Filamentous fungi as cell factories for heterologous protein production

200

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

Filamentous fungi as cell factories for heterologous protein production Peter J. Punt, Nick van Biezen, Ana Conesa, Alwin Albers, Jeroen Mangnus and Cees van den Hondel Filamentous fungi have been used as sources of metabolites and enzymes for centuries. For about two decades, molecular genetic tools have enabled us to use these organisms to express extra copies of both endogenous and exogenous genes. This review of current practice reveals that molecular tools have enabled several new developments. But it has been process development that has driven the final breakthrough to achieving commercially relevant quantities of protein. Recent research into gene expression in filamentous fungi has explored their wealth of genetic diversity with a view to exploiting them as expression hosts and as a source of new genes. Inevitably, the progress in the ‘genomics’ technology will further develop high-throughput technologies for these organisms. Published online: 27 February 2000

Peter J. Punt* Nick van Biezen Ana Conesa Alwin Albers Jeroen Mangnus Cees van den Hondel TNO Nutrition and Food Research Institute, Dept of Applied Microbiology and Gene Technology, P.O. Box 360, 3700 AJ Zeist, The Netherlands. *e-mail: [email protected]

For centuries, humans have used microorganisms for their own benefit. In most of these traditional processes, only in recent times has the presence and role of microorganisms been recognized (e.g. bread and cheese making, and the manufacture of alcoholic beverages). The production of more exotic products, such as soy sauce and several other oriental food ingredients, also involves processes that depend on microorganisms. Only with further technological development was the role of the various microorganisms identified clearly. People then started speaking about biotechnology, with respect to processes involving living organisms. Further research revealed that fungi, including both true filamentous fungi and yeasts, have a very important role in many of these processes. In particular, attention focused on them because they produce high yields of enzymes and metabolites (including antibiotics, organic acids, pigments and other food additives). Based on early biotechnological knowledge, a further use of fungi for the production of enzymes outside of their traditional application was developed. This has led to the identification of strains and the development of improved strains, allowing the production of large amounts of a variety of industrial enzymes. The development of sophisticated molecular genetic approaches, at first in bacteria such as Escherichia coli and Bacillus spp. and soon afterwards also in fungi, established a new stepping stone towards the goal of using fungi in protein production. The new technology allowed rapid progress, by amplifying useful characteristics and annihilating unfavourable ones to modify the genetic http://tibtech.trends.com

make-up of production strains. Furthermore, this technology opened the way to exploit these organisms to produce proteins from non-fungal origins. In this review, we illustrate how filamentous fungi have become preferred host organisms for protein production, and discuss the hurdles that face further refinement of the production of non-fungal proteins by fungal hosts. Filamentous fungi as cell factories for fungal enzymes

Filamentous fungi are capable of producing large amounts of specific proteins. The production level of any protein of interest in naturally occurring strains is usually too low for commercial exploitation. However, impressive improvements in protein yield have been obtained with traditional strainimprovement strategies based on various mutagenesis approaches. For commercial processes, yields of >30 g l−1 of a specific protein are not uncommon. In most of these examples, the originally identified strain already produced considerable levels of the protein. However, in many examples, this will not be the case, and new molecular genetic, in particular gene-transfer, approaches will be needed to improve upon available strains. The general molecular genetic approach for generating protein-overproducing strains consists of the screening for and isolating a fungal host strain that produces the protein of interest. Based on the purified protein, the gene encoding this protein can be cloned using PCR-based approaches, followed by gene characterization and expression in appropriate fungal hosts. Recently, two new gene-cloning approaches have been used relatively successfully. One approach is ‘expression cloning’, which combines simple enzyme assays with the use of a Saccharomyces cerevisiae expression system for fungal cDNAs [1,2]. The limited metabolic scope of laboratory yeast strains allows transformants that express a wide variety of fungal hydrolases to be identified by sensitive enzyme screening. The resulting cDNA clones can easily be characterized and subsequently expressed in more appropriate high-level expression hosts such as Aspergillus spp. [1,2]. Similar strategies are currently being developed for filamentous fungi [3]. Another option is the combined use of database mining and molecular screening. Besides that of S. cerevisiae, at least two fungal genome sequence

0167-7799/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0167-7799(01)02011933-9

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

are publicly available (Neurospora crassa, http://www-genome.wi.mit.edu/annotation/fungi/ neurospora/; A. fumigatus, http://www.tigr.org/tdb/ e2k1/afu1/index.shtml) and the publication of various other fungal genomes is expected shortly. This will allow us to identify, based on sequence comparison, fungal genes encoding proteins with potentially interesting hydrolytic activities. These approaches, which have not been applied to fungal species as yet, will allow new fungal enzymes to be developed. Where specific activities have been described for fungal species, database mining in combination with PCR-based molecular screening, will allow genes from fungal species with preferred enzyme activities to be cloned and then expressed in a suitable host. The choice of expression host strain cannot be made solely on the basis of production yields. Other aspects, such as regulatory issues, have a very important role in this choice. Preferred host strains are chosen from among those for which successful Generally Recognized as Safe (GRAS) and Food-Additive petitions at the Food and Drug Administration (FDA) have been filed. Moreover, patents and intellectual property rights have necessitated searching for expression hosts other than the species traditionally used. Clearly, the successful outcome of many of the studies aimed at improving production of fungal proteins held strong promise for similar research on the overproduction of non-fungal proteins. Filamentous fungi as cell factories for non-fungal proteins

Initial approaches towards the high-level production of proteins of non-fungal origin were identical to the approach followed for fungal proteins. But from the first results, it was clear that initial optimistic promises were not completely met. Although in several cases, production of the protein of interest could be detected at levels comparable to those obtained in various other expression systems, the levels obtained in filamentous fungi were always much lower than those for fungal proteins. One of the most obvious reasons for the low yields was the abundant production of (secreted) proteases by most of the fungal host strains used [4,5]. Since the first use of filamentous fungi as protein production hosts, several research groups have addressed the problem of endogenous protease production, and developed improved proteasedeficient host strains [4,5]. Such strains have largely alleviated the problem of extracellular degradation. However, even in the best protease-deficient strains there are residual levels of protease activity, which for some proteins and under certain conditions might still result in serious losses in protein yield. Another major improvement in the secretion of heterologous proteins was the ‘carrier’ approach, based on the idea that a secreted fungal protein can act as a carrier for more efficient secretion of a http://tibtech.trends.com

201

heterologous protein. Ward et al. [6], who expressed a gene-fusion comprising the genes encoding Aspergillus niger glucoamylase and bovine chymosin, described the first example of this approach in fungi. The yields of chymosin produced in strains carrying this gene-fusion were significantly greater than in strains expressing the unfused chymosin gene. Based on this initial example, further modifications of this approach were carried out (reviewed in [7]). In most subsequent cases, only the catalytic domain of the glucoamylase was used and a proteolytic processing site separated the carrier moiety from the protein of interest. The specific amino acid sequence of this processing site (NVISKR) was derived from the pro-peptide of the glucoamylase protein itself. This sequence represents a target site for a KEX2-type protein processing protease. These type of proteases are active in the secretion pathway of various eukaryotic organisms [8]. Recently, the gene encoding the corresponding protein in A. niger has been cloned and characterized [9]. Surprisingly, our recent results show that in several cases, the presence of the proteolytic processing site between the carrier and the target protein is not essential for the cleavage of the fusion protein, indicating that the KEX2-type processing is not essential for processing in these cases (P.J. Punt et al., unpublished observations). Although there are many examples that show successful use of this fusion system, much of the background to this approach remains unclear. From our own research, and undoubtedly others will have had the same experience, we also have a few examples where the fusion approach completely fails to improve yields of secreted protein [10]. However, the gene-fusion approach still remains the first choice in attempts to produce non-fungal proteins in Aspergillus. Two examples of heterologous protein production

As examples to illustrate the possibilities of fungal hosts for the production of high valued heterologous proteins, we investigated the secretion of the human cytokine interleukin 6 (IL -6), a mammalian protein with medical relevance, and a fungal manganese peroxidase (MnP), a metal-ligand containing peroxidase with extensive biochemical applications for the chemical industry. Interleukin 6

Initial experiments carried out in A. nidulans [11] showed that expression of the human gene in Aspergillus spp. was possible, but the yields of active IL-6 were small (µg l−1 levels). Introduction of the genefusion approach resulted in a considerable (200-fold) increase in the production of active IL-6 [12]. However, the levels obtained were still low (5 mg l−1) compared with many homologous fungal proteins. The initial system was based on A. nidulans, which is not a particularly good protein producer. Moreover, A. nidulans products do not have FDA approval. Therefore, we investigated other more

202

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

Table 1. Production of hIL-6 in yeast and fungal expression hosts

a

Production host

Expression system (promoter)

Secretion system

Production system

Production ⫺1 (mg l )

Refs

S. cerevisiae A. nidulans

GAL1 A. niger glaA

ss (Mf) ss (glaA/hil6) glaA carrier ss (?) glaA carrier glaA carrier

Shake flask Shake flask Shake flask Shake flask Shake flask Shake flask Stirred tank Air lift fermenter Shake flask Shake flask Shake flask Shake flask Shake flask Shake flask Stirred tank Shake flask Stirred tank

30 ⬍0.1 5 1–5 100 20–40 30 50 ⬍0.1 5–10 ⬍0.1 ⬍0.1 2 10 50–150 20–50 200–300

[42] [11] [12] [43] [43] [44] [44] [44] [45] [7] [11] [16] [16] U U U U

A. nidulans alcA alcR constitutive A. nidulans alcA(gdr) alcR constitutive A. nidulans alcA

A. foetidus/awamori

A. awamori exlA

A. niger A. niger glaA A. niger AB1.13 (protease deficient) A. nidulans gpdA A. niger D15 (protease deficient, A. nidulans gpdA non acidifying) A. sojae UV30.1 (protease deficient) A. nidulans gpdA

ss (exlA) glaA carrier ss (glaA/hil6) ss (glaA/hil6) glaA carrier glaA carrier glaA carrier

a

Abbreviations: alcA, alcohol dehydrogenase; alcR, alcohol dehydrogenase regulator; exlA, endoxylanase; gdr, glucose derepressed; glaA, glucoamylase; gpdA, glyceraldehyde-3-phosphate dehydrogenase; hil6, human interleukin 6; Mf, mating factor; mnp, manganese peroxidase; ss, signal sequence; U, unpublished.

suitable hosts for IL -6 production. Among these, A. niger, A. foetidus/awamori and A. sojae were the most attractive candidates. For A. niger and A. sojae, protease-deficient host strains were available in our laboratory [4,13], whereas for A. foetidus/awamori, an interesting alternative expression system based on xylose induction was developed [14]. Berka et al. [15] also describe an A. foetidus expression system. For all three systems, IL -6-producing transformants were obtained and analysed. These data are compared with published data from S. cerevisiae and A. nidulans (Table 1), which shows that using these expression hosts resulted in increased levels of IL -6 compared with A. nidulans, but the amounts of IL -6 produced were still relatively low (mg l−1 levels). In vitro degradation studies in the A. niger protease-deficient host strain (AB1.13; Table 1) showed that the remaining proteolytic activity (against IL -6) in the culture medium was dependent on pH, and virtually absent at non-acidic pH. Therefore, we investigated IL -6 production in A. niger D15pyrG, an AB1.13 derivative, which does not acidify its culture medium. A D15pyrG transformant, carrying several copies of the gene encoding a glucoamylase–IL-6 fusion protein (D15[pAN56–4]H4), was cultivated in parallel with an AB1.13 transformant [16]. The yield and stability of IL-6 during cultivation was improved in the non-acidifying host strain (Table 1). To evaluate the potential of this strain further we carried out controlled batch fermentations with A. niger D15[pAN56-4]H4. To investigate the suggested effect of pH on production yields, a series of two-litre batch fermentations were performed at pH values ranging from 6.0 to 7.5 (Table 2). The biomass yield decreases with increasing pH, whereas the total protein produced is not much different at the different pH values. The highest levels of IL-6 (both in mg l−1 and in mg g−1 biomass) are obtained at pH 7.0. http://tibtech.trends.com

From these results, it is clear that the level of IL-6 depends on the pH of the culture medium. Although differences in pH might exert their effect on the proteolytic system of the fungus, we observed a distinct difference in growth behaviour at different pH values. With increasing pH, the fungal morphology changed from large to small pellets. In controlled fermentations, pellets are favoured over mycelial growth owing to better oxygen transfer. Moreover, small pellets are favoured over larger pellets because in small pellets, the ratio between active mycelium (at the outside of the pellet) and inactive/dead mycelium (in the heart of the pellet) is much higher. Therefore, we could conclude that the cultures carried out at pH 7.0–7.5 perform better because of their less active proteolytic system and/or better morphology. Yield improvements up to 100–200 mg l−1 have been obtained with a multivariable experimental design approach, further optimizing production. Even higher levels of IL-6 have recently been obtained in our laboratory in a newly developed expression host system based on A. sojae [13] (Table 1). Manganese peroxidase

The Phanerochaete chrysosporium MnP is a member of the family of heme-containing peroxidases secreted by white-rot fungi to degrade lignin. These proteins are able to oxidize substrates of high redox potential in a rather nonspecific manner. Therefore, the proteins are of particular interest to the chemical industry, but their industrial application is hampered by limited availability. In their natural host, these proteins are secreted in relatively low amounts only in the stationary growth phase. Attempts to overproduce them in various heterologous systems, including fungal strains, have had limited success [17]. In our initial approach to study the production of MnP in Aspergillus spp., we used the strong, inducible

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

Table 2. Production of biomass, protein and hIL-6 in controlled batch a,b fermentations of A. niger D15[pAN56-4]H4

Batch

⫺1

⫺1

⫺1

pH

DW (g l⫺ )

Protein (g l⫺ )

hIL-6 (mg l⫺ )

Productivity ⫺1 (mg hIL-6 g DW)

6.0 6.5 7.0 7.5

23.1 16.9 13.6 8.4

0.17 0.21 0.20 0.14

43.6 33.8 49.3 29.9

1.89 2.00 3.63 3.55

a

b

Abbreviations: DW, dry weight; hIL-6, human interleukin 6. Batch fermentations were conducted in New Brunswick Scientific Bioflo 2 fermentor systems at 30°C. To maintain dissolved oxygen levels of at least 5%, stirrer speed (started at 400 rpm) was adjusted to 600 or ⫺1 ⫺1 800 rpm, if required. The growth medium contained 50 g l glucose, 13 g l sodium nitrate, ⫺1 ⫺1 5 g l yeast extract, 5 g l tryptone and trace elements. Fermentations were run for at least 48 h and samples were taken at various timepoints. The values presented correspond to timepoints just before glucose levels in the culture became exhausted. hIL-6 levels were determined with a BIACORE biosensor [46].

glucoamylase promoter to drive the expression of the mnp1 gene in an A. niger protease-deficient strain, which was previously shown to be superior for the production of IL-6 [16]. Initial yields of MnP obtained with this system in shake-flask experiments were 5–10 mg l−1 [10] (Table 3). Although this was slightly higher than previously obtained in A. oryzae [18], it was still far too low for high-level production requirements. Time-course analysis of MnP production in [MGG029]pMnp1.I#25, our most productive strain, showed that the extracellular levels of MnP, as was the case for IL-6, diminished when the culture medium became strongly acidified. Therefore, we decided to analyse MnP production under different pH conditions. Because the activity of the glaA promoter used in pMnP1.I is pH dependent [19], we constructed an expression vector where the mnp1 gene was placed under control of the constitutively expressed, pH-independent, glyceraldehyde 3-phosphate dehydrogenase (gpdA) promoter. Consequently, strain [MGG029]pgpdMnp1.I#13 was generated, in which multiple copies of the PgpdAdriven mnp1 gene were present [20]. Both this strain and [MGG029]pMnp1.I#25 (containing multiple copies of the PglaA-driven mnp1) were used in

203

controlled batch fermentations at pHs ranging from 3.0 to 6.0. At pH 4.0, ~2 mg l−1 MnP was obtained in the fermentation of strain [MGG029]pMnp1.I#25. As expected from the expression profile of the glaA promoter, even lower MnP production was observed when this strain was grown at higher pHs. The performance of strain [MGG029]pgpdMnp1.I#13 was considerably better in the whole pH range. At pH 6.0, a production level of >15–25 mg l−1 MnP was reached. This represents a 2–5-fold improvement over the initial results of the shake-flask cultures of strain [MGG029]pMnp1.I#25 [20] (Fig. 1). This example illustrates how production of heterologous proteins can be significantly improved by altering strain and culture conditions. We have also observed that supplementing the culture medium with an external heme source (e.g. hemin of hemoglobin) resulted in a 6–10-fold increase in the production of MnP by [MGG029]pMnp1.I#25 [10]. Heme supplementation might not be commercially feasible in large-scale fermentations. However, addition of hemoglobin-containing slaughterhouse residues might be feasible, although the current outbreak of BSE could cause some concern. Nevertheless, this result shows that specific requirements of the heterologous protein of interest should be considered for further improvements. From both examples described here, it is clear that only a combination of host strain selection, strain development, molecular genetic approaches and production-process development will finally result in commercially attractive protein yields. Alternative host strains

Until now, only a limited number of fungal host species has been explored for recombinant protein production. The field of modern fungal biotechnology is competitive and is attracting considerable interest from industrial parties outside of the traditional fermentation industry, interested in the applications of enzymes and other proteins. Therefore, it is not surprising that several of these parties started to explore the

Table 3. Production of Phanerochaete chrysosporium manganese peroxidase in yeast and fungal a expression hosts ⫺1

Production host

Expression system (promoter) Secretion system

Production system

Production (mg l⫺ ) Refs

P. chrysosporium S. cerevisiae A. oryzae

P. chrysosporium gpd ? A. oryzae amy

Shake flask Shake flask Shake flask ⫹ heme addition Shake flask

5–10 (?) No EC protein 5

[47] [48] [18]

5–10

[10]

100

[10]

No active protein

[10]

2 15–25

[20] [20]

A. niger MGG029 A. niger glaA (protease deficient)

ss (mnp) ss (?) ss (amy) ss (mnp) ss (mnp) glaA carrier

A. nidulans gpdA a

ss (mnp) ss (mnp)

Shake flask ⫹ heme addition Shake flask ⫹ heme addition Stirred tank Stirred tank

Abbreviations: amy, amylase; EC, extracellular; glaA, glycoamylase; gpdA, glyceraldehyde-3-phosphate dehydrogenase; mnp, manganese peroxidase; ss, signal sequence.

http://tibtech.trends.com

Fig. 1. Production of manganese peroxidase (rMnP) in controlled batch fermentations at different pHs. Circles represent strain MGG029pMnp1.I#25; squares represent strain MGG029pgpdMnP1.I#13. Values correspond to 48 h samples. rMnP production was calculated based on ABTS [2,2′-azino-bis (3-ethylbenzthiazoline-6sulfonic acid)] oxidizing activity [56]. See [20] for further details.

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

25

rMnP production (mg l–1)

204

20

15

10 2 1 0

3

4

5

6

pH

immediately cleared from the bloodstream, making them practically useless for therapeutic purposes. Proteins produced in yeasts such as Saccharomyces spp. and Pichia spp. invariably show high mannose-type glycosylation (GlcNac2Man>20). Only the use of genetically modified yeast strains results in the secretion of glycoproteins with reduced mannosylation (mainly GlcNac2Man5), which might be converted into mammalian type glycoproteins [26]. Although relatively little research has been carried out in filamentous fungi, these organisms show no extensive hyper-mannosylation of glycoproteins. In several cases, glycostructures directly compatible with conversion into mammalian structures (GlcNac2Man5) are formed. Maras et al. [27] review, in detail, the fungal glycosylation pathway and approaches to modulate it for the production of human proteins.

TRENDS in Biotechnology

Processing

possibility of using new fungal expression hosts as alternatives to those covered by several patent applications. In the past few years, a clear increase in the development of alternative hosts has been observed, which is probably related to the uncertain outcome of court trials for patent issues. For the choice of organism, clearly several aspects are relevant, such as whether registration of products derived from this organism would encounter major problems before being allowed on the food market. For this reason, species from which products have already been given FDA approval are more attractive. However, for nonfood applications (and if the proteins are of sufficient commercial value) this aspect might be less relevant. Table 4 provides a taxonomy-based overview of various potential new fungal expression hosts in relation to already existing fungal and yeast expression hosts and other species well known as molecular genetic models. Recent research in several groups has focused on developing Fusarium veneratum [21], A. sojae [13], A. japonicus [22], N. crassa [23], Mortierella alpinis [24] and Chrysosporium lucknowense [25] as hosts for protein and metabolite production. In most cases, the evaluation of the true potential of these species is still subject to further research. Production of authentic products

Clearly, as shown for the case of IL-6 and MnP, our first aim was to improve the yields of recombinant protein in the fungal culture supernatant. However, in these and many other examples, the authentic structure of the protein is also relevant, in particular for those proteins where the desired activity or application depends on its molecular structure. Glycosylation

For most proteins that have pharmaceutical potential, correct post-translational modification by N-glycosylation is important. Incorrectly glycosylated (non-complex high mannose-type) proteins are http://tibtech.trends.com

As for IL-6, the gene-fusion approach has proved valuable in the production of sufficient yields of several proteins. The fusion protein shows considerable proteolytic stability, which might indicate that production of fusion proteins, followed by specific in vitro cleavage of the fusion protein in a downstreamprocessing step, could further improve protein yields. This would be particularly useful for the production of small biologically active peptides. A potential drawback of this approach could be that processing of the fusion protein can occur, even in the absence of an engineered processing site or at sites other than those that have been engineered [28,29]. This type of processing is also occurring in the absence of a functional KEX2-like protease (P.J. Punt et al., unpublished), suggesting that extracellular proteases are responsible. The aberrant processing results in a final protein that lacks a few N-terminal amino acids or has a few extra amino acids. Although in the cases reported, this did not effect biological activity, there are also indications that incorrect processing of at least two fungal peroxidases, lignin peroxidase and chloroperoxidase, might result in inactive proteins [10,20]. Protein folding

Besides glycosylation and proteolytic processing, several other post-translational modification steps might be required to produce functional proteins. In particular, processes related to incorrect protein folding, such as incorrect oligomerization and inefficient ligand (metal-ion/heme) incorporation, were suggested to have adverse affects on yields of active protein. In the case of an oligomeric protein such as human TNF-α, no secretion of oligomeric forms of the protein was observed. However, other reasons than lack of oligomerization are suggested to be the major hurdles [30]. In the case of various hemecontaining peroxidases, the addition of hemin to the culture medium is beneficial for the yields of active protein [10,18,31,32]. For this and other types of heterologous proteins, the use of host strains with

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

205

Table 4. Systematics of biotechnologically relevant true fungi Phylum

Subphylum

Class (subclass)

Ascomycota

Saccharomycotina Saccharomycetes

Family

Genus

Saccharomycetales

Candidaceae Dipodascaceae

Candida (G) Geotrichum (F) Yarrowia (F) Eremothecium/Ashbya (G) Hansenula/Pichia Kluyveromyces (G) Saccharomyces (G) Schizosaccharomyces (F) Monascus (F) Aspergillus (G,P) Penicillium (G,P) Talaromyces (P) Cryphonectria/Endothia (G) Tolypocladium (P) Hypocrea/Trichoderma Fusarium/Gibberella (F,G) Acremonium (P) Neurospora (F)

Eremotheciacea Saccharomycetaceae

Pezizomycotina

Eurotiomycetes

Sordariomycetes

Basidiomycota

Zygomycota

Schizosaccharomycetales Eurotiales

Schizosaccharomycetaceae Monascaceae Trichocomaceae

Diaporthales Hypocreales

Valsaceae Clavicipitaceae Hypocreaceae

Sordariales

Sordariaceae

Hymenomycetes Agaricales (Hymenomycetideae)

Zygomycetes

a

a,b

Order

Agaricaceae Pleurotaceae Pluteaceae Tricholomataceae

Auriculariales Aphyllophorales

Auriculariaceae Ganodermataceae

Mucorales

Mortierellaceae Mucoraceae

Agaricus (F) Pleurotus (F) Volvariella (F) Flammulina (F) Lentinula (F) Auricularia (F) Ganoderma (F) Mortierella (G) (Rhizo)mucor (G) Rhizopus (G)

b

Abbreviations: F, food; G, generally recognized as safe (GRAS); P, pharmaceutical. Species from the indicated genus appear on the list of products from microorganisms approved for human food (GRAS status at http://vm.cfsan.fda.gov/~dms/opa-micr.html). Other species are used for the production of pharmaceutical compounds (P) or have a long history in food fermentation or for use as human food (F). For further information see [49,50] and ATCC catalogue (http://www.atcc.org/SearchCatalogs/A_Fungi_Yeasts.cfm). The remaining species are frequently used for recombinant protein production but have no recorded status. The species for which gene transfer system have been described in literature are indicated in bold [24,49,51–55].

improved levels of secretion-related chaperones, such as ER-chaperone BipA, protein disulphide isomerase PdiA, prolyl-peptidyl isomerase CypB and calnexin ClxA, might further improve protein yields. Initial results obtained in fungal strains where the levels of each of these chaperones were changed have indicated a role for these proteins in protein secretion. However, improved levels of secreted protein were only obtained in the case of calnexin overproduction in MnP-producing strains [33,34]. Future prospects

Clearly, the past decade has provided an abundance of new data regarding the use of filamentous fungi as cell factories for heterologous protein production. The basis for various strain improvement strategies lies in the development of molecular tools for several of these organisms. These tools have initiated a detailed analysis of the fungal secretion pathway [33]. From the results obtained so far, it is already clear that this pathway differs from that in S. cerevisiae. Current research in several research groups (including the EUROFUNG consortium; www.eurofung.net) is focused on unravelling the key features of this pathway. As is demonstrated in the two examples provided here, more classical research regarding the http://tibtech.trends.com

development of the production process will also be of enormous importance for the further exploitation of filamentous fungi. Although most current processes rely on submerged fermentation (SmF), the use of traditional solid state fermentation (SSF) processes should be considered. Several results have indicated that the SSF process results in improved levels of various secreted fungal hydrolases [35]. Moreover, the processes involved are probably competitive, particularly when considering process sustainability. Until now, most of the research in the field of SSF has been focused on process and fermenter design [36], treating the organism involved as a black box. Already, the first results of molecular research in SSF, show distinct differences in the level of gene expression and regulation compared with SmF [37]. Another field of research with considerable promise is that of the ‘-omics’ technologies. These technologies deal with an overall analysis of gene expression (transcriptomics), protein (proteomics) and metabolite (metabolomics) production at the level of the complete organism. Obviously, these technologies rely heavily on sophisticated laboratory techniques, development of data analysis and pattern-recognition tools (bioinformatics). Currently, the focus of the fungal research community is still

206

Review

TRENDS in Biotechnology Vol.20 No.5 May 2002

primarily on the acquisition of DNA sequence data coupled with functional analysis of identified genes [38–40]. Also the first reports on fungal proteomic approaches have appeared [41]. In the References 1 Dalboge, H. (1997) Expression cloning of fungal enzyme genes; a novel approach for efficient isolation of enzyme genes of industrial relevance. FEMS Microbiol. Rev. 21, 29–42 2 Dalboge, H. and Lange, L. (1998) Using molecular techniques to identify new microbial biocatalysts. Trends Biotechnol. 16, 265–272 3 Emalfarb, M.A. (2001) World Patent Application WO 01/25468 4 Mattern, I.E. et al. (1992) Isolation and characterization of mutants of Aspergillus niger deficient in extracellular proteases. Mol. Gen. Genet. 234, 332–336 5 van den Hombergh, J.P. et al. (1997) Aspergillus as a host for heterologous protein production: the problem of proteases. Trends Biotechnol. 15, 256–263 6 Ward, M. et al. (1990) Improved production of chymosin in Aspergillus by expression as a glucoamylase-chymosin fusion. Biotechnology 8, 435–440 7 Gouka, R.J. et al. (1997) Efficient production of secreted proteins by Aspergillus: progress, limitations and prospects. Appl. Microbiol. Biotechnol. 47, 1–11 8 Van de Ven, W.J. et al. (1993) Structure and function of eukaryotic proprotein processing enzymes of the subtilisin family of serine proteases. Crit. Rev. Oncog. 4, 115–136 9 Jalving, R. et al. (2000) Characterization of the kexin-like maturase of Aspergillus niger. Appl. Environ. Microbiol. 66, 363–368 10 Conesa, A. et al. (2000) Studies on the production of fungal peroxidases in Aspergillus niger. Appl. Environ. Microbiol. 66, 3016–3023 11 Carrez, D. et al. (1990) Heterologous gene expression by filamentous fungi: secretion of human interleukin-6 by Aspergillus nidulans. Gene 94, 147–154 12 Contreras, R. et al. (1991) Efficient KEX2-like processing of a glucoamylase-interleukin-6 fusion protein by Aspergillus nidulans and secretion of mature interleukin-6. Biotechnology 9, 378–381 13 Heerikhuisen, M. et al. (2001) World Patent Application: WO 01/09352 14 Gouka, R.J. et al. (1996) An expression system based on the promoter region of the Aspergillus awamori 1,4-β-endoxylanase A gene. Appl. Microbiol. Biotechnol. 46, 28–35 15 Berka, R.M. et al. (1995) World Patent Application WO 95/15390 16 Broekhuijsen, M.P. et al. (1993) Secretion of heterologous proteins by Aspergillus niger: production of active human interleukin-6 in a protease-deficient mutant by KEX2-like processing of a glucoamylase-hIL6 fusion protein. J. Biotechnol. 31, 135–145 17 Cullen, D. (1997) Recent advances on the molecular genetics of ligninolytic fungi. J. Biotechnol. 53, 273–289 18 Stewart, P. et al. (1996) Efficient expression of a Phanerochaete chrysosporium manganese peroxidase gene in Aspergillus oryzae. Appl. Microbiol. Biotechnol. 62, 860–864 http://tibtech.trends.com

near future, requirements for ‘-omics’ research in filamentous fungi will undoubtedly be met, allowing the fungal research community to address the various biological questions that it still faces.

19 Withers, J.M. et al. (1998) Optimization and stability of glucoamylase production by recombinant strains of Aspergillus niger in chemostat culture. Biotechnol. Bioeng. 59, 407–418 20 Conesa, A. (2001) Overproduction of fungal peroxidases in filamentous fungi, Thesis, University of Leiden, the Netherlands 21 Royer, J.C. et al. (1996) World Patent Application WO 96/00787 22 Berka, R.M. et al. (1995) World Patent Application WO 95/15391 23 Stuart, W.D. (1997) World Patent Application WO 97/22687 24 Mackenzie, D.A. et al. (2000) Isolation and use of a homologous histone H4 promoter and a ribosomal DNA region in a transformation vector for the oil-producing fungus Mortierella alpina. Appl. Environ. Microbiol. 66, 4655–4661 25 Emalfarb, M.A. et al. (2000) World Patent Application WO 00/20555 26 Chiba, Y. et al. (1998) Production of human compatible high mannose-type (Man5GlcNAc2) sugar chains in Saccharomyces cerevisiae. J. Biol. Chem. 273, 26298–26304 27 Maras, M. et al. (1999) Filamentous fungi as production organisms for glycoproteins of bio-medical interest. Glycoconj. J. 16, 99–107 28 MacKenzie, D.A. et al. (1998) Aberrant processing of wild-type and mutant bovine pancreatic trypsin inhibitor secreted by Aspergillus niger. J. Biotechnol. 63, 137–146 29 Spencer, J.A. et al. (1998) Determinants of the fidelity of processing glucoamylase-lysozyme fusions by Aspergillus niger. Eur. J. Biochem. 258, 107–112 30 Krasevec, N. et al. (2000) Can hTNF-α be successfully produced and secreted in filamentous fungus Aspergillus niger? Pflugers Arch. 439, 84–86 31 Elrod, S.L. et al. (1997) World Patent Application WO 97/47746 32 Conesa, A. et al. (2001) Expression of the Caldariomyces fumago chloroperoxidase in Aspergillus niger and characterization of the recombinant enzyme. J. Biol. Chem. 276, 17635–17640 33 Conesa, A. et al. (2001) The secretion pathway in filamentous fungi: a biotechnological view. Fungal Genet. Biol. 33, 155–171 34 Conesa, A. et al. (2002) Calnexin overexpression increases manganese peroxidase production in Aspergillus niger. Appl. Environ. Microbiol. 68,846–851 35 Pandey, A. et al. (2000) New developments in solid state fermentation: I-bioprocesses and products. Process Biochem 35, 1153–1169 36 Weber, F.J. et al. (1999) A simplified material and energy balance approach for process development and scale-up of Coniothyrium minitans conidia production by solid-state cultivation in a packedbed reactor. Biotechnol. Bioeng. 65, 447–458 37 Ishida, H. et al. (1998) Nucleotide sequence of an alternative glucoamylase-encoding gene (glaB) expressed in solid-state culture of Aspergillus oryzae. J. Ferment. Bioeng. 86, 301–307 38 Yoder, O.C. and Turgeon, B.G. (2001) Fungal genomics and pathogenicity. Curr. Opin. Plant Biol. 4, 315–321

39 Sweigard, J.A. and Ebbole, D.J. (2001) Functional analysis of pathogenicity genes in a genomics world. Curr. Opin. Microbiol. 4, 387–392 40 Hamer, L. et al. (2001) Gene discovery and gene function assignment in filamentous fungi. Proc. Natl. Acad. Sci. U. S. A. 98, 5110–5115 41 Lim, D. et al. (2001) Proteins associated with the cell envelope of Trichoderma reesei: a proteomic approach. Proteomics 1, 899–910 42 Guisez, Y. et al. (1991) Production and purification of recombinant human interleukin-6 secreted by the yeast Saccharomyces cerevisiae. Eur. J. Biochem. 198, 217–222 43 Hintz, W.E. et al. (1995) Improved gene expression in Aspergillus nidulans. Can. J. Bot. 73, 876–884 44 Yadwad, V.B. et al. (1996) Effect of culture conditions and induction strategies on production of human interleuki-6 by a recombinant Aspergillus nidulans strain. Mycol. Res. 100, 356–360 45 Gouka, R.J. et al. (1996) Analysis of heterologous protein production in defined recombinant Aspergillus awamori strains. Appl. Environ. Microbiol. 62, 1951–1957 46 Malmqvist, M. (1999) BIACORE: an affinity biosensor system for characterization of biomolecular interactions. Biochem. Soc. Trans. 27, 335–340 47 Mayfield, M.B. et al. (1994) Homologous expression of recombinant manganese peroxidase in Phanerochaete chrysosporium. Appl. Environ. Microbiol. 60, 4303–4309 48 Pease, E. and Tien, M. (1991) Lignin-degrading Enzymes from the Filamentous Fungus Phanerochaete chrysosporium, Plenum Press 49 Whiteford, J.R. and Thurnston, C.F. (2000) The molecular genetics of cultivated mushrooms. Adv. Microb. Physiol. 42, 1–23 50 Chiu, S-W. et al. (2000) Themes for mushroom exploitation in the 21st century: Sustainability, waste management, and conservation. J. Gen. Appl. Microbiol. 46, 269–282 51 Sato, T. et al. (1998) Transformation of the edible basidiomycete Lentinus edodes by restriction enzyme-mediated integration of plasmid DNA. Biosci. Biotechnol. Biochem. 62, 2346–2350 52 Jain, S. et al. (1992) Development of a transformation system for the thermophilic fungus Talaromyces sp. CL240 based on the use of phleomycin resistance as a dominant selectable marker. Mol. Gen. Genet. 234, 489–493 53 Baron, M. et al. (1992) Efficient secretion of human lysozyme fused to the Sh ble phleomycin resistance protein by the fungus Tolypocladium geodes. J. Biotechnol. 24, 253–266 54 Gietz, R.D. and Woods, R.A. (2001) Genetic transformation of yeast. Biotechniques 30, 816–826 55 Van den Hondel, C.A.M.J.J. and Punt, P.J. (1991) Gene transfer systems and vector development for filamentous fungi. In Applied Molecular Genetics of Fungi (Peberdy, J.F. et al., eds) pp. 1–28, Cambridge University Press 56 Glenn, J.K. and Gold, M.H. (1985) Purification and characterization of an extracellular Mn(II)dependent peroxidase from the lignin-degrading basidiomycete, Phanerochaete chrysosporium. Arch. Biochem. Biophys. 242, 329–341