New Results

An ancient integration in a plant NLR is maintained as a trans-species polymorphism

Helen J. Brabham, Inmaculada Hernández-Pinzón, Samuel Holden, Jennifer Lorang, Matthew J. Moscou
doi: https://doi.org/10.1101/239541

Main text

Plant immune receptors are under constant selective pressure to maintain resistance to plant pathogens. Nucleotide-binding leucine-rich repeat (NLR) proteins are one class of cytoplasmic immune receptors whose genes commonly show signatures of adaptive evolution1,2. While it is known that balancing selection contributes to maintaining high intraspecific allelic diversity, the evolutionary mechanism that influences the transmission of alleles during speciation remains unclear. The barley Mla locus has over 30 described alleles conferring isolate-specific resistance to barley powdery mildew and contains three NLR families (RGH1, RGH2, and RGH3)3. We discovered (using sequence capture and RNAseq) the presence of a novel integrated Exo70 domain in RGH2 in the Mla3 haplotype. Allelic variation across barley accessions includes presence/absence of the integrated domain in RGH2. Expanding our search to several Poaceae species, we found shared interspecific conservation in the RGH2-Exo70 integration. We hypothesise that balancing selection has maintained allelic variation at Mla as a trans-species polymorphism over 24 My, thus contributing to and preserving interspecific allelic diversity during speciation.

Recognition of plant pathogens is mediated by membrane and cytoplasmic immune receptors that recognise pathogen-derived molecules, including effectors4. Cytoplasmic immune receptors recognise secreted pathogen effectors directly, or indirectly through monitoring the molecular status of host proteins targeted by pathogen effectors5. At the population level, plant immunity to pathogens is conferred, in part, through the maintenance of diverse allelic variants at immune receptor loci. Pathogen pressure influences the relative frequency of resistance alleles in plant populations, with selection fluctuating due to pathogen evolution. Mutations that generate novel approaches to manipulate a host by immune suppression or nutrient acquisition, for example, are selected for in the pathogen. Similarly, novel forms of pathogen recognition are selected for in the plant. A recent area of intensive research in plant immunity includes the identification of integrated non-canonical domains in NLRs, which are thought to serve as baits for pathogen effectors, and thus, facilitate efficient immune responses6–8. It is hypothesised that integrated domains are released from purifying selection for host function, and are free to adapt specifically for pathogen recognition6.

Blumeria graminis f. sp. hordei (barley powdery mildew) is a pandemic disease of barley (Hordeum vulgare). A major source of isolate-specific resistance to B. graminis f. sp. hordei is found at the multi-allelic Mla locus, which contains the three NLR families RGH1, RGH2, and RGH33. All known functional alleles of the Mla gene belong to the RGH1 gene family9. Considerable haplotype variation has been identified based on sequencing RGH1 genomic fragments, with alleles sharing >90% protein identity3,9. Specific residues of Mla proteins are under positive selection, particularly in the LRR region, which is known to determine the specificity of B. graminis f. sp. hordei recognition9,10. In the barley accession Morex, the Mla locus contains members of all three NLR families that are organised in three gene-rich regions interspersed with transposable elements3. In addition, a near identical 40-kb tandem duplication includes members from all three NLR families3. Morex contains RGH1bcd, an allele of Mla pseudogenised due to a retrotransposon insertion. We set out to understand the diversity of the RGH gene families across Mla haplotypes. To do so, we performed RNAseq on leaf tissue derived from diverse barley accessions, including elite, landrace, and wild barley. Sequence variation and presence/absence variation in expression of RGH1, RGH2 and RGH3 were identified across accessions (Fig. 1a). RGH1 was the most prevalent of all RGH gene families with variation including intact open reading frames (ORF), pseudogenised genes, and in two cases, absence at the expression level. For all haplotypes sequenced, RGH2 was always present with RGH3. Presence of RGH3 without RGH2 was seen in four accessions, and only in those also containing a pseudogenised Mla allele. Strikingly, we observed the presence of an integrated Exo70 domain within RGH2 in five haplotypes (accessions Baronesse, Duplex, Finniss, HOR 1428, and Maritime). Allelic variants of RGH2 in other accessions contain a coiled coil integrated domain derived from RGH1 or an in-frame full length RGH1 family member. Using sequence capture designed on the entire Mla locus and PacBio circular consensus sequencing11, the integrated RGH2-Exo70 was confirmed in the barley cultivar Baronesse (Fig. 1c). Furthermore, RGH2-Exo70 and RGH3 in Baronesse were found in head-to-head orientation (Fig. 1b), and based on other similar structural examples we hypothesise they function as a pair12. Head-to-head orientation of RGH2 and RGH3 was also observed in other grass species (Fig. 1b).

Fig. 1.
Fig. 1. Intra- and inter-specific variation in RGH1, RGH2, and RGH3. RGH1, RGH2 and RGH3 family members are shown in coloured arrows (blue, yellow, and grey, respectively).

(a) Haplotype variation in leaf expression of RGH1, RGH2, and RGH3 within barley (Hordeum vulgare; Hv), Aegilops tauschii (Aet), and oat (Avena sativa; Avs). Presence of a complete coding sequence is shown in bold arrows, absence with faded arrows, pseudogenised RGH1 with the symbol ψ, integration of Exo70 with purple arrows, and out-of-frame integrated Exo70 with hatched shaded arrows. Haplotypes is defined as the number of haplotypes containing different gene combinations. (b) Genomic structure of the region encompassing RGH2 and RGH3 across Poaceae species showing conserved head-to-head orientation of RGH2 and RGH3. Variation is observed in RGH2 integrated domains with barley accessions Morex (mla)3 and Baronesse (Mla3/Rmo1)9,30 containing RGH2-Exo70 (yellow-purple arrow), oat accession Victoria containing RGH2-Exo70 (yellow-purple arrow), Brachypodium distachyon (Bd) containing RGH2-Receptor-like-kinase (yellow-teal arrow), and rice (Oryza sativa; Os) containing RGH2-Thioredoxin (yellow-light purple arrow). A. sativa accession Victoria and B. distachyon integrated Exo70 are out-of-frame with RGH2, indicated with the hatched shaded arrows. (c) Protein model of non-integrated RGH2 and integrated RGH2-Exo70. Individual domains include coiled coil (CC; red), nucleotide-binding (NB; orange), leucine-rich repeats (LRR; cream), and Exo70 (purple).

Exo70 is one of the eight subunits comprising the exocyst complex, involved in the tethering of post-Golgi secretory vesicles to the plasma membrane prior to SNARE complex-mediated membrane fusion13–15. In plant genomes, the core exocyst subunits of Sec3, Sec5, Sec6, Sec8 and Sec10 are retained in few or single copies, whereas Exo70 has experienced dramatic proliferation into multiple gene families16. Diversification is characteristic of sub-functionalisation or neo-functionalisation17 and Exo70 genes are involved in diverse roles in morphogenesis15, development18, and immunity19. Cvrcóková et al. (2012) comprehensively described members of the exocyst complex throughout land plants using representative species from both monocots and dicots16. We annotated the Exo70 gene families from several sequenced grass species including Brachypodium distachyon, barley, Oryza sativa (rice), Oropetium thomaeum, Sorghum bicolor, Setaria italica, and Zea mays (maize). The Exo70F and Exo70FX clades were found to be greatly expanded within these grasses (Fig. 2a). Using maximum likelihood phylogenetic analysis, we discovered that the integrated Exo70 originated from Exo70F1 (Fig. 2b). Approximately 87% of the coding sequence of Exo70F1 is integrated in RGH2. Members of Exo70F and Exo70FX gene families are known to have a role in immunity. Barley EXO70F-like (ExoFX11b.a) is essential for full penetration resistance to B. graminis f. sp. hordei20. Rice Exo70F3 binds the M. oryzae effector AVR-Pii, and this interaction is essential for Pii (NLR) mediated resistance21,22. The integration of Exo70F1 in RGH2 suggests a potential role in immunity either through localisation of the NLR or as a decoy for effector recognition.

Fig. 2.
Fig. 2. Exo70F1 is the donor of the Exo70 integration in RGH2. (a) Maximum likelihood phylogenetic tree of Exo70 proteins from seven grass species (see below; Supplementary Data 1), including barley RGH2 integrated Exo70 denoted HvRGH2.

Saccharomyces cerevisiae Exo70 gene (YJL085W) was used as an outgroup. Black dots represent bootstrap support greater than 80% based on 1,000 bootstraps. (b) Maximum likelihood phylogenetic tree of Exo70F proteins from seven grass species (see below). The origin of the integrated Exo70 in RGH2 is Exo70F1 family (highlighted in pink). Arabidopsis thaliana Exo70 gene (AT5G50380; AtExo70F1) was used as an outgroup. Orange dots represent bootstrap support greater than 80% based on 1,000 bootstraps. Species included in the analysis were: B. distachyon (Bd), barley (Hv), rice (Os), Oropetium thomaeum (Ot), Sorghum bicolor (Sb), Setaria italica (Si), maize (Zea mays; Zm).

Integration of Exo70F1 in RGH2 was previously observed in the D genome of wheat (Triticum aestivum L.) in cloning Sr33 on chromosome 1D23. Presence of integrated Exo70F1 in wheat and barley prompted us to investigate its conservation across diverse grasses to establish its potential as a trans-species polymorphism2. First, to understand whether allelic variation exists in Aegilops tauschii (the donor of the D genome), we analysed several publically available leaf transcriptomes24. We observed allelic variation in the presence or absence of RGH2 and RGH3 expression, with RGH2 present as the integrated Exo70F1 allele (Fig. 1a). We did not identify an allele of RGH2 without the Exo70F1 integration. We expanded our search to include additional species outside the Triticeae with sequenced genomes or publically available leaf RNAseq data. Sequence analysis found allelic variation and interspecific conservation in the RGH2-Exo70F1 integration across diverse Poaceae species. In oat, (Avena sativa) RGH2 and RGH3 were present in two accessions, whereas RGH1 was absent (Fig. 1a). The integrated Exo70F1 was in frame in the accession Kanota, whereas a single base pair InDel causes an out-of-frame Exo70F1 in Victoria. Integration of Exo70F1 in RGH2 in the genome was confirmed using sequence capture and PacBio sequencing of the oat accession Victoria. Using maximum likelihood phylogenetic analysis, we found that integrated Exo70F1 derived from Triticeae species, A. sativa, and Holcus lanatus form a distinct clade from non-integrated Exo70F1 (Fig. 3). B. distachyon forms the outgroup of this integrated clade, suggesting that integration of Exo70F1 in RGH2 occurred after speciation of B. distachyon but prior to radiation of the Poeae and Triticeae. The RGH2 ortholog in B. distachyon encodes an NLR with a C-terminal integrated receptor-like kinase with intact transmembrane and extracellular lectin domains. RGH2 orthologs in oat and the Ae. tauschii-derived D genome of wheat contain integrated Exo70F1 that is not in frame, suggesting that gene exchange through unequal crossovers, gene conversion, or mutation can interrupt the ORF. The differentiation of integrated and non-integrated Exo70F1 after B. distachyon speciation, but prior to Poeae-Triticeae radiation dates this integration at 24 Mya (CI: 18.4 to 29.8 Mya)25, and is indicative of a trans-species polymorphism.

Fig. 3.
Fig. 3. Exo70F1 integration in RGH2 occurred prior to the Poeae-Triticeae radiation. Maximum likelihood phylogenetic tree was performed on codon aligned Exo70F1 genes from 19 grass species (Supplementary Data 1). Integrated Exo70F1 highlighted in blue.

Branch support was generated using 1,000 bootstraps, with orange dots designating support greater than 80%. Rice Exo70F1 (OsExo70F1; Os01g69230.1) was used as an outgroup. Scale bar shows nucleotide substitutions per site. Species included in the analysis were: Achnatherum splendens (Acs), Ae. tauschii (Aet), Agropyron cristatum (Agc), Agrostis stolonifera (Ags), A. sativa (Avs), B. distachyon (Bd), Bromus inermis (Bin), Dactylis glomerata (Dgl), Festuca pratensis (Fpr), Holcus lanatus (Hla), H. pubiflorum (Hp), barley (Hvu), rice (Os), Melica nutans (Mnu), Nardus stricta (Nst), Poa pratensis (Ppr), Secale cereale (Sec), Stipa lagascae (Sla), Triticum aestivum (TaA, TaB, TaD subgenomes), maize (Z. mays; Zm). Clusters were formed due to 100% identical sequence and are listed in Supplementary Table 1.

We found that all grass Exo70 gene families are under strong purifying selection (Supplementary Table 2). We hypothesise that integration of Exo70F1 into RGH2, and its role as a proposed ‘decoy’ domain, would relax purifying selection compared to non-integrated Exo70F1. To understand the impact on the molecular evolution across Exo70F1, we performed branch-specific tests for variable levels of selective pressure. Evidence of relaxed purifying selection was observed for the integrated Exo70F1 clade (ωa = 0.423), compared to stronger purifying selection for non-integrated domains of Exo70F1 (ω0 = 0.102) (Supplementary Table 3, Supplementary Fig. 1). Evidence of relaxation confirms the release of integrated domains from strong purifying selection that preserves the endogenous activity of their non-integrated counterparts26.

We were interested if, following incorporation, integrated Exo70F1 has altered the evolution of RGH2. To investigate the potential co-evolution of RGH2 and the integrated Exo70F1, we compared the phylogenies of RGH1, RGH2, and RGH3 gene families; and Exo70F1. All RGH gene families predominantly exhibit species specific grouping (Supplementary Fig. 2, 3, 4). Variation is seen in the domain structure of RGH2 homologs; the majority contain an Exo70F1 integration however other domains such as coiled coil, receptor-like kinase (with lectin domains), thioredoxin and protein kinase are observed in orthologs (Supplementary Fig. 3). RGH2 belongs to the NLR clade MIC1 comprising members known to contain diverse integrated domains27. Integrated Exo70F1 have a distinct evolutionary history compared to their non-integrated counterparts. The integrated Exo70F1 clade follows the species phylogeny, with the exception of HvuRGH2HOR1428 and TaDRGH2.2 from the reference genome of wheat (Fig. 3). Within the integrated clade two subclades are supported: one clade containing the Aegilops and barley alleles, and the other comprising wheat, Phalaris arundinacea, Holcus lanatus, and oat (Fig. 3). In contrast, RGH2-Exo70F1 and non-integrated RGH2 alleles do not show the same clade distinction; therefore, integrated status has not altered RGH2 evolutionary history and co-evolution has not occurred (Supplementary Fig. 3).

Until now, it was unclear how selective forces in the plant-pathogen interaction influence the preservation of diverse alleles through plant speciation events because previous evolutionary analyses were limited by high rates of intraspecific and interspecific variation of NLRs. We used the integrated Exo70F1 domain within RGH2 as an evolutionary footprint to track the history of the Mla locus through speciation of the grasses. We observed presence/absence variation in RGH2-Exo70F1 across barley and presence of RGH2-Exo70F1 alleles in five grass species, indicating the passage and retention of ancestral alleles through speciation. This establishes a trans-species polymorphism originating 24 Mya25, and maintained to present day (Fig. 4). Long-lasting trans-species polymorphisms are maintained through balancing selection, as in its absence, mutations within a population are either lost, achieve fixation, or exist in a frequency-dependent manner2. Previous work in plant immunity established time scales of long-term maintenance of polymorphic sites in stress response genes between A. thaliana and Capsella rubella28 over 5 Mya, and presence/absence polymorphisms in RPS5 in A. thaliana due to long-term balancing selection29 over 2 Mya. Using genomes and transcriptomes from species across the Poaceae, we have unravelled the evolutionary history of a shared integrated NLR within the grasses, and shown that allelic diversity is present as a trans-species polymorphism at Mla. The association of the Mla locus with multiple pathogen recognition30 emphasises the impact of diverse pathogen populations on the evolutionary forces shaping immune receptor diversity in the Poaceae.

Fig. 4.
Fig. 4. An evolutionary model of integrated and non-integrated RGH2 transmission during the speciation of Poeae and Triticeae.

Horizontal lines of dots represent the gene pool across a species at a given point in time. Branch points represent speciation events. Coloured dots represent allelic variants of RGH2; blue dots represent non-integrated alleles, yellow dots represent RGH2-Exo70 integrations, and green dots show RGH2-Receptor-like kinase (RGH2-RLK) integration in B. distachyon. Following speciation of B. distachyon, an allele of RGH2 acquires the integrated Exo70F1. Integrated and non-integrated alleles of RGH2 are maintained within the population though balancing selection, and are sustained through the speciation of Avena and other Poeae species, and subsequently the Triticeae. Long-term balancing selection preserves the RGH2 allelic pool as a trans-species polymorphism.

Methods

Plant material

A full inventory of plant species and accessions used in this study are available in Supplementary Data 1.

RNAseq and de novo assembly

First and second leaf tissue was harvested at 10 days after sowing of barley and oat accessions grown in the greenhouse. Tissue was flash frozen in liquid nitrogen and stored at −80 °C. Tissue were homogenized into a fine powder in liquid nitrogen-chilled pestle and mortars. RNA was extracted, purified, and quality assessed as described by Dawson et al. (2016)31. RNA libraries were constructed using Illumina TruSeq RNA library preparation (Illumina; RS-122-2001). Barcoded libraries were sequenced using either 100 or 150 bp paired-end reads. Library preparation and sequencing was performed at either the Earlham Institute (Norwich, United Kingdom) or BGI (Shenzhen, China). Quality of all RNAseq data was assessed using FastQC32 (0.11.5). Trinity33 (2.4.0) was used to assemble de novo transcriptomes using default parameters and Trimmomatic34 for read trimming. Exo70F1, RGH1, RGH2, and RGH3 were identified in de novo assemblies using BLAST+ (v2.2.9)35.

Development of the barley and oat sequence capture

The barley capture library TSLMMHV1 is composed of 99,421 100 mer baits with 2x coverage over the target space. Targeted sequences include repeat masked Mla locus from Morex3, all cloned alleles of Mla9,10,36-38, the Mlo locus39, the Rpg1 locus40, and the rpg4/Rpg5 locus41. In addition, the capture library targets the barley NLR gene space identified in genomic sequence from barley accessions Barke, Bowman, and Morex, full length cDNA derived from barley accession Haruna Nijo, and transcriptomes of barley accessions Abed Binder 12, Baronesse, CI 16153, CIho 4196, Manchuria, Pallas, Russell, and SusPtrit. The capture library TSLMMAS1 is composed of 29,672 100 mer baits with 2x coverage over the target space. Targeted sequences include NLR containing contigs from de novo assembled transcriptomes of oat accessions Kanota and Victoria.

For both libraries, NLR gene space was identified using the following approach. For transcriptomes, TransDecoder42 (v4.1.0) was used to identify and translate ORFs. Using a similar strategy as Jupe et al. (2012)43, we developed a motif set using MEME44 trained on a random randomized proportional sample of NLR45,46 from rice (N=35) and B. distachyon (N=17). The MEME motifs spanned the CC domain (motifs 4, 11, 13, and 15), NB domain (motifs 1, 2, 3, 5, 6, 7, 8, 10, 12, and 14), and the LRR domain (motifs 19, 9, 20, 16, 17, and 18) (Supplementary Data 3). All the identified motifs in the NB are similar to those previously defined by Meyers et al. (2003)47 in A. thaliana. A MAST significance threshold of 1e-20 was selected based on its ability to identify all annotated NLRs within B. distachyon and rice. For barley whole genome shotgun assemblies, all six ORFs were translated for each contig and concatenated into a single peptide sequence for the forward and reverse strand. Translated genomic contigs were scanned using FIMO, which assesses all twenty MEME-generated motifs independently. Contigs were included in the capture if one of two conditions were met: (1) at least one CC and two NB motifs were present or (2) at least two NB and one LRR motifs are present in the translated sequence strand.

Next, redundancy within the sequence capture template was removed. We fragmented the input data set into 100 bp fragments with a scanning window of 25 bp and performed BLASTn onto the entire data set. Any sequence found to have identity of 95% or higher was considered redundant other than the original site. The first occurrence of the sequence would be retained and the others were masked. While this approach removes redundancy in the data set, the inclusion of extensive genomic sequence will introduce repetitive sequence that can produce competition in the sequence capture due to the high copy number of repetitive sequence in the barley genome. Therefore, two approaches were used to remove repetitive sequence in the sequence capture design. All loci were repeat masked based using RepeatMasker48 (v4.0.5) using default and Triticeae-specific repeat databases. As repeat databases are not complete, we applied genomic masking of the capture design. To do so, we fragmented the input data set into 100 bp fragments with a scanning window of 50 bp and performed BLAST onto the barley Morex WGS assembly. A threshold of eight or fewer copies was found selected to balance between copy number variation within NLRs and avoiding the inclusion of repetitive sequence.

DNA extraction and sequencing library preparation

Total genomic DNA was extracted from leaf tissue according to a CTAB method49. In brief, 3 g of leaf tissue were ground on liquid N2 and homogenized with 20 mL of CTAB extraction buffer (2% CTAB, 100 mM Tris-HCl pH8.0, 20 mM EDTA pH8.0, 1.4 M NaCl, 1% β-Mercaptoethanol). Samples were incubated for 30 min at 65°C followed by two chloroform extractions and ethanol precipitation. DNA was then resuspended in 1x TE, 50 μg/mL RNase A solution and incubated for 1 h at 37°C. DNA was subsequently precipitated with 2.5 volumes of ice-cold 95% ethanol and resuspended in 1x TE. Quantification of DNA samples was performed using a Nanodrop spectrophotometer (Thermo Scientific) and the Qubit dsDNA HS Assay Kit (Molecular Probes, Life Technologies). DNA samples were normalized to 3 µg and sheared to an average length of 3-4 kb using a Covaris S2 sonicator with the following settings: Duty Cycle 20%, Intensity 1, Cycle Burst 1000, Time 600 s, Sample volume 200 µL. After sonication, a small aliquot was assayed by gel electrophoresis and additional size selection was carried out using 0.4x Agencourt AMPure XP beads (Beckman Coulter Genomics). Samples were then end repaired followed by 3’dA addition using the NEBNext Ultra DNA Library Prep Kit for Illumina (New England Biolabs). Illumina sequencing adapters were ligated onto the ends and following purification with AMPure XP beads, the DNA was PCR amplified (8 cycles) using indexed PCR primers (NEBNext Multiplex Oligos for Illumina, New England Biolabs) and the Illumina PE1.0 PCR primer. After purification using AMPure XP beads, quality assays were performed with a Bioanalyzer DNA 1000 chip (Agilent) and the Qubit dsDNA assay to determine the average fragment sizes and concentrations.

Target enrichment and sequencing

Enrichment and sequencing was carried out as described by Witek et al. (2016)11. Briefly, DNA sequencing libraries was enriched according to the MYbaits protocol (MYbaits User Manual version 2.3.1) and using MYbaits reagents (MYcroarray). Briefly, 500 ng of the prepped library was hybridized in hybridization buffer (10x SSPE, 10X Denhardt’s solution, 10 mM EDTA, 0.2% SDS) to the biotinylated RNA baits for 20 h at 65°C on a thermocycler. After hybridization bound DNA was recovered using magnetic streptavidin-coated beads as follows: the hybridization mix was added to 30 µL Dynabeads MyOne Streptavidin C1 (Invitrogen, Life Technologies) that had been washed 3 times and resuspended in binding buffer (1 M NaCl; 10 mM Tris-HCl, pH 7.5; 1 mM EDTA). After 30 m at 65°C, beads were pulled down and washed three times at 65°C for 10 m with 0.02% SSC/0.1% SDS followed by resuspension in 30 µL of nuclease-free water. Library was then PCR amplified (26 cycles) using Kapa HiFi HotStart Ready Mix (Kapa Biosystems) and Illumina P5 and P7 primers. The amplified library was size fractionated with the Sage Scientific Electrophoretic Lateral Fractionator (SageELF, Sage Science) using a 0.75% SageELF agarose gel cassette. Fractions with size distribution between 3 and 4 kb were pooled and purified with AMPure PB beads (Pacific Biosciences). Then, library was assembled for PacBio sequencing using the SMRTbell Template Prep Kit 1.0 (Pacific Biosciences) according the 2-kb Template Preparation and Sequencing protocol (www.pacificbiosciences.com/support/pubmap/documentation.html). PacBio RSII sequencing using C4-P6 chemistry was performed at the Earlham Institute (Norwich, UK), using four SMRT cells for each barley accession Baronesse and oat accession Victoria.

PacBio assembly

PacBio circular consensus reads with at least three passes were used for genome assembly. Reads were trimmed to remove the adapter sequence (first and last 70 bp) and size selected to reads less than 4kb. We used Geneious50 (v10.2.3) De Novo Assembly with the following Custom Sensitivity parameters for assembly: don't merge variants with coverage over approximately 6, merge homopolymer variants, allow gaps up to a maximum of 15% gaps per read, word length of 14, minimum overlap of 250 bp, ignore words repeated more than 200 times, 5% maximum mismatches per read, maximum gap size of 2, minimum overlap identity of 90%, index word length 12, reanalyze threshold of 8, and maximum ambiguity of 4.

Phylogenetic analyses

Exo70 coding and protein sequences were accessed from Department of Energy-Joint Genome Institute Phytozome database (https://phytozome.jgi.doe.gov), A. thaliana gene sequence was accessed from TAIR (www.arabidopsis.org), and barley Exo70 gene families from the recently updated 2017 genome51. Substantial curation of the barley Exo70 gene family was required and incorporated gene models from the 2012 genome52. Outgroups for complete Exo70, Exo70F, and Exo70F1 phylogenetic trees were Saccharomyces cerevisiae Exo70 protein (YJL085W), A. thaliana Exo70F1 protein (AT5G50380), and O. sativa Exo70F1 (Os01g69230.1), respectively. MUSCLE53 (v3.8.31) and PRANK54 (v.140603) were used for protein and codon-based sequence alignment using default parameters, respectively. Curation of the multiple sequence alignment for complete Exo70 and Exo70F gene family was used to remove sequences with less than 40% of the breadth of the alignment and to remove positions with more than 60% missing data. Gene families were identified based on bootstrap support in the phylogenetic tree (Fig. 2a), incorporating a previous annotation performed on the Exo70 gene family16 (Supplementary Data 4). We required that 90% sequence coverage for inclusion in the Exo70F1 phylogenetic tree. RAxML55 (v8.2.9) was used for phylogenetic tree construction using the PROTGAMMAJTT and GTRCAT models for protein and coding sequence alignment, respectively. Bootstrap support was determined for all phylogenetic trees, using a convergence test to confirm sufficient sampling.

Exo70F1 homologs were identified from diverse Poales species56 using BLAST+ (Supplementary Data 1). The only species found without a non-integrated Exo70F1 was H. lanatus. Alignment of H. lanatus RNAseq reads to Dactylis glomerata Exo70F1 and de novo assembly using Geneious was used to reconstruct H. lanatus Exo70F1. Multiple sequence alignment and phylogenetic analysis of Exo70F gene families was used to establish Exo70F1 orthology (Supplementary Fig. 5). The site of sequence conservation between integrated and non-integrated Exo70F1 was determined based on codon-based sequence alignment.

Molecular evolutionary analyses

Molecular evolutionary analyses were performed with PAML57 (v4.8) codeml. The F3x4 codon frequency model was used for all analyses. Codon-based sequence alignment using PRANK, phylogenetic analysis using RAxML, and codeml were used to estimate ω (dN/dS) for all Exo70 gene families (Supplementary Table 2). For molecular evolutionary analyses of the non-integrated and integrated Exo70F1 gene family, four hypotheses were tested: H0: a single rate of ω0 across the entire tree, H1: a different rate of ω1 for the initial branch of the integrated Exo70F1, H2: a different rate of ωa for integrated Exo70F1 compared to non-integrated Exo70F1 (ω0), and H3: three different rates of ω: ω0, non-integrated Exo70F1 outside of the Poeae and Triticeae, ωa, integrated Exo70F1, and ωb, non-integrated Exo70F1 from the Poeae and Triticeae (Supplementary Fig. 1, Supplementary Table 3).

Data

All high-throughput sequencing data, de novo transcriptome assemblies, and de novo assembly of the NLR gene space of barley and oat are deposited in the NCBI BioProject PRJNA378334, PRJNA378723, PRJNA422803, and PRJNA422986. De novo transcriptome assemblies for publically available RNAseq data, multiple sequence alignments, and phylogenetic trees in Newick format are available from figshare (https://figshare.com). Scripts, analysis pipeline, and details associated with Exo70 gene family curation can be found on the GitHub repository https://github.com/matthewmoscou/Exo70 (v1.0).

Author contributions

Conceived and designed the experiments: H.J.B., J.L., and M.J.M. Performed transcriptome sequencing and sequence capture: I.H-P. Performed bioinformatic analyses: H.J.B., S.H., and M.J.M. Analysed the data: H.J.B., S.H., J.L., and M.J.M. Wrote the paper: H.J.B. and M.J.M.

Competing interests

The authors declare no competing financial interests.

Acknowledgements

The authors thank Tom Wolpert for excellent advice, enthusiasm, and curiosity, Phon Green for management of plant materials, William Jackson for the development of initial bioinformatic approaches for assembly of NLR gene space, Kamil Witek for advice on sequence capture and PacBio library preparation, Ksenia Krasileva for discussion on NLR-ID evolution, Simon Griffiths and Sophien Kamoun for their critical input, and Ryohei Terauchi and Hiromasa Saitoh for their open dialogue of science. This work was funded using core funding from the Gatsby Foundation and a BBSRC Institute Strategic Programme (BB/P012574/1). Additional support from the Perry Foundation was provided as a PhD studentship to HJB. JL was funded by grant 2016-67013-24736 from the USDA National Institute of Food and Agriculture.

References

  1. 1.↵
    Bergelson, J., Kreitman, M., Stahl, E.A. & Tian, D. Evolutionary dynamics of plant R-genes. Science 292, 2281–5 (2001).
    OpenUrlAbstract/FREE Full Text
  2. 2.↵
    Klein, J., Sato, A., Nagl, S. & O'hUigin, C. Molecular trans-species polymorphism. Annual Review of Ecology and Systematics 29, 1-+ (1998).
    OpenUrlCrossRefWeb of Science
  3. 3.↵
    Wei, F., Wing, R.A. & Wise, R.P. Genome dynamics and evolution of the Mla (powdery mildew) resistance locus in barley. Plant Cell 14, 1903–17 (2002).
    OpenUrlAbstract/FREE Full Text
  4. 4.↵
    Jones, J.D. & Dangl, J.L. The plant immune system. Nature 444, 323–9 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  5. 5.↵
    Jones, J.D., Vance, R.E. & Dangl, J.L. Intracellular innate immune surveillance devices in plants and animals. Science 354(2016).
  6. 6.↵
    Cesari, S. Multiple strategies for pathogen perception by plant immune receptors. New Phytologist (2017).
  7. 7.
    Kroj, T., Chanclud, E., Michel-Romiti, C., Grand, X. & Morel, J.B. Integration of decoy domains derived from protein targets of pathogen effectors into plant immune receptors is widespread. New Phytol 210, 618–26 (2016).
    OpenUrlCrossRefPubMed
  8. 8.↵
    Sarris, P.F., Cevik, V., Dagdas, G., Jones, J.D. & Krasileva, K.V. Comparative analysis of plant immune receptor architectures uncovers host proteins likely targeted by pathogens. BMC Biol 14, 8 (2016).
    OpenUrlCrossRefPubMed
  9. 9.↵
    Seeholzer, S. et al. Diversity at the Mla powdery mildew resistance locus from cultivated barley reveals sites of positive selection. Molecular Plant-Microbe Interactions 23, 497–509 (2010).
    OpenUrlCrossRefPubMedWeb of Science
  10. 10.↵
    Shen, Q.H. et al. Recognition specificity and RAR1/SGT1 dependence in barley Mla disease resistance genes to the powdery mildew fungus. Plant Cell 15, 732–44 (2003).
    OpenUrlAbstract/FREE Full Text
  11. 11.↵
    Witek, K. et al. Accelerated cloning of a potato late blight-resistance gene using RenSeq and SMRT sequencing. Nature Biotechnology 34, 656–60 (2016).
    OpenUrlCrossRefPubMed
  12. 12.↵
    Okuyama, Y. et al. A multifaceted genomics approach allows the isolation of the rice Pia-blast resistance gene consisting of two adjacent NBS-LRR protein genes. Plant Journal 66, 467–79 (2011).
    OpenUrlCrossRefPubMedWeb of Science
  13. 13.↵
    Synek, L. et al. AtEXO70A1, a member of a family of putative exocyst subunits specifically expanded in land plants, is important for polar growth and plant development. Plant Journal 48, 54–72 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  14. 14.
    He, B., Xi, F., Zhang, X., Zhang, J. & Guo, W. Exo70 interacts with phospholipids and mediates the targeting of the exocyst to the plasma membrane. EMBO Journal 26, 4053–4065 (2007).
    OpenUrlCrossRefPubMedWeb of Science
  15. 15.↵
    Heider, M.R. & Munson, M. Exorcising the exocyst complex. Traffic 13, 898–907 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.↵
    Cvrčková, F. et al. Evolution of the land plant exocyst complexes. Frontiers in Plant Science 3, 159 (2012).
    OpenUrl
  17. 17.↵
    Lynch, M. & Conery, J.S. The evolutionary fate and consequences of duplicate genes. Science 290, 1151–5 (2000).
    OpenUrlAbstract/FREE Full Text
  18. 18.↵
    Sekereš, J. et al. Analysis of exocyst subunit EXO70 family reveals distinct membrane polar domains in tobacco pollen tubes. Plant Physiology 173, 1659–1675 (2017).
    OpenUrlAbstract/FREE Full Text
  19. 19.↵
    Gu, Y., Zavaliev, R. & Dong, X. Membrane trafficking in plant immunity. Molecular Plant 10, 1026–1034 (2017).
    OpenUrl
  20. 20.↵
    Ostertag, M., Stammler, J., Douchkov, D., Eichmann, R. & Huckelhoven, R. The conserved oligomeric Golgi complex is involved in penetration resistance of barley to the barley powdery mildew fungus. Molecular Plant Pathology 14, 230–40 (2013).
    OpenUrl
  21. 21.↵
    Fujisaki, K. et al. Rice Exo70 interacts with a fungal effector, AVR-Pii, and is required for AVR-Pii-triggered immunity. Plant Journal 83, 875–87 (2015).
    OpenUrlCrossRefPubMed
  22. 22.↵
    Fujisaki, K. et al. An unconventional NOI/RIN4 domain of a rice NLR protein binds host EXO70 protein to confer fungal immunity. bioRxiv (2017).
  23. 23.↵
    Periyannan, S. et al. The gene Sr33, an ortholog of barley Mla genes, encodes resistance to wheat stem rust race Ug99. Science 341, 786–8 (2013).
    OpenUrlAbstract/FREE Full Text
  24. 24.↵
    Nishijima, R., Yoshida, K., Motoi, Y., Sato, K. & Takumi, S. Genome-wide identification of novel genetic markers from RNA sequencing assembly of diverse Aegilops tauschii accessions. Mol Genet Genomics 291, 1681–94 (2016).
    OpenUrl
  25. 25.↵
    The Timetree of Life, (Oxford University Press, Oxford, UK, 2009).
  26. 26.↵
    Cesari, S., Bernoux, M., Moncuquet, P., Kroj, T. & Dodds, P.N. A novel conserved mechanism for plant NLR protein pairs: the "integrated decoy" hypothesis. Frontiers in Plant Science 5, 606 (2014).
    OpenUrl
  27. 27.↵
    Bailey, P.C. et al. Dominant integration locus drives continuous diversification of plant immune receptors with exogenous domain fusions. bioRxiv (2017).
  28. 28.↵
    Wu, Q. et al. Long-term balancing selection contributes to adaptation in Arabidopsis and its relatives. Genome Biology 18, 217 (2017).
    OpenUrl
  29. 29.↵
    Karasov, T.L. et al. The long-term maintenance of a resistance polymorphism through diffuse interactions. Nature 512, 436–440 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  30. 30.↵
    Inukai, T., Vales, M.I., Hori, K., Sato, K. & Hayes, P.M. RMo1 confers blast resistance in barley and is located within the complex of resistance genes containing Mla, a powdery mildew resistance gene. Molecular Plant-Microbe Interactions 19, 1034–41 (2006).
    OpenUrlCrossRefPubMedWeb of Science
  31. 31.↵
    Dawson, A.M. et al. Isolation and fine mapping of Rps6: an intermediate host resistance gene in barley to wheat stripe rust. Theoretical and Applied Genetics 129, 831–843 (2016).
    OpenUrl
  32. 32.↵
    Andrews, S. FastQC: a quality control tool for high throughput sequence data. (http://www.bioinformatics.babraham.ac.uk/projects/fastqc, 2010).
  33. 33.↵
    Grabherr, M.G. et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nature Biotechnology 29, 644–52 (2011).
    OpenUrlCrossRefPubMed
  34. 34.↵
    Bolger, A.M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–20 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    Camacho, C. et al. BLAST+: architecture and applications. BMC Bioinformatics 10, 421 (2009).
    OpenUrlCrossRefPubMed
  36. 36.↵
    Halterman, D., Zhou, F., Wei, F., Wise, R.P. & Schulze-Lefert, P. The MLA6 coiled-coil, NBS-LRR protein confers AvrMla6-dependent resistance specificity to Blumeria graminis f. sp. hordei in barley and wheat. Plant Journal 25, 335–48 (2001).
    OpenUrlCrossRefPubMedWeb of Science
  37. 37.
    Halterman, D.A. & Wise, R.P. A single-amino acid substitution in the sixth leucine-rich repeat of barley MLA6 and MLA13 alleviates dependence on RAR1 for disease resistance signaling. Plant Journal 38, 215–26 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  38. 38.↵
    Zhou, F. et al. Cell-autonomous expression of barley Mla1 confers race-specific resistance to the powdery mildew fungus via a Rar1 independent signaling pathway. Plant Cell 13, 337–350 (2001).
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    Büschges, R. et al. The barley Mlo gene: a novel control element of plant pathogen resistance. Cell 88, 695–705 (1997).
    OpenUrlCrossRefPubMedWeb of Science
  40. 40.↵
    Brueggeman, R. et al. The barley stem rust-resistance gene Rpg1 is a novel disease-resistance gene with homology to receptor kinases. Proceedings of the National Academy of Sciences, USA 99, 9328–33 (2002).
    OpenUrlAbstract/FREE Full Text
  41. 41.↵
    Brueggeman, R. et al. The stem rust resistance gene Rpg5 encodes a protein with nucleotide-binding-site, leucine-rich, and protein kinase domains. Proceedings of the National Academy of Sciences, USA 105, 14970–5 (2008).
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    Haas, B.J. et al. De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nature Protocols 8, 1494–512 (2013).
    OpenUrl
  43. 43.↵
    Jupe, F. et al. Identification and localisation of the NB-LRR gene family within the potato genome. BMC Genomics 13, 75 (2012).
    OpenUrlCrossRefPubMed
  44. 44.↵
    Bailey, T.L. et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res 37, W202–8 (2009).
    OpenUrlCrossRefPubMedWeb of Science
  45. 45.↵
    Li, J. et al. Unique evolutionary pattern of numbers of gramineous NBS-LRR genes. Molecular Genetics and Genomics 283, 427–38 (2010).
    OpenUrlCrossRefPubMedWeb of Science
  46. 46.↵
    Tan, S. & Wu, S. Genome wide analysis of nucleotide-binding site disease resistance genes in Brachypodium distachyon. Comparative and Functional Genomics 2012, 418208 (2012).
    OpenUrl
  47. 47.↵
    Meyers, B.C., Kozik, A., Griego, A., Kuang, H. & Michelmore, R.W. Genome-wide analysis of NBS-LRR-encoding genes in Arabidopsis. Plant Cell 15, 809–34 (2003).
    OpenUrlAbstract/FREE Full Text
  48. 48.↵
    Smit, A.F.A., Hubley, R. & Green, P. RepeatMasker Open-4.0. (http://www.repeatmasker.org/, 2013–2015).
  49. 49.↵
    Stewart, C.N., Jr.. & Via, L.E. A rapid CTAB DNA isolation technique useful for RAPD fingerprinting and other PCR applications. Biotechniques 14, 748–750 (1993).
    OpenUrlPubMedWeb of Science
  50. 50.↵
    Kearse, M. et al. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28, 1647–9 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  51. 51.↵
    Mascher, M. et al. A chromosome conformation capture ordered sequence of the barley genome. Nature 544, 427–433 (2017).
    OpenUrlCrossRefPubMed
  52. 52.↵
    IBGSC. A physical, genetic and functional sequence assembly of the barley genome. Nature 491, 711–6 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  53. 53.↵
    Edgar, R.C. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32, 1792–7 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  54. 54.↵
    Löytynoja, A. Phylogeny-aware alignment with PRANK. Methods in Molecular Biology 1079, 155–170 (2014).
    OpenUrl
  55. 55.↵
    Stamatakis, A., Ludwig, T. & Meier, H. RAxML-III: a fast program for maximum likelihood-based inference of large phylogenetic trees. Bioinformatics 21, 456–63 (2005).
    OpenUrlCrossRefPubMedWeb of Science
  56. 56.↵
    McKain, M.R. et al. A phylogenomic assessment of ancient polyploidy and genome evolution across the Poales. Genome Biology and Evolution 8, 1150–64 (2016).
    OpenUrlCrossRefPubMed
  57. 57.↵
    Yang, Z. PAML 4: phylogenetic analysis by maximum likelihood. Molecular Biology and Evolution 24, 1586–91 (2007).
    OpenUrlCrossRefPubMedWeb of Science
Back to top
Posted January 08, 2018.
Download PDF

Supplementary Material

Email
An ancient integration in a plant NLR is maintained as a trans-species polymorphism
Helen J. Brabham, Inmaculada Hernández-Pinzón, Samuel Holden, Jennifer Lorang, Matthew J. Moscou
bioRxiv 239541; doi: https://doi.org/10.1101/239541
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area