METHODSFORMEASURINGTHE ACTIVITY OF COMPONENTSOFTHE STREPTOCOCCALFIBRINOLYTIC SYSTEM, AND

STREPTOCOCCALDESOXYRIBONUCLEASE'

By L. R. CHRISTENSEN(From the Department of Microbiology, New York University College of Medicine,

New York City)

(Received for publication September 22, 1948)

Studies of the effects of streptokinase anddesoxyribonuclease from hemolytic streptococci oncertain pathological processes, reported in an ac-companying paper (1), necessitated the develop-ment of quantitative methods suitable for routinelaboratory assay of the various substances knownto affect the streptokinase-plasminogen (strepto-coccal fibrin-lysing) system. Methods were de-vised, or earlier methods modified, for the routineassay of streptokinase (streptococcal fibrinolysin),antistreptokinase (antifibrinolysin), serum inhib-itor (serum trypsin inhibitor), and plasminogen(serum protease, serum fibrinolysin, serum tryp-tase).

MATERIALS

1. Crude streptokinase:2 The method used forproduction of streptokinase is based on one previ-ously published (2), but differs principally in theuse of ethanol as a precipitant rather than am-monium sulfate.

Cultures of Group C strain H46A are grown inthe presence of high glucose concentration, em-ploying intermittent neutralization with NaOHin the manner previously reported (2). Theyields of streptokinase and desoxyribonucleasefrom atypical culture are presented in Figure 1.It will be noted that the peaks of enzyme pro-duction coincide with the peak of glucose utiliza-tion. As incubation is continued, both strepto-kinase and desoxyribonuclease levels decline.For this reason it is best to begin harvesting ofthe culture shortly before glucose utilizationceases.

'This study was supported in part by a grant fromthe Life Insurance Medical Research Fund and the RalphB. Rogers Rheumatic Fever Fund.

2We are greatly indebted to Dr. Benjamin Carey andMr. J. N. Adam, Jr. of the Lederle Laboratories Divi-sion of the American Cyanamid Company for supplyinglarge quantities of crude streptokinase for these studies.

At the end of the growth period the cells areremoved by Sharples centrifugation at about40,000 rpm. The supernatant is adjusted to pH3.84.2 by the addition of glacial acetic or hy-drochloric acid. Cold ethanol at ' 80 C is addedto the supernatant at room temperature to a finalconcentration of 40-50%. After standing over-night in the cold, the practically clear supernatantis siphoned off and the flocculent precipitate, whichcontains the streptokinase and desoxyribonucle-ase, is collected by centrifugation at 2,000-3,000rpm. The precipitate is dissolved in borate bufferwith the aid of alkali, adjusted to pH 7.8-8.0, andlyophilized. This precipitate contains the cellsnot sedimented by Sharples centrifugation, vari-

° STREPTOKINASE* DORNASE'' 5 N NAOH

Z8X10 800o

z <

-I 4XI / 4004

0-O r _

I-

1 2 3 4 5 6INCUBATION TIME

HOURS

FIG. 1. RELATION OF STREPTOKINASE AND DEsoxy-RIBONUCLEASE PRODUCTIONTO GROWTH(ExPRESSED ASML. 5N NAOHADDEDTO NEUTRALIZE AcI FoRiD)

1 Desoxyribonuclease.

ous streptococcal proteins, some media compon-ents, nucleic acids, streptokinase and desoxyribo-nuclease.

2. Purified streptokinase: The purification pro-cedure is a modification of an earlier method (3).A 2-5% solution of crude streptokinase is made

163

L. R. CHRISTENSEN

in borate buffer, pH 7.8-8.0. Cells and insolubledebris are spun out at 16,000 rpm in a refrigeratedangle centrifuge. Occasionally a significant por-tion of the activity is also spun down by thistreatment. If this is the case, the precipitate isresuspended and made alkaline to thymol blue(above pH 9). The solution is then adjusted topH 7.8-8.0 and centrifuged. Following thistreatment the activity remains in solution. Thesupernatants are pooled and diluted with distilledwater to a nitrogen content of about 0.25-0.50mg./ml. and sufficient Pillimer's acetate buffer isadded (4) to reduce the pH to 5.6. Sufficient1%o protamine 8 solution in distilled water isadded to cause the formation of a flocculent pre-cipitate (usually about 1 ml./5-10 mg. nitrogen).The protamine precipitate is spun out in a re-frigerated angle centrifuge at 4,000-5,000 rpm.The supernatant should show no further precipi-tation on testing with protamine, but care shouldbe used to avoid any great excess of protamine,since it will complicate later purification steps.The supernatant is cooled to about 00 and ethanolat - 100 to - 200 is added to a final concentra-tion of 20%b. On standing overnight at - 80 to- 100 a flocculent precipitate forms which isspun off in an angle centrifuge in the cold. Theprecipitate is dissolved in borate buffer, pH 7.8-8.0. Some insoluble material remains in suspen-sion, but in most cases little activity is associatedwith it; and it can be removed by centrifugation,resulting in little or no decrease in total activity,but a decrease of about 50%o in total nitrogen.The clear solution usually contains between 100and 180 units of streptokinase per microgram ofnitrogen. Recovery usually represents about25%b of the original activity of the culture, some50%o of the original activity being lost during theinitial Sharples centrifugation. Desoxyribonu-clease, which has accompanied streptokinase inthe previous purification steps, is separated fromstreptokinase by the 20%o ethanol fractionation.

8 Several sources of protamine have been tested. Inour experience, consistently good results are obtainedwith Squibb protamine. Protamines from other sourcesoften fail to effect purification, and are less efficient inthat much larger quantities are necessary to precipitateinert materials. We wish to express our gratitude toE. R. Squibb and Sons for generous gifts of protamine.

Practically all of the desoxyribonuclease activityremains in the supernatant.

The purified streptokinase is sterilized by fil-tration through Seitz or Selas filters, lyophilizedand stored in the refrigerator. Solutions remainstable for several weeks if not too dilute and ifkept in the cold.

3. Streptococcal desoxyribonuclease: As men-tioned in previous reports (5-7), streptococcalcultures contain a potent desoxyribonuclease.In the original method of purification (2) usingammonium sulfate, both enzymes were found inthe same fraction. However, they are readilyseparated since streptokinase is precipitated at pH5.6 by 20% ethanol or 50% ammonium sulfate,while the majority of the desoxyribonuclease re-mains in solution. In the present study, desoxy-ribonuclease was prepared by bringing the 20%oethanol supernatant of the above protamine-puri-fied streptokinase precipitate to 40-50%o ethanol.The precipitate which forms overnight in the coldis collected by centrifugation and dissolved in thegelatin-magnesium-veronal buffer recommendedfor desoxyribonuclease titration by McCarty (8).The preparations usually contain insoluble ma-terial, with which is associated some of the de-soxyribonuclease activity. The clear supernatantobtained by centrifugation of the material, how-ever, contains a significant proportion of activity.

Up to the present time methods have not beenperfected for the quantitative isolation of the en-zyme from streptococcal cultures and recoveryrepresents only a fraction of the amount in the

TABLE I

Fractionation of streptokinase and desoxyribonucleasefrom streptococcal cultures

Total unitsDesoxyribo-

Fraction Streptokinase nuclease

(1) original culture 1.75 X 106 units 5.25 X 10' units2) after Sharples cent- 1.5 X 106 4.1 X 104

rifugation(3) alcohol precipitate 1.2 X 106* 3.0 X 104

of culture super-natant

(4) 20% alcohol ppt. of 1.38X106 0.1X 104protamine superna-tant

(5) 40%alcohol ppt. of 0.O1X 10' 1.8X104(4) supernatant

* This value is somewhat lower than the true valuebecause sufficient time for reversal of the inactivation,which occurs in the neighborhood of pH 5 (3), was not al-lowed before titration.

164

STREPTOKINASEAND STREPTOCOCCALDESOXYRIBONUCLEASE

original culture. This loss is due apparently tothe instability of the enzyme.

In Table I are presented the data obtained ina typical purification experiment.

4. Bovine fibrinogen: For reasons to be dis-cussed later, human fibrinogen preparations were

considered unsuitable for assay purposes. Ar-mour's Fraction I of bovine plasma,' a commercialproduct, has been found to be quite suitable forour purposes, since it contains no antistreptoki-nase, serum inhibitor or plasminogen activatableby streptokinase. The dry powder is stored in therefrigerator and dissolved in borate buffer, pH7.4-7.6, in appropriate concentrations before use.

Solutions are reasonably stable and generally re-

main usable for six to eight hours if kept cool.5. Plasminogen: No satisfactory preparation

of plasminogen is available. The purer prepara-

tions with which we are familiar, such as HowardFraction III-2,3 (9), contain plasmin as wellas plasminogen and proteolytic products of fibrin.The cruder preparations, prepared by salt frac-tionation (2), isoelectric precipitation (2, 10) or

ethanol fractionation, such as Harvard FractionIII (11), contain a variety of substances, some

of which undoubtedly influence the streptokinase-plasminogen reaction. However, reproduceableresults have been obtained in these laboratoriesby using a single lot of Harvard Fraction III.5For use, the material is made up in borate bufferin a concentration of 0.25%o. The solutions mustbe used immediately, and kept in an ice bath, be-cause some spontaneous plasmin activity will de-velop in a few hours. However, when used as

outlined, we have noted no variation in the plas-minogen titer of the dry material in over a year.

6. Standard antistreptokinase: A pool of Har-vard Fraction II-III was fractionated by Har-vard method 9 (12). Fraction II, the antibodyfraction, was lyophilized and stored in a vacuum

dessicator.6 Assay of the material with strepto-

'We are indebted to Dr. J. B. Lesh of Armour Labora-tories for supplying the Fraction I used throughout thisstudy. Fraction I is an ethanol fraction of bovine plasmaand consists largely of fibrinogen.

5 Weare indebted to Dr. Dwight Mulford, Division ofBiologic Laboratories, Massachusetts State Departmentof Health, for supplying a large lot of Fraction III iso-lated from a single pool of plasma.

6 The author wishes to express his indebtedness to Dr.J. W. Williams and associates of the Department of

kinase indicated that 1 mg. would inhibit the ac-tivity of 10 units of streptokinase, when tested bythe method outlined below. The material was as-signed, therefore, a value of 10 antistreptokinaseunits per mg., and used throughout these studiesas a primary standard to check the constancyof the streptokinase unit and of the antistreptoki-nase titrations.

7. Crystalline trypsin: A solution of commercialcrystalline trypsin, obtained from the Plaut labor-atories, is prepared in N/400 HCl. The stocksolution contains 0.1% of the dry powder, ofwhich about 50%o is magnesium sulfate. Theconcentration of active trypsin in the solution isdetermined by titration with soybean inhibitor, asoutlined below, and appropriate dilutions for use,usually containing 20 p&g. of active trypsin permilliliter, are made up in borate buffer immedi-ately before use.

8. Crystalline soybean inhibitor: A stock solu-tion of Kunitz' crystalline inhibitor 7 (13), con-taining 2 mg./ml. of inhibitor, was prepared inborate buffer. Appropriate dilutions are madeup in borate buffer, pH 7.4-7.6, before use.

9. The preparation of other reagents, such asborate buffer and gelatin buffer, have been de-tailed in previous reports (2, 14).

METHODS

1. Streptokinase assay: Many methods havebeen described for streptokinase assay, utilizingclots prepared from human fibrinogen (15-18).These methods depend on the presence of plas-minogen as a contaminant of the fibrinogen prep-arations, since, as Milstone (10) has shown,highly purified human fibrinogen will not lysein' the presence of streptokinase unless anotherserum factor (lysin factor, plasminogen) is added.In our experience, methods utilizing humanfibrinogen are unsuitable for long-term routineuse. The principal disadvantage is the variationfrom batch to batch in the fibrinogen-plasminogenratio obtained with different methods of prepa-

Physical Chemistry of the University of Wisconsin formaking facilities available and for help in the preparationof various plasma fractions.

7Dr. M. Kunitz kindly supplied us with a preparationof crystalline soybean inhibitor for use as a standard,and made available his assay methods in advance ofpublication.

165

L. R. CHRISTENSEN

ration, and even with the same method. For ex-ample, in our experience ethanol fractions ofplasma, such as Harvard I, contain less plas-minogen than do the salt fractionations previouslyused in this laboratory (2), as shown by the factthat clots treated with streptokinase will lysemuch more readily if the fibrinogen is preparedby salt fractionation rather than the ethanol.Also, human fibrinogen preparations are unstable,due to the spontaneous conversion of plasminogento plasmin. Another difficulty in the use of hu-man fibrinogen is the presence of variable amountsof antibody globulin in the preparations. Whilethis occasions no particular difficulty with mostsamples of normal plasma whatever the methodof fractionation, we have encountered occasionallots of plasma with such high titers of antistrepto-kinase that they were unusable.

For these reasons, we have modified the test bythe substitution of clots prepared from bovinefibrinogen (Armour's Fraction I). The bovinefibrinogen contains no plasminogen or plasmin, asshown, by the fact that such clots, without addedprotease, will remain stable for days in the pres-ence or absence of streptokinase. Antistrepto-kinase and inhibitor titrations indicate the absenceof these substances also. As a source of plas-minogen we have used a single large lot of Har-vard Fraction III. The fibrinogen concentrationused is 0.1%, as in previous reports, and the con-centration of Fraction III is adjusted to give astreptokinase unit equal in value to the one pre-viously employed in this laboratory (3). Thestreptokinase assay ,is performed as follows:

0.1 ml. streptokinase dilution in gelatin buffer0.4 ml. bovine fibrinogen, 0.25% in borate

buffer0.5 ml. Fraction III, 0.25% in borate buffer.

(This solution must be prepared just beforeuse, and should be kept on ice)

0.1 ml. Lederle Hemostatic Globulin,8 diluted1: 3 in borate buffer.

8 Lederle Hemostatic Globulin has recently become un-available. Parke-Davis bovine thrombin may be substi-tuted as well as Upjohn thrombin. They are less stablethan Hemostatic Globulin when in solution in boratebuffer, and solutions must be titrated frequently to deter-mine the minimum amount necessary to form a firm clot.These preparations are more stable in glycerol solution,but we have not used glycerol solutions.

The tubes are incubated at 350 and the lysis timeof each tube of the series noted. The dilutionlysing the clot in 10 minutes contains one unit andis determined by interpolation as outlined in aprevious report (3).

2. Antistreptokinase: The antistreptokinaseunit is defined as the amount of antibody re-quired to inhibit 1 unit of streptokinase to sucha degree that it will not activate the plasminogensufficiently to cause lysis of the standard test clotin 30 minutes. Other methods of antistrepto-kinase assay have been presented (15, 18-20) butno general agreement has been reached on amethod. We have thought it preferable to basethe antibody unit on the method of streptokinasetitration used in this laboratory. In order tohave a primary standard of reference throughoutthe study, a lot of Fraction II, the gammaglobu-lin fraction of human serum, was prepared andlyophilized. This preparation contained 10 anti-streptokinase units per mg., and has been used asa reference standard for streptokinase and anti-streptokinase titrations. Standardization of strep-tokinase with standard antibody is performedas follows:

0.5 ml. streptokinase dilutions in gelatin buffer,pH 7.6-7.8

0.5 ml. standard antibody solution, containing2 antistreptokinase units per ml. (The anti-body solution is heated at 560 for 30-45minutes to destroy the small amount ofof plasminogen and inhibitor which it con-tains.)

The mixture of streptokinase and antibody isincubated at 350 for 30 minutes. At the end ofthis time, the following reagents are added:

0.5 ml. of a 0.25%o solution of Fraction III0.5 ml. of 0.5%o bovine fibrinogen0.1 ml. Hemostatic Globulin, diluted 1: 3.

The tubes are incubated for another 30 minutesand the highest dilution of streptokinase whichfails to lyse the clot is taken as the end point.A solution of crude or purified streptokinasestandardized in this manner should show no sig-nificant change in titer over a period of a week ortwo if the streptokinase concentration is aboveabout 2,000 units per ml. for the crude and above

166

STREPTOKINASEAND STREPTOCOCCALDESOXYRIBONUCLEASE

10,000 units per ml. for the purified. More di-lute solutions are not quite so stable, probablybecause of denaturation occurring in dilute pro-tein solutions. However, in practice, the strep-tokinase solution is standardized every day or two.

In determining the antistreptokinase titer ofserum the serum must be heated at 560 for 30-45minutes to destroy the plasminogen and inhibitorwhich it contains. As may be seen in Table II,

TABLE II

Effect of heating serum at 56° Cfor 30 minutes onserum inhibitor and antistreptokinase

Inhibitor equivalent to Antistreptokiaemicrograms trypsineSerumnumber

before after before afterheating heating heating heating

1 1000 100 128 1282 500 80 64 643 1000 -80* 64 324 667 100 64 325 667 -80* 128 646 667 -80* 16 87 1800 120 2560 25608 800 48 3200 3200

* Inhibitor equivalent to less than 80 pg.heating reduces the inhibitor concentration byapproximately 90%o, and produces no significantchange in the antibody titer. A more valid rea-

son for heating the serum than for the inactivationof inhibitor is to destroy plasminogen, since ifthis is not done, in the case of sera with a lowantibody content, the tubes containing the low-est dilutions of serum will lyse the standard clotbecause of the activation of the plasminogen inthe serum by uncombined streptokinase.

The inactivated serum is titrated for anti-streptokinase as follows:

0.5 ml. inactivated serum dilution0.5 ml. streptokinase solution, diluted in gela-

tin buffer to contain about 1-2 units.

The tubes are incubated for 30 minutes, and theremainder of the titration carried out as above.The end point is taken as the highest dilutionof serum preventing lysis of the clot, and con-

tains antistreptokinase equivalent to the strepto-kinase in the test.

3. Serum inhibitor: A number of methods havebeen devised in the last few years for the quanti-tative estimation of serum inhibitor (14, 18, 21-

23). These methods, however, have certain dis-advantages for routine use.

Kunitz (13) has shown that crystalline soybeaninhibitor undergoes an immediate, stoichiometric,irreversible reaction with trypsin, and that activetrypsin can readily be determined by titration ofthe amount of soybean inhibitor necessary toinhibit the proteolytic activity of the trypsin sam-ple, since the two are equivalent, weight forweight. Schmitz (24) has isolated a smallamount of inhibitor from serum which he con-cluded was probably identical with the pancreaticinhibitor. It has also been shown (14) thatserum and pancreatic inhibitor behave in thesame way toward trypsin and toward plasmin, al-though the two inhibitors behave toward trypsinin a quantitatively and perhaps qualitatively dif-ferent manner than they behave toward plasmin.In view of these facts it was decided to determinethe inhibitory power of serum in terms of itsability to inhibit trypsin rather than plasmin.Additional reasons influencing this decision wereas follows: (a) crystalline trypsin is readily avail-able and easily standardized, whereas no plasminpreparations of comparable purity are known; (b)trypsin is much more susceptible to the action ofinhibitor than is plasmin; and (c) active trypsincan be conveniently stored in a stable solution(in M/400 HCl) while plasminogen and plasminpreparations are very unstable (14, 25, 26).

Standardization of trypsin with soybean in-hibitor is carried out as follows:

0.5 ml. freshly prepared dilutions of stock tryp-sin in borate buffer, pH 7.4-7.6

0.5 ml. soybean inhibitor, usually containing10 Mg.

1.0 ml. bovine fibrinogen, 0.2%o in borate buffer.

The tubes are incubated for 10 minutes at 350.At the end of this time 0.1 ml. of HemostaticGlobulin, diluted 1: 3, is added and the formationof clots noted in 5-10 minutes. The method isbased on Ferguson's (27) procedure for thefibrinogenolytic assay of tryptases. The lowestdilution of trypsin in which a clot forms containstrypsin equivalent to the soybean inhibitor in thetest. It is necessary to allow the trypsin to actin the absence of Hemostatic Globulin becausethe Hemostatic Globulin preparations used con-

167

L. R. CHRISTENSEN

tain some trypsin inhibitor. An advantage tothis procedure, however, is that the addition ofHemostatic Globulin stops all tryptic activity, sothat once the clots are formed, they are stable.Any desired range of trypsin concentration canbe titrated by this method by using appropriateconcentrations of inhibitor in the test. The sen-sitivity obtained depends on the interval betweensuccessive trypsin dilutions.

The inhibitor concentration of serum is deter-mined as above with the substitution of dilutionsof serum for the soybean inhibitor and the sub-stitution of a fixed amount of active trypsin, usu-ally about 10 pg., for the trypsin dilutions. Thetube containing the highest dilution of serum inwhich a clot forms on addition of thrombin con-tains serum inhibitor equivalent to the trypsin inthe test. Results are expressed as microgramsof active trypsin inhibited by 1 ml. serum.

The accuracy of the fibrinogenolytic test wascompared with the method of Kunitz (13) for de-termining trypsin inhibitor, with results as shownin Table III.

TABLE III

Comparison of methods for serum inhibitor determinationInhibitor equivalent to micrograms trypsin

inhibited per ml. serumSerum Kunitz Fibrinogenolytic

number method method

1 520 5002 490 5003 590 660

4. Plasminogen: Plasminogen determinations ofserum are of doubtful accuracy because plasmino-gen cannot be separated quantitatively from serum,nor can the interfering substances, antistrepto-kinase and serum inhibitor, be removed or de-stroyed without loss of plasminogen. However,in samples of sera which are not abnormally highin antibody or inhibitor, the relative concentra-tions of plasminogen can be determined by acti-vating with streptokinase and determining theleast amount of activated serum necessary to lysea standard clot. The details of the determinationare as follows:

0.5 ml. serial dilutions of serum in borate buf-fer

9 As noted above, Hemostatic Globulin is no longeravailable. Preliminary experiments with the more highly-purified bovine preparation of Parke-Davis indicates thatthey contain little or no inhibitor.

0.1 ml. streptokinase solution, previously stand-ardized as outlined below

0.4 ml. bovine fibrinogen, 0.5%o in borate buffer0.1 ml. Hemostatic Globulin, diluted 1: 3.

The highest dilution of serum lysing the clot in 30minutes at 350 represents one unit of plasmin,or plasminogen.

It is necessary to activate the plasminogen inthe presence of substrate, because, if this is notdone, the total protease activity declines rapidly,probably because of the two-fold effect of auto-digestion (25) and combination with serum in-hibitor. Activation in the presence of substrateappears to minimize these effects. Since the ac-tivation of plasminogen by streptokinase appearsto be a catalytic reaction (2, 14, 28-31) it will bemost efficient when an excess of streptokinase isused. A large excess of streptokinase would alsoprevent interference with the reaction by theantistreptokinase contained in variable amountsin practically all human sera. When attempts aremade to use maximal amounts of crude strepto-kinase to activate serum plasminogen, an inhibi-tory effect is noted. This inhibitory effect of un-diluted, concentrated crude streptokinase is due,probably, to the presence in crude streptokinasepreparations of a protease inhibitor (14) whichdecreases in amount as streptokinase is purified.When purified streptokinase is used to activate aserum, maximum protease activity is obtainedwith the highest concentration of streptokinase, asshown in Table IV.

TABLE IV

Plasminogen activity of serum activated with variousconcentrations of purified streptokinase

Purified streptokinase Highest dilution of serumdilution causing lysis in 30 minutes

undiluted 51201:10 51201:100 12801:1000 160

In addition to the absence of an inhibitory effectof high streptokinase concentration when usingpurified material, it can be seen from the data inTable V that the activity obtained is considerablyhigher than with the crude material. It is obvious,therefore, that activation with purified materialis the method of choice. Unfortunately, it hasnot been possible in the past to obtain sufficientpurified material for all requirements of the stud-

168

STREPTOKINASEAND STREPTOCOCCALDESOXYRIBONUCLEASE

ies involved in these reports. Therefore, proteasedeterminations were made with selected batchesof crude streptokinase, titrated to determine thoselots and dilutions which would give the highest

TABLE V

Relation between maximum plasmin activity developedin serum by crude and purified streptokinase

Maximum plasmin activityCrude Purifiedtrepto- strepto-

Serum kinase kinaseL (6-19-48) 1280 3200L (7-16-48) 640 1600S 320 1280B 160 1280F 320 2560

values with normal sera. Such lots were setaside and used in all the protease determinationsreported in the present studies. The results ob-tained, while not representing the maximal val-ues, are useful as a measure of the relative plas-minogen content of sera. The validity of this as-sumption is borne out by the observation (TableV) that a direct relation exists between the maxi-mumamount of protease activity obtained on theactivation of serum with crude and with purifiedstreptokinase.

5. Desoxyribonuclease: The determination iscarried out as described by McCarty (7), exceptthat drop in relative viscosity is measured overa 10-minute instead of a 20-minute period of in-cubation. A unit is defined as the amount of en-zyme necessary to cause a drop of 1 viscosity unitin 10 minutes at 30° in a reaction mixture con-sisting of 2.4 ml. of desoxyribose nucleic acid and0.1 ml. enzyme solution. The substrate consistedof a 0.15-0.2o solution of thymus desoxyribosenucleic acid in veronal buffer. The nucleic acidwas prepared as described by Mirsky (32). Dif-ficulty was encountered with the nucleic acidsolutions in that a spontaneous drop in viscositywas noted. This effect could, be eliminated byheating the nucleic acid solution to 560 for aboutan hour. The heated solution, when used in thereaction mixture described above, had an initialrelative viscosity of 4-5 units.

RESULTS

1. Antistreptokinase: Evidence is abundant thatan antibody to streptokinase appears followingstreptococcal infection (18, 33-36). In Table VI

are presented the results obtained by using themethods described above.

TABLE VI

Antistreptokinase levels in various sera

Antistreptokinase unitsSerum ~~No.Serum samples

Mean Range

Normal 14 36 4-80Infants (under 1 year) 9 8 0-29Rheumatic fever 8 1302 100-4200

(children)Streptococcal pharyn- 3 92 80-115

gitisTuberculous pleurisy 22 87 2-400Dog 3 102 72-156Monkey 1 66Cow (pool) 8Horse (pool) 4Chicken 1 32

Direct evidence that antistreptokinase is pro-duced in humans following injection of strepto-kinase has been obtained in the present series ofstudies and is reported more fully in the accom-panying paper (1). In Table VII are shown theantibody titrations in the serum of two patientsfollowing the intrapleural injection of purifiedstreptokinase. The antibody responses seen in thetwo patients reported in Table VII are perhaps

TABLE VII

Rise in antistreptokinase foUlowing intrapleuralPatient injection of purified streptokinase

M. H. (female)

Patient

C. L. (male)

Date

6/8 (200,000 SK in-jected intrapleurally)

6/116/176/22

5/13 (50,000 units Sk in-jected intrapleurally)

5/17 (100,000 units SK)5/206/96/227/2

ASKunits20

805120

over 10,000

40

5120over 10,000over 10,000

5120

somewhat more rapid than is usual as a primaryresponse to the injection of an antigen. Therapidity of the response, however, is probably dueto the fact that the response is an anamnestic one,since it will be noted that both patients had a titerbefore injection of streptokinase. In fact, all hu-man sera tested have had some antibody, even

169

L. R. CHRISTENSEN

with no known history of streptococcal infection.2. Serum inhibitor: The values obtained for a

variety of sera are presented in Table VIII. Re-sults are expressed as micrograms of active trypsininhibited by 1 ml. of serum.

TABLE VIII

Serum inhibitor levels of serum from man and animals

Serum inhibitor as micro-grams trypsin inhibited

Serum No. of per ml.samples-Mean Range

Normal adult 16 655 100-1200Infants (under 1) 9 613 420-1024Rheumatic children 5 760 570-900Pulmonary tuberculosis 7 1035 660-1500Carcinoma 4 883 600-1333Leukemia 5 1096 842-1350Chicken 1 264Rabbit 1 400Horse (pool) 2 765 670-860Dog 1 365Guinea pig 2 2125 1750-2500

The data presented are obviously too scanty towarrant the drawing of conclusions. However,the finding that the value is elevated in chronicwasting diseases such as tuberculosis and malig-nancies is in accord with many earlier observa-tions on serum inhibitor (see Grob [21] forreferences).

The values for animal sera tested lie withinthe range of normal human values, with the ex-ception of guinea pigs, which, at least in the fewspecimens tested, have an abnormally high value.

TABLE IX

Levels of plasminogen in various sera

Units of plasminogenSerum ~ No. ofSerum specimens

Mean Range

Normal adult 8 572 160-1280Rheumatic children 3 960 320-1280Tuberculosis 6 4053 1280-5120Carcinoma 1 2560Non-rheum. H. D. 5 576 320-640Chicken 1 Less than 10 unitsMonkey 2 Less than 10 unitsDog 3 Less than 10 unitsCattle (pool) 1 Less than 10 units

Of interest are the somewhat higher values forrheumatic children. If these differences are sig-nificant, their relation to the higher plasminogenvalues obtained for rheumatics (Table IX) is of

importance, suggesting the possibility that the in-hibitor might serve a role in relation to plasmino-gen. However, when. one considers the inhibitorand plasminogen values obtained with animalsera, no relationship appears to exist, since theinhibitor level is high while the amount of plas-minogen activatable by streptokinase is negligible.

3. Plasminogen: The deficiencies of the methodfor plasminogen determination have been dis-cussed above. The results obtained are presentedin Table IX.

In spite of the deficiencies of the method, theresults obtained with normal sera have been rea-sonablyuniform, the great majority of them show-ing levels of 320 or 640 units. Retitration of asingle serum, or titrations of successive bleedingsof the same individual show a variation of nomore than one tube in the series. The highervalues obtained with certain of the patients un-der investigation (e.g., tuberculous and carcino-matous) probably represent an increase overnormal, even though the method of titration doesnot give maximal values for any serum. If thesevalues are artifacts and do not represent an in-crease of total plasminogen over normal, it isdifficult to account for the results, since the twosubstances in serum known to influence the re-action, serum inhibitor and antistreptokinase,would both tend to decrease, rather than increasethe plasminogen values.

SUMMARY

A method for purification of streptokinase d*-scribes conditions for obtaining high yields frombacterial cultures. Protamine is 'used to removeimpurities and streptokinase is separted fromdesoxyribonuclease with cold ethanol at 20%oconcentration. The desoxyribonuclease is pre-cipitated with 40-50%o cold ethanol. The yieldof streptokinase is about 25%o and the product soobtained has been used in clinical studies.

Methods are briefly described for the quanti-tative estimation of streptokinase, antistrepto-kinase, serum inhibitor, serum plasminogen, anddesoxyribonuclease. These methods were em-ployed to make a survey study of various sera.There is a marked rise in antistreptokinase ac-tivity following intrapleural injection of strepto-kinase in human beings. Serum from the normal

170

STREPTOKINASEAND STREPTOCOCCALDESOXYRIBONUCLEASE

adult contains more antistreptokinase than serumfrom the infant and the level is strikingly high inrheumatic fever. In various species the concen-tration decreases in the following order: dog,monkey, man, chicken, cow and horse. Seruminhibitor levels show elevations in wasting dis-eases, and in the animals studied the guinea pighas an exceptionally high inhibitor titer. Themethod for plasminogen determination presentsdifficulties but it appears that there may be amarked increase in patients with tuberculosis.The concentration in chicken, monkey, dog, andcattle appears to be low.

Serum can be heated at 560 C for 30 minutesto destroy serum inhibitor but antistreptokinaseis not destroyed.

BIBLIOGRAPHY

1. Tillett, W. S., and Sherry, Sol, The effect in patientsof streptococcal fibrinolysin (streptokinase) andstreptococcal desoxyribonuclease on fibrinous, puru-lent and sanguinous pleural exudations. J. Clin.Invest., 1949, 28, 173.

2. Christensen, L. R., Streptococcal fibrinolysis: A pro-teolytic reaction due to a serum enzyme activatedby streptococcal fibrinolysin. J. Gen. Physiol.,1945, 28, 363.

3. Christensen, L. R., Protamine purification of strep-tokinase and effect of pH and temperature onreversible inactivation. J. Gen. Physiol., 1947,30, 465.

4. Pillemer, Louis, and Hutchinson, M. C., The deter-mination of the albumin and globulin contents ofhuman serum by methanol precipitation. J. Biol.Chem., 1945, 158, 299.

5. Sherry, Sol, Tillett, William S., and Christensen,L. R., Presence and significance of desoxyribosenucleoprotein in the purulent pleural exudates ofpatients. Proc. Exper. Biol. & Med., 1948, 68, 179.

6. Tillett, W. S., Sherry, Sol, and Christensen, L. R.,Streptococcal desoxyribonuclease: Significance inlysis of purulent exudates and production by strainsof hemolytic streptococci. Ibid., 184.

7. McCarty, Maclyn, The occurrence of nucleases inculture filtrates of group A hemolytic streptococci.J. Exper. Med., 1948, 88, 181.

8. McCarty, Maclyn, Purification and properties of de-soxyribonuclease isolated from beef pancreas. J.Exper. Med., 1946, 29, 123.

9. Edsall, John T., The plasma proteins and theirfractionation, in Advances in Protein Chemistry,Vol. III, edited by Anson, M. L., and Edsall, JohnT. Academic Press, Inc., New York, 1947.

10. Milstone, H., A factor in normal blood which par-ticipates in streptococcal fibrinolysis. J. Immunol.,1941, 42. 109.

11. Oncley, J. L., Melin, M., Richert, D. A., Cameron,J. W., and Gross, P. M., Jr. In press.

12. Cohn, E. J., Strong, L. E., Hughes, W. L., Jr.,Mulford, D. W., Ashworth, J. N., Melin, M., andTaylor, H. L., Preparation and properties ofserum and plasma proteins. IV. A system for theseparation into fractions of the protein and lipo-protein components of biological tissues and fluids.J. Am. Chem. Soc., 1946, 68, 459.

13. Kunitz, M., Crystalline soybean trypsin inhibitor. J.Gen. Physiol., 1946, 29, 149; ibid., 30, 291.

14. Christensen, L. R., and MacLeod, Colin M., A pro-teolytic enzyme of serum: characterization, activa-tion, and reaction with inhibitors. J. Gen. Physiol.,1945, 28, 559.

15. Garner, R. L., and Tillett, W. S., Biochemical stud-ies on the fibrinolytic activity of hemolytic strep-tococci: I, II. J. Exper. Med., 1934, 60, 239, 255.

16. Christensen, L. R., Quantitative determination ofactivity of streptococcal fibrinolysin. Proc. Exper.Biol. & Med., 1941, 46, 674.

17. Holmberg, C. G., and Winbald, S., Purification andconcentration of streptococcal fibrinolysin frombroth cultures. Path. et Microbiol. Scand., 1944,21, 833.

18. Kaplan, Melvin H., Studies of streptococcal fibrinoly-sis: II. The inhibition of streptococcal fibrinolysisby antifibrinolysin and antiprotease. J. Clin.Invest., 1946, 24, 337.

19. Anderson, H. C., Kunkel, H. G., and McCarty,Maclyn, Quantitative antistreptokinase studies, etc.J. Clin. Invest., 1948, 27, 425.

20. Massell, B. F., Mote, J. R., and Jones, T. D., Thequantitative relation of fibrinolysin and antifibri-nolysin. J. Immunol., 1939, 36, 45.

21. Grob, D., The antiproteolytic activity of serum. I.The nature and experimental variation of the anti-proteolytic activity of serum. J. Gen. Physiol.,1943, 26, 405.

22. Guest, M. M., Ware, A. B., and Seegers, W. H., Aquantitative study of antifibrinolysin in chickplasma: Increase in antifibrinolysin activity duringpteroylglutamic acid deficiency. Am. J. Physiol.,1947, 150, 661.

23. Croxatto, H., and Croxatto, R., Effect of chloroformon the antitryptic activity of blood plasma. Proc.Soc. Exper. Biol. & Med., 1947, 66, 191.

24. Schmitz, A., Uber den Trypsin-Inhibitor des Blutes.Ztschr. f. physiol. Chem., 1938, 255, 234.

25. Christensen, L. R., The activation of plasminogen bychloroform. J. Gen. Physiol., 1946, 30, 149.

26. Edsall, John T., personal communication.27. Ferguson, J. H., Assay of tryptases by lysis of fibri-

nogen. Proc. Soc. Exper. Biol. & Med., 1943, 52,243.

28. Christensen, L. R., The mechanism of streptococcalfibrinolysis. J. Bact., 1944, 47, 471.

29. Holmberg, C. G., Studies on the splitting of fibrinunder the influence of fibrinolysin from hemolytic

171

L. R. CHRISTENSEN

streptococci. Arkiv fur Kemi, Minerologie o.

Geologic, 1943, 17, 1.30. Kaplan, M. H., Studies of streptococcal fibrinolysis.

I. The dissimilarity of serum protease and trypsinas indicated by the separate specificities of theirkinases, fibrinolysin and enterokinase. J. Clin.Invest., 1946, 25, 331.

31. Astrup, T., 1948, personal communication.

32. Mirsky, A. E., and Pollister, A. W., Chromosin, a

desoxyribose nucleoprotein complex of the cellnucleus. J. Gen. Physiol., 1946, 30, 117.

33. Tillett, W. S., The fibrinolytic activity of hemolyticstreptococci. Bact. Rev., 1938, 2, 161.

34. Mote, J. R., and Jones, T. D., Studies of hemolyticstreptococcal antibodies in control groups, rheu-matic fever, and rheumatoid arthritis, I, II andIII. J. Immunol., 1941, 41, 35, 61, 87.

35. Boisvert, Paul L., The streptococcal antifibrinolysintest in clinical use. J. Clin. Invest., 1940, 19, 65.

36. Commission of Acute Respiratory Diseases, Studiesof streptococcal fibrinolysis. IV. Clinical applica-tion of a quantitative antifibrinolysin test. J. Clin.Invest., 1946, 25, 352.

172