DNA Preparation of Christmas Factor and Use of DNA Sequences

1. Field of the Invention

Factor IX (Christmas factor) is a vitamin K-dependent plasma protein that plays an important role in the middle phase of blood coagulation. Individuals lacking this protein may bleed spontaneously into their skin, soft tissues and joints. This bleeding is often serious in patients even after a minor injury. A deficiency of factor IX (Christmas disease or hemophilia B) affects males primarily, since it is transmitted as a sex-linked recessive trait. The human and bovine proteins contain 12 .gamma.-carboxyglutamic acid residues in their amino-terminal regions. During the coagulation process, factor IX is converted to factor IX.sub.a (a serine protease) by factor XI.sub.a, Fujikawa et al., Biochemistry (1974) 13:4508-4516. This factor then reacts with factor X and by successive steps coagulation occurs.

The amount of factor IX is extremely small and only difficultly obtainable. There is, therefore, difficulty in either obtaining the genetic information from the chromosome or preparing sufficient messenger RNA for successful reverse transcription of a complete sequence for factor IX.

2. Description of the Prior Art

Factor IX has been extensively purified from bovine and human plasma, as described by Fujikawa et al., Biochemistry (1973) 12:4938-4945 and DiScipio et al., ibid. (1977) 16:698-706. Approximately 20% of the amino acid sequence for the human molecule has been determined, DiScipio et al., J. Clin. Invest. (1978) 61:1528-1538. The entire sequence for the bovine molecule has been established. Katayama et al., PNAS USA (1979) 76:4990-4994. Both proteins are single-chain glycoproteins (M.sub.r, 55,000- 57,000) with an amino-terminal sequence of Y-N-S-G-K. Choo et al., Nature (1982) 299:178-180, describe the partial characterization of the human factor IX gene.

Genetic sequences are provided capable of hybridizing to human factor IX, as well as flanking sequences of the human factor IX gene. The sequences can be labeled and can be used for determining mutations, deletions or other modifications in genetic sequences expressing normal human factor IX or modifications thereof. The DNA fragment may also be used in hybrid DNA technology for expression of polypeptides.

The FIGURE is the sequence for factor IX as included in pHfIxl. The nucleotide sequence of the coding strand and the corresponding predicted amino acid sequence are shown. The two arginyl peptide bonds (residues 145 and 180), cleaved during activation of factor IX are shown by the heavy arrows.

DNA sequences are provided capable of binding to portions or all of the gene expressing human factor IX and flanking regions thereof. The sequences can be used in hybridization techniques for detecting genetic deficiencies and mutations involving human factor IX. By employing hybrid DNA technology, the sequences can be used for the production of polypeptides. The DNA sequences can be obtained free of introns, providing a continuous sequence coding for factor IX, precursors to factor IX or fragments of factor IX and their precursors. Also, the DNA sequences may include the sequences flanking the gene expressing human factor IX by themselves or in combination with human factor IX.

The polynucleotide sequences include the chromosomal gene, mRNA which has been matured and is capped and includes a poly(A) 3' chain, cDNA obtained by reverse transcription of fragments of any of the above, where the polynucleotide can be used for hybridization or for expression of a polypeptide having the sequence of all or part of factor IX. Also, mRNA, the combination of DNA sequences coding for a part or all of human factor IX may be joined to a vector for cloning, i.e. replication, or a plasmid provided including regulatory signals for expression, amplification and regulated response to a variety of conditions and reagents.

The nucleic acid sequences are sequences which will homoduplex or heteroduplex to the natural or mutated gene for human factor IX. The sequences may involve either RNA or DNA and will usually be at least about 14 bases long, more usually at least about 18 bases long and may be 3000 bases or longer. Where the fragments are to be used for hybridization, they will generally have fewer than about 1500 bases, usually be fewer than 1200 bases. For hybridization, the fragments may be labeled with a wide variety of labels, such as radionuclides e.g. .sup.32 P, haptens e.g. biotin, fluorescers, and the like. Various techniques may be employed for determining the occurrence of homo- or heteroduplexing, such as the Grunstein and Hogness technique (PNAS USA (1975) 72:3961-3965), the Southern technique (J. Mol. Biol. (1975) 98:503), the Northern technique (U.S. Pat. No. 4,302,204), or the like.

For determining the presence of sequences complementary to sequences encoding the factor IX gene, antibodies may be used which detect double-strandedness of DNA, RNA or a hybrid complex. The cellular polynucleotide to be assayed may be bound to a solid support, e.g. cellulose or nitrocellulose, by conventional means, e.g. heat, covalent diazo linkage, etc., or determined in solution, where a method is provided for detecting the formation of duplexes. The particular mode which is used for detecting the presence of duplexes is not a critical aspect of this invention, the subject invention providing the nucleotide sequences for human factor IX.

The polynucleotide sequences provided in accordance with this invention may be used in a variety of ways. The polynucleotide sequence may include the entire gene or the pseudogene (cDNA), including leader and prosequences and flanking regions. Fragments of the polynucleotide sequence may be taken, such sequences coding for factor IX; sequences coding for either the short or long chain of factor IX.sub.a, factor IX.sub.a.alpha. or factor IX.sub.a.beta. ; sequences coding for the activation peptide; sequences coding for a polypeptide at the N-terminus involved in processing of factor IX; fragments of such sequences or combinations thereof.

It is found that while factor IX.sub.a is strongly conserved as evidenced by the sequences of bovine and human factor IX, the activation peptide has over time been subject to variation and there is substantial variation between the activation peptide involved with human factor IX and the activation peptide involved with bovine factor IX, while factor IX.sub.a in the two cases is substantially conserved.

The following is the complete sequence for factor IX as included in pHfIXl (to be described subsequently). ##STR1## The nucleotide sequence of the coding strand and the corresponding predicted amino acid sequence are shown. The coding strand is inserted and numbered in the same orientation as the ampicillin resistance gene of pBR322. The amino acid sequence corresponding to the entire mature protein is numbered 1-416. The amino acid sequence that corresponds to the leader sequence is represented by minus numbers in the opposite direction. The two arginyl peptide bonds (residues 145 and 180) cleaved during the activation of factor IX are shown by the heavy arrows.

The genetic information for factor IX may be obtained as follows: A primate is injected with antibodies to factor IX, so as to reduce the circulating level of factor IX to substantially below normal. The primate is then sacrificed and the liver rapidly removed. Poly(A)-containing mRNA is isolated and assayed for factor IX with reticulocyte lysate with specific immunoprecipitation of radiolabeled product. By employing the above technique, a multifold enhancement of mRNA level for factor IX is achieved.

The mRNA for factor IX is further enriched by specific immunoprecipitation of liver polysomes with antibodies to human factor IX. The mRNA may then be used to synthesize cDNA employing reverse transcriptase and an appropriate primer e.g. oligo(dT).

Alternatively, by having one or more probes based on a known amino acid sequence, where the probes are at least about 12 bases, preferably at least about 15 bases, a genetic library may be screened for the presence of the desired genetic information. The genetic library may be derived from sheared or restricted genome or from cDNA. Particularly, a human liver cDNA library may be employed. The recombinant plasmids which bind strongly to the probe are isolated and purified by conventional means e.g. cesium chloride gradient centrifugation, and the plasmids digested with an appropriate restriction endonuclease to excise the desired genetic information.

The DNA sequences will then be characterized in a number of ways. One characterization is restriction endonuclease mapping. Another characterization is the nucleotide sequence which may be determined in accordance with conventional ways. The sequence may then be cloned to provide for substantial amounts of the DNA sequences. In addition, the amino acid sequence may be determined from the nucleotide sequence.

The cDNA obtained from the liver cDNA library coding for human factor IX contained 1466 base pairs and is flanked by G-C tails of 11 and 18 base pairs at the 5' and 3' ends, respectively, the tails having been introduced for cohesion and annealing. Nucleotides 12through 149 correspond to a leader sequence of 46 amino acids. The leader sequence contains three potential methionine start sites located at positions -46, -41 and -39. The methionine residues are then followed by a charged amino acid(s) (Arg at a position -45 or Glu at position -37) and a hydrophobic region rich in leucine, isoleucine, and tyrosine. These residues are typical of signal sequences found in most secreted proteins (Blobel et al. (1979) in Soc. for Experimental Biology Symp. XXXIII, Secretory Mechanisms, eds. Hopkins, C. R. & Duncan, C. J. (Cambridge University Press, Great Britain), pp. 9-36). These residues occur just prior to the Y--N--S--G--K sequence which is the amino-terminal sequence of the mature protein circulating in plasma.

Since the R--Y bond is not a typical cleavage site for signal peptidase, it appears likely that the newly synthesized factor IX in liver contains a pro-leader sequence analogous to serum albumin. This suggests that a signal peptidase cleaves at a peptide bond further upstream from the R--Y sequence, such as the A-N sequence (positions -10 and -9), the A-Q sequence (positions -21 and -20), or the S-A sequence (positions -22 and -21).

The mature protein for human factor IX is coded by 1248 pairs (nucleotides 150 through 1397) and is followed by a pair of adjacent stop codons of TAA and TGA. The molecular weight for the protein free of carbohydrate is calculated at 47,079. This is equivalent to a molecular weight of 56,722 upon the addition of 17% carbohydrate (DiScipio et al. (1978), supra). Human factor IX contains 12 glutamic acid residues in the amino-terminal region of the protein that are present as .gamma.-carboxyglutamic acid (Gla) in the mature molecule. These residues are located at positions 7, 8, 15, 17, 20, 21, 26, 27, 30, 33, 36 and 40. The two internal peptide bonds hydrolyzed by factor XI.sub.a during the activation reaction are Arg.sup.145 and Arg.sup.180 -Val.sup.181. Cleavage of these two arginyl peptide bonds results in the formation of factor IX.sub.a (M.sub.r, 43,196), a serine protease composed of a light chain (145 amino acids) and a heavy chain (236 amino acids) joined by disulfide links. The activation peptide (35 amino acids) is composed of residues 146 through 180 and includes four aspartic acid residues, three glutamic acid residues, one arginyl residue and probably modified with a plurality of sialic acid residues. The heavy chain contains the three principle residues involved in the catalytic activity of this serine protease, specifically His.sup.221, Asp.sup.270 and Ser.sup.366.

The synthetic probe that was employed in the screening for the human factor IX plasmid was a nucleotide mixture that included a base sequence of TATTTGCCTTTCAT, which codes for the M--K--G--K--Y sequence in factor IX starting with Met.sup.392. This amino acid sequence is present in both human and bovine factor IX. There are 69 changes in amino acid sequence between human and bovine proteins (Katayama et al. (1979), supra). There is in addition one insertion in bovine (Lys.sup.143) and one deletion in bovine (Asn.sup.259) factor IX. Both proteins contain 22 Cys residues which are present in the same positions in each protein. The largest difference in sequence between the two proteins occurs in the activation peptide where 17 of 35 residues have been changed. The overall identity between human and bovine factor IX was 83%.

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

Two different radiolabeled DNA probes were employed in these experiments. One probe was prepared from baboon liver mRNA that was enriched for factor IX by the following procedures: A young male baboon (5 kilograms body weight) was injected over a period of 48 hours with a total of 73 mg of affinity-purified goat antibodies to human factor IX. This procedure reduced the circulating factor IX clotting activity level to less than 1% of normal. The baboon was then sacrificed and the liver rapidly removed and frozen in liquid nitrogen. Poly(A)-containing RNA was isolated (MacGillivray et al., PNAS USA (1980) 77:5153-5157) and assayed for factor IX with a rabbit reticulocyte lysate (Delham and Jackson, Eur. J. Biochem. (1976) 67:247-256) by specific immunoprecipitation of the radiolabeled product (MacGillivray et al., ibid. (1979) 98:477-485). By this assay, the liver mRNA level for factor IX was elevated approximately 5-fold when compared with a control animal. The mRNA for factor IX was enriched another 20-fold by specific immunoprecipitation of the liver polysomes with affinity-purified goat antibodies to human factor IX employing the procedure of Gough & Adams, Biochemistry (1978) 17:5560-5566. The final factor IX mRNA.-level was approximately 2% of the total as estimated by the reticulocyte translation assay. This mRNA was then used to synthesize a radiolabeled cDNA in the presence of dATP, dGTP, [.alpha.-.sup.32 P]dCTP, [.alpha.-.sup.32 P]TTP, reverse transcriptase, and oligo(dT) as primer (Stein et al., Biochemistry (1978) 17:5763-5772). The specific activity of the cDNA was 5.times.10.sup.7 cpm/.mu.g.

The second probe was a synthetic DNA mixture of 14 nucleotides in length and contained 12 different DNA sequences. These sequences were complementary to the amino acid sequence of Met-Lys-Gly-Lys-Tyr. The DNA mixture contained the following sequences: ##STR2## The DNA mixture was radiolabeled with T4 kinase and [.gamma.-.sup.32 P]ATP to yield a specific activity of approximately 4.times.10.sup.8 cpm/.mu.g (Stein et al., (1978) supra). (Synthetic DNA mixture, P-L Biochemicals, Inc., Milwaukee, Wis.; T4 kinase, Bethesda Research Laboratories, Inc., Gaithersburg, Md.). Approximately 18,000 transformants were screened by a modification of the method of Wallace et al., (Nucleic Acids Res. (1981) 9:879-894).

The modified procedure was as follows: Colonies were grown on tetracycline (12.5 .mu.g/mL) plates overnight at 37.degree. C. (colony size 2-3 mm dia.) and the resulting colonies transferred to Whatman 540 filter paper (826 mm dia.). The filters with colonies up were placed onto chloramphenicol (250 .mu.g/mL) plates, incubated overnight at 37.degree. C. and then dried briefly in air (about 5 min.)

Lysis of the colonies was achieved by placing the filters with colonies up on the Whatman 3 mm paper wetted with 0.5N sodium hydroxide, the filters allowed to stand for 10 min. followed by blotting with clean white diaper cloth. The foregoing procedure was repeated, followed by transferring the filters onto the Whatman 3 mm paper wetted with 0.5M Tris, pH 7.4, the filters allowed to stand for 10 min., followed by blotting with diaper cloth and the procedure repeated. The filters were then transferred onto Whatman 3 mm paper, wetted with 2.times.SSC (1.times.SSC=0.15M NaCl, 0.015M sodium citrate, pH 7.0), the mixture allowed to stand for 10 min., blotted as before and the procedure repeated. After washing the filters briefly in 95% ethanol in a glass tray, the filters were air-dried for 20 min. and then either heated at 80.degree. for 2 hours or 68.degree. for several hours.

Hybridization was achieved by initially prehybridizing in 6.times.NET (1.times.NET=0.15M NaCl, 0.015M Tris-HCl, pH 7.5, 1 mM EDTA), 0.5% NP40 (Shell Oil Co.), 100 .mu.g/mL yeast tRNA, for 2 hours at 55.degree.. (E. coli DNA can be used to replace the yeast tRNA.) After briefly air-drying for about 5-10 minutes, hybridization was performed in 6.times.NET, 0.5% NP40, 250 .mu.g/mL yeast tRNA and about 6-8 ng/mL labeled oligonucleotide at 35.degree. for 15-20 hours. At the end of this time. the filters were washed at 0.degree. C. with 4 changes of 6.times.SSC (250 mL) and then at 35.degree. with two changes of 6.times.SSC for 10 min. The filters were then dried by blotting and exposed to X-ray film with an intensifier screen for 1-5 hours.

The human liver cDNA library (S.L.C. Woo and C. Thirumalachary) contained cDNA inserted into the Pst I site of plasmid pBR322. Four recombinant plasmids that hybridized strongly with the probe were isolated and purified by cesium chloride gradient centrifugation. DNA samples of the positive clone were then digested with Pst I, and the resulting fragments were analyzed by polyacrylamide gel electrophoresis. These inserts were also mapped by restriction endonucleases (Bethesda Research Laboratories, Inc., Gaithersburg, Md.).

Restriction fragments were labeled at the 3' end with [.alpha.-.sup.32 P] cordycepin 5'-phosphate in the presence of terminal deoxynucleotide transferase under conditions specified by the manufacturer (New England Nuclear, Boston, Mass.). They were also labeled at the 5' end with [.gamma.-.sup.32 P]ATP in the presence of T4 kinase after prior treatment of the DNA with bacterial alkaline phosphatase (Worthington Biochemical Co., Freehold, N.J.) or by an exchange reaction of the 5' phosphate group of [.gamma.-.sup.32 O]ATP in the presence of T4 kinase (Maxam and Gilbert, Methods Enxymol. (1980() 65:499- 560). Labeled fragments were then subjected to base modification and cleavage as described by Maxam and Gilbert, ibid., and subjected to electrophoresis on 0.35 mm polyacrylamide gels containing 8.3 M urea. The majority (92%) of the nucleotide sequence was established by two or more sequence experiments, and approximately 80% was determined on both strands.

Four positive clones were identified from a human liver cDNA library of 18,000 recombinant plasmids employing the synthetic oligonucleotide mixture described above and a single-stranded DNA prepared from enriched mRNA for baboon factor IX as probes. With two of these plasmids, the cDNA was readily released by digestion with Pst I, yielding an insert of approximately 1500 base pairs. In preliminary experiments, these two clones were found to be identical by restriction mapping. The other two plasmids were not further examined. The insert from the first clone, designated pHfIXl, was further mapped by restriction endonuclease digestion. Sites for cleavage and end-labeling were chosen from the detailed restriction map, and both 5' and 3' labeling methods were used. The coding strand was used to number the DNA sequence which is presented in the same orientation as the ampicillin-resistance gene of pBR322. The insert has been described above.

In accordance with the subject invention, DNA sequences are provided for hybridization with pro-factor IX, factor IX, factor IX.sub.a, and activation peptide, and for DNA and RNA fragments which can be used in the detection of mutations or other genetic deficiencies concerned with factor IX. The sequences can be used in diagnosing blood clotting deficiencies, such as hemophilia, particularly hemophilia B. By lysing cells as described above and screening the DNA with fragments according to the subject invention, mutations in the factor IX gene may be determined.

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.