STUDIES OF STREPTOCOCCAL FIBRINOLYSIS.II. THE INHIBITION OF STREPTOCOCCALFIBRINOLYSIS BY ANTIFIBRINOLYSIN ANDANTIPROTEASE

Melvin H. Kaplan

J Clin Invest. 1946;25(3):337-346. https://doi.org/10.1172/JCI101714.

Research Article

Find the latest version:

http://jci.me/101714-pdf

STUDIES OF STREPTOCOCCALFIBRINOLYSIS. II. THE IN-HIBITION OF STREPTOCOCCALFIBRINOLYSIS BY

ANTIFIBRINOLYSIN ANDANTIPROTEASE1

By MELVIN H. KAPLAN(From the Respiratory Diseases Commission Laboratory, Regional Station Hospital, Section 2,

Fort Bragg, North Carolina)

(Received for publication September 5, 1945)

The resistance of human plasma clots to dissolu-tion by fibrinolysin is generally considered an im-munological response to infection by ,B-hemolyticstreptococci (1, 2). Consequently, the develop-ment of resistant clots during convalescence hasbeen used as a diagnostic test for streptococcaldisease. However, clot resistance has also beenobserved in individuals where no evidence of strep-tococcal infection could be recognized. Positiveantifibrinolysin tests, for example, have been re-ported in the following instances: 15 per cent ofnormal individuals (3, 4); in a number of un-related disease states, as lobar pneumonia, gono-coccal arthritis, typhoid fever, and acute nephritis(5 to 7); in the newborn (8, 6); and in variousanimal species (9, 1). These findings, in turn,have placed some doubt on the specificity of theantifibrinolysin response.

Recent advances in the study of the mechanismof streptococcal fibrinolysis, however, have madeit necessary to review these non-specific cases ofantifibrinolysis in the light of the new findings.It has been shown (10 to 12) that the lytic factor(13) which participates in streptococcal fibrinoly-sis is the precursor of a proteolytic enzyme inplasma, which is activated by streptococcal fibri-nolysin. Fibrinolysis results from the proteolyticaction of this activated enzyme. Thus two stagesare involved in the process:

(1) Conversion of the lytic factor by fibrinoly-sin into an active protease.

1 This investigation was supported through the Com-mission on Acute Respiratory Diseases, Board for theInvestigation and Control of Influenza and Other Epi-demic Diseases in the Army, Preventive Medicine Serv-ice, Office of The Surgeon General, United States Army,and by grants from the Commonwealth Fund, the W. KKellogg Foundation, the John and Mary R. Markle Foun-dation and the International Health Division of the Rocke-feller Foundation to the Board for the Investigation andControl of Influenza and Other Epidemic Diseases forthe Commission on Acute Respiratory Diseases.

(2) Dissolution of the fibrin clot by the pro-tease.

As was indicated in a preliminary report (12),the inhibition of either of these two stages wouldresult in the apparent resistance of the clot tofibrinolysin. Accordingly, the resistance of aplasma clot to lysis might result from: (a) thepresence in plasma of antifibrinolysin; (b) thepresence of antiprotease; or (c) the deficiency ofan effective lytic factor.

Deficiency of a lytic factor has been reported inthe resistant plasmas of the newborn (14) andof the rabbit (13). The present study is primarilyconcerned with the role and properties of anti-fibrinolysin and antiprotease.

MATERIALS ANDMETHODS

1. Fibrinolysin. Fibrinolysin was prepared by alcoholicprecipitation according to the procedure of Gamer andTillett (15), and stored in the dried or frozen state aspreviously described (16). The source of the fibrinolysinwas a highly active strain of group A P-hemolytic strep-tococcus (no. 98) isolated from the blood of a patientwith erysipelas.

2. Fibrinogen. The product employed was Fraction I(17) of the plasma proteins.2 Clots formed from thisfibrinogen preparation were readily susceptible to fibrinoly-sis, indicating that the lytic factor was present in adequateamounts. The fibrinogen solution used contained 0.60grams of Fraction I in 100 ml. of buffered saline.

3. Serum protease. Preparation of serum protease wasaccomplished- by chloroform treatment of the componentof serun which is precipitated by dilution and acidificationto pH 5.5. This method has been previously described(18). In addition, a component of Fraction III-2 (17)of the plasma proteins 2 was found to be a highly active

2 The products of plasma fractionation employed in thiswork were obtained through the courtesy of Drs. E. J.Cohn, J. T. Edsall, and S. Howard Armstrong, Jr. Theywere developed from blood, collected by the AmericanRed Cross, by the Department of Physical Chemistry,Harvard Medical School, Boston, Mass., under a contract,recommended by the Committee on Medical Research, be-tween the Office of Scientific Research and Developmentand Harvard University.

337

MELVIN H. KAPLAN

source of the protease, and was also employed. Bothpreparations were equally satisfactory.

4. Thrombin. The thrombin preparation wasp a 1:10dilution of a commercial product 3 prepared from rabbitblood.

5. Buffered saline. The buffered saline employed was asolution containing 0.01M phosphate in 0.85 per cent so-dium chloride at pH 7.4.

6. Standard fibrin clot. The standard clot was formedfrom 0.5 ml. of fibrinogen and 0.2 ml. of thrombin madeup to a final volume of 2.2 ml.

7. Assay of serum antiprotease. The titer of serum anti-protease was measured by the highest dilution of serumwhich neutralized the clot-lysing activity of a standardsolution of serum protease. The standard solution ofserum protease adopted was that dilution which permitted0.5 ml. to lyse completely the standard fibrin clot in 30minutes at 370 C. The procedure for the estimation ofserum antiprotease was as follows:

One ml. amounts of twofold dilutions of serum beginningwith %were incubated with 0.5 ml. of standard proteasesolution for 30 minutes at 37° C. The enzyme-serum mix-tures were then incorporated in fibrin clots by the fur-ther addition of 0.5 ml. of fibrinogen and 0.2 ml. of throm-bin. Incubation at 37° C. was carried out for 60 minutes.The reciprocal of the highest serum dilution which com-pletely prevented lysis was taken as the titer of anti-protease.

8. Assay of serum antifibrinolysin. The procedure wassimilar to that described for the estimation of antipro-tease. The titer of serum antifibrinolysin was taken asthe highest dilution of serum which neutralized the fibri-nolytic activity of a standard solution of fibrinolysin.The standard fibrinolysin solution was that concentrationwhich permitted 0.5 ml. to lyse the standard fibrin clotin 30 minutes at 370 C. One ml. amounts of twofolddilutions of serum beginning with %were incubated for30 minutes at 370 C. with 0.5 ml. of fibrinolysin. Toeach tube was added 0.5 ml. of fibrinogen and 0.2 ml. ofthrombin. After incubation for 60 minutes, the tubes wereread for the highest serum dilution completely preventinglysis.

It will be observed that in the methods employed forthe estimation of antiprotease and antifibrinolysin, theunit of lytic activity was the same.

9. Antifibrinolytic sera were obtained from convalescentcases of acute streptococcal pharyngitis or tonsillitis. Thediagnosis was confirmed by the presence of P-hemolyticstreptococcus on throat culture, and by the development ofanother streptococcal antibody, antistreptolysin "O", dur-ing convalescence (19).

A. The fibrinolysin-antifibrinolysin reaction

Although the antifibrinolysin test has been foundoccasionally to be non-specific, it is generally ac-cepted that the clot resistance developed in in-

3 "Hemostatic Globulin," Lederle Laboratories, Inc.

dividuals convalescent from streptococcal diseaseis a specific immunological response to the in-fection (2). While it has not been possible toelicit antifibrinolysin by animal inoculation (20),the immunological nature of the antifibrinolytic re-sponse in man is supported by the observation thatthe development of fibrinolytic resistance is cor-related with the formation of other streptococcalantibodies, such as antistreptolysin "O" (21, 22,8).

It has recently been observed that the level ofantifibrinolysin in serum was also correlated withthe capacity of the serum to inhibit trypsin (23).This observation seemed to contradict the biphasicnature of the fibrinolytic reaction, which suggeststhat antifibrinolysin and antiprotease representseparate activities. The relationship of anti-fibrinolysin to the fibrinolytic mechanism wastherefore investigated. In addition, observationswere extended to the quantitative aspects of thereaction between fibrinolysin and antifibrolysin.

1. The relationship of antifibrinolysin to thefibrinolytic mechanism

Since streptococcal fibrinolysis has been shownto be a two-stage process, it was necessary to de-.termine first whether antifibrinolysin acted byneutralizing fibrinolysin or by blocking the pro-tease system. This problem was approached bycomparing the antifibrinolytic effects observedwhen a given convalescent serum was incubatedwith lytic factor and when the same serum wasincubated with fibrinolysin. If the serum acted onfibrinolysin, it seemed reasonable that incubatingthe serum with the lytic factor in the fibrinogenpreparation would have no appreciable effect onthe resistance of the clot; whereas, incubating theserum with fibrinolysin would result in an in-crease in resistance. The respective capacities ofthe serum to inhibit the lytic factor and fibrinoly-sin were quantitated by testing the serum in pro-gressive twofold dilutions. The inhibitory effectin each case was measured by the highest dilutionof the serum which completely prevented dissolu-tion.

The procedure employed was as follows: One ml.amounts of twofold dilutions of serum were set up intriplicate. To the first set of dilutions were quicklyadded 0.5 ml. amounts of fibrinolysin, 0.5 ml. of fibri-nogen, and 0.2 ml. of thrombin, without prior incubation.

338

II. INHIBITION OF STREPTOCOCCALFIBRINOLYSIS

To the second set of serum dilutions, 0.5 ml. amounts offibrinogen solution (lytic factor) were added and incuba-tion carried out for 15 minutes at 370 C. To these mix-tures were next added 0.5 ml. of fibrinolysin and 0.2 ml.of thrombin. To each tube of the third series of serumdilutions, 0.5 ml. amounts of fibrinolysin were added andincubation of these mixtures carried out for 15 minutesat 370 C. To each mixture was then added 0.5 ml. offibrinogen and 02 ml. of thrombin. The tube of highestserum dilution in which no lysis occurred after one hourat 37° C. was taken as the titer of the serum.

It may be observed from Table I that incubat-ing the serum with the lytic factor did not resultin a significant increase in antifibrinolytic titer.However, when incubation was carried out withfibrinolysin, the titer of the serum was increasedfrom 8 to 256. This marked increase in the anti-fibrinolytic capacity of the serum which resultedfrom incubating previously with fibrinolysin indi-cated that the serum acted specifically on fibri-nolysin.

TABLE I

Effect of incubating serum with fibrinolysin and withlytic factor on the antifibrinolytic titer

Titer of serum* Titer of serum after Titer of serum afterwithout incubation incubation with incubation withwithouincuation lytic factor fibrinolysi

8 16 256

$ Reciprocal of highest dilution of serum which com-pletely prevented lysis.

Subsequent comparative studies of the levelsof antifibrinolysin and antiprotease in the serafrom a large number of individuals, to be pre-sented in the following sections, have similarlyindicated that antifibrinolysin has no effect on theprotease system. The evidence thus indicatedthat the factor responsible for the resistance ofplasma clots to lysis in cases of hemolytic strep-tococcal infection does not affect the serum pro-tease system, but reacts specifically with fibri-nolysin.

2. Quantitative aspects of the fibrinolysin-anti-fibrinolysin reaction

The observation that an increase in antifibri-nolytic titer resulted when a prior incubation ofthe serum with fibrinolysin was carried out sug-gested that neutralization did not occur instan-taneously. It had been observed previously thatplasma clots frequently became more resistant

when the plasma sample was first incubated withfibrinolysin (24). While this increased resistancewas pointed out as a factor which conditioned theresults obtained with the plasma antifibrinolysintest, further investigation of this effect was notcarried out. In the present experiment an at-tempt was made to measure the rate of reactionbetween fibrinolysin and antifibrinolysin.

The effect of incubation was first studied bymeasuring the lytic activity of mixtures of fibri-nolysin and serum which had been incubated pre-viously at 37° C. for various intervals of time.Lytic activity was measured by the time of disso-lution of a standard fibrin clot in which the reac-tion mixtures were incorporated.

The procedure was as follows: Mixtures of 1.0 ml. ofa serum diluted to 1/200 and 0.5 ml. of fibrinolysin wereincubated for 0, 15, 30, and 60 minutes at 370 C. At theend of each period, 0.5 ml. of fibrinogen and 0.2 ml. ofthrombin were added. The lysis time of each clot wasthen measured at 370 C. Lysis time was taken as theperiod between clotting and complete dissolution of thefibrin clot.

The progressive prolongation of the lysis timeresulting from such incubation is demonstrated inTable II. Without prior incubation, the lytic

TABLE II

Prolongation of the lysis time produced by incubatingfor varying periods constant amounts of

fibrinolysin and antifibrinolysin

Lysis time produced afterFbi-Anti- incubating for:

Foilbyrsin- fibrino- Saline inuatn for:lysn in*_ __-__

0 min. 15 min. 30 min. 60 min.

ml. ml. ml. min. min. min. min.0.5 0 1.0 10 10 10 110.5 0.25 0.75 10 15 17 180.5 0.5 0.5 10 23 30 410.5 1.0 0 10 64 90 >120

* The antifibrinolysin solution was a 1/200 dilution of aconvalescent serum.

activity of the fibrinolysin-serum mixtures wasidentical with that of the fibrinolysin-saline con-trol, indicating that no union of fibrinolysin andantifibrinolysin occurred. As the period of in-cubation was lengthened, however, the proportionof fibrinolysin neutralized appeared correspond-ingly increased.

In order to subject the reaction to study inquantitative terms, it was necessary first to devise

339

MELVIN H. KAPLAN

an accurate method of measuring fibrinolysin con-centration. It was found that fibrinolysin concen-tration could be satisfactorily estimated by meansof lysis time if a standard fibrin clot of constantand adequate lytic factor content were employed,and if such measurements were made within arange in which the relationship between lysis timeand fibrinolysin concentration most nearly ap-proached linearity (25). A suitable range of pro-portionality was observed with lysis times between10 and 60 minutes.

Accordingly, by means of a series of precisetitrations, a reference curve was plotted whichrelated lysis time to fibrinolysin concentration(Figure 1). It was found convenient to express

Flo10 111F1111111z 1111111 llll

0 1\1111111111h l lllllllT

>~~~~~~~~~~~6-_Vlf0-rTr Tlrr

into terms of relative fibrinolysin concentration in units,by reading directly from the reference fibrinolysin curve.Deducting the residual units of activity from the 8.0 unitsof fibrinolysin originally present yielded the amount offibrinolysin bound.

In Figure 2 are given the curves for the rate ofreaction between fibrinolysin and antifibrinolysin.The rate of combination is exceedingly rapid dur-ing the initial phase of the reaction, particularlyduring the first 10-minute interval. Subsequently,the reaction proceeds at a progressively diminish-ing rate, and is essentially complete at the end of30 minutes. Equilibrium is reached in approxi-mately 60 minutes.

z 6

yZ4.

38.2.

I.P

5 10 5 20 25 30 35 40 4 50 65 60LYSIS TIME (Ut*4

FIG. 1. STANDARDFIBRINOLYSIN CURVE

fibrinolysin concentration in arbitrary units ofactivity. One unit of fibrinolysin was that amounteffecting lysis of the standard fibrin clot in 60minutes. By means of the reference curve, thelysis time effected by a mixture of fibrinolysin andantifibrinolysin could be converted into terms ofresidual units of fibrinolysin.

The rate of neutralization was then studied bythe following procedure:

To constant amounts of a 1/200 dilution of convalescentserum contained in 1.0 ml. was added 0.5 ml. of fibrinoly-sin equivalent to 8.0 units of activity. Incubation wascarried out for varying periods from 5 minutes to 90 min-utes in a water bath at 370 C. At the end of a givenperiod of incubation, 0.5 ml. of fibrinogen and 0.2 ml. ofthrombin were added, the clotting time noted, and thetime required for dissolution of the clot accurately deter-mined. The lysis time was then a measure of the residualactivity of the fibrinolysin-antifibrinolysin mixture afterthe given period of incubation. Lysis time was converted

5 i0 15 202530354045 5o5560657075 808590TIIME OF INCUBATION (miu=t.

FIG. 2. RATE OF COMBINATION OF FIBRINOLYSIN ANDANTIFIBRINOLYSIN

It will be noted from Figure 2 that the amountof fibrinolysin finally bound was not related stoi-chiometrically to the quantity of antifibrinolysinadded, and consequently could not be reduced toterms of equivalence with a given amount ofserum. For example, while 0.5 ml. of antifibri-nolysin neutralized 7.0 units of fibrinolysin, theaddition of twice as much, or 1.0 ml., neutralizedonly 7.3 units. These observations then indicatedthat the amount of fibrinolysin bound was de-pendent upon the relative proportion of fibrinoly-sin and antifibrinolysin in the reaction mixture.

This fact is clearly demonstrated in Figure 3,in which is given the amount of fibrinolysin finallybound after the addition of varying quantities ofantibody to 8.0 units of fibrinolysin. As the anti-fibrinolysin concentration was increased, each suc-cessive increment combined with progressivelysmaller amounts of fibrinolysin. It was thereforeconcluded that fibrinolysin and antifibrinolysin

340

I

II. INHIBITION OF STREPTOCOCCALFIBRINOLYSIS

a

z I

2'z,.

o-

L-

: :.

IL

0)4z

o.

1 T I ILU I I I I II I I T I I1 1 [ I[

1 1 I T T ILL _LIII8- I I II1 1 1 II 17 1 1x

-t 'IOfF rT rlt-1lTIliU ElI

iL?f II I II2 -1 [ll1

pw

,. I111110.2 0.4 0.6 0.8 1.0

ML. ANTIFIBRINOLYSINADDEDTO &0 UNITS FIBRINOLYSIN

FIG. 3. AMOUNTOF FIBRINOLYSIN BOUNDAFTER ADDINGVARYING AMOUNTSOF ANTIFIBRINOLYSIN TO 8.0 UNITSOF FIBRINOLYSIN

combine in varying proportions which are de-pendent upon the ratio of the concentrations ofthe two substances. The reaction is thus similarto toxin-antitoxin reactions in vitro (26).

When the amount of fibrinolysin bound at theend of 90 minutes was taken as the end-point ofthe reaction, it was found that after 15 minutesincubation, the reaction was 83 to 96 per centcomplete, depending upon the amount of anti-fibrinolysin present; and at the end of 30 minutes,the reaction was 95 to 98 per cent complete.

B. The inhibition of streptococcal fibrinolysisby serum antiprotease

It has been demonstrated that the resistance ofplasma clots to lysis in patients recovering fromhemolytic streptococcal infections is due to thepresence of antifibrinolysin, an antibody specifi-cally neutralizing the action of fibrinolysin. Fromtheoretical considerations of the fibrinolytic mech-anism, it has been suggested that the inhibition offibrinolysis may also result from the presence inplasma of antiprotease.

The nature and physiological significance ofserum antiprotease, sometimes termed serum anti-trypsin, has been reviewed recently (27). Anti-proteolytic activity appears to be a constant prop-erty of human and animal sera; however the sub-stance or substances in serum responsible for thisproperty have as yet not been identified. Themost widely held hypothesis is that antiprotease

is a product of protein hydrolysis. It has beenpointed out, for example, that the antitrypsin ofnormal serum is similar in properties to the tryp-sin-inhibitor produced during the tryptic digestionof proteins (28). Moreover, the antiproteolyticactivity of serum appears to be causally relatedto the products of protein hydrolysis formed bothin the intestine and parenterally after the oral andintramuscular administration of trypsin (27).These findings are also supported by evidencefrom pathological studies. Thus the serum anti-protease level is increased after meals, in patientswith malignant tumors, in tuberculosis and hyper-thyroidism, in anaphylaxis, and in many cases ofacute infection, especially if accompanied by highfever (27).

In the present work, investigation was made ofthe relationship of serum antiprotease to the inhi-bition of fibrinolysin. The association of an in-creased antiprotease level with acute infectionssuggested that this factor might be related in somecases to the non-specific antifibrinolytic tests re-ported. Consequently, comparative studies weremade of the antiprotease and antifibrinolysin levelsof normal human sera, of sera from patients withhemolytic streptococcal infections and pneumo-coccal pneumonia, and of sera from various ani-mal species.

1. The relationship of antiprotease to the fibri-nolytic mechanism

Previous studies have shown that fibrinolysinmay be inhibited by animal and plant antitrypsins(23). These observations were regarded as indi-cating that fibrinolysin and trypsin were similarin properties. However, since streptococcal fibri-nolysis has been shown to be a two-stage reaction,it seemed more likely that the inhibition observedwas due to the blocking of the protease system.This assumption was verified in the following ex-periment.

Ten sera of varied antiprotease level but of lowantifibrinolysin content were selected, and eachserum tested for its relative capacity to inhibitfibrinolysin and serum protease. Antifibrinolyticand antiproteolytic effect was measured by theprolongation of the lysis time of a standard fibrinclot to which the serum was added. It was neces-sary that the sera contain little or no antifibrinoly-

341

-

MELVIN H. KAPLAN

sin, since possible inhibition of fibrinolysin by theantiprotease would be masked by the more pro-nounced effect of the antifibrinolysin. The serawere not incubated with either fibrinolysin or pro-tease prior to clot formation; reagents were addedin rapid succession and clots immediately formed.

The procedure was as follows:

To 1.0 ml. of a 1/10 dilution of each serum was added0.5 ml. of fibrinolysin, and then quickly, 0.5 ml. of fibrino-gen and 02 ml. of thrombin. To a duplicate series ofsera were added 0.5 ml. amounts of serum protease, andclots formed in the same manner. Incubation was carriedout at 370 C. and the lysis times of the clots noted. Therelative capacity of each serum for inhibiting fibrinolysinwas thus directly compared with its relative ability toinhibit the activity of the serum protease.

TABLE III

Relationship of the antifibrinoly!ic capacities ofsera to their antiprotease content

Lysis time Lysis time Anti- Anti-I Serum no. with with fibrinolysin protease

fibrinolysin protease titer titer

min. min.26 C 20 36 <5 519 B 21 41 <5 1025 B 20 43 5 1016 A 26 46 5 4019 A 49 41 <5 2019C 51 48 <5 2025C 56 51 5 4026 B 56 53 5 4025A 56 51 5 4021 A 81 48 <5 40

Table III demonstrates that these propertieswere directly correlated. Sera possessing elevatedlevels of antiprotease were those maximally re-

sistant to the action of fibrinolysin. Since thetrue antifibrinolysin levels of these sera were allnegligibly low, this resistance could have been dueonly to antiprotease. It was concluded that serum

antiprotease is an important factor in the resist-ance of plasma clots to streptococcal fibrinolysis.

2. The antiprotease levels of normal humansera

Sera from 21 normal individuals known to havehad no recent illness or infection were assayedfor their antiprotease and antifibrinolysin titers bythe methods described. As shown in Table IV,variation of the titer of antiprotease in the normalsubjects was slight; none of the sera possessed an

antiprotease titer greater than. 5. In contrast, theantifibrinolysin titers ranged from < 5 to 80. Ap-

TABLE IV

The antiprotease and antifibrinolysin titers ofnormal human sera

Serum no. Antiprotease titer Antifibrinolysin titer

618 5 80619 <5 40620 <5 20621 <5 <5623 <5 10624 <5 10625 5 10626 5 20627 <5 5628 <5 <5629 5 <5630 <5 <5634 <5 <5637 5 <5640 <5 5643 5 10644 5 5645 <5 <5647 <5 <5649 <5 <5650 <5 <5

proximately 70 per cent of the sera possessed anantifibrinolysin titer of less than 10.

No correlation was observed between the anti-fibrinolysin and antiprotease content of these nor-mal sera. However, the considerable variation ofthe antifibrinolysin titer was significant. Thisobservation provided suggestive evidence that theresistant plasma clots reported previously (3, 4)with normal individuals might have been due tothe presence in these cases of significant levelsof antifibrinolysin.

3. Comparison of the antifibrinolysin and anti-protease titers from patients with hemolyticstreptococcal infection

In a recent study of the antifibrinolysin test, itwas reported that sera having high antifibrinolysintiters also possessed high antitrypsin titers (23).However, studies of the mechanism of the fibrino-lytic reaction and its inhibition have thus far indi-cated that antifibrinolysin is -a specific antibodywhose r6le in the antifibrinolytic phenomenon isseparate and distinct from that of antiprotease.Observations were further extended to sera fromknown streptococcal infections.

The antifibrinolysin titers and antiproteasetiters of 65 acute and convalescent sera from 21cases of hemolytic streptococcal pharyngitis ortonsillitis were compared. Hemolytic streptococ-cal infection was confirmed by bacteriological and

342

II. INHIBITION OF STREPTOCOCCALFIBRINOLYSIS

clinical evidence and by the development of strep-tococcal antibodies (antistreptolysin "O") duringconvalescence (19).

TABLE V

Comparison of the antiprotease and antifibrinolysin titersof sera from cases of hemolytic streptococcal infection

(pharyngitis or tonsiUlitis)

Co.Date ofAnti- Anti- Anti-

Cas no. Daerof streptolysin protease fibrinolysinsrm "0" titer titer titer

E-185

E-284

E-361

T-262

T-270

T-202

D-184

E425

D-226

T-229

3-224-134-28

3-275-116-15

4-24-235-11

11-2112-1012-2712-31

11-2111-2812-912-31

11-1911-2812-10

1-3

12-181-51-25

4-105-45-18

12-301-202-10

11-1911-2812-912-31

400500500

83.3250250

125500500

159159250833

159200

20001585

159317500400

83.3159200

20012501250

200625625

62.5159400400

101010

<101020

S1010

20201010

105

105

<5<5

510

<5<5<5

102020

55

<5

<510

520

The results are given in Table V.streptolysin titers are also given for thof establishing the validity of the strept(fection. The antiprotease titers of tranged from < 5 to 20, while the antifititers extended from < 50 to 1400. Ntion of antifibrinolysin with serum arwas observed, either in terms of relaabsolute titers. Moreover, in no patiirise in antiprotease found which was ixcommensurate with the rise in antifibrin

These observations confirm the hypothesis thatantifibrinolysin and antiprotease are distinct sub-stances, each of which, by a different mode ofaction, is capable of effecting resistance to fibri-nolysin. In cases of hemolytic streptococcal infec-tion, this increased resistance is apparently due toa specific immunological response whereby anti-fibrinolysin is elaborated. No correspondingchanges in the serum antiprotease Ievel were

loo found to occur at the same time.10001000 4. The antifibrinolytic properties of sera from

200 cases of pneumococcal pneumonia500 There is some evidence that the changes which

150 occur in the blood during the acute phase of in-600 fection may be responsible for the non-specific600 positive antifibrinolytic tests reported. Such posi-150 tive tests (6, 8) have been observed in the acute150 febrile phase of influenza, poliomyelitis, typhoid300

loo fever, bacillary dysentery, and lobar pneumonia.50 Although these studies have been largely con-

100 fined to patients in the pediatric age group, simi-250 lar observations have been made in some adults,250 particularly those with gonococcal arthritis (5),400 infections due to streptococcus viridans (4, 5),600

lo0 and lobar pneumonia (3, 7). The resistance pro-600 duced in these cases did not appear to be an im-300 munological response to the infection, since maxi-80 mal resistance occurred in the acute phase of the1000

disease and then rapidly fell during convalescence.1400 That these effects might be due to a rise of anti-1400 protease is suggested by the fact that antiproteases

150 are known to be elaborated in the blood during400 acute infection (29, 27).400 Consequently, a study was made of the anti-

50 protease levels of acute phase and convalescent500 sera from 25 cases of lobar pneumonia. The cases500 available for study were soldiers with illnesses of

moderate severity.The anti- Eleven of the 25 cases of lobar pneumonia stud-

Le purpose ied demonstrated antiprotease titers of 20 orDcoccal in- greater in either the acute or convalescent sera.these sera The antiprotease levels of these 11 patients areibrinolysin given in Table VI. The level of antiprotease inJo correla- these cases may be considered appreciably ele-itiprotease vated as compared with the titer of 5 obtainedtive or of with normal sera. Four patients showed a four-ent was a fold or greater fall in titer from the acute to thei the least convalescent phase. The evidence suggests thattolysin. antiprotease may be the factor responsible for the

343

MELVIN H. KAPLAN

TABLE VI

Comparison of the antiprotease and antifibrinolysintiters of selected sera from patients with

pneumococcal pxeumonxi

Antiprotease titer Antifibrinolysin titerSerum no.

Acute Convalescent Acute Convalescent

10 20 20 >640 >64012 20 5 40 4016 40 10 5 <517 40 20 320 32018 20 10 80 8019 20 10 <5 <521 40 10 <5 522 80 40 40 8024 80 80 40 2025 40 40 5 526 40 5 5 <5

resistant clots of some patients with pneumococcalpneumonia.

Those sera containing an elevated antiproteaselevel were also tested for their antifibrinolysintiters. As indicated by Table VI, no correlationbetween the titer of antiprotease and the titer ofantifibrinolysin was obtained. It might be ex-pected from the nature of the fibrinolytic processthat the sera which contained appreciable anti-protease titers would show an increased degree ofresistance to fibrinolysin. This was not found tobe the case. The acute serum of No. 21, for ex-ample, which possessed an antiprotease level of40, had no higher antifibrinolysin titer than theconvalescent serum, the-antiprotease level of whichwas 10. It would appear that the present methodof antifibrinolysin titration was not readily af-fected by the levels of antiprotease in the serastudied.

A possible explanation for this is that theserological method does not permit ready unionof the antiprotease with lytic factor before activa-tion is effected; nor does it readily permit unionof the inhibitor with the activated protease beforethe fibrin clot is attacked. Since the end-point inthe test is based on lysis-prevention, it will be seenthat antiprotease may not be able to prevent fibri-nolysis completely unless present in high concen-tration.

5. The inhibition of fibrinolysin by the sera ofvarious species

The resistance of plasma clots of various animalspecies to the action of fibrinolysin was first ac-counted for by the hypothesis that fibrinolytic

streptococci were species-specific; that is, that agiven streptococcus was able to lyse only the fibrinof the species from which it was isolated (9, 2).Subsequently it was found that the resistance ofrabbit plasma clots could be attributed to the de-ficiency of an effective lytic factor (13), and per-haps also to the presence of an inhibitor (10).Other studies indicated that human fibrin of allthe species tested was most susceptible to fibrinoly-sis, regardless of the source of strain of strepto-coccus (30).

These observations suggested that the differencein susceptibility between human and animal plasmaclots was, in large measure, due to factors in theplasma itself rather than to the fibrin specificityof the streptococcus. In man it has been demon-strated that resistance may result from at leastthree plasma factors: antifibrinolysin, antiprotease,or a deficient lytic factor. It seemed likely thatthese same factors, with the possible exception ofantifibrinolysin, were responsible for the observedresistance of animal plasma clots. In the presentstudy the r6le of antiprotease was examined.

The sera of various animal species including therabbit, guinea pig, horse, cow, and swine wereexamined for their capacity to inhibit the actionof fibrinolysin on human fibrin clots. This prop-erty was compared with the capacity of the samesera to inhibit human serum protease. Inhibitionof fibrinolysin was measured by the lysis time ofa fibrin clot to which was added 0.2 ml. of serumand 0.5 ml. of fibrinolysin. This amount of fibri-nolysin regularly lysed the fibrin clot in 15 min-utes. Antiprotease was determined by serum titra-tion according to the method previously described.

As shown in Table VII, addition of the animalsera rendered the fibrin clot completely resistantto the action of fibrinolysin. In contrast, the ad-dition of normal human serum to the fibrin clotaffected the fibrinolytic rate only slightly. On ex-amination of their antiprotease content, the animalsera all possessed titers of 80 or greater, in strikingcontrast to the level of 5 obtained with normal hu-man serum. Since these sera were incorporatedin fibrin clots which contained adequate amountsof lytic factor, the resistance could be attributedonly to inhibiting substances in the added sera.It was concluded that antiprotease is an importantfactor in the resistance of animal clots to fibrinoly-sin.

344

II. INHIBITION OF STREPTOCOCCALFIBRINOLYSIS

TABLE VII

The inhibition of fibrinolysin and serum proteaseby the scra of various species

Species Lysis time with Antiproteasefibrinolysin titer

Rabbit >24 hrs. 80Guinea Pig >24 hrs. 160Horse >24 hrs. 160Cow >24 hrs. >160Swine >24 hrs. >160Human 20 min. 5Control (Saline) 15 min. 0

DISCUSSION

From theoretical considerations of the mecha-nism of streptococcal fibrinolysis, it has been sug-

gested that the resistance of plasma clots to theaction of fibrinolysin may result:

(1) From the inhibition of fibrinolysin.(2) From the inhibition of the serum protease

system.(3) From a quantitative deficiency of the lytic

factor.The results of the various studies made on fibri-

nolysin inhibition in streptococcal infections, inpneumococcal pneumonia, and in the animal spe-

cies have furthered this hypothesis.In streptococcal infection, the inhibition of fibri-

nolysin is due to antifibrinolysin, a specific anti-body presumably elaborated as an immunologicalresponse to the infection. Antifibrinolysin doesnot enter into any other phase of the fibrinolyticmechanism. The true antifibrinolysin. level of a

given serum is completely independent of its con-

tent of antiprotease.These observations were further confirmed in

studies of the antifibrinolytic properties of sera

from patients with pneumococcal pneumonia. Theabsence in cultures of pneumococcus of a possibleantigenic constituent similar in action to fibri-nolysin (2), as well as the non-immunologicalcharacter of the antifibrinolytic response duringthe disease, indicated that here the basis for theinhibition of fibrinolysin was completely differentfrom that in streptococcal infection. It was ofmarked significance, therefore, that in some ofthe cases of pneumococcal pneumonia. studied, theantiprotease level tended to be elevated particu-larly during the acute stage of the disease.

In a recent review of the relationship of theserum antiprotease level to various disease states

(27), it was reported that antiproteases are liber-ated during the febrile phase of acute infection.In -several cases of lobar pneumonia (31), theantiprotease titer rose markedly just before crisisand then decreased to normal limits within a pe-riod of a week. The present data suggest that thepositive antifibrinolysin tests observed in cases ofpneumococcal pneumonia and other acute febrilediseases may be due in part to the elaboration ofserum antiprotease during the disease processes.The rapid loss of antifibrinolytic properties ob-served in such patients would result from a corres-ponding decrease in antiprotease. Further datarelating the antiprotease level in acute infectionswith the plasma clot antifibrinolysin test aredesirable.

Since the variation of the antiprotease level ofnormal sera is slight, the positive antifibrinolysintests reported with normal individuals are prob-ably not due to this factor. On the other hand,the wide variation of the antifibrinolysin level innormal sera is suggestive evidence that such posi-tive tests may have been due to the presence ofresidual antifibrinolysin elaborated as a resultof streptococcal infection in the past. A criticalanalysis of the occurrence of antifibrinolysin innormal individuals, as well as in the various dis-ease states, has been made possible by the appli-cation of a quantitative method for the estimationof serum antifibrinolysin. The results of this in-vestigation will be reported in a succeeding paper.

SUMMARYAND CONCLUSIONS

The inhibition of streptococcal fibrinolysis byhuman serum may result from the presence ofeither antifibrinolysin or antiprotease.

Antifibrinolysin combines rapidly and specifi-cally with fibrinolysin, and does not affect the ac-tivity of the lytic factor or the serum protease.The reaction between fibrinolysin and antifibri-nolysin is 95 to 98 per cent complete at the end of30 minutes at 370 C. and equilibrium is reachedin approximately 60 minutes. Since the two sub-stances combine in varying multiple proportions,the reaction resembles in vitro toxin-antitoxin re-actions.

Antiprotease prevents streptococcal fibrinolysisby inhibiting the protease system. The antipro-tease content of serum is independent of the levelof antifibrinolysin. No correlation was observed

345

MELVIN H. KAPLAN

either in absolute or relative levels of antiproteaseand antifibrinolysin.

In patients with pneumococcal pneumonia, theantiprotease levels tend to be elevated during theacute phase of the disease. It is suggested thatthe antifibrinolytic response frequently observedin the acute phase of febrile disease is due to thepresence of plasma antiprotease.

The sera of various animal species were foundto possess exceedingly high levels of antiprotease.It is possible that the resistance of animal plasmaclots to the action of fibrinolysin may similarly bedue to inhibition of the protease system.

The author takes pleasure in acknowledging the as-sistance of Dorothy C. Mickle, S/Sgt. Howard E. Duke,S/Sgt. Willard W. Skatrud, T/Sgt. Louis P. Codifer,and Bettie Wooten.

BIBLIOGRAPHY

1. Tillett, W. S., and Gamer, R. L., The fibrinolytic ac-tivity of hemolytic streptococci. J. Exper. Med.,1933, 58, 485.

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

3. Tillett, W. S., Edwards, L. B., and Garner, R. L.,The fibrinolytic activity of hemolytic streptococci;the development of resistance to fibrinolysis follow-ing acute hemolytic streptococcus infections. J.Clin. Invest., 1934, 13, 47.

4. Waaler, E., Development of antifibrinolytic propertiesin blood of patients with rheumatic fever, chronicinfective arthritis and bacterial endocarditis. J.Clin. Invest., 1937, 16, 145.

5. Myers, W. K., Keefer, C. S., and Holmes, W. F., Jr.,The resistance to fibrinolytic activity of the he-molytic streptococcus with special reference to pa-tients with rheumatic fever and rheumatoid arthri-tis. J. Clin. Invest., 1935, 14, 119.

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

7. Waaler, E., Underhelser over de hemolytiske strep-tokokkers fibrinolytiske evne, og over tidstdevae-relsen av antifibrinolysiner i blod. Norsk mag. f.laegevidensk., 1936, 97, 449.

8. Lippard, V. W., and Johnson, P., Beta hemolytic strep-tococcic infection in infancy and in childhood. Am.J. Dis. Child., 1935, 49, 1411.

9. Van Deventer, J. K., and Reich, T., Antihuman fibri-nolytic streptococci. Proc. Soc. Exper. Biol. andMed., 1934, 31, 821.

10. Christensen, L. R., The mechanism of streptococcalfibrinolysis. J. Bact. 1944, 47, 65. Streptococcalfibrinolysis: a proteolytic reaction due to a serumenzyme activated by streptococcal fibrinolysin. J.Gen. Physiol., 1945, 28, 363.

11. Kaplan, M. H., Nature and role of the lytic factor in

hemolytic streptococcal fibrinolysis. Proc. Soc.Exper. Biol. and Med., 1944, 57, 40.

12. Commission on Acute Respiratory Diseases, and Kap-lan, M. H., A quantitative study of the fibrinoly-sin-antifibrinolysin reaction. Science, 1945, 101,120.

13. Milstone, H., A factor in normal human blood whichparticipates in streptococcal fibrinolysis. J. Immu-nol., 1941, 42, 109.

14. Boisvert, P. L., Qualitative differences between posi-tive streptococcal antifibrinolysin tests. Proc. Soc.Exper. Biol. and Med., 1942, 49, 678.

15. Garner, R. L. and Tillett, W. S., Biochemical studieson the fibrinolytic activity of hemolytic streptococci.J. Exper., Med., 1934, 60, 239.

16. 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. In-vest., 1946, 25, 331.

17. Cohn, E. J., Oncley, J. L., Strong, L. E., Hughes, W.L., Jr. and Armstrong, S. H., Jr., The characteriza-tion of the protein fractions of human plasma. J.Clin. Invest., 1944, 23, 417.

18. Tagnon, H. J., Davidson, C. S., and Taylor, F. H. L.,Studies on blood coagulation: a proteolytic enzymeprepared from calcium and platelet-free normal hu-man blood plasma. J. Clin. Invest., 1942, 21, 525.

19. Hodge, B. E., and Swift, H. F., Varying hemolytic andconstant combining capacity of streptolysins,; influ-ence on testing for antistreptolysins. J. Exper.Med., 1933, 58, 277.

20. Van Deventer, J. K., Antigenicity of streptofibrinoly-sin. Proc. Soc. Exper. Biol. and Med., 1935, 33, 17.

21. Stuart-Harris, C. H., Hemolytic streptococcal fibri-nolysis. Brit. J. Exper. Path., 1935, 16, 513.

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

23. Mirsky, I. A., Inhibition of f,-hemolytic streptococcifibrinolysin by trypsin inhibitor (antiprotease). Sci-ence, 1944, 100, 198.

24. Mote, J. R., Massell, B. F., and Jones, T. D., Factorsinfluencing the quantitative relation of fibrinolysinand antifibrinolysin. J. Immunol., 1939, 36, 59.

25. Kaplan, M. H., Unpublished observations.26. Pappenheimer, A. M., Jr., and Robinson, E. S., A

quantitative study of the Ramon diphtheria floccula-tion reaction. J. Immunol., 1937, 32, 291.

27. Grob, D., The antiproteolytic activity of the serum.J. Gen. Physiol., 1943, 26, 405.

28. Hussey, R. G., and Northrop, J. H., Quoted from 27.29. Hort, E. C., Diagnosis of cancer by examination of

the blood. Brit. M. J., 1909, 2, 966.30. LeMar, J. D., and Gunderson, M. F., Studies on strep-

tococcal fibrinolysis. J. Bact., 1940, 39, 717.31. Ascoli, G., and Bezzola, A., Das Verhalten des anti-

tryptischen Vermogens des Blutserums bei der crou-posen Pneumonie. Berl. klin. Woch., 1903, 17, 391.

346