A single plant resistance gene promoter engineered to recognize multiple TAL effectors from disparate pathogens

Abstract

Plant pathogenic bacteria of the genus Xanthomonas inject transcription-activator like (TAL) effector proteins that manipulate the hosts' transcriptome to promote disease. However, in some cases plants take advantage of this mechanism to trigger defense responses. For example, transcription of the pepper Bs3 and rice Xa27 resistance (R) genes are specifically activated by the respective TAL effectors AvrBs3 from Xanthomonas campestris pv. vesicatoria (Xcv), and AvrXa27 from X. oryzae pv. oryzae (Xoo). Recognition of AvrBs3 was shown to be mediated by interaction with the corresponding UPT (UPregulated by TAL effectors) box UPT_AvrBs3 present in the promoter R gene Bs3 from the dicot pepper. In contrast, it was not known how the Xoo TAL effector AvrXa27 transcriptionally activates the matching R gene Xa27 from the monocot rice. Here we identified a 16-bp UPT_AvrXa27 box present in the rice Xa27 promoter that when transferred into the Bs3 promoter confers AvrXa27-dependent inducibility. We demonstrate that polymorphisms between the UPT_AvrXa27 box of the AvrXa27-inducible Xa27 promoter and the corresponding region of the noninducible xa27 promoter account for their distinct inducibility and affinity, with respect to AvrXa27. Moreover, we demonstrate that three functionally distinct UPT boxes targeted by separate TAL effectors retain their function and specificity when combined into one promoter. Given that many economically important xanthomonads deliver multiple TAL effectors, the engineering of R genes capable of recognizing multiple TAL effectors provides a potential approach for engineering broad spectrum and durable disease resistance.

Plant pathogens are a major threat to crop production worldwide and durable disease resistance is a major goal in plant biotechnology (1–3). A key to achieving durable disease resistance is to elucidate the function of effector proteins that various microbial pathogens (bacteria, fungi, oomycetes, and nematodes) secrete into their hosts (4, 5). Although the primary function of microbial effectors is in virulence, some are known to trigger plant resistance (R) gene mediated resistance and have therefore been termed avirulence (Avr) proteins. For example, AvrBs3 from the bacterial phytopathogen Xanthomonas campestris pv. vesicatoria (Xcv) confers both virulence (6) and avirulence (7). AvrBs3-like proteins have been identified in many pathovars of Xanthomonas (8) and Ralstonia solanacearum (9). Owing to their homology to eukaryotic transcription factors, AvrBs3-like proteins are also termed transcription-activator like (TAL) effectors (10–12). A characteristic central domain of TAL effectors is comprised of a variable number of tandemly-arranged, near-perfect copies of a 34/35-amino acid motif and determines virulence and avirulence specificity (8). In addition, TAL effector proteins generally contain C-terminal nuclear localization signals (NLSs) and an acidic transcriptional activation domain (AAD) (8, 11–13). The repeat domain of the TAL effector AvrBs3 interacts with a corresponding UPA (UPregulated by AvrBs3) box in the promoter of the pepper transcription factor UPA20, which induces hypertrophy (i.e., enlargement) of mesophyll cells, as well as to the promoters of other host genes that appear to contribute to susceptibility (14). In addition, AvrBs3 triggers a programmed cell death response, referred to as the hypersensitive response (HR), in pepper plants that contain the cognate R gene Bs3 (15, 16). Certain pepper lines have an allele of Bs3 known as Bs3-E, which confers resistance to strains carrying the AvrBs3 derivative AvrBs3Δrep16 that has a deletion of repeat units 11–14 (15, 17). AvrBs3 and AvrBs3Δrep16 were found to interact specifically with distinct boxes in the Bs3 and Bs3-E promoters, respectively (16). For clarity we herein refer to these binding sites collectively as UPT (UPregulated by TAL effectors), with a subscript designation for the specific TAL effector that targets it. Interaction of AvrBs3 or AvrBs3Δrep16 with their respective UPT sites, UPT_AvrBs3, and UPT_{AvrBs3Δrep16}, induces transcription of the Bs3 or Bs3-E coding sequence (cds), leading to HR. Previously we analyzed multiple in vitro generated UPT_AvrBs3 box mutants and uncovered three AvrBs3Δrep16 inducible box derivatives (16). Notably, the recognition specificity of the mutated boxes always correlated with a loss of the AvrBs3 mediated inducibility and did not result in UPT boxes with dual recognition specificity.

Transcription of the rice R gene Xa27 is specifically induced by AvrXa27, a TAL effector from X. oryzae pv. oryzae (Xoo) (18). The products of rice Xa27 and pepper Bs3 share no sequence homology. However, transcriptional activation of these R genes by their matching TAL effectors suggests that this mechanism of disease resistance is common to both mono- and dicotyledonous plant species.

In the present study, we wanted to clarify if activation of the rice Xa27 promoter by the matching Xoo TAL effector AvrXa27 is mechanistically similar to activation of the pepper Bs3 promoter by the Xcv TAL effector AvrBs3. Furthermore we test if UPT boxes matching distinct TAL effectors retain their specificity and functionality when combined into one complex promoter.

Results

The Pepper Bs3 and Rice Xa27 Resistance Genes Use Identical Mechanisms for Detection of Their Matching TAL Effectors.

Previously we showed that AvrBs3 binds to the pepper Bs3 promoter, resulting in transcriptional activation of the Bs3 cds and subsequent triggering of HR (15, 16). In the current study, we investigated if the combination of the Xcv AvrBs3 protein and the pepper Bs3 promoter can be functionally substituted by the Xoo AvrXa27 protein and the rice Xa27 promoter to facilitate Bs3 gene activation. To do so, we placed the Bs3 cds under transcriptional control of the rice Xa27 or the rice xa27 promoter and tested if these genes were transcriptionally activated by AvrXa27. The constructs (Fig. 1A) were cloned into a plant transformation vector and delivered into Nicotiana benthamiana leaves using Agrobacterium tumefaciens mediated transient transformation, along with avrXa27 or avrBs3 genes driven by the constitutive cauliflower mosaic virus 35S (35S) promoter (Fig. 1B). In previous studies, Xa27 but not xa27 was found to be transcriptionally activated by Xoo strains delivering AvrXa27 (18). Consistent with this observation, we found that delivery of avrXa27 triggered HR in N. benthamiana only in the presence of the Xa27 promoter-driven Bs3 cds, but not in the presence of the xa27 promoter-driven Bs3 cds. Furthermore, delivery of the Xa27 promoter-driven Bs3 cds did not mediate an HR when coexpressed with avrBs3 (Fig. 1B), confirming that recognition was specific for the particular promoter-TAL effector combination, despite >90% identity between the AvrBs3 and AvrXa27 proteins (8). In summary, these findings demonstrate that the combination of the Xcv TAL effector AvrBs3 and the pepper Bs3 promoter can be functionally replaced by the Xoo TAL effector AvrXa27 and the rice Xa27 promoter to activate expression of the Bs3 cds.

Fig. 1.

AvrXa27 activates the rice Xa27 promoter and triggers a Bs3-dependent HR in N. benthamiana. (A) Graphical representation of constructs used for promoter analysis. The Bs3, Xa27, xa27, and the xa27_mut promoters are displayed as yellow, brown, and orange arrows, respectively. A yellow box represents the Bs3 cds. The UPT_AvrBs3, and UPT_AvrXa27 boxes are shown as blue and black boxes, respectively. A black box with a white bar represents the UPT_AvrXa27* box of the xa27 promoter. Black arrows represent transcription start sites (TSS). The scale indicates distances from the ATG start codon. Numbers below the arrows denote the distances between the 3′ end of a UPT box and its TSS. The Bs3 cds is not drawn to scale. (B) The promoter-Bs3 cds and the 35S promoter-driven avr constructs (avrXa27 [Left] or avrBs3 [Right]) were expressed transiently in N. benthamiana leaves via A. tumefaciens. Asterisks (*) mark areas in which the avr construct but no promoter construct was infiltrated. Dashed lines mark the inoculated areas. Four days after infiltration, the leaves were cleared with ethanol to visualize the HR (dark areas).

Functionally-Relevant Nucleotide Polymorphisms Between the Xa27 and xa27 Promoters Are Located Adjacent to the Predicted TATA Box.

The promoters of the pepper Bs3 and Bs3-E genes differ by a single 13-bp insertion-deletion (InDel) polymorphism that determines the recognitional specificity of the R genes (Fig. S1) (15). In contrast, the 1.5 kb putative promoter regions of the rice Xa27 and the xa27 alleles differ by 16 polymorphisms (Figs. S2 and S3). Because the Bs3/Bs3-E InDel is located in the vicinity of the TATA box (Fig. S1), we hypothesized that the functionally relevant Xa27/xa27 promoter polymorphism might have a similar location. Using site-directed mutagenesis, we replaced the polymorphic nucleotides adjacent to the TATA box (one 3-bp InDel polymorphism and an adenine to cytosine substitution) (Fig. S2) in the xa27 promoter by the corresponding nucleotides from the Xa27 promoter. The mutated xa27 promoter (xa27_mut) was fused to the Bs3 cds and the transgene coexpressed with avrXa27 or avrBs3. The xa27_mut promoter construct triggered HR in N. benthamiana when codelivered with avrXa27, but not when codelivered with avrBs3 (Fig. 1B). This indicates that the xa27_mut promoter is functionally identical to the Xa27 promoter and that the specificity-determinant in the Xa27/xa27 promoter is located adjacent to the predicted TATA box. Furthermore, these observations suggest that the xa27 promoter region containing the polymorphisms constitutes the UPT_AvrXa27 box.

The UPT_AvrXa27 Box from the Rice Xa27 Promoter Retains Its Function in the Context of the Pepper Bs3 Promoter.

Previously we delimited the UPT_AvrBs3 box of the Bs3 promoter to an interval of 18 nucleotides and showed that the box retains its function if placed at different positions within the promoters of the pepper Bs3 or tomato Bs4 genes (16). Here, we wanted to test if the UPT_AvrXa27 box from the Xa27 promoter retains its function when transferred into another promoter context. To this end, we introduced the regions encompassing the UPT_AvrXa27 box of Xa27 or the UPT_AvrXa27* box of xa27 into the Bs3 promoter and inserted these sequences in front of the Bs3 cds (Fig. 2A). The former transgene (Xa27/Bs3 [UPT_AvrXa27 box of Xa27 in Bs3 promoter]) mediated HR in N. benthamiana when cotransformed with the avrXa27 gene (Fig. 2B). In contrast, cotransformation with the other construct containing the UPT_AvrXa27* box of xa27 in the Bs3 promoter (xa27/Bs3) did not trigger an AvrXa27 dependent HR (Fig. 2B). These data demonstrate that the UPT_AvrXa27 box can function within different promoter contexts, like the UPT_AvrBs3 box and that the Xa27 promoter but not the xa27 promoter contains a functional UPT_AvrXa27 box.

Fig. 2.

A 16-bp fragment of the rice Xa27 promoter mediates transcriptional activation via AvrXa27. (A) Graphical representation of the analyzed promoter constructs. Yellow arrows and boxes represent the Bs3 promoter and cds, respectively. The UPT_AvrBs3 and UPT_AvrXa27 boxes are shown as blue and black boxes, respectively. A black box with a white bar marks the UPT_AvrXa27* box of the xa27 promoter. Letters on yellow background represent Bs3 promoter sequence. The asterisk (*) marks the position where Xa27/xa27 promoter sequences were inserted. Bold-face black letters mark nucleotide polymorphisms between Xa27 and xa27. Red letters mark the predicted TATA box. The gray background highlights the minimal UPT_AvrXa27 box. “+” and “−” indicates the presence or absence of an HR in N. benthamiana upon codelivery of 35S promoter-driven avrXa27 gene. (B) In planta analysis of UPT_AvrXa27 box deletion constructs. Promoter constructs were delivered together with a 35S promoter-driven avrXa27 gene via A. tumefaciens into N. benthamiana leaves. Asterisks mark areas where only the avr but no R promoter constructs were infiltrated. Dashed lines mark the inoculated areas. Four days after inoculation, the leaves were harvested and cleared with ethanol to visualize the HR (dark areas).

The UPT_AvrXa27 Minimal Box Covers 16 Bps.

Next we determined the boundaries of the UPT_AvrXa27 box by deletion analysis. Of the constructs containing 5′ deletions of the UPT_AvrXa27 box, A-03, A-06, and A-09 still triggered an AvrXa27 dependent HR in N. benthamiana, while A-12 did not (Fig. 2B). Of the constructs containing 3′ deletions, B-03, B-06, B-09, B-10, and B-12 triggered an AvrXa27 dependent HR, while B-13 did not (Fig. 2B). As a result of this deletion analysis, the UPT_AvrXa27 box was delimited to 16 nucleotides (Fig. 2A).

AvrXa27 Binds with High and Low Affinity to the Xa27 and xa27 Promoters, Respectively.

Next we wanted to investigate why AvrXa27 activates the Xa27 promoter but not the xa27 promoter. Electrophoretic mobility shift assays (EMSAs) with a N-terminal 6xHistidine (His)-tagged AvrXa27 fusion protein and biotin-labeled Xa27 and xa27 promoter fragments (Fig. 3A and Fig. S4) showed that AvrXa27 indeed binds with high affinity to the Xa27 promoter fragment and with much lower affinity to the xa27 promoter fragment (Fig. 3B). Importantly, binding of AvrXa27 to labeled Xa27 promoter fragments could be readily out-competed by nonlabeled Xa27 promoter fragments, whereas even a 100-fold excess of nonlabeled xa27 promoter fragments could not out-compete the binding (Fig. 3B). These data demonstrate that AvrXa27 has a significantly higher affinity to the Xa27 promoter than to the xa27 promoter. It is likely that these differing affinities of AvrXa27 for the Xa27 or xa27 promoters provide the molecular basis for the specificity of promoter activation.

Fig. 3.

AvrXa27 binds specifically to the Xa27 promoter. (A) Probes derived from Xa27, xa27, and Bs3 promoters used in EMSAs. Numbering is relative to the transcription start site (TSS). The polymorphisms between the Xa27 and xa27 promoter are shown by white letters and black background. The UPT_AvrXa27 box of the Xa27 and the UPT_AvrBs3 box of the Bs3 promoter are underlined (B) AvrXa27 binds with high and low affinity to the Xa27 and xa27 promoters, respectively. EMSA with AvrXa27 and Xa27- or xa27-derived probes and competition experiment between AvrXa27 and different amounts (in fmol) of a nonlabeled competitor probe in a 6% nondenaturing polyacrylamide gel. Positions of the bound and free probe are indicated on the left. (C) AvrXa27 and AvrBs3 bind specifically to the Xa27 and Bs3 promoters, respectively. EMSA with AvrBs3 and AvrXa27 and Xa27- and Bs3-derived probes in a 6% nondenaturing polyacrylamide gel. Protein amounts are in fmol. Positions of the bound and free probe are indicated on the left.

AvrXa27 Binds to the Rice Xa27 Promoter But Not to the Pepper Bs3 Promoter.

We demonstrated that the Xcv TAL effector AvrBs3 and the Xoo TAL effector AvrXa27 specifically activate the pepper Bs3 and the rice Xa27 promoters, respectively (Figs. 1 and 2). Further EMSA experiments were undertaken to clarify if promoter activation specificity is due to interaction of the TAL effectors with their matching plant promoters. These experiments demonstrated that AvrBs3 interacts with the pepper Bs3 promoter but not with the rice Xa27 promoter (Fig. 3C). Conversely, AvrXa27 interacted with the rice Xa27 promoter but not the pepper Bs3 promoter (Fig. 3C). These data indicate that specific binding to the promoter is the basis for promoter activation in the AvrBs3-Bs3 and the AvrXa27-Xa27 promoter interactions.

UPT Boxes Matching to Distinct TAL Effectors Retain Their Function When Combined into One Complex Promoter.

The finding that the UPT_AvrBs3 and UPT_AvrXa27 boxes can each retain their functions when moved into another promoter suggested that it might be possible to combine functionally distinct UPT boxes to generate a single promoter capable of recognizing multiple Avr proteins. The Bs3 promoter constructs used to delimit the UPT_AvrXa27 box provided an opportunity to test this hypothesis, because these constructs also contain an UPT_AvrBs3 box (Fig. 2A). All of these constructs produced an HR in N. benthamiana when coexpressed with avrBs3, irrespective of which Xa27 or xa27 promoter fragment was inserted into the Bs3 promoter (Fig. S5), demonstrating that UPT boxes with distinct TAL effector specificities can indeed be combined into one complex promoter.

We further explored this concept by introducing the UPT_AvrXa27 box from the rice Xa27 promoter, the UPT_Avrxa27* box from the rice xa27 promoter and the UPT_{AvrBs3Δrep16} box from the pepper Bs3-E promoter, alone or in different combinations, into the pepper Bs3 promoter containing the UPT_AvrBs3 box, and inserted these promoters in front of the Bs3 cds (Fig. 4A). We used A. tumefaciens mediated delivery in N. benthamiana leaves to study recognition specificity of these promoter constructs and found that avrBs3, avrBs3Δrep16, and avrXa27 triggered an HR only when coexpressed with the R gene constructs containing the matching UPT boxes (Fig. 4B). For example, AvrBs3 triggered HR in combination with the promoter constructs Bs3, xa27/Bs3, Xa27/Bs3, Bs3-E/Bs3, Bs3-E/Xa27/Bs3, and Bs3-E/xa27/Bs3 but not in combination with the promoter constructs xa27, xa27_mut, Xa27, and Bs3-E. None of the promoter constructs produced an HR in combination with AvrBs4 (Fig. 4B), which is 96.6% identical to AvrBs3 (8). As expected however, avrBs4 triggered an HR when expressed with its cognate R gene Bs4 (Fig. 4B), which encodes a nucleotide-binding site (NB) leucine rich repeat (LRR) type tomato R protein (19).

Fig. 4.

Tandemly-arranged UPT boxes with distinct TAL effector specificity retain their function when combined into a single promoter. (A) Graphical representation of promoter constructs. Yellow arrows and boxes represent the promoters and cds, respectively of the Bs3 and Bs3-E genes. Bs3 and Bs3-E promoters differ only in their UPT boxes and are therefore displayed in identical color. Green, brown, and orange color represents the promoters (arrows) and cds (boxes) of the tomato Bs4, rice Xa27, and rice xa27 genes, respectively. Pink, black and blue boxes represent the UPT_{AvrBs3Δrep16}, UPT_AvrXa27, and UPT_AvrBs3 boxes, respectively. The black box with the white vertical line represents the nonfunctional UPT_AvrXa27* box of the rice xa27 gene. With exception of the Bs3 cds all elements are drawn to scale. (B) In planta analysis of the promoter constructs. The promoter constructs depicted in (A) were codelivered together with 35S promoter-driven avr genes into N. benthamiana leaves (displayed in boldface letters above each leaf) via A. tumefaciens. Asterisks (*) mark areas in which A. tumefaciens containing only a TAL effector gene were infiltrated. Dashed lines mark the inoculated areas. Four days after inoculation, the leaves were harvested and cleared with ethanol to visualize the HR (dark areas).

We have shown that the placement of the UPT_AvrBs3 box determines the position of the transcription start site (TSS), such that the distance between the TSS and the 3′ end of UPT_AvrBs3 box remains approximately constant (16). On this basis, we anticipated that the complex Bs3-E/Xa27/Bs3 promoter containing the UPT_{AvrBs3Δrep16}, UPT_AvrXa27, and UPT_AvrBs3 boxes would produce different TSSs in combination with the respective TAL effectors AvrBs3Δrep16, AvrXa27, and AvrBs3. Rapid amplification of cDNA ends (RACE) showed that the TSSs of the Bs3-E/Xa27/Bs3 promoter was 311-, 119-, and 59-bp upstream of ATG start codon when coexpressed with AvrBs3Δrep16, AvrXa27, and AvrBs3, respectively (Figs. S6 and S7). The distances between the 3′ end of the UPT boxes and the given TSSs was 45- to 46-, 53- to 54-, and 44-bp for AvrBs3Δrep16, AvrXa27, and AvrBs3, respectively. Notably, these distances observed with the complex Bs3-E/Xa27/Bs3 promoter were not significantly different from the respective distances observed with the native Bs3-E, Xa27, and the Bs3 promoters (Table S1), indicating that the position of the UPT boxes rather than the promoter context defines the TSSs. Taken together, our data indicate that UPT boxes from different plant promoters can be assembled in vitro into one complex promoter in which each UPT box retains its TAL effector specificity.

Discussion

Promoter-Activation Mediated Recognition-The Default Mechanism of Plant R Genes to Detect Matching TAL Effectors?

We have studied the molecular basis of how the TAL effector proteins AvrBs3, AvrBs3Δrep16, and AvrXa27 are recognized by the pepper Bs3, Bs3-E, and rice Xa27 R genes. We determined that these R genes contain distinct promoter motifs (UPT boxes) that are the site of direct physical interaction with and activation by the matching TAL effector proteins. Notably, neither the UPT boxes nor the coding regions of pepper Bs3 and rice Xa27 share any sequence homology, indicating that this mechanism of disease resistance evolved independently in both mono- and dicotyledonous plant species.

With the mechanism of activation elucidated, it is now evident why mutant derivatives of AvrBs3 and AvrXa27 lacking either the NLS or AAD domains fail to activate the Bs3 and Xa27 mediated defense response (18, 20). The dominantly inherited TAL effector R genes Xa7 and Xa10, from rice, also recognize TAL effector proteins but not their NLS and AAD mutant derivatives (21, 22). Thus, Xa7 and Xa10, and possibly other TAL effector R genes, maybe transcriptionally activated by their matching TAL effectors as is the case for pepper Bs3 and rice Xa27.

Transcriptional activation of R promoters is possibly the default mechanism for recognition of TAL effectors by plant immune systems. However, a notable aberration is the tomato R gene Bs4 that mediates recognition of but is not transcriptionally activated by the matching TAL effector AvrBs4 (19, 23). Yet, tomato Bs4 is clearly an exceptional case since it is the only known plant R gene that mediates recognition of NLS-deprived TAL effector mutants (19, 24).

A notable peculiarity of the Bs3 and Xa27 proteins is that they lack any sequence homology to each other or to NB-LRR type R proteins, which appear to represent the most common type of R proteins. Although rice Xa7 and Xa10 are not isolated yet, the physical genomic intervals that contain these R genes have been defined (25, 26). Notably, these intervals lack any genes for NB-LRR proteins or proteins resembling pepper Bs3 or rice Xa27 (25, 26). Thus it seems that TAL effector R proteins are structurally extraordinarily diverse.

These mechanistic insights into how plant R genes mediate recognition of matching TAL effectors suggests alternate cloning strategies for this type of R gene. For example, transcript profiling could be used to identify candidate R genes by virtue of the fact that they are upregulated only upon delivery of a corresponding TAL effector. Chromatin immunoprecipitation (ChIP) and yeast-one hybrid (Y1H) screens might be complementary approaches to identify TAL effector R gene candidates since our studies have shown that a strong interaction between TAL effector and matching UPT box is a prerequisite for transcriptional activation of the Bs3 and Xa27 promoter (Fig. 3). In summary, differential transcript profiling, Y1H screens, and ChIP technology may provide shortcuts to laborious positional cloning methods for the isolation of plant R genes that mediate recognition of TAL effectors.

Recognition of Multiple Pathogen Races Can Be Achieved via a Single Construct.

Most plant R genes are transcribed constitutively and encode NB-LRR proteins that mediate both the recognition of effector proteins, as well as the execution of the defense response. In contrast, the pepper Bs3 and rice Xa27 R genes show a separation of these functions, since recognition of the microbial effectors is mediated through interaction with the UPT box promoter elements, while instigation of the plant's defense response is carried out by the Bs3 and Xa27 proteins. Given that TAL effector R proteins are involved only in the execution of the defense response, but not in Avr perception, may explain their high structural diversity. The molecular basis of NB-LRR mediated Avr recognition is so far poorly elucidated and it has not been possible to combine multiple Avr specificities into one NB-LRR protein. However, with TAL effectors we have now demonstrated that the UPT boxes of three distinct R genes can be combined into one complex R promoter (Fig. 4). A priori, there is no reason to suspect that more than three UPT boxes could not be functionally combined into one complex promoter. Furthermore our approach does not necessarily rely on UPT boxes from known R gene promoters. For example, the Xoo TAL effectors PthXo1, PthXo6, and PthXo7 transcriptionally activate the rice Os8N3, OsTFX1, and OsTFIIAγ1 genes in the context of a compatible interaction (27, 28) and thus, these promoters are likely to contain matching UPT boxes that can be combined into complex R gene promoters. Most recent studies also uncovered how target DNA-specificity of TAL effectors is encoded (29, 30) and thus UPT boxes can now also be predicted in silico, which substantially simplifies the generation of complex promoters. Because several economically significant xanthomonads including Xoo, X. oryzae pv. oryzicola, and X. citri contain multiple TAL effectors it would appear feasible to engineer such a complex R promoter to confer potentially durable disease resistance.

Materials and Methods

Generation of Expression Clones Containing the Bs3 Gene with Promoter Regions of Xa27 and xa27.

The Xa27 and the xa27 promoters were PCR-amplified from genomic rice DNA of the cultivar IRBB27 and IR24, respectively. The amplification was carried out with Phusion high-fidelity DNA polymerase (New England Biolabs) and the primers Xa27–01-fwd-CACC-PR (CACCCTGCAGCTGAACCAAACAGTTTTAGCTCCATCG) and Xa27–01-rev-PR (CACACTGAGACACCCAAGAAGCTGCCTCC). The PCR fragments were cloned into pENTR-D (Invitrogen), sequenced and transferred into the binary-vector pK7-GW-Bs3 (16) by LR recombination (Invitrogen). The promoter xa27_mut was created by Phusion site directed mutagenesis kit (New England Biolabs) using the primers Xa27-mut-01-fwd-PR (AGTGCTATAAATAGAAGAAGAGACCCATAG) Xa27-mut-01-rev-PR (AGTAGTCGTAGTCAACCACAATTCACAAG), that are phosphorylated at their 5′ termini. After sequencing the construct was transferred via LR-recombination in the binary vector pK7-GW-Bs3. pK7-GW-Bs3-derivatives were transformed into A. tumefaciens GV3101 (32) for transient expression assays.

Insertion and Limitation of UPT Boxes.

For the insertion of the UPT_AvrXa27 box, the UPT_AvrXa27* box and the UPT_{AvrBs3Δrep16} box 5′ upstream of the UPT_AvrBs3 box we used the Phusion site directed mutagenesis kit (New England Biolabs). As template we used a pENTR-D, which contains Bs3 promoter and coding sequence (15). For the insertion of the UPT_AvrXa27 box and the UPT_AvrXa27* box we used the primers Xa27-IRBB27inBs3–02-fwd-PR (GTGCTATAAATAGAAGAAGAGACCCATAGAGAGCATCCTGGTTAAACAATGAACACGTTTG) and xa27-IR24inBs3–01-fwd-PR (GTGCTATAAATAGAAGAGACCAATAGAGAGCATCCTGGTTAAACAATGACACGTTTG) in combination with the primer xa27-in3–01-rev-PR (GGTGTGCAAATTGTGGTTTAACCCATAAACTG). For the insertion of the UPT_{AvrBs3Δrep16} box we used the primers 293-bp-ECW-01-fwd-PR (CAATTTTATTATATAAACCTCTCTATTCCACTAAACCATCCTCACAACCAAGTAAACTCAAAGAACTAATCATTGAAC) and box-02–293-rev-PR (CATACTAATTTCATATTTCCCTTGCATAAG). The limitation of the UPT_AvrXa27 box was done using a pENTR-D, which contains the Bs3 promoter with the inserted UPT_AvrXa27 box and the Bs3 coding sequence. The limitation of the 5′ parts of the UPT_AvrXa27 box was also done by site directed mutagenesis using the primers Xa27-Mut-A-3bp-fwd-PR (CTATAAATAGAAGAAGAGACCCATAG), Xa27-Mut-A-6bp-fwd-PR (TAAATAGAAGAAGAGACCCATAGAGAG), Xa27-Mut-A-9bp-fwd-PR, (ATAGAAGAAGAGACCCATAGAG), Xa27-Mut-A-12bp-fwd-PR (GAAGAAGAGACCCATAGAGAGC) in combination with the primer Xa27-Mut-A-01-rev-PR (GGTGTGCAAATTGTGGTTTAACCC). The limitation of the 3′ parts of the UPT_AvrXa27 box was done by site directed mutagenesis using the primers Xa27-Mut-B-3bp-rev-PR (GCTCTCTATGGGTCTCTTCTTC), Xa27-Mut-B-6bp-rev-PR (CTCTATGGGTCTCTTCTTC), Xa27-Mut-B-9bp-rev-PR (TATGGGTCTCTTCTTCTATTTATAGCAC), Xa27-Mut-B-12bp-rev-PR (GGGTCTCTTCTTCTATTTATAGCAC) in combination with the primer Xa27-Mut-B-02-fwd-PR (CTGGTTAAACAATGAACACGTTTGCC). All primers used are phosphorylated at their 5′ termini. After sequencing the constructs were transferred by LR-recombination in the binary-vector pGWB1 (31). pGWB1-derivatives were transformed into A. tumefaciens GV3101 for transient expression assays.

A. tumefaciens Mediated Transient Expression in N. benthamiana.

Agrobacterium strains were grown overnight in yeast extract broth (YEB) medium (5 g bacto beef extract, 1 g bacto yeast extract, 5 g bacto peptone, 5 g sucrose, and 2 mM MgSO₄, pH 7.2, per liter) containing 100 μg/mL each of rifampicin and kanamycin, collected by centrifugation, resuspended in inoculation medium (10 mM MgCl₂, 5 mM Mes, pH 5.3, 150 μM acetosyringone) and adjusted to an OD₆₀₀ of 0.8. Equal amounts of A. tumefaciens strains containing a 35S-promoter driven TAL effector gene and a given promoter construct (fused to the Bs3 coding sequence) were mixed and infiltrated into N. benthamiana leaves by blunt-end syringe infiltration. Leaves were harvested about 4 days post infiltration. For better visualization of the HR, leaves were cleared by incubation in ethanol at 60°C and were dried and photographed.

Electrophoretic Mobility Shift Assay (EMSA).

EMSAs were done as described earlier (15) with the difference that we used a N-terminal 6xHistidine (His) tag instead of a N-terminal GST tag for purification of TAL effector proteins. TAL effector proteins were purified from E. coli BL21-AI with Ni-NTA agarose (Qiagen).

Generation of a Binary Vector that Contains an avrXa27 Gene.

For the generation of a binary vector that contains avrXa27 we used a pKSAvrXa27 (provided by Zhongchao Yin, Temasek, Life Sciences Laboratory) that contains avrXa27. From this vector we amplified the C-terminal BamHI-fragment with the Phusion high-fidelity DNA polymerase (New England Biolabs) using the primers BamHI+CACC+ATG-01-fwd-PR CACCATGGATCCTGGTACGCCCATCGCTGCCGA and BamHI-02-rev-woS-PR GATCGTCCCTCCGACTGAGCCTGACTGAG. The amplified fragment that is flanked by BamHI sites was cloned into pENTR-D (Invitrogen), resulting in pENTR-D-BamHI-avrXa27. After sequencing, the BamHI fragment of pKSAvrXa27 was transferred into pENTR-D-BamHI-avrXa27 resulting in the pENTR-D-avrXa27. The avrXa27 gene was then transferred via LR recombination in the binary vectors pGWB2 or pGWB5 (31). pGWB2 and pGWB5-derivatives were transformed into A. tumefaciens GV3101 for transient expression assays. For EMSA we transferred avrXa27 from pENTR-D-avrXa27 via recombination into pDEST17 (Invitrogen).