Fertile eggs (n = 100; mean weight = 70.04 g, SD = 2.19 g) from 40-week-old broiler hens (Cobb 500) were purchased from a commercial breeder farm (Y. Brown and Sons Ltd., Hod Hasharon, Israel). Eggs were incubated in a Petersime hatchery at the Faculty of Agriculture of the Hebrew University under standard conditions (37.8°C and 56% relative humidity). On embryonic day (E) 10, eggs were candled, unfertilized eggs and dead embryo eggs were removed. Tissue sampling was performed at E11, E13, E15, E17, E19, at hatch (actual time of hatch at the hatchery, 480–483 h in incubation) and at chick placement (31–34 h post hatch, prior to the exposure to feed at the farm). At each sampling day, six embryos/hatchlings were randomly selected, euthanized by cervical dislocation and the following parameters were recorded: embryo yolk free body mass (YFBM), yolk content weight, YS tissue weight (according to Yadgary et al., 2010), liver weight and breast muscle weight. The relative weights of liver, breast muscle and YS tissue were calculated as percent of YFBM. Blood glucose and ketone levels were also recorded. In addition, liver, YS tissue and breast muscle samples were collected for creatine and glycogen content determination and for the evaluation of expression levels of genes involved in creatine synthesis, gluconeogenesis, glycogenesis, and glycogenolysis. Kidney samples were collected for the evaluation of expression of creatine synthesis. Immediately after collection, tissue samples were placed in liquid nitrogen and kept under −80^oc until further processing.

Liver, YS tissue and breast muscle samples were lyophilized and later examined for their creatine and glycogen content by Swiss-BioQuant-AG (Reinach, Switzerland). The concentrations of creatine and glycogen (mg/g dry weight tissue) were determined, and the total amounts (mg) in each tissue were calculated. Due to insufficient material load, breast muscle samples were analyzed only at E15, E17, E19, hatch and chick placement.

Blood glucose and ketones levels were obtained directly from the jugular vein of embryos and determined using FreeStyle Optimum Gluco/Keto meter and test strips (Abbott Diabetes Care Ltd., Witney Oxon, United Kingdom) according to the manufacturer’s instructions.

Total RNA was isolated from 100 mg of tissue (from YS tissue, liver, kidney and breast muscle) using TRI-Reagent (Sigma-Aldrich, St. Louis, MO) according to the manufacturer’s protocol. RNA concentration was determined using a Nano Drop ND-1000 instrument (Thermo Fisher Scientific, Wilmington, DE). Total RNA was treated with DNase using a Turbo DNA-free Kit according to the manufacturer’s protocol (Ambion; Thermo Fisher Scientific, Wilmington, DE). cDNA was created from 1 µg of DNA-free RNA using the qPCRBIO cDNA synthesis kit according to the manufacturer’s protocol (PCRBIOSYSTEMS, London, United Kingdom). Relative mRNA expression was evaluated using gene-specific primers (Table 1) of genes involved in creatine synthesis [arginine-glycine amidinotransferase (AGAT), guanidinoacetate N-methyltransferase (GAMT)], gluconeogenesis [fructose-1,6-bisphosphatase 1 (FBP1), glucose-6-phosphatase 2 (G6PC2)], glycogenesis [glycogen synthase (GYS)], and glycogenolysis [glycogen phosphorylase L (PYGL)] and the housekeeping genes; cytoskeletal protein (β-actin) and hypoxanthine phosphoribosyl transferase 1 (HPRT). Primer sequences were designed using Primer-BLAST software (Ye et al., 2012) based on published cDNA sequences where available and purchased from Sigma-Aldrich (Rehovot, Israel). PCR products were validated by gel electrophoresis in 1.5% agarose gel. Real-time qPCR reactions were conducted in triplicate in a Roche Light cycler 96 (Roche Molecular Systems, Inc., Pleasanton, CA). Each reaction (20 µL) included 3 µL of cDNA sample diluted 1:20 in ultrapure water (UPW, Biological Industries, Beit HaEmek, Israel), 4 µM of each primer, and Platinum SYBR Green qPCR super mix-UDG (Thermo Fisher Scientific, Wilmington, DE). Reaction conditions: preincubation at 95°C for 60°s, followed by 40 cycles of a 2-step amplification cycle of 95°C for 10°s and 60°C for 30°s, ended with a melting curve generated by the following conditions: 95°C for 60°s, 65°C for 60°s, and 97°C for 1 s. Relative mRNA expression was calculated by subtracting the geometric mean of cycle threshold (Ct) values of β-Actin and HPRT reference genes from sample Ct (Pfaffl, 2007).

Data of creatine and glycogen concentrations in tissues and data of blood glucose and ketone levels were subjected to one-way ANOVA analysis with the embryonic day (age) as main effect. Data of creatine and glycogen amount and of gene expression in tissues were subjected to two-way ANOVA analysis with the embryonic day (age), tissue and their interaction as the main effects. Differences between means were tested by using Tukey’s HSD test and considered significantly different with p-value lower than or equal to 0.05 (p ≤ 0.05). Values are presented as mean ± standard error mean (SEM). All statistical analyses were carried out using JMP-pro 16 software (SAS Institute Inc., Cary, NC).

Measurements of creatine concentration (mg/g of dry tissue) in the breast muscle showed a significant increase by 47% (p-value, p = 0.0009) from embryonic day (E) 15 to E17 (Figure 1A, and Supplementary Table S1). Then, throughout E19, hatch and chick placement, creatine levels did not differ significantly between consecutive days. The liver and yolk sac (YS) tissue exhibited similar patterns (Figures 1B, C, and Supplementary Table S1), as creatine concentration increased by 3.8-fold and 3.4-fold respectively from E13 up to the highest value at hatch (p < 0.0001). Different from the breast muscle, creatine concentration decreased in the liver and YS tissue at chick placement by 2-fold and 1.5-fold, respectively. The total amount of creatine in each organ (Figure 1D and Supplementary Table S2) was calculated by multiplying the total dry tissue weight (Suppl. Table 6) by the concentration of creatine (Figures 1A–C and Supplementary Table S1). A two-way ANOVA test revealed a significant interaction between the effects of age and tissue (p < 0.0001). This was mainly due to the breast muscle tissue, as its creatine amount showed a constant elevation pattern unlike the liver and YS tissue. Results also show that creatine is stored mainly in the breast muscle, as its amount was 3-fold higher at hatch and 15-fold higher at chick placement compared with combined amounts of the liver and YS tissue. It should be noted that compared to the liver, the YS tissue had higher amounts of creatine throughout incubation, significantly higher on E17 and at hatch (Figure 1D and Supplementary Table S2).

As for glycogen, its concentration (mg/g of dry tissue) exhibited similar patterns in the breast muscle, liver, and YS tissue (Figures 2A–C, and Supplementary Table S1). Towards E17 an increase in glycogen levels was observed in all tissues, then towards hatch, glycogen levels were significantly decreased reaching the lowest levels at chick placement (p < 0.0001). The total amount of glycogen in each organ (Figure 2D and Supplementary Table S1) was also calculated by multiplying the total dry tissue weight (Supplementary Table S6) by the concentration in tissues (Figures 2A–C and Supplementary Table S1). A two-way ANOVA test revealed a significant interaction between the effects of age and tissue (p < 0.0001). The interaction was mainly due to the YS tissue, as its glycogen amount showed the most dynamic pattern, with a rapid increase up to E17 followed by a sharp decrease. In addition, throughout incubation the YS tissue had the highest amount of glycogen, as it was 27-fold higher on E11 compared to the liver and 3.3-fold higher at hatch and at chick placement, compared to the combined amount in the liver and breast muscle (Figure 2D and Supplementary Table S2).

Glucose (mg/dL) and ketones (β-HBA; mmol/L) levels were measured from E11 to chick placement and exhibited altered patterns. From E11 to E17, glucose levels were relatively constant ranging between values of 123.5–143.6 mg/dL (Figure 3A). From E19 and until hatch and chick placement, blood glucose levels increased significantly by 38% (p < 0.0001), reaching values of 191.9–208.3 mg/dL. Ketone levels (Figure 3B) showed a dynamic pattern, starting on E11 with the lowest level (2.09 mmol), increasing gradually up to the highest value on E17 (4.33 mmol/L; p < 0.0001). On E19, ketones level decreased to 3.24 mmol/L while at hatch levels increased again to 4.06 mmol/L.

The potential of creatine de novo synthesis in the tissues was evaluated by the examination of arginine-glycine amidinotransferase (AGAT), and guanidinoacetate N-methyltransferase (GAMT) (Figure 4 and Supplementary Table S3). Expression analysis of AGAT and GAMT by a two-way ANOVA test revealed significant interactions between the effects of age and tissue (p < 0.0001). The interaction was mainly associated with the varying expression patterns in tissues. In the breast muscle, AGAT expression showed a dynamic pattern, as it was increased from E11 to E19, followed by a decrease at hatch, then, sharply increased again reaching the highest level at chick placement. GAMT expression in the breast muscle also showed a dynamic pattern, increasing from E11 to E19 followed by a decrease at hatch. In the kidney, expression of AGAT remained stable and was not significantly different between consecutive embryonic days, however, GAMT expression increased gradually from E13 along the second half of incubation, reaching highest values at hatch and at placement. The liver and YS tissue showed alternating expression patterns as liver AGAT expression increased towards hatch and chick placement, and YS tissue AGAT expression decreased. As for GAMT expression, in the liver significantly different values were found on E13 as levels were decreased, while at chick placement levels were significantly increased. In the YS tissue GAMT expression increased up to E17, followed by a decrease up to chick placement. Altogether, results demonstrate a shift in expression levels of AGAT and GAMT, decreasing in the extraembryonic YS tissue towards hatch and placement, and increasing in the embryonic liver, breast muscle and kidney.

Gluconeogenesis potential was evaluated by the examination of two genes: fructose-1,6-bisphosphatase 1 (FBP1) and glucose-6-phosphatase 2 (G6PC2) (Figure 5 and Supplementary Table S4). Expression analysis of FBP1 and G6PC2 by a two-way ANOVA test revealed significant interactions between the effects of age and tissue (p < 0.0001). Also here, the interaction is associated with the altering expression patterns in tissues. Results showed daily fluctuation in liver FBP1 expression. In the YS tissue, FBP1 expression was elevated on E13 and significantly decreased on E15, then remained constant throughout incubation up to chick placement. In the breast muscle, FBP1 expression level was relatively low, decreasing between E15 and E17, and increasing again at placement. The variation in expression pattern of G6PC2 was similar to that of FBP1 expression in the liver, breast muscle and YS tissue. In general, when comparing the expression of gluconeogenic genes between the liver, YS tissue and breast muscle within days, the liver showed the highest values throughout incubation.

The potential of glycogen synthesis and breakdown was evaluated by two genes: glycogen synthase (GYS) and glycogen phosphorylase L (PYGL) (Figure 6 and Supplementary Table S5). Glycogen synthase (GYS) has two isoforms/variants: glycogen synthase 1 (GYS1) and GYS2 (Roach et al., 1998). GYS1 is expressed at lower levels in muscle, whereas GYS2 is primarily responsible for glycogen synthesis in the liver (Browner et al., 1989; Nuttall et al., 1994). Accordingly, in the breast muscle, the muscle variant GYS1 was used for expression analysis of glycogen synthase, while in the YS tissue and liver, GYS2 was used. As for, glycogen phosphorylase, the variant PYGL was used for the breast muscle, YS tissue and liver. PYGL was shown to be expressed at the protein level also in the breast muscle (Zambonelli et al., 2016). A two-way ANOVA test revealed that there was no significant interaction between the effects of age and tissue in GYS and PYGL expression. The tissue factor had statistically significant effect on GYS expression (p < 0.0001) with the highest expression levels in the liver, medium values in the YS tissue and lowest in the breast muscle. Similarly, the age factor was also found to be significant (p = 0.00035), with a peak in overall GYS expression at E19, significantly different from all sampling points apart from E17. As for PYGL, the tissue factor had a significant effect (p < 0.0001) with the highest expression levels in the liver, medium values in the YS tissue and lowest in the breast muscle. The age factor was also found to be significant (p = 0.049), with the highest PYGL expression at E17. As in the gluconeogenic genes, expression analysis of glycogen synthesis and breakdown genes between tissues within each day, showed the highest expression levels in the liver, with medium values in the YS tissue and lowest in the breast muscle.