Method for nondestructive detection of contents of EPA and DHA in adductor muscle of scallop based on Raman spectrum

Abstract

The invention provides a method for nondestructive detection of contents of EPA and DHA in adductor muscles of scallops based on Raman spectroscopy, which comprises the following steps: taking a proper amount of gill filament tissues in a survival state of the scallops by using a nondestructive gill filament tissue taking method, preparing into frozen slices, performing Raman laser scanning on the slices to obtain original data, performing pretreatment such as normalization, noise reduction, baseline correction and the like, and determining the content of EPA and DHA in the adductor muscles of the scallops. Selecting a characteristic wavelength, and combining with adductor muscle lipidome data to establish a model for predicting the content of unsaturated fatty acid in adductor muscle in a cross-tissue manner by gill filament Raman data. Compared with an existing method for detecting the content of unsaturated fatty acid in scallops, the method has the advantages that the content of unsaturated fatty acid in rhabdoids of main edible parts of the scallops can be quickly, accurately and simply detected under the condition that the scallops survive, and the detection cost is greatly reduced; an efficient and simple technical means is provided for breeding of shellfish quality characters, and popularization and promotion in comprehensive evaluation and genetic improvement work of aquatic germplasm are facilitated.

Description

Technical Field

[0001] The invention belongs to the technical field of shellfish genetic breeding, and particularly relates to a method for nondestructively detecting the contents of EPA and DHA in scallop adductor muscle based on Raman spectroscopy. Background Art

[0002] The comprehensive quality of aquatic products is determined by nutritional value and sensory quality, among which nutritional value depends on the content and composition of biochemical components. The high content of unsaturated fatty acids in shellfish, especially EPA and DHA, is its key value. EPA and DHA help maintain the healthy function of the brain and retina, regulate cell activity and health, reduce the risk of cardiovascular disease, and are of great significance to human health. Therefore, the content of EPA and DHA can be used as an important indicator for evaluating the quality of shellfish, and is also an important goal for the breeding of high-value shellfish varieties. However, in order to carry out large-scale breeding, it is first necessary to ensure that the DHA and EPA content of the parent individuals is measured while they are alive. Therefore, non-lethal determination of breeding individuals is an important technical means for shellfish breeding.

[0003] In the existing methods, the detection of DHA and EPA is mainly through chemical analysis methods such as gas chromatography-mass spectrometry, but the adductor muscle tissue obtained after the scallop is opened for testing cannot guarantee that the individual is still alive. It is necessary to establish a method for detecting non-lethal unsaturated fatty acids in shellfish. Summary of the invention

[0004] The object of the present invention is to provide a method for nondestructively detecting the EPA and DHA content of scallop adductor muscle based on Raman spectroscopy, which can quickly and non-lethally detect the EPA and DHA content of the adductor muscle, the main edible part of scallop, and has lower detection cost.

[0005] The method for non-destructive detection of EPA and DHA content in scallop adductor muscle based on Raman spectroscopy provided by the present invention comprises the following steps: 1) Take part of the gill filament tissue of the living scallop, freeze-embed the obtained gill filament tissue, make a frozen section with a thickness of 5-10 μm and place it on an aluminum-coated slide; 2) Obtaining the original Raman spectrum: Using a micro-Raman spectrometer, the sample obtained in step 1) is laser scanned, and multiple sites are randomly selected for scanning for each sample to obtain the original Raman spectrum of the site; The parameters of the micro-Raman spectrometer were set as follows: a 532 nm laser was used as the excitation light source, the laser power was 10 mW, the grating was 600 g / mm, the spectral range was set to 191-3945 cm-1, the integration time was 7 s, the objective lens was 100 times, and the spot size was 300 nm; before acquisition, the instrument was calibrated with a silicon wafer, and the silicon wafer signal was calibrated to 520.7 cm-1; 3) Gill Raman data processing: After area normalization of the original Raman spectrum obtained in step 2), the average spectrum was calculated, and then the data was subjected to noise reduction and baseline correction. The gill Raman data with characteristic wavelengths of 935, 1312, 1444, 1655, and 2928 cm-1 were screened, and the prediction models y(EPA)=0.8427x+0.0234 and y(DHA)=1.0028x+0.0004 were used to convert them into the EPA and DHA contents of scallop adductor muscle; where x is the screened gill Raman spectrum data set x.

[0006] The method for performing noise reduction and baseline correction processing on Raman data is SG smoothing and multivariate scattering correction (MSC).

[0007] The present invention discloses a method for non-destructively detecting the EPA and DHA content of scallop adductor muscle based on Raman spectroscopy, which specifically comprises the following steps: firstly, a method for non-destructively obtaining gill filament tissue is used to obtain an appropriate amount of gill filament tissue while the scallop is alive, and a frozen section with a thickness of 5-10 μm is prepared; the section is subjected to Raman laser scanning to obtain raw data; after pre-processing such as normalization, noise reduction, and baseline correction, a characteristic wavelength is selected; and a model for cross-tissue prediction of the unsaturated fatty acid content of adductor muscle using gill filament Raman data is established in combination with adductor muscle lipidome data. Compared with existing methods for detecting the unsaturated fatty acid content of scallops, such as mass spectrometry, the determination method of the present invention does not require pre-treatment of the tissue such as grinding, extraction, vortexing, and centrifugation. Instead, the Raman spectrum scan is directly performed on the sample after the tissue is made into frozen sections, and combined with the experimental method of non-destructive sampling of gill filament tissue, the unsaturated fatty acid content of the skeletal muscle, the main edible part of the scallop, can be quickly, accurately, and simply detected while ensuring the survival of the scallop, and the detection cost is greatly reduced, providing an efficient and simple technical means for the selection and breeding of shellfish quality traits, and is conducive to the popularization and promotion in the comprehensive evaluation of aquatic germplasm and genetic improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is the technical roadmap of the present invention.

[0009] Figure 2 This is the Raman spectrum of the gill filament tissue of the Yesso scallop and the characteristic wavelength position marking map.

[0010] Figure 3 This is the effect diagram before and after spectral preprocessing.

[0011] Figure 4 It is a regression plot of the EPA content predicted by the established prediction model and the true value.

[0012] Figure 5 It is a regression plot of the DHA content predicted by the established prediction model and the true value. DETAILED DESCRIPTION

[0013] Raman spectroscopy detection technology is a method for molecular structure and chemical bond analysis based on the Raman scattering effect that has developed rapidly in recent years. It can provide evidence for distinguishing biological tissues at the molecular level through characteristic information such as Raman frequency shift and intensity that vary from biological molecules. In view of the synergy of material components of different tissues, the present invention establishes a cross-tissue prediction model for shellfish by training the gill Raman spectroscopy data and the DHA\EPA content of adductor muscle to achieve non-destructive determination of the DHA content of shellfish adductor muscle.

[0014] The use of spectral technology to detect the quality of aquatic products relies on the establishment of mathematical models. The detection models are widely used in the detection of freshness, quality grading, and adulteration analysis of aquatic products, but most of them are only for qualitative analysis, and quantitative models are rarely reported. The establishment of existing models is mainly based on the consistency of sample composition and properties. Existing simple models cannot be directly applied to heterogeneous tissues. It is necessary to consider the denoising of tissue heterogeneity information and the selection of cross-tissue characteristic spectra to improve the generalization ability of the model. In addition, the amount of micro-collected gill filament tissue is limited, and the embedding morphology of the tissue must be ensured to collect effective spectral signals. Therefore, pre-processing steps such as tissue embedding and continuous sectioning require fixed pre-judgment of the sample position to obtain high-quality sections of micro-gill filament tissue.

[0015] The present invention optimizes the method from two main parts: gill filament tissue section processing and cross-tissue prediction model, thereby providing a method for non-destructive detection of EPA and DHA content in scallop adductor muscle based on Raman spectroscopy, providing a method for preparing trace gill filaments into frozen sections that are easy to collect signals, and establishing a model for predicting unsaturated fatty acid content through cross-tissue spectroscopy by acquiring Raman spectral data of gill filament tissue and combining it with fatty acid content data of shellfish adductor muscle, thereby achieving non-lethal detection and evaluation of scallop quality, and providing an important method for the breeding of high-quality shellfish varieties.

[0016] The present invention is further described in detail below in conjunction with specific implementation modes.

[0017] Example 1: Establishment of a method for non-destructive detection of EPA and DHA content in scallop adductor muscle based on Raman spectroscopy 1. Preparation of samples to be tested: 60 active scallops ( Mizuhopecten yessoensis ), use pointed tweezers to take appropriate amount of gill filament tissue while ensuring the scallop is alive, freeze-embed the gill filament tissue, make 5-10 μm thick frozen sections and place them on aluminum-coated slides, and store them in a -80 ℃ refrigerator for testing. After dissecting the scallop, take no less than 200 mg of adductor muscle tissue, quickly freeze it in liquid nitrogen, and store it in a -80 ℃ refrigerator for testing.

[0018] 2. The determination steps are as follows: Raman raw data measurement: First, set the measurement parameters, use 532nm laser as the excitation light source, laser power is 10 mW, grating is 600 g / mm, and the spectral range is set to 191-3945 cm -1 , integration time is 7 s, 100x objective lens, spot size is about 300 nm. Before acquisition, the instrument is calibrated with a silicon chip, and the silicon chip signal is calibrated to 520.7 cm -1 The sample to be tested was laser scanned using a Raman microscope with set parameters. Ten points were randomly selected for scanning for each sample, and 10 original Raman spectra were obtained, resulting in a total of 600 original spectral data.

[0019] Lipidome data determination: The adductor muscle tissue samples were taken out from the -80℃ freezer and thawed on ice. They were ground into powder using a grinder (30 Hz, 1 min). 50 mg of the ground sample was accurately weighed into a new EP tube, and 150 μL of methanol solution, 200 μL of methyl tert-butyl ether solution, and 50 μL of 36% phosphoric acid solution were added for extraction. Vortex for 3 min and centrifuge at 4°C and 12000 r / min for 5 min. 200 μL of supernatant was blown dry on a nitrogen blower and 300 μL of 15% boron trifluoride methanol solution was added. Vortex for 3 min and keep in a 60 °C oven for 30 min. Cool to room temperature, accurately add 500 μL of n-hexane solution and 200 μL of saturated sodium chloride solution. Vortex for 3 min, centrifuge at 4°C and 12000 r / min for 5 min, and then 100 μL of n-hexane layer solution was transferred for analysis. Lipidome data were collected using gas chromatography-mass spectrometry (GC-MS).

[0020] 3. Data processing and model building: Origin2019 was used to adjust the area average integral of the original Raman signal data to 1 to eliminate the spectral amplitude differences between different samples. The peak average of the parallel Raman data of the same sample after area normalization was taken as the Raman data corresponding to the sample. The Raman spectrum detection of gill filament tissue is interfered by the biological fluorescence background. At the same time, due to the limitation of its own stability (laser, spectrometer), background noise will be generated, and baseline drift will occur, which will have a great impact on the results of Raman spectrum analysis. Therefore, before further data processing and analysis, it is necessary to perform baseline correction and smoothing on the collected Raman spectrum data to eliminate the influence of baseline drift and noise. After comparing the preprocessing effects of several commonly used preprocessing methods (SG smoothing, median smoothing, standard normal transformation, multivariate scattering correction) separately and in combination, it was found that using TheUnscrambler X 10.4 software, first using SG smoothing for curve smoothing and then using multivariate scattering correction for baseline correction, is the best preprocessing method for gill filament Raman spectrum preprocessing, so this method was used to reduce noise and correct baseline for data in the future.

[0021] MATLAB R2014B was used to select the continuous projection algorithm (SPA) to couple the cross-tissue spectral data. The gill filament spectral characteristic wavelengths with adductor muscle prediction indicators were screened through the correlation between the gill filament Raman data and the adductor muscle lipid content. The screened characteristic wavelengths were 935, 1312, 1444, 1655, and 2928 cm-1, which showed the strong characteristics of the gill filament Raman data and the strong correlation with the adductor muscle lipid group data. The screened gill filament Raman spectral data were used as the independent variable data set x for model establishment. The lipid group data first calculated the total fatty acid content of each adductor muscle sample, and then calculated the proportion of EPA and DHA content compared to the total fatty acid content, which was used as the dependent variable data set y recommended by the model.

[0022] Commonly used models for Raman prediction include support vector machine regression model (SVR), least squares regression model (PLSR), and principal component regression model (PCR). After comparing the accuracy and robustness of the three models, the PLSR model performs best. The number of training sets is set to 60, and the prediction models obtained are y(EPA)=0.8427x+0.0234 and y(DHA)=1.0028x+0.0004.

[0023] Example 2: Model performance test 1. Sample preparation: 40 scallops were randomly selected from the group and appropriate amounts of gill filament tissue and adductor muscle tissue were taken. The gill filament tissue was cryo-embedded and made into 5-10 μm thick frozen sections placed on aluminum-coated slides as samples for Raman spectroscopy. Adductor muscle tissue weighing no less than 200 mg was taken, quickly frozen in liquid nitrogen, and then subjected to a series of pre-treatment processes, and lipidome data were collected using gas chromatography-mass spectrometry (GC-MS).

[0024] 2. Data determination Raman raw data measurement: First, set the measurement parameters, use 532nm laser as the excitation light source, laser power of 10 mW, grating of 600 g / mm, spectral range of 191-3945 cm-1, integration time of 7 s, 100x objective lens, spot size of about 300 nm. Before acquisition, calibrate the instrument with silicon wafer and calibrate the silicon wafer signal to 520.7 cm-1. Use the micro-Raman spectrometer with set parameters to perform laser scanning on the sample to be tested. Randomly select 10 points for scanning for each sample to obtain 10 raw Raman spectra, and a total of 400 raw spectral data are obtained.

[0025] Lipidome data determination: Take out the adductor muscle tissue sample from the -80℃ freezer, thaw on ice, and grind it into powder using a grinder (30 Hz, 1 min). Accurately weigh 50 mg of the ground sample into a new EP tube, add 150 μL of methanol solution, 200 μL of methyl tert-butyl ether solution, and 50 μL of 36% phosphoric acid solution for extraction. Vortex for 3 min and centrifuge at 4°C and 12000 r / min for 5 min. Take 200 μL of the supernatant and blow dry it on the nitrogen blower, and add 300 μL of 15% boron trifluoride methanol solution. Vortex for 3 min and keep it in a 60 °C oven for 30 min. Cool to room temperature, accurately add 500 μL of n-hexane solution and 200 μL of saturated sodium chloride solution. Vortex for 3 min, centrifuge at 4°C and 12000 r / min for 5 min, and then take 100 μL of the n-hexane layer solution for analysis. Lipidome data were collected using gas chromatography-mass spectrometry (GC-MS).

[0026] 3. Model performance evaluation The measured Raman raw data was normalized, that is, the area average integral of the raw Raman signal data was adjusted to 1, and then the curve was smoothed using SG smoothing, and then the baseline was corrected using multivariate scattering correction. The selected characteristic wavelength peaks were selected as the independent variable quantity set x. Using MATLAB R2014B software, the data set x was used as the model input file and applied to the cross-tissue prediction model of EPA / DHA content listed in Table 1. The predicted content values ​​of adductor muscle EPA and DHA content were obtained by running the predict command.

[0027] The coefficient of determination (R) between the EPA / DHA content values ​​predicted by the calculation model and the actual EPA / DHA content values ​​measured by GC-MS 2 ), R 2 It is used to measure the performance of the model in explaining the variance of the target variable. Its value range is between 0 and 1, where 1 indicates perfect fit and 0 indicates that the model cannot explain the variance of the target variable, so as to evaluate the prediction performance of the model. The specific test results are shown in Table 1: Table 1: Prediction model performance test results Prediction object EPA DHA Regression equation y=0.8427x+0.0234 y=1.0028x+0.0004 <![CDATA[R of the full-spectrum prediction model 2 > 0.544 0.617 <![CDATA[R after screening characteristic wavelengths 2 > 0.867 0.860 Compared with the R predicted by the full spectrum of gill filament Raman without characteristic wavelength screening 2 The correlation coefficients of the training models obtained by the traditional training method were only 0.544 and 0.617; the prediction accuracy of the model trained by the screened cross-tissue coupling spectrum was significantly improved, and the R2 of the prediction models of EPA and DHA were calculated to be above 0.86, indicating that the cross-tissue prediction characteristic wavelengths screened by the present invention have strong characteristics and strong correlation with EPA and DHA contents, and efficient and accurate cross-tissue prediction can be performed through gill filament spectra.

[0028] As shown in Table 1, the model established according to the present method for cross-tissue prediction of the unsaturated fatty acid content of adductor muscle tissue using the Raman spectrum of gill filament tissue has a determination coefficient between the predicted value and the true value greater than 0.85, indicating that the model has good prediction performance. Therefore, the determination method of the present embodiment can be applied to the detection of scallop quality traits; the determination method of the present embodiment can quickly, accurately and non-destructively detect the unsaturated fatty acid content of scallops, and greatly reduce the detection cost, and can be promoted and applied in aquatic breeding.

[0029] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0030] The technical solution provided by the present invention is introduced in detail above. Specific examples are used in this article to illustrate the principle and implementation mode of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation mode and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for nondestructive detection of EPA and DHA content in scallop adductor muscle based on Raman spectroscopy, characterized in that: The method comprises the following steps: 1) Take part of the gill filament tissue of the living scallop, freeze-embed the obtained gill filament tissue, make frozen sections and place them on aluminum-coated slides; 2) Obtaining the original Raman spectrum: Using a micro-Raman spectrometer, the sample prepared in step 1) is laser scanned, and multiple sites are randomly selected for scanning for each sample to obtain the original Raman spectrum of the site; 3) Gill Raman data processing: After area normalization of the original Raman spectrum obtained in step 2), the average spectrum is calculated, and then the data is subjected to noise reduction and baseline correction. The gill filament Raman data of the characteristic wavelength is screened, and the prediction models y(EPA)=0.8427x+0.0234 and y(DHA)=1.0028x+0.0004 are used to convert them into the EPA and DHA contents of the scallop adductor muscle; where x is the screened gill filament Raman spectrum data set.

2. The method according to claim 1, characterized in that In the above 1), frozen sections with a thickness of 5-10 μm are prepared.

3. The method according to claim 1, characterized in that The parameters of the micro-Raman spectrometer in the above 2) are set as follows: a 532 nm laser is used as the excitation light source, the laser power is 10 mW, the grating is 600 g / mm, the spectral range is set to 191-3945 cm-1, the integration time is 7 s, the objective lens is 100 times, and the spot size is 300 nm; before acquisition, the instrument is calibrated with a silicon wafer, and the silicon wafer signal is calibrated to 520.7 cm-1.

4. The method according to claim 1, characterized in that In the above 2), a plurality of sites are randomly selected, that is, 10 sites are selected.

5. The method according to claim 1, characterized in that The gill filament Raman data of the characteristic wavelengths in the above 3) are gill filament Raman data of the characteristic wavelengths of 935, 1312, 1444, 1655, and 2928 cm-1.

6. The method according to claim 1, characterized in that In the above-mentioned step 3), the SG smoothing method is used to reduce the noise of the Raman data.

7. The method according to claim 1, characterized in that The baseline correction processing of the Raman data in 3) is performed using multivariate scattering correction.

Images