Tight regulation of plant immune responses by combining promoter and suicide exon elements

Effector-triggered immunity (ETI) is activated when plant disease resistance (R) proteins recognize the presence of pathogen effector proteins delivered into host cells. The ETI response generally encompasses a defensive ‘hypersensitive response’ (HR) that involves programmed cell death at the site of pathogen recognition. While many R protein and effector protein pairs are known to trigger HR, other components of the ETI signaling pathway remain elusive. Effector genes regulated by inducible promoters cause background HR due to leaky protein expression, preventing the generation of relevant transgenic plant lines. By employing the HyP5SM suicide exon, we have developed a strategy to tightly regulate effector proteins such that HR is chemically inducible and non-leaky. This alternative splicing-based gene regulation system was shown to successfully control Bs2/AvrBs2-dependent and RPP1/ATR1Δ51-dependent HR in Nicotiana benthamiana and Nicotiana tabacum, respectively. It was also used to generate viable and healthy transgenic Arabidopsis thaliana plants that inducibly initiate HR. Beyond enabling studies on the ETI pathway, our regulatory strategy is generally applicable to reduce or eliminate undesired background expression of transgenes.

(1) The numbers above the constructs refer to amino acid codon positions. The dexamethasone inducible pTA7001 vector also includes 35S::GVG, a transcription factor that binds the 6xUAS promoter. (2) The p1776 vector has a strong constitutive chimeric octopine and manopine synthase promoter later renamed the "SuperPromoter" (3,4).   FIGURE S6. Resistance gene and effector gene controls for the hypersensitive response. N. benthamiana leaves were spot infiltrated or spot co-infiltrated with the indicated constructs using a needle-less syringe, with three spots on each leaf half. Total Agrobacterium (a-c) OD600 = 0.75 or (d-e) OD600 = 0.9. (a) Bs2-HA alone sometimes produces visible minor chlorosis (pale green color), but not the hypersensitive response. (b) avrBs2 constructs do not initiate a hypersensitive response in the absence of the resistance gene Bs2. (c) The Dex-induced pTA7001 OsL5 or OsL5-6xHis vectors alone do not result in a hypersensitive response. The red dot seen on all leaves near the stem was used to mark infiltrated leaves during the experiment. (d, e) Control infiltrations for ATR1 and RPP1. DsRed2 is added to keep OD600 consistent. RPP1/ATR1Δ51-dependent HR is much stronger in (e) N. tabacum than (d) N. benthamiana. μL or (b, c) 15 μL of crude proteins were run on 4-12% NuPAGE Bis-Tris gels with 1X MOPS buffer, then transferred to nitrocellulose. (a, b) Anti-HA blot shows AvrBs2-HA protein development from induced (a) pTA7001 avrBs2-HA alone or (b) pTA7001 avrBs2-HyP5SM-HA co-infiltrated either with pTA7001 empty vector or pTA7001 OsL5-6xHis. (c) Anti-OsL5-6xHis blot shows development of OsL5-6xHis protein (35.7 kDa) at slightly higher molecular weight than a non-specific band which may potentially be endogenous NbL5 protein (estimated 34.6 kDa). (d) Anti-HA western blot (top) and anti-OsL5-6xHis western blot (middle) comparing the extent of protein expression from oneinput and two-input regulation systems. N. benthamiana leaves were co-infiltrated in 1:1 mixes with the indicated pTA7001 avrBs2 constructs, plus either buffer, pTA7001 empty vector, pTA7001 OsL5, or pTA7001 OsL5-6xHis. The "buffer" control displays the leakiness of pTA7001 without the confounding variable of an additional copy of the GVG transcription factor in pTA7001, while also maintaining equal final OD600 of all avrBs2 constructs. Lanes 3, 6, 11 are empty. Ponceau S stain of RuBisCo large subunit is shown as a loading control below each western blot. M = NEB #P7711 marker. WT = wild-type N. benthamiana.  (a) Infiltrations to test "C59", the multi-gene construct pTKan C59 (which lacks RPP1, but contains Dex-inducible ATR1Δ51-HyP5SM-FLAG and OsL5-6xHis). All sections were co-infiltrated with pEG301 RPP1-3xHA in a 1:1 mix. The mock-induced leaf shows no hypersensitive response from pTKan C59. With Dex, the hypersensitive response is induced. (b) To investigate why the total hypersensitive response seems to be weaker from pTKan C59, N. tabacum was co-infiltrated with 1:1:1 mixes of RPP1, the indicated ATR1 or empty construct, and either pTA7001 OsL5 or additional pTA7001 empty vector (which, although empty after the Dex-inducible promoter, still contains 35S::GVG transcription factor). (c) N. benthamiana was transiently transformed with pTKan C59 and Dex-induced to promote protein expression. Western blots show expression of ATR1Δ51-FLAG (top: anti-FLAG) and OsL5-6xHis protein (bottom: anti-OsL5-6xHis) from four biological replicates. Below each western blot is a Ponceau S stain of RuBisCo. Col-0 WT contains no transgenes, Col-0 ATR1 contains pTKan C59 but no RPP1, WT* is 3860+RPP1 background line for R7, R25, and R21, which contain pTKan C59. The higher band observed for RPP1 is an intron-retaining spliced product. Pre-mRNA, SP-I, and SP-II are observed for ATR-HyP5SM-FLAG. A nested PCR was performed on the PCR product from full-length OsL5, resulting in stronger signal.   Step 1: Choose an insertion site in the sequence of the gene of interest.
The most conserved splice site is the 5' splice site. (6) The HyP5SM cassette must be cloned after an AG dinucleotide in the gene of interest.(5) The following silent mutations may be used if cloning the cassette in frame. Fortunately, dicot plants do not have strong codon preferences for these amino acids. Step 2: Make the three templates for 3-piece-ligation PCR (also called extension PCR).
Make three separate dsDNA templates with a high fidelity polymerase. Template: avrBs2-HA

Primers:
The forward primer (TLG43) anneals to avrBs2 at the insertion site and adds an overhang. The avrBs2 reverse primer (TLG46) adds an XbaI site.
Gel extract each product from an agarose gel. Calculate the molar concentration of the PCR products.
Step 3: Perform 3-piece-ligation PCR with a high fidelity polymerase.
Templates: The three PCR products in a 1:1:1 molar ratio. For avrBs2-HyP5SM-HA (E/V) cloning, the final concentration of each template was approximately 0.15 nM.

Primers:
The avrBs2-HA forward and reverse cloning primers (TLG45, TLG46). The overhangs in the templates also act as primers.
Step 4: The desired full-length product was gel purified, inserted into a cloning vector (Invitrogen pCR2.1 TOPO), and sequenced to confirm that the junctions were correct and that there were no frame shifts. The sequence perfect product was then cloned into a binary vector for plant expression.

II. Cloning the multi-gene plasmids for faster generation of transgenic Arabidopsis thaliana plants
Step 1: Construction of the pTKan expression vector.
Step 3: Construction of the multi-gene plasmids In a MultiSite Gateway reaction (MultiSite Gateway®Pro, Invitrogen), genes or fragments of interest in pDONR221 P1-P4 vector, pDONR221 P4r-P3r vector and pDONR221 P3-P2 vector were connected and incorporated into GW R1-R2 expression vector in a sequential manner. MultiSite Gateway reaction was carried out with the expression vector pTKan-pNOS-DsRed2-tNOS-p35S-GVG-tRbcsE9-p6xUAS-GW-R1R2 and various entry vectors from Step 1 and Step 2. Entry vector combinations for the final expression constructs were listed in the following table. LR Clonase™ II Plus enzyme mix and equal molar of each component construct were used in the MultiSite Gateway reaction according to the manufacturer's manual.