CA2115465A1 - Highly alkaline proteases, dna sequences encoding them and process for their production - Google Patents

Abstract

Abstract of the Disclosure Optimized highly alkaline proteases suitable for use in washing agent formulations in which the original amino acids have been replaced in two, three, four or five of the given positions by different, defined amino acids, and a process for preparing them using microorganisms transformed by mutated DNA sequences obtained by altering DNA sequences encoding highly alkaline protease produced by Bacillus species in specified positions by directed mutagenesis.

Description

2 1 1 ~

HIGHLY ALKAI-INE PROTEASES, DNA SEQUENCES
ENCODING THEM AND PROCESS FOR IHEIR PRODUCTION
FieldoftheInvention The present invention relates to improved highly alkaline proteases in which certain amino acids have been replaced, to D~A sequences which have been altexed by directed mutagenesis and which encode these improved ;`~
proteases, and to vectors which contain these DNA sequences.
The present invention likewise relates to microarganisms lo transformed with these vectors, to a process for preparing optimized highly alkaline proteases, and to washing agents ;~ (i.e. detergents) which contain these improved proteases.
Background of the Tnventior ;~ Highly alkaline proteases are valuable industrial ~` 15 products having advantageou~ uses, particularly in the ;
detergent industry, since they remove impurities which contain protein. In order to be active, these proteins must ~;
not only possess proteolytic activity under washing conditions (pH, temperature) but, in addition to this, also 2~-- must be compatible with other detergent constituents, i.e.
in combination with other enzymes, surfactants, builders, bleaching agents, bleaching agent activators, and other additives and auxiliary substances; that is they must be sufficiently stable towards these constituents and be -~
25 sufficiently active in their presence. - ~
Highly alkaline proteases are special enzymes which are ""',r,~,, ' obtained by cultivating microorganisms, in particular by cultivating Bacillus species, which, like, for example, Bacillus alcalophilus, produce the desired highly alkaline ~
".. . ,- ,.
- 1 - , .' '-,'''' - 2~4~
protease and secrete it into the culture medium, from which the protease can be isolated. These highly alkaline proteases differ from ordinary alkaline proteases which can be obtained by cultivating Bacillus species such as, in particular, B. subtilis, B. amyloliquefaciens and B.
. liche~iformis, for example.
A series of natural and arti~icially (recombinantly) altered alkaline and highly alkaline proteases are known in -the state of the art. Thus, proteases are described, for example in Patent Applications WO 91/00345, EP 328,229 and EP 415,296, whose amino acid sequences have been altered at a plurality of positions by targeted mutagenesis (point mutations). -Nevertheless, there still exists a need for novel highly alkaline proteases which have been further optimi~ed and which exhibit improved properties with regard to their washing performance or stability.
.. - - ~'.'', SummarY of the Invention An aspect of the present invention is to make available novel highly alkaline . ~
pr~teases having properties which were further improved, as well as DNA sequences, :
vectors and transformed microorganisms which were necessary for this purpose, by the process of the present invention for preparing these proteases.
2s The invention relates to highly alkaline proteases which are distinguished by the fact that they have an amino acid sequence which possesses at least 80% homology with the amino acid sequence given in Fig. 1 (SEQ ID NO:l) and differs from this latter amino acid sequence in two, three, four or five of the positions 42, 43, 114, 216 and 249 of ~;
Fig. 1 or in the homologous positions therewith, where, in - the case of a replacement at the positions concerned, '., , ~.

--` 2 1 1 ~

the amino acid located in position 42 is replaced by arginine, the amino acid located in position 43 is replaced by glutamine, the amino acid located in position 114 is replaced by arginine, the amino acid located in position 216 is replaced by -~
serine, g]utamine or alanine, the amino acid located in position 249 is replaced by glutamic acid.
Other and further advantages and features of the invention will be apparent to those skilled in the art from the following detailed description thereof, taken in conjunction with ~ ~ .
the accompanying drawings.
Brief Description of the Drawings The invention will be described in further detail with `
reference to the accompanying drawing figures wherein: ~ -Fig. 1 shows the DNA sequence of the AvaI/HindIII
fragment, containing the structuraI gene of the highly alkaline starting protease from Bacillus alcalophilus HAl, as well as the amino acid sequence of this starting protease;
Fig. 2 shows the restriction map of the plasmid 2S pCLEAN4;
Fig. 3 shows the DNA sequences of the synthetic oligonucleotides used for the directed mutagenesis, with indications of the recognition sites for individual ;~
restriction endonucleases which were eliminated or produced, whereby the nucleotide changes which were made in the ;~
original DNA sequence of the starting protease are -~
designated by indicating the altered nucleotides using small letters;
Fig. 4 shows the restriction map of the vector pCLMUTNl;

., .~ . . . ~

211~6~

Fig. 5 shows the restriction map of the vector pC~MUTC1;
Fig. 6 shows the restriction map of the vector pUBC132;
Fig. 7 shows the restriction map of the vector pALlP;
Fig. 8 shows the restriction map of the expression vectors of the type pALlNC, used for expressing the highly alkaline proteases altered by mutation and for expressing the highly alkaline starting protease;
Fig. 9 shows the restriction map of the vector pALlN;
1~ and Fig. 10 shows the restriction map of the vector pALlC.
De~iled Description These highly alkaline proteases possess molecular weights of 26,000 to 28,000 g/mole, measured by SDS
polyacrylamide gel electrophoresis against reference proteins of known molecular weight. Their pH optimum is for the most part within the range from 8 to 12.5, where pH
optimum is understood to mean that pH range in which the proteases exhibit maximum proteolytic activity. The 2~ proteases according to the invention exhibit good pH
stability.
In accordance with one preferred embodiment of the invention, the proteases have amino acid sequences at least 90~ homologous. in particular, however, at least 95~
homologous, to the amino acid sequence of Fig. 1 (SEQ ID
NO:1). As used herein, the phrase "homologous to the amino acid sequence of Fig. 1" refers to the degree of sequence correspondence or sequence similarity between the respective amino acid sequence and the amino acid sequence given in 3~ Fig. 1 (SEQ ID NO:l). In order to determine homology, the segments of the amino acid sequence of Fig. 1 and of the amino acid sequence to be compared with it which correspond to each other structurally are superimposed with each other such that the maximum structural correspondence is obtained ' - 3a -.
.

;;

between the amino acid sequences, with differences caused by deletion or insertion of individual amino acids being taken into account and compensated for by correspondiny displace~
ments of sequence segments. The number of amino acids which ~, . '' ' ' ~.-'~

' :~

.: " :. .
-~

' " . ':, : .' ;'',.''~,"". '' ~ '" ' ~., ~;'''' ''', . ..- . ...:

'; ' '.''.~ ~

- 3b - ;~
, .. ", .':

~ 2~1~46~
. . .

now correspond to each other in the sequences ("homologous positions"), based on the total number of amino acids contained in the sequence in Fig. 1, then gives the homology in ~. Divergences in the sequences can be due to variation, insertion or deletion of amino acids. Accordingly, in the highly alkaline proteases according to the invention, which proteases were obtained from proteases which were at least 80% homologous with Fig. 1, the amino acid positions designated with reference to Fig. 1 relate to the positions homologous therewith. Deletions or insertions in the proteases which are homologous to the sequence of Fig. 1 can lead to a relative displacement of the amino acid positions, so that the numerical position designations of the amino acid positions corresponding to each other in homologous stretches of amino acid sequences need not be identical.
In the highly alkaline proteases according to the invention, the amino acids originally located in the replacement positions must be replaced by different defined amino acids. Thus, the amino acid located in position 42 is replaced by arginine, the amino acid located in position 43 is replaced by glutamine, the amino acid located in position 114 is replaced by arginine, the amino acid located in position 216 is replaced by either serine, glutamine or alanine, and the amino acid located in position 249 is replaced by glutamic acid. The amino acids originally present in the replacement positions can, in this context, be replaced in two, three, four or five, preferably in two or three, of the aforementioned positions. Thus, the proteases according to the invention exhibit two, three, ;~ 30 four or five of the amino acid replacements N42R, I43Q, N114R, T249E, M216S or M216Q or M216A, based on the positions given in Fig. 1 or on a position homologous therewith. In this patent application, conventional nomenclature is used for designating the mutations in which the amino acids are designated by the single-letter code, with the original amino acid being placed before the --` 211~
. ' ~.

position number and the introduced amino acid being placed after the position number.
Preferably, the highly alkaline protease according to the invention has an amino acid sequence in which the amino acid located in position 216 is replaced by serine, glutamine or alanine and either the amino acid located in position 43 is replaced b~ glutamine or the amino acid located in position 114 is replaced by arginine or the amino acid located in position 249 is replaced by glutamic acid.
Based on the positions given in Fig. 1, or the positions homologous therewith, the proteases according to the invention thus exhibit the amino acid replacements M216S or M216Q or M216A and either I43Q or N114R or T249E.
In accordance with another preferred embodiment the highly alkaline protease according to the invention has an amino acid sequence in which the amino acid located in position 216 is replaced by serine, glutamine or alanine and the amino acid located in position 114 is replaced by arginine and either the amino acid located in position 42 is replaced by arginine or the amino acid located in position 43 is replaced by glutamine. Based on the positions given in Fig. 1, or positions homologous therewith, the protease according to the invention thus exhibits the two amino acid replacements M216S or M216Q or M216A and either N42R or I43Q.
In accordance with a further preferred embodiement, the highly alkaline protease according to the invention has an amino acid sequence in which the amino acid located in position 43 is replaced by glutamine; the amino acid located in position 249 is replaced by glutamic acid, and either the amino acid located in position 114 is replaced by arginine or the amino acid located in position 216 is replaced by ` serine, glutamine or alanine. Based on the positions given in Fig. 1, or positions homologous therewith, the protease according to the invention thus possesses the three amino .. .,:

211~6~
.. . ~

acid replacements I43Q and T249E and either N114R or M216S
or M216Q or M216A.
In accordance with an additionally preferred embodiment the highly alkaline protease according to the invention has an amino acid sequence in which the amino acid located in position 43 is replaced by glutamine; the amino acid located in position 114 is replaced by arginine; the amino acid located in position 249 is replaced by glutamic acid, and the amino acid located in position 216 is replaced by serine, glutamine or alanine. Based on the positions given in Fig. 1, or positions homologous therewith, the protease according to the invention thus exhibits the four amino acid replacements I43Q, N114R, T249E and M216S or M216Q or M216A.
Of the aforementioned preferred replacement combinations in the amino acid sequence of the highly alkaline proteases according to the invention, those are very particularly preferred in which the amino acid located in position 216 is replaced by glutamine. Based on the ~`~ position given in Fig. 1, or the position homologous therewith, this denotes the one amino acid replacement M216Q
~; in the sequence of the protease according to the invention.
In an additional, particularly preferred, embodiment, the highly alkaline proteases according to the invention have an amino acid sequence in which the amino acid located in position 114 is replaced by arginine, and either the ~- amino acid located in position 43 is replaced by glutamine or the amino àcid ~located in position 249 is replaced by glutamic acid. Based on the positions~given in Fig. 1, or positions homologous therewith, the protease according to the invention thus exhibits the two amino acid replacements N114R and either I43Q or T249E.
As regards their washing performance and their storage stability, the highly alkaline proteases according to the invention are clearly superior to the proteases in which the amino acids at the aforementioned positions have not been replaced by the amino acids named for each position. It is 2115~
.
surprising that a marked improvement in the properties can be achieved, in comparison to the starting protease, using the particular combination of the designated amino acids in the aforementioned positions of the amino acid sequence, with positive effects astonishingly being achieved in the combination of the stated amino acid replacements.
In order to obtain the proteases according to the invention, microorganisms are cultivated which have been transformed with an expression-vector which contains a 0 structural gene encoding the respective protease.
The invention therefore also relates to DNA sequences which encode highly alkaline proteases having the above~
described amino acid sequences.
The DNA sequences according to the invention are distinguished by the fact khat they encode a highly alkaline protease which exhibits an amino acid sequence which possesses at least 80% homology with the amino acid sequence given in Fig. 1 and differs from this sequence in two~
three, four or five of the positions 42, 43, 114, 216 and 249 of Fig. 1, or the positions homologous therewith, by, in the event of a replacement at the positions concerned, the amino acid located in position 42 being replaced by arginine, the amino acid located in position 43 being replaced by glutamine, the amino acid located in position 114 being replaced by arginine, the amino acid located in position 216 being replaced by serine, glutamine or alanine, the amino acid located in position 249 being replaced . ...
by glutamic acid.
Those DNA sequences according to the invention are preferred which encode highly alkaline proteases whose amino acid sequences exhibit a homology of more than 90%, in ~
35 particular, however, of more than 95%, with the amino acid --sequence of Fig. 1 (SEQ ID NO:l) and in which two, three, ~ i : ': . .
- 7 ~

~1~5~

four or five of the given positions are replaced. The DNA
sequences according to the invention can, in particular, encode proteases which exhibit the aforementioned amino acid replacements in their sequence, based on the positions in Fig. 1 or on positions homologous therewith.
In addition, the invention relates to washing and/or cleaning compositions which contain at least one highly alkaline protease according to the invention. Such washing and cleaning agents can, for example, be employed for cleaning surfaces in the hygiene and food product fields.
Preferably, the proteases according to the invention are used in detergent formulations for cleaning textiles or tableware. For this purpose, the invention makes available a group of novel highly alkaline proteases having individual properties which are superior to those of previously known proteases, in particular superior washing performance and/or superior enzyme stability, rom which group, depending on the properties (washing performance, temperature resistance, compatibility with other constituents) of the highly alkaline protease to be employed which are especially required, a protease according to the invention can be selected which is particularly suitable ~or the relevant special detergent formulation. The proteases according to the invention can be employed in detergent or cleaning formulations, e.g. in powdered detergents, compact detergents or liquid de~ergents, either individually or, if desired, also in combination with each other, optionally in combination with conventional detergent proteases or with other customary detergent enzymes, such as, for example, amylases, lipases, pectinases, nucleases, oxidoreductases, etc. The proteases according to the invention are used in the detergent formulations in customary amounts for detergent enzymes, in particular in quantities of up to 5 by weight (based on the dry weight of the total composition), preferably in a quantity of 0.2 to 1.5~ by weight.

i ~-`` 2 1 ~ 5 ~

In addition to the detergent enzymes mentioned above, the detergents of the invention mày contain any of the washing agent ingredients which are customary in the state of the art, such as surfactants, bleaching agents or builder substances, as well as customary amounts of other conventional auxiliary substances used in formulating detergents. The auxiliary substances include, for example, boosters, enzyme stabilizers, anti-redeposition agents and/or compatibility agents, complexing and chelating agents, lather regulators and additives, such as optical brightening agents, opacifying agents, corrosion inhibitors, anti-electrostatic agents, dyes, bactericides, bleaching agent activators and peracid bleaching agent pre-stages.
Thus, detergent formulations according to the invention contain, in a typical exemplary composition, based on dry matter, a) at least 1~ by weight of a surfactant or surfactant mixture, b) up to 40% by weight of a builder or builder mixture, c) up to 40~ by weight of a bleaching agent or bleaching agent mixture, preferably a perborate, such as sodium perborate tetrahydrate, sodium perborate monohydrate or sodium percarbonate, d) up to 3~ by weight of at least one protease according to the invention, and e) further constituents, such as auxiliary substances, etc., up to 100~ by weight.
The proteases according to the invention are particularly well suited for use in powdered dishwashing detergents, such as those customarily employed, for example, in automatic dishwashers.
Powdered detergent formulations can be formulated in a conventional manner. For this purpose, for example in powdered or compact detergents, the proteases according to the invention can be mixed in a known manner with the other components of the detergent formulation in the form of g _ -. .-_ 21~6~

granules, prills or pellets, and optionally also provided with surface coatings.
Due to their improved stability, the proteases according to the invention may also be employed very effectively in liquid detergents.
The invention further relates to a process for preparing an optimized highly alkaline protease according to the invsntion using a transformed microorganism which contains a vector having a DNA sequence which encodes an amino acid sequence which possesses at least 80% homology with the amino acid se~uence of the starting protease given in Fig. 1 (SEQ ID NO:1) and differs from this sequence in two, three, four or five, preferably two or three, of the positions 42, 43, 114, 216 and 249 of Fig. 1, with the amino acid located in position 42 being replaced by arginine, the amino acid located in position 43 being replaced by glutamine, the amino acid located in position 114 belng replaced by arginine, the amino acid located in position 216 being replaced by serine, glutamine or alanine, and the amino acid located in position 249 being replaced by glutamic acid. The transformed microorganism is cultured, and the improved highly alkaline protease is isolated from the culture medium.
In order to produce the microorganisms employed in the preparation procedure, the approach can be that a) the DNA sequence encoding the protease (i.e. the structural gene of the protease) is initially isolated from a suitable bacterium which produces a highly ` alkaline protease having an amino acid sequence I possessing at least 80%, preferably more than 90%, in particular, however, more than 95%, homology with the amino acid sequence of Fig. 1 (SEQ ID NO~
b) the nucleotide sequence of this DNA sequence is determined, c) mutations (point mutations) are produced in the DNA
sequence which is now known such that the mutated DNA
.

. .

2 1 1 ~ 11 6 .

sequence now encodes a highly alkaline protease in which amino acids of the original protease are replaced, in aforementioned positi.ons, by an amino acid designated for the res ective position, d) an expression vector is subsequently produced lncorporating the mutated DNA sequence, and e) the resulting expression vector is transformed into a suitable microorganism, which can be finally be employed for preparing the mutated, highly alkaline protease.
The structural genes which encode amino acid sequences of highly alkaline proteases having at least 80~ homology with the amino acid sequence given in Fig. 1 tSEQ ID NO~
can be obtained using known methods. To do this, the chromosomal DNA is isolated, for example, from a bacterium ~ donor bacterium"), in particular from a Bacillus species, ;~ which produces the highly alkaline protease, and this DNA is then partially hydrolysed using suitable restriction endonucleases. -The resulting restriction fragments of the donor D~A
can be fractionated according to size, for example by gel ; electrophoresis, and the fragments of the desired size then recombined with a suitable vector DNA. Advantageously, a plasmid is used as the vector, by means of which plasmid expression can be achieved of the foreign DNA introduced into the host organism employed.
Bacteria, preferably a Bacillus species, can be transformed with the in vitro recombined DNA (vector +
restriction fragments of the donor DNA), and the ; 30 transformants can then be selected in accordance with a known marker property of the vector DNA (e.g. neomycin resistance). Those transformants, among all the resulting transformants, which secrete the highly alkaline protease to ; an increased extent can be sought on protein-containing plates and then isolated. Finally, the plasmid DNA
introduced into such a transformant clone can then be . .
..~. . . ,.:

2115~

isolated from it. Subsequently, the plasmid, which includes the structural gene of the starting protease and additional DNA sequences from the donor DNA sequence, is cut with a number of different restriction endonucleases (restricted);
the resulting DNA fragments are separated according to size by gel electrophoresis, and a restriction map is constructed - on the basis of the banding pattern which has been found.
Knowledge of the restriction map of the plasmid makes it possible to excise from the latter, by cutting with selected restriction endonucleases, a DNA fragment from the donor DNA
sequence, which fragment essentially now only encompasses the structural gene for the highly alkaline protease, the associated pre- and pro-units, and the promoter unit required for expression of the gene.
15By reincorporating this donor DNA sequence, which has now been reduced in size, into a suitable vector, a new, replicable vector can be obtained whose ability to express the highly alkaline starting protease can be tested by transforming a bacterium, in particular a Bacillus species, with this vector, cultivating the resultant transformant and testing it for protease activity. The restriction map in Fig. 2 represents an example of such a reduced vector, which vector has the designation pCLEAN4.
In order to sequence the structural gene of the protease, the aforedescribed vector is first replicated in a suitable microorganism, and the protease gene is subsequently isolated. This gene is then sequenced in accordance with generally known methods for DNA sequencing ~`(e.g. by the method of Maxam and Gilbert, 1980, Methods in Enzymologyl Grossmann L., Moldave K., eds., Academic Press Inc., New York and London, Vol. 65, 499, or by the dideoxy chain termination method of Sanger and Brownlee 1977, Proc.
Natl. Acad. Sci. USA 74: 5473).
The elucldated nucleotide sequence can then be translated, using the genetic code, into the amino acid sequence of the protease.
":

.:

2 1 ~ 5 ~
' In accordance with the invention, the DNA sequence encoding the protease is mutated by replacement of the corresponding codons such that the mutated DNA sequence encodes an optimized highly alkaline protease which possesses an amino acid sequence in which, in two, three, four or five, preferably two or three, of the positions 42, 43, 114, 216 and 249 of the amino acid sequence in Fig. 1 (SEQ ID NO:l), or in one of the positions homologous therewith, the amino acid in question is replaced by one of the aforementioned amino acids.
The point mutations are introduced into the DNA
encoding the highly alkaline protease by means of known methods for directed mutagenesis. For this purpose, circular single-stranded DNA is produced from suitable vectors (phagemids), e.g. from pCLMUTN1 of Fig. 4 or pCLMUTC1 of Fig. 5, optionally with the assistance of a helper phage, which single-stranded DNA contains the whole of the DNA sequence encoding the protease or, preferably, only the DNA sequence in which the mutation is to be undertaken. A hybridizable synthetic oligonucleotide, which contains, at the desired point mutation site, a nucleotide structural unit which is selected such that the associated codon encodes one of the amino acids already indicated for the relevant position, e.g. arginine in position 114, is hybridized with this circular single stranded D~A. In ~ addition, the oligonucleotide is also altered, as compared - with the original nucleotide sequence which is to be hybridized, by one or a few further nucleotide structural units, such that, while the coding of the original amino acid sequence is preserved within the limits of the degeneracy of the genetic code, either a restriction site ; which may be present in the original nucleotide sequence is removed in the synthetic oligonucleotide, or else a further restriction site is introduced into the synthetic oligonucleotide. The removed or introduced restriction site is later used for distinguishing the mutated DNA sequencç

21~6~

from the starting-type DNA sequence usin~g appropriate restriction endonucleases. In a variant, uracylated single-stranded DNA is produced as the template in the process of directed mutagenesis and used for hybridization to the synthetic oligonucleotides. Once the reactions of the process of directed mutagenesis are complete, the uracil-containing DNA single strand, which served as the template for producing mutated DNA strands (vectors), can be removed by treatment with uracil N-glycosylase without any phenotypic selection of mùtants being required. The glycosylase treatment can be carried out either using the isolated en2yme or else with the aid of a suitable microorganism possessing uracil N-glycosylase activity and transformed with mutated vector DNA.
The completion of the partially double-stranded DNA
sequence obtained by hybridization to give the complete double strand is then carried out by adding the necessary nucleotides and using DNA polymerase and DNA ligase. The circular, double-stranded DNA sequence which is produced is subsequently transformed as a vector into a suitable microorganism. Subsequently, the mutated DNA sequences are identified by means of the unique restriction endonuclease recognition sites and isolated from the transformant. If uracylated single-stranded DNA is employed, the replication 25 is then undertaken in, for example, an E. coli strain which -~
preferentially replicates the mutated, non-uracylated DNA
strand of the double-stranded vector produced in the mutation process. This additionally facilitates selection of the mutated DNA vectors.
~30 The synthetic oligonucleotides which are required for ~` -the directed mutagenesis are prepared by known methods (e.g.
according to Beaucage S.L. and Caruthers M.H., 1981, Tetrahedron Letters 22: 1859-1862).
The circular, double-stranded DNA sequences which were obtained by the directed mutagenesis and which possess the introduced mutations represent mutated vectors from which, - 14 - ~

,: ,.'' ~ : ' :::

depending on the circumstances, the whole mutated protease structural gene or the mutated stretch of the protease structural gene can be excised by treating with appropriate restriction endonucleases and introduced ~subcloned~ into an appropriate expression vector. This expression vector is then used to transform appropriate microorganisms, e.g.
Bacil7us species, which are subsequently cultivated under suitable conditions for expressing and isolating the mutated highly alkaline proteases.
In a preferred embodiment of the invention, the whole structural gene is not employed for the directed mutagenesis, but rather only a stretch thereof in which the mutation is to be produced. For this purpose, the N-terminal or C-terminal half of the structural gene, for example, is excised, using appropriate restriction endonucleases, from the vector which is being used to replicate the structural genes and then subcloned into a suitable phagemid. In this way, vectors are obtained which either contain the N-terminal or the C-terminal half of the structural gene and which are first replicated adequately in a suitable microorganismj for example E. coli, and then ;~ ~ subjected to the aforedescribed directed mutagenesis. An advantage of the mutagenesis of stretches of the DNA of the structural gene is that shorter single-stranded DNA
sequences can be used and thus, following the step of hybridization to synthetic oligonucleotides, appreciably fewer nucleotides have to be added to the partial D~A double strand than is the case when using the entire DNA sequence.
This reduces the synthetic requirements and also the danger ,of undesired accidental mutations.
The mutated DNA sequences can be excised from the vector used for producing the mutations with the aid of appropriate restriction endonucleases and then incorporated into vectors possessing corresponding restriction sites, which vectors represent precursors of the actual expression vectors which are required for expressing the highly .

211~

alkaline protease. These vectors are constructed such that, in addition to the appropriate restriction sites (e.g.
consisting of a synthetic linker), they also already contain the regulatory sequences, signal sequences, promoter sequences, and the DNA sequences encoding the pre- and pro-units of the protease, which are required for expressing the protease in a host organism.
Subcloning a mutated DNA sequence into such a vector produces the actual expression vector for an optimized highly alkaline protease. The mutated DNA sequence is incorporated into this expression vector precursor in such a way that an expression vector which has an appropriate reading frame is obtained. Mutated stretches of the DNA
sequence encoding the protease, e.g. a C-terminal or a N-terminal stretch, can be incorporated into vectors which already contain the relevant remaining non-mutated, or, optionally, also mutated (production of multiple mutations) stretch; alternatively, the whole mutated DNA sequence encoding the protease can be incorporated into vectors which, prior to this, do not contain any stretches of this protease DNA sequence. Examples of such precursor vectors of an expression vector which already contain stretches of the protease gene are the vectors having the designations ; pALlN and pALlC, whose restriction maps are reproduced in 2S Fig. 9 and Fig. 10, respectively.
The expression vector precursors for the preferred variant of the invention (mutation in the N-terminal half or in the C-terminal half) can be obtained, for example, by restricting a Bacillus plasmid at an appropriate site and recombining it with a fragment of an E. coli plasmid which contains a marker and sequence parts which are important for replication. Subsequently, where appropriate, those restriction sites which would interfere with the later procedural steps are removed, for example by directed mutagenesis. A new vector is constructed from the resulting plasmid, which vector contains the DNA sequences from the ..

2~ 1 ~4 ~

Bacillus plasmid and the E. coli plasmid which are used for replication, DNA sequences for the promoter, DNA sequences which encode the pre-pro sequence of the protease (obtained, for example, from the plasmid pCLEAN4 of Fig. 2) and a synthetic linker. The restriction map of Fig. 7 represents an example of such a plasmid having the designation pALlP.
In this regard, the synthetic linker is selected such that, following cleavage with appropriate restriction endonucleases, the vector can be combined with the whole original or mutated DNA sequence encoding the mature protease, or with mutated or non-mutated stretches of the structural gene. In order to prepare an expression vector precursor, which, for example, is to be recombined with a mutated N-terminal half of the structural gene, the non-mutated, or optionally already mutated, C-terminal half of the structural gene, for example, is initially inserted, by means of cleaving the synthetic linker, into the vector which is constructed as above and which contains the said Bacillus and E. coli sequences, the promoter and the pre and pro sequences of the protease, as well as the synthetic linker. Subsequently, the mutated N-terminal half of the structural gene of the protease, which is still lacking, is inserted by means of cleaving the synthetic linker once again. In this way, a vector of the type pALlNC of Fig. 8 is obtained. The procedure for the reverse case is analogous.
In this case, the non-mutated, or optionally already mutated, N-terminal half is initially inserted into a vector and the mutated C-terminal half is incorporated at a later stage into the resulting vector, a vector of the t~pe pALlNC
of Fig. 8 likewise being obtained.
Appropriate bacteria, preferably Bacillus species, in particular Bacillus subtilis, B. licheniformis or B.
alcalophilus, are transformed with the above-described expression vectors. The transformants are subsequently cultured in a known manner and the highly alkaline protease 211~fi~
' .

which is formed is isolated from the culture medium. For this purpose, the expression vectors can be transformed either into bacteria which are still capable of forming indigenous protease, or else into protease-deficient bacteria ~which no longer form indigenous protease). In the case of host organisms which form indigenous protease, the highly alkaline protease accordins to the invention can, if desired, be freed from the indigenous proteases which have formed by subsequent purification operations, for example by high resolution liquid chromatography (HPLC). By contrast, such a purification step can be omitted in the case of protease-deficient host organisms, since these organisms are only ~or essentially only) capable of forming the protease according to the invention.
The following examples are intended to illustrate the invention in further detail without, however, limiting its scope.
Unless otherwise indicated, the methods described in Maniatis et al. (Maniatis et al. = T. Maniatis, E.F.
Fritsch, J. Sambrook, Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, 1982) were generally employed.
Starting vectors used here are available commercially and on an unrestricted basis; otherwise, they can be ~; ~ 25 prepared, by known methods from available vectors.
The Bacillus alcalophilus strain employed in Example 1, and designated Bacillus alcalophilus HAl, was deposited on 28 July 1989 with the Deutsche Sammlung von Mikroorganismen (DSM) ~German Collection of Microorganisms) Braunschweig, Germany under the DSM number 5466.
- , :
~. .

' "~.:;
,: .

2~15~5 . .

Example 1: Preparation of a genomic DNA library from B.
alcalophilus and isolation of the gene for the highly alkaline starting protease.
Chromosomal DNA was isolated from the natural isolate 5 Bacillus alcalophilus HA1 (deposited with the German Collection of Microorganisms under the DSM number 5466) in accordance with the method of Saito et al. ~1963, Biochim.
Biophys. Acta. 72: 619-629) and then partially hydrolysed using the restriction endonuclease Sau3A. The restriction . ~.

~ '' .,,''-', :':
`
. .:
~ :

~ '~

.. .
~ , ,.
:

.:

2 ~

fragments were fractionated by electrophoresis on an agarose gel, and the fragments having a size of from 3 to 8 kilobases (KB) were isolated.
The isolated and size-selected DNA fragments from Bacillus alcalophilus HAl were recombined in vitro with plasmid pUB 110 vector DNA (obtained as described in Example 7).
For this purpose, the plasmid pUB110 was first restricted with the restriction endonuclease BamHI and then dephosphorylated using calf intestinal alkaline phosphatase.
Subsequently, 2 ~g of the restricted and dephosphorylated vector DNA were incubated at 16C for 24 hours together with 8 ~g of the Racillus alcalophilus DNA fragments in a total volume of 100 ~l and in the presence of T4 DNA ligase.
Protoplasts of the strain Baci71us subtilis BD224 were transformed with the resulting DNA, which had been recombined in vitro, in accordance with the method described by S. Chang and N. Cohen (1979, Mol. Gen. Genet. 168~
115). The transformants were selected on plates ; 20 containing neomycin and subsequently transferred to skimmed milk agar. Among 13,800 transformants which were examined, one was found which formed a distinctly larger aureole from proteolysis of the skimmed milk. The plasmid DNA was isolated from this clone in accordance with Maniatis et al. ~ ~ 25 The cloned fragment of B. alcalophilus DNA contained in this plasmid had a size of 4.1 KB and contained the entire information for the highly alkaline protease from Bacillus alcalophilus HAl.
In order to simplify the subsequent procedure, the 4.1 KB-sized DNA fragment was first reduced in size. To do l l , '' ' ' this, the recognition sites for restriction endonucleases located on the DNA fragment were identified by cleaving the plasmid with various restriction endonucleases and:~
fractionating the fragments of the restricted DNA by ~ `
electrophoresis on an agarose gel. A 2.3 KB-sized DNA
fragment, obtained by cleaving with the restriction ~: :
- 20 - ~

. . . ~ . , ' .

2~.:.5~
' !

endonucleases AvaI and HindIII, was identified which contained the complete information for the highly alkaline protease and which was used for the subsequent procedure.
For this, the above plasmid containing the 4.1 KB fragment was restricted with the restriction endonucleases AvaI and HindIII. The 2.3 KB-sized DNA fragment was isolated and ligated to the vector pUB131 (obtained as described in Example 7), which vector had likewise previously been cut with AvaI and HindIII.
The resulting plasmid, which was given the designation pCLEAN4, was introduced into the strain B. subtilis BD224.
The transformants were able to secrete the highly alkaline protease, indicating that the AvaI/HindIII fragment contains the complete structural gene for the highly alkaline protease from B. alcalophilus HA1. The restriction map of the plasmid pCLEAN4 is given in Fig. 2.

Example 2: Sequencing of the structural gene for the highly alkaline protease.
In order to prepare single-stranded DNA of the structural gene of the protease, the plasmid pCLEAN4 was cleaved with the restriction endonucleases AvaI and HindIII
and the approximately 2.3 KB-sized AvaI/HindIII DNA fragment (st~uctural gene of the protease) was introduced into the phagemids pBS (+) or pBS (-). The nucleotide sequence of the protease gene contained in the isolated single-stranded phagemids was determined by the dideoxy chain termination method of Sanger et al. (1977, Proc. Natl. Acad. Sci. USA
74: 5463) and by the method of base-speclfic chemical cleavage of the DNA single strand in accordance with Maxam et al. (1980, in Methods in Enzymology, Grossmann L., Moldave K., eds., Academic Press Inc., New York and London, Vol. 65, 499). The nucleotide sequence which was determined, and the deduced amino acid sequence of the protease, are yiven in Fig. 1. The start of the amino acid sequence of the mature highly alkaline protease, in position 211~65 1190 of the nucleotide sequence, wa~ determined by amino~
acid sequencing the N-terminal end of the highly alkaline protease.

5 Exam~le 3: Preparation of mutated DNA sequences by directed ~
mutagenesis. ~-The directed mutations were carried out in constituent DNA sequences of the structural gene of the protease using ~-the "primer extension" technique described by Kunkel, T.A.
(1985, Proc. Natl. Acad. Sci. USA 82: 488-492). For this purpose, the plasmids pCLMUTN1 (preparation as described in example 4) and pCLMUTC1 (preparation as descr.ibed in Example 5) were employed, being initially converted, as described below, into their uracylated, single-stranded analogs. The starting vectors pCLMUTN1 and pCLMUTC1 do not contain the entire DN~ sequence of the structural gene of the protease from B. alcalophilus HAl, but only the N-terminal half (pCLMUTNl) or the C-terminal half (pCLMUTC1) ther~of.
As descendants of a phagemid, these vectors are ; 20 capable, to a certain extent, of forming single-stranded ; `
vector DNA, which it was possible, under the conditions given here, to transfer out of the host organism being used for replication and then isolate.
Each of these vectors was introduced into E. coli CJ236, as the host organism, using the CaCl2 method, in accordance with Maniatis et al. (pp. 250-251).
Since, when vectors are being replicated, the bacterium E. coli CJ236 (uracil N-glycosylase deficient mutant) incorporates the nucleotide uracil instead of thymine into the DNA sequence of the vector, the uracil-containing analogs of the vectors pCLMUTN1 or pCLMUTC1 were obtained by cultivating the above transformants. These uracil-containing vectors cannot be distinguished from the normal thymine-containing vectors in in vitro reactions. The uracil content of the vector DNA does not interfere with in vi tro DNA syntheses since uracil is not mutagenic either in 211~

vi tro or in vivo, and uracil codes in the same way as thymine. Uracylated vectors can be advantageously employed for the subsequent in vitro reactions of the directed mutagenesis. Once the reactions are complete, the uracil-containing DNA single strand, which served as the templatefor producing mutated DNA strands (vectors), can be removed by treating with uracil N-glycosylase without any phenotypic selection of mutants being required. The glycosylase treatment can be carried out either using the isolated enzyme or else using an E. coli strain possessing uracil N-glycosylase activity which has been transformed with vector DNA.
The uracylated single-stranded DNA of the vectors pCLMUTNl and pCLMUTCl, required as template for the directed mutagenesis, was prepared by cultivating E. coli CJ236 bacteria which had been transformed with one of the two vectors, which bacteria had additionally been infected with ;~ the helper phage M13K07 ~obtained from Bio-Rad Laboratories, Richmond, California).
The helper phage itself is scarcely capable of replication and does not exhibit any interfering interaction ; with the vector DNA of the vectors pCLMUTNl or pCLMUTCl.
Its task consists of synthesizing coat proteins for the uracylated, single-stranded vector DNA which is formed.
Coated single-stranded vector DNA is transferred out of the ~; host organism E. coli CJ236 and can be isolated from the culture medium. The qualitative and quantitative yield of (here of uracylated) single-stranded vector DNA is appreciably increased by the cooperation of the helper phage.
1~ ' ' ' j , .
The isolated, uracylated DNA single-strand vectors pCLMUTNIl or pCLMUTCl were hybridized to the synthetic oligonucleotides, prepared in accordance with Example 8, which containecl a mutation site and simultaneously acted as primers for subsequently completing the DNA double strand possessing a mutation.

21~S~
~:
The synthesis of the second DNA strand was carried out in the presence of nucleotides using T4 DNA polymerase, and the subsequent ligation of the newly formed strand was carried out using T4 DNA ligase (Kunkel et al., 1987, Methods in Enzymol. 154, 367-382). The double-stranded vector DNA which was formed was transformed into E. col i MC1061 and the mutated vectors ~ere identified by testing for the corresponding unique restriction endonuclease recognition sites which had either been introduced or removed by the synthetic oligonucleotides.
In order to prepare, for example, two mutations in either the N-terminal or the C-terminal part of the structural gene of the protease, the procedure of this example was repeated, following the introduction of a first mutation (use of a first synthetic oligonucleotide of Example 6) into one part of the structural gene of the protease, in an analogous manner using a further synthetic oligonucleotide of Example 8 for introducing a second mutation into this part of the structural gene of the protease. In this way, mutated vectors of the type pCLMUTN1 or pCLMUTC1 were obtained which possessed, for example, two mutations in either the N-terminal or the C-terminal part of the structural gene of the protease.

Example 4: Construction of the vector pCLMUTN1.
~` The plasmid pCLEAN4, prepared in Example 1, was cleaved using AvaI. The protruding ends ("sticky ends") were filled in to produce double-stranded DNA using the Klenow fragment of B. col i DNA polymerase I (Maniatis et al., p. 114) in the presence of the necessary nucleotides. After subsequent restriction of this DNA with XbaI, the N-terminal fragment of the protease gene, encompassing 1618 base pairs (BP), was isolated and cloned into the SmaI/XbaI site of pBS. The resulting vector was given the designation pCLMUTN1. The restriction map of this vector is given in Fig. 4.
. :
.

~ .' ' .: ' ` 21~4g~
Example 5: Construction of the vector pCLMUTCl. ;
The plasmid pCLEAN4, prepared in Example 1, was cleaved with the restriction endonucleases XbaI and Asp718. The XbaI/Asp718 double-stranded DNA fragment, encompassing 658 BP and containing the C-terminal half of the structural gene of the protease, was cloned into the XbaI/Asp718 site of pBS. The resulting vector was given the designation pCLMUTCl. The restriction map of the vector is given in Fig. 5. -~
1 0 ~ , Example 6: Synthesis of artificial oligonucleotides for the directed mutagenesis.
Synthetic oligonucleotides were prepared in a Cyclone ~-~
synthesizer (Biosearch) by the method of Beaucage S.L. and Caruthers M.H. (1981, Tetrahedron Letters 22: 1859-1862) using ~-cyanoethyl phosphoramidite. The resulting oligonucleotides were purified by elution from polyacrylamide gels and subsequent desalting using Sephadex G25 columns. Examples of the synthesized nucleotide -sequences (SEQUENCE ID NOS:2-6), and of their properties, ; are given in Fig. 3. The sequences of the synthetic oligonucleotides, used in the procedure according to Example ~; 3 for introducing the mutations into the protease gene, were selected such that they fulfilled the following conditions.
- The DNA sequence of the synthetic oligonucleotide was still sufficiently complementary to the corresponding sequence of the protease gene to ensure that its ability to hybridize was adequate.
Replacement of one or more nucleotides within the I codon, encoding the amino acid to be replaced, by other nucleotides such that this mutated codon now encoded a different amino acid (mutation). That codon was employed for the new amino acid which most frequently ; occurred for the corresponding amino acid in the protease gene.

c 211S~6a ' .
- Replacement of further nucleotides within other codons such that, while the original coding of the amino acid was preserved, recognition sequences for restriction endonucleases were thereby either removed from the protease gene or else newly produced in it. These recognition sites were used in the proceduxe according to Example 3 for facilitating screening for the vectors possessing the mutated DN~ sequences for the novel highly alkaline proteases.
,.. ....
Example 7: Isolation and purification of the plasmid pUBllO
and construction of the vector pUB131.
The plasmid pUBllO was isolated from the strain ~: Bacillus subtilis BD366 by the method of T.J. Gryczan et al.
(1978, J. Bacteriol. 134: 318-329) and subsequently purified by cesium chloride density gradient centrifugation in accordance with Maniatis et al. (p. 93). The vector pUBllO
contains uniquely occurring restriction sites for the restriction endonucleases BamHI and EcoRI, a DNA sequence, as a marker, which encodes antibiotic resistance to neomycin, and DNA sequences ("origin of replication") which are required for replication in Bacillus species.
The plasmid pUBllO which was obtained above was restricted with EcoRI and BamHI. The smaller fragment (790 BP) was replaced by a polylinker consisting of 67 base `~
pairs, which polylinker had previously been isolated as an EcoRI/BglII f~agment from the vector M13tgl31. The resulting vector, having the designation pUB131, is thus a descendant of pUBllO in which the approximately 0.8 KB-sized EcoRI/BamHI fragment has been deleted and replaced by a polycloning site.

; Example 8: Construction of the vectors pUBC131 and pUBC132.
The plasmid pUC18 was cleaved with AatII and PvuII. -35 The 1990 base pair-sized fragment containing the ~-lactamase ~ -gene and the E. coli "origin of replication" was isolated.
- 26 - ~ ~

, ' , ' "

2 1 ~

The protruding ends ("sticky ends") were filled in to produce double-stranded DNA using the Klenow fragment of E.
coli DNA polymerase I (Maniatis et al., p. 114) in the presence of the necessary nucleotides. The fragment was subsequently incorporated into the SnaBI site of the vector pUB131, obtained in accordance with Example 7, resulting in the production of the vector given the designation pUBC131.
The 2187 BP EcoRI/BglII fragment of the vector pUBC131, which was obtained above, was subcloned into the EcoRI/BamHI
site of pBS (+), resulting in the production of the vector given the designation pBSREPU. Subsequently, the NcoI or StyI recognition site, which is present in the DNA sequence for the repU polypeptide in the vector pBSREPU (I. Maciag et al. 1988, Mol. Gen. Genet. 212: 232-240), was eliminated by directed mutagenesis in that the nucleotide sequence CCA TGG
was replaced by the nucleotide sequence CCG TGG (both nucleotide sequences encode the amino acid sequence trypto-phan-proline). This was carried out analogously to the procedure used in Example 3. For this purpose, uracylated single-stranded DNA (SEQUENCE ID NO:7) of the vector pBSREPU
was prepared as the template for the directed mutation for eliminating the NcoI or StyI recognition site.
Subsequently, this template was converted, in analogy with the "primer extension" technique described in Example 3, into the completely double-stranded DNA vector using the following synthetic oligonucleotide (SEQUENCE ID NO:8) (prepared and purified in analogously to the process used to prepare the synthetic oligonucleotides of Example 6) NcoI/StyI
Pro Trp Original repU sequence : A~A GTG AGA CCA TGG AGA GAA AA
Synthetic repU sequence: AAA GTG AGA CCg TGG AGA GAA AA

and the vectors, which were now free of NcoI or StyI
recognition sites, were isolated by transforming and cultivating E. coli MC1061. The 1726 BP EcoRI/ApaI fragment of the isolated vector was inserted into the EcoRI/ApaI site of pUBC131. The new vector, whose restriction map is given in Fig. 6, received the designation pUBC132.

5 Example 9: Construction of the plasmid pALlP. ~;
The plasmid pALlP was prepared by ligating the following three elements:
- The 1201 base pair-sized AvaI/NcoI fragment of pCLEAN4;
the fragment contains the promoter and the pre-pro region of the highly alkaline starting protease.
- The following synthetic linker (SEQUENCE ID NOS:9-10) containing individual recognition sites for the restriction endonucleases NcoI, XbaI and Asp718 which permit the introduction of the mutated N-terminal or C-terminal halves of the protease gene from the mutated vectors pCLMUTN1 and pCLMUTC1, respectively, or the introduction of the whole gene of the starting protease ~ from the plasmid pCLEAN4; :
;~ 5' - CCATGGTCTAGAGGTACCA - 3' ~ 20 3' - CAGATCTCCATGGTTCGA~ - 5' ~ ~ A
NcoI XbaI Asp718 HindII
The above double-stranded synthetic linker with protruding 5' ends was prepared by first separately producing the two single-stranded DNA sequences (SEQUENCE ID NOS:9-10) analogously to the synthesis of the synthetic oligonucleotides in Example 6 and then purifying them. The resulting single strands were subsequently hybridized to each other to form the double strand.
- The 5776 base pair-sized AvaI/HindIII fragment from the vector pUBC132 prepared in Example 8; this fragment contains DN~ sequences for replication and for use as selectable markers in E. coli as well as DNA sequences . . :

for replication and for use as selectable markers in B.
subtilis, B. licheniformis and B. alcalophilus. ~ :
The construction of the vector pALlP was carried out in E. coli MC1061 and the vector was isolated from ampicillin-resistant E. coli transformants. The restriction map of theresulting vector is given in Fig. 7.

Example 10: Construction of the plasmids pALlN and pALlC.
The vectors pALlN and pALlC were constructed in E. coli MC1061, the vectors being isolated from ampicillin-resistant E. coli transformants.
The plasmid pALlN was constructed by initially cleaving the vector pCLEAN4 obtained in Example 1 with the restriction endonucleases NcoI and XbaI and subsequently cloning the resulting NcoI/XbaI fragment into the NcoI/XbaI
site of the vector pALlP (prepared in accordance with Example 9). The vector which was prepared contained the N-terminal part of the DNA sequence encoding the mature enzyme, the regulatory elements for transcribing and translating the highly alkaline protease, as well as the signal and processing sequences.
The plasmid pALlC was constructed by first cleaving the vector pCLEAN4, obtained in Example 1, with the restriction endonucleases XbaI and Asp718 and then cloning the resulting 2S XbaI/Asp718 fragment into the XbaI/Asp718 site of the vector pALlP (prepared in accordance with Example 9). The vector which was prepared contained the C-terminal part of the DNA
sequence encoding the mature protease, the regulatory elements for transcribing and translating the highly ;~ 30 alkaline protease, as well as the signal and processing sequences.

Example 11: Construction of the expression vectors pALlNC.
Expression vectors were prepared having mutations in the C-terminal part of the protease DN~ sequence, as were those having mutations in the N-terminal part of the , : , 2 1 1 ~

protease DNA sequence, those having mutations in the N-terminal and C-terminal parts of the DNA sequence, and, for comparative purposes, also those without mutations in the protease DNA sequence.
A. Expression vector having mutations in the N-terminal part of the D~A sequence of the protease The mutated vector pCLMUTN1, which was obtained by directed mutagenesis in accordance with Example 3, was cleaved with the restriction endonucleases NcoI and XbaI.
The isolated 411 base pair-sized NcoI/~baI fragment (mutated N-terminal part of the structural gene of the protease having, respectively, the mutations N42R, N114R, I43Q, N42R/N114R and I43Q/N114R, N42R/I43Q, N42R/I43Q/N114R) was cloned into the NcoI/XbaI site of the plasmid pALlC obtained in accordance with Example 15. The vector obtained in each ~ case represents a complete expression vector having a ; ~ suitable reading frame for expressing a mutated protease.
The vectors were designated as follows~
20 pALN42R = expression vector for the protease having the mutation N42R;
pALI43Q = expression vector for the protease having the mutation I43Q;
pALN114R = expression vector for the protease having ; 25 the mutation N114R;
pALN42R/N114R = expression vector for the protease having the mutations N42R/N114R;
pALI43Q/N114R = expression vector for the protease having -the mutations I43Q/N114R.
pALN42R/I43Q = expression vector for the protease having the mutations N42R/I43Q.
pALN42R/I43Q/
N114R = expression vector for the protease having ;
the mutations N42R/I43Q/N114R.

~ ' ' :''~ ,' 21~5~
, B. Expression vector having mutations in the C-terminal part of the DNA sequence of the protèase The mutated vector pCLMUTCl, which was obtained by directed mutagenesis in accordance with Example 3, was cleaved with the restriction endonucleases XbaI and Asp718.
The isolated 609 base pair-sized XbaI/Asp718 fragment (mutated C-terminal part of the structural gene of the protease having, respectively, the mutations M216Q, T249E
and M216Q/T249E) was cloned into the XbaI/Asp718 site of the plasmid pALlN obtained in accordance with Example 10. The vector obtained in each case represents a complete expression vector having a suitable reading frame for expressing a mutated protease. The vectors were designated as follows:
15 pALM216Q = expression vector for the protease having the mutation M216Q;
pALT249E = expression vector for the protease having the mutations T249E;
pALM216Q/T249E = expression vector for the protease having .
the mutations M216Q/T249E.

C. Expression vector having mutations in the C-terminal and N-terminal part of the protease DNA sequence The expression vector pALN114R, prepared in this example under A., and the expression vector pALM216Q, prepared under B., were in each case cleaved with the restriction endonucleases XbaI and AvaI. The 1612 base pair-sized fragment of the plasmid pALN114R was ligated to the 6388 base pair-sized fragment of the plasmid pALM216Q.
The resulting plasmid was given the designation ,, , , j , , pALN114R/M216Q.
The following expression vectors were prepared in an analogous manner:
pALN114R!M216Q
35 pALI43Q/M216Q
pALN42R/M216Q
. .

2 ~ 6 ~ .~
.:
.
pALI43Q/M216Q/T249E ~`
pALI43Q/N114R/T249E
pALI43Q/N114R/M216Q/T249E
- pALN42R/I43Q/N114R/M216Q/T249E
pALN114R/T249E
pALI43Q/N114R/M216Q
pALN114R/M216S
pALI43Q/M216S/T249E
pALN42R/M216A
pALN42R/M216A/T249E. ~ -D. Expression vector containing the non-mutated DNA
sequence of the starting protease The expression vector containing the non-mutated ~-starting structural gene of the protease was obtained by either cloning the 411 base pair-sized, non-mutated NcoI/XbaI fragment from the plasmid PCLEAN4, obtained in accordance with Example 1, into the NcoI/XbaI site of the -`
~; plasmid pALlC, obtained in accordance with Example 10; or by ~-cloning the 609 base pair-sized XbaI/Asp718 fragment from the plasmid pCLEAN4, obtained in accordance with Example 1, into the XbaI/Asp718 site of the plasmid pALlN, obtained in accordance with Example 10. The vectors thus obtained are complete expression vectors having a suitable reading frame for expressing the non-mutated highly alkaline starting protease.
The restriction map of these vectors of the type pALlNC ~ --; is given in Fig. 8.

Exam~le 12: Preparation of the highly alkaline proteases which have been altered by mutation and, for comparative ;~;
purposes, of the starting protease.
50 ml of preculture medium (20 g of tryptone, 10 g of ` yeast extract, 5 g of NaCl, 75 g of soluble starch and 10 ml of corn steep liquor per liter) were in each case inoculated with a colony of the strains to be tested (in each case B. ~ ~
- 32 - ~ `
. , ` ' :, 21~6~
, subtilis BD224 transformed with one of the pALlNC vectors prepared in accordance with Example 11). The culture was incubated at 37C for 16 hours and at 250 rpm. 50 ml of main culture medium (30 g of soya bean meal, 90 g of potato starch, 10 g of Na caseinate and 10 ml of corn steep liquor per liter) were inoculated with 2.5 ml of this culture. The main culture was incubated uncler conditions which were identical to those used for the preculture. After 72 hours, the cultures were centrifuged. The proteases which had been formed were precipitated from the culture supernatants using acetone and subsequently purified as follows: cation exchanger Mono S, FPLC; elution using an ascending gradient of 20 mmole to 200 mmole ammonium acetate, pH = 6.
The foregoing description and examples have been set forth merely to illustrate the invention and are not intended to be limiting. Since modifications of the disclosed embodiments incorporating the spirit and substance of the invention may occur to persons skilled in the art, the invention should be construed to include everything within the scope of the appended claims and equivalents thereof.
~, :'' ' - 33 - :: :

. ' ,~: .... . ~ '

Claims (19)

1. A highly alkaline protease having an amino acid sequence at least 80% homologous to the amino acid sequence of Fig. 1, and wherein from 2 to 5 of the following substitutions have been effected the amino acid in the position corresponding to position 42 of Fig. 1 is replaced by arginine, the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine, the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine, the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by an amino acid selected from the group consisting of serine, glutamine and alanine, and the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid.

2. A highly alkaline protease according to claim 1, wherein said protease has an amino acid sequence at least 90% homologous to the amino acid sequence of Fig. 1.

3. A highly alkaline protease according to claim 2, wherein said protease has an amino acid sequence at least 95% homologous to the amino acid sequence of Fig. 1.

4. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 216 of Fig 1 is replaced by serine, glutamine or alanine, and the amino acid in the position corresponding to position 43 of Fig 1 is replaced by glutamine.

5. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 216 of Fig 1 is replaced by serine, glutamine or alanine, and the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine

6. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by serine, glutamine or alanine, and the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid.

7. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by serine, glutamine or alanine; the amino acid located in the position corresponding to position 114 of Fig. 1 is replaced by arginine, and the amino acid in the position corresponding to position 42 of Fig. 1 is replaced by arginine.

8. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by serine, glutamine or alanine; the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine, and the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine.

9. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine; the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid, and the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine.

10. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine; the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid, and the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by serine, glutamine or alanine.

11. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine; the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine; the amino acid in the position corresponding to position 249 of Fig.
1 is replaced by glutamic acid, and the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by serine, glutamine or alanine.

12. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position-corresponding to position 216 of Fig. 1 is replaced by glutamine.

13. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine, and the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine.

14. A highly alkaline protease according to claim 1, having an amino acid sequence in which the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine, and the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid.

15. A DNA sequence encoding a highly alkaline protease having an amino acid sequence at least 80% homologous to the amino acid sequence of Fig. 1, and wherein from 2 to 5 of the following replacements have been effected:
the amino acid in the position corresponding to position 42 of Fig. 1 is replaced by arginine, the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine, the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine, the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by an amino acid selected from the group consisting of serine, glutamine and alanine, and the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid.

16. A washing composition comprising a highly alkaline protease according to claim 1, and at least one conventional detergent ingredient.

17. A composition according to claim 16, comprising said highly alkaline protease and at least one further enzyme selected from the group consisting of proteases, lipases, pectinases, amylases, nucleases, oxidoreductases and cellulases.

18. A composition according to claim 16, comprising said highly alkaline protease and at least one further ingredient, selected from the group consisting of surfactants, builders, bleaching agents and bleach activators.

19. A process for preparing a highly alkaline protease comprising the steps of cultivating a transformed microorganism which contains an expressible DNA sequence encoding an amino acid sequence at least 80% homologous to the amino acid sequence of Fig.
1, and wherein from 2 to 5 of the following substitutions have been effected:
the amino acid in the position corresponding to position 42 of Fig. 1 is replaced by arginine, the amino acid in the position corresponding to position 43 of Fig. 1 is replaced by glutamine, the amino acid in the position corresponding to position 114 of Fig. 1 is replaced by arginine, the amino acid in the position corresponding to position 216 of Fig. 1 is replaced by serine, glutamine or alanine, and the amino acid in the position corresponding to position 249 of Fig. 1 is replaced by glutamic acid, whereby said highly alkaline protease is produced by said transformed microorganism; and isolating said highly alkaline protease.