Technical Field

5 This invention applies the art of recombinant DNA technology to the production of desired proteins in microbial hosts, particularly the filamentous fungi including Asperσillus species.

10 Background of the Invention

A variety of gene expression systems have been developed for use with filamentous fungal hosts. Among these are systems that exploit carbon catabolite-repressed promoters, e.g.,

15 glucose-repressed promoters, selected primarily because their ability to drive gene expression can be tightly controlled simply by altering fermentation conditions. One system of this type utilizes the promoter of the Asperqillus nidulans alcA gene, which is repressed in the presence of glucose and is

20 induced by ethanol, threonine or related metabolites under glucose-depleted conditions (see Gwynne et ai. , Biochem. Soc. Trans., 1/7:338). This dual control offers an attractive, phased approach to fermentation, in which the host Asperqillus strain is first cultured in the presence of glucose to maximize

25 biomass, and is then cultured under glucose-depleted conditions in the presence of inducer, to elicit production of the desired protein. Similarly regulated systems have also been developed, and make use of other glucose-repressed promoters such as the promoter of the aldA gene of Asperqillus nidulans (see Pickett

30 et ai. , Gene (1987) !5i:217), the amdS gene of Asperqillus nidulans (see Scazzocchio et al. , 1992, infra), the gla gene of

Asperqillus niqer (see Fowler et ai. , Curr. Genet. (1990) ,

18.:537) and the acvA and ipnA genes of Asperqillus nidulans and

* of Cephalosporium chrysogenum (see Brakhage et aX . , (1992)

4 35 124:3789) .

Although the control over gene expression offered by glucose-repressed promoters is desirable in many instances, their use can complicate the production phase of fermentation. For gene product to form, the host strain must be cultured not

only under glucose-depleted conditions, but also in a medium that employs a carbon source alternative to the more preferred substrate, glucose. It would accordingly be desirable to develop promoters that are relieved of their glucose-repressed function.

Studies of the glucose-mediated mechanism of carbon catabolite repression in Asperqillus and other filamentous fungi have implicated the product of the creA gene as the priniciple represεor. It is now widely accepted (see Scazzocchio, in Asperqillus; Biology and Industrial Applications, Butterworth-Heinemann, 1992 at pp.43-63) that when glucose is present, the creA gene product represses expression of genes involved in utilization of other carbon substrates such as ethanol and lactose, and thereby inhibits utilization of those other pathways when glucose is available. Cloning and analysis of Asperqillus nidulans creA has been reported recently by Dowzer and Kelly in Mol. Cell. Biol. (1991) 11:5701.

Summary of the Invention

The domain at which the creA gene product binds within the promoter regions of glucose-repressed filamentous fungal genes has now been identified. This discovery is exploited, in accordance with the present invention, to disrupt the creA binding domain within glucose-repressed promoters, and thereby provide promoter variants that can drive expression in the presence of glucose.

More particularly, and according to one aspect of the invention, there is provided a recombinant DNA expression cassette useful to achieve expression of protein-encoding DNA in a filamentous fungal host, the cassette comprising the protein-encoding DNA and, linked operably therewith, a promoter variant having a creA binding site that is disrupted to permit expression of the protein-encoding DNA in the presence of glucose.

In embodiments of the present invention, the promoter

variant is derived from the promoter of an Asperqillus gene involved in ethanol metabolism, selected for example from the alcA, aldA and alcR genes of Asperqillus nidulans.

In another of its aspects, the present invention provides a filamentous fungal strain having incorporated therein a recombinant DNA expression cassette comprising heterologous DNA and, linked operably therewith, a promoter variant having a disrupted creA binding site.

According to another aspect of the invention, there is provided a method for producing a desired protein, which comprises the step of culturing in the presence of glucose a filamentous fungal strain having incorporated therein a recombinant DNA expression cassette comprising DNA coding for the desired protein and, linked operably therewith, a promoter variant having a disrupted and spatially preserved creA binding site.

The invention and its preferred embodiments are now described in greater detail with reference to the accompanying drawings.

Brief Reference to the Drawings

Figure 1 illustrates the nucleotide sequence of a region of the Asperqillus nidulans alcA promoter. See also SEQ ID NO: 3 of the Sequence Listing.

Figure 2 illustrates schematically the construction of an alcA promoter variant.

Detailed Description and Preferred Embodiments

The present invention provides a strategy for altering those promoters of filamentous fungal genes that are normally glucose repressed, in order to provide promoter variants that function in the presence of glucose to drive expression of protein-encoding DNA linked operably with the promoter variant. This is achieved by eliminating the creA binding site within the glucose-repressed promoter, either by deletion or more

desirably by nucleotide substitution. As used herein, the term "promoter variant" thus refers to a glucose-repressed promoter in which a creA binding site normally resident therein has been disrupted, allowing the promoter variant to function in the presence of a glucose concentration sufficient to sustain growth of a given filamentous fungal strain.

For purposes of the present invention, promoter variants can in principle be developed from the promoter of any glucose- repressed filamentous fungal gene that incorporates within its promoter sequence the creA binding domain. The creA binding domain is characterized as a GC rich motif, and is more particularly identified as a nucleotide region in which at least six G or C residues are incorporated within a contiguous ten nucleotide stretch. The specific sequence of a creA binding domain can vary widely within this parameter, and may for example consist entirely of contiguous G/C residues, e.g., GCGGGGC, or of G/C residues interrupted by one or more A and T residues.

The creA binding domain within a promoter is located typically upstream of each of the required regulatory elements, i.e. the translational and transcriptional start sites. For glucose-repressed filamentous fungal genes having published sequences, identifying the creA binding domain within the promoter will of course be a simple task of locating the GC rich sequence within that sequence. For genes that have been cloned but for which no sequence is yet known, the creA binding domain can be located after sequencing the 5'-untranslated region of the gene, using techniques established in the art. For other glucose-regulated genes, cloning and then sequencing of the 5'-untranslated region thereof will be necessary but can be achieved also according to procedures well established in the art of molecular biology.

From those glucose-repressed promoters incorporating the creA binding domain, promoter variants that are functional in the presence of glucose are obtained by disrupting the creA binding domain. Disruption can be achieved either by wholesale deletion of the creA binding domain and the region 5' thereof,

by selective deletion from within the promoter of the creA binding domain or a functional portion thereof, or, more preferably, by replacing one or more of the G/C residues within the creA binding domain. Replacement residues are desirably

5 adenine and thymine. Relative to wholesale and selective deletion, G/C replacement has the advantage of preserving the spatial relationship, and function, of other regulatory domains within the promoter variant. Of course, this can be particularly important in the circumstance where expression ιo regulating regions exist within the promoter at a site upstream of the creA binding site.

In some cases, promoters may contain more than one creA binding site, and it is desirable, in order to eliminate glucose-repression thereof, to disrupt each of the creA binding

15 sites within that promoter. This can be achieved, as just mentioned, either by deleting those sites, by replacing residues within those sites, or by using a combination of these disrupting techniques.

According to specific embodiments of the invention, the

20 promoter variant is a variant of an Aspergillus nidulans promoter identified below, in which the creA binding site noted below is disrupted by nucleotide replacement:

25

30

35

-.o ¹ = Gwynne et al. , Gene (1987) 51:205 ² = Picket et al. , Gene (1987) 51:217

= Felenbok et al. , Gene (1988) 73:385

= Scazzocchio et al . , 1992, supra

= Sophianopoulou et al. , Mol. Microbiol. (1989) 3_:705

The creation of promoter variants functional in the presence of glucose can be achieved using these and other promoters by applying now conventional techniques of gene manipulation. For instance, a region of the selected promoter containing the creA binding site can be excised using flanking restriction sites and then incorporated into a vector suitable for amplification and subsequent gene manipulation work. The technigue of site-directed mutagenesis can then be applied to disrupt the targetted creA binding site. To delete the creA binding site and sequences 5' thereof, for example, an oligonucleotide can be designed which introduces any desired restriction site just 3' of the creA site, to allow excision. The same technique can also be applied to introduce a nucleotide substitution within the creA binding site, using an oligonucleotide that is mismatched at the desired location but which otherwise is complementary to and hybridizes with the region targetted for disruption. The convenient technique of polymerase chain reaction (PCR) can also be applied in combination with the site-directed approach, to mutagenize and then amplify either the entire promoter or a selected region thereof. The desired promoter variant may of course also be synthesized de novo by applying methods now standard in the gene synthesis art. Briefly, this entails the successive 3' to 5' coupling of suitably protected nucleotide reagents in an automated synthesizer, and then the recovery by gel purification of the deprotected polynucleotide. The block ligation approach may be employed, whereby "blocks" of oligonucleotide pairs, up to about 80 nucleotides in length, are synthesized and ligated in correct succession by overhang complementarity, as described for example by Wosnick et al. in Gene (1989) 26:153. ι _n an alternative approach, the desired DNA may be synthesized in toto, and then amplified by the PCR technique, using the strategy disclosed for instance by Barnett et al. in Nucl. Acids Res. (1990) 18:3094. Once obtained, the promoter variant may be exploited to drive expression of DNA coding for any desired protein in a filamentous fungal host, in accordance with techniques already

otherwise established for the filamentous fungi (for a review see Ballance, Yeast (1986) 2 _. :229). For this purpose, the present invention provides recombinant DNA constructs in which the promoter variant is linked operably with the protein- encoding DNA. The term "linked operably" means that the promoter is linked with the protein-encoding DNA in a manner enabling expression thereof in the filamentous fungal host. For the purposes of this specification, the protein-encoding DNA is understood to comprise a translational start codon and a translational stop codon; and otherwise encodes either the desired protein end-product per se; a fusion protein in which the desired protein is initially produced in cleavable combination with a carrier protein such as Asperqillus niger glucoa ylase; or a secretable precursor in which the desired protein end-product is coupled with a cleavable signal peptide, such as a signal peptide normally associated with an Aspergillus protein e.g. glucoamylase, or any functional equivalent thereof.

In addition to the promoter and the protein-encoding DNA, the constructs of the present invention may also incorporate a transcriptional terminator at a location 3' of the protein-encoding DNA, although experience has shown that such terminators are not essential components of the constructs. Such transcriptional terminators can be obtained as 0.5-1.0 kb fragments of the 3'non-coding regions of Aspergillus genes, for example of the Aspergillus niger glucoamylase gene.

Once the construct is obtained, it can be introduced either in linear form or in plasmid form, e.g., in a pUC-based or other vector, into a selected filamentous fungal host using a technique such as DNA-mediated transformation, electro- poration, particle gun bombardment, protoplast fusion and the like. To allow for selection of the resulting transformants, the transformation typically also involves a selectable gene marker which is introduced with the expression cassette, either on the same vector or by co-transformation, into a host strain in which the gene marker is selectable. Various marker/host systems are available, including the pyrG, argB and niaD genes

for use with auxotrophic strains of Aspergillus nidulans; pyrG and argB genes for Aspergillus oryzae auxotrophs; pyrG, trpC and niaD genes for Penicillium chrysogenum auxotrophs; and the argB gene for Trichoderma reesei auxotrophs. Dominant selectable markers including amdS, oliC, hyg and phleo are also now available for use with such filamentous fungi as A. niger. A. oryzae. A. ficuum. P. chrysogenum. Cephalosporium acremonium. Cochliobolus heterostrophus. Glomerella cingulata. Fulvia fulva and Leotosphaeria maculans (for a review see Ward in Modern Microbial Genetics. 1991, Wiley-Liss, Inc., at pages 455-495) .

Filamentous fungal strains resulting from the transformation are cultured, according to another aspect of the invention, in the presence of a growth-sustaining concentration of glucose, for the purpose of obtaining the desired protein product. Culturing of the strain can be achieved without the need to establish glucose-depleted growth conditions, since the promoter variant driving expression of the desired gene product is no longer functionally repressed in the presence of glucose. For those promoters whose function is otherwise not controlled, production of the desired protein can be achieved throughout the fermentation period. Phased production can also still be achieved, using promoter variants that are controlled by mechanisms other than glucose-repression. As is hereinafter exemplified, the alcA promoter can be functionally controlled, even after disruption of its creA binding site, using an inducing agent to limit protein production to a certain window within the fermentation period.

Example 1

Creation of an alcA promoter variant

The promoter of the A. nidulans alcA gene is repressed by a combination of the creA gene product and glucose; it is also induced under glucose-depleted conditions by a combination of the alcR gene product and an inducing agent such as ethanol.

To construct an alcA promoter variant that functions in glucose at concentrations sufficient to sustain Aspergillus growth, the alcA promoter was obtained in the manner reported by Gwynne et al. in Bio/Technology (1987) _. 5:713, and manipulated as described below.

Two putative creA binding sites (both GCGGGGC) were first identified by sequence scanning. As shown in Figure 1, these sites reside upstream of the ATG initiation codon of the alcA gene at positions -302 and -167, and both are conveniently within a Sphl/SplI region of the promoter. To provide an alcA promoter disrupted at these sites by nucleotide replacement, a 160bp fragment of the promoter was amplified using the tailed primers noted below:

(Forward)Hindi11 SphI -302

5'GACTGACTAAGCTTGCATGCGGAACCGCACGAGGGTACTACGGAAATTGAC 3'

fSEO ID NO: 1)

(Reverse) EcoRI SplI -167

5'GACTGACTGAATTCCGTACGGGGTAGTACTCGTTTGTGGCTCTCCGTGCG 3'

fSEO ID NO: 2)

Twenty-five cycles of PCR amplification were performed according to the method reported by Scharf, 1990, In: "PCR Protocols: A Guide to Methods and Applicatons". M.A.Innis et al. (Eds.), Academic Press. The Hindlll/EcoRI-tailed amplification product was then sub-cloned into a pTZ18R holding vector and the correct, variant sequence was verified by DNA sequence analysis. The holding vector and the full length alcA gene (pUC8 background) were then each restricted with SphI and SplI, and then selectively ligated to form plasmid palcA ^var -2EB, in which the Sphl/SplI region of wild type alcA promoter is replaced with the variant region containing mutations in the two putative creA binding sites. By this approach, spatial arrangement of the nucleotides within the alcA promoter was

also conserved.

Because expression from the alcA promoter variant requires induction by the product of the alcR gene which is itself glucose-repressed by a creA-mediated mechanism, an Aspergillus nidulans strain that produces the alcR product in a constitutive manner was exploited for expression of the alcA gene from the variant promoter. The selected host strain was constructed from the Glasgow double auxotroph FGSC4(argB-;ura-) by transforming with a plasmid carrying both the A. nidulans argB gene, as selectable marker, and an expression cassette in which DNA coding for the alcR product (see Felenbok et al. , Gene (1989) 23 _. -*385) was placed under expression control of the constitutive promoter of the A. nidulans glyceraldehye-3- phosphate dehydrogenase gene (see Punt et al. , Gene (1988) <5 :49. To accomodate multiple copy integration of cassettes carrying the alcA promoter variant, a strain carrying multiple copies of the constitutive alcR gene was chosen, and designated T2625 (ura ^" , multiple alcR ⁰ ) .

The plasmid palcA ^var -2EB was then introduced into the selected host A. nidulans strain, together with a marker plasmid pFB94 which carries the pyr4 gene of Neurospora crassa (see Buxton and Radford, Mol. Gen. Genet. (1983) 190:403) . Transformation was achieved using the PEG/CaCl ₂ -mediated technique according to the protocol reported by Royer et al.. Mol. Gen. Genet. (1991) 225:168. Putative transformants, containing the pFB94 construct, were first selected without uridine on minimal medium. Mixed spores collected from approximately 80 ura ⁺ transformants (3mm colonies) on a single plate were inoculated into 50ml of selective medium (0.4% 2- deoxyglucose (2-DOG) , 1.0% ethanol , 0.67% yeast nutrient broth) . The culture was incubated at 30°C with agitation (200 rp ) for 3 days. Flocculent colonies were collected from the liquid cultures and isolates derived from single spores were prepared from each of the 2-DOG resistant colonies for further analysis.

The host strain and transformants were next cultured for 48 hours at 30°C in liguid medium containing various

supplements selected to examine effects of carbon source on growth. Results are presented in Table 1 below:

Table 1

Spore suspensions were inoculated into 50 ml of liquid medium supplemented with 0.67% yeast nitrogen base (YNB) and ImM uridine; and optionally ethanol (1.0%), glucose (0.4%) or the non-metabolized analogue, 2-deoxyglucose (2-DOG) (0.4%).

As the table indicates, the host strain was able to grow in the presence of either ethanol or glucose, or a combination thereof. It was unable to grow in the presence only of a non-metabolized glucose analogue 2-DOG, nor could it utilize ethanol in the presence of 2-DOG. Clearly, 2-DOG effectively repressed expression from the wild type alcA gene, the product of which is required for ethanol utilization. On the other hand, the palcA ^var -2EB transformants were able to grow in the presence of ethanol when present either in combination with glucose or when present as the sole carbon source (even when glucose or its 2-DOG equivalent were present) , demonstrating conclusively that expression of the alcA gene product required for ethanol utilization was mediated by the alcA promoter variant.

SEQUENCE LISTING

(1) GENERAL DJTOEMATION:

(i) APPLICANT:

(A) NAME: GIST-EROCADES N.V.

(B) STREET: P.O. Box 1

(C) CITY: Delft (E) OOONIRY: Netherlands

(F) POSTAL CODE (ZIP) : NL-2600 MA

(G) TELEPHONE: 31.15.799111 (H) TELEFAX: 31.15.793957 (ii) TTTI-E OF INVENTION: REDUCTION OF HEIERO DGOUS PROTEINS IN

FILAMENTOUS FUNGI

(iii) NUMBER OF SEQUENCES: 3 (iv) OCMEUTER READABLE POEM:

(A) MEDIUM TYPE: Floppy disk

(B) OCMEUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFIWARE: Patentin Release #1.0, Version #1.25 (EPO)

(vi) ERIOR AE-FIilCATION DATA:

(A) APPLICATION NUMBER: US 07/988,778

(B) FILING DATE: 10-DEC-1992

(2) INFORMATION FOR SEQ ID NO: 1:

(i) SEQUENCE <_HARACTERISTICS: (A) LENGTH: 51 base pairs (B) TYPE: nucleic acid

(C) STRANDEENESS: single

(D) TOPOIDGY: linear

(ii) MDLECUIE TYPE: cENA (iii) HYPOTHETICAL: YES

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1:

GAC GACEAA GCTTCCAIGC GGAACOGCAC GAGGGTACIA GGGAAAITGA C 51

(2) INFORMATION FOR SEQ ID NO: 2:

(i) SEQUENCE αiARACIERISTICS:

(A) LENSIH: 50 base pairs

(B) TYPE: nucleic acid

(C) STRANDEENESS: single (D) TOPOLOGY: linear

(ii) MDLECUIE TYPE: cENA (iii) HYPOTHE_TICAL: YES

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2 :

GACTGACTGA ATTCOGTAOG GGGTAGTACT OGriTTGTGGC TCTCOOTGCG 50

(2) 3-NPOKMAπθN FOR SEQ ID NO: 3:

(i) SEQUENCE OffiRACTERISTTCS: (A) LEENGIH: 360 base pairs (B) TYPE: nucleic acid

(C) STRANDEENESS: double

(D) TOPOIDGY: linear

(ii) MDIECULE TYFE: ENA (genαnic)

(iii) HYPOTHETICAL: NO

(vi) ORIGINAL SOURCE:

(A) ORGANISM: Aspergillus nidulans

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 3: OGCAGCTOGG GAIAGTTOαG ACCTAGGAIT GGATGCATGC GGAAOOGCAC GAGGGOGGGG 60

OGGAAATIGA CACACCACTC CTCTOCACGC AGCCGTTCCA AGAGGIACGC G?IAIAGAGCC 120

GTΓAIAGAGCA GAGACGGAGC ACTΓΓCΓGCT ACΓGTCCGCA OGGGATGTCC GCACGGAGAG lβo

CCACAAACGA GCGGGGCCCC GIACGTGC . TOCIACCOCA GGATCGCATC CTCGCATAGC 240

TGAACATCEA TAIAAAGAOC COCAAGGΓTC TCAGTCTCAC O^CATCATC AACCAACAAT 300 CΆAC_AGTTCΓ C ACTCAG Γ AATΓAGAACA CTΓCCAATCC TA CACCIΌG CATCAAAAIG 360