Study On Changes Of Antioxidant Enzymes And Molecular Hyperspectral Detection Of Tan Sheep Meat During Cold Storage

Tan sheep is one of characteristics and advantageous agricultural products in Ningxia.The autonomous region had included the high-quality development of Tan sheep industry in the "14th Five-Year Plan" to promote the refined,high-end,and green development of Tan sheep industry.Lipid oxidation is an important reason for the deterioration of mutton quality during storage.The content of endogenous antioxidant active molecules and malondialdehyde in meat is a significant parameter reflects the antioxidant capacity and the degree of lipid oxidation of body.The experiment researched the changes of antioxidant activity molecules and malondialdehyde during the different storage times under 4℃ storage conditions,which combined the microhyperspectral technology(400～1000 nm)to detect and research the activity of total superoxide dismutase(T-SOD),catalase(CAT)and malondialdehyde content rapidly,then constructed an optimal prediction model combing with region of interest selection,outlier elimination,spectral optimization and other methods,which provides technical means for low-cost,fast and green testing of cold fresh mutton.The main research contents and results are as follows:(1)The activities of T-SOD,CAT,peroxidase(POD)and the contents of glutathione(GSH)and MDA in meat samples of Tan sheep during storage at 4 ℃(1～10 d)were statistically analyzed.It showed that the MDA content was increasing gradually,indicating that the lipid oxidation degree of mutton continued to deepen.T-SOD activity and GSH content showed a simultaneous changing trend,CAT and POD activity showed an overall downward trend.The MDA content in the front legs of Tan sheep was significantly higher than the hind legs and longissimus dorsi(P<0.05),indicating the degree of lipid oxidation in front legs was higher,which was related to the large difference in the amount of exercise in different parts of the Tan sheep.The correlation analysis showed that the MDA was significantly positively correlated with T-SOD and GSH(P<0.01),T-SOD was significantly positively correlated with GSH and CAT(P<0.01),and CAT was significantly positively correlated with POD(P<0.01).Thus,the antioxidation defense system of Tan sheep meat are form with T-SOD,CAT,POD and GSH jointly.(2)The cytoplasmic region of 180 Tan sheep meat samples at different storage periods were selected as the areas highly correlated with the detection indicators to calculate the original spectral values by the hematoxylin eosin(HE)staining of cell and threshold method.The 8,9,11 abnormal samples of T-SOD activity,CAT activity,MDA content respectively,were eliminated by the combination of the Monte Carlo sampling method and the PLS prediction error.Six pretreatment methods were used to optimize the full band spectral data.The results showed that the normalize method had the most significant optimization effect among T-SOD(Rc=0.8166,Rp=0.8006,RMSEC=0.1255,RMSEP=0.1301).In CAT,the baseline method was the best spectral pretreatment method(Rc=0.8224,RMSEC=0.0700,Rp=0.8202,RMSEP=0.0726).In MDA,the effects of the PLSR model were improved only with baseline(Rc=0.7966,Rp=0.7788,RMSEC=0.0470,RMSEP=0.0590).(3)Establishing a prediction model of T-SOD activity,CAT activity and MDA content based on different bands of sheep meat.CARS-LSSVM was the best prediction model for T-SOD activity,the correlation coefficient of the calibration set and the prediction set was 0.8816 and 0.8647 respectively.The best prediction model for CAT activity was CARS-PLSR,with Rc=0.8340,Rp=0.8403.The CARS-LSSVM model had a relatively good performance in MDA content,with Rc=0.8289,Rp=0.8364.The results represented that the CARS was the best feature wavelength extraction method,which can effectively reduce data redundancy and improve model performance.In addition,it is feasible to simultaneously determine T-SOD,CAT and MDA in mutton by using microhyperspectral imaging.