ELISA analysis was used to quantify the CMG2-Fc protein that was transiently expressed within the leaves of the N. benthamiana intact plants (Figure 1). Bradford assays were also performed to quantify the levels of total soluble protein (TSP) expressed within the N. benthamiana intact plants. Values obtained from the Bradford assays were combined with ELISA data to determine the CMG2-Fc expression level as a percentage of the TSP. The effect of co-expressing certain suppressor proteins on the CMG2-Fc expression level was dramatic, when compared to the baseline agroinfiltration in which CMG2-Fc was tested without co-expression of the PTGS suppressor. Transient expression in intact plants was observed to be higher with p1 co-expression at 3.5 days, resulting in a maximum average production of 0.56 g PBI-220 per kg of leaf fresh weight (FW) and 1.5% of the TSP. With p1, the CMG2-Fc expression level increased ten-fold when compared to the absence of suppressor co-expression, and was statistically different at a 95% confidence level. Co-expression with the p25 PTGS suppressor was also found to be significantly different from the results obtained in the absence of suppressor at 3.5 days. The rest of the suppressor proteins tested had no demonstrated effect on CMG2-Fc yield, since they showed comparable expression levels that did not statistically differ from CMG2-Fc without co-expression. As expected, the negative control consisting of wild type plants infiltrated only with buffer that did not contain Agrobacterium did not show CMG2-Fc expression.

CMG2-Fc expression in intact plants is also evident based on the Western Blot analysis (Figure 2). A major immunoblot band migrated as expected to approximately 47.2 kDa. Dimeric and trimeric bands can also be observed at approximately 94.4 kDa and 141.6 kDa, respectively. It is expected that full-length CMG2-Fc will run a bit higher on a SDS-PAGE because the protein has one N-glycosylation site. The values measured with ELISA allowed us to predict the amount of CMG2-Fc loaded in each lane for the Western analysis (Table 2). In order to compare the relative band intensities, 2000 ng of purified CMG2-Fc were loaded as a positive control, corresponding to an amount of 1 g/kg FW leaf. Western Blot analysis confirmed that CMG2-Fc expression at 3.5 days was strong when co-expressed with p1. The negative control infiltration did not result in any immunoreactive bands.

CMG2-Fc was also quantified from the samples obtained from N. benthamiana detached leaves using ELISA (Figure 3). In this case, the best performer was p19 co-expression at 14 DPI, which showed an average expression level of 0.39 g CMG2-Fc per kg of leaf FW, corresponding to 2% of the TSP. This is a five-fold increase when compared to the absence of suppressor co-expression, but due to the variability observed, was not found to be significantly different using a p-value of 0.05. The 2b suppressor protein slightly increased the CMG2-Fc expression level, although not at a statistically significant level. The negative control infiltration showed no CMG2-Fc expression as expected.

Western Blot analysis proved that CMG2-Fc was also expressed correctly in the N. benthamiana detached leaves (Figure 2). Immunoblot bands can be observed exclusively at the 47.2 kDa level. Western Blot also confirms that the protein of interest was expressed under p19 co-expression, and that CMG2-Fc was present in the detached leaves even at 14 DPI. We also detected CMG2-Fc expression at 3.5, 7, and 14 DPI in both intact plants and detached leaves for several of the other suppressor proteins tested (data not shown). No bands appeared on the negative control lane.

These results show that suppressing the post-transcriptional gene silencing mechanism in N. benthamiana increased the expression level of a recombinant protein expressed transiently under the control of the CaMV 35S promoter. While co-expression of certain suppressor proteins enhanced CMG2-Fc expression, other suppressor proteins had no apparent positive effect on CMG2-Fc accumulation. There have been reports in the literature in which PTGS suppression was not able to increase the expression level of a plant-made recombinant protein under the control of the CaMV 35S promoter [36]. Therefore, it is important to complete the type of screening analysis performed in this project for each new plant system and/or target product of interest, in order to determine the conditions that optimize recombinant protein yields.

Wycoff et al. [37] recently reported a high expression level of CMG2-Fc in Nicotiana benthamiana. Optimized expression enabled these researchers to obtain 0.73 g of CMG2-Fc per kg of leaf FW using the p19 gene silencing suppressor, the highest level reported for this fusion protein. Wycoff et al. evaluated different combinations of CMG2 and IgG1 to develop an improved CMG2-Fc construct. The CMG2-Fc construct utilized in the project reported here is not the current optimized construct, hence the lower expression of 0.56 g of CMG2-Fc per kg of leaf FW, even with gene silencing suppressor co-expression. Regardless, this project provides comparative data on PTGS suppressor co-expression for production of CMG2-Fc, which complements the construct optimization work of Wycoff et al. Both approaches will be important for further optimization of CMG2 production towards expression levels on the order of gram of protein per kilogram of N. benthamiana leaf tissue.

No common trends were identified for the CMG2-Fc expression kinetics using different gene silencing suppressors, which may be due to the diversity of the suppressor proteins tested. Different plant viruses evolved independently to counter the antiviral defense of the plants, as evidenced by their different modes of action against the gene silencing phenomenon. The fact that the suppressor proteins tested in this study target different steps of the RNA interference mechanism may explain the differences observed in the kinetic data. Three of the best performers in this experiment (p19, p21, and 2b) bind dsRNA, which may indicate that pathway specific effects are in play. The p19 protein from tomato bushy stunt virus is a strong silencing suppressor protein and has a unique mechanism by selectively binding to the 21-nt and 22-nt classes of siRNA. The p21 protein from Yellow Beet Virus and the 2b protein from Cucumber Mosaic Virus are both silencing suppressor proteins that can bind siRNA and longer dsRNA without a preference for 21-nt siRNAs. Meanwhile, the p1 protein from Rice Yellow Mottle Virus, which increased CMG2-Fc expression the most when compared to absence of suppressor, has a different mechanism. It inhibits the accumulation of the 24-nt siRNA, but has a less pronounced effect on the accumulation of the 21-nt siRNA [38].

Another possible explanation for the differences in the data may be the vectors used in this experiment. pBIN (11.8 kb) and pCB302 (3.5 kb) are two of the most widely used vectors for Agrobacterium-mediated plant transformation in plants. There is sequence homology between both vectors due to the fact that pCB302 was developed by removing base pairs that were deemed non-essential on pBIN. pCB302 retains all essential elements of pBIN, and its smaller size makes it more versatile for cloning [39]. Still, using two different vectors to drive the expression of the suppressors may have had an effect on the expression level of the suppressors, and therefore, on the accumulation of CMG2-Fc. In fact, the highest expression levels obtained in the experiment were from suppressors that were inserted on pBIN (p1, p19, p25, and 2b), so it is possible that pBIN could be a more suitable vector for this particular system.

During the agroinfection process in plants, the antagonistic forces of plant pathogen invasion and defense are governed by a dynamic set of biological components. There are two possibilities that may explain the observation that the expression level of a CaMV 35S driven recombinant protein is being increased by suppression of PTGS. The first possibility could be that Agrobacterium may be activating the gene silencing mechanism in the plant. Agrobacterium, acting as the plant pathogen in Nicotiana benthamiana, may be producing protein effectors that elicit the production of small RNAs in the plant. The plant might be deploying a gene silencing approach to counter the Agrobacterium infection, and co-expression of gene silencing suppressors may be counteracting the gene silencing response. Previous studies have shown that bacterial effectors can act as suppressors [40] of gene silencing, but no instance has been reported in which these proteins can act as gene silencing activators. Interestingly, a dual function of activation and suppression of gene silencing has been recently found for a viral protein previously identified as a gene silencing suppressor [41]. It has been established that infiltration with Agrobacterium elicits a distinct host response in plants, but the mechanisms are still not well understood [42].

The second possibility for explaining our observations is that the overexpression of a foreign gene is inducing the PTGS mechanism, through a quantitative effect exerted by the RNA derived from the transgene. PTGS has been identified as a plant defense mechanism for aberrant gene expression following agroinfection, such as production of excessive RNA transcripts over a threshold level [43,44]. Several groups have reported that overexpressing a transgene over a certain level causes transgene silencing through activation of PTGS [45–47], including when occurring under CaMV 35S promoter control [48,49]. This evidence supports a theory that PTGS occurs due to the elevated CaMV 35S driven CMG2-Fc expression, regardless of the absence of viral RNAs. Under these conditions, CMG2-Fc expression can be increased when co-expressed with gene silencing suppressors.

It is clear that CMG2-Fc levels change over time, so the same N. benthamiana leaves were sampled along all time points post-infiltration to study the kinetics of production. For some of the transient expression experiments, a high degree of variability was observed between leaves, which is consistent with what others have observed [50]. This variability arose from the differences between the sample replicates (leaves are sampled from different plants), and not from the assays used for protein quantification. It is possible that variations in protein expression are due to the application technique, the sampling method, or protein stability issues. Due to position effects within the leaves, there is inherent variability in the expression of recombinant proteins in plants. A study by Sheludko et al. [51] evaluated sources of variability in the transient expression of a recombinant protein in Nicotiana plants, and determined that factors such as the age of the plant and the location of the leaf within the plant can cause differences in expression. Applying Agrobacterium through syringe infiltration may reduce variability compared with vacuum infiltration, although it is not a method that could be used at large scale. Variability may be reduced by sampling the entire leaf and measuring protein expression from that extract, but this approach would preclude the possibility of examining the expression kinetics. Finally, denaturation, degradation, unfolding, or cleavage of the protein could also have an impact on how much of the protein is measured, as reported for other proteins expressed transiently [52].

A DNA fragment encoding a chimeric human IgG1/IgG2 Fc region (include the upper and middle hinge sequences of human IgG2, followed by the lower hinge, CH2, and CH3 regions of IgG1) was codon-optimized for Nicotiana tabacum expression. Nucleotides encoding the peptide sequence SEKDEL were appended onto the 3′ end to enable accumulation in the plant’s endoplasmic reticulum. This Fc2/1-KDEL sequence was cloned into pMSP-CMG2-Fc_G [37] between the SacI and EcoRI sites, replacing the Fc_G sequence (non-chimeric IgG1 Fc). An AscI digested fragment from pMSP-CMG2-Fc2/1-KDEL was sub-cloned into the Agrobacterium binary vector pBIN+ocs [53], forming pBIN-CMG2-Fc2/1-KDEL. The final expression cassette was comprised of the CaMV 35S promoter, a short 5′ untranslated region from the Nicotiana sylvestris psaDb gene (GenBank E12115), the signal peptide of the 2S albumin storage protein of Arabidopsis thaliana (GenBank P15458) [54], Nicotiana spp. codon-optimized CMG2 gene (amino acids 34 to 220, GenBank AY233452, Fc2/1-KDEL (described above) and the agrocinopine synthase (ags) terminator (GenBank EU181145). The resultant CMG2-Fc2/1-KDEL protein contains 424 amino acids (excluding the signal peptide) having a subunit molecular mass of 47.2 kDa (dimer of 94.4 kDa). The pBIN-CMG2-Fc2/1-KDEL binary expression vector was transformed into A. tumefaciens EHA105 via electroporation [55].

Each of the plant viral suppressor sequences from p1, p19, p25, and 2b were inserted into the pBIN binary vector. Likewise, the DNA sequences of p10, p21, p24, p38, and HcPro were cloned into a mini-binary vector, pCB302 [39]. All of the gene silencing suppressors used in this study were under the control of the CaMV 35S promoter. The binary vectors containing gene silencing suppressor sequences were transformed into A. tumefaciens EHA105, with the helper plasmid (pCH32) by electroporation.

Wild type non-transgenic N. benthamiana was grown from seeds planted in soiled-filled seed trays. Seedlings were transplanted into individual 4-inch pots. A week after transplantation, Osmocote fertilizer (Scotts Miracle-Gro Company, Marysville, OH) was added to ensure optimum health throughout the growth cycle. Plants that were 4–6 weeks old were used for agroinfiltration. In the case of detached leaf agroinfiltration, upper leaves from 4–6 weeks old N. benthamiana plants were harvested immediately before agroinfiltration. All plants were grown in a greenhouse with a 16-h photoperiod and a temperature range of 18 °C (night time low) to 30 °C (day time high).

Agrobacterium cultures were started from −80 °C frozen stocks, and were grown in 5 mL of LB broth. After incubation for 24 h at 28 °C and 250 rpm, the 5 mL cultures were transferred to 250 mL Erlenmeyer flasks containing 125 mL sterile LB media. After incubating for 24 h at 28 °C and 250 rpm, the Agrobacterium cells were harvested by centrifugation at 1890 g for 10 min. Cells were resuspended in an infiltration buffer, containing 100 mM acetosyringone (3′,5′-dimethoxy-4′-hydroxyacetophenone) (Aldrich Chemicals, Milwaukee, WI), 1 mL of 1 M 2-(4-morpholino) ethanesulfonic acid (MES) buffer (pH 5.6), and 10 mM magnesium chloride. The cell density (optical density, OD₆₀₀ obtained using a Milton Roy Spectronic 501 spectrophotometer) of the agroinfiltration solution was adjusted to 0.5 absorbance units for processing intact plants. For agroinfiltration of detached leaves, the agroinfiltration solution was adjusted to 1.0 absorbance units. An Agrobacterium suspension containing the CMG2-Fc construct was mixed 1:1 with an Agrobacterium suspension containing each of the gene silencing suppressor constructs. For the absence of co-expression condition, the suspension was composed solely of Agrobacterium encoding CMG2-Fc. Infiltration buffer without Agrobacterium was used as a negative control for CMG2-Fc expression. All infiltration solutions were incubated for three h in the dark at room temperature. Directly before performing the agroinfiltration, the surfactant Silwet L-77 (GE Silicones, Friendly, WV) was added at a final concentration of 0.01% to reduce the surface tension of the solution and facilitate the agroinfiltration.

Whole N. benthamiana plants or detached N. benthamiana leaves were submerged into the bacterial suspension and subjected to a vacuum of −90 kPa for 1 min, with occasional agitation to release trapped air bubbles. After the vacuum was rapidly released, the plant material was removed from the suspension and allowed to dry. Intact plants were placed in a greenhouse, while detached leaves were placed inside plastic containers prepared as described elsewhere [56].

Leaf discs with a diameter of 14 mm were sampled from agroinfiltrated leaves after 3.5, 7, and 14 days post-infiltration (DPI). Each sample was comprised of two discs obtained from the same leaf. For each condition tested, three individual leaves, from three different N. benthamiana plants, were used for analysis. Phosphate buffer saline (PBS) composed of 137 mM NaCl, 2.7 mM KCl, 10.1 mM Na₂HPO₄ and 1.8 mM KH₂PO₄ (pH 7.4) was used for extraction. PBS was added to the leaf discs at a ratio of 10 μL per mg of fresh weight leaf tissue. Cells were lysed on ice using a plastic pestle attached to a drill (Makita). The cell lysate was clarified by centrifugation at 4 °C and 10,600 g for 20 min. The supernatant was collected and immediately assayed for total soluble protein with the Bradford Assay. Supernatants were stored at −80 °C until required for ELISA and Western Blot analysis.

ELISA was used to quantify the amount of plant-expressed CMG2-Fc. Microplate wells (Costar 3590) were coated with Staphylococcus aureus Protein A (Southern Biotech 7300-01S, Birmingham, AL) diluted at 50 μg/mL in PBS Buffer (pH 7.4) and incubated for 1 h at 37 °C. Blocking was achieved by a 15-min incubation with 5% nonfat dry milk (Safeway) diluted in PBS. All samples and controls were diluted in PBS and applied directly to the coated wells. For each microplate, a standard curve was generated with 5000; 1666.7; 555.6; 185.2; 61.7; 20.6; 6.9; and 2.3 ng/mL of purified CMG2-Fc (supplied by Planet Biotechnology Inc) diluted in PBS. Microplates were incubated with 50 μL of diluted samples and standards at 37 °C for 1 h. Goat anti-human IgG secondary antibody conjugated with horseradish peroxidase (Southern Biotech 2040-05, Birmingham, AL) diluted 1:3000 in PBS was added, and the microplates were incubated at 37 °C for 1 h. Detection was performed with 3,3′,5,5′-tetramethylbenzidine liquid substrate for ELISA. The reaction was stopped with ¹N HCl and absorbance was measured at 450 nm. Each sample was assayed in triplicate, and CMG2-Fc concentrations were interpolated from the linear portion of the standard curve.

The amount of total soluble protein present in the plant extracts was quantified using the Bradford Protein Assay [57]. Samples and controls diluted in PBS were applied directly to the microplate wells (Costar 3598). For each microplate, a standard curve was generated with 0.5, 0.4, 0.3, 0.2, 0.1, 0.05, and 0 mg/mL of bovine serum albumin diluted in PBS. After adding Bradford Protein Reagent (BioRad, Hercules, CA), the microplates were incubated at room temperature for 10 min, and absorbance was measured at 595 nm. Each sample was assayed in triplicate, and total soluble protein concentrations were interpolated from the standard curve.

Protein samples were denatured by heating for five min at 95 °C in sodium dodecyl sulfate (SDS) sample buffer and 1M dithiothreitol (DTT), for reduction of proteins at a final concentration of 90 mM DTT. Samples were fractionated with SDS-PAGE at 200 V for 30 min, and transferred to a 0.45 μm nitrocellulose membrane (GE Healthcare Biosciences, Pittsburgh, PA) at 120 V for 80 min. Blots were blocked with 5% nonfat dry milk diluted in PBS. Immunoblotting was performed using Staphylococcus aureus Protein A, followed by goat anti-human IgG secondary antibody conjugated with alkaline phosphatase (Southern Biotech 2040-04, Birmingham, AL). The blot was developed using the substrate BCIP/NBT (BioRad, Hercules, CA).

Statistical analysis was performed using a two-sample t-test assuming unequal variances in Microsoft Excel.