火腿中风味物质检测方案(感官智能分析)

智能文字提取功能测试中

Food Control 126 (2021) 107873Contents lists available at ScienceDirectFood Controljournal homepage： www.elsevier.com/locate/foodcont Food Control 126 (2021)107873C.-Y. Zhou et al. CONTROLCONTROLFOODCONTROLCONTROLCONTROLCONTROLCONTROLCONTROL 1H NMR-based metabolomics and sensory evaluation characterize tastesubstances of Jinhua ham with traditional and modernprocessing procedures Chang-Yu Zhoub， Yun Bai， Chong Wang， Chun-Bao Li ， Xing-Lian Xu， Dao-Dong Pan a.**Jin-Xuan Cao，"， Guang-Hong Zhou” a State Key Laboratory for Managing Biotic and Chemical Threats to the Quality and Safety of Agro-products， Key Laboratory ofAnimal Protein Food ProcessingTechnology of Zhejiang Province， College of Food & Pharmaceutical Sciences， Ningbo University， Ningbo， 315211， PR China Key Laboratory of Meat Processing and Quality Control， MOE， Key Laboratory of Meat Processing， MOA， Jiangsu Synergetic Innovation Center of Meat Processing andQuality Control， Nanjing Agricultural University， Nanjing， 210095， PR China A RTICLEINFO ABSTRAC T Keywords： Jinhua ham Modern processing Metabolites Partial least square discriminant analysis Taste-active components Taste and richness intensities Jinhua ham is one of high quality dry-cured products， and processing methods play a key role in developing thetaste attributes of Jinhua ham. To investigate the effect of processing methods on sensory attributes and tastesubstances of Jinhua ham， this work highlighted the changes of sensory characteristics， texture parameters， waterdistribution and metabolite profiles， and further discussed the contribution of metabolites to the development oftaste and richness in both traditional-processed and modern-processed hams. The modern-processed hamsshowed higher taste scores and richness intensities than that of traditional processing. The populations ofimmobilized water significantly increased from raw ham to the later stage of dry-ripening， accompanied by adecrease in the populations of free water during the processing of traditional and modern procedures， whilehigher populations (more than 2-fold) of free water were shown in the ham of modern processing at the end ofpost-ripening compared with traditional-processed ham. 1H NMR-based metabolomics revealed that free aminoacids， small peptides and organic acids were the most intense response in developing taste and richness in-tensities of modern-processed ham. Partial least square discriminant analysis and taste-active values furtherdemonstrated that glutamic acid， lactate， glycerol， anserine and creatine were responsible for the higher tasteand richness intensities of modern-processed ham. 1. Introduction Jinhua ham is highly favored by consumers mainly due to its uniqueorganoleptic and flavor characteristics (Zhang， Jin， Wang， & Zhang，2011； Zhou & Zhao， 2007). During the past few decades， the manufac-ture of Jinhua ham has generally followed traditional procedure，beginning in winter and ending in the following autumn under naturalconditions (Zhang， Zhen， Zhang， Zeng， & Zhou， 2010). However， thereare many challenges in the traditional production procedure. Forexample， the flavor quality of the ham cannot be guaranteed when theabnormal weather such as high temperature (more than 20 °C) occurs. Furthermore， the high residues of sodium content (approximate 10%)appear in final product， which is rejected by consumers (Zhou， Wu et al.，2019b； Zhou， Pan et al.， 2019). Consequently， some strategies to reducesodium content during the modern processing of dry-cured ham shouldbe developed. Modern process shows a shorter processing time (6-8months) and more refined processing procedure， which only includesfive processing steps (raw ham preparation， salting， washing， ripeningand post-ripening)， compared with the traditional process (Zhou， Panet al.， 2019). The suitability of raw material and the changes of tem-perature and relative humidity (RH) have been explored to improvetheir qualities during the processing of dry-cured ham (Arnau， Serra， ( * Corresponding author. ) ( ** Corresponding a uthor. ) ( ** Corresponding a uthor. ) E-mail addresses： zhouchangyu@nbu.edu.cn (C.-Y.Zhou)， daodongpan@163.com (D.-D. Pan)， caojinxuan@nbu.edu.cn (J.-X. Cao)， guanghong.zhou@hotmail.com (G.-H.Zhou). ( h ttps： // doi . org/10.1016/ j .foodcont.2021. 1 0 7873 ) ( Received 5 August 2020； Received in revise d for m 2 5 November 2020； Accepted 10 January 2021 Available online 13 January 2021 ) Comaposada， Gou， & Garriga， 2007； Bosse Nee Danz et al.， 2018).Furthermore， low dosage of sodium chloride (about 6%) used at thesalting stage， and the powerful refrigeration units and rooms for con-trolling temperature and RH have been developed to guarantee the safecharacteristics and flavour attributes in many enterprises during themodern processing of Jinhua ham (Zhou， Pan et al.， 2019). However， upto now， the actual substance responsible for the sensory qualities andtaste attributes have not yet been well identified and explored during themodern processing of Jinhua ham. Taste is an important indicator of dry-cured ham， and largely de-termines consumer preferences for the consumption of the products(Zhang et al.， 2018). Amino acids， peptides， organic acids， inorganicsalts and nucleotides are the main taste substances in dry-cured meatproducts (Dashdorj， Amna， & Hwang， 2015； Lopez-Pedrouso et al.，2019； Sforza et al.， 2006). Studies have demonstrated that these tastecompounds mainly derive from protein degradation and lipid oxidation(Toldra， 2006). Most researchers agree that lipid oxidation and proteindegradation which are the result of the activities of endogenous en-zymes， are the major contributors to texture parameters and taste at-tributes of dry-cured ham， and that a large accumulation of hydrolysisproducts does not necessarily serve to enhance sensory and taste char-acteristics in dry-cured ham (Sforza et al.， 2006； Toldra， 2006). Thus，investigating the specific metabolites could be more meaningful to un-derstand the development of taste and sensory attributes in Jinhua hammanufactured by different processing procedures. Metabonomic technologies show great potential to identify metab-olites of dry-cured ham； these identification and quantitation technol-ogies mainly include high performance liquid chromatography， liquidchromatography-mass spectrometer (LC-MS/MS) and high resolutionnuclear magnetic resonance (NMR). NMR spectroscopy has uniqueadvantage of enabling rapid identification and quantification of me-tabolites (Yang，Dai， Ayed， & Liu， 2019)， and it has been well applied inscientific field of meat and meat products in recent years (Castejon，Garcia-Segura， Escudero， Herrera， & Cambero， 2015； Xiao， Ge， Zhou，Zhang， & Liao， 2019； Yang et al.，2019). Zhang et al. (2018， 2019) re-ported that substances including free amino acids， organic acids， nucleicacids and small peptides were systematically identified by NMR indry-cured ham. However， few literatures focus on comparing the com-ponents and profiles of taste substances using 1H NMR and furthercharacterize their contributions to taste and richness intensities of Jin-hua ham manufactured by modern and traditional procedures. Therefore， the aim of the present study was to characterize thecomponents and profiles of taste substances of Jinhua ham using 1HNMR， and to further discuss the contribution of key compounds to tasteand richness development of Jinhua ham during the processing oftraditional and modern procedures. 2. Materials and methods 2.1. The processing and sampling of Jinhua ham The hind legs (average weight， 13.0±0.2 kg； pH， 5.8 ± 0.2) ofDomestic pigs (Large White ×Landrace) were used to the production ofJinhua ham at the Zhejing Provincial Food Company， P.R. China. Theprocessing procedures (traditional and modern procedures) of Jinhuaham were performed according to our previous studies (Zhou， Wu et al.，2019b). Briefly， for the modern processing procedure， hind legs weresalted using 0.15 g KNO3， 0.15 g NaNO2 and 60 g NaCl per kilogram ofraw ham， and then the hams underwent post-salting， washing andripening for 5-6 months in the dry-ripening room， where the RH pro-gressively decreased from 80% to 65% and the ambient temperatureincreased from 6°C to 35 °C； for the traditional processing procedure，hind legs were salted using 0.15 g KNO3， 0.15 g NaNO2 and 80 g NaClper kilogram of raw ham， and then the hams underwent washing，sun-drying and ripening for 6-8 months in the dry-ripening room， wherethe temperature and relative humidity depended on the unique local weather and climate. After ripening， these hams with modern andtraditional processing further underwent post-ripening for 1-2 monthsat room temperature (25°C). The process was terminated when the totalweight loss of ham was approximately 40% of the initial weight in bothtraditional and modern processing. Bicep femoris muscles of twenty hamswere sampled at the raw ham， the end of drying-ripening and the end ofpost-ripening in both traditional and modern processing， respectively.Samples were vacuum-packaged and frozen at -80 °C until analyzed. 2.2. Sensory analysis of ham samples The sensory analysis was performed according to the description ofZhou， Wang et al. (2019a) with some modifications. The sensory attri-butes of ham samples including overall taste， saltiness， sweetness，richness， umami， bitterness， sourness and aftertaste attributes werescored by 10 sensory assessors， balanced in terms of gender and owningrich experience. Rating of these attributes’intensities was performedusing a linear unstructured 1 mm scale anchored at the scales (0：absence of sensation； 5： maximum of sensation intensity). The resultswere expressed as the mean of twenty hams of each group. 2.3. Electronic tongue analysis of Jinhua ham Umami and richness intensities of the water-soluble extraction ofJinhua ham were performed according to the description (Dang， Gao，Ma， & Wu， 2015). Ham samples (10g) were homogenized at 12000 r for2 min in 200 mL distilled water. The homogenates were centrifuged at3000 g for 10 min at 4°C； the supernatant was collected and filteredwith 3 layers of filter paper. Umami and richness intensities of thewater-soluble extraction of Jinhua ham were analyzed using TS-SA402Belectronic tongue (INSENT Inc.， Japan). The sensory intensities of thesesamples were calculated according to the absolute value of the sensorpotential based on the reference solution. The sensor potential ofreference solution was defined as 0， and the values more than the in-tensity (0) of reference solution were considered meaningful. Theaverage values of three times measurements for each sample wereanalyzed using TS-SA402B Library search software (INSENT， Japan).The results of umami and richness intensities were expressed as themean values of twenty replicates. 2.4.Texture analysis of Jinhua ham Textural analysis was performed as described by Lopez-Pedrousoet al. (2018) and Zhou， Wang， Cai et al. (2019). The ham samples (20 ×10 ×10 mm) were measured by texture analyzer (TA-XT Plus； StableMicro Systems， Godalming， UK) equipped the special probe (P 50). Themeasurement of adhesiveness， hardness， springiness and cohesiveness ofthe ham samples was performed according to the description of Zhou，Wang， Cai et al. (2019a). All measurements were performed at 25°C.The results of texture parameters were expressed as the mean values oftwenty replicates. 2.5. Moisture content analysis of Jinhua ham The moisture content was determined according to Gou， Comapo-sada， and Arnau (2004). Briefly，5 g of the minced Bicep femoris muscleswere dried at 105±2°C to a constant weight， and the moisture contentwas defined as gram per 100 g muscles. The results of moisture contentwere expressed as the mean values of twenty replicates. 2.6. Water distribution analysis of Jinhua ham Water distribution analysis of ham samples was performed by Lowfield nuclear magnetic resonance (LF-NMR)， as previously described byGarcia et al. (2015) and Zhou， Wang， Cai et al. (2019a). Sample prep-aration and measurement were performed according to the description of Zhou， Wang， Cai et al. (2019a) at 30°C. The procedure ofCarr-Purcell-Meiboom-Gill sequence (CPMG) was used to measuretransverse relaxation times (T2) and transverse relaxation data wereanalyzed by the algorithm of biexponential fittings. Transverse relaxa-tion times (T2) and distribution populations of bound water (P2b)，immobilized water (P21) and free water (P22) were calculated accord-ing to the measurements of twenty replicates. 2.7. The preparation of metabolites The preparation of ham metabolites (n=7) for each group (raw ham，traditional-processed ham and modern-processed ham) were performedaccording to the description (Zhang et al.， 2018). Briefly， biceps femorismuscle samples (400 mg) of each ham was homogenized at 12000 r for3 min in 600 uL of methanol/water (2：1， v/v). The extracts werecentrifuged at 12000 g for 10 min at 4C. The methanol of the super-natants was removed by vacuum freeze-drying. The extraction wasdissolved in 600 uL of 0.1 M K2HPO4/NaH2PO4 buffer (pH 7.4) con-taining 50% D20， 0.01% NaN3 and 0.001% sodium trimethylsilylpro-pionate and then were centrifuged at 12000 g at 4°C for 10 min. The550 uL supernatants of each extraction were transferred into a 5 mmouter diameter NMR tube and 2， 2-dimethyl-2-silapentane-5-sulfonatewas also added to quantify these metabolites. 2.8. Data analysis of 1H NMR spectra 1H NMR spectra of extraction were collected at 298 K on a BrukerAvance 600 MHz Spectrometer equipped with ultra-low temperaturedetection probe under the operating condition of 600.13 MHz using thestandard Bruker pulse sequence NOESYGPPR1D (RD-90°-t1-90°-tm90°-acquisition). The Free Induction Decay signal was automatically zerofilled， and Fourier transform in processing module and baselinecorrection of data was performed in in Chenomx NMR Suite 8.1 (Che-nomx Inc.， Edmonton， Canada). All these spectra were analyzed againstChenomx Compound Library. The residual water (8 4.7-5.5) signals ofthe spectral regions were removed. A total of 32 metabolites werequantified according to the internal standard from 1H NMR spectra. Allmetabolite concentration information were normalized by weight acrossall replicate samples before being used in the later on multivariableanalysis. 2.9. Characterizing the key taste compounds Components which could contribute to the discrimination betweenmodern-processed ham and traditional-processed ham were identifiedby partial least square discriminant analysis (PLS-DA). The candidatetaste components were further calculated the taste-active values (TAVs)so as to confirm their contributions to taste and richness intensities，according to these descriptions (Haseleu， Lubian， Mueller， Shi， & Koe-nig， 2013； Kranz， Viton， Smarrito-Menozzi， & Hofmann， 2018； Liu， Xia，Wang， & Chen， 2019； Zhang， Ayed， Wang， & Liu， 2019). 2.10. Statistical analysis All values were expressed as the mean ± standard deviation. Theumami， richness， water content， relaxation time (T2)， T2 populations，taste compound profiles and taste-active values were analyzed byDuncan’s multiple range test in one-way analysis of variance of SAS 8.0(SAS Institute Inc.， Cary， NC， USA). Student’s t-test model was alsoperformed to compare these parameters at the same sampling pointbetween traditional and modern processing. The significant level was setas 0.05. Hierarchical cluster analysis and principle component analysis(PCA) were also performed to characterize the components and contentsof metabolites among modern-processed and traditional-processedhams. Partial least square discriminant analysis (PLS-DA) was used tofurther characterize the key taste compounds among modern-processed 3. Results and discussion 3.1. Sensory characteristics of ham samples Sensory characteristics of biceps femoris muscle derived from the hamsamples at the end of post-ripening are shown in Fig. 1. The sensoryscores of sweetness， sourness， aftertaste and bitterness did not showobvious difference between traditional-processed ham and modern-processed ham， while significantly higher scores in overall taste， rich-ness and umami were shown in the ham of modern processing than thatof traditional processing. Interestingly， there was a significant decreasein the saltiness scores of modern-processed ham than that of traditional-processed ham， which could be attributed to the fact that these hams ofmodern processing were salted using lower NaCl content， comparedwith the production of traditional processing (Zhou， Wang et al.， 2019a).Consistently， the intensities of umami and richness of ham samplesobviously increased from the raw ham to the end of post-ripening inboth traditional and modern processing； significantly higher intensitiesof umami and richness in modern-processed ham observed by electrictongue were also shown at the later stage of dry-ripening and the end ofpost-ripening compared with the samples of traditional-processed ham.The sensory results indicated that the ham manufactured by modernprocessing would have more intense overall taste， richness and umamiattributes than that of traditional-processed ham. The attributes of overall taste， richness and umami are extremelysignificant quality characteristics of dry-cured ham (Hersleth， Lengard，Verbeke， Guerrero， & Nas， 2011； Zhang et al.， 2018).Processing tech-nology and procedures have a significant effect on organoleptic qualitiesof dry-cured ham (Toldra， 2006； Zhou， Wu et al.， 2019b). In our study，the sensory results showed that overall taste， richness and umami weremore significant than sweetness， sourness， aftertaste and bitterness indescribing the changes of organoleptic attributes of Jinhua ham betweentraditional and modern processing. The difference of overall taste，richness and umami intensities could be attributed to the fact that theham of modern processing accumulated higher profiles in key tastesubstances which could enhance the intensities of overall taste，richnessand umami. 3.2. The changes of texture parameters in both traditional-processed andmodern-processed hams The profiles of hardness， adhesiveness， springiness and cohesivenessare shown in Table 1 during the processing of traditional procedure andmodern procedure in Jinhua ham. As shown in Table 1， the values ofhardness， adhesiveness and cohesiveness significantly increased fromraw ham to the later stage of dry-ripening， whereas the values ofspringiness correspondingly decreased； no obvious difference wasobserved in hardness， springiness and cohesiveness from the later stageof dry-ripening to the end of post-ripening in both traditional-processedham and modern-processed ham， while the increased profile of adhe-siveness was found from the later stage of dry-ripening to the end ofpost-ripening in both traditional and modern processing. Interestingly，no obvious difference was shown in the values of hardness， springinessand cohesiveness in traditional-processed and modern-processed hamsat the later stage of dry-ripening and at the post-ripening stage，indi-cating that these ham samples of traditional processing and modernprocessing showed similar textural profiles. Texture is an important organoleptic characteristic to evaluate thequality of dry-cured ham (Tomovic et al.， 2013； Zhou， Wang et al.，2019a). Proper texture of dry-cured ham is very popular with consumersand texture parameters of dry-cured ham mainly include hardness，adhesiveness， springiness and cohesiveness， which is closely relatedwith the changes of salt penetration， moisture content and ripening timeduring the processing of dry-cured ham (Arnau et al.， 2007； Fulladosa， Table 1The changes of texture parameters of Jinhua ham. Hardness (N) Adhesiveness Springiness Cohesiveness (Nes) raw ham 208.69± -11.39± 0.57± 0.32±0.04B 20.37B 2.07C 0.08A end of dry- 9824.04± -46.48± 0.46± 0.48± ripening 744.64Aa 2.02Bb 0.05Ba 0.07Aa (T) end of dry- 9586.37 ± -65.64 ± 0.42± 0.45± ripening 547.93Aa 4.45Ba 0.06Ba 0.02Aa (M) end of post- 10047.41± -72.85± 0.45± 0.46± ripening 682.75Aa 2.62Aa 0.03Ba 0.03Aa (T) end of post- 10233.65± -80.81± 0.39± 0.42± ripening 723.32Aa 9.36Aa 0.09Ba 0.06Aa (M) T， M represent the sampling of traditional-processed ham and modern-processedham， respectively. A-C Different letters indicate significant difference indifferent processing stages (P <0.05)； a different letters indicate significantdifference in different groups (P <0.05). and water distribution of dry-cured ham (Morales， Arnau， Serra， Guer-rero， & Gou， 2008). In the present study， the values of hardness from208.69 N in raw ham increased to 10047.41 N in traditional-processedham and to 10233.65 N in modern-processed ham， and no obviousdifference was observed in these parameters of hardness， adhesiveness，springiness and cohesiveness at the final product， which indicated thatmodern-processed procedure did not change the development oftextural parameters of Jinhua ham. 3.3. The changes of moisture and water distribution in both traditional-processed and modern-processed hams The changes of moisture content and water distribution of Jinhuaham during the processing of traditional and modern procedures areshown in Fig. 2. The moisture content of biceps femoris muscle samplesdecreased significantly from the raw ham to the end of post-ripening inboth traditional and modern processing (P <0.001)； no obvious dif-ference was observed at the final product of traditional processing andmodern processing (P >0.05)， and the moisture content of biceps femorismuscle was approximately 49.74% and 52.19% at the final product oftraditional and modern processing procedures， respectively. Bound water (T2b) with relaxation time around 0-5 ms is closelyassociated with macromolecules， immobilized water (T21) is a majorcomponent characterized by relaxation time around 20-60 ms， andextramyofibrillar water (T22， free water) is a slower relaxing component(T22) with a relaxation time around 100-400 ms (Garcia et al.， 2015). Fig. 2. The changes of moisture content and water distribution of Jinhua ham. A-C Identical letters indicate that there is no significant difference in differentprocessing stages (P > 0.05)； a-b Identical letters indicate that there is no significant difference at the same processing stage (P >0.05). The transversal relaxation times and corresponding populations oflow-fieldd NMR gained from these samples of traditional- andmodern-processed hams are shown in Fig. 2. The relaxation times ofwater were significantly affected by ripening time. The transverserelaxation time of immobilized water (T21) and the transverse relaxa-tion time offreee water (T22) of traditional-processedandmodern-processed hams significantly decreased from raw ham to thelater stage of dry-ripening (P < 0.001)， but there was no significantchange from the later stage of dry-ripening to the end of post-ripening inboth traditional-processed and modern-processed hams. The transverserelaxation time of bound water (T2b) did not show obvious differencefrom raw ham to the end of post-ripening in both traditional-processedand modern-processed hams. For the changes of corresponding pop-ulations of water， the populations of immobilized water (T21 Pop)significantly increased from raw ham to the later stage of dry-ripening(P < 0.001)， while the populations of free water (T22 Pop) signifi-cantly decreased from raw ham to the later stage of dry-ripening； therewas no significant change in the populations of immobilized water (T21Pop) and free water (T22 Pop) from the later stage of dry-ripening to theend of post-ripening in both traditional-processed andmodern-processed hams. No obvious changes in the populations ofbound water (T2b Pop) were observed during the whole processing oftraditional procedure and modern procedure， while significantly higherpopulations in T22 Pop were shown in modern-processed ham than intraditional-processed ham at the final product. Moisture content and water distribution play a role in developingsensory attributes including texture and juiciness (Garcia， Rodriguez，deAvila Hidalgo， & Bertram， 2016). High content in moisture usually results in soft texture of dry-cured ham and these products with softtexture commonly are rejected by consumers (Morales et al.， 2008). Inour results， the moisture content of traditional-processed andmodern-processed hams decreased from raw ham to the end ofpost-ripening， and the content of moisture decreased to 49.74% and52.19% at final products， respectively. The change could be explainedby the fact that the dehydration of muscle tissues occurred during thesalting and ripening stage (Bajd， Gradisek， Apih， & Sersa， 2016；BosseNee Danz et al.， 2018). Some studies have demonstrated that thedecrease of moisture content showed a great contribution to the increaseof hardness of dry-cured ham (Alino et al.， 2009). Thus， in our study， theincrease of hardness could be attributed to the decrease of moisturecontent and redistribution of water populations during the processing ofJinhua ham. The mobility of water resulted in the loss of water and ashift from P22 populations to P21 populations in our study. The changesin water populations of ham in the present study agreed well on thereports of fermented sausages (Garcia et al.， 2015). This shift can bemainly explained by the fact that the formation of protein networkinduced by salting and ripening holds more immobilized water (Garciaet al.， 2015； Sun， Holley，& Safety， 2011). Some studies demonstratedthat the populations of free water are related to the development ofjuiciness of dry-cured ham (Bertram， Aaslyng， & Andersen， 2005). In ourstudy， higher populations of T22 Pop in modern-processed ham couldcontribute to the higher juiciness scores of modern-processed ham，which is confirmed by previous study of Zhou， Wu et al.(2019b). Representative 600-MHz -H NMR spectra for biceps femoris muscle ofraw ham， traditional-processed ham and modern-processed ham areshown in Figure S1. From these spectra， 32 metabolites were identified，which mainly included free amino acids，organic acids， small peptides，nucleic acids and sugars. These components were further quantifiedbased on the known concentration of 2， 2-dimethyl-2-silapentane-5-sul-fonate standard. To compare the changes of each metabolite， hierar-chical cluster analysis heatmap (Fig. 3) was performed to explore theinformation of metabolite profiles. Samples from raw ham， traditional-processed ham and modern-processed ham displayed different colordistributions， indicating that there were characteristic differences in theprofiles of metabolite among three groups. Taking the Euclidean dis-tance into consideration， the identified metabolites were classified into 2main classes， that the metabolites of raw ham clustered into one group，and the metabolites of traditional-processed and modern-processedhams clustered into another group， which indicated that these samplesof traditional-processed ham and modern-processed ham showed similarmetabolites in terms of the components and profiles. Furthermore，Principle component analysis (PCA) was also employed to characterizethe 32 metabolites identified in these hams and to further evaluate theeffects of processing procedures on metabolite profiles of Jinhua ham.The score plot of PCA (Fig. 2) showed that 94.3% of the variability wasexplained by the first three principal components， accounting for 81.8%，9.5% and 3.0% of the total variance， respectively. These metabolites offumarate， anserine and ADP were the major contributors of first prin-cipal components of PCA model； lactate/lactic acid， creatine， glycerol，niacinamide， creatine phosphate， adenosine， glycine， hypoxanthine andglucose were the key contributors of secondary principal components ofPCA model； valine， acetic acid， tyrosine， phenylalanine， glutamate/glutamic acid，tryptophan， methionine， leucine， isoleucine and carno-sine were the important contributors of third principal components ofPCA model. Overall，the results of PCA and hierarchical cluster analysisheatmap indicated that free amino acids， organic acids， small peptides，nucleic acids and sugars were main components of metabolites of Jinhuaham (Zhang et al.， 2018)， and that these differential components andprofiles were the source of variation in response to processingprocedures. 3.5. Partial least square discriminant analysis (PLS-DA) and taste-activevalues of key taste substances To get more information on dry-cured ham with different processingprocedures (traditional- and modern-processing procedures)， the rela-tionship between metabolites and taste scores and richness intensities was conducted by PLS-DA. X-matrix was designed as the profile of me-tabolites， and Y-matrix was set as taste scores and richness intensities inthe PLS-DA model (Fig.4). The PC1 and PC2 of PLS-DA model explained99.1% of the variance in the X-matrix and 98.6% of the variance in theY-matrix. These results indicated that the PLS-DA model well explainedthese variates including sensory attributes (taste scores and richnessintensities) and most of metabolites. Variable Importance for the Pro-jection plots (VIP) were employed to evaluate the contribution of me-tabolites towards improving taste and richness of the dry-cured ham. Asis shown in Fig. 4， several metabolites including glutamate/glutamicacid， alanine， leucine， lactate， glycerol， anserine， creatine and creatinephosphate showed a high VIP-value (more than 1)， indicating that theseparameters could have a key contribution to Y-variables (taste scoresand richness intensities). Furthermore， the coefficients plots showedthat taste and richness (Fig. 5) were strongly correlated with glutamate/glutamic acid， alanine， leucine， lactate， glycerol， anserine， creatine andcreatine phosphate， respectively. In order to further evaluate the contributions of these metabolites ontaste and richness， these key metabolites were characterized by theanalysis of taste-active values and taste attributes. As shown in Fig. 6，the taste-active values of lactate， glutamate and alanine were over 30，indicating that these metabolites showed great contributions to the tasteand richness of Jinhua ham. In addition， these metabolites includingleucine， anserine，and glycerol also showed the taste-active values withmore than 1， meaning that they had significant effect on the taste andrichness of dry-cured ham. However， the taste-active values of creatinewere less than 1 in both traditional-processed ham and modern-processed ham， indicated that it did not directly contributed to thebitterness of Jinhua ham. The taste attributes of these metabolites werefurther characterized according to previous studies (Haseleu et al.， 2013；Kranz et al.， 2018； Liu et al.， 2019；Zhang， Ayed， Wang， & Liu， 2019). Asshown in Fig. 6， lactate was the major contributor of sourness ofdry-cured ham； glutamate was the main component of umami taste；glycerol and alanine contributed to the sweet taste； anserine showed apositive effect on meaty taste. Interestingly， lactate， anserine， glyceroland creating showed significantly higher taste-active values in thesesamples of modern-processed ham than that of traditional-processedham. These differences in taste-active values could be the main sourceof the variations of taste and richness between modern-processed hamand traditional-processed ham. The taste and richness development of dry-cured ham was related tothe profile of free amino acids， small peptides， organic acids and nucleicacids (Dashdorj et al.， 2015). Furthermore， free amino acids and smallpeptides were identified as the major components of taste substances ofdry-cured ham (Sforza et al.， 2006). The taste and richness attributes ofdry-cured ham could be associated with the amount of free amino acid Fig.3. Hierarchical cluster analysis heatmap and principal component analysis (PCA) on the metabolites. A1，A2， A3， A4， A5， A6 and A7 represent the samples ofraw ham； D1， D2， D3，D4， D5， D6 and D7 represent the traditional-processed ham samples of the end of post-ripening； H1， H2， H3，H4， H5， H6 and H7 represent themodern-processed ham samples of the end of post-ripening. (Keska & Stadnik， 2017). Sweet taste is associated with the content ofthreonine， serine， alanine， glycine and proline； glutamic acid andaspartic acid are responsible for the umami taste of dry-cured ham，whereas the content of phenylalanine， isoleucine， leucine and methio-nine is related to bitter taste of dry-cured ham (Keska & Stadnik， 2017；Sforza et al.， 2006). It has been well documented that 13 amino acids(tyrosine， leucine， phenylalanine， isoleucine， alanine， glutamic acid，valine， methionine， tryptophan， aspartate， asparagine， taurine andglycine) were identified and quantified by 1H NMR， and these aminoacids significantly increased from the raw ham to the final products inboth traditional-processed and modern-processed hams. In meat andmeat products， glutamic acid is one of the main amino acids， accountingfor 20% of all amino acids in meat proteins. Consistently， in our results，the most abundant free amino acids were glutamic acid， alanine，aspartic acid， valine and leucine； the content of glutamic acid andaspartic acid accounted for 20.25% of all amino acids in bothtraditional-processed and modern-processed hams. According to tasteclassification， the content of umami amino acids including aspartic acidand glutamic acid showed largest increase and followed by sweet aminoacids (alanine， glycine and valine) at the final products in bothtraditional-processed and modern-processed hams. These results indi-cated that the increase of umami and sweet amino acids could beresponsible for the improvement of taste and richness from raw ham topost-ripening of Jinhua ham (Keska & Stadnik， 2017). However， in termsof profiles and kinds of free amino acids， there were no obvious differ-ence between traditional-processed and modern-processed hams， indi-cating that free amino acids were not the main source of difference oftaste and richness between traditional-processed ham andmodern-processed ham. Carnosine (histidine-derived dipeptide) and anserine (B-alanyl-1- 1methylhistidine) have been described as taste compounds (Jung et al.，2013). They are widely distributed in the skeletal muscle and aresignificantly positive correlation to sensory characteristics of musclefood. Suyama and Shimizu (1982) demonstrated that carnosine had anotable buffer effect. Studies also reported that carnosine could enhancethe taste attributes ofmusclefoods (Djenane， Martinez，Sanchez-Escalante， Beltran， & Roncales， 2004). Carnosine and anserinecould prolong the taste sensation in oral cavity (Dashdorj et al.， 2015). Inaddition， both carnosine and anserine are able to break unsaturatedaldehydic products and minimize the rancidity in dry-cured meatproducts (Gianelli， Flores， & Toldra， 2003). Furthermore， some studiesdemonstrated that anserine was related to the meaty taste and richnessof muscle food， for examples， the existence of anserine significantlyincreased the richness and taste attributes of beef broth (Pereira-Lima，Ordonez， de Fernando， & Cambero， 2000).These studies indicated thatanserine and carnosine played a role in developing the taste attributes ofdry-cured ham (Dashdorj et al.， 2015). In our results， higher content ofanserine Was shown in modern-processed ham than intraditional-processed ham， which could contribute to the higher rich-ness intensities and taste scores of Jinhua ham with modern-processedprocedure. traditional-processed ham， which could be attributed to the shorterprocessing procedure of modern-processed ham. Creatine and creatinephosphate played a vital role in the energy delivery process of musclecells (Jung et al.， 2013). In terms of taste attributes， creatine has beenidentified as taste-active compounds in dry-cured ham (Zhang et al.，2018) and stewed beef broth (Pereira-Lima et al.， 2000). Schlich-therle-Cerny and Grosch (1998) demonstrated that it was related tobitter taste when taste-active values of creatine were more than 1 andthat the decrease of creatine contributed to the improvement of richnessintensities of stewed beef broth. In the present study， taste-active valuesof creatine did not reach 1， which indicated that it contributed to theimprovement of richness of traditional-processed and modern-processedhams. Previous studies have reported that the content of nucleotidesgradually decreased to undetectable levels during the processing ofParma ham (Zhang et al.， 2018). In our study， four nucleic acids andtheir derivatives were identified among three groups； uracil， adenosineand hypoxanthine were the main components of derivations of nucleicacids. Hypoxanthine is usually derived from the degradation of inosinemonophosphate (Kuda， Fujita， Goto， & Yano， 2008)， and contributes tothe bitter attribute of dry-cured ham. There was no obvious difference inthe content of uracil， adenosine and hypoxanthine， and these metabolites showed a low level in both traditional-processed ham andmodern-processed ham， which further demonstrated that they were notthe major taste substances of Jinhua ham. Glucose and glycerol are the product of glycogen degradation and fatdegradation， which both contribute to sweetness of food matrix (Dash-dorj et al.， 2015). Glucose and glycerol were also identified in our re-sults， and significantly higher content of glycerol was shown inmodern-processed ham than in traditional-processed ham. The highercontent of glycerol could be attributed to the fact that lipolytic enzymesaccelerated the degradation of fat and further enhanced the accumula-tion of glycerol under modern-processed procedure (Zhang et al.， 2011；Zhang， Zhen， Zhang， Zeng， & Zhou， 2009). Furthermore， the analysis ofPLS-DA and taste-active values demonstrated that glycerol showed asignificant contribution to sweetness， which further indicated that highcontent of glycerol contributed to the improvement of taste scores ofmodern-processed ham (Zhang et al.， 2018). 4.Conclusions The modern-processed ham showed higher taste scores and richnessintensities compared with the traditional-processed ham. The shift fromfree water populations to immobilized water populations and the modern procedure decrease of moisture content could contribute to the development oftexture parameters of Jinhua ham during the modern processing. Freeamino acids， small peptides and organic acids were the most intenseresponse in developing the taste and richness of modern-processed ham；glutamic acid， lactate， glycerol， anserine and creatine were responsiblefor the higher taste scores and richness intensities of modern-processedham. CRediT authorship contribution statement Chang-Yu Zhou： Formal analysis， Validation， Visualization， Datacuration， Writing - original draft. Yun Bai： Project administration，Investigation. Chong Wang： Data curation， Validation. Chun-Bao Li：Project administration，Writing- review & editing， Supervision. Xing-Lian Xu： Project administration， Supervision. Dao-Dong Pan： Meth-odology， Writing- review & editing， Supervision. Jin-Xuan Cao：Conceptualization， Ideas， Methodology， Supervision， Funding acquisi-tion， Writing - review & editing. Guang-Hong Zhou： Methodology，Supervision， Funding acquisition， Writing- review & editing. Declaration of competing interest The authors declare that they have no known competing financialinterests or personal relationships that could have appeared to influencethe work reported in this paper. Acknowledgement This work was supported by National Key Research Program of China(2016YFD0401502)， Modern Agricultural Technical Foundation ofChina (CARS-42-25)， Science Foundation of Zhejiang Province(LR18C200003)， Science andTechnologyPrograms ofZhejiang(2019C02085) and Ningbo (2019C10017). Appendix A. Supplementary data Supplementary data to this article can be found online at https：//doi.org/10.1016/j.foodcont.2021.107873. References Alino， M.， Grau， R.， Toldra， F.， Blesa， E.，Pagan， M. J.， & Barat， J. M. (2009). Influence ofsodium replacement on physicochemical properties of dry-cured loin. Meat Science，83(3)， 423-430. Arnau， J.， Serra， X.， Comaposada， J.， Gou， P.， & Garriga， M. (2007). Technologies to shorten the drying period of dry-cured meat products. Meat Science， 77(1)， 81-89.Bajd， F.， Gradisek， A.， Apih， T.， & Sersa， I. (2016). Dry-cured ham tissue characterization by fast field cycling NMR relaxometry and quantitative magnetization transfer.Magnetic Resonance in Chemistry， 54(10)， 827-834. Bertram， H. C.， Aaslyng， M. D.， & Andersen， H. J. (2005). Elucidation of the relationshipbetween cooking temperature， water distribution and sensory attributes of pork -acombined NMR and sensory study. Meat Science， 70(1)， 75-81. Bertram， H. C.， Oksbjerg， N.， & Young， J. F. (2010).NMR-based metabonomics revealsrelationship between pre-slaughter exercise stress， the plasma metabolite profile attime of slaughter， and water-holding capacity in pigs. Meat Science， 84(1)， 108-113. Bosse Nee Danz， R.， Muller， A.， Gibis， M.， Weiss， A.， Schmidt， H.， & Weiss， J. (2018).Recent advances in cured raw ham manufacture. Critical Reviews in Food Science andNutrition， 58(4)， 610-630. Castejon， D.， Garcia-Segura， J. M.， Escudero， R.， Herrera， A.， & Cambero， M. I. (2015).Metabolomics of meat exudate： Its potential to evaluate beef meat conservation andaging. Analytica Chimica Acta， 901， 1-11. Dang， Y.， Gao， X.， Ma， F.， & Wu， X. (2015). Comparison of umami taste peptides in water-soluble extractions of Jinhua and Parma hams. Lebensmittel-Wissenschaft und-Technologie-Food Science and Technology，60(2)， 1179-1186. Dashdorj， D.， Amna，T.， & Hwang， I. (2015). Influence of specific taste-active componentson meat flavor as affected by intrinsic and extrinsic factors： An overview. EuropeanFood Research and Technology， 241(2)，157-171. Djenane， D.， Martinez， L.， Sanchez-Escalante， A.， Beltran， J. A.， & Roncales， P. (2004).Antioxidant effect of carnosine and carnitine in fresh beef steaks stored undermodified atmosphere. Food Chemistry， 85(3)， 453-459. Fulladosa， E.，Serra， X.， Gou， P.， & Arnau， J. (2009). Effects of potassium lactate and highpressure on transglutaminase restructured dry-cured hams with reduced salt content.Meat Science， 82(2)， 213-218. Garcia， A. B. G.， Larsen， L. B.，Rodriguez， C.， I， M.， Diaz， C.， et al. (2015). Proteolysisprocess in fermented sausage model systems as studied by NMR relaxometry. Journalof Agricultural and Food Chemistry， 63(11)，3039-3045. Garcia， A. B. G.， Rodriguez， M. I. C.， de Avila Hidalgo， M. D. R.， & Bertram， H. C. (2016).Water mobility and distribution during dry-fermented sausages “Spanish type”manufacturing and its relationship with physicochemical and textural properties： Alow-field NMR study. European Food Research and Technology， 243(3)， 455-466. Gianelli， M. P.， Flores， M.， & Toldra，F. (2003). Interactions of soluble peptides and proteins from skeletal muscle on the release of volatile compounds. Journal ofAgricultural and Food Chemistry， 51(23)，6828-6834. Gou， P.， Comaposada， J.， & Arnau， J. (2004). Moisture diffusivity in the lean tissue ofdry-cured ham at different process times. Meat Science， 67(2)， 203-209. Haseleu， G.， Lubian， E.， Mueller， S.， Shi， F.， & Koenig， T. (2013). Quantitative studies and taste reconstitution experiments of the sour and lingering mouthful orosensation in a.debittered extract of traditional Japanese dried and fermented skipjack tuna(hongarebushi). Journal of Agricultural and Food Chemistry，61(13)，3205-3211. Hersleth， M.， Lengard， V.， Verbeke， W.， Guerrero， L.， & Naes， T.(2011). Consumersacceptance of innovations in dry-cured ham： Impact of reduced salt content，prolonged aging time and new origin. Food Quality and Preference， 22(1)， 31-41. Jung， S.， Bae， Y. S.， Kim， H. J.， Jayasena， D. D.， Lee， J. H.， Park， H. B.， et al. (2013).Carnosine， anserine， creatine， and inosine 5'-monophosphate contents in breast andthigh meats from 5 lines of Korean native chicken. Poultry Science， 92(12)，3275-3282. Keska， P.， & Stadnik， J. (2017). Taste-active peptides and amino acids of pork meat ascomponents of dry-cured meat products： An in-silico study. Journal of SensoryStudies， 32(6)， e12301. Kranz， M.， Viton， F.，Smarrito-Menozzi， C.， & Hofmann， T. (2018). Sensomics-basedmolecularization of the taste of pot-au-feu， a traditional meat/vegetable broth.Journal of Agricultural and Food Chemistry， 66(1)， 194-202. Kuda， T.， Fujita， M.， Goto， H.， & Yano， T. (2008). Effects of retort conditions on ATP-' related compounds in pouched fish muscle. LWT-Food Science and Technology，41(3)，469-473. Liu， T.T.， Xia， N.， Wang， Q. Z.， & Chen， D. W. (2019).Identification of the non-volatiletaste-active components in crab sauce. Foods，8(8)， 324. Lopez-Pedrouso， M.， Perez-Santaescolastica， C.， Franco， D.， Carballo， J.， Zapata， C.， &Lorenzo，J.M. (2019). Molecular insight into taste and aroma of sliced dry-curedham induced by protein degradation undergone high-pressure conditions. FoodResearch International， 122， 635-642. ( L op ez - P e d ro u so ， M . ， Pe r ez- S a n taescola s t i c a， C .， F r a n c o， D .， F u l l a d osa， E.， C a r b a l l o ， J. ， Z apata， C . ， e t a l . (2018). Comparative proteomic pr o fili n g o f myofibr i llar p r oteins in P d ry - c u re d h a m w i th d i f f er e nt p r o te o l y s is i n dice s an d a dhe s i v e n e s s. Food C h e m is t ry ， 2 44， 23 8- 2 45 . ) Morales， R.， Arnau， J.， Serra， X.， Guerrero， L.， & Gou， P. (2008). Texture changes in dry-cured ham pieces by mild thermal treatments at the end of the drying process. MeatScience， 80(2)， 231-238. Pereira-Lima， C. I.， Ordonez， J. A.， de Fernando， G.D. G.， & Cambero， M. I. (2000).Influence of heat treatment on carnosine， anserine and free amino acid compositionof beef broth and its role in flavour development. European Food Research andTechnology， 210(3)， 165-172. Ruiz-Ramirez， J.， Arnau， J.， Serra， X.， & Gou， P. (2005). Relationship between watercontent， NaCl content， pH and texture parameters in dry-cured muscles. MeatScience， 70(4)， 579-587. Schlichtherle-Cerny， H.， & Grosch，W. (1998). Evaluation of taste compounds of stewedbeef juice. Zeitschrift fur Lebensmittel-Untersuchung und -Forschung A， 207(5)，369-376. Sforza， S.， Galaverna， G.， Schivazappa， C.，Marchelli， R.， Dossena，A.， &Virgili， R. (2006).Effect of extended aging of Parma dry-cured ham on the content of oligopeptides andfree amino acids. Journal of Agricultural and Food Chemistry， 54(25)，9422-9429. Sun， X. D.， & Holley， R. A. (2011). Factors influencing gel formation by myofibrillarproteins in muscle foods. Comprehensive Reviews in Food Science and Food Safety， 10(1)， 33-51. Suyama， M.， & Shimizu， T. (1982). Buffering capacity and taste of carnosine and its methylated compounds. Bulletin of the Japanese Society of Entific Fisheries， 48， 89-95. Toldra， F. (2006). Biochemical proteolysis basis for improved processing of dry-curedmeats. Advanced technologies for meat processing. Advanced technologies for meatprocessing， 13，329. Tomovic， V. M.， Jokanovic， M. R.， Petrovic， L. S.， Tomovic， M. S.， Tasic， T. A.，Ikonic， P. M.，et al. (2013). Sensory， physical and chemical characteristics of cookedham manufactured from rapidly chilled and earlier deboned M. semimembranosus.Meat Science， 93(1)， 46-52. Xiao， Z.， Ge， C.， Zhou， G.， Zhang， W.， & Liao， G. (2019). (1)H NMR-based metaboliccharacterization of Chinese Wuding chicken meat. Food Chemistry， 274， 574-582. Yang，L.， Dai， B.， Ayed， C.， & Liu， Y. (2019). Comparing the metabolic profiles of raw andcooked pufferfish (Takifugu flavidus) meat by NMR assessment. Food Chemistry， 290，107-113. Zhang，N.， Ayed， C.， Wang， W.， & Liu， Y. (2019b). Sensory-guided analysis of key taste-active compounds in pufferfish (Takifugu obscurus). Journal of Agricultural and FoodChemistry， 67(50)，13809-13816. Zhang，J.， Jin， G.， Wang， J.， & Zhang， W. (2011). Effect of intensifying re ripening onlipolysis and lipid oxidation of Jinhua ham. Lebensmittel-Wissenschaft und -Technologie- Food Science and Technology， 44(2)， 473-479. Zhang， J.， Ye， Y.， Sun， Y.， Pan， D.， Ou， C.， Dang，Y.， et al. (2018). 1 H NMR andmultivariate data analysis of the differences of metabolites in five types of dry-curedhams. Food Research International， 113，140-148. Zhang， J.， Yi， Y.， Pan， D.， Zhou， G.， Wang，Y.， Dang， Y.， et al. (2019a). 1H NMR-basedmetabolomics profiling and taste of boneless dry-cured hams during processing. FoodResearch International， 122， 114-122. Zhang， J. H.， Zhen， Z. Y.， Zhang， W. A.， Zeng， T.， & Zhou， G. H. (2010). Effect ofintensifying high-temperature ripening on proteolysis， lipolysis and flavor of Jinhuaham. Journal of the Science of Food and Agriculture， 89(5)， 834-842. Zhou， C. Y.，Pan， D. D.， Bai， Y.， Li， C. B.， Xu， X. L.， Zhou， G. H.， et al. (2019). Evaluatingendogenous protease of salting exudates during the salting process of Jinhua ham.Lebensmittel-Wissenschaft und -Technologie- Food Science and Technology， 101， 76-82. Zhou， C. Y.， Wang， C.， Cai， J. H.， Bai， Y.， Yu， X. B.， Li， C. B.，et al. (2019a). Evaluating theeffect of protein modifications and water distribution on bitterness and adhesivenessof Jinhua ham. Food Chemistry， 293， 103-111. Zhou， C. Y.， Wu， J. Q.， Tang， C. B.， Li， G.， Dai， C.， Bai， Y.， et al. (2019b). Comparing theproteomic profile of proteins and the sensory characteristics in Jinhua ham withdifferent processing procedures. Food Control， 106， 106694. Zhou， G.， & Zhao， G. (2007). Biochemical changes during processing of traditionalJinhua ham. Meat Science， 77(1)，114-120. 摘要：金华火腿是一种高品质的干腌产品，其加工方法对金华火腿的口感属性的发展起着关键作用。为了研究加工方法对金华火腿感官特性和口感物质的影响，本研究重点研究了金华火腿感官特性、质地参数、水分分布和代谢谱的变化。并进一步讨论了代谢物对传统加工和现代加工火腿口感和丰富性发展的贡献。现代加工火腿比传统加工火腿表现出更高的口感评分和丰富度强度。从生火腿到干熟后期，固定化水的数量显著增加，而在传统和现代加工过程中，游离水的数量减少。现代加工火腿成熟期末游离水含量比传统加工火腿高2倍以上。1H nmr代谢组学研究表明，游离氨基酸、小肽和有机酸对现代加工火腿口感和丰富度的影响最为强烈。偏最小二乘判别分析和味觉活性值进一步表明，谷氨酸、乳酸、甘油、鹅氨酸和肌酸是现代加工火腿较高的口感和丰富度强度的原因。关键词：金华火腿   现代加工   代谢物   偏最小二乘判别分析   Taste-active组件   口感和丰富性强烈