An Agrobacterium-Mediated Transient Expression Method for Functional Assay of Genes Promoting Disease in Monocots

Abstract

Agrobacterium-mediated transient expression (AMTE) has been widely used for high-throughput assays of gene function in diverse plant species. However, its application in monocots is still limited due to low expression efficiency. Here, by using histochemical staining and a quantitative fluorescence assay of β-glucuronidase (GUS) gene expression, we investigated factors affecting the efficiency of AMTE on intact barley plants. We found prominent variation in GUS expression levels across diverse vectors commonly used for stable transformation and that the vector pCBEP produced the highest expression. Additionally, concurrent treatments of plants with one day of high humidity and two days of darkness following agro-infiltration also significantly increased GUS expression efficiency. We thus established an optimized method for efficient AMTE on barley and further demonstrated its efficiency on wheat and rice plants. We showed that this approach could produce enough proteins suitable for split-luciferase assays of protein-protein interactions on barley leaves. Moreover, we incorporated the AMTE protocol into the functional dissection of a complex biological process such as plant disease. Based on our previous research, we used the pCBEP vector to construct a full-length cDNA library of genes upregulated during the early stage of rice blast disease. A subsequent screen of the library by AMTE identified 15 candidate genes (out of ~2000 clones) promoting blast disease on barley plants. Four identified genes encode chloroplast-related proteins: OsNYC3, OsNUDX21, OsMRS2-9, and OsAk2. These genes were induced during rice blast disease; however, constitutive overexpression of these genes conferred enhanced disease susceptibility to Colletotrichum higginsianum in Arabidopsis. These observations highlight the power of the optimized AMTE approach on monocots as an effective tool for facilitating functional assays of genes mediating complex processes such as plant-microbe interactions.

Keywords: Agrobacteria-mediated transient expression; Arabidopsis; barley; chloroplast; functional screening; gain-of-function; protein-protein interaction; rice blast disease; susceptibility.

Figure 1

Comparison of transient expression efficiency between different vectors in barley (var. E9) leaves. (A). Schematic drawings of binary vectors used in this report. Diverse cis-regulatory elements and promoters in T-DNA were selected for GUS (β-glucuronidase) expression. The empty vector pER8:DEST was used as a negative control. Green letters represent different promoter elements regulating GUS transcription: 35S indicates cauliflower mosaic virus (CaMV) 35S promoter; Ubi indicates ubiquitin promoter; XVE, an estradiol-activated chimeric transcription activator, controlled by constitutive promoter G10–90 and targets the LexA operator (LexA); 35S mini indicates −46 to +1 region of the 35S minimal promoter; GVG, a dexamethasone (DEX) activated chimeric transcription activator, target UAS operator; 4×UAS and 6×UAS indicate four or six copies of GAL4 UAS; Gos2 indicates the constitutive promoter of rice GOS2 gene; nos indicates terminator; Hpt and Bar indicate Hygromycin and Basta resistance genes, respectively; RB and LB indicate T-DNA right and left border regions; inverted letters represent the gene orientation (from right to left). (B). GUS staining after transient expression with different vectors in barley (upper) and Nicotiana benthamiana (lower) leaves. AGL1 and individual vectors (0.5 OD₆₀₀) were syringe infiltrated into the first leaves of barley following 1 d of moisturizing treatment, then collected for staining analysis at 4 dpi. N. benthamiana leaves were infiltrated at the same time as a control. For pER8:GUS, pCBEP:GUS+pER8, and pER8:DEST, 100 μM estradiol was infiltrated 2 h after bacterial infiltration; for pINDEX2:GUS and pTA7002:GUS, 20 μM DEX was infiltrated. Scale bars = 1.5 mm. (C). Quantitative fluorescence analysis of GUS activity in barley leaves infiltrated with AGL1 (GUS). Leaf samples were treated and collected as in (B). Data shown are mean ± SD (n = 6). Different letters indicate significant differences at p < 0.01. Three biological replicates were performed.

Figure 2

Optimizing conditions for transient expression in barley (var. E9) leaves. (A). GUS staining in the first (1st) and the second (2nd) leaves following the infiltration of AGL1 (pCBEP:GUS) at 4 dpi. Empty vector (pCBEP:DEST) used as a negative control. Scale bars = 1.5 mm. (B). Quantifying GUS activity under different moisturizing times (0, 1 or 4 d) following the agro-infiltration in the first leaf of barley at 4 dpi. (C). Quantifying GUS activity at different cultivation times (0–8 dpi) with 1 d moisturizing after agro-infiltration. (D). Quantifying GUS activity under 1 or 2 days of dark treatment before collecting samples at 4 dpi. (E). Western blot analysis of FLAG-tagged GUS protein accumulation under dark treatments at 0 dpi (the first line) and 4 dpi (last two lines) with 0 d or 2 d dark treatment. Signals were detected with an anti-FLAG antibody, quantitated, and normalized to the Ponceau loading control. The experiments were syringe-infiltrated with 0.5 OD₆₀₀ of AGL1 (pCBEP:GUS). Data shown are mean ± SD (n = 6). Different letters indicate significant differences at p < 0.01. Three biological replicates were performed.

Figure 3

The flow chart of the AMTE. The details of experimental conditions can be found in Materials and Methods.

Figure 4

Transient expression with pCBEP:GUS in barley, wheat, and rice plants. (A). Expression of GUS in the first leaves of different barley varieties. E9 was used as control. (B). Expression of GUS in the 1st and 2nd leaves of wheat plants (AK58). (C). Expression of GUS in leaf and sheath of rice plants (Nipponbare). The plants were syringe-infiltrated (A,B) or vacuum-infiltrated (C) with 0.5 OD₆₀₀ of AGL1 (pCBEP:GUS) following 1 d moisturizing/2 d darkness treatment; samples were collected at 4 dpi. Expression of empty vector (pCBEP:DEST) used as a negative control. Scale bars = 1.5 mm.

Figure 5

The optimized AMTE was amenable to Split-LUC assay of protein-protein interaction in barley (var. E9). The cLUC-OsbHLH6 and OsMYC2-Nluc (A) and the cLUC-OsJAZ1 and OsMYC2-Nluc (B) were transiently expressed in barley leaves. Samples were assayed at 4 dpi. Co-expressions with cLUC-EV or EV-nLUC were used as the control.

Figure 6

Functional analysis of the chloroplast-related genes OsNYC3, OsNUDX21, OsMRS2–9, and OsAk2 on barley and Arabidopsis. (A). Transient expression of candidate genes enhanced blast disease symptoms on barley leaves. At 2 d after agro-infiltration, barley leaves were spray-inoculated with blast fungus P131 (1 × 10⁵ spores/mL) and incubated in a dark growth chamber at 25 °C for 24 h, followed by a 16-h light/8-h darkness photoperiod. Photographs were taken 3 days after fungal inoculation. (B). Relative transcript levels of candidate rice genes during rice blast disease. Quantitative RT-PCR data were normalized against the rice housekeeping gene ACTIN. Data shown are means ± SD (n = 3). Asterisks indicate significant differences calculated by Student’s t-test (* p < 0.05, ** p < 0.01). Three biological replicates were performed. (C). Overexpression of candidate genes in transgenic Arabidopsis lines promoted disease caused by Colletotrichum higginsianum. Detached leaves of 5-week-old plants were spot-inoculated with Ch-1 (1 × 10⁵ spores/mL) and incubated in a plastic box at 25 °C in darkness. Photos were taken 5 days post-inoculation. The above experiments were repeated three times, and similar results were observed. (D,E). Relative fungal biomass in inoculated barley (D) and Arabidopsis (E) leaves. The relative fungal biomass in samples collected in (A,C) was determined by DNA-based qPCR. Levels of the M. oryzae actin gene were normalized against the barley actin gene, and levels of C. higginsianum β-Tublin were normalized against the Arabidopsis actin gene. A student’s t-test was used to test significance and generate p-values (** p < 0.01).