Inhibition of dengue virus type 2 by hypericin mediated through viral envelope protein interaction

Abstract

Background

Dengue fever continues to exert significant global impact, affecting populations worldwide with considerable public health and economic consequences. There is no antiviral drug for dengue. This study focuses on hypericin, a naturally occurring compound from Hypericum perforatum L. whose anti-dengue properties have been underexplored. We systematically examined its antiviral efficacy against dengue virus (DENV), revealing strong inhibitory effects and clarifying its precise antiviral mechanism.

Methods

The study assessed the efficacy of hypericin against DENV using various scientific methods like plaque assays and Western blotting. We looked into its antiviral mechanism. We used a time-of-addition approach during our research. Moreover, the basic mechanisms involved were studied through molecular docking, surface plasmon resonance (SPR), and co-immunoprecipitation (Co-IP).

Results

This study demonstrated that hypericin exhibits broad-spectrum antiviral activity against DENV-2 in cell lines derived from multiple species. In time-of-addition experiments, it showed inhibitory effects under co-treatment, direct virucidal, and post-treatment conditions. Crucially, hypericin primarily blocked viral attachment and entry stages, thereby effectively reducing intracellular viral load. Mechanistic investigations revealed a interaction between hypericin and the E protein, evidenced by a computational docking score of -7.0 kcal/mol and an experimental SPR-derived Kd of 7.18 µM. Furthermore, Co-IP assays demonstrated that hypericin competitively blocks the association between the E protein and its cellular receptor, HSP70.

Conclusion

As per these findings, the E protein was seen to be a target of hypericin with an antiviral activity against DENV-2 at multiple stages by limiting viral adsorption and viral entry projecting a molecular basis for the candidate molecule as a possible anti-dengue agent.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12985-026-03087-4.

Introduction

DENV from the Orthoflavivirus genus and the Flaviviridae family has a positive-sense ssRNA genome. The dengue virus genome is 11 kb long and codes for ten viral proteins. These are classified into three structural elements which are C, prM and E essential for virion assembly and seven non-structural species, NS1–NS5 mainly involved in replication [1, 2]. The dengue virus consists of four distinct serotypes, identified as DENV-1 through DENV-4. Among these, DENV-2 has become the most common model in the studies of antiviral agents [3, 4]. DENV is a mosquito-borne pathogen that is mainly transmitted by Aedes aegypti and Aedes albopictus. DENV causes dengue fever in humans. The clinical signs and symptoms may develop to situations that endanger life such as dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS), which have an increased risk during secondary heterotypic infection [5–7]. The E protein, which plays a fundamental role in dengue virus attachment and fusion, has long been recognized as a promising target for antiviral development [8–10]. This research urgency is compounded by the escalating worldwide dengue incidence, which carries substantial morbidity and economic costs [11]. Present control methods, focusing on mosquito management and community education, fail to provide adequate containment. Although two vaccines are available, their utilization is hampered by restricted indications and adverse effect risks, and the TV003 vaccine is still in development [12, 13]. With no approved dengue-specific therapeutics, clinical management remains supportive. Preliminary clinical studies of JNJ-1802 and NITD-688 offer some promise [14, 15], yet conclusive evidence is pending. Therefore, creating effective and safe pharmacological options for dengue is an imperative public health priority.

Plant-derived phytochemicals such as polyphenols, tannins, terpenoids, alkaloids, etc. can lead to development of novel anti-dengue drugs which has bioactive properties [16]. Hypericin, a secondary metabolite isolated from Hypericum perforatum L., exhibits a broad spectrum of pharmacological activities [17], such as anticancer [18], antiviral [19], antimicrobial [20], anti-inflammatory [21, 22] and antidepressant [23] effects. The outbreak of the COVID-19 pandemic in late 2019 has spurred work on the antiviral potential of hypericin. Studies indicate that it stops virus from spreading through a number of ways. For example, it binds to SARS-CoV-2 Mpro/3CLpro to stop the virus from replicating [24]. It also stops the virus from entering the cell through binding to structural proteins [25]. Furthermore, it stops transcription of the virus by blocking viruses [26]. While some mechanisms need to be validated further, hypericin is a potential agent against SARS-CoV-2 [27]. It can also inhibit reverse transcriptase and interfere with the assembly and release of the virus (HIV) [28]. Research has shown that Although, the clinical trial outcome was inconsistent because of different raw material sources, the anti-HIV mechanisms of this product may still be scientifically useful [29]. Hypericin also demonstrated antiviral activity against hepatitis C virus, HCV [19], herpes simplex virus type 1, HSV-1 [30] and infectious bronchitis virus, IBV [31], primarily through suppression of key viral enzymes or interference with viral adsorption and penetration. Although these studies have elucidated hypericin’s broad-spectrum antiviral properties, its specific activity and mode of action against dengue virus remain uncharacterized, constituting a critical gap in current research.

Given the current lack of specific drugs for dengue fever, identifying antiviral lead compounds from natural sources holds significant importance. This study focuses on hypericin, a natural compound with diverse biological activities, aiming to evaluate its inhibitory effect against dengue virus serotype 2 at the cellular level, determine its effective time window, and preliminarily reveal its molecular targets and mechanism of action.

Materials and methods

Compounds and Recombinant proteins

The Hypericin analysis reference substance was provided by Macklin Reagent (Shanghai, China), dissolved in dimethyl sulfoxide (DMSO, 0.1%), and kept at -20 °C. The recombinant E DIII protein was kindly given by Professor Zhengling Shang from Guizhou Medical University, China.

Cell lines and viruses

Multiple cell lines were routinely maintained in our facility. C6/36 cells (from Aedes albopictus larvae) were cultured in Roswell Park Memorial Institute 1640 medium (RPMI-1640; GIBCO, USA) with 10% fetal bovine serum (FBS; Procell, China) at 28 °C. In parallel, BHK-21, A549, and Vero cells were grown in Dulbecco’s Modified Eagle’s Medium (DMEM; GIBCO, USA), likewise containing 10% FBS, but at 37 °C in a 5% CO₂ atmosphere. The DENV-2 New Guinea C strain used in this work was amplified in C6/36 cells, then stored at -80 °C to establish a working virus stock.

Cytotoxicity assay

The potential cytotoxic effect of hypericin was assessed in BHK-21, A549, and VERO cells via the CCK-8 method. After establishing a seeding density of 10,000 cells per well in 96-well plates, the cultures were allowed to adhere over 24 h (h) under standard conditions (37 °C, 5% CO₂). A 48 h exposure to varying concentrations of the compound ensued. Subsequently, 10 µl of CCK-8 reagent was introduced into each well, followed by a 30-minute incubation period. Finally, a microplate reader was employed to record the absorbance at 450 nm, from which viability percentages were calculated.

Viral plaque assay

BHK-21 cells were plated in 12-well plates at 4 × 10⁵ cells per well for DENV-2 viral titer determination. After cell adherence, a 10-fold serial dilution of the virus-containing supernatant was made and applied to triplicate wells. Infection was carried out at 37 °C with 5% CO₂ for 2 h to allow viral adsorption. Cells were then cultured for 7 days under the same conditions with a 1.4% agarose overlay in DMEM supplemented with 2% FBS. After fixation, the cells were stained with 1% crystal violet for 1 h, after which plaques were counted and the viral titer was determined in terms of plaque-forming units per milliliter (PFU/ml).

Cytopathic effect assay

BHK-21 cells were seeded in 12-well plates at 2 × 10⁵ cells per well and allowed to incubate overnight, providing sufficient time for them to develop into a confluent monolayer. Following monolayer formation, cells were infected with DENV-2 at a multiplicity of infection (MOI) of 0.01 and incubated for 2 h. Both uninfected (blank control) and virus-infected (positive control) groups were maintained under identical culture conditions at 37 °C with 5% CO₂. Cell morphology was observed daily by light microscopy. When distinct morphological changes became evident between control and infected groups, representative images were captured using an inverted microscope for documentation.

In vitro antiviral detection of hypericin by plaque assay

The antiviral effect of hypericin against DENV-2 was assessed in BHK-21 cells by plaque assay. Following infection with DENV-2 (MOI = 0.01) and a 2 h adsorption period, cells were treated with various concentrations of hypericin. Blank control (uninfected) and virus control (infected, untreated) groups were included for reference. After 48 h of incubation under consistent conditions, culture supernatants were harvested for viral titer quantification using plaque assay. The half-maximal effective concentration (EC₅₀) was derived from dose-response curves analyzed by nonlinear regression in GraphPad Prism.

RNA extraction and real-time fluorescence quantitative PCR (qRT-PCR)

BHK-21, A549, and VERO cells were infected with DENV-2 (MOI = 0.01), a low level chosen due to the high sensitivity of qPCR which enables effective detection even at low viral loads. Following cytopathic effect assay protocols, the infected cells with or without hypericin treatment were harvested at 48 h post-infection (hpi) for total RNA extraction using Trizol reagent. The extracted RNA was reverse-transcribed into cDNA and subsequently subjected to qRT-PCR analysis on a 96-well real-time PCR system using DENV-2-specific primers. The primer sequences used were:

DENV-2 E forward: CAGTCGGAAATGACACAG; DENV-2 E reverse: GCAACCATCTCATTGA; Gold hamster actin forward: CCACCATGTACCCAGGCATT; Gold hamster actin freverse: ACTCCTGCTTGCTGATCCAC; Human GAPDH forward: GTCTTCACCACATGGAGAA; Human GAPDH reverse: ATGGCATGGACTGTGGTCAT; Monkey GAPDH forward: TTGCATCGCCAGCGCATC; Monkey GAPDH reverse: TCGCCCCACTTGATTTTGGA.

Immunofluorescence assay (IFA)

An immunofluorescence assay was performed to analyze DENV protein expression and localization in hypericin-treated infected cells. Given that immunofluorescence has a relatively limited detection capacity for low-abundance proteins, a higher MOI of 0.5 was used for infection to ensure viral protein expression reached detectable levels. Specifically, BHK-21 cells grown in confocal dishes were infected with DENV-2 (MOI = 0.5) for 2 h. After washing, cells were fixed, permeabilized, and blocked, then incubated overnight with primary antibodies against DENV E protein (Invitrogen, MA5-17291, GT643, 1:1000) and HSP70 (Proteintech, 10995-1-AP, 1:1000). Specific binding was detected using fluorophore-conjugated secondary antibodies, with nuclei counterstained by DAPI (Solarbio, C0065). Images were acquired on an OLYMPUS spinning-disk confocal super-resolution microscope.

Western blot (WB)

Following infection with DENV-2 at a MOI of 0.5 for a duration of 2 h, BHK-21, A549, and VERO cell lines were treated with hypericin and then harvested 48 hpi. The experiment employed a MOI of 0.5 and varying concentrations of hypericin (5–10µM) to ensure adequate viral protein expression, with changes in protein abundance detected via Western blot. Total protein extracts were obtained using RIPA lysis buffer (Boster) and quantified via a BCA assay. The proteins were then separated using SDS-PAGE and subsequently transferred to PVDF membranes. Following the blocking step, these membranes were incubated with DENV-2 E protein (GeneTex, GTX127277, 1:5000) and β-actin (HUABIO, HA722023, PSH03-63, 1:50000) primary antibodies, where the β-actin serves as loading control. After washing with TBST thoroughly, the membranes were incubated with a DyLight 800-conjugated secondary antibody (Invitrogen, SA5-35571, 1:5000) at room temperature. The Odyssey Clx imaging system helped visualization of the protein bands.

Drug addition assay

The anti-DENV-2 activity of hypericin was evaluated in a BHK-21 cells. Four distinct administration strategies were implemented to assess its effects: pre-treatment, co-treatment, direct treatment, and post-infection treatment [32]. Cells at 2 × 10⁵ per well in 12-well plates were cultured overnight until confluence, followed by medium removal and Phosphate buffered saline(PBS)washing. (a) Pre-treatment: cells were initially pretreated with hypericin for 2 h at 37 °C in a 5% CO₂ environment. After removal of the compound, the cells were infected with DENV-2 at a MOI of 0.01 for 2 h; (b) co-treatment: equal volumes of the virus and hypericin were mixed and immediately added to the cells for a 2 h incubation; (c) direct-treatment: the virus and hypericin were pre-incubated at 4 °C for 2 h before being added to the cells for a further 2 h incubation; (d) post-treatment: cells were first infected with DENV-2 for 2 h, followed by treatment with medium containing hypericin. All groups were cultured with DMEM containing 2% FBS and 1% antibiotics post-infection and drug treatment, including blank and virus controls. After 48 h at 37℃ and 5% CO₂, supernatants and cellular RNA were collected for plaque assays and qRT-PCR.

Attachment and entry assay

In the attachment assay, DENV-2 was co-incubated with Hypericin and pre-cooled cells at 4 °C for 2 h. For the entry assay, cells were first allowed to adsorb DENV-2 at 4 °C, followed by Hypericin introduction and incubation at 37 °C. Supernatants and total RNA were collected 48 h post-treatment for quantification of viral replication via plaque assay and qRT-PCR [33].

In vitro assessment of virucidal activity

The virucidal effect of hypericin was assessed by pre-incubating DENV-2 with varying concentrations of the compound at 4 °C for 2 h. The mixture was then diluted and applied to cell monolayers for viral adsorption. After incubation, cells were overlaid with methylcellulose-containing DMEM supplemented with FBS and antibiotics, followed by 7 days of culture. Plaques were enumerated post-fixation and staining [34].

Molecular Docking

Molecular docking analysis of hypericin with the DENV-2 envelope protein was performed using AutoDock Vina. The DENV-2 E protein crystal structure (PDB ID: 1OKE) was obtained from the Protein Data Bank and prepared with AutoDock Tools. Structural preparation involved water removal, hydrogen atom addition, receptor designation, and format conversion to pdbqt. Hypericin was searched for in TCMSP, and the corresponding mol2 file was downloaded. It was then preprocessed in AutoDock Tool, with hydrogen atoms added, assigned as a ligand, and saved as a pdbqt file. Protein-ligand interactions were analyzed, and binding scores were calculated. PyMOL (version 2.6.0) was used to visualize the docking structure of hypericin and the E protein.

Surface plasmon resonance (SPR) analysis

Binding of hypericin to DENV-2 E DIII was analyzed by surface plasmon resonance (SPR) using a Biacore 8 K system (GE Healthcare, USA). The DENV-2 E DIII domain was immobilized via GST capture on a CM5 sensor chip (Cytiva, USA) previously activated with EDC/NHS. Anti-GST antibody was covalently coupled to the surface, followed by ethanolamine blocking. A 500 µg/mL ligand solution prepared in sodium acetate buffer was immobilized on the sensor chip at 20 °C. Hypericin samples serially diluted in PBS were injected across the flow cells at 30 µL/min with 150-second contact time. Between analyte cycles, the sensor surface was regenerated using 10 mM glycine-HCl solution (pH 2.0). Experimental binding data were processed through global fitting to a 1:1 Langmuir interaction model implemented in Biacore Insight evaluation software.

Co-immunoprecipitation (Co-IP) assay

DENV-2-infected BHK-21 cells treated with Hypericin were lysed, and protein lysates were incubated with HSP70 antibody (Proteintech, 10995-1-AP, 2 µg) or normal mouse IgG-coated magnetic beads (Beyotime, P2106-1 ml). The immunoprecipitates were washed, eluted, then analyzed by Western blot using specific antibodies against DENV-2 E protein (GeneTex, GTX127277, 1:5000) and β-actin (HUABIO, HA722023, PSH03-63, 1:50000).

Statistical analyses

Statistical evaluations for this study were carried out in GraphPad Prism 8.0. Data from three separate replicates are expressed as means ± standard deviation. When comparing two experimental groups, the two-tailed Student’s t-test was used, whereas one-way ANOVA coupled with Dunnett’s post hoc analysis served for multi-group comparisons. Significance levels were denoted as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.

Results

Effects of hypericin on intracellular replication of DENV-2 infected cells

The cytotoxicity of hypericin was assessed in BHK-21 cells, with CCK-8 assay indicating a CC₅₀ of 29.85 µM (Fig. 1B), thus defining a safe concentration range for antiviral testing. Microscopic analysis showed that 0.8 µM hypericin significantly reduced DENV-2-induced cytopathic effects, with treated cells maintaining morphology similar to uninfected controls (Fig. 1C). Plaque assays confirmed dose-dependent antiviral activity, where 0.4 µM hypericin produced inhibition equivalent to 16 µM ribavirin (Fig. 1D), demonstrating effective viral suppression at sub-micromolar concentrations.

Fig. 1.

Hypericin reduces DENV-2-induced cytopathic effects and viral replication. (A) Chemical structure of hypericin. (B) Cytotoxicity evaluation of hypericin in BHK-21 cells by CCK-8 assay. (C) Suppressive effect of hypericin on DENV-2 (MOI = 0.01)-induced cytopathic changes in BHK-21 cells after 5 days of infection. (D) Inhibition of DENV-2 Virus particle release by hypericin as detected by the plaque assay. Data are representative of three independent experiments (n = 3)

Hypericin reduces RNA and protein expression levels of DENV-2

To determine the inhibitory effect of hypericin on DENV-2, assays for viral RNA (qRT-PCR), protein expression (WB), and infection visualization (IFA) were conducted in BHK-21 cells. Exposure to escalating concentrations of hypericin for 48 h led to a stepwise decline in both E protein mRNA transcripts and E protein synthesis (Fig. 2A-B). The antiviral effect was visually corroborated by IFA, where hypericin-treated cultures exhibited substantially weaker fluorescence compared to the intense signal in untreated, infected controls (Fig. 2C). Together, these results confirm the dose-dependent inhibition of DENV-2 replication by hypericin.

Fig. 2.

The expression of DENV-2 is decreased by hypericin. (A) qRT-PCR analysis measuring DENV-2 E protein mRNA levels following infection at MOI = 0.01. (B) Western blot assessment of DENV-2 E protein expression in infected cells (MOI = 0.5). (C) Immunofluorescence visualization of DENV-2 E protein localization in BHK-21 cells infected at MOI = 0.5. ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001 compared to virus control. Data are from three independent experiments (n = 3)

Hypericin exhibits similar inhibitory effects on DENV-2 in other cell lines

To further investigate whether hypericin exhibits antiviral activity in other cell lines, our study extends this to A549 and Vero cells, where CCK-8 assays show CC₅₀ values of 43.66 ± 0.69 µM and 49.18 ± 1.21 µM (Fig. 3A, B). In addition, hypericin treatment significantly inhibited the DENV-2 E mRNA and E protein levels in A549 cells and Vero cells (Fig. 3C-F), and this inhibition showed a dose-dependent effect. The above results suggest that the antiviral effect of hypericin in inhibiting DENV-2 replication is cell line broad-spectrum and not limited to BHK-21 cells.

Fig. 3.

Hypericin exhibits antiviral activity against other cell lines. (A, B) CCK-8 assay for evaluating the toxicity of hypericin on A549 and Vero cells. (C, D) qRT-PCR detection of DENV-2 E protein mRNA expression in A549 and Vero cells. (E, F) Expression levels of DENV-2 envelope protein in A549 and Vero cell lines were evaluated by Western blot analysis. ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001 compared to virus control. Data are from three independent experiments (n = 3)

The antiviral mode of hypericin

To elucidate the antiviral mechanism of hypericin, we conducted pretreatment, co-treatment, direct treatment, and post-treatment experiments. Pre-treatment aimed to verify whether it inhibits viral attachment and endocytosis by acting on host cell targets (such as surface receptors). Co-treatment and direct treatment were used to determine whether it acts on the virus itself to block viral-cell binding. A post-treatment approach was designed to investigate the compound’s ability to suppress viral replication and safeguard cellular integrity following infection. This strategy helps clarify its potential mechanism during the post-entry phases of the viral life cycle. Significant inhibition was observed at co-treatment, direct-treatment, and post-treatment stages, with the strongest effects during co-treatment and direct-treatment (Fig. 4B, C). In contrast, cell pre-treatment showed no inhibition, indicating that Hypericin does not act via pre-conditioning host cells. These results demonstrate multi-stage inhibition and direct antiviral efficacy.

Fig. 4.

The inhibitory effect of time-course administration of hypericin on DENV-2 in BHK-21 cells. (A) Schematic Diagram of Four Drug Administration Methods. (B) qRT-PCR detection of DENV-2 E protein mRNA expression (MOI = 0.01). (C) The inhibitory effect of time-release administration of hypericin on the release of DENV-2 viral particles. ns: no statistically significant difference; *P < 0.05, **P < 0.01, ***P < 0.001 vs. virus-only control group. Data are from three independent experiments (n = 3)

Hypericin inhibits adsorption and entry of DENV-2 and directly inactivates DENV-2

To investigate the effects of hypericin on the early stages of DENV-2 infection, we conducted anti-attachment and anti-entry assays. In the anti-attachment assay, hypericin was incubated with the virus at 4 °C for 2 h, during which the virus could adsorb to the cell surface but was unable to enter. In the anti-entry assay, viruses were first adsorbed onto cells at 4 °C. After removing unbound viruses, hypericin was added at 37 °C to specifically interfere with viral entry. qRT-PCR results demonstrated that hypericin significantly reduced intracellular viral E protein mRNA levels in a concentration-dependent manner under both experimental conditions (Fig. 5A). Further plaque assays demonstrated a decreasing trend in observed viral particle counts with increasing hypericin concentrations (Fig. 5B). Collectively, these results confirm that hypericin effectively inhibits the initial attachment of DENV-2 to host cells and subsequently blocks its entry process.

Fig. 5.

Hypericin inhibits the adsorption and entry of DENV-2 and directly inactivates the virus. (A) qRT-PCR detection of DENV-2 E protein mRNA expression (MOI = 0.01). (B) Distribution of DENV-2 plaques in anti-attachment and anti-entry assay. Data are from three independent experiments (n = 3)

Binding of hypericin to DENV-2 E protein

Since hypericin directly inhibits DENV-2 adsorption and entry into host cells—as demonstrated in our time-of-addition and viral attachment/entry inhibition assays—and given that the E protein mediates these early steps, we investigated its binding to the DENV E protein. Extensive structural and functional studies have established the ED-III domain as the core region for receptor binding, making it the most direct and plausible target to explain the observed antiviral phenotype [35, 36]. Molecular docking revealed a low binding energy (− 7.0 kcal/mol) between hypericin and the DENV-2 E protein, suggesting stable binding (Fig. 6A). Subsequently, surface plasmon resonance (SPR) confirmed concentration-dependent binding to ED-III, with a dissociation constant of 7.18 µM (Fig. 6B). Together, these results demonstrate a interaction between hypericin and the E protein, thereby elucidating the molecular mechanism underlying its inhibition of viral attachment and entry.

Fig. 6.

Binding of hypericin to DENV-2 E protein. (A) Molecular docking between E protein and hypericin (hydrogen bonds shown as blue dashed lines). (B) SPR analysis determined the binding kinetics between E DIII protein and hypericin

Binding of hypericin to DENV-2 E protein

Given that HSP70 interacts with the dengue E protein to facilitate viral entry [37]. Colocalization of the E protein and HSP70 was observed in the cytoplasm of BHK-21 cells following DENV-2 infection (Fig. 7A), providing further evidence of their direct interaction. Next, we examined hypericin’s effect using Co-IP. Results showed markedly reduced E protein levels in immunoprecipitated complexes from hypericin-treated samples (Fig. 7B). These findings indicate that hypericin can bind to the E protein of DENV-2 and inhibit the interaction between the E protein and the receptor protein HSP70.

Fig. 7.

Hypericin disrupts the interaction of E protein with HSP70. (A) The relationship between E and HSP70 as determined by immunofluorescence analysis. (B) Detection of the relationship between E protein and HSP70 via Co-IP in the presence or absence of hypericin. Data are from three independent experiments (n = 3)

Discussion

DENV is a major arthropod-borne pathogen responsible for hundreds millions of infections annually and a significant global health burden. DHF and DSS has high mortality and has been presenting clinical challenges [38]. At present, the therapeutic landscape lacks specific antiviral drugs, leaving clinical care reliant on supportive measures that cannot directly inhibit viral reproduction or halt the advancement of the disease. Although several vaccines have been licensed, their widespread application is constrained by concerns related to antibody-dependent enhancement (ADE) [39]. To meet the clinical need we identified hypericin as a potent anti-DENV compound with a novel and multi-target mechanism. When compared to other available small-molecule inhibitors directed toward single stages of the viral life cycle, hypericin displays pharmacological characteristics that could allow it to be positioned as a suitable treatment option.

This study reveals for the first time that hypericin exhibits a dual-phase antiviral mechanism involving early blockade and late-stage inhibition. On the one hand, hypericin directly targets the conserved viral E protein, impairing viral attachment and entry. Notably, although no E‑protein-targeted inhibitors have yet been approved for clinical use against dengue, computational comparisons with reported E‑protein binders are informative. Under identical simulation conditions, hypericin demonstrates a binding free energy (− 7.0 kcal/mol) comparable to that of the known inhibitor α‑Mangostin [40] and significantly stronger than that of weaker binders such as Styrylpyrone Derivative (− 1.86 kcal/mol) [8]. This suggests that hypericin possesses E‑protein binding potential on par with established active compounds, providing a robust computational basis for further experimental validation.

On the other hand, hypericin continues to suppress the later stages of the viral replication cycle, including assembly and release. Unlike inhibitors targeting replicase enzymes such as NS3/NS5 [41, 42],—which can be affected by serotype variability and viral mutations—hypericin’s dual mode of action may help overcome the limitations of single‑site inhibitors (e.g., lectins or NS4B inhibitors) and reduce the risk of drug resistance. Furthermore, hypericin exhibits interesting broad‑spectrum activity at concentrations well below its cytotoxic threshold. It thus appears safer and holds greater therapeutic potential than some broad‑spectrum antiviral agents like ivermectin [43], positioning it as a promising lead for the development of next‑generation anti‑dengue therapeutics.

DENV infection starts with the attachment of E protein and entry into host cells [9]. Conformational rearrangements occur in the receptor-binding domain and the hydrophobic pocket of the E protein. In consequence of that, this area is seen as an ideal drug target [44, 45]. We showed with time-course dosing experiments that hypericin works mainly at the viral attachment and entry steps of action mechanism. Further experiments showed that its activity results from the direct interaction of the viral particles without using the host cell pathways. Molecular docking simulation studies show stable interactions of hypericin with numerous key E protein residues. This finding which the surface plasmon resonance (SPR) experiments validated indicated a high-affinity and specific binding between the two. The E protein demonstrates a high degree of conservation across the four known dengue virus serotypes [46]. This property has sound structural biology evidence and clinical application potential for hypericin as a direct-acting antiviral (DAA) with a wide range of antiviral capability.

HSP70, a highly conserved inducible molecular chaperone [47], serves as a critical host factor in DENV infection by translocating to the cell surface and binding the viral receptor complex to mediate viral entry [48]. This study reveals that hypericin specifically inhibits the interaction between the dengue viral E protein and the host factor HSP70, a chaperone known to support multiple stages of the infectious cycle—including viral entry, genome replication, and virion assembly [49]. During early infection, HSP70 normally translocates from the cytoplasm to the plasma membrane, where it binds the E protein to facilitate viral entry [44]. Our results indicate that hypericin interferes with this process, leading to a pronounced decrease in infectious viral particles and subsequent suppression of replication. The antiviral activity of hypericin thus appears to operate through the disruption of a critical host–pathogen interface. Further research should assess its potential as a broad-spectrum agent, particularly by examining whether it affects E protein interactions with additional host molecules.

The present investigation concentrated primarily on the DENV-2 serotype, future work is essential to assess hypericin’s activity against the other three dengue virus serotypes (DENV-1, 3, and 4). The present findings are mostly the outcome of in vitro studies, as such in vivo studies must be carried out for a better assessment of efficacy and safety. Preclinical and clinical data on hypericin are limited at present, and these data are crucial for clinical translation. Future research should provide further insight into the mechanism whereby hypericin suppresses viral replication through modifying the structure and function of the E protein. Further, combination therapies aimed at using hypericin and other antiviral agents could allow for more effective multi-target inhibition with less risk of resistance. Together, these studies will be solidifying the clinical utility of hypericin as an antiviral agent and offer stronger evidence.

Conclusions

In short, hypericin targets the dengue viruss E protein and interferes with its interactions with host HSP70 dampening viral adsorption and viral entry. The research confirms that hypericin is a powerful anti-DENV agent that is capable of direct targeting of the E protein of DENV.

Abbreviations

DENV

Dengue virus

DHF

Dengue hemorrhagic fever

DSS

Dengue shock syndrome

HIV

Human immunodeficiency virus

HCV

Hepatitis C virus

HSV-1

Herpes simplex virus type 1

IBV

Infectious bronchitis virus

HSP70

Heat shock protein

SPR

Surface plasmon resonance

E protein

Envelope protein

BHK-21

Baby hamster Syrian kidney

C6/36

Aedes albopictus larvae cells

A549

Human lung epithelial cells

Vero

Monkey kidney epithelial cells

DMSO

Dimethyl sulfoxide

PBS

Phosphate buffered saline

FBS

Fetal bovine serum

RPMI

Roswell Park Memorial Institute

DMEM

Dulbecco’s Modified Eagle’s Medium

CCK-8

Cell Counting Kit-8

h

hours

min

minutes

MOI

Multiplicity of infection

EC₅₀

Half-maximal effective concentration

CC₅₀

Half-maximal cytotoxic concentration

qRT-PCR

real-time fluorescence quantitative PCR

hpi

Hours post-infection

IFA

Immunofluorescence assay

Co-IP

Co-immunoprecipitation assay

Author contributions

LXQ, MLT, JFQ and JHW designed research; LXQ and MLT performed all the experiments and drafted the manuscript; LXQ, MLT, LBL, JZC and QQX contributed to data analysis; LBL provided suggestions on the manuscript; JFQ and JHW edited this manuscript. All authors reviewed and approved the manuscript.

Funding

The study was supported by the Training Project for High-Level Innovative talents in Guizhou Province, China (Grant No. Qian Ke He Platform Talent-GCC [2022] 0331), the Science and Technology Innovation Talent team of Guizhou Province, China (Grant No. Qian Ke He Platform Talent-CXTD [2022] 004) and the Central Government Guided Local Science Foundation of Guizhou Province (Grant No. Qian Ke He [2025] 024).

Data availability

All data supporting the findings of this study are available within the paper and its Supplementary Information.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors consent to the publication of the manuscript.

Competing interests

The authors declare no competing interests.