The Arabidopsis receptor kinase FLS2 binds flg22 and determines the specificity of flagellin perception

Abstract

Flagellin, the main building block of the bacterial flagellum, acts as a pathogen-associated molecular pattern triggering the innate immune response in animals and plants. In Arabidopsis thaliana, the Leu-rich repeat transmembrane receptor kinase FLAGELLIN SENSITIVE2 (FLS2) is essential for flagellin perception. Here, we demonstrate the specific interaction of the elicitor-active epitope flg22 with the FLS2 protein by chemical cross-linking and immunoprecipitation. The functionality of this receptor was further tested by heterologous expression of the Arabidopsis FLS2 gene in tomato (Lycopersicon esculentum) cells. The perception of flg22 in tomato differs characteristically from that in Arabidopsis. Expression of Arabidopsis FLS2 conferred an additional flg22-perception system on the cells of tomato, which showed all of the properties characteristic of the perception of this elicitor in Arabidopsis. In summary, these results show that FLS2 constitutes the pattern-recognition receptor that determines the specificity of flagellin perception.

Figure 1.

Antibodies Raised against the C Terminus of FLS2 Detect a 175-kD Polypeptide. Polyclonal antibodies raised against the C terminus of FLS2 were used in protein gel blot analysis with extracts from Arabidopsis cell cultures, wild-type ecotypes, and fls2 mutant plants. M_r (kD) indicates the migration of standard molecular mass markers.

Figure 2.

Immunoprecipitation Using Anti-FLS2 Antibodies. (A) Proteins from Arabidopsis cells were solubilized with detergents and immunoadsorbed to anti-FLS2 antibodies or commercial anti-myc antibodies as a control. Immunoprecipitates were analyzed by SDS-PAGE and staining of protein gel blots using anti-FLS2 antibodies. (B) Binding activity of the immunocomplexes was tested by adding ¹²⁵I-Tyr-flg22 without competitor or with 10 μM unlabeled flg22 as competitor. Bars and diamonds indicate means and actual values, respectively, of two replicate measurements. (C) Radioligand binding of the immunocomplex with anti-FLS2 was tested in the presence of increasing concentrations of unlabeled flg22. Dashed lines indicate the concentration required to reduce binding by 50% (IC₅₀ ∼ 5 nM). Inset, specificity of binding to the immunoprecipitate was assayed in the presence of 100 nM flg15 or 100 nM flg22^Atum as biologically inactive analogs. Bars and diamonds indicate means and actual values, respectively, of two replicate measurements.

Figure 3.

¹²⁵I-Tyr-flg22 Specifically Cross-Links to a 175-kD Polypeptide That Is Immunoprecipitated by Anti-FLS2 Antibodies. (A) Leaf extracts or intact cells of the cell culture were incubated with ¹²⁵I-Tyr-flg22 without competitor or with 10 μM unlabeled flg22, the biologically inactive analog flg22^Atum, or the structurally unrelated peptide elf18 as indicated. After binding, cross-linking was initiated by the addition of ethylene glycol bis(succinimidylsuccinate) (EGS). Radiolabeled proteins were analyzed after separation by SDS-PAGE with a PhosphorImager. Equal loading with proteins was checked by Coomassie blue staining of the gels (data not shown). (B) Competition of label cross-linking to the 175-kD polypeptide by different concentrations of unlabeled flg22. Top, PhosphorImager picture showing the relevant part of the gel with the 175-kD band (no other radiolabeled bands were detectable on the rest of the gel). Bottom, quantification of the label in the 175-kD band seen in the top panel. Dashed lines indicate the IC₅₀ value for the inhibition of labeling by flg22. (C) Immunoprecipitation of the 175-kD band that cross-links to flg22. Membrane proteins from cells after cross-linking with ¹²⁵I-Tyr-flg22 were solubilized with detergents and immunoadsorbed to anti-FLS2 antibodies or anti-myc as a control (left) or first denatured by boiling in SDS in the presence of DTT and then immunoadsorbed to anti-FLS2 antibodies (right). Radiolabeled proteins remaining in the supernatant (supern.) or adsorbed to the antibodies on the Sepharose beads (pellet) were separated by SDS-PAGE and analyzed with a PhosphorImager.

Figure 4.

Expression of FLS2 in Tomato Cells. Detergent-solubilized extracts of equal amounts of cells (fresh weight) from tomato, tomato transformed with P_35S:FLS2:3*myc (tomato-P_35S:FLS2), and Arabidopsis were immunoprecipitated with anti-FLS2 antibodies. (A) Analysis of these immunoprecipitates on protein gel blots with anti-FLS2 antibodies. (B) Binding activity of these immunoprecipitates tested for binding of ¹²⁵I-Tyr-flg22 without the addition of competitor or with 100 nM flg15 or 10 μM flg22 as indicated. Bars and diamonds indicate means and actual values, respectively, of two replicate measurements.

Figure 5.

Competition of ¹²⁵I-Tyr-flg22 Binding by flg22 and flg15. Aliquots of intact tomato cells, tomato cells expressing FLS2, and Arabidopsis cells were incubated with ¹²⁵I-Tyr-flg22 and various concentrations of flg22 and flg15.

Figure 6.

Activity of flg22 and flg22-ΔA^16/17, a flg22 Derivative with an Ala Deletion at Position 16 or 17. Alkalinization response in tomato cells, tomato cells expressing FLS2, and Arabidopsis cells treated with different doses of flg22 and/or flg22-ΔA^16/17. Values represent changes in extracellular pH occurring within 20 min of treatment. Initial pH (pH_i) was 4.6 to 4.8 in the different cell cultures. Competitive antagonism of flg22-ΔA^16/17 in nontransformed tomato cells (left, closed circles) was analyzed in cells pretreated for 6 min with 1 μM flg22-ΔA^16/17, and extracellular pH was determined 20 min after the addition of the different doses of flg22.

Figure 7.

Structural Analogs of flg22 as Tools to Distinguish Tomato-Type and Arabidopsis-Type Flagellin Perception. (A) Alkalinization response in cells of tomato, tomato expressing FLS2, and Arabidopsis to treatment with flg22 and flg22-ΔA^16/17 in the combinations indicated. Changes in extracellular pH (ΔpH) occurring within 20 min of treatment with the peptides are plotted as means (bars) and standard deviations of three replicates. (B) Alkalinization response to treatment with flg22 and flg15-Δ7 in the combinations indicated. (C) Alkalinization response to treatment with flg22-Δ2 and flg15-Δ7 in the combinations indicated.