High-efficiency protein expression in plants from agroinfection-compatible Tobacco mosaic virus expression vectors

Abstract

Background: Plants are increasingly being examined as alternative recombinant protein expression systems. Recombinant protein expression levels in plants from Tobacco mosaic virus (TMV)-based vectors are much higher than those possible from plant promoters. However the common TMV expression vectors are costly, and at times technically challenging, to work with. Therefore it was a goal to develop TMV expression vectors that express high levels of recombinant protein and are easier, more reliable, and more cost-effective to use.

Results: We have constructed a Cauliflower mosaic virus (CaMV) 35S promoter-driven TMV expression vector that can be delivered as a T-DNA to plant cells by Agrobacterium tumefaciens. Co-introduction (by agroinfiltration) of this T-DNA along with a 35S promoter driven gene for the RNA silencing suppressor P19, from Tomato bushy stunt virus (TBSV) resulted in essentially complete infection of the infiltrated plant tissue with the TMV vector by 4 days post infiltration (DPI). The TMV vector produced between 600 and 1200 micrograms of recombinant protein per gram of infiltrated tissue by 6 DPI. Similar levels of recombinant protein were detected in systemically infected plant tissue 10-14 DPI. These expression levels were 10 to 25 times higher than the most efficient 35S promoter driven transient expression systems described to date.

Conclusion: These modifications to the TMV-based expression vector system have made TMV vectors an easier, more reliable and more cost-effective way to produce recombinant proteins in plants. These improvements should facilitate the production of recombinant proteins in plants for both research and product development purposes. The vector should be especially useful in high-throughput experiments.

Figure 1

Maps of plasmid T-DNAs used in this study. Cauliflower mosaic virus (CaMV) 35S promoter driven versions of the gfp gene, Tomato bushy stunt virus p19 gene or Tobacco mosaic virus (TMV) vector cDNAs were constructed. All plasmids were based on the binary vector pCB301 backbone. T-DNA border sequences not shown in maps. Open boxes represent open reading frames; black block arrows; 35S promoter (35S Pr); black boxes, CaMV 3' terminator sequence; light grey boxes, Tobacco etch virus 5' non-translated leader sequence (L); dark grey box, ribozyme (Rz); bent arrows, locations of TMV subgenomic promoters. The TMV vectors in pJL36 and pJL43 contain the full complement of TMV genes as well as an additional subgenomic promoter. TMV sequences 5' of the multiple cloning site (MCS) are from the U1 strain of TMV. Virus sequences 3' of the MCS are from the U5 strain of TMV. TMV transcripts are processed by a ribozyme to generate authentic TMV 3' ends. The sequence of the MCS in pJL 36 and 43 are presented. Restriction endonuclease recognition sequences are underlined. Because SapI is non-palindromic (GCTCTTC N1/4) both strands of the MCS of pJL 43 are presented for clarity.

Figure 2

Diagram of a Sticky RICE cloning reaction into pJL 43. Sticky RICE cloning used a mixture of DNA polymerase and ligase (and, optionally, polynucleotide kinase) with specially designed vector and insert (PCR product) DNAs to directionally ligate DNAs. Single stranded 3 nt, 5' overhangs were generated on pJL 43 by digestion the restriction endonuclease SapI (underlined). Vector was treated with phosphatase after digestion to remove phosphates from 5' ends of DNA. I. Purified PCR product [amplified with 5' phosphorylated primers that began with 5'GGCCWW and 5'GCWW (W = A or T)] was added to SapI cut pJL 43. II. A mixture of T4 DNA polymerase, the nucleotides dATP/dTTP and T4 DNA ligase were added to combined vector and PCR product. During this step the 5' overhangs of SapI cut pJL 43 were altered by the T4 DNA polymerase. A single G residue was removed from the 3' end of the left end of the SapI cut vector, to generate a 5' overhang of GGCC. A single A residue was added to the 3' end of the right end of the SapI cut vector, to generate a 5' overhang of GC. Similarly, the 3' to 5' exonuclease activity of T4 DNA polymerase in the presence of dATP and dTTP removed G or C residues from the 3' ends of the PCR product. Complementary 5' overhangs in vector and PCR product (insert) guided annealing of DNAs. III. Annealed DNAs were joined by T4 DNA ligase. Sequences of the PCR product are in bold type. Vector sequences in final joined product are in all caps. The recognition sequences for the restriction endonucleases StuI (AGGCCT) and Hind III (AAGCTT) were generated at the vector-insert junctions.

Figure 3

Effect of p19 on the agroinfection efficiency of a Tobacco mosaic virus-based vector. N. benthamiana leaves were infiltrated with mixtures of A. tumefaciens (A.t.) cells (OD₆₀₀ 1.0, or dilutions thereof) containing the binary plasmid pJL43:GFP and, in some treatments, pJL3:P19. pJL43:GFP has as its T-DNA a 35S promoter driven TMV vector with a green fluorescent protein (gfp) gene insert. When TMV RNA is transcribed from the T-DNA, the TMV vector initiates self-replication and expresses GFP. The GFP-expressing TMV vector can move cell-to-cell and systemically in plants. Photographs taken under UV light, where GFP appears green and non-GFP expressing tissue red. A. Leaves infiltrated with (i) 1:1 mix of A.t./pJL43:GFP + A.t./pJL4 (empty vector); (ii) 1:1 mix of A.t./pJL43:GFP + A.t./pJL3:p19; (iii) 1:1 mix of 1:50 dil of A.t./pJL43:GFP + undiluted A.t./pJL4; (iv)1:1 mix of 1:50 dil of A.t./pJL43:GFP + undiluted A.t./pJL3:p19. Photographed 3 days post infiltration (DPI). B. Fluorescent micrographs of infiltrated leaves, 3 DPI. Labelling same as in A. C. Systemic movement of TMV:GFP in N. benthamiana plant where a lower leaf was infiltrated with A.t./pJL43:GFP cells. Photo taken 11 DPI.

Figure 4

Temporal analysis of GFP expression from TMV vectors. N. benthamiana leaves were infiltrated with an A. tumefaciens/pJL43:GFP cell suspension. Total soluble protein extracts were prepared from infiltrated leaf tissue from 3–7 days post infiltration (DPI) or from plant tissue systemically infected with the TMV:GFP vector (12 DPI). Extracts were separated on a 4–20% SDS PAGE gel and stained with coomassie blue. Lanes: M, MW marker; 2, Healthy leaf extract; 3–7, extracts from A. tumefaciens/pJL43:GFP infiltrated leaves, 3–7 DPI, respectively; 8, extract from systemically infected tissue, 12 DPI. Marker band sizes (in kDa) are listed. Locations of GFP and TMV CP are identified by open and solid triangles, respectively.

Figure 5

Quantitation of transient expression of GFP in plants from 35S promoter or from a TMV vector. Top. Individual N. benthamiana leaves infiltrated with (A) A. tumefaciens (A.t.)/pCB:GFP + A.t./pJL3:P19 cell suspensions or (B) A.t.pJL43:GFP + A.t./pJL3:P19 cell suspensions. (C) Non-infiltrated control leaf. Leaves photographed (4 days post infiltration, DPI) under UV light to visualize GFP. Bottom. Micrograms of GFP produced per gram of infiltrated tissue as estimated by GFP fluorescence assay. Labelling same as in top of figure. Plant extracts prepared 4–5 DPI were analyzed with a plate-based GFP fluorescence assay. Three plants of each treatment were analyzed. Samples were analyzed in triplicate, and values averaged. Purified His-6 tagged GFP was used to generate a standard curve.