ER Quality Control Components UGGT and STT3a Are Required for Activation of Defense Responses in Bir1-1

Abstract

The receptor-like kinase SUPPRESSOR OF BIR1, 1 (SOBIR1) functions as a critical regulator in plant immunity. It is required for activation of cell death and defense responses in Arabidopsis bak1-interacting receptor-like kinase 1,1 (bir1-1) mutant plants. Here we report that the ER quality control component UDP-glucose:glycoprotein glucosyltransferase (UGGT) is required for the biogenesis of SOBIR1 and mutations in UGGT suppress the spontaneous cell death and constitutive defense responses in bir1-1. Loss of function of STT3a, which encodes a subunit of the oligosaccharyltransferase complex, also suppresses the autoimmune phenotype in bir1-1. However, it has no effect on the accumulation of SOBIR1, suggesting that additional signaling components other than SOBIR1 may be regulated by ER quality control. Our study provides clear evidence that ER quality control play critical roles in regulating defense activation in bir1-1.

Introduction

Eukaryotic cells have evolved several quality control mechanisms to monitor the folding of secretory proteins in the endoplasmic reticulum (ER) [1,2]. Correctly folded proteins are allowed to export to their final destinations, whereas misfolded proteins are retained in the ER for additional folding process or degraded by the ER-associated degradation pathway.

One of the well-studied protein folding pathway specific for secreted glycoproteins is the calnexin (CNX)/calreticulin(CRT) cycle, which involves ER-localized lectin-like chaperones CNX/CRT and the UDP-glucose:glycoprotein glucosyltransferase (UGGT) [3]. Following protein translation, preassembled glycan chains (Glc₃Man₉GlcNAc₂) are transferred to the Asn (N)-residues in the N-X-Ser/Thr sequences in acceptor proteins by the oligosaccharyltransferase (OST) complex. Trimming of two glucose residues from the glycan chain by glucosidases generates proteins with monoglucosylated glycans (GlcMan₉GlcNAc₂), which CNX and CRT interact with and assist with folding in the ER. Subsequent removal of the remaining glucose from GlcMan₉GlcNAc₂ leads to dissociation of the client protein from CNX and CRT. Proteins that attained their nature structure can then enter the secretory process, whereas improperly folded proteins are recognized by UGGT and a glucose residue is added back to the Man₉GlcNAc₂ by the enzyme. The monoglucosylated proteins subsequently associate with CNX and CRT to go through another round of folding.

UGGT and CRT3 have been shown to play important roles in the biogenesis of transmembrane receptors in plants. Retention of the defective brassinosteroid receptor bri1–9 protein in the ER requires both UGGT and CRT3 [4,5]. In Arabidopsis uggt and crt3 mutant plants, accumulation of the receptor-like kinase (RLK) EFR, which recognizes bacterial EF-Tu, is reduced [6,7]. Expression of the tobacco protein INDUCED RECEPTOR-LIKE KINASE was also shown to be dependent on NbCRT3 [8]. In tomato, silencing of CRT3a affects the biogenesis of Cf-4 and leads to loss of pathogen resistance mediated by Cf-4 [9]. In addition, loss of function mutations in Arabidopsis STT3a, which encodes the catalytic subunit of the OST complex, also cause reduced EFR protein level and impair its function in plant immunity [7,10].

In Arabidopsis, BAK1-INTERACTING RECEPTOR-LIKE KINASE 1 (BIR1) negatively regulates cell death and defense responses mediated by the RLK SOBIR1 (SUPPRESSOR OF BIR1, 1) [11]. Previous studies showed that activation of defense responses in bir1–1 is also dependent on the β and γ subunits of heterotrimeric G protein as well as several ER quality control (ER-QC) components including CRT3, ERdj3b and SDF2 [12,13]. Here we report that additional ER-QC regulators, UGGT and STT3a, also play important roles in the regulation of defense responses in bir1–1.

Results and Discussion

Identification and characterization of sobir6–1 bir1–1 pad4–1

To identify defense pathways activated in bir1–1, a suppressor screen was carried out in the bir1–1 pad4–1 mutant background [11]. sobir6–1 is one of the mutants identified from the screen. In the sobir6–1 bir1–1 pad4–1 triple mutant, the dwarf morphology of bir1–1 pad4–1 is almost completely suppressed (Fig. 1A). Analysis of the expression levels of defense marker genes PATHOGENESIS-RELATED 1 (PR1) (Fig. 1B) and PR2 (Fig. 1C) in sobir6–1 bir1–1 pad4–1 showed that PR2 expression is significantly lower than in bir1–1 pad4–1. In addition, sobir6–1 bir1–1 pad4–1 supports much higher growth of the oomycete pathogen Hyaloperonospora arabidopsidis (H.a.) Noco2 than bir1–1 pad4–1 (Fig. 1D). These data indicate that the constitutive defense responses observed in bir1–1 pad4–1 is largely suppressed by sobir6–1.

Fig 1. Characterization of the sobir6–1 bir1–1 pad4–1 triple mutant.

(A) Morphology of Col-0 wild type (WT), bir1–1, bir1–1 pad4–1 and sobir6–1 bir1–1 pad4–1. Plants were grown on soil at 23°C and photographed three weeks after planting. (B-C) Expression levels of PR1 (B) and PR2 (C) in WT, bir1–1, bir1–1 pad4–1 and sobir6–1 bir1–1 pad4–1 seedlings compared to ACTIN1. Total RNA was extracted from 12-day-old seedlings grown on half-strength MS plates. (D) Growth of H. a. Noco2 on WT, bir1–1, bir1–1 pad4–1 and sobir6–1 bir1–1 pad4–1 seedlings. Error bars in (B-D) represent standard deviations of three measurements.

SOBIR6 encodes UGGT

The sobir6–1 mutation was mapped to a region between marker T2E12 and T8K14 on Chromosome 1 using a mapping population generated by crossing sobir6–1 bir1–1 pad4–1 (in the Columbia ecotype background) with Landsberg. Further fine mapping analysis narrowed the mutation to a 70 kb region between markers F23N20 and F26A9 (Fig. 2A). To identify the sobir6–1 mutation, PCR fragments covering this region were amplified from genomic DNA of sobir6–1 bir1–1 pad4–1 and sequenced. A single G to A mutation was identified in AT1G71220, which encodes the ER-QC component UGGT. The mutation is located at the junction between the 29^th intron and 30^th exon of UGGT.

Fig 2. Map- based cloning of SOBIR6.

(A) Mapping of the sobir6–1 mutation. Positions of the mapping markers, the gene structure of SOBIR6 and the mutation site in sobir6–1 are shown. The exons are indicated with boxes and introns with lines. The mutation site is located at the junction between the 29th intron and 30th exon. The lower case letters represent nucleotides in the intron and the uppercase letters represent nucleotides in the exon. (B) Morphology of sobir6 bir1–1 pad4–1 alleles and representative F1 plants of indicated crosses for the complementation test. Plants were grown on soil at 23°C and photographed three weeks after planting. (C) Mutations identified in the sobir6 alleles. aa, amino acid. ¹The positions of mutated nucleotide in the coding sequence are listed.

In the same mutant screen, we also identified two additional alleles of sobir6. Both of them failed to complement sobir6–1 (Fig. 2B). Sequence analysis of UGGT in sobir6–2 and sobir6–3 showed that they also contain mutations in the gene. In sobir6–2, a C to T mutation changes Ala₁₄₂₆ to Val. In sobir6–3, a G to A mutation introduces a stop codon in the coding region (Fig. 2C). These data suggest that SOBIR6 is UGGT.

UGGT is required for constitutive defense responses in bir1–1

To test whether sobir6–1 can suppress the constitutive defense responses in bir1–1 in the absence of pad4–1, we isolated the sobir6–1 bir1–1 double mutant from the F2 population of a cross between sobir6–1 bir1–1 pad4–1 and wild type. sobir6–1 bir1–1 is much bigger than bir1–1, but smaller than wild type (Fig. 3A). In sobir6–1 bir1–1, expression of both PR1 and PR2 is greatly reduced compared to that in bir1–1 (Fig. 3B-C). As shown in Fig. 3D, resistance to H.a. Noco2 is also considerably reduced in sobir6–1 bir1–1. These data suggest that UGGT is required for the constitutive defense responses in bir1–1. This is consistent with the requirement of another component of the CNX/CRT cycle, CRT3, for the autoimmune phenotype in bir1–1 [13].

Fig 3. Characterization of the sobir6–1 bir1–1 double mutant.

(A) Morphology of wild type (WT), bir1–1 and sobir6–1bir1–1 plants. Plants were grown on soil at 23°C and photographed three weeks after planting. (B-C) Expression levels of PR1 (B) and PR2 (C) in WT, bir1–1 and sobir6–1 bir1–1 seedlings as normalized with ACTIN1. Total RNA was extracted from 12-day-old seedlings grown on half-strength MS plates. (D) Growth of H. a. Noco2 on WT, bir1–1 and sobir6–1 bir1–1 seedlings. Error bars in (B-D) represent standard deviations of three measurements.

The autoimmune phenotype of bir1–1 is partially suppressed by stt3a-2

Since STT3a is involved in co-translational N-glycosylation of nascent proteins before they enter the CNX/CRT cycle [3], we tested whether STT3a is required for the constitutive defense responses in bir1–1 by crossing stt3a-2 with bir1–1 pad4–1 and isolating the stt3a-2 bir1–1 pad4–1 triple mutant and the stt3a-2 bir1–1 double mutant in the F2 generation. As shown in Fig. 4A, stt3a-2 bir1–1 pad4–1 is larger than bir1–1 pad4–1, but considerably smaller than wild type. In stt3a-2 bir1–1 pad4–1, the expression of both PR1 and PR2 is lower than in bir1–1 pad4–1 (Fig. 4B-C). H.a. Noco2 growth on stt3a-2 bir1–1 pad4–1 is much higher than on bir1–1 pad4–1, but significantly lower than on wild type (Fig. 4D). The stt3a-2 bir1–1 double mutant retained the dwarf morphology of bir1–1, but is larger in size (Fig. 5A). In stt3a-2 bir1–1, the expression of both PR1 and PR2 is greatly reduced compared to that in bir1–1 (Fig. 5B-C). There is a small amount of H.a. Noco2 growth on the stt3a bir1–1 double mutant compared to almost no growth of the pathogen on bir1–1 plants (Fig. 5D). Taken together, the autoimmune phenotype of bir1–1 is partially dependent on STT3a. Our data suggest that STT3a-dependent N-glycosylation is also critical for activation of defense responses in bir1–1.

Fig 4. Characterization of the stt3a-2 bir1–1 pad4–1 triple mutant.

(A) Morphology of wild type (WT), bir1–1, bir1–1 pad4–1 and stt3a-2 bir1–1 pad4–1 plants. Plants were grown on soil at 23°C and photographed 3 weeks after planting. (B-C) Expression levels of PR1 (B) and PR2 (C) in WT, bir1–1, bir1–1 pad4–1 and stt3a-2 bir1–1 pad4–1 seedlings as normalized with ACTIN1. Total RNA was extracted from 12-day-old seedlings grown on half-strength MS plates. (D) Growth of H. a. Noco2 on WT, bir1–1, bir1–1 pad4–1 and stt3a-2 bir1–1 pad4–1 seedlings. Error bars in (B-D) represent standard deviations of three measurements.

Fig 5. Characterization of the stt3a-2 bir1–1 double mutant.

(A) Morphology of wild type (WT), bir1–1 and stt3a-2 bir1–1. Plants were grown on soil at 23°C and photographed three weeks after planting. (B-C) Expression levels of PR1 (B) and PR2 (C) in WT, bir1–1 and stt3a-2 bir1–1 seedlings normalized by ACTIN1. Total RNA was extracted from 12-day-old seedlings grown on half-strength MS plates. (D) Growth of H. a. Noco2 on WT, bir1–1 and stt3a-2 bir1–1 seedlings. Error bars in (B-D) represent standard deviations of three measurements.

sobir6–1 affects the protein level of SOBIR1

Because the constitutive defense responses in bir1–1 are dependent on the RLK SOBIR1 and the accumulation of SOBIR1 is dependent on the ER-QC component CRT3 [13], we further tested whether UGGT and STT3a are also required for SOBIR1 accumulation. A transgenic line expressing the SOBIR1-FLAG fusion protein under its own promoter in wild type background was crossed into sobir6–1 or stt3a-2, a T-DNA knockout mutant of STT3a As shown in Fig. 6A, the SOBIR1-FLAG protein level is considerably lower in sobir6–1 than in wild type background, suggesting that UGGT is also required for the accumulation of SOBIR1-FLAG protein. In contrast, the SOBIR1-FLAG protein levels are similar in stt3a-2 and wild type background.

Fig 6. Accumulation of SOBIR1 protein in sobir6–1 and stt3a-2.

(A) Western blot analysis of SOBIR1-FLAG protein level (top) and RT-PCR analysis of SOBIR1-FLAG expression (bottom) in wild type (WT) and sobir6–1 mutant background. (B)Western blot analysis of SOBIR1-FLAG protein level (top) and RT-PCR analysis of SOBIR1-FLAG expression (bottom) in WT and stt3a-2 mutant background. Protein and RNA samples were extracted from 12-day-old seedlings grown on half-strength MS plates at 23°C. Rubisco was used as protein loading control and ACTIN1 was used as RNA control.

In addition to its role in cell death and defense activation in bir1–1, increasing evidences suggest that SOBIR1 functions as a critical component of receptor-like protein (RLP)-mediated immunity [14,15]. SOBIR1 proteins in tomato interact with two RLPs Cf-4 and Ve1 and are required for Cf-4 and Ve1 mediated immunity. In addition, SOBIR1 functions together with Arabidopsis RLP30 in defense against necrotrophic fungi [16].

Our study provided additional evidence that ER-QC plays important roles in the biogenesis of SOBIR1 and reduced accumulation of SOBIR1 contributes to the suppression of bir1–1 mutant phenotypes by mutations in UGGT. Because biogenesis of SOBIR1 in Arabidopsis is dependent on multiple components of ER-QC, it is likely that accumulation of SOBIR1 proteins in tomato also relies on ER-QC. The compromised Cf-4 and Ve1-mediated immune responses observed in tomato plants when CRT3a was silenced [9,17] might be partially due to reduced accumulation of tomato SOBIR1.

Compared to almost complete suppression of the autoimmune phenotype in bir1–1 pad4–1 by uggt and crt3 mutants, stt3a-2 has a much smaller effect on the morphology as well as defense responses in bir1–1 pad4–1. This can probably be explained by genetic redundancy. In Arabidopsis, there is a close homolog of STT3a named STT3b [18]. It is likely that STT3b can partially compensate the loss of the function of STT3a in N-glycosylation.

As SOBIR1 accumulation is not affected in stt3a-2, the mechanism of how stt3a-2 suppresses the phenotypes of bir1–1 remains to be determined. It is possible that the contribution of STT3a to the biogenesis of SOBIR1 is masked by genetic redundancy between STT3a and STT3b. Since SOBIR1 usually functions together with RLPs, it is likely that one or more RLPs might be involved in the activation of cell death and defense responses in bir1–1 and the suppression of bir1–1 mutant phenotypes by stt3a-2 might be caused by reduced accumulation of the RLPs. Previously UGGT and STT3a were also shown to be required for SA-induced defense responses [7]. It is possible that reduced response to SA also contributes to the suppression of bir1–1 mutant phenotypes by stt3a-2.

Method

Plant material

bir1–1, bir1–1 pad4–1 and the identification of suppressor mutants of bir1–1 pad4–1 have been described previously [11]. stt3a-2 (SALK_058814) was obtained from the Arabidopsis Biological Resource Center and has been described previously [18]. All plants were grown at 23°C under 16h light/8h dark. To isolate the sobir6–1 single mutant and sobir6–1 bir1–1 double mutant, sobir6–1 bir1–1 pad4–1 was crossed with a Col-0 plant. In F2, sobir6–1 single mutant and sobir6–1 bir1–1 double mutant were isolated by PCR genotyping. To generate stt3a-2 bir1–1 double mutant and stt3a-2 bir1–1 pad4–1 triple mutant, stt3a-2 was crossed with bir1–1 pad4–1. stt3a-2 bir1–1 and stt3a-2 bir1–1 pad4–1 were identified in F2 by PCR genotyping.

Mutant characterization

H. a. Noco2 infection was performed on 12-day-old seedlings. The seedlings were sprayed with spore suspension at a concentration of 50,000 spores per ml water. Sprayed plants were covered with a clear dome and kept at 16°C under 12h light/12h dark cycles in a growth chamber. The humility in the growth chamber was approximately 95%. Infection results were scored seven days later as previously described [19].

For gene expression analysis, RNA was extracted from 12-day-old seedlings grown on half-strength MS plates using EZ-10 Spin Column Plant RNA Mini-Preps Kit from Bio Basic Inc. About six seedlings were collected and extracted in each sample. The extracted RNA was reverse transcribed into total cDNA using Easy Script Reverse Transcriptase from Applied Biological Materials Inc. Real- time PCR was performed in triplicate with three independent RNA samples using SYBR Premix Ex Taq IIfrom Takara. Total cDNA was used as a template to determine the expression level of target genes with ACTIN1 as control. Primers used for real-time PCR analysis of ACTN1, PR1 and PR2 have been described previously [20].

Analysis of SOBIR1 protein level in sobir6–1 and stt3a-2

A transgenic line expressing the SOBIR1 protein with a 3xFLAG tag [13] was crossed with sobir6–1 and stt3a-2. In F2, plants that were homozygous for sobir6–1 and stt3a-2 and carried the SOBIR1-FLAG transgene were identified by PCR. For Western blot analysis, about 20 seedlings from half-strength MS plate were collected in liquid nitrogen for each sample. The samples were ground and boiled in 2×SDS gel-loading buffer (100mM Tris-Cl pH6.8, 4% w/v sodium dodecyl sulfate, 0.2% bromophenol blue, 20% w/v glycerol, 200mM DTT). Supernatants were subject to Western blot analysis using the anti-flag M2 antibody (Sigma-Aldrich).

Supporting Information

S1 Table. Sequences of markers used in this project.

(PDF)

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

Natural Sciences and Engineering Research Council of Canada (www.nserc-crsng.gc.ca/index_eng.asp) provided funding for this project. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.