Proteomic-metabolomic combination analysis reveals novel biomarkers of meat quality that differ between young and older ducks

Abstract

In order to explore the difference and its underlying mechanism between young and older ducks, 60-day-old (D60) and 300-day-old (D300) of young ducks and 900-day-old ducks (D900) of older ducks were selected and studied. HE staining indicated that breast muscle fibers in the D900 group were more inseparable than D60 and D300 groups and the greater redness were showed in D300 and D900 groups. Quantitative proteomic analyses were conducted to further identify differences between young and older ducks that 61 proteins overlapped in the comparative analysis of the D900 vs. D60 and D900 vs. D300 groups. Furthermore, metabolomics analysis from the D900 group showed marked differences from the results of the D60 and D300 groups in 31 unique metabolites. In particular, lower guanosine, hypoxanthine, guanine, and doxefazepam levels indicated the increased nutritional value of older ducks. Integrated proteomics and metabolomics analysis showed that purine metabolism was specifically enriched, indicating that NME3, RRM2B, AMPD1, and AMPD3 might mainly affect meat from older ducks. In conclusion, our results indicated that meat from 900-day-old ducks possessed a unique biochemical signature that could provide candidate biomarkers to distinguish young ducks from older ducks.

INTRODUCTION

Duck, as one of the most popular poultry meats, plays an important role in people's lives in China. In recent years, with the improvement in living standards, consumer demand for older ducks has increased because of their high nutritional and medicinal value (Wang et al., 2020a). However, due to their nutritional value and cultivation cost, older ducks retail at a higher price than young ducks, which has caused merchants to replace older ducks with young ducks, leading to economically motivated adulteration events (Everstine et al., 2013). Therefore, it is imperative to identify key biomarkers to distinguish older ducks.

Recently, the emerging technologies of proteomics and metabolomics analyses have been widely used in the identification of animal meat (Liao et al., 2022). He et al. identified 212 proteins in ducks under heat stress using proteomic techniques and found that HSP70 negatively regulated the defensive mechanism against heat stress in ducks (He et al., 2019). Furthermore, the dynamic change of metabolite composition was determined in chicken to evaluate the development of meat quality based on non-targeted metabolomics (Li et al., 2022). The high hypoxanthine and inosine monophosphate (IMP) content in chilled chicken was shown to distinguish between chilled and frozen chicken after cooking (Wang et al., 2022). Meanwhile, Wang et al. found that metabolites associated with histidine, anserine, and NAD may regulate the reduction of wooden breast myodegeneration in chickens (Wang et al., 2020b), suggesting that omics analysis can play important roles in defining novel biomarkers of meat quality (Wang et al., 2022). However, the use of proteomic and metabolomic analyses to distinguish between young and older ducks has not been investigated.

In this study, 60-day-old (D60), 300-day-old (D300), and 900-day-old (D900) White Shaoxing ducks were examined, and the physical properties and muscle fiber characteristics were determined. Label-free-based quantitative proteomics and metabolomics profile analyses were performed to compare differential protein expression and metabolites in young (D60 and D300) and older (D900) ducks. In addition, a conjoint proteomics and metabolomics analysis was performed to provide a more comprehensive perspective to explore the possible candidate biomarkers in older ducks.

METHOD AND MATERIALS

Ethics Approval

The animals used in this study were raised and euthanized in accordance with the national standard guidelines for the ethical review of animal welfare (GB/T 35892-2018) issued by the General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China and the Standardization Administration of the People's Republic of China. All experimental procedures were approved by the Institute of Animal Husbandry and Veterinary Science, Zhejiang Academy of Agricultural Sciences (Hangzhou, China). We also confirm that all efforts were made to minimize animal suffering.

Animals

Twenty-one pure female White Shaoxing ducks (Anas platyrhynchos) were obtained from Zhuji Guowei Technology Co., Ltd. (Shaoxing, China), including seven D60 ducks, seven D300 ducks, and seven D900 900 ducks. All ducks were raised under the same conditions, including water, temperature, and diet. The composition and nutritional level of the basic diet is shown in Table S1.

Tissue Collection

At 60, 300, and 900 d, ducks were anesthetized via venous injection of pentobarbital sodium (100 mg/kg), and breast muscle samples were collected immediately for omics analysis. The remaining breast muscle samples were fixed in 4% formalin for muscle fiber analysis or excised for shear force and pH measurements.

Muscle Fiber Measurements

Hematoxylin and eosin (HE) staining was conducted on breast muscle tissues fixed in 4% paraformaldehyde to observe histomorphological changes. Then, the obtained slices were examined at 200 × magnification using a Nikon E100 microscope (Tokyo, Japan), and photographs were obtained.

pH and Shear Force Measurements

The pH and shear force were measured after slaughter, immediately. The pH of each muscle sample was measured using a pH meter (DELTA 320, Mettler Toledo, Zurich, Switzerland), and the muscle tissue was examined at a depth of 10 mm at 3 points (upper, middle, and lower). Breast muscle sections at different positions (1 cm thickness) were obtained to perform shear force analysis using a digital tenderness meter (GR-151, Warner-Bratzler, Tallgrass Soultions, Inc, Manhattan, KS).

Label-free-based Quantitative Proteomics Analysis

Protein Extraction and Digestion

Protein was obtained from breast muscle samples using lysis buffer, and the protein concentration was measured using a BCA Kit (Beyotime, Shanghai, China). Then, 100 mM TEAB (Sigma,-Aldrich, Darmstadt, Germany) was added to the protein samples, and the mixture was digested with 1:50 trypsin/protein overnight. The proteins were treated with 5 mM dithiothreitol (Sigma) and 11 mM iodoacetamide (Sigma) for reduction at 56°C for 30 min. The digested samples were desalted using a C18 SPE column (Nest Group, Southborough, MA).

Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS) Analysis

The digested peptides were separated using a nanoElute UHPLC system (Bruker Daltonics, Billerica MA). The peptides were dissolved in solvent A, which contained 0.1% formic acid and 2% acetonitrile in water. Then, the peptides were separated using solvent B, which contained 0.1% formic acid and 90% acetonitrile, with a gradient setting of 4 to 20% solvent B over 68 min, 20 to 32% for 4 min, and 80% for 4 min at a constant flow rate of 500 nL/min. The separated peptides were ionized by subjecting them to an NSI ion source and then were analyzed using Orbitrap exploration 480 mass spectrometry. The electrospray voltage applied was 2.30 kV. The precursors and fragments were detected by the Orbitrap using an MS/MS scan range from 400 to 1200 m/z for primary mass spectrometry and 100 m/z for secondary mass spectrometry. The dynamic exclusion time of MS/MS scanning was set to 30 s.

Bioinformatics Analysis of Differentially Expressed Proteins (DEPs)

The obtained data were annotated using the Anas platyrhynchos proteome reference on the UniProt website (https://www.uniprot.org/). Data with an absolute fold change > 1.5 or < 0.67 and a P value < 0.05 (t-test) were considered DEPs. Gene Ontology (GO) terms were mapped using the UniProt-GOA database, and functional enrichment analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) for characteristic DEPs obtained from 900-day-old ducks. The GO terms and KEGG pathways with a corrected P value < 0.05 according to a 2-tailed Fisher's exact test were considered significant. Protein-protein interaction (PPI) analysis (the confidence > 0.7) was performed using the STRING database (https://string-db.org/) to investigate the interactions of DEPs.

Metabolomic Analysis

Breast Muscle Sample Extraction

Breast muscle samples (20 mg) were prepared, and 400 μL of 70% methanol solution was added. After being placed on ice for 15 min, the samples were shaken at 1500 rpm for 5 min and centrifuged at 12,000 rpm and 4°C for 10 min. Then, 300 μL of the supernatant was collected and stored at −20°C for 30 min and then centrifuged at 12,000 rpm and 4°C for 3 min to collect the metabolomic supernatant for LC-MS analysis.

High-Performance Liquid Chromatography (HPLC)-MS Analysis

All breast muscle samples were examined using an LC-MS system (Agilent 1100, Palo Alto, CA) with a Waters ACQUITY UPLC HSS T3 C18 column (1.8 µm, 2.1 mm × 100 mm). The mobile phase flow rate was set to 0.4 mL/min and the injection volume of solvent (0.1% formic acid and 10 mM acetonitrile) was 2 μL. The samples were dissolved in solvent A containing 0.1% formic acid in water and solvent B containing 0.1% formic acid in acetonitrile with a gradient setting of 95:5 V/V at 0 min, 10:90 V/V at 11.0 min, 10:90 V/V at 12.0 min, 95:5 V/V at 12.1 min, and 95:5 V/V at 14.0 min.

Bioinformatics Analysis of Differentially Expressed Metabolites (DEMs)

The raw LC-MS data were converted into mzML format using ProteoWizard software through peak extraction, peak alignment, and retention time correction. Principal component analysis and orthogonal partial least squares discriminant analysis (OPLS-DA) were performed using R software (www.r-project.org). Data with variable importance in the projection ≥ 1.0 and a P value < 0.05 (t-test) were considered DEMs. The DEMs were mapped to the KEGG pathway database (http://www.kegg.jp/kegg/pathway.html). The metabolome analysis included 7 biological replicates.

Statistical Analysis

Statistically significant differences between the D60, D300, and D900 groups were determined by one-way ANOVA. The statistical analysis was performed using GraphPad Prism 8.0 software (GraphPad Software Inc, San Diego, CA). A value of P < 0.05 was considered statistically significant.

RESULTS

Effect of Age on the Characteristics of Duck Breast Muscle

To reveal the difference in duck meat quality and muscle fiber characteristics at different ages, HE staining and analysis of the physical properties of breast muscle were performed. Meat from the D300 and D900 groups displayed greater redness than that from the D60 group, but the redness was not significantly different between the D300 and D900 groups (Figure 1A). Additionally, the pH and shear force in the D900 group were not significantly different from those in the D60 and D300 groups (Figures 1B and 1C). The muscle fiber characteristics were compared to further assess variations among the D60, D300, and D900 groups. As shown in Figure 1D, breast muscle fibers in the D900 group were more inseparable than those in the D60 and D300 groups.

Figure. 1.

Phenotype of the breast muscle of 60-, 300-, and 900-day-old ducks. (A) Schematic of breast muscle at different ages. (B, C) pH value and shear force of breast muscle. (D) Hematoxylin and eosin staining of breast muscle. Vertical bars indicate the mean ± standard error (n = 7). Different letters indicate P < 0.05.

Identification and Comparison of DEPs at Different Ages

Comparative proteomics analyses of the D900 vs. D60 and D900 vs. D300 groups were performed using the label-free proteomic approach to reveal the effect of age on protein changes in duck breast muscle. Based on volcano plots, Venn diagrams, and clustering maps of the DEPs, 616 DEPs (295 upregulated and 321 downregulated) were found among the 3 groups (Figure 2).

Figure 2.

Changes in protein expression during aging in ducks. (A, B) Volcano plots of differentially expressed proteins (DEPs) in the D900 vs. D60 and D900 vs. D300 groups, respectively. (C) Venn diagram showing the number of overlapping DEPs between the D900 vs D60 and D900 vs. D300 groups. (D) Heat map analysis of DEPs in the D60, D300, and D900 groups. Orange and green indicate up- and downregulated proteins, respectively.

Functional Enrichment Analysis of the Characteristic DEPs in the D900 Group

To obtain the characteristic DEPs of the D900 group, 61 proteins among the 616 DEPs were screened (Figures 2C and 3A). Twenty-five identified proteins were upregulated in the D900 group compared with the D60 and D300 groups. Additionally, 36 proteins showed variation trends (Table 1). GO and KEGG enrichment analyses were performed to further explore the potential functions of the 61 significant DEPs in the D900 group. The GO terms were categorized as biological processes, molecular functions, and cellular components (Figure 3B). In the biological process group, DEPs were involved in the carboxylic acid biosynthetic process (GO:0046394), monocarboxylic acid metabolic process (GO:0032787), response to vitamins (GO:0033273), and carboxylic acid metabolic process (GO:0019752). In the cellular component category, the enriched GO terms were mainly involved in the fibrinogen complex (GO:0005577). The DEPs in the molecular function category were mainly related to AMP deaminase activity (GO:0003876) and adenosine-phosphate deaminase activity (GO:0047623). Additionally, the regulatory pathways associated with the 61 DEPs in the D900 group were extracted (Figure 3C). Thirteen enriched pathways were revealed, and these pathways were involved in purine metabolism (map00230), tyrosine metabolism (map00350), pyrimidine metabolism (map00240), fatty acid biosynthesis (map00061), and the PPAR signaling pathway (map03320). which were mainly involved in purine and lipid metabolism. Then, PPI analysis of DEPs was performed to investigate the interactions between the characteristic DEPs of the D900 group, including APOB, RRM2B, and NME3 (Figure 3D).

Figure 3.

Selected differential proteins of the D900 group. (A) Heat map analysis of 61 proteins in the D900 group compared with the D60 and D300 groups. Orange and green indicate up- and downregulated proteins, respectively. (B) Gene Ontology classification. (C) Kyoto Encyclopedia of Genes and Genomes pathway enrichment. (D) Protein-protein interaction network analysis.

Table 1.

The selected list of unique DEPs between D900 vs. D60 and D900 vs. D300.

Proteins	Names	Average fold change		P value
		D900 vs. D60	D900 vs. D300	D900 vs. D60	D900 vs. D300
R0J8G4	PLIN1	3.415	2.809	0.007	0.017
R0LU42	SH3KBP1	9.639	1.620	0.000	0.041
R0LJE0	LOC101801977	1.621	1.825	0.006	0.003
R0JRH0	ACOT2	1.659	1.662	0.000	0.000
R0M884	LOC101798048	1.707	0.647	0.001	0.002
R0KA54	LOC101805370	1.653	1.594	0.016	0.003
R0M4M0	NME3	5.317	4.468	0.001	0.014
R0LSK4	AQP4	2.169	2.297	0.000	0.000
R0JU38	HABP4	2.293	2.516	0.011	0.019
R0KZU9	KRT78	2.993	1.743	0.004	0.049
R0LJW6	CBR4	1.619	1.954	0.004	0.015
R0JY88	MAP3K7CL	2.505	2.453	0.012	0.018
R0KA83	CHMP1B	2.653	1.783	0.001	0.004
R0K265	LOC101790722	2.163	0.656	0.041	0.006
R0KKS9	KRT19	1.732	1.741	0.004	0.025
R0M1G1	RRM2B	5.200	1.937	0.000	0.006
R0L2R8	ACSBG2	3.549	1.795	0.000	0.000
R0LZX5	PLXDC2	1.808	1.595	0.002	0.003
R0JQ97	APOB	58.716	1.886	0.000	0.000
R0M3Q3	LOC101799714	8.162	5.265	0.000	0.000
R0LAC7	TNNC1	1.957	2.471	0.001	0.000
R0J6S6	PABPC1	1.514	1.974	0.009	0.010
R0JGE1	SLC37A4	1.662	2.186	0.009	0.005
R0JLZ8	MIF	1.585	0.643	0.046	0.039
R0JC05	THAP4	1.870	1.566	0.008	0.003
R0JAW6	HPX	0.065	0.286	0.000	0.000
R0L315	Npc2	0.256	0.377	0.022	0.042
R0JMJ2	RPIA	0.524	0.518	0.032	0.048
R0JD26	ARL3	0.120	0.295	0.004	0.043
R0LYW2	AMPD1	0.586	0.647	0.008	0.033
R0LSP8	MRPL32	0.444	2.304	0.023	0.032
R0LST2	AMPD3	0.551	0.535	0.004	0.004
R0LDN0	PRSS12	0.270	0.591	0.000	0.000
R0JS80	FGB	0.260	0.495	0.000	0.000
R0L173	GSN	0.532	0.587	0.001	0.002
R0JIF4	C5	0.119	0.44	0.002	0.023
R0JH54	ERLIN2	0.537	0.450	0.004	0.002
R0LAU3	MRPL45	0.651	1.783	0.005	0.004
R0M159	ASPH	0.523	0.617	0.029	0.019
R0K6M4	FKBP9	0.433	2.202	0.001	0.005
R0LYX6	SVEP1	0.381	0.638	0.029	0.032
R0LDM8	STK26	0.617	0.592	0.014	0.033
R0KUL5	SNX9	0.584	2.358	0.010	0.009
R0L9U8	PSMD5	0.440	0.570	0.009	0.014
R0JS45	SPECC1L	0.301	0.458	0.000	0.002
R0KJX5	EPHX2	0.415	3.634	0.000	0.017
R0M1S7	SCPEP1	0.510	0.469	0.002	0.003
R0LQ88	PSPH	0.647	0.633	0.000	0.000
R0KC12	SUCLG2	0.267	0.553	0.000	0.005
R0KQL3	ME3	0.289	0.385	0.000	0.000
R0K0Y0	ZYX	0.266	0.470	0.001	0.023
R0LIW0	FGG	0.359	0.609	0.000	0.002
R0L2W8	AKR1D1	0.060	0.154	0.000	0.000
R0L8P5	ADH5	0.664	0.590	0.002	0.001
R0LXM8	TPPP3	0.320	0.642	0.000	0.002
R0M2M3	GALK1	0.402	0.464	0.002	0.005
R0M1J7	SPARC	0.101	0.437	0.000	0.014
R0K280	MRPS18A	0.657	1.863	0.006	0.022
R0JPJ8	SERPINH1	0.148	5.031	0.000	0.001
R0LJA8	EFEMP1	0.235	0.345	0.013	0.032
R0KW16	C9	0.229	0.280	0.000	0.000

Metabolic Profiling Analysis of Characteristic Metabolites in the D900 Group

To determine the metabolite levels at different ages, breast muscle metabolomic analysis was performed using HPLC-MS. Notably, a clear separation between the D60, D300, and D900 groups was shown by the OPLS-DA score plots (Supplementary Figure S1A), which suggested the presence of distinct metabolic properties in each pairwise comparison between groups. Correspondingly, a permutation analysis was conducted (Supplementary Figure S1B), which indicated that the principal components in the OPLS-DA models were reliable.

To further examine characteristic metabolites in the D900 group, intersection statistics of the D900 vs. D60 and D900 vs. D300 groups were determined to identify unique metabolites. A total of 31 differential overlapped metabolites (18 metabolites in the positive mode and 13 metabolites in the negative mode) are listed in Figures. 4D and 4E. Interestingly, all distinguishing unique metabolites were significantly downregulated in the D900 group compared with the D60 and D300 groups, including 1 benzene compound, 2 glycerophospholipids, 3 nucleotide metabolomic compounds, 6 fatty acyls, 6 heterocyclic compounds, 3 amino acid metabolomic compounds, 8 organic acid metabolomic compounds, 1 alcohol and amine compound, and 1 aldehyde (Table 2). In particular, the level of 4 unique biomarkers (guanosine, hypoxanthine, guanine, and inosine) decreased with age. Furthermore, KEGG pathway analysis reveals that 31 unique metabolites in the D900 group were enriched in nucleotide metabolism, purine metabolism, histidine metabolism, arginine, and proline metabolism, cysteine and methionine metabolism, (Figures 4C and 4F), which suggested these metabolic pathways played significant roles in flavor regulation. Thirty-one differential metabolites were screened based on positive and negative mode using KEGG pathway analysis

Figure 4.

Overlapping metabolites in a comparative analysis of the D900 vs. D60 and D900 vs. D300 groups. (A) Venn diagram depicting the overlapping metabolites in the positive mode. (B) Correlation plot of the 18 unique metabolites in the positive mode. (C) Kyoto Encyclopedia of Genes (KEGG) and Genomes pathway enrichment of 18 unique metabolites in the positive mode. (D) Venn diagram depicting the overlapping metabolites in the negative mode. (E) Correlation plot of the 13 unique metabolites in the negative mode. (F) KEGG pathway enrichment of the 13 unique metabolites in the negative mode.

Table 2.

List of the significate unique metabolites between D900 vs. D60 and D900 vs. D300 in positive and negative mode.

Description	Formula	Class	D900 vs. D60			D900 vs. D300
			Log2FC	P value	Trend	Log2FC	P value	Trend
Albendazole Sulfone	C12H15N3O4S	Benzene	-2.357	0.037	↓	-3.424	0.001	↓
LPC(O-16:0/2:0)	C26H54NO7P	GP	-1.958	0.000	↓	-1.004	0.047	↓
PC(16:0/2:0)	C26H52NO8P	GP	-1.117	0.002	↓	-1.119	0.046	↓
Guanosine	C19H25Cl2N3O3	Nucleotide	-1.420	0.007	↓	-1.339	0.005	↓
S-Adenosylmethionine	C15H22N6O5S	Nucleotide	-2.520	0.021	↓	-2.434	0.002	↓
Inosine	C10H12N4O5	Nucleotide	-1.834	0.033	↓	-2.627	0.002	↓
Carnitine C4:0	C11H21NO4	FA	-1.500	0.005	↓	-1.403	0.013	↓
Carnitine C6:0	C13H25NO4	FA	-1.559	0.005	↓	-1.660	0.018	↓
Carnitine C5:0	C12H23NO4	FA	-1.582	0.010	↓	-1.827	0.010	↓
Carnitine C22:2	C29H53NO4	FA	-3.208	0.016	↓	-3.262	0.026	↓
Arachidyl carnitine	C27H53NO4	FA	-3.818	0.003	↓	-2.307	0.026	↓
7,7-dimethyl-5,8-Eicosadienoic Acid	C22H40O2	FA	-1.212	0.002	↓	-1.001	0.015	↓
Prometon	C10H19N5O	Heterocyclic	-1.315	0.003	↓	-1.551	0.003	↓
Hypoxanthine	C5H4N4O	Heterocyclic	-1.639	0.032	↓	-2.386	0.002	↓
Guanine	C5H5N5O	Heterocyclic	-1.616	0.005	↓	-1.543	0.002	↓
Doxefazepam	C17H14ClFN2O3	Heterocyclic	-1.302	0.010	↓	-1.709	0.000	↓
Cerberin	C32H48O9	Heterocyclic	-1.480	0.001	↓	-1.428	0.018	↓
Eriocitrin	C27H32O15	Heterocyclic	-1.831	0.040	↓	-2.759	0.001	↓
Nalpha-Acetyl-L-Arginine	C8H16N4O3	Amino acid	-1.689	0.005	↓	-1.537	0.016	↓
delta-CEHC	C14H18O4	Amino acid	-1.367	0.022	↓	-1.739	0.002	↓
Tyr-Tyr-Thr	C22H27N3O7	Amino acid	-2.179	0.031	↓	-3.064	0.010	↓
Urocanic Acid	C6H6N2O2	Organic acid	-2.336	0.042	↓	-2.704	0.005	↓
3-Methyl-2-Oxobutanoic Acid	C5H8O3	Organic acid	-1.819	0.023	↓	-2.562	0.002	↓
L-Arogenate	C10H13NO5	Organic acid	-1.739	0.032	↓	-2.483	0.002	↓
N-(5-Phospho-D-ribosyl) anthranilate	C12H16NO9P	Organic acid	-1.813	0.016	↓	-2.664	0.002	↓
C-8 Ceramide-1-phosphate	C26H52NO6P	Organic acid	-1.077	0.002	↓	-1.105	0.032	↓
3-(2,5-dimethoxyphenyl) propanoic acid	C11H14O4	Organic acid	-1.728	0.030	↓	-2.434	0.002	↓
1-O-(2-Acetamido-2-deoxy-alpha-D-glucopyranosyl)-1D-myo-inositol 3-phosphate	C14H26NO14P	Organic acid	-2.537	0.010	↓	-1.387	0.013	↓
(2S)-2-Hydroxy-2-methyl-3-oxobutanoic acid	C5H8O4	Organic acid	-2.229	0.003	↓	-1.813	0.001	↓
5-[(1E,3E,5E,7E,9E,11E,13E,15E,17E)-18	C40H54O2	Alcohol and amines	-2.409	0.000	↓	-1.077	0.030	↓
(2S,3R,4R,5S)-2	C28H48O3	Aldehyde	-2.048	0.003	↓	-1.338	0.004	↓

Integrative Proteomics and Metabolomics Analysis

Integrative metabolomics and proteomics analysis of breast muscle was conducted to identify the unique regulatory mechanisms in D900 ducks. Based on the comparative analysis results between the D900 vs. D60 and D900 vs. D300 groups, co-enriched pathways with unique metabolites and proteins in the D900 group were determined, and one shared signaling pathway was enriched (Figures 5A and 5B). Interestingly, we found that the purine metabolism pathway, including NME3, RRM2B, AMPD1 and AMPD3, participated in flavor regulation in older ducks (Table 3).

Figure 5.

Integrated analysis of the unique differential proteins and metabolites in Kyoto Encyclopedia of Genes (KEGG) pathways in the D900 group. (A) Venn diagram of the pathways associated with the differential proteins and metabolites. Arrows indicate shared pathways. (B) The number of proteins and metabolites that exhibit co-participation based on the shared KEGG pathways. Blue indicates the number of proteins, and orange indicates the number of metabolites.

Table 3.

The list of the specific proteins and metabolites involved in the shared pathways between metabolomics and proteomics.

			D900 vs. D60			D900 vs. D300
Pathways		Names	Change fold	P value	Trend	Change fold	P value	Trend
Purine metabolism	Protein	NME3	5.317	0.001	↑	4.468	0.014	↑
		RRM2B	5.200	0.000	↑	1.937	0.006	↑
		AMPD1	0.586	0.008	↓	0.647	0.033	↓
		AMPD3	0.551	0.004	↓	0.535	0.004	↓
	Metabolites	Hypoxanthine	-1.639	0.032	↓	-2.386	0.002	↓
		Inosine	-1.834	0.033	↓	-2.627	0.002	↓
		Guanosine	-1.420	0.007	↓	-1.339	0.005	↓
		Guanine	-1.616	0.005	↓	-1.543	0.002	↓

DISSCUSSION

In recent years, consumers have had a growing awareness of animal products and nutrition. In ducks, appropriate aging is necessary to improve the nutritional value of the meat (Wang et al., 2020a). The meat quality of D60, D300, and D900 duck breast muscle was compared in this study. Meat color is an important apparent parameter in consumer purchasing decisions and is used as an indicator of freshness and wholesomeness (Joo et al., 2013; Onk et al., 2019). Jaturasitha et al. found that the meat of slow-growing chickens was redder than that of fast-growing birds (Jaturasitha et al., 2008). In our present study, we found that meat from D300 and D900 ducks were redder than that of D60 ducks. While we found that the meat from D300 ducks was more redder than D900, which may be influenced by heme pigments (Jaturasitha et al., 2008). Additionally, our results showed that the pH of duck meat at D900 showed no significance compared with D60 and D900 ducks, which was consistent with the results of Weng et al., who reported that the pH of meat of 120-day-old geese was not significantly different from that of 70-day-old geese. Skeletal muscle is the main component of meat, and several researchers have confirmed that fiber characteristics are key determinants of meat quantity and quality (Petracci and Cavani, 2012). In the present study, muscle fibers from D900 and D300 ducks were more inseparable than those from D60 ducks. Although significant differences in duck breast muscle were detected at 60, 300, and 900 d, the mechanisms of duck meat at different ages have rarely been examined.

Because of advances in proteomics technology in recent years, high-resolution and high-throughput MS are routinely used to identify and quantify proteomes in animals (Ji et al., 2017; Wang, et al, 2020a). As expected, breast muscle undergoes dramatic changes in protein levels between 60 and 900 d, and 61 DEPs were detected in D900 ducks. Among these, APOB was the most marked DEP and was upregulated in the D900 group compared with the D60 and D300 groups. APOB is considered to play an important role in determining intramuscular fat and overall meat quality (Li et al., 2018). Moreover, higher nutrient concentrations were necessary for flavor-nutrient associations based on appropriate fat content (Revelle and Warwick, 2009). In this study, we found that the APOB protein level was significantly increased at 900 d, indicating that APOB may affect muscle development and meat flavor through protein changes.

The physicochemical properties of meat, such as the nutritional value and sensory attributes, can determine its quality and acceptability. In general, meat quality depends on proteins (Wei et al., 2019), lipids (Jia et al., 2021b), nucleotides (Jia et al., 2021a), peptides (Moya et al., 2001; Fornal and Montowska, 2019), and small molecule metabolites (Li et al., 2020). However, whether changes in metabolites affect duck meat quality was explored in the current study. Some studies have reported that it is vital for the content of nucleotides and their derivatives to change in fresh meat to improve flavor precursors, which may affect the taste, tenderness, and water-holding capacity of meat (Xiao et al., 2019; Fu et al., 2022). In the present study, the guanosine content in the D900 group was significantly lower than those in the D60 and D300 groups, which was consistent with a previous study that found that the guanosine content in meat is highly correlated with flavor, and meat with lower guanosine content has a higher quality (Tikk et al., 2006; Muroya et al., 2019). However, inosine was found to generate a bitter taste (Mateo and Zumalacárregui, 1996; Wang et al., 2019). This study found that inosine was present at low concentrations in the D900 group compared with that in the D60 and D300 groups. This finding suggests that D900 ducks have a higher nutrient content and a more natural taste.

Hypoxanthine is a purine-based organic compound in muscle tissues and is formed during purine catabolism as a product of the action of xanthine oxidase on xanthine. It is occasionally found as a constituent of nucleic acids (Wang et al., 2022). A previous study showed that hypoxanthine is a key biomarker for poultry quality evaluation (Liu et al., 2021). Based on the metabolomic data (Figure 4), the hypoxanthine content showed significant changes in meat from ducks of different ages, with lower hypoxanthine levels in the D900 group compared with those in the D60 and D300 groups. Furthermore, the guanine content was significantly decreased in meat from older ducks, which may be due to the accumulation of nucleotide products during aging (Subbaraj et al., 2016).

Integrated KEGG pathway analysis has been used to directly examine the internal relationship of proteins and metabolites and thus establish a biochemical reaction network system (Klamt and Stelling, 2003). These results illustrate that purine metabolism pathways affects the meat quality in older ducks, providing insight into the metabolic changes that occur during aging. Interestingly, we found that NME3, RRM2B, AMPD1 and AMPD3, which are involved in these pathways, were significantly increased in D900 ducks compared with D60 and D300 ducks. Therefore, up-regulation of NME3 and RRM2B or down-regulation of AMPD1 and AMPD3 could promote flavor enhancement through the purine metabolism pathway.

Taken together, this study was the first to combine metabolomics and proteomics analyses to distinguish differences in the meat between young and older ducks. One shared signaling pathway was enriched based on the comparative metabolomic and proteomic analyses. Among the pathways examined, purine metabolism was uniquely enriched, which regulated flavor enhancement. In addition, NME3, RRM2B, AMPD1, and AMPD3 may also be potential targets to distinguish young and older ducks. This study provides a multi-omics perspective for comparing the quality of meat from ducks at different ages.