|Home | About | Journals | Submit | Contact Us | Français|
Fungal secondary metabolites are important bioactive compounds but the conditions leading to expression of most of the putative secondary metabolism (SM) genes predicted by fungal genomics are unknown. Here we describe a novel mechanism involved in SM-gene regulation based on the finding that, in Aspergillus nidulans, mutants lacking components involved in heterochromatin formation show de-repression of genes involved in biosynthesis of sterigmatocystin (ST), penicillin and terrequinone A. During the active growth phase, the silent ST gene cluster is marked by histone H3 lysine 9 trimethylation and contains high levels of the heterochromatin protein-1 (HepA). Upon growth arrest and activation of SM, HepA and trimethylated H3K9 levels decrease concomitantly with increasing levels of acetylated histone H3. SM-specific chromatin modifications are restricted to genes located inside the ST cluster, and constitutive heterochromatic marks persist at loci immediately outside the cluster. LaeA, a global activator of SM clusters in fungi, counteracts the establishment of heterochromatic marks. Thus, one level of regulation of the A. nidulans ST cluster employs epigenetic control by H3K9 methylation and HepA binding to establish a repressive chromatin structure and LaeA is involved in reversal of this heterochromatic signature inside the cluster, but not in that of flanking genes.
Fungi are well-known producers of bioactive compounds with antibiotic, growth-regulating, toxic, mutagenic, immunosuppressive, enzyme inhibitory and other biological effects (Keller et al., 2005; Pelaez, 2005). The term ‘secondary metabolism’ (SM) denotes biochemical processes underlying the formation of fungal metabolites such as toxins, antibiotics and immune-suppressing agents (Hoffmeister and Keller, 2007). In the ascomycete fungus Aspergillus nidulans, the formation of secondary metabolites is correlated with growth arrest and reproduction (sexual and asexual) and a wealth of information is available on the genetic basis regulating these processes (Calvo et al., 2002; Yu and Keller, 2005). The finding of a nuclear complex including both the light-regulated developmental factor VeA (Calvo et al., 2004; Bayram et al., 2008) and the global regulator of SM, LaeA (Bok and Keller, 2004) establishes a mechanistic link between SM production and morphological differentiation.
Arguably the secondary metabolite synthesis pathway which is best understood is that of sterigmatocystin (ST). Like most other secondary metabolites, the genes required for ST biosynthesis are arranged in a cluster, including the pathway-specific activator gene aflR (Keller and Hohn, 1997). A large body of genetic evidence has led to the identification of transcriptional regulators and signalling pathways involved in expression of the ST cluster genes (reviewed in Yu and Keller, 2005). One additional activator of ST production is LaeA. Bok and Keller (2004) showed that LaeA is a nuclear protein required for efficient transcription of the ST cluster, including the pathway activator aflR. LaeA contains a predicted and a functionally necessary S-adenosyl-methionine (SAM) binding domain (Bok et al., 2006a). Such domains are present in all members of the methylase superfamily (Kozbial and Mushegian, 2005), including histone methyltransferases involved in chromatin modification and epigenetic gene regulation. LaeA is a conserved fungal protein and has been shown to be a global regulator of chemical diversity (Bok et al., 2006a,b; Perrin et al., 2007; Kosalkova et al., 2009), reproduction (Bayram et al., 2008; Kale et al., 2008) and virulence (Bok et al., 2005) in different Aspergillus and one Penicillium species.
The cluster arrangement of secondary metabolite genes may facilitate a chromatin-based co-regulation mechanism. Deletion of the A. nidulans histone deacetylase HdaA or inhibition of other fungal HDACs by trichostatin A leads to over-production of several secondary metabolites (Shwab et al., 2007). Chichewicz and colleagues treated a number of different fungi with HDAC and DNA methyltransferase inhibitors and found clear evidence for enhanced chemical diversity and higher natural product formation (Williams et al., 2008; Fisch et al., 2009). A position effect has been shown to be involved in the expression of some secondary metabolite biosynthetic and regulatory genes. Ectopic expression of aflR bypasses the need for LaeA function and the arginine biosynthetic gene argB is silenced in a laeAΔ strain when placed inside the ST cluster (Bok et al., 2006a). Similar position effects on secondary metabolite gene regulation have been reported for other Aspergillus species as well (Chiou et al., 2002; Roze et al., 2007; Smith et al., 2007). Additionally, our laboratories recently showed that histone H3 lysine 4 di- and trimethylation (H3K4me2/me3), a chromatin mark usually associated with active genes, but also required for telomere silencing in Saccharomyces cerevisiae (Mueller et al., 2006), regulates SM in A. nidulans. A mutant lacking the Bre2 homologous component (termed CclA in A. nidulans) of the COMPASS complex expressed cryptic SM gene clusters in A. nidulans. Notably, the reduced amounts of H3K4me in some SM gene promoters in cclAΔ strains also resulted in low levels of histone H3 lysine 9 trimethylation (H3K9me3), a chromatin mark associated with gene silencing and heterochromatin formation (Bok et al., 2009). Taken together, these results suggest a strong influence of chromatin-mediated regulation on fungal SM and might represent bona fide heritable epigenetic information.
In transcriptionally active chromatin, lysines of histones H3 and H4 (including lysine 9 of histone H3, H3K9) are usually acetylated (Li et al., 2007). Heterochromatic domains are transcriptionally silent and characterized by hypoacetylation of lysines in H3 and H4 (Holbert and Marmorstein, 2005). Lack of acetylation enables mono-, di- or tri-methylation of H3K9 (H3K9me) by the histone methyltransferase Su(var)3-9 in Drosophila (Taddei et al., 2001), or Clr4, its orthologue, in Schizosaccharomyces pombe (Rea et al., 2000; Noma et al., 2001). This histone modification is recognized by one of the principal components of heterochromatin, the heterochromation protein 1 (HP1) (Cryderman et al., 1998; Wang et al., 2000). HP1 recognizes H3K9me through its amino terminal chromodomain that is necessary for both targeting and transcriptional repression (Fanti and Pimpinelli, 2008). Although HP1 is predominantly associated with pericentric heterochromatin, it is also found at many euchromatic sites where it can promote repression in a H3K9 methylation-dependent or independent process. This system has been found to be conserved in numerous organisms, including filamentous ascomycetes, such as Neurospora crassa, where it has been shown to direct DNA methylation (Freitag et al., 2004; Lewis et al., 2009).
To address whether a process analogous to developmentally regulated facultative heterochromatin regulates SM in A. nidulans, we deleted the HP1 (SWI6 in S. pombe) orthologue HepA and the H3K9-specific methylase ClrD (the orthologue of SU(VAR)3-9 in D. melanogaster and of Clr4 in S. pombe) and demonstrate their loss leads to over-expression of several A. nidulans SM genes. Using chromatin immunoprecipitation (ChIP) we demonstrate that repressive histone marks and high levels of HepA are associated with the silent ST cluster and that LaeA plays a role in reversing this heterochromatic state during the onset of SM. This is the first direct experimental support for a model in which secondary metabolite gene clusters are regulated by a metabolically dependent, reversible formation of heterochromatin.
The unique putative homologue of HP1, hepA (gene number AN1905.3), shows a chromatin modifier protein signature, i.e. a 50-amino-acid N-terminal chromo-domain followed by a C-terminal chromo-shadow domain typical for all known HP1 type proteins. A. nidulans HepA shows extended homologies to putative heterochromatin proteins (between 60% and 35% identity, results not shown). Budding yeast S. cerevisiae does not contain a HP1 protein and the best studied fungal HP1 homologues are the S. pombe Swi6 (Ekwall et al., 1996) and Hpo from N. crassa (Freitag et al., 2004). HepA shows limited overall identity (26%) to SWI6, but high similarity in the conserved chromo domain (38% identity) and the chromo-shadow domain (42% identity). Deletion of hepA does not lead to any evident morphological or physiological phenotype and the mutant strain grows with the same rate and sporulates as the isogenic wild-type strain (Figs S1 and S2). This is in contrast to several other systems studied so far in which HP1 deletions have been shown to strongly affect viability (reviewed in Hiragami and Festenstein, 2005). Mutations in S. pombe Swi6 lead to loss of chromosome stability (Allshire et al., 1995; Ekwall et al., 1996) and hpo mutations in N. crassa show pronounced growth defects (Freitag et al., 2004).
In a transcriptome analysis comparing submerged cultures of hepAΔ to an isogenic wild-type strain we identified several upregulated transcripts coding for proteins involved in SM (data not shown). Northern blots under conditions of secondary metabolite production (for details, see Experimental procedures) confirmed the microarray results (Figs 1A and 2A). We tested some genes involved in ST production and found strong overexpression in the hepAΔ mutant for the ST-cluster transcription factor aflR (~7-fold), and for two tested structural genes stcO (~2.5-fold) and stcU (~14-fold). Also genes involved in isopenicillin A production (ipnA) and in terraquinone A biosynthesis (tdiB) were significantly overexpressed. Interestingly, laeA, the gene encoding the general regulator of SM in fungi, also shows a ~2-fold enhanced expression in the hepAΔ strain (Fig. 1A and Fig. S3). Importantly, two genes located immediately outside of the ST cluster and expressed invariantly under primary or secondary metabolism conditions (genes AN7826.3 and AN7801.3; Fig. 1B) did not respond to hepA deletion suggesting that the effect is restricted to genes located inside the ST cluster. Complementation of the hepAΔ strain with the complete hepA gene including promoter and terminator sequences resulted in reversal of the hepA deletion phenotype, as tested for aflR expression (Fig. S4). This demonstrates that the observed hepAΔ phenotype was caused by the hepA deletion.
Higher laeA transcript levels in the hepAΔ mutant could be responsible for overexpression of LaeA target genes as observed in expression studies and metabolite analysis (Fig. 1). We therefore tested the genetic interaction between hepA and laeA and found that in a hepAΔlaeAΔ double mutant the expression of aflR, but not that of stcO, is partially remediated in comparison with a laeAΔ single mutant (Fig. 2A). As the biosynthesis of ST or its immediate precursor, norsolorinic acid (NOR), requires the expression of all genes in the cluster, we wanted to test if deletion of hepA not only remediates aflR expression in a laeAΔ strain but also leads to remediation of metabolite accumulation. This was not the case. In the hepAΔlaeAΔ double mutant the partial remediation of aflR expression did not translate into remediation of ST production after 48 h of growth in liquid GMM. Unexpectedly, when the strains were grown on solid GMM medium for 5 days, metabolite production in the hepAΔlaeAΔ double mutant was clearly far above the levels observed in the laeAΔ single mutant for ST (Fig. 2B) and also for NOR (Fig. S5A). These data indicate that the higher expression levels of aflR caused by the hepA deletion leads to slightly increased ST biosynthesis enzyme levels, the activities of which become only apparent as increased metabolite levels after longer incubation periods. Additionally, or alternatively, some environmental factors present only when cultures are grown on solid media allow full restoration of ST production in the absence of LaeA.
We further tested the role of HepA by measuring HepA occupancy at the ST cluster using ChIP. Because the commercial HP1 antibody considerably cross-reacts with other A. nidulans proteins (see Western analysis in Fig. S6) we always performed ChIP reaction in parallel with a hepAΔ control strain. Two genes inside the cluster (aflR, stcO) and one non-cluster gene were tested at time points when ST genes are not expressed (24 h) or are expressed (48 h). Figure 3A and B show that HepA occupancy at the aflR and stcO promoter regions significantly decreases in cultures of 48 h as compared with cultures of 24 h. A gene immediately telomere-distal to the ST cluster (locus AN7801) shows high HepA levels and these levels do not change when the neighbouring cluster genes are activated (48 h, Fig. 3C). This suggests that a chromatin boundary element exists between these two coding regions and actively separates the ST cluster from the surrounding heterochromatic region. We investigated the presence of HepA in the bidirectional promoter of the primary metabolism niiA-niaD gene cluster in which chromatin remodelling is regulated by different forms of nitrogen (Muro-pastor et al., 1999; Berger et al., 2008). HepA cannot be detected above background levels at this promoter by ChIP at either 24 or 48 h of culture (data not shown). The high HepA levels at the ST locus during the active growth phase hence strongly suggests an important role of heterochromatin in ST gene repression.
HP1 function requires H3K9me3 to mediate heterochromatin formation. Genome analysis of A. nidulans showed one protein (gene number AN1170.3) highly similar to S. pombe Clr4 and N. crassa DIM-5 (Fig. S7). DIM-5 and H3K9 methylation is required in the latter organism to recruit HP1 as well as DIM-2, a DNA methylase leading to epigenetic silencing (Tamaru et al., 2003; Freitag et al., 2004). In accordance with the A. nidulans nomenclature AN1170.3 is called clrD. ClrD shares 41% overall identity with the S. pombe H3K9 methyltransferase and 38% identity with N. crassa DIM-5. ClrD in A. nidulans and DIM-5 in N. crassa are the only two SET-domain proteins in these organisms showing a predicted pre-SET motif (putative Zn-finger) combined with a SET domain (putative methyltransferase domain). Interestingly, ClrD and the putative orthologues from other filamentous fungi lack the N-terminal chromo-domain found in S. pombe Clr4 and other metazoan Su(var)3-9 HMTs (Fig. S7). ClrD is likely the only H3K9 methylase in A. nidulans as the deletion strain showed only background signals in Western and ChIP analysis using the H3K9me3-specific antibody (Fig. S8).
Similarly to what has been observed in the hepA deletion strain, inactivation of clrD does not show any evident morphological or physiological phenotype and the mutant grows with the same rate as the isogenic wild type on solid and in liquid media (Figs S1 and S2). Also similarly to hepAΔ, the clrD inactivated strain displayed higher expression of the ST cluster genes aflR and stcO under conditions of SM (Fig. 4A) which translates to a slightly increased ST production in liquid GMM (Fig. 4B). The expression pattern of the ST cluster-neighbouring genes AN7826.3 and AN7801.3 remains unchanged in the clrDΔ strain, identically to what is seen in hepAΔ. Somewhat differently to the laeAΔhepAΔ strain, which did show some remediation of aflR expression, neither aflR nor stcO transcripts are detectable in the laeAΔclrDΔ background. Consistently, we observed lack of ST production in liquid GMM, but also for this double mutant, a clear remediation of SM production was observed after 5 days' incubation on solid GMM for ST (Fig. 4B) and NOR (Fig. S5).
The slight difference in aflR expression between the laeAΔhepAΔ and the laeAΔclrDΔ strains may indicate that HepA has additional, H3K9me3-independent, functions in ST cluster silencing. However, data from ChIP analysis support a direct involvement of ClrD and histone H3K9me3 levels in the regulation of the ST cluster region. As shown in Fig. 5A and B, trimethylation of H3K9 decreases in the wild type at the aflR and stcO promoters concomitantly with transcriptional activation (compare Fig. 4A) and depletion of HepA (compare Fig. 3A). This decrease in H3K9me3 is restricted to ST cluster genes and not seen at the AN7801.3 locus (Fig. 5C). Importantly, and consistent with the predicted role of H3K9me for HepA binding, HepA levels are reduced approximately to the hepAΔ background in the clrDΔ strain (see Fig. 6A).
A straightforward interpretation of the results presented above is that LaeA counteracts the heterochromatization mediated by the HepA/ClrD complex. We thus tested the effect of laeAΔ on HepA occupancy and H3K9me3 levels at the aflR promoter. Figure 6A shows that HepA occupancy at the aflR promoter is drastically increased in a laeAΔ background both after 24 and 48 h culturing. This increased occupancy correlates strongly with increased H3K9me3 levels (Fig. 6B). Whereas in laeAΔ both H3K9 methylation levels and HepA occupancy are greatly increased, this is not the case in the clrDΔlaeAΔ double mutant, supporting our previous observation that efficient HepA binding requires H3K9me3. However, H3K9me3 levels are also low in hepAΔ strains and this suggests that HepA is required for efficient targeting of ClrD to the aflR promoter. As expected, the absence of the ClrD methylase (in clrDΔlaeA+) leads to low HepA occupancy, but HepA is present at wild type levels when both LaeA and ClrD are not functioning (double clrDΔlaeAΔ). This observation suggests that at the aflR promoter HepA shows also a H3K9me3-independent localization and this might explain why remediation for aflR expression is seen in hepAΔlaeAΔ, but not in clrDΔlaeAΔ.
Taken together, these results suggest a role of LaeA in counteracting H3K9 trimethylation, thus preventing HepA binding and the formation of repressive chromatin. To test whether LaeA is also involved in histone acetylation, an activating euchromatic mark, we analysed by ChIP the acetylation status of histone H3 at lysines K9 and K14 (Fig. 6C). H3K9/K14 acetylation at the aflR promoter is higher at conditions of ST cluster activation (48 h culture). H3K9/K14 acetylation is strongly increased in both the hepAΔ and in clrDΔ strains, regardless whether LaeA is functional or not. These experiments demonstrate that LaeA does not directly participate in the establishment of the activating H3K9K14 acetylation chromatin marks. Moreover, while these crucial marks are correlated with transcriptional activation in the wild type and in the heterochromatin single mutants as well as in the laeAΔhepAΔ double mutant, we see strongly increased acetylation also under conditions which do not promote aflR expression, i.e. in 24 h cultures of clrDΔ and in 48 h cultures of laeAΔclrDΔ. Thus, H3K9K14 acetylation levels are not necessarily associated with transcriptional activation of aflR but are mainly determined by the methylation level of H3K9. From this set of experiments we can conclude that the balance between acetylated and methylated lysine 9 of histone H3 is determined by HepA and ClrD and that LaeA directly counteracts HepA/ClrD function.
Pre-genomic studies and fungal genome sequencing have identified hundreds of clustered genes known or predicted to be involved in production of polyketides, non-ribosomal peptides or alkaloids (Hoffmeister and Keller, 2007). The contiguous arrangement of genes involved in the synthesis of a given compound might facilitate coordinated gene expression and chromatin domain modifiers thus may regulate such large genomic regions by establishing or lifting chromatin-based gene silencing.
The genetic and biochemical evidence presented here for the first time demonstrates that the establishment of heterochromatin by histone H3K9 trimethylation and subsequent binding of heterochromatin protein 1 (HepA) represents an important regulatory mechanism for secondary metabolite gene expression in A. nidulans. Entry of the culture into the stationary growth phase, known to activate SM, generates a signal which leads to upregulation of the general SM regulatory gene LaeA and this protein is subsequently required to turn on transcription of the pathway-specific AflR activator (Bok and Keller, 2004; Bok et al., 2006a). The activation of AflR is a crucial step for the regulation of the ST cluster as it was shown previously that forced ectopic expression of this regulator rendered the ST cluster independent of LaeA. Our data demonstrate that LaeA functions at two levels in aflR gene activation. First, LaeA counteracts the establishment of heterochromatin domains at the aflR promoter and possibly also in other genes in the ST cluster (shown here only for stcO and stcU). Second, LaeA also is required for the transcriptional activation step of aflR because the absence of heterochromatin marks in strains carrying a hepA or a clrD deletion is not sufficient to turn on aflR expression under conditions of primary metabolism (24 h cultures) where laeA is expressed at low levels (compare Fig. S3). Transcriptional activation is often associated with the placement of acetylation marks on chromatin and transcription factors are known to recruit histone acetylation complexes (Li et al., 2007). From our studies it is unlikely that the activating function of LaeA is associated with the H3K9K14 acetylation pathway because these marks are only marginally reduced in the laeAΔ strain and even strongly elevated in strains carrying laeA deletions in combination with hepAΔ or clrDΔ. However, we only have tested a very limited set of activating acetylation marks on histones in this study and a broader analysis of activating histones marks, such as other acetylation marks in histones H3 and H4 or methylation of H3K4, will be required to further characterize the pathway through which activation by LaeA operates.
It is worth noting that both heterochromatic marks, H3K9me3 and HepA occupancy, are strongly elevated already under conditions of primary metabolism in the laeAΔ strain. This indicates that LaeA function is not restricted to the SM phase but generally functions in balancing heterochromatin and euchromatin establishment in the ST cluster. This raises the possibility that specific de-methylases, which must operate to reduce the H3K9me3 level under conditions of SM in the wild type, do not function in the laeAΔ strain. It remains an open question how LaeA prevents hyper-methylation. It could function by stimulating histone H3K9 de-methylases, or by inhibiting the ClrD H3K9 methyltransferase, or both. The presence of a necessary SAM-binding motif in LaeA is no doubt suggestive, but biochemical work will be necessary to mechanistically understand LaeA function and which is the causal link with H3K9 de-methylation.
In conclusion, reversal of heterochromatic marks mediated by LaeA during the onset of SM is required, but is not sufficient, to transcriptionally activate aflR. Interestingly, in cultures grown for 5 days on solid GMM expression of genes within the ST cluster become largely independent from LaeA because ST or NOR levels are only reduced for about 50% in the laeAΔ strains, depending on the genetic background (Figs 2B and 4B and Fig. S5). These results suggest that prolonged growth on solid medium creates an activation signal for another, LaeA-independent pathway. Because under these conditions we see full restoration and even overproduction of ST and NOR in a laeAΔ background by deletion of either hepA or clrD, it is intuitive to propose that also this alternative, LaeA-independent, pathway employs reversal of heterochromatin during the activation process. However, ChIP analysis of 5 day grown liquid batch cultures was impossible due to DNA degradation (data not shown) and it was not within the scope of this study to establish ChIP conditions for cultures grown on solid medium to be able to test this hypothesis.
H3K9me3 marks and HP1 binding have both been associated with the formation of pericentric heterochromatin and gene repression at euchromatic regions. Available data argue that HP1 can trigger repressive chromatin structure once it is targeted to specific promoters in euchromatin and that the resulting chromatin structure resembles heterochromatin (Hediger and Gasser, 2006). The genes located outside of the ST cluster (e.g. gene AN7826.3 and AN7801.3) are not subject to de-repression in the hepAΔ and clrDΔ mutants nor are their repressive chromatin marks (H3K9me3 and HepA occupancy) reversed during ST cluster activation. We noticed that the expression level of these genes is sufficiently high to be detected under the conditions of our experimental set-up and thus the high constitutive levels of H3K9me3 and HepA bound to the promoters of these genes do not prevent gene expression. At the moment we have no information on the function of the putatively encoded proteins. However, the specific transcription and chromatin mark profiles argue for a boundary which insulates the ST, and by implication, other gene clusters, from the neighbouring chromosomal domains. Subtelomeric repeat regions present both telomere distal and proximal to the ST cluster might fulfil this insulation function and may in turn, be constitutively heterochromatic.
Chromatin immunoprecipitation analysis of hepA deletion strains showed decreased levels of H3K9 trimethylation, but neither N. crassa hpo mutants (Freitag et al., 2004) nor S. pombe Swi6 mutants (Nakayama et al., 2001) show diminished H3K9me levels. However, HP1 protein interaction partners were shown to include Drosophila H3K9 methyltransferases (Aagaard et al., 1999; Yamamoto and Sonoda, 2003) and fly as well as mammalian histone deacetylases (Czermin et al., 2001; Vaute et al., 2002). Our findings could therefore be explained if stable maintenance and spreading of repressive chromatin at the ST locus required interaction between HepA and ClrD and only in the presence of HepA recruitment of more H3K9 methylase activity through a putative HepA/ClrD/HDAC interaction would occur. We find rather surprising that neither deletion of hepA or clrD show any obvious phenotypic effects. This is in contrast to the crucial functions these genes play in other organisms in maintaining chromosome configuration and stability and the integrity of centromeres and telomeres (Allshire et al., 1995; Ekwall et al., 1995; Kellum and Alberts, 1995; Fanti et al., 1998; Perrini et al., 2004). In N. crassa and in S. pombe, loss of the single HP1 orthologues led to pronounced growth defects and chromosome loss respectively (Allshire et al., 1995; Freitag et al., 2004). For Aspergillus fumigatus, Palmer and coworkers reported that disruption of AfclrD results in several growth abnormalities, including reduction in radial growth and conidial production (Palmer et al., 2008). In A. nidulans, no reduction in specific growth rates of the hepA and clrD mutants is seen and hence the effect on secondary metabolite gene expression cannot be attributed to general stress symptoms, which have been implicated in triggering secondary metabolite gene expression in various fungi (Yu and Keller, 2005). Because the majority of the filamentous ascomycetes are predicted to contain single HepA/ClrD orthologues, it is tempting to speculate that coordinated silencing secondary metabolite gene clusters during the active growth phase and reversal of the heterochromatic marks by the conserved LaeA proteins is a widespread epigenetic phenomenon in fungi.
Table S1 lists all fungal strains used in this study. A. nidulans strains were grown and manipulated according Pontecorvo et al. (1953). Some strains are not discussed in the text but were used for sexual crosses to obtain the strains of interest. All strains were maintained as glycerol stocks and were grown at 37°C on glucose minimal medium with appropriate supplements (GMM). Transformations were performed as described previously (Tilburn et al., 1983).
All cassettes were constructed by DJ-PCR (Yu et al., 2004) and Table S2 lists all oligonucleotides used for this purpose. Deletion of the hepA and clrD open reading frames were carried out by replacing each of them with the A. fumigatus pyrG gene (pyrGf) and details of knock-out construct generation as well as generation of hepA and clrD complemented strains are given in Supporting information.
Standard techniques were used to analyse and manipulate RNA and DNA. To characterize deletion strains and complementation events in these strains, we used analytical PCR and Southern blotting. Restriction enzyme digests and probes used for Southern analysis, RNA extraction protocols and probes for Northern analysis are described in detail in Supporting information. Growth conditions for expression analysis are described in the legends to figures.
Thin-layer chromatography (TLC) analysis was applied for ST and NOR production of A. nidulans wild-type and mutant strains from either 24 or 48 h liquid shake cultures on GMM or on GMM solid media cultures using published procedures (Bok and Keller, 2004). Details for the SM analysis carried out during this work are given in Supporting information.
Standard deviation was calculated from at least three independent experiments. In TLC analysis of the wild type, where the level was arbitrarily set to 100%, the relative SDrel was calculated proportionally to the absolute densitometry value aV, SDrel = [(100/aV) × SDaV]. For statistical analyses of metabolite production, a probability of type I error of less than 0.01 was considered statistically significant. Where only two treatments were compared, significance of variation was determined using a Student's t-test. Where more than two treatments were compared, analysis of variance was used to determine significance of overall variability among treatments, followed by a Neumann–Keuls test to compare individual pairs of treatments. The Microsoft Excel data analysis package was used to perform analysis of variance and t-tests. Neumann–Keuls tests were performed by hand.
Chromatin immunoprecipitation was carried out as described previously (Bernreiter et al., 2007). Antibodies used for ChIP were purchased from Abcam, Cambridge, UK (rabbit polyclonal to histone H3 trimethyl K9, ab 8898, rabbit polyclonal to HP1, ab24726 and rabbit polyclonal to human C-terminus histone H3 antibody, ab1791) and from Upstate-Millipore, MA, USA, Anti-acetyl K9-K14 histone H3, 06-599. Two micrograms of antibody was used per reaction of 200 µg total protein. To detect the background signal of the immunoprecipitation reaction while using anti-HP1 antibody a hepA-deleted strain was analysed in parallel.
Amplification and detection of precipitated DNA in real-time qPCR was performed with Platinum SYBR Green qPCR SuperMix UDG (Invitrogen) following the instructions of the provider. The relative amounts of DNA were calculated by dividing the immunoprecipitated DNA by the input DNA. Each PCR reaction was replicated. To normalize the amount of DNA precipitated with histone H3-trimethyl K9 and H3-acetyl, the resulting ratio of the precipitation with these antibodies was divided by the resulting ratio of the anti-C-terminus histone H3 precipitation. Two biological repeats were performed for each condition and standard deviation was calculated upon these. Primers aflRpromfwd and aflRpromrev were used to amplify aflR promoter region, stcOpF and stcOpR were used to amplify stcO promoter region, AN7801pF and AN7801pR were used to amplify AN7801 promoter region, niiA_nuc-1_F and niiA_nuc-1_R were used to amplify niiA nucleosome 1 region.
The authors are grateful to Markus Markl and Alberto Pagani (Eurofins-ofi, Vienna) for technical assistance in metabolite analysis and Sotiris Amillis (University of Athens, Greece) for critical reading of the manuscript. Work in Vienna was supported by the Austrian Science Fund (FWF) Grants P19732-B11 and F 3703-B11 (Special Research Area SRA) to J.S. Work at Orsay was supported by the CNRS, the Université Paris-Sud, and the Institut Universitaire de France. Y.R.-D. was supported by a studentship of the Ministère de l'Education Supérieure et de la Recherche and the CONACYT (México). Funding has been provided for this research and publication from the USDA Cooperative State Research, Education and Extension Service (CSREES) project WIS049621, NSF MCB-0236393 and NIH 1 R01 Al065728-01 to NPK.
Additional supporting information may be found in the online version of this article.
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.