Glycan moiety of flagellin in Acidovorax avenae K1 prevents the recognition by rice that causes the induction of immune responses.

Abstract Recognition of pathogen-associated molecular patterns (PAMPs) such as flagellin, a main component of the bacterial flagellum, constitutes the first layer of plant immunity and is referred to as PAMP-triggered immunity (PTI). The rice avirulent N1141 strain of gram-negative phytopathogenic bacterium, Acidovorax avenae, induces PTI including H2O2 generation, while flagellin from the rice virulent K1 strain of A. avenae does not induce these immune responses. Mass spectrometry analyses revealed that total 1,600-Da and 2,150-Da of glycan residues were present on the flagellins from N1141 and K1, respectively. A deglycosylated K1 flagellin induced immune responses in the same manner as N1141 flagellin, suggesting that the glycan in K1 flagellin prevent epitope recognition in rice. We identified three genes in K1 flagella operon, which regulate structural modification of glycan in K1 flagellin. The immature glycan-attached flagellin from three genes deletion mutant, KΔ3FG, induced H2O2 generation in cultured rice cells, whereas the K1 mature-type flagellin did not cause a detectable increase in H2O2. The data indicate that the immature glycan of flagellin from KΔ3FG cannot prevent the epitope recognition in rice.

Specific Induction of PTI by Flagellin of A. avenae

Plants are continuously confronted with diverse potential pathogens, but actual infection occurs only in certain limited cases. Plant can recognize pathogens via two perception systems. In the first layer, plant recognize pathogen-associated molecular patterns (PAMPs) through pattern-recognition receptors, leading to PAMP-triggered immunity (PTI). The second layer immune system, termed effector-triggered immunity (ETI), is caused by recognition of effector proteins, which are injected into the host cell from bacteria cell through the Type III secretion system (T3SS). By definition, PAPMs are conserved across a wide range of microbes, which may or may not be pathogenic. Because these molecules are essential for viability or lifestyle, microbes are less likely to evade host immunity through mutation or deletion of PAMPs, compared with virulence effectors. PAMPs include structures characteristic from pathogens such as β-glucan, polysaccharide chitin, ergosterol, lipopolysaccharides (LPS), flagellin and elongation factor-Tu. Among these PAMPs, flagellin, a main component of the bacterial flagellum, has been the most extensively studied in regards to the recognition mechanism and signal transduction., We previously reported that flagellin from the rice avirulent N1141 strain of gram-negative phytopathogenic bacterium, Acidovorax avenae, induces PTI including H2O2 generation, while flagellin from the rice virulent K1 strain of A. avenae does not induce. Because recombinant flagellins of N1141 and K1 equally induced H2O2 generation in cultured rice cells, we examined posttranslational modification of flagellins from N1141 and K1. Mass spectrometry analysis revealed that N1141 flagellin and K1 flagellin had total 1,600-Da and 2,150-Da of glycan residues, respectively. Therefore, we next clarified whether the glycosylation of flagellins in A. avenae affects the induction specificity of the immune response in rice cells. Both deglycosylated flagellins derived from flagellin glycosyltransferase gene deficient mutants of N1141 and K1 induced H2O2 generation in cultured rice cells. Furthermore, we identified four glycosylated-amino acid residues (178Ser, 183Ser, 212Ser and 351Thr) in K1 flagellin. In the mutants of Ala-substituted flagellins at four glycosylated amino acid position (178Ser/Ala, 183Ser/Ala, 212Ser/Ala and 351Thr/Ala), 178Ser/Ala and 183Ser/Ala K1 flagellins induced the immune response in cultured rice cells, indicating that the glycans at 178Ser and 183Ser in K1 flagellins prevent epitope recognition in rice. Interestingly, mass spectrometry analysis using flagellins from the NΔFlaA-KFlaA (K1 flagellin expression vector-possessing NΔFlaA, N1141 flagellin deficient strain), KΔFlaA-NFlaA (N1141 flagellin expression vector-possessing KΔFlaA, K1 flagellin deficient strain) revealed that the molecular weight of each glycan chain in K1 flagellin is predicted to be 540, while the molecular weight of each glycan chain in N1141 flagellin is estimated to be 550. The K-type flagellin purified from the NΔFlaA-KFlaA strain induced immune responses such as H2O2 generation, while the flagellin from the KΔFlaA-NFlaA strain did not induce H2O2 to the same degree as the K1 wild-type flagellin. These data clearly indicate that the glycan moiety linked by the K1 glycosyltransferase disrupts flagellin recognition by rice that causes the induction of immune responses.

Identification of Genes that Regulate Glycan Modification of K1

To clarify the prevention mechanism of the epitope site by glycan of K1 flagellin, the glycosylation island, a genetic region required for flagellin glycosylation, within the flagellar gene operon of A. avenae K1 were determined with DNA sequencing. The glycosylation island of flagellin in A. avenae K1 consists of four orfs: designated Fgt, Fmt, Fcs, and Fst. Fgt encodes putative glycosytransferase and the Fgt-deletion K1 strain produced deglycosylated flagellin. The proteins encoded by Fmt, Fcs, and Fst had predicted molecular masses of 27,013-, 50,796- and 41,817-Da, respectively, and showed homology to type 12 methyltransferase of Pseudomonas putida W619, carbamoylphosphate synthase of Bacillus pumilus SAFR-032, and sugartransaminase of P. fulva 12-X ().

Structures of the flagellin glycosylation island in the A. avenae K1. The predicted methyltransferase gene (Fmt) are indicated by hatched pentagon, the predicted carbamoylphosphate synthase gene (Fcs) are indicated by gray pentagon, and the predicted sugartransaminase gene (Fst) are indicated by black pentagon.

To examine whether the Fmt, Fcs, and Fst genes are responsible for the specific recognition by rice, we generated the three genes-deletion mutant using homologous recombination and designated KΔ3FG. To determine the glycan structure in the flagellin of KΔ3FG strain, MALDI-TOF MS analysis was performed. The mass spectrum of the K1 wild-type flagellin showed that the molecular mass of the mature-type K1 flagellin is 51,225, which is greater than the calculated molecular mass by approximately 2,150 (). Mass spectrometry analysis also showed that the molecular mass of the KΔ3FG flagellin is 50,495, which are also greater than the calculated masses by approximately 1,420 (). The mass spectrum data revealed that glycan of KΔ3FG are smaller than that of glycan from K1 flagellin. Three genes, Fmt, Fcs, and Fst within flagellin glycosylation island of K1 are involved in structural modification of glycan of K1 strain.

MALDI-TOF MS analysis of intact flagellins of K1 wild type (left) and KΔ3FG (right). Mass spectrometry analysis showed that molecular weights of flagellins purified from the K1 wild type and KΔ3FG are 51,225 and 50,495, respectively.

It previously reported that flagellin from Pseudomonas syringae pv. tabaci 6605 (Pta 6605) was attached glycan chains composed of trisaccharide (modified viosamine (mVio)-rhamnose-rhamnose). Moreover, vioA (dTDP-viosamine aminotransferase), vioB (dTDP-viosamine acetyltransferase), vioM (methyltransferase), vioR (3-oxoacyl-(acyl-carrier protein) reductase), vioS (3-oxoacyl-(acyl-carrier protein) synthase III) and acp (acyl-carrier protein) genes were identified as biosynthetic genes of mVio. Flagellin from ΔvioR mutant of Pta 6605 was attached with rhamnosyl glycans without modified viosamine and flagellin from ΔvioM of Pta 6605 was attached with demethylated mVio-rhamnose-rhamnose. Several common properties including predicted function and sequence homology were observed between two genes (vioR and vioM) of Pta 6605 and three genes (Fmt, Fcs, and Fst of A. avenae K1 strain). These suggest the possibility that three genes (Fmt, Fcs, and Fst of A. avenae K1 strain) are involved in modification of the glycan. In addition, molecular weight of each glycan chain attached on flagellin from KΔ3FG was 730 smaller than that of glycan chain attached on mature flagellin from K1 strain. These data together with the properties of Fmt, Fcs, and Fst genes indicate that Fmt, Fcs, and Fst may be involved in modification of the non-reducing terminal group of glycan chain. Determination of glycan chains attached on flagellin of KΔ3FG will be helpful to confirm the possibility about function of Fmt, Fcs, and Fst genes.

Prevention of Epitope Recognition by the Immature Glycan of K1 Flagellin

To clarify role of Fmt, Fcs, and Fst genes, the motility of KΔ3FG was examined based on a swimming assay on soft agar plates. KΔ3FG strain had a diffuse spreading growth pattern that is characteristic of motile bacteria and was similar to the parental strains, suggesting that the deletion of Fmt, Fcs, and Fst genes dose not affect the swimming motility (). We next examined whether structural modification of K1 flagellin glycan by deletion of Fmt, Fcs, and Fst genes affects the specific induction of PTI in rice. When the cultured rice cells were treated with the immature glycan-attached flagellin from KΔ3FG, H2O2 was rapidly generated, whereas the K1 mature-type flagellin did not cause a detectable increase in H2O2 until 3 h after treatment (). These results clearly indicate that the immature glycan of flagellin from KΔ3FG cannot prevent the epitope recognition in rice. An identification of the glycan structure attached with flagellin in the A. avenae K1 will be important to further understand the specific recognition mechanism of flagellin by rice.

Role of glycan modification in K1 flagellin. (A) Swimming motility of K1wild type and KΔ3FG strains of A. avenae. Bacterial cell densities were adjusted to an OD610 of 1.0 and 2 μl aliquots were inoculated on LB soft agar plates. Photographs were taken after 24 h at 30°C. (B) Time course of H2O2 generation in cultured rice cells that were treated with flagellin purified from the K1 wild type (solid squares) or KΔ3FG (open circles) strains. The error bars in all figures indicate the standard deviation of the mean for five experiments.