High Expression of HMOX1 in Blue-Shelled Chickens is Associated with a TG Haplotype

HMOX1 is an important gene in biosynthesis of the eggshell pigment of blue eggs. Previous studies found that HMOX1 is highly expressed in the shell gland of hens laying blue eggs (BlueH) compared with hens laying brown eggs (BrownH); however, the reasons for the differential expression are unclear. In this study five single nucleotide polymorphism (SNP) in HMOX1 were genotyped in 111 BlueH and 115 BrownH. The association of haplotypes of these SNP with the blue egg phenotype was tested. Haplotype-specific expression of HMOX1 was detected in the shell gland. The interaction of sequence variants and transcription factors was analyzed using electrophoretic mobility shift assay (EMSA). A TG haplotype covering upstream 1.4 kb region of HMOX1 was significantly associated with blue eggs (p<0.05). Furthermore, the birds (n=12) with the haplotype expressed 3.8 fold more transcripts than those (n=12) without the haplotype (p<0.05). After re-sequencing a 2.2 kb region harboring the TG haplotype, a total of 26 SNP were found, of which a SNP was predicted to create a binding site of Nrf2, a transcription factor initiating HMOX1 expression. However, subsequent EMSA failed to confirm the Nrf2-DNA interaction. Taken together, the data suggested that the TG haplotype is not directly involved in regulation of HMOX1 expression; a regulatory mutation located near the haplotype and linked with the haplotype may exist and be responsible for the differential expression of HMOX1.


ABStRACt
HMOX1 is an important gene in biosynthesis of the eggshell pigment of blue eggs.Previous studies found that HMOX1 is highly expressed in the shell gland of hens laying blue eggs (BlueH) compared with hens laying brown eggs (BrownH); however, the reasons for the differential expression are unclear.In this study five single nucleotide polymorphism (SNP) in HMOX1 were genotyped in 111 BlueH and 115 BrownH.The association of haplotypes of these SNP with the blue egg phenotype was tested.Haplotype-specific expression of HMOX1 was detected in the shell gland.The interaction of sequence variants and transcription factors was analyzed using electrophoretic mobility shift assay (EMSA).A TG haplotype covering upstream 1.4 kb region of HMOX1 was significantly associated with blue eggs (p<0.05).Furthermore, the birds (n=12) with the haplotype expressed 3.8 fold more transcripts than those (n=12) without the haplotype (p<0.05).After re-sequencing a 2.2 kb region harboring the TG haplotype, a total of 26 SNP were found, of which a SNP was predicted to create a binding site of Nrf2, a transcription factor initiating HMOX1 expression.However, subsequent EMSA failed to confirm the Nrf2-DNA interaction.Taken together, the data suggested that the TG haplotype is not directly involved in regulation of HMOX1 expression; a regulatory mutation located near the haplotype and linked with the haplotype may exist and be responsible for the differential expression of HMOX1.

IntRoduCtIon
Birds are the only class of vertebrates that have evolved eggshell pigmentation (Cassey et al., 2010).The eggshell colors have been proposed to serve a wide variety of visual, physical, physiological, and behavioral functions, such as cryptic coloring, resisting harmful solar radiation, thermal protection, reinforcing eggshell strength, sexual selection, etc. (Reynolds et al., 2009).Thus, appearance of eggshell colors represents an important event of bird evolution.Blue-green coloration is a common eggshell color among birds.It is present not only in wild birds, but also in domestic fowls (i.e., several indigenous chicken breeds, such as Chilean Araucano, Chinese Dongxiang, and Lushi chickens) (Kennedy & Verves, 1973;Wang et al., 2013).The blue-green coloration is believed to function as a signal of female health to attract males involved in the supervision of the offspring (Morales et al., 2010).
Blue egg is colored by biliverdin, a bile pigment derived from the oxidative degradation of heme (Lang & Well, 1987).Two studies have confirmed that hens laying blue eggs (BlueH) have significant higher levels of biliverdin in eggshell and shell gland than those laying brown eggs (BrownH) (Zhao et al., 2006;Wang et al., 2009).However, molecular mechanisms underlying this difference are poorly understood.Blue-green eggshell color, also called as oocyan, is genetically controlled by a dominant mutant (O) (Punnett, 1933).In the latest study, Wang  (Hagenbuch & Gui 2008), it is speculated that formation of blue egg may be relevant to abnormal transportation of the eggshell pigment in shell gland (Wang et al., 2013).
In addition of SLCO1B3, two studies found higher HMOX1 expression in the shell glands of BlueH than of BrownH (Wang et al., 2010;Wang et al., 2011).Because heme oxygenase-1 (HO-1, protein encoded by HMOX1) is the rate-limiting enzyme catalyzing oxidative degradation of heme into biliverdin (Maines, 1997), the above result implied that HMOX1 can play a crucial role in biosynthesis of eggshell biliverdin.
The aim of this study was to further explore the regulatory mechanisms underlying high expression of HMOX1 in the shell glands of BlueH.The association of HMOX1 haplotype with blue egg phenotype was tested in the groups of BlueH and BrownH.Haplotype-specific expression of HMOX1 was detected in the shell glands.The effect of sequence variants on HMOX1 expression was studied using electrophoretic mobility shift assay.

Birds and Sample Collection
In this study, Dongxiang chickens were used to explore how the regulatory mechanisms of HMOX1 expression work.Dongxiang is a well-known blue egglaying breed from Dongxiang city, Jiangxi province, China.Although blue egg phenotype has been specifically selected in the Dongxiang chickens, the trait has not been fixed in the group, in which most birds lay blue-green eggs, but there are still minor individuals laying light brown eggs.A total of 111 BlueH and 115 BrownH were used for haplotype association analysis.All birds were provided by Hualv Dongxiang chicken protection farm, Jiangxi province, China.
Twelve birds with TG haplotype (TG_) and 12 birds without the haplotype (XX) were euthanized approximately three to five h prior to ovulation.Shell glands were collected and immediately stored in liquid nitrogen for subsequent expression analysis.Use and care of birds in this study was approved by the Northwest A&F University Ethics Committee.

SNP genotyping and haplotype analysis
Five SNPs were used to construct the HMOX1 haplotype.These SNPs, covering over 7 kb of the HMOX1 genome sequence, are sufficient to capture most of the variation in HMOX1, as assessed using the Tagger program of Haploview 4.2 (Broad Institute, Cambridge, MA).Information on the SNPs is summarized in Table 1.
Blood samples were collected by wing vein puncture.Genomic DNA was extracted from the blood using a standard phenol-chloroform method.The SNPs were genotyped by polymerase chain reaction-restriction fragment length polymorphism.Restriction enzymes were summarized in Table 1.All fragments were resolved on 2.5 % agarose gels or 12% nondenaturing polyacrylamide gels and visualized after an ethidium bromide or sliver staining.
All allele frequencies were tested for confirmation with the Hardy-Weinberg equilibrium.Construction and association analysis of the haplotype were completed using Haploview v4.2 under default values.Haplotypes with frequencies lower than 0.01 were discarded.In "sliding windows" analysis, 2-SNP, 3-SNP, 4-SNP window sizes were used to detect whether 3 The reverse primer for rs13866562 was modified to create a specific sequence of restriction enzyme of ApaL I.The modified base is indicated in bold.

High Expression of HMOX1 in Blue-Shelled Chickens is Associated with a TG Haplotype
some regions can carry stronger association with blue egg than others.
Real-time PCR was performed in an ABI PRISM 7900 instrument (Applied Biosystems Corp., Foster, CA).Amplification of 1 μL of cDNA was carried out in a total volume of 20 μL, according to the RealMasterMix (SYBR Green I) manual (Tiangen Corp., Beijing, China).Every sample was evaluated in triplicate.Data were analyzed by the 2^-ΔΔCT method (Livak & Schmittgen, 2001).The XX group was designed as calibrator.GAP-DH was used as endogenous reference to normalize the amount of cDNA added to the PCR.The amount of HMOX1 transcripts, normalized to GAPDH and relative to the calibrator, was given by 2^-ΔΔCT.The forward and reverse primers were 5'-ATCGCATGAAAACAGTC-CAG-3' and 5'-CAAATAAGCCCACGGCGAC-3' for HMOX1 and 5'-ATACACAGAGGACCAGGTTG-3' and 5'-AAACTCATTGTCATACCAGG-3' for GAPDH.

Electrophoretic Mobility Shift Assays (EMSA)
The two biotin-labeled oligonucleotides used in EMSA reaction were 5'-CGCGGCTCTCTGCTCTTCCTCCT-GCTGCCCCCGG-3' and 5'-CGCGGCTCTCTGCTCTA-ACTCCTGCTGCCCCCGG-3', corresponding to two alleles (shown in bold and underlined letters) of TC and AA at TSNP21, respectively.Nuclear proteins were extracted from the shell glands using the Nuclei Protein Extraction Kit (Viagene Biotech Inc, Beijing, China), following the manufacturer's protocol.Protein concentration was determined with the BCA Protein Assay Kit (Cwbio Co. Beijing, China).Five μg of nuclear proteins and 2.5 ng of biotin-labeled probes were incubated at room temperature for 20 min in binding mixture containing 1.5 μL of 10 × binding buffer and 1 μg of poly(dI)•poly(dC) in a total volume of 15 μL.In antibody supershift assays, 2 μg of Anti-Nrf2 rabbit polyclonal antibody (Abcam Co., Shanghai, China) were added to the reaction mixture and incubated for 20 min at room temperature prior to electrophoresis.In competition assays, firstly a 100-fold molar excess of unlabeled oligonucleotides was add-ed to the binding mixture, then incubated for 30 min at room temperature.Secondly, labeled probes were added and subjected to additional 20 min incubation at room temperature.DNA-protein complexes were resolved on a 5 % nondenaturing polyacrylamide gel in 0.5 × TBE at 80 V for 2-3 h.After electrophoresis, the complexes were transferred to a membrane and detected with horseradish peroxidase-conjugated streptavidin.Finally, blots were visualized using EMSA Chemiluminescent system (Viagene Biotech Inc, Beijing, China).

Statistical analysis
For comparison of HMOX1 expression between birds with TG haplotype and those without the haplotype, the null hypothesis that there was no difference between two groups was tested by unpaired t test.Statistical significance was assumed at p < 0.05.All statistical analyses were done by SAS ® V8.2 statistical software package (PC-SAS) (Cary, NC, USA).

Association of HMOX1 haplotypes with blue egg
In order to elucidate whether there are sequence variants responsible for high expression of HMOX1 in shell gland of BlueH, the association of HMOX1 haplotypes, consisting of 5 SNPs, with blue egg phenotype was analyzed.Five major haplotypes were observed in BlueH and BrownH, of which only the GAGTG haplotype showed a significant association with blue eggs (p<0.05,Table 2).In order to explore whether some regions of HMOX1 can carry a stronger association, 4-, 3-and 2-SNP "sliding window" analysis were performed.As the "sliding window" size was reduced, three haplotypes (TGGT, TGG, and TG) showed increasing association with blue egg, with the TG haplotype showing the strongest association with blue eggs (Table 3).  2 BrownH = hens laying brown eggs, BlueH = hens laying blue eggs

Haplotype-specific expression analysis
The TG haplotype covers upstream 1.4 kb region of HMOX1, which is traditionally a key region for the regulation of gene expression.A haplotype-specific expression assay was performed using hens birds (n=12) with the TG haplotype (TG_) and hens (n=12) without the haplotype (XX) to investigate the role of the haplotype in the regulation of HMOX1 expression.The expression data showed that the TG_ group expressed 3.8-fold more transcripts than the XX group in shell gland (p<0.05, Figure 1).

Figure 1 -Haplotype-specific expression analysis in shell gland
TG_ group consisted of birds (n=12) with the TG haplotype, correspondingly XX for ones (n=12) without the haplotype.The results were presented as mean folds of HMOX1 expression relative to the XX group.The TG group expressed approximate 3.8 fold more transcripts than the XX group (* p<0.05).

Effects of sequence variants within the TG-covering region on transcription factor-DNA interactions
A 2.2 kb region of Chr1:51954712-51956937, harboring the TG haplotype, was re-sequenced using 10 BlueH and 10 BrownH to identify crucial regulatory variants.A total of 26 SNPs were found in the region, most of which showed significant association with blue eggs (Table 4).After performing a prediction of transcription factor binding sites (TFBS) using CONSITE software (http://consite.genereg.net/)(LENHARD et al. 2003), 11 SNPs of these SNPs were found to create a putative TFBS (Table 4), with special attention paid to TSNP21 because the blue egg-associated TC allele of the TSNP21 can create a putative TFBS of Nrf2.It has been well established that Nrf2 is a transcription factor initiating HMOX1 expression under oxidative stress (Campbell et al., 2013).
An EMSA assay, therefore, was performed, using nuclear extracts from the shell gland and probes corresponding to TC and AA alleles in order to investigate the authenticity of Nrf2-mediated expression.Similar shifted bands were detected in lanes 2 and 6 with TC and AA probes, suggesting that there was no obvious difference in transcription factorbinding activity between two TSNP21 alleles (Figure 2).After performing a supershift assay with anti-Nrf2 antibody, we also failed to observe supershifted bands in lanes 3 and 7 (Figure 2), which, therefore, implied that the protein-DNA complexes in lanes 2 and 6 may be produced by the binding of other proteins, instead of Nrf2, to the oligonucleotides.Nrf2 is a well-established transcription factor initiating HMOX1 expression under oxidative stress.A putative Nrf2-binding element was created by the TC allele of the TSNP21 locus.EMSA was performed to validate the protein-DNA interaction using nuclear extracts from the shell gland and two probes corresponding to TC and AA alleles of TSNP21.The positions of free probe and shifted DNA-protein complexes are indicated with arrows.In lanes 1 and 5, no shifted band was observed when no nuclear extracts added; in lanes 2 and 6, the same shifted bands were observed using two probes, indicating that there was no obvious difference in transcription factor-binding activity between two alleles; in lanes 3 and 7, the shifted bands were not supershifted by Nrf2 antibody, indicating that the observed complexes in lanes 3 and 7 can consist of other protein, instead of Nrf2, and oligonucleotides; in lane 4 and 8, the shifted bands can be competed out by 100 × molar excess of unlabelled probes (cold probe), confirming that the complex consists of two target probes and unknown proteins.

High Expression of HMOX1 in Blue-Shelled Chickens is Associated with a TG Haplotype
of cis-regulatory elements, such as ARE, HeRE, HSE, CdRE, MARE, MPRE, and (GT)n repeats, were found to be involved in regulation of HMOX1 expression in human, rat and chicken (Martin et al., 2004;Chang et al., 2004;Shibahara et al., 1989;Shibahara et al., 2003;Shan et al., 2004).Thus, it is conceivable that any variants in these cis-elements may be responsible for the differential expression as well.
In the case of trans-acting factors three MAPK pathways of ERK, JNK, and p38 are known to be involved in up-regulation of HO-1 expression (Martin et al., 2004).In addition, trans-acting protein BACH1 is a repressor of Nrf2-mediated HO-1 expression.Its inactivation is a prerequisite for HO-1 induction (Reichard et al., 2007).Thus, any function or expression alterations in these trans-regulatory factors can also become the reasons for the differential expression of HMOX1.
Chicken oviduct is a highly vascular organ.When oocytes leave the ovary and enter the oviduct, the metabolism of the oviduct is stimulated into activity; blood flow through it increases; albumen is largely synthesized and secreted (Kennedy & Ververs, 1973;Baird et al. 1975).Following the process, a large amount of reactive oxygen species will be produced, causing deleterious effects on cellular proteins, lipids, and DNA (Alonso-Alvarez et al., 2004).HO-1 is a crucial protein in resisting oxidative injury (Calay & Mason, 2014).Thus, its high expression in BlueH may imply that appearance of blue egg is a result of bird adaptive evolution in the defense against oxidative stress during lay.
Based on previous studies of HMOX1, this work attempted further to reveal the reasons for the high expression of the gene in BlueH.Here, a TG haplotype was found to represent high expression activity.Because the haplotype appears more frequently (0.099) in BlueH than in BrownH (0.029), the data explained the reasons for high expression of HMOX1 to some extent.However, given the complex regulatory mechanisms of HMOX1, it remains an enormous challenge to define which of the variants is responsible for its differential expression.It, therefore, is necessary further to explore functional variants in an extended region and several trans-acting factors in the future.

ACknowledgeMentS
This study was supported by the National Natural Science Foundation of China (Grant No. 31401051) and China Postdoctoral Science Foundation (2014M550510).

High Expression of HMOX1 in Blue-Shelled Chickens is Associated with a TG Haplotype et
al. (2013) identified the O as a retrovirus (EAV-HP) insertion in SLCO1B3.Because OATP1B3 (gene protein) is a membrane transporter that can mediate transportation of biliverdin

Table 2 -
Frequencies of haplotypes constructed by the five SNPs

Table 3 -
Results of 2, 3 and 4-SNP "sliding window" analysis Only haplotypes significantly associated with blue egg are shown in the table. 1