Transcriptome analysis revealed the effect of selenium yeast on improving eggshell quality in aged laying hens

Egg internal and eggshell quality were deteriorated in aging laying hens. Improving egg and shell quality can prolong the laying cycle. Although, selenium yeast (SY) has been found with the potential to enhance the laying performance and egg quality, the underlying mechanisms have not been investigated. Therefore, we investigated the effect and molecule mechanism of selenium yeast on egg and shell quality in aged laying hens. Three hundred 76-week-old Jing Hong laying hens were divided into four equal treatments and fed with Se-decient diet (SD), 0.15, 0.30, 0.45 mg/kg selenium yeast diet (SY0.15, SY0.30 and SY0.45). At the end of the trial, we evaluated the plasma and tissue selenium content, plasma antioxidant capacity and egg quality. Transcriptomic analysis was performed to investigate the boosting effect of selenium yeast on eggshell quality. The weighted gene co-expression network analysis (WGCNA) analysis was performed to screen out the key candidate genes related with eggshell quality affected by selenium yeast.


Abstract
Background Egg internal and eggshell quality were deteriorated in aging laying hens. Improving egg and shell quality can prolong the laying cycle. Although, selenium yeast (SY) has been found with the potential to enhance the laying performance and egg quality, the underlying mechanisms have not been investigated. Therefore, we investigated the effect and molecule mechanism of selenium yeast on egg and shell quality in aged laying hens.

Methods
Three hundred 76-week-old Jing Hong laying hens were divided into four equal treatments and fed with Se-de cient diet (SD), 0.15, 0.30, 0.45 mg/kg selenium yeast diet (SY0.15, SY0. 30 and SY0.45). At the end of the trial, we evaluated the plasma and tissue selenium content, plasma antioxidant capacity and egg quality. Transcriptomic analysis was performed to investigate the boosting effect of selenium yeast on eggshell quality. The weighted gene co-expression network analysis (WGCNA) analysis was performed to screen out the key candidate genes related with eggshell quality affected by selenium yeast.

Conclusion
Selenium yeast enhanced eggshell strength and reduced the shell translucence by improving antioxidant capacity, selenium status, and regulating the processes of shell mineralization, ion transport and laying period. These ndings provide a novel molecule biomarker for affecting eggshell quality regulated by selenium.
Background Page 3/24 Selenium (Se) is a pivotal micronutrient for animals and humans, and plays vital roles in antioxidant and redox state regulation, immune response, reproductive and multifarious speci c metabolic processes [1].
In last few decades, worldwide researchers and nutritionists have studied this essential nutrient with pleiotropic aspects. In general, the function of Se was executed as selenoprotein, which is formed by selenocysteine (Sec) inserted [2]. To date, 25 human selenoprotein genes [3] and 24 con rmed avian selenoprotein genes [4] have been identi ed.
The current progress in the health impact of upregulation of antioxidant selenoenzymes and cytoprotective properties mediated by Se has been well-known. Dietary Se intakes can affect the synthesis and expression of selenoproteins to modulate de cient or adequate status of Se [5], and appropriate Se status leads to health bene ts, such as decreasing mortality and enhanced immune function. Moreover, early research demonstrated that both Se de ciency and excess can cause adverse health effects which accompanied by a U-shaped relationship. Impairments in antioxidant protection, redox regulation, immunity response, and energy production are associated with Se deprivation [6]. Se de ciency may also cause physiological diseases and economic loss in animals and livestock. In chicken, it has been demonstrated that Se de ciency impaired growth performance [7] and led to exudative diathesis [8], nutritional muscular dystrophy [9,10]. Alarmingly, extra Se supplementation may also have adverse effects on people or animals in adequate Se status, which has been expounded as selenosis. It causes hair and ngernails brittleness, skin and liver damage, and neurotoxicity in livestock [11].
Maintaining egg quality and production performance in aged laying hens for longer laying cycles has aroused an increasing concern in poultry production. In general, the production peak of commercial laying hens usually lasts until 60 weeks, the laying rate and egg quality begin to decline from the middle of production peak period [12]. Aging of hens leads to a deterioration in internal and shell quality of eggs, which is likely to cause the breakage during collection and transport [13]. Thus, overcoming the negative effects of age on egg quality is importance in the late phase of laying cycle. Some previous studies have suggested that the decline in egg quality is associated with nutrients status and oxidant stress in aged laying hens [13,14]. Nowadays as a food additive, Se is widely used for enhancing the antioxidants, immunity, health to obtaining Se-enriched byproducts in pigs, chickens, and other livestock. As the organic form of Se, selenium yeast was used frequently because of its characteristics of being wellabsorbed and biological safety [15,16]. Substantial studies on boiler and laying hens have shown that selenium yeast supplementation maintained poultry health [17], decreased the egg shape index [18], affected shell thickness [19], enhanced antioxidant status and tissue Se deposition [20]. Meanwhile, previous studies have demonstrated that dietary supplementation selenium yeast can improve the productive performance in aging broiler breeder hens [21]. However, the molecule mechanisms of selenium yeast on egg and shell quality in aged laying hens are poorly understand. In this study, the effects of selenium yeast on egg and shell quality, the optimum dose of selenium yeast in the diet of aged laying hens were investigated. Moreover, the potential molecular mechanism of selenium yeast on eggshell quality was explored by using RNA-seq based global transcriptome analysis.

Materials And Methods
Animals, experimental design, diets, and husbandry A total 300 76-week-age Jing Hong laying hens were fed with an average basal Se content of 0.056 mg/kg corn-soybean diet (the composition and nutrient level of the basal diet were shown in Table S1) for 6 weeks (from 76 to 82 week of age) to investigate the effects of Se de cient on aged laying hens and gain the Se-de cient hens. After Se-consumption Jing Hong laying hens with similar laying rate were randomly allocated to 4 treatment groups with 5 replicates (15 chickens per replicate). One group was fed the basal diet only (SD), the remaining groups were supplemented with 0.15, 0.30, 0.45mg/kg selenium yeast (SY) (Alltech, Lexington, KY, USA) for 12 weeks (from 83 to 95 week of age). The Se contents of different diets were shown in Table 1. Batches of the experimental diets were produced every 4 weeks to prevent the feed from mildewing. The hens were housed in an environmentally controlled room maintained at 25℃ and had a daily lighting schedule of 16 h light (from 5 am to 9 pm), 8 h dark (from 9 pm to 5 am). Measurement of egg quality and translucent egg Egg quality was measured both before and after the Se consumption period. In supplementation period, egg quality was measured at 6-week intervals, half of the eggs from each replicate were selected randomly. Eggshell strength, egg weight, egg yolk color, Haugh unit, and albumen height were tested using a digital egg tester (NABEL, DET-6000). Eggshell thickness was measured at the large end, equatorial region, and small end using an Eggshell Thickness Gauge. In the supplementation period, another half of eggs from each replicate were selected for the measurement of the translucent egg at 6week intervals. The eggs were stored for 0, 7, 14 days at same environment (20℃) and classi ed into four score levels, and the scoring method was referenced in this study [22].

Determination of Se content
To determine the Se content of the feed, plasma, ovary, magnum, isthmus, and shell gland. All samples (0.5 g-1 g for feed, ovary, magnum, isthmus, and shell gland and 0.5 ml-1 ml for plasma) were digested in a mixture of HNO 3 and HClO 4 (2:1) for about 2 hours, the mixture was heated at 200℃ until white fumes appeared. Then 5 mL hydrochloric acid solution was added in the mixture and heated until white fumes appeared. After the mixture cooled, 20 mL Ethylene diaminetetraacetic acid (EDTA) was added in the digested sample and adjusted the pH at 1.5-2.0. After which, 3 mL 2,3 DiAminoNaphthalene (DAN) was added in the mixture and heated in the boiled water for 5 minutes. After the mixture cooled, 4 mL cyclohexane was added in the mixture and shook for 8 minutes, then the supernatant was measured by uorescence method using Hitachi 850 uorescence spectrophotometer (Tokyo, Japan) [23].

Measurement of antioxidant enzyme activity
The activity of antioxidant enzymes in plasma can re ect the redox state of the body. The activities of glucokinase and glutathione peroxidase (GSH-Px) and total superoxide dismutase (T-SOD) in plasma samples were determined using the detection kits of GSH-Px and T-SOD (Nanjing JianCheng Bioengineering Institute, Nanjing, China), The activities of total anti-oxidation capacity (T-AOC) in plasma samples were measured by T-AOC Kit (KeyGEN BioTECH, Nanjing, China) following manufacturer's protocol.

Hematoxylin-eosin staining
For histological assessment, the shell gland was xed in 4% paraformaldehyde and embedded in para n. Three-micrometer sections were stained with hematoxylin and eosin (H&E). The entire images were captured by APERIO CS2 (Leica, Germany).
RNA extraction, library preparation, sequencing and RNA-Seq data analyses

Transcriptome Bioinformatic and statistical analysis
Raw transcriptome reads were quality controlled through ltering out low-quality and adaptor sequences, then we got clean data for further downstream analysis. Clean reads were mapped using Hisat2 to the reference genome downloaded from Ensembl (https://asia.ensembl.org/index.html). Quanti cation of gene expression level was conducted by FPKM (fragments per kilobase of exon per million parts mapping) using the featureCounts package, and differentially expressed genes (DEGs) were accessed using DESeq2 R package. The DEGs were identi ed with the following parameters: P-values < 0.05 and the fold change value |log2Ratio| ≥1. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were further analyzed in website DAVID (https://david.ncifcrf.gov/). Short Time-series Expression Miner (STEM) was performed (Default parameters) to investigate the dynamic changes in gene expression supplemented by different dose of selenium yeast in shell gland. Weighted gene co-expression network analysis (WGCNA) was performed to identi ed candidate genes affect eggshell quality regulated by selenium yeast. Pearson correlation (P < 0.05 and |R| > 0.4) was constructed the relationship between modules and Se content. Common genes in signi cant modules and DEGs were performed to GO enrichment analysis, Sankeycharts exhibited the common pathways between above pathways and DEGs pathways to investigate the process affecting eggshell quality by selenium. The genes in common pathways were performed for Protein-Protein Interaction (PPI) analysis on STRING (https://string-db.org/) website (con dence = 0.7).

Statistical analysis
Data was presented as mean ± standard error of the mean (SEM), data was analyzed with one-way ANOVA, followed by Duncan test (SPSS for Windows, version 25; IBM). Statistical differences were considered signi cant at P 0.05. Phenotypic data presentation was carried out using GraphPad Prism (version 7.0, GraphPad Software Inc, San Diego, CA, USA). Every experiment was repeated at least three times.

Results
The effect of selenium yeast intakes on plasma Se status and antioxidant enzyme activity It is well recognized that plasma and serum selenium can re ect the Se utilization status of body [24], and the decreasing expression of systemic selenoproteins caused by dietary Se de cient led to oxidative stress and oxidative stress-related diseases [25]. In our study, the plasma Se status and antioxidant enzyme activity was measured in aged laying hens during the whole experimental period (Fig. 1). During the consumption period (0-6 weeks), the plasma Se content was signi cantly declined from 0.237 µg/mL to 0.068 µg/mL (0-4 weeks) (P 0.01) and then remained a stable low status (4-6 weeks) (Fig. 1a). Moreover, redox state in plasma was altered by dietary Se de cient. There were concomitant decreases in plasma T-AOC, GSH-Px and T-SOD activity in the 2-week Se consumption (P 0.01, P 0.01 and P 0.05, respectively), and then remained relatively stable. Plasma T-AOC activity decreased from 3.716 U/mL to 1.875 U/mL (Fig. 1b). The GSH-Px activity was decreased from 3550.145 U/mL to 1480.000 U/mL (Fig. 1c). Moreover, it led to a decrease from 270.607 U/mL to 225.369 U/mL in plasma T-SOD activity (Fig. 1d). In the supplementation period (6-18 weeks), Se content increased drastically (P 0.01) with a peak around 6-week supplementation and then reached a plateau in supplementation groups (Fig. 1a). Interestingly, the plasma Se content in SY0.45 group recovered to the level before consumption after 8-week supplementation, while other treatments didn't. Similarly, the selenium yeast supplementation promoted antioxidant capacity in plasma. There was no signi cant differences in 4week supplementation in the T-AOC activity detection, but it was higher in the SY groups than that in the SD group after 8-week supplementation (P 0.01) (Fig. 1b). The GSH-Px (Fig. 1c) and T-SOD (Fig. 1d) activity in plasma after 12-week supplementation were signi cantly higher (P 0.05) in the SY0.45 group than the SD group.
Selenium yeast de ciency and supplementation signi cantly affected egg quality and body Se status In the late stage of laying period, changes of nutrients intake have a great impact on laying performance. Some adverse health conditions infertility and reproduction were associated with Se de ciency. Consistent with our studies, Se de ciency reduced the egg quality. The changes in the egg quality during the consumption period were shown in Table 2. With the increasing week of age and Se consumption, egg weight had an increasing trend (P = 0.06), however, eggshell strength was signi cantly reduced (P = 0.04), albumen height, Haugh unit, and yolk color were signi cantly reduced (P 0.01) after 6-week Se consumption. During the supplementation period, eggshell strength was signi cantly higher in SY0.45 group than that in SD group in 6-week and 12-week supplementation and that in the SY0.15 group in 12week supplementation (P 0.05) (Table S2 and Fig. 2a). Eggs randomly collected after 6 and 12-week supplementation were classi ed 4 grades according to the degree of shell translucence ( Fig. 2b and Fig.  S1). Supplementation of selenium yeast could reduce the shell translucence no matter how long time eggs stored and the age of laying hens. In addition, selenium yeast supplementation decreased the egg weight, albumen height, and Haugh unit after 12-week supplementation (Table S2). Furthermore, tissue Se status associated with laying performance including shell gland (Fig. 2c), eggshell (Fig. 2d), magnum (Fig. 2e), isthmus (Fig. 2f) and ovary (Fig. 2g) were detected. The results showed that selenium yeast feeding induced a dose-dependent increase in the body Se contents. The dose range which is bene cial for humans and animals is fairly narrow and has been described as a U-shaped curve.
In this experiment, a low feeding level was choose to measure whether the shell gland of aged laying hens was in toxicity state by H E staining. The results showed that there was no degeneration and necrocytosis of the mucosal epithelium cell layer, glands edema or dissolved in the uterus in each group (Fig. 2h). The results indicated that 0.45 mg/kg selenium yeast supplementation for 12 weeks couldn't cause pathological lesion of aged laying hens.
Dynamic effects of selenium yeast supplementation on gene expression in the shell gland A total of 1308 DEGs were assessed in the SY groups compared with the SD group (fold change 1.00, P 0.05) ( Table S3). The number of DEGs according to up-regulation or down-regulation were displayed in Fig. 3a. In total, 442 DEGs were shared among three groups, and 296, 144, 130 speci c DEGs were harbored in the SY0.15, SY0.30 and SY0.45 group, respectively (Fig. 3b). The biological functions of these DEGs were classi ed by GO term and KEGG pathway enrichment analysis (Fig. 3c, Fig. 3d and Table S4) Transcriptome analysis revealed the potential effects on enhancing eggshell quality supplemented with selenium yeast Egg quality traits are important for laying hens, especially in the late laying period. In our study, eggshell strength in the SY0.45 group was signi cantly higher than that in SD and SY0.15 group. Thus, we used RNA-seq and bioinformatics to explore the potential key genes associated with eggshell strength after selenium yeast supplementation. The Venn diagram showed DEGs from comparisons of SD vs SY 0.45 and SY0.15 vs SY0.45 (Fig. 4a), 78 DEGs were shared in both two comparisons including CEMIP, SDC3, SLC6A17, OVAL, PER2 and PENK. 73 DEGs were unique in the SY0.15 vs SY0.45 including SLC13A5, POMC and CYFIP2, as well as 697 DEGs were unique in the SD vs SY0.45 including EREG, OTOP2, KCNJ2, WNT11, STC2, PTN, SLC26A9, XAF1 and CA2. Functional enrichment analysis results showed that all DEGs were enriched in cell cycle process (regulation of cell proliferation, p53 signaling pathway, cell cycle, and apoptosis), follicle development (regulation of embryonic development, parturition, progesterone-mediated oocyte maturation, and response to estrogen), eggshell mineralization (calcium signaling pathway, glycosaminoglycan biosynthesis, negative regulation of BMP signaling pathway, and cellular zinc ion homeostasis), metabolism (Glutathione metabolism) ( Fig. 4b and 4c). It is interesting to note that trend analysis was performed to reveal the expression patterns of candidate DEGs related to eggshell strength at different selenium yeast dose. Total DEGs were clustered into 8 pro les, of which two signi cant trend pro les (P 0.05), including pro les 1 and 6 (colored block) (Fig. 4d). The expression of 158 genes including CA2, SLC13A5, and OTOP2 decreased with higher dose of Se in pro le 1. In pro le 6, the expression of 123 genes including PENK, WNT7A, and EREG exhibited an obvious increase with the higher dose of Se.
Network analysis is useful to identify genes that are putative hubs of gene co-regulation. Here, we used WGCNA to explore hub genes related to eggshell strength affected by selenium yeast intake. Due to the eggs we collected were not from corresponding laying hens, linear correlation was used to analyze the relationship between shell gland Se content with eggshell Se content and eggshell strength (Fig. 4e), Se content in shell gland had a positive relationship with egg strength (P = 0.0013) and eggshell Se content (P 0.01). The results of WGCNA analysis showed that these genes were segmented into 3 signi cant modules, and for each of these modules, the correlation of the eigengene with shell gland Se content was computed (Fig. 4f). The expression of gene in the Memagenta and MEgreen module was positively correlated with the shell gland Se content, while it in MEpurple was negatively correlated (Table S5). Common genes both in signi cant modules and DEGs were selected to explain the effect of selenium yeast on eggshell strength. Functional enrichment analysis showed that common pathways were implicated in the regulation of cell cycle, cell migration, eggshell mineralization, reproductive organ development and shell gland development (Fig. 4g). The PPI result of genes in common GO pathways showed that candidate genes in previous results including PTN, SDC1, WNT11 and PENK (Fig. 4h).

Discussion
Avian eggshell is made of columnar calcite crystals which protect the eggs from physical damage and microbial contamination, providing calcium sources for developing embryo [26]. Nevertheless, eggshell quality becomes to deteriorate from the middle phase of production with the characteristics of decreasing eggshell strength, increasing ratio of translucent eggs, and increasing numbers of abnormal eggs, which cause substantial economic losses and impaired animal welfare. Numerous studies have identi ed that Se is bene cial for health, egg and shell quality of laying hens. However, the potential molecular mechanism of selenium yeast on shell quality has not been elucidated.
Se de ciency and supplementation are closely related to Se status and antioxidants capacity. In general, plasma Se content and antioxidant enzyme were generally considered as useful biomarkers of both Se status and dietary intake [27]. Antioxidant enzyme including T-AOC and T-SOD acted as the important index for redox state in livestock and poultry. GSH-Px is a Se-dependent enzyme which has abilities for protecting tissues from oxidative damage. In our study, Se de ciency for 6 weeks leads to a signi cant decrease of Se status and antioxidants capacity in aged laying hens. Consistent with other studies, Se de ciency led to a signi cant decrease of Se content, GSH-Px [28], T-AOC [29], and SOD [30] activities in plasma in pigs and poultry. In addition, we found that 71.31% of Se content, 49.60% of T-AOC, 58.31% of GSH-Px, and 16.71% of T-SOD were decreased by Se de cient, respectively. Based on the decline degree of these indicators, Se content in plasma decreased quickly, suggesting that it was more sensitive to Se status, and could be a biomarker of Se status. Selenium yeast supplementation also enhanced the Se status and antioxidant capacity. Consistent with other studies, Se content in plasma and tissues increased with a dose-dependent trends after Se supplementation [31,32]. After 6-week supplementation, plasma Se content in the SY0.45 group was recovered, suggesting that high dose Se supplementation may relieve the harmful effects caused by Se de ciency. Moreover, the effects of Se deposition in different tissues may be different because of complexity of Se absorption and metabolism [33]. We rstly found that the Se deposition hierarchy in tissues associated with laying are isthmus, magnum, ovary, and shell gland, successively. Meanwhile, the activities of GSH-Px, T-AOC, and T-SOD in plasma increased in supplementation period. The data are in agreement with the results previously published that antioxidant enzyme activities dose-depended increased with selenium yeast supplementation [32,34,35]. These ndings suggested that Se content and plasma antioxidant enzyme activity can re ect the Se status and intakes, and plasma Se content can be used as an indicator of body Se status.
Studies clearly indicated that selenium yeast has crucial roles in poultry nutrition and productive. Eggshell strength is vital in ensuring the integrity and safety of the egg contents, it tends to deteriorate with the increasing of bird age [36]. In our study, Se de ciency for 6 weeks signi cantly decreased egg quality in aged laying hens. It has been reported that Se de ciency is detrimental to bone microarchitecture possibly through decreasing antioxidant capacity [37]. Thus, these results suggested that oxidative stress caused by Se de ciency [38] may decrease eggshell strength in conjunction with age. Moreover, 0.45 mg/kg selenium yeast supplementation for 12 weeks signi cantly elevated egg quality including increasing eggshell strength, and decreasing egg weight and translucent eggs. Consistent with previous studies, eggshell breaking strength was signi cantly increased after high does selenium yeast supplementation [39,40]. Meanwhile, translucent eggshell was a problematic issue to affect eggshell appearance and decrease the commercial value of eggs. The reason of eggshell translucent was still unknown, it was inferred to be associated with variations of eggshell membrane [41]. The decreasing of translucent eggs suggested that selenium yeast supplementation may increase the antioxidant capacity of eggs to ameliorate the decline of egg quality. Moreover, increasing egg weight in the last phase of laying cycle is another of the problematic issues [42][43][44], large eggs are more di cult to handle and more prone to break during transport and collection. The results in our study showed that selenium yeast supplementation decreased the egg weight, suggested that selenium yeast supplementation in last stage of laying period played bene cial roles in egg and shell quality.
In our study, the effect of selenium yeast supplementation on shell gland of aged laying hens were obtained by whole transcriptome analysis. The results of transcriptome analysis revealed several novel genes and biological pathways regulates the ions transport, laying period, eggshell calci cation and consequently the eggshell formation. Based on the enrichment pathways, molecular functions, and gene expression trends, thirteen genes were identi ed as potential candidate genes during eggshell formation including CEMIP, SDC3, OVAL, SPP1, SLC6A17, SLC13A5, OTOP2, CA2, POMC, PTN, PENK, WNT11 and EREG.
The growing crystals interact with the shell organic matrix to form a highly ordered microstructure during shell calci cation. Thus, the organic matrix interacts with minerals plays a key role in eggshell formation. We paid particular attention to genes involved in shell mineralization, including CEMIP, SDC3, OVAL, SPP1. CEMIP involved in glycosaminoglycan metabolism and calcium release from endoplasmic reticulum [45]. The up-regulated CEMIP expression (log 2 fold change > 2 in both two comparison) might suggest that high dose selenium yeast supplementation can promote calcium release for eggshell formation. SDC3 protein possesses domains containing some potential glycosaminoglycan attachment sites [46]. Glycosaminoglycan is widely thought to regulate mineral deposition and determine the properties of eggshell [47]. Thus, we inferred that the up-regulation of SDC3 may play an active role in eggshell formation. OVAL is an abundant eggshell matrix protein binding calcium and plays an active role in carbonate formation [48], in coherence with our results, the up-regulation of OVAL might improve the eggshell formation. SPP1 was known as osteopontin, has mineral-binding domains [49] which involved in calcium metabolism and calcium carbonate precipitation [47] for eggshell calci cation. Consistent with other studies, the up-regulation of SPP1 occurred in eggshell calci cation period [50,51]. Similarly, the pathways DEGs enriched in including calcium signaling pathway, glycosaminoglycan biosynthesis and positive regulation of bone mineralization, were also closely associated with eggshell calci cation and formation. Hence, the results suggested that selenium yeast supplementation induced shell calci cation by regulating the expression of CEMIP, SDC3, OVAL, SPP1 and other candidate genes bene cial for eggshell formation.
The chicken eggshell is a highly ordered structure and is mainly composed of calcium carbonate. Furthermore, the eggshell formation process requires a large amount of calcium (Ca 2+ ) and bicarbonate (HCO 3 − ), so the ions transport plays a crucial role. Some DEGs and pathways associated with ions transport were screened out in our study, including SLC6A17, SLC13A5, OTOP2, CA2, voltage-gated calcium channel activity and calcium signaling pathway. SLC6A17 and SLC13A5 as solute carrier family, plays an important role in transporting ions across cell membranes for synthesis of eggshell formation, it had been reported that increasing SLC6A17 expression plays an alanine transport role during eggshell formation, while SLC13A5 plays a more important role in the initiation of synthesis of eggshell formation [52]. The carbonate ion required for eggshell synthesis is catalyzed by carbonic anhydrases (CA), and uterine glandular cells possess the carbonic anhydrase activity sites. Growing evidence has demonstrated that high level of CA2 expression play a pivotal role for conversion of intracellular CO 2 to HCO 3 − in chicken [53,54]. As a member of the otopetrin gene family, OTOP2 gene may have similar functions to OTOP1, which regarded as modulator of cellular calcium in ux [55], to transport calcium across the uterine epithelium for eggshell calci cation [56]. Our ndings inferred that SLC6A17, SLC13A5, OTOP2 and CA2 are the regulators of ions transport affect by selenium yeast.
Moreover, the eggshell formation also depends upon numerous physiological adaptations and processes by the uterine cells, as well as reproductive hormones. In our study, functional enrichment analysis showed that selenium yeast supplementation may affect the response to estrogen, regulation of cell proliferation, regulation of epithelial cell proliferation, and extrinsic apoptotic signaling pathway. POMC plays a role in stimulating the release of cortisol hormone and its up-regulate expression during the mineralization period [52]. Consistent with other studies, a higher expression of POMC was observed in a hard shell egg group [53]. PTN is a developmentally-regulated growth factor and its expression is induced by estrogen [57], PTN and estrogen were reported with a pivotal role in eggshell formation [52]. These ndings suggested that selenium yeast may enhance the expression of reproductive hormones to affect the eggshell formation and quality. Indeed, WGCNA analysis also identi ed some candidate genes which affected the eggshell formation regulated by selenium yeast, including PENK, WNT11, EREG. These upregulation DEGs had been reported in previous studies which had a vital role in eggshell formation [52].
Previous studies have found that trace elements manganese and zinc enhanced eggshell strength by improving biosynthesis of glycosaminoglycan [58] and affecting carbonic anhydrase activity [59], respectively. An increasing number of studies had explored that selenium has a good effect on eggshell quality of laying hens. However, studies on the molecular mechanism of selenium yeast affecting eggshell quality is limited. To date, it is hypothesized that selenium exhibits bene cial regulation on eggshell quality may be directly involved in the process of regulating eggshell formation or the interaction between trace elements. Overall, based on biological functions of the DEGs, we hypothesize that the molecular pathways impacted by selenium on aged laying hens are those related to eggshell mineralization, hormone regulation and ion transduction, suggesting that selenium might play a bene cial role in eggshell formation. Despite that, additional studies are needed to draw conclusive remarks about the molecular mechanisms modulated by selenium yeast in accurate phases of eggshell formation.

Conclusion
In conclusion, the dietary selenium yeast supplementation improved eggshell strength and reduced eggshell translucence in aged laying hens. Moreover, the plasma Se content, plasma antioxidant enzyme activity, tissue Se content were dose depended on the change of dietary selenium yeast intakes. The plasma Se content could be a biomarker to detect Se status. Transcriptomic analysis revealed that selenium affected eggshell quality through the process of eggshell mineralization, ion transport, and laying period. CEMIP, OVAL, SLC6A17, SLC13A5, POMC and PENK were identi ed as candidate genes affecting eggshell quality regulated by selenium.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on request.
Ethics approval and consent to participate The experimental animal protocols for this study were approved by the Animal Care and Use Committee of China Agricultural University (No. AW05060202-1).

Consent for publication
Not applicable.