2019 年 56 巻 2 号 p. 91-95
Amakusa Daioh cross chickens are F1 hybrids of restored Amakusa Daioh sires and Kyushu Rhode dams. In the present study, the association between a single nucleotide polymorphism (SNP; AB604331, g.420 C>A) in the cholecystokinin type A receptor gene and growth traits in Amakusa Daioh cross chicken were investigated. We used 72 male and 72 female birds that had hatched on the same day, were raised in the same chicken house, and were fed the same diet ad libitum from day 0 to 17 weeks (wks) of age. Body weight was recorded at weekly intervals and average daily gain of each week interval was calculated from body weight data. Birds were sacrificed at 17 wks and carcass traits were recorded. SNP genotyping was carried out using the mismatch amplification mutation assay. Associations between the SNP and growth traits were analyzed by a generalized linear model. Body weight from 6 to 17 wks was higher in birds with the A allele than in birds with the C allele, although significant differences in average daily gain traits between birds with A and C alleles were not detected during most of the duration of the experiment. Carcass data showed that birds with the A allele had heavier wings and a smaller proportion of the gizzard than those with the C allele. The g.420 C>A SNP will be useful as a selection marker for parent stock lines to increase the growth performance of Amakusa Daioh cross chickens.
In Japan, fast-growing broilers constitute more than 90% of poultry meat production. However, “Jidori,” as defined in the Japanese Agricultural Standard (JAS, Ministry of Agriculture, Forestry and Fisheries of Japan, 1999), still occupies a certain share in the Japanese market, because Japanese consumers recognize and appreciate its rich taste and flavor. The JAS defines Jidori as cross-bred chickens with over 50% of native-bred hereditary. Most Jidori chickens are bred by crossing indigenous Japanese breeds with highly selected lines that exhibit rapid growth and/or high egg productivity.
Of the indigenous chicken breeds in Japan, “Amakusa Daioh,” native to the Kumamoto prefecture, was one of the largest. Unfortunately, Amakusa Daioh became extinct in the early Showa era (1926–1981). In 2001, the Kumamoto Prefectural Government announced that Amakusa Daioh had been restored by the Animal Husbandry Research Institute, Kumamoto Prefectural Agricultural Research Center (Koshi, Japan) (Matsuzaki et al., 2001). The procedure used was as follows: first, Shamo (Japanese Game), Kumamoto Cochin (a local variety of Cochin breed native to China), and Langshan (native to China) breeds were crossed in accordance with the documented history of the establishment of the Amakusa Daioh breed. Then, the offspring were selected for body size and shape, including external features that were based on an old painting that had existed for seven generations. F1 hybrid chickens from the cross between restored Amakusa Daioh cocks and Kyushu Rhode (a synthetic breed resulting from a cross between Rhode Island Red and White Plymouth Rock) hens are currently commercially available as “Amakusa Daioh cross” in the Japanese market. The meat taste of the Amakusa Daioh cross is recognized and is gaining popularity in Japan.
Although Amakusa Daioh cross chickens are the largest among Jidori chickens, their growth is far slower than that of broilers (Nikki et al., 2012). Thus, breeding for growth traits is a major consideration in Amakusa Daioh cross production. Recently, in the Hinai-dori breed native to the Akita prefecture of Japan, a significant association between a single nucleotide polymorphism (SNP; AB604331, g.420 C>A) in the 5′-untranslated region of the cholecystokinin type A receptor gene (CCKAR) and growth traits was reported (Rikimaru et al., 2013). The objective of the present study was to verify whether the g.420 C>A SNP in CCKAR can be used as a marker to improve growth traits of the Amakusa Daioh cross.
Experimental birds received humane care as outlined in the Guidelines for Proper Conduct of Animal Experiments (Science Council of Japan, 2006).
One hundred forty-four birds (72 male and 72 female) of Amakusa Daioh cross that had hatched on July 2, 2014 were used. The chicks were raised in a battery cage until 2 weeks (wks) of age and then in solid-floored pens in a conventional poultry research house until 17 wks of age. The chicks were fed a starter diet (ME, 3,050 kcal/kg; CP, 23% (wt/wt)) from 0 to 4 wks, a grower diet (ME, 3,210 kcal/kg; CP, 18%) from 5 to 9 wks, and a finisher diet (ME, 3,260 kcal/kg; CP, 18%) from 10 to 17 wks. The chickens had ad libitum access to feed and water throughout the experiment.
The body weight (BW) of the birds was measured at weekly intervals from day-of-hatch (0 d) to 17 wks. The average daily gain (ADG) of each weekly interval was calculated from the BW at each week of age. At 17 wks, 67 birds were randomly selected, fasted overnight, and slaughtered. After they were bled, each carcass was scalded at 63°C for 2 min, and feathers were removed with an automatic picker. The carcass was cooled in ice-cold water for 1 h. Evisceration was done by hand and the head was separated. The carcass was dissected, and the legs and breast were deboned. Each part (breast, thigh, tender, wings, bones, abdominal fat, liver, heart, gizzard) was weighed. The percentage weight of each carcass part was calculated as its weight relative to live body weight at 17 wks.
CCKAR GenotypingBlood was collected from the ulnar vein at 6 wks, spotted onto an FTA filter paper (WB120028; GE Healthcare, Buckinghamshire, UK), and air-dried at room temperature. Genomic DNA was extracted from the FTA paper as described in Rikimaru et al. (2013). The SNP (AB604331: g.420 C>A) in the 5′-untranslated region of CCKAR was genotyped using the mismatch amplification mutation assay as described in Rikimaru et al. (2013).
Statistical AnalysisSNP–trait association analysis was conducted in R (R Core Team, 2016) using the following generalized linear model:
y=µ+Css+Caa+Cdd+e,
where y is the response variable of each phenotype; µ is the intercept; the sex effect (s) is a covariate coefficient with Cs having values of 0 and 1 for male and female, respectively; the additive effect (a) is a covariate coefficient with Ca having values of 2, 1, and 0; and the dominance effect (d) is a covariate coefficient with Cd having values of 0, 1, and 0, for genotypes A/A, A/C, and C/C, respectively; and e is the residual standard error.
The percentage of haplotype variance explained by the model was calculated as described in Rikimaru et al. (2012), using the following formula:
variance percentage=100×(1−Fvar / Rvar),
where Rvar is the residual variance from the reduced model, omitting the additive effect but including sex and dominance effects; Fvar is the residual variance from the full model, including sex, additive, and dominance effects.
In the population of Amakusa Daioh cross chickens, the genotype frequencies of AA, AC, and CC at the g.420 C>A SNP were 0.347, 0.563, and 0.090, respectively. The distribution of the genotypes in the population did not contradict Hardy–Weinberg equilibrium. Allele frequencies of A and C were 0.628 and 0.372, respectively.
Sex, additive, and dominance effects of the SNP on BW traits from 0 d to 17 wks are shown in Table 1. Males were significantly heavier than females from 1 to 17 wks. As for the additive effect, the significant differences between chickens carrying the A allele and those having the C allele from 6 to 17 wks show that the A allele is significantly superior to the C allele in these traits. As for the dominance effect, there were no significant differences between birds with the A allele and those with the C allele throughout the experiment.
| Body weight (BW, g) | Intercept | Sex | Additive | Dominance | Rvar | Fvar | Variance (%) |
|---|---|---|---|---|---|---|---|
| 0 d | 44.3±1.0 | −0.6±0.6 | 0.4±0.5 | −0.3±0.6 | 1573.8 | 1566.1 | 0.5 |
| 1 wk | 118.4±2.5 | −7.5±1.4*** | 2.2±1.3 | 2.3±1.6 | 9818.2 | 9627.1 | 1.9 |
| 2 wks | 238.5±5.0 | −21.8±2.9*** | 5.4±2.6* | 1.9±3.3 | 41486 | 40294 | 2.9 |
| 3 wks | 409.9±8.9 | −52.9±5.1*** | 8.5±4.7 | 2.3±5.8 | 130648 | 127648 | 2.3 |
| 4 wks | 630.8±13.8 | −83.1±7.9*** | 10.9±7.3 | 0.2±9.0 | 310661 | 305725 | 1.6 |
| 5 wks | 873.3±20.9 | −156.6±11.9*** | 20.7±11.0 | −6.9±13.6 | 714854 | 697182 | 2.5 |
| 6 wks | 1128.0±29.0 | −213.0±16.5*** | 36.2±15.3* | −7.1±18.9 | 1400657 | 1346500 | 3.9 |
| 7 wks | 1450.4±33.1 | −297.0±18.8*** | 39.7±17.4* | −13.5±21.5 | 1810545 | 1745458 | 3.6 |
| 8 wks | 1763.6±39.6 | −393.4±22.5*** | 44.8±20.8* | −24.2±25.8 | 2589666 | 2506772 | 3.2 |
| 9 wks | 2068.7±44.7 | −486.6±25.4*** | 57.5±23.5* | −26.7±29.1 | 3324404 | 3188028 | 4.1 |
| 10 wks | 2395.1±52.4 | −592.1±29.7*** | 71.0±27.6* | −26.5±34.1 | 4589994 | 4382028 | 4.5 |
| 11 wks | 2724.2±56.9 | −711.8±32.3*** | 74.2±29.9* | −25.0±37.1 | 5398083 | 5170993 | 4.2 |
| 12 wks | 3009.5±61.4 | −820.5±34.9*** | 85.4±32.3** | −26.3±40.0 | 6322498 | 6021976 | 4.8 |
| 13 wks | 3238.4±64.6 | −899.8±36.7*** | 88.2±34.0* | −21.6±42.1 | 6984011 | 6663441 | 4.6 |
| 14 wks | 3507.9±73.0 | −1017.8±41.4*** | 101.3±38.4** | −35.0±47.5 | 8925085 | 8502412 | 4.7 |
| 15 wks | 3684.9±74.8 | −1025.5±42.5*** | 85.9±39.4* | −41.7±48.7 | 9244574 | 8940468 | 3.3 |
| 16 wks | 3849.7±81.6 | −1035.7±46.4*** | 93.4±42.9* | −56.5±53.2 | 11003424 | 10643990 | 3.3 |
| 17 wks | 4008.1±88.5 | −1053.1±50.3*** | 104.0±46.6* | −36.9±57.7 | 12969202 | 12522910 | 3.4 |
Sex, additive, and dominance effects on ADG in each one week interval are shown in Table 2. As for the sex effect, there were significant differences in ADG between males and females from the 0-d-1-wk interval to the 13–14-wk interval. As for the additive effect, there was a significant difference between the A and C alleles in ADG at the 5–6-wk interval. For the other ADG traits, there were no significant differences between the A and C alleles, whereas the A allele tended to be superior in ADG traits, except for ADG at the 14–15-wk interval. In terms of the dominance effect, there were no significant differences between birds with the A and those with the C allele throughout the experimental period.
| Average daily gain (ADG, g) | Intercept | Sex | Additive | Dominance | Rvar | Fvar | Variance (%) |
|---|---|---|---|---|---|---|---|
| 0 d–1 wk | 10.6±0.3 | −1.0±0.2*** | 0.2±0.2 | 0.4±0.2 | 198 | 195 | 1.3 |
| 1–2 wks | 17.2±0.5 | −2.0±0.3*** | 0.5±0.2 | −0.1±0.3 | 359 | 350 | 2.4 |
| 2–3 wks | 24.5±0.8 | −4.4±0.5*** | 0.5±0.4 | 0.1±0.5 | 1018 | 1010 | 0.8 |
| 3–4 wks | 31.6±1.0 | −4.3±0.6*** | 0.3±0.5 | −0.3±0.6 | 1546 | 1541 | 0.3 |
| 4–5 wks | 34.6±1.5 | −10.5±0.9*** | 1.4±0.8 | −1.0±1.0 | 3816 | 3735 | 2.1 |
| 5–6 wks | 36.4±2.1 | −8.1±1.2*** | 2.2±1.1* | 0.0±1.3 | 7043 | 6840 | 2.9 |
| 6–7 wks | 46.1±1.8 | −12.0±1.0*** | 0.5±1.0 | −0.9±1.2 | 5252 | 5242 | 0.2 |
| 7–8 wks | 44.7±1.9 | −13.8±1.1*** | 0.7±1.0 | −1.5±1.2 | 5809 | 5787 | 0.4 |
| 8–9 wks | 43.6±2.1 | −13.3±1.2*** | 1.8±1.1 | −0.4±1.4 | 7013 | 6878 | 1.9 |
| 9–10 wks | 46.6±2.1 | −15.1±1.2*** | 1.9±1.1 | 0.0±1.4 | 7268 | 7114 | 2.1 |
| 10–11 wks | 47.0±2.5 | −17.1±1.4*** | 0.5±1.3 | 0.2±1.6 | 9895 | 9886 | 0.1 |
| 11–12 wks | 40.8±2.8 | −15.5±1.6*** | 1.6±1.5 | −0.2±1.8 | 12392 | 12288 | 0.8 |
| 12–13 wks | 32.7±2.6 | −11.3±1.5*** | 0.4±1.4 | 0.7±1.7 | 10544 | 10537 | 0.1 |
| 13–14 wks | 38.5±2.8 | −16.8±1.6*** | 1.9±1.5 | −1.9±1.8 | 12744 | 12601 | 1.1 |
| 14–15 wks | 25.3±3.5 | −1.1±2.0 | −2.2±1.9 | −1.0±2.3 | 19971 | 19772 | 1.0 |
| 15–16 wks | 23.5±3.8 | −1.5±2.2 | 1.1±2.0 | −2.1±−0.9 | 23270 | 23223 | 0.2 |
| 16–17 wks | 22.6±4.5 | −2.5±2.5 | 1.5±2.4 | 2.8±2.9 | 32002 | 31906 | 0.3 |
Sex, additive, and dominance effects on the weight of each carcass part at 17 wks are shown in Table 3. Data on the sex effects show that the weights of breast, thigh, tender, wings, bones, abdominal fat, heart, and gizzard in males were significantly higher than those in females. Data of the additive effects showed that birds with the A allele had heavier wings than those with the C allele. As for the dominance effect, there were no significant differences between the weights of carcass parts in birds with the A and those with the C allele.
| Carcass trait (g) | Intercept | Sex | Additive | Dominance | Rvar | Fvar | Variance (%) |
|---|---|---|---|---|---|---|---|
| Breast | 438.2±25.5 | −85.8±14.2*** | 20.0±13.2 | −18.1±16.4 | 218677 | 210968 | 3.5 |
| Thigh | 845.8±39.5 | −287.2±22.1*** | 30.6±20.4 | −35.8±25.4 | 525881 | 507780 | 3.4 |
| Tender | 122.5±6.8 | −20.4±3.8*** | 4.3±3.5 | 1.4±4.4 | 15437 | 15085 | 2.3 |
| Wings | 362.7±15.6 | −114.3±8.7*** | 17.9±8.1* | −8.4±10.1 | 85941 | 79732 | 7.2 |
| Bones | 1149.5±55.4 | −427.7±31.0*** | 34.7±28.7 | 1.6±35.7 | 1024296 | 1001015 | 2.3 |
| Abdominal fat | 115.5±25.2 | 38.0±14.1** | −6.5±13.0 | −26.4±16.2 | 207325 | 206512 | 0.4 |
| Liver | 55.9±5.6 | −3.6±3.1 | 2.4±2.9 | −1.5±3.6 | 10141 | 10032 | 1.1 |
| Heart | 20.6±1.3 | −9.2±0.7*** | −0.1±0.7 | −1.3±0.8 | 556 | 556 | 0.0 |
| Gizzard | 55.8±3.4 | −7.8±1.9*** | −1.8±1.8 | −0.2±2.2 | 3865 | 3799 | 1.7 |
Sex, additive, and dominance effects on the percentage weight of each carcass part at 17 wks are shown in Table 4. Data on the sex effect show that while the percentage weight of thigh, wings, bones, and heart was significantly higher in males than in females, the percentage weight of breast, tender, abdominal fat, liver, and gizzard was significantly lower in males than in females. Data on the additive effect showed that birds with the A allele had a smaller percentage weight of the gizzard than birds with the C allele. As for the dominance effect, there were no significant differences between the relative weights of all carcass parts in birds with the A and those with the C allele.
| Carcass trait (%) | Intercept | Sex | Additive | Dominance | Rvar | Fvar | Variance (%) |
|---|---|---|---|---|---|---|---|
| Breast | 10.8±0.6 | 1.20±0.31*** | 0.14±0.29 | −0.20±0.36 | 100.2 | 99.8 | 0.4 |
| Thigh | 20.8±0.6 | −1.76±0.34*** | 0.11±0.32 | −0.43±0.40 | 123.3 | 123.1 | 0.2 |
| Tender | 3.0±0.2 | 0.47±0.09*** | −0.02±0.08 | 0.15±0.10 | 7.9 | 7.9 | 0.1 |
| Wings | 8.9±0.3 | −0.40±0.17* | 0.18±0.16 | 0.08±0.20 | 31.2 | 30.6 | 1.9 |
| Bones | 27.9±1.2 | −3.53±0.69*** | 0.24±0.64 | 1.09±0.80 | 498.6 | 497.5 | 0.2 |
| Abdominal fat | 2.9±0.6 | 2.02±0.36*** | −0.27±0.33 | −0.70±0.42 | 137.7 | 136.4 | 1.0 |
| Liver | 1.4±0.1 | 0.38±0.08*** | 0.01±0.07 | 0.02±0.09 | 6.6 | 6.6 | 0.0 |
| Heart | 0.5±0.0 | −0.13±0.02*** | −0.02±0.02 | −0.02±0.02 | 0.3 | 0.3 | 2.4 |
| Gizzard | 1.4±0.1 | 0.22±0.05*** | −0.10±0.05* | 0.04±0.06 | 3.2 | 3.0 | 6.5 |
The significant association between CCKAR haplotypes and growth traits was first reported by Rikimaru et al. (2012), who used a resource population produced by crossing low- and high-growth lines of the Hinai-dori breed. Rikimaru et al. (2013) reported a significant difference in allele frequency between the low- and high-growth lines, suggesting that the difference was caused by long-term selection for growth performance. Moreover, Rikimaru et al. (2014) demonstrated that the A allele of the g.420 C>A SNP in CCKAR improves growth traits in the Hinai-dori breed. However, the Hinai-dori breed has not been directly used for commercial purpose; rather, it is used as sires for Hinai-jidori production. Therefore, this is, to our knowledge, the first study showing that the g.420 C>A SNP can be applied to a commercial Jidori chicken.
We cannot exclude that the associations detected in this study might have resulted from linkage disequilibrium between the SNP and other linked DNA polymorphisms directly involved in the regulation of growth traits. CCKAR is located in the distal region of chromosome 4. In this region, CCKAR SNPs other than g.420 C>A (Dunn et al., 2013) and SNPs in other candidate genes (Lyu et al., 2017) have been suggested to affect growth traits by studies on resource populations of White Leghorn×commercial broilers (Dunn et al., 2013) and New Hampshire×White Leghorn (Lyu et al., 2017), respectively. However, we firmly believe that g.420 C>A is a use fulmarker and is practicable for improving growth traits based on similar findings in genetically distinct populations, i.e., the Hinai-dori breed and Amakusa Daioh cross chickens.
The candidate gene CCKAR and its function merit further discussion. Cholecystokinin (CCK) is well known as a gut peptide implicated in the control of appetite (Gibbs et al., 1973). Two receptors for CCK have been identified: CCKAR (Sankaran et al., 1980) and CCK type B receptor (CCKBR) (Innis and Snyder, 1980). CCKAR is predominant in the alimentary tract and CCKBR is more common in the brain (Sankaran et al., 1980; Innis and Snyder, 1980). Based on known CCK functions, it has been suggested that CCKAR polymorphisms might affect appetite. For example, Dunn et al. (2013) reported that birds possessing the high-growth CCKAR haplotype had lower CCKAR mRNA levels and immunoreactive CCKAR in the brain, intestine, and exocrine organs, which correlated with increased levels of orexigenic AGRP (agouti-related protein) in the hypothalamus, and they were resistant to the anorectic effect of exogenously administered CCK, suggesting that their satiety set point had been altered. Meanwhile, Rikimaru et al. (2014) reported that there were no significant differences in feed intake among the three genotypes (A/A, A/C, and C/C) of the g.420 C>A SNP, and the feed conversion ratio between 4 and 10 wks in A/A individuals was significantly higher than that of C/C individuals in the Hinai-dori breed. We think that, depending on the CCKAR haplotype, there are differences in metabolic efficiency rather than in appetite, because the data of Rikimaru et al. (2014) were produced from within a breed, while the data of Dunn et al. (2013) were produced from an interbreed population, and thus, Dunn et al. (2013) might have simply detected a breed difference in CCKAR haplotypes. In addition, the correspondences of haplotypes reported by Dunn et al. (2013) and Rikimaru et al. (2012, 2013, 2014) have not been confirmed, because the haplotypes in Dunn's report were constructed from SNPs located in the downstream region of CCKAR and did not involve g.420 C>A. In any case, further study is needed to explain why CCKAR affects growth traits.
The reasons for the significant associations between the g.420 C>A SNP and gizzard percentage weight and wing weight are unknown. Collins et al. (2014) reported that the gizzard percentage of an old-breed broiler strain (Athens Canadian Random Bred) was higher than that of a modern broiler strain (Cobb 500) slaughtered at the same age. Therefore, the phenomenon observed in this study might simply reflect a negative correlation between live weight and gizzard percentage weight. Another possibility is the involvement of other genes affecting gizzard percentage weight. The osteoprotegerin (OPG) gene located in the proximal region on chromosome 2 has been reported to affect gizzard percentage (Fornari et al., 2014). CCK, CCKAR, and CCKBR have been listed in the gene network of OPG (Fornari et al., 2014). Meanwhile, CCKAR and the other genes located in the distal region on chromosome 4 might affect wing weight, as quantitative trait loci of this trait have been reported in this region (Ikeobi et al., 2004; Nassar et al., 2012).
In conclusion, this study revealed significant associations between a CCKAR SNP and growth traits in Amakusa Daioh cross chickens. As the parent stock lines of the Amakusa Daioh cross were maintained under restricted feeding conditions to avoid obesity and maintain reproductive performance, growth performance and meat productivity have not been tested in these lines. We will use the information obtained in this study for marker-assisted selection in the parent stock lines of the Amakusa Daioh cross, i.e., the restored Amakusa Daioh and the Kyushu Rhode breeds.
This work was financially supported by the Project of the NARO Bio-oriented Technology Research Advancement Institution (the special scheme project on regional developing strategy). We thank the staff of the Animal Husbandry Research Institute, Kumamoto Prefectural Agricultural Research Center (Koshi, Japan) for their kind assistance. The authors declare no conflict of interest.
