Bacteria were grown in LB broth or on LB agar plates at 37°C (E. coli) or at 28°C (A. tumefaciens). If necessary, the media were supplemented with kanamycin (50 μg/ml), rifampicin (50 μg/ml), or gentamycin (25 μg/ml).

The following synthetic oligonucleotides were used: BlnI_35S_F (ATCCTAGGGTCAACATGGTGGAGCACGA), Nos-T_SnaBI_R (ATTACGTAGATCTAGTAACATAGATGACACCGCG), XhoI_35S_F (ATCTCGAGGTCAACATGGTGGAGCACGA), Nos-T_Bst1107I_R (TAGTATACGATCTAGTAACATAGATGACACCGCG), P19-BHI-F (ATGGATCCATGGAACGagCTATACAA), P19-SwI-R (ATATTTAAATTTACTCGCTTTCTTTTTCGAAGG).

The sequence of the P24 gene was codon-optimized in silico for expression in N. benthamiana and synthesized in vitro (Evrogen, Russia). Additionally, this sequence contains BamHI, BsrGI, and SwaI restriction sites for cloning procedures (Additional file 1: Supplementary Figure S1).

As a backbone for construction of expression vectors we used plasmid pEAQselectK (GenBank: GQ497231.1), in which the backbone region is shorter than in pBINPLUS (Sainsbury et al., 2009b). To make this vector suitable for subsequent cloning steps, we first removed restriction sites AscI and SmaI from its multiple cloning site. DNA of pEAQselectK was treated with AscI and SmaI, then the ends were made blunt using Klenow fragment and the DNA was circularized by self-ligation. The resulting plasmid was named pEf.

A synthetic sequence of the P24 gene was cloned in plasmid pNRGFP (Mardanova et al., 2007) instead of the gfp gene using BamHI and BsrGI restriction sites. pNRGFP is a binary vector containing 35S promoter – gfp – NosT terminator expression cassette within the T-DNA region. Thereafter, the 35S – P24 – NosT fragment was amplified by PCR using primers BlnI_35S_F and Nos-T_SnaBI_R, and cloned in the pEf vector at SnaBI and BlnI restriction sites. The resultant plasmid was named pEf-p24.

A similar expression vector, pEf-p19, containing the gene encoding the silencing suppressor P19 from TBSV was also constructed. The P19 coding sequence was amplified from pEAQspecialK-GFP-HT (Sainsbury et al., 2009b) using primers P19-BHI-F and P19-SwI-R, and cloned into pEf-p24 instead of P24 using the BamHI and SwaI restriction sites.

Expression vector pA7248AMV-GFP (Mardanova et al., 2009) contains the 5′- UTR of the PVX genome, the gene for viral RNA polymerase (RDRP), the subgenomic promoter of the gene for PVX 25K transport protein (Sgp1), the reporter gfp gene downstream of the AMV leader sequence, and the 3′-UTR of the PVX genome. This sequence is inserted between the 35S promoter and NosT terminator within the T-DNA region of pBIN19 binary vector. Unique restriction sites AscI and SmaI located at 5′ and 3′ ends of gfp can be used for cloning of a target gene to replace gfp (Figure 1). The sequence comprising the whole expression cassette between the 35S promoter and NosT was amplified by PCR using pA7248AMV-GFP DNA as a template and primers XhoI_35S_F and Nos-T_Bst1107I_R. Cloning of this fragment at XhoI and Bst1107I restriction sites in pEf resulted in expression vector pEf-GFP. The same sequence was cloned into pEf-p24 at XhoI and Bst1107I restriction sites to make the final pEff-GFP vector. The nucleotide sequence of pEff-GFP vector has been deposited in the GenBank database under accession no KY439904. Structures of expression vectors are shown in Figure 1.

Recombinant vector pEff-M2eHBc was constructed for expression of the hybrid protein M2eHBc consisting of an extracellular domain of M2 protein (M2e) of influenza virus strain A/Duck/Potsdam/1402-6/1986 fused to the N-terminus of hepatitis B core antigen (HBc). The plant-optimized synthetic M2eHBc gene was excised from vector pA7248amvM2epHBc (Ravin et al., 2012) and cloned into the pEff vector using the AscI and SmaI restriction sites.

The recombinant binary vectors were transferred from E. coli into A. tumefaciens strain GV3101 using electroporation. The resulting agrobacterial strains were grown overnight with shaking at 28°C. The cells (1.5 ml) were collected by centrifugation for 5 min at 4000 g, and the precipitate was resuspended in 1.5 ml of buffer containing 10 mM MES (pH 5.5) and 10 mM MgSO₄. Leaves of 4- to 6-week old N. benthamiana plants were injected with suspension of agrobacteria (adjusted by buffer to OD₆₀₀ of about 0.2) using a syringe without a needle. In case of co-infiltration, OD₆₀₀ of 0.2 for each agrobacterial strain were used.

Three agroinfiltrated leaves on a single N. benthamiana plant were used for measurements of green fluorescent protein (GFP) expression. The leaves were ground in an extraction buffer [0.4 M sucrose, 50 mM Tris pH 8.0, 10 mM KCl, 5 mM MgCI₂, 10% (v/v) glycerol, 10 mM β-mercaptoethanol] to obtain a homogeneous suspension. The suspension was centrifuged at 14 000 g for 15 min, and the supernatant containing soluble proteins was collected. Total protein content was determined by the Bradford assay kit (Biorad) with bovine serum albumin as standard. Expression of GFP in leaves was visualized under UV illumination and assayed by SDS-PAGE in a 15% (w/v) gel. For comparative quantitative measurements, GFP was assayed fluorometrically. Protein extracts were dissolved in 2.5 ml of phosphate buffered saline (50 mM sodium phosphate, pH 7.0, 200 mM NaCl), and GFP amount was determined with a Fluorat-02 fluorometer (Lumex, Russia), excitation at 390 nm, emission at 510 nm. The background level was determined and subtracted. For each of three analyzed leaves we considered the most efficient expression system vector as a reference (1 relative unit of GFP) and estimated the relative levels of GFP in other samples. The results in diagrams are expressed as the mean value ± standard deviation.

Protein extracts isolated from plant tissues were mixed with SDS-PAGE sample buffer containing β-mercaptoethanol. The samples were boiled for 10 min and subjected to SDS-PAGE on 15% (w/v) gels. After electrophoresis, gels were either stained with Coomassie brilliant blue or the proteins were transferred from the gel onto a Hybond-P membrane (GE Healthcare, USA) by electroblotting. To prevent non-specific binding of antibodies with the membrane, it was treated with 5% (w/v) solution of dry milk in TBS-T (150 mM NaCI, 20mM Tris, 0.1% (v/v) Tween 20, pH 8.0) buffer. The membrane was probed with mouse monoclonal antibodies specific for the M2e peptide and then incubated with rabbit anti-mouse secondary antibodies conjugated with peroxidase. Specific immunoreactive proteins were detected using a Western Blot ECL Plus kit (GE Healthcare, USA). The intensity of the bands in stained gels and blots was determined using Nonlinear.Dynamics.TotalLab.TL120.v2009-NULL software.

Virus-like particles, formed by M2eHBc protein in vivo, were isolated form the soluble protein preparations using the method described in detail by Peyret (2015). Briefly, infiltrated tissue 7 days post-infiltration was blended in a Waring blender with three volumes of 0.1 M sodium phosphate (pH 7.2), filtered over miracloth, and clarified by centrifugation. The supernatant was filtered using a 0.45 μm syringe filter, then ultracentrifuged through a double sucrose cushion. The VLPs were recovered from the bottom of the sucrose cushion, dialysed against PBS, then ultracentrifuged over a Nycodenz density gradient. The particles thus purified were then observed by transmission electron microscopy.

Particle preparations were placed on carbon-coated copper grids and stained with 2% (w/v) uranyl acetate. The grids were examined using an FEI Tecnai 20 transmission electron microscope.

To create an efficient plant expression vector we combined the advantages of different systems, namely, a binary vector reduced in backbone size, a self-replicating PVX-based vector with translational enhancer, and various suppressors of PTGS. Vector pEAQselectK (Sainsbury et al., 2009b) was used as a backbone. This plasmid is only 6974-bp long, it contains nptIII and trfA genes outside of T-DNA region, and nptII within T-DNA along with the multiple cloning site.

The main expression cassette originated from the previously constructed PVX-based vector pA7248AMV-GFP (Mardanova et al., 2009). This vector contains downstream of the 35S promoter an incomplete copy of the PVX genome including the 5′ UTR, RDRP gene, the first promoter of subgenomic RNA Sgp1 and the 3′ UTR. Upon agroinfiltration of plant tissue and transfer of T-DNA into plant cells, the mRNA transcription from the 35S promoter results in synthesis of the viral vector RNA, the viral vector is replicated in the infected cells, the subgenomic RNA encoding the desired gene is synthesized at high level, and the target protein is expressed. Insertion of translational enhancer from 5′-UTR of RNA 4 of AMV immediately upstream of the start codon of the target gene results in additional elevation of protein production (Mardanova et al., 2009). However, this vector system has two drawbacks. The first is its length since it is based on the pBIN19 plasmid that is almost 12 kb long itself. The second is the absence of a cassette for the expression of silencing suppressors in pA7248AMV-GFP. Thus, an efficient expression of the target proteins in plants with this system requires co-infiltration of leaves with another strain of A. tumefaciens to express such suppressors.

Therefore, we first transferred the whole PVX-based expression cassette from pA7248AMV-GFP to pEAQselectK in order to minimize the size of vector. The resulting vector, pEf-GFP (Figure 1), is about 13 kb in contrast to pA7248AMV-GFP which is almost 18 kb. To compare the efficiencies of pEf-GFP and pA7248AMV-GFP, we transformed these vectors into A. tumefaciens strain GV3101, which was used for infiltration of leaves of N. benthamiana plants. Agroinfiltration zones for two vectors were located within one leaf. GFP synthesis was monitored by UV irradiation. The infiltrated plants were grown for 5 days, after which the leaves were collected and the soluble protein extracts were prepared. Synthesized GFP was assayed by SDS-PAGE, and quantified fluorometrically (Figure 2). The results show that the reduction in backbone size increases protein expression levels by about 40% in comparison to initial pA7248AMV-GFP vector. Thus, the pEf-GFP expression vector was used for subsequent vector improvements.

Post-transcriptional gene silencing is known to be one of the most important factors limiting the efficiency of transient systems of recombinant protein expression in plants (Chicas and Macino, 2001). Although it had little negative effect on expression of GFP as shown above, this problem appeared to be more important for the expression of non-model proteins. For example, the use of pA7248AMV vectors for expression of influenza immunogens was only successful when the leaves were co-infiltrated with a silencing suppressor construct (Ravin et al., 2012; Mardanova et al., 2015).

Thus we aimed to insert a silencing suppressor cassette into the pEf-GFP vector. First we evaluated the efficiency of two known silencing suppressors, P24 from GRLaV2 and P19 from TBSV, in our self-replicating vector system. Both suppressor genes were cloned between the 35S promoter and nosT terminator, and these expression cassettes were inserted into the binary vector pEf, derived from pEAQselectK. We then compared GFP expression levels in N. bethamiana leaves infiltrated with (i) only pEf-GFP, (ii) simultaneously with pEf-GFP and pEf-p24, and (iii) simultaneously with pEf-GFP and pEf-p19. Protein samples were isolated 3 days after infiltration. The results show that both silencing suppressors increase GFP expression level about threefold, and that there is no significant difference between them (Figure 3). We chose the P24 for subsequent vector improvements. Thereafter we created the final vector pEff-GFP containing two expression cassettes. The first one contained 35S promoter, 5′-UTR of PVX, RDRP, Sgp1, gfp, 3′-UTR of PVX and Nos-T, and the second cassette contained 35S promoter, p24 and Nos-T. This vector eliminates the need to use co-infiltration, which simplifies manipulations. However, insertion of additional expression cassette in pEf-GFP increases the size of the vector which could have a negative effect on the performance of the expression system. We therefore compared the level of GFP production achieved with pEff-GFP vector alone and with levels obtained by co-infiltration of leaves with pEf-GFP and pEf-p24. GFP production was assayed 3 days after agroinfiltration. We found that GFP production level was similar for two variants (Figure 4). Under optimal conditions, the pEff vector provides a GFP expression level of about 30% of total soluble protein (Figure 4) which corresponds to 1 mg/g of fresh leaf tissue.

Green fluorescent protein is known as a model protein as it is stable and usually well-expressed in plant cells. Thus we examined the efficiency of the pEff system for production of an immunogen of potential clinical interest. We used the hybrid protein M2eHBc consisting of an extracellular domain of influenza virus M2 protein (M2e) fused to HBc at its N terminus. The hybrid protein M2eHBc formed virus-like particles carrying the M2e peptide on the surface and can be used as candidate influenza vaccine (Neirynck et al., 1999). Previously, we expressed M2eHBc protein in plants using the pA7248AMV vector. The protein was expressed up to 1–2% of total soluble protein in N. benthamiana leaves and was found to be able to form virus-like particles in vivo (Ravin et al., 2012). Immunization of mice with such plant-produced particles provided partial protection against the lethal influenza challenge (Ravin et al., 2012).

To produce M2eHBc particles in plants using the newly developed pEff vector, we cloned the M2eHBc sequence into this vector and performed agroinfiltration of N. benthamiana leaves. Protein samples were isolated 5 days after infiltration and analyzed by SDS-PAGE and Western-blotting (Figure 5). Our results show that the pEff expression system provides expression levels of M2eHBc of about 5–10% of total soluble protein, several times higher than obtained previously using the pA7248AMV vector or the non-replicating pEAQ-HT expression system (Thuenemann et al., 2013).

The recombinant M2eHBc particles were purified from agroinfiltrated plants by double sucrose cushion followed by a Nycodenz density gradient as described by Peyret (2015). The M2eHBc was assembled into spherical 30–35 nm particles which were similar to ones formed by the native HBc antigen, as revealed by transmission electron microscopy (Figure 5).