WO1998008954A2 - Production of heme and recombinant hemoproteins - Google Patents

Abstract

The present invention relates to methods of enhancing the production of heme-containing proteins by increasing the accumulation of heme in heterologous host cells. The methods are accomplished by exposing the host cells to increased amounts of endogenous or exogenous δ-aminolevulinic acid (ALA) resulting in an increase in heme production. Enhancing the endogenous production of ALA can be accomplished by adding multiple copies of the hemA gene into the host cell. Methods of producing enhanced amounts of heme are also provided.

Description

PRODUCTION OF HEME AND RECOMBINANT HEMOPROTEINS

Field of the Invention

This invention generally relates to the production of heme and heme-containing proteins. More specifically, the invention relates to methods of enhancing heme production in host cells capable of expressing heterologous, heme-containing proteins.

Background of the Invention

Heme is an iron-containing porphyrin that serves as a prosthetic group in proteins such as hemoglobin, myoglobin and the cytochromes. The biochemical pathway for heme biosynthesis is well known. Except for the initial steps in the formation of δ-aminolevulinic acid (ALA), the pathway is fairly well conserved throughout plants, animals and bacteria.

In E.coli and related bacteria, heme_b (which is also known as protoheme and ferrous protoporphyrin IX) is essential for respiration and for detoxification of reactive species of oxygen. Heme_b serves as an essential cofactor for b-type cytochromes, catalase and peroxidase.

Biosynthesis of heme_b occurs via a complex, branched pathway that involves up to twelve gene products (Fig. 1). In E. coli, ALA, the committed precursor in the heme_b pathway, is formed from the 5-carbon skeleton of glutamate via the C5 pathway. Production of ALA in E. coli occurs in three steps: (1) ligation of tRNA (GTR) synthetase, (2) reduction of the resulting GTR to glutamate 1-semialdehyde (GSA) by GTR reductase, and (3) transamination of GSA to ALA by GSA aminotransferase. Seven additional reactions are required to convert ALA to heme_b including assembly of eight ALA molecules into a cyclic tetrapyrrole, modification of the side chains, and incorporation of reduced iron into the molecule (Warren and Scott, TIBS 15:486-491 (1990)).

In photosynthetic plants, ALA is required for production of hemes and chlorophylls. In vitro studies suggest that the plant C5 pathway is regulated at the biochemical level by heme_b and that heme_b specifically inhibits conversion of GTR to GSA by GTR reductase (Castelfranco et al., Ann. Rev. Plant Phvsiol.. 24:241-278 (1983); Huang & Wang, J. Biol. Chem.. 261 : 13451-13455 (1986)).

In non-recombinant E. coli cells, accumulation of large pools of heme_b pathway intermediates or free heme_b is deleterious to the cells. For example, E. coli cells with mutations in the gene encoding hemH (vis A), are impaired in their ability to insert iron into protoporphyrin IX, and accumulate large pools of protoporphyrin IX, which is light sensitive. Because heme_bhas a propensity to cause oxidative damage to the lipid and protein components of cellular membranes, heme_b is normally found associated with proteins in the cell, rather than as free heme,,. The biochemical and/or genetic mechanism by which E. coli and related bacteria regulate expression of the heme_b pathway is poorly understood. The roles of ALA, GTR reductase, and heme_b in E. coli heme_b pathway regulation are not well known. Moreover, the identity of the gene that encodes GTR reductase has been the source of considerable debate. By analogy with the plant pathway, it has been speculated that the E. coli GTR reductase- catalyzed reaction is a regulatory step in heme_b biosynthesis.

Several GTR reductases have been described in E. coli. Two GTR reductase activities of different molecular masses (85 kDa and 45 kDa) have been purified (Jahn et al., J. Biol. Chem.. 266:2542-2548 (1991)). The 45 kDa enzyme is the product of hemA and was thought to be a minor enzyme, and that the major GTR reductase is a 23 kDa enzyme encoded by hemM (Ikemi et al., Gene. 121: 127-132 (1992)). The purified 85 kDa and 45 kDa GTR reductases have not been shown to be sensitive to heπ^ in vitro (Jahn et al., supra). These results suggest that the E. coli pathway is regulated differently from the plant pathway, or that the /jem/V -encoded enzyme is the relevant GTR reductase. Understanding these mechanisms and being able to better control them provides means for enhancing production of heme and heme-containing proteins (also referred to herein as "hemoproteins") in cells.

Intracellular pools of free heme_b and heme synthetic pathway intermediates are regulated in E. coli to prevent toxicity. However, when E. coli or other cells produce heme or hemoproteins, especially by recombinant DNA methods, such regulation may cause a problem because the limited availability of precursors can limit the amount of heme and or hemoprotein produced. Therefore, a need exists to overcome the normal regulatory mechanism that limits heme production when enhanced heme production is desired. The present invention satisfies this need and provides related advantages as well.

Summary of the Invention

The present invention relates to methods of enhancing the expression of heterologous heme-containing proteins by increasing the amount of endogenous heme available to host cells. Such methods are accomplished by exposing the hosts cells to increased amounts of ALA either by stimulating the production of ALA endogenously or by adding exogenous ALA to the culture medium containing the host cells.

In a particularly useful embodiment, the methods are accomplished by inserting at least one copy, preferably multiple copies, of the hemA gene into a host cell to stimulate the endogenous production of ALA in the heme biosynthesis pathway. Alternatively, ALA can be supplemented directly in a culture of host cells to increase heme production. The host cells of the present invention can be prokaryotic or eukaryotic, such as bacteria, yeast, plant, or animal (vertebrate and invertebrate) cells. Preferably, the host cells are bacteria, for example, E.coli. For E.coli and certain other bacteria, the increased production of heme,, results in the enhanced expression of the heterologous heme-containing protein.

Heme-containing proteins (also referred to herein as "hemoproteins") include, for example, hemoglobin, myoglobin, chlorophyll, siroheme, factor F430 and heme-containing enzymes. Such enzymes include, for example, vitamin B12 catalase and nitric oxide synthetase. Various hemoglobins are also contemplated, including wild-type human hemoglobin and variants thereof, including mutant human hemoglobins such as rHbl.l.

The present invention further provides methods for enhancing heme_b production in a host cell, particularly in E.coli. The methods are accomplished by inserting one or more copies of the hemA gene into the host cell and culturing the transformed host cell to allow production of an enhanced amount of heme_b.

Brief Description of the Figures Figure 1 shows, schematically, the pathway for heme synthesis. The C4 pathway for

ALA production is boxed. Genes and enzymes relevant to this study are indicated.

Figure 2 shows the effect of ALA supplementation and rHbl. l production, on heme pools. Cultures contained no IPTG (D) or 300 mM IPTG (Δ). Each data point represents the average of three independent trials. Experimental variation (standard deviation) is shown by error bars.

Figure 3 shows the features of DNA fragments used to identify promoters of hemA. The start of the hemA coding sequence is indicated by the ATG codon. Arrows mark the transcription initiation sites identified by Verkamp & Chel , J. Bacterio 171 :4728-4735 (1989). The triangle below pSGE864 DNA fragment represents the 72bp DNA segment containing the Al transcription start site that was deleted in the PCR amplification.

Figure 4 shows the effect of ALA concentration on hemA-lacZ expression. Cells were grown in 5 ml cultures and assayed as described in Example 1. Data points represent the average of four independent trials.

Detailed Description of the Invention

The present invention generally relates to the enhanced production of heterologous hemoproteins and heme in host cells. The invention is based on the results of the studies described in the Examples below. The results show:

( 1 ) heme_b is a feedback inhibitor of the heme_b pathway; (2) overexpression of rHbl .l (a genetically fused version of human hemoglobin) and

HemA (glutamyl tRNA (GTR) reductase), an enzyme involved in ALA formation, increased intracellular levels of ALA and heme_b, and increased production of rHbl .1 , thus showing that

HemA is rate-limiting;

(3) multiple copies of the hemM gene encoding another possible GTR reductase had no effect on either ALA pools or the rate of rHbl .1 accumulation; (4) overexpression of rHbl.l with Rhodobacter capsulat s ALA synthase, which catalyzes an alternative C4 pathway for ALA formation, increased intracellular heme_b content, but had no effect on rHbl.l production; and

(5) heme_b does not repress ALA formation, while ALA formation limits heme_b synthesis. The present invention is based on the surprising results of the studies. Therefore, in one aspect of the invention, methods of enhancing the expression of a heterologous hemoprotein are provided. Such methods can be accomplished by first exposing host cells to an increased amount of ALA effective to enhance or increase heme production in the host cell.

The host cells can then be cultured to allow enhanced production of the target hemoprotein. In such methods, the host cells can be exposed to enhanced amounts of ALA either endogenously or exogenously. For exogenous exposure, ALA is added as a supplement to culture medium in which the host cells are grown as described in the Examples below.

Alternatively, the production of ALA can be enhanced by overexpression of the hemA gene as described in more detail below. As used herein, genes are usually designated in lower case, while the corresponding proteins they encode are designated in upper case. For example, hemA refers to the gene, whereas HemA refers to the glutamyl tRNA (GTR) reductase.

Also as used herein, the term "heterologous," when referring to a gene, indicates that the gene has been inserted into a host cell that does not naturally carry the gene, either by way of a stable plasmid or through integration into the genome. When referring to a protein, the term indicates that the protein is the product of a heterologous gene. Heterologous proteins are proteins that are normally not produced by a host cell.

As used herein, the terms "enhanced" and "increased" are used interchangeably and mean a measurably greater amount of expression (i.e., overexpression) or production of a gene or target protein compared to the amount of expression or production of the same gene or protein prior to any manipulation of the host cell or culture medium.

Recombinant systems for producing heterologous proteins or polypeptides, including hemoproteins, are well known in the art. The genes encoding the target protein can be placed in a suitable expression vector and inserted into a microorganism, animal, plant, insect or other organism, or inserted into cultured animal or plant cells or tissues. These host cells, organisms or tissues may be produced using standard recombinant DNA techniques following the teachings of the present invention, and may be grown in cell culture or in fermentations. For example, human alpha and beta globin genes have been cloned and sequenced by Liebhaver et al. fProc. Natl. Acad. Sci. USA. 77:7054-58, 1980) and Marotta et al. fj. Biol. Chem. , 242:5040-53, 1977) respectively. Techniques for expression of both native and mutant alpha and beta globins and their assembly into hemoglobin are set forth in U.S. Patent Nos. 5,028,588, 5,545,727 and 5,599,907, and PCT publication WO 97/04110, all incorporated herein by reference.

Methods for incorporating the desired mutations are well known in the art and include, for example, site-directed mutagenesis. Random mutagenesis is also useful for generating a number of mutants at a particular site. Other recombinant techniques are also known, such as those described in U.S. Patent No. 5,028,588, U.S. Patent No. 5,545,727, U.S. Patent No. 5,599,907, PCT Publications WO 96/40920 and WO 97/04110, all incorporated herein by reference.

The genes can be used to construct plasmids to be inserted into appropriate host cells according to conventional methods or as described in WO 96/40920, incorporated herein by reference. Any suitable host cell can be used to express the novel polypeptides. Suitable host cells include, for example, bacteria, yeast, plant, vertebrate and invertebrate animal cells, including mammalian and insect cells. Host cells in transgenic animals are also contemplated. E. coli cells are particularly useful for expressing desired recombinant hemoprotein.

The transformed host cell is then cultured or fermented until soluble hemoglobin is harvested. After the protein has been expressed, it generally should be released from the cell to create a crude protein solution. This can usually be done by breaking open the cells, e.g., by sonication, homogenization, enzymatic lysis or any other cell breakage technique known in the art. The proteins can also be released from cells by dilution at a controlled rate with a hypotonic buffer so that some contamination with cellular components can be avoided (U.S. Patent No. 5,264,555). Cells also may be engineered to secrete the protein of interest by methods known in the art

After breakage of the cells, or secretion, the target protein is contained in a crude cell lysate or crude cell broth or solution. The protein may be purified according to methods well known in the art. For example, methods for purifying hemoglobin-like proteins are taught in PCT publication WO 95/14038, incorporated herein by reference. The hemoproteins, so-produced, can be used for their known purposes. Heme- containing compounds known in the art include, for example, nitric oxide synthase, myoglobin, chlorophylls (from e.g., plants and bacteria), vitamin B12, catalase, siroheme, factor F430 and various hemoglobins, including those from human, yeast, bacteria, worms, crocodiles, and other sources. Heme-containing proteins, for example the various types of hemoglobins, have many uses, including, for example, for delivery of oxygen or therapeutic uses. Other hemoproteins, for example, P-450 enzymes which can be used for oxidation of drugs, alkaloids, terpenes, pesticides, carcinogens, and other xenobiotic chemicals are also important (Porter & Coo. J . Biol. Chem.. 266:13469-72 (1991); Guengerich, J. Biol. Chem.. 266:10019-22 (1991); see also entire January 1992 issue of FASEB Journal). Cytochrome P-450 enzymes can be used for detoxification of various chemicals.

For example, recombinant hemoglobin can be used for a number of in vitro or in vivo applications. Such in vitro applications include, for example, the delivery of oxygen by compositions of the instant invention for the enhancement of cell growth in cell culture by maintaining oxygen levels in vitro (DiSorbo and Reeves, PCT publication WO 94/22482, herein incorporated by reference). Moreover, the hemoglobins of the instant invention can be used to remove oxygen from solutions requiring the removal of oxygen (Bonaventura and Bonaventura, US Patent 4,343,715, incorporated herein by reference) and as reference standards for analytical assays and instrumentation (Chiang, US Patent 5,320,965, incorporated herein by reference) and other such in vitro applications known to those of skill in the art.

In addition, recombinant hemoglobin can be formulated for use in various therapeutic applications. Example formulations suitable for the recombinant hemoglobin of the instant invention are described in Milne, et al., WO 95/14038 and Gerber et al., WO 96/27388, both herein incorporated by reference. Pharmaceutical compositions can be administered by, for example, subcutaneous, intravenous, or intramuscular injection, topical or oral administration, large volume parenteral solutions, aerosol, transdermal or mucus membrane adsorption and the like.

For example, the recombinant hemoglobins of the present invention can be used in compositions useful as tissue oxygenating therapeutics, as substitutes for red blood cells in any application that red blood cells are used or for any application in which oxygen delivery is desired. The recombinant hemoglobin formulated as oxygen therapeutics can be used for the treatment of hemorrhages, traumas and surgeries where blood volume is lost and either fluid volume or oxygen carrying capacity or both must be replaced. Moreover, because the recombinant hemoglobins of the instant invention can be made pharmaceutically acceptable, they can be used not only as blood substitutes that deliver oxygen but also as simple volume expanders that provide oncotic pressure due to the presence of the large hemoglobin protein molecule.

In a further embodiment, the recombinant hemoglobins of the instant invention can be crosslinked by methods known in the art and used in situations where it is desirable to limit the extravasation or reduce the colloid osmotic pressure of the hemoglobin-based blood substitute. Thus, the recombinant hemoglobins can act to transport oxygen as a red blood cell substitute, while reducing the adverse effects that can be associated with excessive extravasation.

A typical dose of recombinant hemoglobin as an oxygen delivery agent can be from 2 mg to 5 grams of hemoglobin per kilogram of patient body weight. Thus, a typical dose for a human patient might be from a few grams to over 350 grams. It will be appreciated that the unit content of active ingredients contained in an individual dose of each dosage form need not in itself constitute an effective amount since the necessary effective amount could be reached by administration of a number of administrations. The selection of dosage depends upon the dosage form utilized, the condition being treated, and the particular purpose to be achieved according to the determination of those skilled in the art.

Administration of recombinant hemoglobin can occur for a period of seconds to hours depending on the purpose of the hemoglobin usage. For example, as an oxygen carrier, the usual time course of administration is as rapid as possible. Typical infusion rates for hemoglobin solutions as oxygen therapeutics can be from about 100 ml to 3000 ml/hour. In a further embodiment, the hemoglobins of the instant invention can be used to treat anemia, both by providing additional oxygen carrying capacity in a patient that is suffering from anemia, and/or by stimulating hematopoiesis as described in PCT publication WO 95/24213, incorporated herein by reference. When used to stimulate hematopoiesis, administration rates can be slow because the dosage of hemoglobin is much smaller than dosages that can be required to treat hemorrhage. Therefore the recombinant hemoglobins of the instant invention can be used for applications requiring administration to a patient of high volumes of hemoglobin as well as in situations where only a small volume of the hemoglobin of the instant invention is administered.

Because the distribution in the vasculature of hemoglobins is not limited by the size of the red blood cells, the hemoglobins of the present invention can be used to deliver oxygen to areas that red blood cells cannot penetrate. These areas can include any tissue areas that are located downstream of obstructions to red blood cell flow, such as areas downstream of thrombi, sickle cell occlusions, arterial occlusions, angioplasty balloons, surgical instrumentation, any tissues that are suffering from oxygen starvation or are hypoxic, and the like. Additionally, all types of tissue ischemia can be treated using the hemoglobins of the instant invention. Such tissue ischemias include, for example, stroke, emerging stroke, transient ischemic attacks, myocardial stunning and hibernation, acute or unstable angina, emerging angina, infarct, and the like. Recombinant hemoglobin can also be used as an adjunct with radiation or chemotherapy for the treatment of cancer. Because of the broad distribution in the body, the recombinant hemoglobins of the instant invention can also be used to deliver drugs and for in vivo imaging as described in WO 93/08842, incorporated herein by reference.

Recombinant hemoglobins can also be used as replacement for blood that is removed during surgical procedures where the patient's blood is removed and saved for reinfusion at the end of surgery or during recovery (acute normovolemic hemodilution or hemoaugmentation). In addition, the recombinant hemoglobins of the instant invention can be used to increase the amount of blood that can be predonated prior to surgery, by acting to replace some of the oxygen carrying capacity that is donated. The present invention further provides methods for enhancing the production of heme_b. Such methods can be accomplished by inserting one or more copies of the hemA gene into a host cell capable of expressing heme_b. The transformed host cell is then cultured to allow production and accumulation of heme,, Heme, usually in the form of heme-arginate as described in U.S. Patent No.

5,008,388, is a clinically important compound which is administered as treatment for hepatic porphyrias which result from a defect in the human heme synthetic pathway. It is also used to treat myelodysplastic syndrome and has potential for treatment of sickle cell disease, β- thalassemia, and prevention of myelosuppression associated with chemotherapy and AZT treatment (Mustajoki, Br. Med. J. 293:538-39 (1986); Volin, Leukemia Res. 12:423-31 (1988)). Therefore, heme and particularly heme-arginate can be used for hematopoiesis, myelodysplastic syndrome, porphyrias, as well as a natural coloring agent.

The results indicate that ALA formation is a rate-limiting step in E.coli heme_b biosynthesis, and that hemA, not hemM, encodes the major GTR reductase in the pathway. Three enzymes are necessary for conversion of glutamate to ALA by the C5 pathway (Fig. 1 ). Although the roles of GLTX and HemL are well known, the roles of HemA and HemM in the reaction catalyzed by GTR reductase are not. Based on their observation that the hemA gene allowed an E.coli hemM mutant to form only small colonies on medium lacking ALA, and that hemM allowed that mutant to form large colonies, Ikemi et al. suggested that hemM encodes the major GTR reductase in E.coli, and that the GTR reductase encoded by hemA is involved in "an alternative minor pathway for ALA formation." (Ikemi, et al., supra). The existence of two pathways for ALA formation is supported by the fact that there are two GTR reductase proteins in E.coli: a 45-kDa protein encoded by hemA, and an 85-kDa protein whose gene product has not yet been identified. Two facts suggest that hemM may not encode GTR reductase. The E.coli hemM protein encodes a 23-kDa protein in E.coli maxicells, and the deduced amino acid sequence of HemM bears no obvious homology to GTR reductase, or to any proteins whose functions have been described in Ikemi et al., supra.

Chen et al., J. Bacteriol.. 176:2743-2746 (1994), proposed that HemM might form a complex with HemA, and that, in combination, these proteins could form the 85-kDa GTR reductase isolated by Jahn. Chen et al. cloned the region of the E.coli chromosome containing hemA and hemM, and subcloned the two genes on separate high-copy number plasmids. It has been previously reported that overexpression of hemA and hemM together resulted in more ALA accumulation than overexpression of hemA alone in either a hemA or hemM host (Chen et al., supra . Chen also found that overexpression of hemM alone did not result in significant ALA accumulation in either a hemA or a hemM mutant. Based on these findings, Chen et al. proposed that HemA and HemM arc required together for maximal accumulation of ALA. The results of the present studies do not support the hypothesis that hemM encodes the predominant GTR reductase for the C5 pathway of ALA synthesis or that both hemA and hemM are required for maximal accumulation of ALA. A GTR reductase (hemA) mutant was not complemented with a plasmid containing hemM (pSGE1 104). ALA accumulation was enhanced when cells were provided with multiple copies of hemA, but not further enhanced when both hemA+M were provided. Purification of HemM has not yet been reported, and no antibodies or enzyme assays are available for monitoring expression of that protein. It was, therefore, not possible to verify HemM function by enzyme assay or Western analysis. However, indirect evidence can be provided to support expression of hemM from the hemM plasmids evaluated in the present studies. A fused DNA fragment containing the 200 bp region upstream of the hemM start codon and the first 18 codons of the hemM coding sequence to codon 9 of the lacZ gene in pMLB1034 (Weinstock et al., Gene Amplification and Analysis, pp. 26-64 (Elsevier/North-Holland Pub. Co. 1983)) was used to create a hemM-lacZ protein fusion. A wild-type strain containing this hemM-lacZ fusion produced 6118+ 492 Miller units of beta-gal actosidase activity, while a control strain, containing the pMLB1034 vector, produced less than one unit of activity. Both the hemM (pSGE1 104) and Λe -4+M(pSGE1103) containing plasmids used herein contain the complete hemM coding sequence as well as 200 bp of upstream sequence.

Intracellular ALA levels in cultures containing multiple copies of hemA+M were not repressed by exogenously supplied heme_b. These results are in agreement with those of Jahn et al. who demonstrated that purified E.coli GTR reductase is insensitive to heme_b (Jahn et al., J. Biol. Chem.. 266:2542-2548 (1991); Verkamp et al., J. Biol. Chem. 267:3875-8280 (1992)). When considered together with data from the present studies on the effect of heme_b on hemA-lacZ fusions, these results imply that GTR reductase is not regulated at either the genetic or biochemical level by heme_b. In plants, where ALA is synthesized by the C5 pathway, and in Rhodobacter sp., where ALA is produced by the C4 pathway, there is evidence that heme_b is a feedback inhibitor of ALA formation. The present data indicate that heme_b '^{s a} repressor of the E.coli heme pathway, but that ALA formation is not feedback inhibited by heme_b. Using lacZ operons fusions, two promoters were identified for hemA. The presence of two functional promoters for the hemA gene could allow cells to rapidly adjust hemA gene expression during different environmental conditions. Darie and Gunsalus in J. Bacteriol. 176(17):5270-76 (1994) reported that expression of a single-copy hemA-lacZ fusion containing both the hemA, and hemA₂ promoters was approximately 2-fold higher in a hemA mutant during anaerobic growth.

Two different systems for increasing ALA pools— hemA and hemA^RC— were evaluated in the present studies. Both systems allowed for an increase in cellular heme content, but the two systems had different effects on rHbl .1 accumulation. The strong tac promoter that drives the hemA^RC gene in pSGEl 1 10 appears to be more efficient than either the hemA, or hemA₂ promoters, neither of which has ideal -10 or -35 consensus elements (Verkamp et al., 1989, supra.). While not wishing to be bound by any theory, it is possible that the level of HemA^RC protein produced by pSGEl l lO is so high that the cell's capacity to produce rHbl. l is compromised, even though the cell accumulates more heme_b. In support of this theory, others have observed a decrease in protein synthetic capacity in cells expressing genes from very strong promoters (Dong et al., J. Bacteriol.. 177: 1497-1504 (1995)).

It was surprising that strains expressing hemA^RC accumulated very large amounts of PBG following induction. One possible explanation for this result can be found in a related study by Harris et al. in Bioorganic Chem.. 21:209-220 (1993). These investigators found that supplementation of Pseudomonas cultures with large quantities of ALA increased heme pathway flux and led to excretion of high levels of prophyrins, with uroporphyrinogen I, a non-enzymatically formed pathway intermediate which is physiologically non-relevant, appearing as the predominant product. In some strains, a significant amount of this incorrect isomer could be further processed. Based on these findings, Harris et al. proposed that high levels of ALA "saturate" the enzymatic machinery that normally converts four PBG molecules to the physiologically relevant uroporphyrinogen III. It is possible that IPTG induction of hemA^RC provides a sufficiently high quantity of ALA to oveload the enzymatic machinery that converts PBG to heme, for example, by allowing formation of incorrect isomers that cannot be further processed.

The following Examples are intended to illustrate, but not limit, the present invention.

Example 1 Materials and Methods

Bacterial strains and plasmids. Bacterial strains and plasmids used in this study are listed in Table 1. pSGE715 is a pUC-derived plasmid containing a synthetic, to -controlled operon composed of two genetically fused alpha subunits and one beta subunit of human hemoglobin and the lad gene for repression of the tac promoter as described in WO 97/04110. pSGE518 was derived from pRS415 (Simon et al., Gene. 53:35-96 ( 1987)) by replacement of the EcoRl-Smal-BAMHI cloning region with an EcoRΪ-HindIII-EagI-Bg/II-XhoI-Hpai-Bam i polylinker. Table 1

Bacterial strains, plasmids, oligonucleotides, and phages

Table 1 continued

Pl-asmids pAlterEX2 low copy cloning vector with tac promoter, Tc^R pACYC184 low copy cloning vector, Cm^R pSGE518 vector for constructing multicopy lacZ operon fusions, Ap pSGEl 103 pACYCl 84 with 2.1 kb Bamlll-Hindlll PCR fragment containing hemAM pSGE1104 pACYC184 with 0.8 kb Bamlll-Hindlll PCR fragment contain hemM pSGEl l lO pAlterEX2 with 1.2kb Bamlll-Bglll PCR fragment containing R. capsulatus hemA RC pSGE715 high copy plasmid for IPTG-con trolled expression of rHb 1.1 ; Tc pSGE494 pACYC184 with 1.2 kb Bamlll-Hindlll PCR fragment containing hemA pSGE862 pRS518 with 0.5 kb Hindlll-Bglll fragment of pSGE494 containing hemA) and hemA₂ promoters pSGE863 pRS518 containing 0.3 kb Hindlll-Bglll fragment containing hemA_/ promoter pSGE864 pRS518 containing 0.4 kb Hindlll-Bglll fragment containing hemA₂ promoter

Oligonucleotides

TG40 5'-CGGGAATACGGATCCAAATGCACC- (SEQ.ID.NO:l) CTGTAAAAAAAGAAAATGATGTACTGC

TG41 5'-CCCAATAATAAGCTTAAAATCGG- (SEQ.ID.NO:2) GCAGGGGCATAGTGATGACAAGTCC

TG224 5'-GGATATCCGAAGCTTCCACTGTGTC- (SEQ.ID.NO:3) CGCATTATTTCAC

EV53 5' -GAATCTAACGGCTTTCGGCAATTA- (SEQ.ID.NO:4) CTCCAAAAGGTTTCCCGCAGACATG

ACCCTTTTAGCACTCGGT

EV54 5'-GGTCATGTCTGCGGGAAACCTTTTGGA (SEQ.ID.NO:5) Table 1 continued

EV113 5'-GCATGTGTCGGATCCGTCTGCGGG- (SEQ.ID.NO:6) AAATAATAC

TG222 5'-GCATGTGTCGAATTCGAAAGCCGTTA (SEQ.ID.NO:7) GATTCTG

EV98 5'-GCGATACGGGATCCAAGGAGGAAT- (SEQ.ID.NO:8) TTACATATGGACTACAATCTCGCGCTC

EV99 5'-GTGATCGAAAGATCTTCACGCACA- (SEQ.ID.NO:9) GCGCGCCCAGAGCAG

EV50 5'-GCπTTACGCAGATCTTCTTCATT- (SEQ.ID.NO: 10) AAGATTGTG

EV39 5'-CCAATGAATTCAAGCTTAATTAC- (SEQ.ID.NO:l l) TCCAAAAGGGGGCGC

E45 5'-CCAATGAATTCAAGCTTAAAA- (SEQ.ID.NO:12) TCGGGCAGGGG

E57 5'-GTAGAGGCTTTTACGCAGATCTT- (SEQ.ID.NO:13) CΓTCATTAAGATTGTG

PHAGES λRS45 vector for transferring multicopy lacZ fusion to chromosome

Plvir generalized transducing phage

Genetic and recombinant DNA methods. Plvir transductions were performed according to standard methods as described in Silhavy et al., Experiments with Gene Fusions (Cold Springs Harbor Laboratory, 1984). Multicopy hemA-lacZ operon fusions were constructed in pSGE518, transferred to λRS45 by homologous recombination, and then integrated as prophages into the chromosomal attachment site (attλ) according to Simon et al., supra.

Chromosomal DNA for PCR amplification was obtained by standard methods as described by Silhavy et al., supra.. Oligonucleotides for PCR amplification and DNA sequencing were synthesized using an Applied Biosystems model 392 DNA synthesizer

(Applied Biosystems, Inc., Foster City, CA) using reagents obtained from Biogenex (San Ramon, CA) and methods recommended by the manufacturer.

All PCR amplification reactions contained the following reagents: -100 ng of template DNA, 20-50 pmol of each primer, 20 mM Tris-HCl, 10 M KC 1 , 6 mM (NH₄)₂SO₄, 1.5 mM MgCl₂, 0.1% Triton X-100, 0.2 mM of each of the four deoxyribonucleotide triphosphate

(dNTPs), and 2.5 U Pfu polymerase (Stratagene Cloning Systems, La Jolla, CA), in a 100 μl reaction volume. Thermocycling was performed using an Ericomp Twinblock System (Ericomp, San Diego, CA). Cycle conditions and primers are described below. The BamHl-HindlJI fragment in pSGEl 103 was obtained by PCR amplification using primers TG40 and TG224 and SGE1670 (WO 97/04110) chromosomal DNA as the template (other similar strains can also be used). Cycle conditions were as follows: one cycle of 5 min. at 95°C, 30 sec. at 65°C, and 30 sec. at 75°C,; 1 cycle of 10 min at 75°C.

All three hemA promoter fragments (Fig. 3) were amplified using the following cycle conditions (program "J"): one cycle of 5 min. at 95°C, 5 min. at 50°C, 1 min at 72°C; 28 cycles of 1 min. at 94°C, 1 min 50°C, and 30 sec. at 72°C; one cycle of 10 min. at 72°C. The hemA promoter fragment in pSGE863 (any other E. coli having a hemA promoter can be used) was amplified using primers EV50 and EV39. The hemA promoter fragment in pSGE864 (any E. coli having a hemA promoter can be used) was amplified using two sets of primers: EV45 an EV57, and EV53 and EV54. The use of two primer sets allowed removal of a small DNA segment containing the hemA, transcription start site described by Verkamp and Chelm (J. Bacteriol.. 171 :4728-4735 (1989)).

The hemA gene fragment contained in pSGE494 (any other E. coli having a hemA gene fragment can be used) was amplified using program "J" and primers TG40 and TG41. The function of the hemA gene in pSGE494 was verified by complementation assay using an E. coli hemA mutant.

A DNA fragment containing the coding sequence of hemA ^C was obtained by PCR using chromosomal DNA isolated from Rhodobacter strain SB 1003 (Hornberger, Mol. Gen- Genet. 221 :371-78 (1990); any other Rhodobacter strain containing the hemA^RC promoter can be used), primers EV98 and EV99, and program "J" described above. In addition to confirming the DNA sequence of the hemA^RC allows an E. coli ALA auxotroph to grow without ALA supplementation.

Primers and cycle conditions used for amplification of the BamHI-HindlH fragment in pSGEl 104 (any other E. coli strain containing the BamHI-Hindlllfragm nt can be used) were as follows: EV1 13 and TG224; one cycle of 5 min. at 95°C, 5 min. at 60°C, 3 min. at 72°C,

33 cycles of 30 sec. at 95°C, 30 sec. at 60°C, and a final cycle of 8 min. at 72°C.

Plasmid DNA for cloning and DNA sequencing was isolated using the Wizard plasmid isolation kit (Promega, Madison, WI) according to the manufacturer's instructions. Restriction digests, gel electrophoresis, DNA ligations and transformations were performed according to standard methods described in Sambrook et al., Molecular Cloning: A Laboratory Manual (Cold Springs Harbor, 1989). PCR fragments for ligation reactions were purified using a GENECLEAN II kit (Bio 101, Vista, CA). Both strands of DNA fragments generated by PCR were sequenced using the Prism DYE-terminator cycle sequencing system (Applied Biosystems, Foster City, CA) and an Applied Biosystems model 373 automated sequencer (Applied Biosystems, Foster City, CA). DNA sequences were analyzed using programs contained in the MacVector (version 4.5.2.) software package (International Biotechnologies, Inc. New Haven, CT).

Beta-galactosidase assays. Strains were grown overnight in 5 ml of M63 salts (Miller, Experiments in Molecular Genetics. (Cold Springs Harbor, 1972)) supplemented with 0.4% glucose, 0.1% casamino acids, 50 μg/ml proline, 40 μg/ml thiamine, 1 mM MgSO₄, and appropriate antibiotics. Cells were diluted 1 :50 into test tubes containing 5 ml of the same medium, and grown at 37°C with shaking to OD^ 0.4-0.8. Cultures were assayed for β- galactosidase according to the method described by Miller, supra.

Media and growth conditions. Shake flask scale experiments were performed in 50 ml of DM-1 (16) in 250-ml Erienmeyer flasks. Cultures were grown at 37°C with shaking to OD_goo-0.6, induced with 300 μM IPTG and grown for an additional 4 hours. At harvest, final culture densities (measured at OD^) were between 2.8 and 4.0. ALA (Sigma Chemical Co., St. Louis, MO) was solubilized in water, filter sterilized, and added to cultures to yield solutions with the following final concentrations: 0.02 μg/ml, 0.10 μg/ml, 0.39 μg/ml, 1.56 μg/ml, 6.25 μg/ml, 25 μg/ml, and 100 μg/ml. Cell pellets for heme_b measurements were stored at -70°C until analysis

Fermentations were performed at 30°C in 15 liter fermentors (LSL Biolafitte, Inc.

Princeton, NJ). The seed inoculum for 15 liter fermentors was prepared in a two stage process. The primary seed stage was 500 ml of DM59 medium in a 2.5 liter shake flask inoculated with 0.5 ml of a stock culture preserved in 7% (v/v) dimethyl sulfoxide at ~80°C.

The salts used in DM59 medium were as follows: 33 mM KH₂PO₄, 46 mM K₂HPO₄, 13 mM NaH₂PO₄, 18 mM Na-HPO₄, 19 mM (NH₄)₂SO₄, 5.4 mM K₃ citrate, 2.2 mM Na₃ citrate, 4.2 mM MgSO₄, 7.2 mM H₃PO₄. Trace metals were added to DM59 salts to the final concentrations indicated: 0.91 mM FeCl₃, 0.14 mM ZnCl₂, 12 μM CoCl₂, 10 μM Na₂Mo0₄, 0.13 mM MnCl₂, 0.41 mM CaCl₂ and 54 μM CuSO₄. Glucose was added to a final concentration of 1% (w/v). Thiamine was added to a final concentration of 0.32 mg/ml. Antibiotics (see below) were added as needed to maintain plasmids. The primary seed stage . ,.

16 cultures were grown at 30°C with shaking for 8-10 hrs or until the OD₆₀₀=0.81.5, and 400 ml of culture were used to inoculate a secondary seed stage a 2 liter Bioflo III fermentor (New Brunswick, Edison, NJ) containing 1600 ml of DM59 medium supplemented with 1% (w/v) glucose, trace metals (in concentrations described above), and appropriate antibiotics. The pH of the secondary seed stage culture was maintained at 20% in 2 liter fermentors. The secondary stage culture was grown to 00₆₀₀=5-10, and a 500 ml aliquot was used to inoculate the final stage culture: a 15 liter fermentor (LSL Biolafitte, Inc., Princeton, NJ) containing 8 liters of DM59 medium. Glucose was added to 15 liter fermentors at a starting concentration of 2.0 g/liter, and maintained thereafter at 2-10 g/liter. The pH and dissolved oxygen concentrations of 15 liter fermentors were maintained as described above. When cultures reached an OD^-SO, isopropyl B-D-thiogalactopyranside (IPTG) (Sigma Chemical Co., St.

Louis, MO) was added to a final concentration of 100 μM to induce rHbl .l expression.

Heme_b, obtained as bovine hemin (Amesresco, Solon, OH), was dissolved in IN NaOH to a final concentration of 50 mg/ml. Where specified, heme_b was added to fermentors at the time of IPTG induction, and at 3 and 6 hrs. post-induction, as 10 ml, 13 ml, and 17 ml aliquots.

Cell pellets were collected and stored at -70°C until analyzed.

Antibiotics (Sigma Chemical Co., St. Louis, MO) were added to shake flasks and fermentors in the following concentrations: tetracycline, 15 μg/ml; chloramphenicol, 25 μg/ml, and ampicillin, 100 μg/ml. ALA and PBG assays. One ml fermentation samples were resuspended in 10 mM

MES (pH 6.0). Lysozyme, NaCl, and DNAse were added to final concentrations of 500 μg/ml (lysozyme), 100 mM (NaCl) and 60 μg/ml (DNAse), and samples were incubated on ice for 20 min., and then at 37°C for 2 min. Proteinase K (Sigma Chemical Co., St. Louis, MO) was added to a final concentration of 150 μg/ml, and samples were incubated for an additional 20 min. on ice. The sample pH was lowered to 5.5-5.9 by addition of 10% acetic acid.

Samples were freeze-thawed and then centrifuged for 10 min. at 13,000 x g at 4°C. The supernatant was removed and ALA and PBG were separated using a two-column chromatography system (ALA and PBG Column Test Kit) obtained from BioRad (Hercules, CA). ALA and PBG levels were quantified by their reactivity with Ehrlich's reagent, essentially as described in the BioRad test kit.

Heme extraction. Forty ml samples of shake flask cultures were harvested by centrifugation at 13,000 x g for 5 min., and pellets were stored for 1 to 5 days at -200°C. The pellets were thawed on ice, resuspended in one ml acetone- IN HC1 (9: 1 ) and vortexed vigorously for 45 seconds. After a one hour incubation on ice, the samples were vortexed again and centrifuged (13,000 x g) for ten minutes. The supernatants were transferred to fresh tubes, and 25 μl aliquots were analyzed by reversed-phase high-performance liquid chromatography (HPLC). Recovery of heme_b from spiked samples was greater than 90% using these conditions.

Heme_b quantitation. Samples prepared by the method described above were injected onto a Hypersil ODS column (5 micron, 2.1 x 150 mm Alltech Associates, Deerfield, IL) equilibrated with acetonitrile-methanol-H₂O (1: 1:40) and 0.1% (v/v) trifluoroacetic acid (TFA). After injection, the solvent was changed to acetonitrilemethanol (1: 1), 0.1 % (v/v) TFA, with a linear gradient over four minutes. Elution with the second buffer continued for 1 1 minutes at a flow rate of 0.5 ml/min. Elution was monitored at 404 nm with a Hewlett-Packard 1090 Diode Array Detector (Hewlett-Packard, Palo Alto, CA). Bovine hemin (Amresco, Solon, OH) was dissolved in 0.1 N N_aOH, quantified spectrophotometrically (Falk, Porphyrins and Metalloporphyrins. pp.231-246 (Elsevier Pub. Co., 1994)), and used as the standard for heme_b measurements. The heme_b values reported represent total, extractable heme_b present in the cell, including heme_b found associated with rHbl. l as well as other E. coli proteins.

Hemoglobin assay. Samples (1 ml of fermentation broth at) were pelleted and resuspended in 25 mM Na^O_?. Lysozyme and NaCl were added to final concentrations of 1 mM NaCl and 0.75 mg/ml lysozyme, and the samples were incubated first at 4°C (30-40 min), and then at 37°C (3 min). DNAse (60 μg/ml) was added, and the samples were incubated 15 min. at room temperature. The samples were freeze-thawed, treated for 10 sec. with CO gas, and then diluted into a solution of 80 mM Tris-HCl, 2 M NaCl (pH8.0). The crude lysates were heated at 65°C for 4 min., and centrifuged at 13,000 x g for 2 min to remove cellular debris and contaminating proteins. The supernatant fraction containing partially purified hemoglobin was saved, and the hemoglobin was quantified by immobilized metal chelate chromatography using a Biocad Perfusion Chromatography Workstation (PerSeptive Biosystems, Cambridge, MA). The capture column was charged with a solution of 20 mM Zn(OAc)₂, 200 mM NaCl, and equilibrated with 8 mM Tris-HCl, 200 mM NaCl (pH 8.0). Samples were loaded onto the column in the same buffering system used for column equilibration, and the column was washed with 20 mM Tris-HCl, 500 mM NaCl (pH 8.0). Hemoglobin was eluted from the column with 40 mM Tris-HCl, 1 M NaCl, 25 mM EDTA (pH 8.3). Hemoglobin elution was monitored at 412 nm. Example 2 Effect of hemoglobin in expression on the heme pathway

Plasmid pSGE715 is a high-copy number plasmid for IPTG-controIled expression of the alpha and beta subunits of human hemoglobin. Following IPTG induction, E. coli strains containing pSGE715 produce large quantities of fully functional tetrameric hemoglobin variant (rHbl.l; WO 97/041 10). Because each hemoglobin tetramer has the capacity to bind four heme_b groups, induction of rHbl.l is expected to rapidly deplete cellular pools of heme_b. Accordingly, if heme_b is a repressor of E.coli heme_b synthesis, removal of heme_b by rHbl . l should activate heme_b synthesis. The results indicate that heme_b is a negative regulator of the heme_b pathway. Prior to induction of rHbl.l, shake flask cultures of SGE1453 contain approximately 25 pmoles/OD₆₀₀*ml of heme_b in heme_b-containing proteins and as free heme_b (Fig. 2, no ALA, zero hour time point). Four hours following induction of rHbl . l, cultures of SGE1453 have approximately 75 pmoles/OD₆₀₀*ml heme_b (Fig. 2, no ALA, zero time point).

Example 3 Effect of exogenous ALA on heme_b synthesis

The effect of ALA levels on heme_b synthesis was examined by feeding shake flask cultures various quantitites of ALA, and then monitoring cellular heme_b levels four hours after induction of rHbl . l . Maximal heme_b accumulation occurred with 6.25 μg/ml ALA; higher levels of ALA resulted in no additional increase in heme_b pools (Fig. 2). These results indicate that ALA formation limits heme_b synthesis. ALA supplementation resulted in a 3-fold increase in cellular heme_b (Fig. 2). The effects of ALA supplementation and rHbl. l production were synergistic. Cultures expressing rHbl. l, which were supplemented with 6.25 μg/ml ALA, contained approximately 7-fold more heme_b than an uninduced control.

Example 4 Promoters for hemA .

Using the method of SI nuclease protection, two potential transcription start sites, -38 and -131 , and two σ70 consensus promoters (hemA_] and hemA-,) for the hemA gene in aerobically growing E.coli cells have been previously identified (Verkamp & Chelm, JL Bacteriol .. 171 :4728-4735 (1989)). The hemA_x and hemA-, promoters were proposed to lie at nucleotides -45 to -70 and -145 to -173, respectively, relative to the hemA start codon. A divergent transcript that started 83 nucleotides upstream of the hemA₂ promoter on the opposite strand was also detected. This divergent transcript was shown to encode a 23-kDa protein, hemM, and a role for HemM in ALA synthesis was proposed (Chen et al., J. Bacteriol.. 176:2743-2746 (1994); Ikemi, supra.) .

To assess the ability of the two isolated potential hemA promoters to function as heterologous promoters, 244 nucleotides of the hemA coding sequence and 158 to 230 nucleotides of sequence 5-prime of the hemA start codon were fused to the E.coli lacZ gene. Those fusions were recombined into the E.coli chromosome for evaluation. Three different hemA-lacZ fusions were constructed. Because a hemA mutant requires ALA supplementation for growth, the initial analysis of these fusions were performed in a hemA* host (Fig. 3). One fusion (SGE1857) included the hemA_l promoter, one (SGE1858) included the hemA₂ promoter, and one (SGE1859) included both the hemA_x and hemA₂ promoters. Beta- galactosidase activity assays indicated that both the hemA_x and hemA₂ promoters are biologically active, and that the hemA, promoter is the stronger promoter (Table 2). Similar results were obtained when the three hemA-lαcZ fusions were integrated into the genome of a hemA mutant and grown with 0.2 μg/ml ALA.

Table 2

Relative strengths of hemA promoters

Strain" Relevant β-galactosidase Activity

Characteristics (Miller Units)"

SGE1857 hemA .₂-lαc_τ 2022 (±216)

SGE1858 hemA.-lacZ 3049 (±351 )

SGE1859 hemA₂-lαcZ 333 (± 28)

"Cells were grown in five ml cultures and assayed as described in Materials and Methods. ^bValues reported are the average of four independent experiments. Variation is indicated in parentheses as standard error. A control strain containing pSGE518 produced 40

(±12) Miller Units of β-galactosidase.

The effects of exogenous ALA and heme_b on hemA expression were evaluated in hemA mutants containing single-copy hemA-lαcZ fusions. The results show that hem_M but not the hem_M promoter is regulated in response to the intracellular level of ALA. Expression of a hemA_x lαcZ (SGE1861) fusion was activated approximately 2-fold when the concentration of ALA was decreased (Fig. 4), while expression of a hemA₂-lacZ (SGE1862) fusion was unaffected by ALA concentration (Fig. 4). Addition of exogenous heme_b (40 μg/ml) to cultures of SGE1861 (hemA_i lacZ) grown in a glucose minimal medium containing ALA (either 0.2 or 0.05 μg/ml) did not repress expression of any of the hemA-lacZ fusions tested. Thus, heme_b does not regulate hemA gene expression or the hemA promoters tested.

Example 5 Overexpression of hemA and hemM

ALA, which is structurally similar to the dipeptide glycyl-glycine, is taken up by E.coli using a dipeptide permease. Because ALA is normally synthesized by E.coli cells rather than taken up by cells, transport of ALA by the dipeptide permease may be a relatively inefficient method for increasing intracellular ALA pools. Therefore, increasing ALA pools by genetically manipulating enzymes involved in ALA formation, especially GTR reductase, was examined to determine if they have a greater effect on cellular heme_b content than supplementation with exogenous ALA. Accordingly, plasmids for overexpression of the hemA and hemM genes from their native promoters were constructed and evaluated in fermentation cultures (Table 3).

Table 3

Strain Relevant Characteristics

SGE1453 control SGE2664 multicopy hemA SGE2658 multicopy hemA+hemM SGE2658^d multicopy hemA+hemM SGE2680 multicopy hemM SGE2681 multicopy hemA^RC

Cells expressing rHbl. l and containing multiple copies of either hemA alone or hemA and hemM together accumulated large intracellular pools of ALA relative to a control (Table 4). Because ALA accumulated prior to and after induction of rHbl .l expression, multiple copies of hemA (SGE2664) or hemA+hemM (SGE2658) appear to result in constitutive expression of GTR reductase and accumulation of ALA (Table 4). Control strains (SGE1464) had undetectable levels of ALA prior to induction of rHbl .l and only low levels of ALA post- induction. No significant differences in the ALA pools were found when cells containing multiple copies of hemA (SGE2664) and muliple copies of hemA+hemM (SGE2658) were compared (Table 4). ALA pools in SGE2680, which expresses hemM and rHbl . l , were similar to a control strain (SGE1453) both pre- and post-induction (Table 4). Pools of other pathway intermediates (uroporphyrinogen UI, coproporphyrinogen III, and protoporhyrin IX) were indistinguishable in strains SGE1453, SGE2644, SGE2658, and SGE2680. Cells expressing rHbl .l and either HemA (SGE2664) or HemA+HemM (SGE2658) contained more heme_b (Table 5) and accumulated rHbl.l (Table 6) at a faster rate than a control strain. Multicopy hemA+hemM appears not to have a greater effect on rHbl . l accumulation than multicopy hemA. When HemM and rHbl.l were expressed in the same cell (SGE2680), the rate of rHbl. l accumulation was no faster than a strain expressing only rHbl . l (SGE1453). Based on these data, it appears that ALA formation is a rate-limiting step, and that the hernA-cncoded GTR reductase is a rate-limiting enzyme n the E.coli heme_b pathway.

Feeding exogenous heme_b to cells overexpressing hemA+hemM and rHbl . l (SGE2658) had no effect on intracellular ALA levels (Table 4). Because cultures of an isogenic strain, SGE1453, produce approximately 3-fold more rHbl .l when supplemented with heme_b (Table 6), it appears that SGE2658 is capable of heme_b transport, and that ALA formation in E.coli is not repressed by heme_b.

Table 4

Effects of hemA, hemM, and hemA^RC on Intracellular Levels of ALA in

Fermentation Cultures

ALA (-IPTG) ALA (+IPTG)

(pmoles/OD₆₀₀ ⁺ml) (pmoles/OD₆₀₀ ⁺ml) - ^25

1017(± 102) 1690(± 140)

795(± 121) 1735(± 324)

^NT 2208(± 221)

ND ND

ND 391 (± 148) Table 5

Effects of hemA, hemM, and hemA^RC on Intracellular Levels of PBG and heme_b in Fermentation Cultures

PBG (-IPTG) PBG (+IPTG) heme_b (+IPTG)

(pmoles/OD₆₀₀ ⁺ml) (pmoles/OD₆₀₀ ⁺ml) (pmoles/OD₆₀₀ ⁺ml)

ND⁺ ND⁺ lθl(± l l)

277(± 45) 87(± 5) 164(± 19)

457(± 69) 68(± 9) 21 H± 13)

NT 52(± 6) NT

ND ND 90(± 7)

<25 873(± 45) 165(± 13)

Table 6

Effects of hemA, hemM, and hemA on Rates of rHbl.l Accumulation

Strain Relevant Characteristics rHbl.l (+IPTG) (mg/(OD₆₀₀*liter)/hr^a

SGE1453 control 0.284(± 0.054) SGE1453^b control 0.951 (± 0.077) SGE2664 multicopy hemA 0.446(± 0.052) SGE2658 multicopy hemA + hemM 0.592(± 0.032) SGE2680 multicopy hemM 0.186(± 0.032) SGE2681 multicopy hemA^RC 0.341 (± 0.041) ^a Four to 8 replicate fermentations were performed for each strain reported. rHbl.1 measurements were made during the 8 hr. period following induction. Slopes and variation (standard error) were obtained from linear regression lines with correlation coefficients ranging from 0.72 to 0.97. ^b Heme_κ was added to cultures as described in Materials and Methods. Results are the average of four fermentation trials. Example 6 Overexpression of ALA synthase

Although most bacteria synthesize ALA fro glutamate by the C5 pathway, some bacteria, such as Rhodobacter, synthesize ALA by an alternate route, the C4 or Shemin pathway (Fig. 1). In contrast to the C5 pathway which requires three steps, the C4 pathway requires only a single enzymatic step, and is catalyzed by the enzyme ALA synthase (E.C. 2.3.1.37)(Fig. 1). The Rhodobacter gene that encodes ALA synthase has been designated hemA. To avoid confusion with the E. coli hemA, which encodes a different activity, the Rhodobacter hemA gene is referred to herein as hemA^RC. E.coli hemA mutants, which require ALA supplementation for growth, can be complemented by the Rhodobacter /ιemΛ^fiC(Hornberger et al., Mol. Gen. Genet.. 221:371-378 (1990)). Thus, E.coli can produce heme_b using enzymes from either the C4 or C5 pathways. hem^RC was placed under control of the tac promoter to create an inducible system for studying the effect of high-level expression of ALA synthase on heme_b pathway regulation.

ALA and PBG pools in cells expressing hemA^RC (SGE2681) were significantly higher than in a control strain (SGE1453) following induction of HemA^RC and rHbl . l expression (Tables 4 and 5). Although strains expressing HemA^RC and rHbl. l (SGE2681) produced more heme_b, the rates of rHbl. l accumulation for strains SGE1453 and SGE2681 were indistinguishable (Tables 4 and 5). The ratios of ALA and PBG in cells expressing hemA (SGE2644), hemA+M (SGE2658) and hemA^RC (SGE2681) were considerably different. Strains SGE2644 (hemA) and SGE2658 (hemA+M) accumulated very large pools of ALA and smaller pools of PBG, while SGE2681 (hemA^RC) accumulated large pools of both ALA and PBG (Tables 4 and 5). An SGE1453 control strain failed to accumulate PBG either prior to or following induction, and ALA pools were low or undetectable pre- and post-induction.

The foregoing description of the invention is exemplary for purposes of illustration and explanation. It will be apparent to those skilled in the art that changes and modifications are possible without departing from the spirit and scope of the invention. It is intended that the following claims be interpreted to embrace all such changes and modifications.

Claims

Claims:

1. A method for enhancing the production of a heterologous hemoprotein in a host cell, comprising: exposing said host cell to an increased amount of δ-aminolevulinic acid (ALA) effective to increase heme production in said host cell; and culturing said host cell to produce an enhanced amount of said hemoprotein.

2. The method of claim 1 , wherein exposing ALA to said host cell comprises adding exogenous ALA to a culture medium containing said host cell.

3. The method of claim 1, wherein exposing ALA to said host cell comprises enhancing the endogenous production of ALA in said host cell.

4. The method of claim 3, wherein enhancing the endogenous production of ALA comprises inserting at least one hemA gene into said host cell.

5. The method of claim 4, wherein multiple copies of said hemA gene are inserted into said host cell.

6. The method of claims 1-5, wherein said host cell is a bacteria, yeast, plant, vertebrate or invertebrate animal cell.

7. The method of claim 6, wherein said host cell is a bacteria.

8. The method of claim 7, wherein the bacteria is E.coli.

9. The method of claims 7 or 8, wherein said heme is heme_b.

10. The methods of claims 1-5, wherein the heme-containing protein is hemoglobin, myoglobin, chlorophyll, siroheme, factor F430 or a heme-containing enzyme.

11. The method of claim 10, wherein said heterologous heme-containing protein is hemoglobin.

12. The method of claim 11 , wherein said hemoglobin is human hemoglobin or a variant thereof.

13. The method of claim 12, wherein the hemoglobin variant is a mutant of said human hemoglobin.

14. The method of claim 13, wherein said mutant hemoglobin is rHbl.l.

15. A method of enhancing heme_b production in a host cell capable of expressing heme_b, comprising: inserting at least one copy of a hemA gene into the host cell to form a transformed host cell; and culturing the transformed host cell to produce an enhanced amount of heme_b-

16. The method of claim 15, wherein said host cell is E.coli.

17. The method of claim 15, wherein multiple copies of the hemA gene are inserted into the host cell.

18. A host cell transformed with at least one additional copy of a hemA gene, wherein said host cell is capable of expressing heme_b.

19. The host cell of claim 17, wherein multiple copies of the hemA gene are inserted into the host cell.

20. The host cell of claim 17, wherein said host cell is E.coli.