Adhatoda Vasica: A Potential Ayurvedic Intervention Against COVID-19 Associated Impaired Immune Response and Hypoxia-Inflammation Phenotype


 Background: The importance of hypoxia inducible factor-1 α (HIF-1α) stabilization in uncontrolled infection and inflammation is widely accepted. Several inhibitors of HIF signalling are in clinical trials for malignancy, ischemia and inflammatory diseases. Increased hypoxia is being reported to be an important modifier for several pathological features of COVID-19 such as impaired immunity, hyper-inflammation, thrombosis, lung injury and sepsis. Methods: In this study we tested the effect of whole aqueous extract Adhatoda Vasica (AV), that our group has shown to have anti-hypoxic and anti-inflammatory effects, on various outcomes of hypoxic response. Effects of AV were assessed in preclinical mouse models of pulmonary fibrosis, bacterial sepsis and siRNA induced hypoxia-thrombosis phenotype. Therapeutic relevance of AV in current pandemic were also examined through transcriptome and molecular docking analysis. Results: Oral administration AV extract attenuated the increased levels of airway inflammation, collagen content, transforming growth factor-b1 (TGF-b1), IL-6, HIF-1α and improved the overall survival rate in bleomycin treated and Cecum Ligation and Puncture (CLP) induced mice. AV treatment also rescued the prolyl hydroxylase domain 2 (phd2) siRNA induced HIF-1α and associated blood coagulation phenotypes in mice. Transcriptome analysis of lungs of AV treated naïve mice reveal downregulation of hypoxia, inflammation, TGF-b1 and angiogenesis and upregulation of adaptive immunity related genes. These genes and pathways show opposite expression in transcriptome of BALF and PBMCs of SARS-CoV2 infected patient. Molecular docking of AV constituents presents in extract reveal many molecules with low binding energy (≤ -8) to multiple SARS-CoV2 and host target proteins that are relevant in viral entry and replication. Conclusion: Our results provide a scientific rationale for this ayurvedic herbal medicine in ameliorating the hypoxia-hyperinflammation features which could be useful against SARS-CoV-2 infection.


Background
Increased alveolar hypoxic response levels are inevitable consequences of many respiratory disorders such as chronic obstructive pulmonary disease and pulmonary brosis [1,2]. The key player of cellular response to hypoxia is the hypoxia-inducible factor (HIF)-1α and its regulatory protein, the prolyl hydroxylase domain (PHD)-2 enzyme [3]. The induction of HIF-1α is considered to be pro-in ammatory. It leads to transcriptional activation of essential genes implicated in airway remodelling and in ammation, such as vascular endothelial growth factor, transforming growth factor-1b, inducible nitric oxide synthase, interleukin − 17 , and IL-6 [3,4]. Thus, it is not just a consequence of diseases, elevated tissue/cellular hypoxia actively participates in exaggerating the in ammatory response contributing to progressive lung damage/injury. HIF-1α also plays pivotal role in infection, especially in promoting viral and bacterial replication [5]. In the present COVID-19 pandemic by the severe acute respiratory coronavirus 2 (SARS-CoV2) the role of hypoxia response in inducing severe lung in ammation and other outcomes has been one of the most highlighted observation [6][7][8]. Clinically, the interaction of the host and SARS-CoV2 is broadly described in three stages: rst, asymptomatic state; second, a non-severe symptomatic state characterized by upper airway and conducting airway response; third, severe respiratory symptomatic state with the presence of hypoxia, acute respiratory distress syndrome (ARDS) and progression to sepsis [7]. During incubation and non-severe state, a speci c humoral and cell-mediated adaptive immune response is required to eradicate the virus and prevent disease progression to a severe state. Thus, strategies to boost immune responses at this stage are undoubtedly important [7,9]. However, defective immune response causes further accumulation of immune cells in the lungs, progressing to aggressive production of a proin ammatory cytokine such as IL-6, TNF-Alpha resulting in an in ux of immune cells and cytokines that damage the airways/ lung architecture. This vast release of cytokines by the immune system in response to the viral infection and/or secondary infections causes severe in ammation, endothelial dysfunction, sepsis and multi-organ damage [7,9,10]. In addition, recent research also report coagulation abnormalities in severe COVID-19 cases [11]. The relation of hypoxia-coagulation is well known, where we and others also showed the crucial role of hypoxic response in the form of thrombosis and bleeding susceptibility through HIF-1α and vWF axis [12,13]. Thus, medicinal agents that possess immune-boosting and antihypoxic effects could hold a promise for a better therapeutic option to preclude the SARS-CoV2 infection and severity.
We have recently shown an extract of Adhatoda Vasica (AV); an ayurvedic medicine possesses robust anti-hypoxic properties and can reduce severe airway in ammation induced by an augmented hypoxic response in treatment-resistant asthmatic mice [14]. The anti-HIF-1α effect of AV also restores the cellular hypoxia-mediated loss of mitochondrial morphofunction in vitro [14]. As a follow-up we evaluated AV's usefulness in other severe lung pathologies, where hypoxia signalling is pertinent. With hypoxia assuming relevance in COVID19 during this time we further tested it potential in potentially limiting in ammatory outcomes and viral replication using whole-genome expression and molecular docking analysis respectively.

Preparation of plant extract and LC-MS ngerprinting:
Adhatoda Vasica (AV) was collected from Delhi-NCR region, India in the owering season (November to March). Water extract of plant (leaves, twigs and owers) was prepared according to classical method described for rasakriya in Caraka Samhita [15]. The process for the formulation involved preparation of decoction condensation and drying as described in earlier study [16]. Chemical ngerprinting of prepared AV extract was carried out by LC-MS at CSIR-CDRI, Lucknow, India; in two independent experiment.

Animals
The study was designed and performed following guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA) and approved by Institutional Animal Ethics Committee of CSIR-Institute of Genomics & Integrative Biology (IGIB), New Delhi, India. The BALB/c and C57BL/6 male mice (8-10 weeks old) were bred under the pathogen-free condition. They were acclimatized to animal house environment one week before starting the experiments at CSIR-IGIB, New Delhi, India and maintained according to guidelines of CPCSEA. All the surgical procedures were performed under sodium pentobarbital anaesthesia and maximum efforts are taken for minimum suffering of animals.
Grouping and treatment of mice Mice were mainly divided into two groups as Vehicle and treatment according to the experiment. In case of Cecum ligation puncture (CLP) experiment (n = 5-9), BALB/c mice were divided in Vehicle ( distil water and 10% ethanol, oral) and CLP (mice underwent CLP surgery). CLP mice subdivide in CLP+Cyclo A (Cyclosporin A treated CLP mice) and CLP+AV-D2 (Adhatoda extract treated CLP mice) group. For CLP experiment, treatment of AV (130mg/kg dissolved in distilled water, oral) or Cyclo A ( Cyclosporin A, 15mg/kg dissolved in 10% ethanol, orall) was started two days (48hours) before CLP and was continued till the mice survives after CLP ( gure 1). In that some mice (n=3-4) from each group were sacri ced after 20 hours of CLP to assess lung histology and cytokine levels. Similarly, in bleomycin brosis model (n = 5), C57BL/6 mice were divided in Vehicle (i.e. Sham), Bleo (bleomycin treated) and Bleo+AV-D2 (AV 130mg/kg treated Bleo mice). In that AV treatment was done from day 18 to 21, as shown in the schematic ( gure 1A). Bleomycin (3.5 U/kg of mice) was given intratracheally to iso urane-anesthetised C57BL/6 mice on day 0 of the protocol ( gure 1) to induce brotic changes in mice as described previously [17]. For transcriptomic research (n = 4-5), BALB/c mice divided into Vehicle (distil water, oral) and Adhatoda Vasica (AV) extract group. AV group is further subdivided into two according to its dose: AV-D2 (Adhatoda Vasica extract 130 mg/kg, dissolved in distilled water, oral) and AV-D4 (Adhatoda Vasica extract 260 mg/kg, dissolved in distilled water, oral) as described previously [14]. Distil water or AV ( 130mg/kg or 260mg/kg) treatment was given to mice by oral gavage for consecutive four days as represented in the gure. In PHD2 siRNA induced hypoxia model (n = 4-5) BALB/c mice were divided in scrambled siRNA (Scrm siRNA), prolyl hydroxylase domain-2 siRNA (PHD2 siRNA) and AV-D4 treated PHD2 siRNA group (PHD2 siRNA+AV-D4) group. AV-D4 dose (260 mg/kg, dissolved in distilled water, oral) given for four consecutive days and 90µg siRNA (Sigma) administered intranasally which dissolved in ultrapure DNase and RNAse free water with in-vivo jetPEI as the transfection reagent (Polyplus Transfection, France) to iso urane-anesthetised mice on day 1, 3 and 5th of the protocol.
Clotting and bleeding time assay, blood collection and platelet measurement Tail bleeding was measured as described previously [13]. Brie y, anesthetized mice's tail amputated with a sharp scalpel, and bleeding time was then determined by monitoring the duration of animal tail bleeding until it ceased and was kept in a prone position and immersed in PBS. Clotting time measured by the capillary tube method. The mice's tail was cleaned with 70% alcohol and punctured with a 1ml syringe needle. Filled two capillary tubes with free-owing blood from the puncture site after wiping the rst drop of blood. Stop clock started and capillary tubes were broken to see whether a thin brin stand formed between two broken ends. After brin stand is observed, clotting time measured from the average of two capillary tubes. For platelet measurement, blood obtained by cardiac puncture and collected in EDTA coated MiniCollect tubes (Greiner Bio-One Gmbh, kremsmünster, Austria) as described [13]. The whole blood was used to measure total and of active Platelet count and was carried out through ow cytometry using FACSCalibur (BD Biosciences, USA). Brie y, diluted whole blood (1:4) in PBS was incubated with APC conjugated anti-CD62P (eBioscience Inc, San Diego, CA, USA) and FITC conjugated anti-CD41 (eBioscience Inc, San Diego, CA, USA) for 15 min. Matched uorescein-conjugated isotype control antibodies were used simultaneously for staining for comparison. The activity was compared using CellQuest Pro software (BD Biosciences, USA).

Bronchoalveolar lavage uid collection and histopathology
Bronchoalveolar lavage uid (BAL) was collected by instilling 1 ml PBS into the tracheotomised airway and recovered BAL uids were processed to get total leukocyte count, as described previously (). For lung histology, the lungs were excised and xed in 10% buffered formalin. The xed, para n-embedded tissues cut into 5um sections and either stained with hematoxylin and eosin (H&E) to asses in ammation or Masson's trichome (MT) staining to assess collagen content.
Cecum ligation puncture (CLP) procedure Mice were anesthetised by injecting intraperitoneally a solution of 1:1 ketamine (75mg/kg) and xylazine (15mg/kg). The abdomen was shaved, and the peritoneum area was disinfected betadine solution followed by wiping with a 70% alcohol. Under aseptic conditions, a 1 cm midline incision was made, and the cecum carefully exposed with the adjoining intestine. The cecum was then tightly ligated with a 3.0 Mersilk (PROLENE, 8680G; Ethicon) sutures at the base and punctured once with a 19-gauge needle on the same side of the cecum. A small amount of stool extruded to ensure patency of the puncture sites. The cecum then returned to the peritoneal cavity, and the wound was closed with 3.0 Mersilk sutures.
Control mice (i.e. Sham), the cecum was exposed out and then returned to the peritoneum without ligation or puncture. Mice were resuscitated by injecting subcutaneously 1 ml of pre-warmed 0.9% saline solution using a 25G needle. After surgery, animals placed immediately to a cage with exposure to a heating lamp of 150W until they recovered from the anaesthesia. The recovery time is from 30 min to 1 hour. Mice were monitored every 12 hours for survival or euthanised after 20 hours (n=3) for measurement of cytokines while they were fed with their regular diet and water. Two independent experiments recorded the mortality of mice after CLP surgery.

RNA isolation and whole transcriptome analysis:
Total RNA was isolated from mouse lung tissue treated with AV (AVD2 and AV-D4) or distilled water (vehicle) using the RNeasy Plus Mini Kit (Qiagen, CA, USA) following the manufacturer's protocol. For genome-wide expression analysis, the Affymetrix GeneChip MTA 1.0 array was used according to the manufacturer's instruction. For each sample, 250 ng of RNA was quanti ed and hybridized to microarray chips following a series of consecutive steps described in the protocol. After hybridization, microarray chips are then scanned using an Affymetrix GCS 3,000 scanner (Affymetrix, CA, USA) and the signal values are further evaluated using the Affymetrix® GeneChip™ Command Console software. Raw data automatically extracted using the Affymetrix data extraction protocol in the Affymetrix GeneChip® Command Console® Software (AGCC). CEL le import, mRNA level, all analysis, and export of the results were all performed using Affymetrix® Expression Console™ software. A comparative study between the vehicle and the AV treated samples done by using fold-change and p-value, genes considered to the differentially expressed by applying the criteria of signi cance p-value less than or equal to 0.05.

Functional enrichment and Connectivity map analysis
For functional analysis, we used Enrichr (amp.pharm.mssm.edu) tool. For pathway and gene ontology analysis, we examined gene enrichment in Cellular Compartment, Biological Processes, BioPlanet, Wiki, KEGG human pathway and gene set enrichment was considered if P-value less than 0.05 in Enrich r tool.
For connectivity map (CMap) analysis, differentially expressed genes ranked according to fold change and list of top 150 up and down-regulated genes compatible with the CMAP data signatures was used to query the connectivity using clue.io touchstone database.

SARS-CoV-2 transcriptome meta-analysis
We obtained the raw RNA sequencing data from SARS-CoV-2 patients Broncho-alveolar lavage uid (BALF) and peripheral blood mononuclear cells (PBMC) from the authors [18] and analysed them inhouse. Functional enrichment and gene set enrichment was done as described above using the differentially expressed genes. The results were then compared with AV transcriptome data to nd intersections between gene ontologies, enriched gene sets, and connectivity map perturbations between upregulated genes of BALF, PBMC and downregulated genes of AV and between downregulated genes of BALF, PBMC and upregulated genes of AV.

Molecular docking
The complete genome sequence of the novel SARS-CoV-2 virus was obtained from the National Centre for Biotechnology Information (NCBI) nucleotide database (NC_045512.2). The available 3D crystal structures of all the target proteins such as 3CLpro, PLpro, RdRp, S-protein, ACE2 and JAK2 were taken from protein data bank [19]. Others structures (NSP4, NSP7, NSP8, NSP9, NSP13, NSP14, NSP15 and NSP16, and TMPRSS2) were built using homology modeling with suitable templates using Swiss model [20] and I-TASEER web-servers [21]. The active regions of the proteins were identi ed by COACH metaserver and the results were compared with results from CASTp web server [22]. The anti-COVID-19 activity of the compounds extracted from the Adhatoda vasica were investigated using Molecular Docking studies using Schrodinger suite (Maestro) [23] and AutoDock vina packages [24]. In Schrodinger suite, all the target proteins were prepared using protein preparation wizard that included optimization followed by minimization of heavy atoms of proteins. The energy minimized 3D structures of all the ligands were prepared using LigPrep. The best pose of ligands that t well in the protein cavity was carried out using OPLS3 force eld with Glide package in Extra Precision mode (XP) mode. According to the size of binding cavity of the proteins, the coordinates x, y and z of the grid box were chosen with the grid resolution of 1 Å for calculations using AutoDock vina package.

Statistical analysis
Statistical signi cance determined by one-way analysis of variance and analysis was done using GraphPad Prism software. In the case of mice experiment, all data represent mean ± SEM; n= 3-10 in each group and signi cance denoted by *p <0.05, **p <0.01, ***p <0.001. p-value > 0.05 is considered non-signi cant (NS). Signi cance of the survival study determined by Log-rank (Mantel-Cox) test using GraphPad Prism software.

AV treatments inhibits the bleomycin induced pulmonary brosis features as well as increased HIF-1α levels in mice
To test the effect of AV treatment on lung brosis, bleomycin treated mice were orally administered with AV (130mg/kg, AV-D2) as shown in gure 1A. We observed a signi cant increase in TGF-β1 and HIF-1α levels in bleomycin (Bleo) treated mice lung compared to control-Sham mice, which decreased after AV-D2 treatment ( gure 1B, C). Masson's trichrome staining showed a marked increase in collagen deposition in Bleo mice lungs compared to Sham mice ( gure 1D). AV-D2 treatment reduces this increased collagen deposition in Bleo mice ( gure 1D).
AV ameliorates the hallmarks of lung in ammation and injury in mice model of sepsis Mice that underwent CLP (Cecal ligation and puncture) surgery show signi cant increases in IL-6 levels, whereas IFN-g was decreased in lung homogenate after 20 hours of surgery compared to sham mice ( gure 1F, G). AV pre-treatment restored the levels of both cytokines in mice lungs but Cyclo-A (a positive control) pre-treatment reduced only IL-6 levels in mice ( gure 1F, G). Besides, histological analyses showed that CLP went mice lung sections stained with haematoxylin and eosin (H&E) had increased in ammation and blood exudation ( gure 1H). Pre-treatment of AV-D2 and Cyclo A to CLP mice seems effective in reducing these lung histological changes ( gure 1H). CLP surgery also leads to a signi cant decrease in mice survival rate compared to sham group ( gure 1I). Treatment of Cyclo-A or AV-D2 to CLP mice signi cantly increases their survival rate compared to CLP untreated mice ( gure 1I). In CLP+Cyclo A and CLP+AV-D2 group, the mice survival rate after 24 hours is 66.6 and 44.4 %, respectively ( gure J).Though, in both group's survival rate is 33.33% at the end of 142 hours of CLP surgery ( gure J).
AV treatment inhibits hemostatic outcomes of hypoxia induced by PHD2 siRNA in mice Next, to test whether AV anti-HIF-1α effects also prevents the hypoxia-induced blood coagulation phenotype [13], we treated BALB/c mice with AV-D2 (130 mg/kg) and AV-D4 (260 mg/kg) concentration ( gure 2A). Oral administration of AV-D2 and AV-D4 to naïve mice does not cause any signi cant change in body weight, and lung and liver histological architecture ( gure 2B, C), indicating its non-toxic nature. In the case of hemostasis parameter, treatment of AV-D4 dose to naïve healthy BALB/c mice causes a decrease in total as well as activated platelet count. Still, it does not affect mice tail bleeding time ( gure 2D, E). To con rm the above-observed effect of AV-D4 on blood parameters, we induce cellular hypoxia in mice by speci c PHD2 siRNA treatment ( gure 2F), as described previously [13]. PHD2 siRNA treatment leads to a signi cant decrease in blood clotting and tail bleeding time ( gure 2G). It also causes an overall increase in total and activated platelet count in mice blood ( gure 2H). These changes induced by PHD2 siRNA are associated with an increase in blood HIF-1α and vWF levels indicating the platelet aggregation development ( gure 2I, J). Interestingly, AV-D4 treatment to PHD2 siRNA mice causes a signi cant reversal of blood coagulation phenotype in terms of mice's blood clotting time, platelet count (both total and active) and vWF levels ( gure 2 G-I). These effects of AV are associated with the reversal of the increased blood HIF-1α levels ( gure 2J). However, AV treatment does not affect the mice's bleeding time, which was reduced after PHD2 siRNA treatment ( gure 2G).

Modulation of immune response and hypoxia pathway genes: Revealed from lung transcriptome of AV treated mice
Lung transcriptomic analysis showed an upregulation of 1258 genes after AV-D4 treatment and 375 gens after AV-D2 treatment in naive mice. While, 1133 genes in AV-D4 and 262 genes in AV-D2 were downregulated, compared to Vehicle (distilled water) treated mice. We observed enrichment of pathways like IL-2 signaling, T cell signaling, T cell-mediated immunity, natural killer cell-mediated cytotoxicity, Haematopoietic cell lineages in the AV-D4 up-regulated genes. Similarly, biological processes like neutrophil activation and degranulation, neutrophil-mediated immunity, immune response regulation, and cellular defense response was enriched in AV-D4 up-regulated genes ( gure S1A). In AV-D2 up-regulated genes, pathways relevant in mitochondria, T cell signaling, cytotoxic T cell-mediated immune response are enriched ( gure S1A). Whereas, pathways like collagen biosynthesis, extracellular matrix organization, TGF beta regulation, hypoxia, and associated in ammatory MAPK-signaling signi cantly enriched in both AV-D4 and AV-D2 downregulated genes ( gure S1A). Overall, it indicates that AV treatments favour the expression of genes that are important in immunity and adaptive immune response and inhibitory to genes involved in hypoxia associated angiogenesis, brosis, and in ammatory cascade.
Since the above observations made in independent experiments pointed towards AV's role in hypoxiain ammation-coagulation-sepsis axis, we thought it would be worthwhile to investigate the repurposing possibility of AV against SARS-CoV2 infection. We examined transcriptome pro les of BALF and PBMC specimens of SARS-CoV2 patients recently published by Xiong et al [18]. Analysis of differentially expressed genes in BALF and PBMC of SARS-CoV2 patients shows that the genes which are up-regulated during infection are related to viral transcription, viral replication, and viral process ( gure S1B). Besides, pathways like TGF-beta signaling, VEGFR pathway, IL-6 signaling, and HIF-1 signaling are signi cantly enriched in a set of up-regulated genes of BALF or PBMC or both samples ( gure S1B). Whereas downregulated genes are related to pathways like adaptive immune system, IL-2 signaling, mitochondrial relevant function, T cell signaling, local in ammatory response, and processes related to neutrophil activation/degranulation, neutrophil-mediated immunity, and interferon-gamma signaling ( gure S1B).
These ndings show that the pathways and processes enriched in SARS-CoV2 infected individuals overlap with those of AV gene expression but in the opposite direction. Overall, it was observed that functional enrichment patterns in the up-regulated genes of SARS-CoV2 BAL and PBMC samples were seen in down-regulated genes of AV treated mice and vice-versa, suggesting a possible therapeutic application of AV in preventing or reversing the SARS-CoV2 induced transcriptional changes.
Towards this analysis of 'common genes' between AV and SARS-CoV2 i.e. the genes that are up-regulated by AV treatment, but the same were down-regulated during SARS-CoV2 infection or vice-versa, shows pathways/processes such as immune response, brosis, hypoxia response, platelet activation, and aggregation are enriched ( gure 3A, S2). In the case of AV treatment, shared immune function-related genes are up-regulated ( gure 3, S2) and hypoxia, brosis, and hemostasis related genes are downregulated ( gure 2, S2), compared to SARS-CoV2 BAL and PBMC samples as shown in the gure 3A and detailed in gure S2A. These ndings supported with the increased levels of immune cells in bronchoalveolar lavage uid (BAL) of mice treated with AV, compared to vehicle-treated mice ( gure S2F).
CMap analysis identi ed the list of compounds whose expression pattern matches with AV and could be repurposed against SARS-CoV2 infection To identify similarities and differences in the gene expression pattern of AV with other FDA-approved drugs and bio-actives, we mapped the transcriptomic signature of AV using Connectivity Map (CMap) database. A positive score in CMap analysis indicates a similar expression pattern of AV with compared compounds gene expression signature, whereas a negative score indicates an opposite pattern. We also queried the SARS-CoV2 BAL gene signature with CMap and observed that the compounds which are connected positively with AV have a negative score with SARS-CoV2 ( gure 4A) transcriptome data. Noteworthy is the HDAC inhibitor 'Apicidin' which has high positive connectivity with AV gene signature(+91.76 score) ( gure 4A) and a negative connectivity (-79.09 score )with SARS CoV2 BALF signature. ( gure). Apicidin has recently been shown to bind to viral non-structural protein 5 (nsp5) protein and features in the list of potential drugs which could be repurposed against SARS-CoV2 infection [25].

Molecular docking predicts AV could bind SARS-CoV2 and host target proteins
Chemical components present in Adhatoda Vasica (table S1, S2) were examined by molecular docking analysis with SARS-CoV2 and host proteins. Tables 1 and S3 represent the docking results of the constituents of Adhatoda vasica with key target proteins of SARS-CoV-2 virus. The trend in the docking results obtained from Schrodinger suite of packages is almost similar to those of AutoDock vina package. For brevity and further analysis, results from Schrodinger package are only considered. It is observed from the results that all the avonoids derivatives are found to have high negative binding a nity than quinazoline alkaloids analogues. Speci cally, the three compounds namely Luteolin-6,8-di-Cglucoside, Luteolin-6-C-glucoside-8-C-arabinoside and Luteolin-6,8-di-C-arabinoside have higher docking score values (-7.45 kcal/mol for PLpro, -9.43 for S-protein, -15.25 for NSP14, -11.12 for TMPRSS2, -9.82 for ACE2), (-11.59 kcal/mol for 3CLpro, -10.60 for RdRp) and (-10.86 kcal/mol for NSP16, -13.82 for JAK2) respectively (table 1). The key residues of target proteins that contribute more for binding with the compounds of Adhatodavasica are shown in Table S4. The compound Luteolin-6-C-glucoside-8-Carabinoside makes cation-π interaction with residue (His 41) of 3CLpro with higher a nity -11.59 kcal/mol ( gure 4B). The π-π stacking interaction between the Luteolin-6,8-di-C-glucoside and residues (Phe 426 and Phe 506) of NSP14 protein enhances the binding a nity to -15.25 kcal/mol when compared to other compounds as shown in gure 4C and other interaction plots are represented in Supporting material gure S3.

Discussion
Adhatoda Vasica or Vasa has been extensively used in Ayurveda for treating a wide range of in ammatory and respiratory conditions [26]. Even in modern clinical practice, it is recommended for strong bronchodilatory and antitussive effects [26,27]. Also, AV's active ingredients and their derivates such as Bromhexine, Ambroxol are effective against various respiratory ailments like asthma, COPD, and tuberculosis [26]. However, none of the studies have so far studied the effect of AV on molecular markers of asthma/in ammation phenotype in its aqueous extract form prescribed in Ayurveda. For the rst time, we have shown that the anti-asthmatic effect of AV is not only limited to acute asthmatic features but it also alleviates the severe airway in ammation in steroid-nonresponsive asthmatic mice [14]. We show that the AV extract causes inhibition of HIF-1α (key transcription factor in hypoxia) by increasing the expression of its negative regulator, PHD2, thereby modulates hypoxic response, which forms the basis for its diverse therapeutic effects [14]. Further, AV treatment to lung epithelial cells has also been shown to reduce hypoxia's downstream impact such as mitochondrial dysfunction [14]. In this study we examined anti-hypoxic effects of AV in other hypoxia-in ammation prevalent conditions.
Our study shows that AV could reverse the pulmonary brosis (PF) pathological features in the bleomycin mice model (Fig. 1). HIF-1α stabilization is observed in many cell types of PF lungs and causes increases in collagen synthesis, brosis, TGF-β1, VEGF levels and proliferation of broblasts [1,28]. AV treatment reduces the increased expression of HIF-1α protein in bleomycin mice lungs and attenuates increased TGF-β1 and collagen content in mice lungs (Fig. 1A-D). These preliminary results substantiate the hypoxia modulating effect of AV in chronic lung disease conditions such as brosis.
Considering the anti-in ammatory role of AV, we also determined the effect of AV in severe lifethreatening in ammatory conditions, i.e., sepsis, using most commonly used cecal ligation and puncture (CLP) mediated polymicrobial sepsis model (Fig. 1E). AV treatment reduces the increased IL-6 levels in mice, an important proin ammatory cytokine frequently associated with severe in ammatory conditions like sepsis (Fig. 1F). AV treatment is also able to alleviate the increased in ammatory, blood exudates as well as reduction of CLP induced mortality rate in mice (Fig. 1H-J). Interestingly, we observed a decrease in IFN-γ levels upon CLP in mice, which was restored after AV treatment (Fig. 1G). Our observation with the IFN-γ level in CLP mice is not in line with previous studies [29], and it is likely due to the differences in the time of measurement after CLP. However, our results are similar to the clinical ndings that support adjuvant IFN-γ immunotherapy concept to improve the host immune response against infection [30,31].
Hypoxia or HIF-1α activation is also thought to be a possible link between sepsis and thrombosis [32]. We observed that higher concentration of AV could reduce the increased platelet levels in naive mice (Fig. 2D). Further, mice treated with AV also mitigates the HIF-1α induced thrombotic complications and signi cantly prevents the decrease in blood clotting time and reduces the increased levels of platelets, vWF, and HIF-1α in PHD2 siRNA treated mice (Fig. 2F-J). These results suggest that the anti-hypoxic effect of AV could also prevent the development of hypercoagulable state in response to hypoxia.
In the course of our study, we realize that the effects of AV on phenotypic features of the lung and systemic in ammation could also prove bene cial for the present pandemic situations. In SARS-CoV2 infection, the presence of elevated hypoxia response seems to be a consequence of hyper-in ammation that contributes to disease severity [6,7]. We relate the therapeutic relevance of AV for the above-observed effects against severe patho-phenotypes associated with the critical stage of COVID-19, characterized by severe lung in ammation, hypoxemia, angiogenesis, sepsis, and altered coagulation pro le [6,[8][9][10][11].
Therefore, the anti-hypoxic property of AV would be advantageous to attenuate the critical in ammatory stage of COVID-19. Indeed, our transcriptome results show that the innate immune system-related gene ontology terms and pathways critical to adaptive immune response are enriched in overexpressed genes of AV treatment, whereas same processes and pathways were under expressed in COVID-19 patients transcriptome data (Fig. 3A, S1). On the other hand, genes that are signi cantly under-expressed in response to AV treatment but were over-expressed in COVID-19 patient's transcriptome are enriched in functional annotations like collagen biosynthesis, angiogenesis, TGF-β regulation, brosis, hypoxia pathways, and platelet activation and aggregation (Fig. 3A, S1). These results coincide with our biochemical observations in mice and corroborate with results of our pilot-study. Based on these results, we speculate that the treatment with AV could not only help to induce the early-stage antiviral effect but also mitigates the late-stage complications or severe impact of the hypoxia-in ammation axis in COVID-19 subjects.
Furthermore, CMap analysis identi es the positive connectivity of AV with other potential drugs/compounds which recently predicated for SARS-CoV2 prevention such as Apicidin (Fig. 4A). The HDAC2 inhibitor, Apicidin, was predicted to interact with CoV nsp5. Therefore, we screened chemical components of AV identi ed through liquid chromatography against SARS-CoV2 and host target proteins (table S1, S2). We observed quinazoline as well as O and N glycosides e ciently binds with SARS-CoV2 spike protein (Spro), 3C-like proteinase (3CLpro, main protease, NSP5), papain-like proteinase (PLpro), RNA dependent RNA polymerase (RdRp), and other non-structural proteins (nsps) (table 1 and S3). We observed AV constituents interact with host protein such as ACE2, TMPRSS2, and JAK which promote the entry and further in ammatory consequence of virus infection (table 1 and S3). SARS-CoV-2 is a singlestranded RNA virus. More than half of the 5' genome virus encodes for two polyproteins (replicase), namely pp1a and pp1ab. These two proteins are cleaved into 16 nsps, including RdRp by two essential viral proteases, 3CLpro and PLpro. This process makes the formation of the replication complex of SARS-CoV-2. The 3' end of the genome encodes four essential structural proteins, spike (S), envelope (E), matrix/membrane (M), and nucleocapsid (N), along with a set of accessory proteins. SARS-CoV-2 gains entry into the host cell by binding the viral S protein (Spro) to the host ACE2 receptor. The S protein is cleaved into S1 and S2 by a cell-derived protease. S1 binds to ACE2, and S2 is activated by the host serine protease TMPRSS2 and results in membrane fusion. Once inside, SARS-CoV-2 hijacks the host machinery to transcribe, replicate, and translate its RNA genome and structural proteins before being reassembled, encapsulated, and exocytosed from the cell [33]. Thus, strong binding of AV constituents with the above virus and host proteins indicates its potential role in preventing viral entry, replication, and, therefore, its infection. Transcriptomic analysis also shows reduced expression of genes like IFITM1 and IFITM6, which plays important role viral defence.

Conclusion
Treatment of Adhatoda Vasica extract shows the profound effect on the immunity-in ammation axis in such a way that it would be relevant in the initial and later phases of SARS-CoV-2 infection. It shows the multi-dimensional effect at the genome, molecular, cellular, and phenotype level to mitigate the hypoxiahyperin ammation features (Fig. 5). Our study also highlights the likely impact of AV against hypoxia predominant lung in ammation, brosis, sepsis, and blood coagulation symptoms. Moreover in-silico analysis also predict AV role in preventing viral entry and replication, and there it would not only useful in later stages if COVID-19 but also useful in early stages of viral infection. Therefore, it would be worthwhile to test AV on COVID-19 patients. This is a unique study of its kind, wherein we could validate