Potential Anti-Proliferative, Anti-In ammatory and Anti-Viral Components from Sinopodophyllum Hexandrum Explored Using Bio-A nity Ultra ltration with Multiple Drug Targets


 Background: Sinopodophyllum hexandrum (S. hexandrum) is a typical Chinese herbal medicine with numerous components and remarkable pharmacological activities. However, the specific phytochemicals responsible for its anti-proliferative, anti-inflammatory and anti-viral effects remain unexplored.Methods: The integrated analytical strategy combining bio-affinity ultrafiltration with multiple drug targets was developed to rapidly screen and identify bioactive ligands from S. hexandrum. The in vitro anti-proliferative and COX-2 inhibitory assays of bioactive ligands screened were further verified by sulforhodamine B (SRB) cell proliferation and cytotoxicity detection and COX-2 inhibitor screening kits, respectively. Molecular docking analysis was also implemented by the AutoDockTools 1.5.6 software.Results: 10, 7, 9 and 9 phytochemicals were screened out and identified as the potential Topo I, Topo II, COX-2 and ACE2 ligands, respectively. Hereinto, podophyllotoxin and quercetin with higher EF values displayed strong inhibitory effects on A549 and HT-29 cells comparable with etoposide and 5-FU. Furthermore, compared with indomethacin at 0.73 ± 0.07 mM, podophyllotoxin and kaempferol with higher EF values exerted stronger inhibitory effects with IC50 values at 0.36 ± 0.02 mM and 10.49 ± 0.61 mM, respectively. Additionally, the optimal binding sites and mode of action between bioactive ligands and multiple drug targets were determined by molecular docking. Wherein, isorhamnetin showed a stronger affinity to ACE2 with the binding energy of -5.72 kcal/mol and the IC50 value at 63.95 mM, lower than MLN-4760 (-4.27 kcal/mol and 738.62 mM). Conclusions: The integrative strategy combining multiple drug targets and bio-affinity ultrafiltration LC-MS in the present study showed very promising potential for the quick screening and identifying bioactive ligands in S. hexandrum for Topo I, Topo II, COX-2 and ACE2, and some bioactive compounds screened out from this work were verified with other in vitro assays, and even better than those positive drugs of interest. Based on these findings, we then first constructed an interacting network among multi-components and multi-targets. In this way, we showcased a quick and reliable experimental strategy for uncovering the underlying mechanism of the empirical traditional applications of S. hexandrum which could also provide valuable information for better understanding the therapeutic targets and therapeutic ligands of other herbal medicines.

the present study showed very promising potential for the quick screening and identifying bioactive ligands in S. hexandrum for Topo I, Topo II, COX-2 and ACE2, and some bioactive compounds screened out from this work were veri ed with other in vitro assays, and even better than those positive drugs of interest. Based on these ndings, we then rst constructed an interacting network among multi-components and multi-targets. In this way, we showcased a quick and reliable experimental strategy for uncovering the underlying mechanism of the empirical traditional applications of S. hexandrum which could also provide valuable information for better understanding the therapeutic targets and therapeutic ligands of other herbal medicines.

Background
Sinopodophyllum hexandrum (Royle) Ying (S. hexandrum), belonging to Sinopodophyllum genus, Berberidaceae Family, is a rare perennial herbaceous medicinal herb with extensive applications as a traditional Chinese medicine. It is mainly distributed in Yunnan, Tibet, Sichuan, Hubei, Shaanxi, Qinghai, Gansu, Ningxia and other provinces of China [1]. S. hexandrum embodies abundant pharmacological effects such as anti-proliferative, anti-viral, anti-in ammatory, anti-bacterial, insecticide and cytotoxic activities [2]. In recent decades, the main researches on S. hexandrum have been chie y focused on the establishment of chromatographic methods to quantitatively analyze and measure the content of podophyllotoxin, the chemical compositions and pharmacological effects, and how to prompt the sustainable development and utilization of this endangered plant [3][4][5]. So far, numerous chemical components have been isolated and identi ed from S. hexandrum, namely lignans, avonoids, saponins, polysaccharides and tannins, among which lignans are primarily composed of podophyllotoxin [3,6]. However, owing to its complicated and diversi ed chemical components, the speci c bioactive phytochemicals of S. hexandrum responsible for its anti-tumor, anti-in ammatory and anti-viral effects in the empirical applications, their potential target-ligand interactions, and their possible mechanism of action remain unexplored at present. Therefore, an integrative strategy to rapidly screen and identify latent bioactive ligands in S. hexandrum against speci c drug targets is in great need, which can be very helpful to further decipher the material basis for their e cacy in the development and utilization of this medicinal plant.
The bioactive components in traditional Chinese medicine are considered to be the material basis for the prevention and treatment of diseases. Compared with modern drugs, traditional Chinese medicine is characterized to take its pharmacological effects through the interactions among multiple components and multiple targets [7]. For multi-causal diseases regulated by complicated pharmacological networks, singletarget remedies rapidly develop resistance, bring about poor clinical therapeutic effects, and eventually lead to treatment failure on the basis of the strategy of "one disease-one target-one therapeutic drug" [8]. In response to the limitations of single-target drugs, the development of multi-target natural agents, especially with synergistic effects, will provide greater bene ts in enhancing e cacy and lowering drug resistance [9]. Additionally, nearly half of small molecule drugs were developed as enzyme inhibitors according to current statistics. These small molecule drugs exerted their pharmacological effects by inhibiting the biological activities of key enzymes or other signi cant biological macromolecules in a certain process of biochemical reactions in the body [10]. Consequently, the a nity activity between target biological macromolecules and potential small-molecule ligands has become one of the dominating decisive factors in new drug development [11]. Taking the above research ideas into account, four well-known drug targets closely correlated to the empirical applications of S. hexandrum as a traditional Chinese medicine, including DNA topoisomerase I (Topo I), DNA topoisomerase II (Topo II), cyclooxygenase-2 (COX-2) and angiotensinconverting enzyme II (ACE2), were chosen to actively explore the correlation between potential bioactive constituents screened from S. hexandrum, and its multiple pharmacological effects of anti-proliferative, antiin ammatory and anti-viral activities. DNA topoisomerase (Topo) is a type of ribozyme that can control and change the topoisomeric state of DNA. Topo is divided into Topo I and Topo II in accordance with the instantaneous break of single-strand and double-strand during DNA allosteric action [12]. Moreover, Topo, a momentous anti-tumor molecular target, exhibits high-level expression in tumor cells, and numerous links to the action of anticancer drugs, such as DNA replication, DNA damage and repair, gene recombination and transcription, and cell mitosis and differentiation [13,14]. As regards COX-2, it has been considered as a vital target for new anti-in ammatory and auxiliary anti-cancer remedies. Selective COX-2 inhibitors were characterized by retaining all the bene ts of classic non-steroidal anti-in ammatory agents, involving plenty of pathophysiological processes such as in ammation, cancer, and neurodegenerative diseases [15,16]. As a homolog to the carboxypeptidase ACE, ACE2 can act as a basilic functional receptor for the SARS-CoV-2, regulate the renin-angiotensin system (RAS) as a puissant negative regulator, and participate in the absorption of amino acids by the kidneys and intestines [17][18][19].
In vitro screening test is commonly applied to screen enzyme inhibitors from medicinal plants, and its orthodox procedures comprise biometric-guided separation, puri cation and structural identi cation of puri ed compounds followed by the activity test of individual compounds. However, the whole process of operation is quite laborious, cost-prohibitive and time-consuming, and cannot truly re ect the interactions between the receptor and the natural conformation of active ingredients [20,21]. In order to overcome these limitations, a potent and e cient screening strategy based on bio-a nity ultra ltration combined with liquid chromatography mass spectrometry technology (AUF-LC/MS) has been actively developed for rapidly screening and identi cation of potential bioactive ligands from complex systems such as botanical extracts [11,22]. AUF-LC/MS is a membrane separation technology similar to the dialysis method combining ultra ltration device with mass spectrometry. This approach is suitable for high-throughput screening and rapid identi cation of potentially active small molecules from complicated matrixes such as natural product extracts [23,24]. During the AUF-LC/MS procedures, potential small-molecule ligands in a mixture selectively bind to macromolecule target enzymes based on the principle of a nity, the unbound ligands are eluted from the enzyme-ligand complexes, and then the LC-MS/MS is applied to detect and identify the retained ligands from the denatured receptor [23,25]. To the best of our knowledge, the AUF-LC/MS strategy with four drug targets ( Topo I, Topo II, COX-2 and ACE2) was rst introduced and developed in the present study for the quick screening and characterization of their respective ligands from S. hexandrum in an effort to explore the speci c bioactive components responsible for anti-proliferative, anti-in ammatory and anti-viral activities in the traditional applications. In this way, we could further construct a network based on interactions among multi-components in S. hexandrum and multi-targets by evidence-based experimental studies, and a quick and reliable experimental strategy should also be very helpful to uncover the underlying mechanism of the empirical traditional applications of S. hexandrum. More strikingly, this could be a showcased strategy to offer direct experimental evidences on the application of traditional herbal medicine, and facilitate to provide valuable information for better understanding the main therapeutic targets and therapeutic roles of other herbal medicines. was implemented by a Thermo Access 600 HPLC system connected with a TSQ Quantum Access MAX mass spectrometer (Thermo Fisher Scienti c, San Jose, CA, USA). Analytical HPLC was performed applying an Agilent 1220 liquid chromatography with a Waters Symmetry RP-C18 column (250 mm ´ 4.6 mm, 5 mm).

Plant material
The roots of S. hexandrum were purchased from Panzhihua (Sichuan, China), and kindly authenticated by Prof. Guan wan Hu, a professional plant taxonomist from the Key Laboratory of Plant Germplasm Enhancement and Specialty Agriculture (Wuhan Botanical Garden), Chinese Academy of Sciences.

Preparation of samples
The air-dried root powders (200 g) were smashed, soaked in 90% ethanol overnight, and ultrasonically extracted 3 times (30 minutes for each time) at room temperature. Then, the total ltrates were collected and concentrated with a reduced pressure evaporator, and nally to obtain the crude extracts of S. hexandrum.
A nity ultra ltration with Topo I, Topo II, COX-2 and ACE2 The extracts obtained above were screened by a nity ultra ltration method with four drug targets selected (Topo I, Top II, COX-2 and ACE2) according to our previous study with slight modi cations [26]. Brie y, the principle of a nity ultra ltration method primarily involves three steps, including incubation, interception and release. Firstly, 100 µL tested sample solution (8 mg/mL) was mixed with 10 µL Topo I (5 U), Topo II (2 U), COX-2 (4 U) and ACE2 (0.5 µg) in 2.0 mL EP tubes, and incubated at 37 o C for 40 minutes in the dark.
Meanwhile, the incubation procedures of inactivated enzyme solution (obtained by boiling water for 10 minutes) were consistent with the activated enzyme solution. Afterwards, the incubated solutions were turned over to the ultra ltration tubes with 30 kDa ultra ltration membranes, followed by centrifugation at 10,000 rpm for 10 minutes at 25°C. The unbound components were eluted with 200 mL phosphate buffer saline (PBS, pH 7.04) or tri(hydroxymethyl) aminomethane hydrochloride (Tris-HCl, pH 7.80) 3 to 4 times by centrifugation at 10,000 rpm for 10 minutes at 25°C. Critically, 200 mL 90% MeOH-H 2 O (v/v) was added and incubated at room temperature for 10 minutes, and the mixed ltrates were centrifuged 3 to 4 times at 10,000 rpm for another 10 minutes at 25°C to release the potential bioactive components from the enzymeligand complexes. Eventually, those released ltrates were lyophilized and reconstituted with 50 µL MeOH, and later HPLC-UV/ESI-MS/MS technology was applied to analyze these samples.

HPLC-UV/ESI-MS/MS analysis
The HPLC-UV/ESI-MS/MS analysis was executed using a Thermo Access 600 HPLC system connected with a TSQ Quantum Access MAX mass spectrometer. A Waters Symmetry RP-C18 column with a guard column at 30 °C was used for the HPLC analysis at a wavelength of 292 nm with the ow rate at 0.8 mL/min, and the mobile phases were composed of 0.1% formic acid-H 2 O (A) and acetonitrile (ACN, B), and the optimized HPLC elution conditions were set as follows: 0-40 minutes, 5%-95% B. The mass spectrometer collocated with electrospray ionization (ESI) was carried out in the positive ion mode to generate multifarious fragment ions. Brie y, the optimized instrument parameters were practiced as followed: the mass range was scanned from 150 to 1100 (m/z) in the full-scan mode; the drying gas ow rate was set as 6.0 L/min; the capillary temperature and vaporizer temperature were regulated to 250 °C and 350 °C, respectively; the spray voltage, the cone voltage energy and the collision energy were adjusted to 3000 V, 40 V and 20 eV, respectively; the aux gas pressure and the sheath gas pressure were modulated to 10 psi and 40 psi, respectively. Meanwhile, the Thermo Xcalibur ChemStation software (Thermo Fisher Scienti c) was devoted to the acquirement and analysis of all the above data.
In vitro anti-proliferative assays of samples Non-small lung cancer cells (A549) and colon cancer cells (HT-29) were cultured in Dulbecco's modi ed Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin solution under 5% CO 2 at 37°C in a humidi ed incubator. The cell viability was measured by sulforhodamine B (SRB) cell proliferation and cytotoxicity detection kits as described previously [27]. In brief, these two cells suspension (1.0 ´ 10 4 cells) were seeded in 96-well plates and incubated for 24 h. Then, those tested samples (100 mM) dissolved in the dimethyl sulfoxide (DMSO) were added and treated into the two cells.
Thereinto, dimethyl sulfoxide (DMSO) was taken as the blank control, etoposide and 5-uorouracil  were treated as the positive controls. The optical density (OD) value of each well was measured at 540 nm with a microplate reader. The inhibition rate was calculated by the following equation: Inhibition rate (%) = [(OD 1 -OD 2 ) / OD 1 ] × 100% OD 1 and OD 2 represent the absorbance of the blank control and positive control or samples, respectively.
Additionally, all groups were tested in triplicate, and these results were computed by the GraphPad Prism 8 software.
In vitro COX-2 inhibitory assays of samples In vitro COX-2 inhibitory assays of samples were implemented by using COX-2 (human) inhibitor screening kits in the light of the manufacturer's instructions and our previous publication [28]. Brie y, the tested sample solution was rstly diluted with DMSO into a series of various concentration solutions. COX-2 cofactor working solution (50×), COX-2 substrate (50×) and COX-2 probe (50×) were prepared by diluting 10 times with COX-2 assay buffer, respectively. Then, 10 µL COX-2 cofactor working solution, 10 µL COX-2 working solution, 10 µL tested sample solution (0.625-20 mM), and 150 µL Tris-HCl buffer (pH = 7.9) were appended, mixed and incubated in a 96-well black plate for 10 minutes at 37 °C. In addition, 10 µL COX-2 working solution was replaced by the same volume of COX-2 assay buffer as the blank control group. It was also substituted for the equal amount of DMSO as the 100% enzyme activity control group. After that, 10 µL COX-2 probe and 10 µL COX-2 substrate were rapidly added into each well, and incubated in the darkness for 5 min at 37 °C. Finally, the uorescence measurement was conducted by the microplate reader, in which the excitation and emission wavelengths were monitored at 560 nm and 590 nm, respectively. Indomethacin was used as the positive control. The COX-2 inhibitory activity percentage of each sample was computed referring to the following calculation equation: where can be recorded as RFU blank (relative uorescence unit of the blank control group), RFU 100%enzyme (relative uorescence unit of 100% enzyme activity control group), and RFU sample (relative uorescence unit of the tested sample), respectively. Additionally, the data results were measured in three parallels and presented as means ± standard deviation (SD). The half maximal inhibitory concentration (IC 50 ) value stood for the optimal sample concentration when the COX-2 activity was inhibited by 50%, and was calculated by the GraphPad Prism 8 software.

Results
Analysis of the extracts from S. hexandrum before and after ultra ltration with HPLC-UV/ESI-MS/MS analysis Lignans and avonoids are the two primary ingredients from S. hexandrum [3]. Prior to the ultra ltration screening, the extracts from S. hexandrum were rstly subjected to HPLC-UV/ESI-MS/MS analysis in the positive ion mode, and each chromatographic peak was detected and tentatively identi ed in accordance with the retention time (Rt), UV spectra, protonated molecular fragment information ([M + H] + ), characteristic fragment information, the corresponding standards and literature data. After the ultra ltration screening of the extracts above with Topo I, Topo II, COX-2 and ACE2, the resulting samples were analyzed under the same LC-MS conditions. As a result, the chemical structures of 10 potential ligands binding to Topo I, Topo II, COX-2 and ACE2 were identi ed and summarized at great length in Table 1 and Fig. 1. In order to further clarify the mechanism of action for medicinal plants, it is the rst task to screen and identify active ingredients. Therefore, it is necessary to develop a fast, simple and effective approach for targeted screening of active components so as to associate the chemical ingredients and certain pharmacological activities. In addition, there are few reports on multi-targeted screening to date, which cannot meet the growing demand for multi-targeted screening of medicinal plant extracts. It is essential to accelerate the screening of multi-targeted medicinal plant activity for the discovery and development of new drugs. In this study, AUF-HPLC/MS with the characteristics of simplicity, e ciency and sensitivity was developed to screen Topo I, Topo II, COX-2 and ACE2 ligands from S. hexandrum, respectively. The variation in peak area for screened constituents could re ect a speci c binding a nity between the activated and denatured enzymes before and after ultra ltration. The enrichment factor (EF) represented the capacity for those ligands binding to target enzymes, and the calculation was shown as below: where A a , A b and A represent peak areas of each chromatographic peak from the extract of S. hexandrum upon ultra ltration with activated, denatured and without target enzymes (Topo I, Topo II, COX-2 and ACE2), respectively [31]. If the peak area from the activated group was greater than that of the corresponding inactivated group, those constituents were tentatively speculated as potentially bioactive ligands for target enzymes.
As illustrated in Figs. 2, 10, 7, 9 and 9 components from the extract of S. hexandrum exhibited speci c bindings to Topo I, Topo II, COX-2 and ACE2 after AUF screening assay, which were deduced as potential ligands for Topo I, Topo II, COX-2 and ACE2, respectively. And the EF values of each component were shown in

In vitro anti-proliferative and COX-2 inhibitory assays of bioactive ligands screened
In order to explore the correlation between potential bioactive phytochemicals and pharmacological effects, anti-proliferative and COX-2 inhibitory assays in vitro were carried out so as to detect and validate their inhibitory effects of several bioactive ligands screened out targeting Topo and COX-2. For in vitro antiproliferative assay, compounds 7, 9, 10 and 8 with relatively higher EF values exerted strong inhibitory activities (63.24%, 60.48%, 60.70%, and 32.76%, respectively) on A549 cell at the concentration of 100 µM, and were comparable with the positive controls of etoposide and 5-FU (71.13% and 50.44%, respectively).
Moreover, compounds 10 and 7 also showed remarkable inhibitory effects (46.36% and 40.47%, respectively) on HT-29 cell at the concentration of 100 µM, whereas etoposide and 5-FU were 32.24% and 10.95%, respectively. With regard to in vitro COX-2 inhibitory assays, compared with the positive control of indomethacin at 0.73 ± 0.07 µM, compounds 9 and 10 with higher EF values showed signi cant inhibitory effects with IC 50 value at 0.36 ± 0.02 µM and 10.49 ± 0.61 µM, respectively. Hence, it was well worth shing out and identifying potential bioactive ligands from S. hexandrum combining its empirical applications.

Molecular docking simulation
Molecular docking studies were implemented to further simulate the target proteins and several representative compounds with the highest EF values. We applied the AutoDockTools 1.5.6 and Discovery Studio 4.5 Client software with the three-dimensional crystal structures of Topo I, Topo II, COX-2 and ACE2 in this work, respectively. During this simulation procedure, the grid box dimensions and the centroid coordinate for molecular docking of the macromolecular target proteins were shown in Table 2. The molecular docking results of several representative compounds and positive drugs against Topo I, Topo II, COX-2 or ACE2 were also displayed in Fig. 4 and Table 3. As shown in Table 3, podophyllotoxin exhibited a higher a nity to Topo I as its binding energy (BE) and the theoretical IC 50 values were − 6.32 kcal/mol and 23.25 µM, lower than the positive control 5-FU (-3.70 kcal/mol and 1.95 mM) and slightly higher than Topo I inhibitor camptothecin (-7.57 kcal/mol and 2.85 µM).
Considering Topo II, quercetin with a larger EF value displayed a strong a nity to Topo II with the BE of -6.99 kcal/mol and the theoretical IC 50 value of 7.53 µM, which was comparable to Topo II inhibitor etoposide (-7.62 kcal/mol and 2.59 µM). In addition, kaempferol exhibited a higher a nity to COX-2, and its BE and the theoretical IC 50 values were calculated as -7.22 kcal/mol and 5.10 µM, which was not much different from the positive control indomethacin (-9.18 kcal/mol and 186.44 nM). With regard to ACE2, isorhamnetin was discovered with a high binding a nity of -5.72 kcal/mol and the theoretical IC50 value of 63.95 µM, lower than that of ACE2 inhibitor MLN-4760 (-4.27 kcal/mol and 738.62 µM). Above all, molecular docking analysis indicated that some bioactive ligands screened with the largest EF values were provided with relatively lower binding energies and inhibitory effects compared with the positive controls or the same group combining other bioactivity assays in vitro such as anti-proliferative and COX-2 inhibitory assays. The docking results were consistent with the ultra ltration screening and in vitro bioactivity veri cation results, further con rming the feasibility of the molecular docking approach.

Discussion
Ultra ltration analysis of Topo I, Topo II, COX-2 and ACE2 ligands and the in vitro bioactive validation Based on the ndings in Table 1, the intrinsic and intricate correlations between bioactive constituents and multiple targets as well as the pharmacological effects could be partially explored, and the mechanism of action for S. hexandrum could also be further deciphered especially on how to exert its traditional curative effect on in ammatory, cancer and viral diseases.
Among these compounds screened out, compounds 7, 8, 9, and 10 exhibited relatively higher a nity to Topo I or Topo II, which was speculated that these components may target one or two enzymes so as to exert potential anti-proliferative effects. In vitro anti-proliferative assays displayed that these four compounds with higher EF values presented stronger inhibitory effects compared with etoposide and 5-FU. These results indicated that these active components screened out may act on Topo I or Topo II to exert potential antiproliferative effects. With respect to COX-2 inhibitory assays in vitro, compounds 9 and 10 with higher EF values exerted favorable inhibitory activities comparable to indomethacin. These compounds were preliminarily inferred to be the potential anti-in ammatory ingredient group. Concerning the anti-viral effect on ACE2, compound 8 with the highest EF value exhibited high a nity with ACE2, which may be speculated that this compound can be provided with anti-viral activity.
On the one hand, these potential active constituents screened out with AUF-LC/MS were further veri ed to possess anti-proliferative and anti-in ammatory effects through in vitro inhibitory assays. On the other hand, the chemical constituents with larger EF values are tightly bound to target enzymes, and they could be maintained during the interaction with target enzymes. Owing to numerous and diversi ed active ligands screened out and identi ed, several active components may be able to act on one or more target enzymes.
Meanwhile, it was observed that there existed synergistic effects among these potential bioactive constituents, conjointly exerting various pharmacological activities. It was found that compound 8 could potently action all four drug targets to exert anti-proliferative, anti-in ammatory and anti-viral activities; while some compounds exhibited obviously special preference for certain drug targets, like compound 7 for Top I and Top II, compound 9 for Topo II and COX-2 as well as compound 10 for Top I and COX-2. More importantly, this newly integrative strategy combining four drug targets with UF-LC/MS could be used to construct a multi-component and multi-target network based upon experimental evidences as shown in Fig.   3, which could bring insight into the mechanism of action regarding the empirical use of S. hexandrum as a traditional medicine, and promote more new methods to be developed for a better understanding of other traditional herbal medicines.

Molecular docking analysis of representative compounds
Molecular docking is an essential approach to study and predict the interactions between receptors and ligands, further exploring the possible mechanism of bioactive ligands against the target enzymes involved  [33]. Furthermore, the docking simulation results uncovered that three conventional Hbonds were formed between the hydroxyl group of kaempferol and residues Tyr355, Gln192 as well as Gly526 of COX-2, part of which interacted with celecoxib as well [34]. Kaempferol also revealed other interaction forces such as pi-cation interaction with residue Arg513; pi-alkyl interaction with residue Leu352; pi-sigma interaction with residues Ser353 and Val523; and van der Waals interaction with residues Ala516, Ala527, Arg120, His90, Ile517, Leu384, Met522, Phe381, Phe518, Phe529, Ser530 and Tyr385. Moreover, isorhamnetin exerted a relatively strong a nity to ACE2 and was observed to form the conventional H-bonds with residues Glu375 and Ala348 to enhance their a nity against ACE2. There also existed in other important driven forces such as van der Waals force (interaction with residues Asp382, Arg518, Glu398, Gly395, Gly399, Thr347 and Zn804), hydrophobic (interaction with residues His378 and His401) and electrostatic (interaction with residues Arg514 and Glu402) effects in the processes of molecular docking analysis. In addition, it has been reported that isorhamnetin can be used as potential 3CL pro inhibitors to target PIK3CG and E2F1, and then inhibit the replication of SARS-CoV-2 through the PI3K-Akt signaling pathway, and play the anti-viral effect to treat lung injury in COVID-19 by acting on CASP3, CCL2, IL6 or other targets through IL-17 or HIF-1 signaling pathways [35]. In general, these representative components in S. hexandrum exhibited the potential to be developed into the lead candidates responsible for its empirical antiproliferative, anti-in ammatory or anti-viral effects.

Conclusions
S. hexandrum, as a traditional herbal medicine, has long suffered from a lack of experimental evidence regarding the material base and the mechanism of actions regarding its empirical use for the treatment of various diseases due to its complicated and diversi ed chemical components and intricate e cacy. To overcome this tough challenge, the integrative strategy combining four drug targets closely correlated to empirical application of S. hexandrum and a nity ultra ltration LC-MS in the present study showed very promising potential for the quick screening and identifying its multiple bioactive components corresponding to the four respective drug targets selected. Based on these direct experimental evidences, we inferred and constructed for the rst time a ligand-target network among its multi-components and respective multitargets, which could be very conducive to unveil the underlying mechanism of the empirical traditional applications of S. hexandrum. Next, the screening results above were further veri ed using other in vitro bioactivity assays and molecular docking analysis, which con rmed that the integrative strategy could offer direct and reliable experimental evidences for the empirical applications of S. hexandrum. More strikingly, some bioactive compounds corresponding to the therapeutic drug targets obtained in this work were even better than those positive drug controls. Herein, we showcased a quick and reliable experimental strategy for uncovering the underlying mechanism of the empirical applications of S. hexandrum, and could provide valuable information for a better understanding of the main therapeutic targets and therapeutic roles of other traditional herbal medicines.