Vector minimization and infiltration optimization for enhanced protein expression in Nicotiana benthamiana Authors: Kulachatr Panyawechamontri, Hiroyuki Kajiura, Ryo Misaki, Kazuhito Fujiyama Published: 2026 ABSTRACT Key message Vector minimization, combined with optimized infiltration buffer conditions, enhances recombinant protein expression in Nicotiana benthamiana , offering practical improvements in plant-based biotechnology. AbstractNicotiana benthamiana is a widely utilized host for transient expression of recombinant proteins, offering a rapid and flexible system for biotechnology applications. However, achieving high-protein output remains a significant challenge. In this study, we explored how vector minimization and infiltration buffer optimization influence recombinant protein expression. We systematically reduced the size of the commonly used pBI121 vector by eliminating non-essential sequences, generating pBI121Δlac and modified pBI121 (mpBI121) constructs. Using ꞵ-glucuronidase (GUS) and acid sphingomyelinase (ASM) as model proteins, we found that vector minimization significantly enhanced both mRNA and protein expression. The mpBI121 backbone improved GUS and ASM mRNA expression levels by 2.02-fold and 2.05-fold, respectively, versus pBI121. Correspondingly, protein levels increased by 1.76-fold and 1.75-fold. Optimization of infiltration buffer—together with co-expression of the silencing suppressor p19, along with acetosyringone, and ascorbic acid—further improved protein output, producing the highest expression of both proteins. These findings demonstrate that both vector minimization and optimized agroinfiltration conditions enhance recombinant protein expression in plant systems, advancing plant-based biotechnology platforms.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00299-026-03802-z. Introduction Nicotiana benthamiana serves as an ideal host for agroinfiltration-based transient expression, providing a rapid, flexible, and scalable platform for research and biotechnological applications in functional genomics, metabolic engineering, and recombinant protein production (Bally et al. 2018). Its natural susceptibility to Agrobacterium tumefaciens, along with a relatively permissive immune system and strong capacity for transgene expression, makes it especially well-suited for high-throughput and short-term assays. Unlike stable transformation methods, which are labor-intensive and time-consuming, transient expression in N. benthamiana enables gene function studies and protein production within just a few days—an important advantage during a pandemic (Ruocco and Strasser 2022). This system supports the expression of a broad range of proteins, including cytosolic, membrane-bound, and secreted types, without requiring stable transformation, thereby accelerating gene function characterization and biosynthetic pathway elucidation across diverse applications in functional genomics, metabolic engineering, and synthetic biology (Grützner et al. 2024). As a result, N. benthamiana has been widely adopted for the rapid expression of high-value recombinant proteins, including monoclonal antibodies (Sangprasat et al. 2024), enzymes (Sinha et al. 2024), and vaccine components (Ruocco and Strasser 2022), positioning it as a key platform in the expanding field of molecular pharming (Buyel 2018; Chen and Lai 2015). Despite its widespread adoption, maximizing protein yield in transient expression systems remains a significant challenge. While efficient T-DNA delivery and strong promoter-driven transcription are essential for initiating expression, translational efficiency and protein stability ultimately determine accumulation levels. These processes are further influenced by multiple factors, including vector construct design, the efficiency of bacterial transformation, gene silencing and stress responses, and the technical parameters of the infiltration process (Mardanova et al. 2017). Among these, the design and size of the binary vector used for Agrobacterium-mediated transformation play a crucial role. Larger binary vectors often carry additional regulatory elements or selection markers that may not be essential for transient expression, potentially imposing a metabolic burden and reducing transformation performance in either Agrobacterium or the plant host (Thole et al. 2007). This raises the compelling hypothesis that vector minimization, the removal of non-essential sequences, could enhance protein output by improving transformation efficiency and production dynamics in plant cells. Notably, the pGreen/pSoup system demonstrates that substantial vector size reduction is compatible with efficient gene delivery and expression, supporting the rationale for testing minimal vector designs. However, complete removal of selection markers may not be advisable, as marker-free constructs have been associated with transgene instability and gene silencing in some plant systems (Hellens et al. 2000). Beyond vector design, infiltration buffer composition and incubation time have emerged as critical factors influencing the optimization of these parameters which play a more central role than is often recognized in influencing transformation performance and transgene expression (Leuzinger et al. 2013; Sparkes et al. 2006). In this study, we investigated how vector design and optimized agroinfiltration conditions affect recombinant protein expression in N. benthamiana. We propose a two-step, integrated strategy: (1) systematic reduction of a standard plant expression vector size by removing non-essential sequences and (2) optimization of infiltration buffer composition. We evaluated this using β-glucuronidase (GUS), a widely used reporter enzyme, and acid sphingomyelinase (ASM) (Panyawechamontri et al. 2025), representing pharmaceutically relevant recombinant proteins. Our results demonstrate that vector minimization enhances the output of both proteins and that further improvements in ASM yield can be achieved through buffer optimization. Together, these findings provide practical guidance for maximizing protein production and establish a robust framework for advancing transient expression systems in plant biotechnology. Materials and methods Vector minimization and plasmid construction A series of minimized binary vectors was constructed based on the widely used pBI121 backbone by sequentially removing non-essential genetic elements not required for the plasmid’s core function in this context. These include the lac promoter/operator region (872 bp) and the tetracycline resistance (tetR) gene (1626 bp). Removing these non-essential elements reduced the plasmid size while retaining the components necessary for its primary functions, such as gene expression and cloning. Initially, the plasmid was amplified using primers XhoI-lac-F and XhoI-lac-R, followed by XhoI digestion to excise the lac promoter/operator region. The resulting product was ligated to generate pBI121Δlac. Next, pBI121Δlac was used as a template to remove the tetR gene using two primer sets: SalI-tetR1-F/HindIII-tetR1-R and HindIII-tetR2-F/SalI-tetR2-R. The corresponding PCR products were digested with SalI and HindIII and ligated to produce the final minimized vector, pBI121ΔlacΔtetR, also referred to as modified pBI121 (mpBI121). To create a control construct, the β-glucuronidase (GUS) gene in mpBI121 was replaced with the green fluorescent protein (GFP) gene. The GFP fragment, obtained from the pRI201-GFP plasmid (Yamamoto 2023), was digested with HindIII and EcoRI and ligated into the similarly prepared mpBI121 backbone to form the mpBI121-GFP construct. All plasmids were first transformed into Escherichia coli DH5α, followed by introduction into Agrobacterium tumefaciens LBA4404. Primer sequences are provided in Table S1A. Bacterial strains and antibiotic sensitivity test E. coli DH5α and A. tumefaciens LBA4404 were employed for plasmid propagation, with E. coli cultured in 2xYT broth or on agar plates at 37 °C and Agrobacterium grown at 30 °C. E. coli and Agrobacterium were transformed with plasmid constructs via the heat-shock method and electroporation, respectively. To verify the successful deletion of the tetR gene, transformants carrying plasmids pBI121, pBI121Δlac, and mpBI121 were plated on 2xYT agar supplemented with 50 µg/mL each of kanamycin, streptomycin, and rifampicin. A parallel set of plates was prepared using the same antibiotics, with the addition of tetracycline (10 µg/mL). All plates were incubated at 30 °C for 2 days and monitored for colony growth. Engineering vectors for expression of acid sphingomyelinase (ASM) To validate the expression observed with the minimized vector, the GUS gene in pBI121, pBI121Δlac, and mpBI121 backbones was replaced with a truncated ASM fused to an extensin signal peptide (ntExtASM). Additionally, the AtADH 5′ UTR was inserted upstream of the coding sequence, and the NOS terminator was replaced with the HSP terminator to improve transcript stability and translation efficiency. For the pBI121Δlac construct, the plasmid backbone was amplified by PCR using EcoRI-M13-F and HindIII-Splac-R primers (Table S1A) and subsequently digested with EcoRI and HindIII. Meanwhile, the pBI121 and mpBI121 backbones were directly digested with the same enzymes. Using EcoRI and HindIII, the ntExtASM insert was excised from the pRI201-ntExtASM construct as described previously (Panyawechamontri et al. 2025). This fragment was then ligated into the prepared backbones to generate the pBI121-ntExtASM, pBI121Δlac-ntExtASM, and mpBI121-ntExtASM constructs. All plasmids were introduced into E. coli DH5α and A. tumefaciens LBA4404 for further use. Agroinfiltration in Nicotiana benthamiana Wild-type plants were maintained in pots under controlled conditions: 25 °C, 50% relative humidity, and a 16-h light/8-h dark cycle. Plants were watered and provided with a weekly supplement of liquid fertilizer (N-P-K: 6-10-5; HYPONeX, Osaka, Japan) as necessary. Five-week-old plants were used for agroinfiltration. Agrobacterium cultures carrying the desired plasmid constructs were prepared in infiltration buffer (10 mM MES, 10 mM MgSO4, pH 5.5) and adjusted to an OD600 of 0.5 for the infiltration process to assess target gene expression. Blue light–based validation of GFP expression Leaves agroinfiltrated with the mpBI121-GFP and mpBI121 (negative control vector) constructs were harvested and exposed to 470 nm blue light (Digi Print Doc; Biotools, Gunma, Japan). Fluorescence images were captured using a Canon Kiss X5 camera with the following settings: exposure time 0.25 s, aperture F4.5, and ISO 1600. The images were adjusted for brightness (+ 40%) and contrast (− 20%). Fluorescence images were then compared to those of non-infiltrated control leaves. RNA isolation and cDNA synthesis Total RNA was extracted from infiltrated leaf samples at 6 days post-infiltration (6 dpi) using the RNeasy Plant Mini Kit (QIAGEN, Hilden, Germany) following the manufacturer’s protocol. Approximately 0.2 g of leaf tissue was placed into a 2 mL Simport screw cap tube (T332-7, Ina Optica, Osaka, Japan) with two metal beads and immediately frozen in liquid nitrogen. While frozen, the tubes were shaken twice at 1,200 rpm for 30 s using the SH-100 tissue homogenizer (KURABO, Osaka, Japan) to grind the tissue into a fine powder. Genomic DNA contamination in the RNA sample was removed using RQ1 RNase-free DNase (Promega, Madison, WI, USA) as per the manufacturer’s instructions. Complementary DNA (cDNA) was then synthesized from the DNase-treated RNA using the SuperScript™ IV VILO™ Master Mix (Invitrogen; Thermo Fisher Scientific, Waltham, MA, USA) following the provided protocol. The resulting cDNA was stored at − 20 °C. Reverse transcription-quantitative PCR (RT-qPCR) and mRNA expression analysis RT-qPCR was conducted as described in previous studies (Bennett et al. 2015). All cDNA samples were amplified in triplicate, each derived from a separate RNA preparation. The reaction mixture consisted of 5 ng/µL of cDNA, 300 nM CXR reference dye (Promega), 10 pmol/µL of both forward and reverse primers, and 2 × GoTaq® qPCR Master Mix (Promega), bringing the total volume to 20 µL. Amplification was performed in a MicroAmp™ Optical 96-Well Reaction Plate (0.1 mL, Applied Biosystems, Foster City, CA, USA) on a QuantStudio 3 system (Applied Biosystems). The primers used for RT-qPCR are detailed in Table S1B. The thermal cycling conditions included an initial denaturation at 50 °C for 1 min, followed by 95 °C for 10 min, then 40 cycles of 95 °C for 15 s, 55 °C for 30 s, and 60 °C for 1 min. Fluorescence was measured after each cycle. A final dissociation curve was performed with steps at 95 °C for 25 s, 60 °C for 1 min, and 95 °C for 15 s. To ensure consistency, all samples for each reference gene were run on the same plate. The expression levels of target genes were normalized using 18S ribosomal RNA (18S), elongation factor 1-alpha (EF1α), and actin as housekeeping genes. Relative expression was calculated using the 2−ΔΔCt method (Livak and Schmittgen 2001), with three biologic replicates for each analysis. GUS enzymatic activity assay Infiltrated leaves were harvested at 6 days post-infiltration (dpi) and homogenized using an extraction buffer containing 50 mM sodium phosphate pH 7.0, 10 mM ꞵ-mercaptoethanol, 10 mM EDTA, and 0.1% Triton X-100 (Sheshukova et al. 2022). Total soluble protein (TSP) content in the supernatants was quantified using the Bradford assay, and GUS enzymatic activity was measured to assess reporter gene expression. GUS activity was determined by a fluorometric assay, where the enzyme catalyzed the conversion of 4-methylumbelliferyl-β-D-glucuronide (4-MUG; Tokyo Chemical Industry, Tokyo, Japan) into the fluorescent compound 4-methylumbelliferone (4-MU; Tokyo Chemical Industry). TSP supernatants were incubated with 4-MUG at 37 °C for 1 h. The reaction was stopped by adding sodium carbonate buffer, and fluorescence was measured using a spectrophotometer (F-7100; Hitachi Ltd, Tokyo, Japan) with an excitation wavelength of 365 nm and an emission wavelength of 455 nm. The fluorescence intensity of 4-MU was used to calculate GUS activity. The mpBI121-GFP construct was included as a negative control. Recovery of apoplastic wash fluid and ASM expression Infiltrated leaves expressing ASM were harvested at 6 dpi and separated into three groups (whole, intracellular, and apoplast samples), following previously described procedures (Panyawechamontri et al. 2025). Optimizing transient expression condition for ASM To optimize ASM expression, infiltration buffer cases were systematically varied (Table 1). Parameters tested included the co-expression or absence of (1) p19 silencing suppressor (OD600 of 0.5), (2) the addition of 100 µM acetosyringone, and (3) 0.56 mM ascorbic acid (Zhao et al. 2017). Eight buffer combinations were tested to evaluate the impact of each on ASM activity. Equal volumes of Agrobacterium cultures containing the ASM construct and the p19 silencing suppressor were mixed in a 1:1 ratio before infiltration. Bacterial cultures were adjusted to an OD600 of 0.5, combined with the respective infiltration buffers, and incubated at 25 °C for 2 h, with either acetosyringone or ascorbic acid included. When ascorbic acid was present in the buffer, the incubation was conducted in the dark to protect the stability of the reagent. Table 1Infiltration buffer components and experimental cases for transient ASM expression in N. benthamiana, including co-expression of p19 and supplementation with acetosyringone and ascorbic acidV12345678mpBI121-ntExtASM−++++++++p19−−+−−++−+Acetosyringone (100 µM)−−−+−+−++Ascorbic acid (0.56 mM)−−−−+−+++mpBI121++−++−−+−V: negative control vector (mpBI121) Table 1Infiltration buffer components and experimental cases for transient ASM expression in N. benthamiana, including co-expression of p19 and supplementation with acetosyringone and ascorbic acidV12345678mpBI121-ntExtASM−++++++++p19−−+−−++−+Acetosyringone (100 µM)−−−+−+−++Ascorbic acid (0.56 mM)−−−−+−+++mpBI121++−++−−+−V: negative control vector (mpBI121) Infiltration buffer components and experimental cases for transient ASM expression in N. benthamiana, including co-expression of p19 and supplementation with acetosyringone and ascorbic acid V: negative control vector (mpBI121) ASM enzymatic activity assay ASM activity was measured using the Acid Sphingomyelinase Assay Kit (K-3200; Echelon Biosciences, Salt Lake City, UT, USA) following the manufacturer’s protocol. This fluorometric assay quantifies the hydrolysis of sphingomyelin by ASM, generating a fluorescent signal proportional to enzyme activity. Reactions were performed on whole, intracellular, and apoplastic samples, which were incubated at 37 °C for 3 h with gentle shaking. Fluorescence was then measured using a fluorescence spectrophotometer (F-7100, Hitachi Ltd). The enzyme activity was calculated as µU/mg of protein. SDS-PAGE and western blot analysis Protein samples were separated via electrophoresis on a 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) resolving gel with a 6% stacking gel. A prestained molecular weight marker (PageRuler™, Cat#26616; Thermo Fisher Scientific) was included to estimate protein size. Gels were stained with Coomassie Brilliant Blue (CBB) (2.5 mg/mL CBB R-250 in 45% methanol and 10% acetic acid). For Western blot analysis, proteins were transferred from gels onto a polyvinylidene difluoride (PVDF) membrane (Millipore-Sigma, Burlington, MA, USA). The membrane was blocked for 1 h at 25 °C with 5% fat-free skim milk (Nacalai Tesque, Kyoto, Japan) in phosphate-buffered saline with Tween 20 (PBST) solution (1.47 mM KH2PO4, 10 mM Na2HPO4, 2.7 mM KCl, 137 mM NaCl, and 0.05% Tween 20, pH 7.4). The membrane was then incubated with the rabbit polyclonal anti-ASM antibody (Cat# GTX114782; GeneTex, CA, USA) diluted 1:2000 in Can Get Signal™ solution I (TOYOBO, Osaka, Japan). Following three 10-min washes with PBST, the membrane was incubated with ECL™ anti-rabbit IgG labeled with HRP (Cat# NA934V; Cytiva, Marlborough, MA, USA) at 1:10,000 in Can Get Signal™ solution II (TOYOBO) for 45 min at 25 °C. Immunoblot images were acquired with the iBright Imaging System 1500 (Invitrogen) after application of the Immobilon Forte Western HRP substrate (Millipore-Sigma). Recombinant human ASM (rhASM) produced in CHO cells (R&D Systems, Inc., Minneapolis, MN, USA) was used as a positive control, while the agroinfiltrated empty vector (pRI201-AN) served as a negative control. Statistical analysis Data were analyzed for statistical significance using one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test. A p value of ≤ 0.05 was considered statistically significant. Each experiment included at least three biologic replicates. Results are presented as the mean ± standard deviation (SD). Results Engineering plant expression vectors for enhanced recombinant protein expression To assess whether reducing the size of the expression vector enhances recombinant protein expression in plants, the original pBI121 plasmid was systematically modified. The lac promoter/operator region, which regulates ꞵ-galactosidase expression in E. coli but is non-functional and unnecessary in plant expression systems, was deleted to eliminate redundant bacterial control elements and reduce plasmid size, generating pBI121Δlac. Subsequently, the tetR region, which encodes the tetracycline repressor protein and confers tetR resistance, was removed because it is not required for selection or maintenance in A. tumefaciens and contains non-essential DNA. This deletion further minimized the vector, resulting in pBI121ΔlacΔtetR, hereafter referred to as mpBI121. The minimized vector is illustrated in Fig. 1A. Antibiotic sensitivity assays were performed using A. tumefaciens strains harboring pBI121, pBI121Δlac, and mpBI121 to confirm successful deletions of non-essential elements. All transformed strains grew on plates containing kanamycin, streptomycin, and rifampicin, confirming successful transformation. However, only strains carrying pBI121 and pBI121Δlac grew on tetracycline-containing plates, while those harboring mpBI121 did not, indicating the complete removal of the tetR gene. Together, these results validate the precise removal of the lac promoter/operator and tetR regions and clearly demonstrate a substantial reduction in vector size, representing a decisive advance toward a more streamlined and efficient plant expression platform.Fig. 1Vector construction and expression analysis. A Combined schematic representation of minimized vectors (pBI121, pBI121Δlac, mpBI121) and diagrams of corresponding constructs containing the ntExtASM cassette (pBI121-ntExtASM, pBI121Δlac-ntExtASM, mpBI121-ntExtASM). B Schematic map of the mpBI121-GFP construct. C GFP expression in N. benthamiana leaves agroinfiltrated with mpBI121-GFP, visualized under blue light at 0 and 6 dpi, with wild-type (non-infiltrated) and negative control vector (mpBI121) leaves shown for comparison Vector construction and expression analysis. A Combined schematic representation of minimized vectors (pBI121, pBI121Δlac, mpBI121) and diagrams of corresponding constructs containing the ntExtASM cassette (pBI121-ntExtASM, pBI121Δlac-ntExtASM, mpBI121-ntExtASM). B Schematic map of the mpBI121-GFP construct. C GFP expression in N. benthamiana leaves agroinfiltrated with mpBI121-GFP, visualized under blue light at 0 and 6 dpi, with wild-type (non-infiltrated) and negative control vector (mpBI121) leaves shown for comparison The functionality of the mpBI121 expression vector was assessed by replacing the GUS reporter gene with GFP, thereby generating the mpBI121-GFP construct (Fig. 1B). To evaluate whether the reduced plasmid size affected expression efficiency, the mpBI121-GFP and mpBI121 constructs were agroinfiltrated into N. benthamiana leaves, with wild-type (non-infiltrated) leaves serving as controls. Fluorescence was assessed at day 0 (immediately after infiltration) and at 6 days post-infiltration (dpi). As shown in Fig. 1C, no GFP fluorescence was detected at day 0 in either the infiltrated or wild-type leaves. However, by 6 dpi, strong GFP fluorescence was visible only in the infiltrated leaves, while wild-type and mpBI121-infiltrated leaves showed no fluorescence. These results demonstrate that the mpBI121-GFP construct is capable of driving gene expression and that the reduced plasmid size does not affect the vector’s functionality. The mpBI121 backbone therefore remains fully functional for plant gene expression studies. After confirming that the minimized mpBI121 vector remained fully functional for gene expression, we explored how changes to the vector backbone and regulatory sequences might influence expression efficiency in plants. For this purpose, the GUS reporter gene in pBI121, pBI121Δlac, and mpBI121 was replaced with the ntExtASM coding sequence fused to the AtADH 5′ UTR, creating updated expression cassettes. To further optimize transcript processing and stability, the native NOS terminator was subsequently replaced with the HSP terminator. These modified cassettes were successfully subcloned into the three binary vectors, resulting in the constructs pBI121-ntExtASM, pBI121Δlac-ntExtASM, and mpBI121-ntExtASM (Fig. 1A). These constructs were designed to evaluate how vector minimization and regulatory modifications affect transgene expression in plants. Reduced vector size enhances expression at both mRNA and protein levels in N. benthamiana To assess the impact of vector size on transgene expression, we analyzed mRNA expression levels of the GUS gene in agroinfiltrated N. benthamiana leaves using RT-qPCR. Three constructs—pBI121, pBI121Δlac, and mpBI121—were compared, with mpBI121-GFP serving as a negative control. Gene expression was normalized using three plant housekeeping genes [18S, EF1α, and actin (Li et al. 2016)], all of which exhibited consistent trends across samples. Compared to pBI121, GUS expression increased approximately 1.30-fold with pBI121Δlac and 2.02-fold with mpBI121 (Fig. 2A), consistent with slightly lower ΔCt values observed for these constructs. Corresponding GUS enzymatic activity followed the same pattern, increasing from 23.20 ± 0.36 nU/mg in pBI121 to 31 ± 2.2 nU/mg (1.33-fold) with pBI121Δlac and to 40.8 ± 1.4 nU/mg (1.76-fold) with mpBI121 (Fig. 2B). Notably, low GUS activity was observed in the control (mpBI121-GFP) samples, likely reflecting endogenous GUS-like activity in N. benthamiana (Alwen et al. 1992).Fig. 2mRNA and protein expression analysis. A The relative levels of GUS mRNA were measured by RT-qPCR, normalized to plant housekeeping genes: 18S, EF1α, and actin. B GUS activity in leaf extracts at 6 dpi was measured using different vector constructs. G: mpBI121-GFP; P: pBI121; L: pBI121Δlac; M: mpBI121. C Relative expression of ASM mRNA was also quantified by RT-qPCR and normalized to housekeeping genes. D ASM enzyme activity in leaf extracts at 6 dpi from the different constructs: M: mpBI121; PA: pBI121-ntExtASM; LA: pBI121Δlac-ntExtASM; MA: mpBI121-ntExtASM. In the graph, bars labeled with different letters represent significant differences, all of which are highly significant (p < 0.001). E CBB staining and western blot analysis of ASM expression were performed using different vectors, with CHO-derived human ASM (5 ng) as a control (C), probed with rabbit anti-human ASM (1: 2,000). The enzyme activities of GUS and ASM are expressed as nU/mg and µU/mg, respectively. The values represent the mean ± SD from three independent experiments mRNA and protein expression analysis. A The relative levels of GUS mRNA were measured by RT-qPCR, normalized to plant housekeeping genes: 18S, EF1α, and actin. B GUS activity in leaf extracts at 6 dpi was measured using different vector constructs. G: mpBI121-GFP; P: pBI121; L: pBI121Δlac; M: mpBI121. C Relative expression of ASM mRNA was also quantified by RT-qPCR and normalized to housekeeping genes. D ASM enzyme activity in leaf extracts at 6 dpi from the different constructs: M: mpBI121; PA: pBI121-ntExtASM; LA: pBI121Δlac-ntExtASM; MA: mpBI121-ntExtASM. In the graph, bars labeled with different letters represent significant differences, all of which are highly significant (p < 0.001). E CBB staining and western blot analysis of ASM expression were performed using different vectors, with CHO-derived human ASM (5 ng) as a control (C), probed with rabbit anti-human ASM (1: 2,000). The enzyme activities of GUS and ASM are expressed as nU/mg and µU/mg, respectively. The values represent the mean ± SD from three independent experiments To validate these findings, we evaluated a second target gene, ASM, using equivalent construct designs. RT-qPCR analysis revealed a clear, size-dependent increase in ASM transcripts: a 1.36-fold increase with pBI121Δlac-ntExtASM and a 2.05-fold increase with mpBI121-ntExtASM relative to pBI121-ntExtASM (Fig. 2C). No ASM expression was detected in the mpBI121 negative control vector. ASM enzyme activity also increased in parallel with transcript abundance, rising from 72 ± 3.0 µU/mg with the larger vector to 103 ± 1.67 (1.43-fold) and 125 ± 3.4 (1.75-fold) µU/mg with the minimized constructs (Fig. 2D). Protein expression of ASM was confirmed by SDS-PAGE and western blot analysis. Clear bands of about 61 kDa were observed at the expected size, confirming that all constructs produced the target protein (Fig. 2E). All fold-change data from mRNA expression analysis of both GUS and ASM are summarized in Table 2, while the corresponding enzymatic activity data for GUS and ASM are presented in Table 3. These results consistently demonstrate that smaller binary vector constructs enhance both mRNA and protein output of target genes, supporting the conclusion that vector minimization can improve transgene expression in plant systems. Table 2Fold changes in relative mRNA expression of GUS and ASM normalized to endogenous plant controlsTarget genesConstructsFold change relative to control18SEF1αactinGUSpBI1211.02 ± 0.231.03 ± 0.261.00 ± 0.07pBI121Δlac1.30 ± 0.271.32 ± 0.191.28 ± 0.12mpBI1212.00 ± 0.142.06 ± 0.042.01 ± 0.08ASMpBI121-ntExtASM1.00 ± 0.041.00 ± 0.030.96 ± 0.11pBI121Δlac-ntExtASM1.35 ± 0.051.39 ± 0.231.34 ± 0.04mpBI121-ntExtASM2.03 ± 0.332.10 ± 0.092.03 ± 0.05Table 3Summary of the enzyme activity of GUS and ASM in crude extracts from various constructsTarget proteinsCrude extractsEnzyme activityUnitsGUSmpBI121-GFP11.11 ± 0.49nU/mgpBI12123.20 ± 0.36pBI121Δlac30.9 ± 2.2mpBI12140.8 ± 1.4ASMmpBI12141.8 ± 1.5µU/mgpBI121-ntExtASM71.8 ± 3.0pBI121Δlac-ntExtASM103 ± 1.7mpBI121-ntExtASM125 ± 3.4 Table 2Fold changes in relative mRNA expression of GUS and ASM normalized to endogenous plant controlsTarget genesConstructsFold change relative to control18SEF1αactinGUSpBI1211.02 ± 0.231.03 ± 0.261.00 ± 0.07pBI121Δlac1.30 ± 0.271.32 ± 0.191.28 ± 0.12mpBI1212.00 ± 0.142.06 ± 0.042.01 ± 0.08ASMpBI121-ntExtASM1.00 ± 0.041.00 ± 0.030.96 ± 0.11pBI121Δlac-ntExtASM1.35 ± 0.051.39 ± 0.231.34 ± 0.04mpBI121-ntExtASM2.03 ± 0.332.10 ± 0.092.03 ± 0.05 Fold changes in relative mRNA expression of GUS and ASM normalized to endogenous plant controls Table 3Summary of the enzyme activity of GUS and ASM in crude extracts from various constructsTarget proteinsCrude extractsEnzyme activityUnitsGUSmpBI121-GFP11.11 ± 0.49nU/mgpBI12123.20 ± 0.36pBI121Δlac30.9 ± 2.2mpBI12140.8 ± 1.4ASMmpBI12141.8 ± 1.5µU/mgpBI121-ntExtASM71.8 ± 3.0pBI121Δlac-ntExtASM103 ± 1.7mpBI121-ntExtASM125 ± 3.4 Summary of the enzyme activity of GUS and ASM in crude extracts from various constructs Detection and recovery of ASM from apoplastic wash fluid The localization and secretion of ASM were investigated in Nicotiana benthamiana leaves infiltrated with a construct encoding ntExtASM, designed to direct the protein to the apoplast. Following agroinfiltration, leaf tissues were fractionated into whole extract, intracellular fractions, and apoplastic wash fluid. Each fraction was analyzed by SDS-PAGE, Western blotting, and enzyme activity assays to determine the presence and functionality of the expressed protein. A distinct ASM-specific band was detected in the apoplast fraction by western blot, while only a faint signal was observed in the intracellular fraction relative to the whole extract (Fig. 3A). This pattern matched the enzymatic activity assay results, which showed the highest ASM activity in the apoplastic wash fluid (Table 4), moderate activity in the whole sample, and minimal activity in the intracellular sample (Fig. 3B).Fig. 3Recovery of ASM from the apoplast using the mpBI121-ntExtASM construct. A CBB staining and Western blot analysis showing ASM protein in different fractions. W (whole leaf extract), I (intracellular extract), A (apoplastic wash fluid), C (CHO-derived human ASM, 5 ng), were detected using rabbit anti-human ASM antibody (1: 2,000). B Enzymatic activity of ASM measured in the same fractions (W, I, A), confirming successful localization and recovery of functional ASM from the apoplast. Different letters above the bars indicate statistically significant differences, with all comparisons showing high significance (p < 0.001)Table 4Comparison of ASM activity in apoplast fractions between mpBI121 and mpBI121-ntExtASM vectorsCrude extractsASM activity (µU/mg)WholeIntracellularApoplastmpBI12141.8 ± 1.537.05 ± 0.5030.8 ± 4.3mpBI121-ntExtASM125 ± 3.441.2 ± 1.8188.2 ± 3.6 Recovery of ASM from the apoplast using the mpBI121-ntExtASM construct. A CBB staining and Western blot analysis showing ASM protein in different fractions. W (whole leaf extract), I (intracellular extract), A (apoplastic wash fluid), C (CHO-derived human ASM, 5 ng), were detected using rabbit anti-human ASM antibody (1: 2,000). B Enzymatic activity of ASM measured in the same fractions (W, I, A), confirming successful localization and recovery of functional ASM from the apoplast. Different letters above the bars indicate statistically significant differences, with all comparisons showing high significance (p < 0.001) Table 4Comparison of ASM activity in apoplast fractions between mpBI121 and mpBI121-ntExtASM vectorsCrude extractsASM activity (µU/mg)WholeIntracellularApoplastmpBI12141.8 ± 1.537.05 ± 0.5030.8 ± 4.3mpBI121-ntExtASM125 ± 3.441.2 ± 1.8188.2 ± 3.6 Comparison of ASM activity in apoplast fractions between mpBI121 and mpBI121-ntExtASM vectors These results confirm that ASM is successfully secreted to the apoplast, demonstrating the functional effectiveness of the extensin signal peptide. The low intracellular levels further support efficient targeting and minimal retention within the cell. Moreover, the successful recovery of active ASM from the apoplast highlights the reliability of the apoplastic recovery method. Overall, these findings demonstrate that the engineered construct enables efficient secretion and accumulation of functional ASM in the apoplast, offering a practical strategy for targeting recombinant proteins to extracellular compartments in plant systems. Optimization of infiltration buffer components for enhanced protein expression Eight infiltration buffer cases (Table 1) were assessed to determine optimal conditions for ASM production in N. benthamiana. The conditions varied in p19 co-expression, acetosyringone (100 µM), and ascorbic acid (0.56 mM), with the empty vector (mpBI121; V) serving as a negative control vector. ASM activity was predominantly observed in the AWF (apoplastic wash fluid) across all conditions, confirming apoplastic secretion of the recombinant protein (Fig. 4A). The highest ASM levels were associated with the four conditions that included p19 co-expression (cases 5–8), highlighting its critical role in enhancing transient expression by suppressing RNA silencing mechanisms. Case 8, which included p19 co-expression along with acetosyringone and ascorbic acid, produced the highest ASM activity in the apoplast (546 ± 4.3 µU/mg), identifying it as the most effective infiltration buffer composition in this study. This corresponded to a 2.9-fold increase compared with the standard buffer without enhancers (case 1; 190 ± 11 µU/mg), a 16.6-fold increase relative to the vector-only control (33 ± 10 µU/mg), and approximately a 1.2-fold improvement over the next most effective condition (case 5; 523 ± 7 µU/mg). The enhanced performance of case 8 likely results from synergistic mechanisms: p19 suppresses RNA silencing, acetosyringone enhances Agrobacterium virulence gene expression and improves T-DNA delivery, and ascorbic acid reduces oxidative stress, stabilizing the recombinant protein. SDS-PAGE analysis of apoplastic extracts (Fig. 4B) supported these findings, showing the strongest 61 kDa band in case 8, consistent with its elevated ASM activity. These results emphasize the significance of optimizing infiltration buffer components to improve transient protein expression efficiency in N. benthamiana, a key factor for achieving high-protein output in plant-based expression systems. All ASM activity data for all conditions are listed in Table 5.Fig. 4Optimization of the infiltrated buffer for transient ASM expression in N. benthamiana. A Comparison of ASM enzyme activity under different buffer cases. Each data point reflects the average ± SD from three independent biologic replicates. Bars marked with different letters denote significant differences between groups (p < 0.001). B CBB staining and Western blot analysis of whole, intracellular, and apoplast samples. C: CHO-derived human ASM used as a positive control; V: mpBI121 construct used as a negative control vector, probed with rabbit anti-human ASM (1: 2,000)Table 5Summary of ASM activity under different infiltration buffer conditions during agroinfiltrationConditionsASM activity (µU/mg)WholeIntracellularApoplastmpBI121 (V)38.95 ± 5.436 ± 8.833 ± 9.8Case 1120 ± 4.444 ± 5.4190 ± 11Case 2255 ± 2.965 ± 11456 ± 12Case 3183 ± 1250 ± 5.3373 ± 6.4Case 4148 ± 3.847 ± 8.3250 ± 9.1Case 5284 ± 2.877 ± 10523 ± 7.4Case 6270 ± 7.369 ± 6.6496 ± 9.3Case 7237 ± 1353 ± 11413 ± 8.2Case 8314 ± 2.994 ± 6.1546 ± 4.3 Optimization of the infiltrated buffer for transient ASM expression in N. benthamiana. A Comparison of ASM enzyme activity under different buffer cases. Each data point reflects the average ± SD from three independent biologic replicates. Bars marked with different letters denote significant differences between groups (p < 0.001). B CBB staining and Western blot analysis of whole, intracellular, and apoplast samples. C: CHO-derived human ASM used as a positive control; V: mpBI121 construct used as a negative control vector, probed with rabbit anti-human ASM (1: 2,000) Table 5Summary of ASM activity under different infiltration buffer conditions during agroinfiltrationConditionsASM activity (µU/mg)WholeIntracellularApoplastmpBI121 (V)38.95 ± 5.436 ± 8.833 ± 9.8Case 1120 ± 4.444 ± 5.4190 ± 11Case 2255 ± 2.965 ± 11456 ± 12Case 3183 ± 1250 ± 5.3373 ± 6.4Case 4148 ± 3.847 ± 8.3250 ± 9.1Case 5284 ± 2.877 ± 10523 ± 7.4Case 6270 ± 7.369 ± 6.6496 ± 9.3Case 7237 ± 1353 ± 11413 ± 8.2Case 8314 ± 2.994 ± 6.1546 ± 4.3 Summary of ASM activity under different infiltration buffer conditions during agroinfiltration Discussion The development of plant-based expression systems for recombinant protein production has become a central focus in biotechnology, owing to their cost-effective scalability and capacity to produce complex proteins that are challenging to express in other systems (Buyel 2025). However, optimizing key parameters, such as vector design and transformation efficiency, remains a major barrier to improving protein output and quality. In this study, we demonstrate that minimizing the size of the expression vector substantially improves protein expression in N. benthamiana, a widely used host for transient expression. Our results show that reducing the binary vector size from pBI121 to the smaller mpBI121 significantly increased recombinant protein expression by increasing both mRNA levels and protein production levels for the target genes (GUS and ASM). Importantly, although our data clearly demonstrate improved expression following vector minimization, the present study does not allow us to definitively distinguish whether this enhancement is driven primarily by overall vector size reduction or by the removal of specific non-essential bacterial sequences (e.g., lac- or tetR-related elements). The observed effect is therefore likely attributable to a combination of size-dependent and sequence-specific factors. This finding aligns with previous studies, which have shown that smaller, simplified vectors improve cloning efficiency, facilitate more effective T-DNA delivery, and ultimately enhance protein output (Beritza et al. 2024). Recent advancements in plant molecular pharming also suggest that compact binary vectors, when paired with optimized constructs and delivery methods, can outperform larger systems, including viral replicon-based platforms, in terms of robustness and scalability (Sainsbury et al. 2009; Szakasits et al. 2007). Mechanistically, larger vectors impose multiple constraints on transformation efficiency due to difficulties in transferring, replicating, and maintaining DNA in both bacterial and plant cells (Chan et al. 2002). In plant cells, the presence of larger vectors imposes greater demands on the host’s cellular machinery, as the vector competes for resources needed for DNA replication and transcription, leading to reduced T-DNA stability and lower levels of transgene output (Frary and Hamilton 2001). In contrast, smaller vectors promote higher T-strand accumulation and more efficient activation of virulence (vir) genes, especially with acetosyringone induction, thereby enhancing T-DNA mobilization (Xi et al. 2018). Supporting this, previous work has shown that small, high-efficiency binary Ti vectors, such as the pLSU series with co-directional replicons, achieve higher transformation frequencies than larger vectors (Lee et al. 2012). By trimming non-essential sequences outside the T-DNA borders while keeping the internal regions intact, smaller vectors enhance T-strand accumulation and T-DNA transfer, highlighting the impact of overall vector size on transformation performance. While mRNA accumulation occurs after T-DNA transfer and integration, reducing vector size may increase the stability of T-DNA transfer, leading to a higher number of transcriptionally active copies in plant cells, which likely explains the observed increases in mRNA and protein output. Specifically, the minimized mpBI121 vector resulted in a 1.76-fold increase in GUS activity and a 1.75-fold increase in ASM activity compared to the standard pBI121 vector. These results highlight the importance of vector size in reducing the metabolic burden on both Agrobacterium and the plant host, improving transformation efficiency, and promoting more robust protein output. Moreover, our approach, which retains essential elements, such as the 5’ UTR and HSP terminator, ensures functional expression without requiring additional regulatory sequences. This strategy not only confirms the benefits of vector minimization but also provides a practical framework for optimizing plant-based expression systems, strengthening the link between laboratory optimization and industrial protein manufacturing. The composition of the infiltration buffer is another key determinant of transient protein expression. Among the eight buffer cases tested, the combination of p19, acetosyringone, and ascorbic acid consistently yielded the highest ASM expression. This result supports previous studies showing that p19, a viral suppressor of RNA silencing, suppresses post-transcriptional gene silencing (PTGS) to stabilize gene transcripts and allow for sustained protein production (Garabagi et al. 2012; Voinnet et al. 2003). Acetosyringone enhances transformation efficiency by activating Agrobacterium vir genes, promoting efficient T-DNA transfer (Leuzinger et al. 2013), while ascorbic acid reduces oxidative stress during infiltration, improving protein integrity and accumulation (Akram et al. 2017; Ali et al. 2019; Wang et al. 2024). These results expand the current understanding of buffer-driven enhancement in transient expression and provide practical guidance for improving protein output in plant biotechnology. The combination of these buffer components in our study produced the strongest ASM expression observed, underscoring the importance of optimizing multiple parameters in parallel to achieve maximal protein production. While our results are promising, several areas await future research. Although we focused on two recombinant proteins, GUS and ASM, it would be valuable to investigate whether these optimization strategies generalize to other classes of proteins, such as membrane-bound or glycosylated proteins, which may need additional adjustments to the expression system. Furthermore, the scalability of these approaches must be evaluated under conditions that mimic industrial production pipelines. Future studies should assess how well these strategies perform under such conditions and in larger volumes. A deeper understanding of how the buffer components interact with Agrobacterium and the plant host will support additional refinements to the system and further increases in expression efficiency. Additionally, exploring alternative plant species and expression platforms could open new possibilities for improving protein production across diverse applications. While our study focused on eliminating non-essential sequences outside the T-DNA borders, emerging work suggests that optimizing the border-region and internal T-DNA elements can further improve T-DNA delivery and gene expression (Lee et al. 2023; Wang et al. 2023). Investigating such internal modifications in combination with vector minimization could yield additional gains in expression efficiency. Addressing these areas will be key to translating these laboratory-scale advances into reliable systems for industrial-scale protein production. Conclusion This study demonstrates that reducing the size of the expression vector enhances recombinant protein expression in Nicotiana benthamiana. By using two representative target proteins, GUS and ASM, we observed consistent increases in both mRNA accumulation and protein expression when smaller vectors were used. Additionally, we found that optimizing the infiltration buffer, particularly by including p19, acetosyringone, and ascorbic acid, further boosted ASM expression levels. These findings demonstrate that both vector minimization and buffer optimization significantly improve transformation performance and protein output. This dual strategy offers a practical and scalable approach for enhancing plant-based expression systems, supporting applications in the production of therapeutic proteins, enzymes, and vaccines. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 18 KB) Supplementary file1 (DOCX 18 KB)