WO1989002471A1 - RECOMBINANT DNA CONSTRUCTS CONTAINING AN r3 PROMOTER - Google Patents

Abstract

Recombinant constructs of DNA are provided which are useful for the expression of a heterologous polypeptide in a bacterial host. These constructs contain a novel promoter, the r3 promoter. Variations on the basic construct are made which enhance the level of expression directed by the r3 promoter. These variations are the inclusion of sequences containing strong ribosomal binding sites, and containing A-T rich sequences. In addition, the expression directed by the r3 promoter may be controlled by the insertion of an operator sequence into the construct.

Description

RECOMBINANT DNA CONSTRUCTS CONTAINING AN r3 PROMOTER

Background

Technical Field

The present invention relates to the high level expression of useful proteins in bacteria by methods which utilize reco binant DNA compositions.

Background Art

Various proteins and peptides synthesized by living organisms may have important uses in medicine, sci¬ ence, agriculture, and industry. However, the difficul- ties involved in isolating many of these products from their natural sources often causes them to be commercially unavailable. These di ficulties, which include the un¬ availability of the natural source, the low level of the product in the natural source, or potentially dangerous components in the natural source which are not readily removed during isolation procedures, may be overcome by synthesizing the product in prokaryotic organisms which contain the genetic sequence for the desired product in recombinant DNA vectors. Similarly, analogs of natural proteins or peptides which may have commercial use, and which are not available in nature may also be synthesized by prokaryotic organisms containing recombinant DNA vec¬ tors .

The synthesis of a protein or peptide reflects the expression of a structural gene, which includes both transcription of the gene into mRNA, and translation of the mRNA into a polypeptide. The level of expression of a structural gene in bacteria is subject to regulation by regulatory sequences encoded within, the genome. These sequences affect both transcriptional and translational efficiency.

Among the sequences known, to regulate^' transcrip¬ tion are promoter sequences. A promoter is a DNA sequence that directs RNA polymerase to bind to DNA and to initiate RNA synthesis. Different promoters work with different efficiencies. Strong promoters cause mRNAs to be initi¬ ated at high frequency; weak promoters direct the synthesis of rarer transcripts. Examples of strong E. coli promoters include: trp, from the tryptophan operon (Bennett, B.N. et al, J. Mol. Biol. (1976) 121, 113); rRNAB PI, ribosomal RNA promoters (deBoer, H.A. et al, Cell (1979) 11_, 201; Csordas-Toth, E. et al, Nucl. Acids Res. (1979) 1_, 2189; Brosius, J. et al, J. Mol. Biol. (1981) 148, 107); P , from bacteriophage lambda (Maniatis, T. et al, Nature (1974) 25_0, 394); P__ant, from bacteriophage P22 (Youderain, P. et aT7 Cell (1982) 30, 843); and tac, a hybrid promoter derived from trp and lac (de Boer, H.A. et al, Proc. Natl. Acad. Sci. USA (1983) 80, 21) . Examples of weak E. coli promoters are lac, from the lac operon (Maizels, N., Proc. Natl. Acad. Sci. USA (1973) 7_0, 3585); and gal, from the galactose operon (Musso, R. et al, Proc. Natl. Acad. Sci. USA (1977) 7_4, 106) .

It has been recognized that the level of expres- sion of a structural gene within a recombinant DNA vector will be affected by the characteristics of the promoter for the gene. Expression vectors have been constructed using various natural bacterial promoters. See Harris, T.J.R., in Genetic Engineering, ed. R. Williamson, vol 4, Academic Press, New York, 1983, pp. 128-185. The highest reported level of stable protein synthesized in one of these systems was 46% (Tanka, T. et al, Nucl. Acids Res. (1982) 1D: 1741). An expression vector has been constructed which contained a synthetic DNA fragment en- coding the consensus promoter sequence. Rossi,

J.J. et al, Proc. Natl. Acad. Sci. (1983) 8JD, 3203. The level of expression due to this promoter in vivo did not reflect its strength in vitro. Other expression vectors utilizing hybrid promoters have also been constructed. The strength of these hybrid promoters is equal to that of the strong natural E. coli promoters. See de Boer, H.A. et al, in Genes, Structure and Expression, A.M. Kroon, ed. Wiley and Sons, 1983, pp. 205-248.

Since many natural bacterial promoters which may be strong have not yet been identified, cloning vectors have been constructed to select fragments of DNA which contain promoter activity. One such cloning vector which is suitable for the isolation of strong promoters is that described by Brosius. Brosius, J. , Gene (1984) 27, 151. The efficiency of translation is another factor which may govern the level of expression of a structural gene into a polypeptide. Translational efficiency appears to be regulated, at least in part, by characteristics as¬ sociated with the ribosome binding site in the mRNA, which are encoded within the DNA. Ribosome binding is determined by a sequence of 3-12 bases of the mRNA which are complementary in sequence to the 3'-end of the 16S rRNA of the ribosomes (Shine, J. et al, Proc. Natl. Acad. Sci. USA (1974) 7_1, 1342). The distance of the ribosome binding site to the ATG initiation codon is also a factor. Attempts have been made to increase the translational efficiency of expression vectors. In some cases, structural genes have been placed under the control of strong promoters, and the distance between the natural ribosome binding site and the start codon has been systemat±cally altered. Roberts et al, supra; Shepard et al, DNA (1982) 1, 125; Ikehara et al, Proc. Natl. Acad. Sc. USA (1984), 01 , 5996. A disadvantage of this approach is its randomness, which makes it time 5 consuming.

An expression vector has been constructed utilising a synthetic promoter equivalent to the strong promoter- from bacteriophage T5, T5P25, as well as a strong synthetic: ribosome binding site. Jay et al, Proc. Natl.

10. Acad... SS . USA (1984). When the structural gene for gamma—interferon is incorporated into this vector, the level of expression of this protein is 16% of the total E. coli cell protein.

Another type of expression vector has been

15 constructed which contains a strong promoter, trp, a synthetic ribosome binding site, and a unique restriction enzyme site, Sph I, downstream from the ribosome binding site, into which a structural gene may be cloned. In this vector, an ATG codon from the restriction site is added to

20 the structural gene which is inserted therein. Nishi et al, DNA (1983) 2 , 265.

Disclosure of the Invention

One embodiment of the invention is a recombinant 25^' DNA construct. The construct is comprised of a first nucleotide sequence encoding an r3 promoter sequence, wherein the r3 promoter sequence is selected from a group consisting of the nucleotide sequence

30 5'-ACAGAAATTTTTCGCCGTACGCTATTGCGTGACGTAGATTCGTG

ACGTATAGTTACTACAGCTTATTTGTATATAACCACCATCAGGT-3' ,

and mutants thereof, wherein the mutants exhibit promoter activity.

35 Another aspect of the invention is a vector which is comprised of a recombinant DNA construct which has the above described characteristics. This vector is used in a method to produce a recombinant polypeptide. The method consists of providing a population of cells transformed with the vector, growing the transformed cells under conditions which allow expression of the vector, and recovering the polypeptide.

The invention is also embodied by a purified DNA fragment comprised of the sequence

5'-ACAGAAATTTTTCGCCGTACGCTATTGCGTGACGTAGATTCGTGACGTATAGTT ACTACAGCTTATTTGTATATAACCACCATCAGGT-3' .

Description of the Figures

Figure 1 shows pKK232-8, the cloning vector used to isolate the r3 promoter.

Figure 2 shows the nucleotide sequence of the EcoRI-Ba HI fragment containing the r3 promoter, and the 4 promoter regions identified by homology to the -10 and -35 consensus sequences of bacterial promoters.

Figure 3 shows the scheme by which the parental r3 vector, pALr3, was constructed.

Figure 4 shows the map, useful features, and DNA sequence of the synthetic DNA insert in the r3 containing vector, pALlO.

Figure 5 shows the map, useful features, and DNA sequence of the synthetic DNA insert, of the r3 expression vector, pAL12. Figure 6 shows the map, useful features, DNA sequence of the synthetic DNA insert, of the r3 expression vector, pAL13.

Figure 7A shows the synthetic fragments, either A-T or G-C rich, inserted into the EcoRI site of pAL13. Figure 7B shows the sequences upstream of the r3 promoter in pAL13 after insertion of an A-T rich synthetic fragment (r3AT), or a G-C rich synthetic fragment (r3GC) , or no DNA (r3) . Figure 7C shows the sequence of the r3 promoter used in the constructions.

Figure 7D shows the DNA sequences of pAL12 and pAL13 between the r3 promoter and the downstream ATG translation initiation codon of the CAT gene (boxed) . Sequences on each line are contiguous but are grouped to emphasize similarities and differences.

Figure 8A shows the synthetic DNA sequence containing the TRP operon and promoter which was inserted into pTRP233. Figure 8B shows some of the characteristics of pTRP233.

Figure 9 is a flow diagram for the construction of ρAS18 and pAS18-CAT, and pAS19 and ρAS19CAT.

Figure 10 shows the synthetic sequence which encodes acidic human fibroblast growth factor (haFGF) .

Detailed Description of the Invention

The practice of the present invention will employ, unless otherwise indicated, conventional molecular biology, microbiology, and recombinant DNA techniques within the skill of the art. Such techniques are explained fully in the literature. See, e.g., Maniatis, Fritsch & Sa brook, Molecular Cloning: A Laboratory Manual (1982); Miller, Experiments in Molecular Genetics (1972); DNA Cloning, Volumes I and II (D.N. Glover ed. 1985); Oliqonucleotide Synthesis (M.J. Gait ed. 1984); Nucleic Acid Hybridization (B.D. Hames and S.J. Higgins eds. 1985). TRANSCRIPTION AND TRANSLATION (B.D. Hames & S.J. Higgins eds. 1984); IMMOBILIZED CELLS AND ENZYMES (IRL Press, 1986); B. Perbal, A PRACTICAL GUIDE TO MOLECULAR CLONING (1984); and the series, METHODS IN ENZYMOLOGY (S. Colowick and N. Kaplan eds, Academic Press, Inc.).

In describing the present invention, the follow¬ ing terminology will be used in accordance with the definitions set out below.

A "replicon" is any genetic element (e.g., a plas id, a chromosome, a virus) that behaves as an autonomous unit of polynucleotide replication within a cell; i.e., capable of replication under its own control. A "vector" is a recombinant replicon in that another heterologous polynucleotide segment is attached, so as to bring about the replication and/or expression of the attached segment.

A "structural gene" is a polynucleotide sequence which is transcribed, and/or translated into a polypeptide, when placed under the control of appropriate regulatory sequences. The boundaries of the structural gene are determined by a translation start codon at the 5'-terminus and a translation stop codon at the 3'- terminus. A structural gene can include, but is not limited to: prokaryotic sequences; cDNA from eukaryotic mRNA; genomic DNA sequences from eukaryotic DNA, which lack intervening sequences; and even synthetic DNA sequences. A polyadenylation signal and transcription termination sequence will usually be located 3' to the coding sequence.

A "promoter sequence" is a DNA regulatory region capable of binding RNA polymerase and initiating transcription of a downstream (i.e., in the 3' direction) coding sequence. In the present invention, the 3'- terminus of the promoter sequence is upstream (i.e., in the 5' direction) of the sequence encoding the ribosomal binding site, and contains the minimum number of bases or elements necessary to initiate transcription at levels detectable above background. Within the promoter sequence will be found a transcription initiation site (conveniently determined by mapping with nuclease SI), as well as protein binding domains (-consensus sequences) responsible for the binding of RNA polymerase. Although promoter sequences are double stranded DNA molecules, they are described herein with reference to a single strand, using the normal convention of giving only the sequence in the 5' to 3' direction (left to right) along the nontranscribed strand of DNA (i.e., the strand having a sequence homologous to the transcribed RNA) .

A "double-stranded DNA molecule" refers to the polymeric form of deoxyribonucleotides (adenine or A, guanine or G, thymine or T, and cytosine or C, nucleotides) in its normal, double-stranded helix. This term refers only to the primary and secondary structure of the molecule, without reference to any particular tertiary structure. Thus, this term includes double-stranded DNA found, inter alia, in linear DNA molecules, viruses, piasmids, and chromosomes, both in vivo and in vitro. A "ribosomal binding site" is a nucleotide sequence in mRNA to which a ribosome binds, which causes the ribosome to recognize a specific AUG codon within the mRNA as a translation initiation codon. The ribosomal binding site is often called the "Shine-Dalgarno sequence" .

Sequences are "operably linked" when they func¬ tion together to give rise to the expression, including transcription and/or translation, of a single nucleotide sequence. "Transformation" is the insertion of an exogenous polynucleotide into a host cell. The exogenous polynucleotide may be maintained as a plas id, or alternatively, may be integrated within the host genome. A "clone" is a population of cells derived from a single cell. A "heterologous region or gene" within DNA is an identifiable segment of polynucleotide within the larger polynucleotide o.lecule that is not found in association with the larger molecule in nature. Thus, when the heterologous region encodes a mammalian gene, the heterologous region will be flanked by a nucleotide sequence that does not flank it in the genome of the source organism. Another example of a heterologous coding sequence is a construct where the coding sequence itself is not found in nature (e.g., a cDNA where the genomic coding sequence contains introns, or a synthetic sequence having one or more codons different than the native gene) . Allelic variations or naturally occurring mutational events do not give rise to a "heterologous region or gene. "

"Expression" denotes the in vivo process by which a polypeptide is produced from a gene. It involves transcription of the relevant gene into mRNA, and the translation of the mRNA into a polypeptide. "Regulatory sequences" are those sequences that function to control the transcription and translation of a gene. These include the promoter, signals for the initia¬ tion and termination of transcription, for ribosomal bind¬ ing, and for the initiation and termination of transla- tion, and operator sequences. A group of regulatory sequences "control" the expression of a coding sequence when the juxtaposition of the regulatory sequences relative to the coding sequence is such that the coding sequence will be expressed. An "operator" is a sequence located upstream, i.e., 5' to a structural gene encoding a polypeptide, and which regulates the expression of the polypeptide by controlling the level of transcription of that structural gene by binding small molecules which function as repres- sors or inducers. A "mutant" of the sequence encoding the r3 promoter refers to sequences which are derived from the original sequence by substitution of nucleotides and/or by deletion of portions of the original sequence. Mutant sequences may be made synthetically, or by mutagenesis of the original sequence. Promoter activity of mutants may be determined using methods which rely on promoter screen¬ ing vectors, as described below, and which are known to those of skill in the art. The present invention is based, in part, on the discovery of a new strong promoter for gram negative bacteria, the r3 promoter. Briefly, the scheme for isolating the r3 promoter from E. coli DNA was as follows. E. coli DNA was digested with a restriction enzyme, Rsal, and the restriction enzyme fragments were shotgun cloned into the S al site of a promoter cloning vector described in Brosius (1984), Gene 2__, 151. The Brosius vector contains the structural gene and ribosome binding site for chloramphenicol^' acyltransferase, but repiaces the promoter with.a series of unique cloning sites into which random fragments of DNA which potentially contain promoter sequences can be ligated. Insertion of a fragment containing promoter sequences confers chloramphenicol resistance upon the bacteria. In addition, to prevent destabilization of the vector by transcription into the region necessary for vector maintenance in the bacteria, the vector contains strong transcription terminators downstream from the structural gene. The salient features of the Brosius promoter cloning vector, pKK232-8, are shown in Figure 1.

Transformed clones of E. coli K-12 strain MC1061 which contained promoter sequences from the shotgunned

Rsal fragments were selected by their resistance to chloramphenicol (Camr) . The Camr clones contained plasmids with an inserted fragment of DNA. This was evidenced by the increased size of an EcoRI fragment in the plasmids obtained from sixteen randomly chosen Cam^r transformants.

Clones containing strong promoters derived from E. coli DNA were selected by growing transformants harbor¬ ing pKK232-8 derivatives containing the Rsal E. coli DNA fragments on plates containing a relatively high concentration of chloramphenicol, i.e., 500 micrograms/ml, as opposed to the 30 micrograms/ml used in the normal selection- process. Resistance to increased levels of chloramphenicol is the result of increased expression of chloramphenicol acetyltransferase, which in turn reflects the strength of the promoter. Thus, when the transformants harbored pKK232-8 derivatives containing the lac (pKKlac) and tac (pKKtac) promoters, these transformants were resistant to chloramphenicol at concentrations of 250 micrograms/ml and 500 micrograms/ml, respectively. One promoter-containing pKK Rsal-fragment derivative that conferred resistance to 500 micrograms/ml chloramphenicol was named pKKr3. Sizing of the EcoRI fragment in this plasmid indicated that the r3 promoter containing insert in pKKr3 was about 90 base pairs.

The r3 promoter is of greater strength than either the lac or the tac promoter. Assessment of the comparative strengths of the promoters was determined at the level of transcription of the chloramphenicol acetyltransferase gene, as reflected by the steady state level of the mRNA, and at the level of expression of the gene, as reflected by the synthesis of active enzyme. The experimental conditions and results are described infra, in the Examples, section Il.b.. Based upon the results, the r3 promoter is more than twice as strong as the tac promoter, and more than 20 times as strong as the lac promoter. The strength of the r3 promoter is independent of the strain of E. coli K-12 used as host. In addition, the r3 promoter is not sensitive to catabolite repression.

Comparison of the nucleotide sequences of a number of different promoters reveals two highly conserved regions, one located about lObp (-10 region) and the other about 35 bp (-35 region) upstream from the point at which transcription starts. These regions have sequences that vary but are related to a "consensus" sequence. The consensus sequences for the -35 region are TTGACa, and for the -10 region are TAtAaT. (The capital letters and small letters represent bases that occur at that position more than 54% and more than 39% of the time, respectively) . Hawley, D. et/al., Nucl. Acids Res. (1983) 1_1, 2237. The -35 and -10 regions are thought to be important in determining promoter strength because most mutations that increase promoter strength change nonconsensus bases to consensus bases; conversely, most mutations that decrease promoter strength change consensus bases to nonconsensus bases. The spacer region between the conser ed -35 and

-10 regions also contributes to promoter strength in that the number of nucleotides that separate the conserved sequences is important for efficient promoter function. For example, 16 to 19 nucleotides separate the -10 region from the -35 region; mutations altering the spacing between these two conserved regions in a lac promoter and in the -lactamase promoter change the "strength" of the promoter. Stefano, J.E. et al., Proc. Natl. Acad. Sci. USA (1982) 1 1069; Berman, M.L. et al., Proc. Natl. Acad. Sci. USA (1979) 1__, 4303; and Jaurin, B. et/al. Nature (1981) 290, 221. These results indicate either that there is an optimal spacing for all promoters, or that there is an optimal spacing for any individual promoter that is dependent on its particular DNA sequence. The nucleotide sequence of the EcoRI fragment that contains the r3 promoter was determined as described in the Examples section. The sequence containing the r3 promoter fragment and the vector sequences surrounding it that create the EcoRI fragment is shown in Figure 2.

Analysis of the fragment discloses the following features. First, using matches of 3 or more nucleotides to the consensus sequences described by Hawley et al. , supra, six -35^' regions and eight -10 regions were identified. These sequences are indicated in Figure 2 by solid lines with the -10 regions, underlined; and the -35 regions, overlined. Four of the consensus sequence pairs include the consensus base spacer distance of 15-19 nucleotides. In Figure 2, these pairs are numbered 1-4, and will be referred to hereinafter as promoters 1-4. A comparison of -the sequences and distances of these four consensus sequence pairs with the published consensus sequences and spacer distances is also shown in Figure 2.

The site at which transcription is initiated under the control of a promoter may- be determined by map¬ ping using a modification of the SI nuclease method of Berk and Sharp, supra. This method, in conjunction with sequencing of the probe used to detect the transcripts, was used to map the initiation of transcripts of the chloramphenicol acetyltransferase gene which were synthesized under the control of the r3 promoter. The results of the mapping indicate that 90% of the transcripts were initiated at a CAG codon, which is indicated in Figure 2 by three dots over the sequence. A minor fraction of the transcripts, approximately 10%, were initiated approximately 10 bases downstream from this site.

A feature of the r3 promoter is the presence of multiple overlapping consensus sequences, which may be seen in Figure 2. It has been observed that some, but not all, strong promoters contain multiple, often overlapping DNA sequences that match the -35 and -10 regions reason¬ ably well. Transcription analysis indicates that not all ' of these overlapping regions are used as stable RNA polymerase binding sites. Strong promoters containing these multiple overlapping promoter regions are P (Youderain, P. et al, supra) , and the ribosomal RNA promoters (see deBoer et al, supra) . However, strong promoters which do not have discernible multiple -10 and - 35 regions are trp (see Bennett et al, supra) and tac (see deBoer et al, Proc. Natl. Acad. Sci. USA, supra) . Since there are examples of strong promoters which contain these multiple overlapping -10 and -35 regions as well as strong promoters which apparently do not, the importance and relative contribution of such sequences to promoter strength is not clear.

Another feature of the r3 promoter is the pres¬ ence of an A-T rich region upstream towards the 5'-end of the promoter. It has been suggested that A-T rich sequences upstream from the -35 region contribute to the strength of the ribosomal RNA promoters. de Boer et al. Cell, supra; Young, R. et al, Cell (1979) 17, 225; Csordas-Toth, E. et al, supra♦ However, since a comparison of other strong promoters to some weak promot- ers does not reveal any greater degree of A-T richness in the upstream region of the strong promoters, the relative contribution of the upstream sequences to promoter strength is not clear.

Large regions of the naturally occurring r3 promoter region contained within the EcoRI fragment can be mutated without significantly affecting promoter activity. Mutations include nucleotide changes and/or deletions of the original sequence. Many methods for causing mutations of nucleotide sequences are known in the art and include but are not limited to the following. For example, muta- tions may be accomplished by synthesizing promoters using solid phase synthesis, and cloning them. Techniques for the solid phase synthesis of promoters and cloning of these synthetic promoters is known in the art. See, e.g., X. Soberon et al, in Promoters: Structure and Function, page 407; and P.L. de Haseth et al. (1983), Nucl. Acids Res. LI, 773. Mutations may also be accomplished by fragmentation using enzymes. In addition, mutations may be caused using techniques such as site directed mutagenesis, which are known to those of skill in the art. An example of site directed mutagenesis of promoters is given in J.J. Rossi et al. (1983), Proc. Natl. Acad. Sci. j3(), 3203.

After the alterations, mutants of the EcoRI fragment containing promoter activity are isolated using promoter cloning vectors, particularly the promoter clon¬ ing vector of Brosius, supra. A number of promoter clon¬ ing vectors which also may be used are known in the art. See, e.g., McKenney et al. in Gene Amplification and analysis. Volume 2 (J.G. Chirik^'jian and T.S. Papas, eds, Elsevier/North-Holland, N.Y., 1981); R.W. West Jr. et al. (1979), Gene 1_, 271; R.W. West Jr. and R.L. Rodriguez (1982), Gene 2_0_, 291; G. An and J.D. Friesen (1979), J. Bacteriology 140, 400; and H.A. de Boer (1984), Gene 3_0, 251. Moreover, the strength of the cloned altered promoter fragment may be assessed relative to that of the lac promoter and tac promoter, using the techniques described in the Experimental section.

Based upon the above described characteristics of the EcoRI fragment containing the r3 promoter, it is probable that certain features of the fragment are required to retain r3 promoter activity. These features are the following.

First, portions of the fragment extending from the 5'-end of the -35 region of promoter 2 through the CAG transcription initiation site are probably required. Based upon the analysis of start sites for r3-directed transcription, discussed infra in Section II.c, RNA polymerase appears to be using the most optimal promoter, promoter 2, to produce most of the transcripts from the r3 promoter region. The sequences of the -10 and -35 regions of promoter 2, with 3/6 and 5/6 matches to the most conserved promoter consensus bases, respectively, match closer to the conserved sequences than the other three promoter sequences in the EcoRI fragment. These sequences should receive minimal alteration, and based upon the consensus sequence data of Hawley et al., supra, changes within them should be restricted to substitution of the sixth nucleotide in the -35 sequence, and the fifth nucleotide in the -10 sequence.

Promoter 2 also has a 17 base spacer region, which, as discussed above, is considered optimal for most promoters. The base spacer distance should be conserved, or if altered, the distance should be 16 or 18 bases. The CAG major transcription initiation site is 6-8 bases downstream from the -10 region of this promoter. This distance between the -10 region and the transcription initiation site is within the consensus distance of 4-8 nucleotides published by Hawley et al., supra. The EcoRI fragment also contains an transcriptional initiation codon, CAG. Transcription in E. coli is initiated almost exclusively with purine nucleoside triphosphates; moreover, in almost every promoter examined, if a residue signaling a purine start is located either 6 or 7 base pairs from the -10 region, that residue is the point of initiation. P.H. von Hippel et al. (1984), Ann. Rev. Biochem. 5_3, 389. Hence, substitution within the initiation codon should maintain a pyrimidine, either C or T at either position one or two of the initiation codon. The nucleotide sequence downstream of the transcriptional initiation site may contribute somewhat to the strength of the promoter, but. it is unlikely that this sequence is essential to promoter activity. Moreover, since the -10 region of promoter 4 is also beyond the major transcriptional start site, it is also unlikely that the consensus sequences of promoter 4 are essential to the r3 promoter.

The r3 promoter region described herein is employed in expression vectors, and is used to regulate the expression of a heterologous protein in a suitable host. Suitable hosts are gram negative bacteria, examples of which include Escherichia coli, Shiqella, Klebsiella, and Citrobacter, and particular strains of which include Pseudomonas aeruginosa, Serratia marcesans, Pseudomonas putida, and Salmonella typimurium.

One method of using the disclosed r3 promoter is to substitute it for an existing promoter region in a known expression vector which expresses a protein in any of the above listed hosts. Of particular interest are expression vectors which contain weak promoters, when a strong promoter to control expression is desired. These expression vectors containing weak promoters are known in the art. The construction of the parental vector, pALr3, was accomplished by the substitution of the r3 promoter for the tac promoter in the vector pKK 223-3. See the Examples section.

The r3 promoter may also be used to construct new expression vectors. To accomplish this, it is convenient to construct the r3 promoter and appropriate regulatory sequences within an "expression construct"; i.e., a DNA segment comprised of the various regulatory factors required for expression of a heterologous polypeptide. The expression construct may be converted to an expression cassette by modifying regions upstream and downstream (i.e., 5'- to and 3'-to, respectively) to the construct so that they contain restriction sites which allow the cassette to be removed -from one vector and inserted into another vector. These sites may be inserted or constructed by techniques known to those of skill in the art, for example, by site directed mutagenesis.

The expression construct is assembled employing standard recombinant techniques. A basic expression construct comprises, in a 5' to 3' direction on the nontranscribed strand, the r3 promoter which includes the transcription initiation site, a ribosome binding site, a cloning site for a heterologous coding sequence (at least one restriction site), and a transcription termination sequence. The cloning site is preferably comprised of two restriction sites for different endonucleases. This facilitates the orientation of the heterologous coding sequence upon insertion.

An expression construct containing the basic components is contained in the expression vector, pAL 10, the construction of which is described in the Experimental section, and the useful characteristics of which are shown in Figure 4. pAL 10 was constructed by the insertion of a synthetic DNA sequence into the vector pALr3; this sequence encodes a ribosomal binding site, and a transla- tion initiation codon, the latter of which is encoded within an Ncol site. Adjacent to the Ncol site is another restriction site for Smal. If the sequence encoding a peptide or polypeptide which is to be expressed contains its own translation initiation codon, it is inserted into the Ncol-Hindlll digested vector. However, if the sequence lacks a translation initiation codon, it may be inserted into Smal-HinduI vector; translation then progresses from the start codon encoded within the Ncol site. A comparison of the expression capability directed by the r3 promoter and the tac promoter was measured using analogous constructs, both of which contained a sequence encoding an apolipoprotein fragment. This comparison, which is discussed in the Examples sec¬ tion, indicated that approximately twice as much apolipoprotein Al was expressed under the control of the r3 promoter as was expressed under the control of the tac promoter. Expression constructs which yield increased levels of expression relative to that obtained with the basic expression construct may also be assembled. Increased expression may be achieved by increasing the strength of the ribosomal binding site. The strength of ribosome binding is dependent both upon the sequence within the Shine-Dalgarno region (Shine, J., et al. , supra) , and the distance of this region to the ATG initia¬ tion codon. Generally, the optimal distance for various genes is from 7 to 10 nucleotides. Moreover, the sequences in this spacer region are generally A-T rich. See Harris, T.J.R., in Genetic Engineering, supra. Examples of expression constructs containing the r3 promoter and strong ribosomal binding sequences, which give rise to increased expression levels, are contained in the vectors pAL12 and pAL13, the constructions of which are described in the Examples section. It should be noted that a feature which is present in pAL12 which is lacking in pAL13 is the presence of a restriction site for the insertion of operator and repressor sequences. The level of expression may also be increased by the insertion of an A-T rich region upstream from the r3 promoter, although the r3 promoter already includes an A-T rich region. An example of an expression construct which includes an additional A-T rich region is contained in pAL13-CAT-AT, the construction of which is described in the Example section. Insertion of this A-T rich sequence increases the level of expression about 2.2 fold higher than that of the _.parental vector.- The increased expres¬ sion is the result of increased promoter activity. In the cases of high level expression of a heterologous polypeptide, the product polypeptide may be toxic to the host harboring the recombinant vector. In such cases, constitutive expression of the product would lead to death of the host cell. One way of overcoming this problem is to employ transcription promoters which are regulatable, for example, the lac promoter. Maizels, N. , supra.

The expression of polypeptides under the direc¬ tion of the r3 promoter is constitutive in the above described vectors. However, r3 promoter activity may be rendered sensitive to regulation by the insertion of an appropriate operator sequence in the expression construct. Operator sequences which may be used to regulate the expressipn of a heterologous protein -whose transcription is directed by r3 are known to those of skill in the art, some of which have been reviewed in Brosius, supra. An example of an expression construct in which the r3 promoter is regulated by catabolite repression as a result of the insertion of DNA sequence encoding the lac operator, is present in pAS19.

The expression constructs described above are useful for the creation of vectors for the expression of a variety of heterologous genes encoding polypeptides, e.g., interferons, interleukins, tumor necrosis factor, apolipoproteins, growth factors, hormones, enzymes, etc. If the heterologous gene does not contain a sequence which matches a restriction enzyme site in the construct in the vector, the desired sites may be constructed using techniques known to those of skill in the art, e.g., by site directed mutagenesis. For example, an Ndel sites may be constructed on heterologous genes by site directed mutagenesis, see infra in the Examples sec¬ tion.

After a vector is assembled so that it contains the desired expression construct containing the heterologous gene to be expressed, a suitable host may be transformed with the vector by any known means. Suitable hosts include the gram negative bacteria described supra.

Methods of growing transformed gram-negative

10 bacteria to express foreign proteins are known in the art. The method chosen will include conditions to allow the regulation of the r3 promoter, when regulatory sequences such as operators are present in the expression construct. Harvesting and isolation of the heterologous

15 polypeptide product may be by any convenient means, including but not limited^' to centrifugation, chromatography, electrophoresis, dialysis, and extraction. The selection of the appropriate recovery technique will depend on the nature of the protein, and is within the

²⁰ skill of the art.

Described below are examples of the present invention which are provided only for illustrative purposes, and not to limit the scope of the present inven¬ tion. In light of the present disclosure, numerous

²⁵ embodiments within the scope of the claims will be appar¬ ent to those of ordinary skill in the art.

Examples

- - I. Materials and Methods

Enzymes were purchased from commercial sources, and used according to the.manufacturers ' directions.

Radionucleotides and nitrocellulose filters were also purchased from commercial sources. 5 Most of the techniques used to manipulate DNA for the construction of genes and vectors, and for the analysis of RNA and protein products are known within the art. See, e.g., Maniatis, Fritsch & Sambrook, Molecular Cloning: A Laboratory Manual (1982); Miller, Experiments in Molecular Genetics (1972); DNA Cloning, Volumes I and II (D.N. Glover ed. 1985); Oligonucleotide Synthesis (M.J. Gait ed. 1984); Nucleic Acid Hybridization (B.D. Hames and S.J. Higgins eds. 1985). Transcription and Translation (B.D. Hames & S.J. Higgins eds. 1984); Immobilized Cells and Enzymes (IRL Press, 1986); B. Perbal, A Practical Guide to Molecular Cloning (1984); and the series, Methods in Enzymology (S. Colowick and N. Kaplan eds, Academic Press, Inc.). These methods are described in for convenience.

Bacterial host strains are the following. For plasmid production and expression: E. coli K-12 strain MC 1061 (Casadaban, M. , et al. (1980), J. Mol. Biol. 138, 179; E. coli K-12 strain DH1 (Hanahan, D (1983), J. Mol. Biol. 166, 557. For expression: E. coli k-12 strain JA221 (Nakamura, K., et al. (1982), J. Mol. Appl. Genet. I , 289. For growth of M-13: E. coli K-12 strain JM 103 (Messing J., et a. (1982), Nucl. Acids Res. 9_, 309).

Transforma ion was by the method of Cohen, S.N., et al., as detailed in Maniatis, T., Molecular Cloning, supra. The method of Hanahan, D. (1983), J. Mol. Biol. 106, 557, was used for transformation by recombinant M-13 phage.

Plasmids were prepared as follows. For analysis, from 1.5 ml of overnight culture by the boiling method of Holmes, D.S.; and for large scale plasmid preparations, by the SDS lysis method and centrifugation through cesium chloride. Both methods are described in Maniatis et al., supra. Synthetic oligonucleotides were synthesized on a SAM I DNA Synthesizer (Biosearch, San Rafael), using the manufacturers' directions.

Radiolabeling of probes was accomplished by kinasing with [ 32P]-gamma-ATP at specific radioactivity of

3000 Ci/mM) . .

DNA sequencing was by the M-13 chain termination method as described in Sanger and in Messing (1981), in M13mp8, as described in Messing (1982). Sanger, F. (1980), J. Mol. Biol. 143, 161; Messing, J. , et al. (1981), Nucl. Acids Res. 9_, 309; Messing, J., et al. (1982) , Gene _19_, 269.

Site-specific oligonucleotide mutagenesis was as described by Zoller, M. , et al. (1982), Nucl. Acids Res. i , 6487.

Transcriptional start sites were analyzed by the SI nuclease method described by Berk, A.J., et al. (1978), Cell j^, 72i. Steady state levels of mRNA were analyzed by the "dot blot" technique. Analysis of proteins expressed was by the dis¬ continuous gel system of Laemmli, U.K. (1970), Nature 227, 680. More specifically, cells grown in rich medium were harvested, resuspended in the Laemmli gel buffer, boiled for 3-5 minutes, and electrophoresed. Typically, the cells from one o of culture at O.D.„₀=0.5 were resuspended in 100 microliters of Laemmli buffer, and 1-10 microliter aliquots applied to the gel. The proteins were visualized by staining the gel with Coomassie blue. The pattern of the proteins arising from host cells harboring a vector that does not express the heterologous protein was compared to the protein pattern from the cells harbor¬ ing the vector that did express the heterologous protein. Expression that yielded a steady state level of 1% or more of the total protein yielded a new band. Alternatively, if the level of expression was below that which yielded a steady state level of 1%, or if it was desired to confirm the results from the gel analysis, analysis was by immunoprecipitation using the cognate antiserum to the expressed protein. In this case, the proteins are labelled with a radioactive amino acid. The cells are grown in minimal medium, a labeled amino acid is added, and growth is allowed to continue for 30 sec to 30 minutes. The cells are instantly lysed, and the protein precipitated by the addition of trichloracetic acid.

II.a. Isolation of the r3 promoter

Commercially available E. coli DNA, after diges- tion with Rsal, was ligated with Smal digested pKK232-8. pKK232-8 is a promoter-cloning vector described in Brosius (1984), Gene 27_, 151; a map of pKK232-8 is shown in Figure i. Insertion of a promoter into the Smal site of this vector allows the expression of chloramphenicol acetyltransferase, thus rendering the transformed bacteria chloramphenicol resistant. Rsal and Smal are both blunt cutting enzymes. More specifically, digested vector (20ng) was ligated with E. coli DNA (50ng) in a 20 micro- liter reaction. E. coli k-12, strain MC1061 (MC1061) was transformed with the ligation mixture and plated onto L plates containing 30 micrograms/ml chloramphenicol.

Confirmation that chloramphenicol-resistant bacteria harbored pKK232-8 with inserts into the Smal site was obtained as follows. Plasmids were isolated from 16 randomly picked transformants using the boiling method.

The isolated plasmids were digested with EcoRI. The digested DNA was labelled with all four [alpha- 32P]- deoxyribonucleoside triphosphates, using DNA polymerase I,

Klenow fragment. The digested DNA was subjected to electrophoresis on 5% polyacrylamide gels. It was determined from the autoradiographs of the dried gels that only an EcoRI fragment varied in size. Moreover, in all cases, the EcoRI fragment from plasmids that conferred resistance to chloramphenicol was larger than the fragment from the vector alone.

A plasmid with an insert containing a strong promoter was isolated based upon resistance to increased levels of chloramphenicol. Bacteria harboring pKK232-8 with a lac promoter insert (pKKlac), and with a tac - promoter insert (pKKtac), are resistant to 250 micrograms/ ml and 500 micrograms/ml chloramphenicol, respectively. . Bacteria harboring pKK232-8 with E. coli DNA inserts in the Smal site of the vector, were grown in chloramphenicol at 500 micrograms/ml. One promoter-containing pKK232-8 derivative vector that conferred resistance to this level of chloramphenicol was named pKKr3. pKKr3 was digested with EcoRI, and the fragment sizes determined by gel electrophoresis. Based upon its migration in the gels, the size of the r3 insert was about 90 b.p.

II.b. Comparison of r3 promoter to the lac and tac promoters

Il.b.l. Construction of pKKlac and pKKtac

The lac and tac promoters were cloned into the Smal site of pKK232-8 as follows. The lac promoter (Maizels, supra) was isolated as a 250 b.p. Haell-HincII fragment from pUC8 as described by Viera, J. and J. Mess- ing (1982), Gene 19, 259. The tac promoter (deBoer H., et al. , supra) was isolated as a 250 b.p. Hindlll-EcoRI frag¬ ment from ptacll, as described by Amman, E., et al. (1983), Gene 25, 167. Both fragments were blunted with T4 DNA polymerase, Klenow fragment, and deoxynucleoside triphosphates, as described in Maniatis, supra. 50 ng of fragment was ligated into 20 ng of Smal-cut vector. E. coli MC1061 was transformed with the ligation mixture, and plated onto L plates containing 30 micrograms/ml chloramphenicol.

II.b.2. Expression of chloramphenicol acetyltransferase (CAT) from pKKr3 as compared to pKKlac and pKKtac

The purpose of this experiment was to compare the promoter strengths of pKKr3 to that of pKKlac, and pKKtac. The effect of E. coli strain harboring the vec¬ tors was also examined. In addition, the effect of catabolite repression on the expression of CAT as directed by the r3 promoter, compared to the lac and tac promoters, was also examined. Three strains of E. coli K-12, i.e., MC 1061,

JA221, and DH1, were used. Each strain was transformed with each of the vectors, pKKr3, pKKlac, and pKKtac. The transformed -strains were grown in L broth, or in minimal medium with glucose or with glycerol. Isopropyl—D- thiolgalactoside (IPTG) was added at ImM to cultures of

JA221 and DH1 bearing pKKlac and pKKtac. Equal numbers of cells were harvested, and one O.D.-,- of cells in a final volume of 1ml was lysed as described in Talmadge et al. (1980), Proc. Natl. Acad. Sci. USA _ \, 3369. The level of CAT activity in each cell extract was determined by the method of Shaw, W.V. (1975), J. Biol. Chem. 242, 687, tak¬ ing two or three time points in the linear range of the assay. More specifically, cell lysate was diluted 1:10 or 1:100 in 0.14 M Tris-HCl, pH 8; the dilution was dependent upon the level of activity. For each of three time points, diluted cell extract was added to 465 microliters 0.14 M Tris-HCl, pH 8, containing lOmM S-acetylCoA, and incubated at 37°C for five min. The reaction was started by the addition of 6 microCuries 14C-chloramphenicol. At the appropriate times, 180 microliter aliquots were withdrawn and immediately extracted with 1 ml ethyl acetate. The ethyl acetate phase was dried, resuspended in 10 microliters. ethyl acetate, and subjected to thin layer chromatography on silica gel coated aluminum plates, using as solvent, chlorofor :methanol (95:5). The start¬ ing and converted materials were detected by autoradiography of the dried plates. Radioactivity present in the samples was determined by scintillation counting. The units of activity were calculated as nM of conversion per minute per mg of cellular protein.

From the results, summarized in Table 1, it can be seen that when r3 is the promoter, the levels of CAT expressed are 2-fold and 20-fold -higher than when tac and lac, respectively, are the promoters. The results also show that the level of promoter activity is independent of the strain. Also, the levels of r3 and tac promoter activity are essentially independent of the composition of the medium. This is in contract to lac promoter activity, which is known to be catabolite repressed. Magazanik, B. (1962), Cold Spring Harbor Symposium on Quantitative Biol¬ ogy 2_6' 249.

Data in the literature supports the results obtained above. Schottel et al., using an independent construction in which the expression of the CAT structural gene was directed by the lac promoter, found that MC1061 transformants harboring this construct expressed 420 units of CAT activity when the cells were grown in rich medium. Schottel et al. (1984), Gene 2_8, 177. Schottel's results are very close to the average value of 500 units in rich or glucose minimal medium shown in Table 1. deBoer et al. , supra, compared the ability of lac and tac to drive expression of the gene for galactokinase when transformants harboring each construct were grown in rich medium. deBoer's results showed that the levels of activ- ity were 8- to 11-times higher with the tac promoter.

Table 1 shows the same level of difference in lac and tac activities (8-fold in rich or glucose minimal media).

The comparison of the literature results with the results in Table 1 allow the conclusion that r3 is about twice as strong as the very strong promoter, tac. II.b.3. Comparison of r3 promoter strength to lac and tac as determined by analysis of steady state levels of CAT mRNA

The steady state levels of CAT mRNA synthesized in MC1061 harboring pKKr3 and pKKtac were determined by dot blot technique, using as a probe, the EcoRI-Ndel frag¬ ment of pKK232-8 which encodes^" most of the CAT structural gene.

Total RNA was isolated using the guanidinium isothiocyanate method. More specifically, bacteria were harvested and 1 ml at O.D.__cn⁼0.5 was resuspended in 500 microliters of 6 M guanidinium isothiocyanate, 5mM sodium citrate, 0.1 M -mercaptoethanol, 0.5% sarkosyl. The sarkosyl was prepared as described in Maniatis, supra. , and incubated at 90 C for a few minutes prior to use. The lysates were further purified by extraction with phenol:chloroform, and chloroform , and then precipitated with ethanol. The precipitates were resuspended in diethylpyrocarbonate treated water. To determine the amount of hybridizing RNA transcribed from each vector, 5 microliter spots contain¬ ing 800ng, 400ng, 200ng, lOOng, 50ng, 25ng, and 12.5ng of total RNA from each sample were spotted onto a nitro¬ cellulose filter. The filter was probed with the EcoRI- Ndel fragment of pKK232-8, which had been radioactively labeled by the nick translation method of Rigby, P.W.J., et al., J. Mol. Biol. 113, 237. The hybridization condi¬ tions were as follows. Hybridization buffer contained 0.75M NaCl, 0.75M sodium nitrate, 40% formamide, 0.05%SDS, 0.2% bovine serum albumin, 0.02% polyvinyl pyrollidone,

0.1% sodium pyrophosphate, and 20 mg/ml denatured sheared salmon sperm DNA. The filters were incubated with the hybridization buffer for 1 hour at 42°C; probe was added at 5 x 10 5 cpm 32P-labeled probe per ml of hybridization buffer containing, in addition 1% dextran sulfate. The filter was incubated at 42°C for 16 hours, and then washed. Washing was 4 times at room temperature for 5 minutes, followed by 2 times at 5.0 C for 15 minutes. Washing buffer contained 0.3 M NaCl, 0.3 M sodium nitrate, and 0.05% SDS.

After drying and autoradiography, the filters were rewashed with wash buffer for 2 hours at 65 C and hybridized as described to a nick-translated control probe. The control probe is the small PvuII-Ndel fragment of pBR322, a region of pBR322 present on pKK232-8 that is not transcribed. The filters hybridized to the control probe were then washed and autoradiographed, as described above.

To determine radioactivity in the transcripts, both the sample and control autoradiographs were scanned by densitometry. The radioactivity created by hybridiza¬ tion to the control was subtracted from that in the r3 and tac samples. The results showed that the level of RNA from the r3 promoter was twice the level of RNA from the ^tac promoter.

The results concerning promoter strength obtained from an analysis of the CAT mRNA levels support and extend those obtained from an analysis of CAT gene expression directed by r3 as compared to that directed by tac, as shown in section II.b.2. I.e., the increased expression of CAT in transformants harboring pKKr3 relative to that in transformants harboring pKKtac resulted from increased transcription directed by the r3 promoter.

II.c. Analysis of start sites for r3 directed transcrip¬ tion

Transcriptional start sites were mapped by the SI nuclease method of Berk and Sharp, supra, as modified by Weaver and Weissman, supra. The probe was a derivative of a 5'-32P-labeled Pstl-BamHI fragment of pKKr3, which contains the r3 promoter and the start of the CAT gene.

More specifically, total nucleic acids were extracted from MC 1061 harboring either pKKr3, or pKK232- 8, the parental vector. Isolation of the RNA was as described in Section II.b.

The probe was prepared by labeling 10 micrograms of the Pstl-BamHI fragment in a 50 microliter reaction mixture containing [gamma- 32P] adenosine triphosphate. After phenol extraction, the fragment was digested with EcoRI, and isolated by electrophoresis on a 5% polyacrylamide gel and resuspended in 50 microliters 1/10 TE. The resulting probe is labeled only at the BamHI end. Hybridization of. the probe with the samples was as follows. 150 microliters of the ethanol precipitate containing total nucleic acid was centrifuged. Three microliters of kinased probe was added to the pellet, which was dried under vacuum. Probe (1 microliter) without cell nucleic acids was also dried for a probe control. The samples were resuspended in 100 microliters deionized formamide to which 25 microliters of 5x hybridization buffer (0.2 M PIPES, pH 6.4, 5 mM EDTA, pH 8.0, 2 M NaCl) was added. Each sample was divided into 3 40 microliter fractions. All the samples were held at 60 C for 10 minutes, then one-third of each sample was transferred to 37 C, 42 C, or 47 C, for 4 hours. The probe control was transferred to 37°C

The samples were then digested with SI nuclease. More specifically, -the samples which had been incubated at the various temperatures, after the 4 hour incubation, were diluted rapidly into ice-cold SI nuclease buffer (0.28 M NaCl, 0.05 M sodium acetate, pH 4.6, 4.5 mM ZnSO., 20 micrograms/ml single stranded DNA) and held in an ice bath. One unit of SI nuclease was added to each sample; the samples were incubated at 30 C for 90 minutes, and then ethanol precipitated. Analysis was by electrophoresis on an 8% polyacrylamide-8M urea 40 cm gel; size markers were run on the same- gel.

Autoradiographs of the dried gel showed the fol- lowing. When the probe hybridized with RNA from pKKr3, more than 90% of the radioactivity was was in four bands, with the second -of the four the most intense. The bands spanned about 32-35 nucleotides in length on the sizing gel, and reflect the major transcript or transcripts. A minor transcript or transcripts, which, accounted for less than 10% of the radioactivity, was also detected as 4 minor bands about 10 nucleotides smaller than the major bands.

It should be noted that there are limitations to the SI technique. As discussed in Weaver et al., supra. , multiple, closely spaced, products are usually seen in an SI analysis. There are two interpretations of the events leading to these products. First, there are multiple starts at adjacent nucleotides. Second, there is a unique st^*art, and the other bands are generated either by in¬ complete or excessive SI digestion. If there is a clearly prominent band, that band is usually interpreted as the unique start site.

Three factors make the size of the SI product from 2-4 bases smaller than it appears when sized by electrophoresis with a Maxam-Gilbert sequence of the si probe. See Brosius, J. , et al. (1982), J. Biol. Chem. 257, 9205. First, SI nuclease products contain phosphates only at the 5-termini, while the DNA fragments generated by chemical sequencing have phosphates at both the 5'- and 3'-termini. This displaces the si products one-half base higher (making them appear larger) than the DNA fragments of the sequencing ladder. Second, during chemical cleav¬ age, the base that is attacked is removed, leaving the DNA fragment one base smaller. Third, due to steric hindrance of the SI nuclease by the triphosphate at the 5'-end of bacterial mRNA, the DNA probe will be left 1-3 bases too long. Thus, an SI analysis will usually reveal the transcription start site to a region of 1-4 bases smaller than the base indicated by electrophoresis.

The significance of these major and minor start sites will be discussed infra, in Section II.d. In the control samples, i.e., the probe alone, and RNA extracted from pKK232-8, bands smaller than the probe were not detected.

To more precisely identify the nucleotides around these major and minor transcripts, the kinased probe was sequenced by the method of Maxam and Gilbert. This sequence will be discussed infra, in section Il.d.

Il.d. Sequence of the r3 promoter

The nucleotide sequence of the EcoRI fragment "that contains the r3 promoter was determined by the chain termination method of Sanger et al., supra. More specifically, 50 ng of the fragment was ligated into EcoRI-digested, phosphatased, M13mp8 (20ng). The re¬ combinant M13 phage were transformed into E. coli K-12, strain JM103. The orientation of the insert in the re¬ combinant phage isolated from six white plaques was determined; one plaque of each orientation was sequenced by the chain termination method of Sanger et al., supra.

The nucleotide sequence of the r3 promoter frag¬ ment and the vector sequences surrounding it that create the EcoRI-BamHI fragment used to construct the various r3 vectors, is shown in Figure 2. In the figure, matches of . 3 or more nucleotides to the consensus sequences of Hawley et al., supra. , are indicated by solid lines (-10 regions, underlined; -35 regions, σverlined. Six -35 regions and 8 -10 regions were identified. Tabulating the distances between each -10 and -35 region, four pairs have the consensus 15-19 base spacer distance. These pairs are numbered 1-4 in Figure 2. In Figure 2, the sequences in these regions, and their spacer distances, are compared shown and compared to the Hawley consensus sequences and spacer distances.

The analysis by sequencing and matching to the Hawley et al. consensus sequences showed that the fragment containing the r3 promoter contained multiple, overlap¬ ping, -10 and -35 regions. As discussed supra. , in some cases the overlapping of consensus sequences has been indicative of promoter strength.

RNA polymerase appears to be using the most optimal promoter, promoter 2, to produce most of the transcripts from the r3 region. This is concluded from the results of sequencing, analysis of the consensus. sequences, and the SI analysis described in Section II.c. First, the sequences of the -10 and -35 regions of promoter 2, with 3/6 and 5/6 matches to the most conserved prOmoter consensus bases, respectively, are closest among 'the 4 promoters to. the consensus sequences. Promoter 2 also has a 17 base spacer region, which is considered to be optimal for most promoters, as discussed supra.

Second, a comparison of the nucleotide sequence of the kinased probe with the SI products of the RNA from pKKr3 transformed MC1031, analyzed as described in II.c, led to the prediction that the major start for transcrip¬ tion centered around the CAG trinucleotide that is 6-8 bases downstream from the -10 region of promoter 2; this CAG start site is indicated in Figure 2 by three solid circles. This start is within the consensus distance of 4-8 nucleotides from the last base of the -10 region to the transcription initiation site, as observed by Hawley et al. , supra. The initiating codon, CAT, is also close to the consensus CAG for initiation of transcription. Based upon the above described analyses, it was also concluded that RNA polymerase was using a less optimal promoter,, probably promoter 3, to produce the minor transcrip (s) . The fragment containing the r3 promoter contains an uninterrupted A-T sequence from -48 to -57 (+1 being defined as the transcriptional start of promoter 2). This A-T sequence is upstream of an Rsal site within the frag¬ ment, which is extends between 6-10 bases upstream of the start of the -35 region of promoter 2. Since the fragment containing the r3 promoter was isolated from E. coli DNA that had been digested with Rsal, the fragment containing the r3 promoter theoretically should not contain an Rsal site. However, the site could have arisen if the diges- tion with Rsal were incomplete, or alternatively, if two separate DNA fragments were ligated together in the construction of the promoter-containing pKK 232-8 library, as described supra. Whichever occurred, the second Rsal fragment creates the sequences which extend from -42 to - 59, upstream of promoter 2.

As discussed above, A-T rich sequences that are generally centered around -55, but can extend from -40 to -100, may be a feature of strong promoters. Rosenberg et al., supra. , and deBoer et al., supra. Hence, the A-T rich sequence in the fragment containing the r3 promoter may contribute to promoter strength.

Ill. Construction of expression vectors

III.a. Construction of the parental vector, pALr3

The scheme by which pALr3 was constructed is shown in Figure 3. pKK223-3 (Brosius, J., et al. (1984), Proc. Natl. Acad. Sci. USA 8_1, 6929) was digested with EcoRI and PvuII, and filled in with Klenow in the presence of the four deoxynucleoside triphosphates. 20 ng of treated vector was incubated for 30 minutes at 56 C with 1 unit of bacterial alkaline phosphatase. The phosphatased vector was ligated in a 100 microliter reaction volume to promote recircularize the blunt-ended vector. Plasmid from six MC1061 transformants were prepared by the boiling method, and screened for size compared to pKK223-2. Plasmids smaller than pKK223-2 were digested with EcoRI and PvuII. All plasmids had an EcoRI site but had lost the PvuII site; loss of the PvuII site occurs when a blunted EcoRI site is ligated to a PvuII site. One plasmid, called pALl, was digested with EcoRI and BamHI, and phosphatased. Treated vector (20 ng) was ligated in a 20 microliter reaction volume to the EcoRI-BamHI r3 promoter-containing fragment (50 ng) isolated from pKKr3. Insertion of the r3 promoter fragment into the plasmids of MC1061 transformants was analyzed by digestion of plasmids isolated from 1 ml of saturated culture by the method of Holmes et. al. , supra. , with EcoRI and BamHI. One of the inserted plasmids, containing r3, was named pALr3. This vector contains the r3 promoter oriented toward the polylinker region and the strong 5S RNA transcriptional terminators, which are described by Brosius et al. (1984), supra.

Ill.b. Construction of pALlO pALlO was constructed by the insertion of a synthetic DNA fragment into pALr3; the synthetic DNA frag¬ ment encoded a ribosomal binding site, a translational start codon encoded within an Ncol site, and two ad- ditional restriction enzyme sites, one for StuI, and one for Smal. The sequence and features of the synthetic DNA insert are shown in Figure 4. The details of the construction of pALlO are as follows. Two oligonucleotides were synthesized using an automated DNA synthesizer as directed by the manufacturer. The sequences of the oligonucleotides were:

1) 5'- GGCCTTAAGGAGG-3' and

2) 5'-CCCGGGCCATGGGTTAAACCTCC-3' . The oligonucleotides were kinased and ethanol- precipitated; the redissolved oligonucleotides were then hybridized and extended into a blunt-ended synthetic DNA fragment by the metho^'d of Rossi, J.J. (1982), J. Biol.

Che . 257, 9226. More specifically, each oligonucleotide was mixed to a final volume of 40 microliters in distilled water (1.5 micromolar) , boiled for 3 minutes, and held at 15 C for 30 minutes. 6 microliters of a solution contain- ing lOOmM Tris-HCl, pH 7.5, 0.6 M NaCl, 60 mM DTT, and 60 mM MgCl^ was added, and the mixture brought to 60 micro¬ liters by the addition of all four deoxyribonucleoside triphosphates in a final concentration of 60 micrograms/ml per nucleoside triphosphate. The mixture was incubated at room temperature for 30 minutes with 1 unit of T4 DNA polymerase I, Klenow fragment. The sequence of the resulting filled in synthetic DNA insert is shown in Figure 4. pALr3 (20 ng) was digested with BamHI, filled in, phosphatased, and ligated (20 ng) to the synthetic DNA insert (25 ng) . MC1061 transformants were screened for the addition of the 30 b.p. piece by determining whether the plasmids contained an StuI site. The plasmids were digested with StuI, and analyzed on a 0.7% agarose gel. The orientation of the insert was determined by analyzing the size of the fragment generated by EcoRI-StuI diges¬ tion. Candidates were sequenced by the chain termination method. A plasmid containing the synthetic DNA fragment in the correct orientation was named pALlO. The features of pALlO are shown in Figure 4. III.c. Construction of r3 vectors for high levels of expression

III.c.l. Construction of vectors with strong synthetic ribosomal binding sites, pAL12 and pAL13

Two r3 vectors, each of which contains a strong synthetic ribosome binding site (RBS), but which differ in the positioning of a PvuII site relative to the RBS, were constructed. In pAL12, the PvuII site is 5'- to the RBS, while in pAL13, the PvuII site is 3'- to the RBS. The maps and useful features of these vectors are shown in Figure 6 and in Figure 7 for pAL12 and pAL13, respectively. Construction of the vectors was as follows. Two pairs of complementary oligonucleotides were synthesized by solid phase DNA synthesis, as described above; their sequences are shown in Figs. 6 and 7. Each pair of oligonucleotides was kinased, and hybridized, as described in Section Ill.b. The hybridization created a synthetic DNA fragment with BamHI 5'- overhanging ends. pALr3 was digested with BamHI and phosphatased.^' Each synthetic BamHI fragment (25ng) was ligated to BamHI- treated pKKr3 (20 ng) . MC1061 was transformed with the ligation mixtures . Plasmids isolated from the transformants were digested with EcoRI and Ndel. These digests were compared to the digests of pALr3 to determine both the presence and orientation of the inserted DNA. Candidates were sequenced by the chain termination method of Sanger et al. , supra, by cloning the EcoRI-Hindlll fragment into EcoRI-Hindlll digested M13mp8, as described above. A candidate representing each inserted fragment in the appropriate orientation was selected. These were named pAL12 and pAL13. Ill.c.l.a. Comparison of expression of CAT using the strong ribosomal binding sites in pAL 12 and pAL 13, and the natural ribosomal binding site in pKKr3

Ill.c.l.a.l. Construction of a modified CAT gene

Insertion of the CAT gene into the fragments encoding the strong ribosomal binding sites in pAL12 and pAL 13 required the presence of an Ndel restriction site encompassing the translational initiation site. The construction of the Ndel site was accomplished by site directed mutagenesis, which generated two mutations. This was accomplished as follows. The CAT gene (25 ng) , including its RBS, was isolated from pKK 232-8 as a 700 b.p. TaqI fragment according to Alton, N.K., et al. (1978), Nature 282, 864. The TaqI fragment was cloned into Sail-digested M13mp8 (20ng) according to Messing et al., Gene, supra. White plaques were screened for orientation, and the DNA from one of the recombinant phage with the hybridizing orientation was hybridized to the mutagenic oligonucleotide 3'-CCTTCGAGTATACCTCTTTT-5' . The phage DNA and oligonucleotide were reacted with Klenow in the presence of the four deoxyribonucleotides, as described in Zoller, et al., supra. The reaction mixtures were used to transform JM103; transformants were screened for an exact match with the mutagenic oligonucleotide. This was accomplished by probing them with the kinased oligonucleotide, using stringent hybridization and washing conditions. Phage which contained DNA which hybridized to the oligonucleotide under these stringent conditions were sequenced to confirm the presence of the mutation. One of the recombinant phage was isolated as a double-stranded bacteriophage, by the methods described in Sanger, et al., supra, which are incorporated herein by reference. III.c.l.a.2. Expression of CAT activity in pAL12 and pAL13 as compared to pKKr3

The Ndel-Hindlll CAT gene fragment, isolated from the double-stranded bacteriophage of section III.c. l.a.l. , was cloned into the Ndel-Hindlll site of pAL12 and of pAL13. MC1061 was transformed with each of these vectors, and with pKKr3. The transformants, harbor¬ ing pAL12-CAT or pAL13-CAT, were grown in rich medium. CAT activity was determined as described in section II.b.2., and protein expressed as CAT determined by gel electrophoresis in the Laemmli system, as described in Section I.

The results obtained by densitometry of the Laemmli gels showed that the steady state levels of CAT protein in pAL12, pAL13, and pKKr3 were about 30%, 30%, and 10%, respectively.

CAT activity averaged about 51,300 units for pAL12, 30,000 for pAL13, and 10,000 for pKKr3.

The results indicate that the synthetic RBSs.of ρAL12 and pAL13^" are more efficient than the natural CAT gene RBS carried on pKKr3.

III.c.2. Construction vectors with a strong RBS and an A- T rich region upstream The r3 promoter fragment in pAL 13 contains a unique restriction site (EcoRI) at -60 from the start of transcription of the major promoter, promoter 2. An A-T rich DNA fragment was inserted into this site, thus extending the A-T rich region already present at -48 to - 57. This was accomplished as follows.

A self-complementary A-T rich oligonucleotide was synthesized using automated synthesis. The sequence of the oligonucleotide was:

5'-AATTAAAAATTTTT-3' . As a control, a self-complementary G-C rich oligonucleotide of equal length was also synthesized; the sequence of this -oligonucleotide was:

AATTGGGGGCCCCC-3' . Each of these oligonucleotides was hybridized as described in section Ill.b., to form double-stranded synthetic DNA fragments with EcoRI overhangs, as shown in Figure 9a. Each of these synthetic fragments (20 ng) was ligated into the EcoRI site pAL13-CAT (25ng) in a 20 microliter re- action volume. Figures 9B-D shows the resulting sequence of each construction: Figure 9b, the sequence of the vec¬ tor without an insert (r3), the sequence of the vector with an A-T rich region (r3AT), and the vector with a G-C rich region (r3GC; Figure 9c, the sequence of the r3 promoter region (present in all constructions); and Figure 9d, the sequence from the end of the r3 promoter to the translation initiation codon (ATG, shown in a box) of the CAT gene in pAL13CAT. The notations for Figure 9c are the same as those used for Figure 2. The new plasmids with the new A-T insert and with the new G-C insert termed pAL13-CAT-AT and pAL13-CAT-GC, respectively.

To ensure that in the new constructions, the oligonucleotides contained the synthesized sequences, and that the r3 promoter fragment remain unchanged, the sequence of DNA from the Ndel site through the r3 promoter and the inserted oligonucleotide was determined using the method of Sanger et al. More specifically, 3 ml of MC1061 harboring pAL13-CAT, pAL13-CAT-AT, and pAL13-CAT-GC, were grown in rich medium to O.D.-₅₀=0.25 and harvested. Half of each harvested culture was lysed and assayed for CAT activity, as described in section II.b.2. The other half was lysed by the hot guanidine isothiocyanate method, described supra, and the total nucleic acids subjected to an SI analysis as described in section II.c, using 25 ng of the r3 promoter-containing EcoRI-Ndel fragment of pAL13 to probe 20 ng of the pAL13 constructions. Since mRNA is less than 10% of the total RNA, the probe was in excess; thus, the resulting fragments represented a quantitative estimate of the relative amounts of CAT-mRNA. The relative CAT activities, and CAT mRNA levels obtained with pAL13-CAT, pAL13-CAT-GC, and pAL13-CAT-AT are shown in Table 2. The CAT activities were normalized to the values obtained for the parental plasmid, pAL13- CAT. The relative CAT-mRNA levels were determined by scanning the autoradiographs, and normalizing the integrated peak values to those obtained for the parental plasmid, pAL13-CAT.

Table 2

pAL13-CAT pAL13-CAT-GC pAL13-CAT-AT

The results in Table 2 show that the levels of CAT activity and CAT-mRNA were the same when the cells were transformed with either pAL13-CAT, or pAL13-CAT-GC. In contrast, the levels of both CAT activity and CAT-mRNA were 2.2-fold and 3-fold higher, respectively, when the cells were transformed with pAL13-CAT-AT. Thus, insertion of the A-T rich region significantly increased the expres¬ sion of the CAT gene.

Based upon the SI analyses, initiation of mRNA occurred at the same bases with all three constructions; this was the same initiation site as in pKKr3. Since the mRNA sequence was identical with all three constructs, the higher level of mRNA in pAL13-CAT-AT must have resulted from increased transcription due to the insertion of the A-T sequence at -60. pAL12 also contains an EcoRI site at -60. Construction of the vector pAL12-AT is accomplished using the same techniques as were used "for the construction of pAL13-CAT-AT, described above.

Ill.d. Construction of r3 vectors containing the lac operator, pAS19CAT and pAS19

III.d.1. Construction of pAS19 A scheme for the construction of pAS19 is shown in Figure 10. First, pAS14 was constructed from pALr3 by the insertion of a synthetic oligonucleotide_, duplex with the sequence shown in Figure 10. The synthetic oligonucleotide was derived from two complementary oligonucleotides which had been synthesized by solid phase DNA synthesis. The treatment by which the oligonucleotides were kinased and hybridized is described in Section Ill.b. The hybridization created a synthetic DNA fragment with a with a 5 '-BamHI overhanging end, and a 3'-HindllI overhanging end. The sequence of this fragment is the same as that which was used during the construction of pAL13 (see Figure 7 for the fragment used in the construction of pAL13), except for the nucleotides forming the HindiII overhanging end. Insertion of BamHI-HindiII fragment, instead of the BamHI-BamHI fragment, recreates the BamHI site adjacent to the RBS. The synthetic BamHI- HindiII fragment was ligated to pALr3 which had been digested with BamHI and Hindlll. MC1061 was transformed with the ligation mixture. The conditions for ligation, transformation, and screening of the vectors was as described in section III.c.1. A candidate containing the inserted fragment was named pAS14.

A vector which was to be the recipient of a BamHI fragment containing the lac operator sequence (LAC ) , was created by removing the cassette which encoded the r3 promoter linked to the synthetic BamHI-HindiII fragment from pAS14, and inserting it into pSH54. pHS54 is a derivative o.f pTRP233, from which the existing BamHI site has been destroyed by treatment of the parental vec- tor with BamHI, Klenow in the presence of the four deoxyribonucleoside triphosphates, and ligation with T4 DNA ligase.

PTRP233, which is Ndel-, was prepared from pKK 233-2, which is described in detail in Amman, E., et al, Gene (1985) 4_0: 183. The tac promoter of pKK 233-2 was replaced with a synthetic trp promoter of the nucleotide sequence shown in Fig. 8A. The Ndel site of the starting plasmid was eliminated by digesting pKK 233-2 with Ndel, blunting with Klenow, and religating. The Ndel^" product was then digested with EcoRI and PstI, and ligated to an EcoRI/PstI digest of the synthetic trp promoter of Figure 8A to obtain the desired vector, pTRP233. Figure 8B shows some of the characteristics of pTRP233.

Substitution of the r3 promoter cassette for the tryptophan promoter-operator (TRP P/ό) sequence in pSH54 was accomplished as follows. The cassette was removed from pAS14 by treatment of the vector with EcoRI and Hindlll, followed by separation of the fragments by size. The segment of pSH54 which encodes TRP P/O was removed by treatment of pSH54 with EcoRI and Hindlll, followed by separation of the fragments by size. The r3 promoter cas¬ sette, which was the smaller fragment from pAS14, was ligated to the larger fragment derived from pSH54, using T4 DNA ligase. MC1061 was transformed with the ligation mixture. The conditions for transformation were as previ¬ ously described. Selection of a plasmid containing the cassette was by ampicillin selection. A plasmid which contained the r3 promoter cassette was termed pAS18. To create an r3 vector containing LAC , a synthetic DNA fragment containing the sequence was inserted into the unique BamHI site in pAS18. The BamHI site was recreated during the insertion of the r3 promoter cassette. The sequence of the synthetic fragment, which contains 5'- and 3'- overhanging BamHI ends is shown in Figure 10. The complementary oligonucleotides of this fragment were synthesized by automated DNA synthesis, and conditions for annealing were as described for the anneal¬ ing of other oligonucleotide fragments described supra. Prior to insertion of LAC , pAS19 was digested with BamHI, and treated with bacterial alkaline phosphatase (BALP) to prevent self-ligation of parental vectors during ligation with the LAC sequence. Insertion was accomplished react¬ ing the LAC° fragment with the treated vector in the pres¬ ence of T4 DNA ligase. Transformation of MC1061 with the ligation mixture was using transformation conditions previously described. Plasmids isolated from the transformants were screened. One plasmid, which contained the insert in the proper orientation, was termed pAS19.

III.d.2. ^{^}Construction of pAS19CAT and of the control vec¬ tor, PAS18CAT

The synthesis of pAS18CAT and of pAS19CAT was accomplished by the insertion of the CAT gene derived from pAL13-AT-CAT into the site created by Ndel and Hindlll in pASIS, and in pAS19, respectively.

The CAT gene was isolated from pAL13-AT-CAT by digestion of the vector with Ndel and with Hindlll, fol¬ lowed by separation of the fragments on the basis of size. The CAT gene is encoded within the smaller fragment. The construction of pAL13-AT-CAT is discussed in section

III.c.2.. The isolated CAT gene fragment was ligated with pAS18 and pAS19, each of which had been treated with Ndel and Hindlll prior to the ligation. Ligation was ac¬ complished using T4 DNA ligase. MC1061 was transformed with each ligation mixture. Plasmids were isolated from the transformants. Plasmid derivatives from pAS18 and from pAS19, each of which contained the CAT gene, were termed pAS18-CAT _,and pAS19-CAT, respectively.

III.d.3. The effect of catabolite repression on expres¬ sion of the CAT gene in pAS19CAT and its control, pAS18CAT Transformants of E. coli D1210 (lacl^) harboring pAS19CAT, or the control vectors, pAS19, pAS18CAT, or pTAC/CAT, were grown in M9CAA medium at 37°C in the pres- ence or absence of IPTG (ImM). After growth, the bacteria were harvested, and proteins expressed as CAT determined by gel electrophoresis in the Laemmli system and staining, as described in Section I. Electrophoresis was on 12% polyacrylamide gels containing SDS. Under the growth conditions, in the absence of IPTG, the expression of proteins under the control of LAC is repressed. Addition of IPTG to the medium induces the expression of these proteins. In addition, proteins which are not under the control of LAC are expressed constitutively. The results o the expression of the CAT gene are shown in Table 3; "+" symbolizes the presence of a CAT band on the stained gel, and "-" symbolizes the absence of the band.

Table 3

with IPTG without IPTG

pAS19 (r3, LAC°)

pTAC/CAT (TRP P, LAC ,o.) +

pAS18CAT (r3) + +

PAS19CAT (r3, LAC°) + The results showed that the expression of the CAT gene in pAS19CAT was sensitive to catabolite repres¬ sion, and hence was under the control of the synthetic LAC . As expected, the control vector, pAS18CAT, which was the same construction as pAS19CAT but lacked the LAC° sequence, allowed the constitutive expression of the CAT gene- in the transformants. The insensitivity of the r3 promoter to catabolite repression has been discussed earlier. Hence, insertion of the LAC sequence in the vector construct rendered the r3 promoter activity sensi¬ tive to catabolite control.

IV. The expression of heterologous genes inserted into the r3 expression vectors

IV.a. The expression of apolipoprotein Al encoded in PALlO

IV.a.1. Construction of a full-length apolipoprotein cDNA A bacterial vector containing a full-length apolipoprotein Al (ApoAI) cDNA was constructed as described in commonly owned U.S. patent application No. 834,300, which is hereby incorporated by reference. More specifically, plasmid pBL13AI was prepared as described in Seilhamer, J.H., et al. (1984), DNA 309. The plasmid (100 micrograms) was digested to completion with EcoRI for 3 hr at 37°C. The 965 b.p. EcoRI fragment containing the apoAI cDNA was isolated on a 4% (w/v) nondenaturing polyacrylamide gel, as described in Maniatis et al., supra. The apoAI cDNA fragment was excised from the gel, electroeluted, and concentrated by ethanol precipitation. The resulting DNA pellet was dried in vacuo and_ resuspended in water. This EcoRI fragment was further digested with Sau3A for 30 min at 37°C with a DNA-to- enzyme ratio of 1 microgram : 2 units. The resulting Sau3A fragments were separated as described above. The 783 base pair partial fragment was isolated and concentrated as described above. .

Two complementary oligonucleotides containing the coding sequence of the first 8 amino acids of mature apoAI protein were synthesized on an automatic synthesizer. The sequences of the oligonucleotides in the hybridized fragment were:

5' CATG GAC GAA CCG CCG CAG TCT CCG TGG 3'

CTG CTT GGC GGC GTC AGA GGC ACC CTAG

The oligonucleotides were kinased with [gamma- 32P-ATP] using T4 polynucleotide kinase (Maniatis et al. , supra) . The two oligonucleotides (Ing each) were mixed, boiled for 2 min, and allowed to hybridize at RT for 60 in. The hybridized fragment, which contained an Ncol overhang at its 5'-terminus, and a Sau3A overhang at its 3'-terminus, was ligated to 25 ng of Sau3A partial apoAI cDNA fragment with the addition of t4 DNA ligase by the method of

Maniatis et al., supra. After ligation, the mixture was digested to completion with EcoRI and Ncol. Products were separated on a 5% nondenaturing polyacrylamide gel. The 811 base pair fragment corresponding to a full-length cod- ing sequence of mature apoAI protein was excised. The DNA was eluted from the gel slice in lOmM Tris-HCl, pH 8, 1 mM EDTA, 0.4 M NaCl with agitation at 37°C overnight and concentrated by ethanol precipitation. The NcoI-EcoRI fragment was then ligated into pBR329 vector. The result- ing plasmid was designated pFLAI-2.

To confirm the correct DNA sequence through the constructed oligonucleotide site, an Rsal fragment of pFLAI-2 encompassing the first 56 b.p. of the coding region for mature apoAI, as well as 112 base pairs downstream into pBR329, was sequenced by the dideoxy method (Messing, J.(1982), supra and Sanger et al., supra.

An Ncol-HindiII fragment containing the apoAI cDNA was isolated from pFLAI-2 by digestion of the vector to completion with each of these enzymes. The fragment containing the cDNA was purified on a 5% nondenaturing gel and concentrated as described supra.

IV.a.2. Construction of pALlO encoding apoAI The Ncol-Hindlll fragment encoding mature apoAI was ligated into Ncol-HindiII cut pALlO using T4 DNA ligase, and the conditions for ligation and transformation described supra. A plasmid containing the apoAI gene was isolated from transformants, and termed pr3AI.

IV.a.3. Construction of ptacAI

To compare the expression directed by the r3 promoter with that directed by the tac promoter, r3 in pr3AI was replaced with the tac promoter. This was ac- complished as follows. The vector, pr3AI was digested with EcoRI and StuI, and phosphatased. The tac promoter was isolated from pKKtac by digesting the vector with BamHI, filling in the ends with T4 DNA polymerase, Klenow fragment, and the four deoxyribonucleoside triphosphates, and digesting the phenol extracted mixture with EcoRI.

The isolated tac promoter fragment was cloned into the cut pr3AI, creating ptacAI.

IV.a.4. Expression of ApoAI encoded within pr3AI, compared to expression of ApoAI encoded within ptacAI

Transformants of MC1061 harboring either pr3AI, ptacAI, or pALlO (the control vector) were grown in minimal medium supplemented with 40 micrograms/ml isoleucine, leucine and valine to an O.D._ccn=0.5. [ 35S]- methionine (50 microCuries) was added to 1 ml of bacterial culture for 2 minutes, and then the cells were immediately lysed by the addition of 340 microliters 20% trichloracetic acid and held on ice for 15 minutes . The precipitated protein was collected by centrifugation, washed with acetone, dried, and resuspended in 50 micro¬ liters 50 mM phosphate buffer, pH 7.0. Each aliquot of resuspended protein was incubated overnight on ice with 5 microliters of commercially available rabbit anti-human apoAI antiserum. The immune complexes were precipitated by incubation for 30 minutes on ice with 100 microliters of commercially available heat-killed, formalin-fixed, Staphylococcus aureus, and then washed as described by Kessler, supra. The immunoprecipitated complexes were resuspended in 50 microliters of Laemmli gel buffer (Laemmli, supra) and boiled for 5 minutes. 40 microliters of boiled supernatant was electrophoresed on a 12.5% Laemmli gel (Laemmli, supra) and the dried gel autoradiographed.

The autoradiograph showed that a protein of 27,000 MW, the size of native apoAI, immunoprecipitated with the antiserum from extracts of cells harboring plasmids that express apoAI, i.e., pr3AI and ptacAI. This protein was not present in extracts of cells harboring pALlO. The autoradiographs were scanned with a Kontes

Model 800 densitometer, and the relative peak sizes analyzed with a Hewlett-Packard Model 3390A integrator. The results of this analysis showed that there was twice as much bacterially synthesized apoAI from the r3 construction (pr3AI) as compared to the tac construction (ptacAI) .

The above experiment was repeated two times. The level of immunoprecipitated apoAI from the r3 construction as compared to the tac construction were one and a half times higher, and twice as high, in the two experiments.

The above data agree with the data in section II.b.2. and II.b.3., with respect to^' the expression of the CAT gene in r3 and tac containing constructs, i.e., the r3 promoter directs about twice as much expression as does the strong promoter, tac.

IV.b. The expression of a gene encoding fibroblast growth factor inserted into pAL12 and pAL13

IV.b.1. Preparation .of acidic human fibroblast growth factor coding seguence

The coding sequence for acidic human growth fac- tor (haFGF) was prepared as described in commonly owned U.S. patent application no. 809,162, which is hereby in¬ corporated by reference. More specifically, a cDNA library prepared from breast carcinoma mRNA was probed with an 250/AluI probe to obtain cDNA encoding acidic hFGF. The description of the preparation of the prepara¬ tion of the 250/AluI probe is in the above referenced pat¬ ent application. An unspliced cDNA containing the first exon was obtained.

In the alternative, the cDNA sequence informa- tion obtained by Jaye, M. , et a., Science (1986), 233, 541 which is hereby incorporated by reference, was used as a guide for the synthesis of a gene encoding the acidic hFGF. The cDNA clone reported by Jaye et al. was obtained using mRNA from human brain stem, and encodes an acidic hFGF.

The genomic lambdaHAG-9.1 clone was used to provide the 5' portion of the gene. The description of the preparation of lambdaHAG- .1 is in U.S. patent ap¬ plication no. 809,162. To prepare the 5'- portion of the gene, a 1.9kb BamHI fragment was isolated from lambdaHAG9.1 and subcloned into pUC13 to obtain pCBI-101. This intermediate plasmid was then digested with Ncol and BamHI, and the 118 bp fragment containing the codons for the 15 amino acids of the pro sequence along with the first 25 amino acids of the mature "primary: form of acidic hFGF was isolated using a 5% polyacrylamide gel. The location of the Ncol site which contains the ATG that is believed to constitute the start codon at amino acid - 15 from the beginning of the primary sequence is shown in Figure 11, which diagrams the synthetic gene.

The remainder of the coding sequence was synthesized using the synthetic oligonucleotides numbered 1-20 in Figure"11. The synthesis of the individual oligonucleotides uses conventional automated techniques. The oligonucleotides were designed so as to yield the same nucleotide sequence as that reported by Jaye et al., supra, with two exceptions: oligonucleotides 4 and 14 were constructed so as to destroy the Ncol site spanning codon 67 by altering the GCC encoding alanine at codon 66 to GCT, as shown by the asterisk; in addition, oligonucleotides 19 and 20 were modified so as to add Hindlll and EcoRI cleavage sites following the TGA termination codon. Neither of the foregoing changes af¬ fects the amino acid sequence encoded. The synthetic oligonucleotides were ligated to obtain the sequence shown in Figure 11 by kinasing 5 micrograms of each oligonucleotide (except #1 and #20) using standard reaction conditions, annealing the 10 dif¬ ferent complementary oligonucleotide pairs (1 + 11, 2 + 12, etc.), and then ligating the ten oligonucleotide pairs into three segments. These segments were formed sequentially using T4 ligase under standard conditions. To obtain segment A, the pair 1/11 was ligated with 2/12, followed by ligation with 3/13, followed by ligation with 4/14. Segment B was formed by ligation of 5/15 with 6/16, followed by 7/17. Segment C was obtained by ligating 8/18 with 9/19, followed by ligation of the product with 10/20. The three ligated subfragments (A=144 bp, B=108 bp, and C=106 bp) were purified using gel electrophoresis and then sequentially ligated by mixing B and C under standard conditions with T4 ligase, followed by addition of A. The final reaction was extracted with phenol, precipitated with ethanσl, the ethanol precipitate electrophoresed on a 5% acrylamide gel, and the 358 bp fragment A+B+C was eluted. The fragment spanned the BamHI/EcoRI sites, as shown in Figure 11, and its sequence was verified using dideoxy sequencing by subcloning the segment into M13mpl9.

To complete -the coding sequence, the synthetic 358 bp Ba HI-EcoRI synthetic fragment was isolated from the phage or the polyacrylamide gel, its ends kinased, as necessary, and ligated to the 118 bp NcoI-BamHI fragment from pCBI-101. The resultant partially synthetic ^"nucleotide sequence encoding haFGF is shown in Figure 11.

IV.b.2. Construction of pAL12-haFGF and pAL13-haFGF

Prior to inserting the synthetic gene encoding haFGF into pAL12 and into pAL13, the Ncol site at the 5'- terminus of the gene was converted into an Ndel site using site directed mutagenesis. This was accomplished by digesting the synthetic fragment encoding haFGF with Ncol, blunting the end, and then digesting it with EcoRI. This fragment was cloned into Smal-EcoRI digested ml3mpl8, using the technique of Messing et al., supra. White plaques were screened for orientation, and the DNA from one of the recombinant phage with the hybridizing orienta¬ tion was hybridized to the mutagenic oligonucleotide

5'-CTAGAGGATCCCATATGGCTGAAGG-3' . This oligonucleotide hybridizes across the ml3-haFGF gene boundary, and changes the Ncol site (CCATGG) to an Ndel site (CATATG); the underlined portion of the sequence shows the Ndel site. The phage DNA and oligonucleotide were reacted with Klenow in the presence of the four deoxyribonucleotides, as described in Zoller, et al. , supra. The reaction mixtures were used to transform Jml03; transformants were screened for an exact match with the mutagenic oligonucleotide. This was accomplished by probing them with the kinased oligonucleotide, using the hybridization and washing conditions described in Maniatis, supra. Phage which contained DNA which hybrid¬ ized to the oligonucleotide under stringent conditions were sequenced to confirm the presence of the mutation. One of the recombinant phage was isolated as a double- stranded bacteriophage as described in section III.c 1.a.1.

A DNA segment encoding haFGF was inserted into pAL-12 and pAL-13 as follows. The synthetic gene was removed from M13mpl8 by digestion with Ndel and Hindlll. As shown in Fig 11, the Hindlll site in the synthetic gene is just upstream of the 3'-EcoRI site. Both vectors, pAL12 and pAL13, were digested with Ndel and Hindlll, and the Ndel-Hindlll fragment encoding haFGF was ligated into each vector. Vectors which contained the fragment, and which were derived from pAL12 or pAL13, were termed pAL12- haFGF and pAL13-haFGF, respectively.

IV.b.3. Expression of haFGF encoded in pAL12-haFGF and in PAL13-haFGF

The expression of haFGF synthetic gene inserted into pAL12-haFGF and pAL13-haFGF was determined as fol¬ lows. Cultures of E. coli K-12 strain B (Luria et al (1942), Arch. Biqchem. _1, 111) were transformed with pAL12-haFGF or with pAL13-haFGF, or with the control vec- tors, pAL12 or pAL13. The transformants were grown overnight to saturation, and lysates prepared. Aliquots of the cell lysates were subjecte^'d to electrophoresis on polyacrylamide gels in the Laemmli system, as described in Section I. Size markers were run on the same gels. The gels were stained with Coomassie blue. The predicted molecular weight of haFGF is about 17,000.

The results showed that a new protein was expressed in cells transformed with pAL12-haFGF, and in cells transformed with pAL13-haFGF which was not expressed in cells transformed with the parental vectors, pAL12 and pALl3. The 17,000 MW protein was identified as native haFGF based upon the characteristics of the binding to heparin sepharose. While the present invention has been illustrated above by certain specific embodiments, it is not intended that these specific examples limit the scope of the inven¬ tion as described in the appended claims .

Utility

The various embodiments of the invention are useful for the production on a large scale and in relatively highly purified form, of a variety of com- mercially desirable polypeptides. The r3 promoter is a strong promoter in gram negative bacteria, and causes enhanced expression of genes under its control. Thus, recombinant methods utilizing bacteria harboring vectors containing the r3 promoter may give increased yields of commercially desirable polypeptides, which include, for example, interferons, interleukins , hormones, enzymes, growth factors , apolipoproteins , and cellular regulatory factors .

The purified DNA fragment encoding the r3 promoter is useful for the production of recombinant DNA constructs. Constructs containing the r3 promoter are useful for the production of r3 containing expression cas¬ settes. The expression cassettes, in turn, allow for the ready creation of r3 containing expression vectors which exhibit characteristics which are desirable for commercial use, such as the ability to control the expression of the recombinant polypeptide.

Claims

1. A recombinant DNA construct comprising a first nucleotide sequence comprised of an r3 promoter sequence, wherein said r3 promoter sequence is operably linked to a transcription initiation sequence, and wherein said r3 promoter sequence is selected from a group consisting of the nucleotide sequence

5'-ACAGAAATTTTTCGCCGTACGCTATTGCGTGACGTAGATTCGTG ACGTATAGTTACTACAGCTTATTTGTATATAACCACCATCAGGT-3' , and mutants thereof, wherein said mutants exhibit promoter activity.

2. A construct according to claim 1, wherein the mutants are comprised of a nucleotide sequence

5'-TTCGCC-X -TAGATT-3' , wherein X is a spacer sequence consisting of y number of nucleotides, and wherein y is 15 to 19, and preferably is

18, or wherein the mutants are comprised of a nucleotide sequence

5'-TTGCGT-X -TATAGT-3' , wherein X is a spacer sequence consisting of y number of nucleotides, and wherein y is 15 to 19, and preferably is 17, or wherein the mutants are comprised of a nucleotide sequence

5'-GTGACG-X -TACTAC-3' , wherein X is a spacer sequence consisting of y number of nucleotides, and wherein y is 15 to 19, and preferably is

19, or wherein the mutants are comprised of a nucleotide sequence

5'-GTGACG-Xy-TATTTG-3' , wherein X is a spacer sequence consisting of y number of nucleotides, and wherein y is 15 to 19, and preferably is 16.

3. A construct according to claim 1 or claim 2 further comprising a second nucleotide sequence, wherein said sequence terminates transcription, and wherein the first and second sequence are positioned to promote and terminate the transcription of a single RNA molecule.

4. A construct according to claim 3 further comprising a third nucleotide sequence, wherein said third nucleotide sequences is comprised of sequences encoding: a. a ribosomal binding site (RBS); and b. a translational initiation codon; wherein said third sequence is positioned and oriented so that the ribosomal binding site and translational initia¬ tion-site are operably linked to the r3 promoter sequence.

5. A construct according to claim 4, wherein the construct is selected from a group consisting of the construct in pALlO, or pAL12, or pAL13, and mutants thereof.

6. A construct according to claim 5, wherein the construct in pALlO is comprised of a sequence AGGCCTTAAGGAGGTTTAACCCATGGCCCGGG TCCGGAATTCCTCCAAATTGGGTACCGGGCCC, or wherein the construct in pAL12 is comprised of a sequence

AGCTGAGGAAAAAACATATG TCGACTCCTTTTTTGTATAC, or wherein the construct in pAL13 is comprised of a sequence GGAGGAAAAAACATATGCAGCTG CCTCCTTTTTTGTATACGTCGAC.

7. A construct according to claim 4, further comprising a fourth nucleotide sequence, wherein said fourth sequence is comprised of sequence encoding a heterologous structural gene, wherein said fourth sequence is positioned and oriented so that expression of said structural gene is operably linked to the first, second, and third nucleotide sequences.

8. A construct according to claim 7, further comprising a fifth nucleotide sequence, wherein said fifth sequence is an A-T rich oligonucleotide, and wherein said fifth nucleotide sequence is positioned and oriented such that it enhances the transcription of an RNA promoted by the first sequence and terminated by the second sequence.

9. A construct according to claim 7 further comprising a sixth nucleotide sequence, wherein said sixth sequence, is an operator sequence, and wherein said operator sequence is positioned and oriented to act as a regulatory sequence for the expression of the gene encoded within the fourth nucleotide sequence.

10. A vector comprising a recombinant DNA construct, said construct being comprised of: a. a first nucleotide sequence comprised of an r3 promoter sequence, wherein said r3 promoter sequence is selected from a group consisting of the nucleotide sequence

5'-ACAGAAATTTTTCGCCGTACGCTATTGCGTGACGTAGATTCGTG

ACGTATAGTTACTACAGCTTATTTGTATATAACCACCATCAGGT-3' , and mutants thereof, wherein said mutants exhibit promoter activity.

11. A vector according to claim 10, wherein the vector is pALr3, or pALlO, or pr3AI, or pAL12, or pAL13, or pAL12-haFGF, or pAL13-haFGF.

12. A transformed cell harboring the vector of claim 10 or claim 11.

13. A method of producing a recombinant polypeptide comprising: a. providing a population of transformed bacte-

, rial cells harboring the vector of claim 10 or claim 11; b. growing said population of cells under condi¬ tions whereby said polypeptide is expressed; and c recovering said polypeptide.

14. A recombinant polypeptide produced by the method of claim 13.