COAGULATION AND LIQUEFACTION OF SEMEN*
PRoTEoLYTIC ENZYMES AND CITRATE IN PROSTATIC FLUID
BY CHARLES HUGGINS, M.D., AND WILLIAM NEAL, M.D.
(From the Department of Surgery of The University of Chicago, Chicago)
(Received for publication, August 3, 1942)
The semen of man and animals is delivered from the urethra in a liquid
state but differs thereafter in its properties in three principal ways. In the
guinea pig it becomes an elastic solid which is maintained in this condition
both in vitro and in the vagina for many days. The semen of the dog remains
liquid. The ejaculate of normal man solidifies but then liquefies in a few
minutes at room temperature.
The solidification of the semen of the guinea pig is due to the action of an
enzyme, vesiculase (1, 2) derived from a special region of the prostate (3) on
the proteins of the seminal vesicle. No experiments have been reported deal-
ing with the solid-liquid phases of human semen or with the persistence of the
liquid state of canine semen, matters with which the present paper is con-
Our first observation was that a mixture of human semen with blood under-
went coagulation but soon liquefied. Most of the present experiments deal
with various phases of the effects of semen and its components on blood and
its clotting constituents.
Human semen was obtained by manual ejaculation. Following liquefaction, it
was centrifuged and the supernatant fluid used immediately for testing. Canine
semen was obtained by pilocarpine stimulation following the prostatic isolation pro-
cedure of Huggins, Masina, Eichelberger, and Wharton (4); this fluid was passed
through a Seitz filter to obtain a sterile filtrate free from cells.
Fresh blood and citrated plasma were obtained from various species. The plasma
was procured by centrifugalization of blood drawn into a solution of sodiumcitrate,
5 per cent, of which 1 co. was used for each 9 co. of blood; it was recalcified with a solu-
tion of CaC12, 1.5 per cent, of which 0.33 cc. was used for each cubic centimeter of
Testing for lytic activity was carried out by mixing 1 cc. of blood or plasma with
Icc. of semen or other fluid under examination, either undiluted or diluted to 1 cc. with
saline; after mixing, the tubes were stoppered and placed in a water bath at 37°C. and
* This investigation was supported by grants from the Committee for Research in
Problems of Sex, the National Research Council, and from the National Committee
on Maternal Health.
COAGULATION AND LIQUEFACTION OF SEMEN
the times of coagulation and of subsequent liquefaction, when it occurred, were noted.
In stating the dilutions in this paper the actual content of secretion in 1 ce. is given;
saline refers to an aqueous solution of NaCI, 0.15 molar in concentration, and incuba-
tion refers to 37°C. Where citrated plasma was used the specified concentration of
calcium chloride was added according to two plans; either the plasma was recalcified
immediately after adding the fluid to be tested, or after a delay of minutes to hours to
permit a prolonged contact, in the liquid state, of the plasma and the fluid under test.
As control, saliva, bile, cerebrospinal and spermatocele fluids were substituted for
semen and the effects on the dotting and liquefaction of blood noted.
In testing the effect of the lytic factors in semen on the dotting constituents of
human blood, fibrinogen, thrombin, and thromboplastin were prepared from beef
plasma and lung by the method of Smith, Warner, and Brinkhous (5) and prothrombin
by the technique of Seegers and colleagues (6).
Citrate analyses were made by the method of Pucher, Sherman, and Vickery (7)
and throughout this paper the term, citrate, refers to the combined amount of citric
acid and its salts as determined by this method. Trypsin was determined by the
method of Anson (8).
The opacity developing coincident with clot formation induced by recalcification
of citrated plasma, and the effect of semen upon this density were studied in an Evelyn
photoelectric colorimeter. The colorimeter was mounted in an incubator at 37°C.
and all of the solutions were brought to this temperature before using. A 6600 Ang-
strom light filter was used. The total volume in each colorimeter tube was 13.5 ce.
distributed as follows:--
5 cc. of plasma + 2 cc. of semen (saline in the control tubes) + 5 cc. of saline + 1.5
cc. of CaC12.
Following recalcification, readings of the galvanometer were recorded each 12 sec-
onds until a steady state had been reached following coagulation, or until it was
evident that dotting was not occurring.
The Phenomenon of Spontaneous Lysis of Human Semen.--The semen of
normal men after ejaculation forms a solid mass and the container may be
inverted without losing any of the material; within a few minutes softening
and partial liquefaction is observed and within 15 minutes the ejaculate is in
a liquid state save for a few sago-like particles, derived from the seminal
vesicles, which remain for more than 1 hour.
The lytic phenomenon may be observed well through the microscope.
a typical example, at 3 minutes after ejaculation a thin section appeared as
many interlaced bundles of long, clearly defined, parallel, refractile fibres; at
4 minutes the fibres appeared swollen; at 5 minutes the regularity had disap-
peared and the fibres were arranged haphazardly; at 6 minutes there was
large scale liquefaction with movement visible as many of the fibres were
rapidly breaking up. No fibres were seen 8 minutes after ejaculation.
In a man with hyopgonadism due to undescended testes the semen did not
CHARLES HUGGINS AND WILLIAM NEAL
clot; following intramuscular injection of testosterone propionate, 25 rag.
daily for 15 days, normal clotting was observed.
The Effect of Semen and Prostatic Fluid on the Coagulation Time of Blood.--
The clotting time of whole blood, 1 cc. mixed with saline, 1 cc., in 144 instances
ranged from 3 to 10 minutes.
(a) Human semen: Two effects were observed, shortening and prolongation
of the coagulation time of whole blood. Of 23 seminal specimens tested
against whole human blood, 11 prolonged the coagulation time, while 12
caused a decrease of the clotting time as compared to the control saline mix-
tures. In the group in which a mixture of semen lessened the time of coagula-
tion, the effect was evident with as little semen as 0.025 cc. (Table I). This
Effect of Semen and Other Human Secretions on the Coagulation Time of Human Blood
1 cc. of whole blood + 1 co. of secretion, diluted secretion, or saline.
Nature of secretion
Semen, group 1
Semen, group 2
Prostatic fluid heated
Prostatic fluid and CaCls*
Seminal vesicle secretion
Amount of secretion in test
5 5 --
* Calcium chloride, 1.5 per cent, 0.33 cc.
lessened coagulation time is due to thromboplastic substances in appreciable
amounts. In the group with prolonged clotting with semen in 1 cc. amounts
no, or slight clot, was observed in 30 minutes; in this group, semen in 0.2 or
0.1 cc. amounts often lengthened clotting time to, 15 or more minutes.
(b) Human prostatic fluid: All specimens of undiluted fluid tested (from 18
men) inhibited the dotting of human blood. With prostatic fluid 0.1 cc.,
dotting occurred in 14 to 50 minutes in 5 specimens, while no clot was observed
in 2 hours in 13 observations. The coagulation time of blood mixed with
prostatic fluid 0.02 cc. was approximately normal.
Coagulation of the blood was always induced by adding calcium chloride,
1.5 per cent, 0.33 cc. to prostatic fluid, 1 cc. In 5 cases incoagulability, or
greatly prolonged clotting time, was abolished and a clotting time shorter
than normal supervened; in 3 cases the coagulation time was greatly decreased
COAGULATION AND LIQUEFACTION 0~" SEMEN"
but still remained longer than the normal control. In all cases in which semen
caused a prolonged clotting of admixed blood, the addition of the specified
amount of calcium chloride abolished the delay and coagulation occurred more
rapidly than in the control saline-blood mixture.
(c) Dog prostatic fluid: Of 40 specimens of human blood, 1 cc., mixed with
prostatic fluid, 1 cc., the coagulation time was normal or shortened in 32 cases.
In 4 cases clotting did not occur while only slight clot was observed in 13 to 20
minutes with the fluid of 4 dogs in this group. In contrast to the system,
human prostatic secretion-blood, the addition of calcium ions did not initiate
or hasten clotting and the delay was due to a different mechanism, namely, the
disappearance of fibrinogen as it is discussed below.
Citrate Is the Cause of Prolonged Coagulation Time of Mixtures of Blood and
Human Prostatic Fluid or Semen.--It was found that human prostatic fluid
heated in a water bath at 100°C. for 30 minutes still delayed blood coagulation,
while the delay was abolished by adding calcium ions (Table I). Large
amounts of citrate in the prostatic fluid were discovered by Scherst6n (9)
and the finding was confirmed by Dickens (10). We likewise observed large
amounts of this acid, obtaining the following values for citrate in each 100
cc.; 15 specimens of human semen, 140 to 637 rag.; 9 samples of human prostatic
fluid, 480 to 2688 mg.; 20 lots of dog semen, 0. to 2.6 mg.; 2 analyses of human
seminal vesicle secretion, 15 and 22 rag. While Huggins, Scott, and Heinen
(11) found a high calcium content in human semen (6 millimols per liter) and
in human prostatic fluid (30 m~r per liter) the concentration of citrate in
prostatic fluid and some seminal specimens is so high that the calcium ions
are reduced preventing thrombin formation and blood coagulation. It seems
reasonable to conclude that the adverse effect of human prostatic fluid on
blood coagulation is a citrate effect.
Clotting Constituents of Human Semen.--The principal clotting constituents
of blood were prepared from beef plasma and lung. In order to determine
the presence of these or similar substances in semen, various combinations of
them were placed in contact with human semen which had become liquid and
the presence or absence of coagulation and the clotting time determined.
Differences were observed depending on whether calcium ions were added to
induce clotting immediately after adding semen or after preliminary incubation
of semen with the clotting fractions of blood in the liquid state.
(a) Immediate induction of clotting: Human semen was found capable of
substituting for fibrinogen and thromboplastin but not for prothrombin in
inducing clotting (Table II, lines 1-6). The clot formed in the absence of
added fibrinogen, was not firm, indicating that the fibrinogen content of semen
was less than that of the prepared extract. Goldblatt (30) discovered that
human semen contained thromboplastin.
(b) Delayed induction of clotting: Incubation of prothrombin with semen
CHARLES HUGGINS AND WILLIAM NEAL
both for 40 minutes and 18 hours decreased the coagulation time on adding
fibrinogen, thromboplastin, and calcium ions (Table II, lines 7-10).
tion of fibrinogen with semen for 18 hours abolished clot formation on adding
the other principle clotting agents. Neither thromboplastin nor prothrombin
were inactivated by incubation with semen for 18 hours; of the proteins im-
plicated in blood coagulation only fibrinogen was destroyed.
The Effect of Human Semen on the Clotting Constituents of Beef Blood
Fi ri I Throm-
b - Iboplas-
C6. [ CC.
Small amount of clot
A few fibres of fibrin appeared
at 15 sec. Solid clot in 10
The constituents designated * were incubated for 40 minutes at 37°C. and the other
elements were then added.
The constituents designated ~ were incubated for 18 hours and the other elements were
Lyric Action of Human Semen on Normal Human Blood.--The blood of
healthy persons when mixed with certain amounts of human semen or prostatic
fluid, will still clot but subsequently liquefy. With equal concentrations of
blood and semen, lysis occurs in 1 to 5 hours at room temperature, slightly
more rapidly at 37°C. and not at all at 4°C.
saline were mixed with blood, 1 cc., and observed at 18 hours (Table III).
In all instances semen, 0.03 cc., induced lysis; in 4 cases semen, 0.01 co.,
liquefied the clot. Twenty samples of prostatic fluid were studied in the
same way; lysis occurred in all with prostatic fluid, 0.02 cc., in 18 hours, and
in one fluid 0.002 cc. induced lysis.
Human semen did not liquefy beef plasma.
Serial dilutions of semen in
Lysis of dog blood and plasma
COAGULATION AND LIQlYEFACTION OF SEM~EN
clots occurred only twice in 14 experiments and then in no less amounts of
semen that 0.5 cc. and 0.1 cc. Rabbit plasma was resistant to solution.
Lyric Action of Dog Semen on Coagulated Blood of Normal and Diseased Per-
sons.--Lysis of dog blood, clotted after mixing with dog semen, occurred
regularly with semen 0.1 cc. and larger amounts. Solution of clots produced
similarly with rabbit, beef, and rat plasmas occurred with dog semen 1 cc.,
occasionally with 0.5 cc. amounts, never with less than 0.1 cc. of dog semen.
The semen of 10 dogs was tested by serial dilution against the blood of 46
normal persons; the semen was mixed with blood, 1 cc., allowed to clot, and
the experiment terminated at 18 hours.
coagulate, due to destruction of fibrinogen before the clotting was effected,
since adding beef fibrinogen induced prompt clotting.
In 4 instances the blood did not
Where clotting oc-
Lyric Action of Human Semen on Coagulated Blood of Various Species
1 ec. of whole blood -b 1 co. of secretion, diluted secretion, or saline.
Incubation 18 hours, 37°C.
Amount of semen in test
Nature of blood
0.5 cc. 0.03 cc,
Normal human .....
1 cc. .
0.1 cc. 0.05 cc.
L, complete lysis; I, incomplete lysis; 0, no lysis.
* Citrated plasma, recalcified with CaCI~, 1.5 per cent, 0.30 cc.
curred, the minimum amount of dog semen causing lysis was 0.016 cc. (Table
IV). In all cases dog semen, 0.1 cc. lysed blood 1 cc., but smaller amounts
were often ineffective.
In a similar manner, serial dilutions of dog prostatic fluid were tested for
lyric activity against the blood of 24 patients in hospital. Most of the patients
had disease of the urinary tract; 7 of them had benign or malignant tumors
of the prostate gland. Clotting occurred in all of the tests.
men with prostatic cancer, no lysis occurred in 18 hours with prostatic fluid,
1 cc. (Table V). In three patients with febrile illness, lysis occurred with
prostatic fluid, 1 cc., but not with amounts of 0.5 cc. or less. The blood of
19 patients underwent lysis in a normal manner.
General Properties of Fibrinolysis by Semen.--When normal blood, 1 co., was
mixed with saline, 1 cc., allowed to clot, and observed 18 hours later, syneresis
was always observed, and in the serum a small number (perhaps 1 per cent of
the total) of erythrocytes were found lying at the bottom of the tube, detached
In 2 cases, both
CHARLES HUGGINS AND WILLIA~ REAL
Lyric Action of Dog Semen on Coagulated Blood of Various Species
1 co. of whole blood -k- 1 ec. of secretion, diluted secretion, or saline.
Incubation 18 hours, 37°C.
Amount of prostatic secretion in test
Nature of blood
Guinea pig .........
None 1 cc.
0.5 cc. 0.1 cc.
0.05 cc. 0.033 cc.
L, complete lysis; I, incomplete lysis; O, no lysis.
* Citrated plasma, recalcified with CaCI,, 1.5 per cent, 0.33 cc.
The Lyric Action of Semen of the Dog on Human Blood Permitted to Coagulate
Total volume in test 2 co.: I co. of whole blood -t- 1 ec. of prostatic fluid, undiluted or
diluted with saline.
Healthy person (1)
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
Amount of dog semen in test
Carcinoma of prostate; febrile
" " "
c¢ cc tc ct
Benign prostatic hypertrophy; febrile
Tuberculosis of kidney; uremia,
L, lysis; I, partial lysis; 0, no lysis.
from the clot.
to a fibrinolysis.
erythrocytes than with plasma under the same circumstances for then it is
unusual to find any clot retraction.
It is not known whether this erythrocyte detachment is due
Syneresis is always more pronounced in the presence of
COAGULATION AND LIQD-EleACTION OF SEMEN
Fibrinolysis of whole blood by seminal fluids occurs in the absence of bac-
teria and of spermatozoa; the fluid component of fresh semen, free from evi-
dence of infection and passed through a Seitz filter soon after collection, remained
sterile and retained its ability to lyse blood. All samples of semen and prostatic
fluids contained the enzyme. The blood of the donor of prostatic fluid was
always susceptible to fihrinolysis by his fluid. After fibrinolysis the ery-
throcytes are intact and appear normal and the specific blood grouping is not
Maintenance of seminal fluids at 60°C. for 5 minutes did not interfere with
fibrinolytic action, but heating for the same period at 70°C. destroyed lytic
ability. Prostatic fluid dried in air and stored for 3 weeks retained its lytic
property when dissolved in quantity of saline equal to the original volume;
the dried powder heated for 2 hours at 105°C. was still active. Neither toluol,
sulfathiazole crystals in excess, nor merthiolate (sodium ethyl mercuri-thio-
salicylate) inhibited fibrinolysis. Inhibitors of oxidation, NaF or KCN in
0.025 molar concentration, did not interfere with fibrinolysis.
The sterile supernatant fluid following fibrinolysis by semen retained its
lyric activity when added to fresh blood and allowed to clot. Four such pro-
gressive transfers were made without loss of the lyric fraction.
Lyric Action of Human Body Fluids.--Fibrinolysis is not a widespread pro-
perty of body fluids. As control for the lyric effect found in prostatic fluid,
several human fluids were examined in a sterile state. These included; urine
(10 specimens); saliva passed through a Seitz filter, 21 specimens; gall bladder
bile, 6 specimens; hydrocele fluids, 2 lots; 2 spermatocele fluids; 6 normal
cerebrospinal fluids; and the secretion of the seminal vesicle from three men.
No lysis occurred in any.
Fibrinogenase.--When citrated plasma is recalcified, the clear plasma be-
comes opaque as clotting proceeds. This change in light transmission has
been adapted to quantitative study of blood coagulation as a nephelometric
procedure by Kugelmass (12) and especially by Nygaard (13) who devised a
special apparatus for recording the changes. We investigated the opacifica-
tion occurring in progressive dilutions of citrated plasma in a photoelectric
colorimeter at 37°C. with plasma concentrations between 1 and 10 cc. (Fig. 1).
An approximately linear relationship developed when units of density after
clotting are plotted arithmetically against the quantities of fibrin present.
Human citrated blood plasma incubated at 37°C. with prostatic fluid of the
dog, for short periods of time, shows progressive decrease of opacity after
recalcification. In a typical experiment, 5 cc. of citrated plasma were incu-
bated with 2 cc. of prostatic fluid and at intervals of 6 minutes, 5 cc. of saline,
and 1.6 cc. of calcium chloride were added; in the control tubes prostatic fluid
was replaced by saline since the principle electrolytes of dog prostatic fluid
are sodium and chloride in 160 milliequivalent concentration per liter (4).
CHARLES IIUQGINS AND V~LLIAM I~AL
The density resulting from clot formation became progressively less with
increasing incubation (Fig. 2). After incubation of 30 minutes, neither clotting
nor change of density took place. The failure of clot formation was due to
destruction of fibrinogen by dog prostatic fluid. However, no decrease of
density or of clotting capacity occurred when human semen was added under
comparable conditions to the citrated plasma. Indeed, after incubation of
human semen with plasma for 4 hours only a slight decrease of density was
i . !
: - --
90 ' • 2 CC.
I 2 3 4 5 5 7 8 9
F[o. 1. A photoelectric study of the increased density occurring as clot forms in
citrated plasma after recalcification. Progressive increases of density occur with in-
creased content of plasma (cubic centimeters), the total volume of fluid in each tube
being the same. Ordinates: units of galvanometric deflection. Abscissae: time in
minutes after recalcification.
observed (Fig. 3). Human semen possesses only slight power of destroying
Incubation of citrated plasma, 1 cc., with dog prostatic fluid, 1 cc., for 30
minutes abolished clotting on adding CaCh; prompt clotting occurred when
beef fibrinogen, 0.5 cc., was subsequently added.
The differences in activity of canine and human semen on fibrinogen and
fibrin were magnified by dilution of these fluids. Two series of tubes were
arranged, all of which contained citrated human plasma, 1 cc., and either
human or dog semen, 0.5 cc.; in series A, clotting was induced immediately,
by adding calcium chloride, 0.25 per cent, 0.25 cc.; in series B, calcium chloride
30 MINUTE INCUBATION
NORMAL PLASMA CONTROL
I I I 1
7 2 3 4
FIG. 2. Photoelectric study of the effects of incubation of dog semen with human
citrated plasma on clot formation and density occurring after recalcification. A pro-
gressive decrease of clot density occurs with increased periods of incubation until
contact for 30 minutes results in failure of dotting to take place. When dotting
occurs, it begins in all of the tubes containing semen earlier than in the control tubes
of plasma without semen, due to the thromboplastic activity of the secretion. Ordi-
nates: units of galvanometric deflection. Abscissae: time in minutes after recalci-
EO INCUBATION BATION ~4 HOURS
I I | I
7 2 3 4
FzG. 3. A photoelectric study of the effect of incubation of human semen with
citrated plasma on clot formation and density occurring after recalcification. In
contrast to dog semen, only a slight decrease of density occurs after 4 hours incubation;
the thromboplasfic effects are evident as reflected in the increased rate of clotting over
that of the control tube of plasma without semen. Ordinates: units of galvanometric
deflections. Abscissae: time in minutes after recalcification.
CHARLES HUGGINS AND WILLIAM NEAL
was added after 18 hours' incubation. At 18 hours in series A, clots contain-
ing dog semen, 0.1 cc., were only slightly liquefied, and those containing
0.012 cc. were solid, while accompanying clots containing these amounts of
human semen were liquefied (Table VI). At 18 hours in series B, clot could
not be induced by recalcification of any of the tubes containing dog semen
but the addition of calcium chloride to tubes containing less human semen
than 0.1 cc. promptly formed clots which subsequently underwent lysis in
The Effect of Dog and of Human Semen on Lysis of Fibrin and Fibri~gen
1 cc. of plasma + 0.5 cc. of diluted semen
Series A: 0.25 cc. of CaC12 added immediately.
Series B: incubation of fluids for 18 hours before adding CaC12.
Enzymatic action on dotted plasma
Extent of lysis
for 18 hrs.
Enzymatic action on liquid plasma
Time of lysh
+ Did not coagulate I
Did not coagulate
2 to 5 hours. Further, adding fibrinogen, 0.5 cc., to all tubes in series B which
had not clotted after recalcification produced prompt coagulation.
Both human and dog prostatic fluids contain entities capable of destroying
fibrin (fibrinolysin) and fibrinogen (fibrinogenase) but in different concentra-
tions. Human semen contains much fibrinolysin, little fibrinogenase; dog
semen contains little fibrinolysin, much fibrinogenase.
Thrombin in Semen.--The semen of certain dogs has the capacity of clotting
fibrinogen, free from prothrombin, and of producing spontaneous clots in
oxalated or citrated beef and rabbit plasmas. The calcium content of dog se-
men is about 0.3 rnilllmols per liter (4). The thrombin activity is weak and
requires 5 to 60 minutes before clot is evident. Citrated human plasma is not
clotted by dog semen, obviously because human fibrinogen is destroyed by
COAGULATION AND LIQUEFACTION O~ SEMEN
fibrinogenase before the clot can be formed. Human semen does not contain
To 0.2 cc. of rabbit oxalated plasma diluted with 0.8 cc. of saline, 0.5 cc. of dog
semen was added and the tubes maintained at 37°C.; spontaneous clotting was ob-
served in 18 minutes. A similar experiment using human plasma instead of rabbit
plasma resulted in absence of clot at l0 minutes; the addition of 0.25 cc. of CaC12,
0.25 per cent did not induce dotting. Likewise, replacing dog semen with human
semen, 0.5 cc., did not induce clotting in rabbit or human plasma.
To 1.0 cc. of beef fibrinogen, 0.5 cc. of dog prostatic fluid was added, clotting
occurred in 20 minutes; replacing dog prostatic fluid with human semen did not result
in clotting in 5 hours.
Trypsin in Semen.--The semen of both dog and man produced no visible
change on coagulated egg albumin in Mett tubes when incubated overnight
at 37°C. Using the method of Anson (8) small amounts of a proteolytic
enzyme active at pH 7.5, trypsin, were demonstrated in 10 prostatic fluids of
dogs; prostatic fluid, 1 cc. liberated chromogen equivalent to 0.029 to 0.15 mg.
of tyrosin in 15 minutes. Five human semens, in 1 cc. volume liberated from
0.03 to 0.09 rag. of tyrosine, in 15 minutes; two human semens did not contain
The mechanism by which human semen coagulates is not clear, and the
problem is complicated by the difficulty of securing semen prior to clotting.
The finding of fibrinogen and thromboplastin is evidence that seminal clotting
resembles blood clotting; however, thrombin and prothrombin were not demon-
strated after liquefaction. The abundance of citrate, which is sufficient to
bind all of the calcium ions in many semens, provides an obstacle to the in-
ference that the clotting of blood and semen is due to an identical mechanism.
The presence of an active fibrinolytic agent would readily explain liquefaction,
if, as seems likely, fibrin were the cause of the .seminal clot.
The fibrinolysin in semen is derived from the prostate gland, and it was
present in large amount in all of the prostatic fluids examined. This constant
occurrence establishes a new function for the prostate. We have been unable
to find previous observations relative to the action of genital secretions on
fibrin except by Kurzrok and Miller (14), who were unable to demonstrate an
effect of semen on fibrin, the source of which was unspecified. While all of
the selnens became liquid spontaneously in less than 10 minutes in our tests,
the most rapid liquefaction of blood, 1 cc., by semen, 1 cc., occurred in 30 min-
utes. The lysis of blood clot by semen occurs less rapidly than lysis of the
clotted semen itself.
Fibrinolysis without implication of the agent involved has been studied by
several workers since its discovery in serum by Dastre (15). Nolf (16) re-
CHARLES HUGGINS AND WILLIAM HEAL
garded a slow aseptic fibrinolysis to be a natural sequel of plasma coagulation.
Judine (17) observed that the blood of men in profound traumatic shock did
not coagulate; further, that the blood of healthy persons, meeting a violent
death, coagulated rapidly but within several hours became liquid and did not
thereafter coagulate. Macfarlane (18) found that frequently the clotted blood
of patients after surgical operations subsequently underwent complete lysis
under aseptic conditions within 24 hours.
The effect of "chloroform semen" is germane to the present discussion.
Howell (19) and Minot (20) discovered that clotting occurred in oxalated
plasma to which chloroform had been added and Nolf (21) observed that the
serum obtained from such clots is fibrinolytic. These findings have been
confirmed by Tagnon (22, 23) who obtained from such serum, from which the
chloroform had been removed, a globulin with marked fibrinolytic properties;
the attdition of this globulin to fibrinogen produced no clot but effected com-
plete lysis of fibrinogen, while in the presence of prothrombin a clot formed
which sometimes underwent fibrinolysis. The action on blood plasma and
fibrinogen of this globulin resembles the effect of trypsin as described by
Eagle and Harris (24). Ferguson and Erickson (25) in studying the clotting
action of crystalline trypsin on citrated plasma, observed that clots so ob-
tained undergo fibrinolysis within a few minutes; they found that trypsin, in
1 to 2 rag. amounts was optimal for the clotting of 1 cc. of citrated dog plasma.
Many points of similarity occur between the fibrinolysin of semen and that
discovered by Tillett and Garner (26) in hemolytic streptococci. Both agents
lyse normal human blood clot easily but act only after prolonged periods of
time and in high concentration, when at all, on rabbit blood. In disease, at
times, the blood of certain patients is totally resistant to lysis by semen;
Tillett, Edwards, and Garner (27) got the same result with the fibrinolysin
from streptococci in their patients with streptococcal infections. Further the
active principle is demonstrable in dissolved fbrin even after incubation for
18 hours in both cases. Slight differences occur however. Garner and
Tillett (28) found that streptococcal fibrinolysin resisted heating at 100°C.
for 60 minutes, while we observed that semen is inactivated by heating to 70°C.
for 5 minutes. Seminal fibrinolysin retains its activity for months in the re-
frigerator while that from streptococci deteriorates in several weeks.
In relating these observations to the effect of semen in the lysis of fibrin
and fibrinogen, it should be stated that all of the samples of dog semen, and
some human semens contained small amounts of trypsin; other human semens
were free from trypsin. All samples were inactive in destroying beef fibrinogen
Both human and dog semens contain factors capable of inactivating fibrin
and fibrinogen but in different and inverse proportions; human fluid contained
fibrinolysin with greater activity than dog semen, while dog semen exerted
COAGULATION AND LIQIrEFACTION OF SE~EN
far greater fibrinogenase activity than human semen. It was readily possible
by dilution to eliminate the weaker activity and to retain the stronger; this
is evidence for the presence of two distinct proteolytic agents acting on fibrin
and fibrinogen respectively. Garner and Tillet (29) observed that "solutions
of human fibrinogen after brief incubation with fibrinolysin lose the capacity
to form thrombin." The observed quantitative and species differences of
proteolysis in semens do not fall in with the interpretation of these workers,
whose observation on streptococcal filtrate with respect to fibrinogen is the
same as ours on dog semen. We conclude that the agent in semen which
liquefies fibrin resembles closely or is identical with the fibrinolysin of Tillett
and Garner, and that the inactivation of fibrinogen is due to a separate agent,
Certain specimens of human semen shorten the coagulation time of whole
blood because of the presence of active thromboplastic agents, while other
samples prolong its coagulation time. Human prostatic fluid in large amounts
always delays or abolishes blood coagulation. The delay or absence of clot-
ting is counteracted by adding calcium ions and is due to the large concentra-
tion of citrate in prostatic fluid and in some semens.
While most specimens of dog semen shorten the coagulation time of blood
because of their thromboplastic activity, certain specimens render blood in-
coagulable or delay coagulation; in contrast to human semen, this adverse
effect on coagulation is not overcome with calcium ions and is due to a dif-
ferent mechanism, the lysis of fibrinogen. The citrate content of dog prostatic
fluid is small.
Human semen which has become liquefied does not contain thrombin or
prothrombin, but fibrinogen and thromboplastic substances are present. Beef
fibrinogen added to semen is destroyed by incubation for 18 hours, but added
prothrombin and thromboplastic substances are still present after this treat-
ment. Dog semen, in some instances, contains small amounts of thrombin.
The semens of man and dog contain a fibrinolysin for human blood which
seems not to differ greatly from the fibrinolysin associated with hemolytic
streptococci. The blood of the donor of prostatic fluid is susceptible to
fibrinolysis by this fluid. However, the blood of persons with some diseases,
is absolutely resistant to the action of seminal fibrinolysin. In how many
diseases this happens has not yet been determined.
The semens of man and dog both contain an agent capable of inactivating
fibrinogen, but in different amounts. This activity may be called fibrinogenase.
Human semen is rich in fibrinolysin, poor in fibrinogenase; dog semen is rich
in fibrinogenase, poor in fibrinolysin. These species differences, together with
the fact that it is easy by appropriate dilution to retain the stronger proteolytic
CHARLES HUGGINS AND WILLIA.M lqEAL
agent and eliminate the weaker one, imply that fibrinolysin and fibrinogenase
are different entities.
Dog semen, and less constantly human semen, contain very small amounts
All of these proteolytic agents derive from the prostate gland; their secre-
tion in prostatic fluid constitutes a hitherto undescribed function for the
1. Camus, L., and Gley, E., Compt. rend. Soc. biol., 1896, 48, 787.
2. Camus, L., and Gley, E., Compt. rend. Soc. biol., 1897, 49, 787.
3. Walker, G., Bull. Johns Hopkins ttosp., 1910, 9.1, 182.
4. Huggins, C., Masina, M. H., Eichelberger, L., and Wharton, J. D, Y. Exp. Med.,
1939, 70, 543.
5. Smith, H. P., Warner, E. D., and Brinkhous, K. M., J. Exp. Meal., 1937, 66, 801.
6. Seegers, W. H., Smith, H. P., Warner, E. D., and Brinkhous, K. M., Y. Biol.
Chem., 1938, 123, 751.
7. Pucher, G. W., Sherman, C. C., and Vickery, H. B., J. Biol. Chem., 1936, 113, 235.
8. Anson, M. L., J. Gen. Physiol., 1938, 9.2, 79.
9. Scherst~n, B., Skand. Arch. Physiol., 1936, 74, suppl. 9.
10. Dickens, F., Biochem. J., 1941, 35, 1011.
11. Huggins, C., Scott, W. W., and Heinen, J. H., Am. Y. Physiol., 1942, 136, 467.
12. Kugelmass, I. N, Arch. internat, physiol., 1923, 9.1, 139.
13. Nygaard, K. K., Hemorrhagic diseases, St. Louis, The C. V. Mosby Co., 1941.
14. Kurzrok, R., and Miller, E. G., Am. J. Obst. and Gynec., 1928, lfi, 56.
15. Dastre, A., Arch. Physiol., 1893, series 5, 5, 661.
16. Nolf, P., Arch. internat, physiol., 1908, 6, 115.
17. Judine, S. S., Press m~d., 1936, 44, 68.
18. Madarlane, R. G., Lancet, 1937, 1, 10.
19. Howell, W. H., Am. J. Physiol., 1910, 9.6, 452.
20. Minot, G. R., Am. J. Physiol., 1915, 39, 131.
21. Nolf, P., Arch. internat, physiol., 1922, 19, 227.
22. Tagnon, H. J., Science, 1942, 95, 334.
23. Tagnon, H. J., J. Lab. and Clin. Med., 1942, 9.7, 1119.
24. Eagle, H., and Harris, T. N., J. Gen. Physiol., 1937, 9.0, 543.
25. Ferguson, J. H., and Erickson, B. N., Am. J. Physiol., 1939, 19.6, 661.
26. Tillett, W. S., and Garner, R. L., J. Exp. Med., 1933, 58, 485.
27. Tillett, W. S., Edwards, L. B., and Garner, R. L., Y. Clin. Inv., 1934, 13, 47.
28. Garner, R. L., and TiUett, W. S., J. Exp. Med., 1934, 60, 239.
29. Garner, R. L., and Tillett, W. S., J. Exp. Med., 1934, 60, 255.
30. Goldblatt, M. W., Biockem. J., 1935, 9.9, 1346.