Introduction

Plant immunity is complex and involves proteins with roles in pathogen detection and defense activation (Jones & Dangl 2006). The recognition of molecules derived from pathogens/microbes results in elicitation of Pattern Triggered Immunity (PTI) (Monaghan & Zipfel 2012; Bigeard et al. 2015). Successful pathogens can deliver effectors into plant cells to dampen PTI. Perception of these effector molecules by plants results in Effector Triggered Immunity (ETI) (van Wersch et al. 2020; Zhou & Zhang 2020). Further, PTI and ETI are intricately connected and have been shown to mutually potentiate each other (Ngou et al. 2021; Pruitt et al. 2021; Tian et al. 2021; Yuan et al. 2021b,a).

Immune responses need to be tightly controlled since any dysregulation can lead to either an insufficient or an excessive immune output. Autoimmune mutants often display a smaller size proportional to the level of constitutive immunity, making them a valuable resource for genetic analyses. They have been successfully used in genetic screens to identify contributors mediating immunity (Johnson et al. 2012; van Wersch et al. 2016).

One such autoimmune mutant is pi4kβ1,2 (Janda et al. 2014; Šašek et al. 2014; Kalachova et al. 2020). pi4kβ1,2 is mutated in the redundant phosphatidylinositol 4-phosphate (PI4P) biosynthetic enzymes, Phosphatidylinositol 4-Kinase (PI4K) β1 and PI4Kβ2. pi4kβ1,2 plants display autoimmunity-associated dwarfism and enhanced resistance against bacterial and oomycete pathogens (Janda et al. 2014; Šašek et al. 2014; Antignani et al. 2015; Thulasi Devendrakumar et al. 2023). In addition to displaying autoimmunity and dwarfism, pi4kβ1,2 seedlings also grow shorter roots with abnormal root hairs (Preuss et al. 2006; Šašek et al. 2014; Starodubtseva et al. 2022; Thulasi Devendrakumar et al. 2023). Through reverse genetic analysis it has been shown that the defense hormone salicylic acid (SA) and Enhanced Disease Susceptibility 1 (EDS1), a central component required for ETI, are both contributing to pi4kβ1,2’s autoimmunity (Šašek et al. 2014). However, the exact cause of the autoimmunity displayed by pi4kβ1,2 is still unclear (Šašek et al. 2014). Additionally, no genes required for pi4kβ1,2’s short root defect have been identified so far. Thus, we conducted an EMS forward genetic screen aiming to identify novel genes involved in pi4kβ1,2’s autoimmunity (Thulasi Devendrakumar et al. 2024).

Here, we report that Signal Peptide Peptidase (SPP) is required for pi4kβ1,2 autoimmunity and root length defect. SPP is annotated as an intramembrane cleaving aspartic protease (Ponting et al. 2002; Weihofen et al. 2002; Tamura et al. 2008). Arabidopsis SPP has been shown to be ER localized and an essential gene required for the formation of viable pollen (Tamura et al. 2008; Han et al. 2009). Since knockout of SPP leads to lethality, functional characterization of Arabidopsis SPP is lacking. From our pi4kβ1,2 suppressor screen, we identified three spp partial loss-of-function alleles that suppress all the pi4kβ1,2 defects. Further, we uncovered that both PI4Kβ1,2 and SPP contribute to ER stress responses.

Results

Identification of suppressors 79-1, 145-1, and 171-1 from the pi4kβ1,2 suppressor screen

An ethyl methane sulfonate (EMS) mutagenesis based forward genetic screen was carried out to investigate the source of pi4kβ1,2 autoimmunity (Thulasi Devendrakumar et al. 2024). Suppressors 79-1, 145–1, and 171–1 were identified (Fig. 1a). 79-1 and 171-1, and to a lesser degree 145-1 suppressed the autoimmunity-associated dwarfism of pi4kβ1,2 (Fig. 1a, b). Further, they displayed reduced expression of the defense marker gene Pathogenesis related 1 (PR1), and PR2 when compared to pi4kβ1,2 (Fig. 1c, d). pi4kβ1,2 was also reported to accumulate salicylic acid (SA), with increased transcript levels of the SA biosynthetic gene Isochorismate synthase 1 (ICS1) (Wildermuth et al. 2002; Janda et al. 2014; Šašek et al. 2014). The suppressors 79-1 and 171-1 displayed reduced expression of ICS1 and Enhanced Disease Susceptibility 5 (EDS5), a gene that encodes a protein required for the transport of the SA precursor isochorismate from the chloroplast to the cytoplasm (Fig. 1e, f; Rekhter et al. 2019). Consistently, the suppressors also supported more growth of the oomycete pathogen Hyaloperonospora arabidopsidis Noco2 (Ha Noco2) compared to pi4kβ1,2 (Fig. 1g). In addition to suppression of autoimmunity and the associated dwarfism, these suppressors displayed suppression of pi4kβ1,2’s short root phenotype (Fig. 1h). Further, they showed suppression of the root hair defects of pi4kβ1,2 (Fig. 1i). Thus suppressors 79–1, 145–1, and 171–1 show suppression of both the autoimmunity and root defects of pi4kβ1,2.

Fig. 1
figure 1

Characterization of suppressors of pi4kβ1,2. a Morphology of 4-week-old WT, pi4kβ1,2, 79–1, 145–1, and 171–1 plants. b Quantification of rosette weight of 4-week-old plants of the indicated genotypes. The error bars represent SD of the replicates (n = 8). c Pathogenesis Related 1 ( PR1 ) expression in plants of the indicated genotypes. PR1 expression was normalised to the expression level of ACTIN7 . The error bars represent SD of the biological replicates (n = 3). d Pathogenesis Related 2 ( PR2 ) expression in plants of the indicated genotypes. PR2 expression was normalised to the expression level of ACTIN7. The error bars represent SD of the biological replicates (n = 3). e Enhanced Disease Susceptibility 5 ( EDS5 ) expression in plants of the indicated genotypes. EDS5 expression was normalised to the expression level of ACTIN7 . The error bars represent SD of the biological replicates (n = 3). f Isochorismate Synthase 1 ( ICS1 ) expression in plants of the indicated genotypes. ICS1 expression was normalised to the expression level of ACTIN7 . The error bars represent SD of the biological replicates (n = 3). g Quantification of Hyaloperonospora arabidopsidis (Ha) Noco2 growth on plants of the indicated genotypes. The error bars represent SD of the biological replicates (n = 5). h Root lengths of 10-day-old plate-grown seedlings of the indicated genotypes. The error bars represent SD of the replicates (n = 5). From (b to h), the letters indicate significant difference between the different genotypes as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference (HSD) test. Genotypes denoted with the different letters have significant difference (p <0.05). i Bright field microscopy images showing root hair morphology of plate grown 14-day-old seedlings of the indicated genotypes. The scale bar indicates 100 μm

Full size image

Suppressor mutations identified by mapping-by-sequencing

To identify the causal mutations that led to the suppression of pi4kβ1,2 phenotypes, we performed mapping-by-sequencing. The suppressor mutations present in 79–1, 145–1, and 171–1 will hence be referred to as 79–1, 145–1, and 171–1. The suppressors were backcrossed to pi4kβ1,2 to establish their dominance, to determine whether the suppression is caused by a single nuclear mutation, and to generate mapping populations for whole genome sequencing.

F1 plants generated by crossing either 79–1 or 145–1 to pi4kβ1,2 appeared pi4kβ1,2-like, suggesting that the suppressor mutations are recessive. In F2, the 79–1 × pi4kβ1,2 F2 population showed a 127:41 segregation of pi4kβ1,2-like: suppressor-like plants, consistent with the expected 3:1 ratio for recessive mutations (χ2 = 0.032; P value = 0.8586). The 145–1 × pi4kβ1,2 F2 population showed a 109:36 segregation of pi4kβ1,2-like: suppressor-like plants that was also consistent with the expected 3:1 ratio for recessive mutations (χ2 = 0.002; P value = 0.9618). Thus, 79–1 and 145–1 carry single nuclear recessive mutations. For each mutant, tissue was collected from the suppressor-like F2 plants for DNA extraction and whole genome sequencing.

The F1 generated by backcrossing 171–1 to pi4kβ1,2 appeared suppressor-like, indicating that it is dominant. Further, in the 171–1 × pi4kβ1,2 F2 population we observed a 53:151 segregation of pi4kβ1,2-like: suppressor-like plants, consistent with the expected 1:3 ratio for a single dominant nuclear mutation (χ2 = 0.105; P value = 0.7464). F3 seeds from the suppressor-like F2 plants were harvested and planted. Of the 151 F3 populations, 43 displayed no segregation of plants with pi4kβ1,2-like phenotype, demonstrating homozygosity of the suppressor locus. Tissue was collected from these 43 homozygous F3 populations for DNA extraction and whole genome sequencing.

A shared linkage region was identified for these three suppressors on chromosome 2 (Fig. 2a, b, c). Further, they carried different mutations in Signal Peptide Peptidase (SPP; AT2G03120; Fig. 2d) within the identified linkage regions. These mutations all resulted in single amino acid substitutions in SPP (Fig. 2d).

Fig. 2
figure 2

Mapping by next-generation-sequencing of suppressors 79–1, 145–1, and 171–1. a, b, c Linkage map showing the linkage region for the suppressors 79–1 (a), 145–1 (b), and 171–1 (c) to the beginning of chromosome 2. The Signal Peptide Peptidase (SPP) mutation is highlighted using a red data point. d SPP mutations in the suppressors and their resulting amino acid changes

Full size image

Mutations in Signal  Peptide  Peptidase are the causal mutations responsible for the pi4kβ1,2 suppression in 79–1, 145–1 and 171–1

To confirm that the mutations in SPP are causing the pi4kβ1,2 suppression phenotypes, we transformed the 79–1 and 171–1 plants with NP::SPP driven by its native promoter (NP). These NP::SPP transgenic plants displayed a pi4kβ1,2-like size (Fig. 3a, S1a). Further, the autoimmunity and the root length returned to pi4kβ1,2 levels in the transgenic lines (Fig. 3b, c, S1b, c). When the recessive suppressors 79–1 and 145–1 were crossed, the F1 progeny appeared suppressor like (Fig. 3d), further confirming that 79–1 and 145–1 carry mutant alleles of the same gene. In addition, to test that the SPP mutation in 171–1 is the dominant suppressor mutation responsible for the suppression we transformed pi4kβ1,2 with 35S::SPP171−1, cloned from 171–1 genomic DNA. The resulting transgenic plants displayed suppression of pi4kβ1,2 autoimmunity-associated dwarfism, immunity, and root length defect (Fig. 3e-g). These results confirmed that the SPP mutation in 171–1 is the dominant mutation responsible for the suppression of pi4kβ1,2 phenotypes. Further, since the suppressor 171–1 resembles the recessive alleles 79–1, and 145–1, it likely carries a dominant-negative (DN) allele of SPP. Taken together, the mutations in SPP are responsible for the pi4kβ1,2 suppression seen in 79–1, 145-1, and 171–1. We named the spp alleles 79–1, 145–1, and 171–1 as spp-3, spp-4, and spp-5 respectively. As knockout (KO) of this gene is lethal (Tamura et al. 2008; Han et al. 2009), these alleles are likely partial loss-of-function alleles of SPP.

Fig. 3
figure 3

Mutations in SPP are responsible for the suppression phenotypes seen in 79–1, 145–1, and 171–1. a Morphology of 4-week-old WT, pi4kβ1,2, 79–1, and two independent NP::SPP transgenic lines in the 79–1 background. b Quantification of Ha Noco2 growth on plants of the indicated genotypes. The error bars represent SD of the biological replicates (n = 5). c Root lengths of 10-day-old plate-grown seedlings of the indicated genotypes. The error bars represent SD of the replicates (n = 5). d Morphology of 4-week-old WT, pi4kβ1,2, 79–1, 145–1, and 79–1 × 145–1 F1 plants. e Morphology of 4-week-old WT, pi4kβ1,2, 171–1, and two independent 35S::SPP171−1 transgenic lines in the pi4kβ1,2 background. f Quantification of Ha Noco2 growth on plants of the indicated genotypes. The error bars represent SD of the biological replicates (n = 5). g Root lengths of 10-day-old plate-grown seedlings of the indicated genotypes. The error bars represent SD of the replicates (n = 5). In (a to c) and (e to g), the letters indicate significant difference between the different genotypes as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference (HSD) test. Genotypes denoted with the different letters have significant difference (p < 0.05)

Full size image

SPP and PI4Kβ1 are localised to the ER

SPP is reported to be an intramembrane cleaving aspartyl protease and was previously reported to be ER localised in Arabidopsis (Tamura et al. 2008). In order to verify its localization, we generated an SPP-mCitrine fusion. We transiently co-expressed SPP-mCitrine (Yellow; Fig. 4) and the ER marker ER-ck (Cyan; Fig. 4; Nelson et al. 2007) in Nicotiana benthamiana leaves to test whether the two proteins colocalize. As shown in Fig. 4, a network-like localization of the SPP-mCitrine fusion similar to the ER marker ER-ck was observed in confocal microscopy studies (Fig. 4). Colocalization of the SPP-mCitrine and the ER-ck signal was also obvious (Fig. 4). Interestingly, in addition to signals clearly confined to the ER, SPP-mCitrine also displayed punctate localization that did not overlap with the signal of the ER marker (Fig. 4). We further investigated the localization of the PI4Kβs. PI4Kβ1 was previously shown to localize to distinct compartments of the trans Golgi network in Arabidopsis root hair cells (Kang et al. 2011; Antignani et al. 2015). We generated a native promoter driven NP::HA-mNeonGreen(mNG)-PI4Kβ1 construct. The transgene was confirmed to be functional since it could complement the morphological, immune, and root phenotypes of pi4kβ1,2 (Fig. 5a-c). Further, to check the localization of PI4Kβ1 we transiently expressed HA-mNG-PI4Kβ1 in N. benthamiana leaves. We observed a network like localization of PI4Kβ1 (Yellow; Fig. 5d). The signals of the HA-mNG-PI4Kβ1 and ER-ck largely overlapped, indicating that the PI4Kβ1 protein is localized to the ER (Fig. 5d). Thus, both SPP and PI4Kβ1 localize to the ER when transiently expressed in N. benthamiana leaves.

Fig. 4
figure 4

SPP is localized to the ER. Co-localization of SPP-mCitrine with the endoplasmic reticulum (ER) marker ER-ck following transient expression in Nicotiana benthamiana leaf epidermal cells. Clockwise from top-left, images show SPP-mCitrine fluorescence in yellow, ER-ck fluorescence in cyan, the transmission light (TL) image in greyscale, and overlay of mCitrine and CFP channels. The colocalization of yellow and cyan signals appear as green. The confocal images are 25 μm Z-stacks while the TL image is of a single plane. The scale bars are 40 μm

Full size image
Fig. 5
figure 5

PI4Kβ1 is localized to the ER. a Morphology of 5-week-old WT, pi4kβ1,2, and two independent NP::HA-mNeonGreen-PI4Kβ1 transgenic lines in the pi4kβ1,2 background. b Root lengths of 10-day-old plate-grown seedlings of the indicated genotypes. The error bars represent SD of the replicates (n = 5). In (b-c) the letters indicate significant difference between the different genotypes as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference (HSD) test. Genotypes denoted with the different letters have significant difference (p < 0.05). c Quantification of Ha Noco2 growth on plants of the indicated genotypes. The error bars represent SD of the biological replicates (n = 5). d Co-localization of HA-mNG-PI4Kβ1 with the endoplasmic reticulum (ER) marker ER-ck following transient expression in N. benthamiana leaf epidermal cells. Clockwise from top-left, images show HA-mNG-PI4Kβ1 fluorescence in yellow, ER-ck fluorescence in cyan, the transmission light (TL) image in greyscale, and overlay of mNG and CFP channels. The colocalization of yellow and cyan signals appear as green. The confocal images are 25 μm Z-stacks while the TL image is of a single plane. The scale bars are 40 μm

Full size image

Identified SPP mutations result in non-synonymous mutations of highly conserved residues

SPP is a multi-pass membrane protein with 9 transmembrane helices (Fig. 6a; Ponting et al. 2002; Weihofen et al. 2002; Tamura et al. 2008). It is a presenilin type protease with two active aspartic acid residues present within the catalytic YD and the GXGD motifs present on its sixth and seventh transmembrane helices respectively (Fig. 6a; Ponting et al. 2002; Weihofen et al. 2002; Spasic et al. 2006). Arabidopsis SPP is an essential gene since the null spp alleles are lethal (Tamura et al. 2008; Han et al. 2009). To understand the consequence of the spp mutations identified in our screen on SPP function, we generated an amino acid sequence alignment and phylogenetic tree of Arabidopsis and human SPP and SPP-like (SPPL) proteins (Fig. 6b). The Arabidopsis SPP is most closely related to the human SPP (Fig. 6b). The Arabidopsis SPP homolog, SPP-Like (SPPL) 1 shows the highest homology to the Human SPPL3 (Fig. 6b). The other Arabidopsis SPP homologs, SPPL2-5 form a sister clade along with the human SPPL2a, 2b, and 2c (Fig. 6b). The mutation in spp-3 allele leads to the Leu(167)Phe mutation of a fairly conserved amino acid present in the fifth transmembrane domain of SPP (Fig. 6c). Further, spp-4 and spp-5 mutations lead to missense mutations of highly conserved residues in the proximity of the GXGD active site in the seventh transmembrane domain (Fig. 6d). The amino acids mutated in the three spp alleles are largely conserved in all Arabidopsis and human SPP homologs, indicating that they are likely important for the function of SPP and SPPLs. Since complete loss of SPP function leads to lethality, the spp alleles likely have reduced SPP function due to mutations of amino acids that are in close proximity to its active sites.

Fig. 6
figure 6

spp-3, spp-4, and spp-5 mutations lead to missense mutations of conserved amino acids in SPP. a Diagram of SPP domain architecture with labeled active site motifs and the mutations sites. b Phylogeny of Arabidopsis and human SPP and SPPLs. The tree was constructed with full length amino acid sequences. The numbers above the branches indicate the bootstrap values from 1000 bootstrap replicates. The scale denotes branch length. The following abbreviations have been used in the tree to denote the organisms: At- Arabidopsis thaliana and Hs- Homo sapiens. c Multiple sequence alignment showing the region containing the spp-3 mutation present in the suppressor 79–1. d Multiple sequence alignment showing the region containing the spp-4 and spp-5 mutations present in the suppressors 145–1 and 171–1 respectively

Full size image

SPP transcription is up-regulated in pi4kβ1,2

SPP has been shown to be involved in the unfolded protein response (UPR) and endoplasmic reticulum–associated protein degradation (ERAD) in humans, the human parasite Plasmodium falciparum, and the fungal pathogen Ustilago maydis (Harbut et al. 2012; Allen et al. 2014; Chen et al. 2014; Pinter et al. 2019). In Arabidopsis, SPP was shown to be induced upon ER stress (Iwata et al. 2010). Thus, we tested SPP expression in pi4kβ1,2. As shown in Fig. 7a, about 2.5-fold upregulated expression of SPP was observed in pi4kβ1,2. Further, the pi4kβ1,2 suppressors 79–1 and 171–1 showed complete suppression of SPP upregulation (Fig. 7a). However, 145–1, which is comparatively a weak suppressor of pi4kβ1,2 (Fig. 1), displayed SPP expression similar to pi4kβ1,2 (Fig. 7a). These results indicate that SPP is induced in pi4kβ1,2 and that mutations in SPP are capable of suppressing this upregulated SPP expression seen in pi4kβ1,2.

Fig. 7
figure 7

SPP mutations reduce ER stress sensitivity in both WT and pi4kβ1,2. a Expression of SPP in WT, pi4kβ1,2, 79–1, 145–1, and 171–1 normalised to the expression level of ACTIN7. The error bars represent SD of the biological replicates (n = 3). The letters indicate significant difference between the different genotypes as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference (HSD) test. Genotypes denoted with the different letters have significant difference (p < 0.05). b, 21-day-old WT, pi4kβ1,2, 79–1, 145–1, and 171–1 seedlings grown on ½ MS agar plates supplemented with the indicated concentration of DTT. c, d, e, f, 14-day-old WT, spp-3, spp-4, and spp-5 seedlings grown vertically on square ½ MS agar plates supplemented with 0, 1, 1.5, and 2 mM DTT respectively. g, h, i, j, Quantification of the root lengths of the 14-day-old vertical plate grown WT, spp-3, spp-4, and spp-5 seedlings shown in c-f. The error bars represent SD of the replicates (n = 10). The letters indicate significant difference between the different genotypes as determined using a one-way ANOVA with post hoc Tukey’s Honestly Significant Difference (HSD) test. Genotypes denoted with the different letters have significant difference (p < 0.01). k Expression of BiP1,2 in WT, pi4kβ1,2, 79–1, 145–1, and 171–1 normalised to the expression level of ACTIN7. The error bars represent SD of the biological replicates (n = 3). l Expression of BiP3 in WT, pi4kβ1,2, 79–1, 145–1, and 171–1 normalised to the expression level of ACTIN7. The error bars represent SD of the biological replicates (n = 3). m Expression of ERdj3B in WT, pi4kβ1,2, 79–1, 145–1, and 171–1 normalised to the expression level of ACTIN7. The error bars represent SD of the biological replicates (n = 3). In k-m genotypes denoted with the different letters have significant difference (p < 0.05) as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference (HSD) test

Full size image

pi4kβ1,2 displays heightened ER stress sensitivity and the spp mutations reduce the ER stress sensitivity of pi4kβ1,2 and WT

The enhanced SPP expression prompted us to test whether the spp and pi4kβ1,2 plants show altered ER stress response. When grown on ½ MS plates supplemented with the ER stress inducing chemical dithiothreitol (DTT), pi4kβ1,2 exhibited enhanced sensitivity to DTT with seedlings appearing yellow and dying on plates supplemented with 2 mM DTT (Fig. 7b). In comparison, the suppressors 171–1 and 79–1 can survive better on plates supplemented with 2 mM DTT. 145–1 displayed marginal suppression of pi4kβ1,2’s DTT sensitivity (Fig. 7b).

We further tested whether the spp single mutants displayed reduced sensitivity to DTT. As shown in Fig. 7c-j, the spp-3 and spp-5 single mutant seedlings grew longer roots on DTT containing plates compared to WT. However, spp-4 displayed a similar sensitivity to DTT as WT. Further, while expression of the ER-stress marker genes Immunoglobulin-Binding Protein (BiP) 1,2, BiP3 and Endoplasmic Reticulum-localized DnaJ family 3B (ERdj3B) were elevated in pi4kβ1,2, suppressors 79–1 and 171–1 displayed largely suppressed expression of these ER stress marker genes (Fig. 7k-m). 145–1 displayed a similar or mildly suppressed expression of these genes. These results uncover the antagonistic contributions of PI4Kβ and SPP in plant ER stress responses.

spp single mutants have WT level of immunity

SPP was shown to be upregulated in seedlings treated with the immune related plant phytohormone N-hydroxypipecolic acid (NHP; Yildiz et al. 2021) or treatment with pathogen associated molecular patterns flg22 and elf18 (Bjornson et al. 2021) (Table 1). To examine the role of SPP in immunity, infection experiments were performed using the spp single mutants. The spp alleles displayed WT levels of immunity against the virulent bacterial pathogen Pseudomonas syringae pv. maculicola ES4326, the type III secretion system deficient Pseudomonas syringae DC3000 pv. tomato hrcC− which triggers pattern triggered immunity (PTI), or the avirulent Ha Emwa1 which triggers RPP4 mediated effector triggered immunity (ETI) (Fig. 8). Thus, partial loss of SPP does not lead to observable defects in basal immunity, PTI or ETI.

Table 1 SPP is upregulated in response to treatment with NHP, flg22 and elf18
Full size table
Fig. 8
figure 8

SPP single mutants display WT level of immunity against pathogens. a Quantification of Psm ES4326 growth in leaves of the indicated genotypes. Plants were infiltrated with bacterial suspension at OD600 = 0.0001, and bacterial titer was measured on day 0 and day 3 post infiltration. The error bars represent SD of the biological replicates (Day 0: n = 2; Day 3: n = 8). Genotypes denoted with the different letters have significant difference (p < 0.01) as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference test. b Quantification of Pst DC3000 hrcC− growth in leaves of the indicated genotypes. Plants were infiltrated with bacterial suspension at OD600 = 0.001, and bacterial titer was measured on day 0 and day 3 post infiltration. The error bars represent SD of the biological replicates (Day 0: n = 2; Day 3: n = 8). Genotypes denoted with the different letters have significant difference (p < 0.05) as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference test. c Quantification of Ha Emwa1 growth on plants of the indicated genotypes. The error bars represent SD of the biological replicates (n = 5). Genotypes denoted with the different letters have significant difference (p < 0.01) as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference test

Full size image

Human SPP ortholog can complement the phenotypes of the suppressor 79–1

From the phylogenetic tree shown in Fig. 6b, Arabidopsis SPP is the most similar to the human SPP ortholog, HsSPP. In order to test whether HsSPP can perform the functions of Arabidopsis SPP, we performed a heterologous complementation test by transforming the 79–1 suppressor mutant with 35S promoter driven HsSPP cDNA clone (35S::HsSPP). Transgenic plants displayed partial reversion to pi4kβ1,2-like dwarf phenotype (Fig. 9a). Further, the transgenic plants displayed autoimmunity and root length defect close to the pi4kβ1,2 mutant (Fig. 9b, c). These results suggest that human SPP is functionally similar to Arabidopsis SPP.

Fig. 9
figure 9

HsSPP can complement the phenotype of the suppressor 79–1. a Morphology of 4-week-old WT, pi4kβ1,2, 79–1, and two independent 35S::HsSPP transgenic lines in the 79–1 background. b Quantification of Ha Noco2 growth on plants of the indicated genotypes. The error bars represent SD of the biological replicates (n = 5). c Root lengths of 10-day-old plate-grown seedlings of the indicated genotypes. The error bars represent SD of the replicates (n = 5). In (b-c) the letters indicate significant difference between the different genotypes as determined using a one-way ANOVA with post-hoc Tukey’s Honestly Significant Difference (HSD) test. Genotypes denoted with the different letters have significant difference (p < 0.05)

Full size image

Discussion

Signal Peptide Peptidase (SPP) and SPP-like (SPPL) proteins were first identified in humans (Weihofen et al. 2002). Since then, they have been found in animals including mouse (Urny et al. 2003), Caenorhabditis elegans (Grigorenko et al. 2004), Drosophila (Casso et al. 2005), and zebrafish (Krawitz et al. 2005). SPP was shown to be essential for development in C. elegans, Drosophila, zebrafish and mice since the knockout or knockdown mutants were not able to survive to maturity (Grigorenko et al. 2004; Casso et al. 2005; Krawitz et al. 2005; Aizawa et al. 2016). In parallel, Arabidopsis SPP and SPPLs were identified and SPP was shown to be an essential protein localized to the ER (Tamura et al. 2008; Han et al. 2009).

SPP’s initially characterised function was in the cleavage of signal peptides left over in the ER membrane after the action of signal peptidases (Weihofen et al. 2002). In addition, SPP also contributes to the clearing of misfolded membrane proteins (Schrul et al. 2010). Along with signal peptidases, SPP contributes to the sequential processing of type II membrane proteins with large luminal domains during intramembrane proteolysis (Mentrup et al. 2020, 2024). Further, type II tail-anchored membrane proteins with short luminal domains can be directly processed by SPP (Boname et al. 2014; Mentrup et al. 2017). Lastly, SPP was shown to be involved in ER-associated protein degradation (ERAD) for quality control of misfolded proteins that have accumulated in the ER (Loureiro et al. 2006; Lee et al. 2010; Chen et al. 2014).

The role of SPP and SPPLs have been extensively studied in animal systems. However, the functions of SPPs in plants are largely unclear. The Arabidopsis SPP was shown to be essential for the development of viable pollen (Han et al. 2009). Recently, it was shows that a nodule-specific SPP of Medicago truncatula is required for symbiosis with rhizobia (Yang et al. 2023). Here, we found that partial loss-of-function mutations in Arabidopsis SPP lead to suppression of pi4kβ1,2 autoimmunity. In addition, these mutations can suppress its short root and root hair defects. This is noteworthy, since mutation in EDS1 or a gene required for SA biosynthesis only partially suppresses pi4kβ1,2’s immunity and rosette size, but not its root defects (Šašek et al. 2014; Thulasi Devendrakumar et al. 2023). In a previous study, SPP was shown to be transcriptionally upregulated in Arabidopsis plants with induced ER stress (Iwata et al. 2010). Further, SPP was also induced in response to treatment with immune elicitors, indicating a potential role of SPP in mediating immune responses (Bjornson et al. 2021; Yildiz et al. 2021). In this study we found that SPP expression is upregulated in pi4kβ1,2. While pi4kβ1,2 shows high sensitivity to ER stress and upregulation of ER stress marker genes, mutations in SPP can suppress pi4kβ1,2’s heightened ER stress in suppressors 79–1 and 171–1. The spp-4 allele does not seem to suppress the ER stress responses likely because it is a weaker loss-of-function allele. While SPP was seen to be upregulated in response to treatment with immune elicitors, the spp single mutants did not display detectable enhanced susceptibility to pathogens. Thus, it is possible that elevated ER stress is the main defect of pi4kβ1,2, which can be alleviated with reduced SPP activity in both rosette and roots. Its autoimmunity could be downstream of such ER stress, explaining the partial reliance of the autoimmunity of pi4kβ1,2 on SA or EDS1 and further explain the lack of an immune defect in the spp single mutants.

Further, it is interesting that mutations in SPP can lead to the formation of recessive partial loss of function alleles as found in suppressors 79–1 and 145–1, as well as a dominant-negative allele found in suppressor 171–1. All three suppressors carry spp alleles that result in single amino acid substitution mutations in proximity to the SPP transmembrane helices containing the active sites. However, these mutations result in varying degrees of suppression of pi4kβ1,2 phenotypes. All three suppressors show suppression of its autoimmunity and expression of the PR1 and PR2 immune marker genes. However, only suppressors 79–1 and 171–1 show suppression of the expression of ICS1, EDS5, ER stress marker genes, and the DTT induced ER stress sensitivity of pi4kβ1,2. Furthermore, only the spp-3 and spp-5 single mutants (spp alleles from suppressors 79–1 and 171–1 respectively) show reduced sensitivity to DTT induced ER stress. It is still unclear why these mutations result in SPP variants that are dominant-negative or recessive. Human SPP has been shown to function as a tetramer with their N-terminal regions being required for homo-oligomer formation with the catalytic C-termini present at the periphery of the SPP complex (Miyashita et al. 2011). SPP variants that disrupt the overall structure of the complex inhibit its catalytic activity, thereby having a dominant negative effect on SPP function (Miyashita et al. 2011). The dominant-negative allele identified in our screen could have similar mechanism. While these mutations could affect SPP’s function, it is also possible that these amino acid changes may negatively affect the stability of SPP. Further work to study the structure of the Arabidopsis SPP is required to understand the consequence of the amino acid changes on SPP structure and function.

From the phylogenetic tree shown in Fig. 6b, SPP is the most similar to the HsSPP. We further observed that HsSPP can complement the phenotype of suppressor 79–1. This suggests that HsSPP is able to target the Arabidopsis SPP substrate(s) required for pi4kβ1,2 phenotypes, supporting a highly conserved mechanism of SPP function across kingdoms.

We observed that SPP transiently expressed in Nicotiana benthamiana is localized to the ER, agreeing with a previous study (Tamura et al. 2008). However, in addition to the ER localization, we also observed puncta along the ER network. It is currently unclear what this punctate localization is and how it is relevant to SPP function. These puncta could either be membrane-bound organelles involved in the endomembrane system or nanodomains containing a high local concentration of SPP. Future co-localization experiments with other fluorescent markers can help distinguish between these possibilities and provide clues regarding the function of these puncta localized SPP. PI4Kβ1 was shown to be localized to distinct compartments of the trans Golgi network in root hair cells (Preuss et al. 2006; Kang et al. 2011). However, in these studies co-localization with marker constructs was lacking. Here we found that PI4Kβ1 transiently expressed in N. benthamiana is localised to the ER. This difference in localization could be attributed to the different tissues used in the localization studies. It is interesting that both SPP and PI4Kβ1 are ER localized. This shared localization may potentially allow for SPP to be activated by the loss of the PI4Kβs in pi4kβ1,2.

PI4Kβs are involved in the biosynthesis of phosphatidylinositol 4-phosphate (PI4P) (Preuss et al. 2006). Perturbed lipid homeostasis in the ER has been shown to result in the activation of the UPR either due to its effect on ER protein folding and transport or as a direct consequence of the altered lipid composition leading to lipid bilayer stress (Volmer & Ron 2015; Fun & Thibault 2020). Studies have also shown that ER stress, cell death, and plant immunity are intricately linked (Kørner et al. 2015; Ruberti et al. 2015; Manghwar & Li 2022). It is interesting that loss of the ER localized PI4Kβs trigger immune responses and ER stress that is dependent on another ER resident protein, SPP. It is possible that the perturbations in the levels of PI4P lead to heightened ER stress in pi4kβ1,2 and SPP functions in this ER stress response. However, the exact molecular mechanism how this ER stress contributes to pi4kβ1,2 autoimmunity remains unclear. Further work to identify the relevant substrates of SPP will help characterize its role in ER stress and immunity.

Materials and methods

Plant materials and growth conditions

All plants used in this study are in the Arabidopsis thaliana Col-0 ecotype. Col-0 is referred to as WT. The pi4kβ1,2 mutant used in this study has been previously described (Preuss et al. 2006). Col-0 eds1-2 (referred to as eds1-2 in this study) and agb1-2 (referred to as agb1 in this study) that were used as susceptible controls in infection assays were previously described (Zhang et al. 2003; Ullah et al. 2003; Bartsch et al. 2006). All other lines were generated in this study. The spp single mutants were isolated by crossing the original suppressors 79–1, 145–1, and 171–1 to WT. From the segregating F2 populations the spp single mutants, spp-3, spp-4, and spp-5, that lacked the pi4kβ1 and pi4kβ2 T-DNA insertions were isolated. The primers used for genotyping are listed in Table S1. These spp single mutants were used for the root length based DTT sensitivity assays in Fig. 7, and for the infection assays in Fig. 8.

Seeds were soaked in 15% Chlorox® Original Concentrated Bleach supplemented with 0.1% Tween-20 for five minutes for surface sterilization. The seeds were then washed three times with sterile water. The seeds were then suspended in sterile 0.1% agar and stratified by storing them at 4 °C for at least two days.

To measure root length and for the root length based DTT sensitivity assays, standard square plates with ½ Murashige and Skoog (MS) media (pH 5.7) supplemented with 0.5% sucrose, 1% agar, and the appropriate concentration of sterile DTT were used. Sterilized seeds were exposed to light for 6 h and then sown on the plates. The plates were then placed vertically in a growth chamber with 12 h light and 12 h dark conditions. For DTT sensitivity assays based on seedling survival, plates were made with ½ MS media (pH 5.7) supplemented with 0.6% agar, 0.5% sucrose and the appropriate concentration of sterile DTT.

For sowing on soil, stratified seeds were exposed to light for 4 h or overnight. They were then planted on autoclaved Sunshine® Mix #4 aggregate. The flats were covered with a plastic dome for two to three days until the seeds germinated. The dome was then removed, and the plants were grown at 22 °C and long day conditions (16 h light and 8 h dark) for most purposes. For bacterial infection assays, seedlings were grown under long day conditions for two weeks, transplanted into single pots and grown for a further two weeks under short day conditions (8 h light and 16 h darkness) for two more weeks and then used. To grow plants for Hyaloperonospora arabidopsidis (Ha) Noco2 infection assays, the seedlings were grown for 12–14 days after sowing under long day conditions. The plants were then inoculated with the Ha Noco2 spore suspension, covered with a clear dome and then grown for an additional 7 days in a growth chamber set to 18 °C and 12 h light and 12 h darkness.

For all assays, seeds of all the genotypes to be compared were harvested from plants grown at the same time under identical conditions.

EMS mutagenesis and screening M2 populations

Ethyl methanesulfonate (EMS) mutagenesis was carried out as previously described (Li & Zhang 2016). The suppressors 79–1, 145–1, and 171–1 were identified from vertical plate screen of the EMS mutagenized M2 population as described in Thulasi Devendrakumar et al. 2024.

DNA isolation, next generation sequencing and data analysis

Total genomic DNA isolation for Next Generation Sequencing was performed as described in Thulasi Devendrakumar et al. 2024, 2023. The DNA was sequenced by Novogene using the Illumina® NovaSeq™ 6000 sequencing platform. The resulting NGS raw reads were analyzed using a variant discovery pipeline based on GATK Best Practices run on the Compute Canada Cedar cluster (Huang et al. 2019; van der Auwera & O’Connor 2020).

Plasmid construction and primers

The NP::SPP construct was constructed by amplifying the SPP gene and its promoter (1012 bases upstream of SPP’s start codon) and cloning it into a promoterless binary vector. The 35S::SPP-mCitrine construct was constructed by amplifying the SPP gene without the stop codon and cloning it into a 35S promoter driven binary vector with a C-terminal mCitrine tag. The NP::HA-mNG-PI4Kβ1 construct was created by amplifying the PI4Kβ1 gene and its promoter (1633 bases upstream of PI4Kβ1’s start codon) and cloning it into a binary vector upstream of the HA-mNG. Subsequently, PI4Kβ1 gene was amplified and cloned into the vector downstream of the HA-mNG. For these three constructs, Arabidopsis WT DNA was used as template and the amplicons were amplified using NEB Q5® High-Fidelity DNA polymerase. The 35S::HsSPP was cloned by amplifying the HsSPP CDS (Isoform 1) and cloning it into a 35S promoter driven binary vector. RNA was extracted from Caco-2/TC-7 human colon adenocarcinoma cell line. The cDNA generated from this RNA was used as the template to amplify HsSPP CDS. The primers used for cloning are listed in Table S1.

Infection assays

For oomycete Ha Noco2 and Ha Emwa1 infection assays, seedlings were sprayed with the spore suspension in water (50,000 spores/ml). 7 days after inoculation, whole above ground seedling tissue was harvested and vortexed with sterile water. The suspended Ha Noco2 spores were then counted using a hemocytometer. Bacterial infection assays were performed as described in Thulasi Devendrakumar et al. 2023. In brief, Pseudomonas syringae pv. maculicola ES4326 and Pseudomonas syringae pv. tomato DC3000 hrcC− bacterial suspensions were syringe infiltrated into the leaves. Leaf disks were collected on Day 0 (day of infiltration) and Day 3 (three days after infiltration), ground in 10 mM MgCl2, serially diluted and plated on LB plates to measure the colony forming units (cfu) per cm2 of the leaf.

Root length measurement

For regular root length measurement 10-day old seedlings grown on vertical ½ MS plates were used. The seedlings were removed from the agar surface, the root was laid straight and measured manually using a ruler. The distance between the base of the hypocotyl to the root tip was recorded as the root length. For DTT sensitivity assays, the root lengths of 14-day old seedlings were measured.

Quantitative RT-PCR

30–40 mg of leaves were collected from 4-week-old soil grown plants in 2 ml tubes containing 2 glass beads and frozen in liquid nitrogen. The tissue was then ground to a fine powder using a Precellys tissue homogenizer. The total RNA was extracted using the EZ-10 Spin Column kit (BioBasic, Markham, Canada). 1 μg of RNA was then reverse transcribed into cDNA using the ABM EasyScript™ cDNA Synthesis Kit following the manufacturer’s protocol and the cDNA was used for RT-qPCR. TB Green® Premix Ex Taq™ II mastermix was used for the RT-qPCR. BioRad’s CFX Connect™ Real-Time PCR Detection System was used for the qPCR run. The primers used for the reverse transcription and qPCR are listed in Table S1.

Confocal laser scanning microscopy

To detect fluorescence of mCitrine, mNeonGreen (mNG) or Cyan Fluorescent Protein (CFP) fusion proteins confocal laser scanning microscopy was performed using a TCS SP8 Falcon system, with 63 × water immersion objective and LASX 3.5 software (Leica Microsystems, Wetzlar, Germany). The endoplasmic reticulum (ER) marker ER-ck (CFP; Nelson et al. 2007) was used for co-localization. For excitation of mCitrine, mNG, and CFP, a pulsed white light laser was used with laser lines set to 488 nm, 514 nm or 458 nm, respectively. Fluorescence emission was detected using HyD SMD detectors with detection windows ranging from 525 to 560 nm for mCitrine, 500 to 540 nm for mNG, and 465 to 485 nm for CFP. Images were recorded in sequential scanning mode to avoid bleach through of CFP signals into the mCitrine channel.

Phylogenetic analysis and tree construction

The amino acid sequences of the Arabidopsis thaliana (6 proteins) and Homo sapiens (human; 5 proteins) SPP and SPPLs were obtained from TAIR (Arabidopsis) and NCBI (Human). The full-length sequences of the proteins encoded by the most abundant splice variant of each of the genes were aligned using Muscle align in MEGA X. This alignment was then used to predict the evolutionary history of the SPP and SPPLs.

The evolutionary history was inferred using the Maximum Likelihood method and JTT matrix-based model (Jones et al. 1992). The tree with the highest log likelihood (-11827.01) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches (1000 bootstrap replicates). Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the JTT model, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. There were 844 positions in total in the final dataset. Evolutionary analyses were conducted in MEGA X (Kumar et al. 2018).