Plant and prokaryotic TIR domains generate distinct cyclic ADPR NADase products

Abstract

Toll/interleukin-1 receptor (TIR) domain proteins function in cell death and immunity. In plants and bacteria, TIR domains are often enzymes that produce isomers of cyclic adenosine 5'-diphosphate-ribose (cADPR) as putative immune signaling molecules. The identity and functional conservation of cADPR isomer signals is unclear. A previous report found that a plant TIR could cross-activate the prokaryotic Thoeris TIR-immune system, suggesting the conservation of plant and prokaryotic TIR-immune signals. Here, we generate autoactive Thoeris TIRs and test the converse hypothesis: Do prokaryotic Thoeris TIRs also cross-activate plant TIR immunity? Using in planta and in vitro assays, we find that Thoeris and plant TIRs generate overlapping sets of cADPR isomers and further clarify how plant and Thoeris TIRs activate the Thoeris system via producing 3'cADPR. This study demonstrates that the TIR signaling requirements for plant and prokaryotic immune systems are distinct and that TIRs across kingdoms generate a diversity of small-molecule products.

Fig. 1.. Prokaryotic TIRs can trigger in planta cytotoxicity independently of EDS1.

(A) Schematic of prokaryotic Thoeris immunity and plant TIR–immune signaling. In both systems, TIRs sense pathogens and signal immune outputs via NADase activities. Plant TIRs generate multiple metabolites, some of which are known to activate EDS1-dependent outputs (pRib-AMP/ADP and ADPr-ATP/ADP), while potential EDS1-independent outputs remain unknown. BdTIR stimulates both EDS1 and Thoeris outputs; does ThsB stimulate EDS1 outputs (see dashed arrow)? (B and C) Nb WT or eds1^−/− leaves shown approximately 6 dpi with constructs delivering AbTIR, HA-BdTIR, ThsB, or EV [empty vector control; 35S: green fluorescent protein (GFP)]. E/A refers to TIRs containing alanine substitutions at the conserved catalytic glutamate (E) residue. Framed numbers denote leaf replicates per set. (D) Anti-HA immunoblot of N-HA–tagged BdTIR, AbTIR, ThsB, or TcpO-TIR proteins harvested from Nb eds1^−/− leaves at ~40 hours postinfiltration (hpi). (E) Fluorescent NAD⁺ detection assay in Nb eds1^−/− leaves expressing different TIR constructs. NAD⁺ assays were performed at 40 hpi. RFU, relative fluorescence units. NRG1-CC is the CC domain of the CC-NLR protein, NRG1. AbTIR (core TIR from AbTir) residues 157 to 292, TcpO-TIR residues 204 to 341, and SAM_SARM1-TIR residues 478 to 724. Catalytic glutamate (E residue) mutants of TIRs were outlined in dashed gray boxes; AbTIR (E208), BdTIR (E127), and TcpO-TIR (E279). Similar experiments were performed at least three times. Statistical analyses: One-way analysis of variance (ANOVA) and Turkey honestly significant difference (HSD) with CLD (compact letter display) of significance classes. Overlapping letters are ns (nonsignificant) difference (P > 0.05), while separate letter class indicates P < 0.05 or better.

Fig. 2.. Mutagenesis of a ThsB loop region promotes NADase autoactivity, cADPR-isomer production, and stimulation of ThsA-mediated cytotoxicity.

(A) Left: Cryo-EM structure of activated-state TIR domain from plant TIR-NLR, RPP1 [Protein Data Bank (PDB): 7CRC]. Center: Crystal structure ThsB (PDB: 6LHY). Right: Overlay of activated RPP1-TIR with ThsB. The RPP1 BB-loop is colored gray, ThsB loop region is shown in purple, and catalytic glutamates are shown in orange. N, N terminus. (B) Nb eds1^−/− leaf ~6 days postagroinfiltration (dpi) with constructs coexpressing ThsA with WT ThsB, ThsB-mutants, or EV (35S: GFP). ThsA and ThsB were coexpressed at an individual optical density (OD) of 0.40. Positive and negative control SARM1-TIR or EV, respectively, expressed individually at an OD of 0.80. Framed numbers denote leaf replicates per set. (C) Anti-HA immunoblot of N-HA–tagged ThsB variants were transiently expressed in Nb eds1^−/− leaves and harvested ~40 hpi. (D and E) In vitro NAD⁺ consumption by recombinant ThsB-Auto or ThsB-Auto E85Q (catalytic mutant) and detection of cADPR-isomer via LC-MS. (F) In vitro ThsA NADase stimulation by recombinant ThsB-Auto or ThsB-Auto E85Q (catalytic mutant) or glutathione S-transferase (GST)–laden beads alone. (G) LC-MS traces for cADPR isomers (MW 542) in Nb eds1^−/− leaves transiently expressing ThsB-Auto, ThsB-Auto E85Q, or WT ThsB; leaves sampled ~40 hpi. Arrow denotes ThsB-Auto–produced cADPR isomers. Similar experiments were performed at least three times. Statistical analyses: One-way ANOVA and Turkey HSD. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.0005.

Fig. 3.. ThsB-Auto requires TIR-NADase functions to stimulate ThsA-mediated NAD⁺ depletion and cytotoxicity.

(A) Nb eds1^−/− leaf ~5 dpi with constructs expressing WT ThsA or ThsB, ThsA, and ThsB variants, or SARM1-TIR or EV (35S: GFP) controls. ThsA N112A lacks SIR2-type NADase activity; ThsA R371A has an altered SLOG-motif, and ThsB E85Q lacks TIR domain catalytic activity. Constructs expressed at an OD of 0.80. Framed numbers denote leaf replicates per set. (B) Fluorescent NAD⁺ detection assay in Nb eds1^−/− leaves expressing various ThsA or ThsB constructs. NADase assays were performed at 40 hpi. (C) Nb eds1^−/− leaf coexpressing ThsA or ThsA N112A with ThsB-Auto, ThsB-Auto E/Q, or EV, shown ~6 dpi. ThsA and ThsB coexpressed at an OD of 0.40; positive and negative control SARM1-TIR and EV expressed at an OD of 0.80. Catalytic TIR (E/A) mutants outlined in dashedgray; ThsA N112A outlined in dashedblack. (D) Fluorescent NAD⁺ detection assay in Nb eds1^−/− leaves coexpressing ThsA and ThsB combinations. NAD⁺ assays performed at 40 hpi. Catalytic glutamate mutants are outlined in dashedgray; inactive ThsA N112A is outlined in dashedblack. (E) Ion leakage assay in Nb eds1^−/− leaves coexpressing ThsA and ThsB combinations. Leaf discs were collected ~72 hpi, and statistical analyses were performed for final time point. Similar experiments were performed at least three times. Statistical analyses: One-way ANOVA and Turkey HSD with CLD of significance classes. Overlapping letters are ns difference (P > 0.05), while separate letter class indicates P < 0.05 or better.

Fig. 4.. ThsB-Auto does not stimulate EDS1-mediated HR.

BdTIR does stimulate ThsA, but the activated TIR-NLR, RPP1, does not. (A) Nb WT leaf ~5 dpi with constructs expressing HA-BdTIR, ThsB WT, ThsB-Auto, or the previously described ThsB C-terminal mutations. All constructs, including negative control EV, were expressed at an OD of 0.80. Framed numbers denote leaf replicates per set. (B) Nb eds1^−/− leaf ~5 dpi with constructs coexpressing ThsA or ThsA N112A with HA-BdTIR or BdTIR E/A. ThsA with ThsB-Auto and EV controls were also included. All coexpressions contained an OD of 0.40 of each construct, while SARM1-TIR and EV positive and negative controls were at an OD of 0.80. BdTIR E/A lacks TIR-domain catalytic residue. (C) Fluorescent NAD⁺ detection assay in Nb eds1^−/− leaves coexpressing ThsA, HA-BdTIR, or ThsB-Auto combinations, or positive and negative control SARM1-TIR and EV. NAD⁺ assays were performed 40 hpi. (D) Ion leakage assay in Nb eds1^−/− leaves coexpressing different ThsA and ThsB combinations. Leaf discs were collected ~72 hpi, and measurements were recorded every 24 hours for 3 days. Statistical analyses were performed for final time point. (E and F) Nb WT or eds1^−/− leaves ~5 dpi with constructs coexpressing ThsA, ThsB-Auto, or RPP1 TIR-NLR with cognate ATR1 effector. Asterisk denotes RPP1_WsB allele with activating ATR1-Emoy effector, while dagger has nonactivating ATR1-Emwa. ThsA was coexpressed at an OD of 0.40, and RPP1/ATR1 was expressed at an OD of 0.20 each. (G) Fluorescent NAD⁺ detection assay in Nb eds1^−/− leaves coexpressing noted ThsA, RPP1/ATR1, or ThsB-Auto combinations, as well as positive and negative control SARM1-TIR and EV. NAD⁺ assays performed 40 hpi. Similar experiments were performed at least three times. Statistical analyses: One-way ANOVA and Turkey HSD. Overlapping letters are ns difference (P > 0.05), while separate letter class indicates P < 0.05 or better.

Fig. 5.. ThsB-Auto and BdTIR produce different cADPR isomers as major products; ThsA is highly stimulated by 3’cADPR.

(A) LC-MS traces of cADPR isomers from in vitro NADase reactions of ThsB-Auto, HopAM1, AbTIR, or BdTIR, relative to cADPR standard. (B and C) In vitro production of 3’cADPR by ThsB-Auto relative to ThsB-Auto E85Q, HopAM1, AbTIR, or BdTIR. (***P < 0.005, t-test) (D and E) Fluorescent NAD⁺ detection assay in Nb eds1^−/− as previously described, except with AbTIR or HopAM1 coexpressed to stimulate ThsA. NAD⁺ assays were performed at ~40 hpi. Similar experiments were performed at least three times. (F) Model of Thoeris and EDS1-pathway stimulation by signals from plant and prokaryotic TIRs. Statistical analyses: One-way ANOVA and Turkey HSD. Overlapping letters are ns difference (P > 0.05), while separate letter class indicates P < 0.05 or better.