A comparative analysis of the heterotrimeric G-protein Gα, Gβ and Gγ subunits in the wheat pathogen Stagonospora nodorum

Background It has been well established that the Gα subunit of the heterotrimeric G-protein in the wheat pathogen Stagonospora nodorum is required for a variety of phenotypes including pathogenicity, melanisation and asexual differentiation. The roles though of the Gγ and Gβ subunits though were unclear. The objective of this study was to identify and understand the role of these subunits and assess their requirement for pathogenicity and development. Results G-protein Gγ and Gβ subunits, named Gga1 and Gba1 respectively, were identified in the Stagonospora nodorum genome by comparative analysis with known fungal orthologues. A reverse genetics technique was used to study the role of these and revealed that the mutant strains displayed altered in vitro growth including a differential response to a variety of exogenous carbon sources. Pathogenicity assays showed that Stagonospora nodorum strains lacking Gba1 were essentially non-pathogenic whilst Gga1-impaired strains displayed significantly slower growth in planta. Subsequent sporulation assays showed that like the previously described Gα subunit mutants, both Gba1 and Gga1 were required for asexual sporulation with neither mutant strain being able to differentiate either pycnidia nor pycnidiospores under normal growth conditions. Continued incubation at 4°C was found to complement the mutation in each of the G-protein subunits with nearly wild-type levels of pycnidia recovered. Conclusion This study provides further evidence on the significance of cAMP-dependent signal transduction for many aspects of fungal development and pathogenicity. The observation that cold temperatures can complement the G-protein sporulation defect now provides an ideal tool by which asexual differentiation can now be dissected.


Background
Stagonospora (Teleomorph: Phaeosphaeria) nodorum is a necrotrophic fungal pathogen and the causal agent of stagonospora nodorum blotch (SNB) of wheat [1]. Recent studies focused on understanding the molecular basis of the disease has identified the required role of secreted necrotrophic effectors during infection [2]. The interaction of these secreted effector proteins with a corresponding host dominant susceptibility gene results in rapid cell death and the facilitation of a rapid vegetative growth phase in planta.
Whilst the role of the effector proteins in causing disease is clear, it has also been demonstrated that the ability of the pathogen to undergo asexual sporulation is critical for disease progression throughout the growing season [1]. The asexual spores (pycnidiospores) of S. nodorum are formed in asexual structures known as pycnidia [3]. The pycnidiospores are released from the mature pycnidia on the leaf surface by rain splash dispersal leading to new infections on younger leaves. These multiple rounds of successive inoculation by the fungus, and in an inoculum density dependent manner escalates the damaging symptoms of SNB, spreading the disease to the head of the plant.
Recognition of the host by the fungus, followed by its capacity to penetrate the leaf, proliferate and reproduce is likely to require a perception of a range of signals from the host and environment, ultimately influencing disease severity. As such, heterotrimeric G-protein signalling has been the subject of intense research in filamentous fungi and many other biological systems [4].
The Neurospora crassa Gna1 and Gna2 genes were the first reported genes of a G-protein subunit to be cloned in a filamentous fungus [5]. In filamentous fungi, the resulting phenotypic effects of loss and gain of function mutations of the genes encoding the Gα, Gβ and Gγ proteins comprising the heterotrimer, have identified a number of cellular processes under the regulation of the G-protein.
Among others, a commonly described attribute of fungal G-protein-compromised mutants is an effect on sporulation with reports of hyper-sporulation [6], reduced sporulation [7,8] or a complete lack of sporulation [9] across genera.
Reverse genetics studies in S. nodorum have demonstrated the requirement of several signalling pathways for disease development. Strains of S. nodorum lacking the Gα subunit Gna1 [9], the mitogen-activated protein kinase Mak2 [10], a Ca 2+ /calmodulum-dependent protein kinase CpkA [11], or the short-chain dehydrogenase Sch1 [12] all demonstrate a variety of developmental defects including being either severely compromised in sporulation or are unable to do so.
Here, we report the comparison of three mutant strains of S. nodorum with the wild-type strain SN15. All three mutants were compromised in G-protein signalling, with each lacking one of the subunits of the heterotrimer. The Gba1 (Gβ) and Gga1 (Gγ)-lacking strains of S. nodorum, given the strain names gba1-6 and gga1-25, respectively, were created by homologous recombination of the Gba1 and Gga1 genes with a selectable marker. The phenotypic characteristics were then assessed alongside those of the previously described S. nodorum gna1-35 (Gna1 mutant) strain. Consistent with gna1-35, the gba1-6 and gga1-25 strains were less pathogenic on wheat and unable to sporulate asexually. Interestingly, it was found that prolonged incubation of mature plate cultures of gna1-35, gba1-6 and gga1-25 at 4°C would complement the sporulation defect; developing pycnidia and restoring asexual sporulation in these strains. These strains are now helping aid in dissecting the molecular mechanisms underlying the phenotypic defects with the aim of bringing to light better mechanisms of controlling S. nodorum and other fungal pathogens.

Results
Identification and disruption of Gga1 and Gba1 in S. nodorum The genes encoding putative Gγ and Gβ subunits were identified in the S. nodorum genome sequence by blast analysis using related fungal homologues. Using this approach, the genes SNOG_16044 and SNOG_00288 were identified as encoding putative Gγ and Gβ subunits and named Gba1 and Gga1 respectively. As anticipated, BlastP of both Gba1 and Gga1 revealed multiple near identical proteins in closely related fungi. A clustal analysis of these related sequences is shown in Additional file 1: Figure S1.
To investigate the role of the genes in growth and pathogenicity of S. nodorum, Gga1 and Gba1 were disrupted via homologous recombination as described above. The fungal colonies resulting from both transformations were screened by PCR to confirm homologous recombination (Additional file 1: Figure S2). A number of successful mutations were confirmed for both the Gga1 and Gba1 gene disruptions. The putative mutants were selected for copy number determination as described above. All transformants demonstrated by PCR to have undergone homologous recombination had a calculated ratio of the phleomycin resistance gene to single-copy actin gene of between 0.9 and 1.1 indicating that only one copy of the transformation cassette had integrated into the genome. Representative strains for each mutation were chosen for further analysis.
All three G-protein subunits are required for normal growth The phenotypic characteristics of the S. nodorum gga1 and gba1 strains were assessed in vitro. The previously published Gα mutant, gna1-35, was also included for a comprehensive analysis for the each of the three G-protein subunits. The mutant strains gba1-6 and gga1-25 showed a number of phenotypic effects consistent with those described for gna1 by [9]. All three strains were non-sporulating under the standard in vitro culture conditions used to promote asexual sporulation in wild-type SN15. On V8PDA medium, each strain displayed pale pink mycelia, often developing a green colouration towards the centre of the culture. As the strains matured, the mycelia lost the pink and green colouration, becoming white, to display an albino phenotype. On minimal medium containing 25 mM glucose as the sole carbon source, gga1-25 displayed a similar pink colouration, however gna1-35 and gba1-6 both grew albino ( Figure 1).
All mutant strains were found to have reduced radial growth by comparison to wild type, regardless of the carbon source (Figures 1 and 2, Table 1). Differences in the radial growth rate between the mutant strains however were found to be dependent on the available carbon source. S. nodorum gba1-6 showed significantly (p < 0.05) higher radial growth than the other two mutants when provided with arabinose, glucose or sucrose. When provided with fructose however, gba1-6 growth was significantly reduced compared to that on glucose or sucrose. Gna1-35 growth significantly increased compared to most other carbon sources tested, such that when grown on fructose, there was no significant difference in radial growth between gna1-35 and gba1-6. When gba1-6 was provided with arabinose, although growth was equivalent to that measured on fructose, it still retained a higher radial growth than gna1-35 as it does not have the measured increase in growth rate in response to arabinose as it does with fructose. It is evident from this data that fructose resulted in the greatest radial growth for S. nodorum gna1-35, whereas glucose and sucrose resulted in the greatest radial growth for S. nodorum gba1-6. S. nodorum gga1-25 showed significantly less radial growth than all other strains on most carbon sources. On glucose gga1-25 has a radial growth equivalent to that of gna1-35, and on  trehalose the growth was equivalent to both gna1-35 and gba1-6. When casamino acids were added along with glucose, gga1-25 achieved its highest recorded radial growth, which was equivalent to that of gna1-35 and gba1-6 on the same medium ( Figure 2; Table 1).
It has been previously observed that wildtype S. nodorum secretes a light brown coloured pigment when grown on minimal medium (25 mM glucose) under white light ( Figure 3) [9]. By comparison gna1-35 secretes a much darker coloured pigment, with little difference between light and dark grown cultures. Under these conditions, cultures grown in the light showed a pronounced medium discolouration, whilst gba1-6 and gga1-25 both routinely showed pigment secretion when grown in the light and dark. Pigment secretion, as observed by the intensity of the discolouration of the growth medium, was also dependant on the carbon source (Table 1). Whilst gna1-35 had the most pronounced pigment secretion of the strains when grown on glucose, this strain did not show any notable discolouration on any of the other carbon sources tested, nor did gga1-25. On glucose, S. nodorum gba1-6 secreted pigments when the medium was supplemented with lactose, mannitol or arabinose. However, when provided with fructose, sucrose or trehalose, no pigment secretion was noted for this or any other strain.
Gna1, Gba1 and Gga1 are all required for different aspects of pathogenicity on wheat Detached leaf assays (DLAs) were used to compare the differences in pathogenicity of S. nodorum strains on wheat. Figure 4 shows the slowed progression of lesion formation by the mutant strains on wheat compared to the wildtype. After 5 dpi, SN15 causes necrotic flecking of the leaf, whilst the mutant strain gna1-35 produced a chlorotic lesion. The gba1-6 and gga1-25 strain only showed very mild chlorosis on most leaf replicates at the same time after inoculation. The same leaves at 13 dpi infected with gna1-35 or gga1-25 exhibit disease symptoms comparable to those produced by SN15. However, given this extended timeframe disease symptoms of leaves challenged with gba1-6 at this latter stage have not progressed  Wild-type SN15 and the mutant strains gna1-35, gba1-6 and ggaA-25 display carbon source-dependent changes in radial growth rate and pigment secretion as observed 10 days post inoculation (dpi). 1 The letters in brackets shown after the diameter measurements define the significance of the results with growth assays showing the same letter being not significantly different (p < 0.05).
beyond a very mild chlorotic response. Sporulation was not evident for any of the mutants in planta.

Prolonged cold exposure induces pycnidia differentiation
Whilst pycnidial development and the accompanying asexual sporulation of S. nodorum SN15 occurs readily on agar plate media, under the same conditions, the mutant strains gna1-35, gba1-6 and gga1-25 as described above are completely absent of pycnidia formation. It was observed however that the incubation of the strains at 4°C from 8 dpi resulted in the appearance of small dark dots that resembled the initiation of asexual development. A continuation of the incubation of these cultures at the colder temperatures revealed that these conditions appeared to promote the pycnidial development. Toluidine blue stained sections of these spots identified the regions as intertwining mycelia ( Figure 5). Continued incubation of G-protein mutants at the lower temperature allowed the intertwining to progress to the formation of a mycelial knot. Mycelial knot formation is the earliest stage of pycnidia formation, preceding differentiation of the mycelial cells [3]. Subsequent observation of the mycelial knot showed differentiation of the mycelia into pycnidia within four to six weeks at 4°C. This is a significant result as asexual development had not yet been observed in a S. nodorum G-protein signalling mutant. Similar results were also observed for the gbaA and gna1 strains (data not shown).
Pycnidia (accompanied by asexual spore development) in S. nodorum wild-type SN15 developed in a distinct circadian ring pattern within 5 days from inoculation (dpi) of solid minimal medium ( Figure 6). The formation of pycnidia (containing viable spores) in S. nodorum mutant strains gna1-35, gba1-6 and gga1-25 by comparison was evident mainly amongst the outer perimeter of the mycelia after prolonged growth at 4°C. The pycnidia of gna1-35 were heavily pigmented, black in appearance, (Figure 6 & 7) and randomly dispersed amongst the colony's mycelial perimeter. By comparison, gba1-6 which developed lighter, brown-coloured pycnidia, tending to form along the mycelium as it intertwined at the perimeter of the colony. The pycnidia of gga1-25 were comparatively lighter in colour than SN15, gna1-35 or gba1-6, with a light brown-colouration, and although they often developed along the intertwining mycelium like gba1-6, they appeared less confined to this location of development. The pink cirrhus that exudes from pycnidia of S. nodorum SN15 was not evident for any of the mutant pycnidia, and perhaps consequently, spores could only be released by manual disruption. It is significant to note that though that the pycnidiospores released by the mutant were viable (Additional file 1: Figure S3).
The mature pycnidia formed under the cold-inducing conditions were visualized by light microscopy. Toluidine blue-stained longitudinal sections of S. nodorum SN15 pycnidia identified the presence of a number of enlarged cells that form the opening of the ostiole from which the cirrhus of spores are released from the pycnidial cavity ( Figure 8). Analysis of the pycnidia formed by each of the G-protein mutants by cold induction revealed the structures to be comparable to those produced by the wildtype strain under the same conditions. It was observed that the ostiole failed to differentiate on the mutant pycnidia ( Figure 9). This observation was consistent with the requirement that the pycnidiospores within these mutant pycnidia could not be released by water (as typically observed in wildtype pycnidia) and required manual disruption. The pycnidia of gba1-6 and gga1-25 were also nearly always observed as multiple structures fused together and were almost never seen individually (Figure 9). Although the pycnidia of SN15 and gna1-35 often developed fused, it was uncommon for the pycnidia to form (See figure on previous page.) Figure 3 A visual comparison of pigments secreted into the medium by strains of S. nodorum when grown in the dark, compared to those grown under a 12 hour white light cycle. Discolouration of the medium is dramatically intensified in cultures of S. nodorum wild-type SN15 when exposed to light; less so for mutant strains gba1-6 and gga1-25; with little change between the light and dark cultured gna1-35 mutant. Agar cultures are pictured from beneath the petri-dish. indistinct from one another. The pycnidia of gga1-25 and gba1-6 were also comparatively misshapen and less mature in appearance than those of SN15 and gna1-35.

Discussion
The deactivation of the Gα subunit Gna1 from S. nodorum has proven fruitful to further understanding Longitudinal sections of a wax embedded excision of a S. nodorum gga1-25 culture -stained with toluidine blue, is pictured. Slow differentiation of mycelia into pycnidia allowed all stages of development to be captured in an excision from a single culture. Pynidia formation begins with the intertwining of mycelia to form a mycelial knot (A), which is followed by differentiation and enlargement of the cells (Ec), forming a primordium (B through F), which matures into the pycnidium (G), eventually producing pycnidiospores from the conidiogenous cells (Cv) within the pycnidial cavity.
the pathogenesis of this fungal pathogen [9]. The lack of sporulation and reduced pathogenicity of the resulting gna1 strain sparked further investigation into the molecular and phenotypic attributes of this mutant strain largely because a determination of the molecular processes underpinning the phenotype could lead to more targeted control of the pathogen. Subsequent analysis of the gna1 strain to identify downstream-regulated targets and processes has uncovered many interesting aspects of the disease including mycotoxin production.
Based on the success of dissecting the gna1 phenotype, we have extended this study to include the analysis of the Gβ and Gγ subunits of the heterotrimeric G-protein through a reverse genetics approach. Strains of S. nodorum lacking these genes displayed variety of independent phenotypes during growth in vitro. One of the most apparent phenotypic defects under normal growth conditions was the complete lack of pycnidia formation or accompanying asexual sporulation. This phenotype is shared by other S. nodorum strains possessing defects in signalling pathways, and as such, was consistent with earlier findings in S. nodorum [9,11,13]. Along with growth defects in vitro, the mutant strains also exhibited different abilities to cause disease. Lesion formation on leaves inoculated with strains lacking with Gna1 or Gga1 was delayed but appeared comparable to that of the wild type after two weeks post inoculation. Leaves inoculated with Gba1 though failed to elicit any response from the leaves after 5 dpi, and only a very mild chlorotic response was evident after two weeks. This implies that Gba1 has a critical role in disease development in S. nodorum. Given the almost complete lack of symptom development, it could be suggested that Gba1, like StuA [14], has a role in effector regulation. However this is only speculation and requires further analysis.
Nutrient sensing in the S. nodorum gna1, gba1 and gga1 strains Dramatic growth differences between the mutant strains and the wild-type SN15 were noted on agar plate medium. On V8PDA, SN15 grows radially symmetrical with pycnidia forming in distinct circadian bands [15]. The gna1 and gba1 mutant strains both show a similar banding pattern, in mycelial growth, indicating that these strains have not lost the capacity to perceive a light signal.
The radial growth of all three mutant strains 10 dpi was reduced by comparison to SN15 on all tested media. The variation in radial growth of the mutant strains when growing on different carbon sources confirmed that the S. nodorum G-protein(s) play(s) a role in carbon source utilization. In comparison to the wild-type SN15, which displayed a statistically similar radial growth rate when provided with arabinose, fructose, glucose, sucrose or trehalose as a sole carbon source.
The comparatively slower growth of gna1 on sucrose was interesting when considering this strain's slower growth on glucose, but significantly higher growth on fructose. Kraakman et al., (1999) showed that the GPCR Gpr1 binds extracellular glucose in the yeast Saccharomyces cerevisiae and stimulates cAMP synthesis through the Gα subunit Gpa2. Likewise Lemaire et al., (2004) showed both glucose and sucrose induced cAMP signalling through the receptor Gpr1, however it was not fructose-induced. Although deletion of either Gpr1 or Gpa2 did not result in a reduced growth rate in S. cerevisiae, the strains in the study were not limited to a single carbon source [16]. However, an altered capacity to sense and respond to the presence of these sugars will inevitably affect growth. The reduced radial growth rate of the S. nodorum gna1 mutant when solely provided with glucose or sucrose compared to fructose therefore could be due to a reduced capacity for sensing glucose and sucrose and imply similar functions for the S. nodorum Gna1 and yeast Gpa2. It has also been shown that in binding glucose, the GPCR Gpr1 will fail to cause the normal rapid activation of adenylate cyclase if the glucose is not internalized and phosphorylated [16], which may further explain slower growth in response to glucose in strains where the deactivated subunit causes a lesser response to glucose.
Irrespective of the speculated extracellular sensing roles of these G-protein subunits, the difference in growth rates across S. nodorum gna1, gba1 and gga1 strains when provided with these carbon sources can be explained by processes biochemically downstream. Alterations in catabolic processes may have arisen as a result of the mutations. The growth rates of gna1 on each of fructose and glucose, compared to sucrose, for example is consistent with processes downstream of sucrose (α-D-fructofuranosyl α-Dglucopyranoside) hydrolysis, which yields one unit of fructose and one of glucose. Given that gna1 grows faster on fructose, it suggests that glucose may be feeding less efficiently into glycolysis in this strain.
Interestingly the seemingly inherently slower growth rate of S. nodorum gga1 under most conditions is comparable with each of the mutant strains when provided with (See figure on previous page.) Figure 6 Pycnidia development (accompanied by asexual sporulation) in the S. nodorum wild-type strain SN15 is observed in a distinct circadian ring pattern (A and B) within 5 days post inoculation (dpi) of solid minimal medium*. Pycnidia do not develop in the mutant strains during this timeframe. The formation of pycnidia in the S. nodorum mutant strains gna1-35, gba1-6 and gga1-25 is evident amongst the outer mycelia (C -E) from between 3 and 6 weeks incubation of (the initially) non-sporulating (5 dpi) culture at 4°C. S. nodorum strains are pictured growing on nitrocellulose membranes (30 mm diameter)-overlaying minimal medium agar.
trehalose as a sole carbon source. The radial growth rates on trehalose could also implicate all three subunits in processes downstream of extracellular sensing. The hydrolysis of trehalose (α-D-glucopyranosyl-α-D-glucopyranoside) yields two glucose units, yet the growth of gba1 is particularly slower on trehalose than glucose, which may suggest rather than a glycolytic inefficiency as mentioned above, a reduced capacity to hydrolyse trehalose, or even a diminished capacity to sense the signals that would otherwise cue the cell to catabolise trehalose. Changes to trehalose metabolism have been shown to have dramatic effects on sugar metabolism in general, and shown to have severe implications for phytopathogenicity [17,18], so the reduced capacity to use trehalose as a sole carbon source has likely had direct implications on fungal fitness.
Metabolite secretion by the S. nodorum gna1, gba1 and gga1 strains S. nodorum gna1 has been shown to secrete brown pigments comprised of tyrosine, phenylalanine and L-DOPA into the growth medium, first observed in the discolouration of the growth medium [9]. Discolouration at the growth medium is also an attribute of S. nodorum SN15 Figure 7 The observed pigmentation and size of the mutant pycnidium differs significantly between strains. Pictured is a single gna1-35 pycnidium, and pycnidia of the gba1-6 and gga1-25 strains of S. nodorum, amongst the mycelia. Images captured at 40× magnification. Figure 8 A longitudinal section of a wax embedded excision from an asexually sporulating culture of S. nodorum SN15 -stained with toluidine blue. Pictured are pycnidiospores being released from a mature pycnidium. Arrows point to the masses of enlarged cells producing the ostiole, formed in the development of the mature pycnidium, from which the pycnidiospores are released from the pycnidial cavity as a cirrus. Cv, pycnidial cavity; W, pycnidial wall; Ch, cirrus. and the gba1 and gga1 mutant strains. The carbon source dependency and intensity of discolouration of the medium also imply implications at least for primary metabolism, in the mutant strains. The carbon-source dependency of media discolouration identifies differences in metabolite secretion resulting from likely metabolic changes. These could potentially result from the inefficient use of metabolites or products of metabolism due to blockages or even over-active biochemical pathways. Together with the reduced growth rates on different media, the Gna1, Gba1 and Gga1 mutations appear to have introduced metabolic inefficiencies.
In the later observed cultures of S. nodorum gna1, gba1 and gga1, where pycnidia formation was studied, more intense secretions could be seen. It's likely that the intensity of media discolouration was heightened by accumulation over the extended culture period however it may also be that the secretions changed as the cultures' phenotypes changed. It's also possible that the increased concentration of secreted metabolites in the culture medium played a role in triggering the formation of pycnidia in these strains. Either way, the increased presence of secreted metabolites in these strains whilst undergoing pycnidial differentiation adds further interest to the identity of these secreted metabolites.
Pathogenicity and asexual sporulation of the S. nodorum gna1, gba1 and gga1 strains The capacity to rapidly increase fungal inoculum density by releasing spores from pycnidia following infection of the wheat plant by S. nodorum is fundamental to the success and consequently the impact of SNB. S. nodorum gna1, gba1 and gga1 were all unable to sporulate during infection of the wheat leaf, however although this defect may slow disease amplification, sporulation is clearly not a prerequisite for leaf necrosis. The inability for disease caused by infection with the gba1 strain to progress beyond chlorosis however, may implicate necrotrophic effector production in S. nodorum as positively regulated by G-protein signalling through the Gβ subunit Gba1 [14]. It is interesting to note that the requirement of the Gβ and Gγ subunits for infection in different fungal plant pathogens varies. For example, it has been previously demonstrated that GBB1 in Gibberella monoliformis is not required for pathogenicity whist the orthologous protein in the related Fusarium oxysporum is [19,20]. Our data clearly show that gene encoding for the Gβ subunit, Gba1, is required for S. nodorum to cause disease on wheat.
Whilst sporulation was not observed for the gna1, gba1 or gga1 strains in planta, the observations of asexual sporulation described in vitro are of considerable interest. The capacity for the gna1, gba1 and gga1 strains to develop pycnidia during prolonged incubation at 4°C from an already matured, yet non-sporulating culture adds further interest and potential for using these strains to dissect these fundamental processes in S. nodorum.
The physical characteristics of the mutant pycnidia observed in vitro were also of interest. In S. nodorum SN15, differentiation of cells forming the ostiole of the mature pycnidial wall was observed, but was not seen for the mutant pycnidia. It is likely that spores could not be released by the fungal strains because of this defect. Cirrus containing the spores was also observed in SN15, but not in the mutant pycnidia. Without the cirrus, it is unlikely there would be enough turgor pressure to release the spores, even with the formation of a wild-type ostiole, and it may be that this pressure plays a role in the formation of the ostiole in the S. nodorum pycnidium.
The pycnidia of the strains gga1 and gba1 are comparatively misshapen and less mature in appearance than those of SN15 and gna1. However, because these strains do develop viable spores, they may not actually be less mature, but perhaps this manifestation is a consequence of these two strains lacking the capacity to develop such a well-defined pycnidial wall.
In conclusion, this study has demonstrated the critical, and yet independent, roles of the heterotrimeric Gprotein subunits in S. nodorum. Each of these subunits was found to play a role in in vitro and in planta growth, albeit with varied roles. As had been previously observed for the gna1 strain, gba1 and gga1 strains were unable to sporulate when grown under normal growth conditions. However, prolonged incubation of these strains at 4°C appeared to complement the sporulation defect and pycnidia, containing viable pycnidiospores, were differentiated in each of the mutants. The mechanism of how colder temperatures induce sporulation in these mutants is clearly of interest and is the focus of ongoing studies. It should be noted that whilst single event homologous recombination events were demonstrated for each of the mutants generated in this study, future studies will attempt to complement these strains to provide unequivocal proof of the role of these in the above described phenotypes.
The capacity for the S. nodorum strains to grow on different carbon sources was assayed by inoculation of MM agar plates containing different carbon sources by supplementing a stock of MM without the carbon source with 25 mM of the appropriate carbon source. S. nodorum strains were inoculated onto the above media from minimal medium (25 mM glucose) by excising a region of the agar containing approximately 4 mm 2 of the agar surface of the (non-sporulating) growing edge of the mycelia onto the plates. Cultures were grown for 10 days in the dark at 22°C and colony diameters recorded 3, 5 and 10 days from inoculation, and observations of phenotype made. Four replicates were prepared per strain per carbon source assay. All statistical analyses were undertaken using the JMP7 package (SAS Institute). Statistical significance was determined using the Tukey-Kramer analysis.

Plant growth conditions
Plant material and infection conditions Pots (10 cm diam.) containing Perlite (P500) and grade 2 Vermiculite (The Perlite and Vermiculite Factory, WA, Australia) were seeded with five seeds of the wheat variety Amery and grown at 20_ C in a 12 hr day/12 hr night cycle. The pathogenicity of the mutants was assayed on detached leaves from 2-week-old wheat seedlings, using a method modified from that described by Benedikz et al. [9,21]. The distal end (2 cm) of the detached wheat leaves was removed. The next portion (4-5 cm) was embedded into benzimidazole agar, adaxial side up. The leaves were inoculated with small blocks of mycelium (approximately 45 mm 3 ) and incubated in a 12 h light/12 h dark cycle at 22°C to enable disease development.
Additional file 2: Table S1. Sequences of primers used in this study.
Author's contributions JPG carried out most of the experiments and participated in the drafting of the manuscript. RPO and RDG participated in the design of the study and the interpretation of the data. PSS conceived the study, participated in the experiments and wrote the manuscript. All authors read and approved the final manuscript.