This study involved experiments on AG129 mice, which were performed at the International Centre for Genetic Engineering and Biotechnology, New Delhi (ICGEB/IAEC/08/2016/RGP-15), in compliance with the Committee for the Purpose of Control and Supervision of Experiments on Animals guidelines issued by the Government of India.

The Vero cell line was purchased from the American Type Cell Culture, VA, United States. This monkey kidney cell line was maintained using Dulbecco’s Modified Eagle medium (DMEM) supplemented with 10% ΔFBS in a 10% CO₂ humidified incubator at 37°C. World Health Organization reference DENV strains DENV-1 (WP 74), DENV-2 (S16803), DENV-3 (CH53489), and DENV-4 (TVP-360) were received from Dr. Aravinda de Silva’s lab, University of North Carolina, United States. Anti-dengue 2H2 antibody was produced in-house from its hybridoma.

Organic solvents used in the plant extraction and high-performance liquid chromatography (HPLC) analysis were procured from E. Merck Ltd., Mumbai, India, either of analytical or HPLC grade. Before use, solvents were also filtered through a 0.45-μm membrane filter (Millipore, Billerica, MA, United States). The HPLC column RP18e Purospher-STAR (Hibar) (250 × 4.6 mm; 5μm) was used for chemical fingerprinting (E. Merck) and Eclipse 5 μm; 9.4 × 250 mm was used for purification of marker compounds (Agilent). The chemicals and reagents used for standardization and quality control were procured from Sigma-Aldrich, United States. Water for extraction and HPLC analysis was obtained from a high-purity Milli-Q Advantage A10 water purification system (Millipore, Molsheim, France).

The botanical raw material (BRM), i.e., aerial or stem parts of C. pareira and C. hirsutus, were collected by the botanist from Mandawara block, Lalitpur district, Uttar Pradesh, India. Identification of the collected BRM was performed at the Plant Science Division of Council of Scientific and Industrial Research—Indian Institute of Integrative Medicine (CSIR-IIIM), Jammu, India. Duly identified herbarium specimens of C. pareira (accession no. RRLH-23148) and C. hirsutus (accession no. RRLH-23152) were submitted to internationally recognized Janaki Ammal Herbarium at CSIR-IIIM, Jammu. Furthermore, for some studies, BRM of C. hirsutus collected from the same area was procured from the local vendor at Lalitpur. After critical macroscopic and microscopic examination, the botanical identity of the procured BRM samples of C. hirsutus was confirmed at the Plant Science Division of CSIR-IIIM. The duly identified samples of the procured BRM, i.e., aerial (accession nos. CDR-4037 and CDR-4038) and stem (accession nos. CDR-4061, CDR-4064, CDR-4065, and CDR-4078) parts of C. hirsutus, have been submitted to the Crude Drug Repository at CSIR-IIIM, Jammu.

BRM was dried in the shade and pulverized. The grounded plant masses of C. pareira and/or C. hirsutus were separately charged into an extractor at ambient temperature. The extracting solvent used were: (1) Hydro-alcohol, 50:50, (2) Denatured Spirit, and (3) Water. The extraction solvent was added to the pulverized BRM, and the mixture was heated to reflux temperature for about 3 h. The extracted mass was filtered, collected, and stored in a container. The extraction process was repeated from the residue plant mass two more times. All the three filtered extracts were combined, concentrated, and further dried either using rotor vapor, tray under vacuum, or spray drying. The yield of dry extract (drug substance) obtained was 5–14%, depending on the part of the plant used and the extracting solvent.

The phytopharmaceutical drug-based AQCH tablets 100/300/500 mg has been developed as an immediate-release formulation. The dry granulation method has been selected as the manufacturing process based on the hygroscopic nature of AQCH. The stability of AQCH tablets was extensively evaluated, and the drug product was found to be satisfactory for tested up to 6 months in accelerated conditions [40 ± 2°C/75 ± 5% relative humidity (RH) and for tested up to 18 months in long-term conditions (30 ± 2°C/65 ± 5% RH)] (Supplementary Table 3).

Vero cells were seeded in a 96-well plate (∼25,000 cells/well) in 200-μl DMEM + 10% ΔFBS and incubated for 24 h in an incubator adjusted at 37°C and 10% CO₂. Next day, cells were infected with DENV-1, -2, -3, and -4 at a multiplicity of infection of 0.1 (5,000 fluorescence-activated cell sorting infectious unit; FIU) in 100-μl DMEM + 0.5% ΔFBS. After a 2-h incubation of Vero cells with the virus at 37°C and 10% CO₂, virus infection media was aspirated, and a suitable range of extract/drug concentration to be tested was added to the wells in 200-μl DMEM + 0.5% ΔFBS in duplicates. Cells were incubated further for another 46 h in an incubator at 37°C and 10% CO₂. Wells infected with the virus without any subsequent treatment served as virus controls, whereas wells with no infection and no treatment served as cell controls. These experimental controls were utilized for relative virus infection calculations and antibody background signal adjustments, respectively. Post-incubation, cells were trypsinized, fixed using 4% paraformaldehyde, and stained with anti-prM Alexa-488-labeled 2H2 antibody as reported elsewhere (Kraus et al., 2007; Shukla et al., 2020b). Stained cells were assessed for virus infection using a flow cytometer (BD FACS Verse/Lyric), and data were analyzed through Flowjo software. Extract/drug concentration resulting in a 50% reduction in the number of DENV infected cells with reference to the virus control, which represents 100% infection, was defined as the 50% inhibitory concentration (IC₅₀) of the herbal extract/drug tested. All the extracts/drug/compounds prepared during this study were analyzed using this flow cytometry-based virus inhibition assay for assessing their anti-dengue potential.

The in vitro cell cytotoxic index (CC₅₀) of AQCH was evaluated through 3-(4, 5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide (MTT) assay. Vero cells were seeded as described in the flow-cytometry-based virus inhibition assay. Post-24-h incubation at 37°C and 10% CO₂, overlay media was removed, and 200 μl of a suitable concentration range of AQCH prepared in dilution media (DMEM + 0.5% ΔFBS) was added to the wells in duplicates; cells incubated with dilution media alone were kept as cell control and processed in parallel. Cells were incubated further for another 46 h in an incubator at 37°C, 10% CO₂. Post-incubation, 10 μl of 5 mg/ml MTT reagent (procured from Sigma Aldrich, United States) prepared in PBS was added and further incubated for 2 h at 37°C, 10% CO₂. Upon the formation of formazan crystals, the overlay was removed, and 100 μl of dimethyl sulfoxide was added. After the dissolution of crystals in dimethyl sulfoxide, absorbance was taken at 570 nm. The % cell cytotoxicity was calculated for each AQCH concentration with respect to cell control. The concentration of AQCH at which 50% cell cytotoxicity was observed is reported as CC₅₀.

All nuclear magnetic resonance (NMR) spectral data were recorded on a Bruker 400-MHz spectrometer. Chemical shifts (δ) were referenced internally to the residual solvent peak (CD3OD: ¹H δ 3.30, ¹³C δ 49.0 ppm; CDCl³: ¹H 7.26, ¹³C 77.0 ppm), and the reference point was tetramethylsilane (δH and δC: 0.00 ppm). High-resolution electrospray ionization mass spectrometry spectra were recorded on an Agilent 1100 LC-Q-TOF mass spectrometer and HRMS-6540-UHD machines. HPLC purifications were performed on Thermo Scientific Dionex UltiMate 3000 HPLC system with ultraviolet (UV) detector. Column chromatography was performed using silica gel (60–120 and 230–400 mesh); fractions were monitored by thin-layer chromatography (TLC) using pre-coated silica gel plates 60 F254 (Merck). Spots were visualized by UV light or by spraying with H₂SO₄-MeOH, anisaldehyde-H₂SO₄ reagents.

For the HPLC fingerprinting, 100 mg of AQCH (FCH 1909006) was transferred in a 20-ml volumetric flask, added approximately 10 ml of diluent and sonicated/shaken/stirred for 5–10 min to dissolve and made up the volume with diluent and mixed and filtered through a 0.45-m filter for HPLC fingerprinting. It was performed on RP18e Purospher-STAR (Hibar) (250 × 4.6 mm; 5μm) column. The mobile phase containing a buffer (0.1% formic acid in water) and acetonitrile was used at the flow rate of 0.65 ml/min at a column temperature of 30°C at 254-nm wavelength. The volume of injection was 5 μl, and the total run time of the assay was 75 min. A gradient program was used as follows: 0–15 min, 00–05% B; 15–40 min, 05–20% B; 40–55 min, 20–30% B; 55–65 min, 30–60% B; 65–68 min, 60–00% B and 68–75 min, 00% B. For the isolation, a portion of AQCH (500 gm) was suspended in distilled water and partitioned between ethyl acetate (A) and H₂O (B). The aqueous layer (B) was basified with NH₄OH solution (pH 9) and then extracted with chloroform, which resulted in a chloroform layer (4.0 g, C) and aqueous layer (D). The CHCl₃ layer (C) was further purified through repeated column chromatography in neutral alumina eluted with a gradient of CHCl₃-MeOH (100:0–0:100) to obtain Sinococuline (1) as a major constituent along with Magnoflorine (2). The aqueous layer (D) was lyophilized (480.5 g) and suspended in methanol. The methanol soluble portion (400.0 g) was purified by column chromatography (silica gel, 100–200 mesh), eluted with a gradient of CHCl₃-MeOH (100:0–0:100, 500-ml collected volumes of each fraction), and concentrated, giving 50 fractions (Fr.1–Fr.50), and their composition was monitored by TLC, with those showing similar TLC profiles grouped into six major fractions (Fr-1a to Fr-6a). Fraction Fr-2a afforded two UV active compounds as crystals. These crystals containing two UV active compounds were further subjected to semi-preparative C18 reversed-phase HPLC [Eclipse 5 μm; 9.4 × 250 mm; 3 ml/min; gradient of water (B)/acetonitrile (A) over 32 min; 100–80% B (5 min), 80–60% B (5 min), 60–50% B (5 min), 50–40% B (5 min), 40–20% B (5 min), 20–45% B (3 min), 45–70% B (2 min), and 70–100% B (2 min); column oven temperature 25°C] to give 20-Hydroxyecdysone (3) and Makisterone-A (4). The EtOAc soluble fraction (A) was subjected to column chromatography (silica gel, 100–200 mesh), eluted with a gradient of CHCl₃-MeOH (100:0–0:100, 250-ml collected volumes of each fraction) and concentrated, giving 30 fractions (Fr.51–Fr.80). Coniferyl alcohol (5) was obtained from fractions Fr. 66–Fr. 70 in pure. All the isolated compounds 1–5 were identified by detailed spectral analysis using one- and two-dimensional NMR and HRESI-MS data (Supplementary Figure 1) and compared with the reported spectral data (Ishii et al., 1984; Itokawa et al., 1987; Nakasone et al., 1996; Snogan et al., 2007; Taha-Salaime et al., 2019).

AQCH was formulated into tablets of different strengths (100, 300, and 500 mg) using the approved excipients. The accelerated and long-term stability of AQCH and AQCH tablets was assessed by exposing them to different conditions (30 ± 2°C/65 ± 5% RH and 40 ± 2°C/75 ± 5% RH). The in vitro anti-dengue activity and the content of a selected marker compound were evaluated for the stored samples at different time points (1, 2, 3, and 6 months).

A monoclonal anti-flavivirus antibody, 4G2, was produced in-house using hybridoma procured from the American Type Cell Culture. DENV-2 strain S221 used for in vivo studies was obtained from Global Vaccines Inc., NC, United States, and propagated using DMEM adapted C6/36 cell line for preparing virus stocks. The titer of these stocks was estimated using Vero cells in FIU per milliliter, as described elsewhere (Lambeth et al., 2005). AG129 mice were purchased from B&K Universal, United Kingdom, and bred in-house at the International Centre for Genetic Engineering and Biotechnology, New Delhi, for all animal experiments.

Primary DENV infection was established in a type-I and type-II interferon (IFN) receptor knockout, AG129 mice (Shresta et al., 2006; Sarathy et al., 2015) by challenging them with a lethal dose (1 × 10⁵ FIU; intravenous) of DENV-2 S221. Mice were orally gavaged with AQCH, 33 and 100 mg/kg/day (prepared in 0.1% w/v methylcellulose), for 5 days using twice a day (BID), three times a day (TID), and four times a day (QID) dosing regimens starting 30 min post-infection as reported elsewhere (Watanabe et al., 2012). The level of protection conferred by AQCH was monitored through mortality and morbidity scores for 15 days post-lethal challenge. The morbidity score was assessed based on a five-point system: 0.5, mild ruffled fur; 1.0, ruffled fur; 1.5, compromised eyes; 2, compromised eyes with hunched back; 2.5, loose stools; 3.0, limited movement; 3.5, no movement/hind leg paralysis; 4.0, euthanized if the cumulative score was 4; 5.0, assigned to the animal that was dead as reported previously (Ramasamy et al., 2018; Shukla et al., 2020c). The level of protection was scored through % survival for 15 days post-lethal challenge.

The secondary dengue AG129 mouse model was established through intravenous inoculation of immune complexes (IC) of sublethal dose of DENV-2 S221 (2 × 10⁴ FIU) made with neutralizing concentration of 4G2 mAb (10 μg) reported elsewhere (Watanabe et al., 2015; Shukla et al., 2020c, b). IC challenged mice (n = 6) were orally gavaged with different doses of AQCH, 33 and 100 mg/kg/day (prepared in 0.1% w/v methylcellulose), in the BID, TID, and QID dosing regimens for 5 days post-ICs inoculation. The level of protection conferred by AQCH was examined in a similar manner as in the primary dengue AG129 model.

Two groups of IC injected AG129 mice (n = 6) were orally fed with AQCH at 8.25 and 25 mg/kg/dose QID (equivalent to 33 and 100 mg/kg/day, respectively) for 4 days. Serum was collected from each group (n = 3) on day 4 for assessing viral RNA equivalents through real-time quantitative polymerase chain reaction, and post-sera collection vascular leakage was quantified in these mice using Evans blue dye, as detailed in the Supplementary Protocol 1. The remaining three mice from each group were killed on day 4 and small intestines harvested for assessing viral load and levels of tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6) cytokines as elaborated in the Supplementary Material.

Statistical differences between two groups of treated animals were calculated through two-way analysis of variance (with Bonferroni’s correction for multiple comparisons). The Kaplan–Meier survival curves were analyzed by the log-rank test for significance. p-values of ≤0.05 and <0.001 were considered significant and very significant, respectively. All statistical calculations were performed using GraphPad Prism (v8.0) software.

Guided by the Indian Ayurveda literature and our in vitro and in vivo bioassays, we had earlier identified and established that methanolic extract of the aerial parts of C. pareira possesses pan-anti-dengue activity (Sood et al., 2015). C. pareira belongs to the family Menispermaceae, which is historically known to be rich in a variety of alkaloids (Barbosa-Filho et al., 2000). In our previous study (Sood et al., 2015), a total of 19 plants were evaluated for their anti-dengue activity, two of which C. pareira and Tinospora cordifolia belonged to the family Menispermaceae, and both of them were found to possess anti-dengue activity. However, C. pareira, which belongs to the Cocculeae tribe of Menispermaceae, exhibited significantly stronger anti-dengue activity than T. cordifolia, which belongs to the Tinosporeae tribe.

In the present study, we have investigated the anti-dengue activity of plant extracts from aerial parts and stem from C. hirsutus, which has a morphologic resemblance to C. pareira but was found to be microscopically altered in their anatomical characters (Supplementary Table 1). We initiated work on C. hirsutus because of the finding that it is often present with C. pareira in dried collections made from a local vendor.

With methanolic extract of aerial parts of C. pareira as our benchmark, we prepared three batches each of methanolic extract of aerial parts of C. pareira and C. hirsutus in a manner similar to that stated in Sood et al. (2015). In this study, we evaluated all these six methanolic extracts for anti-dengue activity in an in vitro FNT instead of plaque-based bioassay used in Sood et al. (2015). The FNT and plaque-based bioassays are principally similar. However, FNT-based evaluation is advantageous because it is high-throughput and is more stringent, as it uses a higher dose of DENV for the evaluation of the anti-dengue activity. In the FNT assay used in the current study, the Vero cells were infected with 0.1 multiplicity of infection of each DENV serotypes, and post-infection the cells were incubated in media containing extract at various concentrations for 46 h. Post-incubation cells were fixed, permeabilized, and stained with Alexa fluor-labeled anti-dengue mAb, 2H2, reactive to all the four DENV serotypes, which were read in a flow cytometer to determine the percentage of DENV infected cells. This was used to calculate the extract concentration at which 50% of the DENV infection was inhibited (IC₅₀). Upon parallel evaluation of all the six extracts in FNT, it was observed that all the three batches of methanolic aerial C. hirsutus extracts possessed significantly stronger anti-dengue activity against all the four DENV serotypes as compared with methanolic aerial C. pareira extracts (Supplementary Figure 1).

Upon finding C. hirsutus possessing more potent anti-dengue activity as compared with C. pareira, we explored individual preparation of extracts of both aerial (Figure 1A; dashed curves) and stem portions (Figure 1A; solid curves) of C. hirsutus in various solvents such as denatured spirit, hydro-alcohol (50:50), and water. These extracts were evaluated against all the four DENV serotypes by FNT, and their IC₅₀ values were compared (Figure 1B). It was observed that the stem portion of C. hirsutus was significantly more potent than the aerial part, irrespective of the solvent used. Thus, aqueous extract of stem of C. hirsutus, hereafter referred to as AQCH, was advanced further due to regulatory compliance factors.

High-performance liquid chromatography (HPLC) was performed on the AQCH batch (ID: KL/DBE/002/18) post-confirmation of its pan-anti-dengue activity by FNT (Figure 2A) for its chemical profiling. Additionally, the extent of possible in vitro cytotoxicity caused to Vero cells by AQCH was also evaluated by MTT assay, and the CC₅₀ was determined to be ∼90 μg/ml (Figure 2A). The HPLC chromatogram obtained is shown in Figure 2B. This was followed by isolation of five marker compounds using repeated chromatographic methods and characterized using advanced one- and two-dimensional NMR spectroscopic and mass analysis. Marker compounds were identified to be Sinococuline (1), Magnoflorine (2), 20-Hydroxyecdysone (3), Makisterone-A (4), and Coniferyl alcohol (5) (Figure 2C). The physicochemical data of all the five identified marker compounds are given in Supplementary Table 2.

The robustness and consistency of AQCH were monitored by the preparation of various batches of AQCH extract utilizing one of the three drying methods—rotar vapor drying, vacuum tray drying, and spray drying. Irrespective of the method used for drying, the in vitro anti-dengue activity of all the extract batches prepared was comparable (Supplementary Figure 2A). The HPLC chromatograms of three batches corresponding to the three drying methods were observed to be overlapping (Supplementary Figure 2B), with a high degree of consistency in retention times of the five marker compounds (Supplementary Figure 2C). This indicates that the AQCH extract preparation method is consistent and robust, and the choice of drying method does not have any implication on its chemical profiling and biological activity. Thus, spray drying was considered as the method of choice, as it resulted in the formation of free-flowing finer extract in a shorter span of time, which is industrially more compatible. Furthermore, the identified compound markers were assessed for their pan anti-dengue inhibitory potency by FNT (Figure 3), and only Sinococuline was found to have anti-dengue inhibitory activity against all four DENV serotypes (Figure 3A). The other chemical entities characterized in the drug substance, Magnoflorine, 20-Hydroxyecdysone, Makisterone-A, and Coniferyl alcohol, showed almost negligible anti-dengue activity (Figures 3B–E).

To assess the in vivo AQCH efficacy, we tested AQCH in dengue-sensitive AG129 mouse model. AG129 mice are IFN type I and II receptor-deficient mice that allow viral infection to propagate and develop dengue viral symptoms. To develop the DENV disease symptoms in AG129 mice, we optimized the mouse-adapted DENV-2 strain S221 dose, which could develop dengue severe disease symptoms in the absence of any antiviral treatment. Briefly, AG129 mice (n = 6) were challenged with a lethal dose (1.0 × 10⁵ FIU) of DENV-2 S221, which killed all the mice by days 6–7. We utilized this primary DENV infection model for the assessment of AQCH efficacy in vivo. In this experiment, we tested AQCH in three different dosing regimens, i.e., BID (orange curves), TID (purple curves), and QID (blue curve) in Figure 4, at two different doses of 100 mg/kg/day (solid curves) and 33 mg/kg/day (dashed curves) (Figure 4A). It was observed that BID and TID regimens too conferred similar protection, 100% survival at 100 and ∼33% at 33 mg/kg/day, as fed QID for 5 days post-infection. Thus, there was no significant difference in the survival scores of mice between different dosing regimens at either of the two AQCH doses (p > 0.9999). However, the morbidity and weight loss scores of mice were moderately higher under the BID regimen (p = 0.0019) than TID (p = 0.0295) and QID for each of the two doses (Figures 4C,D). Outcomes with control mice groups were observed as expected as 0% survival in the unfed infected group “V” (green curve) and 100% survival in AQCH-fed uninfected group “Only AQCH” (black curve). Among the two doses, significantly lower protection and higher morbidity were observed at 33 mg/kg/day AQCH treatment in the BID and TID dosing regimens (p = 0.01).

Secondary dengue infection is one of the worst scenarios in dengue infection and develops very severe dengue disease through antibody-dependent enhancement (ADE) pathway (Halstead et al., 2010; Guzman et al., 2013). Thus, we investigated the therapeutic potential of AQCH against severe DENV infection using the secondary dengue AG129 mouse model. The IC of DENV-2 S221 were made in vitro by mixing 100% neutralizing concentration of monoclonal antibody 4G2 (10 μg). The 4G2 mAb potentially binds the fusion loop of all four DENV serotypes and tends to enhance DENV infection by following Fcγ receptor-mediated intrinsic ADE, which is manifested as severe dengue infections (DHF/DSS) as evidenced elsewhere (Halstead et al., 2010; Shukla et al., 2020a). We used this ADE model to evaluate AQCH efficacy in vivo. The IC inoculated mice were fed orally with previously primary dengue disease model tested doses of AQCH, viz., 33 (dashed curves) or 100 mg/kg/day (solid curves), using BID (orange curves), TID (purple curves), and QID (blue curves) dosing regimens for 5 days post-IC inoculation (Figure 5). The outcomes with control groups, viz., “Un-infected” (magenta curve), “V” (green curve), “Only AQCH” (black curve), and “IC” (gray curves), were as expected. Unlike the case in the primary dengue AG129 model, the efficacy of AQCH was not found to be similar across the three evaluated dosing regimens. QID was the most effective dosing regimen followed by TID and BID as 100%, 50%, and 0% mice survival rates were observed, respectively, with AQCH treatment at 100 mg/kg/day dose (Figure 5B). Since protection was not observed at 100 mg/kg/day AQCH dose in BID regimen, a lower AQCH dose of 33 mg/kg/day was evaluated only in TID and QID regimens, where it exhibited ∼33% and 50% protection, respectively (Figure 5B), which suggests that BID and TID dosing regimens are not appropriate regimens for treating severely infected AG129 mice. Also, morbidity scores and body weight changes correlated well across BID, TID, and QID regimens at 100 mg/kg/day with the observed survival efficacies (Figures 5C,D).

Severe dengue disease is known to be associated with small intestinal pathologies of enhanced viral load, increased levels of pro-inflammatory cytokines, and vascular permeability (Vejchapipat et al., 2006; Watanabe et al., 2018). Thus, it was of interest to determine the impact of AQCH treatment on small intestinal pathology, which could be examined in secondary dengue disease AG129 mice exhibiting ADE-mediated severe dengue.

It has been reported that small intestinal pathology is a better correlate of protection than serum viremia levels in dengue disease (Watanabe et al., 2018). Thus, we evaluated the effect of AQCH treatment on both the serum viremia and small intestinal viral load. Secondary dengue disease AG129 mice (n = 3) were orally fed with AQCH (8.25 and 25 mg/kg/dose, QID) for 4 days. Mice injected with IC showed symptoms of disease onset such as lethargy, whereas the mice treated with 25 mg/kg/dose QID of AQCH showed no disease symptoms (Supplementary Media File). Viremia levels were determined in serum and small intestinal tissue samples collected on day 4. Mice treatment with AQCH significantly reduced (p < 0.0001) the virus load in both the serum (Figure 6A) and small intestine (Figure 6B). However, unlike the observation with small intestinal viral load, reduction in serum viremia was not found to be dose-dependent, as QID treatment with 8.25 mg/kg/dose (blue empty bar) of AQCH led to a higher reduction in serum viremia than 25 mg/kg/dose, blue solid bar (Figure 6A).

Since vascular leakage is a prominent manifestation of severe dengue infections (DHF/DSS), we assessed intestinal vascular leakage using Evans blue dye, which serves as a marker for albumin extravasation. The leakage was quantified through colorimetric estimation of the dye amount absorbed, which was represented as a fold increase in OD₆₂₀ per gm of tissue’s wet weight in comparison with the “Un-infected” control. AQCH treatment was observed to significantly (p = 0.0015) reduce the vascular leakage at both the dose levels—25 and 8.25 mg/kg/dose QID (Figure 7A, solid and empty blue bars, respectively). Enhanced vascular leakage was observed in the “IC” mice group (Figure 7A, gray bar), as expected. The visual examination of day 4 small intestines of mice also reflected similar findings with distinctive fluid accumulation in the “IC” group, whereas the small intestines of other groups—“V,” “IC + 25 mg/kg/dose AQCH QID,” and “IC + 8.25 mg/kg/dose AQCH QID”—appeared normal and similar to the “Only AQCH” group (Figure 7B).

The manifestation of severe dengue infections occurs during the critical phase of the infection, and levels of pro-inflammatory cytokines, such as TNF-α and IL-6, become elevated, which leads to further disease progression and causes DHF/DSS. To assess the anti-inflammatory activity of AQCH, we evaluated the expression levels of pro-inflammatory cytokines, TNF-α (Figure 7C), and IL-6 (Figure 7D), in small intestines of mice. The results obtained were in correlation with the findings of vascular leakage examination. Treatment with AQCH inhibited the secretion of cytokines in the small intestine in a dose-dependent manner. The “IC” group animals showed the highest expression levels of cytokines (TNF-α and IL-6), which was expected due to 4G2-mediated ADE in secondary dengue infection AG129 mice (Figures 7C,D).

With the finding of strong and consistent in vitro and in vivo anti-dengue potency of AQCH, we formulated the drug substance into 100-, 300-, and 500-mg strengths of AQCH tablets (drug product). The tablets were subjected to accelerated and long-term stability studies along with the AQCH drug substance. Samples from the stability study were analyzed for Magnoflorine content and in vitro anti-DENV-2 activity by FNT to evaluate any deterioration or loss in bioactivity upon storage under the conditions stated. It was found that there was neither a significant change in the Magnoflorine content (Supplementary Table 3) nor any loss of anti-DENV-2 activity (Table 1) of AQCH and AQCH tablets of all the three strengths (100, 300, and 500 mg) under the condition tested up to 6 months. The long-term stability study is ongoing, and samples will be evaluated for up to 3 years of storage.