Development of anti-dengue virus D8 neutralizing monoclonal antibody production in Nicotiana benthamiana

Highlights

• •Plant-produced anti-dengue D8 mAb in ΔXF N. benthamiana neutralized all four dengue virus serotypes.
• •The D8 mAb showed strong NS1 binding and neutralization potency comparable to mammalian-derived HuMAb.
• •Glycosylation profiling revealed predominant GnGn and GnM3 forms, lacking plant-specific glycans.
• •FRNT assay showed highest activity against DENV4, with FRNT₅₀ < 1 μg/mL.

Abstract

Dengue is a life-threatening mosquito-borne viral disease ranging from mild symptoms to severe hemorrhagic fever. In this study, a recombinant anti-dengue D8 monoclonal antibody was produced in Nicotiana benthamiana and characterized for protein integrity and glycosylation by LC-MS. The plant-derived antibody lacked plant-specific β1,2-xylose and core α1,3-fucose residues, displaying a mammalian-like glycan profile. Functional evaluation showed potent cross-neutralizing activity against all four dengue virus serotypes (DENV1–DENV4), with the strongest activity against DENV4 (FRNT50 < 1 µg/mL), followed by DENV2 (5.82 µg/mL), DENV3 (9.49 µg/mL), and DENV1 (28.68 µg/mL), comparable to the mammalian-produced counterpart. The antibody also bound strongly to NS1 proteins of all serotypes, especially DENV2, and demonstrated higher reactivity than mammalian-derived anti-NS1 antibodies. Collectively, our results provide proof-of-concept that a glycoengineered plant platform can generate a functional D8 antibody with mammalian-like glycosylation, robust NS1 binding, and cross-neutralizing activity against DENV1–4, prompting further evaluation in Fc-dependent assays and in vivo models.

1. Background

Dengue virus (DENV) infection remains one of the most widespread mosquito-borne diseases, posing a major global health threat in tropical and subtropical regions. Each year, an estimated 390 million infections occur, with around 96 million resulting in clinical symptoms ranging from mild dengue fever to severe dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS) [1,2]. Severe cases are often associated with vascular leakage and plasma loss, leading to potentially fatal outcomes. Despite decades of research, the development of an effective dengue vaccine remains challenging due to the co-circulation of four distinct serotypes (DENV1–4) and the risk of antibody-dependent enhancement (ADE), where pre-existing cross-reactive antibodies from a primary infection with one serotype facilitate viral entry into Fcγ receptor-bearing cells, leading to more severe disease upon secondary infection with another serotype [3].

Current licensed vaccines, including CYD-TDV (Dengvaxia®) and TAK-003 (Qdenga®), demonstrate limited and variable efficacy across serotypes. CYD-TDV provides incomplete protection, especially in seronegative individuals under nine years old, who are at increased risk of ADE and severe disease [4,5]. TAK-003 offers higher efficacy in seropositive populations but reduced protection against DENV3 in seronegative individuals, and long-term safety data remain limited [6,7]. These constraints, along with high production costs and limited applicability in older adults, underscore the urgent need for new therapeutic and preventive strategies that ensure broad protection without inducing ADE. In the absence of specific antiviral drugs, therapeutic antibodies have gained increasing attention as alternative interventions for dengue. Neutralizing antibodies can block viral entry and fusion, providing serotype-spanning protection; however, preventing antibody-dependent enhancement (ADE) remains a central concern. Among potential targets, the non-structural protein 1 (NS1) has emerged as a promising and safer antigen compared with the viral envelope (E) protein. Unlike E-targeting antibodies that can promote ADE, anti-NS1 antibodies act primarily through Fc-mediated effector mechanisms, such as antibody-dependent cellular cytotoxicity (ADCC) [8], antibody-dependent phagocytosis (ADCP), and complement-dependent cytotoxicity (CDC) [9], to eliminate infected or NS1-expressing cells and reduce vascular leakage.

NS1 is a multifunctional glycoprotein highly conserved among flaviviruses and plays a dual role during infection. Intracellular NS1 supports viral RNA replication and particle assembly, while the secreted form contributes to immune evasion and vascular dysfunction. Circulating NS1 disrupts endothelial cell integrity, activates complement, and induces cytokine release, leading to plasma leakage and severe dengue manifestations such as hemorrhagic fever and shock syndrome [[9], [10], [11]]. Antibodies against NS1 have been shown to block these pathogenic effects by inhibiting endothelial permeability and lowering proinflammatory cytokines such as IL-8 and MCP-1 [12]. Consequently, NS1-targeting antibodies represent a promising therapeutic strategy to confer protection without triggering ADE-associated complications.

Sootichote et al. previously generated a panel of human monoclonal antibodies (HuMAbs) against DENV NS1, identifying several clones with cross-protective activity against dengue and Zika viruses [12]. Among these, clone 8 (D8), originally derived from a patient-isolated hybridoma, emerged as a highly promising candidate due to its potent cross-neutralizing activity against all four DENV serotypes (DENV1–4), derived from their reactivity against viral envelope (E) protein, and strong binding to NS1 protein leading to cross-protection mediated via Fc-effector activity. This dual-specificity mechanism enables direct viral neutralization through E protein binding while reducing NS1-mediated pathogenesis, positioning D8 as an ideal therapeutic for dengue. While mammalian cell lines like CHO cells are the industry standard for mAb production, their high costs and need for sophisticated infrastructure limit accessibility in dengue-endemic low- and middle-income countries. To address this, we selected the D8 antibody sequence for re-expression as a human IgG in a plant-based production system using Nicotiana benthamiana, aiming to validate its functional activity and develop a more accessible alternative to traditional mammalian systems.

Here, we report the expression, purification, and biochemical characterization of a plant-produced anti-NS1 monoclonal antibody (D8) generated in glycoengineered Nicotiana benthamiana ΔXF plants lacking plant-specific xylose and fucose residues. We evaluate its antigen binding and cross-reactivity across DENV-1–4 serotypes, assess its neutralizing activity, and characterize its N-glycan profile to confirm human-like glycosylation. Notably, we identify a marked binding preference for DENV-2 NS1, accompanied by enhanced antiviral potency against DENV-2, highlighting the influence of serotype-specific NS1 epitope presentation on antibody efficacy. Together, these findings provide proof-of-concept that glycoengineered plant systems can produce structurally and functionally competent anti-NS1 antibodies and establish a foundation for future studies assessing Fc-mediated effector functions relevant to dengue immunotherapy.

2. Materials and methods

2.1. Research involving plants

The anti-dengue virus D8 monoclonal antibody (mAb) was produced in Nicotiana benthamiana ∆XF plants, as previously described by Strasser et al. (2008). In brief, the ∆XF plants line was generated by crossing homozygous RNAi Nicotiana benthamiana lines with down-regulated XylT and FucT gene expression. F4 progeny were subsequently selected for stable silencing of both enzymes to produce glycoproteins with human-like N-glycans lacking plant-specific xylose and fucose residues. These mutant plants were grown under controlled environmental conditions at 25 ± 2 °C with a 16-hour light/8-hour dark photoperiod [[13], [14], [15]].

2.2. Construction of plant expression vector

The D8 mAb sequence was obtained from the Center of Excellence for Antibody Research, Faculty of Tropical Medicine, Mahidol University [12]. The sequences were codon-optimized for expression in Nicotiana benthamiana. The variable regions of heavy (VH) and light (VL) chains, containing an N-terminal murine signal peptide with an XhoI site, were synthesized by Genewiz (China). The VH and VL fragments were subsequently digested with BmtI and AflII restriction enzymes, respectively and ligated to human IgG1 heavy chain constant region and the human kappa light chain constant region, respectively. The resulting heavy and light chain constructs were inserted into the pBYR2eK2Md expression vector with XbaI/SacI restriction enzymes [16]. A schematic representation of this vector construct is shown in Fig. 1, which outlines the design for the expression of the D8 mAb in plants. For Agrobacterium-mediated transformation, the heavy and light chain expression vectors were individually introduced into Agrobacterium tumefaciens strain GV3101 using electroporation. Successful transformants were verified by colony PCR and DNA sequencing.

Fig. 1.

Schematic of the pBYR2eK2Md expression vector for D8 monoclonal antibody production.

The vector includes regulatory elements such as the PinII 3′ terminator, TMVΩ 5′-UTR, and P19 silencing suppressor. Transcription is driven by CaMV 35S promoters. The NbPsaK2T1–63 5′-UTR enhances expression in Nicotiana benthamiana. Gene insertion is performed between XbaI and SacI sites. The D8 heavy and light chains consist of signal peptides, variable (VH/VL), and constant (CH/CL) domains. Additional features include the Ext 3′ FL, Rb7 5′ deletion, and C1/C2 origin regions flanked by SIR and LIR sequences.

2.3. Transient expression in N. benthamiana by agroinfiltration

The infiltration procedure was performed as previously described [[16], [17], [18]]. Nicotiana benthamiana plants (4–6 weeks old) were grown under controlled conditions at 25 ± 2 °C with a 16-hour light/8-hour dark photoperiod. Agrobacterium tumefaciens carrying the D8 heavy and light chain constructs were cultured in Luria-Bertani (LB) medium supplemented with 50 µg/mL kanamycin, gentamicin, and rifampicin. The heavy and light chains were co-infiltrated at a 1:1 OD ratio with an infiltration buffer composed of 10 mM MES and 10 mM MgSO₄ (pH 5.5), using syringe infiltration without a needle. Leaf samples were collected at 3 days post-infiltration (dpi) and kept at −80 °C for analysis of recombinant antibody expression.

2.4. Recombinant protein extraction and purification

The leaves were extracted using PBS at a 1:2 (w/v) ratio relative to the fresh weight of the leaves. The homogenate was filtered through clean cotton cloth to remove plant biomass, and the filtrate was centrifuged at 24,700 x g at 4 °C for 20 min. The resulting clear supernatant was further processed by filtration through a 0.45 µm membrane filter (Merck, Ireland). The clarified extract was loaded onto a Protein A resin column (Amintra, Expedon, Cambridge, UK). The column was washed with PBS, and the antibody was eluted with an elution buffer (0.1 M glycine buffer, pH 2.7). The eluate was immediately neutralized with 1.5 M Tris–HCl (pH 8.8) to adjust the pH to 7.4. The purified antibody was dialyzed in PBS cell grade and concentrated using 50 kDa molecular weight cutoff (MWCO) Amicon Ultra-centrifugal filter (Merck, Massachusetts, USA). After purification, the final protein concentration was determined using the Bradford protein assay to ensure quality and yield before storing the protein at −20 °C.

2.5. SDS PAGE and western blotting

SDS-PAGE was performed as previously described [16,19]. Briefly, purified plant-produced D8 mAb was mixed with either reducing or non-reducing loading buffer. Samples were separated on 6 % polyacrylamide gel under non-reducing conditions and a 4–20 % gradient gel under reducing conditions, followed by staining with One-step Blue® (Biotium, USA). Proteins were then transferred onto nitrocellulose membranes (Thermo Fisher Scientific, Waltham, MA, USA). Membranes were blocked with 3 % skim milk for 30 min and probed with HRP-conjugated anti-human gamma chain antibody 1:10,000 (2040–05, Southern Biotech, US) and HRP-conjugated anti-human kappa chain antibody 1:5000 (2060–05, Southern Biotech, US). Signal detection was performed using an ECL chemiluminescent substrate (Cytiva (RPN2106), GER), and protein bands were visualized using X-ray film.

2.6. N-glycan analysis

2.6.1. LC-MS subunit mass analysis

Subunit mass analysis was performed to determine the intact molecular mass and confirm the integrity of the purified protein. The procedure followed the established method described previously(20, 21),with minor adjustments. After protein purification, approximately 20 µg of D8 mAb protein (with a concentration of 1 mg/mL in PBS buffer) was buffer exchanged into 50 mM ammonium bicarbonate (ABC) buffer. It was reduced with 10 mM dithiothreitol (DTT) and incubated at 65 °C for 30 min. The sample was centrifuged at 14,000 rpm for 10 min prior to injection. LC-MS analysis was performed on an Agilent 1290 Infinity II liquid chromatography system coupled with an Agilent 6545XT Q-TOF mass spectrometer. Protein separation was carried out using an Agilent PLRP-S column (1000 Å, 5 µm; 2.1 × 50 mm) at 60 °C. The injection volume was 3 µL. The LC and MS parameters were set according to the previous method [20,21]. A 0.1 % formic acid (FA) in water and acetonitrile were used as mobile phases A and B, respectively. A 12-minute chromatographic run was employed at 0.4 mL min⁻¹. A LC gradient was set as follows: 0 min at 25 % B, 0–1 min from 25–30 % B, 1–3 min, from 30–32 % B, 3–4 min from 32–35 % B, 4–5 min from 35–40 % B, 5–6 min from 40–90 % B, 6–9 min at 90 % B, 9–9.5 min from 90–25 % B and maintaining at 25 % B until 12 min.. The MS was operated in positive mode with mass range of 100–3200 m/z, gas temperature of 350 °C, gas flow of 12 L h^-1, nozzle voltage of 2000 V, skimmer voltage of 65 V, nebulizer at 35 psi and capillary voltage at 4000 V.

2.6.2. LC-MS peptide mapping

Peptide mapping was performed to confirm amino acid sequence coverage and post-translational modifications of the target protein. Approximately 60 µg of protein (1 mg/mL) was reduced with 10 mM dithiothreitol (DTT) at 65 °C for 30 min, followed by alkylation with 25 mM iodoacetamide (IAA) for 20 min in the dark at room temperature. The treated sample was digested with 0.6 µg trypsin at 37 °C for 4 h, and the reaction was terminated with 1 % formic acid (FA). After centrifugation at 14,000 rpm for 10 min, the supernatant was transferred to an LC–MS vial for analysis. Peptide separation and mass spectrometric detection were performed using an Agilent 1290 Infinity II LC system coupled to an Agilent 6545XT Q-TOF mass spectrometer equipped with an AdvanceBio Peptide Mapping column (120 Å, 2.1 × 150 mm, 2.7 µm) at 60 °C and an injection volume of 10 µL. The LC–MS conditions were consistent with those established in previous studies [20,21]. It was conducted using 0.1 % FA in water (mobile phase A) and acetonitrile (mobile phase B) at 0.4 mL min⁻¹, employing an 85-minute gradient elution. A LC gradient was set as follows: 0 min at 0 % B, 0–2 min at 0 % B, 2–35 min from 0–20 % B, 35–55 min from 20–30 % B, 55–65 min from 30–50 % B, 65–70 min from 50–90 % B, 70–75 min at 90 % B, 75–80 min from 90–0 % B and 80–85 min at 0 % B. MS and MS/MS data were acquired in positive mode with a range of 100–1700 and 25–1000 m/z-,respectively and acquisition rates of 5 and 3 spectra s⁻¹, respectively. MS source conditions were a drying gas temperature of 325 °C, gas flow of 13 L h^-1, a nozzle voltage of 500 V, nebulizer at 35 psi and capillary voltage at 4000 V. The top 10 precursor ions per cycle with a minimum threshold of 3000 counts were selected for fragmentation. Other parameters were aligned with previous studies [20,21].

2.6.3. Data analysis

Mass spectrometric data were processed using Agilent MassHunter BioConfirm software version 11.0 to identify intact mass and peptide fragments. Data interpretation and all analysis parameters were performed in accordance with the previously established workflow [20,21].

2.7. Focus reduction neutralization test (FRNT)

The neutralizing activity of monoclonal antibodies (mAbs) against dengue virus (DENV) was evaluated by a focus reduction neutralization test (FRNT) as previously described [18], with minor modifications. Briefly, Vero cells were seeded in 96-well plates at a density of 3 × 10⁴ cells/well and incubated overnight at 37 °C in a 5 % CO₂ atmosphere. Serial two-fold dilutions of each mAb, starting at 64 µg/mL for 7 dilutions until 1 µg/mL, were prepared in culture medium and mixed with each DENV serotype. After incubation at 37 °C for 1 h to allow antibody–virus complex formation, the mixtures were added to Vero cell monolayers and incubated for an additional 2 h to permit viral entry. Following infection, cells were overlaid with a semi-solid medium containing 2 × MEM, 2 % carboxymethyl cellulose (CMC), and 2 % fetal bovine serum (FBS), and incubated at 37 °C for 2 days (DENV4) or 3 days (DENV1–3). Cells were then fixed with 3.7 % formaldehyde for 20 min and permeabilized with 0.1 % Triton X-100. DENV-infected foci were detected using an anti-DENV primary antibody, followed by Alexa Fluor® 488-conjugated goat anti-human IgG (H + L) secondary antibody (1:1000 dilution) (A11013, Thermo Fisher Scientific, USA). Fluorescent foci were visualized using a fluorescence microscope, and the neutralization percentage was calculated based on the reduction of focus-forming units (FFUs) relative to the virus-only control.

2.8. Binding of mAbs to NS1 protein

To confirm the optimal binding of monoclonal antibodies to NS1 protein, antibodies D8 derived from HEK cells and Nicotiana benthamiana by transient expression were studied. DENV NS1 protein of dengue virus serotype 1–4 produced from bacterial E.coli strain was used for binding assay. These proteins were previously tested to show optimal binding with several dengue patients serum samples [22]. The protein from each serotypes were coated at 300 ng/well for 100 μL of the final protein concentration at 3 μg/mL at 4 C° for overnight. On the next day, wells were washed with Tris buffer saline with 0.05 % Tween (TBS-T) for 5 times, and blocked with 1 % BSA in TBS-T for 1 hour at 37 °C. Then, wells were washed again as above, and added with antibodies (HuMAb clone 8 (HEK) and plant-produced D8) at 10 µg/ml. Dengue patient serum (1:1500) was used as control. BSA was used as negative control. After incubation at 37 °C, for 1 hour, wells were washed again, and added with secondary antibody (Goat anti-human IgG-HRP conjugated) (AP112P, Sigma Aldrich, USA.) at 1:15,000 dilution. The wells were incubated for 1 hour at 37 °C. Then, the wells were washed again, and adding with KPL TMB Microwell Peroxidase substrate system 100 µl/well (Sera care, USA.). The interaction signal was read at 450 nm using Glomax Discover (Promega, USA.). All samples were performed in duplicate, and the average from two independent experiments were used for analysis.

2.9. Ethical statement

All experimental procedures using human samples were preapproved by the Ethics Committee of the Faculty of Tropical Medicine (FTM), Mahidol University (protocol number: FTM ECF-019–05). All donors provided a written informed consent before enrollment.

3. Results

3.1. Expression and purification of plant-produced d8 monoclonal antibody

Agrobecterium tumerfaciens containing heavy and light chain of D8 monoclonal antibody (mAb) were co-infiltrated in Nicotiana benthamiana. After 3 days post-infiltration (dpi), the leaves were harvested, homogenized and crude extract was purified using Protein A affinity chromatography. The purified D8 mAb yield was 12.44 µg/g fresh leaf tissue, as determined by the Bradford assay.

The integrity and assembly of the purified D8 mAb were analyzed by SDS–PAGE and Western blotting (Fig. 2). Under non-reducing conditions, a single major band was detected at approximately 150 kDa (Fig. 2A), corresponding to the intact, fully assembled IgG molecule. Western blotting with HRP-conjugated anti-human gamma (Fig. 2B) and kappa (Fig. 2C) antibodies confirmed the presence of both heavy and light chains within the assembled antibody. Under reducing conditions, SDS–PAGE revealed two distinct bands at approximately 50 kDa and 25 kDa (Fig. 2D), representing the heavy and light chains, respectively. These were further validated by Western blot analysis, in which the anti-human gamma antibody (Fig. 2E) specifically recognized the heavy chain, while the anti-human kappa antibody detected the light chain (Fig. 2F). These results confirm that the plant-produced D8 mAb was correctly expressed, assembled, and retained its expected molecular structure consistent with a functional IgG antibody.

Fig. 2.

SDS-PAGE and One-step Blue®staining of D8 mAb after protein A affinity column purification under non-reducing (A) and reducing condition (D). western blot analysis with anti-human gamma (B, E) and anti-human kappa (C, F) for heavy and light chain detection, respectively and M: protein marker, D8: D8 mAb.

3.2. N-Glycan analysis

The subunit mass analysis of the plant-produced D8 monoclonal antibody (mAb) was performed using LC–MS to verify its molecular integrity and glycosylation status (Fig. 3). The light-chain subunit displayed a single major peak at 23,562 Da, consistent with the theoretical mass of a correctly folded, non-glycosylated chain (Fig. 3A). The heavy-chain subunit showed a major non-glycosylated form (50,050.6 Da) along with additional peaks at 51,223.5, 51,352.2, 51,755.6, and 51,917.8 Da, representing glycosylated variants carrying one or more N-linked glycans (Fig. 3B). This profile indicates partial glycosylation of the heavy chain during plant expression, yielding a heterogeneous glycoform mixture.

Fig. 3.

Subunit Mass Analysis of Light and Heavy Chains for D8 mAb; The subunit mass analysis of the light chains for D8 mAb revealed a molecular weight of 23,562 Da, which aligns with the theoretical mass (A). The subunit mass analysis of the heavy chains for D8 mAb exhibited distinct peaks corresponding to non-glycosylated and glycosylated forms. The non-glycosylated heavy chain showed a mass of 50,050.6 Da, while the glycosylated forms were observed at 51,223.5 Da, 51,352.2 Da, 51,755.6 Da, and 51,917.8 Da. (B).

Further characterization of the N-glycan profile at the Asn308 site was performed using LC–MS peptide mapping. The majority of the D8 mAb was non-glycosylated (73.5 %), while the remaining 26.5 % carried plant-type N-glycans (Fig. 4A). The predominant glycoforms included GnM3 (13.3 %) and GnGn (7.4 %), with minor populations of high-mannose (Man8, Man9) and hybrid-type (GnM4X) glycans. These findings indicate that the majority of D8 mAb molecules produced in N. benthamiana are non-glycosylated, while the remaining fraction carries typical plant-derived oligomannose-type glycans.

Fig. 4.

Glycoprofiling of D8 mAb The glycoprofiling at the N308 position of the D8 mAb heavy chain was analyzed. The chromatogram of the EEQYNSTYR peptide showed that the non-glycosylated protein was primarily eluted at 12.23 min, while the glycosylated forms were eluted between 11.49 and 12.30 min (A). Inset table summarizing N-glycosylation data, where 73.5 % of the molecules were not glycosylated. Among 26.5 % glycosylated peptides, the major peak was GnM3 (13.3 %) and GnGn (7.4 %). Man8 and Man9 were observed only in trace amounts (B). The green circles represent Mannose (M) and the blue squares represent Galactosamine (Gn).

3.3. Cross-neutralizing activity of plant-produced D8 mAb and HuMAb clone 8 against DENV

The cross-neutralization activity of the plant-produced D8 monoclonal antibody (mAb), expressed in Nicotiana benthamiana, was evaluated against all four dengue virus serotypes (DENV1–DENV4) using the focus reduction neutralization test (FRNT). The plant-produced D8 mAb exhibited potent and broad neutralizing activity, comparable to the HEK cell-derived HuMAb clone 8 (Fig. 5, Table 1). The plant-produced D8 demonstrated the strongest neutralization against DENV4, with an FRNT50 of 0.78 µg/mL, similar to HuMAb clone 8 (0.89 µg/mL). For DENV2, the plant-produced D8 achieved an FRNT50 of 5.82 µg/mL, slightly more potent than HuMAb clone 8 (12.86 µg/mL). Against DENV3, both antibodies showed comparable neutralizing titers, with FRNT50 values of approximately 9 µg/mL. In contrast, DENV1 was the least efficiently neutralized serotype, with FRNT50 values of 28.68 µg/mL for the plant-produced D8 and 13.37 µg/mL for HuMAb clone 8.

Fig. 5.

The neutralization potential of the antibodies was assessed through the results of at least two independent experiments. A 50 % neutralization threshold is indicated by the dotted line. Data are presented as the mean ± standard deviation (SD).

Table 1.

. FRNT50 of Plant-produced D8 and HuMAb clone 8 against DENV 1- DENV 4.

Mabs	FRNT50 in µg/ml (95 % CI)
	DENV1	DENV2	DENV3	DENV4
Plant-produced D8	28.68	5.82	9.49	0.78
	(22.64 - 38.01)	(4.657 - 7.27)	(7.28 - 12.50)	(0.43 - 1.14)
HuMAb clone 8 (HEK)	13.37	12.86	9.08	0.89
	(10.80 - 16.83)	(10.51 - 15.74)	(7.27 - 11.39)	(0.59 - 1.21)

Overall, these results confirm that the plant-produced D8 mAb retained strong and cross-reactive neutralizing activity against multiple DENV serotypes, demonstrating functional equivalence to its mammalian-cell-derived counterpart. This supports its potential as a cost-effective and scalable candidate for dengue immunotherapy.

3.4. Binding to NS1 protein

The binding of HuMAb clone 8 (produced in HEK cells) and Plant-produced D8 to dengue non-structural protein 1 (NS1) from all four serotypes (DENV1–4) was confirmed using ELISA. Dengue patient serum (white bars) served as a positive control, validating the assay and showing broad reactivity across all serotypes, with the strongest signal observed for DENV2. Both HEK-derived HuMAb clone 8 (gray bars) and Plant-produced D8 (black bars) exhibited cross-reactivity with all four NS1 serotypes, also showing highest binding to DENV2 NS1. Interestingly, Plant-produced D8 displayed markedly higher binding signals (OD450 ∼ 1.7 for DENV2) compared with HEK-derived HuMAb 8 (OD450 ∼ 0.15 for DENV2) across all serotypes. Despite this enhanced ELISA binding, the neutralizing activity of the two antibodies was comparable, indicating that differences in ELISA signal do not necessarily reflect functional potency.

4. Discussion

Antibodies targeting the non-structural protein 1 (NS1) represent a promising therapeutic and vaccine strategy, as it overcome the challenges in dengue vaccine and antibody development of antibody-dependent enhancement (ADE), in which sub-neutralizing antibodies facilitate viral entry into Fcγ receptor–bearing cells and lead to increased disease severity.

In this study, we successfully expressed the anti-NS1 D8 monoclonal antibody in Nicotiana benthamiana and demonstrated its cross-protective activity against all four dengue virus serotypes. These findings align with our previous reports on the plant-produced D54 antibody, which primarily targets the E protein and confers broad protection without inducing the ADE phenomenon [18]. Notably, D8 exhibits a unique dual specificity, targeting both the viral envelope (E) protein and the non-structural protein 1 (NS1), consistent with its original characterization in hybridoma cell [12]. This suggests that plant-produced D8 employs a dual-action mechanism, direct viral neutralization via the E protein and indirect inhibition of pathogenesis by targeting secreted NS1.

NS1 plays a critical role in dengue pathogenesis, particularly in vascular leakage and endothelial dysfunction characteristic of dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS) [12,23]. Typically, NS1 is secreted as a soluble hexamer that directly triggers vascular permeability, while it also exists as a dimeric form on the surface of infected cells for viral replication. In this study, we confirmed that D8 effectively binds to these secreted NS1 proteins which leads to the neutralization of the virus's pathogenic effects. This capability, combined with its direct binding to the E protein, provides D8 with a dual-action neutralizing potential that targets both viral entry and NS1-mediated endothelial disruption. These integrated mechanisms significantly attenuate viral replication and stabilize the endothelial barrier. Nevertheless, while the binding and neutralizing activities are evident, further validation of Fc-mediated effector functions, such as antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC), is necessary to fully characterize the therapeutic efficacy of plant-produced D8.

The human monoclonal antibody D8 activates complement-mediated cytolysis, inhibits viral replication, reduces vascular permeability, and suppresses cytokine and chemokine secretion by targeting NS1 residues 221–299 [12]. This observation is consistent with a separate study in which antibodies raised against a C-terminally truncated NS1 (ΔC NS1), lacking residues 271–352 but retaining the 221–299 region, exhibited reduced binding to endothelial cells and platelets and conferred improved protection, as reflected by shortened bleeding times, compared with antibodies against full-length NS1 [24]. Together, these data suggest that residues 221–299 harbor protective epitopes, whereas more distal C-terminal regions (271–352) contribute to vascular-associated pathogenic effects, underscoring the importance of epitope specificity in achieving safe and effective anti-NS1 antibody responses. This reflects DENV2.

Crucially, our monoclonal antibody exhibits a marked binding preference for DENV-2 NS1 (Fig. 6), indicating serotype-biased recognition. This specificity is likely a result of the antibody’s origin, as the original hybridoma was derived from DENV-2 patient [12]. Another supported evident is likely due to the epitope accessibility within the C-terminal domain of DENV-2 NS1, where serotype-specific sequence polymorphisms and dimeric conformation may favor D8 engagement over other serotypes. Consistent with this, previous studies of anti-NS1 monoclonal antibodies have shown that epitope polymorphisms across DENV serotypes can drive substantial differences in binding affinity despite recognition of overlapping sites. DENV-2 specific sequence features within the C-terminal β-ladder may therefore optimize epitope presentation, as supported by enhanced detection of DENV-2 NS1 dimers reported in related studies [25]. Importantly, this binding preference correlates with increased antiviral potency against DENV-2, indicating that higher affinity engagement of NS1 is functionally relevant. Together, these findings highlight how serotype-specific structural features of NS1 influence antibody efficacy, while further structural and in vivo studies will be required to fully define the underlying mechanisms.

Fig. 6.

The binding activity of HuMAb clone 8 produced from HEK, and plant to recombinant NS1 protein. Data are presented as the mean ± standard deviation (SD). BSA was used as negative control.

Our study is using Nicotiana benthamiana ΔXF plants as the expression system. Glycosylation analysis revealed predominant glycoforms of GnGn and GnM3, which lack plant-specific core fucose and xylose residues. Previous research utilizing ΔXF N. benthamiana for anti-dengue E60 mAbs demonstrated superior neutralization efficacy against DENV2 compared to its counterparts produced in mammalian cell culture and wild-type plants, which mainly present the GnGnXF3 glycan structure. Furthermore, the ΔXF-derived antibody demonstrated a diminished risk of antibody-dependent enhancement (ADE) compared to the mammalian cell-derived antibody [26]. Although the mechanism behind this reduced ADE remains unclear, similar results were observed for an anti-Ebola antibody (13F6), in which ΔXF plant-produced variants carrying homogeneous GnGn glycans displayed enhanced ADCC activity, increased FcγRIIIa binding, and greater antiviral potency than mammalian cell-derived antibodies [27].

These findings support that the absence of core fucose and the homogeneity of GnGn glycans likely contribute to improved FcγR engagement and enhanced effector-mediated clearance. However, Fc glycan structure must be carefully controlled. Afucosylated IgG1 antibodies exhibit a significantly higher binding affinity up to 49-fold for FcγRIIIa [28] enhancing ADCC activity but also potentially increasing ADE risk under certain immune conditions. Clinical data show that dengue patients with elevated afucosylated IgG1 exhibit more severe disease manifestations, including thrombocytopenia, likely due to excessive FcγRIIIa activation [29]. Thus, while glycoengineering in ΔXF plants offers favorable glycan profiles for therapeutic efficacy, further in vivo evaluation is essential to balance enhanced Fc effector functions with potential ADE-related risks for NS1-targeting antibodies.

In this research, the successful expression and functional activity of the anti-dengue D8 mAb in N. benthamiana further demonstrated cross-protection against all four dengue serotypes, highlighting its potential as a broadly protective therapeutic. These findings align with growing evidence that plant-based systems can deliver therapeutically relevant proteins efficiently and offering a viable solution for the production of biologics in dengue-endemic and resource-limited regions.

5. Conclusions

In conclusion, this study demonstrates the successful production of the D8 monoclonal antibody targeting dengue virus NS1 using a glycoengineered plant-based expression system. The plant-produced D8 antibody retained dual specificity for both NS1 and E proteins and exhibited potent cross-protective activity against all four DENV serotypes, comparable to its mammalian-produced counterpart. This dual mode of action likely involves direct inhibition of viral entry through E protein binding, together with mitigation of NS1-mediated vascular pathology by neutralizing secreted NS1. Importantly, the optimized plant-derived glycosylation profile supports functional antibody activity and offers opportunities for Fc glycan tailoring to enhance therapeutic efficacy. Collectively, these findings highlight the plant expression platform as a viable, cost-effective approach for producing functional, cross-protective monoclonal antibodies, supporting further development of D8 as a therapeutic candidate for dengue control.

Funding

This study was supported by Thailand Science research and Innovation Fund Chulalongkorn University (HEA_FF_69_182_3300_023) and the Second Century Fund (C2F), Chulalongkorn University (awarded to Gunt Niyompun).

CRediT authorship contribution statement

Gunt Niyompun: Writing – review & editing, Writing – original draft, Methodology, Investigation. Kaewta Rattanapisit: Writing – review & editing, Writing – original draft, Methodology. Janejira Jaratsittisin: Writing – review & editing, Writing – original draft, Methodology, Investigation, Conceptualization. Pipob Suwanchaikasem: Writing – review & editing, Writing – original draft, Methodology. Rochanawan Sootichote: Writing – review & editing, Writing – original draft, Methodology. Pichamon Sittikul: Writing – review & editing, Writing – original draft, Methodology. Pannamthip Pitaksajjakul: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. Pongrama Ramasoota: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. Waranyoo Phoolcharoen: Writing – review & editing, Writing – original draft, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

Waranyoo Phoolcharoen is a founder/shareholder of Baiya Phytopharm Co., Ltd. Kaewta Rattanapisit, Janejira Jaratsittisin and Pipob Suwanchaikasem are employees of Baiya Phytopharm Co., Ltd. The remaining authors declare no competing interests.

Data availability

No data was used for the research described in the article.