Histochemistry (1992) 97 : 381-388

Histochemistry © Springer-Verlag 1992

Immunocytochemical evidence for the translocation of x-granule membrane glycoprotein IIb/IIIa (integrin O~iib~3 ) of human platelets to the surface membrane during the release reaction H. Suzuki 1, S. Nakamura 2, Y. Itoh 2, T. Tanaka l' *, H. Yamazaki 1, and K. Tanoue 1 1 Department of Cardiovascular Research, The Tokyo Metropolitan Institute of Medical Science, 3-18-22 Honkomagome, Bunkyo-ku, Tokyo 113, Japan z Hitachi ScientificInstruments Center, Nissei Sangyo Co., Ltd., 26-27 Sakamachi, Shinjuku-ku, Tokyo 160, Japan Accepted March 24, 1992 Summary. The localization of glycoprotein (GP) IIb/IIIa (integrin cqiu/~3) in both resting and thrombin-activated platelets was studied immunocytochemically. By the preembedding method where only the GP IIb/IIIa molecules on the surface of platelets were immunostained, the distribution of protein A-colloidal gold label was randomly distributed along the surface membrane of resting platelets at a density of 18.0_+ 2.7 gold particles/ ~tm of membrane. At 15 s after stimulation by 0.1 U/ml of thrombin in an unstirred platelet suspension, the spheroid-shaped platelets with pseudopodia still had normal numbers of e-granules, and the density of gold particles was 19.7 _+3.6 particles/gm. At 5rain, the egranules were no longer present because of the release reaction, and the density of gold particles significantly increased (27.0 _+3.7 particles/lam; p < 0.01). In immunostained ultra-thin frozen sections, the gold particles were detected not only on the surface membrane, including the open canalicular system (OCS), but also on the egranule membranes of resting platelets. At 30 s after thrombin stimulation the e-granules fused with the OCS, resulting in the formation of a swollen OCS, which still had gold particles on its membrane. At 5 rain, the gold particles were detected on the membrane of the swollen OCS located near the surface membrane, while very few gold particles were present on the membrane of the OCS in the central part of the platelets. These results demonstrate that e-granule membrane GPIIb/IIIa translocates to the surface membrane through the membrane of the OCS. Also the translocation of e-granule membrane GPIIb/IIIa gives rise to an actual increase in GPIIb/IIIa on the surface membrane during the release reaction induced by thrombin.

* Present address." Bio-molecular Analysis and Drug formulation Department, BiomedicalResearch Laboratories, Kureha Chemical Industry Co., Ltd., 3-26-2 Hyakunin-cho, Shinjuku-ku, Tokyo 169, Japan Correspondence to. H. Suzuki

Introduction Platelets, when activated by an appropriate stimulus, change their shapes, release the contents of their storage granules, and aggregate with each other (White 1974). Platelet aggregation requires the binding of fibrinogen to its membrane receptor, the glycoprotein (GP) IIb/IIIa complex (Mustard et al. 1979; Bennett and Vilaire 1979). This Ca2+-dependent heterodimer complex of GPIIb and GPIIIa (Jennings and Phillips 1982) is a member of the integrin family, which is also called an integrin eiibfl3 (Hynes 1987). Glycoproteins IIb and IIIa are the most abundant glycoproteins on the surface of platelets, and there are approximately 50000 GPIIb/IIIa complexes on the surface of normal platelets (Lombardo et al. 1985; Coller 1985), representing about 1-2% of the total platelet protein (Jennings and Phillips 1982). Immunocytochemical studies by several investigators have shown that GPIIb/IIIa complexes exist on the plasma membrane including the open canalicular system (OCS) (Polley et al. 1981; Wencel-Drake etal. 1986; Isenberg et al. 1987; Suzuki et al. 1990, 1991). According to evidence obtained by subcellular fractionation (Gogstad etal. 1981; Van der Meulen eta!. 1983) and immunoelectron microscopical studies (Wencel-Drake etal. 1986; Suzuki etal. 1990; Cramer etal. 1990), GPIIb/IIIa complexes are also found on platelet e-granule membranes. Binding assays using monoclonal antibodies against GPIIb/IIIa have demonstrated increases in the number of GPIIb/IIIa complexes on the surface membrane after platelet activation by thrombin (Lombardo et al. 1985; Coller 1985; Niiya et al. 1987). These observations suggest that the binding of these monoclonal antibodies depends on a conformational and/or microenvironmental change in the GPIIb/IIIa complex (Coller 1985). It has also been postulated that the increases of the complex of the surface arise from the translocation of intracellular GPIIb/IIIa localized on the membrane of the OCS and e-granules (Lombardo etal. 1985; WencelDrake et al. 1986). Howewer, the mechanism of the in-

382 crease o f G P I I b / I I I a o n the surface m e m b r a n e after activ a t i o n i n d u c e d b y t h r o m b i n a n d the c o n t r i b u t i o n o f eg r a n u l e m e m b r a n e G P I I b / I I I a to the increase have n o t yet been clarified. I n the p r e s e n t study, b y u s i n g a p r e - e m b e d d i n g imm u n o s t a i n i n g m e t h o d where only G P I I b / I I I a complexes o n the surface m e m b r a n e c a n be i m m u n o s t a i n e d , we have c o n f i r m e d m o r p h o m e t r i c a l l y that G P I I b / I I I a complexes o n the p l a s m a m e m b r a n e increased b y 50% after the release reaction. T h e r e d i s t r i b u t i o n o f e - g r a n u l e m e m b r a n e G P I I b / I I I a o c c u r r e d d u r i n g the release reaction, which was observed i m m u n o c y t o c h e m i c a l l y in ult r a - t h i n frozen sections. This d e m o n s t r a t e d t h a t the eg r a n u l e m e m b r a n e G P I I b / I I I a is t r a n s l o c a t e d to the surface m e m b r a n e via the m e m b r a n e of the swollen OCS, which h a d been f o r m e d b y the f u s i o n o f e - g r a n u l e s a n d OCS. M o r e o v e r , the increase o f G P I I b / I I I a o n the surface m e m b r a n e is caused by the t r a n s l o c a t i o n of e - g r a n ule m e m b r a n e G P I I b / I I I a d u r i n g the release reaction.

Materials and methods

Preparation and platelets Venous blood was drawn from healthy young volunteers, who had not ingested any drugs for at least 10 days, into a plastic syringe containing 1/10 (v/v) volume of 3.8% sodium citrate. The blood was centrifuged at 160 g for 15 rain at room temperature to obtain platelet-rich plasma (PRP). The pH of the PRP was adjusted to 6.5 with 1 M citric acid, the PRP was sedimented by centrifugation at 900 g for 15 rain, and then washed twice as previously described (Suzuki et al. 1987). The washed platelets were finally suspended to a concentration of 5 x 105 platelets/ixl in Tyrode's solution (pH 7.4) consisting of 137 mM NaC1, 2.7 mM KC1, 0.4 mM NaHz PO4, 12raM NaHCO3, 1 mM MgC12, 2raM CaC12, 22 gM Na3C6HsO7, and 0.35% bovine serum albumin (BSA; Sigma Chemical Co., St. Louis, Mo.~ USA) and 0.1% glucose. The platelets were kept for 30 rain at 37° C prior to use. One milliliter aliquots of the platelet suspensions were incubated with bovine thrombin (Sigma) at a final concentration of 0.1 U/ml for 15, 30, and 60 s and 5 min at 37° C without stirring. After the intervals of thrombin stimulation, the incubation was terminated by the addition of a solution (1 ml) containing 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) or 4% paraformaldehyde and 0.2% glutaraldehyde in the same buffer, and the platelets were fixed for 30 rain at room temperature.

Electron microscopy Immunostaining using pre-embedding method. The fixed platelets were centrifuged at 3000 g for 2 rain. The platelet pellets were sequentially suspended and washed three times with 0.1 M phosphate buffer and two times with o.oa M phosphate-buffered saline (PBS, pH 7.4). The platelets fixed with a mixture of paraformaldehyde and glutaraldehyde were incubated with a 1:100 final dilution of rabbit anti-GPIIb/IIIa antibody (Tauoue et al. 1987), for 60 min at room temperature. The platelets fixed with paraformaldehyde alone were incubated with the monoclonal anti-GPIIIa antibody TM83 (Yamamoto et al. 1989), at 30 gg/ml for 60 min at room temperature. The controls were incubated with non-immune rabbit or mouse IgG instead of antibodies. After centrifugation at 3000 g for 2 min, the platelets were washed three times with PBS as described above. The platelets incubated with rabbit anti-GPIIb/IIIa antibody were then incubated with protein A coupled to 5 nm colloidal gold (protein A-gold; 1:10 final dilution; Janssen Life

Science Product, Olen, Belgium) for 60 min at room temperature. When a monoclonal anti-GPIIIa antibody, TM83, was used as the primary antibody, the ptatelets were incubated with rabbit antimouse IgG (10 gg/ml; Cappel, Organon Teknika Co., West Chester, Pa., USA) for 60 min at room temperature, followed by protein A-gold as described above. After the addition of an equal volume of 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4), each suspension was centrifuged at 3000 g for 2 min. The platelet pellets were dissected into blocks of 1 mm 3 or smaller, post-fixed with 1% osmium tetroxide in 0.1 M phosphate buffer (pH 7.4) for 60 min at 4° C, dehydrated with a graded ethanol series, and embedded in Epon. Gray colored ultra-thin sections were prepared, stained with uranyl acetate and lead citrate and then examined with a JEM 100C electron microscope (JEOL Co. Ltd., Tokyo, Japan) at an accelerating voltage of 80 kV.

Quantitative evaluation. The specimens prepared by the pre-embedding method (using rabbit anti-GPIIb/IIIa antibody and monoclonal anti-GPlIIa antibody, TM83) were used for quantitation of GPlIb/IIIa on the surface membrane. Electron micrographs of the unstimulated platelets and the platelets activated by thrombin for 15 s and 5 rain were obtained using a magnification of × 20000 and printed at a final magnification of x 50000. Gold particles on the surface membrane in each platelet were counted. The length of the surface was also measured using a semiautomatic image analyzing system, Nexus 6400 (Kashiwagi Research Co. Ltd., Tokyo, Japan). The density of gold particles (number of gold particles per micrometer) was estimated over the cell surface. Immunostaining using frozen sections. The platelets fixed with a mixture of paraformaldehyde and glutaraldehyde were centrifugated at 3000 g for 2 min. The platelet pellets were dissected into blocks of 1 mm 3 or smaller, then sequentially rinsed three times with 0.1 M phosphate buffer and two times with PBS. Infusion of sucrose into the fixed platelets before freezing was performed by the method of Tokuyasu (1989) with minor modifications as follows. The fixed platelets were sequentially infused with 1 M sucrose in PBS for 60 min, 1.84 M sucrose in PBS for 2 h and then 1.84 M sucrose containing 20% polyvinylpyrrolidone (mol.wt. 10000; Sigma) in PBS for 16 h at 4° C. After freezing in liquid nitrogen, ultra-thin frozen sections were prepared by a Reichert Ultracut N ultramicrotome (Vienna, Austria) with cryo-attachment FC-4E at - 1 1 0 ° C and mounted on nickel grids. The grids were floated onto a droplet of PBS for rinsing and then transferred to a droplet of 0.1% BSA (Sigma) in PBS for blocking. The sections were incubated with rabbit antiGPIIb/IIIa antibody (1 : 1000 dilution with PBS) for 16 h at room temperature. After rinsing five times with PBS, the sections were incubated with protein A-gold (10 nm; 1:30 dilution with PBS;

Fig. 1 A-C. Distribution of glycoprotein (GP) IIb/IIIa on the surface membrane of A unstimulated platelets, B platelets activated by 0.1 U/ml thrombin for 15 s and C 5 rain, detected by the preembedding method. The unstimulated and activated platelets were incubated with the rabbit anti-GPIIb/IIIa antibody followed by protein A-gold (5 nm) before embedding. A An unstimulated platelet has a discoidal shape. The gold labels for GPIIb/IIIa are distributed on the surface membrane (arrows). B The platelet activated by thrombin for 15 s has changed its shape from discoid to spheroid with pseudopodia, but the e-granules in the cytoplasm remain intact in form and number. The gold particles are distributed randomly on the surface membrane, similar to the unstimulated platelets (arrows). C The platelet activated by thrombin for 5 min has no c~-granulesin the cytoplasm. On the contrary, a swollen open canalicular system is formed. There is an obvious increase in the number of gold particles on the plasma membrane (arrows). eG, c~-granules; OCS, open canalicular system; M, mitochondria

383

384

Fig. 2. Distribution of GPIIb/IIIa in unstimulated platelets detected by immunostaining using ultra-thin frozen sections. The sections were incubated with rabbit anti-GPIIb/IIIa antibody followed by protein A-gold (10 nm). The gold labels for GPIIb/IIIa are evident on the membrane of the e-granules (arrows). The gold particles are also distributed on the surface membrane including the OCS. eG, e-granule; OCS, open canalicular system Fig. 3. Distribution of GPIIb/IIIa in the platelets activated by thrombin for 30 s and detected by the immunostaining method using frozen sections. Some of the e-granules are intact and the gold labels for GPIIb/IIIa are present on the granule membrane. Other e-granules have fused with the OCS (arrows) and their contents have moved to the lumen of the OCS, resulting in the release reaction. Gold particles are evident on the membrane of the fused and swollen OCS. eG, e-granule; OCS, open canalicular system

Janssen) for 3 h at room temperature. Control sections were treated with non-immune rabbit IgG and protein A-gold. After sequentially rinsing three times with PBS and five times with distilled water, the sections were stained with uranyl acetate followed by washing and then adsorption staining with a mixture of polyvinylalcohol (mol.wt. 10000; Sigma) and uranyl acetate according to the method of Tokuyasu (1989). The sections were examined with the electron microscope as described above.

Results

Immunostaining using pre-embedding method To assess the surface distribution o f the G P I I b / I I I a c o m plex, a p r e - e m b e d d i n g staining m e t h o d was utilized first,

by which only the G P I I b / I I I a complexes on the surface m e m b r a n e can be labeled by the antibodies. W h e n rabbit a n t i b o d y was used, a b u n d a n t gold labels for the G P I I b / I I I a c o m p l e x could be detected on the entire surface o f b o t h the unstimulated a n d t h r o m b i n - a c t i v a t e d platelets (Fig. 1). U n s t i m u l a t e d platelets were discoid-shaped, with gold particles r a n d o m l y distributed only on the plasma m e m b r a n e s (Fig. 1 A). By careful inspection, the gold particles were f o u n d to be present on the a m o r p h o u s structures and were separated by 10-20 n m f r o m the outer surface o f the p l a s m a m e m b r a n e , as previously reported in detail (Suzuki et al. 1991). The density o f gold particles on the unstimulated platelets were 18.0_+ 2.7 particles/lam o f p l a s m a m e m b r a n e (Table 1). At 15 s after the addition o f t h r o m b i n u n d e r non-stirring condi-

385 Table 1. Labeling density of glycoprotein (GP) IIb/IIIa on the surface of unstimulated and thrombin-activated platelets using rabbit anti-GPIIb/IIla antibody or monoclonal anti-GPIIIa antibody, TM83 Platelets

Unstimulated Thrombin-activated for 15 s Thrombin-activated for 5 min

Gold particle density a Rabbit antibody

Ratio

Monoclonal TM83

Ratio

18.0 _+2.7 (n = 53) 19.7_+3.6 (n = 50) 27.0_+3.7 (n = 50)*

i O0

3.1 _ 0.7 (n = 53) 4.1 _+0.9 (n = 52) 5.8_+0.9 (n = 52) **

100

109 150

132 187

Particle number (mean __SD) per micrometer, determined as described in the ' Materials and methods' *, ** p < 0.01 : significantly different from the values of the unstimulated platelets and the 15 s thrombin-activated platelets

tions, the shapes of the platelets changed from discoids to spheroids with pseudopodia (Fig. 1 B). However, c~granules still existed as intact forms in the cytoplasm. The gold labels for G P I I b / I I I a were distributed randomly on the plasma membrane, similar to the distribution in unstimulated platelets. The density was 19.7_+ 3.6 particles/gm of the plasma m e m b r a n e (Table 1). At 5 min after the activation, the platelets had no c~-granules in the cytoplasm because the release reaction had occurred (Fig. 1 C). At this time a swollen open canalicular system (OCS), which was formed f r o m the fusion of the OCS and c~-granules, was p r e d o m i n a n t in the central part of the cells. Although the distribution patterns of gold particles on the plasma m e m b r a n e remained unchanged, the density of gold particles increased to 27.0 ± 3.7 parti-

cles/gm (Table 1), a value that was significantly higher than those of unstimulated platelets or platelets activated by thrombin for 15 s (p