https://orcid.org/0000-0002-5923-5571
Abstract
Plants possess a well-balanced immune system that is required for defense against pathogen infections. In autoimmune mutants or necrotic crosses, an intrinsic temperature-dependent imbalance leads to constitutive immune activation, resulting in severe damage or even death of plants. Recently, cell wall deposition was described as one of the symptoms following induction of the autoimmune phenotype in Arabidopsis saul1-1 mutants. However, the regulation and function of this deposition remained unclear. Here, we show that cell wall deposits, containing lignin and callose, were a common autoimmune feature and occurred in proportion to the severity of the autoimmune phenotype at reduced ambient temperatures. When plants were exposed to reduced temperature for periods insufficient to induce an autoimmune phenotype, the cell wall deposits were not present. After low temperature intervals, sufficient to induce autoimmune responses, cell wall deposits correlated with a point of no return in saul1-1 autoimmunity. Although cell wall deposition was largely abolished in saul1-1 pmr4-1 double mutants lacking SAUL1 and the callose synthase gene GSL5/PMR4, their phenotype remained unchanged compared with that of the saul1-1 single mutant. Our data showed that cell wall deposition generally occurs in autoimmunity, but appears not to be the cause of autoimmune phenotypes.
Introduction
Plants use various chemical and physical defense mechanisms to protect themselves from pest infestation and diseases. They have evolved mechanisms for the recognition of pathogens to initiate these defenses. At the plasma membrane, pattern recognition receptors perceive pathogen-associated molecular patterns (PAMPs) such as bacterial flagellin and fungal chitin to induce PAMP-triggered immunity (PTI) (Jones and Dangl, 2006; Boller and Felix, 2009; Monaghan and Zipfel, 2012). The triggered defense program involves (i) ion fluxes, particularly of H+, Ca2+, and K+, (ii) an oxidative burst leading to accumulation of H2O2 and other reactive oxygen species, (iii) activation of mitogen-activated protein kinase pathways and calcium-dependent protein kinases, (iv) endocytosis of a number of immune regulators including the FLAGELLIN-SENSING 2 receptor, (v) activation of downstream defense genes, (vi) growth inhibition, likely through inactivation of auxin signaling, and (vii) the deposition of callose at the cell wall. In concert, these immune responses provide efficient pathogen defense. However, several pathogens inhibit PTI by delivering effectors into the host cells (Galán et al., 2014; Macho and Zipfel, 2015). Effector function is monitored by nucleotide-binding leucine-rich repeat receptor proteins (NLRs, also called Nod-like receptors), which initiate effector-triggered immunity (ETI) to combat the pathogen (Li et al., 2015). Unlike PTI responses, ETI results in a cell death program called the hypersensitive response to prevent pathogen establishment and spread.
Regulation of NLR homeostasis is critical for plants. Insufficient NLR activity can lead to increased susceptibility to certain pathogens, while increased accumulation of NLRs can lead to autoimmunity, as in the case of Suppressor of npr1, constitutive 1 (SNC1) (Cheng et al., 2011). Autoimmune mutants such as snc1, chilling sensitive 1 (chs1), and bonzai 1 (bon1) show reduced growth and/or hypersensitive response-induced cell death visible as lesioning in rosette leaves, reactive oxygen species production, expression of defense genes, and induced resistance to pathogens. These phenotypes are dependent on temperature and relative humidity as well as downstream ENHANCED DISEASE SUSCEPTIBILITY 1 (EDS1) and PHYTOALEXIN DEFICIENT 4 (PAD4) signaling (Hua et al., 2001; Jambunathan et al., 2001; Li et al., 2001; Wang et al., 2013; Zbierzak et al., 2013). Recently, we established a novel model system for autoimmunity. Arabidopsis saul1 mutant plants lacking the expression of the SENESCENCE-ASSOCIATED UBIQUITIN LIGASE 1 (SAUL1) gene show all the typical features of autoimmunity (Raab et al., 2009; Vogelmann et al., 2012; Disch et al., 2016; Tong et al., 2017). In addition to the simultaneous yellowing/lesioning of all aboveground organs, a prominent feature of saul1 autoimmunity is cell wall deposition (Disch et al., 2016).
The wall of plant cells is a monitoring system for environmental changes including pathogen infestation and has been linked to biotic stress signaling (Malinovsky et al., 2014; Hamann, 2015; Bacete et al., 2018; Vaahtera et al., 2019; Kozieł et al., 2021). Upon stress, damage-associated molecular patterns may be released from the cell wall and recognized by pattern recognition receptors to induce immune signaling (Ranf et al., 2015; de Azevedo Souza et al., 2017; Stegmann et al., 2017; Bacete et al., 2018; Mélida et al., 2018). Responses to biotic stress involve the coordinated regulation of cell wall integrity and PTI (Escudero et al., 2017; Van der Does et al., 2017; Engelsdorf et al., 2018). Cell wall fortification provides an antimicrobial physical barrier, but modifications to cell wall components that are sometimes accompanied by remodeling of the cytoskeleton in the host cell also affect immune regulation (Opalski et al., 2005; Jin and Mackey, 2017; Bacete et al., 2018; Rui and Dinneny, 2020; Gigli-Bisceglia et al., 2020). Two examples of cell wall changes are the induced biosynthesis and deposition of the β-1,3-glucan polysaccharide callose and the phenylpropanoid-derived heteropolymer lignin. An increase in callose content was observed in Arabidopsis seedlings treated with the bacterial peptide flg22, suggesting that PTI can lead to callose deposition (Gómez-Gómez et al., 1999). This callose formation could be suppressed by the Pseudomonas syringae effector HopA1, which inhibits mitogen-activated protein kinase signaling (Zhang et al., 2007). Another well-studied example is the role of callose in the defense response to fungal infections. Powdery mildew infection induces early callose deposition, mediated by glucan synthases, in the form of cell wall thickenings called papillae that prevent fungal penetration at the site of attack (Jones and Dangl, 2006; Ellinger et al., 2013). Cell wall thickenings at the respective infection sites may contain callose as well as significant amounts of lignin (Underwood, 2012; Malinovsky et al., 2014).
The continuous induction of defense responses in autoimmune mutants leads to increased pathogen resistance, but is established often at the expense of plant growth and fitness (Heil and Baldwin, 2002; Bomblies and Weigel, 2007; Todesco et al., 2010). This is an important issue in plant breeding aimed at introducing resistance to plant pathogens into crop plants. It is therefore important to study the molecular mechanisms of autoimmunity to find ways to uncouple increased resistance from reduced growth in the future. Recently, we showed that cell wall deposition represent an early PAD4-dependent event in the autoimmunity of saul1-1 plants and that these cell wall deposits (CEWDs) contain callose and lignin (Disch et al., 2016). Here, we investigate whether such CEWDs were common in autoimmunity and therefore examined cell wall changes in other autoimmune mutants. Indeed, we demonstrated that CEWDs were a common feature of autoimmunity. We investigated the putative function of CEWDs in autoimmunity by demonstrating the correlation between the degree of temperature-dependent CEWDs and the observed morphological changes in saul1-1 plants. Cell wall deposition was largely abolished in saul1-1 pmr4-1 double mutants, but this reduction did not affect the autoimmune phenotype.
Materials and methods
Plant material and growth conditions
Arabidopsis wild type (WT) and mutant plants were grown on soil and were stratified in the dark for 48 h at 4 °C before being transferred to permissive growth temperatures of 25 or 27 °C for 12 or 16 d prior to a shift to lower temperature of 12, 18 or 20 °C for the indicated times. In recovery experiments, the plants were transferred back to 25 °C after the low temperature treatment. Generally, plants were grown in growth cabinets (models AR-36L3 or AR-36L2, Percival, CLF PlantClimatics GmbH, Wertingen, Germany) with long day conditions (16 h light with a fluence rate of 100 µmol m−2 s−1 and 8 h dark) and a constant relative humidity in the range of 81–89%. The temperature and relative humidity were monitored using a data logger (OM-EL-USB-2-LCD-PLUS, OMEGA Engineering Inc., Deckenpfronn, Germany) and the photon flux density was measured using a light meter (LI-250A with Quantum sensor, LI-COR Biosciences, Bad Homburg, Germany). saul1-1 pmr4-1 double homozygous mutants were generated by crossing homozygous single mutants, and double homozygous F2 and F3 generations were identified by PCR, using the SAUL1 specific primers 5′-TGAGGCCAATCAAATGATTTC-3′ and 5′-TTTCCCCATTCATGAGTGAAG-3′ in combination with the T-DNA insertion (SALK) specific primer Lba 5′-TTCATAACCAA TCTCGATACAC-3′ and the PMR4 specific primers 5′-CAAGGACGGCATTCATAGGT-3′ and 5′-CCGTCTCGCCTC TAGATTCA-3′ to amplify a fragment of PMR4 followed by a restriction digest of the fragment using NheI, indicating the point mutation in PMR4.
Transmission electron microscopy
In all experiments, one of the first two true leaves of plants was taken for analysis by transmission electron microscopy (TEM). For each condition, we analysed three to five leaves from different plants and looked at several cells in the respective cross sections. Leaves were fixed with 2% glutaraldehyde and 2% paraformaldehyde in 75 mM cacodylate buffer (pH 7.0) via vacuum infiltration, washed three times with 75 mM cacodylate buffer (pH 7.0), and post-fixed with 1% osmium tetroxide in 75 mM cacodylate buffer (pH 7.0) overnight at 4 °C. After three washing steps with 75 mM cacodylate buffer (pH 7.0), samples were dehydrated in a gradient of acetone concentrations (30%, 50%, 70%, 90%, 100% on ice and an additional two times 100% at room temperature) for 10 min each. Dehydrated leaves were infiltrated with a 3:1 mixture of 100% Spurr embedding solution for 1 h, with a 1:1 mixture for 1 h and a 1:3 mixture overnight. After two infiltrations with fresh 100% Spurr solution, once for 6 h and once overnight, samples were baked in silicone molds for 16 h at 70 °C. Ultrathin sectioning of samples (70–80 nm) was achieved using an ultramicrotome (Ultracut E, Leica Reichert-Jung, Wetzlar, Germany) equipped with a diamond knife. Sections were collected on copper grids (150 mesh) and subsequently stained for 10 min with 2% uranyl acetate followed by 2% lead citrate. A LEO 906 E transmission electron microscope (Leo, Oberkochen, Germany) equipped with a MultiScan CCD Camera (Model 794) was used for sample analysis and Microscopy Suite software version 2.0.2 (Gatan, Munich, Germany) was used for visualization and analysis of the image data.
Scanning UV microspectrophotometry and confocal laser scanning microscopy
For the detection of lignin and callose in tissue sections of plant leaves we followed the protocols described previously (Disch et al., 2016). In short, samples for scanning UV microspectrophotometry (UMSP) were prepared in the same way as for TEM except for post-fixation with osmium tetroxide. Thin sections were immersed in non-UV absorbing glycerol, and topochemical analyses were based on absorbance at 278 nm measured with a geometrical resolution of 0.25×0.25 µm by using a Zeiss UMSP 80, with analysis using the APAMOS scan program (Carl Zeiss AG). For callose detection, leaves were fixed with 2% paraformaldehyde in 100 mM PIPES, 5–10 mM EGTA, 5 mM MgSO4, pH 6.8. After washing in the same buffer without paraformaldehyde, the samples were dehydrated by ethanol treatment. Gradual infiltration with LR white resin (medium grade acrylic resin; London Resin Company Ltd, Reading, UK) was followed by filling into gelatin capsules and curation at 40–50 °C for 36 h. Semithin sections were stained with 0.01% aniline blue in 150 mM K2HPO4 (pH 9.0). We used the Leica TCS SP8 confocal platform (Leica Microsystems, Wetzlar, Germany). Emission of fluorescence was detected at 461 to 492 nm after excitation at a wavelength of 405 nm. Images were processed by using the Leica Application Suite X (Leica Microsystems).
Gene expression analysis by qPCR
For gene expression, we analysed four biological replicates from above-ground tissues of WT and mutant plants. For RNA isolation and cDNA synthesis we used the innuPREP plant RNA kit (Analytik Jena AG, Jena, Germany) and the QuantiTect reverse transcription kit (Qiagen, Hilden, Germany), respectively. In qPCR experiments samples were treated with the QuantiFast SYBR green PCR master mix (Qiagen) to determine transcription levels with a Rotor Gene Q (Qiagen). Expression levels were standardized to transcription levels of UBI and EIF1. Primers for PCR amplification of PMR4 were 5′-CTTTGCCGGGTTTAACTGCAC-3′ and 5′-AATCCAACATCCCG TCCCTTC-3′.
High-performance liquid chromatography
About 200 mg of Arabidopsis whole plant (above-ground) material was frozen with liquid nitrogen and ground using a TissueLyser (Qiagen, manufactured by Retsch, Hilden, Germany) at 30 Hz for 5 min. Ninety-six percent ethanol was added so that the ground material was covered and incubated at 99 °C for 15 min in a water bath. The mixture was cooled to room temperature and centrifuged at 17 000 g for 10 min. The supernatant was removed and the pellet was washed with 70% ethanol by shaking and mixing thoroughly followed by centrifugation at 17 000 g for 10 min. The washing was repeated seven times. The final pellet was vacuum-dried overnight.
The pellet was resuspended in 200 μl sodium acetate buffer (0.2 M, pH 4.5); 20 μl of α-amylase mix (α-amylase from Bacillus amyloliquefaciens, Sigma-Aldrich, diluted 1:10 in sodium acetate buffer) was added and the mixture was incubated for 1 h at 85 °C and 800 rpm (Thermomixer Compact, Eppendorf). The mixture was then cooled to room temperature and 200 μl of sodium acetate buffer and 20 μl of amyloglucosidase mix (amyloglucosidase from Aspergillus niger, Sigma-Aldrich, diluted 1:10 in sodium acetate buffer) were added and incubated for 2 h at 52 °C and 600 rpm. The enzymes were deactivated by thoroughly mixing with 800 μl of 70% ethanol and inactivation for 10 min at room temperature. The samples were centrifuged for 20 min at 17 000 g. The supernatant was removed and the pellet was washed with 1 ml of 70% ethanol, mixed thoroughly and centrifuged for 5 min at 17 000 g. The washing step was repeated eight times. The pellet was vacuum dried overnight. Of the dried pellet (Alcohol Insoluble Residue, AIR) 5 mg was transferred to a screw cap reaction tube along with two steel balls. The AIR pellet was ground at 30 Hz for 10 min using the TissueLyser. The ground material was then resuspended in 850 μl dd H2O and ground again as described above. The solution was transferred to a new reaction tube and 150 μl of trifluoroacetic acid was added. The mixture was inverted and incubated for 3 h at 100 °C (Thermomixer Compact). After the mixture was cooled to room temperature, it was mixed thoroughly and vacuum dried for 2–3 d. The resulting pellet was resuspended in 1 ml dd H2O, mixed thoroughly and centrifuged at 4 °C and 17 000 g for 10 min. The supernatant was diluted 1:10 and used for HPLC analysis. The hemicellulose composition of Arabidopsis cell walls was determined by HPLC. The analysis was performed using a Dionex CarboPac PA20 column for monosaccharides (Thermo Fisher Scientific, Dreieich, Germany) and a Dionex DC ICS 5000 with an EC detector (Thermo Fisher Scientific). As eluent, a gradient of 2.8–45 mM NaOH was used. The calibration was done prior to each run by measuring a dilution series of standard solutions containing the respective monosaccharides.
Results
A point of no return in temperature-dependent saul1-1 autoimmunity correlates with changes in the structure of cell walls and chloroplasts
CEWDs have recently been shown to be a feature of temperature-dependent autoimmunity in plants, as they were observed as part of the autoimmune response in Arabidopsis saul1-1 mutants (Disch et al., 2016). CEWDs are generally thought to play a role in plant defense (Gómez-Gómez et al., 1999; Otulak-Kozieł et al., 2018). We therefore tested whether we can define a point of no return in saul1-1 autoimmunity that prevents recovery and correlates with the occurrence of CEWDs. Autoimmune responses in saul1-1 can be induced by a decrease of the ambient temperature (Disch et al., 2016). To determine a point of no return, we therefore challenged saul1-1 mutant plants with lower temperature (18 °C) for different periods of time followed by 6 d of recovery at 25 °C. Whereas 24 h at lower temperature did not affect the morphology of developing saul1-1 mutant plants, the identical treatment for 48 h or 72 h strongly reduced growth in saul1-1 (Fig. 1A; Supplementary Fig. S1A). The observed morphological changes correlated with the occurrence of CEWDs, analysed by TEM. Exposure to reduced temperature for 24 h was not sufficient to induce the formation of CEWDs, whereas the degree of CEWDs increased with prolonged exposure to lower temperature for 48 h or more (Fig. 1B). To compare the extent of cell wall deposition, we determined the maximum peak size of the CEWDs. As no CEWDs were formed after 24 h of treatment (plus 6 h of recovery), the mean peak size was zero. Following low temperature treatment, the mean peak size showed significant increases to 0.40±0.19 µm (n=36) and 0.72±0.43 µm (n=10) at 48 h and 72 h of treatment (plus 6 h of recovery), respectively (Fig. 1C). The extended exposure to low temperatures also resulted in changes in the structure of chloroplasts. As described previously (Disch et al., 2016), the shape of chloroplasts became more rounded, and plastoglobules were larger (Fig. 1B). The mean diameter of plastoglobules changed from 0.08±0.01 µm (n=147) at 24 h, to 0.12±0.04 µm (n=263) at 48 h, and to 0.13±0.03 µm (n=139) at 72 h of low temperature treatment followed by 6 d of recovery (Fig. 1D). Our data thus indicated a strong correlation between morphological and structural changes.
saul1-1 autoimmunity shows a point of no return correlating with structural changes. Plants were grown at 25 °C for 12 d and shifted to 18 °C for 24, 48, and 72 h before a recovery phase for 6 d at 25 °C was initiated. (A) Autoimmune phenotypes of WT and saul1-1 mutants at t=24, 48, and 72 h plus 6 d of recovery. (B) Structural changes of chloroplasts and cell wall of WT and saul1-1 mutants at t=24, 48, and 72 h plus 6 d of recovery. (C) Violin plot of a Kruskal–Wallis test for comparing CEWD peak size after differential exposure to low temperature as in (A). (D) Violin plot of a Kruskal–Wallis test for comparing plastoglobule diameter after differential exposure to low temperature as in (A). Asterisks indicate significance levels (*P<0.05, ****P<0.0001). Scale bars represent 1 µm. Arrows indicate the extent of CEWDs.
Though growth of saul1-1 plants was drastically reduced up to 72 h at 18°C, the plants were not dead after the recovery phase of 6 d at 25 °C (Fig. 1A). To identify the duration of low temperature challenge that may lead to the death of saul1-1 plants, we repeated the growth experiment with 1 d increments up to 7 d of challenge at 18 °C, followed by long-term recovery at 25 °C for up to 3 weeks (Fig. 2). Whereas after 1 d of low temperature challenge the saul1-1 plants resembled WT in growth and flowering, 2 d of challenge resulted in a significant reduction of growth. Challenging saul1-1 for 3 d with lower temperature led to a large growth deficit and a delay in flowering. After low temperature exposure for more than 3 d, saul1-1 plants were unable to recover. We identified a point of no return for saul1-1 plants challenged with lower temperature for between 3 and 4 d, right after we observed the formation of large CEWDs in saul1-1 cells (Fig. 1).
saul1-1 autoimmune phenotypes can be recovered after up to 3 d at low temperature. WT (on the left side) and saul1-1 plants (on the right side) were grown at 25 °C for 12 d and shifted to 18 °C for 0–7 d, before a recovery phase at 25 °C for 7, 14, or 21 d was initiated.
Cell wall deposits and changes in the structure of chloroplasts are induced in autoimmune mutants other than saul1-1
The strong correlation between changes in morphology and structural changes in cell wall and chloroplasts prompted us to test whether these CEWDs and changes in chloroplast structure are specific to saul1-1 mutants, or whether they are a general feature of autoimmunity in Arabidopsis. Mutant saul1-1, snc1, cpn1-1, and WT plants were grown for 12 d at a temperature of 27 °C, before autoimmune responses were induced by a temperature shift to 20 °C. Since the induction of autoimmunity in chs1-2 requires lower temperatures (Wang et al., 2013; Zbierzak et al., 2013), we induced autoimmune responses at 12 °C after growing WT, saul1-1, and chs1-2 plants for 16 d at 25 °C. Before the induction of autoimmunity, all mutants were indistinguishable from WT. In both growth experiments, saul1-1 mutants showed a strongly reduced growth and yellowing/lesioning of leaves following the induction of autoimmunity (Fig. 3A, B). Autoimmunity was comparably strong in chs1-2 mutant plants (Fig. 3B). In contrast, growth reduction and yellowing/lesioning were visible, but less pronounced in snc1 and cpn1-1 mutants. In addition, these mutants showed curling of leaves (Fig. 3A). Growth reduction for autoimmune mutants was quantified by measuring the rosette diameter (Supplementary Fig. S1B). In all mutants, autoimmunity was also confirmed on the molecular level by analysing the regulation of the immune marker genes PAD4, EDS16, and PR1. The induction of all three genes was detected for all mutants, though the induction was more pronounced in saul1-1 and chs1-2 (Supplementary Fig. S2). The bak1-1 mutant was included in the analyses, because the BAK1 receptor kinase has been shown to play a role in PTI signaling (Kemmerling et al., 2007). In line with previous data (Wierzba and Tax, 2016), bak1-1 mutant plants showed very little temperature-dependent growth reduction that was reminiscent of reduced growth in ETI (Fig. 3A; Supplementary Fig. S1B). Taken together, saul1-1 and chs1-2 mutants showed much stronger autoimmune responses upon low temperature treatment, compared with the other autoimmune mutants tested.
The autoimmune mutants saul1-1 and chs1-2 show stronger morphological phenotypes in response to lower temperature in comparison with other tested autoimmune mutants. (A) WT, saul1-1, snc1, cpn1-1, and bak1-1 seedlings grown for 12 d at 27 °C before autoimmune responses were induced by a temperature shift to 20 °C. Phenotypes were analysed after 2, 7, and 10 d at lower temperature. Day 0 represents the day of the shift. (B) WT, saul1-1, and chs1-2 plants grown for 16 d at 25 °C before inducing autoimmune responses at 12 °C and analysing the phenotypes after 7 d at low temperature. Day 0 represents the day of the shift.
As a next step, we analysed whether changes in the cellular structure occurred during snc1, cpn1-1, chs1-2, or bak1-1 immune responses by using TEM. Initially, we observed the structure of chloroplasts during autoimmunity. For TEM analysis, plant material was harvested directly prior to the temperature shift (day 0) and at day 7 following the decrease in temperature. While all mutant chloroplasts were indistinguishable from WT chloroplasts in plants grown at higher temperature (Supplementary Fig. S3A), we detected differences following low temperature-induced autoimmunity. In saul1-1 and chs1-2 leaf cells, the shape of chloroplasts changed to more rounded, and the thylakoid membranes disintegrated (Supplementary Fig. S3B, C; cf. Zbierzak et al., 2013; Disch et al., 2016). The most obvious changes in all autoimmune mutants was the size of plastoglobules (Supplementary Fig. S3B, C). The mean diameter of plastoglobules increased in snc1 (0.14±0.03 µm, n=120) and cpn1-1 (0.11±0.04 µm, n=65) chloroplasts when compared with WT chloroplasts (0.07±0.02 µm, n=59), though to a lesser extent than in saul1-1 chloroplasts (0.21±0.02 µm, n=91) (Supplementary Fig. S3D). The changes in the mean diameter of plastoglobules in chs1-2 chloroplasts (0.16±0.04 µm, n=116) were in the same range as for saul1-1 plastoglobules at 12 °C (0.17±0.05 µm, n=91) (Supplementary Fig. S3E). Notably, the size of plastoglobules in saul1-1 chloroplast was larger in plants grown at 20 °C than in plants grown at 12 °C.
In the cell walls no changes in structure were observed in any of the mutants at higher temperatures (Supplementary Fig. S3A). Whereas the cell wall structure in bak1-1 leaves was indistinguishable from WT 7 d after a decrease in temperature, changes in the cell wall structure of snc1 and cpn1-1 leaf cells were observed (Fig. 4A). The mean peak thickness of CEWDs in snc1 (0.63±0.56 µm, n=17) and cpn1-1 (0.56±0.27 µm, n=13) leaf cells was smaller than that in saul1-1 cells (2.59±0.79 µm, n=13; Disch et al., 2016) (Fig. 4B). In contrast, changes in the mean peak thickness of CEWDs of chs1-2 leaf cells (0.91±0.94 µm, n=18) were in the same range as for saul1-1 CEWDs (0.79±0.99 µm, n=15) (Fig. 4C and D). Interestingly, CEWDs were less pronounced in saul1-1 when grown at 12 °C when compared with CEWDs at 18 °C.
Cell wall deposits are a common feature of autoimmune mutants and were more pronounced in saul1-1 and chs1-2 mutants in comparison with other tested autoimmune mutants. (A) WT, saul1-1, snc1, and cpn1-1 mutants grown for 12 d at 27 °C before autoimmune responses were induced by a temperature shift to 20 °C. Samples for TEM analysis were taken 7 d after the temperature shift. (B) Violin plot of a Kruskal–Wallis test for comparing peak size of CEWDs from (A). (C) saul1-1 and chs1-2 plants grown for 16 d at 25 °C before inducing autoimmune responses at 12 °C and taking samples for TEM analysis after 7 d at low temperature. (D) Violin plot of a Kruskal–Wallis test for comparing peak size of CEWDs from the experiment in (C). Asterisks indicate significance level (****P<0.0001). In (A, C), scale bars represent 0.5 µm and arrows indicate the extent of CEWDs.
Recently, we showed that CEWDs in saul1-1 contain callose and lignin (Disch et al., 2016). We therefore investigated the presence of both substances in CEWDs of the other autoimmune mutants. For that purpose, we stained leaf cross sections of all autoimmune mutants with aniline blue. Positive fluorescence signals indicated the presence of callose in CEWDs of snc1, cpn1-1, chs1-2, and saul1-1 cells (Supplementary Fig. S4A, B). For the detection of lignin in the cell walls, we used scanning UMSP, which has been established as a useful technique for the topochemical detection and semiquantitative determination of lignin in situ. It enables a direct imaging of lignin, which displays a characteristic and unique ultraviolet absorbance spectrum (Koch and Schmitt, 2013). Measurements by UMSP detected the presence of lignin in CEWDs of snc1, cpn1-1, and chs1-2 plants (Supplementary Fig. S4C, D). Our data showed that CEWDs, containing callose as well as lignin, and changes in the structure of chloroplasts occurred not only in saul1-1 but also in other autoimmune mutants.
The temperature-dependent degree of CEWDs correlates with the severity of the saul1-1 autoimmune phenotype
Surprisingly, our temperature-shift experiments with saul1-1 plants revealed that CEWDs did not increase proportionally with decreasing temperatures. Instead, shifting plants from 25 °C to 12 °C resulted in less reduced growth and in smaller CEWDs in saul1-1 cells than shifting to 20 °C (Fig. 4A, B; Supplementary Fig. S1C). To confirm this observation, we repeated the experiments with shifts from 25 °C to 20, 18, and 12 °C. As expected, CEWDs were absent in plants grown constantly at 25°C (Fig. 5A). When after 12 d of growth at 25 °C saul1-1 mutants were grown for an additional 7 d at 20 °C and 18 °C, large CEWDs were detected in TEM analysis (Fig. 5B, C). The mean peak sizes were determined to be 1.17±0.90 µm (n=17) at 20 °C and 1.28±0.97 µm (n=20) at 18 °C (Fig. 5E). However, CEWDs in cells of saul1-1 plants shifted to 12 °C for 7 d were much less pronounced, though still present, with a mean peak size of 0.54±0.30 µm (n=16) (Fig. 5D, E). This was significantly lower compared with the peak size at 18 and 20 °C, despite one outlier with a peak size of 4.93 µm. To determine whether initial onset of CEWDs was also dependent on temperature, we analysed cells of saul1-1 mutant plants challenged for 1, 2, and 3 d by decreased temperature. At a temperature of 18 °C CEWDs first occurred after 2 d (Supplementary Fig. S5A), whereas at 12 °C CEWDs developed not before 3 d after the temperature shift (Supplementary Fig. S5B). Taken together, the formation of CEWDs was delayed and diminished at even lower temperatures, and these differences correlated with the severity of the saul1-1 phenotype.
Both cell wall deposits and the severity of the autoimmune phenotype are temperature-dependent in saul1-1 mutants. (A) CEWDs in saul1-1 grown at 25 °C for 12 d and for additional 7 d at 25 °C. Samples for TEM analysis were taken after these 19 d. (B–D) CEWDs in saul1-1 grown at 25 °C for 12 d and shifted to 20 (B), 18 (C) or 12 °C (D) for 7 d. Samples for TEM analysis were taken 7 d after the shift. (E) Violin plot of a Kruskal–Wallis test for comparing peak size of CEWDs of (A–D). Scale bars represent 1 µm. Double arrows indicate the extent of CEWDs. Arrows in (D) point to CEWDs.
GLUCAN SYNTHASE-LIKE5/ POWDERY MILDEW RESISTENT4 is responsible for the majority of cell wall deposits in the saul1-1 mutant
In many biotic interactions callose biosynthesis depends on the callose synthase gene Glucan SYNTHASE-LIKE5/POWDERY MILDEW RESISTENT4 (GSL5/PMR4), hereafter named PMR4 (Jacobs et al., 2003). The pmr4-1 mutation resulted in a decrease of callose deposition in PTI or papillae formation (Nishimura et al., 2003; Ellinger et al., 2013; Leslie et al., 2016). We quantified the expression level of PMR4 during saul1-1 autoimmunity following a decrease in temperature from 25 °C to 20 °C. Whereas at days 1, 2, and 4 after the temperature shift the expression was unchanged in WT and saul1-1 pad4-1 double mutants, PMR4 was induced in saul1-1 (Supplementary Fig. S6A). This induction of PMR4 expression was in a similar range as the 2- to 3-fold induction upon PTI activation through flg22 treatment of WT plants at 22 °C described by Keppler et al. (2018). To test whether this transcriptional upregulation in saul1-1 mutant plants points to a general feature in autoimmunity, we also quantified PMR4 expression in snc1, cpn1-1, and chs1-2 mutants. Indeed, the expression of PMR4 was increased in snc1, cpn1-1, and chs1-2 mutants challenged by a decrease in temperature (Supplementary Fig. S6B, C), indicating that the induction of PMR4 is a general molecular autoimmune response. The degree of PMR4 induction appeared to correlate with the degree of CEWD formation and the degree of the phenotype in the different mutants (cf. Fig. 3).
To study the function of PMR4 in the formation of CEWDs in autoimmunity, we generated a saul1-1 pmr4-1 double mutant. Surprisingly, this resembled the saul1-1 morphology when shifted to 20 °C, thus indicating that the saul1-1 autoimmune phenotype was independent of PMR4 (Fig. 6A). However, the CEWDs in saul1-1 pmr4-1 plants were strongly reduced (Fig. 6B), though not completely absent. Notably, the remaining CEWDs appeared fuzzier and less dense when compared with the PMR4-dependent CEWDs in the saul1-1 single mutant. The mean peak size was determined to be only 0.21±0.09 µm (n=18) (Fig. 6C). The changes in the structure of the chloroplasts in saul1-1 pmr4-1 cells resembled the changes observed in saul1-1 chloroplasts. The mean diameters of plastoglobules was determined to be 0.09±0.02 µm (n=85) in WT, 0.10±0.03 µm (n=131) in pmr4-1, 0.24±0.05 µm (n=99) in saul1-1, and 0.21±0.05 µm (n=193) in saul1-1 pmr4-1 chloroplasts (Fig. 6D). Whereas reduced growth was quantified by measuring the rosette diameter (Fig. 6E), expression of the immune marker gene EDS16 indicated autoimmunity of saul1-1 single and saul1-1 pmr4-1 double mutants on the molecular level (Fig. 6F). Our data showed that PMR4 knock-out resulted in strong reduction of CEWDs in saul1-1 but could not rescue the saul1-1 phenotype. In temperature-dependent saul1-1 autoimmunity, we also observed changes in the content of hemicelluloses, namely increased glucuronic acid content and decreased galactose and rhamnose content (Supplementary Fig. S7). These changes were not detected in saul1-1 pad4-1 double mutants, indicating that they were associated with autoimmunity.
Cell wall deposits but not the autoimmune phenotype of saul1-1 mutants are dependent on PMR4. Plants were grown at 25 °C for 12 d and shifted to 20 °C for 7 d. (A) Autoimmune phenotypes of WT, pmr4-1, saul1-1, and saul1-1 pmr4-1 mutants. (B) Structural changes of chloroplasts and cell walls of WT, pmr4-1, saul1-1, and saul1-1 pmr4-1 mutants. (C) Violin plot of a Kruskal–Wallis test for comparing CEWD peak size from the different plants in (A). (D) Violin plot of a Kruskal–Wallis test for comparing plastoglobule diameter from the different plants in (A). (E) Quantification of the rosette diameter of WT and mutant seedlings (mean ±standard error, n=5–8). (F) Relative expression of the immune marker gene EDS16 in WT and mutant seedlings (mean ±standard error, n=3). In (B), scale bars represent 1 µm and arrows indicate the extent of CEWDs. Asterisks in (C, D) indicate significance levels (***P<0.001, ****P<0.0001). Different superscripts in (E, F) indicate significant differences between groups, whereas identical superscripts denote no significant differences as analysed by Student’s t-test.
Discussion
In this study, in the process of analysing features of low temperature-induced plant autoimmune responses, we identified cell wall deposition to be a general feature of autoimmunity in Arabidopsis. In addition to other features of autoimmunity, such as reduced growth, changed leaf morphology, and immune gene expression, CEWDs were observed in saul1-1, snc1, cpn1-1, and chs1-2 mutants (Figs 1, 4). These deposits only formed at autoimmune permissive temperatures for each mutant line, whereas at higher temperatures plants were indistinguishable from the WT. Interestingly, the extent of deposits correlated with the severity of the autoimmune phenotype. The mutants with most severe phenotype, saul1-1 and chs1-2, exhibited similarly thick CEWDs, whereas snc1 and cpn1-1 were less affected and had less CEWDs. Correspondingly, bak1-1 plants showed very little symptoms of autoimmunity, and no CEWDs could be observed. It appears that CEWDs represent a cellular marker for immune responses that to a certain degree offers a quantitative relation to the morphological phenotypes. However, our double mutant analyses indicated that PMR4/GSL5-dependent CEWDs were not the cause of the saul1-1 autoimmune phenotype.
Temperature dependency is a hallmark of plant autoimmunity and indicative of the regulatory overlap between immune and temperature signaling (Velásquez et al., 2018). Temperature-dependent autoimmunity of Arabidopsis mutants represents a valuable tool to study immune responses that are normally triggered by pathogen effectors secreted into plant cells. Here, we investigated the dynamics of autoimmune responses upon changes in temperature. The extent of CEWDs was reduced at 12 °C when compared with the severity of the phenotype at 18 °C. We also observed a delay in the onset of CEWDs at 12 °C. It has previously been shown that cold stress responses in Arabidopsis were not activated before shifting the temperature below 10°C, whereas daytime temperature shifts to 12 °C resulted in reduced growth by inhibiting photosynthesis (Penfield, 2008; Pyl et al., 2012). The observed differences in saul1-1 at 12 °C may therefore be related to a reduced metabolic potential at lower temperature, as WT plants also grow less at 12 °C. The exposure of plants to low, but non-freezing, temperatures induces rapid metabolic responses that precede the acclimation to a new metabolic status (Korn et al., 2010; Furtauer et al., 2019). In line with this, the differences in CEWDs between 18 °C and 20 °C were very minor, indicating that a difference of only 2 °C may not be sufficient to change the range of plant metabolism significantly. However, the differential response at 12 °C may also be related to a lower immune capacity. It has, for example, been shown that several bacterial effector proteins were secreted effectively at temperatures between 18 °C and 20 °C, but not at lower temperatures (van Dijk et al., 1999). Plants may have evolved to invest in immunity according to the risk of pathogen infection. This idea may be supported by the fact that some bacteria and fungi exhibit sub-optimal growth at elevated temperatures (Pietikäinen et al., 2005). In contradiction to this idea, however, susceptibility of Arabidopsis to a bacterial pathogen has recently been shown to increase with elevated temperature (Huot et al., 2017). Generally, the virulence of pathogens depends as much on their host plant as on abiotic conditions such as temperature (Barrett et al., 2009).
How plants sense temperature changes that may finally lead to autoimmunity is not well understood. In most cases, auto-activation of NLRs as well as hybrid necrosis controlled by NLRs is suppressed at elevated temperatures above 22–30 °C (Bomblies and Weigel, 2007; Alcázar and Parker, 2011; Balint-Kurti, 2019). Temperature sensitivity may be an intrinsic property of NLRs. It has been suggested that auto-activation of the NLR variant scn1-1 is correlated to its localization in the nucleus, because second site mutations abolished its nuclear localization and released autoimmunity (Zhu et al., 2010). Furthermore, the activation of NLRs depends on the chaperone function of HSP90, RAR1, and SGT1B (Zhang et al., 2010). This may indicate that the temperature sensitivity correlates to NLR conformation either directly or through changes in the chaperones involved. The SOC3-dependent autoimmune responses in saul1-1 were indeed also rescued in saul1-1 sgt1b double mutants (Lee et al., 2016). Despite good evidence for NLR temperature sensitivity, it is also possible that the actual temperature sensing does not occur at the level of NLRs. It is still a matter of debate whether transcript and protein levels of EDS1 and PAD4 contribute to temperature dependency of autoimmunity (Wang et al., 2009; Alcázar and Parker, 2011). In addition, components of the sensing mechanism that controls growth at ambient temperatures, such as phytochrome B and PIF4, may be important for immune regulation too (Legris et al., 2016; Gangappa et al., 2017). It is still unclear whether such ambient temperature sensing has a molecular link to NLR activation in autoimmunity. The SUMO E3 ligase SIZ1 may be a good candidate to connect both, because at elevated temperature SIZ1 enhanced growth but also inhibited SNC1-mediated autoimmunity (Hammoudi et al., 2018).
The strong correlation between the temperature-dependent autoimmune phenotype and the severity of CEWDs was additionally supported by the identification of a point of no return in saul1-1 autoimmunity. When shifted to 18 °C for 2 d, CEWDs manifested, and when shifted for 3 or more days, CEWDs were extensive, and autoimmunity could no longer be rescued after transfer back to 25 °C (Figs 1, 2, 5). This indicated the apparent irreversibility of a change within the plant, with CEWDs and disintegration of thylakoid membranes being the most prominent changes. Generally, the CEWDs observed during autoimmunity contained lignin and callose (Supplementary Fig. S4; Disch et al., 2016). Pathogen infection with different bacterial pathogens resulted in the accumulation of lignin (Lee et al., 2001; Mohr and Cahill, 2007). The induction of immune responses has been shown to lead to lignin deposition within a few hours (Malinovsky et al., 2014). Different studies related lignin to immune responses by showing that a decrease in cellulose biosynthesis induced both lignin biosynthesis and defense responses. The respective mutants often showed the induction of defense genes and reduced growth, both hallmarks of autoimmunity (Ellis and Turner, 2001; Ellis et al., 2002; Caño-Delgado et al., 2003). The mechanism of growth reduction is not resolved completely, though it has mostly been related to lignin reduction in the stem vasculature (Besseau et al., 2007; Mir Derikvand et al., 2008; Schilmiller et al., 2009; Li et al., 2010; Bonawitz and Chapple, 2013). However, it was also found in mutant plants with decreased lignin that showed reduced growth without any evidence for collapsed xylem (Anderson et al., 2015). This suggested the presence of other lignin-dependent mechanisms such as possibly immune responses resulting in reduced growth.
We focused our studies on the role of callose contained in CEWDs. It has been shown that the time required for callose deposits to occur in response to flg22 treatment was several hours (Ellinger and Voigt, 2014). We were able to resolve callose deposits in saul1 autoimmunity after 48 h (Disch et al., 2016). In incompatible plant–pathogen interactions, callose is deposited to form papillae at the site of infection to prevent, for example, fungal penetration (Ellinger et al., 2013). Though CEWDs in autoimmunity differed in their appearance from the structure of papillae, they also showed focal accumulation and were not uniformly distributed around the cell. In line with the induction during saul1 autoimmunity of the PMR4/GSL5 gene, which encodes the callose synthase responsible for callose deposition to papillae in plant–fungus interactions (Nishimura et al., 2003; Ellinger et al., 2013), CEWDs were largely abolished in saul1-1 pmr4-1 double mutants. Nevertheless, the observed autoimmune phenotype was not affected in saul1-1 pmr4-1 plants (Fig. 6). This indicated that large callose deposits were not the cause of autoimmunity in saul1-1 mutants. However, we cannot exclude that the remaining PMR4/GSL5-independent CEWDs mediate autoimmunity in saul1-1 mutants. The remaining CEWDs in saul1-1 pmr4-1 mutants may contain lignin, but also additional cell wall polysaccharides (Supplementary Figs S4, S7; Disch et al., 2016). The observed increase in glucuronic acid in saul1-1 autoimmunity may indicate that the attachment of glucuronic acid side chains to xylan is increased. Loss of glucuronic acid side chains to xylan in gux1/2/3 glucuronlytransferase triple mutants has been shown to result in less cell wall thickening (Lee et al., 2012). Xylans are important for secondary wall formation (Cosgrove, 2005; Sarkar et al., 2009). Analysis of secondary cell wall mutants with increased xylose content or modifications of glucuronoxylans and xyloglucans suggested that these changes reduce the Arabidopsis susceptibility of pathogen infection (Delgado-Cerezo et al., 2012). The effects of changes in the structure of cell walls on pathogen resistance have been observed in other Arabidopsis mutants (Miedes et al., 2014; Molina et al., 2021, Wan et al., 2021). The observed decrease in the content of rhamnose and galactose (Supplementary Fig. S7), which are components of cell wall pectins, may indicate that changes in cell wall pectins occur in saul1-1 autoimmunity. Arabidopsis mutants with reduced pectin in cell walls were less resistant to bacterial and fungal pathogens (Bethke et al., 2016). Pectins have important functions in plant development as well as immunity and are the source for oligogalacturonides that are generated upon infection and cell wall damage to induce plant immune responses (Benedetti et al., 2015). Interestingly, cell wall fractions of cell wall mutants that have increased pectin content have been shown to elicit immune responses (Molina et al., 2021). Thus, plant defenses require signals from the cell wall to induce defense pathways. Cell wall damage generally triggers various immune responses, including the production of callose and lignin (Hamann et al., 2009; Hamant and Haswell, 2017; Engelsdorf et al., 2018). In the future, it will be important to analyse the dynamics of changes in cell wall components in plant autoimmunity induced by low temperatures by using different available tools for cell wall analysis (De Lorenzo et al., 2019).
Contrary to cell wall changes, the observed alterations in the structure of chloroplasts may cause autoimmune responses. Yellowing of autoimmune mutant plant leaves, the disintegration of thylakoid membranes, and increased number and size of plastoglobules indicate a strong negative effect on chloroplasts. In the autoimmune mutant chs1-1, the loss of thylakoid membrane integrity and the increase in plastoglobules preceded cell death (Zbierzak et al., 2013). These processes were also related to cell death during leaf senescence and upon oxidative stress (Austin et al., 2006). Plastoglobules may take up lipids derived from degradation of chlorophyll resulting in leaf yellowing and from disintegration of thylakoid membranes (Lippold et al., 2012). The described changes in the structure of saul1-1 chloroplasts occurred within the first 2–3 d and may thus also be early events of autoimmunity in this strong autoimmune mutant (Fig. 1). Recently, it was suggested that SAUL1 may regulate the chloroplast protein Chloroplast-localized Senescence-Associated Protein (CSAP), which is a positive regulator of chloroplast senescence (So et al., 2020). Future experiments may determine whether the alterations in the structure of chloroplasts, which were still present in saul1-1 pmr4-1 double mutants, may indeed be the cause of the autoimmune phenotype.
Supplementary data
The following supplementary data are available at JXB online.
Fig. S1. Rosette diameters of the indicated genotypes in the respective temperature shift experiments.
Fig. S2. The expression of immune marker genes is increased in different autoimmune mutants.
Fig. S3. Low temperature-induced autoimmunity causes structural changes of chloroplasts.
Fig. S4. Cell wall deposits contain callose and lignin in all tested autoimmune mutants but are more pronounced in saul1-1 and chs1-2.
Fig. S5. Structural changes appear after 3 d at 12 °C but already after 2 d at 18 °C.
Fig. S6. The expression of PMR4 is increased in different autoimmune mutants.
Fig. S7. Analysis of hemicellulose composition in saul1-1 plants revealed only small changes in comparison with WT and saul1-1 pad4-1 plants.
Acknowledgements
We thank E. Woelken for support with TEM, D. Paul for support with UMSP, X. Li (University of British Columbia, Canada) for supplying snc1, cpn1-1, and bak1-1 seeds, and P. Dörmann (University of Bonn, Germany) for supplying chs1-2 seeds.
Author contributions
GH, SP, E-MG, and TL performed experiments. GK supervised UMSP experiments and reviewed and edited the manuscript. CV supervised HPLC experiments and commented on the manuscript. SH supervised the project. GH, SP, and SH were responsible for conceptualization and original draft preparation.
Conflict of interest
The authors declare no conflicts of interest.
Funding
This work was funded by State of Hamburg (Hybrids—chances and challenges of new genomic combinations, LFF FV36 to SMP and SH).
Data availability
All data supporting the findings of this study are available within the paper and within its supplementary materials published online.
References
Author notes
These authors contributed equally to this work.
Comments