WO1991000355A1

WO1991000355A1 - Genetic elements useful in production of active protein

Info

Publication number: WO1991000355A1
Application number: PCT/US1990/002987
Authority: WO
Inventors: Eric R. Olson
Original assignee: The Upjohn Company
Priority date: 1989-06-30
Filing date: 1990-06-01
Publication date: 1991-01-10
Also published as: AU5966790A

Abstract

The present invention relates to a recombinant DNA molecule comprising a signal sequence having a nucleotide sequence encoding an amino acid sequence MetArgIleSerLeuLysLysSerGlyMetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla or MetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla. The recombinant DNA molecule of the present invention may further comprise a DNA sequence encoding a fusion polypeptide sequence having a sequence downstream from the signal sequence. The recombinant DNA molecule that is the present invention may further comprise a DNA sequence encoding a heterologous polypeptide and, furthermore, may comprise a linker downstream from said signal sequence and upstream from the DNA sequence which encodes the heterologous polypeptide.

Description

GENETIC ELEMENTS USEFUL IN PRODUCTION OF ACTIVE PROTEIN

FIELD OF THE INVENTION

The present invention relates to a novel protein and the genetic elements therefor which are useful in facilitating production and proper folding of heterologous proteins produced in prokaryotes.

BACKGROUND OF THE INVENTION

When producing heterologous proteins in prokaryotes, several problems arise in making proteins which have normal activity. First, foreign proteins present in the prokaryotic cytoplasmic space are subject to degradation by endogenous proteases. Second, unless formation of disulfide bridges takes place correctly, the heterologous protein will not form the correct secondary and tertiary structure; therefore, it will be inactive. Third, purification procedures used to isolate the protein from the prokaryotic cytoplasm can cause protein degradation or loss of structure.

To decrease the incidence of protein degradation of heterologous peptides by endogenous proteases in Escherichia coli, gene fusions involving bacterial proteins have been used. For example, the lacZ gene has been used in expression of somatostatin and insulin. The E. coli host's endogenous proteases do not degrade the fusion peptide as readily because of the presence of the bacterial protein sequences. However, in vivo, such hybrid proteins often precipitate intracellularly and form granules or so called inclusion bodies.

It is also a common problem that production of heterologous proteins in prokaryotes results in the incorrect folding of the heterologous protein and therefore, biologically non-active molecules are produced. This problem has frequently been observed with peptides which are dependent on a correct formation of disulfide bridges that cannot be formed due to the reducing environment in the cytoplasm of most bacteria. Many foreign proteins accumulate in the E. coli cytoplasm in the form of granules, which are insoluble protein aggregates . After the solubilization procedure and purification, the proteins are often in a non-active form and complex renaturation procedures must follow to produce active protein.

Although granule formation can often simplify the isolation and purification of a desired recombinant DNA product, the use of detergents or denaturing agents is required to dissolve the granules, thereby necessitating a refolding step to obtain the desired product with the correct secondary and tertiary structure. Refolding of proteins can be difficult and the final yield of desired protein products may suffer greatly.

Another potential disadvantage in the cytoplasmic production of foreign proteins in E. coli is that synthesis of the desired proteins is initiated with methionine which sometimes is not removed either in vivo or in vitro. It has been speculated that the methionine derivatives of recombinant DNA products, such as methionyl human growth hormone (met-hGH) may play a role in antibody formation in patients treated with the hormone.

An alternative strategy is to secrete the gene product out of the reducing environment of the cytoplasm. A correct folding of peptides containing disulphide bridges may then be achieved. Secretion also has the advantage of simplifying the recovery of the product, in particular when excretion to the culture medium is obtained. Gram positive bacteria, such as Bacillus subtilis, Staphylococcus aureus and Streptomyces lividans, which efficiently secrete many proteins into the medium, have been used extensively to develop host-vector systems designed to express and secrete recombinant products. Similar systems have also been developed for E. coli.

In order to obtain truly excreting E. coli host-vector systems in which the gene product is transported out of the cell, several approaches have been tried. First, leaky (lky or exc) mutants of E. coli K12 have been isolated and used for optimized extracellular production of alkaline phosphatase and beta-lactamase. Second, secretion vectors based on genes encoding E. coli outer membrane proteins have been constructed. Finally, genes encoding normally extracellular proteins from Gram-positive bacteria have been used to construct expression vectors for E. coli. Although each of these systems have potential, none has yet proved to be an effective system for the general production of heterologous proteins.

Alternative systems have also been developed for E. coli in which the product is recovered from the periplasmic space rather than from the culture medium. The advantages of isolation from the periplasmic space instead of the cytoplasmic fraction from E. coli include avoiding the problems of protein degradation, improper folding and destabilizing purification techniques associated with cytoplasmic production. The E. coli periplasm provides an environment which facilitates efficient disulfide bond formation and proper folding of heterologous peptides. The purification and isolation of the recombinant protein are greatly simplified and the processes needed to refold recombinant proteins derived from the E. coli cytoplasm are eliminated. When heterologous proteins are secreted across the inner membrane into the periplasmic space they can be easily isolated by rupturing the outer membrane and collecting the supernatant following centrifugation. The correctly folded protein with the proper disulfide bonds can be easily purified.

The periplasmic space lies between the inner and outer membranes of gram-negative bacteria. Because of this location, this space should not be thought of as a single homogenous compartment but rather as consisting of several distinct microenvironments created by the two boundary membranes and the peptidoglycan layer. Periplasmic proteins localized within these regions fulfill important functions in the processing of essential nutrients and their transport into the cell and in the biogenesis of the cell envelope.

Periplasmic polysaccharides and other small molecules serve to buffer the cell from changing osmotic and ionic environments and thus help to preserve the more constant internal environment needed for cell growth and viability. Periplasmic proteins, like other envelope proteins , are made as larger precursors containing an aminoterminal signal peptide which is cleaved during export. The amino acid sequence for signal peptides of 13 periplasmic proteins for E. coli is known. These proteins include alkaline phosphatase, eight binding proteins (arabinose and galactose, histidine and lysine-arginineorithine, isoleucine-valine, leucine, maltose, and phosphate), two β- lactamases (AmpC and TEM), and the two subunits of the heat-labile enterotoxin. The signal peptides are involved in the transport of the periplasmic proteins from the cytoplasm to the periplasmic space.

To achieve transport of a heterologous protein to the periplasmic space, it is necessary to attach a signal peptide to its N- terminus. Since practically all proteins that have been identified that have to transverse the inner membrane contain a leader sequence, it is necessary that the heterologous peptide have attached at its amino terminus the amino acid sequence which makes up a signal peptide. The leader sequence is usually removed from the protein by an enzyme (leader peptidase) associated with the inner membrane resulting in the "mature" protein. Removal of the leader peptide is not required for transport. Although the leader sequence is required for the translocation process its exact role is not yet understood. The signal peptide is attached to the heterologous protein by linking the genetic code for the signal peptide upstream from the genetic sequence for the heterologous protein. The fused peptide which results from the translation of such recombined genetic material will be transported to the periplasmic space under appropriate conditions .

An example of a periplasmic protein used to provide genetic information for periplasmic transport is alkaline phosphatase. Alkaline phosphatase of E. coli is synthesized in large amounts under low phosphate conditions and is secreted across the inner membrane to the periplasmic space. Like other secreted proteins, alkaline phosphatase (APase) is synthesized as a precursor form with a signal peptide at the NH₂-terminal. This signal sequence is removed during the secretory process and the mature form is transported to the periplasmic space.

Thus, in a chimeric protein which has no extra amino acid between the end of the alkaline phosphatase signal sequence and the first amino acid of the eukaryotic mature protein, the signal sequence should be processed and the eukaryotic protein should be transported across the inner membrane to the periplasmic space as a mature form. Proteins in the periplasmic space can be extracted easily by the osmotic shock procedure, a method that does not lyse bacteria. Other periplasmic proteins may be used similarly to achieve export of foreign proteins to the periplasmic space.

Transport across the inner membrane through the cell's export pathway is accomplished by fusing the coding sequence of the desired protein in frame with an N-terminal signal sequence of E. coli origin. The signal sequence engages in some way the secretion apparatus composed of several proteins and this initiates the energy dependent transfer of the remainder of the polypeptide across the membrane.

Since inspection of leader sequences reveals no identifiable difference between those of periplasmic and outer membrane proteins that might direct the proteins to their respective compartments, some information specific for sorting is contained in the mature polypep tide. Furthermore, linking the signal peptide to a mature protein does not necessarily result in the transport of the protein to the periplasmic space. The details of the mechanism of how a protein is transported through the inner membrane into the periplasmic space are not known. However, there are several components from both the protein to be transported and the bacterial cell that are crucial for the transport to the periplasm. In addition to the leader sequence, there is evidence that amino acids located in the mature part of the protein are either required or facilitate the translocation process. Since proteins destined for export to different locations have signal sequences of similar structure and function, the more distal steps in export apparently depend on sequences located in the mature portion of the secreted protein. For this reason, attempts at localizing heterologous proteins into the E. coli periplasm have often involved fusing the coding sequence to an E. coli gene coding for a periplasmic protein.

Sequences that are important in the localization of proteins have been termed topogenic sequences. A gene fusion approach has been developed to locate the topogenic sequences present on the polypeptide chain of secreted proteins. By fusing a gene that encodes the normally cytoplasmic enzyme to increasingly larger amounts of a gene coding for a periplasmic protein, a set of hybrid proteins can be produced. These hybrid proteins contain the periplasmic protein signal peptide and increased amounts of the aminoterminal portion of the secreted protein fused to a constant amount of enzymatically active protein, for example β-galactosidase, at the carboxy terminus. The intracellular distribution of the different sized hybrid proteins can indicate where the transport information resides on the polypeptide chain of the secreted protein. This approach assumes that marker enzyme, for example β-galactosidase, Is a passive carrier which does not positively or negatively affect the export information to which it is fused. This is not always the case. Whether the additional sequences present on the larger protein are needed for completion of traversal through or release from the inner membrane or whether they allow the larger protein to assume a water-soluble conformation in the periplasm is variable, dependent an the particular protein.

INFORMATION DISCLOSURE Ohsuye, K. et al., Nucleic Acids Research, Volume 11, No. 5, pp. 1283-1294 (1983) disclose the construction of a chimeric nucleotide sequence coding E. coli alkaline phosphatase (APase) linked to aneo-endorphin (αNE). The APase gene contains a sequence encoding a signal peptide which normally directs transportation to the periplasmic space. In the chimeric construct, the signal peptide was processed but the chimeric protein was not transported to the periplasmic space.

Miyake, T. et al., J. Biochem. 97, 1429-1436 (1985) disclose construction of a secretion vector comprising the promoter and signal sequence from the alkaline phosphatase gene of E. coli. The gene for human interferon-α was inserted downstream from the signal sequence in the vector. The human interferon was secreted into the periplasmic space. In the control experiment, the human interferon gene was inserted in the vector. No human interferon was secreted into the periplasmic space in the control experiment.

Hoffman, C.S. and A. Wright, Proc. Natl . Acad. Sci. USA, 82:5107-5111, (August, 1985), disclose construction of plasmids containing the alkaline phosphatase gene linked to the signal sequences derived from the periplasmic protein β-lactamase and alkaline phosphatase gene linked to the signal sequence of the outer membrane proteins LamB and OmpF. When linked to these heterologous signal sequences, the alkaline phosphatase protein was kinetically processed and transported to the periplasmic space.

Gray, G.L. et al., Gene, 39:247-254 (1985) disclose the transport of human growth hormone (hGH) into the periplasmic space when produced in E. coli. Plasmids containing the E. coli trp promoter and plasmids containing the alkaline phosphatase promoter and signal sequence were used. The hGH coding sequences were inserted into the respective plasmids. In both cases, the protein product was transported to the periplasmic space . The recovery of periplasmic hGH was monomeric and contained the same two disulfide bonds as authentic hGH.

Nagahari, K. et al., EMBO J., 4(No. 13A) : 3589-3592 (1985) disclose the construction of a plasmid containing a structural gene for a human β-endorphin linked downstream from the region of OmpF gene consisting of the promoter region, coding regions for the signal peptide and the N terminus of the OmpF protein. When the fusion protein was produced in E. coli, it was secreted into the culture medium. During secretion the signal peptide was correctly cleaved from the fusion protein.

Oliver, D., Ann. Rev. Microbiol., 39:615-648 (1985) is a review article which refers to protein secretion in E. coli. The article includes a discussion of the structure and function of signal sequences and the role of mature protein sequences in the export of proteins.

Nicaud, J.M. et al., J. of Biotechnology, 3:255-270 (1986) is a review article discussing current status of secretion of foreign protein, including a discussion of the export to the periplasm of heterologous proteins.

Dodt, J. et al., FEBS, 202(No.2) :373-377 (June 1986) disclose construction of a vector containing the E. coli alkaline phosphatase signal peptide linked to the tac-promoter. The gene for hirudin, a proteinase inhibitor, was inserted at a HindIII site located at the end of the signal peptide. The periplasmic fraction revealed the presence of two proteins, one identical to the hirudin product found in nature, and another with similar biological properties.

Becker, G.W. and H.M. Hsiung, FEBS, Vol. 204, No. 1, pp. 145- 150, August 1986, disclose production of human growth hormone under the control of a secretion vector containing a signal sequence for the ompA gene of E. coli. The hGH product was efficiently transported to the periplasmic space where it was purified. The human growth hormone exhibited proper folding and disulfide bridge formation.

Hsiung, H.M. et al., Biotechnology, 4:991-995 (November 1986) disclose the construction of a secretion vector containing genes that code for E. coli ompA signal peptide and human growth hormone. The recombinant fusion protein was expressed in E. coli and transported to the periplasmic space where it was collected. The hGH which was recovered exhibited the correct amino acid terminus, disulfide bonds, and secondary structure.

Oliver, D.B., Periplasm and Protein Secretion, Neidhardt, F.C. et al., Eds., Escherichia coli and Salmonella typhimurium, American

Society of Microbiology, Washington, D.C. (1987), is a chapter covering the periplasm and protein secretion in E. coli. Included is a discussion of periplasmic proteins and the role of the mature protein sequences in directing export.

Randall, L.L. et al., Ann. Rev. Microbiol., 41:507-541 (1987) is a review article describing the theories on mechanisms of protein transport.

Takahara, M. et al., Biotech., Vol. 6, pp. 195-198 (1988) disclose the construction of a plasmid comprising the gene for human superoxide dismutase (hSOD) , a cytoplasmic enzyme , cloned into a high expression secretion vector containing the E. coli ompA gene signal sequence. The cytoplasmic protein was produced and transported to the periplasmic space where it was recovered and found to be in an active state.

None of these references teach or suggest the present invention which relates to the use of genetic sequences derived from the protein fpp gene and useful in the directing of transport of heterologous peptides to the periplasmic space.

SUMMARY OF THE INVENTION

The present invention relates to a recombinant DNA molecule comprising a signal sequence having a nucleotide sequence encoding an amino acid sequence MetArglleSerLeuLysLysSerGlyMetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla or MetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla. The recombinant DNA molecule which is the present invention may further comprise a DNA sequence encoding a fusion polypeptide sequence downstream from the signal sequence. The amino acid sequence of the fusion polypeptide may comprise the amino acid sequence as set forth in Chart 2; or functional fragments of it. The recombinant DNA molecule that is the present invention may further comprise a DNA sequence encoding a heterologous polypeptide. The recombinant DNA molecule that is the present invention may further comprise a linker downstream from said signal sequence and upstream from the DNA sequence which encodes the heterologous polypeptide. The present invention also relates to a plasmid expression vector comprising a recombinant DNA molecule as described above. Furthermore, the present invention relates to host cells transformed with the expression vector. The present invention also relates to a nutrient concentration sensitive Inducible promoter.

DETAILED DESCRIPTION OF THE INVENTION

The present invention provides the necessary genetic information to facilitate the production and proper folding of heterologous proteins in prokaryotes. According to the present invention, genetic sequences are provided which result in a reduction of protein degradation of the heterologous protein in the host prokaryote. Furthermore, the present invention provides genetic information which results in the transport of the heterologous protein into the periplasmic space of the prokaryote. Once in the periplasmic space, the heterologous protein is removed from exposure to cytoplasmic proteases. Additionally, the formation of proper secondary structure is facilitated in the periplasmic space. Finally, purification of the heterologous proteins located in the periplasmic space reduces protein degradation and disruption of secondary structure associated with purification from the cytoplasm.

In the present invention, the genetic information to direct transport of a heterologous protein in E. coli to the periplasmic space is derived from the periplasmic protein fpp . In certain strains of E. coli, e.g., JM101 and HB101, fpp constitutes about 90% of the periplasmic protein in the cell and about 10% of the total cell protein. Synthesis of the protein is repressed in rich media and induced upon shifting the cells into minimal media. Accordingly, one aspect of the present invention relates to the use of the fpp promoter in a recombinant expression vector. The fpp promoter may be inserted in a plasmid. A gene may be operably linked downstream from the promoter. Thus, expression of the gene will be regulated under the control of the promoter. Host cells can be transformed with the recombinant expression vector. In rich medium, the gene under the control of the fpp promoter will not be expressed. However, when the transformed cells are shifted to minimal media , the gene under the control of the fpp promoter will be expressed and the protein encoded by the promoter will be produced. Thus, the fpp promoter is a nutrient concentration sensitive inducible promoter. For the purposes of this specification, the term "nutrient concentration sensitive inducible promoter" means the promoter which endogenously regulates expression of the fpp gene.

The genetic information used in the present invention which is derived from the protein fpp encodes the signal sequence and optionally, at least one amino acid derived from the N terminus of the protein. Furthermore, the present invention may optionally comprise the fpp promoter which is induced by shifting host cells from rich to minimal media. A portion of fpp starting from the N terminal is included when the presence of the signal peptide linked directly to the heterologous protein does not result in protein transport to the periplasmic space. In order to effect the transport of some heterologous proteins, the topogenic sequences derived from fpp are required to be present. The amount of fpp peptide sequence necessary in a fusion polypeptide for efficient transport is variable but can be determined for different heterologous proteins through routine screening. Additionally, the present invention may optionally comprise the 5' untranslated region of the fpp gene. The fpp derived genetic information is used in an expression vector with a gene encoding a desired heterologous protein to provide a transformed bacterial cell the information to produce a heterologous protein and transport that protein to the periplasmic space where it will not be degraded by cytoplasmic proteases and will undergo proper folding. Furthermore, the active form of the protein can be purified from the periplasmic space with relative ease. The fpp derived genetic information can be attached to the gene for the desired heterologous protein with a linker that facilitates removal of the amino acids coded by the fpp derived DNA sequence from the heterologous protein. For the purposes of this specification, the following definitions apply to the following terms:

"Functional equivalents" means nucleotide sequences which differ from those described herein but which are used to encode the same amino acid sequence of the present invention.

"Functional fragments" means fragments of sequences described in the present invention which, although incomplete, yield the same results as the present invention.

"Operably linked" refers to the attachment of nucleotide sequences to form a functional group . For example, a promoter attached upstream from a gene and which exhibits transcriptional control over it is said to be operably linked to the gene. In cases where the attachment must be in the same reading frame to result in a functional group, such as the splicing of nucleotide sequences which are both to be translated, the nucleotide sequences are said to be operably linked if they are in the same reading frame.

"Linker" means the amino acid residue or residues which are recognized by cleavage means such as chemical substances or proteolytic enzymes as a site for severing the peptide bonds at or near that site. Several forms of linkers exist which all serve as a site for seperation of peptides. Examples of cleavage methods and linkers are listed in Table 1. The nucleotide sequence which encodes a linker may be inserted between the nucleotide sequences which encode peptide sequences where it is desired to seperate the peptide products posttranslationally. The linker must be inserted in proper reading frame and the nucleotide sequences flanking the linker must be maintained in proper reading frame.

"Heterologous peptide" is an amino acid sequence, polypeptide or protein which is not endogenously produced by the host. A heterologous peptide is produced by a host if the gene encoding it is introduced into the host by recombinant methods. Examples of heterologous peptides are human nerve growth factor, CD4, IL-1, IL-1 receptor, tissue plasminogen activator and bovine somatotropin.

"Transport sequence" is the genetic information which results in a heterologous peptide being transported to the periplasmic space of a host organism. The transport sequence comprises, at minimum, the fpp derived signal sequence or a functional fragment thereof, and may further comprise a fusion sequence, and a linker.

"Fusion sequence" is a nucleotide sequence derived from a portion of the protein fpp gene 3' of the signal sequence. The fusion sequence is used as part of the transport sequence to provide a heterologous peptide additional amino acid residues so it will be recognized by the protein transport mechanism and transported to the periplasmic space.

"Preleader sequence" is an untranslated nucleotide sequence derived from the 5' untranslated region of the protein fpp gene which is located 3' to the fpp promoter and 5' to the signal sequence.

"Signal sequence" is a nucleotide sequence derived from protein fpp which encodes the signal peptide. The signal sequence is the minimum nucleotide sequence which provides the genetic information that results in the transport of a heterologous peptide which the signal peptide is attached to. The signal sequence is the only component of the transport sequence which is required.

"Signal peptide" is the amino acid sequence encoded by signal sequence. The signal peptide, when attached to a heterologous peptide, can be recognized by the protein transport mechanism. The polypeptide is then transported to the periplasmic space.

"Fusion peptide" refers to the resulting polypeptide when a fusion sequence linked to a heterologous gene is transcribed. A fusion peptide is a heterologous peptide with at least a portion of the N terminus of protein fpp attached to it.

"Heterologous gene" is a gene or nucleotide sequence which encodes a heterologous peptide.

"Variably truncated fusion polypeptide" is a fusion peptide in which the fpp protein portion derived from the N terminus of fpp can comprise the amino acids from 1-173. Depending on the specific heterologous peptide, the number of amino acid residues from the N terminus portion of fpp necessary to facilitate optimum transport to the periplasmic space varies. In some cases, the signal peptide alone is sufficient. For others, the presence of fpp protein sequences is necessary. The amount of fpp protein necessary in the fusion protein varies depending upon the heterologous peptide.

Starting from the N terminal (Amino acid 1), different lengths of

Amino acid seguences may be used, (i.e., 1-2, 1-3, 1-4, ... 1-198, 1- 199, 1-200). Chart 2 shows the amino acid sequence of protein fpp.

A variable fusion polypeptide may comprise any subsequence of this sequence starting from the N-terminus and extending toward the Cterminal.

Chart 1 contains DNA sequences derived from protein fpp of E. coli determined using methods well known to one having ordinary skill in the art. The DNA sequence which encodes fpp as well as the 5'untranslated region and the signal sequence have been determined and are shown in Chart 1. The specific portions of the sequence which encodes the 5' untranslated region, the signal sequence and the protein gene are labeled in Chart 1. The amino acid sequence which is encoded by the fpp nucleotide sequence is also shown in Chart 1.

Additionally, Chart 2 depicts a portion of the amino acid sequence of protein fpp starting at the N terminal (Amino acid 1) extending toward the C terminal to Amino acid 173. This sequence or a subsequence starting from amino acid 1 may be used in chimeric peptides having the fpp signal peptide and a heterologous peptide.

Conventions used to represent plasmids and fragments in Charts 3-5, though unique to this application, are meant to be synonymous with conventional representations of plasmids and their fragments. Unlike the conventional circular figures, the single line figures on the charts represent both circular and linear double-stranded DNA with initiation or transcription occurring from left to right (5' to 3'). Asterisks (*) represent the bridging of nucleotides to complete the circular form of the plasmids. Fragments do not have asterisk marks because they are linear pieces of double-stranded DNA. Endonuclease restriction sites are indicated above the line . Gene markers are indicated below the line.

Chart 3 depicts various embodiments of the present invention. Fragment 1 is a plasmid comprising a promoter, an insert and optionally a 5' untranslated in region intermediate to both components. The insert can consist of DNA fragments shown in Fragments 2, 3, and 4 of Chart 3.

The promoter is operatively linked to the insert. That is, the promoter is functional and is linked to the insert such that it can regulate control expression of the insert. The 5' untranslated region is optionally included and if present, is downstream from the promoter and upstream from the insert. The presence of the 5' untranslated region may be helpful to achieve high levels of expression.

Fragment 2 shows the minimal construction according to the present invention. The signal sequence derived from protein fpp is linked to the heterologous gene sought to be expressed. The two genetic units are fused in proper reading frame such that the signal peptide encoded by the signal sequence is translated and attached adjacent to the heterologous protein encoded by the heterologous gene.

Fragment 3 includes the signal sequence and heterologous gene with a fusion sequence immediate thereto. The fusion, sequence comprises the end terminal portions of the fpp protein directly downstream from the signal sequence. In fragment 3, the signal sequence is linked upstream from the fusion sequence which is linked upstream from the heterologous gene. The three components are linked in proper reading frame such that the signal peptide encoded by the signal sequence is directly adjacent to the amino acid sequence encoded by the fusion sequence which is adjacent to the heterologous protein encoded by the heterologous gene when the entire construction is translated.

Fragment 4 is essentially the same construction as fragment 3 further comprising a linker sequence between the fusion sequence and the heterologous gene. The linker sequence is, as consistent with the 2 and 3 constructs, in proper reading frame with the other genetic components. The linker encodes an amino acid sequence which is recognized by chemicals or enzymes as a proper cleavage site. Thus, after translation, the signal peptide and amino acid sequence encoded by the fusion sequence may be cleaved from the heterologous protein by enzymatic or chemical means.

To practice the present invention, the transport sequence must be operably linked to the heterologous protein gene and inserted downstream from a promoter in an expression vector. At minimum, the transport sequence consists of the signal sequence. In some cases, the amino acid signal sequence will not be removed by the host cell in post translational processing or transport mechanisms.

In some circumstances, the presence of the amino acid signal sequence alone will not result in the transport of heterologous peptide to the preiplasmic space. However, if a portion of the fpp protein from the N terminus is included between the amino acid signal sequence and the heterologous peptide, transport will take place.

This results in a fusion peptide wherein the N terminus of the heterologous polypeptide contains the N terminus of protein fpp.

Attached to the amino end of the fpp N terminus is the signal sequence. Since the transport mechanism is not fully understood, it cannot be predicted how much, if any, of the N terminal portion of protein fpp must be present to achieve optimal transport of any given heterologous peptide to the periplasmic space. However, techniques well known to those skilled in the art make such a determination routine.

All manipulations used to design the heterologous peptide to be transported to the periplasmic space are done at the DNA level and the resulting recombinant molecule is incorpated into an expression vector which is then used to transform suitable prokaryotic host cells. Thus, in constructing the expression vector, the signal sequence must be inserted downstream from the promoter and upstream, adjacent and in proper reading frame with the heterologous gene. In some circumstances it may be preferred to include the 5' untranslated region, termed the preleader sequence, between the promoter and the signal sequence.

To determine the optimal amount of fpp residues in the variably truncated fusion peptide, prior to ligation of the heterologous gene or linker/heterologous gene to the fpp derived DNA, the fpp material may be "chewed back" from 3' to 5' to generate a series of construction having variable numbers of nucleotides encoding N terminal amino acid residues. The techniques used to do this are well known to those skilled in the art and acccordingly, although determining the the optimal number of fpp reidues in a fusion peptide may be time consuming, the procedures are quite routine and no undue experimentation is necessary. Described briefly, the DNA containing the fpp signal sequence and nucleotides encoding the N terminal residues are divided into a number of seperate samples and each sample is treated with an enzyme which digests DNA from 3' to 5' (from the N terminal residue codons toward the signal sequence). A time series is performed wherein the enzyme reaction is halted (for example by raising the temperature of the reaction mixture sufficient to inactivate the enzyme) for separate samples at a series of reaction times. Thus, the samples subjected to longer reaction times have more bases chewed off; that is, the samples subjected to longer reaction times have less N terminal residue codons than the samples which undergo digestion for a short time. All samples are then used in making expression vectors for the heterologous peptide. The vectors are used to transform suitable hosts and the protein in the periplasmic space is harvested. The results are compared and it can be readily determined which construction yields protein most efficiently transported to the space .

The present invention is seen more fully by the Examples below, which are not meant to be limiting.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Example 1 Construction of pfppP.

The plasmid pfppP is a recombinant DNA construct that contains the nucleotide sequence which encodes the "P" domain from human tissue plasminogen activator (tpa) attached downstream from the fpp signal sequence. The construct is under regulatory control of the tac promoter.

To construct plasmid pfppP, the following DNA fragments are isolated as described and ligated together. Chart 4 depicts the construction of pfppP.

a) A DNA fragment containing the beginning of the protein fpp gene and 150 bp upstream of fpp is obtained from the chromosome of E. coli (preferably JM101 or W3110) . The genomic DNA is digested with BamHI restriction enzyme and DNA fragments from 1.2-1.6 kbp isolated from an agarose gel and ligated into any standard E. coli cloning vector that accepts BamHI fragments (preferably plasmid pBR322). The plasmid containing the DNA coding for fpp can be identified by hybridizing the colonies containing the recombinant plasmids with a synthetic oligonucleotide:

5 ' GACCGTCGCAGCAAGTGTTCAGGCTAAAACTCTGGTTTATTGCTCAG

using standard techniques. A BamHI-PvuII DNA fragment can then be isolated that contains 150 bp of untranslated region upstream of fpp, the signal sequence of fpp and 17 codons beyond the signal sequence cleavage site. DNA containing an EcoRI site is added onto the BamHI end, resulting in an EcoRI-PvuII fragment (Fragment 5).

b) Fragment 6 is the plasmid pKK.223-3 (Pharmacia Inc.) linearized by complete digestion with EcoRI and partial digestion with BamHI so that near full-length molecules are generated.

c) Fragment 7 contains a portion of the tpa molecule containing the active site of the molecule (P domain) and some flanking sequences. This fragment is an Fspl-BglII fragment isolated from plasmid pUC1188 (described by Rehberg, E.F. et al., 1989. Protein Engineering, Vol. 2, pp. 371-377).

The three fragments are ligated together using standard techniques, used to transform a competent E. coli containing high levels of lac repressor (preferably JM101) and a recombinant plasmid containing the fragments in the location and orientation shown in Fragment 8 is identified from the resulting plasmids using restriction enzyme analysis and DNA sequencing.

Example 2 Production of fpp-P Domain Fusion Protein in JM101 Cells Which Contain pfppP.

Cells are grown in a media suitable for E. coli growth (LB for rich media and M9 for minimal media are typical examples) to early log phase. Transcription of the fusion gene is induced by IPTG at a concentration of 0.5 mM when the cells are in early log phase. Cells are removed at subsequent times post-IPTG addition, pelleted by centrifugation and extracts prepared from them as follows. If total cellular material is assayed, the cell pellet is resuspended in a small volume of buffer (10 mM tris-HCl, 1 mM EDTA, pH 8.0, for example) and sonicated to disrupt the cells. The resulting mixture is cleared by centrifugation and the resulting supernatant assayed for P domain activity. Proteins in the periplasmic fraction can be recovered following the procedure of Nossal and Heppel (Nossal, N.G. and L.A. Heppel, 1966. The Journal of Biological Chemistry, Vol. 241, pp. 3055-3062). The membrane and cytosolic fractions can be isolated using the procedure of Silhavy et al. (Silhavy, T.J., Casadaban, M.J., Shuman, H.A. and Beckwith, J.R., 1976. Proc. Natl. Acad. Sci. USA, Vol. 73, pp. 3423-3427).

Example 3 Assay For Active P Domain in E. coli Extracts.

Active P domain activity can be assayed by a modification of the method of Verheijen et al., (Verheijen, J.H. et al., 1982. Thromb. Haemostasis, Vol. 48, 266-269) as described by Zaworski and Gill (Zaworski, P.G. and Gill, G.S. 1988. Analytical Biochemistry, Vol. 173, pp. 440-444). Extracts from E. coli can either contain total cellular material or fractions from various portions of the cell (inner membrane, periplasmic, outer membrane, cytoplasmic).

Example 4 Bal31 Digestion of fpp DNA to Identify Sequence That Is Optimal for Transporting a Heterologous Protein Outside of the Cytoplasm.

The fpp gene used in this example is isolated from chromosomal DNA from JM101 that had been digested with HindIII and PstI and cloned into the cloning vector pUC18. The resulting library of plasmids is screened with the oligonucleotide shown in Example 1 for a plasmid that contains the fpp gene missing the first part of its signal sequence. The fpp fragment is excised by digestion with

HindIII and SmaI. As depicted in Chart 5, the plasmid. used for generating deleted fpp fragments to fuse to a heterologous gene is pffpp (Fragment 12). It is composed of pBR322 from the BamHI site to the EcoRI site (Fragment 9), a synthesized DNA fragment from the

EcoRI site to the HindIII site of sequence shown in Chart 5 and containing the start of the fpp signal sequence (Fragment 10) and the

HindIII-SmaI fragment containing the rest of fpp (Fragment 11).

Other plasmids containing fpp and other promoters would also work.

Plasmid pffpp is linearized by the restriction enzyme BglII near the 3' end of the gene and treated with Bal31 for various lengths of time to generate a heterogeneous population of 3' truncated fpp derivatives. The DNA (10μg) is resuspended in a final volume of 100 microliters of 1X Bal31 buffer (12 mM CaC12,12 mM MgC12 , 0.2 mM NaCl, 20 mM Tris-Cl pH 8.0, 1 mM EDTA) and 0.5 unit of enzyme (more or less as determined by the particular batch of enzyme and DNA) and incubated at 30°C. At various times, aliquots are removed, a small amount of the DNA from each fraction Is digested with the enzyme

HindIII and the DNA sized on an agarose gel. DNA from the fractions that gave fragments ranging from 100 bp-1400 bp is pooled, the ends repaired using dNTP's and Klenow and a linker added that contains a sequence to allow for fusing the fpp coding sequence in-frame with the coding sequence of the peptide or heterologous protein of interest. The resulting DNA is cut with another enzyme to liberate the deleted fpp fragments (preferably HindIII, but others will work depending on what vector is subsequently used) and ligated to a plasmid containing the first part of the fpp (from the Met codon to the HindIII site) downstream from a regulatable promoter on a plasmid and the gene of interest such that the fpp fragment would be inserted between the HindIII site and the gene of interest to generate a single coding region in the following order: fpp signal sequencetruncated fpp fragments-heterologous gene. After introducing these molecules into a suitable E. coli host by standard methods, they are screened for ones containing variable lengths of fpp between the signal sequence and the heterologous gene. Representative clones containing a range of fpp segments, each one longer by about 100 bp than the next shorter one and beginning from around 100 bp downstream from the signal sequence cleavage site to the end of fpp are chosen and checked for their ability to direct transport of the fpp-heterologous gene fusion protein outside of the cytoplasm. This procedure will allow for identifying the amount of fpp peptide that is optimal for directing transport of any other peptide. When required, a DNA linker coding for amino acids that would create a cleavage site for a chemical protease cleavage can be inserted between the heterologous gene and the fpp sequences using standard techniques to aid in the isolation and purification of the heterologous protein away from the fpp peptide.

Example 5 Procedure for Fusing a Heterologous Gene to the fpp Signal Sequence .

To transport a heterologous protein out of the cytoplasm using only the signal sequence from fpp (i.e., no other fpp coding sequence) a plasmid is constructed that contains a restriction enzyme site at or near the signal sequence cleavage site that allows for the addition of a heterologous gene at that site. The "P" domain fragment is fused directly to the fpp signal sequence by changing the sequence at the signal sequence cleavage site to include a SacII restriction site. Plasmid pffpp is digested with HindIII and BamHI and the ori containing fragment is ligated with oligonucleotides of the following sequence:

HindIII SacII BamHI

5 'AGCTTGGTCTCAGCCTGGTGGCTATGACCGTCGCAGCAAGTGTTCAGGCCGCGGCCG

3' ACCAGAGTCGGACCACCGATACAGGCAGCGTCGTTCACAAGTCCGGCGCCGGCCTAAG

The resulting vector, pffpp-2 is cut with SacII, digested with mung bean nuclease to- remove the single stranded overhang and then digested with BamHI. The Fspl-BglII fragment containing the P domain coding region described above is then inserted into the SacII-BamHI treated vector. The resulting plasmid will have the following sequence at the fpp signal sequence-P domain junction:

GTT CAG GCC GC

val gin ala ala

signal

sequence P domain

The restriction site chosen to incorporate at the cleavage site can vary depending on the gene to be fused to, as well as if the heterologous protein can have one or two additional amino acids at the amino terminus.

Example 6 Insertion of Linkers Between fpp Sequence and Heterologous

Gene .

Linker sequences between the fpp portion and heterologous gene can be incorporated so that the two portions of the molecule can be separated after expression and preliminary purification of the protein. Several linker sequences that are recognized by different proteolytic treatments can be used. Depending on the sequence of the gene of interest, the linker would have DNA sequences coding for one of the amino acid sequences listed in Table 1 that would be recognized by the companion treatment.

TABLE 1

Linkers Cleavable by Enzymes and Chemical Techniques

Linker Sequence* Cleaved by

-Met-x- Cyanogen bromide

-Asp-Pro-x- Acid

-Asp-Gly-x- Hydroxylamine

-Ile-Glu-Gly-Arg-x- Factor X_a

-Arg(Lys)-x- Trypsin

-Pro α-2 Collagen-x Collagenase

-(Asp)4-Lys-x- Enterokinase

x stands for any given protein sequence.

-27

Claims

1. A recombinant DNA molecule comprising a signal sequence comprising the nucleotide sequence encoding an amino acid sequence selected from the group consisting of MetArglleSerLeuLysLysSerGlyMetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla, MetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla and functional equivalents thereof.

2. A recombinant DNA molecule according to Claim 1 comprising the nucleotide sequence encoding the amino acid sequence MetArglleSerLeu¬

LysLysSerGlyMetLeuLysLeuGlyLeuSerLeuValAlaMetThrValAlaAlaSerValGlnAla.

3. A recombinant DNA molecule according to Claim 2 wherein said recombinant DNA molecule comprises the sequence ATG CGT ATT TCC TTG

AAA AAG TCA GGG ATG CTG AAG CTT GGT CTC AGC CTG GTG GCT ATG ACC GTC GCA GCA AGT GTT CAG GCT.

4. A recombinant DNA molecule according to Claim 1 further comprising a heterologous gene downstream from and in proper reading frame with said signal sequence.

5. A recombinant molecule DNA according to Claim 1 further comprising a DNA sequence encoding a variably truncated fusion polypeptide sequence downstream from and in proper reading frame with said signal sequence.

6. A recombinant DNA molecule according to Claim 5 further comprising a heterologous gene downstream from and in proper reading frame with said variably truncated fusion polypeptide sequence.

7. A recombinant DNA molecule according to Claim 5 further comprising a linker downstream from and in proper reading frame with said variably truncated fusion polypeptide sequence.

8. A recombinant DNA molecule according to Claim 7 further comprising a heterologous gene downstream from and in proper reading frame with said linker.

9. An expression vector comprising a recombinant DNA molecule according to Claim 1.

10. An expression vector according to Claim 9 wherein said recombinant DNA molecule further comprises a heterologous gene downstream from and in proper reading frame with said signal sequence.

11. An expression vector according to Claim 9 wherein said recombinant DNA molecule further comprises a DNA molecule encoding a variably truncated fusion polypeptide sequence downstream from and in proper reading frame with said signal sequence.

12. An expression vector according to Claim 11 wherein said recombinant DNA molecule further comprises a heterologous gene downstream from and in proper reading frame with said variably truncated fusion polypeptide.

13. An expression vector according to Claim 11 wherein said recombinant DNA molecule further comprises a linker downstream from and in proper reading frame with said variably truncated fusion polypeptide sequence.

14. An expression vector according to Claim 13 wherein said recombinant DNA molecule further comprises a heterologous gene downstream from and in proper reading frame with said linker.

15. Host cells transformed with an expression vector according to Claim 9.

16. A method of producing heterologous polypeptides in prokaryotes comprising the steps of:

a) constructing a recombinant DNA molecule according to Claim

1;

b) linking a gene for a desired heterologous polypeptide downstream from and in proper reading frame with said signal sequence of said recombinant DNA molecule;

c) inserting said recombinant DNA molecule linked to said gene into an expression vector;

d) transforming suitable host cells with said vector;

e) harvesting polypeptides from the periplasmic space of said host cells;

f) purifying said desired polypeptide from said polypeptides isolated from said periplasmic space.

17. A recombinant DNA molecule comprising nutrient concentration sensitive inducible promoter.

18. A recombinant DNA molecule according to Claim 17 further comprising a heterologous gene operably linked to said promoter.

19. A recombinant DNA molecule according to Claim 17 wherein said recombinant DNA molecule is an expression vector.

20. A recombinant DNA molecule according to Claim 18 wherein said recombinant DNA molecule is an expression vector.