CA1217718A - Plasminogen activator derivatives - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE
Derivatives of a nonimmunogenic plasminogen activator which comprises at least one polyalkylene glycol group chemically bonded with at least one coupling agent to amino acid side chains of said plasminogen activator, wherein said polyalkylene glycol has a molecular weight of about 200-20,000 and is unsubstituted or is substituted with one or more alkyl, alkoxy or alkanoyl groups or a mixture thereof.
The plasminogen activator derivatives have an extended circulating life in the mammalian bloodstream and also inhibit the formation of thrombus in the same.

Description

~ ~ ~Lt ,~7.~"~

BACKGROUND OF THE INVENTION
Field of the Invention ~ his invention relates to derivatives of human-originated non-immunogenic plasminogen activators, and more particularly to such a derivative comprising at least one polyalkylene glycol attached with at least one coupling agent to amino acid side chains of a plasminogen activator of the type described above, the polyalkylene glycol having a molecular weight in the range of 200 -10 20,000 and optionally containing one or more alkyl, alkoxy and/oralkanoyl groups as substituents. Further, the inveniion is concerned with a process for producing such deri-vativesand with a thrombolytic agent containing such derivativeS .
Description of the Prior Art It is known that human tissues contain a variety of substances which activate plasminogen into a fibrinoly-tic enzyme or plasmin. Among such known substances, 'he most representative is a plasminogen activator, i.e.
20 urokinase, which is formed in the kidney tissue and excre~d into urine. Urokinase may be obtained by isolation and purification from human urine, tissue culture or genetic engineering. As fibrinolytic enzyme activators which hav~
nowadays found widespread commercial utility, there exist proteins originated from hemolytic Streptococcus and urokinase which is an enzyme originated from human urine. In view of its immug~nic behavior towards humans,urokinase re~ulting from such source is favorably employed for clinical application.

:~`

7~ ;

-2-Urokinase originated from human urine is believed to con-tain both high molecular weight urokinase (molecular weight: 54,000) and low molecular weight urokinase (molecu-lar weight: 33,000). Uroklnase has been used, in recent years, as a thrombolytic agent or an adiuvant for carcino-static substances, and its consumption for clinical application`is increasing year by year.
However, urokinase is unstable under certain conditions since it is an enzyme and loses its enzymatic activity, for example, in the course of extraction, isolation and purification from a urokinase-bearing raw material, for example, urine; during the lyophilization processing in preparing dosable formulations; during tne heat treatment for deactivating viruses, or when it is placed in a diluted state in a dripping bottle and kept for a prolonged time period in such a diluted state at room temperature for clinical application. This physically unstable nature of urokinase has created a serious problem in preparing and formulating urokinase on an industrial scale or in actually using the same for clinical purposes.
Human albumin has been employed as an additive to urokinas~
so as to improve its stability. However, this can be by no means a break-through solution to the problem just dis-cussed because pure albumin, i.e. a globulin fraction, is difficult to obtain without immunogenic contamination;
pure albumin is expensive; albumin and urokinase form a complex of a high molecular weight under virus deacti-~ Z~7 7~

vating conditions in which urokinase is subjected to heattreatment at 60C for 10 hours together with albumin added to stabilize urokinase; and such stabilizer if added may be effective to a certain extent for protecting urokinase from losing its enzymatic activity upon the lyophilization but cannotPreVentits loss of activity upon actual clinical use.
The physiological activity of urokinase when admi-nistered intravenously to living bodies is promptly retarded by protease inhibitors present in blood (~2-macroglobulin, and a2-plasmin inhibitors and the like), and the metabolic rate of urokinase per se is ve~y hig;~, -es~l.-ing in extremely shortened half-life which does not exceed even several minutes. Nothing has been heretofore propos-d to solve the problem of s~ort half-life of urokinase in blood. - -_ The present inventors have carried out extensive research with a view toward developing derivatives of human-oriyinated non-immunogenic plasminogen activators which will overcome the above-noted drawbac~s of the prior art techniques. As a result, they have succeeded in ind-ing plasminogen activator derivatives whicn are stable and hardly retarded by inhibitors present in blood and hence achi2ve prolonged half-life in blood, thereby leading to the present discovery.
::' SUMMARY OF T~E INVENTION
-An object of the present invention is to provide a derivative of a human-originated non-immunogenic ~ .

7~

plasminogen activator which is stable and exhibits prolong-ed fibrinolytic activity when administered to living bodies.
Another object of the invention is to provide a process for preparing the novel plasminogen activator derivative.
A further object of the invention is to provide a therapeutically acceptable thrombolytic agent comprising the novel plasminogen activator derivative.
These and other objects and advantages of the inven-tion can be attained by the provision of a derivative of a human-originated ncn-immunogeniC plasminogen activator, comprising at least one polyalkylene glycol attached with at least one coupling agent to amino ~cid side chains of the plasminogen activator, the polyalkylene glycol having a molecular weight in the range of 200 -20,000 and option-ally containing one or more alkyl,alkoxy and/or alkanoyl groups as substituents.

l.Z~7~3 -4a-Thus in one aspect the present invention provides a composition for inhibiting the formation of thrombus in a mammalian bloodstream, which composition comprises an effective amount of a nonimmunogenic plasminogen activator comprising at least one polyalkylene glycol moiety chemically bonded with at least one coupling agent to the amino acid side chains of said plasminogen activator, wherein the polyalkylene glycol has a molecular weight in the range of 200-20,000 and is unsubstituted or is substituted with one or more alkyl, alkoxy or alkanoyl groups or a mixture thereof, together with a pharmaceutically acceptable carrier.
BRIEF DESCRIPTION OF THE DR~WINGS
A more complete appreciation of the present inven-tion and many of the attendant advantages thereof will be readily obtained as the invention becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:
FIG. 1 shows the optimum pH ranges of modified high molecular weight urokinase (PEG-DCI-UX) and unmodified high molecular weight urokinase (molecular weight: 54,000) as lZ~7~
_5_ measured in terms of the amidase activity with a synthetic substrate (S-2444), where the open circles correspond to PEG-DCT-UX and the closed circles to unmodified urokinase;
FIG. 2 shows the optimum pH ranges of modified low molecular weight urokinase (PEG-DCT-L-UK) and unmodified low molecular weight urokinase (molecular weight: 33,000) as measured in terms of the amidase activity with S-2444, where the open circles correspond to PEG-DCT-L-UK and the closed circles to unmodified urokinase;
FIG. 3 diagrammatically shows the stability of PET-DCT-UK and unmodified uro~inase at room temperature in either physiological saline or a Ringer solution, where the solid and broken lines indicate the physiological saline and the Ringer solution, respectively, and where the open circles correspond to PEG-DCT-UK and the closed circles to unmodi-fied urokinase; - _ FIG. 4 shows the stability of PEG-DCT-L-UK and un-modified urokinase (molecular weight: 33,000) against freezing, followed by thawing operation repeated 2, 4 and 6 times, where the open circles correspond toPEG-DCT-L-VK
and the closed circles to unmodified urokinase;
FIG. S shows the varied amounts, as a function of time, of a plasmin inhibitor in a plasma fraction Fl of each of two groups of healthy rabbits, one group adminis-tered with PEG-DCT-UK and the other group with unmodified urokinase (molecular weight: 54,000), where the open cir-cles correspond to the PEG-DCT-UK-administered group and the closed circles to the unmodified urokinase-administered group;

-6- 1~7~
FIG. 6 shows the varied amounts, as a function of time, of plasminogen in a plasma fraction F3 of each of two groups of healthy rabbits, one group administered with PEG-DCT-UK and the other group with unmodified urokinase (mole-cular weight: 54,000), where the open circles correspond to the PEG-DCT-UK-administered group and the closed circles to the unmodified urokinase-administered group; and FIG. 7 shows the varied amounts, as a function of time, of plasminogen in a plasma fraction F3 of each of two groups of healthy rabbits, which plasma was treated w1th an acid and then with a base to inactivate a plasmin inhibitor, one group administered with PEG-DCT-UK and the other grou~
with unmodified urokinase (molecular weight: 54,000), where the open circles correspond to the PEG-~CT-UK-administered group and the closed circles to the unmodified urokinase-administered group.
Figure 8 illustrates the substrate specificity of PEG-Modified UK's. _ _ Figure 9a illustrates the stability of modified Urokinase to UK-Inhibi~ors in Human Plasma.
Figure 9b illustrates the effect of Placental UK-Inhibitor on the Urokinase Activity.
Figure 10 illustrates UK activity in Plasma after UX
injection.
Figure 11 illustrates the relative radioactivity in Blood (%1 for 125I-M-Md and 125I-Native-UK.

:

J ~'~

~7- lZ~77~
Figure 12a illustrates a comparison of the immunogenicity of M-Md with that of native UK by Schultz-Dale test.
Figure 12b illustrates a comparison of the immunogenicity o~ M-Md with that of native UK by Passive Cutaneous Anaphylaxis (PCA) reaction.
Figure 13a illustrates a compaxison of the Plasmin Inhibitor in Fl from Citrated Dog Plasma at certain time intervals after injection of Native UK or PEG-UK into a dog.
Figure 13b illustrates a comparison of the Plasminogen in F3 from Citrated Dog Plasma at certain time interYals after injection of Native UK or PEG-UK into a dog.
Figure 13c illustrates a compari50n of the FDP in Dog Serum at certain time intervals after i.v. Injection of Urokinase or PEG-UK in~o a dog.
Figure 14 illustrates a comparison of FDP levels in dogs bearing artificial thrombus in A. femoralis. Dog A had native UK injected into a proximal branch of the thrombus, while Dog B had PEG-UK injected into the same.

- DESCRIPTION OF THE PREFERRED E~ODIMENTS
By the term "human-originated non-immunogenic plas-minogen activator" as used herein are encompassed not on;y urokinase but also tissue plasminogen activators obtained from human tissues such as uterine, tumor and the like.
These human tissue plasminogen activators also contain those obtained by tissue culture or genetic engineerins.
It should be noted that no limitations are imposed on the molecular weishts of these activators so long as they are obtained in the above-described manner. For example, as urokinase which is a plasminogen activator origina.ed from human urine, high molecular weight urokinase (molecular .

7~3 weight: S4,000) and low molecular weight urokinase (molecu-lar weight: 33,000) may be used solely or in combination.
Suitable polyalkylene glycols which may be used in the invention include a polyethylene glycol and a polypro-pylene glycol. In the case of the polypropylene glycol, both straight-chain polypropylene glycols such as those represented by ~O~CH(CH3)CH2O]nH and branched-chain poly-propylene glycols such as those represented by CH3CH2C-{CH20~cH2cH(cH3)o~nH}3 or HIOCH(CH3)CH2~nOCH{cH2~ocH2cH-(C-~3)]nOH}2.
The molecular weights of the polyalkylene glycols may range from 200 to 20,000. Particularly preferred molecular weights are in the range of 500 - 10,000.
The polyalkylene glycols each may o?tionally contain one or more alkyl, alkoxy and/or alkanoyl groups as substituent groups. Typical example,s of the alkyl groups are methyl, ethyl, propyl,stearyl and the like. Typical examples of the alkanoyl groups are acetyl, propibnyl, stearoyl and the like. As a preferred polyalkylene glycol, an unsubsti-tuted or methyl-substituted polyalkylene glycol is useful.
However, of particular interest is the use of methoxy-polyethylene glycol as the polyalkylene glycol. Chemical modification of urokinase(UK) with activated methoxypoly-ethylene glycol(PEG) of MW 5,000 increased the stability of the urokinase and imparted a markedly extended circulation life in rabbits and rats. Also of interest is the surprising fact that PEG-UK has a superior thrombolytic ability as compared to that ~`;

:~Z~77~
g .
of native UK. It appears likely that the superiority of PEG-UK to native UK with respect to the fibrinolytic activation is from the PEG chains which protect the UK
molecule from interation with inhibitors, thus extending its circulation life.

Suitable coupling agents which may be used in the invention and are adapted to attach a polyalkylene glycol to a non-immunogenic plasminogen activator, for example, urokinase, include those capable of reacting with amino acid side chains of the protein to be modifiedand forming chemical bonds therebetween, for example,.acyl azide, cyanuric halides, p-diazoniumbenzyl ether, 3-(p-diazonium-phenoxy)-2-hydroxypropyl ether, dihalogenosuccinic anhydride and the like. The following partial formulae may be given as examples of the coupling structures between a polyalkylene glycol and urokinase through these coupling agents.

-~ -UK

X
_o- ~ -UK

OH
-O-CH2- ~ N2-UK

-o-cH2cH(OH)cH2 ~ -r~2-UK

wherein X is a halogen atom, and UK is a residual part of the urokinase ~olecule.
The novel urokinase derivative according to the invention can ~e prepared by reacting a coupled product of at least one corresponding polyalkylene glycol and at least one coupling agent with urokinase, the polyalkylene glycol having a molecular weisht in the range of 2Q0 -20,000 and optionally containing one or more alkyl , alkoxy and~or alkanoyl groups as substituents.
Typical examples of the polyalkylene ~lycol-cou~lins agent coupled product include oolyalkylene glycol-~,6-.~2~

dichloro-1,3,5-triazine, polyalkylene glycol-4,6-difluoro-1,3,5-triazine, polyalkylene glycol-4-chloro-6-hydroxy~
1,3,5-triazine, polyalkylene glycol-~-(bro~ocarbonyl)-monopropionate, polyalkylene glycol-azidocarbonyl methyl etller, polyalkylene glycol-(p-diazoniumbenzyl) ether, polyalkylene glycol-3-(p-diazoniumphenoxy)-2-hydroxypropyl ether and the like.
When reacting the polyalkylene glycol-coupling agent coupled product with uroki~ase, it is necessary to choose such reaction conditions that the enzymatic acti-vity is held to a minimum loss. Namely, it is desirable to carry out the reaction at low temperatures, for example, at a temperature of 0C to room temperature in an a~ueous solution such as a buffer. Preferred reaction time may range from several minutes to 5 hours. The pH of the buffer is preferably within such a range that the enzymatic activity of urokinase is not lowered, namely, 2 - 10, preferably 5 - 9. However, the preferred pH range may vary depending on the reactivity of each coupling agent employed and/or the nature of an amino acid residue. The modification degreesof amino acid side chains of the plasminogen activator can be controlled by changing the concentration of the polyalkylene glycol activated witn the coupling agent in a reaction medium.
By wày of illustration, monomethyl ether polyethylene glycol-4,6-dicnloro-1,3,5-triazine having an average molecular weight of 5,000 was reacted at pH 7.0 with hig~

lZ~7~

molecular weight urokinase to obtain novel urokinase derivatives, while changing the concentration of the former reactant to 0.4, 4.0 and 6.0 mM, respectively. Unmodified -amino groups of lysine of the resultant urokinase deri-vatives were quantitatively determined using sodium ~,4,6-trinitrobenzenesulfonate. Their modification degrees were investigated on the basis of the results obtained by the quantitative analyses. The modification percentages of the ~-amino groups of lysine, which were reactive with sodium 2,4,6-trinitrobenzenesulfonate, were 6-7% at 0.4~M, about 40% at 4.0 mM and about 60~ at 6.0 mM. In addition, the molecular weights of these reaction products were determined by SDS polyacryl amide gel electrophoresis.
The average molecular weights of the reaction products were about 60,000 at 0.4 mM and about 120,000 at 4.0 m~1. This finding is substantially in conformity with the results obtained above by the quantitative analyses of the E-amino groups of lyslne. Accordingly, it is necessary to conduct the reaction between the polyalkylene glycol activated with the coupling agent and the plasminogen activator at a high pH level and at a high concentration of tne former reagent when increased modification degrees of the plasminogen activator are desired. On the other hand, where decreased modification degrees are preferable, tne xeaction should be effected at a relatively low pH level and at a low con-centration of the coupled product. The modification degrees .. . .

i~`77 ~

may of course be changed by controlling the reaction time.
By suitably combining these reaction conditions, it is possible to obtain the intended novel plasminogen activator derivatives which are stable and have prolongedfibrinolytic activities.
The polyalkylene glycol activated with the coupling agent can be obtained in the following manner. A terminal-substituted polyalkylene glycol-4,6-dihalogeno-1,3,5-triazine is obtained by reacting its corresponding mono-substituted polyalkylene glycol with a cyanuric halide in an anhydrous solvent and in the presenc,e of a base.
A polyalkylene glycol-4-halogeno-6-hydroxy-1,3,5-triazine is formed by first reacting its corresponding polyalkyler,e glycol with cyanuric halide and then treating the resultant reaction product with water. A polyalkylene glycol-acetoazide is formed by reacting an-ani~n of its corre-sponding polyalkylene glycol with ethyl chloroacetate, followed by treating the reaction product with hydrazine, and finally activating the resulting hydrazide with nitrous acid. A polyalkylene glycol-p-diazoniumbenzyl ether or a polyalkylene glycol-3-(p-diazoniumphenoxy)-2-hydroxypropyl ether is obtained by reacting its corresponding polyalky-lene glycol with p-nitrobenzyl chloride or p-nitrophenyl glyceryl ether and, after reducing the nitro group into an amino group, diazotizing the resultant product with nitrous,acid.
After completion of the reaction of the . . .

lZ177~

polyalkylene glycol activated with the coupling agent and the plasminogen activator, the isolation and purification of the reaction product may be effected in a biochemical manner known per se in the art, for example, by using singly or in combination gel filtration, dialysis, ion-exchange chromatography, affinity chromatography and the - like. Preferably, the conjugate is kept as a solution containing a buffer or a physiological salt, by freezins the solution below -20C, or by lyophilizing the solution.
The optimum pH levels for the novel urokinase derivatives prepared above vary depending on the molecular weights and types of polyalkylene glycols emplo~ed, the types of coupling agents employed, the modification degrees of urokinase, the modification conditions and the measure-ment conditions such as substrates used. ~ihen measured in terms of the amidase activity using a synthetic substrate S-2444, the optimum pH level is within the range of about 8 to 9. In the case of nigh molecular weight urokinase modified by reaction with monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine having an average molecular weight of 5,000 at pH 7.0 and at a temperature of 0C for 3 hou-s, while using the latter reactant at a concentration of 4.0 mM (hereinafter abbreviated as PEG-DC~-UK), the optimum pH as measured in terms of the amidase activity is 8.2 as shown in FIG. 1. The optimum pH of low molecular weight urokinase modified under the same conditions (hereinarter abbreviated as PEG-DCT-L-UK) is, as measured in terms of `

lZ;~ 7~

the amidase activity, 8~2 as shown in FIG. 2.
The novel urokinase derivatives according to the invention have extremely enhanced stability compared with unmodified urokinase. For instance, FIG. 3 shows the results of the residual urokinase activity obtained relative to the passage of time when urokinase was dilu~ed with a Ringer solution or a physiological solution to sucn a concentration as used by dripping administration for clinical application and then allowed to stand at room temperature. As a result, it has been found that PEG-DCT-UK dissolved at a concentration of 105.3 iu/ml in the Rinser solution maintains an activity of 73.4~ at the initiation of the experiment even after 6 hours. In addition, PEG-DCT~UK retains an initial activity of 79.4~ in the physio-logical saline. On the other hand, unmodified urokinase dissolved at a concentration of 76.I iu~l in a Ringer solution keeps only 33.3% in its initial activity even after 2 hours. In a physiological saline, the residual activity is as low as 26.3%.
FIG. 4 illustrates the results obtained by comparing the stability of PEG-DCT-L-UK and unmodified low moiecula~
weight urokinase, both in the course of freezing, followed by thawing operation. The results confirm th~t th2 supe-b stability of PEG-DCT-L-UK is attained even under such conditions.
Furthermore, the novel urokinase derivatives according to the invention have prolonged urokinase ac~i-lZ:~`7~3 vities in blood as compared with unmodified urokinase.For example, the effectiveness of urokinase was investigated by administering 8,000 iu/kg of each of PEG-DCT-UK and unmodified urokinase to healthy rabbits by intravenous injection, sampling blood periodically before the dosage and after the lapse of 4 hours, and measurins the amounts of the plasmin inhibitor and plasminogen present in the citrated plasma. The results are shown in FIGS. 5, 6 and 7. From these results, it has been con-firmed that PEG-DCT-UK alters to a substantial extent biochemical parameters such as a plasmin inhibitor and plasminogen and maintains the parameters at lowered values for an extended period of time as compared with unmodified urokinase.
Therefore, the ~;~ urokinase derivatives accorâing to the invention are excellent plasminog~n activators which, while retaining plasminogen-activating potency as unmodified urokinase does, have surmounted the shortcomings of urokinase such as poor stability and short half-life in blood.
The urokinase derivatives according to the invention may be used as pharmaceutical products for the treatment of a variety of diseases stenlmed from hyper-coagulability of blood such as arterial and venous throm-boses, coronary artery clotting, myocardial infarction, intracerebral infarction, pulmonary embolism, nephritis and the like. These urokinase derivatives can be suitAbly . . .

L77~

administered by intravenous injection or dripping or by an oral route. The intravenous injection is particulary preferred. As dosable forms, preferably usable is a lyophilized form. When supplied eithex in a neat state or with a physiological salt such as a salt of a Ringer solution or sodium chloride, the lyophilized products may be used for clinical application by dissolution with sterile distilled water or further dilution with sterile distilled water in order to adjust their osmotic pressure, prior to the actual use. Since the novel urokinase de-i-vatives according to the invention are stable,they do not require a stabilizer such as albumin. However, the addi-tion of such a stabilizer does not cause any problem or inconvenience. An excipient may also be added in subjec~-ing the novel urokinase derivatives to lyophilization.
The above description genera-lly ~escribes the present invention. A more complete understanding can be obtained by reference to the following examples which are provided for purposes of illustration only and are not intended to be limiting. The modification procedure may be carried out in any one of the preparation steps of plasminogen activators.
Example 1 Monomethyl ether polyethylene glycol-4,6-dicnloro-1,3,5-triazine Polyethylene glycol monomethyl ether having an average molecular weight of 5,000 (25.0 g; O.OS mole) was ~ 77:~

dissolved with warming in dry benzene (200 ml). After cooling the resultant solution to room temperature, anhydrous sodium carbonate (5.0 g) and cyanuric chloride (2.75 g; 0.015 mole) were added, and the mixture was stirred overnight at room temperature. After com?letion of the reaction, the reaction mixture was filtered, and petroleum ether t600 m1) was added to the filtrate. The resulting precipitate was collected by suction filtration and washed with a small amount of petroleum ether. The precipitate was purified by being reprecipitated three times from dry benzene and petroleum ether to remove excess cyanuric chloride, thereby stoichiometrically obtaining monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine as white powder. The product after hydrolyzed showed qualitative reaction characteristics of chlorine ions (AgNO3).
- Similarly, polyethylene glycol monomethyl ethers having average molecular weights of 550, 700, 2,000 and 20,000, respectively, were each reacted with cyanuric chloride to stoichiometrically obtain monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazines having the corresponding average molecular weights.
Example 2 Monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazines (average molecular weights o~ their polyethylene glycol moieties: 10,000 and 15,000) Polyethylene glycol monomethyl ether having an ~77~

average molecular weight of 10,000 (25.0 g; 0.025 mole) was taken up with warming in dry benzene (200 ml~. After cooling the resultant solution, anhydrous sodium carbonate (2.5 g) and cyanuric chloride (1.38g; 0.0075 mole) were added. The resultant mixture was stirred overnight at 33C. After completion of the reaction, undissolved matter was removed by filtration. The filtrate was added with n-hexane (about 600 ml) to induce reprecipitation. Dry benzene (100 ml) was added to the precipitate which was then warmed and dissolved. n-~exane (500 rll) was then added to induce reprecipitation. This procedure was repeated three times.- The thus obtained precipitate was dried overnight at 50C in vacuo, thereby stoichiometri-cally obtaining monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety: lO,~OOO)~as while powder.
Polyethylene glycol monomethyl ether having an average molecular weight of 15,000 was similarly reacted with cyanuric chloride to stoichiometrically obtain monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety:
15,000).
Example 3 Monostearyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine Polyethylene glycol monostearyl ether having an average molecular weight of 3,200 (lo .0 g; 0.005 mole) was 77~

dissolved in dry benzene (200 ml), followed by the addi-tion of anhydrous sodium carbonate (5.0 g) and cyanuri~
chloride (2.75 g; 0.015 mole) under stirring. The resul-tant mixture was stirred overnight at room temperature.
After completion of the reaction, undissolved matter was removed by filtration. The filtrate was added with n-hexane (about 600 ml) to induce reprecipitation. Dry benzene (100 ml) was added to the precipitate to dissolve the latter. n-Hexane (500 ml) was added to induce reprecipitation. This procedure was repeated three times, thereby stoichiometrically obtaining monostearly ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety:

3,200) as white powder.
Example 4 Polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazine Polyethylene glycol having an average molecular weight of 6,000 (33.6 g) was dissolved with warming in dry benzene (150 ml). After cooling the resultant solution, anhydrous sodium carbonate (1.6 g) and cyanuric chloride (0.74 g) were added, and the resultant mixture was stirred overnight at room temperature. Thereafter, water (1.0 ml) was added, followed by stirring the mi~ture at room tem-perature for 6 hours and then at 40C for an overnight period. Undissolved matter was removed by centrifuge (2,000 ppm; 10 minutes), and the supernatant was subjected 7 7~

to condensation under reduced pressure. The residue was taken up with warming in dry benzene, and the soIvent w~s then evaporated. This procedure was further repeated twice. The residue was dried under reduced pressure, thereby obtaining polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety: 6,000).
Polyethylene glycols having different average molecular weights of 1,000 and 4,000 were each reacted in the same manner as above with cyanuric chloride and water, resulting in the stoichiometric formation of polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazines having their correspondins average molecular weishts.
Example 5 Stearoylpolyethylene glycol-4,6-dichloro-1,3,5-triazine - Polyethylene glycol monostearate having an average molecular weight of 2,700 (6.765 g) was dissolved in anhydrous benzene (100 ml), followed by the addition of anhydrous sodium carbonate (2.5 g). While stirring the resultant mixture, cyanuric chloride (1.38 g~ was further added. The resulting mixture was stirred overnight at room temperature and then filtered. The filtrate was concentrated under reduced pressure, and the residue was dried under reduced pressure, thereby stoichiometrically obtaining stearoylpolyethylene glycol-4,6-dlchloro-1,3,5-~riazine (the average molecular weight of the polyethylere 77~

glycol moiety: 2,700) as a white waxy substance.
Example 6 Polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine Polypropylene glycol having an average molecular weight of 1,000 (4.0 g) was taken up in anhydrous benzene (50 ml), followed ~y the addition of anhydrous sodium carbonate ~1.27 g). Cyanuxic chloride (0.552 g) was further added with stirring. After stirring the resultant mixture overnight at room temperature, water (1 ml) was added. The mixture was stirred at room temperature for further 6 hours. The reaction mixture was filtered, and the filtrate was concentrated under reduced pressure.
The residue was added with anhydrous benzene and anhydrous sodium sulfate. The mixture was thereafter stirred at room temperature for 10 minutes. ~ter~~iltration of the mixture, the solvent was evaporated from the filtrate.
The residue was dried under reduced pressure, thereby stoichiometrically obtaining polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (the average molecular weight of the polypropylene glycol moiety: 1,000) as colorless viscous oil.
Following the same procedure as described above, there were obtained a polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (the average molecular weight of the polypropylene glycol moiety: 4,000) and anothe~ poly-propylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (the ~z~

average molecular weight of the polypropylene glycol moiety: 10,000).
The polypropylene glycols employed in the above examples were of a straight-chain type, namely, those respresented by the formula HO[CH(C~3)CH2O]nH.
Example 7 Polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine Following the same procedure as in Example 1, poly-propylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (the average molecular weight of the polypropylene glycol moiety: 4,000) was stoichiometrically obtained as a white substance from polypropylene glycol having an average molecular weihgt of A,000 ~8.0 g), anhydrous benzene t80 ml), anhydrous sodium carbonate (0.636 g) and cyanuric chloride (0.368 g).
. The polypropylene glycol used in this example was of a branched-chain type, namely, those re~resented by the formula CH3CH2C{CH2O[CH2CH(CH3)O]nH}3-Furthermore, using polypropylene glycol representedby the formula, CH2O[CH2cH(cH3)O]nH
CHO[CH2CH(CH3)O]nH
CH2O[CH2CH(cH3)O~nH
and naving an average molecular weight of 3,000, there was obtained polypropylene glycol-4-chloro-6-hydroxy-1,3 "-triazine (the average molecular weight of the polypropylene ~2~7~

glycol moiety: 3,000).
Example 8 ~lonomethyl ether polyethylene glycol methoxycarbohydrazide Polyethylene glycol monomethyl ether having an average molecular weight of 5,000 (13.3 g; 0.0027 mole) was taken up in anhydrous tetrahydrofuran (400 ml) under nitrogen. A small amount of diphenylacetic acid was added as an indicator, and n-butyl lithium was dropped under ice-cooling until the reaction solution turned to pale yellow.
Thereafter, ethyl chloroacetate (5 ml; 0.0~7 mole) was dropped at room temperature, and the resultant solution was stirred overnight and then refluxed for one hour.
After completion of the reaction, the solvent was driven off under reduced pressure,and the residue was taken up in aqueous acetone(200 ml). The resul~nt solution was treated with charcoal. After filtration and subsequent concentration of the filtrate, benzene was added. After removing water by azeotropic distillation, the residue was reprecipitated from benzene and n-hexane to obtain yellowish powder. The thus obtained powder was dissolved in methanol (150 ml) and tnen added with hydrazine hydrate (15 ml?. The mixture was heated overnight under reflux conditions. After driving off the solvent under reduced pressure, water (150 ml) was added, and excess hydrazine was removed by azeotropic distillation. Water, still remaining in the residue, was removed azeotropically 7~

together with benzene. The residue was again dissolved in benzene and dried with anhydrous sodium sulfa~e.
Thereafter, the solvent was driven off, and the residue was dissolved in warm benzene. The benzene solution was treated with charcoal and then concentrated. The reaction product was reprecipitated twice from benzene and n-hexane, treated with charcoal again and with silica gel and then reprecipitated from benzene and n-hexane, thereby obtain-ing monomethyl ether polyethylene glycol methoxycarbo-hydrazide as white powder. In the same manner, a variety of hydrazides were obtained using polyethylene glycol monomethyl ethers havlng average molecular weights of 550, 700, 2,000, 10,000, 15,000 and 20,000, respectively, as starting materials.
Example 9 Monomethyl ether polyethylene glYCo1-4, 6-dichloro-1,3,5-triazine-modified urokinase (PEG-DCT-UK) A 0.1 M phosphate buffer of pH 7.0 (2.0 ml) W2S
added under ice-cooling to 0. 61 ml of a urokinase solution (molecular weight: 54,000; 66,300 iu/ml). Thereafter, monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety: 5,000) was further added in such an amount as to bring the concentration of the triazine to 4 mM.
The mixture was reacted under ice-cooling for 3 hours.
After completion of the reaction; the reaction solution i;~177~

was transferred into a dialyzing tube, and excess triazine derivative was removed by dialysis. The dialysis was carrled out under ice-cooling for 3 hours against a 0.1 M
phosphate buffer (pH 7.2) and then for further one hour against physiological saline. The content in the tube ~7as then filled up to 4 ml and stored in a frozen state at -80C. The urokinase activity of the thus obtained mono-methyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase (PEG-DCT-U~) was determined to be 4,550 iu/ml by the fibrin plate method. Therefore, the total activity was 13,200 iu. Since the activity of the starting urokinase was 40,443 iu, an activity drop of 55% was recognized.
The above procedure was repeated except that the monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight~~of the polyethylene glycol moiety: 5,000) was replaced by monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazines (the average molecular weights of the polyethylene glycol moieties: 550, 700 and 2,000, respectively), thereby obtaining modified urokinases having urokinase activities ....
Of 9,800, 10JOOO and 6,500 iu/ml as determined by the fi~rin plate method.
Example 10 Monomethyl ether polyethylene ~lycol-4,6-dichloro-1,3,5-triazine-modified low molecular ~ .
weight urokinase (PEG-DCT-L-U~) A 0.1 M phosphate buffer of pH 7.0 (4.7 ml) was ~ .
.

1~77~

added under ice-cooling to 0.2 ml of a low molecular weight urokinase solution (molecular weight: 33,000; 567,828 iu/
ml). Thereafter, monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety: 5,000) was further added in such an amount as to bring the concentration of the triazine derivative to 4 mM. The mixture was reacted under ice-cooling for 3 hours. After completion of the reaction, the reaction solution was transferred into a dialyzing tube, and excess triazine derivative was removec by dialysis. The dialysis was carried out under ice-cooling for 3 hours against a 0.1 M phosphate buffer (pH
7.2) and then for further one hoùr against physiological saline. After the dialysis, the content was filled u? to 8 ml and stored in a frozen state at -80C. The urokinase activity of the thus obtained monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified low molecular weight urokinase (PEG-DCT-L-UK) was determined to be 10,300 iu/ml by the fibrin plate method. Thus, the total activity was 82,400 iu. Since the activity of the starting urokinase was 113,566 iu, an activity drop of 27.4~ was recognized.
Example 11 Monomethyl ether polyeth~lene ~lycol-4,6-dichloro-1,3,5-triazine-modified urokinase By repeating the procedure of Example 9 except t~at the concentration of each monomethyl ether polyethvlene .

~77~

glycol-4,6-dichloro-1,3,5-triazi~e was changed to 0.4 mM, there were obtained, with modification degrees different from that achieved in Example 9, monomethyl ether poly-ethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinases (the average molecular weights of the poly-ethylene glycol moieties: 550, 700, 2,000 and 5,000, respectively). Their urokinase activities were 8,670, 8,900, 6,770 and 8,670 iu/ml as measured by the fibrin plate method.
Example 12 Monomethyl ether polyethylene carbomethylazide-modified urokinase (1) To 200 mg of the monomethyl ether polyethylene glycol methoxycarbohydrazide prepared in Example 8 (the average molecular weight of the polyethylene glycol moiety:
5,000) were added under ice-coolin~ 1 ~_hydrochloric acid (2 ml) and then a 0.008 N aqueous solution of sodium nitrite (1 ml). The resultant mixture was stirred at room temperature for 20 minutes, and a 1 N aqueous solution of sodium hydroxide (2 ml) was further added to neutralize the mixture. The resultant solution of monomethyl ether polyethylene glycol carboxymethylazide was stored at 0C.
(2) A 0.1 M phosphate buffer of pH 8.0 (3.656 ml) was added to 1 ml of a urokinase solution (molecular weight: 33,000; 45,600 iu/ml). Thereafter, 0.435 ml of the monomethyl ether polyethylene glycol carbomethylazide solution prepared in the procedure (1) above was added .. .. .

~Z~7~3 -under mild stirring. The mixture was reacted at room temperature for 2 hours. The resulting reaction solution was then transferred into a dialyzing tube and dialyzed under ice-cooling for 4 hours against a 0.1 M phosphate buffer (pH 7.2). The content was filled up to 8 ml. The resultant solution was stored in a frozen state at -80C.
The urokinase activity of the thus obtained monomethyl ether polyethylene glycol carbomethylazide-modified uroki-nase was determined to be 7,300 iu/ml by the fibrin plate method. Compared with unmodified urokinase, an activity increase of 28% was recognized.
E~ample 13 Monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase To 0.61 ml of a urokinase solution (molecular weight: 54,000; 101,167 iu/ml) were added under ice-cool`ng -a 0.1 M phosphate buffer of pH 7.0 (2.0 ml) and then under mild stirring monomethyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety: 10,000). The triazine derivative was added in such an amount as to bring the concentration to 0.1 mM. The mixture was reacted for 3 hours under ice-cooling. After completion of the reaction, the reaction solution was transferred into a dialyzing tube, and excess triazine derivative was removed by dialy-sis. The dialysis was carried out under ice-cooling for 3 hours against a 0.1 M phosphoric acid bùffer added wi-n . . .

~Z~'77~3 .

0.035 v/v % of ethyl amine (pH8.0) and then for further 2 nours against a O.lM phosphate buffer (pH 7.2). The con-tent was adde~ with 3 w/v % bovine serum albumin (0.1 ml) and then filled up to 4.0ml with a 0.1 M phosphate buffer (pH 7.0~. The resultant solution was stored in a frozen state at -80C. The activity of the thus obtained mono-methyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase was determined to be 1~,0~8 iu/ml by the fibrin plate method. Therefore, the total activity was 40,11~ iu, and the activity drop was 35%. Similar modi-fied urokinases were obtained by changing the concentration of the monometnyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine (the average molecular weight of the poly-ethylene glycol moiety: lO,OOO)to 0.4and 1.0 ~1, respectively.
Their urokinase activities were 8,7~4 and 6,634 iu/ml.
Furthermore, monomethyl ethe~ por~ethylene glycol-

4,6-dichloro~1,3,5-triazine (the average molecular weight of the polyethylene glycol moiety: 15,000~ and urokinase were similarly reacted, thereby obtaining monomethyl ether polyethylene glycol-4,6-dLchloro-1,3,5-triazine-modified urokinàse (the average molecular weight of the polyethylene glycol moiety: 15,000). The activities of the modified urokinases obtained by changing the concentration of the triazine derivative to 0.1, 0.4 and 1.0 ~I, respectively, were 11,388, ~,580 and 7,176 iu/ml.

'771~

Example 14 Monostearyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase To 0.2 ml of a urokinase solution (molecular weight:
54,000; 101,167 iu/ml) were added under ice-cooling a 0.1 M
phosphate buffer of pH 7.0 (0.66 ml) and then monostearyl ether polyethylene glycol-4,6-dichloro-1,3,5-txiazine (the average molecular weight of the polyethylene glycol moiety:
3,200). The triazine derivative was added in such an amount as to bring the concentration to 2 mM. The mixture was reacted under ice-cooling for 3 hours. After comple-tion of the reaction,~the reaction solution was transferred into a dializing tube and subjected to dialysis to remove excess triazine derivative. The dialysis was carried ou~
under ice-cooling for 4 hours against a 0.1 M phosphate buffer of pH 7.2. The content was ~dde~ with a 3.0% aque-ous solution of bovine serum albumin (0.1 ml) and then filled up to 4.0 m~ with a 0.1 M phosphate buffer of pH
7Ø The resultant solution was stored in a frozen state at -80C. The urokinase activity of the thus prepared monostearyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase (the average molecular weight of the polyethylene glycol moiety: 3,200) was determined to be 2,826 iu/ml by the fibrin plate method. Since the total activity was 11,304 iu, the activity of the thus modified urokinas~e was 55.9% of that of t~e starting urokinase. The activities of tle modified urokinases 7 ~

obtained by changing tne concentration of the monostearyl ether polyethylene glycol-4,6-dichloro-1,3,5-triazine to 4, 6 and 8 mM, respectively, were 2,351, 1,430 and 1,059 iu/ml.
Example 15 .
Polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinase To 1 ml of a urokinase solution (molecular weight:
54,000; 45,600 iu/ml) were added under ice-cooling a 0.05 M
phosphate buffer of pH 9.2 (4.0 ml) and then polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (2.34 mg). The average molecular weight of the polyethylene glycol moiety of the triazine derivative was 6,000. The mixture was reacted under ice-cooling for 3 hours. After completion of the reaction, the reaction solution was transferred into a dialyzin~ tube, and excess txiaz~ne derivative was removed by dialysis. The dialysis was carried out under ice-cooling for one hour against a 0.05 M phosphate buffer of pH 9.2 and then for further 3 hours against a 0.1 M
phosphate buffer of pH 7.2. The content was filled up to 8.0 ml with a 0.1 M phosphate buffer of pH 7.2. The resultant solution was stored in a frozen state at -80C.
The urokinase activity of the thus obtained polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinase (the average molecular weight of the polyethylene glycol moiety: 6,000) was determined to be 460 iu/ml by the fibrin plate method.

~'77~'~

Similarly, there were obtained other polyethylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinases (the average molecular weights of the polyethylene glycol moieties: 4,000 and 1,000, respectively).
Example 16 Stearoyl polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase .
- To 0.5 ml of a uro~inase solution (molecular weisht:
54,000; 101,167 iu/ml) were added under ice-cooling a 0.1 phosphate buffer of pH 7.0 (1.5 ml) and then 0~05 ml of a dioxane solution of stearoyl polyethylene glycol-4,6-dichloro-1,3,5-triazine (270 mg/ml) (the average molecular weight of the polyethylene glycol moiety: 2,700). The mixture was reacted under ice-cooling for 3 hours. ~fter completion of the reaction, the reactio~ solution was transferred into a dia~yzing ~ube~and-~ùbjected to dialysis to remove excess triazine derivative. The dialysis was carried out under ice-cooling for 4 hours against a 0.1 M
phosphate buffer of pH 7.2. The content was filled up to ~ ml with a pnosphate buffer. The resultant solution was stored in a frozen state at -80C.
The thus obtained stearoyl polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinase was found, a result of the urokinase activity measurement by the fibrin plate method, to have an activity of la6% of that of the starting urokinase.

~Z177~

Other stearoyl polyethylene glycol-4,6-dichloro-1,3,5-triazine-modified urokinases having different modi-fication degrees were obtained by reaction with urokinase using a dioxane solution of stéaroyl polyethylene glycol-4,6-dichloro-1,3,5-triazine (270 mg/ml) in amounts of 0.1 and 0.2 ml, respectively. The average molecular weight of the polyethylene glycol moiety of the triazine derivative was 2,700. The thus prepared modified urokinases had activities of 106 and 101% of that of the starting urokinase.
Example 17 Polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinase To 0. 5 ml of a urokinase solution (molecular weight:
54,000; 101,167 iu/ml) were added under ice-cooling a Ool M phosphate buffer of pH 7.0 (I.5 m~) and then 0.05 ml of a dioxane solution of the polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (1-00 mg/ml) obtained in Example 6. The average molecular weight of the polypro-pylene glycol moiety of the triazine derivative was 1,000.
The mixture was reacted under ice-cooling for 3 hours.
Ater completion of the reaction, the reaction solution was transferred into a dialyzingtube and subjected to dialysis to remove excess triazine derivative. The dialy-sis was carried out under ice-cooling for 4 hours against a 0.1 M phosphate buffer of pH 7.2. The content was filled up to 5 ml with a phosphate buffer and then stored lZ~77~

in a frozen state at -80C. The thus obtained polypro-pylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinase was found to retain an activity of 96.4~ of that of the starting urokinase by the fibrin plate method.
Other polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinases having different modi-fication degrees were similarly obtained by using a dioxane solution of polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine (100 mg/ml) in amounts of 0.1 and 0.2 ml, respectively. The average molecular weight of the polypropylene glycol moiety of the triazine deriva-tive was 1,000. The modified urokinases showed activities of 99.3 and 106.8% in comparison with that of the starting unmodified urokinase.
ExamPle 18 ;
Polypropylene glycol-4-cnloro-6-~droxy-- . 1,3,5-triazine-modified urokinase Polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine-modified urokinases having different modification degrees were obtained exactly in the same manner as in Example 3 with use of a dioxane solution containing 400 mg/
ml of the polypropylene glycol-4-chloro-6-hydroxy-1,3,5-triazine obtained in Example 7 in amounts of 0.05, 0.1 and 0.2 ml, respectively. The average molecular weight of the polyethylene glycol moiety of the triazine derivative was 4,000. Their urokinase activities were found to be Rg,7, 100.0 and 98.9% of the that of the 7 ~

starting urokinase by the fibrin pla~e method.
Example 19 Optimum pH for modified urokinase as measured in terms of amidase activity PEG-DCT-UK obtained in Example 9 and unmodified urokinase were each diluted with physiological saline containing 0.1% human albumin to obtain two solutions of 167 iu/ml. Each of the solutions were divided into por-tions of 0.3 ml which were added with 50 mM tris-HCl buffers of different pH levels (each 1.0 ml) and tnen with a synthetic substrate S-2444 (pyrGlu-Gly-Arg-p-nitro anilide made by Kabi Corporation). The resultant mi:~ture;
were incubated at 37C for 10 minutes. The reactions were then stopped by the addition of 30% acetic acid. Their absorbance was measured at 405 nm. As a result, the optimum pH levels for PEG-DCT-UK and unmodified urokinase were 8.2 and 8.5, respectively, as measured in terms of the amidase activity when the synthetic substrate S-2444 was employed. The results are shown in FIG. 1.
The optimum pH levels for PEG-DCT-L-UK obtained in Example 4 and unmodified low molecular weight urokinase were determined similarly in terms of the amidase activity and we~e found to be 8.2 and 8.4, respectively. The results are shown in FIG. 2.

~2~7'71~

Example 20 Stabilit of modified urokinase at room Y
temperature PEG-DCT-UK obtained in Example 9 was diluted with either a Ringer solution or physiological saline to a concentration of 105.3 iu/ml. For the sake of comparison, unmodified urokinase was also diluted with either a Ringer solution or physiological saline to a concentration of 76.1 iu/ml. 3.5 ml of each of the thus diluted solutions was allowed to stand at room temperature (27~C) for 6 hours. The residual urokinase activity was periodically measured by the fibrin pla~e method until such time tha~
6 hours elapsed. The extent of activity loss of the modified urokinase according to the invention was by far smaller than that of unmodified urokinase. The results are shown in FIG. 3.
-Example 21 -Stability of modified urokinase against freezing and thawing processing PEG-DCT-L-UK obtained in Example 10 was diluted with physiological saline to a concentration of 2Q0 iu/ml.
For the sake of comparison, unmodified low molecular weight urokinase was diluted with physiological saline to a concentration of 200 iu/ml. Each of the thus diluted solutions was frozen to -80C and subsequently thawed at room temperature. The freezing and thawing operations were repeated 0, 2, 4 and 6 times. The residual urokinase 1;Z~'77;~

activity was measured using a synthetic substrate S-244~
(made by Kabi Corporation). The extent of activity loss of the modified urokinase according to the invention was extremely low as compared with that of unmodified low molecular weight urokinase. The results are shown in FIG. 4.
Example 22 Eight healthy male rabbits (Japanese white) each having a body weight of 2.6 - 3.1 kg were divided into two groups. One group was dosed with 8,000 iu/kg of PEG-DCT-UK and the other with 8,000 iu/kg of unmodified urokinase, both through the ear veins. Blood was collec-ted from the ear veins before the dosage and at the time of 1, 2 and 4 hours after the dosage. Blood samples W21-e added with a sodium citrate solution to separate plasma.
After wash-out of the modified and unmo~ified urokinases for 18 days, the two groups were crossed with each other, and similar experiments were carried out to take plasma samples. A part of each of the plasma samples thus obtained was separated and subjected to affinity column chromatography using Lysine Sepharose 4B, thereby obtain-ing a plasmin inhibitor fraction (fraction Fl) and a plasmin and plasminogen fraction (fraction F3). Chan~es in the amount of the plasmin inhibitor in the fraction Flalong the passage of time were determined by adding plasmin to the fraction Fl and measuring the amount of residual plasmin which was not adversely affected b~ the inhibitor, with a svnthe'ic ,tl 7-.lB

substrate S-2251 (made by Kabi Corporation). The plasmin inhibitor from the PEG-DCT-UK-administered group showed its apparent decrease and slower recovexy. In the unmodi-fied urokinase-administered group, the decrease in the plasmin inhibitor was rather less, and its time course changes were not made clear. The results are shown in FIG. 5.
On the other hand, the amount of plasminogen in the fraction F3 was determined by adding urokinase to the fraction F3 to produce plasmin and measuring the amount of the thus produced plasmin by a synthetic substrate S 2251. In the PEG-DCT-UK-administered group, plasminogen decreased to about 50% after 2 hours and thereafter re-covered gradually. However, in the unmodified urokinase-administered group, plasminogen decreased up to about 65~
after one hour but did not proceed ~to decrease any fur.her.
-There was observed a tendency of plasminogen being recov~
ered even 2 hours after the dosage. The results are illustrated in FIG. 6.
Furthermore, a part of each plasma was sampled ana treated with an acid to pH 5.2 so as to deactivate the inhibitor therein. It was then neutrali~ed, and the a~ount of plasminogen present in the acid treated plasma was measured by adding urokinase to the acid treated plasma to produce plasmin and measuring the amount of the thus produced plasmin using a synthetlc SuDstrate S-2251.
In the PEG-DCT-UK-administered gro~lp, plasminogen decreased to about 70% after 2 hours but showed its gradual recovery thereafter. In the unmodified urokinase-administered group, however, the amount of plasminogen was decreased to about 84% after one hour but was not decreased any further.
Thereafter, the amount of plasminogen was increased gra-dually. The results are also shown in FIG. 6. Incidental-ly, no plasmin activity was detected from the fraction F3.
Example 23 Lyophilized_product suitable for the preparation of injectable formulations The PEG-DCT-UK solution obtained in Example 9 was concentrated by a membrane filter (Amicon ~el: trademark~
and then added with physiological saline and a O.OS M
phosphate buffer. The resultant mixture was filtered aseptically using a membrane filter. The filtrate was poured in portions into sterilized vials and then lyophi-lized. The fibrinolytic activity of the lyophilized product of PEG-DCT-UK suitable for the preparation of injectable formulations was found to be 57,000 iu/vial by the fibrin plate method.
Example 24 L o hilized product suitable for tne preparation . Y P
of injectable formulations The PEG-DCT-UK solution obtained in Example 9 was concentrated by ~ membrane filter (~micon Gel: trademark) and then added with physiological saline. The resultant mixture was filtered aseptically using a membrane filter.

; 7 ~r~

- The filtrate was poured in portions into sterilized vials and then lyophilized. The fibrinolytic activity of ~he thus obtained lyophilized product of PEG-DCT-UK suitable for the preparation of injectable formulations was found to be 67,000 iu/vial by the fibrin plate method.
In the following experiments Urokinase (MW 54,000) was obtained from Japan Chemical Research Co., Ltd. PEG
#5,000 (MN = 4,700, MW/MN = l.08) was obtained from Nishio Industry Co., Ltd. Cyanuric chloride was obtained from Kanto Chemical Co., Inc. Fibrinogen and thrombin fcr fibrin plate was obtained from Sigma Chemical Co. and Mochida Pharmaceutical Co., Ltd.
respectively. S-2251 and S-2444 were obtained from Kabi Diagnostica. TNBS (trinitrobenzene sulfonic acid sodium salt) was obtained from Wako Pure Chemical Industry, Ltd. Lysine-Sepharose 4B and Sephadex G-200 were obtained from Pharmacia Fine Chemicals. Nal25I
was obtained from Amersham. GGA-MCA (L-glutaryl-glycyl-L-argir.ine-4-methylcoumarin-7-amide) was obtained from Protein Research Founda~ion.
Urokinase dissolved in O.lM phosphate buffer (P~), pH 7.0, was reacted with PEG #5,000 activated with cyanuric chloride by the method of A~uchowski et al at * trade mark.

3~2~i7~'~

4 for 4 hr. Concentration of PEG was adjusted to 1.7 mM for L-Md, 2 mM for M-Md and 5 mM for H-Md which represent low, medium and high degrees of modification, respectively. Dialysis against 0.05 m PB, pH 5.0, containing 0.9% ~aCl followed by concentration and gel filtration using Sephadex G-200 afforded modified UR's. After adjusting pH to 7.0, UK'5 were lyophilized in small portions for experimental convenience. Lysine residues in M-Md were determined with TNBS by the method of Habeeb. UK's were assayed either by the standard fibrin plate method or by the synthetic substrate method. Plasma was fractioned into Fl, F2 and F3 with Lysine-Sepharose 4B by the method of Igarashi et al. Native UK and M-Md were labelled with 12SI by the procedure of Hunter and Greenwood.
Immunological properties of native UK and M-Md were compared according to the procedures of Arai et al.

Example 2S - SUBSTRATE SPECIFICITY OF PEG-MODIFIED UK's Activation of PEG #5,000 with cyanuric chloride followed by reaction with native UK caused a covalent attachment of PEG's on the lysine residues of UK
protein. Three types of modified UK's, L-Md, M-Md and H-Md, were prepared by changing PEG concentration during the reaction. Their enzymatic activitie~ depend on the assay method which sugges~s that PEG chains ~Z~ 7~

presen~ steric hindrance to bulky substrat~s such as plasminogen in fibrin plate while not to the low molecular weight ones such as GGA-MCA.
Attention was specifically focused on M-Md and the pH Optimum of M-Md was found to be 8.16 and that of native UK to be 8.51 using S-2444 as a substrate in 0.05M Tris HCI buffer at 37C. The molecular weight (MW) of M-Md was calculated to be about 120,000 daltons from the result of lysine residue determination with TNBS. As PEG is a kind of neutral detergent the apparent MW obtained from SDS-PAGE of gel filtration does not reflect the true MW. FIGURE 8 illustrates the substrte specificity of PEG-Modified UK's.

Example 26 - EFFECT OF UK INHIBITORS ON THE ACTIVITY OF
~ATIVE UK AND L-MD.
It is known that in human plasma, protease inhibitors play an important role in retaining blood fluidity. Physiological activity of native UK when administered intraveneously is rapidly retarded not only by the interaction with these inhibitors but also by the fragmentation. The effect of UK inhibitors on the UK activity was examined using Fl, inhibitor fraction from human plasma, and placental UK
inhibitor. In the first experiment, U~s were incubated wieh the former at 37 then residual UK was assayed with S-2444. In the next experiment, placental UK

-:^

771~

inhibitor of different dilution was added to UKs then residual UK was assayed with a combination of plasminogen and S-2251.
In both cases, L-Md retained its activity better than native UK did. Neutral and chemically inert PEG
chains seem to protect UK from the action of UK
inhibitors. FIGURE 9 shows the stability of modified Urokinase to UK-Inhibitors in Human Plasma. ~IGURE 9b shows the effect of placental UK-Inhibitor on the Urokinase activity.

Example 27 - EXTENSION OF CIRCULATING LIVES OF UKs IN RABBITS
The short half-life (~/2) of native UK in the circulatory system often becomes a limitation of UK
therapy. As the ~20f native UK does not exceed several minutes, prolonged drop infusion is necessary to maintain the high plasmin level. Half lives of native and modified UKs were measured in rabbits by an injection of one of them followed by the determination of UK activity in the plasma drawn, periodically, from an ear vein.
As a re~ult, the half-life of native UK in the first phase was calculated as 4-5 minutes, L-Md as 30-40 minutes, M-Md as 80-100 minutes and H-Md as 110-150 minutes. ~20f modified UKs were extended about 10 to 30 times that of the native UK depending upon the 77~

respective degree of modification. In the case of M-Md or H-Md, a one compartment model is approximate enough to describe the pharmacokinetics, which suggests tha~
distribution to a second compartment is suppressed.
FIGURE 10 shows the UK activity of L-Md, M-Md and H-Md in Plasma after UK injection.

Example 28 - BEHAVIOR O~ 125I-NA'rIVE UK AND 125I-M-MD
IN RATS
Native UK and M-Md were labeled with 125I. Then, their ~2in the circulatory system and their distribution among organs were investigated in rats.
Extension of ~2f UK by PEG modification was also observed and was more prominent in these experiments.
Rapid decay of radioactivity in blood of rats injected with 125I-native UK was observed. On the contrary, radioactivity was retained well with 125I-M-Md. In the first phase, ~20f native UK was 1.1 min. while that of M-Md was 89.9 min.
Native UK administered intravenously accumulates in the liver and kidneys and other organs resulting in a short ~/2~ With M-Md, accumulation in these organs was suppressed probably due to less interaction of the M-Md molecule~ with organ cells because of the neutral and inert PEG chains on UK.
Tables 1 and 2 ~how the relative radioactivity in various tissues after 125I-native UK and 125I-M-Md ~L77~

injection, respectively. FIGURE 11 illustrates the relative radioactivity in Blood (~) for 125I-M-Md and 125I-native UK-RELATIVE RADIOACTIVITY AF~ER
125I-NATIVE UK INJ_CTIO~

Time Organ 5 min.10 min.20 min. 40 min.12 hr Brain 0.050.03 0.04 0.010.01 Lung 0.470.62 0.31 0.100.01 Thymus 0.030.02 0.03 0.030.01 Liver 27.8932.60 23.06 2.590.46 Heart 0.140.26 0.17 0.170.10 Kidney 3.686.82 2.40 0.400.25 Adrenal 0.05 -0.07 0.05 0.020.00 Spleen 0.39 1.32 0.67 0.050.02 RELATIVE RADIOACTIVITY AFTER
125I_M_MD INJECTION __ Time organ 5 min. 10 min.20 min~40 min.12 hr Brain 0.17 0.20 0.20 0.220.11 Lung 1.24 2.30 1.31 1.370.71 Thymus 0.15 0.21 0.15 0.240.11 Liver 6.84 8.7511.55 7.472.91 Heart 0.93 1.73 1.38 1.020.53 Kidney 2.21 3~10 3.83 3.270.68 Adrenal 0.06 0.07 0.08 0.060.04 Spleen 0.45 0.40 0.63 0.510.21 EXamP1e 29 - IMMUNOGENICITY AND TOXICITY OF M-MD
The changes in immunogenicity ~y chemical modification of proteins are of interest.

7~3 Immunogenicity of M-Md was compared with that of native UK by ASA reaction, Schultz-Dale test, PCA reaction and Ouchterlony method. Both native UK and M-Md were immunogenic for guinea pigs by ASA reaction after sensitization of them with the respective UK and FCA.
However, the immunological response to M-Md was weaker than that of native UK in the other three tests.
The results summarized in FIGURES 12a and 12b suggest that immunological determinants in M-Md are common with those in native UK. Furthermore, antigen production and/or reactivity with antigen are/is suppressed in M-Md compared with native UK. As native UK itself does not cause immunological response in man, M-Md is not expected to either.
The toxicity of M-Md was found to be extremely low. None of the mice died even after a shot of M-MD
at a dosage of 1,000,000 ~/kg nor afte~ repetitive injection of 500,000 ~/kg/day of M-Md.
FIGURE 12a shows a comparison of the immunogenicity of M-Md with that oE native UK by Schultz-Dale test~ FIGURE 12b shows a comparison of the immunogenicity of M-Md with that of native UK by Passive Cu~aneous Anaphylaxis (PCA) reaction.
In ~ummary, the instability of native UK has created a serious problem in formulating it for clinical purposes. The foregoing examples illustrate that covalent attachment of PEG chains to native UK
greatly increases resistance to inhibition of the UK by UK inhibitors. Moreover, covalent attachment of PEG
chains to native UK also prolonged ~2f the UK in the circulatory system. It seems likely that the PEG
chains may surround the UK molecule to form, in effect an inert capsule to protect the UK molecule from inhibitors and other proteases.
In the next set of examples attention was focused on _ vivo experiments using dogs, where the thrombolytic ability of PEG-modified UK (PEG-UK) was compared wlth that of native UK. In these examples, the PEG-UK used is the same as M-Md in Examples 25-29. The unit for UK activity was determined by the standard fibrin plate (F.P.) method. Plasmin was obtained from Sigma Chemical Co. FDP latex for dog FDP
was obtained from MBL Co., Ltd. Terufus~on~, a blood transfusion set type 1, obtained from Terumo Co., Ltd.
was connected with a three-way stopcock of type PX2-50 from Top Co., Ltd. using ATOM 6 Fr extension tubing.
Other materials used in the following examples are the same as used in Examples 25-29.
T~e haemostatic index for plasmin inhibitor, Fl, was determined by an addition of plasmin to the Fl fraction followed by the assay of residual pla~min with S-2251. F3, plasminogen, was determined with a ~Z~,t - 7 ~

combination of UK and S-2251. FDP was determined wi~h latex sensiti~ed with antigen to dog fibrinogen.
For the shun~ preparation, a dog was anaesthetized with ketamine hydrochloride then A. femoralis and V.
femoralis were exposed in the inguinal region. A
saline solution-filled shunt with the three-way stopcock was connected between them. Blood was allowed to run by opening the stopcock while monitoring the flow with a magnetic bloodflow meter. Either native UK
or PEG-UK dissolved in saline solution was administered from the stopcock. To maintain the anaesthesia, sodium barbiturate was injected.
For the preparation of artifical thrombus, A.
fermoralis of an anaesthetized dog was exposed by the same procedure. A transfusion catheter was fixed at a branch of A. femoralis, then the distal portion of the A. femoralis was clamped with two b~lldo~ clamps at a distance of 2 cm from each other and was emptied, then washed with saline solution. Dry oxygen was passed therethrough at a rate of 2 l/min for 10 minutes to injure intima. After injection of a fibrinogen solution followed by thrombin solution, the formed thrombus was allowed to age for 10 minutes. Then, the bulldog clamps were removed. Thrombus formation was confirmed angiographically by an injection of 10 ml of 60% Urografin~ from the catheter. UK dissolved in 16~
- ~o-saline was injected from the same catheter and angiographies were taken periodically thereafter.

Example 30 - CHANGES IN Fl, F3 AND FDP AFTER UK
INJECTION
Changes in the haemostatic indices are good indications of UK treatment. Often used are Fl (plasmin inhibi~or), F3 (plasminogen) and FDP (fibrin and/or fibrinogen degradation product). Native UK or PEG-UK was injected into a dog at the superficial vein of the fore leg at dosages shown in FIGURES 13a, b and c. Blood was sampled from the opposite side at certain time intervals and the Fl, F3 and FDP contents of the samples were determined.
A decrease in both Fl and F3 levels is a direct proof of plasminogen activation. The consumption of F
and F3 indicates that plasmin activated by UK from plasminogen (F3) was deactivated by its inhibitor (Fl) and was removed from the circulatory system. Another proof is a prominent increase in FDP. In a dog injected with PEG-UK, the FDP level was kept higher for several hours although its enæymatic activity to activate plasminogen was lower than that of native UK. These findings reveal clearly that chemical modification with PEG amplifies the physiological activity of UK.

7'~

FIGURE 13a illustrates a comparison of the Plasmin Inhibitor in Fl from Citrated Dog Plasma at certain time intervals after injection of native UK or PEG-UK
into a dog.
FIGURE 13b illustrates a comparison of the Plasminogen in F3 from Citrated Dog Plasma at certain time intervals after injection of native UK or PEG-UK
in~o a dog.
FIGURE 13c illustrates a comparison of the FDP in Dog Serum at certain time intervals after i.v.
injection of UK or PEG-UK into a dog.

Example 31 - THROMBOLYTIC EFFECT OF PEG-UK I~
EXPERIMENTAL THROMBOSIS
Thrombolytic activities of native UK and PEG-UK
were compared in dogs with experimental thrombosis. A
shunt having a fine filter was fixed between A.
femoralis and V. femoralis and then blood was run through it. After a complete stop of flow (several minutes) either native UK or PEG-UK was administered through the three-way stopcock at a dose of 10,000 ~/kg. As a result, the flow of blood in the dog with PEG-UK recovered while not in the dog having native UK administered even after repeated flushes of saline ~olution.

1' ;LI'i7.~3 -~2-Example 32 - INHIBITION OF THROMBUS FORMATION WITH
PEG-UK
The inhibitory effect of PEG-UK on the thrombus formation was investigated as compared with native UK
with a saline solution-filled shunt prepared by the same procedures de~cribed above. In this experiment either native UK or P~G-UK was administered prior to the blood circulation. The dose was 20,000 ~/kg in both cases. Results obtained are as follows:

1) In the dog without UK, the blood flow stopped within 5 minutes.
2) In the dog with native UK, the blood flow stopped within 5 minutes.
3) In the dog with PEG-UK, the blood flow continued during the experiment for >1 hour.
At this stage, the three filters were washed with saline. Complete inhibition of thrombus formation was observed in PEG-UK, which is probably due to the thrombin inhibition by FDP.

Example 33 - LYSIS OF ARTIFICIAL THROM~US WITH PEG-UK
USING ANGIOGRAM
The thombolytic effect of PEG-UK was evaluated in dogs bearing artifical thrombus in A. femoralis. After confirmation of thrombus formation, either native UK
(dog "A" ) or PEG-UK (dog "B") was injected from the transfusion catheter fixed on a proximal branch of the ~7~

thrombus at a dose of 10,00~ ~/kg. Angiographies were taken periodically to observe ~he recovery of blood flow. It is concluded from these angiographies that PEG-UK dissolved the artifical ~hrombus formed in A.
femoralis while native UK was unable to do so. FIGURE
14 illustrates a comparison of FDP levels in dogs bearing artifical thrombus in A. femoralis. Dog A had native UK injected into a proximal branch of the thrombus, while dog B had PEG-UK injected into the same.

Example 34 - INHIBITORY EFFECT ON THROMBUS FORMATION
Using a dog, a shunt was established under anesthesia by a blood transfu~ion filter ("Terufusion"
blood transfusion Set Type-l) between A. femoralis and V. femoralis and a blood flowmeter ("MF-26", rectangular wave electromagnetic bl~od f~-owmeter manufactured by Nippon Koden K.K.) was connected at the vein side. The filter had in advance been filled with physiological saline and, after administration of a drug, the blood was caused to flow through the shunt by opening a three-way stopcock. Without administration of the drug, thrombus was formed in the filter within 4-5 minutes after the blood had started flowing through the filter, whereby any further flow of the blood ~hrough the shunt was preven~ed. The time period until 1~ o 7 ~

the stoppage of the blood flow due to the formation of thrombus was not prolonged at all even when 200,000 units of unmodified urokinase were administered. On the contrary, thrombus was not formed nor was the blood flow rate reduced at all even after 40 minutes after the initiation of the blood flow through the shunt when PEG-DCT-UK, which pertains to the present invention, had been administered.
In summary, Examples 30-34 dealt with the thrombolysis by PEG-UK wherein the thrombolytic ability thereof was compared with that of native UK using two thrombosis models. The superiority of PEG-UK to native UK with respect to fibrinolytic activation is due to the PEG chains which appear to protect the UK molecule from deactivating interactions with inhibitors. This protection serves to extend the UK circulating life.
Having fully described the pre5ent-~nvention, it will be apparent to one of ordinary skill in the art that many changes and modifications can be made thereto without departing from the spirit or scope of the invention as set forth herein.

Claims (5)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OF PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS.

1. A method of extending the circulating life of a plasminogen activator in a mammalian bloodstream which comprises chemically modifying said plasminogen activator by bonding at least one polyalkylene glycol moiety there-to by means of a coupling agent to the amino acid side chains of said plasminogen activator.

2. The method of Claim 1, wherein said plasminogen activator is urokinase.

3. The method of Claim 1, wherein said mammalian bloodstream is a human bloodstream.

4. The method of Claim 1, wherein said polyalkylene glycol moiety is a methoxypolyethylene glycol moiety.

5. The method of Claim 4, wherein said methoxy-polyethylene glycol has a molecular weight of about 5,000.