farrowed

DtoL

DtoL

LtoD

LtoD

~

8 7 10 9 9 8 3 5 5 5 4 4 3 6

9 10 10 9 7 9 3 5 5 9 4 5 2 1

83

88

7

6

6

8 9 8 5 3 4 5 4 1 4 3 5

2 7 5 5 7 2 5 5 7 4 4 2 1

73

62

171

135

2 2 2 3 2 2 0 2 3 3 1 3 0 4

~

0 0 1 1 0 3 2 4 4

14 13 18 29 6 24 4 37

1 3 1 0 0

24 36 27 0 31

29 20 49 (30%)e

34

297

G.l/litter

&Number of embryo-harvesting surgeries. bNumber of embryo transfer surgeries completed. CD = Duroc. d~ = Landrace. ePercentage of transfers.

to 10 pigs per pen. Growth rate was determined in all barrows and gilts and the pigs remained here from 12 wk of age until the day before slaughter. A 16% CP corn-soybean meal diet and water were freely available throughout the growing-finishing period. Experimental pigs weighed a mean of 38.4 kg, SD = 7.4 kg, after a 7-d adjustment period in the growingfinishing house and weighed a mean of 103.5 kg, SD = 8.2 kg, at the end of the growth test. The age of pigs and length of test were recorded when the pigs were weighed. Two hundred twentyfour pigs from ET and 191 non-ET contemporaries were slaughtered for carcass evaluation (58% of the pigs, Table 3). Pigs were considered to be contemporaries when sired by the same boars and born in the same month. Approximately two gilts were slaughtered from each ET litter and one gilt from each non-ET litter.

Carcass measurements were made on 73% of the barrows, balanced as far as possible for ET and non-ET litters within each genotype. Gilts and barrows slaughtered for carcass data collection were randomly selected from available animals on a within-litter basis. Pigs were slaughtered in either the University of Illinois Meat Science Laboratory, Emge Packing in Anderson, IN, or Behrman’s packing plant in Germantown, IL. Live weight was recorded before leaving the farm the day before slaughter (mean = 111.2 kg, SD = 8.6 kg) and hot carcass weight was taken immediately after slaughter. Carcass traits were measured 24 h postmortem on the right side of each carcass. Traits analyzed were carcass length, average backfat thickness at the first rib, last rib, and last lumbar vertebra, backfat thickness at the 10th rib, and longissimus

Table 2 . Number of litters analyzed and litter size born alive by treatment group Litter size born alive

No. of litters by panty Genotypea

Treatmentb

DxL LxD DxL LxD Total

non-ET non-ET ET ET

20 20

-

-

-

-

-

-

1 7 19

2

3

2 4

5 12

5 6

2 10 -

-

&D = Duroc, L = Landrace; breed of sire x breed of dam. bET = embryo transfer.

Total 19 47 20 29

115

Mean

SD

8.3 7.8 6.0

3.5 2.7 2.5 2.8

6.1

7.2 pigdlitter -

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

September 1988 October 1988 November 1988 December 1988 January 1989 February 1989 April 1989 May 1989 June 1989 July 1989 September 1989 October 1989 November 1980 December 1980 Total

pigs born alive

2352

WILKEN ET AL. Table 3. Number of pigs for growth and carcass traits Growth rate Treatmentb

D x L L x D D x L LxD Tota1

non-ET nonET ET ET

*D bET

-

Barrows

Gilts

65 158 40

74 166 50

98

78 349

71 361

Total

Barrows

Gilts

Total

10

149

47 91 46 72

35 45 61

65 126 91 133

710

256

159

415

139 324

= Duroc, L = Landrace: breed of sire x breed of dam. =

embryo transfer.

muscle area at the loth rib. Fat standardized lean and lean gain per day of age were estimated as functions of hot carcass weight, loth rib backfat thickness, longissimus muscle area, and age using equations published by Grisdale et al. (19841. Statistical Analysis. The following linear model was assumed in analyzing average daily gain:

where Yijklmno = a n observation; Bi = fixed effect of the ith finishing barn (i = 1, . . ., 6); Pj = fixed effect of the jth parity [j = 1, . . ., 41, where j = 4 represented parities 2 4; Fk = fixed effect of the kth farrowing (k = 1, . . ., 13); T1 = fixed effect of the lth genotype-treatment combination (1 = D-L reciprocal cross - EThon-ET class = 1, . , ., 4); S m = fixed effect of the mth sex (gilt, barrow); (BSlim = fixed effect of the interaction between the ith barn and the mth sex; (FSIkm = fixed effect of the interaction between the kth farrowing season and the mth sex; P1 = partial linear regression of ADG on litter size born alive, fijklmno; P2 = partial linear regression of ADG on on-test weight, Xijklmno; p3 = partial linear regression of ADG on off-test weight, Zijklmno; ljkln = random effect Of the nth litter of rearing, the ljkln - iid N (0,$1; aijklmno = random effect of the oth animal, the aijklmno; - N (0,Ma) and uncorrelated with random litter effects, where A = Wright’s numerator relationship matrix; and eijklmno = random residual, eijklmno - iid N (0, and uncorrelated with the random additive and litter effects. Because litter size in which a pig is reared may exert a negative maternal effect on growth traits (Haley, 19891 it was included in statistical models as a covariable. Regressions calculated separately for ET and non-ET classes in preliminary analyses were not detected as being different; hence, a common regression was assumed for both groups. On-test weight was included as a covariable in the

4)

model for ADG because the objective was to compare growth rate for a fixed weight interval &e., between 38 and 104 kg). All pigs did not go ontest a t the same weight, or come off-test a t the same weight; hence, both on- and off-test weights were included as covariables. Of the 10 possible two-factor interactions, six were not considered due to missing cells. Interactions between sex and both genotype-treatment combination and parity were not detected (P > .2O) in a preliminary analysis that assumed the above model without animal or litter effects (SAS, 1985). These interactions were therefore not included in the final model for ADG. Similarly, fixed-model analyses were conducted to determine two-factor interactions to be included in statistical models assumed for other traits. The linear model assumed in analyzing age at 103.5 kg BW was as above, excluding the regression on on-test weight. Preliminary fixed-model analysis detected a genotype-treatment x sex interaction (P c .05)for carcass length, 10th rib backfat thickness, longissimus muscle area, estimated fat-standardized lean, and estimated lean per day of age. The linear model assumed in analyzing these traits was as follows:

where P, Bi, pip Fk, T1, s m , (BS)im, ljkh aijklmnol and eijklmno are as defined above and (TSl1m = fixed effect of the interaction between the lth genotype-treatment combination and the mth sex and p4 = partial linear regression of the dependent variable on off-farm weight, Wijklmno. Experimental hypotheses of interest relating to the presence of reciprocal cross differences, prefertilization and(or1 fertilization effects, and postfertilization effects are given in Table 4. To test for these effects it was necessary to examine all possible differences among the four genotypetreatment means (D x L ET, D x L non-ET, L x D ET, L x D non-ET). In addition, for the traits analyzed

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

Genotype&

Carcass traits

RECIPROCAL CROSS DIFFERENCES IN PIGS

2353

Table 4. Coefficients for hypotheses tested Genotypeltreatment GLSMa Treatmentb

Reciprocal cross Reciprocal cross (he-)fertilization [he-)fertilization Postfertilization Postfertilization

No ET ET D uterus L uterus D x L embryo L x D embrvo

DXL-L

LxD-D

DxL-D'

-1 0 0

1 0 1 0 0

0 -1

0 1

-1

0 1 0

-1

-1 0

LxD-L'

0 1 0

1

-1

&L = Landrace, D = Duroc, breed of sire x breed of dam - maternal environment; and GLSM generalized least squares mean. bET = embryo transfer. CET.

assuming Model [21, genotype-treatment means and linear contrasts were examined separately within sex. Variance components were estimated by REML from a dataset consisting of growth data on 5,649 barrows and gilts and carcass data on 960 barrows (Lo et al., 1992). Pigs were purebred D and L and reciprocal F1 crosses. Best linear unbiased estimates (BLUE) of estimable functions of fixed effects were obtained from Henderson's (1973) mixed-model equations (MME) assuming REML estimates of variance components to be the true variances. Tests of hypotheses involving fixed effects were carried out using the t-test.

Results Covan'ables. For every one pig increase in birth litter size there was a n associated increase of .36d in age to 103.5 kg BW (P < .05), a decrease of .57 mm in loth rib backfat and .49 mm in average backfat thickness (P c .Ol), and a .33 cm2 increase in longissimus muscle area (P < .OS). Lo et al. [1992) reported the regression of age at 103.5 kg on

=

number of live pigs 24 h postfarrowing to be .49 d per pig, but found no effect on carcass traits. Smaller litters associated with the ET procedure increased variation in litter size in the present study, which may explain the effects detected. Both on-test and off-test weights affected growth rate and regressions of carcass traits on off-farm weight were detected (P < .011. Reciprocal Cross Differences. Generalized least squares means for off-test age, ADG, hot carcass weight, and average backfat thickness in the four genotype-embryo transfer treatment classes are presented in Table 5. Genotype-treatment generalized least squares means are presented by sex in Tables 6 (barrows) and 7 (gilts) for traits for which a genotype-treatment class x sex interaction was detected. Reciprocal cross differences, tested separately for both ET and nonET pigs, were not detected for growth rate, carcass weight, or average backfat thickness (Table 8). These traits exhibited no genotype-treatment x sex interaction (P > .2O) and differences were therefore averaged over sex. Genotype-treatment x sex interactions were detected for carcass length, loth rib backfat thickness, longissimus muscle area, estimated fat

Table 5 . Genotype - treatment least squares means f standard errors for growth and carcass traits Genotype Traita AGE ADG HCWT AVBF

D x L

-L

155.77 .87 81.22 31.85

(non-EmC f 1.18 f .03 f .83 f .85

-

embryo transfer (ET) treatment combinationb

L x D

-

D (non-ETIC D x L

155.30 .89 79.98 32.09

f f f f

.88 .02

.66 .71

-

159.50 .88 81.01 30.18

D (Emd k 1.57 k .04 f

.86

k 1.01

-

L (Ernd

156.93 .88 80.68 30.46

f 1.33 f .03

L x D

f f

.80 .97

&AGE = days to 103.5 kg BW; ADG = average daily gain from 38.4 kg to 103.5 kg BW in kilograms; HCWT = hot carcass weight in kilograms; and AVBF = average of first rib, last rib, and last lumbar vertebra backfat thicknesses, in millimeters. bD = Duroc and L = Landrace, breed of s u e x breed of dam - uterine and postpartum environment. 'Non-embryo transfer. dEmbryo transfer.

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

Effect

WILKEN ET AL.

2354

Table 6 . Genotype - treatment least squares means k standard errors for carcass traits and lean growth rate in barrows

-

Genotype Traita

D x L . L (non-ETIC L x D 80.32 27.25 34.95 43.41 .29

f

D (nonETfC

~

80.30 33.04 31.07 40.64 .27

.43

f 1.22 f .84 f 1.42 f .12

f .35 f .99 f .69 f 1.18 f .09

D x L . D (Emd 80.32 28.99 37.25 43.44 .28

f 51 f 1.44 f 1.01 f 1.73 f .13

L x D . L (ET)d 81.30 25.59 33.41 42.18 .27

f .49 f 1.38 f .99 f 1.70 f .12

&LENG = carcass length, in centimeters; BFTR = backfat thickness at the loth rib, in millimeters; LMA = longissimus muscle area at the loth rib, in square centimeters; FSL = estimated fat standardized lean (adjusted to a 10% lipid content); and LDOA = estimated fat standardized lean per day of age, in kilograms/day. bD = Duroc and L = Landrace, breed of sire x breed of dam - uterine and postpartum environment. CNon-embryo transfer. dEmbryo transfer.

standardized lean, and estimated lean gain per day of age and means were therefore contrasted on a within-sex basis (Tables 9 and 101. Non-ET barrows from D sires and L dams had 3.9 cm2 larger longissimus muscle area and 5.8 mm less backfat at the 10th rib (P < .001), yielding 2.8 kg more estimated fat-standardized lean (P < ,101, than did barrows produced by L sires and D dams (Table 9). Barrows from ET sired by D boars also had 3.8 cm2 larger longissimus muscle area (P < .001) than did barrows from ET sired by L boars. There were no differences detected for these barrows in backfat thickness at the 10th rib, however (P > .lo). Reciprocal cross differences in carcass length were small and of marginal statistical significance. Differences were 0 cm (non-ET1and 1.0 cm (ET) for the barrows and 1.3 cm (non-ET)and .8 cm (ET) for the gilts (Table 10). It is important to note

in interpreting levels of significance in Tables 8 to 10 that probabilities are given on a per-comparison basis. With four treatments and six contrasts, each evaluated at o! = .05,the probability of at least one erroneous inference is .26. Reciprocal cross differences for carcass length were not, therefore, judged to be evident in this study. Similarly, convincing differences were not detected for growth rate (Table 81, fat-standardized lean, or lean per day of age (Tables 9 and 10). The only reciprocal cross differences detected in this study were for backfat and longissimus muscle area measured a t the 10th rib in barrow carcasses. Other researchers have also reported that, in reciprocal crosses involving the D breed, carcass characteristics of the offspring were superior when D is the sire and not the dam. Young et al. (1976) found D x Yorkshire (Y) F1 barrows to have 2.2 mm less backfat and 2.2 cm2 larger

Table 7. Genotype - treatment least squares means f standard errors for carcass traits and lean growth rate in gilts Genotype Traita LENG BFTR LMA FSL LDOA

Dx L

. L (nonETIC

80.97 25.13 35.52 44.18 .28

f .53 f 1.51

f 1.01 1.72 f .15 It

- embryo

L x D

-

transfer [ETI treatment combinationb

D (nonETIC

82.23 25.26 36.02 43.49 .28

f .40 f 1.14 f .78 f 1.33 f .ll

Dx L

-

81.25 25.71 38.17 43.64 .27

D (Emd f 52 f 1.48 f 1.04 f 1.77 f .01

L x D 82.04 23.87 38.12 44.09 .28

-

L f f

f f f

[ETP .49 .38 .99 1.69 .12

&LENG = carcass length, in centimeters; BFTR = backfat thickness a t the 10th rib, in millimeters; LMA = longissimus muscle area at the 10th rib, in square centimeters; FSL = estimated fat standardized lean (adjusted to a 10% lipid content); and LDOA = estimated fat standardized lean per day of age, in kilograms/day. bD = Duroc and L = Landrace, breed of sire x breed of dam - uterine and postpartum environment. CNon-embryo transfer. dEmbryo transfer.

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

LENG BFTR LMA FSL LDOA

embryo transfer (ET) treatment combinationb

2355

RECIPROCAL CROSS DIFFERENCES IN PIGS

Table 8. Differences among least squares means f standard errors for traits for which no treatment x sex interaction was detected Difference* Reciprocal cross Traitb -.47 .02 -1.24 .24

Fertilization effect 2

f 1.19 f .03 f .81 f .91

-2.56 .01 -33 .28

SE

3

f 1.46 f .04 f .70 f .82

Postfertilization effect 4

-4.20** .01 -1.03 1.91

f 1.57 f .03 f .91 f 1.11

1.18 .01 -.54 -1.39

8

5

1.83 f .04 f .97 f 1.16 f

3.73* f .01 It -.21 f -1.87 f

1.83 .04 .99 1.11

-1.63 .01 -.70 1.63

f 1.32 f .03 f .81 f 1.09

&Difference1 = mean (L x D - D1 - mean (D x L - L) testing for a reciprocal cross difference, no ET, where L = Landrace, D = Duroc, breed of sire x breed of dam - maternal environment, ET = embryo transfer; Difference 2 = mean (L x D . L) - mean (D x L - D1 testing for a reciprocal cross difference, ET; Difference 3 = mean (L x D - D1- mean (D x L . D) testing for fertilization effect, D uterus; Difference 4 = mean (L x D - L1- mean (D x L - L 1 testing for fertilization effect, L uterus; Difference 5 = mean (D x L . D1- mean (D x L - L) testing for postfertilization effect, D x L embryo; and Difference 6 = mean (L x D - D) - mean (L x D L1 testing for postfertilization effect, L x D embryo. bAGE = days to 103.5 kg BW, d; ADG = average daily gain from 38.4 to 103.5 kg BW, in kilograms; HCWT = hot carcass weight, in kilograms; and AVBF = average of first rib, last rib, and last lumbar vertebra back fat depths, in millimeters. ~

*P < .05. *+P < .01.

longissimus muscle area than Y x D F1 barrows. McLaren et al. (19871found that D x L barrows had 3.9 mm less backfat and 3.4 cm2 larger longissimus muscle area than their reciprocal cross counterparts. Bereskin and Davey (1978) studied 48 D-Y reciprocal cross barrows and 48 gilts and found the F1 from Y dams to have 2.8 cm2 larger longissimus muscle area and 2.2 mm less average backfat than F1 barrows and gilts from D dams. Analyses of variance detected reciprocal cross difference x sex interactions for lean cuts gain (P < .01) but not for longissimus muscle area or backfat (P > .lo). Barrows sired by D boars had 18 g/d greater lean cuts growth rate than did barrows sired by Y boars. Reciprocal cross gilts

did not differ for lean cuts gain. Longissimus muscle area and backfat means by genotype and sex were not reported. Lo et al. (1992) analyzed carcass data on 496 F1 D-L barrows from the same population used in the present study, including 131 of the 138 non-ET barrows reported here. They found that D-sired F1 barrows had 5.9 cm2 larger longissimus muscle area and 6.3 mm less backfat at the 10th rib than did L-sired F1 barrows. Reciprocal cross differences of 3.9 cm2 for 10th rib longissimus muscle area (ET and non-ET barrows) and 5.8 mm for 10th rib backfat thickness (non-ET barrows only) in the present study are, therefore, consistent both with published results from other studies and with previous estimates obtained from a larger sample

Table 9. Differences among least squares means f standard errors in barrows Differencea f SE Reciprocal cross Traitb LENG BFTR LMA FSL LDOA

1

Fertilization effect

2

3

-.02 f .44 .os* f .44 5.80*** f 1.28 .60 f 1.28 -3.88*** f .91 -3.84*** f .87 -2.77+ f 1.55 -1.26 f 1.48 -.02

f

.ll

.oo

f

.12

f .56 .98+ -.03 f 1.59 2.34 4.08* -8.18++* f 1.14 -1.54 f 1.95 -1.22 -2.80 f .14 -.01 -.01

Postfertilization effect

4

.59 1.68 f 1.19 f 2.03 f .15

f f

5

6

.57 f 1.82 2.30* f 1.13 .04 f 1.94 -.01 f .15

-l.OOt f .55 3.45* k 1.54

.07 1.74

It

-2.34* It 1.14 -1.54 f 1.95 -.01 It .13

&Difference1 = mean (L x D . D) - mean (D x L . L) testing for a reciprocal cross difference, no ET, where L = Landrace, D = Duroc, breed of sire x breed of dam - maternal environment, ET = embryo transfer; Difference 2 = mean (L x D - L1- mean (D x L - Dl testing for a reciprocal cross difference, ET; Difference 3 = mean (L x D - D) - mean (D x L . D) testing for fertilization effect, D uterus; Difference 4 = mean (L x D - L) - mean (D x L - L 1 testing for fertilization effect, L uterus; Difference 5 = mean (D x L . D) - mean (D x L - L1 testing for postfertilization effect, D x L embryo; and Difference 8 = mean (L x D D) - mean (L x D - L) testing for postfertilization effect, L x D embryo. bLENG = carcass length, in centimeters; BFTR = back fat depth at the 10th rib, in millimeters; LMA = longissimus muscle area at the loth rib, in square centimeters; FSL = estimated fat standardized lean (adjusted to a 10% lipid content); and LDOA = estimated fat standardized lean per day of age, in kilograms/day. ~

+P < . l o . *P e .OS. ***P < ,001.

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

AGE ADG HCWT AVBF

1

~t

2356

WILKEN ET AL.

Table 10. Differences among least squares means -+ standard errors in gilts Difference* f SE Reciprocal cross Traitb 1.26" .14 .50 -.67 -.OO

.60 f 1.72 f 1.16 f 1.97 f .17 5

4

2

3

.79' f .47 -1.84 f 1.33 -.05 f .92 .44 f 1.56 .01 f .13

.Q8 f .62 -.45 f 1.76 -2.14+ f 1.23 -.16 f 2.11 .01 f .l6

1.08 -1.28 2.59* -.07

-.oo

.67 f 1.91 f 1.76 f 2.26 f .18

f

Postfertilization effect 5

6

.28 f .66 .59 5 1.87 2.64* f 1.28 -.51 f 2.18 -.01 f .18

.19 5 B O 1.39 f 1.71 -2.10t f 1.23 -.60 f 2.10 -.OO f .15

&Difference1 = mean (L x D - D) - mean (D x L . L) testing for a reciprocal cross difference, no ET, where L = Landrace, D = Duroc, breed of sire x breed of dam. maternal environment, ET = embryo transfer; Difference 2 = mean (L x D - L) - mean (D x L - D) testing for a reciprocal cross difference, ET; Difference 3 = mean (L x D - D) -mean (D x L - D) testing for fertilization effect, D uterus; Difference 4 = mean (L x D - L) - mean (D x L - L 1 testing for fertilization effect, L uterus; Difference 5 = mean (D x L - D) - mean (D x L - L) testing for postfertilization effect, D x L embryo; and Difference 6 = mean (L x D - D) - mean (L x D - L) testing for postfertilization effect, L x D embryo. bLENG = carcass length, in centimeters; BFTR = back fat depth at the loth rib, in millimeters; LMA = longissimus muscle area at the loth rib, in square centimeters; FSL = estimated fat standardized lean (adjusted to a 10% lipid content); and LDOA = estimated fat standardized lean per day of age, in kilograms/day.

+P < .lo. *P < .OS.

of barrows produced by crossing the lines used in the present experiment. Prefertilization and(or) Fertilization Effects. Barrows sired by D boars reared in a D postfertilization environment (ET) had 4.1 nun less 10th rib backfat and 6.2 om2 greater longissimus muscle area (P < .05) than barrows sired by L boars gestated and reared by D dams (non-ET)(Table 91. Barrows sired by D boars reared by L dams (nonET) had 2.3 mm less backfat and 1.5 cm2 greater longissimus muscle area (P > ,101 than barrows sired by L boars reared by L dams (ET). Only one (pre-)fertilization effect difference was significant for gilts (Table 101, and the estimate for the alternative postfertilization environment was of similar magnitude but the opposite sign. PostfertiZization Effects. Postfertilization effects were similar for barrows and gilts (Tables 9 and 101, but three of four contrasts were significant (P c .05)for the barrows, compared with one of four for the gilts. Although these results must be considered with caution (probabilities are biased downward, as discussed above], transferred embryos of the same genotype seemed to have greater longissimus muscle area than nontransferred embryos. This might reflect failure of the statistical model to adequately account for litter size differences in ET and non-ET groups or, possibly, selection pressure on embryos before implantation resulting from the ET technique might have indirectly selected for growth factors. Results for thickness of backfat a t the 10th rib indicate that barrows with L postfertilization environments were leaner than those with D postfertilization environments (Table 9).Postfertilization effects were generally less than (pre-Ifertilization effects in contributing to reciprocal cross differences.

Discussion Reciprocal cross differences result from events occurring before or a t fertilization or in the postfertilization environment provided by the dam. Results presented confirm previous findings that reciprocal cross differences exist for backfat thickness and longissimus muscle area for D - L F1 pigs. Further, these differences seem to be associated with events occurring before or a t fertilization. Maternal effects might be cytoplasmically inherited via the ovum in mammals (Brumby, 19601. Mitochondrial DNA (mtDNA), which codes for structural proteins of the electron transport chain, is maternally inherited via the cytoplasm in mammals (Hutchinson et al., 1974) and has, therefore, been suggested as one possible cause of (prelfertilization effects. Restriction fragment length polymorphism analysis of mtDNA obtained from both D and L pigs representing six (two D and four L) independent maternal lineages represented in the lines used in this study, however, demonstrated the mtDNA sequences of the two breeds to be at least 99.1% conserved (Wilken et al., unpublished data). Additionally, Lo et al. (1992)reported means for carcass traits obtained from 464 purebred D and L and 496 F1 D x L and L x D barrows from the same population used in the present study. Pigs sired by D boars had 4.4 cm2 larger longissimus muscle area and 7.3 mm less loth rib carcass backfat than did pigs sired by L boars (P < .Oil Crossbred barrows with D sires had 5.9 cm2 larger longissimus muscle area and 6.3 mm less loth rib backfat than did F1 pigs with L sires (direct genetic effects, P c .001). Reciprocal cross barrows from D sires in the present study had 3.9 cm2 larger

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

LENG BFTR LMA FSL LDOA

1

Fertilization effect

2357

RECIPROCAL CROSS DIFFERENCES IN PIGS X

XDYD

L

?

XLXL

Ldl XLYL

X

D

?

XDXD

Figure 1. Sex linkage under the experimental design of this study. D = Duroc, L = Landrace.

affecting carcass composition. For a paternal effect to exist under sex linkage the mechanism of control would almost certainly have to be Y-linked. Little information is available concerning genes on the Y chromosome of mammals. In humans, however, at least three Y-linked genes have been mapped, a n immune system protein gene (H-Y),the ZFY (zinc-finger protein) gene, and the sex-determining region Y (SDY) gene. Might a Y-linked gene exist that prevents demethylation of autosomal maternal genes that influence carcass composition in D pigs? Reciprocal cross differences for longissimus muscle and backfat thickness do exist for D-L barrows, are comparable in magnitude to direct breed effects, seem to be largely established either before or at fertilization, and seem to be Y-linked. Before definite conclusions can be made concerning the mechanism(s1 controlling these differences, more research is needed in the area of molecular genetics to better understand the mechanisms governing gene control and expression. Mapping the porcine genome using genetic markers (Haley et al., 1990; Schook et al., 1990, 19911 will provide a “genetic road map” with which to detect genes affecting quantitative traits.

Implications The objective of this study was to partition reciprocal cross differences for growth and carcass traits in F1 Duroc-Landrace pigs into effects occurring at or before fertilization vs postfertilization effects. Reciprocal cross differences were detected for backfat thickness and longissimus muscle area at the loth rib in barrows; barrows sired by Duroc boars had less backfat and larger longissimus muscle area than barrows sired by Landrace boars. Results of embryo transfer indicate that these differences were established either before or at fertilization, and detection in barrows only indicates that they are Y-linked. More research is needed to determine the role the sire and dam play in determining body composition of their offspring.

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

longissimus muscle areas and 5.8 mm less 10th rib backfat than did barrows sired by L boars, indicating a paternal effect in that crossbreds were more like their breed of sire than their breed of dam. Cytoplasmic inheritance seems highly unlikely, therefore, to be contributing to the observed effects. Two additional ways in which parental contributions to the embryo might differ are genomic imprinting and sex linkage. Genomic imprinting is a mechanism whereby gene expression is preferentially “switched” on or off depending on whether the gene’s gamete of origin was the sperm or the egg. Evidence in mice indicates that autosomal genes may demonstrate different patterns of expression depending on their methylation status (Reik et al., 1987; Sapienza et al., 1987). Differential methylation of DNA has, therefore, been suggested as a possible mechanism of genomic imprinting in mammals. Swain et al. (1987) studied the methylation pattern of a n autosomal transgene in mice. Results demonstrated that when the transgene was inherited from the male it was expressed in the heart and no other issue. When the transgene was inherited from the female, however, it was not expressed. This pattern of expression was shown to correlate precisely with the parentally imprinted methylation state evident in all tissues of the progeny. It was therefore suggested that methylation of the transgene was acquired by its passage through the female parent and eliminated during gametogenesis in the male. These observations provided direct molecular evidence that autosomal gene expression can be influenced by the sex of the parent from which the gene is inherited. Even though genomic imprinting represents a possible mechanism controlling maternal or paternal effects in mammals, it is not consistent with the results of this study. Under genomic imprinting, reciprocal cross offspring might perform differently depending on the breed of their sire or dam. However, one would not expect this difference to depend on the sex of the offspring. Reciprocal cross differences for longissimus muscle area and loth rib backfat thickness were evident only in the barrows, suggesting sex linkage. For a n X-linked genekl, one would expect barrows to be similar to their dam (Figure 1). Gilts, however, would be expected to be similar either to the sire or to the dam, depending on random barr body formation. In the case of a Y-linked genek) barrows would be expected to be more similar to the sire breed and reciprocal cross gilts would not be expected to differ in performance. Our results are therefore consistent with Y-linked inheritance

2358

WILKEN ET AL.

Literature Cited

100:180.

Haley, C. S . , A. Archibald, L. Anderson, A. A. Bosma, W. Davies, M. Fredholm, H. Geldermann, M. Groenen, I. Gustavsson, L. Ollivier, E. M. Tucker, and A. Van de Weghe. 1990. the pig gene mapping project - PiGMaP. Proc. 4th World Congr. Genet. Appl. Livest. Prod. XII:67. Henderson, C. R. 1973. Sire evaluation and genetic trends. In: Proc. Anim. Breed Genet. Symp. in Honor of Dr. Lush. ASAS and ADSA, Champaign, IL. Hohenboken, W. D. 1985. Maternal effects In: A. B. Chapman (Ed). General and Quantitative Genetics. pp 135-146. Elsevier Science Publishing, New York. Hutchinson, C. A,, 111, J. E. Newbold, S . S . Potter, and M. H. Edgell. 1974. Maternal inheritance of mammalian mito-

248.

Sapienza, C.. A. C. Peterson, J. Rossant, and R. Balling. 1987. Degree of methylation of transgenes is dependent on gamete of origin. Nature (Lond.) 328:251. SAS. 1985. SAS User’s Guide: Statistics. SAS Inst. Inc., Cary, NC. Schneider, J.F., L. L. Christian, and D. L. Kuhlers. 1982.Crossbreeding in swine: Genetic effects on pig growth and carcass merit. J. h i m . Sci. 54:747. Schook, L. B., J. E. Beever, P. A. Clamp, H. A. Lewin, and D. G. McLaren. 1990. Status of the pig gene map. In: J. E. Womack (Ed.) Mapping the Genomes of Agriculturally Important Animals. Published for Texas A&M Univ. by Cold Spring Harbor Laboratory Press, NY. Schook, L. B., H. A. Lewin, and D. G. McLaren. (Ed.) 1991.Gene Mapping Techniques and Applications. Marcel Dekker, New York. Swain, J., T. A. Stewart, and P. Leder. 1987. Parental legacy determines methylation and expression of a n autosomal transgene: A molecular mechanism for parental imprinting. Cell 50:719. Willham, R. L. 1972. The role of maternal effects in animal breeding: 111. Biometrical aspects of maternal effects in animals. J. Anim. Sci. 35:1288. Young, L. D., R. K Johnson, I. T. Omtvedt, and L. E. Walters. 1976. Postweaning performance and carcass merit of purebred and two-breed cross pigs. J. Anim. Sci. 42:1124.

Downloaded from https://academic.oup.com/jas/article-abstract/70/8/2349/4632020 by Iowa State University user on 03 February 2019

Ahlschwede, W. T., and 0. W. Robison. 1971. Prenatal and postnatal influences on growth and backfat in swine. J. h i m . sci. 3210. Bereskin, B. 1983. Performance of selected and control lines of Duroc and Yorkshire pigs and their reciprocal crossbred progeny. J. Anim. Sci. 57:867. Bereskin, B., and R. J. Davey. 1978. Genetic, sex and diet effects on pig carcass traits. J. Anim. Sci. 40:1581. Bereskin, B., C. E. Shelby, and L. N. Hazel. 1971.Carcass traits of purebred Durocs and Yorkshires and their crosses. J. h i m . sci. 32:413. Brumby, P. J. 1900.The influence of the maternal environment on growth in mice. Heredity 14:l. Dziuk, P. J. 1971.Obtaining eggs and embryos from sheep and pigs. In: J. C. Daniel, Jr. (Ed.) Methods in Mammalian Embryology. pp 76-85. W. H. Freeman and Co., San Francisco, CA. Grisdale, B., L. L. Christian, H. R. Cross, D. J. Meisinger, M. F. Rothschild, and R. G. Kauffman. 1984.Revised approaches to estimate lean of pork carcasses of known age or days on test. J. Anim. Sci. 58:335. Haley, C. S. 1989. Maternal effects on performance traits which are mediated via litter size. J. Anim. Breed. GBnBt.

chondrial DNA. Nature (Lond.) 25 1:536. Lo, L. L., D. G. McLaren, F. K. McKeith, R. L. Fernando, and J. Novakofski. 1992. Genetic analyses of growth, real-time ultrasound, carcass, and pork quality traits in Duroc and Landrace pigs: I. Breed Effects. J. h i m . Sci. 70:2373. McLaren, D. G., D. S . Buchanan, and R. K. Johnson. 1987. Individual heterosis and breed effects for postweaning performance and carcass traits in four breeds of swine. J. h i m . Sci. 04:83. Reik, W., A. Collick, M. L. Norris, S . C. Barton, and M. A. Surani. 1987. Genomic imprinting determines methylation of parental alleles in transgenic mice. Nature (Lond.) 328: