Introduction

Filamentous fungi play very important roles in nature [100]. They form part of the ecological systems for degradation of plant and animal decaying material. On the other hand fungi contribute to the maturation processes of cheese and other fermented products for the food industry and play a key role in the pharmaceutical industry as producers of antibiotics, antitumor agents, anticholesterolemics and immunomodulators [24, 66, 101].

An outstanding group of secondary metabolites produced by a few fungal species are the β-lactam antibiotics [1]. In this article we focus our attention on the characterization of the Velvet complex of Penicillium chrysogenum and its role on the biosynthesis of penicillin and PR-toxin, two metabolites of this fungus which are known to be regulated by LaeA and other components of the velvet complex.

P. chrysogenum is an excellent model system to study the molecular mechanisms of control of secondary metabolites biosynthesis because of the availability of its complete genome sequence and detailed proteome information [44, 45, 84, 91].

The penicillin biosynthetic pathway and the penicillin gene cluster

The penicillin biosynthetic pathway starts in the cell cytosol with the non-ribosomal formation of the tripeptide (l-α-aminoadipyl)-l-cysteinyl-d-valine (ACV) from l-α-aminoadipic acid, l-cysteine and l-valine. This first step is catalyzed by the ACV synthetase [99], encoded by the pcbAB gene [26, 63]. In the second step the tripeptide LLD-ACV, in which the valine has the D-configuration, is converted to isopenicillin N (IPN) by the IPN synthase [81], encoded by the pcbC gene. IPN is the first bioactive β-lactam compound of the pathway. Then, IPN enters the peroxisome [72], presumably by an MFS transporter [30], and the l-α-aminoadipyl side chain is replaced by a hydrophobic side chain in a reaction catalyzed by the IPN acyltransferase (IAT), encoded by the penDE gene [2, 4, 65]. The IAT is located inside the peroxisomal matrix [35, 72]. This transacylation reaction requires previous activation of the side chain that is achieved by specific peroxisomal aryl-CoA or acyl-CoA synthetases [54, 57, 94]. In recent years considerable advances have been made on our knowledge of the enzymology and the molecular mechanisms of the four penicillin biosynthetic enzymes [reviewed by 69].

In P. chrysogenum, Penicillium nalgiovense, Aspergillus nidulans and Aspergillus oryzae, three of the genes responsible for penicillin biosynthesis (pcbAB, pcbC and penDE) are clustered with other ORFs forming an amplifiable 56.8 kb DNA unit [32, 90], which is present in several copies in high penicillin producing strains [31, 32, 38, 90]. Two of the penicillin biosynthetic genes, pcbAB and pcbC, are transcribed from divergent (bidirectional) promoter regions. The control of expression of genes from promoter regions containing divergent promoters has been associated with reorganization of the heterochromatin structure that allows easier interaction of those divergent promoter regions with the RNA polymerase and the transcriptional factors [34, 43].

The accumulated evidence in several filamentous fungi shows that the regulation of many pathways for secondary metabolites, such as antibiotics, is frequently associated with cellular differentiation processes [15, 98]. Several studies have provided insights of other factors that link production of secondary metabolites and morphological differentiation processes [52].

Pathway-specific and global regulators

The biosynthesis of penicillin and some other secondary metabolites is regulated by global regulators that integrate signals for response to nitrogen, carbon, and phosphate limitation. In some secondary metabolite gene clusters, there is a specific regulator that controls the adjacent genes, e.g., AflR in the aflatoxin gene cluster in Aspergillus parasiticus [19] and A. nidulans [29]. However, in many other cases there are no specific cluster-situated regulators and it is unclear how the expression of these clusters is controlled. Furthermore, evidence for regulators located in a cluster that control the expression of a different cluster has been reported recently [9].

Regulation of those clusters of secondary metabolites that lack cluster-situated specific regulators, seems to be mediated by global regulators that influence many genes, such as PacC, NreA, CreaA, CPCR1 and FKH1, among others [21, 27, 28].

Two of these global regulators are LaeA and VelA, commonly designated as core components of the velvet complex. This multicomponent complex was initially identified in A. nidulans and Aspergillus fumigatus. This protein complex affects heterochromatin rearrangements and, therefore, the study of the velvet complex has great importance for understanding the expression of genes encoding secondary metabolite gene clusters.

The model velvet complex in Aspergillus nidulans

The core of the velvet complex in A. nidulans is composed of three subunits VelA (initially name VeA), VelB and LaeA [5]. This complex is involved in the stress signaling cascade and affects production of several secondary metabolites and also participates in the fungal response to light. Of these three components LaeA plays a key role in the regulation of secondary metabolism through heterochromatin reorganization [11]. An even more complex regulation network was established when the multisubunit velvet complex was shown to link the differentiation and secondary metabolism processes [5, 8, 16, 51, 55, 80, 86]. The founding member of this complex was VelA [48]. Under dark conditions, this protein forms a heterotrimeric complex with VelB, and the global regulator of secondary metabolism LaeA [5, 10]. A third velvet protein, VosA, also interacts in the dark with VelB, and it is proposed that the heterodimer VosA-VelB represses asexual spore formation and controls spore maturation and trehalose biosynthesis [6, 74]. Shuffling of VelB between the VelB-VosA dimer and VelA-LaeA complexes is controlled by the LaeA protein, which therefore, plays a key role in the dynamics of complexes regulating secondary metabolism and development in A. nidulans [6]. In contrast, the function of VelC, the fourth member of the velvet superfamily, is still unclear [7]. The LaeA and other components of the Velvet complex are conserved in different fungi although there are still important differences in the percentage of conserved amino acids (Table 1) [15]. The sequence differences suggest a significant degree of plasticity that was confirmed in heterologous complementation experiments.

Table 1 Genes of P. chrysogenum encoding LaeA and the core components of the velvet complex
Full size table

The P. chrysogenum laeA gene

The laeA gene of P. chrysogenum (Pc-laeA) was first isolated and characterized by Kosalková et al. [56]. It encodes a protein with 61 % amino acid identity to the orthologous protein of A. nidulans. The laeA gene is expressed well in P. chrysogenum during the growth phase in cultures in complex media under conditions of penicillin production. Transcripts of laeA are already present in batch cultures at 24 and 48 h of cultivation. This gene contains an intron (nucleotide positions 390–445) as shown by RT–PCR studies. The deduced LaeA protein sequence contains an S-adenosylmethionine binding site and the protein has been proposed to have methyltransferase activity [40, 56].

The laeA gene is present as a single copy both in the wild type P. chrysogenum NRRL 1951 strain and in improved penicillin production strains P. chrysogenum Wis 54-1255 and ASP-78. Unlike the penicillin gene cluster (pcbAB-pcbC-penDE) which is amplified in tandem in a 56.8 kb region in the high production strains [31, 32, 73], the laeA gene is located outside of the amplifiable region and is present in an identical restriction fragment in all the strains tested [56], indicating that this chromosomal region has not been reorganized during strain improvement.

Overexpression of the laeA gene in P. chrysogenum was achieved by locating the laeA gene under the control of the Aspergillus awamori gdh promoter. Interestingly, overexpression of this gene resulted in a fourfold increase in the transcript level of the pcbC and penDE genes in the transformed strain (Fig. 1). This led to an increased production of 20 % in penicillin as compared to the parental strain.

Fig. 1
figure 1

Transcription studies in a laeA mutant of P. chrysogenum. Transcription of the penicillin biosynthesis genes pcbC and penDE and the roquefortine dmaW gene of P. chrysogenum in the parental Wis54-1255 strain (white bars; control, C), the laeA overexpressing strain (gray bars, overexpressed OE), and the laeA silenced mutant (black bars, silenced S). Data are standardized with respect to the expression of the actin gene expression in the case of pcbC and penDE and in relation to lys1 in the case of dmaW [56]

Full size image

Studies on the effect of laeA inactivation on pigment production, penicillin biosynthesis and differentiation were performed using a mutant silenced in laeA. For this purpose a transformant containing the RNAi interference system carrying an exonic fragment of laeA was constructed [89]. Northern analysis revealed that expression of the laeA gene in this silenced transformant was drastically reduced. The silenced transformant lacks the typical green pigment of P. chrysogenum and also it lacks the brown pigment that is observed on the reverse side of colonies growing on solid medium.

There was a reduction of about 50 % in the sporulation of the silenced transformants indicating that laeA has also an important role in the control of asexual differentiation.

The silenced transformant showed a drastically reduced production of penicillin, 59 and 47 % production with respect to the parental strain at 72 and 96 h of cultivation [56]. Interestingly, there was no reduction of the production of roquefortine C, another metabolite produced in small amounts by P. chrysogenum [36]. It has been reported that in A. nidulans different secondary metabolites are regulated either positively or negatively by LaeA to a different extent. Therefore, it is possible that the lack of regulation of the roquefortine biosynthesis is due to the low level expression of the roquefortine C gene cluster, probably due to its localization site in the chromosome. Differences in expression have been associated with the localization of the secondary metabolite gene cluster in the genome [77].

The distinct effect of LaeA on penicillin and roquefortine C biosynthesis was confirmed by transcriptional studies using probes internal to the pcbC and penDE genes that encode two enzymes of the penicillin biosynthesis pathway, and also the rpt (synonimus dmaW) gene that encodes the dimethylallyltriptophan synthase, the first enzyme of the roquefortine C biosynthesis [36, 67].

The transcriptional studies in the laeA silenced mutant showed differences in the effect of LaeA on the early and late genes of the pathway: expression of the pcbC gene is drastically reduced, by a factor of fivefold at 48 h, whereas the expression of penDE is reduced in the order of twofold (Fig. 1). Indeed, expression of penDE occurs from a different promoter regulated by a different set of transcriptional factors [64]. Other secondary metabolites, still not identified chemically, were also regulated by LaeA as shown by HPLC studies. Some of these are regulated in a positive manner, whereas others are down regulated in the laeA mutant [56]. Some of those unknown yellow secondary metabolites have been identified recently as members of the sorbecillinoids family [84].

Genes for primary metabolism do not seem to be significantly affected by LaeA. The transcription of lys1 gene, that encodes homocitrate synthase, the first enzyme of the lysine pathway, was tested in the laeA mutant as compared to the parental strain. There was no difference of expression in the laeA mutant. A likely explanation is that primary metabolism genes do not require the same transcriptional factors as do the secondary metabolism, which interact with LaeA.

Effect of laeA and velA in the high producing strain P. chrysogenum P2

Following the work of Kosalková et al. [56] on the laeA gene other researchers proceeded with the identification of additional components of P. chrysogenum velvet complex. Hoff et al. [40] used in those studies a P2niaD strain, derived from the P2 Panlabs strain, which is a high penicillin producing strain with 5–6 copies of the amplified 56.8 kb DNA region (containing the penicillin gene cluster and ancillary biosynthetic genes). For this purpose they cloned the P. chrysogenum velA gene by similarity with the orthologous velA gene of A. nidulans described previously [51]. The P. chrysogenum velA is similar to the A. nidulans orthologue, but lacks certain domains, for example the NLS (nuclear localization sequence) domain is not clearly present in this particular protein. Mutants disrupted in the velA gene were obtained and tested for differences in penicillin production. The velA mutant showed a reduced production of penicillin, as occurs also with the laeA mutant, indicating that both, VelA and LaeA, are positive regulators of penicillin biosynthesis.

The genes pcbC and penDE were drastically reduced in their expression in the velA mutant [40]. Pairwise comparisons of gene expression in the P2niaD parental strain and the velA mutant revealed that several genes of primary metabolism were also affected to a different extent; the main differences were observed in genes related to cell wall precursor biosynthesis and differentiation, particularly to spore formation. Regarding secondary metabolism, other than penicillin biosynthesis, there were some genes that were overexpressed in the velA mutant while others were downregulated; this means that velA has also an important role in regulation of secondary metabolism, either repressing or enhancing gene expression of these secondary metabolism genes.

Hoff and coworkers [40] also studied the role of laeA in the P2niaD strain. This gene was 100 % identical to that cloned by Kosalková et al. [56] in P. chrysogenum Wis54-1255. Mutants in laeA were obtained in P. chrysogenum P2niaD by standard gene replacement procedures. These mutants were totally defective in LaeA, unlike from the silenced mutants of Kosalková et al. [56]; their penicillin production level was drastically reduced from 5 g/l to about 0.5 g/l, i.e., a reduction of 90 % in the penicillin titre. These results confirm the previous observation obtained with the silenced mutant. In conclusion, both laeA and velA act as positive regulators of penicillin biosynthesis. In addition, complementation of the deleted laeA mutant with an intact copy of the wild type restored penicillin production to normal level.

Effect of VelA on sporulation and differentiation

It is known that velA in A. nidulans has an effect on sporulation but this effect is distinct in different fungi. In P. chrysogenum the comparison of the parental strain with the velA mutant revealed that, in the parental strain the conidiation is light-dependent whereas in the velA mutant the number of spores is similar in both dark and light conditions. On the other hand the laeA mutant of P. chrysogenum P2niaD shows a very clear reduction in conidiation. Microscopical observation revealed that the parental strain P. chrysogenum P2niaD forms chains of spores, whereas the laeA mutant forms only one spore in each phialide, which means that the reduction of conidiation affects the formation of conidia but not the number of conidiophores.

Heterologous complementation studies: LaeA and VelA plasticity

As mentioned above there are significant differences in the amino acid sequences of both the LaeA and the VelA proteins in different fungi. However, complementation of P. chrysogenum mutants using the Fusarium fujikuroi laeA and velA genes, showed that there is a heterologous complementation of the effect of the disruption of these genes on production of penicillin [40].

The conservation of amino acids between P. chrysogenum and F. fujikuroi LaeA proteins is of 33 % and that of VelA is of 38 %. Heterologous complementation suggests that despite the differences in the amino acid sequences the catalytic sites (e.g. methyltransferase) or, protein-to-protein interacting motifs of these proteins appear to be functional in heterologous multicomponent complexes.

Surprisingly, in contrast to the previous work [40, 56, 62] on the important role of LaeA and VelA, Veiga et al. [93] described in another high producing strain derived from the DSM culture collection that the deletion of velA and laeA genes does not have too much effect on penicillin production. The reason is that these authors performed their experiments under glucose limited continuos culture conditions, and therefore, these results suggest that the LaeA and VelA effect is greatly influenced by the culture conditions, particularly by glucose limitation. These results point to a possible involvement of the CreA regulator in the LaeA mediated control of gene expression.

Additional components of the velvet complex

Recently, Kopke et al. [55] described the characterization of three additional components of the velvet complex in P. chrysogenum, namely PcVelB, PcVelC, and PcVosA. Using yeast two-hybrid analysis and bimolecular fluorescence complementation these authors demonstrated that all velvet proteins are part of an interacting-protein network.

Studies using single and double disrupted mutants indicate that several velvet subunits have opposing roles in the regulation of penicillin biosynthesis. LaeA was shown to act as a positive regulator, in agreement with the initial observation of Kosalková et al. [56], while PcVelB represses this biosynthetic pathway [55]. In addition, PcVelB and PcVosA promote conidiation, while PcVelC has an inhibitory effect. Further genetic analyses showed that light-dependent spore formation depends not only on PcVelA but also on PcVelB and PcVosA.

Novel insight in the velvet complex of P. chrysogenum

Effect of diamines on laeA expression and penicillin biosynthesis

The biosynthesis of fungal secondary metabolites is frequently induced by plant-derived external elicitors, e.g., alginate in P. chrysogenum [37], that form part of an attack/defense cross-talk between plant and fungi. Secondary metabolism appears to be also regulated by endogenous inducers, which may work in a way similar to that of bacterial autoinducers.

Several autoinducer molecules have been studied in relation to the formation of secondary metabolites; jasmonic acid in plants and some basidiomycetes secondary metabolites [78] and conidiogenone in Penicillium cyclopeum [82].

A few years ago Jorge Martín and coworkers [62] described that an autoinducer isolated from the culture broth of P. chrysogenum and Acremonium chrysogenum was able to stimulate the biosynthesis of penicillin. This autoinducer was purified and identified by NMR and mass spectrometry as 1, 3-diaminopropane and the genes for the biosynthesis of 1, 3-diaminopropane were identified in P. chrysogenum genome. Further studies showed that 1,3 diaminopropane and spermidine are able to activate the expression of the penicillin biosynthesis genes pcbC and penDE, and therefore, the effect of these diamines is due to an increase of the transcription of these genes or to the increase of the stability of the messenger RNAs. In P. chrysogenum no inducing effect was exerted by γ-butyrolactone, jasmonic acid, or the penicillin precursor δ-(l-α-aminoadipyl)-l-cysteinyl-d-valine.

It is unclear if the effect of the diaminopropane autoinducer is specifically related to the transcription of a single secondary metabolite gene cluster or is a more general phenomenon related to heterochromatin-associated expression of genes for many secondary metabolites.

Diaminopropane induced transcription of the pcbAB, pcbC, and penDE genes and penicillin biosynthesis when added at inoculation time, but its effect was smaller if added at 12 h and it had no effect when added at 24 h, as shown by Northern analysis and lacZ reporter quantification of promoter activity. Addition of pure 1, 3-diaminopropane stimulated the production of penicillin by about 100 % as compared to the titres of the control culture [62].

Interestingly, our studies showed a clear control of laeA gene expression by the diamines 1,3 diaminopropane and spermidine [62]; this means that the transcript level of laeA, and therefore of the penicillin biosynthetic genes, is under the control of these diamines. Furthermore, the diamines reverse the effect of laeA silencing in P. chrysogenum. It is not entirely clear how the diamines revert the effect of laeA silencing but it is known that the amino groups of 1,3-diaminopropane and other diamines are charged positively and interact with nucleic acids in the cells [83, 92]. Polyamines are known to modulate binding of certain proteins (e.g., estrogen receptors) to Z-DNA-forming regions [88]. This raises the question of whether there are DNA structures similar to Z-DNA in the promoters of some secondary metabolite genes. Polyamines alter the specific DNA–protein interactions, particularly that of heat shock proteins and activator proteins induced by heat shock. Therefore, it is likely that 1, 3-diaminopropane modulates the expression of penicillin biosynthesis genes by interacting with the DNA and other activator proteins in the promoters of the pcbAB, pcbC, and penDE genes.

Control of laeA gene expression by other regulators has been described. AflR, a regulator of the Zn2-Cys6 class, located in the sterigmatocystin gene cluster, represses expression of the laeA gene and subsequently the biosynthesis of sterigmatocystin, penicillin and lovastatin [10]. In conclusion, certain transcriptional factors and/or autoinducers regulate expression of laeA in different fungi and, therefore, subsequentially control secondary metabolites biosynthesis.

Changes in the Velvet complex during the strain improvement programs

Interestingly, there were some significant differences in the nucleotide sequences of the velA and laeA genes of the high producing strain of DSM as compared to other P. chrysogenum strains, resulting in amino acid differences of the encoded proteins. For example, the absence in the VelA protein of a sequence for protein degradation, and also changes in the S-adenosylmethionine binding site of LaeA. Since the velvet complex is a multisubunit structure, the impact of strain improvement programs on the core component of the Velvet complex were studied comparing the wild type P. chrysogenum NRRL1951 with the P. chrysogenum orthologs of relevant genes in high producing strains: LreA, LreB, FphA, KapA, VelB, and VosA were all identical in the genomes of strains NRRL1951 (wild type), Wisconsin 54-1255 and the high producing strain DS17690. However, important mutations were found in Pc16g14010 and Pc13g13200 genes that encode LaeA and VelA, respectively [84].

During the strain improvement programs the laeA gene acquired a nucleotide substitution at position 850 (T to C) which resulted in a substitution of a lysine for glutamic acid (Lys284Glu). A second mutation was introduced during development of the high producing strain DS17690 and caused a Gly338Ser substitution. In addition, a non-sense mutation was found in the velA gene; a C to T substitution at nucleotide position 943 formed a stop codon resulting in the formation of a truncated VelA protein (from 562 to 315 amino acids) [84]. These results indicate that the velvet complex has been drastically mutated in the strain improvement programs, and this likely has impacted the expression of several secondary metabolite encoding genes. This explains the lack of yellow `pigment in the penicillin overproducing strain and changes in other secondary metabolites. Surprisingly, when the transcriptomes in laeA and velA disrupted mutants of P. chrysogenum DS17690 were compared with the transcriptome of the reference strain, only 23 genes were differentially transcribed between the reference strain and both laeA and velA mutants. Eleven of these 23 genes, located in two different clusters, are downregulated in both mutants and appear to encode proteins related to secondary metabolites. One of these set of genes encode proteins of unknown function and the second set, corresponding to genes Pc12g06310 to Pc12g06370, includes a gene for aristolochene synthase [93]. Aristolochene is a precursor of the PR-toxin (see below) and the entire pathway for biosynthesis of the PR-toxin of P. chrysogenum has been identified recently [39]. It is important to note that the genes known to be expressed differentially in the laeA and velA mutants of DS17690 strain, correspond exactly to the PR-toxin cluster (Fig. 2).

Fig. 2
figure 2

PR-toxin genes in P. chrysogenum strains. Comparative alignment of: a the gene cluster (Pc12g06310 to Pc12g06370) described to be under regulated in the laeA knock-out mutant derived from the high producing strain DS17690 [93] and, b the PR-toxin gene cluster (prx1 to prx11) of P. chrysogenum Wis54-1255 [39], where Pc12g06290 corresponds to a pseudogene. Note that the genes in A and B are the same and corresponds to those of the PR-toxin gene cluster. The prx2 gene encodes the aristolochene synthase

Full size image

The biosynthesis of PR-toxin in P. chrysogenum: expression of a near-silent gene cluster is up-regulated by LaeA and VelA

The PR-toxin is a bicyclic sesquiterpene belonging to the eremophilane terpenoid class [95]. The PR-toxin molecule, containing 17 carbon atoms derives from three molecules of isopentenyldiphosphate and an acetyl group. Four other compounds related to PR-toxin named eremofortines A,B,C and D, were proposed to be intermediates in the biosynthesis of PR-toxin [71] and this was supported by recent evidence on the enzymes encoded by the PR-toxin gene cluster [39].

Aristolochene is a non-oxygenated 15-carbon molecule that is formed by direct cyclization of farnesyl-diphosphate. The aristolochene synthase was isolated from P. roqueforti and it was shown to convert in vitro farnesyl-diphosphate to aristolochene [41, 79]. The aristolochene synthase was later purified from A. terreus [17]. The ari1 gene encoding this enzyme was initially cloned from P. roqueforti [17, 18, 79].

Until recently it was unknown whether a PR-toxin gene cluster exists in P. chrysogenum, since this fungus does not produce a significant amount of PR-toxin, although might synthesize traces of it [33]. The silencing of the PR-toxin (prx) genes decreases PR-toxin biosynthesis in P. roqueforti but resulted in a large overproduction of mycophenolic acid, an antitumor compound synthesized by a different pathway [23], suggesting a cross-talk of PR-toxin and mycophenolic acid pathways.

The P. chrysogenum eleven gene cluster (Pc12g0_6260 to Pc12g0_6370) that includes the above mentioned prx genes and a 14-TMS drug/H+ antiporter, has been reported to be very poorly expressed in P. chrysogenum under conditions of penicillin production (strongly aerated cultures) [91], i.e., it behaves as a near-silent gene cluster [68]. We found that this near-silent gene cluster is able to produce PR-toxin in P. chrysogenum under static culture conditions on hydrated rice medium. Of particular note, the production of PR-toxin was 2.6-fold higher in P. chrysogenum npe10, a penicillin nonproducing strain deleted in the 56.8 kb amplifiable region that includes the pen gene cluster, than in the Wisconsin 54-1255 parental strain from which the npe10 mutant strain derives [39]. Similar results were reported by Harris et al. [38] using a mutant deleted only in the three penicillin biosynthetic genes, derived from the high penicillin producing strain DS17690. To further investigate this cross-talk phenomenon we used a P. chrysogenum mutant lacking only the IPN synthase (pcbC gene). This mutant does not produce penicillin but again produces increased levels of PR-toxin. These results provide another example of cross-talk between secondary metabolite pathways in this fungus.

Cross-regulation of secondary metabolite biosynthetic pathways is well known in Streptomyces [60, 70, 94] but only a few examples have been described in filamentous fungi [9, 68].

The cross-regulation of secondary metabolites opens a new interesting approach to improve biosynthesis of secondary metabolites in filamentous fungi. Deletion or silencing of some specific gene clusters or more specifically of some positive or negative regulatory genes may improve the biosynthesis of other secondary metabolites.

A wealth of secondary metabolites in different fungi are regulated by LaeA and the velvet complex

In this review we have concentrated in the analysis of the regulation of penicillin and PR-toxin biosynthesis by LaeA and other components of the velvet complex in P. chrysogenum. In addition, in the last few years several other filamentous fungi have been studied and the number of fungal metabolites described to be regulated by LaeA has increased greatly (Table 2). These metabolites include secondary metabolites such as mevastatin, cyclopiazonic acid; pigments such as the azaphilones involved in red rice production, primary metabolites such as citric acid, and enzymes such as endoglucanases or cellulases. Complex developmental processes and plant/animal pathogenicity are also regulated by LaeA. These complex differentiation or pathogenicity effects are the result of integration of the regulatory effect of LaeA and the velvet components on different biosynthetic processes.

Table 2 Examples of secondary metabolites, primary metabolites and biological processes regulated by LaeA and the velvet complex in a variety of filamentous fungi
Full size table