NO175317B - DNA molecule encoding a thrombolytic protein - Google Patents

Abstract

Thrombolytic proteins with tissue plasma type activity. The proteins are characterized by modification within the 94 amino acid N-terminus, and / or by Arg-275, and / or by one or more of the N-linked glycosylation sites. Methods for the preparation of these proteins are described in the same manner as therapeutic compositions containing them.

Description

The present invention relates to a DNA molecule which codes for a thrombolytic protein with tissue plasminogen activator-type activity.

These proteins are active thrombolytic agents that have improved fibrinolytic profiles compared to native human t-PA. This can be manifested as an increased affinity to fibrin, reduced reactivity with inhibitors of t-PA, faster thrombolysis rate, increased fibrinolytic activity and/or prolonged biological half-life. It is also intended that the proteins according to the invention can be more easily produced in a more homogeneous form than the native human t-PA can be produced. An improved overall pharmacokinetic profile is characteristic of the proteins.

The structure of native human t-PA can be seen as comprising an amino (N-) terminus of approx. 91 amino acid residues, two so-called "pretzel" regions and at the carboxy terminus, a serine protease-type region. The applicant has found that the N-terminus contains several subregions that play functional roles, including in fibrin binding and in vivo clarification of the protein. Recently, the recovery of another form of t-PA which lacks the native N-terminus and the first pretzel region has been indicated, see EP-publ. 0196920. According to this report, the truncated form of t-PA beginning with Ala-160 in native human t-PA is fibrinolytically active.

As described in greater detail above, the present invention provides novel protein analogs of human t-PA that retain both the pretzel regions of native human t-PA, but also contain modifications within the N-terminus. While the modifications in certain embodiments involve deletions in the N-terminus, the first coil region remains intact and the N-terminal deletion is never greater than 94 amino acids. Most embodiments involve significantly less omission and/or amino acid substitution. By retaining more of the structure of native human t-PA, the aim is that the proteins according to the invention selectively retain more of the desired biological activities for native human t-PA and may be less immunogenic than the more drastically modified analogues of t-PA. Therefore, the proteins according to the invention have improved fibrinolytic and pharmacokinetic profile compared to both native human t-PA and the truncated Ala-160 t-PA, as well as other modified forms of t-PA.

The polypeptide chain of native human t-PA also includes four consensus Asn-linked glycosylation sites. It has been shown that two of these sites are characteristically glycosylated in t-PA form from melanoma-derived mammalian cells, that is, at Asn-117 and Asn-448. Asn-184 is sometimes glycosylated and Asn-218 is characteristically not glycosylated. t-PA from melanoma-derived mammalian cells, such as Bowes cells, is also referred to herein as "native" or "natural" human t-PA.

According to this, the present invention relates to a DNA molecule which codes for a thrombolytic protein with tissue plasminogen activator-type activity, and the molecule is characterized by an amino acid sequence essentially the same as the peptide sequence in human t-PA, where: a) the sequence starting with Cys- 6 up to and including Ile-86 have been omitted, b) at least one of the consensus N-linked glycosylation sites has been modified to something other than a consensus N-linked N-glycosylation site, and where appropriate c) modification of the enzymatic cleavage site Arg has been carried out -175/Ile-276 in that Arg is omitted or

replaced by another amino acid.

In one embodiment of the invention, it relates to a DNA molecule as described above and which is characterized in that the consensus N-linked glycosylation site at ASn-117 is modified so that Asn-117 is replaced by Gln-117.

In a further embodiment of the invention, it concerns a DNA molecule as described above and which is characterized by the fact that the three consensus N-linked glycosylation sites at amino acid positions 117-119, 184-186 and 448-450 are modified so that Asn-117, Asn -184 and Asn-448 are replaced with Gin.

In a further embodiment, the invention relates to a DNA molecule as described above and which is characterized in that a) the sequence from and including Gly-(-3) to and including Thr-91 is omitted and b) at least one of the consensus N-linked glycosylation sites are modified to other than a consensus N-linked glycosylation site.

The proteins according to the invention differ in structure from human t-PA in that they contain modifications in the peptide sequence (i) at up to three of the Asn-linked glycosylation sites present in native t-PA; (ii) in the N-terminus of the proteins corresponding to the 94 amino acid mature N-terminus of native t-PA; and/or (iii) at the proteolytic cleavage site spanning Arg-275 and Ile-276. These features of the proteins according to the invention are described in greater detail below. Notwithstanding the various modifications, the numbering of the amino acids as shown in the one-letter coding sequence in Table 1 is retained.

A. Modifications at the N-terminus

In a feature of the invention, the proteins are characterized by the omission of 1-95 amino acids within the peptide region spanning Gly-(-3) or Ser-1 up to Thr-91, in relation to native human t-PA. In one embodiment, for example, Cys-51 to Asp-87 in native t-PA is omitted. In two other specific embodiments, Cys-6 through Ser-50 and Cys-6 through Ile-86, respectively, are omitted. In other embodiments, more conservative modifications are present in the N-terminal region of the proteins. For example, certain proteins according to the invention contain one or more amino acid omissions or substitutions within one or more of the following, more distinct sub-regions:

These and other modifications within the T-terminus spanning Gly-(-3) to Thr-91 are described in greater detail below.

B. Modifications at N-linked glycosylation sites

Furthermore, the protein variants according to the invention need not contain any N-linked carbohydrate parts or may only be partially glycosylated in relation to natural human t-PA. A "partially glycosylated" protein, as this term is used herein, means a protein that contains fewer N-linked carbohydrate moieties than the fully glycosylated native human t-PA. This absence of glycosylation or only partial glycosylation stems from the amino acid substitution or omission at one or more of the consensus N-linked glycosylation recognition sites present in the native t-PA molecule. The applicant has found that variant proteins according to the invention with such a modification on one or more N-linked glycosylation sites retain t-PA-type thrombolytic activity with greater fibrinolytic activity in certain cases, can be more easily produced in a more homogeneous form than native t-PA and in many cases have longer in vivo half-lives than native t-PA.

N-linked glycosylation recognition sites are today believed to comprise tripeptide sequences that are specifically recognized by the appropriate cellular glycosylation enzymes. These tripeptide sequences are either asparagine-X-threonine or asparagine-X-serine, where X is usually any amino acid. Their localization within the t-PA peptide sequence is shown in Table 1. A number of amino acid substitutions or omissions in one or more of the three positions for a glycosylation recognition site result in non-glycosylations of the modified sequence. As an example, both Asn-117 and Asn-184 of t-PA are both replaced with Thr in one embodiment and with Gin in another. In at least one case of the double Gln replacement, the resulting glycoprotein, (Gln-117-Gln-184), should contain only one N-linked carbohydrate moiety (at Asn-448) instead of two or three such moieties as is the case in native t-PA. The person skilled in the art will recognize here that analogous glycoproteins with the same Asn-448 monoglycosylation can be prepared by omitting amino acids or the substitution of other amino acids in positions 117 and 184 and/or by omitting or substituting one or more amino acids at other positions within the respective glycosylation recognition sites, for example at Ser-119 and Ser-186 as mentioned above, and/or by centrifugation, or more particularly by omission, at one or more of the "X" sites in the tripeptide sites. In another embodiment, Asn at positions 117, 184 and 448 is replaced with Gin. These resulting variants should not contain any N-linked carbohydrate moieties, but two or three such moieties as is the case in native t-PA. In other embodiments, potential glycosylation sites are modified individually, for example by replacing Asn, for example, with Gin in position 117 in a currently preferred embodiment, in position 184 in another embodiment and in position 448 in a further embodiment. The present invention includes such non-glycosylated, monoglycosylated, diglycosylated and triglycosylated t-PA variants.

Exemplary modifications of one or more of these three consensus N-linked glycosylation sequences, R<*>, R<* and R^, as found in various embodiments of the invention, are set forth below:

C. Modification of Arg-275/ Ile-276- cleavage site

According to a feature of the invention, the variants are optionally modified at the proteolytic cleavage site spanning Arg-275 and Ile-276 by omission of Arg-275 or substitution of another amino acid, preferably an amino acid different from Lys or His. Thr is today a particularly preferred replacement amino acid for Arg-275 in various embodiments of the present invention. Proteolytic cleavage at Arg-275 of native t-PA gives the so-called "two-ls jede" molecule as known in the art. Proteins according to the invention which are characterized by modification on this cleavage site can be more easily produced in a more homogeneous form than the corresponding protein without the cleavage site modification and can possibly very importantly have an improved fibrinolytic profile and pharmacokinetic properties.

In one embodiment, the proteins are characterized by a peptide sequence substantially the same as the peptide sequence of human t-PA where Arg-275 is omitted or replaced with another amino acid, preferably different from lysine or histidine, and at least one of consensus Asn-linked glycosylation sites is omitted or modified to other than a consensus Asn-linked glycosylation sequence.

Examples of proteins of this embodiment are set forth in Table 1 below. By the expression "characterized by a peptide sequence substantially the same as the peptide sequence of human t-PA", as used herein, is meant the peptide sequence of human t-PA, a peptide sequence encoded by a DNA sequence encoding human t-PA or a DNA sequence capable of hybridizing to it under stringent hybridization conditions.

Thus, the proteins according to the invention include analogues of t-PA which are characterized by the various modifications or combinations of modifications as described here, and which may also contain other variations such as allelic variations or further omissions, substitutions or insertions of amino acids which still retain thrombolytic activity, as long as the DNA encoding these proteins (before the modification according to the invention) is still capable of hybridization to a DNA sequence encoding human t-PA under stringent conditions. In a second embodiment, the proteins are characterized by a peptide sequence substantially the same as the peptide sequences of human t-PA in which one or more amino acids are omitted in the N-terminal region from Gly-(-3) through Thr-91 and in which (a) a or more Asn-linked glycosylation sites are optionally omitted or otherwise modified to other than a consensus Asn-linked glycosylation site, and/or (b) Arg-275 is optionally omitted or replaced with another amino acid, preferably different from lysine or histidine. Examples of proteins according to this embodiment are shown below:

Illustrative proteins with N-terminal deletions

The following proteins have the peptide sequence shown in Table 1, where R<1>, R<2> and R<3> mean the wild-type tripeptide sequences, but where the N-termini (Gly-(-3) even Thr-91) is replaced with:

This embodiment includes a subgenus of proteins in which 1 to about 94 amino acids are omitted from the region of Gly-(-3) through Thr-91 and one or more of the Asn-linked glycosylation sites are omitted or otherwise modified to something other than a consensus Asn-linked glycosylation sequence as described previously. Further included is a subgenus of compounds wherein 1 to about 94 amino acids are omitted from the region of Gly-(-3) through Thr-91, and where Arg-275 is omitted or replaced with another amino acid, preferably different from lysine or histidine. A further subgenus of this embodiment is characterized by the omission of 1 to approx. 94 amino acids from the region Gly-(-3) to Thr-91 inclusive, omission or modification of one or more of the Asn-linked glycosylation sites (see for example the table on page 6) and omission of Arg-275 or its replacement by another amino acid. Examples of proteins of these subgenera are given in tables 2 and 2.5 below.

This embodiment also includes a subgenus of proteins in which the N-terminal omission comprises an omission of 1 to about 45 amino acids from the region Ser-1 through Ser-50. Also included is a subgenus of proteins in which 1 to approx. 45 amino acids are omitted from the region Ser-1 through Ser-50 and one or more glycosylation sites are modified as described previously. A further subgenus comprises proteins with an omission of 1 to about 45 amino acids from the range Ser-1 through Ser-50, in which Arg-275 is omitted or replaced with another amino acid. Additionally included is a subgenus with an omission of 1 to approx. 45 amino acids from the range Ser-1 to Ser-50 and where both of (a) one or more glycosylation sites, and (b) Arg-275, are optionally modified as indicated above. Examples of proteins in these subgenera are listed in Table 3 below, as well as in

tables 2 and 2.5.

Specific proteins according to the invention can be designated by a three-part statement comprising a compound number from Table 2 followed by a statement of the N-terminus and then the identification of position 275. For example, compound No. 2-11/N-6/Arg means a protein where the 3-glycosylation site is omitted ("2-11", see Table 2), C-36 through C-43 are omitted (N-terminus #N-6) and Arg-275 is retained.

Table 3: Examples of proteins with an omission of 1 -

about. 45 amino acids in the region Ser-1 through Ser-50 and a modification of one or both of (a) Arg-275 and (b) at least one N-linked glycosylation site (for the general sequence, see Table 1 ).

11 luster proteins are son, defined in Table 2, but with the fading N-termini as replacements for the wild-type sequence of Gly-(-3) through Thr-91:

Referring to the modified N-termini as indicated in Table 3, it should be clear that more than one amino acid can be omitted. Where several amino acids are omitted, they may be close together or be separated by one or more amino acids. In the designation of compounds with such N-termini, the size of the omission is indicated by an "An" where "n" is the number of omitted amino acids. For example, N-terminus #24 where one amino acid is omitted as shown in Table 3 may be called "N-24A1", where two amino acids are omitted, for example S-1 and Y-2 the N-terminus is called a "N-24A2", and so on. Where combinations of omissions are carried out, for example where S-1, Y-2 and 1-5 are omitted, the N-terminus may be designated as "N-24A2, N-28A1". As indicated in the text following Table 2.5, specific compounds are indicated by a three-part code comprising a compound number from Table 2 followed by an indication of the N-terminus element #, for example from Table 2.5, with a subsequent identification of the status of position 275. Compound 2-26/N-24A2, N-28A1/- thus indicates the protein that all three glycosylation sites; R-275; S-1, Y-2 and 1-5 are omitted. This embodiment further includes a subgenus of proteins wherein 1 to about 41 amino acids are omitted from the region Cys-51 through Thr-91. Proteins in this subgenus can optionally be modified so that Arg-275 is omitted or replaced with another amino acid, preferably different from lysine or histidine. Proteins of this subgenus can, as an alternative to or in addition to the modification at Arg-275, be modified so that (a) one or more N-linked glycosylation sites are renounced as described above, and/or (b) one or more amino acids are omitted in the region GLy-(-3) through Ser-50. Examples of proteins of this subgenus correspond to those listed in Tables 2 and 3, but contain

N-termini such as those listed in Table 4 below.

Table 4: Examples of proteins with an omission of 1 to approx. 41 amino acids from the region Cys-51 through Thr-91 (for the general sequence, see Table 1).

Illustrative proteins are as indicated in Table 2, but with the following N-termini in place of the wild-type sequences in Gly-(-3) through Thr-91:

With reference to table 4 above, it should be pointed out that several subclasses of proteins are described. For example, proteins containing an omission of 1 to 37 amino acids (individually, consecutively or in combination) from Cys-51 through Asp-87 are indicated in the same way as proteins containing an omission of 1 to 37 amino acids from Cys-51 to and with Arg-87 and an omission of one or more amino acids within the range of Gly-(-3) through Cys-51 (see N-76). With reference to compounds containing an N-terminus selected from N-termini N-76 through N-III, it should be understood that more than one amino acid may be omitted. For example, N-77 where an amino acid is omitted, as shown in Table 4, may be referred to as "N-77A1". Where six amino acids are omitted, for example S-52 through F-57, this N-terminus is referred to as "N-77A6", and so on. Specific proteins are indicated as described according to Tables 2 and 3. Examples where three glycosylation sites and Arg-275 are omitted are shown below:

Illustrative compounds in which no modifications are present at the glycosylation sites, but which contain various omissions, are shown below:

Illustrative compounds with the same omissions as above, but also modified at the first N-linked glycosylation site, here by replacement of Asn-117 with Glin, are set forth below:

2-l/N-424/Arg

2-l/N-425/Arg

2-l/N-426/Arg

2-l/N-427/Arg

2-l/N-428/Arg

2-1/N-63A2,N-92A3/Arg

2-1/N-63A1,N-92A2/Arg

2-1/N-63A1,N-92A1/Arg

Illustrative compound with the same omissions as above, but which is also modified at all three glycosylation sites, here by replacing Asn<*>s with Gins, is indicated below: 2-7/N-424/Arg

2-7/N-425/Arg

2-7/N-426/Arg

2-7/N-427/Arg

2-7/N-428/Arg

2-7/N-63A2,N-92A3/Arg

2-7/N-63A1,N-92A2/Arg

2-7/N-63A1,N-92Al/Arg

Thus, this embodiment further includes a subgenus of proteins, where one or more omissions of less than approx. 20 amino acids are present in the region Gly-(-3) through Thr-91. Proteins of this subgenus can also be modified at Arg-275 and/or at one or more of the Asn-linked glycosylation sites. Examples of proteins of this subgenus correspond to those indicated in Tables 2 to 4 inclusive, but contain instead of the described wild-type N-terminus an N-terminus of the type described in Table 5 below. Additional examples of compounds of this subgenus are also indicated above by their three-part notation. In a third embodiment, the proteins are characterized by a peptide sequence essentially the same as the peptide sequence of human t-PA in which different amino acids are inserted instead of 1-94 of the amino acids in the region Gly-(-3) up to and including Thr-91. This embodiment includes a subgenus of compounds characterized by substitution of one or more amino acids within the aforementioned N-terminus and by modification at Arg-275 as previously mentioned. Also included is a subgenus of compounds that are characterized by an above-mentioned replacement of one or more amino acids within this N-terminus, and modification, as previously mentioned, at one or more of the consensus Asn-linked glycosylation sites. A further subgenus of this embodiment is characterized by substitution of one or more amino acids within this N-terminus and modification as previously described, both at Arg-275 and at one or more of the N-linked glycosylation sites. In one aspect of this embodiment, the amino acid substitution or substitutions are within the range of Gly-(-3) through Ser-50, with or without modification at Arg-275 and/or one or more of the N-linked glycosylation sites. In another feature of the invention, the amino acid substitution(s) lies within the range Cys-51 to Thr-91, also here with or without modification at Arg-275 and/or one or more of the N-linked glycosylation sites. In a further move, 1 becomes approx. 11, preferably 1 to approx. 6 amino acids replaced within one or more of the following areas, again with or without another modification mentioned above:

In a further feature of the invention, substitution takes place in one or more of the following areas: R-7 up to and including S-20, W-21 up to and including Y-33, N-37 up to and including 0-42 and H- 44 through S-50. In a further feature of this embodiment, the N-terminus is modified once again by substitution of 1 to ca. 11, and preferably 1 to approx. 6 amino acids in one or more of the areas stated above, and is further modified by omitting 1 to 93 and preferably 1 to approx.

45 and most preferably 1 to approx. 15 amino acids.

Illustrative amino acid substitutions are shown in Table 6-A below, and exemplary proteins are listed in Table 6-B. Of the replacement amino acids for R-40, A-41 and O-42 as indicated in Table 6-A, S is a preferred replacement for R-40, V and L preferred replacements for A-41 and Q-42, respectively. It should be pointed out that proteins according to the invention comprising the substitutions identified for R-40, A-41 and O-42 in Table 6-A are at the present time preferred alone, or, as in other aspects and subgenera according to the invention, in combination with other substitutions and/or omissions within the N-terminus and/or modifications on R-275 and/or at least one glycosylation site. To the extent that the present proteins are modified by substitution rather than deletion, the proteins retain more of the native t-PA conformation and selectively retain more of the desired biological properties of native t-PA.

Table 5: Protein examples with one or more omissions on

less than approx. 20 amino acids within the range of Gly-(-3) through Thr-91 (for the general sequence, see Table 1).

Illustrative proteins are defined in Table 2, but with the following N-termini instead of the wild-type sequence in Gly-(-3) through Thr-91:

And with reference to tables 2.5, 3 and 4: N-l

N-2

N-6 through N-ll N-15 through N-17 N-24A1 through Al9 N-27A1 through Al9 N-44A1 through Al9 N-73A1 through Al9 N-95A1 through and with Al9 NlllAl through A5

Table 6-B: Protein Examples Containing Substitution for One or More Amino Acids in the Range Gly-(-3) through Thr-91 (For the General Sequence, See Table 1)

Illustrative proteins are as indicated in Table 2, but with the following N-termini in place of the wild-type sequences in Gly-(-3) through Thr-91:

Proteins according to the invention which comprise amino acid substitution can be indicated by a three-part code as described earlier. Of course, it should be clear that more than one wild-type amino acid can be substituted. By designating the compounds with such N-termini, it is intended to indicate the number of substitutions by "sn" where "n" is the number of substituted amino acids, for example with amino acid substitutions such as (but not limited to) those listed in Table 6-A . For example, N-terminus #N-122sl indicates N-terminus 122 as indicated in Table 6-B, while N-terminus #N-122s4 indicates the N-terminus where S-1 is replaced by G and the following three wild-type amino acids are replaced with other amino acids. Proteins of this embodiment containing multiple amino acid substitutions may be indicated by a string of N-terminus designations indicating specific substitutions, all as follows:

Illustrative compounds with multiple N-terminal substitutions with Asn-117 replaced by Gin and Arg-275 replaced by Thr:

A subgenus of special interest is characterized by substitution of one or more of Y-67 to S-69 with possible omission of, and/or replacement of, one or more amino acids from and including Gly-(-3) to and including L -66, with or without modifications as indicated above on one or more of the glycosylation sites and/or on Arg-275.

In a feature of the invention, the proteins contain at least one so-called "complex carbohydrate" sugar part which is characteristic of mammalian glycoproteins. As exemplified in greater detail below, such "complex carbohydrate" glycoproteins can be produced by expressing a DNA molecule encoding the desired polypeptide sequence in mammalian host cells. Suitable mammalian host cells and methods for transformation, cultivation, amplification, evaluation and product production and purification are known in the art, reference should be made to, for example, Gething and Sambrook, "Nature" 293:620-625 (1981), or alternatively, Kaufman et al., "Molecular and Cellular Biology", 5(7):1750-1759 (1985) or Howley et al. in US-PS 4,419,446.

A further feature of the invention involves t-PA variants as indicated above in which each carbohydrate moiety is a processed portion of the original doicol-linked oligosaccharide characteristic of insect cell-produced glycoproteins as opposed to a "complex carbohydrate" substituent characteristic of mammalian glycoproteins, including mammalian-derived t-PA. Such insect cell-type glycosylation is referred to herein for simplicity as "high mannose" carbohydrate. For purposes of the present application, complex and high mannose carbohydrates are as defined by Kornfeld et al., "Ann. Rev. Biochem.", 54:631-64 (1985). "High mannose" variants according to the invention are characterized by a variant polypeptide backbone as described above and which contains at least one occupied N-linked glycosylation site. Such variants can be produced by expressing a DNA sequence that codes for a variant in the insect host cells. Suitable such as well as methods and substances for transformation/transfecting, insect cell cultures, evaluation and product procuration and purification which can be used by carrying out this aspect of the invention are known in the art. The glycoprotein produced in this way also differs from natural t-PA and from t-PA previously produced by recombinant construction techniques in mammalian cells in that the variants in this aspect of the invention do not contain terminal sialic acid or galactose substituents on the carbohydrate parts or other protein modifications that are characteristic of mammalian-derived glycoproteins.

The proteins according to the invention which do not contain N-linked carbohydrate moieties can also be produced by expressing a DNA molecule encoding the desired variant, for example compounds 1-6 through 1-11 inclusive in Table 1, in mammalian, insect, yeast - or bacterial host cells, where eukaryotic host cells are currently preferred. As indicated above, suitable mammalian and insect host cells and additionally suitable yeast and bacterial host cells as well as methods and substances for transformation/transfection, cell cultivation, evaluation and product production and purification which can be used in this feature of the invention are also known in the art .

In addition, as should be apparent to the person skilled in the art, the present invention also includes other t-PA variants which are characterized by, instead of amino acid omission in the area of Gly-(-3) or Ser-1 up to and including Thr-91, of one or more amino acid substitutions within this region, especially in the region Arg-7 through Ser-50, or by a combination or omission and substitution. cDNAs encoding these compounds can be easily prepared, for example by methods that are highly analogous to the mutagenesis procedures described here using suitable mutagenesis oligonucleotides. These cDNAs can optionally be mutagenized on one or more of the codons for R<1>, R<2> and R<3> and/or Arg-275, and can be inserted into expression vectors and expressed in host cells by the methods described herein . It is clear that these proteins will share the advantageous pharmacokinetic properties of other compounds according to the invention and possibly avoid undue antigenicity when administered in pharmaceutical preparations analogous to those described here.

As can be seen from the above, all variants of the invention are produced by recombinant techniques using DNA sequences that encode analogues that may also contain fewer or possibly no potential glycosylation sites compared to natural human t-PA and/or the omission or replacement of Arg-275. Such DNA sequences can be prepared by conventional site-directed mutagenesis of DNA sequences encoding t-PA.

The DNA sequences encoding t-PA have been cloned and characterized, see, for example, D-Pennica et al., "Nature" (London) 301:214

(1983) and R. Kaufman et al., "Mol. Cell. Biol.", 5(7):1750

(1985). One clone, ATCC 39891 encoding a thrombolytically active t-PA analog is unique in that it contains a Met residue at position 245 instead of Val. Characteristically, the DNA sequence encodes a leader sequence that is processed, i.e. recognized and removed by the host cell, followed by the amino acid residue in the full-length protein, starting from Gly.Ala.Arg.Ser.Tyr.Gin.. Depending on the medium and host cell in which the DNA sequence is expressed , the thus produced protein can begin with a Gly.Ala.Arg amino terminus or be further processed so that the first three amino acid residues are proteolytically removed. In the latter case, the mature protein has an amino terminus comprising: Ser.Tyr.Gln.Leu... t-PA variants with these amino termini are thrombolytically active and are covered by the invention. Variants according to the present invention also include proteins with either Met-245 or Val-245, as well as other variants, for example allelic variations or other amino acid substitutions or omissions, which still retain thrombolytic activity.

The invention also encompasses compounds as described above which contain the additional modification in the polypeptide region above Asn-218 through Thr-220. Specifically, compounds in this embodiment are further characterized by an amino acid other than Asn or a peptide bond at position 218 and/or an amino acid other than Pro or a peptide bond at position 219 and/or an amino acid other than Ser or Thr or a peptide bond at position 220 Thus, compounds of this embodiment lack the consensus N-linked glycosylation site that is characteristically not glycosylated in t-PA produced by melanoma-derived mammalian cells.

As mentioned above, DNA sequences encoding individual variants according to the invention can be produced by conventional site-directed mutagenesis of a DNA sequence encoding human t-PA or analogues or variants thereof.

Such methods of mutagenesis include the M13 system of Zoller and Smith, "Nucleic Acids Res." 10:6487-6500 (1982); "Methods Enzymol." 100: 468-500 (1983); and DNA 3:479-488 (1984), using single-stranded DNA and the method according to using heteroduplex DNA. Various examples of oligonucleotides used according to such methods to cause omissions in the N-terminus or to convert an asparagine residue into, for example, threonine or glutamine, are shown in Table 7. It should of course be clear that the DNA encoding each of the glycoproteins according to the invention can be analogously produced by the person skilled in the art by site-directed mutagenesis by using one or more suitably chosen oligonucleotides. Expression of DNA by conventional means in a mammalian, yeast, bacterial or insect host cell system yields the desired variant. Mammalian expression systems and the various variants thereby obtained are currently preferred.

The mammalian cell expression vectors described herein can be synthesized by techniques well known to those skilled in the art. The components of the vectors such as bacterial replicons, selection genes, enhancers, promoters and the like can be obtained from natural sources or synthesized by known procedures, see, for example, Kaufman et al., "J. Mol. Biol.", 159:51-521 ( 1982); Kaufman, "Proe. Nat. Acad. Sic", 82:689-693 (1985).

Established cell lines including transformed cell lines are suitable as hosts. Normal diploid cells, cell lines derived from in vitro culture of primary tissue, as well as primary explants (including relatively undifferentiated cells such as hematopoietin stem cells) are also suitable. Candidate cells do not need to be genotypically deficient in the selection gene as long as the selection gene acts dominant.

The host cells are preferably established mammalian cell lines. For stable integration of vector DNA into chromosomal DNA and for subsequent amplification of the integrated vector DNA, both by conventional methods, CHO (Chinese hamster ovary) cells are today preferred. Alternatively, vector DNA may contain all or part of the bovine papilloma virus genome (Lusky et al., "Cell", 36:391-401 (1984) and carried in cell lines such as C127 mouse cells as a stable episomal element. Other useful mammalian cell lines include , but are not limited to HeLa, COS-1 monkey cells, mouse L-929 cells, 3T3 lines derived from Swiss, Balb-c, or NIH mice, BHK or HaK hamster cell lines, and the like.

Stable transformants are then assessed with a view to expression of the product by standard immunological or enzymatic analyses. The presence of DNA encoding the variant proteins can be detected by standard procedures such as Southern blotting. Transient expression of DNA encoding the variants within a few days after introduction of expression vector DNA into suitable host cells such as COS-1 monkey cells is measured without selection by activity or immunological analysis of the proteins in the culture medium.

In the case of bacterial expression, the DNA encoding the variants can be further modified to contain different codons for bacterial expression as known in the art and is preferably operably linked in frame to a nucleotide sequence encoding a secretory leader polypeptide that allows bacterial expression, secretion and processing of the mature variant protein, also as known in this technique. The compounds that are expressed in mammalian, insect, yeast or bacterial host cells can then be recovered, purified and/or characterized with regard to physicochemical, biochemical and/or clinical parameters, all by methods known per se.

These compounds have been found to bind ten monoclonal antibodies directed against human t-PA and can thus be recovered and/or purified by immunoaffinity chromatography using such antibodies. Furthermore, these compounds have t-PA-type enzymatic activity, for example, compounds according to the invention actively activate plasminogen in the presence of fibrin to force fibrinolysis, measured in an indirect assay using plasmin chromogenic substrate S-2251 as known in the art.

The invention also encompasses preparations for thrombolytic therapy and which comprise a therapeutically effective amount of a variant as described above in admixture with a pharmaceutically acceptable parenteral carrier. Such preparations can be used in the same way as described for human t-PA and must be usable for humans and lower animals such as dogs, cats and other mammals known to be victims of thrombotic cardio-vascular problems. It must be clear that the preparations are used both for treatment and, where appropriate, also for the prevention of thrombotic conditions. The exact dosage and method of administration is determined by the attending physician depending on the potency and pharmacokinetic profile of the compound in question as well as various factors that modify the action of the medicants, such as body weight, sex, diet, time of administration, drug combination, reaction sensitivities and the severity of the case in question.

The following examples are given to illustrate embodiments of the invention. It should be clear that these examples are illustrative and the invention is not to be considered limited to these, but to what is stated in the accompanying claims

In each of the examples involving insect cell expression, the nuclear polyhedrosis virus used was the L-1 variant of Autographa californica, and the insect cell line used was the Spodoptera frugiperda IPLB-SF21 cell line (Vaughn, J.L. et al., "In Vitro" (1977) 13, 213-217). Cell and viral manipulations are as indicated in the literature (Pennock, G.D., et al., supra, Miller, D.W., Safer, P. and Miller, L.K., "Genetic Engineering", vol. 8, pp. 277-298, J.K. Setlow and A. Hollaender, ed. Plenum Press, 1986). The RF-ml3 vectors, mpl8 and mpll, are commercially available from New England Biolabs. However, the person skilled in the art will recognize that other strains, strains, host cells, promoters and vectors containing the relevant cDNA as described above can also be used when carrying out each embodiment according to the invention. The DNA manipulations used are, unless otherwise noted, according to Maniatis et al., "Molecular Cloning: A Laboratory Manual" (Cold Spring Harbor, NY 1982).

As those skilled in the art will recognize, oligonucleotides can easily be constructed for use by omitting one or more amino acids or inserting another (that is, replacement) amino acid at a desired site by omitting one or more codons, respectively by replacing the codon with the desired replacement amino acid in the oligonucleotide.

Other mutagenesis oligonucleotides can be constructed, based on an approx. 20-50 nucleotide sequence spanning the desired site, with replacement or omission of the original codon(s) one wishes to replace.

Plasmid derivatives

Mutagenesis of cDNAs on codons for the different amino acids was carried out using an appropriate restriction fragment of cDNA in M13 plasmids according to Zoller and Smith. The omissions within this cDNA were carried out by deletion mutagenesis using a suitable restriction fragment, for example the Sac1 fragment, to the cDNA either in M13 vectors or by a heteroduplex deletion in plasmid pSVPA4.

This plasmid was constructed to allow expression of t-PA glycoprotein in mammalian cells. This plasmid was prepared by first removing the DNA encoding the SV40 large T polypeptide from the plasmid pspLT5 (Zhu, Z. et al., 1984, "J. Virology", ELL: 170-180). This was accomplished by performing a total XhoI digestion followed by partial Bam-HI restriction endonuclease digestion. The SV40 large T coding region of pspLT5 was replaced with human t-PA coding sequence by ligating a cohesive SalI/BamHl t-PA coding restriction fragment, isolated by degradation plasmid J205 (ATCC no. 39568) with SalI and BamHl to the origin -XhoI/BamH1 cut vector pspLT5, prepared as described above. As a result, t-PA will be transcribed in this vector under the control of the Sv40-sen promoter introduced into mammalian cells. This final construct is designated pSVPA4.

Plasmid pLDSG is an amplifiable vector for expressing t-PA in mammalian cells such as CHO cells. pLDSG contains a mouse DHFR cDNA transcription unit using the adenovirus type 2 major late promoter, MLP, the simian virus 40 (SV40) enhancer and the origin of replication, the SV40 late promoter (in the same orientation as adenovirus MLP), a gene encoding tetracycline resistance and a cDNA encoding human t-PA (Val-245) in the correct orientation for adenovirus type 2 MLP. The preparation of pLDSG from pCVSVL2 (ATCC No. 39813) and a t-PA encoding cDNA is described in detail as is cotransformation with, and amplification of, pLDSG in CHO cells, see Kaufman et al., "Mol. and Cell. Bio.", 5(7):1750-1759

(1985).

Plasmid pWGSM is identical to pLDSG except that the cDNA inserts encode Met-245 human t-PA. pWSGM can be constructed using cDNA from plasmid J205 (ATCC No. 39568) or pIVAPA/1 (ATCC No. 39891). Throughout the specification, pWGSM and pLDSG are used interchangeably, although, as previously indicated, the former vector produces Val-245 proteins and the latter Met-245 proteins.

pIVPA/1 (ATCC No. 39891) is a baculoviral translocation vector containing a cDNA encoding t-PA. pIVPA/1 and mutagenized derivatives thereof are used to insert a desired cDNA into a baculoviral genome so that the cDNA will be under the transcriptional control of the baculoviral polyhedrin promoter.

Heteroduplex mutagenesis

Mutagenesis via heteroduplex DNA of specific regions in the t-PA expression plasmid, pSVPA4, involves the following steps:

Preparation of ampicillin-sensitive pSVPA4 DNA

1. Plasmid pSVPA4 (15 µg) was completely linearized with Pvul. This mixture was extracted with phenol/chloroform and this DNA was precipitated using two volumes of ethanol with 0.1 M NaCl present. 2. This DNA was resuspended in a 21 µl water, 1 µl dNTB solution (containing 2 mM dATP , dGTP, dTTP, dCTP), 2.5 µl 10X nick translation buffer (0.5 M Tris-Cl pH 7.5, 0.1 M MgSC>4, 10 mM DTT, 500 pg/ml) and 0.5 µl or 2 units of DNA polymerase 1 large fragment from New England Biolabs. This mixture was incubated at room temperature for 30 minutes and then phenol/chloroform extracted, followed by ethanol precipitation as described above. 3. The precipitated DNA was resuspended in 0, 2 pg/pl when adding 75 pl of water.

Preparation of ampicillin-resistant pSVPA4 DNA

1. Plasmid pSVPA4 (15 pg) was digested with Sac I which cuts this plasmid twice in the t-PA coding sequence and gives two restriction fragments, a 1.4 kbp t-PA coding restriction fragment plus the parent vector. After restriction digestion, 1 µl or 28 units of calf intestinal alkaline phosphate from Boehringer Mannheim was added and then incubated at 37°C for 5 minutes. The two bands were separated by pouring the mixture onto a 0.7% agarose gel. The parent vector restriction fragment was excised from the gel and extracted by adsorption onto silica at 4°C, followed by elution in 50 mM Tris/1 mM EDTA at 37°C for 30 minutes. The eluted DNA was adjusted to a final concentration of 0.2 pg/µl.

Heteroduplex maturation

1. Mix 6 µl or 1.2 µg of ampicillin-sensitive pSVPA4 DNA with 6 µl or 1.2 µg of ampicillin-resistant pSVPA4 DNA. 2. Add an equal volume (12 µl) of 0.4 M NaOH. Incubate at room temperature for 10 minutes. 3. Slowly add 4.5 volumes or 108 µl of 0.1 M Tris-Cl pH 7.5/20 mM HC1. 4. 50 picomoles or 5 µl of phosphorylated mutagenic oligonucleotide was added to 45 µl of heteroduplex mixture. 5. This mixture was incubated at 68°C for two hours and then slowly cooled to room temperature.

Mutagenesis

1. Each mutagenesis reaction was adjusted to the following concentrations by adding 7 µl to the heteroduplex mixtures, 2 mM MgCl/0.2 mM ATP/α60 µm dATP, dTTP, dGTP, dCTP/4 mM DTT/40 units/ml Klenow -fragment E. voli DNA polymerase I (B.R.L.), 2000 units/ml T4 DNA ligase (N.E.B.). This mixture was incubated at room temperature in

2 hours.

2. The reaction mixture was then phenol-chloroform-extracted followed by ethanol precipitation. The precipitated DNA was resuspended in 12 µl of 50 mM Tris-Cl/1 mM EDTA. 4 µl of this was used to transform competent HB101 bacteria. 3. Ampicillin-resistant colonies were screened with 1 x 10^ cpm/ml of a <32>p-labeled probe oligonucleotide in 5X SSC, 0.156 SDS, 5X denhardf's reagent and 100 pg/ml denatured

salmon sperm DNA.

4. The filters were washed with 5X SSC, 0.1% SDS at a temperature 5°C below the calculated melting temperature of the oligonucleotide probe. 5. DNA was prepared from positively hybridizing clones and analyzed initially by digestion with various restriction enzymes and agarose gel electrophoresis. DNA was transferred to nitrocellulose and the filter was primed and hybridized to probes to ensure that mutagenic oligonucleotides were introduced into the correct

fragment.

6. The DNA was then retransformed into E. coli and ampicillin-resistant clones were screened for hybridization to the probe oligonucleotide. 7. The final mutations were confirmed by DNA screening according to Sanger.

Production of mutagenized cDNA' is: the M13 method

The following schematic restriction map shows a cDNA encoding human t-PA (above) with cleavage sites indicated for specific endonucleases indicated below.

The initiation codon, ATG, and the cDNA regions encoding (a), R<*>, R<2> and R<3> are indicated. Thus, mutagenesis at the N-terminus can be carried out using the SacI fragment or the Bglll/Narl fragment. Mutagenesis at Arg-275 and/or at R<1> and/or R<2> can be carried out by, for example, using the SacI fragment or the BglII/SacI fragment. Mutagenesis of R<3> can be carried out using an EcoRI/XmaI or EcoRI/ApaI fragment. The choice of restriction fragment may be determined based on considerations of expediency when using particular vectors for mutagenesis and/or for expression vector construction.

In general, the cDNA restriction fragment to be mutagenized can be excised from the full-length cDNA present, for example in pWGSM, pIVPA/1 or pSVPA4, using indicated endonuclease enzymes, and then mutagenized, for example, with the oligonucleotides shown in Table 7 or other oligonucleotides indicated for the desired mutagenesis.

Exemplary mutagenized cDNA fragments that can thus be produced are shown in table 8 below. After mutagenesis, the fragment, with or without further mutagenesis, can then be excised from the M13 vector and ligated back into an expression vector containing full-length or partial cDNA previously cleaved with the same enzyme(s) used for excision of the mutagenized fragment from the M13 vector. With this method, full-length cDNA, mutagenized as desired, can be reassembled using one or more mutagenized fragments as restriction fragment cassettes.

cDNAs encoding the following illustrative compounds (see

the table, page 9 as well as tables 2.0, 2.5 & 3) can be prepared from the mutagenized fragments from table 8 as follows:

The plasmids pIVPA or pSVPA4, in addition to being usable as expression vectors, can also be used as a "depot" in the construction of cDNA with any desired permutation of mutagenized sites. Thus "pIVPA/A" or "pSVPA4/a" mutagenized (via M13 or heteroduplexing) containing a desired modification in the cDNA region encoding the N-terminal region can be degraded with Narl (partial) and Xmal (Smal) (total ) to remove the cDNA region encoding the protein region spanning R<1>, R<2> and R<3>. A second pIVPA or pSVPA4 plasmid, if desired mutagenized (via M13 or heteroduplexing), at any combination of Arg-275, R<1->, R<2-> and R<3-> coding regions , can then be digested with Narl (total) and Xmal (Smal) (total) and the Narl/Xmal (Smal) fragment can then be identified, isolated and ligated into Narl/Xmal (Smal)-digested pIVPA/A or pSVPA4/ A. Such use of the Narl/ZmaI (SmaI) restriction fragment cassette allows, for example, the construction of the desired mutagenized cDNAs in pIVPA or pSVPA4. The mutagenized cDNA can then be transferred, for example, as a BglII/XmaI restriction fragment cassette into BglII/XmaI digested pWGSM for mammalian expression, if desired.

EXAMPLES

EXAMPLE 1

PRODUCTION OF GLN-117 OMISSION VARIANTS

A. Preparation of Gln-117 blunt cDNA.

cDNA molecules encoding the polypeptide sequence of compounds 2-1/N-21/Arg, 2-1/N-22/Arg, and 2-1/N-23/Arg were prepared using the oligonucleotide-directed mutagenesis method of Zoller and Smith. In particular, the mutagenesis vector RF M13/t-PA, containing the t-PA gene, was constructed from the mammalian t-PA expression plasmid pSVPA4. RF M13/t-PA was constructed by first digesting PSVPA4 to completion with restriction endonuclease Sac1. About. The 1436 base pair (bp) SacI fragment encodes a large portion of the polypeptide sequence of t-PA and includes the nucleotide sequence encoding the consensus N-linked glycosylation sites comprising asparagines 117, 184 and 218. This 1436 bp (hereafter referred to as 1.4 kbp) fragment was purified by preparative agarose gel electrophoresis.

The SacI fragment of this t-PA cDNA, obtained as a SacI fragment above, was ligated into a linearized double-stranded RF M13mpl8 DNA vector previously digested with SacI. The ligation mixture was used to transform transformation-competent JM101 bacterial cells. The M13 plaques containing t-PA-derived DNA prepared from transformed cells were identified and isolated by analytical DNA restriction analysis and/or plaque hybridization. Radiolabeled oligonucleotides (∼17-mer, with positive polarity) derived from within the Sac1 restriction sites of the t-PA coding nucleotide sequence listed in Table I were used as probes when filter hybridization was used to detect viral plaques containing t-PA DNA . All oligonucleotides were prepared by automated synthesis with an Applied Biosystems DNA synthesizer according to the manufacturer's instructions.

Several of the positive plaques detected by restriction or hybridization analysis were then further cloned by conventional plaque purification. Purified M13/t-PA bacteriophage obtained from the plaque purification procedure was used to infect JM101 cells. These infected cells yielded cytoplasmic double-stranded "RF" M13/t-PA plasmid DNA. The infected cells also produce bacteriophage in the culture medium containing single-stranded DNA complementary to the 1.4 kbp Sac1 fragment of t-PA and to M13 DNA. Single-stranded DNA was purified from the M13/t-PA-containing phage isolated from the culture medium. This single-stranded M13/t-PA DNA was used as a template in a mutagenesis reaction according to Zoller and Smith using oligonucleotide #3 in Table 7. This mutagenesis event changes the Asn codon to a Gln codon at position 117 of the then the following coding strand of DNA was obtained by changing the DNA sequence from "AAC" to "CAG". After the mutagenesis reaction, this DNA was transformed into the bacterial strain JM101. To identify mutagenized cDNAs, the transformant plaques were DNA hybridized using radiolabeled oligonucleotide #4 from Table 7. All exemplified oligonucleotides in Table 7 are of positive polarity, that is, represent portions of a coding rather than a non-coding strand of DNA. All hybridization-positive plaques were further purified by subsequent secondary infections of JM101 cells with M13-competent mutagenized DNA.

RF M13/t-PA plasmid DNA was purified from JM101 cells infected with purified M13 phage containing mutagenized t-PA cDNA. The thus obtained RF M13/t-PA plasmid containing the Gln-117 mutagenized Sac1 restriction fragment of t-PA DNA. This can then be mutagenized further, again according to Zoller and Smith, but using the oligonucleotides as described below. The oligonucleotides described below were designed to induce an omission ("deletion") in the cDNA region encoding the N-terminal domain.

Deletion mutagenesis 1; Oligonucleotide #8 in Table 7 induced a cDNA omission encoding Cys-6 through Ser-50. After this second mutagenesis reaction, this DNA was transformed into JM101 cells. To identify mutagenized cDNAs, the transformant plaques were screened as above, but using radiolabeled oligonucleotide #9 in Table 7. Hybridization-positive plaques can be further purified by subsequent secondary infection of JMlOl cells with M13 phage containing the twice mutagenized t-PA cDNA . The cDNA prepared below which contains this mutagenized restriction fragment encodes compound 2-1/N-21/Arg in which Ile-5 is covalently linked to Cys-51 by a peptide bond.

Deletion Mutagenesis 2: Oligonucleotide #10 in Table 7 induced a cDNA deletion encoding Cys-6 through Ile-86. After this second mutagenesis reaction, this DNA was transformed into JM101 cells. To identify mutagenized cDNAs, the transformant plaques were screened as above, but using radiolabeled oligonucleotide #11 in Table 7. Hybridization-positive plaques can be further purified by subsequent secondary infection of JM101 cells with M13 phage containing the twice mutagenized t-PA cDNA . The cDNA prepared below which contains this mutagenized restriction fragment encodes compound 2-1/N-22/Arg in which Ile-5 is covalently linked to Asp-87 by a peptide bond.

Deletion Mutagenesis 3: Oligonucleotld #12 1 Table 7 induced a cDNA deletion encoding Cys-51 through Asp-87. After this second mutagenesis reaction, this DNA was transformed into JM101 cells. To identify mutagenized cDNAs, the transformant plaques were screened as above, but using radiolabeled oligonucleotide #13 from Table 7. Hybridization-positive plaques can be further purified by subsequent secondary infection of JM101 cells with M13 phage containing the twice mutagenized t-PA cDNA . The cDNA prepared below which contains this mutagenized restriction fragment encodes compound 2-1/N-23/Arg in which Ser-50 is covalently linked to Thr-88 by a peptide bond.

Each of these mutagenized restriction fragments can then be ligated back into the mammalian expression vector pSVPA4 as a SacI cassette by methods analogous to those described in Example 3B, or prepared for insertion into the insect cell expression vector pIVPA/1 (ATCC No. 39891) as a Bglll/SacI -cassette derived from modified RF M13/t-PA DNA.

B. Production of vectors used for expression of

high<y>mannose Gln- 117- deletion variants.

The purified RF M13/t-PA containing modified and truncated t-PA cDNA, prepared as described above, can be degraded with the restriction endonucleases BglII and SacI. The approximately 1.2 kbp BglII/SacI restriction fragment was purified by conventional preparative gel electrophoresis. The BglII/SacI fragment thus obtained constitutes a mutagenized cassette lacking a 5' and 3' portion of DNA encoding the amino and carboxy termini of the translated protein.

Insect expression vector pIVPA/1 (ATCC No. 39891) contains a wild-type t-PA cDNA insert operably linked to a polyhedrin promoter together with baculovirus flanking DNA sequences. pIVPA/1 was digested with BglII and SacI to excise a t-PA coding region spanning the N-terminus and R<*> and R<2>. The Bglll/SacI cassettes containing the mutagenized, N-terminally modified t-PA cDNA fragments can each then be ligated to pIVPA/1 expression vector DNA which has been previously purified after digestion with Bglll and SacI. The resulting plasmids, pIVPA/A FBR; Gln-117, pIVPA/A FBR/a EGF; Gln-117; pIVPA/A EGF, Gln-117 shall contain the mutagenized cDNAs encoding the compounds 2-l/N-21/Arg, 2-l/N-22/Arg and 2-l/N-23/Arg, now operational bound to the polyhedrin promoter. The nucleotide sequence of each mutagenized cDNA insert can be confirmed by supercoiling sequencing with plasmid as substrate, see E.Y. Chen et al., 1985, "DNA" 4(2): 165-170.

B. Introduction of the mutagenized cDNA into the insect virus. Each of the pIVPA plasmids containing the mutagenized cDNAs can be introduced into the insect virus by cotransfection with wild-type AcNPV in Spodoptera cells. 1 pg of purified Autographa californica NPV DNA and 10 pg of the desired pIVPA DNA are introduced into Spodoptera cells growing on tissue culture plates by a calcium phosphate transfection procedure (Potter, K.N. and Miller, L.K., "J. Invertebr. Path." (1980) , 36 431-432). The simultaneous introduction of these DNAs into the cells resulted in a double recombination event between the pIVPA plasmid containing the mutagenized cDNA<*>s and the viral DNA in the regions of homology between the two; that is, the polyhedra region from the progeny virus from the recombination event contains the mutagenized cDNA insert from the pIVPA plasmid.

Isolation of viruses containing the nucleotide sequence that encodes the proteins according to the invention.

The progeny virus present in the medium above the transfected cells plaques on a fresh monolayer of cells at several different dilutions. These plaques are analyzed and recombinants are selected based on the PIB-minus phenotype as follows: A virus that has lost its polyhedrin as is the case with a virus containing a mutagenized cDNA will not produce PIBs. Plaques that appear to lack PIB are selected, excised and amplified on healthy cells. The supernatant of these cells is then analyzed for t-PA-type enzymatic activity. Positive analysis results suggest that the glycoprotein has indeed been produced.

An alternative method of vulrus cleansing via the plaque-raising protocol differs slightly from the steps described above and will be described below. The progeny virus from transfection plaques by appropriate dilution on cell culture plates. A nitrocellulose replica of the cell monolayer and virus plaques is produced. The agarose overlay is removed from the plate as a virus source after the results of the following steps have been obtained.

The nitrocellulose filter is probed with radioactive DNA fragments that are representative of the gene that is placed in the viral chromosome. Those containing the foreign gene are considered positive. The hybridized probe is removed. The filter is probed again with radioactive DNA that is representative of a part of the viral chromosome that is removed by substitution with foreign DNA. Those who still have a polyhedra ring are considered positive.

The hybridization probe is removed and the filter is probed again with a radioactive DNA fragment that identifies viral plaques regardless of the state of the polyhedra. A suitable fragment may be the EcoRI I fragment. These are judged as progeny viruses. Plaques are selected that are positive for the foreign gene DNA probe, negative for the polyhedra ring probe and positive for the viral DNA probe. These are strong candidates for the desired genotype.

C. Preparation and characterization of high mannose

glycoprotein.

Antibodies have been used to demonstrate the presence of variant proteins in the extracellular media of infected cells. Recombinant virus, prepared as above, is used to infect cells used in normal TC-100 nutrient saline, but instead of the standard media supplement of 1056 fetal calf serum, this was replaced with a 5056 egg-yolk enrichment (to 156 total volume). Previous experiments have shown a more intact protein under these conditions. The supernatant from the infected cells is fractionated on an affinity column which has attached a monoclonal antibody to natural human t-PA. The protein specifically retained by the column is eluted and analyzed for t-PA enzymatic activity. A fixed amount of activity units of this and control t-PA preparations was separated on an acrylamide gel. This gel is then stained with a silver-based reagent to show the protein pattern. This reveals that, after infecting the insect cells, the virus leads to the extracellular production of a protein with t-PA-type activity.

Radiolabeled protein is prepared for further characterization by first incubating Spodoptera frugiperda cells infected with the virus (m.o.i=1) for 48 hours. The culture plates are then filled with methionine-deficient media. Methionine-deficient media supplemented with <35>S-methionine is then added to the culture plates. The cell cultures are incubated for 4 hours. The supernatant containing the radiolabeled glycoprotein can be analyzed by SDS-PAGE, 7.556, against wild-type (that is, full-length fully-glycosylated) t-PA, produced in an analogous manner 1 insect cells and against mammalian t-PA produced, for example, by the method of R. Kaufman et al., "Mol. Cell. Biol." 5(7):1750

(1985) but in the presence of tunicamycin (non-glycosylated). The partially glycosylated stub proteins produced in example 1 should have increased gel mobility compared to the fully glycosylated analogue and the non-glycosylated full-length analogue.

EXAMPLE 2

PRODUCTION OF OTHER PROTEINS ACCORDING TO THE INVENTION

A. Preparation of other cDNAs

The mutagenesis methods according to example 1 can be used with other conventionally produced synthetic oligonucleotides that modify the original t-PA DNA sequence to produce proteins that are modified in the N-terminal region and/or optionally modified at N-linked glycosylation sites and/or at Arg-275 with the appropriate codon change(s) described previously, see for example "Preparatlon of Mutagenized cDNAs: M13 Method" as well as routes (a) - (h) above.

For example, cDNAs encoding compounds D-6, D-1, and D-3 can be prepared using the Sac1 restriction fragment in M13 t-PA and mutagenizing with oligonucleotides #8, 10, and 12, respectively, but not with oligonucleotide #3. Arg-275 can be omitted or replaced, for example with Thr, when using the oligos 14 and 15 respectively. Vector construction, transfection and expression can be carried out as in example 1 for insect cells as described below in example 3 for mammalian cells.

Single-stranded DNA from the M13 mutagenesis vector (RF M13/t-PA), prepared as in Example 1, can also be used as a template to mutagenize in a site-specific manner at Arg-275 and/or at the glycosylation site R<1> or R< 2> or both. The region encoding the consensus tripeptide comprising Asn-218 can be similarly mutagenized. An interactive process can be used to produce multiple modifications of the protein at these sites. After identification and purification of the M13 phage containing a modified R^-site, for example, a single-stranded DNA containing this modified site can be purified from the phage and used as a template to initiate a second round of mutagenesis within the R<2->site and/ or at Arg-275. This process can be repeated until all desired modifications have been achieved. Then cDNA encoding the compounds 2-2/N-23/Arg, 2-2/N-21/Arg and 2-2/N-22/Arg can be prepared by the method of Example 1, but by replacing the mutagenesis oligonucleotide # 5 with oligonucleotide #3 and to probe oligonucleotide #6 for oligonucleotide #4. cDNA encoding the compounds 2-6(N-21/Arg, 2-6/N-22/Arg and 2-6/N-23/Arg can be prepared by twice mutagenizing the Sac1 fragment as described in Example 1 and by in addition to mutagenizing and probing with oligonucleotides #5 and #6 Vector construction, transfection and expression are as indicated in Example 1 for insect cells or as described below for mammalian cells, see lanes (a)-(h) above.

The RF M13/t-PA mutagenesis vector does not contain the DNA sequence encoding R<3>, the N-linked glycosylation site of t-PA most proximal to the carboxy terminus of the protein. To therefore carry out DNA modifications at this site, a new M13/t-PA mutagenesis RF vector called M13/t-PA:R1-Xmal was created. This vector was constructed by completely digesting M13 vector M13mp11 with EcoRI and XmaI. R1/Xma1 digested M13 vector was ligated to a purified EcoRI/Xma1 t-PA restriction fragment (approximately 439 bp, hereafter called 0.4 kbp) encoding a polypeptide region encompassing the R<3> glycosylation site. This 0.4 kbp restriction fragment was purified after digestion of the plasmid pWGSM with EcoRI and XmaI. Mammalian expression plasmid pWGSM encoding the t-PA gene is identical within the 439 bp EcoRi/XmaI fragment to the plasmid pLDSG described by Kaufman et al., "Mol. Cell. Biol.", 5, 1750-1759

(1985).

The ligation mixture was used to transform competent bacterial JM101 cells. Various plaques were picked and analyzed for the presence of the 0.4 kbp t-PA EcoRI/XmaI fragment by standard DNA restriction fragment analysis. Double-stranded RF M13 DNA was purified from cells containing the 0.4 kbp t-PA fragment. This DNA was named the RF M13/t-PA:RI-XmaI mutagenic vector. As previously indicated in Example IA, this vector when transformed into competent JMlOl cells can be used to produce M13/t-PA:RI-XmaI phage from which single-stranded M13/t-PA:RI-XmaI DNA can be purified . This single-stranded DNA can be used as a template in the site-directed mutagenesis reaction to modify the t-PA DNA and the N-linked glycosylation site R<3>.

Modified R<3> coding sequence can be used to replace the wild-type R<3> sequence present in either modified pIVPA/1 as prepared in Example 1 (blunt and/or modified at R<1> and/or R<2> ) or wild-type pIVPA/1 plasmid DNA. This can be carried out by first carrying out a total Sacl/ApaI degradation of the R<3->modified Ml3/t-PA:RI/XmaI mutagenesis plasmid vector and isolation of the thus produced R<3->modified 165 base pair t- PA restriction fragment. The insect expression vector pIVPA/1 or pIVPA/1 plasmid DNA modified, for example, as in Example 1, can similarly be completely digested with SacI and ApaI to excise the 165 bp wild-type t-PA restriction fragment encoding the unmodified R<3->seat. Ligation of the purified insect expression vector lacking the 165 bp fragment to the modified R<3 >165 bp fragment yields a new insect expression vector. Expression of the vector yields a blunted protein that is modified at the R<3> site as well as at any or all other consensus N-linked glycosylation sites present in natural

t-PA and/or at Arg-275.

The pIVPA plasmid containing the modified cDNA can also be used to generate the Bglll/ApaI fragment of the modified t-PA cDNA spanning the omission reglon 1 N-terminal domain as well as the region encoding R<*>, R<2> and R<3> or the Nar I/Xmal fragment spanning R<*>, R<2> and R<3>. Any of these fragments can be inserted into mammalian expression vectors such as pSVPA4 or pWGSM as described in Example 3.

EXAMPLE 3

PRODUCTION OF COMPOUNDS D-6, D-1 AND D-3 IN MAMMALIAN CELLS

A. Preparation of cDNA

cDNA molecules encoding the polypeptide sequences of compounds D-6, D-1, and D-3 were prepared using the mutagenesis oligonucleotides #8, 10, and 12, respectively, and the SacI fragment of t-PA cDNA as a template by the M13 method of Example 1 or by heteroduplex mutagenesis (Moranaga heteroduplex mutagenesis protocol; both, supra). Mutants selected by DNA hybridization using probe oligonucleotides 9, 11 and 13 respectively were confirmed by DNA

sequence analysis to be correct in the modified DNA sequence.

B. Preparation of modified t-PA vector

Each modified cDNA prepared in Example IA (a, Gln-117) or 3A (a) was first removed from the M13 mutagenesis vector RF M13/t-PA by total digestion of the vector with Sacl. It approx. The 1.4 kbp restriction fragment in each mutagenized cDNA was purified by gel electrophoresis and then ligated into pSVPA4 as follows. First, pSVPA4 was digested with Sacl to remove the wild-type t-PA 1.4 kbp restriction fragment. The remaining portion of Sac1-digested pSVPA4 was then ligated to the 1.4 kbp restriction fragment of the mutagenized cDNA. This ligation event can give two orientations of the inserted fragment. The correct orientation in each case can be identified using EcoI and PvuII as enzymes by conventional analytical restriction enzyme analysis. This replacement allows the SacI fragment to be used as a cassette fragment between the RF M13/t-PA mutagenic vector and the pSVPA4 mammalian expression vector. Modified M13 SacI fragments (blunt and optionally modified at R<1> and/or R<2>) can be inserted into SAcI-digested pSVPA4 DNA which is previously or subsequently modified at R<3>, if this is desired. Alternatively, DNA previously modified at R<1>, R<2> and R<3> can be excised from vectors such as pIVPA or pSVPA4 as a NarI/ApaI or NarI/XmaI fragment. The thus obtained fragment can be inserted into vectors such as pSVPA4 or pWGSM, previously broken down with Narl (partially) and ApaI or XmaI (totally). By this method, any combination of N-terminal deletion and/or substitution and/or glycosylation site mutagenesis and/or Arg-275 mutagenesis can be achieved.

C. Transfection of COS (SV40-transformed African green monkey liver) cells

COS-1 cells (ATCC CRL 1650) were transfected by the method of Lopata, M.A. et al., "Nucl. Acids Res.", 12:5707-57171

(1984) with the vectors prepared in Example 3B, i.e. modified pSVPA4. Serum-containing medium was replaced with serum-free medium 24 hours after transfection and conditioned medium was analyzed both for the presence of plasminogen-activating activity, using the chromogenic substrate S-2251, or the presence of t-PA antigen by an ELISA assay, 48 and 72 hours after transfection.

D. Viral propagation 1 CV1 (African green monkey kidney) cells

Modified complex carbohydrate protein can be prepared by infecting CV1 cells (ATCC CCL 70) with SV40 viral stock propagated as described by Gething and Sambrook ("Nature", 293:620-625. 1981). This was accomplished by first completely digesting modified pSVPA4 with the restriction endonuclease BamH1 to remove the bacterial shuttle vector pXf3 from the SV40 viral DNA. Prior to transfection of this DNA into CV1 cells along with helper virus SV40-rINS-pBR322 DNA (described below), BamH1 linearized SV40/t-PA DNA is circularized by ligation at low DNA concentrations (1 µg/ml). This process was repeated with insulin containing the SV40 vector SV40-rINS/-pBR322 (Horowitz, M. et al., 1982, "Eukaryotic Viral Vectors", pp. 47-53, Cold Spring Harbor Laboratory). The bacterial shuttle vector pBR322 in SV40-rINS-pBR322 was removed by a total EcoRI digestion. The linearized insulin/SV40 viral DNA was then circularized by ligation at a DNA concentration of 1 pg/ml. It is necessary to transfect CV1 cells with circularly ligated pSVPA4 and SV40-rINS DNAs, at equimolar amounts, to generate viral stocks. SV40-rINS is used to provide "late" SV40 proteins, while pSVPA4 provides the "early" SV40 proteins necessary for viral DNA production, while also encoding the proteins according to the invention. As a result, when cells are transfected with both these DNAs as described by Gething and Sambrook, SV40 virus containing DNA from both viral vectors is produced. Subsequent infection of CV1 cells with amplified virus has produced proteins with t-PA-type activity that can be analyzed 72 hours after infection as described in example 3C.

EXAMPLE 4

MANUFACTURE OF OTHER PROTEINS

cDNAs that encode different proteins according to the invention are prepared by the methods according to examples 1, 2 and 3. The Bglll/Xmal restriction fragment cassette can then be excised from both the pIVPA or pSVPA4 vector containing the cDNA that encodes the blunt protein with or without modification on one or more glycosylation sites. The excised BglII/XmaI fragment can then be ligated into the BglII/XmaI cut pSVPA4 or pWGSM for introduction into mammalian cells. Expression of such cDNA in mammalian host cells, for example According to Example 3 or by the method according to Kaufman et al., supra (CHO host cells) or by the method according to Eowley et al., US-PS 4,419,446 (1983) (BPV- expression systems) yield the corresponding mammalian-derived blunt proteins. Thus, cDNAs encoding the compounds 2-1/N-21/Arg (α FBR, Gln-117) and 2-1/N-22/Arg (α FBR/EGF, Gln-117) were prepared and injected into pSVPA4 as described above. cDNA encoding the compound 2-1/N-23/Arg (a EGF, Gln-117) was prepared using mutagenesis oligonucleotide 12 and probe oligonucleotide #13 (Table 7), but by the heteroduplex method described above with pSVPA4 as in was previously mutagenized at position 117 (as above) as a template. Similarly, cDNAs encoding compounds D-2 (α FBR) and D-3 (α EGF/FBR) were prepared by M13 mutagenesis as described above and inserted as the SacI fragment into SacI-digested pSVPA4. cDNA encoding compound D-6 (a EGF) was prepared by the heteroduplex method as described above using pSVPA4 as template and mutagenesis oligonucleotide #12, and probing with oligonucleotide #13.

To prepare the cDNAs encoding the proteins for amplification and expression in mammalian cells, the cDNA contained in pSVPA4 or pIVPA was excised as a BglII/XmaI fragment and ligated into purified BglII/XmaI digested pWGSM. In each case, the resulting pWGSM vector was introduced into CHO cells and amplified according to Kaufman, above. The transformed and enhanced CHO cells produce the compounds D-6, D-1, D-3, 2-1/N-23/Arg, 2-1/N-21/Arg and 2-1/N-22/Arg which were detected in the culture medium by human t-PA specific antibodies. The compounds can then be recovered and purified by immunoaffinity chromatography.

EXAMPLE 5

Example 4 can be repeated using cDNA encoding the proteins modified within the N-terminus and/or at R-275 with or without modification at R<1>, R<2> and/or R<3> to produce the desired protein in CHO cells. Mutagenic cDNA<*>s can be prepared as described above. Thus, cDNAs encoding the compounds 2-7/N-23/Arg, 2-7/N-21/Arg and

2-7/N-22/Arg prepared in pIVPA as described in Example 1. The cDNAs can then be excised as a Bglll/Xmal fragment and ligated into purified, Bglll/Xmal digested pWGSM, and the resulting vector transformed and amplified in CHO cells as in example 4 to thereby produce the compounds

|2-7/N-23/Arg, 2-7/N-21/Arg and 2-7/N-22/Arg.

Claims (4)

1. DNA molecule encoding a thrombolytic protein with tissue plasminogen activator type activity, characterized by an amino acid sequence substantially the same as the peptide sequence of human t-PA, where: a) the sequence from Cys-6 to Ile-86 is omitted, b) at least one of the consensus N-linked glycosylation sites is modified to something other than a consensus N-linked N-glycosylation site, and where appropriate c) modification of the enzymatic cleavage site Arg-275/Ile-276 has been carried out in that Arg is omitted or replaced by another amino acid. 2. DNA molecule according to claim 1, characterized in that the consensus N-linked glycosylation site at Asn-117 is modified so that Asn-117 is replaced by Gln-117. 3. DNA molecule according to claim 1, characterized in that the three consensus N-linked glycosylation sites at amino acid positions 117-119, 184-186 and 448-450 are modified so that Asn-117, Asn-184 and Asn-448 are replaced with Gin . 4. DNA molecule according to claim 1, characterized in that a) the sequence from and including Gly-(-3) to and including Thr-91 is omitted and b) at least one of the consensus N-linked glycosylation sites is modified to something other than a consensus -N-linked glycosylation site.