Quantitative trait loci segregating in crosses between New Hampshire ...

doi: 10.1111/j.1365-2052.2012.02344.x

Quantitative trait loci segregating in crosses between New Hampshire and White Leghorn chicken lines: II. Muscle weight and carcass composition M. K. Nassar, Z. S. Goraga and G. A. Brockmann Breeding Biology and Molecular Genetics, Department of Crop and Animal Sciences, Humboldt-Universita¨t zu Berlin, Invalidenstrabe 42, D-10115, Berlin, Germany

Summary

In order to identify genetic factors influencing muscle weight and carcass composition in chicken, a linkage analysis was performed with 278 F2 males of reciprocal crosses between the extremely different inbred lines New Hampshire (NHI) and White Leghorn (WL77). The NHI line had been selected for high meat yield and the WL77 for low egg weight before inbreeding. Highly significant quantitative trait loci (QTL) controlling body weight and the weights of carcass, breast muscle, drumsticks–thighs and wings were identified on GGA4 between 151.5 and 160.5 cM and on GGA27 between 4 and 52 cM. These genomic regions explained 13.7–40.2% and 5.3–13.8% of the phenotypic F2 variances of the corresponding traits respectively. Additional genome-wide highly significant QTL for the weight of drumsticks–thighs were mapped on GGA1, 5 and 7. Moreover, significant QTL controlling body weight were found on GGA2 and 11. The data obtained in this study can be used for increasing the mapping resolution and subsequent gene targeting on GGA4 and 27 by combining data with other crosses where the same QTL were found. Keywords inbred chicken line, intercross, linkage, muscle, selection.

Introduction Carcass composition and in particular muscle yield and quality are important traits that influence the nutritional and economic value of a chicken. An increased understanding of the molecular basis behind complex traits like muscle development and carcass composition, which cannot be measured directly on live chickens, helps researchers to identify genes and gene variants contributing to the trait variation and supports breeders in using this information in breeding programmes. As a first step in gene discovery, often a linkage or association study is performed to map genomic loci contributing to the trait of interest.

Address for correspondence G. A. Brockmann, Breeding Biology and Molecular Genetics, Department of Crop and Animal Sciences, Faculty of Agriculture and Horticulture, Humboldt-Universita¨t zu Berlin, Invalidenstraße 42, D-10115 Berlin, Germany. E-mail: [email protected] Accepted for publication 27 November 2011

Genomic regions influencing carcass parts, such as weights of breast muscle, wings, drumsticks and thighs, have been mapped in different crosses between or within diverse chicken breeds (chickenQTLdb; http://www.animal genome.org/cgi-bin/QTLdb/GG/index). Because different chicken strains or lines can harbour different variants of genes that control carcass-related traits, additional crosses between chickens that differ extremely in phenotype can add new information about genetic determinants of these traits. If inbred lines of diverse selection lines are used for gene-mapping experiments, there is a high probability that alleles contributing to the trait variation segregate in the mapping population and thus can be identified. This study aimed at the identification of additional genetic variation contributing to differences in muscle yield and carcass composition in chickens. Our study takes advantage of phenotypically extreme lines with a high degree of inbreeding. We used the dual purpose line New Hampshire (NHI), which had been selected for high meat yield at the age of 20 weeks, and the layer line White Leghorn (WL77), which had been selected for low egg weight during the laying period (Goraga et al. 2010). NHI chickens had about twice as heavy breast muscles and

© 2012 The Authors, Animal Genetics © 2012 Stichting International Foundation for Animal Genetics

1

2

Nassar, Goraga and Brockmann 2.4 times heavier drumsticks–thighs than WL77 chickens did. In reciprocal F2 intercrosses between NHI and WL77, we performed a linkage analysis to detect genomic loci responsible for the big differences between the two chicken lines. Recently, the same population has been used to identify quantitative trait loci (QTL) for egg-laying performance (Goraga et al. 2011). Here, we present QTL for body weight and carcass composition in 20-week-old males. We identified QTL for muscle weight and different carcass parts that have not been reported before, and we could confirm several QTL from other mapping studies. Therefore, the NHI 9 WL77 chicken crosses provide another resource for fine mapping and subsequent gene discovery of economically importance genes, for example in combined analyses of several crosses, where the QTL segregate.

Materials and methods Chicken lines and mapping populations The chicken lines NHI and WL77, as well as the population and pedigree structure for QTL mapping used in this study, have been described in detail previously (Goraga et al. 2010, 2011). NHI and WL77 lines were inbred after selection for either high meat yield at 20 weeks of age or low egg weight respectively. The inbreeding levels were about 86% and 100% in NHI and WL77 respectively. Detailed information on the population and pedigree structure is given in Table S1. In this study, we analysed all 278 F2 males.

Phenotypes We dissected the animals at 20 weeks of age, which was also the age of selection for high meat yield in the NHI line. Five to six males of the parental lines, 25 F1 and all 278 F2 males were weighed, slaughtered and dissected after a fasting period of 10 h. Feathers, head, feet, shanks, all inner organs and visceral white adipose tissues were removed. The carcass was defined as the remaining empty body with bones and muscles covered with skin. Afterwards, the skin and neck were removed. Left and right breast muscles were separated from the bones and weighed; the sum was named breast muscle weight. The summed weights of the left and right wings (wings weight) and the sum of the weights of drumsticks and thighs (drumsticks–thighs weight) were recorded. After removing all mentioned parts, the residual carcass with breast bones and the dorsal part was weighed. Relative weights were calculated as a percentage of carcass weight.

Statistics All analyses were performed using SAS (SAS Institute 2008). To test significant differences either between NHI, WL77

and F1 populations or between the reciprocal crosses, an analysis of variance was preformed. Pairwise comparisons were made using the Tukey–Kramer test. Pearson’s correlation coefficients between traits were estimated in the F2 populations. Heritabilities were estimated in the F2 generation by the analysis of sire and dam variance components (PROC VARCOMP) using the restricted maximum-likelihood (REML) method. Genotype effect plots were drawn with least square means (LSM). A Tukey–Kramer test was performed to test phenotypic differences between genotype classes of the nearest marker to a QTL peak.

Genotypes, linkage map construction and QTL mapping All animals of the pedigree were genotyped at 123 marker loci on 25 chromosomes. QTL mapping was performed with a pedigree-specific linkage map as described previously (Goraga et al. 2011). Prior to QTL analysis, the phenotypic data were corrected for hatch (four levels in the WL77 9 NHI cross and five levels in the NHI 9 WL77 cross) and family (number of hens that were mated with the same cock; two levels in the NHI 9 WL77 cross and three levels in the WL77 9 NHI cross). Therefore, QTL effects account for phenotypic F2 variation after correction. The standard model for the analysis of the combined reciprocal crosses included cross (two levels) as a fixed effect. To test whether the direction of cross affected the QTL, cross was included as an interactive covariate in the QTL model. The effect of cross was considered significant if the difference between the full model and the reduced model DF was 4.6. An F-value difference of 4.6 corresponds roughly to a significance level of P 0.05 at a LOD difference of 2.0 (Li et al. 2006). The QTL positions are given for the peak F-values as the cM distance from the first marker on the chromosome, which was placed at 0 cM. To translate genetic positions (cM) into physical positions (Mb), we used the Mb positions of markers according to the genomic chicken sequence (http://www.ensembl.org/Gallus_gallus/, Ensembl 60 WASHUC2 and http://www.ncbi.nlm.nih.gov/, NCBI build 2.1). The direction of genetic effects is given as NHI allele effect compared to WL77. If the NHI effect is larger than the WL77, the effect is positive; otherwise negative. Pedigree-specific significance thresholds were derived by random QTL scans with 1000 permutations of the data (Churchill & Doerge 1994). The following thresholds were taken from the resulting F-test statistic distribution: ‘genome-wide highly significant’ (a = 0.01) corresponded to an F-value of 10.36, ‘genome-wide significant’ (a = 0.05) corresponded to F = 8.47, and ‘genome-wide suggestive’ corresponded to ‘chromosome-wide significant’ (a = 0.05) and ranged between 3.3 F 6.4 for the different chromosomes. The 95% confidence interval of a single QTL was estimated using a parametric bootstrap analysis

© 2012 The Authors, Animal Genetics © 2012 Stichting International Foundation for Animal Genetics, doi: 10.1111/j.1365-2052.2012.02344.x

QTL for muscle weight and carcass composition with 1000 iterations (Visscher et al. 1996). The percentage of phenotypic F2 variance explained by an identified QTL was calculated for each trait as a reduction of sums of squares in the full model compared to the reduced model.

Results Phenotypic characteristics of NHI, WL77, F1 and F2 animals NHI chickens were about twice as heavy as WL77 were at the age of 20 weeks (Table 1). The weights of breast muscle and drumsticks–thighs were about 2- and 2.4-fold higher respectively in NHI chickens when compared to those of WL77 chickens. The F1 and F2 means of body weight and weights of carcass, breast muscle, drumsticks– thighs and residual carcass shifted towards the mean values of NHI animals, suggesting dominance components in the mode of inheritance. The F2 variance was about 1.5to 2.2-fold as high as the respective F1 variance. In the comparison of the reciprocal crosses, NHI 9 WL77 had a better performance than did WL77 9 NHI. The F2 chicken population with a NHI grand-sire showed significantly higher mean values of all measurements and had higher variances. The heritabilities for body weight and carcass composition traits ranged from 0.21 to 0.31, indicating genetic determinants for all traits. In both reciprocal crosses, all carcass composition traits were highly correlated with all others (r 0.74) (Table 2).

QTL effects In the combined analysis of both reciprocal crosses, we detected genomic regions affecting body weight and carcass traits on 14 chromosomes (Table 3). Each trait was controlled by four to nine QTL, which at least surpassed the suggestive level of significance. QTL on chromosomes 1, 2, 4, 5, 7 and 27 were highly significant for at least one trait. Individual QTL accounted for 2.9–40.2% of the corresponding phenotypic F2 variance. The most significant (20.06 F 67.56) genomic region affecting body weight and all carcass traits was mapped on GGA4 with peak F-value positions at 155 and 156 cM (76.93 and 77.55 Mb respectively). The 95% confidence interval for the different traits (Table 3, Fig. S1) was 4–9 cM long. The nearest marker to the peak F-value position was UMA4.034 at 153.3 cM (75.89 Mb). The locus accounted for 13.7–40.2% of the phenotypic F2 variance of the corresponding traits. The genetic effect on all traits was additive (Table 3) with the NHI allele as the increasing allele. Chickens homozygous for the NHI allele at the GGA4 QTL were on average 300 g (P < 0.0001) heavier and had 34 g (P < 0.0001) more breast muscle than did homozygous WL77 chickens. The direction of

the QTL effect was consistent in both reciprocal crosses, but the magnitude was higher in the NHI 9 WL77 cross direction (Figs 1 and S2). The second highest QTL effect was found on GGA27 and contributed to highly significant differences in body weight (F = 11.51) and the weights of breast muscle (F = 13.37), drumsticks–thighs (F = 16.12) and residual carcass (F = 10.61) and, to a lesser extent, wings (F = 8.48) and carcass weights (F = 7.20) (Table 3, Fig. S1). The peak QTL positions were located between 46 and 52 cM (3.61 and 3.87 Mb respectively) with confidence intervals covering almost the whole chromosome. The nearest marker to the peak F-value position is ADL0376 at 52 cM (3.87 Mb) at the distal end of the chromosome. The locus explained 5.3–13.8% of the phenotypic F2 variance of the corresponding traits. At this locus, two different high alleles, N1 and N2, were segregating in the NHI line at the marker closest to the peak F-value position (Figs 1 and S2). The allelic effects of both alleles were almost additive in combination with the low WL77 allele (W). When the two NHI alleles (N1 and N2) were inherited together, the effect tended to be additive for all traits, except for the weight of wings, where the higher allele was dominant over the lower allele in the WL77 9 NHI cross. Despite the wide F-value curve across GGA27, a scan for two QTL provided no evidence for a second QTL. The NHI line was also the origin of high alleles for additionally identified highly significant and significant QTL that influenced body weight on GGA2 and 11 and the weight of drumsticks–thighs on GGA1, 5 and 7. These QTL explained 6.9–12.1% of the phenotypic F2 variance of the corresponding traits (Table 3). Testing cross as a factor interacting with the QTL provided statistical evidence for six out of 14 genomic regions that had significantly different effects in the reciprocal crosses (Table 3). For example, the significant QTL (F = 10.28) for weight of drumsticks and thighs on GGA1 was detected in the NHI 9 WL77 cross, but not in the reciprocal cross. Thus, such QTL is specific to the cross in which it was identified.

Discussion Owing to the large phenotypic differences between NHI and WL77 chickens and the high correlation between body weight and different carcass traits, we expected to detect QTL regions that had effects on more than one trait. The identified QTL clearly illustrate that many loci across the genome have responded to the selection procedure for high meat yield in the NHI line. Most alleles inherited from the NHI line were associated with higher performance for these loci. Furthermore, this study confirms that most genetic effects contributing to body weight and carcass traits act additively (Ikeobi et al. 2004; Hocking 2005; Jacobsson et al. 2005; Nadaf et al. 2009).


3

4

Nassar, Goraga and Brockmann Table 1 Phenotypic characterization of parental lines, F1 and F2 males of the reciprocal crosses at the age of 20 weeks. Mean (SD)1 Parental lines

Measurement Absolute values, g Body weight Carcass weight Breast muscle weight Drumsticks and thighs weight Wings weight Residual carcass weight Relative values,3 % Breast muscle weight Drumsticks and thighs weight Wings weight Residual carcass weight

NHI n=6

F2 2

F1 NHI 9 WL77 n = 14

WL77 n=5

WL77 9 NHI n = 11

NHI 9 WL77 n = 127

WL77 9 NHI n = 151

2786 1915 342 690 251 390

(162)a,A (122)a,A (36)a,A (53)a,A (18)a,A (29)a,A

1442 909 178 292 110 187

(73)b,C (23)b,C (11)b,B (16)b,C (6)c,C (10)b,B

2551 1679 312 554 186 358

(122)a (95)a (20)a (51)a (9)b (31)a

2385 1603 300 523 183 349

(141)B,** (118)B (19)B,* (39)B (11)B (50)A

2426 1666 310 552 198 356

(245) (199) (36) (76) (20) (38)

2278 1552 290 509 184 330

(198)*** (148)*** (31)*** (60)*** (17)*** (30)***

17.8 36.0 13.1 20.3

(1.2)a,A (1.0)a,A (0.9)a,A (0.5)a,B

19.6 32.1 12.1 20.5

(1.0)b,B (1.1)b,B (0.4)b,B (0.8)a,B

18.6 33.0 11.1 21.4

(0.7)ab (2.1)b (0.4)c (1.8)a

18.7 32.7 11.5 21.7

(0.6)AB (2.0)B (0.6)B,* (2.1)A,*

18.4 33.1 11.9 21.4

(1.0) (1.5) (0.7) (1.0)

18.7 32.8 11.9 21.3

(1.1)* (1.4) (0.6) (0.8)

NHI, partially inbred New Hampshire; WL77, inbred White Leghorn. SD, standard deviations were calculated for each population separately. 2 Data were corrected for hatch and family as fixed effects, in each cross separately. 3 Values relative to carcass weight. a,b,c Significant differences in absolute or relative trait values between parental and F1 of the cross NHI 9 WL77 for the same trait (P 0.05). A,B,C Significant differences in absolute or relative trait values between parental and F1 of the cross WL77 9 NHI for the same trait (P 0.05). *P 0.05, **P 0.01 and ***P 0.001 refer to significant differences between reciprocal crosses (ANOVA followed by Tukey–Kramer test). 1

Table 2 Pearson’s correlation coefficients (P < 0.0001) among measurements in the F2 populations.

Measurement

Body weight

Carcass weight

Breast muscle weight

Drumsticks and thighs weight

Wings weight

Residual carcass weight

Body weight Carcass weight Breast muscle weight Drumsticks and thighs weight Wings weight Residual carcass weight

0.29 0.98 0.82 0.92 0.85 0.91

0.98 0.27 0.85 0.94 0.88 0.93

0.89 0.90 0.23 0.74 0.74 0.75

0.94 0.94 0.82 0.31 0.86 0.83

0.87 0.87 0.77 0.85 0.24 0.81

0.92 0.93 0.82 0.87 0.88 0.21

NHI 9 WL77 (above diagonal) and WL77 9 NHI (below diagonal). NHI, partially inbred New Hampshire; WL77, inbred White Leghorn. Heritability estimates are given in the diagonal for combined crosses and were estimated by analysis of sire and dam variance components (PROC VARCOMP) using the restricted maximum-likelihood method.

Although most of the detected QTL explained a small part of the phenotypic F2 variance, two genomic regions harboured strong effects on the analysed traits. The QTL on GGA4 and 27 explained 13.7–40.2% and 5.3–13.8% of the phenotypic variance respectively. These QTL affected many of the correlated traits and might represent either pleiotropic effects of the underlying gene or different genes in the QTL region that affect traits independently of each other. This is supported by the fact that if carcass weight was included as a covariate in the model, the QTL for breast muscle weight on GGA4 and 27 were lost (data not shown). In this study, the biggest effect on all traits was detected in the distal part of GGA4. This genomic region has been repeatedly mapped for body weight and carcass composi-

tion traits in crosses between high- and low-growth selected lines (Jacobsson et al. 2005; Park et al. 2006; Nadaf et al. 2009) and in crosses between different breeds, for example White Leghorn (layer) with either broiler (meat-type chicken) (Sewalem et al. 2002; Carlborg et al. 2004; Ikeobi et al. 2004; Schreiweis et al. 2005; Zhou et al. 2006a,b), Rhode Island Red (Tuiskula-Haavisto et al. 2002), Red Junglefowl (Kerje et al. 2003) or Oh-Shamo (native Japanese breed) (Tsudzuki et al. 2007). The same genomic segment on GGA4 was also found in crosses between broiler and Fayoumi (native Egyptian breed) (Zhou et al. 2006a,b) or layer lines (Ambo et al. 2009). Albeit the traits in our study were measured at 20 weeks of age, the coincidence of QTL intervals with those detected in many other studies, in which traits were anal-


QTL for muscle weight and carcass composition Table 3 Positions and effects of QTL for body weight and carcass composition in 20-week-old males of the reciprocal crosses between partially inbred New Hampshire (NHI) and inbred White Leghorn (WL77).

Measurement

GGA

cM (Mb)1

95% confidence interval (cM)

Marker2

F3

DF4

Additive effect (SE)5

Dominance effect (SE)5

% P6

Body weight

1 2 4 5 11 13 14 20 27 2 4 7 11 13 27 2 4 7 13 14 27 28 1 4 5 7 8 12 13 14 27 4 7 12 27 1 4 5 10 11 13 27

118 94 155 105 40 105 10 14 52 95 156 13 44 107 50 99 155 32 106 8 51 33 103 156 107 31 21 0 100 61 51 155 1 0 47 124 156 111 44 46 109 46

28.5–311.5 43.0–109.5 153.0–157.0 32.0–118.0 8.5–58.0 67.0–113.0 0.0–69.0 0.0–14.0 16.0–52.0 35.0–177.0 153.0–159.0 0.0–36.0 28.5–57.0 57.5–115.0 10.5–52.0 44.5–156.0 151.5–160.5 0.0–36.0 9.0–113.5 0.0–69.0 20.0–52.0 8.0–34.0 41.5–203.0 152.0–159.0 87.5–115.0 0.0–36.0 0.0–68.0 0.0–10.0 0.0–112.0 3.5–69.0 37.0–52.0 152.5–158.0 0.0–36.0 0.0–13.0 4.0–52.0 1.0–324.0 153.0–158.0 3.5–135.0 15.0–88.0 16.0–63.0 41.5–118.0 18.0–52.0

MCW0112 MCW0062 UMA4.034 ADL0166 ADL0308 MCW0216 MCW0136 ADL0034 ADL0376 MCW0062 UMA4.034 MCW0183 ADL0308 MCW0216 ADL0376 MCW0062 UMA4.034 MCW0236 MCW0216 MCW0136 ADL0376 ADL0299 MCW0112 UMA4.034 ADL0166 MCW0236 ADL0154 LEI0131 MCW0216 LEI0066 ADL0376 UMA4.034 MCW0178 LEI0131 ADL0376 MCW0112 UMA4.034 ADL0166 MCW0067 ADL0308 MCW0216 ADL0376

7.32† 11.10** 58.14** 7.40† 9.42* 7.82† 7.79† 3.80† 11.51** 5.83† 32.50** 6.68† 6.12† 6.40† 7.20† 7.82† 20.06** 5.97† 6.00† 5.76† 13.37** 5.62† 12.85** 67.56** 13.83** 10.95** 4.71† 5.74† 5.65† 6.70† 16.12** 55.23** 6.23† 4.52† 8.48* 6.58† 40.95** 5.58† 6.27† 5.87† 7.13† 10.61**

1.10 5.53 28.81 3.24 3.51 2.30 3.01 0.84 4.73 2.09 16.31 2.62 1.52 2.04 2.58 3.89 9.33 1.88 1.33 2.47 6.29 2.56 4.55 34.11 6.26 4.68 2.36 2.54 2.11 1.43 7.19 27.49 2.60 1.91 7.01 0.54 20.51 2.54 2.96 2.71 2.54 4.64

57.2 69.1 168.6 25.0 30.4 17.4 87.8 17.3 69.9 43.5 107.9 55.3 19.8 1.5 48.6 10.6 16.1 8.3 2.5 14.5 12.2 8.2 24.5 55.7 9.8 22.3 13.2 3.6 4.5 4.8 25.0 14.0 6.3 3.3 5.2 8.1 23.6 3.2 6.8 4.7 3.3 11.3

36.0 49.5 9.7 98.0 109.3 114.0 63.1 53.0 9.3 34.4 21.5 0.5 67.4 89.2 2.1 6.9 1.4 6.2 16.4 8.1 3.1 2.3 14.2 6.5 44.7 4.1 6.7 21.1 26.6 29.6 8.4 0.3 3.2 3.0 1.6 8.6 1.2 17.6 8.4 11.0 16.8 4.8

5.4 8.0 31.3 5.5 6.9 5.8 5.8 2.9 8.3 4.4 20.2 4.9 4.5 4.7 5.3 5.8 13.7 4.5 4.5 4.4 9.6 4.3 11.3 40.2 12.1 9.8 4.5 5.4 5.3 6.3 13.8 29.9 4.6 3.4 6.0 4.9 24.3 4.2 4.7 4.4 5.3 7.7

Carcass weight

Breast muscle weight

Drumsticks and thighs weight

Wings weight

Residual carcass weight

(68.84) (57.85) (76.93) (55.40) (14.61) (14.33) (2.26) (3.14) (3.87) (58.17) (77.55) (22.58) (15.91) (14.87) (3.78) (59.44) (76.93) (28.24) (14.61) (2.02) (3.82) (4.18) (63.48) (77.55) (55.72) (27.89) (20.08) (4.90) (13.00) (8.34) (3.82) (76.93) (19.96) (4.90) (3.66) (70.90) (77.55) (56.37) (10.10) (16.56) (14.89) (3.61)

(16.4) (16.7) (15.7) (16.7) (16.8) (18.6) (26.5) (14.8) (14.7) (14.6) (13.5) (15.4) (13.5) (15.8) (12.8) (2.9) (2.5) (2.7) (3.0) (4.3) (2.4) (2.5) (5.1) (4.8) (5.1) (4.8) (4.5) (4.4) (5.7) (5.4) (4.5) (1.3) (1.8) (1.3) (1.3) (2.8) (2.6) (3.0) (2.4) (2.5) (3.0) (2.6)

(26.0) (26.2) (21.9) (28.2) (28.0) (29.3) (62.3) (20.8) (21.0) (23.4) (19.3) (24.1) (21.3) (25.0) (19.3) (4.8) (3.6) (4.1) (4.8) (10.1) (3.6) (3.6) (9.3) (6.7) (9.1) (7.6) (6.6) (6.3) (8.5) (8.5) (6.6) (1.9) (2.8) (1.8) (2.0) (4.6) (3.7) (5.6) (3.7) (3.8) (4.7) (3.8)

1

Chromosomal location is given as pedigree-specific cM position (Mb); first marker on each chromosome was set at 0 cM. Marker closest to the chromosomal position with the highest F-value. 3 F-value is F-statistic for QTL using standard model. 4 DF-values are in boldface when QTL interaction with cross was considered significant if the difference between the F-value of the standard model and the cross interaction model was 4.6, which corresponded to a LOD score of 2.0 in this F2 population. 5 The direction of QTL effects is given as NHI allele effect compared to WL77. 6 Phenotypic F2 variance (%) explained by the QTL; QTL effect given as reduction of the residual sum of squares fitting 1 vs. 0 QTL. **Highly significant at 1% genome-wide level (F 10.36), *Significant at 5% genome-wide level (F 8.47), † Significant at 5% chromosome-wide level (3.3 F 6.4), which is assumed suggestive at the genome-wide level. 2

ysed at younger ages, shows that these genomic loci affect the traits during a long developmental period. The later age analysed in our study contributed to the detection of an additional QTL effect for residual carcass weight that has not been reported in other crosses. The confidence interval

of the QTL region on GGA4 harbours 86 genes (Ensembl release 60), among them are the genes encoding the fibroblast growth factor-binding proteins 1 and 2 (FGFBP1 and FGFBP2), which might influence carcass and muscle development.


5

Nassar, Goraga and Brockmann

Breast muscle weight (g)

360

NHI x WL77 WL77 x NHI

***

330

**

300

** *

270

***

NN

NW

WW

240 UMA4.034 @ GGA4 (75.89 Mb)

360 Breast muscle weight (g)

340

NHI x WL77 WL77 x NHI

***

*

320

**

300 **

280

***

WW

N 2W

WW N2 N2

N1W

N 2 N2 N 1 N1

N1N2

260 N1N1

6

ADL0376 @ GGA27 (3.87 Mb) Figure 1 Exemplary genotype effect plots of the markers UMA4.034 and ADL0376 nearest to the QTL peaks on GGA4 and 27 affecting all carcass traits respectively in 20-week-old males of the F2 populations of the reciprocal crosses between NHI and WL77. Values are LSM ± SE. N: New Hampshire allele, W: White Leghorn allele, N1 and N2 belong to alternative alleles in NHI chickens at marker ADL0376. In the cross NHI x WL77, genotype N1N1 did not occur. *P 0.05, **P 0.01 and ***P 0.001 refer to significant differences between genotype classes (ANOVA followed by TukeyKramer test). See Figure S2 for other traits.

The highly significant QTL affecting body weight and carcass traits on GGA27 in our crosses confirms QTL effects from other crosses for body weight (Sewalem et al. 2002; Kerje et al. 2003; Schreiweis et al. 2005; Ambo et al. 2009), drumsticks–thighs weight (Ankra-Badu et al. 2010) and wings weight (Ikeobi et al. 2004). In addition, our data are the first to provide evidence that this QTL region contains a gene or genes affecting breast muscle weight. Effects were also found for residual carcass and carcass weights. The confidence interval comprises the whole microchromosome harbouring 191 genes (Ensembl release 60). Among them is the chicken growth hormone (GH) gene, for example, which was associated with muscle fibre characteristics in a F2 White Plymouth Rock 9 Xinghua cross (Lei et al. 2007). Other functional candidates include the type II growth hormone-releasing hormone receptor (Type II GHRHR), the membrane protein palmitoylated 3 (MAGUK

p55 subfamily member 3) (MPP3), the homeobox B9 (HOXB9), and the signal transducers and activator of transcription 5B (STAT5B) genes. The fact that the QTL regions on GGA4 and 27 have been identified in many crosses between layers and meattype chickens is likely due to the joined phylogenetic origin of either layer or broiler chicken lines (Rubin et al. 2010). As the magnitudes of the genetic effects reported for the different mapping populations vary, it can be assumed that more than two alleles very likely occur in the different breeds. In our cross, we found different alleles for the QTL on GGA27 in the NHI line. Males of the NHI 9 WL77 cross with the father descending from the NHI line were heavier than those of the reciprocal cross. The phenotypic differences between the two reciprocal crosses originate in part from residual genetic variation among the NHI animals as a result of incomplete inbreeding. This could also affect potential QTL–genetic background interaction effects. Evidence for this is given by different allelic effects of the QTL on GGA27. The inbreeding coefficient in the NHI line was 0.86 at the time the cross-bred experiment was performed. Additional cross differences may have been caused by genetic variation of the W chromosome and mitochondrial DNA between the parental NHI and WL77 hens. Furthermore, epigenetic effects on early growth during the embryonic development owing to different egg sizes of NHI and WL77 hens, and subsequently F1 hens, cannot be ruled out. In addition to significant and highly significant QTL, we reported genome-wide suggestive QTL that have not been reported before in other crosses, for example the QTL for carcass weight on GGA11 and 13, for breast muscle weight on GGA14 and 28, and for drumsticks–thighs weight on GGA8, 12 and 14. The QTL for wings weight on GGA12 and for residual carcass weight on GGA5, 10, 11 and 13 are also novel. The suggestive QTL for body weight and the weights of carcass, breast muscle and wings on GGA1, 2, 5, 7, 13, 14 and 20 are in agreement with findings in other QTL studies (Carlborg et al. 2004; Ikeobi et al. 2004; Jacobsson et al. 2005; Nadaf et al. 2009; Tercˇicˇ et al. 2009). In conclusion, this project was the initial step for discovering the genetic basis for the big phenotypic differences between the NHI and WL77 lines in muscle weight and carcass composition. QTL with large effects that have been reported previously in other studies could be confirmed. Furthermore, several QTL effects were detected that have not been reported before. The highly significant QTL on GGA4 and 27, which are responsible for variability in muscle and carcass composition traits, make a search for causative candidate genes feasible. The mapping resolution could be efficiently increased by combining data from two or more crosses where QTL on GGA4 and 27 were also found or by association studies in advanced intercross populations, for example.


QTL for muscle weight and carcass composition

Acknowledgements This work was supported by a grant from the Yousef Jameel Foundation. We acknowledge Jacob Litke of Wesleyan University in Middletown, Connecticut, for his comments.

References Ambo M., Moura A.S.A.M.T., Ledur M.C. et al. (2009) Quantitative trait loci for performance traits in a broiler 9 layer cross. Animal Genetics 40, 200–8. Ankra-Badu G.A., Shriner D., Le Bihan-Duval E. et al. (2010) Mapping main, epistatic and sex-specific QTL for body composition in a chicken population divergently selected for low or high growth rate. BMC Genomics 11, 107. Carlborg O., Hocking P.M., Burt D.W. & Haley C.S. (2004) Simultaneous mapping of epistatic QTL in chickens reveals clusters of QTL pairs with similar genetic effects on growth. Genetical Research 83, 197–209. Churchill G.A. & Doerge R.W. (1994) Empirical threshold values for quantitative trait mapping. Genetics 138, 963–71. Goraga Z., Nassar M., Schramm G.P. & Brockmann G.A. (2010) Phenotypic characterization of chicken inbred lines that differ extremely in growth, body composition and egg production traits. Archiv Tierzucht 53, 337–49. Goraga Z.S., Nassar M.K. & Brockmann G.A. (2011) Quantitative trait loci segregating in crosses between New Hampshire and White Leghorn chicken lines: I. Egg production traits. Animal Genetics 43, 183–9. Hocking P.M. (2005) Review of QTL mapping results in chickens. Worlds Poultry Science Journal 61, 215–26. Ikeobi C.O.N., Woolliams J.A., Morrice D.R., Law A., Windsor D., Burt D.W. & Hocking P.M. (2004) Quantitative trait loci for meat yield and muscle distribution in a broiler layer cross. Livestock Production Science 87, 143–51. Jacobsson L., Park H.B., Wahlberg P., Fredriksson R., Perez-Enciso M., Siegel P.B. & Andersson L. (2005) Many QTLs with minor additive effects are associated with a large difference in growth between two selection lines in chickens. Genetical Research 86, 115–25. Kerje S., Carlborg O., Jacobsson L., Schutz K., Hartmann C., Jensen P. & Andersson L. (2003) The twofold difference in adult size between the Red Junglefowl and White Leghorn chickens is largely explained by a limited number of QTLs. Animal Genetics 34, 264–74. Lei M., Luo C., Peng X., Fang M., Nie Q., Zhang D., Yang G. & Zhang X. (2007) Polymorphism of growth-correlated genes associated with fatness and muscle fiber traits in chickens. Poultry Science 86, 835–42. Li R., Tsaih S.W., Shockley K., Stylianou I.M., Wergedal J., Paigen B. & Churchill G.A. (2006) Structural model analysis of multiple quantitative traits. PLoS Genetics 2, e114. Nadaf J., Pitel F., Gilbert H. et al. (2009) QTL for several metabolic traits map to loci controlling growth and body composition in an F2 intercross between high- and low-growth chicken lines. Physiological Genomics 38, 241–9. Park H.-B., Jacobsson L., Wahlberg P., Siegel P.B. & Andersson L. (2006) QTL analysis of body composition and metabolic traits in an intercross between chicken lines divergently selected for growth. Physiological Genomics 25, 216–23.

Rubin C.J., Zody M.C., Eriksson J. et al. (2010) Whole-genome resequencing reveals loci under selection during chicken domestication. Nature 464, 587–91. SAS Institute (2008) SAS User’s Guide: Statistics, Version 9.2. SAS Institute, Inc., Cary, NC, USA. Schreiweis M.A., Hester P.Y. & Moody D.E. (2005) Identification of quantitative trait loci associated with bone traits and body weight in an F2 resource population of chickens. Genetics Selection Evolution 37, 677–98. Sewalem A., Morrice D., Law A., Windsor D., Haley C., Ikeobi C., Burt D. & Hocking P. (2002) Mapping of quantitative trait loci for body weight at three, six, and nine weeks of age in a broiler layer cross. Poultry Science 81, 1775–81. Tercˇicˇ D., Holcman A., Dovcˇ P., Morrice D.R., Burt D.W., Hocking P.M. & Horvat S. (2009) Identification of chromosomal regions associated with growth and carcass traits in an F3 full sib intercross line originating from a cross of chicken lines divergently selected on body weight. Animal Genetics 40, 743–8. Tsudzuki M., Onitsuka S., Akiyama R., Iwamizu M., Goto N., Nishibori M., Takahashi H. & Ishikawa A. (2007) Identification of quantitative trait loci affecting shank length, body weight and carcass weight from the Japanese cockfighting chicken breed, Oh-Shamo (Japanese Large Game). Cytogenetic and Genome Research 117, 288–95. Tuiskula-Haavisto M., Honkatukia M., Vilkki J., de Koning D.J., Schulman N.F. & Maki-Tanila A. (2002) Mapping of quantitative trait loci affecting quality and production traits in egg layers. Poultry Science 81, 919–27. Visscher P.M., Thompson R. & Haley C.S. (1996) Confidence intervals in QTL mapping by bootstrapping. Genetics 143, 1013–20. Zhou H., Deeb N., Evock-Clover C.M., Ashwell C.M. & Lamont S.J. (2006a) Genome-wide linkage analysis to identify chromosomal regions affecting phenotypic traits in the chicken. I. Growth and average daily gain. Poultry Science 85, 1700–11. Zhou H., Deeb N., Evock-Clover C.M., Ashwell C.M. & Lamont S.J. (2006b) Genome-wide linkage analysis to identify chromosomal regions affecting phenotypic traits in the chicken. II. Body composition. Poultry Science 85, 1712–21.

Supporting information Additional supporting information may be found in the online version of this article. Figure S1 F-value curves across GGA4 (left) and 27 (right) pertaining to one QTL scans for body weight and carcass composition traits in 20-week-old male F2 populations. Figure S2 Genotype effect-plots of the markers UMA4.034 and ADL0376 nearest to the QTL peaks on GGA4 (left) and 27 (right) respectively in 20-week-old male F2 populations of the reciprocal crosses between NHI and WL77. Table S1 Population structure and number of animals in the pedigree. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.


7