American Journal of Hematology 39:25-31 (1992)

Deficiency of P62, a Putative Collagen Receptor, in Platelets From a Patient With Defective Collagen-Induced Platelet Aggregation Ryukichi Ryo, Akinori Yoshida, Wataru Sugano, Mutsumi Yasunaga, Katsuji Nakayama, Katsuyasu Saigo, Masashi Adachi, Nobuo Yamaguchi, and Minoru Okuma Blood Transfusion Service and Department of Laboratory Medicine, Kobe University School of Medicine, Kobe (R.R., A.Y., W.S., M.Y., K.N., K.S., M.A., N.Y.) and Department of Internal Medicine, Kyoto University School of Medicine, Kyoto (M.O.), Japan.

Recently, we described a platelet antibody against a putative collagen receptor (P62), which was found in a patient with idiopathic thrombocytopenic purpura (ITP) (Blood 69:1712). We now report a deficiency of the P62 receptor in a young man whose platelets showed defective collagen-inducedplatelet aggregation. He had a mild bleedingtendency and slight thrombocytopenia. The results of coagulation and fibrinolysis studies were normal. The patient's platelets were partially unresponsive to collagen, although aggregation in response to ADP, thrombin, ristocetin, and calcium ionophore (A23187) was almost normal. Adhesion of his platelets to bovine collagen was markedly reduced. Addition of collagen caused no synthesis of thromboxane (TX)B, in platelet rich plasma (PRP) from this patient. Furthermore,collagen produced no rise of cytosolic free calcium ([Ca2+]i) in fura24oaded platelets. In contrast, thrombin caused TXB, formation and an increase of [Ca2+]iin his platelets. These results suggest defective interaction between the platelets and collagen. The IgG from the ITP-patient induced irreversible aggregation in normal PRP, but caused no aggregation of the young man's platelets. lmmunoblot studies showed that normal platelets had antigens with a molecular weight of 62 KDa under reducing conditions and of 57 KDa under nonreducing conditions. In contrast, the young man's platelets had no P62 band, although GPla/lla and thrombospondin were normally present. These results indicate that impaired collagen-induced aggregation in the patient's platelets was due to a deficiency of P62 and confirm that P62 may play a crucial role as a collagen receptor in platelet activation. Key words: thrombocytopathy, thrombospondin, glycoprotein la/lla, Western blotting

INTRODUCTION The interaction of platelets with collagen is thought to be the initial event in hemostasis after damage to vascular endothelium. Collagen-platelet interactions result in the release reaction and subsequent aggregation. Recent studies have shown that these processes are mediated by a receptor for collagen on platelet membranes [ 1 4 ] . A number of platelet proteins have been shown to bind to collagen and have been suggested as potential platelet collagen receptors [5-81. Recently we reported a patient with idiopathic thrombocytopenic purpura (ITP) and a platelet antibody which inhibited collagen-induced aggregation in normal platelet rich plasma (PRP). The antibody recognized a putative collagen receptor with a molecular weight of 62kDa under reducing conditions (P62) [9]. Now we report a 0 1992 Wiley-Liss, Inc.

patient with platelets that are unresponsive to collagen and lack P62. The present findings also suggest that P62, which is recognized by the ITP patient's IgG, plays an important role as a collagen receptor.

MATERIALS AND METHODS With the informed consent of the patient and control subjects, PRP and platelet poor plasma (PPP) were prepared as described previously [9]. The platelets were washed twice in 150 mmol/l NaC1, 5 mmol/l glucose, 10

Received for publication December 7, 1990; accepted July 16, 1991. Address reprint requests to Dr. Ryukichi Ryo, Blood Transfusion Service, Kobe University Hospital, Chuo-ku, Kobe 650, Japan.

26

Ryo et at.

mmol/l Tris-HC1, pH 7.4, containing 15% ACD and then suspended in HEPES-buffered saline (10 mmol/l HEPES, 145 mmol/l NaCl, 5 mmol/l KCl, 0.5 mmol/l Na,HPO,, 6 mmol/l glucose pH 7.4). Platelet Aggregation and ATP Release

Phillips and Agin [ 141. After electrophoresis under nonreducing conditions, the lane was cut out and incubated in the reducing sample buffer containing 20% dithiothreitol (Bio Rad, Richmond, CA) for 15 minutes at 60°C. The strip of acrylamide were placed on top of the stacking gels of a second slab (6.5% acrylamide).

Platelet aggregation was determined by a turbidometric method using PRP and washed platelet suspensions (thrombin effect only) at a density of 15X 104/pl and an aggregometer. ADP (Sigma, St. Louis, MO), A23187 (Calbiochem, CA), thrombin (Mochida Pharmacuetical Comp., Tokyo), acid soluble collagen from bovine tendon (Hormon-Chemie, Munich, Germany), human type I, 111, collagen (Fuji Chem. Ind., Tokyo), bovine achilles tendon collagen type I (Sigma) and ristocetin (H. Lundbeck & Comp., Copenhagen) were used as the platelet agonists. Platelet release reaction was estimated by measurement of the release of ATP using a Lumiaggregometer (Chrono-Log, Havertown) by the fireflyluminescence method.

Platelet antigens on the nitrocellulose membranes were incubated with antiserum diluted to 1X 100, and stained with avidin-biotin complex (ABC) reagents (Vector, Burlingame, CA). Peroxidase activity was detected by adding 10 ml of 0.3% 4-chloro-l-naphthol in methanol, 25 pJ of 30% H,02 and 50 ml of PBS. To analyze platelet glycoproteins, biotinylated lectins including ricinus communis agglutinin, concanavalin A, and wheat germ agglutinin (Vector) were used [15]. Each lectin was diluted to 0.008 mg/ml in PBS. The nitrocellulose membranes were washed three times with 0.9% NaCl and incubated with ABC reagent.

Measurement of Thromboxane (TX)B,

Preparation of the Patient's Serum and IgG

The platelet counts were adjusted to 1X105/pl in plasma or in HEPES-buffered saline, and the TXB, content of the supernatant was assayed directly using a TXB, RIA kit (Amersham, Tokyo) 5 minutes after agonist-induced aggregation under continuous stirring on the aggregometer as described previously [lo].

With the informed consent of the patient, the serum was obtained as described previously [9]. The IgG was purified by affinity chromatography with protein-G Sepharose 4 Fast Flow (Pharmacia, NJ). The Fab fragment of the IgG against P62 inhibited collagen-induced aggregation. The other character of P62 IgG is described in previous paper [9].

Platelet Cytosolic Free Calcium Ion ([Ca2+]i)

The [Ca2']i was measured using a fluorescent Ca2+ indicator, fura 2-AM (Dojindo Laboratories, Kumamoto, Japan) as described previously [ l l ] . In brief, fura-2 loaded platelets were resuspended in HEPES-Tyrode solution, and transferred to the cuvette. The external Ca2' was adjusted by adding 1 mM CaCl, or 3mM EGTA as required. The fura-2 fluorescence of the platelets was recorded using a dual excitation wavelength spectrofluoremeter (CAFIOO, NihonBunko Co., Tokyo, Japan) with excitation at 340 and 380 nm and emission at 500 nm. Values of [Ca2']i were calculated from fluorescence signals, as described by Tsien et al. [12]. The platelet aggregation was also monitered with CAF 100. SDS-PAGE and Two-Dimensional Nonreducing/ReducingSDS-PAGE of Platelet Proteins

Platelets were dissolved in SDS sample buffer of the method of Laemmli [13]. Reduced and nonreduced samples were subjected to SDS-PAGE (7.5% acrylamide) and then were blotted onto nitrocellulose membranes (0.45 pm, Bio Rad, Richmond, CA) using an electrophoretic blotting apparatus (Sartorius). Two-dimensional electrophoresis was performed as described by

lmmunoblotting

CASE REPORT

A 26-year-old man of normal stature was referred to us for investigations of spontaneous superficial bleeding in Sept. of 1986. Physical examinations revealed numerous petechiae and multiple ecchymoses over the four extremities and the trunk. He had no previous history of a bleeding tendency. Thereafter, he developed occasional episodes of spontaneous epistaxisis and ecchymoses over the extremities without any disposing factors every three or four months. He denied ingestions of any drugs before he noted hemorrhagic diathesis. His parents and two brothers are healthy. There was no parental consanguinity. His mother has two babies without any severe bleeding complications. We have no chance to examine his mother's platelet function or coagulation parameters. RESULTS

The patient's platelet count varied from approximately 10X lo4 to 13X 104/pl. The appearance of the platelets was normal after May-Giemsa staining. The mean platelet volume was 9.9 f P (normal mean 2 2 SD; 10.3 1.8 f P ) . The slight thrombocytopenia has been always ob-

*

Thrombocytopathy With Defective Collagen Receptor

27

-100

-x

C

P 0 -)

-50

6C

li

+m c

x

3 LO

/

I

Collagen 5 &rnl

roo-

1 min

H

Fig. 1. Aggregation patterns and [Ca2’]i mobilization of patient’s and control PRP and/or washed platelets in response to thrombin and collagen [Ca2’]i was measured as described in Methods. The effect of thrombin was observed in washed platelets, and that of collagen in PRP.

served from September 1986 to December 1990. His megakaryocytes had a normal morphology and normal counts. Bleeding time was 13 minutes with Duke’s method (normal values: within 5 minutes) and clot retraction was normal. The bleeding time was tested at least three times for the last 4 years. Platelet adhesion was decreased to 13% with Mant’s method [I61 using bovine tendon achilles collagen type I (normal mean *2 SD; 61.4 t 7.2%). The platelet adhesion was observed with PRP anticoagulated with 0.1 vol. 0.2% EDTA in 0.533% saline as described by Mant [16]. Coagulation parameters, including a-PTT, prothrombin time, a,-plasmin inhibitor, and FVIII-related antigen were within normal limits. Platelet Aggregation

Primary and secondary aggregation of the patient’s PRP induced by 10 p M ADP was almost normal, and primary aggregation was reversible with 3 pM ADP. Ristocetin (1.5 mg/ml) induced normal aggregation of PRP. The response of the patient’s washed platelets to 1 U/ml of thrombin was entirely normal (Fig. 1). In contrast, acid soluble collagen at concentrations ranging from 3 pg/ml to 20 pg/ml failed to induce any aggregation of PRP. Aggregation curves for the patient’s platelets and control platelets in response to collagen are shown in Figures 1 and 2. The increased collageninduced turbidity in the patient’s platelets was observed

Control

Patient

Fig. 2. Aggregation patterns and ATP release of patient’s and control PRP in response to collagen. Platelet aggregation and ATP release were monitored as described in Methods.

with CAFlOO (Fig. 1). This effect may imply that collagen could bind to one or more platelet receptors. On the contrary, no change of the turbidity in the patient’s platelets was noted with Lumiaggregometer (Fig. 2). This discrepancy may depend on the aggregometers used. The defect of collagen-induced platelet aggregation was similarly observed with human collagen. The IgG against P62 prepared from the previously reported ITP patient (P62 IgG) [9] induced irreversible aggregation of normal PRP, whereas it failed to induce aggregation of the patient’s PRP (Fig. 3). ATP Release

ATP release during platelet aggregation occurred in response to A23187 (10 pM) in normal PRP and in the patient’s PRP (Fig. 4). In contrast, acid soluble collagen (10 pg/ml) and P62 IgG (300 pg/ml) induced ATP release in normal PRP, but failed to induce ATP release in the patient’s PRP (Figs. 2, 3). Platelet [Ca2’]i Elevation

Thrombin (1 U/ml) caused [Ca2’]i elevation in normal washed platelets and the patient’s platelets, whereas

28

Ryo et at. P62IsG(300ps/ml)

A23187 (10pM)

P62IgG(300ps/ml)

A23187(10pM) t

I -

1002

Control

Control

Patient

Fig. 3. Aggregation patterns and ATP release of patient's and control PRP in response to P62 IgG. The studies were performed and expressed as described in Methods.

Patient

Fig. 4. Aggregation patterns and ATP release of patient's and control PRP in response to A23187.

TABLE I. Thrornboxane 9 2 Synthesis

TXB2 (ng/ LO8 platelets) Platelets PRP

Washed platelets

Agonist Collagen Saline Thrombin Saline

Patient

2 pgiml 20 pgiml

(-)

(-)

0.5 U/ml

1

7.6 1 54 4.0

Control 1.6 f 2.98 46.7 f 9.0 (-)

44 f 8.6 9.2 f 1.7

Thromboxane B2 synthesis in the supernatantsof PRP with collagen and washed platelets with thrombin was measured as described in Methods. All valuesper 10' platelets are shown. The values in normal controls are mean k SD of at least 5 determinations.

collagen had no effect on [Ca2+]i (Fig. 1). It appeared that the patient's platelets were hyperresponsive to thrombin as judged by Ca2' release. Platelet TXB, Formation

Thrombin (0.5 U/ml) promoted TXB, formation in the patient's washed platelets, whereas collagen (2 pg/ml) had no effect in the patient's PRP, although collagen at a higher concentration of 20 pg/ml caused a slight stimulation of TXB, formation (Table I).

Analysis of Platelet Proteins by lmmunoblotting

An immunoblotting analysis of P62 expression by the platelets of a normal control and patient is shown in Figure 5 . Under nonreducing conditions, one labeled band with a molecular weight of 57 KDa was specific for the serum containing the antibody against P62 in the normal control platelets. Numerous other nonspecific bands were visible even with normal serum. It should be noted that the serum was useful to identify P62 in the platelets. On the other hand, the 57 KDa band was

29

Thrombocytopathy With Defective Collagen Receptor

Reduced

NONREDUCED

Non-Reduced

____.__f

M. W.

(k Da)

200 11697 66 -

43 -

NONREDUCED

1

2

3

4 kOa

Fig. 5. lmmunoblot of isolated platelets from patient and a normal individual stained with the serum containing antiP62. The platelets (75 pg) were electrophoresed on 7.5% SDS gels under reduced or nonreduced conditions, transferred to nitrocellulose, and stained with the serum containing anti-P62. Lane 1 and lane 3, immunoblots of platelet proteins from a normal individual under reducing and nonreducing conditions, respectively. Lane 2 and lane 4, immunoblots from the patient under the same conditions. The arrow indicates the specific antigen recognized by the serum containing anti-P62.

lacking in the patient’s platelets. Under reducing conditions, a band with a molecular weight of 62 KDa was specifically labeled by the serum containing the antibody against P62 in normal platelets, but no binding of P62 antibody to the patient’s platelets was found under reducing conditions. The lack of P62 identified by the serum containing the antibody against P62 was not observed in the platelets from at least 27 normal controls. Deficiency of P62 expression by the patient’s platelets was confirmed by two-dimensional gel electrophoresis followed by immunoblotting (Fig. 6). The defect of P62 in the patient’s platelets has been repeatedly confirmed at least four times since September 1986. Analysis of Platelet Glycoproteins Using Two-Dimensional Gel Electrophoresis and Avidin-Biotin-ConjugatedLectins

The patient’s platelet glycoproteins were selectively stained with lectins including ricinus communis agglutinin, concanavalin A, and wheat germ agglutinin (Fig. 7). The platelet glycoproteins were judged according to Kehrel’s criteria [15]. GPIa, GPPIb, GPIc, GPIIa,

rn

Fig. 6. lmmunoblots of isolated platelets (75 pg) from the patient and a normal individual using two-dimensional gel electrophoresis followed by an incubation with the serum containing anti-P62. First dimension, SDS-PAGE, 7.5% acrylamide under nonreducing condition; second dimension, SDS-PAGE, 7.5% acrylamide under reducing conditions followed by immunoblot. A: Normal individual; B: patient. The antigen recognized by the serum containing anti-P62 in the patient’s platelets was absent as indicated by the arrow.

GPIIb, GPIIIa, GPIV, and thrombospondin in the patient’s platelets were stained as clearly as in normal control platelet. The classical analysis of platelet proteins by one-dimensional SDS-PAGE (7.5% acrylamide) followed by periodic acid-Schiff staining also confirmed the presence of GPIa in the patient’s platelets. DISCUSSION

The platelet aggregation and release reaction studies in our patient showed that his platelets were unresponsive to collagen. This defect of collagen activation was evidently not due to a failure of thromboxane synthesis and calcium mobilization, since both were normally observed in

Ryo et al.

30

may be involved in collagen-induced platelet activation. In fact, several candidates for collagen receptor have been proposed [l-3,5,7,8]. More recently, Moroi et al. reported the patient, whose 200 platelet lacks P62 as well as collagen-induced platelet aggregation and adhesion [ 191. The defect in the patient’s platelets reported by Moroi seem to be almost the same 116 as that in our patient’s platelets. However, we should note some differences between two patients. First, Moroi 97 could not immunologically detect P62 in the platelets under reducing conditions, whereas we could. The Moroi’s patient had normal platelet counts, whereas our patient had slight thrombocytopenia. The reason why the patient showed thrombocytopenia remains unclear. We assume that one of characteristic features in the patient with platelets deficient in P62 is slight thrombocytopenia. Fig. 7. lmmunoblots of isolated platelets (75 pg) from the patient using two-dimensionalgel electrophoresisfollowed Furthermore, Moroi demonstrated no evidence that the by an incubation with avidin-biotin conjugated lectins. la, defect of P62 is related to impaired collagen-induced GPla; Ib, GPlb; Ic, GPlc, Ila, Glla; Ilb, GPllb, Illa, GPllla; IV, calcium mobilization and thromboxane formation in the GPIV; TSP, thrombospondin. platelets. Anyway, the present studies indicate that impaired collagen-induced aggregation in the patient’s platelet was due to a deficiency of P62 in platelet and that response to thrombin and the calcium ionophore A23 187. P62 may play a crucial role as a collagen receptor in These results implied that the interaction of collagen with platelet activation. the patient’s platelet via the membrane receptor might be impaired. To investigate the platelet membrane, we observed the effect of P62 IgG [9] which has been REFERENCES proposed to recognize a putative collagen receptor, on I . Chiang TM, Kang AH: Isolation and purification of collagen orl(l) receptor from human platelet membrane. J Biol Chem 257:7581platelet aggregation. P62 IgG failed to produce platelet 7586, 1982. aggregation in the patient’s platelets in contrast to its 2. Kotite NJ, Cunningham LW: Specific adsorption of a platelet memeffect on normal platelets, suggesting that the antigen brane glycoprotein by human insoluble collagen. J Biol Chem recognized by P62 IgG may have been lacking in our 26 1 :8342-8347, 1986. patient. To confirm the absence of P62, the patient’s 3 . Santoro SA: Identification of a 160,000 dalton platelet membrane protein that mediates the initial divalent cation-dependent adhesion of platelet proteins were analyzed by immunoblotting . The platelets to collagen. Cell 46:913-920, 1986. P62 antigen was not detected by one-dimensional gel 4. Coller BS, Beer JH, Scudder LE, Steinberg MH: Collagen-platelet electrophoresis and two-dimensional gel electrophoresis interactions: Evidence for a direct interaction of collagen with platelet followed by immunoblotting. Thus, the impaired interGPIa/IIa and indirect interaction with platelet GPIIb/IIIa mediated by action of the patient’s platelets with collagen resulted adhesive proteins. Blood 74:182-192, 1989. 5 . Bensusan HB, Koh TL, Henry KG, Murray BA, Culp LA: Evidence from the absence of P62. A patient with defective that fibro-nectin is the collagen receptor on platelet membranes. Proc collagen-induced platelet aggregation was for the first Natl Acad Sci USA 75:5864-5868, 1978. time reported by Nieuwenhuis et al. [6,17]. 6. Nieuwenhuis HK, Akkerman JWN, Houdijk WPM, Sixma JJ: Human The authors concluded that the defect of collagen blood platelets showing no response to collagen fails to express surface aggregation was attributable to the absence of GPIa. glycoprotein la. Nature 318:47@472, 1985. 7. Saito Y , lmada T, Takagi J, Kikuchi T, Inada Y : Platelet factor XIIl. Kunicki [ 181 confirmed that GPIdGPIIa complex plays a The collagen receptor? J Biol Chem 261:1355-1358, 1986. crucial role in collagen-induced aggregation using a 8. Kehrel B, Balleisen L, Kokott R, Mesters R , Stenzinger W , Clemetmonoclonal antibody directed against GPIa. Furtherson KJ, van de Loo J : Deficiency of intact thrombospondin and more, Kehrel [8] described a case of aquired thrombomembrane glycoprotein la in platelets with defective collagen-induced cytopathy with defective collagen-induced aggregation, aggregation and spontaneous loss of disorders. Blood 7 1: 1074-1078, 1988. in which GPIa and thrombospondin were deficient. 9. Sugiyama T, Okuma M , Ushikubi F, Sensaki S , Kanaji K, Uchino H: However, our patient’s platelets were found to express A novel platelet aggregating factor found in a patient with defective both GPIa and thrombospondin, and to lack P62. Thus, collagen-induced platelet aggregation and autoimmune thrombocythe defective collagen-induced platelet activation in our topenia. Blood 69:1712-1720, 1987. present patient is not attributable to a deficiency of GPIa 10. Ryo R, Yoshida A, Adachi M, Sugano W, Yasunaga M, Yoneda N, Yamaguchi N: Cytosolic calcium mobilization and thromboxane or thrombospondin. Therefore, a number of receptors

MI kb

-

Nonreduced

Thrombocytopathy With Defective Collagen Receptor synthesis in a human megakaryocytic leukemia cell. Exp Hematol 18:271-275, 1990. II. Adachi M, Ryo R, Yoshida A , Teshigawara K, Yamaguchi N, Hoshijima M, Takai Y, Sato T: Elevation of intracellular calcium ion by prostaglandin El and its inhibition by protein kinase C in a human megakaryocyte leukemia cell line. Cancer Res 49:3805-3808, 1989. 12. Tsien RY, Rink TJ, Poenie M: Measurement of cytosolic free CaZ+in individual small cells using fluorescence microscopy with dual excitation wavelengths. Cell Calcium 6:145-157, 1985. 13. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685, 1970. 14. Philips DR, Agin PP: Platelet plasma membrane glycoproteins. Evidence for the presence of nonequivalent disulfide bonds using nonreduced-reduced two-dimensional gel electrophoresis. J Biol Chem 252:2121-2126, 1977. 15. Kehrel B, Kokott R, Balleisen L, Stenzinger W, Clemetson KJ, van de Loo J: Selective staining of platelet glycoproteins using two-

16.

17.

18.

19.

31

dimensional O’Farrell gel electrophoresis and avidin-biotin-conjugated lectins. Thromb Haemost 58:96&963, 1987. Mant MJ: Platelet adherence to collagen: A simple, reproducible, quantitative method for its measurement. Thromb Res 1 1:729-737, 1977. Nieuwenhuis HK, Sakariassen KS, Houdijk WPM, Nievelstein, PFEM, Sixma JJ: Deficiency of platelet membrane glycoprotein Ia associated with a decreased platelet adhesion to subendothelium: A defect in platelet spreading. Blood 68:692-695, 1986. Kunicki TJ, Nugent DJ, Staats SJ, Orchekowski RP, Wayner EA, Carter WG: The human fibroblast class extracellular matrix receptor mediates platelet adhesion to collagen and is identical to the platelet glycoprotein la-IIa complex. J Biol Chem 263:45 1-519, 1988. Moroi M, Jung SM, Okuma M, Shinmyozu K: A patient with platelets deficient in glycoprotein VI that lack both collagen-induced aggregation and adhesion. J Clin Invest 84:1440-1445. 1989.