Survival of rabbit platelets exposed to immune complexes

EXPERIMENTAL

AND

MOLECULAR

PATHOLOGY

52, 109-121 (1990)

Survival of Rabbit Platelets Exposed to Immune Complexes A. PARBTANI,’

R. L. KINLOUGH-RATHBONE,~ A. CHAHIL,~ M. RICHARDSON,’ AND J. FRASERMUSTARDS

‘Division of Nephrology, Victoria Hospital, London, Ontario; and ‘Department of Pathology, McMaster University, 1200 Main Street West, Hamilton, Ontario, Canada L8N 325 Received June 27, 1989, and in revised form September 18, 1989 Rabbits injected with human serum albumin (HSA) formed detectable immune complexes after 5 days; complex formation was maximal between 11 and 14 days after which the complexes were cleared from the circulation. Platelets from control rabbits or HSA-injected rabbits had a reduced survival upon injection into rabbits in which complexes were forming. Platelets from HSA-animals tended to survive for a longer period upon injection into control rabbits than when they were injected into HSA-rabbits, raising the possibility that some of the immune complexes may have eluted from their surface. Platelets prepared from either control animals or from HSA-treated animals at the time when complexes were being cleared from the circulation (M-21 days) did not have a shortened life span in HSA- or control rabbits. When platelet survival was reduced, it could not be attributed to platelet accumulation at sites of vessel wall injury or to accumulation in kidneys damaged by immune complexes, since the tissues (aorta and kidney) appeared to be morphologically normal and free of thrombi. The reduction in platelet survival likely results from the interactions of immune complexes with the surface of platelets leading to the platelets being recognized as “foreign” and cleared from the circulation by the reticuloendotheliaf system. 01990 Academic F’ress. Inc.

INTRODUCTION Shortened platelet survival has been observed in a variety of immunological disorders (Harker, 1978; Kinlough-Rathbone ef al., 1983; Shulman and Jordan, 1987)but the mechanisms involved have not been established. Vessel wall injury is a common feature of many of the conditions that have been associated with shortened platelet survival (Kinlough-Rathbone et al., 1983). Vascular injury in association with immune complex formation has been found in animals injected with serum albumin. In kidneys, the injuries reported range from minimal glomerular changes to severe proliferative glomerulonephritis (Dixon et al., 1958; Germuth, 1953; Kniker and Cochrane, 1%5; Neild et al., 1984). Swelling and separation of endothelial cells and endothelial cell denudation have been observed in the aortae and other arteries of animals with serum sickness (Germuth, 1953; Kniker and Cochrane, 1965; Sharma and Geer, 1977; Neild et al., 1984). In some cases,leukocytes and platelets have been observed at focal sites of injury around vessel orifices and branches (Kniker and Cochrane, 1968; Sharma and Geer, 1977). It has been proposed that platelets contribute to vascular injury by releasing materials such as serotonin, histamine, and permeability factors from their granules and by forming products of arachidonic acid degradation. In immune complex diseasesassociated with vascular injury, platelets may be cleared from the circulation because (a) they are consumed at sites of vessel injury, (b) they are modified during their interaction at injury sites so that when they return to the circulation they are cleared by the reticuloendothelial system, or (c) they interact directly with immune complexes in the circulation leading to their alteration and 109 0014-4&30/!90 $3.00 Copyri&t Q 1990 by Academic Press. Inc. AU rights of reproduction in any form reserved.

110

PARBTANI

ET AL.

rapid clearance. In this study we examined the possible roles of vessel injury and platelet interaction with immune complexes as factors influencing platelet survival in rabbits in which immune complexes were formed in the circulation by the injection of human serum albumin (HSA). MATERIALS AND METHODS Animals

Male New Zealand White rabbits weighing between 2.2 and 3.0 kg were used throughout the study. Platelet

and White Blood Cell Counts

Blood was collected from each rabbit through a marginal ear vein into EDTAcontaining tubes (Vacutainer; Becton-Dickinson; New Jersey) and platelet and white cells were counted using an S-plus Coulter counter (Coulter Electronics of Canada, Ltd., Oakville, Ontario). Serum Albumin

In initial experiments, rabbits were injected with bovine serum albumin (BSA, Fraction V; Miles Laboratories, Elkhart, IN: 250 mg/kg). However BSA caused a fall in platelet and white cell counts and the animals became pyrexic, possibly because of endotoxin contamination of the BSA. In subsequent experiments, pyrogen-free human serum albumin (HSA; Canadian Red Cross Society; manufactured by Connaught Laboratories, Ltd., Willowdale, Ontario) was injected in place of BSA. In experiments in which radioactive HSA was used, the HSA was labeled with 1251(Amersham, UK) by the method of McFarlane (1958). Induction

of Immune

Complex Disease in Rabbits

Rabbits were injected through a marginal ear vein with HSA (250 mg/kg) mixed with 5 pg of 12$HSA (1 x lo7 cpm). Control rabbits were injected with an equal volume of pyrogen-free, sterile saline (0.9% sodium chloride; Squibb Canada, Inc., Montreal). Clearance of 12’Z-HSA from the Circulation

The clearance of 1251-HSAwas determined by measuring the radioactivity in a l-ml sample of blood collected from an ear vein 2 hr after the HSA injection and at daily intervals from Day 1to Day 17. The amount of i2’I-HSA in the 2-h sample was arbitrarily assigned a value of 100% and all other values were related to it. Detection of Circulating Immune Complexes

Antibodies

to HSA and Circulating

Antibodies to HSA were measured using the interfacial ring test (Hudson and Hay, 1980) and circulating immune complexes were measured by the method of Farr (1958). Preparation

of Tissues for Transmission

and Scanning Electron

Microscopy

Anesthetized rabbits were killed by perfusion-fixation (Groves et al., 1982) and samples of aorta and renal cortex were prepared for scanning (aorta) and trans-

IMMUNE

COMPLEXES

AND

mission (renal cortex) electron microscopy et al., 1982). Antiserum

PLATELET

(Richardson

SURVIVAL

and Moore,

111

1980; Groves

to HSA (anti-HSA)

Rabbits were given an initial intramuscular injection of 500 pg of HSA in Freund’s complete adjuvant (GIBCO Laboratories, New York, NY) followed by three injections of HSA (500 pg each) in Freund’s incomplete adjuvant at 2-week intervals. Four days after the last injection, 5 ml of blood was collected and the serum was tested for the presence of antibodies to HSA by the interfacial ring test (Hudson and Hay, 1980). Rabbits whose serum was positive for the precipitating antibodies to HSA were bled the following day and the serum obtained was tested again for the presence of antibodies to HSA and screened for contaminating antibodies to other human serum proteins by double diffusion in gel (Ouchterlony and Nilsson, 1986). The antiserum showed only one precipitating band and this in turn showed complete identity with a single precipitin line obtained when tested against normal human serum. The serum containing HSA antibodies was decomplemented by heating at 56°C for 30 min; it was then centrifuged at 8000g for 2 min in an Eppendorf centrifuge (Brinkman, Rexdale, Ontario). The supematant antiserum was stored in 2-ml aliquots at -70°C. Determination Equivalence

of the Antibody Content of the anti-HSA Serum and the Zone of for the HSA:Anti-HSA Immune Complexes

The antibody content of the anti-HSA serum was determined by the method of Hudson and Hay (1980), except that the amount of antigen (HSA) used to react with 100 pl of the anti-HSA serum ranged between 0 and 100 pg. The point of equivalence was considered to be the point at which the amount of antigen used gave the maximum amount of precipitate when reacted with 100 p,l of the antiHSA serum. The anti-HSA serum used for the experiments described in this paper gave the maximum precipitation when 100 t~l of anti-HSA was reacted with 7.5 pg of HSA. The quantity of immunoglobulin present in the anti-HSA serum was approximately 1.2 mg/ml. Preparation

of Platelets

Suspensions of washed rabbit platelets were prepared as described previously (Ardlie et al., 1971). For studies of platelet secretion, platelets suspended in calcium-free Tyrode solution were labeled with [14C]serotonin (14C-5-hydroxytryptamine-3’-creatinine sulfate 50 pCi/pmole; Amersham/Searle, Arlington Heights, IL) (Greenberg et al., 1975). In most experiments platelets were simultaneously labeled with N+%G4 (2W500 uCi/mg of chromium; Amersham/ Searle; 50 &i of %Zr was used to label platelets from the blood of each rabbit). The labeled platelets were washed in calcium-free Tyrode solution and resuspended at a final concentration of 0.5 x lo6 platelets/p1 in platelet-poor plasma (PPP) prepared from blood collected into heparin (5 U/ml; Hepalean, Harris Laboratories, Brantford, Ontario). In studies of the distribution of platelets on Stractan density gradients, some of the washed platelets were resuspended in Tyrode solution containing albumin and apyrase (Ardlie et al., 1971).

112

PARBTANI

In Vitro Platelet Aggregation Immune Complexes

ET AL.

and Release Induced by HSA:Anti-HSA

Suspensions of platelets doubly labeled with ‘lCr and [14C]serotonin were resuspended in heparinized platelet-poor plasma (0.5 x lo6 platelets/$). (Immune complexes would not aggregate rabbit platelets in the absence of a source of complement (Henson and Ginsberg, 1981).) One milliliter of platelet suspension containing 5 p& imipramine (Sigma Chemical Co., St. Louis, MO) was stirred at 37°C for 1 min in a Payton aggregation module (Payton Associates, Scarborough, Ontario) and then 100 ~1 of anti-HSA serum was added. One minute later, 1251HSA was added at a concentration sufficient to form immune complexes either at antigen-antibody equivalence, at antibody excess, or at antigen excess. Six minutes after addition of the antigen (‘251-HSA), the mixture was centrifuged at SOOOg for 2 min. The amounts of [14C]serotonin and “Cr in the supematant fluid were measured and the extent of release of [14C]serotonin and lysis (51Cr) was expressed as a percent of the 14Cand 51Crin the original platelet suspension. The 1251 in the pellets and the supematant was measured to determine the amount of 1251-HSAassociated with platelets. In some experiments, platelets that had been incubated with immune complexes were layered onto Stractan density gradients (Cieslar et al., 1979) to determine the density distribution of platelets and immune complexes (see later). For antigen control, anti-HSA was replaced by 100 t.~lof decomplemented normal rabbit serum, and for the antibody control, HSA was replaced by an equal volume of heparinized platelet-poor plasma. Determination

of the Association

of Zmmune Complexes

with the Platelets

The association of immune complexes with platelets suspended in heparinized platelet-poor plasma was estimated by measuring the amount of 1251-HSAassociated with the platelets, expressed as a percentage of the total amount of 1251HSA that was present in the reaction mixture. Initially this was done by centrifuging the platelets after they had reacted with the immune complexes and counting the “‘1 associated with the pellet and in the supemate. To ensure that the radioactivity measured was not solely attributable to cosedimentation of immune complexes with the pelleted platelets, the platelet suspensionsin Tyrode-albumin or in heparinized platelet-poor plasma in the presence and absence of immune complexes were separated on Stractan density gradients (Cieslar et al., 1979). The platelets labeled with “Cr were reacted with “‘1-HSA:anti-HSA immune complexes and 2.5 ml of the mixture was layered over 2.5ml layers of 18, 15, and 13.5% Stractan. The tubes were then centrifuged at 1600g for 45 min. Five fractions (2 ml each) were separated and the radioactivity (12’1 and 51Cr) was measured in each fraction and expressed as a percentage of the total radioactivity of each isotope. The top fraction was designated as fraction 1 and the bottom fraction as fraction 5. 1251-HSA,“Cr-labeled platelets, and 1251-HSA-anti HSA immune complexes were also layered separately on the Stractan gradients to determine their individual distributions. Platelet

Survival Studies

Platelets were labeled with Na, “Cr04 in calcium-free Tyrode solution containing 0.35% human albumin (150 t&i of 51Cr was used to label platelets obtained

IMMUNE

COMPLEXES

AND

PLATELET

SURVIVAL

113

from the blood of each rabbit). The labeled platelets were washed once with calcium-free Tyrode-albumin solution and resuspended at a final concentration of 2 x lo6 platelets/p.1in platelet-poor plasma prepared from blood collected into acid citrate-dextrose. “Cr-labeled platelets (3.2 x 10” platelets) were injected into each rabbit through a marginal ear vein. Samples of blood were collected 2, 3, 21, 23,45,69, and 96 hr after the platelets were injected, and the radioactivity in each sample was determined. The radioactivity in the 2-hr sample was assigned a value of 100% and the radioactivity in the other samples was expressed as a percentage of this. Platelet survival was calculated using the gamma function described by Murphy and Bolling (1978). The recovery of “0-labeled platelets in the circulation 2 hr after the injection of ‘lCr-labeled platelets was calculated by expressing the radioactivity in the 2-hr sample as a percentage of the injected radioactivity and based on an estimated blood volume in rabbits of 58 ml/kg (Reimers et al., 1973). Survival of 51Cr-labeled platelets was studied in HSA-induced immune complex disease as follows: 1. Platelets from normal rabbits were injected into control rabbits or into recipient rabbits 7, 8, 10, 14, and 21 days after the latter had received an injection of HSA. 2. Platelets were harvested from rabbits 7, 8, 10, 14, and 21 days after an HSA injection; the harvested platelets were then injected into normal rabbits or into rabbits that had been injected with HSA for the equivalent periods. RESULTS Effects

of BSA and HSA on Platelet

and White Blood Cell Counts

In preliminary experiments, rabbits were injected with BSA (250 mg/kg). Both platelet and white cell counts fell after the injection of BSA (Figs. la and lb). The white cell count then increased and was significantly higher than the preinjection value (Fig. lb). The initial thrombocytopenia and the leukocytosis persisted for up to 3 days. Because of the possibility that the BSA might be contaminated with endotoxin, the pyrogenicity of the BSA was assessed.BSA (50 mg/kg, one-fifth the dose used in our experiments) caused a 2.4 ? 0.2”C (mean 2 SD, n = 6) increase in the temperature of rabbits. When pyrogen-free HSA was used in place of BSA, these changes in platelet and white cell counts were not observed (Figs. 2a and 2b) and there was no appreciable increase in the temperature of the rabbits (0.08 f O.OS”C,n = 6). Subsequently all experiments were done using HSA. In Vitro Studies Platelet

aggregation

and release reaction

in response

to immune

complexes.

The addition to washed rabbit platelets resuspended in heparinized platelet-poor plasma of anti-HSA serum followed by HSA at a ratio that produced antigenantibody equivalence caused extensive platelet aggregation and some release (7%) of [14C]serotonin (Fig. 3a). Platelet aggregation was less extensive when antigen and antibody were added at a ratio that produced 10 times antibody excess (Fig. 3b) or 5 times antigen excess. Platelet aggregation did not occur when the antigen and antibody were added at 20 times antibody excess or at 10 times antigen

114

PARBTANI

ET AI.,.

PLATELET COUNT (x 108/L) 1

Q

600-

1 1

+

460t + 200

+ t 0 !C; 2 HR BEFORE 5 INJ WHITE CELL COUNT (x 100/L)

10

15

b

1612+

6-

o-

?+ 2hi BEr$:RE

I 5

I 10

I 15

DAYS AFTER BSA INJECTION 1. (a) Platelet count (mean f SEM) in blood obtained before and after injections into rabbits of BSA (250 mg/kg). Asterisks indicate the counts which were significantly different (P < 0.05) from the counts obtained with blood collected before the injections of BSA. (b) White cell counts (mean f SEM) in blood obtained before and after injections of BSA (250 mgikg). Asterisks indicate the counts which were signiticantly different (P < 0.05) from the counts obtained with blood collected before the injections of BSA. FIG.

excess. Addition of antigen or antibody alone did not cause platelet aggregation or release of [14C]serotonin. Association of antigen (‘25Z-ZiZSA)with platelets. There was little association of 1251-HSA with platelets in the absence of anti-HSA (1.35 + 0.1% of the added 12’I-HSA; mean + SEM, n = 13). The association of 12’I-HSA with stirred platelets was significantly greater when antigen and antibody were added at equivalence (15.7 k 0.5% of the added ‘251-HSA, n = 15) or at 10 times antibody excess (5.7 + 0.3%, n = 4). (The percentage of immune complexes associated with platelets at antigen-antibody equivalence was always greater in unstirred samples (36.3 + 5.8%, n = 4) than with platelets that were stirred and allowed to aggregate (19.9 + 3.6%, n = 4).

To ensure that the radioactivity pelleted with the platelets actually represented the association of antigen-antibody complexes formed at equivalence, the mixture was separated on Stractan density gradients. The distribution on the gradient of

IMMUNE

COMPLEXES

AND PLATELET

(x 1OVL) 600

115

SURVIVAL

a

1

600 ltu”Sttx/xitA ““o”J t BELyRE

5

WHITE CELL (x log/L) 1

COUNT

15

10

b

6 4J OJ I , BEFORE INJ

I 5 DAYS

I 10

AFTER

HSA

I 15 INJECTION

FIG. 2. (a) Platelet count (mean 2 SEM) in the blood of rabbits before and after injections of HSA (250 mg/kg). (b) White cell count (mean * SEM) in the blood of rabbits before and after injections of HSA (250 mg/kg).

washed 51Cr-labeled platelets suspended in Tyrode-albumin is shown in Table I; most of the platelets separated in the fourth fraction. When 1251-HSA in Tyrodealbumin was centrifuged on the gradient only a small amount of this material entered the gradient. In contrast, antigen-antibody complexes formed at equivalence were centrifuged through the gradient and appeared in the fifth fraction; when the complexes were incubated with the %--labeled platelets suspended in LIGHT TRANSMISSION 90 b

ANTI-HSA

0 01

8

01

6

8

MINUTES

FIG. 3. (a) Washed rabbit platelets resuspended in heparinized PPP aggregated in response to anti-HSA and HSA added at a ratio that produced antigen-antibody equivalence. Zero and 100 percent light transmission were initially set using platelet suspension and platelet suspending medium, respectively. (b) Washed rabbit platelets suspended in heparinized PPP aggregated poorly in response to antigen and antibody at a ratio that produced 10 times antibody excess.

116

PABBTANI

ET AL.

TABLE I The Association of Immune Complexes with Platelets in the Absence and in the Presence of a Source of Complement Percentage in platelet subpopulations

jiCr Platelets

“‘1-HSA (no platelets) Experiment I-Tyrode

Fraction 1

‘2JI-HSA + anti-HSA (no platelets) albumin

70.0 + 3.0

12.3 ” 5.4

I 9.8 + 4.3 Cr 2.2 f 0.3

25.0 k 2.5

5.0 f 1.5

I 8.0 + 5.0 Cr 1.8 k 0.3

3.5 + 0.5

3.0 f 0

I 4.0 + 1.4 Cr 10.0 k 1.5

0.7 2 0.05

2.0 + 0

I 4.5 * 1.5 Cr 69.0 2 5.0

0.2 + 0.01

78.0 + 6.5

I 73.0 2 10.0 Cr 17.0 + 6.0

220 3k 1.0 18 2 0 4 68 f 3.0 5 9 + 2.0 Experiment 2-Heparinized Fraction 1

PPP

76.0 + 2.5

62.0 + 4.4

I 13.0 f 2.7 Cr 18.0 k 3.5

20.0 + 2.1

25.0 f 2.5

I 8.2 + 2.3 Cr 8.3 * 1.9

3.0 2 0.4

6.0 f 0.9

I 3.6 + 1.0 Cr 3.4 + 1.0

0.8 f 0.3

2.7 f 0.7

I 70.0 ” 6.4 Cr 68.0 + 5.2

0.5 f 0.3

2.0 + 0.5

I 4.5 r 2.2 Cr 2.0 + 0.3

8 f 1.9 4 + 1.1 5 + 2.3 4 78 -t 6.5 5 4*

1.3

Wr Platelets + ‘=I-HSA + anti-HSA

Note. Mean values + SEM from five to seven experiments. In these experiments the ‘*‘I-HSA and anti-HSA were added at equivalence.

Tyrode-albumin, their distribution on the gradient was almost the same as when platelets were not present (Table I). However, when a similar study was carried out using heparinized plasma as the suspending medium, neither ‘=I-HSA alone nor ‘*%HSA in the presence of anti-HSA centrifuged through the gradient. However, in the presence of %rlabeled platelets, a large percentage of the *“I in the presence of anti-HSA appeared in the same fraction on the gradient as the platelets (Table I). In Vivo Studies Clearance of ‘25Z-HSA from the circulation and the amount of circulating imfrom the circulation and the mune complexes. The rate of clearance of “‘I-HSA appearance of circulating immune complexes are shown in Fig. 4. Immune complexes could be detected in the circulation between Days 5 and 7, reached a peak

IMMUNE

COMPLEXES

AND

PLATELET

SURVIVAL

117

PERCENT 100-q

So-

125i - HSA

INJECTED

60-

125 I- HSA

i

5

lb

AS COMPLEX

15

DAYS AFTER HSA INJECTION FIG. 4. The closed circles (0) indicate the amount of “%HSA (mean f SEM) in the circulation of rabbits after the injection of 250 mgkg HSA. The amount of “‘1-HSA in samples collected 2 hr after the HSA injection was taken as 100%. The closed triangles (A) indicate the percentage of “‘1-HSA (mean 2 SEM) in the serum samples in the form of immune complexes. Antibodies to HSA were detectable between Days 7 and 9 and could still be detected in the circulation (in very small amounts) up to Day 17 after the HSA injection.

between Days 11and 14, and were cleared from the circulation between 15 and 17 days after the HSA injection. Free antibodies to HSA were detected in the circulation between Days 7 and 9 and were still present 17 days after the injection of HSA. Survival of %4abeled platelets. “Cr-labeled platelets were prepared from HSA-treated rabbits at a time when the amount of immune complexes in the circulation was increasing (7 to 10 days). When these ?Jr-labeled platelets were injected into another group of rabbits that had been injected with HSA 7-10 days previously the platelets had a shorter than normal life span (Table II). Platelets from normal animals also had a shortened survival when they were injected into animals that had circulating complexes (7-10 days after HSA). Platelets harvested from HSA-rabbits at 7-g days had a shortened survival upon injection into control rabbits. If the platelets were harvested from HSA rabbits at 10days, their survival was reduced in HSA rabbits, but there was no significant reduction in their survival upon injection into control rabbits. Platelets prepared from either control animals or from HSA-treated animals at a time when complexes were being cleared from the circulation (14 to 21 days) did not have a shortened platelet survival when injected into HSA-treated rabbits at 14-21 days or into control rabbits (Table II). Morphology. Extensive examination of aortae and kidneys by scanning and

118

PABBTANI

ET AL.

TABLE II Effect of HSA Injections on Platelet Survival Platelet survival (hr) Platelets from HSA-rabbits injected into HSA-rabbits

Control platelets injected into HSA-rabbits

HSA-platelets injected into control rabbits

7and8 (pooled)

52.5 2 6.1 (8) P < o.oos*

51.1 ” 7.1 (9) P < 0.005*

61.8 f 5 (10) P < o.os*

10

58.0 +- 5.9 (10) P < 0.02’

52.9 + 3.5

69.4 ” 7.1 (13)

P < 0.001*

70.4 2 13.2 (4)

73.7 2 2.1 (9

Days after HSA injection

14 and 21

(16)

67.8 f 7.8

(6)

Note. Mean values + SEM. The numbers in parentheses indicate the number of animals in each experiment. The recoveries of slCr-labeled platelets in the circulation 2 hr after their injection into rabbits that had HSA injections 10 days previously were 73.1 2 2.0% (HSA platelets into HSArabbits), 85.8 + 2.5% (control platelets into HSA-rabbits), and 90.6 + 4.5% (HSA platelets into control rabbits). * Significance of the difference in platelet survival compared with the survival of control platelets injected into control rabbits. Platelets from control rabbits survived for 72.6 * 2.7 hr (n = 23) upon injection into other control rabbits.

transmission electron microscopies showed no apparent morphological or microthrombi (Figs. 5a and Sb).

changes

DISCUSSION The results of these studies show that the production of immune complexes in rabbits by injections of human serum albumin that is free of contamination by bacterial endotoxin leads to (a) little change in circulating platelet or white cell counts and (b) a reduction in platelet survival in animals with circulating immune complexes that is not associated with detectable vascular injury or thrombosis. Platelets

and Immune

Complexes

The in vitro studies show that the complexes formed in these studies interacted with washed rabbit platelets only when they were suspended in heparinized plasma that contains complement. It has been established previously that a source of complement is necessary for immune complexes or aggregated IgG to interact with rabbit platelets (Henson and Ginsberg, 1981). Under the conditions of the present experiments we did not obtain any evidence of platelet lysis. On the basis of the in vitro studies in which it was shown (a) that immune complexes formed at antigen-antibody equivalence are able to interact with platelets and cause them to aggregate and (b) that a large percentage of 12’1-HSA separates with ‘%r-labeled platelets on Stractan density gradients provided anti-HSA is present, it might be expected that immune complexes would interact with platelets in vivo at the time when circulating immune complexes were detectable and that the interaction of platelets with these complexes could lead to a reduction in platelet survival. The observation that rabbit platelets react maximally with immune complexes when

IMMUNE

COMPLEXES

AND

PLATELET

SURVIVAL

119

Fro. 5. (a) Scanning electron micrograph of the endothelial surface of a segment of abdominal aorta of a rabbit 10 days after an injection of HSA. Occasional leukocytes and RBCs are observed but the endothelial cells appear to be morphologically normal (X 1250). (b) Transmission electron micrograph of a glomerulus from a rabbit 10 days after an injection of HSA. No thrombi are observed (X 1500).

120

PARBTANI

ET AL.

the complexes are at antigen-antibody equivalence is in keeping with findings obtained with pig or human platelets (Clark et al., 1980, 1982). Platelet

Survival

When platelets were harvested from animals during the period in which there were identifiable circulating immune complexes and injected into animals in which complexes continued to form for several more days, the platelets survived for a shorter period than normal platelets injected into normal animals. Platelets harvested from animals after immune complexes were cleared from the circulation survived normally in animals which no longer had circulating complexes. Platelets from normal rabbits had a shortened platelet survival upon injection into animals with circulating immune complexes. Taken together, these data imply that the interaction of immune complexes with platelets leads to their clearance from the circulation. However, when platelets from animals with circulating immune complexes were isolated, labeled, washed, and infused into normal animals, they survived in the circulation for a longer period than when they were injected into HSA-treated animals. This may mean that the immune complexes elute from platelets when the platelets are washed and resuspended in normal plasma. The mechanisms responsible for the reduction in platelet survival caused by exposure to immune complexes in vivo have not been established. The maximum loss of platelets from the circulation occurred during the time when immune complexes were forming. The lack of morphological evidence in this study of vascular injury in animals with circulating immune complexes indicates that the shortened platelet survival is unlikely to have resulted from platelet accumulation at sites of vessel wall injury. Nor can the loss of platelets be attributed to their accumulation in kidneys damaged by the immune complexes since the kidneys appeared to be morphologically normal and platelet aggregates were not observed in the microvasculature. In several of the previous studies that have examined the effect of immune complex formation in experimental animals, the animals have received an injection of endotoxin before the injection of BSA (Kniker and Cochrane, 1965; Neild et al., 1984). Many commercial preparations of BSA also contain endotoxin. Thus it is possible that some of the features observed could be attributed to the effects of endotoxin. The use of endotoxin-free HSA in the present studies would have avoided this effect and may account, in part, for the lack of detectable vessel wall damage and thrombi in the microvasculature in these animals. The shortened platelet survival is more likely a consequence of the rapid clearance of platelets from the circulation by the reticuloendothelial system and may be influenced, in part, by the transient sequestration of platelets in the lungs. Under other conditions, it has been shown that when blood cells have immune complexes on their surface, they are rapidly cleared from the circulation (Shulman and Jordan, 1987). It is possible that immune complexes may associate with the platelet surface in such a way that the platelets are recognized as foreign and cleared from the circulation. Indeed, should this occur, it might provide some protection for the vessel wall against immune complex-induced damage since the complexes would be cleared from the circulation with the platelets. It can be concluded from these experiments that the formation of immune complexes does not necessarily cause vascular injury and that the reduction in platelet survival in animals with circulating complexes does not therefore result from the consumption of platelets at sites of vessel injury but is more likely due

IMMUNE

COMPLEXES

AND

PLATELET

SURVIVAL

121

to the interaction of platelets with immune complexes that alters the platelets sufficiently to lead to their rapid clearance from the circulation. ACKNOWLEDGMENT This work was supported by a grant-in-aid (MT 1309) from the Medical Research Council of Canada.

REFERENCES ARDLIE, N. G., PERRY,D. W., PACKHAM, M. A., and MUSTARD, J. F. (1971). Influence of apyrase on stability of suspensions of washed rabbit platelets. Proc. Sot. Exp. Biol. Med. 136, 1021-1023. CIESLAR, P., GREENBERG, J. P., RAND, M. L., PACKHAM, M. A., KINLOUGH-RATHBONE, R. L., and MUSTARD, J. F. (1979). Separation of thrombin-treated platelets from normal platelets by densitygradient centrifugation. Blood 53, 867-874. CLARK, W. F., TEVAARWERK, G. J. M., and REID, B. D. (1982). Human platelet-immune complex interaction in plasma. J. Lab. Clin. Med. 100, 917-931. CLARK, W. F., TEVAARWERK, G. J. M., BANDALI, K., LINDSAY, R. M., and LINTON, A. L. (1980). Platelet aggregation and release associated with immune complex formation in pig plasma. .Z. Lab. Clin.

Med.

96, 654-665.

DIXON, F. J., VA~QUEZ, J. J., WEIGLE, W. O., and COCHRANE, C. G. (1958). Pathogenesis of serum sickness. Arch. Pathol. 65, 18-28. FARR, R. S. (1958). A quantitative immunochemical measure of the primary interaction between I*BSA and antibody. J. Infect. Dis. 103, 239-262. GERMUTH, F. G. (1953). A comparative histologic and immunologic study in rabbits of induced hypersensitivity of the serum sickness type. J. Exp. Med. 97, 251-282. GREENBERG, J., PACKHAM, M. A., CAZENAVE, J.-P., REIMERS, H.-J., and MUSTARD, J. F. (1975). Effects on platelet function of removal of platelet sialic acid by neuraminidase. Lab. Invest. 32, 476a4.

GROVES, H. M., KINLOUGH-RATHBONE, R. L., RICHARDSON, M., J@RGENSEN,L., MOORE, S., and MUSTARD, J. F. (1982). Thrombin generation and fibrin formation following injury to rabbit neointima. Studies of vessel wall reactivity and platelet survival. Lab. Invest. 46, 60542. HARKER, L. A. (1978). Platelet survival time: Its measurement and use. In “Progress in Hemostasis and Thrombosis” cr. H. Spaet, Ed.), Vol. 4 pp. 321-347. Grune & Stratton, New York. HENSON, P. M., and GINSBERG, M. H. (1981). Immunological reactions of platelets. In “Platelets in Biology and Pathology 2” (J. L. Gordon, Ed.), pp. 265-308. Elsevier/North-Holland, Amsterdam. HUDSON, L., and HAY, F. C. (1980). “Practical Immunology,” 2nd ed. Blackwell, Oxford. KINLOUGH-RATHBONE, R. L., PACKHAM, M. A., and MUSTARD, J. F. (1983). Vessel injury, platelet adherence and platelet survival. Arteriosclerosis 3, 529-546. KNIKER, W. T., and COCHRANE, C. G. (1968). The localization of circulating immune complexes in experimental serum sickness. The role of vasoactive amines and hydrodynamic forces. J. Exp. Med. 127, 1 W-136.

KNIKER, W. T., and COCHRANE, C. G. (1%5). Pathogenic factors in vascular lesions of experimental serum sickness. J. Z&p. Med. 122, 83-98. MCFARLANE, A. S. (1958). Efficient trace labelling of proteins with iodine. Nature (London) 182, 53. MURPHY, E. A., and BOLLING, D. R. (1978). Platelet survival. In “Platelet Function Testing” (H. J. Day, H. Hohnsen, and M. B. Zucker, Eds.), No. (NIH) 78;1087, pp. 561-567. DHEW, Washington, DC. NEILD, G. H., IVORY, K., and WILLIAMS, D. G. (1984). Glomerular thrombosis and cortical infarction in cyclosporin-treated rabbits with acute serum sickness. Brit. J. Exp. Pathol. 65, 133-144. OUCHTEIUONY, c)., and NILSSON, L.-A. (1986). Immunodiision and immunoelectrophoresis. In “Handbook of Experimental Immunology,” (D. M. Weir, Ed.), 4th ed., Vol. 1, 32.1-32.50. Blackwell, Oxford. REIMERS, H.-J., BUCHANAN, M. R., and MUSTARD, J. F. (1973). Survival of washed rabbit platelets in vivo. Proc. Sot. Exp. Biol. Med. 142, 1222-1225. RICHARDSON, M., and MOORE, S. (1980). Preparation of large curved biological surfaces for scanning electron microscopy. Artery 6, 409417. SHARMA, H. M., and GEER, J. C. (1977). Experimental aortic lesions of acute serum sickness in rabbits. Amer. J. Pathol. 88, 255-266. SHULMAN, N. R., and JORDAN, J. V. (1987). Platelet kinetics. In “Hemostasis and Thrombosis. Basic Principles and Clinical Practice,” (R. W. Colman, J. Hirsh, V. J. Marder, and E. W. Salzman, Eds.), 2nd ed., pp. 43 l-45 1. Lippincott, Philadelphia.