Identification and characterization of the Sfp-type PPTase gene Ffppt1 reveals involvement in lysine biosynthesis

In order to identify the NpgA/CfwA-encoding homolog in the F. fujikuroi genome, a BlastP analysis was performed using the A. fumigatus PptA sequence. One protein sequence with 32% identity to PptA (e value = 2.6 e⁻²⁸) was found and designated FfPpt1 (GenBank accession number HE614113). RT-PCR revealed an open reading frame of 876 bp spanning one intron of 50 bp. Targeted gene replacement using a nourseothricin resistance cassette yielded three transformants (designated ΔFfppt1 T8, T13, and T14). The loss of the Ffppt1 ORF was verified by diagnostic PCR (Fig. S1), and additional integrations of the resistance cassette into the genome were excluded by Southern blot (Fig. S1). In contrast to the wild type, the mutants were unable to grow on minimal medium without addition of lysine (Fig. 2A). Since all transformants exhibited the same phenotype, ΔFfppt1 T8 was arbitrarily chosen for further experiments. This mutant was complemented by re-integration of the Ffppt1 wild-type gene copy into the genome. Minimal medium without lysine was used for selection, and PCR verified re-integration of Ffppt1 for the transformants designated ΔFfppt1^C (Fig. S1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Influence of FfPpt1, FfSre1 and FfFtr3/FfFet3 on lysine biosynthesis, iron homeostasis and oxidative stress.

A: Growth ability of the indicated strains on solidified Czapek Dox (CD) medium supplemented as indicated. Representative pictures were taken after 3 days of incubation at 28°C in the dark. B: Phylograms of ferroxidases and iron permeases from F. fujikuroi (Ff), characterized proteins from F. graminearum (Fg) and A. fumigatus (Af), as well as homologous sequences from F. oxysporum (FOXG) and F. verticillioides (FVEG) obtained from the Broad Institute database were created as described in Methods. Scale bars represent character changes. C and D: Northern blot analysis using indicated genes as probes and rRNA visualization as loading control. The indicated strains were grown as described in Methods. (−); addition of water, (+) addition of FeCl₃ to a final concentration of 1 mm.

https://doi.org/10.1371/journal.pone.0037519.g002

To determine which of the pleiotropic defects of the ΔFfppt1 mutant are due to the non-functional α-aminoadipate reductase resulting in lysine auxotrophy, we performed a targeted deletion of the α-aminoadipate reductase gene Ffaar1 (GenBank accession number HE614114). The deduced protein showed 49% identity to Aar1 from C. graminicola by BlastP analysis (e value = 0). Three out of 20 obtained transformants (ΔFfaar1 T13, T14 and T20) proved loss of the target ORF by diagnostic PCR (Fig. S2A). The ΔFfaar1 T13 mutant was used as a control for all further analyses.

Growth tests with the ΔFfppt1, ΔFfppt1^C, and ΔFfaar1 strains and the wild type were carried out on minimal medium containing either no lysine or 10 mm lysine. The ΔFfppt1^C strain showed wild-type-like growth on all media, whereas the ΔFfaar1 and ΔFfppt1 strains did not grow on medium without lysine (Fig. 2A).

FfPpt1 contributes to a functional iron uptake system that is controlled by the GATA-type transcription factor FfSre1

In addition to the growth defect on lysine-deficient medium, growth of the ΔFfppt1 strain was severely restricted on medium containing lysine in the presence of the iron chelator BPS and H₂O₂, respectively (Fig. 2A). This restriction was overcome when 2 µm of the siderophore ferrichrome (FEC) was added to the medium containing BPS/lysine and H₂O₂/lysine, respectively (Fig. 2A). These data suggest that siderophore-assisted iron uptake allows the F. fujikuroi wild type to grow in the presence of BPS, and that loss of siderophore production affects sensitivity against H₂O₂ in the ppt1 mutant. The ability of ΔFfppt1 to grow on medium containing lysine and only trace amounts of iron (CD minimal medium contains 3.6 µm FeSO₄) without the addition of siderophores suggested that a reductive iron uptake system is present in F. fujikuroi. BlastP analysis using the sequence of the plasma-membrane-localized Fet1 of F. graminearum against the genome database of F. fujikuroi revealed three proteins designated FfFet1 (89% identity, e-value = 0), FfFet2 (66% identity, e-value = 0), and FfFet3 (56% identity, e-value = 0). All three F. fujikuroi ferroxidase encoding genes revealed the presence of an adjacent iron permease encoding gene (designated Ffftr1, Ffftr2, and Ffftr3, respectively) each sharing its promoter region with the adjacent Fet-encoding gene. Searching the available Fusarium genome sequences revealed that F. verticillioides and F. oxysporum each possess three of these pairs similar to F. fujikuroi. Phylogenetic analysis showed that the A. fumigatus FtrA/FetC cluster groups together with FfFtr3/FfFet3, but no respective homologs seem to be present in the F. graminearum genome (Fig. 2B). To test whether the respective genes are expressed in an iron-dependent manner, we grew the wild type and the Ffppt1 mutant under iron deficient conditions before addition of ferric chloride (FeCl₃) or water. Northern analyses revealed that all of the genes coding for putative F. fujikuroi ferroxidases and iron permeases are induced by iron starvation in the wild type and the Ffppt1 mutant (Fig. 2C and D). Furthermore, the expression of the F. fujikuroi gene Ffnps6 coding for the homolog of the F. graminearum NRPS Nps6 responsible for extracellular siderophore production [50] revealed an identical expression pattern (Fig. 2C and D). Expression signals of the SreA-encoding homologous gene Ffsre1 in the wild type and the Ffppt1 mutant were found under iron excess only (Fig. 2C).

Since in F. graminearum the proteins Ftr1/Fet1 were shown to constitute a bona fide plasma membrane iron uptake system [51], [52] and Ftr2/Fet2 were suggested to function as a vacuolar iron transport system due to their localization [51], we wanted to investigate the role of newly identified FfFtr3/FfFet3 with respect to iron acquisition in the wild type and the siderophore-deficient Ffppt1 mutant. Targeted gene replacement of FfFtr3/FfFet3 was performed in both the wild-type strain and the ΔFfppt1 strain. Diagnostic PCR revealed three of eleven transformants to have lost the ORFs of both genes, respectively (Fig. S2B). These transformants were designated ΔΔFfftr3/fet3 (T3, T6 and T7) and ΔΔΔFfppt1/ftr3/fet3 (T1, T2 and T3) of which ΔΔFfftr3/fet3 T3 and ΔΔΔFfppt1/ftr3/fet3 T1 were arbitrarily chosen for further investigation. To learn more about the regulation of the genes putatively involved in iron acquisition in F. fujikuroi we also deleted the gene Ffsre1, encoding the homolog of the the GATA-type transcription factor Sre1 from A. fumigatus. This transcription factor was shown to be involved in regulation of iron homeostasis [44]. Of the eleven transformants obtained, seven were proven to have lost the ORF by diagnostic PCR (Fig. S2C) and designated ΔFfsre1 (T1, T2, T3, T4, T6, T10, T11) of which T1 was arbitrarily chosen for further experiments.

As expected, the Ffppt1/ftr3/fet3 triple mutant was unable to grow without the addition of lysine and showed restricted growth in the presence of H₂O₂ (Fig. 2A). Interestingly, the Ffftr3/fet3 double and the Ffppt1/ftr3/fet3 triple mutant exhibited a slightly less restricted growth in the presence of H₂O₂ compared to the wild type and the Ffppt1 single mutant, respectively (Fig. 2A). When additional 2 µm of the siderophore FEC was present, the growth defect of the Ffppt1 single mutant could partially be overcome whereas the Ffftr3/fet3 double and the Ffppt1/ftr3/fet3 triple mutant were restored to wild-type-like growth (Fig. 2A). In contrast, in the presence of BPS the Ffppt1/ftr3/fet3 triple mutant showed a more severe growth defect compared to the Ffppt1 single mutant (Fig. 2A). The overall picture of growth ability of the mutants on the different media indicates that FfPpt1, FfFtr3 and FfFet3 participate in iron acquisition and mediate H₂O₂ tolerance in F. fujikuroi. The Ffsre1 deletion mutant showed restricted growth compared to the wild type on all media tested (Fig. 2A) and no growth when 1 mm FeCl₃ was present (Fig. S3). To learn more about the role FfSre1 plays in regulation of genes involved in iron metabolism and to investigate whether transcriptional deregulation can be observed when the genes encoding FfFtr3/FfFet3 are missing, we performed northern blot analyses of the wild type and the ΔΔΔFfppt1/ftr3/fet3, ΔFfppt1, ΔFfsre1 and ΔΔFfftr3/fet3 strains. Similarly to the observations from the previous iron-shift experiment, signals of Ffnps6, Ffftr1, Fffet1, Ffftr2 and Fffet2, were only visible under nitrogen starvation conditions in the Ffppt1/ftr3/fet3 and Ffftr3/Fffet3 mutants and the parental strains (Fig. 2D). The fact that signals for Ffnps6, Ffftr1, Fffet1, Ffftr2 Fffet2, Fffet3 and Ffftr3 were detectable in the Ffsre1 mutant even when 1 mm FeCl₃ was present (Fig. 2D) indicates that FfSre1 acts as a repressor of genes involved in iron metabolism in F. fujikuroi.

FfPpt1 is involved in conidiogenesis and sexual development

A. nidulans npgA/cfwA mutants showed delayed and reduced spore formation with an altered morphology at a range of 32°C to 37°C, but could be restored when contiguously grown to the wild type [34]. Since deletion mutants of the respective homologs in several filamentous fungi were also affected in conidiogenesis or conidia morphology, we investigated the effect of the Ffppt1 deletion in F. fujikuroi. As to our knowledge none of the previous studies on PPTase mutants in any filamentous fungus compared sporulation ability to that of an α-aminoadipate reductase mutant, we included the Ffaar1 mutant in our analysis. Similarly to the observations in other fungi, sporulation was severely reduced, but morphologically unaltered, in the Ffppt1 mutant compared to the wild type, the ΔFfppt1^C and ΔFfaar1 strains (Fig. 3A and B). However, when the Ffppt1 mutant was grown contiguously to the wild type separated by water-permeable cellophane membrane sporulation was partially restored (Fig. 3A). Interestingly, this partial restoration was not observed when the individually cultivated Ffppt1 mutant was supplemented with FEC indicating that iron limitation is not responsible for the sporulation defect (Fig. 3A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Influence of FfPpt1 on conidiogenesis and sexual mating recognition.

A: Spores produced of indicated strains per cm² after 10 days of incubation on solidified V8 medium in constant light conditions. Blue: strain IMI58289; red: strain C-1995. FEC: medium was supplemented with 2 µm ferrichrome; contiguous incubation: strains were incubated as described in Methods. B: DIC images of spores produced by the indicated strain. Strains were incubated and spores were collected as described in Methods. C: Representative photographs of sexual crossings of indicated strains as described in Methods.

https://doi.org/10.1371/journal.pone.0037519.g003

To study the influence of FfPpt1 on sexual development, a gene replacement mutant of the Ffppt1 locus was generated in the F. fujikuroi strain C-1995 that carries the MAT1-2 idiomorph using the same targeted deletion strategy. Similarly to the Ffppt1 mutant in the wild-type strain IMI58289, the deletion mutant in C-1995 was lysine auxotrophic (Fig. S4) and exhibited drastically reduced sporulation ability (Fig. 3A). When performing sexual crosses, the wild-type strain IMI58289 and all mutant strains generated in this background (ΔFfppt1, ΔFfppt1^C and ΔFfaar1) carrying the MAT1-1 idiomorph were able to interact with the wild-type strain C-1995 resulting in the formation of dark purple perithecia (Fig. 3C and S5). However, when the Ffppt1 locus was missing in strain C-1995 carrying the MAT1-2 idiomorph, no recognition took place when contiguously grown with any of the strains of the opposite mating type (wild-type IMI58289, ΔFfppt1, ΔFfppt1^C and ΔFfaar1). Subsequently, colonies from both mating partners did not come in close contact resulting in a lack of perithecia (Fig. 3C and S5). Supplementation with FEC did not restore formation of sexual structures (Fig. S5).

Deletion of Ffppt1 results in loss of PKS- and PKS/NRPS-derived products and in transcriptional down-regulation of distinct secondary metabolite cluster genes

Since deletion of Sfp-type PPTase-encoding genes in several fungal organisms resulted in absence of PKS-, NRPS- and PKS/NRPS-derived metabolites, we assessed production of F. fujikuroi metabolites in the Ffppt1 mutant. The wild type and the Ffppt1 mutant were cultivated under bikaverin-, fusarubin-, and fusarin-stimulating conditions, respectively. The wild type exhibited the typical coloration for each of the three metabolites in the respective induction medium, whereas the Ffppt1 mutant appeared colorless in all three experiments (Fig. 4). Analysis of the culture filtrates using high performance liquid chromatography (HPLC) coupled to an diode array detector (DAD) for bikaverin and fusarubins and to an ultraviolet light (UV) detector for fusarin detection, respectively, confirmed the absence of bikaverins, fusarubins and fusarins in the Ffppt1 mutant in contrast to the wild type (Fig. 4). Surprisingly, northern blot analyses revealed that genes encoding the PKS key enzymes for bikaverin and fusarubin production, respectively, are negatively affected on transcriptional level when Ffppt1 is deleted (Fig. 4). Furthermore, other bikaverin and fusarubin cluster genes are affected in the same manner in the Ffppt1 mutant in contrast to the wild type, the Ffaar1 and the Ffppt1 add-back strains (Fig. 4 and S6A). As expected, deletion of Ffppt1 does not affect the expression of all PKS- or NRPS-encoding genes tested: expression of fus1, encoding the PKS/NRPS hybrid responsible for fusarin production, was not repressed in the Ffppt1 mutant although no fusarins could be detected in the culture filtrate (Fig. 4). Similarly, expression of the NRPS-encoding Ffnps6 and PKS-encoding Ffpks6 genes was not effected in the Ffppt1 mutant compared to the wild type (Fig. 2 and Fig. S6B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Involvement of FfPpt1 in PKS- and PKS/NRPS-derived secondary metabolite production and gene regulation.

HPLC-UV chromatograms (bikaverins (510 nm) fusarubins (450 nm) and fusarins (363 nm)) in relative units (mAU) of indicated strains incubated as described in Methods. For HPLC conditions see Methods. Northern blot analyses of indicated strains from the same culture conditions probed with indicated cluster genes and rRNA visualization as loading control.

https://doi.org/10.1371/journal.pone.0037519.g004

Some F. fujikuroi strains are able to produce the mycotoxin moniliformin, which was assumed to be a PKS-derived metabolite [53]. To test this hypothesis we deleted the ppt1 locus in the highly moniliformin-producing wild type strain MRC2276. Analyses using HPLC coupled to Fourier transformation mass spectrometry (FTMS) showed that the wild-type strain as well as the Ffppt1 deletion mutant was able to form moniliformin in detectable amounts (Fig. S6C), suggesting that no Sfp-type PPTase activity is required for moniliformin production.

Influence of FfPpt1 on the production of sesqui- and diterpenes

To show if the loss of PKS-and NRPS-derived products in the Ffppt1 deletion mutants have an effect on biosynthesis of terpenes, strains IMI58289 (GA high-producing) and C-1995 (GA low-producing) were investigated for their production of diterpenoid GAs, the GA precursor ent-kaurene and the sesquiterpene alcohol α-acorenol. The latter was recently identified as the main volatile sesquiterpene produced by F. fujikuroi [8]. Quantification of GA in culture extracts by HPLC-DAD revealed no significant increase of GAs in the IMI58289/ΔFfppt1 mutant compared to IMI58289 (Fig. 5A), while deletion of Ffppt1 in the strain C-1995 resulted in a dramatic increase of GA production compared to the parental strain C-1995. Accordingly, GC-MS analysis of headspace extracts obtained by use of a closed loop stripping apparatus (CLSA) demonstrated that deletion of Ffppt1 in strain C-1995 led to a significant increase in ent-kaurene and α-acorenol production (Fig. 5B). These findings on the secondary metabolite level coincided with the expression levels for the genes ggs2 and cps/ks encoding the first two enzymes of GA biosynthesis: the signals were dramatically increased in the Ffppt1 mutant of strain C-1995 compared to the wild type, but only marginally altered in the IMI58289 Ffppt1 mutant compared to its parental strain (Fig. 5C). Surprisingly, a significantly decreased production of GAs was obtained in the Ffaar1 deletion mutant. In full agreement with this finding, expression signals for cps/ks and ggs2 were also significantly reduced compared to the Ffppt1 mutant and the wild-type IMI58289 (Fig. 5C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Effect of Ffppt1 deletion on terpene-derived metabolites.

A: HPLC quantified amounts of GA₃ (red) and the sum of GA₄ and GA₇ (blue) in mg per l culture and mycelium dry weight in mg of indicated strains. Data are given as means and standard deviations of two biological replicates. For cultivation and HPLC conditions see Methods. B: Quantified amounts of α-acorenol (red) and ent-kaurene (blue) of indicated C-1995 strains by GC-MS. Data are given as means and standard deviations of three biological replicates. For cultivation and GC-MS conditions see Methods. C: Northern blot analysis of the first GA cluster genes of the indicated strains. rRNA visualization as loading control.

https://doi.org/10.1371/journal.pone.0037519.g005

FfPpt1 is a pathogenicity factor in hydroponic rice cultures

To investigate whether the loss of PKS- and NRPS-derived compounds by deleting Ffppt1 affects rice root infections, the wild-type strain and the Ffppt1 mutant were transformed with the vector pHphDsRed conveying constitutive expression of the red fluorescent protein (DsRed). The DsRed-tagged wild type and Ffppt1 mutant were inoculated onto roots of germinated rice seedlings in the presence or absence of lysine and lysine plus BPS, respectively. Fluorescence microscopy showed that the wild type was able to penetrate and infect rice root cells under all conditions tested (Fig. 6). The Ffppt1 mutant established infection patterns in the presence of lysine only. No interaction of the Ffppt1 mutant with the rice roots was observed in the absence of lysine or the presence of lysine when BPS was supplemented concurrently. The hyphae show a non-oriented growth on the root surface and do not penetrate (Fig. 6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Fluorescence microscopy of Ffppt1 and wild-type strains during rice root infection assays.

Representative fluorescent microscopy pictures of indicated strains in rice root infection assays performed as described in Methods. Gamborg B5 Medium was supplemented as indicated.

https://doi.org/10.1371/journal.pone.0037519.g006

FfPpt1 is essential for lysine biosynthesis and involved in iron acquisition

In S. cerevisiae it was proven that the Sfp-type PPTase Lys5p is essential for activating the apo α-aminoadipate reductase Lys2p by 4′phosphopantetheinylation and therefore the enzyme is essentially involved in lysine biosynthesis [29]. As expected, targeted deletion of the homologous gene ppt1 in F. fujikuroi resulted in lysine auxotrophic mutants demonstrating that Ffppt1 is essential for lysine biosynthesis. The high degree of amino acid conservation of FfPpt1 to Lys5p makes it very likely that FfPpt1 activates the F. fujikuroi α-aminoadipate reductase Aar1 in the same mode of action as Lys5p activates Lys2p. The involvement of FfPpt1 in lysine biosynthesis is consistent with the observations made in several filamentous ascomycetes [30], [33], [36]–[39], [41]. Since Ffppt1 mutants were not viable without lysine supplementation it is suggested that FfAar1 cannot be post-translationally modified by the mitochondrial AcpS-type PPTase homolog.

A. nidulans npgA/cfwA mutants exhibit an iron uptake deficiency additional to a lysine auxotrophy since the production of NRPS-derived intra- and extra-cellular siderophores is abrogated and no additional reductive iron uptake system is present [35]. Similarly, the production of the NRPS-derived siderophores of C. graminicola is also dependent on Sfp-type PPTase activity, but respective PPTase mutants are able to grow without siderophore addition indicating the existence of a reductive iron uptake system [36]. In fact, absence of reductive iron acquisition systems has not been reported for any other fungal species other than A. nidulans [42]. Accordingly, the Ffppt1 mutants were able to grow without the addition of siderophores in the presence of lysine indicating the existence of such reductive iron uptake system also in F. fujikuroi. Furthermore, the ability of the F. fujikuroi wild type strain to grow in the presence of the iron chelator BPS, which specifically inhibits the reductive iron uptake system, strongly indicates the existence of NRPS-derived siderophores in F. fujikuroi. Since in the closely related species F. graminearum two NRPS-encoding genes were recently shown to be responsible for production of the extracellular siderophore triacetyl fusarine C [50] and the intra-cellular siderophore ferricrocin [54], respectively, the homologous genes in F. fujikuroi, Ffnps6 and Ffnps2, are very likely required for siderophore production. However, the nature of the F. fujikuroi siderophores has to be elucidated in future experiments. Nevertheless, the involvement of FfPpt1 in siderophore biosynthesis is evident since respective mutants were severely restricted in growth when grown in the presence of BPS and lysine, resembling the situation in A. nidulans and C. graminicola [35], [36]. This growth defect could be restored when FEC, which functions as intra- and extra-cellular siderophore in Schizosaccharomyces pombe [55], was supplemented, suggesting that it can be utilized by F. fujikuroi. Similar observations of FEC utilization were made in A. nidulans [56].

Ftr3/Fet3 are involved in iron metabolism of F. fujikuroi

Since Ffppt1 mutants were able to grow without siderophore addition in the presence of lysine but not when additional BPS was present, the existence of reductive iron uptake systems was suggested in F. fujikuroi. In F. graminearum two reductive iron uptake systems were identified, each consisting of a ferroxidase and an iron permease [51], [52]. Blast searches performed in the recently obtained F. fujikuroi genome database revealed the existence of three gene pairs each encoding an iron permease and a ferroxidase that share one promoter region. Phylogenetic analysis shows that two of them, FfFtr1/FfFet1 and FfFtr2/FfFet2, are closely related to the characterized proteins from F. graminearum, and that the newly identified proteins FfFtr3 and FfFet3 group together with FtrA and FetC from A. fumigatus, respectively. Similarly to the ftrA deletion mutant of A. fumigatus [43], F. fujikuroi Ffftr3/Fffet3 deletion mutants were not restricted in growth compared to the wild type, most likely due to the production of siderophores. However, when Ffftr3/Fffet3 were deleted in a Ffppt1 mutant background the growth defect in the presence of BPS was more dramatic, indicating that FfFtr3/FfFet3 are involved in iron acquisition but can be complemented by another iron uptake system. Generation of triple mutants of Ffppt1 and Ffftr3/Fffet3 together with either Ffftr1/Fffet1 or Ffftr2/Fffet2 could resolve the question if the identified putative reductive iron uptake systems have redundant functions. Furthermore, it would be interesting to investigate if FfFtr3/FfFet3, which are not present in F. graminearum, could restore iron transport in Fgftr1/Fgfet1 mutants.

FfSre1 controls iron metabolism

Similarly to the expression of ftrA and fetC observed in A. fumigatus [44] and genes involved in reductive iron uptake in H. capsulatum [49], all of the six genes putatively involved in reductive iron uptake in F. fujikuroi were shown to be expressed under iron deficient conditions and repressed upon iron addition, strongly suggesting their role in iron metabolism. Furthermore, expression of Ffnps6 encoding the homolog of the F. graminearum NRPS, responsible for extracellular siderophore production [50], revealed the same expression pattern thereby indicating a role of FfNps6 in iron homeostasis. The homolog in A. fumigatus, sidD, was also shown to be repressed by the addition of iron [44] underlining our hypothesis. In A. fumigatus and H. capsulatum the GATA-type transcription factor-encoding genes sreA and sre1, respectively, are expressed under iron sufficiency and act as repressors of genes involved in iron acquisition [44], [49]. The F. fujikuroi homologous gene Ffsre1 is also expressed under iron sufficient conditions. Deletion resulted in deregulation of all six genes putatively involved in reductive iron uptake as well as Ffnps6 when iron was supplemented, indicating that FfSre1 executes comparable repressing functions as SreA/Sre1 in A. nidulans, A. fumigatus and H. capsulatum, respectively [44], [49], [57], [58]. Interestingly, Ffsre1 deletion mutants were not viable in the presence of constant iron excess, which is likely due to a lack of repression, leading to increased iron flux into the cells. The increased intracellular iron pool might function in Fenton/Harber Weiss chemistry generating oxidative stress to a toxic level. However, under physiological iron concentrations Ffsre1 mutants exhibited a decreased sensitivity to H₂O₂ compared to the wild type, which might either be due to an increased intracellular siderophore concentration possibly scavenging free iron from Fenton/Harber Weiss chemistry and/or by specifically providing more intracellular iron as cofactor for the iron-dependent catalase known to detoxify H₂O₂. Supporting this hypothesis, addition of FEC to Ffppt1 mutants increases H₂O₂ resistance. This hypothesis is consistent with findings in A. nidulans and A. fumigatus where loss of intracellular siderophores increases the liable intracellular iron pool leading to a decreased oxidative stress resistance [59], [60]. Future studies focusing on the involvement of FfSre1 in iron metabolism and oxidative stress responses in Fusarium species will help to elucidate this complex context. Here, investigating the role of the homolog of the bZIP transcription factor HapX will be of special interest. In A. fumigatus HapX was shown to be a transcriptional repressor of genes involved in iron consuming pathways and an activator of genes involved in iron acquisition under iron deficiency. The HapX encoding gene itself is transcriptionally repressed by SreA under iron replete conditions [61].

Ppt1 controls secondary metabolism in F. fujikuroi

The findings that putatively NRPS-derived siderophore biosynthesis is affected in Ffppt1 mutants led us to investigate a broader PKS- and PKS/NRPS-derived secondary metabolite spectrum of F. fujikuroi. Similarly to the findings in several filamentous ascomycetes [33]–[39], F. fujikuroi ppt1 mutants were not able to produce any PKS-derived products such as bikaverins, fusarubins or PKS/NRPS-derived fusarins. This is in agreement with the fact that Sfp-type PPTases are essential for 4′-phosphopantetheinylation of ACPs and PCPs in PKSs and NRPSs, respectively [25], [26]. However, production of moniliformin, that was suggested to be produced in a PKS-dependent manner [53], was not altered in Ffppt1 mutants, indicating a biosynthetic pathway that is independent of Sfp-type PPTase activity in F. fujikuroi. Production of TC-derived secondary metabolites such as the diterpenoid GAs and the sesquiterpene alcohol α-acorenol was not negatively affected when ppt1 was deleted in F. fujikuroi. However, when the α-aminoadipate reductase-encoding gene Ffaar1 was deleted a significant reduction of GAs could be observed. A similar negative effect on secondary metabolism was found when the glutamine synthetase-encoding gene glnA was deleted in F. fujikuroi [62]. How FfAar1 is involved in GA metabolism remains elusive and awaits clarification in the future.

The effect of ppt1 deletion on GA and α-acorenol production in F. fujikuroi was shown to be strain-dependent. The ppt1 knock-out in strain C-1995 that produces very low amounts of GAs resulted in a significant increase in GA and α-acorenol production (ca. 15-fold increase), while the deletion in the highly GA-producing strain IMI58289 had almost no effect. However, since the Ffaar1 mutant in IMI58289 resulted in a significant decrease in GA production, but the Ffppt1 mutant produced wild-type-like amounts although FfAar1 is not functional, an increase in GA production can be observed that might be restricted to a wild-type level due to saturated enzyme activity in downstream reactions of the biosynthetic pathway in IMI58289. In summary, we postulate that the increased flux of acetyl-CoA precursor molecules is channeled into the terpene metabolism due to a block of PKS and NRPS pathways. A similar phenomenon of redirected secondary metabolite carbon flux occurs in Taraxacum brevicorniculatum when the cis-prenyltransferases responsible for natural rubber production are silenced [63].

An interesting observation is the specific transcriptional down-regulation for the two non-reducing PKS-encoding genes, bik1 and fsr1, as well as additional genes from the corresponding clusters that was not observed for any other secondary metabolite key enzyme tested in Ffppt1 mutants. A similar effect has not been reported in any other filamentous fungi investigated for secondary metabolism in Sfp-type PPTas deficient mutants so far. Although the nature of this transcriptional effect cannot be resolved, it is supported by the findings that a transcriptional interdependency of the bikaverin and the fusarubin structural enzymes was observed in earlier studies [9], [22]. Future experiments will have to provide evidence whether this phenomenon is related to non-reducing PKS gene clusters in general or specific for the bikaverin and fusarubin gene clusters in F. fujikuroi.

FfPpt1 is involved in asexual and sexual development most likely by inactivating PKS and/or NRPS pathways

We noted that Ffppt1 mutants revealed a significant reduction of conidiation that was independent from FEC supplementation, but could be restored when the mutant was grown contiguously to the wild-type strain. These data suggest that possible PKS- and/or NRPS-derived metabolites can function as diffusible conidiation signals in F. fujikuroi. In A. nidulans, sporulation is also suggested to be induced by a diffusible signal that involves the activity of the glutamine synthetase-like enzyme FluG [64]. The npgA/cfwA and fluG mutants showed a similar restoration of sporulation in contiguous growth experiments [34], [64]. Recently it was shown that the sporulation defect of A. nidulans fluG mutants could be rescued by the concomitant addition of specific TC- and PKS-derived products [65]. Further evidence for the existence of diffusible sporulation signals comes from Ascochyta pisi, where the metabolite P310/mycosporine was shown to induce sporulation [66], [67]. Mycosporines are produced by many fungal and marine organisms [68]. In cyanobacteria it was proven that this group of metabolites involves a NRPS during biosynthesis [69]. From the data obtained in this study it is intriguing to speculate that sporulation in F. fujikuroi also involves either a mycosporine-like NRPS product similar to A. pisi and/or a distinct mixture of secondary metabolites as reported in A. nidulans.

Apart from the defect in conidiation of F. fujikuroi ppt1 mutants, we also observed a defect in formation of sexual structures when Ffppt1 was deleted in the MAT1-2 mating partner and crossed with a wild-type MAT1-1 strain, whereas the vice versa situation led to formation of dark purple perithecia. In several ascomycetes, two NRPS-independent peptide pheromone/receptor systems that underlie transcriptional control of the MAT1-1 and MAT1-2 idiomorphs are a prerequisite for mating recognition in heterothallic species [70]. Although the MAT1-1 and MAT1-2 idiomorphs as well as the genes encoding the pheromone/receptor systems have recently been identified in heterothallic Fusarium species including F. fujikuroi [71] an involvement in mating recognition awaits experimental proof. Interestingly, in the homothallic species F. graminearum, which contains both MAT idiomorphs, deletion of one idiomorph prevents self-fertilization but mutants maintain the ability to outcross [72]. However, recent studies in F. graminearum revealed that the genes coding for the pheromone/receptor systems are not essential for self-fertilization and outcrossing [73], [74]. Our results suggest that FfPpt1 is involved in sexual recognition, disclosing the possibility that the MAT1-2 idiomorph specifically controls a PKS- and/or NRPS-derived metabolite or its receptor. And although differential cDNA screening and microarray analyses of MAT1-2 deletion mutants of F. verticillioides and F. graminearum did not reveal any apparent Sfp-type PPTase target or putative receptor-encoding gene to be transcriptionally controlled by the MAT1-2 idiomorph [75], [76], our hypothesis should thoroughly be investigated in F. fujikuroi.

FfPpt1 is a pathogenicity factor during rice root infection of F. fujikuroi

The data obtained from pathogenicity assays of fluorescently labeled F. fujikuroi ppt1 mutants in hydroponic rice cultures indicate that lysine production and iron uptake are essential for the establishment of infection of the rice roots. However, when lysine was supplemented, wild-type-like infection structures of the ppt1 mutant could be observed, indicating that the reductive iron uptake systems of F. fujikuroi are sufficient for iron acquisition during infection. An essential role for reductive iron uptake was reported in the smut fungus U. maydis during maize infection [77], whereas ftrA mutants of A. fumigatus were not affected in virulence in a murine model of invasive aspergillosis [43]. Interestingly, in F. graminearum, which lacks the Fet3/Ftr3 homologous system, siderophore-assisted iron uptake is essential for full pathogenicity on wheat [52]. Our data further suggest that no PKS- and NRPS-derived metabolites of F. fujikuroi are essential for primary invasion of rice roots. This stands in contrast to the observations reported for Sfp-type PPTase mutants of the hemibiotrophic plant pathogens C. graminicola and M. oryzae that were unable to cause primary infections [36], and C. sativus that showed strongly reduced primary infections on unwounded host plant leaves, respectively [41]. However, T. virens Sfp-type PPTase mutants were able to colonize Solanum lycopersicum roots in a wild-type-like manner when lysine was supplemented [39]. The theory that plant pathogenic fungi and bacteria need PKS- and NRPS-derived secondary metabolites for establishment of full virulence is reported in several species [78]–[83]. Whether this is also true for F. fujikuroi needs to be investigated in a more sophisticated pathogenicity assay in the future. It is possible that the species-specific production of GAs in F. fujikuroi is the main determinant of host specificity and is essential for primary infection of rice roots. The fact that F. fujikuroi strains lacking the global regulator velvet are defective in both GA production and virulence support our suggestion on the role of TC-derived GAs for the infection of rice roots [84]. Examples for the essential role of other TC-derived secondary metabolites during infection come from the gray mold fungus B. cinera [85] and the more closely related species F. graminearum [86]. Analyses of the roles different F. fujikuroi secondary metabolites play during primary infection of rice will be a major task in the future.

Summarizing, our studies describe the first investigations centering on a Sfp-type PPTase in the genus Fusarium. We show unequivocally that FfPpt1 is essentially involved in primary metabolism of lysine biosynthesis and in PKS-, PKS/NRPS- and NRPS-derived secondary metabolites such as bikaverins, fusarubins, fusarins, and most likely siderophores. Surprisingly, moniliformin production was not affected indicating that the biosynthetic pathway does not involve a Sfp-tye PPTase-dependent PKS. Furthermore we provide strong evidence that deletion of Ffppt1 causes re-channelling of carbon flux into the terpene metabolism which results in increased GA and α-acorenol production. Therefore, Ffppt1 mutants provide a reasonable strain improvement strategy for terpene-derived secondary metabolite production. Pathogenicity assays using hydroponic rice cultures revealed that lysine biosynthesis and iron acquisition, but not PKS and NRPS secondary metabolism is essential for establishing primary infections of F. fujikuroi on rice roots. Additionally, the results obtained disclose the possibilities that PKS- and/or NRPS-derived metabolites might function as diffusible conidiation signals and also might play a crucial role in mating recognition in dependency on a specific MAT idiomorph in heterothallic Fusarium species.

Furthermore, our studies revealed the existence of a third previously unidentified putative reductive iron uptake system consisting of FfFtr3 and FfFet3 that is closely related to the only reductive system, FtrA/FetC, in A. fumigatus. Functional characterization using targeted deletion of both genes, Ffftr3/Fffet3, provides strong evidence that they are involved in iron acquisition and under transcriptional repression of the GATA-type transcription factor FfSre1 under iron-replete conditions. Deletion of the first Fusarium homolog of this GATA-type transcription factor strongly indicates its involvement in regulation of iron homeostasis and oxidative stress resistance, providing evidence for conserved regulation mechanisms between Fusarium and Aspergillus species in this respect.

Fungal strains and culture conditions

The wild-type strains F. fujikuroi IMI58289 (Commonwealth Mycological Institute, Kew, UK), F. fujikuroi C-1995 (kindly provided by J.F. Leslie, Kansas State University), and the moniliformin-producing strain F. fujikuroi MRC2276 (kindly provided by W. F. O. Marasas, Research Institute for Nutritional Diseases, South Africa) were used for ppt1 knock-out experiments. For all cultures, F. fujikuroi was preincubated at 28°C for 48 h in 300 ml Erlenmeyer flasks with 100 ml Darken medium (DVK) [87] on a rotary shaker at 180 rpm. For RNA isolation and secondary metabolite analyses, 0.5 ml DVK were used to inoculate synthetic ICI (Imperial Chemical Industries Ltd., UK) media [88] containing either 6 mm glutamine (GAs and bikaverins), 60 mm glutamine (fusarins) or 6 mm NaNO₃ (fusarubins). For cultivation including Ffppt1 mutants, lysine was added to all media to give a final concentration of 1 mm. The cultures were incubated at 28°C on a rotary shaker at 190 rpm for 3, 5, 7 or 10 days. For iron shift experiments FeCl₃ was added to a final concentration of 1 mm and incubated for 2 h. For moniliformin analyses, the strains were grown at 28°C for 14 days on cracked corn as previously described [89]. Headspace analyses were performed with agar plate cultures grown on complete medium (CM) [90] amended with 1 mm lysine after three days of incubation. For protoplasting, 0.5 ml of the starter culture was transferred into Erlenmeyer flasks with 100 ml ICI medium containing 6 mm (NH₄)₂SO₄ and 10 g/L fructose instead of sucrose and incubated at 28°C on a rotary shaker at 190 rpm for 18 h. For DNA extraction, fungal strains were grown for 3 days at 28°C on cellophane sheets (Alba Gewürze, Bielefeld, Germany) placed on solidified CM. For sporulation assays solidified 20% (v/v) vegetable juice (V8) (Campbell Foods, Puurs, Belgium) containing 30 mm CaCO₃ and 10 mm lysine was used. Additionally solidified V8 contained 2 µm ferrichrome (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) as indicated. After 10 days incubation under constant light spores were washed of the plates, filtered and counted using a hemocytometer. For growth tests 5% (w/v) Czapek Dox (CD) medium (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) was used and complemented with 10 mm lysine, 100 µM BPS, 2 µM FEC as indicated. For additional growth test, solidified CM without iron was used containing 10 mm lysine and 1 mm FeCl₃ when indicated. Sexual crossings were performed on carrot agar containing 10 mm lysine plus 2 µm FEC when indicated as described by Klittich and Leslie [91]. Microscopy of perithecia was performed using a SteREO Discovery.V20™ microscope equipped with an AxioCam MRc (Carl Zeiss MicroImaging GmbH, Jena, Germany). Spores were visualized using an Axio Imager.M2 (Carl Zeiss MicroImaging GmbH, Jena, Germany). Differential interference contrast (DIC) was used for bright field images captured with uniform exposure time using an AxioCam MRm. Images were processed uniformly using AxioVision Rel. 4.8 (Carl Zeiss MicroImagingGmbH, Jena, Germany).

Standard molecular methods

DNA and RNA analysis used standard techniques [92]. Fungal DNA or RNA was prepared by first grinding lyophilized mycelium into a fine powder with a mortar and pestle and then dispersing it in extraction buffer as described by Cenis [93]. DNA for Southern hybridization experiments was prepared following the protocol of Doyle and Doyle [94]. For Southern blot analysis, genomic DNA was digested with the indicated restriction enzymes (Fermentas GmbH, St. Leon-Rot, Germany), fractionated in 1% (w/v) agarose gels, and transferred to Nytran® nylon transfer membranes (Whatman Inc., Sanford, ME, USA) by downward blotting [95]. ³²P-labelled probes were prepared using the random oligomer-primer method and membranes were hybridized according to the protocol of Sambrook et al. [92].

Total F. fujikuroi RNA was isolated using the RNAgents total RNA isolation kit (Promega GmbH, Mannheim, Germany). Samples of 20 µg of total RNA were transferred to Hybond-N⁺ membranes after electrophoresis on a 1% (w/v) agarose gel containing 1% (v/v) formaldehyde, according to Sambrook et al. [92]. Northern blot hybridizations were accomplished by the method of Church and Gilbert [96]. cDNA was synthesized from 1 µg of total RNA and the SuperScript II reverse transcriptase (Invitrogen, Groningen, The Netherlands) according to the manufacturer's instructions.

All primers used for PCR were obtained from Eurofins GmbH (Ebersberg, Germany) (Table S1). PCR reactions contained 25 ng DNA, 5 pmol of each primer, 200 nm dNTPs, and 1 unit of BioTherm™DNA polymerase (GeneCraft GmbH, Lüdinghausen, Germany) and were initiated with a 4 min soak at 94°C followed by 36 cycles of 1 min at 94°C, 1 min at 56 to 65°C, 1–3 min at 70°C, and a final soak for 10 min at 70°C. PCR products were cloned into pCR®2.1-TOPO® vector using the TOPO TA Cloning® kit (Invitrogen, Groningen, The Netherlands) and transformed into Escherichia coli (Invitrogen). Plasmid DNA from E. coli was extracted using the GeneJET™ Plasmid Miniprep Kit (Fermentas GmbH, St. Leon-Rot, Germany) and sequenced using the BigDye® Terminator v3.1 Cycle Sequencing Kit and the ABI Prism® 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) according the manufacturer's instructions. DNA and protein sequence alignments were done with DNA STAR (Madison, WI, USA). Sequence homology searches were performed using the NCBI database server. Protein homology was based on BlastX searches [97]. Phylogenetic analysis was performed using the web-based tool at www.phylogeny.fr [98]. The nucleotide and protein sequences were deposited in GenBank under accession number HE614113 (ppt1), HE614114 (aar1), HE614115 (fet1), HE614116 (ftr1), HE614117 (ftr2), HE614118 (fet2), HE614119 (ftr3), HE614120 (fet3), HE614121 (nps2), HE614122 (nps6) and HE614123 (sre1), respectively.

Plasmid construction

The putative F. fujikuroi ppt1 gene and flanking regions were amplified using the primer pairs ppt1-F and ppt1-R which were based on the putative F. verticillioides ppt1 (FVEG_01894.3) sequence. The F. fujikuroi ppt1 knock-out plasmid pΔppt1 was created by sequentially cloning the 550 bp 5′ Ffppt1 flank (generated with primers ppt1-5′F/ppt1-5′R) and the 770 bp 3′ Ffppt1 flank (generated with primers ppt1-3′F/ppt1-3′R) into pNR1 using SacI/XbaI and HindIII/SalI restriction sites, respectively [99], such that the nourseothricin resistance cassette was flanked by F. fujikuroi genomic sequence. For generating a complementation construct, a 1.8 kb fragment including 5′ and 3′ non-coding regions was amplified with primers ppt1-Prom-F/ppt1-Term-R. The plasmids pΔaar1, pΔftr3/fet3 and pΔsre1 were assembled using yeast recombinational cloning as essentially described for Neurospora crassa deletion vectors [100] and recently established for F. fujikuroi vectors [84]. The 5′ and 3′ flanks of Ffftr3/fet3 and Ffsre1 were amplified using primer pairs “gene”-5′-F1/-R1 and “gene”-3′-F1/-R1, respectively. Plasmid DNA from S. cerevisiae was extracted using the GeneJET™ Plasmid Miniprep Kit (Fermentas GmbH, St. Leon-Rot, Germany) with slight modifications: cells were resuspended in 300 µl Resuspension Solution plus 100 µl glass beads, lysed by addition of 600 µl Lysis Solution and neutralized with 450 µl Neutralization Solution. DNA fragments used for deletion of Ffftr3/fet3 and Ffsre1 were prepared by PCR using primers “gene”-5′-F1 and “gene”-3′-R1 and 1 µl of pΔftr3/fet3 or pΔsre1, respectively, as template. The plasmid pHphDsRed was constructed by ligating the HindIII/XbaI fragment of pChap-GFP [101] containing the hygromycin resistance cassette into HindIII/XbaI restricted pPgpd-DsRed [102].

Fungal transformations

Preparation of protoplasts from F. fujikuroi mycelium was carried out as described [103]. Briefly, 10⁷ protoplasts of F. fujikuroi strains were transformed with 10 µg of the replacement cassette of the vector pΔppt1 or PCR products obtained of pΔftr3/fet3 and pΔsre1, respectively, as described above. Transformed protoplasts were regenerated for 6–7 days at 28°C in a complete regeneration agar (0.7 m sucrose, 0.05% yeast extract) containing 100 µg/ml nourseothricin and 1 mm lysine in case of targeted Ffppt1 deletion (Werner-Bioagents, Jena, Germany) or 100 µg/ml hygromycin and 1 mm lysine (Calbiochem, Darmstadt, Germany) in case of targeted Ffaar1, Ffftr3/fet3 and sre1 deletion. For complementation of ΔFfppt1 strains, 10 µg of the genomic Ffppt1 PCR fragment was used for transformation as described above, but without addition of lysine in the regeneration media. Protoplasts of the IMI58289 wild-type strain and the Ffppt1 mutant were transformed with 20 µg pHphDsRed and transformed protoplasts were regenerated in complete regeneration agar containing 100 µg/ml hygromycin and 1 mm lysine.

The homologous integration events in transformants targeting replacement of Ffppt1 with the nourseothricin resistance marker were verified by PCR using primers ppt1- F and ppt1- R targeting outside the replacement fragment in combination with pLOF-OliP and Tub-T, respectively. In case of hygromycine resistant transformants, targeted replacement was verified using a primer outside the replacement fragment in combination with a primer targeting the hygromycine resistance cassette (“gene”-F1d/pCSN44-trpC-T and “gene”-R1d/pCSN44-trpC-P) in case of ΔFfaar1, ΔFfftr3/fet3 and ΔFfsre1. The absence or presence of the wild-type gene loci in deletion and add-back strains was verified by PCR using primer pairs targeting the replaced coding region (“gene”-WT-F1/-R1).

Virulence assays

Infection assays of single plants of Oryza sativa spp. japonica c.v. Nipponbare were performed as described previously [84]. Gamborg B5 Medium (Duchefa Biochemie, Haarlem, The Netherlands) solution was supplemented with 10 mm lysine and 10 mm BPS when indicated. Microscopy was performed using an Axio Imager.M2 (Carl Zeiss MicroImaging GmbH, Jena, Germany). DIC was used for bright field images and DsRed fluorescence was detected using filterset 38 (excitation band pass 470/40 nm, color splitter 495, emission band pass 525/50 nm). Images were captured with uniform exposure time using an AxioCam MRm and were processed uniformly using AxioVision Rel. 4.8 (both Carl Zeiss MicroImaging GmbH, Jena, Germany).

Chemical analysis

Gibberellic acids GA₃ and GA_4/7 were extracted from 20 ml culture filtrate after 7 days of incubation in ICI medium containing 6 mm glutamine. Extraction was performed using Sep Pak C18 cartridges (Waters GmbH, Eschborn, Germany) from which GA₃ was eluted with 2 ml 20% acetonitril (ACN) (LGC/Promochem GmbH, Wesel, Germany) and GA_4/7 were eluted with 2 ml 55% ACN. GA₃ and GA_4/7 amounts were measured by HPLC-DAD analysis using a Merck-Hitachi System (Merck KGaA, Darmstadt, Germany) consisting of a gradient pump (L-7100), an autosampler (L-7200) and a Diode Array Detector (L-245). As column a Lichrospher 100 RP-18 column (5 µm; 250 mm×4 mm; Merck KGaA) was applied. HPLC conditions were as follows: solvent A: 0.05% H₃PO₄ (Merck KGaA), pH 3; solvent B: ACN. The subsequent gradient was applied: 15 min 15% B; in 20 min to 40% B; in 2 min to 15% B. Data analysis was carried out using EZChrom Elite Version 3.3.2 SP1 (Scientific Software, Inc.). Quantification of GAs was performed by generating a calibration line using different dilutions of 1 mg/ml GA₃ and GA_4/7 standards (DKSH GmbH Hamburg, Germany). GA amounts were calculated per 1 l culture filtrate and 1 g dry weight applying peak areas of the different samples and the gradient of the calibration line.

Fusarin, bikaverin and fusarubin production was analyzed as previously described [9], [104].

For moniliformin analyses fungal cultures were extracted as previously described [105] for 1 h on a rotary shaker at 190 rpm at 28°C. For analyses 1 ml of extract was evaporated under a stream of nitrogen at 40°C, dissolved in 150 µl 5% Methanol (v/v) and analyzed by HPLC-FTMS using chromatographic conditions as previously described [106].

The volatiles released by agar plate cultures were collected and analyzed as previously described [8]. Briefly, the volatiles emitted by the fungal cultures on solidified CM were collected by use of a closed loop stripping apparatus (CLSA). Therefore, a circulating air flow was directed through a charcoal filter (Chromtech GmbH, Idstein, Precision Charcoal Filter, 5 mg) in a closed apparatus containing the fungal culture for 24 h. The charcoal filter was extracted with 30 µl of analytically pure dichloromethane and the obtained solutions were immediately analyzed by GC-MS and stored at −80°C. GC-MS analyses were carried out on a HP6890 GC system connected to a HP5973 Mass Selective Detector fitted with a HP-5 fused silica capillary column (25 m×0.22 mm, 0.25 µm film, SGE Inc.). Conditions were as follows: inlet pressure: 77.1 kPa, He 23.3 ml min⁻¹; injection volume: 1 µl; injector: 250°C; transfer line: 300°C; electron energy: 70 eV. The GC was programmed as follows: 50°C (5 min isothermic), increasing at 10°C min⁻¹ to 320°C, and operated in splitless mode (60 s valve time); carrier gas (He): 1.0 ml min⁻¹. Quantification was carried out by peak integration with the MSD Chem Station software (Agilent) of three replicate samples and is given as arithmetic means ± standard deviations, normalized to 100% for the production of the C-1995 wild-type strain.