WO2005087922A1 - Altered expression in filamentous fungi - Google Patents

Abstract

Agaricus bisporus serine protease is selectively expressed in the stipe, with maximum expression observed after 5 days at 18 °C, the promoter thereof being effective to permit selective expression of a heterologous protein in the fruiting body. Transformants can be selected without having to induce fruiting, by forcing expression using humic proteins. Post-transcriptional gene suppression is useful to suppress the serine protease to allow harvesting of heterologous proteins from the fruiting body without degradation by the serine protease.

Description

ALTERED EXPRESSION IN FILAMENTOUS FUNGI

The present invention relates to the suppression of proteases in filamentous fungi, methods for such transformation, expression of heterologous proteins in such fungi, and processes for the collection of the resulting expression products.

The cultivated mushroom, Agaricus hisporus, is a high value crop with a world value at €4.5bn. This value can be rapidly lost after crop harvest as a result of natural deteriorative processes affecting product quality which can be readily detected by potential consumers as brown discoloration, loss of texture, maturation or cap opening and changes in flavour.

Transformation technologies in the cultivated button mushroom A. hisporus have only recently been optimised (Chen et al., 2000; Mikosch et al, 2001). Very few fransgenes have been expressed in this fungus, and these have been heterologous genes, such as hygromycin phosphotransferase (Chen et al, 2000, Mikosch et al, 2001) and green fluorescent protein. To date, no attempts to overexpress, downregulate or knock out a native A. hisporus gene have been published.

Mushroom senescence is the breakdown of the fruiting body resulting in loss of texture and structure in the mushroom. In cropped mushrooms, this process begins shortly after harvest (ca. 1-2 days). When the mushrooms are left unharvested, this process occurs at the end of the fruit body's life. Several reports indicate nutrient recycling from the stipe to the cap in this process (Hammond and Nichols, 1975; Ajlouni et al, 1992; McGarry and Burton, 1994) which corresponds with studies in Schizophyllum commune (Lilly et al, 1991) and Postia placenta (Micales, 1992). The mushroom fruit body becomes nutritionally isolated on harvest, triggering recycling of nutrients to allow continued maturation of the sporophore for basidiospore production. In the later stages of development in unpicked mushrooms, similar deprivation of nutrients may occur. Removal of mushroom caps from stipes has no effect on senescence, suggesting that onset of degradation may be related to the loss of a signal from the mycelia, rather than cap-related signalling (Burton et al, 1997a). The mechanisms by which senescence is initiated are not fully understood.

Serine proteinase is a 27kDa protein implicated in post-harvest and age-related senescence of mushroom sporophores. Serine proteinase was isolated and characterised as the predominant proteolytic enzyme in senescent tissue (Burton et al. , 1993). Once harvested, mushroom fruit bodies rapidly senesce, with the protein content in both the stipe and cap of the mushroom decreasing, although protein loss is proportionally higher in stipe tissue, where protein loss coincides with an increase in serine proteinase (Burton et al, 1997a). At its height, 5 days post harvest, serine proteinase accounts for 3% of soluble protein in the stipe tissue, suggesting a role in the nutrition of isolated sporophores (Burton et al, 1993).

Serine proteinase degrades proteins, and the protein levels in the tissues of mushrooms stored for 5 days after harvest declines to one tenth and one fifth in stipe and cap respectively, of the original levels at the time of harvest. Serine proteinase has been shown to activate the enzyme, tyrosinase (also known as polyphenol oxidase or PPO) which is responsible for tissue browning.

Serine proteinase is also implicated in mycelial growth with experiments demonstrating it to be the dominant extracellular proteinase produced when mycelia are cultured on compost or in liquid media containing protein (Burton, et al, Mycological Research, 101, pgs 1341-1347 (1997)). Thus, serine proteinase appears to be significant in both mycelial nutrition and senescence of the mushroom fruit body.

Although mushrooms are largely considered to be a food crop, they also have the potential to be used as a system for the production of heterologous proteins for use as pharmaceuticals, industrial enzymes, bulk proteins etc. This bioprocessing in mushrooms involves the transformation of A. hisporus cells with the gene encoding the protein of interest.

Attempts to express heterologous proteins in filamentous fungi have met with numerous problems, not least being the difficulty associated with transforming the fungi in the first place. For example, WO95/02691 relates to transforming a mushroom mycelium and fruit bodies through methods such as electroporation with suitable vectors. Although this method works on a small scale, it is not especially efficient.

WO98/45455 relates to the possibility of transforming moulds, such as Agaricus hisporus, with the Agrobacterium tumefaciens bacterium, which causes crown gall tumours at the wound site of infected dicotyledonous plants. This bacterium is well known for its ability to transform plants, but it has only recently been established that it can also transform filamentous fungi.

WO 96/41882 discloses the expression of hydrophobins by the hyp A, hyp B, hyp C and hyp D genes. These are naturally occurring fungal products, and are expressed during fruiting. Heterologous expression is suggested, in connection with enhancing the flavour and/or nutritional content of the fruiting bodies.

Mol. Gen. Genet. (1993), 238: 91-96, provides a reporter-gene system in S. commune, for the expression of hydrophobins.

The problem still remains that the expression of heterologous genes in substantial amounts will generally substantially reduce the growth potential of the mycelium and, therefore, the harvest of the fruiting body, and prohibits the expression of any substance which is, in any way, toxic to the growth of the fungus.

Heterologous protein yields in mycelial fungi are generally low (Gouka et al, 1997). The 'foreign' nature of the DNA seems to have an impact on production level as DNA composition and GC content can be very different. In most cases, the production of heterologous protein is lower than for homologous proteins. For example the production of egg lysosyme is five times less than the production of a resident glucoamylase (Radzio and Kuck, 1997).

The ability to introduce novel genes into mushrooms through transformation is now a reality for the major edible crop species. Conventional transformation technologies have been shown to work in the principal cultivated mushroom, Agaricus hisporus, Lentinus edodes, and species oϊPleurotus the oyster mushrooms (Challen et al 2000). More recently Agrobacterium mediated transformation has been used to transform A. hisporus at high frequency. Until the development of Agrobacterium transformation technologies for A. bisporus were restricted to the use of protoplasts coupled with chemical or electroporative mediated uptake of DNA. Ballistic delivery of transforming DNA (biolistics) is protoplast independent and also has potential for mushroom transformation (Moore et al, 1995).

Many of the problems associated with pharma production in green plants are not relevant to mushrooms. Mushrooms are grown world wide in purpose built structures which enables the containment of crops of pharma mushrooms with no risk of contaminating other food crops or the environment.

Current treatments for many conditions/diseases depend on human and viral proteins. These are currently produced in animal cell cultures in sophisticated bioreactors (Houdebine 2000, Lubon &.Palmer, 2000). The two problems associated with the current systems are cost and capacity. Drugs produced in animal cell cultures in bioreactors are extremely expensive. There are not enough bioreactors to meet current let alone increasing future demand. Around 75% of current mammalian cell fermentation capacity is tied up in the production of just four pharmaceuticals (Pew Initiative, 2002).

Potential yields of a transgenic protein from a green plant system, tobacco, (Cramer 1999), and the mushroom, A. bisporus, can be compared on an annual basis. For tobacco, about 550 tonnes of biomass can be produced per hectare compared with mushrooms, where 3000 tonnes per hectare are achievable. Potential yields of a transgenic protein can be readily calculated.

A particularly advantageous system is disclosed in our co-pending PCT Application PCT/GB2003/004716. Successive selection procedures identify the transformed cultures producing full-length, active protein at high yields. The selected culture may then be grown vegetatively first on grain to produce mushroom spawn which in turn is used to inoculate mushroom compost. When the compost is fully colonised, the production of mushroom fruitbodies is induced by the colonisation of mushroom casing and the initiation process (reduced temperatures, reduced carbon dioxide levels and high levels of irrigation). The advantages of mushrooms for bioprocessing are low capital costs, rapid lead times, scaleable production, containment and high yields (300kg per tonne compost or 3200 tonnes/ha/year).

Heterologously produced protein may be removed from transgenic mushrooms by harvesting, homogenisation, separation of filtrate from pulp, followed by purification procedures. The major problem of protein extraction/purification is the release of endogenous proteinases at the time of cell disruption which cause hydrolysis of the protein of interest. For mushrooms, the enzyme serine proteinase is a major concern because of its broad substrate specificity and its high activity after harvest. Any reduction in serine proteinase activity in the mushroom would offer major advantages to mushrooms in the bioproduction of heterologous proteins. Standard procedures to inhibit serine protease would be unlikely to solve the problem, as there appear to be several isozymes.

The traditional approach for eliminating expression of a target gene is to disrupt the desired gene with a marker (e.g. antibiotic resistance) and transform with this cassette. The cassette integrates into the host genome in a homologous fashion, replacing the functional gene with the disrupted one. This approach was not used, as knowledge of flanking sequences is necessary, and this is currently limited, for Agaricus bisporus Sprl.

In addition, for such a disruption to be successful, all copies of the target gene need to be affected, which is contraindicated when using sections of gill tissue from the fruit body, which contain many nuclei. Targeted gene disruption in all the nuclei of a resultant transformant would be exceedingly unlikely, and would allow any intact Sprl genes to mask disrupted copies.

In addition, it is likely that Sprl is part of a gene family. Burton et al. (1997) isolated two serine proteinase isozymes, and the predicted N-terminal region amino acid sequence derived from the Sprl cDNA (Kingsnorth et al, 2001) does not correspond exactly to the N-terminal amino acid sequence deduced by Burton et al, (1993). Thus, homologous disruption of the target serine proteinase gene may be possible, but the phenotypic effect could be masked by proteinase activity from a non- disrupted isozyme.

Post-transcriptional gene silencing (PTGS) has been investigated in various organisms, including in the model fungus Neurospora crassa. (Hutvagner and Zamore, 2002; Faugeron, 2000). Although PTGS is not completely understood, it is thought that an "aberrant RNA" species triggers a degradation of all similar RNA species in the cell. PTGS appears to spread through organisms and, in fungi, this mechanism is known as quelling, and heterokaryons containing quelled and non- quelled nuclei have been shown to exhibit a quelled phenotype (Cogoni et al, 1996).

Transformation and overexpression of proteases have been carried out in several fungi (with varied results), such as Candida albicans (Orozco et al, 2002), Trichoderma harzianum (Delgado-Jarana et al, 2002), Aspergillus oryzae (Cheevadhanarak et al, 1991), A.fumigatus (Dunne et al, 2000; Reichard et al, 2000), Metarhizium anisophilae (St Leger et al, 1996) and Cryphonectria parasitica (Choi et al, 1993). Downregulation of a proteinase was reported by Zheng et al. (1998) in oryzae.

No attempts to use PTGS in A. bisporus or other Homobasidiomycetes have been reported.

US-A-6352841 discloses filamentous fungi transformed to express heterologous proteins, and wherein the host has been genetically modified to express significantly reduced levels of a metalloprotease and an alkaline protease. There is no suggestion that anything other than standard mutational procedures can be used, either to suppress the protease or to enhance expression of the heterologous protein. Thus, expression of the protein is constitutive, and protease suppression is compromised if isozymes of the proteases exist. Only Aspergillus spp. are exemplified.

We have now established that it is possible to substantially inhibit protease expression in filamentous fungi by post-transcriptional gene suppression (PTGS).

Thus, in a first aspect, there is provided a filamentous fungus in which one or more proteases are suppressed by post-transcriptional gene suppression.

It is particularly preferred that a protease to be suppressed is a metalloproteinase, preferably a matrix-metalloproteinase, preferably a human matrix- metalloproteinase.

Aspartate and cysteine proteases are also preferred. However, it is particularly preferred that the protease is a serine protease.

It is particularly preferred that the filamentous fungus is one that produces harvestable fruiting bodies, and is preferably a Basidiomycete. Homobasidiomycetes are preferred, especially Agaricus bisporus. While the invention extends to other forms of filamentous fungus, the most preferred is A. bisporus, as there is a wealth of experience in cultivating this species.

It will be appreciated that, where the terms "mushroom" and "A. bisporus" are used herein, these terms include other fungi of the invention, unless otherwise apparent or specified. It will be appreciated that the term "filamentous fungus", as used herein, extends to fungal preparations, such as spawn, other mycelial preparations, dried fruiting bodies, and spores, for example.

PTGS may take any suitable form in the fungi of the present invention. Essentially, all that is necessary is that the fungus carry DNA corresponding to the protease to be suppressed, this DNA being capable of being at least partially transcribed.

This PTGS DNA is preferably incorporated into the genome of the fungus and is under the control of a suitable promoter. While it is not essential that the promoter be situated in the immediate proximity of the DNA to be transcribed, it is generally preferred that the promoter be directly associated with the DNA, and that a suitable marker, in a preferred embodiment, be situated beyond the transcribable DNA.

The transcribable DNA corresponds to the DNA of the protease to be suppressed. In this regard, the DNA may take various forms, such as are illustrated in the accompanying Example. In particular, sense or antisense DNA conesponding to all or a part of the gene to be suppressed may be provided, under the control of a suitable promoter. A stop codon may be situated such as to prevent translation. A terminator may also be provided, especially where it is desired to restrict a length of the transcription product. If there is a marker, then it is preferred that the transcription product be separate, or cleavable from the mRNA transcription product. This can be achieved, for example, by the use of suitable stop and start codons, or providing the suppression RNA in the form of a cis-acting ribozyme RNA.

In the event that the transcribable DNA is sense DNA, then it is preferred that this not encode the entire protease. When selecting transformants, however, full length sense DNA may be employed, as transformation of fungal cells, especially A. bisporus, frequently appears to involve truncation events.

In general, without being restricted by theory, it is assumed that PTGS usually occurs through the agency of antisense RNA binding with protease mRNA to yield dsRNA, which is rapidly cleaved in situ, thereby preventing, or substantially inhibiting, expression of the protease. Thus, it is only necessary that short lengths of DNA, corresponding to the protease to be suppressed, be transcribed. Suitable lengths of DNA to be transcribed vary from as low as about 20 nts (nucleotides) up to 1000 or more nts. However, especially where it is desired to provide a marker, it is preferable to restrict the length of the transcription product that will interfere with protease expression to between about 25 and 250 nts, with 25 to 100 and 25 to 50 nts being increasingly preferable from the perspective of avoiding, or minimising, truncation events.

The PTGS DNA may be under the control of the promoter of the protease to be suppressed, especially when this is serine protease, for example. Thus, the RNAi will only be produced by a fungus when the serine protease is produced, and this can be advantageous, insofar as a level of serine protease will be expressed prior to ^" effective suppression by the resulting RNAi.

However, in general, it is preferred that the PTGS DNA be transcribed constitutively, although it is preferred that this not be a metabolic drain, and be sufficient to substantially suppress protease production.

It is not essential to completely suppress protease expression and, in general, levels of expression of up to 10 percent of normal can be tolerated, although it is prefened that levels of expression of 1 percent, or less, be achieved.

The advantage of using PTGS in the present invention includes the ability to suppress production of protease isozymes. In this regard, it is preferred to select a region of the gene common to the various isozymes, thereby permitting RNAi to suppress the various isozymes.

A preferred RNAi corresponds to a sense or antisense form of SEQ ID NO. 1.

Kadotani et al, (Mol. Plant-Microbe Int., Vol 16, 2003) shows that RNAi works in M. oryzae (formerly M. grisea). However, RNAi is only discussed in relation to this specific fungus. There is no suggestion that RNAi would work in any other fungal species, and in particular that it would work in filamentous fungi, let alone against proteases.

RNAi has been shown to work in other fungi, see, for instance WO 01/53475A2. Liu et al, (Genetics, Vol 160, 2002) discloses that RNAi gives rise to specific gene disruption in C. neoformans, a heterobasidiomycete. Liu et al makes no mention of proteases at all.

Sanceau et al, (J. Biol. Chem. Vol. 278, 2003) discloses that RNAi can be used against proteases, but this is only in relation to humans, and specifically in cancer, such as Ewing's sarcoma.

RNAi is known in humans and also known in trypanosomes, even to prevent expression of proteases (Srinivasula et al, and LaCount et al, both J. Biol. Chem. Vol. 278, 2003, WO03/064621 A and US2003/0235900A1).

However, as mentioned above, no attempts to use PTGS in A. bisporus or other Homobasidiomycetes have been reported. Indeed, as there is no suggestion that PGTS is effective in filamentous fungi, it would not be expected to do so. It is, therefore, particularly suspiring that PGTS has been found to work, not only in filamentous fungi, but against proteases in said fungi.

Filamentous fungi are, of course, substantially different from the fungi disclosed in the art, such as M. oryzae. Indeed, even within filamentous fungi, Hetero- and Homo- basidiomycetes are also significantly different. Hetero- and Homo- basidomycetes are two different and distinct sub-classes of the fungal subdivision of Basidomycotina.

C neoformans is an encapsulated heterobasidomycete yeast-like fungus. C neoformans and other Heterobasidomycetes (such as Ustilago maydis) are morphologically and genetically distinct from the Homobasidiomycete fleshy-type fungi of a preferred embodiment of the present invention. Many of the genetics techniques applied to Heterobasidomycetes have been proven to be not readily transferable to Homobasidiomycetes.

Although some genetic processes may be conserved between these two fungal sub-classes, there are often substantial differences in their genetic architecture, for instance. Indeed, Hargreaves et al (1999, Antisense & Nucleic Acid Drug Development, 9, pp. 101-104) failed to demonstrate PTGS in the Heterobasidomycete U. maydis despite evidence of silencing transcripts being transcribed.

Mushrooms of the present invention demonstrate reduced senescence, and increased shelf-life, but are particularly useful in the production of heterologous proteins.

Where a heterologous protein is encoded by a fungus of the invention, then it may be harvested from the mycelium or fruiting body, and suffer less digestion from the suppressed protease. It will be appreciated that the present invention specifically envisages the production of proteins that would otherwise be susceptible to digestion by the suppressed protease. It is also particularly advantageous that the protein to be harvested be expressed in a fungal fruiting body.

A heterologous protein, or other gene product, for expression in fungi of the invention may be under the control of the protease promoter, especially the A. bisporus serine protease promoter. In particular, all or part of the promoter of SEQ ID NO. 3 is preferred, or an effective analogue thereof.

When under the control of the serine protease promoter, the heterologous protein will be primarily expressed in the fruiting body, with maximum expression noted at Day 5, or thereabouts, when stored at 18°C, as is characteristic of the expression of serine protease. This is particularly advantageous from the point of view of being able to assay the effectiveness of transformation with the heterologous protein.

In this regard, a transformant for commercial use will tend to express optimal levels of the protein in the fruiting body after about 5 days, although it will be appreciated that purification of the protein may be begun at any desired point, either before or after this time. Leaving the fruiting body for longer than 5 days will generally allow even greater expression of the heterologous protein, but other processes involved in the degeneration and senescence of the mushroom may hinder or reduce recovery. Purification of the protein significantly prior to 5 days will generally be inefficient, as the amount of heterologous protein will not be maximised under normal circumstances. It is a particular advantage that transformants can be encouraged to express the heterologous protein during the mycelial stage, rather than having to encourage the fungus to produce fruiting bodies. In this regard, as noted by Burton et al. (1997, supra), exposure to humic proteins encourages the expression of serine protease so that, if heterologous proteins are under the control of the serine protease promoter, then exposure of transformants to humic proteins can be used to assay the effectiveness of transformation.

It will also be appreciated that fruiting bodies may be harvested and kept under cool conditions, prior to storage at 18°C, if it is desired, for any reason, to process batches at times other than around 5 days after harvest.

Preferred fungi of the invention are those wherein a protease, especially serine protease, is suppressed by PTGS, and a heterologous protein is expressed, especially during the production of the fruiting body.

The majority of edible mushrooms are Homobasidiomycetes (although truffles and morels are in the Ascomycota, however). Cultivated edible Homobasidiomycetes are decayers that have been domesticated, such as button mushrooms (Agaricus bisporus), shiitake (Lentinula edodes), oyster mushrooms (Pleurotus ostreatus), and others. Accordingly, preferred Homobasidiomycetes according to the present invention include edible mushrooms, being those that are generally considered to be safe to eat by humans, especially shiitake and oyster mushrooms. These may be cultivated, in particular commercially.

The present invention also relates to wild-collected edible species of Homobasidiomycetes, such as porcini (Boletus edulis), chanterelles (Cantharellus cibarius), and matsutake (Tricholoma matsutake). However, as most of the wild- collected edible species are mycorrhizal (making them more difficult to cultivate), cultivated or cultivatable Homobasidiomycetes are preferred.

It will be understood that the present invention encompasses all Homobasidiomycetes fungi, not just the illustrative examples given above.

As disclosed in co-pending PCT Application PCT/GB2003/004716, at least three genes are switched on, or otherwise subjected to elevated levels of expression, at around the veil-break stage of fruiting body development, and that heterologous DNA under the control of the expression mechanisms of these fungal genes can be selectively expressed at this stage of development of the fungus, rather than during growth of the mycelium.

Thus, in a preferred aspect, an Spr-quelled filamentous fungus of the invention is transformed with a heterologous sequence of DNA, the fungus being capable of expressing the heterologous DNA, and wherein the heterologous DNA is under the control of a filamentous fungus transcription promoter active substantially only during stage 1, or later, of the development of the fruiting body of the fungus.

In this embodiment, it is a particular advantage that little or no metabolic energy need be diverted from mycelium growth, thereby maximising fruiting body mass and concomitant tissue capable of expressing the heterologous gene once it is switched on.

It is a further advantage that little or none of the heterologous gene product is expressed during vegetative growth of the mycelium, thus enabling the production of substantially any substance capable of expression in the filamentous fungus in question, even if that substance, either alone or in combination, results in the death or stasis of the fungus. The promoters allow the synchronous switching on of the gene at a time of rapid growth and high metabolism so that, by the time any potentially toxic effects become apparent, harvestable quantities of the substance are available. Where the gene product is a regulator of mushroom growth, for example, then such considerations are not generally necessary.

It is particularly preferred to harvest the expression product of the heterologous DNA, and the product may be purified, if and as desired, by any suitable means, such as are well known in the art.

There is further provided a method for the production of a substance expressible by a DNA sequence, wherein the sequence is operably associated with a filamentous fungus transcription promoter active substantially only during stage 1 , or later, of the development of the fruiting body of the fungus, the sequence and promoter being expressibly incorporated in a filamentous fungus, the fungus being cultured to fruition and the product being harvested. In this embodiment, it is particularly prefened that the DNA encode a protein to be harvested. The present invention also provides a transformed filamentous fungus. The fungus is preferably transformed to express heterologous DNA. Preferably, the heterologous DNA is under the control of a protease promoter. Preferred promoters are discussed elsewhere, but a particularly preferred promoter is a Serine protease promoter, especially abstl or rafe.

Accordingly, there is also provided a method for selecting potential transformants, wherein the DNA to be expressed is preferably under the control of filamentous fungus protease promoter, preferably a serine protease promoter, especially abstl or rafe. The method preferably comprises culturing fungal cells to form a mycelium, exposing the mycelium to sufficient humic stimulus to stimulate serine protease production in untransformed A. bisporus, and assaying for the presence of the expression product thereafter.

The three genes associated with selective promotion that have been identified are expressed substantially only during development of the fruiting body, and particularly during stages 4 to 7 (veil-break onwards). Without being constrained by theory, it is likely that these genes are associated with the massive water uptake required for the expansion of the fruiting body and its maturation. Preferably, the promoter is expressed at or around the veil-break stage, preferably during stages 2-7, more preferably 4-7.

These three genes, so far identified, are abstl, rafe and mag2. The expression product of abstl appears to be involved in the transport of sugars, whilst the expression product of rafe is a putative riboflavin aldehyde forming enzyme. The expression product of mag2 is a so far unidentified moφhogenesis associated protein.

In relation to the genes of the preferred promoters of the invention:

abstl : Up-regulated (more than 100 fold) during mushroom development, abundantly expressed through stages 4 - 7 (later stages of mushroom development), ca. 0.6% transcripts at stage 4, represented by 20 clones in the differential library, 1.7 kb transcript

rafe: Up-regulated (up to 50 fold) during mushroom development, abundantly expressed through stages 4 - 7, 0.7 kb transcript mag2: Up-regulated (up to 30 fold) during mushroom development, ca. 0.6% transcripts at stage 4, represented by seven clones in the differential library, comparable levels of expression in stipe and cap tissue, ca. 0.7 kb transcript

It will be appreciated that the Basidiomycetes, including members of the Agaricales, of which A. bisporus is one, share the exceedingly rapid development of the fruiting body in common. Without being constrained by theory, it is envisaged that this development is as a result of a rapid increase in osmotic pressure in the cells of the immature fruiting body, thereby causing a rapid influx of water into the cells. The resulting sudden expansion of the cells expands the fruiting body up to several hundred times its original size.

One or more sugar transport mechanisms are switched on at the early stages of fruiting body development, and abundant expression product is noted, especially by stage 4. These genes form a prefened subject of the present invention, and especially the control element associated therewith, but it will be appreciated that any gene selectively expressed, or with greatly enhanced expression, during development of the fruiting body is useful in the present invention.

The abstl gene is up-regulated, by more than 100-fold, during mushroom development, and is abundantly expressed through stages 4 to 7, and represents about 0.6%) of the transcripts detected at stage 4. The transcript is about 1.7kb in length.

The gene product of rafe is up-regulated by about 50-fold during mushroom development, and is abundantly expressed from stage 4 onwards, as with abstl. The transcript is 0.7kb in length.

The expression product of mag2 is up-regulated by about 30-fold during mushroom development, and represents about 0.2% of the transcripts at stage 4. Unlike abstl and rafe, the expression of mag2 appears to be comparable in both the stipe and cap tissues. The length of the transcript is about 0.7kb.

The control elements, and especially the promoters, of these and other genes expressed during the development of the fruiting body are particularly useful in the present invention. Elements from genes associated with sugar transport are particularly preferred. It is an advantage of the present invention that heterologous genes can be expressed at selected stages of sporophore development, where these genes might otherwise be harmful to the fungus. Expression of the heterologous genes can be selected to occur substantially only during growth of the fruiting body so that, unless the gene product is acutely toxic, then large amounts of the gene product can be expressed which would otherwise harm or hinder growth of the fungus.

Owing to the commonality of the fruiting body growth process, the promoters of any one filamentous fungus, switched on during the fruiting body growth cycle, may be employed in other filamentous fungi in the context of the present invention, in order to express heterologous genes.

In the accompanying Sequence Listing, SEQ ID NO's 10 and 11 are the promoter sequences associated with abst 1 and rafe, respectively.

The promoter may be used in association with other suitable control sequences, such as terminators. A suitable terminator may be as shown in the accompanying sequences, or may be the Aspergillus nidulans trpC terminator, for example. Other terminators are well known in the art.

The terminator sequences of abst 1 and rafe are provided as SEQ ID NO's 12 and 13, and it has been found that it is desirable to use terminators generally associated with the promoters, so that it is prefened to use the abst 1 terminator sequence with the abst 1 promoter sequence, and likewise for rafe. However, it will be appreciated that the present invention also envisages any suitable expression system comprising the promoter, and any suitable terminator may be employed, as desired.

It will also be appreciated that the promoter sequences and terminator sequences of the invention are preferably those as listed as SEQ ID NO's 10 and 11, and 12 and 13, respectively, and sequences comprising these sequences, as well as sequences hybridising with these sequences, preferably under conditions of 60°C stringency or higher, provided that promoter activity is retained in the by the sequence or the sequence to which it hybridises. Mutations and naturally occurring variants of the sequences are encompassed, and it may be, for example, appropriate to introduce a restriction site or sites for ease of manipulation. Provided that promoter activity is retained, there is no restriction on how much the promoter may be modified. Similar considerations apply to the terminator sequences.

Preferably, promoter sequences and terminator sequences has greater than 75% sequence homology, more preferably 90%, more preferably 95%, more preferably 99% and most preferably greater than 99.5% sequence homology, with SEQ ID NO's 10 and 11, and 12 and 13, respectively.

The promoter regions may be used in their entirety when preparing heterologous genes for expression in filamentous fungi. Alternatively, it may be prefened to use consensus sequences from these regions. There is no especial advantage to using consensus sequences, except that these may be shorter. Otherwise, it is sufficient to supply the promoter upstream of the desired heterologous gene. Being a promoter, there is also no requirement that it be in the coπect reading frame, just within the appropriate promoter distance.

It is generally preferable that the transformed fungus also expresses a linked selectable marker. Any marker known in the art may be used, and may be excised once a faithful strain has been generated. However, it is generally preferable that the transformed fungus maintains a marker to ensure that the desired heterologous product is still produced, and to ensure that there is no reversion to wild type. In this respect, it is prefened that the marker have no significant negative effect on either the fungus or the product. Such markers may normally be selected from resistance markers, in order that the growth medium contain amounts of an antifungal agent ensuring that only transformed fungus can grow successfully.

Suitable markers include the hygromycin resistance cassette and the benomyl resistance tubulin gene.

Suitable methods for transforming filamentous fungi are as described above with respect to WO95/02691 and WO98/45455, which disclosures are incoφorated herein by reference.

In general, the desired control sequences are ligated with the appropriate heterologous expression sequences and prepared for insertion into a suitable preparation of the fungus, such as protoplasts, all by methods well known in the art. The resulting organism can then be grown by standard methods, and prepared as spawn after cultivation of the resulting mycelium. Spawn has the advantage that it can be stored inert for relatively long periods of up to about a year, although it is generally prefened to use it within about 4 months.

Spawn may be produced in any recognised manner, such as by growing the mycelium on sterile agar and introducing the culture to autoclaved grain. The grains may then be stored at elevated temperature to encourage colonisation, and then kept at reduced temperatures until needed.

It will be appreciated that the heterologous gene for incoφoration may be in the form of cDNA or genomic DNA. cDNA is prefened, as it is generally shorter and more easy to handle.

It will also be appreciated that the heterologous gene insert should encode the sequence desired, including leader sequences and cleavage sequences, if required.

It will also be appreciated that greater expression may be achieved if fungal codons are used in place of mammalian codons, although expression will still occur, and such substitution is not necessary.

It will further be appreciated that heterologous genes may need to be expressed in the form of a cassette, for example, in order to produce the required product. In general, it is prefened to require as few heterologous gene products as possible, as the greater the number, the more likely it is that the fungal metabolism will interfere in some way, and it is generally desirable to minimise unpredictability.

Thus, although not limited thereto, it is generally prefened to limit the number of heterologous expression products to one, two or three, preferably one or two, and preferably one, other than any marker. The marker is preferably linked to the heterologous gene, such as downstream of the gene and also under the control of the fungal promoter, so as best to indicate successful and/or continuing stable transformation.

It is particularly prefened that products such as peptides be the target, as these can be harvested relatively simply. Thus, enzymes and antibodies are particularly useful, although conformational proteins, such as vaccine antigens, and active peptides such as interferons are also useful. Accordingly, heterologous genes suitable for expression in the filamentous fungi include those whose expression results in the production of: antibodies, including other diagnostic material; secondary metabolites, such as lectins, pesticidal compounds such as Bacillus thuringiensis toxin (Bt toxin); therapeutic compounds such as vaccines, steroids, heterocyclic organic compounds; biological macromolecules, such as interferon, endostatin and insulin; and medical enzymes, such as thrombolytics and cerebrosidases.

In the context of the present invention, the term "heterologous" includes native DNA not normally associated with selective expression, especially heightened expression during sporophore production. In such a respect, the native gene becomes heterologous insofar as its expression pattern is altered. Such expression may generally serve one of two puφoses. The first is generally to obtain large/greater amounts of native protein, such as by transforming the filamentous fungi with extra copies or modified copies of a native gene or genes. The second may be used instead to affect/control the characteristics of mushroom crop production, such as by altering the timing of crop, flushing pattern, yield, growth rate and/or final size of the mushroom sporophore. This latter may also suitably be achieved by the introduction of heterologous DNA from other species, if desired.

The crops are preferably allowed to go to full cap development, where possible, in order to maximise expression of the heterologous gene, although the skilled person will appreciate the best stage for harvesting any given product. The resulting caps may then be processed in any suitable manner to extract and/or purify the product, or the caps may otherwise be employed or processed, as desired.

Where the product is potentially dangerous, standard procedures may be employed between crops to entirely sterilise the area, such as steam sterilisation and swabbing of the walls, as described above.

The present invention will now be illustrated with reference to the accompanying figures, in which:

Figure 1 : Alignment of sequenced Sprl cDNA and published Sprl cDNA Comparison of sequenced Sprl sequence (top line, SEQ ID NO 1) with published Sprl sequence (bottom line, SEQ ID NO 2). On sequencing the Sprl cDNA present in plasmid pKING03, discrepancies were discovered between this sequence and that published (Accession, no. Y13805) as illustrated in bold type. These differences affect the amino acid sequence of the protein, as illustrated in Figure 2.

Figure 2: FAST amino acid sequence comparison of Agaricus bisporus Sprl with other fungal proteases

The predicted amino acid sequence of sequenced and published Sprl is compared, with other fungal proteases. Regions shown shaded with black have 100% homology, regions in darker shading have >75% homology, and the lighter shaded regions show >50 % homology. Similar amino acids are also shaded. Differences in amino acid sequence between the published and sequenced Sprl are shown in boxes.

Figure 3: PCR analysis of putative recombinant pAN7-lA:S/wi plasmids

3 A. Diagrammatic representation of PCR primer positions in recombinant plasmids pANsense and pANstop. 3B. Diagrammatic representation of PCR primer positions in recombinant plasmids pANantisense. 3C. PCR of putative recombinant plasmids pANsense, pANstop and pANantisense.

Figure 4: PCR analysis of putative recombinant pBU004:S/w7 plasmids

4 A. PCR of putative recombinant plasmids p004sense and p004stop using primers 004-pl and Sprl-pz, which respectively anneal within the promoter region and the Sprl region of recombinant plasmids to yield a lkb PCR product. 4B. PCR of putative recombinant plasmid p004antisense using primers 004-pl and Sprl-px, which respectively anneal within the promoter region and the Sprl region of the recombinant plasmid to yield a 1.2kb PCR product. Figure 5: Diagrammatic representation of successful cloning strategy 3 for the construction of hph and Sprl containing binary plasmids

Step A. The hph cassette from phph004 was cloned from pBluescript to pBluescriptll using Sac I and Kpn I. Step B. The hph cassette was then cut with Bss HII and Kpn I and cloned into similarly digested p004Sprl plasmids, which linked the Sprl and hph cassettes (Step C). Step D. The whole fragment was then excised with Bgl II and Spe I and ligated to pGREEN cut with Bam HI and Spe I

Figure 6: Transfer of phph004 cassette to pBluescriptll backbone.

6A. Plasmid phph004 digested with BssH II, which does not cut within this plasmid, and hence DNA was visualised on an agarose gel as an undigested smear. 6B. Putative plasmid phph004BSII digested with BssH II, which excises the 2.3kb hph cassette from the 2.9kb plasmid backbone, confirming successful construction.

Figure 7: Diagrammatic representation of positions of PCR primers and fragment sizes used for analysis of putative Sprl Agaricus bisporus transformants

7A. Expected PCR fragment sizes with pGRsensehph or pGRstophph transformed mycelia using primer pairs 004-pl / Spr-z (1075bp) and Spr-x / Spr-z (750bp). 7B. Expected PCR fragment sizes with pGRantihph transformed mycelia using primer pairs 004-pl / Spr-x (1230bp) and Spr-x / Spr-z (750bp).

Figure 8 PCR screening for hygromycin resistance marker and endostatin transgene.

Lane contents are as follows: (1) 1 kb marker; (2 and 10) Positive control for hygromycin/endostatin transgenes; (3 and 11) water control; (4 and 12) A. bisporus non-transformed control; (5 - 9) hygromycin gene (987 bp) and (13 - 17) endostatin gene (574 bp) fragments from transformed mycelia. The present invention will now be illustrated further, by reference to the following, non-limiting Examples.

EXAMPLE 1

A serine proteinase cDNA (Sprl) has previously been cloned and partially characterised and was available for use (Kingsnorth et al, 2001).

PTGS may be exploited by providing a construct with two opposing copies of the transgene, designed to form a double stranded RNA molecule, or haiφin loop, and trigger gene silencing (Cogoni et al, 2001). Another, more straightforward, method is to transform the transgene in the opposing direction to the native gene (antisense), thus forming a double stranded RNA species on expression of the native gene. As the mechanisms of PTGS are not completely understood, and work of this type has not previously been conducted on a Homobasidiomycete fungus, several constructs were devised. For Sprl, an "antisense" construct was designed, as well as a "sense" construct (normal orientation) and a "stop" construct (normal orientation, but containing a stop codon to prevent translation).

The method selected to transform Agaricus bisporus with serine proteinase (Sprl) cDNA, was to flank the DNA with regulatory sequences and to link this cassette to a marker, in this case, a hygromycin phosphotransferase cassette, for positive selection of transformants. Three construct types were designed: sense, stop, and antisense. The sense construct has the unadulterated Sprl cDNA cloned between the A. bisporus gpdll promoter and Aspergillus nidulans trpC terminator. The stop construct is the same, with the exception that a stop codon is introduced to prevent translation. The antisense construct has the unadulterated cDNA, but in the reverse direction.

PCR primers were designed to introduce necessary modifications and restriction sites at the cDNA ends to facilitate cloning. This requires an Nco I site at the 5' end, and a BamH I site at the 3' end of the gene to be cloned. However, the Sprl cDNA already contains an internal Nco I site, so Afl HI or BspH I sites were engineered instead to give compatible cohesive ends with Nco I digested DNA. Primers are shown in Table 1.

Table 1: Primers used to amplify Sprl cDNA

Primer Sequence SEQ ID Details NO.

Sprl-pl ccc gtc atg atg cat ttc tct ttg tct 28 Sense 5' BspH I site

Sprl-p2 ccc eea tec gca aaε tεt ata ttc ctt 29 Sense/Stop 3' BamH I site

Sprl-ρ3 ccc etc atg ate cat ttc tct tAg tct 30 Stop 5' BspH I site

Sprl-p4 ccc εεa tec ace εac gat gca ttt etc 31 Antisense 3' BamH I site

Sprl-p5 ccc aca tgt gca aag tgt ata ttc ctt 32 Antisense 5' Afl III site gac

Engineered restriction sites are shown underlined. Sprl cDNA sequence is shown in bold font. A base change from T to A in primer Sprl-p3 to create a stop codon is represented in upper case type.

The primers detailed were used to amplify the Sprl cDNA from pKING03, which consists of the cDNA cloned into pBK-CMV. Primer pairs used to yield "sense", "stop", and "antisense" versions of the cDNA are shown in Table 2.

Construct Primer pair Details type Sense Sprl -pi and Sprl-p2 coding cDNA with 5' BspH I site and 3' BamH I site Stop Sprl-p3 and Sprl-p2 non-coding cDNA with 5' BspH I site and 3' RαmH I site Antisense Sprl -p4 and Sprl -p5 antisense cDN A with 5' Afl III site and 3' BamH I site

Table 2: Primer pairs used to amplify Sprl cDNA for three construct types

PCR was carried out as described in section 2.2.3.4, but with a thermal cycle as follows: 95°C, 5 min; (94°C, 1 min; 45°C, 1 min; 72°C 1 min) x 30; 72°C, 15 min.

PCR products were separated by gel electrophoresis. The purified PCR products were cloned into the PCR vector pGEM-T-Easy. Recombinant DNA was isolated from transformed E. coli cultures and confirmed by sequencing.

Vector System pGEM-T Easy developed by Promega is suitable to clone PCR products (in the size range of 80 bp to 3.5 kb) with 3' A overhangs generated by Taq polymerase during PCR. The pGEM-T Easy is a linearised vector with 5'T overhangs used for ligating the 3' A overhangs in the PCR product. After a standard PCR reaction, the products are purified (e.g. using Qiagen spin columns) to remove the PCR components and then ligated with the pGEM-T Easy vector. The ligation mix can then be transformed into a suitable E. coli strain and blue/white screening can be used to identify positive transformants containing the cloned PCR product.

Using this system, PCR products can be ligated and transformed rapidly (approximately within 2 hours of amplification). The presence of inserts can be distinguished using blue white screening, and to determine insert size, the 5' and 3' flanking Not I restriction sites can be used, since Not I is an eight base cutter and is not frequently found within genomes. If necessary there are other unique flanking restriction sites to pick from or use vector primers for PCR based screening of putative white colonies (positive transformants). On sequencing recombinant Sprl clones, disparities were discovered between the obtained sequences and the published Sprl sequence (Kingsnorth et al, 2001; Genbank accession number Y13805). These discrepancies were verified by sequencing of the Sprl cDNA region from plasmid pKing03. Comparison of these sequences is shown in Figure 1. The amino acid sequences of the published and sequenced Sprl cDNA were compared with similar fungal proteases (Figure 2), revealing that A. bisporus Sprl is 11 amino acids longer than previously thought, and is more similar to other fungal proteases than previously realised.

The Sprl cDNA was used to replace the hygromycin phosphotransferase gene (hph), in plasmid pAN7-l A, and the luciferase gene in plasmid pBU004. Plasmid pAN7-l A contains the Aspergillus nidulans gpd promoter, and pBU004 has the Agaricus bisporus gpdll promoter region, with both plasmids utilising the terminator region from A. nidulans trpC. These promoters are known to be effective in A bisporus transformation. The Sprl fragments were excised from cloning plasmid pGEM-T-Easy, using the appropriate enzymes, and ligated into Nco I / BamH I digested toolkit plasmids to yield the following; pANsense, pANstop, pANantisense, p004sense, p004stop, p004antisense. Although this cloning strategy was seemingly straightforward, problems were experienced with low DNA yields and rapid death of E. coli colonies containing recombinant plasmids.

In Figure 2, the predicted amino acid sequence of sequenced and published Sprl is compared, with other fungal proteases. Regions shown shaded with black have 100% homology, regions in darker shading have >75% homology, and the lighter shaded regions show >50 % homology. Similar amino acids are also shaded.

Differences in amino acid sequence between the published and sequenced Sprl are shown in boxes. There are minor discrepancies at positions 208, 291, 293 and 329, but the main difference is the additional 11 amino acids at the end of the sequence. This additional length shows the serine proteinase amino acid sequence to be similar in length and constitution to other fungal proteases, but it was previously thought to be significantly shorter.

Amino acid comparisons are shown between: Agaricus bisporus sequenced serine proteinase (A.b. Sprl), published A. bisporus serine proteinase, accession no. Y13805 (A.b. Sprl (pub)); Metarhizium anisopilae var. acridum serine proteinase PR1A, accession no. CAB63911 (M.a.PRIA); Verticillium chlamydosporium var. chlamydosporium alkaline serine protease, accession no. CAD20578 (V.c. ASpr); M. anisopilae cuticle degrading protease precursor Prl, accession no. P29138 (M.a. chem.).

PCR analysis of putative recombinant pAN7-lA:S >ri plasmids is shown in Figure 3. pAN7-l A:Sprl plasmids were produced by replacing the hygromycin phosphotransferase gene (hph) in pAN7-l A with Sprl cDNA in sense, stop and antisense formats. Putative recombinant plasmids were analysed by PCR using primers gpd-pl, Sprl-px and Sprl-pz.

Figure 3 A is a diagrammatic representation of PCR primer positions in recombinant plasmids pANsense and pANstop. Successful construction of these plasmids would yield 700bp and 920bp products.

Figure 3B is a diagrammatic representation of PCR primer positions in recombinant plasmids pANantisense. Successful construction of this plasmid would yield 700bp and l.lkb products.

Figure 3C shows PCR of putative recombinant plasmids pANsense, pANstop and pANantisense. All of the samples produced the 700 bp product expected from the Sprl-px and Sprl-pz primers, which conesponds with a similar product when pKing03 (which contains the Sprl cDNA) was used as the template. pANsense and pANstop produced an additional 920bp product as illustrated in panel A, confirming successful construction. pANantisense had an additional l.lkb fragment as illustrated in panel B, confirming successful construction.

PCR analysis of putative recombinant pBU004:Sjσri plasmids is shwn in Fuigure 4. pBU004:S r/ plasmids were produced by replacing the luciferase gene in pBU004 with Sprl cDNA in stop, sense and antisense formats. Putative recombinant plasmids were analysed by PCR using primer 004-pl with Sprl-px or Sprl-pz.

Figure 4 A shows PCR of putative recombinant plasmids p004sense and p004stop using primers 004-pl and Sprl-pz, which respectively anneal within the promoter region and the Sprl region of recombinant plasmids to yield a lkb PCR product. The presence of this product confirmed successful construction of these plasmids. Figure 4B shows PCR of putative recombinant plasmid p004antisense using primers 004-pl and Sprl-px, which respectively anneal within the promoter region and the Sprl region of the recombinant plasmid to yield a 1.2kb PCR product. The presence of this product confirmed successful construction of this plasmids.

Cloning Sprl constructs into binary plasmid: Strategy 1

Once successfully linked to the desired regulatory elements, Sprl cassettes needed to be transfened to a binary plasmid containing a hygromycin resistance cassette for primary transformant selection. The initial strategy devised was as follows: Hpa I and Kpn I restriction sites would be engineered at the termini of the Sprl cassettes using PCR, and used to clone the cassettes into similarly digested pGRhph004 (pGRhph004 consists of phph004 cloned into the binary plasmid pGREEN). Primers used are shown in Table 5.3. Problems were encountered while attempting to amplify the Sprl cassettes by PCR. PCR products of the desired size were either absent or formed smeared patterns on the agarose gel. PCR was carried out at different annealing temperatures and with more cycles, with different template DNA concentrations, and with longer PCR primers. None of these conditions produced PCR products suitable for further cloning. In addition, PCR with a different taq polymerase was attempted, according to the manufacturer's instructions (AmpliTaq Gold, Perkin Elmer). This sensitive polymerase is added to the PCR reaction once the samples have been heated. Given that the presence of proteases can be inhibitory to the polymerase enzyme, it is possible that thus Sprl could have been inhibiting the PCR reaction. Pre-heating the PCR reaction mixture destroys any proteases before the enzyme is introduced. However, these PCR reactions also failed to yield satisfactory PCR products.

Cloning Sprl constructs into binary plasmid: Strategy 2

Difficulties were experienced when attempting to introduce restriction sites into Sprl cassettes by PCR in Strategy 1, so a new strategy was devised whereby restriction sites would be introduced by initially cloning the Sprl plasmids into a polylinker plasmid, pSLl 180. This cloning was possible with the plasmids containing the Agaricus bisporus gpdll promoter (p004sense, p004stop and p004antisense) using restriction enzymes Sac II and Kpn I. However, the constructs containing the Aspergillus nidulans gpd promoter (pANsense, pANstop and pANantisense) were not amenable, and cloning attempts with these plasmids were discontinued.

Primer Sequence SEQ ID NO. Hρa-gpd-004 cccgttaacgaagaagaattcaga 15 Hpa-gpd-pan cccgttaacgaattcccttgtatc 16 Kpn-tφC cccggtacctcgagtggagatgtg 17 Hpa-004-long cccgttaacgaagaagaattcagaggtccg 18 Hpa-pan-long gctgttaacggcacacaggctcaaatcaat 19 Kpn-tφC-long caaggtacctcgagtggagatgtggagtgg 20

Table 3: Primers designed to amplify promoter-S/w-i-terminator cassettes and introduce Hpa I and Kpn I restriction sites

No PCR products were seen when using primers designed to amplify the whole promoter-S_jpri -terminator region were used (Hpa-pan-long and Kpn-tφC-long or Hpa-004-long and Kpn-tφC-long where appropriate), although a conectly sized fragment was produced when plasmid phph004 was used as a PCR template. Internal Sprl primers (Sprl-px and Sprl-pz) produced 700bp PCR products as expected.

Cloning the p004Sprl plasmids into pSLl 180 allows the blunt cutting enzyme Sea I to become available for cloning. The Sprl cassettes could then be excised using Sea I and Kpn I and cloned into pGRhph004 digested with Hpa I and Kpn I.

The initial transfer of Sprl cassettes into pSLl 180 was successful, but attempts to clone the Sprl fragments into the binary plasmid pGRhph004 failed. Difficulties were experienced with preparation of sufficient quantities of DNA. The DNA that was produced was often of dubious quality and this may have caused problems with the blunt ended ligation of DNA molecules. The failure to produce pure DNA may have been in some way related to problems experienced when culturing E. coli colonies containing Sprl plasmids. When cultured on nutrient agar, colonies were smaller and malformed in comparison to colonies carrying other plasmids, and liquid cultures were visibly less turbid. In addition, colonies stored on agar medium at +4°C rapidly became unviable, suggesting the Sprl cDNA presence causes premature cell death. Problems of this type have been reported previously (Kingsnorth et al, 2001) and these difficulties were noted with plasmids containing all three orientations of the Sprl cDNA.

Cloning Sprl constructs into binary plasmid: Strategy 3

Given the difficulties experienced when attempting to clone blunt-ended Sprl DNA fragments, a third strategy was devised, using only sticky-ended cloning. This strategy involves the combination of the Sprl cassette and hygromycin resistance cassette, then transfer of this combined fragment into the binary plasmid pGREEN and is illustrated in Fig. 5.

Firstly, the hph cassette from phph004 was excised from the pBluescript backbone using restriction enzymes Kpn I and Sac I, and cloned into similarly digested pBluescriptll. These two backbone plasmids are very similar, with the exception that pBluescriptll has additional BssH II sites flanking the multiple cloning nest (Fig. 5.16). The hph cassette was then excised from the new backbone using enzymes Kpn I and BssH II. This fragment was then ligated to similarly digested pSL004Sprl plasmids (pSL004sense, pSL004stop and pSL004antisense), thus linking the hph and Sprl cassettes (to create pSLsensehph, pSLstophph, pSLantihph). E. coli colonies containing recombinant molecules were identified by colony hybridisation and further analysed by restriction digest.

The appropriate fragments were then excised using Bgl II and Spe I and ligated to pGREEN cut with BamH I and Spe I to form binary plasmids pGRsensehph, pGRstophph and pGRantihph. These binary plasmids were transformed into Agrobacterium tumefaciens strains LBA1126 and AGL-1.

This strategy was successful in the construction of binary plasmids pGRsensehph, pGRstop-hph and pGRantisense-hph, which contain a hygromycin phosphotransferase resistance cassette and Sprl cassette in the binary vector pGREEN. The promoter-A 7 2-terminator cassette from phph004 was excised from the pBluescript backbone and cloned into pBluescriptll using restriction enzymes Sac I and Kpn I to create phph004BSII. This transfer introduces two BssH II restriction sites flanking the multiple cloning nest, and hence the hph cassette.

Figure 6 shows the transfer of phph004 cassette to pBluescriptll backbone. Column A shows plasmid phph004 digested with BssH II, which does not cut within this plasmid, and hence DNA was visualised on an agarose gel as an undigested smear. Column B shoes putative plasmid phph004BSII digested with BssH II, which excises the 2.3kb hph cassette from the 2.9kb plasmid backbone, confirming successful construction.

Transformation of Agaricus bisporus with Sprl plasmids

Agaricus bisporus was transformed with plasmids pGRsensehph, pGRstophph and pGRantihph, with both Agrobacterium tumefaciens strains AGL-1 and LBAl 126, as described in section 2.2.6.2. Hygromycin resistant colonies were subcultured and used for further analysis. Colonies were subcultured alternately onto hygromycin containing media, or media without antibiotic, to confirm stability of transformation. The number of antibiotic resistant colonies produced was low in comparison with a positive control, pBGgHg (Table 4)

Agrobacterium Plasmid No. of hygromycin % of gill tissue strain resistant colonies pieces transformed

LBAl 126 0 0

LBAl 126 pBGgHg 80 80%

LBAl 126 pGRsense-hph 1 <1%

LBAl 126 pGRstop-hph 7 6.6%

LBAl 126 pGRanti-hph 10 8.7% AGL-1 0 0

AGL-1 pBGgHg 72 72%

AGL-1 pGRsense-hph 7 7.1%

AGL-1 pGRstop-hph 10 9.0%

AGL-1 pGRanti-hph 4 4.3%

Table 4: Transformation rates of Agaricus bisporus transformed with Sprl plasmids Colonies which hybridised to both DNA species, i.e. 11,12 and 34 (putative pSLsensehph), 101 and 111 (putative pSLstophph) and 129 (putative pSLantihph) were selected for further analysis by restriction digest.

Plasmid DNA was prepared from these colonies and digested with restriction enzymes Bgl II and Spe I. In recombinant molecules, these enzymes excised the Spτl-hph cassette as a 4.2kb fragment, leaving the pSLl 180 backbone as a 3.2kb fragment. This pattern was seen for colonies 11 and 34 (pSLsensehph) and for colonies 101 and 111 (pSLstophph), confirming successful construction. pSLantihph was not successfully created in this reaction.

E. coli colonies containing putative recombinant plasmids with Sprl -antisense and hph cassettes were identified by colony hybridisation (not shown).

DNA was prepared from these colonies and digested with restriction enzymes Bgl II and Spe I. In the parent molecule pSL004antisense, the Sprl cassette is excised as a 2.4kb fragment, with the pSLl 180 backbone as a 3.2kb fragment. In recombinant molecules, these enzymes excised the Sprl -hph cassette as a 4.2kb fragment, leaving the pSLl 180 backbone as a 3.2kb fragment, confirming successful creation of pSLantihph.

On digestion with Hpa I and Sac I, binary plasmids pGRsensehph and pGRstophph would be expected to produce DNA fragments of 3.9kb, 640bp, and 325bp from the Sprl-hph cassette, along with a 4kb pGRΕΕN backbone fragment. For pGRantihph, the fragment sizes would be 3.35kb, 1.2kb, 325bp, and 4kb due to the Sprl cDNA being in the opposite orientation. There are also a number of very small backbone fragments which would not be visible on an agarose gel. The 325bp fragment is characteristic of the cassette being present in the binary plasmid pGRΕΕN. In pSLsensehph, pSLstophph, and pSLantihph, where the backbone is the polylinker plasmid pSLl 180, there is no Hpa I site, and so cassette fragments will be present, along with a 3.2kb backbone fragment but there is no 325bp fragment. phph004 would be linearised in this digestion. Digestion of pGRΕΕN will produce the 4kb backbone fragment and the characteristic 325bp fragment.

Digestions were performed and visualised by agarose gel electrophoresis. Putative recombinant plasmids were digested with Hpa I and Sac I.

In the restriction analysis of plasmids pGRsensehph and pGRstophph, all putative plasmids produced the expected fragment pattern, including the 325bp fragment indicative of the pGREEN backbone and the 640bp fragment produced from the Sprl cassette. The remaining fragment visible on the gel image resolved further with time, into two fragments of 3.8kb and 4kb, those of the Sprl -hph cassette and pGREEN backbone respectively.

In the restriction analysis of plasmid pGRantihph, all putative plasmids produced the expected fragment pattern, including the 325bp fragment indicative of the pGREEN backbone and the 1.2kb fragment produced from the Sprl cassette. In addition, the 3.35kb Sprl -hph fragment and 4kb pGREEN fragments were present.

Analysis of Agaricus bisporus Sprl transformants by polymerase chain reaction

Agaricus bisporus was transformed with Sprl plasmids pGRsensehph, pGRstophph, and pGRantihph. DNA was prepared from all the transformants generated (8 for pGRsense-hph, 17 for pGRstop-hph, and 14 for pGRanti-hph) and used for PCR analysis. Primers used are provided in Table 5, with a diagrammatic representation of primer positions and fragment sizes in Fig. 7.

Prim Sequence SSEEQ ID NO. PCR product details er gdhl cgccgcggggaat 21 Genomic primers that amplify a 500 bp ggaattacgccgct region of glutamate dehydrogenase A from cggg Agaricus bisporus gdh2 ggcgggatccgtg 22 aggaaggaccatg gtgta

Hyg gcgtggatatgtcct 23 Internal primers which anneal within

1 gcggg hygromycin phosphotransferase and

Hyg ccatacaagccaac 24 produce a 600bp fragment

2 cacgg

Spr- gccaacttcaaggc 25 Primers anneal within Sprl cDNA to yield

X caaggt a 750bp PCR product

Spr-z agaatgaagcacg 26 agcgtcg

004- cccgcgtctcgaat 27 Anneal within the promoter and Sprl pl gttctc regions of pGRsense-hph and pGRstop-hph Sprz as above 27 regions of pGRsense-hph and pGRstop-hph to yield a 1075 bp product

004- as above 26 Anneal within the promoter and Sprl pl regions of pGRanti-hph to yield a 1230 bp Sprx as above 27 product

Table 5: PCR primers used in analysis of Agaricus bisporus Sprl transformants

DNA from all transformants, and from A. bisporus wild type mycelia, was confirmed as suitable for PCR by analysis with genomic primers gdhl and gdh2, which anneal within the A. bisporus glutamate dehydrogenase A gene. All colonies were also confirmed as transformed by PCR with primers hygl and hyg2, which anneal within the hygromycin transferase gene used in transformation for primary antibiotic selection. Transformants were also assessed for presence of bisporus serine proteinase 1 (Sprl) cDNA using primers Spr-x and Spr-z, which amplify a 750bp region of the transgene. However, not all transformant colonies yielded PCR products, with only 3 of 8 pGRsensehph transformants, 6 of 17 pGRstophph transformants, and 7 of 14 pGRantihph transformants being PCR positive with these primers. Primers Spr-x and Spr-z anneal within the Sprl cDNA, and may also anneal within the genomic Sprl gene, assuming that no introns interrupt the primer sequences.

PCR with pGRantihph transformants exhibited additional PCR fragments of ca 1.3kb, and this product was also observed when wild type DNA was used as a template. A. bisporus genes contain frequent introns of ca. 50bp in length, so this 1.3kb PCR product is likely to have been produced from the genomic Sprl gene. One of the pGRantihph transformants, Al, yielded Sprl cassette specific PCR products but no genomic product, which may suggest that a gene replacement event has occuned. However, PCR with pGRsensehph and pGRstophph transformed mycelia did not yield this additional fragment, and nor did wild type controls in these reactions despite use of the same wild type sample as utilised in the pGRantihph PCR. PCR reactions with pGRantihph transformants were conducted at a slightly lower annealing temperature (50°C as opposed to 55°C) which may have accounted for the presence of the genomic product. Additional PCR analysis was performed on all transformants using 004-pl, which anneals within the bisporus gpdll promoter, and either Spr-z (for pGRsensehph and pGRstophph transformants) or Spr-x (for pGRantihph transformants). These primer pairs are specific for the transforming cassettes only and no PCR products would be expected from the remaining genomic DNA. Some of the colonies that were PCR positive with primer pair Spr-x and Spr-z did not yield a product in this additional PCR reaction, but all colonies which produced PCR products with the cassette specific primers had also previously yielded the 750 bp PCR product with primers Spr-x and Spr-z.

These PCR results suggest difficulties with the PCR conditions used and these experiments would have been repeated had time not been limited. Primers Spr-x , Spr-z and 004-pl were all 20bp in length, which is possibly too short for use with genomic DNA. In addition, the differing results obtained when using Spr-x and Spr-z between experiments suggests that the PCR conditions may require optimisation. Thus, while presence of PCR products in these reactions confirm transformation, absence of amplified fragments cannot be taken as a negative result. However, there was good conelation between PCR experiments with different primer pairs, which may indicate that the PCR results observed could be accurate. Presence of PCR products in some cases with the Sprl specific primers but not with cassette specific primers could indicate T-DNA truncation.

In the above experiments, the results were as follows:

When DNA was extracted from hygromycin resistant colonies of Agaricus bisporus transformed with hph-Sprl plasmids, all transformant colonies tested, and wild type DNA, produced the expected fragment, using genomic primers gdhl and gdh2 which amplify a 500bp region of the A. bisporus glutamate dehydrogenase gene.

Using primers hygl and hyg2, which amplify a 600bp fragment from the hygromycin phosphotransferase transgene (hph), all the transformant colonies tested produced the expected fragment, along with a positive control of the plasmids used for the transformation. As expected, PCR of wild type DNA with these primers yielded no PCR product.

DNA from hygromycin resistant colonies of Agaricus bisporus transformed with plasmid pGRsensehph was subjected to PCR performed with two primer pairs which anneal within the Sprl cassette. Using Spr-x and Spr-z, which amplify an 750bp region of the Sprl cDNA, three transformants, SE2, SE6 and SE7, yielded a 750bp PCR product conesponding to the presence of Sprl cDNA. However, no transformants were seen when PCR was performed with 004-pl and Spr-z, which anneal in the bisporus gpdll promoter region and Sprl cDNA region of plasmid pGRsensehph respectively to yield a 1075bp PCR product.

With DNA from hygromycin resistant colonies of Agaricus bisporus transformed with plasmid pGRstophph, PCR was performed with two primer pairs which anneal within the Sprl cassette. PCR performed with Spr-x and Spr-z, which amplify an 750bp region of the Sprl cDNA, showed no transformants, while PCR performed with 004-pl and Spr-z, which anneal in the A bisporus gpdll promoter region and Sprl cDNA region of plasmid pGRsensehph, respectively, to yield a 1075bp PCR product, showed transformants ST6, ST7, ST14 and ST15 having a 1075bp PCR product conesponding to the presence of the promoter-Sprl cassette.

With DNA from hygromycin resistant colonies of Agaricus bisporus transformed with plasmid pGRantihph, PCR was performed with two primer pairs which anneal within the Sprl cassette. PCR performed with Spr-x and Spr-z, which amplify an 750bp region of the Sprl cDNA, showed transformants Al, A5, A5, A8, A9, AlO and Al 1, all having a 750bp PCR product conesponding to the presence of Sprl cDNA. In addition, all transformants with one exception yielded a further ca. 1.3kb fragment which could conespond with intron-containing genomic Sprl . PCR performed with 004-pl and Spr-z, which anneal in the A bisporus gpdll promoter region and Sprl cDNA region of plasmid pGRsensehph, respectively, to yield a 1075bp PCR product, also showed transformants Al, A5, A5, A8, A9, AlO and Al 1 as yielding a 1075bp PCR product conesponding to the presence of the promoter- Sprl cassette.

Qualitative assay of proteinase activity in Agaricus bisporus Sprl transformants

Agaricus bisporus was transformed with Sprl plasmids pGRsensehph, pGRstophph, and pGRantihph, and several hygromycin resistant colonies were isolated for each. Transformation was confirmed by PCR with primers which anneal within the hygromycin phosphotransferase (hph) region of the constructs. PCR was also performed using primers specific for the Sprl plasmid region, but not all the transformant colonies yielded products.

All the A bisporus transformants were assessed for proteinase activity using a qualitative plate assay, which involved subculture of transformant colonies onto nutrient agar with an overlay containing either milk or gelatin. Plates were incubated at 25°C and examined after 10 days and 17 days. The milk layer forms an opaque overlay, so proteinase activity could be discerned by a clear zone sunounding the colony at 10 and 17 days. Digestion of gelatin could only be assessed by staining with amido black, which kills the mycelia, and so was only performed after 17 days incubation. Clearing zones were measured from the edge of the mycelial colony to the edge of the clearing zone. Where clearing zones were inegular, an average was taken. Wild type mycelia, and Agaricus bisporus GFP (Green Fluorescent Protein) transformants were also tested.

Two separate experiments were performed in which different concentrations of protein were used. Initially, a 5ml overlay of either 1% milk or 0.2% gelatin was used, but the milk clearing zone was occasionally difficult to distinguish. In a second experiment, a 10ml overlay of either 2% milk or 4% gelatin was used, which made the milk and gelatin clearing zones easier to distinguish, but reduced the size. Several of the A bisporus transformants failed to produce clearing zones on these media in both experiments.

All wild type cultures tested, and all non-Sprl transformants exhibited clearing zones on both types of media. In contrast, several Sprl transformed Agaricus bisporus cultures had no clearing zones. Three of eight pGRsensehph transformants (SE1, SE4, SE5) had no clearing zone on either medium, and one colony (SE7) was only able to digest gelatin. Five of 17 pGRstophph transformants (ST4, ST6, ST8, ST9 and ST11) had completely reduced clearing zones on both media, with a further four transformants (ST7, ST 15, ST 16, ST 17) diminished in their ability to digest either milk or gelatin. Five of 14 pGRantihph transformants (Al, A3, A5, AlO and A13) had no clearing zone on either medium, and two further transformants (A2, A14) were only able to digest gelatin. There was no consistent conelation between colonies with reduced proteinase activity and those which produced DNA fragments in PCR experiments with Sprl primers. However, five of the colonies that yielded Sprl PCR products had diminished clearing zones, namely pGRstophph transformants 4 and 6, and pGRantihph transformants 1, 5 and 6.

Discussion

Development of transformation constructs containing serine proteinase 1 (Sprl) cDNA

Three transformation constructs were devised for Agrobacterium tumefaciens mediated transformation of Agaricus bisporus with native serine proteinase 1 (Sprl) cDNA, namely pGRsensehph, pGRstophph, and pGRantihph, which contained unadulterated Sprl cDNA, untranslatable Sprl cDNA, and Sprl cDNA in the antisense direction respectively. These binary plasmids also contain a hygromycin phosphotransferase (hph) cassette for primary selection of putative transformants. Problems were encountered during development of these constructs, including difficulties in culture of E. coli containing Sprl -plasmids, a lack of ability to use Sprl DNA as a template for PCR reactions, difficulties in preparation of Sprl DNA from E. coli cultures, and problems with restriction digestion of any resulting DNA. Kingsnorth et al, (2001) also noted a deleterious physiological effect in growth of E. coli containing Sprl cDNA. This is unlikely to be due to overexpression of serine proteinase, as these difficulties were experienced for plasmids containing all three orientations of Sprl. The literature does not reveal other fungal serine proteinases causing such problems, but they may be unreported.

Transformation of Agaricus bisporus with Sprl plasmids

Agaricus bisporus was transformed with plasmids pGRsensehph, pGRstophph, and pGRantihph, with low, in comparison with a positive control, but successful, transformation rates seen for each. This low rate does not seem to be related to the Agrobacterium tumefaciens strain used, or constituents of the plasmids, as similarly low rates were experienced in transformation of A bisporus to GFP expression (data not shown). The size of the T-DNA is also unlikely to be a factor, as the positive control used (pBGgHg) is of a similar magnitude (Chen et al, 2000). The main difference between these plasmids is the binary construct backbone used, which is pCAMBIA in pBGgHg, whereas pGREEN is used here. However, other workers have reported high transformation rates when using pGREEN. During use, it was noticed that the size of pGREEN increased by ca. 600-700bp, which is similar to problems noted by the manufacturers, who state that pGREEN has a tendency to incoφorate E. coli chromosomal DNA in the ColEI replication region (www.pgreen.co.uk). There is a possibility that this decreased stability may have a detrimental effect on transformation rate and moving transformation cassettes to the more stable pGREENIl, or another binary plasmid, may increase transformation rates.

Integration of T-DNA into the Agaricus bisporus genome

Agaricus bisporus was transformed with plasmids pGRsense-hph, pGRstophph, and pGRantisense-hph and a number of stable hygromycin resistant colonies were isolated for each, from which DNA was prepared and used for PCR analysis. Successful transformation of all these colonies was confirmed by PCR with primers specific for the hygromycin phosphotransferase gene present in the plasmid. However, only some of these colonies were found to be PCR positive when tested with primers specific for the serine proteinase region of the constructs. In addition, some of the transformants yielded PCR products with primer pair Spr-x and Sprz, which anneal within the Sprl cDNA, but failed to amplify when an additional primer annealing with the promoter region was used. pGRantihph transformants yielded a genomic product in most cases, but one transformant had a cDNA specific PCR product but no genomic product, suggesting a possible gene replacement event.

Thus, PCR products confirm the presence of the Sprl cDNA, but the absence of PCR products cannot be taken to represent lack of integration. Conelation was observed when different PCR pairs were used with the same template DNA, indicating possible truncated transfer of T-DNA during the Agrobacterium transformation process.

In the hph-Sprl constructs used during these experiments, the Sprl cassettes were at the right border of the T-DNA, with the hph region next to the left border. In examination of T-DNA transfer in the Agrobacterium mediated transformation of the ectomycoπhizal fungus Hebeloma cylindrosporum, 85% of left border sequences were recovered, and only 15% of right border sequences (Combier et al, 2003). Thus it is possible that many of the transformation cassettes could have been truncated at the right border in the Sprl region, which is consistent with the PCR results observed.

Thus, in order to confirm transformation, it is preferable that transformation cassettes have the antibiotic selection component next to the right border.

All transformant colonies were tested for proteinase activity in a plate based assay, with some having complete reduction of visible proteinase activity, but these downregulated colonies do not conespond with Sprl PCR positive colonies, in most cases. This suggests that some Sprl DNA has been transfened to the A bisporus genome, and is mediating RNA based silencing, and that size of the DNA integrated is not significant, especially given that the small interfering RNA (siRNA) that is postulated as a mediator of RNA degradation, are often only ca. 25 nt in size (Cogoni and Macino, 1999; Chicas and Macino, 2001; Cogoni, 2001). Truncated transfer of T-DNA would deprive Sprl of the control of the A bisporus gpdll promoter, but a genomic promoter or the other gpdll promoter in the construct could be causing transcription.

Downregulation of serine proteinase 1 (Sprl) of Agaricus bisporus

Agaricus bisporus was transformed with plasmids pGRsensehph, pGRstophph, and pGRantihph, and a number of stable hygromycin resistant colonies isolated for each (8 for pGRsense-hph, 17 for pGRstop-hph and 14 for pGRantisense-hph). These colonies were qualitatively screened for proteinase activity on media containing either milk or gelatin by measurement of a protein clearing zone around the mycelia. Several colonies exhibited complete absence of this clearing zone on both media, with others diminished in their ability to digest either milk or gelatin. No obvious increase in proteinase activity was observed, although in a few colonies the clearing zone appeared to have increased, but the assay used was qualitative. In addition, it is now known what other proteinases may have been active in the assay used, and these could have masked any overexpression. EXAMPLE 2 Development of transformed strains

Five candidate genes were transformed into mushroom strains: (i) Green Fluorescent Protein (sGFP), a reporter gene giving visual evidence of transformation (Chiu et al 1996); (ii) Chymosin, an industrial enzyme used in the dairy industry for cheese-making (Dunn-Coleman et al 1991); (iii) The Bacillus thuringiensis CrylAc insecticidal delta endo-toxin (Gleave et al 1992); (iv) Endostatin, an anti-angiogenic protein shown to inhibit the growth of blood supply to solid cancerous tumors (Read et al 2001); and (v) Cyanovirin -N (in glycosylated and non-glycosylated forms), a microbicidal protein active in preventing transmission of HIV/ AIDS (Mori et al 1998).

Novel Agrobacterium binary constructs were prepared using pGREEN (Hellens et al 2000) containing the hygromycin resistance pAN7-l unit (Punt et al 1987; Challen et al 2000) and with the heterologous protein as a divergently transcribed pair.

Fungal and bacterial strains

The following strains were used in these studies: Sporophores of A bisporus A15, a commercial hybrid variety, were harvested from the HRI mushroom unit. Routine cloning procedures were performed in E. coli strain DH5α. Agrobacterium tumefaciens strains were obtained from HRI collections (LBAl 126, EHA105, GV2260, GV3101) or via Professor M. Guiltinan, Pennsylvania State University (AGL1).

Molecular methods

Agrobacterium mediated transformation of A bisporus was carried out using protocols described elsewhere (Challen et al 2000; Chen et al 2000; Leach et al 2004, these proceedings). Transformants recovered on hygromycin supplemented agar (25μg/ml) and were analyzed using standardized molecular procedures. RESULTS AND DISCUSSION

To date transformants have been recovered for four of the genes (sGFP, Chymosin, CrylAc, Endostatin) and shown to contain the transgenes. An example gel showing the presence of the endostatin transgene is shown in Figure 8.

Mushrooms may not have been considered previously for biomanufacturing because their molecular biology and genetics is less intensively studied than plants and has been developed more slowly. The advantages of mushrooms can be seen by comparison with competing systems in Table 1. All of the alternatives to mushrooms have serious disadvantages that may be critical to profitable operations. These range from high costs of infrastructure and production to lack of containment, safety issues, long lead times and poor flexibility in production scheduling.

Mushroom production and picking can be automated (Reed et al 1995) and there is the potential to develop industrial-style production methods to maximise biomass production and minimise costs. Lead times are short and the production of fruited mushrooms from a newly created transgenic strain can be expected in as little as 10-12 weeks.

REFERENCES

Baldauf, S.L., Palmer, J.D. (1993) Animals and Fungi Are Each Other's Closest Relatives - Congruent Evidence from Multiple Proteins. Proc. Natl. Acad. Sci. USA 90: 11558-11562. Burton, K.S., Sreenivasaprasad, S., Challen, M.P., Elliott, TJ. (2003) Selective expression in filamentous fungi. UK Patent application no. 0225390.4. Challen, M., Gregory, K., Sreenivasaprasad, S., Rogers, C, Cutler, S., Diaper, D. et al. (2000) Transformation technologies for mushrooms. Mushr. Sci. 15: 165- 172. Chen, X., Stone, M., Schlagnhaufer, C, Romaine, C.P. (2000) A fruiting body tissue method for efficient Agrobacterium-mediated transformation of Agaricus bisporus. Appl Environ. Microbiol 66: 4510-4513. Chiu, W.-L., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H., Sheen, J. (1996) Engineered GFP as a vital reporter in plants. Curr. Biology 6: 325-330. Cramer, C.L. (1999) Post-harvest bioproduction of active human enzyme therapeutics in transgenic tobacco. Abstract, International Molecular Farming Conference, London, Ontario. Cullen, D., Gray, G.L., Wilson, L.J., Hayenga, K.J., Lamsa, M.H., Rey, M.W. et al. (1987) Controlled expression and secretion of bovine chymosin in Aspergillus nidulans. Bio/Technology 5: 369-376. De Groot, M.J.A., Bundock, P., Hooykaas, P.J.J., Beijersbergen, A.G.M. (1998) Agrobacterium tumefaciens-mediated transformation of filamentous fungi. Nature Biotech. 16: 839-842. Dunn-Coleman, N.S., Bloebaum, P., Berka, R.M., Bodie, E., Robinson, N., Armstrong, G. et al. (1991) Commercial levels of chymosin production by Aspergillus. Bio/Technology 9: 976-981. Elliott, TJ., Burton, K.S., and Challen, M.P. (2001) Farming fungi - the potential of cultivated mushrooms. In Molecular Farming. OECD Workshop, La Grande Motte (France) September 3-6, 2000. Balazs, J.P.T.E. (ed): INRA, Paris, pp. 249-260. Fischer, R., and Emans, N. (2000) Molecular farming of pharmaceutical proteins. Transgenic Res. 9: 279-299. Fox, J.L. (2003) Puzzling industry response to Prodigene fiasco. Nature Biotech. 21: 3-4. Gallie, D.R. (1998) Controlling gene expression in transgenics. Current Opinion PI Biol A: 166-172. Gleave, A.P., Hedges, R.J., Broadwell, A.H., Wigley, P.J. (1992) Cloning and nucleotide sequence of an insecticidal crystal protein gene from Bacillus thuringiensis DSIR732 active against three species of leafroller (Lepidoptera: Tortricidae). New Zealand J. Crop Horticul Sci. 20: 27-36. Gouka, R.J., Punt, P J., Van den Hondel, C. (1997) Efficient production of secreted proteins by Aspergillus: Progress, limitations and prospects. Appl. Microbiol. Biotech. 47: 1-11. Hellens, R.P., Edwards, E.A., Leyland, Ν.R., Bean, S., Mullineaux, P.M. (2000) pGreen: a versatile and flexible binary Ti vector for Agrobacterium-mediated plant transformation. Plant Mol. Bio. 42: 819-832. Hileman, B. (2003) Prodigene and Starlink incidents provide ammunition for critics. Chem. Engineering News 81: 23. Houdebine, L.M. (2000) Transgenic animal bioreactors. Transgenic Res. 9: 305-320. Kusnadi, A.R., Nikolov, Z.L., Howard, J.A. (1997) Production of recombinant proteins in transgenic plants: Practical considerations. Biotech. Bioengineering 56: 473-484. Leach, K., Odon, V., Zhang, C, Kim, H.K., Henderson, J., Warner, P. et al. (2004) Progress in Agaricus bisporus transformation: Agrobacterium methodologies and development of novel marker genes. Mushr. Sci. 16: these proceedings. Lubon, H., and C, P. (2000) Transgenic animal bioreactors - where we are. Transgenic Res. 9: 301-304. Moore, AJ., Challen, M.P., Elliott, T.J., Warner, P.J. (1995) Ballistics for the delivery of transforming DNA to mushrooms. Mushr. Sci. 14: 63-70. Mori, T., Gustafson, K.R., Pannell, L.K., Shoemaker, R.H., Wu, L., McMahon, J.B., and Boyd, M.R. (1998) Recombinant production of cyanovirin-N, a potent human immunodeficiency virus inactivating protein derived from a cultured Cyanobacterium. Protein Expr. Purificat. 12: 151-158. Pew (2002) Pharming the Field: A look at the benefits and risks of bioengineering plants to produce pharmaceuticals. In: Pew Initiative on Food and Biotechnology, 36pp. Punt, P J., Oliver, R.P., Dingemanse, M.A., Pouwels, P.H., Van den Hondel, C.A.M.J.J. (1987) Transformation of Aspergillus based on the hygromycin B resistance marker from Escherichia coli. Gene 56: 117-124. Radzio, R., and Kuck, U. (1997) Synthesis of biotechno logically relevant heterologous proteins in filamentous fungi. Process Biochem. 32: 529-539. Read, T.A., Sorensen, D.R., Mahesparan, R., Enger, P.O., Timpl, R, Olsen, B.R. et al. (2001) Local endostatin treatment of gliomas administered by microencapsulated producer cells. Nature Biotech. 19: 29-34. Reed, J.N., Crook, S., He, W. (1995) Harvesting mushrooms by robot. Mushroom Science 14: 385-391. Romaine, C.P. (2002) Gene transfer technology for mushrooms: the power and potential for significant crop improvement. Mushr. News July 2002: 4-14. Schuren, F.H J. (2002) Heterologous protein production in Mycelial Fungi. In The Mycota X Industrial applications. Esser, K.B., Berlin, J.W. (eds). Heidelberg: Springer- Verlag, pp. 389-403. USDA (2002) Guidance notes for industry. Drugs, Biologies and medical devices from bioengineered plants for use in humans and animals. Draft document published by USDA and other agencies: 26 pp. Ward, M. (2002) Expression of antibodies in Aspergillus niger. Genet. Engineering News 21: 22. Zeng, Q.D., Morales, A.J., Cottarel, G. (2001) Fungi and humans: closer than you think. Trends Genet. 17: 682-684.

SEQUENCES

SEQ ID NO. 1 : sequenced sprl nucleotide sequence, according to the present invention (see SEQ ID NO. 14 for amino acid sequence).

SEQ ID NO. 2: previously published sprl nucleotide sequence of 1323 bp (NCBI Accession No. Yl 3805)

SEQ ID NO 3 : sprl promoter PCR 877bp fragment with engineered SacII-NcoI restriction sites.

SEQ ID NO 4 : serine proteinase promoter and 5' end of the gene: ATG start codon positioned at nucleotide 1248

SEQ ID NO 5: 1499 bp raw sequence data for Sprl gene (up to first intron). The ATG for the gene is at position 1189.

SEQ ID NO 6: 3022 bp raw sequence data for Spr2 complete. ATG is also at the position 1189.

SEQ ID NO 7: previously published sprl protein sequence (NCBI Accession No. Y13805).

SEQ ID NO. 8: previously published sprl nucleotide sequence of 1280 bp (NCBI Accession No. Yl 3805).

SEQ ID NO. 9: previously published revised sequence of sprl of 1325 bp

SEQ ID NO's 10 and 11 : promoter sequences associated with abst 1 and rafe, respectively.

SEQ ID NO's 12 and 13: terminator sequences of abst 1 and rafe, respectively.

SEQ ID NO. 14: sequenced sprl amino acid sequence, according to the present invention.

SEQ ID NOs. 15-32: PCR primers as per Tables 1, 3 and 5.

Claims

CLAIMS:

1. A filamentous fungus in which one or more proteases are suppressed by post- transcriptional gene suppression.

2. A fungus according to claim 1, wherein the protease to be suppressed is a serine protease.

3. A fungus according to claim 1, wherein the protease to be suppressed is sleeted from the group consisting of cysteine proteases, aspartate proteases and human matrix-metalloproteinases .

4. A fungus according to any preceding claim, that produces harvestable fruiting bodies under suitable growth conditions.

5. A fungus according to any preceding claim, which is a Basidiomycete.

6. A fungus according to any preceding claim, which is a member of the Homobasidiomycetes.

7. A fungus according to any preceding claim, which is Agaricus bisporus.

8. A fungus according to any preceding claim, wherein the DNA encoding the post-transcriptional gene suppression is associated with a marker.

9. A fungus according to any preceding claim, wherein the DNA encoding the post-transcriptional gene suppression is functionally associated with a promoter and a marker, and wherein the marker is distal to the promoter, in relation to the post- transcriptional gene suppression DNA.

10. A fungus according to any preceding claim, wherein the post-transcriptional gene suppression DNA is sense or antisense DNA conesponding to all or a part of the genetic sequence to be suppressed.

11. A fungus according to claim 10, wherein the post-transcriptional gene suppression DNA is sense DNA, and does not encode the entire protease to be suppressed.

12. A fungus according to any preceding claim, wherein the DNA encodes RNAi conesponding to a sense or antisense portion of SEQ ID NO. 1.

13. A fungus according to any preceding claim, transformed to express a heterologous protein.

14. A fungus according to claim 13, wherein the protein is expressed in a fruiting body of the fungus.

15. A fungus according to any preceding claim, transformed to express heterologous DNA, which is under the control of the promoter of abstl.

16. A fungus according to any of claims 1 to 14, transformed to express heterologous DNA, which is under the control of the promoter of rafe.

17. A fungus according to claim 15, wherein the promoter comprises the sequence of SEQ ID NO 10, or a mutation or variant thereof, or a sequence which hybridises thereto under conditions of at least 60°C stringency.

18. A fungus according to claim 16, wherein the promoter comprises the sequence of SEQ ID NO 11, or a mutation or variant thereof, or a sequence which hybridises thereto under conditions of at least 60°C stringency.

19. A fungus according to any preceding claim, wherein the heterologous DNA and/or PTGS DNA is operably linked with a terminator comprising the sequence of SEQ ID NO 12, or a mutation or variant thereof, or a sequence which hybridises thereto under conditions of at least 60°C stringency.

20. A fungus according to any of claims 1 to 18, wherein the heterologous DNA and/or PTGS DNA is operably linked with a terminator comprising the sequence of SEQ ID NO 13, or a mutation or variant thereof, or a sequence which hybridises thereto under conditions of at least 60°C stringency.

21. A fungus according to any of claims 1 to 18, wherein the heterologous DNA and/or PTGS DNA is operably linked with a promoter comprising the sequence of SEQ ID NO. 10 and a terminator comprising the sequence of SEQ ID NO 12, or a mutation or variant of either, or a sequence which hybridises thereto under conditions of at least 60°C stringency.

22. A fungus according to any of claims 1 to 18, wherein the heterologous DNA and/or PTGS DNA is operably linked with a promoter comprising the sequence of SEQ ID NO. 11 and a terminator comprising the sequence of SEQ ID NO 13, or a mutation or variant of either, or a sequence which hybridises thereto under conditions of at least 60°C stringency.

23. A filamentous fungus, optionally as defined in any of claims 1 to 18, transformed to express a heterologous protein, or other gene product, wherein the coding DNA for the protein or other product is under the control of the serine protease promoter.

24. A fungus according to claim 23, wherein the promoter is the promoter of SEQ ID NO. 3.

25. A fungus according to any preceding claim, wherein the heterologous DNA is native DNA.

26. A fungus according to any preceding claim, transformed to express heterologous DNA encoding: antibodies, including other diagnostic material; secondary metabolites, such as lectins, pesticidal compounds such as Bacillus thuringiensis toxin (Bt toxin); therapeutic compounds such as vaccines, steroids, heterocyclic organic compounds; biological macromolecules, such as interferon, endostatin and insulin; and medical enzymes, such as thrombolytics and cerebrosidases.

27. A fungus according to any of claims 1,2 and 4-26, transformed to express heterologous DNA, wherein the heterologous DNA is under the control of a Serine protease promoter.

28. A method for selecting potential transformants according to claim 27, the DNA to be expressed being under the control of filamentous fungus serine protease promoter, comprising culturing fungal cells to form a mycelium, exposing the mycelium to sufficient humic stimulus to stimulate serine protease production in untransformed A bisporus, and assaying for the presence of the expression product thereafter.

29. A method for selecting potential transformants of a filamentous fungus, wherein the DNA to be expressed is under the control of filamentous fungus serine protease promoter, comprising culturing fungal cells to form a mycelium, exposing the mycelium to sufficient humic stimulus to stimulate serine protease production in untransformed A bisporus, and assaying for the presence of the expression product thereafter.

30. A method for harvesting an expression product from a fungus according to any of claims 1-27, comprising harvesting fruiting bodies from the fungus on or about day 5, after storage at approximately room temperature and pressure.