Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

The groundwork for modern fungal biotechnology was laid in the beginning of the twentieth century, accompanied by advances in microbiology, biochemistry and fermentation technology. The pioneering works of Jokichi Takamine (production of amylase from koji mold Aspergillus oryzae, 1894), James Currie (development of fungal fermentation for citric acid production, 1917) and Alexander Fleming (discovery of penicillin production by Penicillium notatum, 1928) stimulated scientists to further explore fungal metabolic capacities and, moreover, prompted engineers to develop large-scale and controlled production processes for filamentous fungi. Improvements of fungal capacities to produce metabolites of interest were, however, mainly restricted to classic mutagenesis techniques. The development of recombinant DNA technologies for filamentous fungi, shown for the first time in 1979 for Neurospora crassa, was a milestone in obtaining insights into the molecular basis of product formation and to improve traditional fungal fermentations by genetic engineering.

The industrial relevance of filamentous fungi is based on their high capacity to produce primary and secondary metabolites as well as for secreting proteins, having a wide spectrum of activity such as hydrolases and proteases (Conesa 2001; Punt et al. 2002). Additionally, the fungal glycosylation machinery is capable of providing a more ‘mammalian-like’ glycosylation pattern to proteins compared to the commonly used yeast hosts (Karnaukhova et al. 2007; Nevalainen et al. 2005; Ward et al. 2004), making filamentous fungi very attractive for the production of proteins used in medical applications. Secretion is, in particular, related to fungal morphology as it is thought to take place at the growing fungal tip (Fischer et al. 2008; Khalaj et al. 2001; Torralba et al. 1998). One focus of current research is thus on fungal morphology to improve the secretory capacity of industrially used fungi (Grimm et al. 2005; Meyer et al. 2008; Papagianni et al. 2003).

The following three sections are devoted to three important areas of research of genetic engineering in filamentous fungi, all aiming at the improved application of these organisms in biotechnology.

Section II summarizes current genomic approaches for filamentous fungi and discusses their benefit for the identification of new commercially interesting products.

Section III highlights the progress made in the post-genomic era, concerning new omic techniques as well as related challenges and future perspectives.

Section IV deals with genetic and metabolic engineering tools applicable nowadays in filamentous fungi and highlights the progress made for different transformation techniques and gene knockout/knockdown approaches.

2 Fungal Genomics: Advances in Exploring Sequence Data

Identification of putative new and industrial applicable enzymes or secondary metabolites, e.g. of medical interest, requires evaluation of the genetic potential of a given organism. Sequence data has to be screened for genes coding for desired enzymes involved in the pathway of interest. To meet the demand for sequence information, more and more fungal genomes have been sequenced since the first genome of the yeast model organism Saccharomyces cerevisiae was published (Goffeau et al. 1996).

Generally, genome sequencing projects continue to unravel the genetic capabilities of many organisms, resulting in more than 800 fully sequenced genomes currently being available, 25 of which are fungal genomes. More than 2500 sequencing projects are ongoing, including more than 300 fungal projects (http://www.genomesonline.org/gold_statistics.htm), whereby industrial and medically interesting fungi are mainly in the focus of genome sequencing projects (Table 18.1).

Table 18.1 Genome sequencing status of selected filamentous fungi
Full size table

These genomes have been partially or completely sequenced, or they are currently being re-sequenced. However, even fully sequenced genomes such as for S. cerevisiae are still undergoing correction with regard to sequence details, splicing or improved annotation (Jones 2007). Comparative genomics is a very useful tool to compare less well-studied species with better understood model organisms and to clarify hypothetical genes by computational annotation. Furthermore, they help to identify differences between closely related species as regards pathogenicity, secondary metabolism or other properties (reviewed by Jones 2007). For example, the genome of Trichoderma reesei is considered to be a promising candidate for biofuel production from cellulose fibers. It should eventually become a monosaccharide-producing factory through the use of rational metabolic engineering (Martinez et al. 2008).

‘Gene mining’ as a tool for unravelling new proteins masked in fungal genomes is an approach which focuses on the industrial exploitation of currently unknown proteins and enzymes. For example, over 200 new proteases have been identified during annotation of the A. niger genome, of which two have already been commercialized – a protease preventing chill-haze in beer (Lopez and Edens 2005) and a protease used for the production of sport drinks (Edens et al. 2005).

Detailed analysis of the A. niger genome combined with the reconstruction of the metabolic networks identified many gene duplications, transporters and intra- and extracellular enzymes, providing new insights into the efficient citric acid production by A. niger (Pel et al. 2007). The A. oryzae genome was found to be extremely rich in genes involved in biomass degradation and primary and secondary metabolism (Abe et al. 2006; Kobayashi et al. 2007). New α-amylases and α-glucosidases were found. The genome sequence revealed 134 peptidase genes, in contrast to 18 peptidases hitherto known (Kobayashi et al. 2007). Comparative genomics led to improved functional annotation of the genome of A. nidulans and deeper insights in pathway modelling (David et al. 2008).

Furthermore, the so-called ‘genome-based strain reconstruction’, based on a comparison of high-producing strains with the wild-type strain, led to reconstructed strains superior to former production strains (Adrio and Demain 2006; Ohnishi et al. 2002).

An absolute novelty decrypted by genome data and first metabolic approaches was the identification of (silent) gene clusters for mycotoxin production in fungal strains which have been regarded to be safe for many years. Species such as A. oryzae, A. niger and A. sojae were shown to produce or have the opportunity of mycotoxin synthesis dozing in the genome (Jones 2007; Pel et al. 2007). Dependent on sequencing results, these facts indicate that using these fungi (and other species) could be more risky than initially thought and might thus require new risk assessments (Abe et al. 2006).

‘Metagenomics’ or ‘environmental sequencing’ involves the sequencing of whole ecosystems, independent of specific organisms. This boosts the amount of sequence data available (Raes et al. 2007). Metagenomics can be used to identify and analyze fungal communities (O’Brien et al. 2005), explicitly for chosen proteins (mostly polysaccharide-modifying enzymes, proteases, nitrilases; Blackwood et al. 2007; Streit and Sehmitz 2004), or to sequence fungi which cannot yet be grown in culture and therefore cannot be analyzed by traditional means. Environmental sequencing will play a major role in the decryption of complex communities and synergisms among different organisms in their natural habitats.

‘Functional sequencing’ allows insight into the real natural behavior over and above the limiting conditions of cultivating strains under laboratory conditions and provides excellent opportunities to identify new and as yet unknown proteins and enzymes.

3 Post-Genomic Approaches to Unravel the Metabolism of Filamentous Fungi

New ‘omic’ tools allow fast and easy decryption of fungal genomes and the metabolites produced. These tools can be used to explore new commercially interesting products and to improve metabolic fluxes by rational design. The main characteristic of the so-called ‘post-genomic’ era is the need for methods and tools to analyze the huge amount of data produced. The suffix ‘omics’ was created in the 1990s in the field of bioinformatics and marks the realization of the importance of information processing in biology; omics is a general term for a broad discipline of science and engineering analyzing the interaction of biological information in various ‘omes’. These include the genome, transcriptome, proteome, metabolome, expressome, interactome and many more defined fields, such as glycome, ionome, lipidome or even physiome and reactome.

In most cases, genomics, transcriptomics and proteomics together with metabolomics are used to identify theoretical knockout events or metabolic fluxes in pathways of interest. The main focus is on gathering information for engineering metabolic networks to manipulate the regulatory mechanisms of the entirety of biological processes.

The term ‘systems biology’ is often used in combination with omics and is the ‘biology’ that focuses on complex systems in life. It is a holistic approach which allows analysis of the topology of biochemical and signalling networks involved in cellular responses; in addition it is able to capture the dynamics of the response. While different omics deliver only a piece of the puzzle, it is hoped that systems biology will eventually map the complete picture (Siliang Zhang 2006).

3.1 Transcriptomics

RNA-based technologies have been developed enormously in the past few years, whereby two different strategies have emerged: (i) strategies based on the knowledge of DNA sequences (microarrays) and (ii) strategies which do not require any sequence information, such as suppression subtractive hybridization (SSH) and serial analysis of gene expression (SAGE Breakspear and Momany 2007). The first 50 fungal microarrays were reviewed by Breakspear and Momany (2007), describing the development of fungal array approaches. Interestingly, they highlight the lack of basidiomycete microarray experiments (only three have been performed to date).

Comparative transcriptional analysis of A. oryzae using DNA microarrays indicated the potential of new proteins identified; and it is as a tool which can be used to develop industrial systems (Abe et al. 2006). Anderson et al. (2008b) used a comparative transcriptomic approach where one Gene Chip was developed for transcriptome analysis of triplicate batch cultivations of A. nidulans, A. niger and A. oryzae.

They were able to identify 23 genes conserved across Aspergillus spp. (mainly sugar transporters and enzymes) and 365 genes which were differently expressed in only two of the Aspergilli. Thus, such a cross-species study based on transcriptome analysis can be used for the identification of new enzymes or even strains. In a former publication the authors described the use of transcriptome data to perform metabolic network modelling in A. niger, which is explained in detail in Sect. III.C (Andersen et al. 2008).

SSH and SAGE approaches also helped to understand the infection process of plant pathogen fungi and mycoparasitic fungi. (Carpenter et al. 2005; Gowda et al. 2006; Larraya et al. 2005; Matsumura et al. 2005; Salvianti et al. 2008; Schulze Gronover et al. 2004; Yu et al. 2008). Furthermore, transcriptomic analysis of different culture conditions will help to identify genes involved in product formation and may lead to overproducing strains which are ‘debugged’ with respect to bottlenecks or unnecessary genes (Foreman et al. 2003). Also, insights in energy and primary metabolism have been achieved by transcriptional analysis (Maeda et al. 2004). The method of transcriptional profiling has also been used to reveal previously unknown, but relevant pathways (Rautio et al. 2006; van der Werf 2005).

3.2 Proteomics

Transcriptomic approaches are mostly based on the straightforward development of microarray chips, whereas proteomic fields require analytical techniques such as mass spectrometry (MS) or nuclear magnetic field resonance. The latest review of Aspergillus proteomics provides a summary of all recent information on this technique (Kim et al. 2008).

In 10 different studies using Aspergillus proteomics, a total of 28 cell surface proteins, 102 secreted and 139 intracellular proteins were identified. A proteome map highlighting proteins identified in major metabolic pathway is summarized by Kim et al. (2008).

For Phanerochaete chrysosporium, more than 1100 intracellular and 300 mitochondrial proteins were resolved in 2D gels and the metabolic flux shift was determined by comparative proteomics for growth on vanillin. By using vanillin, which is the decomposition product of lignin, the lignin degradation by white rot fungi can be studied. Many more proteomic approaches have been performed using fungi to identify new products for industrial applications, as recently reviewed by Kim et al. (2007).

3.3 Metabolomics

It is assumed that around 10 000 secondary metabolites may be produced by fungi, examples of which are found in Aspergillus and Penicillium; however, less than 10% are known (Smedsgaard and Nielsen 2005). Much understanding of the central metabolism and regulation in less-studied filamentous fungi can be learned from comparative metabolite profiling and the metabolomics of yeast and filamentous fungi (Smedsgaard and Nielsen 2005). Metabolomics is considered to be more sensitive than transcriptomics or proteomics due to the measuring of metabolite concentrations caused by environmental perturbations that may not affect transcription or protein levels, an example being enzyme activity levels (Oliver et al. 1998). The use of metabolomics means that methods are available that can simultaneously determine over 1000 charged metabolites using capillary electrophoresis-MS (Soga 2007). Even techniques which simulate the whole cell metabolism have been developed (Ishii et al. 2004).

A. nidulans represents an important model organism for studies of cell biology and gene regulation. Kouskoumvekaki et al. (2008) initiated a metabolomics approach in recombinant Aspergilli for clustering and classification. More than 450 detected metabolites were analyzed and resulted in the identification of seven putative biomarkers by which classification into genotype was possible. Thus, metabolite profiling is a powerful tool for the classification of filamentous fungi as well as for the identification of targets for metabolic engineering (Kouskoumvekaki et al. 2008).

A metabolic model integration of the bibliome, genome, metabolome and reactome was constructed for A. niger (Andersen et al. 2008). A complete metabolic network was presented which shows great potential for expanding the use of A. niger as a high-yield production platform. By the multi-omic approach, the use of precursors was identified to increase productivity. In order to combine the knowledge of the underlying metabolic system and the analysis of the data, novel computational approaches and methods that go beyond commonly used statistical techniques are required (Van Dien and Schilling 2006).

Progress has been made using methods for analyzing quantitative metabolomics data in the context of the entire network. Methods based on Gibbs energies of formation, the second law of thermodynamics and on known direction reactions within the cell allow dynamic analysis of metabolites in the cell (Kummel et al. 2006).

Mainly, metabolic flux analysis (MFA) has become a fundamental tool in metabolic engineering to elucidate the metabolic state of the cell and has been applied to various biotechnological processes (Spegel et al. 2007). ‘Conventional MFA’ is based on mass balances using stoichiometric constraints coupled to extracellular product formation rates and is a widely used technique (Vallino and Stephanopoulos 2000).

Applying the full range of omics technologies for strain improvement, it would be desirable if the ‘concept of sample’ was shared among these technologies (in particular, to focus on a biological sample that is prepared for use in a specific omics assay; Morrison et al. 2006). A common data file format should be found to enable fast and secure file sharing and to reduce redundant data and the potential for errors. The main issue as regards combining the analysis of metabolomics data with other omics results is the data integration problem (Mendes 2006).

There is also the call for standards in sample collection and processing to obtain reliable results and to avoid significant error in omics data. It is essential to use the same biological sample when confronted with diverse omic approaches (Martins et al. 2007; Weckwerth et al. 2004).

This has resulted in many omics standardization initiatives aiming at the development of new concepts to overcome the problems of data confusion and sample correlation (Jones et al. 2007; Morrison et al. 2006). Also claimed is the necessity to develop common databases for storage of metabolomics data (Kouskoumvekaki et al. 2008).

4 Metabolic Engineering: Finding the Optimum Genetic Strategy

For successful optimization of a fungal metabolism, several issues of engineering strategy have to be taken into account. Every approach depends on the organism of choice, the target to be engineered and the aim of the intervention. From various genetic engineering tools available nowadays for filamentous fungi, the most suitable has to be chosen in order to specifically and efficiently improve the strain of interest.

4.1 Choosing the Right Transformation Technique

Highly sophisticated genetic methods are available nowadays for filamentous fungi; however, safe and suitable transformation techniques are fundamental prerequisites for genetic engineering approaches. Different transformation techniques enable scientists to design and develop rational metabolic engineering strategies for industrially important fungal species. Linear DNA or plasmids are transferred into the fungal cell, either by chemical treatment of protoplasts using Ca2+ and polyethylene glycol (PEG; Spegel et al. 2007), by physical treatment such as electroporation and biolistic procedures or by using Agrobacterium tumefaciens-mediated transformation (AMT; Casas-Flores et al. 2004; Meyer 2008; Michielse 2005; Ruiz-Diez 2002; reviewed by Fincham 1989).

Historically, PEG-mediated transformation of protoplasts was the first transformation technique established in 1978 for the budding yeast S. cerevisiae (Hinnen et al. 1978). This method was afterwards used in filamentous fungi (Fincham 1989; Punt et al. 1987; Ruiz-Diez 2002; Stahl 1987). Progress in the past few decades was achieved in establishing alternative transformation methods to overcome the limits related to protoplast formation, which is highly dependent on the quality of the cell wall-degrading enzyme preparation and often only leads to low transformation rates (Michielse 2005).

In most cases, asexual spores such as conidia or sporangiospores are used as they are considered to be the most favorable constituent for transformation. However, if these are not available, the whole mycelium has to be transformed, making the procedure more difficult and laborious. In addition, the multinuclear state of some conidia, mycelia and protoplasts hampers the selection of transformants and makes the selection process tedious (Deed 1989; Farina et al. 2004). When comparing the transformation rates of yeast and filamentous fungi, it becomes evident that transformation of yeast is usually much more efficient (Ruiz-Diez 2002).

New systems such as electroporation have been developed. This is actually a method often used for the transformation of filamentous fungi (reviewed by Ruiz-Diez 2002). There are three possible fungal structures that can be transformed by means of electroporation: protoplasts, conidia or young germlings (Chakraborty et al. 1991; Ruiz-Diez 2002).

Biolistic transformation was introduced in 1987 (Klein et al. 1992; Sanford 1987) and has been applied to a number of filamentous fungi (Ruiz-Diez 2002). This method is especially beneficial for those organisms which cannot be transformed by AMT. The use of A. tumefaciens as a mediator of foreign DNA into fungal cells is a tool which has been adapted for many different fungal organisms in the past few years (Michielse 2005). The transformation rate is up to 100–1000 times higher than with protoplastation and can be used for most fungi which cannot be transformed by other methods or where protoplasts do not regenerate sufficiently (Casas-Flores et al. 2004; MacKenzie 2004; Meyer et al. 2003). However, even this method is not suitable for all fungi. There are also reports on less successful attempts, for example in A. niger (Michielse 2005).

AMT and all other transformation techniques cannot be used for every fungal organism without first being adapted. This may be explained by major differences found in fungal genomes such as A. nidulans (Galagan et al. 2005), A. oryzae (Machida et al. 2005) and A. fumigatus (Nierman et al. 2005). Analysis of their genomes shows that there is only 68% of amino acids identity shared by all three species (Galagan et al. 2005), an evolutionary distance comparable to that between human and fish (Dujon et al. 2004; Fedorova et al. 2008). Roughly 70% of A. nidulans genes could be mapped to a syntenic block with either A. fumigatus or A. oryzae, with about 50% of A. nidulans in conserved synteny across all three species (Galagan et al. 2005), suggesting that this could be one reason for the observation that each transformation method has to be adapted and optimized for every single species and even strain.

Apart from transformation approaches, sexual crossing is a useful tool to improve the phenotype of filamentous fungi. Unfortunately, most medically or industrially interesting fungi lack a sexual cycle, precluding them from classic genetics. However, insights into the genome of A. fumigatus and A. oryzae revealed that both are potentially capable for sexual reproduction (Dyer and Paoletti 2005; Galagan et al. 2005; Paoletti et al. 2005), suggesting that classic genetic tools can also be established for these fungi. These findings challenge the whole taxon of fungi imperfecti and even highlight the power of genome analysis for taxonomical (re-)classification.

A very important and interesting question as regards the different transformation systems is the fate of the introduced DNA. DNA which has been introduced will be either maintained autonomously (occurs rarely) or integrated into the genome via homologous or heterologous recombination. Homologous recombination targets the foreign DNA to regions showing sufficient homology, whereby the DNA becomes integrated into the genome as a single or tandem copy. In contrast, heterologous recombination events occur randomly, resulting in single or multi-copy integrations (de Groot et al. 1998; Malonek and Meinhardt 2001; Mullins and Kang 2001).

In the case of AMT, mainly single-copy integration events occur, whereas mainly multi-copy integrations are observed after transformation by PMT. This observation can have an important impact on the choice of a suitable transformation system for the design of a metabolic engineering strategy. AMT can be the method of choice for targeted integration into the genome and PMT can be used for ectopic and multi-copy integration to improve protein expression in the strain of interest (Meyer 2008). The latter strategy has indeed been shown to be a powerful tool for protein overexpression in Aspergillus and Trichoderma (Askolin et al. 2001; Lee et al. 1998; Verdoes JC 1995).

In addition, it was found that AMT can increase homologous recombination frequency, for example in A. awamori (Michielse et al. 2005). This observation pointed to new application opportunities of AMT as a suitable method for directed and insertional mutagenesis (Betts et al. 2007; Lee and Bostock 2006; Michielse 2005; Sugui et al. 2005).

5 Enhancing Gene-Targeting Efficiency

Metabolic engineering can be used to delete genes of unwanted side-pathways, thereby redirecting the metabolic flux to the required product-forming pathway. In addition, targeted integration of genes to a genomic locus known to strengthen transcription is one strategy to enhance protein expression and thereby to improve productivity of an industrial process. Besides, gene targeting is also the method used for functional genomics. The mode of integration of foreign DNA is determined by two competing processes – homologous recombination (HR) and the non-homologous end-joining (NHEJ) pathway (Dudasova et al. 2004; Krogh and Symington 2004). The low rates of HR events usually observed can render some filamentous fungi unattractive for many industrial applications.

However, advances in fungal gene targeting were recently achieved by suppressing the NHEJ pathway (for a review, see Meyer 2008). As summarized in Table 18.2, a dramatic increase in HR efficiency was reported when strains were used in which the NHEJ pathway was inactivated (Goins et al. 2006; Ishibashi et al. 2006; Kooistra et al. 2004; Krappmann et al. 2006; Meyer et al. 2007; Nayak et al. 2006; Ninomiya et al. 2004; Poggeler and Kuck 2006; Takahashi et al. 2006).

Table 18.2. Genetic tools applicable to filamentous fungi for gene targeting and silencing
Full size table

Phenotypic analysis of defective NHEJ strains revealed that these strains showed higher sensitivity to various toxins and irradiation (da Silva Ferreira et al. 2006; Meyer et al. 2007; Ninomiya et al. 2004). Therefore, some unexpected growth behavior could appear, and the more elegant solution would be a transient silencing of the NHEJ pathway, as recently reported for Candida glabrata and A. nidulans (Nielsen et al. 2008; Ueno et al. 2007).

5.1 RNA-Based Tools for Metabolic Engineering

RNA technology is an attractive alternative to DNA-based methods to silence gene expression post-transcriptionally and thereby to control unwanted metabolic pathways. Different RNA techniques such as antisense RNA, RNAi and hammerhead ribozymes have been shown to be valuable tools for filamentous fungi (Table 18.2). These methods provide the advantage of not deleting a gene, thereby bypassing the possibility of lethal or other unwanted effects on the organism.

RNA-based methods are especially valuable when: (i) gene-targeting approaches fail, (ii) multiple copies of a gene of interest are present in the genome or (iii) isogenes might compensate for the knockout of the deleted gene (Akashi et al. 2005). For example, silencing a whole gene family using only a single antisense RNA construct has been described for A. oryzae (Yamada et al. 2007). Another advantage of the RNAi mechanism is its locus independence due to its mediation by a mobile trans-acting signal in the cytoplasm. Consequently, this mechanisms can be used in fungi which have multi-nuclear hyphae or a low targeting efficiency, even in heterokaryotic fungal strains, RNA-based downregulation of genes is possible (de Jong et al. 2006; Nakayashiki 2005).

Antisense RNA gene silencing is performed by using single-stranded RNA which is complementary to an mRNA strand transcribed within the cell. Formation of a complementary mRNA hybrid physically blocks the translation machinery and thereby stops translation of the endogenous mRNA.

Successful gene silencing using artificial antisense constructs have been reported for different filamentous fungi (Bautista et al. 2000; Blanco and Judelson 2005; Kitamoto et al. 1999; Lombrana et al. 2004; Ngiam et al. 2000; Zheng et al. 1998). For example, antisense silencing of the protease aspergillopepsin B resulted in a reduction of 10–70% of protease levels and a 30% increase in heterologous thaumatin production in A. awamori (Moralejo et al. 2002).

However, antisense-mediated reduction of gene expression to zero levels has not been reported to date. Nevertheless, exactly this phenomenon can be used for knocking-down gene expression instead of knocking it out.

For instance, the wide-domain transcription factor CreA, the key component of carbon catabolite repression in Aspergillus (Dowzer and Kelly 1991; Ruijter and Visser 1997; Shroff et al. 1997), negatively regulates a number of industrially important enzymes. Bautista et al. (2000) reported partial suppression of creA expression in A. nidulans by its antisense molecule (about 50% reduced expression was estimated) yielding a partial alleviation of glucose repression and thereby a substantial increase of the productivities of intra- and extracellular glucose-repressible enzymes.

Another RNA-based technology for filamentous fungi is the use of catalytic RNA molecules, termed ribozymes. The spliceosome and ribosomes are two examples for naturally occurring ribozymes. The so-called hammerhead ribozyme is the smallest and best-studied class of catalytic RNAs (Akashi et al. 2005; Mueller et al. 2006). The hammerhead ribozyme and other ribozymes are antisense RNA molecules. They function by binding to the target RNA moiety through Watson–Crick base pairing and inactivate it by cleaving the phosphodiester backbone at a specific cutting site. The substrate-recognition arms of hammerhead ribozymes are engineered so that the arms are rendered complementary to any chosen RNA, enabling the ribozyme to bind to its target. The functionality of hammerhead ribozymes as a tool for RNA-based technology on gene expression has been shown for bacterial, yeast, plant and mammalian systems (Akashi et al. 2005; Bussiere et al. 2003; Isaacs et al. 2006). Mueller et al. (2006) recently provided a proof of principle for filamentous fungi.

Congruent to other systems, it has been shown that ribozymes, targeting the 5' region of a substrate mRNA can lead to complete gene silencing in A. giganteus, whereas ribozymes targeting the 3' region only lead to a partial reduction (about 20–50%).

RNA interference (RNAi) is a naturally occurring post-transcriptional gene-silencing phenomenon, first described in Caenorhabditis elegans and thereafter in other organisms such as N. crassa (quelling; Romano and Macino 1992), plants (‘co-suppression’; Napoli et al. 1990) and animals (‘RNA interference’; Elbashir et al. 2001a). The concept of RNAi can be used for artificial gene silencing in nearly all organisms, even in filamentous fungi. By this method, double-stranded RNA (dsRNA) trans-genetically delivered to the fungal interior, is cleaved by Dicer (type-III-ribonuclease) into 21–26 nt small interfering RNA (siRNA). Dicer is always associated with Argonaute proteins (which bind targeted si/mRNA) and acts by generating siRNA molecules which in turn target mRNAs to be silenced. Dicer cleaving products get incorporated into the ribonucleoprotein complex (RISC; Bernstein et al. 2001; Hammond et al. 2000). Homologous mRNAs are subsequently recognized and degraded via complementary base pairing by means of incorporated siRNA in the RISC (Elbashir et al. 2001b; Zamore et al. 2000). In some organisms, a RNA-dependent RNA polymerase (RdRP) can use the antisense siRNA to prime the conversion of endogenous mRNA into dsRNA amplifying the silencing signal (Forrest 2004).

The most effective way of post-transcriptional gene silencing in filamentous fungi can be achieved using ectopically integrated RNAi constructs which usually code for ‘double-stranded RNA’ molecules. These molecules are self-complementary hairpin RNAs, formed by an inverted repeat which is interrupted by a spacer sequence, and are identical to part of the endogenous sequence being targeted (Mouyna et al. 2004). Insertion of an intron in the spacer sequence greatly increases the silencing efficiency in N. crassa, possibly due to an enhanced export of the hairpin from the nucleus during splicing (Goldoni et al. 2004). Remarkably, some fungal strains, e.g. Ustilago maydis, Candida albicans and S. cerevisiae lack components of the RNAi-silencing machinery, indicating that this tool is not applicable for these organisms (Nakayashiki 2005). As summarized in Table 18.2, specific inhibition of gene expression by RNAi has been shown to be suitable for a multitude of filamentous fungi, such as A. nidulans (Hammond and Keller 2005; Khatri and Rajam 2007), A. fumigatus (Bromley et al. 2006; Henry et al. 2007), A. oryzae (Yamada et al. 2007), Coprinus cinereus (Namekawa et al. 2005; Walti et al. 2006), Fusarium solani (Ha et al. 2006), Magnaportae oryzae (Caracuel-Rios and Talbot 2008; Kadotani et al. 2003; Nakayashiki et al. 2005), N. crassa (Goldoni et al. 2004) and Schizophyllum commune (de Jong et al. 2006). Similar to antisense strategies, RNAi-induced silencing of fungal gene expression was most often found to be incomplete (maximal reduction up to 10% of wild-type level) and full knockout phenotypes were seldom observed.

A striking disadvantage of RNAi-based gene silencing is the instability of the silencing construct and the possibility of co-silencing unwanted genes (‘off targets’), showing partial sequence homology to the target gene. Many transformants lose the RNAi construct after prolonged cultivation.

Chimeric constructs with two genes in tandem can result in very different silencing efficacies (Goldoni et al. 2004; Henry et al. 2007; Nakayashiki et al. 2005). In the case of A. fumigatus, it has been reported that approximately 50% of the transformants lack the complete RNAi construct or part of it after prolonged cultivation (Henry et al. 2007). One possible explanation for this phenomenon might be the loss of one of the inverted repeats after the first mitotic event (Henry et al. 2007).

Still, advances in using RNAi approaches have been achieved during recent years. For example, success has been reported by using inducible RNAi constructs in A. fumigatus (Khalaj et al. 2007) or by using new uptake methods of artificial siRNA constructs in A. nidulans (Khatri and Rajam 2007). Nevertheless, the possibility of off-target effects can form an obstacle and future work on the approach has to address the question what is the optimum sequence length and the minimum homology to avoid any unwanted co-silencing.

6 Concluding Remarks and Prospects

The fungal post-genomic era is still in its infancy; however, the real power of the omic approaches is the possibility to analyze cellular processes and responses on different levels, including DNA, RNA, protein and metabolite levels (Shulaev et al. 2008).

By integrated analysis of these levels, several important features of metabolic regulation has been and will be identified (Le Lay et al. 2006; Shulaev et al. 2008). Future strategies will combine all fields of omics and will allow the unravelling of the dynamics of cellular metabolic activities in various filamentous fungi. The identification of new proteins, enzymes, pathways and their involvement in metabolic networks as well as ‘classic genetic’ techniques and new metabolic engineering approaches will eventually enable scientists to develop optimum production strains. The use of the new and highly sophisticated omic methods developed in the fungal post-genomic era will open the floodgates towards high-throughput analysis and efficient rational metabolic engineering approaches. The knowledge gained leads to the improvement of industrial biotechnological processes and will help to meet the increasing need for sustainable fungal bio-products. In order to avoid bottlenecks in the post-genomic era, standardized methods and protocols have to be established for sampling, sample processing, sample collection, sample handling and integration into databanks.