Glycosidase and glycan polymorphism control hydrolytic release of immunogenic flagellin peptides

INTRODUCTION Immunogenic flagellin fragments are a signature of bacterial invasion in both plants and animals. Plants recognize flagellin fragments via flagellin sensitive 2 (FLS2), a model receptor kinase that is highly conserved among angiosperms. However, little is known about events upstream of flagellin perception by FLS2. The flagellin fragments recognized by FLS2 are buried in the flagellin polymer structure and require hydrolytic release before recognition can occur, yet the hydrolases releasing these elicitors remain to be identified. Uncovering their identity is a daunting task because the extracellular space of plants (the apoplast) contains hundreds of uncharacterized glycosidases and proteases. RATIONALE We reasoned that pathogenic bacteria would suppress plant hydrolases that are important for immunity. To identify suppressed hydrolases in the apoplast of infected plants, we applied activity-based protein profiling with the use of chemical probes that irreversibly label the active site of hydrolases. We applied this strategy to study the infection of the tobacco relative Nicotiana benthamiana with the bacterial pathogens Pseudomonas syringae pv. tabaci ( Pta 6605), P. syringae pv. syringae ( Psy B728a), and a virulent mutant of P. syringae pv. tomato [ Pto DC3000(Δ hQ )]. RESULTS Glycosidase activity profiling of apoplastic fluids isolated from Pto DC3000(Δ hQ )-infected plants revealed that the activity of β-galactosidase 1 (BGAL1) is suppressed in the apoplast during infection. BGAL1 suppression is caused by a heat-stable, basic, small inhibitor molecule that is produced by the bacteria under the control of hrpR/S/L virulence regulators. Null mutants of N. benthamiana lacking BGAL1 generated by genome editing have substantially reduced apoplastic β-galactosidase activity and are more susceptible to Pto DC3000(Δ hQ ), demonstrating that BGAL1 contributes to immunity. When investigating how BGAL1 functions in immunity, we discovered that treatment of Pto DC3000(Δ hQ ) and Pta 6605 bacteria with apoplastic fluids containing BGAL1 results in the release of an elicitor that triggers a burst of reactive oxygen species in leaf discs, a signature immune response in plants. The released elicitor is flagellin derived because the triggered immune response requires both the FLS2 receptor in the plant and the flagellin-encoding fliC gene in the bacteria. More precisely, treatment of purified flagella with apoplastic fluids containing BGAL1 facilitates the release of immunogenic peptides from flagellin. The flagellin polymer of both Pto DC3000(Δ hQ ) and Pta 6605 is O-glycosylated with glycans consisting of several rhamnose residues and a terminal modified viosamine (mVio). Mutant Pta 6605 bacteria carrying nonglycosylated flagellin, or carrying rhamnosylated flagellin lacking mVio, trigger the plant immune response when treated with apoplastic fluids, irrespective of BGAL1 presence, thus demonstrating that BGAL1 requires mVio for its function in immunity. Addition of a protease inhibitor cocktail to apoplastic fluids blocks the release of the flagellin elicitor from nonglycosylated flagellin, implicating apoplastic proteases in elicitor release acting downstream of BGAL1. Consistent with a specific role of BGAL1 in elicitor release, bgal1 null mutants of N. benthamiana show increased susceptibility only to bacterial strains carrying mVio. Treatment of Psy B728a with apoplastic fluids containing BGAL1 does not facilitate release of the flagellin elicitor because its flagellin carries a different glycan moiety lacking mVio, thus providing protection against recognition. CONCLUSION Glycosidase BGAL1 acts upstream of proteases in the apoplast of N. benthamiana to release immunogenic peptides from glycosylated flagellin, but only on glycosylated flagellin containing mVio. P. syringae strains use both BGAL1 inhibitors and glycan polymorphism to suppress BGAL1 function and escape recognition. Glycan polymorphism is common to bacterial pathogens, indicating a general role for flagellin glycans in evading recognition of bacterial pathogens by both plants and animals.