Introduction
The plant immune system is multi-layered and complex. It traditionally comprises three modules; microbe associated molecular pattern (MAMP)-triggered immunity (MTI), effector-triggered immunity (ETI) and systemic acquired resistance (SAR) (Jones and Dangl, 2006; Shineet al. , 2019). The initial layer of defence, MTI, provides broad-spectrum defence against a diverse range of pathogens and has recently been shown to be involved in potentiating ETI responses, which can in turn reinforce MTI (Lu and Tsuda, 2021; Ngou et al. , 2021; Nguyen et al. , 2021; Yuan et al. , 2021). Classical pathogen cell surface receptors comprise transmembrane receptor-like kinases (RLKs) or receptor-like proteins (RLPs) including FLAGELLIN SENSING 2 (FLS2), EF-Tu RECEPTOR (EFR) and CHITIN ELICITOR RECEPTOR KINASE 1 (CERK1-2) which, respectively, detect flagellin and elongation factor thermo-unstable (EF-Tu) from bacterial pathogens and chitin from fungi (Yu et al. , 2017). However, an increasing number of MAMPs associated with a diverse range of pathogens have been identified (Noman, Aqeel and Lou, 2019). In addition, cell surface receptors can detect plant derived damage associated molecular patterns (DAMPs) found within extracellular spaces. Amongst DAMP receptors are the well-characterised RLKs, PEP RECEPTOR 1 (PEPR1) and PEPR2 which detect plant elicitor peptides, Peps. PEPR1, recognises Peps1-6 while PEPR2 recognises only Pep1 and Pep2 (Yamaguchi, Pearce and Ryan, 2006; Yamaguchi et al. , 2010). These Peps are cleaved from the C-terminus of plant PROPEPs during cell damage and the transcripts of PROPEP1-3 are induced by defence-related hormones methyl salicylate (MeSA) and methyl jasmonate (MeJA) (Huffaker, Pearce and Ryan, 2006; Yamaguchi et al. , 2010).
The pattern recognition receptors (PRRs), FLS2, EFR and PEPR1/2, are cell membrane localised and contain extracellular leucine rich repeat (LRR) surfaces where their ligands bind. Upon peptide detection by PRRs, co-receptors are recruited and bind to PRRs (and in some cases the ligand). The well characterised co-receptor Brassinosteroid Insensitive 1 (BRI1)-associated receptor kinase 1 (BAK1) belongs to the somatic embryogenesis receptor-like kinase family (SERK) which contains five members, one of which, SERK4/BKK1 (BAK1-LIKE 1), has high sequence similarity to BAK1 and has functional redundancy (He et al. , 2007). While BAK1 was first identified as a co-receptor for the Brassinosteroid receptor BRI1, involved in cell growth and division, it has become widely known for its role in plant immunity as plants containing the reduced function bak1-5 allele have impaired FLS2, EFR and PEPR receptor function (Roux et al. , 2011; Schwessingeret al. , 2011). In contrast, bkk1-1 still exhibits a reactive oxygen species (ROS) burst and MAP Kinase (MPK3, MPK4 and MPK6) activation, that is comparable to wild type plants, when treated with flg22 or elf18. However, the bak1-5/bkk1-1 plants show minimal ROS and no MAPK (mitogen-activated protein kinase) activation in response to these PAMPs (Zipfel et al. , 2006; Roux et al. , 2011).
MTI triggers rapid calcium signalling, ROS and MAPK signalling cascades all of which involve plasma membrane to nuclear signalling (Noman, Aqeel and Lou, 2019). Microbes successful in colonisation secrete effectors to inter- or intracellular locations, which can dampen MTI signalling. Examples of such effector triggered suppression (ETS) include the AvrPto effector from Pseudomonas syringae which interacts with the PRRs FLS2 and EFR to dampen MTI in Arabidopsis thaliana (Xianget al. , 2008) and AvrE from P. syringae and the maize pathogen Pantoea stewartii subsp. Stewartia which targets protein phosphatase 2A (PP2A) complexes in order to dampen MTI (Jinet al. , 2016).
Effectors collectively target an array of plant immune signalling components, many of which still remain elusive. Some effectors are directly or indirectly recognised by cytoplasmic receptors, most often belonging to the nucleotide-binding leucine-rich repeat receptors (NLRs) class, activating a second immune response, ETI (Jones and Dangl, 2006). There are three major classes of NLRs, the first two classically defined by their N-terminal; Toll-like, Interleukin-1 receptor domain TIR-NLRs (TNLs), coiled-coil domain CC-NLRs (CNLs). More recently the Resistance to Powdery Mildew 8 (RPW8) CC-NLR class (RNLs) (Jones, Vance and Dangl, 2016; Zhong and Cheng, 2016) have been described which act as “helper” NLRs for TNL and CNL “sensor” NLRs (Lu and Tsuda, 2021; Nguyenet al. , 2021; Maruta et al. , 2022). Interaction of an effector and NLR is usually associated with the macroscopic development of the hypersensitive response which restricts pathogen growth.
Classically, MTI research has centred around signal transduction pathways originating from the plasma membrane and activating nuclear transcription however, it is becoming increasingly recognised that chloroplasts are a key hub of immune signalling (Kachroo, Burch-Smith and Grant, 2021; Littlejohn et al. , 2021). Chloroplasts play a central role in integrating environmental signals and maintaining cellular homeostasis via retrograde signalling (de Souza, Wang and Dehesh, 2017; Breeze and Mullineaux, 2022). Relevant to host immune signalling, chloroplasts are also the site of chloroplastic ROS (cROS) generation and synthesis of defence hormone precursors, jasmonic acid (JA), salicylic acid (SA) and abscisic acid (ABA) (Littlejohn et al. , 2021). A key early MTI response is the rapid ROS generation, an apoplastic localised respiratory burst, primarily generated by RBOHD, a member of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase homologue (RBOH) family (Miller et al. , 2009). Activating MTI using an effector secretion deficient strain of P. syringaepv. tomato strain DC3000 (DC3000hrpA ) also rapidly generates cROS production in A. thaliana , which is attenuated in the virulent DC3000 strain, shortly after effector delivery (de Torres Zabala et al. , 2015).
Concomitant with differences in cROS production during infection between the P. syringae strains DC3000 and DC3000hrpA , global transcriptome profiling of A. thaliana revealed significant alterations of nuclear encoded chloroplast genes (NECG s). Remarkably, NECGs, represent ~10% of all differentially upregulated genes and ~30% of those significantly down regulated (de Torres Zabala et al. , 2015) during early MTI responses despite NECGs collectively account for only ~14% of the transcriptome. Superimposed on this, effector delivery (2-3 hour post infection; hpi) caused transcriptional reprogramming of NECGs , suggesting ETS also targets NECGexpression (de Torres Zabala et al. , 2015). These molecular signatures are reflected by physiological changes between DC3000 and DC3000hrpA challenge as evidenced by quantifying net photosynthetic CO2 assimilation (Asat) and chlorophyll fluorescence imaging parameters associated with electron transport during photosynthesis. DC3000 but not DC3000hrpAchallenge induced a decrease in CO2 assimilation, maximum dark-adapted quantum efficiency (Fv/Fm ), maximum operating efficiency of photosystem II (PSII) (Fv′/Fm ) and the efficiency with which light absorbed by PSII is used for quinone acceptor (QA) reduction and linear electron transport (Fq′/Fm ) (de Torres Zabalaet al. , 2015). In addition, DC3000 infection elicited an increase in Non-Photochemical Quenching (NPQ) and PSII redox state (qL; (Fq′/Fv′)/(Fo′/F′ )) compared to DC3000hrpA (de Torres Zabala et al. , 2015). qL estimates the percentage of open PSII centres and the oxidation state of the primary PSII QA (Baker, 2008). An increase in qL suggests a decrease in electron transport from PSII. Thus, virulent pathogens can radically alter chloroplast physiological functions as part of their virulence strategy.
De novo induction of the plant hormone ABA during DC3000 infection contributes to ETS (de Torres-Zabala et al. , 2007) and was also recently shown to play a significant role in modulating chloroplast function. DC3000 induced suppression ofFv/Fm was accelerated by co-infiltration of 10 µM ABA, effectively phenocopying DC3000 challenge of the Arabidopsis ABA hypersensitive protein phosphatase 2C (PP2C) abi1/abi2/hab1 triple mutant. By contrast, the ABA deficient Arabidopsis aldehyde oxidase 3 (aao3 ) mutant restricted DC3000 suppression ofFv/Fm (de Torres Zabala et al. , 2015). Collectively these data show that the chloroplast is targeted early in pathogen infection and prior to bacterial multiplication, one of the earliest events being suppression of cROS.
This study focussed on how well characterised PRRs and co-receptors modulated chloroplast physiology, including accessing whether diverse signalling pathways converged to similarly modulate chloroplast function. Here we comprehensively examine chlorophyll fluorescence dynamics and the impact on attenuating chloroplast cROS. We show that pre-treatment of receptor mutant plants with MAMP and DAMP peptides generally offer protection against effector modulation of chlorophyll fluorescence but surprisingly, fls2 plants pre-treated with chitin fail to provide such protection. The double mutant of the MTI co-receptors bak1-5/bkk1-1 exhibits a remarkable decrease inFv/Fm compared to control plants during infection, underlining the importance of MTI mediated signalling in underpinning chloroplast immunity. Expanding these findings to better understand the role of ABA and abiotic stress in chloroplast immunity we found that high light overrides the protection offered by MAMPs on wild-type plants.
Materials and Methods
Arabidopsis growth conditions. Arabidopsis thalianaseeds were sown in a compost mix comprising Levingston F2 compost + sand (LEV206):vermiculite (medium grade) mixed in a 6:1 ratio. Plants were grown in a controlled environment growth chamber under a 10 h day (21 °C; 120 µmol m-2s-1) and 14 h night (21 °C) with relative humidity of 65% for 5–6 weeks prior to use.
Arabidopsis peptide treatment. Pre-treatment of plants was conducted 16 h prior to bacterial challenge. Co-infiltration experiments were conducted by mixing the peptide or hormone of interest with the bacterial culture to attain the required final concentration and OD600 prior to infiltration. Concentrations of peptides or hormones were as follows; 1 µM of flg22, elf18, Pep1, Pep2 and Pep3, 100 µg/ml of Chitin (Sigma - C9752) and 10 or 100 µM ABA. H2O was used as mock for pre-treatments.
Bacterial growth, maintenance and inoculation.Pseudomonas syringaestrains were grown on solid Kings B media containing appropriate antibiotics as described (Truman, de Zabala and Grant, 2006). For inoculation, overnight cultures were grown with shaking (200 rpm) at 28 °C. Cells were harvested (2,500 g × 7 min), washed and re-suspended in 10 mM MgCl2. Cell density was adjusted to OD600 0.15 (∼0.75 × 108 colony forming units (cfu) ml-1) for fluorescence imaging and confocal microscopy or OD600 0.0002 for growth assays. All growth assays and ROS imaging experiments were performed at least three times. All fluorescence imaging experiments were performed at least four times.
Chlorophyll fluorescence imaging. Photosystem II chlorophyll fluorescence imaging of Arabidopsis rosettes was performed with a CF Imager (Technologica Ltd, Colchester, UK). Normal light cycle;plants were placed in the chamber for 40 min post-inoculation and then dark adapted for 20 min. This was followed by a saturating light pulse (6,349 µmol m­-2s-1 for 0.8 s) to obtain maximum dark-adapted fluorescence (Fm ). Actinic light (120 µmol m-2s-1 – the same as plant growth light intensity) was then applied for 15 min, followed by a saturating pulse to obtain maximum light adapted fluorescence (Fm′ ). The plants remained in actinic light for a further 24 min, then returning to a dark period of 20 min. This cycle (59 min duration) was repeated 23 times. High light cycle; plants were placed in the chamber for 40 min post-inoculation and then dark adapted for 20 min. This was followed by a saturating light pulse (6,349 µmol m­-2s-1 for 0.8 s) to obtain maximum dark-adapted fluorescence (Fm ). High light (650 µmol m-2s-1) was then applied for 15 min, followed by 3 saturating light pulses 5 minutes apart to obtain maximum light adapted fluorescence (Fm′ ). The plants remained in high light for a further 150 min then returned to a 20 min dark phase. This cycle (200 min duration) was repeated 8 times.Fm , Fm′ andFo (minimal fluorescence with fully oxidized PSII centres) were used to calculate chlorophyll fluorescence parameters related to photosystem II: Fv/Fm(maximum dark-adapted quantum efficiency) and non-photochemical quenching (NPQ). These values were calculated as described by (Baker, 2008).
Bacterial growth measurements. Three leaves per plants (6 plants total) were syringe infiltrated with bacteria, OD600 0.0002, and placed either under high light (450 µmol m-2s-1 or 600 µmol m-2s-1) or normal light (120 µmol m-2s-1) for 4 days. Three independent leaf discs per plant were excised and homogenised using a Tissue Lyser (Qiagen). Serial dilutions were spotted on Kings B media and colonies were counted 24 hpi.
Confocal microscopy. Col-0 plants were pre-treated with either water or peptide 16 h prior to bacterial challenge, then 3.5 hpi leaves were detached and floated, adaxial surface upwards, in a solution of 10 mM MgCl2 containing 10 μM 2′7′-dichlorodihydrofluorescein diacetate (H2DCF-DA; Enzo) for 40 min, then washed for 20 min in 10 mM MgCl2 before imaging. Samples were mounted in perfluorodecalin (Littlejohn et al. , 2010) and images were captured on a Zeiss 880 using a 40× oil immersion lens. Argon laser excitation at 488 nm and an emission window of 512–527 nm was used to capture the dichlorofluorescein (DCF) signal. Chloroplast fluorescence was measured at 659–679 nm.