46 Biochimica et Biophysica Acta, 1012 (1989) 46-56 Elsevier

BBA 12469

The role of microtubules in platelet secretory release

Susan Berry 1, Do lo re t t a D. Dawick i 1, Ka i l a sh C. A g a r w a l 2 and M a n f r e d Steiner 1

I Division of Hematology / Oncology, Memorial Hospital of Rhode Island and 2 Department of Medicine and Section of Biochemical Pharmacology, Brown University, Providence, RI (U.S.A.)

(Received 22 November 1988)

Key words: Platelet microtubule; Digitonin; Platelet permeabilization; (Human)

The role of microtubules in platelet aggregation and secretion has been analyzed using platelets permeabilized with digitonin and monoclonal antibodies to a (DMIA) and .8 (DMIB) subunits of tubulin. Permeabilized platelets ,,:ere able to undergo aggregation and secretory release. However, threshold doses of agonists capable of eliciting a second wave of aggregation and the platelet release reaction were higher than in control platelets exposed to dimethyl sulfoxide, the solvent for digitonin. Both antibodies to a and ,8 tubulin caused a further increase in the threshold concentration of agonists and inhibited the secretory release of permeabilized platelets, but were ineffective using intact p|atelets. Neither moneclonal antibody inhibited polymerization or depolymerization of platelet tubulin in vitro. Antibodies to platelet actin and myosin also exhibited an inhibitory activity on platelet aggregation albeit less severe than that observed with the antibodies to a and .8 tubulin. There was evidence of an interaction between DMIA and DMIB and the antibodies to actin and myosin. The interaction of platelet tubulin and myosin was investigated by two different methods. (1) Coprecipitation of the proteins at low ionic strength at which tubulin by itself did not precipitate and (2) affinity chromatography on columns of immobilized myosin. Tubulin freed of its associated proteins (MAPs) by phosphoceHulose chromatography bound to myosin in a molar ratio which approached 2. Piatelet actin competed with tubulin for 1 binding site on the myosin molecule. MAPs also reduced the binding stoichiometry of tubulin/myosin. Treatment of microtubule protein with p-chloromercuribenzoate or colchicine did not influence its binding to myosin. DMIA and DMIB inhibited the interaction of tubulin and myosin. This effect could also be demonstrated by reaction of electrophoretic transblots of extracted platelet tubulin with the respective proteins. We interpret these results as evidence for an interference of the two monoclonal antibodies to the tubulin subunits (DMIA and DMIB) with the translocation of microtubule protein from its submembranous site to a more central one during the activation process.

Introduction

Microtubules have been shown to play an important role in maintaining the discoid shape of platelets [1,2]. Depolymerization of microtubules leads to the loss of their discoid shape [3-5]. Stimulation of platelets by agonists is associated with a distinct change in shape from discoid to spherical with pseudopodal extensions. Disruption of microtubule assembly-disassembly by colchicine or other tubulin ligands was able to inhibit serotonin release and the second wave of platelet aggre-

Abbreviations: Pipes, 1,4-piperazinediethanesulfonic acid; DMSO, di- methyl sulfoxide; BSA, bovine serum albumin; PC-tubulin, purified tubulin; SDS, sodium dodecyl sulfate; PBS, phosphate-buffered saline; PCA, perchloric acid.

Correspondence: M. Steiner, Division of Hematology/Oncology, Me- morial Hospital of Rhode Island, 111 Brewster Street, Pawtucket, RI 02860, U.S.A.

gation [6-9]. D20, a microtubule stabilizing agent, on the other hand, was reported to overcome this effect of colchicine [7]. However, the exact mechanism by which microtubules participate in the process of platelet activation remains unclear. We have previously shown that agonistic stimulation of platelets results in a rapid but transient dissociation of the circumferential bundle of microtubules [10]. Similar findings have been ob- tained by Kenney and Chao [11].

In the studies reported here we address the question of the role of microtubules in the activation process of platelets in greater delail. Permeabilized platelets which allowed entry of monoclonal antibodies to a and .8 tubulin monomers were used to elucidate the function of microtubules in the activation process. Platelet aggre- gation and release from electron dense granules and a granules were significantly reduced by these antibodies. Their effects were further potentiated by antibodies to actin or myosin. Our experiments suggest that the tubu- fin antibodies DMIA and DMIB interfere with the

0167-4889/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

interaction of myosin and actin with platelet tubulin. Microtubules are important modulators of the release reaction. Their translocation from the periphery to a central position during platelet stimulation may be mediated :Jy the interaction of these two proteins witl~ tubulia.

Materials and Methods

Materials Digitonin, ADP, PAF, p-chloromercudbenzoate, col-

chicine and Sepharose 2B and mouse IgG were obtained from Sigma Chemical Co., MO. Collagen was purchased from Helena, TX. Arachidonic acid was a product of NuChek Prep, MN. Platelet actin was received from Calbiochem, CA. Monoclonal antibodies to a (DM1A) and fl tubulin (DM1B) of chick brain (mouse IgG) were purchased from Amersham Corp., IL. Monoclonai anti- body to platelet actin (mouse IgG) was obtained from Chemicon, CA, and polyclonal anti-human platelet myosin (rabbit IgG) was purchased from Biomedical Technology, MA. Bovine thrombin was obtained from Parke-Davis, NJ. Gold-conjugated anti-mouse and anti-rabbit IgG, and the silver-enhancement kit were from Janssen Life Sciences Products, Belgium. The LDH assay kit was a product of Eastman Kodak, NY. The silver-stain kit was obtained from Bio-Rad, CA.

Isolation and permeabilization of platelets Blood was obtained from normal human volunteers

who had abstah~ed from any medication affecting platelet function for a minimum of 10 days. Platelets were isolated and washed as previously described [12]. For the extraction of tubulin, platelets were suspended in 0.1 M Pipes buffer (pH 6.9) containing 4 mM EGTA, 2 mM MgSO 4 and 1 mM GTP. For the extraction of myosin at~d actin, platelets were suspended in 0.14 M NaCl, 0.3~ sodium citrate, 1 mM EDTA, and 3 mM dithiothreitol, washed and pellets prepared.

Permeabilization of platelets with digitonin was per- formed as previously described [13]. Control platelet suspensions were incubated with an equal amount of dimethyl sulfoxide (DMSO), the solvent for digitonin, for the same period of time.

Isolation and preparation of proteins Microtubule protein was prepared from human

platelets. Throughout the paper, the term 'microtubule protein' refers to protein obtained after two cycles of temperature-dependent polymedzation-depolymefiza- tion. Tubulin was freed of its associated proteins (MAPs) by passage of the two-cycle microtubule protein pre- paration through a column of phosphocellulose, pre- equilibrated with 5 mM imidazole-HCl buffer (pH 6.6) containing 50 mM KCI. Purified tubulin (PC-tubulin) was eluted in the void fraction. MAPs were eluted by

47

raising the KCI concentration of the elution buffer to 0.8 M. The protein fractions were desalted by passage through Sephadex G-25 columns pre-equilibrated with the starting buffer. All protein samples were con- centrated to approx. 1 mg/ml by uitrafiltration.

Myosin and actin were extracted from outdated hu- man platelets according to Adelstein et al. [14] and Probst and Luscher [15], respectively. Purity of all pro- teins was analyzed by SDS-polyacrylamide gel gradient ele~:trophoresis and by immunologic characterization with specific antibodies of the Western blots prepared from such gels.

Electrophoresis methods SDS-polyacrylamide gradient gel eL'ctrophoresis was

performed as previously described [16]. A series of high and low molecular weight standards (Mr 200000-Mr 14400) was included with each electrophoretic sep- aration. Proteins were electrophoreticaUy transblotted either onto nitrocellulose or polyvinylidene difluoride sheets (PVDF; Immobilon). Completeness of transfer was verified by inclusion of a series of pre-stained molecular weight standards.

Urea SDS-polyacrylamide gel electrophoresis of cyanogen bromide cleaved peptides of tubulin was per- formed according to Swank and Munkres [17].

Gels were usually stained with a commercially availa- ble silver stain kit. To obtain quantitative results, the amount of protein applied to the gels ranged between 5-15 #g/gels. In this range linearity between ab- sorbance peak areas (633 #m) and protein concentra- tion could be verified. For quantification of peaks the gels and Western Blots were scanned using a laser densitometer (LKB Instruments).

Co-precipitation assay Purified tubulin, microtubule protein and myosin

were dialyzed against 20 mM Tris-HCl buffer (pH 7.6) containing 0.4 M KCI. The proteins were mixed at ratios of 1:3 to 1:5 (tubulin/myosin) by weight and incubated for 20 rain at 30°C. The protein mixtures were then diluted to 1/10 ionic strength by addition of ice-cold distilled water and centrifuged at 5000 × g for 15 rain. The resulting pellet was washed once with the dilute Tris-HCl buffer (pH 7.6) and was then prepared for electrophoresis as previously described.

For certain experiments tubulin preparations were preincubated with monoclonal antibodies to the ot and/or fl tubulin subunit, DMIA and/or DM1B (1 : 50 final dilution) at 30 °C for 30 min. To these incubations was then added myosin in the above weight ratio and the co-precipitation assay performed.

Preparation of myosin affinity column Myosin was immobilized on Sepharose 2B which had

been activated with cyanogen bromide. The cyanogen

48

bromide agarose was washed with coupling buffer, 0.05 M sodium pyrophoshate (pH 6.2) containing 0.5 M KCI. Myosin equilibrated with this buffer was then mixed with the agarose (6-10 mg/ml agarose) and stirred gently in the cold for 16-24 h. Free myosin was removed by several washes with coupling buffer. Activated groups remaining on Sepharose-myosin were blocked by incubation for 2 h with 0.5 M ethanolamine and 0.05 M sodium pyrophosphate buffer (pH 8.0) at 4°C. The conjugate was washed free of ethanolamine and suspended in Pipes buffer (pH 6.9) containing 0.5 mM GTP and 0.5 mM ATE In general, the columns contained between 0.5 to 1.0 ml of myosin-agarose. After pre-equilibration with this buffer, the affinity column was ready. The volume of the sample of micro- tubule protein or tubulin applied to the columns did not exceed their void column. After addition of the protein, the outflow was clamped for 20 rain. Usually, these columns were operated at 4 o C, although other tempera- tures were used occasionally. Retained protein was eluted from the affinity column with the above Pipes buffer containing 0.3 M KCI.

The amount of myosin on the column was estimated from the difference between the total amount applied and the sum total of the myosin removed by the washes. A total of 2-3.5 mg myosin/ml agarose was obtained from the various columns prepared. Because of the presence of GTP and ATP in the elution buffer, protein measurements for this series of experiments were made by determining the maximal fluorescence emission of tryptophan residues ia the range of 320-350 nm after stimulating with light of wavelengths of 295 nm. A standard curve was prepared with bovine serum al- bun.in in the elution buffer.

In vitro polymerization-depolymerization of platelet tubu- lin

Tubulin was extracted from human platelets as described before using a temperature-dependent, poly- merization-depolymerization technique [18]. Two-cycle tubulin was used to evaluate the effect of monoclonal antibodies on the polymerization and depolymerization of tubulin. Assembly and dissassembly of microtubules was followed by recording the changes in light transmis- sion at 350 run in a temperature-controlled cuvette holder, using a recording double-beam spectrophotome- ter. For these experiments tubulin was prepared in 0.1 M Pipes buffer containing 4 mM EGTA, 2 mM MgSO4 and 1 mM GTP (pH 6.9) at a concentration of 1.5 mg/mL

Treatment of permeabilized platelets with monoclonal antibodies

Platelets suspended in PBS with BSA and glucose were incubated for 30 rain at 30 °C with DM1A and DMIB and with antibodies to actin or myosin. Control

platelets were incubated with an equal volume of mouse plasma diluted in the same proportion as the antibodies. The incubation was terminated by centrifugation of the platelet suspensions. The platelet pellets were recon- stituted with platelet-poor plasma of the individual donors to a concentration of approx. 3- 108/ml.

Measurement of platelet aggregation The aggregability of control and antibody-treated

platelets was tested using collagen, 5,8,11,14-eico.~ satetraenoic acid (arachidonic acid), ADP, 1-O-alkyl-2- acetyl-sn-glyceryl(3)phosphorylchoiine (platelet activat- ing factor, PAF) and thrombin as agonists. The minimal effective concentration producing complete aggregation (primary and secondary wave) was established for each agonist using control platelets (treated with digitonin dissolved in DMSO). Aggregation profiles of control and experimental samples were evaluated with respect to initial slope of the aggregation curve and the maxi- mum height attained after 4 min.

The secretory phase of platelet aggregation was analyzed by following the release of ATP using a Lum~aggregometer (Chronolog, PA) and the luciferin- luciferase method [19]. ot granule release was evaluated by measuring N-acetylglucosaminidase activity [20].

Statistical differences between control and experi- mental samples were evaluated by Student's t-test.

Nucleotide analysis Platelet suspensions were incubated with digitonin or

DMSO. After the incubation the suspensions were centrifuged (2000 x g for 10 rain) and the platelets were resuspended in 0.05 M Tris-HC1 (pH 7.2) containing 0.14 M NaCI (0.1 ml). An equal volume of ice-cold 6~ (v/v) perchloric acid (PCA) was added and stirred rapidly on a Vortex mixer. The precipitated proteins were removed by centrifugation at 12 000 × g for 5 rain. The supernatants were neutralized with K2CO3, and precipitated KCIO4 was removed by centrifugation. The neutralized PCA extracts were used directly for the assay of nucleotides by high-pressure liquid chromatog- raphy (Varian LCS-1000) [21]. Nucleotide concentra- tions were estimated by comparing the peak areas with those of known nucleotide standards.

Evaluation of interaction of monoclonai antibodies with transblotted platelet tubulin

Tubulin extracted from human platelets by two cycles of temperature-dependent polymerization-depolymeri- zation was separated from its MAPs by SDS-poly- acrylamide gradient gel electrophoresis. The resolved proteins were transblotted electrophoretically onto nitrocellulose sheets, using the method of Towbin et al. [22]. The blotted proteins were then reacted with ap- propriate dilutions of monoclonal antibodies to a and fl tubulin. In other experiments the nitrocellulose sheets

were first treated with platelet actin, platelet myosin or normal human plasma. Myosin and actin dissolved in 20 mM Tris-HCl buffer containing 0.1% bovine serum albumin and 0.14 M NaCI (pH 8.2) (BSA-TBS buffer) were incubated with the nitrocellulose blots of tubulin or of cyanogen bromide cleaved tubulin peptides at concentrations of I or 2/~M (final concentration) for 2 h at 37 ° C. After washing the nitrocellulose strips three times with the same buffer, the blots were incubated with monoclonal antibody to a or fl tubulin (1 : 50, final dilution) or with antibody to platelet actin or myosin. After a 2 h incubation, the primary antibody solution was decanted and the nitrocellulose strips were washed three times with 0.1% BSA-TBS buffer. The presence of the primary antibody on the blots was recognized by incubatiag the strips with gold-conjugated anti-mouse or anti-~abbit IgG (1:190, final dilution). After two more w,~shing steps with 0.17o BSA-TBS buffer and two brief fi~sing steps with distilled water, the nitrocellulose strips were subjected to silver enhancement using the IntenSE II kit of Janssen Life Sciences Products. In- tensification was allowed to proceed until the back- ground began to darken (usually within 10-20 min). The strips were then washed with H 2 0 and left to dry.

Results

Platelet myosin was able to precipitate actin and microtubule protein at low ionic strength either together or each protein alone. Purified tubulin could also be precipitated under these experimental conditions (Fig. 1). Without myosin, neither microtubule protein nor PC-tubulin precipitated. The interaction between myosin and microtubule protein was temperature-dependent, iucreasin8 from a tubulin/myosin ration of 0.7 (mol/mol) at 4°C to 1.5 at 30 ° C. Maximal precipita- tion occurred within 10 min and the pH of interaction was optimal between 6.4 and 7.8. All subsequent experi- ments were performed at pH 7.6. Washing of the pre- cipitate up to four times did not change its composition as analyzed by SDS-polyacrylamide gel electrophoresis.

As the above method of evaluating tubulin binding to myosin proved to be quite cumbersome, we devel- oped an alternate technique utilizing immobilized platelet myosin. The principal advantage of this method was the ability to reutilize the columns and thereby economize on the myosin available for these experi- ments. The immobilization of myosin on Sepharose did not significantly reduce its ability to interact with mi- crotubule protein or PC-tubulin. Operating the columns at 30 °C and at a weight ratio of 3 : 1 (myosin/tubulin), 1.4 mol of microtubule protein was bound per mol of myosin. The elution buffer with 0.5 mM GTP and ATP removed all of the bound tubulin (Fig. 2). Raising the molarity of the buffer to 0.8 M KCI did not elute any more protein from the affinity columns.

49

B

Fig. 1. Tubulin/myosin co-precipitation analyzed by SDS-poly. acrylamide gel electrophoresis. (A) Microtubule protein, acti~: and myosin were incubated at approximately equal molar concentrations for 20 min at 30°C. (B) Actin and myosin, (C) microtubule protein and myosin, and (D) actin and microtubule protein were incubated as in (A). After centrifugation the precipitates were washed twice, solubilized and subjected to electrophoresis. The silver-stained gels were scanned at 633 nm. The absorbance profiles are shown in the figure. Details of the experimental design are provided unden Materi- als and Methods. M, myosin; T, tubulin; A, actin; L I and L 2, myosin

light chains; T.D., position of tracker dye (bromophenol blue).

4

u

m

2 n,,

¢,

, . . , 200

~-o 1 1 6

- ~ 9 Z

~ o 45

31

21 1 4

8 12 16 20

Fraction number

24

Fig. 2. Affinity chromatography of microtubule protein on columns of immobilized platelet myosin. To a 0.5 ml agarose column containing 2.2 mg myosin per ml was added 1.6 mg microtubule protein. The arrow indicates the change over from washing buffer to high ionic strength (0.3 M KCI) elution buffer. Protein was measured by fluores- cence emission of tryptophan residues. The inset demonstrates the analysis of the column eluate by SDS.polyacrylamide gradient gel electrophoresis. The sample was reduced and alkylated [30] and the gel stained with a silver stain. The arrow in the inset identifies the

position of the tubulin heterodimer.

50

1.5

• •

~ 1.0

C

'ID

} as (9

0.2 0,4 O.6 Mi¢ rotubule protein (mg / ml )

Fig. 3, Stoichiometry of tubulin-myosin interaction. A 15 FM myosin solution was incubated with increasing concentrations of microtubule protein in the absence (O) or presence (o) of 10 tiM platelet actin. Co-precipitation assays were carried out as described under Materials

and Methods.

The stoichiometry of myosin-tubulin interaction was examined both by the co-precipitation and the affinity column method. Both techniques produced similar re- sults. Increasing the concentration of microtubule pro- tein which interacted with a constant amount of myosin produced a curve showing saturation characteristics (Fig. 3). A tubulin/myosin molar ratio of about 1.5:1 was

2,0 |

m

w

C

"~ 1,5

E

' O

ID

0,5

0.2 0.4 0.6 Tubulin (mg/ml)

Fig. 4. Stoichiometry of tubulin-myosin interaction in the presence (O) and absence (®) of MAPs. Equal volumes of myosin (15/zM), PC-tubulin (varying concentrations) and MAPs (0.3 mg/ml) was incubated and co-precipitation assays performed as described under

Materials and Methods.

approached as the concentration of microtubule protein in the reaction mixture was raised. The addition of actin produced a reduction in this ratio to approx. 0.9.

The interaction of PC-tubulin with myosin produced a molar ratio that approached 2.0 (Fig. 4). In the presence of a constant concentration of MAPs the molar ratio of PC-tubulin/myosin showed a marked reduc- tion.

Polymerizability and ligand binding of tubulin are sensitive indicators of its structural integrity. We there- fore examined the structural requirements for the inter- action of tubulin with platelet myosin. Tubulin al- kylated with p-chloromercuribenzoate and colchicine bound to microtubule protein were prepared. Binding of these tubulins to myosin was tested by the usual methods. There was no significant difference in the molar ratio of tubulin/myosin that was approached compared to that obtained with normal untreated tubu- lin (data not shown).

Incubation of normal or DMSO treated, but non- permeabillzed platelets, with monoclonal antibodies to a and fl tubulin or equal volumes of mouse plasma failed to show any change in their aggregability when tested with platelet agonists. Permeabilized platelets, on the other hand, were profoundly affected by mono- clonal antibodies to a and fl tubulin, but not by equivalent dilutions of mouse plasma or by non-im- mune mouse IgG. Representative examples of the ag- gregation responses tested at threshold levels of the individual agonists are shown in Fig. 5. Aggregation induced by PAF, arachidonic acid, collagen, ADP and thrombin was severely reduced by preincubation of digitonin-treated platelets with antibody to a or fl

E: 6Q 8 3o

A

~O.3o,1 ~ ~ 4 3 G

• i a i ,

¢

1

m m = j a |

* * = n |

D / 1

2

- I

=

• I g a i ii 0 1 2 3 4 5 0 1 2 3 4 5

T ime ( ra in)

Fig. 5. Effect of monoclonal antibody to a [3] or B [4] tubulin is compared with control platelet-rich ~dasma treated with DMSO [1] and with digitonin-permeabl,7~ed p!utelets that were incubated either with mouse plasm~ equa', i , volume to that of the monoclonal antibodies used or with 0.5 CM mouse 18G [2]. Platelets were stimu- lated with thrombin (0.15 U/ml) (A), PAF (2.5/~M) (B), ADP (10 ~M) ((2) and collagen (3.0 ~,g/ml) (D). The time of addition of the

a,~onist is indicated by the arrow.

TABLE I

Aggregation response of permeabilized platelets to tubufin antibodies

51

Agonist Cone. Aggregation response a

Slope: C c old #e ad+fle

ADP 15/LM 1.8±0.12 Collagen 1.6 pg/ml 2.3 + 0.19 PAF 2.0 pM 2.6 ± 0.2 Thrombin 0.15 U/ml 1.8 ± 0.14 Arachidonic acid 490/L M 2.4 ± 0.21

O.T. (~) b

Slope: C ~

ADP 15 pM 65 ± 3.5 Collagen 1.6 pg/ml 78 ± 4.6 PAF 2.0 ~M 59 ± 3.0 Thrombin 0.15 U/ml 58 ± 2.8 Arachidonic acid 490/t M 68 ± 3.9

1.3±0.1 t 1.1±0.1 t 0.9±0.08 f 0.8±0.09 f 0.5±0.07 f 0.4±0.03 f 1.8±0.21 t 1.6±0.12 t 1.4±0.12 t 0.8±0.1 f 0.6±0.1 f 0.5±0.08 t 1.6±0.1 t 1.2±0.1 f 1.0±0.1 f

ad #e a d + f l e

33±2.9 f 25±2.4 f 21±2.0 t 35±3.0 g 28±2.8 t 23±2.1 t 32±2.8 t 30±3.8 t 25±3.2 t 38±3.0 f 35±3.1 t 29±2.8 t 35±3.1 f 30±2.1 f 21±1.9 f

" M e a n ± S .D. of five experiments. b O.T. (%), change in optical transmission as percentage of platelet-poor plasma or blank reading (for thrombin only). c C, control platelet suspension treated with mouse plasma (1:50 dilution) or mouse IgG (0.5 pM). d a, same as in legend to Table II!. e fl, same as in legend to Table Ill. f P < 0.005.

TABLE 11

A TP release ofpermeabilized platelets treated with tubulin antibodies

Agonist Cone. ATP release per 10 n platelets (~mol)

c b otc fld otc+fl d

ADP 15 pM !.87-1-0.15 1 22±0.25 ¢ Arachidonic acid 495 p M 1.75 + 0.18 0.45 + 0.1 t Collagen 1.6 ttg/ml 1.78 + 0.20 0.38 + 0.1 t

0.73±0.15 f 0.20±0.05 t 0.25±0.08 t

0.52±0.10 t 0.11±0.03 t 0.19±0.~ t

a Mean + S.D. of three experiments. b C, control platelet suspension treated with mouse plasma (1:50 diiution) or mouse lgG (0.5 ~M). c a, platelet suspension treated with monoclonal antibody to a tubuha (final dilution 1 : 50). d /L platelet suspension treated with monoclonal antibody to fl tubulin (final dilution 1:50). e p < 0.025. t P < 0.005.

TABLE III

Secretory release of permeabilized platelets treated with tubulin antibod- ies a

ATP release/10 li platelets N - A G I u b

(pmol) (~)

Control 1.7 ±0.1 28.7±2.1 a 0.5 ±0.06 c 8.2±0.9 c fl 0 .36±0.~ c 5.1±0.4 ¢ a + f l 0.28±0.02 c 4.0± 0.4c

" Mean+ 1 S.D. of three experiments. Secretory ~c'(e-,se was induced by 0.15 U/ml thrombin. Control, control tfl,ttelet s~spension treated with mouse plasma (final dilution ~ : 30) or mouse IgG (0.5 pM). a, platelet suspension treated with antibody to a tubul~n (final dilu- tion 1:50). fl, platelet suspension treated with antibody to P, tubulin (final dilution 1 : 50).

b N-acetylglucosaminidase activity expressed as percentage of total present in platelets.

c p < 0.0005.

t u b u l i n ( T a b l e I). In all expe r imen t s a n t i b o d y to fl

t u b u l i n p r o v e d to be a s l ight ly more effect ive i n h i b i t o r

t h a n a n t i b o d y to a tubu l in . M e a s u r e m e n t s o f the sec re to ry p la te le t release agreed

wi th the agg rega t i on f ind ings . A r a c h i d o n i c acid-, A D P -

a n d c o l l a g e n - i n d u c e d A T P release were s ign i f i can t ly

redt:,~;ed b y ,~ncubation w i th m o n o c l o n a l a n t i b o d i e s to

t u b u l i n ( T a b l e II). Re lease f rom a g ranu les was

e v a l u a t e d b y m e a s u r e m e n t o f N-ace ty lg lucosamin ida se

ac t iv i ty in the s u p e r n a t a n t o f the aggregated pla te le ts .

M o n o c l o n a l a n t i b o d i e s to b o t h a and /] t u b u l i n in-

h i b i t e d t h r o m b i n - i n d u c e d a g r anu l e release to a s imi la r

e x t e n t as t he re lease o f A T P f rom e lec t ron dense gran-

ules ( T a b l e I l l ) . O n the bas is o f these resul ts we c~>nsidered the

pos s ib i l i t y t h a t the t u b u l i n an t ibod ie s i ~ e r f e r e d wi th

the p o l y m e r i z a t i o n - d e p o l y m e r i z a t i o n of p la te le t mic ro -

52

03

~o.2 o

0.1

! !

0 5 10 15 20 25 T ime (ra in)

Fig. 6. Temperature-dependent polymerization and depolymerization of control (A) and monoclonai antibody (to a tubulin) treated (B) tubulin extracts of human platelets. Final concentration of tubulin in the medium was 1.5 mg/ml. Optical absorbance was followed in a temperature-controlled double-beam recording spectrophotometer at 350 nm, Polymerization was induced by increasing the temperature from 4 to 37 o C, The arrow indicates the time at which the tempera- turc of the cuvette holder was switched from 37 to 4°C. Identical experiments were obtained using monoclonal antibody to fi tubulin. Both monoclonal antibodies were used at the concentrations which

inhibited platelet aggregation and release.

25

~) o

:~ 5 o

o _u 25

0

A,/ 21 b

d

0 1 2 3 4 5

d

• ii | i • ii

o 1 2 3 ,4 5 T ime (rain)

Fig. 7. Typical aggregation profiles of permeabilized platelets treated with various combinations of antibodies to platelet actin (b in panels A and B) or platelet myosin (b in panels C and D). The effect of monoclonal antibody to fl tubulin is shown in panels A and B (c) and of monoclonal antibody to a tubulin in panels C and D (c). The combined effect of the various antibody pairs is shown in curve d. The arrow indicates the time of application of the agonist (ADP, 5 ttM in panel A; PAF 1.5/tM in panel B; collagen 1.6/tg/ml in panel C; arachidonic acid 495 ttM in panel D). Controls are represented by

c u r v e a .

tubules. We, therefore, performed experiments to de- termine the polymerizability of isolated tubulin in the presence of monoclonal antibody to a or fl tubulin. Turbidimetric measurements of tubulin polymerization, induced by raising the temperature to 37 ° C, did not show any difference between control tubulin prepara- tions exposed to equal volumes of mouse plasma and monoclonal antibody-treated preparations (Fig. 6). De- polymerization induced by cooling the microtubule pre- parations to 4°C was equally unaffected by the pres- ence of antibodies to the tubulin subunits. In view of these findings, we excluded the possibility that these monoclonal antibodies inhibited polymerization or de- polymerization.

The question whether the monoclonal antibodies in- terfered with the translocation of microtubules from their submembranous position to a central location was investigated by the use of monoclonal antibodies to actin and myosin. Antibody to myosin exhibited inhibi- tory action on platelet aggregation which equalled that of monoclonal antibody to a tubulin with agonists such as collagen (Fig. 7) and PAF (not shown). It was weaker with other agonists such as arachidonic acid and ADP (not shown). The combined use of myosin and tubulin antibodies produced inhibition equal or slightly greater than the sum total of the individual responses to these two monoclonal antibodies. Antibody to platelet actin also inhibited platelet aggregation. As shown in Fig. 7 it produced a similar depression of aggregation as anti- body to fl tubulin when ADP and PAF were the agonists. With arachidonic acid and collagen as platelet stimu- lants, antibody to actin was slightly less effective than

antibody to fl tubulin but equalled the effect of anti- body to a tubulin (data not shown). The combined use of antibodies to tubulin and actin resulted in inhibition

B

D

L

m

Fig. 8. Analysis of the effect of tubufin antibodies DM1A and DM1B on the interaction of tubulin and myosin. Tubulin preparations were preincubated with DMIA and DM1B for 30 min. Myosin was then added in equimolar concentration to tubulin and the co-precipitation assay performed as described under Materials and Methods. (A) Control without antibodies; (B) both DMIA and DM1B were used in the preincubation period; and (C) DMIA and (D) DMIB were preincubated with the tubulin preparation. The letter code is the same

as used in Fig. 1.

m ' % , ~ ; i ¸ ~!i;

E ~O

<

53

The individual monoclonal antibodies to a and fl tubu- lin yielded only partial inhibition of myosin binding. As this assay was unable to show any co-precipitation of tubulin and actin, the effect of the two monoclonal antitubulin antibodies was not further investigated in this system. This is in contrast to the co-precipitation of actin and tubulin observed when actin is incubated with microtubule protein (Fig. 1D). These results suggest that actin interacts with tubulin indirectly via one or more MAPs.

The interaction of myosin and actin with platelet tubulin was also examined by reacting electrophoreti- cally resolved two-cycle tubulin preparations that had been transblotted onto nitrocellulose with actin or myosin in concentrations of 1 or 2 #M, respectively. We attempted to prove the existence of an interaction be- tween the proteins by demonstrating a reduction of the binding of the specific antibody to a or ~ tubulin. The interference of myosin pretreatment with the subse- quent reaction of antibodies to c~ and p tubulin is shown in Fig. 9. Both tubulin monomers showed a reduction in the binding of the specific antibody, slightly more with fl than with a tubulin. Actin pretreatment of the tubulin transblots produced less visible evidence of an inhibition of binding to either the a o r / / subunit using this method (data not shown).

In further experiments we were able to extend these findings by preparing cyanogen bromide peptides of both a and /~ tubulin monomers. The peptides were resolved by urea SDS-polyacrylamide gel electrophore- sis (Fig. 10). After transblotting the separated peptides onto nit 'cellulose, they were reacted with monoclonal antibodies to ~ and fl tubulin, respectively. One major reactive peptide could be identified in both tubulin

Fig. 9. Effect of platelet myosin on the reaction of platelet tubufin with monoclonal antibody to a (a and b) and fl tubulin (c and d). Platelet tubulin prepared by two cycles of temperature-dependent polymerization-depolymerization was reduced, alkylated with iodoa- cetamide and subjected to SDS-polyacrylamide gradient gel electro- phoresis. Transblots onto nitrocellulose sheets were reacted with 2 /~M platelet myosin. Untreated controls (a and c) and myosin pre- treated preparations (b and d) were then reacted with monoclonal antibody to a and fl tubulin. The presence of antibody was demon- strated with gold-conjugated anti-mouse IgG followed by silver in-

tensification.

of platelet aggregation equal to or greater than that caused by the individual antibodies.

The effect of myosin and actin on the interaction of the monoclonal antibodies DM1A and DMIB with tubulin was also examined in a cell-free fluid-phase system. Utilizing the co-precipitation assay, it was pos- sible to determine that combined use of these antibodies blocked the interaction of myosin with tubulin (Fig. 8).

0.5

0.7[

50 75 100 125

Distance ( ram}

0 3

Fig. 10. Densitometric scans of peptides prepared by cyanogen treat- ment of a ( . . . . . . ) and fl ( ~ ) tubulin. The peptides were resolved by urea SDS-polyacrylamide gel electrophoresis and visual- ized with a silver stain. The abscissa represents the distance (mm) of the laser from the position of the tracker dye (bromophenol blue). The anodal end of the gel is at the left. The origin is the same for Figs.

l lB-14.

54

, t A ~ . •

a ~ b I 32 r

i

c E 29

2,6

..-:..-....:.-_:. .................................................

77 84 91 98 Distance (mm)

Fig. 12. Laser densitometric scans of cyanogen bromide cleaved peptides of a tubulin separated on urea SDS-polyacrylamide gels, transblotted and reacted with 1 ~tM platelet actin (solid line and interrupted line) or with buffer (dotted line). The solid and dotted lines represent scans of peptides reacting with monoclonal antibody to a tubulin while the interrupted line represents a scan of peptides reacting with antibody to actin. Antibodies were visualized by gold- conjugated secondary antibodies as described under Materials and

Methods.

37

~ 3 2

2 7

77 84 91 9e Distance (ram)

Fig. 11. (A) Reaction of cyanogen bromide cleaved a (a) and/3 (b) tubulin subunits resolved by urea SDS-polyacrylamide gel electro- phoresis and transblotted onto nitrocellulose sheets with monoclonal antibody to a and/3 tubulin, respectively. Antibody reacting with specific peptides was visualized with gold-conjugated anti-mouse IgG which was intensified by silver enhancement. (B) Laser densitometric scans of cyanogen bromide cleaved tubulin peptides separated by urea SDS-polyaerylamide gel electrophoresis, transblotted onto nitrocellu- lose sheets and then reacted with monoclonal antibody to a or/3 tubulin. Peptides reactive with the respective monoclonal antibody were visualized with gold-conjugated anti-mouse IgG intensified by silver enhancement, a tubulm peptides ( )/3 tubulin peptides

(- . . . . . ).

monomers , a tubul in had an addi t ional three or four epi topes reactive with the monoc lona l an t ibody while tubul in showed only one (Fig. 11A and B).

The in te rac t ion of plate le t myos in and act in wi th cyanogen b r o m i d e pept ides of tubu l in was also investi- gated. Plate le t act in was able to block the react ion of the monoc lona l an t ibody to a tubu l in with a specific cyanogen b r o m i d e pept ide of this tubul in subuni t (Fig. 12). M y o s i n in terfered with the b ind ing of the same monoc lona l a n t i b o d y to a d i f ferent pep t ide of a tubu l in (Fig. 13). The evidence of a s imilar interference by ac t in and myos in of the b inding of monoc lona l an t ibody to tubul in pep t ides was not as clearcut. There was a slight change in the relative s ta ining in tens i ty of the two

3.3

i

E 29 ":"i

o ~ ! ii

R m m 77 84 91 98

Distance (ram)

Fig. 13. Laser densitometric scans of cyanogen bromide cleaved peptides of a tubufin pretreated with 2 ~M platelet myosin ( ) or buffer only ( . . . . . . ). Electrophoretic resolution of the peptides, transblotting, reaction with specific monocional antibodies to a tubuo fin and visualization with gold-conjugated anti-mouse IgG were done

as described in the legend to Fig. 9.

55

27

2.5

2 7

25__

L f~ ~ " ' - 7 7 8 4 91 9 8

D i s t o n c e ( m m )

Fig. 14. Comparison of cyanogen bromide peptides of ot (lower panel) and/3 (upper panel) tubulin reacting with platelet actin which were identified by specific platelet actin antibody. Resolution of the peptides, transblotting and staining procedures are described under

Materials and Methods.

major p tubulin peptides (data not shown). However, pretreatment of the transblotted peptides with either platelet myosin (data not shown) or actin gave evidence of the presence of these proteins on specific tubulin peptides (Fig. 14). Reaction of such peptides with anti- body to myosin or actin revealed distinct bands either at the site of or close to the peptides reacting with the monoclonal antibodies to a and fl tubulin.

Discussion

Monoclonal antibodies to a and p tubulin were able to interfere with the aggregation and secretory release of permeabilized platelets. Specific ligands of microtubules such as colchicine [7,9], vincristine [9] and nocodazole [6] have been previously shown to interfere with the platelet-release reaction. As these agents are well known to block microtubule assembly, we suspected that a similar mechanism was responsible for the antibody associated inhibition of platelet release. Our experi- ments, however, gave clear evidence that neither of the monoclonal antibodies was able to block assembly or disassembly of microtubules. Both polymerization and depolymerization proceeded at equal rates in the pres- ence or absence of monoclonal antibody to a and fl tubulin. For these reasons we turned our attention to proteins that interact with tubulin as possible targets for the interference of the monoclonal antibodies with the platelet release reaction.

Stoichiometric associations of myosin with tubulin have been noted in certain tissues [23,24]. Evidence for the interaction of platelet tubulin with myosin was obtained in a call-free system as well as in platelets that were permeabilized to allow entry of antibodies to tubulin. This interaction could be shown not only by

the traditional precipitation method [24], utilizing low ionic strength at which myosin as well as actomyosin are aggregated, but also under non-aggregating condi- tion with myosin immobihzed on agarose. Actin as well as MAPs appear to interfere with the myosin-tubulin interaction. Although direct interactions of tubulin and actin have not been observed, indirect associations through MAPS have been reported [25]. As our MAP preparation normally contains a small amount of actin, we believe that it may be responsible for the observed inhibitory effect. Of the two tubulin molecules that can bind to myosin, one binds to a site at which it has to compete with actin, whereas the other is reserved for tubulin. Using the monoclonal antibodies DM1A and DM1B, we were able to completely block the interac- tion of tubulin and myosin. Thus, both ~x and ~ tubulin appear to have binding sites for myosin.

The structural requirements for an interaction with myosin do not seem to be very stringent as fully al- kylated tubulin as weil as colchicine bound tubulin both can bind to myosin without any significant hindrance. Having found a broad pH range at which binding can take place, it appears that the hydrogen ion concentra- tion is not an important factor in the interaction be- tween tubulin and myosin. A distinct temperature de- pendence of the binding process was noted.

The interaction of tubulin with such cellular proteins could be of significance for platelet function as a trans- location of the microtubules to a more central position appears to take place during the activation process [26,27]. The slight potentiation of inhibition of aggre- gation by the combined use of monoclonal antibody to either a or fl tubulin and antibody to myosin or actin was suggestive of an interaction of tubulin with these two platelet proteins. More definitive evidence was ob- tained when electrophoretically separated and trans- blotted cyanogen bromide peptides of a and ~ tubulin were reacted with platelet myosin or actin.

The interference of DM1B and DMIA with the binding of tubulin to platelet myosin could also be confirmed in a system in which all proteins were in the fluid phase. Thus, the results of the Western blot experi- ments could be confirmed by a method that did not require tubulin to undergo denaturation by a detergent.

Previous experiments by at least two other laborato- ries [28,29] have shown that the monoclonal antibodies DM1A and DM1B were directed against peptides of the C-terminal end of both ~x and ~ tubulin subunits. In our studies a small number of peptides reacted with these antibodies, more in a than in ~ tubulin. Both tubulin monomers showed only one clearly predomi- nant a~tibody-reactive peptide. In this respect our re- sults are very similar to those reported recently by Breitling and Little [28] and Serrano etal. [4].

We interpret our results as evidence that monoclonal antibodies to the C-terminal end of both the a and

56

tubulin subunits interfere with the binding of myosin and actin by microtubules. We believe that this interac- tion is an integral part of the activation process of platelets and is essential for the secretory release phase of platelet activation.

Acknowledgement

This study was supported by research grant HL 19323 of the National Heart, Lung and Blood Institute.

References

1 Behnke, O, (1965) J. Uitrastruct. Res. 13, 469-477. 2 White, J,O. (1971) in The Circulating Platelet (Johnson, S.A., ed.),

pp. 45-121, Academic Press, New York. 3 Behnke, O. (1966) J. Cell Biol., 34, 677-701. 4 Serrano, L., Wandosell, F. and AviIa, J. (1986) Anal. Biochem.

159, 253-259. 5 White, J,G. and Krivit, W. (1967) Blood 30, 625-635. 6 Jung, S.M., Yamazaki, H., Tetsuka, T. and Moroi, M. (1981)

Thromb. Res. 23, 401-410. 7 Menche, D., Israel, A. and Karpatkin, S. (1979) J. Clin. Invest. 66,

284-291. 8 White, J,G. (1969) Am. J. Pathol. 53, 281-291. 9 White, J.G. (1969) Am. J. Pathol. 54, 467-478.

10 Steiner, M. and Ikeda, Y. (1979) J. Clin. Invest. 63, 443-448.

11 Kenney, D.M. and Chao, F.C. (1980) J. Cell. Physiol. 103, 289-298. 12 Landolfi, R., Mower, R.L. and Steiner, M. (1984) Biochem.

Pharmacol. 33, 1525-1530. 13 Lineberger, B., Dawicki, D.D., Agarwal, K., Kessimian, N. and

Steiner, M. (1989) Biochim. Biophys. Acta 1012, 36-45. 14 Adelstein, R.S., Pollard, T.D. and Kuehl, W.M. (1971) Proc. Natl.

Acad. Sci. USA 68, 2703-2707. 15 Probst, E. and Luscher, E.F. (1972) Biochim. Biophys. Acta 278,

577-584. 16 Steiner, M. and Luscher, E.F. (1986) J. Biol. Chem. 261, 7230-7235. 17 Swank, R.T. and Munkres, K.D. (1971) Anal. Biochem. 39,

462-477. 18 lkeda, Y. and Steiner, M. (1976) J. Biol. Chem. 251, 6135-6141. 19 Feinman, R.D., Lubowsky, J., Charo, I.F. and Zabinski, M.P.

(1977) J. Lab. Clin. Med. 90, 125-129. 20 Li, Y.T. (1966) J. Biol. Chem. 241, 1010-1012. 21 Agarwai, K.C. and Parks. R.E., Jr. (1975) Biochem. Pharmacol. 24,

2239-2248. 22 Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad.

Sci. USA 76, 4350-4354. 23 Fujii, T., Kondo, Y., Kumasaka, M., Hachimori, A. and Ohki, K.

(1982) Pharm. Bull. 30, 4134-4139. 24 Shimo-Oka, T., Hayashi, M. and Watanabe, Y. (1980) Biochem-

istry 19, 4921-4926. 25 Griffith, L.M. and Pollard, T.D. (1978) J. Cell. Biol. 78, 958-965. 26 Behnke, 0. (1970) Stand. J. Haematol. 7, 123-140. 27 White, J.G. (1968) Blood 31, 604-622. 28 Breitling, I. and Little, M. (1986) J. Mol. Biol. 189, 367-370. 29 Serrano, L., Wandoseil, F. and Avila, J. (1986) Anal. Biochem.

159, 253-259.