Haplotype Analysis Sheds Light on the Genetic Evolution of the Powdery Mildew Resistance Locus Pm60 in Triticum Species

Abstract

Wheat powdery mildew (Blumeria graminis f. sp. tritici, Bgt, recently clarified as B. graminis s. str.), is one of the most destructive diseases of wheat. Pm60 is a nucleotide-binding leucine-rich repeat (NLR) gene that confers race-specific resistance to Bgt. Allelic variants (Pm60, Pm60a, and Pm60b) were found in Triticum urartu and T. dicoccoides, the wild progenitors of wheat. In the present study, we studied the diversity of the Pm60 locus in a large set of wheat germplasm and found 20 tetraploid wheats harboring the Pm60 alleles, which correspond to three novel haplotypes (HapI–HapIII). HapI (Pm60 allele) and HapII (Pm60a allele) were present in domesticated tetraploid wheats, whereas HapIII (Pm60a allele) was identified in wild tetraploid T. araraticum. A sequence comparison of HapII and HapIII revealed that they differed by three SNPs and a GCC deletion. Results of the phylogenetic analysis revealed that HapII was more closely related to the functional haplotype MlIW172. Infection tests showed that HapII-carrying lines display a partial resistance response to Bgt#GH, while HapI was susceptible. Our results provide insights into the genetic evolution of the Pm60 locus and potential valuable alleles for powdery mildew resistance breeding.

1. Introduction

Wheat powdery mildew (Blumeria graminis f. sp. tritici, Bgt), recently clarified as B. graminis s. str. [1], is one of the most destructive foliar diseases that pose severe yield loss during the past decades worldwide [2]. Planting of Bgt-resistant cultivars is the most effective and environmentally safe strategy to control this disease [3]. However, wheat cultivars are prone to lose their resistance due to the rapid evolution of the pathogen and the emergence of new virulent Bgt isolates [4]. To date, over 69 loci for resistance to powdery mildew (Pm1–Pm69) have been formally documented and genetically mapped on different chromosomes in wheat [5,6]. Several Pm genes were identified from wheat wild relatives, such as Triticum urartu [7], T. dicoccoides [8], and Aegilops biuncialis [9].

Powdery mildew resistance gene Pm60 was identified from T. urartu accession PI428309 and characterized as nucleotide-binding leucine-rich repeat (NLR) protein [7]. Three allelic variants were found in T. urartu populations and designated as Pm60, Pm60a, and Pm60b. The coding region of Pm60a has a 240-nucleotide deletion, whereas Pm60b contains a 240-nucleotide insertion, as compared with the Pm60 sequence in PI428309, which resulted in the loss or addition of two of the leucine-rich repeat (LRR) motifs of Pm60. Although the deletion or insertion did not affect the resistance to Bgt isolate E09 [7], the deletion of the two LRR motifs in Pm60a substantially narrowed the resistance spectrum [10]. For example, Pm60 was highly resistant to 43 out of 54 Bgt isolates tested, whereas Pm60a was only resistant to 5 of them [10]. A nonfunctional allele of Pm60a, designated as Pm60a’, was presented in multiple susceptible T. urartu accessions. The Pm60a’ sequence was 98.52% identical to Pm60a, with 58 single nucleotide polymorphisms (SNPs) and a 3-nucleotide deletion [11].

The wild emmer wheat (T. dicoccoides) harbors the Pm60 allele (TdPm60), which is an ortholog of Pm60 from T. urartu. TdPm60 was found to be present in the wild emmer donor lines of resistance genes PmG16 [8], MlWE18 [12], MlIW72 [13], and MlIW172 [14]. Among them, MlWE18 and MlIW172 were confirmed to be the functional alleles of Pm60 by using ethyl methanesulfonate (EMS) mutagenesis and genetic transformation [12,14]. TdPm60 constitutes a strong candidate for PmG16 mildew resistance [8]. The results of the haplotype analysis of Pm60 alleles revealed diversifications in sequence variation in the locus and presence/absence polymorphisms in wild emmer wheat populations [14]. MlWE18 and MlIW172, together with additional eight haplotypes (Hap1–Hp8) of TdPm60, were defined in wild emmer [14], which shared four common single nucleotide variations between the Pm60 alleles from T. dicoccoides and T. urartu. A different resistance spectrum of functional Pm60 alleles was observed when inoculated with 20 Bgt isolates sampled from different locations of China [14], implying that these Pm60 alleles and haplotypes are promising for breeding powdery mildew resistance wheat cultivars.

Previous studies focused on the characterization of Pm60 alleles in either diploid T. urartu or tetraploid T. dicoccoides accessions. Whether there are further unidentified Pm60 alleles in other wheat species has not been thoroughly assayed. In the present study, we screened a large collection of wheat germplasm with Pm60 gene-based primers. This screening revealed that 20 genotypes harbor one of the Pm60 alleles. The results of the sequence analysis identified three novel Pm60 haplotypes from tetraploid wheats. Our results shed light on the evolution of the Pm60 locus and provide potential valuable alleles for powdery mildew resistance breeding.

2. Materials and Methods

2.1. Plant Materials

Two hundred thirty-eight genotypes of different wheat species (Table S1) were used in the current study, including diploid Ae. searsii, Ae. bicornis, and Ae. sharonensis; tetraploid sources T. timopheevii, T. armeniacum, T. carthlicum, T. dicoccon, T. dicoccoides, T. durum, T. palaeocolchicum, T. polonicum, T. turanicum, and T. turgidum; as well as hexaploid wheat T. spelta, T. sphaerococcum, T. aestivum, T. compactum, T. macha, T. vavilovii, and T. zhukovskyi. The seeds of those wheat accessions were originally obtained from the National Small Grains Collection (NSGC, USDA) (PI and CItr numbers) and Triticeae Research Institute, Sichuan Agricultural University, Chengdu, China (AS numbers).

2.2. Genomic DNA Extraction and Marker Analysis

Wheat seeds were wrapped in a moist filter paper and put in a chamber at a temperature of 4 °C for 3 days and transferred into a growth chamber at a temperature of 10 °C during the dark period (8 h) and 20 °C during the light period (16h) for 10 days. Total genomic DNA was prepared from seedling leaves using the cetyltrimethylammonium bromide (CTAB) method as described by Zhang et al. [15]. The DNA samples were checked for integrity by 0.8% agarose gel electrophoresis, followed by staining with GelRed Nucleic Acid Gel Stain (Biotium, Hayward, CA, USA). DNA concentration was assessed using a NanoDrop One spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Each PCR mixture (20 μL) contained 80–100 ng of template DNA, 300 nM of each of forward and reverse primers, and 10 μL of Taq Master Mix (P112-02, Vazyme Biotech Co., Nanjing, China). PCR amplification with the functional molecular marker for Pm60 (M-Pm60-S1) [11] was programmed as 94 °C for 5 min, 35 cycles of 94 °C for 35 s, 57 °C for 35 s, and 72 °C for 1 min, terminated after an extension at 72 °C for 7 min. Alternatively, the high-fidelity ExTaq polymerase (Takara, Dalian, China) was used for the amplification of the entire Pm60 region. Touchdown PCR amplification was carried out as described by Huang et al. [16]. PCR products were separated using 1.2% agarose gel electrophoresis, stained with GelRed Nucleic Acid Gel Stain (Biotium).

2.3. Sequence Analysis of Pm60

Pm60 alleles were previously isolated from T. urartu and T. dicoccoides accessions. To characterize whether Pm60 alleles are present in other wheat species, we screened a collection of 238 wheat germplasm using the PCR-based molecular marker M-Pm60-S1, which covers the LRR regions and can differentiate the Pm60, Pm60a, and Pm60b alleles by the size of amplified products (Figure 1A). The accessions positive for the amplification of the marker M-Pm60-S1 were selected to obtain the entire Pm60 sequences (from start to stop codons) using four primer pairs amplifying overlapping fragments (Figure 1A and Table S2), which were developed based on the published nucleotide sequence around the Pm60 locus in T. urartu [7]. The four PCR primer sets were also used for the sequencing of the Pm60 alleles in the present study. The sequencing was performed by Sangon Biotechnology Company (Chengdu, China). Sequences were assembled and subjected to multiple alignments using DNAMAN software version 6.0 (Lynnon Biosoft, San Ramon, CA, USA). The final sequences were manually corrected, and SNPs were visually checked on a chromatogram to ensure their quality.

2.4. Phylogenetic Analysis

The published Pm60 alleles and homologous sequences from cereals were selected for phylogenetic analyses. The Pm60 homologs in the phylogeny tree were identified from the National Centre for Biotechnology Information (NCBI) database and the Ensembl Plants website. Multiple sequence alignments were performed with BioEdit [17], and a neighbor-joining tree was constructed using MEGA software by the bootstrap method with 1000 replicates.

2.5. Powdery Mildew Resistance Response

Five Bgt isolates including Bgt#1-25, Bgt#15-9-4, Bgt#WJ, Bgt#SZS, and Bgt#GH were used to test the response of wheat accessions harboring different Pm60 haplotypes in the present study. Isolates Bgt#1-25 and Bgt#15-9-4 were kindly provided by Wenqi Shi from the Institute of Plant Protection and Soil Fertility, Hubei Academy of Agricultural Sciences. The isolates Bgt#WJ and Bgt#SZS were newly purified from bread wheat in the experimental fields of Sichuan Agricultural University and Sichuan Academy of Agricultural Sciences, respectively. The isolate Bgt#GH was purified from bread wheat in the greenhouse of Sichuan Agricultural University. Wheat accessions PI352335, PI355465, PI415152, and CItr7685 carrying HapI and accessions PI190949, PI113392, PI124494, and AS285 carrying HapII of Pm60 were randomly selected for the powdery mildew resistance test. Wild emmer wheat accession G18-16 (TdPm60, highly resistant) [8] and Chinese Spring (highly susceptible) were used as controls. Wheat plants at the two-leaf stage were tested for resistance to powdery mildew as previously described by Li et al. [8]. In brief, the first leaf of a seedling plant was excised into three segments with approximately 2 cm in length and cultured in petri dishes harboring 8 g/L agar with 50 mg/L benzimidazole (Sigma-Aldrich, St. Louis, MO, USA). Dishes were inoculated with fresh spores of powdery mildew isolates and were transferred to a growth chamber with 75% humidity and kept in 20 °C/18 °C under a 16 h light/8 h dark photoperiod, respectively. Resistance results were recorded at 7 and 10 days post inoculation (dpi). Disease symptoms were recorded based on a 0–4 infection type (IT) scale as described by Wu et al. [14]. The IT scores were classified into two groups, of which 0–2 were considered as resistant (0 = immunity, 0 = necrotic flecks, 1 = highly resistant, and 2 = resistant) and 3–4 as susceptible reactions (3 = susceptible and 4 = highly susceptible).

3. Results

3.1. Distribution of Pm60 in Wheat Species

The Pm60 functional marker M-Pm60-S1 could amplify 591 bp for Pm60a, 831 bp for Pm60, or 1071 bp for Pm60b, whereas there was no amplification product for the absence of Pm60 alleles [11]. As a result, 20 out of 238 accessions were positive for the M-Pm60-S1 marker, whereas 218 showed no amplification (Table S1). The results showed that tetraploid wheat species, including T. carthlicum, T. dicoccon, T. armeniacum, T. dicoccoides, T. durum, and T. turanicum, were harboring the Pm60 gene. Based on the size of the PCR products, we concluded that these tetraploid wheats may contain Pm60 and Pm60a alleles (Figure 1B and Table S1).

3.2. Sequence Variation of Pm60

The genomic sequence of the entire Pm60 gene was amplified, using four overlapping primer pairs (Table S2), in the 20 tetraploid wheat accessions that may harbor Pm60 alleles. The obtained sequences were compared with the published Pm60 sequences and revealed three novel (HapI–III) haplotypes (Table S3, GenBank accession numbers OP893800-OP893802) and two known haplotypes MW375699 and MW375701 (Table S1), which were previously reported by Wu et al. [14]. HapI was detected in T. dicoccon (n = 6) and T. turanicum (n = 1) accessions; HapII was identified in T. carthlicum (n = 2), T. durum (n = 4), and T. turanicum (n = 4) accessions; HapIII was only found in T. armeniacum accession PI361859 (Table S1).

The size of the HapI sequence was 4,362 bp in length, which showed 99.75% similarity to TuPm60 with eight SNPs and a GCC deletion. The sizes of the HapII and HapIII sequences were 4,215 and 4,122 bp, respectively, which both had a 240 nt deletion that was the same as the previously reported TuPm60a. The HapII and HapIII sequences were 99.71% and 99.56% identical to TuPm60a, respectively. Twelve and fifteen SNPs were identified in HapII and HapIII, respectively, as compared with TuPm60a. Only three SNPs and a GCC InDel were detected between HapII and HapIII (Table S3 and Figure 2).

3.3. Phylogenetic Analysis of Pm60

To determine the phylogenetic relationship between Pm60 proteins and homologs in the annotated genome assemblies of domesticated wheat (T. aestivum and T. durum) and their wild relatives (A. speltoides, A. longissima, A. searsii, T. urartu, and T. dicoccoides) (Table S4), an unrooted tree was constructed using the neighbor-joining method (Figure 3). The results of the phylogenetic analysis indicated that all published Pm60 proteins from T. urartu accessions and wild emmer accessions as well as three Pm60 haplotypes in this study were classified into one clade. HapII was more closely related to the functional haplotype MlIW172. Four homologs (TraesCS7B01G479200.1, Ae.searsii.TE01.7S01G0924600.1, Ae.longissima.TL05.4S01G0749100.1, and Ae.sharonensis.TH02.7S01G0968600.1) were clustered very close to this reported Pm60. TRIUR3_00771 and TRIUR3_00770 from T. urartu acc. G1812, Ae.speltoides.TS01.7B01G0976700.1 from Ae. speltoides, TRIDC7AG077150.1 from T. dicoccoides acc. Zavitan, and TraesCS7D01G542200.1 from T. aestivum cv. Chinese Spring (CS) were classified on a cluster with longer branches than the Pm60 proteins. TRITD7Av1G277190.1 and TRITD7Bv1G230590.1 from T. durum acc. Svevo, and BRADI_1g29658v3 and BRADI_2g38192v3 from Brachypodium distachyon, were located as an outgroup in the phylogenetic tree.

3.4. Powdery Mildew Responses

To test if any of the newly identified Pm60 haplotypes confers resistance to powdery mildew, a leaf infection test at the seedling stage was initially performed in eight randomly selected wheat accessions carrying either HapI or HapII of Pm60 using different Bgt isolates. HapIII was not involved in the inoculation test due to the unavailable seeds of T. armeniacum accession PI 361859. All accessions were susceptible (IT = 3–4) to Bgt isolates Bgt #1-25, Bgt#15-9-4, Bgt#WJ, and Bgt#SZS, whereas PI190949, PI113392, PI124494, and AS285 carrying HapII of Pm60 showed partial resistance to Bgt#GH (IT = 2;

Table 1. Reactions of wheat accessions with different Pm60 haplotypes to five Bgt isolates at 10 days post inoculation.

□	□	HapI				HapII				□
Bgt Isolates	G18-16 (TdPm60)	PI352335	PI355465	PI415152	CItr7685	PI190949	PI113392	PI124494	AS285	Chinese Spring
Bgt#1-25	ND	4	4	4	3	4	4	4	3	4
Bgt#15-9-4	ND	4	4	4	4	3	4	4	3	4
Bgt#SZS	0	3	4	4	4	4	3	4	4	4
Bgt#WJ	0	3	4	4	4	3	4	4	4	4
Bgt#GH	0	4	4	3	4	2	2	2	2	4

Susceptible wheat cultivar Chinese Spring was used as the control. ND, not determined.

and Figure S1).

4. Discussion

The powdery mildew resistance gene Pm60 isolated from the diploid T. urartu encodes a typical CC-NB-LRR protein that has at least three functional alleles, including TuPm60, TuPm60a, and TuPm60b [7]. Pm60 were further identified in T. dicoccoides, which are the orthologs of Pm60 from T. urartu [8,14]. In the present study, we have studied the diversity of Pm60 genes in a large set of wheat accessions and found that Pm60 alleles were not only present in wheat progenitor species T. urartu and T. dicoccoides but also distributed in cultivated tetraploid wheats.

Pm60 is an ancient gene present in T. urartu and T. dicoccoides species, where it shows a high level of presence/absence polymorphism [8,11,14]. TuPm60 was the most prevalent one, and TuPm60a, TuPm60a’, and TuPm60b were the less prevalent alleles in T. urartu accessions [11]. The T. urartu accessions from Turkey were shown to have all forms of Pm60 alleles. The wild emmer wheat population had functional Pm60 alleles, which are only present in the natural populations of T. dicoccoides from the Southern Levant, and almost all of them belong to the TdPm60 form [8,14]. The TdPm60a and TdPm60b alleles are rare and were identified only once [8,14]. In the current study, we found Pm60 homologs present in 20 tetraploid wheats but not in the hexaploid wheat accessions tested. The presence of Pm60 homologs in cultivated tetraploid wheats implicates that Pm60 may be involved in wheat domestication events. The results of the sequence analysis revealed only three new Pm60 haplotypes (HapI–HapIII). HapI was similar to Pm60. HapII and HapIII had a 240 nt deletion, which was the same as the previously reported Pm60a. Among them, HapI was present only in domesticated tetraploid wheats T. dicoccom and T. turanicum; HapII was found in tetraploid wheats T. durum, T. turanicum, and T. carthlicum; whereas HapIII was only identified in wild tetraploid T. araraticum accession PI361859.

Phylogenetic analyses indicated that all proteins encoded by the Pm60 gene were clustered into one clade, suggesting the evolutionary conservation of the Pm60 locus in Triticum species. T. dicoccun (emmer wheat, A^uA^uBB) is the domesticated form of T. dicoccoides and is believed to have been domesticated probably in southeast Turkey. Cultivated emmer was later evolved into the free-threshing ears of durum wheat [18,19]. Therefore, we hypothesized that the HapI in emmer originated possibly from the northern populations of T. dicoccoides [20] with an uncharacterized TdPm60 form. HapII in cultivated T. durum was likely derived from emmer wheat. HapIII in T. araraticum (A^uA^uGG) might be independently inherited from T. urartu by the polyploidization event. Further studies are still needed to test these hypotheses.

A previous study showed that the nonfunctional allele Pm60a’ with 58 SNPs and a 3-nucleotide deletion was detected in several susceptible T. urartu accessions, implying that these SNPs and deletion could be important for the resistance function [11]. In the present study, eight to fifteen SNPs were identified in between new Pm60 haplotypes (HapI–HapIII) and functional TuPm60a. Besides, HapI and HapIII sequences had 3 nt deletion in the interdomain region between the CC and NB-ARC domains of Pm60 as compared with HapII and TuPm60. HapII and HapIII sequences had a 240 nt deletion, which was the same as the previously reported TuPm60a [7]. We found that tetraploid wheat accessions carrying HapI or HapII sequences were susceptible to four tested Bgt isolates, whereas HapII (Pm60a allele) carrying lines showed partial resistance (IT=2) to Bgt#GH, suggesting that HapII may be functional. HapII differed from HapI by 10 SNPs and a 3 bp InDel as well as 240 nt deletion, which resulted in the loss of the two leucine-rich repeat motifs (Table S3). A recent finding demonstrated that deletion of the two leucine-rich repeat motifs in TuPm60a substantially narrowed the resistance spectrum [10]. It seems that these SNPs and a 3 bp InDel between HapII and HapIII are vital for the resistance function. However, we cannot rule out the possibility of the presence of other Pm genes in the HapII-carrying lines.

Pyramiding of several resistance genes or resistance alleles has the potential to improve resistance in terms of durability and spectrum [21,22]. TdPm60, Pm60, and Pm60b have been successfully introduced from wheat donor species into common wheat using the durum wheat as a ‘bridge’ [3,8]. The developed introgression lines carrying different Pm60 genes conferred high resistance to Bgt in hexaploid wheat genetic backgrounds, which facilitated the utilization of Pm60 in wheat resistance breeding. A previous study observed a different resistance spectrum of Pm60 alleles in T. dicoccoides (MlIW172 and MlWE18) and in T. urartu (Pm60, Pm60a, and Pm60b) when inoculated with 20 Bgt isolates [14]. It seems that single Pm60 cannot provide durable protection against wheat powdery mildew. Therefore, allele pyramiding of Pm60 or pyramiding of Pm60 with other disease resistance genes/alleles is valuable for breeding disease resistance wheat cultivars.