Stagonospora Nodorum: Cause Of Stagonospora Nodorum Blotch Of Wheat

Transcript

   MOLECULAR PLANT PATHOLOGY   (2006) 7  (3), 147–156DOI: 10.1111/J.1364-3703.2006.00326.X© 2006 BLACKWELL PUBLISHING LTD   147  BlackwellPublishingLtd  Pathogen profile  Stagonospora nodorum   : cause of stagonospora nodorum blotch of wheat  PETER S. SOLOMON, ROHAN G. T. LOWE, KAR-CHUN TAN, ORMONDE D. C. WATERS AND RICHARD P. OLIVER*  Australian Centre for Necrotrophic Fungal Pathogens, Division of Health Sciences, Murdoch University, Perth, WA 6150, Australia    SUMMARY   Stagonospora nodorum    is an important pathogen of wheat andrelated cereals, causing both a leaf and glume blotch. This reviewsummarizes recent advances in our understanding of taxonomy,control and pathogenicity of this species.   Taxonomy: Stagonospora    (syn. Septoria    ) nodorum    (Berk.)Castell. and Germano [teleomorph: Phaeosphaeria    (syn.   Leptosphaeria    ) nodorum    (Müll.) Hedjar.], kingdom Fungi, phylumAscomycota, subphylum Euascomycota, class Dothideomycetes,order Pleosporales, family Phaeosphaeriaceae, genus Phae- osphaeria    , species nodorum    .   Host range: Wheat, Triticum aestivum    , T. durum    , Triticale,are the main hosts but other cereals and wild grasses havebeen reported to harbour S. nodorum.   Disease symptoms arelens-shaped necrotic lesions on leaves, girdling necrosis on stems(especially the nodes, hence ‘   nodorum    ’) and lesions on glumes.Mature lesions produce pycnidia scattered throughout thelesions, especially as tissue senesces.   Useful websites: http://ocid.nacse.org/research/deephyphae/htmls/asco_taxlist_spat.html (taxonomic information),http://ohioline.osu.edu/ac-fact/0002.html (disease information),http://wwwacnfp.murdoch.edu.au/ (ACNFP homepage),http://www.broad.mit.edu/annotation/fungi/stagonospora_nodorum/index.html (genome sequence homepage),http://cogeme.ex.ac.uk/efungi/ (genome sequence annotation   and analysis).   INTRODUCTION   Stagonospora    (syn. Septoria    ) nodorum    (Berk.) Castell. and Germano[teleomorph: Phaeosphaeria    (syn. Leptosphaeria    ) nodorum    (Müll.)Hedjar.] is a major pathogen of wheat and other cereals. In wheat,the disease is called stagonospora nodorum blotch, septorianodorum blotch and glume blotch. The first of these terms willbe used hereafter. It should not be confused with septorialeaf blotch, caused by Septoria tritici    , whose teleomorph is   Mycosphaerella graminicola    (Palmer & Skinner, 2002).   S. nodorum    causes a necrotic leaf blotch as well as discolorationof the head in the symptom known as glume blotch. Resistancein wheat is at best partial and is often polygenic, hence resist-ance breeding is time-consuming and difficult. The pathogenis, however, well contained if appropriate fungicide regimes areused. These can be used not only on foliar outbreaks but also tocontrol spread through infected seed.   S. nodorum    can be cultured in vitro    and is experimentally ame-nable to a range of molecular techniques, including homologousgene replacement. In recent years a small number of genes havebeen analysed and a nascent picture of the molecular basis of pathogenicity is emerging.In this review, we summarize recent research, focusing onpapers published after the volume ‘Septoria on Cereals; a studyof pathosystems’ published in 1999 (Lucas et al    ., 1999). We hopethis review will set the stage for the detailed analyses that willfollow the elucidation of the complete genome sequence andwill encourage others to consider this organism in their research.   TAXONOMY   Molecular techniques have provided a firm foundation forthe phylogenetic placement of Septoria nodorum    (Cunfer andUeng, 1999). ITS sequencing has shown that Septoria tritici    and S. nodorum    are not closely related (Goodwin and Zismann,2001). Septoria tritici    was placed among the Mycosphaerellaeand Cladosporia, whereas Septoria nodorum    was moved to   Stagonospora nodorum    (Berk.) Castell. and Germano [teleo-morph: Phaeosphaeria    (syn. Leptosphaeria    ) nodorum    (Müll.)Hedjar.]. The higher order taxonomy is still controversial but arecent review based on full 18S rRNA gene sequences places   Stagonospora    in the Pleosporales, a grouping that appears to bemonophyletic (Cunfer and Ueng, 1999; Goodwin, 2004). Similar  *  Correspondence   : Tel.: +61 893607404; E-mail: [email protected]    148   P. S. SOLOMON et al.   MOLECULAR PLANT PATHOLOGY   (2006) 7  (3), 147–156© 2006 BLACKWELL PUBLISHING LTD   results have been obtained using mating type genes (Bennett   et al    ., 2003). These data put Stagonospora    closest to the phy-topathogens Leptosphaeria    , Pleospora    and the Phoma herbarum    group that includes the Ascochyta. Other close genera include   Venturia    and a tight group that includes Cochliobolus    , Pyreno- phora    and Alternaria    (Fig. 1). It is intriguing to note that so manydevastating pathogens are so closely related.   HOST RANGE   Although most closely associated with wheat, S. nodorum    is alsopathogenic on barley (   Hordeum vulgare    ), and isolates from eachof these hosts have been shown to be capable of infecting theother (Osbourn et al    ., 1986). In addition, there are reports of theisolation of S. nodorum    from a wide range of wild grasses, whichthe pathogen may use as alternative hosts (Williams and Jones,1973). The oat pathogen Stagonospora avenae    is believed to bea distinct species, although it can occasionally be isolated fromwheat (Cunfer and Ueng, 1999).   SYMPTOMS   The broad features of the infection process have long beenestablished (Bird and Ride, 1981). Multiple germ tubes extend fromthe ends and middle of the spore and attempt to penetrate bothdirectly via the cuticle and opportunistically through stomata.Penetration through the cuticle has been associated with the devel-opment of swellings both at the hyphal tip and on lateral branches.Perhaps these are examples of a simple appressorium although,in our experience, these are only rarely seen and are not essentialfor direct penetration of the cuticle. Initial leaf effects includeyellowing at the site of infection and tip burn of the leaf (Fig. 2A).The chlorotic front expands to form oval-shaped lesions on the leaf,with the majority of hyphae running parallel to the vasculatureof the leaf. Within the chlorotic areas, small regions of necrosisare often evident. After a week or so under conditions of con-trolled humidity, pycnidia will begin to form in the lesion.Pycnidia are translucent at very early stages of development butrapidly darken and expand to form pale brown growths (Douaiher   et al    ., 2004). The release of pycnidiospores, usually in a massof pink-pigmented cirrus, is preceded by the swelling of a singlepoint on the pycnidial surface forming a protuberance whichruptures the cuticle. Pycnidia can form throughout the lesion,with no particular pattern prevalent (Fig. 2B). Once the entire leaf collapses and chlorosis is total, the fungus quickly ramifiesthrough the tissue and asexual sporulation begins en masse    .Successive rounds of asexual reproduction and dispersal candistribute S. nodorum    throughout the plant architecture. Furtherspread within the wheat canopy can lead to glume blotch, as thefungus attacks the glumes directly.   IMPORTANCE AND GEOGRAPHICAL RANGE   Stagonospora nodorum    is a major pathogen on cereals in majorwheat-growing regions of the world (Wiese, 1987). The pathogen Fig. 1 Outline taxonomic placement of Stagonospora nodorum  . Adapted from Goodwin (2004). Fig. 2 (A)   Macro symptoms on wheat seedling leaves 7 days post-infection by Stagonospora nodorum  ; (B)   sporulation occurring on infected detached leaves.    Stagonospora nodorum   149  © 2006 BLACKWELL PUBLISHING LTD  MOLECULAR PLANT PATHOLOGY   (2006) 7  (3), 147–156   causes up to 31% loss of yield in wheat (Bhathal et al    ., 2003). Adetailed survey of pathologists’ opinions has been carried out inAustralia (Brennan and Murray, 1998). This survey revealed that   Stagonospora nodorum    causes 5–15% losses across the wholeof the Western Australian wheatbelt, with heavier losses in thoseareas of higher rainfall. It is the most important pathogen bysome margin in this area. In the rest of Australia, losses wereestimated to vary from zero to 5% (Fig. 3). This pattern is intrigu-ing as many other areas of Southern Australia share essentialfeatures of their climate with Western Australia (WA). It could bespeculated that S. nodorum    is more successful in WA because therainfall periodicity more closely aligns with the sporulation/colonization cycle of the pathogen.In some Western European countries it has been reported thatstagonospora nodorum blotch has been replaced in recent yearsby septoria leaf blotch as the most important foliar necrotrophicdisease of wheat (Eyal, 1999b; Hardwick et al    ., 2001). This hasbeen attributed to differential responses by the pathogens tofactors such as cultivars and fungicides employed, and to alteredcultural practices or climatic conditions (Eyal, 1999b). PCR analysisof archived wheat straw samples covering 160 years since 1843suggests a significant correlation with atmospheric SO  2   and therelative abundance of S. nodorum    and S. tritici    (Bearchell et al    .,2005) in southern England. However, the reasons for theseobserved trends, and whether they are coincidental or the resultof some interaction, are not clear. As it is not known why theswitch in dominant pathogen occurred, there is a great need tomaintain vigilance in case S. nodorum    was to re-emerge.   LIFE CYCLE   Infected seeds and ascospores are considered to be the majorprimary inocula for stagonospora nodorum blotch (Shah et al    .,1995, 2000). Long-range air dispersal is primarily mediated bysexual ascospores during winter (Bathgate and Loughman, 2001).The release of ascospores from stubble of the previous year’s cropis initiated by low temperature, rainfall and high relative humidity(Arseniuk et al    ., 1998; Bathgate and Loughman, 2001). Second-ary spread of the pathogen occurs via splash-dispersed pycnid-iospores (Shah et al    ., 2001). It is likely that 2–4 cycles of asexualinfection are necessary for the fungus to produce significant dis-ease and, in particular, to infect the wheat heads (Shah and Berg-strom, 2002). There is circumstantial evidence that the asexualsporulation coincides with the pattern of 2–3 days of rainfall every8–12 days that is a feature of the winter weather pattern in Med-iterranean climates (Bathgate and Loughman, 2001).The fungus can be readily propagated in vitro    on a variety of synthetic and derived media. Sporulation occurs within 4–8 daysand is promoted by near-UV light. The sexual phase is difficult toreproduce in the laboratory. Many laboratories have not been able toestablish conditions under which ascospores can be produced. Inthe laboratories which do report success, the process takes severalmonths (Czembor and Arseniuk, 2000; Halama, 2002). Hence, sexualcrossing is not currently a practical method of genetic analysis.   EPIDEMIOLOGY   Stagonospora nodorum    epidemics occur frequently in affectedregions. Correct application of fungicide at a suitable time willeffectively control most outbreaks (Stover et al    ., 1996). Owingto the reliance on the accurate timing of fungicide application,disease forecasting is a critical part of successful control strategies.Artificial neural networks have been used to model stagonosporanodorum blotch (De Wolf and Francl, 2000), but the authors alsoconcluded that more work was required to produce a successfulforecasting system. Future studies should determine at whattime point modelling should begin and at what point fungicideapplication should be considered. Fig. 3 Estimated losses in wheat yield due to Stagonospora nodorum   in Southern Australia. Colours indicate areas with the given percentage losses, x = % loss (Brennan and Murray, 1998).    150   P. S. SOLOMON et al.   MOLECULAR PLANT PATHOLOGY   (2006) 7  (3), 147–156© 2006 BLACKWELL PUBLISHING LTD   Wind dispersal of sexual ascospores and rain splash dispersalof asexual pycnidiospores would suggest that a population of    S. nodorum    might consist of localized clones. However, in allpopulations tested, isolates of S. nodorum    were found to beextremely diverse. Keller et al    . (1997) reported that gene flowwas unrestricted between populations and was consistent withrandom mating events. Similarly, Caten and Newton (2000)analysed variation found within localized populations andfound a coarse patchwork of clones within a single field. Resultswere also reported from a Western Australian population,which also exhibited a high degree of genotypic diversity (Murphy   et al    ., 2000).The wide dispersal of ascospores via wind would suggest thatsexual crossing would be ecologically compulsory in naturalpopulations of the fungus. This would predict a more or less equalpreponderance of the two mating type idiotypes MAT1-1   and   MAT1-2.   Halama (2002) analysed distributions of mating typesusing tester strains in a small but world-wide collection, andreported a marked preponderance of MAT1-1   alleles in all thepopulations sampled. Interestingly, Bennett et al    . (2003) testedtwo New York populations using degenerate PCR primers andfound one contained roughly equal ratios of MAT1-1   and MAT1- 2    , whereas the other had increased frequency of MAT1-1   isolates.To help resolve this conundrum, Solomon et al    . (2004b) devel-oped S. nodorum-    specific PCR primers for rapid determination of mating type. To avoid re-sampling clones, isolates were derivedfrom ascospores collected in wind-traps. This sample was foundto contain a ratio of 14 MAT1-1   to nine MAT1-2    alleles, notsignificantly different from a 1 : 1 ratio. It was also concluded thatboth mating types had equal virulence; therefore, infection rateswould not skew the sampling method. Overall, it appears thatmating type ratios are not significantly skewed in the WesternAustralian or New York State populations and suggestions thatother populations may be skewed could be due to small samplesizes and the failure to identify members of clones.Epidemics can be initiated by either wind-borne ascospores orvia infected seed. Under controlled environmental conditions,Shah and Bergstrom (2000) showed frequent transmittance of    S. nodorum    from infected seed to the first leaf of the seedling.Transmission occurred most efficiently at 9 °   C where 37% of firstleaves were carriers of disease. At temperatures of 13–21 °   C,which are more representative of field conditions, first leaf trans-mission rates were around 10%. The authors concluded that theobserved rate of seed to leaf transmission could initiate significantdisease outbreaks early on in a crop’s growth without requiringwindborne ascospores. In a subsequent study, Shah et al    . (2001)showed that diverse S. nodorum    populations were present in NewYork wheat fields prior to tiller elongation or disease development.The results led to the conclusion that foliar epidemics of stagonosporanodorum blotch do not rely solely on immigrant S. nodorum    ascospores arriving after canopy closure. The study did notdifferentiate between initial inoculum derived from wind-borneascospores and that from infected seed. Ideally, future studieswould use controlled conditions to document disease progressionsolely from infected seed or solely from wind-borne ascosporesto help determine which situation prevails in field populations.The sowing of seed known to be free of S. nodorum    into a fieldsituation would rapidly reveal the extent to which disease wasdependent on seed inoculum or wind-borne ascospores.   CHEMICAL CONTROL   Fungicide treatment of temperate wheat is a critical part of cropmanagement. Initially, carboxin was used to treat wheat seedsagainst infection, but it has been replaced in the last 10–15 yearsby systemic triazole ergosterol biosynthesis inhibitors. Sundin   et al    . (1999) reported that difenoconazole and triadimenol weremore effective. Strobilurins are the latest family of fungicides tobe introduced. In addition to a direct antifungal effect, due toinhibition of mitochondrial respiration, strobilurins have beenfound to increase flag leaf lifespan significantly (Gooding et al    .,2000). As sporulation of the fungus on senescent tissue is acritical phase in the development of disease, this effect of strobilurins is likely to be very important. It should be noted, how-ever, that strobilurin resistance has become common in otherwheat pathogens and so its use will require careful management.   GENETIC CONTROL   In many parts of the world, genetic control of disease is the onlyavailable option. The degree of resistance varies in differentcultivars ranging from highly susceptible to moderate resistance.No complete resistance has been found in the existing wheatgene pool (Aguilar et al    ., 2005). Crosses of wheat lines haveshown that resistance is inherited as a quantitative trait. In arecent study of bread wheat, four quantitative trait loci (QTLs)on chromosomes 2B, 3B, 5B and 5D were found (Czembor et al    .,2003) whereas a study of Swiss cultivars found glume blotchresistance QTLs on 4BL and 3BS (Schnurbusch et al    ., 2003). Widecrosses involving Aestivum tauschii    have found that resistance iscontrolled by single genes, but it is not yet clear whether thesegenes can usefully be introgressed in bread wheat (Loughman   et al    ., 2001). Partial resistance has also been found in durum andtriticale (Cao et al    ., 2001; Oettler and Schmid, 2000). In thesecircumstances, the identification or cloning of resistance geneswould be highly problematic. These difficulties are compoundedby reports that resistance to leaf infection may be controlled bydifferent genes to those resisting glume blotch (Wicki et al    .,1999). Such observations need to be factored in to any screeningprogramme involving resistance testing. In a recent study it wasdemonstrated that the parental wheat lines in the InternationalTriticeae Mapping Initiative demonstrate differential sensitivity    Stagonospora nodorum 151 © 2006 BLACKWELL PUBLISHING LTD MOLECULAR PLANT PATHOLOGY   (2006) 7 (3), 147–156 to a toxic compound (SnTox1—believed to be proteinaceous)derived from S. nodorum   strain Sn 2000 (Liu et al  ., 2004a,b).In this work it was demonstrated that toxin insensitivity co-segregated with a major disease-resistance QTL on the short armof chromosome 1B and was a recessive characteristic. This resistancewas most obvious on younger plants, so is primarily involved inseedling resistance. These results show that this toxin is actingas a host-selective compound and insensitivity to the toxin maymake a significant contribution to disease resistance. Thesecharacteristics are very similar to the Pyrenophora tritici  – repentis  ToxA system, and we await reports of this toxin structure withinterest.Measuring partial resistance is a major problem in plantpathology. Such resistance is useful in the field and indeed manypathologists argue that it is more durable than complete resist-ance (Fraaije et al  ., 2001). However, difficulties in phenotypinghave made progress slow. In a commodity crop such as wheat,and in diseases not known to produce toxins that affect theconsumer, the ultimate test of resistance is the degree of yieldloss induced by the disease. This is a demanding measurement tomake in terms of both cost and time. The yield benefit conferredby fungicide can be measured, but this assumes that the onlypathogen inhibited by the fungicide is the one under study, in thiscase S. nodorum  . Such a precondition is rarely achieved evenif artificially inoculated leaves are used. Growth chambers canreduce the impact of extraneous diseases but are expensive foryield trials. Furthermore, fungicides such as the strobilurins havebeen shown to have growth-promotion effects independent of disease control. The use of the 1000 grain weight (RGW) doesreduce the scale and cost of the trial, but relies on the assumptionthat grain number does not alter, and in any case grain quality isan important economic characteristic of the crop.A new approach to measuring resistance is to measure fungalbiomass. This has the advantage of being objective. Fraaije et al  .(2001) designed multiplexed PCR tests for S. nodorum   and meas-ured fungal biomass in infected flag leaves. They demonstrated agood correlation with adult disease scores. It will be interestingto see if such tests can be applied to seedlings and whether fungalbiomass proves to correlate with yield loss. MOLECULAR TECHNOLOGY Transformation of S. nodorum   was achieved early in the develop-ment of the technique for filamentous fungi (Cooley et al  ., 1988).Several selectable markers (hygromycin, phleomycin, benomyl)have been used, co-transformation is frequent and heterologousexpression of genes is straightforward. The general frequency of transformation events is moderate but workable at about 100 per µ g of DNA or per 10 6  protoplasts.Gene targeting and homologous recombination was demon-strated by Howard et al  . (1999) and since then only a smallnumber of genes have been disrupted. The relative frequencyof homologous vs. ectopic insertion is shown in Table 1. Thefrequency of homologous recombination is workable provided1–2 kb of flanking DNA is in the construct.In most of these experiments, the disruption construct consistsof the selectable marker flanked by 0.3–5 kb of the gene to bedisrupted. The construction of such vectors can be inordinatelytime consuming. Solomon et al  . (2003) utilized an in vitro   trans-position system to insert the fungal selectable marker gene intoa Ptr2   cDNA clone. The system is fast and can be used with anyselectable marker for which donor clones are available. It hasbeen used several other times to generate disruption constructsfor genes encoding the G α  subunit of a heterotrimeric G-protein( Gna1 ), malate synthase ( Mls1 ), glyoxylase ( Gox1 ), mannitol1-phosphate dehydrogenase ( Mpd1 ) and a mitogen-activatedprotein kinase ( Mak2  ) (Solomon and Oliver, 2004; Solomon et al  .,2004a,c, 2005a,b).Howard et al  . (1999) detected gene silencing in S. nodorum  .Of the 14 non-nitrate-utilizing mutants created by attemptedgene replacement, six were found to contain ectopic insertionsof the NIA1  disruption vector. The endogenous NIA1  locus wasintact and was not influenced by DNA methylation effects. Theonly reason cited for loss of NIA1  expression was gene silencingcaused by the additional 1.1 kb of NIA1  sequence contained inthe disruption construct; however, molecular evidence for thiswas not presented.A number of cDNA and genomic libraries have been madeand these currently provide the resource for the construction of  Table 1 The relative frequency of homologous vs. ectopic insertion events for selected publications on inactivated genes in Stagonospora nodorum. GeneReference5 ′  flank (kb)3 ′  flank (kb)Total transformantsHomologous gene replacementsRate % Nia1 Howard et al  . (1999)14.826623 Odc1 Bailey et al  . (2000)1115032 Snp1 Bindschedler et al  . (2003)1+1+791013 Ptr2  Solomon et al  . (2003)0.51.2372465 Gox1 Solomon and Oliver (2004)0.30.84549 Mls1 Solomon et al  . (2004a)0.90.84824 Gna1 Solomon et al  . (2004c)0.60.55024 Mpd1 Solomon et al  . (2005a)0.30.86023 Mak2  Solomon et al  . (2005b)0.21.2101144