• Open Access

The wheat Lr34 gene provides resistance against multiple fungal pathogens in barley

Authors


Correspondence (fax +61 2 6246 4950;

email Evans.Lagudah@csiro.au)

and

(fax +41 44 634 8204;

email bkeller@botinst.uzh.ch)

Summary

The Lr34 gene encodes an ABC transporter and has provided wheat with durable, broad-spectrum resistance against multiple fungal pathogens for over 100 years. Because barley does not have an Lr34 ortholog, we expressed Lr34 in barley to investigate its potential as a broad-spectrum resistance resource in another grass species. We found that introduction of the genomic Lr34 sequence confers resistance against barley leaf rust and barley powdery mildew, two pathogens specific for barley but not virulent on wheat. In addition, the barley lines showed enhanced resistance against wheat stem rust. Transformation with the Lr34 cDNA or the genomic susceptible Lr34 allele did not result in increased resistance. Unlike wheat, where Lr34-conferred resistance is associated with adult plants, the genomic Lr34 transgenic barley lines exhibited multipathogen resistance in seedlings. These transgenic barley lines also developed leaf tip necrosis (LTN) in young seedlings, which correlated with an up-regulation of senescence marker genes and several pathogenesis-related (PR) genes. In wheat, transcriptional expression of Lr34 is highest in adult plants and correlates with increased resistance and LTN affecting the last emerging leaf. The severe phenotype of transgenic Lr34 barley resulted in reduced plant growth and total grain weight. These results demonstrate that Lr34 provides enhanced multipathogen resistance early in barley plant development and implies the conservation of the substrate and mechanism of the LR34 transporter and its molecular action between wheat and barley. With controlled gene expression, the use of Lr34 may be valuable for many cereal breeding programmes, particularly given its proven durability.

Introduction

Breeding for durable disease resistance is fundamental to sustainable agriculture as the emergence of virulent pathogen races causes serious yield losses. One of the few durable disease resistance genes identified in crop plants is Lr34 from bread wheat. It was first described by Dyck et al. (1966) as a leaf rust (Puccinia triticina) resistance gene that is most effective in the flag leaf of adult plants during the critical grain-filling stage. Lr34-conferred resistance is also observed in seedlings when exposed to cold temperatures (Rubiales and Niks, 1995; Singh and Gupta, 1992), although this is likely to result from slower fungal growth or Lr34-dependent accumulation of a growth-inhibiting substance, rather than increased Lr34 expression in cold-treated seedlings (Risk et al., 2012). The race nonspecific leaf rust resistance provided by Lr34 is characterized as a durable, quantitative, slow-rusting response. In addition, the Lr34 gene provides partial resistance to stripe rust [Yr18 (Singh, 1992b)], stem rust [Sr57 (Dyck, 1987; Hiebert et al., 2009; Kolmer et al., 2011; McIntosh et al., 2011)] and powdery mildew [Pm38 (Lillemo et al., 2008; Spielmeyer et al., 2005)]. The Lr34 locus was further found to be genetically associated with tolerance to barley yellow dwarf virus (BYDV) and resistance to the spot blotch disease (Lillemo et al., 2013; Singh, 1993). Thus, Lr34 is one of the very few known genes that confer broad-spectrum, multipathogen resistance.

Lr34 encodes a putative adenosine triphosphate–binding cassette (ABC) transporter belonging to the ABC subtype G (formerly PDR), subclass of transporters (Krattinger et al., 2009). The developmentally regulated resistance was found to correlate with elevated Lr34 transcript levels in flag leaves of adult plants (Risk et al., 2012). Several allelic forms of the Lr34 gene have been identified in the wheat germplasm, with two alleles occurring at a high frequency (Cao et al., 2010; Dakouri et al., 2010; Lagudah et al., 2009; McCallum et al., 2012). The gain-of-function resistance allele (Lr34res) differs by only two amino acid changes from the other allelic forms associated with susceptibility (Lr34sus). Thus, the resistance conferred by Lr34res may be the result of altered activity or substrate specificity. Besides the Lr34 copy on chromosome 7D, two additional copies of Lr34 were identified in hexaploid wheat, and to date, there is no evidence that either of the alternate copies confers pathogen resistance.

The presence of Lr34 in wheat is associated with leaf tip necrosis (LTN) (Singh, 1992a), a type of necrosis that forms predominantly in the tips of wheat flag leaves, and is dependent on expression of Lr34res (Risk et al., 2012), as well as on genetic background and environmental conditions (Shah et al., 2010, 2011). The LTN phenotype correlates with senescence-like processes, as evidenced by expression of the senescence-related marker gene HvS40 (Krattinger et al., 2009) and the presence of senescence-specific chlorophyll catabolites (Krattinger et al., 2009; Risk et al., 2012). Senescence is an integral developmental process in cereals resulting in major metabolic changes. It involves transcriptional reprogramming and protein degradation and leads to nutrient mobilization from the leaves to the filling grain as reviewed by Gregersen et al. (2008). In Lr34res-containing wheat, the LTN phenotype seen in the flag leaves coincides with plant heading. The presence of Lr34res and LTN has only a small effect on yield in uninfected plants (Singh and Huerta Espino, 1997).

Previous studies have shown that the successful transfer of a resistance gene between related species is relatively rare, but possible in some special cases [reviewed by (Wulff et al., 2011)]. Comparative analysis of other grass species for the presence of Lr34 homologues and orthologs identified Lr34-like genes in rice and sorghum, but not in maize, barley or Brachypodium (Krattinger et al., 2011). The diploid nature of barley, the close phylogenetic relationship with wheat, as well as the absence of an Lr34 ortholog make barley a useful organism to study the molecular function of the wheat Lr34 resistance gene. Moreover, wheat stem rust (P. graminis f.sp. tritici) (Park et al., 2009) and barley-specific pathogens such as barley leaf rust (P. hordei) and powdery mildew (Blumeria graminis f.sp. hordei) are all virulent on barley cv. Golden Promise and can be used to test the functionality of wheat Lr34res in this heterologous system. In this study, we transformed barley cv. Golden Promise with genomic Lr34res and Lr34sus and two cDNA Lr34res constructs (cLr34res). In genomic Lr34res lines, we found increased resistance to multiple fungal pathogens, early LTN as well as early activation of senescence and defence genes.

Results

Transgenic Lr34res expression causes premature LTN in barley seedlings

The complete genomic sequence of the Lr34res, including 2.4-kb native wheat promoter and 1.5 kb native wheat terminator sequence, was introduced in the genetic background of barley cv. Golden Promise by Agrobacterium tumefaciens-mediated transformation.

Four independent transgenic events expressing the full-length genomic Lr34res sequence: BG8, BG9, B23 and B32 were all found to have comparable phenotypes (see SI for characterization of transgenic material). Collectively, they develop a severe leaf necrosis in seedlings at the 2–4 leaf stages (12–16 days postgermination––dpg), which starts with the first leaf at the leaf tip and is soon accompanied by the formation of chlorotic lesions on the leaf blade (Figures 1a and S1a). The premature leaf chlorosis continues throughout plant development with lower leaves dying as younger leaves emerge (Figure 1b). As a consequence, plants grown in standard glasshouse conditions (SGC, see SI for specifics) have fewer fully green leaves at maturity, compared with the nontransgenic control lines or untransformed wild-type cv. Golden Promise (Figure 1c), and are compromised in their growth and grain set. Interestingly, we found that plants of B23 and B32 grown at a constant temperature of 13 °C performed much better than plants grown in SGC. We hypothesize that those plants that produced little or no viable grains in SGC were predominantly homozygous for the transgene. Initial observations suggested that all barley transformants carrying the full-length genomic Lr34res insert underwent premature senescence. To test whether the LTN phenotype was indicative of Lr34res, genomic DNA blots were carried out on multiple T2 and T3 plants (Figure S2). This confirmed a cosegregation of the LTN phenotype with the insertion of the full-length genomic Lr34res in both heterozygous and homozygous plants; thus, we used LTN as a phenotypic marker for Lr34res, and the subsequent analysis was carried out on a mixture of heterozygous and homozygous plants.

Figure 1.

Heterologous expression of Lr34res in barley results in a severe LTN phenotype. (a) The first five leaves from BG8 (+Lr34res) and control (no Lr34res) grown in standard glasshouse conditions (SGC). (b) The last five leaves from the flag leaf down from B23 (+Lr34res) and control grown at 13 °C. Scale bars represent 10 mm. (c) BG8 and control line grown in SGC. (d) Representative plants from B32 (+Lr34res) and control grown at 13 °C.

To quantify the phenotypic effects of Lr34res expression in barley, transgenic plants from lines BG8, BG9, B23 and B32 were grown in SGC and compared with their respective controls for biomass (dry weight), plant height, number of heads and total grain weight (Table S1). There was a reduction in total grain weight in lines containing Lr34res, significantly so in the BG8, BG9 and B32 line. The total grain weight was improved when transgenic lines, B23 and B32, were grown at a constant temperature of 13 °C (Table S1). These transgenic plants exhibited a significant increase in height and only a small reduction in total grain weight per plant compared with their controls (Figures 1d, S1b, Table S1).

Transgenic Lr34res expression confers multipathogen resistance to leaf rust, stem rust and powdery mildew in barley at seedling stage

Lr34res was originally classified as a partial leaf rust resistance gene conferring resistance predominantly in adult wheat plants, with cold treatment at the seedling stage also resulting in Lr34-based resistance (Risk et al., 2012; Singh and Gupta, 1992). Interestingly, expression of Lr34res in transgenic barley confers strong resistance to barley leaf rust in seedlings without cold treatment. Transgenic plants inoculated at the 2–3 leaf stage (12 dpg) exhibited a strong resistance response to the invading pathogen on second and third leaves at 7–10 days postinfection (dpi), whereas the corresponding leaves of the controls were completely susceptible (Figures 2a and S3a). Analysis of infection sites (5 dpi) revealed a strong reduction in colony development in transgenic lines B23 and B32 when compared to their corresponding controls. Higher resistance levels were also observed in more mature BG8 transgenics inoculated 6 weeks postgermination when compared to the control, thus confirming Lr34res also functions as an adult plant resistance gene in barley (Figure S3b).

Figure 2.

Heterologous expression of Lr34res provides multipathogen resistance in barley seedlings. (a) Representative leaves of transgenic lines and control lines (B32 and B32 sib shown) grown in SGC and inoculated with P. hordei 9 dpi, scale bar = 10 mm. Inset, WGA-FITC-stained infection sites 5 dpi, scale bar = 0.5 mm. (b) Representative leaves of transgenic lines and control lines (BG8 and BG8 sib shown) inoculated with B. graminis f.sp. hordei 6 dpi, scale bars = 10 mm. (c) Leaves of transgenic lines and control lines (B32 and B32 sib shown) grown at constant 22 °C 16 h light/8-h dark and inoculated with P. graminis f.sp. tritici, 8 dpi, scale bar = 10 mm. Inset, WGA-FITC-stained infection sites 9 dpi, scale bars = 1 mm.

Lr34res-based resistance against powdery mildew has only been described in field experiments in wheat (Lillemo et al., 2008; Spielmeyer et al., 2005). We found that barley powdery mildew infections performed on younger plants (at 14, 21 and 28 dpg) of transgenic lines BG8, B23 and B32, and the corresponding controls showed a reduction in the severity of the infection in the transgenic lines (no or very small sporulating colonies) 6 dpi compared with their controls (Figures 2b and S3c). The resistance was particularly apparent in the first and second leaf of plants infected 14 dpg, but was less pronounced in younger leaves of the plants infected at later stages. The powdery mildew resistance was also observed in a plate assay, where inoculated leaf segments of second and third leaves detached from 14 dpg BG8 seedlings showed resistance (Figure S3e).

In addition to the two barley-specific fungal pathogens, the susceptibility of barley cv. Golden Promise to wheat stem rust allowed us to test Lr34res effectiveness with a third biotrophic pathogen in our heterologous system. In wheat, Lr34res provides partial resistance to stem rust in adult plants in some cultivars (Dyck, 1991; Kolmer et al., 2011). Although wheat stem rust grows more slowly on barley and is less virulent than barley leaf rust, transgenic seedlings; BG8, B23 and B32 infected 16–20 dpg showed some intermediate resistance without cold treatment. Analysis of WGA-FITC-stained stem rust colonies confirmed that both transgenics and controls contained a number of infection sites that failed to grow, but those that became established were larger and more frequent in the controls (Figures 2c and S3d). In summary, heterologous expression of Lr34res in barley provides resistance against three biotrophic barley pathogens. In contrast to Lr34res-containing wheat, this resistance is also effective in seedlings (12–28 dpg) without cold treatment, suggesting barley exhibits a higher level of resistance than is observed in wheat.

Expression of the full-length Lr34res cDNA or the genomic Lr34sus sequence in transgenic barley is low compared with genomic Lr34res and does not cause LTN or confer leaf rust resistance

Several independent transgenic cLr34res lines with the Lr34res cDNA expressed under the control of the native promoter showed neither LTN nor resistance to leaf rust. These lines contained the Lr34res cDNA gene either untagged (Bc lines) or extended by a GFP tag (Bc-GFP lines). Analysis by quantitative PCR (qPCR) revealed that transcript levels in the Bc lines were, on average, ~20- to 80-fold lower than in the transgenic Lr34res lines BG8 and BG9, respectively (Figures S4a–c). Thus, the lack of Lr34res-associated phenotypes in these lines is likely to be the result of lower transcript expression levels. Transgenic barley lines containing the genomic Lr34sus allele (BGS lines), under control of the native promoter and 3'UTR sequences, also showed neither LTN nor resistance to leaf rust across multiple lines. Quantitative expression analysis of Lr34sus transcript levels in the fourth leaf of several independent BGS lines (T1 plants) revealed that these lines express the Lr34sus transcript at ~12- to 38-fold lower levels than the transgenic Lr34res lines BG8 and BG9, respectively. Expression was at a comparable level to both Thatcher Lr34 (Th34) and Thatcher Lr16 (Th16) wheat seedlings (Lr34res and Lr34sus, respectively), but lower than in the wheat flag leaves (Figures S4d–f). The characterization of transgenic Bc, Bc-GFP and BGS lines is described in detail in the SI.

Lr34res expression in the heterologous barley system mirrors Lr34res expression in wheat

As Lr34res expression in barley causes a severe premature LTN phenotype in seedlings, we compared Lr34res transcript levels in barley seedlings with levels in transgenic wheat [Ta54 (Risk et al., 2012)] seedlings and with Th34 seedling or flag leaves (endogenous Lr34res). We found that Lr34res transcript levels in barley seedlings are generally higher than in Th34 wheat seedlings. However, BG8 seedling expression is comparable with the endogenous expression levels of Th34 flag leaves and BG9 expression with transgenic wheat Ta54 seedlings (Figure 3a). Yet, in contrast to transgenic Lr34res barley, Ta54 seedlings do not exhibit premature LTN. As the development of the LTN, general plant vigour and total grain weight varied between the four different transgenic lines, we assessed Lr34res expression in the fourth leaf from each line. The line B32 showed significantly higher expression levels than lines BG8 and B23 (Figure 3b). Interestingly, this high-expressing line also showed higher reduction in biomass and total grain weight (Figure 3b and Table S1). Thus, it is likely that there is an association between Lr34res transcript levels and impact on plant biomass and total grain weight. T-DNA copy number analysis showed lines BG9, B23 and B32 all had a single insertion, whereas BG8 has two copies (Figure S5), but shows the lowest level of expression (Figure 3b). This suggests that gene silencing may have an influence on the level of expression in BG8.

Figure 3.

Lr34res transcript levels in transgenic lines. (a) Comparison of Lr34res transcript levels between wheat and transgenic barley. Th34-SL and FL = cv. Thatcher Lr34 with endogenous Lr34res, fully emerged seedling leaf 4 and flag leaf, respectively. Ta54 = transgenic Bobwhite Lr34res wheat. SibTa54 = nontransgenic Bobwhite Lr34res wheat. Sib Hv = barley negative control. Fully emerged seedling leaf 4 was used for Ta54, SibTa54, Sib Hv and the BG lines. Data are relative to one of three controls: Actin, Tubb4 (β-tubulin 4) and RLILP (68 kDa protein HP68) ± SEM from four biological replicates relative to Th34-SL = 1. (b) qPCR analysis of Lr34res transcript levels in leaf 4 of transgenic barley lines (26 dpg) grown in SGC. Letters a and b indicate lines with equivalent expression levels (least significant difference, < 0.05). The table below the graph shows the reduction in biomass and total grain weight (in %) of each line compared with its corresponding control (grown at SGC). The reduction is calculated from the means of a minimum of six plants per line, see Table S1). (c) qPCR analysis of Lr34res transcript levels in transgenic barley line B32 grown at either 13 °C or under SGC. For both (b) and (c), data were normalized using Actin and Tubb4, and the data are ± SEM from four biological replicates (pooled sibs, n = 6) relative to a B32 sample = 1.

We then wanted to determine whether Lr34res expression during development of transgenic barley mirrored that of wheat, specifically if expression is highest in flag leaves. Similar to wheat, Lr34res expression was higher in flag leaves than in leaf 4 of transgenic barley seedlings (Figure 3c). Growing the plants at 13 °C did not result in a consistent reduction in Lr34res expression, but did reduce Lr34res-associated phenotypes (reduced LTN, increased total grain weight––Table S1). Thus, expression of Lr34res depends on the developmental stage of transgenic barley, but lower temperature does not seem to have an influence. In summary, barley developmental patterns and levels of Lr34res transcripts are comparable to wheat.

Lr34res expression in barley initiates premature activation of senescence marker genes and promotes increased expression of pathogenesis-related (PR) genes

Lr34res expression causes LTN in the flag leaves of wheat (Shah et al., 2011; Singh, 1992a), although the mechanism resulting in this phenotype remains unclear. Semi-qRT expression analysis of the senescence marker gene HvS40 (Krupinska, 2002) showed up-regulation in transgenic BG8 seedlings only at 12 dpg, but not at 8 dpg (Figure S6a), thus correlating the onset of LTN with an induction of HvS40. To further test whether the premature LTN seen in barley was correlated with senescence-related genes, we carried out qPCR on the third and fourth leaves of B23 and corresponding control 26 dpg using two additional senescence markers; the serine carboxypeptidase III precursor (CP-MIII) and the papain-like cysteine protease (Pap8) both associated with nutrient reallocation in senescing barley (Cantu et al., 2011; Cao et al., 2011; Parrott et al., 2010, 2007). RNApol II small subunit and Sec61p α subunit are stably expressed during senescence (Parrott et al., 2010) and were used as control genes. Significantly higher expression levels of CP-MIII, Pap8 and HvS40 were seen in the chlorotic third leaf of transgenic B23 plants compared with corresponding control leaf (Figure 4a). This confirms that the necrosis seen in the transgenic seedlings correlates with senescence. The levels of CP-MIII or Pap8 in the fourth green leaf did not differ between the control and the transgenic lines. However, HvS40 was up-regulated in the fourth leaf, indicating that HvS40 is activated before either CP-MIII or Pap8 and possibly acts earlier in the senescence cascade.

Figure 4.

Early LTN correlates with the induction of senescence-related genes, and uninfected Lr34res transgenic lines show elevated PR transcript levels. (a) qPCR analysis of CP-MIII, Pap-8 and HvS40 transcripts known to be up-regulated during senescence. Leaves were harvested 26 dpg (13 °C) from B23 (+Lr34res) and B23 sib lines. Leaf 3: necrotic and Leaf 4: green. Both Leaf 3 and Leaf 4 expressed Lr34res (Figure S6b). Data were normalized using RNApol II and Sec61p α. Data are ± SEM from four biological replicates relative to B23-Leaf 3-A = 1 (** < 0.001, * < 0.01 when compared to control line). (b) qPCR analysis of three PR gene transcript levels in leaf 4 of transgenic seedlings 26 dpg grown in SGC. Data were normalized using Actin and Tubb4. Data are mean ± SEM from 3–5 biological replicates and are relative to a BG8 sib sample = 1 (** < 0.001, * < 0.02 when compared to the corresponding control line).

In addition to the developmental processes, senescence genes are also often up-regulated in response to plant stress (Weaver et al., 1998). Likewise, due to complex hormone networks, some defence genes are triggered by senescence (Jukanti et al., 2007). As transgenic barley seedlings exhibited multipathogen resistance, we tested our uninfected transgenic lines for expression of three PR genes (PR1, PR2 and PR3) previously found to be up-regulated in wheat only postinfection (Hulbert et al., 2007; Risk et al., 2012). In contrast to wheat, Lr34res barley seedlings showed a significant increase in PR gene expression without infection in two lines. Interestingly, the Lr34res-low-expressing line B23 showed the highest levels of PR gene expression (Figures 3b and 4b).

Discussion

In this study, we demonstrate the functional transfer of the genomic sequence of the wheat Lr34res gene into barley. There, it confers partial resistance against pathogens that infect barley seedlings. The use of disease resistance genes that have been transferred from alien species to wheat has been recognized already in conventional wheat breeding during the last century. The 1BL-1RS wheat rye translocation provides an example of such transferred pathogen resistance [Sr31; Lr26; Yr9; and Pm8 (McIntosh et al., 2012)], and similarly Pm21, a durable powdery mildew resistance gene was transferred from the wild relative Haynaldia villosa to wheat as a translocation (Cao et al., 2011). In contrast to conventional breeding, the transgenic approach allows the transfer of a single gene between a wider range of sexually incompatible species. The functional interspecies gene transfer of disease resistance genes through transgenic approaches has been achieved within the Solanaceae family, where the tomato Pto gene was transferred to tobacco (Rommens et al., 1995). In the Poaceae, only the maize Rxo1 and the rice Xa21 genes have been functionally transferred to taxonomically diverse species (Mendes et al., 2010; Zhao et al., 2005). Both these genes confer resistance against bacterial pathogens, whereas Lr34 is effective against multiple fungal pathogens. To date, the only report on a fungal resistance gene functionally transferred between cereals comes from barley Mla6, which in a single-cell transient assay conferred resistance to the wheat nonhost pathogen B. graminis f.sp. hordei (Bgh) (Halterman et al., 2001). Other attempts to express resistance genes in a heterologous species were hampered by ‘restricted taxonomic functionality’ (RTF), which is thought to arise from the absence of essential components within the defence signalling pathways in the recipient species (McDowell and Woffenden, 2003; Tai et al., 1999). For example, the RPW8 gene, which conferred resistance against several powdery mildew isolates in Arabidopsis, was found to exert its function in tobacco, but not in tomato, presumably due to RTF (Xiao et al., 2003). Likewise, the maize Rp1-D gene conferring race-specific rust resistance in maize showed no resistance against pathogens of wheat and barley (Ayliffe et al., 2004). In summary, only a few resistance genes, mostly conferring resistance to bacterial pathogens, have been functionally transferred between sexually incompatible species, and this is predominantly in noncereal species. The successful transfer of Lr34res resistance to barley shows that all components involved in the biosynthesis of the putative substrate and the proteins involved in signal transduction are also present in barley. This finding is of importance as it will allow further genetic dissection of Lr34res-mediated resistance in this diploid species.

Lr34 confers broad-spectrum resistance in wheat to several biotrophic pathogens, and it is likely that it provides defence against diseases through a general physiological mechanism. In barley, the Lr34res-associated phenotypes (pathogen resistance and LTN) are produced earlier in plant development and are stronger than in wheat. Although both BG9 and Ta54 [a wheat line with the Lr34res transgene (Risk et al., 2012)] had similar levels of Lr34res expression (Figure 3a), Ta54 seedlings do not exhibit premature LTN, suggesting that additional regulatory mechanisms are present in wheat which reduce the effect of Lr34res overexpression. In addition, the early initiation of PR gene expression in barley prior to infection suggests that a comparable level of Lr34res expression can have a more profound effect in barley than in hexaploid wheat. Possibly, post-transcriptional mechanisms regulating Lr34res expression and protein stability in wheat are either weaker or absent in barley. Alternatively, differences in the metabolism, in particular, the amount or availability of the substrate of the LR34res transporter or the physiological susceptibility to the transported substrate between wheat and barley may account for the more extreme effect of Lr34res in barley.

The introduction of the cLr34res into barley did not result in LTN or pathogen resistance in either Bc or Bc-GFP lines (Figures S4a,b). As this coincided with reduced transcript levels (Figure S4c), we hypothesize that the cDNA transgene transcript may be subject to reduced expression due to intron-associated regulatory mechanisms that can ultimately affect mRNA stability, reviewed in Le Hir et al. (2003) and Rose (2008). As misregulation of Lr34res expression has a potentially detrimental effect on plant health, it may be that some of the 24 introns present in Lr34 contribute to the fine-tuning of transcriptional regulation.

Similarly to the cLr34res lines, we found that expression of the genomic Lr34sus sequence does not cause any of the Lr34res-associated phenotypes. Previously, Krattinger et al. (2009) found by semi-qRT that transcript levels of Lr34res and Lr34sus in wheat flag leaves are comparable, suggesting the two gain-of-function mutations in Lr34res are responsible for resistance in wheat, possibly through altered substrate specificity/affinity. Using a more sensitive qPCR assay, we show the transcript level of Lr34sus in transgenic barley seedlings is equivalent to that of Th16 seedlings (Lr34sus), but less than Lr34res barley seedlings. Thus, at this stage, we cannot exclude the possibility that the reduced Lr34sus transcript level is responsible, at least in part, for the susceptible phenotype in transgenic barley. Clearly, the differential expression level of transgenic Lr34res and Lr34sus needs further investigation.

In summary, our findings show that the introduction of genomic Lr34res is effective against the host-specific pathogens of barley. It remains to be seen whether this is also the case in other grass species. Unlike R genes, many of which are notoriously fleeting in their durability, Lr34res has been shown to be effective over a long time period and may provide additional advantages in cereal cultivar longevity when employed with other R genes (Brun et al., 2010; Wulff et al., 2011). If functional Lr34res expression can be controlled by appropriate cell-specific or pathogen-inducible promoters to avoid pleiotropic effects on growth and yield, the transfer of Lr34res into other agronomically important crops such as rice or maize has great potential as a new source of durable resistance in such species.

Experimental procedures

Production of stable transgenic Lr34 barley lines

The genomic Lr34res sequence was cloned into binary vectors at both University of Zürich (UZH) and CSIRO. The genomic Lr34sus construct was cloned at UZH, the cLr34res constructs were cloned at CSIRO and UZH, see SI for specifics. Stable transformation of barley cv. Golden Promise was as described by Hensel et al. (2009) or Tingay et al. (1997) with the callus-forming media modified to include 1.25 mg/L CuSO4 pentahydrate using the A. tumefaciens strain AGL-1. Identification of stable cLr34res transformants and plant growth conditions are described in detail in the SI. ‘Control’ refers to a corresponding nontransgenic sibling (sib) line of each event.

Identification of stable genomic Lr34res transformants

Primary transformant lines derived from p6U:gLr34res (UZH) were identified by PCR screens and genomic DNA analysis. Genomic DNA was isolated from leaves as described by Stein et al. (2001). Transgenic plants containing the Lr34res construct were identified using the PCR-based Lr34res marker cssfr1 (L34DINT9F/L34PlusR) (Lagudah et al., 2009) and the hygromycin phosphotransferase (HPT)/ubiquitin marker (primers GH-Ubi-F3/GH-Hyg-R2, Table S2). Transgene copy number was determined by genomic DNA analysis on cssfr1 (Lagudah et al., 2009)-positive plants using 10 μg genomic DNA digested with EcoRI or NruI. EcoRI-digested DNA was probed with a 32P-labelled probe covering the HPT gene of p6U (GH-Ubi-F3/GH-Hyg-R2), and the NruI-digested DNA was probed with a 32P-labelled Lr34 promoter PCR amplicon (Lr34PDRPROMF/Lr34PDRPROMR). At CSIRO, full-length insertion lines derived from pWBVec8:gLr34res were identified by digesting 12 μg genomic DNA with NotI and probing with the Lr34 promoter (Lr34PDRPROMF/Lr34PROMR) and 3'UTR probes consecutively (Lr343UTRF/Lr343UTRR). T-DNA copy number was established using genomic DNA blots where 10 μg of XbaI-digested DNA was probed with the HPT selectable marker (MNhpt5'_5+/MNhpt3'_515). The sequences of all primers referenced are given in Table S2.

Nine from 10 p6U:gLr34res-derived T0 events were positive for both Lr34res- and HPT-specific PCR-markers; however, only 5 T0 events produced progeny. From these 5 T0 transformants (BG1, BG6, BG8, BG9 and BG10), only BG6, BG8 and BG9 contained the full-length transgene insertion, whereas BG1 and BG10 had truncated copies of Lr34res. Genomic DNA blot analysis revealed that BG6 and BG8 are from the same gene transfer event, and BG9 is an independent event. Of the 36 hygromycin-resistant lines produced at CSIRO, 27 were found to have transgenic events, of which 5 (B1, B6, B7, B23 and B32) were positive for both the promoter and the 3'UTR of Lr34. The pWBVec8:gLr34res-derived T0 transformants: B1, B7, B23 and B32 all showed LTN, whilst transformant B6 did not. Subsequent analysis of B6 T1 plants using a PCR screen [L34SPF/L34DINT13R2 (Lagudah et al., 2009)] revealed an internal deletion within the Lr34res coding sequence. Genomic DNA blot analysis revealed that lines B1, B7 and B23 were derived from a single gene transfer event, henceforth referred to as B23. B32 proved to be an independent event. T-DNA copy number analysis by genomic DNA blot revealed lines BG9, B23 and B32 contained a single insertion of Lr34res, whereas BG8 had two copies. Finally, integration of the transgene in transformant lines was confirmed by full-length amplification of the Lr34res cDNA as previously described in Risk et al. (2012).

Pathogen infections and visualization

Leaf rust infections on seedlings and mature plants of transgenic barley were performed as previously described by Risk et al. (2012) using (at UZH) the P. hordei pathotype 1.2.1 (Qi et al., 1998), which is avirulent on plants with resistance genes Rph3, 7, 12 and virulent on lines containing Rph1, 2, 4, 5, 6, 8 and 9 (Shtaya et al., 2006), and (at CSIRO) P. hordei pathotype 4653P+ (PBIC culture number 990492), which is avirulent on plants with resistance genes Rph3, 5, 7, 10, 11, 14, and 15 and virulent on lines with resistance genes Rph1, 2, 4, 6, 8, 9, 12, 13, and 19. See figure legends for specifics.

Powdery mildew inoculations were performed on transgenic Lr34res barley 14, 21 and 28 days postgermination (dpg) (2–6 leaf stage). Pregerminated grains were transplanted and grown under SGC (see SI for specifics) until inoculation with B. graminis f.sp. hordei (Bgh). Plants were maintained in an incubator under SGC until evaluation of the resistance response. For the plate assay, 5-cm segments were cut from the second and third leaf of 14 dpg plants and placed on 0.5% phytoagar plates containing 30 ppm benzimidazole. Inoculation and incubation of the plates were performed as described above.

Stem rust inoculations were carried out as the leaf rust infections using P. graminis f.sp. tritici race 98-1, 2, 3, 5 and 6 (Park et al., 2009).

Pathogen infection sites were stained with wheat germ agglutinin-fluorescein isothiocyanate (WGA-FITC) and visualized according to Risk et al. (2012).

Quantitative PCR-based expression analysis

RNA extraction, cDNA synthesis and qPCR experiments were carried out as previously described by Risk et al. (2012) with the following changes: cDNA was synthesized using 2.5 μm oligo (dT)23-N or 1 μL of random primers (50 ng/μL) and 100 nm Lr34RT_r1, and qPCR was performed at an annealing temperature of 62 °C. Table S3 describes the primers used in expression analysis of Lr34res, cLr34res and Lr34sus (Figures 3a–c, S4c,f), senescence marker genes and PR genes (Figures 4a,b). Significance was assessed using REST 2009 software (Pfaffl et al., 2002; Vandesompele et al., 2002).

Acknowledgements

This work was supported by an Advanced Investigator grant of the European Research Council (ERC-2009-AdG 249996, Durable resistance), the Swiss National Science Foundation grant 3100A-12706/1, a Marie Curie International Outgoing Fellowship within the 7th European Community Framework Program (PIOF-GA-2009-252731, Dures) and the Grains Research and Development Corporation: grant #CSP00099, Australia. We thank Ming-Bo Wang, Plant Industry, CSIRO, for kindly providing pWBVec8. The authors thank Carola Bollmann, Sibylle Freist and Sabine Sommerfeld for their excellent technical assistance.

Ancillary