Bipartite and tripartite Cucumber mosaic virus-based vectors for producing the Acidothermus cellulolyticus endo-1,4-β-glucanase and other proteins in non-transgenic plants

Abstract

Background: Using plant viruses to produce desirable proteins in plants allows for using non-transgenic plant hosts and if necessary, the ability to make rapid changes in the virus construct for increased or modified protein product yields. The objective of this work was the development of advanced CMV-based protein production systems to produce Acidothermus cellulolyticus endo-1, 4-β-glucanase (E1) in non-transgenic plants.

Results: We used two new Cucumber mosaic virus (CMV)-based vector systems for producing the green fluorescent protein (GFP) and more importantly, the Acidothermus cellulolyticus endo-1, 4-β-glucanase (E1) in non-transgenic Nicotiana benthamiana plants. These are the inducible CMVin (CMV-based inducible) and the autonomously replicating CMVar (CMV-based advanced replicating) systems. We modified a binary plasmid containing the complete CMV RNA 3 cDNA to facilitate insertion of desired sequences, and to give modifications of the subgenomic mRNA 4 leader sequence yielding several variants. Quantitative RT-PCR and immunoblot analysis showed good levels of CMV RNA and coat protein accumulation for some variants of both CMVin and CMVar. When genes for E1 or GFP were inserted in place of the CMV coat protein, both were produced in plants as shown by fluorescence (GFP) and immunoblot analysis. Enzymatic activity assays showed that active E1 was produced in plants with yields up to ~ 11 μg/g fresh weight (FW) for specific variant constructs. We also compared in vitro CMV genomic RNA reassortants, and CMV RNA 3 mutants which lacked the C' terminal 33 amino acids of the 3A movement protein in attempts to further increase E1 yield. Taken together specific variant constructs yielded up to ~21 μg/g FW of E1 in non-transgenic plants.

Conclusions: Intact, active E1 was rapidly produced in non-transgenic plants by using agroinfiltration with the CMV-based systems. This reduces the time and cost compared to that required to generate transgenic plants and still gives the comparable yields of active E1. Our modifications described here, including manipulating cloning sites for foreign gene introduction, enhance the ease of use. Also, N. benthamiana, which is particularly suitable for agroinfiltration, is a very good plant for transient protein production.

Figure 1

(A) Diagram of the E1 constructs and (B) Modified RNA 4 leader sequences containing restriction enzyme sites compared to unmodified wild type (wt) leader sequences. (A) Gene structure of the E1 used in this report. The E1 gene contains the rice amylase signal peptide in the upstream of its ORF and a 6 histidine-tag in its C-terminus. RAmySP; rice amylase signal sequence. This Figure is not to scale. (B) Restriction endonuclease sequences are shown as italicized and underlined, the CP start codon (AUG) is shown in bold. Modified leader sequences are located between the wildtype leader sequence and AUG start codon in subgenomic RNA4. Wt, wildtype leader sequence; 2+, modified #2 leader sequences addition to the wildtype leader sequence; 6+, modified #6 leader sequences addition to the wildtype leader sequence; 8+, modified #8 leader sequences addition to the wildtype leader sequence.

Figure 2

Relative RNA and coat protein accumulation levels for CMVinII and CMVarII variants having wt or modified RNA 4 leader sequences. After 6 days post-infiltration, infiltrated leaves were harvested and total RNAs and proteins were extracted as described in Methods. For CMVinII variants (see Table 2), estradiol was applied 18 hrs after infiltration using a cotton swab. For CMVinII variants, two leaves each from 3 plants for each treatment (n = 6) were used for 1^st, 2^nd biological trials, and three leaves each from 2 plants for 3^rd, 4^th trials (n = 6) yielding four different biological trials (total n = 24). For CMVarII variants (see Table 2), two leaves each from 3 plants for each treatment (n = 6) were used for analysis. These experiments were replicated three times (total n = 18). For immunoblot analysis, each lane was loaded with 15 μg of total protein. Real-time PCR reactions were performed using an ABI 7500 with gene specific primer sets. Relative Ct values were calculated to the respective 18S rRNA relative concentration. Relative folds were calculated to the repective wildtype CMVinII and CMVarII value, respectively. Group superscript letters next to the numbers represent different statistical groups, means with the same letter are not significantly different by the Bonferroni (Dunn) t test using the SAS 9.1 program. Panel (A) CMVinII variants; H , healthy uninfiltrated leaf; pR1R2, pCMVinII RNA 1 and RNA 2 only; CMVinII wt , wt leader; CMVinII 2 , #2 leader; CMVinII 6, #6 leader; CMVinII 8 , #8 leader; CMViva, pCMV (see Methods). Panel (B) CMVarII variants; H, healthy, uninfiltrated leaf; R1 + R2, pCMVarII RNA 1 and RNA 2 only; CMVarII wt , wild type leader; CMVarII 2 , #2 leader; CMVarII 6 , #6 leader; CMVarII 8, #8 leader; pQ123, pCassQ123 (see Methods).

Figure 3

GFP fluorescence for CMVinII and CMVarII wt and leader sequence variants in N. benthamiana plants.CMVinII G and CMVarII G (see Table 2) columns are indicated for both 6 and 10 days after infiltration. Rows show relative GFP fluorescence for leader sequence variants. A. tumefaciens cells containing respective plasmids were mixed just before leaf infiltration. CMVinII G plants then had estradiol induction treatments 18 hrs after infiltration. Photos were taken under long wavelength UV light.

Figure 4

Immunoblot analysis of E1 produced in infiltrated N. benthamiana leaves. N. benthamiana leaves were co-infiltrated with A. tumefaciens containing plasmids corresponding to the CMVinII E and CMVarII E variants (see Table 2) shown. CMVinII E leaves were treated with estradiol 18 hrs after infiltration. Leaves were harvested 6 days post-infiltration. Soluble proteins were extracted as described in Methods and concentrations were determined by Bradford assay. Numbers above each lane correspond to the specific RNA 3 variant used. Each sample represents 15 μg total protein extract per lane. Prestained size marker (Benchmark, Invitrogen, Carlsbad, CA, U.S.A.) shows 80 kDa band.

Figure 5

Comparison of GFP fluorescence from CMVarI G and CMVarII G variants. N. benthamiana leaves were co-infiltrated with A. tumefaciens containing plasmids corresponding to the CMVarI and II G variants (see Table 2) as shown at Panel A. Numbers for leaves in panel at right indicate specific RNA 3 variant. Photographs were taken 6 days post-infiltration under UV light shown as Panel B.

Figure 6

Comparison of GFP fluorescence CMVarI 33 G and CMVarII 33 G MP deletion mutants. N. benthamiana leaves were co-infiltrated with A. tumefaciens containing plasmids corresponding to the CMVarI and II 33 G variants (see Table 2) as shown at left Panel A. Numbers for leaves in Panel B at right indicate specific RNA 3 variant containing MP deletion. Photographs were taken 6 days post-infiltration under UV light shown as Panel B.