A landscape of resistance gene analogs in sour cherry (Prunus cerasus L.)

Abstract

Objective

This research aims to analyze the presence and distribution of resistance genes in the avium and fruticosa subgenomes of Prunus cerasus through computational methods and bioinformatics tools.

Results

Analysis of genome and transcriptome sequencing data revealed a total of 19,570 transcripts with at least one resistance gene domain in Prunus cerasus subgenome avium and 19,142 in Prunus cerasus subgenome fruticosa. Key findings include the identification of 804 “complete” resistance gene transcripts in Prunus cerasus subgenome avium and 817 in Prunus cerasus subgenome fruticosa, with distinct distributions of resistance gene classes observed between the subgenomes. Phylogenetic analysis showed clustering of resistance genes, and unique resistance proteins were identified in each subgenome. Functional annotation comparisons with Arabidopsis thaliana highlighted shared and unique resistance genes, emphasizing the complexity of disease resistance in cherry species. Additionally, a higher diversity of RLKs and RLPs was observed, with 504 transcripts identified and 18 showing similarity to known reference genes.

Introduction

Resistance breeding against biotic stressors is increasingly important particularly because of the demand for resilient and high-quality varieties and less pesticide application. While breeding to develop such resilient varieties in annual crops can take between 10 and 20 years [1], it can last several decades in perennial crops such as fruit trees [2]. Initial successes in resistance breeding have been achieved in recent years in various fruit crops such as apple or peach [3, 4]. Breeding resistant sour cherries particularly against major pathogens like cherry leaf spot (Blumeriella jaapii), brown rot (Monilinia laxa), and bacterial canker (Pseudomonas syringae pv. morsprunorum) is one of the primary goals in sour cherry breeding [5, 6]. While many common cultivars, such as ‘Schattenmorelle,’ are susceptible to these pathogens, other sources of resistance have been identified, offering potential for breeding more tolerant or resistant varieties [6,7,8]. Several breeding programs have successfully released cultivars that exhibit high levels of resistance or tolerance to these diseases [5, 8, 9]. The success of a resistance-breeding program depends on the availability of not only a wide variety of genetic resources but also tools for inheritance studies of resistance genes and loci such as genomic information in the form of genome sequences and genetic maps with high-throughput genome-wide genetic markers. With the help of this information, breeders can specifically identify and characterize potential resistance (R) genes in candidate regions and develop diagnostic markers for targeted selection. Genome sequences for sour cherry varieties have recently been published [10,11,12]. The sour cherry (Prunus cerasus L.) is a tetraploid species (2n = 4x = 32) that originated from hybridization between an unreduced pollen grain of diploid sweet cherry (P. avium, 2n = 2x = 16) and a tetraploid ground cherry (P. fruticosa, 2n = 4x = 32). Therefore the genome of sour cherry is characterized by the presence of two subgenomes: P. cerasus subgenome avium and P. cerasus subgenome fruticosa. These subgenomes contribute to the species estimated 599 Mbp genome. The two main sour cherry cultivars (‘Schattenmorelle’ in Europe and ‘Montmorency’ in USA) have been sequenced and several differences were found in terms of subgenome composition and genomic complexity [10, 12]. ‘Montmorency’ is trigenomic, containing two distinct subgenomes from a Prunus fruticosa-like ancestor (A and A’) and two identical subgenomes from a Prunus avium-like ancestor (BB). In contrast, for ‘Schattenmorelle’ only two consensus subgenomes from the P. avium-like ancestor (B) and one from P. fruticosa-like ancestor (A) were assembled. However, there is no information about the nature of the resistance genes in the sour cherry genome.

Plants possess an innate immune system that essentially consists of two branches [13] to defend against pathogens, one of which is based on the recognition of pathogens by resistance proteins encoded by R-genes in individual cells. The major classes of plant resistance genes are grouped based on their functional domains e.g. NBS, LRR, CC, TIR, kinase etc. [14]. The main class of resistance proteins found in plants has an NB - nucleotide-binding site and a leucine-rich repeat (LRR) domain. NB-LRR genes can generally be divided into 3 types, which differ in the presence of a Toll/interleukin-1 receptor domain at the N-terminal end. NB-LRR genes lacking a TIR domain usually have a coiled-coil domain CC instead. A third type is RPW8-NBS-LRR, which act in downstream defense signal transduction [15, 16]. The TM-LRR is divided into 2 types, receptor like kinases (RLK) and receptor like proteins (RLP) [15, 16]. While the occurrence of resistance genes in the genomes of various Prunus species has been investigated [17,18,19], this has not yet been done for sour cherries. The aim of this research note is to examine the data published by Wöhner and colleagues [11] regarding the classes of resistance genes present in the ‘Schattenmorelle’ sour cherry genome (subgenome avium and subgenome fruticosa) and to provide their positions.

Methods

Genome and transcriptome sequencing, along with structural and functional annotation, as well as InterProScan analysis, were previously performed as described in Wöhner et al. [10]. InterProScan results for P. cerasus from previously published data [11] were downloaded from the OpenAgrar repository (www.openagrar.de) and filtered in Microsoft Excel. Pfam and Superfamily tabsheets were filtered for domain-specific entries according to [20]. Entries labelled with Pfam identifiers PF18052 indicated a CC domain, PF00931 indicated the NB domain, and PF01582 indicated the TIR domain. Leucine-rich repeats were filtered out of the Superfamily tab sheet using the identifiers SSF52058, SSF52047, and “leucine-rich repeat”. Kinase domains were filtered using the GO:0004672 term in the Pfam tabsheet, and TRANSMEMBRANE or SIGNAL_PEPTIDE was used for filtering specific domains in Phobius obtained data. The LysM protein motif was filtered using IPR018392 in data obtained from ProSiteProfiles tabsheets. The data was sorted and assigned to transcript IDs obtained from the annotation of P. cerasus ‘Schattenmorelle’ for P. cerasus subgenome avium and P. cerasus subgenome fruticosa. Only full-length genes were summarized, and the number of genes per resistance gene class was determined. For annotation of NB-LRRs, a BLAST database was created using proteins obtained from UniProt, filtered by Gene Ontology terms (0051707 - response to other organism; 0006952 - defense response; taxonomy Embryophyta). The final transcripts of full-length resistance gene candidates were blasted against this database using BLAST p [21,22,23] on the Galaxy server (www.usegalaxy.org). For annotation of RLKs and RLPs, functional characterized reference genes from [15, 16] and Uniprot proteins filtered to GO:0051707 (response to other organism), GO:0006952 (defense response), GO:0004672 (kinase domain) and taxonomy Embryophyta were used to create a BLAST database. Final RLP/RLK transcripts were blasted against this database. All final full-length sequences (581 NB-LRR and 2157 RLP/RLK transcripts in total) were aligned separately using MAFFT (gap extend penalty 0.0, gap open 1.53, matrix BLOSUM62) on Galaxy [24]. The final alignment was used for phylogenetic tree construction using the Neighbor-Joining method [25,26,27]. The optimal tree is shown, with the percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) shown next to the branches. This analysis was performed with MEGA11 [28]. The final tree was color-labelled using the Interactive Tree of Life tool [29]. The full-length resistance transcripts were displayed using MapChart software [30].

Results and discussion

A comprehensive analysis of resistance genes in P. cerasus subgenome avium and P. cerasus subgenome fruticosa revealed intriguing insights into the putative genetic basis of disease resistance in cherry species. A total of 19,570 transcripts with at least one single domain (CC, TIR, RWP8, NB, LRR, SP, TM, LysM, K) were identified in the P. cerasus subgenome avium, while 19,142 transcripts were found in the P. cerasus subgenome fruticosa. Among these, 804 (4.1%) were detected as “complete” (TIR-NB-LRR, CC-NB-LRR, RPW8-NB-LRR, RLK, RLP) in P. cerasus subgenome avium, and 817 (4.3%) in P. cerasus subgenome fruticosa. Classification of the resistance genes showed distinct distributions among the subgenomes. Since multiple splice variants may occur within one gene, we determined the number of genes and not transcripts per resistance gene type. In P. cerasus subgenome avium, 103 CC-NB-LRR, 10 RPW8-NB-LRR, 69 TIR-NB-LRR, 20 LysM, 198 RLK and 114 RLP genes were identified, whereas P. cerasus subgenome fruticosa comprised 111 CC-NB-LRR, 13 RPW8-NB-LRR, 74 TIR-NB-LRR, 21 LysM, 212 RLK and 112 RPL genes (Fig. 1). Clustering for phylogenetic analysis was performed with the transcripts of all genes determined. Phylogenetic analysis revealed clustering of resistance genes, with 34 RPW8-NB-LRR (red), 266 TIR-NB-LRR (blue), and 281 CC-NB-LRR (yellow) genes observed in a rooted tree (Fig. 2, Table S1). The plotting of the identified full-length resistance genes revealed 16 clusters (≥ 2 genes) in P. cerasus subgenome avium and 18 in P. cerasus subgenome fruticosa (Figure S1). Notably, a RPW8-NB-LRR cluster was identified on chromosome 7 in both subgenomes. Comparing the annotations obtained from Arabidopsis thaliana, several resistance proteins were found in both avium and fruticosa subgenomes, while some were unique to each of the subgenomes. In both subgenomes, annotations for Disease Resistance-like Proteins (DSC1 and DSC2) [31] and the Suppressor of npr1-1, Constitutive 1 (SNC1) gene [32], as well as Disease Resistance Protein TAO1 [33] were identified. Unique resistance proteins were also found in each subgenome. For instance, only the avium subgenome contained a homolog to Disease Resistance Protein RML1B, conferring resistance to Leptosphaeria maculans in Arabidopsis thaliana [34]. Similarly, only the fruticosa subgenome featured Disease Resistance Protein RPP13-like Protein 1, involved in recognizing specific pathogens such as Pseudomonas syringae [35]. The presence of shared and unique resistance genes underscores the complexity of disease resistance in cherry species. A higher diversity is given for RLKs and RLPs, since these play not only crucial roles in immunity but also in plant development and growth [36]. A total of 2,191 RLPs and RLKs were identified and BLAST analyses against the uniprot for the obtained proteins led to the identification of 504 RLP/RLK transcripts in P. cerasus. BLAST against the 22 references summarized by [9] revealed 18 transcripts as best hits (Table 1; Fig. 3). The highest percentage identity was obtained for PCE_F_Chro1G0077800.1 with 83.8% to PTO interacting protein 1 (Pti1) from Solanum lycopersicum [31] and for PCE_F_Chro4G0199100.1 with 81.7% to PBS1 from Arabidopsis thaliana [38]. The tomato gene Pti1 encodes a serine/threonine kinase, phosphorylated by PTO, and plays a pivotal role in the hypersensitive response (HR)-mediated signaling cascade [37]. The Arabidopsis PBS1 resistance gene encodes a protein kinase essential for RPS5-mediated plant defense, with AvrPphB cleavage of PBS1 and kinase activity both necessary for triggering RPS5-mediated resistance, while also contributing to PAMP-triggered immunity (PTI) signaling downstream of FLS2 [36]. The analysis of resistance genes in P. cerasus subgenomes reveals crucial insights for breeding programs [39,40,41]. The identification of 19,570 and 19,142 transcripts in the subgenome avium and subgenome fruticosa, respectively, with notable clusters of resistance genes like RPW8-NB-LRR, TIR-NB-LRR, and CC-NB-LRR, provides a strong basis for targeted breeding. Unique genes, such as RML1B in the avium subgenome and RPP13-like in the fruticosa subgenome, offer potential for developing cultivars with enhanced resistance to specific pathogens. The diversity of RLKs and RLPs further supports breeding efforts aimed at improving disease resistance and overall plant health [42, 43]. These findings can guide the development of more resilient sour cherry varieties, reduce chemical treatments, and advance crop sustainability. Building on these findings, functional validation of the identified resistance genes is crucial to confirm their roles in disease resistance and to understand their potential applications in breeding programs. Identification of resistance loci through QTL mapping or GWAS is essential for pinpointing genomic regions associated with resistance. Finally, fine mapping will narrow down candidate genes, qRT-PCR will validate their expression, and functional characterization via transient expression or stable transgenesis will determine their resistance efficacy [44,45,46,47].

Fig. 1

Number of identified full length resistance genes in the two subgenomes (Prunus cerasus subgenome avium and Prunus cerasus subgenome fruticosa) of P. cerasus ‘Schattenmorelle’. The arrow indicates the presence of RPW8-NB-LRR genes of chromosome 7 in both subgenomes

Fig. 2

A phylogenetic tree of the 581 identified full length resistance transcripts (RPW8-NB-LRR – red; TIR-NB-LRR – blue; CC-NB-LRR – yellow)

Table 1 RLKs and RLP homeologs from other studies found in P. Cerasus ‘Schattenmorelle’

Fig. 3

A phylogenetic tree of the 22 reference RLKs and RLPs Sehkwal et al. [15] and 18 transcripts from P. cerasus ‘Schattenmorelle’

Limitations

It is important to acknowledge that while our study has provided valuable insights into the presence and distribution of resistance genes in the P. cerasus subgenomes avium and fruticosa; there are inherent limitations that should be considered. One key limitation is that our analysis was primarily based on computational methods, utilizing genome and transcriptome sequencing data along with bioinformatics tools for annotation and classification of resistance genes. While this approach allowed us to identify a significant number of candidate resistance genes, it is essential to recognize that the presence of these genes in the dataset does not inherently imply functional proof of their effect on known resistances. Functional validation through experimental approaches is crucial to confirm the role of these genes in mediating specific disease resistances in cherry species. Additionally, while we compared our findings with annotations from Arabidopsis thaliana to infer potential functions of the identified resistance genes, it is important to acknowledge that functional roles may vary amongst different plant species and environments. Subgenome bias and dosage effects may complicate the interpretation of functional validation results, particularly if gene dosage influences the observed phenotypes. The susceptibility of the Schattenmorelle cultivar suggests that the identified resistance genes might not be specific to the mentioned pathogens (e.g. Monilinia laxa, Blumeriella jaapii etc.), requiring validation in other genotypes (transgenesis). High synteny across Prunus species underscores the need for comparative studies to identify structural variations in candidate genes that may affect their function and resistance potential.