Improving agroinfiltration-based transient gene expression in Nicotiana benthamiana

Abstract

Background: Agroinfiltration is a simple and effective method of delivering transgenes into plant cells for the rapid production of recombinant proteins and has become the preferred transient expression platform to manufacture biologics in plants. Despite its popularity, few studies have sought to improve the efficiency of agroinfiltration to further increase protein yields. This study aimed to increase agroinfiltration-based transient gene expression in Nicotiana benthamiana by improving all levels of transgenesis.

Results: Using the benchmark pEAQ-HT deconstructed virus vector system and the GUS reporter enzyme, physical, chemical, and molecular features were independently assessed for their ability to enhance Agrobacterium-mediated transformation and improve protein production capacities. Optimal Agrobacterium strain, cell culture density and co-cultivation time for maximal transient GUS (β-glucuronidase) expression were established. The effects of chemical additives in the liquid infiltration media were investigated and acetosyringone (500 μM), the antioxidant lipoic acid (5 μM), and a surfactant Pluronic F-68 (0.002%) were all shown to significantly increase transgene expression. Gene products known to suppress post-transcriptional gene silencing, activate cell cycle progression and confer stress tolerance were also assessed by co-expression. A simple 37 °C heat shock to plants, 1-2 days post infiltration, was shown to dramatically increase GUS reporter levels. By combining the most effective features, a dual vector delivery system was developed that provided approximately 3.5-fold higher levels of absolute GUS protein compared to the pEAQ-HT platform.

Conclusions: In this paper, different strategies were assessed and optimised with the aim of increasing plant-made protein capacities in Nicotiana benthamiana using agroinfiltration. Chemical additives, heat shock and the co-expression of genes known to suppress stress and gene silencing or stimulate cell cycle progression were all proven to increase agroinfiltration-based transient gene expression. By combining the most effective of these elements a novel expression platform was developed capable of producing plant-made protein at a significantly higher level than a benchmark hyper-expression system.

Keywords: Agroinfiltration; Hyperexpression; Nicotiana benthamiana; Transient; pEAQ-HT.

Fig. 1

Schematic representation of pEAQ-GSN, pSPECIAL and pNEEDS vectors. pEAQ-GSN was assembled by introducing the uidA reporter gene and small synthetic intron (syntron) between the CPMV 5′ and 3′ UTR translation enhancers in pEAQ-HT [31]. pSPECIAL is based on pEAQ-GSN with a downstream expression cassette encoding the truncated CMV 2b (1-94) silencing suppressor protein. pNEEDS is a pBIN-Plus binary vector comprising two expression cassettes encoding the AtBAG4 stress tolerance protein under the transcriptional control of the nos promoter and the TYDV Rep/RepA cell cycle control gene products under the transcriptional control of the truncated CaMV (− 90) promoter. 35SP = Cauliflower mosaic virus 35S promoter; CPMV 5′ UTR = Cowpea mosaic virus RNA‐2 5′UTR; uidA = gene encoding GUS; syntron = synthetic intron; CPMV 3′ UTR = Cowpea mosaic virus RNA‐2 3′UTR; nosT = nopaline synthase terminator from Agrobacterium; TBSV p19 = Tomato bushy stunt virus p19 silencing suppressor gene; 35ST = Cauliflower mosaic virus 35S terminator; CMV 2b (1-94) = Cucumber mosaic virus truncated 2b silencing suppressor gene (amino acids 1-94); nosP = nopaline synthase promoter from Agrobacterium; AtBAG4= Arabidopsis BAG4 gene; ∆35SP = truncated Cauliflower mosaic virus 35S (− 90) promoter; TYDV Rep/RepA = Tobacco yellow dwarf virus Rep/RepA gene encoding both Rep and RepA

Fig. 2

Effects of agrobacteria strain and cell culture density on transient GUS expression via agroinfiltration. a Agrobacteria strains AGL1, C58C1 and LBA4404 harbouring pEAQ‐GSN were infiltrated into N. benthamiana leaves. Leaves were sampled at 0, 2, 4, 6 and 8 days post infiltration (dpi). b Agrobacterium strain AGL1 harbouring pEAQ‐GSN was infiltrated into N. benthamiana leaves at increasing concentrations OD₆₀₀ = 0.001, 0.01, 0.1, 0.5, 1.0, and 1.5. TSP was extracted for GUS fluorometric enzyme assays. Columns represent relative levels of mean GUS enzyme activities and bars represent ± SE. (†) indicates the reference treatment and (*) indicates data significantly different to the reference (p < 0.05)

Fig. 3

Effects of chemical additives on transient GUS expression. Agrobacteria strain AGL1 harbouring pEAQ‐GSN were infiltrated in MMA media containing different concentrations of chemical additives, a Lipoic acid, b ascorbic acid, c acetosyringone, d Pluronic F‐68, and e PVP, into N. benthamiana leaves. Leaves were sampled 4 dpi and TSP extracted for GUS fluorometric enzyme assays. Columns represent relative levels of mean GUS enzyme activities and bars represent ± SE. (†) indicates the reference treatment and (*) indicates data significantly different to the reference (p < 0.05)

Fig. 4

Effects of heat shock treatment and co‐expressing a stress tolerance gene on transient GUS expression. a Agrobacteria strain AGL1 harbouring pEAQ‐GSN were infiltrated into N. benthamiana leaves and the whole plants heat shocked (37 °C for 30 min), at either 0, 1, 2, or 3 dpi. b Co‐transformation with a vector capable of expressing the Arabidopsis BAG4 gene product. Leaves were sampled 4 dpi and TSP extracted for GUS fluorometric enzyme assays. Columns represent relative levels of mean GUS enzyme activities and bars represent ± SE. (†) indicates the reference treatment and (*) indicates data significantly different to the reference (p < 0.05)

Fig. 5

Effects of co‐expressing suppressors of gene silencing on transient GUS expression. Agrobacteria strain AGL1 harbouring the vector p35S‐GSN were co‐infiltrated into N. benthamiana leaves with various suppressors of gene silencing, including CMV 2b, CMV 2b (1-94) truncation, PRSV HC-Pro, TBSV p19, TLCV TrAP and both the CMV 2b (1-94) truncation and TBSV p19 together. Leaves were sampled 4 dpi and TSP extracted for GUS fluorometric enzyme assays. Columns represent relative levels of mean GUS enzyme activities and bars represent ± SE. (†) indicates the reference treatment and (*) indicates data significantly different to the reference (p < 0.05)

Fig. 6

Effects of co‐expressing virus‐derived genes encoding cell cycle regulatory proteins. Agrobacteria strain AGL1 harbouring pEAQ‐GSN were co‐infiltrated into N. benthamiana leaves with a TYDV Rep/RepA or RepA genes and the TYDV RepA gene containing a LxCxK mutation in the RB-binding motif, and b cell cycle regulatory genes from related circular ssDNA plant viruses, under the transcriptional control of the truncated CaMV 35S (− 90) promoter (Δ35S). Leaves were sampled 4 dpi and TSP extracted for GUS fluorometric enzyme assays. Columns represent relative levels of mean GUS enzyme activities and bars represent ± SE. (†) indicates the reference treatment and (*) indicates data significantly different to the reference (p < 0.05)

Fig. 7

Comparison of GUS expression afforded by pEAQ-GSN versus pSPECIAL + pNEEDS vectors. Agrobacteria strain AGL1 harbouring pEAQ‐GSN or pSPECIAL and pNEEDS were infiltrated into N. benthamiana. Leaves were sampled 4 dpi and GUS levels measured by a fluorometric MUG enzyme assays, b ELISA and c SDS‐PAGE. M = Novex^® Sharp Pre‐stained Protein Standard; 1 = pBIN‐Plus in MMA; 2 = pEAQ‐GSN in MMA; 3 = pSPECIAL + pNEEDS in MMA‐LP with heat shock (2 dpi); 4 = Purified GUS standard (0.3 μg) (GUS Type VII‐A; Sigma‐Aldrich G7646). Columns in a represent relative levels of mean GUS enzyme activitiy (μmol 4-MU/mg protein/min) and bars represent ± SE. (*) indicates a significant difference (p < 0.05). Columns in b represent mean GUS concentrations and bars represent ± SE