Genome analysis of poplar LRR-RLP gene clusters reveals RISP, a defense-related gene coding a candidate endogenous peptide elicitor

Abstract

In plants, cell-surface receptors control immunity and development through the recognition of extracellular ligands. Leucine-rich repeat receptor-like proteins (LRR-RLPs) constitute a large multigene family of cell-surface receptors. Although this family has been intensively studied, a limited number of ligands has been identified so far, mostly because methods used for their identification and characterization are complex and fastidious. In this study, we combined genome and transcriptome analyses to describe the LRR-RLP gene family in the model tree poplar (Populus trichocarpa). In total, 82 LRR-RLP genes have been identified in P. trichocarpa genome, among which 66 are organized in clusters of up to seven members. In these clusters, LRR-RLP genes are interspersed by orphan, poplar-specific genes encoding small proteins of unknown function (SPUFs). In particular, the nine largest clusters of LRR-RLP genes (47 LRR-RLPs) include 71 SPUF genes that account for 59% of the non-LRR-RLP gene content within these clusters. Forty-four LRR-RLP and 55 SPUF genes are expressed in poplar leaves, mostly at low levels, except for members of some clusters that show higher and sometimes coordinated expression levels. Notably, wounding of poplar leaves strongly induced the expression of a defense SPUF gene named Rust-Induced Secreted protein (RISP) that has been previously reported as a marker of poplar defense responses. Interestingly, we show that the RISP-associated LRR-RLP gene is highly expressed in poplar leaves and slightly induced by wounding. Both gene promoters share a highly conserved region of ~300 nucleotides. This led us to hypothesize that the corresponding pair of proteins could be involved in poplar immunity, possibly as a ligand/receptor couple. In conclusion, we speculate that some poplar SPUFs, such as RISP, represent candidate endogenous peptide ligands of the associated LRR-RLPs and we discuss how to investigate further this hypothesis.

Keywords: gene clustering; immunity; ligands; poplar; receptors; wounding.

Figure 1

Poplar and arabidopsis LRR-RLP families evolved independently. (A) Alignment of the protein sequences of the 82 and 45 LRR-RLP from poplar and A. thaliana, respectively. Blue and red colors indicate a minimum of 50 and 90% of amino acid identity per position, respectively. The canonical domains of tomato Cf-9-like LRR-RLPs described by Fritz-Laylin et al. (2005) are indicated above the alignment. The asterisks mark the domains that are variable in size and position within the alignment; they have been arbitrarily adjusted to correspond to the sequence of poplar LRR-RLP9 (see text). The C3 and D domains, conserved in all sequences, are indicated in red and have been used for performing the phylogenetic analysis presented in (B). LRR, Leucine-Rich Repeat; Cys, Cysteine; TM, Trans-Membrane. (B) Phylogenetic tree of poplar and A. thaliana LRR-RLP families. The analysis was done with the C3-D domains presented in (A). Poplar sequences are highlighted in green, whereas arabidopsis sequences are in blue. Main nodes with Approximate Likelihood-Ratio Test (aLRT) values superior to 0.7 are marked with an asterisk.

Figure 2

Poplar LRR-RLP genes cluster with genes coding small-proteins of unknown function. (A) Phylogenetic tree of the 82 LRR-RLP from poplar constructed with the C3-D domains. Almost all sequences gather into four main clades (a, b, c, and d) indicated by gray boxes. IDs have been colored according to their presence on similar chromosomes (chr.) or scaffolds (scf.), as follow: chr.3: orange; chr.5: green; chr.11: blue; chr.12: red; chr.15: cyan; chr.16: dark-yellow; scf.39: pink; scf.46: gray; scf.64: purple; any other chr. or scf.: black. Sequence IDs marked with an asterisk indicate that the corresponding gene model is represented in clusters depicted in (B). (B) The nine main clusters (C) or super-clusters (SC) of LRR-RLP genes are depicted, approximately facing their corresponding phylogroup in (A). All genes present within LRR-RLP clusters (i.e., any genes at less than 25 kb of an LRR-RLP gene) are represented. This threshold has been extended to 50 kb in the case of SC3.

Figure 3

Poplar SPUFs have limited paralogs and homologs in other plants. Poplar SPUF sequences were used for homology searches against the predicted poplar proteome on the Phytozome portal, as well as the non-redundant protein database at the NCBI website. The dataset used, homology search details and homologs identified in other plants are detailed in Table S1B.

Figure 4

Some families of SPUFs are gathered into the same LRR-RLP clusters. The left-hand part of the figure presents a phylogenetic tree computed from poplar SPUF sequences. Sequence IDs of SPUFs are colored as their clustered LRR-RLPs in Figure 2. Vertical bars indicate close sequences for which gene models are physically associated within LRR-RLP clusters. Thick bars indicate examples presented in details in Figure 5D (black bar) or on the right-hand part of the figure (green and cyan bars). On the right-hand of the figure, the green rectangle presents a sub-part of Super-Cluster 5 (SC5), whereas the cyan rectangle presents the Cluster 15 (C15) as well as a SPUF/LRR-RLP pair from chromosome 12 (Chr. 12). The color code for arrows is as described in Figure 2. The last numbers of gene model IDs are indicated. Red bars and corresponding numbers indicate the percentage of amino acid identity between the protein products of the SPUF genes. The inserts in the black rectangles present SPUF alignments. Blue and red colors indicate a minimum of 50 and 90% of amino acid identity per position, respectively.

Figure 5

The gene expression of some pairs of LRR-RLP/SPUF genes is coordinated. (A,B) The average levels of gene expression measured in poplar leaves (non-wounded condition) and in excised leaf disks (wounded condition) with oligoarrays are compared for the considered LRR-RLP (A) and SPUF (B) genes. Error bar: SE, n = 6. The linear regression and associated values are indicated for each graph. Colored points are further discussed in the text and in the parts (C,D). The complete transcriptome dataset is available in Table S1C. (C) Cluster 11 (C11), a sub-part of Super-Cluster 5 (SC5), and Cluster 15 (C15) associated with a pair of LRR-RLP/SPUF genes from chromosome 12 (Chr.12) are depicted (for color code and gene model IDs, see Figure 2 and the legend of Figure 4). Green (C11), orange (SC5), and yellow (C15 + Chr.12) dots under gene models indicate a co-expression of LRR-RLP and SPUF genes in poplar leaves, as shown in (A,B). (D) RISP and LRR-RLP genes from scaffold 580 (Phytozome ID Poptr_0580s00210 and Poptr_0580s00200, respectively), are the closest paralogs of RISP-Like and LRR-RLP9 genes from chromosome 9 (Phytozome ID Poptr_0009s11500 and Poptr_0009s11510), respectively. Bars and associated numbers indicate the percentage of amino acid identity. Red and blue dots under RISP and LRR-RLP9 genes indicate their co-induction as shown in (A,B).

Figure 6

Physical association between RISP and a chimera of LRR-RLP580 and LRR-RLP9. (A) Schematic representation of the location of RISP, LRR-RLP580, and LRR-RLP9 on P. trichocarpa genomic sequence (Phytozome v2.2). The primers used to amplify and sequence the fragment containing the LRR-RLP associated with RISP are indicated. Note that the primer 1 is specific to the RISP gene whereas primers 2, 3, and 4 could not discriminate between LRR-RLP9 (blue) and LRR-RLP580 (red) genes. The primers 5 and 6 were designed to amplify the extracellular domain of both LRR-RLP9 and LRR-RLP580. (B) The large 8 kb fragment amplified using the 1–4 primer pairs was used as template to amplify the two shorter fragments (using 1–2 and 3–4 primer pairs) that were used for sequencing.