Re: 그린란드 상어고기, 생선, 육류의 발효탐구!!

[문형철]

누룩소금으로 

생선, 돼지고기, 오리고기 등을 발효해서 먹는 체험을 하고있는 중

https://www.coupang.com/vp/products/6807323360?itemId=16092779382&vendorItemId=87008521132&q=%EB%88%84%EB%A3%A9%EC%86%8C%EA%B8%88&searchId=6fee48283164790&sourceType=search&itemsCount=36&searchRank=17&rank=17&traceId=mm30nihu

천잠 미네랄 쌀 누룩소금1kg 유리판 천일염 김장 양념소금 경기광주맛집 육즙 업소용 핑크소금

왜 그린란드 상어만 위험할까?

일반적인 어류도 TMAO를 가지고 있지만,

그린란드 상어는 극도로 차갑고 수압이 높은 심해에서 수백 년을 살기 위해 

다른 어종보다 수십 배 높은 농도의 TMAO를 축적하고 있습니다.

그래서

별도의 발효 과정 없이 생으로 먹는 것은 매우 위험합니다.

초록 요약 (Abstract)

본 연구는
아이슬란드 전통 발효 식품인 하카를(hákarl)의 미생물 생태계를
문화-비의존적 방법으로 조사하였다.

16S rRNA 유전자 고처리량 시퀀싱(HTS)을 통해
발효 과정에서 미생물 군집이 초기 해양 관련 분류군(Photobacterium, Pseudoalteromonas)에서
혐기성 포자형성균(Tissierella)으로,
후기에는 저온성 세균(Pseudomonas)으로 천이하는 것을 확인하였다.

이러한 변화는
트리메틸아민 N-옥사이드(TMAO)와 요소(urea)의 해독과 직결되어
제품을 식용 가능하게 만든다.

자생 미생물군집이 주도하는 자연 발효의 특성을 밝힘으로써
식품 안전성과 제어 발효 가능성을 제시하였다.

알파-다양성 감소와 알칼리성·혐기성 조건에서의
군집 천이가 핵심 지표이다.

서론 및 배경

그린란드 상어(Somniosus microcephalus) 고기는
심해 환경 적응을 위해 TMAO와 요소를 고농도로 함유하고 있어
생식 시 트리메틸아민뇨증 유사 증상이나 요독증을 유발한다.

전통 하카를 제조법은
상어 살코기를 자갈에 묻어 6~12주 발효한 후 2~4개월 건조하는 방식으로,
미생물이 TMAO와 요소를 암모니아 등 휘발성 물질로 분해하여
pH를 9 수준으로 높이고 독성을 제거한다.

서론에서는 하카를을 수르스트뢰밍(surströmming), 라크피스크(rakfisk) 등 다른 발효 어류와 비교하며, 기존 연구가 주로 배양-의존적 방법에 치중해 Clostridium botulinum 위험성만 지적한 점을 지적한다. 본 연구는 메타지노믹스를 통해 시간에 따른 미생물 천이를 최초로 규명하고, 해독 메커니즘과 향미 형성에 기여하는 핵심 균군을 가설화하였다. EU 식품 유산 보호(PDO/TSG)와 자생 발효 식품의 안전성 관리 측면에서 의의가 크다.

재료 및 방법

• 시료 채취: 아이슬란드 Bjarnarhöfn의 전통 작업장에서 3회 독립 생산된 하카를을 채취. 상어 필레(약 20 kg)를 천공 플라스틱 박스에 담아 자갈 속에서 35일 발효(4~8°C). Day 0(원료), 7, 14, 21, 28, 35일에 무균 채취(n=3, 균질화).
• 이화학 분석: pH 측정(초기 6.5 → 최종 9.0), TMAO/요소 HPLC 정량.
• DNA 추출: PowerFood Microbial DNA Isolation Kit + 비드-비팅(포자 파괴). NanoDrop 및 아가로스 젤로 품질 확인.
• PCR-DGGE: 16S rRNA V3 영역 증폭(GC-357f/907r), 30-60% 우레아-포름아미드 구배 DGGE. 밴드 절취 후 BLAST 동정(97% 유사도).
• 고처리량 시퀀싱(HTS): V1-V3 영역 증폭(27F/534R), Illumina MiSeq(2×300 bp, 약 50,000 reads/sample). QIIME v1.9.1 파이프라인으로 OTU 클러스터링(97%), Greengenes 분류. 알파-다양성(Shannon, Chao1), 베타-다양성(UniFrac), PERMANOVA·ANOVA 통계.
• 대조군: 추출·PCR 음성 대조, 반복 실험으로 재현성 확보.

문화-비의존적 접근으로 비배양 미생물을 포괄적으로 포착하였다.
결과

• 이화학 변화: pH 6.3±0.2 → 8.9±0.1, TMAO 100 mmol/kg → 검출한계 이하(암모니아 축적과 상관).
• DGGE: 초기 Photobacterium phosphoreum·Pseudoalteromonas, 중기 Tissierella creatinophila 우세, 후기 Pseudomonas·Atopostipes 지속.
• HTS 미생물 다양성:
  • 알파-다양성: Shannon 지수 4.5 → 2.0 (알칼리·혐기 조건으로 선택 압력).
  • 베타-다양성: PCoA에서 시간에 따른 명확한 클러스터링.
  • 분류군 조성: Day 0 – Gammaproteobacteria 60-70% (Photobacterium 25%, Pseudoalteromonas 20%); Day 7-14 – Firmicutes (Tissierella 최대 50%); Day 21-35 – Tissierella 40-60%, Pseudomonas 10-20%.
• 통계: PERMANOVA p<0.01, LEfSe로 Tissierella를 후기 발효 바이오마커로 확인.

결과는 초기 해양균 → 발효 혐기균 → 건조 저온균으로의 명확한 천이를 보여준다.

토론
미생물 천이는 환경 변화와 일치한다. 초기 TMAO 환원균(Photobacterium, Pseudoalteromonas)이 tdm 유전자를 통해 TMAO를 TMA로 전환하여 pH 상승과 암모니아 생성을 주도하며, Tissierella는 크레아티닌 분해와 단백질 가수분해로 감칠맛·암모니아 향을 형성한다. Pseudomonas는 건조 안정성에 기여하나 과도하면 부패 위험이 있다.
타 발효 어류(태국 pla-ra: Lactobacillus 우세, 스웨덴 수르스트뢰밍: Halanaerobium)와 비교 시, 하카를은 저염·고pH 조건으로 인한 클로스트리디아 프로필이 독특하다. 안전성 측면에서 병원균 검출량은 낮으나 C. botulinum 위험 관리가 필요하다. 본 연구는 Tissierella 스타터 개발 등 품질 균일화의 과학적 근거를 제공한다.
한계: HTS 프라이머 편향, 기능적 메타지노믹스(메타트랜스크립토믹스) 부재. 치즈 숙성(블루치즈 Pseudomonas)과의 유사성을 통해 극한 발효 연구의 지평을 넓혔다.

https://www.youtube.com/watch?v=9_ZWOjVd3J4&t=458s

https://www.mdpi.com/2304-8158/12/9/1768

Application and Effect of Pediococcus pentosaceus and Lactiplantibacillus plantarum as Starter Cultures on Bacterial Communitie

1. 초록 및 배경

• Zhayu(자유)는 중국 후난성의 전통 발효 생선 제품으로, 소금에 절인 풀잉어(grass carp) 필레를 밥가루, 고추, 향신료와 섞어 고상 발효시킨다. 자연 발효는 미생물 불균일성, 계절 변동 등으로 품질이 불안정해 산업화가 어렵다.
• 목적: 소금 내성 및 항균성이 강한 Pediococcus pentosaceus P1과 Lactiplantibacillus plantarum L1 (치즈 생선에서 분리)을 스타터로 사용해 발효를 촉진하고, 미생물 군집, 휘발성 향기 화합물, 안전성, 질감을 개선하는 효과를 평가.

2. 실험 설계 (Materials and Methods)

• 시료: 풀잉어 필레를 소금 절임 후 밥가루(20%), 적효모 쌀가루(4%)와 스타터(10^6 CFU/g) 혼합. 그룹: CK(자연 발효), P1, L1, PL(혼합).
• 발효: 32°C, 10일 (0, 2, 4, 6, 8, 10일 샘플링).
• 분석: pH, TCA 가용성 펩타이드(단백질 분해), 미생물 수(총균, 장내세균, 황색포도상구균), 단백질 분해(SDS-PAGE), 지방 산화(TBARS), 질감(TPA), 수분 상태(LF-NMR), 미생물 군집(16S rRNA 시퀀싱, V3-V4), 휘발성 화합물(HS-SPME-GC-MS).

3. 주요 결과 (Results)

• 발효 속도 및 안전성:
  • pH 급강하 (L1 그룹: 2일 만에 4.28 vs. CK 5.22).
  • LAB 수: 4일째 10^7 CFU/g 이상 (CK보다 빠름).
  • 장내세균 및 황색포도상구균: 40% 감소 (<1 log CFU/g).
  • 지방 산화(TBARS): 45% 저하 (1.26 mg/kg vs. CK 2.03).
• 단백질 분해 및 질감:
  • 미오신(200 kDa), 액틴(43 kDa) 가수분해 가속 → TCA 펩타이드 1.3배 증가 (L1: 1037.5 µmol/g vs. CK 861.5).
  • 질감: 경도(hardness)와 저작성(chewiness) 향상 (1.25배); 결합수 증가로 부드러움 개선.
• 미생물 군집 변화 (16S 시퀀싱):
  • CK: 다양성 높음 (Shannon 지수 4.347), Pantoea(22.59%), Klebsiella(18.77%) 우세 (부패균).
  • 스타터 그룹: LAB 지배 (Pediococcus 80% in P1, Lactiplantibacillus 72% in L1) → 다양성 감소, 부패균 75% 감소 (L1에서 Klebsiella 특히).
• 휘발성 향기 화합물:
  • 80종 식별. L1 그룹: 2,3-부탄디올(자극적 알코올 냄새) 감소 (Klebsiella와 양의 상관).
  • 2-펜틸퓨란, 2-부틸-3-메틸피라진 3배 이상 증가 → 독특한 향과 풍미 강화.

4. 논의 및 함의 (Discussion)

• 스타터 LAB는 산 생성, 박테리오신으로 병원균 억제하고, 단백질 분해로 질감 개선. 미생물 변화가 향기와 직결 (Lactiplantibacillus가 옥탄산, 퓨란과 양의 상관; 알코올/산과 음의 상관).
• 유사 발효 어류(수안유 등)와 비교: LAB가 부패균 억제 공통. 하카를(hákarl)과의 연계: 원 논문(하카를 미생물 연구)을 인용하며, LAB가 발효 어류의 안전성과 풍미를 향상시키는 보편적 메커니즘으로 지지.
• 산업적 가치: L1 균주가 가장 효과적 (안전+풍미). 혼합 스타터로 더 최적화 가능.

5. 결론 (Conclusions)

• LAB 스타터(특히 L. plantarum L1)는 Zhayu의 안전성(병원균↓, 산화↓), 질감(단단+쫄깃), 풍미(자극적 냄새↓, 매력적 향↑)를 크게 개선. 미생물-향기 상관 분석으로 발효 최적화의 과학적 근거 제공. 전통 발효 식품의 현대화에 기여.

•  
•  

24 April 2023

Application and Effect of Pediococcus pentosaceus and Lactiplantibacillus plantarum as Starter Cultures on Bacterial Communities and Volatile Flavor Compounds of Zhayu, a Chinese Traditional Fermented Fish Product

Dongmei Xu1,2,

Yongle Liu1,2,

Xianghong Li1,2,

Faxiang Wang1,2,

Yiqun Huang1,2 and

Xiayin Ma1,2,*

Abstract

Zhayu is a type of traditional fermented fish product in China that is made through the fermentation of salted fish with a mixture of cereals and spices. Inoculation fermentation was performed using Pediococcus pentosaceus P1, Lactiplantibacillus plantarum L1, and a mixture of two strains, which were isolated from cured fish in Hunan Province. Compared with the natural fermentation, inoculation with lactic acid bacteria (LAB) accelerated the degradation of myosin and actin in Zhayu, increased the trichloroacetic acid (TCA)-soluble peptide content by about 1.3-fold, reduced the colony counts of Enterobacteriaceae and Staphylococcus aureus by about 40%, and inhibited their lipid oxidation. In the texture profile analysis performed, higher levels of hardness and chewiness were observed in the inoculation groups. In this study, the bacterial community and volatile flavor compounds were detected through 16S high-throughput sequencing and headspace solid-phase microextraction–gas chromatography–mass spectrometry (HS-SPME-GC-MS). Inoculation with L. plantarum L1 reduced around 75% abundance of Klebsiella compared with the natural fermentation group, which was positively correlated with 2,3-Butanediol, resulting in a less pungent alcohol odor in Zhayu products. The abundances of 2-pentylfuran and 2-butyl-3-methylpyrazine were increased over threefold in the L1 group, which may give Zhayu its unique flavor and aroma.

Keywords:

 Zhayu; LAB fermentation; microbial communities; texture; volatile flavor compounds

1. Introduction

Fermentation has been widely used in fish processing in order to prolong the fish products’ shelf lives and develop unique flavors [1]. In China, there are a variety of traditional fermented fish products produced in ethnic minority areas, such as Chouguiyu [1,2], Suanyu [3], and Yucha [4]. Zhayu is a type of solid fermented fish product from Hunan Province, China. In the traditional natural fermentation of Zhayu, washed fish fillets were cubed and combined with salt, rice flour, cayenne, and other seasonings, and after that, the mixture was pressured and cultured in a solid form [5]. However, along with globalization and market opening, natural fermented Zhayu products face many difficulties in achieving large-scale industrial production. The quality of Zhayu products is influenced by a variety of factors, such as environmental microorganisms, seasonal variation, and manual operation experience [6,7], leading to long fermentation cycles and the high variability of product quality.

According to relevant reports, inoculation fermentation is a frequently employed technique for raising the quality of fermented fish products, and starter cultures used for fermentation are commonly selected from Lactobacillus, Pediococcus, Staphylococcus, Saccharomyces, and related genera [8,9,10]. The physiological functions and characteristics of LAB were crucial to the fermentation process. Spoilage microorganisms could be inhibited by antibacterial metabolites produced by LAB, including organic acids and bacteriocins [11,12]. Studies have proved that LAB could encourage the development of fish products with improved flavors [13].

The flavor of fermented foods was correlated with the microbial composition during fermentation [14]. High-throughput sequencing based on the 16S rRNA gene amplicon has been utilized to research microbial communities in fermented fish products [15,16,17]. The relationship between microbial succession and flavor development has been investigated in Suanyu inoculated with diverse starter cultures [18]. A further study investigated how diverse cultures affected the characteristic flavor of Suanyu [19]. In Chinese fish sauce, Wang et al. [20] investigated the dynamics of volatile flavor chemicals and their association with microbial populations. As for research on Zhayu, Yang et al. [8] found that the addition of flavorzyme could enhance the synthesis of alcohols, aldehydes, and esters in Suanzhayu. An et al. [5] found that Zhayu inoculated with L. plantarum and P. acidilactici showed rising acidity and a pleasant odor, and the amount of terpenoids, esters, acids, and S-containing compounds were increased in Zhayu. However, the relationship between flavor quality and microbial composition remains to be exploited, and strains with good fermenting properties need to be identified.

Previously, two strains of LAB (P. pentosaceus P1 and L. plantarum L1) were chosen from the traditional natural fermentation of cured fish and identified as having good fermentation properties, such as salt tolerance and antibacterial ability. In the current investigation, fermented Zhayu samples were generated by inoculation with L. plantarum, P. pentosaceus, or a mixture of the two strains. The safety and textural properties of Zhayu were eval‎uated, and the bacterial communities and volatile flavor compounds of Zhayu were investigated. This study aimed to ascertain the impact of inoculation with LAB species on the quality of Zhayu and provide information on the relevance between microbial variation and flavor enhancement.

2. Materials and Methods2.1. Materials and Chemicals

The grass carp (Ctenopharyngodon idellus) and rice flour were bought from a partial market in Changsha City, Hunan Province. Red yeast rice flour (Jiangsu Weipinhui Food Co., Ltd. Taizhou, China) was purchased from the supermarket of Changsha University of Technology. The molecular weight marker used was the protein ladder (range 10–200 kDa, Thermo Fisher Scientific Inc. MA, USA). The de Man, Rogosa, and Sharpe (MRS) medium, Violet Red Bile Glucose Agar (VRBA), and Mannitol Salt Agar (MSA) were bought from Huankai Microbial Co., Ltd. (Guangzhou, China). All additional chemicals used were of analytical quality.

2.2. Pre-Culture of LAB

P. pentosaceus P1 and L. plantarum L1 were isolated from cured fish and preserved in the laboratory of the School of Food and Biological Engineering, Changsha University of Science and Technology. The LAB strains were revitalized twice in MRS broth at 37 °C for 12 h. Then, culture solutions were centrifuged at 4 °C for 10 min with a centrifugal force of 11,100× g. Then, the bacterial cell pellets were re-suspended in sterile saline water and adjusted to a cell density of 106 CFU/mL.

2.3. Preparation of Zhayu Samples

Fresh grass carps (weight: 2.5 ± 0.5 kg) were slaughtered and stored in crushed ice immediately. White dorsal muscle was removed and processed into blocks of 3 × 3 × 2 cm3. The grass carp pieces were immersed in brine solution (1:25 g/g, salt to fish; 1:10 w/v, salt to water) at 4 °C for 24 h and then patted dry with paper towels. Following this, the grass carp pieces were patted with paper and dried for 2 h at 45 °C in an electric blast drying oven (DHG-9140A, Shanghai Jinghong Instruments Co., Ltd., Shanghai, China). Based on our previous study, the carp pieces were mixed with 20% (g/g) rice flour, 4% (g/g) red yeast rice flour, and 3% (w/v) LAB starter cultures (Figure 1). Samples were fermented in 50 mL jars with lids sealed tightly at 32 °C in a constant temperature incubator (LHS-250SC, Shanghai Yiheng Technology Co., Ltd., Shanghai, China) and named CK (without starter cultures), P1 (inoculated with P. pentosaceus P1), L1 (inoculated with L. plantarum L1), and PL (inoculated with the 1:1 mixture of P. pentosaceus P1 and L. plantarum L1). As shown in Figure 1. The samples were taken from separate containers at 0, 2, 4, 6, 8, and 10 d during the fermentation process for inspection. At each sampling location, three parallel samples were collected for analysis.

Figure 1. Preparation of Zhayu samples.

2.4. Determination of pH and Trichloroacetic Acid (TCA)-Soluble Peptides

The pH values of Zhayu were measured according to the national standard GB 5009.237-2016 [21]. Samples (1 g) were homogenized with 10 mL of 0.75% (w/v) KCl solution and then measured with an electronic pH meter (Ohaus International Trading Co., Ltd., Shanghai, China).

According to the method of Hatairad et al. [22], the TCA-soluble peptide contents of Zhayu products during fermentation were examined. Samples (2.0 g) were homogenized with 18 mL of 5% (w/v) TCA, and the mixture was extracted at 4 °C for 1 h while being vibrated every 15 min. The solution was centrifuged at 4000× g and 4 °C for 15 min, and then the supernatant was collected and mixed with Folin reagent to produce a dark blue complex. The TCA-soluble peptides were observed at 500 nm using the UV spectrophotometer (TU-1901, Beijing Pu-Analysis General Instrument Co., Ltd., Beijing, China). The results were reported as µmol tyrosine/g sample. Each determination was made three times.

2.5. Determination of Microbial Counts

Five grams of each sample was aseptically removed from the jars and homogenized for 1 min in 45 mL of 0.9% (w/v) saline solution. A series of gradient dilutions were prepared, and 100 μL of bacterial suspensions was dispersed on plates for microbial counting. LAB was cultured on MRS medium at 37 °C for 24 h, Enterobacteriaceae were cultured on VRBA at 37 °C for 24 h, and Staphylococcus aureus strains were cultured on MSA at 37 °C for 24 h. The number of colonies was expressed as log colony forming units (CFU) per gram of Zhayu sample.

2.6. SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

The SDS-PAGE procedure is modified from Li et al. [23] in certain ways. Briefly, the extracted protein solutions were adjusted to the same concentration with 1× loading buffer solution and then heated at 100 °C for 5 min prior to electrophoresis. The molecular weight marker used was protein ladder (range 10–200 kDa). The concentrated gel and separated gel were 5 g/mL and 120 g/mL. The electrophoresis equipment (DYCZ-25D, Beijing Liuyi Biotechnology Co., Ltd., Beijing, China) was run at a continuous current of 20 mA for the concentrated gel and 40 mA for the separated gel, which had concentrations of 5 g/mL and 120 g/mL. They were stained with Coomassie brilliant blue R250, followed by destaining to visualize the gels. Results from scanning were obtained utilizing a Bio-5000 Plus GEL imaging scanner (Shanghai Microtek Technology Co., Ltd. Shanghai, China).

2.7. Determination of Thiobarbituric Acid-Reactive Substances (TBARS)

Two-gram portions of Zhayu samples were combined with 10 mL of trichloroacetic acid solution (75 g/L trichloroacetic acid, 1 g/L propylgallate, and 1 g/L EDTA). The homogenate was filtered through double-layer filter paper, and 5 mL of the filtrate was combined with 5 mL of the 0.02 mol/L 2-thiobarbituric acid solution and heated in boiling water (90 °C) for 30 min. After cooling with ice water, the mixture was analyzed using an ultraviolet spectrophotometer (Beijing General Instrument Co., Ltd. Beijing, China) at 532 nm [24]. The TBARS values were expressed as mg of malondialdehyde (MDA) per kg of sample.

2.8. Texture Profile Analysis (TPA)

Texture profile analysis (TPA) of Zhayu samples (1.5 cm radius and 2.0 cm in thickness) was performed using the Texture Analyzer TA.XT.plus (Stable Micro Systems Ltd., Godalming, UK). According to Wang et al. [13], the following experimental parameters were chosen: a trigger force of 0.049 N, a compression ratio of 30%, a pre-test speed of 1.0 mm/s, a test speed of 1.0 mm/s, and a post-test speed of 2.0 mm/s. Using a cylinder probe P/50, six parallel experiments were conducted before each test.

2.9. Low-Field Nuclear Magnetic Resonance (LF-NMR)

Samples were cut into 10 mm × 10 mm × 10 mm (1 g) cubes along the fiber direction and placed in nuclear magnetic resonance tubes (25 mm in diameter). The transverse relaxation data were measured using a MesoMR23-060V-I LFNMR analyzer (Niumag, Ltd., Shanghai, China) according to the method of Qin et al. [25]. The analyzer was operated at a resonance frequency of 18 MHz at 32 °C. The Carr–Purcell–Meiboom–Gill (CPMG) pulse sequence was used to measure the T2. The time delay used for the T2 measurement was between 90 °C and 180 °C, with pulses of 14 ms and 24 ms. There were 200 measurement points for the CPMG measurements. Data from 4000 echoes were collected over the course of 8 scan repeats, with a 9000 ms repeat interval between each scan. The multi-exponential decay curve was obtained from NMR relaxation processing by conducting multi-exponential fitting analysis with Nuimag’s Multi Exp Inv Analysis software.

2.10. Analysis of Bacterial Diversity

Samples were obtained in a sterile environment and kept at −80 °C. The entire genome of DNA in the samples was extracted using the CTAB/SDS method. In 1% (w/v) agarose gels, DNA concentration and purity were examined. DNA was diluted to a concentration of 1 ng/L using sterile water.

The 16S rRNA genes were amplified in different locations (16S V3-V4) using a particular primer (16S V4: 515F-806R) and barcodes. Amplifications were performed employing an initial denaturation cycle at 98 °C for 1 min, followed by 30 cycles of denaturation (98 °C for 10 s), annealing (50 °C for 30 s), elongation (72 °C for 30 s), and a final 5 min extension at 72 °C. All PCR mixtures contained 15 L of Phusion® High-Fidelity PCR Master Mix (New England Biolabs, London, UK), 0.2 µM of each primer, and 10 ng of target DNA. PCR products were resolved by a Qiagen Gel Extraction Kit (Dusseldorf, Germany). Using the NEBNext® UltraTM IIDNA Library Prep Kit, sequencing libraries were created (Cat No. E7645) and assessed using the Agilent Bioanalyzer 2100 system and the Qubit@ 2.0 Fluorometer from Thermo Scientific (Waltham, MA, USA). Ultimately, 250 bp paired-end reads were produced after the library was sequenced on an Illumina NovaSeq device. Sequences were combined using FLASH and quality filtered before being grouped into operational taxonomic units (OTUs) by UPARSE at a 97% similarity threshold, and chimaeras were detected using UCHIME.

2.11. Determination of Volatile Compounds by HS-SPME-GC-MS Analysis

Each sample (5 g) was chopped and put into a headspace vial with a 20 mL capacity and then placed in a robotic arm TriPlus RSH (CTC Analytics, Basel, Switzerland) autosampler incubation chamber and heated at 40 °C for 10 min for volatile compound enrichment. The extraction was performed using a headspace solid-phase microextraction (HS-SPME) extraction needle (50/30 μm DVB/CAR/PDMS Supelco, MA, USA) for 20 min. Using the GC-Orbitrap-MS system (Q Exactive GC, Thermo Fisher, MA, USA), the volatile chemicals were examined on a TG-5 column (0.25 × 0.25 × 0.32, 30 m, Fisher Scientific, MA, USA).

The GC-MS started out at 40 °C for 2 min and was then raised to 250 °C at a rate of 8 °C/min and held for 2 min, for a total running cycle of 32 min. The MS was run in positive electron impact ionization EI+ mode with a solvent delay for the first 0.5 min, scanning from ion mass fragments 35–475 m/z, an interscan delay of 0.1 s, and a resolution of 60,000 at FWHM (Full Width at Half Maximum). The carrier gas was helium of 99.996% purity (Shandong, China). The velocity of the helium gas was set at 1.2 mL/min.

For the retention times and retention index of volatile flavor compounds, we relied on computer matching with the reference of the MS library of NIST 14.0 (mass-spectral similarity match ≥ 80). By computing the percentages of GC peak regions, the compounds were given a numerical value.

2.12. Data Analysis

For each experiment, at least three technical replicates were examined using the statistical program SPSS 26.0. The results were presented as mean standard deviation (SD). Using Welch’s ANOVA (α = 0.05), the Dunnet-T3 post-hoc test was used to examine the significance of the data; a probability value of p < 0.05 was deemed significant. SIMCA-P 14.0 software (Umetrics, Umeå, Sweden) was used to perform orthonormal partial least squares-discriminant analysis (OPLS-DA) on GC-MS metabolomic data. Characteristic metabolites were defined as variables in the projection, with variable importance (VIP) > 1. GraphPad Prism 8.0.2 was used to examine the significant differences and Pearson correlation coefficient and plot all heatmap construction.