Figure 1. The Pik-1/Pik-2 orthologues are distributed across diverse species of grasses.

(A) The maximum likelihood (ML) phylogenetic trees of Pik-1 (left) and Pik-2 (right) orthologues. The trees were calculated from 927- and 1239-nucleotide-long codon-based alignments of the NB-ARC domain, respectively, using RAxML v8.2.11 (Stamatakis, 2014), 1000 bootstrap method (Felsenstein, 1985), and GTRGAMMA substitution model (Tavaré, 1986). Best ML trees were manually rooted using the selected clades (marked with grey circles) as outgroups. The bootstrap values above 70% are indicated with grey triangles at the base of respective clades; the support values for the relevant nodes are depicted with numbers. The scale bars indicate the evolutionary distance based on nucleotide substitution rate. The Pik-1 integration clade is shown in pink. Genetically linked genes are linked with lines, with colours indicating plant subfamily: Oryzoideae (purple), Pooideae (dark green), or Panicoideae (light green); the continuous lines represent linkage in a head-to-head orientation, the dashed line indicates linkage in a tail-to-tail orientation. The interactive trees are publicly available at: https://itol.embl.de/tree/14915519290329341598279392 and https://itol.embl.de/tree/14915519290161451596745134. (B) Schematic illustration of the Pik locus in selected species. The schematic gene models of Pik-1 (blue) and Pik-2 (grey) are shown. The integrated heavy metal-associated (HMA) domain is marked with pink. The coordinates of the regions presented in this figure are summarised in Supplementary file 1E. (C) Comparisons of pairwise d_S rates calculated for the Pik-1 and Pik-2 receptors. The rates were calculated using Yang and Nielsen, 2000 based on 972- and 1269-nucleotide-long codon-based alignments of the NB-ARC domains of Pik-1 and Pik-2, respectively; only positions that showed over 70% coverage across the alignment were used for the analysis. The comparisons were categorised to reflect species divergence (shapes) and colour-coded to illustrate percentage identity of d_S values (% identity). The coefficient of determination (R²) was calculated for each dataset using R v3.6.3 package. (D) Summary of identified Pik-1 and Pik-2 homologues in plant species included in this study. The phylogenetic tree was generated using TimeTree tool (Kumar et al., 2017). The number of pairs correspond to the number of Pik-1/Pik-2 genes in a head-to-head orientation separated by intergenic region of various length. ^**The species harbours a truncated gene between Pik-1 and Pik-2; ^*the species has likely lost the HMA domain; Pik-1–HMA: Pik-1 with the HMA domain; Pik-1: Pik-1 without the HMA integration; BOP: Bambusoideae, Oryzoideae, Pooideae; PACMAD: Panicoideae, Arundinoideae, Chloridoideae, Micrairoideae, Aristidoideae, Danthonioideae.

Figure 1—figure supplement 1. Pik-1 and Pik-2 orthologues fall into two well-supported clades.

(A) Phylogenetic tree of CC-type NLRs of Zea mays, Sorghum bicolor, Setaria italica, Triticum aestivum, Hordeum vulgare, Brachypodium distachyon, Oryza brachyantha, and Oryza sativa. The maximum likelihood (ML) tree was calculated based on 241-amino-acid-long alignment of the NB-ARC domains of 3062 of CC-NLRs amended with 35 known and functionally characterised NLRs from grasses using RAxML v8.2.11 (Stamatakis, 2014) with bootstrap values (Felsenstein, 1985) based on 1000 iterations and the best-scoring JTT likelihood model (Jones et al., 1992). The best ML tree is shown. The scale bar indicates the evolutionary distance based on site substitution rate. The clades constituting Pik-1 and Pik-2 orthologues are marked with blue and grey triangles, respectively. Branches corresponding to the reference NLRs are labelled. The interactive tree is publicly available at: https://itol.embl.de/tree/8229133147365371602863457. (B) The ML phylogenetic trees of Pik-1- (left) and Pik-2-related sequences (right) constructed based on 957- and 1218-nucleotide-long codon-based alignments of the sequences of the NB-ARC domain, respectively, using RAxML v8.2.11 (Stamatakis, 2014), 1000 bootstrap method (Stamatakis, 2014), and GTRGAMMA substitution model (Tavaré, 1986). Best ML trees were manually rooted using the selected clades (marked with grey circle) as outgroups. Bootstrap values above 70% are marked with grey triangles at the base of respective clades. The scale bars indicate the evolutionary distance based on nucleotide substitution rate. The interactive trees are publicly available at: https://itol.embl.de/tree/8229133147449491602864812 and https://itol.embl.de/tree/8229133147449511602864812.

Figure 1—figure supplement 2. Genotyping of Oryza brachyantha accession.

(A) Nucleotide alignment of Pikp-2, the ObPik-2 (Ob locus) gene, and the ObPik-2 coding sequence (Ob cds) from the reference genome (Chen et al., 2013), illustrating 46-bp-long deletion and the primers used for the genotyping. (B) Gel electrophoresis of ObPik-2 fragments amplified from different O. brachyantha accessions (labelled above). The symbols next to the accession numbers mark sequences that carry the 46 bp deletion (*), harbour 4 bp deletion (**), carry 1 bp deletion (^×), and do not carry any deletions (•) were used for amplification of the full-length gene (^#). Water and Pikp-2 were used as a negative and positive control, respectively. The left and the right lanes show molecular size markers, labelled on the left.

Figure 1—figure supplement 3. Pik-1 and Pik-2 orthologues from Oryza spp. fall into K- and N-type clades.

The phylogenetic tree shown in Figure 1A, illustrating the divide between the N- (dark grey) and K-type (light grey) Pik genes. The trees were manually rooted using the selected clades (marked with grey circle) as outgroups. The bootstrap values above 70 are indicated with grey triangles at the base of the respective clades; the support values for the relevant nodes are depicted with numbers. The scale bars indicate the evolutionary distance based on nucleotide substitution rate.

Figure 1—figure supplement 4. Schematic representation of selected Pik clusters in wheat (T. aestivum), sorghum (S. bicolor), and foxtail millet (S. italica).

The schematic presents gene models and genetic locations of Pik-1 (blue), Pik-2 (grey), and other NLR genes (purple). Non-NLR genes are shown in light green. The coordinates of the regions presented in this figure are summarised in Supplementary file 1F.

Figure 1—figure supplement 5. Random pairwise comparisons of d_S rates calculated for the Pik-1 and Pik-2 receptors.

The synonymous (d_S) rates were calculated using Yang and Nielsen, 2000 and presented in Figure 1C. The random datasets for d_S values were generated by name shuffling in the existing dataset and random sampling from it 1000 times (left panel). The coefficient of determination (R²) was calculated for every random pairing and the R² distribution was plotted (right panel), as implemented in R v3.6.3 package. If less than 5% of the R² for the random dataset is bigger than the R² for the real dataset, then, according to the null model, the observed difference is very rare and can be accepted as significant with p<0.05.

Figure 1—figure supplement 6. Genetically linked Pik-1 and Pik-2 have similar molecular age.

Comparisons of pairwise d_S rates calculated for the Pik-1 and Pik-2 receptors. The rates were calculated using Yang and Nielsen, 2000 based on 972- and 1269-nucleotide-long codon-based alignments of the NB-ARC domains of Pik-1 and Pik-2, respectively; only positions that showed over 70% coverage across the alignment were used for the analysis. The pairwise comparisons of d_S rates are presented as a heatmap. The comparisons were ordered based on the Pik-1 phylogenetic relationship, shown on the left. The list of genes used for the pairwise comparisons is summarised in Supplementary file 1G.