THE AMERICAN JOURNAL OF CLINICAL PATHOLOGY
Vol. 48, No. 1
Copyright © 1967 by The Williams & Wilkins Co.
Printed in U.S.A.
EVIDENCE OF HEMOLYSIS IN THE INITIATION OF HEMOSTASIS
HARLAN J. PEDERSON, B.S., THOMAS H. TEBO, B.S., AND SHIRLEY A. JOHNSON, PH.D.
Research Service, Wood Veterans A dminislraiion Hospital, Wood, Wisconsin, and Department of Physiolo
Marquette University School of Medicine, Milwaukee, Wisconsin
diphosphate in the blood shed 20 sec. after
transection of mesenteric arterioles in
guinea pigs. The adenosine diphosphate
appears to accumulate in shed blood, for no
evidence of adenosine monophosphate or
adenosine was observed and the total
amounts of adenosine triphosphate and
adenosine diphosphate remained constant in
the blood collected at different times during
bleeding.
From the above data, our group has concluded that both adenosine diphosphate and
thrombin appear simultaneously in the
early stages of hemostasis.
We have no evidence that adenine
nucleotides are added to the shed blood
from the damaged vessel wall, as was suggested by Born.1 No difference was found
between the total amount of adenine
nucleotides in the shed blood in comparison
with the total amount in circulating blood
sampled by heart puncture. 0
The quantity of adenosine diphosphate
observed in whole shed blood exceeded that
obtainable from platelets, with the result
that we concluded that much of it must be
of red blood cell origin. Most of the fibrin
observed in the early stages of hemostasis
was in close proximity to red blood cells. On
the basis of these two observations, we
postulated that both the newly evolved
adenosine diphosphate and the partial
thromboplastin of red blood cells may contribute substantially to the initiation of
hemostasis. The known physiologic mechanism that could most likely account for the
release of these two substances from red
blood cells was hemolysis. Hellem and colleagues3 had postulated earlier that microhemolysis might account for the availability
of adenosine diphosphate in hemostasis.
The present study was carried out to
determine whether hemolysis occurred when
blood was shed through the transected
vessel wall. First the ultrastructure of experimental hemolysis was analyzed, to en-
Received October 11, 1966.
This study was supported in part by Research
Grant No. HE-06033, U. S. Public Health Service.
02
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
Older concepts of hemostasis, based on
the work of many investigators, postulated
the initial cohesion of platelets to be clue
primarily to the liberation of adenosine
diphosphate on the platelet surface. The
appearance of thrombin, which renders the
aggregation irreversible, was believed to
occur only after the platelets had aggregated.
This brief summary was enlarged by
Spaet and Zucker11 into a hemostatic theory
composed of three steps which closely
resembled the concept of Owren.8 The initial
reaction was platelet adhesion to connective
tissue fibers, resulting in the release from
the platelets of a small amount of adenosine
diphosphate which, in the presence of calcium, brought about cohesion of the platelets into masses. This aggregation of platelets was made more durable and compact
by the action of thrombin, which also
brought about additional release of adenosine nucleotides.
Not all research groups, however, place
the role of the blood coagulation mechanisms so late in the hemostatic processes.
Troup and Luscher12 state that hemostasis
is probably initiated by release of tissue
thromboplastin from damaged endothelial
cells, resulting in formation of thrombin on
the platelet surface.
Marr and associates6 have found ultrastructural evidence of fibrin, often in close
proximity to collagen and red blood cells,
15 to 20 sec. after transection of small
mesenteric arterioles in guinea pigs. Welldeveloped loci of fibrin fibers can be identified by the characteristic periodicity of 240
A 30 sec. after the transection of the vessel.
Marr and associates7 have also established
that 50% of the total blood adenosine triphosphate is broken down to adenosine
July 1967
HEMOLYSIS AND HEMOSTASIS
63
T "
R8C
\>i
\
FIG. 1. This election micrograph shows a red cell clump surrounding a platelet aggregate in a hemostatic plug 1 min. after bleeding began. Note the reduced electron density. X S400.
able us to identify similar ultrastructure in
hemostatic plugs in vivo. The amount of
hemolysis in circulating blood collected by
heart puncture was then compared with the
amount of hemolysis in blood shed from
transected mesenteric vessels. An attempt
was made to obtain evidence of a hemolysin.
MATERIALS AND METHODS
Benzidine base (Hartman-Leddon Company, Inc.). The purification procedure of
this reagent was carried out as described by
Hanks and co-workers.2
Glacial acelic acid (Fisher Scientific Company). This preparation was used as supplied.
Hydrogen peroxide, 30% (Fisher Scientific
Company). This preparation was diluted to
3 % beforeuse.
Heparin, Liquaemin sodium 10 (Organon,
Inc.). This preparation was used as supplied.
Electron microscopy. The materials used for
electron microscopy have been described
previously.4 In brief, the tissue was embedded in Epon S12 according to the procedure of Luft6 with 3 parts of Epon A and 7
parts of Epon B. The sections were viewed on
an EMU 3G double condenser RCA electron
microscope.
Formation of in vivo hemostatic plug. A
small mesenteric vessel of a guinea pig was
transected and allowed to bleed for approximately 55 sec. The bleeding end of the vessel
was removed and placed in a vial of cold
s-collidine buffered osmium at exactly 1 min.
after transection. After 10 min. the tissue
was divided into 1-mm. blocks and replaced
in the osmium for 1 hr. Subsequent dehydration, penetration, and embedding were carried out according to technics described in a
previous publication.4
Preparation of red blood cells for study of
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
RBC
, &
64
TABLE 1
NUMBER OF HEMOLYZED RED BLOOD CELLS
COUNTED WITH THE ELECTRON MICROSCOPE
Type of
No.
No.
Red Blood Cell Specimen Counted Hemolyzed Hemolysis
In vivo plug
In 0.35% saline
In 0.9% saline
Vol. 48
PEDERSON ET AL.
1073
1145
1371
1059
1122
21
%
98.69
97.99
1.53
RESULTS
Ultrastructural Evidence of Hemolysis in
in Vivo Hemostatic Plugs
Red blood cells clumped in in vivo hemostatic plugs fixed after 1 min. of bleeding
displayed reduced electron density. The red
blood cells showed reduced electron density
of varying degrees and tended to be spherical
in shape, with irregular or serrated membranes, as illustrated in Figure 1. Table 1 is
a comparison of the number of hemolyzed
red blood cells (identified in electron micrographs by reduced electron density) found in
the in vivo clump with the number of red
blood cells hemolyzed when they were suspended in 0.35 % solution of saline and 0.9 %
solution of saline for 10 min. Nearly all of
the red blood cells in the in vivo clump were
hemolyzed.
Figure 2 is an electron micrograph taken
from the same in vivo clot, but at the
periphery of the red cell clump. It may be
seen that the free red blood cells directly
around the outer limits of the clumped red
cells showed normal electron density. Their
shape was irregular or flattened, as is usual
for circulating red blood cells, instead of
spherical, and the membranes tended to be
smooth. All of the in vivo clots studied
shared this common characteristic of a centrally located clump of hemolyzed red blood
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
experimental hemolysis. Guinea pig blood was
collected by heart puncture and added to
3.2% solution of sodium citrate in a ratio of
9:1. When the blood was centrifuged at
3000 X g for 20 min. and the plasma was
drawn off, the packed red blood cells were
divided into two equal parts. One half was
suspended in an equal volume of 0.35%
solution of saline and the other half in an
equal volume of 0.9% solution of saline.
After the tubes were inverted to facilitate
mixing, the suspensions were centrifuged for
20 min. at 3000 X g and the packed red
blood cells were fixed and embedded as
described.
Determination of hemoglobin in various
•plasmas. Circulating blood was obtained by
heart puncture through a 20-gauge needle on
a syringe containing heparin, final concentration 1000 units per ml. The shed blood was
obtained immediately after transection of a
mesenteric arteriole by collecting the blood
in either capillary pipettes or a syringe and
needle similar to those used to sample
circulating blood. The specimen of circulating blood was obtained by heart puncture
either before or after the specimen of shed
blood. The blood was centrifuged at 3000 X
g for 20 min. and the plasma was removed.
The plasma hemoglobin was determined by
use of the sensitive methods of Hanks and
associates,2 which is based on the fact that
hemoglobin catalyzes the oxidation of benzidine by hydrogen peroxide.
Attempt to detect a hemolysin or hemolysins
in plasma of shed blood. Circulating blood
was collected by heart puncture and shed
blood was collected from transected mesenteric vessels in heparin as described previously. Red blood cells separated from the
circulating blood were incubated in a water-
bath at 37 C. for 10 min. with plasma
separated from shed blood in an attempt to
detect a hemolysin liberated from the damaged cells of the transected arteriole into the
plasma of the shed blood. Both the plasma
and the whole blood were kept at 4 C. until
the incubation period. In some experiments
the red blood cells were incubated with the
plasma in a ratio of 1:1; in others the ratio
was 1:9.
Hemoglobin determinations were carried
out on the following control and test combinations: plasma from the heart puncture;
plasma from the shed blood; plasma from an
incubation mixture of whole blood and
plasma from shed blood; plasma from an
incubation mixture of whole blood and
plasma from circulating blood; and shed
blood plasma and plasma from circulatingblood.
July 1967
HEMOLYSIS AND HEMOSTASIS
G5
cells surrounded by free, electron-dense red
cells."
Electron Density of Red Blood Cells Suspended
in 0.9% Saline Compared with Those
Suspended in 0.S5 % Saline
Figure SA shows red blood cells that were
suspended in 0.9% solution of saline before
fixation for electron microscopy. Nearly all
exhibited normal electron density (Table 1).
Notice that they were irregular and flattened
in shape in comparison with the hemolyzed
cell, which was spherical (Fig. 3, A and B).
Figure 'SB is an electron micrograph of red
blood cells that were suspended in 0.35%
solution of saline before fixation. Nearly all
of the cells displayed varying degrees of
reduced electron density (Table 1). They all
tended to be spherical in shape and looked
much like those found in the in vivo clump.
Chemical Determination of Hemoglobin
in Circulating and. Shed Blood
The results of a chemical determination
of hemoglobin in shed blood from mesenteric
vessels compared with circulating blood collected by heart puncture are shown in Table
2. The plasma hemoglobin from the shed
blood of the mesenteric vessels was always
higher (mean 11.4 ± 4.6) than the hemoglobin from the heart puncture (mean 5.1 ±
1.1), whether the heart puncture was performed before or after the transection of the
vessels. Both ultrastructural and biochemical
data show that a significant amount of
hemolysis takes place in the early stages of
hemostasis when blood is shed.
Attempt to Detect a Plasma Hemolysin
Since we had shown that hemolysis results
from a major injury to the vessel wall, it
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
FIG. 2. In this electron micrograph on the edge of a hemostatic plug, some clumped red blood cells
with reduced electron density, and a free red blood cell with the normal amount of electron density may
be seen. The second electron dense red blood cell has just joined the clump. This tissue was fixed after
1 min. of bleeding. X 8400.
66
Vol. Z8
PEDERSON ET AL.
TABLE 2
COMPARISON OP HEMOLYSIS IN CIRCULATING
BLOOD COLLECTED HY HEART PUNCTURE AND
IN SHED BLOOD FROM MESENTERIC VESSELS*
Type of Sample
Plasma Hemoglobin
Mean
S.D.
mg./lOO ml.
Blood from heart puncture
5.1
1.1
Shed bloodfrom mesenteric vessel 11.4
4.1
•N
10; p = 0.001.
was of interest to determine the mechanism
of hemolysis. Attempts to demonstrate presence of a hemolysin in the plasma were
unsuccessful. Red blood cells incubated with
the plasma from shed blood did not significantly add to the hemoglobin level of the
plasma alone. The ratio of red blood cells to
plasma was altered from 1:1 to 1:9 in an
attempt to create conditions in which the
concentration of hemolysin was higher in
comparison with the amount of red blood
cells.
DISCUSSION
Although hemolysis has held the interest
of investigators for several decades, the
actual means whereby hemoglobin leaves the
intact red blood cells is not understood. We
have presented conclusive evidence that
hemolysis takes place in hemostatic plug
formation, but have not elucidated the tic-
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
FIG. 3a (left). This electron micrograph shows electron dense red blood cells that have been suspended
in 0.9% NaCl. Only one red blood cell of reduced electron density can be seen. X 2200.
FIG. 36 (right). The red blood cells shown in the electron micrograph were suspended in 0.35% NaCl
solution for 30 min. Most of the red blood cells exhibit reduced electron density and spherical shape,
typical characteristics of hypotonic hemolysis. X2200.
July 1967
HEMOLYSIS AND HEMOSTAS1S
07
Fig. 4.
HEMOSTASIS
(bleeding resulting from a major injury)
Vessel Wall Releases
)
Tissue Tpln.
(
Activation of Prothrombin
>
THROMBIN
t
Degradation of ATP in RBC's
t
RBC's, on contact with a foreign surface,
hemolyze releasing
I
(4 min.) Confluence of Fibrin Loci
ADP
0.2uM/ml.
t
Few Platelet Aggregates
\
Extensive Platelet Aggregation
Platelet ' D e g r a n u l a t i o n
ADP
( Disintegration 1.2 u M/ml.
t
Hemostasis
Achieved
Advanced Platelet Disintegration
RBC Entrapped in Fibrin Network Emanating from Platelet Aggregation
(15 min.)
(Clot Retraction in Progress)
FIG. 4. This figure illustrates a possible mechanism by which hemolysis of red blood cells may contribute both adenosine diphosphate and partial thromboplastin to the initial stages of hemostasis.
tual liemolytic mechanism. The method of
Hanks and associates2 for measurement of
hemoglobin in plasma is very sensitive, but
if great care in the collection of the blood is
taken, small amounts of hemoglobin can be
accurately detected in plasma.
The possibility that damaged cells in the
vessel wall released a substance, a hemolysin,
into the plasma that hemolyzed nearby red
blood cells was considered. Using the technics described in this manuscript, we were
unable to find evidence of a hemolysin in the
plasma of shed blood. A relation between
the injured vessel wall and the hemoglobin
in the plasma of shed blood, which could be
demonstrated by other means, possibly
exists.
Although we do not know what aspect of
hemostasis initiates the hemolysis described
here in shed blood, several mechanisms are
possible. Perhaps a newly formed, as yet
unidentified component, or simply contact
with a foreign surface in hemorrhage, alters
the red blood cell membrane sufficiently to
bring about hemolysis. Further investigation
of this phenomenon is in progress.
From the above data, it can be seen that
the injured vessel wall may have two purposes, serving as a source of tissue thromboplastin which initiates both the coagulation
mechanisms and degradation of adenosine
triphosphate; and as hemolysin or as a
foreign surface responsible for hemolysis.
Both functions may play a part in vivo.
Some components of the coagulation
mechanisms (activated by tissue thromboplastin) might initiate reduction of adenosine
triphosphate to adenosine diphosphate in
red blood cells. Upon contact with the injured vessel wall, the red cells could hemolyze, releasing both the preformed adenosine
diphosphate and the partial thromboplastin
of Shinowara10 and Quick and associates,9
bringing about further activation of pro-
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
Partial Tpln.
i
Activation of Prothrombin
»
(3-15 sec.) THROMBIN
r—
(15 sec.) Fibrin Formation
I
(Prothrombin
(30 sec.) Several(Fibrin Loci (activated by
(1 min.) Expansion of Fibrin Loci(released platelet 3
I
PEDERSC sT ET Ah.
68
thrombin to thrombin. The oncoming platelets then flow into an environment rich in
both thrombin and adenosine diphosphate,
and extensive platelet aggregation takes
place.
Based on these findings, a schema for
hemostasis is presented (Fig. 4).
SUMMARY
REFERENCES
1. Born, G. V. R.: Aggregation of blood platelets
by adenosine diphosphate and its reversal.
Nature, 194: 927-929, 1962.
2. Hanks, G. E., Cassell, M., Ray, R, N., and
Chaplin, H.: Further modification of the
benzidine method for the measurement of
hemoglobin plasma. J. Lab. & Clin. Med.,
56: 486-498, I960.
3. Hellem, A. J., Borchgrevink, C. F., and Ames,
S. B.: The role of red cells in haemostasis:
the relation between haematocrit, bleeding
time and platelet adhesiveness. Brit. J.
Haemat., 7: 42-50,1961.
4. Johnson, S. A., Balboa, R. S., Pederson, IT. J.,
and Buckley, M.: The ultrastructure of
platelet participation in hemostasis.
Thromb. et Diath. Haemorrh., IS: 65-S3,
1965.
5. Luft, J. H.: Improvements in epoxy resin embedding methods. J. Biophys. Biochem.
Cytol., 9: 409-414,1961.
6. Marr, J., Barboriak, J. J., and Johnson, S. A.:
Relationship of appearance of adenosine
diphosphate, fibrin formation and platelet
aggregation in the haemostatic plug in vivo.
Nature, 205: 259-262, 1965.
7. Marr, J., Tebo, T. H., and Johnson, S. A.:
The dependency of adenosine triphosphate
degradation on the coagulation mechanisms
in hemostasis of whole blood. Nature, 211:
1308, 1966.
8. Owren, P. A.: The haemostatic plug. Conference Internal. Comm. Blood Clotting Factors, Gleneagles, Scotland. Thromb. et
Diath. Haemorrh. Supp. 13: 325-333, 1903.
9. Quick, A. J., Georgates, J. G., and Hussey,
C. V.: The clotting activity of human erythrocytes: theoretical and clinical implications.
Am. J. Med. Sc, 228: 207-213, 1954.
10. Shinowara, G. Y.: Enzyme studies on human
blood. XL The isolation and characterization of thromboplastic cell and plasma components. J. Lab. & Clin. Med., 38: 11-22,
1951.
11. Spaet, T. H., and Zucker, M. B.: Mechanism
of platelet plug formation and role of adenosine diphosphate. Am. J. Physiol., 206:
1267-1274, 1964.
12. Troup, S. B., and Liischer, E. F.: Hemostasis
and platelet metabolism (editorial). Am.
J. Med., S3:161-165,1962.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 5, 2016
Some red blood cells in an in vivo hemostatic plug fixed 1 min. after transection of a
small mesenteric arteriole in a guinea pig
were found to display reduced electron density when they were compared with normalappearing red blood cells. Hemolysis was
quantitated biochemically in circulating
blood drawn by heart puncture and in blood
shed from transected mesenteric arterioles,
and it was established that hemolysis does
occur in shed blood. I t was postulated that
hemolysis is the mechanism whereby both
adenosine diphosphate and partial thromboplastin are released from red blood cells in
the initial stages of hemostasis.
Vol. 48