Quantitative detection method of kaempferol-3-O-rutinoside in raspberry

Abstract

The invention discloses a quantitative detection method of kaempferol-3-O-rutinoside in raspberry. The quantitative detection method comprises the following steps: drying a raspberry sample, pulverizing and sieving; detecting content data of the kaempferol-3-O-rutinoside by using a high performance liquid chromatography; collecting near infrared spectrum data of the powder; preprocessing the near infrared spectrum data; screening a characteristic wavelength from the preprocessed near infrared spectrum data; the method comprises the following steps: constructing a data set by correlating content data and characteristic wavelength data of kaempferol-3-O-rutinoside in powder; establishing a quantitative analysis model by adopting a partial least square method, and training the quantitative analysis model by adopting the data set; drying and powdering a raspberry sample to be detected, collecting near infrared spectrum data for pretreatment, screening out data at characteristic wavelengths, and inputting the data into the trained quantitative analysis model to obtain content data of kaempferol-3-O-rutinoside in the raspberry sample to be detected. The method disclosed by the invention is simple, convenient, rapid and non-destructive.

Description

Technical Field

[0001] The present invention relates to the field of pharmaceutical technology, and particularly to a method for quantitatively detecting kaempferol-3-O-rutinoside in Rubus chingii Hu. Background Art

[0002] Rubus chingii Hu. is the dried fruit of the plant Rubus chingii Hu. of the genus Rubus in the Rosaceae family. As a traditional medicinal plant, it has a long history of edible and medicinal use in China. It is recorded in "Compendium of Materia Medica" that it has the effects of tonifying the liver and kidney, consolidating essence and reducing urination, and improving eyesight.

[0003] Modern research shows that Rubus chingii Hu. is rich in various antioxidant substances such as vitamin C, flavonoids, kaempferol-3-O-rutinoside, etc. These components endow Rubus chingii Hu. with powerful biological activities such as antioxidant, anti-inflammatory, and anti-cancer. In addition, polyphenolic compounds such as kaempferol and quercetin in Rubus chingii Hu. also show potential for the prevention and treatment of various chronic diseases such as cardiovascular diseases and diabetes.

[0004] Kaempferol-3-O-rutinoside (NFR) is a flavonol rutinoside compound, and its chemical structure is as follows:

[0005]

[0006] Research shows that NFR has potential therapeutic effects on ischemic cerebral infarction and has a certain protective effect on nerve cells.

[0007] The content of medicinal components is one of the key indicators for evaluating the quality of genuine medicinal materials, which is directly related to the curative effect. For example, the chemical components of Artemisia annua L. vary significantly due to geographical location, and various ginsenoside components with different pharmacological properties are produced in ginseng due to different processing methods. Rubus chingii Hu. at different growth stages has different potential markers. These studies show that the chemical components of plants are affected by various factors.

[0008] Therefore, it is necessary to adopt some appropriate evaluation methods to distinguish the quality differences of medicinal materials to ensure their stability. In the Chinese Pharmacopoeia and the prior art, the content detection method for NFR usually adopts high performance liquid chromatography (such as: HPLC method for determining the content of kaempferol-3-O-rutinoside in Daphniphyllum calycinum Benth., Sichuan Journal of Traditional Chinese Medicine, Vol. 35, No. 2, 2017: 57-59; LC-MS / MS method for determining the concentration of kaempferol-3-O-rutinoside in plasma and its application in pharmacokinetic studies of rats, Chinese Journal of Clinical Pharmacology and Therapeutics, 2006.11(5):491-496). High performance liquid chromatography is time-consuming and requires a large amount of organic reagents for sample preparation, which has a greater impact on the health of operators and the quality of products.

[0009] Therefore, it is of great value to use a rapid method to determine the NFR content in raspberries to ensure their quality and effectiveness. Summary of the Invention

[0010] The present invention provides a quantitative detection method for kaempferol-3-O-rutinoside in raspberries, which is simple, rapid and non-destructive.

[0011] The technical solution of the present invention is as follows:

[0012] A quantitative detection method for kaempferol-3-O-rutinoside in raspberries, comprising:

[0013] (1) After drying the raspberry sample, powder it and sieve it.

[0014] (2) Use high performance liquid chromatography to detect the content data of kaempferol-3-O-rutinoside in the raspberry sample powder.

[0015] (3) Collect the near-infrared spectrum data of the raspberry sample powder, and the near-infrared spectrum scanning range is 10000 cm -1 ~4000 cm -1 ;

[0016] (4) Pretreat the near-infrared spectrum data of the raspberry sample powder; the pretreatment is Savitzky-Golay smoothing treatment followed by first derivative treatment.

[0017] (5) Use the characteristic wavelength selection method to screen the characteristic wavelengths from the pretreated near-infrared spectrum data.

[0018] (6) By correlating the content data of kaempferol-3-O-rutinoside in the raspberry sample powder obtained in step (2) and the characteristic wavelength data obtained in step (5), construct a data set;

[0019] Use partial least squares method to establish a quantitative analysis model, and use the data set to train the quantitative analysis model.

[0020] (7) Dry and powder the raspberry sample to be tested, collect the near-infrared spectrum data of the raspberry sample powder to be tested for pretreatment, and screen out the data at the characteristic wavelengths, and input them into the trained quantitative analysis model to obtain the content data of kaempferol-3-O-rutinoside in the raspberry sample to be tested.

[0021] In step (1), after drying and powdering the raspberry sample, sieve it through a 50-100 mesh sieve.

[0022] In step (3), the near-infrared spectrum scanning range is 10000 cm -1 ~4000 cm -1, with a total of 32 scans and a resolution of 8 cm -1 ; The spectra of each sample were measured multiple times, and the average value was used for variable analysis.

[0023] In step (5), the competitive adaptive reweighted sampling algorithm was used to screen the characteristic wavelengths from the preprocessed near-infrared spectral data, including:

[0024] (i) An 80% random sample was used to construct a calibration set for the partial least squares regression model;

[0025] (ii) The wavenumber subset with the minimum root mean square error of cross-validation (RMSECV) in the partial least squares model was selected as the screened characteristic wavelength.

[0026] Furthermore, the number of the characteristic wavelengths is 21; the characteristic wavelengths include 4011.211 cm -1 , 4042.066 cm -1 , 4180.916 cm -1 , 4227.199 cm -1 , 6896.197 cm -1 , 8408.114 cm -1 , 8816.949 cm -1 , 8951.942 cm -1 , 8975.084 cm -1 , 9063.793 cm -1 , 9067.65 cm -1 , 9086.935 cm -1 , 9318.351 cm -1 , 9322.208 cm -1 , 9329.922 cm -1 , 9472.628 cm -1 , 9626.905 cm -1 , 9642.333 cm -1 , 9866.035 cm -1 , 9931.604 cm -1 , 9974.029 cm -1 .

[0027] During the training process of the quantitative analysis model, the evaluation parameters related to the model include the correlation coefficient R 2 , the root mean square error RMSE, and the relative analysis error RPD. The calculation formulas are as follows:

[0028]

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] The present invention adopts near-infrared spectroscopy technology and combines chemometrics technology to rapidly analyze the content of kaempferol-3-O-rutinoside in raspberries. Compared with traditional methods, the sample preparation process of the method of the present invention is convenient, the sample is non-destructive, there is no risk of chemical reagent pollution, and the results are accurate. Description of the Drawings

[0031] Figure 1 It is a scatter plot of the measured values and analytical values of kaempferol-3-O-rutinoside in the training set under the SG+FD+CARS model in Example 2;

[0032] Figure 2 It is a scatter plot of the measured values and analytical values of kaempferol-3-O-rutinoside in the test set under the SG+FD+CARS model in Example 2. Detailed Embodiments

[0033] The present invention will be further described in detail below in conjunction with the drawings and embodiments. It should be noted that the following embodiments are intended to facilitate the understanding of the present invention and do not limit it in any way.

[0034] Example 1: Near-infrared Quantitative Model for Analyzing the Content of Kaempferol-3-O-rutinoside in Raspberries

[0035] 1. Instruments and Materials

[0036] Fourier transform near-infrared spectroscopy analyzer (Thermo Scientific TM Antaris TM II FT-NIR); The HPLC system is the Agilent 1100 series, and the chromatographic column is a Boston BosChrom ODS (4.6×250mm, 5μm, C18) column; Kaempferol-3-O-rutinoside (batch number: MUST-24120712). Acetonitrile (HPLC grade); Methanol (analytical grade) and phosphoric acid (analytical grade, 85%); Raspberry samples, sourced from many places in Zhejiang Province such as Lin'an, Chun'an, and Lishui.

[0037] 2. Sample Preparation

[0038] The raspberry samples are dried, powdered, and passed through a 50-mesh sieve. About 0.5 g of the raspberry powder is precisely weighed and placed in a stoppered conical flask. 50 ml of 70% methanol is precisely added, and the weight is weighed. It is heated under reflux for 1 hour, cooled, and then weighed again. The lost weight is made up with 70% methanol, shaken well, and filtered to obtain the solution.

[0039] 3. Spectral Data Acquisition

[0040] Take 2 g of the raspberry samples that have been dried overnight and place them in a round sample cup (40 mm in diameter and 25 mm in height), and use Thermo Scientific TM Antaris TM II FT-NIR analyzer for analysis. All spectra were measured with the background air spectrum as a blank control. The scanning range of the samples was 4000 - 10000 cm -1 , with 32 scans in total, and the resolution was set to 8 cm -1 . The spectra of each sample were measured 6 times, and the average value was used for variable analysis to reduce the error of non-uniform samples.

[0041] 4. Data preprocessing

[0042] Before performing model analysis, it is necessary to preprocess the NIR data, which can effectively reduce the effects of high-frequency random noise, baseline variation, path length difference, and light scattering. The preprocessing methods adopted in the present invention include Raw data, Normalized Difference Vegetation Index (NDH), Multiplicative Scatter Correction (MSC), Savitzky-Golay (SG) smoothing, Standard Normal Variate (SNV) transformation, First Derivative (FD), Second Derivative (SD), and the combination of the two. Compare the results of different preprocessing methods and select the best preprocessing method.

[0043] 5. Establishment of quantitative analysis model

[0044] When establishing a quantitative analysis model for near-infrared spectroscopy, the Partial Least Squares (PLS) analysis method is a commonly used quantitative model. To evaluate the analytical ability of the model, the following parameters are generally used: the correlation coefficient of the training set (Rc 2 ), the correlation coefficient of the test set (Rp 2 ), the Root Mean Square Error of Calibration (RMSEC) of the training set, the Root Mean Square Error of Prediction (RMSEP) of the test set, and the Ratio of Performance to Deviation (RPD).

[0045] The present invention proves the feasibility of using NIR spectroscopy combined with the PLS model and various preprocessing methods to determine kaempferol-3-O-rutinoside in raspberries. Different preprocessing methods were used to compare the analysis results of the model after feature variable selection under the PLS model. The combination of smoothing followed by first derivative processing was used as the best preprocessing method for kaempferol-3-O-rutinoside. The results comparison under the PLS model is shown in Table 1. Kaempferol-3-O-rutinoside showed the best performance in the SG + FD + PLS model (Rp 2 = 0.5857, RMSEV = 0.0280, RPD = 1.5870).

[0046] Table 1 Analysis results of the PLS model for the content of kaempferol-3-O-rutinoside using different data preprocessing methods

[0047]

[0048] Example 2: Near infrared quantitative model for analysis of kaempferol-3-O-rutinoside content in raspberries

[0049] 1. Instruments and Materials

[0050] Same as Example 1.

[0051] 2. Sample Preparation

[0052] Same as Example 1.

[0053] 3. Spectral Data Collection

[0054] Same as Example 1.

[0055] 4. Data Preprocessing

[0056] A combination of normalization, multivariate scatter correction, smoothing, standard normal variate change processing, and first-order derivative processing was selected.

[0057] 5. Characteristic wave number screening

[0058] The original NIR spectrum of a raspberry sample contains 1557 spectral bands, which have collinearity and high dimensionality problems. In some cases, suitable methods can identify the most effective variables to reduce input variables and improve the accuracy and stability of the model. The competitive adaptive reweighted sampling (CARS) algorithm is a feature variable selection method that combines the Monte Carlo sampling method with the regression coefficient of the PLS model. The algorithm is suitable for high-dimensional spectra. The CARS method optimizes the selection by gradually evaluating, analyzing, filtering and removing each wavelength point in the spectrum. The specific steps are as follows: (1) 80% of random samples are used to construct the calibration set of the PLS regression model; (2) The exponential decreasing function (EDF) is used to eliminate the wavenumbers with lower regression coefficients; (3) Adaptive reweighted sampling (CARS) is used to select the wavenumbers with higher regression coefficients. Finally, the wavenumber subset with the smallest root mean square error of cross validation (RMSECV) in the PLS model is selected.

[0059] 6. Establishment of quantitative analysis model

[0060] (1) The pretreatment method was the first-order derivative after normalization, and then CARS was used to screen the characteristic wavenumbers to construct a near-infrared quantitative model for the analysis of kaempferol-3-O-rutinoside content in raspberry.

[0061] (2) The preprocessing method was the first-order derivative after multivariate scattering correction, and then CARS was used to screen the characteristic wavenumbers to construct a near-infrared quantitative model for the analysis of the kaempferol-3-O-rutinoside content in raspberry.

[0062] (3) The preprocessing method is the first derivative after smoothing, and then CARS is used to screen the characteristic wavenumbers to construct a near-infrared quantitative model for the analysis of kaempferol-3-O-rutinoside content in raspberries.

[0063] (4) The preprocessing method is the first derivative after standard normal variate transformation, and then CARS is used to screen the characteristic wavenumbers to construct a near-infrared quantitative model for the analysis of kaempferol-3-O-rutinoside content in raspberries.

[0064] The PLS results after the above 4 different preprocessing methods combined with CARS screening of characteristic wavenumbers are shown in Table 2.

[0065] Table 2 PLS results after different preprocessing methods combined with CARS screening of characteristic wavenumbers

[0066]

[0067] The preprocessing method is the combination of normalization followed by the first derivative. The near-infrared spectral characteristic wavelengths are 15, and the model is selected as partial least squares method to construct a near-infrared quantitative analysis model for kaempferol-3-O-rutinoside in raspberries. Its test set (Rp 2 = 0.7499, RMSEP = 0.0243, RPD = 2.0573).

[0068] The preprocessing method is the combination of standard normal variate transformation followed by the first derivative. The near-infrared spectral characteristic wavelengths are 34, and the model is selected as partial least squares method to construct a near-infrared quantitative analysis model for kaempferol-3-O-rutinoside in raspberries. Its test set (Rp 2 = 7524, RMSEP = 0.0193, RPD = 2.0528).

[0069] The preprocessing method is the combination of smoothing followed by the first derivative. The near-infrared spectral characteristic wavelengths are 21, and the model is selected as partial least squares method to construct a near-infrared quantitative analysis model for kaempferol-3-O-rutinoside in raspberries. Its test set (Rp 2 = 0.7586, RMSEP = 0.0232, RPD = 2.0789).

[0070] Example 3: Near-infrared quantitative model for the analysis of kaempferol-3-O-rutinoside content in raspberries

[0071] 1. Instruments and materials

[0072] The same as in Example 1.

[0073] 2. Sample preparation

[0074] The same as in Example 1.

[0075] 3. Spectral data acquisition

[0076] The scanning range of the sample is one or more segments from 4000 to 10000 cm -1 The other conditions are the same as those in Example 1.

[0077] 4. Data preprocessing

[0078] Select the preprocessing method of smoothing followed by first derivative processing.

[0079] 5. Establishment of quantitative analysis model

[0080] Adopt the preprocessing method of smoothing followed by first derivative. Select the near-infrared range of the raspberry sample as 4000 cm -1 ~6000 cm -1 . Adopt the partial least squares method to correlate the preprocessed data with the kaempferol-3-O-rutinoside content in raspberries, and construct the near-infrared quantitative model 1 for analyzing the kaempferol-3-O-rutinoside content in raspberries.

[0081] Adopt the preprocessing method of smoothing followed by first derivative. Select the near-infrared range of the raspberry sample as 6000 cm -1 ~8000 cm -1 . Adopt the partial least squares method to correlate the preprocessed data with the kaempferol-3-O-rutinoside content in raspberries, and construct the near-infrared quantitative model 2 for analyzing the kaempferol-3-O-rutinoside content in raspberries.

[0082] Adopt the preprocessing method of smoothing followed by first derivative. Select the near-infrared range of the raspberry sample as 8000 cm -1 ~10000 cm -1 . Adopt the partial least squares method to correlate the preprocessed data with the kaempferol-3-O-rutinoside content in raspberries, and construct the near-infrared quantitative model 3 for analyzing the kaempferol-3-O-rutinoside content in raspberries.

[0083] Adopt the preprocessing method of smoothing followed by first derivative. Select the near-infrared range of the raspberry sample as 4000 cm -1 ~8000 cm -1 . Adopt the partial least squares method to correlate the preprocessed data with the kaempferol-3-O-rutinoside content in raspberries, and construct the near-infrared quantitative model 4 for analyzing the kaempferol-3-O-rutinoside content in raspberries.

[0084] Adopt the preprocessing method of smoothing followed by first derivative. Select the near-infrared range of the raspberry sample as 4000 cm -1 ~10000 cm-1 Using the partial least squares method, the preprocessed data was correlated with the kaempferol-3-O-rutinoside content in raspberries to construct a near-infrared quantitative model for analyzing the kaempferol-3-O-rutinoside content in raspberries 5.

[0085] Table 3 Results of the PLS model for different band selections of near-infrared in raspberries

[0086]

[0087] The band selection was 8000 cm -1 ~10000 cm -1 , the preprocessing method was first derivative processing after smoothing, the model selection was the partial least squares method, and a near-infrared quantitative analysis model for kaempferol-3-O-rutinoside in raspberries was constructed, and its test set (Rp 2 =0.5355, RMSEP = 0.0337, RPD = 1.4596).

[0088] The band selection was 4000 cm -1 ~8000 cm -1 , the preprocessing method was first derivative processing after smoothing, the model selection was the partial least squares method, and a near-infrared quantitative analysis model for kaempferol-3-O-rutinoside in raspberries was constructed, and its test set (Rp 2 =0.5851, RMSEP = 0.0328, RPD = 1.5316).

[0089] The band selection was 4000 cm -1 ~10000 cm -1 , the preprocessing method was first derivative processing after smoothing, the model selection was the partial least squares method, and a near-infrared quantitative analysis model for kaempferol-3-O-rutinoside in raspberries was constructed, and its test set (Rp 2 =0.5857, RMSEP = 0.0280, RPD = 1.5870).

[0090] The above-described embodiments have described in detail the technical solutions and beneficial effects of the present invention. It should be understood that the above is only a specific embodiment of the present invention and is not used to limit the present invention. Any modifications, supplements, equivalent replacements, etc. made within the scope of the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A quantitative detection method for kaempferol-3-O-rutinoside in raspberries, characterized in that, Including: (1) Drying the raspberry samples, pulverizing them, and sieving them; (2) Detecting the content data of kaempferol-3-O-rutinoside in the raspberry sample powder by high performance liquid chromatography; (3) Collect the near-infrared spectral data of the raspberry sample powder, and the near-infrared spectral scanning range is 10,000 cm -1 ~4,000 cm -1 ; (4) Preprocessing the near-infrared spectral data of the raspberry sample powder; the preprocessing is Savitzky-Golay smoothing followed by first derivative processing; (5) Screening characteristic wavelengths from the preprocessed near-infrared spectral data by a characteristic wavelength selection method; (6) Constructing a data set by correlating the content data of kaempferol-3-O-rutinoside in the raspberry sample powder obtained in step (2) and the characteristic wavelength data obtained in step (5); Establishing a quantitative analysis model by partial least squares method and training the quantitative analysis model with the data set; (7) Drying and pulverizing the raspberry sample to be measured, collecting the near-infrared spectral data of the raspberry sample powder to be measured for preprocessing, screening out the data at the characteristic wavelengths, and inputting them into the trained quantitative analysis model to obtain the content data of kaempferol-3-O-rutinoside in the raspberry sample to be measured.

2. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 1, characterized in that, In step (1), after drying and pulverizing the raspberry samples, they are sieved through a 50-100 mesh sieve.

3. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 1, characterized in that, In step (3), the near-infrared spectrum scanning range is 10000 cm -1 ~ 4000 cm -1 , with a total of 32 scans and a resolution of 8 cm -1 ; The spectra of each sample are measured multiple times, and the average value is used for variable analysis.

4. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 1, characterized in that, In step (5), the competitive adaptive reweighted sampling algorithm is used to screen characteristic wavelengths from the preprocessed near-infrared spectral data.

5. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 4, characterized in that, Step (5) includes: (i) Constructing a calibration set of a partial least squares regression model using 80% of the random samples; (ii) Selecting the wavenumber subset with the minimum root mean square error of cross-validation in the partial least squares model as the screened characteristic wavelengths.

6. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 5, characterized in that, The number of the characteristic wavelengths is 21.

7. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 6, characterized in that, The characteristic wavelengths include 4011.211 cm -1 、4042.066 cm -1 、4180.916 cm -1 、4227.199 cm -1 、6896.197 cm -1 、8408.114 cm -1 、8816.949 cm -1 、8951.942 cm -1 、8975.084 cm -1 、9063.793 cm -1 、9067.65 cm -1 、9086.935 cm -1 、9318.351 cm -1 、9322.208 cm -1 、9329.922 cm -1 、9472.628 cm -1 、9626.905 cm -1 、9642.333 cm -1 、9866.035 cm -1 、9931.604 cm -1 、9974.029 cm -1 。 8. The quantitative detection method of kaempferol-3-O-rutinoside in raspberries according to claim 1, characterized in that, During the training process of the quantitative analysis model, the evaluation parameters related to the model include the correlation coefficient R 2 , root mean square error RMSE, and relative analysis error RPD. The calculation formulas are as follows:

Images