(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Use of the Puccinia sorghi haustorial transcriptome to identify and characterize AvrRp1-D recognized by the maize Rp1-D resistance protein [1]

['Saet-Byul Kim', 'Department Of Plant Pathology', 'Center For Plant Science Innovation', 'University Of Nebraska-Lincoln', 'Lincoln', 'Nebraska', 'United States Of America', 'Department Of Entomology', 'Plant Pathology', 'North Carolina State University']

Date: 2024-12

The common rust disease of maize is caused by the obligate biotrophic fungus Puccinia sorghi. The maize Rp1-D allele imparts resistance against the P. sorghi IN2 isolate by initiating a defense response that includes a rapid localized programmed cell death process, the hypersensitive response (HR). In this study, to identify AvrRp1-D from P. sorghi IN2, we employed the isolation of haustoria, facilitated by a biotin-streptavidin interaction, as a powerful approach. This method proves particularly advantageous in cases where the genome information for the fungal pathogen is unavailable, enhancing our ability to explore and understand the molecular interactions between maize and P. sorghi. The haustorial transcriptome generated through this technique, in combination with bioinformatic analyses such as SignalP and TMHMM, enabled the identification of 251 candidate effectors. We ultimately identified two closely related genes, AvrRp1-D.1 and AvrRp1-D.2, which triggered an Rp1-D-dependent defense response in Nicotiana benthamiana. AvrRp1-D-induced Rp1-D-dependent HR was further confirmed in maize protoplasts. We demonstrated that AvrRp1-D.1 interacts directly and specifically with the leucine-rich repeat (LRR) domain of Rp1-D through yeast two-hybrid assay. We also provide evidence that, in the absence of Rp1-D, AvrRp1-D.1 plays a role in suppressing the plant immune response. Our research provides valuable insights into the molecular interactions driving resistance against common rust in maize.

The common rust disease of maize is caused by the obligate biotrophic fungus Puccinia sorghi. Resistance to common rust is controlled by race-specific dominant NLR (nucleotide-binding domain and leucine-rich repeats) genes and by a variety of non-race-specific quantitative trait loci. The maize Rp1-D is a coiled-coil-NLR protein conferring race-specific resistance that includes a rapid localized programmed cell death, hypersensitive response (HR). In this study, to identify AvrRp1-D from an avirulent P. sorghi IN2, we employed the isolation of haustoria, facilitated by a biotin-streptavidin interaction, as a powerful approach. This method proves particularly advantageous in cases where the genome information for the fungal pathogen is unavailable, enhancing our ability to explore and understand the molecular interactions between maize and P. sorghi. The haustorial transcriptome is generated through this technique in combination with bioinformatic analyses. We identified two closely related genes, AvrRp1-D.1 and AvrRp1-D.2, which triggered an Rp1-D-dependent defense response in Nicotiana benthamiana. AvrRp1-D-induced Rp1-D-dependent HR was further confirmed in maize protoplasts. We demonstrated that AvrRp1-D.1 interacts directly and specifically with the leucine-rich repeat domain of Rp1-D through yeast two-hybrid assay. Our research provides valuable insights into the molecular interactions driving resistance against common rust in maize.

Funding: This work was supported by grants from National Institute of Food and Agriculture (NIFA) (award #2022-67013-36504) to RD, PB-K, S-BK, the United States Department of Agriculture, Agricultural Research Service (USDA-ARS) research project 5020-21220-014-00D to MH and National Science Foundation (NSF IOS-2126256) to E.P, J.C. This work was supported by the National Research Foundation (NRF) of Korea grant (2018R1A5A1023599, SRC) to E.P and D.C. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

In the present study, we generated a transcriptome from P. sorghi haustoria and, using a combination of biochemical and functional approaches, identified the effector responsible for the activation of Rp1-D, which we termed a putative AvrRp1-D. We have further characterized aspects of the AvrRp1-D/Rp1-D interaction. Among other things, we identified the regions of both proteins that are important for the specific interaction and showed AvrRp1-D functions to suppress the plant immune response when Rp1-D is absent.

Resistance to maize common rust (CR), caused by the obligate biotrophic fungus P. sorghi, is controlled by race-specific single dominant NLR genes termed Rp genes [ 20 ] and by a variety of non-race-specific quantitative trait loci (QTL) [ 21 ]. CR resistance conferred by Rp genes is generally associated with HR. Over 25 Rp genes have been identified in maize, fourteen of which were mapped to the Rp1 locus (Rp1-A to Rp1-N) [ 22 ]. Subsequently, it was reported that some of these specificities are likely closely linked rather than allelic [ 23 , 24 ]. Rp1-D is a CC-NLR protein conferring race-specific HR-mediated resistance [ 25 , 26 ]. Due to its complex nature, the Rp1 locus is highly unstable and recombinogenic [ 24 ]. The Rp1-D haplotype carries Rp1-D and eight paralogues, called rp1-dp1 to rp1-dp8, seven of which encode full NLR proteins [ 27 ]. A number of novel alleles derived from intragenic recombination between homologs at the Rp1-D locus have been identified [ 28 ]. Rp1-D21, an autoactive allele that causes spontaneous HR in maize and N. benthamiana in the absence of pathogens [ 5 , 29 , 30 ], is a chimeric gene derived from intragenic recombination between rp1-dp2 and Rp1-D [ 27 , 30 ].

As obligate biotrophs, rust fungi are difficult to manipulate, which has constrained research on identifying and characterizing their effectors and Avr genes. Nevertheless, significant progress has been made, especially since genomes of several rust fungi have become available [ 13 ]. In maize, two Avr-genes have been identified from P. polysora, a causal agent of southern corn rust. AvrRppC was cloned by sequencing avirulent isolates against the corresponding resistance gene RppC [ 14 ]. Another gene, AvrRppK was identified through the co-expression with RppK in maize protoplasts and transgenic plant carrying AvrRppK lacking its signal peptide [ 15 ]. Co-expression of almost all of these cloned rust Avr genes with their corresponding R-genes in Nicotiana benthamiana triggers HR cell death [ 10 , 16 – 19 ].

Most bacterial pathogen effectors can be identified by the presence of a type-III secretion signal [ 8 ], while oomycete effectors often possess conserved domains such as the RxLR motif that are also believed to mediate translocation into the host cell [ 9 ]. Fungal effectors do not possess comparable identifying features and instead have been identified using a combination of bioinformatic and differential expression approaches. Most biotrophic fungi, including rusts, produce infection structures called haustoria that penetrate the host cell wall, invaginating the host plasma membrane, forming an intimate connection through which molecules are exchanged in both directions. Most effectors are believed to be expressed in haustoria [ 10 ] and introduced into the host cell via these structures. Haustorial transcriptomes have therefore often been used as a starting point for the bioinformatic detection of effector candidates [ 10 , 11 ]. Other criteria used to identify candidate effectors include small size, high cysteine content, potential secretion signals, and taxonomical specificity. A number of specialized tools have been developed for this purpose [ 12 ].

Most R-genes encode cytoplasmic proteins carrying NLRs (nucleotide-binding domain and leucine-rich repeats). Most NLRs carry either coiled-coil (CC) or Toll-interleukin receptor (TIR) domains at their N-termini. In general, the N-terminal domains appear to be responsible for activating cell death pathways, while the NB-ARC and LRR domains are generally associated with regulating the activity of the R-protein [ 5 – 7 ].

In plants, pathogen recognition occurs via two broadly defined and intimately connected systems, pattern-triggered immunity (PTI) and effector-triggered immunity (ETI). The PTI response is triggered by recognition of microbe-derived molecules known as microbe- or pathogen-associated molecular patterns (PAMPs) by receptors located in the plant cell membrane known as pattern recognition receptors (PRRs). ETI is mediated by intracellular resistance (R)-proteins. Pathogens secrete a class of molecules (usually proteins), known as effectors, into the host cell or apoplast to enhance the pathogenesis process. In adapted pathogens, a subset of these effectors suppresses host PTI and allows infection. R-proteins detect the presence of specific effectors, known as avirulence (Avr) proteins, either through direct interaction or via the action of the effectors on another molecule that is monitored (or ‘guarded’) by the R-protein [ 1 ]. Effector detection triggers the ETI defense response that includes many of the same responses noted for PTI as well as, often, a “hypersensitive response” (HR), a rapid localized cell death response at the point of pathogen penetration [ 2 ]. The ETI and PTI responses are integrated in ways that are not entirely understood, such that each activation appears to potentiate the other response [ 3 , 4 ].

Results

Assembly of the P. sorghi haustorial transcriptome Consistent with our previous study, we observed that infection with P. sorghi isolate IN2 induced HR on the infected leaves of the H95: Rp1-D maize genotype, while H95 and W22 genotypes showed disease symptoms characterized by pustule formation at 5 days post-inoculation (dpi) (Fig 1A). Based on these observations, we hypothesized that P. sorghi IN2 produces an avirulence protein recognized by Rp1-D, which we named AvrRp1-D, and that the recognition of AvrRp1-D by Rp1-D activates defense responses and HR. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Isolation of Puccinia sorghi IN2 haustoria and de novo assembly of transcripts expressed in haustoria. (A) Hypersensitive response on a leaf of the maize line H95:Rp1-D infected with an avirulent P. sorghi IN2 and disease symptoms on the susceptible maize lines H95 and W22 infected with the same isolate. The photos were taken at 6-day post inoculation (dpi). (B) A mixture of chloroplasts and haustoria imaged during the haustoria purification process. Haustoria appear green due to concanavalin-A conjugated with GFP that is specifically bound to them. Chloroplasts appear red due to autofluorescence. (C) Overview of the process used to identify effector candidates expressed in haustoria of P. sorghi IN2 and their screening in N. benthamiana. (D) The heatmap shows the expression pattern of 1,761 secreted proteins, hierarchically clustered using one minus Pearson correlation and complete linkage method. Read counts of each gene were used for the heatmap. Blue and red colors indicate expression levels. The red color indicates a higher expression. https://doi.org/10.1371/journal.ppat.1012662.g001 Since most rust Avr genes are expressed in haustoria, we decided to characterize the haustorial transcriptome. Haustoria were isolated from severely infected W22 leaves at 4 dpi, before sporulation, utilizing a procedure involving biotinylated concanavalin A and streptavidin (Fig 1B) and following RNA isolation and transcriptome sequencing as detailed in the methods section. In a previous study, we conducted deep RNA sequencing of a time course of leaves of the near-isogenic maize lines H95 (susceptible) and H95:Rp1-D (resistant) infected with P. sorghi IN2 [31]. Examining these data, we found that 0.10–0.66% of the derived contigs mapped to P. sorghi genome in H95:Rp1-D and 0.08–44.03% in H95. To investigate the variation within the fungal RNA-seq dataset, we combined the RNA-seq data conducted on H95:Rp1-D and H95 lines at 12, 24, and 120 hours, along with the haustoria samples. After removing sequences matching maize from the dataset, we determined that at 12 and 24 hours, only ∼1% of the data was of fungal origin. However, by 120 hours, over 30% of the data was of fungal origin, reflecting the increase in fungal biomass in the tissue during this time course. A principal component analysis (PCA) was then performed, demonstrating that the three or four biological replicates from each sample were well clustered, thus confirming the robustness of the dataset (S1 Fig). De novo assembly identified 72,538 transcript contigs (S1 Data). We used these data to narrow down our AvrRp1-D candidate gene set, as explained below. While 72,538 effector genes were predicted from the haustorial sample, fewer were predicted from samples derived from infected tissues (22,747, 21,932, and 49,433 at 12, 24, and 120 hpi, respectively).

Identification of AvrRp1-D candidates from P. sorghi IN2 To narrow down the list of 72,538 transcript contigs that might function as AvrRp1-D candidates, we used SignalP4.1 and TMHMM v2.0 (S1 Table). After filtering, 2,687 transcript contigs were identified. Among them, we excluded 926 transcript contigs that were expressed at 0 dpi, which might be maize genes, which left a total of 1,761 predicted proteins as potential effector candidates (Fig 1C). To further refine our search for AvrRp1-D candidate genes, we referred to previously generated RNA-seq data from P. sorghi infected H95 maize lines [31]. We reasoned that effectors would be highly expressed in H95 lines at 12 and 24 hours after infection and enriched in haustoria (Fig 1D). Based on the transcript expression data, we identified the top 251 fungal gene transcripts that were most highly expressed on average over the two-time points as potential effector candidates (S2 Table).

High-throughput screening of effectors in N. benthamiana identifies an AvrRp1-D candidate It has been reported that Rp1-D primarily localizes in the plant cell cytosol when transiently expressed in N. benthamiana [6]. Many previous studies have demonstrated that co-expression of wheat NLRs with their cognate rust Avr proteins in N. benthamiana causes HR [14, 15, 32, 33]. We hypothesized that this might also be the case for Rp1-D and the putative cognate effector. Therefore, we cloned effector candidates lacking their signal peptide sequence into a potato virus X (PVX) vector optimized for in planta expression in N. benthamiana [34]. All clones were sequenced, and the sequencing results revealed that 9 clones represented alternative spliced transcripts with early stop codons. The sequence analysis indicated that, of the 241 de novo assembled sequences, 202 showed approximately 94% or higher identity with the original putative effector sequences. Consequently, we successfully transformed 241 effector candidates into Agrobacterium GV3101 for subsequent experiments. For high-throughput screening of the functional AvrRp1-D, agrobacterium carrying an Rp1-D expression construct was infiltrated using a needless syringe into entire leaves of N. benthamiana. Subsequently, agrobacterium containing PVX vectors carrying genes coding for the effector candidates were introduced into these Rp1-D infiltrated leaves at 1 dpi using a toothpick to prick the leaves at specific points (S2 Fig). Transient expression of 48 of the 241 effector candidates induced cell death in the leaf when co-expressed with Rp1-D. However, we considered the possibility that these effectors might cause Rp1-D-independent cell death or be affected by viral proteins produced in the PVX expression system [34]. To address these possibilities, we re-cloned the 48 effector candidates into a binary vector, pGWB17, driven by the 35S promoter. These candidates were then co-expressed with and without Rp1-D using agroinfiltration in N. benthamiana. In these experiments, two effector candidates, 4A12 and 597, encoding proteins of unknown function, induced Rp1-D independent cell death (S3 Fig). One candidate, 3E3, caused robust and consistent cell death in N. benthamiana when co-expressed with Rp1-D but not when expressed alone. By aligning the cloned gene sequence with the de novo assembled sequence of 3E3, we noted that our original 3E3 clone constituted a partial transcript. We determined the 3’ region of the complete 3E3 gene using 3’ RACE and showed that the full-length 3E3 lacking signal peptide also induced cell death in N. benthamiana dependent on the co-expression of Rp1-D (Fig 2A to 2C). Based on this data, 3E3 was designated as a putative AvrRp1-D. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. AvrRp1-D.1 effector induces Rp1-D-dependent cell death and suppresses chitin-mediated ROS production. (A) Overview of the process for identifying effector candidates conferring cell death in N. benthamiana. (B) Transient expression of Rp1-D:3xHA with AvrRp1-D.1:4xcMYC or empty vector (EV). HA is hemagglutinin. Rp1-D21:3xHA and empty vector were used as positive and negative controls, respectively. Representative leaf was photographed at 4 dpi. The white dashed circles indicate agroinfiltrated areas. 15 individual plants were infiltrated and showed similar results. (C) Protein expression of Rp1-D and AvrRp1-D.1 transiently expressed in N. benthamiana. Total protein was extracted from agro-infiltrated leaves at 36 hours post infiltration (hpi), and anti-HA or anti-cMYC antibody was used to detect the expression of the fused proteins. The sizes of the proteins are indicated on the right. Ponceau S staining of the Rubisco subunit showed equal loading of protein samples. (D) Luminescence detected in protoplasts co-expressing Rp1-D with AvrRp1-D.1 or expressing each gene alone (co-transformed with an empty vector). Lower luminescence indicates increased cell death. Rp1-D21, an autoactive allele of Rp1-D, is used as a positive control for HR activation. Six independent biological replicates were tested. The asterisk indicates a significant difference between Rp1-D/EV and Rp1-D/AvrRp1-D.1 expression. (paired t-test, *p<0.05). (E) AvrRp1-D.1 suppresses chitin-mediated ROS production. The indicated constructs were transiently expressed in N. benthamiana. 48 hpi leaf discs were collected and challenged within chitin, and relative luminescence was monitored using a luminol-based assay. Super Yellow Fluorescent Protein (sYFP2) was used as a reference control. Error bars represent the standard error of the mean (SEM). Three independent experiments were performed with similar results. https://doi.org/10.1371/journal.ppat.1012662.g002 While cloning the full-length AvrRp1-D gene from the cDNA of H95 infected with P. sorghi IN2, we also cloned a variant of AvrRp1-D.1, AvrRp1-D.2, which encoded a protein with a 21-amino acid deletion at the N-terminus and five non-synonymous changes when compared to AvrRp1-D.1. AvrRp1-D.2 induced Rp1-D-dependent cell death in N. benthamiana (S4 Fig). The genomic sequences of the AvrRp1-D.1 and AvrRp1-D.2 genes each comprise four exons with the entire gene spanning 3026 bp in each case. AvrRp1-D.1 (NCBI# OR593746) and AvrRp1-D.2 (NCBI# OR593747) encodes 912 and 891 amino acids, respectively. When they were blasted in NCBI, no similarity to any proteins from other fungal species was found. AvrRp1-D is predicted to have two nuclear localization signals (NLS), one near the N-terminus and one near the C-terminus. As both AvrRp1-D.1 and AvrRp1-D.2 induced the same Rp1-D-dependent cell death and encoded almost identical proteins, we decided to use AvrRp1-D.1 for further experiments.

Expression of AvrRp1-D.1 in maize also causes Rp1-D-specific cell death To verify that AvrRp1-D.1 also conferred Rp1-D-dependent cell death in maize, we used a protoplast system. Rp1-D and AvrRp1-D.1 genes were co-transfected with the firefly luciferase gene into protoplasts isolated from the susceptible maize line H95. Co-expression of AvrRp1-D.1 with Rp1-D and luciferase led to a reduction in luminescence, indicative of cell death, while AvrRp1-D.1 or Rp1-D alone co-expressed with luciferase did not elicit a similar response (Fig 2D). This finding indicates that AvrRp1-D.1 can induce Rp1-D-dependent cell death in both N. benthamiana and maize, suggesting a conserved interaction between the two proteins across expression in different organisms. We also observed that the full-length AvrRp1-D.1, including the signal peptide sequence of 22 amino acids, did not induce cell death in the presence of Rp1-D, indicating the intracellular expression of AvrRp1-D.1 is required for Rp1-D-dependent cell death (S5 Fig).

AvrRp1-D.1 suppresses chitin-mediated ROS production The production and accumulation of reactive oxygen species (ROS) is one of the earliest plant defense responses activated by pathogens [35, 36]. As such, virulent pathogens often secrete effectors either to the host apoplast or cytosol to suppress ROS production, thereby circumventing host immune responses [37]. To assess whether AvrRp1-D.1 suppresses an early defense response, we transiently expressed cMYC-tagged AvrRp1-D (without the predicted signal peptide) in N. benthamiana and tested its ability to suppress PAMP-triggered ROS burst. N. benthamiana leaf discs transiently expressing AvrRp1-D.1 were treated with chitin, and ROS production was monitored over time. We transiently expressed Super Yellow Fluorescent Protein (sYFP2) as a reference control. Transient expression of AvrRp1-D.1 in N. benthamiana suppressed chitin-mediated ROS accumulation by approximately 60% compared to the sYFP2 control (Fig 2E).

Co-expression of six other Rp1 alleles with AvrRp1-D results in little or no cell death Sun et al. [27] identified eight additional Rp1 alleles and Rp1-D at the Rp1-D locus on chromosome 10, designated Rp1-dp1 through Rp1-dp8. All showed a high level of amino acid similarity with Rp1-D, ranging from 88% to 94%, except for Rp1-dp4, which contains an early stop codon in the NB-ARC domain (S3 Table). Of these nine alleles, only Rp1-D conferred race-specific resistance. We would expect therefore that our candidate AvrRp1-D.1 would not confer cell death in N. benthamiana when co-expressed with any of these other seven alleles. We amplified and cloned Rp1-dp1, Rp1-dp2, Rp1-dp5, Rp1-dp6, Rp1-dp7 and Rp1-dp8 into expression vectors. We were not able to amplify Rp1-dp3. When co-expressed with AvrRp1-D.1, Rp1-D induced strong cell death as previously observed, and Rp1-dp2 induced weaker cell death, while expression of the other alleles [Rp1-dp1, 5, 6, 7, 8] with AvrRp1-D.1 did not confer any noticeable cell death (Fig 3). Likewise, AvrRp1-D.2 also triggered cell death when co-expressed with Rp1-D and, more weakly, with Rp1-dp2 (S6 Fig) but not with the other alleles. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. AvrRp1-D.1 specifically activates the Rp1-D allele. (A) Rp1-dp1, -dp2, -dp5, -dp6, -dp7, and -dp8 fused with a C-terminal tag 3xHA were co-infiltrated with AvrRp1-D.1 fused with a C-terminal 4xcMYC or EV in N. benthamiana. A representative leaf was photographed at 4 dpi. 9 individual plants were infiltrated and showed similar results. (B) The table to the right indicates the relative strength of the HR induced by the co-expression of AvrRp1-D.1 with each Rp1 allele (+++ is the strongest HR). Rp1-D21 was used as a positive control. (C) Protein expression of Rp1 alleles and AvrRp1-D.1. Total protein was extracted from agro-infiltrated leaves at 36 hpi, and anti-HA or anti-cMYC antibody was used to detect the expression of the fused proteins. The sizes of the proteins are indicated on the right. Ponceau S staining of the Rubisco subunit showed equal loading of protein samples. https://doi.org/10.1371/journal.ppat.1012662.g003

Analysis of Rp1-D chimeras identifies regions important for activation by AvrRp1-D A previous study [5] demonstrated that the interaction between the NB-ARC domain and the LRR domain enables the induction of HR triggered by the auto-active Rp1-D21. This was accomplished using a series of chimeric proteins generated through domain swapping between Rp1-D and Rp1-dp2 (S7 Fig). To identify the specific domain of Rp1-D that interacts with AvrRp1-D.1, we co-expressed AvrRp1-D.1 with all these chimeric genes, which did not induce auto-active cell death (Fig 4A and 4B). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Investigation of regions of Rp1-D responsible for interaction with AvrRp1-D.1. (A)-(B) All constructs tested were described previously (5). The amino acid position of the recombination site of each construct was indicated above the construct. The red line indicates a mutated amino acid position. The domain swap constructs were fused with the C-terminal tag 3xHA, and AvrRp1-D.1 was fused with the C-terminal tag 4xcMYC. The strength of HR resulting from co-expression in N. benthamiana is shown. 9 individual plants were infiltrated and showed similar results. The table to the right indicates the relative strength of the HR induced by the co-expression of AvrRp1-D.1 with each Rp1 chimera or mutant (+++ is the strongest HR). Rp1-D21 was used as a positive control. (C) Yeast two-hybrid assay using strains co-expressing Rp1-D, -dp2, -dp6, and -dp7 fused to the GAL4 activation domain (AD) with AvrRp1-D.1 fused to the GAL4 DNA binding domain (BD) on control media lacking leucine and tryptophan (-LW) or selective media additionally lacking histidine (-LWH). Growth on selective media indicates protein-protein interactions. Interaction between T antigen with GAL4 activation domain and Lam or p53 with GAL4 DNA binding domain were used as negative/positive control, respectively. AD- or BD-binding proteins were detected using anti-GAL4 and anti-GAL4 (DBD). (D) Schematic diagram of the parts into which Rp1-D was divided for the yeast-two-hybrid assay shown in (E). (E) Yeast two-hybrid assay using strains co-expressing each part of Rp1-D or Rp1-dp7 shown in the schematic above fused to AD with AvrRp1-D.1 or Lam fused to BD on control media lacking leucine and tryptophan (-LW) or selective media additionally lacking histidine (-LWH). Pictures were taken 5 days after plating. https://doi.org/10.1371/journal.ppat.1012662.g004 Among these, the V17 chimera, which was largely identical to Rp1-D but carried the C terminal amino acids 1090–1292 from the LRR domain of Rp1-dp2, lost the ability to induce cell death upon co-expression with AvrRp1-D.1. On the other hand, three chimeric proteins (V1, V7, and V13), containing the last portion of the LRR domain (amino acids 1200–1292) from Rp1-D, induced strong cell death in the presence of AvrRp1-D.1, similar to Rp1-D itself. Interestingly, V18, which carried part of the LRR domain (amino acids 1200–1276) from Rp1-D also did not induce cell death with AvrRp1-D.1. Wang et al. [5] showed that when transiently expressed in N. benthamiana, the single K1184N mutation, which replaced the Rp1-dp2 with the Rp1-D amino acid at that spot, in construct V1 (with a combination of amino acids 1–1200 from Rp1-dp2 and amino acids 1200–1292 from Rp1-D), produced an auto-active cell death while V1 did not. This result suggested that the amino acid 1184 is important for regulating activity. Interestingly, however, Rp1-dp2 (K1184N) did not confer HR, suggesting that additional polymorphisms between Rp1-D and Rp1-dp2 in the C-terminal region are required to confer autoactivity. When AvrRp1-D.1 was co-expressed with Rp1-dp2-K1184N, it conferred stronger cell death than when co-expressed with Rp1-dp2, providing further evidence that this specific amino acid is important for the regulation of Rp1-D activity (Fig 4A). V16, comprising the CC-NB-ARC domain and the N-terminal region of the LRR domain of Rp1-D (amino acids 1–775) and the LRR domain of Rp1-dp2 (amino acids 776–1292), caused moderate cell death when co-expressed with AvrRp1-D. The findings imply that the C-terminal LRR domains of Rp1-D likely play a role in recognizing AvrRp1-D.1. However, the fact that V16 induced a stronger Rp1-D dependent HR compared to Rp1-dp2 suggests that additional portions of Rp1-D are crucial for AvrRp1-D.1 recognition. To investigate this further, we created three chimeric mutants (V1_dp7, V16_dp7, and V17_dp7) which mirror the V1, V16 and V17 chimeras except that the portions from Rp1-dp2 are replaced with portions from Rp1-dp7 (Fig 4). Compared to V7 and V16, both V1_dp7 and V16_dp7 lost the ability to induce AvrRp1-D.1 dependent HR. These results further underline that Rp1 activation through specific recognition is a complex process involving domains throughout the protein. Both Rp1-D and Rp1-dp2 contain so-called MHD and LHD motifs within the NB-ARC domain (S7 Fig). We previously demonstrated that the mutation of the LHD motif in Rp1-dp2 (Rp1-dp2 (D517V)) induced autoactive cell death, while other MHD and LHD mutants, including Rp1-D (H517A), Rp1-D (D518V), Rp1-D (H521A), and Rp1-D (H517AD518V), did not [5]. To determine whether the mutations are involved in AvrRp1-D.1-dependent cell death, we co-expressed the four non-autoactive MHD mutants with AvrRp1-D.1 or AvrRp1-D.2 (S8 Fig). Co-expression of Rp1-D-D518V or Rp1-D-H517AD518V with AvrRp1-D did not induce cell death, suggesting that amino acid 518 might be important for AvrRp1-D dependent activation. In summary, our experiments suggest that amino acids 1090–1292 within Rp1-D, and amino acid 1184 within Rp1-dp2 have a crucial role in the control of its activation by AvRp1-D.1.

Rp1-D and Avr-Rp1-D.1 physically interact in an allele-specific manner To establish the direct interaction between Rp1-D and AvrRp1-D.1, we conducted a yeast two-hybrid assay. Rp1-dp2, Rp1-dp6, and -dp7 were used as controls since they induced lower levels of cell death (Rp1-dp2) or no cell death (Rp1-dp6, Rp1-dp7) when co-expressed with AvrRp1-D.1. Unexpectedly, the full-length proteins of Rp1-D, Rp1-dp2, Rp1-dp6, and Rp1-dp7 all demonstrated weak interactions with AvrRp1-D.1 (Fig 4C) in the yeast two-hybrid system. To further explore the specific domains responsible for the interaction, we tested the individual domains of the Rp1 protein for their interactions with AvrRp1-D.1. The Rp1-D protein was divided into four fragments: CC-NB-ARC, LRR1 (amino acids 528–775), LRR2 (amino acids 776–1046), and LRR3 (amino acids 1047–1292) (Fig 4D). Interestingly, while the LRR2 domains from both Rp1-D and Rp1-dp7 showed strong interactions with AvrRp1-D.1, only the Rp1-D LRR3 domain interacted with AvrRp1-D.1 (Fig 4E). These findings suggest that the LRR3 domain of Rp1-D may be responsible for the direct recognition of AvrRp1-D.1. This is consistent with our findings detailed above that amino 1090–1292 within Rp1-D plays a crucial role in the recognition of AvrRp1-D.1.

AvrRp1-D.1 is nuclear localized, and its N-terminal region is required for Rp1-D activation To gain insights into which region of AvrRp1-D.1 activates Rp1-D, we generated ten deletion mutants (Fig 5A). Each deletion clone was co-expressed with Rp1-D in N. benthamiana. Interestingly, all the deletion clones that carried the N-terminal region amino acids 23–229, namely del-2, -3, -4, -5, and -6, triggered Rp1-D-dependent cell death at 3 dpi with del-3 inducing the strongest cell death response (Fig 5). AvrRp1-D.1 is predicted to have two nuclear localization signals (NLS) at both N-terminus and C-terminus. The co-expression results suggest that the AvrRp1-D.1 N-terminus, which includes a predicted NLS, may play a crucial role in its recognition by Rp1-D and subsequent induction of cell death. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. The N-terminus of AvrRp1-D.1 is essential for the recognition of Rp1-D in N. benthamiana. (A) Schematic representation of truncations of AvrRp1-D.1 with the corresponding amino acid positions on the right. The blue bar at the top is a schematic of the entire AvrRp1-D.1 protein and the grey bars below illustrate the portion that is expressed by each deletion construct. The table to the right indicates the relative strength of the HR induced by the co-expression of AvrRp1-D.1 with each Rp1 allele (+++ is the strongest HR). (B) Co-expression of deletion constructs of AvrRp1-D.1 with Rp1-D or EV in N. benthamiana. Pictures show the abaxial side of the leaves at 4 dpi. 9 individual plants were infiltrated and showed similar results. (C) Protein expression of deletion constructs of AvrRp1-D.1 and Rp1-D from experiment shown in (B). Total protein was extracted from agro-infiltrated leaves at 32 hpi, and anti-HA or anti-cMYC antibody was used to detect the expression of the fused proteins. The sizes of the proteins were labeled on the right. * indicates the target band. Ponceau S staining of the Rubisco subunit showed equal loading of protein samples. (D) The subcellular localization of truncations of AvrRp1-D.1. AvrRp1-D.1 deletion constructs fused with the C-terminal enhanced yellow fluorescent protein (EYFP) were transiently expressed in N. benthamiana. Confocal images were assessed at 30–36 hpi. Confocal micrographs are of single optical sections and the scale bar is 20μm. https://doi.org/10.1371/journal.ppat.1012662.g005 To examine the localization of AvrRp1-D.1 and the possible role of localization in its recognition by Rp1-D, we expressed AvrRp1-D.1 and its truncated forms fused with C-terminal EYFP in N. benthamiana. We observed that AvrRp1-D.1 and del-3, -4, and -5 localized in the nuclei, del-2, -7, and -8 localized predominantly in the nuclei together with weak cytosolic localization, and del-1, -6, -9, and -10 localized predominantly in the cytosol (Fig 5D). Since del-6 did not localize to the nucleus, there was no clear correspondence between nuclear localization and the ability to induce HR via interaction with Rp1-D.

[END]
---
[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1012662

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.