Novel induced mlo mutant alleles in combination with site-directed mutagenesis reveal functionally important domains in the heptahelical barley Mlo protein

Abstract

Background: Recessively inherited natural and induced mutations in the barley Mlo gene confer durable broad-spectrum resistance against the powdery mildew pathogen, Blumeria graminis f.sp. hordei. Mlo codes for a member of a plant-specific family of polytopic integral membrane proteins with unknown biochemical activity. Resistant barley mlo mutant alleles identify amino acid residues that are critical for Mlo function in the context of powdery mildew susceptibility.

Results: We molecularly analyzed a novel set of induced barley mlo mutants and used site-directed mutagenesis in combination with transient gene expression to unravel novel amino acid residues of functional significance. We integrate these results with previous findings to map functionally important regions of the heptahelical Mlo protein. Our data reveal the second and third cytoplasmic loop as being particularly sensitive to functional impediment by mutational perturbation, suggesting that these regions are critical for the susceptibility-conferring activity of the Mlo protein. In contrast, only mutations in the second but not the third cytoplasmic loop appear to trigger the Endoplasmic Reticulum-localized quality control machinery that ensures the biogenesis of properly folded membrane proteins.

Conclusion: Our findings identify functionally important regions of the polytopic barley Mlo protein and reveal the differential sensitivity of individual protein domains to cellular quality control.

Figure 1

Accession CGN0524 harbors a natural candidate mlo allele. A. Southern blot analysis of natural mlo candidate accessions. Genomic DNA of indicated barley accessions was digested with either Eco RV or Hin dIII, blotted onto a nylon membrane and probed with a radiolabelled full-length Mlo cDNA fragment. White arrowheads indicate fragments of the wild-type Mlo copy, black arrowheads point to the prominent mlo-11-characteristic signals (see also [11]). Approximate sizes of these fragments were calculated based on the DNA sequence of the Mlo genomic locus [42] and the arrangement of the mlo-11-specific repeats [11]. B. Single cell complementation of accession CGN0524 by transient gene expression. Leaves of CGN0524 were either bombarded with a plasmid encoding the GUS reporter protein or co-bombarded with the GUS reporter plasmid and a plasmid harboring the wild type Mlo cDNA. Fungal entry rates in transformed (GUS-stained) cells were scored at 48 hours post inoculation. Results show the mean ± standard deviation of n = 3 experiments for expression of GUS alone and n = 5 experiments for expression of GUS + Mlo. The asterisk indicates a statistically significant difference (p < 0.01) from the GUS control according to Student's t-test.

Figure 2

Yemenite accessions 5589 and 5590 harbor the natural mlo-11 allele. Genomic DNA of the indicated barley lines was used as a template for PCR amplification (40 cycles; extension 1 minute at 72°C) using either the oligonucleotide combination ADUP7A/Mlo6 (diagnostic for the presence of the mlo-11 repeat structure [11]) or Mlo6/Mlo10 (indicating presence/absence of a MITE associated with the 5'-terminal repeat in the majority of mlo-11 haplotypes [11,30]). PCR products were separated by agarose gel electrophoresis and visualized via ethidium bromide staining in combination with UV transluminescence. Cultivar Ingrid (Mlo genotype) and line back cross Ingrid (BCI) mlo-11 served as controls for wild type and mlo-11 genotypes, respectively. Note that lines RAH995 and Ab1089 were included as additional controls.

Figure 3

Biological activity of Mlo variants with single amino acid substitutions in the second and third cytoplasmic loop. A. Pairwise amino acid alignment of the third cytoplasmic loop of barley Mlo and AtMLO2. mlo-27 (G318E), mlo-29 (P334L) and mlo-33 (A306T) denote previously known barley mlo mutant sites. Highlighted in black are identical amino acid residues, in grey residues with similar biochemical properties. Lines indicate amino acids selected for site-directed mutagenesis. B. Relative protein accumulation of Mlo protein variants. Protein accumulation was determined using the dual luciferase assay using wild type Mlo as a positive control (set as 100%) and the Mlo-1 mutant variant (characterized by the W162R amino acid substitution) as a negative control. Values represent mean ± standard deviation based on 4 independent experiments with 2 technical replicates each. Asterisks indicate a statistically significant difference (p < 0.05) from Mlo wild type according to Student's t-test. C. Functional assay of Mlo protein variants. Leaf segments of the powdery mildew resistant barley line BCI mlo-3 were co-bombarded with the pUbi-GUS reporter plasmid and a plasmid encoding the indicated Mlo protein variant. Host cell entry was scored in GUS-stained cells attacked by powdery mildew sporelings. Expression of wild type Mlo served as a positive control, expression of GUS alone as a negative control. Values represent mean ± standard deviation based on at least 3 experiments with each ca. 100 investigated GUS-positive cells/construct. Asterisks indicate a statistically significant difference (p < 0.05) from the GUS control according to Student's t-test.

Figure 4

Pairwise amino acid alignment of the second cytoplasmic loop of barley Mlo and AtMLO2. Highlighted in black are identical amino acid residues, in grey residues with similar biochemical properties. Asterisks denote amino acids identified in powdery mildew resistant Arabidopsis Atmlo2 mutants [17].

Figure 5

A map of amino acid residues that are critical for Mlo function in the context of powdery mildew susceptibility. Schematic representation of the barley Mlo protein and the relative position of loss-of-function single amino acid substitution mutations and a two amino acid deletion mutant (mlo-10). The light grey horizontal rectangle depicts the lipid bilayer, the grey vertical rectangles the seven transmembrane domains and the black bends the extracellular and cytoplasmic loops. NH₂ and COOH denote the amino and carboxyl-terminus, respectively. A. Position of induced barley mlo mutations and relative protein accumulation of the respective protein variants. Colored dots designate barley mlo mutant alleles as defined in additional file 1. The color code indicates protein variants with wild type-like accumulation (green), intermediate accumulation (yellow), and severely affected accumulation (red), while white indicates lack of data about protein accumulation of this variant. B. Position of site-directed mlo mutants and relative protein accumulation of the respective protein variants. Colored dots designate the site-directed barley mlo mutants generated in this (Figure 3) and previous studies [21,22,25]. Only mutant variants with at least 50% reduction in fungal entry rate are shown. The color code indicates protein variants with wild type-like accumulation (green), intermediate accumulation (yellow), and severely affected accumulation (red).

Figure 6

Location of mlo mutant sites within the predicted consensus secondary structure of the Mlo second and third cytoplasmic loop. Consensus secondary structures of the second and third cytoplasmic loop regions of monocot and dicot Mlo orthologs were calculated with the JPRED web tool (see Methods for details). Amino acids are given in the single letter code ("OrigSeq") and the final predicted secondary structure("Jnet") is indicated by a red H (α helix), a yellow E (β sheet) or dashes (random coil) below the sequence. The color code (highlighted residues) indicates amino acids that lead to non-functional (<50% entry rate compared to wild-type Mlo) mutant variants with wild type-like accumulation (green), intermediate accumulation (yellow), or severely affected accumulation (red). "hmm" indicates the Jnet hidden Markov model prediction; "Jnet_25", "Jnet_5" and "Jnet_0" signify the predicted relative solvent accessibility using 25%, 5% and 0% cut-off values. The prediction indicates whether a residue is buried ("B") or exposed ("-") at each of the three relative solvent accessibility cut-offs. "Jnet Rel" specifies the prediction accuracy for each position, ranging from 0 to 9 (the higher the better).