An evaluation method for distinguishing quality of different species of momordica grosvenori

Abstract

The application discloses an evaluation method for distinguishing different quality of momordica grosvenori of different origins. The application adopts high performance liquid chromatography to establish the fingerprint of momordica grosvenori, carries out gradient elution by taking water-acetonitrile as a mobile phase, and detects the wavelength of 203 nm to obtain the HPLC fingerprint of momordica grosvenori; the obtained chromatogram is introduced into a traditional Chinese medicine chromatographic fingerprint similarity evaluation system to generate a control fingerprint, different origins of momordica grosvenori are distinguished through similarity comparison and / or chemical pattern recognition analysis; and chemical markers with VIP>1 such as 5-hydroxymethyl furfural, momordica grosvenori flavonoid, momordica grosvenori saponin V and momordica grosvenori saponin IIIA2 are screened. The application can effectively distinguish two main origins of green skin fruit and red hair fruit, overcomes the defects that traditional trait identification is subjective and it is difficult to distinguish similar appearance samples, and provides an objective, accurate and stable technical means for origin identification, quality consistency evaluation and authenticity identification of momordica grosvenori.

Description

Technical Field

[0001] This invention belongs to the field of traditional Chinese medicine quality evaluation technology, specifically relating to an evaluation method for distinguishing the quality of different varieties of Luo Han Guo (Siraitia grosvenorii), and more particularly to a method for distinguishing between green-skinned and red-haired Luo Han Guo varieties based on high-performance liquid chromatography fingerprinting combined with chemical pattern recognition technology. Background Technology

[0002] Monk fruit ( Siraitia grosvenorii Monk fruit (Siraitia grosvenorii) is the dried fruit of the cucurbitaceae plant Siraitia grosvenorii, a specialty medicinal and edible resource from Guilin, Guangxi, my country. It possesses properties such as clearing heat and moistening the lungs, relieving sore throat and hoarseness, and promoting bowel movements. It is widely used in traditional Chinese medicine preparations, health foods, and food additives. The main active ingredients of monk fruit are mogrosides, which exhibit various biological activities including high sweetness, low calories, antioxidant properties, and blood sugar reduction.

[0003] Currently, the main sources of monk fruit circulating in the market include two categories: "green-skinned fruit" and "red-haired fruit." Green-skinned fruit is of superior quality, with a high content of monk fruit saponins, and is recognized in the industry as a high-quality source. Red-haired fruit, due to its relatively lower internal quality, has poorer market acceptance. However, after drying and processing, the two types of monk fruit have extremely similar appearances, making it difficult to accurately distinguish them using traditional methods of visual identification. This has led to the phenomenon of red-haired fruit being passed off as green-skinned fruit, and inferior products being sold as superior ones, seriously affecting the quality stability and clinical efficacy of monk fruit medicinal materials and products.

[0004] In the existing technology, the quality evaluation methods of monk fruit mainly include morphological identification, thin-layer chromatography identification, and content determination of single or a few indicator components (such as monk fruit saponin V). These methods have the following shortcomings: (1) Morphological identification is highly subjective and it is difficult to distinguish different seed sources with similar appearances; (2) The detection of single or a few indicator components cannot fully reflect the overall characteristics of the chemical composition of monk fruit and it is difficult to effectively distinguish samples from different seed sources; (3) There is a lack of systematic and objective chemical fingerprint evaluation methods, which cannot provide a reliable basis for judging the authenticity of monk fruit seed sources.

[0005] High-performance liquid chromatography (HPLC) fingerprinting technology can comprehensively reflect the overall distribution characteristics of various chemical components in medicinal materials. Combined with chemical pattern recognition methods (such as cluster analysis, principal component analysis, and orthogonal partial least squares discriminant analysis), it can effectively uncover differential markers between samples from different sources and has shown promising application prospects in the identification and quality evaluation of various Chinese medicinal materials. However, there are currently no systematic reports on combining HPLC fingerprinting with chemical pattern recognition technology to distinguish the quality of different seed sources (green-skinned fruit and red-haired fruit) of Luo Han Guo.

[0006] Therefore, establishing a quality evaluation method that can comprehensively, objectively, and accurately distinguish between different varieties of monk fruit is of great significance for regulating the monk fruit market order and ensuring the quality of monk fruit-related products. Summary of the Invention

[0007] In order to overcome the shortcomings of existing technologies that rely on appearance for identification, which is highly subjective, and that single or a few indicators cannot fully reflect the chemical characteristics of monk fruit, and cannot effectively distinguish between different varieties of monk fruit such as green-skinned and red-haired monk fruit, this invention provides an evaluation method for differentiating the quality of monk fruit from different varieties based on high-performance liquid chromatography fingerprinting combined with chemical pattern recognition technology.

[0008] Specifically, the technical problems to be solved by this invention include: (1) establishing an HPLC fingerprint method that can comprehensively characterize the overall features of the chemical components of monk fruit; (2) achieving accurate classification of monk fruit from different sources by combining fingerprint similarity evaluation with chemical pattern recognition (cluster analysis, principal component analysis, orthogonal partial least squares discriminant analysis); and (3) screening and confirming key chemical markers that distinguish different sources, providing a scientific basis for the identification and quality evaluation of monk fruit sources.

[0009] The method of this invention is simple to operate, has good reproducibility and high distinguishability. It can effectively solve the problem of difficulty in distinguishing between different species of monk fruit that have similar appearances after drying and processing, prevent inferior products from being passed off as superior ones, and ensure the quality stability and clinical efficacy of monk fruit medicinal materials and related products. It has good practical application value.

[0010] The present invention provides a method for evaluating the quality of monk fruit from different sources, comprising the following steps: S1. Obtain the test solution of the monk fruit sample to be tested; S2. The test solution is detected by high performance liquid chromatography to obtain the chromatogram of the monk fruit sample to be tested; S3. Compare the chromatogram with the standard fingerprint spectrum of monk fruit, and distinguish monk fruit from different sources through chemical pattern recognition analysis.

[0011] The different seed sources involved in this invention include green-skinned fruit and red-haired fruit; Green-skinned and red-haired varieties are the two main seed sources circulating in the current Luo Han Guo (monk fruit) market. After drying and processing, the fruits are extremely similar in appearance and difficult to distinguish with the naked eye. However, there are significant differences in their internal quality: green-skinned fruits have a higher content of active ingredients such as mogrosides and are of superior quality, recognized as a high-quality seed source within the industry; while red-haired fruits have relatively lower internal quality and are less accepted in the market. Due to the difficulty in distinguishing them by appearance, there is a phenomenon in the market of red-haired fruits being used to impersonate green-skinned fruits, seriously affecting the quality stability and clinical efficacy of Luo Han Guo medicinal materials and products. Therefore, this invention uses green-skinned and red-haired fruits as the main distinguishing factors, and uses high-performance liquid chromatography fingerprinting combined with chemical pattern recognition technology to explore the differences in the overall outline of their chemical components, achieving accurate differentiation.

[0012] It should be emphasized that the method of the present invention is not limited to the two seed sources of green-skinned fruit and red-haired fruit. For other seed sources (such as long-barn fruit, Lajiang fruit, etc.), the method of the present invention can also be used to evaluate and distinguish the quality of Luo Han Guo. This is a simple extension of the technical solution of the present invention and should also fall within the protection scope of the present invention.

[0013] The chemical markers used in the evaluation method of this invention to distinguish between green-skinned fruit and red-skinned fruit include one or more of 5-hydroxymethylfurfural, mogroflavin, mogroside V, and mogroside IIIA2. Among them, 5-hydroxymethylfurfural is usually related to Maillard reaction and thermal degradation during processing. Its content varies significantly among different species of monk fruit after processing, with a relatively high content in green-skinned monk fruit, which can be used as one of the distinguishing indicators. Mogroside V is a flavonoid component, and its distribution varies significantly among different species, with a relatively high content in red-skinned monk fruit, which contrasts sharply with green-skinned monk fruit. Mogroside V is the main sweet and active ingredient in monk fruit, and its content is significantly higher in green-skinned monk fruit than in red-skinned monk fruit. It is a core indicator for evaluating the quality of monk fruit and a key marker for distinguishing the two species. Mogroside IIIA2 is a mogroside component, and its accumulation pattern varies among different species. Its content is lower in green-skinned monk fruit, showing an inverse trend with mogroside V. Using it in combination can enhance the accuracy of differentiation.

[0014] In practical applications, the content difference or peak area ratio of one or more of the above-mentioned markers can be used as auxiliary discrimination indicators. A single marker (such as mogroside V) already has good distinguishing ability, while combining multiple markers can further improve the discrimination accuracy and stability. Those skilled in the art can flexibly choose the combination of markers according to the actual detection conditions and discrimination requirements, all of which fall within the protection scope of this invention.

[0015] It should be noted that the screening and confirmation of the above-mentioned chemical markers are based on the OPLS-DA model and VIP analysis established in this invention. The differences in the markers are a specific manifestation of the overall differences in the internal chemical components of different strains of *Siraitia grosvenorii*. This invention does not exclude other unidentified common peaks that can also be used as auxiliary markers. All chromatographic peaks with VIP > 1 screened according to the method of this invention fall within the scope of the technical concept of this invention.

[0016] In the evaluation method of this invention, in step S1, the preparation method of the test solution is as follows: take dried monk fruit, crush it, add organic solvent to extract it, filter it, and the solution is obtained. The organic solvent is methanol, and the extraction is ultrasonic extraction for 30-60 minutes.

[0017] In the evaluation method of this invention, in step S2, the standard fingerprint spectrum of Luo Han Guo is obtained by detecting Luo Han Guo samples of known origin using the high performance liquid chromatography method in step S2, and using the control fingerprint spectrum generated by the Chinese medicine chromatographic fingerprint spectrum similarity evaluation system. The conditions for the high-performance liquid chromatography method include: The chromatographic column was packed with octadecylsilane-bonded silica gel. Mobile phase: Gradient elution was performed using pure water as mobile phase A and acetonitrile as mobile phase B. The flow rate is 0.8-1.2 mL / min; The column temperature is 25-35 ℃; The detection wavelength is 203 nm; The injection volume is 5-20 μL.

[0018] From 0 to 10 min, mobile phase A was maintained at 88% and mobile phase B at 12%. In 10-15 minutes, mobile phase A decreased from 88% to 85%, while mobile phase B increased from 12% to 15%. Over 15-30 minutes, mobile phase A decreased from 85% to 80%, while mobile phase B increased from 15% to 20%. Over 30-40 minutes, mobile phase A decreased from 80% to 75%, while mobile phase B increased from 20% to 25%. Over 40-60 minutes, mobile phase A decreased from 75% to 72%, while mobile phase B increased from 25% to 28%. Over 60-65 minutes, mobile phase A decreased from 72% to 70%, while mobile phase B increased from 28% to 30%. Over 65-85 minutes, mobile phase A decreased from 70% to 15%, while mobile phase B increased from 30% to 85%. For 85-90 min, mobile phase A is maintained at 15% and mobile phase B is maintained at 85%.

[0019] In the evaluation method of this invention, the chemical pattern recognition analysis includes one or more of cluster analysis (HCA), principal component analysis (PCA), and orthogonal partial least squares discriminant analysis (OPLS-DA).

[0020] When the chemical pattern recognition analysis is performed as cluster analysis, the area of ​​common peaks is used as the variable, and Ward's minimum variance method (i.e., the sum of squared deviations method) and Euclidean distance are used to construct a hierarchical clustering dendrogram. Cluster analysis is an unsupervised pattern recognition method, whose advantage lies in that it does not rely on pre-grouped information, but only on the natural classification based on the chemical characteristic distance between samples. The dendrogram allows for a visual observation of the clustering trend of different batches of monk fruit samples, determining whether the samples are naturally grouped according to their provenance. If the green-skinned fruit samples and the red-haired fruit samples cluster into different branches, and the samples within each group are closely clustered, it indicates that there are significant chemical differences between the two sources, and the clustering results can serve as a preliminary basis for provenance differentiation.

[0021] In the case of principal component analysis (PCA), the common peak area is imported as a variable into SIMCA software (or other multivariate statistical analysis software) for unsupervised pattern analysis, generating score plots and loading plots. PCA reduces the original multiple variables into a few principal components through dimensionality reduction, thus displaying the overall distribution of the samples in two-dimensional or three-dimensional space. The score plot can intuitively reflect the separation trend between different samples: if green-skinned fruit and red-skinned fruit cluster in different regions on the score plot without obvious overlap, it indicates that there is a stable difference in the overall chemical composition profile between the two sources. The loading plot is used to identify variables (i.e., chromatographic peaks) that contribute significantly to the principal components, assisting in the search for potential differential components. PCA does not rely on sample classification labels, and the results are objective and reproducible, making it a commonly used method for verifying the distinguishing ability of fingerprint spectra.

[0022] When the chemical pattern recognition analysis is performed using orthogonal partial least squares discriminant analysis, the common peak area is imported into SIMCA software to establish an OPLS-DA model. The VIP value is calculated, and peaks with VIP > 1 are selected as chemical markers to distinguish different provenances. VIP > 1 indicates that the variable makes a significant contribution to the model's classification and can serve as a key indicator for quality evaluation. Through OPLS-DA, not only can accurate discrimination of different provenances of *Siraitia grosvenorii* be achieved, but the key chemical components causing these differences can also be revealed, providing a scientific basis for establishing provenance-specific quality standards.

[0023] In the evaluation method of this invention, cluster analysis, principal component analysis, and orthogonal partial least squares discriminant analysis can be used individually or in combination. Preferably, unsupervised preliminary exploration is first conducted using PCA or HCA to confirm the natural clustering trend of the samples, and then a supervised discriminant model is constructed using OPLS-DA to further verify the discrimination effect and screen biomarkers. The three methods corroborate and complement each other, enabling a comprehensive and systematic evaluation of the quality differences of Luo Han Guo from different sources. Those skilled in the art can choose one or more of these methods for data analysis according to actual needs, all of which fall within the protection scope of this invention.

[0024] The evaluation method of the present invention also includes a step of evaluating the similarity of the sample to be tested: the chromatogram of the sample to be tested is compared with the standard fingerprint of the sample to be tested. If the similarity is higher than the intra-group threshold, it is determined to be the same source; if the similarity is lower than the inter-group threshold, it is determined to be different sources.

[0025] This invention, through extensive experimental research, has revealed significant differences in within-group and between-group similarities among different strains of *Monk Fruit*. Taking the green-skinned variety and the red-haired variety as examples: The similarity within the same variety (such as between different green-skinned fruits) is relatively high, usually not less than 0.90 (actual data: the similarity within the green-skinned fruit group is 0.908-0.987, and the similarity within the red-haired fruit group is 0.988-0.999). The similarity between different varieties (between green-skinned fruit and red-haired fruit) is significantly reduced, usually not higher than 0.92 (actual data: inter-group similarity is 0.626-0.922).

[0026] Based on the aforementioned differences, this invention sets intra-group and inter-group thresholds for similarity determination. If the similarity between the chromatogram of the sample to be tested and the standard fingerprint chromatogram of a known species is higher than or equal to a preset intragroup threshold (e.g., 0.90, which can be adjusted according to the actual verification results), then the sample to be tested is determined to belong to that species. If the similarity between the chromatogram of the sample to be tested and the standard fingerprint chromatograms of all known provenances is lower than the intergroup threshold (e.g., 0.85), then the sample to be tested is determined to be different from all known provenances and may be from other provenances. If the similarity is between the intra-group threshold and the inter-group threshold, a comprehensive judgment can be made by combining chemical pattern recognition (such as OPLS-DA) or chemical marker content.

[0027] The similarity evaluation steps described in this invention are simple to operate and quick to calculate, requiring no complex multivariate statistical analysis to preliminarily determine the provenance, making it particularly suitable for rapid screening of large batches of samples. Furthermore, similarity evaluation and chemical pattern recognition analysis complement each other; similarity evaluation provides qualitative identification of the overall profile, while chemical pattern recognition (especially OPLS-DA and VIP screening) provides quantitative confirmation of differential markers. The combination of both significantly improves the accuracy and reliability of provenance identification.

[0028] The method for establishing HPLC fingerprints to distinguish the quality of green-skinned and red-haired monk fruit, which is involved in this invention, also falls within the scope of protection of this invention.

[0029] The evaluation method of this invention and the HPLC fingerprint spectrum established therefrom can be widely applied in various fields such as the identification of the germplasm of Luo Han Guo, the evaluation of quality consistency, and the identification of authenticity. Specific analysis is as follows: Application in the identification of monk fruit provenance: This invention establishes standard HPLC fingerprint chromatograms of monk fruit from different provenances, and combines similarity evaluation and chemical pattern recognition analysis to accurately identify the provenance of monk fruit samples. Specifically, the chromatogram of the sample is compared with the standard fingerprint chromatograms of green-skinned and / or red-haired monk fruit, the similarity is calculated, and combined with chemical pattern recognition methods such as cluster analysis, principal component analysis, or orthogonal partial least squares discriminant analysis, the sample can be determined to belong to green-skinned, red-haired, or other provenances. This method overcomes the shortcomings of traditional trait identification, which is highly subjective and difficult to distinguish between samples with similar appearances, providing a reliable technical means for the objective identification of monk fruit provenance.

[0030] Application in the Quality Consistency Evaluation of Luo Han Guo (Monk Fruit): The HPLC fingerprint established in this invention can comprehensively reflect the overall distribution characteristics of various chemical components in Luo Han Guo, and can be used to evaluate the quality consistency of different batches of the same variety. By comparing the similarity of the chromatogram of the batch sample to be tested with the standard fingerprint of the same variety, if the similarity is higher than a preset threshold (e.g., 0.90), it indicates that the batch sample has good quality consistency; otherwise, it suggests that there may be changes in the source of raw materials, fluctuations in processing technology, or adulteration risks. This method is applicable to the batch-to-batch quality stability evaluation of Luo Han Guo medicinal materials, processed slices, and related products, providing a scientific basis for production enterprises and quality supervision departments.

[0031] Application in the identification of genuine and counterfeit Luo Han Guo (monk fruit): This invention can also be used to identify genuine and counterfeit Luo Han Guo. There may be instances in the market where other fruits (such as bitter fruit, fruits of other cucurbitaceae plants) are sold as Luo Han Guo. By using the HPLC fingerprint established in this invention, the chromatogram of the sample to be tested is compared with the standard fingerprint chromatogram of Luo Han Guo. If the number of chromatographic peaks, retention time, relative peak area, and other characteristics show significant differences from the standard fingerprint chromatogram, or if the similarity is lower than a preset threshold, it can be determined to be a counterfeit. This method is objective, accurate, and repeatable, effectively combating adulteration and counterfeiting, and ensuring the standardized and orderly market for Luo Han Guo medicinal materials.

[0032] Flexibility of application: It should be noted that the above applications can use the evaluation method of this invention alone, or the HPLC fingerprint spectrum established by this invention alone as a comparison standard; it can be applied to the quality control of monk fruit raw materials, and can also be extended to the traceability and quality evaluation of raw materials for downstream products such as traditional Chinese medicine, health food and food containing monk fruit. Attached Figure Description

[0033] Figure 1 HPLC overlay chromatograms of different sources of monk fruit; S1 (pentamethylfurfural); S2 (monoklavin); S6 (monoklavin V); S7 (semaphorin); S9 (monoklavin IV); S10 (monoklavin IIIA2); S11 (11-deoxymonoklavin V).

[0034] Figure 2 The results of cluster analysis of monk fruit from different sources.

[0035] Figure 3 The results of principal component analysis of monk fruit from different cultivars.

[0036] Figure 4 The results of orthogonal partial least squares discrimination analysis for different sources of Luo Han Guo.

[0037] Figure 5 The results are from 200 permutation tests.

[0038] Figure 6 A bar chart of VIP values.

[0039] Figure 7 Comparison of peak areas of different components between green-skinned fruit and red-skinned fruit; S1 (pentahydroxymethylfurfural); S2 (mogroflavin); S5 (unidentified); S6 (mogroflavin V); S10 (mogroflavin IIIA2).

[0040] Figure 8 This is an HPLC overlay of three batches of monk fruit from unknown sources. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. These embodiments are only for explaining the invention and are not intended to limit the scope of the invention.

[0042] Example 1: Preparation of test solution and reference solution Take dried monk fruit, crush it and pass it through a No. 4 sieve (sieve aperture 250±9.9 μm). Accurately weigh about 2 g of powder and place it in a stoppered conical flask. Accurately add 10 mL of methanol, seal tightly, weigh, and sonicate for 45 min (power 250 W, frequency 40 kHz). Cool to room temperature, weigh again, make up the weight loss with methanol, shake well, filter through a 0.22 μm microporous membrane, and collect the filtrate to obtain the test solution.

[0043] Accurately weigh appropriate amounts of the following reference standards: 5-hydroxymethylfurfural (batch number: PS021217, purity ≥98.0%), sarmandin I (batch number: PU0291-0025, purity ≥98.5%), mogroside IV (batch number: PU1164-0025, purity ≥98.0%), 11-deoxymogroside V (batch number: PU0337-0025, purity ≥98.0%), mogroside III-A2 (batch number: PU0330-0010, purity ≥98.0%), mogroside V (batch number: DSTDL002502, purity ≥98%), and mogroside flavonoid (batch number: DSTDL003001, purity ≥98%). Dissolve each in methanol and dilute to 10 mL to prepare a solution with a concentration of 0.5 mg / mL. -1 Prepare the reference solution for later use.

[0044] Example 2, Chromatographic Detection Conditions Chromatographic column: Octadecylsilane-bonded silica column (4.6 mm × 250 mm, 5 μm); Mobile phase: Pure water as mobile phase A, acetonitrile as mobile phase B, gradient elution according to the table below; Flow rate: 1.0 mL·min -1 Column temperature: 30 ℃; Detection wavelength: 203 nm; Injection volume: 10 μL.

[0045] Table 1 HPLC mobile phase elution conditions

[0046] Example 3: Validation of fingerprinting methodology 1. Specificity test The test solution (represented by Qingpiguo Q1), the mixed reference solution, and the blank solvent (methanol) were injected and analyzed according to the conditions of Example 2. The results showed that the blank solvent had no obvious chromatographic peaks at the corresponding retention times and did not interfere with the sample analysis; the retention times of the common peaks in the chromatogram of the test solution were consistent with those of the corresponding components in the chromatogram of the reference solution, indicating good specificity.

[0047] 2. Precision test Take the same batch of monk fruit samples (Q1), prepare the test solution according to the method in Example 1, and inject the sample six times consecutively under the conditions in Example 2, recording the chromatograms. Using monk fruit saponin V (S6) as the reference peak, calculate the relative retention time and relative peak area of ​​each common peak. The results are shown in Tables 2-5. The relative retention time RSD is <0.11%, and the relative peak area RSD is <5.00%, indicating good precision.

[0048] Table 2. Common peak retention times in precision tests

[0049] Table 3. Relative retention times and RSDs of common peaks in precision tests

[0050] Table 4. Peak area of ​​total peaks in precision test

[0051] Table 5. Common peak relative peak area and RSD in precision test

[0052] 3. Stability test Samples (Q1) from the same batch of monk fruit were used to prepare test solutions according to the method in Example 1. The solutions were injected and analyzed at 0, 2, 4, 8, 16, and 24 h. Using monk fruit saponin V as a reference peak, the relative retention time and relative peak area of ​​each common peak were calculated. The results are shown in Tables 6-9. The relative retention time RSD < 0.15% and the relative peak area RSD < 5.00%, indicating that the test solution exhibited good stability within 24 h.

[0053] Table 6. Peak retention times in stability tests

[0054] Table 7. Relative retention times and RSDs of common peaks in stability tests

[0055] Table 8. Peak area of ​​total peaks in stability experiment

[0056] Table 9. Common peak relative peak area and RSD in stability test

[0057] 4. Repeatability test Six sample solutions of the same batch of monk fruit (Q1) were prepared in parallel according to the method in Example 1, and injected for detection according to the conditions in Example 2. Using monk fruit saponin V as the reference peak, the relative retention time and relative peak area of ​​each common peak were calculated. The results are shown in Tables 10-13. The relative retention time RSD was <0.10%, and the relative peak area RSD was <5.00%, indicating good repeatability.

[0058] Table 10. Common peak retention times in repeatability tests

[0059] Table 11 Relative retention times and RSDs of common peaks in repeatability tests

[0060] Table 12. Peak area of ​​common peaks in repeatability tests

[0061] Table 13. Relative peak areas and RSD of common peaks in repeatability tests

[0062] Example 4: Establishment and Similarity Evaluation of HPLC Fingerprints of Luo Han Guo from Different Provenances Four batches of green-skinned fruit (Q1-Q4) and four batches of red-skinned fruit (H1-H4) were collected. Test solutions were prepared according to Example 1, and the samples were injected and analyzed under the conditions of Example 2. Chromatograms were recorded. All chromatograms were imported into the "Traditional Chinese Medicine Chromatographic Fingerprint Similarity Evaluation System (2012 version)" software. Using Q1 as the reference chromatogram and a time window width of 0.10 min, Mark peak matching was performed using the multi-point correction method, and the median method was used to generate the control fingerprint (R).

[0063] Figure 1 This is a superimposed HPLC chromatogram of different strains of *Siraitia grosvenorii* (chromatograms of 8 batches of samples). The main chromatographic peaks are marked in the figure: S1 (5-hydroxymethylfurfural), S2 (mogroside), S6 (mogroside V), S7 (symmenoside I), S9 (mogroside IV), S10 (mogroside IIIA2), and S11 (11-deoxymogroside V).

[0064] The similarity calculation results are shown in Table 14. The intra-group similarity of red shiitake fruit is 0.988-0.999, the intra-group similarity of green shiitake fruit is 0.908-0.987, while the similarity between red shiitake fruit and green shiitake fruit is 0.626-0.922, which is significantly lower than the intra-group similarity, indicating that there are significant differences in the overall profile of chemical components between the two sources of monk fruit.

[0065] Table 14. Similarity of HPLC fingerprint chromatograms of Luo Han Guo samples from different sources

[0066] Example 5: Cluster Analysis (HCA) Using the peak area of ​​11 common peaks in the fingerprint spectra of 8 batches of monk fruit samples as variables, Ward's minimum variance method (sum of squared deviations method) was used, and Euclidean distance was used as the similarity measure. SPSS 26.0 software was used to construct a hierarchical clustering dendrogram.

[0067] Figure 2 Dendrograms were generated for cluster analysis of different provenances of *Siraitia grosvenorii*. The results showed that when a suitable distance threshold was set, the eight batches of samples clearly clustered into two major categories: H1-H4 (red-skinned fruit) clustered into the first category, and Q1-Q4 (green-skinned fruit) clustered into the second category. Samples within each group were tightly clustered, and the boundaries between groups were clear. The clustering results were completely consistent with the actual provenance classification of the samples, indicating that different provenances of *Siraitia grosvenorii* possess significant categorical characteristics in terms of overall chemical composition content, and can be accurately distinguished through chemical pattern recognition.

[0068] Example 6: Principal Component Analysis (PCA) The peak areas of 11 common peaks from 8 batches of monk fruit samples were used as variables and imported into SIMCA 14.1 software for unsupervised principal component analysis. The model was automatically fitted, and score plots and loading plots were drawn.

[0069] Figure 3 Principal component analysis (PCA) score plots for different provenances of *Siraitia grosvenorii* are shown. The results indicate good model explanatory power, with samples exhibiting a clear grouping trend on the score plots: red-skinned fruit (H1-H4) is concentrated on one side, while green-skinned fruit (Q1-Q4) is concentrated on the other side, with no significant overlap between the two groups, demonstrating a significant separation trend. The PCA score plots visually reflect a stable difference in the overall distribution of chemical components between the two provenances, further validating the HPLC fingerprint and cluster analysis results, indicating that provenance is a crucial factor influencing the differences in chemical components of *Siraitia grosvenorii*.

[0070] Example 7: Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) and VIP Analysis Based on unsupervised PCA analysis, and grouping samples by provenance (green-skinned fruit group – red-haired fruit group), the peak areas of 11 common peaks from 8 batches of samples were used as variables and imported into SIMCA 14.1 software to establish a supervised OPLS-DA discriminant model. Unit variance scaling was employed, and the model type was OPLS-DA. Simultaneously, 200 permutation tests were performed to verify the model's reliability.

[0071] Figure 4 OPLS-DA score maps of different provenances of *Siraitia grosvenorii* were generated. The results showed that the red-skinned and green-skinned varieties were completely separated on the score map, with tight clustering within each group and clear distinction between groups, achieving accurate identification of samples from different provenances.

[0072] Figure 5 The graph shows the results of 200 permutation tests. All randomly arranged Q² values ​​on the left are lower than the original values ​​on the right, and the Q² regression line intersects the vertical axis in the negative region, indicating that the model is not overfitting and has good predictive ability.

[0073] Model parameters: R 2 X=0.899, R 2 Y=0.976, Q 2 =0.938, indicating that the model has good fitting and predictive ability.

[0074] Calculate the variable importance projection (VIP value) for each variable, and use VIP>1 as the criterion to screen the characteristic peaks that contribute significantly to the differentiation of seed sources.

[0075] Figure 6 This is a bar chart of VIP values. Five chromatographic peaks with VIP > 1 were selected, listed in descending order of VIP value as follows: S6 (monk fruit saponin V), S2 (monk fruit flavonoids), S1 (5-hydroxymethylfurfural), S10 (monk fruit saponin IIIA2), and S5 (unidentified).

[0076] Figure 7 A comparison of peak areas of differentially expressed components between green-skinned and red-skinned monk fruit is shown. The results indicate that the contents of 5-hydroxymethylfurfural and mogroside V in green-skinned monk fruit are significantly higher than those in red-skinned monk fruit (P < 0.01), while the contents of mogroside flavonoids and mogroside IIIA2 are significantly lower in green-skinned monk fruit (P < 0.01). These components can serve as potential chemical markers for distinguishing between green-skinned and red-skinned monk fruit, providing a scientific basis for quality identification and specific quality evaluation of monk fruit from different provenances.

[0077] Example 8: Identification of the provenance of unknown samples Three batches of dried monk fruit samples (numbered U1, U2, and U3) were taken for testing. Test solutions were prepared according to Example 1, and the samples were injected and analyzed under the conditions of Example 2. Chromatograms were recorded. Figure 8As shown. The chromatograms were imported into the chromatographic fingerprint similarity evaluation system for traditional Chinese medicine, and similarity calculations were performed with the control fingerprint chromatograms of green tangerine peel and red tangerine peel established in Example 4, respectively. The results are as follows: Table 15 Similarity of HPLC fingerprint chromatograms of the samples of monk fruit to be tested

[0078] Table 16 Similarity between the HPLC fingerprint chromatograms of the tested Luo Han Guo samples and the control chromatograms

[0079] After confirmation with the sample provider, U1 and U2 were confirmed to be red-skinned fruit, and U3 was confirmed to be green-skinned fruit. The identification results were consistent with the actual situation.

[0080] The above embodiments fully present the specific operation steps, methodological verification data, fingerprint spectrum establishment, multiple chemical pattern recognition analysis results, and practical application cases of the present invention, and clearly introduce the corresponding accompanying drawings, which meets the writing requirements of the "detailed implementation" section of the patent specification.

Claims

1. A method for evaluating the quality of monk fruit from different sources, comprising the following steps: S1. Obtain the test solution of the monk fruit sample to be tested; S2. The test solution is detected by high performance liquid chromatography to obtain the chromatogram of the monk fruit sample to be tested; S3. Compare the chromatogram with the standard fingerprint spectrum of monk fruit, and distinguish monk fruit from different sources through chemical pattern recognition analysis.

2. The evaluation method according to claim 1, characterized in that: In step S3, the chemical pattern recognition analysis includes one or more of cluster analysis, principal component analysis, and orthogonal partial least squares discriminant analysis.

3. The evaluation method according to claim 1 or 2, characterized in that: In step S2, the standard fingerprint spectrum of Luo Han Guo is obtained by detecting Luo Han Guo samples of known origin using the high performance liquid chromatography method described in step S2, and using a similarity evaluation system for chromatographic fingerprint spectrum of traditional Chinese medicine.

4. The evaluation method according to any one of claims 1-3, characterized in that: In step S1, the test solution is prepared by crushing dried monk fruit, adding organic solvent for extraction, and filtering. The organic solvent is methanol, and the extraction is ultrasonic extraction for 30-60 minutes.

5. The evaluation method according to any one of claims 1-4, characterized in that: In step S2, the conditions for the high-performance liquid chromatography include: The chromatographic column was packed with octadecylsilane-bonded silica gel. Mobile phase: Gradient elution was performed using pure water as mobile phase A and acetonitrile as mobile phase B. The flow rate is 0.8-1.2 mL / min; The column temperature is 25-35 ℃; The detection wavelength is 203 nm; The injection volume is 5-20 μL; From 0 to 10 min, mobile phase A was maintained at 88% and mobile phase B at 12%. In 10-15 minutes, mobile phase A decreased from 88% to 85%, while mobile phase B increased from 12% to 15%. Over 15-30 minutes, mobile phase A decreased from 85% to 80%, while mobile phase B increased from 15% to 20%. Over 30-40 minutes, mobile phase A decreased from 80% to 75%, while mobile phase B increased from 20% to 25%. Over 40-60 minutes, mobile phase A decreased from 75% to 72%, while mobile phase B increased from 25% to 28%. Over 60-65 minutes, mobile phase A decreased from 72% to 70%, while mobile phase B increased from 28% to 30%. Over 65-85 minutes, mobile phase A decreased from 70% to 15%, while mobile phase B increased from 30% to 85%. For 85-90 min, mobile phase A is maintained at 15% and mobile phase B is maintained at 85%.

6. The evaluation method according to any one of claims 1-5, characterized in that: When the chemical pattern recognition analysis is a cluster analysis, the area of ​​the common peak is used as a variable, and the Ward minimum variance method and Euclidean distance are used to construct a hierarchical cluster dendrogram. During principal component analysis, the chemical pattern recognition analysis uses the common peak area as a variable to import into SIMCA software for unsupervised pattern analysis, and plots score plots and loading plots are generated. When the chemical pattern recognition analysis is performed using orthogonal partial least squares discriminant analysis, the area of ​​the common peak is imported into the SIMCA software as a variable to establish an OPLS-DA model, the VIP value is calculated, and peaks with VIP > 1 are selected as chemical markers to distinguish different species.

7. The evaluation method according to any one of claims 1-6, characterized in that: The different seed sources include green-skinned fruit and red-haired fruit; Chemical markers that distinguish between green-skinned fruit and red-skinned fruit include one or more of 5-hydroxymethylfurfural, mogroflavin, mogroside V, and mogroside IIIA2.

8. The evaluation method according to any one of claims 1-7, characterized in that: The evaluation method also includes a similarity evaluation step for the sample of monk fruit to be tested: the chromatogram of the sample to be tested is compared with the standard fingerprint of monk fruit to be tested. If the similarity is higher than the intra-group threshold, it is determined to be the same source; if the similarity is lower than the inter-group threshold, it is determined to be different sources.

9. A method for establishing an HPLC fingerprint for distinguishing the quality of green-skinned and red-haired monk fruit, comprising the following steps: Samples of monk fruit, specifically green-skinned and red-haired varieties, were taken separately to prepare test solutions. The chromatogram is recorded by performing the high-performance liquid chromatography (HPLC) under the conditions described in any one of claims 1-8. The chromatograms were imported into the chromatographic fingerprint similarity evaluation system for traditional Chinese medicine. Using the typical sample chromatograms of green jujube or red jujube as reference chromatograms, the median method was used to generate the reference fingerprint chromatograms of green jujube or red jujube through multi-point correction and Mark peak matching.

10. The application of the evaluation method according to any one of claims 1-8 or the HPLC fingerprint spectrum established according to claim 9 in the identification of Luo Han Guo (Siraitia grosvenorii) germplasm, quality consistency evaluation or authenticity identification.

Images