Discovery of Potential Anti-Microbial Molecules and Spectrum Correlation Effect of Ardisia crenata Sims via High-Performance Liquid Chromatography Fingerprints and Molecular Docking

Ardisia crenata Sims, an important ethnic medicine, is recorded in the Chinese Pharmacopoeia for treating laryngeal diseases and upper respiratory tract infections. This study aimed to evaluate the antimicrobial effect of extracts and potential antimicrobial compounds of A. crenata Sims. It was found that the roots of A. crenata Sims have a potential inhibitory effect on Candida albicans and Aspergillus flavus, with MICs of 1.56 mg/mL and 0.39 mg/mL, and the leaves of A. crenata Sims have a potential inhibitory effect on Pseudomonas aeruginosa and Staphylococcus aureus, with MICs of 3.12 mg/mL and 6.77 mg/mL, respectively. Meanwhile, five compounds including one catechin and four bergenins were obtained from roots. These components were identified on the fingerprint spectrum, representing chromatographic peaks 16, 21, 22, 23, and 25, respectively. Among these, 11-β-d-glucopyranosyl-bergenin and (−)-gallocatechin showed potential inhibition for Staphylococcus aureus and Pseudomonas aeruginosa with MIC of 0.26 and 0.33 mg/mL, respectively. The roots, stems, and leaves of A. crenata Sims are very similar in chemical composition, with large differences in content. Principal component analysis (PCA) and Hierarchical cluster analysis (HCA) showed that 16 batches of A. crenata Sims could be divided into four main production areas: Guizhou, Jiangsu, Guangxi, and Jiangxi. Furthermore, molecular docking results showed that 11-β-d-glucopyranosyl-bergenin had a better affinity for Casein lytic proteinase P (ClpP), and (−)-gallocatechin possessed a strong affinity for LasA hydrolysis protease and LasB elastase. These findings suggest catechin and bergenins from A. crenata Sims can be used as antimicrobial activity molecules.


Introduction
As the most common human bacterial pathogen, Staphylococcus aureus often asymptotically colonizes the nasal mucosa of humans, causing superficial infections of the skin and mucosa, and even life-threatening systemic infections [1].People with damaged skin and mucosal barrier or impaired immune systems are particularly vulnerable to S. aureus infection, which can cause a variety of diseases, including pneumonia, sepsis, serious skin infection, and respiratory system infection [2,3].Among the various virulence factors of S. aureus, Casein Hydrolase (ClpP) is the key virulence factor that determines the pathogenicity of S. aureus and plays a crucial role in the pathogenicity of bacteria.Therefore, ClpP has been identified as a new candidate antibacterial target for screening and discovering inhibitors of important virulence factors of S. aureus [4,5].Pseudomonas aeruginosa can colonize various surfaces and tissues with strong adaptability, invasiveness, and pathogenicity and cause various acute and chronic infections such as burn wounds, urinary tract infections (UTI), and lung infections [6].This can be attributed not only to its highly endogenous nature and acquired resistance but also to various virulence factors [7].P. aeruginosa can secrete different kinds of extracellular proteases, such as LasA protease, LasB elastase, alkaline protease, and protease IV.Among them, AprA and LasB can alter the mucosal cilia clearance rate, degrade lung tissue, and disrupt the host immune system, thereby strongly promoting lung diseases [8].P. aeruginosa is also the main cause of chronic lung infection in patients with cystic fibrosis.The emergence and spread of widely resistant or multidrug-resistant P. aeruginosa isolates pose significant risks to human health [9].
Antibiotics are one of the greatest inventions of the 20th century and are widely used in the treatment of infectious diseases.However, approximately 50% of antibiotics are abused and misused globally every year, leading to strong antibiotic resistance in bacteria [10,11].Traditional Chinese medicine (TCM) has been used clinically for thousands of years, with characteristics such as less toxicity, fewer side effects, and multiple targets of action.It has played a crucial role in human efforts to overcome major epidemics.The effective active ingredients of Chinese herbal medicine mainly contain polysaccharides, essential oils, and phenolic compounds [12][13][14].These components often have certain antibacterial effects and are not prone to developing compound resistance, and they even reverse the compound resistance of bacteria.They have received increasing attention in clinical and scientific research [15].Ardisia crenata Sims, a plant of the genus Ardisia, also called zhu sha gen is mainly distributed in Guizhou, China [16].It is also used as an important Miao medicine called ba zhua jin long with antimicrobial, anti-viral, anti-inflammatory, and anti-tumor effects [17,18].In addition, A. crenata Sims has a significant antibacterial effect on type A, type B hemolytic streptococcus, and Staphylococcus aureus [19].Clinically, as the main drug of Kaihoujian spray (child type), it is mainly used to treat respiratory tract infections, tonsillitis, rheumatic bone pain, and other diseases without general toxicity or adverse effects [20,21].At present, studies have found that A. crenata Sims mainly contains coumarins, triterpenoid saponins, flavonoids, and other chemical components [22,23].Especially, bergenin as one kind of main coumarin, has inhibitory effects on the growth of microbes [24].
Modern analytical techniques have played an important role in the quality identification of TCM, including the detection of hydrazine in real water and soil samples from the growing areas of TCM [25].In particular, integrated metagenomics and metatranscriptomics sequencing can also be used to examine the abundance of microbial consortiums and their metabolites [26][27][28].Macromolecular phase separation was also used to deliver bioactive compounds [29].The chromatography-mass spectrometry technology can detect the active components of TCM and food, as well as analyze and confirm the structure of unknown active substances [30].In addition, research has shown that the anti-mold secondary amine bond of soy protein can effectively and environmentally improve the anti-mold properties of its adhesive [31].
In this paper, we established the fingerprints of A. crenata Sims roots, stems, and leaves, and assigned their common peaks and characteristic peaks.Meanwhile, five compounds were isolated and purified from A. crenata Sims.The contents of these compounds were also determined to illustrate the differences in different medicinal parts of A. crenata Sims.Furthermore, we also compare the antimicrobial activity of different batches of A. crenata Sims.Moreover, these compounds were evaluated by molecular docking analysis, obtaining good antimicrobial activity to elucidate the possible mechanism.

Anti-Microbial Activity Evaluation of A. crenata Sims
The different parts of A. crenata Sims were used to evaluate antimicrobial activities against two kinds of fungi and six kinds of bacteria.As shown in Table 1, the extracts of root, stem, and leaf exhibited inhibitory activity on these test strains with an inhibition zone diameter (IZD) of 6.08~20.84mm.The results showed that the roots of A. crenata Sims had good activity against Candida albicans and Aspergillus flavus.Leaves had good antibacterial activity against Pseudomonas aeruginosa and Staphylococcus aureus.In addition, the results indicated that the Minimum inhibitory concentration (MIC) values of root to Candida albicans and Aspergillus flavus were 1.56 mg/mL and 0.39 mg/mL, respectively.The MIC values of leaves against Pseudomonas aeruginosa and Staphylococcus aureus were 3.12 mg/mL and 6.77 mg/mL, respectively (Table 2).

Analysis of HPLC Fingerprint
The HPLC fingerprint of 16 batches and the reference fingerprint from A. crenata Sims are presented in Figures 1 and 2. Ten common peaks are shown as peaks

Analysis of HPLC Fingerprint
The HPLC fingerprint of 16 batches and the reference fingerprint from A. crenata Sims are presented in Figures 1 and 2. Ten common peaks are shown as peaks

Analysis of HPLC Fingerprint
The HPLC fingerprint of 16 batches and the reference fingerprint from A. crenata Sims are presented in Figures 1 and 2. Ten common peaks are shown as peaks

Analysis of HPLC Fingerprint
The HPLC fingerprint of 16 batches and the reference fingerprint from A. crenata Sims are presented in Figures 1 and 2. Ten common peaks are shown as peaks

Analysis of HPLC Fingerprint
The HPLC fingerprint of 16 batches and the reference fingerprint from A. crenata Sims are presented in Figures 1 and 2    The similarity analysis was conducted with A. crenata Sims in the S1 production area as a reference (Table 4).As a result, the root similarity was more than 0.97 in different producing areas, while, many differences in the fingerprints of the leaves and stems of A. crenata Sims from different regions.The aboveground part of A. crenata Sims was different due to being affected by the environment.As shown in Figure 3, HCA analysis found that the roots of A. crenata Sims from 16 production areas have high similarity, and the differences in chemical composition of stems and leaves due to different climatic environments in different production areas.The results of PCA analysis showed that the roots, stems, and leaves of A. crenata Sims in the four production areas could be divided into four categories, representing the major production areas, including Guizhou, Jiangsu, Guangxi, and Jiangxi (Figure 4).

PCA and HCA Analysis of 16 Batches of A. crenata Sims
As shown in Figure 3, HCA analysis found that the roots of A. crenata Sims from 16 production areas have high similarity, and the differences in chemical composition of stems and leaves due to different climatic environments in different production areas.The results of PCA analysis showed that the roots, stems, and leaves of A. crenata Sims in the four production areas could be divided into four categories, representing the major production areas, including Guizhou, Jiangsu, Guangxi, and Jiangxi (Figure 4).

Effect of Compounds 1-5 on Anti-Microbial Activities
The isolated and identified compounds were used to evaluate the antimicrobial activities against the tested strains.As shown in Table 5, compounds 1-5 exhibited inhibitory activity on these microorganisms with inhibition zone diameters (IZD) of 6.11~9.28mm.The results showed that 11-β-D-glucopyranosyl-bergenin, 11-α-D-galactopyrnside-bergenin, and 11-O-galloybergenin had good activity against Staphylococcus aureus.(−)-gallocatechin and bergenin had good antibacterial activity against Pseudomonas aeruginosa.In addition, the results indicated that the MIC values of these compounds against Staphylococcus aureus and Pseudomonas aeruginosa ranged from 0.26 to 0.39 mg/mL, respectively.

Effect of Compounds 1-5 on Anti-Microbial Activities
The isolated and identified compounds were used to evaluate the antimicrobial activities against the tested strains.As shown in Table 5, compounds 1-5 exhibited inhibitory activity on these microorganisms with inhibition zone diameters (IZD) of 6.11~9.28mm.The results showed that 11-β-D-glucopyranosyl-bergenin, 11-α-D-galactopyrnside-bergenin, and 11-O-galloybergenin had good activity against Staphylococcus aureus.(−)-gallocatechin and bergenin had good antibacterial activity against Pseudomonas aeruginosa.In addition, the results indicated that the MIC values of these compounds against Staphylococcus aureus and Pseudomonas aeruginosa ranged from 0.26 to 0.39 mg/mL, respectively.

Spectrum-Effect Relationship
The spectrum-effect relationship between chromatographic peaks and anti-microbial activity was established by GRA and PLSR models.As a result, GRA analysis showed that the correlation data of all the common peaks were greater than 0.7.This indicated that the antimicrobial activities of A. crenata Sims were caused by the compounds represented by all these peaks (Table 6).
VIP value > 1 was used as the standard to screen the key components of the antimicrobial effect of roots from A. crenata Sims.As a result, PLSR analysis showed that peaks of 19, 21, 26, 27, and 29 were the main components of A. crenata Sims against Candida albicans (Figures 5a and 6a).Peaks of 19, 25, 26, 27, and 29 were the main components of roots from A. crenata Sims against Aspergillus flavus (Figures 5b and 6b).Peaks 16 and 22 were the main components of leaves from A. crenata Sims against Pseudomonas aeruginosa (Figures 5c and 6c).Peaks 19, 23, and 8 were the main components of leaves from A. crenata Sims against Staphylococcus aureus (Figures 5d and 6d).These results indicated that the antimicrobial effect of A. crenata Sims was jointly influenced by multiple components.

Quantitative Analysis of Anti-Microbial Ingredients in A. crenata Sims
As shown in Table 7, the HPLC analysis showed that the contents of bergenin, and (−)-gallocatechin were much higher than 5.5 mg/g in the roots.Moreover, these two components in roots were much higher quantities than in the stems and leaves.In

Quantitative Analysis of Anti-Microbial Ingredients in A. crenata Sims
As shown in Table 7, the HPLC analysis showed that the contents of bergenin, and (−)-gallocatechin were much higher than 5.5 mg/g in the roots.Moreover, these two components in roots were much higher quantities than in the stems and leaves.In addition, 3.84 mg/g of 11-O-galloybergenin was found in the leaves.Meanwhile, 0.38 mg/g of 11-β-D-glucopyranosyl-bergenin was found in the stems, which was much higher than the content in the roots and leaves.The chromatograms of standards solution of these compounds were showed in Figures S1-S5 (Supplementary Materials).

Sample Preparation
The roots, stems, and leaves of A. crenata Sims were dried and crushed into powder.Sample powders (2.0 g) were accurately weighed and extracted with 50 mL methanol in a stoppered Erlenmeyer flask with an ultrasonic multi-frequency cleaning machine (frequency 40 kHz) for 40 min.The obtained extracts were filtered, concentrated under reduced pressure and vacuum dried.Then, the extracts were mixed with 10% DMSO to configure with a concentration of 100 mg/mL as the test solution for the anti-bacterial experiment.The 11-α-D-galactopyranoside-bergenin, 11-β-D-glucopyranosyl-bergenin, bergenin, 11-O-galloybergenin, and (−)-gallocatechin were dissolved in 30% MeOH-H 2 O to for content determination.

Validation of Methodology
The precision test was evaluated by six consecutive injections of the same sample (S3) solution, and the repeatability was evaluated by repeating six times with samples (S3) from the same place of origin.The stability tests were analyzed within 0, 2, 4, 6, 8, 10, 12, and 24 h, respectively.

Analysis of HPLC Fingerprint
The similarity of the Chinese medicine chromatographic fingerprints was analyzed and evaluated using the 2012A version system under the optimized HPLC conditions.Hierarchical cluster analysis (HCA) and multivariate principal component analysis (PCA) were used to divide the samples into different groups based on the similarity of their measured properties [40].

Anti-Bacterial Activity Evaluation
The antimicrobial doses were set according to the pharmacological dosage of A. crenata Sims in Kaihoujian spray (child type) and the antimicrobial concentration gradient selected in the pre-experiment.The blank drug-sensitive paper (6 mm × 1 mm) was soaked in the compounds solution (0.4 mg/mL) and extracts solution of roots, stems, and leaves (100 mg/mL) of A. crenata Sims and the sterile 10% DMSO solution was used as the blank control.The ceftazidime and nystatin were set as positive drugs with a concentration of 1.0 mg/mL.Candida albicans and Aspergillus flavus were placed in an incubator at 28 • C while the other six bacteria were cultured in an incubator at 37 • C for 24 h to observe the growth of the bacteria and fungi.The diameter of the inhibition zone was measured three times, and the data were recorded.
The minimum inhibitory concentration (MIC) of extracts and compounds from A. crenata Sims was performed according to the two-fold serial dilution method.The dilution concentrations of roots, stems and leaves extracts for S. aureus were 55.50 to 0.22, 66.90 to 0.26, and 54.20 to 0.21 mg/mL, respectively.The dilution concentrations of roots, stems and leaves extracts for B. subtilis were 55.00 to 0.21, 66.30 to 0.26, and 60.20 to 0.24 mg/mL, respectively.The dilution concentrations of root, stem, and leaf extracts for E. faecalis were 66.60 to 0.26, 71.40 to 0.28, and 70.60 to 0.27 mg/mL, respectively.The dilution concentrations of root, stem, and leaf extracts for E. coli were 66.60 to 0.26, 71.40 to 0.28, and 70.60 to 0.27 mg/mL, respectively.The dilution concentrations of root, stem, and leaf extracts for P. aeruginosa were 54.70 to 0.21, 60.40 to 0.24, and 100.00 to 0.19 mg/mL, respectively.The dilution concentrations of root, stem, and leaf extracts for P. vulgaris were 67.80 to 0.26, 67.50 to 0.26, and 55.50 to 0.22 mg/mL, respectively.The dilution concentrations of root, stem, and leaf extracts for C. albicans were 100 to 0.19, 98 to 0.19, and 100 to 0.19 mg/mL, respectively.The dilution concentrations of root, stem, and leaf extracts for A. flavus were 100 to 0.19, 88.6 to 0.34, and 98.8 to 0.38 mg/mL, respectively.The dilution concentration of 11-β-D-glucopyranosyl-bergenin for S. aureus was 1.04 to 0.03 mg/mL and (−)-gallocatechin for P. aeruginosa was 1.32 to 0.04 mg/mL.The dilution concentration of ceftazidime for bacteria was 0.64 to 0.02 mg/mL and nystatin for fungi was 0.84 to 0.025 mg/mL.Bacteria were cultured at 37 • C for 24 h and fungi were kept at 28 • C for 48 h, respectively.Then, 10 µL of 2,3,5-triphenyl tetrazolium chloride (TTC) was added to the plates and incubated.The MIC was determined as the highest dilution of extracts and compounds exhibiting no growth visibility of bacteria and fungi.All the tests were performed in replicates three times.
3.9.Spectrum-Effect Relationship 3.9.1.Gray Relational Analysis (GRA) GRA analysis could be used to determine the contribution of fingerprint-shared peaks to anti-microbial activity.Sixteen batches of A. crenata Sims were used as reference sequences to determine the inhibition zone diameter of the four sensitive strains.The common peak area data in the HPLC fingerprint of the corresponding batches of A. crenata Sims were taken as the comparison sequence.The gray correlation analysis method was used to establish a spectral efficacy correlation mathematical statistical model and the correlation degree of each common peak to the efficacy indicators was calculated, and the resolution coefficient was ξ = 0.5 [41].

Partial Least Squares Regression (PLSR)
The peak area of each common peak in the fingerprint of A. crenata Sims was set as the independent variable (X) and the antibacterial activity of A. crenata Sims against strains as the dependent variable (Y), using these, the regression models were built sequentially.Then SIMCA-P 14 was used for PLSR analysis and the regression coefficient of X to Y and the variable importance projection (VIP) value were calculated [42].

Molecular Docking Analysis
Molecular docking is currently one of the important means for studying the interaction between small molecules and proteins in traditional Chinese medicine.It can be used to identify targets with a high affinity for speculating the mechanism of traditional Chinese medicine in treating diseases.In this paper, the key target proteins related to antimicrobial activity were used for docking with ingredients of A. crenata Sims.The molecular docking analyses of anti-microbial compounds to target proteins were conducted according to the Ligand docking module of Schrödinger Suite 2021-1 (Schrödinger, LLC, New York, NY, USA).The Crystal structures of ClpP PR (PDB ID: 3V5e), LasA PR (PDB ID: 3IT7), LasB PR (PDB ID: 3DBK), DNA ligase (2XCQ), DNA gyrase (3JSN) and MurF ligase (4CVL) were chosen for the docking analysis [43].

Conclusions
In summary, we innovatively established fingerprints, for the first time, of the roots, stems, and leaves of A. crenata Sims from different origins, and screened the main regions as sources of medicinal materials.In addition, we also speculated on the antimicrobial active ingredients in the chemical composition of A. crenata Sims via combining antimicrobial experiments with fingerprint analysis.Furthermore, we isolated and identified five phenolic compounds and quantified them from A. crenata Sims.Among these, 11-β-Dglucopyranosyl-bergenin and (−)-gallocatechin showed better inhibition for Staphylococcus aureus and Pseudomonas aeruginosa, respectively.Moreover, 11-β-D-glucopyranosyl-bergenin had a much better affinity to ClpP PR and (−)-gallocatechin showed the best affinity to LasA PR and LasB PR.These results confirmed that unvalidated molecular docking analysis may suggest a possible mechanism of antimicrobial activity.And the phenolic components could be used as anti-microbial activity molecules for developing new anti-microbial agents.In our next work, we will conduct molecular biology experiments to verify the antimicrobial target proteins screened by molecular docking.

C 21 H 20 O 13 Figure 1 .Figure 1 .
Figure 1.The HPLC reference fingerprint of the roots (a), stems (b), and leaves (c) from A. crenata Sims. Figure 1.The HPLC reference fingerprint of the roots (a), stems (b), and leaves (c) from A. crenata Sims.

Figure 2 .
Figure 2. The HPLC fingerprint of the roots (a), stems (b), and leaves (c) from A. crenata Sims.Figure 2. The HPLC fingerprint of the roots (a), stems (b), and leaves (c) from A. crenata Sims.

Figure 2 .
Figure 2. The HPLC fingerprint of the roots (a), stems (b), and leaves (c) from A. crenata Sims.Figure 2. The HPLC fingerprint of the roots (a), stems (b), and leaves (c) from A. crenata Sims.

Figure 13 .Table 9 .
Figure 13.The molecular docking results of 11-O-galloybergenin (A) and bergenin (B) on SQS PRs.Table 9. Molecular docking score of the active compounds on key target proteins of fungi.Compounds SQS PRs (7WG1)

Table 2 .
MIC of different parts from A. crenata Sims on different strains (x ± s, n = 3, mg/mL).

Table 3 .
Structures of compounds 1-5 from the roots of A. crenata Sims.

Table 3 .
Structures of compounds 1-5 from the roots of A. crenata Sims.

Table 3 .
Structures of compounds 1-5 from the roots of A. crenata Sims.

Table 3 .
Structures of compounds 1-5 from the roots of A. crenata Sims.

Table 3 .
Structures of compounds 1-5 from the roots of A. crenata Sims.

Table 3 .
Structures of compounds 1-5 from the roots of A. crenata Sims.

Table 4 .
The fingerprint similarities of root, stem, and leaf from A. crenata Sims.
2.4.PCA and HCA Analysis of 16 Batches of A. crenata Sims

Table 6 .
Grey correlation analysis between the fingerprint of A. crenata Sims and anti-microbial activity.

Table 7 .
The contents of active ingredients in different parts (mg/g) of A. crenata Sims.

Table 8 .
Molecular docking score of the active compounds on key target proteins of bacteria.
a Positive control.Molecules 2024, 29, x FOR PEER REVIEW 12 of 21

Table 9 .
Molecular docking score of the active compounds on key target proteins of fungi.
a Positive control.