Hybrid DNA synthesis of epidermal growth factor

Abstract

DNA sequences and methods of obtaining DNA sequences which include a sequence encoding for mammalian epidermal growth factor are provided. The DNA sequences may be used in cloning and expression vectors for production of DNA and RNA and for producing polypeptides including mammalian epidermal growth factor.E. coli strains HB 101 (pmegf1) and HB 101 (pmegf10b), and bacteriophage .lambda.hEGF34 were deposited at the A.T.C.C. on May 10, 1983 and given Accession Nos. 39357, 39358 and 40070, respectively.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

Epidermal growth factor (EGF) is a polypeptide of 53 amino acids that has been characterized in both mice and humans. It is a potent mitogen for a variety of cells, such as fibroblasts, glia, epithelial, endothelial and epidermal cells, both cultured and in vivo. EGF is also a potent inhibitor of gastric acid secretion. EGF was first isolated from male mouse submaxillary glands, where it exists in inexplicably high levels.

In glandular homogenates, EGF is found as a 74,000 dalton complex of two molecules of EGF (Mr 6045) and two molecules of a binding protein (Mr 29,300), a kallikrein-like arginyl enteropeptidase The amino acid sequence of mouse submaxillary EGF has been determined and the synthesis of EGF and a larger 9000 dalton precursor with a carboxy terminal extension has been demonstrated in cultured submaxillary glands.

Human EGF, which appears to be similar if not identical to urogastrone, is also found in urine in larger forms of 28,000 and 30,000 Daltons that do not dissociate on sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

Isolating both the DNA or RNA encoding for EGF and particularly a putative EGF precursor protein is extremely difficult for a number of reasons. Even where the peptide is abundant, the amount of messenger RNA is extremely small. Immunoprecipitation of in vitro translation products, even under strongly denaturing conditions fails to detect a precursor protein, possibly due to the huge size of the precursor and/or the masking of its antigenic determinants on the native peptide to which antibodies were made. Because of the physiological importance of EGF there is substantial interest in being able to obtain DNA sequences encoding for EGF and the EGF polypeptide precursor. In addition, since it is known that a number of hormones are generated by proteolytic processing from larger precursors the cDNA and derived amino acid sequence of the EGF precursor could reveal "cryptic", previously unknown polypeptide hormones and/or growth factors.

2. Description of the Prior Art

A human genomic cDNA library in bacteriophage .lambda. is described in Lawn et al., Cell (1978) 15:1157-1174. Savage et al., J. Biol. Chem. (1972) 247:7612-7621 report the amino acid sequence of mouse EGF. Sporn et al., Science (1983) 219:1329-1331 and Assoian et al., J. Biol. Chem. (1983) 258:7155-7160 describe transforming growth factors (TGF). See particularly Gray et al., Nature (1983) 303:722-725.

SUMMARY OF THE INVENTION

DNA and RNA are provided encoding for mammalian EGF, polypeptide precursors thereof, and numerous other polypeptides also encoded for by the DNA sequence which includes the segment encoding for EGF. The DNA sequences may be used for production of mammalian EGF, precursors of mammalian EGF, and associated related polypeptides. Employing radiolabeled hybridization probes, messenger RNA encoding for the mouse EGF precursor is detected, isolated and used for the production of cDNA. The cDNA is sequenced and a fragment employed for hybridization with human DNA under conditions where mismatched heterologous hybrids can be detected. In this manner, a DNA sequence encoding for a large precursor peptide encompassing human EGF or urogastrone and numerous, related polypeptides is detected and isolated.

DESCRIPTION OF SPECIFIC EMBODIMENTS

In accordance with the subject invention, DNA and RNA sequences are provided which include a segment encoding for EGF, particularly mouse and human, as well as the expression products of these sequences, including propolypeptides and peptides, particularly peptides having one or more physiological functions of EGF, or peptides having one or more other hormonal or growth factor regulatory functions.

The DNA sequences of interest are single or double stranded ranging from about 60 bases or base pairs (bp) to about 5 thousand base pairs (kbp), where sequences encoding for a physiologically active polypeptide will generally range from about 60 bp to about 1000 bp, which may include introns. Generally, the DNA sequences will have open reading frames (involving one or more exons) encoding for polypeptides ranging from about 20 amino acids to polypeptides about 1000 amino acids, where the sequence will include at least 2, usually at least 5, more usually at least 10 bp, outside the segment encoding EGF. Polypeptides of particular interest will generally be of from about 20 to 250 amino acids, more usually from about 20 to 175 amino acids, particularly 30 to 60 amino acids.

The DNA segments encoding for polypeptides of interest or mature polypeptides may be located in any region in the DNA sequences described in this invention. Of particular interest are sequences bordered by basic amino acids, i.e. arginine and lysine, more particularly when joined to a second basic amino acid, or alanine, leucine, aspartic or glutamic acid or amide thereof. Amongst these sequences of interest are seven previously unknown EGF-like polypeptides with amino acid sequences homologous to EGF. The DNA sequences obtained in accordance with this invention were obtained by the following experimental design.

A mammalian cDNA or genomic DNA library is screened with a plurality of radiolabeled hybridization probes for detection of a sequence encoding for an amino acid sequence present in EGF. A plurality of probes are employed providing for the various possible redundant codons encoding for the oligopeptide. In the subject method a cDNA library from mouse submaxillary gland cells was probed. Plasmids binding strongly to the probes are isolated and the several overlapping cDNA inserts sequenced. The mouse EGF encoding cDNA has about 4800 bases. Mouse EGF is encoded for by nucleotides 3281-3440.+-.5, with an open reading frame encoding for 1217.+-.5 amino acid residues and a protein of approximately 130-140 kilodaltons (kdal), particularly about 133 kdal.

The mouse cDNA may then be used to probe a human cDNA or genomic DNA bank. Conveniently a restriction fragment may be employed of about 500 to 1500 bp. Particularly, a BstEII-PvuII fragment of about 1213.+-.5 bp, may be employed The hybridization is carried out under conditions which facilitate the detection of mismatched, heterologous hybrids. The BstEII-PvuII fragment encodes mouse EGF (53 amino acids) and 286 amino acids before and 66 amino acids after the EGF moiety. Clones which hybridized to the probe were isolated and the human DNA inserts characterized.

Once the DNA sequence is isolated, it can be used in a variety of ways: For production of synthetic DNA sequences, either in whole or in part, for replication, or for the production of messenger RNA or expression of the precursor protein incorporating EGF, fragments of such protein or of EGF, or analogs of EGF, differing by one or more amino acids, usually by not more than about five amino acids from the naturally occurring EGF amino acid sequence.

Various DNA sequences are of particular interest in encoding polypeptides which can be obtained from the cDNA sequence. These DNA sequences are set forth in the argument map set forth in the experimental section, along with the human EGF sequence. The polypeptide sequences of interest include, but are not limited to, seven previously undescribed, EGF-like polypeptides identified on the basis of the homology of their amino acid sequences to EGF, especially the positional relationship(s) of the several cysteine residues. (See diagram in experimental section.) These sequences are frequently bounded by one or more basic amino acids.

Once the desired DNA sequence encoding for a protein or peptide of interest, e.g. EGF or its homologues, has been isolated, it can be joined with other DNA sequences for replication and expression. A wide variety of vectors are available for unicellular microorganisms, particularly for bacteria and fungi, where the DNA sequence encoding the poly(amino acid) of interest may be replicated and/or expressed.

Various hosts of interest include E. coli, S. cerevisiae, B. subtilis, mouse 3T3 cells, or the like. Conventional vectors include replication systems derived from R6-5, ColEI, the 2 .mu.m plasmid from yeast, RK plasmids, or the like. Alternative replication systems may be derived from viruses or phage, such as lambda, SV40, etc. In some instances, it will be desirable to have two different replication systems, where different functions may be achieved in different hosts. These vectors, referred to as shuttle vectors, frequently employ a replication system for E. coli and a replication system for a higher organism, e.g. yeast, so that amplification of the gene or cloning may be achieved in the bacterium, while expression may be achieved in the higher organism with appropriate processing, e.g. glycosylation.

Conveniently included with the replication system is at least one marker, which allows for selection or selective pressure to maintain the DNA construct containing the subject DNA sequence in the host. Convenient markers include biocidal resistance, e.g. antibiotics, heavy metals and toxins; complementation in an auxotrophic host; immunity; etc. The DNA sequence including the fragment encoding for a polypeptide having epidermal growth factor physiological properties or fragments of such sequence may be replicated in a cloning vector, which is capable of replication in a unicellular microorganism, such as bacteria and yeast. The DNA may also be used in an expression vector for expression of a polypeptide of interest, which may be mammalian EGF, particularly mouse or human, other physiologically active polypeptides present in the sequence, e.g. other hormones or growth factors, fragments thereof or analogs thereof differing by from about one to five amino acids.

The open reading frame of the DNA sequence allows for the production of a large polypeptide. The large polypeptide may be treated in a variety of ways. The large polypeptide may be partially digested with a variety of proteases either individually or in combination. Illustrative endopeptidases include trypsin, pepsin, membrane dipeptidases, esteropeptidases or the like. The resulting fragments may then be separated by charge and/or molecular weight by any conventional means, e.g. filtration, sedimentation, chromatography, electrophoresis, or the like and then tested for physiological activity. Of particular interest are growth factors acting as mitogens or differentiation regulators. Based on the activities observed, the various fractions may be further purified by bioassays to obtain pure active factors.

The DNA sequences of this invention can be used in a variety of ways. Fragments can be used as probes for detecting complementary sequences in genomic DNA or in messenger RNA for detecting mutations and/or deletions in genomic DNA of hosts. The sequences can be used for expressing the polypeptides encoded for by the sequence.

The following examples are offered by way of illustration and not by way of limitation:

EXPERIMENTAL

Methods

cDNA Synthesis and Construction of Recombinant Plasmids

To construct the cDNA library, polyA-containing RNA was isolated from the submaxillary glands of 60-day-old male Swiss-Webster mice. ds cDNAs were prepared and inserted into the PstI site of a pBR322 derivative using the dGdC tailing technique (Chirgwin et al., Biochemistry (1979) 18:5294-5299; Goodman and MacDonald, Methods in Enzymol. (1980) 68:75-90). Resultant tetracyline-resistant transformants of E. coli HB101 were stored at -70.degree. C. in microtiter dishes (Gergen et al., Nucl. Acids Res. (1979) 7:2115-2136; Ish-Horowicz and Burke, ibid. (1981) 9:2989-2998).

Oligonucleotides were then synthesized by solid-phase phosphoramidite methodology as described in copending application, Ser. No. 457,412, followed by isolation from 20% acrylamide gels modification of the method described in Beancage and Camthers, Tetrahedron Lett. (1981) 22: 1859-1862. Dodecamers were prepared which were complementary to the strand coding for amino acids 17 to 23 (lacking the last 5'-nucleotide) of mouse EGF cDNA. The fractions had the following sequences, where after the addition of the eleventh nucleotide, two pools were prepared, one terminating in A and the other terminating in G and after addition of the seventeenth nucleotide, the two pools were further divided with addition of the eighteenth nucleotide, with two of the pools now terminating in G and the other two pools terminating in A. In this manner, a total of four pools were obtained, where each pool had a plurality of eicosamers of differing compositions at positions 3, 6 and 9. ##STR1##

The different sequences are required because of the uncertainty as to the specific codon usage due to the redundancy of the genetic code for amino acids.

Synthetic oligonucleotides were labeled with adenosine 5'-(.gamma.-.sup.32 P) triphosphate (ICN, crude preparation, 7000 Ci/mmol, lCi=3.7.times.10.sup.10 Bq) by a polynucleotide kinase reaction (Wallace et al., Nucl. Acids Res. (1981) 9:879-894). The labeled oligonucleotides were separated from unincorporated (.gamma.-.sup.32 P) triphosphates by chromatography on a C-18 Sep-Pak.TM. column (Waters Associates, Inc.) as follows: The crude labeling mixture was applied (disposable syringe) to the Sep-Pak cartridge which was then washed wtih 20 ml of water to elute the unincorporated adenosine 5'-(.gamma.-.sup.32 P) triphosphate. The radiolabeled oligonucleotide was then eluted with 1:1 (v/v) methanol: 0.1M triethylammonium acetate (pH 7.3) and the eluate evaported to dryness. The specific activity of the probe was of the order of 10.sup.8 -10.sup.9 cpm/.mu.g.

Transformants were grown on Whatman 541 filter paper, the plasmids amplified in situ with chloramphenicol and the DNA immobilized on the filters (Gergen et al., supra). .sup.32 P end-labeled probes were used to search the library. Additional screening was with nick translated cloned cDNA, with the same filters being used repeatedly.

The sequence of cDNA inserts were determined by the Maxam and Gilbert method.

mRNA Size Estimation

Glyoxylated total RNA from male and female mouse submaxillary glands was separated on 2% agarose gels, transferred to nitrocellulose and hybridized with a nick translated .sup.32 P-labeled PstI-PstI fragment of the EGF cDNA insert. After washing, the RNA was autoradiographed at -b 70.degree. C. using an intensifying screen. Glyoxylated HindIII .lambda. and .phi.174 RF HaeIII digested DNA fragments were used as size markers.

Results

Screening of the cDNA Library

5000 transformants were initially screened, where 11 colonies yielded strong signals with pool 4 probes, the pools having the nucleotides at positions 12 and 18 of G and A, respectively. Weaker, but definitely positive signals were obtained with pool 3 which had the nucleotides G and G, respectively. Pools 1 and 2 gave no positive signals. The largest clone was 1800 bp. Terminal restriction and other fragments of this clone were used to screen the original 5000 plus 7500 additional colonies (12,500 total) and yielded additional overlapping cDNA colonies which did not contain the EGF sequence. Since it was subsequently determined by DNA sequence analysis (vide infra) that these overlapping clones lacked the 5'-terminal region of the mRNA, another cDNA library was synthesized using an oligonucleotide primer complementary to nucleotides 1032-1051 (see argument map, infra) as follows:

3'-CCGCTTCCTTCGGTGCGAAT-5' and this library was then screened as described above. The relative abundance of the cDNA clones in the initial library suggests that EGF mRNA comprises about 0.2% of the polyA.sup.+ m RNA from this tissue.

mRNA Sequence

The size of mouse EGF mRNA was determined by Northern analysis of mRNA from adult male and female glands to be about the same size as 28s ribosomal RNA, approximately 4800 bases. The mRNA in the male gland was at least ten-fold greater in abundance than in the female gland. The nucleotide sequence of overlapping cDNA clones provided 4750 bp of sequence as follows: ##SPC1##

This includes the exact 53 amino acid residue sequence of mouse EGF (nucleotides 3281-3440), a translational start codon AUG (nucleotides 354-356) and a stop codon TAG (nucleotides 4005-4007). An open reading frame throughout the sequence which encodes for 1217 amino acid residues and a protein of approximately 133 kdal. Also, seven additional EGF-like polypeptides are identified on the basis of the homology of their amino acid sequences to EGF, especially the positional relationship of their cysteine residues, as shown below: ##SPC2##

Human EGF Gene

A .sup.32 P-labeled (O'Farrell, Focus (1981) 3:1-3) 1,213 bp BstEII-PvuII fragment of mouse submaxillary EGF cDNA clone, pmEGF10, was hybridized to a human genomic DNA library (Lawn et al., Cell (1978) 15:1157-1174) in bacteriophage .lambda. (available from Dr. T. Maniatis, Harvard University) using conditions which facilitate the detection of mismatched-heterologous hybrids. The BstEII-PvuII fragment of pmEGF10 encoded mouse EGF (53 amino acids) and 286 amino acids amino terminal to and 66 amino acids carboxy terminal to the EGF moiety. The hybridization conditions were 50% formamide, 5X SSC, 10% dextran sulfate, 20 mM sodium phosphate, pH 6.5, 100 .mu.g/ml sonicated, denatured salmon testes DNA, and 0.1% sodium dodecyl sulfate at 30.degree. C. (Wahl et al., Proc. Natl. Acad. Sci. USA (1979) 76:3683-3687). The filters were washed for one hour at 50.degree. C. in 1M NaCl (Perler et al., Cell (1980) 20:555-566) before autoradiography. Four of the approximately 10.sup.6 phage screened hybridized to the probe. Characterization of the human DNA inserts in these phage indicated that they represented overlapping DNA segments from the same region of the human genome. The partial sequence of the human DNA in .lambda.hEGF35, corresponding to the exons encoding EGF or uragastrone, indicated that these phage contained portions of the human EGF gene.

The human EGF gene was sequenced and the mouse and human sequences compared. The amino acid sequences of EGF from the two species are described by Carpenter, In: Tissue Growth Factors, Handbook of Experimental Pharmacology, R. Baseraga (ed.), Vol. 57, Springer-Verlag, Berlin, 1981, p. 94.

In accordance with the subject invention, polynucleotide sequences are provided which encode for a large polypeptide which includes the amino acid sequence of EGF. The large polypeptide can be used as a source of polypeptides having physiological activity. In particular, seven additional EGF-like polypeptides are identified. The DNA sequences can be used for production of the large polypeptide or fragments thereof by employing recombinant DNA technology and inserting the polypeptide sequence downstream from an appropriate promoter in a functioning episomal element. The episomal element may then be introduced into an appropriate host for replication and expression of the desired polypeptide.

Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it will be obvious that certain changes and modifications may be practiced within the scope of the appended claims.

Claims

1. A composition comprising DNA molecules containing a nucleotide sequence encoding human epidermal growth factor (EGF) and substantially free of DNA molecules that do not contain said nucleotide sequence, said EGF-encoding nucleotide sequence comprising: ##STR2##

2. The composition of claim 1 whrein said DNA molecules are episomal elements.

3. A composition comprising cells containing an intron free DNA sequence encoding human epidermal growth factor (EGF) substantially free of cells not containing said EGF-encoding DNA sequence, said EGF-encoding DNA sequence comprising: ##STR3##

4. The composition of claim 3 wherein said cells are bacteria.

5. The composition of claim 3 wherein said cells are yeast.