TaAbc1, a Member of Abc1-Like Family Involved in Hypersensitive Response against the Stripe Rust Fungal Pathogen in Wheat

To search for genes involved in wheat (Triticum aestivum L.) defense response to the infection of stripe rust pathogen Puccinia striiformis f. sp. tritici (Pst), we identified and cloned a new wheat gene similar to the genes in the Abc1-like gene family. The new gene, designated as TaAbc1, encodes a 717-amino acid, 80.35 kD protein. The TaAbc1 protein contains two conserved domains shared by Abc1-like proteins, two trans-membrane domains at the C-terminal, and a 36-amino acid chloroplast targeting presequence at the N-terminal. Characterization of TaAbc1 expression revealed that gene expression was tissue-specific and could be up-regulated by biotic agents (e.g., stripe rust pathogen) and/or by an abiotic stress like wounding. High-fold induction was associated with the hypersensitive response (HR) triggered only by avirulent stripe rust pathotypes, suggesting that TaAbc1 is a rust-pathotype specific HR-mediator. Down-regulating TaAbc1 reduced HR but not the overall resistance level in Suwon11 to CYR23, suggesting TaAbc1 was involved in HR against stripe rust, but overall host resistance is not HR-dependent.


Introduction
Plant hypersensitive response (HR), visualized as rapid death of host plant cells at the infection sites, is a means of plant defense response to microbial pathogens [1][2][3][4]. This process is under intrinsic fine genetic control similar to the programmed cell death (PCD) in animal cells. In mammals, PCD is defined as apoptosis, a complex physiological process that removes unwanted or harmful cells from a body [5]. In plants, HR is triggered by incompatible interactions of host plants and corresponding pathogens. Plant HR may deplete nutrient supply and effectively restrict pathogen growth, especially for biotrophs that only survive and reproduce on living cells. Accompanying rapid cell death, a set of defense responses including cell wall reinforcement, accumulation of reactive oxygen intermediates, antimicrobial compounds and pathogenesis-related proteins is also observed.
Decades of studies in animal and plant systems have revealed that mitochondria play an important role in integrating cell-death signals in regulating PCD and HR [6]. Several genes associated with the death signal pathway through mitochondria have been isolated and characterized. Among those genes, CABC1, a human mitochondrial gene that is highly similar to the yeast nuclear gene Abc1 (for Activity of bc1 complex), has been shown to play an important role in the p53-induced apoptotic pathway [7].
The first Abc1 gene was cloned and characterized in the yeast Saccharomyces cerevisiae, and was shown to suppress the cytochrome b mRNA translation defect [8]. Up to now, Abc1-like genes have been identified in nuclei mitochondria and chloroplasts [7][8][9] and have been shown to fulfill diverse functions. The yeast Abc1 gene was located in the nucleus but has a dual function in mitochondria; it controls the correct folding of cytochrome b and the assembly of bc1 complex in the mitochondrial respiration chain [10]. Abc1 was lately found to be necessary for yeast coenzyme Q synthesis [11][12]. The Abc1-like gene family has also been described as a new family of putative protein kinases [13]. A plant Abc1-like gene, the Arabidopsis AtOSA1 gene, located in chloroplasts, was found to be involved in balancing oxidative stress generated by Cd 2+ , hydrogen peroxide (H 2 O 2 ), and light [9].
In this study, we were searching for genes involved in wheat (Triticum aestivum L.) defense response to the stripe rust pathogen Puccinia striiformis f. sp. tritici (Pst). This fungal pathogen can cause wheat stripe rust disease, and is a worldwide problem threatening wheat production [14]. We isolated a wheat gene that shares the common features of Abc1-like genes. Here, we present the results on molecular characteristics of this newly identified wheat Abc1like gene, its expression profiles in response to Pst infection and abiotic elicitors and its involvement in wheat defense response to a group of rust fungal pathogens.

Identification and isolation of the TaAbc1 gene
From a cDNA library constructed from the wheat cultivar Suwon11 infected with a virulent Pst pathotype CYR31 [15], we selected an EST clone for further characterization in this study due to its strong up-regulated expression in the incompatible interaction of Suwon 11 infected with Pst pathotype CYR23. The clone contains a 597-bp cDNA fragment that is highly homologous to the Abc1 (for activity of bc1 complex) like gene family in rice (Oryza sativa) (GenBank accession no. AK120335.1). After database mining, we obtained a 2,675-bp wheat sequence in silico. The sequence was highly homologous to a cDNA clone (J013059I22) from O. sativa Japonica. To verify the gene sequence in wheat, a pair of primers TaAbc1-Fp/Rp (Table 1) was designed and used to amplify the candidate gene from wheat mRNA. A 2,550-bp cDNA sequence was successfully amplified from Suwon11. The sequence includes a 2,154-bp open reading frame (ORF), a 123-bp 59-untranslated region (UTR), and a 173-bp 39-UTR predicted using the DNASTAR software (www.dnastar.com). The encoded protein shares an 87% similarity to the Abc1-like protein of O. sativa (GenBank accession: BAD22484.1). We designated the wheat gene as TaAbc1 (GenBank accession: JX944037).
The TaAbc1 protein was predicted to have 717 amino acids, a molecular weight of 80.35 kD and an isoelectric point (pI value) of 9.38. The protein includes two conserved domains shared by Abc1-like proteins, two trans-membrane domains located between positions 659-679 and 685-703, and a 36-amino acid N-terminal chloroplast targeting presequence (Fig. 1). The two conserved domains are the domain of 121-amino acids from the region of 267-387 that shares the characteristics of Abc1-like proteins; and the other of 321-amino acids of a putative kinase domain from the region of 276-596.

TaAbc1 gene expression
Real-time quantitative PCR was used to measure the relative transcript abundance level of the gene. For assaying the gene expression in different wheat organs, we measured the TaAbc1 transcripts in leaves, stems and roots at the seedling stage of Suwon11 without pathogen challenge. The gene expression patterns appeared to be tissue-specific. The highest TaAbc1 transcript abundances were detected in stems. Normalizing the level of TaAbc1 transcript abundances in leaves as 1, stems had about 50% higher level than that in leaves. In contrast, roots had the lowest --only about 11% of the level in leaves (Fig. 2).
To study the expression profiles of TaAbc1 during the course of rust infections in both incompatible and compatible interactions, we challenged Suwon11 with two Pst pathotypes, CYR23 and CYR31. Suwon11 was resistant to CYR23, and susceptible to CYR31, therefore, Suwon11/CYR23 represented an incompatible interaction, and Suwon11/CYR31 a compatible interaction. In addition, we included in the study two additional incompatible interactions with a USA spring wheat genotype Scholar with a leaf rust resistance gene Lr47, inoculated with pathotype PBJL of the leaf rust pathogen (P. triticina) and pathotype QFCSC of the stem rust pathogen (P. graminis f. sp. tritici). TaAbc1 transcript abundances were measured in the leaf tissues collected at 5 time points (Fig. 3) challenged with different rust species and pathotypes in two different wheat cultivars. In both incompatible and compatible interactions, the relative expression of TaAbc1 was up-regulated and peaked at 24 hours post inoculation (hpi) and declined at 48 hpi. The transcripts remained at higher levels than each of their corresponding 0-hpi-controls at 72 hpi in Suwon11 challenged with both stripe rust pathotypes, while the levels in Scholar+Lr47 were soon back to the control level at 48 hpi (Fig. 3). The striking difference was that the up-regulated level of TaAbc1 transcripts was much higher (as much as 7-fold) in the incompatible interaction of Suwon11/CYR23, and increased only up to 2-fold in the rest of the interactions by comparison to each of their corresponding controls at 0 hpi (Fig. 3). By 120 hpi, the TaAbc1 expression level resumed the level of its 0-hpi-control in the incompatible interaction but was still slightly higher than the 0hpi-control in the compatible interaction. In summary, TaAbc1 expression increased and decreased dramatically over the course of   120 hours post inoculation in the incompatible interaction of Suwon11/CYR23; whereas, in the rest of interactions, TaAbc1 was only slightly increased and decreased over the same time period.
Since TaAbc1 was up-regulated and peaked at 24 hpi with rust challenge in both incompatible and compatible interactions, we asked the question whether the gene can be induced by a signal molecule. We treated Suwon11 with four different signal molecules including salicylic acid (SA), jasmonic acid (JA), ethanol (ET) and abscisic acid (ABA) at the seedling stage. We collected leaf tissues at 5 different time points up to 24 hours post treatment (hpt) and used 0 hpt as the control for each corresponding treatment (Fig. 4A). We normalized the level of TaAbc1 transcript abundances at 0 hpt as 1 for each treatment. As shown in Fig. 4A, TaAbc1 was slightly down-regulated at 2 hpt in all except the JA treatment, and then was lingering around the 0 hpt level up to 12 hpt. At 24 hpt, the TaAbc1 transcript abundances were increased about 3-fold only in the leaves treated with JA and ET, but the magnitudes were significantly smaller than those induced by stripe rust (Fig. 3). No significant changes in TaAbc1 expression occurred with the SA and ABA treatments.
TaAbc1 expression was induced by rust inoculations but was unchanged with SA treatment. To search for the signal for TaAbc1 induction by rust, we treated wheat plants with 100 mM H 2 O 2 because oxidative burst is one of the earliest host responses upon pathogen detection [16]. Real-time PCR revealed a significant 6fold increase in TaAbc1 transcript level in the H 2 O 2 treated plant as early as at 2 hpt (Fig. 4B), then a decrease to twice the 0-hpt level by 12 hpt. A second small boost of TaAbc1 level was seen at 48 hpt, about a 3-fold increase in the treated plants than in the control (Fig. 4B). Cell death was visualized at 72 hpt on the H 2 O 2 treated leaves (data not shown).
To understand the induction by exogenous treatment of JA and ET, we wounded wheat plants by a sterilized scissors and measured TaAbc1 transcript abundant level at 5 time points as shown in Fig. 4C. As much as a 7-fold increase in TaAbc1 expression level was detected at 6 hpt, and then the transcripts dramatically declined to an undetectable level by 24 hpt (Fig. 4C).
Down-regulating TaAbc1 affects host hypersensitive response triggered by an avirulent pathotype of Pst Rapid induction of TaAbc1 gene expression was associated with incompatible interaction. To test if TaAbc1 has any functional role in wheat defense response to stripe rust, we knocked down the endogenous TaAbc1 gene in Suwon11 using the barley stripe mosaic virus induced gene silencing (BSMV-VIGS) assay. The silencing vector was constructed to contain a 160-bp fragment of TaAbc1, designated as BSMV:TaAbc1. As controls, a construct carrying only the BSMV genome, designated as BSMV:00, and a construct carrying a 120-bp wheat phytoene desaturase (PDS) gene, designated as BSMV:PDS, were included in the study. In addition, Suwon11 seedlings inoculated with BSMV inoculation buffer were also included as a mock control. Nine days post viral RNA inoculations, plants inoculated with BSMV:PDS started to show photobleaching (Fig. 5A), indicating BSMV induced gene silencing was initiated. Twenty-four each of the TaAbc1 silenced plants and control plants were then challenged with each of the two stripe rust pathotypes CYR23 and CYR31. Fungal infection types were assessed at 15 days post rust inoculation. As shown in Fig. 5B, Suwon11 had a resistant response to CYR23, exhibited with a high density of necrosis resulting from hypersensitive response (HR) on the mock and BSMV:00 control leaves. TaAbc1 silenced plants showed the same level of resistance to CYR23 as the controls, but the density of necrosis was significantly less than that on the controls (Fig. 5B). Meanwhile, none of the treatments changed Suwon11 susceptible response to CYR31, and no detectable difference was observed among BSMV:TaAbc1 treated plants and controls. To make sure the silencing assay was effective, leaf tissues were collected from both BSMV:TaAbc1 and BSMV:00 treated plants inoculated with either CYR23 or CYR31 right after fungal inoculation at 0, 48 and 120 hpi. Real-time PCR revealed that TaAbc1 transcript abundances from all the BSMV:TaAbc1 treated plants were reduced to about 20% to 40% of the relative levels of the BSMV:00 treated control plants at the time of rust inoculation and up to 120 hpi (Fig. 6). The results confirmed the silencing of TaAbc1 was effective, and knocking down TaAbc1 reduced HR in Suwon11, triggered by the avirulent pathotype CYR23 (Fig. 5B).
Hypersensitive cell death was further examined at 48 and 120 hpi under a florescence microscope as shown in Fig. 7. Autoflorescence was the evidence of cell death, visualized as a necrotic area on leaf surface, and was not detectable on the leaves inoculated with the virulent pathotype CYR31 at either 48 hpi (  (Fig. 7G). Weak auto-florescence was visible at 48 hpi on the leaves in which TaAbc1 gene was knocked down (Fig. 7D). The florescence areas were also expanded at 120 hpi ( Fig. 7H) but were relatively smaller and weaker than those on the leaves of mock and BSMV:00 inoculation ( Fig. 7F and 7G). Fungal development and host responses in CYR31-inoculated plants were similar to what have been described for Suwon11-Pst compatible interaction [17]. We further measured the size of necrotic area per infection site under a microscope. In an incompatible interaction with CYR23, Suwon11 had an average size of 4 mm 2 necrotic area per infection site on the control leaves at 48 hpi, and only about 1.6 mm 2 on the leaves of TaAbc1 silenced plants, which was significantly (P,0.01) smaller than the control (Table 2). At 120 hpi, average sites of necrotic areas became bigger, about 14 mm 2 per infection site on the control, and 8 mm 2 on the TaAbc1 silenced leaves. However, there was no detectable difference observed by the naked eye in resistance level to CYR23 between the controls and TaAbc1 silenced plants (Fig. 5B). Under microscopy, fungal hyphae were significantly (P,0.01) longer in the TaAbc1 silenced leaves compared with the leaves of the two controls at both 48 and 120 hpi ( Table 2). These results suggested that HR had a positive

Discussion
We isolated and characterized a new gene involved in wheat resistance to stripe rust pathogen. The gene resembles the structures of a group of Abc1-like genes in animals and plants. In the Arabidopsis genome, there are 17 putative Abc1-like genes [9]. Among them, At3g07700 shares 72% amino acid identity with  TaAbc1. However, so far, only two Abc1 representatives have been characterized in plants. One is At4g01660, localized in mitochondria [8,18], and is a homolog of the yeast Abc1 gene [18]. The encoded protein shares only 28% amino acid identity with TaAbc1.The second one, AtOSA1, shares 46% amino acid identity with TaAbc1 (Fig. S1), and is the first member of Abc1 family to be localized in the chloroplasts with a function of balancing oxidative stress [9]. Mutants with T-DNA insertion on AtOSA1 were more sensitive to H 2 O 2 . TaAbc1 and AtOSA1 exhibited the same tissue-specific expression patterns that predominantly expressed in leaves and stems, with only a very low level in roots. In addition, both TaAbc1 and AtOSA1 proteins carry an N-terminal chloroplast targeting presequence [9]. These results tend to suggest that the TaAbc1 may locate and execute its function in chloroplasts.
TaAbc1 plays an important role in stress response. The gene transcript level followed a complex kinetics, with as much as 2.5,7-fold induction in response to different stresses, but cell death was only observed on leaves when the induction was as much as 7-fold up-regulated. Our observations imply the existence of a threshold, responding to the combination of stress signals initiated by either a biotic agent (e.g. stripe rust) and/or an abiotic stress like wounding, above which a cell commits itself to suicide. All three rust inoculations which up-regulated the TaAbc1 transcript abundances in leaves peaked at 24 hpi. However, 7fold induction was only seen in the incompatible interaction of Suwon11/CYR23. The expression of TaAbc1 was not affected by SA treatment, suggesting that TaAbc1 was positioned either upstream of SA or in an SA-independent manner. Similar levels of 2.5,3-fold induction were detected in the compatible interaction of Suwon11/CYR31 as well as the other two incompatible interactions of Scholar+Lr47/PBJL and Scholar+Lr47/QFCSC, where HR was associated with resistance response. It was tempting to speculate that the 2.5,3-fold induction was an induction level responding to the basal defense triggered by a rust fungal molecule pattern. An exogenous treatment of JA/ET up-regulated TaAbc1 transcription but the increase was not as much as that detected in the directly wounded plants. JA and ET are two key signaling molecules during plant defenses against necrotrophic pathogens and insects as well as wounding [19]. Induction with exogenous treatment by JA/ET may suggest the involvement of TaAbc1 in wounding signaling.
Down-regulating TaAbc1 reduced HR, elicited by the incompatible interaction between Suwon11 and Pst pathotype CYR23. In addition, TaAbc1 transcripts were highly up-regulated by 100 mM H 2 O 2 . These observations invited the hypothesis that TaAbc1 could be a commonly shared cell-death signal used in wheat defense response to rust fungal pathogens. However, the   TaAbc1 transcript abundance level responding to leaf and stem rust in the two incompatible interactions tested did not support this hypothesis. Our results tend to suggest that TaAbc1-mediated HR was rust species and pathotype specific only in the incompatible interactions triggered by stripe rust. It is unclear yet whether all HR seen in incompatible interactions between wheat and stripe rust pathotypes go through TaAbc1. With current rather limited observations, we cannot exclude the possibility that TaAbc1mediated HR is R-gene specific. Microscopic observations did reveal the timing differences of HR triggered by different rusts. HR associated with resistance to stripe rust fungus appeared at 18 hpi [17], earlier than observed in resistance to leaf or stem rust pathogens in wheat. In general, HR triggered by leaf or stem rust pathogens appeared at 48 hpi (Huang unpublished data). Conserved mechanisms and mediators regulating cell death have been found in animals and plants, but studies also pointed out multiple signaling pathways and different organelles executing programmed cell death [6]. In plants, in addition to mitochondria, the plastid organelles also participate in HR-associated signaling [20,21]. It is not surprising to learn that HR induced by different rust pathogens used different mediators located in different organelles.
Silencing TaAbc1 reduced HR triggered by Suwon11/CYR23 incompatible interaction. Significantly longer fungal hyphae were observed under a microscope in TaAbc1 knocked-down plants compared to the controls. However, there were no visible differences in overall stripe rust resistance between the TaAbc1 silenced plants and the controls. The results suggest that HR is an important but not an essential defense response to stripe rust in wheat. HR reinforces growth arrest of stripe rust pathogen via depletion of nutrient supply, but overall host defense is not HRdependent. Accumulating results from studies in wheat and other plant systems show that not all R-gene-triggered resistance requires cell death [4]. Potato Rx-mediated resistance to potato virus X (PVX) is not associated with HR [22]. Wheat leaf rust resistance gene Lr34-mediated resistance has no visible HR [23]. In Arabidopsis, the dnd1 (for defense with no death) mutant expresses resistance when the HR is blocked [24]. Undoubtedly, plant defense response signaling is not a simple straight line; the networks could be parallel, branching or converging. Although HR is a very common rust pathogen resistant phenotype in wheat, it is not an obligatory reaction. Our study suggested that signaling pathways to HR and other defense responses induced by CYR23 in wheat are branched.
The function of TaAbc1 in resistance toward stripe rust was probably related to reactive oxygen species (ROS). TaAbc1 transcripts were highly up-regulated by exogenous treatment of 100 mM H 2 O 2 . AtOSA1 T-DNA-insertion atosa1 mutants showed increased sensitivity toward H 2 O 2 and high light. AtOSA1 transcript levels were up-regulated in wild type after 1 mM cadmium treatment that induced oxidative stress although the mechanism of how the stress is induced by this metal is still obscured [9]. Cd 2+ can inhibit electron transferring in membranes and induces ROS formation [25]. AtOSA1 probably acts as part of a signal transduction pathway to balance oxidative stress. In the incompatible interaction of Suwon11/CYR23, the generation and accumulation of ROS was detected at infection sites and is associated with the occurrence of HR [17]. In conclusion, we infer that TaAbc1 plays an important role mediating hypersensitive response. However, the mechanism of TaAbc1 in wheat defense response and the relationship with ROS need to be further investigated.

Plant materials
Two wheat (T. aestivum L.) cultivars Suwon11 (Chinese spring wheat) and Scholar+Lr47 (USA spring wheat), two stripe rust Pst pathotypes CYR23 and CYR31, one leaf rust pathotype PBJL and one stem rust pathotype QFCSC were the biological materials used in this study. Seedlings of Suwon11 and Scholar+Lr47 were grown and maintained followed the procedure described by Kang and Li [26]. Suwon11, containing the stripe rust-resistance gene YrSu [14], is highly resistant to CYR23 but is highly susceptible to CYR31 [27]. Scholar+Lr47 is highly resistant to leaf rust PBJL and stem rust QFCSC.

Cloning and sequence analysis
To clone the TaAbc1 gene, in silico cloning was performed as previously described by Zhang et al. [28]. The rice gene sequence extracted from GenBank was initially used as the seed probe to screen the wheat EST database in GenBank using BLASTN analysis. The sequence was assembled as described by Xia et al [29]. To verify the assembled sequence, primers TaAbc1 (TaAbc1-Fp/Rp, Table 1) were designed to amplify the full length cDNA of TaAbc1. The template was a mixture of the first strand cDNA samples extracted from leaves of Suwon11 post inoculated with CYR23 (an incompatible combination) at 12, 24 and 48 hpi. The amplified PCR products were cloned into a pGEM-T Easy Vector (Promega, Madison, WI, USA) and sequenced with an ABI PRISM 3130XL Genetic analyzer (Applied Biosytems, USA

Rust inoculations and chemical treatments
Stripe rust inoculations were conducted at Northwest A & F University, China. Leaf and stem rust inoculations were conducted at Plant Growth Center, Montana State University, USA. For stripe rust inoculation, freshly collected urediniospores were applied with a paintbrush to the surface of primary leaves of 7day-old wheat seedlings. After inoculation, plants were incubated for 24 hours in dark in a 100% humidity dew chamber and were subsequently transferred to a growth chamber with a 16 hour photoperiod. Mock stripe rust control inoculation was carried out with ddH 2 O. For leaf and stem rust inoculations, rust urediniospores were suspended in Soltrol 170 isoparaffin (Chempoint, Bellevue, WA). The mocks were carried out with Soltrol 170. For details, visit http://vimeo.com/48605764 to view the videotaped protocols.

RNA extraction and quantitative real-time PCR (qRT-PCR)
For tissue-specific expression analyses of TaAbc1, intact root, stem, and leaf tissues of 2-week-old wheat seedlings were respectively sampled from the same plants at the same time for the same time point. Sampled tissues for the time-course study were immediately frozen in liquid nitrogen, and stored at 280uC prior to extraction of total RNA. Three independent biological replications were performed for each experiment.
RNA for real-time PCR was isolated with Trizol Reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. First-strand cDNA was synthesized with the GoScript Reverse-Transcription System (Promega Corp., Madison, WI, USA). Primer design and qRT-PCR reactions were conducted as described by Wang [31]. Quantitative real-time PCR was performed using a 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The wheat elongation factor TaEF-1a gene (GenBank accession number Q03033) was used as an internal control for qRT-PCR analysis. The gene-specific primers, TaEF-F/R (amplifying a 85-bp TaEF-1a fragment) and TaAbc1-fp/rp (amplifying a 167-bp TaAbc1 fragment) for qRT-PCR, are listed in Table 1. Dissociation curves were generated for each reaction to ensure specific amplification. Threshold values (Ct) generated from the ABI PRISM 7500 Software Tool (Applied Biosystems) were used to quantify relative gene expression using the comparative 2 2DDCT method [32].

BSMV-mediated gene silencing
Plasmids used for gene silencing were based on the constructs described by Holzberg and associates [33]. A cDNA fragment (120 bp) of the wheat phytoene desaturase gene TaPDS was obtained by reverse-transcriptase polymerase chain reaction (RT-PCR). A RNA-derivative clone (BSMV:PDS) was created using BSMV:GFP (green fluorescent protein) cDNA as the starting material. The GFP coding sequence in BSMV: GFP was replaced by a PDS fragment in an antisense orientation resulting in BSMV:PDS. BSMV:TaAbc1 was constructed by the same approach; a 160 bp TaAbc1 cDNA fragment amplified by a pair of primers TaAbc1-VIGS-F/R (Table 1), derived from the 39untranslated region was inserted.
Capped in vitro transcripts were prepared from linearized plasmids that contain the tripartite BSMV genome [34] using the mMessage mMachine T7 in vitro transcription kit (Ambion, Austin, TX, USA) following the manufacturer's instructions. The BSMV inoculum was made by combining an equimolar ratio of a, b, and c transcripts at a 1:1:1 ratio with excess inoculation buffer (named as FES) containing a wounding agent. The third leaf of a three-leaf wheat seedling was inoculated with BSMV transcripts by gently rubbing the leaf surface with a gloved finger [33,35,36]. As a control, 24 seedlings were inoculated with 16 FES buffer as a mock [37]. Three independent sets of inoculations were performed, with a total of 72 seedlings inoculated for each of the three BSMV viruses (BSMV:00, BSMV:PDS, and BSMV:TaAbc1). Post viral inoculation, wheat plants were maintained in a growth chamber at 2362uC, and examined for symptoms at regular intervals. Once photobleaching was observed, three independent sets of inoculations including CYR23, CYR31 and sterile water (as a mock), were performed. Infection types of stripe rust were examined at 15 days post rust inoculation. The fourth leaves corresponding to the photobleached areas of BSMV:PDS infected plants were divided for histological observation and real-time PCR assay.

Histological observation of fungal growth and host response
Wheat leaves infected with BSMV were sampled at 0, 48 and 120 hpi with Pst and stained as described by Wang [17]. Cleared leaf segments were examined under an Olympus BX-51 microscope (Olympus Corp., Tokyo) for infection sites and lengths of infection hyphae. Auto-fluorescence of attacked mesophyll cells was observed as a necrotic death area by epifluorescence microscopy (excitation filter, 485 nm; dichromic mirror, 510 nm; and barrier filter, 520 nm). At least 50 infection sites were examined on each of five randomly selected leaf segments per treatment. Only the infection sites with fungal appressoria formation over stomata were considered to have been successfully penetrated. The necrotic leaf area was measured with a calibrated eyepiece micrometer and corresponding areas (square micrometers) calculated according to the formula p 6 length 6 width/4. Standard deviations and Tukey's test for statistical analysis were performed with the SPSS software (SPSS, Inc. Chicago, USA). Figure S1 Amino acid comparison between TaAbc1 and AtOSA1 on the two conserved domains of Abc1-like proteins.