Are PECTIN ESTERASE INHIBITOR Genes Involved in Mediating Resistance to Rhynchosporium commune in Barley?

A family of putative PECTIN ESTERASE INHIBITOR (PEI) genes, which were detected in the genomic region co-segregating with the resistance gene Rrs2 against scald caused by Rhynchosporium commune in barley, were characterized and tested for their possible involvement in mediating resistance to the pathogen by complementation and overexpression analysis. The sequences of the respective genes were derived from two BAC contigs originating from the susceptible cultivar ‘Morex’. For the genes HvPEI2, HvPEI3, HvPEI4 and HvPEI6, specific haplotypes for 18 resistant and 23 susceptible cultivars were detected after PCR-amplification and haplotype-specific CAPS-markers were developed. None of the tested candidate genes HvPEI2, HvPEI3 and HvPEI4 alone conferred a high resistance level in transgenic over-expression plants, though an improvement of the resistance level was observed especially with OE-lines for gene HvPEI4. These results do not confirm but also do not exclude an involvement of the PEI gene family in the response to the pathogen. A candidate for the resistance gene Rrs2 could not be identified yet. It is possible that Rrs2 is a PEI gene or another type of gene which has not been detected in the susceptible cultivar ‘Morex’ or the full resistance reaction requires the presence of several PEI genes.


Introduction
Barley (Hordeum vulgare L.), one of the most important crop species worldwide is threatened by the hemibiotrophic, haploid ascomycete Rhynchosporium commune (formerly named Rhynchosporium secalis (Oudem.)J.J. Davis) [1][2][3], which causes a necrotic foliar disease (scald or leaf blotch) on infected plants [4].Recent studies revealed that the genus Rhynchosporium which consists of two species R. secalis and R. orthosporum can be further divided based on phylogenetic criteria [5].R. secalis separates into three monophyletic groups that are also characterized by their host spectra: R. secalis infecting rye and triticale, R. commune infecting cultivated barley and other Hordeum spp. as well as Bromus diandrus (brome grass) and R. agropyri, infecting Agropyron spp.[5].
Regarding the resistance of barley against R. commune, research has led to the identification of altogether 16 major resistance genes or alleles (R genes), however only nine of them have been mapped until now [6,7], which include four major resistance loci such as the Rrs1 complex on chromosome 3H with at least 11 known alleles [8,9], the Rrs2 locus on 7HS [10,11], Rrs13 on chromosome 6H [12] and the Rrs15 locus on 2H [13].
Our knowledge is still limited regarding the identity of the mapped resistance genes and their possible functions and resistance mechanisms.For the Rrs1 mediated resistance, prevention of subcuticular hyphae growth has been observed to be a major factor in the defence reaction against R. commune [1].Jørgensen et al. [14] found that the Rrs2 mediated resistance reaction in the cultivars 'Digger' and 'Osiris', which both carry the Rrs2 gene, is characterized by the prevention of fungal penetration of the cuticle through an accelerated and larger formation of papillae and halos in the cell walls compared to susceptible cultivars.The authors also suspect the involvement of chemical compounds in the inhibition process.
Pectins account for ca.30% of the cell walls of dicot and non-graminaceous monocot plants and for between 5-10% of the walls of grasses.They are a family of complex polysaccharides whose backbones are highly enriched in α-1,4-linked-D-galacturonic acid residues, with various degrees of methylesterification [15].Pectin is secreted in highly methyl-esterified form and subsequently deesterified by pectin methylesterases (PMEs; EC 3. 1.1.11).PME activity is controlled by pectin methylesterase inhibitors (PMEI or PEI), which belong to the family of invertase inhibitors [15].PMEs as well as PMEIs appear to occur as large gene families.In Arabidopsis thaliana, 66 ORFs have been annotated as putative full-length PMEs and 35 ORFs in Oryza sativa [16].The structural details of PME-PMEI interaction have been studied by crystallographic methods in several plant species and the inhibition apparently occurs through the formation of an inactive 1:1 complex between the two proteins [17][18][19][20].The role of PMEs in plants have been implicated in several processes, such as cell wall extension, fruit maturation, senescence, pollen tube growth and seed germination [15,21,22].Furthermore, PMEs are involved in the systemic spread of tobacco mosaic virus [23] as well as tobamovirus [24] and seem to stimulate virus-induced gene silencing [25].An involvement of cell walls and PME action in resistance to pathogens has been discussed [15,16,[26][27][28].
Our previous results of fine mapping, in which we used a segregating population 'Atlas' (resistant) × 'Steffi' (susceptible) of 9,179 F 2 plants, assigned the scald resistance gene Rrs2 to an area of suppressed recombination on chromosome 7HS [29].Based on these fine mapping results we decided to carry out a sequence survey of this suppressed recombination region in order to pinpoint interesting genes depicting haplotype variation between resistant and susceptible cultivars.Analysis of several genes, identified within the region of interest, led to the discovery of a family of putative PECTIN ESTERASE INHIBITORs (PEI) whose haplotypes differed between resistant and susceptible cultivars.This feature and the fact that these genes co-segregated with the resistance gene Rrs2 made them possible candidates for the Rrs2 gene.The goal of the present research work was to elucidate the role of the putative PECTIN ESTER-ASE INHIBITOR (PEI) genes, in the process of mediating resistance to R. commune in barley.

The PEI-gene family in barley
The Rrs2 gene which confers resistance to R. commune was fine mapped by Hanemann et al. [29].It is located in a co-segregating genomic region on barley chromosome 7HS.In previous research we had established two BAC contigs for the co-segregating region, which consisted of BACs from the susceptible cultivar 'Morex' [29].In this genomic region a number of genes had been identified including genes coding for putative pectinesterase inhibitor domain containing proteins [29].Further analysis of BAC sequences that became available during the past two years [30] led to the discovery of additional PEI family members in the neighborhood of already known PEI candidates (Fig 1).Gene ontology description predicted enzyme inhibitor activity (GO:0004857) for pectin esterase activity (GO:0030599) for genes HvPEI1, HvPEI2, HvPEI3, HvPEI4 and HvPEI6, therefore we refer to this gene family as putative pectin esterase inhibitor genes.For HvPEI6 additionally an ABC transporter related InterPro description was identified with molecular functions of ATP binding (GO:0005524) and ATPase activity (GO:0016887), while for HvPEI5 no gene ontology function was identified (S1 File).
Cluster analysis revealed a defined structure of the family of PECTIN ESTERASE INHIBI-TOR genes found in the Rrs2 region.By combining database information from different sources (e.g.http://www.harvest-web.org),we were able to identify corresponding expressed unigenes or cDNA clones for a total of six PEI genes (Fig 2).As exceptional case, HvPEI5 showed no differences between resistant and susceptible cultivars and hence was not included in further analyses.Although the ViroBLAST (IPK Barley BLAST Server, http://webblast.ipk-gatersleben.de/barley/)search assigned the gene to chromosome 7H we were not able to locate it within the investigated chromosomal region (Fig 1).The gene HvPEI1 had a presumed susceptible variety-specific haplotype H2 in the Rrs2 carrying (resistant) variety 'Atlas46' and a presumed resistant variety-specific haplotype H1 in several susceptible varieties and was therefore also excluded from further analysis.

Analysis of transgenic plants
For over-expression of identified PEI candidate genes, the barley cv.'Golden Promise' (GP) was used because of its amenability to genetic transformation and its susceptibility to Rhynchosporium infection.The full-length coding sequences of HvPEI2, HvPEI3, HvPEI4 of resistant cv.'Atlas' were first cloned for each candidate separately into the binary vector p6U and afterwards stably transformed by using Agrobacterium-based transformation technology.For HvPEI2, HvPEI3 and HvPEI4, 25, 41 and 17 independent primary transgenic plants were generated, respectively.For all investigated HvPEI2, HvPEI3 and HvPEI4 over-expression (OE)plants, a significant increase in expression of the respective genes was confirmed by quantitative RT_qPCR in transgenic progeny in comparison to non-transgenic control plants (cv.'GoldenPromise').In several plants with high copy numbers, such as PEI2_OE BG361-E15 lines, the over-expression levels were reduced, which may be an effect of gene silencing.(S2 Fig) .The gene HVPEI6 was only discovered later during the analysis of the respective genomic region; therefore for this gene copy only sequence and haplotype analysis was conducted, but it had not been included in the series of production of transgenic lines which had been finished for the other genes at the time point of discovery of HvPEI6.
The copy number in the progeny was assessed by applying TaqMan™ PCR using a probe specific for the HPT gene.In a T0-plant each transformed gene is usually only present as a single copy per integration site, this means it is present in heterozygous state.Therefore the T1-generation of transformed plants is segregating with respect to copy number, because the transformed sites are not genetically fixed yet.The presence of the transformed PEI genes from 'Atlas' was furthermore verified by applying gene and allele-specific CAPS-marker analysis.
Several T1 OE-plants for HvPEI2, HvPEI3 and HvPEI4 were tested for resistance to R. commune isolate 'LfL 12F' in an infection test in the greenhouse at the 'Bayerische Landesanstalt' in Freising, Germany.Overall, none of the transformed lines reached the resistance level of the resistant cultivar 'Atlas' of 0.78 (Table 2).In a disease scoring range of 0 (most resistant) to 4 (most susceptible) the transformed OE-plants ranged from 2.79 to 3.86 on average.Usually a scoring from 0 to 2 is considered as resistant, while a scoring of 3 and 4 is considered as susceptible according to Jackson and Webster [31].Comparing the transformed OE-lines to nontransformed 'Golden Promise', with an average resistance level of 3.84, several of the transformed OE-lines were significantly different from 'Golden Promise' (Table 2).Thereby, four out of five OE-lines for HvPEI4 reached a significance level of P<0.001.However, none of the transformed plants reached a scoring level to be considered as resistant.From these results, it can be concluded that none of the HvPEI2, HvPEI3 and HvPEI4 genes alone seems to represent the Rrs2 gene of 'Atlas', since the resistance level of the OE-lines is rather still in the susceptible range.Some improvements of resistance could be observed, especially with construct HvPEI4, however this may be attributed to somaclonal variation.Since the investigated PEI genes originate from the susceptible cultivar 'Morex', which does not carry Rrs2, it is possible that in the resistant cultivar 'Atlas' additional genes are present, which were not detected and therefore not tested here.These either could be additional PEI genes or other genes which are not present in 'Morex'.

Involvement of pectin esterase inhibitors in plant defense reactions
Several reports indicate a potential role of PEI or PMEI genes in plant defense reactions against pathogens: The PECTIN METHYL ESTERASE INHIBITOR gene CaPMEI1 from pepper (Capsicum annuum L.) was isolated and functionally characterized [32,33].Virus-induced gene silencing of CaPMEI1 in pepper conferred enhanced susceptibility to Xanthomonas campestris pv.vesicatoria.Ectopic expression of CaPMEI1 in Arabidopsis led to enhanced resistance to Pseudomonas syringae pv.tomato, but not to the biotrophic pathogen Hyaloperonospora parasitica [32].Furthermore, the promoter of CaPMEI1 induced a high level of GUS reporter activity in tobacco leaf tissue, driven by pathogen infection as well as by ethylene and methyl jasmonate treatment, suggesting an involvement in host defense response [33].
Table 2. Results of resistance tests with the Rhynchosporium isolate LfL12F comparing control plants of non-transformed cv.'Golden Promise' and T 1-plants of HvPEI2, HvPEI3 and HvPEI4 overexpression (OE) lines.Since the individual plants of the T 1 -events are still segregating the estimation of copy numbers of the respective gene copies from 'Atlas' are indicated: mc = multi-copy, 1c = one copy.The disease scoring ranged from 0 (most resistant) to 4 (most susceptible).

Standard deviation
No. of plants The over-expression of PMEIs in Arabidopsis resulted in an increase of 16% of pectin methylesterification in cell walls of the transformed plants [34].Transformed plants were more resistant to the necrotrophic fungus Botrytis cinerea.The reduced symptoms caused by the fungus on transgenic plants were related to its impaired ability to grow on methylesterified pectins [34].
Differences in the methyl ester distribution of homogalacturons were reported for near-isogenic wheat lines resistant and susceptible to the wheat stem rust fungus Puccinia graminis f. sp.Tritici, which may result in differences for the enzymatic generation of endogenous suppressors resulting in different outcomes of the host-pathogen interaction [35].
The fruit-specific Fragaria × ananassa PECTIN METHYL ESTERASE FaPME1 was expressed in wild strawberry Fragaria vesca, resulting in a reduced degree of esterification of the oligogalacturonides from the transgenic fruits as compared to the wild type fruits [36].The transgenic F. vesca lines had constitutively activated pathogen defense response, resulting in higher resistance to the necrotrophic fungus Botrytis cinerea.It was speculated if oligogalacturonides released from the cell wall may play a role in eliciting defense responses [36].
Transgenic durum wheat lines expressing the PMEI from Actinidia chinensis (AcPMEI) showed a significant increase in the degree of methyl esterification and a significant reduction of disease symptoms caused by the fungal pathogens Bipolaris sorokiniana and Fusarium graminearum [37].The expression of the TaPMEI gene in wheat was monitored by RT-PCR and high transcription was detected in salicylic acid and hydrogen peroxidase treatments [38].An enzyme assay indicated that PME was completely inhibited by TaPMEI.
From the previous examples, two possibilities for the involvement of PMEIs in plant defense mechanism become obvious.The first is an increase of PMEI resulting in repressed PME activity, reduced de-methylesterification and therefore in highly esterified cell walls, which are less prone to the attack by pectin-degrading enzymes of potential pathogens [32,34,37].Another possibility is the involvement of elicitor molecules which may be released by partial degradation of cell walls [36] and subsequent elicitation or enhancement of the plant defense response.In the given examples, the plant PMEIs seem to interact with the own host PMEs.To our knowledge, so far no example exists where a plant PMEI interacts with a fungal or microbial PME.
The concept of 'effector-triggered defence' (ETD) has been suggested for apoplastic fungal pathogens, such as R. commune [39].ETD responses are characterized by a relatively slow response to the pathogen, not associated to a fast hypersensitive host cell death response, whereby effectors of apoplastic pathogens are recognized at the cell surface [38].

Conclusions from own experiments
In previous research, the Rrs2 gene was placed in a linkage block of suppressed recombination which was found in all investigated varieties carrying Rrs2 and indicated a common origin for this resistance gene [29].This linkage block had previously been defined by SNPs of non-specified genomic sequences [29].The present results of haplotype analysis confirmed that genes HvPEI2, HvPEI3, HvPEI4 and HvPEI6 are present within the linkage block, which made these genes possible candidates for Rrs2 (Table 1) and the testing of this hypothesis in over-expression plants was initiated.As result, none of the tested candidate genes HvPEI2, HvPEI3 and HvPEI4 alone conferred a high resistance level in transgenic over-expression plants, though an improvement of the resistance level was observed especially with OE-lines for gene HvPEI4 (Table 2).Therefore, we consider the following hypotheses about the identity of the Rrs2 gene: 1. Rrs2 is identical to HvPEI6 which has not been tested yet in a transgenic approach.
2. The resistance of Rrs2 requires the action of two or more PEI genes.
3. Since all available sequence information comes from the susceptible cultivar 'Morex', it is possible that in resistant cultivars, such as 'Atlas' additional PEI genes or other genes are present in the Rrs2 region which cannot be found in 'Morex'.
In conclusion, the question posed in the title cannot conclusively be answered by the current results.We can conclude, that neither HvPEI2, HvPEI3 or HvPEI4 alone represents the resistance gene Rrs2.However the presence of additional PEI genes in the resistant cultivar 'Atlas' cannot be excluded.It is also possible that several linked genes are required for resistance expression.
Several fungal isolates were employed for practical reasons.These included Sachs 147-1 and LfL12F (both obtained from Günther Schweizer, LfL Freising, Germany) and UK7 (obtained from Wolfgang Knogge, IPB Halle, Germany).In all experiments, the two barley varieties 'Atlas' (carrying Rrs2) and 'Steffi' (susceptible), which were the parents of the original mapping population [29], were included as controls.

Haplotype analysis and marker development
New putative PEI genes were identified using two different strategies: First, a keyword search was performed within the HarvEST barley database (http://www.harvest-web.org),version 1.83 (Assembly 35) and sequences of resulting candidates were verified for chromosome position with the ViroBLAST tool (http://webblast.ipk-gatersleben.de/barley/viroblast.php), applying default BLAST parameter settings [40].The second approach was based on a software application, an automated annotation system for rice: RiceGAAS ( [41]; http://ricegaas.dna.affrc.go.jp/), which possesses programs for gene prediction as well as BLAST search options for identification of homologous sequences and gene ontology predictions.Available sequence information of different BAC contigs was analyzed regarding gene content using the graphical user interface.
All PEI candidate genes were amplified by PCR from genomic DNA of cultivar 'Atlas' (Rrs2 +, resistant) using specific forward and reverse primers (S2 Table ).Primer development was done with the program PrimerSelect™ which is part of the software suit DNASTAR1 Lasergene version 10 (DNASTAR, Inc., Madison, Wisconsin, USA).
Haplotype analysis and alignment of nucleotide sequences was performed with the program Sequencher™ version 5.2.3 (Gene Codes Corporation, Ann Arbor, MI, USA).Alignments were analyzed for presence of restriction sites at SNP positions with the Sol Genomics Network tool (http://solgenomics.net/tools/caps_designer/caps_input.pl)[42].
Restriction digestions of PCR products were carried out in a 10-μl volume using 5 μl of PCR product, 1U of the respective restriction endonuclease and an incubation time of 2h at recommended optimal temperature.The restricted fragments were separated on 1-2% 1 x TAE agarose gels.

Preparation of over-expression constructs
Sequences of candidate PEI genes were amplified by PCR from genomic DNA of cultivar 'Atlas' (Rrs2+, resistant) using specific forward and reverse primers (S2 Table ) and cloned into TOPO vector pCR2.1 (Invitrogen, Karlsruhe, Germany) according to manufacturer's instruction manual for sequence verification.Afterwards correct sequences were transferred into the entry vector pUBI-ABM (DNA-Cloning-Service, Hamburg, Germany) via a digestion and ligation-based cloning strategy that ensured a proper in-frame orientation of PEI candidates.Clones that were tested positive for presence of the gene of interest were used for further cloning.A fragment that includes the maize UBIQUITIN1 promoter, the gene of interest as well as a transcription termination sequence (NOPALINE SYNTHASE terminator) was transferred into destination vector p6U (DNA-Cloning-Service, Hamburg, Germany) using the rare cutter SfiI, which has a 13 nucleotide recognition site including five variable nucleotides in central position [43].Barley 'Golden Promise' immature embryos were stably transformed with PEI over-expression vectors using Agrobacterium tumefaciens according to [44,45].

Transcript abundance and fungal growth (RT-qPCR)
The assessment of candidate gene expression as well as fungal growth, on stably transgenic barley plants (OE), was performed by quantification via reverse transcription real-time PCR as described by Himmelbach et al. [46].For this purpose, total RNA and DNA were extracted from control or treated primary leaves as well as first and second leaves at different time points after inoculation with R. commune spores.RNA isolation was performed using TRIzol Reagent (Life Technologies GmbH, Darmstadt, Germany), which is a monophasic solution of phenol, guanidine isothiocyanate and other components and which allows to perform sequential precipitation of RNA, DNA and proteins from a single sample [47].After homogenization of the samples with TRIzol Reagent, chloroform was added and the homogenate was allowed to separate into a clear upper aqueous layer containing RNA, an interphase, and a red lower organic layer containing DNA and proteins.RNA was precipitated from the aqueous layer using isopropanol.The clear stratification in diverse phases prevented the contamination with genomic DNA.In case of analysis of target gene expression, cDNA was synthesized from 1 μg total RNA by using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA) in a 20 μl reaction.
An aliquot of 1 μl cDNA or DNA was taken as template for real-time PCR on a 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) by using the GoTaq1 qPCR Master Mix Kit (Promega) according to the manufacturer's manual, except that the total reaction volume was reduced to 10 μl.The primer pairs specific for the respective genes of interest (qRT2-PEI2_F/R, qRT-PEI3_F7/R7, qRT-PEI4_Fa/R) were designed by the software program Lasergene (DNASTAR) (S3 Table ) and deployed in a final concentration of 1 μM.For amplification of DNA, a standard two-step, 40-cycle program, followed by a dissociation step was used.After initial incubation at 50°C (2 min) and extended activation of the polymerase at 95°C (10 min), 40 cycles with 95°C (15 sec) and 62°C (30 sec) were applied.
Transcript as well as DNA abundance was determined relative to standard curves and normalized to a gene encoding the barley UBIQUITIN-CONJUGATING ENZYME (UBC).Data analysis was performed with the sequence detection software (Version 2.2.2;Applied Biosystems) that is provided with the Fast Real-time cycler.Statistical validation was ensured by three biological replicates.
The verification of transgenic barley plants was carried out both via TaqMan1 PCR for copy number estimation and RT-qPCR for candidate gene expression.The latter method was already described above whereas the multiplex real-time PCR (TaqMan1) was performed according to Yang et al. [48] by adapting the SYBR Green assay to TaqMan1 chemistry: The amplification reaction (15 μl) contained 1 μl genomic DNA, 7.5 μl Maxima Probe/ROX qPCR Master Mix (Thermo Scientific), and forward and reverse primers that are specific for the HYGROMYCIN PHOSPHOTRANSFERASE selectable marker gene (HPT) or the barley UBC gene, respectively.Probes were dual-labeled with a fluorescent reporter dye attached to the 5'end of the oligonucleotide and the highly efficient dark quencher BHQ1 (Biosearch Technologies, Inc., Petaluma, California, USA) attached to the 3'-end (S3 Table ).In case of HPT the 6-carboxy-fluoroscein (FAM™) was used, whereas for UBC 6-carboxy-4',5'-dichloro-2',7'dimethoxyfluoresceine (JOE™) was applied.Final primer concentration was adjusted to 0.5 μM and probe concentration to 0.2 μM.
Amplification conditions were derived from a two-step, 40-cycle program described above that is complemented with a third step at 72°C (30 sec), while omitting the dissociation step.Each sample was quantified in triplicate.
To calculate transgene copy numbers, a quantitative method was used which allowed comparing the absolute quantification reaction for the candidate gene HPT with the endogenous reference gene UBC.Transgene copy numbers were estimated using a homozygous HPT-transgenic control plant, verified as described by [44], as calibrator sample and UBC as endogenous control.Each reaction comprised three technical replicates.For statistical analysis the software package SigmaPlot11.0was used.

Fig 1 .
Fig 1. BAC contig of the Rrs2 region carrying various PEI-genes.The horizontal bar represents the genetic map based on 9,179 F 2 -plants with the numbers of recombinant plants between six genetic markers indicated according to [29].The hatched areas represent the location of recombination events and could also harbor the Rrs2-gene.The blue boxes represent BAC-clones and the green arrows indicate the approximate locations of the PEI genes.doi:10.1371/journal.pone.0150485.g001

Fig 2 .
Fig 2. Cluster analysis of the PEI gene family in barley.Phylogenetic analysis of six detected PEI family members was performed by using database information of known expressed unigenes (HarvEST Assembly 35), BAC sequence information as well as sequence information from 41 analyzed barley cultivars.In addition data from the Gene Index Project (Barley database 12.0) were also included.doi:10.1371/journal.pone.0150485.g002

Fig 3 .
Fig 3. Haplotype analysis of detected PEI genes on chromosome 7HS.Full-length sequences of all detected PECTIN ESTERASE INHIBITOR members were analyzed in different barley varieties either carrying the Rrs2 resistance gene or not.Information regarding susceptibility to Rhynchosporium commune is according to [29].doi:10.1371/journal.pone.0150485.g003

Fig 4 .
Fig 4. CAPS-markers for different PEI genes.As a result of haplotype analysis various SNPs were discovered within the ORF of HvPEI2, HvPEI3, HvPEI4 and HvPEI6.Polymorphisms were used for CAPSmarker development that in turn could be applied for haplotype discrimination.Examples on the gel photo illustrate results of PCR amplification as well as digestion of the amplicon, with the appropriate restriction enzyme (given in brackets), for different PEI family members.As an exception for HvPEI6 only the results of digestion are presented.Primers that were used for each PCR reaction are given in brackets as well.Additional information regarding distinguished haplotypes is shown as abbreviation on top of lanes with digested fragments (e.g.H1 = haplotype 1, H2 = haplotype 2, et cetera).M = 1kb+ DNA Ladder (Thermo Scientific™ GeneRuler, catalog number #SM1333).doi:10.1371/journal.pone.0150485.g004