125I anti-immunoglobulin binding assay for the detection and characterization of anti-platelet antibodies

25I Anti-Immunoglobulin Binding Assay or the Detection and Characterization )f Anti-Platelet Antibodies ean Kirkley and John W. Fabre*

LBSTRACT: The binding assay as described in this paper is a vewj z,ersa, % system, and in this J:ud~ it has been evaluated specifically for the detection of allo- or autoantibodies to platelets in ma,. The basic assay involves the incubation of a standard number of platelets u'ith dilutions of te, t sera and the detection of platelet bound immunogiobulin by a secor/d incuba:ion u'ith ~'--'!la&hd. immunoadsorbent purified rabbit F(ab') anti-h~'lman F(ab'J~ (RAHL Of most importance, by varying the number of target platelets in the titr~ttions and looking for binding plateaus, one can readily define conditions of optimum sensitivity for particular serum /platelet combinations. In addition, the assay can be used in conjunction with quantitative absorptions to subdicid¢ complex sera into subspecificities and to give at, estimate of the relative amounts of particular antigens on different platelets or other tissue m" cell suspensions. One can also use saturating concentrations of R A H i;: the second incubation, in which case the amount of platelet bound radioactivity is directly relaIed to the amount of first antibody bound to the platelets, and this can be manipulated to give information about serum antibody concentrations and amounts of antigen on the target tissue. The problem of ABO antibodies in this system, optimal conditions for platelet storage for the assay, and techniques for reducing assay backgrounds resulting from immunoglobulin adsorbed to the platelet surface are all evaluated. ~BBREVIATIONS

FCS RBC WBC

foetal calf se
PBS RAH

phosphate buffered saline immunoadsorbent purified rabbit F(ab')., anti-human F(ab')-,

INTRODUCTION The detection of antibodies to platelet antigens is of importance both to characterize platelet specific antigen systems, and also to identify pathological conditions where allo- or autoimmunization to platelet antigens results in thrombocytopenia. There is no entirely satisfactory assay for anti-platelet antibodies, since most systems depend on various phenomena secondary to the binding of antibody, such as complement fixation [1, 2], release of p l a t e l e t factors [3], agglutination [4], and others [5, 6]. Various binding techniques have From the Nuffield Department of Surgery. University of Oxford. John Radcliffe Hospital. Headington. Oxford OX3 9DU, England. Address requests for reprints to Dr. J.W. Fabre. Nuffield Department of Surgery. John Radcli/l~ Hospital. Oxford OX3 9DU. England. *WellcomeSenior Clinh'al Research Fellow. Received 1980.

Human Immunology 4, 369-381 (1980) ~) Elsevier North Holland, In,.-., 1980 52"Vanderbilt Ave.. N e w Yor~, NY 1001.7

369 0198-8859t~04369- | ~$2.25

i70

J~an Kirkley and John W. Fabre been described, including a histochemical technique using p,~roxidase labeled anti-human immunoglobulin [7], an assa~ for platelet bound immunoglobulin using inhibition of 125I labeled anti-human immunoglobulir, [8], and an assay where the proportion of added '251 labeled anti-human immunoglobulin which is platelet bound is measured [9]. These assays, as described, are technically complicated, are not easily interpreted, and lack the versatility necessary for analysis of complex sera. The assay described in this paper is a relatively simple '2"~I antiimmunoglobulin binding assay which has been used extensively with a variety of different targets such as lymphocytes [10-12], erythrocytes [13], and tissue homogenates [ 14]. It is an easily interpreted and very versatile system which can be used both in direct binding assays, and also with quantitative absorptions for the analysis of complex sera with multiple speci:~cities.

~[ATERIALS A N D M E T H O D S Preparation of platelets. All procedures were carried out at 4"C or in ice, using siliconized glassware or plastic. Blood was collected into EDTA-containing tubes, diluted with 3 volumes of 0.9% N~CI, and centrifuged at 200 x g for 20 min. The platelet rich supernatant was removed, leaving a l-cm margin above the cell pellet, and centrifuged again as above to ensure complete depletion of erythrocytes (RBC*) and leucocytes (WBC*). The platelet rich supernatant was removed as above, and the platelets spun down at 2000 x g for 15 rain. The platelets were resuspe: Jed in 0.9% NaCl to approximately the original volume of blood arm counted on a haemocytometer using a phase-contrast microscope. The yield was generally l0 s platelets/ml of blood. The platelets were then spun down again a,~ above, and resuspended to J0S/ml in Ca ++ and MS ++ free phosphate buffered saline (PBS*) (Gibco-Biocult, Paisley, Scotland) with 0.1°~ sodium azide. They were then left overnight at 4"C for processing as targets in the binding assay the following day. This was done purely for convenience.

Sera. Normal sera came from healthy, nontransfused males. Broadly crossreactive HLA antisera (GRO, McG and STO) were a gift of Dr. A. Ting of the Tissue Typing Laboratory of this department. These sera were inactivated by heating at 56°C for 30 rain. lmmunoadsorbent purified rabbit F(ab')z anti-human F(ab')z (RAH*) was prepared as follows. Human immunoglobulins were precipitated from normal human serum three times with 16% NazSO4 at room temperature. The immunoglobulins were pepsin degraded for 20 hr at 37°C in 0.1 M Na acetate buffer, pH 4.5, using an enzyme/substrate ratio of 4%. The pepsin-degraded material was passed through a G-200 column (Pharmacia, Sweden) and the F(ab')2 peak was pooled. The F(ab')z was shown to be pure by sodium dodecyl sulphate electrophoresis in polyacrylamide disc gels [ 15]. One hundred milligrams of F(ab'h was coupled to 10 ml of CNBr activated sepharose 4B essentially as described by the manufacturers (Pharmacia, Sweden). Rabbits were hyperimmunised with the F(ab')z in Freund's complete adjuvant. The antiserum was ultracentrifuged to remove cellular debris, heat inactivated at 56°C for 30 rain, and then passed through the human F(ab')2 column. When the column was saturated, as indicated by Ouchterlony plates, rb,e column was washed, and the adsorbed antibody eluted with I M propionic acid. The eluted antibody was immediately ueutralised using 2 M Tris, and was pepsin degraded and filtered through G-200 as above. It was stored at 1 mg/ml at -40oc. Pepsin degradation and/or gel filtration of eluted antibody has been recommended to remove aggregates caused by the low p H elution [ 10].

Binding Assay for Platelet Antibodies

37 !

The R A H was iodinated using the chloramine T method, the reaction being stopped with excess tyrosine instead of reducing agents, to maintain the integrity of the labeled F(ab')z. Briefly 1 izl of t2n[ (I0 ~ttl of Code IMS 30, Radiochemical centre, Amersham) was ~,laced at the: bottot~ of a glass tube and to this was added 20/zl of freshly prepared chloramine T at 2 mg/ml in 0.3 M phosphate buffer, p H 7.3, followed by 25 #l (25/zg) of RAH. After 2 rain, the reaction was stopped by the addition of 50 ~tl o f tyrosine at 0.5 mg/ml in the above-mentioned phosphate buffer. After a further 2 rain, I00 #1 o f I0% bovine serum albumin (BSA) in PBS was added, and unbound iodine was separated from iodinated R A H on a prepacked 9-ml G-50 column gPD-10, Pharmacia, Sweden). t2~I R A H was stored at 4°C in 0.5% BSA 0.15 M NaCI, 0.025 M Tris. 0.07% Na azide, p H 7.4, for up to 3 wk.

Binding assay. This was essentially as described by Morris and Williams [ 10] for the analysis of rabbit anti-rat brain xenosera ~ssayed on thymus cells, with a few modifications. All procedures were carried out at 4°C or on ice. The platelets to be used as targets were prepared the day before the assay as described above, and stored overnight at 4°C at 109/ml. in all except one experiment (Figure 3), the immunoglobulin nonspecifically adsorbed to the surface of the platelets was blocked by adding R A H to a concentration of 10 /zg/ml of platelet suspension, and incubating for half an hour. As will be discussed, this was not essential, but was done as it improved results slightly. The platelet count was checked with a haemocytometer and was almost al,s,a vs unchanged from the previous day. The platelet suspension was then diluteJ by adding approximately 10-15 volumes of Ca ++ and Mg ++ free PBS, the platelets were spun down and resuspended to 10'~/ml (unless otherwise stated) in C.5~-c BSA in Ca ++ and Mg ++ free PBS. Serum diltltions were made in 0.5% BSA in Ca ++ and Mg +* free PBS, and 25 #tl of each dilution was transferred in duplicate to LP3 tubes (Luckham, .Ltd., Sussex). Twenty-five microliters o f the platelets suspension were added, incubated for 1 hr on ice, and shaken after 30 min. The platelets were then washed twice in 0.1% BSA in 0.9% NaCl by centrifugadon at 2000 × g for 15 rain, and to the pellet of the second wash 100 g,l o f t2~I labeled R A H was added, containing approximately 200,000-300,000 cpm. After resuspension this was incubated for 1 IV further on ice, and the platelets were washed twice as above. After the second wash, the platelets were resuspended in approximately 0.6 ml of 0.9% NaCl, transferred to fresh LP3 tubes, and the platelet bound radioactivity was measured in a Packard gamma counter. As performed above, the binding assay is termed a "'trace assay", as the ~"~! R A H is not present in saturating concentrations [ 10]. In one experiment (Figure 2) a "saturating" assay was used, and here the second incubation was with unlabeled R A H at 25 ~tg/ml, with r'nI R A H at approximately 400,000 cpm per assay.

Preparation of erythrocytes (RBC). Pure suspensions of RBC were preFared at room temperature from heparinized blood by washing twice in 0.9% NaCl, removing the bully coat and upper RBC layer each time. This was followed by centrifuging over triosil-ficoll of specific gravity 1.090 for 15 min at 800 x g. The pure erythrocyte pellet was then washed twice in 0.9~?F NaCI, and adjusted to 3 × 10Y/ml using a Coulter counter model DN.

Absorption analysis. For absorption analysis, a serum dilution was chosen which gave conditions o f antigen excess in the assay system, as discussed later. Equal

~.

Jean Kirkley and John W. Fabre volumes (80/zl) of the seruw at the chosen dilution were mixed with tripling dilutions of the RBC suspensions at initial concentrations of 3 x 10Vrul. This was incubated for 1 hr on ice; the RBC were pelleted by centrifugation; and the supernatant serunl was ,qssayed in the binding assay as described above.

Storage ofplatelets. As previously indicated, platelets were routinely steered at lO~/ml overnight at 4°C in Ca ++ and Mg ++ free PBS with 0.1% Na az~de. In one experiment, the platelets were stored in this way for 14 days. For longer term storage, the platelets were frozen in one of two ways. First, they were simply placed in -80°C freezers at a concentration of 5 x 10S/ml in the above buffer. Second, the platelets at 10Vml were mixed with an equal volume of 10% dimethyl sulphoxide (DMSO), 20% foe~l calf serum (FCS*) (Gibco-Biocult, Paisley, Scotland) in medium 199 (Gibco-Biocult, Paisley, Scotland), and placed in the -80°C freezer.

3I?LTS :ial Screening of Broadly Cross-Reactive Anti-HLA Sera on Platelets Three multiparous sera from blood group O individuals (GRO, MeG and STO) were titrated against three different blood group O platelets. All three sera reacted with the three platelet targets, and the results of one set of titrations are given in Figure 1. It can be seen that the sera gave good binding above background and that the amount of binding decreased as the sera were diluted, giving a typical titration curve. The strongest of the three sera, GRO, wa~ used for subsequent analyses. T.he amount of binding of this serum to different blood group O platelets varied, and in some cases the binding was substantially better than that shown in Figure 1. ice" versus "'Saturating" Assays

If the second incubation of the platelets is done with concentrations of RAH which will saturate all of the first antibody bound to the platelets~ then the amount Of radioactivity bound to the platelets is directly related to the amount of first antibody bound [10~. This can be used to yield useful information in a variety of ways. For exampl~~, if the first antibody is also bound un:ler saturating conditions (as indicated by ~ titration plateau) the amount of radioactivity bound is directly related to the amount of antigen in the target suspepsion. Concentrations of second antibody of 20 ltg/ml or above have been calculated to be saturating in most assay systems [10] and we have used concentrat.~ons of RAH of 25 /~g/ml. The main disadvantage of the saturating assay is that it is less sensitive than the trace assay, as shown in Figure 2, since the ~eSl labeled R A H must compete with a large excess of unl lbeled RAH for the firs~ antibody bound to the l:latelets. Because of the gre~ter sensitivity of the trace assay, we have used it throughout this study. mpts to Reduce Assay Background Because o f Surface Lunoglobulin on Plaeelets Platelets have immunoglobulin adsorbed on to their surfaces. The RAH will bind to this immunoglobulin as well as to any antibodies specifically bound to target antigens on the platelet membrane. Binding to adsorbed imtaunoglobulin obviously increases the assay background. In an attempt to minimize ~his, the

nding Assay for Platelet Antibodies

373

20

18"

m-----e McG

16

~ ' ~ k

D......~J STO NORHAL SERUM

.......

".

I x 10-3

12

t ASSAY

I0

8

6

~

'1

.*''"

........;.~ .~ ....,,.-:,.,.: ........................................... o..........iii : :!!; "0.

......... '~........... o " " "

2"

"

4

2

0

!

I/2

I/4

I,~8 1/16 SERUM DILUTION

1/32

1/64

1/128

GURE l Titration of ,roadly cross-reactive anti-HLA sera and a normal serum--all from ~od group O individuals,. m blood group O platelets.

plate.¢ts were ~r~incubated with unlabeled R A H to block the adsorbed immunoglobulin, as o~tlined in Materials and Met;~ods. The results are given in Figure 3 and show that blocking the adsorbed immunoglobulin did decrease assay backgroundl Though blocking is obviously not essential, it did consistently reduce back :rounds by about a half, and we therefi,re used it routinely. :her A t t e m p t s to Reduc.* Assay B a c k g r o u n d s It was found that ~ Lshing the platelets four times instead of twice after the first incubation (platelet: plus serum) did not diminish background at all (figure not shown). ,e Effect o f Concentratic , o f the T a r g e t Platelets: Determination o f ~say Conditions for O p t i n turn Sensitivity Given that the assay background is fairly constant over the range of antiserum dilutions tested, it is lear that the maximum specific binding above background will be seen when the target and,;ens on the platelets are covered with specific antibody, i.e. when conditions of specific antibody excess exist in the assay system. This wal tested by measuring the binding of the serum G R O ov,:r a wide series of dilutic~ns against target platelets at three concentrations, and the

574

Jean Kirkley and John W. Fabre

8

~x

O

NORMAL

SSRUM

TRACE ASSAY

7

• .....

SATURATING

ASSAY

z 6

~x x

CPM

X 10 -3

x

5

BOUND PER ASSAY

4

3 ""'-

2 "-~___<>._-.4X.~'~ 1

0

N

~,;3

,Ao xho

,)4o ,/8o

1A6o

1132o ~)64o-~/nso ,)2s6o

SERL~I DILUTION

FIGURE 2 Comparison of "trace" and "saturating" assays using serum GRO and a normal blood group O serum on O platelets.

results are given in Figure 4. Looking at the curves for platelets at the lowest concentration (0.3 x 109/ml) it can be seen that the amount of binding to the platelets did not dimin:sh over the range of serum dilutions from 1/10 to 1/30. This plateau suggests conditions o f specific antibody excess in this region, since the amount of antibody binding to the target platelets remained constant in spite of changes in the amount and concentration o f antibody in the assay system. It is interesting that it is in this region o f antibody excess that the highest ratio o f specific to background binding (approximately three) was obtained. Since there w ~ specific antibody excess with the s2rum at a 1/30 dilution and platelets at 0.3 x 10°/ml, it is likely that similar conditions would exist with the serum at a 1/10 dilution and the platelets at I × 109/ml, since ~he amount of both antibody and target antigen is increased by three. Again, the ratio of specific to background binding at that point w ~ about three. Obviously, any increase in the amount of target antigen would lead to conditions of antigen excess for serum d,:lutions greater than 1/10, and one can see that with platelets at 3 x 109/ml the maximum ratio of specific to background binding was only tWO.

It is clear that if one wishes to obtain maximum sensitivity, one should manipulate assay conditions to obtain a titration plateau.

Binding Assay for Platelet Antibodies

375

20 ~'\ 18

N~",

~x 16 14

x ~ x~ ~

"

i

GRO

--

NO RNAL SERUM BLOCKED

.....

NOT BLOCKED

12

CPM x 10-3 BOUND PER ASSAY

10

"a,.

4 2

o z/~o

~.¢3o

1/~o

1/~7o SERUM DILUTION

' 1/810

' 1/~4~

FIGURE 3 The effect of blocking immunoglobulin adsorbed to the platelet surface by incuSetion with immunoadsorbent puril']ed rabbit anti-human F(ab')~ prior to the assay. Blood group O platelets were used as targets.

Storage o f Platelets The effects o f different techniques for storing th,: platelets (see Materias arid Methods) are shown in Figure 5. It can be seen that optimum results were obtained with fresh platelets ,q.e., those stored overnight at 4°C) and chat M1 three storage metho0.s tested resulted in increased assay background without a proportionate increase in specific binding. This might have been the resuk o f ertry by 12sI R A H into damaged platelets during the second incubation o f tbe assay. The worst results were obtained with freezing in r.he absence o f DMSO; the best with simply keeping the platelets at 4°C. Whether or not periods ~f storage beyond 14 d;~ys at 4°C were acceptable was not tested. The Problem of ABO Antibodies In all the data presented above, sera from blood group O individuals w~'rg~ tested against blood group O platelets. In Figure 6 the results of testing ser~ .from nontransfused males o f blood group O, A, B, and AB aghast O, A~, B, and A,B platelets are shown. It can be seen that the serum from the group O person reacted quite strongly with A, and A,B platelets, but not with B or O

Jean Kirkley and John W. Fabre

;

20

~', x

x "

18

~ "",,.

16

t x 10 - 3

GRO

0

NORMAL SERUM

.............. 0 . 3

• ~.dk'

14

•

"x

~

~

~

.

X 10 9 PLATELETS/ML

1 X 10 9

..... 3 X 109

PLATELETS/ML

PLATELETS/ML

12

--

~--0-

. . . . . . . . .

-o

6

4 o...

2

,. . . . . . . . . . . . . . . .

........ o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~

..........

"""A ................................. •

............... o ............................... o

0

~Ao

i)30

I/~o

z/2"~o

I/e~o

SERUM D'r L U T I O N

GURE 4 The effect of varying the concentration of platelets in the target cell suspension on e sensitivity of the binding assay. Blood group O platelets were used as targets.

pl~ telets. This suggested that anti-A antibodies were being detected. The ABO antibodies in the A and the B sera did not give any binding. Figure 7 gives the results of testing mainly O and At sera from nontransfused males against B, O, At, and A2 platelets. It can be seen that three O and two Aj sera did not bind above background to B platelets, suggesting that anti-B antibodies are unlikely to raise problems in this assay. Of the three O sera tested against AI platelets, two gave bioding higher than background. Thus, three of four O sera tested gave binding to At platelets. N o n e of the sera gave detectable binding to A:~ platelets. tbsorption Analysis Suggesting That Interaction o f N o r m a l Sera with qatelets is D u e to A B O Antibodies For absorption analysis, one needs conditions o f antigen excess in the assay system, so that any removal of antibody is immediately detectable. The two O sera which reacted most strongly with A, platelets (Figures 6 and 7) were absorbed undiluted with tripling dilutions of O, A,, As, and B erythrocytes, and assayed on At platelets. After addition of the RBC suspension, the sera were at a one-half dilution, which represents conditions of antigen excess in both cases, ~ince the titration curve slopes downwards beyond the one-tudf dilution. The sera gave virtually identical results, and the results of one of the analyses are

~inding Assay for Platelet Antibodies 22

dk'-'..:,

20

~

18 16

377

"..N, .. ',, "'. ", ......... xx,dk\ X, • \ ',. \ "... ~ ".. xx •.. • x '"-...~ '.. \ "- ~

14 PM X I0 -3 OUND ER ASSAY

""

"'...~

~'\

'"..

\\,

"~".~.""..

12

o

I0

O.... "".., "'"" '~2;..

o-

•.....

.

.

.

.

GRO

o

NORMALSERUM

--

FRESH

.....

14 DAYS AT

4°C

............FROZEN WITH DMSO

~\

~'~,,..

.

•

.......... FROZEN WITHOUT DMSO

xx

x~. "A... ~\ .. " . ~

-'~,....

8

"..... "... ", ~: {~.~..

..........\

2 0 1/3.3

1/10

1/30

1/90

1/270

1/810

SERUM DILUTION

I G U R E 5 Comparison of various methods of storing platelets on their performance in the inding assay. Blood group O platelets were used as targets.

given in Figure 8. It can be seen that AI and A._, RBC could remove all the antibody activity, bat that B and O RBC could not remove any activity at all. It is interesting that the A2 RBC could absorb down to background levels, the only difference being that one needed approximately three times as many A, as At RBC to remove the antibody activity. Thus the O sera bound to A, but not B or O platelets, and the antibody activity could be removed by A but not B or O RBC. This pattern makes it virtually certain that it is ABO antibodies which are being detected. )ISCUSSION The binding assay as described in this paper is a simple, reproducible, and versatile assay for the measurement o f anti-platelet antibodies It shares with other binding assays the advantage :hat detection depends only on the interaction o f the ontibody with its target antigen, i.e., that there is no reliance on secondary phenomena, which are often dependent on the class of the antibody being assayed, and are not quantitatively related to the amount of antibody bound. In addition, it has two other important advantages. Fix~t, as shown in Figure 4, the most senstive assay conditions for particular serum and platelet combinations can be readily roland by manipulating the ratio o f serum to platelets to obtain a titration plateau. It is here, under conditions o f specific

B

Jean Kirkley and John W. Fabre

I2 I

0 pLATELETS

I

'°t |

B PLATELETS

I

' ' .....

'

""":"

o

--.o

~

~10 -3

AI PLATELETS

ND ASSAY

|B PLATELET5

............... o.." ...~ _ ~ . . ........o--

- -

"-"~ .

Y'"~..o---'-'::~".":'7""

....

p..s-~ " "~ :'~'~'2~.:

""~.,0

0 1,1

1/3

1/10

I:30

I/'90N I.~3 SERUM DILUTION

I)10

|/30

1/90

GURE 6 Binding of sera from nontransfuscd normal male volunteers of blood group O e), A (o . . . . o), B ([7 ............D), and AB (am.......... ai) to platelets of various ABO blo~gl ~ups, as indicated. antibody excess, that maximum specific/background binding ratios are obtained, as mentioned in the Results section. As backgrounds tend to rise slightly with the higher serum concentrations, the ideal is to obtain this p!ateau at serum dilutions o f l/5 or higher. The second important advantage is that the assay can readily be used for quantitative absorption analyses. As shown in Figure 8, the relative amounts o f an antigen on different tissues or cell suspensions can be measured, and tissue homogenates can be used in the absorptions without risk o f interfering with the assay [e.g., 10, 12]. O f more importance, is the fact that complex sera can be subdi-lided into components by absorption down to plateaus above background

~inding Assay for Platelet Antibodies

2 I

379 0 PI2tTELETS

B P~TELETS

PM X 10-3 :)UND JR ASSAY A, P L A T E L E T S

6

0 N

'GURE

7

1/2

1/4

I/8

1/16 1/32 1164 1/128 N 112 1/4 SERUM DILUT'rON

1/8

1/16 I132 1164 11128

Binding of sera from nontransfused normal male volunteers of blood ~roup

C~ _~, • •, e -) and A. ([3. . . . . [] . A . . . . A, o . . . . . ~) to platele~s of rious ABO blood groups as indicated. The autologous serum was included for three platelet :gets as follows: blood group B (¢¢ ~:), blood group O ('.). and blood group A: . . . . . . . . . . . . .

....

x).

by tissues carrying only some o f the specificities being recognized [ 11, 12]. This is either not possible or is not so clearly demonstrable in other assay systems. After preliminary analysis, judicious choice o f assay target and/or preliminary serum absorption usually allows the different components o f the serum to be analyzed separately [ 11]. The usefulness of this assay for the detection of anti-platelet antibodies in various clinical situations will depend on the concentration and type o f the antibodies to be assayed, the density and amount o f the target aatigen(s), and also on the relative sensitivity o f this assay when compared to those currently in use. When compared to other binding assays, it is likely to be at least as sensitive, especially if the assay conditions are manipulated to obtain maximum sensitivity (Figure 4). An important related point is the degree o f binding, in terms o f specific/background ratios, that should be considered positive, h i:i difficult to make absolute rules on this point, just as it is difficult to sa')" absolutely the degree o f cytotoxicity above background that should be taken as positive in cytotoxicity assays. Much depends on the system under study, and to some degree, the skill and care o f the person performing the assays. In general terms, a specific/background ratio of 1.5 is likely to be significant and a ratio o f

Jean Kiridey and John W. Fabre

22 2o

ab.- __ : . . . . . . . . . . . . . . . . . . . . . . . . . . . .

18

-

22.?"

.......• ..........

16

..

,.o....°.o..'"'J .at''"

, o

ASsAyDX 10-3 1412

, ..,..,....af'""'""''~

°.m"

8

6

~ Af RBC ......... A:, RBC

4

............... B

.....

2 o

i/3

~'9 1:2~ l/s1 OF RBC FOR ABBORPTION

RBC

0 RBC i/'243

DZLUTZON

URE O Quantitative absorption of serum from a nontransfused normal male volunteer of d group O (the serum shown in Figure 6). The s e r u m was absorbed with RBC suspensions as m as starting concentrations of 3 x 10~/ml and assayed on platelets of blood group AI.

2 certainly significant. If, however, one performs titrations, greater confidence can be placed in lower ratios, (e.g., the 1/I0 dilution o f the blood group O serum on AIB platelets in Figure 6). Betause undiluted serum occasionally gave spuriously low counts, as discussed below, a negative result with undiluted serum should be treated with caution. As regards A B O antibodies, many more sera would need to he studied to see how frequently these antibodies interact with platelets. However, it is clear from the results presented that anti-A antibodies on At plateiets are definitely a problem, and chat to be safe one should match the serum and target platelets for compatibility of the A B O system. Another technical point that should be noted is that some undiluted sera and some at a one in two dilution seemed to cause platelet lysis, with consequently spuriously low counts bound at these serum dilutions. Whether or not this was caused by inadequate decomplementation of these sera is unknown. In any case, most sera behaved normally at these dilutions, as can be seen by scanning the figures o f the results presented. The assay can be manipulated in a number o f ways. For example, the second antibody used in these studies was anti-human F(ab')z, which Is essentially directed at the light chains o f immunoglobulin. This was designed so as to detect auto- or alloantibodies irrespective o f the class o r subclass o f the antibody~ However, class-specific second antibodies would aUow t h e antibody response to be subdivided quantitatively into its various components. Xenosera

ling Assay for Platelet Antibodies

3481

can also be analyzed, simply by using anti-immunoglobulin sera directed at the immunized species~ and this has in fact been the main application o f this assay in other systems. Other blood cell types can o f course be used as targets; the only difficulty is a higher background when studying cells or autoantibodies with B lymphocyte targets [ 13] because o f the surface immunoglobulin o f B cells.

ACKNOWLEDGMENT

This work was supported by grants f~om the Medical Research Council of the United Kingdom and the Wellcome Trust.

FERENCES I. Colombani J, D'Amoro J, Gabb B, Smith G, Svejgaard A: International agreement on a microtechnique of platelet corrkpiement fixation. Transplant Proc 3:12 l, 197 I. 2. Lizak GE, Grumet FC: A new micromethod for the in vitro detection of antiphtelet antibodies: C-FDA thrombocytotoxicity. Human |mmunol 1:87, 1980. 3. Henrich D, Gutschank S, Mueiler-Eckhardt C: HLA antibody induced t4C-serotonin release from platelets. A methodolc!gical analysis. Vox Sam 33:65, 1977. 4. van Leeuwen A, Eernisse JG, van ~ood JJ: A new leucocyte group with two alleles: leucocyte group five. Vox Sang 9:431, 1964. 5. Coombs RRA, Bedford D: The A ;;~nd B antigens on human platelets demonstrated by means of mixed erythrocyte-platelet agglutination. Vox Sam 5:11 l, 1955. 6. Dixon R, Rosse W, Ebbert L: Quantitative determination of antibody in idiopathk thrombocytopenic purpura. Correlation of serum and platelet-bound antibody with clinical response. New Eng Med J 292:230, 1975. 7. Tate DY, Sorensen RL, Gerrard JM, White JG, Kirvit W: An immuneenzyme histochemicai technique for the detection of platelet antibodies from the serum of patients with idiopathic (autoimmune) thrombocytopenia purpura (ITP). Br Haematol 37:265, 1977. 8. blcMillan R, Smith RS, Longmi~e RL, Yelenousky R, Reid RT, Craddock CG: lmmuno#obulins associated with human platelets. Blood 37:316, 197 I. 9. Soulier JP, Patereau C, Drouet J: Platelet indirect radioactive Coombs test. Its utilisation for Pla, grouping. Vox Sang 29:253, 1975. 10. Morris RJ, Williams AF: Antigens on mouse and rat lymphocytes recognised by rabbit antiserum to rat brain: the quantitative analysis of a xenogeneic antiserum. Eur lmmunol 5:274, 1975. I I. Fabre JW, Williams AF: Quantitative serological analysis of rabbit anti rat lymphocyte serum and preliminary biochemical characterisation of the major antigen recognized. Transplant 23:349, 1977. 12. Dalchau R, Fabre JW: Identification and unusual tissue distribution of the canine and human homologues of Thy-I (O). J Exp Med 149:576, 1979. 13. Hart DNJ, Fabre JW: Quantitative studies on the tissue distribution of la and SD antigens in the DA and Lewis rat strains. Transplant 27: 110, 1979. 14. H.u't DNJ, Fa!;~'eJW: Kidney specific aUo~ and auto-antibodies in the alloandbody response to r~ t kidney: the use of kidney homogenate as a target for serologk~ analysis. Ciin Exp 40:11 l, 1980. 15. Fairbanks G, ~tech TL, Wallach DJH: Eiectruphoretic analysis of major polypepddes of the human q~rythrocyte membrane. Biochem 10:2606, 197 I.