Heating temperatures affect meat quality and vibrational spectroscopic properties of slow- and fast-growing chickens

Abstract

This study determined the effect of water bath cooking (70°C and 90°C for 40 min) and the extreme heat treatment by an autoclave (121°C for 40 min) on the quality of breast meat of a fast-growing chicken, commercial broiler (CB), and slow-growing chickens, Korat chicken (KC), and Thai native chicken (NC) (Leung Hang Khao), by vibrational spectroscopic techniques, including synchrotron radiation-based Fourier transform infrared (SR-FTIR) microspectroscopy and Fourier transform Raman (FT-Raman) spectroscopy. Taste-enhancing compounds, including inosine-5ˊ-monophosphate (IMP) and guanosine-5ˊ-monophosphate (GMP), were better retained in cooked KC and NC meats than in cooked CB meat (P < 0.05). The high heat treatment at 121°C depleted the amount of insoluble collagen in all breeds (P < 0.05). Shear force values of slow-growing chicken meat were not affected by high heating temperatures (P > 0.05). In addition, the high heat treatment increased protein carbonyl (P < 0.05), while no effect on in vitro protein digestibility (P > 0.05). SR-FTIR microspectroscopy performed better in differentiating the meat quality of different chicken breeds, whereas FT-Raman spectroscopy clearly revealed differences in meat qualities induced by heating temperature. Based on principal component analysis (PCA), distinct characteristics of chicken meat cooked at 70°C were high water-holding capacity, lightness (L*), moisture content, and predominant α-helix structure, correlating with Raman spectra at 3,217 cm⁻¹ (O–H stretching of water) and 1,651 cm⁻¹ (amide I; α-helix). The high heating temperature at 90°C and 121°C exposed protein structure to a greater extent, as evidenced by an increase in β-sheets, which was well correlated with the Raman spectra at 2,968 and 2,893 cm⁻¹ (C–H stretching), tryptophan (880 cm⁻¹), tyrosine (858 cm⁻¹), and 1,042, 1,020, and 990 cm⁻¹ (C–C stretching; β-sheet). SR-FTIR and FT-Raman spectroscopy show potential for differentiation of chicken meat quality with respect to breeds and cooking temperatures. The marked differences in wavenumbers would be beneficial as markers for determining the quality of cooked meats from slow- and fast-growing chickens.

INTRODUCTION

Heat treatments are commonly applied in meat processing to achieve microbiological safety and palatability. Different thermal processing techniques can induce changes in meat quality, and improper heat treatments may cause the physicochemical properties, such as appearance, taste, and tough texture, as well as the nutritional properties of meat products, to deteriorate (Wattanachant et al., 2005; Kavitha and Modi, 2007; Qi et al., 2017). Heat-induced denaturation of meat protein and shrinkage of meat fibers and connective tissue influence the texture and cooking loss of cooked meat (Palka and Daun, 1999). Heat treatment also decreases inosine-5ˊ-monophosphate (IMP) content, which is a major umami-related compound in chicken meat due to leaching, degradation, and reaction with other compounds (Kavitha and Modi, 2007). The nucleotide content in slow-growing chickens is typically higher than the nucleotide content in fast-growing chickens, such as broilers (Tang et al., 2009; Jung et al., 2011). However, little information is available on changes in nucleotide content in different chicken breeds upon heating. Thermal process can also affect the digestibility negatively or positively depending on the processing conditions. Higher processing at temperature above 100°C can induce unfavorable structural changes in meat proteins, such as severe formation of intermolecular aggregates and cross-links in proteins which reduce their susceptibility to digestive enzymes and decrease the digestibility (Zhang et al., 2020). In addition, heat treatments generate free radicals, inducing protein oxidation and meat quality deterioration (Estévez and Xiong, 2019). Hence, understanding the effect of thermal processes on meat quality of each chicken breed would be vital to obtain high-quality cooked products.

Vibrational techniques, including infrared and Raman spectroscopy, have been applied as a rapid and nondestructive method for monitoring structural changes in various food proteins (Calabrò and Magazù, 2012; Han et al., 2019). Synchrotron-based Fourier transform infrared (SR-FTIR) microspectroscopy has been applied as a novel technique to study biological tissues as well as meat products (Yu, 2005; Yu et al., 2005). SR-FTIR takes advantage of extremely bright synchrotron light, which is capable of exploring at the cellular level within the microstructure of intact biological tissues at ultraspatial resolution (Yu, 2005). Moreover, Raman spectroscopy has been carried out to investigate structural changes in pork myofibrillar proteins induced by thermal treatments (Xu et al., 2011) and to predict the optimal cooking temperature and cooking time of pork meat (Berhe et al., 2014). Thus, structural changes in meat subjected to various thermal treatments, as revealed by SR-FTIR and FT-Raman spectroscopy, would provide insightful information about the changes at the molecular level.

Korat chicken (KC) is a crossbred between Leung Hang Khao sires (Thai native chicken; NC) and SUT 101 dams (a crossbreed between broiler and layer chicken). According to a previous study, KC and NC had unique textures with high nutritional quality compared to broiler meat (Katemala et al., 2021, 2022). Thus far, information regarding the changes in quality attributes of the KC and NC meat during heating is unknown. Therefore, the present study aimed to investigate the effect of heat treatments on the breast meat quality of slow-growing chickens, KC and NC Leung Hang Khao, and a fast-growing chicken, commercial broiler (CB). Cooking temperatures of 70°C and 90°C were achieved using a water bath, while extremely high thermal treatment of 121°C was carried out using an autoclave. Although heating condition of 121°C is not applied in commercial sterilization of food products, it can be used to illustrate the effect of extreme thermal treatment on chicken meat qualities. Vibrational spectroscopic techniques, including SR-FTIR and FT-Raman spectroscopy, were applied to gain molecular changes of cooked meats under various thermal treatments. In addition, correlations of chicken meat parameters and vibrational spectral data were established.

MATERIALS AND METHODS

Ethics Statement

All procedures used in the present study were approved by the Ethics Committee on Animal Use of Suranaree University of Technology (SUT), Thailand (document ID: U1-02631-2559).

Animals, Sample Preparation, and Heating Procedures

A total of 120 of each genotype, 1-day-old mixed-sex KC and NC (Leung Hang Khao), were raised under the same conditions in 3 separate pens (40 birds/pen/5 m²) in an indoor facility at Suranaree University of Technology Farm (Nakhon Ratchasima, Thailand). The birds were fed ad libitum with the same commercial diet for starter (0–3 wk), grower (4–6 wk), and finisher (7 wk to slaughter) stages, containing 21, 19, and 17% crude protein, respectively. The birds had free access to water and to the outdoor environment. When the birds reached the commercial size at their market age of 10 wk for the KC and 16 wk for the NC, 36 male chickens (12 chicks/pen) of each breed were randomly selected and subjected to total feed withdrawal for 12 to 15 h and weighed (KC: 1.40–1.82 kg and NC: 1.40–2.04 kg). The birds were slaughtered at a commercial slaughterhouse (Nakhon Ratchasima, Thailand) and processed under commercial conditions using electrocution as the stunning system, conventional neck cut, bled, scalded, plucked, and eviscerated. Then, the carcasses were packed in an ice box and brought to the laboratory within 1 h. The whole breast meat samples were collected 24 h postmortem in a 4°C-chiller, and skin, bone, visible connective tissue, and fat were removed. Breast meat samples from 6-wk-old male broilers (CBs) with a live weight of 2.90 to 3.00 kg were obtained from a commercial chicken meat processing plant (Charoen Pokphand Foods Public Company Limited, Nakhon Ratchasima, Thailand). The CB samples were also stored for 24 h in a 4°C-chiller before analysis.

Whole breast meat samples from each treatment were cut into 4.0 × 4.0 × 0.5 cm pieces and divided into 200 g of raw chicken meat, which were packed in a nylon/polyethylene vacuum bag and heated in a water bath set at 70°C and 90°C for 40 min. For extremely high cooking temperature, samples were packed in a nylon/polyethylene vacuum bag and heated in an autoclave set at 121°C for 40 min (Liang et al., 2022). Temperature was monitored by inserting a thermocouple into the middle of 200-g sample packed in nylon/polyethylene vacuum bags, which attained the set temperature in each treatment. Heating was carried out in duplicate for each cooking temperature. Subsequently, cooked samples were immediately placed on ice for 10 min and weighed to determine cooking loss. The moisture content, water-holding capacity (WHC), and shear force were measured within 24 h. The remaining samples were vacuum-packed and kept at 4°C for color measurement within 48 h. For SR-FTIR spectroscopy measurements, the samples were embedded in optimal cutting temperature compound (OCT), snap-frozen in liquid N₂, and stored at −80°C until preparation. The remaining samples were minced, vacuum-packed, and stored at −80°C until further chemical analysis and FT-Raman spectroscopy measurements. Before analysis, frozen meats were thawed in a refrigerator at 4°C for 12 to 18 h.

Cooking Quality Measurement

The moisture content was determined according to AOAC (2010). The color was recorded using a colorimeter (Hunter Associates Laboratory, Reston, VA). All measurements were made in the CIE L*a*b* color space using the D65 illuminant and a 10° observer through a 25-mm aperture. The instrument was standardized with a light trap (black hole) and white tiles before measurement. The color values were expressed as L* (lightness), a* (redness/greenness), and b* (yellowness/blueness). WHC was determined according to Ryoichi et al. (1993) with some modifications. The minced sample (2 g) was placed into a centrifugation tube with filter paper (No. 4; Whatman International Ltd., Maidstone, UK) and then centrifuged at 6,710 × g for 10 min at 25°C. Absorbed moisture in the filter paper was determined. The results were calculated as the amount of water retained in the meat sample per 100 g of water present in the sample before centrifugation (%). Cooking losses were calculated from differences in the weight of raw and cooked meats after heat treatments.

Nucleotides

The nucleotide contents of cooked meat samples were measured according to Kim et al. (2012) with slight modifications. Nucleotides were extracted by homogenizing meat samples (5 g) with 50 mL of 7.5% cold perchloric acid (Ultra Turrax T25; Ika, Werke GmbH & Co., Staufen, Germany). The mixture was centrifuged at 2,000 × g for 5 min at 4°C. The extract was then mixed with 0.6 M neutralizing buffer (pH 7.6; KH₂PO₄ + K₂HPO₄) for 10 min and filtered through a 0.45-µm nylon filter. Nucleotides were separated on a C18 reverse-phase column (Hypersil ODS C18, 4.6 × 150 mm, 3 μm particles, Thermo Scientific, Waltham, MA) equipped with an HPLC system (HP 1260, Agilent Technologies, Inc., Santa Clara, CA). The injection volume was 10 μL and was eluted with mobile phase A containing 150 mM potassium dihydrogen phosphate (KH₂PO₄) and 150 mM potassium chloride (KCl), pH 6, and mobile phase B containing mobile phase A mixed with 20% acetonitrile at a flow rate of 0.5 mL/min. The composition of the mobile phase in gradient elution mode was 97% A for 0 to 5 min, then reduced to 91% A for 5 to 10 min, 80% A for 10 to 15 min, finally 100% B for 15 to 20 min, and finally maintained at 100% B for 5 min. The column temperature was maintained at 25°C, and detection was monitored at 254 nm. The quantities of inosine-5′-monophosphate (IMP), guanosine-5ˊ-monophosphate (GMP), adenosine triphosphate (ATP), adenosine diphosphate (ADP), adenosine-5ˊ-monophosphate (AMP), inosine, and hypoxanthine were calculated using external standards (Sigma-Aldrich Co., St. Louis, MO). In addition, the % retention of IMP and GMP was calculated from the following formula:

$$(\%)\text{Retention}=\left[\mathrm{A}/\mathrm{B}\right]\times 100$$

where A is the nucleotide content in the sample after heating, and B is the nucleotide content in a sample before heating.

Textural Properties

The shear force of cooked samples with a size of 1.0 cm × 2.0 cm × 0.5-cm pieces was evaluated using a Texture Analyzer (TA.XT. Plus, Stable Micro Systems, Surrey, UK) equipped with a Warner-Bratzler shear apparatus attached to a 25-kg load cell at a 2 mm/s crosshead speed (Wattanachant et al., 2004). The force required to move the blade to shear through the longitudinal axis of muscle fibers was measured. The mean value of the 9 replicates for each treatment is presented.

The total collagen content was determined by alkaline hydrolysis as described by Reddy and Enwemeka (1996). Samples were hydrolyzed with 7 M sodium hydroxide (NaOH) at 120°C for 40 min. The hydrolysate was neutralized with 3.5 M sulfuric acid (H₂SO₄), filtered, and reacted with chloramine T solution and Ehrlich's reagent. Absorbance was measured at 550 nm using a spectrophotometer (Jenway, Bibby Scientific Ltd., Stanffordshore, UK). The amount of hydroxyproline was determined, and the total collagen content was calculated using a coefficient of 7.25 (Bergman and Loxley, 1963).

The insoluble collagen content was determined according to the method of Liu et al. (1996). Meat samples were homogenized with 25% Ringer's solution. The homogenates were cooked at 77°C for 70 min in a water bath and centrifuged at 2,300 × g at 4°C for 30 min (Sorvall Legend MACH 1.6R, Thermo Electron LED GmbH, Lengensellbold, Germany). The extraction was repeated twice, and then the residues were dried overnight at 105°C. The hydroxyproline content of the dried residue was determined, and the collagen content was calculated as described above to represent the insoluble collagen content.

Protein Oxidation

Protein oxidation was measured according to Mercier et al. (2004) with minor modifications. Cooked samples were homogenized in 20 mM phosphate buffer (pH 6.5) using an Ultraturrax homogenizer at 10,000 rpm for 30 s (Ultra turrax T25, Ika Werke GmbH & Co., Staufen, Germany). Carbonyl groups were detected with 2,4-dinitrophenylhydrazine (DNPH) to form protein hydrazones. Two equal aliquots of meat homogenate were precipitated with 20% trichloroacetic acid (TCA) and centrifuged at 4,500 × g for 5 min at 4°C. One pellet was treated with 0.2% (w/v) DNPH dissolved in 2 M HCl, and the other was treated with 2 M HCl as a blank. Subsequently, the samples were further centrifuged for 5 min at 4,500 rpm at 4°C. DNPH was removed by washing with 10% TCA, and then a mixture of ethanol:ethyl acetate (1:1) was added until a clear solution was obtained. The pellets were dried with N₂ gas and finally solubilized in 20 mM sodium phosphate buffer (pH 6.5) containing 6 M guanidine hydrochloride. The amount of carbonyl was read at 370 nm for protein hydrazone and expressed as nanomole of carbonyl per mg protein using the extinction coefficient of 21.0 mM/cm. Bovine serum albumin was used as a protein standard.

In Vitro Protein Digestibility

Protein digestibility was assessed according to Minekus et al. (2014) with slight modifications. Cooked meat samples (2 g) were homogenized in 4 mL of simulated gastric fluid (SGF). The homogenate was adjusted to pH 3.0, and then pepsin solution (454.44 units/mg solid, porcine gastric mucosa; P7000, Sigma-Aldrich Co.) and CaCl₂ were added to achieve 2,000 U/mL and 75 µM of the final mixture. The mixture was incubated in a shaking water bath at 37°C and 150 rpm for 2 h. Digestion was ceased by adjusting the pH to 7.0. The gastric digested sample (∼5 mL, pH 7.0) was mixed with 5 mL simulated intestinal fluid (SIF). The pH was readjusted to 7.0 using 1 M NaOH before the addition of 100 µL of pancreatin (5.57 U TAME/mg solid) (P7545, Sigma-Aldrich Co.) was added to achieve 100 U/mL of final digestion mixture, and then CaCl₂ was added to reach 0.3 mM. Intestinal digestion was performed in a shaking water bath at 37°C and 150 rpm for 2 h, stopped by heating in a water bath at 95°C for 10 min, and placed in an ice bath for 10 min. The final volume of digested samples was recorded. After centrifugation at 4,500 × g and 4°C for 35 min, the supernatants were collected and stored at −20°C. The digested samples were analyzed for α-amino groups produced during digestion using the 2,4,6-trinitrobenzene sulfonate (TNBS) method as described by Alder-Nissen (1979). The control (blank) was treated similarly but used deionized water instead of a meat sample. The total α-amino group was determined by hydrolyzing samples with 6 N HCl at 120°C for 24 h before analysis of α-amino groups. The in vitro protein digestibility (%) was calculated using the following equation:

$$\text{Digestibility}(\%)=\left[\left(\mathrm{A}{\mathrm{N}}_{\mathrm{s}}-\mathrm{A}{\mathrm{N}}_b\right)/\mathrm{A}{\mathrm{N}}_{\mathrm{t}}\right]\times 100$$

where AN_s is the α-amino group of digested samples, AN_b is the α-amino group content of the blank, and AN_t is the total α-amino group content in a sample.

Vibrational Spectroscopy

SR-FTIR Measurement. Infrared spectra were collected at room temperature using an SR-FTIR (Hyperion 2000, Bruker Optics Ltd., Ettlingen, Germany). Cooked meat samples were embedded in optimal cutting temperature (OCT) compound, snap-frozen in liquid N₂, and stored at −80°C until cut into 6-µm thick sections by a cryomicrotome (Microm HM525; Thermo Fisher Scientific, Walldorf, Germany). Then, the specimens were placed on barium fluoride (BaF₂) windows and dried in a vacuum chamber overnight before SR-FTIR analysis. The samples were measured in transmission mode under an infrared microscope 15 × objective lens equipped with an MCT D315 detector cooled with liquid nitrogen. Synchrotron radiation from the storage ring at beamline BL4.1 (Infrared Spectroscopy and Imaging) of the Synchrotron Light Research Institute (SLRI) (Nakhon Ratchasima, Thailand) was used for infrared emission. The measurement was performed in the mapping mode in the area of 200 × 200 µm² over wavenumbers ranging from 4,000 to 800 cm⁻¹ using an aperture size of 10 × 10 µm with 64 scans coadded at a resolution of 4 cm⁻¹. The spectra were subtracted from the background air spectrum in each scan. Spectral acquisition and instrument control were performed using OPUS Software 7.2 (Bruker Optics Ltd., Ettlingen, Germany). At least 30 spectra were collected from each sample and averaged to represent one spectrum. Three replications of independent lots were carried out for each chicken breed and cooking temperature. The spectrometer was purged with N₂ to reduce spectral contributions from water vapor. The background spectrum of BaF₂ windows was recorded. The baseline was estimated using automated background removal. IR spectra were also preprocessed by the water compensation method in OPUS 7.2 software (Bruker Optics Ltd.).

FT-Raman Measurement. Raman data were acquired using a Bruker Vertex 70 FT-Raman spectrometer (Bruker, Karlsruhe, Germany). Frozen samples were thawed at 4°C overnight and equilibrated to room temperature before being packed in a sample holder for Raman spectroscopy measurement. Spectra were recorded over the range of 4,000 to 400 cm⁻¹ at a resolution of 4 cm⁻¹ and 256 scans. Sulfur was used to calibrate the Raman frequency. Spectra were excited with the 1,064 nm Nd:YAG laser line with 500 mW of laser power. FT-Raman spectral acquisition and instrument control were performed using OPUS 7.2 software (Bruker Optics Ltd.). A total of 30 spectra were collected per sample.

Spectra Processing and Analysis

Spectra were processed using the Savitzky-Golay algorithm with 17 smoothing points and baseline correction and then normalized using extended multiplicative signal correction (EMSC) for the IR spectra in the ranges of 3,801 to 2,704 and 1,802 to 899 cm⁻¹ and for Raman in the ranges of 3,801 to 2,704 and 1,803 to 399 cm⁻¹. The relative integral peak area of the second derivative spectra of SR-FTIR and integral peak intensities of the FT-Raman spectra were analyzed using OPUS 7.2 software (Bruker Optics Ltd.)

Before analysis of protein secondary structure, IR and Raman spectra were preprocessed by smoothing by applying 13 points, vector normalization against the amide I region (1,700−1,600 cm⁻¹), and baseline scattering correction. The secondary structure of proteins was determined as the percentage of α-helix, β-sheet, β-turn, and random coil by curve fitting in 1,700 to 1,600 cm⁻¹ (amide I region) of SR-FTIR and FT-Raman spectra using appropriate Gaussian and Lorentzian functions in OPUS 7.2 software.

Principal Component Analysis

Principal component analysis (PCA) with spectral ranges of 3,800 to 2,704 cm⁻¹ and 1,800 to 899 cm⁻¹, representing clear peak separation, was performed using Unscrambler X 10.5.1 (CamonAnalytics, Oslo, Norway) to identify significant variations in the spectral dataset. Seven principal components (PCs) were chosen for analysis. Subsequently, the selected high loading wavenumbers obtained from the loading plot, secondary protein structure (%), and meat quality traits were subjected to PCA with the weighting method of 1/standard deviation (SD) to evaluate correlation among variables.

Statistical Analysis

All analytical experiments were performed in triplicate. Data were evaluated statistically using the SPSS version 16.0 program (SPSS Inc., Chicago, IL). Analysis of variance (ANOVA) was conducted to determine the significance among treatments (breed and heating temperature) on the parameters determined. Differences between treatment means were analyzed for significance (P < 0.05) using Tukey's test.

RESULTS AND DISCUSSION

Cooking Qualities

High heat treatment at 121°C led to a decrease in the L* (lightness) value and an increase in the b* (yellowness) value in all 3 chicken breeds (P < 0.05, Table 1). The cooked KC meat exhibited the lowest L* value at a cooking temperature of 121°C (P < 0.05). KC and NC meat samples cooked at 121°C were found to have the highest a* (redness) value, whereas the lowest a* value was found in KC and NC meat cooked at 90°C (P < 0.05). Differences in chicken meat color during heating might be related to the different contents and chemical states of myoglobin and an effect of light scattering from denatured proteins (Purslow et al., 2020). High thermal treatment at 121°C induced the formation of denatured globin hemichrome (ferrihemochrome), which has the dull-brown appearance of cooked meat (Suman et al., 2016). KC and NC meat displayed more redness and yellowness (P < 0.05) but lower lightness than CB meat after heating at 121°C possibly due to higher cooking loss in CB meat (P < 0.05, Table 1), which leads to higher loss of soluble myoglobin. A higher myoglobin content was reported in older birds (Ortiz et al., 2021), consistent with more redness in cooked KC and NC meat, which are older than CB.

Table 1.

Cooking qualities of breast meat from 3 chicken breeds under different thermal treatments (mean ± SD).

Heat treatments (°C)	Chicken breeds	Moisture (%)	L*	a*	b*	WHC (%)	Cooking loss (%)
	CB	72.51 ± 0.95a	80.73 ± 2.88a	2.31 ± 0.54cd	16.23 ± 1.02bc	43.23 ± 2.72bc	20.90 ± 2.99c
70	KC	72.61 ± 0.11a	79.67 ± 1.57a	2.33 ± 0.82cd	14.56 ± 1.39d	52.22 ± 0.08a	5.59 ± 0.33e
	NC	72.09 ± 0.09a	79.34 ± 0.76a	1.82 ± 0.59de	15.30 ± 1.55cd	52.56 ± 1.28a	11.67 ± 0.49d
	CB	67.90 ± 0.52b	79.43 ± 2.46a	2.38 ± 0.74cd	16.25 ± 1.43bc	37.02 ± 1.96e	32.87 ± 0.89ab
90	KC	68.51 ± 0.20b	78.69 ± 2.19a	1.51 ± 0.47e	14.59 ± 1.40d	43.50 ± 0.14bc	23.00 ± 1.57c
	NC	68.23 ± 0.29b	79.25 ± 2.18a	0.54 ± 0.55f	15.44 ± 1.04cd	45.94 ± 2.95b	24.47 ± 0.87c
	CB	66.07 ± 0.62c	76.09 ± 2.81b	2.95 ± 0.50bc	17.11 ± 1.45b	37.28 ± 2.91de	37.15 ± 0.91a
121	KC	65.74 ± 0.26cd	72.11 ± 2.55c	4.35 ± 0.97a	21.39 ± 1.97a	38.38 ± 0.23de	32.50 ± 0.67ab
	NC	64.99 ± 0.30d	74.38 ± 1.53b	3.62 ± 0.93b	21.75 ± 1.98a	40.51 ± 1.39cd	33.59 ± 1.01ab

n = 6.

Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken.

^a–fMean value in the same column with different superscripts differ significantly (P < 0.05).

Thermal treatment caused a decrease in moisture content and WHC in all 3 chicken breeds (P < 0.05, Table 1). A higher WHC in cooked slow-growing chickens (NC and KC) could be related to a higher content of collagen (P < 0.05), which denatures and retains more water (Maximo and Cunha, 2010). Lower WHC in cooked CB meat is associated with higher cooking loss (Table 1). Higher cooking loss in CB meat might be associated with its higher lipid content that could be lost to a greater extent during cooking (Puente et al., 2019; Katemala et al., 2022). Chumngoen and Tan (2015) also reported similar finding that CB breast meat cooked in a 85°C water bath until the core temperature reached 80°C exhibited higher cooking loss than Taiwan native chicken. Considering these results, low heat treatment showed better color associated with high moisture retention. Our results revealed that breast meat of slow-growing chickens showed higher WHC and lower cooking loss than fast-growing chicken meat.

Nucleotides

IMP is a major nucleotide found in chicken meat and is considered as one of major precursors responsible for umami taste (Jayasena et al., 2013). IMP content of all treatments ranged from 2023.03 to 3218.97 mg/100 g of dry samples (Table 2). Inosine was the second most abundant nucleotide, which showed higher levels in meat from all breeds cooked at 70°C than those cooked at 90°C and 121°C. Higher contents of hypoxanthine were also observed in all breast meats cooked at 90°C and 121°C. ATP, AMP, and ADP were detected at low concentrations in all samples (Table 2). Degradation products of IMP are inosine and phosphate group, and inosine can further be degraded to hypoxanthine and ribose (Tikk et al., 2006). Katemala et al. (2021) reported that the nucleotide content in CB, KC, and NC varied with breed, age, and muscle types. This study reveals that cooking temperature is a significant factor affecting IMP degradation. The mild heating temperature in a 70°C-water bath resulted in the lowest hypoxanthine content concomitant with the highest inosine content (P < 0.05, Table 2). These results suggested that degradation of IMP to inosine and hypoxanthine occurred to a lesser extent in mild cooking (70°C). It was also observed that IMP retention was higher in slow-growing KC meat than in fast-growing CB at both cooking temperatures of 70°C and 90°C (P < 0.05, Figure 1). It has been documented that indigenous chickens, such as KC in Thailand (Katemala et al., 2021), Korean native chicken (KNC) in Korea (Jung et al., 2011), Hinai-jidori chicken in Japan (Rikimaru and Takahashi, 2010), and slow-growing Wenchang and Xianju genotypes in China (Tang et al., 2009) contained higher IMP compared to that of commercial broilers.

Table 2.

Nucleotide contents of cooked breast meats from slow- and fast-growing chickens under different heat treatments (mean ± SD).

Heat treatments (°C)	Chicken breeds	Nucleotides content (µg/g sample, db)
		IMP	GMP	ATP	ADP	AMP	Inosine	Hypoxanthine
	CB	2023.03 ± 288.41b	147.78 ± 22.86d	89.20 ± 27.54ab	146.06 ± 84.93cd	110.67 ± 8.39ab	1586.18 ± 175.74b	126.06 ± 5.29b
70	KC	2981.04 ± 186.90a	109.94 ± 6.05de	71.41 ± 5.82b	378.29 ± 9.64a	119.33 ± 4.16a	1555.96 ± 111.09b	25.58 ± 5.07c
	NC	2429.80 ± 336.00b	60.64 ± 9.35e	ND	329.02 ± 27.79b	101.01 ± 3.91b	1798.91 ± 122.85a	91.18 ± 11.56b
	CB	2437.81 ± 179.66b	141.13 ± 20.87d	87.52 ± 33.55b	121.41 ± 19.70cde	71.55 ± 8.56cd	930.92 ± 93.67e	904.20 ± 137.40a
90	KC	3218.97 ± 99.83a	115.06 ± 7.48de	96.57 ± 5.01ab	170.41 ± 6.56c	76.37 ± 6.54cd	1029.13 ± 48.99de	745.83 ± 279.15a
	NC	3074.19 ± 303.76a	221.75 ± 42.81c	34.90 ± 20.22c	102.17 ± 16.11de	78.73 ± 11.52c	1134.74 ± 191.21cd	744.07 ± 71.95a
	CB	2092.62 ± 211.08b	240.99 ± 13.33c	76.98 ± 32.43b	89.63 ± 12.53e	69.14 ± 11.14cde	1036.25 ± 22.13de	969.36 ± 260.00a
121	KC	3017.16 ± 308.92a	1560.99 ± 72.27a	113.70 ± 10.42a	111.44 ± 13.11de	65.74 ± 2.83de	1043.20 ± 56.01de	969.36 ± 259.10a
	NC	2038.58 ± 475.87b	1427.49 ± 44.24b	19.53 ± 1.27cd	86.50 ± 5.27e	58.11 ± 1.03e	1288.63 ± 63.68c	969.36 ± 259.10a

n = 6.

Abbreviations: CB, commercial broiler; db, dry basis; KC, Korat chicken; NC, Thai native chicken; ND, Not detected.

^a–eMean value in the same column with different superscripts differ significantly (P < 0.05).

Figure 1.

Retention of inosine-5ˊ-monophosphate in breast meat samples of 3 chicken breeds cooked under different heat treatments. Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken. ^a-b in different superscripts indicate differences in mean values (P < 0.05). Bars indicate standard deviation.

Collagen Content

The cooked NC meat exhibited the highest total collagen content at a low heating condition at 70°C (P < 0.05, Figure 2A). The total collagen content of cooked NC meat decreased with temperature, while total collagen content of CB and KC appeared to be constant at heating treatments of 70°C to 121°C. Cooking at 60°C to 70°C causes shrinkage and gelatinization of connective collagen, while cooking at higher temperatures of 80°C to 121°C leads to higher loss of soluble collagen content, yielding more compact structure (Palka, 1999). Significant loss of total and insoluble collagen content of NC meat cooked by an autoclave and by a 90°C heating (P < 0.05, Figure 2A and B) implied extensive disintegration of cross-linked collagen under high thermal treatments. Our study revealed that slow-growing chickens contained higher amounts of insoluble collagen content in breast meat than did fast-growing CB meat at all temperatures tested except at 121°C (P < 0.05, Figure 2B), possibly due to the breed and age variation. Thermally stable cross-linked collagen increases with the age of the animal, causing toughness of meat (Petracci and Cavani, 2012). KC and NC were 4 and 10 wk older than CB, respectively, which likely possessed higher stable cross-linked collagens, corresponding to higher contents of insoluble collagen at heating temperatures of 70°C and 90°C (P < 0.05, Figure 2B). This also coincided with higher shear force of NC when compared to CB. Jeong et al. (2020) also demonstrated that 12-wk-old Korean native chicken and 72-wk-old laying hen exhibited higher insoluble collagen content in raw breast meat than did CB, and the formers showed higher shear force value after being subjected to retort sterilization at 121°C. Unfortunately, total collagen and insoluble collagen content of cooked meat samples were not determined in their study. Our study revealed that high cooking temperature at 121°C resulted in the highest loss of insoluble collagen content in all breeds (P < 0.05). Lin et al. (2022) demonstrated that extended cooking of pig trotters at 95°C resulted in greater reduction of collagen cross-links, leading to softer texture. Cross-linked collagen appeared to be critical parameters governing textural properties of cooked meat, which should be further investigated in cooked meat of these 3 breeds.

Figure 2.

Total collagen (A) and insoluble collagen (B) content of cooked chicken breasts from 3 breeds under different heat treatments. Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken. ^a–e in different letters indicate differences between means (P < 0.05). Bars indicate the standard deviation. db, dry basis.

Textural Properties

At 70°C and 90°C, the shear force values of all 3 species were comparable (P > 0.05, Figure 3). In addition, shear force values of the KC and NC meat samples cooked at 121°C and 90°C were comparable (P > 0.05). Significant reduction in shear force was observed in CB meat cooked at 121°C when compared to slow-growing chickens (P < 0.05). The lower insoluble collagen content in CB resulted in a higher loss of mechanical strength from cross-linking collagen. The fibers in CB meat are weaker and pull apart more easily, leading to a lower shear force value than KC and NC meat, which are stabilized and reinforced by the cross-linking of collagen. Wattanachant et al. (2005) also reported that the perimysium and endomysium of broiler muscles melted after cooking at 80°C; however, only slight disintegration was observed in the indigenous chicken muscles. KC and NC meat samples were tougher than CB meat. The muscle structures of the KC and NC chicken muscles appeared to be less affected by high cooking temperatures when compared to the CB muscles.

Figure 3.

Shear force values of cooked breast meat samples of 3 chicken breeds under different heat treatments. Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken. ^a–c of different superscripts indicate differences in mean values (P < 0.05). Bars indicate standard deviation.

Protein Oxidation

The carbonyl content of cooked meat increased with heating temperatures in all breeds, particularly in KC meat after heating at 121°C (P < 0.05, Figure 4), indicating that protein oxidation was induced by thermal treatment. High thermal treatments generate more free radicals and decrease antioxidant protection, leading to more oxidation (Traore et al., 2012). Higher carbonyl group formation during heating was observed in slow-growing chickens than in fast-growing CB meat. Heating leads to myoglobin denaturation and oxidative cleavage of hematin pigment, resulting in the release of iron from the heme molecules, which further promotes the formation of free radicals (Traore et al., 2012). The higher protein oxidation in KC and NC might be due to the higher heme content in older chickens, which catalyzes lipid and protein oxidation to a greater extent.

Figure 4.

Carbonyl content of cooked breast meat samples from 3 chicken breeds under different heat treatments. Abbreviations CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken. ^a–d in different superscripts indicate differences in mean values (P < 0.05). Bars indicate the standard deviation. db, dry basis.

In Vitro Protein Digestibility

KC meat cooked at 90°C revealed higher protein digestibility than CB and NC meat cooked at 121°C (P < 0.05, Figure 5). Thermal denaturation under mild conditions partially exposes cleavage sites to digestive enzymes and might improve digestibility by increasing susceptibility to gastrointestinal proteases (Zhang et al., 2020; Mitra et al., 2022). Mitra et al. (2022) showed that the digestion rate of pork cooked at a lower temperature (58°C for 72 min) in both the gastric and intestinal phases was higher than the digestion rate of pork cooked in an oven at 160°C for 72 min. Under extreme heating conditions, a high level of protein oxidation also leads to cross-linking and aggregation, reducing proteolysis (Rysman et al., 2016). An increase in carbonyl content of samples cooked at higher heating temperatures, as shown in Figure 4, could partly contribute to lower digestibility. Moreover, heating induces shrinkage of protein, allowing inter- and intraprotein interactions and resulting in a denser protein structure, which can reduce enzyme accessibility and protein digestibility (Zhang et al., 2022).

Figure 5.

In vitro protein digestibility of cooked breast meat samples from 3 chicken breeds under different heat treatments. Abbreviations CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken. ^a–c in different superscripts indicate differences in mean values (P < 0.05). Bars indicate standard deviation.

SR-FTIR and FT-Raman Spectroscopy Coupled With Principal Component Analysis

Typical SR-FTIR and FT-Raman spectra collected from cooked chicken meats at different heat treatments give overall fingerprints, as shown in Figure 6A and B, respectively. A summary of tentative assignments of the discriminatory SR-FTIR and FT-Raman bands following the literature is given in Table 3. The amide I region at 1,700 to 1,600 cm⁻¹ is a more prominent feature, showing differences among samples with more distinct SR-FTIR spectra than their FT-Raman counterparts (Figure 6A and B). Obvious differences in SR-FTIR spectra were also observed at ∼1,250 cm⁻¹ (Böcker et al., 2007) and 1,080 to 1,050 cm⁻¹ (Deniz et al., 2018), and those in FT-Raman spectra were observed at ∼3,200 cm⁻¹ (Herrero, 2008b), associated with a random coil structure in the amide III region, C–O stretching of glycogen, and O–H stretching of water, respectively.

Figure 6.

Average SR-FTIR (A) and FT-Raman spectra (B) of cooked chicken breasts from 3 chicken breeds under different heat treatments. Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken.

Table 3.

Bands assignment of vibrational spectra of cooked breast meats from 3 chicken breeds under different thermal treatments.

Wavenumber (cm−1)
IR	Raman	Vibrational modes	Molecules
	3217, 3208	O–H stretching	Water
2929, 2862	2968, 2893	C–H stretching	Protein, lipid
1747	1747	C=O stretching	Lipid
1657–1648	1657–1645	Amide I (α-helix)	Protein
1695–1674, 1640–1610	1680–1665, 1640–1612	Amide I (β-sheet)	Protein
1657–1642	1665–1660	Amide I (random coil)	Protein
1686–1662	1690–1680	Amide I (β-turn)	Protein
	1553	Tryptophan	Protein
1564		Aggregated β-sheet	Protein
1462		CH2 bending	Lipid
	1454, 1450, 1340	C–H bending	Protein
1307		Amide III (α-helix)	Protein
	1261	Amide III (random coil)	Protein
1250		Amide III (random coil)	Protein
1080–1050		C–O stretching	Glycogen
	1042, 1020, 1007	C–C stretching (β-sheet)	Protein
	990, 960	C–C stretching (random coil)	Protein
	938	C–C stretching (α-helix)	Protein
	880	Tryptophan	Protein
	858, 850, 830	Tyrosine	Protein
	720, 645, 620	C–S stretching	Protein
	530, 527	S–S stretching (gauche-gauche-trans)	Protein

Peak assignments were obtained from Dàvila et al., 2006 ; Marinkovic and Chance, 2006; Böcker et al., 2007; Herrero, 2008a,b; Berhe et al., 2014; Deniz et al., 2018; Candoğan et al., 2021.

Changes in the peak area of the second derivative spectra of SR-FTIR and the peak intensities of the FT-Raman bands occur from cooked chicken meats at different heat treatments, as shown in Table 4. The bands of SR-FTIR at 1,307 cm⁻¹ assigned to α-helical structures of amide III regions of myofibrillar proteins decreased concomitantly with the increased absorption band at 1,250 cm⁻¹ assigned to random coils from amide III (Böcker et al., 2007), suggesting a loss of the native protein with high heat treatment temperature. In addition, the absorption of SR-FTIR bands at 1,080 to 1,050 cm⁻¹ indicated that C–O stretching of glycogen was predominantly found in cooked NC meat (Deniz et al., 2018). The FT-Raman spectrum of water at 3,208 cm⁻¹ (O–H stretching) (Herrero, 2008b) decreased with high heat treatment temperature, and the lowest intensity was found in KC and NC meat at a heating temperature of 121°C (P < 0.05, Table 4). This result is consistent with a decrease in moisture content in cooked chicken meat as shown in Table 1. In contrast, the cooked chicken meat at high heat treatment temperature showed a positive contribution from the random coil structure (1,261 and 960 cm⁻¹) (Dàvila et al., 2006; Berhe et al., 2014) and the ratio of the tyrosyl doublet (850 and 830 cm⁻¹) (Herrero, 2008b), indicating an increase in protein unfolding. The tryptophan (Trp) residue band vibrations at 1,553 and 1,340 cm⁻¹ and aliphatic residues assigned at 1,450 cm⁻¹ (CH₂ and CH₃ bending) (Herrero, 2008b) of Raman spectra decreased with heating temperature, suggesting an increase in hydrophobic interactions in cooked meat at high heating temperatures. The Raman integral area at 938 cm⁻¹ assigned to C–C stretching of the α-helix structure (Herrero, 2008b) of meat cooked at lower heating temperatures showed greater values than meat cooked at 121°C. Pedersen et al. (2003) reported that the WHC in pork meat was predicted by the Raman spectra regions at 3,128 to 3,071 and 951 to 876 cm⁻¹ measured within an hour after slaughter. This result is consistent with our study in Table 1, in which WHC was negatively correlated with heating temperature (P < 0.05). Moreover, high intensities of Raman spectra at 720, 645, 620 cm⁻¹, and 530 cm⁻¹ assigned to the C–S bond of the methionine and cysteine residues and S‒S stretching of disulfide bonds in the “gauche-gauche-trans” (g-g-t) conformation (Herrero, 2008b) were predominant in meat cooked at 70°C. Overall, the spectral data revealed the most obvious changes in protein structure upon heat treatment temperature with conversion of α-helices into random coils as well as intense hydrophobic interactions among unfolded muscle proteins upon heating.

Table 4.

Integral peak area (×10⁻²) of second derivative spectra of SR-FTIR and integral peak intensities of FT-Raman spectra obtained from cooked breast meats from 3 chicken breeds under different heat treatments (mean ± SD).

Wavenumber (cm−1)	Thermal treatment
	70°C			90°C			121°C
	CB	KC	NC	CB	KC	NC	CB	KC	NC
SR-FTIR
1080–1050	0.13 ± 0.05ab	0.14 ± 0.01ab	0.22 ± 0.11a	0.09 ± 0.01b	0.09 ± 0.02b	0.10 ± 0.02b	0.08 ± 0.01b	0.09 ± 0.01b	0.12 ± 0.00ab
1307	0.15 ± 0.01ab	0.15 ± 0.01ab	0.19 ± 0.04a	0.13 ± 0.01b	0.10 ± 0.03bc	0.13 ± 0.00b	0.06 ± 0.00c	0.10 ± 0.03bc	0.06 ± 0.00c
1250	0.23 ± 0.12c	0.50 ± 0.06bc	0.42 ± 0.04bc	0.57 ± 0.19b	0.49 ± 0.15bc	0.31 ± 0.00bc	0.56 ± 0.09b	0.56 ± 0.16b	0.85 ± 0.02a
FT-Raman
3208	0.77 ± 0.11a	0.84 ± 0.05a	0.73 ± 0.04a	0.44 ± 0.05bc	0.35 ± 0.02bc	0.50 ± 0.02b	0.28 ± 0.20c	0.02 ± 0.03d	0.00 ± 0.00d
1553	0.22 ± 0.01cd	0.26 ± 0.01a	0.24 ± 0.01ab	0.20 ± 0.00d	0.24 ± 0.00ab	0.25 ± 0.01a	0.20 ± 0.01d	0.22 ± 0.01bc	0.21 ± 0.01cd
1450	3.32 ± 0.11ab	3.40 ± 0.13a	3.12 ± 0.14bc	2.63 ± 0.15e	2.95 ± 0.03cd	2.99 ± 0.06cd	2.78 ± 0.17de	2.79 ± 0.05de	2.60 ± 0.04e
1340	0.19 ± 0.00ef	0.33 ± 0.01a	0.29 ± 0.01b	0.20 ± 0.03de	0.25 ± 0.01c	0.25 ± 0.01c	0.17 ± 0.01f	0.23 ± 0.01cd	0.24 ± 0.00c
1261	0.46 ± 0.05bc	0.41 ± 0.04bc	0.43 ± 0.04bc	0.51 ± 0.05b	0.70 ± 0.02a	0.70 ± 0.02a	0.68 ± 0.05a	0.77 ± 0.01a	0.72 ± 0.01a
960	0.00 ± 0.00e	0.00 ± 0.00e	0.00 ± 0.00e	0.04 ± 0.01cd	0.05 ± 0.00bcd	0.03 ± 0.01d	0.06 ± 0.00ab	0.07 ± 0.00a	0.06 ± 0.01abc
938	0.13 ± 0.01cd	0.32 ± 0.01a	0.25 ± 0.02b	0.05 ± 0.00ef	0.08 ± 0.01de	0.16 ± 0.07c	0.00 ± 0.00f	0.05 ± 0.00ef	0.05 ± 0.00ef
720	0.15 ± 0.01a	0.16 ± 0.01a	0.15 ± 0.01a	0.12 ± 0.01bc	0.14 ± 0.01ab	14.04 ± 0.02ab	0.10 ± 0.01c	0.11 ± 0.00c	0.10 ± 0.01c
645	0.13 ± 0.00ab	0.14 ± 0.01a	0.13 ± 0.01abc	0.11 ± 0.00c	0.12 ± 0.01abc	0.12 ± 0.01abc	0.12 ± 0.01bc	0.12 ± 0.00abc	0.11 ± 0.01c
620	0.08 ± 0.00ab	0.09 ± 0.01a	0.07 ± 0.01bc	0.06 ± 0.01cd	0.07 ± 0.00bc	0.07 ± 0.01bc	0.06 ± 0.00cd	0.06 ± 0.00cd	0.05 ± 0.01d
530	0.22 ± 0.01b	0.27 ± 0.01a	0.25 ± 0.00a	0.16 ± 0.01c	0.21 ± 0.01b	0.22 ± 0.01b	0.13 ± 0.01d	0.17 ± 0.01c	0.16 ± 0.00c
Tyr db*	0.92 ± 0.05c	1.00 ± 0.11c	0.93 ± 0.07c	1.01 ± 0.06c	0.97 ± 0.00c	0.86 ± 0.16c	1.42 ± 0.07b	1.67 ± 0.11a	1.60 ± 0.05ab

Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken.

Tyr db, Tyrosine doublet = peak intensity at 850 cm⁻¹/830 cm⁻¹.

^a–fMean values with different superscripts in the same row within each vibrational spectroscopy technique differ significantly (P < 0.05).

Quantitative information about the secondary structure of the protein was obtained from the amide I profile of SR-FTIR spectra. β-Sheet is the main protein secondary structure found in cooked chicken meat, followed by α-helix structure (Table 5). In addition, the small amount of random coils and β-turns in chicken meat appeared to increase with heating temperature. The secondary structure of protein obtained from amide I of SR-FTIR spectra revealed that α-helix decreased, while β-sheet increased with heating temperature. NC meat contained the lowest α-helix content and the highest β-sheet structure upon cooking at 121°C (P < 0.05, Table 5). A high heating temperature of 121°C decreased the α-helix structure to a greater extent than a mild heating temperature in water bath of 70°C (P < 0.05). Changes in the amide I band obtained from SR-FTIR were more noticeable than the changes in the amide I band in the FT-Raman spectra (Figure 6A and B). This study revealed that the amide I profile of SR-FTIR spectra showed better differentiation of the protein secondary structure of cooked chicken meats from these 3 breeds. Differences in protein secondary structure profiles and the ratio of α-helix and β-sheet partially explain their meat quality. Beattie et al. (2004) showed a positive correlation between α-helical and tenderness, whereas β-sheet structure was positively correlated with meat toughness, coinciding with the higher shear force value of cooked NC meat than the β-sheet structure in CB and KC meat at high heat treatment (121°C) (Figure 3).

Table 5.

Change of secondary structures of cooked breast meats from 3 chicken breeds under different heat treatments obtained from SR-FTIR spectra (mean ± SD).

Cooking temperature (°C)	Chicken breeds	Secondary structures (%)
		α-Helix	β-Sheet	Random coil	β-Turn
	CB	32.82 ± 1.42a	33.57 ± 0.96d	17.69 ± 1.41abc	15.91 ± 1.40bcde
70	KC	27.95 ± 0.48cd	37.83 ± 0.97b	15.13 ± 0.39c	19.09 ± 1.50ab
	NC	27.37 ± 0.25cd	38.47 ± 1.84b	21.06 ± 2.97a	13.10 ± 2.23e
	CB	31.23 ± 0.88ab	34.82 ± 0.59cd	16.73 ± 0.97bc	17.22 ± 0.95abcd
90	KC	26.57 ± 1.30cde	37.49 ± 1.32bc	16.79 ± 1.30bc	19.15 ± 0.99ab
	NC	25.29 ± 1.41de	39.82 ± 1.49b	20.19 ± 1.94ab	14.69 ± 0.94de
	CB	29.08 ± 0.72bc	37.49 ± 0.36bc	15.11 ± 0.70c	18.32 ± 0.62abc
121	KC	24.23 ± 1.27e	38.32 ± 0.35b	17.73 ± 1.02abc	19.72 ± 0.26a
	NC	21.76 ± 1.10f	43.60 ± 0.88a	18.90 ± 1.08abc	15.74 ± 1.28cde

Abbreviations: CB, commercial broiler; KC, Korat chicken; NC, Thai native chicken.

^a–fMean values with different superscripts in the same column differ significantly (P < 0.05).

PCA was carried out to classify and correlate the relationship between vibrational spectral data (SR-FTIR and FT-Raman) and the quality of cooked chicken meat from different heat treatments. The results of the score plot of the first and second principal components showed that PCA clearly distinguished among different heating temperatures and chicken breeds (Figure 7A). Two main principal components (PCs) were extracted that accounted for 67% of the total variability (PC-1 explained 44% of the variance, and PC-2 explained 23% of the variance, Figure 7A). Meats cooked at high cooking temperatures at 90°C and 121°C are separated from those cooked at low (70°C) temperature along PC-1, while the differences between NC from CB and KC meat are visible in PC-2 (Figure 7A). PC-1 showed stepwise structural changes in meat cooked with different methods at 70°C to 121°C. To assess the spectral features and meat quality aspects of the cooked chicken meat samples presented in PCA score plots, a correlation loading plot was generated, as shown in Figure 7B. The correlation loading plot (Figure 7B) for PC-1 reveals that cooked chicken meats at low temperature (70°C) present better moisture retention (moisture content and WHC), lightness (L*), α-helical structure obtained from the amide I region of SR-FTIR and FT-Raman spectra and Raman spectra at 1,651 cm⁻¹, which was assigned to the α-helix conformation according to Herrero (2008b). The quality parameter correlated with the Raman spectra of O–H stretching of water at 3,217 cm⁻¹ (Herrero, 2008b), which can infer to the moisture content in chicken cooked at 70°C, was higher than the chickens cooked at high temperatures at 90°C and 121°C (Table 1). Moreover, the high intensity of the Raman spectrum at 527 cm⁻¹ was assigned to S‒S stretching of disulfide bonds in the g-g-t conformation (Herrero, 2008b), which was predominant in meat cooked at 70°C. The Raman band attributed to C=O stretching of fatty acids at 1,747 cm⁻¹ (Herrero, 2008a) is related to the fat present in the cooked chicken meats. According to Aaslyng et al. (2003), cooked pork samples generally experienced minimal cooking loss when cooked at moderate to low temperatures, which was attributed to a decreased loss of lipids at mild cooking temperature.

Figure 7.

Plots of scores (A) and correlation loading (B) of principal component (PC)-1 vs. PC-2 from average spectra of cooked chicken breasts from 3 chicken breeds under different heat treatments (70°C, 90°C, 121°C). Abbreviations: a*, redness; b*, yellowness; CB, commercial broiler; GMP, guanosine-5ˊ-monophosphate retention (%); IMP, inosine-5ˊ-monophosphate retention (%); IR, SR-FTIR spectra; KC, Korat chicken; L*, lightness; NC, Thai native chicken; R, FT-Raman spectra; WHC, water-holding capacity.

According to PCA analysis, a positive correlation was found between chicken meat cooked at high heat treatments (90°C and 121°C) and cooking loss, yellowness (b*) and random coil structure obtained from FT-Raman spectra (Figure 7B). These quality parameters showed a strong correlation with Raman bands in the region of C–H stretching vibrations (2,968, 2,893 cm⁻¹), but they negatively correlated with aliphatic hydrophobic residues assigned at 1,454 (C–H bending), and 1,340 cm⁻¹ (Trp) (Herrero, 2008b). The positive correlation between the high cooking temperature and the Raman peaks at 1,042, 1,020, and 990 cm⁻¹ could be ascribed to the C–C stretching vibrations of the β-sheet structure (Dàvila et al., 2006 ; Herrero, 2008b). Moreover, high heating temperatures correlated with high intensity Raman bands at 880 and 858 cm⁻¹, assigned to tryptophan and tyrosine, respectively (Herrero, 2008b). These results indicated exposure of tryptophan and tyrosine concomitant with an increase in β-sheet structures upon heating at high temperatures, which was correlated with tougher texture. Berhe et al. (2014) also reported that cooking loss of pork Longissimus thoracis can be predicted by Raman bands at 1,340 and 856 cm⁻¹. The present study provides information that cooking both fast- and slow-growing chickens at high temperatures of 90°C and 121°C induced significant unfolding of the protein structure, exposing more hydrophobic residues, resulting in more β-sheet and random coil structures.

Cooked slow- and fast-growing chicken meats were clearly differentiated by PC-2 (Figure 7A). The cooked KC and CB meats were characterized by SR-FTIR spectra at 1,668 cm⁻¹ assigned to β-turn (Marinkovic and Chance, 2006), 1,564 cm⁻¹ assigned to aggregated β-structure (Böcker et al., 2007), 1,747 cm⁻¹ and 1,462 assigned to C=O stretching and CH₂ bending, respectively (Candoğan et al., 2021). Cooked NC meat contained high GMP and a random coil structure, as well as observed from the SR-FTIR spectra of C–H stretching at 2,929 and 2,862 cm⁻¹ (Candoğan et al., 2021). FT-Raman spectra were capable of differentiating meat quality from different cooking temperatures, while SR-FTIR spectra differentiated chicken meat from different genotypes. SR-FTIR and FT-Raman spectroscopy can potentially be used to rapidly identify and control the quality of cooked chicken meat products in manufacturing processes.

CONCLUSIONS

The breast meat of slow-growing chickens (KC and NC) subjected to high heat treatment at 121°C showed a darker color than CB meat. Taste-enhancing compounds, including IMP and GMP, showed higher retention in both slow-growing chickens. The shear force values of slow-growing chickens were higher than the shear force values of CB meat at all heat treatments. A high heat treatment of 121°C resulted in the highest loss of insoluble collagen content in all 3 chicken breeds. Protein oxidation was predominant in cooked slow-growing chicken meats. Based on PCA of SR-FTIR and FT-Raman spectral data and meat quality traits, low-heat treatment (70°C) cooked chicken meat samples were accompanied by a high moisture content and predominant α-helix structure observed in the FT-Raman spectra at 3,217 cm⁻¹ (O–H stretching of water) and 1,651 cm⁻¹ (amide I; α-helix), respectively. High temperatures of 90°C and 121°C resulted in high cooking loss and cooked meat with a more random coil structure, as revealed by FT-Raman bands at the region of C–H stretching (2,968, 2,983 cm⁻¹), C–C stretching of the β-sheet structure (1,042, 1,020, 990 cm⁻¹), and aromatic side-chain amino acids (880 cm⁻¹; Trp; 858 cm⁻¹; Tyr). Cooked KC and CB meat showed similar characteristics, with high turns, coils, and aggregated β-sheet structures. Cooked NC meat showed high SR-FTIR spectra of C–H stretching (2,929, 2,862 cm⁻¹) associated with the exposure of hydrophobic amino acid residues. These results revealed further applications of vibrational spectroscopy as a nondestructive and rapid tool for differentiating slow- and fast-growing chicken meats cooked at mild and extreme heat treatments.