mlo-based powdery mildew resistance in hexaploid bread wheat generated by a non-transgenic TILLING approach

Abstract

Wheat is one of the most widely grown cereal crops in the world and is an important food grain source for humans. However, wheat yields can be reduced by many abiotic and biotic stress factors, including powdery mildew disease caused by Blumeria graminis f.sp. tritici (Bgt). Generating resistant varieties is thus a major effort in plant breeding. Here, we took advantage of the non-transgenic Targeting Induced Lesions IN Genomes (TILLING) technology to select partial loss-of-function alleles of TaMlo, the orthologue of the barley Mlo (Mildew resistance locus o) gene. Natural and induced loss-of-function alleles (mlo) of barley Mlo are known to confer durable broad-spectrum powdery mildew resistance, typically at the expense of pleiotropic phenotypes such as premature leaf senescence. We identified 16 missense mutations in the three wheat TaMlo homoeologues, TaMlo-A1, TaMlo-B1 and TaMlo-D1 that each lead to single amino acid exchanges. Using transient gene expression assays in barley single cells, we functionally analysed the different missense mutants and identified the most promising candidates affecting powdery mildew susceptibility. By stacking of selected mutant alleles we generated four independent lines with non-conservative mutations in each of the three TaMlo homoeologues. Homozygous triple mutant lines and surprisingly also some of the homozygous double mutant lines showed enhanced, yet incomplete, Bgt resistance without the occurrence of discernible pleiotropic phenotypes. These lines thus represent an important step towards the production of commercial non-transgenic, powdery mildew-resistant bread wheat varieties.

Keywords: Blumeria graminis; Mlo; Targeting Induced Local Lesions in Genomes; hexaploid bread wheat; plant disease resistance; powdery mildew.

Figure 1

Exon 9 of TaMlo is the target for the TILLING screen. The scheme illustrates the predicted topology of the 7 transmembrane domain TaMlo proteins (top) and the experimentally determined common exon‐intron structure of the corresponding TaMlo homoeologues (bottom). The 11 exons are indicated by dark boxes, introns by orange lines. The third cytoplasmic loop of TaMlo, depicted in green and encoded by exon 9, was the target for TILLING. Arrows below the TaMlo gene model specify the amplicons of the 1st and 2nd round PCR. A scale bar, shown in red, is given below exon 11 (100 bp).

Figure 2

TILLING‐derived variants of Tamlo homoeologues exhibit different levels of functionality in a transient gene expression assay. Detached 8‐day‐old leaves of the powdery mildew‐resistant barley mlo‐3 mutant were co‐bombarded with a GUS reporter plasmid and a plasmid encoding the indicated TaMlo protein variant (WT or mutant version under transcriptional control of the maize Ubiquitin1 promoter). Expression of GUS alone and GUS plus WT barley Mlo, driven by the maize Ubiquitin1 promoter, were used as negative and positive controls, respectively, for restoration of Bgh susceptibility. Host cell entry was scored at 48 hours post inoculation (h p.i.) in GUS‐stained cells attacked by powdery mildew sporelings and results visualized as box plots. Centre lines show the medians; upper and lower box limits indicate the 25th and 75th percentiles respectively; upper and lower whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, respectively, and outliers are represented by dots. Numbers at the bottom of the boxplots indicate the number of biological replicates per sample (n). One biological replicate was typically composed of six leaves with 150 scored cells. Asterisks indicate a statistically significant difference to the respective WT (barley Mlo or TaMlo) with ***P < 0.001, **P < 0.01 and *P < 0.5 as determined by a Generalized Linear Model (GLM) test. Statistics were performed and boxplots generated with R software.

Figure 3

Bgh infection phenotypes of barley mutants mlo‐38 and mlo‐29 and their respective parental lines. Seven‐day‐old leaves were inoculated with Bgh isolate K1. (a) The macroscopic phenotype was recorded at 7 days post inoculation (d.p.i.). A scale bar shown in white is given in the lower right corner (1 cm). (b) Host cell entry was scored at 48 h p.i. Centre lines show the medians; upper and lower box limits indicate the 25th and 75th percentiles respectively; upper and lower whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, respectively. Shown are data of n = 3 independent biological replicates with 3–4 leaves (typically 200 scored interaction sites per leaf) per replicate. Asterisks indicate a statistically significant difference to the respective parental background with ***P < 0.001 as determined by a GLM test. Statistics were performed and boxplots established by R software.

Figure 4

Bgt infection phenotypes of wheat WT (cv. Cadenza), single, double and triple Tamlo mutants. Ten‐day‐old leaves were inoculated with Bgt isolate JA82. (a) Host cell entry was scored at 72 h p.i. Centre lines show the medians; upper and lower box limits indicate the 25th and 75th percentiles respectively; upper and lower whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles, respectively, and outliers are represented by dots. Numbers at the bottom of the boxplots indicate the number of biological replicates per sample (n). One biological replicate was typically composed of four leaves with 200 scored cells. Letters indicate genotypes whose data are significantly (P < 0.05) different from genotypes labelled with other letters, as determined by pair‐wise testing with a Games–Howell post hoc test. Statistics were performed and boxplots generated with R software. (b) First leaves of germinated seedlings were fixed with surgical tape to a polycarbonate platform and inoculated with Bgt conidiospores. The macroscopic phenotype was recorded at 6 d.p.i. A scale bar shown in white is given in the lower right corner (1 cm).

Figure 5

Transcript accumulation of TaMlo homoeologues in five bread wheat tissues at different developmental stages. Transcript levels of each homoeologue were summed up and then averaged across replicates, ± standard errors of the mean. FPKM: Fragments per kb per million mapped reads; Z: Zadoks developmental scale (Zadoks et al., 1974).