EP0517790A1 - Method for enhancing recovery of recombinant proteins - Google Patents

Abstract

Method for improving the quantity of recombinant protein recovered after fermentation of bacteria containing vectors encoding the recombinant protein. The process is not limited by the required induction of the cloned gene by the addition of exogenous inducers. Thanks to the methods of the invention, the production of recombinant bacterial proteins can exceed a level of 20% of the total soluble protein expressed. The methods of the invention are demonstrated on bacterial cytochrome P450 and on bacterial electron transfer proteins usually difficult to express at commercially achievable levels.

Description

METHOD FOR ENHANCING RECOVERY OF RECOMBINANT PROTEINS

The Government may own certain rights in this invention pursuant to National Institutes of Health Grants GM19036-20.

The present invention relates to a method for enhancing the recovery of recombinant protein from cultured cells. In particular, the invention concerns a method for enhancing recombinant protein recovery from cultured bacterial cells containing such proteins on a plifiable DNA vectors. In more particular aspects, the invention applies to a method for enhancing recovery of bacterial he e proteins such as cytochrome P450_BM-₃, and the related electron transfer proteins such as putidaredoxin reductase and putidaredoxin.

Heme enzymes and related proteins represent a superfamily of proteins which are known to catalyze a variety of oxidative reactions on diverse substrates including endogenous prostaglandins, leukotrienes, steroids and fatty acids as well as exogenous drugs, carcinogens, and pesticides. The ability of this family of proteins to insert an oxygen atom into various organic compounds has been, therefore, the focus of intense investigation. The production of fatty acids from hydrocarbons, the detoxification of pollutants and new routes to pharmacologically-active compounds are but a few of the possibilities arising from reactions catalyzed by these proteins.

Because of the potential commercial usefulness of the products and processes, investigators have strived to obtain commercially useful quantities of such heme proteins. However, the commercial exploitation of the specific reactions catalyzed by the individual members of this family of proteins has been stalled by the absence of adequate sources of the purified enzymes.

Fatty acid monooxygenase (cytochrome P450 or simply P450) represents one member of this class of commercially important enzymes. These monooxygenases have been purified from a variety of mammalian tissues (Kupfer 1980), plants (Kolattukudy 1969; Croteau and Kolattukudy 1975a; Croteau and Kolattukudy 1975b) , yeasts (Tulloch et al. 1962; Heinz et al. 1969), and bacteria (Peterson et al. 1966; Miura and Fulco 1974) . Because of the sequence similarities between the medically important mammalian microsomal P450 (family IV) and the bacterial fatty acid monooxygenase from Bacillus meqaterium (cytochrome P450_BM-. ₃) , a large amount of effort has been expended in characterizing the bacterial enzyme.

In a series of studies on bacterial fatty acid oxidation in B. meqaterium. Fulco and his co-workers have demonstrated the hydroxylation of long chain fatty acids and their respective amides and alcohols by the soluble, carbon monoxide-sensitive, NADPH-dependent monooxygenase in these cells (Miura and Fulco 1974; Miura and Fulco 1975; Hare and Fulco 1975; Ho and Fulco 1976). Unlike previously recognized P450 systems, the bacterial reactions ( onooxygenation and reduction) in this strain were shown to be catalyzed by a single protein of molecular weight 119,000 (Nahri and Fulco 1986). The gene for this bifunctional enzyme has been cloned (Wen and Fulco 1987) , sequenced (Ruettinger et al. 1989) , and expressed in Escherichia coli (Nahri et al. 1988) .

Even though this enzyme is inducible in Bacillus by the inclusion of barbiturates in the growth medium, the amount of protein which is present in these cells is quite low (0.5% of the soluble protein). As a consequence of the low level of enzyme in these cells, recovery of pure enzyme is difficult and the yields are poor (Narhi et al. 1986) . Thus, earlier studies characterizing the ability of these enzymes to metabolize commercially important substrates were difficult to carry out. This is particularly true since the inducibility of the expression of the cloned gene was shown to be limited to Bacillus cells and not to be feasible in other standard bacterial host cells amenable to large scale production (Wen and Fulco 1987) .

The study of the electron transfer reactions from reduced pyridine nucleotides to P450 is also an area of intense investigation. In certain soil bacteria (i.e., Pseudomonas putida) , the oxidation of camphor requires the participation of protein components in addition to the cytochrome an FAD-containing flavoprotein, putidaredoxin reductase and the iron sulfur protein putidaredoxin (Katagiri et al. 1968). In contrast to the P450_BM-_S system which is like mammalian icrosomal P450, the bacterial electron transfer system in P. putida is very similar to the one functional in higher organisms - in particular, it appears similar to the one active in the adrenal cortex mitochondrial metabolism of steroids or the steroid oxidizing enzymes in other endocrine tissues. Here again, research into the mechanism of electron transfer from NADH to P450_CAM in the metabolism of commercially important substrates has been significantly hindered by the limited availability of the proteins responsible for the transfer of electrons to the cytochrome (Geren et al. 1986) . Only low level expression of putidaredoxin reductase has been feasible to date [Roome et al. (1983); Unger et al. (1986)].

Although enzymes such as putidaredoxin reductase may be purified from the wild type strain of P. putida. there are a number of drawbacks to this approach (Roome and Peterson 1988) . It is, for example, difficult to separate putidaredoxin reductase from P450_CAM due to the similarity in size and ionic charge and, the level of putidaredoxin reductase in the wild type cell is significantly below that of P450_CAM. Thus, the cloning and expression of these proteins distinct from one another in a non-pseudomonad is necessary.

While enhanced protein recovery in Bacillus and in

Pseudomonas is desired, the technology for doing so lags significantly behind the technology for use of E. coli as a host cell in large scale fermentations. For instance, numerous methods are available for enhancing the stability of cloned genes in E. coli which are not readily available in other strains of bacteria. In particular, because certain mutant strains of E. coli (DH5 and DH5α) carry deletions in certain genes responsible for aspects of DNA metabolism (recA) , these mutant strains have been found to be useful hosts for maintaining recombinant DNA molecules which contain large inserts (Hanahan 1984) such as those associated with the cytochrome P450 family of enzymes.

Recently, workers in the field interested in enhancing the yield of vector recombinant plasmid and cosmid DNA have utilized these recA-strains of E. coli to study the effect of certain growth media on plasmid yield. When compared to standard media such as Luria Broth (LB) , a culture medium referred to as "Terrific

Broth" (TB) was shown to yield 4 to 7 times more DNA when the plasmids were harvested from DH5α cells grown overnight (Tartof and Hobbs 1987) . However, media such as TB have not been demonstrated to provide substantial increases in protein recovery for recombinant proteins.

In fact, a recent catalog (Bethesda Research Laboratories - Life Technologies, Inc., "The Guide to Frozen Competent Cells - New Applications, New Hosts, New Protocols," (1989)) describes certain advantages of using DH5α and related strains of E. coli in combination with the TB media and methods of Tartof and Hobbs (1987) . The commercial supplier does not recommend DH5α as an expression host. Rather, an expression system is recommended utilizing a vector with a lacl^q promoter (DH5α F*IQ™) • Hosts containing this marker are known to overproduce the lac repressor which regulates transcription from the lac promoter and, therefore, regulates alpha-complementation or expression from the lac promoter.

Methods for enhancing the recovery of recombinant protein for commercially important enzymes such as the cytochrome P450 enzymes and the related putidaredoxin reductase and putidaredoxin are needed. Desirably, such methods would work with other proteins as well. Preferably, such methods would not require expression in the strain from which the recombinant protein was derived since this severely limits the ability to produce large quantities of the recombinant protein where the strain is not readily adapted to scaled-up production. Additionally, such methods would be most useful if they did not require the addition of exogenous inducers of the cloned gene. Commercially useful amounts of these proteins should lead to their use in a wide variety of reactions which require the addition of an atom of oxygen to organic substrates.

The present invention provides a surprisingly simple method for enhancing recovery of recombinant proteins from cultured bacterial cells. The method is not limited to induction by exogenously added inducers such as barbiturates or synthetic inducers of bacterial gene promoters (e.g., inducers for the ,9-galactosidase gene of the lac operon in E. coli) . Furthermore, the method allows for the enhanced expression of heterologous proteins in E. coli. the workhorse standard of the fermentation industry and research laboratories.

A method of enhancing the expression of a recombinant protein is described which is surprisingly simple and straightforward. One may use the method for any desired recombinant protein by obtaining bacterial cells which contain a recombinant DNA vector encoding the desired recombinant protein and then inoculating a protein expression enhancing medium with these bacterial cells. The cells are then cultured in the protein expression enhancing medium to at least the late stationary phase of cell growth in order to allow the cells to express the desired recombinant protein and to allow recovery of the recombinant protein with high efficiency. For the purposes of the invention, a protein expression enhancing medium is a medium which when used in combination with the methods of the invention allows comparable levels of recovery.

The method allows the expression of recombinant proteins to significantly higher levels than heretofore possible without specific genetic manipulation of the regulatory genes encoded along with the recombinant protein. Using the method of the invention, it is possible to obtain recovery of a desired recombinant protein at levels which represent at least fifteen percent of the recombinant protein by wet weight of the total soluble protein from the bacterial host cells.

Further refinements in the method of the invention allow recoveries in excess of 22 percent of the recombinant protein by wet weight of the total soluble protein from the bacterial host cells.

More specifically, for the first time, the present invention allows for the production of certain heme -1-

proteins and of certain electron transfer proteins of bacteria in commercially feasible quantities. Recovery of these proteins using the method of the present invention may exceed 20% of the total soluble protein of the cultured cells.

The invention may successfully enhance the expression of a bacterial cytochrome P450. In a preferred embodiment, enhanced expression of the cytochrome P450_BM.₃ derived from B. meqaterium may be realized. Since this particular cytochrome P450 represents a commercially useful enzyme and since expression of this enzyme in commercially useful quantities has, heretofore, been extremely limited, the methods of the invention provide a means of achieving a long-felt need in the area of expression of cytochrome P450_p. This is particularly true in the instance of the enhanced expression of the cytochrome P 50_BM-₃ derived from B. meqaterium since this enzyme exhibits considerable similarity to certain human cytochrome P450s.

The methods of the invention may also be applied to obtain enhanced expression of bacterial electron transfer proteins. In certain preferred embodiments, the expression of bacterial electron transfer proteins may be realized. In one embodiment, a bacterial redoxin reductase may be recovered in significantly higher amounts than was possible using prior art techniques. Specifically, the methods of the invention were applied to the bacterial redoxin reductase, putidaredoxin reductase, derived from £. putida. In another related embodiment, the bacterial redoxin, putidaredoxin, which is derived from P. putida. was demonstrated to be amenable to the expression enhancement method of the invention. The protein expression enhancing medium of the invention is very rich in nutrients compared to the media typically used in the prior art. In a general sense, the protein enhancing medium should possess several important characteristics. First, the medium should be rich enough in energy source molecules, protein synthesis substrates, and other nutrients to achieve normal logrythmic growth of bacterial cells. Such a medium should also provide for significant increases in the copy number of the vectors encoding the recombinant protein in order to achieve a low ratio of lac repressor molecules in relation to the copies of vector molecules. In addition, the medium should be rich enough so that following completion of logrythmic growth and amplification of vector molecules, when the bacterial cells have entered the stationary phase of growth or when the cells are otherwise stopped from growing, that at least enough of the energy source, molecules, protein synthesis substrates and other nutrients remain unused in order that unrestricted protein synthesis is possible.

Furthermore, the energy source molecule should be a molecule like glycerol such that, while providing ample reserve energy following logrythmic growth for unrestricted protein synthesis, the molecule will not prevent the required release from inhibition by lac repressor of the lac operon providing regulation for the recombinant protein's transcription. Thus, for instance, one could not use as the principal energy source molecule, glucose, sucrose or lactose.

For instance, the expression enhancing medium is at least three times richer than Luria-Bertani (LB) medium. For the purposes of the invention, the richness of the medium refers to the fact that the protein expression enhancing medium contains much more Bactotryptone and Bacto-yeast extract than encountered in the prior art media. Where LB medium would typically contain 10 grams of Bactotryptone and 5 grams of Bacto-yeast extract in a liter of final volume, the protein expression enhancing medium of the invention may contain as much as 12 grams or more of Bactotryptone and as much as 24 grams or more of Bacto-yeast extract. Of course, other sources of tryptone and of yeast extract may be substituted in place of the industry standards, Bactotryptone and Bacto-yeast extract. Moreover, the richness of the medium may be duplicated in a variety of ways including some of the media described in the catalog of the American Type

Culture Collection (Fifteenth Edition, 1982) , e.g. #276, Achromobacter pestifer medium with glycerol; #344, Wheat Peptone Agar (minus the agar) ; or #636, Py Medium (minus the agar) . Any of these formulations or other rich media not listed above may be substituted for TB medium as described. Indeed, it would be possible to duplicate the richness of the medium using a completely chemically- defined medium; however, the use of a completely chemically defined medium would be very expensive to prepare and would obviate at least one of the desirable features of the invention: ease of use and cost. However, such media are expressly included in the term protein expression enhancing medium as used in combination with the method of the present invention.

The protein expression enhancing medium, in certain preferred embodiments, may contain buffering agents. In one such embodiment, the buffering agents used are KH₂P0 and K₂HP0₄. It will be appreciated by those of skill in the art that a variety of buffering systems and a variety of different buffering capacities may be used successfully in combination with the methods of the invention. However, the addition of buffering agents in combination with the methods of the invention are disclosed herein and encompass any such addition to achieve a buffering of the protein expression enhancing medium in excess of the buffering capacity inherent in the other ingredients of the medium.

Further preferred embodiments of the invention include the addition of glycerol to the protein expression enhancing medium. In one such embodiment, glycerol is added at a concentration of at least 0.2% volume of glycerol per volume of medium. Glycerol may be added at a concentration of at least 0.7% volume of glycerol per volume of medium in other embodiments. It is assumed here that one of skill in the art will maximize the concentration of glycerol in the protein expression enhancing medium in combination with the other teachings of the invention.

While a variety of permutations of the protein expression enhancing medium may be practiced within the methods of the invention, TB medium has been shown to work effectively. TB medium is routinely prepared by adding 100 milliliters of a sterile solution of 0.17 molar KH₂P0₄ and 0.72 molar K₂HP0₄ to a sterile solution containing 12 grams of Bacto-tryptone, 24 grams of Bacto- yeast extract, 4.0 milliliters of glycerol and water to a final volume of 900 milliliters.

The methods of the invention may be applied to any bacterial cells capable of enhanced protein expression. In a preferred embodiment the bacterial cells will be those which carry a mutation of the recA gene. However, in a more preferred embodiment, the bacterial cells used in combination with the methods of the invention designed to enhance expression of a desired recombinant protein will be cells of the industrial standard E. coli. Preferably, a mutation of the recA gene will be present in such E. coli cells. In a specific embodiment, the methods of the invention used in combination with E. coli cells carrying a mutation of the recA gene will ^"be practiced using a DH5α strain of E. coli. A variety of such strains will be known to those of skill in the art to possess certain advantages in specific instances. The use of such strains in combination with the methods of the invention are expressly encompassed here. Furthermore, the mutation of the recA (recA" for the purposes of the invention) may include any such mutation of the recA gene locus as long as a recA^" phenotype is achieved.

The surprising and unexpected results realized using the method of the invention to enhance expression of protein is obtained by culturing of the bacterial cells to at least the late stationary phase of cell growth. In certain embodiments, it will be found advantageous to extend the growth for at least 16 hours. In another specific embodiment, the growth period will be extended for at least 24 hours. It is assumed here that one of skill in the art will maximize the extended growth period in the protein expression enhancing medium in combination with the other teachings of the invention.

In one sense, the method of the invention may be ,viewed as a significant improvement over standard methods for expressing recombinant proteins. The prior art methods relied on culturing bacterial cells containing DNA segments encoding such a protein in a rich medium under conditions appropriate to express that protein. Using the prior art methods, maximal protein production not dependent on addition of inducers to the medium is typically attained prior to the late stationary phase of cell growth. This procedure results in a significant reduction in the cell mass recovered from the growth medium even though the yield of enzyme per gram of cells is increased. The surprising improvement realized by applying the method of the invention consists of culturing the bacterial cells in a protein expression enhancing medium and allowing the cells to grow to at least the late stationary phase of cell growth in order to allow the bacterial cells to express the protein.

The method of the invention may be successfully applied using TB medium. Using the methods of the invention, expression of the recombinant protein may be routinely realized to levels representing at least fifteen percent of the recombinant protein by wet weight of the total soluble protein from the bacterial cells.

Figure 1: Restriction Map of the BamHl to StuI Fragment Containing the Genes Encoding Putidaredoxin Reductase (PdR) and Putidaredoxin (Pd) . The region encoding PdR is indicated by the longer of the two heavy bars while the short heavy bar represents the region coding for Pd. The fragments which were subcloned for sequencing are represented by the light bars from BamHl to Sail. Sail to Nrul, and Nrul to StuI. The actual segments which were sequenced and the direction of sequencing are represented by the light bars with arrows.

Figure 2: Nucleotide and Deduced Amino Acid Sequence of Putidaredoxin Reductase and Putidaredoxin. As indicated in Fig. 1, PdR is the first protein sequence and Pd is the second. Other regions of the sequence are described in the text.

Figure 3: Subcloning Strategy for the Expression of Putidaredoxin Reductase. The relative position of the lacZ promoter is indicated by the heavy arrow on the circumference of the plasmid.

A surprisingly simple method of enhancing the recovery of cloned proteins is described in the present invention which includes inoculating bacterial cells which contain a recombinant protein encoded in DNA vector such as a plasmid into a medium designed for enhanced recovery of vector DNA (e.g., TB medium). The culture is allowed to grow for at least 16 hours or to at least the late stationary phase of typical logorithmic bacterial growth. Following the extended growth in TB medium, it is possible to obtain recoveries of the recombinant protein in excess of 20% of the total soluble protein of the cells.

The discovery of the enhancement in protein expression by the inventors is illustrative of the surprising nature of the invention. The inventors were interested in the expression of recombinant bacterial heme proteins, proteins which had presented significant problems in expression studies in the past. The prior art existing at the time of the discovery suggested many ways of enhancing expression of such proteins including, in the instance of certain bacterial cytochrome P450s, the use of barbiturate additives to the medium. The inventors had applied these techniques with only limited success and certainly anticipated difficulty in commercializing any such recombinant proteins due to the inability of recovering commercially feasible amounts of such proteins. In fact, as described more fully herein, use of the prior art standard media for expressing recombinant proteins, even when used with growth periods extended to the extent taught by the present invention, failed to provide substantially enhanced recovery of the recombinant proteins. It was well substantiated in the art that protein production substantially ceased after cells in culture reached stationary phase [Nolan, R.D. (1986), Wolf (1985), Kleiner et al. (1988)].

Other existing art taught the use of certain rich media, in particular TB medium, to enhance the recovery of DNA vectors. Even so, the art in this area suggested only that overnight growth be accomplished. Moreover, the art in this area did not state nor suggest use of this rich medium to enhance protein expression in the manner of the present invention. According to the prior art, one should maximize the amount of protein produced per gram of cells for the lowest cost. In most instances, the cost of using rich media was prohibitive when applying these considerations. In fact, rich media were typically only used where the protein which was desired was transcriptionally linked to a promoter corresponding to a gene which is turned on only in later growth stages (i.e., sporulation genes in Bacillus subtilis) . In typical prior art, where fermentations of E. Coli using rich medium were being carried out and where there is no apparent temporal regulation of gene expression, very late log phase cultures exhibited the general problems associated with cell death including contamination and overgrowth of contaminating microorganisms as well as increased proteolysis.

While attempting to enhance the recovery of DNA vectors containing certain recombinant bacterial heme proteins, the inventors inadvertently allowed the incubation period taught by the prior art for using TB medium to enhance DNA vector recovery to be greatly extended. Though there was no teaching in the prior art to lead them to do so, the inventors analyzed the resulting overgrown cell culture for the recombinant protein expecting poor recovery at best. Surprisingly, a significant increase in the recovery of the recombinant protein was observed. It was possible to recover 1 gram of homogenous, enzymatically active protein from 300 grams wet weight of cells.

In preferred aspects, then, the method of the invention will routinely provide recovery of the cloned protein of at least fifteen percent total soluble protein from the cultured cells. By manipulating the incubation time, the amount of aeration of the culture and the concentration of glycerol in the medium, one of skill in the art of fermentation may maximize the amount of the recovered protein for each individual case.

The invention has been found to have particular utility when applied to the cloning and expression of bacterial heme proteins such as bacterial cytochrome P450. In a preferred embodiment, the method of the invention was applied to enhance the recovery of cytochrome P450_BM-₃ derived from B. meqaterium. The invention methods allowed, for the first time, recovery of commercially significant quantities of this important monooxygenase in amounts exceeding 1 g purified protein/300 g wet weight of cells.

Additionally, the method of the present invention has been applied to recovery of cloned bacterial electron transfer protein. Bacterial redoxin reductase and bacterial redoxin are electron transfer proteins which are used to transfer electrons from an electron donor molecule such as NADH to the oxidizing cytochrome P450_cam. The method of the invention was applied to the bacterial redoxin reductase derived from P. putida. putidaredoxin reductase. In this application, the inventors additionally sought to separate the associated bacterial electron transfer protein, putidaredoxin also derived from P. putida. from the putidaredoxin reductase closely linked on the P. putida chromosome. Likewise, the heme protein of this electron transfer system, P450_cam, could be expressed in high levels in the preferred expression system of the invention. In this manner, and using the methods of the invention, these bacterial electron transfer proteins could be expressed at commercially feasible levels. The studies which gave the impetus for the current invention arose from a pressing need for commercially feasible quantities of the heme proteins briefly described above. The activation of molecular oxygen for incorporation into drugs, steroids, and carcinogens is the focus of intense interest since the discovery of the role of cytochrome P450 in this diverse set of reactions. The application of molecular cloning techniques to the studies of these enzymes has resulted in considerable detailed information including the sequence determination of more than 100 cytochrome P450s. The study of the electron transfer reactions required to provide the cytochrome with the necessary redox potential to affect these reactions has also been intensely pursued.

For instance, the oxidation of camphor by extracts from P. putida requires the participation of three protein components. These have been identified as an FAD-containing flavoprotein, putidaredoxin reductase, and an iron-sulfur protein, putidaredoxin, as well as the cytochrome P450. What makes these studies even more intriguing, is the fact that this electron transfer sequence is very similar to the one functional in the human adrenal cortex mitochondrial metabolism of steroids.

These studies into the mechanism and potential commercial utility of electron transfer from NADH to cytochrome P450 has been severely hindered by the limited availability of the enzymes involved. For instance, initial experiments on the characterization of the reaction catalyzed by P450_BM-.₃ had, at the time of the present invention, been accomplished using limited amounts of enzyme isolated from B. meqaterium. Even though this enzyme is induced by the inclusion of barbiturates in the growth medium, the amount of this enzyme which is present in the organism is still quite low (0.5% of the total soluble protein) and the earlier metabolite and mechanistic studies were difficult to evaluate because of the availability of only such limited amounts of the enzyme. Coupled with the cloning and sequencing of a recombinant form of the cytochrome P450_BM. ₃, the methods of the present invention may be used to alleviate the constraints due to limited quantities of enzyme.

Such an advance as represented by the methods of the invention are critical in several regards. As applied to the enhanced expression of the bacterial cytochrome P450_BM-₃, for instance, the significantly increased availability of the enzyme will provide numerous avenues for commercialization of the enzyme reaction. P450_BM_₃, being a soluble, catalytically self-sufficient enzyme with an extremely high turnover number in the w- hydroxylation of fatty acids should offer an excellent opportunity to enzymatically manipulate a wide variety of commercially important fatty acids. Moreover, study of these reactions with ample supply of the enzyme permitted by the invention, provide a potential means for extrapolating the bacterial enzyme results to the medically important human w-hydroxylases. For instance, studies using the enzymes in the multiple oxidation reactions of long chain fatty acids will likely illuminate mechanisms of carbon-carbon cleavage reactions such as cholesterol side chain cleavage.

Longstanding research and commercial application involving other electron transfer proteins has been hindered as well. For instance, limited availability of putidaredoxin reductase and putidaredoxin has blocked progress in commercialization of reactions involving these electron transfer proteins. It is well known among those of skill in the art that expression of proteins encoded by the lac operon is controlled by the lac repressor. There are strains of E. coli which contain the lacIQ gene which results in the super production of this gene product, the repressor.

These are the strains which are usually utilized for the "controlled" expression of recombinant proteins. In this instance, the chemical IPTG is added to the growth medium to bind the repressor and relieve the inhibition of the lac operon. The DH5a strain of E. coli. utilized in a preferred embodiment herein, only produces moderate amounts of the repressor and is referred to as "leaky." Thus, there is a small amount of synthesis of the proteins encoded by the lac operon even in the absence of IPTG. The over-production of proteins described in the present invention is suggested to result from a combination of factors: (1) significant increase in the copy number of the plasmid during late log phase; (2) decrease in synthesis of cellular protein and with it the lac repressor which accompanies late log phase of growth; (3) release of the inhibition of the lac operon because the ratio of repressor molecules to plasmid molecules is reduced; and, (4) enhanced protein synthesis from the lac operon which contains the genes for the desired proteins.

The examples to follow illustrate the use of the methods of the invention using E. coli DH5α cells, TB medium and extended growth to at least late stationary phase to enhance recovery of the bacterial heme protein, cytochrome P450 derived from B. meqaterium. and of the bacterial electron transfer proteins, putidaredoxin reductase and putidaredoxin derived from P. putida. These examples illustrate various preferred embodiments for carrying out the invention. The studies set forth below were conducted in part through the application of standard laboratory practices of the inventors as well as procedures developed by the inventors or otherwise found to work well in the practice of the invention. Various modifications, rearrangements of the steps, substitutions, and the like will be apparent to the skilled artisan in light of the examples to follow.

EXAMPLE I: The Effect of Incubation Time, Glycerol

Content and Aeration on Protein Recovery

The effect of various parameters on protein recovery was studied using the test protein putidaredoxin reductase (PdR) . All experiments were carried out in TB medium. By allowing incubation to proceed well past the point at which stationary phase has been reached in cell growth on TB medium (approximately 6 to 8 hours) , and well beyond the time at which the prior art teaches to harvest for the purposes of increased plasmid production, a surprising increase in the cloned protein expressed as a percentage of the total soluble protein of over 20% was realized (Table I) .

Table I: The Effect of Incubation Time on % Recombinant Protein in the Total Soluble Protein

TIME (HRS) %PdR¹

16 10

20 13

24 23

30 16

Additionally, by varying the glycerol content of the TB medium, protein recovery after extended incubation times was maximized (Table II) .

Putidaredoxin reductase Table II: The Effect of Glycerol Content of the Medium on % Recombinant Cloned Protein in Total Soluble Protein Harvested at 24 Hours Incubation.

GLYCEROL CONCENTRATION f% v/v) %PdR

0.2 10

0.5 20 0 0.7 15

1.0 5

Finally, aeration effects on the recovery of recombinant 5 proteins suggest that will less aeration lessen the recovery, even the larger volumes of cultures maintain substantial increases in the recovered protein (Table III) . In fact, with controlled aeration in a 14 or 150 1 fermentor, the enhanced recovery of protein described in 0 the invention was also observed with yields of 7-10 g of wet cell paste per liter of culture medium.

Table III: The Effect of Aeration on % Recombinant Protein in Total Soluble Protein When 5 Cells are Grown in a 250 ml Flask and Harvested at 24 Hours

MEDIA VOLUME fml) %PdR D 50 20 100 17 200 17

5 EXAMPLE II: Expression of the Cytochrome P450^,-₃ of Bacillus mecraterium in E. coli

P450_BM-₃ is contained on an 14 Kbp plasmid which contains significant coding and noncoding regions for 0 other proteins (Ruettinger et al. 1989). Initial efforts to obtain stable expression of P450_BM-₃ in either the JM103 or strain JM109 of E. coli provided with the original clone from A.J. Fulco failed due to "instability" of the plasmid (at levels of 1.0 mg P450/ g cells). The plasmid was moved to a commercial recA^" strain of E. coli (DH5α) which was shown to provide stable, high level expression of P-450_BM-₃ when used in conjunction with TB medium. The expression of P-450^-₃ under these conditions is not under the control of lacZ nor is it induced by barbiturates and yet realization of substantial increases in protein recovery may be had.

Materials: Unlabeled fatty acids were obtained from Nu Chek Prep, Inc., P.O. Box 295, Elysian, MN 56028,

U.S.A.). DE-52 anion exchange resin was purchased from Whatman BioSystems Ltd, Maidstone, Kent, England. All other reagents used were of purest grades available. The E. coli strain DH5α (F-endAl. hsdR17(r_k ^",m_k ⁺) , supE44. thi-1. recAl. qyrA96. relAl, A (arqF-laczya)U169. ø80dlacZAM15) was obtained from Bethesda Research Laboratories, Life Technologies, Inc., P.O. Box 6009, Gaithersberg, MD 20877, U.S.A. (BRL) as competent cells.

Preparation of P450_FH_.: The E. coli clone (JM109,

BM3-2A) containing the 9.2 kbp plasmid encoding P450_BM.₃ was obtained from Dr. A.J. Fulco, Department of Biological Chemistry, University of California, Los Angeles, California. The plasmid was isolated by conventional techniques (Birnboim and Doly 1979) and used to transform competent cells of E. coli strain DH5α. The transformed cells were isolated and the plasmid DNA of selected clones was examined by restriction analysis to demonstrate that the original plasmid was unchanged (Sambrook et al. 1989) . Cultures of cells were maintained in 2xYT media containing 50 μg per mL ampicillin. For longer term storage of cells, the media was made 7.5% in glycerol and the cells were stored at - 70^*C. Several different media were tested for their effect on the expression of P450_BM-₃ as shown in Table IV. After overnight growth at 36'C on the selected media, the cells were harvested by centrifugation, washed with 20 mM MOPS buffer (morpholinopropane sulfonate) , pH 7.4, containing 20 mM KCl, 2 mM dithiothreitol (Buffer A) and resuspended in this buffer. The P450_BM_₃ content of the whole cells was determined by difference absorbance spectrophotometry using 91 mM^"1cm^"1 as the molar absorptivity (O'Keeffe et al. 1978) . As shown in Table IV, P450_BM-₃ is expressed in normal growth media without the addition of IPTG (isopropyl-ø-D-thiogalactoside) to relieve the repression of the lacZ gene promoter (Wen and Fulco 1987) . The level of expression of P450^.₃ which was obtained in DH5α with either LB or 2xYT media was 2-4 fold higher than that reported by Narhi et al. (1988) and repeated in our laboratory for similar growth conditions in E. coli strain JM109. Extension of the time of incubation of the cells to late stationary phase in TB media resulted in a remarkable increase in the level of expression of P450_EM-₃ with this enzyme representing approximately 20% of the soluble protein of these cells.

Table IV

"Sambrook et al. 1989. Tartof and Hobbs 1987.

°Cells when grown overnight (14 hrs.) ^dCells when grown for 24 hours For purification of P450_BM.₃, the cells were grown for 24 hrs in TB media in a 10 L or 150L fer enter. The cells were harvested by centrif gation, washed and resuspended in Buffer A. The cell suspension, containing a ratio of 1 g of cell paste to 4 iL of buffer, was subjected to four freeze-thaw cycles followed by lysozyme treatment (0.5 mg per mL) at 4'C for 1 hr, to lyse the cells. DNase A (1 g per mL) and magnesium chloride (8mM) were added to the suspension to hydrolyze the DNA and decrease the viscosity of the solution. The suspension was incubated at 4'C for an additional hour. The lysate was centrifuged at 20,000 RPM for 1 hour and the supernatant solution was subjected to ammonium sulfate precipitation. The fraction which precipitated between 40-80% of saturated ammonium sulfate contained P450_BM-₃. The precipitate was dialyzed, diluted to a protein concentration of 60 mg per mL, and loaded on a DE-52 ion exchange column which had been equilibrated with 20 mM MOPS and 2mM DTT (dithiothreitol), pH 7.4 buffer (bed volume of 1 L) . The column was washed with 3 L of the equilibration buffer and the enzyme was eluted with a linear gradient of potassium chloride (0.1 to 0.5 M in equilibration buffer) . Most of the extraneous protein was removed by the initial wash and the P450_BM_₃ was eluted by the gradient as a single peak. Fractions containing at least 15 nmole of P450 heme per mL were pooled and concentrated by ultrafiltration. The protein was further purified by gel exclusion chromatography on a BioGel A- 1.5M column (1 L bed volume) . A representative sample of the purification procedure, starting with lOOg wet weight of cells, is summarized in Table V.

In agreement with the previously published data, the molecular weight of P450_BM.₃ was found to be approximately 119,000 (Nahri and Fulco 1986). SDS polyacrylamide gel electrophoresis of up to 8 μg of the protein per lane on an analytical gel, gave a single band by silver stain techniques (data not shown) . The content of P450 per mg of protein was estimated by three different techniques as shown in Table VI. Pure P450_BM_₃ should contain 8.3 nmoles of P450 heme per mg of protein based on a molecular weight of 119,000. Because of the inherent unreliability of protein estimation techniques and the fact that the ratio of heme to FAD to FMN is l (Nahri et al. 1988) , we believe that these preparations contain essentially all of the protein as holoenzyme.

Oxygen Uptake Determinations: Oxygen consumption was measured with a Clark type oxygen electrode immersed in a sealed reaction chamber containing 1.6 mL of 50 mM MOPS, pH 7.4, and 0.4 μM P450^-₃. The desired concentration of fatty acid in 50 mM potassium carbonate was added to the chamber and the solution preincubated at 25'C for about 5 min before the addition of NADPH. The oxygen concentration in the reaction solution was determined and the recorder calibrated using beef heart electron transport particles and NADH as has been previously described (Estabrook 1967) .

Table V

Purification of P450_EM-₃

Total P450 Total P450 Percent

Step Protein⁸ nmole/mg fumoles) Recovery

Cell Lysate ND ND 11

40-80%

Ammonium Sulfate 14.8 0.55 8.2 74

DE52 Chromotography 1.8 3.3 5.8 52

Bio-Gel A-1.5M 0.70 7.9 5.5 50

^aThe protein was estimated by the Warburg and Christian (1941) method for the impure fractions while the amount of protein in the pure P450_BM-₃ was estimated by the Lowry method (Lowry et al, 1951) .

Table VI

Protein Estimation of P450_BH-₃ nmoles P450 per mg protein

5.0 6.3 7.9

Putidaredoxin of Pseudomonas putida in E. coli,

Materials: The plasmids pIBI24 and pIBI25 which are derived from pEMBL plasmids, were obtained from International Biotechnologies, Inc. Restriction enzymes and bacteriophage M13mpl8 and M13mpl9 were obtained from BRL. P. putida (ATCC17453) was obtained from the American Type Culture Collection. The cell line was stored in media containing 7.5% glycerol at -80^*C. The nucleotide sequencing kit Sequenase Ver. 1.0 was obtained from U.S. Biochemicals, Corp., P.O. Box 22400, Cleveland, OH 44122, U.S.A.. Ampicillin was obtained from Sigma Chemical Co., P.O. Box 14508, St. Louis, MO 63178, U.S.A.. All other reagents and chemicals were of the highest purity available. Mutagenic oligonucleotides and DNA sequence primers were synthesized on an Applied Biosystems 380A oligonucleotide synthesizer and purified by Sep-Pak C18 (Waters Associates, Milford, MA 01757, U.S.A.) column chromatography.

Bacterial Growth: Stock cultures of E. coli strain DH5α which harbored the appropriate plasmids were grown in 2xYT media containing 50 μg per mL ampicillin. For long term storage of the cell lines, the cell suspension was made 7.5% in glycerol, frozen in a dry ice/acetone bath, and stored at -80^*C until used.

For expression of either putidaredoxin reductase or putidaredoxin, the desired cell line was grown for 16 to 24 hours in TB media Tartof and Hobbs (1987) containing 50 μg per mL ampicillin. The cells were harvested by centrifugation at 8,000 rpm in a Beck an J21 centrifuge (Beckman Instruments, Inc., Mail Station E-06-A, 2500 Harbor Blvd., Box 3100, Fullerton, CA 92634-9989, U.S.A.). The cells were broken by gentle sonication in a Branson Sonifier, Eagle Road, Danbury, CO 06810, U.S.A. for 60 sec and the cell debris and broken cells were removed by centrifugation. The amount of enzyme present in the extract was determined as described below.

Preparation of Whole Cell DNA from P. putida: The plasmids encoding catabolic enzymes of pseudomonads are typically rather large and difficult to prepare (Hansen and Olsen 1978) . For the present studies, whole cell DNA was prepared by a variation of a published procedure (Lorence 1984) . Cells of P. putida (1 L) which had been grown on d-camphor to stationary phase, were harvested by centrifugation. The cells were washed with 275 mL of 10 mM sodium phosphate buffer, pH 7.0 to remove media and excess camphor and its metabolites. The cells were resuspended in 100 mL of 10 mM Tris buffer, pH 8.0, containing 1.0 mM EDTA, and 20% sucrose and stored on ice. The cell suspension was made (TE buffer) and recentrifuged. The cells were finally resuspended in 100 L of lysis buffer containing 50 mM Trischloride, pH8.0, 50mM EDTA, made 2 mg per mL in lysozyme and the incubation on ice continued for 30 min.

The cells were lysed by the addition of SDS (sodium dodecyl sulfate) to a final concentration of 4% and the mixture was heated to 70^*C and mixed gently for 30 min. Proteinase K was added to a final concentration of 0.1 mg per mL and the heating continued for an additional 60 min. Potassium acetate was added to a final concentration of 0.5M and the incubation at 70'C continued for 15 min. After cooling to room temperature, the cell debris was removed by centrifugation at 17,000 rpm for 20 min in a JA20 rotor in a Beckman J21 centrifuge (Beckman Instruments, Inc., Mail Station E-06- A, 2500 Harbor Blvd., Box 3100, Fullerton, CA 92634-9989, U.S.A.) .

The DNA was precipitated from the supernatant solution by the addition of PEG8000 to a final concentration of 10%. The solutions were mixed by gentle inversion and stored at 4'C overnight. The precipitate was collected by centrifugation at 12,000 rpm in the JA20 rotor for 15 min at 4'C, rinsed with cold 95% ethanol, and resuspended in 8 ml of TE buffer. The suspension was made 0.1 mg per mL in RNase A and heated at 60'C for 30 min. The sample was cooled and extracted with: 1) 1 volume of phenol; 2) 1 volume of phenol, 1/2 volume of chloroform:isoamyl alcohol (24:1); and, 3) 1 volume of chloroform:isoamyl alcohol. The DNA was precipitated with 1/2 volume of 7.5 M ammonium acetate and 2 volumes of ethanol. The precipitate was collected by centrifugation and dissolved in 1 volume of TE buffer and the ethanol-ammonium acetate precipitation was repeated. The precipitate was washed with a small volume of 70% ethanol and dried in a vacuum desiccator. The DNA was resuspended in 2 mL of TE buffer.

Library Construction and Screening: Whole cell DNA was cleaved with BamHl and StuI. ligated into similarly cleaved pIBI25, and transformed into DH5α competent cells. Approximately 20,000 recombinant cells were plated onto 2xYT plates containing ampicillin (50 μg per mL) , transferred to 85 mm nitrocellulose filter disks (Schliecher and Schuell, BA85) (Gelman Sciences, Inc., Ann Arbor, MI 48106, U.S.A.), and screened by hybridization using a polynucleotide kinase-labeled oligonucleotide complementary to a portion of the N-terminal sequence of putidaredoxin reductase between the BamHl and Hindlll sites of the P450_CAM-containing fragment isolated by Unger et al. (1986) . as probe. Filter disks were washed using empirically determined conditions to remove non- specifically bound probe. Colonies which had hybridized to the probe were visualized by overnight exposure of X- ray film at -80'C in the presence of a DuPont Chronex Lightning Plus intensifying screen (Dupont) . Selected colonies were removed from plates with sterile toothpicks and grown overnight in 2xYT media containing ampicillin (50 μg per mL) . The identity of the clone was confirmed by restriction analysis or by Southern hybridization of the isolated plasmid DNA with the probe.

Plasmid Constructions: Plasmid DNA was isolated for restriction enzyme analysis by the alkaline-lysis method of Birnboim and Doly 1979. Selected restriction fragments were fractionated on a 1.5% low melting point agarose gel, the desired fragment was excised, melted at 65'C in a final volume of 0.4 mL of TE containing 100 mM NaCl, and extracted two times with TE-saturated phenol. The extracted fragment was precipitated twice with ethanol, and analyzed by agarose gel electrophoresis. Purified fragments were ligated into similarly cleaved pIBI24 or pIBI25, and transformed into DH5α competent cells according to procedures recommended by the manufacturer. Recombinant clones were identified as white colonies in the presence of the chromogenic substrate X-gal, and screened by restriction enzyme analysis. In most cases, the orientation of the insert in the ligation reaction was controlled by the nonidentical cohesive ends of the insert which matched the vector. The procedure used for the ligation of target DNA into either the plasmid vector or into the replicative form of the M13mpl8 or M13mpl9 bacteriophage was identical and has been described previously (Sambrook et al. 1989) .

The Southern hybridization procedure used to identify DNA fragments containing desired sequences was performed essentially as described (Sambrook et al. 1989) . The washing conditions to remove excess radiolabeled oligonucleotide probe were determined empirically. The preparation of ³²P-labeled oligonucleotide probes has been described Sambrook et al. 1989.

Nucleotide Sequence Determination: Selected restriction fragments were purified from low melting point agarose, ligated into similarly cleaved M13mpl8 and M13mpl9, and transformed into DH5αF' competent cells. Recombinant clones were identified as clear plaques in the presence of X-gal, and screened by restriction enzyme analysis of replicative form DNA isolated by the alkaline-lysis procedure. Single-strand phage DNA was isolated from infected cell cultures by polyethylene glycol precipitation, followed by SDS-Proteinase K digestion, phenol extraction and ethanol precipitation. The nucleotide sequence of the purified ssDNA was determined by the dideoxynucleotide chain termination method (Sanger et al. 1977) using a USBiochemicals Sequenase kit (U.S. Biochemical Corp., P.O. Box 22400, Cleveland, OH 44122, U.S.A.) containing a modified T7 DNA polymerase (Tabor and Richardson 1987) version 1.0. In those instances where the sequence determined for both the coding and non-coding strand were not in agreement, the determinations were repeated with deaza-dGTP to resolve band compression artifacts (Barnes et al. 1983; Gough and Murray 1983; Mizusawa et al. 1986). In each instance, the discrepancies were resolved by this procedure.

Oligonucleotoide Directed Mutagenesis:

Oligonucleotide directed mutagenesis was performed using the two primer method of Zoller and Smith (1984) . In each case, the single base change was inserted in the middle of a 21mer which would hybridize with the single stranded DNA at the desired location. The newly synthesized double stranded DNA was used to transform DH5αF* competent cells. Recombinant plaques were transferred and fixed to the filters as described (Sambrook et al. 1989) and mutants were identified by hybridization to the polynucleotide kinase-labeled mutagenic oligonucleotide. Rather than using the standard wash procedure, tetramethylammonium chloride was used to accentuate the difference in melting temperature between the probe and the wild-type and changed sequences (Wood et al. 1985) . In each case, the bacteriophage was purified until all of the plaques on a given plate would hybridize with the probe. To ascertain that only the desired base change occurred, each mutant was completely sequenced following plaque purification.

Putidaredoxin Reductase and Putidaredoxin Determinations: The standard assay for the amount of putidaredoxin reductase and putidaredoxin takes advantage of the ability of the enzymes to catalyze the reduction of cytochrome c. This reaction is dependent on the presence of the reducing agent NADH, and both enzyme components (Roome 1983) . In the typical assay either putidaredoxin reductase or putidaredoxin was in excess while the other component was limiting. Under these conditions, the rate of reduction of cytochrome c was linearly dependent on the concentration of the limiting component. Although there is a slight background of

NADH-dependent reduction of cytochrome c in whole cell extracts from E. coli. the increase in rate upon the addition of the limiting component was at least four-fold greater than the background.

The standard assay for putidaredoxin reductase contained the following components in 20 mM MOPS buffer, pH 7.4: 0.1 mM NADH, 10 μM cytochrome c and 5.5 μM putidaredoxin. An appropriate dilution of the cell-free extract was added to the reaction mixture prior to the addition of the putidaredoxin. The rate of reduction of cytochrome c was compared to a standard curve to determine the amount of putidaredoxin reductase present in the cell extract.

The standard assay for putidaredoxin contained the following components in 20 mM MOPS buffer, pH 7.4:0.1 mM NADH, 10 μM cytochrome c, and 1 nM putidaredoxin reductase. In this instance, the cell extract was added prior to the addition of putidaredoxin reductase. The concentration of putidaredoxin in the whole cells was also determined by EPR spectroscopy and the signal compared to a known standard. The concentration of cytochrome P450_cam in cell free extracts was determined by standard procedures (O'Keeffe et al. 1978).

Putidaredoxin Reductase and Putidaredoxin Cloning and Sequence Determination: The cloning of PdR² and Pd was aided by the publication in 1986 of the nucleotide sequence of P450_cβπ which included the N-terminal 153 nucleotides of putidaredoxin reductase (Unger et al. 1986) . Publication of the preliminary restriction map for the 4.4 and 2.6 kbp Hindlll fragments which contained the regulatory region, an alcohol dehydrogenase, P450_cam and both of these proteins Unger et al. (1986) indicated their general position within these fragments. To assist in clone selection, an oligonucleotide was synthesized which was complimentary to a portion of the N-terminal sequence of putidaredoxin reductase between the BamHl and Hindlll sites. The whole cell DNA from £. putida was cleaved sequentially with BamHl and StuI and the digested DNA was ligated into BamHI-Sma I cleaved pIBI25. The resulting reaction mixture was used to transform E. coli strain DH5α. The oligonucleotide probe was used to select clones which contained the 2.2 kbp BamHl-StuI fragment. The preliminary restriction map published by Unger et al. 1986 indicated that the coding sequence for putidaredoxin should span the StuI restriction site.

The 2.2 kbp BamHl-StuI fragment was subcloned into both M13mpl8 and M13mpl9 for sequence determination. The BamHI-Sall. Sall-Nrul, and Nrul-StuI (the EcoR I site from the polylinker region of the plasmid vector was actually used to clone the 3' end of this piece of DNA) fragments shown in Fig. 1 were cloned into the appropriate bacteriophage. Either the universal primer or synthetic oligonucleotides were used to prime the synthesis of the complimentary strand of DNA for the sequence determination. As indicated in Fig. 1, essentially all of the DNA was sequenced on both strands. The restriction map shown in this figure was deduced from the determined nucleotide sequence.

The nucleotide sequence of the BamHl-StuI fragment is shown in Fig. 2. The deduced amino acid sequence for both putidaredoxin reductase and putidaredoxin are also shown in this figure. There are several points to be made about these sequences: (1) the amino acid sequence of the N-terminus of putidaredoxin reductase is in perfect agreement with that reported (Unger et al. 1986a; Unger et al 1986b;Romeo et al. 1987) ; (2) the deduced amino acid sequence of putidaredoxin exactly matches the published sequence (Tsai et al. 1971; Tanaka et al. 1974) except for one amino acid; (3) the reported sequence had a glutamine residue at amino acid position 15 while the nucleotide sequence indicates that this amino acid is really a glutamic acid residue; and, (4) There are no open reading frames in the region 3' to putidaredoxin which might code for a protein of greater than 60 amino acids.

As would be expected for putidaredoxin, the differences between the deduced and reported sequences are minor and explainable on the basis of the removal of the initiating methionine residue in post-translational processing of the protein. There is one additional glutamic acid residue and one fewer glutamine residue. The calculated molecular weight of this protein including the two iron and two acid-labile sulfur atoms of the active site is 11,726 daltons. The deduced and reported composition of putidaredoxin reductase are similar enough to lead one to the conclusion that this is probably the same protein. The calculated molecular weight of 46,215 agrees well with the reported value of 48,500 (Roome et al. 1983). Expression of Putidaredoxin Reductase: To ascertain the validity of the identification of the coding region as putidaredoxin reductase, this fragment was subcloned into pIBI24 in the correct orientation for transcription directed by the lacZ promoter. To obtain a clone of E. coli which would express only putidaredoxin reductase, the BamHl-StuI/EcoRI fragment was digested with both BamHl and Mlul and the resulting 1.5 kbp fragment of DNA was purified by electrophoresis in low melting point agarose. The fragment was recovered and ligated into the plasmid pIBI24 as shown in Fig. 3 which resulted in the insertion of this fragment in the appropriate orientation with respect to the lacZ promoter. This particular piece of DNA was chosen because it created a termination codon in the 3-galactosidase reading frame which would prevent the formation of a fusion protein between the β- galactosidase and putidaredoxin reductase. The putative Shine-Dalgarno site is located 3• to this termination codon (see Fig. 2) . The Mlul restriction site was chosen for the 3' termination of this DNA fragment because it was within the coding region of putidaredoxin. The presence of putidaredoxin reductase and putidaredoxin in the same E. coli cell line might result in cell death due to leakage of electrons from NADH through the reductase to putidaredoxin which has a greater sensitivity to oxidation by molecular oxygen than do the other components of this electron transfer system. In fact, when both putidaredoxin reductase and putidaredoxin were subcloned together in the correct orientation for transcription, no colonies were found (data not shown) .

As can be seen in Table VII, the E. coli clone, which was isolated following transformation with the plasmid construct, would express active putidaredoxin reductase. This level of putidaredoxin reductase is similar to that observed in wild type P. putida (Roome et al. 1983) . Effect of the Start Codon on the Level of Expression of Putidaredoxin Reductase: The start codon for putidaredoxin reductase is the rare initiation codon GTG (Unger et al. 1986a; Unger et al. 1986b). This codon has been presumed to be important in the post-transcriptional regulation of protein abundance. To test this hypothesis, the G at position 1 was changed to an A by site-directed mutagenesis in the single-stranded bacteriophage M13mpl8 which contained the BamHI-Sall fragment. This mutant clone was isolated and completely sequenced to ascertain whether there were any changes other than the desired one. The BamHl-Xhol fragment containing the desired nucleotide change was purified, from low melting point agarose, from the replicative form of the bacteriophage, and ligated into the plasmid described above for putidaredoxin reductase expression as shown in Fig. 3. Previously, the wild type BamHl to Xhol fragment had been removed from this plasmid by restriction digestion and low melting point agarose gel electrophoresis. The plasmid was transformed into E. coli. DH5α and clones which were able to grow in the ampicillin- containing media were selected. Samples of the cells were grown up and the sequence of the double stranded DNA was determined with the expected change in nucleotide sequence present (data not shown) . These cells were grown in TB media and their putidaredoxin reductase content was determined as shown in Table VII. The average of three determinations gave a content of 7.4 mg of putidaredoxin reductase per gram wet weight of cells which is approximately an 18 fold increase over the level in JC. coli cells containing the wild type gene.

Expression of Putidaredoxin: To obtain the expression of putidaredoxin-independent of putidaredoxin reductase, the 856 bp Narl-S al fragment of the original BamHl-StuI clone was subcloned into the AccI and Smal sites in the polylinker region of M13mpl8. ^"A single base change of a G to a T resulted in the formation of a Hindlll site at about 1160 bp in the putidaredoxin reductase coding sequence (see Fig. 2) . The mutagenic oligonucleotide was used as a probe for the clones with the anticipated base change as described in supra Wood et al. 1985. The 580 bp fragment of DNA from the new Hindlll site to the Smal site was purified from the replicative form of M13mpl8 by low melting point agarose gel electrophoresis and ligated into pIBI25 which had been cleaved with the same restriction enzymes. The level of expression of putidaredoxin in this cell line and control cells is shown in Table VII.

Table VII

Expression of Putidaredoxin Reductase and Putidaredoxin in E. coli"

Clone PdR Pd P450

PdR (GTG) 0.4 0 0

PdR (ATG) 7.4 0 N.D.

Pd 4.8

N.D. means not determined. ^aThe amount of each of these proteins is expressed as the number of mg of protein per g wet weight of cells.

REFERENCES CITED

The following references are specifically incorporated herein by reference to the extent that they disclose or teach methods and or compositions useful in the practice of the invention.

1. Barnes, W.M. , Bevan, M. , and Son, P.H. (1983) Meth . Enzym . 101, 98-122. 2. Birnboim, H.C. and Doly, J. (1979) Nucleic Acids Res . 7, 1513-1523.

3. Croteau, R. , and Kalattukudy, P.E. (1975a) Arch . Biochem . Biophys . 170, 61-72.

4. Croteau, R. , and Kolattukudy, P.E. (1975b) Arch . biochem . Biophys . 170, 73-81.

5. Geren, L., Tuls, J., O'Brien, P., Millett, F., and Peterson, J.A. (1986) J. Biol . Chem . 261, 15491- 15495.

6. Gough, J.A. and Murray, N.E. (1983) J. Mol . Biol . 166, 1-19.

7. Hanahan, D. (1984) in DNA Cloning, A Practical Approach, vol. I, D.M. Glover, ed. IRI press, Oxford, p.109.

8. Hansen, J.B. and Olsen, R.H. (1978) J. Bacteriol . 135, 227-238.

9. Hare, R.S. and Fulco, A.J. (1975) Biochem . Biophys . Res . Commun . 65, 665-672.

10. Heinz, E. , Tullock, A.P., and Spencer, J.F.T. (1969) J. Biol . Chem. 244, 882-888. 11. Ho, P.P. and Fulco, A.J. (1976) Biochim . Biophys . Acta 431, 249-256.

12. Katagiri, M. , Ganguli, B.N. , and Gunsalus, I.e. (1968) J. Biol . Chem . 243, 3543-3546.

13. Klein et al. (1988) J. Gen . Microbiol . 134, 1779- 1784.

14. Kolattukudy, P.E. (1969) Plant Physiol . 44, 315-317.

15. Kupfer, D. (1980) Pharmac . Ther. 11, 469-496. 16. Lorence, M.C. (1984) Ph.D. Dissertation, The University of Texas at Dallas, Dallas, Texas.

17. Miura, U. and Fulco, A.J. (1974) J. Biol . Chem . 249, 1880-1888. 18. Miura, U. and Fulco, A.J. (1975) Biochim . Biophys . Acta 388, 305-317.

19. Mizusawa, S., Nishimura, S., and Seela, F. (1986) Nucleic Acids Res . 14, 1319-1324.

20. Narhi, L.O., Wen, L-P., and Fulco, A.J. (1986) J. Biol . Chem . 261, 7160-7169.

21. Narhi, L.O., Wen, L-P., and Fulco, A.J. (1988) Mol . Cell . Biochem . 79, 63-71.

22. Nolan (1986) in Overproduction of Microbiol. Metabolites: Strain Improvement and Process Control Strategies, (Vanek and Hostalek, eds.) pp. 215-231,

Butterworths, Boston.

23. O'Keeffe, D.H., Ebel, R.C., and Peterson, J.A. (1978) Meth . Enzym . 51, 151-157.

24. Peterson, J.A. , Basu, D., and Coon, M.J. (1966) J. Biol . Chem . 5162-5164.

25. Romeo, C, Moriwaki, N. , Yasunobu, K.T., Gunsalus, I.e., and Koga, H. (1987) J. Prot . Chem . 6, 253-261.

26. Roome, P.W. , Philley, J. , and Peterson, J.A. (1983) J. Biol . Chem . 258, 2593-2598. 27. Roome, P.W. , Jr., and Peterson, J.A. (1988) Arch. Biochem . Biophys . 266, 32-40.

28. Ruettinger, T. , Wen, L-P., and Fulco, A.J. (1989) J. Biol . Chem . 264, 10987-10995.

29. Sambrook, J. , Fritsch, E.F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, Second

Edition, pp. 1.98-1.99, Cold Spring Harbor Laboratory Press, Cold Spring Harbor.

30. Sanger, F. , Miklen, S. and Coulsen, A.R. (1977) Proc . Natl . Acad . Sci . USA 74, 5463-5467. 31. Tabor, S. and Richardson, C.C. (1978) Proc . Natl . Acad . Sci . USA 84, 4767-4771. 32. Tanaka, M. , Haniu, M. , Yasunobu, K.T., Dus, K. , and Gunsalus, I.C. (1974) J. Biol . Chem . 249. 3689-3701.

33. Tartof, K.D. and Jobbs, C.A. (1987) Focus 9:2, 12.

34. Tsai, R.L. , Dus, K. , and Gunsalus, I.C. (1971) Biochem . Biophys . Res . Commun . 45, 1300-1306.

35. Tulloch, A.P., Spencer, J.F.T., and Gorin, P.A.J. (1962) Can . J. Chem . 40, 1326-1337.

36. Unger, B.P., Gunsalus, I.C. and Sugar, S.G. (1986a) J. Biol . Chem . 261, 1158-1163. 37. Unger, B.P., Sugar, S.G., and Gunsalus, I.C. (1986b) in The Bacteria: A Treatise On Structure and Function, vol. X, The Biology of Pseudomonas (Sokatch, J.R. , ed.) pp. 557-589, Academic Press, Inc. San Diego. 38. Wen, L-P. and Fulco, A.J. (1987) J. Biol . Chem . 262, 6676-6682 39. Wolf (1985) in the Molecular Biology of Bacterial

Growth (Schaechter, et al., eds.) pp. 202-211, Jones and Bartlett Publishers, Inc., Boston. 40. Wood, W.I., Gitschier, J. , Lasky, L.A. and Lawn, R.M. (1985) Proc . Natl . Acad . Sci . USA 82, 1585- 1588. 41. Zoller, M.J. and Smith, M. (1984) UNA 3, 479-488.

Modifications and alterations will become apparent to one of skill in the art in light of the foregoing disclosure. It is intended by the inventors that all such modifications and changes be included within the scope of the subject matter which the inventors of the present patent regard as their invention, and which are defined by the following claims.

Claims

CLAIMS :

1. A method of enhancing the expression of a recombinant protein, comprising:

obtaining bacterial cells which include a recombinant DNA vector encoding the recombinant protein;

inoculating a protein expression enhancing medium with the bacterial cells;

culturing the cells in the protein expression enhancing medium to at least late stationary phase of cell growth to allow the cells to express the recombinant protein; and

recovering the recombinant protein so expressed.

2. The method of claim 1 where the expressed recombinant protein comprises at least fifteen percent, by wet weight, of the total soluble protein from the bacterial cells.

3. The method of claim 1 where the recombinant protein comprises a bacterial heme protein.

4. The method of claim 3 where the recombinant protein comprises a bacterial cytochrome P450.

5. The method of claim 4 where the bacterial cytochrome P450 comprises a cytochrome P450_BM-₃ derived from Bacillus megaterium.

6. The method of claim 1 where the recombinant protein comprises a bacterial electron transfer protein.

7. The method of claim 6 where the bacterial electron transfer protein comprises a bacterial redoxin reductase.

8. The method of claim 6 where the bacterial electron transfer protein comprises a bacterial redoxin.

9. The method of claim 7 where the bacterial redoxin reductase comprises a putidaredoxin reductase derived from Pseudomonas putida.

10. The method of claim 8 where the bacterial redoxin comprises a putidaredoxin derived from Pseudomonas putida.

11. The method of claim 1 where the protein expression enhancing medium remains rich enough in energy source molecules, protein synthesis substrates and nutrients required for bacterial growth in order to allow not only the achievment of logrythmic growth of bacterial cells but also to allow amplification of the copy number of vector molecules encoding the recombinant protein and to allow protein synthesis significantly unrestricted by lack of energy source molecules, protein synthesis substrates and nutrients required by the bacterial cells once the stationary phase of growth has been achieved.

12. The method of claim 11 where the medium comprises a medium at least three times richer than Luria-Bertani medium.

13. The method of claim 11 where the protein expression enhancing medium further comprises buffering agents.

14. The method of claim 12 where the buffering agents comprise KH₂PO_A and K₂HP0₄.

15. The method of claim 12 where the protein expression enhancing medium further comprises glycerol.

16. The method of claim 14 where the medium comprises at least 0.1% and as much as about 2.0% volume of glycerol per volume of medium.

17. The method of claim 15 where the glycerol is added at a concentration of at least 0.2% volume of glycerol per volume of medium.

18. The method of claim 15 where the glycerol is added at a concentration of at least 0.5% volume of glycerol per volume of medium.

19. The method of claim 1 where the protein expression enhancing medium is TB.

20. The method of claim 1 where the bacterial cells further comprise Escherichia coli cells.

21. The method of claim l where the bacterial cells comprise recA^~ cells.

22. The method of claim 19 where the recA bacterial cells comprise recA^~ Escherichia coli cells.

23. The method of claim 20 where the Escherichia coli cells which carry a deletion of the recA gene comprise a DH5-α strain of Escherichia coli.

24. The method of claim 1 where the bacterial cells are cultured for at least 16 hours.

25. The method of claim 22 where the bacterial cells are cultured for at least 24 hours.

26. In a method for expressing recombinant proteins which method includes culturing bacterial cells containing DNA segments encoding such proteins in a medium under conditions appropriate to express that protein, where the improvement comprises culturing the bacterial cells in a protein expression enhancing medium and allowing the cells to grow to at least the late stationary phase of cell growth in order to allow the bacterial cells to express the protein.

27. The method of claim 24 where the protein expression enhancing medium is TB.

28. The method of claim 24 where the expressed recombinant protein comprises at least fifteen percent, by wet weight, of the total soluble protein from the bacterial cells.