A total of 785 female Shaoxing ducks were randomly selected and raised in separate cages under similar environmental conditions and diet. The egg production traits of all ducks were recorded throughout the egg production process in terms of age at first egg, egg weight at 34 weeks old and total number of eggs at 34 weeks old and 72 weeks old. Egg collection for egg quality measurements took place at 34 weeks old. The following traits were recorded: egg shape index, shell thickness, shell strength, albumen height, Haugh unit score and eggshell color.

The blood samples of all the 785 individual ducks were collected from the wing vein using vacuum tubes containing dipotassium ethylene diamine tetraacetate (EDTA-K₂) as an anticoagulant for further SNP analysis. To investigate the expression of the multiple forms of melatonin receptor mRNA, 6 ducks representing two groups (3 of each) according to the age at first egg (at 128 days of age or after 150 days of age) were carefully selected at age of 150 days. Ovaries were sampled after slaughter. These experiments were conducted in accordance with the Law of the People’s Republic of China on Animal Protection.

Based on the complete DNA sequence of Anas platyrhynchos genomic DNA sequence (NCBI accession no. NW_004676748.1), four pairs of primers were designed to amplify the target regions for SNP genotyping using the Primer Premier 6.0 software. Primer sequences are listed in Table 1.

Polymerase chain reaction (PCR) was carried out in a total volume of 15 μL consisting of 1 μL genomic DNA, 0.15 μL (10 μM) of each primer, 1.5 μL 10× buffer, 1.5 μL (25 mmol/L) MgCl₂, 0.3 μL (10 mmol/L) dNTPs, and 1.5 U Taq DNA polymerase. PCR conditions were as follow: 95°C for 3 min, followed by 35 cycles of 95°C for 15 s, 55°C for 15 s, 72°C for 30 s, and a final extension of 72°C for 10 min. The PCR products were sequenced commercially using ABI 3730XL automated sequencer (Applied Biosystems, Carlsbad, CA, USA) after being purified and extended. SNPs were identified by looking for multiple peaks at the same base pair.

The genotypic frequencies were calculated for each individual and the Hardy-Weinberg equilibrium was analyzed using the Chi-square test of PopGene Version 1.32. Pairwise tests for linkage disequilibrium (LD) were performed for each SNP using the SHEsisPlus online software platform (http://shesisplus.bio-x.cn/SHEsis.html). The egg production traits were compared among the genotypes. The association between the SNPs and different traits in 785 ducks were analyzed using SPSS 22 with the model Y = μ+G+L+G×L+e, where Y is the dependent variable (analyzed traits), μ is the overall mean, G is the genotype with a variation for the candidate gene, L is the fixed effect of breed, G×L is the interaction between the genotype and duck population (a fixed effect), and e is the random error. The significance of the least squares means was tested with lest significance difference test.

Total RNA was extracted from ovary samples using Trizol reagent according to the manufacturer’s protocol. cDNA was synthesized from the total RNA according to the manufacturer’s protocol of TransScript First-Strand cDNA Synthesis SuperMix (TransGen, Beijing, China). Quantitative real-time PCR was performed using ABI 7500 (Applied Biosystems, Foster City, CA, USA). The mRNA expression levels were determined by Applied Biosystems real-time PCR using the MTNR1A, MTNR1B, and MTNR1C mRNA-specific primers and the β-acting gene as the internal control (Table 1). The 20 μL amplification reaction contained 10 μL of SYBR Green Universal PCR Master Mix, 2 μL cDNA, 0.4 μM of each primer and nuclease free water up to 20 μL. Thermal parameter used to amplify the template started with an initial denaturation at 94°C for 3 min followed by 40 cycles of 94°C for 10 s and annealing at 60°C for 30 s. The relative expression levels of the genes test were calculated using the 2^−ΔCt method. Groups with different letters were significantly different at p<0.05.

Seven SNPs were identified in all three melatonin receptor gene exons respectively by direct sequencing (Figure 1), their corresponding allele and genotype frequencies are presented in Table 2. Hardy-Weinberg equilibrium tests showed that the alleles of MTNR1A g. 268C>T, MTNR1C g. 10C>T, and MTNR1C g. 108C>T were in Hardy-Weinberg equilibrium (p>0.05), and others deviated from the Hardy-Weinberg equilibrium (p<0.05) (Table 1). The LD tests of Shaoxing duck population showed that two SNPs (g. 41C>T, g. 161T>C) identified in MTNR1B gene were completely linked (D′ = 1.0 and r² = 1.0) (Supplementary Table S1), providing only three possible haplotypes.

The association between the SNPs and egg production traits are shown in Table 3. For the MTNR1A g. 268C>T locus, age at first egg of the genotype CT was significantly less than that of the genotype TT (p<0.05). The g. 108C>T locus of MTNR1C was significantly associated with age at first egg and total number of eggs during 34-week egg-laying period (p<0.05). For MTNR1C g. 108C> T, the egg number of the genotype CT at 34 weeks old was significantly higher than that of the genotype CC (p<0.05), and ducks with the genotype TT exhibited significantly earlier age at first egg than that of the genotype CC (p<0.05). No significant association was found among other SNPs and egg weight, total number of eggs at 34 weeks old, total number of eggs at 72 weeks old or age at first egg (p>0.05).

As shown in Table 4, there were no significant association of 7 SNPs with egg quality trait. Furthermore, relation of SNPs with the color of eggshell was analyzed using chi-square test. When white eggshell was compared with blue eggshell, no significant differences in genotypes of MTNR1B or MTNR1C were observed except MTNR1A (p>0.05). The genotype TT and CT of g. 268C>T in MTNR1A were more common in Shaoxing ducks laying white eggshell compared with blue eggshell (Supplementary Table S2), suggesting MTNR1A g. 268C>T SNP may affect the eggshell color.

The expression of MTNR1A, MTNR1B, and MTNR1C between early-mature (first egg at 128 days of age) group and late-mature (first egg after 150 days of age) group showed different patterns (Figure 1). The late-mature group had significantly higher MTNR1A mRNA expression (0.0684±0.0017) compared to the early-mature group (0.0491±0.0060) (p<0.05) (Figure 2a), whereas there was no significant difference of MTNR1B mRNA expression between the early-mature group (1.491±0.4580) and the late-mature group (0.869±0.2435) (p> 0.05) (Figure 2b). The MTNR1C expression level of ovary was significantly higher in the early-mature group (2.4328±0.3936) than in late-mature group (p<0.05) (0.4756±0.0595) (Figure 2c).