Spectrum–effect relationship between ultra‐high‐performance liquid chromatography fingerprints and antioxidant activities of Lophatherum gracile Brongn

Abstract Lophatherum gracile Brongn. is a medicinal and edible plant resource as well as a natural additive in the functional food market. To better understand its characteristics and efficacy, a method combining chromatographic fingerprints and antioxidant activity was proposed. A total of 21 common peaks were confirmed from liquid chromatography fingerprints and were identified as 14 flavonoids and 7 phenolic acids by ultra‐high‐performance liquid chromatography (UHPLC) coupled with quadrupole Orbitrap mass spectrometry (Q‐Orbitrap/MS). Their antioxidant activities were evaluated by 1,1‐diphenyl‐2‐trinitrophenylhydrazine (DPPH), 2,2'‐diazide‐bis (3‐ethylbenzothiazoline‐6‐sulfonic acid) diammonium salt (ABTS), and ferric reducing antioxidant power (FRAP) assay. The results showed that all of the test samples had moderate to high antioxidant effects, with IC50 values ranging from 5.2 to 16.1 mg/ml and 1.2 to 2.8 mg/ml for DPPH and ABTS assays, and the FeSO4 concentrations of 1.84–4.20 mmol/L for the FRAP assay. The spectrum–effect relationship between UHPLC fingerprints and antioxidant activity was investigated through Pearson correlation analysis and Grey relational analysis (GRA) to identify the antioxidant constitutes in Lophatherum gracile Brongn. The results showed that 11 compounds were greatly associated with the antioxidant activity with a correlation degree >0.80, which can be used as the quality marker of Lophatherum gracile Brongn.


| INTRODUC TI ON
Modern pharmacological studies and clinical practices have demonstrated that free radicals produced by normal organ activities will cause chain reactions after exceeding a certain amount in the body, and a series of biological reactions will eventually lead to the occurrence of a variety of diseases, such as cancer, atherosclerosis, hypertension, cataracts, arthritis, and rheumatoid arthritis (Moreno-Ortega et al., 2020;Raja et al., 2020;Wu et al., 2020). Therefore, antioxidant research has become the forefront of medical research and a hot spot for disease prevention and treatment.
Antioxidants, including both chemically synthesized and naturally occurring antioxidants, have the ability to capture and neutralize free radicals, hence reducing the damage to the organism (Hu et al., 2020). Several chemical synthetic antioxidants like 2,6-di-tert-butyl-p-cresol (BHT) and butylated hydroxyanisole (BHA) have been discontinued in some countries due to nephrotoxicity, potential mutagenesis, and teratogenicity concerns (Moretti et al., 2020;Muğlu, 2020). At the same time, the annual growth rate of natural antioxidants' consumption was 15.6% (Marazza et al., 2012), which indicates that the discovery and development of nontoxic natural antioxidant foods and pharmaceutical products from natural sources has always been a common goal for biology, medicine, and food science research.
Lophatherum gracile Brongn. (L. gracile) is a perennial herb with dual medicinal and culinary qualities that is extensively used by traditional Chinese medical physicians to treat fever and inflammatory diseases. As a natural additive in the functional food industry, it is a significant ingredient of many herbal teas including "Wong Lo Kat herbal tea" (Ge et al., 2013;Ma et al., 2020). Flavonoids and phenolic acids are plant-based antioxidants, and many have been identified from L. gracile, including chlorogenic acid, isoorientin and orientin, isovitexin, cynaroside, and luteolin (Guo et al., 2019), which have antioxidant, antidiabetic, anticardiovascular, anti-inflammatory, and other properties (Gong et al., 2015;He et al., 2016). For example, they can be used to make facial masks and skin creams, as well as for processing and pickling foods. Besides, antioxidants of L. gracile leaves have been approved as a food antioxidant by the Chinese Food Additive Standardization Committee, and have also been applied in the development of food, health care products, medicine, and cosmetics in North America, Asia, and other regions (Kuei-HungLai et al., 2020).
The leaf of L. gracile has been officially listed in the Chinese Pharmacopoeia as a crude drug, however, there is no quality control stated except the character identification of the herb. Because few investigations have focused on the quality control of L. gracile, it is vital to investigate the quality safety and efficacy of L. gracile.
The fingerprint technique has become a routine technique for the quality control of traditional Chinese herbal medicine (TCHM), but the similarity between chromatograms does not accurately reflect their closeness in drug efficacy (Dabić et al., 2020;Xu et al., 2020;Zhang, Wang, et al., 2019;Zhang, Ding, et al., 2019). Therefore, a spectrum-effect relationship approach has recently been applied to improve the quality assessment of TCHM. A method based on the spectrum-effect relationship was developed for discovering quality markers of the herb Meconopsis integrifolia . Ding and his colleagues established the HPLC fingerprint of different batches of Panax notoginseng to determine the promoting blood circulation pharmacodynamic indices of Panax notoginseng, and to explore the spectrum-effect relationship between the chemical components and the efficacy of promoting blood circulation (Ding et al., 2021). Liu et al. performed an activity-calibrated chemical standardization approach for quality assessment of Salvia miltiorrhiza Bge (Liu et al., 2017). Those studies indicate that using the spectrum-effect relationship approach to link the "quality" with the potential clinical effects of TCHM is a worthwhile tool.
In this study, the chromatographic fingerprint was established through UHPLC analysis to completely evaluate the quality consistency of 15 batches of L. gracile, and the common peaks were identified via UHPLC-Q-Orbitrap/MS analysis. Three methods, including 1,1-diphenyl-2-trinitrophenylhydrazine (DPPH) method, 2,2'-diazide-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammo- This study provides a certain reference basis for the clinical rational use and quality control of L. gracile.

| Reagents and materials
Chromatographic-grade acetonitrile and methanol were obtained from Fisher Scientific. Formic acid in MS grade was provided by Sigma-Aldrich. Distilled water was produced via a Millipore water purification system (Millipore). All other chemicals were of analytical grade. For standards, luteolin (111,605), and chlorogenic acid (110,018) were purchased from NIFDC (National Institutes for Food and Drug Control, Changchun, Jilin, China).  Table S1.

Novelty impact statement
The contribution of specific chemical compounds to antioxidant activity was identified by the spectrum-effect relationship-based approach, which provides us a deeper insight into the joint effect of multiple ingredients of Lophatherum gracile Brongn. This study is of great significance for elucidating the pharmacodynamic material basis, screening key quality markers related to medicinal effect, and ensuring the safety and rational application of medicinal herbs.

| Sample solution preparation
Fifteen batches of L. gracile sample solution were prepared. An aliquot of 1.0 g fine powder of each sample was properly weighed and ultrasonically extracted with 20 ml of methanol/water (70/30, v/v) for 20 min at room temperature. The mixture was filtered through a 0.22μm membrane filter. The filtrate was transferred into a sample vial for UHPLC and UHPLC-Q-Orbitrap/MS analysis. Besides, each powder was accurately weighed and extracted using the same method for the DPPH, ABTS, and FRAP assay.

| UHPLC analysis
The UHPLC analysis was performed on a Waters ultra-high-

| UHPLC-Q-Orbitrap/MS analysis
Chromatographic separation was conducted using the same method as UHPLC analysis. And the elute was introduced into a Q-Orbitrap-MS (Thermo Fisher Scientific) equipped with an electrospray ionization source under the negative ion mode. The parameters of the ion source were set to 40 Arb for sheath gas flow, 10 Arb for auxiliary (aux) gas flow, and 1 Arb for sweep gas flow.
The S-Lens radio frequency (RF) was 55%. The capillary voltage was set to −3.5 kV with a capillary temperature of 350°C. Full MS data were acquired at the centroid mode from m/z 150 to 2000 Da, using the resolution of 70,000, with the automatic gain control (AGC) target of 1 × 10 6 and maximum injection time (IT) of 100 ms.
The tandem mass spectrum was obtained in Full-MS/ddMS2 mode using the following settings: 17,000 for resolution, 1 × 10 5 for automatic gain control (AGC) of the target, 50 ms for maximum IT, 5 for Loop count, 5 for TopN, 4.0 m/z for Isolation window, and 25, 35, 55 for stepped normalized collision energy (NCE).

| Antioxidant activities
2.5.1 | DPPH assay All L. gracile samples were tested for DPPH radical scavenging activity according to the previous report (Guo et al., 2019) with minor modifications. In brief, an aliquot (100 μl) of each sample was added to 100 μl of DPPH solution to initiate the reaction, and 70% methanol was used as the blank solution. After 30 min of incubation in the dark at room temperature, the absorbance was measured at 517 nm. Each measurement was performed in triplicate. The DPPH quenched rate was calculated according to the following equation: Where A blank and A sample represent the absorbances of the blank sample and test sample, respectively.

| ABTS assay
The ABTS radical scavenging activity of each sample was determined using Mohammad Noshad's technique (Noshad et al., 2021) with slight modifications. The same volumes of 7.4 mM ABTS solution and 2.60 mM K 2 S 2 O 8 were mixed and kept at room temperature for 12 h under dark conditions to prepare the stock solution.
Before use, the ABTS working solution was obtained by adding 70% methanol to the stock solution until the absorbance reached 0.70 at 734 nm. Thereafter, 50 μl of each sample solution was mixed with the 800 μl of ABTS working solution and was kept at ambient temperature for 10 min. After that, the absorbance at 734 nm against the blank sample (70% methanol) was measured and recorded. All measurements were done in triplicate. The ABTS activity of each sample was calculated according to the following formula:

| FRAP assay
The ability to reduce ferric ions was measured based on previous articles (Krivokapic et al., 2021;Tadic et al., 2021) with minor revisions. The stock solutions included 300 mM acetate buffer, pH 3.6, 10 mM 2,4,6-Tris(2-pyridyl)-s-triazine (TPTZ) solution in 40 mM HCl and 20 mM FeCl 3 .6H 2 O solution. The working solution was prepared by mixing 25 ml of acetate buffer, 2.5 ml of TPTZ solution, and 2.5 ml of FeCl 3 .6H 2 O solution and then warming at 37°C before use.
Each sample solution (50 μl) was used to react with 200 μl of FRAP reagent at room temperature for 10 min and the absorbance of the reaction mixture was measured at 593 nm. The 100 mmol/L FeSO 4 was used as the standard solution for calibration. All tests were carried out in triplicate. The total antioxidant capacity of reducing ferric ions of each sample was expressed by the amount of substance produced which represents that Fe 3+ -TPTZ is reduced to Fe 2+ -TPTZ.

| Data processing and multivariate statistical analysis
The raw data from UHPLC were converted into.AIA format file and then imported into the Similarity Evaluation System for Chromatographic for peak alignment and similarity analysis. The reference fingerprint was generated by the system using the Median method from the general comparison of the chromatograms of 15 L. gracile extracts and the similarity values between the reference fingerprint and the chromatograms of different sample extracts were calculated by the cosine of vectorial angle method using this software.
The Pearson correlation analysis and GRA were performed using the Hiplot web tool and SPSSAU (Version 21.0, online application software), respectively, to find the spectrum-effect relationship between common peaks and antioxidant activities and further to characterize the quality of L. gracile. proposed fragmentation pathway is shown in Figure S1. Collectively, this ion at RT 7.71 min_m/z 447.0945 was assigned to isoorientin.

| DPPH free radical scavenging activity
The studies using varying concentrations (5-20 mg/ml) of L. gracile extracts were first conducted, and the result is shown in Figure S2a.
The antioxidant activity of L. gracile was evaluated using the halfmaximal inhibitory concentration (IC 50 ) values. The better the antioxidant activity, the lower the IC50 levels. In the case of samples from Dazhou city, our results exhibited an outstanding DPPH scavenging activity with IC 50 ranging from 5.2 to 5.4 mg/ml, which is lower than those of the samples from Anhui province (7.9-8.8 mg/ ml) and Chengdu city (12.7-16.1 mg/ml). The scavenging activity of extracts from L. gracile samples against the stable radical DPPH was evaluated at a concentration of 12.5 mg/ml and the result is illustrated in Figure 3a. The samples collected from Dazhou city (S6-S10) showed a high scavenging activity (>75%). The scavenging activity of samples collected from Anhui province exhibited the scavenging activity of 40%-50%, which was higher than the sample harvest from Chengdu city ranging from 25% to 30%.

| ABTS free radical scavenging ability
The IC 50 value of these samples was investigated using varying concentrations (1-10 mg/ml). As observed with the ABTS assay, lower IC 50 values (1.2-1.5 mg/ml) were obtained from the sample S6-S10 that was collected from Dazhou city ( Figure S2b). The ABTS radical scavenging activity is based on the reduction of ABTS + to ABTS through a hydrogen atom transfer mechanism, in the presence of antioxidant agents (Zhao et al., 2017). The scavenging activity of various extracts (10 mg/ml) of L. gracile is presented in Figure 3b. Better scavenging activity is obtained from the samples collected from Dazhou city (>70%), which is in agreement with the results obtained by the DPPH assay. A similar scavenging activity was observed for the sample harvest from Anhui province and Chengdu city, ranging from 35% to 70%.

| Ferric ion reducing antioxidant power (FRAP) assay
The FRAP assay measures the substance's ability to reduce Fe 3+ to Fe 2+ through electron transfer. According to the above procedure,

| Studies on the spectrum-effect relationship analysis
In the present study, the spectrum-effect relationships between chromatographic fingerprint peaks and antioxidant activity of L. gracile samples were investigated by Pearson correlation analysis and GRA.

| Grey relational analysis
The GRA was also used to investigate the relationship between the antioxidant effects and 21 common peaks found in L. gracile samples, and further to identify the primary active compounds responsible for these effects. After data standardization, the correlation coefficient was calculated. The correlation coefficient was utilized to determine the association between chemical components and antioxidant capacity, the greater the correlation coefficient, the stronger the association between the chemical composition and the antioxidant capacity. As listed in Table 2, peaks with a correlation degree >0.80 were ranked in the following sequence: P13>P6>P10>P 16>P2>P1>P4>P8>P11>P3>P14>P19>P20 for DPPH assay, P13>
Since ancient times, L. gracile has been utilized as an antioxidant to prevent food deterioration. The main active ingredients have been identified as flavonoids, polysaccharides, and phenolic acids.
The presence of hydroxyl groups in the structure of the phenolic compounds of L. gracile substances is often linked to their antioxidant activity (Gong et al., 2015).
In this study, the antioxidant effects of L. gracile were investi-  Discovering bioactive molecules from herbal medicine is difficult, due to its property of multiple chemical components and multiple targets. The most prevalent methods concentrate on compound separation and single component activity, which are time-consuming and fail to reveal the complex roles of numerous components in herbal medicines (Liu et al., 2017). The spectrum-effect relationship analysis is an innovative and dependable technology that integrates chromatographic fingerprints with pharmacological effects by chemometrics, which could be used to investigate the correlations between bioactive components and efficacy, as well as to identify the major bioactive elements in herbal medicines Zhang, Ding, et al., 2019).
The antioxidant properties of L. gracile, according to the data, were dominated by several compounds rather than a single sub- The characterization and quantification of all active chemical ingredients of TCHM face substantial hurdles due to the variety and complexity of the chemical composition of TCHM. At present, the application of chemical fingerprints to evaluate the quality of TCHM has been widely accepted and used in many countries. However, the assessment of pharmacodynamic effects will be overlooked when only analyzing the type and content of its chemical components to evaluate the quality of TCHM. The investigation of spectrum-effect relationships, which combine the chemical composition of TCHM with its pharmacodynamic activity, allows for a more thorough assessment of drug quality.

| CON CLUS ION
The UHPLC fingerprints of various L. gracile were established in this investigation, and 21 common peaks were chosen. DPPH, ABTS, and FRAP tests were used to assess the antioxidant properties of several L. gracile samples. The antioxidant chemicals were then discovered TA B L E 2 Grey relational coefficient (GRA) between fingerprints and antioxidant activities of L. gracile by looking at the spectrum-effect relationship between UHPLC fingerprints and their antioxidant activities. The Pearson correlation analysis and GRA results showed that 11 compounds that were identified as 6 flavonoids and 5 phenolic acids contributed most to the antioxidant effects of L. gracile and were selected as potential quality markers. The established spectrum-effect relationship-based approach is of great significance for elucidating the pharmacodynamic material basis, screening core quality markers related to medicinal effect, and ensuring the safety and rational application of traditional herbal medicines.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the authors.