Two Novel InDels within the 5’UTR of SIRT1 are Associated with Growth Traits in Chickens

Background: SIRT1, a NAD + dependent histone deacetylase, is involved in lipid metabolism, glucose metabolism, apoptosis, and insulin secretion. However, the function of the SIRT1 gene in chickens has not been elucidated. Results: In our study, we identied two novel InDels (c.-1552_-1553insCG and c.-450_-451delCG) in the 5’UTR of the chicken SIRT1 gene. After genotyping 1,141 chickens from 7 breeds, we found that the wild type genotypes for both sites were the most common. An association study using 860 chickens from a Gushi ×Anka F2 resource population showed that c.-1552_-1553insCG was signicantly correlated with growth traits and serum lipid indicators. The insertion genotype was most highly associated with body weight in 0-, 2-, and 4-week old chickens, and with shank length and shank circumference in 4-week and 8-week old chickens. The wild type genotype at this site was most highly associated with serum lipid indicators. In contrast, c.-450_-451delCG was signicantly correlated with muscle ber diameter. We also analyzed SIRT1 gene expression in chickens with different InDel genotypes and found that SIRT1 expression in muscle and fat tissue was signicantly higher with heterozygous genotypes at both sites, relative to expression in chickens with the corresponding homozygous genotypes. Finally, we analyzed the effects of different haplotypes on SIRT1 promoter activity. The results showed that promoter activity depends on haplotype, with haplotype HapII exhibiting the highest activity. Conclusion: We conclude that the SIRT1 gene is associated with chicken growth traits and that the two InDels inuence SIRT1 promoter activity in chickens. recessive black-bone CS, Changshun chicken; YF1, H-line of Yufen I layers; AA, AA broiler; Allele I, insertion or deletion allele; Allele II, wild type allele. P (HWE), P for Hardy–Weinberg equilibrium.


Background
Sirtuin 1 (SIRT1), a member of the sirtuin protein family, encodes a homolog to the yeast Silence information regulator 2 (Sir2) protein. Seven members of the sirtuin family have been identi ed, all of which share a highly conserved structure [1]. However, sirtuin family members differ in their subcellular locations, suggesting that their functions are not identical. While functional studies have not yet been completed for most family members, SIRT1 is well understood in mammals [2]. SIRT1, also known as NADdependent protein deacetylase sirtuin-1, coordinates several important functions such as the cell cycle, metabolism, and autophagy [3,4]. SIRT1 is also involved in the regulation of many physiological processes in animals, particularly lipid metabolism, glucose metabolism, and insulin secretion [5].
Previous studies have focused on SIRT1 as a regulator of lipid metabolism [6][7][8]. The results suggest that SIRT1 regulates fatty acid metabolism and also inhibits lipogenesis. SIRT1 increases fatty acid oxidation and inhibits fat production by activating the AMPK/LKB1 pathway in human hepatoma cells [9]. SIRT1 also deacetylates sterol regulatory element binding transcription factor 1 (encoded by SREBP-1c) in the mouse liver during fasting [10]. SIRT1-knockout mice exhibit liver steatosis and enlargement after feeding on a diet containing 11% fat, suggesting that changes in SIRT1 activity affect liver lipid deposition [11]. Li et al. found that SIRT1 activates LXR and then up-regulates the expression of Abca1, which is necessary for reverse cholesterol transportation and the synthesis of high-density lipoprotein [12]. Plasma cholesterol levels are also signi cantly reduced in SIRT1-de cient mice. SIRT1 not only affects adipogenesis in the liver, but also regulates adipogenesis in adipose tissue. In white adipose tissue, Picard et al. [13,14] found that SIRT1, in combination with NcoR and SMART, inhibits PPARγ activity, down-regulates the expression of genes related to fat synthesis, inhibits the differentiation of preadipocytes, promotes fat mobilization, reduces fat deposition, and provides energy. Moreover, SIRT1 deacetylates Foxo1, regulates triglyceride hydrolase (Atgl), and also regulates fat decomposition [15]. Overall, the evidence indicates that SIRT1 plays an important role in the regulation of lipid pathways, and that a SIRT1 gene de ciency might compromise a variety of critical cellular functions. SIRT1 has been investigated more extensively in mice and humans than in chickens. The chicken SIRT1 gene is located on chromosome 6 and consists of 9 exons and 8 introns, encoding 756 amino acids (CDS = 2271 bp, including the stop codon) (NCBI accession number: NM_001004767.1). The aims of this study were to identify SIRT1 gene variations, analyze their potential relationship to growth and carcass traits, and evaluate the genetic variations for utility as molecular tools for marker-assisted selection in the poultry industry.

Methods
Animals and phenotypes 860 chickens from the Gushi × Anka F2 resource population were used to assess biochemical phenotypes as described previously [16][17][18][19][20], which were raised in Henan poultry germplasm resources innovation engineering research center. Growth traits included body weight and shank length at 0, 2, 4, 6, 8, 10, and 12 weeks, and shank circumference at 4, 8, and 12 weeks. Carcass traits were collected after 860 F2 chickens were euthanized at the age of 12 weeks, including length, girth, width, and density of breast and leg muscle ber. All the chickens were injected intraperitoneal with 2 ml 5% Pentobarbital (No. 57330 of Chinese Academy of Sciences, Beijing Siyuan Technology Co., Ltd.). Carotid artery bleeding was done after 2 ~ 3 min of no respiration. Serum biochemical indicators included total serum cholesterol, triglyceride, alkaline phosphatase, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and glucose.
Blood samples were collected from the jugular vein after chickens were killed. Blood samples were also obtained from Xichuan black-bone chickens (XC), Lushi chickens (LS), Changshun chickens (CS), the Hline of Yufen I layers (H-line, YF1), recessive white chickens (RW), and AA broilers (AA). Genomic DNA was isolated from blood samples using phenol-chloroform and used for genotyping.
InDel discovery and genotyping SIRT1 polymorphisms were detected using pooled DNA sequencing. Two pairs of primers (two 5'UTR1 primers and two 5'UTR2 primers; see Table 1) were designed using the published DNA sequence (NCBI accession number: NM_001004767.1) to amplify and sequence regions within the 5' anking region of the chicken SIRT1 gene. Pooled samples represented groups of 50 chickens from the Gushi × Anka F2 resource population. The pools were generated by combining 80 ng of DNA from each chicken, which were used to identify the InDels. InDels in PCR products were identi ed based on the position of peak patterns in the sequencing chromatograms. For genotyping, PCR products containing the InDels were digested with DdeI (Takara, Dalian, China) and separated by electrophoresis through a 2.5% agarose gel. Genetic diversity was assessed using Hardy-Weinberg equilibrium (HWE), heterozygosity (He), effective allele numbers (Ne), polymorphism information content (PIC), and xation index. Data analysis was conducted using Genepop (v1.3.1) [21,22].

Association analysis between SIRT1 polymorphisms and phenotypes
Associations between the two SIRT1 InDel genotypes and growth, carcass, and serum biochemistry traits were evaluated with a general linear model using SPSS 24.0. A model was tted as follows: Where Y ijklm is the phenotype value of traits, µ is the overall mean, G i is the xed effect of InDels genotypes, S j , f l , and H k are the xed effects of gender, family, and hatch, and e ijklm is the random residual error e ~ N (0, δ e 2 ).

SIRT1 expression in chickens with different InDel genotypes
H-line chickens with different InDel genotypes were obtained from Henan poultry germplasm resources innovation engineering research center. Five chickens were randomly selected for each genotype. Sebum, abdominal fat, breast muscle, leg muscle, heart, liver, lungs, kidneys, and hypothalamus were harvested, rapidly frozen in liquid nitrogen, and stored at -80 °C. Total RNA was extracted with Trizol (Invitrogen, Waltham, MA, USA), and RNA quality was assessed by electrophoresis and spectrophotometry. cDNA was synthesized using an RNA reverse transcription Kit (Takara, Dalian, China), following the manufacturer's instructions.
SIRT1 gene expression in the 9 different tissues was analyzed by qPCR, using β-actin expression as a reference. SIRT1 expression in chickens with different genotypes was also analyzed. qPCR was conducted using TB Green Premix Ex Taq™ (Takara, Dalian, China) and a Roche LightCycler®96 thermocycler. Primers for qPCR were designed using the SIRT1 and β-actin mRNA sequences (Table 1).

SIRT1 promoter activity reporter plasmids and luciferase assay
Promoter fragments from the chicken SIRT1 gene were ampli ed using primers 1 (+), 2 (-), ID (+), and ID (-) containing NheI and HindIII sites. Fragments representing four haplotypes were ampli ed and cloned into the pGL3-basic plasmid vector. The resulting constructs were designated HapI, HapII, HapIII, and HapIV. All primers used are listed in Table 1.
The human embryonic kidney cell line (293T) was used to validate the activity of different haplotypes of the SIRT1 promoter. Cells were seeded into 24-well plates. After incubation for 24 hours, the cells were cotransfected with 800 ng recombinant SIRT1 promoter plasmid and 100 ng pRL-TK (Renilla luciferase control reporter vector) using Lipofectamine 2000 (Invitrogen, Waltham, MA, USA). Luciferase activity was detected 48 hours after transfection using the Dual-Glo Luciferase Assay System (Promega, USA). All assays were repeated three times with three culture replicates each.

Statistical analysis
Relative mRNA expression was calculated using the 2 −ΔΔCt method. Differences between groups were tested for signi cance by one-way ANOVA, with a post-hoc test (Duncan) at the 5% level. Results were expressed as mean ± S.E. GraphPad Prism 5 was used to visualize the data (GraphPad Software, San Diego, California USA).

Results
Identi cation of InDels in the SIRT1 promoter Two novel InDels were detected by sequencing the DNA from all chickens in the Gushi × Anka F2 resource population (860 individuals). Relative to the translation start site (coordinate 0), the two InDels were located at c.-1552_-1553insCG (a CG insertion) and c.-450_-451delCG (a GC deletion), as shown in Fig. 1.
Three genotypes were detected for each polymorphic site using polymerase chain reaction-restriction fragment length polymorphisms (PCR-RFLP) to generate markers. For the c.-1552_-1553insCG site, fragments of 136, 78, 32, and 26 bp were generated for the homozygous insertion; 136, 104, 78, 32, and 26 bp for the heterozygous (insertion/wild type) genotype, and 136, 104, and 32 bp products for the homozygous wild type genotype (Fig. 2a). For the c.-450_-451delCG site, fragments of 150 and 55 bp were generated for the homozygous deletion, 205, 150, and 55 bp for the heterozygous (deletion/wild type) genotype, and 205 bp for the homozygous wild type genotype (Fig. 2b).
Allele and genotype frequencies for the identi ed InDels Allele and genotype frequencies were calculated after genotyping the Gushi × Anka F2 resource population, four indigenous chicken breeds from Henan, and two commercial breeds (seven populations total; Finally, the frequency of the deletion genotype was quite low in 6 of the populations (F2, RW, LS, XC, CS and AA).   Effects of InDel genotypes on SIRT1 expression patterns qPCR analyses revealed that SIRT1 is expressed in many tissues. The highest expression levels were found in lung, liver, and heart, and the lowest expression was in sebum and abdominal fat (Fig. 3). To explore how SIRT1 gene expression is affected by InDels within the 5'UTR region, SIRT1 expression was analyzed in chickens with different genotypes from the H-Line of YuFen I layers. Signi cant differences in expression were observed in breast muscle, leg muscle, sebum, and abdominal fat (Figs. 4 and 5). For the c.-450_-451delCG site, heterozygous chickens showed the highest expression levels (relative to homozygous chickens) in muscle and fat tissues. No signi cant difference was found between the two homozygous genotypes in either breast muscle or abdominal fat. Similarly, for the c.-1552_-1553insCG site, the highest SIRT1 expression level was observed for the heterozygous genotype in muscle and fat tissues. However, chickens with a homozygous insertion genotype had higher levels of breast muscle, sebum, and abdominal fat than did chickens with the wild type genotype.

SIRT1 haplotypes and SIRT1 promoter activity
The InDel genotypes described above combine to form four haplotypes, designated HapI, HapII, HapIII and HapIV (Fig. 6). To determine if SIRT1 promoter activity was dependent on haplotype, we used a luciferasebased plasmid reporter assay and analyzed expression in transfected human embryonic kidney (293T) cells. Promoter activity was signi cantly higher with the HapII haplotype than for the other three Discussion SIRT1 is the most studied gene in the sirtuin family [23]. It is expressed in a wide variety of cell types and is involved in the regulation of biological processes including glycolipid metabolism, energy metabolism, and cell senescence [24]. SIRT1 regulates insulin sensitivity in liver cells by negatively regulating protein tyrosine phosphatase 1B [24], inhibits lipid-related gene transcription by deacetylating SREBP-1c [25], and reduces fatty liver formation by reducing fat deposition in liver cells. Although many studies have focused on SIRT1 in mammals, SIRT1 in chickens has rarely been investigated. Our study investigated the relationship between two InDels in the 5'UTR of the chicken SIRT1 gene and chicken growth traits, and explored the effect of the InDels on SIRT1 expression and promoter activity.
China has the largest number of native chicken breeds of any nation, but these native breeds typically do not perform as well as commercial lines. The genetic improvement of these breeds will require the development of new DNA markers to enable e cient selective breeding strategies. Using evidence from DNA pool sequencing and PCR-RFLP analyses, we detected two mutations (c.-1552_-1553insCG and c.-450_-451delCG) in the 5'UTR region of the SIRT1 gene. The most common genotype for the c.-1552_-1553insCG site in commercial layer lines and native chicken breeds was the wild type genotype, while the insertion genotype was more common in commercial broiler lines. The contrast suggests that these populations experienced different selection power during breeding. The results of our genetic differentiation analysis are consistent with this hypothesis. However, for the c.-450_-451delCG site, the high-frequency genotype was the wild type genotype. The deletion genotype was detected in only three populations (F2, RW and YF1), all of which were recombinant or synthetic lines. Further studies will be required to investigate the origin of this variation.
An association study was performed to examine the two InDels vs. growth and carcass traits in the F2 resource population. A clear correlation was observed between the c.-1552_-1553insCG locus vs. growth traits and some lipid indicator traits, consistent with the biological function of the SIRT1 gene [26][27][28].
Signi cant correlation was also found between the c.-450_-451delCG locus and muscle ber diameter. Muscle ber traits are useful quantitative indicators of meat tenderness. There is substantial agreement that muscle ber properties are determined by genetic factors [29][30][31][32]. Fulco et al. demonstrated that SIRT1 protein binds to the promoter region of skeletal muscle differentiation genes and induces the deacetylation of histone H3 residues [33]. Microarray analysis shows that SIRT1 overexpression speci cally inhibits the expression of MyoD and Myf5, both of which are marker genes for myocyte differentiation. Decreased endogenous SIRT1 increases the expression of marker genes for myocyte differentiation, whereas SIRT1 inhibitors enable the differentiation of skeletal muscle satellite cells [34]. In summary, various studies show that the SIRT1 gene is not only involved in lipid metabolism but also in muscle development, in agreement with the results of our association analysis.
Our tissue expression experiments show that SIRT1 expression in lung, liver, and heart is higher than in muscle and fat. These results are not consistent with previous reports. Ren et al. studied the tissue expression patterns for several sirtuin family members (SIRT1 ~ 7) in 30-week old Hy-line layers and found that the SIRT1 gene is expressed at the highest levels in liver and fat tissue [19]. A possible reason for the discrepancy is age. For layers, 30 weeks is the peak laying age, whereas the 75-week old chickens in our study were in the late laying stage, and had accumulated a large amount of abdominal fat. This may account for the different SIRT1 expression patterns.
We also studied SIRT1 expression patterns associated with the InDels within the 5'UTR region of SIRT1. The results show that SIRT1 expression differs signi cantly depending on InDel genotype. Combining these data with our association analysis results, we hypothesize that SIRT1 inhibits the growth of pectoral bers and participates in lipid transport. Furthermore, promoter activity assays demonstrate that promoters with the HapII genotype are 1.5 times more active than promoters with the other haplotypes tested. This suggests that the insertion mutation (c.-1552_-1553insCG) might signi cantly increase transcription of the SIRT1 gene.

Conclusions
In summary, we identi ed two novel InDels in the 5'UTR region of the SIRT1 gene and investigated the association between InDels and chicken growth traits. Our results reveal a novel molecular marker with potential application for chicken breeding, and establishes a foundation for further study of the SIRT1 gene in chickens.