WO2013160365A1 - Stereoselective hydroxylation of benzenes - Google Patents

Description

STEREOSELECTIVE HYDROXYLATION OF BENZENES

The present invention relates to novel cytochrome P450 monooxygenase with modified (higher) substrate selectivity and/or (re)activity towards aromatic hydroxylation of benzenes. In particular, the invention provides nucleic acid sequences coding for the novel enzymes, expression constructs and vectors comprising these sequences, transformed microorganisms expressing these novel enzymes, and processes for the stereoselective hydroxylation of substituted benzenes useful as building blocks in vitamin E synthesis. Regioselective hydroxylation of aromatic rings leading to enantiopure products is of high interest for producing precursors of pharmaceuticals, flavors or vitamins. Especially, the demand for phenol and alkylphenols as feedstock for resins, plastics and bisphenol-A is constantly increasing. Until today, phenols are mainly produced through a three-step cumene process (the so-called "Hock-process") or alternatively isolated from natural sources such as tar coal, biomass or gasification processes. The latter require often intensive extraction procedures with high energy consumption, especially for separating the isomeric

alkylphenols. A main disadvantage of the Hock-process is the low product yields: only about 5% of the initially used benzene is converted to phenol. Novel synthetic routes to phenols mainly focus on development of metal-catalysts generating reactive oxygen species for the hydroxylation of benzene educts.

Substituted benzenes are also used as building blocks such as e.g. in vitamin E synthesis, wherein 2,3,6-trimethylphenol is converted via oxidation and catalytic hydrogenation into trimethylhydroquinone (TMHQ), in important intermediate in vitamin E synthesis (for more detail see e.g. Ullmann's Encyclopedia of Industrial Chemistry, 6^th, completely revised edition, Volume 38, Wiley-VCH, 2002). Improving the known vitamin E synthesis is advantageous in order to save costs and resources.

Direct hydroxylation with enzymes can be an attractive alternative reducing reaction and purification steps, as well as waste generation and process energy demands. Cytochrome P450 monooxygenases (CYPs), a protein superfamily with already more than 5000 cloned members, catalyze oxygenation reactions with molecular oxygen at room temperature. Direct aromatic ring hydroxylation is a synthetically attractive reaction which has been reported for several P450 monooxygenases. An example of such P450 monooxygenase is the cytochrome P450 CYP102A1 from Bacillus megaterium commonly referred to as the P450 BM3. P450 BM3 is a water-soluble enzyme of 1 18 kDa. As natural fusion protein, P450 BM3 contains its heme and reductase domains on a single polypeptide chain. Very recently, a model has been proposed for regioselective aromatic hydroxylation for substituted benzenes catalyzed by P450 BM3 (Whitehouse et al. , ChemBioChem 2010, 1 1, 2549-2556). However, since they are difficult to express in the established host systems and are rather sensitive to inactivation in the isolated state, these enzymes have so far hardly been used for

biotechnological processes (Bernhardt, J. Biotechnol. 2006, 124, 128-145).

For example, direct aromatic hydroxylation of p-xylene to 2, 5-Dimethylphenol (2,5-DMP) is of high interest for the production of temperature stable polymers (>500 K) as well as a building block for so-called ^"next-generation^" plastics. Only a few enzymes (toluene 4-monooxygenases; microsomal P450 monooxygenases) were reported catalyzing the aromatic hydroxylation of p-xylene to the corresponding phenol (Iyer et al, Biochemistry 1997, 36, 7136-7143).

Unfortunately, enzymes such as toluene 4-monooxygenase suffer from low activity (k_cat 36 min^"1 ), stability (four-protein complex) as well as selectivity (20% aromatic and 80% benzylic hydroxylation) using p-xylene as substrate.

Stereoselective aromatic ring hydroxylation can be furthermore interesting in the synthesis of vitamin E. Thus, there is a strong need for novel biocatalysts to be used in industrial processes which are capable of direct aromatic hydroxylation of substituted benzenes, in particular hydroxylation of p-xylene or pseudocumene, into the respective alkylphenol, wherein the biocatalyst is modified in such a way that it shows increased enzymatic activity, coupling efficiency, product stability and stereoselectivity.

Surprisingly, we found several novel P450 mutants which show an increased activity of up to 30 times towards aromatic hydroxylation of methylbenzenes, in particular p-xylene or pseudocumene, compared to the activity of a wild-type P450. Furthermore, the coupling efficiency and the selectivity for the production of alkylphenols could be increased. The inventors have identified at least 5 amino acid positions which are involved in the specific enzyme activity, coupling efficiency and/or selectivity of the enzyme. In particular, these at least 5 positions are highly important for aromatic hydroxylation of a given substrate. The present invention provides isolated polynucleotides encoding the mutant polypeptides, i.e. P450 enzymes, of the present invention, nucleic acid constructs, recombinant expression vectors, and recombinant host cells comprising the polynucleotides, and to methods of producing the polypeptides. Novel polynucleotides, polypeptides and recombinant host cells can be used in a process for producing alkylphenols from substituted benzenes, e.g., production of 2,5-DMP from p-xylene. Useful substrates include p-xylene, m-xylene, o- xylene, toluene or pseudocumene. Preferred is the catalysis of pseudocumenes or p-xylene using the novel enzymes.

In one embodiment, the present invention is directed to a modified enzyme having P450 monooxygenase activity, comprising at least one mutation, in particular amino acid substitution, wherein the at least one mutation leads to an increase in the monooxygenase activity and/or selectivity and/or coupling efficiency, i.e. is said to be a functional mutation, and wherein the at least one mutation is at one or more amino acid positions selected from the group consisting of amino acid positions corresponding to positions 47, 51 , 87, 330, 401 and combinations thereof of the amino acid sequence of Bacillus megaterium P450 BM3 as shown in SEQ ID NO:2, wherein SEQ ID NO:2 shows the wild-type sequence. The corresponding polynucleotide encoding the wild-type P450 BM3 is shown in SEQ ID NO: 1 (for both sequences, see Figure 4). Preferably, the wild- type P450 BM3 according to SEQ ID NO:2 is mutated wherein the mutated or modified enzyme comprises a mutation in one or more of the amino acid residues R47, Y51 , F87, A330 and/or 1401 . Particularly preferred are amino acid substitutions but also other forms of mutations known to the skilled person are possible. In a preferred embodiment, the modified/mutated P450 monooxygenase of the present invention comprises combinations of mutations, in particular amino acid substitutions. Such combinations may include three of the positions mentioned above, e.g. a substitution of amino acid residues corresponding to amino acids R47/Y51 /I401 on SEQ ID NO:2, more preferably a combination of substitutions R47S/Y51W/I401M or R47E/Y51W/I401M.

In a further preferred embodiment, the mutated enzyme comprises a

combination of 2 mutations mentioned above, such as e.g. amino acid

substitutions on positions corresponding to positions R47/Y51 on SEQ ID NO:2, more preferably a combination of substitutions R47S/Y51W or R47E/Y51W. The modified enzyme according to the present invention preferably contains at least 1 , at least 2, at least 3, at least 4 or at least 5 mutations, e.g.

substitutions, on one of the above-identified positions when compared with the amino acid sequence of the corresponding non-modified monooxygenase as exemplified by SEQ ID NO:2. The respective positions corresponding to amino acid residues shown in SEQ ID NO:2 are listed in Table 1 .

All these modifications mentioned above leading to a mutant P450

monooxygenase, in particular a mutant P450 BM3, are used in a process for direct aromatic ring hydroxylation of substituted benzenes, such as preferably p- xylene or pseudocumene. Particularly useful amino acid substitutions are the ones selected from R47S, R47E, Y51W, I401M or the mutation pattern shown in Table 1 for mutants M1 , M2, M3, and functional equivalents thereof. Residue F87 proofed to be essential for efficient coupling and aromatic hydroxylation, in particular with p-xylene or pseudocumene as substrate. It is also possible to substitute the original amino acids on the above-mentioned positions by other functional equivalent amino acids.

As used herein, "functional equivalents" are to be understood as mutants which exhibit, in at least one of the abovementioned sequence positions, an amino acid substitution other than the one mentioned specifically, but still lead to a mutant which, like the mutant which has been mentioned specifically, show the same properties with respect to the wild -type enzyme and catalyze at least one of the abovementioned hydroxylation reactions.

"Functional equivalents" also encompass the mutants which can be obtained by one or more additional amino acid additions, substitutions, deletions and/or inversions, it being possible for the abovementioned additional modifications to occur in any sequence position as long as they give rise to a mutant with a modified profile in the above sense.

The mutant enzymes, in particular the recombinant P450 BM3 mutants, show increased specific activity in the hydroxylation of a given substrate such as e.g. p-xylene or pseudocumene, increased coupling efficiency, and increased

(enantio)selectivity for production of the respective alkylphenols. The activity increase is in the range of about at least 5, 10, 20-fold but can be up to 30-fold and more, such as e.g. obtained with mutant M2 catalyzing the aromatic hydroxylation of p-xylene or pseudocumene in comparison to the wild-type enzyme. Regarding the coupling efficiency, this could be increased to more than about 40, 50%, such as 60 and even 70%, compared to the wild-type enzyme.

In one embodiment, aromatic hydroxylation of p-xylene or pseudocumene using a modified enzyme according to the present invention, the selectivity for 2,5 DMP can be increased to above 98% or more, such as even 100%, compared to a process using wild-type P450 BM3.

As used herein, the term "specific activity" or "activity" with regards to enzymes means its catalytic activity, i.e. its ability to catalyze formation of a product from a given substrate. The specific activity defines the amount of substrate consumed and/or product produced in a given time period and per defined amount of protein at a defined temperature. Typically, specific activity is expressed in μητιοΐ substrate consumed or product formed per min per mg of protein. Typically, μιηοΐ/min is abbreviated by U (= unit). Therefore, the unit definitions for specific activity of mol/min/(mg of protein) or U/(mg of protein) are used interchangeably throughout this document.

In a preferred embodiment, the hydroxylation of methylated benzenes, such as e.g. p-xylene or pseudocumene, is catalyzed by mutant M2, leading to 30-fold increased activity compared to wild-type P450-BM3, excellent coupling

efficiency (46%) and selectivity (>98%) for the respective product such as e.g. production of 2,5-DMP in the case of p-xylene as substrate. A catalytic constant (Kcat) in the range of 1953 min^"1 can be achieved and an increase of 46% regarding the coupling efficiency.

As used herein, the terms "enzyme", "P450 monooxygenase" or "cytochrome P450 monooxygenase" are used interchangeably herein in connection with the description of the present invention.

A wild-type P450 monooxygenase may include any P450 monooxygenase including enzymes isolated from microorganisms, yeast or mammals that are used as starting point for designing mutants with increased activity, increased coupling efficiency and/or (enantio)selectivity according to the present invention. Preferably, the wild -type enzyme is selected from the P450

monooxygenase of Bacillus, Geobacillus, Erythrobacter, Burkholderia,

Herpetosiphon, Ralstonia, Bradyrhizobium, Azorhizobium, Streptomyces,

Rhodopseudomonas, Rhodococcus, Delftia, Saccharopolyspora, Comamonas, Burkholderia, Cupriavidus, Variovorax, Fusarium, Gibberella, Aspergillus or Amycolatopsis, more preferably selected from Bacillus megaterium, Bacillus subtilis, Bacillus licheniformis, Bacillus weihenstephanensis, Bacillus cereus, Bacillus anthracis, Burkholderia sp. 383, Erythrobacter litoralis, Geobacillus sp. Y412MC10, Herpetosiphon aurantiacus, Ralstonia eutropha, Ralstonia pickettii, Ralstonia metallidurans, Bradyrhizobium japonicum, Azorhizobium caulinodans, Streptomyces avermitilis, Rhodopseudomonas palustris, Rhodococcus ruber, Rhodococcus sp. NCIMB 9784, Delftia acidovorans, Saccharopolyspora erythraea, Comamonas testosteroni, Burkholderia mallei, Cupriavidus taiwanensis,

Variovorax paradoxus, Fusarium oxysporum, zeae, Gibberella moniliformis, Aspergillus fumigatus or Amycolatopsis orientalis, most preferably Bacillus megaterium, Bacillus subtilis, Bacillus licheniformis, Bacillus

weihenstephanensis, Bacillus cereus, Bacillus anthracis, in particular Bacillus megaterium or Bacillus subtilis, such as Bacillus megaterium (P450 BM3) shown in SEQ ID NO:2 or homologous sequences thereof showing the same enzymatic activity. An example of a useful mammalian enzyme is the human CYP2D6.

"Wild-type" in the context of the present invention may include both P450 monooxygenase sequences derivable from nature as well as variants of synthetic P450 enzymes. The terms "wild-type P450" and "non-modified P450" are used interchangeably herein. A "mutant", "mutant enzyme", or "mutant P450 enzyme", such as P450 BM3 enzyme, may include any variant derivable from a given wild-type enzyme/ P450 (according to the above definition) according to the teachings of the present invention and being more active/ efficient in coupling/selective than the respective wild-type enzyme. For the scope of the present invention, it is not relevant how the mutant(s) are obtained; such mutants may be obtained, e.g. , by site-directed mutagenesis, saturation mutagenesis, random

mutagenesis/directed evolution, chemical or UV mutagenesis of entire cells/organisms, and other methods which are known in the art. These mutants may also be generated, e.g. , by designing synthetic genes, and/or produced by in vitro (cell-free) translation. For testing of specific activity, mutants may be (over-)expressed by methods known to those skilled in the art. Different test assays are available, such as e.g. pNCA-assays, 4-AAP-assay or pNTP-assay which are all known to the skilled person. The terms "mutant P450 enzyme" and

"modified P450 enzyme" are used interchangeably herein.

The invention also relates to nucleic acid sequences coding for one of the monooxygenases according to the invention. Preferred nucleic acid sequences are derived from SEQ ID NO:1 or homologous sequences thereof, which have at least one nucleotide substitution which leads to one of the functional amino acid mutations described above. The invention moreover relates to functional analogs of the nucleic acids obtained by addition, substitution, insertion and/or deletion of individual or multiple nucleotides, which furthermore code for a

monooxygenase having the desired improvements mentioned above.

The invention also encompasses those nucleic acid sequences which comprise so- called silent mutations or which are modified in comparison with a specifically mentioned sequence in accordance with the codon usage of a specific origin or host organism, and naturally occurring variants of such nucleic acid sequences.

The invention also encompasses modifications of the nucleic acid sequences obtained by degeneration of the genetic code (i.e. without any changes in the corresponding amino acid sequence) or conservative nucleotide substitution (i.e. the corresponding amino acid is replaced by another amino acid of the same charge, size, polarity and/or solubility), and sequences modified by nucleotide addition, insertion, inversion or deletion, which sequences encode a

monooxygenase according to the invention having a modified

activity/selectivity/coupling efficiency, and the corresponding complementary sequences. The invention furthermore relates to expression constructs comprising a nucleic acid sequence encoding a mutant according to the invention under the genetic control of regulatory nucleic acid sequences; and vectors comprising at least one of these expression constructs. Preferably, the constructs according to the invention encompass a promoter 5'- upstream of the encoding sequence in question and a terminator sequence 3'- downstream, and, optionally, further customary regulatory elements, and, in each case operatively linked with the encoding sequence. Operative linkage is to be understood as meaning the sequential arrangement of promoter, encoding sequence, terminator and, if appropriate, other regulatory elements in such a manner that each of the regulatory elements can fulfill its intended function on expression of the encoding sequence. Examples of operatively linkable

sequences are targeting sequences, or else translation enhancers, enhancers, polyadenylation signals and the like. Further regulatory elements encompass selectable markers, amplification signals, replication origins and the like.

In addition to the artificial regulatory sequences, the natural regulatory sequence can still be present upstream of the actual structural gene. If desired, this natural regulation may be switched off by genetic modification, and the expression of the genes may be enhanced or lowered. However, the gene construct may also be simpler in construction, i.e. no additional regulatory signals are inserted upstream of the structural gene and the natural promoter with its regulation is not removed. Instead, the natural regulatory sequence is mutated in such a way that regulation no longer takes place and the gene expression is increased or reduced. One or more copies of the nucleic acid sequences may be present in the gene construct.

In principle, all natural promoters with their regulatory sequences can be used. In addition, synthetic promoters may also be used in an advantageous fashion. The skilled person will know which and how the promoters are to be used. Examples of suitable promoters are: cos, tac, trp, tet, trp-tet, Ipp, lac, Ipp-lac, laclq, T7, T5, T3, gal, trc, ara, SP6, l-PR or l-PL promoter, which are

advantageously employed in Gram-negative bacteria; and Gram-positive promoters amy and SP02, the yeast promoters ADC1 , MFa, Ac, P-60, CYC1 , GAPDH or the plant promoters CaMV/35S, SSU, OCS, lib4, usp, STLS1 , B33, nos or the ubiquitin or phaseolin promoter. Particular preference is given to using inducible promoters, for example light- and in particular temperature-inducible promoters, such as the PrP1 promoter.

The abovementioned regulatory sequences are intended to allow the targeted expression of the nucleic acid sequences and of protein expression. Depending on the host organism, this may mean, for example, that the gene is expressed or overexpressed only after induction has taken place, or that it is expressed and/or overexpressed immediately.

An enhancement of the regulatory elements may advantageously take place at the transcriptional level by using strong transcription signals such as promoters and/or enhancers. In addition, translation may also be enhanced by improving, for example, mRNA stability.

An expression cassette is generated by fusing a suitable promoter with a suitable monooxygenase nucleotide sequence and a terminator signal or polyadenylation signal. Suitable cloning techniques are described, for example, in T. Maniatis, E. F. Fritsch and J. Sambrook, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. (1989) and in T. J. Silhavy, M. L. Berman and L. W. Enquist, Experiments with Gene Fusions, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. (1984) and in Ausubel, F. M. et al., Current Protocols in Molecular Biology, Greene Publishing Assoc. and Wiley Interscience (1987).

For expression in a suitable host (micro)organism, the recombinant nucleic acid construct or gene construct is advantageously inserted into a host-specific vector which allows optimal gene expression in the host. Vectors are well known to the skilled person and include, for example, phages, viruses, such as SV40, CMV, baculovirus and adenovirus, transposons, IS elements, plasmids, cosmids, and linear or circular DNA. These vectors can be replicated autonomously in the host (micro)organism or chromosomally.

The vectors according to the invention allow the generation of recombinant (micro)organisms which are transformed, for example, with at least one vector according to the invention and which can be employed for producing the mutants. The above-described recombinant constructs according to the invention are advantageously introduced into a suitable host system and expressed. The skilled person knows about preferred cloning and transfection methods in order to bring about expression of the abovementioned nucleic acids in the expression system in question. Suitable systems are known to the skilled person.

Expression of the nucleic acids encoding the mutant monooxygenases according to the present invention can be done in any host system, including

(micro)organisms, which allows expression of the nucleic acids according to the invention, their allelic variants, and their functional equivalents or derivatives. Examples of suitable host (micro)organisms are bacteria, fungi, yeasts or plant or animal cells. Preferred organisms are bacteria such as those of the genera Escherichia, such as, for example, Escherichia coli, Streptomyces, Bacillus or Pseudomonas, eukaryotic microorganisms such as Saccharomyces cerevisiae, Aspergillus, Pichia pastoris, Hansenula polymorpha or Yarrowia lipolytica and higher eukaryotic cells from animals or plants.

The invention furthermore provides a process for preparing a monooxygenase according to the invention, which comprises cultivating a monooxygenase- producing host, e.g. microorganism, if appropriate inducing the expression of the monooxygenase, and isolating the monooxygenase from the culture. If desired, the monooxygenase according to the invention can thus also be produced on an industrial scale.

The microorganism can be cultivated and fermented by known methods.

Bacteria, for example, can be grown in a TB or LB medium at 20 to 40° C and a pH of 6 to 9. Suitable cultivation conditions are described in detail in e.g.

Maniatis et al. (supra).

Fermentation media may further contain suitable carbon substrates. Suitable substrates may include but are not limited to monosaccharides such as glucose and fructose, oligosaccharides such as lactose or sucrose, polysaccharides such as starch or cellulose or mixtures thereof and unpurified mixtures from

renewable feedstock. It is contemplated that the source of carbon utilized in the present invention may encompass a wide variety of carbon containing substrates and will only be limited by the choice of organism. The fermentation of the microorganism can be performed as batch, fed-batch or as continuous process. The use of so-called "resting cells", i.e. cells that are not growing any further, is also within the scope of the present invention.

To isolate the recombinant protein, it is particularly advantageous to use vector systems or oligonucleotides which extend the cDNA by certain nucleotide sequences and thus code for modified polypeptides or fusion proteins which serve to simplify purification. Suitable modifications of this type are, for example, so-called "tags" which act as anchors, such as, for example, the modification known as hexa-histidine anchor, or epitopes which can be recognized as antigens by antibodies (described, for example, in Harlow, E. and Lane, D., 1988, Antibodies: A Laboratory Manual. Cold Spring Harbor (N.Y. ) Press). These anchors can be used to attach the proteins to a solid support such as, for example, a polymer matrix, which can, for example, be packed into a chromatography column, or to a micro titer plate or to another support.

These anchors can also at the same time be used to recognize the proteins. It is also possible to use for recognition of the proteins conventional markers such as fluorescent dyes, enzyme markers which form a detectable reaction product after reaction with a substrate, or radioactive markers, alone or in combination with the anchors for derivatizing the proteins.

In one aspect of the invention, a process for direct aromatic hydroxylation of substituted benzenes is provided, comprising:

a1 ) culturing a recombinant host, e.g. microorganism, according to the above definition in a culture medium, in the presence of an exogenous (added) substrate or an intermediately formed substrate, which substrate can be hydrolyzed by the modified monooxygenase according to the invention, or a2) incubating a substrate-containing reaction medium with an isolated modified enzyme according to the invention, and optionally

b) isolating the hydroxylation product formed or a secondary product thereof from the medium.

With regards to the final concentration of the substrates and /or products to be used in a hydroxylation reaction as described herein, a concentration in the range of about 10 to 100 mM and 1 to 1000 mM might be workable.

The skilled person knows which reaction conditions to be used with a mutant enzyme of the present invention. Isolation of the product including e.g. 2,5 DMP or TMHQ is also known. The products formed by a process according to the present invention, i.e. using a modified P450 enzyme in a aromatic hydroxylation reaction with one of the herein-mentioned substrates, in particular

pseudocumene or p-xylene, can be used in industrial processes, e.g. production of polymers or synthesis of vitamin E wherein these products serve as building blocks which result upon further conversion /rearrangement steps to vitamin E. In principle, any substituted benzene can be used for aromatic hydroxylation reaction catalyzed by the modified enzyme according to the present invention. Particularly, a suitable substrate is selected from methylated or methoxylated benzenes, wherein the methylation includes 1 , 2, 3, or all C-atoms of the benzene ring. Examples are toluene, p-/o-/m-xylene, pseudocumene, o-/m-/p- cresol, bromo-/chloro- or nitrobenzenes such as, e.g., p-/m-/o-nitrotoluene or bromoanisole. Possible hydroxylation steps are depicted in Figure 5, wherein the mutant enzyme can be used for step I and/or II, i.e. for introduction of 2 hydroxyl-groups.

Substrates useful in vitamin E synthesis are for example p-xylene (hydroxylated to 2,5 DMP) or pseudocumene (hydroxylated to 2,3,5 or 2,3,6-tremethylphenol), as well as m-xylene, o-xylene or toluene in a process described above using one of the mutated enzymes, wherein the use of M2 is particularly useful. A preferred reaction is the hydroxylation of p-xylene or pseudocumene, in particular catalyzed by M2. When starting from p-xylene or pseudocumene, the mutated enzyme can catalyze the first and second hydroxylation, preferably the first hydroxylation reaction (see Figure 5).

The term ^"vitamin E^" is used herein as a generic descriptor for all tocol and tocotrienol derivatives exhibiting qualitatively the biological activity of a- tocopherol (l UPAC-I UB Recommendation 1981 , Eur. J. Biochem. 123, 473-475, 1982).

The present invention is now described in greater detail with reference to figures 1 to 5 and the following examples. The work leading to this invention has received funding from the European Community^'s Seventh Framework Program (FP7/2007-2013) under Grant Agreement No. 212281 .

Figure 1 : Gas chromatograms show the hydroxylation of p-xylene by P450 BM3 wild-type (3), M1 (4) and M2 (5). As standards 2,5-dimethylphenol (2) as well as 4-Methylbenzylalcohol (1 ) were employed with an identical GC-program. The retention time is blotted on the x-axis [min] , the FID signal is blotted on the y- axis. Further explanations are given in the text.

Figure 2: Kinetic characterization of wild-type P450 BM3 ( A ), variants M1 (·) and M2 (■) for determination of k_cat and K_M with p-xylene as substrate. All measurements were performed in triplicate with less than 10% standard deviation. Fitting of kinetic parameters was achieved using Origin 7.0 software and Hill equation fitting (OriginLab Corporation, Northampton, MA, USA). The amount of p-xylene is blotted on the x-axis [mM], the productivity is blotted on the y-axis [nmol 2,5-DMP*nmol^"1 P450*min^"1] . For further explanation see the text.

Figure 3: Gas chromatograms show the hydroxylation of toluene (1 ), m-xylene (2) and o-xylene (3) by P450 BM3 variant M2 (R47S/Y51W/I401 M). All potential products (a- and aromatic hydroxylation) were available as standards and employed with an identical GC-program (data not shown). The retention time is blotted on the x-axis [min], the FID signal is blotted on the y-axis. For further explanation see the text. Figure 4: Nucleotide sequences (bottom line) and amino acids sequence (upper line) of the wild-type cytochrome P450 BM3 monooxygenases, corresponding to SEQ I D NO: 1 and 2, respectively (see text) .

Figure 5A: Hydroxylation of p-xylene to 2, 5 dimethylphenol (Step I ) and further to 2,5 dimethylhydroquinone (Step II ), wherein Step I and II are hydroxylations which might be catalyzed by P450. For further explanation see the text.

Figure 5B: Hydroxylation of pseudocumene to 2,3,5 trimethylphenol (Step I) and further to 2, 3,5 trimethylhydroquinone (Step II ), wherein Step I and II are hydroxylations which might be catalyzed by P450. Example 1 : General methodology

The 3.1 kb P450 BM3 gene from Bacillus megaterium (CYP102A1 ) was cloned into the 2 kb pET22b-derived pALXtreme-1 a plasmid (Blanusa et al. , Anal. Biochem 406, 141 -146, 201 1 ). For expression of the P450 BM3 monooxygenase the pALXtreme-1 a plasmid was transformed into chemical competent E. coli BL21 Gold (DE3)-derived E. coli laql^Q1 Gold (DE3) cells. Preparation of chemical competent cells was achieved following a standard protocol generating 1 0⁸ clones per μξ pUC19 (Inoue et al. , Gene 1990, 96, 23-28). All plasmid

preparations were performed using the QIAGEN QIAprep Spin Miniprep Kit (Hilden, Germany). All chemicals used were of analytical grade and purchased from Sigma Aldrich (Steinheim, Germany), abcr (Karlsruhe, Germany) or AppliChem (Darmstadt, Germany). Oligonucleotides used for mutagenesis were obtained from Eurofins MWG Operon (Ebersberg, Germany) and produced at HPLC purity. Enzymes were received from New England Biolabs (Frankfurt, Germany) and Sigma Aldrich (Steinheim, Germany) . dNTPs were purchased from Fermentas (St. Leon-Rot, Germany).

For protein activity measurement of the P450 BM3 mutant libraries, frozen pellets of grown expression cultures stored for 4h at -20° C in 96-deep-well plates were resuspended by retro-pipetting in 150 μΐ 50 mM potassium phosphate buffer (KPi ) pH 7.5. Additional 1 50 μΐ 50 mM KPi pH 7.5 were added containing 5 mg/ml lysozyme. Plates were incubated for 1 h at 900 rpm and 37° C in an Infors HT Microtron shaker to ensure entire cell lysis. Plates with lysed cells were centrifuged for 15 min at 3200 g and 4° C in an Eppendorf 581 OR centrifuge to spin down cell debris. The clear supernatant was used for assaying

monooxygenase activity by measuring NADPH depletion at 340 nm in a Tecan Sunrise MTP reader (Groeding, Austria; Glieder & Meinhld, Methods Mol. Biol 230, Clifton, N.J. , USA, 157-170, 2003) . 96-well quarz glass MTPs (Hellma, Muellheim, Germany) were used during all measurements with p-xylene since the substrate reacts with most poly-propylene and poly-styrene based materials. A reaction mixture in MTP format contained 1 .5 % DMSO, 200 mM p-xylene, 10 to 50 μΐ cell lysate and 0. 16 mM NADPH in a final volume of 250 μΐ 50 mM KPi buffer pH 7.5. The amount of lysate had to be adjusted in a way that linear depletion of NADPH could be monitored reliable for 2 min. The reaction was stopped by pipetting 25 μΐ quenching buffer (4 M urea in 0.1 M NaOH) in the reaction mixture, followed by addition of 20 μΐ 4-aminoantipyrine (4-AAP) (5 mg/ml) and 20 μΐ potassium peroxodisulphate (5 mg/ml) for phenolic product quantification (Wong et al. , J. Biomol. Screening 2005, 10, 246-252). After 30 min incubation at 750 rpm on an Eppendorf MixMate MTP shaker (Hamburg, Germany) the amount of 2,5-DMP was quantified at 509 nm absorbance in a Tecan Sunrise MTP reader (Wong et al. , supra). Mutants for next rounds of saturation were selected based upon best performance for NADPH depletion as well as highest formation of 2,5-DMP. Improved variants of P450 BM3 were sequenced at MWG Eurofins DNA (Ebersberg, Germany) and analyzed using Clone Manager 9 Professional Edition software (Scientific & Educational Software, Cary, NC, USA).

Example 2: Generation of P450 BM3 mutants

Altogether five amino acid residues were chosen for iterative NNK saturation (R47, Y51 , F87, A330 and 1401 ) . Amino acid residues R47 and Y51 were

simultaneously saturated meanwhile sites F87, A330 and 1401 were saturated as single positions. Oligonucleotides were designed for optimal PCR product yields according to recommendations of user's manual from Stratagene's QuikChange Site-directed mutagenesis Kit (Stratagene, La Jolla, CA, USA). A standard PCR set-up contained 20 ng plasmid DNA, 5 U Phusion DNA polymerase, 1 x Phusion DNA polymerase buffer, 0.2 mM dNTP mix and 0.4 μΜ of forward and reverse primer (Table 2). All PCRs were performed in thin wall PCR tubes (Saarstedt, Nuembrecht, Germany) and using an Eppendorf Mastercycler proS (Hamburg, Germany). The cycling protocol was performed according to Stratagene's QuikChange Site-directed mutagenesis Kit manual instruction (Stratagene, La Jolla, CA, USA). Temperatures and incubation times during each PCR cycle were adjusted to provider's recommendation for Phusion DNA polymerase (New England Biolabs, Frankfurt, Germany). Annealing temperature during each PCR cycle was adjusted to 55 °C for 30 s. Successful PCR amplifications were verified on 0.8 % agarose gel according to a standard protocol (Maniatis et al. , 1982 1 - 545). Digestion of template DNA was achieved by addition of 20 U Dpnl to 50 μΐ PCR mixture followed by incubation at 37° C for 3 h. The PCR products were transformed without further purification into chemically competent E . coii Gold (DE3) laql^Q1 cells using a heat shock transformation protocol (Hanahan et al. , Methods in enzymology 1991 , 204, 63-1 13). Recovered cells were plated and incubated overnight at 37° C on LB agar plates containing 100 Mg/ml kanamycin. Obtained colonies were transferred with sterile toothpicks into Greiner BioOne 96-well microtiterplates (MTP) (Frickenhausen, Germany) pre-filled with 100 μΐ LB media and 100 Mg/ml kanamycin. The plates were incubated for 14 h at 900 rpm and 37° C in an Infors HT Microtron shaker (Bottmingen, Switzerland) followed by addition of 100 μΐ of 50% glycerol (W/V). All MTPs were stored at - 80° C prior to expression of mutant libraries. For successful oversampling, a library size of 180 clones for single site-saturated positions (F87, A330 and 1401 ) and 2000 clones for double site- saturation (R47/Y51 ) was generated to achieve a confidence level of 99% for all possible codons (Firth et al. , Nucleic Acids Res. 2008, 36, W281 -285). The results are shown in Table 1 .

Table 1 . Comparison of the amino acids in the wild-type P450-BM3 (WT) and in mutant enzymes (M1 to M4), wherein the amino acid substitution is indicated. The positions are according to the ones in SEQ ID NO:2. WT M1 M2 M3 M4

R47 R47S R47S R47E

Y51 Y51W Y51W Y51W

F87 F87 F87 F87

A330 A330 A330 A330P A330

1401 1401 I401 M 1401

Table 2. Oligonucleotides used for NNK saturation of targeted amino acid residues.

Example 3: Expression and purification of P450 BM3 mutants

Expression of P450 BM3 mutant libraries was achieved in Eppendorf 96-deep-well plates (Hamburg, Germany) as described by Nazor et al., Protein Eng. , Des. Sel. 2008, 21, 29-35. The grown cultures were centrifuged for 15 min at 2900 g and 4° C in an Eppendorf 5804R centrifuge (Hamburg, Germany) for removal of residual expression media. Obtained pellets were washed in 50 mM KPi pH 7.8 and spin down again applying the same centrifugation program (15 min; 2900 g; 4° C). The supernatant was removed and pellets were stored at -20° C for 12 h. For purification, frozen cell pellets were resuspended in 10 ml 50mM KPi buffer pH 7.8 prior to disruption in an Avestin EmulsiFlex-C3 high-pressure homogenizer (Ottawa, ON, Canada) by applying three cycles of 1500 bar pressure. Cell debris was removed by centrifuging homogenized samples for 30 min at 16000 g and 4° C in a Sorvall RC-6 Plus ultracentrifuge (Thermo Scientific, Rockford, IL, USA). Removal of small insoluble particles was achieved by pressing the supernatant through a 0.45 μΜ membrane filter. The filtered cell lysate was loaded on a Kronlab TAC15/ 125PE5-AB-2 column (YMC Europe GmbH, Dinslaken, Germany) prefilled with Toyopearl DEAE 650S anion exchange matrix (Tosoh Bioscience, Stuttgart, Germany) . An AKTAprime Plus chromatography system with UV- detection (GE Healthcare, Muenchen, Germany) was used for purification and collection of eluted protein samples. A standard protocol was applied to purify P450 BM3 wild-type as well as the variants M1 and M2 (Schwaneberg et al. , J. Chromatogr. , A 1999, 848, 149-159. ). Eluted protein samples were collected and purity was estimated on a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Fractions containing active and approximately 90% pure protein, were pooled in an Amicon Ultra-4 centrifugation tube (30 kDa cutoff; Millipore, Schwalbach, Germany) and concentrated to a final volume of 2 ml. Concentrated protein samples were desalted and equilibrated in 50 mM KPi pH 7.5 buffer using a PD-10 gel-filtration column (GE Healthcare, Muenchen, Germany). Finally enzyme samples were shock-frozen for long time storage in liquid N₂ and lyophilized in an Alpha 1 -2 LD plus freeze-dryer (Christ, Osterode am Harz, Germany) for 48 h at -54° C. Example 4: Kinetic characterization of P450 BM3 variants

Lyophilized P450 BM3 variants were resuspended in 50 mM KPi pH 7.5 buffer and the concentration of active P450 BM3 was assayed applying a standard CO quantification protocol for P450 monooxygenases (Omura & Sato, J. Biol. Chem. 1964, 239, 2379-2385). Kinetic characterization (KM, k_cat) of wild type P450 BM3 and improved variants was performed using the 4-AAP assay for quantification of 2,5-DMP. Therefore a mixture of 3 ml 50 mM KPi pH 7.5 containing p-xylene (1 to 30 mM), 1 .5 % DMSO (V/V), catalase (1200 U/ml) and P450 BM3 monooxygenase (WT: 0.5 μΜ; M1 : 0.1 33 μΜ; M2: 0.036 μΜ) was pre-incubated for 5 min in an air- closed 25 ml glass-flask shaking on an Eppendorf MixMate (Hamburg, Germany) at 750 rpm. Substrate conversion by P450 BM3 was induced by adding 330 μΐ NADPH (5 mM). Every 10 s 180 μΐ sample were taken from the flask and pipetted onto 25 μΐ quenching buffer (0.1 M NaOH and 4 M urea) prefilled into a 96-well quarz-glass MTP. Finally 20 μΐ of 4-AAP (5 mg/ml) and 20 μΐ of potassium peroxodisulphate (5 mg/ml) were added (Wong et al. , supra). Color development was achieved by shaking the plate for 30 min at 750 rpm on an Eppendorf MixMate. The amount of produced 2,5-DMP was quantified by measuring absorbance at 509 nm in a Tecan Sunrise MTP reader. Fitting of kinetic

parameters was achieved using Origin 7.0 software (OriginLab Corporation, Northampton, MA, USA). All products from P450 BM3 hydroxylation by GC could be confirmed (Figure 1 ). Coupling efficiency of the generated variants was determined by applying the same set up assay mixture concentrations (no catalase was added) that were used for kinetic characterization of P450 BM3 variants. The final substrate concentration was adjusted to 20 mM p-xylene. After 5 min pre-incubation in 5 ml glass cuvettes (Hellma, Muellheim, Germany) the reaction was induced by addition of 166 μΐ 5 mM NADPH. Depletion of NADPH was measured in a Varian Cary 50 spectrophotometer (Agilent Technologies, Darmstadt, Germany) at 340 nm. After one minute incubation, 180 μΐ assay mixture were quenched by transfer into a pre-filled 96-well quarz glass MTP (25 μΐ; 0.1 M NaOH and 4 M urea). Product quantification was achieved via 4-AAP assay as described for screening and kinetic characterization (Wong et al. , supra). Additionally the formation of H₂0₂ by the peroxide-shunt of P450 BM3 was determined as reported by a standard protocol using horse-reddish peroxidase (HRP) (Morawski et al. , Protein Eng. 2000, 13, 377-384). All measurements for kinetic

characterization as well as determination of coupling efficiency were done in triplicate. The results are shown in Table 3. Table 3. Catalytic parameters for hydroxy lation of p-xylene by P450 BM3 (wild- type, M1 and M2).

/(cat : [nmol 2,5-DMP^*nmol^"1 P450^* min^"1]; _M : [mM]; k_eff : [mM^"1*min^"1]; C : coupling efficiency [%];D : 2,5-dimethylphenol [%]; M : 4-methylbenzylalkohol [%].

The wild-type P450 BM3 showed a low activity for p-xylene (k_cat 68 min^"1 ; 45% coupling efficiency), producing 2, 5-DMP with high selectivity (>98%). Variant M1 (R47S/Y51W) showed more than a 7-fold higher product formation (k_cat 500 min^"1 2,5-DMP) compared to the wild-type P450 BM3. Besides the increase in activity, variant M1 showed a remarkable catalytic performance since the coupling efficiency increased by 20% (total coupling efficiency of 54%) accompanied by a slightly decreased K_M (from 7.9 to 7.1 mM). Lower coupling efficiencies were also achieved when using o-xylene as well as m-xylene as substrates (12 and 29%). Variant M2 (R47S/Y51W/I401M) showed an up to 5-fold improved activity compared to M1 . Purified M2 displayed in total a 30-fold higher catalytic activity (kcat 1953 min^"1 ) than the wild-type P450 BM3 as well as a 46% increased coupling efficiency (in total 65%) as well as high selectivity (>98%) for 2, 5-DMP production while the K_M is reduced by 15% to 6.1 mM. The high k_cat value makes the M2 variant an efficient monooxygenase catalyst (k_en: 320.2 mM^"1*min^"1 ), as depicted in Figure 2.

Aromatic hydroxylation of pseudocumene with around 98% selectivity to 2,3,6- trimethylphenol and 2,3,5-trimethylphenol can be achieved with M2. Example 5: Product detection via gas-chromatography

Qualitative product detection using toluene and all three xylene isomers as substrates with P450 BM3 variant M2 was performed via gas chromatography. Therefore 3 ml reaction mixtures were set up as described above for kinetic characterization in a 25 ml shake flask. After addition of 330 μΐ of 10 mM NADPH the mixture was shaken for 30 min at 750 rpm on an Eppendorf MixMate.

Subsequently, additional 330 μΐ of 10 mM NADPH were supplemented followed by 30 min shaking at 750 rpm. Two-phase extraction of products was realized by adding 1 ml chloroform (containing 1 0 mM phenol as internal standard) to the reaction mixture. The reaction mixture was vortexed thoroughly, centrifuged (13000 g; 5 min; Eppendorf 5415R centrifuge, Hamburg, Germany) and separated into a water and organic solvent fraction. The latter was dried over water-free MgS0₄. Extraction was repeated once and both organic fractions were pooled in GC vials and injected into a Shimadzu GC-2010 Plus gas chromatograph

(Duisburg, Germany) using isothermally at 120° C an Optima-17 MS column (Macherey Nagel, Dueren, Germany). As standards all expected phenols (o-/m- /p-cresol), di-methylphenols (DMP) (2,3-DMP, 2,4-DMP, 2,5-DMP, 2,6-DMP, 3,4- DMP, 3,5-DMP) as well as a- hydroxy lated benzenes (benzylalcohol; 2-, 3-, 4- methylbenzylalcohol) were injected applying the identical GC program. The results are shown in Figure 3.

Using M2 with pseudocumene as substrate, two significant peaks were detected after 21 .6 and 22.2 min, respectively, indicating formation of 2,3,6- trimethylphenol and 2, 3,5-trimethylphenol, respectively (data not shown).

Claims

Claims

1. A modified P450 monooxygenase capable of stereoselective hydroxylation of benzene derivatives, wherein the amino acid sequence comprises one or more mutation(s) on a position corresponding to position(s) 47, 51 , 87, 330, 401 , and/or combinations thereof of a Bacillus megaterium P450 monooxygenase according to SEQ ID NO:2.

2. A modified P450 monooxygenase according to claim 1 , wherein the mutation(s) is/are selected from mono- or polyamino acid substitutions R47S, R47E, Y51W, I401M, preferably a combination of R47S/Y51W, R47E/Y51W, R47S/Y51W/I401M, R47E/Y51W/I401M and functional equivalents thereof.

3. A DNA sequence comprising a DNA sequence coding for a P450

monooxygenase as claimed in claim 1 or 2.

4. A vector comprising the DNA sequence of claim 3.

5. The vector of claim 4 which is an expression vector.

6. A host cell which has been transformed by a DNA sequence as claimed in claim 3 or the vector of claim 4 or 5.

7. A process for the aromatic hydroxylation of benzene derivatives, comprising:

a1 ) culturing a host cell according to claim 6 in a culture medium, in the presence of an exogenous substrate or an intermediately formed substrate selected from benzene derivatives, or

a2) incubating the substrate-containing reaction medium with an isolated modified P450 monooxygenase according to claim 1 , and optionally

b) isolating the hydroxylation product formed or a secondary product thereof from the medium.

8. A process according to claim 7 for aromatic hydroxylation of di- and tri- methylated benzenes, in particular p-xylene or pseudocumene.

9. The process as claimed in claim 7 or 8, wherein the mutant P450 monooxygenase comprises one of the following mono- or polyamino acid substitutions selected from R47S, R47E, Y51W, I401M, or mutation pattern as shown in Table 2 for P450 BM3 mutants M1 and M2.

10. A process for the production of vitamin E, wherein the hydroxylation of a substrate selected from benzene derivatives is catalyzed by a modified P450 monooxygenase according to claim 1.