Linking pattern recognition and salicylic acid responses in Arabidopsis through ACCELERATED CELL DEATH6 and receptors

Abstract

The Arabidopsis membrane protein ACCELERATED CELL DEATH 6 (ACD6) and the defense signal salicylic acid (SA) are part of a positive feedback loop that regulates the levels of at least 2 pathogen-associated molecular patterns (PAMP) receptors, including FLAGELLIN SENSING 2 (FLS2) and CHITIN ELICITOR RECEPTOR (LYSM domain receptor-like kinase 1, CERK1). ACD6- and SA-mediated regulation of these receptors results in potentiation of responses to FLS2 and CERK1 ligands (e.g. flg22 and chitin, respectively). ACD6, FLS2 and CERK1 are also important for callose induction in response to an SA agonist even in the absence of PAMPs. Here, we report that another receptor, EF-Tu RECEPTOR (EFR) is also part of the ACD6/SA signaling network, similar to FLS2 and CERK1.

Abbreviations

ACD6

ACCELERATED CELL DEATH 6

PM

plasma membrane

ER

endoplasmic reticulum

SA

salicylic acid

BTH

benzo (1, 2, and 3) thiadiazole-7-carbothioic acid

PRRs

pattern responses receptors/co-receptors

FLS2

FLAGELLIN SENSING 2

BAK1

BRASSINOSTEROID-ASSOCIATED KINASE 1

CERK1

CHITIN ELICITOR RECEPTOR/LYSM domain receptor-like kinase 1

EFR

EF-Tu RECEPTOR

PAMPs

pathogen-associated molecular patterns

BABA

β-aminobutyric acid

LecRK-VI.2

LECTIN RECEPTOR KINASE-VI.2.

The multipass membrane protein ACCELERATED CELL DEATH 6 (ACD6) localizes at the plasma membrane (PM) and endoplasmic reticulum (ER) and acts in a positive feedback loop with the defense signal salicylic acid (SA).^1,2 Arabidopsis plants that lack ACD6 have a kinetic defect in SA accumulation during infection and they are less responsive to the SA agonist benzo (1, 2, and 3) thiadiazole-7-carbothioic acid (BTH).^3,4 Plants with extra copies of ACD6 are more responsive to the SA agonist BTH, whereas plants with a gain of function version of ACD6 (acd6-1) have high SA levels and constitutively active defenses.^2,5,6 A key break-through toward understanding the mechanism of ACD6 came from examining its relationship with pattern responses receptors/co-receptors (PRRs). The abundance of 3 PRRs, including FLAGELLIN SENSING 2 (FLS2), BRASSINOSTEROID-ASSOCIATED KINASE 1 (BAK1) and CHITIN ELICITOR RECEPTOR (LYSM domain receptor-like kinase 1, CERK1) is positively regulated by ACD6. These proteins are also associated with ACD6.^2,4 Plants lacking ACD6 have reduced responses to the ligands for FLS2 and CERK1: flg22, a peptide derived from bacterial flagellin^7,8 and chitin, a fungal cell wall component,⁹ respectively. Thus, a major function of ACD6 is to regulate receptors.

FLS2, BAK1 and CERK1 levels are dynamically regulated by SA, which is reflected in changes in the sensitivity of plants to the respective ligands at different times after BTH treatment. Longer treatments (≥24 h) with BTH cause these proteins to accumulate to high levels. Additionally, plants lacking any one of these proteins show less induction of callose, a cell wall modification, in response to BTH. Thus, PRRs are multifunctional: they can respond both to a specific ligand as well as an SA agonist. The outputs of the responses to different stimuli are not identical, but they share callose deposition changes.⁴

A key question is whether an additional Arabidopsis PRR, EF-Tu RECEPTOR (EFR), like FLS2 and CERK1, is important for responsiveness to SA and affected by ACD6 activation.

EFR is a PRR that responds to elf18, a peptide elicitor that is derived from EF-Tu.¹⁰ Here, we report that EFR is part of the ACD6/SA signaling network and ACD6 is needed for maximal responsiveness to elf18.

All experiments were performed as described in Tateda et al.,⁴ except here we also used elf18 as a pathogen-associated molecular pattern (PAMP) and EFR antibody¹¹ for Western blot analysis. Sequence data from this article can be found in the EMBL/GenBank data libraries under accession numbers as At4g14400 (ACD6) and At5g20480 (EFR). Arabidopsis thaliana Columbia (Col) wild-type accession, acd6-1 (dominant gain of function allele of ACD6),⁶ acd6-2 (loss of ACD6 mutant, SALK_045869¹²) and efr (loss of EFR mutant, SALK_044334¹⁰) plants were grown as described^12,13 in 16 h light/8 h dark conditions for 21 d and then harvested for various assays.

ACD6 regulates several receptors. To determine if ACD6 might also regulate EFR, we sought to monitor EFR levels. In wild-type plants, EFR protein accumulated to very low levels (Fig. 1), so it was not possible to accurately assess EFR accumulation in ACD6 loss of function plants. However, in gain of function acd6-1 plants, EFR protein accumulated to higher levels than in wild type, similar to the behavior of the receptors/co-receptors FLS2, BAK1 and CERK1.⁴ This suggests that ACD6 also regulates EFR directly or indirectly (e.g., possibly due to high levels of signal molecules such as SA, etc.)

Figure 1.

EFR protein accumulates to a high level in plants with a dominant gain of function allele of ACD6 (acd6-1). Microsomal fractions were isolated from wild-type (Col) and acd6-1 plants as described.⁴ To detect EFR levels, membrane proteins separated by 10% of SDS-PAGE were transferred to a PVDF membrane, treated with primary EFR antibody (1:2500) overnight at 4C, and secondary horseradish peroxidase conjugated anti-rabbit antibody (Thermo Scientific) used at 1:1000 (3 h, room temperature). SuperSignal West Pico Stable Peroxidase and SuperSignal West Femto Stable Peroxidase (Thermo Scientific) were used to detect the signals. CBB: coomassie blue stained. This experiment was repeated 3 times with similar results.

If ACD6 has a regulatory role for EFR function, then plants lacking ACD6 should have a reduced response to the EFR ligand elf18. To test this, we treated wild type and acd6-2 loss of function plants with elf18 and monitored callose deposition. As with other PAMPs (flg22 and chitin),⁴ the response of acd6-2 to elf18 was reduced compared to wild type (Fig. 2). This indicates that ACD6 has a role in determining the sensitivity of Arabidopsis to the EFR ligand.

Figure 2.

Plants lacking ACD6 show reduced callose deposition in response to elf18. Leaves of wild-type (Col) and acd6-2 plants infiltrated with 1 µM elf18 were collected 18 h later and stained with aniline blue. Callose was quantified by manually counting deposits and presented as percentage of deposits in Col. Bars show standard error of 3 independent experiments analyzed together (n > 24). Asterisk indicates significant difference from Col, P < 0.05, student's t-test.

Finally, we recently showed that receptors regulated by ACD6 are needed for the callose response to the SA agonist BTH. EFR was also strongly needed for callose deposition in response to BTH, since efr plants had a weaker response than wild type (Fig. 3).

Figure 3.

Plants lacking EFR show reduced callose deposition in response to the SA agonist BTH. Leaves of wild-type (Col) and efr plants infiltrated with 100 µM BTH were collected 24 h later and stained with aniline blue. Eight random fields were photographed from 2 leaves of 3 independent plants. Callose was quantified as in Figure 2. Bars show standard error of 3 independent experiments analyzed together (n > 24). Asterisk indicates significant difference from Col, p < 0.05, student's t-test.

ACD6, SA and PRRs comprise a regulatory network. ACD6 positively regulates basal levels of FLS2 and CERK1. In response to a stimulus such as flg22, chitin or BTH, plants lacking ACD6 show less callose. In acd6-1 and BTH-treated wild type, FLS2 and CERK1 proteins levels are elevated, resulting in increased responsiveness to flg22 or chitin. FLS2 and CERK1 are also partially required for BTH-induced callose deposition. These findings suggest (1) the positive feedback loop of ACD6 and SA regulates PAMP responsiveness by regulating receptors and (2) PRRs have multiple roles in signaling.⁴ Data reported in this study suggest that EFR is part of the ACD6/SA regulatory network and is needed for PAMP-independent signaling, similar to FLS2 and CERK1. A model summarizing our findings is shown in Figure 4.

Figure 4.

Regulation of receptors and co-receptor by ACD6 and positive feedback loop of SA signaling. Receptor/co-receptor levels and ligand responsiveness in plants that are wild type or express ACD6-HA (similar phenotypes as wild-type) (A), loss of ACD6 function mutant, acd6-2 plant (B) and gain of ACD6 function allele, acd6-1 or ACD6-1-HA plant (HA tagged ACD6-1-expressing plant, similar phenotypes as acd6-1) (C). ACD6 and ACD6-1 proteins form complexes with FLS2, CERK1 and BAK1 in ACD6-HA and ACD6-1-HA plants.^2,4 (Several proteins, including FLS2 form complexes with both ACD6-HA and ACD6-1-HA,² suggesting that the components of complexes from ACD6-HA and ACD6-1-HA might be very similar.) Lower levels of FLS2 or CERK1 cause plants to be less responsive to flg22- or chitin-treatment in acd6-2 plants, respectively.⁴ Additionally, ACD6, FLS2, CERK1 and BAK1 are all partially required for BTH-induced callose deposition.⁴ In this report, acd6-2 plants show less callose deposition in response to elf18 treatment (in Fig. 2) and a loss of EFR mutant (efr) also shows less callose deposition in response to BTH treatment (in Fig. 3). Thus, EFR's role and regulation appears similar to that of other receptor/co-receptor(s) (FLS2, BAK1, CERK1) with respect to the functional relationship with ACD6. In addition to ACD6 affecting receptor/co-receptor levels in the basal state, receptor/co-receptor proteins accumulate to high levels in plants with high SA signaling (such as in acd6-1 and/or BTH-treated wild type). This results in potentiation of responses to specific ligands.^2,4 From Figure 1, EFR is part of the ACD6/SA regulatory network, as was shown for FLS2, BAK1 and CERK1.

Plants lacking ACD6 are hypersusceptible to Pseudomonas syringae and P. syringae lacking a type III secretion system and less responsive to multiple PAMPs.^3,4 In contrast, a dominant gain of function allele of acd6-1 or natural alleles of ACD6 from different Arabidopsis accessions show higher resistance against to Pseudomonas syringae, Hyaloperonospora parasitica, Golovinomyces cichoracearum, and Hyaloperonospora arabidopsidis.^1,4,6,12 The changes in susceptibility of loss and gain function mutants/allelic variants, respectively, might be caused by changes in the accumulation of multiple PRRs.

Responsiveness to microbe-derived molecules and elicitors (e.g. flg22, chitin, phytophthora elicitor(s) and elf18) are potentiated by SA, BTH and β-aminobutyric acid (BABA).^4,15-19 Might BABA cause similar regulatory effects as SA/BTH? BABA application induces resistance to several types of pathogens in a variety of plants,^20-22 similar to SA and BTH.^23-27 The LECTIN RECEPTOR KINASE-VI.2 (LecRK-VI.2) is required for BABA-primed resistance to bacteria and responsiveness to flg22.¹⁸ In addition to BABA priming, LecRK-VI.2 is required for flg22- or elf18-induced callose deposition.^18,19 It's not known if BABA affects PRR levels or trafficking to the plasma membrane as SA signaling does,^2,4 but we speculate that such regulation may explain some effects of BABA. Additionally, it seems likely that LecRK-VI.2 or other receptor-like kinases also might be involved in SA signaling, something that will be interesting to investigate in the future.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank the Ohio State Stock Center, Dr. Yusuke Saijo for providing EFR antibody, Dr. Robert Dietrich for providing BTH and Drs. Delphine Chinchilla and Joanna Jelenska for useful discussions.

Funding

This work was supported by grants to JTG from the NIH (R01 GM54292) and NSF (IOS 0822393). CT was supported by JSPS postdoctoral fellowships for Research Abroad.