Innate immunity in plants can be triggered by microbe- and pathogen-associated molecular patterns. The pathogen-associated molecular pattern-triggered immunity (PTI) is often suppressed by pathogen effectors delivered into the host cell. Plants can overcome pathogen suppression of PTI and reestablish pathogen resistance through effector-triggered immunity (ETI). An unanswered question is how plants might overcome pathogen-suppression of PTI during ETI. Findings described in this paper suggest a possible mechanism. During Pseudomonas syringae pathovar tomato (Pst) DC3000 infection of Arabidopsis, a host ADP ribosylation factor guanine nucleotide exchange factor, AtMIN7, is destabilized by the pathogen effector HopM1 through the host 26S proteasome. In this study, we discovered that AtMIN7 is required for not only PTI, consistent with the notion that Pst DC3000 degrades AtMIN7 to suppress PTI, but also ETI. The AtMIN7 level in healthy plants is low, but increases posttranscriptionally in response to activation of PTI. Whereas DC3000 infection led to degradation of AtMIN7, activation of ETI by three different effectors, AvrRpt2, AvrPphB, and HopA1, in Col-0 plants blocks the ability of Pst DC3000 to destabilize AtMIN7. Further analyses of bacterial translocation of HopM1 and AtMIN7 stability in HopM1 transgenic plants show that ETI prevents HopM1-mediated degradation of AtMIN7 inside the plant cell. Both AtMIN7 and HopM1 are localized to the trans-Golgi network/early endosome, a subcellular compartment that is not previously known to be associated with bacterial pathogenesis in plants. Thus, blocking pathogen degradation of trans-Golgi network/early endosome-associated AtMIN7 is a critical part of the ETI mechanism to counter bacterial suppression of PTI.
Keywords: type III effector, disease, GTPase
Through evolution, plants have developed a sophisticated immune system against microbial infections. A major branch of innate immune signaling in plants is initiated as a consequence of the detection of specific microbe- and pathogen-associated molecular patterns (MAMPs and PAMPs) by cognate plasma membrane (PM)-localized pathogen recognition receptors (PRRs) (1). Collectively referred to as PAMP-triggered immunity (PTI) (2, 3), this branch of host defense signaling is believed to represent the first line of defense against pathogen infection. Another branch of innate immune signaling is activated when individual pathogen effectors are recognized by disease resistance (R) proteins, resulting in effector-triggered immunity (ETI). In addition, the plant immune responses are regulated by several plant hormones, including salicylic acid (SA), jasmonic acid (JA), and ethylene (4, 5). At present, the molecular connections between different branches of the plant immune system are not well understood.
An important virulence mechanism of Gram-negative bacterial plant pathogens involves the assembly of a type III protein secretion system (T3SS) (6). The T3SS enables bacteria to deliver effector proteins into the host cell to manipulate host cellular functions. Identification of the host targets of effectors have begun to give a glimpse into an impressive array of host cellular processes manipulated by bacteria during pathogenesis (7, 8). The tomato and Arabidopsis pathogen Pseudomonas syringae pathovar (pv.) tomato (Pst) strain DC3000 produces about 30 effectors, including two large-size effectors, AvrE (1,795 amino acids) and HopM1 (712 amino acids). AvrE and HopM1 do not share sequence similarities but are functionally redundant (9, 10). HopM1 and AvrE belong to two of the most widely distributed effector families in P. syringae and/or other bacterial plant pathogens. Importantly, HopM1, AvrE, and orthologs of AvrE have a critical virulence role in diverse plant/bacterial interactions, and their encoding genes are located in the conserved effector locus (CEL) that is physically linked to the T3SS gene cluster in a pathogenicity island (9, 11–13). These observations suggest that coacquisition of a T3SS and the HopM1 and AvrE families of effectors might be a key step in the evolution of P. syringae and several other bacterial pathogens in plants. Thus, elucidating the virulence functions of HopM1 and AvrE is expected to shed light on an important aspect of bacterial pathogenesis in plants.
Recent studies have shown molecular “arms races” between the plant immune system and pathogen virulence effectors. PTI is often suppressed by pathogen effectors delivered into the host cell (14). However, plants can overcome pathogen suppression of PTI and reestablish pathogen resistance through ETI. A major unanswered question is how plants overcome pathogen suppression of PTI during ETI. In a previous study, we identified several host targets of HopM1 in Arabidopsis (15). Among them, AtMIN7 is an ADP ribosylation factor (ARF) guanine nucleotide exchange factor (GEF) protein. HopM1 interacts with and mediates the degradation of AtMIN7 in a proteasome-dependent manner (15). Importantly, atmin7 mutant plants significantly rescue the virulence defect of the Pst DC3000 ΔCEL mutant, which is greatly reduced in the ability to multiply in wild-type Col-0 plants due to deletion of both hopM1 and avrE genes (12, 15). However, the underlying basis for the hypersusceptibility of the atmin7 mutant to the Pst DC3000 ΔCEL mutant and the role of AtMIN7 in various branches of plant innate immunity have remained unknown.
In this study, we performed live cell imaging to show that HopM1 and AtMIN7 are localized in the trans-Golgi network/early endosome (TGN/EE) of plant cells. We found that AtMIN7 plays a broad role in PTI, ETI, and SA-regulated immunity in Arabidopsis. Most significantly, whereas Pst DC3000 infection causes degradation of AtMIN7, activation of ETI by effectors AvrRpt2, AvrPphB, and HopA1 in Col-0 plants completely blocks the ability of Pst DC3000 to degrade AtMIN7. Further analyses show that activation of ETI blocks AtMIN7 degradation in transgenic plants expressing HopM1 inside the plant cell and does not affect bacterial translocation of HopM1 into the host cell. These results have revealed an example in which ETI counters bacterial virulence by blocking pathogen degradation of an immunity-associated ARF-GEF protein in the TGN/EE.
Results HopM1 and AtMIN7 Are Localized in the TGN/EE Compartment.HopM1 has been shown to be in the endomembrane fraction when transgenically expressed in Arabidopsis (15). However, the specific endomembrane organelle in which HopM1 resides is not known. For more precise analysis of endomembrane organelle localization, we constructed HopM1-GFP and YFP-HopM1 fusions and expressed them transiently in Nicotiana benthamiana leaves or stably in Arabidopsis plants. When transiently expressed in N. benthamiana leaves, HopM1-GFP caused necrosis, similar to 6×His-HopM1 (15), although it took longer for HopM1-GFP (about 24 h) to cause necrosis than 6×His-HopM1 (about 10 h) after induction with 30 μM dexamethosone (DEX; SI Appendix, Fig. S1A). HopM1-GFP also destabilized AtMIN7, although it was less potent compared with 6×His-HopM1 (15) (SI Appendix, Fig. S1B). These results suggest that HopM1-GFP is functional, but less active than 6×His-HopM1.
Confocal examination of leaf cells showed that both HopM1-GFP and YFP-HopM1 were localized in small intracellular punctate structures (Fig. 1 and SI Appendix, Fig. S2). These punctate structures are membrane derived because they can be stained by the membrane-binding dye FM4-64 (16) (Fig. 1). Further colocalization with specific subcellular protein markers revealed that the HopM1-associated punctate structures were positive for VHA-a1 (a marker protein for the TGN/EE compartment; Fig. 1) but not for ARA6 (a marker protein for late endosome) or ST (a marker protein for Golgi; SI Appendix, Fig. S2). Thus, the localization of HopM1-GFP is likely TGN/EE specific.
Fig. 1.HopM1 is localized to the trans-Golgi network/early endosome (TGN/EE). (Top) HopM1-GFP is localized in intracellular punctate structures 8 h postinduction with 30 μM DEX in DEX::hopM1-GFP transgenic Arabidopsis. These structures could be stained with FM4-64, indicating that they are membrane derived. (Middle) Transiently expressed HopM1-GFP colocalizes with the TGN/EE marker VHA-a1-RFP in leaf cells of Nicotiana benthamiana. (Bottom) Transiently expressed HopM1-GFP colocalizes with AtMIN7-DsRed2 in leaf cells of Nicotiana benthamiana. (Scale bars: 10 μm.) Dashed lines in bright field images indicate borders between leaf cells.
The TGN/EE localization of HopM1 in leaf cells is interesting because AtMIN7 has recently been localized to the TGN/EE in Arabidopsis root cells (17). We conducted colocalization experiments with HopM1-GFP and AtMIN7-DsRed2 (Fig. 1), which we found to be functional because it complemented the atmin7 mutation in Arabidopsis (SI Appendix, Fig. S3). HopM1-GFP (expressed from the DEX-inducible promoter) and AtMIN7-DsRed2 (expressed constitutively from the CaMV 35S promoter) were found to colocalize 6–8 h after DEX treatment (Fig. 1) before the AtMIN7-DsRed2 signal disappeared. This finding is consistent with the previously observed physical interaction between HopM1 and AtMIN7 in vivo (15), and indicates that HopM1 acts in the TGN/EE compartment.
AtMIN7 Has an Important Role in PTI and SA-Regulated Immunity.HopM1 was previously shown to suppress plant basal immune responses, such as SA-regulated callose deposition in the cell wall (9), suggesting that its host targets (such as AtMIN7) play a role in regulating PTI and/or SA-regulated immunity. To test this hypothesis, we pretreated wild-type Col-0 plants with flg22, a 22-aa MAMP derived from bacterial flagellin (18) or benzothiadiazole (BTH), a synthetic activator known to trigger SA-dependent immunity in plants (19). Compared with water pretreatment, flg22- and BTH-pretreated Col-0 plants exhibited greatly reduced disease symptoms in response to Pst DC3000 infection (Fig. 2). In contrast, flg22- and BTH-treated atmin7 plants showed more disease symptoms following Pst DC3000 infection, and contained significantly increased Pst DC3000 populations compared with flg22- and BTH-pretreated Col-0 plants, respectively (Fig. 2). Thus, AtMIN7 is required for effective PTI and SA-regulated immunity.
Fig. 2.Role of AtMIN7 in BTH- and flg22-triggered immunity. Arabidopsis Col-0 plants were sprayed with H2O, 2 μM flg22, or 30 μM BTH. Plants were dip-inoculated with Pst DC3000 bacteria (1 × 108 cfu/mL) 24 h later. (A) Disease symptoms (chlorosis and necrosis) at day 4 and (B) bacterial populations in infected leaves were recorded. *P < 0.05 and **P < 0.01 between Col-0 and atmin7 plants. (C) AtMIN7 protein amounts in Arabidopsis Col-0leaves 9 h after infiltration with H2O, 30 μM BTH, or 2 μM flg22. One to three closely migrating bands of molecular weights between 200 and 250 kDa could be detected in immunoblotting by the AtMIN7 antibody (Upper). A portion of a Coomassie Blue-stained gel loaded with the same protein samples used in immunoblotting is shown to indicate equal loading (Lower).
To investigate the possibility that flg22 and BTH might induce the expression of the AtMIN7 gene, we treated Col-0 plants with water, flg22, or BTH. However, none of these treatments resulted in reproducible differences in the steady-state RNA level of AtMIN7 (SI Appendix, Fig. S4A). We next examined whether the AtMIN7 protein level is regulated in response to activation of PTI and SA-regulated immunity. In contrast to the RNA level, the AtMIN7 protein level was regulated in various treatments (Fig. 2C). Specifically, the basal AtMIN7 protein level in water-treated healthy plants was very low. Treatment with flg22 and BTH increased the AtMIN7 protein level significantly (Fig. 2C). Thus, expression of AtMIN7 is under posttranscriptional control during PTI and SA-regulated immunity.
Because AtMIN7 is one of the ARF-GEF proteins, which are key regulators of vesicle trafficking, it could control one or more steps in PTI and SA-regulated immunity pathways. Our analyses indicate that AtMIN7 does not affect the expression of FRK1 (a marker gene for PTI) and/or PR1 (a marker gene for SA-regulated immunity) in response to flg22, BTH, or bacteria (SI Appendix, Fig. S4B), but affects callose deposition and the secretion of a specific subset of BTH-inducible proteins (SI Appendix, Fig. S5). Taken together, these results suggest that AtMIN7 acts downstream or independent of immune gene expression to control protein secretion and callose deposition.
A Prominent Role of AtMIN7 in Effector-Triggered Immunity.One of the most important tenets of the current models depicting the evolution of the plant immune system is that ETI can overcome pathogen suppression of PTI and reestablish plant resistance (2, 3). How plants reestablish a resistant state during ETI in the context of apparently widespread suppression of PTI-associated responses by effectors is not known. Because AtMIN7 is degraded by HopM1, we initially thought that Arabidopsis might bypass AtMIN7 and activate an alternative ARF-GEF–regulated vesicle traffic pathway during ETI. Accordingly, we did not expect that AtMIN7 would be required for ETI. Contrary to this expectation, however, we found that AtMIN7 is required for ETI. Specifically, Pst DC3000 expressing AvrRpt2 activates a strong ETI in Col-0 plants, which carry the cognate disease-resistant protein RPS2 (20), resulting in greatly reduced Pst DC3000(avrRpt2) multiplication and disease symptoms compared with Pst DC3000 infection (Fig. 3A and SI Appendix, Fig. S6A). In contrast, Pst DC3000(avrRpt2) multiplied to high levels and showed marked disease symptoms in atmin7 plants. Pst DC3000(avrRpt2) also induced a significantly higher level of callose deposits in Col-0 leaves compared with that in atmin7 leaves (Fig. 3B), suggesting that AtMIN7 is required not only for PTI-associated callose deposition (SI Appendix, Fig. S5), but also for callose deposition during ETI. In contrast, AvrRpt2-triggered hypersensitive response (HR), a rapid, localized cell death response, was similar in timing or frequency between Col-0 and atmin7 leaves (SI Appendix, Table S1).
Fig. 3.The role of AtMIN7 in effector-triggered immunity. (A) Multiplication of Pst DC3000(avrRpt2) in Col-0 and atmin7 plants. Arabidopsis plants were dip inoculated with bacteria at 1 × 108 cfu/mL. Bacterial populations were determined at day 4. *P < 0.05 and **P < 0.01 between Col-0 and atmin7 plants. (B) Callose deposition in leaves of Col-0 and atmin7 plants. Arabidopsis leaves were hand infiltrated with 1 × 108 cfu/mL bacteria. Leaves were stained for callose deposition 9 h after treatment. Average numbers of callose depositions per field of view (0.9 mm2) are presented with SDs displayed as errors. *P < 0.05 and **P < 0.01 between Col-0 and atmin7 plants. (C) Responses of Col-0 and atmin7 plants to nonhost P. syringae strains. Plants were dip inoculated with Pst DC3000 or a nonhost pathogen, P. syringae tabaci 11528 or P. syringae pv. phaseolicola NPS3132, at 1 × 108 cfu/mL. Bacterial populations were determined at day 4. *P < 0.05 between Col-0 and atmin7 plants.
To test whether the requirement of AtMIN7 for ETI is specific to AvrRpt2, we performed experiments with AvrPphB, an effector from P. syringae pv. phaseolicola that triggers RPS5-dependent immunity in Col-0 plants (21). As expected, Pst DC3000(avrPphB) multiplied poorly in Col-0 plants (SI Appendix, Fig. S6B). However, AvrPphB-triggered immunity is severely compromised in the atmin7 mutant, as evidenced by increased multiplication of Pst DC3000(avrPphB) and more prominent disease symptoms in atmin7 plants, compared with those in Col-0 plants (SI Appendix, Fig. S6C). These results suggest that AtMIN7 may be generally required for the execution of ETI. atmin7 plants maintained resistance to incompatible bacterial pathogens P. syringae pv. phaseolicola NPS3132, which infects bean, and P. syringae pv. tabaci 11528, which infects tobacco. These incompatible bacteria multiplied poorly in Col-0 and atmin7 mutant Arabidopsis; no reproducible population differences were detected in atmin7 vs. Col-0 plants (Fig. 3C).
Prevention of AtMIN7 Protein Degradation During ETI.We found the requirement of AtMIN7 for AvrRpt2- and AvrPphB-triggered immunity very intriguing, because AtMIN7 is destabilized by Pst DC3000, which is the recipient of the avrRpt2 and avrPphB genes in these experiments. We considered the possibility that ETI might induce the expression of the AtMIN7 gene in Col-0 plants to compensate for HopM1-mediated degradation of the AtMIN7 protein. To examine this possibility, we infected Col-0 plants with the ΔCEL mutant, Pst DC3000, or Pst DC3000(avrRpt2), and examined for AtMIN7 transcript levels by RT-PCR. None of the treatments resulted in observable differences in the steady-state RNA level of AtMIN7 (Fig. 4A). Consistent with results shown in SI Appendix, Fig. S4A, in these experiments Col-0 plants treated with water, flg22, or BTH also did not result in obvious differences in the steady-state RNA level of AtMIN7 (Fig. 4A).
Fig. 4.(A) RT-PCR analysis of AtMIN7 and ACTIN1 gene expression. Arabidopsis Col-0 leaves were hand infiltrated with 1 × 108 cfu/mL bacteria, 30 μM BTH, or 2 μM flg22. Total RNA samples were purified 9 h after treatment and subjected to RT-PCR assays (25 cycles). (B–D) Western blot analysis of the AtMIN7 protein. Arabidopsis Col-0 leaves were hand infiltrated with H2O, 1 × 108 cfu/mL bacteria, 30 μM BTH, or 2 μM flg22. atmin7 leaves infiltrated with distilled water were used as a negative control. Total protein samples were extracted from leaves 9 h after treatment and loaded onto SDS/PAGE gel. AtMIN7 was detected with AtMIN7 antibody (α-AtMIN7).
We next examined whether the AtMIN7 protein level is regulated in response to ETI. Again, the basal AtMIN7 protein level in water-treated healthy plants was very low, and treatment with flg22 and BTH increased the AtMIN7 protein level significantly (Fig. 4B). Inoculation with the T3SS-defective hrcC mutant and the ΔCEL mutant also increased the AtMIN7 protein level significantly, consistent with activation of PTI and SA-regulated immunity by these mutants (9, 22). Pst DC3000 infection and infection by ΔCEL mutant (hopM1) resulted in marked reduction of the AtMIN7 protein level (Fig. 4B and SI Appendix, Fig. S4C), consistent with HopM1-mediated degradation of AtMIN7 observed previously (15). Remarkably, Pst DC3000(avrRpt2), which triggers ETI in Col-0, prevented AtMIN7 degradation (Fig. 4 B and C). In several experiments, we noticed slightly altered mobility of AtMIN7 in Pst DC3000(avrRpt2)-infected leaves, compared with that in other treatments (Fig. 4C). However, whereas stable AtMIN7 was observed in all Pst DC3000(avrRpt2) infection experiments, the mobility change was not observed in every experiment, presumably because of the large size of AtMIN7 (between 200 and and 250 kDa) and variable SDS/PAGE conditions.
To determine whether prevention of AtMIN7 degradation is unique to AvrRpt2-triggered immunity, we studied Pst DC3000(avrPphB) and Pst DC3000(hopA1). HopA1 is an effector from P. syringae pv. syringae and triggers the RPS6-dependent ETI pathway in Col-0 plants (23). Like in Pst DC3000(avrRpt2)-inoculated leaves, AtMIN7 degradation was prevented in leaves inoculated with Pst DC3000(avrPphB) or Pst DC3000(hopA1) (Fig. 4D). Thus, prevention of AtMIN7 degradation appears to be common in different ETI pathways.
ETI Does Not Affect Type III Translocation of HopM1.Recent studies show that preactivation of PTI can inhibit type III translocation of effectors by P. syringae (24, 25). We therefore considered the possibility that ETI may also block type III translocation of effectors, including HopM1, as a mechanism to prevent AtMIN7 degradation. To test this possibility, we conducted adenylate cyclase (CyaA)-based type III translocation assay. Pst DC3000(avrRpt2), at the inoculums of 1 × 108 cfu/mL, caused ETI-associated HR cell death at 12 h postinfiltration under our experimental conditions. Also, we could observe clear degradation of AtMIN7 by Pst DC3000 and stabilization of AtMIN7 during ETI within 9 h after bacterial infiltration (Fig. 4). For these reasons, we monitored type III translocation of HopM1 for 9 h after bacterial infiltration to avoid potentially nonspecific cellular damages that occur during the HR. Similar time courses were used in other studies (25). Col-0 leaves were infiltrated with Pst DC3000(hopM1-cyaA), Pst DC3000(avrRpt2, hopM1-cyaA), or Pst DC3000(avrPphB, hopM1-cyaA), and the adenylate cyclase activity (as indicated by the cAMP level) was measured, following previously established procedures (24, 25). As expected, cAMP levels increased over time in Pst DC3000(hopM1-cyaA)-infiltrated leaves, indicative of HopM1-CyaA translocation into the host cell (Fig. 5A). As a negative control, Col-0 leaves infiltrated with the T3SS-defective hrcC mutant did not have detectable CyaA activity. In contrast, cAMP levels in leaves infiltrated with Pst DC3000(avrRpt2, hopM1-cyaA) or Pst DC3000(avrPphB, hopM1-cyaA) were similar to that in leaves infiltrated with Pst DC3000(hopM1-cyaA) (Fig. 5A). We also measured the bacterial population levels, and no significant difference was observed within the experimental period (SI Appendix, Fig. S7). These results suggest that AvrRpt2- and AvrPphB-triggered ETI does not block type III translocation.
Fig. 5.(A) Detection of bacterial translocation of HopM1-CyaA during bacterial infection. Bacterial strains harboring a shcM-hopM1-cyaA reporter plasmid were hand infiltrated into Arabidopsis Col-0 leaves at 1 × 108 cfu/mL. Samples were collected at 0 and 9 h postinfiltration. **P < 0.01 indicates significant difference in cAMP levels between Pst DC3000-infected leaves and hrcC mutant-infected leaves. (B–E) Western blot analysis of the AtMIN7 protein in plants. (B) Col-0 gl1 and 6×His-HopM1 Arabidopsis leaves were coinfiltrated with 1 × 108 cfu/mL bacteria plus DEX (final concentration of 10 nM). (C) Col-0 gl1 and 6×His-HopM1 Arabidopsis leaves were coinfiltrated with 1 × 108 cfu/mL bacteria plus DEX (final concentration of 200 nM) or with H2O plus DEX (final concentration of 200 nM). (D) 10 nM DEX was infiltrated into leaves of 6×His-HopM1 plants 9 h before infiltration with H2O or 1 × 108 cfu/mL bacteria. (E) H2O or 1 × 108 cfu/mL bacteria were infiltrated into leaves of 6×His-HopM1 plants 3 h before infiltration of 10 nM DEX. Total protein samples were extracted from leaves 9 h after the last leaf infiltration. AtMIN7 (indicated by arrow) was detected using an AtMIN7 antibody (α-AtMIN7).
ETI Blocks AtMIN7 Degradation by HopM1 Expressed Inside the Plant Cell.The lack of an effect of ETI on type III translocation of HopM1 raises the intriguing possibility that AtMIN7 degradation might be blocked inside the plant cell. To directly test this possibility, we examined AtMIN7 degradation in HopM1 transgenic plants (15). In these plants, 6×His-HopM1 is expressed under the DEX-inducible promoter and 3–10 nM DEX was sufficient to restore the growth of the ΔCEL mutant in HopM1 transgenic plants without dramatically increasing hrcC mutant growth (SI Appendix, Fig. S8) (15). This result suggests that this DEX concentration induces a near physiological level of HopM1 inside the host cell.
We coinfiltrated 10 nM DEX with Pst DC3000 or Pst DC3000(avrRpt2) into Col-0 gl1 and HopM1 transgenic plants and analyzed AtMIN7 protein levels by Western blot. As expected, Pst DC3000 infection resulted in degradation of AtMIN7 in both Col-0 gl1 and HopM1 transgenic plants (Fig. 5B). AtMIN7 remained stable not only in Col-0 gl1 plants, but also in HopM1 transgenic plants infected with Pst DC3000(avrRpt2) (Fig. 5B). This result provides more direct evidence that ETI can stabilize AtMIN7 inside the host cell.
The level of AtMIN7 was somewhat lower in HopM1 transgenic plants than that in Col-0 gl1 plants in both Pst DC3000 (avrRpt2) and Pst DC3000 infections (Fig. 5B), likely because HopM1 transgenic plants produced a higher-than-physiological level of HopM1 protein delivered by bacteria. This observation raises the possibility that out-of-context induction of HopM1 in the transgenic plants might overwhelm ETI. We conducted experiments to test this hypothesis. HopM1 transgenic plants were coinfiltrated with DC3000(avrRpt2) and a 20-fold higher level of DEX (200 nM). We found that the 200 nM DEX treatment eliminated the ability of ETI to protect AtMIN7 from HopM1-mediated degradation (Fig. 5C). We next investigate whether preinduction of HopM1 could also reduce the ability of ETI to prevent AtMIN7 degradation. HopM1 transgenic plants were treated with 10 nM DEX 9 h before Pst DC3000(avrRpt2) infection. Indeed, pretreatment of 10 nM DEX diminished the ability of ETI to protect AtMIN7 (Fig. 5D).
Next, we examined whether preinduction of PTI could inhibit HopM1-mediated AtMIN7 degradation. HopM1 transgenic plants were infiltrated with the hrcC mutant 3 h before treating the same plants with 10 nM DEX to induce the expression of HopM1. In these experiments, Pst DC3000(avrRpt2) infection was used as a control. We found that preinduction of PTI could slightly inhibit HopM1-mediated degradation in HopM1 transgenic plants, compared with water treatment (Fig. 5E). In all experiments, ETI had a stronger protective effect than PTI. Taken together, these results indicate that there might be a “race” between HopM1-mediated AtMIN7 degradation and plant defense-mediated AtMIN7 stabilization inside the plant cell, and that timing and/or strength of plant defense and the amount of HopM1 are important in determining the final extent of AtMIN7 degradation.
DiscussionIn this study we found that AtMIN7 is required for PTI and SA-regulated immunity. These findings are consistent with the hypothesis that HopM1-mediated degradation of AtMIN7 is part of a bacterial virulence strategy to suppress PTI and SA-regulated immunity in susceptible plants (9, 15), and support the current models depicting the host-pathogen arms race (2, 3). A major unresolved issue in the current host-pathogen arms race models, however, is how plants reestablish pathogen resistance during ETI in the context of pathogen destruction of PTI-associated components. Do plants restore/repair existing components or bypass pathogen-suppressed components entirely and activate alternative defense pathways during ETI? Of note, PTI and ETI activate overlapping immune responses, such as immune gene expression (26) and callose deposition (27 and this study). Here, we present experimental evidence that AtMIN7 is also required for ETI. Most importantly, we found that ETI blocks Pst DC3000 degradation of AtMIN7 in resistant plants, illustrating an example in which a pathogen-targeted component of PTI and SA-regulated immunity is protected during ETI. Using the CyaA-based assay, we detected normal translocation of the HopM1-CyaA fusion in ETI-activated leaves (Fig. 5A). Moreover, we found that ETI could stabilize AtMIN7 even when HopM1 is directly produced in transgenic plants, independently of bacterial production and translocation of HopM1 (Fig. 5B). Taken together, these results suggest that ETI prevents AtMIN7 degradation inside the plant cell. Our experiments do not rule out the possibility that type III translocation of effectors and/or general bacterial physiology might be affected at some late stage of infection, which may be important as an additional host preventive mechanism to protect host proteins from the action of bacterial effectors during ETI.
Our data also begin to suggest a race between HopM1-mediated AtMIN7 degradation and plant defense-mediated AtMIN7 stabilization inside the plant cell. Timing and/or strength of plant defense and the amount of HopM1 seem to be important in determining the final extent of AtMIN7 degradation. For example, inoculations with the hrcC mutant and the ΔCEL mutant increase the AtMIN7 protein level posttranscriptionally (Fig. 4), consistent with activation of PTI and SA-regulated immunity by these mutants during infection (9, 22). Pst DC3000 or expression of HopM1 in the ΔCEL mutant degrades AtMIN7 during infection (Figs. 4 and 5 and SI Appendix, Fig. S4C), suggesting that the physiological amount of HopM1 delivered by Pst DC3000 or the ΔCEL mutant (hopM1) is sufficient to degrade the amount of AtMIN7 that is accumulated during infection by the ΔCEL mutant. However, plants appear to have evolved ETI with proper speed and strength that are sufficient to prevent degradation of AtMIN7 by the physiological level of HopM1 during infection. Interestingly, ETI-mediated protection of AtMIN7 is diminished if HopM1 production in transgenic plants is artificially induced using a high dose of DEX or 9 h before activation of ETI (Fig. 5 C and D). Conversely, preactivation of PTI could slightly inhibit HopM1-mediated degradation of AtMIN7 in HopM1 transgenic plants (Fig. 5D). We have shown that HopM1 degradation of AtMIN7 is dependent on the host 26S proteasome (15). Future research should determine whether modulation of AtMIN7 levels during various forms of plant immunity is achieved through modification of AtMIN7 (as indicated by altered mobility in some experiments; Fig. 4), HopM1, and/or the 26S proteasome activity. Also, it would be exciting to determine whether prevention of AtMIN degradation is part of a more global plant mechanism to protect/restore PTI components during ETI.
Based on normal PTI and SA-associated nuclear gene expression in the atmin7 mutant (SI Appendix, Fig. S4), but altered protein secretion and callose deposition patterns (SI Appendix, Fig. S5), we speculate that AtMIN7 likely defines a regulatory step controlling downstream vesicle trafficking, the endpoint of various plant immune pathways. Surprisingly, we found that secretion of PR-1 protein was not affected in the atmin7 plants in response to BTH activation. We also found no effect of the atmin7 mutation on the development of HR (SI Appendix, Table S1). Instead, we detected a group of proteins of 17–28 kDa in the IWFs of BTH-activated Col-0 leaves, but not in BTH-activated atmin7 leaves. Thus, secretion of PR1 and HR development are not sufficient for plant resistance to Pst DC3000 in the atmin7 mutant, and AtMIN7 probably controls the secretion of a different subset of defense-associated proteins. In addition, we observed significantly reduced callose deposition in the cell wall of atmin7 mutant leaves in response to flg22, ΔCEL mutant, or Pst DC3000(avrRpt2) (15) (Fig. 3 and SI Appendix, Fig. S5). AtMIN7 has also been shown to be involved in the recycling of auxin efflux carrier PIN1 in root cells, and atmin7 plants exhibit some root development phenotypes (19). Auxin has been implicated in modulating Arabidopsis susceptibility to Pst DC3000 (28–30). Taken together, we speculate that AtMIN7 may control the traffic of multiple immunity-associated cargoes that collectively contribute to Arabidopsis resistance to Pst DC3000.
Manipulation of regulators of host vesicle traffic has emerged as one of the most important virulence strategies in the pathogenesis of mammalian pathogenic bacteria (31, 32). In recent years, an increasing number of vesicle trafficking components have been discovered as regulators of plant immunity and pathogenesis (33, 34). HopM1 was the first effector from an extracellular bacterial pathogen shown to target host ARF-GEF-regulating vesicle traffic (15). A recent paper reports that an effector (EspG, which shares no sequence similarity to HopM1) from the extracellular human pathogen EHEC O157:H7 also targets ARF-regulated traffic (35), suggesting exciting conceptual parallels between plant and human bacterial pathogenesis. Confocal microscopic imaging conducted in this study (Fig. 1) and a previous study (19) localized HopM1 and/or AtMIN7 to the TGN/EE. The requirement of AtMIN7 for PTI, ETI, and SA-regulated immunity suggests that the TGN/EE is a critical common subcellular organelle for the execution of multiple plant immune pathways.
Materials and MethodsAll experiments reported in this paper were performed three or more times with similar results. Confocal imaging was performed with leaf tissues from Nicotiana benthamiana and Nicotiana tabacum plants (for transient expression) or Arabidopsis (stable expression). Analysis of AtMIN7 and bacterial infections followed the protocols described previously (15). Callose staining was conducted following the protocol reported by Hauck et al. (22). Methods for other experiments and information about bacterial strains and plant materials can be found in SI Appendix.
Supplementary MaterialSupporting Information
AcknowledgmentsWe thank Krithika Shanmugasundaram, who was involved in the construction of HopM1-GFP and YFP-HopM1 fusion constructs as part of Michigan State University's High School Honors Science/Mathematics/Engineering Program. We thank Dr. Federica Brandizzi and Dr. Melinda Frame for assistance with confocal microscopy, and James Alfano for providing the Gateway CyaA fusion vector pML123::CYA and a hopA1 plasmid. Karen Bird edited the manuscript. This work was supported by funding from the National Institutes of Health (AI060761), the Chemical Sciences, Geosciences and Biosciences Division, Office of Basic Energy Sciences, Office of Science, US Department of Energy (DE-FG02-91ER20021), and the National Science Foundation (IOB 0444915) (to S.Y.H.).
Footnotes ReferencesThis section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary MaterialsSupporting Information
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4
