Molecular and functional analyses of a maize autoactive NB-LRR protein identify precise structural requirements for activity

Abstract

Plant disease resistance is often mediated by nucleotide binding-leucine rich repeat (NLR) proteins which remain auto-inhibited until recognition of specific pathogen-derived molecules causes their activation, triggering a rapid, localized cell death called a hypersensitive response (HR). Three domains are recognized in one of the major classes of NLR proteins: a coiled-coil (CC), a nucleotide binding (NB-ARC) and a leucine rich repeat (LRR) domains. The maize NLR gene Rp1-D21 derives from an intergenic recombination event between two NLR genes, Rp1-D and Rp1-dp2 and confers an autoactive HR. We report systematic structural and functional analyses of Rp1 proteins in maize and N. benthamiana to characterize the molecular mechanism of NLR activation/auto-inhibition. We derive a model comprising the following three main features: Rp1 proteins appear to self-associate to become competent for activity. The CC domain is signaling-competent and is sufficient to induce HR. This can be suppressed by the NB-ARC domain through direct interaction. In autoactive proteins, the interaction of the LRR domain with the NB-ARC domain causes de-repression and thus disrupts the inhibition of HR. Further, we identify specific amino acids and combinations thereof that are important for the auto-inhibition/activity of Rp1 proteins. We also provide evidence for the function of MHD2, a previously uncharacterized, though widely conserved NLR motif. This work reports several novel insights into the precise structural requirement for NLR function and informs efforts towards utilizing these proteins for engineering disease resistance.

Fig 1. Sequence alignment of Rp1-D, Rp1-dp2 and Rp1-D21.

The CC (coiled-coil), NB-ARC (nucleotide binding), ARC1 (APAF1, R gene products and CED-4), ARC2 and LRR (leucine-rich-repeat) domains were indicated by bars with cyan, green, dark blue, pink and red colors, respectively. The conserved motifs (EDVID, P-loop, GLPL, RNBS-D, MHD1 and MHD2) were indicated by orange bars and labeled below the sequences. The recombination point of Rp1-D21 was labeled by cyan bar. Patches 1 and 2 are the major difference regions between Rp1-D and Rp1-dp2 in the ARC2 domain. The positions of the mutations in the intragenic Rp1-D21 suppressor from maize were labeled by blue color. The landmark positions of recombination points in the constructs listed in Fig. 6 were labeled by numbers with red color. The black and light blue shaded regions represent 100% and above 50% similarity of the amino acids, respectively.

Fig 2. EMS mutagenesis screening for mutants that lost the hypersensitive response (HR) induced by Rp1-D21.

(A) The growth phenotype of Rp1-D21 heterozygotes in the field compared to the Rp1-D21 mutant in which the HR phenotype is suppressed (arrow). (B) Twelve missense (in black) and two nonsense (in red) point mutations that abolished the HR phenotype of Rp1-D21 were identified in different domains (CC, NB-ARC and LRR) of Rp1-D21.

Fig 3. Rp1-D21 triggers a hypersensitive response phenotype when transiently expressed in N. benthamiana.

(A) Rp1-D21, Rp1-D and Rp1-dp2 proteins fused with a C-terminal EGFP tag were agro-infiltrated into N. benthamiana, with an empty vector (EV) as a negative control. A representative leaf was photographed at 3 days post infiltration (dpi, left). Total protein was extracted from agro-infiltrated leaves at 30 hours post infiltration (hpi), and anti-GFP antibody was used to detect the expression of the fused proteins (right). The sizes of the proteins were labeled on the right. Equal loading of protein samples was shown by Ponceau-S staining of the Rubisco subunit (below, right). (B) Rp1-D21, Rp1-D and Rp1-dp2 proteins fused with a C-terminal 3×HA tag were agro-infiltrated into N. benthamiana, with empty vector (EV) as a negative control. A representative leaf was photographed at 3 dpi (left). Total protein was extracted from agro-infiltrated leaves at 30 hpi, and anti-HA antibody was used to detect the expression of the fused proteins. Equal loading of protein samples was shown by Ponceau-S staining of Rubisco subunit (below, right). These experiments were repeated three times with the same results.

Fig 4. Functional characterization of the MHD2 motif and the P-loop motif.

(A) Multiple sequence alignment of the MHD motifs from Rp1-D, Rp1-D21 and Rp1-dp2. Positions of the MHD1 and MHD2 motifs are indicated. The numbers indicated the positions of the aspartate (D) corresponding to Rp1-dp2. (B) The MHD1 and MHD2 point mutations in Rp1-dp2 background were generated and transiently expressed in N. benthamiana. (C) Investigating the MHD2 point mutations in V1 and V16, two recombinant constructs that were not autoactive as shown in Fig. 6. (D) Point mutations in P-loop motif of Rp1-D21 and Rp1-dp2(D517V) caused the loss of the ability to induce HR. EV (empty vector) as a negative control. The picture of hypersensitive response (HR) phenotype was taken at 3 days post-infiltration and total protein was detected by anti-HA antibody from agro-infiltrated leaves at 30 hours post infiltration. Equal loading of protein samples was shown by Ponceau-S staining of Rubisco subunit. The experiments were repeated three times with the same results.

Fig 5. Investigating the functional domains of Rp1-D21 and Rp1-D to trigger a hypersensitive response (HR).

(A) Schematic diagram of the Rp1-D21 and Rp1-D domain structure and the derived fragments used for agro-infiltration of N. benthamiana. The different domains are indicated using different colors: CC (purple); NB-ARC (red); LRR (dark blue). The positions of the amino acids that form the domain boundaries are indicated on the top and the abilities to induce (+) or not induce (-) HR are listed on the right of each construct. (B) The HR phenotype of the constructs fused with a C-terminal EGFP tag and transiently expressed in N. benthamiana. Representative leaves were photographed at 3 days post infiltration (dpi, left). Total protein was extracted from agro-infiltrated leaves at 30 hours post infiltration (hpi), and anti-GFP antibody was used to detect the expression of fused proteins. Equal loading of protein samples was shown by Ponceau-S staining of Rubisco subunit (right, below). The results shown were from different domains of Rp1-D21, and similar results were observed from different domains of Rp1-D. (C) The phenotype of the constructs fused with a C-terminal 3×HA tag and transiently expressed in N. benthamiana. A representative leaf was photographed at 3 dpi (left), and the same leaf was cleared by ethanol (middle). Total protein was extracted from agro-infiltrated leaves at 30 hpi, and anti-HA antibody was used to detect the expression of the fused proteins. Equal loading of protein samples was shown by Ponceau-S staining of Rubisco subunit (right). The results shown were from different domains of Rp1-D21, and similar results were observed from different domains of Rp1-D. The experiments were repeated three times with the same results.

Fig 6. Schematic diagram of domain swap constructs between Rp1-D and Rp1-dp2 used for transient expression in N. benthamiana.

For ease of interpretation, the gene structure has been divided into CC (coiled coil), NB-ARC (nucleotide binding) and LRR (leucine rich repeat). The amino acid position of the recombination site of each construct was indicated above the construct. For each construct, either with no tag, 3×HA tag or EGFP tag, the strength of the hypersensitive response (HR) resulting from transient expression in N. benthamiana is shown. HR was scored on a 0 (no HR) to 5 (strong HR) scale according to Slootweg et al. (2013). If there was no score recorded, the particular experiment was not performed. Where known, the abilities of constructs to induce (+) or not induce (-) HR in maize were indicated. In some cases the number of single amino acid polymorphisms (SAAP) between constructs is indicated. All experiments were performed three times with similar results. (A) Constructs with the N-terminus from Rp1-D and C-terminus from Rp1-dp2. (B) Constructs designed to delimit the domains of Rp1-D or Rp1-dp2 important for self-inhibition. (C) The reciprocal constructs: Rp1-D21 with V16, and V3 with V17.

Fig 7. Investigating the functional regions of amino acids (AAs) from 1170–1200 and the C-terminal end of Rp1-D21.

(A) Investigating the point mutations in the region of AAs 1170–1200 which are important for inducing hypersensitive response (HR). Top: Multiple sequence alignment of Rp1-D, Rp1-D21, Rp1-dp2, V1 and V2. The black, red and light blue shaded regions represent 100%, 75% and above 50% similarity of the amino acids, respectively. The polymorphic amino acids are boxed, and the position of 1184 (corresponding to 1182 of Rp1-D21) is labeled above the sequences. Bottom left: The HR phenotype of V1, V2 and V1-derived point mutants as indicated. A representative leaf was photographed at 3 days post infiltration (dpi). Bottom right: Total protein was extracted from agro-infiltrated leaves at 30 hours post infiltration (hpi), and anti-HA antibody was used to detect the expression of the fused proteins. Equal loading of protein samples was shown by Ponceau-S staining of Rubisco subunit. (B) Investigating the phenotype conferred by various recombinants and point mutants at the C-terminal end of Rp1-D21. Top: Multiple sequence alignment of the C-terminus of Rp1-D, Rp1-D21 and Rp1-dp2. The polymorphic amino acids are boxed. Bottom left: Schematic diagram of the constructs used. The K1184N mutation is indicated by a red vertical bar in Rp1-dp2(K1184N), V1(K1184N) and V18. Bottom middle: The HR phenotype of the constructs transiently expressed in N. benthamiana. A representative leaf was photographed at 3 dpi. Bottom right: Total protein was extracted from agro-infiltrated leaves at 30 hpi, and anti-HA antibody was used to detect the expression of the fused proteins. Equal loading of protein samples was shown by Ponceau-S staining of Rubisco subunit. The experiments were repeated three times with the same results. (C) The final model presenting the functional regions important for autoactivity. In this model, white represents regions derived from Rp1-dp2 and black the region from Rp1-D that need to be present in the same construct to result in the autoactive HR in N. benthamiana. Regions in grey can be derived from either paralog. The combination of AA 458–651 from Rp1-dp2 and AA 1184–1292 (especially N1184 and the last 16 AAs in red vertical bars) from Rp1-D are important for HR induction.

Fig 8. Investigating the function of Patch 1 and 2 from the ARC2 domain for Rp1-D21 autoactivity.

(A) Sequence alignment of Rp1-D, Rp1-D21 and Rp1-dp2 in Patch 1 and 2 regions from the ARC domain. Positive (+), negative (-) and neutral (n) charges were labeled. RNBS-D, MHD1 and MHD2 motifs are indicated by orange bars and labeled below the sequences. (B) Homology modeling of NB-ARC_D21 based on the crystal structure of APAF1 (PDB: 1z6t). The P-loop, MHD, GLPL, RNBS-A, RNBS-D and Walker B motifs are color-coded in magenta, yellow, orange, pink, green and blue, respectively. Two boxed regions (Patch 1 and 2) are the major polymorphic regions between Rp1-D and Rp1-dp2 and are labeled in red. The positions of the suppressor mutations T260I and E312K identified by EMS mutagenesis are indicated in green. The N- and C-termini are indicated as N and C, respectively. (C) The surface electropotential of NB-ARC_D21 and NB-ARC_D. The positive and negative charges are indicated in blue and red, respectively. The positions of Patch 1 and Patch 2 are indicated by pink dots in the yellow oval boxes. (D) Schematic diagram and the hypersensitive response phenotype of the constructs indicated. (E) Protein expression was detected by anti-HA antibody at 3 days post-infiltration. Ponceau-S staining of Rubisco subunit was used as loading control. The experiment was repeated three times with the same results.

Fig 9. Investigation of self-association of full length Rp1-D21, Rp1-D, Rp1-dp2, the separate domains of Rp1-D21, and the interaction of full length Rp1-D21 with Rp1-dp2.

(A) Self-association of 3×HA- and EGFP-tagged Rp1-D21 (left panel), Rp1-D (middle panel) and Rp1-dp2 (right panel). Arrows indicate the target bands. (B) Self-association of CC_D21, NB_D21 and LRR_D21. (C) Left: Co-IP after co-expression of Rp1-D21 and Rp1-dp2. Right: HR induced by co-expression of Rp1-D21:EGFP with Rp1-D21:HA compared to HR induced by co-expression of Rp1-D21:EGFP with Rp1-dp2:HA. EGFP- and 3×HA-tagged constructs were transiently co-expressed in N. benthamiana and samples were collected at 30 hours post infiltration for Co-IP assay. Protein extract was immunoprecipitated by anti-GFP microbeads and detected by anti-GFP and anti-HA antibodies. The experiments were repeated three times with the same results.

Fig 10. Schematic diagram of the intra-molecular interactions and HR phenotype of the constructs indicated, summarizing the data shown in S7 Fig. in which the various domains were co-expressed and co-immunoprecipitated in trans.

The gene structure has been divided into CC, NB-ARC and LRR domains. The amino acid position of the recombination site of each construct was indicated above the construct. For each construct, the strength of the hypersensitive response (HR) resulting from transient expression of each full length molecule in N. benthamiana was scored on a 0 (no HR) to 5 (strong HR) scale. The ability to form or not to form intra-molecular interactions among different domains was indicated by arrows or arrows with crosses, respectively.

Fig 11. Model of the intra-molecular interactions controlling auto-inhibition or auto-activation in Rp1 proteins.

In the auto-inhibited state (left), the NB-ARC domain inhibits CC-induced HR via intra-molecular interaction involving the ARC2 domain. In the auto-activated state (right), the LRR suppresses an inhibitory effect of the NB-ARC on the CC through direct interaction with the NB-ARC which destabilizes the CC/NB-ARC interaction. Blue and pink triangles indicate ADP and ATP, respectively. Arrows indicate intra-molecular interactions. The positions of amino acids (D82, R125, N1184 and the C-terminal 16 amino acids) which are important for activity are labeled in red. The rationale behind the models is explained in the main text.